based on invited talks during dae-brns national …ila.org.in/kiran/kiran_25_03.pdf1 vol. 25, no. 3,...

Vol. 25, No. 3, December 2014A Bulletin of the Indian Laser Association

Based on invited talks during

DAE-BRNS National Laser Symposium (NLS-23)

Editor

Prof. Manoranjan P. Singh RRCAT, Indore

Editorial Board

Prof. A.K. Gupta SCTIMST,

Thiruvananthapuram

Dr. A.K. Maini LASTEC, New Delhi

Prof. S. Maiti TIFR, Mumbai

Prof. S.C. Mehendale RRCAT, Indore

Prof. V.P.N. Nampoori CUSAT, Kochi

Prof. B.P. Pal IIT, Delhi

Prof. Reji Phillip RRI, Bangalore

Prof. Asima Pradhan IIT, Kanpur

Prof. B.P. Singh IIT, Bombay

Prof. B.M. Suri BARC, Mumbai

Prof. C. Vijayan IIT, Madras

Editorial Committee (RRCAT, Indore)

Dr. C.P. Paul Dr. C.P. Singh

Mr. H.S. Patel Dr. S. Verma

Dr. G.J. Singh Dr. B.N. Upadhyay

Dr. Pankaj Misra Dr. S. Sendhil Raja

ILA Executive Committee Editorial Team of

Cover Photo:

Top Left: Schematic view of five lasers impinging on a single ion (See details on page no. 15).

Top Right: Oscillator amplifier setup.(See details on page no. 26)

Bottom: All-fiber based laser characterization. (See details on page no. 57)

President

Prof. S.K. Sarkar BARC, Mumbai

Vice President

Prof. L.M. Kukreja RRCAT, Indore

Gen. Sec. I

Prof. P.K. Dutta IIT, Kharagpur

Gen. Sec. II

Prof. K.S. Bindra RRCAT, Indore

Treasurer

Dr. S. Verma RRCAT, Indore

Regional Representatives

Dr. S.K. Bhadra CGCRI, Kolkata

Prof. M.P. Kothiyal IIT, Madras

Prof. D. Narayana Rao Univ. Hyderabad

Prof. H. Ramachandran RRI, Bangalore

Dr. A.K. Razdan LASTEC, New Delhi

Web Committee

Chairman:

Prof. P.A. Naik RRCAT, Indore

Webmaster:

Mr. Rajiv Jain RRCAT, Indore

A Bulletin of the Indian Laser Association

Contents

Vol. 25, No. 3, December 2014

Page No.

From the Editor 1

From the ILA President 2

1. Optical Coherence Tomography: Emerging Technology Trends and Applications 4Hrebesh M. Subhash

2. Trapped Ytterbium Ion for Optical Frequency Standards in India 13

P.K. Mukhopadhyay, P.K. Gupta, C.P. Singh, A.J. Singh, S.K. Sharma, K.S. Bindra and S.M. Oak

4. Investigations on the Efficacy of Chlorin p Conjugates for Photodynamic Treatment of 316

Cancer and Bacterial InfectionAlok Dube, Khageswar Sahu, Mrinalini Sharma and P.K. Gupta

5. Multi-Step Photoionization Spectroscopy of Lanthanides and Actinides: 39Measurements of Atomic Parameters Relevant to Isotope-Selective Photoionization ProcessesVas Dev

6. Advanced Fiber Optic Sensors for Nuclear and Industrial Applications 47Sanjay Kher

8. Fiber Laser Technology– Current Status and Activities by CSIR-CGCRI 56

9. Development of Continuously Tunable Mid-infrared Source and its Application in 61Laser Assisted Aerodynamic Isotope SeparationD.J. Biswas, J.P. Nilaya, M.B. Sai Prasad, S. Daga, G. Chakraborty, Ayan Ghosh, R.C. Das, A. Tak and A.K. Nayak

10. Ultrafast Dynamics in Nanostructured Materials 63J. Jayabalan, S. Khan, A. Singh, R. Chari, P. Zhou, C. Streubühr, K. Sokolowski-Tinten, Zi-An Li, M. Farle and U. Bovensiepen

11. Photoionisation of Atomic/Molecular Snow: Creating Highly Charged Matter using 71Low Intensity Laser PulsesS. Das, P. Sharma and R.K. Vatsa

12. Application of Laser Raman Spectroscopy to the Study of Actinides, Anomalous 75Thermal Expansion Materials and Ancient PaintingsT.R. Ravindran, A.K. Arora, K. Kamali, C. Ravi and T.N. Sairam

S. De, A. Rastogi, N. Batra, S.Panja and A. Sen Gupta

3. Development of Novel Ytterbium Doped Fiber Oscillators with Output in Diverse 20Temporal Format for Seeding of Multi-Stage Power Amplifier

7. Ultrafast Single Shot Coherent XUV Imaging using High Order Harmonics 53H. Singhal, K.H. Lee, S.B. Park and C.H. Nam

Ranjan Sen, Mrinmay Pal, Atasi Pal, Anirban Dhar, Maitreyee Saha, Sourav Das Chowdhury, Nishant Kumar Shekhar, Debasis Pal and Aditi Ghosh

1


rdThe 23 DAE-BRNS National Laser Symposium is going to take place at

Department of Physics, Sri Venkateswara University, Tirupati, during

December 3-6, 2014. This issue of Kiran is based on some of the invited talks

to be delivered during the symposium. We thank the authors for sending their

articles in time.

We hope you will find this valuable and interesting.

- Manoranjan P. Singh

From the Editor....

2


Dear Fellow Members and Readers,

Greetings from the Executive Council of Indian Laser Association (ILA).

At the outset let me express my gratitude to each and every members of ILA

for their unstinted support and cooperation in the growth of the Association.

As you are aware ILA is an organization devoted to promote education,

advancement and applications of laser science and technology in India.

Presently preparations are in full swing for organizing the 23 DAE-BRNS

National Laser Symposium (NLS-23) at SV University, Tirupati, A.P. from

December 3-6, 2014. NLS is a well established annual scientific conference

where scientists and engineers working in the area of lasers and photonics

participate from all major Indian laser establishments, research institutes

and universities. We expect nearly 500 participants in NLS-23 from all over

the country and abroad. I extend my best wishes for the success of NLS-23.

Formed 25 years back, ILA today is a well-knit vibrant family with members

totaling over 1000 with 14 Corporate members from all over the country.

Dissemination of scientific information to its members and others in this area

has been one of the important activities being pursued by the society. We

regularly arrange ILA Short Courses on topics of current interest just before

NLS to educate and train the future generation of researchers in this field. We

publish a quarterly magazine KIRAN, regularly since 1990. ILA is also

promoting active interaction with other professional bodies / institutions

especially in the field of Engineering, Biology and Medicine. More details of

our activities can be obtained from the web pages of ILA.

It gives me great pleasure to pen down a few thoughts about R&D activities

in the area of Laser and photonics in this special issue. The role of light in our

lives is both pervasive and primordial. Ultraviolet light probably had a role

in the very origins of life, and light-driven photosynthesis underlies all but

the most primitive of living things today. For humans, sight is the most

crucial of the senses for perceiving the world around us. Indeed, the highly

evolved vertebrate eye is one of the most exquisite light detectors ever

created. Yet light is influencing the way we live today what we could never

have imagined just a few decades ago. Laser and photonics demonstrate

powerfully the ties of fundamental science and technologies to the society ;

indeed, UNESCO has recently adopted a resolution declaring 2015 to be the

International Year of Light (IYL 2015). IYL has been the initiative of a large

consortium of scientific bodies to bring together many different stakeholders

rd

From the ILA President....

3


including scientific societies and unions, educational institutions,

technology platforms, non-profit organizations and private sector partners.

The importance of raising global awareness about how light-based

technologies can promote sustainable development and provide solutions to

global challenges in energy, education, agriculture, health care and security.

It has revolutionized medicine, opened up international communication via

the Internet, and continues to be central to linking cultural, economic and

political aspects of the global society. We have also plan to participate along

with other Indian organizations like Optical Society and request our

members to send their valued proposals for this celebration.

I hope that NLS-23 will provide a stimulating environment among the peer

and the young to open up newer frontiers. Once again, as President of ILA, I

extend most warm and hearty welcome to all the delegates and wish the

symposium every success in achieving its intended objectives. At the end, I

urge that researchers and corporates in this area to become member of ILA

and work with zeal and zest towards the progress of this exciting field.

Dr. Sisir K. Sarkar

4


reference optical delay line is in the other arm. The light interferes at the detector only when light reflected from the sample is matched in optical path length with that reflected from the scanning reference mirror. A single scan of the reference mirror thus provides a one-dimensional depth-reflectivity profile of the sample. Two-dimensional cross-sectional images are formed by laterally scanning the incident probe beam across the sample. The reconstructed OCT image is essentially a map of the changes of reflectivity that occurs at internal interfaces, similar to the discontinuities in acoustic impedance in ultrasound images.

As a multifunctional clinical diagnostic and monitoring technique, OCT has become a well-established tool in many areas including ophthalmology, dermatology, gastrointestinal endoscopy, intravascular imaging and oncology among others. OCT can be implemented using either time-domain (TD) or frequency-domain (FD) methods in which the interference signals generated by combining reference signals with light scattered by the target are detected. In recent years, the development and advancement of OCT technology have become an active area of research, which include development of optical sources, detectors and various type of beam delivery systems and patient interfaces for clinical imaging applications. Another important trend is the extension of OCT for functional imaging applications such as angiography, vibrometry, elastography etc. Presently, there has been an increasing interest in the development of cost-effective, compact and easy to use OCT platform for Point of Care (POC) diagnostic applications, which can enable rapid and accurate diagnosis and monitoring with reduced cost and time associated with health care services. The following sessions will describe the new trends in the developments of different specific devices and subsystems, which include optical sources, optical detection systems, and optical configurations for functional imaging and signal processing, and affordable OCT platforms for POC and personal care applications.

Different OCT schemes

In OCT, the coherence-gated information about the elementary volume of the scatters within the obscuring

Abstract

Optical coherence tomography (OCT) has become a well-established noninvasive optical imaging modality with a wide range of applications in biology, clinical medicine and material applications. OCT technology enables noninvasive real-time, simultaneous imaging of both three-dimensional cellular resolution tissue morphology as well as assessment of depth resolved function, which can significantly improve early medical diagnosis, contribute to a better understanding of disease pathogenesis, and enhance the monitoring of therapy. This manuscript will briefly and selectively review the current trends and state-of-the-art in the more prominent areas of activity in the OCT technology and its applications, which includes trend in the development of various specific devices and subsystems, such as optical sources, beam delivery devices, and signal processing techniques for various imaging applications. The topic also discusses recent advances in functional extensions of OCT technologies such as, OCT-based angiography techniques, OCT-based vibrometry, photo-thermal OCT for targeted imaging, and spectroscopic OCT etc.

Keywords: optical coherence tomography, medical imaging, optical instrumentation, biomedical optics, image processing.

Introduction

Optical coherence tomography (OCT) has emerged as a novel, non-invasive, optical imaging modality based on low coherence interferometry. It was first conceived in 1990 by Dr. Naohiro Tanno, a professor at Yamagata University, and then perfected in 1991 by Massachusetts Institute of Technology team headed by Prof. James Fujimoto [1]. OCT enables the non-invasive, non-contact imaging of cross-sectional structures in biological tissues and materials with high resolution. In principle, OCT is an optical analogue to clinical ultrasound. In OCT, the temporally gated optical pulse remitted from scattering sites within the sample is localized by low-coherence interferometry (LCI). This is typically achieved with a Michelson interferometer. Traditional time-domain OCT operates using broadband optical source with the sample rests in one arm of the interferometer and a scanning

Optical Coherence Tomography: Emerging Technology Trends and Applications

Hrebesh M. SubhashTissue Optics and Microcirculation Imaging Facility,National Biophotonics & Imaging Platform Ireland,

National University of Ireland, Galway City, Co Galway, IrelandE-mail: [email protected]

5


scattering specimen can be obtained from either the time domain measurement principle (TD-OCT) or the frequency domain (FD-OCT) measurement principle.

Time-Domain OCT

The time-domain OCT uses a variable optical delay scanning as shown in Fig.1 (a), the wave number dependent photo detector current is captured using a single detector where the reference mirror is scanned to match the optical path from reflections within the sample. Depending on the scanning modality and the signal detection scheme employed, the time-domain OCT can be further classified into three: 1) point-scan (flying scan OCT), 2) Line field (Linear-OCT) and Full-field (Wide-field) OCT [2].

Frequency-Domain OCT

Frequency domain OCT (FD-OCT) [3] is a variant

interferometric imaging modality. It has been widely attracted in the biomedical imaging field due to its higher sensitivity and imaging speed compared to conventional TD-OCT. The principle of FD-OCT relies on the transformation of the OCT time varying signal along the optical axis, termed as A-scan, into the frequency domain. The basic physics behind FD-OCT is based on the inverse scattering theorem. According to Wiener-Khintchine theorem the spectral power amplitude of the back scattered wave equals the Fourier transform of the axial distribution of the object scattering potential. FD-OCT has the advantage that the full sample depth information is obtained in a parallel manner such that no moving parts are necessary. Based on the implementation the FD-OCT can be divided into two classes, spectral domain OCT (SD-OCT) and swept-source OCT (SS-OCT). In SD-OCT the optical frequency components are captured simultaneously with a dispersive element and a linear detector, on the other hand in the SS-OCT the optical frequency components are captured by a single detector in a time encoded sequence by sweeping the frequency of the optical source. Recent studies have shown that Fourier domain OCT can provide signal to noise ratio that is more than 20dB better than the conventional TD-OCT.

Spectral Domain OCT

A typical implementation of fiber optics based SD-OCT setup is shown in Fig. 2(a). The back scattered low coherence light is mixed with a reference beam by the 2X2 fiber optics based Michelson interferometer, then the grating based spectrometer at the output of the interferometer separates each spectral components detected using a linear array detector. The SD-OCT detects the spectrally resolved interference signal with a spectrometer that consists of a high efficiency diffraction grating and a high-speed line camera. Fig. 2 (c) shows the reflectivity profile of the different target positions in the ocular medium and Fig. 2 (d) shows the detected intensity spectrum by the camera. In order to suppress autocorrelation, self-cross correlation, and camera noise artifacts, first, all the spectral interferogram in each slice along the x-direction (B-scan) were ensemble-averaged at each wavelength to obtain a reference spectrum, this background spectrum then subtracted from each A-scans. Since the Fourier transform relates the physical distance

(z) with the wave number k=2p/l, however, the spectra obtained with the SD-OCT spectrometer is not necessarily evenly spaced in k-space. In order to obtain a proper depth profile, the subtracted spectral

interferograms are then remapped from l-space to k-space by use of the spline interpolation method, as shown in Fig. 2(e). Due to the Fourier relation (Wiener-Khintchine theorem between the auto correlation and the

Fig.1: (a) Schematic of a typical fiber optics based time-domain OCT setup (b) Reflectivity profile of target position in the ocular medium (c) detected interference signal by the photodetector (d) detected signal with reference signal filtered out (e) demodulated and envelop detected OCT signal.

6


hundreds of kHz has been achieved, which drastically improves the imaging speed of the SS-OCT to compete with SD-OCT and TD-OCT in terms of imaging speed. Fig. 2 (b) shows the typical implementation of SS-OCT systems, which employ a broadband, rapid frequency-swept laser source and InGaAs photodetectors to perform Fourier domain OCT imaging without the use of a spectrometer. With such a wavelength-swept source, interference signals at individual wavelengths can be measured sequentially with high spectral resolution. This spectrally resolved data acquisition is central to frequency-domain ranging. This method offers significantly higher sensitivity than the time-domain ranging method used in conventional OCT.

Trends in OCT Subsystems and Beam Delivery Devices

In general, OCT system can be considered from a modular view point in terms of various integrated hardware and software functionalities such as imaging engine, low coherence light source, beam delivery and probes, computer control and image processing etc. Depending upon the applications, there are many variant embodiments of the interferometer and imaging engines such as polarization sensitive, Doppler flow imaging, optical angiography, spectroscopy, frequency scanning, spectral radar and parallel detection. The new developments, applications and advancements of OCT technology rely on these embodiments [1, 3].

Optical Source: The characteristics of the optical source determine the general performance of the system. For example, the short coherence length of the optical source determines the axial resolution and the nominal wavelength determines the achievable penetration depth,

of the system. The broader the line-width Dl of the source, the smaller the coherence length of the source is. There are four main criteria that need to be considered when choosing a light source for OCT. These parameters are wavelength, bandwidth, single transverse mode power, and stability. In the decade, a lot of research has been performed in the field of development of optical sources specifically to meet the high requirements in application for OCT. There are mainly two types of optical sources widely used for OCT applications, which include broadband light source for TD-OCT and SD-OCT systems and tunable sources, or swept source laser for SS-OCT systems. There are for the moment few main applications for which commercial systems are available, ophthalmology, cardiovascular, dermatology and gastroenterology. In ophthalmology, for retina imaging, the main bandwidth is that around 840 nm while for anterior segment imaging, the main bandwidth is 1300nm. Recently, OCT imaging in the 1050-1060nm wavelength region is getting much attention in the field of

Fig. 2: (a) Experimental Schematic of a SD-OCT set-up (b) Experimental Schematic of a SS-OCT set-up (c) reflectivity profile within the anterior chamber of an eye (d) detected spectral interferogram by the camera (A-scan) (e) Remapped Spectral interferogram in K-space with back ground subtraction (f) Fourier transformed spectrum.

spectral power density) the depth resolved information can be immediately reconstructed by a Fourier-transformation from the remapped spectra, without movement of the reference arm, as show in Fig. 2(f).

Swept source OCT

The concept of FD-OCT can also be implemented using a tunable laser source over a broad spectral range in conjunction with a single detector. The FD-OCT of this type has been called swept source OCT (SS-OCT). Despite the difference in system configuration, both SD-OCT and SS-OCT have a common net result: the OCT signal is sampled in spectral domain and a SNR improvement is gained because of the Fourier reconstruction. In SS-OCT the time required to tune the wavelength determines the time to produce an A-scan. Recently swept source with sweeping frequency of few

7


broad spectral bandwidths (5 m axial resolution) centered around 1,300 nm wavelength, for deeper imaging penetration in highly scattering tissue. Both of these laser sources, however, are large, require additional pump laser sources, water-cooling, and an experienced operator to align and maintain them.

Swept source lasers: Swept-source lasers are tunable light source emitting one wavelength at a time, rapidly swept over a broad spectral range. The promising capabilities of frequency swept light sources for OCT imaging have gained intense interest in their development and enabled a quantum leap in the speed and sensitivity of OCT technology development. There are three major concepts to achieve high speed tuning depending on the method used for wavelength selection inside the laser cavity of the wavelength swept source: one based on a fast rotating polygonal mirror, the second based on a diffraction grating on a mechanically resonant galvo-scanner and the third using a fiber Fabry-Perot tunable filter (FFP-TF). For very high tuning speeds and to overcome limitations given by the buildup time of lasing in the cavity, the technique of Fourier Domain Mode Locking (FDML) has been introduced [4]. FDML lasers can achieve ultrahigh sweep rates up to 5.2 MHz by buffering or multiplexing the sweeps. FDML works optimally at 1.3 μm and 1.5 μm wavelengths where optical fiber dispersion and loss are negligible. However, dispersion can be compensated using fiber Bragg gratings to improve performance at 1 μm and 1.3 μm wavelengths. Another recent development is the miniaturization of external cavity tunable lasers using microelectromechanical systems (MEMS) technology and which lead to an increase in sweep rates enabling OCT imaging up to ~ 1 MHz axial scan rates. Presently commercial SS-lasers and OCT engines are available at wavelengths around 840 nm, 1060 nm, 1310 nm and 1550 nm. However, most technologies require that the MEMS filter bandwidth be broad enough to tune multiple longitudinal modes in order to reduce excess noise associated with mode competition. Consequently, the coherence length of MEMS-tunable short cavity lasers can be limited. The reduction of laser cavity length to achieve single longitudinal mode operation significantly improves SS-OCT performance. This can be achieved using a new configuration called vertical-cavity surface emitting laser (VCSEL) technology [5]. Although VCSELs were developed in late 1970s, applications were limited to photonics. Recently, OCT imaging using MEMS-tunable VCSELs at 1300 nm and 1060nm were reported. However, the mechanical movement, which represents the key aspect around which the above mentioned SS-laser technology designs are based on, can also represent the limiting factor of the technology itself, limiting, by consequence, laser performance and imaging

μophthalmology, where imaging the deeper layers of retina are very crucial of diagnosing retinal pathologies. OCT systems for dermatology, cardiovascular and gastroenterology utilize the 1300nm bandwidth, for less scattering.

Superluminiscent Diodes (SLD): Broadband sources are mainly SLD and current SLDs operate at various wavelengths, such as 0.8 μm, 1 μm, 1.3 μm and 1.55 μm and all of these wavelengths have been used for OCT in different applications. At the 830 nm range, the SLD's are based on Al Ga As emitters and single SLD with up to 70 nm bandwidths and 3-15mW fiber pigtailed output power (corresponding to around 3.3μm axial resolution in the tissue) are commercially available (Superlum Ltd, Russia). However, using multiplexed SLD technology (synthesized optical source), SLD with a bandwidth more than 130nm and optical power up to 15mW are commercially available (Denselight Semiconductors, Singapore). While in some specific areas, like ultra-high resolution OCT, there are strong competitors to SLDs, namely femtosecond lasers/supercontinuum sources, for most of practical applications SLDs are now considered as the most attractive emitters due to their small size, easy to use, and much cost effective compared to alternative light sources.

Supercontinuum (SC): SC lasers are new type of light sources based on nonlinear optical phenomenon. Supercontinuum light is generated by invoking high optical nonlinearity in a material. Typically, mode-locked pulsed laser sources are used in the near infrared (1,064 nm)in the nanosecond, picosecond, or femtosecond range, ensuring high peak powers to drive the nonlinear effect in a material, which ''breaks'' the pulse out into a SC spectrum. The SC combine high brightness with broad spectral coverage, a combination offered by no other technologies. Over the last decade, the development of supercontinuum lasers has been driven by both technological advances and ever-evolving market requirements especially in the field of biomedical imaging. Presently, turnkey supercontinuum fiber lasers specifically designed for OCT applications are commercially available (NKT Photonics, Denmark and Fianium, UK).

Kerr-lens mode-locked (KLM) lasers: KLM are another class of broadband optical source getting more attention in the field of OCT. The titanium: sapphire laser is tunable from 0.7μm to 1.1μm and can produce not only broad spectral bandwidths for high-resolution imaging, but also high output powers for fast image acquisition. Ti: Al O 2 3

laser are optimized for short coherence length and were demonstrated to achieve sub-2-μm resolution with a power exceeding 100mW. The chromium: forsterite laser has also been used to generate high output powers and

8


robustness and better spectral with high-speed. On-chip spectrometers based on digital planar holography (DPH) is another competing technology for the development of affordable SD-OCT system [8]. The idea of (DPH) is to let the light travel inside a hologram for thousands of wavelengths in order to increase significantly the possibility for light processing. This technology allows easy writing of an arbitrary computer-generated hologram onto a planar waveguide with very long light pathway and for creating and reproducing photonic transfer functions of very high-resolution spectrometer-on-chip. Arial type detectors are mainly used with full-field OCT (FF-OCT) configuration in which tomographic images in the en face orientation are acquired with an area camera and by illuminating the whole field of view using a low-coherence light source. By using the latest advances in CCD/CMOS detectors, high performance aerial cameras are commercially available for specific applications of FF-OCT.

Beam delivery system: The main objective of the beam deliver system is to deliver and focus the probing beam light onto the target tissue surface and to collect the backscattered light back in to the detector system. There are a broad range of beam delivery systems available depending upon the intended OCT application.

i) Bench top/Hand-held scanner: Bench top and hand-held scanners are the most widely used scanners for lateral scanning applications, especially for dermatological applications and research grade systems. The key feature of hand-held scanners is to provide convenient access to clinical environment for tissue examination. Galvanometric scanners are generally used for the deflection of the scanning mirror. Another approach is beam deflectors based on polymorph or bimorph actuators. Recently, to enable a more portable OCT instrument, MEMS (Micro-Electro-Mechanical Systems) based hand-held scanners are getting much attraction [9].

ii) Endoscopic probe: One of the primary design consideration of endoscopic probe is to minimize the probe diameter. To this end, the single-mode fiber used for the fiber optics OCT systems is ideally suited for this purpose. Other important technical characteristics of an OCT probe are scanning range, field of view, speed, and flexibility. Endoscopic OCT probes can be divided into two groups based on their scan mode implantation-side-imaging and forward-imaging probes. Side-imaging probes can be further divided into two categories: (a) circumferential and (b) linear (or translational) scanning probes. Forward imaging OCT endoscopes emit and collect light in front of the probe to provide structural information in the forward direction of the probing catheter. The implementation of a forward imaging

quality. The akinetic, all-semiconductor laser is an innovate technology which involves no mechanical moving parts to generate the sweep [6]. The laser is based on integrated semiconductor opto-electronic design without the need of external cavity coupling. The all-semiconductor laser cavity is ~2 mm in length and is monolithically-constructed within the semiconductor. The all-semiconductor design enables a full electronic control of laser operation. The akinetic, all-semiconductor technological approach and design allow the akinetic laser to overcome most of the limitations encountered with mechanical sweep-based design implementations.

Optical sensors and detectors: The selection of the right detection system is one of the crucial parameter for the design of an OCT system. The optical detection system for OCT can be classified into point, line and areal type detectors according to the OCT system configuration. Convectional point scan OCT based time-domain and frequency domain based swept source configurations use point detectors. Balanced detection (BD) is a commonly used detection method in many optics experiments with the need for increased signal-to-noise (SNR). In comparison to the unbalanced configuration, the BD configuration in TDOCT can yield a SNR in excess of 6 dB since the signal current is double that of the single detection and also because of the additional capability of balanced reception to suppress relative intensity noise (RIN). For fast TDOCT systems, the use of the balanced configuration can result in a better than 10 times improvement in the SNR. The use and advantages of BD in OCT have been amply demonstrated in TDOCT and SSOCT setups. Presently a range of BD are commercially available specifically for OCT applications from vendors like Thorlabs Inc (USA), Newport Corporation (US) and Santec Corporation (Japan). Linear array based detectors are generally used with Fourier-domain spectral OCT systems and time-domain line scan OCT systems. Point scan SD-OCT is a widely adapted and commercially successful OCT technology and currently a wide range linear array based spectrometers are commercially available in both NIR and short-wave NIR range based on silicon CCD/CMOS camera and InGaAs photodetector array, respectively, which include Bayspec Inc (USA), Wasatch Photonics (USA) and Bioptigen Inc (USA). However, traditional optical spectrometers are often large, bulky, expensive, and delicate. Recently nanophotnics based miniature and affordable spectrometers are specifically designed to meet these challenges in SD-OCT. Tornado Spectral System, Canada developed an on-chip spectrometer based on a planar light wave circuit (PLC) called OCTANE (Optical Coherence Tomography Advanced Nanophotonic Engine) [7], which can provide small form factor,

9


Optical Doppler tomography or Doppler OCT (D-OCT) [12] is an example of a purely phase-based technique for imaging flow velocity of moving particles in a highly scattering medium. In D-OCT, based on SD or SS implementation, the blood circulation is evaluated by taking the phase difference between adjacent A-line scans in B-frame. Although the D-OCT algorithm is capable of imaging and quantification of flow velocity in relatively large blood vessels, the velocity of the dynamic component of small vessels are underestimated due to the presence of static scattering. Moreover, this technique is sensitive to the Doppler angle and is unable to detect the flow components perpendicular to the scanning beam. Phase variance and Doppler variance are other alternative phase-based methods developed for the visualization of microcirculation. Unlike D-OCT, these methods are insensitive to Doppler angle. These methods are capable of detecting both transverse and axial flow, and do not require any non-perpendicular beam of incidence. Optical micro-angiography (OMAG) [13] and rapid volumetric angiography are examples of techniques that utilize the complex field of both amplitude and phase of the OCT signal. OMAG utilizes a modified Hilbert-transform-based algorithm to separate the dynamic scatters from static tissue background. By applying the OMAG algorithm along the slow scanning axis, high-sensitivity imaging of capillary flow can be achieved. For obtaining high sensitivity, OMAG requires the removal of bulk motion artifact by resolving the Doppler shift. Recently, the flow sensitivity of OMAG has been enhanced using a new processing and scanning protocol termed ultra-high-sensitive OMAG, which utilizes the OMAG algorithm in the C-scan direction to obtain high-sensitivity flow map. To date, based on SDOCT technology, a couple of OCT angiography techniques utilizing the complex field of the OCT signal have been

proposed by various research groups. However, phase-based methods are more susceptible to the axial movement of bulk tissue and other sources of motion artifacts such as galvanometer jitter, physiological motion, and thermal drift, which require more sophisticated methods of the bulk motion phase correction. On the other hand, magnitude-based angiography techniques are purely based on the amplitude of the OCT signal. Techniques termed speckle-variance OCT and split-spectrum amplitude decorrelation angiography are examples of magnitude-based methods. Unlike phase-based technique, magnitude-based techniques are insensitive to bulk phase changes and, therefore, do not require any sophisticated phase correction methods. Correlation mapping OCT (cm-OCT) is another purely magnitude-based angiography technique developed by our group, which takes advantage of the time-varying speckle effect, which

system within a narrow probe is more technically challenging. Generally fiber–GRIN lens probe assembly in conjunction with PZT cantilever were used for accomplishing the forward scans. Recently, forward-imaging probes based on MEMS at the distal end of endoscope also used for this application. The side scanning OCT system, the probe light is emitted from and collected at the side of the endoscope. The general design of such a probe consists of a rod mirror or prism attached to a rotation assembly to deflect the emitted light from the optical fiber tip out of a window on the side of the probe. The rotation was achieved by attaching the proximal end of the flexible probe to a motor and a gear systems. Besides rotation, the linear translation of the OCT probe can also be used to achieve side imaging by back and forth translation of the probe to achieve 3D imaging and this is called pull back system. This is a key requirement for intravascular imaging OCT (IV-OCT) applications. IV-OCT with synchronous motor based circumferential scanning was recently demonstrated with a scan speed up to 3200 fps [10]. A circumferential scanning endoscopic capsule was recently designed for use in screening patients for Barrett's esophagus, a major precursor to esophageal cancer. It has several advantages over traditional endoscopy [11].

Trends in Functional Extensions of OCT Imaging

In conventional OCT, the contrast mechanism relies upon the spatial variation of the coherent back scatters within a cross-sectional plane or volume of tissue or materials. However, a number of other intrinsic and extrinsic properties of the tissue have been demonstrated to provide information about the morphological changes by altering the amplitude, phase or polarization states of the probing beam. Exploiting these properties, thus providing novel imaging contrast mechanisms, enhances the clinical and biomedical applications for OCT to achieve functional imaging that can reveal more details about the tissue dynamics and physiology. Some of the important functional modes of OCT are Angiographic OCT, polarization sensitive OCT, Doppler OCT, spectrometric OCT, OCT vibrometry, photothermal OCT and differential absorption OCT.

Angiographic OCT: OCT-based angiography techniques utilize the red blood cell scattering dynamics as the contrast mechanism, which exhibits phase or amplitude fluctuations over time, while the static tissue scattering is relatively constant over time. There are mainly three categories of OCT-based angiography methods, which are essentially based on the utilization of the complex nature of the OCT signal to obtain the microcirculation map: phase-based techniques, magnitude-based techniques, and techniques that use the complex data, incorporating both magnitude and phase information.

10


locations in a depth-resolved manner. Recently, there is a growing interest in the field of optical vibrometry based on optical coherence tomography (OCT) for various type of clinical and pre-clinical studies [15]. Fourier-domain OCT vibrography is a unique tool for studying middle ear and inner era function. Feasibility of OCT vibrography for studying the function of cochlear transduction was recently demonstrated. OCT vibrometry is an indispensable tool for studying the middle ear function [16].

Polarization-sensitive OCT: Most of the body tissues such as muscle, tendons and nerve fibers contain collagen and elastin, which exhibit birefringence when they form a specific structure. Birefringence depicts a change in polarization state of light due to the refractive index difference for light polarized in two orthogonal planes. The propagation of light through such birefringence sample may alter the optical polarization state of reflected light. Therefore, polarization sensitive measurements at the output of the interferometer can provide depth-correlated information about the birefringence nature of the material or tissue specimen. As well as providing added contrast, changes in birefringence may indicate changes in functionality or structure of the tissue. Since the first report of functional polarization sensitive OCT (PS-OCT) system [3], a variety of PS-OCT configurations have been investigated. They all differ in complexity, capabilities and signal processing schemes. The most complete information about the polarization properties of a biological specimen is given by system capable of producing depth resolved Muller matrix elements. These configurations can account for the depolarization as well as the changes in total, linear and circular degree of polarization of the probe beam during propagation in tissue.

Spectroscopic OCT: Spectroscopic OCT is an alternative mode of OCT, which provides further access to the composition and functional state of the specimen. Spectroscopic OCT can be implemented in a variety of ways. One approach is based on spectral ratio imaging of OCT images using two or more spectral bands with wavelength division multiplexer, which combine lights with different wavelength and then electronically distinguish the resulting signal by their different Doppler shift resulting from the reference arm scan. By use of state-of-the-art ultra-broadband femtosecond Ti:Al2O3 lasers, spectroscopic imaging over the wavelength range from 650 to 1000 nm has been reported by Morgner et. al [17]. Another implementation is based on estimation of depth resolved spectral of the source spectrum, in which the modification of the source spectrum caused by the sample can be measured directly from the Fourier domain

is normally dominant in the vicinity of vascular regions compared to static tissue region. It utilizes the correlation coefficient as a direct measurement of decorrelation between two adjacent B-frames to enhance the visibility of microcirculation.

Photo-thermal OCT: OCT has several advantages over other clinical imaging technologies, as OCT enables real-time, in situ and in vivo visualization of tissue microstructures with image resolution scales approaching those of histopathology. However, as OCT relies on detecting the coherence scattering from the refractive index boundaries of tissue microstructures, OCT imaging provides only limited molecular information of clinically relevant structures. This is due to the fact that OCT is intrinsically insensitive to incoherent scattering phenomenon such as fluorescence and spontaneous Raman scattering, which are the key to molecular imaging. Thus, there is a great interest in enhancing the utility of OCT for molecular imaging through the incorporation of extrinsic contrast agents such as iron oxide particles, proteins, dyes and various types of gold nanoparticles. Photo-thermal OCT (PT-OCT) [14] is an alternative method to obtain targeted imaging contrast with the photothermal heating phenomenon, where photon absorption by an imaging target of interest (e.g., an absorbing nanoparticle) leads to a temperature change in the environment surrounding the target. A heating laser is collinearly coupled with the imaging beam to provide local temperature change in the surrounding environment of the target. The induced local temperature changes cause thermoelastic expansion of the sample and shifts in the local index of refraction, which can be directly detected using the OCT phase sensitive method. Recently, a range of contrast agents such as, gold nanoparticles, dye loaded PLGA nanoparticles, carbon nanotubes and various types of gold nanoparticles were demonstrated with PT-OCT method.

OCT Vibrometry: Noninvasive depth-resolved imaging technique for measuring acoustic vibration are potentially important for studying the structural dynamics of physical and biological systems in many areas of research including biology and medicine. Conventional laser Doppler vibrometry (LDV) techniques have a series of undoubted advantages, such as high sensitivity, non-contact nature of the probe, and the unparalleled characteristics offered by the use of coherent monochromatic laser light. However, LDV is incapable of providing high-resolution depth-resolved cross-sectional map of structural and vibrational information. This is because the long coherence length of the laser attributes interference over a long range and yields a complex signal that is not assigned to target

11


which can enable a greater than 100x reduction in size and cost. MR-OCT architecture promises to fit into a robust, cost-effective design (~€10), which can be largely solid state, and can be implemented by the optics and process technology used for the production of CD/DVD ROM pick-up head technology (CD/DVD PUH) to address a variety of high volume applications.

Conclusion

OCT is a high resolution, non-invasive, 3D- imaging technique with great potential in both clinical and fundamental research application in many areas. Its wide range of current applications ranges from medical diagnosis and surgical guidance to the characterization of polymer micro-structures and reading of multi-layered storage media etc., which indicate that OCT will play a major role in practical scientific innovation and research in years to come. Therefore, it is essential that OCT technologies are developed further as an enabling measurement technology. The exceptionally high spatial resolution and velocity sensitivity, the functional extension of OCT technique can simultaneously provide tissue structure, blood perfusion, birefringence, and other physiological information has great potential for basic biomedical research and clinical medicine.

References

1. Hrebesh M. S and R K. Wang, in “Advanced Biophotonics: Tissue Optical Sectioning,” ISBN: 9781439895813 CRC Press (2012).

2. Hrebesh M. S, Advances in Optical Technologies, Volume 2012 (2012), Article ID 435408.

3. Hrebesh M.S and Ruikang K. Wang, in “Biomedical Optical Imaging Techniques: Design and Applications, Ron Liang, editor, Springer (2012).

4. R. Huber, M. Wojtkowski, and J. G. Fujimoto, Opt. Express 14, 3225-3237 (2006)

5. Jayaraman V., Cole G.D., Robertson M., Uddin A. and Cable A., Electron. Lett. 48, 867–868 (2012).

6. M. Bonesi, M.P. Minneman, J. Ensher, B. Zabihian, H. Sattmann, P. Boschert, E. Hoover, R. A. Leitgeb, M. Crawford, and W. Drexler, Opt. Express 22, 2632-2655 (2014).

7. Arthur N., Kyle P., Nicolás S D, Bradley S.S. and Arsen R.H., Proc. SPIE 8934, XVIII, 89340F (March 4, 2014).

8. C. Peroz, C. Calo, A. Goltsov, S. Dhuey, A. Koshelev, P. Sasorov, I. Ivonin, S. Babin, S. Cabrini, and V. Yankov, Opt. Lett. 37, 695-697 (2012).

processing of cross-correlation interferometric data.

Nano-Sensitive OCT: Depth resolved label-free detection of structural changes with nanoscale sensitivity is an outstanding problem in the biological and physical sciences and has significant applications in both the fundamental research and healthcare diagnostics arenas. Nano-sensitive OCT (ns-OCT) [18] is a new approach for obtaining label-free depth resolved sensing technique to detect structural changes at the nanoscale. In ns-OCT, the structural components of the 3D object, which is spectrally encoded in the raw OCT interferogram are transformed from the Fourier domain into each voxel of the 3D OCT image without compromising sensitivity. Spatial distribution of the nanoscale structural changes in the depth direction is visualized in just a single OCT scan. This label free approach provides new possibilities for depth resolved study of pathogenic and physiologically relevant molecules in the body with high sensitivity and specificity. It offers a powerful opportunity for early diagnosis and treatment of diseases. Experimental results show the ability of this approach to differentiate structural changes of 30 nm in nanosphere aggregates, located at different depths, from a single OCT scan, and structural changes less than 30 nm in time from two OCT scans. Application for visualization of the structure of human skin in vivo is also demonstrated.

Trends in Affordable OCT Platform

In recent years, there has been an increasing interest in the development of cost-effective, compact and easytouse OCT platform for POC diagnostic applications, which can enable rapid and accurate diagnosis and monitoring with reduced cost and time associated with healthcare services. Among them, integrated optics based FD-OCT is an emerging technology and has the potential to provide a small form factor and be more cost efficient. However, integrated optics technology is still in its infancy and the technology is not yet well-established for commercial applications of OCT technology. Unlike bulk or fiber optics, several technological difficulties need to be overcome in the field of integrated optics. Firstly, the insertion losses in an integrated optics system are considerably larger compared to fiber or micro-optics based systems. Secondly, coherent systems are extremely sensitive to reflections and these artifacts can drastically affect system sensitivity and imaging performance. Moreover, the commercial availability of robust and low cost integrated optics compatible light sources and array detectors are another challenging issue to be resolved. Recently, a small form factor on chip spectrometer called OCTANE is developed based on silicon photonics for SD-OCT application by Tornado Spectral Systems, Canada [7]. Another competing technology is Compact Imaging Inc.'s multiple reference OCT (MRO™) [19],

12


McCarty, Opt. Lett. 37, 981-983 (2012).

15. Hrebesh M.S., Anh N.H., R.K. Wang, S.L. Jacques and Alfred L.N., Biomed. Opt. 17(6), 060505 (2012). DOI: 10.1117/1.JBO.17.6.060505.

16. Hrebesh M.S., N. Choudhury, S.L. Jacques, R.K. Wang, F. Chen and A.L. Nuttall, J. Biomed. Opt. 18(3), (2013).

17. U. Morgner, W. Drexler, F.X. Kärtner, X.D. Li, C. Pitris, E. P. Ippen, and J. G. Fujimoto, Opt. Lett. 25, 111-113 (2000).

18. S.A. Alexandrov, Hrebesh M.S., A. Zam and M. Leahy, Nanoscale 6, 3545 (2014).

19. Roshan D., Hrebesh M.S., Kai N., Josh H., Carol W. and Martin L., Biomed. Opt. Express 5, 2870-2882 (2014).

9. C.D. Lu, M.F. Kraus, B. Potsaid, J.J. Liu, W. Choi, V. Jayaraman, A.E. Cable, J. Hornegger, J.S. Duker, and J.G. Fujimoto, Biomed. Opt. Express 5, 293-311 (2014).

10. T. Wang, W. Wieser, G. Springeling, R. Beurskens, C.T. Lancee, T. Pfeiffer, A.F.W. van der Steen, R. Huber and G. van Soest, Opt. Lett. 38, 1715-1717 (2013)

11. M.J. Gora, Nat. Med., 19, 238–240 (2013); doi:10.1038/nm.3052.

12. Hrebesh M.S., International Journal of Optic, March 2011. DOI:10.1155/2011/29368

13. Hrebesh M.S. and R.K. Wang, in “Microcirculation Imaging Book, Martin J. Leahy, editor, ISBN-10:3-527-32894-7 Wiley-VCH, Weinheim (2012).

14. Hrebesh M.S., Hui Xie, Jeffery W. Smith, Owen J.T.

13


structure constant, electron-to-proton mass ratio also require precise clock. Experimental verification of such fundamental sciences could be revealed using clocks

-18having fractional accuracies (Δν/ν ) better than 10 .o

Accurate measurement of the base unit of time, the second, plays a major role in metrology. Time interval, and its reciprocal, frequency, can be measured with orders of magnitude higher resolution and less uncertainty than any of the other physical quantities. Thus definitions of many other physical quantities depend on time and there accuracies depend on accuracy of second. In view of all this, operation of highly accurate and reliable time and frequency standards is a priority for any country and in India, CSIR-NPL has the mandate of maintaining the Indian Standard Time (IST). In SI unit, “one second" is defined as the time required for 9,192,631,770 cycles between doubly splitted hyperfine

133 133ground states of cesium ( Cs) atoms. Currently Cs-atomic fountain clocks serve as the primary frequency

-16 standards with Δν/ν ~10 [4] and here we are operating o1 3 3 - 1 5the Cs-atomic fountain with Δν/ν ~10 [5]. o

Developments of optical frequency standards, in last one decade, pave the way for atomic clocks with fractional

-18uncertainty ~10 . Currently we are developing an optical frequency standard using laser cooled and trapped

171 +single ytterbium-ion ( Yb ) in a Paul trap. In this paper, we give an overview of our developments.

From Earth's Rotation to Quantum Oscillators

Frequency standards are referenced to a well characterised periodic “tick-tock” event that repeats at a constant rate and is least perturbed. Such periodic events are produced by a device called the resonator. Such tick-tock occurs at a rate called the resonance frequency f, which is the reciprocal of the period of oscillation, T. Any clock needs an oscillator and a counter to count the number of ticks / tocks. The quest for accurate and stable timekeeping has been one of man's favourite pursuits in last few centuries as depicted in Fig. 1(a), which shows the improvement of the time keeping. The evolution from the oldest mechanical clocks that were capable of showing one second accuracy in few minutes, to the present atomic clocks which maintains the time with one second accuracy over the age of the universe (about 14

Abstract

Accuracy of the atomic clocks based on transitions in -18optical regime go up to 10 and thus they have

applications in various fields ranging from state-of-the art technology to precision experiments for searching fundamentals in science. At CSIR-NPL, we have started developing an atomic clock at the optical wavelength using trapped and laser cooled single ytterbium-ion. Ytterbium-ion has three narrow transitions suitable for using them as clock; out of them the octupole transition at wavelength 467 nm is the most preferable one for an accurate frequency standard and that we will be probing. We have opted the suitable design of the Paul trap of the end cap geometry for trapping the ion. This will create nearly pure quadrupole trapping potential and at the same time this will provide enough optical access to the trap centre. We have also estimated systematic uncertainties that are expected from our experiment. Our estimation predicts the octupole transition of ytterbium provides order of magnitude better frequency standard than its quadrupole transition at the wavelength 435 nm. Currently the design of the trap has been finalized and it is in the process of fabrication. We are working for setting up the lasers required for photoionization of the atom and laser cooling of the ion. In this paper we give an overview of the atomic frequency standards and development of our ytterbium-ion optical frequency standard experiment.

Keywords: Atomic clock, optical frequency standards, ion trap, laser cooling, field programmable gate array.

Introduction

Time with different levels of accuracies are used in our society in various applications ranging from accurate positioning of satellites, detecting the location of enemy missiles or targets, satellite based navigation, communication, operations of electric power grids, exploring the fundamentals of science [1-3]. Grand unification of the Standard model by developing quantum theory of gravity can be pursued through precision experiments for testing Einstein equivalence principle, local position invariance, local Lorenz invariance, distinguishing general relativity from other metric theories of gravity and so on. Accurate measurement of geodesy, temporal constancy of fine

Trapped Ytterbium Ion for Optical Frequency Standards in India

*S. De , A. Rastogi, N. Batra, S.Panja and A. Sen Gupta

CSIR-National Physical Laboratory, Dr. K. S. Krishnan Marg, New Delhi – 110012.*E-mail: [email protected]

14


billion years) took place over a period of 700 years.

Early astronomers noted rotation of the Earth's on its axis as a natural oscillator and defined 1 s as 1/ 86400 of the duration of a solar day. However, seasonal fluctuation of a day becomes the bottleneck for accurate timekeeping. Galileo's pendulum clocks served as precise timekeepers

thtill early 20 century with uncertainties of few seconds in one year. The invention of quartz oscillators set a new level of accuracy but soon it was realized that their resonant frequencies are uniquely determined by their dimensions and fabrication process influenced by temperature, air pressure, vibrations and gravity [6].

9 Quartz oscillators provide accuracy of 1 part in 10 and they have poor long term stability due to aging. Atomic systems, unlike crystal oscillators, offer electronic transitions as frequency references which are identical in all atoms of the same species and also mostly unperturbed. Thus, up till now atomic systems provide most accurate tick-tock and have been used as frequency

thstandards since mid-20 century. The first atomic clock was developed in 1949 based on the inversion transition in ammonia at 23.8 GHz and the first cesium- clock was

built in 1955 at NPL-UK. In 1967, the SI unit of second was redefined in terms of 9.2 GHz microwave transition

133of Cs [7]. However, the accuracy increases further for clocks operating in optical frequencies, i.e., few hundreds of THz.

Atomic Clocks

Atomic clocks use excitation between their quantized energy levels as resonance frequency ν , which can be o

measured precisely for a narrow transition linewidth. An atomic clock involves locking an external oscillator to a narrow atomic transition and precisely counting the atomic excitations, which is nothing but the resonance frequency of the external oscillator. The time required for ν excitations is 1 s [8]. The excitation frequencies could o

be either in the microwave or in the optical regimes. The stability of the clock increases either with number of atoms N or with the integration time τ. A measure of the stability is given by the Allan deviation as

(1)

A microwave frequency standard has so far achieved its -16 133best fractional systematic accuracy at 2×10 in a Cs

fountain clock [9].Operating frequencies of the microwave and optical clocks differ by at least four orders of magnitude and hence their accuracies. Figure 1b shows the advancement in the accuracies of the microwave and optical atomic clocks. There are two contenders for optical frequency standards: ensemble of atoms stored in optical lattice and single ion trapped in an ion trap. Optical lattice clocks have a very good signal-to-

5noise ratio (S/N) in a shorter τ due to ~10 atoms. However, continuously changing density and hence collisional shift due to decay from the trap leads to poor long term stability. In comparison, trapped single ion clock shows better long term stability but this requires longer τ to acquire reasonable S/N. Table 1 shows best reported optical clocks based atomic and ionic species.

Ytterbium-ion has three ultra-narrow optical transitions 2 2suitable for clocks. Out of them | S ; F=0, m =0> - | D ; 1/2 F 3/2

2F=2, m =0> quadrupole transition (E2) and the | S ; F=0, F 1/22m =0> - | F ; F=3, m =0> octupole transition (E3) have F 7/2 F

natural linewidths 3.02 Hz and 1 nHz, respectively (Fig. 2 a).At CSIR-NPL we are building an optical clock using

171 +the E3-transition of Yb [19] as it offers several advantages: m =0 states associated to the clock transition F

are insensitive to the first order Zeeman shifts and the E3-transition offers highest sensitivity for measuring the temporal variation of the fine structure constant among all the species [20].

Fig.1: (a) Historical improvements in the accuracy of clocks particularly the duration in which clocks show 1 s inaccuracy. (b) Advancement atomic clocks operating in microwave and optical frequencies.

15


heated oven and photoionize them using a pair of laser beams at wavelengths 399 nm and 369.5 nm. The ions will be trapped in a Paul trap and laser cooled to mK temperature using. Initially, few ions would be trapped; then, by gradually decreasing the trap depth, the high energy ones are made to escape from the trap. This process continues until a single ion is left in the trap which can be confirmed from the detected fluorescence level. Finally, that laser cooled ion will be interrogated by the sub-Hz linewidth clock laser.

Energy Levels of Ytterbium-ion

Laser cooling is required to reduce kinetic energy of the ion so that its motion gets confined within a sub wavelength spatial extension known as the Lambe-Dicke regime [21]. This is necessary for the cancellation of the

2 2first order Doppler shift. The strong | S ; F=1> - | P ; 1/2 1/2

F=0> transition at wavelength 369.5 nm will be used for laser cooling of the ion (Fig. 2a). The off-resonant scattering of that light populates the F=0 ground state and recovering them to the cooling cycle requires a light which is 14.7 GHz up shifted from the cooling laser. The

2excited P state branches to the lower lying, long lived 1/22 2D and F states; thereby bringing the ions out of the 3/2 7/2

laser cooling cycle. Additional pairs of repump lasers at 935 nm and 760 nm are thus required for bringing the ions

2 2back to the cooling cycle. The | S ; F=1> - | P ; F=1> 1/2 1/2

transition will be driven to optically pump the laser 2cooled ion to | S ; F=0, m =0>state. This state selection, 1/2 F

Table 1: List of optical frequency standards and their accuracies based on laser cooled and trapped single ion and neutral atoms.

Single Trapped Ytterbium-Ion Optical Frequency Standards

171 +The elaborate experimental setup for the Yb optical frequency standards consist of an ion trap mounted inside of an ultra high vacuum (UHV) chamber, optical arrangement for five lasers at different wavelengths, large optics setup for all these lasers and their frequency stabilization, high efficiency and resolution imaging system for detecting single ion, a sub-Hz linewidth highly stabilized clock laser for probing the clock transition and frequency comb for using it as a frequency standard. We plan to use Ytterbium atoms coming from a resistively

Species Δν/ν Wavelength of clock Referenceo

transition (nm)199 + -17Hg 1.9 × 10 282 [10]171 + -17Yb 7.1 × 10 467 [11]88 + -17Sr 2.3 × 10 674 [12]40 + -16 Ca 6.5 × 10 729 [13]115 + -13 In 2.35 × 10 237 [14]27 + -18Al 8.6 × 10 267 [15]171 -18Yb 1.6 × 10 578 [16]88 -18Sr 6.4 × 10 698 [17]199 -15Hg 5.7 × 10 266 [18]

171 +Fig. 2: (a) Energy levels of Yb which are relevant for our experiments. Dotted lines indicate ultra-narrow clock transitions. (b) Schematic view five lasers impinging a single ion with an accuracy of ~10 μm. The ion is trapped using electric fields V (t) and its T

position is fine-tuned using the compensation electrodes kept at potentials E - E , a high numerical aperture lens collects the 1 4

fluorescence for detection.

16


Ion Trapping

Trapping a particle at a particular point requires a force which vanishes at that point and increases in all directions linearly with its position. Confinement of an ion of charge Q requires either an alternating electric field (Paul trap [22]) or a combination of electric and magnetic fields (Penning trap [23]). In a Paul trap, the trapping potential rotates at the frequency of the applied ac field and in a Penning trap, the ion rotates at the cyclotron frequency, as a result the ion feels a time averaged potential minima in all 3D. We plan to use a Paul trap since any external magnetic field cause systematic shift. This requires a dc field U and an ac V Cos ω t to alter the local minima of the rf

potential Ø(x,y,z,t) along the radial and axial directions [24], which trap the ion in 3D if ω is faster than the rf

motion of the ion. At any position r the ion of mass m experiences a force . To produce a restoring force that increases linearly in all directions from the trap centre, a harmonic potential is required. A suitable linear combination of the ac and dc fields V (t) = T

U+V Cos ω t, a can produce a nearly pure harmonic rf

potential as

.. (2)

where R is the radius of trap volume. However, in reality, a pure harmonic potential is nearly impossible to produce in an experiment. Perturbations from a pure quadrupole potential can be estimated from multipoles of higher orders and for an axially symmetric trap, quadrupole, dodecapole, sedecapole and so on contribute. Any higher order potential couples the ionic motion along different axes thereby leading to complicated nonlinear resonances in addition to deviation from linearly increasing force from the trap centre. Thus, it is critical to have a nearly harmonic potential for any precision experiment. Trajectory of the trapped ion along the axial (z) and radial (x, y) directions depend on the stability parameters, a = -2a and q = 2q , which depends on the z x,y z, x,y

dc and ac voltages respectively. The trapping of ions become stable when magnitudes of the stability parameters are much less than 1. Figure 3a shows phase diagram between the ac and dc stability parameters for the radial and the axial directions. Their overlapped region signifies radial and axial stability.

We are using axially symmetric Paul trap of end cap geometry [25].The dominant perturbation in the trapping potential comes from the octupole term, although, other higher orders also contribute in the systematic effects for our experiment. We have performed detailed analysis of the trapping potential for various trapping geometry up to

th10 order multipole. We first performed numerical calculations to find out spatial dependence of the potentials at different geometry of the trapping

before probing the clock transition, requires a 2.1 GHz blue detuned light from the cooling laser. Finally, the laser cooled ion is interrogated using ~Hz linewidth laser at 467 nm to drive the E3-transition. Figure 2b shows schematic of the trapping of ion and its intersection with all five lasers as described above. The ion is spatially confined using end cap electrodes of cylindrical symmetry. The minimum of the trapping potential is accurately tuned using the compensation electrodes E ,..E so that it overlaps with all the lasers. The 1 4

fluorescence of the ion at wavelength 369.5 nm will be collected using a high numerical aperture lens.

Fig. 3: (Colour Online) (a) Phase diagram showing stability regions in axial (blue) and radial (green) directions. The region with simultaneous stability along all directions is shown in inset. (b) Trap potentials: numerical estimation by CPO software (green), fitted harmonic potential (red) and anharmonic potential (yellow). The difference between harmonic and harmonic potentials (dashed red).(c) Three dimensional electric field pattern generated due to the trap potential.

17


Systematic Shifts

The accuracy of a frequency standard is decided by errors in the measured atomic transition frequency due to the systematic effects and statistical uncertainty. The systematic effects in a trapped ion frequency standard arise due to interaction of the ion with external electric and magnetic fields and due to the environmental parameters such as temperature. Therefore, it is important to determine these systematic shifts precisely in order to improve the accuracy of the realized frequency standard. We have estimated some of the systematic shifts

171 +for both the E2 and E3-transitions of Yb [22]. In this section, we briefly mention these shifts and also discuss possible remedies.

1. Electric quadrupole shift of the energy levels is the dominating source of all the systematic uncertainties. It arises due to interaction of the atomic quadrupole moment with an external electric field gradient. The

2 171 + ground states | S ; F=0> of Yb is spherically 1/2

symmetric, thus, it acquires zero quadrupole moment, 2 2however, the excited states | D ; F=1> and | F ; 3/2 7/2

F=3>have a non-zero electric quadrupole moment. The harmonic component of the trapping potential gives a constant electric field gradient; however, a spatial dependence comes from the anharmonic part. The estimated electric quadrupole shifts are 26.25 Hz and -0.5 Hz for the E2 and E3 transitions, respectively. This shift averages out to zero while probing the clock transition along three mutually orthogonal directions. Thus we have designed the UHV chamber accordingly which will allow us to quantize the ion along three mutually orthogonal directions and interact with it by the clock laser along those directions. These shifts, which are three orders of magnitude smaller due to the electric field gradient, arise due to the anharmonic potential, which adds uncertainty in the cancellation of the quadrupole shift. Hence identifying suitable trap geometry with minimum anharmonicity is essential for a precision measurement.

2. Doppler Shift arises due to the relative motion between the photons and the ionic frames of reference. The effect of first order Doppler shift gets cancelled for laser cooled ion at mK temperature. However, the second order Doppler shift is still significant for our application. Under the influence of this confining potential, the trapped ion executes fast excursions at driving radio frequency, called micromotion that are superposed over slow harmonic oscillations called secular motion. While laser cooling greatly reduces the secular motion, the micromotion of the ion contributes significantly to the second order Doppler shift. Other than the temperature dependence of the micromotion, ions acquire excess of it due to a true non-zero phase difference between the ac

electrodes. For this, we used commercial software: charge particle optics (CPO) which uses boundary element method. Then nature of the potentials are

thcharacterized by fitting them up to 10 order and then we find trajectory of the ions corresponding to those potentials which then helps in estimating systematic uncertainties [26]. Figure 3b shows potentials of our trap along the axial direction. Figure 3c shows calculated 3D electric field pattern at a particular instant of time due to the trapping potential and direction of the field lines alters in next half cycle of the ac. Since the ac is at very high frequency than the oscillation frequency of the ion, they feel a time averaged quadrupole field at all times and along all the directions. This exercise helps us to identify the suitable geometry of the ion trap electrodes as shown in Fig. 4. The trap consists of a pair of coaxial electrodes facing each other. The inner electrode is made of tantalum rod of diameter 1 mm and trapping face is machined at

o10 . The outer electrode is made of tantalum tube of inner and outer diameters 1.4 mm and 2 mm respectively,

owhich will have its face machined at 45 . Opposite electrodes have tip-to-tip separation 0.6 mm and the ion overlapped with all five lasers will be trapped at the centre. In order to achieve this, precision machining and fine adjustment will be required. The electrodes will be mounted on a trap holder made out of molybdenum and macor spacer will be used for electrical isolation between the coaxial electrodes. The trap holder will be mounted on a 8-pin UHV compatible electrical feedthrough to mount the ion trap inside the vacuum chamber.

Fig. 4: (left) CSIR-NPL trap mounted on an electrical feedthrough. (right) Design parameter of the electrodes.

18


jittering and mechanical noise due to their soft landing.

We have standardized a dc power distributor which have eight outputs through XLR connectors, where each of these outputs contain ± 24 V and + 5 V. The module contains a dc power supply as well as they are connected to uninterrupted power supply (UPS), since our experiments are designed to operate over long duration without having any interruption.

We intend to monitor and log laser power at different stages in our huge optical setup. For that we have designed photodiode amplifier circuit and readout system, which uses Ethernet connectors and network cables for biasing the photodiodes as well as for acquiring signal and displaying it.

As mentioned in Sec.2.1, this experiment uses five lasers at different wavelengths. We are using extended cavity diode lasers (ECDL) which have tendency to drift in their lasing frequencies owing to environmental fluctuations. Hence, laser frequency stabilization systems are crucial to our experiments. Although, the hardware electronics typically needed in a laser frequency locking system, are commercially available. We have taken an effort of developing an indigenous, cost effective, portable and most importantly all digital laser frequency stabilization system using field programmable gate array technology (FPGA). Use of FPGA allows an entire frequency stabilization unit to be realized inside a small silicon chip

2of few mm area. We have used VHDL to implement the behavioural modules that mimic the laser frequency locking electronics and have succeeded in demonstrating the frequency locking.

Conclusion

We are building the first optical frequency standards 2 2using the | S ; F=0, m =0> - | F ; F=3, m =0> ultra 1/2 F 7/2 F

narrow transition of ytterbium-ion. The clock will operate on a single ion which will be trapped in a Paul trap and laser cooled to mK temperature. The experimental setup consists of an elaborate laser & optics assembly, ion trap in an ultra high vacuum, high resolution imaging for detection of tiny fluorescence, sub-Hz linewidth highly stabilized clock laser, data acquisition and frequency comb for using it as a frequency standard. As of now, we have designed the ion trap, the ultra high vacuum chamber and optics associated laser cooling. We have already fabricated some cub-components of the experiment such as electromagnetically shielded helical resonator of high quality factor that would inductively couple the radio frequency source to the trapping electrodes, various electronic modules and atomic oven for producing nearly collimated ytterbium atomic beam.

voltages applied at the two electrodes and also due to the stray electric fields that ion experiences. Our analysis estimates that the ac phase difference and the stray

o electric fields need to be controlled better than 0.5 and 20 mV/ mm, respectively in order to build a frequency

-17.standard with fractional accuracy ~10

3.DC Stark Effect arises due to the interaction of the induced electric dipole moment (EDM) of an atom with an external electric field. The sources of the external electric fields are the stray fields and the electromagnetic radiation at finite surrounding temperature. Over the time of operating the ytterbium oven, atoms sprayed on the electrodes and the differential work function between the ytterbium and the electrode metal results to a patch potential. The electric field due to black body radiation (BBR) at the finite temperatures of the apparatus also produces a dc electric field. At the room temperature the BBR shifts amounts to be 0.36 Hz and 0.068 Hz for the E2 and E3-transitions respectively.

4. Zeeman Shift arises due to the interaction of atomic and nuclear magnetic moments with an external magnetic field. In an experiment stray magnetic field and the Earth's magnetic field are the major sources of this shift. The E2 and E3-clock transitions are insensitive to the first order Zeeman effect since ground and excited states have m = 0. In New Delhi the Earth's magnetic field will F

produce 52.40 Hz and 5.4 Hz shifts for the E2 and E3-transitions respectively. This indicates magnetic field compensation coils will be essential for cancelling magnetic fields in all three directions. Also the entire experiment needs to be built with non-magnetic elements together with magnet shielding.

Instrumentation

We indigenize lots of instrumentation in the field of machine design, vacuum technology, lasers & optics, computer networking & data acquisition and electronics in order to build this state-of-the-art experiment. In this section, we give a glimpse of some in-house developed electronic modules.

Blocking and unblocking of laser lights synchronized to an experimental sequence is an important requirement in almost any of the laser based experiments. We have developed low cost mechanical shutters modified from old and unused electromagnetic relay, hard disc drive (HDD) and dc motor. In addition we have also designed a universal driver for low noise operation of these shutters. The unique feature of the driver is that it puts out high, medium and low currents using the pulse width modulation technique, which are used for overcoming the inertia, fast sweeping and soft landing of the shutters, respectively. Thus all the shutters operate at response times less than 0.5 ms but also produces negligible

19


12. P. Dubé, et. al, Phys. Rev. A 87, 023806 (2013).

13. Gao KeLin, Chinese Science Bulletin 58, 853 (2013).

14. Y. H. Wanga, et. al, Optics Comm. 273, 526 (2007).

15. C. W. Chou, et. al, Phys. Rev. Lett. 104, 070802 (2010).

16. N. Hinkley, et. al, Science 341, 1215 (2013).

17. Chr. Tamm, et. al, Phys. Rev. A 80, 043403 (2009).

18. J. J. McFerran, et. al, Phys. Rev. Lett. 108 183004 (2012).

19. S. De, et. al, Current Science 106, 1348 (2014).

20. V. A. Dzuba, et. al, Phys. Rev. A 77, 012515 (2008).

21. R. H. Dicke, Physs Rev. A 89, 472 (1953).

22. W. Paul, et. al, Verkehrminist Nordrhein-Westfalen 415 (1958).

23. F. M. Penning, Physica 3, 873 (1936).

24. F. G. Major, et. al, Charged Particle Traps, Springer Series on Atomic, Optical and Plasma Physics (2010).

25. C. A. Schrama, et. al, Opt. Comm. 101, 32 (1993).

26. N. Batra, et. al, manuscript communicated, arXiv:1405.5399 (2014).

Acknowledgement

We acknowledge funding from CSIR-NPL for building the single trapped ion optical frequency standard (STIOS) experiment and also funding from DAE-BRNS for developing the FPGA baser laser frequency locking system.

References

1. J. K. Webb, et. al, Phys. Rev. Lett. 82, 884, (1999).

2. S. Blatt, et. al, Phys. Rev. Lett. 100, 140801 (2008).

3. C. W. Chou, et. al, Science 329, 1630 (2010).

4. J. Gueena, et. al, IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control 59, 391 (2012).

5. P. Arora, et. al, IEEE Transactions on Instrumentation and Measurement 62, 2037, (2013).

6. F. G. Major, Springer-Verlag, New York (1998).

7. Resolution 1 of the 13-th Conference Generale des PoidsetMesures (CGPM) (1967).

8. J. C. Bergquist, et. al, Physics Today (2001).

9. S. R. Jefferts, et. al, Japanese Journal of Applied Physics 43, 2803 (2004) .

10. T. Rosenband, et. al, Science 319, 1808 (2008).

11. N. Huntemann, et. al, Phys. Rev. Lett. 108, 090801 (2012).

20


industrial material processing, aerospace, biomedical, defense and other branches of science and technology due to their high electrical to optical conversion efficiencies, low thermal load, reliable fiber geometry and ability to provide high gains. Further, since these systems generally employ external cascaded fiber amplifier chains within Master Oscillator Power Amplifier (MOPA) configuration, the performance of the seed laser usually remain unaffected by the power scaling process. This provides the flexibility to configure the amplifier system for wide range of applications requiring narrow line width high power CW beam or high average power pulsed beam with pulse duration in the femtosecond to microsecond regime, by employing appropriate seed laser source. Hence development of appropriately-conditioned (e.g. narrow linewidth or accurately pulsed) seed laser is required for the successful development of the amplifier system for a specific application.

We have recently initiated activities on the development of fiber oscillator and amplifier systems and developed novel Yb-doped fiber oscillators (YDFL) with output in diverse temporal formats suitable for seeding the power amplifier. This includes: broadly tunable narrow line-width CW all-fiber YDFL based on multimode interference filter, passively Q-switched YDFL based on fiber optic ring resonator producing pulses in microsecond to 100s of nanosecond range, mode-locked YDFL in all-normal-dispersion setup to generate clean ultrashort pulses in the pico-second and femtosecond regime and mode-locked YDFL in ultra-long figure8 cavity generating shaped pulses such as flat-top pulses with tunability in pulse duration in the range of 3ns- 100ns, step-like long pulses, bound-pulses and burst-mode pulses with adjustable number of pulses in a burst. In this article the physical basis and state of the art of building these lasers with amplification characteristics of some of the lasers are discussed.

Narrow Linewidth Widely Tunable CW Yb-doped Fiber Laser

High power, narrow line-width, widely tunable laser sources ytterbium (Yb) doped fiber lasers and amplifiers

Abstract

In this article novel ytterbium (Yb) doped fiber oscillators with output in four different temporal formats such as continuous wave (cw), Q-switched, modelocked and shaped pulses for seeding of multistage power amplifier are described. The all-fiber cw laser incorporates a narrow band transmission filter based on multi-mode interference effect and produces narrow linewidth ( ~0.05nm) output with wide tunability ( 1038 nm -1070 nm). Q-switching in Yb-fiber laser is achieved by a novel saturable absorber based on fiber-optic ring resonator in combination with nonlinear polarization rotation. In the Q-switched mode of operation output pulse duration can be varied in the range of 1μs-167 ns by changing the pump power with corresponding repetition rate in the range of 45kHz-84kHz. Modelocking Yb-doped fiber laser is done in all-normal-dispersion configuration to produce a train of ultrashort pulses at 37 MHz repetition rate. The pulses are extracted after the nonlinear polarization rejection port to obtain clean temporal profile. The FWHM pulse duration at the output port of the laser was measured to be 5ps which are compressed to ~150 fs duration by removing the chirping in the pulses with the help of a grating pair in near littrow configuration. Shaped pulses are generated by mode locking of Yb-doped fiber laser in ultra-long figure-8 cavity configuration. The interplay of gain, dispersion, self phase modulation, nonlinear loop mirror and nonlinear polarization rotation generates a diverse mode-locking states producing pulses with diverse structures such as flat-top pulses with tunability in pulse duration in the range of 3ns- 100ns, step-like long pulses, bound-pulses and burst-mode pulses with adjustable number of pulses in a burst. The physical basis and state of the art of building these lasers with amplification characteristics of some of the lasers are discussed.

Keywords: Fiber laser, Yb-doped, cw, Q-switched, modelocked

Introduction

In recent times ytterbium (Yb)-doped fiber lasers and amplifiers are being increasingly recognized as the preferred high-power sources for many applications like

Development of Novel Ytterbium Doped Fiber Oscillators with Output in Diverse Temporal Format

for Seeding of Multi-Stage Power Amplifier*P.K. Mukhopadhyay , P.K. Gupta, C.P. Singh,

A.J. Singh, S.K. Sharma, K.S. Bindra and S.M. OakSolid State Laser Division, RRCAT, Indore

*E-mail: [email protected]

21


Design of the MMI filter

The basic construction of MMI filter and its mounting are shown in Fig.(1b). The MMI filter is formed simply by splicing single mode fibers on either ends of a multimode fiber of length L. As the light from the input SMF enters the MMF segment it will excite several higher order modes in the MMF which interfere and generate images of the SMF core at several equidistant locations inside the MMF as shown by the bright dots in the middle panel of Fig.1(b). The exact location of these self-images depends on the wavelength of the input beam. If the second SMF is spliced precisely at such one of the self-image location pertaining to a particular wavelength then coupling loss of the beam to the second SMF for this wavelength will be minimum whereas all other neighboring wavelengths will suffer high loss. The principle of working of MMI filter is explained by several researchers [8-10], hence, here we focus on the design of the MMI filter for single wavelength operation of Yb-doped fiber laser with narrow spectral width and wide wavelength tunability. Through numerical simulation of the MMI filter we find the optimum length of the MMF which can support

particularly in all-fiber format are required for many applications such as high resolution spectroscopy, sensing, laser cooling, metrology and biomedical technology etc. To date, various methods have been proposed to achieve wavelength tuning in Yb-doped fiber laser, including using specialized band-pass tuning filter [1], diffraction grating pair [2], passive multiple-ring cavity [3], array of fiber Bragg gratings [4], Mach-Zehnder interferometer [5] and recently using an acousto-optic tunable filter [6]. However, many of the above systems employ expensive specially made fiber-optic components or use bulk optical components deviating from the all-fiber format. In recent times band-pass optical filter based on single mode-multimode-single mode (SMS) fiber structure [7] has generated a considerable interest for wavelength tuning in fiber laser due to its many advantages such as low cost, ease of fabrication and compatibility for all-fiber integration. Such filter works on the basis of multimode interference (MMI) effects in the multimode fiber (MMF) segment and its peak transmission wavelength can be tuned by controlling external parameters like temperature or strain of the MMF segment. In this work, we developed an all-fiber Yb-doped laser producing cw output with 0.05 nm of spectral width and more than 30 nm of wavelength tunability using an MMI filter and demonstrated a simple wavelength tuning mechanism with the help of a standard polarization controller.

The Laser Setup

The schematic of the all-fiber laser setup is shown in Fig1a. The pump source (LD) was a fiber Bragg grating (FBG) stabilized single mode fiber (SMF) coupled laser diode with a maximum CW output power of 450 mW at 975nm (Lumics, Model LU0975M450). The pigtailed fiber end of the laser diode after the FBG was connected to a 80 cm long single clad Yb-doped fiber with mode field diameter (MFD) of 6.0 μm at 1060 nm (YB501, CorActive) through a 975/1060 nm WDM coupler. The other end of the Yb-doped fiber is fusion spliced to a 90/10 fiber coupler. A polarization insensitive SMF based fiber isolator (Opto-link Corporation) is spliced to the 90% power transmission port of the coupler for unidirectional ring cavity operation. All these fiber-optic components used in the system are based on standard SMF(Corning HI 1060, MFD 6.2 μm at 1060 nm). A segment of step-index multimode fiber (MMF) with core and cladding refractive index of 1.4504 and 1.4271 respectively is placed between the output port of the isolator and signal port of the WDM to complete the unidirectional ring cavity. Since, both ends of the MMF is spliced to SMFs, this SMS combination (MMIF in Fig.1a) acts like a bandpass filter and influences the spectral characteristic of the laser.

Fig. 1: (a) Schematic of the laser setup. WDM: wavelength division multiplexer, PC: in-fiber polarization controller, LD: laser diode, MMIF: multimode interference filter (b) construction of MMI filter (top), multimode interference effect in the MMF segment (middle) and tuning mechanism of MMI filter using polarization controller (bottom)

22


order to p = 8, corresponding to ~67 nm as can be int

seen from Fig. 2A (a). Once the self-imaging order is fixed, the core diameter of the MMF dictates the transmission bandwidth of the MMI filter. The transmission bandwidth (Δλ ) of the MMI filter is FWHM

estimated at the full width at half maximum of the transmission curve obtained from the Eq.(1) at λ=1070 nm. The computed variation of Δλ as a function of the FWHM

MMF core diameter at p = 8 is shown in Fig. 2A (b). It can be seen that Δλ decreases monotonically with the FWHM

core diameter of the MMF. However, since the HI 1060

Δλ

narrow transmission bandwidth 'Δλ ' (full width at FWHM

half maximum) and wide wavelength separation between the successive transmission peaks of the filter known as the self-imaging wavelength interval 'Δλ ' [10]. int

Assuming an ideal alignment in the SMS structure the wavelength response of the MMI filter can be obtained as [8]

(1)

where T is the transmission of the MMI filter for the wavelength λ, m and M are the index and total number of the radial modes excited in the MMF by the input SMF,

thc is the field excitation coefficient for the m order mode m

[9] and β is the longitudinal propagation constant for this m

mode. In Eq.(1) L is the length of the MMF segment chosen such that L = p Z , where p (p =1, 2, 3…) is the im

order of the self-images of the input field formed by the interference of the guided modes in the MMF and Z is im

the re-imaging distance. The re-imaging distance can be obtained from the diameter (d) and refractive index (n ) core

2of the core of the MMF as [8,9] Z = 4 x n x d / λ , im core 0

where λ is the design wavelength. It is to be noted that the 02ratio L/d is a constant for a given λ and corresponds to a 0

particular order of the self-image. Hence a longer length of the MMF is required for a larger core diameter in order to create a particular order of the self-image at the end of the MMF.

Eq.(1) is solved numerically for λ =1070 nm to find the 0

wavelength response of the transmission of the MMI filter from which Δλ and Δλ are estimated for int FWHM

different values of p and d following the method introduced in Ref. 8 and 9. The simulation results are summarized in Fig.2A. The points in Fig. 2A (a) show the computed variation of Δλ as a function of the self int

imaging order p. It was found through the numerical simulation that Δλ depends only on the design int

wavelength and the order of the self-image and can be expressed as Δλ = λ /2p (dashed line in Fig. 2A (a)). int 0

Thus Fig. 2A (a) provides a guideline to the maximum tolerable self-imaging order for which single wavelength operation can be obtained irrespective of the core diameter and the numerical aperture of the MMF. For 80 cm long Yb-doped single-mode fiber the gain peak usually occurs at ~1030 nm under the pumping wavelength at 975 nm but the gain spectrum spans over a wide range, from ~1010 nm to ~1120 nm. Hence, for the design wavelength of 1070 nm for the MMI filter, Δλ int

should be more than 60 nm to ensure single wavelength operation. This requirement restricts the self-imaging

Fig. 2: A Computed variation of (a) self-image wavelength interval as a function of the order of the self-image, (b) transmission bandwidth of the MMI filter as function of the MMF core diameter for self-imaging order of 8. (c) Wavelength response of MMI filter with d = 100 μm at p = 8 and (d) expanded view of the transmission peak at 1070 nm. B. Recorded spectra (a) without any MMI filter, (b) with MMI filter using 43.4 cm long MMF with 100 μm core diameter, inset: spectra in dB scale. (c) Expanded view of the recorded spectra at 1070 nm.(d) recorded spectra with 65 cm long MMF with 100 μm core diameter.

23


was measured at the maximum pump power. It can be seen from Fig.2B (a) that the output spectral profile without the MMF segment was very broad ranging from 1045 nm to 1064 nm with distinct peaks at 1048 nm, 1057 nm and at 1061 nm. However, with the MMF segment the laser operates with single spectral peak at ~ 1070 nm with narrow spectral width as can be seen from Fig. 2B(b). The inset of Fig. 2B(b) shows the spectrum in dB scale confirming that there is no other appreciable peak present in the spectrum. The maximum output power was measured to be ~80 mW with the MMI filter. Fig. 2B(c) shows the expanded view of the spectrum from which the FWHM spectral width was measured to be 0.05 nm. We also have operated the laser with a longer segment (~65cm) of MMF in the MMI filter corresponding to a self-imaging wavelength interval of ~44nm which is much shorter than the width of the gain spectrum of Yb-doped fiber. With this long segment of MMF the laser operates at two distinct wavelengths as can be seen from the recorded spectra shown in Fig. 2B (d).

Initially the actuator knob of the polarization controller 0was kept vertically upward direction (referenced as the 0

position) and was turned towards one direction for tuning the output wavelength. As the knob is turned, the MMF is mechanically strained leading to a change in the propagation constant of the guided modes which causes the shift in output wavelength. In Fig. 3(a) we show the superimposed spectral profiles of the laser obtained for the different turning angles of the actuator knob from its vertical position. The shortest wavelength obtained was

01038 nm with the maximum turning angle of 120 and the longest wavelength was 1070 nm at the vertical position of the PC knob. Fig. 3(b) shows the tuning curve of the laser as a function of turning angle of the PC knob. It can be seen that output wavelength decreases linearly with the turning angle. A tuning range of more than 30 nm was obtained by this technique. However the knob could not

0be turned more than 120 due to the mechanical limitation of the polarization controller.

SMF fiber has a cladding diameter of 125 μm, the cladding diameter for the MMF also should be close to 125 μm for successful and low loss fusion splicing of the SMFs to the MMF which restricts the allowable core diameter of the MMF < 105 μm. Hence, the core diameter of the MMF segment was chosen to be 100 μm which is easily available commercially. The computed wavelength response of the MMI filter with 100 μm MMF at p = 8 is shown in Fig 2A (c) which shows the presence of a single transmission (100%) peak over the entire gain spectrum of the Yb-doped fiber and some other transmission peaks of much lower heights resulting from the imperfect interference of the modes in the MMF. The expanded view of the transmission peak is shown in Fig. 2A (d). It can be seen that the peak transmission is centered at 1070 nm with a transmission bandwidth of 0.4 nm. The required length of the MMF segment is estimated to be 43.376 cm (p =8, Z = 54.22 mm). im

However, precise cleaving at this length may not be necessary as the filter is placed inside the laser and the lasing wavelength will be adjusted accordingly to the peak transmission wavelength of the filter.

For tuning of the peak transmission wavelength, the optical length of the MMF needs to be modified in a controlled way around its design length. In our setup, the wavelength tuning was achieved simply by applying a localized stress in the MMF. For that purpose the MMF segment was held straight by clamping its ends on metal blocks for wavelength stability and an in-fiber polarization controller (PLC-003 PolaRiteTM) is placed near the middle of the MMF as shown in the bottom panel of Fig.1b. The stress on the MMF was applied locally by slight tightening of the actuator knob of the polarization controller and then turning the actuator block in the sideway direction. The stress in the fiber leads to modification of the indicatrix of the MMF which in turn leads to the change of the propagation constant of the fiber modes. Since the propagation constant depends on the wavelength, the stress on the MMF lead to a shift of the wavelength enabling the tuning of the peak transmission wavelength of the MMI filter. This method is proved to be an efficient and simple technique for wide tunability of the Yb-doped fiber laser using MMI filter as demonstrated in the experiment.

Performance of the Laser

The output from the laser is taken through the 10% port of the fiber coupler and the spectral profile of the laser output was characterized with the help of an optical spectrum analyzer (Agilent, Model no. 86124B). The spectral characteristics of the laser are summarized in Fig.2B. First we operated the laser without the MMI filter by removing the MMF segment. The laser had a threshold of ~100 mW and a maximum output power of ~100 mW

Fig. 3: (a) Superimposed spectral response of the tunable MMI Yb-doped fiber laser at various angle position of the actuator knob of the polarization controller, (b) tuning curve of the laser as a function of the turning angle of the actuator knob of the polarization controller.

24


(WDM1) is connected between the LD and the WDM2 for the protection of the laser diode from damage due to the amplified spontaneous emission (ASE). The maximum pump power delivered in the core of the Yb-doped fiber was measured to be ~330 mW. One end of the Yb-fiber was spliced to the output port of the WDM and at the other end, a standard SMF (HI1060, 105 cm long) was connected. At the signal port of WDM a 300 cm long single mode fiber (SMF) was spliced. The free ends of the two SMFs are connected to in-fiber collimators (COL1 and COL2). The total cavity length including the free space between the collimators was ~570 cm. A polarizing beam splitter (PBS) is placed near COL1. The PBS1 in combination with the two infiber polarization controllers (PC1 and PC2) attached to the SMFs defines the polarization state of the laser which can be varied arbitrarily by varying the pressure or torsion in the polarization controllers. A fraction of the circulating power is coupled out from the cavity at PBS as the output. A bulk optical isolator (ISO) was placed in the free space for unidirectional ring cavity operation. The fiber optic ring resonator (FORR) is placed after the PC2. The ring interferometer is made from a 90:10 single window coupler (SWC) by connecting the 10% output port to the auxiliary port in the input side. The FORR in combination with the PBS and polarization controllers act like a saturable absorber with delayed response which prevent the mode-locking operation but can lead to Q-switching operation due to its nonlinear transmission behavior. The Q-switching operation was initiated by adjusting the polarization controllers and the output pulses from the PBS were detected with the help of a fast photodiode (rise

Passive Q-Switching of YDFL Based on Fiber Optic Ring Resonator (FORR)

Passive Q-switching of ytterbium (Yb) doped fiber laser using in-fiber saturable absorber is of great practical importance as it offers the possibility for all-fiber format for the laser setup as compared to the traditional approach of active Q-switching using acousto-optic modulator [11] or passive Q-switching using bulk Cr4+:YAG crystal [12]. In the recent past passive Q-switching operation in ytterbium (Yb) doped fiber laser using in-fiber components has been demonstrated in a number of ways like by using chromium or samarium doped fiber [13,14], GT wave fiber [15], holey fiber [16] etc. However all of these systems employed a speciality fiber-optic component which not only make the system complex and expensive but also impose stringent restriction on the operating range of the pump power beyond which the laser exhibits instabilities and often splitting in the Q-switched pulses. On the other hand stimulated Brillouin scattering (SBS) induced self pulsing operation in Yb-doped fiber laser can be obtained in a very simple configuration, however, the laser power should be sufficiently high to initiate the SBS process [17]. Further the self Q-switched pulses arising due to SBS are often irregular and accompany with satellite pulses and significant intra-pulse cw signal. Hence there is a need for self-pulsing technique in Yb-doped fiber laser in a simple configuration which can be easily integrated with all-fiber components as well as capable of producing stable Q-switched pulses for a wide range of operating parameters.

In this work, we demonstrate stable and sustainable self Q-switching operation in Yb-doped fiber laser with the help of a saturable absorber comprising of a polarizing beam splitter (PBS), polarization controller and a fiber optic ring resonator. By adjusting the polarization controllers stable Q-switching operation is readily initiated which remained stable throughout the operating range of the pump power and no significant modulation in the Q-switched pulses due to self mode-locking is observed. At the maximum pump power of ~300mW coupled to the gain fiber stable Q-switched pulse train of ~80 kHz repetition rate with 167 ns of pulse duration was obtained. The average power was measured to be ~130 mW corresponding to 1.6 μJ of pulse energy.

Experimental Setup and Results

The schematic of the laser setup is shown in Fig.4(a). The laser comprised of 70 cm long single clad single mode Yb-doped fiber with mode field diameter of 6.0 μm. It was pumped in-core by a FBG stabilized single mode fiber coupled laser diode (LD) at 976 nm with the help of a 980/1060 WDM combiner (WDM2). Another WDM

Fig. 4: (a) Schematic of the laser setup for passive Q-switching of YDFL using FORR based saturable absorber. (b) recorded Q-switched pulse train with repetition rate 75 kHz

25


fall time of 300 ns and 700 ns respectively. At the maximum pump power the average output power was measured to be ~130 mW corresponding to 1.6 μJ of pulse energy.

Generation of Clean Femtosecond Pulse from All-Normal-Dispersion Mode-locked Yb-doped Fiber Laser

Ytterbium (Yb) doped mode-locked fiber laser in all normal dispersion ( ANDi) configuration has attracted a great deal of current interest due to its simplicity in construction and configurability with all-fiber integration. Further, the shape of the mode-locked pulses from the ANDi laser is dissipative soliton type which can tolerate a large variation of gain and loss and hence highly suitable for energy scaling in external amplifier. Since its first demonstration in 2006 [18], considerable progress has been made during the recent past to understand the pulse shaping dynamics in ANDi laser as well as to improve its performance [18-22]. However, the pulses from most of these systems exhibit large amount of side-lobes when they are compressed to femtosecond duration and the pulse quality degrades further on amplification. One of the reasons for poor pulse quality from the ANDi laser is that in most of the systems the output is taken either before or at the nonlinear polarization rejection port (NPR) in order to extract maximum pulse energy, though, the influence of nonlinearity like self phase modulation (SPM) on the spectra of the pulses is very strong at those locations. It has been indicated in [22] that by implementing an output coupler just after the NPR it is possible to obtain a considerably cleaner pulse, however, no attention has been paid to maximize the pulse energy or on its amplification characteristics from this port.

In this work, we present the results of our experimental studies on the spectral, temporal and amplification characteristics of the pulses taken from an output port with a variable coupling ratio implemented just after the NPR in an ANDi laser. By adjusting the out-coupling ratio more than 3nJ of pulse energy is obtained from this port which is comparable or higher to that achievable from the NPR under the same pumping power. The pulses are compressed to ~160fs duration with a clean temporal profile without any side-lobes. Further, the pulse energy was scaled-up to ~7 nJ in an amplifier segment without any significant degradation of the pulse quality.


The schematic of the mode-locked Yb-doped fiber laser setup under all-normal dispersion configuration is shown in Fig.6. The laser comprised of 70 cm long single clad single mode Yb-doped fiber with mode field diameter of 6.0 μm. It was pumped in-core by a FBG stabilized single mode fiber coupled laser diode (LD) at 976 nm with the

time 1 ns) and displayed on an oscilloscope. We also monitored the optical spectrum of the pulses with the help of a wavelength meter.

As the pump power exceeds the threshold of ~150 mW, stable train of passively Q-switched pulses are readily observed. Fig 6(b) shows the recorded pulse train from the laser. It can be seen that the pulse train are highly stable with less than 5% amplitude and timing jitter. The corresponding spectral peak is located at 1030 nm with less than 1nm spectral width (FWHM). The variation of the pulse repetition rate and the pulse duration as a function of the pump power is shown in Fig.5(a). It can be seen that the repetition rate is ~40 kHz near the threshold and correspondingly the pulse duration was ~ 1.2 μs. However as the pump power is increased the repetition rate increases and the pulse duration decreases almost linearly with the pump power. At the maximum pump power of 330 mW the measured repetition rate was ~80kHz and correspondingly pulse duration was measured to be 167 ns. Fig.5(b) shows the recorded pulse shape at the maximum pump power. It can be seen that the pulse shape is very smooth without any modulation in the profile resulting from the self modelocking. However, the pulse shape is quite asymmetric with a measured rise and

Fig. 5: (a)Variation of pulse width and repetition rate as a function of pump power from the laser.(b) Recorded pulse shape of the passively Q-switched laser

26


based on NPR becomes effective and the mode-locking becomes self-starting.

The laser cavity was made with all-normal-dispersion 2elements (net GVD ~0.12 ps ) and for the purpose of

pulse width management a narrow band interference filter (BPF) with 10 nm bandwidth and peak transmission at 1060 nm is placed after the Isolator. When a highly chirped pulse passes through an appropriately chosen narrowband transmission filter it leads to shortening of the pulse by cutting its wings as the spectral components outside the transmission band of the filter are placed at the wings of a heavily chirped pulse. Thus the narrowband interference filter not only helps to control the pulse width but also leads to strong self amplitude modulation by transmitting the peak of the pulse and cutting its wings.

The HWP was first oriented for minimum coupling loss through the OP port with pump power set to its maximum value. The mode-locking operation was then easily achieved by adjusting the PC1 and PC2. At this condition the HWP was gradually rotated until the power coupled out through the OP port reaches to maximum without losing the mode-locking. Fig. 7 (a) shows the variation of the output power from the NPR (squares) and OP port (circles) as a function of the pump power with the above mentioned settings of the HWP, PC1 and PC2. It can be seen from Fig. 7(a) that under the low pumping power the output is continuous wave (CW) but becomes Q-switched mode-locked (QML) as the pump power is increased. In both these regime the power at the NPR is much higher than that at the OP port. However as the pump power is increased beyond 270 mW the laser gets mode-locked and correspondingly the output power at OP port increases by manifold than that at the NPR. Fig. 7(b) and 7(c) show the pulse train recorded in an oscilloscopes in two different time scales ( 20 ns and 10s respectively). From Fig. 7(b) the pulse repetition rate was measured to be ~35 MHz and it can be seen from the long range trace (Fig. 7c) that the amplitude of the mode-locked pulses is highly stable. Nearly 115 mW of average power is obtained at the output port corresponding to a pulse energy of ~3.2 nJ. On the other hand, the maximum pulse energy from the NPR rejection port was measured to be ~ 2.5nJ with minimum coupling loss through the OP port.

help of a 980/1060 WDM combiner. The maximum pump power delivered in the core of the Yb-doped fiber was measured to be ~330 mW. One end of the Yb-fiber was spliced to the output port of the WDM and at the other end, a standard SMF (HI1060, 105 cm long) was connected. At the signal port of WDM a 300 cm long single mode fiber (SMF) was spliced. The free ends of the two SMFs are connected to in-fiber collimators (COL1 and COL2). The total cavity length including the free space between the collimators was ~570 cm. A polarizing beam splitter (PBS1) is placed near COL1. The PBS1 in combination with the two in-fiber polarization controllers ( PC1 and PC2) attached to the SMFs act like a fast saturable absorber based on nonlinear polarization rotation (NPR). A fraction of the circulating power is coupled out from the cavity at PBS1 as NPR rejection. An output coupler (OP) is implemented after the PBS1 with the help of a half wave plate (HWP) and a polarizing beam splitter (PBS2). By adjusting orientation of the C-axis of the HWP with respect to the polarization direction the output coupling ratio can be varied for a wide range. A bulk optical isolator (ISO) was placed in the free space for unidirectional operation so that the saturable absorber

Fig. 6: Schematic of the ANDi fiber-laser oscillator and amplifier (top) and photograph of the oscillator amplifier setup under operation (bottom)

Fig. 7: (a) Output vs. input from NPR and OP port. (b) and (c) Oscilloscope traces of pulse train in short and long range.

27


coupler. Nearly 20 mW of signal power from the OP port, co-propagating with the pump, is coupled to the amplifier with the help of an aspheric lens. Several directional couplers (SWC) are placed at various locations in the amplifier to monitor the spectral profile of the pulses. Fig 9(a) shows the variation of the output power from the amplifier as a function of the input pump power. It can be seen that the output power varies linearly with the pump power with a slope of ~ 45%. At a pump power of 500 mW, around 245 mW of average signal power was obtained. This corresponds to amplified pulse energy of 7nJ. Fig. 9(b) shows the recorded spectral profile at the maximum pump power. The spectral width was measured to be ~21nm and some ripples can be found in the spectra due to the strong SPM at the amplifier, however, the ASE signal at the gain peak at 1030nm is quite negligible as can be seen from the spectra in dB scale in the inset of Fig. 9(b). The intensity autocorrelation trace of the compressed pulses from the amplifier is shown in Fig. 9(c) and the corresponding envelope of the fringes in interferometric AC is shown in the inset. The dotted line in Fig. 9(c) is the Gaussian-fit to the AC trace and the FWHM pulse width was measured to be ~164 fs. No side-lobes are observed in the AC traces, however, a slight hump can be seen at the wings of the pulse which could be due to the uncompensated nonlinear chirping resulting from the SPM in the amplifier.

Generation of Shaped Pulses

High energy shaped pulses like long flat top, step like, bound or burst mode type pulses are required for many applications such as all- optical square wave clocks, precision micro-machining, laser micro-welding, optical sensor and laser ablation etc. Recently, passive mode locking in fiber lasers has attracted a great interest in long-pulse generation. After the first demonstration of nanosecond square-wave pulse generation based on the figure-eight structure (Figure-8) [23] and the nonlinear polarization rotation [24] (NPR) in passive mode locked fiber lasers, the research work in this direction have attracted a large interest in enhancing the tunable range of nanosecond square wave pulse duration [25-30]. In this work, a mode locked fiber laser is designed in an ultra long Figure-8 shaped resonator configuration to produce nanosecond square pulses by the combined action of NPR and the non linear optical loop mirror (NOLM). The

Fig. 8: (a), (b): Recorded spectral profile and (c), (d) auto correlation trace of compressed pulse from the NPR and OP port.

The spectra from the NPR and OP port with the optimum output coupling are shown in Fig. 8(a) and (b) respectively. It can be seen that the spectral shape from the NPR rejection port is highly structured with peaked edges whereas that from the OP port has diminished peaks at the edges and a smooth dome near the center. These are in close agreement with the simulated spectral profiles before and after the PBS1 as shown in Fig. 6. However, the spectral width was measured to be nearly same ( ~17 nm) from both the ports with a central wavelength at 1060 nm.

Since the laser was made of all-normal dispersion components the pulses are chirped and hence compressed externally using a grating pair (600 lines/mm) in near littrow configuration. The separation between the grating pair was adjusted to ~32 cm to obtain the minimum pulse duration and the temporal profiles were characterized by non-collinear intensity autocorrelation. Fig. 8 (c) and (d) show the autocorrelation (AC) trace of the pulses from the NPR rejection and the OP port respectively taken after maximizing the respective power. The pulse duration from the NPR rejection port was measured to be 130fs (FWHM) which was 30fs shorter than that from the OP port, however, it was accompanied with many distinct side lobes with amplitude ranging from 5-40% of the peak of the pulse. On the other hand, the temporal profile from the OP port is very smooth and clean without any side-lobes as can be seen from Fig. 8(d).

The amplifier segment is comprised of a 1m long Yb-doped single mode fiber pumped in-core by a fiber coupled laser diode at 976nm with the help of WDM

Fig. 9: (a): Output vs input power, (b): spectral profile and (c) autocorrelation trace of compressed pulse from the amplifier.

28


bandwidth 10nm) was placed. The signal port of WDM2 was spliced with collimator (COL2) to complete the ring resonator via fiber based polarization insensitive fiber isolator, a single window coupler (SWC1, 90:10) and another polarization controller (PC2). The output ports of the SWC1 were fusion spliced together via a long length of SMF to act as a nonlinear optical loop mirror (NOLM). The length of the SMF used in the NOLM was 200m and 300m separately. The ISO placed between BPF and PBS makes the ring cavity unidirectional which assist in self-staring of the mode locking. This completes the design of fiber ring resonator in figure-8 cavity configuration. The rejected signal at PBS is characterized as an oscillator output.

The mode locking in the above resonator is attributed to the combined effect of nonlinear polarization rotation (NPR) and nonlinear optical loop mirror (NOLM). A long fiber length used in NOLM helps in pulse shaping of the mode locked pulses by the peak power clamping effect in the figure8 cavity and results in the square-shaped nanosecond pulses whose pulse duration is adjustable by changing either the pump power coupled into YDF or by changing the orientation of PC. To obtain the mode-locked pulses the LD pump power was increased and self starting mode locked pulses could be obtained by adjusting the setting of PC's. The threshold pump power for the mode locking was measured to be ~70mW. A typical recorded pulse train at 200mW pump power coupled in YDF is shown in Fig.11a. The repetition rate of the mode locked pulses was 850kHz and 594.8kHz for SMF lengths of 200m and 300m respectively which corresponds to the fundamental repetition rate for the designed resonator length. We also recorded the stability of ML pulses for variation in its peak power and FWHM pulse duration. Over a time period of 60minutes, the variation in peak power and FWHM pulse duration was measured to be 9% and 2% respectively. As the LD pump power was increased, the pulse duration of the mode locked pulses was increased.

Figure 11(b) shows the variation of pulse duration of ML pulses with the pump power coupled into YDF for 200m

width of the flat-top pulses can be varied continuously from 3ns to more than 90ns by changing the pump power (without altering the height of the pulses). Further laser provide a platform for diverse modelocking regime to generate different types of shaped pulses as discussed in this section.


The schematic of the mode-locked (ML) Yb-doped fiber (YDF) oscillator set-up is shown in Fig.10. The laser consists of 70cm long single clad single mode Yb doped fiber (mode field diameter 6mm) pumped in-core by a fiber Bragg grating (FBG) stabilized single mode fiber (SMF) coupled laser diode (LD) with a maximum CW output power of 450mW at 975nm (Lumics, Model LU0975M450) through wavelength division multiplexer (WDM2). Another WDM1 was spliced with LD to protect it from the backward propagating amplified spontaneous emission (ASE). The other end of the Yb-doped fiber was fusion spliced with fiber collimator (COL1) in which a polarization controller (PC) was implemented. After COL1 a polarizing beam splitter (PBS), polarization insensitive isolator (ISO) and a band pass filter (BPF, central wavelength 1060nm, FWHM

Fig. 10: Schematic of passively mode-locked Yb-doped oscillator in ultra-long fig-8 cavity configuration to generate flat-top ns and other shaped pulses (top) and photograph of the laser (bottom)

Fig. 11: (a) Recorded oscilloscope trace of the ML pulse train, (b) Variation of ML pulse duration with pump power coupled into YDF, inset shows pulse shapes at two power levels

29


12(c') and (d'). It can be seen as the number of bound pulses decreases the secondary spectral peak approaches closer to the main peak leading to a broadening of the main spectra. Fig.12(e) shows a burst mode of two pulses obtained at lower pump power. Under this operating regime as the pump power is increased the number of pulses in a burst increases. A typical bust mode containing five pulses is shown in Fig. 12(f). The corresponding spectra under burst mode is found to be broad, however, shifting towards the longer wavelength side with pump power as can be seen from Fig.12(e') and (f'). The modelocking operation was found to be reasonably stable at each of these states. The interplay of gain, dispersion, spectral filtering due to finite gain bandwidth, self phase modulation and cross phase modulation along with the nonlinear polarization rotation and nonlinear loop mirror are believed to generate such diverse modelocking states, however, further studies are necessary in this direction.

Summary

In summary, we have developed fiber laser oscillators in different temporal format for seeding multistage fiber amplifier. The main features of these seed sources are summarized as follows:

(i) An all-fiber broadly tunable narrow linewidth Yb-doped ring fiber laser based on MMI filter is developed. By using ~ 43.4 cm long 100 μm core diameter multimode fiber in the SMS structure single wavelength output with a measured spectral width of 0.05 nm is achieved. The MMI filter transmission wavelength was tuned with the help of an in-fiber polarization controller. With this simple mechanism more than 30 nm ( 1038 nm -1070 nm) of tuning range is demonstrated. The laser is constructed with off-the shelf components and can be easily reproduced. This laser is suitable in developing high power narrow line width tunable CW source in MOPA configuration.

(ii) Stable and sustainable passive Q-switching of Yb-doped fiber laser using saturable absorber based on fiber optic ring resonator is demonstrated. The Q-switched pulse train is highly stable with less than 5% timing and amplitude jitter. The pulse duration from the laser can be varied from 1μs to ~200 ns by changing the pump power in the oscillator. At the maximum pump power of 330 mW, 1.6 μJ pulses with 167 ns duration are obtained.

(iii) We have generated clean 5ps pulses and 160fs pulses after compression from mode-locked Yb-doped fiber laser in all-normal-dispersion configuration by out-coupling the beam after the NPR port. By adjusting the output coupling ratio

of SMF in the NOLM. It can be seen that the laser was mode-locked at a coupled pump power of ~30mW and flat top mode-locked pulse width of ~3ns was obtained near the threshold. However as the coupled pump power is increased the duration of mode-locked pulses increases linearly with coupled pump power. At the maximum coupled pump power of 230 mW around 32 ns of pulse duration was obtained. In the inset of Fig.11(b) we also plot recorded pulse profile under low power (red, left corner of the plot) and at under higher power (blue, right corner of the plot). It can be seen that the pulses are reasonably flat in shape with a sharp rise and relative slow fall time. It is worth to mention here that while pump power is increased the peak of the pulse changes only slightly. The maximum average power which could be extracted out of the oscillator was measured to be ~33mW which corresponds to pulse energy of the ML pulses to be ~56nJ.

In order to obtain longer pulse duration the SMF of the NOLM segment was replaced with a longer length of 300m length. The repetition rate with this longer NOLM was measured to 594.6 kHz. The threshold of the laser was increased slightly and the pulse duration near the threshold of was measured to be ~10ns. In this configuration the maximum pulse duration of ~100ns was obtained at the maximum coupled pump power. It has been observed that by adjusting the polarization controllers and the pump power the laser operates in diverse modelocking states producing a wide variety of shaped pulses. Some of the representative temporal profiles and the corresponding spectral output are shown in Fig.12. Fig. 12 (a) and (a') shows a long (~60ns) flattop pulse and the corresponding spectrum respectively. It can be seen that the spectrum of the long flat top pulse is narrow with 0.3 nm FWHM spectral width. A step like pulse and the corresponding spectrum are shown in Fig.12(b) and (b') respectively. It can be seen that under this condition a smaller spectral peak accompanies the main spectral component. Fig.12(c) and (d) show the bound states of group of pulses and two distinct pulses respectively. The corresponding spectra are shown in Fig.

Fig. 12: Recorded temporal (left) and spectral (right) profile of the pulses obtained from the laser.

30


13. Bernard Dussardier, Jérôme Maria, and Pavel Peterka, Appl. Opt. 50, E20-E23 (2011)

14. A Fotiadi, A. Kurkov, and I. Razdobreev, OSA Technical Digest Series (Optical Society of America, 2007), paper CMC4.

15. S. K. Turitsyn, A. E. Bednyakova, M. P. Fedoruk, A. I. Latkin, A. A. Fotiadi, A. S. Kurkov, and E. Sholokhov, Opt. Express 19, 8394-8405 (2011)

16. G. A. Sanchez, A. M.Rios, I. T.Gomez, R. S. Aguilar, and J.M. E.. Ayala, Rev. Mexicana De Fisica 54, 1 (2008).

17. S. V. Chernikov , Y. Zhu, J. R. Taylor, and V. P. Gapontsev, Opt. Lett. 22, 298 (1997)

18. A Chong , J. Buckley, W. Renninger and F. Wise, Opt. Exp. 14, 10095 (2006).

19. A Chong ,W. H. Renninger and F. W. Wise, Opt. Lett. 32, 2408 (2007).

20. K. Kieu and F. W. Wise, Opt. Exp. 16, 11453 (2008).

21. P. K. Mukhopadhyay , K. Ozgoren, I Budunoglu, and F.Ilday, IEEE J. Sel. Top. Quantum Electron 15,145 (2009)

22. A Chong , W. Renninger and F. Wise, JOSA B 25, 140 (2008).

23. D. J. Richardson , R. I. Laming, D. N. Payne, V. Matsas, and M. W. Phollips, Electron. Lett. 27 , 542 (1991).

24. V. J. Matsas, T. P. Newson, and M. N. Zervas, Opt. Commun. 92 , 61 (1992).

25. L.M.Zhao, D.Y.Tang, T.H. Cheng, C Lu,Opt. Commun. 272 (2007).

26. X. Li , X Lieu, X Hu, L wang, H Lu, Y Wang, Opt. Lett. 35,19 (2010).

27. C G Liang , GUChun, XU Li Xin, W Ting, M Hai, Chin.Phy.Lett. 28, 12 (2011).

28. G Chen , C Gu, L Xu, H Zheng, H Ming, Chin.Opt.Lett. 9, (2011).

29. X. Zhang C. Gu, G. Chen, B. Sun, L. Xu, A. Wang, and H. Ming, Opt. Lett. 37,8 (2012).

30. W. Li, Q. Hao, M. Yan and H. Zeng, Opt. Exp. 17,12 (2009)

more than 3nJ of pulse energy is obtained which are further amplified to 7nJ energy without degrading the pulse quality. The laser can be used for biomedical applications where clean femtosecond pulses of low energy are required or can serve as the excellent seed source for power amplifier.

(iv) Stable mode-locked pulses were generated in ultra-long fig-8 cavity configuration. The pulses are flat-top in shape and the pulse duration can be varied continuously in the range of 3ns to ~100 ns by changing the pump power. The pulse repetition rate is less than 1 MHz with individual pulse energy of 50-100nJ. Diverse modelocking regimes under ultra-long cavity configuration is observed producing differently shaped pulses.

References:

1. A Hideur, T Chartier, C Ozkul and F Sanchez, Opt. Lett. 26, 1054 (2001).

2. M. Auerbach, P. Adel, D. Wandt, C. Fallnich, S. Unger, S. Jetschke, H.-R. Müller, Opt. Exp. 10, 139 (2002).

3. F Yin, S Yang, H Chen, M Chen, S Xie, IEEE Photon. Technol. Lett. 23, 1658 (2011).

4. J.A. Alvarez, A. M. Rios, I. T. Gomez, and H.L. Offerhaus, Laser Phys. Lett. 4, 880 (2007).

5. Y Meng, S Zhang, X Wang, J Du, H Li, Y Hao, X Li Opt. Las. Eng. 50, 303 (2012).

6. R. Royon J. Lhermite, L. Sarger, and E. Cormieri, Opt. Exp. 21, 13818, (2013).

7. W. S. Mohammed, P. W. E. Smith and X. Gu, Opt. Lett. 31, 2547 (2006).

8. Q. Wang et.al. J. Light-wave Technol. 26, 512 (2008).

9. W. S. Mohammed, G. Farrell and W. Yen, J. Light-wave Technol. 22, 469 (2004)

10. X. Zhu, A. Schülzgen, H. Li, L. Li, L. Han, J. V. Moloney, and N. Peyghambarian, Opt. Exp. 16, 16632 (2008).

11. J. Limpert, N. Robin, S.Petit, I. Honninger, F. Salin, P. Rigail, C. Honninger and E. Mottay, Appl. Phys. B 81, 19 (2005)

12. L. Pan, I. Utkin, and R. Fedosejevs, in OSA Technical Digest Series (Optical Society of America, 2007), paper JTuA70.

31


photodynamic treatment, photosensitizer conjugate, wound infection

Introduction

Photodynamic therapy (PDT) makes use of light and a photosensitizer to induce cellular damage in target tissue through generation of reactive oxygen species. Active R&D on the use of PDT for the treatment of cancer started ea r ly s even t i e s w i th the deve lopmen t o f hematophorphyrin derivative (HpD) [1]. The product known as Photofrin is the first clinically approved PDT drug and it has been used widely for the treatment of various types of cancer such as oesophagus, lung, bladder and brain. Apart from Photofrin there are only two other photosensitizers 5-aminolevulinic acid (ALA, a porphyrin precursor) and Foscans (temporfin, meta-tetrahydroxyphenyl chlorin) which have been approved for clinical use in USA and Europe for the treatment of skin cancer and cancer of the head and neck, respectively [2].

Most of the photosensitizers used in PDT are tetrapyrolles compounds generally classified under porphyrins and chlorins. The main difference in porphyrin and chlorin lies in that the one of the four pyroll rings lack a double bond and this single chemical difference provides chlorin better photophysical properties such as higher absorption coefficient at longer wavelength regions (660-800 nm) and better triplet yield [3]. In this respect, naturally occurring chlorophylls which are substituted derivative of chlorins have been considered very useful for the preparation of suitable photosensitizer for PDT. A variety of chlorophyll derivatives have been synthesized and evaluated for their PDT efficacy [4]. On the basis of chemical nature these derivatives can be broadly categorized in two classes: the pheophorbide and its derivatives which are hydrophobic and chlorin e (Ce ) and its derivatives which are 6

hydrophilic. Since pheophorbide and its derivatives are not water soluble, these require suitable carrier system such as liposomes, nanoparticles or other formulations for effective delivery in target cells. Currently, pheophorbide-a derivative HPPH (2-[1-hexyloxyethyl]-

6

Photodynamic therapy (PDT) because of its high selectivity is receiving considerable attention for the treatment of cancer. It exploits selective localization of a photosensitive drug in tumor which when photoexcited leads to the generation of reactive oxygen species and results in destruction of the tumor. PDT is also being actively investigated for antibacterial applications to control and manage antibiotic resistant infections of skin, superficial wounds and dental cavities. Efficacy of PDT puts several demands on the photosensitizer. These include low dark toxicity, high selectivity in tumor localization, absorption at longer wavelength (~ 700 nm) which can penetrate deeper in the tissue. Our earlier studies on the use of chlorin p (Cp ) showed that it is a 6 6

good choice and could be used to treat tumor volumes of 3up to ~ 150 mm . However, for larger volume tumors the

efficacy of PDT was compromised due to inadequate uptake of the photosensitizer. To address this issue, we have conjugated Cp to histamine (Cp -his) to enhance its 6 6

uptake in tumors which over express histamine receptors. Studies in cancer cells and Hamster cheek pouch tumor model with Cp -his showed ~10 times higher uptake in 6

oral cancer cells, enhancement in phototoxicity, improvement in tumor selectively and complete

3regression of larger tumor of volumes up to 1000 mm . Details of these studies and the studies carried out on mechanism of uptake, and localization of Cp -his in cells 6

will be presented.

For antimicrobial applications, we have conjugated Cp 6

with polylysine (Cp -pl), a cationic peptide which is able 6

bind more effectively to the cell wall of Gram-negative bacteria. With Cp -pl, the bactericidal effect of PDT on 6

both Gram positive Methicilin resistant S. aureus (MRSA) and Gram negative P. aeruginosa could be enhanced substantially. Studies on the use of Cp -pl for 6

photodynamic treatment of P. aeruginosa infected wound in mice showed that PDT with Cp -pl could control the 6

bacterial infection and promoted healing of wounds by reducing inflammation and enhancing collagen synthesis. Details of these studies will also be presented.

Key words: chlorophyll derivative, oral cancer,

Investigations on The Efficacy of Chlorin p 6

Conjugates for Photodynamic Treatment of Cancer and Bacterial Infection

Alok Dube*, Khageswar Sahu, Mrinalini Sharma and P.K. GuptaLaser Biomedical Applications & Instrumentation Division,

Raja Ramanna Centre for Advanced Technology, Indore*E-mail: [email protected]

32


for tumors beyond this size the regression was only partial [8].

Studies on PDT of Cancer using Cp -Histamine 6

Conjugate

For enhancing uptake of Cp in tumor cells and thus 6

improve its PDT efficacy for larger tumors we investigate the possibility to use its histamine conjugate for its targeted delivery in cancer cells through histamine receptors. The receptor mediated delivery of photosenstizer to achieve higher selectivity and PDT efficacy is based on the fact that as compared to normal cells, the tumor cells typically have increased expression of cell surface receptors for a particular ligand which can be growth factors or regulatory bio-molecules. For example, epidermal growth factor receptor (EGFR) is expressed in various normal tissues where it has a regulatory role in cell proliferation and it is highly expressed in a number of solid tumors. Photosensitizer conjugated to epidermal growth factor (EGF) as well as various other molecules such as folic acid, low density lipoprotein (LDL), transferrin, insulin have been investigated in various studies [9]. The need for investigating various ligand to target the anticancer drug is based on the fact that the expression level of various receptors can vary in various types of tumors. The over-expression of histamine receptors has been reported for several types of malignancies, e.g. breast carcinoma, melanoma and adrenocortical cancer [10]. It is also pertinent to note that the tumor vasculature is generally more permeable and vascular permeability is regulated through histamine receptor. Further, histamine receptors are also expressed on immature myeloid cells (iMCs) which accumulate within tumors and help in tumor growth. Therefore it is expected that the conjugation with histamine could provide better option for targeted delivery of photosensitizer as compared to EGFR or folate receptors which primarily target the tumor cells.

2-devinyl pyropheophorbide-a) with trade name Photochlor® is under clinical trials for PDT of lung, skin, head and neck, and esophageal cancers [2]. The water soluble derivative Ce and mono-L-aspartyl Ce (MACE) 6 6

have been widely investigated for PDT of cancer [2]. Currently MACE is undergoing phase II trials in USA for glioma and phase III trials for metastatic colorectal cancer and hepatoma [2]. Chlorin p (Cp ) is another 6 6

water soluble derivative which differs from Ce in that it 6

has a carboxylic group attached directly to the pyroll ring that 13 carbon position rather through a methylene (CH2)

group as in Ce . The photodynamic activity of Cp against 6 6

cancer cells was reported in 1986 by Hoober et al [5]. Although, some derivatives of Cp e.g. lysyl Cp , 6 6

cycloimide Cp and their methyl esters have also been 6

investigated for PDT of cancer [6], Cp remained less 6

explored. This could be perhaps because of easier synthesis of Ce from pheophorbide in one step as 6

compared to the synthesis of Cp which requires three 6

step procedure. While Ce can be obtained by direct 6

alkaline hydrolysis of pheophorbide, for the synthesis of Cp the pheophorbide is first converted into purpurin-18, 6

which because of hydrophobic character can be easily purified by column chromatography and then through one step mild hydrolysis of purpurin-18, pure Cp can be 6

obtained. Although both Ce and Cp have almost same 6 6

absorption properties (ε ~ 25000 at 654 nm, singlet oxygen yield 0.7 in D2O), the advantage of Cp over Ce 6 6

is that it contains only single long chain carboxylic group which make it easier to prepare conjugates in pure form. For example, MACE is synthesized by coupling L-aspartic acid to the side chain carboxylic group of Ce via 6

an amide bond. However, since there are two long chain carboxylic groups in Ce , the direct coupling leads to the 6

formation of two types of conjugates. Therefore, the synthesis of pure MACE from Ce is bit more involved 6

and complex.

Our studies on the efficacy of Cp for PDT of cancer in 6

hamster cheek pouch model showed that Cp6

administered systemically at dose of 1.5 mg/kg body weight accumulated preferentially in small size tumors

3(~60 mm ), cleared rapidly from the skin and led to complete tumor regression after PDT [7]. However, for relatively large tumors its uptake was poor which compromised the PDT efficacy. Therefore, to improve PDT efficacy of Cp we investigated the effect of drug 6

dose on PDT-induced tumor regression and mode of tumor damage. Results showed that as compared to lower drug dose (2.0 mg/kg body weight) which led to complete

3regression of tumor of volume upto 50-80 mm , the higher dose of Cp (4 mg/kg body weight) could induce 6

3tumor regression in tumor of volume up to 130 mm but Fig. 1: Chemical structure of Cp and Cp -his6 6

33


results suggested that the conjugating Cp with histamine 6

can help improve the effectiveness of PDT in oral cancer cells by enhancing its intracellular delivery.

After confirming the efficacy of Cp -his in vitro [11], we 6

investigated the use of Cp -his for PDT of tumors in 6

Hamster cheek pouch model. To determine the tumor selectivity and systemic clearance of Cp -his, its 6

accumulation in tumors, normal tissue and skin was monitored by in vivo fluorescence spectroscopy. For measurements from tumors, the fluorescence of endogenous porphyrins was subtracted to resolve

Histamine can be easily conjugated to Cp through amide 6

bond formation between 17'-carboxilic group of Cp and 6

amino group of the histamine using standard carbodiimide coupling reaction. The chemical structures of Cp and its conjugate are shown in fig. 1. Since the 6

over-expression of histamine receptors has been well documented for breast carcinoma [10], first we studied the cellular uptake and phototoxicity of the conjugate and free Cp in breast cancer cells MCF7. Results showed that 6

the cellular uptake of Cp -his was ~2.5 times higher than 6

free Cp and as a result the phototoxicity was also 6

enhanced by a factor of ~3.

Since results with Cp -his in MCF7 cells were 6

encouraging, we carried out detailed studies to investigate the use of Cp for PDT of oral cancer. For in 6

vitro studies on uptake and phototoxicity of Cp -his, we 6

used human oral cancer cells NT8e and 4451 which are derived from tumor specimen of the upper aerodigestive tract and a recurrent tumor in the lower jaw, respectively. Results showed that in both the cell lines the accumulation of Cp -his was ~ 10 times higher as 6

compared to Cp (fig 2a). We also investigated whether 6

intracellular uptake of Cp -his is mediated through 6

histamine receptors. Surprisingly, no inhibition of the cellular uptake of Cp -his was observed in the presence of 6

histamine which is a natural ligand of histamine receptor. Since increasing concentration of histamine from 1 mM to 5 mM also had no inhibitory effect on the uptake of Cp -his, it suggested the possibility that the binding of 6

Cp -his with histamine receptor is stronger as is the case 6

for antagonists. Therefore, the cellular uptake of the Cp -6

his was measured in presence of H1 and H2 histamine receptor antagonist namely, pheniramine and ranitidine. In both the cell lines ranitidine at 100 µM led to significant inhibition (~30%) in the cellular uptake of Cp -his (fig 2b). These results suggested that the uptake 6

of Cp -his is mediated by H2 receptor. Using western blot 6

technique it was also confirmed that H2 receptor is expressed in both oral cancer cell lines (fig 2b, inset) similar to MCF 7. Further, the cellular uptake of Cp -his 6

was observed to decrease upon incubation of cells at lower temperature (~10C) which confirmed that receptor mediated endocytosis play a role in its uptake.

To assess the PDT effectiveness, the cells were treated with either Cp -his or Cp (5 µM) and then irradiated with 6 6

2red light at varying dose (0-38 kJ/m ). Results showed that in both the cell lines, as compared to Cp much higher 6

phototoxicity (~50%) could be induced with Cp -his even 62at lower light dose (~12 kJ/m ) (fig 2c.) At higher light

2dose (28 kJ/m ), while the phototoxicity induced by Cp -6

his was ~95%, with Cp it was much less (~50%). These 6

Fig. 2: Time dependent cellular uptake of Cp and Cp -his (5 6 6

μM) in 4451 and Nt8e cells (a), the effect of histamine receptor antagonist on the cellular uptake of Cp and Cp -his (b) inset- 6 6

histamine H2 receptor protein bands detected in two cell lines by western blot, and percent phototoxicity induced in 4451 and Nt8e cells after treatment with Cp and Cp -his (5 μM) and 6 6

varying doses of red light (c).

34


fluorescence band of photosensitizer. The peak (~670 nm) fluorescence intensity was used to determine the relative level of Cp -his in various tissues. In Table 1, the 6

ratio of photosensitizer level in tumor Vs normal mucosa for four different animals is presented. For Cp -his the 6

ratio of tumor Vs normal mucosa level was ~9.5 which is ~3 times higher as compared to Cp suggesting higher 6

tumor selectivity for Cp -his. The fluorescence 6

measurements from skin showed that Cp -his clear 6

rapidly from the skin and its level at 72h decreased to ~80% relative to peak level at 3 h (fig 3). These results showed that the tumor selectivity and clearance of Cp -6

his is better as compared to clinically approved mTHPC (meta(tetrahydroxyphenyl)chlorin) and HPPH for which

tumor vs. normal tissue ratio of ≥ 2 and prolonged retention in the skin taking 10 days to clear by ~80% has been reported in hamster model [12,13].

Table 1. Ratio of fluorescence intensity in tumor and normal mucosa for Cp -his and Cp (3.0 mg/kg body 6 6

weight). ** For Cp -his the ratio of tumor: normal 6

mucosa is significantly higher than for Cp (P value 6

<0.001).

To assess the PDT efficacy of Cp -his, the tumors were 62 irradiated with red light at dose of 100 J/cm at 3 h after

Cp -his administration. The PDT-induced cellular 6

damage in tumor was assessed at 24 h post PDT by histology using eosin and hematoxylin staining and the results are shown in fig 4, left panel. The control tumor is stained dark with hematoxylin staining (fig. 4a, left panel) which is because of the presence of intact nucleus in tumor cells. In contrast, absence of hematoxylin staining in treated tumor (fig. 4b, left panel) indicates extensive cellular damage. The photographs of tumor in a representative animal before and after PDT are shown in fig 4, right panel. The size of this tumor before PDT was

3~520 mm and the mucosa surrounding the tumor had prominent vasculature (fig 4a, right panel). One week after PDT, the tumor reduced to ~95% of its original size and the mucosal vasculature also attains a normal morphology (fig. 4b, right panel). During follow up, the regressed tumor did not show further growth and the morphology of mucosa and vasculature around the tumor was completely normal (fig. 4c, right panel). To determine extent of tumor regression, the physical volume of the tumor before and one week after PDT was measured and data are presented in Table 2. All the tumors of small size regressed completely and larger

3tumors (>500 mm ) regressed to ~95%.

Results of these studies demonstrated that the coupling of Cp to histamine helped improve the PDT of tumors in 6

Fig. 3: The level of Cp -his (circle) and Cp (square) in the 6 6

abdominal skin of the hamster at different time interval. Inset – Representative fluorescence spectra of Cp -his collected from 6

skin of the hamster at 4 h, 24h, 48h and 72h after administration of 3 mg/kg body weight. Spectra of Cp were similar.6

Fig. 4: Left panel - Photomicrographs showing histology of untreated tumor tissue (a) and tumor tissue subjected to PDT (b). Magnification 400X, Bar- 400 m. Right panel - Photographs showing PDT-induced tumor regression in a representative animal. (a) Tumor before PDT and (b) one week after PDT (c) One month after PDT.

35


considered remote. In addition, PDT can inactivate bacterial virulent factors [21, 22], biofilms [23] which also can contribute to faster healing.

The bactericidal efficacy of PDT is governed by the chemical characteristics of the photosensitizer (PS) used particularly its charge and hydrophobicity. Generally cationic photosensitizers are more suitable for APDT because these are preferentially taken up by bacterial cells. This is because the bacterial outer envelope is composed of predominantly anionic phospholipids while the mammalian cell membrane outer leaflet comprises of mostly zwitterionic phospholipids. Also, the transmembrane potential of bacteria is more negative (-120 to -150 mV) than that of mammalian cells (-90 to -110 mV). Amphiphilic photosensitizers are preferred over hydrophobic or hydrophilic PS because these can be easily solubilized in aqueous media and can also partition efficiently into microbial membrane [24]. Cationic phenothiaziniums, like methylene blue and toluidine blue have been the most widely investigated PS for APDT. However, these have relatively lower triplet yield [25] and are substrates for multidrug efflux pumps [26]. Chlorophyll derivatives like Cp although attractive 6

because of good triplet yield, amphiphilicity and strong absorbance in the red (660 nm) region [7] being negatively charged it has not received much interest for APDT. To exploit the photophysical properties of Cp6 for APDT, we have conjugated it with a cationic peptide 'poly-L-lysin' and investigated the use of this conjugate (Cp -pl) for photodynamic inactivation of Gram positive 6

Methicilin resistant S. aureus (MRSA) and Gram negative P. aeruginosa. Since apart from antibacterial effect, APDT can also have some effect on wound healing processes such as inflammation and collage remodeling, we have investigated the possible effect of APDT on these processes using biochemical assays, histology and Polarization sensitive Optical Coherence Tomography (PSOCT).

Efficacy of Cp -pl for APDT6

Results on uptake of free Cp and Cp -pl in MRSA and P. 6 6

aeruginosa are shown in figure 5. For MRSA as well as P. aeruginosa the uptake of Cp -pl was found to be 5 times 6

higher than Cp at 2 µM and 4 µM, respectively. It was 6

also observed that, as compared to P. aeruginosa (fig. 5b), the uptake of free Cp was higher in MRSA and showed 6

saturation beyond 4µM (fig. 5a). This difference was expected because of the difference in the nature of the outer wall in Gram positive and Gram negative bacteria.

Measurements on cell survival of bacteria showed that while Cp in dark (1.0-8.0 µM) led to no toxicity, for Cp -6 6

pl the cell viability decreased significantly (~1.0 log) in both types of bacteria (fig. 6 a,b). For MRSA subjected

hamster model [14]. The efficacy of Cp -his for oral 6

cancer in human requires that H2 receptor is over-expressed in tumor. Therefore we analyzed H2 receptor expression in few biopsy samples of human OSCC tissue by immunohistochemistry and observed that expression of H2 receptor in human OSCC is very similar to hamster tumor. In addition, pronounced staining was also observed in blood vessels and matrix cells comprising host immune cells as expected due to role of histamine receptor in vascular permeability and immune-modulation. This would suggest that conjugation of Cp 6

with histamine open new possibility for its targeted delivery and improving efficacy for PDT of oral tumors.

Table 2: Tumor volume before and one week after photodynamic treatment using Cp -his at 3.0 mg/kg body 6

2weight and light dose of 100 J/cm .

Antimicrobial PDT for Treatment of Skin and Wound Infections

Chronic wounds represent a major health care burden particularly in diabetic and immunocompromised conditions because of the growing emergence of antibiotic resistance in bacteria [15-17]. Chronic wounds generally contain multiple species of microbial organisms including drug resistant Gram-positive bacteria such as Staphylococcus aureus, Streptococcus pneumoniae, Enterococcus spp. and drug resistant Gram-negative bacteria e.g. Acenatobacter baumannii, Klebsiella pneumoniae, and Pseudomonas aeruginosa .

Antibiotic resistance in bacteria can arise mainly because of presence of enzymes for antibiotics degradation or modification and efflux pump to prevent drug accumulation in bacteria. To overcome the problems of antibiotic resistance, alternative strategies such as phenotypic modification, antimicrobial peptides, bacteriophage therapy and antimicrobial photodynamic therapy (APDT) have been investigated [18, 19]. Of these, APDT is particularly attractive [20], because, while bacteria are expected to develop resistance mechanisms against the other approaches [18], the possibility of developing resistance against APDT is

36


2to PDT with red light (~ 25 J/cm ), the pre-treatment with 1.0 µM Cp -pl was enough to induce substantial decrease 6

(~ 4 logs) in cell survival (fig. 6 a). Under similar conditions, 4.0 and 8.0 µM Cp led to decease in cell 6

survival only by ~ 0.9 log and ~ 2.3 log, respectively (fig. 6a). For P. aeruginosa, PDT using Cp -pl at concentration 6

of 2.5 and 5.0 µM of Cp -pl led to decrease in cell 6

survival by 3.5 log and 5.0 log, respectively and in comparison Cp (1.0-8.0µM) did not lead to any 6

significant decrease in cell survival (fig. 6b). We also checked whether Cp -pl mediated PDT can exert any 6

cytotoxic effect on keratinocytes, the cells which are responsible for formation of epithelial layer in wound healing. From the results presented (fig. 6c) it is evident that there is no significant phototoxicity induced in keratinocytes (HaCaT) at Cp -pl concentration and red 6

light dose (point marked by arrow) which results in 2-3 log decrease in survival of bacteria. These results, suggest that use of Cp -pl can help improve the antibacterial 6

efficacy of APDT without having any significant toxicity

to host cells.

We also investigated the efficacy of APDT using Cp -pl 6

for improvement of healing of P.aeruginosa infected wounds in mice. The results showed that treatment of

Fig. 5: Uptake of Cp and Cp -pl by MRSA (6.a) and P. 6 6

aeruginosa (6. b) incubated with different concentrations of Cp and Cp -pl in dark for 15 minutes. Values represent mean ± 6 6

standard deviation.

Fig. 6: Surviving fraction of MRSA (a) and P.aeruginosa (b) and HaCaT cells (c) treated with Cp and Cp -pl at the indicated 6 6

concentrations for 15 minutes in the dark and then exposed to red light (660 nm ± 40nm).

37


collagen present in extracellular matrix. In chronic wound infections, excessive collagen degradation by bacterial proteases [28] or bacterial protease induced activation of latent host collagenases [29] like matrix metalloproteases can severely impair wound healing. Therefore, inactivation of bacteria and the bacterial proteases following APDT are expected to restore collagen remodeling in wounds. We investigated the effect of Cp -pl mediated topical APDT on collagen 6

remodeling in wounds infected with P. aeruginosa and S. aureus (MRSA). Collagen remodeling was monitored by measurement of tissue retardance using Polarization Sensitive Optical Coherence Tomography (PSOCT) as well as by measurements on hydroxyproline content and matrix metalloproteases (MMP-8 and MMP-9) level in wounds. Fig. 8 a-d shows the representative OCT, PSOCT and histology images of P. aeruginosa infected wounds without treatment and wounds subjected to APDT. The OCT images of untreated infected wounds (Fig. 8 a) clearly show the presence of crust and edematic region. In PSOCT image of untreated infected wounds (Fig. 8 b) the polarization contrast is considerably reduced as compared to untreated uninfected wounds

thwhich suggest degradation of collagen structure. At 18 day after APDT, the re-epithelialization in wound tissue is clearly visible (Fig. 8 c) and there is increase in polarization contrast in PSOCT image (Fig. 8 d) which suggest restoration of collagen structure. The histology of wound tissue correlated with these observations (fig 8 e-f). To quantify the collagen remodeling, the PSOCT images were analyzed to compute the tissue retardance and the results are shown in fig 8 g. Results showed that as compared to both untreated wounds and wounds treated with Cp -pl alone, in infected wounds treated with both 6

Cp -pl and red light the value of retardance were higher 6

by a factor of ~2 (fig. 8 g). APDT also led to increase in hydroxyproline content and decrease in MMP-8, 9 levels. For wound infected with MRSA the results were similar. These results together suggest that APDT expedites healing in bacteria-infected wounds by enhancing epithelialization and collagen remodelling [30].

Conclusion

Photodynamic treatment of oral cancer and infected wounds has been investigated using conjugates of Cp . 6

Results showed that conjugation of Cp to histamine 6

enhanced the cellular uptake, tumor selectivity and efficacy of PDT-induced tumor regression. For antimicrobial PDT, conjugation of Cp with polylysine 6

helped to improve the PDT-induced inactivation of bacteria, particularly for P. aeruginosa, reduction in bacterial load and hyperinflammatory conditions in infected wounds because of which the healing of wounds could be accelerated significantly.

2wounds with Cp -pl and light (~120 J/cm ) led to >1.5 log 6

decrease in the bacterial load at 24 h and subsequently a further decrease by >3 logs was observed at 72 h after PDT (fig. 7 a). APDT with Cp -pl also showed significant 6

acceleration of wound healing by reducing the time of wound closure from ~20 days to ~15 days (fig 7b). It may be noted that though the treatment of wounds with silver nitrate (AgNO ) also led to significant antibacterial effect 3

(fig. 7 a), it did not show any improvement in wound closure (fig. 7 b). From the histology of wound tissue and measurements on biomarkers of inflammation (IL-6 and TNF-α), it was observed that the inflammation is substantially reduced after APDT [27].

An important aspect of wound healing is remodeling of

Fig. 7: Effect of APDT on bacterial counts in wounds in different groups at 24 hr, 48 and 72 hr post treatment time(a). Effect of APDT on wound closure of P.aeruginosa infected wounds of mice (b).

38


References1. J. Moan, Q. Peng, Anticancer Res. 23, 3591 (2003).2. P. Agostinis, K. Berg, K.A. Cengel, T.H. Foster,

A.W. Girotti, S.O. Gollnick, S.M. Hahn, M.R. Hamblin, A. Juzeniene, D. Kessel, M. Korbelik, J. Moan, P. Mroz, D. Nowis, J. Piette, B.C. Wilson, J. Golab, CA Cancer J. Clin. 61, 250 (2011).

3. M. Palma, G.I. Cárdenas-Jirón, M.I. Menéndez Rodríguez, J. Phys. Chem. A 112:13574 (2008).

4. E.S. Nyman, P.H. Hynninen, J. Photochem. Photobiol. B 73,1 (2004).

5. J.K. Hoober, T.W. Sery, N. Yamamoto, Photochem. Photobiol. 48,579 (1988).

6. D. Kessel, K. Woodburn, C.J. Gomer, N. Jagerovic, K.M. Smith, J. Photochem. Photobiol. B 28,13 (1995).

7. A. Dube, S. Sharma, P.K. Gupta, Oral Oncol. 42,77 (2006).

8. A. Dube, S. Sharma, P.K. Gupta, Oral Oncol. 47,467 (2011).

9. B. Chen, B.W. Pogue, P.J. Hoopes, T. Hasan, Crit. Rev. Eukaryot. Gene Expr. 16,279 (2006).

10. E.S. Rivera, G.P. Cricco, N.I. Engel, C.P. Fitzsimons, G.A. Martín, R.M. Bergoc, Sem. Cancer Biol. 10,15 (2000).

11. A. Parihar, A. Dube, P.K. Gupta, Cancer Chemother. Pharmacol. 68:359 (2011).

12. S. Andrejevic, J.F. Savary, P. Monnier et al, J. Photochem. Photobiol. B 36,143 (1996).

13. K. Furukawa, H. Yamamoto, D.H. Crean, H. Kato, T.S. Mang, Las. Surg. Med. 18,157 (1996).

14. A. Parihar, A. Dube, P.K. Gupta, Photodiagnosis Photodyn. Ther. 10, 79 (2013).

15. J.R. Mekkes, M.A.M Loots, A.C. Van Der Wal, J.D. Bos, Br J Dermatol. 148, 388 (2003).

16. S. Enoch, D.J. Leaper, Surgery (Oxford); 26,31 (2008).

17. C.J. Schofield, G. Libby, G.M. Brennan, R.R. MacAlpine, A.D. Morris, G.P. Leese, Diabetes Care. 29, 2252 (2006).

18. P. W. Taylor, P.D. Stapletona, J. P. Luziob , Drug Discovery Today. 7,1086 (2002).

19. S.M. Mandal, A. Roy, A.K. Ghosh , T.K. Hazra , A. Basak , O.L. Franco, Front Pharmacol. 5,105 (2014).

20. T. Maisch , Mini Rev. Med. Chem. 9: 974 (2009).21. S.Tubby, M.Wilson, and S. P.Nair, BMC

Microbiol. 9, 211 (2009).22. M. Sharma, H. Bansal, P.K. Gupta, Curr. Microbiol.

50,277 (2005).23. M. Sharma, L.Visai, F.Bragheri, I. Cristiani, P.

K.Gupta, P. Speziale, Antimicrob. Agents Chemother. 52, 299 (2008).

24. K. Matsuzaki, Biochim. Biophys. Acta. 1788,1687 (2009).

25. E. Alves , M.A. Faustino, M.G. Neves, A. Cunha, J. Tome, A. Almeida, Future Med. Chem. 6,164 (2014).

26. G.P. Tegos, M.R. Hamblin, Antimicrob. Agents Chemother. 50,196 (2006).

27. K. Sahu, M. Sharma, H. Bansal, A. Dube, P.K. Gupta, Lasers Med. Sci. 28, 465 (2013).

28. L.W. Heck, K. Morihara, W.B. McRae and E.J. Miller, Infect. Immun. 51,115 (1986).

29. D.J.Harrington, Infect. Immun. 64,1885 (1996).30. K. Sahu, M. Sharma, P. Sharma, Y. Verma, K.D Rao,

H.Bansal, A. Dube and P.K.Gupta Photomed. Laser Surg. 32, 23 (2014).

Fig. 8: Effect of APDT on collagen remodeling of P.aeruginosa infected wounds. OCT images (a & c), PSOCT images (b & d). The images of untreated wound (a,b) and wound subjected to PDT (c,d). Image size in OCT images: 1.5 mm x 3 mm. Tissue histology depicting Masson's trichrome staining in untreated (e) and treated wound (f). Scale bar: 100 µm. Letter 'E' indicate epithelium layer and * indicate granulation tissue. Fig 8g shows the retardance calculated from PSOCT images of

thwounds. Results shown are for wounds at 18 day post wounding.

39


In a multi-step photoionization, the number of steps is decided by the ionization potential (IP) of the element to be studied and the lasers. The ionization potential of lanthanides and actinides is in the range of 5-7 eV [9,14]. Consequently, for an isotope-selective photoionization process based on visible lasers whose photon energy is ~2 eV, a three-step resonant ionization is most appropriate. In the three-step photoionization process (Fig.1), the atom in the ground state or in the lowest metastable level

st ndis excited up in the ladder sequentially by 1 and 2 step -1lasers to one of the energy levels in the 15000-19000 cm

-1 region and in the 30000- 36000 cm region respectively, rdbefore it is finally photoionized in the 3 step, probably

through an autoionization state (AI), by another laser. Therefore, the information on the atomic parameters in these energy regions is essential for isotope- selective photoionization of these elements.

Isotope shift and hyperfine structure of all the levels in an excitation sequence are required to set the spectral bandwidth of the lasers to achieve the desired selectivity in the process. Till recently, most of the isotope shift measurements reported in the literature were performed employing high-resolution spectroscopy techniques such as ring - dye - laser - induced laser fluorescence spectroscopy [19] as well as conventional Fourier-transform spectroscopy [20]. These measurements are mainly restricted to low-lying energy levels and are done

Abstract

Multi-step resonant photoionization spectroscopy combined with mass spectrometry is an excellent tool for investigating the complex atomic structure of heavy elements like lanthanides and actinides especially in the high energy regions which are not easily accessible with the conventional spectroscopic techniques. In this article, various laser spectroscopic techniques based on multi-step photoionization spectroscopy used for the measurements of atomic parameters relevant to isotope-selective photoionization process are discussed with some examples from the literature and from the recent measurements carried out in our laboratory. A novel laser spectroscopic technique of isotope selective photoionization of odd-isotopes using broadband lasers by exploiting polarization selection rules is also discussed.

Introduction

Atomic parameters such as isotope shift, hyperfine structure, radiative lifetime, branching ratio, transition probability and photoionization cross-section are the most fundamental characteristics of an atomic system. The information on these atomic parameters is of great interest in many areas of basic science and in laser-based applications such as isotope-selective photoionization processes, elemental ultra trace analysis, etc. Over the last three and half decades there has been a continuing interest in atomic vapour laser isotope separation (AVLIS) of lanthanides and actinides, especially of uranium [1-5]. Since the AVLIS process is based on isotope-selective multi-step photoionization in atomic vapour stream, the precise information on these parameters is a pre-requisite in selecting an appropriate multi-step photoionization scheme for the AVLIS process. Multi-step laser resonant ionization spectroscopy [6-18] is an excellent tool to study the complex atomic structures of these elements and has played an important role in the measurement of these parameters, especially in the high energy regions which are not easily accessible with the conventional spectroscopic techniques.

Multi-Step Photoionization Spectroscopy of Lanthanides and Actinides: Measurements of Atomic Parameters

Relevant to Isotope-Selective Photoionization ProcessesVas Dev

Laser and Plasma Technology Division,Bhabha Atomic Research Centre, Trombay, Mumbai – 400085, India

E-mail: [email protected]

Fig. 1: Typical three-step photoionization pathway.

40


estimation of transition probabilities. The other laser-based methods such as laser saturation of transition, Rabi frequency measurement and Autler-Towns splitting, etc [28,29] are not significantly affected by these uncertainties, but put stringent requirements on the laser parameters such as spectral width and temporal profile of the laser pulse, etc.

Carlson et al. [30] have developed an alternative method for measuring absolute transition probability based on the delayed photoionization technique which is not subjected to above difficulties. Using this method, they have measured branching ratios of uranium transitions at 4362.1 and 4393.6 Å wavelengths in the blue region of uranium spectra and radiative lifetimes of the upper levels of these transitions independently. Hackel et al. [31] have refined this to accurately measure the transition probability of a uranium transition at 6395.4 Å in the red region of uranium spectrum. Recently, we have measured the radiative lifetimes, branching ratios and transitions probabilities of uranium by employing the said technique [32, 33]. Miyabe et al. [34] have exploited the same to measure the absolute transition probabilities of gadolinium. Bisson et al. [35] have extended it further to measure the excited-state-to-excited-state transition probability of cerium. In this article, various laser spectroscopic techniques based on multi-step photoionization used for the measurements of these parameters are discussed with some examples from the literature and from the recent measurements carried out in uranium and samarium in our laboratory. A novel laser spectroscopic technique of isotope-selective photoionization of odd-isotopes using broadband lasers by exploiting polarization selection rules is also discussed.

Experimental

A typical experimental set-up used for a three-step photoionization spectroscopy is shown in Fig.2. Basically, it consists of three pulsed dye lasers pumped by the second harmonic of two Nd:YAG lasers at 532 nm, a high temperature oven assembly coupled to an indigenously built time of flight mass spectrometer (TOFMS), a U-Ne hollow cathode discharge tube (HCDT), boxcar averagers, Fabry-Perot (FP) etalon, photodiode, digital oscilloscope, etc. The spectral width, repetition rate and temporal pulse width of the lasers are

-10.08 cm , 20 Hz, 8 ns (FWHM) respectively. Temporal sequencing of laser pulses from different lasers is achieved using an electronic delay generator to an accuracy of ± 1 ns. The laser beams are apertured individually and spatially overlapped in the interaction zone using beam steering optics. Uranium vapour is generated by resistively heating a few hundred mg of

0uranium metal in a high temperature oven at ~ 1600 C in

to understand the nuclear structure. However, in the high energy region where most of the levels suffer from considerable configuration mixing, measurements of isotope shifts are inadequate to provide sufficient information on the nuclear structure. Hence, there are only very few measurements in the energy region above

-130,000 cm . Multi-step photoionization in an atomic beam combined with mass spectrometry has been developed as an alternative to high-resolution spectroscopy technique to measure the isotope shifts of the high-lying levels of uranium by Miyabe et al. [21]. We have recently used this technique to measure the isotope shifts of the high-lying levels of samarium [22].

Radiative lifetimes of the atomic levels in the excitation ladder should be longer than the laser pulse duration and branching ratios of the transitions to the respective originating levels should be large for effective utilization of atoms in the process. More importantly, the transition probabilities which determine the laser fluence needed in each step for effective atomic photoionization should be high and must be known accurately.

Transition probabilities are conventionally obtained either from the measurement of relative line strengths in the emission spectra acquired from arcs, sparks or discharge lamps or from the measurements of absolute absorption [23-25]. The accuracy of these measurements is limited by the uncertainties associated with the source temperature, relative emission intensities over wide wavelength range, atomic vapour density, density profile, path length, self absorption, etc. In recent years, one of the most commonly used method to obtain the transition probability is based on the measurements of branching ratio (β) and radiative lifetime (τ), where β is generally determined from the relative intensities of emission lines measured over a wide wavelength range using Fourier transform spectrometer and τ is measured independently [26, 27]. The absolute transition probability (A) is related to these two atomic parameters by a well-known relation, A = β/τ, where β is the branching ratio of the transition and τ is the lifetime of the upper level. In this method, the complete understanding of the emission spectra is necessary for accurate measurement of the absolute transition probability. With the presence of many optically active electrons in the outer most shell of these elements, the electronic structure is quite complex and, hence, this method is not advantageous for heavy elements. Petit et al. [28] have measured the β-values of

ndatomic uranium transitions originating from a 2 excited energy level using laser-induced fluorescence (LIF) and independently measured the τ-value of this level using delayed photoionization technique. However, one of the major drawbacks of this method is that a large number of undetected infrared transitions may result in over

41


level to a level of interest in the energy region 15500--119000 cm . The population in the excited level is

monitored by ionizing the excited atoms by stepwise two-photon photoionization process, using the second dye laser (probe laser) at various time delays between the pump and probe lasers. To avoid the effect of the excited atoms drifting out of the interaction volume due to atomic beam flow velocity, the pump laser beam diameter was kept large compared to that of the probe laser. The experimental results are fitted using the exponential decay formula

i(t) = i exp(-t/τ),0

where τ is the lifetime, i(t) is the two-color three-photon photoionization signal at delay time t and i 0

photoionization signal corresponding to no delay. A typical population decay curve of uranium level at

-1 15631.85 cm is shown in Fig. 4. These measurements have yielded the radiative life time of 630±30 ns for this level which is close to the value of 607±20 ns reported in

ndthe literature. Similarly the lifetimes of the 2 step -1excited levels in the energy region 30000-36000 cm are

measured by a three-color three-step delayed photoionization method. As shown in Fig.3b, the excited level of interest was pumped by stepwise excitation using two pulsed lasers synchronized in time and the population of the excited level is monitored by ionizing the excited atoms using the third dye laser (probe laser) at various time delays between the pump and probe lasers.

-7 a vacuum chamber maintained at 2x10 mbar pressure, whereas in the case of samarium, the operating

0temperature of the oven is of the order of ~600 C. The uranium atoms effusing out of the oven through an orifice of 1.5 mm diameter are further apertured at a distance of 20 mm downstream by placing another fixed aperture of diameter 2 mm to form a well-collimated atomic beam. Typical atomic number density in the laser interaction

7 -3region is ~ 10 cm . The atomic absorption line width in the interaction zone is about 50 MHz. Spatially overlapped laser beams interact with the atomic beam in a cross configuration and the resultant photo-ions produced are extracted and introduced into the TOFMS by a dc electric field of 120 V/cm, and finally detected by a micro-channel plate (MCP) detector. A part of the laser beam is made to pass through a U-Ne hollow cathode

-1discharge tube and a Fabry-Perot (FP) etalon of 0.5 cm free spectral range for the purpose of laser wavelength calibration by simultaneously recording optogalvanic signal and FP etalon transmission fringes detected with a photodiode.

Results and Discussion

Lifetime Measurement

stThe lifetime measurements of 1 step levels are carried ndout using two pulsed dye laser systems and those of 2

step levels are performed using three independent pulsed dye laser systems. The Schematics of photoionization sequences for lifetime measurements are shown in Fig. 3. As shown in the figure, atoms are excited to the upper level of interest by pump lasers, and subsequently ionized

stby probe lasers. To measure the lifetimes of the 1 excited levels, the first dye laser is used as a pump laser to excite the uranium atoms from the ground or the first metastable

Fig. 2: Schematic of the experimental set up. MO-master oscillator, EDG-electronic delay generator, DL- dye laser, BS- beam splitter, HCDT-hollow cathode discharge tube , BD- beam dump, BCA- box car averager, PC/CR- personnel computer/ chart recorder, TOFMS –time-of-flight mass spectrometer, DSO- digital storage oscilloscope, MCP- micro channel plate detector, VA- variable aperture, FPE- Fabry Perot etalon and PD- photo diode. Inset: cross sectional view of the laser beams in the interaction zone.

Fig. 3: Schematic of the photoionization sequence for lifetime measurements of a) even-parity levels of uranium in the energy

-1region 15000 – 19000 cm , b) odd-parity levels in the energy -1region 30000 – 36000 cm . The ground level of uranium is of

odd-parity.

42


depicted in Fig. 5b, where S represents the 0

photoionization signal of the probe laser in the absence of the pump laser, S(0) is the signal of the probe laser when both the pump and probe lasers are synchronous, i.e. at t =

0, and S(¥ ) is the probe signal at infinite time delay

Branching Ratio Measurement

The measurement of branching ratio is performed using three pulsed dye lasers. One dye laser is used as a pump and the other two are used as a probe. The level schematics of excitation pathways for the pump and probe lasers to measure branching ratios by delayed photoionization method are shown in Fig. 5a. The

-1uranium atoms from level 1 (0 or 620.32 cm ) are excited to level 2 by the pump laser which is in resonance to the transition between levels 1 and 2. The population in level 1 is probed via two-color three-photon photoionization process using two synchronized probe lasers. The resultant photoionization signal is recorded as a function of delay between the pump and the probe lasers. Due to finite radiative lifetime (τ) of the excited level (2), the excited atoms decay back to various lower energy levels including level 1. The population dynamics of levels 1 and 2 just after the pump laser pulse can be written as

(1)

(2)

By solving these equations, we get

(3)

where N (0) and N (0) are the populations in levels 1 and 1 2

2 respectively just after the pump laser pulse set at t = 0. In the case of complete saturation and equal level degeneracies, N (0) = N (0) = N /2, where N is the 1 2 10 10

initial population of level 1 just before the pump laser pulse. The photoionization signal produced by the probe lasers is directly proportional to the population of the level 1. Due to repopulation of the level 1 by spontaneous decay of the level 2, the probe signal increases with time delay between the pump and probe lasers. The delayed photoionization signal of the probe from the level 1 is

Fig. 4: Population decay curve of uranium atomic level at -115631.85 cm .

Fig. 5: a) Schematic of the excitation pathways for the pump and probe lasers to measure the branching ratio of the transition between levels 1 and 2. b) Representative probe signal versus time delay between the pump and the probe laser pulses. c). Repopulation curve of the ground level of uranium

-1from the excited atomic level at 15631.85 cm . The solid line is a fit of the data using Eq.(3) for β = 0.57 ±0.05 and τ = 592 ± 60 ns.

(a)

(b)

(c)

43


which is the solution of the two-level rate equations. In a multi-step photoexcitation process when lasers pulses from different lasers are disconnected in time, the two-level approximation of the atom is quite valid.

Isotope Shift Measurement

For isotope shift measurement, we have chosen samarium as a representative of these elements.

144Samarium has seven naturally abundant isotopes Sm, 147 148 149 150 152 154Sm, Sm, Sm, Sm, Sm and Sm with an abundance of 3.1, 15, 11.3, 13.8, 7.4, 26.6 and 22.6 % respectively. The isotope shift of high-lying levels of Sm

154 144between Sm and Sm isotopes has been measured by using two-color resonance photoionization mass spectrometry. A typical time-of-flight mass spectrum of

Sm ion produced by nonselective photoionization of Sm in an atomic beam using single-color three-photon technique at 570.68 nm wavelength is shown in Fig. 7. It may be noted that the isotopic composition of the mass peaks shown in the figure slightly deviates from the natural abundance because the laser is preferentially

144tuned towards Sm to enhance the weak signal of the minor isotope. For measuring the isotope shift, the wavelength of the first-step laser was fixed in between the

144 154resonances of Sm and Sm isotopes and the second-step laser wavelength was scanned. The mass selective two-color photoionization spectra of the naturally

between the pump and probe lasers, i.e. at t → . The

branching ratio can then be written as β = [S( )-S(0)]/[S -0

S(0)], where [S -S(0)] represents the depopulation of 0

level 1 at time t = 0 and [S(¥ )-S(0)] is the repopulation of

the level 1 at time t → ¥ . Repopulation of the level 1 due to the radiative decay from the level 2 is about 99.3 % at t =

5τ, which is a good approximation for t → ¥ .

This method is illustrated by measuring the branching ratio of the uranium transition at 6395.42 Å. The probe signal for various time delays is shown in Fig. 5c. The value of the branching ratio is obtained by fitting the repopulation data using Eq. (3). The measured β value equal to 0.57 ± 0.05 matches very well with the value reported by Hackel et al [29]. Using multi-step delayed photoionization method, we have measured lifetimes and branching ratios of uranium with an accuracy better than 10% [32,33]. By combining the experimentally measured values of radiative lifetimes and branching ratios, the absolute transition probabilities are determined.

Photoionization Cross-section Measurement

The photoionization cross-section (σ) from the highly ndexcited level (2 excited level) of uranium atoms is

measured by saturation method. In this method, the photoionization signal is monitored as a function of the third-step laser fluence while the fluence of other two lasers kept constant. Fig.6 shows the dependence of photoionization signal from second excited level at

-1 rd35578.25 cm on the 3 step laser fluence (φ) at λ =568.42nm. The saturation in photoionization signal at 3

higher fluence is quite apparent. The photoionization cross-section value is obtained by fitting the saturation curve with a simple expression S = S [1- exp(-σφ)] 0

¥

¥

Fig. 6: Dependence of the photoionization (PI) signal vs laser fluence at λ =568.42 nm from the second excited level at 3

-135578.25 cm of uranium.

Fig. 8: Typical mass-resolved two-colour photoionization spectra of Sm based on the first- step transition 292.58 –

-117810.85 cm . FP: Fabry Perot etalon transmission fringes.

Fig. 7: Typical single-colour three-photon TOF mass spectrum of Sm at λ = 570.68 nm.

44


Two-color three-step excitation scheme, 0 (J=0) → -1 -1 +15650.5 cm (J=1) → 33116.8 cm (J=1) → Sm , identified

for the selective excitation of the odd isotopes of Sm, is shown in Fig.9. First laser excites the Sm atom from the

-1ground (J=0) level to 15650.5 cm (J=1) level and second laser excites it further up in the ladder to an intermediate

-1level at 33100.8 cm (J=1) before it is finally ionized by another photon from the same laser.

The basic principle of Sm isotope separation technique based on polarization selection rules is schematically shown in Fig.10 (a&b). The figure illustrates the possible excitation routes for an even isotope and an odd isotope from the ground state of Sm I for the excitation lasers linearly polarized in the same direction. The symbols J and m denote the total angular momentum quantum J

number and the magnetic quantum number respectively. As shown in the Fig. 10(a), each energy state of an even isotope has 2J+1 degenerate magnetic sublevels. The dipole transitions for the linearly polarized (π) radiations are allowed between magnetic sublevels of same mJ

except for the symmetry forbidden transition ∆J =0,

∆m =0 and m : 0↔ 0. Thus, forbidding the second step J J

transition (J = 1, m = 0 → J = 1, m = 0). Where as in case of j j

an odd isotope, due to non-zero nuclear spin I (I=7/2) there is hyperfine interaction and the transitions to the excited level E (J=1) are possible as shown in Fig. 10(b). 2

Hence, isotope-selective excitation of the odd isotopes is possible when appropriate combinations of J-values and laser polarizations are used. Using linearly polarized broad band lasers (polarization of λ || λ ) two-color three-1 2

154 144abundant isotopes Sm and Sm were recorded simultaneously by fixing the gates of the two boxcar averagers on the respective mass peaks in the mass spectrum as shown in Fig.7. Fig.8 shows a part of the mass resolved two-color spectra. The transition isotope shift, i.e., the difference in the transition energies between 154 144Sm and Sm isotopes is measured from the mass-resolved two-color photoionization spectra. The sign of

154the isotope shift value is positive if Sm line lies on the 144higher wave number side with respect to Sm. The level

isotope shifts are obtained by adding the measured second-step transition isotope shift with that of the first-step transition [36].

Polarization-based Selective Excitation of Odd-isotopes using Broad Band Lasers

It is known that in a laser isotope separation (LIS) process isotope-selective photoionization of the selected isotope is generally achieved by utilizing the isotopic shift or the difference in the hyperfine structure of different isotopes using narrow band lasers. This method puts stringent conditions on the laser bandwidths and the atomic beam parameters, especially when the isotope shifts are small and the hyperfine spectra are overlapping. An alternative isotope separation method based on polarization selection rule using broad band lasers has been previously applied for the separation of odd and even

isotopes of Zr, Gd and Yb where the isotope shifts are small and the hyperfine spectra are complex [1,37,38]. In this method, odd isotopes with non-zero nuclear spin are selectively excited while the even isotopes with zero nuclear spin are prohibited from excitation using two parallel linearly polarized laser beams. However, this method is applicable to elements having states with low J values. In this section we present the experimental demonstration of isotope-selective photoionization of atomic samarium using two-color resonance ionization polarization spectroscopy with broad band lasers.

Fig. 9: Excitation scheme for the selective photoionization of odd isotopes of Sm.

Fig. 10: Transitions between magnetic sublevels of Sm for both the excitation lasers linearly polarized in the same direction. (a) even isotopes (b) odd isotopes for the chosen scheme. Forbidden transition is labeled with a cross.

45


photon photoionization of Sm was performed for a chosen scheme and a selective photoionization of odd

147 149isotopes ( Sm and Sm) with an isotopic selectivity of more than 100 was demonstrated in the TOF mass spectrum as shown in Fig. 11.

Conclusions

In this article, we have described multi-step photoionization spectroscopy technique and many of its variants used for measuring various atomic parameters. In particular, multi-step delayed photoionization spectroscopy and its application to measure radiative lifetimes, branching ratios and transition probabilities have been described. This method has a remarkable advantage in that the lifetime, branching ratio and other related radiative quantities can be measured with the same experimental setup simply by changing the laser wavelength and the pulsing sequence. The

154 144measurements of isotope shifts between Sm and Sm of high-lying atomic levels of Sm employing two-colour RIMS has also been described. Moreover, a novel laser spectroscopic technique of isotope-selective photoionization of odd-isotopes of samarium using broadband lasers by exploiting polarization selection rules has also been described.

Acknowledgments

The author expresses his sincere thanks to his colleagues in Applied Spectroscopy Section and Spectroscopic Diagnostic & Laser Control Section of Laser & Plasma Technology Division upon whose work this article is written.

References

1. C. Haynam, B. Comaskey, J. Conway, J. Eggert, J. Glaser, Ed Ng, J.A. Paisner, R.W. Solarz and E.F.

Worden, Proc. SPIE, 1859, Laser Isotope Separation 24 (1993)

2. P.T. Greenland, Laser isotope separation, Contemporary Physics, 31, 405 (1990)

3. R. Avril, A. Petit, J. Radwan and E. Vors, Proc. SPIE, 1859, Laser Isotope Separation 38 (1993)

4. P.R.K. Rao, Curr.Sci , 85, 615(2003)

5. A.N. Tkachev and S.I. Yakovlenko, Quantum Electron 33, 581 (2003)

6. M Miyabe, M Oba and I Wakaida, J. Phys. B: At. Mol. Opt. Phys. 31, 4559(1998).

7. L.R. Carlson, J.A. Paisner, E.F. Worden, S.A. Johnson, C.A. May and R.W. Solarz, J. Opt. Soc. Am. 66, 846 (1976)

8. R.W. Solarz, C.A. May, L.R. Carlson, E.F. Worden, S.A. Johnson, J.A. Paisner, and L.J. Radziemski, Phys. Rev. A 14, 1129 (1976)

9. E.F. Worden, R.W. Solarz, J.A. Paisner, and J. G. Conway, J. Opt. Soc. Am. 68, 52 (1978)

10. A. Coste, R. Avril, P. Blanchard, J. Chatelet, D. Lambert, J. Legre, S. Liberman and J. Pinard, J. Opt. Soc. Am. 72, 103 (1982)

11. V.K. Mago, B. Lal, A.K. Ray, R. Kapoor, S.D. Sharma and P.R.K. Rao, J Phys B: At Mol Phys. 20, 6021 (1987)

12. P.N. Bajaj, K.G. Manohar, B.M. Suri, K. Dasgupta, R. Talukdar, P.K. Chakraborti and P.R.K. Rao, Appl. Phys. B 47, 55 (1988)

13. K.G. Manohar, P.N. Bajaj, B.M. Suri, R. Talukdar, K. Dasgupta, P.K. Chakraborti and P.R.K. Rao, Appl. Phys. B 48, 525 (1989)

14. E.F. Worden, L.R. Carlson, S.A. Johnson, J.A. Paisner, and R.W. Solarz, J. Opt. Soc. Am.B 10, 1998 (1993)

15. M. Miyabe, M. Oba and I. Wakaida, J Phys B: At Mol. Opt. Phys. 33, 4957 (2000)

16. T. Jayasekharan, M.A.N.Razvi and G.L. Bhale, J Opt Soc Am B 17, 1607 (2000).

17. Vas Dev, M. L. Shah, A.K. Pulhani and B.M. Suri, Appl. Phys. B 80, 587 (2005)

18. B.A. Bushaw, S. Raeder, S.L. Ziegler and K .Wendt, Spectrochim Acta Part B 62, 485 (2007)

19. M-K Oh, W. Choi, J-H Jeon, M. Lee, Y. Choi, S. Park, J-H Lee and K. An, Spectrochim. Acta, Part B 59, 1919 (2004).

Fig. 11: Selective photoionization signal of the odd isotopes of Sm.

46


20. J. Blaise and L.J. Radziemski, J. Opt. Soc. Am. 66, 644 (1976).

21. M. Miyabe, M. Oba and I. Wakaida, J. Phys. Soc. Jpn.70, 1315 (2001).

22. A.U. Seema, P.K. Mandal, Asawari D. Rath and Vas Dev, J. Quant. Spectrosc. Radiat. Transf. 145, 197 (2014).

23. C.H. Corliss and W.R. Bozman, Natl. Bur. Std. Mono. No. 53 (1962)

24. C.H. Corliss, J. Res. Natl. Bur. Std. 80A, 1 (1976)

25. R. Kapoor and G.D. Saksena, J. Opt. Soc. Am. B 6, 1623 (1989)

26. S.E. Bisson, E.F. Worden, J.G. Conway, B. Comaskey, J.A.D. Stockdale and F. Nehring, J. Opt. Soc. Am. B 8, 1545 (1991)

27. J.E. Lawler, G. Bonvallet and C. Sneden, Astrophys J. 556, 452 (2001)

28. A. Petit, R. Avril, D. L'Hermite and A. Pailloux, Phys. Scr. T100, 114 (2002)

29. P.T. Greenland, D.N. Travis and D.J.H. Wort, J. Phys. B: At. Mol. Opt. Phys. 24, 1287 (1991)

30. L.R. Carlson, S.A. Johnson, E.F. Worden, C.A.

May, R.W. Solarz and J.A. Paisner, Opt. Commun. 21, 116 (1977)

31. L.A. Hackel and M.C. Rushford, J. Opt. Soc. Am. 68, 1084 (1978)

32. P.K. Mandal, R.C. Das, A.U. Seema, A.C. Sahoo, M.L. Shah, A.K. Pulhani, K.G. Manohar and Vas Dev, Appl. Phys. B, 116, 407 (2014)

33. R.C. Das, P.K. Mandal, M.L. Shah, A.U. Seema, D.R. Rathod, Vas Dev, K.G. Manohar, B.M. Suri, J. Quant. Spectrosc. Radiat. Transfer 113, 382 (2012).

34. M. Miyabe, I. Wakaida and T. Arisawa, Z. Phys. D 39, 181 (1997)

35. S.E. Bisson, B. Comaskey and E.F. Worden, J. Opt. Soc. Am. B 12, 193 (1995)

36. H. Brand, B. Seibert, and A. Steudel, Z. Physik A - Atoms and Nuclei 296, 281 (1980).

37. N. Hideeaki, K. Takaaki, H. Yasunobu and T. Shi- geki, J. Nucl Sci. Tech. Suppl. 6, 101 (2008) and references therein.

38. G.I. Bekov, A.N. Zherikhin, V.S. Letokhov, V.I. Mishin and V.N. Fedoseev, JEPT Lett. 33, 450 (1981).

47


or single mode fiber systems. With multi mode fibers, the large core size, high numerical aperture and high collection efficiency allows low power sources such as LEDs to implement the sensor systems. Single mode fiber systems typically use a laser, specialized couplers, modulators and spectrum analyzers but provide specific advantages for interferrometric /wavelength encoding systems. They are further categorized as point measurement systems or distributed sensing systems. In distributed mode, the measuring parameter of interest is monitored at many locations along the fiber (spatial mapping). A wide range of techniques such as Intensity, wavelength encoding and polarization provide powerful sensing capabilities. In the following sections, I would describe specialized advanced sensors and optical fibers for nuclear, biological, telecommunication and industrial applications.

Fiber Optic Raman Temperature Radar (FORTR)

Many areas of temperature measurements require large area of coverage with high localization accuracy. Raman optical fiber based distributed temperature sensors are equipped with the ability of providing temperature values as a continuous function of distance along the fiber. FORTR can be employed for real-time temperature monitoring of entire pipelines for oil, gas and fire detection in power cables, bore holes, tunnels and critical installations, induction furnaces and process control Industries.

Abstract

Fiber optic sensors have been the focus of strong research and development efforts for last thirty years. On-line measurement sensors for accumulated dose in reactors, distributed temperature measurement by Raman Radar, micro-nano cavity engineered fibers/gratings for gas sensing and grating based dynamic strain sensors for health integrity monitoring provide unique enabling technologies. In this article, I would attempt to describe important features of these sensors with a focus on our own research on these systems.

Keywords: Fiber, Sensors, Nuclear, Radar

Introduction

The dramatic reduction of transmission loss in optical fibers coupled with important developments in the area of light sources (Lasers) and detectors have brought phenomenal growth of fiber optic industry. The information revolution of internet, face book, twitter, Google all owe its existence to optical fibers and fiber communication with almost zero loss, high speed, almost infinite bandwidth. Dense wavelength division multiplexing technology involving at least eight wavelengths in 1550 nm low loss window through one single mode fiber has indeed resulted in an enormous increase in available bandwidth for data transfer. The trillion dollar industry of optical fiber network has spurred the growth and advancement in all spheres of science and technology especially fiber sensors and specialized fibers.

Optical sensors and related devices which also form a backbone of integrated optical fiber network provide several advantages over traditional electrical sensors: high accuracy and performance, immunity to EMI, easy installation and resistance to harsh mechanical/chemical environments. Industry's demand for high performance optical sensors drove the race to engineer structures that are compatible with existing optoelectronics and that control and guide light. This has created niche market for some specialized needs like sensors for high nuclear environment, distributed high temperature sensing, monitoring dynamic strain in fighter jet-wings or overhead railway lines.

The sensors are based on either multimode fiber systems

Advanced Fiber Optic Sensors for Nuclear and Industrial Applications

Sanjay KherFiber Sensor Lab, SSLD, RRCAT, Indore-452013, India

E-mail: [email protected]

Fig. 1: Schematic diagram of fiber Raman sensor

48


as sensor cable has been proposed. This scheme has the potential to offer sub-centimeter spatial resolution sensor, below 1°C temperature resolution over a distance of few hundreds of meters [3].

Dynamic Strain/Force Sensors

These sensors are mainly based on technology of fiber gratings. Fiber gratings are optical components of significant interest due to their multitude of applications in communications and fiber optic sensing. The fabrication of grating in fiber or waveguide relies upon the ability to introduce periodic longitudinal variation in refractive index.

The fiber gratings are classified based upon the period of the grating. Short period gratings or fiber Bragg gratings, have a sub-micron period and act to couple light from the forward-propagating mode of the fiber to a backward, counter-propagating mode. The long period grating (LPG) has a period in the range of 100 μm to 1 mm and act to couple light from propagating core mode to co-propagating cladding modes.

The axial strain sensitivity of LPGs may be shown to be given by expression:

(1)

The basic principle of FORTR is based on optical time domain reflectometry involving time of flight measurement in conjunction with Raman scattering. The sensing fiber is coupled to short laser pulses and the backscattered optical anti-stokes (AS) and stokes (St) components are monitored. The ratio of AS signal to St Signal is used to find the temperature (Fig. 1).

To determine the unknown temperature profile with certain accuracy for complete fiber length, appropriate measures are to be devised and implemented to address several error causing issues. These issues include: the difference in theoretical and experimental values of the ratio at various temperature values, non identical fiber attenuation along the fiber length for Raman AS and St signals due to the difference in their wavelengths, noise induced into the Raman signals from various sources, and variation in amplitude of AS and St signals with time due to slow variations/drifts in laser power and laser-fiber coupling. The system is calibrated using the arrangement shown in Fig. 2.

In order to address the above issues, a discrete wavelet transform (DWT) based dynamic self-calibration and denoising technique is used and implemented. The DWT based technique is simpler, more automatic and provides a single solution to address all the above issues simultaneously. The DWT technique takes care of the difference in optical attenuation for AS and St signals by using their trend and also denoises the AS and St Signals while preserving spatial locations of peaks [see, ref. 1, 2]. Fig. 3 shows the temperature sensitive AS signal after DWT based processing for 2.5 meter oven heated length for a sensor cable which can be heated up-to 500 °C. The RRCAT, Indore has developed a field portable sensor unit with a spatial resolution of 1 meter and temperature range from 25 to 400 °C. The intelligent, user friendly software supervises the entire process of data acquisition, signal processing and display of temperature profile along the fiber. The system has been configured to produce temperature profile data, zone data and alarm data.

Recently, a distributed sensor using a superconducting nanowire single photon detector and chalcogenide fiber

Fig. 2: Experimental setup for calibration of Raman Sensor

Fig. 3: Anti-Stokes signal as recorded for various temperature o ovalues (28 C- 400 C). Change in anti-Stokes signal value at

hot zone location clearly shows the possibility of temperature measurement in standard 50/125 micron gold coated fiber.

Fig. 4: Ray optical diagram depicting coupling of light in fiber gratings and spectral response

49


the catenary is the contact force between pantograph and the high voltage (25 KV) overhead contact wire.

Austrian Research production agency and CGCRI, Kolkata (see fig. 7) are actively working to develop dynamic strain/force sensors based on packaged Bragg gratings for real-time condition monitoring of the pantographs and overhead HV line infrastructure. The technology of such sensors will also be applicable for monitoring the bridges and strain in objects near inaccessible locations.

Micro-nano Engineered Fibers for Gas Sensing Applications

Optical fiber technology is playing very crucial role in environmental and safety monitoring. The detection of methane for search of life on extra-terrestrial objects, real time leak detection of hydrogen on space rockets and launch vehicles are some of the potential applications. The availability of good quality CO laser systems and 2

focused Ion beam sources for micro-tapering, hole drilling, cavity milling, nanometer thick metal coating have opened a new field of Micro/nano fiber based sensors. In this article, I will describe a technique based on structured Photonic crystal fibers for fast response, high sensitive gas sensor.

Many gas molecules have absorption lines in the high transmission window of silica fiber, i.e. (0.8 -1.8 μm) corresponding to vibration or vibration-rotation

Where δ n =( n - n ) is the differential effective index, eff eff cl

ordinal m has been dropped for the sake of simplicity. The two terms on the right side can be divided into material (first term) and waveguide (second term) contributions. The axial strain sensitivity comprises material and waveguide effects, the material effects being the change due to transverse dimension of the fiber (Poisson effect) and strain optic effect while the waveguide effect arises from change in dispersion( dλ/ dΛ). The FBGs typically offer strain sensitivity of 1pm/ με while an LPG can offer 5-15 pm/ με depending on design optimization. The devices have an inherent self referencing capability and can be embedded in most materials.

Major applications of fiber gratings are:

1. Fiber Bragg grating network system for condition monitoring of railway Catenary-pantograph structure.

2. Strain monitoring in inaccessible areas, bridges and fighter jet wings

Recently, you must have read news in English daily: The times of India news dated Sept. 22, 2014, 'Sunday woes: CR services hit again”. The pantograph of a local train got entangled with the wire, overhead wire snapped, and paralyzing rail services for three hours delaying at least 10 long distance trains. The pantograph and catenary is seen in Fig. 5.

Fig. 5: The (asymmetrical) 'Z'-shaped pantograph of the electrical pickup. This pantograph uses a single-arm design (Source: Wikipedia).

The catenary (overhead contact line) is a vital part of the railway infrastructure (Fig 6a., b). The vertical and horizontal position as well as the tension of the overhead contact lines needs to be in certain limits that comply with international standards. For the moving electric train, that establishes contact to the catenary via his pantograph, this ensures minimal losses, limited wear, tear and a reduced risk of disruption of the current transmission to the train engines power unit. Irregularities need to be detected in time to avoid severe damage. A measure for the quality of

Fig. 6: Schematic of pantograph, catenary and possible locations of sensors.

Fig. 7: Extended view of pantograph (turned upside down) with possible locations for strain sensors.

50


For a practical sensor the response time should be smaller than one minute. To achieve a higher sensitivity with fast response time, recently, techniques have been developed for drilling periodic transverse micro-holes (Openings) along the length of the fiber [Fig.10 (a, b)]. The ArF laser at 193 nm, fs laser, CO laser and Focused Ion Beam 2

(FIB) systems have been used for V grooving, hole drilling and profiling the PCF fibers. A number of precision holes, V-grooves are drilled transversely through the silica/air cladding with a depth up-to the hollow core of HC-PCF so that the gas can diffuse freely without any necessity of pressure filling.

Other useful PCF fibers of type SCFs have a wavelength scale solid silica core supported by thin filaments connected to our cladding of few large size air-holes. Fabrication of three micro-channels by use of fs infra-red laser has been demonstrated with full access to evanescent field [6]. RRCAT has demonstrated the capability of micro-machining in PCF fibers using precision focused CO laser as shown in Fig. 11[8].2

Fig. 11: The Micrograph of V-Grove fabricated in PCF fiber

A fast response Methane sensor with multiple transverse Micro-channels has been reported. The sensor is 7 cm long with 7 micro-channels in a PCF fiber with a diffusion response time of three seconds [7].

Fiber Grating Based Gamma Radiation dose Sensors (Nuclear, Space and Medical Applications)

The presence of ionizing radiation makes nuclear facilities extremely hazardous, not only for human operators, but also for all kinds of devices and systems.

-4Total gamma radiation dose level may span from 10 Gy 7for personal dosimetry to 10 Gy near fusion reactor core.

1 Gy = 1 Gray, which is S.I. unit for absorbed dose. 1 Gy corresponds to 1 Joule of energy absorbed per kilogram of material. 1 Gy corresponds to 100 rad. In ITER, during the plasma burn, the diagnostic systems installed inside the vacuum vessel will have to withstand a total dose

9approaching 10 Gy.

transitions. The strength of gas absorption lines is used to perform quantitative measurement of gas concentration. Photonic crystal fiber (PCF) or holey fiber is a novel class of optical fibers that include hollow-core photonic band gap fibers (HC-PBFs) and suspended core fibers (SCFs). The cross section micrograph of HC-PCF is shown in Fig. 8. The central core is relatively large and hollow. The light in this fiber is guided through Photonic band-gap effect.

Fig. 8: Micrograph of NKT PCF HC-1550-02 [Google open source through NKT Photonics]

The core and cladding holes of hollow core fibers can be readily filled with gases. Since the optical mode is guided, the interaction length between the optical field and gas in the core is limited by the fiber loss. This fiber has pass band of 1490-1680 nm and is suitable for detection of gases like Methane (1667nm), CO 2

(1573nm), CO (1567nm). Such fibers allow confinement of an optical mode and gas phase materials simultaneously within the hollow core. They can be coiled to very small diameters (<1 cm) for high sensitivity point detection. The laser based gas sensing system is shown in Fig. 9. However, the simple configuration of Fig. 9 offers long response time of few minutes as normal diffusion of gas in micro-holes from single open end is a slow process.

Fig. 10 (b): Sensor PCF Fiber with periodic openings of size l for use in Fig. 10(a)

Fig. 10 (a): Fiber Optic Gas sensor system with transverse micro-structure

Fig. 9: Schematic of PCF fiber based Fiber Optic Gas sensor system

51


has designed, fabricated and packaged such sensors [see fig. 14] for real-time gamma dose sensing applications.

Fig. 14: Packaged Gamma dose sensor developed at RRCAT, Indore

The parameters of the packaged sensor are as follows:

• Gamma dose range:1- 10 kGy : Resolution: 0.4 kGy

Sensitivity: 0.53 nm/ kGy10-1000 kGy: Maximum error; ± 15%

Conclusion

Distributed Raman backscatter, micro-channeling and in-fiber gratings constitute three versatile platforms for realizing a variety of advanced fiber optic sensors. In this article, I have provided a glimpse of such sensors for nuclear and industrial applications. The treatment is by no means exhaustive, as the intention has been to give a summary of some of the key developments in the area.

References

1. Manoj Saxena, R. Arya, Sanjay Kher et al, Measurement, 47, 345(2014).

2. Manoj Saxena, R.B. Pachori, Sanjay Kher et al., Opt. Laser Technol. 65, 14 (2015) .

3. Gabriele Bolognini, Arthur Hartog, Optical fiber technology 19,678 (2013).

Interaction of ionizing radiation with optical fibers and gratings leads to several physical processes that can be used for radiation dosimetry. Increase of attenuation (optical density), luminescence, optically stimulated luminescence (OSL), thermo luminescence, scattering and radiation induced index change have been used to design dose sensors for dose ranges up-to 100 kGy. The attenuation based sensors based on specialty doped fibers reach saturation level above 10 kGy.

The wavelength encoded sensors based on fiber gratings can solve many measurement problems such as radiation induced broadband transmission loss in optical fibers, source fluctuation etc. Most Bragg grating based sensors, reported till date, are either less sensitive or reach saturation level near 50-150 kGy depending on composition and grating writing techniques Therefore, commercially available fiber optic dose sensors seem unreachable for very high level dose of 100 KGy- 10 MGy. We have developed a one-of-a-kind long period fiber grating (LPFG) sensor for sensing high level

2gamma dose up-to 1 MGy with a high dynamic range (10 6Gy to 10 Gy).

Specialty long period fiber gratings were fabricated using CO laser (shown schematically in Fig.12). This method 2

is adaptable for writing LPGs in wide range of single mode fibers.

Fig. 12: Experimental set up for LPG inscription

Our studies based on in-situ radiation dose measurements suggest that long period gratings in Boron doped fiber can indeed be used for dose level up-to 1 MGy [8, 9 and 10]. We have experimentally studied in-situ the effect of high gamma dose on long period fiber grating (LPFG) parameters such as resonance dip wavelength shifts, amplitude and width of bands, room temperature annealing and effect of mode orders.

It has been observed that there is a monotonous increase in resonance wavelength of the grating upto total dose of 1MGy (Fig. 13). This shift is calibrated for sensing the radiation dose. It has been shown that the grating modulation (Dip strength) and widths are not affected remarkably due to high level of gamma dose and hence devices based on this scheme are useful for dose measurement applications at critical locations. RRCAT

Fig. 13: Real time gamma induced spectral transmission spectra of LPG

52


9. Kher S.; Chaubey S.; Kashyap R.; Oak S.M.; Turnaround-Point long period fiber gratings (TAP-LPGs) as high radiation dose sensors, IEEE Photon. Techn. Lett. 24, 742 (2012).

10. Kher S, Chaubey S., Oak S.M., J. Instrumentation Science & Technology, Taylor & Francis, 41,135 (2013).

4. Richard wagner, Dietnor Makz et al., proceedings thof EWSHM-7 European workshop on structural

health monitoring (2014).

5. Somnath Bandopadyaya, CGCRI Progress report 2014.

6. A. van Brakel, C. Grivas et al., Opt.Express, 15 , 873 (2007).

7. Y.L.Hoo, S.J. Liu et al., IEEE Photonics Technol. Lett. 22,296 (2010).

8. Sanjay Kher, Manoj Saxena, Smita Chaubey, S.M.Oak., Sensors & Transducers Journal, 116,112(2010).

53


two approaches namely Coherent Diffractive Imaging (CDI) and XUV holography. We will discuss these techniques in following sections.

Coherent Diffractive Imaging (CDI)

In the case of Coherent Diffractive Imaging (CDI) first the diffraction pattern generated by the sample is captured. If one knows the phase and intensity distributions in the Fourier space one can reconstruct the resulting image simply by its Fourier transform (FT). However the diffraction pattern only gives the intensity in the Fourier space and not the phase. The phase problem is then solved by Gerchberg-Saxton algorithm [4]. In order to implement this algorithm one needs to have some information about the object, it is known as the space domain constraint of the image. For example the triangle shape in Fig. 1 can be used as space domain constraint. The constraint of the image can also be determined from its autocorrelation because the image must lie totally inside its autocorrelation and can be iteratively adjusted to match the real object known, such algorithm is known as shrink wrap algorithm [5]. This algorithm does not require the initial knowledge of the object but the convergence of the algorithm can be slower. Hence the plus point of CDI is that is requires low flux, on the flip side is that either one should know the image constraint a priori or one must predict the image constraint dynamically during the reconstruction, which makes the image reconstruction highly algorithm dependent.

XUV Holography

In contrast with CDI the XUV holography can reconstruct the input image directly from its FT. The basic principle behind this technique is that if we have some

Introduction

Imaging of an object with simultaneous high temporal and spatial resolution is an challenging area of research. Time resolved x-ray Diffraction (TXRD) utilizing ultrafast x-ray source [1] and x-ray imaging from quasi continuous synchrotron sources can give high temporal or spatial resolution separately. In order to combine the advantage of both areas we need an ultrashort high fluence source. This source combined with high throughput x-ray optics and sensitive, low noise detectors will facilitate the realization of ultrafast x ray microscopy studies. Moreover since the x-ray spectral range do not usually facilitate the refractive /reflective optics hence good spatio-temporal coherence is another essential requirement of illuminating source. Since high order harmonics (HOH) generated by the interaction of ultrashort laser pulses with the low density gas/plasma plume targets inherit the coherence and temporal profile of the parent laser, they are ideal candidate to serve as the illuminating source for the imaging applications in extreme ultraviolet (XUV) spectral range [2]. The basic requirement towards the single shot microscopy using HOH is to increase the flux of the harmonics. In this paper we present our study on the direct single shot imaging using HOH. We first discuss various techniques of ultrafast XUV imaging. Then we will present our results on ultrafast single shot direct imaging using HOH [3].

Ultrafast XUV Imaging Techniques

There are two main approaches of XUV imaging using coherent quasi-monochromatic radiation known as direct imaging and indirect imaging. In direct imaging scheme a diffractive focusing element such as zone plate is used to focus the light scattered by the sample, whereas in indirect imaging schemes the diffraction pattern of the XUV light scattered by the object is used to reconstruct the image of the object. In both the cases the XUV radiation is first loosely focussed on the object to increase the fluence of scattering signal. Moreover the object is prepared by special techniques such that it has high contrast between transparent and opaque areas at the desired wavelength. The indirect imaging uses mainly

Ultrafast Single Shot Coherent XUV Imaging using High Order Harmonics

1* 2 3 2,4H. Singhal , K.H. Lee , S.B. Park and C.H. Nam

1Raja Ramanna centre for Advanced Technology Indore 452013 India,2Center for Relativistic Laser Science, Institute for Basic Science (IBS), Gwangju 500-712, Korea,

3Korea Advanced Institute of Science and Technology4Advanced Photonics Research Institute, GIST, Gwangju 500-712, Korea,


Fig. 1: Phase retrieval from the modulus of Fourier transform (FT). a) input image b) reconstructed image from the modulus of FT after 100 iterations, c) after 1000 iterations.

54


but results in larger blurring. The intensity of CC with dot is much smaller than the line hence the line is usually taken as the reference in such experiment. We have also de-convoluted the CC image from our knowledge of complete FT (intensity and phase) of the CC and the reference, same is shown in Fig. 2 e). Finally the line cut along the red marks shown in Fig 2d) and 2e) is shown in Fig 2f). We can see a blurring in the case of images obtained by differentiative transform, whereas the images obtained from deconvolution suffers from high frequency noise. If one takes a sharper reference the blurring can be suppressed but the intensity of the cross correlation signal will also go down. Hence, the size of the reference will serve as a tread off between the signal strength and resolution.

Ultrafast Single Shot Direct Imaging

Fig. 3: Experimental setup for single shot direct imaging using th38 harmonic order of Ti:Sapphire laser

We have experimentally realized the ultrafast single shot direct imaging using high order harmonic radiation. The

thharmonic radiation used in this experiment was 38 order of the Ti_sapphire laser (~21 nm). The experimental setup is shown in Fig. 3. Efficient harmonic radiation was generated in a 6 mm long plume of helium gas jet using two color Ti:sapphire laser pulses. The flux obtained in the present case is similar to the one reported in ref. 2. The noise of the visible light is decoupled efficiently from the XUV imaging system by placing two setups (i. e. harmonic generation and XUV microscope) in different vacuum chambers isolated by an 150 nm thick aluminium XUV filter. We have used a zone plate to image the sample on a x ray CCD. The advantage of direct imaging is the real time production of images. The drawback of this technique is 1) the x ray flux is blocked by the zones of the zone plate and 2) the resolution is similar with the width of outer most zone of the zone plate. To overcome these problems we have made a phase reversal zone plate (PRZP) using electron beam lithography. The width of outermost zone in this zone plate is ~100 nm and the

known reference with the unknown object, and obtain the diffraction pattern by uniformly illuminating the sample, then the inverse FT of the diffraction pattern will contain 1. Autocorrelation of the sample 2. Autocorrelation of the reference and 3. Cross correlation (CC) of the sample and reference. If the autocorrelations and CC images are separated spatially then we can remove autocorrelation signals by applying suitable transform on the CC images,We retrieve the input image. If the reference is a point then the CC is convolution of point with image and if point is sufficiently small the CC is same as input image, if the reference is a line then the CC is convolution along the line direction and the differentiation along that direction will result in the input image. It is pertinent to note here that separation of line and image should be larger than the sum of their extents in that direction . In this case the autocorrelation and cross correlation images are spatially separated and can be separated easily. However in this case the source should illuminate a wider sample uniformly, which means a loose focusing on the target and in turn a decreased diffraction efficiency. We have simulated the XUV holography process for the demonstrational purposes and same is depicted in Figure 2. Figure 2 a) shows the input image with reference, Fig. 2 b) shows the diffraction pattern intensity and Fig. 2c) shows the filtered CC signal. One can see that there are two CC pattern both side of centre. The differentiation of this CC gives four images of the input sample but with

0opposite phase (hence rotated 180 ). For these images to be spatially separated the reference line should be equal to or longer than the height of the sample. Moreover a certain amount of blurring is introduced by the reference image depending upon its size. The intensity of the cross correlation signal also depend on the size of the reference image hence larger reference results in more intense CC

Fig. 2: Reconstruction of image from XUV holography. a) input image with reference, b) diffraction pattern c) Cross correlation signal d) reconstruction by differentiating cross correlation signal along vertical direction e) Reconstruction by deconvolution of the image from knwon reference f) the line profile aling the shown portions of d) and e) showing the blurring introduced by the thickness of the reference line.

55


Summary

In summary we have discussed various aspects of XUV imaging. We present our simulation results on coherent diffraction imaging and XUV holography. Here we present first experimental realization of single shot

thultrafast shot direct imaging. The microscope use 38 harmonic order of Ti:sapphire laser as illumination source and generate single shot images with a spatial resolution of <140 nm, which is very close to the theoretical resoultion of the zone plate (~100 nm) in the present case.

References

1. V. Arora, S. Bagchi, M. Gupta, J. A. Chakera, A. Gupta, P. A. Naik, P. Chaddah, and P. D. Gupta, J. Appl. Phys. 114, 023302 (2013).

2. I. J. Kim, C. M. Kim, H. T. Kim, G. H. Lee, Y. S. Lee, J. Y. Park, D. J. Cho, and C. H. Nam, Phys. Rev. Lett. 94, 243901 (2005).

3. K. H. Lee, S. B. Park, H. Singhal, and C. H. Nam, Opt. Lett. 38, 1253 (2013).

4. J. R. Fienup, Appl. Opt. 21, 2758 (1982).

5. S. Marchesini, H. He, H. N. Chapman, S. P. Hau-Riege, A. Noy, M. R. Howells, U. Weierstall, and J. C. H. Spence, Phys. Rev. B, 68 140101 (2003).

zones plate consists of ~400 zones. With the efficient decoupling of visible signal from XUV radiation we were able to decrease the losses of XUV radiation. An aperture is installed in the path of XUV pulse to further reduce the optical noise (since the divergence of harmonic beam is much smaller than the divergence of laser pulse). A pair of flat-concave multilayer mirrors is used to focus the XUV radiation on the test sample of grating pattern. The total reflectivity of this mirror pair is ~12%. By adopting these techniques and using low noise deep cooled CCD we have increased the sensitivity of the microscope high enough to capture the single shot images from HOH [3]. Figure 4 shows the image of the test pattern of the 1 μm half period grating for a) three shots and b) single shot. One can see that the image in the case of single shot exposure is quite clear and sharp. The single shot imaging eliminate all the errors due to beam pointing stability.

Fig. 4: Images of 1 μm half period grating for a) three shots and b) single shot. The inset shows the SEM image of the grating.

Figure 5 shows the different grating patterns of half period a) 100 nm half period and b) 200 nm half period imaged by the microscope. One can see that grating pattern shown in Figure 2 a) is unresolved whereas in 2 b) the grating pattern is clearly resolved. The resolution of the microscope for single shot imaging is estimated to be ~140 nm using 10-90 Rayleigh criterion. This estimated resolution is very close to the theoretical limit of resolution of the zone plate ( ~100 nm in the present case) which indicate high degree of spatial coherence and low bandwidth of the harmonic radiation.

Fig. 5: Images of a) 100 μm half period grating (unresolved) and b) 200 nm half period grating (resolved). The inset shows the SEM image of the gratings.

56


necessary in industry and medical instrumentation.

Lasers near 2 μm are of significance due to their applicability in lidar, gas sensing, spectroscopy and medical applications. thulium (Tm)-doped fiber lasers, which can emit in this wavelength regime, are attractive for medical surgery owing to the strong absorption of water at 1.94 μm. For tissue surgery, the laser interaction with the tissue due to absorption is of interest, which leads to heat generation. In the case of soft tissues which have >60% water, Tm-doped fiber lasers are a potential candidate for the purpose of surgery. The broad fluorescence spectra of the Thulium (Tm) ion in silica glass allows the fabrication of a widely tunable fiber laser with tuning range from 1.7 to 2.1 μm, thus providing an excellent overlap with the strong absorption lines of targeted gases such as CO , CH . under pumping at 1600 2 4

nm and 800 nm [4].

At CSIR-CGCRI, Kolkata, a state of the art facility has been established for making RE-doped optical fiber/preforms for high power laser application through the vapor phase deposition technique in addition to the conventional solution doping method [5]. A series of Yb doped fiber with double clad configuration has been fabricated and utilized to design high power fiber laser at around 1 µm for industrial application. Additionally work is in progress to develop Tm doped fber laser at around 2 µm for environmental gas sensing and medical application. The present paper gives an overview of the activities being pursued in this direction to build-up indigenous capability.

Fabrication of RE–doped Optical Fiber:

Several methods have been followed to incorporate RE ions into the silica glass structure, among them solution doping method is the most popular due to its simplicity and versatility. However, it suffers from poor repeatability, low dopant uniformity and reached its limit regarding increase of core diameter which is essential for high power laser fibers. The search for an alternate process led to the development of vapor phase doping technique involving supply of the RE and Al precursor vapors at high temperature and simultaneous deposition with silica. In this process, the chances of clustering and

Introduction:

Fiber laser technology has shown dramatic progress in the recent past. The advantages of fiber laser – compactness, high average power, excellent beam quality, easy handling capability, high efficiency and low maintenance cost – have allowed them to be superior against conventional solid–state lasers. As a result, they are now leading contenders for a wide range of applications covering industrial, strategic and medical fields requiring powers from a few watts to several kilowatts. The market demand of fiber laser worldwide is predicted to touch $1.6 billion in 2015, capturing 30% of the total market share (Source: Strategies Unlimited).

The key part of a laser module is the rare earth (RE)-doped active fiber. Rapid advances in raising the output power of continuous wave (CW) fiber lasers to the multikilowatt level [1] have required novel active fiber design strategies. It is necessary to suppress nonlinear scattering processes and optical damage while maintaining good beam quality to achieve higher output power from a single optical fiber. Thus, efficient operation of fiber laser systems demands the use of RE-doped fibers with large mode area (LMA) structure, high RE concentration, uniform dopant distribution and good optical properties with low attenuation [2].

For lasing in the 1 μm window, Ytterbium (Yb)-doped fibers are used in view of their high efficiency, wide pump absorption band, broad gain-bandwidth and high power scaling capability and is fast emerging as a promising candidate for high-power all-fiber laser system which can compete with the solid-state laser systems in terms of efficiency and output power and at the same time, can offer the attractive advantages of a fiber-based system. The availability of low-cost and high-brightness fiber-coupled semiconductor pump diodes that emit light in the 0.9–1.0 μm absorption band of Yb-doped double clad optical fiber, is making the use of fiber lasers even more attractive. The Yb-doped fiber laser has the potential to generate pulses as short as 4.5 fs with peak power of 100 kW [3] and the continuous wave (CW) power has been scaled well beyond 10 kW level [2]. Such high brightness fiber has potential application in material processing, metal cutting, solar cell scribing, welding and marking

Fiber Laser Technology– Current Status and Activities by CSIR-CGCRI

Ranjan Sen*, Mrinmay Pal, Atasi Pal, Anirban Dhar, Maitreyee Saha, Sourav Das Chowdhury, Nishant Kumar Shekhar, Debasis Pal and Aditi Ghosh CSIR–Central Glass & Ceramic Research Institute, 196 Raja S. C. Mullick Road, Kolkata–700032


57


phase separation are reduced significantly. CSIR-CGCRI established a state of the art facility for vapour phase deposition and has successfully demonstrated the process of fabricating large core, low refractive index and uniformly RE–doped preforms in reliable manner [5].

Refractive index profile (RIP) of the preforms was measured by using preform analyzer (PK2600). The numerical aperture (NA) of the preforms is in the range of 0.08–0.12. The distribution of Yb O and Al O in a 2 3 2 3

preform as detected by electron microprobe analysis (EMA) is shown in Fig.1. The EMA curves show negligible center dip, usually formed due to the evaporation of dopants during collapsing of the preform. The doping levels of the core material are in the range of 0.1–0.75 mol% Yb O and 1.0–2.5 mol% Al O . It has 2 3 2 3

been observed that the variation of dopants concentrations along the longitudinal direction is negligible up to the preform length of 42 cm. The typical RIP of the fiber is shown in Fig. 2.

The preforms were shaped by grinding to double D-,

hexagonal and octagonal shape to enhance the pump absorption of the core through deviation of the circular geometry of the pump cladding in order to avoid helical modes that do not pass through the core. The low index acrylate-coated fiber was drawn employing a typical fiber drawing tower by heating the preform in a furnace at a temperature of around 2000 °C.

The fibers exhibited low Photo Darkening Induced Loss (PDIL) and proficiently delivered 105 W of CW power at 1060 nm emission wavelength (in free space resonator configuration). The laser was in operation for three hours continuously without any degradation in output. Further, it was also tested for more than one month regularly for its performance. It did not show any degradation over this period.

In order to develop fiber laser at around 2 µm, a set of both Tm and Tm/Yb-doped single-mode optical fiber in single-clad as well as double clad configuration with different host compositions, Tm-ion concentrations and proportions of Yb to Tm were also designed and fabricated by using the MCVD coupled with solution doping method [6] to use as gain medium for Tm doped fiber laser.

Yterbium Doped Fiber Laser at 1 µm:

Laser operation in CW mode:

An experimental set-up (Fig. 3) has been built-up for measurement of fiber laser characteristics in 'all-fiber' configuration. Multiple numbers of single emitter multimode pump laser diodes of pump power 10 watt at 975 nm were combined through a pump combiner to excite the double-clad Yb-doped fiber. Each pump laser diode was placed on a proper heat-sink. One high reflector (99%) FBG of bandwidth 3 nm and another low reflector (20%) FBG of bandwidth 1 nm having central wavelength at 1064 nm were spliced at both ends of the Yb-fiber to make a laser cavity. Since the Yb-fiber has special waveguide design (D-shaped, hexagonal), care was taken for proper splicing, particularly the perfect cleaving of double D-shaped Yb fiber. Splicing between the Yb-fiber and FBG is very crucial part and it was done properly by matching the waveguide parameters of both the fibers. The Yb-fiber length was optimized to get the maximum laser power at available pump power.

Fig. 1: EMA results of Yb O and Al O distribution in the core 2 3 2 3

of a preform

Fig. 2: Refractive index profile of a Yb-doped fiber Fig. 3: All-fiber based laser characterization

58


Laser operation in actively Q-switched Pulsed mode:

In actively Q-switched lasers, the pulse energy and duration vary with repetition rate because of both gain recovery variation and thermal effects. Shortening of the opening-time of the Q-switch leads clipping of the pulse trailing peaks. Simultaneous fine setting of the pump power, open-time and rise-time of the Q-switch allowed generating 'clean pulse'. The dynamics of these lasers involves several time constants which are all of the same order of magnitude: round-trip time, pulse duration and rise-time of the Q-switch.

A 0.5 m long matched single clad passive fiber was spliced to the output FBG fiber to deliver laser output power. The spliced region was covered by high-index gel. Critical study was done for dissipation of accumulated pump power in the fiber which otherwise creates unwanted issues like fiber fuse, polymer resin burning, splicing damage and destroy the components. Finally, laser output power was collected through the isolator to prevent the back reflection. Detailed experimental study has been done to get the maximum lasing efficiency with stable laser output power.

The CW output power achieved so far is 21W at 1064 nm in 6 m length of a double D shapedYb doped fiber having core diameter of 11 µm, core NA of 0.09, clad diameter of 125 µm and clad NA of 0.46 which was pumped by multimode 975 nm pump diodes. The spectrum of laser and the variation of laser power with pump power is shown in Fig. 4 and 5 respectively.

The intensity profile at 1064 nm was Gaussian in nature 2and measured M value was less than 1.4 which indicates

almost single-mode nature of output laser (Fig. 6).

Fig. 4: Lasing emission @ 1064nm

Fig. 5: Lasing efficiency

Fig. 6: Intensity profile of the laser beam

Fig. 8: 100 ns Q-switched pulse

Fig. 7: Schematic of Q-switched MOPA

59


In addition, a laser cavity under multimode pumping at 808 nm has been designed by utilizing Tm-doped active fiber in double clad configuration. In this configuration, six numbers of laser diodes are used through a pump combiner to scale up the laser power. Laser output power of 3.3 W at 1.95 µm (Fig. 12) has been developed in this configuration. The slope efficiency is 23.4% (Fig.13).

As shown in the Fig.8, the setup basically comprises two distinct parts, one is seed source and other is power amplifier which is referred as “Master Oscillator Power Amplifier (MOPA)” configuration.

The seed source consists of a 4 m long Yb doped double clad fiber (core diameter 11 µm, core NA of 0.09, clad diameter of 125 µm) which was pumped by a single multimode laser diode at 975 nm via a (2+1):1 combiner. One end of the Yb-fiber is spliced with a high reflecting FBG of reflectivity 99.9% with central wavelength of 1064 nm. The output coupler FBG has reflectivity of 50%. Active Q-switching is done by placing a fiber-coupled acousto-optic modulator (AOM, operating wavelength: 1030 to 1090 nm) inside the cavity which modulated the cavity loss. The total cavity length was about 10 m which corresponds to the cavity round-trip time of the order of 100ns. The pulse input to the AOM is controlled by a RF function generator via an AOM modulator. The signal from the seed is fed to the power amplifier through an isolator which was used as a preventive measure for back-reflections. The power amplifier comprises of a 8 m long octagonal Yb doped double clad fiber having core diameter 15 µm, core NA 0.08 that was pumped by two multimode laser diodes at 975 nm via a (2+1):1 combiner. The unused pump dumping is achieved by splicing a passive single-clad fiber. The output power is measured in a power meter and the time domain signal is analysed in an oscilloscope (Tektronix -DPO7254C). Q-switched pulse of width 100 ns (shown in Fig. 8) with peak power at about 1.9 kW at 1064 nm was generated.

Thulium Doped Fiber Laser at 2 µm:

An all-fiber Tm doped fiber laser with an output power of 11 mW has been designed to tune centering wavelength of 1.997 µm. The laser has been used successfully for the detection of CO gas. The schematic of the laser based 2

sensor system is shown in Fig. 9. The spectrum of the tunable laser is shown in Fig. 10.

The designed Tm doped fiber laser in single clad configuration is effective in offering a stable means of detection of CO using its optical spectrum and at a 2

minimum detectable concentration level of 345 ppm. The spectrum of the tunable laser is illustrated in Fig. 11 when the chamber was under 1.5 bar CO and 2 bar N . The 2 2

system designed laser is superior in terms of the selection of the operating wavelength and its fit to a specific absorption line of the gas showing a stronger absorption cross-section, in a particular longer wavelength absorption band. The detected CO spectrum in the range 2

of 1.998 μm has been shown to be comparable with the spectral features available from the HITRAN data base.

Fig. 9: Schematic of the gas sensor system using the Tm-doped tunable fibre laser (HR: High reflective FBG; LR: low reflective FBG; BBS: broadband source)

Fig. 10: 3-D trace of normalized laser spectrum (at ~1.997 µm) as a function of negative strain or compression.

Fig. 11: Intensity variation of the tunable Tm-doped fiber laser exposed in the gas cell

60


Conclusion

The paper presents the review of research work on fiber laser carried out at CSIR-CGCRI. Starting from preform fabrication, Yb and Tm doped double clad fibers of specified design and compositions have been fabricated successfully. The vapour phase doping technique has been demonstrated in a reliable manner for making large core preforms/fibers, suitable for high power laser application. The fibers exhibited good lasing performance and are therefore being utilized to build-up 'all-fiber' lasers-both CW and pulsed at 1 µm and 2 µm regions. Work is in progress to make compact laser system at different power level targeting specific applications with the aim of developing the technology indigenously.

References:

1. Richardson D J, Nilsson J and Clarkson W A J. Opt. Soc. Am. B 27, B63–92 (2010).

2. Limpert J, Roser F, Klingebiel S, Schreiber T, Wirth C, Peschel T, Eberhardt R and Tunnermann A IEEE J. Sel. Top. Quantum Electron. 13, 537(1995).

3. Andres M V, Cruz J L, Diez A, Perez-Millan P and Delgado-Pinar M Laser Phys. Lett. 5, 93(2008)

4. R M Percival, D Szebesta, C P Seltzer, S D Perrin, S T Davey and M Louka, IEEE J. Quantum Electron. 31, 489 (1995).

5. Saha M, Pal A and Sen R Photonics Technol. Lett. 26, 58 (2014)

6. Atasi Pal, Anirban Dhar, Shyamal Das, Shu Ying Chen, Tong Sun, Ranjan Sen, Kenneth T V Grattan Opt. Express 18, 5068(2010).

The signal power can be increased with further availability of pump power as the output did not exhibit a roll-over at the current pump powers. Such laser has potential importance in soft tissue surgery.

Fig. 12: Spectrum of laser at 1.95 µm

Fig. 13: Variation of laser power with pump power

61


from the resonant to the non-resonant species through collisions reduces. Unlike the case of MLIS following the multi-photon dissociation route, the requirement on the intensity of the laser source is much less stringent here, thereby making this process accessible to even cw laser sources.

CO laser that operates in the mid infrared region and 2

tunable over ~9 to 11 µm region, albeit discretely, is the most attractive choice to impart vibrational excitation to the rarer isotopic species either directly or by way of converting its emission into other wavelengths through non-linear or optical pumping routes. The discrete tunability of the parent laser, however, comes in the way of resonant excitation as the possibility of exact coincidence between the absorption centre of the resonant isotopic species and the emission centre of the coherent source is extremely rare. The first demonstration of dramatic enhancement of enrichment factor on combining laser with aerodynamic separation, however, made use of a discretely tunable CO laser that 2

fortunately had reasonable overlap with the absorption centre of appropriate SF isotope, the working molecule.6

124Among other isotopes, separation of Sn in its molecular 124form received our attention. Importance of Sn as

detector element in neutrino-less double beta decay experiment is well recognized for its better sensitivity. The organo-tin compounds with their Sn-H stretching frequency (fundamental mode) lying in the region of 3

-11800-1900 cm have good absorption in the 5 µm spectral region that can be reached by frequency doubling of the emission of a CO laser and thus appeared to be the 2

suitable working molecule. Continuous tunability of 5 µm radiation generated this way will be automatic if the emission of the CO laser can be made continuously 2

tunable. The dark region in the emission feature of a CO 2

laser can be reduced by increasing its operating pressure that, in turn, increases the broadening of the gain. Operation at a pressure of ~10 atmosphere results in a broadening of the gains on the neighbouring ro-vibrational transitions to the extent that they merge with each other providing the prospect of a near continuous

The first laser that flashed to life in 1960, fired the imagination of scientists in many countries towards using it to shrink the size of the usually huge industrial enrichment plants. The idea was to exploit the extraordinary purity of laser light to selectively excite the rarer of the isotopes. The resulting agitation should ease the identification of this precious isotope thereby facilitating its separation. The approach based on this simple concept although has been extensively researched over last many decades but failed to establish its credentials beyond laboratory experimentation. The other approach that has gained momentum, in particular over the last decade or so, exploits the purity and brightness of laser light by combining it with an existing separation scheme that take advantage of the difference of one or another physical properties of the isotopes. To this end the laser assisted aerodynamic isotope separation scheme has established its clear advantage over other methods that work on the similar line.

It is well known that when a gaseous mixture expands through an orifice, the heavier species tends to concentrate near the centre of the expanding gas mixture while the lighter one moves outwardly. This thus presents a possibility of separating the heavier and lighter species. The spatial separation of the species understandably depends on their relative mass difference. The conventional aerodynamic separation that is based on this principle can thus be very effective in separating lighter isotopes e.g., hydrogen and deuterium. The enrichment factor that is achievable under this scheme for heavier isotopes like uranium, on the other hand, is too small to have any practical utility. This drawback can be overcome by shining the expanding jet with an appropriate coherent beam of light. As the gas cools, the molecules in the jet forms van der Waal's clusters. Selective laser excitation of one of the isotopes prevents it from forming clusters thereby giving rise to a larger mass difference between the excited isotope and the non resonant isotope that forms clusters, resulting in more efficient spatial separation of the isotopes. Diluting the molecular isotopic mixture with a buffer gas like Ar further enhances the overall separation efficiency as now the possibility of transferring the vibrational excitation

Development of Continuously Tunable Mid-infrared Source and its Application in

Laser Assisted Aerodynamic Isotope Separation

D.J. Biswas, J.P. Nilaya, M.B. Sai Prasad, S. Daga,G. Chakraborty, Ayan Ghosh, R.C. Das, A. Tak and A.K. Nayak

Laser & Plasma Technology Division, Bhabha Atomic Research Centre, Mumbai – 400 085

62


tunable coherent source over 9 to ~11 µm. By making use of a commercial multi-atmosphere CO laser and with a 2

judicious interplay of the operation voltage and the Q of the resonator, we have achieved near continuous tunability. When this laser is used as a pump source for extending the wavelength to other regions of mid-infrared region following nonlinear routes, its continuous tunability gets automatically translated to the generated coherent radiation source. We have developed a continuously tunable 5 µm coherent radiation by SH conversion of the emission of the multi-atmosphere TE CO laser in an AgGaSe crystal. The continuous 2 2

tunability of the SH has been demonstrated over most of the frequency doubled spectral region of the pump laser. A very significant improvement in the SH conversion efficiency could be achieved by gainful exploitation of an unstable cavity that was integrated with the stable pump cavity. This integration allowed transport of the pump radiation into the external cavity that at the same time exhibited high 'Q', prerequisite for multi-pass of the unabsorbed pump through the crystal. The unstable cavity not only ensured longer interaction length in the crystal without exposing it to high optical flux thereby safeguarding the crystal even in the pulsed operation, but also eliminated the feedback problem. Further, it was experimentally found that a significant advantage in second harmonic generation could be achieved when the nonlinear crystal is illuminated with alternate high and low regions of intensity along its length as against the uniform illumination case maintaining the same average intensity. The advantage is attributed to the square dependence of the generated second harmonic on the intensity of the pump.

The synthesis and purification of the working molecule, Dimethyl Stannane (DMS), was carried out in house and the purity of the synthesized compound ascertained by NMR studies. In absence of any spectroscopic data on DMS in the literature, it was imperative to carry out its spectroscopic studies as a precursor to applying laser assisted aerodynamic separation scheme to separate tin isotopes. As DMS is both heavy and polyatomic, many low lying vibrational levels, both pure and combination modes, are populated at the ambient temperature. As a matter of fact, only ~3% of the population is actually available at the ground vibrational state at room temperature. Spectroscopic studies towards gaining information of isotope specific vibrational-rotational spectral information is possible only if the gas is appropriately cooled to ensure that significant fraction of population is arrested in the ground vibrational state. However, the reduced vapour pressure at the required low temperatures renders spectroscopic measurements impractical in static cooled conditions. The absorption experiments in DMS were therefore carried out under dynamically cooled conditions by gainfully employing a nozzle based expansion scheme. The absorption in the free jet of DMS-He at various downstream distances was measured by employing the continuously tunable 5.4 µm laser. At certain downstream location, the dynamic cooling obtained in the jet allowed the resolution of the vibrational-rotational transitions leading to isotope selective absorption. This data would be of importance in the characterisation of cluster formation process that in turn would facilitate efficient repression of cluster formation of the desired isotope and the application of this scheme for isotope separation experiments.

63


processes in material, like excitation cross-section, loss of coherence among carriers, thermalization rate, electron-phonon coupling, carrier transport, radiative and nonradiative decay mechanisms etc. get modified. The rapid progress in computing technology, quantum information manipulation and the basic requirement in understanding the properties of nanostructured materials demands knowledge on the carrier dynamics at femtosecond time scales.

Metallic nanostructures have the ability to guide and manipulate light in nanometer length scales. This light manipulating ability of metal nanostructures is being utilized for several applications in metamaterials and plasmonics [1]. The response of free-electrons confined within a nanospace (the metal nanoparticle) to the applied field is maximum at a specific wavelength called the localized surface plasmon resonance (LSPR) [2]. At the LSPR, the field strength in and around the particle can be much higher than that of the applied field. The field enhancement at the LSPR shows up even more strongly in optical nonlinear response of metal particles due to the higher-order intensity dependence of the nonlinear response [3-4]. Thus it is much easier to observe the nonlinear response of metal nanostructures compared to other materials. The emerging field of nonlinear plasmonics and nonlinear metamaterials has the potential to revolutionize material science with new optical materials which can provide optimized properties [5-6]. Understanding the effect of LSPR on the higher-order nonlinear response of metal nanostructures and isolation of its instantaneous and delayed response to light are essential for designing metal nanostructure for faster computation and communication based on controlling light with light. Transient absorption measurements on metal nanoparticles performed using a femtosecond laser can be used for understanding such dynamical material response at ultrafast time scales.

Designing semiconductors nanostructures which make use of new quantum process can also be used for improving communication speeds. Recently, ultrafast all optical switching up to 10 Gbit/sec has been demonstrated in a semiconductor nanostructure which

Abstract

Understanding changes in the material response due to the reduction in size of materials down to nanometer scale is essential for designing new materials for different applications. Rapid progress in computing technology, quantum information manipulation and the basic requirement in understanding the properties of nanostructured materials demands knowledge on the carrier dynamics at femtosecond time scales. In this short report on the invited talk, our recent results on ultrafast dynamics measurements carried out on metal nanoparticles, semiconductor quantum wells, Bi thin films and high temperature superconductor nanoflakes using different pump-probe techniques are described. By modeling the transient absorption data measured on metal nanoparticles with a two-temperature model and a generalized T-matrix approach, it has been shown that the metal nanocolloid has intrinsic contribution to various orders of nonlinearities. Enhanced by the local field, such higher-order nonlinear response of metal nanoparticles shows its effects at much lower intensities. We find that the transient reflectivity measurement on near-surface quantum well shows coherent oscillatory signal apart from the usual incoherent ultrafast response. It will be shown that carriers tunneling towards the surface states are responsible for the coherent beating observed in this system. An optical beam could be used to directly probe the electronic states of the system while direct probing of lattice dynamics requires ultrashort x-ray or electron pulse. Method of performing time-resolved transmission electron diffraction experiment, making use of a femtosecond laser will be presented. A review of recent results on ultrafast electron diffraction experiments performed on Bi & BSCCO nanofilms will be described.

Keywords: Ultrafast spectroscopy, Pump-Probe Technique, Nanostructures, Carrier Dynamics, and Lattice Dynamics.

Introduction

At present there is an enormous progress in material fabrication and manipulation at nanometer length scales. Due to the size reduction various fast dynamical

Ultrafast Dynamics in Nanostructured Materials1* 1 1 1 2 2J. Jayabalan , S. Khan , A. Singh , R. Chari , P. Zhou , C. Streubühr ,

2 3 3 2K. Sokolowski-Tinten , Zi-An Li , M. Farle and U. Bovensiepen

1Ultrafast Spectroscopy Laboratory, Laser Physics Applications Section, Raja Ramanna Centre for Advanced Technology, Indore-452013, India.

2Faculty of Physics, University of Duisburg-Essen, Duisburg 47057, Germany.3Faculty of Physics and Center for Nanointegration Duisburg-Essen (CeNIDE),

University Duisburg-Essen, D-47048 Duisburg, Germany.* E-mail: [email protected]

64


makes use of surface quantum wells [7]. In a near-surface quantum well (NSQW), the quantum well (QW) is in close proximity to the top surface of the nanostructure. In such a structure the carriers excited inside the QW can tunnel to the surface and recombine non-radiatively [8]. This tunneling pathway to the surface provides a faster decay channel to the carriers, which reduces the lifetime of carriers in quantum wells. This can be extremely beneficial for ultrafast optoelectronic applications [9]. Studies on the coherence and dephasing time of these tunneling carriers is essential for their coherent manipulation and for ultrafast quantum device applications based on the near-surface QW structures. Usually, oscillatory signal observed over a smoothly varying optical response of the material after excitation by an ultrafast pulse provides information about the coherent beating process happening inside the materials [10]. Such beating signals observed in transient reflectivity measurements could be used for studying the coherence of tunneling carriers in a near-surface QW.

Time-resolved all-optical pump–probe techniques like transient absorption and transient reflection are rather well established and can also provide a sensitive method to analyze some properties of the lattice by probing the electronic response [3]. However, optical data provide limited quantitative information about the state of lattice and it is difficult to directly deduce the complicated electron-lattice interactions from optical data. Complete understanding of the lattice dynamics is essential in many cases, especially in cases of materials having complex crystal structures like cuprites, where the mechanism of high temperature superconductivity is unknown. The time-varying diffraction pattern of the lattice can provide direct structural dynamics of the lattice. It is expected that a direct measurement of structural dynamics following excitation by an ultrafast optical pulse may provide information about the electron-lattice coupling in these materials [11-12]. All of the previous direct lattice dynamics studies in Bismuth Strontium Calcium Copper Oxide (BSCCO) were done in reflection geometry which suffers a serious drawback of interaction of pump-laser generated electrons with the probe electrons. For more accurate information on the lattice dynamics, the measurement should be done in transmission geometry. So far there are no reports on the measurement of lattice dynamics through time-resolved transmission electron diffraction setup (TR-TED).

In this talk, our recent results on ultrafast dynamics measurements performed various nanostructured materials are presented. Studies on three different nanomaterials using three different pump-probe techniques are described in different sections as follows:

1. Transient absorption measurements on metal

nanoplatelet/water colloids,

2. Transient reflectivity measurements on semi-conductor quantum wells and

3. TR-TED measurements on Bi and BSCCO thin films in transmission geometry.

Origin of Optical Nonlinearities in Metal Nano-particles

As described above the nonlinear response of metal nanoparticles is expected to be strong and fast at the LSPR. The aim of this work was to determine the physical process responsible for the sub-ps nonlinear optical response. The silver nanoplatelets colloidal solution on which the transient absorption measurements were performed was prepared by a wet chemical method [13]. Figure 1 shows the extinction spectrum of the sample. The average diameter and thickness estimated from Transmission Electron Microscope (TEM) images of the nanoplatelets were 49 and 7 nm, respectively (see inset of Fig.1). The transient absorption measurements were performed using a 190 fs, 82 MHz Ti:Sapphire laser at 778 nm.

The transmissivity (ΔT/T) of the silver nanocolloid was measured at different pump peak intensities in the range

-2 -2of 0.08 MWmm to 1.07 MWmm and is shown in Fig. 2. Just after the excitation of the sample by the pump pulse (time = 0) the transmissivity of the sample increases and then recovers with a fast and a slow components. For lower pump intensities the recovery is faster and exponential. As the pump intensity increases the recovery slows down and also becomes nonexponen-tial. The transient absorption (Δ ) data can be obtained from the transmissivity using the relation, ΔT/T = -Δαd, where d is the thickness of the sample. The measured time and intensity dependent transient absorption data could be fit to a third-order polynomial,

(1)

and the time dependence of the coefficients α , α , and α 2 4 6

was obtained. Here t is the delay between the pump and pp

probe pulses. These coefficients α , α , and α are 2 4 6

proportional to the imaginary part of the third-, fifth-, and seventh-order nonlinearities, respectively and hence these nonlinear coefficients can be estimated from a single experiment [4].

Figure 3 shows the time dependence of α , α , and α 2 4 6

extracted from the transient absorption measurements. All the coefficients shows a fast initial rise followed by recovery to a small value of the same sign which has a relatively much longer decay time. In addition α and α 4 6

peaks again shows a peak of opposite sign before reaching the small value. The maximum value of these

α

65


material formed by metal nanoparticles dilutely dispersed in a transparent host medium, part of the energy in the pulse is absorbed by the free-electrons in the metal. The initial nonthermal electrons thermalize in the first few hundreds of femtoseconds by electron–electron scattering [3]. After thermalization the temperature of the electrons (T ) is much higher than that of the lattice e

−11 −1 −24coefficients α , α , and α are −1.1 × 10 mW , 9 × 10 2 4 6 3 −2 −36 5 −3m W , and −3 × 10 m W . Further simple exponential

fit to the time dependence shows that decay times of α , 2

α , and α are 0.9 ± 0.04 ps, 0.41 ± 0.05 ps and 0.27 ± 0.06 4 6

ps respectively. Several interesting relationships between the nonlinear coefficients can be noted here. First the magnitude of α , and α are of the order of square and cube 4 6

of the magnitude of α . Second the decay time of α , and α 2 4 6

are nearly one half and one third of the decay time of α . 2

Finally, the peak change of α is positive while the peak 4

change of α and α are negative. These relations between 2 6

the various nonlinear coefficients hint at a common origin for all the different orders of nonlinearities in metal nanoparticles.

When a femtosecond pulse (shorter than the thermalization time of electrons) falls on a composite

Fig. nanoplatelet colloid at different pump peak intensities. Note that as the pump intensity increases the decay time increases.

2: The measured change in the transmissivity of the silver

Fig. 1: Extinction spectrum of the silver nanoplatelet sample. Red vertical line shows the wavelength at which transient absorption measurement is performed. Inset shows a typical TEM picture of the sample.

Fig.

of silver nanoplatelets in water obtained from the experimental transient absorption data. Solid lines: guide to the eyes.

3: The extracted time dependence of (a) α , (b) α , and (c) α 2 4 6

Fig.

α of silver nanocomposite material, after the thermalization of 6

electrons, based on a two-temperature model.

4: The theoretical time dependence of (a) α , (b) α , and (c) 2 4

66


metal nanoparticle is different from that of just metal due to dielectric confinement effects. The change in absorption of a metal nanoparticle embedded in a transparent host medium can be written as [15]

(7)

Substituting for the Δε' , the real part of change in m

dielectric constant of metal from Eqn.(5), the change in absorption at probe frequency can be written as,

(8)

where

(9)

(10)

(11)

where

(12)

These nonlinear absorption coefficients α , α , and α are 2 4 6

proportional to the imaginary part of the third-, fifth-, and seventh-order nonlinearities of the effective composite material respectively. Figure 4 shows the estimated time dependence of α , α , and α using the above equations. 2 4 6

The wavelength of the pump and probe are assumed to be the same and are on the blue side of the LSPR peak. The value of τ (= 1/n) is taken to be 0.9 ps. The model shows that the third-order nonlinear absorption coefficient α 2

recovers exponentially which is also the case in experiment. α and α increases in magnitude first and 4 6

decays to zero then peaks again with opposite sign before reducing finally to zero. This changing sign behavior also matches with the experimental results. Finally, the decay time of higher-order nonlinear absorption coefficients estimated based on the model is lower than that of the lower orders also shown by the experimental results. Thus the extended two-temperature model described here correctly explains the signs and time dependence of the measured nonlinear absorption coefficients.

In the past, open aperture z-scan based experiments performed on metal nanoparticles show an increase or decrease in the transmission depending on the wavelength and these results have been interpreted based on absorption saturation or two-photon absorption. However, the origin of such change in transmission in case of a metal nanoparticle colloid is due to the change in temperature of electrons due to the absorbed light. Figure 5 shows the dispersion of α , α , and α around the LSPR 2 4 6

of silver nanospheres in water based on the model described above. The α which is proportional to the 2

temperature (T ) which is still nearly at the room l

temperature (T ). At later times, the hot electrons cool by o

heating up the lattice though electron–phonon scattering. The time dependence of the electron and lattice temperature can be described by coupled differential equations [14]. Assuming an instantaneous internal thermalization of electrons to a temperature T and for exc

small absorbed power density the coupled equations can be analytically solved to obtain [14]:

(2)

where T is the temperature of the thermalized eq

electron–lattice system G is the electron–phonon coupling constant. The constant γ is defined through the 0

dependence of the specific heat capacity of electrons in metal on its temperature given by C = γ T . The maximum e 0 e

mechange in the temperature of the electrons (ΔT = T -T ) e exc o

depends on the amount of energy absorbed and is given by,

(2)

where ΔQ is the amount of absorbed energy by the metal per unit volume. Since the specific heat capacity of the lattice is much larger than that of electrons the T will be eq

nearly equal to that of room temperature, with this approximation the Eqn. 2 can be written as,

(4)

where ΔT is the change in the temperature of the electron e

at any time t and n = G/(γ T ). The solution of the above 0 o

equation is given by the Lambert W function. For small arguments the Lambert W function can be expanded using Taylor series. For the intensity regime where a linear relation exists between the change in dielectric constant of metal and the temperature of electrons, the change in the dielectric constant of the metal (Δε ) in m

terms of the field inside the particle (E ) can be obtained L

from Eqn.4 as [15],

(5)

with

(6)

where the local field (E ) inside the particle is related to L

the applied field (E ) by E = f(ω)E , n and k are the real 0 L 0 m m

and imaginary part of the refractive index of the metal respectively. Note that the first, second and third term in Eqn.(5) are related to the third-, fifth- and seventh-order nonlinearity of the metal since the intensity is

2proportional to |E | . However the nonlinear response of 0

67


imaginary part of third-order nonlinearity has a positive value on the red-side of LSPR this will result in an increase in absorption with intensity while on the blue-side the same process will result in reduction of absorption with intensity.

Using the above described experiment we have also measured the volume fraction dependence of the delayed third- and fifth-order nonlinearity of silver metal nanoparticles and are shown in Fig.6. Results show that the hot-electron contribution to the third- and fifth-order nonlinearity of metal nanoparticles is much higher than their instantaneous values (within the pulse duration) [16, 17]. It should be noted that the metal nanoparticle colloid shows different nonlinearities at different times (Fig.3). Further, even the third-order nonlinear response is maximum only after 500 fs. This is due to the fact that the change in transmission reaches its maximum only after the thermalization of electrons which takes about 500 fs in silver [3]. Hence the nonlinearity measured using a laser of pulse duration higher than the thermalization time of electrons would be much higher than that

Fig.

around its LSPR at ~380 nm. α is the linear absorption 0

coefficient.

5: Dispersion of α , α , and α of a metal nanosphere 2 4 6

Fig. rd thcorrected (open points) imaginary part of 3 - (circle) and 5 -

order (square) nonlinearities of the silver nanoplatelet colloid.

6: The measured (solid points) and pump depletion effect

measured with shorter pulse durations. Further, we find that the effect of pump depletion on the measured third- and fifth-order nonlinearities is strong and carrier dynamics measurements on metal particles should also consider the effect of pump-depletion to interpret the transient absorption data [16, 17].

Coherence of Tunneling Carriers in a Near-Surface Quantum Well

The aim of this study was to look for the effect of the proximity of surface to a quantum well on the dynamics of photo-generated carriers. Using femto-second transient reflectivity measurements in a near-surface quantum well, we could identify a coherent interaction of carriers mediated by the surface states. The sample, GaAs P /Al Ga As near surface single-quantum 0.86 0.14 0.7 0.3

well (NSQW), is grown by low-pressure metal-organic vapor phase epitaxy on a GaAs substrate [18]. The main energy levels of the NSQW corresponding to the lowest transitions as determined from photoreflectance measurement are given in the inset of Fig.7. e (lh , hh ) 1 1 1

the first electron (light hole, heavy hole) level, hh is the 2

second heavy hole level and the SS represents surface state. The error in the estimation of energy QW transitions is ± 2 meV which comes from the fitting of photoreflectance spectrum. The GaAs substrate is opaque to the photon energy of quantum well levels hence transient absorption technique cannot be used for studying the carrier dynamics in this system. To study the carrier dynamics in NSQW, transient reflectivity

Fig. exponential fit to the data. Inset: Layer structure and energy levels (see the text for details. (b) The |ΔR/R| after subtraction of exponential fit. The solid line is the best fit to the data with F.

7: (a) The as measured |ΔR/R| of NSQW. The line is the

68


In a three level system beating occurs when the carriers from two different levels are coherently coupled via a single third state. In this case of NSQW the beating occurred due to coherent coupling of holes tunneling from lh and hh states to the surface states and this 1 2

process is confirmed by the following. First the oscillation period matches the energy difference between the states lh and hh . These two states have a higher 1 2

tunneling probability to the surface states. Secondly the oscillations were observed only when both of these holes states are simultaneously excited with the femtosecond pulse. Beating disappears if the laser wavelength is tuned away from any of these levels. Thirdly similar careful measurements on a bulk GaAs and a sample with single layer of AlGaAs on GaAs did not show any such beating at various photon energies. Fourthly, the beating was not observed when transient reflectivity measurements were performed on a similar QW sample with 50 nm top-barrier, where the tunneling probability to the surface state is much less.

We have also studied the effect of carrier density on the coherence of tunneling carriers by measuring the beating at various pump fluences (Fig.8). We find that the oscillation time period is independent of pump fluences since the energy gap between lh and hh is not expected to 1 2

change with the pump fluences. The amplitude of oscillation and the dephasing rate (1/τ ) increases linearly with the pump fluence. At higher pump fluences the number of tunneling carriers is more and hence it increases linearly. The observed increase in the dephasing rate with increasing pump fluence in NSQW suggests that the inter-sub-band carrier-carrier scattering is the main cause for the observed dephasing.

Ultrafast Lattice Dynamics in Nanofilms

This section describe the ultrafast electron diffraction measurement in transmission geometry on Bi nanoflim and on BSCCO nanoflake. A 50 fs, 800 nm and 1-200 kHz repetition rate with maximal ~1 mJ/pulse homemade laser was used with a time-resolved transmission electron diffraction setup for the lattice dynamics studies (Fig. 9). Part of the laser beam was frequency tripled by using two BBO crystals. This third-harmonic beam was used to generate the ultra-short electron probe pulse by exciting a 20 nm Au photocathode and the electron beam was subsequently accelerated using an anode at 30 kV. After passing through an aperture in the anode, the electron beam was focused on to the sample using a magnetic lens. The electron diffraction pattern of the sample was recorded using an MCP, phosphor screen, and CCD camera combination. The whole electron beam generation and detection along with a sample were kept

-7below 1x10 mbar pressure. The 800 nm fundamental

Fig. fluence. The solid lines are the best fit to the experimental data with an oscillating and decaying function. The coherence decay time (τ) and the excitation fluence are also shown.

8: The dependence of the observed beating on excitation

measurements were performed with a ~70 fs, 82 MHz, tunable Ti:Sapphire laser (spectral width: 28 meV) at

oroom temperature (23 C) in a standard degenerate pump-probe geometry. The change in the reflectivity of the sample was measured by chopping the pump beam and monitoring the change in the probe energy using a lock-in-amplifier.

The magnitude of transient reflectivity signal (|ΔR/R|) for NSQW measured at a photon energy 1.577 eV is shown in Fig.7. With a spectral width of 28 meV (FWHM) the laser beam excites the three lowest resonances e -lh , e -hh and 1 1 1 1

e -hh of the NSQW. With the arrival of the pump pulse 1 2

|ΔR/R| increases and reaches a maximum near the end of the pump pulse. This peak is then followed by a double decay behavior which is typical of a AlGaAs/GaAs heterostructure. Note that a periodic variation of |ΔR/R| appears around the decaying signal (Fig. 7). To obtain only this oscillating signal ( ) from the experimental data the measured |ΔR/R| after 0.15 ps was first fit to a single exponential function. is then obtained by subtracting this best fit function to the experimental data (See Fig.7).

The oscillations found in the could be fit well with the function, F given by

(13)

where A, T, φ and τ are the amplitude, period, phase and decay time of the oscillations respectively. For excitation

-2with 26 μJcm the value of T and τ are found to be 120 ± 4 fs and 147 ± 10 fs respectively [19].

69


intensity due to Debye-Waller effect. Thus the measured time constant gives the rate at which electron-lattice scattering is taking place. This kind of experiment first needs the successful preparation of nanofilm to get the diffraction pattern. For studying the lattice dynamics using TR-TED setup, nanoflakes of BSCCO samples were prepared by micromechanical cleavage technique from bulk. These nanoflakes are then deposited on a TEM grid for TR-TED measurements. The electron diffraction pattern was obtained in TR-TED setup by optimizing the rate at which the electrons are falling on the nanoflakes. Similar to the Bi case (Fig.10), the diffraction spot of BSCCO reduces by about 3% with a time constant of ~5

-2ps at a pump fluence of 3 mJcm .

Conclusion

It has been shown that different kinds of time-resolved pump-probe techniques making use of femtosecond pulsed lasers could be used for studying various processes in nanostructured materials. Measurement of transient absorption at different pump intensities was used to understand the origin of nonlinearities in metal nanoparticles. Using two-temperature model it has been shown that the time dependence of the electron temperature could directly explain the origin and time dependence of various orders of nonlinearities in metal nanoparticles. Transient reflectivity measurements were used to demonstrate coherent beating in near-surface quantum well and to study its origin and carrier density dependence. We show that the beating is caused by holes tunneling from QW states to the surface states. The electron-lattice scattering time constant in Bi and in BSCCO nanofilms has been studied using a time-resolved electron diffraction setup in transmission geometry.

Acknowledgements: The authors acknowledge Dr. T. K. Sharma, Dr. S. Pal for providing the semiconductor quantum well samples and useful discussions. Further they would also like to acknowledge the support from Dr. H. S. Rawat, Head, Laser Physics Applications Section and Dr. S. M. Oak, Head, Solid State Lasers Division. The TR-TED work has been supported by the European Union, Seventh Framework Programme, under the project GO FAST, grant agreement No. 280555. M.F.

References

1. J. A. Schuller, E. S. Barnard, W. Cai, Y. Chul Jun, J. S. White, and M. L. Brongersma, Nature Mater. 9, 193 (2010).

2. J. Jayabalan, A. Singh, R. Chari, and S. M. Oak, Nanotechnology 18, 315704 (2007).

3. J. Jayabalan, A. Singh, R. Chari, S. Khan, H. Srivastava, and S. M. Oak, Appl. Phys. Lett. 94,

beam was used for exciting the sample. The delay between the pump and electron probe beam was varied by changing the path length of the pump beam. The sample's lattice dynamics was then monitored by measuring the diffraction pattern obtained from the probe electron pulse. The overall time resolution achievable with this setup is limited by the electron probe duration to < 0.7 ps.

For identifying the zero time as well as the spatial overlap the TR-TED experiment was first performed on a Bi nanofilm of 22 nm thickness. The sample was excited by

2800 nm pulse of fluence ~0.5 mJ/cm . Electron diffraction patterns at various pump-probe delays were recorded. Several scans were performed and averaged to improve the signal to noise ratio. The time evolution of a single diffraction spot intensity was then analyzed from the recorded diffraction images. Figure 10 shows the time dependence of diffraction intensity of [2 -1 -1] spot. After excitation the intensity of the diffraction spot reduces exponentially and levels to about 98% of the original intensity. The decay could be fit well with a single exponential decay with a time constant 4-6 ps. As soon as the pump beam falls on the sample it excites the electrons and the temperature of the electrons increases. The electrons then cool down by heating the lattice through electron-lattice scattering. The increase in temperature of the lattice causes the reduction of diffraction spot

Fig. 9: Schematic of the ultrafast electron diffraction setup.

Fig. versus delay time of a BSCCO nanoflake. line: Exponential function with a time constant 4-6 ps.

10: Normalized diffraction intensity of the [2 -1 -1] - spot

70


12. Dynamics of Size-Selected Gold Nanoparticles Studied by Ultrafast Electron Nanocrystallography, C. Y. Ruan, Y. Murooka, R. K. Raman, and R. A. Murdick, Nano Lett. 7, 1290 (2007).

13. G. S. Metraux and C. A. Mirkin, Adv. Mater., 17, 412 (2005).

14. F. Vallee, Non-Equilibrium Dynamics of Semiconductors and Nanostructures (CRC Press, 2005), Chap. 5, pp. 101–142.

15. J. Jayabalan, J. Opt. Soc. Am. B, 28, 2448 (2011).

16. J. Jayabalan, A. Singh, S. Khan and R. Chari, J. Opt., 15, 055203 (2013).

17. J. Jayabalan, A. Singh, S. Khan, and R. Chari, J. Appl. Phys., 112, 103524 (2012).

18. S. Pal, S. D. Singh, S. Porwal, T. K. Sharma, S. Khan, J. Jayabalan, R. Chari, and S. M. Oak, Semicond. Sci. Technol. 28, 035016 (2013).

19. S. Khan, J. Jayabalan, R. Chari, S. Pal, S. Porwal, T. K. Sharma and S. M. Oak, Appl. Phys. Lett., 105, 073106 (2014).

181902 (2009).

4. D. Rativa, R. E. de Araujo, and A. S. L. Gomes, Opt. Express 16, 19244 (2008).

5. M. Lapine, V. I. Shadrivov, and Y. S. Kivshar, Rev. Mod. Phys. 86, 1093, (2014).

6. M. Kauranen and A. V. Zayats, Nature Photonics 6, 737 (2012).

7. A. Bazin, K. Lengl, M. Gay, P. Monnier, L. Bramerie, R. Braive, G. Beaudoin, I. Sagnes, R. Raj, and F. Raineri, Appl. Phys. Lett. 104, 011102 (2014).

8. J. M. Moison, K. Elcess, F. Houzay, J. Y. Marzin, J. M. Gerard, F. Barthe, and M. Bensoussan, Phys. Rev. B 41, 12945 (1990).

9. S. Ohta, O. Kojima, T. Kita, and T. Isu, J. Appl. Phys. 111, 023505 (2012).

10. F. Rossi and T. Kuhn, Rev. Mod. Phys. 74, 895 (2002).

11. R. K. Raman, R. A. Murdick, R. J. Worhatch, Y. Murooka, S. D. Mahanti, T. R. T. Han, and C. Y. Ruan, Phys. Rev. Lett. 104, 123401 (2010).

71


generation of multiply charged atomic ions were observed in different atomic and molecular clusters using

9 2nanosecond laser pulses of intensity ~ 10 W/cm [9-12]. 6Since electric field (~10 V/cm) associated with the

gigawatt intense laser pulse is not large enough, possibility of field ionisation can be ruled out. In order to understand the mechanism of formation of multiply charged ions at such low laser intensity conditions, study has been carried out on diethyl ether clusters by varying different experimental parameters.

Diethyl ether is an aliphatic ether, used as a common solvent in several industrial processes, namely production of cellulose plastic, etc. In addition, ethers have astrochemical importance, since they have been observed to be present in the interstellar medium. Hence, it is necessary to understand photochemistry of diethyl ether which has profound applications in various fields. Earlier, photochemistry of diethyl ether has been studied using single photon ionisation by Botter et. al.[13] Based on their studies, accurate ionisation energy (9.60 eV) has been determined along with appearance energy of different fragments of diethyl ether monomer. Castlemann and co-workers studied photochemistry of diethyl ether clusters where effect of solvation in chemical reaction (intracluster ion molecule reaction) has been studied [14]. Moreover, generation of multiply charged atomic ions of carbon and oxygen has been also observed in photoionisation of diethyl ether cluster at 532

9 11 2nm with laser intensity ~10 -10 W/cm [15]. In the present work, diethyl ether cluster has been ionised using

9 2nanosecond laser pulse of intensity ~ 10 W/cm to understand the mechanism of multiple charge state formation at gigawatt laser intensity conditions. Firstly, effect of laser wavelength on formation of multiply charged atomic ions has been studied by varying selected wavelengths in UV and visible region. Secondly, role of clusters for generation of higher charge state has been probed by comparing mass spectra of diethyl ether monomer and cluster at 532 nm.

Experimental Section

Details of the experimental setup have been described in our earlier publications and only a brief description

Abstract

Diethyl ether clusters have been photoionised using second (532 nm), third (355 nm) and fourth (266 nm)

9 2harmonic of Nd:YAG laser having intensity ~ 10 W/cm . The ions produced as a result of photoionisation, were detected using time-of-flight mass spectrometer. Using 266 and 355 nm, singly charged fragment ions were observed along with hydrogenated molecular ion peak. At 532 nm, along with singly charged fragment ions, multiply charged atomic ions of carbon and oxygen were observed up to +4 and +3 state respectively. Generation

9 2of multiply charged ions was unusual at ~ 10 W/cm and has been qualitatively explained based on the three stage cluster ionisation mechanism. According to this mechanism, ionisation was initiated via multiphoton process followed by electron impact ionisation which causes formation of multiply charged ions inside the cluster.

Keywords: Photoionisation, Clusters, Diethyl ether, Multiply charged atomic ions

Introduction

Photoionisation of atoms/molecules using laser pulses leads to various photo physical/chemical processes such as excitation, dissociation (in case of molecule), ionisation etc. When atomic/molecular clusters which are aggregates of atoms or molecules having size intermediate between individual atoms/molecules and bulk materials [1,2], interact with laser pulses, new photo physical processes can occur. Thus, atomic/ molecular clusters add a new facet in laser-cluster interaction studies and in recent times a lot of studies have been devoted to this area of research. Efficient laser-cluster interaction leads to generation of multiply charged atomic ions, [3] energetic electrons, [4] emission of x-rays [5] and even neutrons in tabletop setup [6]. Mostly these observations were confined in picoseconds and femtosecond laser pulses having intensity in the range of

12 20 2 9~10 - 10 W/cm [7, 8]. The electric field (~ 10 V/cm) associated with the high intensity laser is of the order of atomic potential of the cluster constituents which cause field ionisation of the cluster and subsequently form highly charged atomic species. However, in our group

Photoionisation of Atomic/Molecular Snow: Creating Highly Charged Matter using Low Intensity Laser Pulses

S. Das, P. Sharma and R.K. Vatsa*Chemistry Division, Bhabha Atomic Research Centre, Mumbai 400085, India


72


relevant to this work is given here [16-18]. Neutral diethyl ether clusters were generated via supersonic expansion of C H OC H vapors seeded in argon at a 2 5 2 5

pressure of 3 bar using a pulse valve (0.8 mm nozzle diameter and 300 μs pulse duration). The supersonic jet so produced was skimmed at a distance of 5 cm from the pulsed nozzle. The distance between the skimmer and the ionization region is 17 cm. Ionization was carried out using second (532nm), third (355 nm) and fourth (266 nm) harmonic of a pulsed Nd:YAG nanosecond laser (Quanta System, Italy; 10 ns, GIANT G790-10). The ions so formed were accelerated and guided into a 100 cm field free region using a home built Wiley–McLaren assembly and detected using a channel electron multiplier (CEM) detector (Dr. Sjuts Optotechnik GmbH, Germany). Typical voltages applied to the repeller and extractor grids were 2700 and 2000V, respectively. The ion signal from the CEM detector was recorded on a digital storage oscilloscope and averaged for 500 shots. The averaged signal was finally transferred to a computer for further processing. The mass resolution of the instrument is ~300.

Results and Discussions

Figure 1 (a) represents the mass spectra of diethyl ether cluster upon photoionisation by 266 nm laser pulses of

9 2intensity ~ 10 W/cm . The major fragments of diethyl + + + + +ether clusters are CH , C H , CH OH , C H O , C H O , 3 2 5 2 2 3 2 5

+C H OCH at m/z 15, 29, 31, 43, 45 and 59 respectively 2 5 2

in figure 1 (a). Apart from fragment ions, hydrogenated molecular ion peak was also observed at m/z 75. Fragmentation of diethyl ether cluster at laser intensity ~

9 210 W/cm can be explained based on multiphoton ionisation process. In order to determine photon dependency, laser power dependency study has been carried out. This study shows two photon dependence for different fragment ions at this wavelength indicating ionization mediated via Rydberg states at ~ 9.32 eV.

Figure 1 (b) represents mass spectra of diethyl ether clusters at 355 nm. The fragments at m/z 15, 29, 31, 43, 45

+ + + + +and 59 represents CH , C H , CH OH , C H O , C H O , 3 2 5 2 2 3 2 5

C H OCH respectively. However, no molecular ion peak 2 5 2

was observed in the mass spectra. Laser power dependence study has been carried out for different fragments and two photon dependency was found at this wavelength. Thus, at 355nm, diethyl ether molecules are excited to an electronic state ~ 7 eV which is dissociative in nature. As a result, no molecular ion peak was observed in the mass spectrum. After dissociation, the fragments undergo ionisation via multiphoton process.

Figure 2 (a) represents mass spectrum of diethyl ether 9 2clusters at 532 nm with laser intensity ~ 10 W/cm .

Fragment ions were observed at m/z 1, 15, 29, 31, 43, 45,

+ + + + + +59, 74 due to H , CH , C H , CH OH , C H O , C H O , 3 2 5 2 2 3 2 5+ +C H OCH and C H OC H respectively. Along with 2 5 2 2 5 2 5

fragment ions, dimer of diethyl ethyl was also observed at m/z 148. In addition to these singly charged ions, multiply charged atomic ions of oxygen (up to 3+) and carbon (up to 4+) were also observed. Ionisation energy

4+of highest observed charge state is ~ 64.5 eV (C ) which is much higher than the photon energy (2.33 eV) of 532 nm. For observation of such high charge state, absorption of ~ 28 photons is required which is extremely improbable based on the multiphoton ionisation probability. Laser power dependency study shows absorption of 5 photons dependence for different multiply charged atomic ions. This 5 photon excitation leads to diethyl ether molecule to an autoionising state (~ 11.65 eV) above ionisation potential ( ~ 9.60 eV). Earlier, excitation of diethyl ether molecule to such autoionisng state has been reported by Botter et al. [13] However, 5 photon excitation is not sufficient to ionise the cluster to

Fig. 1: Time-of-flight mass spectrum of diethyl ether clusters at 9 2(a) 266 nm and (b) 355 nm with laser intensity ~ 10 W/cm .

73


In order to understand the multiple ionisation of diethyl ether clusters at gigawatt intense laser field, a three stage cluster ionisation mechanism has been proposed [19, 20]. This mechanism assumes that the primary step is multiphoton ionisation of the cluster constituents. Multiphoton ionisation of diethyl ether cluster results in generation of few electron-ion pairs within the cluster. The electrons generated on the surface of the cluster escape leaving behind a net positive charge on the cluster, which then acts as a potential barrier for the remaining inner electrons inside the cluster. As a result, the electrons which have left the parent diethyl ether ion, but cannot leave the cluster (referred as quasi free electrons) and are forced to interact with the optical field of the laser pulse. These quasi-free electrons under the influence of Coulomb field within the cluster, keep extracting energy from the laser pulse presumably via inverse Bremsstrahlung (IBS) process, through electron-ion and electron-neutral collisions. Once the electron energy exceeds the ionisation energy of singly charged ion, secondary ionisation by these energetic electrons leading to generation of doubly and triply charged carbon or oxygen ions can take place. In spite of increasing energy of the electrons due to IBS process, the electrons are retained within the cluster because of increasing Coulomb potential which arises due to the small but finite escape probability of energetic electrons. This process of electron energisation and ionisation to the next higher charge state of carbon and oxygen continues until a stage comes when Coulombic repulsive energy overcomes the total cohesive energy of the cluster and the multiply charged cluster violently explodes resulting in formation of multiply charged atomic ions with large kinetic energy.

The overall energy gained by the quasi free electrons in the total time span starting from initial ionisation till the disintegration of cluster is dictated by the product of ponderomotive energy and the total number of effective electron-ion/neutral collision frequency and can be written as [21] –

(1)

Here ν is the collision frequency of the inner ionized 14 15electrons and is of the order of ~ 10 -10 Hz and U is p

ponderomotive energy of electrons given by the equation-

(2)

2where I is expressed in W/cm and λ in mm. From equation 2, it is clear that as wavelength (λ) increases, ponderomotive energy (U ) of electrons increases p

quadratically for a given laser intensity. As a result, net energy extraction by the electron from the optical field is

4+ 4+give rise to such high charge state such as C or O . Thus, some efficient mechanism is operating to extract large energy from the optical field to the cluster and cause such high charge state formation.

In order to understand the role of clusters in generation of multiply charged atomic ions under our experimental conditions, similar photoionisation study has been carried out in diethyl ether monomer. Figure 2(b) represents the time-of-flight mass spectrum of diethyl ether monomer at 532 nm. The mass spectrum contains only singly charged fragment ions similar to cluster spectra however no multiply charged atomic ions were observed. Thus, comparing time-of-flight mass spectra of monomer and clusters it can be concluded that cluster is playing a crucial role for generation of multiply charged

9 2atomic ions at 532 nm of laser intensity ~ 10 W/cm .

Fig. 2: Time-of-flight mass spectrum of diethyl ether (a) 9clusters and (b) monomer at 532 nm with laser intensity ~ 10

2W/cm .

74


Chem. Res. 19, 413 (1986).

2. A.W. Castleman, Jr. and R.G. Keesee, "Clusters: Properties and Formation", Ann. Rev. Phys. Chem. 37, 525 (1986).

3. T. Ditmire, J.W.G. Tisch, E. Springate, M.B. Mason, N. Hay, R.A. Smith, J. Marangos and M.H.R. Hutchison, Nature 54, 386 (1997).

4. L.M. Chen, J.J. Park, K.H. Hong, I.W. Choi, J.L. Kim, J. Zhang and C.H. Nam, Phys. Plasmas, 9, 3595 (2002).

5. S. Dobosz, M. Schmidt, M. Perdrix, P. Meynadier, O. Gobert and D. Normand, JETP Lett. 68, 485 (1998).

6. T. Ditmire, J. Zweiback, V.P. Yanovsky, T.E. Cowan, G. Hays and K.B. Wharton, Nature 398, 489 (1999).

7. G. Karras and C. Kosmidis, Phys. Chem. Chem. Phys. 14, 12147 (2012).

8. V. Kumarappan, M. Krishnamurthy and D. Mathur, Phys. Rev. Lett. 87, 085005 (2001).

9. P. Sharma, R.K. Vatsa, S.K. Kulshreshtha, J. Jha, D. Mathur and M. Krishnamurthy, J. Chem. Phys. 125, 034304 (2006).

10. S. Das, P.M. Badani, P. Sharma, R.K. Vatsa, D. Das, A. Majumder and A.K. Das, Rapid Commun. Mass Spectrom. 25, 1028 (2011).

11. S. Das, P.M. Badani, P. Sharma and R.K. Vatsa Chem Phys Lett 552, 13 (2012).

12. P.M. Badani, S. Das, P. Sharma and R.K. Vatsa, Int. J. Mass Spectrom. 348, 53 (2013).

13. R. Botter, J.M. Pechine and H.M. Rosenstock, International Journal of mass spectrometry and ion phys. 25 7 (1977).

14. S. Wei, W.B. Tzeng and A.W. castleman Jr., J. Chem. Phys. 95, 5080 (1991).

15. N. Zhang, W. Wang, W. Zhao, F. Han and H. Li, Chem Phys 373, 181 (2010).

16. S. Das, P. Sharma, A. Majumder and R.K. Vatsa, J. Indian Chem Soc. 87, 165 (2010).

17. P.M. Badani, S. Das, P. Sharma and R.K. Vatsa, Rapid Commun. Mass Spectrom. 26, 2204 (2012).

18. S. Das, P.M. Badani, P. Sharma and R.K. Vatsa, RSC Advances, 3, 12867 (2013).

19. N. Zhang, W. Wang, H. Cang, H. Wang and H. Li, Chem. Phys. Lett. 469, 14 (2009).

20. W. Wang, H. Li, D. Niu, L. Wen and N. Zhang, Chem. Phys. 352, 111 (2008).

21. T. Ditmire, T. Donnelly, A.M. Rubenchik, R.W. Falcone and M.D. Perry, Phys. Rev. A: At., Mol., Opt. Phys. 53, 3379 (1996).

higher at longer wavelength. Thus, higher level of ionisation and charging is expected at longer wavelength, which is qualitatively in agreement with the present experimental observations.

Based on the above proposed three stage cluster ionisation mechanism, role of clusters for generation of multiply charged atomic ions can also be explained. The electrons which are produced at the surface of the cluster due to multiphoton ionisation will leave the cluster at the early part of the laser pulse. However, the quasi free electrons which are caged in the cluster interact with the laser pulse and get energised. These electrons cause multiple ionisations in the cluster via electron impact ionisation. In case of monomer, initial ionisation takes place via multiphoton ionisation and cause singly charged fragment ions. Since there are no caged or quasi free electrons in this process, further ionisation does not occur in monomer. Thus, presence of cluster is essential for generation of multiply charged atomic ions under gigawatt intense laser field.

Conclusions

Photoionisation of diethyl ether cluster has been studied 9as a function of laser wavelength at laser intensity ~ 10

2W/cm . At 266 and 355 nm, only singly charged fragment ions have been observed in time-of-flight mass spectrometer. On the contrary, at 532 nm where individual photon energy is lower, along with singly charged fragment ions, multiply charged atomic ions of

4+ 3+carbon (up to C ) and oxygen (up to O ) have been observed in the mass spectrum. Thus, based on our study, it has been concluded that laser wavelength play an important role in determining the photoionisation phenomena of diethyl ether clusters and longer wavelength interact efficiently and cause multiply charged atomic ions at gigawatt intense laser field. Multiple charge state formation at such low laser intensity conditions can be explained based on three stage cluster ionisation mechanism. According to this mechanism, initial multiphoton ionisation is responsible for generation of singly charged fragment ions. However, multiple charge state has been formed due to electron impact ionisation of the quasi free electrons within the cluster. Moreover, by comparing monomer spectra at 532 nm, role of clusters for generation of these multiply

9 2charged ions at laser intensity ~ 10 W/cm was also established. Thus, by careful optimisation of the laser and cluster condition one can dictate the efficiency of laser-cluster interaction and produce highly charged matter using low intensity laser pulses.

References

1. A.W. Castleman, Jr. and R.G. Keesee, "Clusters: Bridging the Gas and Condensed Phases", Accts.

75


reports of Raman spectra of metals in the literature since their spectral intensities are quite weak. Raman spectra of a few elemental metals (with more than one atom per primitive cell) such as Be and Ga have been reported at ambient conditions; Zr and the hcp phase of Fe have been studied at high pressures in diamond anvil cells.

On the whole however, there is lack of experimental work on actinide metals due to their toxicity, radioactivity and scarcity, and this limits our understanding of these technologically important materials [2]. The present work represents the possibility of bringing into the realm of a versatile laboratory technique i.e. Raman spectroscopy, what has hitherto been studied mainly at large experimental facilities such as nuclear reactors and synchrotrons.

Experimental Details

Natural uranium metal discs of about 5 mm diameter and 1 mm thickness, cut from a cylindrical, vacuum induction melted and cast rod [3] were used in this study. The discs were ground and polished with emery papers of gradually finer mesh, finally polishing the surface with 1 μm diamond paste. Even though mirror finish was obtained the sample had a large number of surface scratches as could be seen under optical and scanning electron microscopes (Figure 1). A high throughput Renishaw micro Raman spectrometer (model Invia) was employed to record the spectra using the 514 nm, 785 nm and 325 nm laser excitations.

Results and Discussion

Unpolished and poorly polished discs exhibited Raman bands characteristic of UO and oxidized UO (Figure 2, 2 2

-1spectrum a) at 450 and 580 cm respectively [4] that disappeared on polishing the metal to a silvery mirror finish (Figure 2, spectrum b). In the frequency region

-1 -1below 150 cm down to 50 cm , however, a number of sharp Raman bands characteristic of the rotational modes of N and O [5] appeared (Figure 2, spectra a and b). 2 2

They are due the molecules from the ambient air adhering to the metal surface. This is in the same wave number region in which uranium Raman bands are expected. Thus it becomes a non-trivial problem to detect the Raman spectra of uranium metal. Techniques such as

Abstract: Raman spectroscopy was an esoteric technique that was practiced in very few specialized laboratories around the world till the advent of lasers in early 60's. Use of lasers to excite the feeble Raman spectra in materials led to a revival of interest in this technique, and further developments in instrumentation such as CCD cameras and the marriage of microscope and Raman spectrometer has literally brought this technique to the centre stage, surpassing even x-ray diffraction in its versatility in many applications. After a brief introduction, this talk will highlight some of our work such as (i) surface enhanced Raman spectroscopy of uranium metal, (ii) spectroscopic resolution of the structure of Zn(CN) , (iii) mechanism of thermal 2

expansion in anomalous thermal expansion materials such as NaZr (PO ) and (iv) study of pigments in ancient 2 4 3

paintings.

Raman Spectroscopy of Uranium by a Surface Enhanced Raman Scattering (SERS) Technique

Study of metals by Raman spectroscopy is a challenging problem due to the low penetration of visible laser into meta ls and hence low sampl ing volumes . Technologically important actinide metals such as Uranium and Plutonium have not been investigated using Raman spectroscopy due to poor signal intensities. We have employed a SERS technique to successfully record the Raman spectrum of Uranium by coating a thin layer of gold on it. Temperature dependence of the mode frequency is in agreement with results reported from inelastic neutron scattering measurements. There is a possibility that this technique can be used to study other metals and materials of weak Raman intensity.

Current Interest on Structural Properties of Uranium

At ambient conditions uranium has orthorhombic structure with 2 atoms per primitive cell. Group theoretical considerations show that three Raman active modes (A +B +B ) are expected to be present in this -g 1g 3g

uranium. Lattice dynamical calculations and inelastic neutron scattering measurements [1] showed the presence of zone center optical modes about 80, 100 and

-1120 cm , but there are no reports of Raman spectra of uranium in the literature. In fact there are not many

Application of Laser Raman Spectroscopy to the Study of Actinides, Anomalous Thermal Expansion Materials and Ancient PaintingsT.R. Ravindran*, A.K. Arora, K. Kamali, C. Ravi and T.N. Sairam

Materials Science Group, Indira Gandhi Centre for Atomic Research, Kalpakkam 603 102*E-mail: [email protected]

76


polishing and encapsulating the metal in an argon atmosphere in a glove box, or even vacuum annealing the

Osample in a quartz tube at 300 C for two hours followed by vacuum sealing did not dispense with the adsorbed air molecules, as determined from the recorded Raman spectra. A layer of gold was coated on the sample to see if SERS effect could be invoked from the proximity of gold of a few nanometers thickness.

Gold layers of 5 nm and 15 nm thickness were deposited by pulsed laser deposition and thermal evaporation respectively on two different, polished discs of uranium. The skin depth of the laser excitation wavelength of 514 nm in gold is about 20 nm, and hence the photons could penetrate into the uranium disc through thin layers of gold. Laser excitation of dark scratch lines and pits on the sample gave rise to Raman spectra expected of uranium metal (Figure 3). Several sets of spectra at 20-121 points

in circular meshes of ≥ 1 μm step size were recorded from both samples. Well resolved Raman bands about 82 (ν ), 1

-1108 (ν ) and 126 cm (ν ) were obtained with laser 2 3

excitations of 514 nm (Figure 3a) and 785 nm (Figure 3b). All sharp bands of N and O disappeared, indicating 2 2

that the adhesion of N and O to gold surface is quite low. 2 2

Several bands of different uranium oxides such as UO 2+x

and U O [6] was observed in the extended range of wave 3 8

Fig. 1: Typical optical micrograph (a) and SEM image (b) of gold coated surfaces of uranium metal.

Fig. 3: Raman spectra (background subtracted) of uranium coated with 5 nm gold film with (a) 785 nm and (b) 514nm laser excitation at different spots on the sample. Higher intensity of spectra with 785 nm excitation is due to larger sampling volume and laser intensity. Three spectra from smooth surfaces are shown at the bottom of (b).

Fig. 2: Polished uranium sample showing rotational Raman -1modes of adsorbed air below 150 cm (a), Raman bands of UO 2

and UO on surface oxidation (b), and the three bands of 2+x-1uranium metal at 81, 107 and 127 cm on coating with a 15 nm

Au film, in addition to the bands of uranium oxides. The relative intensities of the three uranium bands are different in this sample compared to those of the 5 nm Au coated sample.

77


active and there are 2 (F ) IR active modes.1u

We carried out Raman and IR spectroscopic measurements on Zn(CN) (>99.8%, Alfa Aesar) and 2

obtained the mode frequencies. The spectra are shown in Fig.4 a (Raman spectrum) and Fig.4 b (IR spectrum). The results are summarized in the Table 1.

Table 1. Raman and IR spectra observed and expected for the two structures of Zn(CN)2

Though the number of observed Raman and IR modes for the disordered structure are closer to what is expected from factor group analysis, one cannot conclude that the structure of Zn(CN) is 'disordered' solely on this basis. 2

This is because of the fact that though group theory predicts a certain number of modes to be active in the IR

-1 numbers up to 800 cm (Figure 2, spectrum c). Laser excitation of smooth golden yellow surface however did not result in uranium spectra (as in Figure 3 a, three spectra at the bottom).

Conclusion: To summarize, we have successfully recorded the Raman spectra of Uranium metal by a sub-SERS technique by coating 5 nm and 15 nm gold layers on the surface of the metal. Naturally formed thin oxide layer on the metal before gold coating provides the dielectric base to localize the surface plasmons of gold that enhance the Raman signal. More details can be found in [7].

Resolution of the Structure of Zn(CN) Through 2

Vibrational Spectroscopy

Introduction: Zn(CN) exhibits a negative thermal 2

expansion (NTE) coefficient twice as large as that of zirconium tungstate, as determined from structural data at 14 and 305 K [8]. Raman spectroscopic measurement of the mode Gruneisen parameters of the various phonons is important to delineate their contributions to thermal expansion and understand the mechanism of NTE, as was done in the case of Zr(WO ) [9-11]. Based on the 4 2

topologies of framework materials, Zn(CN) was argued 2

to support a large number of low frequency rigid unit phonon modes that contribute to NTE [12]. In this context it is of significance to classify the phonons of different symmetries and probe the phonon spectrum of Zn(CN) . 2

Two different cubic structures have been reported to fit well to the diffraction patterns of Zn(CN) . The 'ordered' 2

1structure with space group P43m (T ) consists of ZnC a d 4

tetrahedron linked to neighbouring four ZnN tetrahedra 4

with CN bonds [13]. It was pointed out that exchanging C and N atoms did not significantly alter the refinement parameters, and this disordered structure with space

4group Pn3m (O ) fitted equally well to the diffraction h

pattern [8].

Results and Discussion: In the present work we attempt to resolve the ambiguity about the structure, from vibrational spectroscopic considerations. Factor group analyses of the phonon symmetries in both the possible structures of Zn(CN) are carried out and compared with 2

experimental results. The unit cells of both the structures contain two formula units and hence 3 degrees of freedom for acoustic and 27 degrees of freedom for optical phonons. For the ordered structure, the following irreducible representation of the optical phonons is

Opticalobtained: Γ = 2A (R) + 2E(R) + 2F + 5F (R, IR). 1 1 2

There are 9 Raman modes 5 IR modes for this 'ordered' structure. On the other hand, for the 'disordered structure,

OpticalΓ = A (R) + E (R) +F + 3F (R) + A + E + 2F (IR) + 1g g 1g 2g 2u u 1u

F . Out of these, 5 modes (A , E and F ) are Raman 2u 1g g 2g

Fig. 4: (a) Raman spectra and (b) Infrared spectrum of Zn(CN) . Observe the mutual exclusion of Raman and IR 2

-1modes, except for the CN stretch mode around 2220 cm .

78


also its structural stability at high pressures.

Experimental and Computational Details: NZP was synthesized by a sol-gel technique. In situ high pressure Raman spectra were recorded up to 20 GPa from a symmetric diamond anvil cell (DAC), using a micro Raman spectrometer (Renishaw, UK, model Invia) with 514 nm laser excitation. High pressure x-ray diffraction experiments were carried out at ambient temperature using a Mao-Bell type DAC in an angle dispersive mode at several pressures up to 19 GPa. The XRD patterns were indexed with POWD program. Theoretical calculation of phonon spectrum was carried out using VASP (Vienna Ab initio Simulation Package). To compute the phonon spectrum over the entire Brillouin zone and to carry out mode assignments, Phonopy program was used. Visual representations of the different phonon modes were obtained by plotting the eigen vectors of displacement of the various atoms in the unit cell using the VESTA visualization software.

Results and Discussion: NaZr (PO ) crystallizes in a 2 4 36rhombohedral structure with space group R-3c (D ) with 3d

six formula units in the crystallographic unit cell. Our group theoretical analysis predicts 25 Raman active and 27 IR active modes from NZP. Raman spectrum of NZP at ambient conditions exhibits 17 distinct Raman bands (Fig. 5), similar to that reported earlier. Raman bands in

-1the range 325-1084 cm arise from the internal vibrations -1of phosphate ions. Lattice modes below 270 cm arise

from the translations of the Zr and the translations and librations of the PO ions.4

Atomic displacement plot -computed by density functional perturbation theory (DFPT)- of the lowest

-1energy Raman active E mode at 113 cm (Fig. 6a) is seen g

to be a combination of the PO librations and Zr 4

translations that contribute negatively to thermal expansion – consistent with the present experimental and

-1computational results. The E mode at 127 cm (Fig. 6b) g

seems to exhibit coupled rotation of PO tetrahedra and 4

ZrO octahedra. The Gruneisen parameter of this mode is 6

experimentally determined to be negative.

From Raman spectroscopic measurements, it is seen that as the pressure is increased, the low energy bands at 72

-1and 112 cm that arise from PO librations and Zr 4

translations soften. There are several distinct changes in the spectra above 5.5 GPa, giving clear indications of a structural phase transformation about 6 GPa from rhombohedral to a new phase. There are no distinct changes in the spectra of the new phase up to 20 GPa. The mode Gruneisen parameters γ = (B /ω )(dω /dP), where i 0 i i

B is the bulk modulus for several internal modes are 0

significantly positive which indicates that these modes

and in Raman, one does not always observe all the modes, due to insufficient intensity (i.e., poor scattering efficiency) or accidental degeneracy of modes.

It is seen from Figure 1 and Table 1 that the disordered structure is more probable in the present case, since the Raman and IR modes are found to be mutually exclusive.

-1Though the 2218 cm IR mode corresponding to the CN stretch is not expected for the disordered structure, this well-known mode due to the net dipole moment of the CN species appears with weak intensity compared to other IR modes. This would mean that the disorder in this case is not a 'complete disorder', i.e., the fractional occupancy of C and N is not each 0.5, but it may be something like 0.4 and 0.6, and this leads to a partially or locally ordered

structure, giving rise to the C≡N mode in Raman as well

as in IR spectra.

Conclusion: We have addressed a long standing controversy in the structure of Zn(CN) , and using 2

spectroscopic techniques, resolved the structure to be disordered. It is noteworthy that this problem could not be satisfactorily solved using x-ray diffraction or neutron diffraction techniques earlier by other research groups. More details can be found in [14].

Thermal Expansion and Structural Stability of NaZr (PO ) Studied by Raman Spectroscopy, First 2 4 3

Principles Calculations and X-ray Diffraction

Introduction: NaZr (PO ) (NZP) is the prototype of a 2 4 3

broad family of compounds with a framework structure that are highly stable and flexible. Corner sharing PO 4

tetrahedra and ZrO octahedra form this framework, with 6

the Na ions placed in the interstitial sites. NZP family of compounds are well known for their ultralow thermal expansion properties, superionic conductivity, and they have also been considered as a host for nuclear waste immobilization due to their ability to accommodate most of the radioactive ions of widely different sizes at the interstitial Na-sites or at the octahedral Zr- sites. High temperature x-ray diffraction and differential thermal analysis on NZP showed the absence of phase transitions up to ~1000 K. However, there were no reports of high pressure investigations of NZP in the literature.

NZP exhibits anisotropic thermal expansion, with α = c-6 -1 -6 -123.5x10 K and α = -5x10 K , with an overall low a

-6 -1positive thermal expansion coefficient α =4.5x10 K av

from 293-1273 K. There are no previous studies of the phonons and vibrational modes responsible for this behaviour. This is a report on our recent investigations of the contribution of the various phonon modes to the thermal expansion of NZP through experiments, calculations using density functional theory (DFT), and

79


thermal expansion. High pressure XRD confirms two phase transitions around 5 and 6.7 GPa which are indexed to a rhombohedral (R3) and an orthorhombic phase respectively. Raman spectroscopy confirms these two phase transformations. The behaviour and nature of phonon modes of the high pressure phases could be studied. Two soft E modes in the ambient phase are found g

to be responsible for polyhedral tilt transition at 5 GPa. The high pressure structure is found to remain stable up to 20 GPa, in contrast to other related framework structures that exhibit negative thermal expansion and pressure induced amorphization. Density functional perturbation theoretical calculations of phonon spectrum of NZP were carried out using VASP code. Mode assignments were carried out using Phonopy combined with VASP computed zone-centre phonon frequencies and eigen

-6 -1vectors. Thermal expansion coefficient (7.510 K ) obtained from the computed quasi-harmonic mode frequencies is in good agreement with the reported experimental value.

More details can be obtained from [17-18].

contribute positively to thermal expansion. Negative contribution to thermal expansion is by the two low

-1frequency modes at 72 and 112 cm that correspond to librations of PO tetrahedra and Zr translation 4

respectively, reminiscent of other NTE systems such as Zr(WO ) and Zn(CN) .4 2 2

The phonon spectrum was computed at a lower volume of 3512.96 Å (corresponding to 2.65 GPa) in addition to that

at equilibrium volume. The Γ-point phonon frequencies at this volume and the equilibrium volume were employed to obtain the mode Gruneisen parameters. The γ for all 70 optical modes except the lowest energy i

-1Raman active E band (ω = 113 cm , ω = 72 g calculated experimental -1cm ) mentioned above are positive. Using Einstein's

specific heat equation for the various modes (i=1 to 70) the total specific heat C was obtained. Thermal V

expansion coefficient α=(γ C )/(3V B ), where av V m 0

γ =½Σp C γ /C , p are the degeneracies of the respective av i i i V i i

phonon branches at the Brillouin zone centre, V is the m

molar volume [15-16]. The bulk modulus B was 0

computed by fitting the DFT-calculated energy vs. volume data to the Vinet universal equation of state. Our calculations involve a set of seven energy-volume data and results in a B = 44.7 GPa.0

We also performed in-situ high pressure x-ray diffraction measurements on NZP to obtain the bulk modulus and to examine the phase transformation indicated by Raman spectroscopy. Up to 4.8 GPa the patterns could be indexed to a hexagonal structure R-3c with monotonically decreasing volume. Several new peaks (at

O2θ=9.8, 11.2, 15.1, 21.1 ) emerge around 5.4 GPa, some old peaks (2θ=16.1, 24.9) disappear, and the pattern is quite different from the starting phase around 9.1 GPa. POWD analysis of these new patterns result in an orthorhombic structure, with a 5% reduction in volume around 5.4 GPa. Later, a more detailed study of the nature of the structural phase transformation and stability under pressure using a combination of x-ray diffraction at BL-11 in Indus-2, RRCAT, and Raman spectroscopy revealed that there are two structural phase transformations in NZP at close intervals of pressure at 5 and 6.7 GPa (Fig.7). Phase II between 5 and 6.7 GPa could be indexed to R3 space group and above 6.7 to an orthorhombic space group Pbcn.

Conclusion: To conclude, we have investigated the thermal expansion behavior of NaZr (PO ) using Raman 2 4 3

spectroscopy as a function of pressure at ambient temperature; the individual modes that make negative and positive contributions to thermal expansion in this material are identified, indicating that the PO tetrahedral 4

librations and Zr translations contribute to negative

Fig. 5: Raman Spectra of NZP at different pressures. The three spectra at the top are in the pressure reducing run. There is a dramatic change in the spectra at 5.5 GPa, and the change is reversible on reducing the pressure.

-1Fig. 6: Atomic displacements of (a) the 113 cm mode showing a combination of the PO librations and Zr translations; (b) the 4

-1127 cm mode exhibiting coupled rotation of PO tetrahedra 4

and ZrO octahedra6

80


for each spectrum varied between 1 and 60 s depending on the signal strength.

One example of the study of a Mogul painting (Figure 8) is presented here:

Figure 9 shows the spectrum at various points on the above painting:

-1 • Spectrum A: spot A: 1345, 1460 and 1584 cm : Indian-yellow: MgC H O .5H O)19 16 11 2

-1• B: 121 and 548 cm : red-lead (Pb O )3 4

-1• C and D:139 and 277 cm : Massicot: Reddish yellow mineral form of PbO

-1• E: 1331 cm : cobalt-yellow (K [Co(NO ) .xH O])? 3 2 6 2-1BUT 305 and 821 cm absent. So, it is not KCoNO.

-1• 1446 and 1622 cm : Studio Red: Synthetic pigment(year 1860): Border might have been 'restored'?

Study 2: The Bhimbetka rock shelters are an archaeological World Heritage site located in the central Indian state of Madhya Pradesh. Earlier studies have suggested that some of these shelters were inhabited by hominids like homo erectus more than 100,000 years ago. Some of the Stone Age rock paintings found among the Bhimbetka rock shelters are approximately 30,000 years old (Paleolithic Age).

Raman Spectroscopic Studies on Ancient Indian Art, Off-site and On-site

Two sets of works carried out by us will be presented in this talk: (1) A micro-Raman study of three miniature

thpaintings of 17 century from the collections of the Madras museum. This is the first study of its kind in India. (2) on-site and off-site Raman spectroscopic investigations of ancient rock-paintings in the World Heritage site Bhimbetka near Bhopal, India. This again is the first study of its kind in India, and the only the third on-site study of ancient paintings in the world. The Raman spectra are assigned and the pigments are identified. A comparison is made between the pigments used in these paintings. Based on the history of pigments present, the possibilities of retouching/repainting and issues relating to the dating/authentication are discussed.

Study 1: Three miniature paintings, two belonging to the Rajput school and the third belonging to Mogul school were were studied. The paintings were removed from the glass frame and placed on a precision x-y translation in a micro-Raman spectrometer (LabRam-HR800, Horiba-Jobin-Yvon). Raman spectra of pigments were recorded from several spots on the paintings. Different laser wavelengths such as 633, 488 and 785 nm were used for the study. Laser powers were set between 2 and 15 mW so as to avoid damage due to laser heating. The recording time

Fig. 7: Pressure dependence of volume per NZP formula unit (Z). The solid lines show the fitting with the third order Birch Murnaghan equation of state for the three different phases.

Fig. 9: Raman spectra of pigments at spots A-E in Fig. 1

thFig. 8: 17 century Mogul painting: Rani Durgavati

Fig. 10: Flower with white and yellow petals

81


References

1. W. P. Crummet, H. G. Smith, R. M. Nicklow and N. Wakabayashi, Phys. Rev. B 19, 6028 (1979)

2. K. T. Moore, Rev. Mod. Phys. 81, 235 (2009)

3. A. K. Rai, S. Raju, B. Jayaganesh, E. Mohandas, R. Sudha, and V. Ganesan, J. Nucl. Mater. 383, 215 (2009)

4. G. C. Allen, I. S. Butler, and N. A. Tuan, J. Nucl. Mater. 144, 17 (1987)

5. L. C. Hoskins, J. Chem. Educ., 52, 568 (1975)

6. D. Manara and B. Renker, J. Nucl. Mater. 321, 233 (2003)

7. T. R. Ravindran and A. K. Arora, Journal of Raman Spectroscopy 42, 885 (2011)

8. D. J. Williams, D. E. Partin, F. J. Lincoln, J. Kouvetakis, and M. O'Keefe, J. Solid State Chem. 134, 164 (1997).

9. T. R. Ravindran, A. K. Arora and T. A. Mary, Phys. Rev. Lett. 84 3879 (2000)

10. T. R. Ravindran, A. K. Arora and T. A. Mary J. Phys.: Condens. Matter 13, 11573 (2001)

11. T. R. Ravindran, A. K. Arora and T. A. Mary Phys. Rev. B 67, 064301 (2003)

12. A. L. Goodwin and C. K. Kepert, Phys. Rev. B. 71, 140301(R) (2005).

13. B. F. Hoskins and R. Robson, J. Am. Chem. Soc. 112, 1546 (1990)

14. T. R. Ravindran, A. K. Arora and T. N. Sairam, J. Raman Spectrosc. 38, 283 (2007)

15. K. Kamali, T. R. Ravindran, C. Ravi, Y. Sorb, N. Subramanian and A. K. Arora, Phys. Rev. B 86, 144301 (2012)

16. K. Kamali, T.R. Ravindran, N.V. Chandra Shekar, K.K. Pandey, S.M. Sharma, Journal of Solid State Chemistry 221 285 (2015)

17. T.R. Ravindran, A.K. Arora, S. Ramya, R.V. Subba Rao, and Baldev Raj, Journal of Raman Spectroscopy 42, 803 (2011)

18. T. R. Ravindran, A. K. Arora, M. Singh, and S. B. Ota, Journal of Raman Spectroscopy 44, 108 (2013)

Raman spectroscopic studies on some of these paintings were carried out on-site using a portable Raman spectrometer (BWTek: Model InnoRam) with a 785 nm diode laser for the excitation of the spectra. In addition, tiny fragments of pigments (100-200 micrometer in size) extracted from some of the artworks were also studied in laboratory using a micro-Raman spectrometer and analyzed using energy-dispersive x-ray analysis for elemental composition. Based on the Raman spectra and the elemental analysis mineral-based pigments such as calcite, gypsum, hematite, whewellite, and goethite could be identified.

A comparison of the spectra recorded on-site using a light-weight portable spectrometer with those using laboratory equipment will also be made and discussed.

• Raman spectrum recorded with a micro Raman spectrometer off-site shows the presence of α-quartz (SiO ), gypsum (CaSO .2H O) and calcite 2 4 2

(CaCO )3

• in-situ spectrum shows only the gypsum band

• Presence of both gypsum and calcite in the spectra of extracted sample and absence of calcite in the on-site spectrum (Fig. 3b) suggests that gypsum is present only as a weathering product in the top layer.

Conclusion: Raman spectroscopic analyses show that thIndian artists of 17 century were familiar with a number

of mineral-based pigments such as red-lead, lead-white, vermilion and litharge. Although the anatase phase of TiO2 is considered as a rare mineral and its first synthesis dates to 1908, it is found to be present in one of the paintings. Raman spectroscopic studies carried out on the pigments used in the paintings in three of the rock shelters of Bhimbetka have revealed the presence of minerals such as calcite, hematite, gypsum, whewellite, and goethite. An evidence of the presence of a possible organic binder is also found in one of the extracted pigment sample. This is the first report of any on-site Raman study on a pre-historic site in India that contains rock paintings starting from the period of the cave dwellers to recent times.

More details can be found in [17-18].

Acknowledgements

Financial support from Department of Science & Technology, Govt. of India is gratefully acknowledged to carry out the paintings Project. We thank Dr. S. B. Ota, Regional Director – Central Zone, Archaeological Survey of India, and Dr. Manager Singh, Scientist, ASI, Aurangabad, for fruitful technical collaboration.

82


Full Name :

Address for correspondence :

PIN Code

Tel. No. Fax No.

E-mail Address :

Other Address (Res./Office) :

PIN Code

Tel. No. Fax No.

Date of Birth :

Academic Qualifications :and Award / Honours received

Present Position :

Fields of Specialization :

Type of membership requested : Life (Fee : ` 1,500/-For Indian Residents ; ` 5,000/- For Non Residents) / Corporate (Fee : ` 10,000/-) / Student (Fee : ` 500/-)

Any particular field in which : Writing articles / Giving popular talks / Local organization / Othersyou would like to contribute (Please Specify)to ILA activities

Membership Payment : Cheque# / Bank Draft No.

DATE SIGNATURE

Send completed application form along with payment to : General Secretary II, ILASolid State Laser Division Raja Ramanna Centre for Advanced TechnologyPO : CAT, INDORE - 452 013 (M.P.)E-mail : [email protected]

# Make Cheque / Draft payable to Indian Laser Association. Drafts should be payable at Indore.# For outstation cheque please add Rs. 35 upto 1000/- rupees and additional Rs. 4.5 per extra thousand rupees for bank charges. Combined payment is acceptable.

FOR ILA OFFICE

Membership type and No. :

Membership Receipt No. :

Any other remarks : (General Secretary II)

MEMBERSHIP FORM

INDIAN LASER ASSOCIATION

Fig. 8 on page no. 38 Fig. 6 on page no. 26



Issues of Kiran and Membership form are available at www.ila.org.inFor circulation among ILA members only.

(Not for sale)Printed by : Burhani Offset Printers, Indore. M- 99811-74052.A Bulletin of the Indian Laser Association


Fig. 2 on page no. 6


Fig 1 on page no. 21

based on invited talks during dae-brns national …ila.org.in/kiran/kiran_25_03.pdf1 vol. 25, no. 3,...

Documents