applications of excimer laser in nanofabricationchouweb/publications/211 xia... · applications of...

Appl Phys A (2010) 98: 9–59DOI 10.1007/s00339-009-5455-0

I N V I T E D PA P E R

Applications of excimer laser in nanofabrication

Qiangfei Xia · Stephen Y. Chou

Received: 7 October 2009 / Accepted: 15 October 2009 / Published online: 5 November 2009© Springer-Verlag 2009

Abstract This paper addresses novel applications of anexcimer laser (308 nm wavelength, 20 ns pulse duration)in nanofabrication. Specifically, laser assisted nanoimprintlithography (LAN), self-perfection by liquefaction (SPEL),fabrication of metal nanoparticle arrays, and the fabricationof sub-10-nm nanofluidic channels are covered. In LAN, apolymeric resist is melted by the laser pulse, and then im-printed with a fused silica mold within 200 ns. LAN hasbeen demonstrated in patterning various polymer nanos-tructures on different substrates with high fidelity and uni-formity, and negligible heat effect on both the mold andthe substrate. SPEL is a novel technology that uses selec-tive melting to remove fabrication defects in nanostructurespost fabrication. Depending on the boundary conditions,SPEL is categorized into three basic types: Open-SPELthat takes place with surface open, Capped-SPEL wherea cap plate holds the top surface of the nanostructuresand Guided-SPEL where a plate held a distance above thestructure guides the molten materials to rise and form anew structure with better profile. Using SPEL (in less than200 ns), we have achieved a reduction of line edge rough-ness (LER) of Cr lines to 1.5 nm (3σ ) (560% improve-ment from the original), which is well below what the pre-vious technologies permit, and a dramatic increase of theaspect ratio of a nanostructure. We have used SPEL to makesub-25-nm smooth cylindrical NIL pillar molds and smooth-ing Si waveguides. Excimer laser is also used to make metal

Q. Xia current address: Information and Quantum Systems Lab,Hewlett-Packard Laboratories, 1501 Page Mill Road, Palo Alto, CA94304, USA.

Q. Xia (�) · S.Y. ChouNanostructure Laboratory, Department of Electrical Engineering,Princeton University, Princeton, NJ 08544, USAe-mail: [email protected]

nanoparticles. Monolayers of particles are fabricated on var-ious substrates (silicon, fused silica and plastics) by expos-ing thin metal films to a single laser pulse. Periodic nanopar-ticle arrays have been fabricated by fragmentation of metalgrating lines. The periodicity of these nanoparticles can beregulated by surface topography such as shallow trenches.Finally, an excimer laser pulse has been used to melt the topportion of 1D and 2D Si gratings to seal off the top surface,forming enclosed nanofluidic channel arrays. The channelwidth has been further reduced to 9 nm using self-limitedthermal oxidation. DNA stretching using 20 nm wide self-sealed channels is also demonstrated.

PACS 06.60.Jn · 81.07.-b · 81.16.Rf · 81.16.Nd · 87.85.Rs

1 Introduction

Since its invention in 1960 [1, 2], laser has been widely usedin many fields, such as scientific research, medical treat-ment, industry and military. Excimer lasers are an importantfamily of gas lasers that were firstly commercialized in the1970s [3]. The most common excimer lasers use rare-gasmonohalides such as KrF, XeCl, and have a wavelength inultraviolet (UV) range.

The applications of excimer laser are very broad withnew applications continuously coming out. This paper is notintended to give an overview on this expanding field butrather focuses on certain novel applications in nanofabrica-tion that have been developed by the authors at PrincetonUniversity as part of one of the author’s PhD thesis [4].

This paper is organized as follows: after a brief intro-duction of our general experimental setup, we will discuss(1) laser assisted nanoimprint lithography (LAN), an ultra-fast imprint process that patterns nanostructures in polymers

mailto:[email protected]

10 Q. Xia, S.Y. Chou

within 200 ns with negligible heating effect on both the moldand substrate; (2) SPEL, which uses the excimer laser to re-move fabrication defects and to enhance the nanostructureprofile; (3) the fabrication of metal nanoparticle monolayersand periodic metal nanopaticle arrays using a single excimerlaser pulse exposure in combination with nanoimprint litho-graphy (NIL); and (4) the fabrication of sub-10-nm nanoflu-idic channels using laser melting and self-limiting thermaloxidation, along with the application of the channels instretching λ-phage DNA molecules.

2 Laser setup and molten time measurements

We use a XeCl excimer laser in this study (Lambda Physik,model number COMPex 102). The laser tube is filled to3200 mbar with premixed gases. The gas composition (pres-sure percentage) is as follows: 0.13% HCl, 0.02% H2, 2.35%He, 1.87% Xe, and Ne as the balance. The maximum outputenergy is 200 mJ/pulse and the maximum repetition rate is20 Hz.

Figure 1 shows the main experimental setup used through-out this study. After a laser pulse comes out of the laser tube,a beam delivery system (BDS) directs the laser beam to thesample. This BDS consists of two mirrors, an attenuator,a homogenizer, a condenser lens, a metal aperture, and animaging doublet. A mirror at one end of the optical trackreflects the laser pulse coming out of the tube to an atten-uator, which is used to modulate the laser energy continu-ously from 10% to 90%. The homogenizer, together withthe condenser, is used to generate a mesa-shaped spatial in-tensity profile to achieve much more uniform distribution.The spatial plane of this intensity profile is called the ho-mogenized plane. A metal aperture is mounted at the ho-mogenized plane to block off the edges of the beam and sidelobes, resulting in a nearly flat top spatial intensity profilefor the laser pulse. Another mirror is placed at the other endof the optical track to direct the modified laser pulse to the

Fig. 1 Excimer laser setup used for this paper. This setup includes anexcimer laser, a beam delivery system, a cw HeNe monitoring laserand a sample stage

imaging doublet, which consists of a set of objective lensesthat focuses the energy into a smaller spot so as to achievethe required energy intensity for melting different materials.

An x–y–z stage is used to mount the samples. In mostof our experiments, the stage is moved manually although amotor-driven stage can be used when precise alignment inlaser scanning is required. A continuous wave (cw) HeNelaser (λ = 633 nm, maximum output power 1 mW, beamdiameter 0.5 mm) is used as a monitor laser. This monitorlaser serves several purposes: (1) To align the laser pulseand the sample; (2) To act as a light source for time-resolvedreflectivity measurement and other in situ measurements asdescribed later in the text.

One important parameter in laser melting is the moltentime of a material for a given laser fluence. To measure themolten time, time-resolved reflectivity (TRR) measurementswere carried out for several semiconductors. The setup isshown in Fig. 2. The HeNe laser beam is incident on a spotof the surface to be melted by the excimer laser. The re-flected signal from the surface, which is much higher inmolten material than its solid phase for semiconductors wascollected by a high speed Si p-i-n detector (Newport, 818-BB-21A, rise time <500 ps, fall time <500 ps) with a bandpass filter (Andover Corp, 633FS02-25, center wavelength633 nm, bandwidth 1.0±0.2 nm) mounted in front of the de-tector. The signal is sent to an oscilloscope (Tektronix, TDS220, bandwidth 100 MHz, sample rate 1 GS/s) for analysis.The detector and oscilloscope together have a time resolu-tion better than 10 ns.

Figure 3 shows a typical reflectivity curve during meltingsilicon captured from the oscilloscope screen. In the solidstate, the reflectivity stays low and constant. When the UVlaser pulse reaches the Si surface, it melts the surface layerimmediately and the reflectivity jumps to a much highervalue within 1 ns [5, 6]. When the surface layer is at a moltenstate, the reflectivity stays constantly high before droppingto its previous level upon resolidification.

Using this technique, the molten times for different mate-rials at different excimer laser fluence have been measured.Figure 4 shows the molten time as a function of laser flu-ence for different types of semiconductors including a bareSi wafer (p-type, (100) orientation, resistivity 10–20 � cm),an SOI wafer (SIMOX, 100 nm thick (100) silicon devicelayer and 380 nm thick buried oxide), a Ge wafer, and SiGeon Si or SiO2.

The molten time increases with laser fluence for all cases.This is due to the fact that with a higher laser fluence (be-low the ablation threshold), a thicker layer is melted whichneeds a longer time to re-solidify. It is also interesting to notethat under the same laser fluence, the molten time for SOI islonger than that for silicon. This trend is also observed forSiGe on SiO2 as compared to that on Si. This is believed to

Applications of excimer laser in nanofabrication 11

Fig. 2 Schematic illustration ofthe TRR setup. A HeNe laserbeam (λ = 633 nm) is incidentat an angle on the UV laserexposed surface of asemiconductor. The reflectedsignal passes through a red filterand is collected by an ultrafastphotodetector. The capturedsignal is analyzed by anoscilloscope

Fig. 3 A typical TRR curve captured from an oscilloscope screen. Thereflectivity stays low when Si is at solid state, then jumps to a muchhigher value within 1 ns when the surface is melted. After a certainperiod of time, it drops to its normal value when the Si resolidifies.The width of the curve is defined as the molten time, which is about70 ns in this case. The tail in the curve is probably due to the slow cooldown process at the low temperature end in the solid state

result from the fact that the oxide layers have acted as ther-mal barriers, making the heat dissipation speed into the bulksubstrate slower than the substrate without a SiO2 barrier.

In addition to semiconductors, the TRR technique is alsosuitable for measuring molten time of metals and other ma-terials which possess a difference in reflectivity betweentheir solid and liquid phases. For example, using a 633 nmwavelength HeNe laser, the molten times for Al, Al-basedalloys, and TiN-based anti-reflective coatings have beenmeasured [7]. However, the difference in the reflectivity atsolid and liquid states for metallic materials is not as signif-icant as for semiconductors. For example, upon melting Al,

the reflectivity decreases from 92% to 85%, while that forSi increases from 34.7% to 73.4% [7]. Note that the reflec-tivity for molten metal is lower than that of its solid phase(rather than higher as in the case for semiconductors). Thisis probably due to the phonon scattering of the incident lightin the metal examples, which is more noticeable when thetemperature increases. This scattering results in a lower re-flectivity [5].

3 Laser assisted nanoimprint lithography (LAN)

In traditional thermal NIL, both the mold and substrate areheated with the resists [8–10]. When the mold and substratesare made from different materials, the thermal expansiondifference will cause misalignment between the two. For ex-ample, the thermal expansion coefficients of Si and SiO2

are 2.6 × 10−6 and 0.5 × 10−6 (◦C)−1, respectively [11]. Ifa SiO2 mold and a Si substrate are used during imprintingat 100◦C, the difference in thermal expansion would be 2.1×10−4, leading to a global misalignment of about 21 µm fora 4-inch wafer. Although a mold made from the same mate-rial as the substrate has been demonstrated to minimize thethermal expansion difference [12], this is not practical sincenot every substrate material can be used for mold. A methodthat can greatly reduce the heating of the substrates and themolds is indispensable.

A new development in NIL is using a laser pulse as theheating source [13]. In this section, an excimer laser pulseis used to melt the polymeric resists in NIL. Since thereis no direct heating of the underlying substrate, laser as-sisted nanoimprint lithography (LAN) reduces the thermalexpansion effects. LAN is demonstrated to be capable ofproducing nanostructures in various polymers on different


Fig. 4 Molten time measuredby TRR technique for Ge, Si,SOI, and SiGe on Si and SiO2.(a) (100) Ge wafer. (b) Si andSOI, the SOI wafer has a devicelayer of 100 nm and buriedoxide layer of 380 nm. The Si isa bare p-type (100) wafer.(c) SiGe on SiO2 and Si, theSi0.6Ge0.4 is 74 nm thick, andthe oxide layer is 170 nm thick.Dots are measured data andlines are a linear fit for the data

substrates using only a single light pulse. The imprint timefor LAN is less than 200 ns, which is measured using real-time monitoring by scattering of light (RIMS) [14]. Numer-ical simulation shows negligible heating of the substrate andmold, and little distortion of both.

3.1 Principle of LAN

The principle of LAN is shown in Fig. 5.A resist thin film is spin coated on a substrate (Si or

quartz). The resist-covered substrate is then baked at 60–70◦C on a hot plate for about 30 minutes to drive out the

residual solvent. A fused silica (FS) mold with nanostruc-tures is applied against the resist film with some pressure.A single XeCl excimer laser pulse passes through the moldand melts the polymer film, during which time the mold isimmediately imprinted into the resist. The mold is then sep-arated, leaving a negative pattern in the resist.

Based on the working principle, it can be seen that LANis an ultrafast process, taking place in nanoseconds. Sincethe laser pulse is so short, heating of the substrate is neg-ligible. The mold does not directly absorb the laser energybecause its bandgap (8 eV) [11] is much larger than the pho-ton energy of the 308-nm wavelength UV laser (4 eV). The


Fig. 5 Schematic illustration of LAN principle. (a) Bring mold andresist into contact under pressure, (b) UV laser irradiation; the resistmelts upon laser exposure, (c) mold is pressed into resist layer, (d) re-sist becomes rigid again; (e) mold separation. The mold is made fromfused silica, which is transparent to the laser pulse. A single laser pulseis enough for the imprint process

same principle may work for UV-NIL as well since manyUV-curable resists will cross link at this wavelength.

3.2 Fabrication of imprint mold

Fused silica (FS) is used as the mold material for several rea-sons. It is transparent to the UV laser. The reflectivity at theFS/air interface is as low as 3% at normal incidence. FS isalso mechanically strong, thermally and chemically stable,and does not react with the resists. The processing ability ofFS using standard clean room techniques also favors it as anideal candidate for mold material.

In mold fabrication, the original master mold was pat-terned using interference lithography, followed by patterntransfer into silicon with thermally-grown oxide or bare sil-icon wafers [15]. 100 nm half-pitch gratings on a Si mastermold were transferred to a 2-inch fused silica wafer usingtraditional thermal-NIL and reactive ion etching (RIE). TheRIE was carried out in a Plasma Therm 2486 machine with10 sccm CHF3, 1.5 sccm O2 at power density of 60 mW/cm2

with a pressure lower than 3 mTorr. Cr was used as the etch-ing mask since it has very high selectivity to FS (>10:1).The etching rate was about 3–4 nm/min and the as-etchedFS trenches (depth = 90 nm) have vertical sidewalls andflat bottoms. After stripping off the Cr using a CR-7 so-lution, the etched fused silica wafer was then diced into1 × 1 mm2 pieces using a wafer dicing saw. The dicedmolds were cleaned in a solution (NH4OH:H2O2:deionizedH2O = 1:1:5) at 80◦C for 15 min (RCA #1 cleaning), fol-lowed by treatment with an anti-adhesion agent (1H, 1H,2H, 2H-perfluorodecyltrichlorosilane, FDTS, Alfa Aesar)for better mold release during the separation. The finishedmolds have 200 nm period gratings (100 nm line/spacing,90 nm trench depth) over the entire mold area. A scanningelectron microscope (SEM) image of the mold is shown inFig. 6 [16].

Fig. 6 SEM image of the 200 nm period grating mold made fromfused silica for LAN. The mold is made by NIL and RIE. The gratingshave a line width of 100 nm and height of 90 nm [16]

Fig. 7 SEM image of NPR-69 gratings on a Si substrate by LAN usinga single laser pulse with a fluence of 400 mJ/cm2 [16]

3.3 Imprint results

A single laser pulse was sufficient to imprint the nanopat-terns of the mold into the resist with high fidelity. For ex-ample, with a laser fluence not lower than 350 mJ/cm2, uni-form gratings of the 200 nm period in the mold were trans-ferred completely to a thermoplastic resist, NPR-69, whichhas a glass transition temperature of about 100◦C. The im-printed gratings had vertical sidewalls and flat surfaces andbottoms (Fig. 7) corresponding to those in the mold. Ex-cellent pattern transfer by LAN was also observed in otherresists, but at different laser fluences depending on the glasstransition temperature and laser absorbance of the resist. Forexample, with NPR-46, a single laser pulse with a fluence of560 mJ/cm2 was used to duplicate the pattern from the moldto the resist (Fig. 8).

Resists on different substrates were also successfully im-printed by LAN. For NPR-69 on a quartz substrate, wefound that at least 500 mJ/cm2 laser fluence was needed for agood imprint, higher than that on a Si substrate (Fig. 9). The


difference in required fluence is attributed to the differencein the laser absorbance of the resists on different substrates(21.4% for a 200 nm NPR-69 on a Si substrate, and 14.3%for that on a quartz substrate). This will be discussed in de-tail in Sect. 3.4. The other possible reason for the differentfluences is that the quartz substrate does not absorb the 308-nm laser pulse, but a Si substrate does and transfers the heatto the resist on top of it. Hence the incident energy was usedmore efficiently with resists on the Si substrate.

Fig. 8 Atomic force microscope (AFM) image of NPR-46 gratingsproduced on a Si substrate. The height of the gratings is 90 nm, whichcorresponds exactly to the depth of the trench in the mold. A singlelaser pulse with a fluence of 560 mJ/cm2 was used [16]

Fig. 9 SEM image of NPR-69 gratings made on a quartz substrate byLAN with a single laser pulse of 500 mJ/cm2 [16]

In LAN, the molds with anti-stick coating could be usedfor several times without loss in grating quality during im-print. For example, the 1st and 3rd imprint with the samemold—without in between mold cleaning—resulted in thesame quality grating lines of NPR-46 (Fig. 10).

The imprinted area could be as large as that of the moldand was quite uniform. An imprinted spot in NPR-69 wasexamined under SEM after LAN at five different locations(four corners and one in the center), as shown in Fig. 11.From the top view SEM images, it can be seen that theimprint quality is high and uniform without noticeable de-fects. The pitch of the imprinted gratings was analyzed byfast Fourier transform (FFT) analysis using image process-ing software (Image Pro Plus, from Media Cybernetics, Inc.,MD), as shown on the right panel of Fig. 11. From theFFT images, it can be concluded that the pitch of the im-printed gratings at the five different spots is very uniform.This demonstrates both the uniformity of the imprint duringLAN and the uniformity of the mold fabrication process forLAN.

3.4 Modeling of laser absorption in resists during LAN

Light absorption by the resist and substrate can be cal-culated using the following model. For example, with aSiO2/polymer/substrate sandwich, the light path in the sys-tem can be depicted as in Fig. 12.

A simple case for this problem is when the substrate andthe mold are both made of the same materials, such as SiO2.Consider a normal incidence (θi = 0), the refractive indicesfor both the mold and the substrate are equal and they areboth real numbers. This is similar to a Fabry-Perot etalon[17] except that the system here absorbs energy in the in-termediate layer. In this case, the fractional absorption IA

Ii

is [4]

IA

Ii

= 1 − ARA∗R

AiA∗i

− AT A∗T

AiA∗i

, (1)

where Ai , AR and AT are the amplitude of the incident,reflected and transmitted light, respectively.

Fig. 10 SEM images of 200 nmNPR-46 period gratingsproduced on a Si substrate by asingle laser pulse (560 mJ/cm2)

using the same mold.(a) Imprint result after mold isused for the first time and(b) after the third time. There isno obvious difference in gratingquality in the two pictures [16]


Fig. 11 SEM images (leftpanel) and corresponding FFTimages (right panel) fordifferent locations of animprinted NPR-69 spot on a Sisubstrate. The locationsexamined are: (a) Upper leftcorner, (b) upper right corner,(c) center, (d) lower left cornerand (e) lower right corner. Thegratings are uniform over thewhole imprinted spot which isabout 1 mm by 1 mm


Fig. 12 Schematic light path during the LAN process (θi = 0 in LAN).The incident light passes through the transparent mold and reaches theresist layer and the substrate. Part gets reflected back, part is absorbedby the resist layer, and part goes into the substrate

Fig. 13 Calculated fractional absorption of a NPR-69 resist with anabsorption coefficient of 8 × 10−4 nm−1 on silicon and quartz (SiO2)substrates. On a Si substrate the resist absorbs more energy than on aquartz substrate under otherwise identical conditions

However, when the substrate is Si, the above equation(1) does not apply since in this case, n1 �= n3. Also at the308-nm wavelength, the refractive index for Si is a complexnumber. In order to solve the problem, commercially avail-able software such as GSolver (Grating Solver DevelopmentCompany, TX [18]) was used for the calculation.

For NPR-69 resist which has an absorption coefficient of8 × 10−4 nm−1, the fractional absorption as a function offilm thickness is plotted in Fig. 13 for both quartz and Sisubstrates. For a 200-nm thick NPR-69 film, the fractionalabsorption on quartz and silicon is 14.3% and 21.4%, re-spectively. In this calculation, the indices at this wavelengthare 1.49 for SiO2 [19]; 5.015 + 3.650i for Si; and 1.57 forNPR-69.

3.5 Simulation of the substrate/mold heatingand deformation

For the molds and substrates made from SiO2, the temper-ature remains nearly unchanged during LAN because heatconduction is poor (two orders of magnitude less than Si)[20] and there is no direct absorption of laser energy. For Sisubstrates, the heating is also greatly reduced since LAN hasa much shorter processing time (hundreds of nanoseconds)than traditional thermal NIL. To better understand the heat-ing of Si substrates, numerical simulation using ABAQUS(ABAQUS Inc., RI [21]) was carried out. Also simulatedwas the deformation of the mold and substrate during theLAN process.

The case simulated was that of a 200-nm thick NPR-69 on a Si substrate exposed to a laser pulse with a flu-ence of 400 mJ/cm2. In this simulation, the thermal diffu-sion length for Si was calculated using tabulated physicalproperties [20] to be 1.3 µm for a 20 ns laser pulse. Thetemperature profile at 2 µs after the laser pulse reached thesurface is shown in Fig. 14. It can be concluded that thethermally affected zone is mainly localized at the Si surface.This is reasonable since Si is a much better thermal conduc-tor than SiO2. Also, only a very thin surface layer in Si isthermally affected, while the bulk of the substrate remainsat room temperature. This is beneficial to the inhibition ofthermal expansion since the bulk Si wafer at room tempera-ture serves as a constraint, limiting the thermal expansion.

To better visualize the temperature distribution, temper-ature as a function of time for different distances from thesurface is plotted in Fig. 15. It illustrates how the tempera-ture drops quickly with distance from the surface.

For example, after a certain period of time, the tempera-ture at a depth of 14.3 µm is below 50◦C. Only a thin layernear the surface (about 3% of a 500 µm thick Si wafer) wasthermally affected during the process, while the bulk of theSi wafer remained at nearly room temperature. The figurealso illustrates how the Si surface temperature drops be-low the polymer glass transition temperature (about 100◦C)within 500 ns, which suggests that the whole imprint processtime was shorter than 500 ns. It should be pointed out herethat the heating of the Si substrate could be further reducedby tailoring the resist’s chemical composition to increase itslaser absorbance (see Sect. 3.6 for example).

Because the bulk of the Si wafer is at room tempera-ture during the imprint process, the lateral thermal expan-sion of the surface layer is efficiently constrained. This sup-ports good overlay alignment. Numerical simulation of thestress distribution shows that the stress is localized at thecorner of the mold/substrate interface. The maximum prin-cipal stress is about 15.3 MPa for an extreme high pressurecase (2700 psi applied pressure). This is far less than theYoung’s Modulus of Si (107 GPa) and SiO2 (70 GPa) [11].


Fig. 14 Temperature profile2 µs after the laser pulse in atypical LAN process. The laserfluence is 400 mJ/cm2. From thecross section-temperaturedistribution, the highest surfacetemperature is about 60◦C after2 µs, indicating a fastheating/cooling process

Fig. 15 Simulation results of temperature evolution in the surfacelayer of a Si substrate after the incidence of a single laser pulse(400 mJ/cm2). Each curve represents the temperature as a functionof time at different distances from the Si surface. The inset showsthe model geometry in which we assume a 200 nm polymer film ona 500 µm thick Si substrate [16]

Numerical simulation of the deformation of a Si substrateshows that the maximum lateral displacement is also limitedto the corner area within 1 µm range (Fig. 16). The max-imum in-plane principal strain is about 5 × 10−4 (0.05%),thus the local displacement of the substrate and the molddue to the pressure is less than a nanometer.

3.6 Resist engineering for LAN

Commercial polymers such as polystyrene (PS, molecularweight 67.5 K), poly (methyl methacrylate) (PMMA, Mw15 K), and nanoimprint resists (homemade NPR series) on

different substrates were tested for LAN. Although satisfac-tory results were obtained for all the resists, it is worthwhilenoting that tailoring the chemical composition of the resistscan improve the light absorption, hence further reduce thesubstrate heating. For this purpose, NPR-series resists withvarying amounts of UV light absorber have been developed.

One of the best UV absorbers is pyrene (98%, Aldrich),which shows strong absorbance from 200–350 nm as ourUV light absorber. The absorption spectrum of pyrene is(98%, Aldrich) is shown in Fig. 17 [22], with the mole-cular structure depicted in the inset. To prepare the re-sist, the chemical ingredients (pyrene, commercial polymerpowders such as PS, etc.) were dissolved in cholorobenze(HPLC grade, Aldrich). The amount of pyrene used for thisstudy was 1–10% of the combinational weight of other solidchemicals. The solution was then stirred with a magneticspin bar at 300 rpm overnight. Thin resist films with differ-ent thicknesses were spin coated on fused silica (FS) wafers,followed by baking on a 70◦C hot plate for 15 min to driveout the residual solvent.

The absorption properties of the resists were character-ized by their absorption coefficients, which were measuredusing a simple set-up illustrated in Fig. 18. In this set-up,the laser pulse was introduced perpendicular to the FS/resistsample from the wafer side. The intensity of the incidentlight (Ii ), and that of the transmitted light after a film ofthickness L (IL) were measured by averaging reading from10 laser pulses.

The relationship between Ii and IL will give us informa-tion about the absorption coefficients (α) according to thefollowing equation:

IL = Ii · exp(−α · L). (2)


Fig. 16 Maximum in-planeprincipal strain distributionduring the LAN process. Itoccurs at the corner of themold/substrate interface. Themaximum value is about5 × 10−4, which is 0.05%

Fig. 17 The absorption spectrum of pyrene dissolved in cyclohexane.The inset shows the chemical structure of pyrene. Reproduced fromreference [22]

Fig. 18 Setup for measurement of resists’ absorption coefficients. The308-nm wavelength light goes in from the fused silica side with a nor-mal incident angle. The light intensity is measured after the resist thinfilm (IL) and directly from the light source (Ii ) for different resist filmthicknesses (L), which are used for calculating the absorption coeffi-cient (α)

Fig. 19 Measured absorption coefficients of polystyrene (Mw 67.5K)doped with varying amount of pyrene. With more pyrene, the absorp-tion coefficient increases noticeably

The addition of pyrene greatly enhanced the absorptioncoefficients of the resist. IL/Ii as a function of thicknessL is plotted in a semi-logarithmic chart for pyrene dopedpolystyrene (Mw 67.5 K) in Fig. 19. The slope of the linearfit line represents the absorption coefficient for each case,which increased significantly with the addition of pyrene.For example, with 5 wt.% pyrene, the absorption coeffi-cient is about 4.32 × 10−4 nm−1, while increasing pyreneto 10 wt.% results in an absorption coefficient of about8.66 × 10−4 nm−1. With the same method, the absorptioncoefficients of other resists were also measured. Similarly,the addition of pyrene has resulted in higher absorbance.

The improvement in the absorption coefficient resultedin less heating of the substrate during LAN. With the dif-


Fig. 20 Absorbance of resists with different absorption coefficients(α) on SiO2 substrates

ference in the absorption coefficients due to the alternationof the chemical composition of the resists, the percentageof the light that is absorbed by the resist layer is going tochange. As a result, the percentage of light that goes into thesubstrate and heats the substrate will change, too. As an ex-ample, Fig. 20 gives the absorption properties (that is, frac-tional absorption intensity, Ia/Ii , where Ia is the absorbedintensity by the resist films) for resists with different ab-sorption coefficients on fused silica substrates (SiO2) calcu-lated using a model developed elsewhere [4]. For a 200 nmthick resist film on silica substrate, an increase of absorp-tion coefficient from 4 × 10−4 nm−1 to 10 × 10−4 nm−1

will result in a fractional absorption increase from 7.5% to18.8%. The thickness of the imprint resists plays an impor-tant role. Under otherwise identical conditions, the absorp-tion is improved with a thicker resist film. These conclusionsalso hold for Si substrates.

3.7 Measurement of imprint time for LAN

It is important to understand the speed of LAN. From thesimulation results (e.g., Fig. 15), the Si surface temperatureremained above 100◦C for 500 ns, which suggests that theimprint should be done within 500 ns since the glass tran-sition temperature of the resist is about 100◦C. However, toget an accurate imprint time, direct measurement is neces-sary.

Recently, real-time monitoring by scattering of light(RIMS) has been proposed and demonstrated for flow char-acterization of both thermal and photocurable polymers ormonomer mixtures in NIL [14, 23]. In RIMS, a surface reliefdiffraction grating is created on the imprint mold and the dif-fracted light intensity from the grating—which depends on

resist filling of the grating trenches—is monitored continu-ously during the imprint process. Previous work has demon-strated that RIMS is capable of monitoring NIL processeswhich finish in less than 1 second [23]. Here, RIMS is usedfor measuring the processing time of LAN.

3.7.1 Principle of RIMS

In RIMS, the intensity of diffracted light depends on the fill-ing ratio of the polymeric resist to the mold trench. As theresist flows into the mold trenches, the diffracted signal de-cays monotonically because the imprint resist has an indexalmost the same as that of the mold (1.46 at 633 nm in ourexperiments). The diffraction will disappear when the resistfills the mold grating trenches completely (Fig. 21).

A rigorous model (scalar diffraction model) [24] has beenapplied to RIMS [14, 25], and it clearly shows the relation-ship between the normalized diffraction intensity and thefilling ratio of the mold trench. It is also possible for RIMSto monitor the case where the refractive indices of mold andresist are not the same. In this case, the intensity does notlinearly decrease with the filling ratio, nor does it reach zerowhen the trench is fully filled with resist due to the refractiveindex contrast.

3.7.2 Experimental details

A schematic of the RIMS experimental setup is shown inFig. 22. The substrate, resist and mold in our experimentswere prepared as follows: A clean silicon substrate with na-tive oxide was spin coated with a 210 nm thick resist that

Fig. 21 Schematic illustration for the operational principle of RIMS.A laser beam is incident on the grating area and the intensity of −1storder diffraction is monitored continuously. Before imprint (a), the in-tensity of the −1st order is strongest. During imprint when the mold ispressed into the resist (b), the intensity is reduced, and when the moldis completely pressed into the resist (c), it falls to zero. In this figure,it is assumed that the refractive indices of the mold and resist are thesame. Reproduced based on description in [14]


Fig. 22 Schematic illustrationof the RIMS experimental setup.A UV laser spot (λ = 308 nm)covers the mold/resist/substratesandwich, which was pressed bytwo parallel metal plates (notshown here). A HeNe laserbeam (λ = 633 nm) is incidenton the mold at an angle of 60◦to the mold surface. The anglebetween the HeNe laser beamincident plane and the gratinglines was 45◦. The diffractedsignal (−1st order) passesthrough a red filter and iscollected by an ultrafastphotodetector. The capturedsignal is analyzed by anoscilloscope

was custom-made by modifying the commercially availableNanonex resist NXR-1023 with a UV light absorber. Theresist film was then baked at 70◦C for 30 min to drive outthe residual solvent. The mold was an optical grade fusedsilica wafer (0.5 mm thick, UV-transparent, refractive index1.46) having parallel 300 nm wide grating lines of 950 nmpitch and 150 nm depth over the entire mold area (1.2 mmby 1.2 mm) fabricated by NIL and RIE. The main reason foretching the mold to 150 nm deep is that at this depth, thediffraction intensity will be monotonously reduced with thefilling of the trench (Fig. 23). This makes the interpretationof experimental data straightforward. The finished mold wascoated with FDTS monolayers for easy mold release afterimprint.

During the imprint, the mold was placed on top of the re-sist and pressed into the resist by two parallel plates. Therewas a hole in the center of the top plate that was largeenough for the laser beams to pass through. An excimerlaser pulse with a spot size large enough (3 mm by 3 mm)to cover the mold area was used to melt the resist. A HeNelaser (632.8 nm wavelength, continuous wave, beam diam-eter 0.5 mm) was used to monitor the resist imprint by theRIMS process. The incident angle of the HeNe laser beamwas 60◦ relative to the surface of the silica mold. The anglebetween the incident plane and the grating lines was 45◦.This arrangement ensured that the diffracted beam was di-rected along a different path than the reflected beam. The−1st order diffraction signal was collected by a high speedSi p–i–n detector with a band pass filter mounted in frontof the detector. The signal was sent to an oscilloscope foranalysis. The specifications for the detector, filter and os-

Fig. 23 Calculated diffraction efficiency as a function of SiO2 moldtrench depth for a 950 nm pitch fused silica mold using GSOLVERsoftware. The protrusion width is 300 nm. Based on this plot, the moldin our experiment was etched to 150 nm deep so that a monotonouschange in the diffraction signal could be observed

cilloscope have been described earlier in Sect. 2 during theTRR measurement.

3.7.3 Measurement results

Figure 24 is a typical curve of the diffracted light intensityas a function of time obtained during the real-time moni-toring of the ultrafast imprint process [26]. Before imprint,the grating mold and resist were in contact at room temper-ature and the resist was rigid with almost no deformation.


Fig. 24 Typical curve ofnormalized diffraction intensityvs time during RIMS. Thiscurve was obtained whenimprinting a 210 nm thick NXR1023 thin film on Si substrate ata laser fluence of 380 mJ/cm2.Three zones are defined on thecurve that represent the stagesbefore, during, and afterimprinting. The schematics ofpolymer filling into the gratingtrenches in each of the threestages are drawn as insets. Bluearrows indicate the start and endpoints of imprint. The wholeimprint process takes about200 ns [26]

No polymer filled into the trenches at this stage and the dif-fraction intensity stayed constant (Fig. 24, zone I). As theresist was heated by an excimer laser pulse, it became softand started to flow into the grating trenches under the ex-ternal pressure, causing a monotonic reduction of the effec-tive trench depth and hence a reduced diffraction intensity(Fig. 24, zone II). After completion of the imprint process,the trenches were completely filled with resist and the dif-fraction intensity remained constant near zero, due to therefractive index match between the mold and resist (Fig. 24,zone III). The entire imprint process was completed in about200 ns.

Furthermore, a spike was observed in the diffracted lightcurve (between zone I and zone II in Fig. 24). Since thespike had a pulse width almost the same as the excimer laser(20 ns), and the height could be reduced greatly by improv-ing the UV filtering, it was determined that it came fromthe scattering of the excimer laser. The spike, which marksthe time of the excimer laser beam, is useful to establish thetime relationship of laser melt and imprint [26]. Since thediffracted light intensity drops immediately after the spike, itreconfirms our previous observation (that the resist melts inless than 20 ns after the excimer laser pulse), which was ob-tained by TRR measurement of resists using the same pho-todetector and oscilloscope as in this study [27].

Based on the measured imprint time, the average speedof the imprint is estimated to be 0.5 m/s. The imprint speedis related to several factors such as viscosity and surface ten-sion of the polymer melt, imprint pressure and mold geome-try. Faster imprint speeds can be achieved by using polymersthat intrinsically have low viscosity. Higher laser fluence canbe used to raise the imprint temperature and lower the vis-cosity of the polymer melt. Raising the imprint temperaturealso decreases the surface tension of the polymer melt [28],resulting in increased imprint speeds. In addition, accordingto recent experimental and numerical simulation results, a

better mold geometry and a thicker resist film may also behelpful for polymer flow in NIL [29, 30].

3.7.4 Imprint fidelity during RIMS for LAN

To confirm that the trenches in the mold have been fullyfilled with resist during RIMS for LAN, both the mold andimprinted resist were examined using SEM and AFM. Geo-metrical features such as flat tops and vertical sidewalls werefully transferred from the mold into the resist (Figs. 25a, b).More importantly, the height of the resist lines was the sameas the trench depth in the mold (Figs. 25c, d), which suggeststhat the mold was completely pressed into the resist within200 ns. Further SEM characterization (Figs. 25e, f) showedthat the line width of the protruding part in the mold wasthe same as the trench width in the resist (both are 300 nm).This demonstrates the high fidelity of the ultrafast imprintprocess.

3.7.5 Picosecond capability of RIMS

In addition to measurement of imprint time for LAN, theexperiments here demonstrated the capability of RIMS formonitoring nanosecond imprint processes. However, the ul-timate time resolution of RIMS could be even higher. Thiscan be achieved using an ultrafast photodetector togetherwith an ultrafast oscilloscope. For example, a commerciallyavailable high-speed photodetector can offer as high as a60 GHz bandwidth (PX-D7, Newport Corp., Impulse re-sponse 7 ps, wavelength range 400–900 nm) [31]. And, adigital storage oscilloscope with bandwidth up to 15 GHz(TDS6154C, Tektronix Inc., 28 ps rise time, 40 GS/s samplerate) is also on the market [32]. Using this or other advancedequipments may extend the capacity of RIMS to monitor ul-trafast processes to tens of picoseconds.

In summary, we have proposed and demonstrated thelaser assisted nanoimprint lithography in which 100 nm


Fig. 25 AFM and SEM imagesof the mold used [(a), (c), and(e)] and the imprinted resist[(b), (d), and (f)]. The flat topsand bottoms and verticalsidewalls in the mold were fullytransferred to the resist duringimprint. Cross-sectionalanalyses of the mold (c) and theimprinted resist (d) showed thatmold grating height was thesame as resist trench depth (bothat 150 nm), indicating a fullpattern transfer. Linewidth ofthe raised area in the mold (e)was also the same as the resisttrench width (f) (300 nm),suggesting high processfidelity [26]

wide (200 nm pitch) grating lines have been fabricated uponexposure to a single laser pulse. Since the pulse duration isvery short, the heating of the substrate and mold is negli-gible. Numerical simulations have confirmed that the tem-perature increase in the substrate/mold and the deformationof both are minimal. This could be helpful in reducing themisalignment due to thermal mismatch of the mold and sub-strate. The imprint time of LAN has been measured to beabout 200 ns using real-time imprint monitoring by scatter-ing of light (RIMS). The resist for LAN can be chemicallytailored, resulting in even less heating of the substrate. Thistechnique could be used in direct patterning of electronicand optical devices.

4 Self perfection by liquefaction (SPEL)

4.1 Introduction

Fabrication defects are unavoidable in most nanofabricationand ultimately determine the fabrication limit. Although alot of research has been carried out to correct the defects inthe process, subsequent fabrications add new defects. Previ-ously, isothermal heating to reflow and smooth polymericresist lines has been used to reduce line edge roughness

(LER). However, the resist is only an intermediate materialfor patterning that will be removed after a pattern transfer.And even with a perfected resist profile, the pattern transferfrom the resist to a hard material will introduce new fabrica-tion defects. For example, using isothermal heating, NPR-69lines on a Si substrate were heated at 80◦C for 5 min, whichresulted in smooth polymeric lines (Fig. 26a). However, af-ter RIE (with 10 sccm SF6, 40 sccm CHF3, 10 sccm Ar at15 mTorr, power density 425 mW/cm2, 2 min), the resultingSi lines have rough edges (Fig. 26b). This line edge rough-ness is introduced by the RIE process, which suggests thata smoothing process as the final stage in nanofabrication isindispensable.

Recently, a drastically different approach that removesdefects in nanostructure geometry within nanoseconds wasproposed [33, 34]. The new approach, termed self-perfectionby liquefaction (SPEL), removes defects by selectively melt-ing nanostructures with defects for a short period of time(e.g., hundreds of nanoseconds) while guiding the flow ofthe molten materials with a set of boundary conditions.Each of the molten structures reshapes itself into a desir-able geometry and then re-solidifies, and maintains the newshape. Using SPEL, we reduced the LER of Si and Cr grat-ing lines as much as 5.6 fold in less than 200 ns. For ex-ample, the 3σ LER of 70 nm wide Cr lines was reduced


Fig. 26 Even using smoothpolymeric lines (a) as masks forRIE, the resultant Si lines can bevery rough (b) due to the defectsintroduced by the RIE process.The 120 nm wide NPR-69 lineswere smoothed by isothermalheating at 80◦C for 5 min

from 8.4 nm to below 1.5 nm. This is well below the 3 nm“redzone limit” as shown in the International TechnologyRoadmap for Semiconductors (ITRS) [35]. We also dis-covered, for the first time, that if a plate is placed a dis-tance above the metal or semiconductor nanostructures dur-ing SPEL, the molten material self-rises to the top plate, andreshapes itself into new structures. These new structures notonly have smooth edges, vertical sidewalls, and flat tops,but also narrower width (>2.15-fold reduction) and greaterheight (>2.10-fold) than the original (hence a >4.50-foldaspect ratio improvement). The melting was achieved usinga single nanosecond excimer laser pulse.

In this section, the principle of SPEL is firstly introduced.Next, different forms of SPEL are demonstrated with exam-ples. The mechanisms of SPEL are discussed according todifferent models. Finally, a few applications of SPEL are in-troduced.

4.2 Principle of SPEL

The principle of SPEL is based on the fact that surface ten-sion in a liquid will smooth out the rough edges of a liq-uid line and change a non-circular shape of a liquid dot intoa circle. This is because a smooth edge or circular shaperepresents an energy minimum in thermal equilibrium [36].In SPEL, the defective nanostructures are selectively meltedfor a short period of time under a set of simple boundaryconditions. In the molten state, the nanostructures reshapethemselves into better geometries for energy minima. Whenthe structures re-solidify, the perfect shapes are maintainedand the set of boundary conditions can be removed. Themelting can be achieved using a single nanosecond excimerlaser pulse, which would melt only a thin surface layer ofa material in less than 1 ns [6] and keep other parts of thematerial at a low temperature and in the solid phase. This al-lows for modification of the surface layer while leaving thebulk unchanged.

According to the differences in the boundary conditions,SPEL can be categorized into three basic forms (Fig. 27).(1) Open space SPEL (O-SPEL): nanostructures are placedon a substrate without any additional boundary conditions.

During this process, the surface tension and the interactionbetween the molten material and the substrate cause the ma-terial to flow to minimize the surface energy. As a result,the rough edges of a line are smoothed, and a circular dot isformed from a non-circular one. However, in O-SPEL, de-pending on the melting time and energy, the sidewall and thetop surface can become rounded and the line width can bewidened. (2) Capped SPEL (C-SPEL): a flat plate is placedin contact with the top of the to-be-perfected structures. Thiscan be done by placing a single transparent place on top ofall nanostructures, or one individual plate on the top of eachstructure. Under the boundary conditions set by the sub-strate and the top plate(s), the molten material reflows intonew shapes with smooth and vertical sidewalls, flat tops andthe same height as originals. (3) Guided SPEL (G-SPEL): aplate is placed a gap above the to-be-perfected structures. Inthis case, the interaction between the molten structures andthe top plate can make the molten structure rise up to reachthe top plate. Consequently, a greater height and a narrowerline width in addition to smooth edges, vertical sidewallsand flat tops are achieved.

Based on the principle, it can be concluded that SPEL hasthe following novelties.

1. It removes defects post fabrication. Although smoothingof etching barriers (e.g., resist lines) has been known formany years [37], it is only intermediate material. Evenwith a perfected resist profile, the transfer from the resistto a hard material will introduce new fabrication defects,as shown in Fig. 26. From this point of view, SPEL is ca-pable of removing intrinsic fabrication defects that haveno manufacturable solutions yet.

2. It is an ultrafast, low-temperature process. Using of anultrafast laser pulse makes it possible that only the sur-face layer is thermally affected. The bulk of the mater-ial will stay at low temperature without being damaged.Isothermal heating has been demonstrated for polymerreflows and smoothing [37], but such an approach is notencouraged for materials such as semiconductors, met-als, or hard dielectrics because these materials have highmelting temperatures. This high temperature will destroy


Fig. 27 Working principle of three forms of self-perfection by liquefaction (SPEL). (a) Open SPEL, (b) Capped SPEL, and (c) Guided SPEL for(I) lines and (II) squares or dots. See text for more detail [33]

other parts of the devices as well as the substrates. Un-like other smoothing techniques, SPEL avoids a globalisothermal high temperature process.

3. It is a selective process. Selectivity of SPEL has a two-fold meaning. First, it is materials selective. The heat-ing in SPEL is performed by a pulsed laser, which isabsorbed by only a thin surface layer of the materialswith a bandgap smaller than the photon energy [38], sothat other materials (such as SiO2) on the surface or un-derneath will be kept at much lower temperature duringSPEL. Second, this process is area selective. For exam-ple, in our study, the laser spot is about 3 mm by 3 mmand can be adjusted within a certain range. A selectivemelting of certain areas of a wafer can also be achievedwith a mask (materials either reflecting or absorbing laserenergy) that is placed either directly on the wafer or a dis-tance away. This allows us to selectively expose the areawith defects while leaving other components intact.

4. The perfecting is achieved by itself under simple bound-ary conditions. The profile of the nanostructures can beenhanced as well under certain conditions. Previous workusing laser heating to cause material reflow—such asplanarization of phosphosilicate glass (PSG) [39] andsmoothing of SiO2 optical disks with a continuous wave(cw) CO2 laser [40], and metal planarization [41] and Silarge grain growth promotion using pulsed lasers [42]—did not use any guiding in laser melting. As a result, thesmoothed structures usually have sloped and curved side-walls, and a rounded line top, which are undesirable and

detrimental to nanofabrication and nanodevices. The useof guiding conditions to achieve a desired self-perfectionnot only opens new avenues to overcome these problems,but also presents interesting physics at nanoscale. Fur-thermore, different boundary conditions (such as combi-nation of different surface properties) may lead to evenbetter self-perfection.

4.3 Open space SPEL (O-SPEL)

4.3.1 Experimental

In order to demonstrate the principle of SPEL in open space,several types of samples were prepared: (1) metal pads of200 to 225 nm pitch with rough edges; (2) 200 nm pitchmetal lines with rough edges; (3) Si lines on SOI wafers oron Si wafers, with a zigzag profile or rough sidewalls. Tomake the metal nanostructures, NIL was first performed onthermoplastic resists, followed by oxygen RIE to remove theresidual resist layer. Metallization was then carried out usingan electron gun evaporator, followed by a liftoff process. Theroughness in the resultant nanostructures was mainly fromthe RIE and liftoff processes.

To fabricate the Si nanostructures, NIL was first used forpatterning the resists. Two sets of molds were used. Onewas a defective mold that has severe line edge roughnessfrom failed interference lithography. Lines on the mold wereabout 70 nm wide (200 nm pitch) and had a zigzag shape.The other was a mold having 950 nm pitch and 250 nm


wide lines with much better edge smoothness. After NILand O2 RIE, SiOx was deposited on the samples using evap-oration. After liftoff, chlorine based RIE with 10 sccm Cl2,20 sccm Ar, 100 mW/cm2, 8 mTorr was used to intention-ally create sidewall roughness for demonstration purposes.Then samples were dipped in diluted HF briefly to removethe SiOx mask before being cleaned in RCA #1 solution. Inboth cases, such a “bad” mold or RIE recipe offers an ex-cellent test of the effectiveness of SPEL in perfecting severeedge (sidewall) roughness. During SPEL experiments, sam-ples were exposed to single laser pulses with proper laserfluence. The characterization was done using an SEM andan AFM.

4.3.2 Results

Si lines Before O-SPEL, the original 70 nm wide, 100 nmhigh Si grating lines on a SOI wafer had a zigzag shape(Fig. 28). Similarly, zigzag shaped 70 nm wide, 70 nm highSi grating lines were made on a (100) bare Si wafer aswell (Fig. 30). These Si lines had severe edge roughness,which came from the “bad” imprint mold and RIE in pat-tern transfer, as discussed earlier. However, after O-SPEL

with a single excimer pulse with a laser fluence of 545 and440 mJ/cm2, respectively, the zigzag edges became smooth(Figs. 28b, 30b). Using digitized SEM images and fractalanalysis [43], we found the 3σ line edge roughness (LER)of the grating lines, which is defined as 3 times of the rootmean square (RMS) of variation in fractal analysis, is re-duced from an original 19.5 nm to 3.6 nm for Fig. 28a andfrom 14.4 nm to 3.3 nm for Fig. 30a, respectively, represent-ing a LER improvement (simply defined as 3σbefore/3σafter)of 542% and 436% (Table 4). As shown in Table 4, the SEMimage resolution is 1.7 nm/pixel for Fig. 28, hence the ac-tual LER after O-SPEL can be less than calculated. A closeexamination of the LER in the SEM image (Fig. 29) showsthat the maximum variation has been reduced from 16.1 nmto 3.5 nm, and that there are high frequency digitizing noisesfrom the SEM imaging process superimposed on the actualLER.

To demonstrate the ability of O-SPEL to remove the side-wall roughness, 250 nm wide Si grating lines in both Si andSOI wafers were fabricated (Figs. 31a, 32a).

After O-SPEL with a single excimer pulse with a laserfluence of 880 and 490 mJ/cm2, respectively, the sidewallroughness in both cases was reduced (Figs. 31b, 32b). The

Fig. 28 70 nm wide and100 nm high zigzag Si lines onSOI (a) were smoothed duringO-SPEL using a single pulse of545 mJ/cm2 (b) [33]

Fig. 29 Digitized edge of zigzag SOI lines before (a) and after (b) O-SPEL. The edge profiles are from the high resolution SEM images of the70 nm wide SOI lines shown in Figs. 28a and b [33]


Fig. 30 70 nm wide, 70 nmhigh zigzag Si lines on Si(a) were smoothed duringO-SPEL with a single pulse of440 mJ/cm2 (b) [33]

Fig. 31 250 nm wide, 150 nmhigh Si lines with sidewallroughness (a) smoothed out byO-SPEL using a single laserpulse of 880 mJ/cm2 (b) [33]

Fig. 32 250 nm wide, 190 nmhigh Si lines on 380 nm thickburied oxide (SIMOX SOI) withsidewall roughness (a)smoothed out by O-SPEL with asingle pulse of 490 mJ/cm2 (b)

Fig. 33 (a) 70 nm wide, 40 nmhigh rough edged Cr lines on22 nm thermal oxide capped Siwafer. (b) After O-SPEL with asingle laser pulse of320 mJ/cm2, the lines becomeultra-smooth [33]

difference in the laser fluence needed for Si was higher thanthe SOI lines even though the lines in SOI were higher thanthat on Si. This fact can be explained by the effect of SiO2

as a thermal barrier layer as discussed earlier in Sect. 2. Itshould be noted that in these cases, the cross section of thelines changed from rectangular to semi-circular.

The Si root sizes (line width on a top view image) ofthe Si lines after O-SPEL almost stayed at the same as be-fore O-SPEL. However, if higher laser fluence was used, theroot size of the Si lines could be larger than the original.

For example, with a laser fluence of 1200 mJ/cm2 for the Silines shown in Fig. 31a, the resulting round cross-sectionedSi lines had a root size of about 520 nm, much larger thanthe original 280 nm. This is because high fluence results inlonger molten time and hence longer flow distance, so largerroots are developed.

Metal lines For metal lines, two types of nanostructureswere fabricated, namely, 70 nm wide Cr grating lines(Fig. 33a) and 75 nm wide Au grating lines (Fig. 35a). Both


Fig. 34 Digitized edge profiles of the Cr lines before (a) and after(b) O-SPEL. The edge profiles are from the high resolution top-downSEM images of the 70 nm Cr lines shown in Figs. 33a and b. Note

that O-SPEL reduces 3σ LER of 70 nm Cr lines from 8.4 nm to below1.5 nm, the same order of the noise from SEM imaging [33]

Fig. 35 (a) 75 nm wide, 40 nmhigh rough edged Au lines on22 nm thermal oxidized cappedSi wafer. (b) After O-SPEL witha single laser pulse of150 mJ/cm2, the lines becomeultra-smooth

Fig. 36 (a) 10 nm thick, 75 nmside Cr pads with rough edgeson a 300 nm SiO2 capped Siwafer. (b) After O-SPEL with asingle laser pulse of 1.7 J/cm2,the original rough-edged padswere turned into smooth rounddots of 50 nm in diameter and22 nm in height [33]

are 40 nm high and on 22 nm thick thermal SiO2 capped Siwafers. After O-SPEL with a single pulse of ∼320 mJ/cm2

and ∼150 mJ/cm2 for Cr and Au, respectively, the roughlines were smoothed (Fig. 33, Fig. 35). The edge profilesbefore and after O-SPEL of Cr are shown in Figs. 34a andb, respectively.

It is worthwhile to note that for the Cr lines, the 3σ LERbefore and after OSPEL was reduced from 8.4 nm to 1.5 nm

(see Table 4). This is well below the 3 nm “red-zone limit”and even smaller than the targeted LER for the year 2013.

Metal dots To demonstrate the capability of O-SPEL inturning the rough edged metal dots into nearly perfect roundones, Cr squares of 10 nm thick and 75 nm×75 nm area withrough edges (Figs. 36a, 37a) were fabricated on a 300 nmthick thermal oxide capped Si wafer. A single laser pulse of


Fig. 37 AFM images of Cr dotsbefore (a) and after (b) O-SPEL.These AFM images correspondto SEM images shown inFigs. 36a and b, respectively

Fig. 38 (a) Cr rectangles of145 nm × 110 nm with roughedges. These Cr rectangles are22 nm thick, 225 nm in pitchand were made on a 100 nmthick thermal SiO2 capped Siwafer. (b) After O-SPEL with asingle pulse of 300 mJ/cm2, therectangles became round shapedwith a diameter of 90 nm andheight of 47 nm. Laser meltingdid not change the center tocenter spacing

Fig. 39 FFT patterns for Crrectangles before (a) and after(b) O-SPEL. Both patternsshowed that the periodicitystayed the same (225 nm) beforeand after O-SPEL

1.7 J/cm2 turned the squares into nearly perfectly round dotswith a 50 nm diameter and 22 nm height (Figs. 36b, 37b).

O-SPEL changed the shape of rough-edged metal dotsinto round ones and reduced their sizes. It also kept theperiodicity of the dots before and after perfecting. Thesize of the nanostructures became more uniform as well.For example, Cr nanodots of rectangular shape and sizeapproximately 145 nm × 110 nm were made on 100 nmthick thermal oxide on a Si wafer (Fig. 38a). After O-SPEL with a single laser pulse of 300 mJ/cm2, the rec-tangles turned into circles with a diameter of 90 nm(Fig. 38b).

The size and pitch distributions before and after O-SPELwere analyzed in Image Pro Plus [44]. A statistical sizeanalysis was carried out for 150 dots before and after O-

SPEL. The sides of the rectangles before O-SPEL were143.7 ± 6.1 nm and 110.5 ± 3.8 nm, with a length/widthratio of 1.29 ± 0.06. After O-SPEL, the circles had a di-ameter of 90.3 ± 1.8 nm, and the “length/width” ratio was1.08 ± 0.04, suggesting that the circles were in nicely roundshape. From these data, it can be concluded that both the sizeand size distribution became smaller after O-SPEL, with astandard deviation improved by a factor of ∼3.4. Figure 39shows the fast Fourier transform (FFT) of the Cr rectanglearrays. It was found that the pitch of the Cr rectangles beforeand after O-SPEL did not change; both were still 225 nm,suggesting that laser melting did not change the center tocenter spacing.

In addition to removing the defects, another important ad-vantage of open-SPEL is that reduces the original size. The


final dot diameter after SPEL can be estimated from the con-stant material volume V and contact angle θ (Figs. 40a, b).If we assume the dot is a segment of a perfect sphere, itsvolume is

V = πh

6

(3r2 + h2) = πR3

3

(2 − 3 cos θ + cos3 θ

), (3)

where h is the height of the cap, r the diameter of the cap, R

the diameter of the hosting sphere, and θ the contact angle.Because the h of partial sphere is often many times larger

than the initial pattern thickness, the diameter of the finalpartial sphere formed by melting the pattern will be smallerthan the lateral dimension of the initial pattern. The effective

Fig. 40 Schematic of effective metal dot diameter when the contactangle between the dot and the substrate is smaller than 90◦ (a); andlarger than or equal to 90◦ (b). (c) Calculated dependence of metal dotdiameter on the material volume for the materials with contact angles(to the substrate) of 60◦ and 140◦, respectively [63]

diameter, d , of the partial sphere is given by

d =⎧⎨

⎩

2r = 2[ 3V

π(2−3 cos θ+cos3 θ)] 1

3 sin θ, for θ < 90◦,

2R = 2[ 3V

π(2−3 cos θ+cos3 θ)] 1

3 , for θ ≥ 90◦.(4)

The dependence of the effective metal dot diameter onthe volume of the metal for materials with different contactangles (we use 60◦ and 140◦ in the calculation) was plottedin Fig. 40c. Clearly, with less material and/or a larger contactangle, the dot size becomes smaller.

Following the “scaling rule”, round Au dots of 10 nmdiameter were fabricated on a Si wafer capped with 200 nmthick thermal oxide (Fig. 41). The structure before O-SPELwas made by NIL, RIE and an e-beam evaporator (4 nmAu was deposited on 1 nm thick Cr on the substrate), with200 nm pitch and 46 nm width. After single pulse O-SPELwith a laser fluence of 640 mJ/cm2, the Au pads were turnedinto nanodots of 10 nm in diameter (Fig. 41).

4.3.3 Discussion on the O-SPEL speed

O-SPEL is an ultrafast process which finishes within hun-dreds of nanoseconds for Si and metals due the high surfacetension and low viscosity of the molten materials. The self-perfection time in free space can be estimated using level-ing theory. In this theory, the smoothing time is proportionalto the surface tension of the liquid, while inversely propor-tional to the viscosity. For a liquid layer with a thickness ofh and a surface variation period of L, the time (T ) requiredfor the amplitude of the surface wave to decay to 1/e of itsinitial amplitude is given by [45]

T = 3ηL4

16π4γ h3, (5)

where η and γ are the viscosity and the surface tension ofthe molten material, respectively.

The viscosity and surface tension data for different ma-terials at their molten states are listed in Table 1. From thistable, it can be noticed that the molten Cr and Si has a vis-cosity about 1,000 to 1,000,000 times lower and a surface

Fig. 41 10 nm Au dotsfabricated by single pulseO-SPEL of 640 mJ/cm2. Theoriginal structures were 46 nmin size fabricated by depositing4 nm thick Au on 1 nm Cr onthe substrate patterned by NIL(a) and the final round dots’ sizewas 10 nm in diameter (b)


Table 1 Viscosity and surface tension for molten materials

Viscosity Surface tension References

(mPa s, cp) (mN m−1)

Silicon 0.58 720 [46, 47]

Chromium 5.70 1,642 [48, 49]

Gold 5.13 1,145 [11, 49]

PMMA (3K) 105–106 31 [50]

Water 1 73 [11]

Fig. 42 Smoothing time as a function of viscosity for semiconductors,metals and polymers. A smaller viscosity at molten state favors a fastersmoothing process. The film thickness is 50 nm and the surface vari-ation period is 1 µm for the calculations. Adapted from reference [51]with extended data range

tension of 20–100 times higher than a molten polymer. Thisleads to a smoothing time 3–8 orders of magnitude shorter.For example, for smoothing L = 1 µm in a 50 nm thick film,it would take 52 ns, 16 ns and 51 seconds for molten Cr, Si,and PMMA, respectively.

To better understand the difference in smoothing timefor different materials, plots of smoothing time as functionsof viscosity and surface tension are shown in Fig. 42 for a50 nm thick film with a surface variation period of 1 µm [51].The approximate ranges of smoothing time for metals, Siand polymers are marked on the plots using ellipses. It isclear that the smoothing time needed for metals and semi-conductors is much shorter than that for polymers. A shortertime in SPEL helps the self-perfection of desired structureswithout degrading other structures adjacent and underneath.

Limitations to O-SPEL do exist. For example, the crosssections of the nano-structures are altered after SPEL. Theline width after O-SPEL may change depending on the laserfluence and melting time. Under certain conditions, the in-stability in liquid may play some adverse roles and should be

controlled. However, O-SPEL still suggests a new approachfor removing nanofabrication defects.

4.4 Capped SPEL (C-SPEL)

As discussed earlier, one limitation of O-SPEL is thatthe repaired structures have sloped and curved sidewallsand rounded tops. In some cases—such as coupling ofwaveguide to fiber optics—this is actually beneficial forthe applications. However, a rectangular cross section withflat tops and vertical sidewalls is desired for pattern trans-fer fidelity (e.g., during RIE). Different from O-SPEL, thecapped SPEL (C-SPEL) can keep the top surface of a to-be-perfected structure flat and the sidewall vertical.

There are two types of C-SPEL (Fig. 27). One uses a FSplate on top of the nanostructures to be repaired (which isin contact with the structures to hold them during melting)and the other uses a Cr/SiO2 double layer on top of eachstructure to block the laser light from reaching the top ofnanostructures.

4.4.1 Experiments

To test the principle of C-SPEL, both Cr and Si grating lineswith rough edges were fabricated using NIL, RIE, metal-lization and liftoff. The samples and their preparation pro-cedures are listed in Table 2. For schemes 1 and 2, a fusedsilica plate was placed in contact with the defective Si or Crnanostructures with some pressure. For scheme 3, Cr/SiO2

double layer was used. This double layer was removed byCR-7, diluted HF (1:10) sequentially after C-SPEL.

4.4.2 Results

Figure 43 shows the SEM images before and after C-SPEL of the 280 nm wide Si lines on a SOI substrateguided by a single fused silica plate. The laser fluenceused is 480 mJ/cm2. Comparison before and after C-SPELshows that the linewidth and height did not change, whilethe roughness on the sidewall had been removed. Frac-tional LER analysis showed that the 3σ LER had been re-duced from 9.6 nm to 4.2 nm. Similarly, for Cr lines, thelinewidth and height did not change before and after CSPEL(406 mJ/cm2, Fig. 44), with a 3σ LER improvement from17.7 nm to 7.5 nm.

For 200 nm wide Si lines guided by individual caps(Cr/SiO2)—one on each line—a single laser pulse of390 mJ/cm2 was used. After C-SPEL, the top of these struc-tures were flat and the sidewalls were vertical (Fig. 45). Thefractal analysis shows that the 3σ LER had been reducedfrom 11.1 nm to 5.4 nm.


Table 2 Samples and fabrication procedures for C-SPEL

Scheme Structures Fabrication procedures

1. Single plate for all Si lines Si lines: 280 nm wide; 200 nm high; on500 nm thick buried oxide (SIMOX SOI)

(1) Thermal imprint NPR-69; (2) O2 RIE residual layer;(3) Evaporate SiOx ; (4) Liftoff; (5) RIE Si (10 sccm Cl2,20 sccm Ar, 8 mTorr with 100 mW/cm2); (6) Diluted HFdip.

2. Single plate for all Cr lines Cr lines: 280 nm wide; 62 nm high; on220 nm thick thermal oxide

(1) Thermal imprint NPR-69; (2) O2 RIE residual layer;(3) Evaporate 62 nm Cr; (4) Liftoff.

3. Individual plate for each Si line Si lines: 200 nm wide; 140 nm high on Si (1) Grow 45 nm thick thermal oxide; (2) Thermal imprintNPR-69; (3) O2 RIE residual layer; (4) Evaporate 45 nmCr; (5) Liftoff; (6) RIE SiO2(33 sccm CF4, 7 sccm H2 at50 mTorr and 425 mW/cm2 for 1 min); (7) RIE Si(40 sccm CHF3,10 sccm Ar, 10 sccm SF6 at 15 mTorrand 425 mW/cm2 for 2 min and 40 secs).

Fig. 43 Using a fused silica plate in contact with the Si lines of280 nm width, 200 nm height, 950 nm pitch on 500 nm thick SiO2(SIMOX SOI), the rough lines in (a) were smoothed (b) without chang-

ing the line height, flat top and vertical sidewalls. A single laser pulseof 480 mJ/cm2 was used in C-SPEL. The 3σ LER was reduced from9.6 nm to 4.2 nm [33]

Fig. 44 C-SPEL of Cr linesusing a FS plate. (a) Initialstructure is 280 nm wide, 62 nmhigh, with a 3σ LER of17.7 nm. (b) After C-SPEL witha single pulse of 406 mJ/cm2,the width and height of the Crlines was preserved but the 3σ

LER was reduced to 7.5 nm [33]

Fig. 45 SEM images of 200 nmwide, 140 nm high Si linesbefore (a) and after (b) C-SPELusing 45 nm Cr/45 nm SiO2double layer as the cap and asingle laser pulse of390 mJ/cm2 [33]


4.4.3 Discussion

The C-SPEL is made possible by the energy minimiza-tion principle. Consider a line with width w, height h andlength l. The surface tension at the molten state is γ . Thecross sections corresponding to the original, O-SPEL, andC-SPEL are illustrated in Fig. 46. In order to simplify theproblem, let’s just consider a special case where the crosssection after O-SPEL is a half-circle of radius r . This meansthe contact angle of the molten materials with the substrateis 90◦, which is very close to that of molten Si on SiO2

(87◦ [52]).The surface energy for the above three cases can be ex-

pressed as

Ea = γ · (w + 2h)l, (6)

Eb = γ · πrl, (7)

Ec = γ · 2hl. (8)

Due to volume conservation, we have

whl = 1

2πr2l, (9)

which gives the relationship between radius r and w, h as

r =√

2wh

π. (10)

The equation for surface energy after O-SPEL (case (b) inFig. 46) reduces to

Eb = γ · πrl = γ · π√

2wh

πl = γ · √2πwhl.

Since (w + 2h)2 = w2 + 4h2 + 4wh ≥ 4wh + 4wh =8wh > 2πwh, the surface energy after O-SPEL (case (b)) issmaller than that before O-SPEL (case (a)), which is reason-able for all the O-SPEL experiments.

In order to favor C-SPEL over O-SPEL, γ · 2hl < γ ·πrl

should hold. This leads to the following geometrical condi-tion:

w >2

πh (= 0.64h). (11)

According to (11), as long as the width/height ratio of theoriginal line is larger than 0.64, C-SPEL will be favorable

to O-SPEL considering the energy minimum principle. Forthe nanostructures tested in our experiments, this criterion issatisfied for all cases (Figs. 43 to 45).

4.5 Guided SPEL (G-SPEL)

In guided-SPEL, local spacers are used to keep a plate fixedabove the nanostructures (Fig. 27), the molten nanostruc-tures rise against the surface tension until they reach theplate. This leads to smooth edges, vertical sidewalls and flattops, and also to narrower linewidths and greater line heights(and hence higher aspect ratios) than the original structures.The entire melting, rising-up and reshaping took less than200 ns for silicon and chromium materials.

4.5.1 Experimental details

Samples and fabrication Four types of nanostructureswere fabricated for G-SPEL experiments: (1) Si dots; (2)Cr dots; (3) Si lines; and (4) Cr lines. Their geometries arelisted in Table 3. The Si nanostructures were fabricated onan epitaxy SOI wafer which had a 50 nm thick Si devicelayer and 200 nm thick buried SiO2 by thermal NIL, O2

RIE, SiOx etching mask deposition, liftoff and Cl-based RIEwith a final quick HF dip to remove the residual SiOx mask.The Cr nanostructures were deposited on NIL patterned re-sists by evaporation followed by liftoff. All the as-fabricatednanostructures had rough edges which were introduced bythe fabrication process.

Spacer fabrication To control the gap between the topplate and the substrate, silicon oxide (SiOx ) spacers weredeposited onto a fused silica (FS) wafer through a shadowmask using the electron beam evaporator. The SiOx islandswere about 500 µm by 500 µm in size and had differentheights. The distance between each spacer was 700 µm inone direction and 1200 µm in the other.

Another type of spacer was fabricated by etching into theFS wafer using evaporated Cr as an etching mask, followedby CR-7 etching and RCA #1 cleaning. Although this fab-rication process took more steps than the first approach, thespacers material is denser than the evaporated SiOx so theycould be cleaned under RCA #1 without degradation aftereach G-SPEL experiment and used repeatedly.

Fig. 46 Schematic cross section for original line (a) and those during O-SPEL (b) and C-SPEL (c). The surface energy is calculated using thefigures here. For simplification, the cross section after O-SPEL is considered as a semi-circle


Table 3 Samples and results for Si and Cr G-SPEL

Samples Geometries Gap Dose G-SPEL O-SPEL

(mJ/cm2)

Si dots 90 nm × 100 nm 22 nm 400 78 nm diam. 85 nm diam.

50 nm height 73 nm height 62 nm height

Cr dots 140 nm × 110 nm 40 nm 420 70 nm diam. 80 nm diam.

20 nm height 62 nm high 48 nm height

Si lines 285 nm width 40 nm 595 175 nm width NA

50 nm height 90 nm height

Cr lines 280 nm width 70 nm 406 130 nm width NA

62 nm height 130 nm height

Fig. 47 SEM images of Si andCr dots before and after O-SPELand G-SPEL. As-fabricated Sisquares (90 nm × 100 nm and50 nm tall) (a) becamesemi-spheres (85 nm diameterand 62 nm tall) if O-SPEL wasused (b), but became cylindersif G-SPEL (22 nm gap) wasused (c). The cylinders had a78 nm diameter and 73 nmheight (150% of originalheight), in addition to flat topsand vertical sidewalls. Similarly,the fabricated Cr squares(140 nm × 110 nm and 20 nmtall) (d) became semi-spheres(80 nm diameter and 48 nm tall)after O-SPEL (e), but becamethe cylinders that had a 70 nmdiameter (50% of original inlateral size) and 62 nm height(310% of original height) inaddition to flat tops and verticalsidewalls (f) after G-SPEL(40 nm gap) [33]

G-SPEL experiments The substrate and the top plate weresandwiched between two metal plates which applied pres-sure via a set of screws that kept a conformal contact. Thegap between the top plate and the surface of the nanostruc-tures was kept constant by the spacers. In all cases, a singlelaser pulse was sufficient to achieve G-SPEL. There was alaser fluence window for G-SPEL; for example, fluence of370 to 450 mJ/cm2 worked best for G-SPEL of Cr dots. Thenanostructures before and after the laser pulse were charac-terized using both SEM and AFM.

4.5.2 Results

The results for the G-SPEL of Si and Cr, together with theexperimental conditions are summarized in Table 3.

Cr and Si dots An O-SPEL with a single laser pulse ex-posure made the original dots round (Figs. 47b, e), whilea G-SPEL with the same laser fluence resulted in cylinders(Figs. 47c, f). For Si, with a single pulse of 400 mJ/cm2, theoriginal pads of 90 nm by 100 nm and 50 nm tall (Fig. 47a)


became (1) smooth semi-spheres of 85 nm diameter and62 nm tall if an O-SPEL was used (Fig. 47b); and (2) cylin-ders of smooth and vertical sidewall, flat top, 78 nm diam-eter, and 73 nm tall (150% of original height) if a G-SPELwas used (Fig. 47c). Similarly, a single pulse of 420 mJ/cm2

made the 140 nm by 110 nm by 20 nm Cr pads (1:7 as-pect ratio) (Fig. 47d) into semi-spheres of 80 nm diameterand 48 nm tall (Fig. 47e) in O-SPEL; but a G-SPEL withthe same laser fluence resulted in cylinders of 70 nm diame-ter, 62 nm tall (310% of original height, and an aspect ratioof almost 1), in addition to flat tops and vertical sidewalls(Fig. 47f). AFM characterization of the original Cr dots andthose after O-SPEL and G-SPEL, is shown in Fig. 48, whichcorresponds to the right column in Fig. 47.

Cr and Si lines Figure 49 shows SEM of Cr lines and Silines before and after a single laser pulse G-SPEL (guidedby a single quartz plate with 70 and 40 nm gaps above theoriginal structures, respectively). Figures 49a and b showthat the original 280 nm wide Cr lines of 62 nm thick onSiO2 surface became 130 nm wide (46% of original width)and 130 nm tall (210% of original height) after G-SPEL,and hence a 452% (4.5 times) increase in aspect ratio wasachieved. The 3σ LER was reduced from 17.4 nm to 5.4 nm(322% improvement). For 285 nm wide silicon lines of50 nm height, G-SPEL made them 170 nm wide and 90 nmhigh. A single laser pulse of 406 mJ/cm2 and 595 mJ/cm2

was used for Cr and Si lines, respectively. All G-SPEL oc-curred in a time frame of hundreds nanoseconds.

4.5.3 Discussion

Mechanism for self rising during G-SPEL Previously, aself-rise-up of material under a plate placed a gap awaywas observed only in continuous polymer films [53, 54].A model based on electrostatic interaction was developedwhich explained the experimental date well [53–55]. Herewe are trying to fit that model to our observation with metaland Si in order to find out if the phenomena is governed bythe same physics.

According to reference [55], the time for the material(polymer in that case) to rise up can be calculated using thefollowing formula:

t = ηγ d3

ε20(�Φ)4

, (12)

where t is the time for the material to rise up, d is the initialgap between the material and the top plate (hence electricfield ∼�Φ/d), η is viscosity, γ is surface tension, and ε0 isvacuum permittivity (ε0 = 8.85 × 10−12 C2/Nm2).

If the rising time is known, one can calculate the poten-tial difference between the molten material and the plate,�Φ , using the same formula. Take Cr and Si for instance,

for t ∼ 200 ns (observed in G-SPELs), d ∼ 7 × 10−8 m (Cr)and 4 × 10−8 m (Si), plugging in the viscosity and surfacetension date (in Table 1) shows that the required �Φ is 22and 7 V for G-SPELs in Cr and Si, respectively. These aremuch larger than possible in the experiments since the workfunction difference is only a few volts. This discrepancy sug-gests that the model for polymers may not be applicable toSi and metals.

The possible reasons for the discrepancy could be as fol-lows. First, Si and metals have much higher surface tensionthan molten polymers, hence require a much higher pullingforce, which might be out of range for that model. Second,the original model was developed for a continuous film, inwhich case an assumption that the film thickness is muchsmaller than its lateral dimension was made [55]. However,in the G-SPEL experiments, the thickness and width are onthe same order.

Although other mechanisms at nanoscale need to be ex-plored in order to understand the underlying physics, theshort self-perfection time makes SPEL very useful in manyapplications. It allows for the self-repair and self-perfectionof a desired structure without damage to degradation ofother structures on the substrates or the substrate itself.

Parameters of G-SPEL Several parameters are crucial to asuccessful G-SPEL. First of all, an appropriate gap size be-tween the guiding plate and the nanostructures is necessary.The gap size should be large enough to allow the molten ma-terials to rise after being melted, since when a wide line/dotbecomes narrower, the height always changes. However, ifthe gap size is too large, there would not be enough forceto pull the molten materials upwards. According to (12),the rising time is proportional to the third power of the dis-tance (gap size), which means even with enough electrosta-tic force, the rising time needed is very sensitive to the gapsize.

The geometries of the original nanostructures also mat-ter. Our experimental data showed that under certain circum-stances, the width and height of the lines played an impor-tant role. For example, with 280 nm wide, 62 nm thick Crlines, the resulting structures in G-SPEL were raised lines.However, metal lines with the same height but narrowerwidth (e.g., 90 nm, 50 nm, 40 nm) resulted in partial or totalfragmentation of the lines into nanoparticles.

The thickness of the lines plays a similar role. For ex-ample, reducing the film thickness from 62 nm to 35 nm(as shown in Fig. 49) for Cr results in partial fragmentation(Fig. 50) instead of G-SPEL.

4.6 Line edge roughness (LER) analysis

One of the advantages of SPEL is its ability to smooth outLER at the final stage. To quantify the improvement for the


Fig. 48 AFM images of Cr dots before and after SPELs. (a) and (d), original Cr pads of 20 nm high. (b) and (e), 48 nm high spherical dots afterO-SPEL of 420 mJ/cm2. (c) and (f), 62 nm high cylinders after G-SPEL of 420 mJ/cm2


Fig. 49 In G-SPEL with a70 nm gap and a fluence of406 mJ/cm2, the original Crlines of 280 nm wide and 62 nmheight on SiO2 (a) become130 nm wide (46% of original)and 130 nm tall (210% oforiginal height) due to materialrise-up during G-SPEL), a452% (4.5 times) increase inaspect ratio (b). In G-SPEL witha 40 nm gap and a fluence of595 mJ/cm2, the original Silines of 285 nm wide and 50 nmtall (c) become 175 nm wide(61% of original) and 90 nmheight (180% of original height)(d) [33]

Fig. 50 With the same laserfluence and same line width asin Fig. 49 (i.e., 280 nm) and a70 nm gap, 35 nm thick Cr lineson thermal oxide (a) becamedots which are connected witheach other rather than rising upinto narrower lines (b)

LER of Si and Cr lines, a fractal analysis based on high-resolution SEM images of the lines was carried out. The firststep was acquiring high-resolution SEM images of the grat-ing lines. The edges of the lines were then recognized byMatlab which uses Otsu’s threshold selection method [56].The algorithm in the LER analysis is based on the work ofConstantoudis et al. in reference [43]. The basic idea is asfollows. First, the position of each point along the edge isrecognized (the interval is 1 pixel in our analysis) and astraight line for the average position of the points is con-structed by the least square method. Next, the distance fromeach point on the edge to the linear fit line is measured, andthe standard deviation of the measured distance is consid-ered the LER (1σ).

However, the σ value is related to the vertical dimensionof roughness (in the direction perpendicular to the fit line)and gives no information about its spatial complexity. It issuggested that parameters for the description of LER shouldinclude: (1) the σ value, (2) the correlation length ξ , and(3) the roughness exponent α [43]. The correlation length

ξ defines a representative lateral dimension of a rough lineedge. If the distance between two edge points is within ξ ,the heights at these two points can be considered correlated.However, if the separation of two edge points is much largerthan ξ , then we can say that the heights at these two pointsare independent of one another.

The 3σ LER before and after SPEL was measured usinga MatLab program. The results, together with the improve-ment of ξ , are listed in Table 4 [33].

From Table 4, the analysis of SEM images of the Cr andSi grating lines before and after open-, capped-, and guided-SPEL show that SPEL can reduce 3σ LER by as much as afactor of 5.6, such as from 8.4 nm to far less than 1.5 nm for70 nm wide Cr lines. For each case, the ξ value increasedafter SPEL, which means the high frequency LER was im-proved significantly. It has to be pointed out that the parame-ters depend heavily on the imaging processing, the qualityof the SEM image, and other factors. As a result, the real σ

could be smaller.


Table 4 Measured line edge roughness (LER) before and after SPEL for Si and Cr lines [33]

Samples Image Res. 3σ LER Improvement ξ (nm) Fluence

(nm/pixel) (nm) (mJ/cm2)

Open-SPEL

Cr (Fig. 33) Original 0.4 8.4 5.6 281.7

After 0.4 1.5 5.6 387.6 320

Si (Fig. 28) Original 1.7 19.5 5.4 30

After 1.7 3.6 5.4 60 545

Si (Fig. 30) Original 0.4 14.4 4.4 19.6

After 0.4 3.3 4.4 216.6 440

Si (Fig. 31) Original 1.7 9.6 2.3 23.0

After 1.7 4.2 2.3 38.9 880

Capped-SPEL

Si (Fig. 43) Original 1.7 9.6 2.3 17.7

After 1.7 4.2 2.3 35.4 480

Si (Fig. 45) Original 1.7 11.1 2.1 37.2

After 1.7 5.4 2.1 125.6 390

Cr (Fig. 44) Original 3.4 17.7 2.4 27.6

After 3.4 7.5 2.4 117.3 406

Guided-SPEL

Cr (Figs. 49a, b) Original 3.4 17.4 3.2 24.1

After 3.4 5.4 3.2 79.3 406

4.7 Applications of SPEL

4.7.1 Sub-25-nm smooth cylindrical NIL molds

Fabrication of NIL molds with very small feature sizes istechnically challenging. One traditional method involves theuse of electron beam lithography (EBL) [8] to pattern a poly-mer thin film on a substrate. After developing the polymericresist, a thin layer of metal is deposited, followed by a liftoffprocess. The patterned metal layer serves as a hard maskduring a pattern transfer process using RIE. Although EBLhas been successfully demonstrated in making NIL moldswith lines of 10 nm width (35 nm pitch) [57] and dots of10 nm diameter (40 nm pitch) [58], it is a serial processthat is time-consuming and expensive. In addition, it is dif-ficult to control the critical dimension and structure profiledue to proximity effects and wet chemical processing (i.e.,development). Other methods such as interference lithogra-phy [59], proton beam writing [60], focused-ion-beam writ-ing [61], and X-ray lithography [62] have also been used inNIL mold fabrication, but they encounter problems similarto EBL. Furthermore, for all the aforementioned mold fab-rication methods, defects are intrinsically generated due tothe noise in lithography and the use of sequential processessuch as developing and etching.

In this section, SPEL was used to fabricate round andsmooth NIL pillar molds in SiO2 with sub-25-nm feature

size over large areas. The first step is to turn non-idealshaped metal nanopads into round dots using SPEL with asingle laser pulse. Next, RIE is carefully carried out into thesubstrate, achieving features as small as 25 nm over waferscale.

Principle The principle of our method is shown in Fig. 51[63]. First, metal pads are patterned on a substrate usingNIL, metallization and a liftoff process. The edges of thepads are usually rough (Fig. 51a), which results from thefabrication environment and fundamental fabrication prin-ciples. Next, we do SPEL for these metal pads using alaser pulse. The metal is selectively and rapidly heatedinto a molten state, and the surface tension turns the padsinto round droplets (Figs. 51b, c). The round shape ofthese droplets is preserved after the metal is resolidified(Fig. 51d). They are then used as hard masks for RIE(Fig. 51e). After stripping off the metal masks, round pillarswith smooth sidewalls remain on the substrate (Fig. 51f).

Experiment and results To test the principle, an NIL-patterned array of Cr rectangles was deposited on a fusedsilica substrate. These rough-edged pads were about 140 nmby 110 nm in size, and 20 nm thick (Fig. 52a). During thefollowing SPEL process, a single excimer laser pulse of420 mJ/cm2 was used to turn these pads into near-perfect


Fig. 51 Schematic of our process to make cylindrical pillar molds.(a) Metal nanostructures with geometrical defects. (b) The nanostruc-tures in (a) are melted upon exposure with a single laser pulse. (c) Themolten material reflows into round dots to minimize the surface en-ergy. (d) After re-solidification, the rough-edged pads turn into smoothround dots. (e) These round metal dots are used as a hard mask for RIE.(f) The metal mask is stripped off before the pillar mold is ready foruse [63]

circular dots with a final diameter of 80 nm (Fig. 52b). Fur-ther AFM measurements showed that the height of thesedots was about 48 nm. These round Cr dots were then usedas a hard mask during RIE of the fused silica, using a gasmixture of 10 sccm CHF3 and 1.5 sccm O2 under a pres-sure of <3 mTorr with a power density of 60 mW/cm2 ina Plasma Therm 2486 etcher. Round pillars with a height ofabout 200 nm were created after RIE and metal stripping us-ing Cr-7 etchant (Fig. 52d). With this etching recipe, thesepillars had smooth sidewalls. The periodicity of the metalnanodots was maintained as well (Fig. 52f). As a compari-son, the pillars which were etched using the unsmoothed Crpads as etching mask are also shown in Figs. 52c, e. Theyinherited the rough edges from the original Cr mask.

Successful imprints were achieved using the as-fabricatedround pillar molds (Figs. 53 and 54). In both UV and thermalimprint, the holes in resists are round in shape and uniformin diameter.

Large-area NIL mold have been made by step and repeatexposure. In order to cover the whole area, a stitching area(200 µm wide) between each laser spot was intentionallyleft (Fig. 55a). We have successfully made a mold with aneffective size of 1 inch by 1 inch in a FS wafer (Fig. 55b).It is worth noting that the Cr dots were of the same high

Fig. 52 (a) Cr squares of140 nm by 110 nm and 20 nmthickness on a fused silicasubstrate. (b) After SPEL, theirregular squares in (a) becamenearly perfect round dots of80 nm diameter and 48 nmheight. (c) Rough pillars etchedinto FS using Cr pads in (a) asan etching mask. (d) Smoothround pillars etched into FSusing round Cr dots in (b) asetching mask. (e) and (f) are topview images of pillars in(c) and (d), respectively, whichshow the etching fidelity duringRIE [63]


quality and same size in the areas that experienced only oneor multiple pulses (Figs. 55c, d), suggesting high processingtolerance of our process. The current area of the mold is notthe limit of our technology, as more step and repeat exposurewill result in a larger area.

Discussions SPEL has several advantages in makingsmaller feature size molds with fewer defects. First, asmaller feature size structure can be made from a structurethat is initially relatively large in size. This extends the ca-pability of lithographic tools into a smaller size regime. Sec-ond, the size can be tuned by tailoring the amount of materi-als deposited on each site before SPEL, as well as the inter-facial properties between the metal and the substrate. Third,

Fig. 53 Imprint in PUV-30 with the round FS pillar mold, resulting inround holes in the resist with uniform size

in addition to shrinking the size of the original nanostruc-tures, the size variation of the metal dots is also reduced atthe same time while keeping the original periodicity. Fourth,the shape of the metal dots (masks) can be almost perfectlyround due to the high surface tension of molten metal. Sincethe process is ultrafast, the ideal shape of a liquid dropletis preserved after re-solidification. This gives an ideal shapefor the etched pillars. Fifth, the mold area is relatively largeand the process has high throughput. Sixth, as an ultrafasttechnology, the laser pulse can heat the materials on all kindsof substrate selectively, so the resolution of the approach isnot limited by the substrate (for EBL, the electron chargingeffect is substrate-sensitive), hence one cannot make verysmall feature sizes on non-conductive substrates.

Fig. 54 25 nm diameter hole array in thermal plastic resist (NPR-69)imprinted using a Si pillar mold that was fabricated by SPEL on Crdots and RIE [63]

Fig. 55 (a) Schematic of stepand repeat exposure withoverlap (shadowed area) inSPEL; (b) Photograph of a FSmold which is about 1 inch insize; (c) and (d) are SEMimages from spot A and B in(a), which shows that the doubleexposure in the overlapped areaduring SPEL results in the samehigh quality round metal dots asetching masks [63]


As discussed in earlier sections, the shape and size of theCr dots can be changed; cylinders with a smaller diametercan be created if G-SPEL has been conducted for the orig-inal Cr pads. As a result, smaller FS pillar with better side-wall profile can be achieved after RIE.

In summary, SPEL has been used to make sub-25 nmround pillar molds with smooth sidewalls. The molds wereused for successful nanoimprinting. The size of the finalstructure can be tuned by tailoring the amount of mask ma-terial deposited at each site, and the interfacial propertiesbetween the metal and the substrate. This method works fordifferent substrates with high throughput and low cost, andit extends the capability of other lithographic methods intoa new regime.

4.7.2 Smoothing of Si waveguides

Integrating optical components on a chip with high packingdensity using existing Si fabrication technology is currentlythe topic of extensive research [64]. However, as device fea-ture sizes shrink, geometric defects such as sidewall rough-ness introduced in the fabrication process, will become moreprominent and will adversely affect light propagation prop-erties. Previous methods for reducing sidewall roughness,such as anisotropic wet etching [65, 66], thermal oxidation[67], or a combination of the two [65, 68], are either lim-ited to certain crystalline facets of semiconductor materi-als, or often involve harsh processing conditions and hightemperatures. These methods may also affect other compo-nents and/or materials on the same chip that are inadver-tently processed at the same time. In this section, the appli-cation of SPEL in smoothing Si waveguides and its implica-tion in the reduction of optical propagation loss is discussed.

Methods and experiments Micro-scale (4–10 µm) Si wave-guides were fabricated on a bonded SOI wafer (1.5 µm thickSi layer and 1 µm thick buried oxide) [69]. A thin layer ofthermal SiO2 was grown on the surface of the SOI wafer,followed by waveguide patterning using photo-lithography.A 20 nm thick Cr layer was evaporated using an electronbeam evaporator, followed by a liftoff process. The Cr pat-terns were used as etching masks for RIE the thin SiO2 layer.

After removal of Cr, the silicon device layer was etched us-ing chlorine base RIE with the thermal SiO2 as a mask.The wafers were then briefly dipped in a diluted HF solu-tion to remove the top oxide layer. Finally, the samples werecleaned in RCA #1 solution at 80◦C for 15 min before thesamples were ready for SPEL experiments.

Results Multiple pulses were used for the best smooth-ing results for the microscale waveguides. For example, therough sidewalls of a 4 µm wide waveguide (Fig. 56a) weresmoothed out after exposure to 20 pulses of 900 mJ/cm2

with a repetition rate of 1 Hz (Fig. 56b). LER analy-sis showed that roughness was reduced from 13 to 3 nm(Fig. 57).

It should be pointed out that the profile of the Si lineschanged from square to semi-circular. This might be helpfulin optical coupling between the Si waveguides and roundoptical fibers. However, in the case when the square profileneeds to be preserved, C-SPEL can serve this purpose.

Calculation of propagation loss The propagation loss is afunction of different sidewall roughness and roughness au-tocorrelation lengths. An SOI strip waveguide with a rec-tangular cross section 500 nm wide and 200 nm high wasconsidered. The propagation loss was then plotted as afunction of σ and Lc (autocorrelation length). For exam-ple, Fig. 58 shows the waveguide loss contours for a Siwaveguide 200 nm high and 500 nm wide. It clearly showsthat the waveguide transmission loss reduces with smallercorrelation length and smoother surface. Figure 59 plots theloss as a function of waveguide sidewall roughness for a200 nm high Si waveguide with a width of half and quarterwavelength used in telecommunication (i.e., 1.55 µm), as-suming an autocorrelation length of 5 nm. It indicates thatfor a 220 nm wide waveguide, a reduction of roughnessfrom 13 nm to 3 nm can result in a decrease of propaga-tion loss from 53 to 3 dB/cm. This means the transmittedpower could increase by 5 orders of magnitude for a 1 cmlong Si waveguide.

As in the previous section, wafer scale SPEL has beenachieved by a step and repeat exposure system for Siwaveguides. An exposure overlap of about 200 µm widewas used between each pulse spot. The smoothing results

Fig. 56 (a) A 4 µm wide Siwaveguide on SiO2 with roughsidewalls; (b) the waveguidewas smoothed after exposure to20 pulses with 900 mJ/cm2 at arepetition rate of 1 Hz [69]


Fig. 57 Line edge profile alongthe length direction of a 4 µmwide Si waveguide on SiO2before (a) and after (b) SPEL.The 1σ LER in (a) and (b) are13 and 3 nm, respectively [69]

Fig. 58 Waveguide loss contours for a Si waveguide of 200 nm highand 500 nm wide. The light wavelength used was 1.55 µm (in air) [69]

in the overlapping area had no difference with other areas,suggesting no rigorous alignment is required in SPEL for alarge area. It was found that a fluence window was about25% for smoothing the nanoscale Si gratings, indicating ahigh tolerance for energy fluctuation from pulse to pulse forSPEL. The features make the SPEL process simple to im-plement.

Fig. 59 Calculated waveguide transmission loss versus the roughnessfor Si waveguides with a width of half (220 nm) and quarter (110 nm)wavelength. In this calculation, the waveguides are 200 nm high withan autocorrelation length of 5 nm. The light wavelength used was1.55 µm (in air) [69]

In summary, we have used SPEL to smooth the side-wall roughness of Si waveguides. Our experimental resultsshowed that waveguide sidewall roughness was reducedfrom 13 to 3 nm using this technique. With this reduction inthe sidewall roughness, transmission loss in the waveguide


(200 nm by 500 nm) decreased from 53 to 3 dB/cm ac-cording to our calculation. As an ultrafast, highly selectivemethod, SPEL will be increasingly important when the sizeof silicon waveguides shrink for higher density nanopho-tonics. The advantages of SPEL will become more promi-nent as waveguide sizes shrink to allow for single opticalmode transmission and increased waveguide packing den-sity, since the effect of sidewall roughness in increasingtransmission loss becomes greater with smaller waveguidesizes. In addition, as chips become hybridized with optical,electronic, and other functions, chip fabrication technologymust keep pace. SPEL is an excellent candidate for highlyselective fabrication of ultra-smooth surfaces.

4.8 Summary

Three forms of SPEL and their applications have been ex-perimentally demonstrated in this section. Using O-SPEL,rough edges of nanoscale lines have been smoothed and non-ideal shaped nanopads have been turned into a nearly perfectshape. With O-SPEL, 3σ LER of rough edged 70 nm wideCr lines have been reduced from 8.4 nm to 1.5 nm, which iswell beyond the “red-zone limit” on the current technologyroadmap. With C-SPEL, the LER can be removed and thecross-section profile can be maintained. C-SPEL preservesthe same height, flat tops and vertical sidewalls of the nanos-tructures. Furthermore, during G-SPEL the molten nanos-tructures rise upon their own and reach the plate, reshapingthemselves into structures that have not only smooth edgesbut also vertical sidewalls, flat tops and narrower width forboth lines and dots. Applications of SPEL have been ex-emplified by making smooth round NIL pillar molds andsmooth Si waveguides that have propagation loss orders ofmagnitude lower.

The novelty and major contribution of this work areas follows. (a) Different from previous work in isother-mal smoothing of photoresists, LER and other defects innanostructures of high melting temperature materials (met-als and semiconductor) are repaired directly. (b) Differ-ent from other work such as lithographically induced self-assembly (LISA) [53] which works with a continuous film,the current work uses of a second surface (the plate above) toguide the reshaping of pre-patterned structures. (c) It is ob-served for the first time that liquid metals and semiconduc-tors flow upward against surface tension between the twoplates. Previously, similar behavior was only observed inpolymer systems. (d) The current work provides a method toenhance the nanostructures profile into smaller feature sizeand higher aspect ratio. This extends the lithography capa-bility to a new regime with only a set of simple boundaryconditions.

5 Nanoparticle arrays fabricated by pulsed lasermelting

5.1 Introduction

Metal nanoparticles have wide applications in catalysis [70],environmental remediation [71], DNA detection [72], highdensity data storage [73], and electronic and optical devices[74, 75]. In certain applications, one and only one layer (i.e.monolayer) of metal nanoparticles is required. To achievemetal nanoparticle monolayers, drop-casting [73] or spin-coating [76] of nanoparticle colloidal solutions on a sub-strate followed by solvent evaporation, or self-assembly us-ing a Langmuir–Blodgett technique [77] have been used.However, these techniques involve wet chemical reactionsand lengthy processing times that increase cost. Further-more, they often require thermal annealing to improve theadhesion of the particles to the substrate [73]. This not onlyincreases the cost, but is also incompatible with substrateslike plastics that cannot withstand high temperatures. Previ-ously, an excimer laser has been used to change the shapeof noble metal islands on a quartz substrate into sphericalnanoparticles [78–81]. Usually, multiple laser pulses wereused for this process.

In this section, a simple method for manufacturing metalnanoparticle monolayers using a single laser pulse to meltmetal thin films and metal lines on various substrates is dis-cussed. The method starts with a thin metal film depositedon a substrate or metal lines patterned by NIL, followedby melting using a single XeCl excimer laser pulse. Dur-ing laser melting, the originally continuous metal thin filmor metal lines break into a monolayer of partial sphericalnanoparticles or an array of metal particles due to the dewet-ting of the metal from the substrate surface. Various metals(e.g., noble metals (Ag, Au, Pt), magnetic metals (Fe, Co,Ni, Cr), refractory metals (W), etc.) on different substrates(silicon, fused silica and plastics) were studied. The effect ofsubstrate material, film thickness, laser fluence, and ambientconditions on the formation of nanoparticles were explored.

5.2 Metal nanoparticles from continuous films

5.2.1 Experiments

The substrates used in this study were silicon, fused silicaand plastic (e.g., 1/16′′ thick polycarbonate sheets). Siliconand fused silica wafers were first cleaned using a RCA #1solution at 80◦C for 30 min. Some silicon wafers were thendipped in diluted HF acid (1:50) for one minute to removethe native oxide before the metal deposition. The plasticswere cleaned in boiling DI water with a small amount ofMicro-90 cleaning solution for 10 min, followed by a 30-min ultrasonic bath, and then rinsed in running DI water


for 10 min. Noble metals (Au, Ag, Pt), magnetic metalsand alloys (Fe, Co, Ni, Cr, permalloy), and refractory metals(W) were investigated. Thin metal films (2 to 20 nm thick)were deposited on cleaned substrates using an electron beamevaporator under a base vacuum better than 2 × 10−6 Torr.All the as-deposited thin films were continuous except forthree cases: Ag and Au films on Si, and 2.1 nm thick Pt filmon fused silica, as examined by an SEM.

The excimer laser was used to melt the metal films onsubstrates. In order to cover a large area, a step and repeatexposure scheme was used. A single pulse was found to besufficient to fragment a metal film into nanoparticles. Mostof the melting experiments were carried out in air, while oth-ers were done in liquids such as acetonitrile to avoid oxida-tion of the nanoparticles and to study the effect of interfacialenvironments. Size distribution of the nanoparticles was an-alyzed from SEM pictures using commercial software (Im-age Pro-Plus) [44]. UV-Vis absorption spectra were mea-sured for Ag nanoparticle monolayers on UV-grade fusedsilica using a UV-Vis spectrophotometer scanning from 200to 800 nm. The chemical composition of some nanoparticleswas analyzed using energy dispersive X-ray analysis (EDX).

5.2.2 Results

After a single laser pulse exposure, continuous thin films ofmost metals (except Ti on a Si substrate) were turned intonanoparticle monolayers. Part of the metallic thin films, ex-perimental conditions and results are tabulated in Table 5.

Pure metal nanoparticles For almost all the pure metals(Au, Ag, Pt, Fe, Co, Ni, W, etc.) tested in this study, sub-100 nm diameter nanoparticles were achieved; provided thata proper substrate and laser fluence were used. For exam-ple, 5 nm Pt was deposited on fused silica, Si (with na-tive oxide) wafers. After a single laser pulse of 680 and

740 mJ/cm2, respectively, monolayers of Pt nanoparticleswere achieved (Figs. 60a, b). These particles have almostthe same size distribution (Figs. 60d, e), 9.8 ± 4.2 nm for(d) and 10.1 ± 4.4 nm for (e). However, for a 5 nm thick Ptfilm on the silicon wafer without native oxide, the resultantstructures are not particles (Fig. 60c).

Metal nanoparticles on plastic substrate Metal nanopar-ticle monolayers on plastic substrates were also fabricated.A single laser pulse of 160 mJ/cm2 incident on a 10 nmthick Ag film on a polycarbonate substrate produces a mono-layer of Ag nanoparticles (Fig. 61). There is no signifi-cant damage to the substrate due to the short pulse widthof the laser and the fact that polycarbonate does not sub-stantially absorb the 308 nm wavelength UV laser (<1%for a 20 µm thick sheet [82]). Recently, low-temperatureplasma enhanced chemical vapor deposition (PECVD) hasbeen introduced to grow carbon nanotubes (CNT) on varioussubstrates [83, 84]. Our current result suggests that smallermetal catalyst particles can be made on plastic substrates.This may create an opportunity for making smaller and moreuniform CNTs for devices on plastics, such as flexible dis-plays.

Alloy metal nanoparticles In addition to pure metals,permalloy nanoparticle monolayers were also fabricated bythe same technique. These nanoparticles have been used ascatalysts for carbon nanotube growth by chemical vapor de-position, increasing the diameter uniformity and yield.

Figure 62a shows a 2 nm thick permalloy film on a Sisubstrate after exposure to a single laser pulse with a fluenceof 1 J/cm2. The particles have a diameter distribution cen-tered at about 10 nm (10.4 ± 3.8 nm) (Fig. 62b). After CNTgrowth with 1 LPM (liters per minute) ethylene at 700◦Cfor 10 min, the typical TEM image shows uniform CNTs

Table 5 Partial list of metallic thin films and resultant particle sizes

Category Film Thickness Substrate Fluence Particle size

Material (nm) (mJ/cm2) (nm)

Nobel Aua 10 Fused Silica 194 106.6 ± 55.8

Agb 5 Siliconc 310 25.2 ± 12.6

Pt 5 Fused Silica 680 9.8 ± 4.2

Magnetic Co 5 Silicon 736 21.5 ± 5.3

Ni 5 Silicon 1000 23.1 ± 10.1

Permalloy 5 Silicon 1000 23.2 ± 6.8

Refractory W 6 Silicon 640 15.9 ± 8.9

aWith 2 nm Ti as sticking layerbAg films are exposed to laser in acetonitrilecSi substrates in this table are all with native oxide


Fig. 60 SEM micrograph of the metal nanostructures produced from5 nm thick Pt film on (a) fused silica substrate, (b) silicon with na-tive oxide, and (c) silicon without oxide, with a laser fluence of 680,740 and 740 mJ/cm2, respectively. Only in (a) and (b), the metal film

became a monolayer of Pt nanoparticles, while in (c) there were noparticles formed. (d), (e) are the diameter histograms for nanoparticlesin (a) and (b), respectively. Note they are both centered at 10 nm withsimilar distribution


Fig. 61 Ag nanoparticles fabricated by a 160 mJ/cm2 laser pulse inci-dent on a 10 nm thick silver film on a polycarbonate substrate

(Fig. 62c). One hundred arbitrary selected CNTs from theTEM images are measured, and the diameter distribution iscentered around 10 nm (10.9±1.9 nm) (Fig. 62d). For com-parison, the typical image and diameter distribution of CNTsgrown in the unexposed area are also shown (Figs. 62e, f).CNTs grown directly on the as-deposited film had a diame-ter distribution centered at 15 nm (14.8 ± 2.5 nm). Compar-ison between Figs. 62d and 62f shows that CNTs grown onthe nanoparticles produced by laser exposure have smallerdiameters and a narrower size distribution. Laser exposureFe films also enhanced the yield for SWNTs growth whenmethane was used as the precursor gas.

Particle formation and alloying with a single pulse Bi-layer metal films can be turned into alloy nanoparticles withsingle pulse laser exposure as well. For example, a 4 nmthick Cu film was deposited on a silicon substrate, followedby a 5 nm thick Ni film deposition. A laser pulse with flu-ence of 1130 mJ/cm2 breaks the thin film into nanoparticleswith an average size of about 48 nm. To determine the chem-ical composition of the nanoparticles from bi-layer metals,EDX was used on single nanoparticles. In order to avoidthe strong Si background signal in EDX, larger particles ofheavier metals were made from thicker metal films. For ex-ample, 6.6 nm Ni and 4.4 nm W thin films were deposited ona Si substrate, followed by exposure to a single laser pulse(500 mJ/cm2). A typical EDX spectrum of a single particleis shown in Fig. 63, which exhibits both Ni and W peaks,indicating that the particles are NiW alloys rather than pureNi or W. This suggests that nanoparticle formation and al-loying can be achieved at the same time during a single laserpulse exposure. Similar results were achieved for Pt–Co andPt–Ni systems, which are widely used as catalysts for fuelcell applications [85].

Large-area nanoparticles for optical applications Al-though the laser spot is only several mm in size, a largercoverage area can be achieved by a step and repeat exposuresystem. A 5 nm thick Ag film was deposited on a piece of0.5 mm thick UV grade fused silica (2 cm by 1 cm). Step andrepeat exposure was carried out for Ag films immersed inacetonitrile, a solvent that does not absorb 308 nm UV light.Each spot was exposed to a single pulse of ∼310 mJ/cm2.The average particle size was about 25 nm. A typical UV-Visspectrum of the resulting nanoparticle monolayers (Fig. 64)exhibited an absorption peak at 395 nm.

5.2.3 Discussion

Effect of substrate Nanoparticle monolayers were ob-tained for metal thin films on fused silica and silicon sub-strates covered with native oxide. However, this is not thecase for silicon substrate that had been dipped in diluted HFacid. This suggests that molten Pt de-wets with SiO2, butnot with Si, which might be a result of reactions betweenthe molten Pt and Si substrate (Fig. 60).

Dependence of particle size on film thickness The parti-cle size is dependent on the film thickness. Under otherwiseidentical conditions, a thicker film usually results in largerparticles. In order to explore the correlation between filmthickness and particle size, Pt thin films with thickness of2.1, 5.1, 10.0, 16.1 and 20.8 nm were deposited on fusedsilica substrates and exposed to single laser pulses with thesame laser fluence (∼650 mJ/cm2). There was a linear rela-tionship between the average particle size and the film thick-ness except for the 2.1 nm case (Fig. 65). It was suggestedthat the particle formation is a spinodal dewetting process[86] in which the particle size should be a monotonic in-creasing function of film thickness. The disruptive case forthe 2.1 nm thick film might be a result of the fact that thefilm is actually isolated islands or networks and not contin-uous. Upon laser exposure, the islands shrink into spheresrather than breaking into even smaller ones. The typical laserinduced shape/size change described in references [78–81]occurs.

Effect of laser fluence The laser fluence required to ob-tain nanoparticles varies from metal to metal. However, foreach particular metal (alloy), there is a laser fluence win-dow in which the resultant nanoparticles have similar sizedistributions. The fluence window can be as large as 100%for many metals or alloys on both silicon (with native ox-ide) and fused silica substrates (Fig. 66). This means thatas long as the laser fluence is high enough to melt the thinfilm without noticeable evaporation, varying the laser flu-ence does not change the size distribution of the nanopar-ticles. This finding indicates that our approach has a high


Fig. 62 (a) SEM image of nanoparticles resulting from a 2 nm thickpermalloy film after exposure to a single pulse laser with a fluenceof 1 J/cm2; (b) histogram of the diameter for permalloy nanoparticlein (a); (c) a typical TEM image of MWNTs grown with nanoparticle

catalysts shown in (a), and (d) the corresponding MWNT diameter his-togram; (e) a typical TEM image of MWNTs grown using as-deposited2 nm thick permalloy film as a catalyst, and (f) their diameter distribu-tion histogram


Fig. 63 Typical EDX spectrum of a single particle made fromNi/Wbi-layer metal thin film, suggesting that the particle is an alloy.The inset SEM image shows the NiW particles used for the analy-sis which were intentionally made large for the ease of EDX analysis.Scale bar: 200 nm

Fig. 64 UV-Vis absorption spectrum for a Ag nanoparticle monolayeron a fused silica surface with an absorption peak at 395 nm. Inset is theSEM image of the nanoparticles which have an average size of 25 nm

tolerance for energy fluctuation from pulse to pulse, typi-cally a problem for gas lasers. Two other important pointscan be drawn from Fig. 66. First, under otherwise identicalconditions, a 5 nm thick Ag film results in larger particlesthan those for a 5 nm thick Co or permalloy film. This mightbe attributed to the difference in the surface tension betweenAg and Co (or permalloy). Since Co has a higher surfacetension than Ag [87, 88], the liquid droplets have a strongertendency to shrink into a smaller size. Second, a 5 nm thickpermalloy thin film results in a larger particle size than that

Fig. 65 Dependence of the average Pt particle diameter on the initialfilm thickness for fused silica substrate. Particle diameter is propor-tional to the film thickness from 5.1 to 20.8 nm, but not at 2.1 nm

Fig. 66 The average particle diameter as a function of laser fluencesfor different metal thin films on silicon or fused silica substrates. Notethat particle diameter is insensitive to laser fluence within a certainrange

of a 2 nm thick film, confirming that the particle size is re-lated to the original film thickness.

5.3 Nanoparticle arrays by laser fragmentation of metallines

As shown in Table 5, the nanoparticles formed by expos-ing a continuous metal film to laser pulse have a wide sizedistribution, and exhibit almost no periodicity. In order tocontrol the particles periodicity and size distribution, NIL isfirst used to pattern the metal films. Since the lines are iso-lated from each other, this method is expected to improve


Fig. 67 Au nanodots formed by fragmentation of a blank thin film(left column) and a 200 nm pitch, 100 nm line width grating (rightcolumn) on fused silica substrates. (a) SEM image of nanodots from acontinuous film; (b) histogram of the particle size distribution for (a),the size is 106.6 ± 55.8 nm; (c) FFT image for (a), showing no regularperiodicity for the particles in (a); (d) SEM image of nanodots fromAu lines; (e) histogram of the particle size distribution for (d), the size

is 65.2 ± 10.4 nm; (f) FFT image for (d), showing a periodicity of220 ± 70 nm along the original grating line direction and the original200 ± 9 nm grating period in the orthogonal direction. In both casesthe films are 10 nm thick (with 2 nm Ti adhesion layer), and the laserfluences are 194 mJ/cm2. Insets in (a) and (d) are schematics of thestarting structures (blank thin film and nanogratings) on substrates [89]

both the periodicity and the size distribution of the resultantnanoparticles.

To demonstrate the idea, 10 nm thick Au film was de-posited on a fused silica substrate, with a 2 nm thick Ti filmas the sticking layer [89]. Meanwhile, 10 nm Au/2 nm Timetal lines of 200 nm pitch and 100 nm width were fab-ricated on fused silica substrates using NIL, metal evapo-

ration and liftoff. Single laser pulses of 194 mJ/cm2 wereused for exposing both samples. With a continuous film,the particle size is 106.6 ± 55.8 nm with a broad distrib-ution of period (Figs. 67a, b, c). However, with the samelaser fluence, the diameter of the nanodots formed from the200 nm pitch gratings (10 nm Au/ 2nm Ti on fused silica)was 65.2 ± 10.4 nm, both the size and size distribution have


been improved. The period along the original grating linedirection is 220 ± 70 nm and that perpendicular to the grat-ing line is 200 ± 9 nm (the same as the original grating pe-riod) (Figs. 67d, e and f). The relative standard variations (1sigma) of the size and pitch distribution were significantlybetter than those of nanodots from a blank thin film.

The fragmentation of molten metal lines into dot arrayscan be explained using the Rayleigh instability theory [90].According to this theory, a liquid cylinder of a radius R willbecome unstable and starts to break into periodic dropletsof a critical wavelength, λc. If it is perturbed along thecylinder longitudinal direction, the critical wavelength canbe calculated using λc = 2πR and the maximum (equilib-rium) wavelength is λm = 2

√2πR [91]. We find the pitch

of the nanodots along the original grating lines agrees withRayleigh instability model. For example, 80 nm wide Crlines of 200 nm pitch broke into the dots with an averageperiod of about 250 nm along the original grating line direc-tion. The ratio of the nanodot pitch to the original gratingline width is 3.12, which is close to the predicted value (π).

5.4 Periodicity engineering of nanoparticle arrays

To further improve the periodicity along the nanograting di-rection, two approaches were proposed. The first approachutilizes differences in wettability and the second approachtakes advantage of surface topography. Figure 68 shows theprinciple of the first approach. Lines of adhesion material(A) are patterned and deposited on a substrate (Fig. 68a).Then, another set of lines are patterned at an angle to thefirst ones. The angle can be adjusted according to the appli-cation. The particle material (B) is then deposited, formingthe crossbar structure (Fig. 68b). This set of lines is eitherin contact with the substrate, which is non-wetting to mate-rial B, or the adhesion lines, which are wetting to materialB. During the fragmentation process (e.g., pulsed laser melt-ing), the molten material B will preferably flow to the crosspoints where A and B join due to the better adhesion at thoseareas, hence forming nanoparticle arrays with regular peri-odicity predetermined by the lines of A and B (Fig. 68c).

Another approach is to use surface topography differ-ences, as shown in Fig. 69. Shallow trenches are first pat-terned and etched into the substrate (Fig. 69a), followed bymetal line deposition at an angle to the trenches (Fig. 69b).During the fragmentation process, the molten material tendsto flow to the trench to minimize the system energy and staysat the cross points of the trench and the original metal lines,resulting in a regular metal nanoparticle array with period-icity determined by the shallow trenches and the pitch of theoriginal metal lines.

To test the principle, both Au and permalloy nanograt-ings were fabricated. For Au lines, fused silica substratewith 10 nm deep, 70 nm wide and 200 nm pitch shallow

Fig. 68 Principle of periodicity engineering using difference of wet-tability. (a) Patterning lines of adhesion material are deposited on asubstrate; (b) Lines of particle material are deposited with an angle(adjustable) to the first layer of lines; (c) During the fragmentationprocess, the molten particle material directly in touch with the non-wetting substrate flows to the cross points of the two materials due tothe difference in the wettability, maintaining a regular periodicity de-termined by the two layers of lines

trenches were used. Then 2 nm thick Ti and 10 nm thick Aulines were deposited using NIL, O2 RIE, metal evaporationand liftoff. Similarly, 7 nm thick permalloy lines were de-posited on 210 nm thick SiO2 capped Si wafer with 10 nmdeep trenches that were 70 nm wide and had 200 nm pitch.The metal lines are nearly perpendicular to the trench di-rections in both cases. After exposing the samples to a sin-gle laser pulse of 325 and 790 mJ/cm2 for Au and permal-loy, respectively, regular arrays of metal nanoparticles weremade on the substrates (Figs. 70a, c). The periodicity alongthe original metal lines and trenches is shown in the FFTimages (Figs. 70b, d). A closer look at the SEM images(insets in both Figs. 70a and 70c) clearly indicates that theresultant nanoparticles are sitting in the cross points of theoriginal metal lines and the shallow trenches, favoring min-imum system energy. The fast Fourier transfer (FFT) analy-sis showed the periodicity of nanodots in the original metalline direction was 200.0 ± 11.3 nm and in the normal di-rection was 200.0 ± 6.9 nm (Fig. 70b) for Au, and thosefor permalloy were 200.0 ± 7.4 nm and 200.0 ± 5.8 nm,respectively (Fig. 70d). Compared with the nanodot arraysby fragmentation of a blank Au film (Fig. 67a, c) (whichhad no certain periodicity), and those by fragmentation ofAu nanogratings on a flat surface (Fig. 67d, f) (which had a


Fig. 69 Principle of periodicity engineering using surface topography.(a) Patterning of shallow trenches on a substrate; (b) Lines of particlematerial are deposited with an angle (adjustable) to the trench direc-tion; (c) During the fragmentation process, the molten particle materialflows into the trenches and resolidifies at the cross points of the trenchand the metal lines due to a lower energy on those sites, maintaining aregular periodicity determined by the original metal lines and the shal-low trenches [89]

pitch of 220 ± 70 nm in the original grating line direction),the pitches of the nanodot arrays in Fig. 70 were much moreuniform because they were predetermined by the pitches ofthe shallow trenches on the substrate surface and the origi-nal metal nanogratings. It is also interesting to note that inFig. 70, the permalloy has a more uniform periodicity thanAu in both directions. There are also some missing dots inthe Au nanodots array and some cloud in the FFT image(Figs. 70a, b), a possible reason for this is that the migrationof Au at molten state because it has higher mobility.

The results in Fig. 70 suggest that although the pitch ofnanodots depends heavily on the original nanowire widthaccording to Rayleigh instability theory, they can be regu-lated using substrate surface topography. As a result, the fi-nal nanostructure will be determined by the interplay of thenatural instability process and the substrate surface topogra-phy. Figure 71 summarized the possible nanostructures af-ter fragmentation of nanogratings on different surface reliefstructures. When the pitch of the shallow trenches is smallerthan the maximum instability wavelength (2

√2πR), there

will be only one dot at each trench/grating intersection point.And the pitch of the nanodots along the original grating di-rection is determined by the original trench pitch. On theother hand, if the pitch of the surface relief structure is larger

than the maximum wavelength 2√

2πR, one shall expectthat there will be multiple dots formed at one intersectionpoint. This is important because it suggests that complexnanodot arrays can be formed by engineering the surfacetopography.

In summary, we have used a single laser pulse to fab-ricate various kinds of metal nanoparticle monolayers onsilicon, fused silica and plastic substrates. The particle sizewas mainly determined by the film thickness, while otherfactors, such as the laser fluence, have little effect on thesize distribution. Nanoparticle formation and nano-alloyingcan be achieved during a single laser pulse exposure of bi-layer metal thin films. A large coverage area was achievedusing a step and repeat method, and optical properties ofAg nanoparticles on fused silica were studied. To get metalnanoparticles with better periodicity, metal nanogratingswere used as the starting structure. This approach not onlyimproves the size distribution, but also introduced a certaindegree of periodicity along the original grating direction ac-cording to Rayleigh instability. The periodicity of the nan-odot arrays was further engineered using a surface reliefstructure, resulting in a regular 2D array of nanodots withregular periodicity along both directions. As a simple fabri-cation method, it could be extended to other metals and haswide applications in many areas such as magnetics, plas-monics, surface enhanced Raman scattering and other pho-tonic devices.

6 Sub-10-nm self-enclosed nanofluidic channels

6.1 Introduction

Nanofluidic channels are important tools in the emergingbionanotechnology field for manipulation and analysis ofbiomolecules at the single molecule level [92]. Fabricationof extremely small channels (and arrays) requires state-of-the-art nanopatterning techniques. Equally important is thesealing of trenches into functional channels. For the pat-terning of nanoscale trenches, EBL [93, 94] or focused ion-beam (FIB) milling techniques [95] have been widely used.However, these methods suffer from low throughput andare expensive. The advent of NIL [8–10] has greatly alle-viated some of these problems. Using NIL, low-cost waferscale fabrication of high density nanofluidic channel arraysfor DNA manipulation and analysis has been demonstrated[96–98].

The other challenge, i.e., sealing of nanochannels is notas easy as it sounds. Soft elastomers such as PDMS wereused to seal the channels [99], resulting in a uniform sealingover a larger area. However, channels might become cloggedsince the soft material can be easily pressed into the channel.Wafer bonding techniques [100], on the other hand, providea rigid seal, but are limited because they require defect free


Fig. 70 SEM and FFT images for regular Au and permalloy nanodotarrays on pre-patterned surfaces with shallow trenches. (a) Au nan-odot arrays fabricated by fragmentation of Au gratings (200 nm pitch,100 nm width, 10 nm thick, with 2 nm thick Ti as adhesion layer)on pre-patterned fused silica substrate. The dot size is 80.1 ± 8.1 nm.Laser fluence: 325 mJ/cm2. (b) FFT image for (a). (c) Permalloy nan-

odot arrays fabricated by fragmentation of permalloy gratings (200 nmpitch, 70 nm width, 7 nm thick) on a pre-patterned 210 nm thickSiO2 capped Si wafer. The dot size is 72.5 ± 6.8 nm. Laser fluence:790 mJ/cm2. (d) FFT image for (c). The substrates have trenches of200 nm pitch, 70 nm linewidth which are 10 nm deep [89]

Fig. 71 Summary of the relationship between the surface trench pitchand the resulting nanodot pitch. (a) For a flat surface, Λ = ∞, the pitchof the nanodot arrays along the original grating direction (λ) is about2πR, agreeing well with Rayleigh instability; (b) for the surface grat-

ing with a pitch Λ < 2√

2πR, λ is determined by the trench pitch Λ,λ = Λ; (c) for the surface trenches with Λ > 2

√2πR, λ < Λ, resulting

in particle doublet (or triplet, etc.) arrays [89]

surfaces. Cracks might be introduced when there is a mis-match in the thermal expansion between the substrate andthe cover material. Sealing of nanochannels has also beendone by deposition of sealing materials over sacrificial lay-ers such as polysilicon or polymers followed by wet etch-ing or thermal decomposition [101, 102]. These methodsusually require long etching times to remove all the sacri-ficial material and become increasingly difficult for smallernanochannels where the flow rates are very low. Sealing

techniques such as non-uniform deposition [96] and imprintof trench molds into thin polymer films [98] address some ofthese concerns, although they may not be suitable for com-plicated biochips with different functional devices.

In this section, a self-sealing technique to make enclosednanofluidic channel arrays from NIL patterned silicon nan-otrenches and pillar arrays is introduced. During the sealingprocess, a single laser pulse is used to melt the top layerof the nanostructures, inducing lateral flow which seals the


nanostructures before resolidification. The channel dimen-sions can be controlled by the laser fluence with fine sizecontrol by secondary thermal oxidation.

6.2 Self-sealing of Si trenches

The principle of our method is shown in Fig. 72 [103]. Sil-icon nanostructures (made by NIL) (Fig. 72a) are exposedto a UV laser pulse, which melts the top portion of thestructures (Fig. 72b). Due to the surface tension, the orig-inal rectangular cross section of the etched nanostructuresbecomes circular in shape during the reflow process. Withsufficient melting, adjacent structures join together to forman enclosed structure (Fig. 72c), and re-solidify (Fig. 72d).As an option, thermal oxidation can be carried out on thesestructures to further shrink the size (Fig. 72e), and to makethe cover optically transparent.

The fabrication of the self-enclosed Si channels includedthe following main steps: First, fabrication of the nanoscaletrenches (or pillars) on Si using NIL and RIE; second, lasermelting to seal the trenches and form channels; third, sizereduction using thermal oxidation. Fabrication of nanoscaletrench and pillar arrays started with clean p-type (100) sili-con wafers capped with 30 nm thermal oxide. After RCA #1cleaning, the wafers were first spin coated with a 180 nmthick thermoplastic imprint resist (NXR 1020, NanonexCorp, NJ) and then baked at 70◦C for 15 min to drive out theresidual solvent. Silicon master molds (made by interferencelithography and RIE) having parallel lines of 200 nm pitchor pillars of 970 nm pitch over 4 inch wafers were used toimprint the resists. The patterns on the imprinted resist werethen transferred to the thermal oxide layers by RIE in a Plas-matherm SLR 720 RIE system (Plasma-Therm/Unaxis, St.

Fig. 72 Schematics of self-sealing process for enclosed nanochannels.(a) Si nanostructures (lines of pillars) made by NIL; (b) the top portionof the Si structures is melted by laser; (c) the molten Si flows sidewardand joins the neighboring lines (pillars); (d) after resolidification, en-closed channels are formed; (e) channel size shrinking using thermaloxidation [103]

Petersburg, FL) using a CHF3/O2 chemistry, with 10 sccmCHF3, 1.5 sccm O2, 5 mTorr base pressure and 70 mW/cm2

power density. The patterned thermal SiO2 served as a hardmask for silicon etching using deep RIE (STS, Newport,UK). The Si etching had several cycles depending on thedesired depth. Each cycle started with an etch step (30 sccmSF6, 6 sccm O2, 600 W ICP power, 12 W platen power,15 mTorr base pressure, 7 sec), followed by a passivationstep (85 sccm C4F8, 600 W ICP power, 15 mTorr base pres-sure, 5 sec). The etched samples (200 nm pitch trenches withdifferent widths and depths, and 700 nm diameter pillars of970 nm pitch and 2 µm height) were dipped into diluted HFto strip the SiO2 mask, followed by a further cleaning usingRCA #1 for 15 min.

The fabricated structures were then exposed to the ex-cimer laser with a spot size of 3 × 3 mm2. In each case,a single laser pulse with an appropriate fluence was usedto melt the top of the structures just enough to result in auniform seal over the exposed area. Some channel sampleswere then put into a Tystar oxidation tube (Tystar Corp, CA)for a standard wet thermal oxidation process at 1,000◦C. Af-ter oxidation, the samples were cleaved 1 mm away from theedge of the sealed area and perpendicular to the channels toexpose the channel openings for SEM imaging and DNAstretching experiments.

A single laser pulse was found sufficient to melt the topportion of the fabricated structures and form enclosed 1Dor 2D nanochannels. Figure 73a shows the starting struc-ture, which consists of 100 nm wide trenches of 200 nmpitch and 640 nm depth. After exposure with a single laserpulse of 790 mJ/cm2, the top liquid layer joins together toform a cap, and forming enclosed channels of 100 nm width,∼250 nm height (Fig. 73b). Similarly, 970 nm pitch pillars

Fig. 73 (a) 100 nm wide, 200 nm pitch Si lines (640 nm high). (b) Asingle laser pulse of 790 mJ/cm2 turns nanolines in (a) into enclosedSi channels. (c) 700 nm diameter Si pillars (970 nm pitch, 2 µm high)were turned into channels upon exposure to a single laser pulse (d) of765 mJ/cm2. Insets in (b) and (d) are magnified images of the crosssection of the nanochannels [103]


Fig. 74 Effect of laser fluence on channel sealing quality. (a) With560 mJ/cm2, only the pillar tips are melted. Without enough moltenmaterial, the tips are turned into individual round dots. (b) With635 mJ/cm2, the top is partially sealed but with several holes on thesurface. (c) A completed sealing is achieved with a laser fluence of765 mJ/cm2. The starting structures are 700 nm diameter Si pillars of2 µm height, 970 nm pitch and 270 nm spacing

(270 nm spacing, 2 µm high) (Fig. 73c) are sealed, form-ing an interconnected network of 2D nanochannels (270 nmwide and 400 nm high) (Fig. 73d) upon exposure to a sin-gle pulse of 765 mJ/cm2. The laser fluence was optimizedto achieve successful nanochannel sealing. Laser fluencesranging from 250 to 1100 mJ/cm2 were tried for the 200 nmpitch Si trenches.

We found that laser fluence below 600 mJ/cm2 did notresult in neighboring molten silicon lines joining together,while fluence higher than 880 mJ/cm2 resulted in partial ab-lation of the lines (Fig. 74). An optimal processing windowfor the 200 nm pitch silicon lines was found to be 690–790 mJ/cm2. The molten time vs laser fluence for a flat(100) silicon wafer is plotted in Fig. 4. For a laser pulse of750 mJ/cm2, the molten time is about 60 ns. Since full seal-ing is completed while the silicon is in a liquid state, the es-timated flow speed of the molten silicon in this case is about

Fig. 75 With narrower trenches in Si, the top surface is flat after lasersealing. (a) 70 nm wide, 840 nm deep trenches of 200 nm pitch in Si.(b) With a single laser pulse of 656 mJ/cm2, the trenches are sealed intoenclosed channels with a flat top. (c) Tilt view of the sealed sample in(b), which clearly shows a flat and smooth top surface [103]

2.25 m/s for the 270 nm spacing Si pillar structures whereSi from each side of the gap needs to travel 135 nm in 60 ns.

The surface morphology after laser sealing was foundto be related to the trench size. For example, the surfaceof sealed 100 nm wide/200 nm pitch channels was rugged(Fig. 73b). However, with the same pitch, laser sealing of70 nm wide Si trenches (Fig. 75a) resulted in a flat andsmooth top surface (Figs. 75b, c). This is believed due tothe fact that in the latter case the distance needed for themolten Si to flow is shorter. In that case, a self-planarizationprocess of molten Si under surface tension (to minimize thesurface area) before it re-solidifies made the top surface flat.

6.3 Shrinking the channel sizes

6.3.1 By laser fluence

One way to control the channel size is by using differentlaser fluences, which will result in different molten times


Fig. 76 Different laser fluences result in different channel sizes. With665 mJ/cm2, the channel size by self-sealing of pillars (in Fig. 73c)is 270 nm by 750 nm (a), while that for a 765 mJ/cm2 laser pulse is270 nm by 400 nm (b) [103]

and different melting depths [104]. Higher laser fluencewill result in a molten layer that penetrates deeper into theSi surface. As a result, a smaller sized channel (reducedheight) can be made with a higher fluence. For example,for the starting pillar structures shown in Fig. 73c, the re-sulting channel size was 270 nm by 750 nm (Fig. 76a) afterbeing exposed with a laser fluence of 665 mJ/cm2, whilefor a 765 mJ/cm2 laser pulse, the channel was 270 nm by400 nm (Fig. 76b). A similar effect was observed with 1DSi trenches. For example, with 840 nm deep (70 nm wide,200 nm pitch) Si trenches, a laser pulse of 656 mJ/cm2 re-sulted in nanochannels of 450 nm height, while increasingthe fluence to 756 and 856 mJ/cm2 reduced the channelheight to 400 and 300 nm, respectively (Fig. 77).

6.3.2 By thermal oxidation

For fine size control, thermal oxidation can be used to fur-ther reduce the size of the nanochannels. Turning the seal-ing layer to transparent silicon oxide also makes the chan-nels suitable for fluorescence analysis of biomolecules suchas DNA. Self-sealed silicon nanochannels were put into theoxidation furnace for wet oxidation at 1,000◦C for differ-ent lengths of time. After oxidation, the samples were cut1 mm apart from the laser spot edge for SEM cross-sectionalimaging. Channel widths were measured after 30 to 180 minoxidation (Fig. 78). The 100 nm wide channels shrunk to50, 20, and 9 nm after 30, 60 and 90 min oxidation, respec-tively. The size reducing rate from 30 min to 90 min sloweddown. Further increasing the oxidation time to 180 min didnot continue to reduce the size.

6.4 Discussion on channel oxidation

6.4.1 Mechanism of the “self-limiting” oxidation

It is very interesting that the channel size did not changewith oxidation being prolonged from 90 min to 180 min

Fig. 77 Starting with 70 nm wide, 840 nm deep trenches in Si (seeFig. 75a for the original structure), a fluence of 656 mJ/cm2 results inchannel height of 450 nm (a); while with a laser fluence of 756 mJ/cm2

and 857 mJ/cm2, the channel height is 400 nm (b) and 300 nm (c),respectively [103]

(Fig. 78) for the 100 nm wide enclosed channels. In or-der to understand this “self-limiting” behavior, the samplesthat have been oxidized for different lengths of time werecleaved and dipped into diluted HF (1:100) for 2.5 min todelineate the SiO2/Si interface, as shown in Fig. 79. Notethat the channel sizes shown in Fig. 79 were a little bit largerthan their corresponding ones in Fig. 78 due to the HF etch.After 30 min thermal oxidation, there was a Si core existingin between each channel (Fig. 79a). With the oxidation timeincreased to 60 min, the Si core size was reduced signifi-cantly (Fig. 79b). After 90 min, the Si in between the chan-nels was consumed completely, and the enclosed channelsare all above the SiO2/Si interface line (Fig. 79c). Further


Fig. 78 Plot of channel size as a function of wet oxidation time. Insetsare the SEM cross-section images for 100, 50, 20 and 9 nm channels,respectively. Note that increasing the oxidation time from 90 min to180 min does not reduce the channel size further [103]

Fig. 79 Cross-sectional images of laser-sealed channels after oxi-dation of different durations (a) 30 min; (b) 60 min; (c) 90 min;(d) 180 nm, and 2.5 min diluted HF etch of oxide. The structure beforeoxidation is shown in the first inset in Fig. 78. The SiO2/Si interfacesare clearly shown after HF dip. Note that the channel sizes were a lit-tle bit larger due to the HF etch. Increasing the oxidation time from90 min to 180 min has pushed the SiO2/Si interface further into the Sibulk. Scale bars: 100 nm [103]

increasing the oxidation time from 90 min to 180 min didnot further reduce the channel diameter, but only grew anadditional layer of oxide below the channels (Fig. 79d).

6.4.2 Geometrical factors for the Si channels oxidation

The oxidation behavior of the enclosed Si channels dependsheavily on their geometries. One example is the channelwidth. For channels of 70 nm by 350 nm (cap thickness270 nm), 90 min wet oxidation at 1000◦C has resulted insmaller channels <12 nm in size (Fig. 80a). However, therewas still Si around the channels, which was further oxidizedwhen the oxidation time was increased to 180 min. As aresult, the channels were filled and did not exist any more(Fig. 80b). These results confirmed that the Si supply con-trols the final size/existence of the channels. The cap layerwhich encloses the channels also matters for the oxidationbehavior. Without the cap layer, the open channels were ox-idized much faster and in a less controllable fashion. Forexample, 70 nm wide, 200 nm pitch Si trenches of 500 nmdeep were oxidized at 1,000◦C (wet oxidation) (Fig. 81a).After only 5 min, the trenches were almost filled with SiO2,leaving 15 nm wide channels at the bottom (Fig. 81b). The15 nm channels disappeared when the samples were oxi-dized for 60 min (Fig. 81c). Similar phenomenon was ob-served for the 100 nm wide open trenches in Si as well. Thecap layer has provided an enclosed environment in whichthe oxygen supply is lower than the outside surface, thus theoxidation rate is reduced.

Other geometrical factors such as the thickness of thecap, the aspect ratio of the channel, etc. also play importantroles in oxidation. Exploration in these parameters will notonly help understanding the process, but also produce usefultransparent structures/devices for different applications.

6.5 DNA stretching demonstration

To demonstrate the continuity and optical transparency ofthe channels, we tested the stretch of λ-phage DNA, sus-pended in a buffer solution into the 20 nm wide, 60 nmtall nanochannels fabricated by self-sealing and oxidation(Fig. 78, 1 h oxidation). Oxidized samples were cleaved at

Fig. 80 (a) 90 min oxidation for the 70 nm by 350 nm enclosed Sitrenches results in <12 nm wide channels. Note that there is still Siaround the channel. (b) Further increasing the oxidation time to 180 nmmakes the channels disappear. These results confirmed that the Si sup-ply controls the final size/existence of the channels [103]


Fig. 81 (a) 70 nm wide,200 nm pitch open trenches of500 nm deep in Si. (b) After5 min wet oxidation at 1000◦C,the trenches are almost filledwith oxide except for 15 nmwide channels at the bottom. (c)After 60 min oxidation, thetrenches are totally filledwithout any channels left [103]

Fig. 82 CCD image of λ-phage DNA stretching inside of nanochan-nels which have oxidized for 1 h. The channel size is about 20 nm.The DNA molecules have been effectively stretched. This experimentdemonstrates that the nanochannels are continuous [103]

least 1 mm away from the edge of the sealed area and per-pendicular to the channels to expose the channel openings.λ phage DNA (48.5kbp, New England BioLabs) at a concen-tration of 1 µg/ml in 0.5 × TBE buffer (0.045 M tris-base,1 mM EDTA with 0.045 M boric acid) were loaded intothe channels by capillary action. The DNA was labeled withTOTO-1 (Molecular Probes, 1 dye molecule per 10 bp) andimaged using a Pentamax ICCD camera (Roper Scientific,NJ) on a Nikon Eclipse TE2000 microscope using laser flu-orescence microscopy (488 nm excitation/514 nm emission)with a Nikon 100× (oil) objective (NA = 1.4). Figure 82shows a typical image for the DNA. It can be concluded thatthe DNA molecules have been successfully loaded into thechannel and effectively stretched, which means the channelsare continuous.

The self-sealing method has several advantages. It is ul-trafast, simple, and cost effective. It does not involve a lotof complicated equipment and processing, extra material forthe seal or exceptionally flat surfaces. Due to its high-speednature, it is regarded as a low-temperature sealing process,so the sealing method is viable for substrates that do notwithstand high temperature processes. It brings negligiblethermal effect to other components on the same chip, too.Our method is highly selective to materials and only meltsthose which absorb the specific wavelength of a UV laser.Other materials on the same chip which do not absorb that

wavelength are not damaged. The process is area selective.The laser spot size is adjustable, so we can expose onlythose areas that need to be sealed while leaving other ar-eas untouched. This flexibility is necessary for complicatedbiochips that have different functional devices on one chip.Channel size can be controlled by different laser fluence orby oxidation. And lastly, this method can be extended tochannels made from other materials using a laser of a dif-ferent wavelength. For example, a CO2 laser can be used forsealing SiO2 trenches since it absorbs the laser pulse [105].

In summary, a simple method that can self-seal nanoflu-idic channels was developed. Using a single laser pulse, 1Dand 2D enclosed silicon nanochannels were fabricated. Thesize could be controlled by using different laser fluence orthermal oxidation. Nanochannel arrays with a feature sizedown to 9 nm were prepared using this method. The oxi-dation for certain structures was found to be self-limiting,and the mechanism was explored. DNA stretching using theoxidized channels (20 nm wide) was also demonstrated.

7 Concluding remarks

In this paper, a host of novel technologies have been de-veloped using a pulsed excimer laser together with NIL. Inall the cases, the laser pulse has been used as an ultrafastheating source which melts only a surface layer of a mater-ial, leaving negligible thermal effect on the substrate and/orother components on the same chip.

With laser assisted NIL (LAN) process, 100 nm wide(200 nm pitch) grating lines have been fabricated upon ex-posure to a single light pulse. Since the pulse duration isvery short, the heating of the substrate and mold is greatlyreduced. This has been verified by numerical simulations.The imprint time of LAN has been measured to be about200 ns.

Self-perfection by liquefaction (SPEL) has been pro-posed and demonstrated as a new paradigm to remove fab-rication defects and enhance nanostructure profiles. Threeforms of SPEL, namely, O-SPEL (in open space), C-SPEL(with a top plate in contact) and G-SPEL (with a top platea distance above), have been discussed. Using O-SPEL,


rough-edged Si and metal lines have been smoothed out andnon-ideal shaped nanopads have been turned into nearly per-fect shapes. From this point of view, the most importantachievement of O-SPEL is that 3σ LER of rough edged70 nm wide Cr lines was reduced from 8.4 to 1.5 nm,which is well beyond the capability of current semiconduc-tor industry. With C-SPEL, the cross-section profile can bemaintained while the LER is removed. The first observa-tion and exploration of G-SPEL has major impact both inscientific research and in industrial applications. Applica-tions of SPEL in making smooth round NIL pillar molds,and smooth Si waveguides that have propagation loss ordersof magnitude lower after SPEL, have been demonstrated.

Excimer laser was also used to make various kinds ofmetal nanoparticle monolayers on different substrates. Formetal thin film on a blank substrate, the particle size ismainly determined by the film thickness. For better period-icity, metal nanogratings were used as the starting structure.A surface relief structure with shallow trenches has beenintroduced to regulate the periodicity of the nanoparticlesalong the direction of the original metal lines.

Finally, a simple method that can self-seal nanofluidicchannels has been developed. Using a single laser pulse, 1Dand 2D self-enclosed silicon nanochannels have been fabri-cated. The size can be controlled by using different laser flu-ence or fine tuned using thermal oxidation, leading to SiO2

nanofluidic channels with 9 nm in. The self-limiting oxida-tion behavior of enclosed Si channels was studied, and DNAstretching using 20 nm wide channel arrays has been demon-strated.

It has to be pointed out that the current paper only demon-strates a limited number of applications of excimer laser innanofabrication. As a powerful tool, excimer laser has beendemonstrated in other fields and broader applications are yetto be developed.

Acknowledgements This work was partially supported by the USOffice of Naval Research (ONR) and the Defense Advanced ResearchProjects Agency (DARPA). We appreciate contributions from manyformer group members in the NanoStructure Laboratory (NSL) atPrinceton University, including but not limited to Chris Keimel, Dr.Zhaoning Yu, Dr. Wei Wu, Dr. Keith Morton, Dr. Haixiong Ge, Dr.Xingyu Huang, Dr. Dr. Zengli Fu, Dr. He Gao, Dr. Shufeng Bai, Dr.Xiaogang Liang, Dr. Patrick Murphy. We would also like to thank Ms.Dandan Zhang for some of the schematic illustrations.

References

1. T.H. Maiman, Stimulated optical radiation in ruby. Nature 187,493–494 (1960)

2. A.L. Schawlow, C.H. Townes, Infrared and optical masers. Phys.Rev. 112, 1940–1949 (1958)

3. J.J. Ewing, Rare-gas halide lasers. Phys. Today 31, 32–39 (1978)4. Q.F. Xia, Nanostructure Engineering Using Pulsed Laser

and Nanoimprint Lithography (Princeton University, Princeton,2007)

5. J. Boneberg, J. Bischof, P. Leiderer, Nanosecond time-resolvedreflectivity determination of the melting of metals upon pulsedlaser annealing. Opt. Commun. 174, 145–149 (2000)

6. J. Liu, H. Kurz, N. Bloembergen, Picosecond time-resolvedplasma and temperature-induced changes of reflectivity andtransmission in silicon. Appl. Phys. Lett. 41, 643–646 (1982)

7. P.G. Carey, B.J. Woratschek, F. Bachmann, Progress toward ex-cimerlaser metal planarization and via hole filling using in situmonitoring. Microelectron. Eng. 20, 89–106 (1993)

8. S.Y. Chou, P.R. Krauss, P.J. Renstrom, Imprint of sub-25 nmvias and trenches in polymers. Appl. Phys. Lett. 67, 3114–3116(1995)

9. S.Y. Chou, P.R. Krauss, P.J. Renstrom, Imprint lithography with25-nanometer resolution. Science 272, 85–87 (1996)

10. S.Y. Chou, P.R. Krauss, W. Zhang, L.J. Guo, L. Zhuang, Sub-10 nm imprint lithography and applications. J. Vac. Sci. Technol.B 15(6), 2897–2904 (1997)

11. D.R. Lide, CRC Handbook of Chemistry and Physics, 79th edn.(CRC Press, Boca Raton, 1998)

12. W. Zhang, S.Y. Chou, Multilevel nanoimprint lithography withsubmicron alignment over 4 in. Si wafers. Appl. Phys. Lett. 79,845–847 (2001)

13. S.Y. Chou, C. Keimel, J. Gu, Ultrafast and direct imprint ofnanostructures in silicon. Nature 417(6891), 835–837 (2002)

14. Z.N. Yu, H. Gao, S.Y. Chou, In situ real time process characteri-zation in nanoimprint lithography using time-resolved diffractivescatterometry. Appl. Phys. Lett. 85, 4166–4168 (2004)

15. Z.N. Yu, Subwavelength Gratings and Applications (PrincetonUniversity, Princeton, 2003)

16. Q.F. Xia, C. Keimel, H.X. Ge, Z.N. Yu, W. Wu, S.Y. Chou, Ul-trafast patterning of nanostructures in polymers using laser as-sisted nanoimprint lithography. Appl. Phys. Lett. 83, 4417–4419(2003)

17. A. Yariv, Optical Electronics, 4th edn. (Saunders College Pub-lishing, Philadelphia, 1991)

18. GSolver, http://www.gsolver.com/19. E.D. Palik, Handbook of Optical Constants of Solids (Academic

Press, New York, 1985)20. S.A. Campbell, The Science and Engineering of Microelectronic

Fabrication, 2nd edn. (Oxford University Press, Oxford, 2001)21. ABAQUS, http://www.abaqus.com/22. H. Du, R.C.A. Fuh, J.Z. Li, L.A. Corkan, J.S. Lindsey, Pho-

tochemCAD: A computer-aided design and research tool in pho-tochemistry. Photochem. Photobiol. 68, 141–142 (1998)

23. Z.N. Yu, H. Gao, S.Y. Chou, New developments in real-time im-print monitoring by scattering-of-light (RIMS), in The Third In-ternational Conference on Nanoimprint and Nanoprint Technol-ogy (NNT04). Vienna, Austria, 2004

24. R.L. Morrison, Symmetries that simplify the design of spot arrayphase gratings. J. Opt. Soc. Am. A 9, 464–471 (1992)

25. H. Gao, Nanoimprint Methods, Monitoring and Processes(Princeton University, Princeton, 2007)

26. Q.F. Xia, Z.N. Yu, H. Gao, S.Y. Chou, In situ real time monitor-ing of nanosecond imprint process. Appl. Phys. Lett. 89, 073107(2006)

27. Q. Xia, TRR measurement for polymers. Unpublished, 200328. D.W. van Krevelen, Properties of Polymers, 3rd edn. (Elsevier

Science, Amsterdam, 1990)29. H.C. Scheer, H. Schulz, A contribution to the flow behaviour of

thin polymer films during hot embossing lithography. Microelec-tron. Eng. 56, 311–332 (2001)

30. H.D. Rowland, A.C. Sun, P.R. Schunk, W.P. King, Impact ofpolymer film thickness and cavity size on polymer flow duringembossing: toward process design rules for nanoimprint lithog-raphy. J. Micromech. Microeng. 15, 2414–2425 (2005)

31. Newport high-speed detector, http://www.newport.com

http://www.gsolver.com/

http://www.abaqus.com/

http://www.newport.com


32. Tektronix digital storage oscilloscope, http://www.tektronix.com

33. S.Y. Chou, Q.F. Xia, Improved nanofabrication through guidedtransient-liquefaction. Nat. Nanotechnol. 3, 295–300 (2008)

34. S.Y. Chou, Q.F. Xia, Self-repair and enhancement of nanos-tructures by liquificaiton under guiding conditions. US Patent7282456, 2007

35. International Technology Roadmap for Semiconductors, 2006Update. http://www.itrs.net/Links/2006Update/FinalToPost/08_Lithography2006Update.pdf, 2006: p. Table 77b

36. C. Herring, Some theorems on the free energies of crystal sur-faces. Phys. Rev. 82, 87–93 (1951)

37. Z.N. Yu, L. Chen, W. Wu, H.X. Ge, S.Y. Chou, Fabricationof nanoscale gratings with reduced line edge roughness usingnanoimprint lithography. J. Vac. Sci. Technol. B 21(5), 2089–2092 (2003)

38. P.L. Silvestrelli, A. Alavi, M. Parrinello, D. Frenkel, Ab initiomolecular dynamics simulation of laser melting of silicon. Phys.Rev. Lett. 77, 3149–3152 (1996)

39. M. Delfino, T.A. Reifsteck, Laser activated flow of phosphosil-icate glass in integrated circuit devices. Electron Device Lett.3(5), 116–118 (1982)

40. D.K. Armani, T.J. Kippenberg, S.M. Spillane, K.J. Vahala, Ultra-high-Q toroid microcavity on a chip. Nature 421(6926), 925–928(2003)

41. D.B. Tuckerman, A.H. Weisberg, Planarization of gold and alu-minum thin-films using a pulsed laser. IEEE Electron. DeviceLett. 7(1), 1–4 (1986)

42. T. Sameshima, Self organized grain growth larger than 1 µmthrough pulsed-laser-induced melting of silicon films. Jpn. J.Appl. Phys. 2 32(10B), L1485–L1488 (1993)

43. V. Constantoudis, G.P. Patsis, A. Tserepi, E. Gogolides, Quan-tification of line-edge roughness of photoresists. II. Scaling andfractal analysis and the best roughness descriptors. J. Vac. Sci.Technol. B 21(3), 1019–1026 (2003)

44. Image Pro Plus, http://www.mediacy.com45. S.E. Orchard, On surface levelling in viscous liquids and gels.

Appl. Sci. Res. A11, 451–464 (1962)46. Y. Sato, Y. Kameda, T. Nagasawa, T. Sakamoto, S. Moriguchi,

T. Yamamura, Y. Waseda, Viscosity of molten silicon and thefactors affecting measurement. J. Cryst. Growth 249(3–4), 404–415 (2003)

47. P.H. Keck, W. Vanhorn, The surface tension of liquid silicon andgermanium. Phys. Rev. 91(3), 512–513 (1953)

48. R.F. Brooks, A.T. Dinsdale, P.N. Quested, The measurement ofviscosity of alloys—a review of methods, data and models. Meas.Sci. Technol. 16(2), 354–362 (2005)

49. B.J. Keene, Review of data for the surface tension of pure metals.Int. Mat. Rev. 38(4), 157–192 (1993)

50. J. Brandrup, E.H. Immergut, Polymer Handbook, 3rd edn. (Wi-ley, New York, 1989)

51. S.Y. Chou, Q.F. Xia, Ultrafast selective self-repair of nanostruc-tures by liquefaction, in The 49th International Conference onElectron, Ion and Photon Beam Technology and Nanofabrication(EIPBN) (AVS, Orlando, 2005)

52. E. Yablonovitch, T. Gmitter, Wetting angles and surface tensionin the crystallization of thin liquid films. J. Electrochem. Soc.131, 2625–2630 (1984)

53. S.Y. Chou, L. Zhuang, Lithographically induced self-assemblyof periodic polymer micropillar arrays. J. Vac. Sci. Technol. B17, 3197–3202 (1999)

54. S.Y. Chou, L. Zhuang, L.J. Guo, Lithographically induced self-construction of polymer microstructures for resistless patterning.Appl. Phys. Lett. 75, 1004–1006 (1999)

55. L. Wu, S.Y. Chou, Electrohydrodynamic instability of a thin filmof viscoelastic polymer underneath a lithographically manufac-tured mask. J. Non-Newtonian Fluid. Mech. 125, 91–99 (2005)

56. N. Otsu, A threshold selection method from gray level his-tograms. IEEE Trans. Syst. Man Cybern. 9, 62–66 (1979)

57. W. Zhang, Development of Large-Area and Multilevel Nanoim-print Lithography and the Applications in MOSFETs (PrincetonUniversity, Princeton, 2001)

58. S.Y. Chou, P. Krauss, Imprint lithography with sub-10 nm fea-ture size and high throughput. Microelectron. Eng. 35, 237–240(1997)

59. S. Park, H. Schift, H.H. Solak, J. Gobrecht, Stamps for nanoim-print lithography by extreme ultraviolet interference lithography.J. Vac. Sci. Technol. B 22(6), 3246–3250 (2004)

60. K. Ansari, J.A. van Kan, A.A. Bettiol, F. Watt, Fabrication ofhigh aspect ratio 100 nm metallic stamps for nanoimprint litho-graphy using proton beam writing. Appl. Phys. Lett. 85(3), 476–478 (2004)

61. K. Watanabe, T. Morita, R. Kometani, T. Hoshino, K. Kondo, K.Kanda, Y. Haruyama, T. Kaito, J. Fujita, M. Ishida, Y. Ochiai,T. Tajima, S. Matsui, Nanoimprint using three-dimensional mi-crolens mold made by focused-ion-beam chemical vapor deposi-tion. J. Vac. Sci. Technol. B 22(1), 22–26 (2004)

62. I. Junarsa, P.F. Nealey, Fabrication of masters for nanoimprint,step and flash, and soft lithography using hydrogen silsesquiox-ane and x-ray lithography. J. Vac. Sci. Technol. B 22(6), 2685–2690 (2004)

63. Q.F. Xia, S.Y. Chou, Fabrication of sub-25 nm diameter pillarnanoimprint molds with smooth sidewalls using self perfectionby liquefaction (SPEL) and reactive ion etching. Nanotechnology19, 455301 (2008)

64. L.C. Kimerling, L.D. Negro, S. Saini, Y. Yi, D. Ahn, S. Akiyama,D. Cannon, J. Liu, J.G. Sandland, D. Sparacin, J. Michel,K. Wada, M.R. Watts, Monolithic silicon microphotonics. Top.Appl. Phys. 94, 89–119 (2004)

65. K.K. Lee, D.R. Lim, L.C. Kimerling, J. Shin, F. Cerrina, Fabri-cation of ultralow-loss Si/SiO2 waveguides by roughness reduc-tion. Opt. Lett. 26(23), 1888–1890 (2001)

66. N. Tsumita, J. Melngailis, A.M. Hawryluk, H.I. Smith, Fabri-cation of x-ray masks using anisotropic etching of (110) Si andshadowing techniques. J. Vac. Sci. Technol. B 19, 1211–1213(1981)

67. J. Arentoft, T. Sondergaard, M. Kristensen, A. Boltasseva,M. Thorhauge, L. Frandsen, Low-loss silicon-on-insulator pho-tonic crystal waveguides. Electron. Lett. 38, 274–275 (2002)

68. W.H. Juan, S.W. Pang, Controlling sidewall smoothness for mi-cromachined Si mirrors and lenses. J. Vac. Sci. Technol. B 14,4080–4084 (1996)

69. Q.F. Xia, P.F. Murphy, H. Gao, S.Y. Chou, Ultrafast and selectivereduction of sidewall roughness in silicon waveguides using self-perfection by liquefaction. Nanotechnology 20, 345302 (2009)

70. G. Schmid, Large clusters and colloids-metals in the embryonicstate. Chem. Rev. 92, 1709–1727 (1992)

71. W.X. Zhang, Nanoscale iron particles for environmental remedi-ation: An overview. J. Nanopart. Res. 5, 323–332 (2003)

72. W. Fritzsche, T.A. Taton, Metal nanoparticles as labels for het-erogeneous, chip-based DNA detection. Nanotechnology 14,R63–R73 (2003)

73. S. Sun, C.B. Murray, D. Weller, L. Folks, A. Moser, Monodis-perse FePt nanoparticles and ferromagnetic FePt nanocrystal su-perlattices. Science 287, 1989–1992 (2000)

74. H.-M. So, J. Kim, W.S. Yun, J.W. Park, J.-J. Kim, D.-J. Won,Y. Kang, C. Lee, Molecule-based single electron transistor. Phys-ica E 18, 243–244 (2003)

75. Z.C. Wang, G. Chumanov, WO3 sol–gel modified Ag nanopar-ticle arrays for electrochemical modulation of surface plasmonresonance. Adv. Mater. 15, 1285–1289 (2003)

76. G. Leo, Y. Chushkin, S. Luby, E. Majkova, I. Kostic, M. Ul-meanu, A. Luches, M. Giersig, M. Hilgendorff, Ordering of

http://www.tektronix.com

http://www.tektronix.com

http://www.itrs.net/Links/2006Update/FinalToPost/08_Lithography2006Update.pdf

http://www.itrs.net/Links/2006Update/FinalToPost/08_Lithography2006Update.pdf

http://www.mediacy.com


free-standing co nanoparticles. Mater. Sci. Eng. C, Biomim.Supramol. Syst. 23, 949–952 (2003)

77. S.W. Chen, Langmuir-blodgett fabrication of two-dimensionalrobust crosslinked nanoparticle assemblies. Langmuir 17, 2878–2884 (2001)

78. J. Bosbach, D. Martin, F. Stietz, T. Wenzel, F. Trager, Laser-based method for fabricating monodisperse metallic nanoparti-cles. Appl. Phys. Lett. 74, 2605–2607 (1999)

79. T. Wenzel, J. Bosbach, A. Goldmann, F. Stietz, F. Trager, Shap-ing nanoparticles and their optical spectra with photons. Appl.Phys. B 69, 513–517 (1999)

80. M. Kawasaki, M. Hori, Laser-induced conversion of noble metal-island films to dense monolayers of spherical nanoparticles.J. Phys. Chem. B 107, 6760–6765 (2003)

81. D.Q. Yang, M. Meunier, E. Sacher, Excimer laser manipula-tion and pattering of gold nanoparticles on the SiO2/Si surface.J. Appl. Phys. 95, 5023–5026 (2004)

82. S. Gupta, D. Choudhary, A. Sarma, Study of carbonaceousclusters in irradiated polycarbonate with UV-VIS spectroscopy.J. Polym. Sci. Part B, Polym. Phys. 38, 1589–1594 (2000)

83. S. Hofmann, C. Ducati, J. Robertson, B. Kleinsorge, Low-temperature growth of carbon nanotubes by plasma-enhancedchemical vapor deposition. Appl. Phys. Lett. 83, 135–137 (2003)

84. S. Hofmann, C. Ducati, B. Kleinsorge, J. Robertson, Directgrowth of aligned carbon nanotube field emitter arrays onto plas-tic substrates. Appl. Phys. Lett. 83, 4661–4663 (2003)

85. M.K. Min, J.H. Cho, K.W. Cho, H. Kim, Particle size and alloy-ing effects of Pt-based alloy catalysts for fuel cell applications.Electrochim. Acta 45, 4211–4217 (2000)

86. J. Bischof, D. Scherer, S. Herminghaus, P. Leiderer, Dewettingmodes of thin metallic films: nucleation of holes and spinodaldewetting. Phys. Rev. Lett. 77, 1536–1539 (1996)

87. J. Lee, W. Shimoda, T. Tanaka, Temperature dependence of sur-face tension of liquid Sn–Ag, In–Ag and In–Cu alloys. Meas. Sci.Technol. 16, 438–442 (2005)

88. W.J. Yao, X.J. Han, M. Chen, B. Wei, Z.Y. Guo, Surface tensionof undercooled liquid cobalt. J. Phys., Condens. Matter 14, 7479–7485 (2002)

89. Q.F. Xia, S.Y. Chou, The fabrication of periodic metal nanodotarrays through pulsed laser melting induced fragmentation ofmetal nanogratings. Nanotechnology 20, 285310 (2009)

90. L. Rayleigh, On the instability of jets. Proc. Lond. Math. Soc. 10,4–13 (1878)

91. F.A. Nichols, W.W. Mullins, Surface- (interface-) and volume-diffusion contributions to morphological changes driven bycapilliarity. Trans. Metall. Soc. AIME 233, 1840–1848 (1965)

92. R.H. Austin, J.P. Brody, E.C. Cox, T. Duke, W. Volkmuth,Stretch genes. Phys. Today 50(2), 32–38 (1997)

93. H. Asoh, K. Nishio, M. Nakao, A. Yokoo, T. Tamamura, H. Ma-suda, Fabrication of ideally ordered anodic porous alumina with63 nm hole periodicity using sulfuric acid. J. Vac. Sci. Technol.B 19(2), 569–572 (2001)

94. C.K. Harnett, G.W. Coates, H.G. Craighead, Heat-depolymerizable polycarbonates as electron beam patternablesacrificial layers for nanofluidics. J. Vac. Sci. Technol. B 19(6),2842–2845 (2001)

95. D.M. Cannon, B.R. Flachsbart, M.A. Shannon, J.V. Sweedler,P.W. Bohn, Fabrication of single nanofluidic channels inpoly(methylmethacrylate) films via focused-ion beam millingfor use as molecular gates. Appl. Phys. Lett. 85(7), 1241–1243(2004)

96. H. Cao, Z.N. Yu, J. Wang, J.O. Tegenfeldt, R.H. Austin, E. Chen,W. Wu, S.Y. Chou, Fabrication of 10 nm enclosed nanofluidicchannels. Appl. Phys. Lett. 81(1), 174–176 (2002)

97. W. Reisner, K.J. Morton, R. Riehn, Y.M. Wang, Z.N. Yu, M.Rosen, J.C. Sturm, S.Y. Chou, E. Frey, R.H. Austin, Statics anddynamics of single DNA molecules confined in nanochannels.Phys. Rev. Lett. 94(19) (2005)

98. L.J. Guo, X. Cheng, C.F. Chou, Fabrication of size-controllablenanofluidic channels by nanoimprinting and its application forDNA stretching. Nano Lett. 4(1), 69–73 (2004)

99. H.P. Chou, C. Spence, A. Scherer, S. Quake, A microfabricateddevice for sizing and sorting DNA molecules. Proc. Natl. Acad.Sci. USA 96(1), 11–13 (1999)

100. M. Stjernstrom, J. Roeraade, Method for fabrication of microflu-idic systems in glass. J. Micromech. Microeng. 8(1), 33–38(1998)

101. S.W. Turner, A.M. Perez, A. Lopez, H.G. Craighead, Monolithicnanofluid sieving structures for DNA manipulation. J. Vac. Sci.Technol. B 16(6), 3835–3840 (1998)

102. W.L. Li, J.O. Tegenfeldt, L. Chen, R.H. Austin, S.Y. Chou,P.A. Kohl, J. Krotine, J.C. Sturm, Sacrificial polymers fornanofluidic channels in biological applications. Nanotechnology14(6), 578–583 (2003)

103. Q.F. Xia, K.J. Morton, R.H. Austin, S.Y. Chou, Sub-10 nm self-enclosed self-limited nanofluidic channel arrays. Nano Lett. 8,3830–3833 (2008)

104. D.H. Auston, J.A. Golovchenko, A.L. Simons, C.M. Surko,T.N.C. Venkatesan, Dynamics of Q-switched laser annealing.Appl. Phys. Lett. 34(11), 777–779 (1979)

105. A.D. McLachlan, F.P. Meyer, Temperature-dependence of theextinction coefficient of fused-silica for Co2-laser wavelengths.Appl. Opt. 26(9), 1728–1731 (1987)

applications of excimer laser in nanofabricationchouweb/publications/211 xia... · applications of...

Documents