Raman Spectroscopy through the Indenter Working as an Optical Objective
2019 Volume 60 Issue 8 Pages 1433-1435
Details
2019 Volume 60 Issue 8 Pages 1433-1435
The transparent indenter which can used as an optical objective were tested to obtained a spectra during the indentation. A special device which comprises the transparent indenter and actuator was developed and embedded into the Raman spectrometer. An indentation into the silicon sample was performed and phases that exist under the load and without it were identified.
Fig. 4 Raman spectra obtain during and after the unloading part of the indentation.
Instrumented hardness testers have become a common tools for the investigation of mechanical properties of thin films and heterogeneous materials. These devices use a method that allows to determine local hardness and elastic modulus by means of the analysis of the force vs depth curve.1) In a number of cases such kind of measurements have an artifacts, which among other things, may be caused by phase transformations occurring underneath the indenter.2) These events are often identified by the elbow or pop-out features on the unloading curve. Such an information by itself can hardly be sufficient for the deep analysis of the processes that take place in the material during the indentation and one studies the residual imprints by means of the Raman spectroscopy.3) However, the phases after the unloading can differ from those existing under the peak load. A number of works was done to measure the spectra simultaneously under the loading or unloading.
Article4) describes the scheme in which an indentation is performed at the vicinity of the transparent sample edge, meanwhile the Raman spectra is measured in the perpendicular direction at the different side of the sample near the indentation spot. An analogous approach is given in Ref. 5. In work6) indentation was performed at the top side of the transparent sample, while the spectra were collected at the bottom side, suggested scheme was applicable only for the transparent or thin samples.
The purpose of the current work was investigation of the possibility to use the earlier suggested indenter-objective7) for the in-situ collection of Raman spectra during an indentation. As it was mentioned in work,7) the indenter has the form of the Berkovich pyramid on the both ends, which are rotated at 60 degrees to each other. Such kind of indenter-objective allows to obtain an optical image of the surface outside the contact region directly through the tip. Corresponding measurements were performed without the use of liquids with an conventional optical microscope with an illuminates through the objective.
In order to test the indenter-objective a special load application unit was designed capable of being used with standard laboratory Raman spectrometer. This unit consisted of linear translation stage with a stepper motor, tenso-resistive load cell, sample holder and a housing. Schematic picture of the unit is shown on the Fig. 1.
Schematics of the load application unit.
The load was applied by means of linear translator, which pushes the sample into the indenter-objective, fixed in the flat upper part of the housing. Applied load was monitored using the load cell.
A silicon sample with ⟨100⟩ surface orientation was chosen. As it is known from the literature, this material shows a number of phase transformation during the unloading process.3,8) The sample thickness was about 0.4 mm. An indentation was performed up to the maximum load of 50 N, the spectra were collected through the 20× (NA = 0.4) objective at the load of 3 N, 10 N, 20 N, 30 N, 40 N and 50 N during the loading and unloading process.
As it was described in Ref. 7 an optical image that is directly seen through the indenter–objective consists of three sectors, so that the full-fledged image of the surface can be obtained by the linear translation these sectors. The scheme of an indented surface seen through the indenter is shown on the Fig. 2(a), black areas corresponds to the imprint left on the surface. A coherent picture, which is obtained by the linear transformation of the sectors “1”, “2” and “3”, is shown on the Fig. 2(b). As long as the possibility to obtain the full-fledged pictures was discussed earlier, in this work we’ve concentrated on one of the sectors (e.g. on sector “1” according to scheme on the Fig. 2(a) and Fig. 2(b)). The image of surface seen through this sector under the load of 50 N is shown on the Fig. 2(c). A lot debris correspond to the particles of silicon, which appear to crack even at small loads, at least at the load of 3 N when the first picture of loaded surface was taken.
(a) Scheme of the indented surface observed through the indenter-objective (b) coherent image obtained by linear transformation (c) silicon surface as it is seen through one of the sector, when 50 N is applied.
Raman spectra were measured directly through the diamond indenter simultaneously with loading and unloading. During the loading Si-I peak, which is observed for the undeformed sample at ∼520 cm−1, becomes skewed and moves to the higher frequencies. The skewness of the peak under the small loads can be explained by the small imprint size, so that the light is collected both from the loaded and free material volumes. As the load increases the whole spot is focused in the indented area and one can see a single symmetric peak, shifted up to the about 540 cm−1 when the maximum load of 50 N is achieved. Using the data given in Ref. 9 one can conclude that this 20 cm−1 peak shift corresponds to the pressure about 5 GPa, which is roughly coincides with silicon yield strength.
Raman spectra obtained during the loading part of indentation.
Another peak that is seen on almost all of the loading curve has the Raman frequency shift of 374 cm−1. That frequency according to the work10) can be attributed to the Si-II phase, which exist only in the loaded state. The overall tendency of the data to increase at high frequency values is explained by the fact the spectra is collected through the diamond tip, which has a Raman shift frequency of 1332 cm−1.
As the load is decreased back from 50 N to 0 N new phases appear. The transformation of the spectra is shown on the Fig. 4. Identification of peaks can be using high-pressure cell and residual imprint spectra. As it follows from the review of phase transformations in silica,10) slow decompression at room temperature leads to Si-II → Si-XII transformation, as the relatively low pressures of about 2 GPa Si-XII → Si-III transition occur.11) The peaks at the unloaded “0 N” state can be identified using the residual indent spectrum given in Ref. 12, which is quite similar to that obtained in this work. However, one should notice that the identification according to the work12) has a debatable issue with the Si-XII peak: the existence of the Si-XII phase at the unloaded state and room temperature is thought to be unproved13) and the corresponding peak at roughly 350 cm−1 is attributed to the Si-III phase. In order to underline this feature we’ve mark the peaks with “Si-XII/Si-III” label (Fig. 4). Peak frequencies determined over spectra from all loads are given in the Table 1.
Raman spectra obtain during and after the unloading part of the indentation.
The obtained results show somewhat apparent fact that the indenter-objective can be used for the simultaneous Raman and indentation measurement, which is expectable as long as it was shown that one can observe the surface through this indenter with an ordinary optical microscope. One apparent limitation of the suggested innovation is the scattering of light in the diamond tip, even though the objective was focused on the surface the tip’s diamond Raman shift line (1332 cm−1) is present. On this point silicon is somewhat convenient sample as long as most of its Raman peaks lies far behind the diamond line. We also admit that for the present time quality and resolution of the measurement are limited by the tip manufacturing accuracy and the use of objective magnification higher than 20× is definitely unreasonable as long as the image becomes even more distorted. However, we left the issue of accuracy beyond the scope of the current work and concentrated mainly on the general opportunity to conduct simultaneous indentation and in-situ spectrum measurements directly through the same tip, which was shown to be possible.
An indenter objective that was described elsewhere7) can be used simultaneous indentation and Raman spectroscopy measurement. Such kind of approach allows in-situ investigation of phases both under the applied load or after it’s removal. The approach is not limited to transparent sample or “near edge” measurement; however each spectrum has a peak, that corresponds to the tip’s material, which in this case was diamond. Such kind of feature has a negligible effect for the investigation of material’s Raman peaks significantly separated from the indenter’s one, as it was shown for the case of silicon.
This work was supported by the Ministry of Education and Science of the Russian Federation (project ID RFMEFI57717X0274; the agreement No. 14.577.21.0274); the work was done using the Shared research facilities “Research of Nanostructured, Carbon and Superhard Materials” FSBI TISNCM.
