Infrared photoresponse of GeSn/n-Ge heterojunctions grown by molecular beam epitaxy Sangcheol Kim, Nupur Bhargava,* Jay Gupta, Matthew Coppinger, and James Kolodzey Department of Electrical and Computer Engineering, University of Delaware, Newark, Delaware 19716, USA *
[email protected] Abstract: Heterojunction devices of Ge1-xSnx / n-Ge were grown by solid source molecular beam epitaxy (MBE), and the mid-infrared (IR) photocurrent response was measured. With increasing Sn composition from 4% to 12%, the photocurrent spectra became red-shifted, suggesting that the bandgap of Ge1-xSnx alloys was lowered compared to pure Ge. At a temperature of 100 K, the wavelengths of peak photocurrent were shifted from 1.42 µm for pure Ge (0% Sn) to 2.0 µm for 12% Sn. The bias dependence of the device response showed that the optimum reverse bias was > 0.5 volts for saturated photocurrent. The responsivity of the Ge 1-xSnx devices was estimated to be 0.17 A/W for 4% Sn. These results suggest that Ge1-xSnx photodetectors may have practical applications in the near/mid IR wavelength regime. ©2014 Optical Society of America OCIS codes: (040.5160) Photodetectors; (060.4510) Optical communications; (200.4650) Optical interconnects.
References and links 1. 2. 3. 4.
5. 6. 7. 8.
9. 10.
11. 12. 13.
J. Werner, M. Oehme, M. Schmid, M. Kaschel, A. Schirmer, E. Kasper, and J. Schulze, “Germanium-tin p-i-n photodetectors integrated on silicon grown by molecular beam epitaxy,” Appl. Phys. Lett. 98(6), 061108 (2011). S. Su, B. Cheng, C. Xue, W. Wang, Q. Cao, H. Xue, W. Hu, G. Zhang, Y. Zuo, and Q. Wang, “GeSn p-i-n photodetector for all telecommunication bands detection,” Opt. Express 19(7), 6400–6405 (2011). R. Soref, “Mid-infrared photonics in silicon and germanium,” Nat. Photonics 4(8), 495–497 (2010). S. Takeuchi, A. Sakai, K. Yamamoto, O. Nakatsuka, M. Ogawa, and S. Zaima, “Growth and structure evaluation of strain-relaxed Ge1xSnx buffer layers grown on various types of substrates,” Semicond. Sci. Technol. 22(1), S231–S235 (2007). J. Michel, J. Liu, and L. Kimerling, “High-performance Ge-on-Si photodetectors,” Nat. Photonics 4(8), 527–534 (2010). P. Moontragoon, Z. Ikonic, and P. Harrison, “Band structure calculations of Si–Ge–Sn alloys: achieving direct band gap materials,” Semicond. Sci. Technol. 22(7), 742–748 (2007). D. Jenkins and J. Dow, “Electronic properties of metastable GexSn1-x alloys,” Phys. Rev. B 36(15), 7994–8000 (1987). R. Roucka, J. Mathews, C. Weng, R. Beeler, J. Tolle, J. Menéndez, and J. Kouvetakis, “Development of high performance near IR photodiodes: A novel chemistry based approach to Ge-Sn devices integrated on silicon,” IEEE J. Quantum Electron. 47(2), 213–222 (2011). G. Chang, S. Chang, and S. Chuang, “Strain-Balanced GezSn1-z-SixGeySn1-x-y Multiple-Quantum-Well Lasers,” IEEE J. Quantum Electron. 46(12), 1813–1820 (2010). H. Tseng, K. Wu, H. Li, V. Mashanov, H. Cheng, G. Sun, and R. Soref, “Mid-infrared electroluminescence from a Ge/Ge0.922 Sn0.078/Ge double heterostructure p-i-n diode on a Si substrate,” Appl. Phys. Lett. 102(18), 182106 (2013). J. Gupta, N. Bhargava, S. Kim, T. Adam, and J. Kolodzey, “Infrared electroluminescence from GeSn heterojunction diodes grown by molecular beam epitaxy,” Appl. Phys. Lett. 102(25), 251117 (2013). N. Bhargava, M. Coppinger, J. Gupta, L. Wielunski, and J. Kolodzey, “Lattice constant and substitutional composition of GeSn alloys grown by Molecular Beam Epitaxy,” Appl. Phys. Lett. 103(4), 041908 (2013). S. Kim, J. Gupta, N. Bhargava, M. Coppinger, and J. Kolodzey, “Current-Voltage Characteristics of Ge/Sn Heterojunction Diodes Grown by Molecular Beam Epitaxy,” IEEE Electron Device Lett. 34(10), 1217–1219 (2013).
#202262 - $15.00 USD Received 29 Nov 2013; revised 25 Mar 2014; accepted 21 Apr 2014; published 1 May 2014 (C) 2014 OSA 5 May 2014 | Vol. 22, No. 9 | DOI:10.1364/OE.22.011029 | OPTICS EXPRESS 11029
14. N. Bhargava, J. Gupta, T. Adam, and J. Kolodzey, “Structural Properties of Boron-Doped Germanium-Tin Alloys Grown by Molecular Beam Epitaxy,” J. Electron. Mater. 43(4), 931–937 (2014). 15. M. Coppinger, J. Hart, N. Bhargava, S. Kim, and J. Kolodzey, “Photoconductivity of Germanium Tin Alloys grown by Molecular Beam Epitaxy,” Appl. Phys. Lett. 102(14), 141101 (2013).
1. Introduction In recent years there has been an increased interest in GeSn alloys due to possible applications in optoelectronics, microelectronics and telecommunications [1, 2]. Previous studies suggested that the incorporation of Sn in Ge lowers the bandgap energy, which is attractive in mid-infrared (IR) applications that require band gaps lower than about 0.6 eV [3]. Some useful properties of GeSn alloys such as the possibility of indirect-to-direct bandgap crossover at Sn composition near 6%, compatibility with silicon processing technology, and bandgap control by strain [4,5] have drawn great attention to the development of GeSn-based mid-IR optoelectronic devices [6–11]. With recent research efforts to understand the properties of GeSn alloys, the characteristics of optoelectronic devices such as infrared detectors and emitters are becoming better understood [11–14]. This report examines the properties of Ge1-xSnx / n-Ge heterojunction devices as infrared light detectors at various compositions and bias conditions. 2. Experimental detail The undoped Ge1-xSnx alloys were grown in an EPI (Veeco) 620 molecular beam epitaxy (MBE) system on 76-mm diameter (100) oriented n-type Ge substrates with resistivities of 0.005 – 0.02 Ω-cm. Ultra high purity Ge (triple zone refined) and Sn (6N purity) were thermally evaporated from effusion cells with pyrolytic boron nitride (pBN) crucibles. The ntype doping used a custom effusion cell having a pBN baffle for preferential phosphorus evaporation from solid GaP pellets. Prior to the undoped Ge 1-xSnx alloy growths, n-type Ge buffer layers were grown and doped to concentrations of about 3x10 18 cm3 at substrate temperatures of 420 °C for coherent epitaxy. Three different samples of undoped Ge1-xSnx alloys with different Sn compositions (sample A: 4%, sample B: 9.8%, and sample C: 12%) and pure Ge (sample D) without Sn were prepared. All Ge 1-xSnx alloys samples were grown using the identical Sn deposition rate of ~0.80 Å/min while varying the Ge deposition rate by adjusting the Ge effusion cell temperature, to change the composition. The compositions of Sn were obtained through Rutherford back-scattering spectrometry (RBS). Channeling RBS measurements indicated that most of the Sn was substitutional in the Ge lattice [12]. The thickness of the Ge1-xSnx layers were measured by stylus profilometry to be ~150, 80, and 55 nm for samples A (SGC 591), B (SGC 599), and C (SGC 609), respectively and the thickness of epitaxial undoped Ge was ~150 nm for sample D. Standard optical photolithography was used to fabricate the Ge 1-xSnx / n-Ge heterojunction devices with structure as shown in Fig. 1(b). Evaporated metals of Al (300 nm thick) for the Ge1-xSnx topside, and Ti/Ag (30/300 nm thick) for the bottom n-Ge substrate side were used as electrical contacts. To prepare devices for measurements, the metalized samples were diced using a diamond saw into several chips with a top surface area of either 2mm x 2mm or 1mm x 1mm. After device fabrication, current-voltage measurements were carried out with a Keithley 2400 Source Meter, with the results described below. Photocurrent (PC) measurements were carried out using Fourier transform infrared (FTIR) spectroscopy in the step-scan mode using a Thermo Nicolet Nexus 870 FTIR spectrometer equipped with a Globar IR light source, as shown in Fig. 1(a). After mechanical light chopping (Thorlabs MC 1000) of the optical beam, the light was directed onto the Ge1-xSnx devices, which were mounted in the cryostat and cooled by a closed loop cooling system. An n-Ge wafer filter was placed in front of the samples to reduce the current contribution from the Ge substrate absorption, except for the
#202262 - $15.00 USD Received 29 Nov 2013; revised 25 Mar 2014; accepted 21 Apr 2014; published 1 May 2014 (C) 2014 OSA 5 May 2014 | Vol. 22, No. 9 | DOI:10.1364/OE.22.011029 | OPTICS EXPRESS 11030
pure Ge sample (D). The generated PC signal was delivered to the FTIR electronics from a lock-in amplifier.
Fig. 1. (a) Schematic of the experimental setup for photocurrent spectroscopy using an IR source (Globar) and optical chopper. The Ge1-xSnx device is held at temperature T on a cryostat. A lock-in amplifier and Thermo-Nicolet 870 FTIR were used to measure the spectral response. (b) Diagram of the Ge1-xSnx / n-Ge heterojunction device structure showing top and bottom metal contacts.
3. Results and discussion The inset in Fig. 2 shows the dark current-voltage characteristics of typical Ge1-xSnx / n-Ge heterojunction devices at room temperature. The large variation in the magnitude of the dark current between forward and reverse bias indicated good rectifying behavior. At a given forward voltage, the dark current increased with increasing Sn composition, suggesting higher conductivity of the materials and/or greater injection of charge carriers, perhaps caused by a smaller energy bandgap [13].
#202262 - $15.00 USD Received 29 Nov 2013; revised 25 Mar 2014; accepted 21 Apr 2014; published 1 May 2014 (C) 2014 OSA 5 May 2014 | Vol. 22, No. 9 | DOI:10.1364/OE.22.011029 | OPTICS EXPRESS 11031
0.8
5 4
Current(A)
Photoresponse(a.u)
6 Sn 12% Sn 9.8% Sn 4%
0.6 0.4 0.2 0.0 -0.4
3
0.0
0.4
Voltage(V)
0.8
-1.0 -0.5 0.0 0.5
2 1 0
0.5
0.6
0.7
0.8
Photon Energy(eV)
0.9
Fig. 2. Photocurrent response of a Ge0. 902Sn0.098/ n-Ge heterojunction device at 100K (sample B, with 9.8%Sn): Increasing reverse-biased voltage resulted in larger amplitude of photocurrent that saturated at higher reverse voltages. The inset shows the dark current-voltage characteristics of several Ge1-xSnx devices with different Sn compositions.
Figure 2 shows the spectral response of the photocurrent of a typical device of sample B (9.8% Sn) for several different biases at a temperature of 100 K. The spectral photocurrent showed a narrow peak at photon energies below 0.8 eV. The reduced photocurrent above 0.8 eV was attributed to the absorbing n-Ge wafer filter inserted in front of the devices. The photocurrent increased as the reverse bias voltage increased slightly. At reverse voltages above 0.5 Volts, however, the photocurrent started to saturate, which suggested that all the photo-generated carriers were swept out to the electrodes. The photocurrent decreased with forward bias, because the forward current opposes the collection of photo-generated carriers. The total current through the device is the sum of the dark current and the photocurrent. Figure 3 shows the spectral photoresponse of Ge1-xSnx/ n-Ge heterojunction devices with different Sn compositions at a temperature of 100 K, and reverse bias of 0.5 Volts. The peak output current from the devices varied from 3µA (sample C 12%) to 18µA (sample A 4%), using the n-Ge filter in the FTIR beam path as described above. The photon energies of the peak spectral photocurrents were shifted from 0.872 eV (1.42 µm) for pure Ge (sample D 0% Sn) to 0.619 eV (2.0 µm) for 12% Sn (sample C) at 100 K. The lower energy progression of the spectral peaks was due to the bandgap narrowing as the Sn composition increased, which is consistent with previously reported models [5, 6], and observations [15].
#202262 - $15.00 USD Received 29 Nov 2013; revised 25 Mar 2014; accepted 21 Apr 2014; published 1 May 2014 (C) 2014 OSA 5 May 2014 | Vol. 22, No. 9 | DOI:10.1364/OE.22.011029 | OPTICS EXPRESS 11032
Photoresponsce(A)
25 20 15
12% Sn 9.8% Sn 4% Sn pure-Ge
10 5 0 0.6
0.7
0.8
0.9
Photon Energy(eV)
Fig. 3. Spectral photoresponse in mid-IR regime of Ge1-xSnx/ n-Ge heterojunction devices, with different Sn compositions, at 100K. The response peaks lie at 0.619eV (sample C 12% Sn), 0.681eV(sample B 9.8% Sn), 0.765eV(sample A 4% Sn), and 0.872eV(pure Ge sample D, 0% Sn) respectively.
The photocurrents of the Ge1-xSnx / n-Ge heterojunction devices generated from the Globar IR source were compared to the photocurrent of a calibrated Newport 818 Ge detector (responsivity 0.81 A/W at peak wavelength) for calibration. The responsivities at the peak response for samples A (4% Sn), B (9.8% Sn), and C (12% Sn) were calculated to be 0.17 A/W, 0.13 A/W, and 0.03 A/W, respectively, at the temperature of 100K with 0.5V reverse bias. The relatively lower responsivity of the higher Sn composition device (sample C) was possibly due to the much thinner Ge1-xSnx layer compared to other samples. The relatively low responsivities of all Ge 1-xSnx devices compared to the Ge reference detector were attributed to the: a) relatively thin absorbing layer of the Ge1-xSnx, b) attenuated photons reaching the absorption layer of the devices, c) relatively large area of Al metal on the surface that blocked the active region (the metal covered ~64% of the surface), and d) some surface reflection from Ge1-xSnx layer. 4. Conclusion In summary, the properties of Ge1-xSnx / n-Ge heterojunction devices grown by MBE were investigated using electrical and photocurrent measurements. The 12% Sn sample (C) showed a photoresponse peak at 2.0 µm. The shift in wavelength in 12% Sn sample (C) from pure Ge was attributed to the bandgap decrease with Sn in the Ge 1-xSnx alloys, which can be very useful for extending the response of the devices into the mid-IR. The bias dependences indicated that the photoresponse saturated with a reverse bias of 0.5 volts. The Ge 1-xSnx alloys with lower Sn composition showed more photocurrent (higher responsivity) possibly due to the thicker absorption layer.
#202262 - $15.00 USD Received 29 Nov 2013; revised 25 Mar 2014; accepted 21 Apr 2014; published 1 May 2014 (C) 2014 OSA 5 May 2014 | Vol. 22, No. 9 | DOI:10.1364/OE.22.011029 | OPTICS EXPRESS 11033
Acknowledgments This work was financially supported by the AFOSR under award number: FA9550-09-10688, by Voltaix LLC under award number 12A01464, and by gifts from IBM Corporation, IR Labs, and Voltaix Corporation. Special thanks to T. Adam, D. Beatson, S. DeVore, M. Kim, M. Pikulin, R. Soref, and S. Zollner for suggestions on the manuscript and useful discussions.
#202262 - $15.00 USD Received 29 Nov 2013; revised 25 Mar 2014; accepted 21 Apr 2014; published 1 May 2014 (C) 2014 OSA 5 May 2014 | Vol. 22, No. 9 | DOI:10.1364/OE.22.011029 | OPTICS EXPRESS 11034