It is expected that the short circuit current of bulk heterojunction solar cells will rise with increasing active layer thickness due to increased absorption. According to d > ld = p x т x E, on the other hand, electrical losses due to recombination are expected when the thickness of the active layer exceeds the drift distance of the charge carriers, where E = Voc/d is the electric field. Based on these arguments, the thickness dependence of the short circuit current of bulk heterojunction solar cells should give an estimate of the drift distance of the charge carriers, which can then be compared to the рт product determined by the photo-CELIV technique.
As has been emphasized, the performance of bulk heterojunction solar cells is morphology dependent; therefore the proposed comparative study can only be meaningful if the morphology of the active layer is unchanged by changing the thickness of the active layer. The same solution of MDMO-PPV:PCBM 1:4 by weight (chloroform, 0.5 mg polymer/ml solution-1) was used to prepare all the active layers of bulk heterojunction solar cells with varying thickness by varying the spin speed during the spin coating between 1000 and 6000 rpm. It was expected that this procedure would be more suitable for the proposed thickness dependent experiment as compared to e. g. varying the total concentration, or the type of solvent. The thickness of the active layer was around 125 nm at 6000 rpm, and increased nonlinearly to 280 nm at 1000 rpm. The films prepared at lower spin speeds were somewhat less uniform, therefore eight devices were measured, and the results averaged. The average power conversion efficiency of the bulk heterojunction solar cells versus the film thickness of bulk heterojunction solar cells is shown in Figure 10.27. It is around 2.5 % using a thickness of 125 nm, which drops to approx. 1.7 % when the thickness is increased to 280 nm.
The parameters Isc, Voc, FF, and the injection current density at +2 V are shown in Figure 10.28 for comparison. The average short circuit current density increases slightly as the thickness of the active layer is increased. It is maximal at 225 nm (~6 mAcm-2), and drops slightly when at 280 nm. The measured Voc is constant ~800 mV for all active thickness. The fill factor, on the other hand, drops constantly from 0.6 to 0.4 as the thickness of the active layer is increased, which corresponds nicely to the reduced injection current density at +2 V. From
Figure 10.29 The current density versus voltage curves of bulk heterojunction solar cells with varying active layer thicknesses under illumination.
this data it is evident that the drop in the power conversion efficiency at 280 nm is attributed mainly to the reduced fill factor, which decreases more strongly than the slight increase in the short circuit current.
The current density versus voltage curves under illumination are shown in Figure 10.29 for various active layer thicknesses. The J-V curves are analyzed by a modified version of a simple one diode model shown in Figure 10.15, which has been introduced by Schilinsky et al. to explain the illumination intensity dependent J-V curves of bulk heterojunction solar cells based on P3HT:PCBM blends . The model takes into account the applied voltage (Vext) dependent reduction of the charge collection of the photogenerated charge carriers at the electrodes. As the external voltage is increased towards flat band conditions, the electric field in the device is reduced. Accordingly, only charge carriers created within the reduced drift distance ld = ц x т x (Voc – Vext)/d will contribute to the short circuit photocurrent. In other words, the photocurrent (jsc) in Equation 10.8 is not constant, but depends on the applied voltage (Vext).
Important predictions from that model are that the sign of the photocurrent changes when Vext > Voc, and that the short circuit current is reduced when Vext is raised towards Voc. Moreover, the high fill factor of devices with 125 nm thickness indicates that only a small electric field (Voc – Vext)/d is sufficient to collect most of the photogenerated charge carriers from the whole device (the photocurrent reaches ~5 mA cm-2 at 0.6 V, corresponding to 0.2 V/125 nm electric field). On the other hand, the photocurrent density steadily increases to ~0.1 V (corresponding to a field of 0.7 V/226 nm) when the device thickness is increased to 226 nm.
The цт product of the long lived charge carriers can be calculated using the values determined by the photo-CELIV technique as:
ц x TB = 2 x 10-4 cm2 V-1 s-1 x 2 x 10-6 s = 4 x 10-8 cm2V-1 (10.12)
The maximum thickness of the active layer (d < ld) can be calculated as:
d = Уp X г X Voc = У8 X 10-10 cm2 V-1 x 0.8 V = 180nm (10.13)
According to this result, the majority of the long lived charge carriers can be collected at external electrodes in photovoltaic devices not exceeding 180 nm active layer thickness, or in other words, bimolecular recombination of the charge carriers should not limit the short circuit current of the device. The calculated drift distance of the long lived (ps-ms) charge carriers is in accordance with experimental observations shown in Figure 10. 28, which implies that the majority of long lived charge carriers can be collected in operational bulk heterojunction solar cells. These conclusions are also supported by observed light intensity dependence of the short circuit photocurrent density . Scaling factors close to one are typically observed, which indicates that the short circuit current is not limited by second order recombination processes, such as bimolecular recombination.
The science and technology of bulk heterojunction solar cells has been reviewed focusing on advanced characterization techniques and novel materials. The various microscopic (AFM, SEM, TEM) tools are used to study the correlation between the nanomorphology of the interpenetrating phase separated network and the efficiency of the bulk heterojunction solar cells. The various mobility techniques (FET, space charge limited current measurements, and ToF ) are used to determine the charge carrier mobility within the bicontinous interpenetrating network of the electron and hole transporting phases. It has been demonstrated that the novel technique of photoinduced charge carrier extraction by linearly increasing voltage (photo – CELIV) can be used to simultaneously determine the charge carrier mobility (p) and lifetime (t) in bulk heterojunction solar cells. The obtained рт values are in accordance with the observed performance of bulk heterojunction solar cells with varying active layer thicknesses. The major limitation of bulk heterojunction solar cells is their limited absorption in the lower energy part of the solar spectra. Recent developments including low bandgap polymers and light absorbing fullerenes have been summarized.