N‐Type Doping and Passivation of Black Phosphorus Using Pentacene for Infrared Polarization‐Sensitive Photodetection and Imaging

Few‐layer black phosphorus (FL‐BP) shows great potential in infrared polarization‐sensitive photodetection due to its narrow bandgap and strong in‐plane anisotropy. However, FL‐BP is easy to be oxidized and degraded, which severely restricts its applications. Herein, a simple and scalable in situ interfacial passivation method based on pentacene film is reported to enhance the ambient stability of FL‐BP as well as achieve n‐type doping for improved polarization‐sensitive photoimaging. High crystal quality FL‐BP flakes with intrinsic anisotropy are prepared by electrochemical exfoliation. The morphology and Raman spectra of FL‐BP/pentacene exhibit excellent stability 28 days after passivation. In addition, the hole mobility is significantly enhanced and dark current is substantially suppressed due to the interfacial passivation and n‐type doping effect. The FL‐BP/Pentacene based photodetector shows a broadband response from visible to near‐infrared with good stability. The photoresponsivity under 915 nm illumination is calculated as 2.2 mA W−1 with an anisotropy ratio of 1.29. Furthermore, the polarization imaging under the irradiation at zigzag and armchair directions directly reveals the strong infrared polarization‐sensitive imaging capability of FL‐BP/pentacene. This work also opens up new avenues to enhance the photoresponse and stability of various optoelectronics based on similarly unstable 2D materials.


Introduction
Infrared polarization-sensitive photodetectors exhibit significant advantages in acquiring target feature information and suppressing interference, [1] hence present broad prospects for industrial inspection, environmental monitoring and remote sensing, etc. [2] However, traditional infrared polarization-sensitive photodetectors inevitably suffer from their large volumes due to the complicated optical components and cooling equipment.The appearance of twodimensional materials with intrinsic anisotropy provides an effective way to obtain the highly integrated infrared polarization-sensitive photodetectors. [3]everal eminent materials including ReS 2 , [4] CaTe, [5] GeSe 2 , [6] ZrS 3 , [7] and antimonene [8] have been explored in the application of polarization-sensitive photodetection.However, the band structure, light absorption rate and carrier mobility of such materials severely restrict their application and performance in infrared polarized-light detection.As a novel 2D material with in-plane anisotropy, BP has a bandgap range of 0.3-2.0eV that can be tuned by adjusting the thickness.It exactly complements the gap between the bandgap of graphene and transition metal dichalcogenides (TMDCs), and overcomes the limitations of photoresponse spectral ranges. [9]Unlike most multilayer TMDCs with indirect bandgaps, BP with different thicknesses always possesses a direct bandgap. [10]The different light absorption capacity at zigzag and armchair directions of FL-BP guarantees its significant advantages in room temperature infrared polarization-sensitive photodetection. [11]The preparation of large size FL-BP flakes is a key requirement for industrial applications.However, the chemical vapor deposition method is challenging with difficulties in process control and results in low crystal quality. [12,13]echanical stripping exhibits low efficiency, [14,15] while liquidphase exfoliation leads to small material dimensions. [16,17][20] However, the action of ion intercalation may cause potential damage to 2D nanosheets.
Although FL-BP crystals have many advantages, they are easy to be oxidized and degraded under environmental conditions, which has become a critical problem restricting their research and application. [21]Therefore, the passivation protection of BP has aroused much attention recently, in which the approaches can be divided into three types: physical encapsulation, [22][23][24] chemical functionalization, [25] and doping. [26,27]Physical encapsulation is a common used method due to its advantages of facile operation and large material selectivity. [28]Al 2 O 3 film prepared by atomic layer deposition(ALD) as passivation layer has been explored as a promising way to protect BP. [29] However, in the process of Al 2 O 3 deposition, the oxygen precursor always leads to the degradation of BP.Besides, directly encapsulating BP with few-layer 2D materials (graphene [30] or h-BN [31] ) also has been investigated, but it is hard to achieve the scalable fabrication.Recently, the organic small molecule pentacene (C 22 H 14 ) has emerged as a promising hole-transport layer in perovskite optoelectronics, [32,33] which can be easily deposited on the surfaces of various materials.The pentacene film prepared by evaporation exhibits uniform particle size, compact arrangement, and dense structure.It not only passivates the surface defects in perovskite, but also effectively modulates the energy band alignment at the interface, and facilitates the hole migration.However, the application of pentacene for surface passivation and modulation of unstable 2D materials is still rarely reported.
Herein, we demonstrate a simple and scalable in situ interfacial passivation method based on pentacene film to enhance the ambient stability of FL-BP, which is further used to prepare a long-term stable infrared polarization-sensitive photodetector.First, electrochemical exfoliation method was used to prepare FL-BP flakes with high crystal quality, strong in-plane anisotropy and lateral size of 70 μm.The pentacene film was then deposited on the surface of FL-BP flakes for passivation, and electron doping was realized.The photodetector based on FL-BP/Pentacene exhibits excellent photodetection performance and ambient stability, along with the capability of infrared polarization-sensitive imaging.Our work also opens up new avenues to enhance the photoresponse and stability of various optoelectronics based on similar unstable 2D materials.

Results and Discussion
Figure 1a displays a photograph of FL-BP nanosheets dispersed in DMF solution with a distinct Tyndall effect, in which the nanosheets are prepared by electrochemical exfoliation (Figure S1, Supporting Information).The intercalation reaction was conducted in a typical two-electrode system containing 0.6 mg mL −1 Tetrabutylammonium acetate (CH 3 COOTBA) /dimethylformamide (DMF) solution, in which the bulk BP crystal and platinum sheet were used as the cathode and anode, respectively. [34]A large number of TBA + ions were inserted in the internal layer of the BP crystal with applied bias potential, resulting in a significant volume expansion while weakening the van der Waals forces between the layers. [35]However, massive ion insertion may cause potential damage to FL-BP nanosheets.Electrochemical exfoliation using TBAHSO 4 [35] and phytic acid [36] is always coupled with the formation of hydrogen bubbles at the BP interface.Although the exfoliation yield is promoted, the domain size is substantially reduced.On the contrary, the absence of hydrogen bubbles production in the CH 3 COOTBA intercalation process avoids the destruction of P─P bonds, [34] thereby facilitating the synthesis of FL-BP nanosheets with a large domain size.The details of experimental process are shown in Experimental Section.
The well dispersed FL-BP nanosheets in DMF then were transferred onto the SiO 2 /Si substrate without overlapping (Figure S2, Supporting Information).Figure 1b displays a typical optical microscopy image of the as-prepared FL-BP with a lateral size ≈70 μm.X-ray photoelectron spectroscopy (XPS) was used to investigate the chemical states of FL-BP crystals as shown in Figure 1c and Figure S3 (Supporting Information).The two sharp peaks located at ≈129.6 and 130.5 eV can be ascribed to the 2p 3/2 and 2p 1/2 state of P, which is agree well with the previously reported values of BP nanosheets, [37] indicating an excellent crystal quality of the exfoliated FL-BP nanosheets.Additionally, no peaks of oxidized BP (i.e., PO x ) groups was observed in freshly as-obtained FL-BP crystals at around 133.0 eV. [38]Raman spectra of FL-BP nanosheets under 532 nm excitation is shown in Figure 1d.As expected, the spectra show three characteristic peaks at ≈360.0, 439.0 and 466.9 cm −1 , which can be attributed to the out-of-plane (A g 1 ) and in-plane (B 2g and A g 2 ) vibration modes, respectively. [39]The intensity of A g 1 mode is related to the covalent functionalization of BP molecule.The intensity ratio of A g 1 /A g 2 can be used for qualitative analysis of the oxidation degree of FL-BP crystals.A ratio of A g 1 /A g 2 between 0.4 and 0.6 can prove that the FL-BP nanosheets are unoxidized. [40]The value of as-obtained FL-BP flakes is 0.53, indicating that our preparation method effectively avoids the oxidation.Angle-resolved polarized Raman (ARPR) was performed on FL-BP flake by rotating the laser's polarization angle with 30°steps from 0°to 360°t o further identify the anisotropic in-plane lattice arrangement.Figure 1e displays the intensity mapping of Raman as a function of polarization angle.The three characteristic peaks exhibit clear polarization angle-dependent properties and periodicity.Besides, isolating A g 2 separately allows for a more intuitive observation of its angle-dependent characteristics (Figure S4, Supporting Information).The anisotropic crystal structure of BP guarantees its ability of polarization-sensitive photodetection. [41]he lattice structure of FL-BP flakes after delamination was characterized by high-resolution transmission electron microscopy (HRTEM).A typical FL-BP flakes on TEM grid is shown in Figure S5 (Supporting Information).Although the high energy electron beam irradiation induces a lot of defects in FL-BP films, the high-resolution TEM image (Figure 1f) clearly reveals that the nanosheet possesses a single-crystal structure with the interplanar spacing of 0.26 nm consistent with the d-spacing of (111) planes. [42]No significant dislocation can be observed in the asobtained FL-BP flakes.High-angle annular dark-field (HAADF) image of FL-BP nanosheet is shown in Figure 1g.Besides, the collected element mapping of P shows a uniform distribution with red color in Figure 1h and directly reveals the homogeneous and phase-pure nature of as-obtained FL-BP flakes.
Previous studies have shown that FL-BP is prone to be oxidated under ambient conditions. [43,44]Herein, a layer of pentacene film was deposited on the surface of bare FL-BP flakes by vacuum thermal evaporation to improve the ambient stability.The structure diagram after passivation is shown in Figure 2a.The surface morphology of freshly prepared FL-BP was observed after transferred to SiO 2 /Si substrate.Morphology variation over time is recorded as shown in Figure 2b,c.Freshly prepared FL-BP flakes exhibit a clean surface, clear edges and no bubbles.Two days later, visible bubbles appeared on the surface of bare FL-BP because of the producing of P x O y in oxidation process (Figure S6, Supporting Information). [45]After 6 days, the number of bubbles increases rapidly, and some of bare FL-BP nanosheets disappeared, only the boundary can be observed.
The thickness of pentacene film as passivation layer is set at 15 nm.The evaporation rate of pentacene is as slow as 0.2 Å s −1 , which is conducive to the transverse spreading of crystalline grains on the surface.The AFM scanning was conducted on an area of 1.5 × 1.5 μm 2 as shown in Figure 2d.The pentacene film prepared by evaporation exhibits high crystalline quality, with a particle size of ≈136 nm and a compact arrangement.The root-mean-square (RMS) roughness of pentacene film surface is 3.1 nm.The dense film effectively isolates water and oxygen molecules from the surrounding environment.Meanwhile, the hydrophilic or hydrophobic properties of the substrate greatly affect the microstructure of pentacene film.If the substrate is hydrophilic, pentacene will grow in a flat patch shape rather than rod-like. [46]The microstructure of pentacene grown on the surface of FL-BP in our preparation exhibits a flat patch shape, indicating the hydrophilicity of the FL-BP surface, which is obviously unfavorable for its stability.After the freshly prepared bare FL-BP was transferred to the substrate, it was quickly put into the thermal evaporation system to minimize the influence of ambient conditions during the experiments.As shown in Figure 2b,e, the color of the substrate changed from purplish red to bluishpurple under the microscope after pentacene film deposited.
The morphologic evolution of FL-BP/Pentacene after passivation over time then was studied.Freshly prepared FL-BP/Pentacene nanosheets had a clean and smooth surface (Figure 2e).The oxidation of FL-BP/Pentacene nanosheets is not significant after 4, 8, and 22 days (Figure S7, Supporting Information).The surface of the FL-BP/Pentacene is clean and there are no noticeable bubbles.Until day 28, a few slight bubbles appeared on the surface of the FL-BP/Pentacene nanosheets, but its structure is still intact.Furthermore, an aging analysis of the FL-BP flakes structure before and after passivation was carried out by Raman spectroscopy.During the aging test, the wavelength and intensity of the excitation laser source are maintained.Since the optical response of the FL-BP is highly anisotropic, the polarization direction of the excited laser has a significant effect on the Raman intensity.We keep the direction of FL-BP flake constant with the laser polarization angle in the experiments to ensure that the change of peak intensity exactly reflect the degradation of the flake instead of the variation of polarization angle.The long-period Raman spectroscopy of bare FL-BP flakes are shown in Figure 2g.The intensity of three characteristic peaks of bare FL-BP flakes decreased significant after 2 days.By the sixth day, they even disappeared completely.On the contrary, the position of the characteristic Raman peaks of FL-BP/Pentacene didn't change and its intensity didn't decrease significantly for up to 22 days.After 28 days, the intensity of characteristic peaks just decreased slightly (Figure 2h).Another previous study reported that the intensity of A g 1 mode decreased with increasing covalent functionalization of BP molecule due to the breakdown of the phosphorus bond in the interlayer. [47]The normalized intensity of A g 1 mode before and after passivation FL-BP flakes over time is shown in Figure 2i to qualitatively analyze BP degradation.The blue curve represents the decay of the bare FL-BP A g 1 mode over time, which is normalized based on the A g 1 mode intensity of the bare FL-BP at day 0. The red curve represents the decay of the FL-BP/Pentacene A g 1 mode over time, which is normalized based on the A g 1 mode intensity of the FL-BP/Pentacene at day 0. The A g 1 intensity of bare FL-BP flakes dropped sharply and disappeared almost completely on the sixth day, and that means it oxidizes quickly and has poor stability.On the contrary, the intensity of the FL-BP/Pentacene A g 1 mode remained above 90% and stabilized over a period of up to 22 days.This indicates that the passivation method as-adopted can effectively isolate water molecule and oxygen from FL-BP nanosheets in the ambient and improve the stability of FL-BP/Pentacene flakes, which lay a foundation for the subsequent preparation of optoelectronic devices based on FL-BP flakes.
Besides, the Raman spectrum of bare pentacene film and FL-BP/Pentacene hybrid film is also collected (Figure S8, Supporting Information).There is a strong molecular vibration peak at 1371 cm −1 , and two weak peaks at 1533 and 1579 cm −1 belonging to benzene ring stretching.The vibration peaks at 1158 and 1171 cm −1 correspond to the E 2g mode of benzene ring, which is caused by C-H deformation.These Raman peaks are agree well with the previous reports for pentacene film. [48]Due to strong interlayer coupling between pentacene and FL-BP flakes, the Raman intensity of FL-BP/Pentacene hybrid film is significant weakened, compared to the bare pentacene.Photoluminescence (PL) spectra can be used to confirm the optical bandgap of individual materials and the transport of photoinduced charge across heterogeneous interface.Figure 3a shows the PL spectrum of pentacene film and FL-BP/Pentacene film under 532 nm laser excitation, and the fixed laser power is 1 mW.The inset in Figure 3a is the region where the PL spectra were collected.Due to the limitations of the equipment's operating range, we cannot collect the PL peak of FL-BP, which has been reported appearing in mid-infrared range.The PL peak of bare pentacene appears at 672 nm, corresponding to a band gap of ≈1.85 eV.This peak corresponds to the singlet transition (S1→S0) of pentacene. [49]owever, the PL intensity was significantly decreased or even disappeared in FL-BP/Pentacene film.This indicates that exciton migration occurs at the interface between pentacene and FL-BP flake, leading to the quenching of PL. [50] In order to study the doping effect of pentacene on FL-BP flakes, we prepared a typical three-terminal field effect transistor (FET) device.The output characteristic curve of FL-BP device before and after passivation under 0 V gate bias is shown in Figure 3b.We can observe that the output current density of FL-BP/Pentacene FET is significantly reduced, compared with bare FL-BP FET.For photodetectors, the reduction of dark current is beneficial to improve the performance of photodetection.Figure 3c shows the transfer characteristic curve of FL-BP devices before and after pentacene passivation.We can clearly observe the device transition from p-type to ambipolar behavior after chemical doping using pentacene.When the gate voltage varied from −60 to 60 V with a bias voltage of 2 V, the source-drain current increases along both negative and positive scanning directions, manifesting hole and electron transport, respectively.In addition, the current increases much faster in the negative gate voltage region than in the positive gate voltage region, indicating a bipolar behavior dominated by hole transport. [51]The threshold voltage of FET devices before and after passivation is negatively offset from −22 to -36 V (Figure S9, Supporting Information).At the same time, the current minimum is negatively offset.All these indicate that pentacene modified FL-BP flakes exhibit a significant n-type doping effect, which can be ascribed to the interfacial electron transfer caused by large difference in work function between pentacene and FL-BP.
In the transfer curve, the increasing slope of the linear region on the negative gate voltage side proves that the hole mobility of the modified FET device is improved.The reason for this phenomenon is that the modification of pentacene inhibits the oxidation reaction and reduces the surface scattering. [52]The improvement of current and mobility lays a foundation for the subsequent construction of high-performance photodetectors.Noteworthy is that the current density collected from bare pentacene devices (Figure S10, Supporting Information) is 5 orders of magnitude lower than FL-BP/Pentacene devices, indicating that the performance improvement can be attributed to the doping effect of the FL-BP channel rather than the conductivity of pentacene films.Regarding the reason for the above phenomenon, the doping mechanism can be explained by the change of Fermi level.Prior to the application of source-drain bias, the alignment of energy levels between FL-BP flakes and the metal electrodes reaches an equilibrium as the physical contact is built.A non-negligible Schottky barrier between the metal contact and FL-BP flakes can be observed as shown in Figure 3e.The FL-BP flakes has a small bandgap, and the transistors always show p-type behavior. [53]In this case, the pentacene film serves as an electron donator to regulate the Fermi level of FL-BP flakes, which leads to the conversion of FL-BP from p-type to ambipolar.In other words, the Fermi level is pulled toward a higher energy due to doping, which subsequently narrows the Schottky barrier, reduces the contact resistance, and enhances the electrical performance of the FL-BP/Pentacene devices. [54]o explore the photodetection performance of FL-BP/Pentacene flakes after passivation, the photoconductive devices were prepared on the SiO 2 /Si substrate as schematically illustrated in Figure 4a.Details on devices fabrication can be found in the Experiments Section.The thickness of FL-BP flake in the channel is revealed as 22.3 nm as shown in the AFM image (Figure 4b).All photoresponse tests in this work were performed under ambient conditions.Laser irradiation further accelerates the oxidation reaction of bare FL-BP flakes under ambient conditions, resulting in an extremely unstable test signal of photodetector based on bare FL-BP flakes (Figure S11, Supporting Information).Figure 4c shows the I-V curve of the FL-BP/Pentacene photodetector at a bias of −1 to 1 V, where the inset is an optical photograph of the device.The wavelength of the incident laser is 915 nm and the intensity are varying from 1.72 to 277.6 μW.The current increases gradually with the increase of incident light power, and exhibits an obvious response even under weak light intensity conditions.In order to explore the working range of the photodetector, a series of lasers with different wavelengths (520, 915, 1310 nm) were used to test the photoresponse of the FL-BP/Pentacene photodetector at 1 V. From Figure 4d, it can be observed that the photodetector exhibits an obvious photoresponse to incident light of different wavelengths, covering the visible to near-infrared range.Although the laser power is approximately the same, the photocurrent decreases with increasing wavelength.The reason is that the photon energy carried by different laser wavelengths varies, which affects the number of photo-generated carriers excited by the laser.
Subsequently, the photoresponse under different power densities was investigated using 915 nm laser (Figure 4e).When the incident power varies between 0.87 and 277.6 μW, the photocurrent increases with the increase of the laser power.We further study the relationship among responsivity, photocurrent and incident power density, as shown in Figure 4f.Obviously, the photocurrent increases with increasing incident optical power density.The responsivity is defined as the rate of photocurrent and incident power, the formula is R = I ph /P.Due to the suppressed scattering between photoexcited charge carriers, the responsivity is significantly improved when the incident optical power density is reduced.The responsivity reaches 2.2 mA W −1 with an incident optical power density of 39 W cm −2 .Due to equipment limitations, we are unable to fully exploit the potential of the photodetector.Even though the responsivity is comparable to previously reported FL-BP photodetectors (Table S1, Supporting Information).The low responsivity may be attributed to the defects caused by the electrochemical exfoliation process.In addition, the cleanliness of the contact interface and other molecules adsorbed on the channel also affect responsivity.For instance, CH 3 COOTBA molecules adsorbed on the surface of FL-BP flakes during electrochemical intercalation are thought to be the main factor of trapping holes.Additionally, we have tested the transfer characteristics of the FET device as a function of optical intensity (Figure S12, Supporting Information).The curve of the FL-BP/Pentacene device shifts not only vertically but also horizontally with increasing optical intensity, indicating the presence of a localized charge trapping-induced photogating effect.To assess the response speed of the FL-BP/Pentacene photodetector, a switching laser was used to test the transient photoresponse, as shown in Figure 4g.The response (decay) time is defined as the time between the normalized response value increasing (decreasing) from 10% (90%) to 90% (10%).The response and decay times of as-obtained photodetector are 62 and 41 ms under 915 nm illumination.In Figure 4h, the incident laser is switched on/off repeatedly every 5s at a bias voltage of 1 V to measure the temporal response of the photodetector in the near infrared region.The longitudinal axis is the normalized photocurrent to evaluate the long-term operation stability of the FL-BP/Pentacene photodetector.The two insets in Figure 4h are the magnified images of the first 7 cycles and the last 7 respectively.It is obvious that the current switching behavior of the device is maintained well for up to 1300 s, indicating that the device can achieve reversible transition between high and low conduction status.Besides, the photoresponse of the FL-BP/Pentacene photodetector is monitored for a long time, and its normalized photocurrent curve varying with the date is shown in Figure 4i.The stability of the photoresponse remained well until a significant decline occurred on the 16th day.The stability of the device under ambient conditions has been greatly improved compared with the pure BP photodetector, [34] which is very important for the long-term practical application of the device.Compared to previously reported FL-BP flakes passivated using different mechanisms (Table S2, Supporting Information), our approach utilizing pentacene encapsulation demonstrates the advantages of simplicity and scalability, while also exhibits excellent stability.This provides a promising prospect for the industrialization of photodetectors based on FL-BP flakes.
Due to the additional polarization angle detection factor, polarization-sensitive photodetectors can obtain more dimensional optical signals from the target. [55]Figure 5a shows the schematic diagram of the polarization-sensitive photodetection test.The incident light is converted to linearly polarized light by a polarizing plate and a half-wave plate, and then irradiated to the photosensitive region.The angle of polarized light is tuned by rotating half-wave plate, and the position of photodetector remains unchanged in the test system.Figure 5b shows the real-time normalized photoresponse of the device under the rotating polarized light irradiation (blue curve).Obviously, photocurrent exhibits evident periodicity and excellent stability with the variation of polarization angle.Figure 5c shows polarizationdependent photocurrent and fitting curve under 1 V bias and 915 nm laser irradiation.The photocurrent anisotropy ratio (I ph-max /I ph-min ) is calculated as 1. 29.
In order to further demonstrate the polarization imaging ability of the FL-BP/Pentacene photodetector, the device is used as a single pixel in the imaging sensor.Figure 5d schematically illustrates the infrared polarization imaging test system.The light passes through a mask with the hollow letter "MSE", a polarizing plate and a half-wave plate, and then irradiate on the photode-tector.The mask can be controlled by a 2D motorized stage and move continuously line by line from bottom to top.The moving step of motorized stage is set as 300 μm, which is also the spacing between pixels.As a result, the pattern of the measured object can be obtained by collecting the photocurrent of the device simultaneously.Figure 5e displays the "MSE" polarization image (216*86 pixels, after 5x interpolation) at zigzag and armchair direction, respectively.The images directly reveal the tremendous difference in photocurrent of the devices under the irradiation of the polarized light at zigzag and armchair direction, demonstrating the powerful polarized light imaging capability of the FL-BP/Pentacene photodetector.As the device exhibits high responsivity and excellent operating stability under polarized light, we believe it has great potential in the field of highly integrated infrared polarized photodetection.

Conclusions
In summary, we demonstrate a simple and scalable in situ interfacial passivation method based on pentacene film to enhance the ambient stability of FL-BP and achieve n-type doping, which is further used to prepare a long-term stable infrared polarization-sensitive photodetector.FL-BP flakes with a lateral size of 70 μm are prepared by electrochemical exfoliation using TBA + cations as intercalation ion.TEM images reveals that the FL-BP flakes have excellent crystal quality, and ARPR spectrum indicates strong in-plane anisotropy.The passivation is carried out by pentacene film on the surface of FL-BP flakes.The stability of FL-BP flakes after passivation are up to 28 days according to the surface morphology and Raman monitoring.The threshold voltage of the FET based on FL-BP/Pentacene migrates negatively and the dark current decreases, which indicates that the FL-BP flakes had a n-type doping effect by pentacene.The photodetector based on FL-BP/Pentacene has a broadband response from visible to near-infrared, with a responsivity of 2.2 mA W −1 under the illumination of 915 nm and a good photoresponse stability.Due to the strong in-plane anisotropy of FL-BP, the photocurrent of the photodetector based on FL-BP/Pentacene has obvious periodicity and high stability, where the anisotropy ratio is calculated as 1.29.The infrared polarization imaging sensing ability of FL-BP flakes is verified by a single pixel imaging system excited by infrared polarized laser.The results directly reveal the great potential of FL-BP/Pentacene as a high-performance and ambient stable infrared polarization-sensitive photodetector.

Experimental Section
Electrochemical Exfoliation of BP Flakes: BP flakes were prepared by electrochemical exfoliation in a typical two-electrode system.BP bulk (Six-Carbon >99.999%, about 20 mg) and platinum sheet electrodes were used as cathode and anode, respectively, with a spacing of 2 cm.The electrolyte were obtained by dissolving 0.006 g tetra-n-butyl-ammonium acetate (CH 3 COOTBA) (Aladdin 97%) in 10 ml DMF solution.In order to ensure adequate electrolytic process, intercalation time was set as 30 min at 20 V.After electrolysis was complete, the exfoliation solution was transferred to the centrifugal tube for centrifugation.The centrifugal speed was 1000 rpm for 10 min.Then the unexfoliated and thick-layered BP crystals were removed and the supernatant was retained.The resulting FL-BP/DMF solution can be transferred to the substrate for characterization and device preparation.
Devices Fabrication: A 30 mm diameter container was filled with deionized water, and FL-BP flakes dispersion was dripped onto the surface by pipette (10 μL).With the assistant of surface tension of the water, the FL-BP flakes were dispersed at the water-air interface and then transferred to the substrate.The substrate was n-doped silicon wafer covered with 285-nmthick silicon dioxide.The substrate with FL-BP flakes then was annealed in a tube furnace filled with the mixture of argon/hydrogen (900/100 sccm).Cr/Ag (8/50 nm) electrode was deposited on the substrate by electron beam evaporation using Cu grid as a physical mask.A passivation layer of 15 nm pentacene film was deposited on the device surface by vacuum thermal evaporation at the evaporation rate of 0.2 Å s −1 to obtain the FL-BP/Pentacene based transistors.
Characterization and Measurement: All optical images were carried out by an optical microscope (KEYENCE, VHX-1000).The oxidation of FL-BP flakes was analyzed by X-ray photoelectron spectroscope (Shimadzu Axis Ultra DLD), and the binding energy of C 1s peak (284.6 eV) was used for energy calibration.Raman and PL spectra were collected using a Raman spectroscopy (LabRAM HR800) with an excitation wavelength of 512 nm.
The lattice structure and element mapping of FL-BP flakes were characterized by field emission transmission electron microscopy (FETEM, Tecnai G2 F30, FEI).The thickness of FL-BP flakes and the morphology of pentacene films were observed by atomic force microscope (Bruker Dimension EDGE).The photoelectric properties of the device were collected by Keithley 2636B digital source meter coupled with four-probe station (SCG-O-2, Semishare Co., Ltd) on a damped vibration isolation optical platform (HGZZ0906Z Wuhan Red Star Yang Science and Technology Co., Ltd) under ambient conditions.The incident light was driven by a laser controller (ITC 4001, Thorlabs) and was modulated by a waveform generator (Keysight 33210A).The polarized imaging capability was verified by a single pixel scanning system (Metatest E2-SCS).

Figure 1 .
Figure 1.a) Physical image of FL-BP flakes dispersion as-obtained by intercalation.b) Optical image of a FL-BP flake on 285 nm SiO 2 /Si substrate.Scale bar: 10 μm c) Raman spectra of FL-BP flake.d) Polarized Raman intensity mapping of FL-BP flakes as a function of polarization angle of excited laser and wave number.e) High-resolution XPS spectra of P 2p states.f) High-resolution TEM image of an exfoliated FL-BP nanosheet.g, h) HAADF scanning TEM image of FL-BP nanosheet and the Element mapping of P from the corresponding area.

Figure 2 .
Figure 2. a) Structure diagram of FL-BP flakes passivated by pentacene film.b) 0 day and c) 6 days later optical image of fresh FL-BP flake.d) AFM image of the microstructure of pentacene film.e) 0 day and f) 28 days later optical image of FL-BP/Pentacene flake.g) Raman spectrum evolution of fresh FL-BP flakes.h) Raman spectrum evolution of FL-BP/Pentacene flakes.i) The normalized peak intensity of A g 1 mode as a function of time.

Figure 3 .
Figure 3. a) PL spectra of pentacene film and FL-BP/Pentacene nanosheet (inset: FL-BP flake with pentacene patches.Scale bar is 5 μm).(b) I-V characteristics of FL-BP transistor before and after passivation.c) Transfer characteristics obtained from FL-BP and FL-BP/Pentacene FET at V sd = 2 V. d) Schematic of the proposed charge transfer at BP/pentacene interface.Band diagram of the e) fresh and f) pentacene-doped FL-BP FETs at equilibrium.

Figure 4 .
Figure 4. a) Schematic diagram of a FL-BP/Pentacene based photodiode.b) AFM image of FL-BP covered with pentacene in the channel.c) I-V characteristics of FL-BP/Pentacene based photodetector with different intensities under 915 nm laser.The inset is an optical morphology of FL-BP/Pentacene phototransistor.Scale bar is 40 μm.d) Photoresponse of the photodetector under different laser wavelengths from 520 to 1310 nm.e) Photoresponse at 915 nm with power varying between 0.87 to 277.6 μW.f) The photocurrent and responsivity as functions of the laser illumination power density at wavelength of 915 nm.g) The photocurrent normalized rise and fall curves in a period of measurement.h) Long-term stability of normalized photocurrent in laser periodic on/off illumination.i) Normalized photocurrent stability curve under different days.

Figure 5 .
Figure 5. a) Schematic diagram of polarized photodetection system.b) Normalized photocurrent dependent on polarization angle for V ds = 1 V. c) Photocurrent polar coordinate diagram and the corresponding fitting curve at 915 nm, V ds = 1 V. d) Schematic diagram of a single pixel polarization imaging system.e) Imaging patterns of "MSE" under 915 nm polarized light incident at zigzag and armchair direction, respectively.