Solar Energy Materials & Solar Cells 144 (2016) 273–280
Contents lists available at ScienceDirect
Solar Energy Materials & Solar Cells journal homepage: www.elsevier.com/locate/solmat
Light intensity dependence of External Quantum Efficiency of fresh and degraded organic photovoltaics E.A. Katz a,b,n, A. Mescheloff a, I. Visoly-Fisher a,b, Y. Galagan c a Department of Solar Energy and Environmental Physics, Swiss Institute for Dryland Environmental and Energy Research, J. Blaustein Institutes for Desert Research, Ben-Gurion University of the Negev, Sede Boqer Campus, 84990, Israel b Ilse Katz Inst. of Nano-Science and Technology, Ben-Gurion University of the Negev, Be’er Sheva 84105, Israel c Holst Centre, PO Box 8550, 5605KN Eindhoven, The Netherlands
art ic l e i nf o
a b s t r a c t
Article history: Received 30 June 2015 Received in revised form 8 September 2015 Accepted 14 September 2015
The effect of light intensity on the External Quantum Efficiency (EQE) of encapsulated bulk heterojunction organic photovoltaics (OPV) is presented. The measurements were applied to devices based on poly(3-hexylthiophene) (P3HT) blended with the fullerene derivative phenylC61-butyric acid methyl ester (PCBM) in as-produced and various degradation states. The degradation of current collection in the OPV devices is shown to enhance the sub-linear dependence of the short-circuit current on light intensity, and the corresponding EQE decrease with increasing incident light intensity. On the other hand, fresh cells and cells exposed to a low photon dose demonstrated an increase in the fullerenerelated part of the EQE with increasing light intensity, i.e. a super-linear dependence of the photocurrent in this spectral range. Generation of traps in PCBM was proposed as the underlying mechanism for this effect. Perusal of our results suggests that (1) EQE dependence on the incident light intensity should be always taken into account in measuring spectral response of fresh OPV and especially of degraded devices; (2) intensity-dependent characterization provides an insight to the degradation mechanisms of OPV and can help to separate degradation in absorption/generation from degradation of the charge collection in the cell. & 2015 Elsevier B.V. All rights reserved.
Keywords: Organic photovoltaics External Quantum Efficiency Degradation Traps Current collection
1. Introduction Efficiency of organic photovoltaics (OPV) has recently surpassed 10% [1]; however, better understanding of the physical mechanisms controlling the device performance is essential for further development. A number of protocols for accurate OPV efficiency assessment has been suggested [2–6]. It is well accepted that such assessment should include recording the irradiance spectrum and measurements of External Quantum Efficiency (EQE) of a reference cell and the tested OPV cell [7,8]. EQE is a dimensionless parameter quantifying the number of electron–hole pairs collected at the device contacts per incident photon at a given wavelength. The Short-Circuit Current Density (JSC) can be calculated as the integral of the product of the measured EQE and the number of incident photons NP at each wavelength λ, typically deduced from the AM1.5G spectrum [9], over the irradiating n Corresponding author at: Department of Solar Energy and Environmental Physics, Swiss Institute for Dryland Environmental and Energy Re search, J. Blaustein Institutes for Desert Research, Ben-Gurion University of the Negev, Sede Boqer Campus, 84990 Midershet Ben-Gurion, Israel. E-mail address:
[email protected] (E.A. Katz).
http://dx.doi.org/10.1016/j.solmat.2015.09.020 0927-0248/& 2015 Elsevier B.V. All rights reserved.
wavelength range: Z J SC ¼ qEQE λ N P λ dλ
ð1Þ
where q is the elementary charge. Thus, a verification of photovoltaic measurements can be assumed when the current density calculated by (1) agrees with the measured JSC. In a recent commentary in Nature Photonics [7] Zimmermann et al. reviewed 1262 papers reporting current–voltage (I–V) curves of OPV, of which only 375 papers reported EQE. Comparison of integrated EQE data with the reported JSC values demonstrated that 37% of the publications overestimated JSC. Significant differences between these values can occur in OPV devices in which the photocurrent generation and collection depend non-linearly on the incident light intensity. Therefore, light intensity dependence of EQE is very informative for studying the operating mechanisms in OPV devices, and intensity-dependent losses should be taken into account. The published data for such behavior in OPV are very limited and restricted only to the case of linear or sub-linear dependence of JSC on light intensity [10–13]. In polymer–fullerene bulk heterojuction OPV cells [14] light is absorbed by the polymer/fullerene photoactive layer resulting in the formation of a neutral excited state (binding energy 0.5 eV)
274
E.A. Katz et al. / Solar Energy Materials & Solar Cells 144 (2016) 273–280
fullerene derivative phenylC61-butyric acid methyl ester (PCBM) in as-produced state and at the various states of accelerated degradation. To the best of our knowledge, published data on evolution of light intensity dependence of the EQE with the OPV degradation are absent. The degradation experiments were performed using concentrated sunlight [19–20]. We demonstrate that the degradation of charge collection in the OPV devices enhances the sub-linear dependence of JSC on light intensity and the corresponding EQE decrease with increasing incident light intensity. On the other hand, both fresh cells and cells exposed to a low photon dose demonstrated an increase in the fullerene-related EQE signal with increasing light intensity, i.e. a super-linear behavior of the photocurrent in this spectral range. Generation of traps in PCBM was proposed as the underlying mechanism for this effect. Perusal of our results suggests that (1) EQE dependence on the incident light intensity should be always taken into account in measuring the spectral response of fresh OPV and especially of degraded devices; (2) it provides an insight to the degradation mechanisms of OPV which cannot be deduced from traditional I–V tracing and constant-intensity EQE measurements.
2. Experimental 2.1. Sample preparation
Fig. 1. (a) Schematic energy diagram of an OPV cell of normal architecture (including the active layer, a hole transport layer and the contacts) illustrating the five processes whose multiplied efficiencies determine the EQE: (i) absorption, (ii) exciton generation, (iii) exciton diffusion, (iv) charge separation, and (v) charge collection. (b) Absorption spectrum of a P3HT:PCBM blend and EQE spectrum for a cell with the same active layer.
[15]. Free carriers can be generated by exciton dissociation at the donor–acceptor interface, leaving the electron on the fullerene acceptor and the hole on the conjugated polymer donor. Efficient photocurrent generation requires that the donor and acceptor materials form interpenetrating and continuous networks, “phase separated” on the scale of the exciton diffusion length ( 10 nm [16]). Following exciton dissociation into free carriers, the electrons and holes are conducted through the respective moieties toward the respective transport layers and electrodes. Accordingly, the EQE may be divided into five sub-efficiencies that correspond to the stages of the charge generation and collection in bulkhetrojunction OPV cells (Fig. 1a): EQEðλÞ ¼ ηAb ðλÞ ηGe ðλÞ ηED ðλÞ ηCS ðλÞ ηCC ðλÞ
ð2Þ
where ηAb is the photon absorption efficiency, ηGe is the efficiency of exciton generation, ηED is the efficiency of exciton diffusion towards the donor/acceptor interface, ηCS is the efficiency of the exciton separation and ηCC is the charge collection efficiency. Recently EQE measurements have been suggested as an important supplementation to I–V measurements for OPV stability assessment [17,18]. Since each absorbing material in the OPV device can be identified by corresponding peaks in the light absorption and EQE spectra (Fig. 1b), analysis of EQE evolution can be a useful tool for tracking different degradation paths observed in OPV materials and devices. In this paper we present experimental results of the effect of light intensity on the EQE of encapsulated bulk heterojunction OPV based on poly(3-hexylthiophene) (P3HT) blended with the
A series of ITO-free solar cells were prepared with the following layer sequence: glass substrate/Ag-grids/PEDOT:PSS/P3HT:PCBM/ LiF:Al/encapsulation (Fig. S1 in Supplemental information) [21–23]. The Ag-grids had a hexagonal (honeycomb) configuration (Fig. S1b). The materials used for device preparation were highly conducting Orgacon PEDOT:PSS (Agfa-Gevaert), Poly-(3-hexylthiophene) (P3HT) (Plextronics) and [6, 6]-phenyl-C61-butyric acid methyl ester (PCBM) (Solenne BV). The active layer (220 nm thick) was spin-casted (1000 rpm) in a glove-box. All devices were encapsulated with a thin film barrier with low oxygen and water vapor transition rates (10 6 g m 2day 1) [24]. To prevent side leakage, the organic layers were structured via laser ablation prior to encapsulation. The active area of the devices was 3.76 cm2. Samples with the layer sequence glass/P3HT:PCBM/encapsulation layer were used for studying UV–vis light absorption of the P3HT:PCBM active layer. The thickness and composition of the P3HT:PCBM and encapsulation layers were similar to those in the solar cells. 2.2. Accelerated degradation experiments Outdoor sunlight was concentrated and transferred indoors using the unique “solar furnace” developed in-house [25,26]. In this system a dual-axis tracking flat heliostat reflects sunlight into the laboratory, where a flat mirror (with a hole at its center) tilted at 45° redirects the light upward to a 526 mm-diameter paraboloidal dish with numerical aperture of 0.4, whose focal plane is just below the tilted mirror (Fig. S2). The light intensity was moderated by a louvered shutter between the heliostat and the flat indoor mirror. It should be noted that the spectrum measured at “noon time 72–3 h” at Sede Boqer (Lat. 30.8°N, Lon. 34.8°E, Alt. 475 m), where the lab is located, is very close to the AM 1.5G spectrum. Flux uniformity was achieved using kaleidoscopes placed between the paraboloid focal point and the cell (Figs. S2c and S3). Increase in the kaleidoscope exit area results in deconcentration of the sunlight delivered to the cell. The degree of concentration can be varied gradually from 0 to, for example, 100 and 10,000 sun for 1 cm2 and 1 mm2 cells, respectively. The incident power of concentrated sunlight was measured with a pyrometer of 5% accuracy,
E.A. Katz et al. / Solar Energy Materials & Solar Cells 144 (2016) 273–280
2.3. Current–voltage measurements Measurements of I–V curves were performed with uniform illumination of the cell by concentrated natural sunlight during I–V tracing only ( o1 s) to avoid excessive degradation and temperature variations. I–V measurements were limited to clear-sky periods at “noon time 7 2 h”. The light spectrum on the cell was nearly invariant and close to the AM1.5.
EQE of the OPV tested cells (EQEtested) was calculated by measurement of Isc of the tested cell and a reference device (with a known EQE spectrum) at every wavelength (λ) as: EQEtested ¼
I tested ðλÞ U EQEref erence I ref erence ðλÞ
ð3Þ
Measuring EQE with lock-in detection enables applying a white light bias during the measurements. As mentioned above, measuring EQE with light bias with intensity equal to 1 sun is ideal since it mimics the realistic working conditions of the cell. In our EQE measurements, the light bias intensity was varied from 0 to 2 sun. This value was estimated by comparison of the Isc measured under light bias to that measured under a known solar illumination intensity. The intensity of the monochromatic modulated light was controlled by varying the voltage supplied to the halogen lamp (Vsuppl). Fig. S5 illustrates three different light power profiles of the monochromatic beam for Vsuppl ¼ 7, 9 and 12 V.
3. Results and discussion 3.1. Evolution of UV–vis light absorption spectrum of the P3HT– PCBM photoactive layer Encapsulated P3HT:PCBM/glass samples were exposed to concentrated sunlight at various light intensities (up to 4800 sun, 1 sun ¼100 mW/cm2) and exposure doses (up to 3600 sun h) [19]. Examples of UV–vis light absorption spectra measured after exposure to various doses are shown in Figs. 2a and S6. Fig. S7 demonstrates the absorption coefficient spectrum (cm 1) of the fresh sample. Fig. 2a depicts the results of exposure to the maximum dose (3600 sun h). This dose was achieved after merely 12 h of illumination by concentrated sunlight with intensity of 300 sun 0.9 0.8
Absorption [a.u.]
and the sunlight concentration was calculated taking into account the cell illuminated area. Sunlight exposure dose was quantified in units of sun*hours (1 sun h ¼360 J/cm2). The “Solar furnace” can operate in continuous-irradiation or flash-like modes. The latter is achieved by inserting a reflective rotating disk (chopper) of diameter 264 mm with a 30 50 mm2 aperture close to its circumference for sample illumination, positioned just above the focal plane (Figs. S2c and S4a). The disk's rotation frequency can be varied up to 25 71 Hz, and the shortest temporal window for constant sample irradiance is 1.1 ms. The light power Pin during “light on” periods is equal to that during continuous-illumination. According to the geometry of the chopper disk, the sample is irradiated during 5.4% of the cycle time (Fig. S4b). It means that if a sample was illuminated by chopped light for X min, the overall experiment duration was 18.5X min (see also Table S1 in Supplemental information). The sample temperature was measured by a thermocouple (T type) connected to the rear side of the OPV cells, using silver paste. Modulation of concentrated light by a chopper was found to suppress overheating of the cell by concentrated sunlight by means of effective dissipation of the heat, accumulated during the “light on” period, during the non-irradiated period (for further details see [20]). This, together with thermal bonding to the top of a thermoelectric cooled plate, allowed performing the photodegradation experiments with very high sunlight concentration at almost constant temperature (for example, at 4800 sun the measured temperature was only 30 °C higher than the temperature set by the thermoelectric controller) [20]. This approach allows performing accelerated degradation experiments with independently controlled temperature and illumination intensity. In further research it can constitute a basis for the development of a scaling relation between the degradation under 1 sun and under concentrated sunlight via rapid systematic determination of the acceleration factors, towards cross-validated accelerated stability testing. However, such systematic study is out of the scope of the present study and is irrelevant for the main messages of our paper. During exposure to concentrated sunlight the cells were kept at open-circuit conditions.
275
0.7 0.6
as produced
0.5
after 3 hours
0.4
after 6 hours
0.3 0.2
after 9 hours
0.1
after 12 hours
0 285 315 345 375 405 435 465 495 525 555 585 615 645 675 705
Wavelength [nm] 0.86
2.4. Spectroscopic measurements UV–vis absorption spectra of the P3HT:PCBM photoactive blend films and the EQE spectra of OPV cells were measured using a PV spectral response measurement system (“TECHNOEXAN” Ltd., Ioffe Physical-Technical Institute, Russia). The spectra were recorded as a function of wavelength by focusing light from a halogen lamp by means of a parabolic mirror on the input slot of a grating monochromator with automatic grating switching. In front of the slot, a chopper rotated by a synchronous motor was placed in order to use lock-in amplified detection of the signals. The overall spectral range is covered using two gratings: P1 (for the range 320– 540 nm) and P2 (400–1200 nm range). In all reported measurements the grating change between P1 and P2 occurred at 500 nm. The modulated light passes through the monochromator to an output slot equipped with a diaphragm. The chopper's frequency was set to 75 Hz.
Absorption [a.u.]
0.85 0.84 0.83 0.82 0.81 0.8 0.79 0
1000
2000
3000
Exposure dose [Sun*Hours] Fig. 2. (a) UV–vis light absorption (1 I/I0) spectra of an encapsulated P3HT:PCBM/ blend in as-produced state and after its exposure to 300 sun (dose up to 3600 sun h). The spectra were recorded for the sample as-produced and after every 3 h of the exposure. (b) Dependence of light absorption at 345 nm (PCBM-related peak) on the exposure dose.
276
E.A. Katz et al. / Solar Energy Materials & Solar Cells 144 (2016) 273–280
but it corresponds to 1.6 years of outdoor operation under natural sunlight in Sede Boqer (Negev desert, Israel) [27]. No degradation of P3HT-related absorption (e.g., the main absorption peak at 510 nm and the 554 and 605 nm shoulders, indicating ordering of P3HT domains [28]) was observed. These results demonstrate (1) the absence of oxygen/moisture penetration into the sample with this encapsulation method, and (2) the high photochemical stability of P3HT in the P3HT:PCBM blend, where pure photolysis of P3HT in the absence of oxygen occurs at a very low rate [29] and the presence of PCBM in the blend further suppresses it [30]. This observation is in accordance with previous data about long term stability of well-encapsulated P3HT:PCBM cell under exposure to 1 sun [31–33]. Slight degradation of the PCBM-related peak (345 nm) was noted ( 7% for the dose of 3600 sun h) (Fig. 2b).
Current [mA]
15 10 5 0 0
200
400
Voltage [mV]
80 as produced after 1 hour after 2 hours after 3 hours
70 60 Isc [mA]
3.2. Effect of white light bias intensity on the EQE and its evolution with sunlight exposure
cell as produced after 1 hour after 2 hours after 3 hours
20
50 40 30 20 10 0 0
1
2
3
4
5
6
Suns
0.6 0.5
EQE
Contrary to light absorption and EQE measurements, which may be performed on small areas (2 2 mm2 in our system), the I– V measurements should be carried out under illumination of the entire cell area (3.76 cm2). Analysis of the photovoltaic performance of inorganic (non-excitonic) solar cells under localized illumination is quite simple [34] while in the case of OPV it becomes more complicated due to the voltage-dependent photocurrent and the corresponding strong dependence of JSC on the illuminated area [35,36]. In our solar concentrator, the highest illumination level for the area of 3.76 cm2 is 150 sun. Accordingly, I–V characterization could be performed only following cell exposure to 150 sun or less, while EQE could be measured following any exposure level. I–V and EQE measurements of cells, employing the same P3HT: PCBM active layer, after 3 h of exposure to 20 sun (dose of 60 sun h) and 40 sun (dose of 120 sun h) did not reveal any significant degradation (not shown). This stability is in accordance with the behavior recently reported for similar ITO-free OPV cells (see [37; Figs. 2 and 3]). However, exposure to 145 sun for various times (maximum dose of 435 sun h) resulted in considerable degradation in the I–V characteristics of the cell (Fig. 3a). After the first hour of exposure to 145 sun, the fill factor (FF) degraded strongly (from 0.45 to 0.31) while Isc decreased slightly (by less than 5%). After two hours of such exposure, a substantial degradation of both FF (down to 0.25) and Isc (by more than 50%) was recorded. A slight monotonic decrease in the open-circuit voltage (VOC) during the two hours of exposure (by 5%) was noted. The overall efficiency of the cell was found to degrade from 1.4% for the fresh sample to 0.29% after two hours of the exposure. Subsequent exposure did not result in any additional degradation of the cell performance measured under 1 sun (Fig. 3a). Fig. 3b demonstrates Isc values measured under various sunlight concentrations. A sub-linear character of the Isc dependence on the light intensity was found to be more pronounced after exposure to concentrated sunlight. In conventional inorganic solar cells such behavior can be explained by a monotonic increase in the cell's series resistance (RS). Indeed, an increase in RS is known to result initially in a decrease of the FF. As the FF reaches its minimal value of 0.25, Isc starts to significantly decrease and exhibit a sub-linear dependence on the light intensity. In OPV, in addition to the series resistance limitation, a number of mechanisms can be suggested to be responsible for degradation in the charge collection efficiency, including bi-molecular recombination [13], enhancement of the voltage dependence of the photocurrent [35], space charge limitations [38–40], and unbalanced transport of holes and electrons, also known as the “m τlimited” process [39]. One cannot distinguish between these
0.4 0.3 as-produced after 1 hour after 2 hours after 3 hours
0.2 0.1 0 300
350
400
450
500
550
600
650
700
Wavelength [nm] Fig. 3. PV characterization of P3HT:PCBM-based cells following exposure to 145 sun for various times (dose up to 435 sun h): (a) I–V curves measured under 1 sun, (b) light intensity dependence of Isc, and (c) EQE spectra.
mechanisms merely by the analysis of I–V curves and dependence of the main PV parameters on the light intensity. We will therefore use the terms “charge collection degradation” or “RS limitation” to refer to this characteristic behavior. Contrary to the degradation noted by the I–V curves, EQE spectroscopy (without white light bias) revealed very minor changes after 3 h of exposure to 145 sun (Fig. 3c). This finding is in accordance with the photochemical stability of the P3HT:PCBM active layer observed even after exposure to much higher solar doses (3600 sun h), as reported above. To explain the discrepancy between the substantial (450%) degradation of Isc measured under 1 sun and the stability of the EQE spectra, we postulate that photocurrent generation in the photoactive layer was not affected by the exposure to concentrated sunlight. However, it enhanced RS losses in the device and, as a result, caused the Isc degradation. Indeed, the intensity of the monochromatic probe beam used in the EQE measurements is at least 4 orders of magnitude smaller than that used for I–V tests under 1-sun illumination (100 mW/ cm2). Therefore, effects of series resistance or other current collection losses were not detected by the EQE recording. This suggestion was confirmed by EQE measurements with light bias (Fig. 4). In particular, EQE spectra of degraded cells measured with light bias with intensity of 1 sun exhibited the expected degradation and resolved the discrepancy. Now, let us analyze the effect of light bias on the cell EQE in more details. Adding moderate light bias intensities ( 0.5 and 1 sun) to the as-produced cell was found to cause no change in the EQE values (Fig. 4a). This is consistent with linear light intensity dependence of Isc for the as-produced cell up to intensity of
0.6
0.6
0.5
0.5
0.4
0.2 0.1 400
no bias 0.5 suns bias 1 suns bias 1.14 suns bias
0.3 0.2 0.1
0 300
500
0 300
600
400
0.6
0.5
0.5
0.4
EQE
EQE
0.6
after 1 hour,no bias
0.2 0.1
after 1 hour, 1 suns bias
0.1 0 300
0 500
600
700
after 60 min,no bias
0.3 0.2
400
600
0.4
after 1 hour, 0.5 suns bias
after 1 hour, 2 suns bias
300
500
Wavelength [nm]
Wavelength [nm]
0.3
277
0.4
no bias 0.5 suns bias 1 suns bias 1.14 suns bias
0.3
EQE
EQE
E.A. Katz et al. / Solar Energy Materials & Solar Cells 144 (2016) 273–280
after 60 min, 0.5 suns bias after 60 min, 1 suns bias after 60 min, 2 suns bias
400
500 Wavelength [nm]
600
Wavelength, [nm] 0.6
0.6
0.5
0.5
0.4 EQE
EQE
0.4 0.3 0.2
after 2 hours, no bias after 2 hours, 0.5 suns bias after 2 hours, 1 sun bias after 2 hours, 2 suns bias
0.1 0 300
400
500
600
Wavelength [nm] Fig. 4. EQE spectra measured with various light bias intensities for (a) as-produced cell, and after its exposure to (b) 145 sun for 1 h and (c) 2 h.
1 sun (Fig. 3b). After passing a critical bias intensity value ( 1.14 sun) the EQE values started to decrease slightly (in accordance with the sub-linear behavior of Isc above 1 sun found even in the as-produced cell, Fig. 3b). Introducing light bias to the degraded cell after 1 h of exposure to 145 sun (Fig. 4b), and especially after 2 h (Fig. 4c), resulted in enhanced EQE deterioration even with low light bias intensity (0.5 sun). The EQE decrease with light bias intensity is equivalent to the sub-linearity of Isc. Both trends became stronger for the degraded cells due to the degradation in charge collection (increase in RS). We speculate that this can be due to degradation of the charge collecting layer(s), e.g. PEDOT:PSS. Recently a dominant role of the increase of the sheet resistance of the PEDOT:PSS hole transport layer in photodegradation of encapsulated ITO-free bulk heterojunction OPV under UV-illumination was demonstrated [41]. In our exposure experiments, the intensity of UV-light was high due to optical concentration of natural sunlight. In order to analyze the light bias effect on the EQE values associated with photogeneration of the PCBM and P3HT moieties of the photoactive layer, we normalized the spectra shown in Fig. 4 in the following manner. For every spectrum measured under light bias, we calculated the ratio of the EQE measured at λ ¼ 600 nm under light bias to that measured at this wavelength without the light bias. Then, we divided all other EQE values by this ratio. As a result, all normalized spectra measured under light bias have EQE
after 120 min, no bias
0.3
after 120 min, 0.5 suns bias
0.2
after 120 min, 1 sun bias
0.1
after 120 min, 2 suns bias
0 300
400
500 Wavelength [nm]
600
Fig. 5. Normalized EQE spectra measured with various light bias intensities for (a) as-produced cell, and after its exposure to (b) 145 sun for 1 h and (c) 2 h.
values at λ ¼ 600 nm equal to those measured at this wavelength without light bias (Fig. 5). The particular value of 600 nm was chosen as the wavelength associated with photogeneration by P3HT. The normalized spectra for the as-produced cell coincide in the entire spectral range (Fig. 5a). On the other hand, for the degraded cell (Fig. 5b and c), the normalized EQE values in the spectral range associated with the photogeneration of PCBM (around 350 nm) were found to increase with the light bias intensity. Such behavior is characteristic of super-linear dependence of the photocurrent on the light intensity. This effect was even more evident in EQE measurements with variation of the light intensity of the monochromatic beam (see Section 3.3). Normalizing the EQE spectra enabled filtering this effect from the sub-linear behavior of the Isc caused by the current collection degradation. We postulate that filling of traps in PCBM by photogenerated carriers is the mechanism responsible for the super-linear dependence of PCBMrelated photocurrents on light intensity (see Section 3.3). Fig. 6 shows the EQE evolution after local illumination of the OPV cell by 300 sun for various times (illuminated area was 0.55 cm2). 12 h of such irradiation corresponds to the dose of solar exposure of 3600 sun h. This is the highest dose used in our experiments. It should be noted that even the first period of such exposure corresponds to a dose of 900 sun h, higher than the doses used during exposure to 145 sun described above.
278
E.A. Katz et al. / Solar Energy Materials & Solar Cells 144 (2016) 273–280
0.7
0.65
0.6
0.6 as produced after 3 hours after 6 hours after 9 hours after 12 hours after 12 hours+light bias
0.4 0.3 0.2
0.55 EQE
EQE
0.5
as produced after 9.7 minutes after 16.2 minutes
0.5 0.45
0.1 0 250
0.4 300
350
400
450
500
550
600
650
700
0.35 250
Wavelength [nm]
300
350
Fig. 6. EQE spectra of the as-produced cell and after various time of its exposure to 300 sun (dose up to 3600 sun h).
400
450
500
550
600
650
700
Wavelength [nm]
EQE
0.7 0.6
0.6
0.5
0.5
0.4
EQE
0.7
0.4 0.3
0.3 0.2
Beam 7V Beam 9V Beam 12 V Beam 12V+light bias
0.2 0.1
0.1 0 250
0
Beam 7V Beam 9V Beam 12 V Beam 12V+light bias(4.5V) 350
450
550
650
Wavelength [nm] 250
350
450
550
650
Wavelength [nm] Fig. 7. EQE spectra measured with various intensities of the monochromatic modulated beam for the as-produced cell.
EQE significantly decreased in the entire spectral range after each period of the exposure. A comparison between the EQE degradation with the corresponding evolution of UV–vis light absorption spectra of the P3HT:PCBM blend (Fig. 2) reveals that the cell's photocurrent degradation does not result from the degradation of light absorption in the photoactive layer. Adding light bias to the measurement of the cell in its final degraded state resulted in a strong EQE decrease at all wavelengths, which indicates that the degradation of charge collection plays a significant role. 3.3. Effect of the intensity of monochromatic light on the EQE and its evolution with sunlight exposure Controlling the intensity of the monochromatic modulated beam during the EQE measurements was achieved by varying the power supplied to the halogen lamp. In all experiments described above this voltage was fixed as Vsuppl ¼12 V. In this section we report EQE measurements with three different values of Vsuppl (7, 9 and 12 V). An increase in Vsuppl resulted in an increase of the intensity of the monochromatic modulated beam in the entire spectral range (Fig. S5). Fig. 7 shows the effect of Vsuppl variation on the EQE spectrum for the as-produced cell. While the “red” part of the spectrum, associated with photogeneration in P3HT, was not affected by the light intensity, a significant increase of the PCBM-related EQE peak with the light intensity was noted. This corresponds to superlinear dependence of the photocurrent generated in the PCBM part of the photoactive layer on incident light intensity, as discussed
Fig. 8. EQE spectra of the cell exposed to 1850 sun: (a) spectra measured with Vsuppl ¼ 12 V as a function of the exposure time; (b) spectra measured with various intensities of the monochromatic modulated beam for the cell exposed for 16.2 min (a dose of 500 sun h).
above. Such super-linearity was previously observed in devices with high density of carrier traps in the active layer, for example in solar cells based on amorphous and nanocrystalline silicon [42–44]. At a higher light intensity, more traps become populated, resulting in reduced recombination and super-linear increase of the photocurrent. We can explain the observed effect by the presence of charge traps in the PCBM crystallites and their population by photogenerated carriers [45]. Since a similar effect, observed by the normalization of EQE spectra measured with various intensities of light bias, was found to be more significant following photo-degradation (Fig. 5), EQE measurements with variation of the monochromatic light intensity were also applied to a degraded cell. Fig. 8 shows the EQE degradation after a short chopped illumination by 1850 sun (doses up to 500 sun h), and the light-intensity dependence of the EQE spectra. The EQE spectra following exposure (Fig. 8a) show a reduction of the main P3HT absorption peak at 510 nm, while the “polymer shoulder” ( 600 nm) did not decrease and even slightly increased, probably due to improved crystallization during the exposure. The PCBM peak ( 340 nm) showed a significant degradation. This degradation is much stronger than that of the PCBM-related light absorption (Fig. 2). Hence PCBM related degradation of the OPV performance originates not from a decrease in light absorption, but rather by the degradation of its electronic properties [31], for example due to PCBM dimerization [46–48]. Perusal of the dependence of EQE of this cell in its degraded state (after exposure to a dose of 500 sun h) on the incident light intensity (Fig. 8b) suggests additional trap generation in PCBM during the sunlight exposure. Indeed, the effect of
E.A. Katz et al. / Solar Energy Materials & Solar Cells 144 (2016) 273–280
varying the beam intensity on the EQE became more pronounced, compared to the as-produced cell, and was extended to longer wavelengths. The EQE spectra of the degraded cell measured at various light intensities coincide only for λ 4 615 nm, i.e., the photocurrent super-linearity is observed practically for the entire spectral range of PCBM absorption [49]. Furthermore, the EQE spectrum of the degraded cell measured with white light bias is higher in the same spectral range than the curve measured with the same monochromatic power (12 V) without the bias light.
4. Conclusions We demonstrated that investigating the effect of intensity of monochromatic light and white light bias on the EQE spectra of OPV cells in various degraded states is a powerful tool for the analysis of OPV photodegradation. Accelerated degradation of encapsulated P3HT:PCBM films and OPV cells was achieved by exposure to concentrated sunlight using various light intensities (up to 4800 sun) and exposure doses (up to 3600 sun h). Absorption measurements of encapsulated PCBM: P3HT blend films revealed no P3HT photobleaching and slight degradation of the PCBM optical absorption (by few percents). This demonstrates the very high photochemical stability of the P3HT: PCBM photoactive layer in the studied samples. However, a corresponding exposure of the OPV cells resulted in significant deterioration of their photovoltaic performance correlated with increasing exposure photon dose in the degradation experiments. A joint analysis of the light intensity dependence of the cell I–V curves and EQE spectra resolved this contradiction. This analysis showed that the degradation in the cell's photoactive layer was not significant and the photocurrent generation (controlled by the product of absorption, exciton generation and diffusion efficiencies) did not change. Slight degradation of the current generation in the PCBM moiety of the photoactive layer was found to track the corresponding trend in PCBM photo-bleaching. The deterioration of the charge collection in the cells was therefore suggested to be the main reason for the observed degradation of the cell performance, which was exhibited in the sub-linear behavior of photocurrent and EQE decrease with increasing incident bias light intensity. Surprisingly, fresh cells and those after exposure to low solar doses exhibited a super-linear dependence of the photocurrent and EQE in the fullerene-related spectral range on the incident light intensity. This behavior was demonstrated by EQE measurements with varying intensities of monochromatic light or by normalizing EQE spectra measured with white light bias. Generation of traps in PCBM was suggested as the underlying mechanism for this super-linear dependence. Our results show that intensity-dependent measurements can separate degradation in absorption/generation from degradation of the charge collection in the cell, which cannot be separated when measurements are performed only at 1 sun illumination. Intensity-resolved EQE provides unique information about degradation of different device parts/materials, which cannot be obtained without wavelength- and intensity-dependent measurements, and therefore can allow better understanding of degradation mechanisms in OPV.
Acknowledgments This work was performed, in part, in the framework of the “Largecells” project that received funding from the European Commission's Seventh Framework Program (FP7/2007-2014). EAK,
279
IVF and YG acknowledge the support of the European Commission's StableNextSol COST Action MP1307. We thank the Adelis fund for partial support of this research.
Appendix A. Supplementary material Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.solmat.2015.09.020.
References [1] Y. Liu, J. Zhao, Z. Li, C. Mu, W. Ma, H. Hu, K. Jiang, H. Lin, H. Ade, H. Yan, Aggregation and morphology control enables multiple cases of high-efficiency polymer solar cells, Nat. Commun. 5 (2014) 5293. [2] J. Rostalski, D. Meissner, Photocurrent spectroscopy for the investigation of charge carrier generation and transport mechanisms in organic p/n-junction solar cells, Sol. Energy Mater. Sol. Cells 61 (2000) 87–95. [3] E.A. Katz, D. Faiman, S.M. Tuladhar, J.M. Kroon, M.M. Wienk, T. Fromherz, F. Padinger, C.J. Brabec, N.S. Sariciftci, Temperature dependence for the photovoltaic device parameters of polymer–fullerene solar cells under operating conditions, J. Appl. Phys. 90 (2001) 5343–5350. [4] J.K.J. van Duren, A. Dhanabalan, P.A. van Hal, R.A.J. Janssen, Low-bandgap polymer photovoltaic cells, Synth. Met. 121 (2001) 1587–1588. [5] J.M. Kroon, M.M. Wienk, W.J.H. Verhees, J.C. Hummelen, Accurate efficiency determination and stability studies of conjugated polymer/fullerene solar cells, Thin Solid Films 403–404 (2002) 223–228. [6] V. Shrotriya, G. Li, Y. Yao, T. Moriarty, K. Emery, Y. Yang, Accurate measurement and characterization of organic solar cells, Adv. Funct. Mater. 16 (2006) 2016–2023. [7] E. Zimmermann, P. Ehrenreich, T. Pfadler, J.A. Dorman, J. Weickert, L. SchmidtMende, Erroneous efficiency reports harm organic solar cell research, Nat. Photonics 8 (2014) 669–672. [8] S.A. Gevorgyan, J.E. Carle, R. Sondergaard, T.T. Larsen-Olsen, M. Jorgensen, F.C. Krebs, Accurate characterization of OPVs: device masking and different solar simulators, Sol. Energy Mater. Sol. Cells 110 (2013) 24–35. [9] ASTM Standard G173 03, 2012, Standard Tables for Reference Solar Spectral Irradiances: Direct Normal and Hemispherical on 37° Tilted Surface1, ASTM International, West Conshohocken, PA, 2012, DOI: 10.1520/G0173-03R12. [10] D.J. Wehenkel, K.H. Hendriks, M.M. Wienk, R.A. Janssen, The effect of bias light on the spectral responsivity of organic solar cells, Org. Electron. 13 (2012) 3284–3290. [11] I. Gonzalez-Valls, D. Angmo, S.A. Gevorgyan, J.S. Reparaz, F.C. Krebs, M. LiraCantu, Comparison of two types of vertically aligned ZnO NRs for highly efficient polymer solar cells, J. Polym. Sci. B: Polym. Phys. 51 (2013) 272–280. [12] T.J.K. Brenner, Y. Vaynzof, Z. Li, D. Kabra, R.H. Friend, C.R. McNeill, White-light bias external quantum efficiency measurements of standard and inverted P3HT:PCBM photovoltaic cells, J. Phys. D: Appl. Phys. 45 (2012) 415101. [13] J.A. Bartelt, Z.M. Beiley, E.T. Hoke, W.R. Mateker, J.D. Douglas, B.A. Collins, J.R. Tumbleston, K.R. Graham, A. Amassian, H. Ade, J.M.J. Fréchet, M.F. Toney, M.D. McGehee, The importance of fullerene percolation in the mixed regions of polymer–fullerene bulk heterojunction solar cells, Adv. Energy Mater. 3 (2013) 364–374. [14] C.J. Brabec, N.S. Sariciftci, J.C. Hummelen, Plastic solar cells, Adv. Funct. Mater. 11 (2001) 15–26. [15] V.I. Arkhipov, H. Bassler, Exciton dissociation and charge photogeneration in pristine and doped conjugated polymers, Phys. Status Solidi A 201 (2004) 1152–1187. [16] J.J.M. Halls, K.P.R.H. Friend, S.C. Moratti, A.B. Holmes, Exciton diffusion and dissociation in a poly(p-phenylenevinylene)/C-60 heterojunction photovoltaic cell, Appl. Phys. Lett. 68 (1996) 3120–3122. [17] G. Teran-Escobar, D.M. Tanenbaum, E. Voroshazi, M. Hermenau, K. Norrman, M.T. Lloyd, Y. Galagan, B. Zimmermann, M. Hösel, H.F. Dam, M. Jørgensen, S. Gevorgyan, S. Kudret, W. Maes, L. Lutsen, D. Vanderzande, U. Wü rfel, R. Andriessen, R. Roösch, H. Hoppe, A. Rivaton Gülash, Y. Uzunog lu, D. Germack, B. Andreasen, M.V. Madsen, E. Bundgaard, F.C. Krebs, M. LiraCantu, On the stability of a variety of organic photovoltaic devices by IPCE and in situ IPCE analyses – the ISOS-3 inter-laboratory collaboration, Phys. Chem. Chem. Phys. 14 (2012) 11824–11845. [18] H. Waters, N. Bristow, O. Moudam, S.W. Chang, C.J. Su, W.R. Wu, U.S. Jeng, M. Horie, J. Kettle, Effect of processing additive 1,8-octanedithiol on the lifetime of PCPDTBT based organic photovoltaics, Org. Electron. 15 (2014) 2433–2438. [19] T. Tromholt, E.A. Katz, B. Hirsch, A. Vossier, F.C. Krebs, Effects of concentrated sunlight on organic photovoltaics, Appl. Phys. Lett. 96 (2010) 073501. [20] I. Visoly-Fisher, A. Mescheloff, M. Gabay, C. Bounioux, L. Zeiri, M. Sansotera, A.E. Goryachev, A. Braun, Y. Galagan, E.A. Katz, Concentrated sunlight for accelerated stability testing of organic photovoltaic materials: towards decoupling light intensity and temperature, Sol. Energy Mater. Sol. Cells 134 (2015) 99–107.
280
E.A. Katz et al. / Solar Energy Materials & Solar Cells 144 (2016) 273–280
[21] Y. Galagan, J.E.J.M. Rubingh, R. Andriessen, C.C. Fan, P.W.M. Blom, S.C. Veenstra, J.M. Kroon, ITO-free flexible organic solar cells with printed current collecting grids, Sol. Energy Mater. Sol. Cells 95 (2011) 1339–1343. [22] Y. Galagan, E.W.C. Coenen, S. Sabik, H.H. Gorter, M. Barink, S.C. Veenstra, J.M. Kroon, R. Andriessen, P.W.M. Blom, Sol. Energy Mater. Sol. Cells 104 (2012) 32–38. [23] Y. Galagan, E.W.C. Coenen, B. Zimmermann, L.H. Slooff, W.J.H. Verhees, S.C. Veenstra, J.M. Kroon, M. Jorgensen, F.C. Krebs, R. Andriessen, Adv. Energy Mater. 4 (2014) 1300498. [24] F.V. Assche, H. Rooms, E. Young, J. Michels, T.V. Mol, G. Rietjens, P.V.D. Weijer, P. Bouten, Thin-film barrier on foil for organic LED lamps, in: Proceedings of AIMCAL Fall Technical Conference, Myrtle beach, SC, USA, 2008, pp. 1152–1173. [25] J.M. Gordon, D. Babai, D. Feuermann, A high-irradiance solar furnace for photovoltaic characterization and nanomaterial synthesis, Sol. Energy Mater. Sol. Cells 95 (2011) 951–956. [26] A. Braun, B. Hirsch, A. Vossier, E.A. Katz, J.M. Gordon, Temperature dynamics of multijunction concentrator solar cells up to ultra-high irradiance, Prog. Photovolt. 21 (2013) 202–208. [27] D. Faiman, D. Feuermann, P. Ibbetson, A. Zemel, A. Ianietz, V. Liubansky, I. Setter, A. Forshman, S. Suraqui, Data Processing for the Negev Radiation Survey: Part II–Typical Meteorological Year (TMY), Version 5, Israel Ministry of Energy and Water publication RD-20-12 (2012). [28] P.J. Brown, D.S. Thomas, A. Köhler, J.S. Wilson, J.-S. Kim, C.M. Ramsdale, H. Sirringhaus, R.H. Friend, Effect of interchain interactions on the absorption and emission of poly(3-hexylthiophene), Phys. Rev. B: Condens. Matter 67 (2003) 064203. [29] M. Manceau, A. Rivaton, J.-L. Gardette, S. Guillerez, N. Lemaitre, Light-induced degradation of the P3HT-based solar cells active layer, Sol. Energy Mater. Sol. Cells 95 (2011) 1315–1325. [30] A. Dupuis, A. Tournebize, P.-O. Bussi`ere, A. Rivaton, J.-L. Gardette, Morphology and photochemical stability of P3HT:PCBM active layers of organic solar cells, Eur. Phys. J. Appl. Phys. 56 (2011) 34104. [31] C.H. Peters, I.T. Sachs-Quintana, J.P. Kastrop, S. Beaupre´, M. Leclerc, M.D. McGehee, High efficiency polymer solar cells with long operating lifetimes, Adv. Energy Mater. 1 (2011) 491–494. [32] M.O. Reese, A.M. Nardes, B.L. Rupert, R.E. Larsen, D.C. Olson, M.T. Lloyd, S.E. Shaheen, D.S. Ginley, G. Rumbles, N. Kopidakis, Photoinduced degradation of polymer and polymer–fullerene active layers: experiment and theory, Adv. Funct. Mater. 20 (2010) 3476–3483. [33] R. Tipnis, J. Bernkopf, S. Jia, J. Krieg, S. Li, M. Storch, D. Laird, Large-area organic photovoltaic module – fabrication and performance, Sol. Energy Mater. Sol. Cells 93 (2009) 442–446. [34] E.A. Katz, J.M. Gordon, W. Tassew, D. Feuermann, Photovoltaic characterization of concentrator solar cells by localized irradiation, J. Appl. Phys. 100 (2006) 044514. [35] A. Manor, E.A. Katz, T. Tromholt, B. Hirsch, F.C. Krebs, Origin of size effect on efficiency of organic photovoltaics, J. Appl. Phys. 109 (2011) 074508.
[36] A. Manor, E.A. Katz, R. Andriessen, Y. Galagan, Study of organic photovoltaics by localized concentrated sunlight: towards optimization of charge collection in large-area solar cells, Appl. Phys. Lett. 99 (2011) 173305. [37] Y. Galagan, A. Mescheloff, S.C. Veenstra, R. Andriessen, E.A. Katz, Reversible degradation in ITO-containing organic photovoltaics under concentrated sunlight, Phys. Chem. Chem. Phys. 17 (2015) 3891–3897. [38] V.D. Mihailetchi, H.X. Xie, B. de Boer, L.J.A. Koster, P.W. Blom, Charge transport and photocurrent generation in poly (3-hexylthiophene): methanofullerene bulk-heterojunction solar cells, Adv. Funct. Mater. 16 (2006) 699–708. [39] V.D. Mihailetchi, J. Wildeman, P.W.M. Blom, Space-charge limited photocurrent, Phys. Rev. Lett. 94 (2005) 126602. [40] J. Szmytkowski, The influence of the thickness, recombination and space charge on the loss of photocurrent in organic semiconductors: an analytical model, J. Phys. D: Appl. Phys. 40 (2007) 3352. [41] S.B. Sapkota, M. Fischer, B. Zimmermann, U. Würfel, Analysis of the degradation mechanism of ITO-free organic solar cell under UV radiation, Sol. Energy Mater. Sol. Cells 121 (2014) 43–48. [42] S. Prezioso, S.M. Hossain, A. Anopchenko, L. Pavesi, M. Wang, G. Pucker, P. Bellutti,, Superlinear photovoltaic effect in Si nanocrystals based metal– insulator–semiconductor devices, Appl. Phys. Lett. 94 (2009) 062108. [43] S. Hegedus, H. Lin, A. Moore, Light induced degradation in amorphous silicon studied by surface photovoltage technique: a comparison of lifetime vs. space charge effects, J. Appl. Phys. 64 (1988) 1215. [44] P. Chatterjee, P. McElhenry, S. Fonash, Influence of illumination conditions on the spectral response of amorphous silicon Schottky barrier structures, J. Appl. Phys. 67 (1990) 3803–3809. [45] B. Park, N.-H. You, E. Reichmanis, Exciton dissociation and charge trapping at poly(3-hexylthiophene)/phenyl-C61-butyric acid methyl ester bulk heterojunction interfaces: photo-induced threshold voltage shifts in organic fieldeffect transistors and solar cells, J. Appl. Phys. 111 (2012) 084908. [46] T.E. Shubina, D.I. Sharapa, C. Schubert, D. Zahn, M. Halik, P.A. Keller, S.G. Pyne, S. Jennepalli, D.M. Guldi, T. Clark, Fullerene van der waals oligomers as electron traps, J. Am. Chem. Soc. 136 (2014) 10890–10893. [47] A. Distler, T. Sauermann, H.-J. Egelhaaf, S. Rodman, D. Waller, K.-S. Cheon, M. Lee, D.M. Guldi, The effect of PCBM dimerization on the performance of bulk heterojunction solar cells, Adv. Energy Mater. 4 (2014) 1300693. [48] H. Zhang, A. Borgschulte, F.A. Castro, R. Crockett, A.C. Gerecke, O. Deniz, J. Heier, S. Jenatsch, F. Nüesch, C. Sanchez-Sanchez, A. Zoladek-Lemanczyk, R. Hany, Photochemical transformations in fullerene and molybdenum oxide affect the stability of bilayer organic solar cells, Adv. Energy Mater. 5 (2015) 1400734. [49] A.B. Belay, W. Zhou, R. Krueger, K.O. Davis, Ü. Alver, N.S. Hickman, Effect of UV-ozone exposure on PCBM, IEEE J. Photovolt. 2 (2012) 148–153.