large molecular weight polymer solar cells with strong chain …walter.hu/research... ·...

Large Molecular Weight Polymer Solar Cells with Strong ChainAlignment Created by Nanoimprint LithographyYi Yang,*,† Kamil Mielczarek,‡ Anvar Zakhidov,†,‡,∥ and Walter Hu*,†,§

†Department of Materials Science and Engineering, ‡Department of Physics, and §Department of Electrical Engineering, TheUniversity of Texas at Dallas, Richardson, Texas 75080, United States,∥Center of Energy Efficiency, National University of Science and Technology “MISiS”, Leninsky Prospekt, 4, Moscow, Russia

ABSTRACT: In this work, strong chain alignment in largemolecular weight polymer solar cells is for the first timedemonstrated by nanoimprint lithography (NIL). Thepolymer crystallizations in nonimprinted thin films andimprinted nanogratings with different molecular weightpoly(3-hexylthiophene-2,5-diyl) (P3HT) are compared. Wefirst observe that the chain alignment is favored by mediummolecular weight (Mn = 25 kDa) P3HT for nonimprinted thinfilms. However, NIL allows large molecular weight P3HT (>40kDa) to organize more strongly, which has been desired forefficient charge transport but is difficult to achieve through any other technique. Consequently P3HT/[6,6]-penyl-C61-butyric-acid-methyl-ester (PCBM) solar cells with large molecular weight P3HT nanogratings show a high power conversion efficiencyof 4.4%, which is among the best reported P3HT/PCBM photovoltaics devices.

KEYWORDS: polymer solar cell, morphology, charge transport, nanoimprint lithography, molecular weight dependent chain alignment

1. INTRODUCTION

Polymer solar cells have potential to provide a low cost,lightweight, and flexible source of renewable energy.1,2 However,this type of solar cell has relatively low power conversionefficiency (PCE) when compared to those of other inorganiccounterparts.3 To improve PCE, one needs to simultaneouslycontrol the polymer donor−fullerene acceptor phase separationwithin the short exciton diffusion length (∼10 nm) and increasethe charge mobility, especially hole mobility, which is importantfor efficient charge separation and transport.4,5 The bulkheterojunction (BHJ) has provided a promising solution tothese two issues through its large donor−acceptor interface areaas well as its annealing assisted polymer chain ordering.6−8

However, the discrete and randomly distributed phases in thisstructure cause charge recombination, resulting in poor chargeseparation and transport.9 Moreover, the chain ordering and holemobility in this type of solar cell are greatly dependent on theintrinsic properties of the polymer donor materials, for example,the polymer molecular weight (MW).10−14 Although low MWpolymers are easy to crystallize, largeMWpolymers are preferredfor efficient charge transport in devices, due to the latter beingable to travel further along the chains before having to hop a largedistance to other molecules.15 However, it is found that, for manypolymers, such as the widely used poly(3-hexylthiophene-2,5-diyl) (P3HT), hole mobility initially increases or remainsconstant when MW increases within the low range (Mn ∼ 10−20 kDa), but starts decreasing when it further increases to thehigh range (>30 kDa). A mediumMW (20−30 kDa) is thereforefound to be optimal for solar cell applications.10,16,17 It is believedthat the long chains of large MW polymers are easy to tangle and

cause twisting of the polymer backbone, decreasing intrachaintransport by creating more traps and/or reducing interchaincharge hopping by allowing less overlap of conjugated seg-ments.16,18,19 To better improve charge transport, newapproaches that can overcome the chain entanglement problemenabling a strong chain alignment in large MW semiconductingpolymers are needed.In this work, strong chain alignment in large molecular weight

P3HT is for the first time demonstrated by nanoimprintlithography (NIL). The chain alignments in nonimprinted thinfilms and imprinted nanogratings made from medium to largeMW (Mn ∼ 25−50 kDa) P3HT are compared using grazingincident X-ray diffraction (GIXRD) and UV−vis lightabsorption. We found that nonimprinted thin films with mediumMW (25 kDa) show the strongest chain alignment, consistentwith literature reports. However, NIL causes stronger polymerordering in large MW (43 kDa) samples. Then, BHJ andnanoimprinted P3HT/[6,6]-phenyl-C61-butyric-acid-methyl-ester (PCBM) solar cells with different MWP3HT are fabricatedto compare its influence on device performance. Consistent withthe GIXRD results, higher PCEs for BHJ structure are observedfrom medium MW P3HT solar cells, but nanoimprinted deviceswith large MWs are more efficient. The highest PCE ofnanoimprinted solar cells is 4.4%, which is among the bestreported for P3HT/PCBM photovoltaics.

Received: January 6, 2016Accepted: March 7, 2016

Research Article

www.acsami.org

© XXXX American Chemical Society A DOI: 10.1021/acsami.6b00192ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

www.acsami.org

http://dx.doi.org/10.1021/acsami.6b00192

2. EXPERIMENTAL SECTION2.1. Nanoimprint Lithography, GIXRD, and UV−Vis Measure-

ments of P3HT Nanogratings. The number-average molecularweight, Mn, and polydispersity (given by Mw/Mn where Mw is theweight-average molecular weight) of the P3HT batches used in thisstudy are given in Table 1. Since polymer regioregularity has been shown

to affect charge mobility, all polymers used in this study have highregioregularities over 96%. It should be noted that P3HT MW startingwith 25 kDa is used in this study for two reasons: the previous studieshave shown that BHJ P3HT solar cells withMW∼ 25 kDa give the mostefficient hole transport, and highest PCE.10,16,17 Choosing the sameMWwould then allow us to compare the performance of nanoimprintedsolar cells with efficient nonpatterned BHJ reference devices. Second, wehave found that MW needs to be at least 25 kDa for P3HT to beinsoluble in dichloromethane (DCM), an orthogonal solvent used tospin-coat acceptor materials such as PCBM into the imprintednanogratings when making solar cells.20

The schematic of the nanoimprint process is shown in Figure 1a,b. A110 nm P3HT (Reike Metal, Ltd.) thin film was spin-coated ontosubstrates (silicon or glass) from solution in chlorobenzene and broughtinto contact with a silicon nanograting mold treated with anantiadhesion monolayer 1H,1H,2H,2H-perfluorodecyltrichlorosilane(FDTS). The nanograting samples were imprinted at 170 °C and 50MPa for 600 s. Figure 1c,d shows that the nanogratings on a Si mold aretransferred into P3HT film with excellent fidelity, with height h = 170nm, width w = 60 nm, and spacing p = 80 nm. As shown in eq 1, the ratioof imprinted nanostructure interface area (A) to nonimprinted flat thinfilm (A0) is defined by interface enhancement factor (IEF).

21,22 The IEFis 3.43 for these imprinted nanogratings, which is the largest we canachieve in this work.

= = ++

A Ah

w pIEF / 1

20

(1)

Nonimprinted reference samples were pressed by a flat mold at thesame temperature, pressure, and duration. It should be noted that in theGIXRD measurements we chose Si substrate because it exhibits nopeaks within the range of interest of P3HT (5°−25°). Additionally,silicon substrate gives the same type of P3HT chain orientation as ITOor PEDOT:PSS coated ITO, and is therefore widely used in literature tosimulate crystallization in solar cells.7,23−25

The chain alignment of nonimprinted thin films and imprinted P3HTnanogratings with various MWs as listed in Table 1 are studied usingGIXRD (Rigaku Ultima III diffractometer). According to literaturestudies, the edge-on configuration is the dominant chain orientation inflat P3HT thin films, and it changes to vertical orientation in imprintednanostructures.3,25−28 The edge-on and vertical orientations can beproperly analyzed using two kinds of GIXRD configurations, i.e., out-of-plane and in-plane, respectively. As shown in Figure 2a1,a2, in the out-

of-plane GIXRD, the detector is rotated vertically with respect to theP3HT surface with a scan axis of 2θ, so that the strong vertical diffractionsignals from the large polymer backbones with spacing a in edge-onorientation can be collected. However, the diffraction signals from thesmall P3HT hexyl side chains with distance c are very weak, making out-

Table 1. Number-Average Molecular Weight (Mn), WeightAverage Molecular Weight (Mw), Polydispersity (PDI), andRegioregularity (RR) of P3HT Samples Used in This Study

P3HT batch no. Mn (kDa) Mw (kDa) PDI RR (%)

B1 25 50 2.0 >96B2 35 70 2.0 >96B3 43 90 2.1 >96B4 50 110 2.2 >96

Figure 1. Schematic process flow of nanoimprint of P3HT: (a) a thin layer of P3HT was deposited on a substrate, and (b) P3HT nanogratings ofnegative replication are transferred from the mold by nanoimprint lithography. SEM images of (c) Si mold and (d) imprinted P3HT nanogratings.

Figure 2. (a1) Schematic of out-of-plane GIXRD setup with scan axis of2θ for (a2) edge-on chain alignment; (b1) in-plane GIXRD setup withscan axis of ϕ/2θχ for (b2) vertical chain alignment in P3HT. Theincident X-rays in both setups are at a small incident angleω to the planeof sample surfaces.

ACS Applied Materials & Interfaces Research Article

DOI: 10.1021/acsami.6b00192ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

B


of-plane GIXRD difficult for vertical alignment characterization.Nevertheless, it can be analyzed by in-plane measurement, whichreveals the lateral chain orientation. As illustrated in Figure 2b1,b2, inthis type of measurement, both sample stage and detector are rotatedhorizontally with scan axes of ϕ and 2θχ, respectively, so that thehorizontal diffraction signals from the large backbones with spacing aand π−π stacking spacing b in the vertical orientation can be detected. Itshould be noted that, in P3HT solar cells, the active layer is normallyvertically sandwiched between the anode and the cathode. It is thereforepreferable for their molecules to align with an orientation which allowsfor a smaller hole hopping distance along the vertical electric fielddirection and thus larger hole mobility. As shown in Figure 2, the verticalhopping distance c (∼0.38 nm) in vertical orientation is much shorterthan a (∼1.69 nm) in edge-on orientation, making the former moreefficient for hole transport. In both GIXRD setups, a wavelength of0.154 nm and angular spectrum of 3°−30° with the incident angle ω =0.5° with respect to the plane of sample surface were used. It should benoted that this incident angle had been optimized in our preliminaryexperiment, so that the X-ray could go deep into the 170 nm highnanogratings to collect the crystallite information as much as possible,while not penetrating into the substrate and causing intensity loss. Theirradiation area of the X-ray beam was 15 mm × 5 mm, smaller than theimprinting area (25 mm × 25 mm).

The light absorbance of nonimprinted thin films and imprintednanogratings on glass substrates wasmeasured by UV−vis (Agilent 8453spectrometer) with wavelength 350−800 nm.

2.2. Solar Cell Fabrication and Characterization. To study theimpacts of molecular weight dependent chain alignments on the deviceperformance, P3HT:PCBM blend (BHJ) and imprinted P3HT/PCBMsolar cells with different MWs of P3HT as listed in Table 1 werefabricated. For both types of solar cells, 20 nm PEDOT:PSS (CLEVIOSP VP Al 4083, H. C. Starck, Inc.) was spin-coated onto the patternedITO coated glass substrates (Luminescence Technology) and baked at150 °C for 15 min. Low conductivity PEDOT:PSS was chosen tominimize the measurement error from device areas due to the lateralconductivity of PEDOT:PSS. Also, 1 nm LiF and 100 nm Al werethermally evaporated onto the active layer as the cathode for all devices.For the BHJ solar cells, the active layer is made of 200 nm blendedP3HT:PCBM (Nano-C) from 1:1 P3HT:PCBM solution in chlor-obenzene, followed by being pressed by a flat Si mold at the sameconditions as imprinted ones (170 °C and 50 MPa for 600 s). For NILdevices, 120 nm PCBM was spin-coated onto P3HT nanogratings fromdichloromethane which serves as an orthogonal solvent. In this work, wecould not obtain a clear cross-section image after PCBM deposition dueto the challenge in sample preparation. Nevertheless, we can derive thatthe contact between P3HT and PCBM was well-formed from the high

Figure 3. (a) Out-of-plane GIXRD measurements of nonimprinted P3HT thin films and (b) in-plane measurements of imprinted P3HT nanogratingswith direction initially (b1) parallel and (b2) perpendicular to the incident X-ray beamwith molecular weights of 25 kDa (F1 andNIL1), 35 kDa (F2 andNIL2), 43 kDa (F3 and NIL3), and 50 kDa (F4 and NIL4). The inset figures in part a show the magnified views of the (100) peaks, corresponding toedge-on oriented crystallites. The inset figures in parts b1 and b2 show the magnified views of the (100) and (010) peaks representing vertically orientedcrystallites.



C


PCE (4.4%) generated from these devices. Otherwise, the PCE wouldbe very poor if there was an infiltration or dissolving issue as we observedbefore. The top PCBM thickness after infiltration f was estimated to be∼20 nm using

= −·+

f Lp h

w p (2)

where L is the spin-coated PCBM thickness, and w, p, and h are P3HTnanograting width, spacing, and height, respectively. In our preliminarystudies, the impact of PCBM spin-coating thickness was investigated,and we found the device performance started significantly droppingwhen L was less than 100 nm (no top PCBM left according to thisequation), proving this estimate approach is reliable. After the solar cellswere made, their current density−voltage (J−V) characteristics weremeasured using Air Mass 1.5 global solar simulated light (AM. 1.5G)calibrated using an NREL traceable KG5 color filtered siliconphotodiode (PV Measurements Inc.) to an intensity of 100 mW/cm2.Four solar cell pixels with an active area of 9 mm2 each were formed oneach substrate. To reduce the experimental errors, 10 batches of solarcells were made in this work. The average and standard deviation ofdevice characteristics were calculated from these 10 batches (10 × 4devices in total for each molecular weight).

3. RESULTS AND DISCUSSION3.1. Effects on Chain Alignment. Nonimprinted P3HT

thin films F1−F4 with four different MWs as listed in Table 1 arefirst characterized by out-of-plane GIXRD. Before measure-ments, all thin films are pressed by a flat Si mold at the samenanoimprint conditions. As shown in Figure 3a, these foursamples show large (100) peaks at 5.2°, corresponding to latticeconstant a and edge-on orientations, but with differentintensities, indicating that crystallite density in planar P3HTthin films is highly dependent on the MW. The (100) peakintensity decreases as the MW increases from F1 (25 kDa), F2(35 kDa), F3 (43 kDa), to F4 (50 kDa), suggesting that itbecomes more difficult for P3HT chains to organize when thechains become longer. It is consistent with the observations inliterature that long polymer chains are more prone toentanglement.29 People have suggested that this may cause the

alignment of neighboring conjugated segments to be moredifficult to achieve, resulting in a reduced mobility.16,19 It istherefore reasonable to speculate that the sample F1 withmedium MW P3HT is optimal for the BHJ solar cells.To characterize the lateral chain arrangement with respect to

the imprinted P3HT nanogratings using in-plane GIXRD, asshown in Figure 3b, the nanograting direction was initiallyadjusted to be parallel and perpendicular, respectively, to theincident X-ray beam. As shown in Figure 3b1, all nanostructuresshow large (100) peaks and ultralow (010) peaks whenmeasuredin the parallel direction. While in the perpendicular direction, all(100) peaks are quenched, and larger (010) peaks are present asshown in Figure 3b2. These results indicate that significantamounts of P3HT chains are vertically aligned after nanoimprint,with hexyl side chain spacing a perpendicular to and π−πstacking b along the grating direction. The tiny (010) peaks inFigure 3b1 and (100) peaks in Figure 3b2 are believed to comefrom the finite alignment variations. However, their intensitiesare much lower, suggesting that the vertical chain configurationas described is dominant within the imprinted nanogratings. Itcan also be seen that large MW sample NIL3 (43 kDa) showslarger peak intensity than medium MW samples NIL1 (25 kDa)and NIL2 (35 kDa) in Figure 3b1,b2, suggesting that NIL allowslonger P3HT chains to align at a much stronger degree, which isdifferent from what is observed in nonimprinted thin films. Thisshows for the first time that strong chain alignment in largemolecular weight P3HT can be realized. We believe thisphenomenon can be attributed to two combined mechanismsoccurring during NIL: when polymer is squeezed, forcing it toflow into the mold cavities, the molecular confinement enablesentangled polymers to reorganize, and this reorganization isfavored by longer polymer chains with large MW due to its lowerviscosity at the nanoscale.10 Second, high pressure upward flowin the molten state and hydrophobic interaction betweensidewalls of the Si mold and P3HT hexyl side chains allow thereorganized polymer chains to crystallize.3,10−12,14,27,30 Inaddition, it is noted that when MW increases further to 50 kDa

Figure 4. Normalized integrated intensities (peak area) of (a) (100) peaks in Figure 3a for nonimprinted P3HT thin films, and (b1) (100) peaks inFigure 3b1 and (b2) (010) peaks in Figure 3b2 for imprinted nanogratings with different molecular weights.



D


in NIL4, the peak intensity starts going down. This indicates thatwhen the P3HT chains become very long, NIL under the currentconditions is not effective enough to induce strong verticalalignment.To better quantitatively understand the impact of MW on the

chain alignment in both types of samples, we plot the normalizedintegrated intensities (area) of (100) peaks in Figure 3a,b1 aswell as (010) peaks in Figure 3b2 (indication of P3HT crystallitedensity), respectively, as a function of MW. As shown in Figure4a, the integrated peak intensity decreases almost linearly withMW for nonimprinted thin films. While for the imprintednanogratings in Figure 4b1,b2, it first increases linearly and moresharply with MW (NIL1 to NIL3) but drops at NIL4 in both aand b directions. These results demonstrate that longer P3HTchains are less organized in regular nonimprinted thin films butcan be effectively aligned by NIL until the MW becomes verylarge.In XRD spectra, the integrated intensity of each peak is

proportional to the P3HT crystallite density. The crystallite sizeL can be obtained by the Scherrer formula

λθ

∼Δ θ

L0.9

cos( )2 (3)

where Δ2θ is the full width half-maximum of the peak.31−33

Applying eq 2 to the (100) peaks in Figure 3a,b1, respectively,one can compare the crystallite size La in nonimprinted P3HTthin films and La* in imprinted nanogratings with different MWsalong the hexyl side chain direction a. As summarized in Table 2,

the crystallite size follows the same trend as peak intensity forboth types of P3HT samples; i.e., F1 shows the largest crystallitesize among all nonimprinted thin films, and NIL3 does so amongall imprinted nanograting samples. Moreover, when we comparethe crystallite sizes between thin films and nanogratings with thesame MW, the latter are always larger from MW 25 to 50 kDa.The largest crystallite size (∼17 nm) in nanogratings with MW43 kDa in NIL3 is also larger than the largest (∼14 nm) innonimprinted thin film with MW 25 kDa in F1. Combining all ofthese results, we can speculate that nonimprinted and imprintedsolar cells with these two MWs are likely the most efficient whenthe same types of devices are compared, respectively. It is alsoreasonable to speculate that the most efficient imprinted solarcells have a higher PCE than the most efficient BHJ ones due tothe former’s larger crystallite density, crystallite size, as well asvertical chain orientation.In literature studies, people have found that a better polymer

chain alignment is usually associated with a smaller bandgap andred-shift of the absorbance spectra. To further investigate theeffect of MW on the chain alignment, here we measure the lightabsorption of nonimprinted and imprinted samples. As shown inFigure 5a, the degree of red-shift increases with increasing MWfor nonimprinted thin film samples from F1 to F4, while that forimprinted samples in Figure 5b first increases, reaching thelargest on NIL3 with 43 kDa, and starts decreasing when MWfurther increases to 50 kDa on NIL4. They are very consistent

with the GIXRD results. Moreover, when plotted together asshown in Figure 5c, NIL3 shows a larger red-shift and broaderabsorption shoulder than the nonimprinted F1 (25 kDa), whichagain indicates that the NIL induced chain alignment for polymerwith large MW is likely stronger than that of nonimprinted thinfilms with medium MW.

3.2. Effects on Solar Cell Performance. To investigate theimpacts of P3HTMW and chain alignment on the photovoltaicsperformance, BHJ and nanoimprinted solar cells with thesedifferent MWs of P3HT in Table 1 are fabricated. Their J−Vcharacteristics are shown in Figure 6a,b, respectively. Voc, Jsc, FF,and PCE of these two types of devices extracted from the J−Vcurves are listed in Tables 3 and 4, respectively. Their overalldependences on the P3HTMW are illustrated in Figure 7. Let usfirst discuss the impacts ofMWon the BHJ solar cells. The resultsshow that Jsc has monotonic decreasing correlation with theincreasing P3HT MW. Referring to the GIXRD measurementresults, which show a monotonic decreasing correlation ofcrystallite density and size with increasing MW in Figure 4a andTable 2, we can speculate that the lowest photocurrent in thelargest MW sample F4 would be likely due to its weakest chainordering and inefficient charge transport. This speculation can befurther confirmed by the change of FF. In literature studies,people have found that FF is highly dependent on the holetransport for P3HT solar cells and can decrease with the reducedchain ordering.24,34,35 Here the constantly decreasing FF withincreasing MW agrees well with the trend in GIXRD measure-ments. The degree of chain alignment in different MWs can befurther revealed by the change of Voc. When the chain alignmentbecomes weaker, it would decrease the highest occupiedmolecular orbital (HOMO) level of conjugated polymers towiden their effective bandgap, and thus increase the Voc.

36−38 InFigure 7a, Voc keeps increasing from F1 to F4 (0.62−0.64 V),confirming that the bandgap becomes larger and the polymerchain ordering is lowering.Regarding the nanoimprinted solar cell devices, Table 4 and

Figure 7b illustrate that larger MW NIL3 (43 kDa) show higherJsc, FF, and PCE than the lower MW NIL1 (25 kDa) and NIL2(35 kDa). The performance reduces when the MW increasesfurther to the ultralarge NIL4 (50 kDa). Moreover, the Voc firstdecreases with MW from NIL1, NIL2, reaches the lowest atNIL3, and then starts increasing at NIL4. We believe thesefindings are consistent with the GIXRD results as shown inFigure 3b and light absorption results in Figure 5b, where NIL3shows the strongest chain alignment. These findings togetherdemonstrate that nanoimprint lithography is an effectivetechnique to generate stronger chain alignment and moreefficient devices at larger MW. It is also noted that the PCE ofNIL3 with large MW 43 kDa is 4.4%, higher than the mostefficient nonimprinted BHJ device F1 with mediumMW 25 kDa(PCE of 3.45%). Although NIL3 shows a red-shift and broaderabsorption shoulder when compared to those of F1, which mayhave positive impact on exciton generation, it would befarfetched to attribute such a great increase in PCE (∼1%) tolight absorption enhancement only. Instead, similar to reports inthe literature, we tend to believe they are signs of a smallerbandgap due to better chain alignment, allowing for better chargetransport and efficiency, as further supported by NIL3’s largercrystallite size (Table 2), lower Voc, and higher FF than F1. TheNIL defined bicontinuous interdigitized heterojunction maybenefit the charge transport as well, since the phase separation inBHJ solar cells is difficult to precisely control.

Table 2. Effects of P3HT Molecular Weight on the CrystalliteSizes in Nonimprinted P3HT Thin Films (La) and ImprintedP3HT Nanogratings (La*)

P3HT MW 25 kDa 35 kDa 43 kDa 50 kDa

La (nm) 14.23 13.74 12.85 11.53La* (nm) 14.82 15.31 16.96 15.67



E


From these results we believe that the top-down nano-structured polymer solar cells by nanoimprint lithography can bean effective device architecture by enabling the benefits of largerMW polymers. The Jsc, FF, and PCE of nanoimprinted solar cellsusing large MW 43 kDa P3HT show higher PCE than BHJ

devices using the optimal MW 25 kDa. The performance of thistype of solar cells can be further improved with stronger NILinduced chain alignment, so that we can use even larger MW(e.g., NIL4 or larger) P3HT to fully unleash its potential inefficient charge transport. In our previous work, we have

Figure 5. Light absorbance of (a) nonimprinted P3HT thin films and (b) imprinted nanogratings with molecular weights of 25 kDa (F1 and NIL1), 35kDa (F2 and NIL2), 43 kDa (F3 and NIL3), and 50 kDa (F4 and NIL4). (c) Light absorbance comparison between nonimprinted F1 and imprintedNIL3.

Figure 6. J−V characteristics of the P3HT and PCBM solar cells fabricated on (a) P3HT:PCBM blend and (b) PCBM coated P3HT nanogratings withdifferent P3HT molecular weights: 25 kDa (F1 and NIL1), 35 kDa (F2 and NIL2), 43 kDa (F3 and NIL3), and 50 kDa (F4 and NIL4).

Table 3. Performance of P3HT:PCBM Photovoltaic Devices with P3HT and PCBM Blend

device F1 F2 F3 F4

Voc (V) 0.62 ± 0.01 0.63 ± 0.01 0.63 ± 0.01 0.64 ± 0.01Jsc (mA/cm

2) 9.26 ± 0.54 8.15 ± 0.32 7.58 ± 0.27 6.87 ± 0.48FF 0.60 ± 0.01 0.57 ± 0.01 0.55 ± 0.02 0.53 ± 0.02PCE (%) 3.45 ± 0.12 2.92 ± 0.06 2.63 ± 0.05 2.33 ± 0.09

Table 4. Performance of P3HT/PCBM Photovoltaic Devices with Imprinted P3HT Nanogratings

device NIL1 NIL2 NIL3 NIL4

Voc (V) 0.63 ± 0.01 0.62 ± 0.01 0.61 ± 0.01 0.62 ± 0.01Jsc (mA/cm

2) 7.75 ± 0.45 8.79 ± 0.32 10.98 ± 0.28 9.68 ± 0.39FF 0.58 ± 0.01 0.60 ± 0.01 0.66 ± 0.02 0.61 ± 0.01PCE (%) 2.83 ± 0.11 3.27 ± 0.05 4.42 ± 0.08 3.66 ± 0.06



F


demonstrated that the chain alignment becomes stronger whenthe interface area of imprinted polymer nanostructure increases,due to the stronger polymer-to-mold interaction duringimprint.39 Therefore, a higher device efficiency can be achievedby further decreasing the width while increasing the height ofP3HT nanostructures, i.e., increasing the interface area. Inaddition to charge transport, the increase in interface area wouldalso improve the exciton separation between P3HT and PCBM.There is a lot of room to increase this interface area since the IEFis less than 3.5 in this study.

4. CONCLUSIONIn this work, strong chain alignment in large molecular weightsolar cell polymer P3HT is for the first time demonstrated bynanoimprint lithography. Consequently P3HT/PCBM solarcells made from large MW P3HT nanogratings show a high PCEof 4.4%, which is attributed to the efficient charge separation andtransport. The results indicate that the top-down nanostructuredpolymer solar cells created by nanoimprint lithography can be aneffective device architecture by enabling the benefits of largeMWpolymers.

■ AUTHOR INFORMATIONCorresponding Authors*E-mail: [email protected].*E-mail: [email protected].

NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTSThis work is supported by the National Science Foundation(NSF) (Grant ECCS-0901759), Welch Foundation Grant AT-

1617, and the Ministry of Education and Science of the RussianFederation in the framework of Increase CompetitivenessProgram of NUST “MISiS” (K2-2015-014). The authorsgratefully acknowledge M. Stefan from Department ofChemistry, J. Hsu from Department of Materials Science andEngineering, and P. Zang from Department of ElectricalEngineering at UT Dallas for their support and helpfuldiscussions.

■ REFERENCES(1) Lewis, N. S. Toward Cost-Effective Solar Energy Use. Science 2007,315, 798−801.(2) Service, R. F. Outlook Brightens for Plastic Solar Cells. Science2011, 332, 293−293.(3) Chang, J.-F.; Sun, B.; Breiby, D. W.; Nielsen, M. M.; Solling, T. I.;Giles, M.; McCulloch, I.; Sirringhaus, H. Enhanced Mobility of Poly(3-hexylthiophene) Transistors by Spin-Coating from High-Boiling-PointSolvents. Chem. Mater. 2004, 16, 4772−4776.(4) Kroeze, J. E.; Savenije, T. J.; Vermeulen, M. J. W.; Warman, J. M.Contactless Determination of the Photoconductivity Action Spectrum,Exciton Diffusion Length, and Charge Separation Efficiency inPolythiophene-Sensitized TiO2 Bilayers. J. Phys. Chem. B 2003, 107,7696−7705.(5) Haugeneder, A.; Neges, M.; Kallinger, C.; Spirkl, W.; Lemmer, U.;Feldmann, J.; Scherf, U.; Harth, E.; Gugel, A.; Mullen, K. ExcitonDiffusion andDissociation in Conjugated Polymer Fullerene Blends andHeterostructures. Phys. Rev. B: Condens. Matter Mater. Phys. 1999, 59,15346−15351.(6) Yu, G.; Gao, J.; Hummelen, J. C.; Wudl, F.; Heeger, A. J. PolymerPhotovoltaic Cells - Enhanced Efficiencies via a Network of InternalDonor-Acceptor Heterojunctions. Science 1995, 270, 1789−1791.(7) Ma, W. L.; Yang, C. Y.; Gong, X.; Lee, K.; Heeger, A. J. ThermallyStable, Efficient Polymer Solar Cells with Nanoscale Control of theInterpenetrating Network Morphology. Adv. Funct. Mater. 2005, 15,1617−1622.

Figure 7.Overall impacts of P3HTmolecular weight on (a) Voc, (b) Jsc, (c) FF, and (d) PCE of P3HT:PCBM blend and imprinted P3HT/PCBM solarcell performance.



G

mailto:[email protected]

mailto:[email protected]


(8) Li, G.; Shrotriya, V.; Huang, J. S.; Yao, Y.; Moriarty, T.; Emery, K.;Yang, Y. High-Efficiency Solution Processable Polymer PhotovoltaicCells by Self-Organization of Polymer Blends.Nat. Mater. 2005, 4, 864−868.(9)Watkins, P. K.;Walker, A. B.; Verschoor, G. L. B. Dynamical MonteCarlo Modelling of Organic Solar Cells: The Dependence of InternalQuantum Efficiency on Morphology. Nano Lett. 2005, 5, 1814−1818.(10) Rowland, H. D.; King, W. P.; Pethica, J. B.; Cross, G. L. W.Molecular Confinement Accelerates Deformation of EntangledPolymers during Squeeze Flow. Science 2008, 322, 720−724.(11) Lovinger, A. J.; Davis, D. D. Electron-Microscopic Investigation ofthe Morphology of a Melt-Crystallized Polyaryletherketone. J. Appl.Phys. 1985, 58, 2843−2853.(12) Kanazawa, H.; Stejny, J.; Keller, A. Polymerization and Structureof Poly(Sulphur Nitride) Prepared under High Pressure. J. Mater. Sci.1991, 26, 1635−1639.(13) Osaka, I.; McCullough, R. D. Advances in Molecular Design andSynthesis of Regioregular Polythiophenes. Acc. Chem. Res. 2008, 41,1202−1214.(14) Beck, V. A.; Shaqfeh, E. S. G. Ergodicity-Breaking and theUnraveling Dynamics of a Polymer in Linear and Nonlinear ExtensionalFlows. J. Rheol. 2007, 51, 561−574.(15) Kline, R. J.; McGehee, M. D.; Kadnikova, E. N.; Liu, J.; Frechet, J.M. J.; Toney, M. F. Dependence of Regioregular Poly(3-hexylth-iophene) Film Morphology and Field-Effect Mobility on MolecularWeight. Macromolecules 2005, 38, 3312−3319.(16) Ballantyne, A. M.; Chen, L.; Dane, J.; Hammant, T.; Braun, F. M.;Heeney, M.; Duffy, W.; McCulloch, I.; Bradley, D. D. C.; Nelson, J. TheEffect of Poly(3-hexylthiophene) Molecular Weight on ChargeTransport and the Performance of Polymer:Fullerene Solar Cells. Adv.Funct. Mater. 2008, 18, 2373−2380.(17) Liu, F.; Chen, D.; Wang, C.; Luo, K.; Gu, W.; Briseno, A. L.; Hsu,J. W. P.; Russell, T. P. Molecular Weight Dependence of theMorphology in P3HT:PCBM Solar Cells. ACS Appl. Mater. Interfaces2014, 6, 19876−19887.(18) Kline, R. J.; McGehee, M. D.; Kadnikova, E. N.; Liu, J.; Frechet, J.M. J. Controlling the Field-Effect Mobility of Regioregular Poly-thiophene by Changing the Molecular Weight. Adv. Mater. 2003, 15,1519−1522.(19) Hiorns, R. C.; de Bettignies, R.; Leroy, J.; Bailly, S.; Firon, M.;Sentein, C.; Khoukh, A.; Preud’homme, H.; Dagron-Lartigau, C. HighMolecular Weights, Polydispersities, and Annealing Temperatures inthe Optimization of Bulk-Heterojunction Photovoltaic Cells Based onPoly(3-hexylthiophene) or Poly(3-butylthiophene). Adv. Funct. Mater.2006, 16, 2263−2273.(20) Yang, Y.; Mielczarek, K.; Aryal, M.; Zakhidov, A.; Hu, W.Nanoimprinted Polymer Solar Cell. ACS Nano 2012, 6, 2877−2892.(21) Yang, Y.; Aryal, M.; Mielczarek, K.; Hu, W.; Zakhidov, A.Nanoimprinted P3HT/C60 Solar Cells Optimized by ObliqueDeposition of C60. J. Vac. Sci. Technol., B: Nanotechnol. Microelectron.:Mater., Process., Meas., Phenom. 2010, 28, C6M104−C106M107.(22) Yang, Y.; Lee, K.; Mielczarek, K.; Hu, W.; Zakhidov, A.Nanoimprint of Dehydrated PEDOT:PSS for Organic Photovoltaics.Nanotechnology 2011, 22, 485301.(23) Hlaing, H.; Lu, X.; Hofmann, T.; Yager, K. G.; Black, C. T.; Ocko,B. M. Nanoimprint-Induced Molecular Orientation in SemiconductingPolymer Nanostructures. ACS Nano 2011, 5, 7532−7538.(24) Chen, D.; Zhao, W.; Russell, T. P. P3HT Nanopillars for OrganicPhotovoltaic Devices Nanoimprinted by AAO Templates. ACS Nano2012, 6, 1479−1485.(25) Aryal, M.; Trivedi, K.; Hu, W. Nano-Confinement Induced ChainAlignment in Ordered P3HT Nanostructures Defined by NanoimprintLithography. ACS Nano 2009, 3, 3085−3090.(26) Yokoyama, T.; Seki, K.; Morisada, I.; Edamatsu, K.; Ohta, T. X-Ray Absorption Spectra of Poly-P-Phenylenes and Polyacenes:Localization of π Orbitals. Phys. Scr. 1990, 41, 189−192.(27) Kline, R. J.; McGehee, M. D.; Toney, M. F. Highly OrientedCrystals at the Buried Interface in Polythiophene Thin-Film Transistors.Nat. Mater. 2006, 5, 222−228.

(28) Shinohara, T.; Higaki, Y.; Hoshino, T.; Masunaga, H.; Ogawa, H.;Okamoto, Y.; Aoki, T.; Takahara, A. Buried Nano-Structure andMolecular Aggregation State in Ordered Heterojunction Poly(3-hexylthiophene)-Based Photovoltaics. Jpn. J. Appl. Phys. 2014, 53,05FH09.(29) Cowie, J. M. G. Polymers: Chemistry and Physics of ModernMaterials; Nelson Thornes: Cheltenham, 1991.(30) Yamamoto, K.; Ochiai, S.; Wang, X.; Uchida, Y.; Kojima, K.;Ohashi, A.; Mizutani, T. Evaluation of Molecular Orientation andAlignment of Poly(3-hexylthiophene) on Au (111) and on Poly(4-vinylphenol) Surfaces. Thin Solid Films 2008, 516, 2695−2699.(31) Zhokhavets, U.; Erb, T.; Hoppe, H.; Gobsch, G.; Sariciftci, N. S.Effect of Annealing of Poly(3-hexylthiophene)/Fullerene Bulk Hetero-junction Composites on Structural and Optical Properties. Thin SolidFilms 2006, 496, 679−682.(32) Patterson, A. L. The Scherrer Formula for X-Ray Particle SizeDetermination. Phys. Rev. 1939, 56, 978−982.(33) Erb, T.; Zhokhavets, U.; Gobsch, G.; Raleva, S.; Stuhn, B.;Schilinsky, P.; Waldauf, C.; Brabec, C. J. Correlation between Structuraland Optical Properties of Composite Polymer/Fullerene Films forOrganic Solar Cells. Adv. Funct. Mater. 2005, 15, 1193−1196.(34) Shih, C. F.; Hung, K. T.; Wu, J. W.; Hsiao, C. Y.; Li, W. M.Efficiency Improvement of Blended Poly(3-hexylthiophene) and 1-(3-methoxycarbonyl)-propyl-1-phenyl-(6,6)C(61) Solar Cells by Nano-imprinting. Appl. Phys. Lett. 2009, 94, 143505.(35) Lee, J. H.; Kim, D. W.; Jang, H.; Choi, J. K.; Geng, J.; Lung, J. W.;Yoon, S. C.; Jung, H.-T. Enhanced Solar-Cell Efficiency in Bulk-Heterojunction Polymer Systems Obtained by Nanoimprinting withCommercially Available AAO Membrane Fitters. Small 2009, 5, 2139−2143.(36) Yang, Y.; Mielczarek, K.; Zakhidov, A.; Hu, W. Efficient LowBandgap Polymer Solar Cell with Ordered Heterojunction Defined byNanoimprint Lithography. ACS Appl. Mater. Interfaces 2014, 6, 19282−19287.(37) Peet, J.; Kim, J. Y.; Coates, N. E.; Ma, W. L.; Moses, D.; Heeger, A.J.; Bazan, G. C. Efficiency Enhancement in Low-Bandgap Polymer SolarCells by Processing with Alkane Dithiols.Nat. Mater. 2007, 6, 497−500.(38) Di Nuzzo, D.; Aguirre, A.; Shahid, M.; Gevaerts, V. S.; Meskers, S.C. J.; Janssen, R. A. J. Improved Film Morphology Reduces ChargeCarrier Recombination into the Triplet Excited State in a Small BandgapPolymer-Fullerene Photovoltaic Cell. Adv. Mater. 2010, 22, 4321−4324.(39) Yang, Y.; Mielczarek, K.; Aryal, M.; Zakhidov, A.; Hu, W. Effectsof Nanostructure Geometry on Nanoimprinted Polymer Photovoltaics.Nanoscale 2014, 6, 7576−7584.



H


large molecular weight polymer solar cells with strong chain …walter.hu/research... ·...

Documents