Temperature-dependent Luminescence Properties of Digital-alloy In(Ga1−zAlz)As

Article information

. 2018;27(3):56-60
Publication date ( electronic ) : 2018 May 31
doi : https://doi.org/10.5757/ASCT.2018.27.3.56
aDepartment of Physics, Kangwon National University, Chuncheon 24341, Republic of Korea
bCenter for Opto-Electronic Convergence Systems, Korea Institute of Science and Technology, Seoul 02792, Republic of Korea
*Corresponding author: E-mail: myryu@kangwon.ac.kr
received : 2018 May 30, rev-recd : 2018 June 01, accepted : 2018 June 01.

Abstract

The optical properties of the digital-alloy (In0.53Ga0.47As)1−z/(In0.52Al0.48As)z grown by molecular beam epitaxy as a function of composition z (z = 0.4, 0.6, and 0.8) have been studied using temperature-dependent photoluminescence (PL) and time-resolved PL (TRPL) spectroscopy. As the composition z increases from 0.4 to 0.8, the PL peak energy of the digital-alloy In(Ga1−zAlz)As is blueshifted, which is explained by the enhanced quantization energy due to the reduced well width. The decrease in the PL intensity and the broaden FWHM with increasing z are interpreted as being due to the increased Al contents in the digital-alloy In(Ga1−zAlz)As because of the intermixing of Ga and Al in interface of InGaAs well and InAlAs barrier. The PL decay time at 10 K decreases with increasing z, which can be explained by the easier carrier escape from InGaAs wells due to the enhanced quantized energies because of the decreased InGaAs well width as z increases. The emission energy and luminescence properties of the digital-alloy (InGaAs)1−z/(InAlAs)z can be controlled by adjusting composition z.

I. Introduction

InGaAlAs quaternary alloys have been of much interest for optoelectronic devices operating at 1.0 to 1.65 mm [15]. The digital-alloy technique has been used to grow ternary or quaternary alloys by molecular-beam epitaxy (MBE). This technique overcomes intrinsic problems of the conventional MBE growth such as change of cell temperature, growth interruption, and additional source cell, etc. [6,7]. Defects such as alloy disorder due to an intermixing of Ga and Al atoms at the interfaces between InGaAs well and InAlAs barrier can be created in the digital-alloy InGaAlAs grown with (In0.53Ga0.47As)1−z/(In0.52Al0.48As)z short-period superlattices (SPSs) [8,9]. In order to reduce defects, which act as non-radiative recombination centers, many efforts such as rapid thermal annealing process and adjusting the thickness of InGaAs well and/or InAlAs barrier have been proposed [810].

Temperature- (T-) dependent optical processes give rise to various thermal dynamics of carriers in semiconductors such as carrier redistribution between energy states and/or carrier escape from active region to barrier [11,12] and can calculate the interband energy spacing by analyzing the T-dependent integrated photoluminescence (PL) intensity [13]. In this paper, the luminescence properties of digital-alloy In(Ga1−zAlz)As grown by MBE have been investigated by using a T-dependent PL as a function of composition z in digital-alloy (In0.53Ga0.47As)1−z/(In0.52Al0.48As)z (z = 0.4, 0.6, and 0.8). At low (high) temperature region, the thermal activation energies are extracted to be 3.87 (95.2), 8.89 (86.4), and 15.97 (70.5) meV for the sample with z = 0.4, 0.6, and 0.8, respectively. In order to better understand the recombination dynamics of the digital-alloy InGaAlAs, T-dependent time-resolved PL (TRPL) have been also carried out as a function of temperature.

II. Experiments

The digital-alloy In(Ga1−zAlz)As samples were grown on n-InP substrate by MBE system. A 100 nm-thick In0.52Al0.48As cladding layer was deposited on top of InP substrate at 510°C. The digital-alloy In(Ga1−zAlz)As layers, which are the lattice-matched to InP substrate were grown at 510°C and followed the growth of 25 nm-thick In0.52Al0.48As capping layers at the same temperature. The composition z of In(Ga1−zAlz)As layers are varied from 0.4 to 0.8. The thickness of In0.53Ga0.47As (In0.52Al0.48As) layer for z = 0.4, 0.6, and 0.8 are 9.8 (6.6), 6.6 (9.8), and 3.75 (15) Å, respectively. The interface and superlattice periodicity of digital-alloy InGaAlAs were clearly observed by cross-sectional transmission electron microscopy [14]. More detailed information for the growth conditions of the samples can be found in Refs. 6 and 7.

The PL and TRPL measurements were carried out using an FLS 920 spectrometer (Edinburgh Instruments). T-dependent luminescence measurements in the range of 10 to 300 K were performed using a helium closed-cycle cryostat. The PL and TRPL spectra of the digital-alloy In(Ga1−zAlz)As samples excited by using a 532-nm continuous-wave laser and a 634-nm pulsed diode laser (pulse width = 93 ps), respectively, were measured using a near-infrared photomultiplier tube detector. The PL decay profiles were measured by using a time-correlated single photon counting system.

III. Results and Discussion

Figure 1(a) shows the PL spectra of the digital-alloy (In0.53Ga0.47As)1−z/(In0.52Al0.48As)z samples measured at 10 K. The PL peak energy for the (In0.53Ga0.47As)1−z/(In0.52Al0.48As)z samples with composition z of 0.4, 0.6, and 0.8 is 1.117, 1.252, and 1.409 eV, respectively. As z increases from 0.4 to 0.8, the PL peak energy at 10 K is blueshifted about 292 meV. This blue-shift is attributed to the enhanced quantization energy caused by the decrease of well width from 9.8 to 3.75 Å with increasing z from 0.4 to 0.8, respectively. The weak PL peak around 1.441 eV in Figure 1(a) comes from the InP substrate [6]. As seen in Figure 1(a), the sample with z = 0.6 exhibits the strongest PL intensity, which indicates the enhanced carrier confinement into unit well regions due to reduced unit well width and increased InAlAs barrier’s thickness. The sample with z = 0.8 exhibits slightly reduced PL intensity compared with that of sample with z = 0.6. Figure 1(b) shows the full width at half maximum (FWHM) of the PL spectrum for the samples. The sample with z = 0.6 shows the narrowest FWHM (27.0 meV), while the sample with z = 0.8 exhibits much broad FWHM (36.4 meV). The decreased PL intensity and broadened FWHM of the sample with z = 0.8 compared with those of the sample with z = 0.6 can be attributed to the poor crystalline quality due to the high aluminum (Al) contents. With z increases from 0.6 to 0.8, the Al contents in the digital-alloy InGaAlAs are increased due to an intermixing of Ga and Al in interface of InGaAs wells and InAlAs barriers [15].

Figure 1

(a) PL spectra and (b) full width at half maximum of digital-alloy (In0.53Ga0.47As)1−z/(In0.52Al0.48As)z samples with composition z of 0.4, 0.6, and 0.8 measured at 10 K.

Figure 2(a) shows the integrated PL intensities of the digital-alloy samples as a function of inverse temperature. The T-dependent integrated PL intensities give the thermal activation energy of carriers by fitting them to an Arrhenius equation with I(T) = I0/[1+∑Aiexp(−Ei/kBT)], where I0 is the PL intensity at 0 K, Ai is the fitting coefficient, Ei is the thermal activation energy, and kB is the Boltzmann constant [16]. The activation energies E1 and E2 are calculated by fitting the PL data at low temperature (T ≤ 80 K) and high temperature range (80 K < T ≤ 300K), respectively. The calculated activation energies E1 and E2 are shown in inset of Figure 2(a) as a function of composition z in digital-alloy (In0.53Ga0.47As)1−z/(In0.52Al0.48As)z. The estimated E1 is 3.87, 8.89, and 16.0 meV and the E2 is 95.2, 86.4, and 70.5 meV for the sample with z = 0.4, 0.6, and 0.8, respectively. The E1 is attributed to the carrier confinement energy at potential fluctuations in digital-alloy InGaAlAs, and the E2 is ascribed to the energy difference between the ground states of the InGaAs well and InAlAs barrier. The increase of E1 with increasing z from 0.4 to 0.8 is due to larger potential fluctuations caused by higher Al contents.

Figure 2

(a) Temperature-dependent integrated PL intensities of the digital-alloy InGaAlAs as a function of inverse temperature and the schematic band diagrams for the sample with (b) z = 0.4 and (c) z = 0.8. The inset in (a) shows the activation energies E1 and E2 as a function of z estimated from the temperature-dependent integrated PL intensities.

The schematic band diagrams of the digital-alloy InGaAlAs samples with z = 0.4 and 0.8 are represented in Figure 2(b) and 2(c), respectively. The bandgap energies of In0.53Ga0.47As and In0.52Al0.48As layers at 300 K are 0.75 and 1.45 eV, respectively, and the valence-band and conduction-band offset are 0.2 and 0.5 eV, respectively [17]. The ground state (GS) transition energy and activation energy of the sample with z = 0.4 are 1.039 eV and 95 meV, respectively, shown in Figure 2(b) while the GS energy (1.335 eV) and activation energy (70 meV) of the sample with z = 0.8 are shown in Figure 2(c). These transition and carrier activation energies were obtained by the T-dependent PL measurement. Note that the energy separation between the GS of the InGaAs well and continuum states of the InAlAs barrier in the conduction band decreases from 316 to 45 meV with increasing z from 0.4 to 0.8, respectively, as shown in Figure 2(b) and 2(c). As z increases from 0.4 to 0.8, the InGaAs well width decreases, and thus the quantized energy in the well increases with increasing z. As a result, the carrier activation energy E2 decreases.

T-dependent PL peak energies of the digital-alloy InGaAlAs samples are calculated by fitting each set of PL spectra with Gaussian or exponentially modified Gaussian peak, and the results are shown in Figure 3(a). The solid lines are calculated using Varshni’s type equation with Eg(T) = Eg(0)−αT2/(T+β), where α = 4.0 × 10−4 eV/K, and β = 226 K, and Eg(0) is used the PL peak energy of each sample measured at 10 K [18,19]. The calculated curves for the InGaAlAs samples with z = 0.4 and 0.6 agree well with the PL data as shown in Figure 3(a) while the PL peak energies of the sample with z = 0.8 also agrees well with the calculated curve except for the intermediate temperature range (100 K < T < 250 K). The fast bandgap shrinkage of the sample with z = 0.8 is ascribe to larger potential fluctuations and thermally easier escape of carriers from InGaAs wells due to the shallow potential barriers as shown in Figure 2(c). Figure 3(b) displays the FWHM of the digital-alloy InGaAlAs samples as a function of temperature. The FWHM of the samples increases with increasing temperature. The sample with z = 0.8 exhibits much broader FWHM compared with the other two samples, which is attributed to larger potential fluctuations and weak carrier confinement.

Figure 3

(a) PL peak energies and (b) FWHM of the digital-alloy InGaAlAs samples as a function of temperature. The solid lines in (a) are estimated using Varshni’s type equation with Eg(T) = Eg(0)−4.0 × 10−4T2/(T+226), where Eg(0) is the PL peak energy of each samples at 10 K.

T-dependent TRPL measurements have been performed to explore the carrier dynamics in the digital-alloy InGaAlAs samples. The PL decay curves measured at 10 and 140 K are plotted in Figure 4(a) and (b), respectively. As the composition z increases from 0.4 to 0.8, the PL decay becomes faster at 10 K while the decay becomes slower at 140 K. The decrease of PL decay time with increasing z at 10 K can be explained by the easier carrier escape from the InGaAs well due to the enhanced quantized energies as shown in Figure 2(b) and 2(c). The PL decay curves are fitted by a two-exponential function, I(t) = A1exp(−t/τ1) + A2exp(−t/τ2), in order to estimate decay times. τ1 and τ2 are the fast and slow decay times, respectively, and A1 and A2 are the contribution of the corresponding parts to the total PL intensity. Table 1 shows the estimated PL decay times and amplitudes of the samples measured at 10 and 140 K. As z increases from 0.4 to 0.8, the fast decay time τ1 at 10 K decreases from 1.20 to 0.88 ns while the τ1 at 140 K increases from 0.71 to 1.32. On the other hand, the decay time τ2 for all samples are pretty similar as shown in Table 1. In previous emission-photon energy-dependent decay times and excitation-power-dependent PL studies, the τ1 and τ2 are related to the exciton transition and donor-acceptor pair transition, respectively [10].

Figure 4

The PL decay curves of the digital-alloy InGaAlAs samples measured at (a) 10 K and (b) 140 K taken at the PL peak energy for each sample.

Estimated PL decay times (τ1, τ2) for the digital-alloy InGaAlAs samples with composition z of 0.4, 0.6, and 0.8. The PL decays at 10 and 140 K were measured at the PL peak energy at each temperature.

The decay times of the digital-alloy samples were measured at the PL peak as a function of temperature shown in Figure 5(a). The PL decay times of the samples with z = 0.4 and 0.6 increase up to 30 and 60 K, respectively, and then decrease with increasing temperature further, while the PL decay of the sample with z = 0.8 increases up to 70 K and then is constant up to 140 K. T-dependent radiative (τR) and nonradiative (τNR) lifetimes can be estimated by using the relationships of η(T) = τPL(T)/τR ≈ I(T)/I0, and 1/τPL(T) = 1/τR(T) + 1/τNR(T), where η is the internal quantum efficiency, I is the integrated PL intensity, and tPL the PL decay time. η(10 K) = 1 are assumed by using I(10 K) = I0. Figure 5(b) shows the T-dependent radiative decay time τR and the T-dependent nonradiative decay time τNR are plotted in the inset of Figure 5(b). The τR of the samples with z = 0.6 and 0.8 increase continuously from 10 to 140 K while the τR of the sample with z = 0.4 increases up to 40 K and then shorten at higher temperatures. On the other hand, the τNR for all three samples decreases steadily with increasing temperature, a typical property of semiconductors [2022]. As seen in Figure 5, the τR at low temperatures (T ≤ 30, 60, and 70 K for the samples with z = 0.4, 0.6, and 0.8, respectively) well represents the PL decay times while at higher temperatures the PL decay times are mainly described by the τNR. The increase in the PL decay time at low temperatures are attributed to the thermally-activated carriers from potential fluctuations.

Figure 5

Temperature-dependent (a) PL decay time τ1 and (b) radiative lifetime τR of the digital-alloy InGaAlAs samples. The inset in (b) shows the temperature-dependent nonradiative lifetime τNR of the digital-alloy InGaAlAs samples. Each PL decay times are obtained at the PL peak energy for each temperature.

IV. Conclusions

The temperature-dependent PL and TRPL spectra of the digital-alloy InGaAlAs grown with (In0.53Ga0.47As)1−z/(In0.52Al0.48As)z short-period superlattices as a function of composition z (z = 0.4, 0.6, and 0.8) have been investigated at temperature ranging from 10–300 K. As z increases from 0.4 to 0.8, the activation energy E1 estimated at low temperatures (T ≤ 80 K) increases while the E2 at high temperatures (80 K < T ≤ 300 K) reduces. The increased E1 is attributed to larger potential fluctuations due to high Al contents with increasing z, and the reduced E2 is ascribed to the shallow potential barrier caused by the increased quantized energy due to the narrow quantum well width as z increases. The emission energy and luminescence properties are controlled with adjusting the well and barrier widths of digital-alloy InGaAlAs.

Acknowledgments

This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology (NRF-2017R1A2B4003744). Time-resolved photoluminescence measurements were performed at the Central Lab of Kangwon National University.

References

1. Cinguino P, Genova F, Rigo C, Cacciatore C, Stano A. Appl Phys Lett 50:15151987;
2. Kollakowski St, Böttcher EH, Strittmatter A, Bimberg D. Electron lett 34:5871998;
3. McDougall SD, Kowalski OP, Marsh JH, Aitchison JS. IEEE Photon Technol Lett 11:15571999;
4. Gready D, Eisenstein G, Ivanov V, Gilfert C, Schnabel F, Rippien A, Reithmaier JP, Bornholdt C. IEEE Photon Technol Lett 26:112014;
5. Rudno-Rudziński W, Syperek M, Maryński A, Andrzejewski J, Misiewicz J, Bauer S, Sichkovskyi VI, Reithmaier JP, Schowalter M, Gerken B, Rosenauer A, Sęk G. Phys Status Solidi A 215:17004552018;
6. Song JD, Heo DC, Han IK, Kim JM, Lee YT, Park S-H. Appl Phys Lett 84:8732004;
7. Woo JT, Kim JH, Kim TW, Song JD, Park YJ. Phys Rev B 72:2053202005;
8. Song JD, Yu JS, Kim JM, Bae SJ, Lee YT. Appl Phys Lett 80:46502002;
9. Cho I-W, Byun HR, Ryu M-Y, Song JD. J Korean Vac Soc 22:3212013;
10. Cho I-W, Ryu M-Y, Song JD. Thin Solid Films 636:3922017;
11. Zhang YC, Huang CJ, Liu FQ, Xu B, Wu J, Chen YH, Ding D, Jiang WH, Ye XL, Wang ZG. J Appl Phys 90:19732001;
12. Wu Y-F, Lee JC, Nee T-E, Wang J-C. J Lumin 131:12672011;
13. Liu W-S, Chu T-F, Huang T-H. Optics Express 22:309632014;
14. Heo D, Song JD, Han IK, Choi WJ, Lee YT. IEEE J Quantum Electron 49:242013;
15. Offermans P, Koenraad PM, Wolter JH, Song JD, Kim JM, Bae SJ, Lee YT. Appl Phys Lett 82:11912003;
16. Leroux M, Grandjean N, Beaumont B, Nataf G, Semond F, Massies J, Gibart P. J Appl Phys 86:37211999;
17. Heo D, Kim G-H, Song JD. J Korean Phys Soc 69:12252016;
18. Varshni YP. Physica 34:1491967;
19. Gaskill DK, Bottka N, Aina L, Mattingly M. Appl Phys Lett 56:12691990;
20. Kim HY, Ryu M-Y, Kim JS. J Lumin 132:17592012;
21. Kong LM, Cai JF, Wu ZY, Gong Z, Niu ZC, Feng ZC. Thin Solid Films 498:1882006;
22. Song JH, Lee JW, Yu PW, Ryu M-Y, Zhang J, Kuokstis E, Yang JW, Asif Khan M. Solid State Commun 127:6612003;

Article information Continued

Figure 1

(a) PL spectra and (b) full width at half maximum of digital-alloy (In0.53Ga0.47As)1−z/(In0.52Al0.48As)z samples with composition z of 0.4, 0.6, and 0.8 measured at 10 K.

Figure 2

(a) Temperature-dependent integrated PL intensities of the digital-alloy InGaAlAs as a function of inverse temperature and the schematic band diagrams for the sample with (b) z = 0.4 and (c) z = 0.8. The inset in (a) shows the activation energies E1 and E2 as a function of z estimated from the temperature-dependent integrated PL intensities.

Figure 3

(a) PL peak energies and (b) FWHM of the digital-alloy InGaAlAs samples as a function of temperature. The solid lines in (a) are estimated using Varshni’s type equation with Eg(T) = Eg(0)−4.0 × 10−4T2/(T+226), where Eg(0) is the PL peak energy of each samples at 10 K.

Figure 4

The PL decay curves of the digital-alloy InGaAlAs samples measured at (a) 10 K and (b) 140 K taken at the PL peak energy for each sample.

Figure 5

Temperature-dependent (a) PL decay time τ1 and (b) radiative lifetime τR of the digital-alloy InGaAlAs samples. The inset in (b) shows the temperature-dependent nonradiative lifetime τNR of the digital-alloy InGaAlAs samples. Each PL decay times are obtained at the PL peak energy for each temperature.

Table 1

Estimated PL decay times (τ1, τ2) for the digital-alloy InGaAlAs samples with composition z of 0.4, 0.6, and 0.8. The PL decays at 10 and 140 K were measured at the PL peak energy at each temperature.

composition z τ1 (ns) τ2 (ns)
10 K 140 K 10 K 140 K
0.4 1.20 0.71 6.20 7.40
0.6 1.01 0.83 5.90 6.30
0.8 0.88 13.32 5.70 6.01