Spray coating technique has been established as a promising substitute for the traditional coating methods in the fabrication of organic devices in many reports recently. Control of film morphology at the microscopic scale is critical if spray-coated devices are to achieve high performance. Here we investigate electrospray deposition protocols for the fabrication of poly(3,4-ethylenedioxythiophene):poly(styrenesulfonate) (PEDOT:PSS) thin films with a single additive system under ambient conditions at room temperature. Critical deposition parameters including solution composition, applied voltage, and relative humidity are discussed systematically. Optimized process for preparing homogenous PEDOT:PSS thin films is applied to all-electrospray-coated organic photovoltaic cells and contributes to a power conversion efficiency comparable to that of the corresponding all-spin-coated device.

1. Introduction

There has been intensive research on fabrication methods of thin films fabrication for cost-effective organic electronic devices in the past decade [15]. Spin coating is the most widely used wet process of making devices in laboratories due to its simplicity and reproducibility. However, this method cannot meet the requirement of high-throughput roll-to-roll process and have limitations when applying to large-area devices and flexible substrates [68]. Besides, this method is hard to create multilayer devices since the subsequently superposed thin films often dissolve in the previously coated thin films. Therefore, a variety of printing methods are established, for example, screen printing [9], doctor blading [10], and inkjet printing [1113]. Among the alternatives to spin coating, electrospray deposition is attractive due to its simplicity and versatility. It has been demonstrated as a technique capable of fabricating large-area organic electronic devices with multiple layers and patterns [1418]. Comparing with spin coating, this technique works with much more dilute solutions with a high material utilization efficiency and circumvents the solubility problem for conjugated polymer thin films [1921].

Most research works about spray-coated devices are focused on the fabrication of the active layer. Our group also published several papers on the electrospray deposition of active layer for organic light-emitting diodes and organic photovoltaic cells [17, 2225]. However, systemic investigations on other functional layers are addressed seldom. As a potential substitute for metal in the electrode, poly(3,4-ethylenedioxythiophene):poly(styrenesulfonate) (PEDOT:PSS) has been widely applied to organic devices as a hole transport layer. It can improve the reproducibility of the device characteristics by smoothing (indium tin oxide) ITO spikes [26]. PEDOT:PSS is an aqueous dispersion with high surface tension, which makes it difficult to atomize into fine droplets during spray deposition and results in inhomogeneous thin films on substrate. Girotto et al. try to enhance the coverage of the substrate and reduce the roughness of resultant films by optimizing the ink formulations with a two-solvent system [27]. Kim et al. used a substrate heating method to improve the morphology of PEDOT:PSS films [28].

Here we reported a method of improving PEDOT:PSS film morphology under ambient conditions at room temperature by simply adjusting the deposition parameter and physical properties of PEDOT:PSS solution. Electrospray deposition of PEDOT:PSS films is studied by independently varying the applied voltage between nozzle and substrate and the volume fraction of the single additive in the precursor solution. Relative humidity, which is a factor affecting the deposition process in practical application, is also discussed. To demonstrate the feasibility of fabricating all-spray-coated organic photovoltaic cells (OPVs), optimized deposition parameters of PEDOT:PSS layers are applied to prepare all-spray-coated OPVs.

2. Material and Methods

PEDOT:PSS inks were prepared by adding dimethyl sulfoxide (99.5%, Wako Pure Chemical Industries, Ltd.) (DMSO) with a volume fraction varying from 60% to 90% into PEDOT:PSS (Clevios P VP AI 4083, H.C. Starck). To make the electrospray-coated PEDOT:PSS thin films comparable, the volume of the PEDOT:PSS solution is kept as a constant for all PEDOT:PSS inks. Preparation of the P3HT:PCBM ink was described in our previous report [17]. Poly(3-hexylthiophene-2,5-diyl) (P3HT, Lumtec) and (6,6)-phenyl-C61 butyric acid methyl ester (PCBM, Lumtec) were used as the donor and the acceptor material of the bulk heterojunction OPVs, respectively. P3HT and PCBM (P3HT : PCBM = 6 : 5, w/w) were dissolved in o-dichlorobenzene (99%, Wako Pure Chemical Industries, Ltd.) with the concentration of 1 mg/mL. Acetonitrile (99.5%, Wako Pure Chemical Industries, Ltd.) was added into the mixture as an additive solvent to a concentration of 10 vol%. All inks were stirred overnight to achieve complete dissolution prior to electrospray deposition. ITO-coated glass substrates were successively cleaned in solid cleaner, deionized water, acetone, and isopropyl alcohol for 10 min and were treated in ultraviolet ozone cleaner for 20 min before deposition.

A schematic diagram for the electrospray deposition setup is shown in Figure 1. A glass capillary was fabricated using a puller (PC-10, Narishige) and a microforge (MF-900, Narishige). The inner diameter of the glass capillary was approximately 100 μm and 50 μm for the deposition of PEDOT:PSS inks and the P3HT:PCBM ink, respectively. The positive high voltage was generated by a high voltage power source (HJPQ-30P1, Matsusada Precision) and was applied to the ink inside the glass capillary through a copper wire. The ITO-coated glass substrate was grounded and used to collect the atomized droplets. The distance between capillary tip and ITO substrate was 6 cm. The electrospray deposition setup is placed in a home-built glove box in which the relative humidity is controlled by an air compressor consisted of an oil-free pump (Bebikon, Hitach, Japan) and a heatless air dryer (Super Pack QSQ010A, Orion, Japan). All experiments were performed under ambient conditions at room temperature unless otherwise noted.

OPVs discussed in this paper have the same architecture, ITO/PEDOT:PSS/P3HT:PCBM/Al. Their PEDOT:PSS layer and the active layer were fabricated by either electrospray deposition or spin coating. The thickness of both spin-coated and electrospray-coated PEDOT:PSS thin films was kept at about 20 nm by adjusting spin speed and deposition time. The thickness of P3HT:PCBM layers was kept at around 100 nm for comparison. More deposition details of the P3HT:PCBM layers can be found in our previous report [23]. The topography of the PEDOT:PSS thin films was analyzed using AFM (Seiko, SPA300HV). The performance of the OPVs was measured under simulated illumination at AM 1.5 G, 100 mW/cm2 with a Keithley 2400 source meter controlled by a computer program.

3. Results and Discussion

Control of droplet size is critical for electrospray deposition to get desired surface morphology since thin films form with the drying of the droplets. De La Mora and Loscertales [29] proposed the following equations to estimate the droplet diameter and spray current of electrohydrodynamic atomization in the cone-jet mode through the liquid cone:where is droplet diameter (m), is a function of the liquid permittivity, is liquid flow rate (m3s−1), is permittivity of a vacuum (CV−1m−1), is relative permittivity of the liquid (CV−1m−1), and is conductivity (Sm−1). This equation indicates that droplet size can be controlled by changing flow rate, relative permittivity, or conductivity of liquid. Although direct measurement of droplet size is rather difficult, the effect of the three parameters on droplet size could reflect finally on the morphology of coated thin films. Here the relative permittivity and conductivity of inks were turned through a single additive system. As the liquid inside of the capillary is driven mainly by a high-voltage-induced electrical force, the flow rate of an ink in cone-jet mode can be controlled by the applied voltage between the liquid inside of capillary and the target substrate. To make the protocol of this paper simple and clear, the applied voltage was varied to match the ink with different relative permittivity and conductivity. The morphology of spray-coated thin films was correlated with ink properties and deposition parameters.

The conductivity of PEDOT:PSS films increases several orders of magnitude after adding high boiling point solvents and polar compounds, such as DMSO, ethylene glycol, or sorbitol [3032]. DMSO has a surface tension (43.54 mN·m−1) much smaller than that of water (72.86 mN·m−1) [33], which facilitates its atomization. Here DMSO was selected as the additive of PEDOT:PSS to adjust the conductivity, relative permittivity, and surface tension of PEDOT:PSS inks. The relationship between the applied voltage and the corresponding flow rate for a PEDOT:PSS ink is shown in Figure 2(a). The onset of spray in cone-jet mode moved gradually from 10 kV to 6 kV as the volume fraction of DMSO was increased by steps from 60% to 90%, which indicates that DMSO is an effective additive for depressing the atomization energy required for the electrospray deposition of PEDOT:PSS. For each PEDOT:PSS ink, three data points after the onset of spray were collected for comparison. It is apparent that the flow rate has a tendency to increase with the applied voltage. The exception case of 60% DMSO may be attributed to a transition of spray mode resulting from insufficient DMSO for keeping the spray working in con-jet mode. It worth noting that the volume fraction of DMSO affects not only onset of spray but also the flow rate. The PEDOT:PSS ink with more DMSO produces a comparable or higher flow rate than the ink with few DMSO at the same applied voltage. The root mean square (RMS) roughness of the PEDOT:PSS thin films in Figure 2(b) was calculated from their 50 μm × 50 μm AFM topographies. We can found a positive correlation between the RMS roughness of coated thin films and the flow rate. It can be explained from (1) that the droplet size is always increase with the flow rate for a given liquid. Representative topographies of 80 vol% DMSO samples are shown in Figure 3. As the increase of the flow rate, the size of features on the surface of 80 vol% DMSO samples becomes larger. This result is coherent with what (1) predicted. A minimum RMS roughness of about 5 nm is obtained at the smallest flow rate of 0.25 μL/min, which indicates that homogeneous PEDOT:PSS thin films were obtained under this condition. The smoothing effect of low flow rate is weakened when the volume fraction of DMSO increases to 90 vol%. Both the flow rate and the surface roughness vary slightly when changing the applied voltage, which may be due to the over diluting of the PEDOT:PSS ink. We can conclude that thin films of low RMS roughness can be formed with small droplets resulting from a rather low flow rate.

One can also compare the RMS roughness of PEDOT:PSS thin films coated from inks at almost the same flow rate from Figure 2. For 70 vol%, 80 vol%, and 90 vol% DMSO samples, a flow rate around 5 μL/min is obtained at the applied voltage of 10 kV, 9 kV, and 8 kV, respectively. The RMS roughness for the three samples reduces with increasing the fraction of DMSO. As the relative permittivity and the conductivity of the PEDOT:PSS inks change with the fraction of DMSO, the size of droplets produced from these inks at around 5 μL/min differs. The small RMS roughness could result from the reduction in the size of droplets. In addition, low surface tension of DMSO could also contribute to the smoothing effect since the droplets containing more DMSO tend to have larger contact with the surface. The topography of the three samples can be found in Figure 3. There are a lot of pinholes in Figure 3(e), which indicates that the distribution of PEDOT:PSS droplets is ununiform. These pinholes disappear when the fraction of DMSO increases to 90%; however, many aggregates come out from the surface. A representative topography for the 60 vol% DMSO sample is shown in Figure 3(d). Features of several micrometers in size make the surface of the PEDOT:PSS thin film inhomogeneous. The big feature size could be due to the high flow rate and instability of spray during the transition of spray mode. In general, changing the fraction of DMSO in the PEDOT:PSS inks could be a method of adjusting the morphology of electrospray-coated PEDOT:PSS thin films, but not as direct as changing the flow rate.

As the PEDOT:PSS used is a aqueous solution and all the electrospray deposition of PEDOT:PSS inks above was performed in ambient air, water vapor in the air may influence the deposition process. Here an investigation on the effect of relative humidity was also conducted. As shown in Figure 4, the level of the flow rate for the PEDOT:PSS inks of different fraction of DMSO keeps the same at different relative humidity. For all the PEDOT:PSS inks tested, the flow rate decreased with the increment of relative humidity. It is worth noting that the effect of relative humidity at 10% and 60% on flow rate enhances for the PEDOT:PSS inks having less DMSO. This implies that the more the water in the inks is, the more the sensitive flow rate is in electrospray deposition at a very low or high relative humidity. By comparison, the flow rate varies not a lot during electrospray deposition of PEDOT:PSS inks at a moderate relative humidity (20% and 40%). The AFM topographies of the 80 vol% DMSO samples are shown in Figure 5. It is rather clear from the four topographies that the surface of the PEDOT:PSS thin films becomes more and more homogeneous as the increment of relative humidity. One possible explanation is that the PEDOT:PSS droplets evaporate slowly at high relative humidity and have a high probability of merging with other droplets to form a smooth surface. The RMS roughness values of the four films were extracted from their topographies and plotted in Figure 4. A similar correlation between flow rate and RMS roughness can be found when comparing to the electrospray deposition in ambient air, which implies that the decreasing of RMS roughness could partially attribute to the shrinking of droplet size induced by the low flow rate.

As described in the introduction, electrospray deposition could be used to substitute for the spin coating process in the fabrication of semiconducting devices. The optimized deposition parameters for PEDOT:PSS inks were applied to the fabrication of solution-processed OPVs. The current density-voltage characteristics and the corresponding performance data for the OPVs in this work are shown in Figure 6 and Table 1, respectively. Comparing devices B and C, we find that the negative effect of inhomogeneity caused by the electrospray deposition on device performance became weak when we applied the electrospray deposition technique to the fabrication of the second layer (P3HT:PCBM layer) instead of the first layer (PEDOT:PSS layer). Device C was fabricated using the same processes as device D except for the processing of the P3HT:PCBM layer. However, at 1.68%, device C has lower power conversion efficiency (PCE) than device D. This indicates that the inhomogeneity of the electrospray-coated PEDOT:PSS thin films was exaggerated by the subsequent spin-coated P3HT:PCBM layer and finally affected the whole device. Both the PEDOT:PSS and the P3HT:PCBM layer of device D were made by electrospray deposition, which resulted in an intermediate PCE of 2.10% as compared to devices B and C. It is worth noting that both the open voltage and the fill factor of device D are higher than those of device B and C, which may be due to the fusion of the electrospray-coated PEDOT:PSS and P3HT:PCBM layers. These results suggest that the intrinsically formed interfacial boundaries between the electrospray-coated blend pancakes resist charge transport and limit the efficiency of electrospray-coated OPVs. The PCE of devices B and D are comparable but still lower than that of the all-spin-coated device (device A). In general, the performance of OPVs is primarily influenced by the inherent morphology of electrospray-coated thin films when applying electrospray deposition to fabrication of OPVs.

4. Conclusions

Control of the morphology of electrospray-coated PEDOT:PSS thin films in cone-jet mode is archived by adjusting flow rate and ink properties via applied voltage and the fraction of the additive DMSO. The RMS roughness of the PEDOT:PSS thin films is dominated by the flow rate of the PEDOT:PSS inks tested in this paper. Homogeneous PEDOT:PSS thin film was coated from the ink having 80 vol% DMSO at a low flow rate. PEDOT:PSS inks of different fraction of the additive DMSO can produce thin films of different feature size at the same flow rate. For the electrospray deposition of aqueous solutions like PEDOT:PSS, flow rate becomes very sensitive at both low and high relative humidity.

Moreover, we found incompatibility between electrospray deposition and spin coating in the fabrication of OPVs when electrospray deposition is used to the fabricating the first layer of OPVs. The PCE of the all-electrospray-coated OPV archived 2.10%. However, it is still lower than that of the all spin-coated one due to the interfacial boundaries formed intrinsically during electrospray deposition. All the findings above are rather important for the practical application of electrospray deposition, especially in large area and high-throughput production of PEDOT:PSS thin films.

Competing Interests

The authors declare that they have no competing interests.


The authors gratefully acknowledge the support by the JSPS KAKENHI Project (no. 26420267). The authors also thank Dr. Tammo Reisewitz for his comments on the scientific writing and thank Miss. Asumi Suzuki for her assistance in the experiment.