Exploration of GaInT1P and Related T1-Containing III-V Alloys for Photovoltaics

Transcription

1 NREL/CP Exploration of GaInT1P and Related T1-Containing III-V Alloys for Photovoltaics D.J. Friedman, S.R. Kurtz, and A.E. Kibbler National Renewable Energy Laboratory Presented at the National Center for Photovoltaics Program Review Meeting Denver, Colorado September 8-11, 1998 National Renewable Energy Laboratory 1617 Cole Boulevard Golden, Colorado A national laboratory of the U.S. Department of Energy Managed by Midwest Research Institute for the U.S. Department of Energy under contract No. DE-AC36-83CH10093 Work performed under task number PV October 1998

2 NOTICE This report was prepared as an account of work sponsored by an agency of the United States government. Neither the United States government nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States government or any agency thereof. Available to DOE and DOE contractors from: Office of Scientific and Technical Information (OSTI) P.O. Box 62 Oak Ridge, TN Prices available by calling Available to the public from: National Technical Information Service (NTIS) U.S. Department of Commerce 5285 Port Royal Road Springfield, VA or or DOE Information Bridge Printed on paper containing at least 50% wastepaper, including 10% postconsumer waste

3 Exploration of GaInTlP and Related Tl-containing III-V Alloys for Photovoltaics D. J. Friedman, Sarah R. Kurtz, and A. E. Kibbler National Renewable Energy Laboratory, 1617 Cole Blvd., Golden, CO Abstract. This paper discusses the results of an attempt to grow GaInTlP for application as a 1-eV material for the third junction of a GaInP/GaAs/3rd-junction high-efficiency solar cell. Although early indications from the literature were promising, we are unable to produce crystalline homogeneous material, and so we conclude that this material is not a promising candidate for such applications as photovoltaics. INTRODUCTION Future-generation extensions of the GaInP/GaAs two-junction solar cell call for adding a third junction under the GaInP/GaAs stack, with the possibility of adding a fourth junction of Ge underneath (1). For optimal device efficiency and materials quality, a semiconductor lattice-matched to GaAs and with a 1-eV bandgap is desired for this third junction. Although GaInNAs has received the most attention in this regard (2), other materials have been considered as well. In 1994, van Shilfgaarde on theoretical grounds proposed that In 1 x Tl x P would have a lattice constant essentially equal to that of InP, and a band gap ranging from the 1.3 ev band gap of InP down to 0 ev with increasing Tl content, thus making this material of interest for a range of infrared-detector applications (3). Shortly thereafter, Asahi extended this proposal to include GaInTlP and GaInTlAs, the idea being essentially that Tl could be substituted for In with little change in lattice constant and a significant decrease in band gap, thus extending the range of band gaps available for materials lattice-matched to GaAs and InP (4). One alloy composition of Asahi s proposal is of special interest for us: the material Ga 0.5 Tl 0.5 P would be lattice-matched to GaAs and have a band gap of 1 ev. Although Asahi was interested in infrared detectors, these characteristics also satisfy the requirements for our 3rd-junction solar cell as described above. Encouragingly, Asahi s group presented several papers claiming successful growth of these Tl-containing alloys (4-7). However, in the meantime, Berding extended van Schilfgaarde s theoretical analysis to conclude that TlP and TlAs would be unstable, and that there would be extremely high group-v vapor pressures over InTlP and InTlAs; in other words, that these materials may not be growable (8). Furthermore, Antonell reported a lack of success in attempting to grow these materials (9,10). We undertook a study, the 1

4 results of which are reported here, to explore the growth of these alloys to resolve this inconsistency and evaluate them for use as the third junction in III-V multijunction solar cells. EXPERIMENT Materials Growth: The materials growth (or attempted growth) was performed by solid-source molecular-beam epitaxy (MBE). Arsenic and phosphorus was Tl BEP (torr) Tl oven temperature ( C) /(Tl oven temperature( K)) FIGURE 1. Tl BEP as a function of oven temperature. obtained from valve crackers, with a typical beam-equivalent pressure (BEP) of 10 5 torr. The group-iii sources Ga, In, and Tl were obtained from conventional Knudsen cells, at BEPs typically in the 10 7 to 10 6 torr range. Of these materials, only the Tl is unconventional. The Tl starting material as loaded into the Knudsen cell was an ingot specified to be of % purity. As Asahi reported, the Tl proved to be easy to evaporate with a well-behaved BEP. Figure 1 shows a plot of the Tl BEP over the range of interest for materials growth, indicating good reproducibility and an easily obtainable oven temperature. Epilayer Measurements: Surface morphology for each of the epilayers grown was examined using Nomarski optical microscopy and/or scanning electron microscopy (SEM). Chemical compositions of some of the samples were measured by electron probe microanalysis (EPMA). Band gaps were determined by measuring the photocurrent from a junction between a liquid electrolyte and the epilayer surface (11). RESULTS We attempted to grow Tl-containing alloys over a wide range of growth conditions. The sections below illustrate the trends in the general results. High-temperature growth of GaTlP: We first attempted to grow GaTlP under what would be considered fairly conventional MBE growth conditions for phosphides. A growth temperature T g =450 C was used, with the BEP ratio BEP(Tl)/BEP(Ga)=3 that is, the Tl flux was high enough so that if the Tl incident upon the growth surface incorporated into the growing semiconductor, the Ga/Tl composition ratio in the as-grown material would be on the order of 1. Nomarski microscopy of the resulting surface showed a specular surface. The spectral response of the epilayer is shown in Fig. 2. There is no evidence of any 1-eV direct-gap semiconductor: the only band edge visible is from the GaAs substrate, with the rest of the spectrum being suggestive of GaP. 2

5 spectral response (a.u.) GaAs band edge from substrate sample: EA photon energy (ev) FIGURE 2. Spectral response of nominally- GaTlP epilayer grown at T g =450 C. spectral response (a.u.) ev Ga 0.5In 0.5P band edge sample: EA photon energy (ev) FIGURE 3. Spectral response of nominally- GaInTlP epilayer grown at T g =450 C. High-temperature growth ot GaTl x In 1 x P 2 : As a complementary parallel approach, we attempted to incorporate Tl by starting with Ga 0.5 In 0.5 P and substituting a fraction of the In with Tl. At a growth temperature of 450 C, we again obtained a specular surface. The resulting spectral response is shown in Fig. 3. It is indicative of the presence only of Ga 0.5 In 0.5 P, with no evidence for any band gap depression due to Tl incorporation. Intermediate-Temperature Growth of GaTlP: In an attempt to enhance the Tl incorporation in the material, an attempt was made to grow GaTlP at the lower temperature of 350 C. The resulting surface morphology was noniform, with metal droplets covering the surface as shown in Fig. 4a. The irregular objects on the surface are Tl droplets; EPMA indicated that in the area of the surface not covered by droplets, there is no Tl within measurement error (~ <1%). The spectral response, which is especially weak for this sample, is shown in Fig. 4b FIGURE 4a. SEM micrograph of GaTlP surface for T g =350 C. Spectral response (a.u.) GaAs band edge from substrate sample: EA photon energy (ev) FIGURE 4b. Spectral response for sample of Fig. 4a. Intermediate-temperature growth otfgatl x In 1 x P 2 : A similar attempt was made to enhance Tl incorporation in GaTl x In 1 x P 2 by growing at T g =350 C. The resulting surface morphology had a significant concentration of metal droplets. No spectral reponse could be measured for this sample, due to the metal on the surface. For 3

6 FIGURE 5. Nomarski micrograph of GaInTlP surface for T g =350 C (sample ID: EA375). Note the metal droplets. surface roughness (arb. units comparison, Ga 0.5 In 0.5 P was grown under conditions the same except for the absence of the Tl flux, and a specular surface was obtained growth anneal epilayer desorbs sample temperature ( C) roughn. T g sample EA398 run time (minutes) FIGURE 6. Evolution of GaTlP surface roughness on annealing. Stability of Low-Tg-Grown GaTlP: Finally, a layer of GaTlP was grown at room temperature, an extremely low growth temperature far too low to obtain crystalline material. The sample was then slowly heated up under a phosphorus flux as the roughness of the epilayer was monitored in-situ by monitoring the intensity of diffusely-scattered laser light from the surface. The roughness and T g are plotted vs run time in Fig. 6. The sudden spike in the roughness is indicative of desorption of the epilayer, showing the material to be unstable even at these relatively low temperatures. DISCUSSION The above results can be summarized by noting that essentially no incorporation of Tl is obtained at growth temperatures high enough for good surface and bulk crystalline quality. This is consistent with Berding s theory and Antonell s experimental results. Also, more recently there have been additional attempts by others to grow Tl-containing III-V compounds, none of which report successful growth of homogeneous crystalline Tl-containing material [(12-14). Without new and independent evidence to the contrary, it appears that the original reports (4-7) may need to be reexamined. It may be noted that the original reports deduced the presence of Tl indirectly, from x-ray diffraction, and more conclusive, chemical, evidence has not yet been presented by these workers. 4

7 CONCLUSIONS The initial promise of the Tl-containing III-V materials does not seem to have been realized: growth of homogeneous crystalline material is elusive if it is possible at all. Thus, these materials are not promising candidates for photovoltaic applications. REFERENCES 1. S. R. Kurtz, D. Myers and J. M. Olson, in 26th IEEE Photovoltaic Specialists Conference, 1997, pp J. F. Geisz, D. J. Friedman, J. M. Olson, C. Kramer, A. Kibbler and S. R. Kurtz, Proceedings of the 1998 NREL/SNL PV Program Review (this proceedings). 3. M. vanschilfgaarde, A.-B. Chen, S. Krishnamurthy and A. Sher, Appl. Phys. Lett. 65, 2714 (1994). 4. H. Asahi, K. Yamamoto, K. Iwata, S. Gonda and K. Oe, Jpn. J. of Appl. Phys. Lett. 35, 876 (1996). 5. K. Yamamoto, H. Asahi, M. Fushida, K. Iwata and S. Gonda, J. Appl. Phys. 81, 1704 (1997). 6. H. Asahi, M. Fushida, K. Yamamoto, K. Iwata, H. Koh, K. Asami, S. Gonda and K. Oe, J. Cryst. Growth 175/176, 1195 (1997). 7. H. Asahi, M. Fushida, H. Koh, K. Yamamoto, K. Asami, S. Gonda and K. Oe, Proceedings of the 1997 International Conference on InP and Related Materials 448 (1997). 8. M. Berding, M. vanshilfgaarde, A. Sher, M. J. Antonell and C. R. Abernathy, J. Electron. Mater. 26, 683 (1997). 9. M. J. Antonell, C. R. Abernathy, A. Sher, M. Berding and M. vanschilfgaarde, Proceedings of the 1997 International Conference on InP and Related Materials 444 (1997). 10. M. J. Antonell, C. R. Abernathy, A. Sher, M. Berding, M. vanschilfgaarde, A. Sanjuro and K. Wong, J. Cryst. Growth 188, 113 (1998). 11. S. R. Kurtz and J. M. Olson, in IEEE Photovoltaic Specialists Conference(IEEE, New Orleans, Louisiana), 1987, pp M. D. Lange, D. F. Storm and T. Cole, J. Electron. Mater. 27, 536 (1998). 13. J.-S. Liu, J.-S. Wang and H.-H. Lin, J. Cryst. Growth 192, 372 (1998). 14. D. I. Lubyshev, W. Z. Cai, G. L. Catchen, T. S. Mayer and D. L. Miller, Proceedings of the 24th International Symposium on Compound Semiconductors (1997). 5