3-1 Introduction
In this chapter, we discuss our study to eliminate the degradation caused by the unstable interface between ITO and PEDOT: PSS. As mentioned in section 1-4-2, acidic PEDOT: PSS may etch the In2O3 in ITO substrate, followed by the diffusion of indium.
The effects of the chemical reaction of ITO with the PEDOT: PSS layer have been reported. Kawano et al. reported that with water absorption into the PEDOT:
PSS, the resistivity of the PEDOT: PSS/blend layer interface increased due to ITO reaction with PEDOT: PSS.35 Additionally Chen et al. showed that a decrease of ionization potential (IP) of PEDOT: PSS was detected due to dedoping of the PEDOT resulting from a reaction of PSS with ITO in the presence of water.52
The effects of the ITO/PEDOT: PSS reaction have been reported for polymer light emitting diode, which showed lower brightness and current density as a consequence of such reaction. However, the effect of this reaction on solar cells has not been studied. Thus, it was our objective to address this unexplored issue..
Several approaches have been developed to prevent the occurrence of the ITO/PEDOT: PSS reaction. One approach is to add glycidoxypropyl trimethoxysilane
41
(GPS) to PEDOT followed by annealing, upon which the PEDOT film becomes a crosslinked network,53 which prevents indium diffusion from the ITO electrode.
Another approach is to treat the ITO surface with compounds such as (NH4)2Sx54 and self-assembly monolayer,55which can diminish the reactivity of ITO/PEDOT: PSS interface. However, the practicalities of these approaches are limited by their complicated process, poor uniformity, and inadequate suface coverage.
For this reason, we selected atomic layer deposition (ALD) to form a blocking layer. ALD deposits inorganic thin film through chemical reaction with alternative precursors, showed in Figure 3-1. The advantages of ALD process are: (1) Large area thickness uniformity; (2) Accurate thickness control within a monolayer; (3) Capability of depositing defect-free films; (4) 100 % step coverage. Through these benefits, we believe it has great potential to form a well protecting layer.
For the material of blocking layer, hafnium oxide (HfO2) thin film by ALDhas been reported with good hydrophobia and anticorrosive. We inserted HfO2 in P3HT:
PCBM solar cells as a blocking layer. Its well-coverage and anticorrosive have the potential against acidic PEDOT: PSS. The insert layer in PEDOT also has been reported with an improved PCE and current density in solar cells’ performance.57 Thus, we expected the ALD HfO2 inserted in structure may effectively promote the performance also retarded the degradation of solar cells.
42
Two structures introduced for studying: HBH structure (ITO/PEDOT:
PSS/HfO2/PEDOT: PSS/active layer/Ca/Al) and BH structure (ITO/ HfO2/PEDOT:
PSS/active layer/Ca/Al). Here, H is abbreviated from hole transporting layer, which is the function of PEDOT: PSS; B is the abbreviation of blocking layer.
43
Figure 3-1 Schematic illustration of an ALD growth cycle (1–4) leading to the formation of a binary oxide film of metal (○) and oxygen (●). L refers to the precursor ligand.56
44
3-2 Experimental Section
3-2-1 FabricationThe device structure is showed in Figure 3-1. The fabrication of devices was as same as section 2-2-1 except the thickness of PEDOT: PSS layer with and without HfO2 was controlled at 60nm, annealing temperature of PEDOT: PSS was at 150oC for 150min.
ALD HfO2 thin films were grown in Cambridge Nanotech Savannah 100 ALD system. Tetra-dimethyl amino hafnium (TDMAHf) and water (H2O) were selected as the precursor of HfO2. Deposition process was operated at 150oC of chamber temperature. ALD recipe of HfO2 was showed below:
Table 3- 1 The operating condition of ALD deposited HfO2. Pulse time Exposure time
[s]
The procedure of durability test was shown in Figure 3-X. Semi-cells with PEDOT: PSS and active layer were storage at 28oC and 60% relative humidity for several weeks. Before metal cathode evaporated on device, a 100oC annealing for 1hr was utilized.
45
Figure 3-2 Device structures: (A) conventional structure; (B) HBH structure;
(C) BH structure. (H: hole transporting layer; B: blocking layer)
46
47
b) a)
c)
Figure 3- 3 Storage condition and process of degradation test: a) storage; b) discharge residual water; c) evaporation of cathode.
3-2-2 Measurement
The current-density-voltage (J-V) characteristics of the devices were measured in air with a Keithley 2400 source meter under simulated AM1.5G irradiation (100 mW cm−2) from a xenon-lamp-based solar simulator (Oriel 92250A-1000);
Thickness of HfO2 was measure by Alpha-Step surfcorder (Kosaka Lab. Ltd.
Surfcorder ET-3000). Thickness of HfO2 per cycle was measure from the HfO2 films with 500 cycles, deposited on ITO substrates.
Auger Electron Spectroscopy (AES) surface profile of PEDOT was obtained by VG Scientific,Microlab 350 and operated at 3kV.The sample was in the structure of ITO/PEDOT, storing in dry box for three months.
Depth profile of x-ray photoemission spectroscopy (XPS) were measure by Thermo Scientific K-Alpha with the sputter energy of 3keV. The XPS sample was stored at 28oC and 60% relative humidity with the structure: ITO/ (HfO2, 1nm)/PEDOT: PSS (~60nm)/active layer (~100nm).
48
3-3 Results and Discussion
3-3-1 Verification of Indium Diffusion
To verify the phenomenon about indium diffusion in PEDOT: PSS layer, the surface survey of the main elements of PEDOT: PSS was detected by Auger Electron Spectroscopy (AES). In the range of electron energy from 100eV to 600eV, the components of sulfur (S) at 152eV, carbon (C) at 272eV, and oxygen (O) at 503eV are the elements what the PEDOT: PSS composed of.56 And indium (In) at 404eV is the target which we judge whether the reaction between ITO and PEDOT: PSS occurred.
After storing at 28oC and 60% relative humidity for three months, the indium signal contained in the surface of PEDOT: PSS films in Figure 3-4. The original and 1st differential data of AES indicates that indium which released from the reaction of ITO and PEDOT: PSS diffused through the entirely PEDOT: PSS to the surface.
This result corresponds with previous literatures and confirms the importance of prohibiting indium’s diffusion.
49
100 200 300 400 500 600
Figure 3-4 AES survey of top PEDOT: PSS stored in air for three months: origin data (bottom) and 1st differential data (top).
50
3-3-2 HBH Structure
Firstly, we optimized the thickness of HfO2 blocking layer. The thickness of HfO2 each cycle was 1.3Å. As shown in Figure 3-5 and Table 3-2, the best thickness of HfO2 in HBH structure was 1.9 nm (15 cycles). With the optimized thickness, the performance in HBH structure had a better performance in both Jsc and PCE than the conventional one. The PCE and Jsc were increased from 10.00 mA/cm2 and 3.34% to 11.33 mA/cm2 and 3.46% respectively. This enhancement of Jsc and PCE was due to the promoted inner electrical field, which has been reported in Al2O3 formed HBH structure.57 Tsai et al. indicate that upper PEDOT and original one in HBH structure have different work functions of 5.6 eV and 5.43eV. Larger difference of work function between cathode and anode can provide a higher electrical filed, assisting in carriers transporting for higher current density.
As an insulator, thicker HfO2 over 2nm (15 cycles of ALD) prevented carrier transporting via tunneling. As shown in Figure 3-5 and Table 3-2, device performance deteriorated severely into a quite low PCE of 0.39% and Jsc of 3.15 mA/cm2. Thus we chose the critical thickness – 1.9 nm of HfO2 (15 cycles by ALD) as our thickness of blocking layer in HBH structure, expected its advantages of enhanced current density and good protection.
To clarify the effect of out-diffused indium from ITO to the performance of solar
51
cells, devices without metal cathode stored at 28oC and 60% relative humidity to allow the chemical reaction between ITO and PEDOT: PSS to occur. Other decay-causing factors such as light and residual water absorbed in storage were prevented by storing in dark and annealing after storage respectively.
The storing test is showed in Figure 3-6 and Table 3-3. Degradation of conventional devices without HfO2 had a 12.6% degraded PCE from 3.72% to 3.25%
after 3 weeks storage. Both Jsc and Voc decreased in the range of 10%. We interpreted the descents of conventional cells resulted from the unstable interface of PEDOT: PSS and other unavoidable factors: surface contamination, morphology change. We expected that the HBH cells we introduced might efficiently obstruct indiums across through the HfO2.
However, unfortunately HBH cells had a 37.6% degraded PCE from 3.76% to 2.32%, which was worse than that of conventional devices (Figure 3-7 and Table 3-3).
Based on all constant controls, the worse performance of HBH cells, we explained, was attributed to the destruction of HfO2 by the swelling of PEDOT: PSS. As we know, the swelling of polymer may increase its surface roughness,58 which is easily higher than what the inorganic thin film can bear. Also, the roughness and film quality of PEDOT: PSS are larger responsible for its transporting function.59 Instead of serving as a protection, the crumbled HfO2 layer worsened the film quality and thus
52
decreased the FF and Jsc. Moreover, the original increase current density by larger electrical field was diminished by this destruction HfO2.
Although we obtained worse-durability device with HBH structure, we still placed high hope on the protection of ALD HfO2. Compare to grow films on soft polymer of PEDOT: PSS, ALD HfO2 deposited on solid ITO is more firm and stable.
Thus, the structure inserting HfO2 between ITO and PEDOT: PSS was introduced. We expected it can prevent the occurrence of the chemical reaction between ITO and PEDOT: PSS positively.
53
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Figure 3-5 The current-voltage characteristic for conventional cells and HBH structure with different thickness HfO2.
Table 3- 2 The device performance for conventional cells and HBH structure with different thickness HfO2.
HBH structure
JSC VOC FF PCE
Thickness of HfO2
[mA cm-2]
0.0 0.1 0.2 0.3 0.4 0.5 0.6
Figure 3-6 The current-voltage characteristic of conventional cells for observing the degradation in 3 weeks.
0.0 0.1 0.2 0.3 0.4 0.5 0.6
Figure 3-7 The current-voltage characteristic of HBH structure devices for observing the degradation in 3 weeks.
55
Table 3- 3 The device performance of durability test for conventional cells and HBH structure stored in 3 weeks.
conventional structure
JSC VOC FF PCE
Storage time
[mA cm-2]
[weeks] [mV] [%] [%]
0 9.64 0.61 64.0 3.76 -
1 9.76 0.59 63.8 3.67 97.6%
3 9.18 0.58 61.0 3.25 86.4%
HBH structure
JSC VOC FF PCE
Storage time
[mA cm-2]
[weeks] [mV] [%] [%]
0 10.77 0.62 55.8 3.72 -
1 9.83 0.59 52.6 3.05 82.0%
3 8.50 0.55 50.0 2.32 62.4%
56
3-3-3 BH structure
BH device was in the structure of ITO/HfO2/PEDOT: PSS/active layer/Ca/Al.
First step, we optimized the thickness of HfO2 in the architecture. As shown in Figure 3-8 and Table 3-4, the critical thickness of HfO2 in BH structure was 0.9nm (7 cycles of ALD). As the same mechanism mentioned in section 3-3-2, both Jsc and PCE in BH cells were improved.
The protection of blocking layer in BH structure also was examined by durability test shown in Figure 3-9 and Table 3-5. During 5 weeks storage at 28oC and 60%
relative humidity, conventional device degraded 16.0% in PCE and 15.0% in Jsc, which was proportional degraded with the data in 3-3-2. Compare to conventional devices, BH cells had a slower degrade rate with 10% decay. The variation of 9.0%
decay between BH and conventional cells came from the unstable interface releasing indium ions. (Conventional structure: 84.0% and BH structure: 90.1%) We supposed the achievement attributed to the protection of ALD HfO2 which was effective against
acidic PEDOT :PSS etching ITO substrate.
To evidence the function of HfO2 blocking layer, we observed the concentration distribution of ITO/ (HfO2)/PEDOT/P3HT: PCBM thin films by x-ray photoemission spectroscopy (XPS). The elements we detected were C, O, S, which are the main elements of PEDOT and active layer and In, Tin (Sn) which constructed ITO. The
57
signal of Sn is as the demarcation line of ITO substrate. As shown of XPS in Figure 3-10a and 3-10b, after 2 weeks storage, these two figures seemed to resemble in the concentration distribution and hard to distinguish the interface of ITO and PEDOT layer. However, based on the distribution of Sn (Figure 3-10c), structure with an HfO2
protection had a larger indium slope of 0.147 than the conventional one with a 0.122 slope. We presume the minor slope means a more considerable diffusion occurred in concentration profile. In other words, BH structure device had a shorter diffusion distance of indium, contributed to the protection of HfO2.
Thus, the function of HfO2 blocking layer was established as a blocking layer protecting the unstable interface meanwhile retarding the degradation by indium.
More than 9% decay can be diminished by this ALD HfO2 blocking layer in 5weeks storage.
58
59
Table 3- 4 The device performance for conventional cells and BH structure with different thickness HfO2.
BH structure Thickness of HfO2
[nm]
Figure 3-8 The current-voltage characteristic for conventional cells and BH structure with different thickness HfO2.
0 1 2 3 4 5 6
Figure 3-9 The decay plot of PCE for different structure devices in 5 weeks storage:
conventional structure (square), HBH structure (circle) and BH structure (triangle).
Table 3- 5 The device performance of durability test for conventional cells and BH structure stored for 5 weeks.
conventional structure
0 200 400 600 800 P3HT:PCBM/ PEDOT ITO
BH structure
Figure 3- 10 XPS depth profile of two structures: a) conventional structure; b) BH structure; c) the comparison of Indium distribution.
61
3-4 Summary
In this chapter, we confirmed the existence of indium in PEDOT: PSS layer during a long-term storage. Also, the effect from this unstable interface was demonstrated by the gradual decay of conventional device. The concept of blocking layer was examined for reducing the damage caused by indium to device performance.
We tested ALD HfO2 layer as the blocking layer with two device structure: HBH structure and BH structure. The HBH structure provided a more unstable performance.
We interpret the decay was resulted from the swelling PEDOT destroyed the 2nm HfO2. Far from a protection, crumbled HfO2results in a poor film-quality, which may decreased the performance of PEDOT: PSS.
The other architecture - BH structure had a better performance in long term storage. It was attributed in the anticorrosive HfO2, effectively prevented the chemical reaction between unstable ITO and PEDOT: PSS. The effects of HfO2 blocking layer also could be proved by XPS depth profile that BH structure had a shorter diffusion distance of indium.
62