Optical properties of InP nanowires on Si ... - Semantic Scholar
APPLIED PHYSICS LETTERS 92, 013121 共2008兲
Optical properties of InP nanowires on Si substrates with varied synthesis parameters Linus C. Chuang, Michael Moewe, Shanna Crankshaw, and Connie Chang-Hasnaina兲 Department of Electrical Engineering and Computer Sciences, and Applied Science and Technology Group, University of California at Berkeley, Berkeley, California 94720, USA
共Received 18 October 2007; accepted 17 December 2007; published online 8 January 2008兲 We report the effect of synthesis parameters on the physical appearance and optical properties of InP nanowires 共NWs兲 grown on Si substrates by metal-organic chemical vapor deposition. A strong dependence on the group V to III precursor ratio is observed on the NW shape and, consequently, its photoluminescence 共PL兲. Narrow, uniform-diameter NWs are achieved with an optimized V/III ratio. The uniform NWs exhibit PL widths as low as 1.4 meV. Their peak wavelength does not vary much with excitation, which is important for NW lasers on Si. These characteristics are attributed to the one-dimensional density of states in uniform-diameter NWs. © 2008 American Institute of Physics. 关DOI: 10.1063/1.2832643兴 The integration of III-V compound materials with Si has been an important research area for monolithic integration of semiconductor diode lasers and Si-based electronic circuits.1–4 However, past attempts have not been successful because of poor laser reliability due to high defect densities resulting from a large lattice mismatch and process incompatibility with complementary metal-oxide semiconductor integrated circuits due to high temperatures required for epitaxial synthesis.5,6 III-V compound nanowires 共NWs兲 grown on Si substrates have recently drawn much attention because they provide a means of circumventing these difficulties for integration.7–11 Recently, we showed dislocation-free III-V NWs with excellent optical properties grown at substantially lower temperatures 共430– 470 ° C兲 using a metal-organic chemical vapor deposition 共MOCVD兲 system.7 In this paper, we report the effect of the precursor V/III ratio on the shape and optical properties of InP NWs grown on 共111兲 Si substrates. We show that the V/III ratio can be used to tailor the NW shape and optical properties. In particular, we report the growth of NWs with uniform diameters along the axial direction with a record narrow photoluminescence 共PL兲 peak of 1.4 meV and a large blueshift of 178 meV due to quantization. These uniform NWs also have less power dependence for their PL emission peak wavelength. This wavelength stability is important for critical applications such as NW lasers. The PL intensity can also be maximized when using high V/III ratios. Five InP NW on 共111兲 Si samples were grown with V/III ratios equal to 15, 30, 67, 180, and 240. The Si substrates were first cleaned and then chemically deoxidized with buffered oxide etch followed by Au nanoparticle 共NP兲 dispensing. Colloidal Au NPs with an average of 20 nm diameter were used as the catalysts for vapor-liquid-solid 共VLS兲 NW growth.12 The growth procedure was the same as Ref. 7 using a MOCVD system. The group V and III precursors were tertiarybutylphosphine 共TBP兲 and trimethylindium 共TMIn兲, respectively. The TMIn mole fraction was held at 1.9 ⫻ 10−5 in a 12 l / min hydrogen carrier gas flow for all the growths. The TBP mole fraction was varied to attain the five a兲
Author to whom correspondence should be addressed. Tel.: 1-510-6424315. Electronic mail: [email protected].
0003-6951/2008/92共1兲/013121/3/$23.00
V/III ratios: 15, 30 67, 180, and 240 共samples A, B, C, D, and E, respectively兲. The V/III ratios we quote above are the supplied gas-phase mole ratios which are the input experimental parameters for the NW syntheses in this work. The growth time was 3 min for all the samples and the growth pressure was 76 Torr. The NW shape was characterized by field-emission scanning electron microscopy 共FE-SEM兲. Optical properties were characterized by microphotoluminescence 共-PL兲 measurements at both room temperature 共RT兲 and at 4 K. Figure 1 shows the FE-SEM images of the five InP NW/ 共111兲 Si samples. With increasing V/III ratio, a significant NW shape change was observed. With a low V / III= 15, NWs did not grow due to insufficient phosphorus 关Fig. 1共a兲兴. Many indium-rich balls, whose composition was determined by SEM energy dispersive spectroscopy, were observed. The improvement for NW formation is seen for sample B with V / III= 30 关Fig. 1共b兲兴. NWs on sample C, with V / III= 67, have uniform diameters along the entire NW lengths 关Fig. 1共c兲兴. Further increasing the V/III ratio resulted in tapered NWs with wider bases and narrow tips, shown in Figs. 1共d兲 and 1共e兲, with V/III ratios equal to 180 and 240, respectively. The tapering is attributed to an increase in the thin-film deposition rate on NW sidewalls compared to that of the vertical VLS growth. The dependence of thin film growth rate on V/III ratio has been previously reported for GaAs material in conventional thin film growth.13 Here, we observed the sidewall thin-film growth mechanism for InP NW growth, which has a similar V/III ratio dependence, as that shown in Ref. 13. The two tapered NW samples, D and E, appear to have slightly thinner tips than sample C. This might be due to the increased Au-catalyst diffusion into the NWs during the growth for the higher V/III ratio conditions.14 This phenomenon served as a secondary effect to make the NWs more tapered. Between the straight NW sample C 共V / III= 67兲 and the tapered NW sample D 共V / III= 180兲, two more V/III ratios, 90 and 120, were tested. While the V / III= 90 NW sample still looks straight, the V / III= 120 sample begins to show some taperness. The onset of NW tapering is then deduced as between V / III= 90 and 120.
92, 013121-1
© 2008 American Institute of Physics
013121-2
Chuang et al.
Appl. Phys. Lett. 92, 013121 共2008兲
FIG. 2. 共Color online兲 共T = 4 K兲 Representative -PL spectra for various V/III ratio growth conditions 共samples A–E兲 with excitation power equal to 1.8 W. Gray line indicates the 4 K InP bulk bandgap 共1.423 eV兲. Inset shows a single NW spectrum from a straight and narrow NW 共from sample C, V / III= 67兲, with a full width at half maximum 共FWHM兲 of 1.4 meV and blueshift of 178 meV from the bulk bandgap. For sample C, the high-energy side of the spectrum is spiky due to the contribution of these individual thin and straight NWs.
FIG. 1. 20° tilt FE-SEM images for five InP NW/共111兲Si samples, sample A, B, C, D, and E, with V/III ratios equal to 共a兲 15, 共b兲 30, 共c兲 67, 共d兲 180, and 共e兲 240, respectively. 共a兲 and 共b兲 represent insufficient V/III ratios and, hence, indium-rich growth conditions. 共c兲 represents a growth condition for yielding nontapered NWs. 共d兲 and 共e兲 represent the high V/III ratio growth conditions which yield tapered-NWs due to the stronger thin-film NW sidewall deposition.
-PL characterization was performed on all samples at 4 K and RT 共300 K兲 using a diode-pumped solid state laser at 532 nm focused to a ⬃2 m spot. Figure 2 shows the 4 K -PL spectrum comparison for the five samples in Fig. 1. The emission wavelengths of samples A and B are expected to be very close to the bulk InP bandgap since there are very few NWs on both samples but only some larger InP blobs, which might contribute to this PL emission. The PL peak of sample C with uniform NWs shows a blueshift of 40 meV from the bulk InP bandgap due to quantum confinement.7 Single-wire peaks from the narrowest NWs are visible on the high-energy side of the ensemble spectrum. For example, a peak with a 178 meV blueshift and linewidth of 1.4 meV was observed 共see inset of Fig. 2兲, which is the narrowest linewidth reported for a III-V NW.7,15 Sample D has a similar PL blueshift as sample C. However, the spiky features at the high energy side are not observed for sample D. This can be explained by the following. At low temperature, the carrier diffusion length is longer.16 Hence, carriers originally generated at the tips of the very narrow NWs can diffuse to the wider parts of the NWs where they see less quantum confinement 共smaller photon energy兲 and recombine there. For sample E, the carrier diffusion phenomenon is more pronounced for the strongly tapered NWs. As a consequence, for sample E, not only have the spiky features disappeared, the
PL peak shifts significantly to the redder side of sample C, at 20 meV blueshift of the InP bulk bandgap. The RT -PL spectra 共Fig. 3兲 show significant differences with those at 4 K. First of all, the fine features are no longer visible for sample C. Secondly, sample E has the same PL peak energy and linewidth as samples C and D, instead of being 20 meV redder at 4 K. This is particularly interesting since the NW base is wider than 100 nm and emission at the bulk bandgap is expected. We attribute this observation to luminescence from the narrow NW tips, where at RT confined carriers recombine radiatively before being able to diffuse to the wider base region. Third, samples A to D all show an extra 60 meV blueshift from the bulk InP bandgap compared to the amount of blueshift at 4 K 共see Fig. 2兲. The origin of this extra 60 meV blueshift is under further investigation. The PL intensity for both 4 K 共Fig. 2兲 and RT 共Fig. 3兲 measurements increases with V/III ratio. A similar trend is observed for InP epilayers,17,18 where it has been shown that high V/III ratios result in epilayers with lower defect densities.18,19 Figure 4共a兲 shows the peak energy position as a function of the excitation power at 4 K. When the excitation power is swept from 600 nW to 1.6 mW, over a three-decade power increase, the straight 共V / III= 67兲 and most tapered NWs
FIG. 3. 共Color online兲 RT PL spectrum for the various V/III ratios with excitation power equal to 400 W. Gray line indicates RT InP bulk bandgap 共1.340 eV兲. All samples are blueshifted with respect to the bulk bandgap. No fine features are seen due to thermal broadening.
013121-3
Appl. Phys. Lett. 92, 013121 共2008兲
Chuang et al.
mized to 67, nontapered InP NWs were grown. These NWs show a record narrow PL peak and weak excitation-power dependence, resembling features of ideal one dimensional structures. In our experiments, non-tapered InP NWs could be synthesized for V/III ratios ranging from 60 to 90, hence, offering a reasonable growth window. We also showed that the PL peak intensity could be increased with the increase of V/III ratio.
FIG. 4. 共Color online兲 T = 4 K 共a兲 -PL peak energy position versus excitation power for straight InP NWs 共sample C, V / III= 67兲 and strongly tapered InP NWs 共sample E, V / III= 240兲, respectively. Sample C shows a smaller wavelength dependence of only a 9 meV blueshift when excitation power is increased from 600 nW to 1.6 mW, while sample E shows a 40 meV blueshift. 共b兲 -PL peak intensity vs excitation power from 600 nW to 1.6 mW for various V/III ratios. The higher the V/III ratio, the larger the PL peak intensity is. Samples C and D show some intensity saturation at higher excitation powers while sample E does not.
共V / III= 240兲 show a blueshift of 9 and 40 meV, respectively. The small power-dependent blueshift for the straight NWs reflects the one-dimensional density of states nature of these NWs. The wavelength stability over a large excitation power range is important for critical devices such as lasers. The large blueshift for the tapered NWs is attributed to the effect of decreased carrier diffusion length at higher pumping, resulting in emission from the narrower part of the tapered NWs. The -PL peak intensity versus excitation power was also studied, as shown in Fig. 4共b兲. The higher V/III ratio, as described earlier, resulted in a larger PL peak intensity. For example, sample E is more than ten times brighter than sample C at 1.6 mW excitation. This effect might be also due to the larger NW volume since higher V/III ratio increases the sidewall deposition rate. The PL peak intensity shows some saturation at higher excitation powers for samples C and D, while there is no saturation observed for E within this range, likely due to the NW volume difference. For example, the NW volume of sample E is estimated as 2.5 ⫻ 10−15 cm3 and the NW volume ratio between samples E and D is ⬃6. The number of available electronic states is proportional to the volume which makes sample E remain unsaturated at higher pumping levels. In conclusion, we demonstrate the V/III ratio effects on both the shape and optical properties for InP NWs grown on 共111兲 Si substrates. The higher the V/III ratio is, the more tapered the NWs become. The PL intensity increases dramatically with the V/III ratio. When the V/III ratio is opti-
This work was supported by DARPA HR0011-04-10040 共CONSRT兲 and HP-CITRIS grants. The authors acknowledge the fellowship support from the UC Berkeley EECS Fellowship Program, NSF-IGERT Program, and NSF Graduate Research Fellowship Program. Linus C. Chuang and Michael Moewe contributed equally to this work. 1
M. Razeghi, The MOCVD Challenge, Volume 1: A Survey of GaInAsP-InP for Photonic and Electronic Applications 共Hilger, Bristol and Philadelphia, 1989兲. 2 Z. Mi, J. Yang, P. Bhattacharya, P. K. L. Chan, and K. P. Pipe, J. Vac. Sci. Technol. B 24, 1519 共2006兲. 3 G. Balakrishnan, A. Jallipalli, P. Rotella, S. Huang, A. Khoshakhlagh, A. Amtout, S. Krishna, L. R. Dawson, and D. L. Huffaker, IEEE J. Sel. Top. Quantum Electron. 12, 1636 共2006兲. 4 M. E. Groenert, A. J. Pitera, R. J. Ram, and E. A. Fitzgerald, J. Vac. Sci. Technol. B 21, 1064 共2003兲. 5 S. F. Fang, K. Adomi, S. Iyer, H. Morkoç, H. Zabel, C. Choi, and N. Otsuka, J. Appl. Phys. 68, R31 共1990兲. 6 R. M. Lum, J. K. Klingert, R. B. Bylsma, A. M. Glass, A. T. Macrander, T. D. Harris, and M. G. Lamont, J. Appl. Phys. 64, 6727 共1988兲. 7 L. C. Chuang, M. Moewe, C. Chase, N. P. Kobayashi, C. Chang-Hasnain, and S. Crankshaw, Appl. Phys. Lett. 90, 043115 共2007兲. 8 T. Mårtensson, C. P. T. Svensson, B. A. Wacaser, M. W. Larsson, W. Seifert, K. Deppert, A. Gustafsson, L. R. Wallenberg, and L. Samuelson, Nano Lett. 4, 1987 共2004兲. 9 E. P. A. M. Bakkers, M. T. Borgström, and M. A. Verheijen, MRS Bull. 32, 117 共2007兲. 10 S. S. Yi, G. Girolami, J. Amano, M. S. Islam, S. Sharma, T. I. Kamins, and I. Kimukin, Appl. Phys. Lett. 89, 133121 共2006兲. 11 N. P. Kobayashi, L. VJ, M. S. Islam, X. Li, J. Straznicky, S.-Y. Wang, R. S. Williams, and Y. Chen, Appl. Phys. Lett. 91, 113116 共2007兲. 12 R. S. Wagner and W. C. Ellis, Appl. Phys. Lett. 4, 89 共1964兲. 13 T. Kikkawa, H. Tanaka, and J. Komeno, J. Appl. Phys. 67, 7576 共1990兲. 14 J. B. Hannon, S. Kodambaka, F. M. Ross, and R. M. Tromp, Nature 共London兲 440, 69 共2006兲. 15 M. H. M. van Weert, O. Wunnicke, A. L. Roest, T. J. Eijkemans, A. Yu Silov, J. E. M. Haverkort, G. W. ‘t Hooft, and E. P. A. M. Bakkers, Appl. Phys. Lett. 88, 043109 共2006兲. 16 S. M. Sze, Physics of Semiconductor Devices 共Wiley, New York, 1981兲. 17 D. Kasemset, K. L. Hess, K. Mohammed, and J. L. Merz, J. Electron. Mater. 13, 655 共1984兲. 18 C. C. Hsu, J. S. Yuan, R. M. Cohen, and G. B. Stringfellow, J. Cryst. Growth 74, 535 共1986兲. 19 D. Keiper, R. Westphalen, and G. Landgren, J. Cryst. Growth 204, 256 共1999兲.
Optical properties of InP nanowires on Si substrates with varied synthesis parameters Linus C. Chuang, Michael Moewe, Shanna Crankshaw, and Connie Chang-Hasnaina兲 Department of Electrical Engineering and Computer Sciences, and Applied Science and Technology Group, University of California at Berkeley, Berkeley, California 94720, USA
共Received 18 October 2007; accepted 17 December 2007; published online 8 January 2008兲 We report the effect of synthesis parameters on the physical appearance and optical properties of InP nanowires 共NWs兲 grown on Si substrates by metal-organic chemical vapor deposition. A strong dependence on the group V to III precursor ratio is observed on the NW shape and, consequently, its photoluminescence 共PL兲. Narrow, uniform-diameter NWs are achieved with an optimized V/III ratio. The uniform NWs exhibit PL widths as low as 1.4 meV. Their peak wavelength does not vary much with excitation, which is important for NW lasers on Si. These characteristics are attributed to the one-dimensional density of states in uniform-diameter NWs. © 2008 American Institute of Physics. 关DOI: 10.1063/1.2832643兴 The integration of III-V compound materials with Si has been an important research area for monolithic integration of semiconductor diode lasers and Si-based electronic circuits.1–4 However, past attempts have not been successful because of poor laser reliability due to high defect densities resulting from a large lattice mismatch and process incompatibility with complementary metal-oxide semiconductor integrated circuits due to high temperatures required for epitaxial synthesis.5,6 III-V compound nanowires 共NWs兲 grown on Si substrates have recently drawn much attention because they provide a means of circumventing these difficulties for integration.7–11 Recently, we showed dislocation-free III-V NWs with excellent optical properties grown at substantially lower temperatures 共430– 470 ° C兲 using a metal-organic chemical vapor deposition 共MOCVD兲 system.7 In this paper, we report the effect of the precursor V/III ratio on the shape and optical properties of InP NWs grown on 共111兲 Si substrates. We show that the V/III ratio can be used to tailor the NW shape and optical properties. In particular, we report the growth of NWs with uniform diameters along the axial direction with a record narrow photoluminescence 共PL兲 peak of 1.4 meV and a large blueshift of 178 meV due to quantization. These uniform NWs also have less power dependence for their PL emission peak wavelength. This wavelength stability is important for critical applications such as NW lasers. The PL intensity can also be maximized when using high V/III ratios. Five InP NW on 共111兲 Si samples were grown with V/III ratios equal to 15, 30, 67, 180, and 240. The Si substrates were first cleaned and then chemically deoxidized with buffered oxide etch followed by Au nanoparticle 共NP兲 dispensing. Colloidal Au NPs with an average of 20 nm diameter were used as the catalysts for vapor-liquid-solid 共VLS兲 NW growth.12 The growth procedure was the same as Ref. 7 using a MOCVD system. The group V and III precursors were tertiarybutylphosphine 共TBP兲 and trimethylindium 共TMIn兲, respectively. The TMIn mole fraction was held at 1.9 ⫻ 10−5 in a 12 l / min hydrogen carrier gas flow for all the growths. The TBP mole fraction was varied to attain the five a兲
Author to whom correspondence should be addressed. Tel.: 1-510-6424315. Electronic mail: [email protected].
0003-6951/2008/92共1兲/013121/3/$23.00
V/III ratios: 15, 30 67, 180, and 240 共samples A, B, C, D, and E, respectively兲. The V/III ratios we quote above are the supplied gas-phase mole ratios which are the input experimental parameters for the NW syntheses in this work. The growth time was 3 min for all the samples and the growth pressure was 76 Torr. The NW shape was characterized by field-emission scanning electron microscopy 共FE-SEM兲. Optical properties were characterized by microphotoluminescence 共-PL兲 measurements at both room temperature 共RT兲 and at 4 K. Figure 1 shows the FE-SEM images of the five InP NW/ 共111兲 Si samples. With increasing V/III ratio, a significant NW shape change was observed. With a low V / III= 15, NWs did not grow due to insufficient phosphorus 关Fig. 1共a兲兴. Many indium-rich balls, whose composition was determined by SEM energy dispersive spectroscopy, were observed. The improvement for NW formation is seen for sample B with V / III= 30 关Fig. 1共b兲兴. NWs on sample C, with V / III= 67, have uniform diameters along the entire NW lengths 关Fig. 1共c兲兴. Further increasing the V/III ratio resulted in tapered NWs with wider bases and narrow tips, shown in Figs. 1共d兲 and 1共e兲, with V/III ratios equal to 180 and 240, respectively. The tapering is attributed to an increase in the thin-film deposition rate on NW sidewalls compared to that of the vertical VLS growth. The dependence of thin film growth rate on V/III ratio has been previously reported for GaAs material in conventional thin film growth.13 Here, we observed the sidewall thin-film growth mechanism for InP NW growth, which has a similar V/III ratio dependence, as that shown in Ref. 13. The two tapered NW samples, D and E, appear to have slightly thinner tips than sample C. This might be due to the increased Au-catalyst diffusion into the NWs during the growth for the higher V/III ratio conditions.14 This phenomenon served as a secondary effect to make the NWs more tapered. Between the straight NW sample C 共V / III= 67兲 and the tapered NW sample D 共V / III= 180兲, two more V/III ratios, 90 and 120, were tested. While the V / III= 90 NW sample still looks straight, the V / III= 120 sample begins to show some taperness. The onset of NW tapering is then deduced as between V / III= 90 and 120.
92, 013121-1
© 2008 American Institute of Physics
013121-2
Chuang et al.
Appl. Phys. Lett. 92, 013121 共2008兲
FIG. 2. 共Color online兲 共T = 4 K兲 Representative -PL spectra for various V/III ratio growth conditions 共samples A–E兲 with excitation power equal to 1.8 W. Gray line indicates the 4 K InP bulk bandgap 共1.423 eV兲. Inset shows a single NW spectrum from a straight and narrow NW 共from sample C, V / III= 67兲, with a full width at half maximum 共FWHM兲 of 1.4 meV and blueshift of 178 meV from the bulk bandgap. For sample C, the high-energy side of the spectrum is spiky due to the contribution of these individual thin and straight NWs.
FIG. 1. 20° tilt FE-SEM images for five InP NW/共111兲Si samples, sample A, B, C, D, and E, with V/III ratios equal to 共a兲 15, 共b兲 30, 共c兲 67, 共d兲 180, and 共e兲 240, respectively. 共a兲 and 共b兲 represent insufficient V/III ratios and, hence, indium-rich growth conditions. 共c兲 represents a growth condition for yielding nontapered NWs. 共d兲 and 共e兲 represent the high V/III ratio growth conditions which yield tapered-NWs due to the stronger thin-film NW sidewall deposition.
-PL characterization was performed on all samples at 4 K and RT 共300 K兲 using a diode-pumped solid state laser at 532 nm focused to a ⬃2 m spot. Figure 2 shows the 4 K -PL spectrum comparison for the five samples in Fig. 1. The emission wavelengths of samples A and B are expected to be very close to the bulk InP bandgap since there are very few NWs on both samples but only some larger InP blobs, which might contribute to this PL emission. The PL peak of sample C with uniform NWs shows a blueshift of 40 meV from the bulk InP bandgap due to quantum confinement.7 Single-wire peaks from the narrowest NWs are visible on the high-energy side of the ensemble spectrum. For example, a peak with a 178 meV blueshift and linewidth of 1.4 meV was observed 共see inset of Fig. 2兲, which is the narrowest linewidth reported for a III-V NW.7,15 Sample D has a similar PL blueshift as sample C. However, the spiky features at the high energy side are not observed for sample D. This can be explained by the following. At low temperature, the carrier diffusion length is longer.16 Hence, carriers originally generated at the tips of the very narrow NWs can diffuse to the wider parts of the NWs where they see less quantum confinement 共smaller photon energy兲 and recombine there. For sample E, the carrier diffusion phenomenon is more pronounced for the strongly tapered NWs. As a consequence, for sample E, not only have the spiky features disappeared, the
PL peak shifts significantly to the redder side of sample C, at 20 meV blueshift of the InP bulk bandgap. The RT -PL spectra 共Fig. 3兲 show significant differences with those at 4 K. First of all, the fine features are no longer visible for sample C. Secondly, sample E has the same PL peak energy and linewidth as samples C and D, instead of being 20 meV redder at 4 K. This is particularly interesting since the NW base is wider than 100 nm and emission at the bulk bandgap is expected. We attribute this observation to luminescence from the narrow NW tips, where at RT confined carriers recombine radiatively before being able to diffuse to the wider base region. Third, samples A to D all show an extra 60 meV blueshift from the bulk InP bandgap compared to the amount of blueshift at 4 K 共see Fig. 2兲. The origin of this extra 60 meV blueshift is under further investigation. The PL intensity for both 4 K 共Fig. 2兲 and RT 共Fig. 3兲 measurements increases with V/III ratio. A similar trend is observed for InP epilayers,17,18 where it has been shown that high V/III ratios result in epilayers with lower defect densities.18,19 Figure 4共a兲 shows the peak energy position as a function of the excitation power at 4 K. When the excitation power is swept from 600 nW to 1.6 mW, over a three-decade power increase, the straight 共V / III= 67兲 and most tapered NWs
FIG. 3. 共Color online兲 RT PL spectrum for the various V/III ratios with excitation power equal to 400 W. Gray line indicates RT InP bulk bandgap 共1.340 eV兲. All samples are blueshifted with respect to the bulk bandgap. No fine features are seen due to thermal broadening.
013121-3
Appl. Phys. Lett. 92, 013121 共2008兲
Chuang et al.
mized to 67, nontapered InP NWs were grown. These NWs show a record narrow PL peak and weak excitation-power dependence, resembling features of ideal one dimensional structures. In our experiments, non-tapered InP NWs could be synthesized for V/III ratios ranging from 60 to 90, hence, offering a reasonable growth window. We also showed that the PL peak intensity could be increased with the increase of V/III ratio.
FIG. 4. 共Color online兲 T = 4 K 共a兲 -PL peak energy position versus excitation power for straight InP NWs 共sample C, V / III= 67兲 and strongly tapered InP NWs 共sample E, V / III= 240兲, respectively. Sample C shows a smaller wavelength dependence of only a 9 meV blueshift when excitation power is increased from 600 nW to 1.6 mW, while sample E shows a 40 meV blueshift. 共b兲 -PL peak intensity vs excitation power from 600 nW to 1.6 mW for various V/III ratios. The higher the V/III ratio, the larger the PL peak intensity is. Samples C and D show some intensity saturation at higher excitation powers while sample E does not.
共V / III= 240兲 show a blueshift of 9 and 40 meV, respectively. The small power-dependent blueshift for the straight NWs reflects the one-dimensional density of states nature of these NWs. The wavelength stability over a large excitation power range is important for critical devices such as lasers. The large blueshift for the tapered NWs is attributed to the effect of decreased carrier diffusion length at higher pumping, resulting in emission from the narrower part of the tapered NWs. The -PL peak intensity versus excitation power was also studied, as shown in Fig. 4共b兲. The higher V/III ratio, as described earlier, resulted in a larger PL peak intensity. For example, sample E is more than ten times brighter than sample C at 1.6 mW excitation. This effect might be also due to the larger NW volume since higher V/III ratio increases the sidewall deposition rate. The PL peak intensity shows some saturation at higher excitation powers for samples C and D, while there is no saturation observed for E within this range, likely due to the NW volume difference. For example, the NW volume of sample E is estimated as 2.5 ⫻ 10−15 cm3 and the NW volume ratio between samples E and D is ⬃6. The number of available electronic states is proportional to the volume which makes sample E remain unsaturated at higher pumping levels. In conclusion, we demonstrate the V/III ratio effects on both the shape and optical properties for InP NWs grown on 共111兲 Si substrates. The higher the V/III ratio is, the more tapered the NWs become. The PL intensity increases dramatically with the V/III ratio. When the V/III ratio is opti-
This work was supported by DARPA HR0011-04-10040 共CONSRT兲 and HP-CITRIS grants. The authors acknowledge the fellowship support from the UC Berkeley EECS Fellowship Program, NSF-IGERT Program, and NSF Graduate Research Fellowship Program. Linus C. Chuang and Michael Moewe contributed equally to this work. 1
M. Razeghi, The MOCVD Challenge, Volume 1: A Survey of GaInAsP-InP for Photonic and Electronic Applications 共Hilger, Bristol and Philadelphia, 1989兲. 2 Z. Mi, J. Yang, P. Bhattacharya, P. K. L. Chan, and K. P. Pipe, J. Vac. Sci. Technol. B 24, 1519 共2006兲. 3 G. Balakrishnan, A. Jallipalli, P. Rotella, S. Huang, A. Khoshakhlagh, A. Amtout, S. Krishna, L. R. Dawson, and D. L. Huffaker, IEEE J. Sel. Top. Quantum Electron. 12, 1636 共2006兲. 4 M. E. Groenert, A. J. Pitera, R. J. Ram, and E. A. Fitzgerald, J. Vac. Sci. Technol. B 21, 1064 共2003兲. 5 S. F. Fang, K. Adomi, S. Iyer, H. Morkoç, H. Zabel, C. Choi, and N. Otsuka, J. Appl. Phys. 68, R31 共1990兲. 6 R. M. Lum, J. K. Klingert, R. B. Bylsma, A. M. Glass, A. T. Macrander, T. D. Harris, and M. G. Lamont, J. Appl. Phys. 64, 6727 共1988兲. 7 L. C. Chuang, M. Moewe, C. Chase, N. P. Kobayashi, C. Chang-Hasnain, and S. Crankshaw, Appl. Phys. Lett. 90, 043115 共2007兲. 8 T. Mårtensson, C. P. T. Svensson, B. A. Wacaser, M. W. Larsson, W. Seifert, K. Deppert, A. Gustafsson, L. R. Wallenberg, and L. Samuelson, Nano Lett. 4, 1987 共2004兲. 9 E. P. A. M. Bakkers, M. T. Borgström, and M. A. Verheijen, MRS Bull. 32, 117 共2007兲. 10 S. S. Yi, G. Girolami, J. Amano, M. S. Islam, S. Sharma, T. I. Kamins, and I. Kimukin, Appl. Phys. Lett. 89, 133121 共2006兲. 11 N. P. Kobayashi, L. VJ, M. S. Islam, X. Li, J. Straznicky, S.-Y. Wang, R. S. Williams, and Y. Chen, Appl. Phys. Lett. 91, 113116 共2007兲. 12 R. S. Wagner and W. C. Ellis, Appl. Phys. Lett. 4, 89 共1964兲. 13 T. Kikkawa, H. Tanaka, and J. Komeno, J. Appl. Phys. 67, 7576 共1990兲. 14 J. B. Hannon, S. Kodambaka, F. M. Ross, and R. M. Tromp, Nature 共London兲 440, 69 共2006兲. 15 M. H. M. van Weert, O. Wunnicke, A. L. Roest, T. J. Eijkemans, A. Yu Silov, J. E. M. Haverkort, G. W. ‘t Hooft, and E. P. A. M. Bakkers, Appl. Phys. Lett. 88, 043109 共2006兲. 16 S. M. Sze, Physics of Semiconductor Devices 共Wiley, New York, 1981兲. 17 D. Kasemset, K. L. Hess, K. Mohammed, and J. L. Merz, J. Electron. Mater. 13, 655 共1984兲. 18 C. C. Hsu, J. S. Yuan, R. M. Cohen, and G. B. Stringfellow, J. Cryst. Growth 74, 535 共1986兲. 19 D. Keiper, R. Westphalen, and G. Landgren, J. Cryst. Growth 204, 256 共1999兲.
