Improved Reliability of Small Molecule Organic Solar Cells by Double Anode Buffer Layers

1 Department of Applied Physics, National University of Kaohsiung, Kaohsiung 81148, Taiwan 2Department of Electronic Engineering, Fortune Institute of Technology, Kaohsiung 83160, Taiwan 3 Institute of Microelectronics, Department of Electrical Engineering, National Cheng-Kung University, Tainan 70101, Taiwan 4Department of Electro-Optical Engineering, Southern Taiwan University of Science and Technology, Tainan 71005, Taiwan


Introduction
Organic solar cells (OSCs) have strong potential advantages compared with conventional inorganic solar cells, which could expand the range of applications, including low-cost, light-weight, colorful, mechanically flexible, and large-area devices.Now, a milestone in efficiency of OSCs has been achieved in recent years, allowing for 10% power conversion efficiency (PCE) by tandem structures even significant to 12% [1], which makes its commercial industrialization very promising.Many works also tend to design the appropriate configuration for carrier generation and transportation to attain high efficiency and stability.At the same time, a lot of organic semiconducting materials have extremely interesting properties in terms of photocurrent generation and reveal very high absorption coefficients in the visible region of the solar spectrum for even <100 nm films, making them promising compounds for photovoltaic devices [2].Copper phthalocyanine (CuPc) was employed as a donor in the first bilayer heterojunction OSC reported in the literature over two decades [3].Their well-known capability stems from the sensitizer by photo-induced electron transfer to acceptors.It was coupled with the fullerenes (C 60 or C 70 ) when the photo-induced electron transfer was employed.However, on the contrary, the reliability studies and the scientific explanations of degradation phenomena of OSCs device have not been further researched.Reliability is particularly difficult to discuss because both device parameters and experimental factors would intensively affect the degradation of device, as well as the possibility for interplay among these factors.Some groups have tested for reliability of OSCs by applying indium tin oxide (ITO) as an anode, the poly(3, 4-thylenedioxythiophene)-poly(styrenesulfonate) (PEDOT : PSS) or metal oxide (MoO 3 , WO 3 ,V 2 O 5 ) as the anode buffer layer (ABL), which is penetrative and collects the holes generated in the active layer, and active donor/acceptor layers with a low work function metal aluminium (Al) as the cathode on top of the cell for electron collection [4][5][6][7].Besides, it was indicated that the degradation of OSCs device is determined by the following three important elements, including photodegradation, ambient degradation (H 2 O and O 2 ), and thermal degradation of the photoactive organic layer [8][9][10][11][12][13][14][15].We focus on the impact of ambient degradation because the absorption of water would lead to degradation of device performance, reflecting on series resistance (  ).To achieve planar heterojunction (PHJ) of small molecule organic solar cells (SM-OSCs) and prevent it from water and oxygen damage, the pentacene layer was introduced between the MoO 3 and CuPc layer which provided improved cell stability and kept the performance of device, owing to material characteristics [16,17].The PEDOT : PSS layer was substituted for MoO 3 because of wide energy gap semiconductor with high work function [5][6][7].The shortage of organic materials is narrow absorption range in visible region, which leads to the low short-circuit current density ( sc ).Therefore, to get the highest  sc for the best performance, the optimization of the device structure is necessary [18,19].In this work, we show an important presence of double ABL to improve the reliability of device and further generally elucidate the impact of the interfacial layers on device lifetime and photovoltaic parameters.In addition, the  sc variation of device with different ABL material and thicknesses is also investigated in detail.

Experiment
In this study, the vacuum-deposited OSCs device was fabricated on patterned ITO-glass substrate.The ITO-glass substrates with a sheet resistance of 7 Ω/sq were sequentially cleaned by ultrasonic treatment in acetone, isopropyl alcohol, and deionized water for 5 min each and finally dried by highpurity nitrogen (N 2 ) gas before deposition.The organic materials used in the device as MoO 3 (99%), pentacene (99%), CuPc (99%), C 60 (99.95%), and BCP (99%) were purchased from Sigma-Aldrich which were without further purification.All organic materials were successively deposited onto the substrates by thermal evaporation at a deposition rate of 0.02-0.04nm/s below the pressure of 2.3 × 10 −6 Torr.The Al as a cathode was deposited to 100 nm thickness at a deposition rate of 0.02-0.15nm/s by thermal evaporation through a shadow mask, giving an active area of 6 mm 2 .And the structure of device was ITO/buffer/CuPc (20 nm)/C 60 (30 nm)/BCP (10 nm)/Al (100 nm), where the buffer layers were any or all formation of MoO 3 .The films thickness and deposition rate were monitored by using an oscillating quartz thickness monitor.
The configurations of OSC device and the energy level diagram in this study are shown in Figure 1.The currentvoltage characteristics were measured with a Keithley 2400 source meter, under an illumination of 100 mW/cm 2 with an AM 1.5 G sun simulator.The light intensity was measured using a calibrated Si-solar cell.Devices were further measured by atomic force microscope (AFM) to analyze the surface morphology and by spectrophotometer (Hitachi UV-3010, Japan) to gauge the absorption spectrum of device.All measurements were carried out in air with encapsulation.HOMO and LUMO of the organic materials was determined by absorption spectra.Figure 2 shows the current densityvoltage ( − ) characteristics of devices under illumination of 1-sun AM 1.5 G solar radiation.The Devices from A to D (ITO/buffer layer/C 60 (30 nm)/BCP (10 nm)/Al (100 nm)) consisted of both MoO 3 and pentacene materials.Corresponding device parameters, series resistance (  ), and performance were listed in Tables 1-3. Figure 3 shows the absorption spectra of all devices.The ABL is playing an important role in device because it has not only decreased the effect of non-Ohmic electrodes on device characteristics and revealed properties of the organic heterojunction, but also influenced the ability of carrier transportation in device.

Results and Discussion
We have completed the optimized parameters of Devices A and B under the premise of using MoO 3 and pentacene as the ABL materials, respectively.First, for Device A using MoO 3 material, the best performance was achieved without MoO 3 and it was compared to relative articles [20].It is found that the MoO 3 material was not good for basic CuPc/C 60 OSCs using Al as the cathode, although its advantages of wide energy gap semiconductor with high work function have been reported in many works [5][6][7].The  sc of device decreased from 3.41 mA/cm 2 to 2.84 mA/cm 2 and  oc from 0.39 V to 0.36 V when the MoO 3 layer thickness was added from 0 nm to 5 nm in  1.To further elucidate the distinction on MoO 3 and pentacene, we have focused on the absorption spectrum of used materials in this work.Before measuring these devices, the bare glass substrate is measured as a baseline.The absorptions of C 60 contribution (350-520 nm) and CuPc contribution (550-800 nm) were nearly the same for all devices.The  sc of device with pentacene layer between ITO anode and CuPc layer slightly increased due to the small decrease of   .It is attributed to the extra absorption of pentacene layer from 660 nm to 680 nm in Figure 3. On the other hand, comparing to Device A, the performance of Device B was enhanced from 0.56% to 0.79%.The reason for improved performance was the material characteristics of pentacene.The pentacene layer built an internal electric field at the interface of pentacene/CuPc that can enhance the hole extraction from CuPc layer to the ITO anode.The physical mechanism of formation of internal electric field within pentacene layer can be elucidated by the energy level diagram of Figure 1.The differences of the highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) for pentacene and CuPc are 0.2 eV (5.0 and 5.2 eV, resp.) and 0.5 eV (3.0 and 3.5 eV, resp.).The difference of energy level is the main cause of the internal electric field, leading to the improved efficiency.

OSC Reliability.
Although the internal electric field was formed at the interface of pentacene/CuPc to strengthen hole extraction [21], the only single ABL for pentacene layer was obviously unhelpful and innocuous to enhance the reliability of Device B. To further enhance the cell performance, we incorporated a pentacene layer between the MoO 3 and the CuPc layer for accomplishing double ABL idea to increase hole extracting ability.Besides, we also completed Devices C and D to discuss the sequence problem of both MoO 3  and pentacene with optimized thickness in Table 3.In order to obtain reliable and meaningful experimental results for comparison studies, the lifetime data of all devices over 10 days was observed in Figure 4. Device reliability has a significant status for photoelectric application.The organic-electrode interfaces commonly used in OSCs are inherently unstable and they exhibit photoreduced chemical changes in some studies by using X-ray photoelectron spectroscopy (XPS) measurement [22][23][24].The efficiency and reliability are also influenced by the size of the active area [25].In general, enlarging the active area leads to lower efficiency but higher reliability.This suggested that the degradation is also influenced by air diffused from peripheral edges for our devices.It encourages us to gain a clearer understanding of the basic role of hole extraction layer and to elucidate the phenomena involved in the interfacial degradation mechanism.The - characteristics of all devices were measured in the same condition after encapsulation (fixed water and oxygen).The degradation of parameters of Devices C and D was plotted in Figure 4.The preferable reliability of Device C with the 5 nm MoO 3 and 6 nm pentacene layer was also obtained in Table 3.Without ABL, the lifetime of standard CuPc/C 60 cells is only one day.And the relative parameters were decreased from 3.41 mA/cm 2 to 2.12 mA/cm 2 for  sc , 0.54 to 0.37 for FF, and 0.72% to 0.36% for PCE.In Table 3, Device C, using the double ABL (MoO 3 /pentacene), displays better excellent reliability than that of Device D (pentacene/MoO 3 ).Device D has lower  sc on any thickness of double ABL, comparing to Device C in Table 3, because the thickness of MoO 3 was too thick to transfer hole to anode fluently.Rapid degradation and significant decreases in  sc and  oc were also detected.It is clear that the devices are very unstable with single ABL.Within 24 hours, the PCE values of the standard device without ABL reduced to 0.36% because of a strong decrease in  sc ,  oc , and FF.In contrast, about 5 days later, the PCE of Device B decreases from 0.67% to 0.51% and the PCE of Device C still remains at about 0.786%.Although Device C has slightly increased from 0.78% to 0.81% after 2 days, it has been proved again that the parameters of cell would rise a little [26].If device lifetime is defined as the degradation time of PCE from PCE 0 to half of it, the lifetime for Devices A to D can be estimated to be 3 days, 8 days, 10 days, and less than 3 hours, respectively.It is remarkable that the reliability of device of OSCs based on CuPc/C 60 can be ameliorated from 1 day to 10 days by using the scarce double anode buffer layer.The reason may be attributed to the pentacene layer that can separate moisture from ITO anode and the MoO 3 layer can protect organic material from oxidized reaction.The holes can cross MoO 3 more smoothly by multiple tunneling steps between the gap states introduced by oxygen vacancies [27].At the same time, the difference of energy level can also form the internal electric field at the interface of pentacene/CuPc [21].It could become another reason why the reliability of Device C is better than that of D. Device D is a good idea because it has matched the energy level to transport carriers for each other.However, the performance of Device D has not only poor efficiency but weak reliability with different pentacene thicknesses in Table 3.So, the advantage of small molecule OSCs based on CuPc/C 60 was accomplished by depositing the double ABL, MoO 3 , and pentacene layer on the ITO-glass substrate sequentially, resulting in the improved device lifetime from 1 to 10 days.

Conclusion
In this work, we have observed the effect of each and double ABL on OSCs, which is based on CuPc/C 60 structure, by using MoO 3 and pentacene material.The performance of devices with different ABL thicknesses has been investigated.The excellent reliability and best performance of device were achieved simultaneously at 5 nm MoO 3 and 6 nm pentacene, which are attributed to reaching the water and oxygen resistance property to maintain the carrier collection efficiency in device.And the device lifetime extends to 10 days.Besides, gaining the highest  oc of device not only needs the deepest HOMO level of donor but also needs the highest work function of anode.The  oc of device by using ABL is better than that of device using alone layer or without it.As a result, the  oc is greatly improved from 0.36 V to 0.41 V by inserting the MoO 3 and pentacene layer sequentially between the ITO and CuPc layer, leading to a steady state in efficiency of device over 10 days.It is thus hinted that the presence of electrode buffer layer would be important in reliability and stability of OSCs because of the movement of generated exciton on organic-electrode interface.

Figure 1 :
Figure 1: Energy level diagram for the constituent materials in Devices A to D.

3. 1 .Figure 2 :
Figures 1(a) and 1(b) show the energy levels of the OSC and the device structure.The difference between the

Figure 3 :
Figure 3: The optical absorption spectra of constituent materials in Devices A to D (insert diagram shows the contribution for pentacene layer from 660 nm to 680 nm).

Figure 4 :
Figure 4: The parameter, (a) PCE, (b)  sc , (c)  oc , and (d) FF, as a function of time for Devices C and D.

Table 2
, which is attributed to   with thicker MoO 3 layer.Both of the surface morphology and the   decreased as the MoO 3 layer thickness increased and it also responds to the decreased PCE.Second, for pentacene material, we replaced MoO 3 to enhance the device performance.The performance parameters of device with different thicknesses of pentacene layer are summarized in

Table 1 .
According to the downward trend of both  sc and PCE in Table2, t, it was found that MoO 3 does not have good reliability for conventional OSCs (CuPc/C 60 ) device although the ITO roughness has been improved.At the same time, we also discovered that the device cannot be enhanced intensively on PCE with the increase of the thickness of pentacene monolayer in Table