Micron 67 (2014) 50–64
Contents lists available at ScienceDirect
Micron journal homepage: www.elsevier.com/locate/micron
Raman-in-SEM, a multimodal and multiscale analytical tool: Performance for materials and expertise Guillaume Wille ∗ , Xavier Bourrat, Nicolas Maubec, Abdeltif Lahfid BRGM, 3 Avenue Claude Guillemin – BP 36009, 45060 ORLEANS Cedex 2, France
a r t i c l e
i n f o
Article history: Received 5 February 2014 Received in revised form 20 June 2014 Accepted 20 June 2014 Available online 28 June 2014 Keywords: Raman-in-SEM Contamination Cathodoluminescence Micro-Raman spectroscopy EBSD Mineralogy
a b s t r a c t The availability of Raman spectroscopy in a powerful analytical scanning electron microscope (SEM) allows morphological, elemental, chemical, physical and electronic analysis without moving the sample between instruments. This paper documents the metrological performance of the SEMSCA commercial Raman interface operated in a low vacuum SEM. It provides multiscale and multimodal analyses as Raman/EDS, Raman/cathodoluminescence or Raman/STEM (STEM: scanning transmission electron microscopy) as well as Raman spectroscopy on nanomaterials. Since Raman spectroscopy in a SEM can be influenced by several SEM-related phenomena, this paper firstly presents a comparison of this new tool with a conventional micro-Raman spectrometer. Then, some possible artefacts are documented, which are due to the impact of electron beam-induced contamination or cathodoluminescence contribution to the Raman spectra, especially with geological samples. These effects are easily overcome by changing or adapting the Raman spectrometer and the SEM settings and methodology. The deletion of the adverse effect of cathodoluminescence is solved by using a SEM beam shutter during Raman acquisition. In contrast, this interface provides the ability to record the cathodoluminescence (CL) spectrum of a phase. In a second part, this study highlights the interest and efficiency of the coupling in characterizing micrometric phases at the same point. This multimodal approach is illustrated with various issues encountered in geosciences. © 2014 Elsevier Ltd. All rights reserved.
1. Introduction In materials and environmental sciences, elucidating the structure of materials at the micrometric scale is a crucial step in controlling the nature and properties of such materials. The challenge is for example to identify the use of a resource and/or to understand solid/liquid/gas interaction processes. This requires the use of different and complementary analytical methods to obtain detailed information of a material. Scanning electron microscopy (SEM) combined with conventional techniques such as energydispersive X-ray spectroscopy (EDS), wavelength-dispersive X-ray spectroscopy (WDS) for elemental analysis, electron backscattered diffraction (EBSD) or cathodoluminescence (CL) is one of the tools widely used to characterize materials. It provides numerous information concerning morphology, elemental composition, distribution of phases and crystallographic orientation.
∗ Corresponding author at: BRGM, Laboratories Direction, 3 Avenue Claude Guillemin – BP 36009, 45060 ORLEANS Cedex 2, France. Tel.: +33 02 38 64 35 22; fax: +33 02 38 64 37 11. E-mail address: [email protected] (G. Wille). http://dx.doi.org/10.1016/j.micron.2014.06.008 0968-4328/© 2014 Elsevier Ltd. All rights reserved.
Micro-Raman spectroscopy is another analytical technique widely used in materials sciences. Raman spectroscopy measures the vibrational frequency effect of molecular bonds in the material, on a laser beam (Raman and Krishnan, 1928). The resulting Raman spectrum is not only unique for a given compound, but is also sensitive to the local environment. The spectrum reveals structural data: it provides details on the chemical and structural properties (nature of the functional groups, symmetry groups, lattice defects, etc.) and the crystallinity of mineral compounds. This method offers a variety of analytical possibilities which include, among others, identification of different polymorphs, determination of different oxidation states and crystallographic orientation, evaluation of temperature and stress effects on molecular structure, etc. Therefore, micro-Raman spectroscopy provides complementary chemical and structural information to that obtained from SEM and is often used in parallel with SEM (Frost et al., 2007; Stefaniak et al., 2009). In both techniques, the data can be collected at a similar scale (micrometre), which represents an advantage for the characterization of a small area of a sample or for the analysis of microscopic objects. However, although micro-Raman spectroscopy has proven to be an excellent tool for characterizing samples, it can be limited by its
G. Wille et al. / Micron 67 (2014) 50–64
51
Fig. 1. Raman/SEM facility at BRGM, Orleans. (a) FE-LV-SEM Tescan Mira 3 XMU and Renishaw InVIA Raman microspectrometer; (b) zoom on the SEMSCA interface device used to pilot the optical laser beam inside the SEM (left: chamber door; in the foreground is the dual detector: backscattered electron YAG scintillator/polychromatic cathodoluminescence (350–650 nm); background: LVSTD and EDS detectors).
visualization mode (Maubec et al., 2013). Indeed, since geological samples often show a complexity and heterogeneity in the nature and distribution of phases and in particle size, the use of optical microscopy in association with the Raman spectrometer can make phase or particle recognition difficult or impossible. Thus, transparent particles, chemical zonation (heterogeneous distribution of minor elements in a unique mineral phase) or small areas are not easily seen using light illumination and the location of these kinds of particles or areas of interest is often difficult in micro-Raman spectrometry in contrast to SEM. One of the consequences, when the two tools are used in parallel, is obtaining elemental (X-ray spectroscopy) and molecular data (Raman spectroscopy) at different points, which may give ambiguous results. To overcome this difficulty, Truchet and Delhaye (1986, 1988) were among the first to propose and describe a system that makes it possible to perform Raman analyses directly within an electron microscope. The Raman effect is unaffected by the operating environment, so the technique works equally well under environmental, low, high or ultra-high vacuum in the SEM for example. The technical challenge is to drive a laser light beam in the electron microscope onto the same particle that is analyzed by the various SEM detectors (including EDS/WDS X-ray spectrometers). Such a system has been developed over a decade and consists of a coupling between a Raman spectrometer and a SEM chamber. This innovation has several advantages linked to the observation of the sample with SEM performance: greater resolution and depth of focus than in an optical microscope, imaging modes related to composition and structure (BSE, CL), and realization of chemical (EDS) and structural (Raman) analyses at the same scale, the same area and without moving the sample. This also saves time compared to characterization of a sample with separate devices. To date, the system has successfully been used for the characterization of biological samples (Jarvis et al., 2004; Jarvis and Goodacre, 2004), individual particles (Worobiec et al., 2010; Stefaniak et al., 2014), forensic applications (Otieno-Alego, 2009) and mineralogical samples (Maubec et al., 2012, 2013; Stefaniak et al., 2009). Four instruments are available on the market based on two different principles. Two systems are “on-axis” (i.e. Raman scattering collection is done under the SEM pole piece), from manufacturers Renishaw and Horiba. The Horiba system is based on the use of a cathodoluminescence mirror coupled to a Raman spectrometer, and the Renishaw system use a curved mirror specifically designed for Raman spectroscopy in the SEM. Two stand-alone systems are proposed by the companies Hybriscan and Witec (in association with the SEM manufacturer Tescan). These system are different from the 2 other because Raman spectra are collected “off-axis” (i.e. Raman scattering signal is collected after shifting the stage from the SEM position to the Raman position). Such system is described by Van Apeldoorn et al. (2005) under the name “CRSEM” (confocal Raman SEM).
However, the acquisition of a Raman spectrum in a SEM is subject to the influence of various phenomena related to electron microscopy, both in terms of electron-matter interactions and in terms of artefacts related to the SEM technique. Therefore, this paper reports, firstly, the results of the influence of SEMrelated phenomena on Raman spectra performed in a coupled Raman-in-SEM system in order to propose appropriate experimental conditions, then presents some examples illustrating the interest of using this system. 2. Materials and methods 2.1. Instrumentation The different tests of this work were performed on a field emission gun scanning electron microscope (LV-FE-SEM): MIRA 3 XMU (manufacturer TESCAN, Brno – Czech Republic) under low vacuum conditions (10 < P < 30 Pa). SEM images were collected by both a low vacuum special SE detector (low vacuum Tescan secondary detector (LVSTD) (Jacka et al., 2003) and a backscattered electrons (BSE) detector (doped-YAG scintillator BSE Autrata-type detector – Autrata et al., 1986). Cathodoluminescence (CL) images were obtained using a panchromatic cathodoluminescence detector (350–650 nm) (dual BSE/CL detector from Tescan). Orientation maps were collected using an EDAX PEGASUS EDS/EBSD system with a DIGIVIEW IV camera and OIM DC 6.4 software (manufacturer EDAX, Mahwah – USA). Collections were obtained on non-coated samples at HV = 25 kV, under low vacuum conditions (P = 20 Pa). Samples were polished beforehand using a high quality protocol including a colloidal silica final step: sample is first polished using SiC abrasive paper, then polished using monocrystalline diamond (6, 3, 1 and 0.25 m grain size), and finally polished using colloidal silica suspension (
Contents lists available at ScienceDirect
Micron journal homepage: www.elsevier.com/locate/micron
Raman-in-SEM, a multimodal and multiscale analytical tool: Performance for materials and expertise Guillaume Wille ∗ , Xavier Bourrat, Nicolas Maubec, Abdeltif Lahfid BRGM, 3 Avenue Claude Guillemin – BP 36009, 45060 ORLEANS Cedex 2, France
a r t i c l e
i n f o
Article history: Received 5 February 2014 Received in revised form 20 June 2014 Accepted 20 June 2014 Available online 28 June 2014 Keywords: Raman-in-SEM Contamination Cathodoluminescence Micro-Raman spectroscopy EBSD Mineralogy
a b s t r a c t The availability of Raman spectroscopy in a powerful analytical scanning electron microscope (SEM) allows morphological, elemental, chemical, physical and electronic analysis without moving the sample between instruments. This paper documents the metrological performance of the SEMSCA commercial Raman interface operated in a low vacuum SEM. It provides multiscale and multimodal analyses as Raman/EDS, Raman/cathodoluminescence or Raman/STEM (STEM: scanning transmission electron microscopy) as well as Raman spectroscopy on nanomaterials. Since Raman spectroscopy in a SEM can be influenced by several SEM-related phenomena, this paper firstly presents a comparison of this new tool with a conventional micro-Raman spectrometer. Then, some possible artefacts are documented, which are due to the impact of electron beam-induced contamination or cathodoluminescence contribution to the Raman spectra, especially with geological samples. These effects are easily overcome by changing or adapting the Raman spectrometer and the SEM settings and methodology. The deletion of the adverse effect of cathodoluminescence is solved by using a SEM beam shutter during Raman acquisition. In contrast, this interface provides the ability to record the cathodoluminescence (CL) spectrum of a phase. In a second part, this study highlights the interest and efficiency of the coupling in characterizing micrometric phases at the same point. This multimodal approach is illustrated with various issues encountered in geosciences. © 2014 Elsevier Ltd. All rights reserved.
1. Introduction In materials and environmental sciences, elucidating the structure of materials at the micrometric scale is a crucial step in controlling the nature and properties of such materials. The challenge is for example to identify the use of a resource and/or to understand solid/liquid/gas interaction processes. This requires the use of different and complementary analytical methods to obtain detailed information of a material. Scanning electron microscopy (SEM) combined with conventional techniques such as energydispersive X-ray spectroscopy (EDS), wavelength-dispersive X-ray spectroscopy (WDS) for elemental analysis, electron backscattered diffraction (EBSD) or cathodoluminescence (CL) is one of the tools widely used to characterize materials. It provides numerous information concerning morphology, elemental composition, distribution of phases and crystallographic orientation.
∗ Corresponding author at: BRGM, Laboratories Direction, 3 Avenue Claude Guillemin – BP 36009, 45060 ORLEANS Cedex 2, France. Tel.: +33 02 38 64 35 22; fax: +33 02 38 64 37 11. E-mail address: [email protected] (G. Wille). http://dx.doi.org/10.1016/j.micron.2014.06.008 0968-4328/© 2014 Elsevier Ltd. All rights reserved.
Micro-Raman spectroscopy is another analytical technique widely used in materials sciences. Raman spectroscopy measures the vibrational frequency effect of molecular bonds in the material, on a laser beam (Raman and Krishnan, 1928). The resulting Raman spectrum is not only unique for a given compound, but is also sensitive to the local environment. The spectrum reveals structural data: it provides details on the chemical and structural properties (nature of the functional groups, symmetry groups, lattice defects, etc.) and the crystallinity of mineral compounds. This method offers a variety of analytical possibilities which include, among others, identification of different polymorphs, determination of different oxidation states and crystallographic orientation, evaluation of temperature and stress effects on molecular structure, etc. Therefore, micro-Raman spectroscopy provides complementary chemical and structural information to that obtained from SEM and is often used in parallel with SEM (Frost et al., 2007; Stefaniak et al., 2009). In both techniques, the data can be collected at a similar scale (micrometre), which represents an advantage for the characterization of a small area of a sample or for the analysis of microscopic objects. However, although micro-Raman spectroscopy has proven to be an excellent tool for characterizing samples, it can be limited by its
G. Wille et al. / Micron 67 (2014) 50–64
51
Fig. 1. Raman/SEM facility at BRGM, Orleans. (a) FE-LV-SEM Tescan Mira 3 XMU and Renishaw InVIA Raman microspectrometer; (b) zoom on the SEMSCA interface device used to pilot the optical laser beam inside the SEM (left: chamber door; in the foreground is the dual detector: backscattered electron YAG scintillator/polychromatic cathodoluminescence (350–650 nm); background: LVSTD and EDS detectors).
visualization mode (Maubec et al., 2013). Indeed, since geological samples often show a complexity and heterogeneity in the nature and distribution of phases and in particle size, the use of optical microscopy in association with the Raman spectrometer can make phase or particle recognition difficult or impossible. Thus, transparent particles, chemical zonation (heterogeneous distribution of minor elements in a unique mineral phase) or small areas are not easily seen using light illumination and the location of these kinds of particles or areas of interest is often difficult in micro-Raman spectrometry in contrast to SEM. One of the consequences, when the two tools are used in parallel, is obtaining elemental (X-ray spectroscopy) and molecular data (Raman spectroscopy) at different points, which may give ambiguous results. To overcome this difficulty, Truchet and Delhaye (1986, 1988) were among the first to propose and describe a system that makes it possible to perform Raman analyses directly within an electron microscope. The Raman effect is unaffected by the operating environment, so the technique works equally well under environmental, low, high or ultra-high vacuum in the SEM for example. The technical challenge is to drive a laser light beam in the electron microscope onto the same particle that is analyzed by the various SEM detectors (including EDS/WDS X-ray spectrometers). Such a system has been developed over a decade and consists of a coupling between a Raman spectrometer and a SEM chamber. This innovation has several advantages linked to the observation of the sample with SEM performance: greater resolution and depth of focus than in an optical microscope, imaging modes related to composition and structure (BSE, CL), and realization of chemical (EDS) and structural (Raman) analyses at the same scale, the same area and without moving the sample. This also saves time compared to characterization of a sample with separate devices. To date, the system has successfully been used for the characterization of biological samples (Jarvis et al., 2004; Jarvis and Goodacre, 2004), individual particles (Worobiec et al., 2010; Stefaniak et al., 2014), forensic applications (Otieno-Alego, 2009) and mineralogical samples (Maubec et al., 2012, 2013; Stefaniak et al., 2009). Four instruments are available on the market based on two different principles. Two systems are “on-axis” (i.e. Raman scattering collection is done under the SEM pole piece), from manufacturers Renishaw and Horiba. The Horiba system is based on the use of a cathodoluminescence mirror coupled to a Raman spectrometer, and the Renishaw system use a curved mirror specifically designed for Raman spectroscopy in the SEM. Two stand-alone systems are proposed by the companies Hybriscan and Witec (in association with the SEM manufacturer Tescan). These system are different from the 2 other because Raman spectra are collected “off-axis” (i.e. Raman scattering signal is collected after shifting the stage from the SEM position to the Raman position). Such system is described by Van Apeldoorn et al. (2005) under the name “CRSEM” (confocal Raman SEM).
However, the acquisition of a Raman spectrum in a SEM is subject to the influence of various phenomena related to electron microscopy, both in terms of electron-matter interactions and in terms of artefacts related to the SEM technique. Therefore, this paper reports, firstly, the results of the influence of SEMrelated phenomena on Raman spectra performed in a coupled Raman-in-SEM system in order to propose appropriate experimental conditions, then presents some examples illustrating the interest of using this system. 2. Materials and methods 2.1. Instrumentation The different tests of this work were performed on a field emission gun scanning electron microscope (LV-FE-SEM): MIRA 3 XMU (manufacturer TESCAN, Brno – Czech Republic) under low vacuum conditions (10 < P < 30 Pa). SEM images were collected by both a low vacuum special SE detector (low vacuum Tescan secondary detector (LVSTD) (Jacka et al., 2003) and a backscattered electrons (BSE) detector (doped-YAG scintillator BSE Autrata-type detector – Autrata et al., 1986). Cathodoluminescence (CL) images were obtained using a panchromatic cathodoluminescence detector (350–650 nm) (dual BSE/CL detector from Tescan). Orientation maps were collected using an EDAX PEGASUS EDS/EBSD system with a DIGIVIEW IV camera and OIM DC 6.4 software (manufacturer EDAX, Mahwah – USA). Collections were obtained on non-coated samples at HV = 25 kV, under low vacuum conditions (P = 20 Pa). Samples were polished beforehand using a high quality protocol including a colloidal silica final step: sample is first polished using SiC abrasive paper, then polished using monocrystalline diamond (6, 3, 1 and 0.25 m grain size), and finally polished using colloidal silica suspension (
