Wavelength-Tunable InGaAs-Capped Quantum-Dot Infrared Photodetectors

IEEE PHOTONICS TECHNOLOGY LETTERS, VOL. 22, NO. 4, FEBRUARY 15, 2010 227

Wavelength-Tunable InGaAs-Capped Quantum-Dot

Infrared Photodetectors

Wei-Hsun Lin, Student Member, IEEE, Chi-Che Tseng, Student Member, IEEE, Kuang-Ping Chao,

Shu-Cheng Mai, Shih-Yen Lin, Member, IEEE, and Meng-Chyi Wu, Senior Member, IEEE

Abstract—Quantum-dot infrared photodetectors (QDIPs) with

InGaAs capping layers are investigated. Compared with the stan-dard QDIP with 2.5-mono-layer (ML) InAs QDs, the detection wavelength is shifted from 6 to 7.9 m for an 8-nm InGaAs-capped QDIP. By decreasing the QD coverage from 2.5 to 2.0 ML, an even longer detection wavelength 10.4 m is observed, which is attributed to the higher energy levels of the QD excited states resulted from the smaller QDs. By further increasing the capping layer thickness to 12 nm, longer detection wavelengths with broad response 10–18 m is observed for the InGaAs-capped QDIP.

Index Terms—Quantum-dot infrared photodetectors (QDIPs).

I. INTRODUCTION

C

photodetectors (QWIPs), advantages of quantum-dot in-frared photodetectors (QDIPs) like high-temperature operation and normal incident absorption have been widely investigated [1]–[5]. However, for most of the QDIPs, the detection wave-lengths are limited in the midwavelength infrared (MWIR, 3–5 m) range. To improve this disadvantage, reports regarding to the InAs QDs embedded in InGaAs quantum-well structures (DWELL) have been proposed [6]–[9]. The devices have exhib-ited the long-wavelength infrared [(LWIR) 8–12 m] detection. However, for such devices, an additional InGaAs layer prior to the QD growth is always required to achieve the devices operated at longer detection wavelengths [10]. With the more complicated structures, device parameter optimization such as underlying InGaAs thickness and growth conditions would be required. Therefore, a simpler QDIP structure with tunable detection wavelengths from MWIR to LWIR ranges would be advantageous for the application of multicolor detection.

In this letter, ten-period QDIPs with InGaAs capping layers are investigated. Compared with the detection wavelength 6 m of the standard QDIP with 2.5-mono-layer (ML) InAs QDs,

Manuscript received September 11, 2009; revised November 19, 2009; ac-cepted November 20, 2009. First published January 08, 2010; current version published January 27, 2010. This work was supported in part by the National Science Council, Taiwan under Grant NSC 98-2221-E-001-001.

W.-H. Lin, K.-P. Chao, S.-C. Mai, and M.-C. Wu are with the Institute of Electronics Engineering, National Tsing Hua University, Hsinchu 300, Taiwan (e-mail: [email protected]; [email protected]; [email protected]; [email protected]).

C.-C. Tseng is with the Institute of Photoics Technologies, National Tsing Hua University, Hsinchu 300, Taiwan (e-mail: [email protected]).

S.-Y. Lin is with the Research Center for Applied Sciences, Academia Sinica, Taipei 11529, Taiwan, and also with the Department of Photonics, National Chiao-Tung University, Hsinchu 300, Taiwan, and also with the Institute of Optoelectronic Sciences, National Taiwan Ocean University, Keelung 20224, Taiwan (e-mail: [email protected]).

Digital Object Identifier 10.1109/LPT.2009.2037727

the insertion of the additional 8-nm In Ga As layer of a InGaAs-capped QDIP would result in a detection wavelength shift to 7.9 m. The results suggest that the dominant transi-tion mechanism in the InGaAs-capped QDIPs is from the QD excited state to the InGaAs QW ground state. By decreasing the QD coverage from 2.5 to 2.0 ML, a longer detection wave-length 10.4 m is observed. The phenomenon is attributed to the higher QD excited states resulted from reduced QD sizes such that a reduced energy difference between the QD excited state and the InGaAs QW ground state is obtained. By further increasing the capping layer thickness to 12 nm, even longer de-tection wavelengths ranging from 10 to 18 m are observed for the InGaAs-capped QDIP.

II. EXPERIMENTS

The samples investigated in this letter are grown on (100)-ori-ented semi-insulated GaAs substrates using a Riber Compact 21 solid source molecular beam epitaxy (MBE) system. For all the samples, 600- and 300-nm GaAs layer with n-type doping cm are grown as bottom and top contact layers. Four samples with ten-period (a) 2.5-ML InAs QDs/50-nm GaAs,

(b) 8-nm In Ga As/2.5-ML InAs QDs/42-nm GaAs, (c)

8-nm In Ga As/2.0-ML InAs QDs/42-nm GaAs, and (d)

12-nm In Ga As /2.0-ML InAs QDs/38-nm GaAs are

prepared, which are denoted as Samples A, B, C, and D, respec-tively. Standard photolithography and chemical wet etching are adopted to fabricate devices with 100 100 m mesas. An edge-coupling scheme is adopted for the spectral response measurements of all the devices. The measurement system for the spectral response consists of a Perkin Elmer Spectrum 100 Fourier Transformation infrared spectroscopy coupling with a Janis cryostat and a current preamplifier [4].

III. RESULTS ANDDISCUSSION

Normalized 10 K spectral responses of devices A and B op-erated at 2.0 V are shown in Fig. 1(a). As show in the figure, peak responses of 6 and 7.9 m are observed for Samples A and B, respectively. Significant detection wavelength red shift to the LWIR range is observed for Device B with an additional InGaAs capping layer. Also shown in the figure is the narrower response full-width at half-maximum (FWHM) of Device B. To explain the phenomenon, the 10 K spectral responses of Device B at 0.6, 1.0, and 1.4 V are shown in Fig. 1(b). As shown in the figure, the peak responses of the device would shift from MWIR to LWIR range with increasing applied voltages. The results suggest that there are two transition mechanisms involved in the spectral re-sponse measurements of Device B. At low applied voltages,

228 IEEE PHOTONICS TECHNOLOGY LETTERS, VOL. 22, NO. 4, FEBRUARY 15, 2010

Fig. 1. (a) Normalized 10 K spectral responses of Devices A and B at 2.0 V and (b) the 10 K spectral responses of Device B at 0.6, 1.0, and 1.4 V, respectively.

Fig. 2. The 10 K PLE spectrum of Sample B with its PL peak energy 1.054 eV as the detection wavelength.

dominant transition is the one at the MWIR range while the LWIR response would dominate at higher applied voltages.

To further investigate the transition mechanisms of Device B, the 10 K photoluminescence excitation (PLE) spectrum of Sample B is shown in Fig. 2 with its PL peak energy 1.054 eV as the detection wavelength. Four peaks are observed in the spectrum, which are the first excited state of the QDs , the second excited state of the QDs , the QW ground state in the InGaAs layer , and the wetting layer state [11]. As shown in the figure, the energy differences of

and are 0.236 and 0.143 eV

(5.25 and 8.67 m), respectively. The values are close to the observed dominant response peaks 6 and 7.9 m at low and higher applied voltages, as shown in Fig. 1(b). The results sug-gest that the LWIR response of Device B is resulted from the

transition between and states [11]. With

in-creasing applied voltage over Device B, the dominant transition

Fig. 3. (a) The 10 K spectral response of Device C at 2.0 V and (b) the 10 K PLE spectrum of Sample C with its PL peak energy 1.151 eV as the detection wavelength.

would change from to , which

is attributed to the lower energy of the state compared with the state such that the tunneling probability of the photo-excited electrons at the state is lower at low ap-plied voltages.

Although LWIR responses are already observed for Device B, devices with even longer detection wavelengths are still re-quired. In this case, an efficient approach to decrease the energy difference between the QD excited state and the InGaAs QW ground state is to push the QD excited state to higher energy levels. To achieve this goal, smaller QDs are required. There-fore, Sample C with lower InAs coverage 2.0 ML is prepared. The 10 K spectral response of Device C at 2.0 V is shown in Fig. 3(a). As shown in the figure, a peak response at 10.4 m with responsivity of 1.17 A/W at 2.0 V is observed, while a much weaker peak is observed at 5.7 m. The high responsivity of the device at 10.4 m suggests optimized growth conditions are adopted for the InGaAs-capped structures. Without an ad-ditionhal InGaAs layer grown prior QD growth, the device has exhibited a long detection wavelength up to 10.4 m. The 10 K PLE spectrum of Sample C is shown in Fig. 3(b) with its PL peak energy 1.151 eV as the detection wavelength. As shown in the figure, similar peak positions are observed for the energy levels and of Samples B and C, which suggest that the and states would not change significantly with the QD coverage. On the other hand, the QD first excited state has been raised up to 1.235 eV. In this case, the dominant tran-sitions responsible for Device C should be

and transitions. The energy differences for the

two transitions are 0.122 and 0.214 (10.16 and 5.79 m), re-spectively. As shown in Fig. 3(a), both transitions are observed

LIN et al.: WAVELENGTH-TUNABLE InGaAs-CAPPED QDIPs 229

Fig. 4. (a) The 10 K spectral response of Device D at 2.0 V and (b) the 10 K PLE spectrum of Sample D with its PL peak energy 1.114 eV as the detection wavelength.

while the dominant one at 2.0 V is the

tran-sition, which is similar to the behaviors of Device B at high ap-plied voltages.

To push the detection wavelengths beyond 10.4 m, one ef-fective approach is to grow a thicker InGaAs capping layer such that a lower state would be obtained. The 10 K spec-tral response of Device D at 2.0 V is shown in Fig. 4(a). As shown in the figure, two major responses are observed with one at 7 m and the other with broad detection wavelengths between 10–18 m. The 10 K PLE spectrum of Sample D is shown in Fig. 4(b) with its PL peak energy 1.114 eV as the detection wavelength. As shown in the figure, the energy differences of

and are 0.069 and 0.137 eV

(18 and 9.1 m), respectively. Therefore, the broad response between 10 and 18 m of Device D should be attributed to the

summation of and transitions.

As for the peak at 7 m, one possible mechanism responsible for the response is the summation of transitions between

to or . As shown in Fig. 4(b), the energy

differ-ences are 0.143 and 0.211 eV (8.7 and 5.9 m), respectively. However, since no similar responses are observed for Devices B and C, further investigations are required to confirm the attri-bution.

IV. CONCLUSION

Tunable detection wavelengths from MWIR to LWIR ranges of InGaAs-capped QDIPs have been achieved. To achieve LWIR responses, two different approaches of (a) uprising the QD ex-cited state by reducing the QD sizes and (b) lowering the In-GaAs state by increasing the capping layer thickness are both proved to be effective. The detection wavelengths of InGaAs-capped QDIPs could be easily tuned from 6 to 10 m or even longer wavelength. The high responsivities of the devices have also demonstrated that the simpler InGaAs-capped QD struc-tures can be easily prepared without severe strain accumulation. The results are advantageous for the development of multicolor QDIP FPAs at both MWIR and LWIR ranges.

REFERENCES

[1] S. F. Tang, S. Y. Lin, and S. C. Lee, “Near-room-temperature opera-tion of an InAs/GaAs quantum-dot infrared photodetector,” Appl. Phys.

Lett., vol. 78, pp. 2428–2430, Apr. 2001.

[2] L. Jiang, S. S. Li, N. T. Yeh, J. I. Chyi, C. E. Ross, and K. S. Jones, “In Ga As/GaAs quantum-dot infrared photodetector with operating temperature up to 260 K,” Appl. Phys. Lett., vol. 82, pp. 1986–1988, Mar. 2003.

[3] P. Bhattacharya, X. H. Su, S. Chakrabarti, G. Ariyawansa, and A. G. U. Perera, “Characteristics of a tunneling quantum-dot infrared pho-todetector operating at room temperature,” Appl. Phys. Lett., vol. 86, pp. 191106–191108, May 2005.

[4] S. T. Chou, M. C. Wu, S. Y. Lin, and J. Y. Chi, “The influence of doping density on the normal incident absorption of quantum-dot in-frared photodetectors,” Appl. Phys. Lett., vol. 88, pp. 173511–173513, Apr. 2006.

[5] S. F. Tang, C. D. Chiang, P. K. Weng, Y. T. Gau, J. J. Ruo, S. T. Yang, C. C. Shih, S. Y. Lin, and S. C. Lee, “High-temperature operation normal incident 2562 256 InAs/GaAs quantum dot infrared photode-tector focal plane array,” IEEE Photon. Technol. Lett., vol. 18, no. 8, pp. 986–988, Apr. 15, 2006.

[6] S. D. Gunapala, S. V. Bandara, C. J. Hill, D. Z. Ting, J. K. Liu, S. B. Rafol, E. R. Blazejewski, J. M. Mumolo, S. A. Keo, S. Krishna, Y. C. Chang, and C. A. Shott, “Long-wavelength infrared (LWIR) quantum dot infrared photodetector (QDIP) focal plane array,” in Proc. SPIE, 2006, vol. 6206, p. 62060J.

[7] S. Krishna, D. Forman, S. Annamalai, P. Dowd, P. Varangis, T. Tumo-lillo, A. Gray, J. Zilko, K. Sun, M. Liu, J. Campbell, and D. Carothers, “Demonstration of a 3202 256 two-color focal plane array using InAs/ InGaAs quantum dots in well detectors,” Appl. Phys. Lett., vol. 86, pp. 193501-1–193501-3, May 2005.

[8] G. Ariyawansa, A. G. Unil Perera, G. S. Raghavan, G. Von Winckel, A. Stintz, and S. Krishna, “Effect of well width on three-color quantum dots-in-a-well infrared detectors,” IEEE Photon. Technol. Lett., vol. 17, no. 5, pp. 1064–1066, May 2005.

[9] L. Höglund, P. O. Holtz, H. Pettersson, C. Asplund, Q. Wang, H. Malm, S. Almqvist, E. Petrini, and J. Y. Andersson, Appl. Phys. Lett., vol. 93, pp. 203512-1–203512-3, Nov. 2008.

[10] S. Raghavan, D. Forman, P. Hill, N. R. Weisse-Bernstein, G. von Winckel, P. Rotella, S. Krishna, S. W. Kennerly, and J. W. Little, “Normal-incidence InAs/In Ga As quantum dots-in-a-well detector operating in the long-wave infrared atmospheric window (8–12m),” J. Appl. Phys., vol. 96, pp. 1036–1039, Jul. 2004. [11] W. H. Lin, C. C. Tseng, K. P. Chao, S. C. Mai, S. Y. Lin, and M. C. Wu,

“InGaAs-capped InAs–GaAs quantum-dot infrared photodetectors op-erating in the long-wavelength infrared range,” IEEE Photon. Technol.