Open Access
Add to BibSonomy
Abstract
The InAs1−xSbx ternary alloy band gap nonlinearly depends on the composition, which provides the opportunity for use of this material in devices operating in a wide range of infrared radiation. We present experimental results for InAs1−xSbx samples for Sb composition from 0.1 to 0.8. The most common way to determine it is by using a high resolution X-ray diffractometer. In a previous article, we showed that energies of folded longitudinal acoustic and folded transverse acoustic Raman peaks are linearly correlated with mole fraction. In this work, we will illustrate how to determine mole fraction using peak energy and calculate the bowing parameter for InAs1−xSbx at 300 K.
© 2019 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Devices for the detection of infrared radiation (motion detectors, smoke and gas detectors) have more and more applications in today [1,2]. Systems based on HgCdTe are well known and widely used. The AIIIBV group technology is more often alternative to HgCdTe-based devices. Advantage of covalent bonding contribution comparing to ionic bonding in HgCdTe makes InAs1−xSbx more stable which the epitaxial layers gives greater mechanical and chemical stability [1,2].The InAs1−xSbx ternary alloy band gap depends on the composition, which gives the opportunity to use this material in devices operating in a wide range of infrared radiation [3].
High Resolution X-Ray Diffraction (HRXRD) is the most commonly used technique to determine mole fraction [4,5]. However this technique is not always optimal, what is discussed in this article in experimental part. Another way to determine mole fraction may be Raman spectroscopy. One can obtain mole fraction using Longitudinal Optical (LO) intensities [6], but only for mole fraction lower than 0.5.
The other idea is to use Raman low energy bands commonly called Disorder Activated Longitudinal Acoustic (DALA) and Disorder Activated Transverse Acoustic DATA [7–11]. In literature in low energies region except DALA there were also Folded Longitudinal Acoustic (FLA) peaks observed as a consequence of long distance ordering [12,13]. In previous article we shown DALA and DATA bands in InAs1−x Sbx may be interpreted as FLA and Folded Transverse Acoustic (FTA) peaks [8,10,11]. We shown the energy of both FLA and FTA peaks decreases with increasing Sb mole fraction [11]. Here we present how to use FLA and FTA energies to determine Sb mole fraction. After determination mole fraction using FLA we may calculate bowing parameter for InAs1−xSbx in 300 K [14].
2. Materials and methods
In this work we present experimental results of InAs1−xSbx samples (Table 1) bulk material grown in Molecular Beam Epitaxy (MBE) VIGO/MUT laboratory on 2 inch (001) GaAs substrate. The sample was obtained using a RIBER Compact 21-DZ solid-source molecular beam epitaxy (MBE) system [15]. Crystallographic properties were measured by High-resolution HRXRD diffractometer of PANalytical X'Pert MRD3. For each sample (Sample no) Sb mole fraction obtained using HRXRD (XSb HRXRD) as well as using FLA (XSb FLA) and FTA (XSb FTA) are presented. We also added energies of FLA (FLA) for each sample, FTA (FTA) peaks for most of samples and bandgap energy obtained using PL (PL). Experiments were performed using Bruker Vertex 70v FT-IR spectrometer, MCT photodetector and lock-in. The whole system was working in Step Scan mode [16]. As a pump beam we used 637 nm line laser chopped mechanically with frequency of 1000 Hz. All samples presented in the article were measured in room temperature. Raman spectra were acquired using Renishaw InVia Raman Microscope equipped with Eclipse filter, x100 objective and 532 nm laser. We used about 50 μW laser power and 30 seconds measurement time to avoid sample heating.
Table 1. The InAs1−xSbx samples parameters.
3. Results and discussion
In Fig. 1 typical Raman spectra for InAs0.05Sb0.96 and InAs are presented. We assigned Longitudinal Acoustic (LO), as well as 2*Transverse Acoustic (2TA) peaks for InAs. For InAs0.05Sb0.96 we marked InSb LO, InAs LO and second order 2LO, 2TO and TO + LO peaks. There are also TO peaks in the spectrum but it is hard to assign them not losing readability of figure. Two low energy peak usually called DALA and DATA are assigned as FLA and FTA as we explained in previous article [11].
We are not sure what the phonon branches generating FLA and FTA peaks are. FLA (Fig. 2 a) and FTA (Fig. 2.b) energies for all measures samples are marked as black squares. In literature for Cu-Pt zone folding was observed and L-point becomes Γ point [17]. We assigned theoretical value for LA and TA branches at L point for both InAs and InSb as blue squares [18]. Red line represents linear fit of experimental points while blue line has the same slope as red but is shifted in y-scale by 18.5 cm−1. Obtained blue line almost perfectly connects blue theoretical points. Similar operation we performed for FLA experimental points (Fig. 2.b). In this case blue line in shifted by 7 cm−1 with respect to red line and also connects theoretical blue points.
Our experimental points for FLA are close to obtained in literature assign as DALA [19]. It is hard to interpret why theoretical blue line is shifted with respect to experimental red line, however in other article [13] average LA phonon energy in L point also does not fit to theoretical value.
We decided to use red lines as theoretical curves to calculate mole fraction using FLA and FTA peaks energy. Transforming equations obtained from linear fit (Fig. 2 a and Fig. 2 b) we may calculate mole fraction as:
and using FTA Assuming last steps: The question is why mole fraction obtained using Eqs. (1) and (2) may work better than obtained using HRXRD, which was used to obtain Eqs. (1) and (2). Choosing points with mole fraction below 0.2 (Fig. 2) we can observe that red line behaves like average for group of points. We may not obtain this average if we would have only mole fraction from HRXRD. Average from 12 numbers is one number. However if we have some additional information about points (FLA and FTA energies) we may use additional axis to perform average and that makes Eqs. (1) and (2) possibly better.Having dependency between these values we can calculate mole fraction for all sample using FLA and FTA energies and plot band gap energy vs mole fraction using all three methods (Fig. 3). To verify if obtained mole fraction values are better than obtained using HRXRD we compare experimental results of bandgap value for all samples versus mole fraction obtained using HRXRD, FLA and FTA. For each points we plot theoretical curve (3) that is commonly accepted in literature [14].
Where Eg stands for bandgap energy for InAsSb, InAs and InSb, x is Sb mole fraction and C bowing parameter. One can see that for all our figures curves are similar. On the other hand points obtained from Raman method are much better fit to theoretical curve 3 (R2 > 0.9) than using HRHXR (R2 < 0.9). The matching parameter C for HRXRD, FLA and FTA are 0.69, 0.69 and 0.68 respectively. This value is close to obtained by other groups [20–22].Disagreement between experimental results and parabolic curve obtained for Fig. 3 may be also possibly explained by differences if ordering for each sample. Ordering has influence on the bandgap energy [14] and differences in ordering may cause errors in fitting in Fig. 3(a). One of the way to measure ordering is to compare bandgap value obtained using PL for different excitation power [23]. In the Fig. 4 we present PL spectra for sample where we observe highest energy shift for different excitation power measured in 300 K (Fig. 4). The difference between peaks’ energies is 3 meV. This means any differences in ordering for presented samples does not have strong influence on band gap measurements.
Fig. 4. PL spectra for sample #10139 performed in 300 K with 40 mW (blue curve) and 200 mW (red curve) laser power.
Presented in this article idea to determine mole fraction using Raman low energy peaks (FLA and FTA) may be applied to other III-V alloys. However for each type of analyzing material one have to have enough high sample quality to be able to distinguish FLA or/and FTA peaks. For example in [8,9] quality of Raman spectra is too low to analyze FLA peaks, on the other hand in [12] authors could measure it.
4. Conclusions
In this article we showed an alternative way to calculate mole fraction value for InAs1−xSbx using FLA and FTA. We calculated and compared bowing parameter using mole fraction determination by HRXRD, FLA and FTA. In all cases obtained parameter is very similar, however fit to data points is much better for FLA and FTA than HRXRD calculated mole fraction. Obtained parameter 0.69 is close to previously reported.
Funding
Narodowe Centrum Badań i Rozwoju (TECHMATSTRATEG1/347751/5/NCBR/2017.).
Acknowledgments
This paper has been completed with the financial support of the - grant no. TECHMATSTRATEG1/347751/5/NCBR/2017.
References
1. J. C. Woolley, J. Warner, J. C. Woolley, and J. Warner, “Preparation of InAs-InSb Alloys,” J. Electrochem. Soc. 111(10), 1142–1145 (1964). [CrossRef]
2. A. Rogalski, “Infrared detectors: Status and trends,” Prog. Quantum Electron. 27(2-3), 59–210 (2003). [CrossRef]
3. J. Piotrowski and A. Rogalski, High-Operating-Temperature Infrared Photodetectors (SPIE, 2007).
4. M. Erkuş and U. Serincan, “Phonon frequency variations in high quality InAs1-xSbx epilayers grown on GaAs,” in Applied Surface Science (North-Holland, 2014), 318, pp. 28–31.
5. M. A. Marciniak, R. L. Hengehold, Y. K. Yeo, and G. W. Turner, “Optical characterization of molecular beam epitaxially grown InAsSb nearly lattice matched to GaSb,” J. Appl. Phys. 84(1), 480–488 (1998). [CrossRef]
6. K. Murawski, K. Grodecki, D. Benyahia, A. Wysmolek, B. Jankiewicz, and P. Martyniuk, “X-ray and Raman determination of InAsSb mole fraction for x <0.5,” J. Cryst. Growth 498, 137–139 (2018). [CrossRef]
7. K. Radhakrishnan, T. H. K. Patrick, H. Q. Zheng, P. H. Zhang, and S. F. Yoon, “InP/InxGa1-xAs (0.53 ≤ x ≤ 0.81) high electron mobility transistor structures grown by solid source molecular beam epitaxy,” J. Cryst. Growth 207(1-2), 8–14 (1999). [CrossRef]
8. Z. C. Feng, A. A. Allerman, P. A. Barnes, and S. Perkowitz, “Raman scattering of InGaAs/InP grown by uniform radial flow epitaxy,” Appl. Phys. Lett. 60(15), 1848–1850 (1992). [CrossRef]
9. K. Radhakrishnan, T. H. K. Patrick, H. Q. Zheng, P. H. Zhang, and S. F. Yoon, “Study of doping concentration variation in InGaAs/InP high electron mobility transistor layer structures by Raman scattering,” J. Vac. Sci. Technol., A 18(2), 713–716 (2000). [CrossRef]
10. Y. T. Cherng, K. Y. Ma, and G. B. Stringfellow, “Raman scattering in InAs1-xSbx grown by organometallic vapor phase epitaxy,” Appl. Phys. Lett. 53(10), 886–887 (1988). [CrossRef]
11. K. Grodecki, K. Murawski, K. Michalczewski, B. Budner, and P. Martyniuk, “Raman scattering of InAsSb,” AIP Adv. 9(2), 025107 (2019). [CrossRef]
12. T. Mori, T. Hanada, T. Morimura, K. Shin, H. Makino, and T. Yao, “Surface structure of InGaAs/InP(0 0 1) ordered alloy during and after growth,” in Applied Surface Science (North-Holland, 2004), 237(1–4), pp. 230–234.
13. H. M. Cheong, S. P. Ahrenkiel, M. C. Hanna, and A. Mascarenhas, “Phonon signatures of spontaneous CuPt ordering in Ga0.47In0.53As/InP,” Appl. Phys. Lett. 73(18), 2648–2650 (1998). [CrossRef]
14. S. P. Svensson, W. L. Sarney, H. Hier, Y. Lin, D. Wang, D. Donetsky, L. Shterengas, G. Kipshidze, and G. Belenky, “Band gap of InAs1-xSbx with native lattice constant,” Phys. Rev. B: Condens. Matter Mater. Phys. 86(24), 245205 (2012). [CrossRef]
15. L. O. Bubulac, A. M. Andrews, E. R. Gertner, and D. T. Cheung, “Backside-illuminated InAsSb/GaSb broadband detectors,” Appl. Phys. Lett. 36(9), 734–736 (1980). [CrossRef]
16. M. Motyka, G. Sęk, J. Misiewicz, A. Bauer, M. Dallner, S. Höfling, A. Forchel, G. Sȩk, J. Misiewicz, A. Bauer, M. Dallner, S. Höfling, A. Forchel, G. Sęk, J. Misiewicz, A. Bauer, M. Dallner, S. Höfling, and A. Forchel, “Fourier transformed Photoreflectance and photoluminescence of mid infrared GaSb-based type II quantum wells,” Appl. Phys. Express 2(12), 126505 (2009). [CrossRef]
17. Z. Jinghua, T. Xiaohong, and T. Jinghua, “Atomic ordering of AlInP grown by MOVPE at different temperatures in pure ambient N2,” CrystEngComm 11(6), 1068–1072 (2009). [CrossRef]
18. L. Lindsay, D. A. Broido, and T. L. Reinecke, “Ab initio thermal transport in compound semiconductors,” Phys. Rev. B: Condens. Matter Mater. Phys. 87(16), 165201 (2013). [CrossRef]
19. Y. T. Cherng, K. Y. Ma, and G. B. Stringfellow, “Raman scattering in InAs1-xSbx grown by organometallic vapor phase epitaxy,” Appl. Phys. Lett. 53(10), 886–887 (1988). [CrossRef]
20. I. Vurgaftman, J. R. Meyer, and L. R. Ram-Mohan, “Band parameters for III-V compound semiconductors and their alloys,” J. Appl. Phys. 89(11), 5815–5875 (2001). [CrossRef]
21. E. H. Steenbergen, O. O. Cellek, D. Lubyshev, Y. Qiu, J. M. Fastenau, A. W. K. Liu, and Y.-H. Zhang, “Study of the valence band offsets between InAs and InAs1-xSbx alloys,” Proc. SPIE 8268, 82680K (2012). [CrossRef]
22. S. R. Kurtz, R. M. Biefeld, and A. J. Howard, “Magneto-optical determination of light-heavy hole splittings in As-rich, InAsSb alloys and superlattices,” Appl. Phys. Lett. 67(22), 3331–3333 (1995). [CrossRef]
23. H. Dumont, L. Auvray, and Y. Monteil, “Optical signature of atomic ordering in In0.53Ga 0.47As/InP: Photoluminescence properties and IR response,” in Optical Materials (North-Holland, 2003), 24(1–2), pp. 309–314.
