Dominik Kufer and Gerasimos Konstantatos *

Transcription

1 Highly sensitive, encapsulated MoS 2 photodetector with gate controllable gain and speed. Dominik Kufer and Gerasimos Konstantatos * D. Kufer and Prof. G. Konstantatos. ICFO-Institut de Ciencies Fotoniques, The Barcelona Institute of Science and Technology, Castelldefels (Barcelona), Spain (Gerasimos.konstantatos@icfo.es)

2 1.) Device overview The encapsulation approach was tested on several samples. The manuscript and SOM display data of four MoS 2 and one MoSe 2 device which are summarized in Table T1. Device MoS 2 _ 1 MoS 2 _ 2 MoS 2 _3 MoS 2 _4 MoSe 2 Layer thickness monolayer bilayer bilayer monolayer Multi-layer Dimension (WxL) 5 um x 3 um 7 um x 4 um 12.5 um x 2.4 um 6.0 um x 1.5 um 11.5 um x 6.5 um Oxide HfO 2 HfO 2 HfO 2 Al 2 O 3 HfO 2 Mobility (pristine) Mobility (encapsulated) 0.5 cm 2 /Vs 6.4 cm 2 /Vs 3.3 cm 2 /Vs 10 cm 2 /Vs 5.9 cm 2 /Vs 12 cm 2 /Vs 21.5 cm 2 /Vs 8.1 cm 2 /Vs 52 cm 2 /Vs 16 cm 2 /Vs V DS 5 V 3 V 1 V 5 V 5 V Responsivity R (at strong gate Vg) 434 A/W 400 A/W 563 A/W 406 A/W 16 A/W Detectivity D* 7.7 * Jones 7*10 10 Jones 1.7*10 12 Jones 3.8* Jones 1.0*10 11 Jone s Decay time (in depletion) 7 ms (41 mw/cm2) 10 ms (41 mw/cm2) 180 ms (26 mw/cm2) 55 ms (3.5 mw/cm2) 5 ms (3 uw/cm2) Table T1. Characteristics and figure of merit of all displayed devices.

3 2.) Atomic layer deposition and AFM images To assure a full coverage of the nanoflakes by HfO x the atomic layer deposition process was performed in 300 cycles leading to approximately 30 nm of oxide thickness. Atomic force microscope (AFM) was used to control the coverage of the nanoflakes. Figure S1. AFM images of two different devices after HfO x encapsulation. MoS 2 device 1 (a) and MoSe 2 device (b) with entirely covered surface. 3.) Encapsulation with Al 2 O 3 Figure S2 demonstrates the device performance of monolayer MoS 2 encapsulated with a 30 nm thick Al 2 O 3 film. The device improvement after encapsulation demonstrates very similar trends as HfO 2 encapsulated devices and proves that this approach is generalizable to other encapsulant materials.

4 Figure S2. MoS 2 vs. MoS 2 /Al 2 O 3 device performance. (a) Linear transfer curve I DS -V G before and after ALD encapsulation. Inset: logarithmic plot. V DS = 0.1 V. (b) Corresponding field-effect mobility. The mobility µ is determined in the linear regime of the device where the transfer curve I DS -V G shows the highest slope (At V G ~ 25 V). The encapsulation improves mobility by a factor of ~ 5 from 10 cm 2 /Vs to 52 cm 2 /Vs (c) Power dependent responsivity after Al 2 O 3 encapsulation. The responsivity improved by more than an order of magnitude compared to the previous unprotected device. V DS = 5 V. (d) Decay curves before and after encapsulation after switching of a light pulse with 3.5 mw/cm 2 illumination intensity. 4.) MoSe 2 device Figure S3 summarizes a full set of measurements for MoSe 2 to demonstrate that the encapsulation approach can be applied on other 2-dimensional materials with similar trends and improvement.

5 Figure S3. Dataset of the MoSe 2 device. (a) Linear transfer curve I DS -V G before and after ALD encapsulation. Inset: logarithmic plot. V DS = 1 V. (b) Power dependent responsivity before and after HfO 2 encapsulation. V DS = 1 V. (c) Power dependent responsivity for different backgate voltages V G. V DS = 5 V. (d) Decay times and responsivity depending on V G for low power of 3uW/cm 2. (e) Noise power density for different V G. (f) Dark current I DS, responsivity R and measured detectivity D* as a function of V G.

6 5.) Device stability under humid conditions In order to confirm the higher stability due to the oxide encapsulation, a series of control experiments under harsh, humid environment was performed. Therefore two monolayer devices were prepared, pristine MoS 2 and encapsulated MoS 2 /HfO 2, which underwent the same humidity conditions. Both samples were characterized in ambient conditions first and then consecutively exposed to different humidity conditions inside a climatic chamber with relative humidity (RH) of 40%, 60% and 80%. Both samples were kept inside the chamber under stable RH values for around one hour and measured right afterwards. The results are depicted in Figure S6 for the pristine (a) and the encapsulated device (b). The pristine device is clearly affected by the harsh environment and shows increase of hysteresis and strong reduction of on-current. The initial mobility of 3.7 cm 2 /Vs in air is reduced to 1.7 cm 2 /Vs, 1.4 cm 2 /Vs and 0.9 cm 2 /Vs after exposure to 40%, 60% and 80% respectively. The protected device on the other hand shows minor impact of the harsh humidity conditions. The I DS -V G curve stays free of hysteresis and mobility doesn t change. Also the photoresponse and decay dynamics are quite stable after the humidity treatment and show responsivity on the order of 30 A/W (measured at Vg = -40 V) with decay times of ms, while the pristine device measured under same conditions gave responsivity of 4 A/W with corresponding decay times of 2-3 s.

7 Figure S4. MoS 2 vs. MoS 2 /HfO 2 device performance after harsh humidity treatment. The top panel illustrates linear I DS -V G curves of the pristine (a) and the protected device (b) measured before and after different RH treatments of 40%, 60% and 80%. The bottom panel shows the corresponding decay dynamics.

8 6.) Noise measurements Figure S5. Measured noise power density at two different backgate voltages. (a) At V G = -20 V a clear 1/f component can be seen for low frequency. (b) At V G = -40 V the noise floor of the measuring unit is reached and the noise power density of the device cannot be determined.