Nanoimprinted Perovskite NanogratingPhotodetector with Improved EfficiencyHonglei Wang,† Ross Haroldson,‡ Balasubramaniam Balachandran,§ Alex Zakhidov,∥ Sandeep Sohal,∥

Julia Y. Chan,¶ Anvar Zakhidov,‡,⊥ and Walter Hu*,†,#

†Department of Electrical Engineering, ‡Department of Physics, §Department of Mechanical Engineering, and ¶Department ofChemistry and Biochemistry, The University of Texas at Dallas, Richardson, Texas 75080, United States∥Department of Physics, Texas State University, San Marcos, Texas 78666, United States⊥Energy Efficiency Center, National University of Science and Technology, MISiS, Moscow 119049, Russia#ASIC and System State-Key Laboratory, Institute of Microelectronics, Fudan University, Shanghai 200433, China

*S Supporting Information

ABSTRACT: Recently, organolead halide-based perov-skites have emerged as promising materials for optoelec-tronic applications, particularly for photovoltaics, photo-detectors, and lasing, with low cost and high performance.Meanwhile, nanoscale photodetectors have attractedtremendous attention toward realizing miniaturized opto-electronic systems, as they offer high sensitivity, ultrafastresponse, and the capability to detect beyond the diffractionlimit. Here we report high-performance nanoscale-pat-terned perovskite photodetectors implemented by nano-imprint lithography (NIL). The spin-coated lead methyl-ammonium triiodide perovskite shows improved crystal-linity and optical properties after NIL. The nanoimprinted metal−semiconductor−metal photodetectors demonstratesignificantly improved performance compared to the nonimprinted conventional thin-film devices. The effects of NILpattern geometries on the optoelectronic characteristics were studied, and the nanograting pattern based photodetectorsdemonstrated the best performance, showing approximately 35 times improvement on responsivity and 7 timesimprovement on on/off ratio compared with the nonimprinted devices. The high performance of NIL-nanogratingphotodetectors likely results from high crystallinity and favored nanostructure morphology, which contribute to highermobility, longer diffusion length, and better photon absorption. Our results have demonstrated that the NIL is a cost-effective method to fabricate high-performance perovskite nanoscale optoelectronic devices, which may be suitable formanufacturing of high-density perovskite nanophotodetector arrays and to provide integration with state-of-the-artelectronic circuits.

KEYWORDS: perovskite, nanoimprint lithography, nanogratings, photodetector, nanoscale, crystallinity

Photodetectors that can convert optical signals into anelectrical signal play an important role in a variety ofapplications, such as optical communication, digital

imaging, and environment monitoring.1−4 Nanoscale photo-detectors enable the opportunity to integrate high-densitydevices with state-of-the-art integrated circuits, while simulta-neously offering high sensitivity, ultrafast response due to highsurface-to-volume ratio, and reduced conductive channeldimensions.5−7 Imaging systems with nanoscale pixels arecapable of achieving resolution beyond the diffraction limit.8

Materials compatible with conventional silicon electronics orflexible substrates are especially attractive. Most of thenanostructured photodetectors reported to date are inorganic-based materials such as carbon nanotubes,9 group II−VI,10,11and group III−V compounds.12 These materials are synthesized

via relatively expensive methods such as vapor−liquid−solid orepitaxy, which have difficulty in system level integration.13 Thedevelopment of their practical applications has thus beenhampered.Recently, hybrid organolead halide perovskites, such as

CH3NH3PbI3 type (lead methylammonium triiodide: MAPbI3),have emerged as a promising material for advancedoptoelectronic devices due to their advantages of long carrierdiffusion length, high carrier mobility, and large opticalabsorption coefficient over a broad wavelength range.14−17

Received: August 16, 2016Accepted: November 15, 2016Published: November 15, 2016

Artic

lewww.acsnano.org

© 2016 American Chemical Society 10921 DOI: 10.1021/acsnano.6b05535ACS Nano 2016, 10, 10921−10928

Moreover, these hybrid organic−inorganic perovskites aresolution processable for cost-effective fabrication. Amongcurrently reported perovskite photodetectors,4,18−28 the planarmetal−semiconductor−metal (MSM) structure has attractedinterest due to its simplicity and integration advantages. Thefirst perovskite MSM photodetector was reported by Hu etal.,24 who utilized an ITO−perovskite−ITO structure andachieved a photoresponsivity of 3.49 A/W and 0.0367 A/W at365 and 780 nm respectively, with a 3 V bias. Since then, efforthas been made to further improve the MSM photodetectorperformance, which is greatly influenced by the charge carrierdiffusion length and mobility and could be improved withbetter morphology and higher crystalline quality. Lian et al.4

and Saidaminov et al.25 have reported a single-crystallineperovskite photodetector with significantly improved respon-sivity compared to polycrystalline counterparts.Toward the study of nanoscale hybrid organic−inorganic

perovskite, Horvath et al.29 reported the synthesis of MAPbI3nanowires with a mean diameter of 50 and 400 nm and lengthup to 10 μm using a simple slip-coating process. The as-fabricated MSM photodetector, however, shows a lowresponsivity of 5 mA/W. Zhuo et al.30 also reported thesynthesis of porous perovskite nanowires using Pb-containingprecursors. However, these methods lack precise control of thenanowire position and dimensions. Meanwhile, organoleadhalide perovskite is not compatible with conventional photo-lithography that is widely used to fabricate nanoscale electronicand optoelectronic devices and circuits.31 Therefore, creatingnanoscale perovskite optoelectronic devices by well-control-lable, low-cost fabrication methods while simultaneouslymaintaining high optoelectronic performance remains a majorchallenge.In this work, we report high-performance nanoscale

perovskite photodetectors based on highly crystalline nano-grating structures implemented by nanoimprint lithography(NIL). NIL has been previously applied to pattern organicmaterials to obtain highly ordered nanostructure morphologywhile simultaneously improving optoelectronic properties byinducing polymeric chain alignment32,33 and thus enhancingcarrier mobility.34 However, to our best knowledge, NIL hasnot been reported on organohalide perovskites to date. Herewe demonstrate the use of NIL processes to successfully definemicro- and nanostructures on perovskite thin films into activedevice components, while simultaneously improving theperovskite film crystalline quality and corresponding optoelec-tronic performance.

RESULTS AND DISCUSSION

Characterization of Nanoimprinted and Nonim-printed Perovskite Thin Films. A modified solvent-engineer-ing method reported by Jeon et al.35 was used for MAPbI3perovskite deposition. This method utilizes a mixture of γ-

butyrolactone (GBL) and N,N′-dimethyl sulfoxide (DMSO) assolvents of methylammonium lead halide perovskite for spin-coating followed with a toluene drip while spinning, whichallows formation of a homogeneous perovskite layer afterthermal annealing. In this study, (100) Si with 100 nm thickthermal oxide was used as the substrate. After spin-coating andtoluene-dripping, the sample was annealed at 100 °C for 10min, which subsequently drives out the solvents and forms aperovskite thin film with a thickness of 265 nm. One of thesamples was imprinted, as illustrated in Figure 1. Threedifferent Si molds were used to compare different structures. Aflat surface (Figure 1a), nanopillar (Figure 1b), and nanograting(Figure 1c) structures were first treated with an antiadhesionmonolayer of 1H,1H,2H,2H-perfluorodecyltrichlorosilane(FDTS). The molds were then placed on the perovskite thinfilm on different areas in a single process to ensure the sameconditions applied to all molds. The samples were imprintedwith a profile of conditions illustrated in Figure S1. The NILwas done at a temperature of 100 °C and a pressure of 7 MPafor 20 min. Reference samples with as-spin-coated perovskitethin films without NIL were also prepared.The morphology of both the nonimprinted (Figure 2a) and

imprinted films (Figure 2b−d) was studied by scanning

Figure 1. Schematic of NIL process with (a) Si flat mold, (b) Si nanopillar mold, and (c) Si nanograting mold.

Figure 2. Scanning electron micrograph of (a) nonimprintedperovskite thin film and perovskite thin film imprinted with (b) a Siflat mold, (c) a Si nanopillar mold, and (d) a Si nanograting mold.(e) SEM image of freestanding perovskite nanorods peeled off fromthe surface.

ACS Nano Article

DOI: 10.1021/acsnano.6b05535ACS Nano 2016, 10, 10921−10928

10922

electron microscopy (SEM). The well-defined perovskitenanohole structures (Figure 2c) and nanograting structures(Figure 2d) were negative replications of the Si nanopillar moldand nanograting mold, respectively. The nanoholes have apatterned diameter and pitch of 275 and 600 nm, respectively,while the patterned width and pitch of nanogratings are 270and 600 nm, respectively. The SEM cross-section image revealsa depth of 315 nm with almost no residual layer for thenanoholes and a structure depth of 300 nm and residual layerthickness of 130 nm for the nanogratings. No obviousperovskite volume change is observed after NIL. The resultshave demonstrated that although perovskite is an ionic solidand does not have a glass transition behavior like polymers, itcan be successfully patterned by NIL, as it can deform and fill inmold cavities upon applied heat and pressure. According to themechanical properties study by Jing,36 the hybrid perovskitelayer has a low shear modulus and is suitable for fabrication of aflexible device or compliant device with large deformationdemand. Our results here demonstrate that NIL is a convenientand cost-effective technique to fabricate perovskite nanostruc-tures.Morphologic improvement by NIL is also revealed. From

both top the view and cross-section view, the nonimprintedfilm (Figure 2a) shows small crystal grains and clear grainboundaries, while the nanoimprinted film (Figure 2b−d) showsa larger grain size and smoother surface. The nanogratings,particularly, have shown almost invisible grain boundaries.Figure 2e demonstrates free-standing nanorods stripped duringmold release, and the nanorods appear to have a dramaticmorphologic difference compared with the residual layer. It islikely that with multidimensional confinement during nano-

imprinting the nanorods have been formed with significantlybetter crystallinity than the underlying residual layer.To confirm the improved structural order and crystallinity,

another set of nonimprinted perovskite and imprintedperovskite samples were analyzed with X-ray diffraction(XRD), and the results are presented in Figure S2. The sharp(220) reflection indicates the crystallinity of the imprintedMAPbI3 films. The crystallite sizes of the nonimprinted andnanoimprinted films were determined to be 68 and 188 nm,respectively.Both microscopic images and diffraction analysis have

confirmed the positive effect of NIL on perovskite crystallinity.A simple hypothesis is that during nanoimprinting withelevated temperature and pressure the perovskite small grainsslide toward the mold cavities and collide with each other,which forms bigger grains and defects such as dislocations,disclinations, and vacancies are reduced. Another explanation isthat the grain boundaries have been pushed to the perovskiteand mold interface during nanoimprinting and therefore crystalgrains are redefined based on the mold structures. Previously,improved crystallinity was also reported by Xiao et al.37 using ahot-pressing process as the post-treatment of a perovskite thinfilm, and the treated perovskite showed improved photo-luminescence lifetime and solar cell efficiency. Matsushima etal.38 reported a similar perovskite morphologic treatmentprocess by using a hot isostatic pressure method. Thetemperature was also found to have an influence on perovskitemorphology.39 Therefore, we expect that with an optimizedNIL process by carefully choosing temperature, pressure, andtime, the morphology and crystallinity of the perovskite thinfilm could be further improved.

Figure 3. (a) UV−vis transmission and (b) reflectance spectra of nonimprinted, flat-imprinted, nanohole, and nanograting MAPbI3 films. (c)Steady-state photoluminescence and (d) time-resolved photoluminescence of MAPbI3 thin film, nanogratings, and nanoholes.

ACS Nano Article

DOI: 10.1021/acsnano.6b05535ACS Nano 2016, 10, 10921−10928

10923

We then investigated the optical properties of bothnonimprinted and nanoimprinted films by characterizing thetransmission (Figure 3a), reflectance (Figure 3b), photo-luminescence (PL, Figure 3c), and PL lifetime (Figure 3d) ofperovskite films on glass substrates. The imprinted films withnanoholes and nanogratings show significantly reducedreflectance for the whole spectrum and reduced transmissionin the wavelength range of 550 to 800 nm. The results indicatewith proper mold nanostructure design for NIL favorablephoton trapping in perovskites for higher optical absorption isfeasible. Figure S3 shows iridescence of the patterned nanoholesample due to 2D photonic crystal effects, suggestingreasonable uniformity of NIL patterning across the 1 cm2

imprinted area, while the area appearing dark representsincomplete NIL or defects.The photoluminescence emission peak is located around 780

nm (1.59 eV) (Figure 3c). The nanoimprinted perovskitesdemonstrate better spontaneous emission properties withapproximately 3 times improvement for nanogratings and 4times improvement for nanoholes compared with the non-imprinted thin films. The time-resolved photoluminescenceacquired using a time-correlated single photon countingmethod (excitation laser wavelength 435 nm, pulse width 100fs, repetition rate 1 MHz), as presented in Figure 3d,demonstrates improved lifetime by NIL: lifetime increasesfrom a perovskite thin film (35 ns) to imprinted nanogratings(42 ns) and nanoholes (50 ns).Characterization of Nanoimprinted and Nonim-

printed Perovskite Photodetectors. To evaluate theoptoelectronic performance of perovskites by NIL, we

fabricated metal−semiconductor−metal photodetectors byevaporating 300 nm thick gold electrodes with a 25 μmelectrode gap on both imprinted and nonimprinted perovskitesamples. The width of the electrode gap is 100 μm. On thebasis of the film morphology, four types of devices werestudied: the conventional nonimprinted thin-film perovskitephotodetector (tf-PSPD), flat mold imprinted perovskitephotodetector (flat-PSPD), nanohole perovskite photodetector(nanohole-PSPD), and nanograting perovskite photodetector(nanograting-PSPD), as shown in Figure S4. The photo-electrical characteristics of four different types of photo-detectors were obtained under the same test configuration atroom temperature in air. The schematic illustration is shown inFigure 4a with a nanograting-PSPD. One should note that for ananograting-PSPD, the electrode pairs were deposited so thatthe current flow is along the gratings under applied electricalfield. Figure 4b shows an example of an I−V curve of thenanograting-PSPD in the dark and under halogen lightillumination with irradiance varying from 0.11 to 7.27 mW/cm2. The linear current−voltage behavior indicates a goodohmic contact between gold and perovskite. In the dark state,the device has a resistance of tens of gigaohms. Underillumination, large amounts of electron−hole pairs aregenerated due to photon absorption and are subsequentlyextracted by the electrical field, which causes a dramaticincrease of conductance. At the same voltage, the photocurrentincreases gradually with incident light density, as illustrated inthe inset graph of Figure 4b. The corresponding I−V curve forthe tf-PSPD is presented in Figure S5. Figure 4c shows thetemporal current of the nanograting-PSPD under 7.27 mW/

Figure 4. Structure and characteristics of perovskite metal−semiconductor−metal photodetectors. (a) Schematic illustration of thenanograting-PSPD. (b) I−V characteristics of nanograting-PSPD with 0.11 to 7.27 mW/cm2 halogen light illumination. The inset shows thecurrent as a function of irradiance. (c) Temporal current characteristics of nanograting-PSPD at 7.27 mW/cm2 halogen light illumination witha bias voltage of 1 V. The light was switched on and off for 10 cycles for the repeatability test. (d) I−V characteristics of tf-PSPD, flat-PSPD,nanohole-PSPD, and nanograting PSPD at 7.27 mW/cm2 halogen light illumination in logarithmic scale. The inset shows the same curves inlinear scale.

ACS Nano Article

DOI: 10.1021/acsnano.6b05535ACS Nano 2016, 10, 10921−10928

10924

cm2 halogen lamp illumination with a bias of 1 V. The light wasswitched on and off for 10 cycles during the test. An on/offcurrent ratio of more than 1000 was achieved with a darkcurrent as low as 30 pA while illuminated with a current ofmore than 40 nA. To compare the optoelectronic performancebetween the tf-PSPD, flat-PSPD, nanohole-PSPD, and nano-grating-PSPD, their I−V characteristics under 7.27 mW/cm2

halogen light illumination were plotted in Figure 4d. One couldclearly see that under the same conditions the flat-PSPD,nanohole-PSPD, and nanograting-PSPD all exhibit a largeimprovement in photocurrent compared with the tf-PSPD. Thenanograting-PSPD has the highest photocurrent, over 35 timesthat of the tf-PSPD at a bias of 1 V. To obtain a more reliableperformance comparison, multiple devices were tested for eachtype. The corresponding I−V curves under 7.27 mW/cm2

halogen light illumination are shown in Figure S6.Responsivity is widely used to evaluate the efficiency of a

photodetector responding to an optical signal. It is defined asthe ratio of the photocurrent to the illumination power, asexpressed by

=×

RI

L Sph

light (1)

where Llight is the incident light power density, S is the effectivearea, and Iph represents the photocurrent as given by

= −I I Iph illuminated dark (2)

where Iilluminated and Idark are the current with and withoutillumination, respectively. Besides responsivity, the on/off ratiothat is represented by illuminated current divided by the darkcurrent is another important parameter for photodetectors. Thecalculated photodetector parameters along with their geo-metries are summarized in Table 1. All the nanoimprinted

devices showed significantly enhanced performance; that is, theresponsivity R and on/off ratio are over 10 times and 5 timeshigher than those of the nonimprinted devices, respectively.Particularly, in the dark environment, Idark has also increased by3 times for the flat-PSPD, 2.4 times for the nanohole-PSPD,and 5 times for the nanograting-PSPD, which should beattributed to the improvement of charge carrier transport in theNIL films under an applied electrical field.To study whether thermal annealing during NIL was the

primary cause of performance enhancement of nanoimprinteddevices, another set of perovskite thin-film samples wasprepared and the corresponding MSM photodetectors were

fabricated. One sample was annealed at 100 °C for 10 min,while another was annealed for 30 min at the same temperatureduring the perovskite thin-film preparation process. Theirphotodetector performance was characterized and is presentedin Figure S7. The devices with 30 min annealing show relativelyworse performance, suggesting possible degradation due tolong-time thermal treatment. The results indicate that theimprovement of nanoimprinted devices were caused by thecombined effect of elevated pressure and temperature duringNIL, especially with confined nanostructures.We believe the improved crystallinity of the imprinted

perovskite film is one of the primary causes of the performanceenhancement for the nanoimprinted photodetectors. Severalprocesses including photon absorption, electron−hole gen-eration, carrier transport, and recombination would determinethe photodetector performance.40 The results of SEM (Figure2) and XRD (Figure S2) have shown that NIL has induced theformation of larger and ordered grains, and thus the crystallinityof perovskite has improved, which plays a positive role inmultiple processes. First, the carrier transport and mobilitywould increase,25 as the charge carriers encounter lessscattering at the grain boundaries or defects. Therefore, boththe illuminated current and dark currents increase significantlyin the NIL imprinted samples. Second, it will result in a longerelectron−hole recombination lifetime,37 which is verifiedthrough PL lifetime tests (Figure 3d). Improvements in bothmobility and carrier lifetime contribute to longer diffusionlength. These effects contribute to the dramatically improvedphotocurrent and thus responsivity for the nanoimprintedphotodetectors. The comparable on/off ratio between the flat-PSPD and nanohole-PSPD indicates that charge carriertransport might be the primary cause of the inferiorperformance of the nanohole-PSPD. The vertical nanoholearrays could hinder the carrier transport since the chargecarriers suffer from severe surface scattering driven by theelectrical field. The gratings, on the other hand, largelyenhanced the carrier transport due to well-aligned conductivechannels along the electrical field and ordered crystal grains toreduce surface and grain boundary scattering. The nanogratingstructure is also suitable for photon management due to theone-dimensional photonic band gap effect.41,42 Therefore, thenanograting-PSPDs deliver the best performance in thesesamples with 35-fold higher responsivity and 7-fold higher on/off current ratio than the tf-PSPD. It is noted that due to thelimitation of the mold depth, the nanograting-PSPD has aresidual layer of 130 nm that also contributes to the total devicecurrent. From the SEM images of Figure 2d and e, the residuallayer may not have a high crystallinity as compared to thegratings. Therefore, it is reasonable to assume that the residuallayer has a similar performance to that of the flat-PSPD, and itscontribution to the total current should be less than 20%. Byfurther optimizing the mold depths, we expect the residual-layer-free nanogratings to have even higher performance.

High-Performance Perovskite Nanograting Photo-detectors. To further evaluate the nanograting-PSPDs, theywere tested along with the tf-PSPD under monochromatic LEDillumination. The currents of the nanograting-PSPD, tf-PSPD,and a commercial Si photodiode were measured and plottedagainst the LED forward current in Figure S8. Both the tf-PSPD and nanograting-PSPD were biased at 1 V, while the Siphotodiode was reverse biased at 10 V. The nanograting-PSPDshows much higher current than the tf-PSPD. One should notethat the Si photodiode has the largest current, as its effective

Table 1. Performance and Geometry Comparison of tf-PSPD, Flat-PSPD, Nanohole-PSPD, and Nanograting-PSPDDevices

parametertf-

PSPDflat-PSPD

nanohole-PSPD

nanograting-PSPD

thickness (nm) 265 265 315 300, residue 130pattern width (nm) N/A N/A 275 270pattern pitch (nm) N/A N/A 600 600responsivitya 1 15.4 11.9 34.7on/off ratioa 1 5.2 5.0 6.9dark currenta 1 3.0 2.4 5.0aValues of responsivity, on/off ratio, and dark current are normalizedto the results of the tf-PSPD. Responsivity and on/off ratio under 7.27mW/cm2 illumination were extracted from raw data as shown inFigure S6 and Table S1.

ACS Nano Article

DOI: 10.1021/acsnano.6b05535ACS Nano 2016, 10, 10921−10928

10925

radiant-sensitive area is over 3000 times that of our perovskitephotodetectors, while the current per effective illuminated areaunder the same irradiance is of concern here. The irradiancewas evaluated with the Si photodiode, and the correspondingphotodetector current versus irradiance is plotted in Figure S9.Figure 5 shows the calculated photodetector responsivity versusirradiance at λ = 466 nm (Figure 5a) and λ = 635 nm (Figure5b) with a bias voltage of 1 V. It is observed that generally witha decrease in the light intensity the responsivity increased. Theresults were in agreement with the literature.4,26,27 Theperformance of the nanograting-PSPD was superior to that ofthe tf-PSPD, similar to the results of the halogen lightillumination tests. At λ = 466 nm (Figure 5a), the tf-PSPDhas only R = 0.16 A/W, while the nanograting-PSPD has R =3.23 A/W under 1 μW/cm2 illumination. With 2 nW/cm2

irradiance, the nanograting-PSPD has R = 24.1 A/W, which is100 times that of the commercial Si photodiode (0.12 A/W).Similarly, at an illumination of λ = 635 nm (Figure 5b), thenanograting-PSPD has R = 6.2 A/W, which is over 30 timesthat of the tf-PSPD and 20 times that of the commercial Siphotodiode under 1 μW/cm2 irradiance. At 4.5 nW/cm2

irradiance, the responsivity of the nanograting-PSPD hasincreased to 58.5 A/W, which is 100 times more than that ofthe commercial Si photodiode (0.3 A/W). Both devices showbetter response at λ = 635 nm than at 466 nm. The imprintednanograting-PSPDs also outperform the previously reportedhybrid perovskite nanowire29,30,43,44 and thin-film photo-detectors.24,26,45,46

CONCLUSIONIn summary, we report the use of nanoimprint lithography todefine ordered perovskite nanostructures as active device areas,while NIL simultaneously improves their crystallinity andoptoelectronic performance. NIL was conducted on perovskitethin films with flat, nanopillar, and nanograting molds. Planarmetal−semiconductor−metal photodetectors were fabricatedon the perovskite films with different morphologies, and theiroptoelectronic performance was characterized. All of thenanoimprinted devices demonstrated significantly improvedperformance compared to nonimprinted devices, while thenanograting devices are the best with an average of 35 timesimprovement in responsivity and 7 times improvement in on/off current ratio under 7.27 mW/cm2 halogen lightillumination. The nanograting-PSPD has a high responsivityvalue of 24.1 A/W at 2 nW/cm2 LED illumination of λ = 466

nm and 58.5 A/W at 4.5 nW/cm2 illumination of λ = 635 nmwith a bias voltage of 1 V. Such performance is ∼30 timesbetter than the tf-PSPD and more than 100 times better thanthe commercial Si photodiode. The performance enhancementis likely due to NIL-induced higher crystallinity; particularly,the nanograting structure is favorable for better photonabsorption and charge carrier transport. Further improvementon the nanograting-PSPD performance is expected byoptimizing the nanograting geometries and the nanoimprintingconditions. Additionally, nanograting-based photodetectors andsolar cells with vertical P−I−N structures are in preparation.Our study demonstrated that NIL is a simple yet effective wayto fabricate high-performance nanoscale optoelectronic devicesusing emerging hybrid perovskite materials, which are suitablefor electronic circuit integration and manufacturing.

METHODSPerovskite Thin-Film Preparation. The perovskite solution was

prepared by dissolving a 1:1 molar ratio of PbI2 and CH3NH3I in a 7:3volume ratio of γ-butyrolactone:N,N′-dimethyl sulfoxide solventmixture in a N2 glovebox. The resulting concentration was 1.2 M.The solution was heated for 24 h at 60 °C. The solution was then spin-coated onto the Si substrates with 100 nm thick thermal SiO2 that waspreviously ultrasonically cleaned with acetone and treated by oxygenplasma. A two-step spin-coating process was performed for 22 s at1000 rpm and then 22 s at 5000 rpm. A 350 mL amount of anhydroustoluene was dropped on the film after 12 s in the second spin-coatingstep. The sample was then annealed on a hot plate at 100 °C for 10min, during which solvents were evaporated and a dense and uniformMAPbI3 film was formed with a thickness of about 265 nm.

Nanoimprinting of Perovskite Films. The Si flat, nanopillar, andnanograting molds were first treated with FDTS in n-heptane solventfor 5 min and then cleaned with acetone and blow dried with N2. Themolds were then annealed at 100 °C for 20 min. Monolayer FDTS wasformed on the Si molds, which served an antiadhesive purpose in theNIL process. The Si flat mold, nanopillar mold, and nanograting moldwere then placed on the perovskite thin-film-coated substrate atdifferent areas in a single process. The imprint utilized a multistepprocess: 90 s at a temperature of 35 °C and a pressure of 2 MPa, 180 sat a temperature of 55 °C and a pressure of 4 MPa, 180 s at atemperature of 75 °C and a pressure of 6 MPa, and then, importantly,1200 s at a temperature of 100 °C and a pressure of 7 MPa. Thepressure was kept at 7 MPa while the chamber was cooled to atemperature of 35 °C. The nanoimprinting process was then finished,and perovskite nanostructures were formed as a negative replication ofthe Si molds.

Fabrication of a Metal−Semiconductor−Metal Photodetec-tor. SiO2 (100 nm) was thermally grown on a (100) Si wafer.

Figure 5. Irradiance-dependent responsivity plot at (a) λ = 466 nm and (b) λ = 635 nm. Red curve represents the nanograting-PSPD and bluecurve represents the tf-PSPD. The right y-axis illustrates the relative responsivity normalized to the Si photodiode. Both perovskitephotodetectors were biased at 1 V, while the Si photodiode was reverse biased at 10 V.

ACS Nano Article

DOI: 10.1021/acsnano.6b05535ACS Nano 2016, 10, 10921−10928

10926

Perovskite thin films and nanoimprinted samples were prepared asdescribed previously. A 300 nm thick gold film was deposited on theperovskite samples in an e-beam evaporator using a shadow mask. Thegap between the gold electrode pairs was 25 μm in length and 100 μmin width. The effective photodetector area was 2.5 × 10−5 cm2.Characterization of Photodetector Devices. A Keithley 4200

and a Cascade probe station were used to characterize the perovskitephotodetectors. The devices under a dark environment and differentillumination conditions were tested. A 150 W halogen lamp was usedfor illumination for all the devices. A blue LED with a peak wavelengthof 466 nm and a red LED with a peak wavelength of 635 nm were usedfor the responsivity test of the nanograting-PSPD and tf-PSPD devices.The illumination light intensity was calibrated with a commercial Siphotodiode.

ASSOCIATED CONTENT*S Supporting InformationThe Supporting Information is available free of charge on theACS Publications website at DOI: 10.1021/acsnano.6b05535.

Nanoimprint process profile, X-ray diffraction, opticalimage, photodetector 3-D schematics, and optoelectriccharacteristics (PDF)

AUTHOR INFORMATIONCorresponding Author*E-mail: walter.hu@utdallas.edu.

ORCIDHonglei Wang: 0000-0002-5079-625XJulia Y. Chan: 0000-0003-4434-2160NotesThe authors declare no competing financial interest.

ACKNOWLEDGMENTSThis work is partially supported by the National ScienceFoundation (Nos. ECCS-0901759 and ECCS-0955027).Financial support of the Welch Foundation under grant AT-1617 is also highly appreciated. W.H. acknowledges supportfrom the 1000 Talent Program of Shanghai, China. Partialfinancial support from the Ministry of Education and Science ofthe Russian Federation in the framework of IncreaseCompetitiveness Program of NUST “MISiS” (No. K2-2015-014) is also appreciated. The authors acknowledge D. Saraninand A. Ishteev for help with preparation of perovskite films, Y.Ren for X-ray data collection, and Y. Yang and M. Aryal forhelpful discussions on nanoimprint lithography.

REFERENCES(1) Tang, L.; Kocabas, S. E.; Latif, S.; Okyay, A. K.; Ly-Gagnon, D.-S.; Saraswat, K. C.; Miller, D. A. B. Nanometre-Scale GermaniumPhotodetector Enhanced by A Near-Infrared Dipole Antenna. Nat.Photonics 2008, 2, 226−229.(2) Xia, F.; Mueller, T.; Lin, Y.-m.; Valdes-Garcia, A.; Avouris, P.Ultrafast Graphene Photodetector. Nat. Nanotechnol. 2009, 4, 839−843.(3) Lopez-Sanchez, O.; Lembke, D.; Kayci, M.; Radenovic, A.; Kis, A.Ultrasensitive Photodetectors Based on Monolayer MoS2. Nat.Nanotechnol. 2013, 8, 497−501.(4) Lian, Z.; Yan, Q.; Lv, Q.; Wang, Y.; Liu, L.; Zhang, L.; Pan, S.; Li,Q.; Wang, L.; Sun, J.-L. High-Performance Planar-Type Photodetectoron (100) Facet of MAPbI3 Single Crystal. Sci. Rep. 2015, 5, 16563.(5) Hayden, O.; Agarwal, R.; Lieber, C. M. Nanoscale AvalanchePhotodiodes for Highly Sensitive and Spatially Resolved PhotonDetection. Nat. Mater. 2006, 5, 352−356.

(6) Guo, F.; Yang, B.; Yuan, Y.; Xiao, Z.; Dong, Q.; Bi, Y.; Huang, J.A Nanocomposite Ultraviolet Photodetector Based on InterfacialTrap-Controlled Charge Injection. Nat. Nanotechnol. 2012, 7, 798−802.(7) Peng, L.; Hu, L.; Fang, X. Low-Dimensional NanostructureUltraviolet Photodetectors. Adv. Mater. 2013, 25, 5321−5328.(8) Gramotnev, D. K.; Bozhevolnyi, S. I. Plasmonics beyond TheDiffraction Limit. Nat. Photonics 2010, 4, 83−91.(9) Avouris, P.; Freitag, M.; Perebeinos, V. Carbon-NanotubePhotonics and Optoelectronics. Nat. Photonics 2008, 2, 341−350.(10) Kind, H.; Yan, H.; Messer, B.; Law, M.; Yang, P. NanowireUltraviolet Photodetectors and Optical Switches. Adv. Mater. 2002, 14,158−160.(11) Li, L.; Wu, P.; Fang, X.; Zhai, T.; Dai, L.; Liao, M.; Koide, Y.;Wang, H.; Bando, Y.; Golberg, D. Single-Crystalline CdS Nanobeltsfor Excellent Field-Emitters and Ultrahigh Quantum-EfficiencyPhotodetectors. Adv. Mater. 2010, 22, 3161−3165.(12) Wang, J.; Gudiksen, M. S.; Duan, X.; Cui, Y.; Lieber, C. M.Highly Polarized Photoluminescence and Photodetection from SingleIndium Phosphide Nanowires. Science 2001, 293, 1455−1457.(13) Wagner, R. S.; Ellis, W. C. Vapor-Liquid-Solid Mechanism ofSingle Crystal Growth. Appl. Phys. Lett. 1964, 4, 89.(14) Lee, M. M.; Teuscher, J.; Miyasaka, T.; Murakami, T. N.; Snaith,H. J. Efficient Hybrid Solar Cells Based on Meso-SuperstructuredOrganometal Halide Perovskites. Science 2012, 338, 643−647.(15) Green, M. A.; Ho-Baillie, A.; Snaith, H. J. The Emergence ofPerovskite Solar Cells. Nat. Photonics 2014, 8, 506−514.(16) Stranks, S. D.; Eperon, G. E.; Grancini, G.; Menelaou, C.;Alcocer, M. J. P.; Leijtens, T.; Herz, L. M.; Petrozza, A.; Snaith, H. J.Electron-Hole Diffusion Lengths Exceeding 1 Micrometer in AnOrganometal Trihalide Perovskite Absorber. Science 2013, 342, 341−344.(17) Xing, G.; Mathews, N.; Sun, S.; Lim, S. S.; Lam, Y. M.; Graetzel,M.; Mhaisalkar, S.; Sum, T. C. Long-Range Balanced Electron- andHole-Transport Lengths in Organic-Inorganic CH3NH3PbI3. Science2013, 342, 344−347.(18) Dou, L.; Yang, Y.; You, J.; Hong, Z.; Chang, W.-H.; Li, G.; Yang,Y. Solution-Processed Hybrid Perovskite Photodetectors with HighDetectivity. Nat. Commun. 2014, 5, 5404.(19) Dong, R.; Fang, Y.; Chae, J.; Dai, J.; Xiao, Z.; Dong, Q.; Yuan,Y.; Centrone, A.; Zeng, X. C.; Huang, J. High-Gain and Low-Driving-Voltage Photodetectors Based on Organolead Triiodide Perovskites.Adv. Mater. 2015, 27, 1912−1918.(20) Liu, C.; Wang, K.; Yi, C.; Shi, X.; Du, P.; Smith, A. W.; Karim,A.; Gong, X. Ultrasensitive Solution-Processed Perovskite HybridPhotodetectors. J. Mater. Chem. C 2015, 3, 6600−6606.(21) Fang, Y.; Huang, J. Resolving Weak Light of Sub-picowatt perSquare Centimeter by Hybrid Perovskite Photodetectors Enabled byNoise Reduction. Adv. Mater. 2015, 27, 2804−2810.(22) Su, L.; Zhao, Z. X.; Li, H. Y.; Yuan, J.; Wang, Z. L.; Cao, G. Z.;Zhu, G. High-Performance Organolead Halide Perovskite-Based Self-Powered Triboelectric Photodetector. ACS Nano 2015, 9, 11310−11316.(23) Li, D.; Dong, G.; Li, W.; Wang, L. High Performance Organic-Inorganic Perovskite-Optocoupler Based on Low-Voltage and FastResponse Perovskite Compound Photodetector. Sci. Rep. 2015, 5,7902.(24) Hu, X.; Zhang, X.; Liang, L.; Bao, J.; Li, S.; Yang, W.; Xie, Y.High-Performance Flexible Broadband Photodetector Based onOrganolead Halide Perovskite. Adv. Funct. Mater. 2014, 24, 7373−7380.(25) Saidaminov, M. I.; Adinolfi, V.; Comin, R.; Abdelhady, A. L.;Peng, W.; Dursun, I.; Yuan, M.; Hoogland, S.; Sargent, E. H.; Bakr, O.M. Planar-Integrated Single-Crystalline Perovskite Photodetectors.Nat. Commun. 2015, 6, 8724.(26) Guo, Y.; Liu, C.; Tanaka, H.; Nakamura, E. Air-Stable andSolution-Processable Perovskite Photodetectors for Solar-Blind UVand Visible Light. J. Phys. Chem. Lett. 2015, 6, 535−539.

ACS Nano Article

DOI: 10.1021/acsnano.6b05535ACS Nano 2016, 10, 10921−10928

10927

(27) Lee, Y.; Kwon, J.; Hwang, E.; Ra, C.-H.; Yoo, W. J.; Ahn, J.-H.;Park, J. H.; Cho, J. H. High-Performance Perovskite-Graphene HybridPhotodetector. Adv. Mater. 2015, 27, 41−46.(28) Deng, H.; Yang, X.; Dong, D.; Li, B.; Yang, D.; Yuan, S.; Qiao,K.; Cheng, Y.-B.; Tang, J.; Song, H. Flexible and SemitransparentOrganolead Triiodide Perovskite Network Photodetector Arrays withHigh Stability. Nano Lett. 2015, 15, 7963−7969.(29) Horvath, E.; Spina, M.; Szekrenyes, Z.; Kamaras, K.; Gaal, R.;Gachet, D.; Forro, L. Nanowires of Methylammonium Lead Iodide(CH3NH3Pbl3) Prepared by Low Temperature Solution-MediatedCrystallization. Nano Lett. 2014, 14, 6761−6766.(30) Zhuo, S.; Zhang, J.; Shi, Y.; Huang, Y.; Zhang, B. Self-Template-Directed Synthesis of Porous Perovskite Nanowires at RoomTemperature for High-Performance Visible-Light Photodetectors.Angew. Chem., Int. Ed. 2015, 54, 5693−5696.(31) Wang, G.; Li, D.; Cheng, H.-C.; Li, Y.; Chen, C.-Y.; Yin, A.;Zhao, Z.; Lin, Z.; Wu, H.; He, Q.; Ding, M.; Liu, Y.; Huang, Y.; Duan,X. Wafer-Scale Growth of Large Arrays of Perovskite MicroplateCrystals for Functional Electronics and Optoelectronics. Sci. Adv.2015, 1, e1500613.(32) Yang, Y.; Mielczarek, K.; Aryal, M.; Zakhidov, A.; Hu, W.Nanoimprinted Polymer Solar Cell. ACS Nano 2012, 6, 2877−2892.(33) Yang, Y.; Lee, K.; Mielczarek, K.; Hu, W.; Zakhidov, A.Nanoimprint of Dehydrated PEDOT:PSS for Organic Photovoltaics.Nanotechnology 2011, 22, 485301.(34) Zhou, M.; Aryal, M.; Mielczarek, K.; Zakhidov, A.; Hu, W. HoleMobility Enhancement by Chain Alignment in Nanoimprinted Poly(3-Hexylthiophene) Nanogratings for Organic Electronics. J. Vac. Sci.Technol., B 2010, 28, C6M63−C6M67.(35) Jeon, N. J.; Noh, J. H.; Kim, Y. C.; Yang, W. S.; Ryu, S.; Seok, S.I. Solvent Engineering for High-Performance Inorganic−OrganicHybrid Perovskite Solar Cells. Nat. Mater. 2014, 13, 897−903.(36) Feng, J. Mechanical Properties of Hybrid Organic-InorganicCH3NH3BX3 (B = Sn, Pb; X = Br, I) Perovskites for Solar CellAbsorbers. APL Mater. 2014, 2, 081801.(37) Xiao, J.; Yang, Y.; Xu, X.; Shi, J.; Zhu, L.; Lv, S.; Wu, H.; Luo, Y.;Li, D.; Meng, Q. Pressure-Assisted CH3NH3PbI3 MorphologyReconstruction to Improve The High Performance of PerovskiteSolar Cells. J. Mater. Chem. A 2015, 3, 5289−5293.(38) Matsushima, T.; Fujihara, T.; Qin, C.; Terakawa, S.; Esaki, Y.;Hwang, S.; Sandanayaka, A. S. D.; Potscavage, W. J.; Adachi, C.Morphological Control of Organic-Inorganic Perovskite Layers by HotIsostatic Pressing for Efficient Planar Solar Cells. J. Mater. Chem. A2015, 3, 17780−17787.(39) Zhu, L.; Shi, J.; Lv, S.; Yang, Y.; Xu, X.; Xu, Y.; Xiao, J.; Wu, H.;Luo, Y.; Li, D.; Meng, Q. Temperature-Assisted ControllingMorphology and Charge Transport Property for Highly EfficientPerovskite Solar Cells. Nano Energy 2015, 15, 540−548.(40) Sum, T. C.; Mathews, N. Advancements in Perovskite SolarCells: Photophysics behind The Photovoltaics. Energy Environ. Sci.2014, 7, 2518−2534.(41) Mellor, A.; Hauser, H.; Wellens, C.; Benick, J.; Eisenlohr, J.;Peters, M.; Guttowski, A.; Tobías, I.; Martí, A.; Luque, A.; Blas̈i, B.Nanoimprinted Diffraction Gratings for Crystalline Silicon Solar Cells:Implementation, Characterization and Simulation. Opt. Express 2013,21, A295−A304.(42) Yao, Y.; Wang, Y.; Liu, H.; Li, Y.; Song, B.; Wu, W. Line WidthTuning and Smoothening for Periodical Grating Fabrication inNanoimprint Lithography. Appl. Phys. A: Mater. Sci. Process. 2015,121, 399−403.(43) Deng, H.; Dong, D.; Qiao, K.; Bu, L.; Li, B.; Yang, D.; Wang,H.-E.; Cheng, Y.; Zhao, Z.; Tanga, J.; Song, H. Growth, Patterning andAlignment of Organolead Iodide Perovskite Nanowires for Optoelec-tronic Devices. Nanoscale 2015, 7, 4163−4170.(44) Li, Y.; Wang, X.; Wu, S.; Ci, H.; Xu, H.; Li, X.; Sun, H.; Zhang,Z.; Cao, A.; Guo, X.; Li, Y. Large-Scale Aligned CrystallineCH3NH3PbI3 Perovskite Array Films. J. Mater. Chem. A 2015, 3,18847−18851.

(45) Zhang, Y.; Du, J.; Wu, X.; Zhang, G.; Chu, Y.; Liu, D.; Zhao, Y.;Liang, Z.; Huang, J. Ultrasensitive Photodetectors Based on Island-Structured CH3NH3PbI3 Thin Films. ACS Appl. Mater. Interfaces 2015,7, 21634−21638.(46) Sutherland, B. R.; Johnston, A. K.; Ip, A. H.; Xu, J.; Adinolfi, V.;Kanjanaboos, P.; Sargent, E. H. Sensitive, Fast, and Stable PerovskitePhotodetectors Exploiting Interface Engineering. ACS Photonics 2015,2, 1117−1123.

ACS Nano Article

DOI: 10.1021/acsnano.6b05535ACS Nano 2016, 10, 10921−10928

10928