Field-induced exciton dissociation in PTB7-based organic solar cells

The physics of charge separation in organic semiconductors is a topic of ongoing research of relevance to material and device engineering. Herein, we present experimental observations of the field and temperature dependence of charge separation from singlet excitons in PTB7 and ${\mathrm{PC}}_{71}\mathrm{BM}$, and from charge-transfer states created across interfaces in $\mathrm{PTB}7/{\mathrm{PC}}_{71}\mathrm{BM}$ bulk heterojunction solar cells. We obtain this experimental data by time-resolving the near infrared emission of the states from 10 K to room temperature and electric fields from 0 to $2.5{\mathrm{MV}\mathrm{cm}}^{-1}$. Examining how the luminescence is quenched by field and temperature gives direct insight into the underlying physics. We observe that singlet excitons can be split by high fields, and that disorder broadens the high threshold fields needed to split the excitons. Charge-transfer (CT) states, on the other hand, can be separated by both field and temperature. Also, the data imply a strong reduction of the activation barrier for charge splitting from the CT state relative to the exciton state. The observations provided herein of the field-dependent separation of CT states as a function of temperature offer a rich data set against which theoretical models of charge separation can be rigorously tested; it should be useful for developing the more advanced theoretical models of charge separation.

Figure 1

Current-voltage characteristics of the investigated $\mathrm{PTB}7/{\mathrm{PC}}_{71}\mathrm{BM}$ device measured in a solar simulator and corresponding solar cell parameters.

Figure 2

Kinetic model of a carrier escaping from the Coulomb potential via incoherent hopping between discrete sites with an average spacing $r_0$ [44]. After optical excitation, site No. 1 is populated either directly in case of singlet excitons, or indirectly via a CT reaction from a singlet precursor state. From the first site, electrons and holes can either recombine or a carrier can hop to an adjacent site. $a_j$ denotes the hopping rate from site $j$ to site $j+1$. The field-induced dissociation rate $k_{\text{diss}}$ results from the ensemble of individual hopping rates $a_j$. The charges are considered to be free when the escaping carrier reaches a site at a critical distance, $n{r}_0$. Spatial and energetic disorder are taken into account by $x_j$ and $E_j^0$ that are random displacements from the spatial lattice and the energy levels. Full details of the model are found in the text.

Figure 3

Recorded PL spectra (a)–(c) and transients (d)–(f) for different electric fields at a sample temperature of 10 K. The black curves represent control measurements without applied bias. The data in (a) and (d) were obtained from a neat PTB7 film under 450 nm excitation. The data in (b) and (e) are attributed to PCBM excitons, which can be selectively probed in the PTB7/PCBM blend film with an excitation wavelength of 400 nm. (c) and (f) show PL spectra and transients of the PTB7/PCBM blend for an excitation wavelength of 705 nm. In this case, the recorded PL signal is a superposition of singlet and CT emission. The inset in (c) shows the field-dependent spectra of only the CT signature, obtained by integrating the recorded TRPL data over the time interval between 100 and 2000 ps after optical excitation.

Figure 4

Field-induced PL quenching of the neat PTB7 film emission and for the PTB7/PCBM device after selective excitation of the PCBM domains and the intermixed phases, respectively. The sample temperature for the quenching curves presented in (a) was 10 K. In (b), quenching curves of the PTB7 emission are presented for 290 and 80 K. The dotted curves show the modeled quenching behavior, as predicted by the Onsager-Braun model [with parameters ${\epsilon }_0=3,\mu {\tau }_r={10}^{-17}{\mathrm{V}}^2{\mathrm{m}}^{-1}$, and $r_0=1.04\mathrm{nm}$ according to Eqs. (6) and (8)].

Figure 5

Temperature-dependent luminescence-quenching curves for PTB7 and PCBM singlet excitons. The PL of the neat PTB7 film was recorded after 450-nm excitation. PCBM excitons were probed in the PTB7/PCBM blend, after excitation at 400 nm. The symbols represent experimental data. Solid lines are theoretical quenching curves, as predicted by the model suggested by Rubel et al. [44]. The parameter sets we used for fitting the data are summarized in Table 1.

Figure 6

Comparison of experimental transients recorded under an external field of $2\times {10}^8{\mathrm{Vm}}^{-1}$ and simulated transients according to Eq. (9).

Figure 7

Measured CT intensity quenching showing the field and temperature dependence of CT separation at the PTB7/fullerene interface (symbols). Simulation results for separation of carriers by hopping between sites in a disordered but flat potential (solid curves) reproduce the data well, suggesting that the Coulomb potential is strongly screened by the interface.

Figure 8

Schematic of the energetics for (a) separation of singlet states and (b) separation of CT states. $E_1$ denotes the effective activation barrier, the gray-shaded area corresponds to the expected fluctuations of the energy levels described by the disorder parameter sigma, and $x_0$ is the initial separation of the charges populating a CT state. In case of the CT states, the range of spatial disorder is indicated by the darker gray areas and the parameter $\delta r$.

Figure 9

Experimental CT transients and reconstructed transients from the modeled dissociation rates, using Eq. (9), both plotted over one order of magnitude. (a) shows the field-dependent decay of the CT PL for a temperature of 150 K. (b) summarizes the temperature-dependent decay characteristics at zero field.