Supporting Information. Silicon Nanowire - Silver Indium Selenide Heterojunction Photodiodes

Transcription

1 Supporting Information Silicon Nanowire - Silver Indium Selenide Heterojunction Photodiodes Mustafa Kulakci 1,2, Tahir Colakoglu 1, Baris Ozdemir 3, Mehmet Parlak 1,2, Husnu Emrah Unalan 2,3,*, and Rasit Turan 1,2 1 Department of Physics, METU, Ankara (06800) Turkey. 2 Center for Solar Energy Research and Applications, METU, Ankara (06800) Turkey. 3 Department of Metallurgical and Materials Engineering, METU, Ankara (06800) Turkey. The size of our devices was 0.5 x 0.5 cm 2 and there are 4 indium circular top contacts in 1 mm diameter. The devices were mount on a chip holder using conductive silver paste and indium top contacts were wired to the pads using 25 µm gold wire. For the photoresponse measurements four indium contacts wired to the pads were shorted together into one device and the illuminated light were covered the hole sample. Photoresponse area was 0.5 x 0.5 cm 2 minus the area of four-indium contacts. A schematic of the devices fabricated in this work is shown below in Figure S1. Figure S1. Schematic of the fabricated devices in this work.

2 Line analysis results are provided in Figure S2. 2 nanowires within the AgInSe 2 film can be realized in the image, clearly showing the conformality of the coating. Figure S2. Line scan across the Si nanowire AgInSe 2 heterojunction. Frequency dependent C-V measurement is a very useful method for obtaining interface properties of the junctions. At low frequencies surface state density can be obtained as N ss = C ss / qa, where A is the diode area and C ss is surface state capacitance value, which can be determined from vertical axis intercept of capacitance frequency plots. Frequency dependent C-V measurements have been performed at room temperature for both samples. The variation of capacitance as a function of frequency at zero bias is given in the Figure S3 (a) and (b) below for samples on planar Si and Si nanowires, respectively. The defect densities of the heterojunctions with nanowire decorated and planar Si substrates are calculated as ~2.92x10 12 cm -2 V -1 and ~3.24 x10 12 cm -2 V -1, respectively. This clearly represents improved film quality in AgInSe 2 upon deposition over the nanowires. For the calculations area of both devices was assumed to be the area of the indium top contact. Although

3 results seems comparable, it should be considered that the actual junction area of the device built on nanowire structure is much more than the indium top contact due to the three dimensionality of the junction. Therefore, one can easily guess that the interface state density is much lower in the case of nanowire based heterojunction per unit area. Figure S3. Room temperature frequency dependent capacitance measurement for AgInSe 2 thin films on (a) planar Si and (b) Si nanowires.

4 The I-V curve of the devices under illumination is given in Figure S4 below, complementing the I-V characteristics under dark in Figure 4 (b). The measurement was taken using a halogen lamp with an intensity of 950 W/cm 2. Open circuit voltages of the devices were found as 158 mv and 107 mv for the heterojunction devices with Si nanowires and planar Si, respectively, under illumination. This is an indication of good interface between AgInSe2 and Si nanowire junction compared to the planar one. In addition, although very weak to be considered as a solar cell, the photoconversion efficiency of the heterojunction devices with nanowires was much higher than that of the planar device. Figure S4. Semilogaritmic plot of the illuminated I-V characteristics of the heterojunction diodes. For comparison of responsivity of AgInSe 2 /Si NW heterojunction with p- n junction diode we have fabricate a Si diode and measured its photoresponse. Below in Figure S5, responsivity curve of a Si homojunction diode is given, while a heterojunction diode is presented in our manuscript in Figure 5 (c). It is clear that the responsivity of a homojunction Si diode is much better than that of AgInSe 2 / Si NW one. However, when the curves are compared we have seen that high energy photoresponse of the heterojunction diode is better than that of the homojunction Si diode. The response curve at high-energy tail extended to 300 nm, which is a 100 nm wider response

5 window, compared to the homojunction Si diode. With the optimization of both nanowire synthesis and AgInSe 2 film formation (higher doping density and mobility also included) and passivation, we believe that the responsivity of the heterojunction could be enhanced much further. Figure S5. Photoresponse curve of a homojunction Si diode.