MEH-PPV and PCBM Solution Concentration Dependence of Inverted-Type Organic Solar Cells Based on Eosin-Y-Coated ZnO Nanorod Arrays

e in�uence of polymer solution concentration on the performance of chlorobenzene(CB-) and chloroform(CF-) based inverted-type organic solar cells has been investigated. e organic photoactive layers consisted of poly(2-methoxy-5-(2-ethyl hexyloxy)-1,4-phenylenevinylene) (MEH-PPV) and (6,6)-phenyl C61 butyric acid methyl ester (PCBM) were spin coated from CF with concentrations of 4, 6, and 8mg/mL and from CB with concentrations of 6, 8, and 10mg/mL onto Eosin-Y-coated ZnO nanorod arrays (NRAs). Fluorine doped tin oxide (FTO) and silver (Ag) were used as electron collecting electrode and hole collecting electrode, respectively. Experimental results showed that the short circuit current density and power conversion efficiency increased with decrease of solution concentration for both CB and CF devices, which could be attributed to reducing charge recombination in thinner photoactive layer and larger contact area between the rougher photoactive layer andAg contact. However, the open circuit voltage decreased with decreasing solution concentration due to increase of leakage current from ZnO NRAs to Ag as the ZnO NRAs were not fully covered by the polymer blend.e highest power conversion efficiencies of 0.54 ± 0.10% and 0.87 ± 0.15% were achieved at the respective lowest solution concentrations of CB and CF.


Introduction
Rapid development in enhancing the performance of organic solar cells (OSCs) for the past few years by utilizing bulk heterojunction (BHJ) devices has enabled power conversion efficiency (PCE) up to 10.6% to be achieved [1].In the case of BHJ conventional device structure, an active layer which consists of polymer donor and acceptor material is normally sandwiched between indium tin oxide (ITO) anode and aluminium (Al) cathode [2].Poly (3,4-ethylenedioxythiophene): poly(styrenesulfonate) or PEDOT:PSS which plays role as conductive hole transport layer (HTL) is commonly introduced in between organic photoactive layer and ITO anode [3].However, due to strong acidic nature of PEDOT:PSS, instability between the interface of ITO/PEDOT:PSS could create defect site and thus leading to deteriorated device performance [4,5].Moreover, oxidation of low work function Al electrode (4.28 eV) in air could also lead to instability at the interface between metal and organic photoactive layer [6].In order to solve these issues, inverted con�guration with low work function metal or metal oxides materials as electron transporting layer and high work function metal (e.g., silver, gold) as hole collection electrode have been introduced to improve the stability of the device [6,7].Lately, a lot of attention towards inverted OSCs with various metal oxide nanostructures as electron transporting layer have been developed, such as ZnO [8][9][10][11][12][13][14], TiO 2 [15][16][17][18][19], alkali-metal oxide cesium carbonate (Cs 2 CO 3 ) [20,21], and others [22,23].e environmental-friendly and low-cost ZnO nanorods arrays (NRAs) can be grown at low temperature (below 100 ∘ C) by hydrothermal method [24].e ZnO NRAs not only have good electron collecting and transporting property in the inverted OSCs hybridized with the polymer blend, but also can provide larger interface surface area to improve photocurrent and efficiency [25].
Recently, much effort has been focused on utilizing blend �lm of poly(3-hexylthiophene) (P3HT) and (6,6)-phenyl C 61 butyric acid methyl ester (PCBM) as photoactive layer in ZnO NRAs-based inverted OSCs [19][20][21][22][23]. Various P3HT-PCBM processing parameters such as solvent selection, solution concentration, and spin coating speed have been investigated to optimize the solar cell performance [25,26].Takanezawa et al. [25] reported that thicker organic photoactive layer deposited by using higher solution concentration or lower spin coating speed results in better power conversion efficiency as a consequence of improved optical absorption and short circuit current density.Meanwhile, Chou et al. [26] demonstrated that organic photoactive layer drying time could be lengthened by lowering the spin coating speed, hence leading to improved photovoltaic performance as a result of better polymer crystallinity, increased thickness, and enhanced in�ltration of photoactive layer.However, little attention has been given to poly[2-methoxy-5-(2 ′ethylhexyloxy)-p-phenylenevinylene] (MEH-PPV) blended with PCBM as photoactive layer in the similar OSCs.MEH-PPV has broad optical absorption region 400-600 nm and also higher hole mobility compared to electron mobility [27].
e present work reports on the effects of MEH-PPV and PCBM solution concentration on the performance of inverted-type OSCs based on Eosin-Y-coated ZnO NRAs.Eosin-Y organic dye, a low-cost alternative to expensive Rucomplexes dye, has good absorption property and is able to increase the wettability of ZnO surface [28].Different solution concentrations of MEH-PPV and PCBM were prepared from chlorobenzene (CB) and chloroform (CF).e short circuit current density and power conversion efficiency increased, whereas the open circuit voltage decreased with the reduction of solution concentration.e optimum power conversion efficiencies of 0.54 ± 0.10% and 0.87 ± 0.15% were achieved at the respective lowest solution concentration of CB and CF devices.

Device Fabrication.
Inverted OSCs with �uorine doped tin oxide (FTO)/Eosin-Y-coated ZnO NRAs/MEH-PPV and PCBM(1:2)/Ag structure were constructed.e FTO (∼15 Ω/sq) substrates were ultrasonically cleaned by using acetone and 2-propanol each for 15 min.e ZnO seed layer was spin coated on top of precleaned substrates for 3 times from an equimolar solution (0.2 M) of zinc acetate dehydrate (Zn(O 2 CCH 3 ) 2 ⋅(H 2 O) 2 ) and diethanolamine (DEA) in ethanol followed by annealing at 300 ∘ C for one hour in air.e hydrothermal growth of ZnO NRA was done by immersing ZnO seed coated substrates into an equimolar aqueous solution (40 mM) of zinc nitrate hexahydrate (Zn(NO 3 ) 2 ⋅6H 2 O) and hexamethylenetetramine (HMT) at 90 ∘ C for 45 min in a laboratory oven.e dye coating was performed by spin coating a thin layer of Eosin-Y from a 0.3 mM acetone solution onto ZnO NRAs.e MEH-PPV with molecular weight 40,000-70,000 g mol −1 and PCBM were purchased from Sigma-Aldrich and Nano-C, respectively.MEH-PPV and PCBM (1:2 by weight) were dissolved in two different solvents, namely chloroform (CF) and chlorobenzene (CB) at various concentrations.CF solutions with concentration of 4 mg/mL to 8 mg/mL and CB solutions with concentration of 6 mg/mL to 10 mg/mL were spincoated on top of Eosin-Y-coated ZnO NRAs.e spinning speed and spinning time were �xed at 1000 rpm and 60 s, respectively.Finally, the silver (Ag) contact (200 nm) was deposited by using magnetron sputtering under vacuum (∼ 10 −3 Torr) by �owing Argon (Ar) gas at 45 cc/min.During sputtering process, a slow deposition rate ∼0.17 nm s −1 was used in order to avoid the damage of organic photoactive layer.Furthermore, in order to oxidize silver, pure oxygen gas was introduced during sputtering process into the vacuum chamber at a �ow rate of 5 cc/min.It was believed that oxidation of Ag will increase its work function from 4.3 to 5.0 eV, which in turn facilitate hole collection from donor materials to Ag and improve the cells performance [29].Four devices for each solution concentration were fabricated.

Device
Characterization.e absorption spectra of blend �lms consisted of organic photoactive layer and Eosin-Ycoated ZnO NRAs were investigated using a Perkin Elmer Lambda 900 UV-Vis spectrophotometer.e cross-sectional images of organic photoactive layer deposited on Eosin-Y-coated ZnO NRAs were recorded by using Carl Zeiss Supra 55VP Field Emission Scanning Electron Microscope (FESEM).e surface morphology of the organic photoactive layers was investigated by using Ntegra Prima NT-MDT scanning probe microscope under ambient condition at %RH ∼68% and room temperature 26 ∘ C. e - characteristics of the devices were measured with Keithley 237 SMU under illumination at 100 mW/cm 2 from a solar simulator with AM 1.5 � �lter.e active area of the devices was de�ned as 0.07 cm 2 by using mask.e incident photon to current conversion efficiency (IPCE) values was measured using Newport IPCE system at a chopping frequency of 10 Hz. e above device fabrication and measurements were conducted under ambient atmosphere (without encapsulation).

Results and Discussion
Figure 1 shows the optical absorption properties of blend �lms of organic photoactive layer and Eosin-Y-coated ZnO NRAs.Broad absorption from 400 nm up to 600 nm is correlated with MEH-PPV [30], whereas absorption below 400 nm is contributed by PCBM and ZnO NRAs [31].e absorption increased with solution concentration for both CF and CB.It has been reported that enhanced absorption in the long wavelength range of the polymer indicates the increase in crystallinity and hole mobility of the polymer [32].However, the shape of absorption spectra of MEH-PPV and PCBM blend �lm (in the range of 400-600 nm) showed no notable changes with different solvents and solution concentration, suggesting no signi�cant different in crystallinity and hole mobility of the polymer.e Eosin-Y-coated ZnO NRAs exhibited absorption in visible region from 440 nm to 560 nm (inset of Figure 1) and two absorption peaks at 500 nm and 530 nm corresponding to the monomeric and dimeric forms of Eosin-Y were observed [33].e intensity of Eosin-Y absorption was much lower than that of MEH-PPV, suggesting Eosin-Y had a little contribution to the overall absorption of device.However, the existence of thin layer dye on top of ZnO NRAs may enhance the charge transfer efficiency at the MEH-PPV/ZnO NRAs interface and also improve the wetting state of ZnO NRAs [34].
Figure 2 shows typical cross-sectional images of devices prepared from CB and CF at different concentrations.ZnO nanorods with diameter ranging from 50 nm to 80 nm and with length up to 300 nm were obtained.e total thickness of the blend �lms prepared from CB and CF reduced from ∼560 nm to ∼460 nm and from ∼500 nm and ∼400 nm, respectively, with decrease of solution concentration.is is consistent with absorption result shown above where the absorption increased with increasing solution concentration, indicating formation of thicker blend �lm.In addition, it can be clearly seen that the in�ltration of organic blend prepared from CB into the spacing between Eosin-Y-coated ZnO nanorods was better than that prepared from CF, as indicated by the increased number of voids appeared in the interspaces between ZnO nanorods.e polymer prepared from a higher boiling point solvent has longer time to effectively in�ltrate into the spacing between ZnO nanorods and gives enough time for polymer chain to self-organized [35].erefore, the poor in�ltration could be attributed to lower boiling point of CF (∼61 ∘ C) as compared to CB (∼131 ∘ C). e - characteristics of typical devices under illumination of a simulated AM 1.5 sunlight at 100 mW/cm 2 are shown in Figure 3(a), whereas the statistical data for photovoltaic parameters obtained from a series of four individual devices for each parameter are plotted in Figure 3(b).e short-circuit current density ( sc ), �ll factor (FF) and power conversion efficiency (PCE) increased with decreasing solution concentration for both CB-and CF-based devices, whereas the open circuit voltage ( oc ) exhibited opposite behavior.Both CB-and CF-based devices showed the highest PCE at respective lowest solution concentration.e device prepared from CB at concentration of 6 mg/mL exhibited the highest PCE of 0.54 ± 0.10% with  sc of 3.45 ± 0.17 mA cm −2 ,  oc of 0.34 ± 0.02 V and FF of 0.48 ± 0.02.In the case of CF, the highest PCE of 0.87 ± 0.15% was achieved at the concentration of 4 mg/mL, with  sc of 4.76 ± 0.17 mA cm −2 ,  oc of 0.35 ± 0.03 V, and FF of 0.51 ± 0.02.However, further reduction of polymer solution concentration results in very low  oc < 0.2 V and low  sc < 0.75 mA cm −2 for both CB and CF devices (not shown here).e series resistance (  ) and shunt resistance ( sh ) were determined from the slopes at  oc and  sc in the - curve, respectively.e enhancement of  sc with decreasing solution concentration could be correlated with the organic photoactive layer thickness.It is generally accepted that increase in thickness results in more excitons due to higher absorption.However, the charge recombination rate also increases with thickness, resulting in lower  sc .
In contrast to study obtained on P3HT and PCBM-based inverted-type organic solar cell [25], increase in thickness did not lead to increase in  sc , but to a decrement in present work.is suggests that charge recombination process dominated the variation of  sc with respect to thickness.It is believed that the probability of photogenerated electrons and holes to transport to ZnO NRAs and Ag, respectively, became higher with thinner photoactive layer.e   decreased signi�cantly from 397.24 ± 30.87 to 25.40 ± 6.91 Ω cm 2 and from 220.26 ± 11.53 to 15.97 ± 2.47 Ω cm 2 as the polymer solution concentration reduced for CB and CF devices, respectively.is result indicates the decrease of bulk resistance across the organic photoactive layer with decreasing thickness, leading to higher number of electron and hole extracted from the device.Besides, the increase of  oc with the solution concentration for both solvents could be due to suppression of leakage current as indicated by larger  sh .e signi�cant decrement of   with respect to  sh leads to increment in FF. erefore, the increase in both  sc and FF was great enough to offset the reduced  oc , which in turn improved the overall device performance.e overall photovoltaic performance of all devices is summarized in Table 1.
Figure 4 shows IPCE spectra of the devices measured from wavelength 350 to 800 nm with maximum IPCE of ∼16% and ∼20% at 500 nm was achieved at respective lowest solution concentration for CB and CF device.Despite the increase of absorption with solution concentration, the opposite trend was observed from IPCE spectra.e IPCE result is consistent with  sc which increased with decreasing solution concentration and further con�rms that relatively thick organic photoactive layer on top of Eosin-Y-coated ZnO NRAs causes the increase of charge recombination in the organic photoactive layer and results in lower  sc even though the absorption increased.e IPCE spectra of the CB and CF devices with the highest solution concentration exhibited low photon conversion efficiency at wavelength (500 nm) corresponding to absorption peak of MEH-PPV.For thicker photoactive layer deposited at high solution concentration, most of the photons with wavelength close to 500 nm were absorbed in the region close to FTO/Eosin-Y-coated ZnO NRAs.Meanwhile, photons with wavelength away from 500 nm were more uniformly absorbed throughout the photoactive layer.e distortion of IPCE spectrum of thick photoactive layer indicates that photocurrent generation in the bulk region close to Ag was more efficient than that in the region close to ZnO NRAs [36].is result also suggests that the probability of holes generated in the region close to ZnO NRAs to be trapped or recombine before reaching Ag was higher.
Figure 5 shows the topographic images of surface morphology of organic photoactive layer deposited on Eosin-Y-coated ZnO NRAs with different solvent concentrations.e root mean s�uare (rms) roughness signi�cantly increased from 2.58 to 14.75 nm and from 4.11 to 17.77 nm, respectively, as the CB and CF solution concentration decreased.e rough surface obtained from lower solution concentration indicates ZnO NRAs were not fully covered by polymer  blend.�is �nding is in agreement with the value of  sh which increased with solution concentration.Besides, this observation is also well in line with the cross-sectional FESEM images from which a wavy surface re�ecting the ZnO NRAs underneath was clearly observed when organic photoactive layer was deposited from a low solution concentration.Meanwhile, the rougher surface also provides larger contact area between organic photoactive layer and metal contact, and thus leading to better charge collection efficiency [37,38].Even though better photovoltaic performance can be achieved by depositing a thinner organic photoactive �lm, the �lm must be thick enough to prevent the direct leakage current from ZnO NRAs to Ag electrode.

Conclusions
e effect of solution concentration of MEH-PPV and PCBM deposited onto Eosin-Y-coated ZnO NRAs on the performance of inverted-type organic solar cell has been investigated.e short circuit current density and power conversion efficiency greatly enhanced decrease of solution concentration for both chlorobenzene-and chloroformbased devices, mainly due to reduce of charge recombination in thinner organic layer and larger contact area between the rougher organic photoactive layer Ag contact.In increase of leakage current from ZnO NRAs to Ag at lower solution concentration results in decrease of open circuit voltage.e highest power conversion efficiencies of 0.54 ± 0.10% and 0.87 ± 0.15% were achieved at the respective lowest solution concentration of CB and CF.It has been shown that the solution concentration played important role in improving performance of invertedtype organic solar cell.

F 1 :F 2 :
e absorption spectra of blend �lms of Eosin-Y-coated ZnO NRAs and organic photoactive layer deposited from (a) CB and (b) CF solvents at various concentrations.Inset shows the absorption spectra of Eosin-Y-coated ZnO NRAs in the region 440-560 nm.Cross-sectional FESEM images of devices prepared from CB at concentrations (a) 10 mg/mL (b) 6 mg/mL and prepared from CF at concentrations (c) 8 mg/mL, (d) 4 mg/mL (scale bar: 100 nm).

2 )F 3 :
(a) - characteristics of devices prepared from different CB and CF solution concentrations, (b) photovoltaic parameters of devices as a function of solution concentration.

F 5 :
AFM images of blend �lms of Eosin-�-coated ZnO NRAs and organic photoactive layer prepared from CB at concentrations (a) 10 mg/mL, (b) 6 mg/mL and prepared from CF at concentrations (c) 8 mg/mL, (d) 4 mg/mL.T 1: Summary of photovoltaic parameters of devices at various solution concentrations.