WIT

ec G

mb

H, L

ise-

Mei

tner

-Str

. 6, 8

90

81

Ulm

, Ger

man

y

ph

one

+49

(0) 7

31 14

0 7

00

, fax

+49

(0) 7

31 14

0 7

0 2

00

info

@W

ITec

.de,

ww

w.W

ITec

.de

APPLICATION NOTE

Correlative Raman Imaging: Geoscience Applications

Confocal Raman microscopy is a powerful method for high-resolution, non-destructive and label-free analysis of geological samples. It can identify the chemical components and visualize their distribution in 2D or 3D.

Sensitivity

A high confocality increases the signal-to-noise ratio by reducing the background. With the UHTS Series, WITec developed lens-based, wavelength-optimized spectrometers with a spectral resolution down to 0.1 cm-1 relative wavenumbers.

Resolution

Lateral resolution is physically limited to ~200 nm, depending on the wavelength of the incident light.

Speed

The more sensitive a system is, the shorter the acquisition time for a single spectrum. WITec’s Ultrafast Raman Imaging reduces acquisition times for single Raman spectra down to well below 1 ms.

The Raman principle

The Raman effect is based on the inelastic scattering of light by the molecules of gaseous, liquid or solid materials. The interaction of a molecule with photons causes vibrations of its chemical bonds, leading to specific energy shifts in the scattered light. Thus, any given chemical compound produces a particular Raman spectrum when excited and can be easily identified by this individual “fingerprint.”Raman spectroscopy is a well-established, label-free and non-destructive method for analyzing the molecular composition of a sample.

Raman imaging

In Raman imaging, a confocal microscope is combined with a spectrometer and a Raman spectrum is recorded at every image pixel. The resulting Raman image visualizes the distribution of the sample’s compounds. Due to the high confocality of WITec Raman systems, volume scans and 3D images can also be generated.

No need for compromises

The Raman effect is extremely weak, so every Raman photon is important for imaging. Therefore WITec Raman imaging systems combine an exceptionally sensitive confocal microscope with an ultra-high throughput spectrometer (UHTS). Precise adjustment of all optical and mechanical elements guarantees the highest resolution, outstanding speed and extraordinary sensitivity - simultaneously!

This optimization allows the detection of Raman signals of even weak Raman scatterers and extremely low material concentrations or volumes with the lowest excitation energy levels. This is an unrivaled advantage of WITec systems.

Calcite

Organics

Fluorite

Quartz

Plagioclase

Titanium oxide

Unidentified phase

Figure 1: Large area EMP and Raman analysis of a diorite sample.

(A) Overview of the polished rock section. The orange rectangle indicates the scan position. (B - D) EMP maps showing elemental distribution for Al (B), Si (C) and Ca (D). (E) Corresponding Ra-man image. (F) Combined false-color Raman image of the diorite sample in (A). The colors correspond to the spectra in (B). (G) Corresponding Raman spectra. Scale bars 1 mm.

F

Raman spectroscopy has long been applied in geoscience, for example for the identification and characterization of minerals, or in the observation of mineral phase transitions in high and ultra-high pressure/temperature experiments. In most cases, measurements have been carried out with a micro-Raman setup, i.e. information was obtained from single or multiple points of interest on a sample. This way, few details of the spatial distri-bution of components, mineral phases or chemical variations could be observed, even though this information may contri-bute significantly to the understanding of a sample’s complexity.

By means of confocal Raman imaging, such characteristics can be evaluated from large-area scans on the centimeter scale to the detailed investigations with sub-micron resolution. WITec confocal Raman microscopes of the alpha300 series allow for such measurements with very high sensitivity and resolution. Raman imaging is a tool that provides information comple-mentary to data obtained by e.g. electron microprobe (EMP), energy dispersive X-ray analysis (EDX) or secondary ion mass spec-trometry (SIMS). These techniques acquire quantitative, semi-quantitative elemental and isotopic data. In addition, confocal Raman imaging contributes the distributi-on of chemical components over a defined sample area. Furthermore, considering that most geo-materials are transparent from the NUV to VIS and NIR to some degree, this information can be obtained three dimensionally due to the confocal setup of the microscopes. The following examples highlight some key analytical features of this technique for geoscience applications.

A sample from the deep biosphere of Äspö, Sweden, was studied using the lar-ge-area scan mode of a confocal Raman microscope (WITec GmbH), aiming to characterize secondary cleavage fillings in a 1.8 to 1.4 billion-years-old diorite.

A section (Fig. 1A) was obtained from a drill core from borehole KJ 0052 F01, run by SKB (Swedish nuclear fuel and waste management) and drilled from a tunnel at ~450 m below the surface with a sampling depth in the core of 12 m. A large-area scan of 8 x 2 mm2 was performed on this polished section with 800 x 200 pixels and an integration time of 36 ms per spectrum using a 532 nm excitation wavelength.

G

A

B

C

D

E

Al

Si

Ca

Large area Raman analysis: overview and detailsConfocal Raman imaging

Figure 2: Variations in different phases of the diorite sample.

(A, B) The plagioclase phase: Two individual spectra (A) show distinct distribution within the plagio-clase phase (B). (C) Regions containing rutile (grey rectangle) and quartz (red rectangle) are marked in the combined Raman image (see Fig. 2) and further analyzed (D, E). (D) The quartz phase: Distinct regions are reflected in the variation of relative peak intensities in the spectra. Note edges around plagioclases. (E) The rutile/Ti-phases contain two different materials in two distinguishable configu-rations. Scale bars 1 mm.

Sample courtesy of C. Heim and V. Thiel, EMP data courtesy of A. Kronz, Geobiology Group, University of Göttingen, Germany.

E rutile phase

Intensities integrated from 140 to 300 cm-1 (Fig. 1E) are compared to EMP maps of Al, Si and Ca (Fig. 1B-D), and indicate a similar overall chemical contrast assignment of mineral phases and their gross distributi-on over the scanned area. The color-coded Raman image (Fig.1E) and corresponding spectra (Fig. 1F) depict the general assig-nment of mineral phases and their gross distribution over the scanned area.

In addition to the mineralogical context information, organic components were identified, spectrally characterized and located, trapped between two generations of hydrothermal precipitates (fluorite and calcite, Fig.1E), indicating at which point in time a “deep biosphere” was active within these rocks.

Cluster analysis of the data set using the WITec Project software package revealed discrete areas of variation in the mineral phases (Fig. 2). This is illustrated by the plagioclase phase, in which two distinct spectra show different distributions within the sample regions identified as plagio-clase (Fig. 2 A, B). Four distinct regions were identified for the quartz phase based on variations in relative peak intensity (Fig. 2D). Note the discrete edges around plagioclases. The rutile phase is represen-ted by the blue and yellow spectra and corresponding regions and an unidenti-fied titanium-containing phase and its variation is represented by the red and green spectra and corresponding regions (Fig. 2E).

D quartz phaseA plagioclase phase

CB

Figure 3: Carbonate composition in ALH84001 and BVC xenoliths.

(A) Carbonate distribution integrated over the ~1090 cm-1 peak. Scale bar 20 µm. (B) Center dis-tribution of the carbonate signal for the images in (A). (C) Peak center shifts of the carbonate ~330 cm-1 peak (light color = higher wavenum-ber). ALH84001 in red, BVC in blue. The letters a, b, c, d and numbers 1, 2 represent distinct chemical zones in the carbonates. (D) Spectra corresponding to the images in (C).

Images are a courtesy of Meteoritics and Plane-tary Sciences; from Steele et al., 2007.

a

b

c

d1

2

BVCALH84001

High spectral and spatial resolution is necessary to observe chemical variations when they occur in micron-sized featu-res. The spectral resolution is required to distinguish minor compositional chan-ges, and the spatial resolution allows the depiction of such changes at the required (sub-)micron level. This is illustrated by small pancake-shaped carbonate structu-res, so-called globules, from the martian meteorite ALH84001 and a terrestrial analogue from Svalbard (mantle peridotite xenolith from the Bockfjorden volcanic complex, BVC) (see Steele et al., 2007 for further description). Fig. 3 shows the results of confocal Raman iaging studies

undertaken with a high spectral resolu-tion setup of the carbonate composition in the ALH84001 and the BVC xenoliths. The two images in Figure 4A show the overall carbonate distribution with relative intensity integrated over the peaks around 1090 cm-1 for both samples. The spectra showing the peak positions are depicted in Figure 3B. Various carbonate reference samples were measured to demonstrate chemical variation, including calcite (Ca), siderite (Si) and magnesite (Ma) (Fig. 3D). The images in Fig. 3C represent shifts of the carbonate ~330 cm-1 peak and clearly delineate discrete zones of chemical vari-ation separated from one another on the

(sub-)micron scale (the lighter colors show a peak shift toward higher wavenumbers). The letters a, b, c, d represent distinct chemical zones in the carbonate globule of the martian meteorite, the numbers 1 and 2 show zoning in the BVC carbonates. The conclusions drawn from these observa-tions contribute further to understanding the processing of volatile and biologically relevant compounds (CO2 and H2O) on Mars and illustrate the importance of high spectral and spatial resolution in confocal Raman microscopy studies (for a detailed discussion see Steele et al., 2007).

BVCALH84001

A

C

B

D

Raman imaging with high resolution spectral and spatial information

Figure 4: Image collage of an IDP.

(A) Video image of the IDP. (B) Combined false-color Raman image of (C - E). Colors as indicated in the spectra in (F). (C) Relative intensity distribution re-sulting from integration over the two peaks around 200 - 300 cm-1 (blue spectrum in F). (D) Relative intensity map of combined hematite / carbon D band phases (1280 - 1400 cm-1, red spectrum in F). (E) Relative intensity map of hematite phases (green spectrum in F). (F) Corresponding Raman spectra.

Valuable compositional and structural information may be contained even in the smallest sample volumes. In order to uncover such information it is not only a requirement to be able to spectrally and spatially resolve chemical features at the highest resolution, it is furthermore essential to work at the highest possible sensitivity. Confocal Raman microscopes of the WITec alpha series provide this level of sensitivity due to the highest possible th-roughput of Raman-shifted photons from the sample surface to the detector. This allows for the analysis of very small and delicate samples without the risk of dama-ging them. In geosciences such materials can, for example, be aerosols or interstellar dust particles (IDPs).

Fig. 5 shows a collage of confocal Raman images of an IDP (for further detail see Toporski et al., 2004). The video image gives an overview of the particle (Fig. 4A). The combined false-color Raman image (Fig. 5B) shows that distinct regions for the components shown in Figures 4C to 4E can be separated from one another. Further unidentified mineral phases (Fig. 4C) are separated from hematite/car-bon-bearing phases (Fig. 4D) and hematite phases (Fig. 4E) and could be associated with distinct regions on the sample. The corresponding Raman spectra are shown in Figure 4F.

To allow for such differentiation and to enable the separation of labile compo-nents and at the same time avoid sample alteration due to excessive excitation laser energy, high sensitivity is a prerequisite.

This has also been demonstrated on particles returned from comet 81P/Wild2 by the NASA Stardust spacecraft (Rotundi et al., 2008). The study of small, delicate and unstable samples further benefits from Raman imaging being non-destruc-tive, so that samples can subsequently be investigated using other more invasive micro-analytical techniques (e.g. SIMS, ToFSIMS, TEM, etc.).

Raman imaging with high sensitivity: revealing hidden details

(D)

(E)

FA

B

C

D

E

(C)

surface (Fig. 5A). As can be seen in the Raman image (Fig. 5B), the garnet shows significant variation in relative peak inten-sities across the scanned area. The corre-sponding spectra (Fig. 5C) (magenta, red and orange) are normalized to the peak near 900 cm-1. Differences are apparent between the red and the magenta spectra (peaks near 850 and 360 cm-1). The orange spectrum possibly represents a Carbonate phase in direct proximity to, or as part of, the inclusion. This variation is confirmed in the depth scan (Fig.5E). Here too, spectra (Fig. 5E) are normalized to the peak near 900 cm-1 and differences are apparent between the red and the magenta spectra (peaks near 850 cm-1 and near 360 cm-1; Fig. 5F). Interestingly, the aqueous phase sho-wed significant variation, which can clearly be seen from the spectra shown.

and processing (WITec Software Suite), the data from the image stack is used to gene-rate the 3D image, such as the example on the right. Applications from various fields will be illustrated in the following.

Fluid inclusions were discovered in ultra -high pressure (diamond-grade) garnets from the Kokchetav Massif in northern Kazakhstan (samples and data courtesy of Andrey Korsakov, Siberian Branch of the Russian Academy of Science, Novosibirsk, Russia). Characterization of the inclusi-on contents regarding composition and petrographic context in this case may pro-vide information on micro-thermometry. Due to confocality, Raman imaging allows for the non-destructive and non-invasive analysis of such features. The garnet was polished so that the inclusion remained intact some tens of microns below the

Figure 5: Fluid inclusion in garnet.

(A) Video image showing the sub-surface inclusion. The black line indicates the plane of the X-Z scan in (C). (B) Combined false-color Raman image. (C, D) Raman spectra of the garnet and carbonate pha-se (red, orange, pink) and of the aqueous phase (green, cyan, blue). (E) Combined false-color Raman image for the X-Z scan along the black line in (A) and corresponding spectra (F, G). Scale bars 10 µm.

2D Raman microscopy visualizes the distribution of chemical compounds, e.g. on the sample’s surface or in a focal plane within the sample (x-y-plane), while 3D Raman imaging enables more complex chemical analysis of the components‘ distribution in three dimensions (x, y and z). This is particularly advantageous for the investigation of comprehensive or bulky samples, complex emulsions or mixtures, geological specimens and components in living organisms.

In order to generate 3D images, confocal 2D Raman images from different focal planes are acquired by automatically scanning through the sample along the z-axis. After data acquisition, evaluation

3D Raman imaging: unveil the information trapped inside

A

B

E

C

GF

D

in Figs. 5D and G, both for the X-Y and the X-Z scans. Variations may be attributed to the presence of H2O in liquid and gaseous phases, or the presence of hydrous silicate phases. Changes in garnet peak intensities appear only around the inclusion. This may either indicate a differential stress regime attributable to the influence of the fluid inclusion, or may be due to slight varia-tions in the garnet composition adjacent to the inclusion.In the previous example, 3D information from a garnet sample was gained by recording one Raman image in the X-Y plane and one orthogonal to it

in the X-Z plane. However, an even more complete piture of the material distributi-on can be obtained from a 3D representa-tion of the sample. This is possible using 3D Raman imaging, where 2D images are recorded from several focal planes. A 3D representation of the investigated sample volume is then created from these image stacks.

Fig. 6 exemplarily shows the 3D Raman image of a fluid inclusion in garnet. It reveals the fluid water inclusion (blue) surrounded by garnet (red), calcite (green) and mica (cyan).

Figure 6: 3D confocal Raman image of a fluid inclusion in garnet. Water inclusion (blue) in garnet (red), calcite (green), mica (cyan). Scan range: 60 x 60 x 30 µm3.

Topographic confocal Raman imaging with TrueSurface

The TrueSurfaceTM is an optical profilo- metry technology pioneered by WITec. It allows confocal Raman microscopy to be applied to roughly textured and tilted samples by keeping their surfaces in con- stant focus. Thus, it is extremely useful for geological samples.

Fig. 7 shows topographic Raman imaging on an inclined and rough rock sample from the Doushantuo formation (China). The height profile (C) was recorded simulta-neously with the Raman spectra at each image point and illustrates the tilted sample plane. The confocal 3D Raman image (D) shows the chemical sample information overlaid on the height profile.

Figure 7: Raman imaging of a rough rock

(A) A cap carbonate sample from the Doushan-tuo formation (Yichang, China) was investiga-ted. From the enframed area in the white light image (B) was the height profile taken (C) and a complete Raman image generated (D). The colours correspond to those in the spectrum (E) with red = ankerite, blue = barite, green = dolo-mite, pink = Al2[SO4]3 and yellow = unidentified phase.

A

ED

C

B

Sample courtesy of Joachim Reitner, Dept. of Geobiology, Göttingen University, Germany.

Figure 8: AFM images of silici-fied bacteria.(A) Topography AFM image. Cell structures can not be recogni-zed from the topography. (B) The Digital Pulsed Force Mode of the Atomic Force Microsco-pe image shows the different adhesion levels and reveals the bacterias‘ cell structures. Scale bars 1 µm.

Atomic force microscopy is a non-destruc-tive imaging technique for surface charac-terization on the nanometer scale. Besides high-resolution topography images, local material properties such as adhesion or stiffness can be investigated by analyzing

the tip-sample interaction forces. Figure 11 shows AFM images of silicified bacteria. Cellular structures can not be distingu-ished in the topography image (Fig. 8A), but are revealed by the image of different adhesion levels (Fig. 8B).

Atomic morce microscopy of fossils

AFM is combinable with confocal Raman imaging in one instrument. Thus, AFM and Raman images can be obtained from the same sample area, making it possible to correlate the structural and chemical information.

Due to its non-destructive and label-free characteristics confocal Raman imaging is ideally suited for correlative imaging techniques, which greatly enhance the insight provided by an investigation.

Correlative imaging techniques gain more and more importance in many fields of applications due to the fact that more characteristics of a specimen can be analyzed with one instrument. The efficient workflow of correla-tive microscopy techniques is an additional advantage that saves time and money.

Raman imaging can be combined with either various methods of light microscopy, with atomic force micros-copy (AFM) and scanning electron microscopy (SEM).

WITec has developed microscopes that integrate confocal Raman imaging and optionally AFM, scanning near-field optical microscopy (SNOM) or SEM into one instrument for highly sensitive and efficient correlative microscopy.

A

The power of correlative Raman imaging: maximize gain of information

B

same sample area as the SEM image. It clearly emphasizes that high resolution optical images require proper focusing, whereas SEM is completely insensitive to surface roughness, which is why it is difficult to retrieve the sample area when using two stand-alone instruments.

RISE (Raman Imaging and Scanning Electron) Microscopy is the combination of confocal Raman Imaging and Scanning Electron Microscopy (SEM). It incorporates the sensitivity of the non-destructive, spectroscopic Raman technique along with the atomic resolution of electron microscopy in one instrument. Due to an intelligent positioning system RISE micros-copy enables diffraction limited confocal Raman imaging from exactly the same sample area as the SEM image.

Mineral phases of a rock section from a drill core were analyzed in the next example. The dominant rock type is diorite. The sample had no need to be treated or vaporized but was only sectioned with a diamond saw under water. This reduces the risk of contamination of the surface, which is of great importance not only for geological drill cores but also for objects derived from the deep sea.

Figure 10A shows the SEM image of a small sample area using the back-scatte-red electron detector of the SEM. The same area was analyzed using EDX (Fig. 9B). The distribution of elemental groups indicates the presence of three distinct minerals. Single Raman spectra could be acquired from the three different areas (Fig. 9C) three of which show the characteristic Raman bands for quartz, epidote and pla-gioclase. This is in good agreement with the elemental composition obtained by EDX. Raman spectra of the three primary minerals were evaluated using cluster analysis. In addition to these spectra, four others were detected, two of which could indicate different grain orientations within the phases of epidote and plagioclase respectively. The spatial distribution of the minerals is shown in the color-coded RISE image (Fig. 9D) where the colors of the Raman image match the colors of the spectra. Not only mineral phase distributi-on could be detected but also small grains within a mineral phase. Figure 9E shows the white light image acquired from the

Figure 9: A mineral phase of a diorite rock section.

(A) SEM image. (B) Overlay of SEM and EDX images. Three element groups could be distinguished with EDX: Si, O (orange); Ca, Fe, Al (grey-purple); Na (green). (C) Raman sing-le spectra acquired from the three distinct regions with the characteristic Raman bands of quartz (brown), epidot (red) and plagioclase (green). (D) Raman spectral image overlaid with the SEM image. (E) Light-microscopy image from the same sample area. Image parameters: 100 x 100 µm2, 150 x 150 pixels = 22,500 spectra, integ-ration time 80 ms per spectrum.

Sample courtesy by C. Heim, Geoscience Centre GZG, Dept. Geology, University of Göttingen, Germany.

RISE imaging of a geological sample

A

C

E

B

D

Rotundi A. et al. (2008) Combined micro-Ra-man, micro-Infrared, and field emission scanning electron microscopy analysis of comet 81P/Wild2 particles collected by Stardust. Meteoritics and Planetary Science, 43, 367-397.

Steele A., Fries M., Amundsen H., Mysen B., Fogel M., Schweizer M and Boctor N. (2007) Com-prehensive imaging and Raman spectroscopy of carbonate globules from Martian meteo-rite ALH84001 and a terrestrial analog from Svalbard. Meteoritics and Planetary Science, 42, 1549-1566.

Toporski J., Fries M., Steele A., Nittler L. and Kress M. (2004) Sweeping the skies: Stardust and the origin of our Solar System imaging and Micros-copy, 04/2004, 38-40.

Figure 10: Raman-SEM image of hematite.

The sample consists of several crystal forms of hematite (red, blue, green, orange, pink) and of goethite (light blue, cyan). The Raman image was overlaid onto the SEM image.

Raman image parameters: 50 x 50 µm2; 150 x 150 spectra, integration time: 150 ms/spectrum, 10 mW laser power.

A piece of hematite (Fe2O

3) was analyzed

first with the SEM, then with the integra-ted confocal Raman microscopy function of the RISE microscope. SEM depicts structural characteristics, but cannot differentiate between the oxides present in the hematite. However, the Raman spectra indicate the occurrence of several crystal forms of hematite and of goethite (FeO(OH)). The distribution of these mine-rals is shown in the Raman image that has been overlaid onto the SEM image (Fig. 10).

RISE imaging of hematite

References

WITec HeadquartersWITec GmbHLise-Meitner-Str. 6D-89081 Ulm . GermanyPhone +49 (0) 731 140700Fax +49 (0) 731 14070200info@witec.dewww.witec.de

WITec North AmericaWITec Instruments Corp.130G Market Place Blvd.Knoxville . TN 37922 . USAPhone 865 984 4445Fax 865 984 4441info@witec-instruments.comwww.witec-instruments.com

WITec JapanWITec K.K.Mita 2-3227, Chome, Tama-ku,Kawasaki-shi, Kanagawa-ken 214-0034JapanPhone +81 44 819 7773info@witec-instruments.bizwww.witec.de/jp

WITec South East AsiaWITec Pte. Ltd.25 International Business Park#03-59A German CentreSingapore 609916Phone +65 9026 5667shawn.lee@witec.biz

WITec ChinaWITec Beijing Representative Offi ceUnit 507, Landmark Tower 18 North Dongsanhuan RoadBeijing, PRC., 100004Phone +86 (0) 10 6590 0577Info.China@witec-instruments.comwww.witec.de/cn

alpha300 Raman-AFM-SNOM

alpha300 access

apyron automated Raman-AFM

alpha300 Raman-AFM

WITec Microscopes

alpha300 R

WITec HeadquartersWITec GmbHLise-Meitner-Str. 6D-89081 Ulm . GermanyPhone +49 (0) 731 140700Fax +49 (0) 731 14070200info@WITec.dewww.WITec.de

WITec North AmericaWITec Instruments Corp.130G Market Place Blvd.Knoxville . TN 37922 . USAPhone 865 984 4445Fax 865 984 4441info@WITec-Instruments.comwww.WITec-Instruments.com

WITec South East AsiaWITec Pte. Ltd.25 International Business Park#03-59A German CentreSingapore 609916Phone +65 9026 5667shawn.lee@WITec.bizwww.WITec.de

WITec ChinaWITec Beijing Representative OfficeUnit 507, Landmark Tower 18 North Dongsanhuan RoadBeijing, PRC., 100004Phone +86 (0) 10 6590 0577info.China@WITec-Instruments.comwww.WITec.de/cn

WITec JapanWITec K. K.1-1-5 Furo-cho, Naka-ku,Yokohama City, Kanagawa Pref. 231-0032 JapanPhone +81(0)45 319 4277info@WITec.jpwww.WITec.de/jp