Challuri Vijay Kumara,
Georgia Sfyriab,
Dimitrios Raptisac,
Elias Stathatosd and
Panagiotis Lianos*ac
aFORTH/ICE-HT, P.O. Box 1414, 26504 Patras, Greece. E-mail: lianos@upatras.gr; Fax: +30 2610 997803; Tel: +30 2610 997513
bPhysics Department, University of Patras, 26500 Patras, Greece
cDepartment of Chemical Engineering, University of Patras, 26500 Patras, Greece
dElectrical Engineering Department, Technological-Educational Institute of Western Greece, 26334 Patras, Greece
First published on 5th December 2014
Low cost copper phthalocyanine has been used as hole-transporting material for the construction of organolead halide solid state perovskite solar cells. The cells were assembled and tested under ambient conditions. They achieved a power conversion efficiency of 5.0% using copper phthalocyanine, which appears to have potential to replace the currently used organic hole transporters. The present work has also examined the possibility of upscaling by construction of small cell modules.
Organometal halide perovskites have attracted a strong scientific interest as components of solar cells due to their excellent optical and conductivity properties. Their chemical structure can be represented by AMX3, where A is usually an organic ammonium cation, M a metal and X is a halide anion.12 By making unprecedentedly fast steps, researchers of organometal halide perovskite solar cells have managed to reach solar cell efficiencies that overpassed traditional photovoltaic solar cells. Thus in 2009, Kojima et al.13 published a 3.8% power conversion efficiency using perovskites as the sensitizing material for DSSCs, while Im et al.14 obtained 6.5% efficiency with the same type of materials. In 2012, an inorganic perovskite, CsSnI3, was used by Chung et al.15 as a hole conductor for solid state DSSCs. By employing CH3NH3Pbl2Br perovskite and spiro-OMeTAD as the hole transporting material Kim et al.16 reached an efficiency of 9.7% while Heo et al.8 reported 12% efficiency using CH3NH3PbI3, TiO2 and a polymeric hole conductor. Snaith and co-workers reduced the processing temperature of the bulk absorber layer in CH3NH3PbI3−xClx perovskite solar cells and achieved power conversion efficiencies of up to 12.3%.17 Gratzel and coworkers recently reported 15% efficiency with CH3NH3PbI3 perovskites and spiro-OMeTAD using a two-step deposition method18 while Snaith and coworkers with vacuum deposited CH3NH3PbClxI3−x reported efficiency of 15.4%.19 Gratzel, Park and coworkers reported champion efficiency of 17.01% with two step spin coating procedure by size-controlled growth of CH3NH3PbI3 cuboids.20 Finally, a world record efficiency of 19.3% has been achieved by Y. Yang and co-workers for the system ITO-polyethyleneimine/Y:TiO2/CH3NH3PbI3−xClx/spiro-OMeTAD/Au.21
Concerning hole transporting materials, molecular Spiro-OMETAD,22 polymeric transporters8,23 and inorganic materials24,25 have been used to make solid state solar cells including perovskite solar cells. Most of the hole transporting materials do not show absorption in the Visible and NIR region. Some of the polymers, like polythiophene, do absorb visible light and may participate in the sensitization process, even though, in the presence of organometal halide perovskites this type of sensitization is of minor importance. The main problem with these materials is their high cost, which dictates a need for the search of alternative hole conductors. Inorganic hole conductors may offer inexpensive alternatives and they have been recently studied with satisfactory results.24,25 In the present work, a low cost and easy to manipulate molecular hole conductor, copper phthalocyanine (CuPc) has been studied as a component of organometal halide perovskite solar cells. CuPc is, thermally and chemically, very stable. CuPc has already been used as a p-type semiconductor and attracted attention because of its flexibility, color, stability, semiconducting property and cost-effective applications.26–28
The molecular structure of copper phthalocyanine is shown in Fig. 1. Its visible absorption lies between 500 to 800 nm (Fig. 2) and for this reason it has a blue color. CuPc possesses a large π-conjugated system that makes it difficult to dissolve in most commonly used organic solvents. Thus it is rarely studied by traditional spin coating techniques. CuPc has been used as material in vacuum deposited OPV devices due to high absorption coefficient and long exciton diffusion length.29–32 The CuPc/C60 combination is among the most common small molecule organic solar cells and has exhibited satisfactory performance.33
![]() | ||
Fig. 1 Molecular structure of Cu phthalocyanine and device layer structure of the perovskite solar cell. |
![]() | ||
Fig. 2 UV-vis absorption spectra of (1) CuPc thin film and (2) TiO2/CH3NH3PbClxI3−x film (after subtracting absorption due to titania alone). |
In the present work CuPc has been deposited by vacuum deposition. It has been used as hole-transporting material to make small laboratory size solar cells but also to construct small perovskite solar cell modules. One serious drawback of organometal halide perovskites is their vulnerability to humidity that necessitates synthesis under dry conditions and encapsulation for further protection from humidity. For sure, this fact will increase cost when perovskite solar cells make it to the market. In addition, the reported efficiencies refer to tiny cells of only a few square millimeters of active surface. It is well known that the efficiency drops when the size of the cell increases. It is then necessary not only to race for higher efficiencies but also to appreciate the difficulties of upscaling. For this reason in the present work, we have studied perovskite solar cells, which were partly made under ambient conditions and under a limited upscaling. We believe that this effort will set lower limit conditions and define realistic terms for the construction of perovskite solar cells.
![]() | ||
Fig. 3 Energy level diagram of perovskite solar cells with CuPc hole conductor.35,36 |
![]() | ||
Fig. 4 FESEM image of a cross section of the TiO2/CH3NH3PbClxI3−x/CuPc/Au solar cell. The scale bar is 200 nm. Pkt stands for perovskite. |
The composition of the CH3NH3PbClxI3−x perovskite deposited under the present conditions was examined by using EDX spectra. A characteristic example is presented by Fig. 5. Based on these data, the closest approximate structure of the presently obtained perovskite sensitizer can be represented by the following formula: CH3NH3PbClI2. This means that the spontaneous uptake of chlorine under the present conditions is smaller than the one present in the precursor solution (see Experimental section). Indeed, the atomic ratio of Pb:
Cl
:
I in the precursor solution was 1
:
2
:
3 but the corresponding ratio detected by EDX was 1
:
1.1
:
2.2. Obviously, the majority of the presently formed perovskites involve one chlorine and two iodine atoms.
![]() | ||
Fig. 5 Energy-dispersive X-ray spectroscopy analysis (EDX) of a sensitized photoanode. Atomic percent: (O)50.7; (Ti)25.1; (Pb)3.6; (Cl)4.1; and (I)7.9. |
The active components of the cell have been also characterized by XRD at various materials combinations, as seen in Fig. 6. The lower XRD spectrum gives the usual peaks expected for pure titania films,37 however, the prevailing lines belong to the FTO substrate. This is because the presently used titania film is very thin and gives small contribution. When the perovskite was added, clearly new and well defined peaks showed up, indicating that CH3NH3PbClxI3−x obtains a crystalline structure.38 On the contrary, addition of CuPc did not produce any new peaks. For this reason, CuPc was also deposited alone on a titania film also by vacuum evaporation and again did not produce any new lines. We then concluded that this material is amorphous. A very broad peak appearing around 14° may indicate some limited small nanoparticle formation but apparently this is not the rule.
The current–voltage (J–V) characteristics of the small unit solar cells measured under simulated solar radiation are shown in Fig. 7, while the corresponding photovoltaic parameters are summarized in Table 1. The best performing cell demonstrated a short circuit current density (Jsc) of 16.3 mA cm−2, an open circuit voltage (Voc) of 0.75 V and a fill factor of 40% leading to a PCE of 5%. Table 1 also shows average values obtained over several cells made under the same conditions. In general, cells gave a satisfactory open-circuit voltage, justified by the relatively large difference between the two charge acceptor levels, i.e. the conduction band of titania and the LUMO level of CuPc.35 This difference is 1.2 Volts (Fig. 3) indicating that there is still room for improvement. The current density was relatively large, as it is usually obtained with CH3NH3PbClxI3−x perovskites, which absorb light in the whole visible range and into the near IR. Indeed, as seen by the IPCE spectrum and the cell absorption spectrum shown in Fig. 8, the cell is active over the whole visible and NIR and this justifies high currents. A small but well distinguished peak of the IPCE spectrum in the NIR indicates some contribution to sensitization by CuPc, in addition to its main role as hole transporter. However, the major sensitizer is perovskite (cf. Fig. 2). The fill factor of the cell was relatively small and this is the main reason for the rather limited cell efficiency, compared to the efficiencies obtained with perovskite solar cells made with spiro-OMeTAD as hole transporter. Apparently the cell has a low shunt resistance39 and this must be eventually improved. Of course, the present data were recorded under ambient conditions without encapsulation while perovskite was synthesized also under ambient conditions of mild humidity. This adds to limitations in cell performance but it also offers a way to set lower limits for cell fabrication conditions.
![]() | ||
Fig. 7 Current–voltage characteristics of the solar cells with Cu phthalocyanine as the hole transporting material measured under ∼100 mW cm−2 simulated solar radiation. |
Cell | Voc (V) | Jsc (mA cm−2) | FF (%) | η (%) |
---|---|---|---|---|
Most efficient cell | 0.75 | 16.3 | 40 | 5.0 |
Maximum current cell | 0.57 | 19.9 | 39 | 4.4 |
Average over several cells | 0.71 | 14.7 | 35 | 3.7 |
Unit cells were combined to form cell modules, as seen in Fig. 9. Fig. 9 also shows the architecture of cell combinations. 6 unit cells were combined in parallel and two rows were combined in series. The photovoltaic characteristics of these cells are summarized in Table 2. The total current produced by each row of 6 cells was 8.9 mA. The actual current of each unit cell of Table 1 was 2.4 mA. Therefore, 8.9 mA are substantially smaller than the expected 6 × 2.4 = 14.4 mA. This loss of current is, however, expected and is observed when upscaling is attempted on small unit cells. The lower current is due to several reasons: overestimation of the current produced by a unit cell, which is usually larger than the area defined by the electrodes; small conductivity of the FTO electrodes; lateral diffusion of current and loss by trapping, etc. In the present case a loss of voltage and a drop of the fill factor were also observed by parallel connection, which is most probably due to fabrication deficiencies under the present conditions. On the other hand, connection of two rows in series did give a voltage multiplication and current preservation. Overall, the results are encouraging and it is obvious that interesting photovoltaic characteristics can be obtained by these small cell modules.
![]() | ||
Fig. 9 Photograph of cell module comprising 6 unit cells connected in parallel with two rows in series and schematic representation of device architecture. |
Device description | Voc (Volts) | Isc (mA) | FF (%) |
---|---|---|---|
6 unit cells connected in parallel and two modules in series | 1.08 | 8.8 | 28 |
6 unit cells connected in parallel in first module | 0.61 | 8.9 | 28 |
This journal is © The Royal Society of Chemistry 2015 |