Short-Wave Near-Infrared Linear Dichroism of Two-Dimensional

Sep 19, 2017 - Polarized detection has been brought into operation for optics applications in the visible band. Meanwhile, an advanced requirement in ...
2 downloads 10 Views 3MB Size
Article pubs.acs.org/JACS

Cite This: J. Am. Chem. Soc. 2017, 139, 14976-14982

Short-Wave Near-Infrared Linear Dichroism of Two-Dimensional Germanium Selenide Xiaoting Wang,† Yongtao Li,† Le Huang,† Xiang-Wei Jiang,† Lang Jiang,‡ Huanli Dong,‡ Zhongming Wei,*,† Jingbo Li,*,† and Wenping Hu*,‡,§ †

State Key Laboratory of Superlattices and Microstructures, Institute of Semiconductors, Chinese Academy of Sciences & College of Materials Science and Optoelectronic Technology, University of Chinese Academy of Sciences, Beijing 100083, China ‡ Beijing National Laboratory for Molecular Sciences, Key Laboratory of Organic Solids, Institute of Chemistry, Chinese Academy of Sciences, Beijing 100190, China § Tianjin Key Laboratory of Molecular Optoelectronic Sciences, Department of Chemistry, School of Sciences, Tianjin University & Collaborative Innovation Center of Chemical Science and Engineering (Tianjin), Tianjin 300072, China S Supporting Information *

ABSTRACT: Polarized detection has been brought into operation for optics applications in the visible band. Meanwhile, an advanced requirement in short-wave near-infrared (SW-NIR) (700−1100 nm) is proposed. Typical IV−VI chalcogenides2D GeSe with anisotropic layered orthorhombic structure and narrow 1.1−1.2 eV band gappotentially meets the demand. Here we report the unusual angle dependences of Raman spectra on high-quality GeSe crystals. The polarization-resolved absorption spectra (400−950 nm) and polarization-sensitive photodetectors (532, 638, and 808 nm) both exhibited wellreproducible cycles, distinct anisotropic features, and typical absorption ratios αy/αx ≈ 1.09 at 532 nm, 1.26 at 638 nm, and 3.02 at 808 nm (the dichroic ratio Ipy/Ipx ≈ 1.09 at 532 nm, 1.44 at 638 nm, 2.16 at 808 nm). Obviously, the polarized measurement for GeSe showed superior anisotropic response at around 808 nm within the SW-NIR band. Besides, the two testing methods have demonstrated the superior reliability for each other. For the layer dependence of linear dichroism, the GeSe samples with different thicknesses measured under both 638 and 808 nm lasers identify that the best results can be achieved at a moderate thickness about 8−16 nm. Overall, few-layer GeSe has capacity with the integrated SW-NIR optical applications for polarization detection.

1. INTRODUCTION Linear dichroism (LD), as a form of electromagnetic spectroscopy, refers to the difference in the absorption of polarized light with the polarization direction parallel or perpendicular to an orientation direction of a monocrystal.1−4 LD allows for the realization of optics applications such as optical switches, near-field imaging, and polarized detection for intrinsically anisotropic materials, which has been brought into operation in the visible band.5 Meanwhile, an advanced requirement in SW-NIR (700−1100 nm) is proposed, which could broaden the practical fields to medical care, night vision and perspective, remote sensing, thermal imaging polarimetry due to its thermal effect, nondestructive property, and strong penetration capacity.6,7 For the linear dichroic samples, previous research works have mainly centered on the one-dimentional nanostructure. But the device preparation and aligned process are limited by their aspect ratios. Recently, two-dimensional (2D) layered systems such as metallic graphene,8−10 insulating hexagonal boron nitride,11,12 and the family of transition metal dichalcogenides (TMDCs) have been extensively developed, due to the absence © 2017 American Chemical Society

of dangling bonds along the c-axis, high specific surface area, and confined electronic systems.13−17 But owing to the commonly appearing highly symmetric structure, most of the 2D materials are isotropic within the plane for their linear optical properties.18,19 Germanium selenide (GeSe), a typical binary IV−VI chalcogenide, crystallizes in a highly anisotropic layered orthorhombic crystal structure (the space group Pcmn− 18 D16 2h, lower symmetry than the space group Bmab−D2h of black 20−22 phosphorus). It has closely arranged band gaps in the range of 1.1−1.2 eV, which makes its applicable LD wavebands distributed within the wavelength of 1100 nm (vis/short-wave near-infrared (SW-NIR) band). And it also possesses a high absorption coefficient at the wavelength close to the band edge of GeSe, due to the high joint density of states near the band extrema.23−25 Its pecularity of high vapor pressures (growth temperature below the melting point of 670 °C) promotes readily prepared, low-power, and high-quality production for large-scale thin films or crystals. In addition, GeSe possesses the Received: June 23, 2017 Published: September 19, 2017 14976

DOI: 10.1021/jacs.7b06314 J. Am. Chem. Soc. 2017, 139, 14976−14982

Article

Journal of the American Chemical Society basic requirements of environmental stability, nontoxicity, and earth-abundant constituent elements for device applications.26,27 In view of the above points for 2D GeSe, more efforts are needed in surveying its in-plane anisotropic structure related characteristics such as LD, to realize its polarized application in the vis/SW-NIR band for various fields. Herein, we obtained high-quality 2D GeSe crystals by thermal sublimation deposition, and the angle dependences of Raman spectra for the Ag and B3g active modes were measured in parallel or cross-polarization configurations. An unusual Raman response for the totally symmetric mode intensities was observed, which confirmed its application in determining the crystal orientation. The experimental polarization-resolved absorption spectra (400−950 nm) based on GeSe nanoflakes exhibited strong absorption capacity, distinct anisotropic feature, and typical absorption ratio αy/αx ≈ 1.09 at 532 nm, 1.26 at 638 nm, and 3.02 at 808 nm. The wavelength coverage of the polarization peak (620−900 nm) was further obtained by the extracted dichroic ratio and LD amplitude plots. Then the 532, 638, and 808 nm polarization-sensitive photodetectors were obtained and exhibited well-reproducible cycles along with the polarized angle and remarkable dichroic ratio (Ipy/Ipx ≈ 1.09 at 532 nm, 1.44 at 638 nm, and 2.16 at 808 nm. The LD amplitude as the other critical parameter was extracted as 1.42 μA at 808 nm under 95 mW cm−2 illumination. To summarize, the polarized measurement for GeSe exhibited an optimum anisotropic response around 808 nm (SW-NIR band). Besides, the angle-resolved photocurrent presented similar forms (twolobed shape) in the polar coordination and a homologous trend to the absorption spectra, indicating that the two testing methods have demonstrated a superior reliability for each other. It follows that the polarized absorption spectra can systematically and accurately set the direction for sensitivity to the incident light in different polarization orientations. For the layer dependence of LD, the theoretical calculation confirmed that the increased layer number has decreased the dichroic ratio. But when this case was coupled with practical factors, the GeSe samples with different thicknesses (3.46, 8.63, 15.2, and 57 nm) were measured and identified that the best results can be achieved at a moderate thickness about 8−16 nm. Thus, few-layer GeSe potentially realizes the integrated SW-NIR optical applications for polarization detection.

Figure 1. (a) Left: Configuration view of the experimental setup for the synthesis of bulk GeSe crystals via the CVT process. Right: XRD patterns of GeSe precursor powder (purple line) and bulk GeSe flakes (black line). (b) HRTEM image of an as-grown GeSe nanosheet and the corresponding atomic arrangement structure of GeSe (inset). (c) Corresponding SAED pattern of the nanosheet and recrystallized at low temperature; thus the crystallinity increased from the polycrystalline to the monocrystalline state.

the SAED pattern (Figure 1c) shows the spot pattern with the surface normal oriented along the [100] crystal zone axis, identical to the [100] sheet orientation observed by XRD in Figure 1b. TEM-EDX elemental maps indicate the uniform distribution throughout the structure (Figure S1). Furthermore, it can be determined from the corresponding EDX spectrum and XPS spectra (Figure S2) that the atomic composition was around Ge:Se = 52:48. 2.2. Vibrational Characteristics. The atomic structure of bulk GeSe from the side view could be depicted as shown in Figure 2a. Raman spectroscopy, as a powerful spectroscopic technique, can be commonly used for crystal orientation identification and vibration symmetry of chemical bond evaluation.28,29 The anisotropy of GeSe can be first explored by orientation-dependent Raman spectroscopy, where the sample was rotated within the plane every 10° by means of a horizontal rotating table.30,31 By inserting a linear polarization analyzer into the detection system, the scattered light polarization was taken along the directions parallel or perpendicular to the incident light polarization. Figure 2b and c present a sequence of Raman spectra for GeSe nanoflakes (Figure S4a) at different rotation angles and the corresponding contour color map under the parallel or cross-polarized configuration. Two Raman peaks were observed at around 150 and 188 cm−1 for the GeSe nanoflakes, corresponding to the B3g mode and Ag mode, regardless of the polarization.32 The contour color map further displayed that the relative intensities of the Ag-like (188 cm−1) mode and the B3g-like (150 cm−1) mode had strong correlation with the sample rotation angles. The angle-resolved variation in the Raman peak intensities was also plotted in the polar coordinates and overlapped with the respective Raman mapping as marked in Figure 2d−g, which matched well with each other. Stars are measured data without background correction, and solid curves correspond to the nonlinear fitted data by the equation in Table 1. It can be clearly seen that whether parallel or perpendicular polarization configuration, the peak intensity at the B3g and Ag modes exhibited a marked

2. RESULTS AND DISCUSSION 2.1. Synthesis and Characterization. The bulk GeSe crystals were synthesized by a chemical vapor transport (CVT) method with GeSe powders as the precursors in a horizontal two-zone tube furnace (more details are described in the Experimental Section). The detailed crystal structures of the GeSe source powder and bulk GeSe flakes were characterized by X-ray diffraction (XRD) patterns (Figure 1a) and indexed to the orthorhombic unit cell (JCPDS No. 48-1226, a = 10.825 Å, b = 3.833 Å, c = 4.388 Å, Pnma 62) without other impurity phases. The results implied that the source powder was sublimated into gaseous GeSe molecules directly at high temperature. Figure 1b and c exhibit the high-resolution TEM (HRTEM) image of the exfoliated GeSe nanosheets and corresponding selected-area electron diffraction (SAED) pattern. The HRTEM image reveals a lattice spacing of 2.98 Å and an intersection angle of approximately 98°, consistent with the {011} set of planes of the orthorhombic GeSe crystal structure (see the atomic arrangement structure in Figure 1b inset). And 14977

DOI: 10.1021/jacs.7b06314 J. Am. Chem. Soc. 2017, 139, 14976−14982

Article

Journal of the American Chemical Society

Figure 2. (a) Atomic structure of bulk GeSe from the side view. Angle-resolved Raman spectra and the corresponding contour color map of the GeSe nanoflakes measured in the (b) parallel and (c) cross-polarization configurations. (d−g) Polar plots of the extracted measured and fitted peak intensities of the Ag (188 cm−1) and B3g (150 cm−1) modes under the polarization configurations (parallel or cross). Stars are measured data without background correction, and solid curves correspond to the nonlinear fitted data by the equation in Table 1. The angle-resolved variations in peak intensities are overlapped with the respective Raman mappings.

can be expressed by I ∝ |ei·R·es|2, where ei and es are the unit polarization vectors of the incident and scattered light, respectively.35−38 For GeSe (a = 10.825 Å, b = 3.833 Å, c = 4.388 Å), its layer stacking is along the lattice constant a (set as the X-axis) (Figure 2a).39,40 The constant b is along the zigzag direction (defined as the Z-axis), and c is the armchair direction (defined as the Y-axis), so ei and es are in the Y−Z plane. Thus, by setting the incident polarization as θ to the crystal axis, ei = (0, cos θ, sin θ) and es = (0, cos θ, sin θ)T (parallel-polarization configurations) and (0, −sin θ, cos θ)T (cross-polarization configurations). From the Raman tensor elements in Table S2, non-null components of the B1g (B2g) mode lie in the X−Z (X−Y) plane, so that only the Ag and B3g modes can be observed by backscattering on the Y−Z plane.41 And combined with the above analysis, the scattering intensities of the Ag and B2g modes under parallel and cross-polarization configuration are presented in Table S3. It is the usual Raman approach that assumes the Raman tensor elements to be real. But obviously, the measured data for the Ag mode do not match well with the fitted curves based on the above Raman intensity equations. Especially under the ei ∥ es configuration, the measured data present a secondary maximum at 50° and 230°. Actually for absorptive materials, this unusual angular dependence is linked with its LD, and we need to allow for the impact of light absorption on the Raman tensor.42,43 Thus, the relevant elements in this work have real and imaginary parts and can be uniformly written as k = |k|eiφk = |k|(cosφk+ isinφk).35 Substituting the complex values of the Raman tensor into I ∝ |ei·R·es|2, the scattering intensity of the Ag mode is modified as in Table 1, in which the phase difference φcb is φc minus φb. For any value of φcb, it can be perceived that the intensities and angles for the maximum and

Table 1. Modified Raman Scattering Efficiency of Ag Modes in GeSe Ag mode

B3g mode

parallel

cross

parallel

cross

(|b| sin2 θ + |c| cos φcb cos2 θ)2 + |c|2 sin2 φcb cos4 θ

[(|b| − |c| cos φcb)2 + |c|2 sin2 φcb] sin2 θ cos2 θ

f 2 sin2 2θ

f 2 cos2 2θ

periodic variation feature (90° or 180°). Only the Ag mode in the parallel configurations yielded an unusual two-lobed shape (a maximum at 50° or 230° and a secondary maximum at 140° or 320°), whereas three other cases exhibit a four-lobed shape (90° as a minor cycle). Then a sequence of Raman spectra at different rotation angles for a thinner GeSe sample (Figure S4b) was carried out to identify the variation with sample thickness. Its polarized Raman spectra exhibited similar forms in the polar coordination to that of thicker GeSe nanoflakes (Figure S5a−d). Although the Ag mode in the parallel configurations still yielded an unusual two-lobed shape (a maximum at 20° or 200° and a secondary maximum at 110° or 290°), a secondary maximum just had a weak bulge, probably due to the weak peak intensity parallel to the substrate signal. And it was observed from the corresponding contour color map (Figure S5e,f) that Raman peak signals at modes of 150 and 188 cm−1 were apt to be confused with the substrate signal, which enlarged the impact of the apparatus error. To comprehend the angle-resolved Raman nature of GeSe, the second-order Raman tensor R for a specific Raman mode was taken into consideration. The group theory analysis indicates that with the incident light normal to the GeSe layer plane, active Raman tensors can be denoted as in Table S2.33,34 And the generalized form of Raman scatter intensity 14978

DOI: 10.1021/jacs.7b06314 J. Am. Chem. Soc. 2017, 139, 14976−14982

Article

Journal of the American Chemical Society

Figure 3. (a) Electronic band structures for bulk GeSe with the HSE06 functional method. (b) Calculated absorbance α along the x-axis (armchair direction) and y-axis (zigzag direction) for 1L, 3L, and 6L GeSe. (c) Experimental polarization-resolved absorption spectra with the spectral range 400−950 nm and the microscope image (inset) of GeSe. (d−f) Evolution of the absorbance plotted in the polar coordination at different wavelengths (532, 638, 808 nm), which were scaled out by the dashed line in (c). The black lines are fitted results by the sinusoidal function α(δ) = αy cos2(δ + φ) + αx sin2(δ + φ).

Their absorption spectra along the x- and y-axis have presented a similar trend in the range of 400−1400 nm. And their dichroic ratio and LD amplitude, as critical parameters for LD, were extracted from Figure 3b and displayed in Figure S9a. The variation tendency of theoretical absorbance, even the bipeak distribution, was almost in agreement with our experimental absorption spectra (Figure 3c). There existed a discrepancy that as the layer number increased, the intersection of the absorption spectra along the x-axis and y-axis moved toward the near-infrared region. This situation also applied to the location at which the most remarkable dichroic ratio and LD amplitude emerged. And the most obvious difference was that the increased layer number has decreased the dichroic ratio and slowly increased the LD amplitude at 620−1100 nm, which has identified the thickness dependence in theory. Then the experimental polarization-resolved absorption spectra (400−950 nm) based on GeSe nanoflakes were obtained and displayed in Figure 3c. The evolution of absorbances was plotted in the polar coordination at typical wavelengths (532, 638, 808 nm), which exhibited a two-lobed shape (Figure 3d−f), and the corresponding absorption ratio αy/αx ≈ 1.09 at 532 nm, 1.26 at 638 nm, and 3.02 at 808 nm. The spatial structure of the wave functions of several bands near the CBM and VBM (Figure S8) was calculated to physically explain anisotropic optical absorption to the external in-plane polarization. Its dichroic ratio and LD amplitude were extracted from Figure 3c and are displayed in Figure S9b. It was observed that GeSe nanoflakes displayed a polarization peak within the visible and SW-NIR spectrum (620 to 900 nm). This directly suggests that the LD originates from anisotropy of the GeSe crystal structure itself.46,47

secondary maximum intensity peaks do not change. And intensities have always been |c|2 and |b|2 at θ = 0° and θ = 90°, while angles have always been in the armchair and zigzag directions. Thus, the Ag (188 cm−1) mode can be employed to determine the crystalline orientation of the GeSe flakes, and the armchair direction of our test sample is oriented about the 50° direction (Figure 2d). TEM analysis has been further conducted and confirmed that Ag modes reached a maximum intensity when the polarization of the incident laser was parallel to the armchair crystallographic direction of GeSe (Figure S6). This result validates the assignment of the armchair and zigzag directions. 2.3. Theoretical and Experimental Polarization-Resolved Absorption Spectra. The Raman spectroscopy of GeSe has provided an initial assessment for its high in-plane crystal anisotropy, which prompts us to further probe it for inplane absorbance along different directions. The electronic band structure of GeSe was calculated by the hybrid functional method (HSE06), which exhibits a narrow band gap of 1.70 eV for 1L GeSe (Figure S7b) and 1.34 eV for bulk GeSe (Figure 3a). Thus, the laser, with excitation energy higher than the band gap, will induce the production of electron−hole pairs. Then the optical absorbance α of 1L, 3L, and 6L GeSe along the xaxis (zigzag direction) and y-axis (armchair direction) was calculated theoretically (Figure 3b). The optical absorption coefficient was calculated from the dielectric function by the expression α(w) = √2w[ ε12(ω) + ε22(ω) − ε1(w)]1/2, where w is the frequency of the incident photon and ε1 and ε2 are the real part and imaginary part of the dielectric function.44,45 14979

DOI: 10.1021/jacs.7b06314 J. Am. Chem. Soc. 2017, 139, 14976−14982

Article

Journal of the American Chemical Society

Figure 4. (a) 3D schematic structure of the polarized photodetection device. The laser was set to pass through the Glan-Taylor prism (polarizer), and then the light polarization was controlled by the half-wave plate. (b) Drain current Id as a function of time with light switched on/off at Vd = 1, 2, and 3 V. (c) Anisotropic response in Iph under the 808 nm laser described via the 2D colormap (Iph is denoted by the color bar, the drain voltage Vd as x-axis, and δ as y-axis). (d) Evolution of photocurrent plotted with probe polarization from 0° to 360° at Vd = 2 V (red dots). The black lines are the fitting results by the sinusoidal function Iph(δ) = Ipy cos2(δ + φ) + Ipx sin2(δ + φ).

2.4. Linear-Polarization-Sensitive Photodetectors with Different Thicknesses. The absorption spectra of GeSe in theory and experiment have provided a general analysis for its anisotropic absorption region, which prompts us to further probe it in linear-polarization-sensitive photodetectors. Then the photosensitive test equipment (Figure 4a) was reassembled, and the modulated laser beam was first set to pass through a Glan-Taylor prism (polarizer) and half-wave plate (HWP) and finally illuminated on the GeSe channel. It was conducted by rotating the HWP every 5° to change the polarization direction while keeping the incident power constant. It needs to be noted that if the vibration plane of the linear-polarized incident light makes an angle of θ to the main section of the HWP, that of the transmission light goes around the angle of 2θ from the original orientation. So the adjacent polarization angle for the sample is 10° apart. The experiments based on samples with four different thicknesses were first measured to identify thickness dependence at both 638 and 808 nm, while the polarized measurement at 532 nm has not been conducted due to its relatively unobvious anisotropic response. The LD amplitude was inevitably disturbed by an external factor in the experiment, such as the sample channel area and incident light power. So the dichroic ratio and LD amplitude/(P*S) were both

calculated and displayed in Tables S4 and S5. LD amplitude/ (P*S) refers to the LD amplitude per unit power (P) of the incident light per unit area (S) of one device channel. Based on our comprehensive experimental study of the thickness dependence under different wavelengths (as shown in Figures S10, S11, S13), they showed that the best results can be achieved at a moderate thickness about 8−16 nm. Thicker or thinner samples both showed much lower dichroic ratios. The optical absorbance α of 1L, 3L, and 6L GeSe along the x-axis (zigzag direction) and y-axis (armchair direction) displayed theoretically that the increased layer number has decreased the dichroic ratio (Figure S9a). Thus, there was no doubt that thicker samples showed much lower dichroic ratios, which was caused by the internal layer structure of GeSe. As for thinner samples, the dichroic ratio and LD amplitude/(P*S) of fivelayer (3.46 nm thick) GeSe have presented weaker polarized photodetection, and its extracted photocurrent has been in the slow downward trend. Weak absorption capacity (low photocurrent value) may be the main inducement for enlarged apparatus error during measurement, so it is inevitable that a little fluctuation and metal contact resistance played a role in changing the dichroic ratio. For amplitude/(P*S), the downward trend for thicker samples was tempered at 638 nm, even on the slow rise at 808 nm, which was similar to the 14980

DOI: 10.1021/jacs.7b06314 J. Am. Chem. Soc. 2017, 139, 14976−14982

Article

Journal of the American Chemical Society

3. CONCLUSIONS In conclusion, we have synthesized bulk GeSe crystals by the CVT method, then in-plane anisotropy essentially and the crystalline orientation of GeSe were preliminarily determined by angular-dependent Raman spectroscopy. The experimental polarization-resolved absorption spectra (400−950 nm) based on GeSe nanoflakes exhibited strong absorption capacity, a distinct anisotropic feature, and typical absorption ratio αy/αx ≈ 1.09 at 532 nm, 1.26 at 638 nm, and 3.02 at 808 nm. Then the 532, 638, and 808 nm polarization-sensitive photodetectors exhibited a well-reproducible cycle along with the polarized angle and remarkable dichroic ratio (Ipy/Ipx ≈ 1.09, 1.44, 2.16). Generally speaking, the optimum wavelength was around 808 nm (SW-NIR band). In addition, the angle-resolved photocurrent presented similar forms (two-lobed shape) in the polar coordination and homologous trend to the absorption spectra. For the layer dependence of LD, when the theoretical calculation was coupled with practical factors, the GeSe samples with different thicknesses measured under both 638 and 808 nm lasers identify that a moderate thickness (about 8− 16 nm) contributes to the acquisition of optimum results for anisotropic photodetection. These results have implied that few-layer GeSe has the capacity with the integrated SW-NIR optical applications for polarization detection.

theoretically calculated absorption spectra displayed in Figure S9a. This was possibly inspired by the thickness and dichroic ratio. Thus, it can be illustrated that moderate thickness contributes to the acquisition of optimum results for anisotropic photodetection. According to our above disscussion, which showed the optimum results only appeared at a moderate thickness, the 8.63 nm GeSe was selected as a typical case for complete and systematic research. The time-resolved photoresponse of a 8.63 nm thick (about 13 layers) GeSe nanoflake (Figure S12) was first measured combined with the 532 nm laser source, and it has verified the high (Rλ of 4.25 AW−1) and repeatable response with different Vd (1, 2, 3 V) under 143 mW cm−2 illumination (Figure 4b). Figure S3 described the transfer and output curves of the GeSe device on a 30 nm Al2O3 substrate and revealed a p-type transport character. Then 532, 638, and 808 nm polarization-sensitive photodetectors for GeSe have been measured to investigate the wavelength dependence, and their evolution of photocurrent with probe polarization was described via a 2D colormap (Iph is denoted by the color bar, the drain voltage Vd as x-axis, and δ as y-axis) and displayed in Figure S13. Their comparison diagrams clearly exhibited a remarkable dichroic ratio (Ipy/Ipx ≈ 1.09, 1.44, 2.16), regardless of the shape of the samples and the laser output power. Figure 4c described separately the remarkable anisotropic response via the 2D colormap under 95 mW cm−2 illumination of a typical 808 nm laser. And its evolution of photocurrent at Vd = 2 V was further extracted as a function of the polarization angle in the polar coordinate (Figure 4d). The angle-resolved photocurrent exhibited similar forms (two-lobed shape) in the polar coordination and a homologous trend to the polarized absorption spectra (400−950 nm) (the corresponding absorption ratio αy/αx ≈ 1.09 at 532 nm, 1.26 at 638 nm, 3.02 at 808 nm) as in Table 2. The dichroic ratio was not totally



The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/jacs.7b06314. TEM-EDS and XPS spectrum of the GeSe nanoflakes; transfer and output plots of the devices; polarized Raman spectra of 21.3 nm thick GeSe nanoflakes; Raman scattering efficiency in GeSe; calculated absorbance, corresponding dichroic ratio and LD amplitude for 1L, 3L, and 6L GeSe and experimental polarization-resolved absorption spectra, corresponding dichroic ratio and LD amplitude of GeSe; the contrast for GeSe with four different thicknesses at 638 and 808 nm; 532, 638, and 808 nm polarized photodetectors for the 8.63 nm thick GeSe; Experimental Section (PDF)

Table 2. Summary of the Anisotropic Absorption Ratio of 2D GeSe Flakes and Dichroic Ratio in the Device 532 nm 638 nm 808 nm

absorption ratio αy/αx

dichroic ratio Ipy/Ipx

1.09 1.26 3.02

1.09 1.44 2.16

ASSOCIATED CONTENT

S Supporting Information *



AUTHOR INFORMATION

Corresponding Authors

*[email protected] *[email protected] *[email protected]

identical to the absorption ratio, possibly due to the device processing technology for polarization photodetectors or apparatus error during measurement or diverse sample thickness. The polarization dependence of the photocurrent (Iph) yields a two-lobed polar plot and can be perfectly fitted by the equation Iph(δ) = Ipy cos2(δ + φ) + Ipx sin2(δ + φ).48 δ is the angle with respect to the 0° reference direction, φ is the angle between the y-direction and the 0° reference, and the fitting curves give φ = 50° for probe polarization. Ipy, Ipx, and Iph(δ) refer respectively to the photocurrent along the y-axis, xaxis, and δ directions, in which Ipy (Ipx) corresponds to 2.65 and 1.23 μA here.49−51 Its dichroic ratio was 2.16, and the LD amplitude like the other critical parameters was extracted as 1.42 μA at 808 nm. The above results accurately and expediently identify the orientation-dependent absorption and subsequent photocurrent detection via LD, further revealing the intrinsic anisotropy of the GeSe nanoflakes.

ORCID

Le Huang: 0000-0003-3189-2171 Zhongming Wei: 0000-0002-6237-0993 Wenping Hu: 0000-0001-5686-2740 Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS The authors acknowledge National Natural Science Foundation of China (61622406, 11674310, 51633006, 61571415, 51502283, 11574304, 61201105, 91222203, 91433115), M i n i s t r y o f S c i e n ce a n d T e c h n o l o g y o f C h i n a (2017YFA0207500, 2016YFB04001100, 2016YFB0700700, 2014CB643600, 2013CB933403), and the Strategic Priority Research Program of the Chinese Academy of Sciences 14981

DOI: 10.1021/jacs.7b06314 J. Am. Chem. Soc. 2017, 139, 14976−14982

Article

Journal of the American Chemical Society

Xing, D. Y. Phys. Rev. B: Condens. Matter Mater. Phys. 2015, 92, 054110. (29) Nagler, P.; Plechinger, G.; Schuller, C.; Korn, T. Phys. Status Solidi RRL 2016, 10, 185. (30) Hafeez, M.; Gan, L.; Li, H. Q.; Ma, Y.; Zhai, T. Y. Adv. Mater. 2016, 28, 8296. (31) Lorchat, E.; Froehlicher, G.; Berciaud, S. ACS Nano 2016, 10, 2752. (32) Xia, F. N.; Wang, H.; Jia, Y. C. Nat. Commun. 2014, 5, 4458. (33) Wu, J. X.; Mao, N. N.; Xie, L. M.; Xu, H.; Zhang, J. Angew. Chem. 2015, 127, 2396. (34) Chenet, D. A.; Aslan, O. B.; Huang, P. Y.; Fan, C.; van der Zande, A. M.; Heinz, T. F.; Hone, J. C. Nano Lett. 2015, 15, 5667. (35) Ribeiro, H. B.; Pimenta, M. A.; de Matos, C. J. S.; Moreira, R. L.; Rodin, A. S.; Zapata, J. D.; de Souza, E. A. T.; Castro Neto, A. H. ACS Nano 2015, 9, 4270. (36) Zhang, S.; Yang, J.; Xu, R. J.; Wang, F.; Li, W. F.; Ghufran, M.; Zhang, Y. W.; Yu, Z. F.; Zhang, G.; Qin, Q. H.; Lu, Y. R. ACS Nano 2014, 8, 9590. (37) Zhao, H.; Wu, J. B.; Zhong, H. X.; Guo, Q. S.; Wang, X. M.; Xia, F. N.; Yang, L.; Tan, P. H.; Wang, H. Nano Res. 2015, 8, 3651. (38) Wolverson, D.; Crampin, S.; Kazemi, A. S.; Ilie, A.; Bending, S. J. ACS Nano 2014, 8, 11154. (39) Deringer, V. L.; Stoffel, R. P.; Dronskowski, R. Phys. Rev. B: Condens. Matter Mater. Phys. 2014, 89, 094303. (40) Zhang, X.; Tan, Q. H.; Wu, J. B.; Shi, W.; Tan, P. H. Nanoscale 2016, 8, 6435. (41) Chandrasekhar, H. R.; Humphreys, R. G.; Cardona, M. Phys. Rev. B 1977, 16, 2981. (42) Low, T.; Rodin, A. S.; Carvalho, A.; Jiang, Y. J.; Wang, H.; Xia, F. N.; Castro Neto, A. H. Phys. Rev. B: Condens. Matter Mater. Phys. 2014, 90, 075434. (43) Yamamoto, M.; Wang, S. T.; Ni, M. Y.; Lin, Y. F.; Li, S. L.; Aikawa, S.; Jian, W. B.; Ueno, K.; Wakabayashi, K.; Tsukagoshi, K. ACS Nano 2014, 8, 3895. (44) Li, L. Y.; Wang, W. H.; Liu, H.; Liu, X. D.; Song, Q. G.; Ren, S. W. J. Phys. Chem. C 2009, 113, 8460. (45) Saha, S.; Sinha, T. P.; Mookerjee, A. Phys. Rev. B: Condens. Matter Mater. Phys. 2000, 62, 8828. (46) Vlachos, S. V.; Lambros, A. P.; Economou, N. A. Solid State Commun. 1976, 19, 759. (47) Siapkas, D. I.; Kyriakos, D. S.; Economou, N. A. Solid State Commun. 1976, 19, 765. (48) Hong, T.; Chamlagain, B.; Lin, W. Z.; Chuang, H. J.; Pan, M. H.; Zhou, Z. X.; Xu, Y. Q. Nanoscale 2014, 6, 8978. (49) Fei, R. X.; Yang, L. Nano Lett. 2014, 14, 2884. (50) Ge, S. F.; Li, C. K.; Zhang, Z. M.; Zhang, C. L.; Zhang, Y. D.; Qiu, J.; Wang, Q. S.; Liu, J. K.; Jia, S.; Feng, J.; Sun, D. Nano Lett. 2015, 15, 4650. (51) Wu, K. D.; Chen, B.; Yang, S. J.; Wang, G.; Kong, W.; Cai, H.; Aoki, T.; Soignard, E.; Marie, X.; Yano, A.; Suslu, A.; Urbaszek, B.; Tongay, S. Nano Lett. 2016, 16, 5888.

(XDPB0603, XDB12030300). Z.W. and L.J. acknowledge financial support from the “Hundred Talents Program” of Chinese Academy of Sciences (CAS). J.L. thankfully acknowledges financial support from the CAS/SAFEA International Partnership Program for Creative Research Teams. The authors thank Prof. Jun-Wei Luo, Prof. Jun Zhang (Institute of Semiconductors, CAS), and Prof. Wei Ji (Renmin University of China) for helpful discussions.



REFERENCES

(1) Zhang, E. Z.; Wang, P.; Li, Z.; Wang, H. F.; Song, C. Y.; Huang, C.; Chen, Z. G.; Yang, L.; Zhang, K. T.; Lu, S. H.; Wang, W. Y.; Liu, S. S.; Fang, H. H.; Zhou, X. H.; Yan, H. G.; Zou, J.; Wan, X. G.; Zhou, P.; Hu, W. D.; Xiu, F. X. ACS Nano 2016, 10, 8067. (2) Nordén, B. Circular Dichroism and Linear Dichroism; Oxford University Press: USA, 1997. (3) Qiao, J. S.; Kong, X. H.; Hu, Z. X.; Yang, F.; Ji, W. Nat. Commun. 2014, 5, 4475. (4) Eaton, W. A.; Hofrichter, J. Methods Enzymol. 1981, 76, 175. (5) Pant, A.; Torun, E.; Chen, B.; Bhat, S.; Fan, X.; Wu, K.; Wright, D. P.; Peeters, F. M.; Soignard, E.; Sahin, H.; Tongay, S. Nanoscale 2016, 8, 16259. (6) Fu, Q. B.; Wang, J. M.; Lin, G. N.; Suo, H.; Zhao, C. J. Anal. Methods Chem. 2012, 2012, 728128. (7) Wang, W. L.; Li, C. Y.; Tollner, E. W.; Gitaitis, R. D.; Rains, G. C. J. Food Eng. 2012, 109, 38. (8) Mak, K. F.; Sfeir, M. Y.; Wu, Y.; Lui, C. H.; Misewich, J. A.; Heinz, T. F. Phys. Rev. Lett. 2008, 101, 196405. (9) Lee, C.; Wei, X. D.; Kysar, J. W.; Hone, J. Science 2008, 321, 385. (10) Bonaccorso, F.; Sun, Z.; Hasan, T.; Ferrari, A. C. Nat. Photonics 2010, 4, 611. (11) Dean, C. R.; Young, A. F.; Meric, I.; Lee, C.; Wang, L.; Sorgenfrei, S.; Watanabe, K.; Taniguchi, T.; Kim, P.; Shepard, K. L.; Hone, J. Nat. Nanotechnol. 2010, 5, 722. (12) Alem, N.; Erni, R.; Kisielowski, C.; Rossell, M. D.; Gannett, W.; Zettl, A. Phys. Rev. B: Condens. Matter Mater. Phys. 2009, 80, 155425. (13) Novoselov, K. S.; Jiang, D.; Schedin, F.; Booth, T. J.; Khotkevich, V. V.; Morozov, S. V.; Geim, A. K. Proc. Natl. Acad. Sci. U. S. A. 2005, 102, 10451. (14) Chhowalla, M.; Shin, H. S.; Eda, G.; Li, L. J.; Loh, K. P.; Zhang, H. Nat. Chem. 2013, 5, 263. (15) Liu, H.; Du, Y. C.; Deng, Y. X.; Ye, P. D. Chem. Soc. Rev. 2015, 44, 2732. (16) Zhou, X.; Gan, L.; Tian, W. M.; Zhang, Q.; Jin, S. Y.; Li, H. Q.; Bando, Y.; Golberg, D.; Zhai, T. Y. Adv. Mater. 2015, 27, 8035. (17) Zeng, H. L.; Dai, J. F.; Yao, W.; Xiao, D.; Cui, X. D. Nat. Nanotechnol. 2012, 7, 490. (18) Aslan, O. B.; Chenet, D. A.; van der Zande, A. M.; Hone, J. C.; Heinz, T. F. ACS Photonics 2016, 3, 96. (19) Li, H. Q.; Wang, X.; Xu, J. Q.; Zhang, Q.; Bando, Y.; Golberg, D.; Ma, Y.; Zhai, T. Y. Adv. Mater. 2013, 25, 3017. (20) Xue, D. J.; Tan, J. H.; Hu, J. S.; Hu, W. P.; Guo, Y. G.; Wan, L. J. Adv. Mater. 2012, 24, 4528. (21) Shi, G. S.; Kioupakis, E. Nano Lett. 2015, 15, 6926. (22) Appelbaum, I.; Li, P. K. Phys. Rev. B: Condens. Matter Mater. Phys. 2016, 94, 155124. (23) Mukherjee, B.; Cai, Y. Q.; Tan, H. R.; Feng, Y. P.; Tok, E. S.; Sow, C. H. ACS Appl. Mater. Interfaces 2013, 5, 9594. (24) Taniguchi, M.; Johnson, R. L.; Ghijsen, J.; Cardona, M. Phys. Rev. B: Condens. Matter Mater. Phys. 1990, 42, 3634. (25) Vaughn, D. D.; Sun, D.; Levin, S. M.; Biacchi, A. J.; Mayer, T. S.; Schaak, R. E. Chem. Mater. 2012, 24, 3643. (26) Xue, D. J.; Liu, S. C.; Dai, C. M.; Chen, S. Y.; He, C.; Zhao, L.; Hu, J. S.; Wan, L. J. J. Am. Chem. Soc. 2017, 139, 958. (27) von Rohr, F. O.; Ji, H. W.; Cevallos, F. A.; Gao, T.; Ong, N. P.; Cava, R. J. J. Am. Chem. Soc. 2017, 139, 2771. (28) Feng, Y. Q.; Zhou, W.; Wang, Y. J.; Zhou, J.; Liu, E. F.; Fu, Y. J.; Ni, Z. H.; Wu, X. L.; Yuan, H. T.; Miao, F.; Wang, B. G.; Wan, X. G.; 14982

DOI: 10.1021/jacs.7b06314 J. Am. Chem. Soc. 2017, 139, 14976−14982