- Email: [email protected]
Natural drying effect on active layer for achieving high performance in polymer solar cells
Renewable Energy 74 (2015) 11e17
Contents lists available at ScienceDirect
Renewable Energy journal homepage: www.elsevier.com/locate/renene
Natural drying effect on active layer for achieving high performance in polymer solar cells Ziyang Hu a, *, Jianjun Zhang b, *, Like Huang a, Jingyang Sun a, Ting Zhang a, Hongyun He a, Jing Zhang a, Houcheng Zhang a, Yuejin Zhu a a b
Department of Microelectronic Science and Engineering, Ningbo University, Zhejiang 315211, China College of Electronic Information and Optical Engineering, Nankai University, Tianjin 300071, China
a r t i c l e i n f o
a b s t r a c t
Article history: Received 30 June 2013 Accepted 18 July 2014 Available online
The self-organization of the polymer in solar cells based on regioregular poly(3-hexylthiophene)(RRP3HT):[6, 6]-phenyl C61-butyric acid methyl ester is studied systematically as a function of the room temperature (RT) (varied from 300 K to 290 K). Optimal self-organized structures within the RRP3HT:PCBM films are achieved by varying spin speed and time as well as the temperature at which the spin casting process occurs. These blend films are characterized by UVevis absorption spectroscopy, atomic force microscopy, and X-ray diffraction measurements. The optimum device efficiency can be achieved in naturally dried devices when spin coating within the temperature range 292e294 K. Both the power conversion efficiency (PCE) and fill factor (FF) of the optimum devices show a plateau region, with PCEs exceed 4% and FFs close to 0.70. For RT < 290 K, the corresponding devices show a wide distribution of performance parameters for the unhomogeneous active layer. While for RT > 296 K, the short current density, FF and PCE of the corresponding devices are gradually decreased, suggesting that there is a major change in the ordered structure of the polymer. Based on the results, it is demonstrated that high performance device can be achieved just by natural drying the active layer at RT condition in air condition without further thermal treatments. © 2014 Elsevier Ltd. All rights reserved.
Keywords: Natural drying Polymer solar cell P3HT:PCBM High performance Self-organization
1. Introduction Bulk heterojunction (BHJ) polymer solar cells (PSCs), whose active layer is comprised of semiconducting polymer and fullerene, afford the potential advantages of easy processability, low material costs and mechanical flexibility [1e6]. Generally, the performance of BHJ PSCs can be maximized by controlling the active layer morphology, because efficient photoinduced charge generation, transport, and collection at each electrode chiefly depend on the nanoscale morphology of the active layer [2,3]. Charge generation takes place at the donor/acceptor heterojunction, so the donor and acceptor need to be well-mixed so as to maximize the interface area of the heterojunction. On the other hand, charge transport requires continuous donor and acceptor domains throughout the active layer, so that electrons and holes can flow smoothly to their respective electrodes. Particularly, the transport of holes is often
* Corresponding authors. Department of Microelectronic Science and Engineering, Ningbo University, Zhejiang 315211, China. E-mail addresses: [email protected], [email protected] (Z. Hu), [email protected] (J. Zhang). http://dx.doi.org/10.1016/j.renene.2014.07.034 0960-1481/© 2014 Elsevier Ltd. All rights reserved.
the factor that limits the obtainable photocurrent, as the hole mobility of most used donors is lower than the electron mobility of commonly used acceptors [4e6]. Therefore, the optimal morphology of the active layer has been recognized as one that contains large donor/acceptor interface area, continuous donor and acceptor domains, and high molecular ordering or crystallinity of the donor. Over the last decade, many research groups have systematically analyzed external treatments in order to optimize the BHJ morphology of regioregular poly(3-hexylthiophene) (RR-P3HT):[6, 6]-phenyl C61-butyric acid methyl ester (PCBM) films for highefficiency solar cells. These treatments enable both components to evolve self-organized crystallization and achieve a well-ordered interpenetrating network of individual materials. The most notable of these treatments involve: (1) post thermal annealing (TA) of the active layer to activate rearrangement of the morphology [7e12]; (2) slowing down the drying rate (solvent annealing, SA) to allow the active layer to a desirable morphology [13e18]; and (3) the use of cosolvent and additive to manipulate the active layer morphology [19e22]. Although impressive efficiencies are achieved with these morphology-controlling techniques, the optimal
12
Z. Hu et al. / Renewable Energy 74 (2015) 11e17
morphology can only be attained with careful control of the processing conditions. Consequently, implementations of identical processing conditions have often resulted in wide-ranging device performance values and time-consumed. Furthermore, all highestperformance devices have been fabricated and characterized entirely in inert atmospheres [7e11,13e19], suggesting the importance of device processing in an oxygen-free environment. However, it is difficult for most laboratories to maintain the oxygen free environment in the whole fabricating process. On the other hand, device fabrication in air condition may decrease the cost for mass production of organic solar cells. In our previous report [23], high-performance and air-processed polymer solar cells were achieved by room temperature drying of the active layer. In this paper, we present a systematic study of the natural drying effect of the BHJ active layer afforded by spin-coating speed and time at room temperature (RT). Similar to TA and SA approach, we use the term “natural drying (NA)” to describe this process here. The RR-P3HT:PCBM active layer was investigated by NA treatment combined with the variable spin-coating speed (Ss) and time (Ts) from 300 K to 290 K. By systematically varying Ss and Ts at different RT, we have conducted a comprehensive investigation of the evolved morphology of the RR-P3HT:PCBM films and the performance of the PSCs. The solvent evaporation rate was controlled by RT, Ss and Ts. The spin coating conditions at different RT were then optimized in terms of the cell efficiency. Post SA was applied in conjunction with the NA process to further manipulate the morphology. The resultant morphology was analyzed with UVevis spectroscopy, X-ray diffraction (XRD), and atomic force microscopy (AFM), which was in turn correlated to the device characteristics. 2. Experimental details
drying at RT after spin coating. The Ss and Ts of the spin coating process were alternatively changed to optimize the device performance at a given temperature. As an example, spin coating for 10 s at 300 rpm and spin coating for 12 s at 230 rpm may result in the same device performance for natural drying at the same RT. For solvent annealing, the spin coated films (either liquid or solid, depending on Ss, Ts and RT) were then left in a grease-sealed cylinder (volume: 10 cm3) at the same RT. At the same time, small amounts of chlorobenzene solvent were added into the cylinder with a lid in order to prevent rapid evaporation of the solvent. 2.3. Device characterizations The P3HT:PCBM composite films used for optical and morphological characterization were prepared using the same method as used for the device fabrication. The samples for absorption measurements were prepared by the same procedure before the cathode deposition step, and then measured using a Varian Cary 5000 UVevis spectrophotometer. AFM images were obtained with a Digital Instruments Multimode scanning probe microscope operated in the tapping mode. All the films with the same thickness as measured by a Dektek profilometer, were prepared for measurement. The current-densityevoltage (JeV) characteristics of the devices were measured with a Keithley 2400 source meter under simulated AM1.5G irradiation (100 mW/cm2) in air condition. For JeV measurements, the light intensity was calibrated using a reference silicon solar cell. The hole mobility of the active layer was obtained from the dark JeV characteristics of hole-only devices operated in the space-charge limited current (SCLC) regime, where the devices were obtained by replacing the LiF/Al cathode of the regular devices with thermally evaporated MoO3/Ag.
2.1. Device fabrication
3. Results and discussion
The layer structure of the solar cells was glass/indium tin oxide (ITO)/poly(3,4-ethylenedioxythiophene):poly(styrene sulfonate) (PEDOT:PSS)/P3HT:PCBM/LiF/Al. PEDOT:PSS (Baytron P VP AI 4083), regioregular (90e93%) P3HT (MW ¼ 68000 g/mol, purity: 99%), and PCBM were purchased from H.C. Starck, Rieke Metals, Inc., and Nano C, Inc., respectively. All materials for the fabrication of the polymer solar cells were used as received. Before device fabrication, the ITO coated glass substrates were cleaned by sequential ultrasonic treatment in detergent, deionized water, acetone, and isopropyl alcohol. A thin layer (~40 nm) of PEDOT:PSS was spin-coated to modify the ITO surface and then the substrates were baked at 403 K for 20 min. The blend solutions of P3HT and PCBM (weight ratio of 1:0.8, 20 mg/ml in chlorobenzene) were spin-coated onto PEDOT:PSS-coated ITO glass substrates in air condition. LiF (0.8 nm)/Al (100 nm) electrodes were deposited by thermal evaporation under high vacuum with shadow masks determining the photoactive areas (6 mm2).
The experiments have been designed based on several considerations. Firstly, both fast drying and slow drying processes transform the polymer solution into a solid film. To obtain optimized performance, TA is commonly performed after fast drying the film, which requires the inert atmosphere [7e11]. NA is free of thermal treatment after fast drying the film, so the detrimental reactions can be avoided during post treatment. Therefore, the high performance can be expected to achieve in air condition. Secondly, the P3HT chains would induce the self-organization or crystallization by slow drying the film, which results in the improved performance [13,14,18]. The evaporation rate of the solvent is subject to the influence of temperature, the vapor pressure, the surface to volume ratio, the flow rate of air flowing through the surface. Therefore, it is realizable that the P3HT self-organization is performed by controlling RT, the selected solvent, Ss and Ts. Thirdly, chlorobenzene (CB) has been deliberately chosen as the solvent because its moderate boiling point allows a wide window of spin-coating process at RT condition. By systematically varying Ss and Ts at different RT, we have conducted a comprehensive investigation of the evolution of the RR-P3HT:PCBM films and the device performance. At elevated temperature, the solvent will be removed more rapidly and the film will lose its crystallinity. Here, the representative experimental conditions were chosen and the corresponding data are presented. RT was changed from 300 K to 290 K. The four combinations of Ss and Ts ((800 rpm, 30 s), (500 rpm, 20 s), (300 rpm, 10 s) and (230 rpm, 12 s)), which were named by #1, #2, #3, and #4, respectively, are selected. At the same time, The SA treatment was also performed for comparison. The SA process was described in the experimental section. RT was first fixed at 298 K, and conducted a comprehensive investigation of the evolution of
2.2. Natural drying Natural drying treatment of the P3HT:PCBM composite films were carried out before the deposition of the LiF/Al electrodes. The P3HT:PCBM solution was stirred on the heating plate for ~24 h at 283 K. In order to achieve the uniform films, prior to spinning the substrate was static and coated with the P3HT:PCBM solution completely. Immediately, the active layer was obtained by spincoating the solution in a single spinning process with a fixed spinning speed. This process was performed at different RT from 300 K to 290 K, as measured by a hygrothermograph (Testo, Ger.) in the open lab. The resulting P3HT:PCBM film was then natural
Z. Hu et al. / Renewable Energy 74 (2015) 11e17
the RR-P3HT:PCBM films and the device performance. The same trend is expected if RT is increased. At 298 K with the four combinations of Ss and Ts, all the device performance showed similar results when the active layers were treated with SA and NA processes, respectively. Even Ss and Ts were alternated, the performance parameters were almost unchanged and poor (PCE < 2%, FF < 0.50, Jsc < 6 mA/cm2, Voc ¼ 0.77~0.67 V). Since RT is too high, CB solvent evaporates quickly during the spin coating and also after the spin coating. The self-organization of the polymer P3HT must be realized in the presence of certain solvent. So SA and NA are useless for improved device performance. However, when the active layer was TA in air at 353 K for 15 min, the device performance was greatly improved to be 3%. The annealing condition is contrast to that of the majority of the literatures in the protected atmosphere (423 K, 30 min). It is probably that the polymer may react with water and oxygen in air condition, thereby reducing performance due to heating at high temperature for a long time. At 298 K, slowing the evaporation rate by adjusting the spin coating conditions the device performance increased slightly. The PCEs of the NA devices increased from 1.16% to 1.53%, and the PCEs of the SA devices increased from 1.28% to 1.70%, respectively. Since Ss and Ts are reduced, the solvent evaporation time is prolonged. Therefore, the P3HT chains can realize the self-organized during the insufficient time. It can be imaged that the complete realization of the P3HT self-organization is difficult over 298 K. Based on the above discussion, we can expect that the improved device performance can be realized by gradually reduced RT for the sequence experiments. Aforementioned experiments also show that the reduced Ss and Ts extend the solvent evaporation time, which results in the well realization of the self-organization. Thus, we carried out a series of experiments at 296 K by alternating Ss and Ts. At 296 K, at a high Ss and long Ts, the device performance still showed similar results when the active layers were treated with SA and NA processes, respectively. However, at a low Ss and short Ts, the SA device performance has been significantly improved, while the improvement of the NA device is very slightly. At the same Ss and Ts, the SA device showed reduced Voc, increased
13
FF and Jsc, compared to the NA devices as shown in Fig. 1. The reduced RT make solvent evaporation slows down, so that the time of realizing the self-organization of P3HT chains was increased. Especially, the evaporation of the residual solvent in a sealed glass dish would further slow, which suggests the performance of the SA device is better than that of the NA device. At 296 K, along with the reduced Ss and shortened Ts, the device performance presented significantly increased trend. The SA device with a Voc of 0.64 V, an FF of 65%, a Jsc of 9.01 mA/cm2, and a PCE of 3.05% was obtained. The performance is better than that of the TA device. Under the same #3 condition, the SA device showed reduced Voc, increased FF and Jsc, compared to the NA devices. Since RT is lowered, the total time of the solvent evaporation continues to prolong, so the P3HT chains staying in the solvent have a longer time to realize the self-organized. The improvement of the NA is not obvious, while the improvement of the SA device performance is prominent. It can be concluded that the P3HT chains realized the self-organization even after the spin coating process. The above experiments also demonstrate that RT plays a main role on controlling the solvent evaporation rate, and the Ss and Ts play an auxiliary or secondary role. It is suggested that the device performance would be further improved by lowering RT, reducing Ss and shorting Ts. Therefore, we further conducted the following experiments at 294 K in order to find the best experimental conditions. At 294 K, even at a high Ss and long Ts, both the use of NA and SA process has a significant impact on the device performance. At a low Ss and short Ts, the SA device performance has a significant improvement over the TA device performance. Under the #3 condition, the SA device performance parameters are listed as follows: Voc ¼ 0.63 V, FF ¼ 0.67, Jsc ¼ 9.58 mA/cm2, PCE ¼ 4.04%. The NA device performance parameters are listed as follows: Voc ¼ 0.65 V, FF ¼ 0.63, Jsc ¼ 8.05 mA/cm2, PCE ¼ 3.29%. The NA device performance also exceeds that of the TA device. Since the RT keeps reducing, the self-organization of the P3HT chains continues to increase. The solvent evaporation rate of NA slowed down with the decreased RT, but still faster than that of SA process. At 294 K, with the lowered Ss and shortened Ts, the overall performance continues to maintain an increasing trend. The device
Fig. 1. Performance parameters as a function of RT (from 298 K to 292 K) combined with the alternating Ss and Ts.
14
Z. Hu et al. / Renewable Energy 74 (2015) 11e17
Voc continues to reduce, FF continues to rise, Jsc continues to increase, and PCE continues to improve as shown in Fig. 2. However, the quality of the active layer is also greatly affected by the reduced RT, which results in a wide distribution of the performance parameters. So we decided to reduce RT to 292 K, but appropriately increased Ts to obtain a uniform film. At 292 K, even at a high Ss and long Ts, both the use of NA and SA process has a significant impact on the device performance. At a low Ss and short Ts, the SA and NA device performance is almost the same. The Jsc, FF and PCE of the SA device increased obviously to 10.78 mA/cm2, 0.68 and 4.62%, respectively. Under the #4 condition, the best NA device performance parameters are listed as follows: Voc ¼ 0.60 V, FF ¼ 0.72, Jsc ¼ 11.07 mA/cm2, PCE ¼ 4.78%. Here, one of the best device performances was obtained in air condition at RT without any annealing [12,17,18]. As Ss and Ts are reduced, the NA device performance continues to increase quickly, while the SA device performance shows slight improvement. Since RT is reduced, the CB solvent evaporation rate continues to slow down, the solvent staying in film continue to increase, so that the P3HT chains have enough time to realize self-organized. At 292 K, the solvent evaporation time of the NA process is close to that of the SA process, so the NA device showed similar performance to the SA device. According to the above rule and conclusions, we conducted the series of experiments at 291 K, the experimental results are similar to that conducted at 292 K. But the film-forming property of the active layer starts to deteriorate. When RT is reduced to 290 K, coupled with the low Ss and short Ts, the film uniformity can not be guaranteed, and the surface of the film appears obvious pointshaped or flower-shaped spots. The distribution of performance parameters was very wide as shown in Fig. 2. By adjusting RT, Ss and Ts to control the solvent evaporation rate, the function of the self-organization of the active layer were well realized. For RT > 298 K, it is difficult to adjust Ss and Ts to control the solvent evaporation. When RT was lowered to 294 K, the use of SA process can achieve good control of the solvent evaporation rate; thereby the device performance is significantly improved.
When RT is reduced to 292 K, the use of NA process also can achieve very good control the rate to achieve the well self-organization, thereby obtaining the optimized performance. Decreasing with RT, although the self-organization of the active layer can be realized, the uniformity of the active layer is difficult to obtain, which results in a wide distribution of performance parameters. So we think that in the vicinity of 292 K, NA process is the best choice for high device performance, while the SA process has a wide technical window. From a macroscopic physical view, we discussed the three factors how to control the solvent evaporation rate, thereby affecting the self-organization of the active layer, which leads to the different device performance. We believe that the improved device performance originates from three important contributions, enhanced absorption in the active layer, increased charge-carrier mobility and improved the contact between the active layer and the electrode interface. In order to understand the change in the absorption spectra of the active layer in different process conditions, we have tested two groups of absorption spectrum. Fig. 3(a) and (b) are the absorption spectra of the active layer using NA and SA treatment at 298 K and 294 K, respectively. At 298 K, the absorption spectrum of the active layer is almost the same even the Ss and Ts were alternated. Because the solvent evaporation is too fast, the SA process is difficult to maintain the slow drying of the active layer, which is also consistent with previous analysis. However, at 294 K, the absorption spectra of the active layer vary greatly, especially the use of the SA process. In all the cases, the vibronic features are clearly observed, indicating that both SA and NA process induce the ordering of RRP3HT in the blend film. Compared to Fig. 3(a), three distinct electronic vibration absorption peaks corresponding exactly to pep* band gap transition are presented in Fig. 3(b) [13,15]. At 294 K, three peaks of the SA blend films are the most obvious, especially at ~610 nm, indicating the highest degree of crystallinity of the P3HT [13]. The corresponding cell also presents the largest Jsc and highest PCE. We carried out the testing of XDR to further characterize the degree of crystallinity. As shown in the inset of Fig. 3(b), the
Fig. 2. Performance parameters derived from eight #3 devices.
Z. Hu et al. / Renewable Energy 74 (2015) 11e17
15
Fig. 3. UVevis absorption spectra for the blend films at 298 K (a) and 294 K (b), the inset of Fig. 3(b) is the XRD profiles of the blend films.
preferred orientation of the (100) direction of the P3HT alkyl chains is reflected by the XRD peak intensity. The strongest XRD peaks indicate the highest degree of crystallization, which is correspondence to the absorption spectrum of the above tests. In Fig. 3(b), we also notice that another absorption peak of PCBM material at 350 nm is more acute and obvious, compared to Fig. 3(a). At low RT, the absorption intensity of PCBM with the reduced Ss and shortened Ts also increases, indicating that PCBM also can be crystallized when the solvent evaporation rate is slow. PCBM crystallization will cause excessive PCBM congregate at the air interface [16,24]. This phenomenon is also beneficial to improve the device performance [16]. The higher order and crystalline of the closer P3HT chains suggest a shorter hopping distance of the charge carriers between P3HT backbones, which could be beneficial to the higher Jsc. We have separately prepared single-carrier devices, and calculated the carrier mobility in the active layer by space charge limited current model using the formula:
J¼
9 v2 mε 8 d3
(1)
where ε is the dielectric constant, m is the carrier mobility, d is the thickness of the active layer. The single electron and hole carrier device structures are ITO/PEDOT:PSS/P3HT:PCBM/LiF/Al and ITO/ PEDOT:PSS/P3HT:PCBM/MoO3/Ag, respectively. The electron and hole mobilities in the active layer were drawn from the tested dark JeV curves, as shown in Table 1. At high RT under #1 condition, the solvent evaporation is too fast to arrange the P3HT molecular chains, so the hole mobility is lower than the electron mobility, which directly leads to lower device FF and Jsc. When the difference between the hole mobility and the electrons mobility is gradually reduced, the device FF and Jsc significantly improve. At 292 K under #3 condition, the highest hole mobility induced the balance carrier transport, which is consistent with the highest Jsc and FF of the PSC. In parallel, the series resistance (RS) was reduced from around 3.5 to 1.2 U cm2. The enhanced mobility causes a lower RS. However, the shunt resistance (RSh) was sharply increased from around 0.9 to Table 1 Hole and electron mobilities in the films, and RS, RSh of the corresponding devices at 294 K. Hole motility (cm2 V1 s1) #1 #2 #3 #3
(NA) (NA) (NA) (SA)
4.5 4.9 6.1 3.2
105 104 104 103
Electron motility (cm2 V1 s1) 1.4 1.2 1.4 1.8
103 103 103 103
RSh (kU cm2)
RS (U cm2)
0.9 1.3 5.2 15.5
3.5 2.5 1.9 1.2
15.5 kU cm2. The reduced RS and enhanced RSh provide the evidence of the improved FF. For high device performance, it requires that the active layer not only has a bicontinuous carrier transport channel, but also has a good intimate contact. The AFM height images of these P3HT:PCBM films at different RT are shown in Fig. 4. At 298 K, the film surface with an RMS roughness of 1.12 nm is presented. Decreasing with RT, the film shows more and more rough surface with RMS roughness from 4.05 to 10.39 nm. The increased roughness is also considered to be a signature of polymer self-organization [13]. Previous studies reported that rough surface morphology can result in the low device FF, lower than 0.60 [25]. Obviously, our result is contrast to this case. The devices with the rougher surface morphology have the higher FF, more than 70%. In Fig. 4(a), we can see that the film surface presents many “V” shaped spine glitch snuggled each other. However, in Fig. 4(d), these spine glitches completely disappeared, replaced by ups and downs of the passivated “U” shaped surfaces. For the deposited metal electrode, “U” shaped surface is more beneficial for close contact between the metal atom and the polymer. The “U“ shaped and rough surface of the active layer may reduce the efficient distance of the charge transport. The rough surface of the active layer may also increase the optical path in the active layer, therefore, the enhanced light absorption causes the increased Jsc [26,27]. The enhanced absorption spectrum of the active layer, balanced carrier mobility and favorable surface topography changes contribute to the increased the Jsc and FF, eventually lead to the improved PCE. However, we also note that Voc is gradually decreased. The possibility mechanism is the formation of a continuous band structure instead of molecular energy levels due the strong interchain interactions, which leads to a reduced effective bandgap. The changed Voc has also been reported previously in the literature for TA [7,8,11,12] and SA [13e15] P3HT:PCBM systems. Recently, Vandewal et al. pointed out that the theoretically maximum Voc is directly related to the energy of the charge transfer (CT) state [28,29]. Voc is linearly proportional to the HOMOeLUMO gap of organic solar cells, which is related to the energy of the charge transfer states [27,28]. The enhanced orderly arrangement of the conjugated P3HT chain length makes the effective radius of the interface polaron increase. Therefore, the charge-transfer complexes was effectively reduced, thus increase the separation efficiency of the charge-transfer complexes [30]. So, increasing with the CT energy, the device Voc is reduced, but the device Jsc and FF was greatly improved. As shown in our data, the dramatic change of Voc from 0.77 V to ~0.60 V, but Jsc and FF improved greatly, which is also consistent with the literature [31]. It seems to be no fundamental limit leading to a trade-off between Jsc and Voc in organic solar cells. The concern that Voc and Jsc have fundamentally conflicting requirements for optimization, just like in the P3HT:PCBM
16
Z. Hu et al. / Renewable Energy 74 (2015) 11e17
Fig. 4. AFM height images of the blend films at different RT: (a) 298 K (RMS: 1.12 nm), (b) 296 K (RMS: 4.05 nm), (c) 294 K (RMS: 7.92 nm), and (d) 292 K (RMS: 10.39 nm).
system. Thus, an optimization of the organic solar cell performance in view of the CT state requires careful material synthesis and interface design. Although the P3HT:PCBM films with the same thickness were prepared for measurements, the variation of the film thickness influences the performance of the devices. At the same Ss and Ts, the film thickness is increasing with the reduced RT. As the film thickness increases from ~80 nm (#1, 298 K) to ~230 nm (#4, 292 K), the Jsc decreases sharply along with a prominent decrease in Voc from 0.77 V to ~0.60 V. However, we think that the profound variations in photovoltaic performance were resulted from the pconjugated structure of RR-P3HT in the films, which is optimally developed when RT approximate 292 Ke294 K. Recently, Z. Li et al. reported that light-induced PCBM oligomerization have influenced organic solar cell device stability and performance [32]. During our experiments, all the films are fabricated under illumination conditions, the device performance may inadvertently influenced by the light exposure. 4. Conclusion We have systematically varied the Ss and Ts of spin coating at different RT in RR-P3HT:PCBM solar cells to study the influence of the self-organization of the polymer on the performance of the solar cells. Our results indicate that the p-conjugated structure of RR-P3HT in the films is optimally developed when RT approximate
292 Ke294 K. The corresponding PSCs performance shows a plateau region, with PCEs exceed 4% and FFs close to be 0.70. The relatively high efficiencies of these devices can be attributed to the enhanced absorption spectrum, balanced carrier mobility and favorable surface topography of the active layer. The findings of this study offer a facile and effective solution to achieve high efficiency base on RR-P3HT:PCBM system in air condition without other thermal treatment.
Acknowledgments This work was supported by the Natural Science Foundation of Zhejiang Province (Grant No. LQ13F050007), the National Science Foundation of China (Grant Nos. 11304170, 11174163, 51302137, 11347173), the Foundation of Zhejiang Educational Commission (Grant Nos. Y201326905, Y201326937), and the Natural Science Foundation of Nignbo City (Grant Nos. 2013A610033, 2013A610139). The author Z. Hu would like to thank the sponsored by K.C. Wong Magna Fund in Ningbo University.
References [1] Lizin S, Leroy J, Delvenne C, Dijk M, Schepper ED, Passel SV. A patent landscape analysis for organic photovoltaic solar cells: identifying the technology's development phase. Renew Energy 2013;57:5e11.
Z. Hu et al. / Renewable Energy 74 (2015) 11e17 [2] Blom PWM, Mihailetchi VD, Koster LJA, Markov DE. Device physics of polymer:fullerene bulk heterojunction solar cells. Adv Mater 2007;19: 1551e66. [3] Dennler G, Scharber MC, Brabec CJ. Polymer-fullerene bulk-heterojunction solar cells. Adv Mater 2009;21:1323e38. [4] Günes S, Neugebauer H, Sariciftci NS. Conjugated polymer-based organic solar cells. Chem Rev 2007;107:1324e38. [5] Krebs FC. Roll-to-roll fabrication of monolithic large-area polymer solar cells free from indium-tin-oxide. Sol Energy Mater Sol Cells 2009;93: 1636e41. [6] Li G, Shrotriya V, Yao Y, Huang J, Yang Y. Manipulating regioregular poly(3hexylthiophene):[6,6]-phenyl-C61-butyric acid methyl ester blendsdroute towards high efficiency polymer solar cells. J Mater Chem 2007;17:3126e40. [7] Padinger F, Rittberger RS, Sariciftci NS. Effects of postproduction treatment on plastic solar cells. Adv Funct Mater 2003;13:85e8. [8] Ma W, Yang C, Gong X, Lee K, Heeger AJ. Thermally stable, efficient polymer solar cell with nanoscale control of the interpenetrating network morphology. Adv Funct Mater 2005;15:1617e22. [9] Kim Y, Cook S, Tuladhar SM, Choulis SA, Nelson J, Durrant JR, et al. A strong regioregularity effect in self-organizing conjugated polymer films and highefficiency polythiophene:fullerene solar cells. Nat Mater 2006;5:197e203. [10] Kim H, So WW, Moon SJ. The importance of post-annealing process in the device performance of poly(3-hexylthiophene):methanofullerene polymer solar cells. Sol Energy Mater Sol Cells 2007;91:581e7. [11] Li G, Shrotriya V, Yao Y, Yang Y. Investigation of annealing effects and film thickness dependence of polymer solar cells based on poly(3hexylthiophene). J Appl Phys 2005;98:043704e8. [12] Nam CY, Su D, Black CT. High-performance air-processed polymerefullerene bulk heterojunction solar cells. Adv Funct Mater 2009;19:3552e9. [13] Li G, Shrotriya V, Huang J, Yao Y, Moriarty T, Emery K, et al. High-efficiency solution processable polymer photovoltaic cells by self-organization of polymer blends. Nat Mater 2005;4:864e8. [14] Zhao Y, Xie Z, Qu Y, Geng Y, Wang L. Solvent-vapor treatment induced performance enhancement of poly(3-hexylthiophene):methanofullerene bulkheterojunction photovoltaic cells. Appl Phys Lett 2007;90:043504e6. [15] Li G, Yao Y, Yang H, Shrotriya V, Yang G, Yang Y. “Solvent annealing” effect in polymer solar cells based on poly(3-hexylthiophene) and methanofullerenes. Adv Funct Mater 2007;17:1636e44. [16] Jo J, Na SI, Kim SS, Lee TW, Chung Y, Kang SJ, et al. Three-dimensional bulk heterojunction morphology for achieving high internal quantum efficiency in polymer solar cells. Adv Funct Mater 2009;19:2398e406. [17] Lin C, Lin EY, Tsai FY. Enhanced thermal stability and efficiency of polymer bulk-heterojunction solar cells by low-temperature drying of the active layer. Adv Funct Mater 2010;20:834e9.
17
[18] Wu Z, Song T, Jin Y, Sun B. High performance solar cell based on ultra-thin poly(3-hexylthiophene):Fullerene film without thermal and solvent annealing. Appl Phys Lett 2011;99:143306e8. [19] Peet J, Kim JY, Coates NE, Ma WL, Moses D, Heeger AJ, et al. Efficiency enhancement in low-bandgap polymer solar cells by processing with alkane dithiols. Nat Mater 2007;6:497e500. [20] Chen FC, Tseng HC, Ko CJ. Solvent mixtures for improving device efficiency of polymer photovoltaic devices. Appl Phys Lett 2008;92:103316e8. AJ, Meerholz K. Morphology control in solution-processed bulk-het[21] Moule erojunction solar cell mixtures. Adv Funct Mater 2009;19:3028e36. [22] Kim KC, Park JH, Park OK. New approach for nanoscale morphology of polymer solar cells. Sol Energy Mater Sol Cells 2008;92:1188e91. [23] Hu Z, Zhang J, Zhu Y. High-performance and air-processed polymer solar cells by room temperature drying of the active layer. Appl Phys Lett 2013;102: 043307e10. [24] Wang T, Dunbar ADF, Staniec PA, Pearson AJ, Hopkinson PE, MacDonald JE, et al. The development of nanoscale morphology in polymer:fullerene photovoltaic blends during solvent casting. Soft Matter 2010;6:4128e34. [25] Vak D, Kim SS, Jo J, Oh SH, Na SI, Kim J, et al. Fabrication of organic bulk heterojunction solar cells by a spray deposition method for low-cost power generation. Appl Phys Lett 2007;91:081102e4. [26] Lee JH, Kim DW, Jang H, Choi JK, Geng J, Jung JW, et al. Enhanced solar-cell efficiency in bulk-heterojunction polymer systems obtained by nanoimprinting with commercially available AAO membrane filters. Small 2009;5: 2139e43. [27] Shih CF, Hung KT, Wu JW, Hsiao CY, Li WM. Efficiency improvement of blended poly(3-hexylthiophene) and 1-(3-methoxycarbonyl)-propyl-1phenyl-(6,6)C61 solar cells by nanoimprinting. Appl Phys Lett 2009;94: 143505e7. [28] Vandewal K, Gadisa A, Oosterbaan WD, Bertho S, Banishoeib F, Severen IV, et al. The relation between open-circuit voltage and the onset of photocurrent generation by charge-transfer absorption in polymer:fullerene bulk heterojunction solar cells. Adv Funct Mater 2008;18:2064e70. €s O, Manca JV. On the origin of the [29] Vandewal K, Tvingstedt K, Gadisa A, Ingana open-circuit voltage of polymerefullerene solar cells. Nat Mater 2009;8: 904e9. [30] Hoofman RJOM, de Haas MP, Siebbeles LDA, Warman JM. Highly mobile electrons and holes on isolated chains of the semiconducting polymer poly(phenylene vinylene). Nature 1998;362:54e6. [31] Brabec CJ, Durrant J. Solution-processed organic solar cells. MRS Bull 2008;33: 670e5. [32] Li Z, Wong HC, Huang Z, Zhong H, Tan CH, Tsoi WC, et al. Performance enhancement of fullerene-based solar cells by light processing. Nat Commun 2013;4:2227e33.
Contents lists available at ScienceDirect
Renewable Energy journal homepage: www.elsevier.com/locate/renene
Natural drying effect on active layer for achieving high performance in polymer solar cells Ziyang Hu a, *, Jianjun Zhang b, *, Like Huang a, Jingyang Sun a, Ting Zhang a, Hongyun He a, Jing Zhang a, Houcheng Zhang a, Yuejin Zhu a a b
Department of Microelectronic Science and Engineering, Ningbo University, Zhejiang 315211, China College of Electronic Information and Optical Engineering, Nankai University, Tianjin 300071, China
a r t i c l e i n f o
a b s t r a c t
Article history: Received 30 June 2013 Accepted 18 July 2014 Available online
The self-organization of the polymer in solar cells based on regioregular poly(3-hexylthiophene)(RRP3HT):[6, 6]-phenyl C61-butyric acid methyl ester is studied systematically as a function of the room temperature (RT) (varied from 300 K to 290 K). Optimal self-organized structures within the RRP3HT:PCBM films are achieved by varying spin speed and time as well as the temperature at which the spin casting process occurs. These blend films are characterized by UVevis absorption spectroscopy, atomic force microscopy, and X-ray diffraction measurements. The optimum device efficiency can be achieved in naturally dried devices when spin coating within the temperature range 292e294 K. Both the power conversion efficiency (PCE) and fill factor (FF) of the optimum devices show a plateau region, with PCEs exceed 4% and FFs close to 0.70. For RT < 290 K, the corresponding devices show a wide distribution of performance parameters for the unhomogeneous active layer. While for RT > 296 K, the short current density, FF and PCE of the corresponding devices are gradually decreased, suggesting that there is a major change in the ordered structure of the polymer. Based on the results, it is demonstrated that high performance device can be achieved just by natural drying the active layer at RT condition in air condition without further thermal treatments. © 2014 Elsevier Ltd. All rights reserved.
Keywords: Natural drying Polymer solar cell P3HT:PCBM High performance Self-organization
1. Introduction Bulk heterojunction (BHJ) polymer solar cells (PSCs), whose active layer is comprised of semiconducting polymer and fullerene, afford the potential advantages of easy processability, low material costs and mechanical flexibility [1e6]. Generally, the performance of BHJ PSCs can be maximized by controlling the active layer morphology, because efficient photoinduced charge generation, transport, and collection at each electrode chiefly depend on the nanoscale morphology of the active layer [2,3]. Charge generation takes place at the donor/acceptor heterojunction, so the donor and acceptor need to be well-mixed so as to maximize the interface area of the heterojunction. On the other hand, charge transport requires continuous donor and acceptor domains throughout the active layer, so that electrons and holes can flow smoothly to their respective electrodes. Particularly, the transport of holes is often
* Corresponding authors. Department of Microelectronic Science and Engineering, Ningbo University, Zhejiang 315211, China. E-mail addresses: [email protected], [email protected] (Z. Hu), [email protected] (J. Zhang). http://dx.doi.org/10.1016/j.renene.2014.07.034 0960-1481/© 2014 Elsevier Ltd. All rights reserved.
the factor that limits the obtainable photocurrent, as the hole mobility of most used donors is lower than the electron mobility of commonly used acceptors [4e6]. Therefore, the optimal morphology of the active layer has been recognized as one that contains large donor/acceptor interface area, continuous donor and acceptor domains, and high molecular ordering or crystallinity of the donor. Over the last decade, many research groups have systematically analyzed external treatments in order to optimize the BHJ morphology of regioregular poly(3-hexylthiophene) (RR-P3HT):[6, 6]-phenyl C61-butyric acid methyl ester (PCBM) films for highefficiency solar cells. These treatments enable both components to evolve self-organized crystallization and achieve a well-ordered interpenetrating network of individual materials. The most notable of these treatments involve: (1) post thermal annealing (TA) of the active layer to activate rearrangement of the morphology [7e12]; (2) slowing down the drying rate (solvent annealing, SA) to allow the active layer to a desirable morphology [13e18]; and (3) the use of cosolvent and additive to manipulate the active layer morphology [19e22]. Although impressive efficiencies are achieved with these morphology-controlling techniques, the optimal
12
Z. Hu et al. / Renewable Energy 74 (2015) 11e17
morphology can only be attained with careful control of the processing conditions. Consequently, implementations of identical processing conditions have often resulted in wide-ranging device performance values and time-consumed. Furthermore, all highestperformance devices have been fabricated and characterized entirely in inert atmospheres [7e11,13e19], suggesting the importance of device processing in an oxygen-free environment. However, it is difficult for most laboratories to maintain the oxygen free environment in the whole fabricating process. On the other hand, device fabrication in air condition may decrease the cost for mass production of organic solar cells. In our previous report [23], high-performance and air-processed polymer solar cells were achieved by room temperature drying of the active layer. In this paper, we present a systematic study of the natural drying effect of the BHJ active layer afforded by spin-coating speed and time at room temperature (RT). Similar to TA and SA approach, we use the term “natural drying (NA)” to describe this process here. The RR-P3HT:PCBM active layer was investigated by NA treatment combined with the variable spin-coating speed (Ss) and time (Ts) from 300 K to 290 K. By systematically varying Ss and Ts at different RT, we have conducted a comprehensive investigation of the evolved morphology of the RR-P3HT:PCBM films and the performance of the PSCs. The solvent evaporation rate was controlled by RT, Ss and Ts. The spin coating conditions at different RT were then optimized in terms of the cell efficiency. Post SA was applied in conjunction with the NA process to further manipulate the morphology. The resultant morphology was analyzed with UVevis spectroscopy, X-ray diffraction (XRD), and atomic force microscopy (AFM), which was in turn correlated to the device characteristics. 2. Experimental details
drying at RT after spin coating. The Ss and Ts of the spin coating process were alternatively changed to optimize the device performance at a given temperature. As an example, spin coating for 10 s at 300 rpm and spin coating for 12 s at 230 rpm may result in the same device performance for natural drying at the same RT. For solvent annealing, the spin coated films (either liquid or solid, depending on Ss, Ts and RT) were then left in a grease-sealed cylinder (volume: 10 cm3) at the same RT. At the same time, small amounts of chlorobenzene solvent were added into the cylinder with a lid in order to prevent rapid evaporation of the solvent. 2.3. Device characterizations The P3HT:PCBM composite films used for optical and morphological characterization were prepared using the same method as used for the device fabrication. The samples for absorption measurements were prepared by the same procedure before the cathode deposition step, and then measured using a Varian Cary 5000 UVevis spectrophotometer. AFM images were obtained with a Digital Instruments Multimode scanning probe microscope operated in the tapping mode. All the films with the same thickness as measured by a Dektek profilometer, were prepared for measurement. The current-densityevoltage (JeV) characteristics of the devices were measured with a Keithley 2400 source meter under simulated AM1.5G irradiation (100 mW/cm2) in air condition. For JeV measurements, the light intensity was calibrated using a reference silicon solar cell. The hole mobility of the active layer was obtained from the dark JeV characteristics of hole-only devices operated in the space-charge limited current (SCLC) regime, where the devices were obtained by replacing the LiF/Al cathode of the regular devices with thermally evaporated MoO3/Ag.
2.1. Device fabrication
3. Results and discussion
The layer structure of the solar cells was glass/indium tin oxide (ITO)/poly(3,4-ethylenedioxythiophene):poly(styrene sulfonate) (PEDOT:PSS)/P3HT:PCBM/LiF/Al. PEDOT:PSS (Baytron P VP AI 4083), regioregular (90e93%) P3HT (MW ¼ 68000 g/mol, purity: 99%), and PCBM were purchased from H.C. Starck, Rieke Metals, Inc., and Nano C, Inc., respectively. All materials for the fabrication of the polymer solar cells were used as received. Before device fabrication, the ITO coated glass substrates were cleaned by sequential ultrasonic treatment in detergent, deionized water, acetone, and isopropyl alcohol. A thin layer (~40 nm) of PEDOT:PSS was spin-coated to modify the ITO surface and then the substrates were baked at 403 K for 20 min. The blend solutions of P3HT and PCBM (weight ratio of 1:0.8, 20 mg/ml in chlorobenzene) were spin-coated onto PEDOT:PSS-coated ITO glass substrates in air condition. LiF (0.8 nm)/Al (100 nm) electrodes were deposited by thermal evaporation under high vacuum with shadow masks determining the photoactive areas (6 mm2).
The experiments have been designed based on several considerations. Firstly, both fast drying and slow drying processes transform the polymer solution into a solid film. To obtain optimized performance, TA is commonly performed after fast drying the film, which requires the inert atmosphere [7e11]. NA is free of thermal treatment after fast drying the film, so the detrimental reactions can be avoided during post treatment. Therefore, the high performance can be expected to achieve in air condition. Secondly, the P3HT chains would induce the self-organization or crystallization by slow drying the film, which results in the improved performance [13,14,18]. The evaporation rate of the solvent is subject to the influence of temperature, the vapor pressure, the surface to volume ratio, the flow rate of air flowing through the surface. Therefore, it is realizable that the P3HT self-organization is performed by controlling RT, the selected solvent, Ss and Ts. Thirdly, chlorobenzene (CB) has been deliberately chosen as the solvent because its moderate boiling point allows a wide window of spin-coating process at RT condition. By systematically varying Ss and Ts at different RT, we have conducted a comprehensive investigation of the evolution of the RR-P3HT:PCBM films and the device performance. At elevated temperature, the solvent will be removed more rapidly and the film will lose its crystallinity. Here, the representative experimental conditions were chosen and the corresponding data are presented. RT was changed from 300 K to 290 K. The four combinations of Ss and Ts ((800 rpm, 30 s), (500 rpm, 20 s), (300 rpm, 10 s) and (230 rpm, 12 s)), which were named by #1, #2, #3, and #4, respectively, are selected. At the same time, The SA treatment was also performed for comparison. The SA process was described in the experimental section. RT was first fixed at 298 K, and conducted a comprehensive investigation of the evolution of
2.2. Natural drying Natural drying treatment of the P3HT:PCBM composite films were carried out before the deposition of the LiF/Al electrodes. The P3HT:PCBM solution was stirred on the heating plate for ~24 h at 283 K. In order to achieve the uniform films, prior to spinning the substrate was static and coated with the P3HT:PCBM solution completely. Immediately, the active layer was obtained by spincoating the solution in a single spinning process with a fixed spinning speed. This process was performed at different RT from 300 K to 290 K, as measured by a hygrothermograph (Testo, Ger.) in the open lab. The resulting P3HT:PCBM film was then natural
Z. Hu et al. / Renewable Energy 74 (2015) 11e17
the RR-P3HT:PCBM films and the device performance. The same trend is expected if RT is increased. At 298 K with the four combinations of Ss and Ts, all the device performance showed similar results when the active layers were treated with SA and NA processes, respectively. Even Ss and Ts were alternated, the performance parameters were almost unchanged and poor (PCE < 2%, FF < 0.50, Jsc < 6 mA/cm2, Voc ¼ 0.77~0.67 V). Since RT is too high, CB solvent evaporates quickly during the spin coating and also after the spin coating. The self-organization of the polymer P3HT must be realized in the presence of certain solvent. So SA and NA are useless for improved device performance. However, when the active layer was TA in air at 353 K for 15 min, the device performance was greatly improved to be 3%. The annealing condition is contrast to that of the majority of the literatures in the protected atmosphere (423 K, 30 min). It is probably that the polymer may react with water and oxygen in air condition, thereby reducing performance due to heating at high temperature for a long time. At 298 K, slowing the evaporation rate by adjusting the spin coating conditions the device performance increased slightly. The PCEs of the NA devices increased from 1.16% to 1.53%, and the PCEs of the SA devices increased from 1.28% to 1.70%, respectively. Since Ss and Ts are reduced, the solvent evaporation time is prolonged. Therefore, the P3HT chains can realize the self-organized during the insufficient time. It can be imaged that the complete realization of the P3HT self-organization is difficult over 298 K. Based on the above discussion, we can expect that the improved device performance can be realized by gradually reduced RT for the sequence experiments. Aforementioned experiments also show that the reduced Ss and Ts extend the solvent evaporation time, which results in the well realization of the self-organization. Thus, we carried out a series of experiments at 296 K by alternating Ss and Ts. At 296 K, at a high Ss and long Ts, the device performance still showed similar results when the active layers were treated with SA and NA processes, respectively. However, at a low Ss and short Ts, the SA device performance has been significantly improved, while the improvement of the NA device is very slightly. At the same Ss and Ts, the SA device showed reduced Voc, increased
13
FF and Jsc, compared to the NA devices as shown in Fig. 1. The reduced RT make solvent evaporation slows down, so that the time of realizing the self-organization of P3HT chains was increased. Especially, the evaporation of the residual solvent in a sealed glass dish would further slow, which suggests the performance of the SA device is better than that of the NA device. At 296 K, along with the reduced Ss and shortened Ts, the device performance presented significantly increased trend. The SA device with a Voc of 0.64 V, an FF of 65%, a Jsc of 9.01 mA/cm2, and a PCE of 3.05% was obtained. The performance is better than that of the TA device. Under the same #3 condition, the SA device showed reduced Voc, increased FF and Jsc, compared to the NA devices. Since RT is lowered, the total time of the solvent evaporation continues to prolong, so the P3HT chains staying in the solvent have a longer time to realize the self-organized. The improvement of the NA is not obvious, while the improvement of the SA device performance is prominent. It can be concluded that the P3HT chains realized the self-organization even after the spin coating process. The above experiments also demonstrate that RT plays a main role on controlling the solvent evaporation rate, and the Ss and Ts play an auxiliary or secondary role. It is suggested that the device performance would be further improved by lowering RT, reducing Ss and shorting Ts. Therefore, we further conducted the following experiments at 294 K in order to find the best experimental conditions. At 294 K, even at a high Ss and long Ts, both the use of NA and SA process has a significant impact on the device performance. At a low Ss and short Ts, the SA device performance has a significant improvement over the TA device performance. Under the #3 condition, the SA device performance parameters are listed as follows: Voc ¼ 0.63 V, FF ¼ 0.67, Jsc ¼ 9.58 mA/cm2, PCE ¼ 4.04%. The NA device performance parameters are listed as follows: Voc ¼ 0.65 V, FF ¼ 0.63, Jsc ¼ 8.05 mA/cm2, PCE ¼ 3.29%. The NA device performance also exceeds that of the TA device. Since the RT keeps reducing, the self-organization of the P3HT chains continues to increase. The solvent evaporation rate of NA slowed down with the decreased RT, but still faster than that of SA process. At 294 K, with the lowered Ss and shortened Ts, the overall performance continues to maintain an increasing trend. The device
Fig. 1. Performance parameters as a function of RT (from 298 K to 292 K) combined with the alternating Ss and Ts.
14
Z. Hu et al. / Renewable Energy 74 (2015) 11e17
Voc continues to reduce, FF continues to rise, Jsc continues to increase, and PCE continues to improve as shown in Fig. 2. However, the quality of the active layer is also greatly affected by the reduced RT, which results in a wide distribution of the performance parameters. So we decided to reduce RT to 292 K, but appropriately increased Ts to obtain a uniform film. At 292 K, even at a high Ss and long Ts, both the use of NA and SA process has a significant impact on the device performance. At a low Ss and short Ts, the SA and NA device performance is almost the same. The Jsc, FF and PCE of the SA device increased obviously to 10.78 mA/cm2, 0.68 and 4.62%, respectively. Under the #4 condition, the best NA device performance parameters are listed as follows: Voc ¼ 0.60 V, FF ¼ 0.72, Jsc ¼ 11.07 mA/cm2, PCE ¼ 4.78%. Here, one of the best device performances was obtained in air condition at RT without any annealing [12,17,18]. As Ss and Ts are reduced, the NA device performance continues to increase quickly, while the SA device performance shows slight improvement. Since RT is reduced, the CB solvent evaporation rate continues to slow down, the solvent staying in film continue to increase, so that the P3HT chains have enough time to realize self-organized. At 292 K, the solvent evaporation time of the NA process is close to that of the SA process, so the NA device showed similar performance to the SA device. According to the above rule and conclusions, we conducted the series of experiments at 291 K, the experimental results are similar to that conducted at 292 K. But the film-forming property of the active layer starts to deteriorate. When RT is reduced to 290 K, coupled with the low Ss and short Ts, the film uniformity can not be guaranteed, and the surface of the film appears obvious pointshaped or flower-shaped spots. The distribution of performance parameters was very wide as shown in Fig. 2. By adjusting RT, Ss and Ts to control the solvent evaporation rate, the function of the self-organization of the active layer were well realized. For RT > 298 K, it is difficult to adjust Ss and Ts to control the solvent evaporation. When RT was lowered to 294 K, the use of SA process can achieve good control of the solvent evaporation rate; thereby the device performance is significantly improved.
When RT is reduced to 292 K, the use of NA process also can achieve very good control the rate to achieve the well self-organization, thereby obtaining the optimized performance. Decreasing with RT, although the self-organization of the active layer can be realized, the uniformity of the active layer is difficult to obtain, which results in a wide distribution of performance parameters. So we think that in the vicinity of 292 K, NA process is the best choice for high device performance, while the SA process has a wide technical window. From a macroscopic physical view, we discussed the three factors how to control the solvent evaporation rate, thereby affecting the self-organization of the active layer, which leads to the different device performance. We believe that the improved device performance originates from three important contributions, enhanced absorption in the active layer, increased charge-carrier mobility and improved the contact between the active layer and the electrode interface. In order to understand the change in the absorption spectra of the active layer in different process conditions, we have tested two groups of absorption spectrum. Fig. 3(a) and (b) are the absorption spectra of the active layer using NA and SA treatment at 298 K and 294 K, respectively. At 298 K, the absorption spectrum of the active layer is almost the same even the Ss and Ts were alternated. Because the solvent evaporation is too fast, the SA process is difficult to maintain the slow drying of the active layer, which is also consistent with previous analysis. However, at 294 K, the absorption spectra of the active layer vary greatly, especially the use of the SA process. In all the cases, the vibronic features are clearly observed, indicating that both SA and NA process induce the ordering of RRP3HT in the blend film. Compared to Fig. 3(a), three distinct electronic vibration absorption peaks corresponding exactly to pep* band gap transition are presented in Fig. 3(b) [13,15]. At 294 K, three peaks of the SA blend films are the most obvious, especially at ~610 nm, indicating the highest degree of crystallinity of the P3HT [13]. The corresponding cell also presents the largest Jsc and highest PCE. We carried out the testing of XDR to further characterize the degree of crystallinity. As shown in the inset of Fig. 3(b), the
Fig. 2. Performance parameters derived from eight #3 devices.
Z. Hu et al. / Renewable Energy 74 (2015) 11e17
15
Fig. 3. UVevis absorption spectra for the blend films at 298 K (a) and 294 K (b), the inset of Fig. 3(b) is the XRD profiles of the blend films.
preferred orientation of the (100) direction of the P3HT alkyl chains is reflected by the XRD peak intensity. The strongest XRD peaks indicate the highest degree of crystallization, which is correspondence to the absorption spectrum of the above tests. In Fig. 3(b), we also notice that another absorption peak of PCBM material at 350 nm is more acute and obvious, compared to Fig. 3(a). At low RT, the absorption intensity of PCBM with the reduced Ss and shortened Ts also increases, indicating that PCBM also can be crystallized when the solvent evaporation rate is slow. PCBM crystallization will cause excessive PCBM congregate at the air interface [16,24]. This phenomenon is also beneficial to improve the device performance [16]. The higher order and crystalline of the closer P3HT chains suggest a shorter hopping distance of the charge carriers between P3HT backbones, which could be beneficial to the higher Jsc. We have separately prepared single-carrier devices, and calculated the carrier mobility in the active layer by space charge limited current model using the formula:
J¼
9 v2 mε 8 d3
(1)
where ε is the dielectric constant, m is the carrier mobility, d is the thickness of the active layer. The single electron and hole carrier device structures are ITO/PEDOT:PSS/P3HT:PCBM/LiF/Al and ITO/ PEDOT:PSS/P3HT:PCBM/MoO3/Ag, respectively. The electron and hole mobilities in the active layer were drawn from the tested dark JeV curves, as shown in Table 1. At high RT under #1 condition, the solvent evaporation is too fast to arrange the P3HT molecular chains, so the hole mobility is lower than the electron mobility, which directly leads to lower device FF and Jsc. When the difference between the hole mobility and the electrons mobility is gradually reduced, the device FF and Jsc significantly improve. At 292 K under #3 condition, the highest hole mobility induced the balance carrier transport, which is consistent with the highest Jsc and FF of the PSC. In parallel, the series resistance (RS) was reduced from around 3.5 to 1.2 U cm2. The enhanced mobility causes a lower RS. However, the shunt resistance (RSh) was sharply increased from around 0.9 to Table 1 Hole and electron mobilities in the films, and RS, RSh of the corresponding devices at 294 K. Hole motility (cm2 V1 s1) #1 #2 #3 #3
(NA) (NA) (NA) (SA)
4.5 4.9 6.1 3.2
105 104 104 103
Electron motility (cm2 V1 s1) 1.4 1.2 1.4 1.8
103 103 103 103
RSh (kU cm2)
RS (U cm2)
0.9 1.3 5.2 15.5
3.5 2.5 1.9 1.2
15.5 kU cm2. The reduced RS and enhanced RSh provide the evidence of the improved FF. For high device performance, it requires that the active layer not only has a bicontinuous carrier transport channel, but also has a good intimate contact. The AFM height images of these P3HT:PCBM films at different RT are shown in Fig. 4. At 298 K, the film surface with an RMS roughness of 1.12 nm is presented. Decreasing with RT, the film shows more and more rough surface with RMS roughness from 4.05 to 10.39 nm. The increased roughness is also considered to be a signature of polymer self-organization [13]. Previous studies reported that rough surface morphology can result in the low device FF, lower than 0.60 [25]. Obviously, our result is contrast to this case. The devices with the rougher surface morphology have the higher FF, more than 70%. In Fig. 4(a), we can see that the film surface presents many “V” shaped spine glitch snuggled each other. However, in Fig. 4(d), these spine glitches completely disappeared, replaced by ups and downs of the passivated “U” shaped surfaces. For the deposited metal electrode, “U” shaped surface is more beneficial for close contact between the metal atom and the polymer. The “U“ shaped and rough surface of the active layer may reduce the efficient distance of the charge transport. The rough surface of the active layer may also increase the optical path in the active layer, therefore, the enhanced light absorption causes the increased Jsc [26,27]. The enhanced absorption spectrum of the active layer, balanced carrier mobility and favorable surface topography changes contribute to the increased the Jsc and FF, eventually lead to the improved PCE. However, we also note that Voc is gradually decreased. The possibility mechanism is the formation of a continuous band structure instead of molecular energy levels due the strong interchain interactions, which leads to a reduced effective bandgap. The changed Voc has also been reported previously in the literature for TA [7,8,11,12] and SA [13e15] P3HT:PCBM systems. Recently, Vandewal et al. pointed out that the theoretically maximum Voc is directly related to the energy of the charge transfer (CT) state [28,29]. Voc is linearly proportional to the HOMOeLUMO gap of organic solar cells, which is related to the energy of the charge transfer states [27,28]. The enhanced orderly arrangement of the conjugated P3HT chain length makes the effective radius of the interface polaron increase. Therefore, the charge-transfer complexes was effectively reduced, thus increase the separation efficiency of the charge-transfer complexes [30]. So, increasing with the CT energy, the device Voc is reduced, but the device Jsc and FF was greatly improved. As shown in our data, the dramatic change of Voc from 0.77 V to ~0.60 V, but Jsc and FF improved greatly, which is also consistent with the literature [31]. It seems to be no fundamental limit leading to a trade-off between Jsc and Voc in organic solar cells. The concern that Voc and Jsc have fundamentally conflicting requirements for optimization, just like in the P3HT:PCBM
16
Z. Hu et al. / Renewable Energy 74 (2015) 11e17
Fig. 4. AFM height images of the blend films at different RT: (a) 298 K (RMS: 1.12 nm), (b) 296 K (RMS: 4.05 nm), (c) 294 K (RMS: 7.92 nm), and (d) 292 K (RMS: 10.39 nm).
system. Thus, an optimization of the organic solar cell performance in view of the CT state requires careful material synthesis and interface design. Although the P3HT:PCBM films with the same thickness were prepared for measurements, the variation of the film thickness influences the performance of the devices. At the same Ss and Ts, the film thickness is increasing with the reduced RT. As the film thickness increases from ~80 nm (#1, 298 K) to ~230 nm (#4, 292 K), the Jsc decreases sharply along with a prominent decrease in Voc from 0.77 V to ~0.60 V. However, we think that the profound variations in photovoltaic performance were resulted from the pconjugated structure of RR-P3HT in the films, which is optimally developed when RT approximate 292 Ke294 K. Recently, Z. Li et al. reported that light-induced PCBM oligomerization have influenced organic solar cell device stability and performance [32]. During our experiments, all the films are fabricated under illumination conditions, the device performance may inadvertently influenced by the light exposure. 4. Conclusion We have systematically varied the Ss and Ts of spin coating at different RT in RR-P3HT:PCBM solar cells to study the influence of the self-organization of the polymer on the performance of the solar cells. Our results indicate that the p-conjugated structure of RR-P3HT in the films is optimally developed when RT approximate
292 Ke294 K. The corresponding PSCs performance shows a plateau region, with PCEs exceed 4% and FFs close to be 0.70. The relatively high efficiencies of these devices can be attributed to the enhanced absorption spectrum, balanced carrier mobility and favorable surface topography of the active layer. The findings of this study offer a facile and effective solution to achieve high efficiency base on RR-P3HT:PCBM system in air condition without other thermal treatment.
Acknowledgments This work was supported by the Natural Science Foundation of Zhejiang Province (Grant No. LQ13F050007), the National Science Foundation of China (Grant Nos. 11304170, 11174163, 51302137, 11347173), the Foundation of Zhejiang Educational Commission (Grant Nos. Y201326905, Y201326937), and the Natural Science Foundation of Nignbo City (Grant Nos. 2013A610033, 2013A610139). The author Z. Hu would like to thank the sponsored by K.C. Wong Magna Fund in Ningbo University.
References [1] Lizin S, Leroy J, Delvenne C, Dijk M, Schepper ED, Passel SV. A patent landscape analysis for organic photovoltaic solar cells: identifying the technology's development phase. Renew Energy 2013;57:5e11.
Z. Hu et al. / Renewable Energy 74 (2015) 11e17 [2] Blom PWM, Mihailetchi VD, Koster LJA, Markov DE. Device physics of polymer:fullerene bulk heterojunction solar cells. Adv Mater 2007;19: 1551e66. [3] Dennler G, Scharber MC, Brabec CJ. Polymer-fullerene bulk-heterojunction solar cells. Adv Mater 2009;21:1323e38. [4] Günes S, Neugebauer H, Sariciftci NS. Conjugated polymer-based organic solar cells. Chem Rev 2007;107:1324e38. [5] Krebs FC. Roll-to-roll fabrication of monolithic large-area polymer solar cells free from indium-tin-oxide. Sol Energy Mater Sol Cells 2009;93: 1636e41. [6] Li G, Shrotriya V, Yao Y, Huang J, Yang Y. Manipulating regioregular poly(3hexylthiophene):[6,6]-phenyl-C61-butyric acid methyl ester blendsdroute towards high efficiency polymer solar cells. J Mater Chem 2007;17:3126e40. [7] Padinger F, Rittberger RS, Sariciftci NS. Effects of postproduction treatment on plastic solar cells. Adv Funct Mater 2003;13:85e8. [8] Ma W, Yang C, Gong X, Lee K, Heeger AJ. Thermally stable, efficient polymer solar cell with nanoscale control of the interpenetrating network morphology. Adv Funct Mater 2005;15:1617e22. [9] Kim Y, Cook S, Tuladhar SM, Choulis SA, Nelson J, Durrant JR, et al. A strong regioregularity effect in self-organizing conjugated polymer films and highefficiency polythiophene:fullerene solar cells. Nat Mater 2006;5:197e203. [10] Kim H, So WW, Moon SJ. The importance of post-annealing process in the device performance of poly(3-hexylthiophene):methanofullerene polymer solar cells. Sol Energy Mater Sol Cells 2007;91:581e7. [11] Li G, Shrotriya V, Yao Y, Yang Y. Investigation of annealing effects and film thickness dependence of polymer solar cells based on poly(3hexylthiophene). J Appl Phys 2005;98:043704e8. [12] Nam CY, Su D, Black CT. High-performance air-processed polymerefullerene bulk heterojunction solar cells. Adv Funct Mater 2009;19:3552e9. [13] Li G, Shrotriya V, Huang J, Yao Y, Moriarty T, Emery K, et al. High-efficiency solution processable polymer photovoltaic cells by self-organization of polymer blends. Nat Mater 2005;4:864e8. [14] Zhao Y, Xie Z, Qu Y, Geng Y, Wang L. Solvent-vapor treatment induced performance enhancement of poly(3-hexylthiophene):methanofullerene bulkheterojunction photovoltaic cells. Appl Phys Lett 2007;90:043504e6. [15] Li G, Yao Y, Yang H, Shrotriya V, Yang G, Yang Y. “Solvent annealing” effect in polymer solar cells based on poly(3-hexylthiophene) and methanofullerenes. Adv Funct Mater 2007;17:1636e44. [16] Jo J, Na SI, Kim SS, Lee TW, Chung Y, Kang SJ, et al. Three-dimensional bulk heterojunction morphology for achieving high internal quantum efficiency in polymer solar cells. Adv Funct Mater 2009;19:2398e406. [17] Lin C, Lin EY, Tsai FY. Enhanced thermal stability and efficiency of polymer bulk-heterojunction solar cells by low-temperature drying of the active layer. Adv Funct Mater 2010;20:834e9.
17
[18] Wu Z, Song T, Jin Y, Sun B. High performance solar cell based on ultra-thin poly(3-hexylthiophene):Fullerene film without thermal and solvent annealing. Appl Phys Lett 2011;99:143306e8. [19] Peet J, Kim JY, Coates NE, Ma WL, Moses D, Heeger AJ, et al. Efficiency enhancement in low-bandgap polymer solar cells by processing with alkane dithiols. Nat Mater 2007;6:497e500. [20] Chen FC, Tseng HC, Ko CJ. Solvent mixtures for improving device efficiency of polymer photovoltaic devices. Appl Phys Lett 2008;92:103316e8. AJ, Meerholz K. Morphology control in solution-processed bulk-het[21] Moule erojunction solar cell mixtures. Adv Funct Mater 2009;19:3028e36. [22] Kim KC, Park JH, Park OK. New approach for nanoscale morphology of polymer solar cells. Sol Energy Mater Sol Cells 2008;92:1188e91. [23] Hu Z, Zhang J, Zhu Y. High-performance and air-processed polymer solar cells by room temperature drying of the active layer. Appl Phys Lett 2013;102: 043307e10. [24] Wang T, Dunbar ADF, Staniec PA, Pearson AJ, Hopkinson PE, MacDonald JE, et al. The development of nanoscale morphology in polymer:fullerene photovoltaic blends during solvent casting. Soft Matter 2010;6:4128e34. [25] Vak D, Kim SS, Jo J, Oh SH, Na SI, Kim J, et al. Fabrication of organic bulk heterojunction solar cells by a spray deposition method for low-cost power generation. Appl Phys Lett 2007;91:081102e4. [26] Lee JH, Kim DW, Jang H, Choi JK, Geng J, Jung JW, et al. Enhanced solar-cell efficiency in bulk-heterojunction polymer systems obtained by nanoimprinting with commercially available AAO membrane filters. Small 2009;5: 2139e43. [27] Shih CF, Hung KT, Wu JW, Hsiao CY, Li WM. Efficiency improvement of blended poly(3-hexylthiophene) and 1-(3-methoxycarbonyl)-propyl-1phenyl-(6,6)C61 solar cells by nanoimprinting. Appl Phys Lett 2009;94: 143505e7. [28] Vandewal K, Gadisa A, Oosterbaan WD, Bertho S, Banishoeib F, Severen IV, et al. The relation between open-circuit voltage and the onset of photocurrent generation by charge-transfer absorption in polymer:fullerene bulk heterojunction solar cells. Adv Funct Mater 2008;18:2064e70. €s O, Manca JV. On the origin of the [29] Vandewal K, Tvingstedt K, Gadisa A, Ingana open-circuit voltage of polymerefullerene solar cells. Nat Mater 2009;8: 904e9. [30] Hoofman RJOM, de Haas MP, Siebbeles LDA, Warman JM. Highly mobile electrons and holes on isolated chains of the semiconducting polymer poly(phenylene vinylene). Nature 1998;362:54e6. [31] Brabec CJ, Durrant J. Solution-processed organic solar cells. MRS Bull 2008;33: 670e5. [32] Li Z, Wong HC, Huang Z, Zhong H, Tan CH, Tsoi WC, et al. Performance enhancement of fullerene-based solar cells by light processing. Nat Commun 2013;4:2227e33.