μ-Raman spectroscopy for stress analysis in high power silicon devices

Verification of Si wafer first‐order phonon peaks for reliable calibration of Raman microscopes

“…The slope (À0.0206 ± 0.0001 cm À1 / C) was comparable with that reported in the literature 6,8,9,[15][16][17][18][19][20] ; the slightly smaller value in the present study might be due to the relatively lower temperature range examined. 6,17 Based on Equation (1), the wavenumber at 23 C was calculated as 520.63 cm À1 ± 0.00 cm À1 and is consistent with the value of NMIJ CRM 5606-a (520.45 cm À1 ± 0.28 cm À1 ) reported previously. 5 [Correction added on 12 March 2024, after first online publication: The Equation 1 has been updated.…”

“…4,5 On the other hand, it is also known that the position of the first-order phonon peak of the Si wafer shifts depending on the material conditions such as the doping type, dopant concentration, temperature and stress. [6][7][8][9][10][11] Even when using a Si wafer with a low dopant concentration, the first-order phonon peak shifts to a lower position with a higher temperature and tensile stress, whereas it shifts to a higher position with a lower temperature and compressive stress. It is also known that the main factor of the peak shift can be identified from the changes in the peak width, which only changes with the temperature.…”

“…It is also known that the main factor of the peak shift can be identified from the changes in the peak width, which only changes with the temperature. 6,8 Furthermore, a higher laser power introduces a lower shift of the peak position due to an increase in the temperature produced by the laser power. 7,[12][13][14] The Si wafers to use for the calibration of Raman microscopes are generally provided by the manufacturer along with the calibration procedures; however, it is difficult to specify the calibration temperature.…”

“…Although there are many reports on linear relationships between the position of the first-order phonon peak of Si wafers and the temperature at a slope of about À0.02 cm À1 /K or cm À1 / C, not only are there differences in the values of the slope among the literature, but the wavenumber at 0 C or 0 K, which is important in order to know the wavenumber at a certain temperature, is unclear in most literature. 6,8,9,[15][16][17][18][19][20] Furthermore, there is no information about the comparability of Si wafers for use as a calibrant although Si wafers are available from different manufacturers and are usually well polished and have different crystal orientations. To use the first-order phonon peak of Si wafers as a reliable calibrant and improve the data reliability over Raman microscopes, it is important to clarify not only the universality and comparability of the wavenumbers obtained from different manufacturers and crystal orientations at a certain temperature but also the influence of the laser power on the shift of the wavenumber.…”

Determination of stress in silicon wafers using Raman spectroscopy

“…Raman spectroscopy has proven its power for the analysis of stress states in semi-conductor materials in the past. 5,6 Unstrained silicon has three degenerated Raman modes with the same frequency at 520.5cm −1 . 8 These three Raman modes, two transverse optical modes and one longitudinal mode, are polarization dependent.…”