Efficiency enhancement of P3HT:PCBM solar cells containing scattering Zn-Al hydrotalcite...

Organic Photonics and PhotovoltaicsResearch Article • DOI: 10.2478/oph-2013-0001 • OPP • 2013 • 1-10

Efficiency enhancement of P3HT:PCBM solar cellscontaining scattering Zn-Al hydrotalcite nanoparticlesin the PEDOT:PSS layerAbstractIn this article we report on a new hybrid (organic-inorganic)composite material based on hydrophilic, electrically inert andsemi-transparent hydrotalcite (HT) nanoparticles and a pH-neutral formulation of PEDOT:PSS. The application of thiscomposite material as electrically and optically active bufferlayer in P3HT:PC61BM bulk heterojunction (BHJ) solar cellsis reported. Two different synthetic routes are explored to ob-tain HTs having discoid shape, with a diameter of around 150-200 nm and a thickness of ∼20 nm, to be easily embedded in∼50 nm thick PEDOT:PSS films. The good affinity betweenHTs and the sulfonate groups of the PEDOT:PSS allows toobtain homogeneous HTs/PEDOT:PSS films, for HT concen-trations of 0.25% and 0.50% by weight (vs. PEDOT:PSS). Atthese particle loads the electrical and morphological featuresof doped and undoped PEDOT:PSS films are nearly identical,while providing a significant effect on the visible light scat-tering properties of the composite films. We demonstrate∼12% improvement in power conversion efficiency (PCE)for P3HT:PC61BM solar cells incorporating HTs in the PE-DOT:PSS layer, which mainly originates from increased short-circuit current densities (JSC ).

Keywordscomposite film • hydrotalcite • light scattering • bulk-heterojunction OPV • PEDOT:PSS

PACS: 78.66.Sq, 82.70.Uv, 78.35.+c, 81.07.Pr, 88.40.jr© Versita sp. z o.o.

Margherita Bolognesi1, Marta Tessarolo2, Tamara Posati1∗,Morena Nocchetti3, Valentina Benfenati4, Mirko Seri4‡,Giampiero Ruani2§, Michele Muccini2

1 Laboratory MIST E-R, Via P. Gobetti, 101 - Bologna I-40129, Italy

2 Consiglio Nazionale delle Ricerche, Istituto per lo Studio deiMateriali Nanostrutturati (CNR-ISMN),Via P. Gobetti,101 - Bologna I-40129, Italy

3 Dipartimento di Chimica and Centro di Eccellenza MaterialiInnovativi Nanostrutturati (CEMIN), Università di Perugia,Via Elce di Sotto 10, 06123 Perugia, Italy

4 Consiglio Nazionale delle Ricerche, Istituto per la Sintesi Organicae la Fotoreattività (CNR-ISOF) Via P. Gobetti,101 - Bologna I-40129, Italy

Received 2/20/2013Accepted 4/24/2013

1. Introduction

Organic photovoltaic (OPV) solar cells have being in-tensively studied because of their promising large-scalefabrication by printing techniques, easily transferrable tocontinuous roll-to-roll industrial processes [1, 2]. The ac-tive materials engineering is rapidly leading to break-through the 10% module efficiency limit, necessary for theOPV technology to be adopted in the industrial world [3–6]. However, the main limiting factor for the OPV cellsperformances is the difficulty to maximize external quan-tum efficiencies (EQEs), because of the organic active

∗E-mail: t.posati@bo.ismn.cnr.it‡E-mail: mirko.seri@isof.cnr.it§E-mail: g.ruani@bo.ismn.cnr.it

†Electronic Supplementary Information (ESI) available.

layer thickness limitations. This is mainly due to the mis-match between the common light absorption depth (hun-dreds of nanometers) and the exciton diffusion length (tensof nanometers) in the photoactive film. The bulk heterojuc-tion (BHJ) approach attempts to solve this intrinsic prob-lem, but carrier transport through the contorted BHJ mor-phology remains the bottleneck that limits the maximumsuitable thickness of the active layer [7–9].

This limit could be overcome if technologically-viable, ro-bust, cheap light management techniques are employed.Light managing techniques for optoelectronic devices suchas solar cells span from collector mirrors, [10] fiber basedand micro lenses, [11, 12] light concentrators [13], lumi-nescent concentrators [14–16], anti-reflection coatings [17]and surface plasmon resonators [18–22]. Many of thesetechniques exploit the use of additional layers and/orhighly ordered and periodic nano-structures able to trap

1

Authenticated | g.ruani@bo.ismn.cnr.it author's copyDownload Date | 11/8/13 10:29 AM

M. Bolognesi et al.

the light into the active layer enhancing its total lightabsorption. Nevertheless, the advanced techniques usedto prepare or manipulate these highly ordered micro- andnano-structures are often limited to lab-scale use [23, 24].Moreover, surface plasmon resonators and solar concen-trators often require the use of rare or expensive materials,such as gold, silver, rare earth ionic complexes and metalquantum dots [25].

The simplest light trapping approach that could enhancethe active layer absorption and cell photocurrent is theemployment of light scattering centres in the device struc-ture. The randomization of the light propagation pathwayinside the solar cell stack results in an increased inter-action between the solar radiation and the photoactiveabsorbing layer. This, in principle, might lead to a sub-stantial enhancement of active layer absorption, deviceshort-circuit photocurrent and PCE [26, 27].

In the dye sensitized solar cells (DSSCs) technology, theuse of a scattering material in the device structure is aconsolidated technique since more than fifteen years [28–33]. Indeed, the introduction of a TiO2,layer with largenanoparticles [28–31] or the dispersion of nanozeolitesinto the standard TiO2 nanoporous layer has shown sig-nificant improvements in terms of performance and life-time stability [32, 33]. On the other hand, controlledlight scattering in BHJ OPVs is a relatively recent tech-nique. In particular, research has focused on the use ofmetal oxide (TiO2, ZnO, MoO3, . . . ) [34–37] or metallic (Au,Ag,. . . ) [18–22] nanostructured materials (often nanopar-ticles) dispersed in the buffer layer or in the active layerof the OPV stack. However, these materials allowed theenhancement of the OPV performances by mainly exploit-ing a number of effects, including plasmon resonance andbuffer layer work-function tuning, rather than light scat-tering.

Hydrotalcites are among the most attractive cheap andnon-toxic micro- and nano-structured inorganic materialsand have been already used in a wide range of applica-tions in the fields of polymeric nanocomposites, [38–43]heterogeneous catalysis, [44–47] photochemistry [48–52]and dye sensitized photovoltaics [53–56].

In detail, HTs have the general formula[M(II)1−xM(III)x (OH)2]x+[Anx/n]mH2O, where M(III) cationsare typically Al, Cr, Fe, or Ga and M(II) are typicallyMg, Zn, Ni, Co, or Cu, while An− is a charge balancinganion (organic, inorganic or metallorganic) with n asionic valence and x is the M(III) molar fraction. They canbe easily synthesized in a variety of dimensions, withcrystallite diameters ranging from some micrometers [57]to nanometers [58–61]. For their chemical nature, HTsreveal a high affinity with the sulfonate groups of the PSS

Fig 1. a) Schematic representation of a standard BHJ OPV and b) aBHJ OPV with the composite HTs/PEDOT:PSS film acting asscattering layer.

(poly(4-styrenesulfonate)) [62], enabling their potentialuse in composite materials with PEDOT:PSS.

One of the most common type of HTs are the Zn(II)-Al(III) based HTs (ZnAl-HTs), that can be easily syn-thesized in the form of crystalline nanoplatelets witha diameter of about 150-200 nm and a thickness of∼20 nm [60, 61]. This high aspect-ratio could allow apreferential in-plane self-orientation of HTs in the com-posite ZnAl-HTs/PEDOT:PSS material, thus leading tohomogeneous and smooth films with a maximized scatter-ing cross-section per particle.

On the basis of the hydrophilicity, electrical inertnessand semi-transparency properties of ZnAl-HTs, we reporthere on their use in the PEDOT:PSS layer of OPV so-lar cells. P3HT:PC61BM based BHJ solar cells incor-porating pristine PEDOT:PSS or ZnAl-HTs/PEDOT:PSSfilms have been fabricated with the standard OPV struc-ture (Figure 1), revealing an interesting and effectivebi-functionality of the ZnAl-HTs/PEDOT:PSS compositefilm, both as hole transporting and light scattering layer.Two different synthetic routes have been used to prepareZnAl-HTs: the first one is based on a variation of the ureamethod, using a mixture of water and ethylene glycol assolvent [54], while the second one is based on the dou-ble microemulsion process [60, 61]. Both methods led tocrystalline ZnAl-HTs (called respectively ZnAl-HTwg andZnAl-HTµE ) with similar dimensions but different aggre-gation tendencies.

Since HTs are metal hydroxides, which can be dissolvedin acid environments, a pH-neutral PEDOT:PSS formu-lation is used in this work. The morphological, op-tical and electrical properties of the composite ZnAl-HTwg/PEDOT:PSS and ZnAl-HTµE /PEDOT:PSS filmshave been studied for different HTs concentrations andcorrelated with the corresponding OPV performances.

2

Authenticated | g.ruani@bo.ismn.cnr.it author's copyDownload Date | 11/8/13 10:29 AM

Efficiency enhancement of P3HT:PCBM solar cells…

2. Experimental section2.1. Materials and methods

All reagents, regioregular poly-3-hexyl-thiophene(P3HT) from Rieke Metals, [6,6]-phenyl-C61-butyric acidmethyl ester (PC61BM) from American Dye Source,poly(3,4-ethylenedioxythiophene)-poly(styrenesulfonate)(PEDOT:PSS), Orgacon™ N-1005 (Sigma-Aldrich),Zn(NO3)2·6H2O, Al(NO3)3·9H2O, cetyltrimethylammo-nium bromide (CTAB), urea and various solvents werepurchased from commercial sources and used withoutfurther purification unless otherwise specified. Syntheticdetails and the procedure for the preparation of colloidalaqueous dispersions of ZnAl-HTs nanoparticles are givenin ESI†.

2.2. HTs/PEDOT:PSS thin-films preparation andcharacterization

pH-neutral PEDOT:PSS was used after previous filtrationwith a 0.45 µm filter. The ZnAl-HTs dispersions in waterwere sonicated for 10 minutes before use. The concentra-tion of ZnAl-HTµE and ZnAl-HTwgstarting solutions were6.6 mg/mL (for lower concentrations mixed solutions) or15 mg/mL (for higher concentrations). To avoid dilutioneffects, small volumes (at maximum, a volume of 67 µLin 1 mL, used for the 5 wt% ZnAl-HTs/PEDOT:PSS so-lution) of these ZnAl-HTs dispersions were added to thePEDOT:PSS aqueous solution to prepare mixed ZnAl-HTs/PEDOT:PSS blends with the following nanoparticlesloads: 0.25%, 0.50%, 1%, 5% (weight % vs. PEDOT:PSSsolid content). Each solution was sonicated at roomtemperature for 10 minutes, then ZnAl-HTs/PEDOT:PSSfilms were prepared by spin coating (on pre-cleaned andplasma-treated glass substrates) at 4000 rpm for 1 minuteand subsequent annealing at 120C for 10 minutes. Thethicknesses of the various films were measured with a pro-filometer (KLA Tencor, P-6) and all resulted ∼50 nm thick.HTs and HTs/PEDOT:PSS films were imaged with a scan-ning electron microscope SEM, ZEISS LEO 1530 FEG,after metallization with gold. Atomic Force Microscopy(AFM) images were taken with a Solver Pro (NT-934MDT) scanning probe microscope in tapping mode.For XRD analysis, the films were prepared by drop cast-ing blends with HTs loads of 5 wt% for both ZnAl-HTµEand ZnAl-HTwg , in order to achieve detectable varia-tions. X-ray diffraction (XRD) patterns were recorded witha Philips X’PERT PRO MPD diffractometer operating at40 kV and 40 mA, step size 0.0170 2θ degree and 20 sstep scan, using the CuKα radiation and an X’Celeratordetector.The total transmittance (T%) of the doped and undopedPEDOT:PSS films was measured with a Perkin-Elmer

Lambda-9 UV-vis spectrophotometer equipped with an in-tegrating sphere.The descriptive scattering parameters of the doped andundoped PEDOT:PSS films were determined experi-mentally from laser in-line transmission (LT) measure-ments [63]. Haze parameter H [64, 65] have been de-rived by measuring the linear and total transmission ofa monochromatic light passing through the sample. The488 nm emission of an Ar+ Innova 90 Coherent laserpasses perpendicularly through the sample (the power isset at 40 mW in order to have a high signal to noise ratiobut avoiding any possible distortion due to local heating ofthe sample) and the transmitted light is collected througha lens by a Ophir laser power meter. The transmittedlight collected after passing through a pinhole positioned50 cm after the sample, with a narrow (φ = 1 mm) or wide(φ = 20 mm) aperture, are called Tlin (in-line transmittedlight) and Ttot (total transmitted light) respectively.The conductivity of doped and undoped PEDOT:PSS filmswas measured through the four-point probe technique ap-plied to devices with glass/ITO electrodes arranged in thegeometry reported in ESI† [66].

2.3. BHJ OPV Device fabrication and character-ization

Patterned ITO-coated glasses (Rs ∼10 Ω/sq, ITO rough-ness RMS <1 nm) were cleaned in sequential sonicatingbaths for 10 min, twice in acetone and once in isopropanol,then cleaned in an ozone-plasma chamber for 10 min.Next, the PEDOT:PSS or HTs/PEDOT:PSS layer wasspun-cast from the corresponding acqueous suspensionson the ITO surface at 4000 rpm, for 1 minute, and sub-sequently annealed at 120C for 10 minutes (as for thefilms on glass). Samples were then transferred inside theglove box (<0.1 ppm of O2 and H2O). A P3HT:PC61BM 1:1(wt/wt) solution was previously prepared in glove box witha total concentration of 60 mg/mL in dry chlorobenzene:orthodichlorobenzene (1:1 v/v) and left under stirring andheating at 70C overnight. The P3HT:PC61BM solu-tion was then spun-cast on top of the ITO/PEDOT:PSSor ITO/ZnAl-HTs/PEDOT:PSS surface at 1000 rpm for10 seconds. The wet active layers were slowly driedand then annealed at 110C for 10 minutes resulting in∼350 nm thick films. To complete the device fabrica-tion, LiF/Al cathodes (0.6 nm and 100 nm respectively)were next deposited sequentially without breaking vac-uum (∼3×10−6 Torr) using a thermal evaporator directlyconnected to the glove box. The current–voltage (I–V)characteristics of complete OPV devices were recordedby a Keithley 236 source-measure unit under simulatedAM1.5G illumination of 100 mW/cm2 (Abet TechnologiesSun 2000 Solar Simulator). The light intensity was de-

3

Authenticated | g.ruani@bo.ismn.cnr.it author's copyDownload Date | 11/8/13 10:29 AM

M. Bolognesi et al.

termined by a standard silicon photodiode filtered with aKG5 color glass filter to bring spectral mismatch to unity.During testing, each device was illuminated through acalibrated mask (exactly 6 mm2), to avoid any excess ofphotocurrent generated from the parasitic device regionsoutside the overlapped electrode area. All solar cells weretested inside the glove box in oxygen and moisture free en-vironment. External Quantum Efficiency (EQE) was mea-sured with a home built system on encapsulated devices:monochromatic light was obtained with a Xenon arc lampfrom Lot-Oriel (300 Watt power) coupled with a Spectra-Pro monochromator. The photocurrent produced by thedevice passed through a calibrated resistance (50 Ω) andthe voltage signal was collected with a Lock-In DigitalAmplifier SR830. Signal was pulsed by means of anoptical chopper (∼500 Hz frequency). A calibrated UV-enhanced Silicon photodiode was used as reference.

3. Results and discussion

3.1. HTs/PEDOT:PSS thin-film properties

Before comparing the OPV properties of doped andundoped PEDOT:PSS based solar cells, we investi-gated the structural properties of HTs once dispersedin the polymeric matrix, together with the result-ing HTs/PEDOT:PSS composite thin-film morphological,electrical and optical properties. Some key parameters,required for these films to be used as scattering layers inOPV devices, were analyzed: i) the chemical and struc-tural stability of the HTs in the PEDOT:PSS matrix; ii)the homogeneous distribution of the HTs in the film, com-bined with a preferential in-plane self-orientation; iii) theHTs/PEDOT:PSS composite film surface regularity andflatness, to favor the subsequent deposition of the pho-toactive layer; iv) the maintenance of the optical (highvisible light transmission) and electrical (high conduc-tivity) properties of the HTs/PEDOT:PSS thin-film. Tothis end, a variety of techniques, such as x-ray diffrac-tion (XRD), scanning electron microscopy (SEM), atomicforce microscopy (AFM), visible light total transmissionand electrical conductivity, have been used for deep in-vestigations.Figure 2a shows the X-ray diffraction (XRD) patterns ofpristine ZnAl-HTµE and ZnAl-HTwg powders comparedto the ZnAl-HTµE and ZnAl-HTwg films, while in Fig-ure 2b are reported those of the ZnAl-HTs/PEDOT:PSShybrid films prepared with a HTs load of 5 wt% (requiredfor detectable variations) compared to the undoped PE-DOT:PSS film. It can be seen that in the composite filmsthe main reflections at 8.0 Å and 8.7 Å (Figure 2b) are thesame as in the HTs pristine films (Figure 2a) and compat-ible with the presence of bromide and nitrate anions in

the interlayer region [41, 60]. The presence of these re-flections indicates that the lamellar structure of HTs ispreserved in the polymer matrix, excluding any delami-nation process due to the HTs/PEDOT:PSS interactions.Moreover, the very weak reflection detected at 4.4, cor-responding to a d spacing of 20.6 Å, indicates a partialintercalation of some PSS species in the interlayer regionof HTs [62], confirming the good affinity between sulfonategroups of the PSS and HTs. In addition, unlike the powderspectra (Figure 2a), where the HTs are randomly arranged,the absence of any in-plane reflections (h,k 6= 0) at highangles evidences a favourable in-plane orientation of theHTs nanoplatelets in the composite film [67, 68].Scanning electron microscopy (SEM) was used to char-acterize the topography of the blend films. In agreementwith the previous considerations, the SEM image of theZnAl-HTµE /PEDOT:PSS film with 0.50 wt% of HTs (seeFigure 3b) shows that the HTs nanoparticles are homo-geneously dispersed in the PEDOT:PSS matrix. In addi-tion, the preferential in plane orientation of the plateletsis confirmed, since the majority of the HTs are embeddedin the bulk PEDOT:PSS layer and only a few superficialprotrusions are present. However, the SEM image of theblend ZnAl-HTwg/PEDOT:PSS with 0.50 wt% HTs load(see Figure 3d) shows some nanoparticle aggregates, in-dicating the slightly higher aggregation tendency of ZnAl-HTwg, compared to ZnAl-HTµE , in agreement with theimages of pristine ZnAl-HTwg (Figure 3c) and ZnAl-HTµE(Figure 3a) films.The morphological properties of the doped and undopedPEDOT:PSS based films have been also investigated byAtomic Force Microscopy (AFM, Figure 4). The topo-graphical analysis of ZnAl-HTµE /PEDOT:PSS films re-veal that ZnAl-HTµE loads of 0.50 wt% and 1 wt% (Fig-ure 4b and 4c) have a weak influence on the film roughness(RMS of ∼5.0 and 5,6 nm,respectively), which is very sim-ilar to that of the undoped PEDOT:PSS film (Figure 4a,RMS ∼5.0 nm). However, by further increasing the ZnAl-HTµE content up to 5 wt%, the RMS sharply increases to16.8 nm, due to the formation of micrometric aggregates(Figure 4d). A similar trend, with a higher dependence onthe HTs loads, was found for ZnAl-HTwg based compos-ites, for which the RMS increases from ∼5.1 nm to ∼10 nmand ∼30 nm for particles loads of 0.50 wt% (Figure 4e),1 wt% (Figure 4f) and 5 wt% (Figure 4g) respectively, inaccordance with the higher aggregation tendency of ZnAl-HTwg with respect to ZnAl-HTµE .Finally, the total transmittance (T%) of the PE-DOT:PSS film and of the most highly doped (5 wt%)HTs/PEDOT:PSS films (see Figure S1 in the ESI†), mea-sured with a spectrophotometer equipped with an inte-grating sphere to take into account the contributions of

4

Authenticated | g.ruani@bo.ismn.cnr.it author's copyDownload Date | 11/8/13 10:29 AM

Efficiency enhancement of P3HT:PCBM solar cells…

Fig 2. X-ray diffraction patterns (XRD, on glass substrates) of: a) ZnAl-HTµE and ZnAl-HTwg powders compared to the ZnAl-HTµE and ZnAl-HTwgfilms; b) ZnAl-HTs/PEDOT:PSS hybrid films prepared with a HTs load of 5 wt% vs. PEDOT:PSS, compared to the reference PEDOT:PSSfilm.

Fig 3. SEM images of: a) pristine ZnAl-HTµE ; b) 0.50 wt% ZnAl-HTµE /PEDOT:PSS; c) pristine ZnAl-HTwg and d) 0.50 wt% ZnAl-HTwg/PEDOT:PSSbased films deposited on glass substrates.

both unscattered and scattered light, results unvaried inthe whole visible range (350 – 850 nm). This excludesthe possibility that any variation in the photovoltaic re-sponse of the solar cells could be due to a change intransparency of the buffer layer following the inclusionof HTs. Analogously, the conductivity experiments on theHTs/PEDOT:PSS films revealed that any influence on thePEDOT:PSS conductivity due to the presence of HTs inthe film (up to 5 wt% of ZnAl-HTµE or ZnAl-HTwg) can beneglected (see Figure S2 in ESI†).

The positive results obtained from the morphological, elec-trical and optical analysis on the HTs/PEDOT:PSS filmsprompted us to test them in BHJ OPVs. In particu-lar, the smooth and regular surfaces observed for theHTs/PEDOT:PSS films with lower HTs concentrations(0.25 and 0.50 wt%) is not expected to induce morphologi-cal changes of the overlying P3HT:PC61BM layer (as ev-idenced by AFM, see Figure S3 in ESI†). In addition, theHTs in-plane preferential orientation is favorable for max-imizing the HTs/PEDOT:PSS film light scattering prop-erties.

3.2. Scattering properties of HTs:PEDOT:PSSfilms

To investigate the scattering effects of theHTs/PEDOT:PSS films, laser in-line transmissionmeasurements (LT, Figure 5a) were performed on dopedand undoped PEDOT:PSS films deposited on glass. Ingeneral, with this method an approximation of the diffusedtransmittance (Tdif ) of a semi-transparent sample is givenby the difference between the film total transmittance(Ttot) and linear transmittance (Tlin). An approximatedhaze parameter (Ht) for transmitted light can thenbe determined as Ht = (Ttot–Tlin)/Ttot = Tdif /Ttot .In this way, the enhanced light scattering propertiesof the ZnAl-HTs/PEDOT:PSS films, compared to thebare PEDOT:PSS film, can be evaluated through thedifference between their Ht and the Ht of the referencePEDOT:PSS film (this ratio can be called as differentialhaze, or ∆Ht). Therefore, the Ht parameters of undopedPEDOT:PSS and HTs/PEDOT:PSS composited filmson glass have been calculated, and the corresponding∆Ht are reported in Figure 5b. As expected, the ∆Ht

values of the glass/ZnAl-HTµE /PEDOT:PSS films atvarious particle concentrations increases with respect

5

Authenticated | g.ruani@bo.ismn.cnr.it author's copyDownload Date | 11/8/13 10:29 AM

M. Bolognesi et al.

Fig 4. AFM images (20x20 µm) of the films on glass substrates: a) pristine PEDOT:PSS (RMS = 5.0 nm); b) ZnAl-HTµE /PEDOT:PSS, 0.50 wt%of HTs vs. PEDOT:PSS (RMS = 5.1 nm); c) ZnAl-HTµE /PEDOT:PSS, 1 wt% of HTs vs. PEDOT:PSS (RMS = 5.6 nm); d) ZnAl-HTµE /PEDOT:PSS, 5 wt% of HTs vs. PEDOT:PSS (RMS = 17 nm); e) ZnAl-HTwg/PEDOT:PSS, 0.50 wt% of HTs vs. PEDOT:PSS (RMS= 5.1 nm); f) ZnAl-HTwg/PEDOT:PSS, 1 wt% of HTs vs. PEDOT:PSS (RMS = 10 nm); g) ZnAl-HTwg/PEDOT:PSS, 5 wt% of HTs vs.PEDOT:PSS (RMS = 30 nm).

Fig 5. a) Scheme of the laser in-line transmission measure-ments; b) ∆Ht parameters (versus glass) calculatedfor the ZnAl-HTs/PEDOT:PSS films (left y axis, filledsquares) compared to the Isc values of the ITO/(ZnAl-HTs/PEDOT:PSS)/P3HT:PC61BM/LiF/Al correspondingdevices (right y axis, empty squares), at different ZnAl-HTsconcentrations (0, 0.25, 0.50, 1 and 5 wt% of ZnAl-HTs vs.PEDOT:PSS). Data relative to ZnAl-HTµE are reported in blueand data relative to ZnAl-HTwg in red.

to the glass/PEDOT:PSS reference film, following thesame trend of the corresponding devices JSC measuredat 1 sun (also reported, for comparison, in Figure 5b).In particular, a maximum ∆Ht is obtained for a particleconcentration of 0.50 wt%, while at higher concentrations(1 and 5 wt%) ∆Ht gradually decreases.

Analogously, the ∆Ht of the glass/ZnAl-HTwg/PEDOT:PSS films increases with respect tothe bare glass/PEDOT:PSS film, reaching a maximumvalue for 0.25 wt% particle concentration and graduallydecreases for higher ZnAl-HTwg loads, again followingthe Isc trend of the corresponding OPV devices. It has tobe noted that the decrease in the ∆Ht parameter for highparticles concentrations, both for ZnAl-HTµE and ZnAl-HTwg, is likely due to the formation of the micrometricaggregates previously observed in the morphologicalanalysis of the highly doped PEDOT:PSS films (SEMand AFM images, Figures 3 and 4, respectively).

These measurements validate our hypothesis, confirmingthat the addition of the ZnAl-HTs to the PEDOT:PSSlayer enhances the scattering properties of the film forthe transmitted light.

3.3. BHJ OPV devices performances

The potential of HTs as effective scattering cen-ters in OPV devices was investigated in bulk-heterojunction (BHJ) solar cells using P3HT:PC61BMas active layer. The device structure employedis glass/ITO/PEDOT:PSS/P3HT:PC61BM/LiF/Al (Fig-ure 1). Details for the device fabrication and character-ization are given in the experimental section. Different

6

Authenticated | g.ruani@bo.ismn.cnr.it author's copyDownload Date | 11/8/13 10:29 AM

Efficiency enhancement of P3HT:PCBM solar cells…

Fig 6. J-V plots of the best performing P3HT:PC61BM based BHJ OPVdevices prepared with doped or undoped PEDOT:PSS films.

HTs loads (0.25 wt%, 0.50 wt%, 1 wt%, 5 wt% HTs vs. PE-DOT:PSS) have been tested.Table 1 summarizes the photovoltaic response data of thecorresponding BHJ OPV devices (maximum and averagevalues calculated over ∼15 devices), including Voc, JSC ,FF, and PCEs. Representative current density-voltage(J-V) plots of the best performing solar cells, measuredunder standard illumination AM1.5G, are shown in Fig-ure 6. The optimized reference solar cell, based on pris-tine pH-neutral PEDOT:PSS (∼50 nm), affords a JSC of7.32 mA/cm2, a Voc of 0.54 V, a FF of 66% and a PCE of2.58%. The greatest OPV performances were achieved forthe devices prepared with 0.50 wt% of ZnAl-HTµE and0.25 wt% of ZnAl-HTwg dispersed in the PEDOT:PSSlayer. In fact, when the ZnAl-HTµE /PEDOT:PSS com-posite film (with HTs 0.50 wt%) is used in the solar cells,the best device exhibits a JSC of 8.21 mA/cm2, a VOC of0.54 V, a FF of 67% and a PCE of 2.98%, which correspondsto a ∼16% PCE improvement with respect to the referencecell (∼12% on average). A similar but less marked trendis observed for ZnAl-HTwg based devices, for which thebest performance is peaked at a particle load of 0.25 wt%,leading to JSC , Voc, FF and PCE for the best device of7.67 mA/cm2, 0.54 V, 66% and 2.73%, respectively, corre-sponding to an improvement of ∼12% with respect to thereference cell (∼6% on average). Noticeably, the mainfactor determining the efficiency improvements for the de-vices with HTs/PEDOT:PSS layers is the enhancement ofthe short-circuit current density, in perfect agreement withthe light scattering properties of the doped buffer layer.By further increasing the HTs content (up to 5 wt%) inthe PEDOT:PSS films, the JSC , FF and PCE significantlydecrease. This could be ascribed to several factors, suchas: i) the presence in relatively high percentage of electri-cally inert HTs, increasing the series resistance within the

Fig 7. External Quantum Efficiency (EQE) spectra of the best per-forming P3HT:PC61BM based BHJ OPV devices prepared withdoped or undoped PEDOT:PSS films.

OPV device; ii) a reduced nanoscale self-organization ofthe overlying photoactive layer (P3HT:PC61BM), inducedby the rough and irregular surface of the highly dopedbuffer layer.It should be noted that within all films, the highest ∆Ht

has been obtained for the ZnAl-HTwg/PEDOT:PSS filmwith 0.25 wt% nanoparticles concentration. However thecorresponding ZnAl-HTwg/PEDOT:PSS based solar celldoesn’t show the highest photocurrent, within all devices.This discrepancy could be explained through the higheraggregation tendency of the former particles comparingto the latter, as discussed before (Figure 3). The possiblepresence of aggregates, embedded in the composite filmbulk or at the ITO/PEDOT:PSS interface, for low particlesloads (25 wt% and 0.50 wt%), cannot be excluded throughSEM or AFM surface analysis (Figures 3b and 3d, Fig-ure 4b and 4e). Therefore it is reasonable to assume thatthe limiting factor for the photocurrent generation of theHTwg/PEDOT:PSS based device is the higher content ofaggregates in the HTwg/PEDOT:PSS film comparing tothe HTµE /PEDOT:PSS one, which could partially limitthe charge extracting ability of the former buffer layercomparing to the latter.A further evidence of the effectiveness of the ZnAl-HTs/PEDOT:PSS films in enhancing the devices pho-tocurrent through light scattering can be given by thestudy of their spectral response. Figure 7 shows the Ex-ternal Quantum Efficiency (EQE) spectra of the best per-forming solar cells with ZnAl-HTs/PEDOT:PSS films, incomparison with the reference cell.The resulting EQE values are gradually enhanced passingfrom the reference device (maximum EQE = 0.56), to thecells based on 0.25 wt% of ZnAl-HTwg (maximum EQE

7

Authenticated | g.ruani@bo.ismn.cnr.it author's copyDownload Date | 11/8/13 10:29 AM

M. Bolognesi et al.

Table 1. Summary of the photovoltaic performances of the P3HT:PC61BM based BHJ OPV devices prepared with different buffer layers: pristinePEDOT:PSS, or PEDOT:PSS doped with different loads of ZnAl-HTwg or ZnAl-HTµE . The reported parameters are the values averagedover ∼15 devices and the maximum values (in brackets).

ZnAl-HTs type ZnAl-HTs/PEDOT:PSS wt%

JSC(mA/cm2)

Voc(V)

FF(%)

PCE(%)

– 0 7.21 (7.32) 0.54 (0.54) 63 (66) 2.47 (2.58)

ZnAl- HTwg

0.25 7.55 (7.67) 0.54 (0.54) 64 (66) 2.63 (2.73)0.50 7.64 0.54 65 2.681 7.52 0.54 62 2.515 5.98 0.56 57 1.93

ZnAl- HTµE

0.25 7.89 0.55 62 2.680.50 8.04 (8.21) 0.54 (0.54) 64 (67) 2.76 (2.98)1 7.17 0.54 65 2.545 6.42 0.54 60 2.09

= 0.58), to the one with 0.50 wt% of ZnAl-HTµE (max-imum EQE = 0.60). Convolution of these EQE spectrawith the AM 1.5 solar spectrum gave calculated shortcircuit current densities in good agreement with thoseobtained from the I-V measurements. Moreover, no sig-nificant alteration in the shape of the EQE spectra ofthe ZnAl-HTs/PEDOT:PSS based devices, with respectto the reference cell, is registered. This further confirmsthat, for low HTs loads, no significant changes are in-duced by the underlying PEDOT:PSS doped layer onthe P3HT:PC61BM morphology (see also Figure S3 inESI†), since they would likely result in absorption andEQE spectral modifications.These results demonstrate the efficacy of theHTs/PEDOT:PSS layer in enhancing the propaga-tion length and trapping of the incident light into theBHJ OPV device, leading to a significant active layerabsorption enhancement and consequent increase of Jsc,EQE and PCE.

4. Conclusion

In conclusion, two series of PEDOT:PSS films dopedwith ZnAl-HTwgor ZnAl-HTµE were employed inP3HT:PC61BM based BHJ OPV devices to investigatetheir scattering properties. The results demonstrate thatZnAl-HTs could effectively scatter a portion of the inci-dent light into larger angles through the active layer of theOPV solar cell, enhancing its light absorption and improv-ing the device JSC and PCE. Indeed, the HTs/PEDOT:PSSfilms with the highest haze for transmitted light (∆Ht)resulted in devices with improved performances, withPCEs going from 2.58%, for the device with pristine PE-DOT:PSS, to 2.73% and to 2.98% for the devices withZnAl-HTwg/PEDOT:PSS and ZnAl-HTµE /PEDOT:PSSfilms, respectively.

The extraordinary versatility of HTs, given by their rela-tively simple synthetic procedures and by the possibilityto combine a wide variety of metals and counter-ions intheir chemical structure, paves the way to the engineer-ing of novel HTs-based composites for applications in theOPVs technology. Indeed, we expect that device perfor-mances can be further improved by a rational and finetuning of the optical, electrical and chemical properties ofthe HTs, taking advantage of the results and considera-tions reported in this study.This approach might allow a potential and effective im-plementation of HTs/PEDOT:PSS composite films in therecombination layer of tandem solar cells. Indeed, the ac-tive layers of both the bottom and top cells could benefitof the scattering effect induced by the doped intermediatelayer [69, 70].In addition, with a similar approach, the use of otherhydrophilic scattering nanoparticles in the PEDOT:PSSlayer of BHJ OPVs, such as nanozeolites, have been ex-plored in cooperation with SAES Getters R&D (in July2012) with promising results, which are nonetheless outof the main focus of this publication.

AcknowledgementThe authors wish to thank Dr. Franco Corticelli (fromCNR-IMM) for the precious technical assistance with theSEM measurements and Mr. Paolo Mei for the precioustechnical contribution to the 4-points probe equipment.This work was partially supported by Programma Opera-tivo FESR 2007-2013 of Regione Emilia-Romagna – At-tività I.1.1., Progetto Premiale CNR 2012 – Produzione dienergia da fonti rinnovabili (Iniziativa CNR per il Mezzo-giorno, L. 191/2009 art. 2 comma 44).

8

Authenticated | g.ruani@bo.ismn.cnr.it author's copyDownload Date | 11/8/13 10:29 AM

Efficiency enhancement of P3HT:PCBM solar cells…

References

[1] C. J. Brabec, S. Gowrisanker, J. J. M. Halls, D. Laird,S. J. Jia, S. P. Williams, Adv. Mater., 22, 3839 (2010).

[2] A. Mishra, P. Baeuerle, Angew. Chem., Int. Ed., 51,2020 (2012).

[3] W. Tracy, available from:www.konarka.com/index.php/site/pressreleasedetail/konarkas_power_plastic_achieves_world_record_-83_efficiency_certification_fr

[4] R. Nigel, available from: www.heliatek.com/wp-content/uploads/2011/12/111205_PI_Heliatek-with-efficiency-record-for-organic-solar-cell_EN.pdf

[5] T. Pearl, available from: www.polyera.com/newsflash/polyera-achieves-worl_d-record-organic-solar-cell-performance

[6] M. A. Green, K. Emery, Y. Hishikawa, W. Warta, E. D.Dunlop, Prog. Photovolt. Res. Appl., 20, 12 (2012).

[7] G. Li, V. Shrotriya, Y. Yao, Y. Yang, J. Appl. Phys., 98,043704 (2005).

[8] K.M. Coakley, M. D. McGehee, Chem. Mater., 16,4533 (2004).

[9] A. Abbotto et al., J. Polym. Sci. A Polym. Chem., 50,2829 (2012).

[10] T. Tromholt, In: F. C. Krebs (Ed.), Stability andDegradation of Organic and Polymer Solar Cells,First Edition, (John Wiley & Sons Ltd, UK, 2012)243.

[11] J. Liu, M. A. G. Namboothiry, D. L. Carroll, Appl. Phys.Lett., 90, 133515 (2007).

[12] B. O’Connor, D. Nothern, K. P. Pipe, M. Shtein, Opt.Express, 18, A432 (2010).

[13] M. O’Neill, S. M. Kelly, Adv. Mater., 23, 566 (2011).[14] M. G. Debije, P. P. C. Verbunt, Adv. Energy Mater., 2,

12 (2012).[15] S. Quici et al., Chem. Mater., 21, 2941 (2009).[16] L. Armelao et al., ChemPhysChem, 11, 2499 (2010).[17] H. K. Raut, V. A. Ganesh, A. S. Nair, S. Ramakrishna,

Energy Environ. Sci., 4, 3779 (2011).[18] I. Kim, D. S. Jeong, T. S. Lee, W. S. Lee, K.-S. Lee, J.

Appl. Phys., 111, 103121 (2012).[19] D. D. S. Fung et al., J. Mater. Chem., 21, 16349 (2011).[20] M.-G. Kang, T. Xu, H. J. Park, X. Luo, L. J. Guo, Adv.

Mater., 22, 4378 (2010).[21] A.Y. Mahmoud, J. M. Zhang, D. L. Ma, R. Izquierdo ,V.

V. Truong, Org. Electron., 13, 3102 (2012).[22] D. H. Wang et al., Angew. Chem., Int. Ed., 50, 5519

(2011).[23] K.-J. Byeon, H. Leea, Eur. Phys. J. Appl. Phys., 59,

10001 (2012).[24] Y. Xu, F. Zhang, X. Feng, Small, 7, 1338 (2011).[25] D.-H. Ko, J. R. Tumbleston, A. Gadisa, M. Aryal, Y.

Liu, R. Lopez, E. T. Samulski, J. Mater. Chem., 21,

16293 (2011).[26] B. Jacobs, G. De Mey, Solid-State Electron., 21, 1191

(1978).[27] S. Hore, P. Nitz, C. Vetter, C. Prahl, M. Niggemann,

R. Kern, Chem. Commun., 15, 2011 (2005).[28] J. Ferber, J. Luther, Sol. Energy Mater. Sol. Cells, 54,

265 (1998).[29] S. Hore, C. Vetter, R. Kern, H. Smit, A. Hinsch, Sol.

Energy Mater. Sol. Cells, 90, 1176 (2006).[30] M. Planells et al., Energy Environ. Sci., 4, 1820

(2011).[31] A. Reynal, E. Palomares, Eur. J. Inorg. Chem., 4509

(2011).[32] S. Guarnera, A. Bonucci, S. Perissinotto, R. Gian-

nantonio, G. Lanzani, A. Petrozza, RSC Adv., 3, 2163(2013).

[33] SAES GETTERS SPA, International PatentPCT/EP2010/067667, 17.11.2010.

[34] L. Chen, P. Wang, F. Li, S.Yu, Y. Chen, Sol. EnergyMater. Sol. Cells, 102, 66 (2012).

[35] C.C. Oey et al., Thin Solid Films, 492, 253 (2005).[36] B. Riedel, J. Hauss, M. Aichholz, A. Gall, U. Lemmer,

M. Gerken, Org. Electron., 11, 1172 (2010).[37] Y.-J. Lin, Y.-C. Su, J. Appl. Phys., 111, 073712 (2012).[38] U. Costantino, F. Montanari, M. Nocchetti, F. Canepa,

A. Frache, J. Mater. Chem., 17, 1079 (2007).[39] F. R. Costa, M. Saphiannikova,U. Wagenknecht, G.

Heinrich, Adv. Polym. Sci., 210, 101 (2008).[40] G. G. Aloisi et al., Mater. Chem. Phys., 123, 372

(2010).[41] U. Costantino et al., ACS Appl. Mater. Interfaces, 1,

668, (2009).[42] D. G. Evans, X. Duan, Chem. Commun., 5, 485 (2006).[43] L. Liu, J. C. Grunlan, Adv. Funct. Mater., 17, 2343

(2007).[44] A. Vaccari, Appl. Clay Sci., 14, 161 (1999).[45] M. Turco et al., J. Catal., 228, 43 (2004).[46] T. Montanari et al., Catal. Today, 152, 104 (2010).[47] G. Busca, U. Costantino, T. Montanari, G. Ramis, C.

Resini, M. Sisani, Int. J. Hydrogen Energy, 35, 5356(2010).

[48] L. Latterini, M. Nocchetti, U. Costantino, G. G. Aloisi,F. Elisei, Inorg. Chim. Acta, 360, 728 (2007).

[49] L. Latterini, M. Nocchetti,G. G. Aloisi, U. Costantino,F. De Schryver, F. Elisei, Langmuir, 23, 12337 (2007).

[50] W. Shi, M. Wei, J. Lu, D. G. Evans, X. J. Duan, J. Phys.Chem. C, 113, 12888 (2009).

[51] T. Posati et al., Cryst. Growth Des., 10, 2847 (2010).[52] T. Posati, F. Costantino, L. Latterini, M. Nocchetti,

M. Paolantoni, L. Tarpani, Inorg. Chem., 51, 13229(2012).

[53] X. Wang et al., J. Mater. Chem. A, DOI:

9

Authenticated | g.ruani@bo.ismn.cnr.it author's copyDownload Date | 11/8/13 10:29 AM

M. Bolognesi et al.

10.1039/c0xx00000x[54] M. Bastianini et al., Inorg. Chem., 51, 2560 (2012).[55] L. Teruel, Y. Bouizi, P. Atienzar, V. Fornes, H. Garcia,

Energy Environ. Sci., 3, 154 (2010).[56] J. H. Lee, J. Chang, J.-H. Cha, D.-Y. Jung, S. S. Kim,

J. M. Kim, Chem. Eur. J., 16, 8296 (2010).[57] U. Costantino, F. Marmottini, M. Nocchetti, R. Vivani,

Eur. J. Inorg. Chem., 1439 (1998).[58] D. O’Hare, G. Hu, J. Am. Chem. Soc., 127, 17808

(2005).[59] D. O’Hare, G. Hu, N. Wang, J. Davis, J. Mater. Chem.,

17, 2257 (2007).[60] F. Bellezza, A. Cipiciani, U. Costantino, M. Nocchetti,

T. Posati, Eur. J. Inorg. Chem., 2603 (2009).[61] F. Bellezza, M. Nocchetti, T. Posati, S. Giovagnoli, A.

Cipiciani, J. Colloid Interface Sci., 20, 376 (2012).[62] L. Li, R. Ma, Y. Ebina, N. Iyi, T. Sasaki, Chem. Mater.,

17, 4386 (2005).[63] R. Apetz, M. P. B. van Bruggen, J. Am. Ceram. Soc.,

86, 480 (2003).[64] C.-H. Tsai et al., Org. Electron., 12, 2003 (2011).[65] J. Krč, M. Zeman, F. Smole, M. Topič, J. Appl. Phys.,

92, 749 (2002).[66] D. Schroder, Semiconductor material and device char-

acterization, 3rd edition (John Wiley & Sons Ltd, NewJersey, 2006).

[67] L. Wang, C. Li, M. Liu, D. G. Evans, X. Duan, Chem.Commun., 2, 123 (2007)

[68] Z. Liu et al., J. Am. Chem. Soc., 128, 4872 (2006).[69] G. Dennler et al., Adv. Mater., 20, 579 (2008).[70] J. Yang et al., ACS Nano, 5, 6210 (2011).[71] S. Sista, Z. Hong, L.-M. Chen, Y. Yang, Energy Env-

iron. Sci., 4, 1606 (2011).

10

Authenticated | g.ruani@bo.ismn.cnr.it author's copyDownload Date | 11/8/13 10:29 AM