PHYSICAL REVIEW B 89, 155304 (2014)
Magnetophotoluminescence line-shape narrowing through interactions between excited states in organic semiconducting materials Lei He,1 Mingxing Li,1 Augustine Urbas,2 and Bin Hu1,* 1
Department of Materials Science and Engineering, University of Tennessee–Knoxville, Knoxville, Tennessee 37996, USA 2 AFRL/RXBN, Air Force Research Laboratory, Wright-Patterson Air Force Base, Dayton, Ohio 45433, USA (Received 3 July 2013; revised manuscript received 17 February 2014; published 7 April 2014) We find that interactions between intermolecular excited states can cause a line-shape narrowing in magnetophotoluminescence in an organic composite containing N,N-dimethylaniline and pyrene in the liquid state. The line-shape narrowing indicates that interactions between intermolecular excited states can decrease the spin-exchange interaction within intermolecular excited states. Our analysis shows that interactions between intermolecular excited states can occur through long-range Coulomb interaction, midrange spin-orbital interaction, and short-range spin interaction, with the consequence of line-shape modification in the development of magnetic field effects. Our experimental results reveal a parameter, the interactions between intermolecular excited states, involved in the development of magnetic field effects in organic semiconducting materials. DOI: 10.1103/PhysRevB.89.155304
PACS number(s): 78.20.Jq, 75.30.Et, 81.05.Fb, 75.85.+t
I. INTRODUCTION
Organic semiconducting materials can exhibit magnetic field effects (MFEs) on photoluminescence [1,2], electroluminescence [1,3,4], electrical current [1,3,5–8], and photocurrent [1,2,9–12]. Experimental studies have found that magnetic field effects can be conveniently obtained from intermolecular excited states, regarded as intermolecular electron-hole pairs, in organic semiconducting materials [1–12]. Therefore, using intermolecular excited states becomes an important approach in developing magnetic field effects in organic semiconducting materials. In general, spin-dependent interactions can exist both within individual intermolecular excited states and between intermolecular excited states in the development of magnetic field effects. However, earlier studies have been limited to interactions within intermolecular excited states. Within intermolecular excited states, an important mechanism for magnetic field effects relies on the magnetic field–dependent singlet/triplet ratio through magnetic perturbation of the singlet-triplet intersystem crossing [1,13–15]. Magnetic field– dependent singlet/triplet ratios can be realized when an applied magnetic field perturbs the equilibrium of singlet-triplet intersystem crossings established by competition between the spin-exchange interaction and internal magnetic interaction (hyperfine or spin-orbital coupling) within individual intermolecular excited states [1,13–15]. Therefore, the spinexchange interaction [14,16] and internal magnetic interaction [17–20] within intermolecular excited states can essentially determine the magnetic field effects in organic semiconductors. It should be pointed out that the interactions between intermolecular excited states can affect the spin-exchange and magnetic interactions within individual intermolecular excited states through long-range Coulomb interaction, midrange spin-orbital interaction, and short-range spin interaction. As a consequence, the interactions between intermolecular excited states can be an important factor in the development of magnetic field effects in organic semiconducting materials. In this work, we use magnetophotoluminescence from
*
[email protected]
1098-0121/2014/89(15)/155304(6)
intermolecular charge-transfer (CT) states in liquid solution, as a well-controlled system, to investigate the influence of interactions between intermolecular excited states on the development of MFEs. II. EXPERIMENTAL METHOD
The light-emitting intermolecular CT states were generated under photoexcitation in a typical donor:acceptor system in liquid solution containing N,N-dimethylaniline (DMA) and pyrene dissolved in N,N-dimethylformamide (DMF) [14,15]. Here, DMA and pyrene function as donor and acceptor, respectively. The liquid solution was placed in an electrically generated magnetic field with photoexcitation by a laser beam of 325 nm. The photoluminescence was detected by a spectrofluorometer through an optical fiber. The magnetophotoluminescence is defined as MFE = (IB − I0 )/I0 , where IB and I0 denote the photoluminescence intensities in situations with and without a magnetic field, respectively. The magnetophotoluminescence in this work was detected at the peak emission of the intermolecular CT states. Fitting of magnetophotoluminescence to experimental data was conducted by using Origin 8.0 software with the Nonlinear Curve Fit function. III. RESULTS AND DISCUSSION
The photogenerated intermolecular CT states in the DMA:pyrene:DMF system are strongly photoluminescent between 430 and 630 nm. We note that the photoluminescence spectrum at high photoexcitation intensity (1000 mW/cm2 ) is almost identical to that at low photoexcitation intensity (50 mW/cm2 ), except that the peak emission wavelength at high photoexcitation is slightly blue-shifted by about 2 nm. Applying a magnetic field does not cause any detectable changes in the photoluminescence spectrum in the sample solutions. In our study, the magnetophotoluminescence shows no dependence on the emission wavelength measured within the major portion of the intermolecular CT emission. Figure 1 shows the normalized magnetophotoluminescence curves from the intermolecular CT states in the
155304-1
©2014 American Physical Society
LEI HE, MINGXING LI, AUGUSTINE URBAS, AND BIN HU
PHYSICAL REVIEW B 89, 155304 (2014)
FIG. 1. (Color online) Normalized magnetophotoluminescence curves for pyrene (0.6 mmol):DMA (12.6 mmol):DMF (0.8 ml) liquid solution under low and high photoexcitation intensities. The solid symbols are experimental data. The solid lines are fitted curves from Eq. (1). The inset shows magnetophotoluminescence curves before normalization.
DMA:pyrene:DMF liquid solution under low and high photoexcitation intensities. The photoluminescence intensity first increases and then saturates with increasing field strength, generating a magnetophotoluminescence signal. Because only singlet excited states can radiatively decay in the studied system, this magnetophotoluminescence indicates that a magnetic field can perturb the singlet-triplet intersystem crossing in the intermolecular CT states, leading to an increase of the singlet/triplet ratio in intermolecular CT states [13–15]. From Fig. 1, we can see an interesting phenomenon: increasing photoexcitation intensity causes a line-shape narrowing in magnetophotoluminescence. The line-shape narrowing can be shown by the decrease of the half-width (B1/2 ) of magnetophotoluminescence curves upon increasing photoexcitation intensity, as indicated in Table I. Specifically, for the DMA (12.6 mmol):pyrene (0.6 mmol):DMF (0.8 ml) liquid solution, increasing the photoexcitation intensity from 50 to 1000 mW/cm2 decreases the B1/2 from 31.5 to 26.0 mT. In general, the line-shape narrowing can be attributed to the following two possibilities upon increasing photoexcitation intensity. First, increasing the photoexcitation intensity can increase the density of intermolecular CT states. In this case, the interactions between CT states are responsible for the line-shape narrowing. Second, increasing the photoexcitation intensity may cause an increase in photo-induced ions. Magnetoresistance studies have found that the interaction TABLE I. The B1/2 values of magnetophotoluminescence and parameters from fitting magnetophotoluminescence using Eq. (1) for the pyrene:DMA (12.6 mmol):DMF (0.8 ml) liquid solution under low and high photoexcitation intensities. Npyrene (mmol) P (mW/cm2 ) B1/2 (mT) 0.15 0.6 2.4
50 1000 50 1000 50 1000
36.1 26.8 31.5 26.0 27.3 24.7
A1
B1
A2
B2
0.285 0.121 0.276 0.186 0.250 0.202
5.61 5.60 5.56 5.57 5.56 5.63
0.724 0.889 0.730 0.824 0.764 0.817
54.8 33.8 45.1 32.4 37.1 30.4
between ions (charges) and triplet CT states can change the line shape [7,8]. However, experimental studies have shown that adding ions into the DMA:pyrene system does not change the line shape of magnetophotoluminescence [21]. Theoretically, the triplet ion reaction would not affect the line shape in magnetophotoluminescence if the triplet-triplet annihilation does not appreciably occur in the generation of photoluminescence, which is the case in our work. Thus, we can suggest that the interactions between CT states are the primary factor responsible for the line-shape narrowing in magnetophotoluminescence upon increasing photoexcitation intensity. The line shape of magnetic field effects has been extensively discussed by considering polaron pairs based on Lorentzian and non-Lorentzian functions [6,22]. Here, we ) use the changing rate of the singlet/triplet ratio [ ∂(S/T ] as ∂B an explicit parameter to describe the line shape. In this case, we can clearly see how line-shape narrowing can reflect the effects of the interaction between intermolecular CT states on spin interaction within CT states. Specifically, the line shape of magnetophotoluminescence essentially reflects how fast the singlet/triplet ratio [ TS ] changes with the applied magnetic field when the magnetic field perturbs the singlet-triplet intersystem crossing within individual intermolecular CT states. Therefore, the line shape of magnetophotoluminescence is controlled ) by the changing rate of the singlet/triplet ratio [ ∂(S/T ]. Here, ∂B the changing rate of the singlet/triplet ratio means the rate with which the singlet/triplet ratio changes with the applied magnetic field. We know that in intermolecular CT states, the spin-exchange and internal magnetic interactions disallows and allows the singlet-triplet intersystem crossing, respectively [13–16]. The competition between the spin-exchange interaction and internal magnetic interaction leads to a dynamic equilibrium in the singlet-triplet intersystem crossing with a certain singlet/triplet ratio. When an applied magnetic field is comparable to the internal magnetic interaction, it can break the dynamic equilibrium in the singlet-triplet intersystem crossing, with the consequence of changing the singlet/triplet ratio, to generate magnetophotoluminescence [13–15]. In the DMA:pyrene liquid solution, the internal magnetic interaction mainly comes from a hyperfine interaction with the field strength less than 10 mT [23–25], due to very weak spinorbital coupling from the absence of heavy elements. In our experiments, the magnetic field effects were measured in the field range above hyperfine field strength (>10 mT). In this field range, the magnetic field–dependent singlet/triplet ratio is essentially controlled by competition between the spin-exchange interaction, responsible for the spin-conserving process, and the applied magnetic field, responsible for the spin-mixing process [16]. Therefore, in our measurements, the spin-exchange interaction is the key parameter in determining ) the line shape [ ∂(S/T ] of magnetophotoluminescence [16]. ∂B In intermolecular CT states, the spin-exchange interaction can range from a negligible value to a value on the order of milli-electron volts, depending on the electron-hole separation distance [16,26]. In the DMA:pyrene:DMF liquid solution, the electron-hole separation distance in intermolecular CT states can be roughly estimated by the average separation distance between DMA and pyrene molecules by considering the molecular volume occupation. An excessive quantity
155304-2
MAGNETOPHOTOLUMINESCENCE LINE-SHAPE NARROWING . . .
of DMA molecules over pyrene molecules are noted in the DMA:pyrene system. The occupation volume of each DMA molecule can then be used to estimate the average separation distance between DMA and pyrene molecules [27]. The estimated DMA-pyrene separation distance is 0.42 nm. It should be pointed out that the estimated DMA-pyrene separation distance is different from the electron-hole separation distance in an intermolecular CT state due to the fluid nature of the liquid state. Specifically, the electron-hole separation distance is mainly determined by the attractive Coulomb interaction between the electron and the hole, not by the DMA-pyrene separation distance, in a given solvent dielectric background. Nevertheless, from the short DMApyrene separation distance (0.42 nm) estimated from solution concentration, we can still suggest that the spin-exchange interaction in intermolecular CT states has a considerable value. This spin-exchange interaction essentially controls the line shape of magnetophotoluminescence. A narrower line shape in magnetophotoluminescence (i.e., a higher changing ) rate of singlet/triplet ratio [ ∂(S/T ]) corresponds to a smaller ∂B spin-exchange interaction within intermolecular CT states [16]. Here, our magnetophotoluminescence studies indicate that increasing the interactions between intermolecular CT states can decrease the spin-exchange interaction within individual intermolecular CT states, leading to a line-shape narrowing in magnetophotoluminescence. We further use the line shape to analyze the fundamental effects from the interactions between intermolecular CT states on the development of magnetic field effects. In general, the line shape in magnetic field effects can be described by Lorentzian or non-Lorentzian functions [6,17–20,22]. Here, we consider that a magnetic field can first compete with the hyperfine interaction and then with the spin-exchange interaction to perturb the equilibrium of singlet-triplet intersystem crossing in the development of magnetophotoluminescence. This consideration suggests that two Lorentzian functions should be combined to describe hyperfine and spin-exchange components in the line shape of magnetophotoluminescence [Eq. (1)], MFE = A1
B2 B2 + A , 2 B 2 + B12 B 2 + B22
(1)
PHYSICAL REVIEW B 89, 155304 (2014)
5.57 mT). Also of note from our curve fitting is that increasing photoexcitation intensity increases the contribution (A2 ) from the spin-exchange interaction but decreases the contribution from hyperfine interaction (A1 ) to the magnetophotoluminescence, due to the interactions between intermolecular CT states. Our curve fitting results show that combining two Lorentzian functions can give the best fit to experimental magnetophotoluminescence data, as shown in Fig. 1. It should be noted that the spin-exchange interaction in the intermolecular CT states can have a distribution due to the certain distribution of the electron-hole separation distance in intermolecular CT states. In general, there are three typical situations for the electron-hole separation distance. First, when the electron-hole separation distance is too small, the strong spin-exchange interaction does not allow the magnetic field to perturb the singlet-triplet intersystem crossing through spin mixing [16]. Second, when the electron-hole separation distance is too long, the spin-exchange interaction becomes negligible, and a magnetic field competes with the hyperfine interaction to perturb the singlet-triplet intersystem crossing, generating magnetophotoluminescence in the hyperfine regime (10 mT). In our study, the spin-exchange–related line shape can be fitted by using only the Lorentzian function. This suggests that the spin-exchange interaction has a narrow distribution for the intermolecular CT states, which can be shown as an average value with curve fitting, when a magnetic field is able to compete with the spin-exchange interaction to perturb the singlet-triplet intersystem crossing in intermolecular CT states. It should be noted that increasing the density of intermolecular CT states through photoexcitation intensity leads to a decrease in the amplitude of magnetophotoluminescence (inset in Fig. 1). In general, the magnetophotoluminescence amplitude AMFE can be expressed as AMFE =
where B1 and B2 are two constants related to hyperfine and spin-exchange interactions, respectively, within individual intermolecular CT states. The coefficients A1 and A2 represent the contributions from hyperfine and spin-exchange interactions to the magnetophotoluminescence, respectively. The best fit on the experimental data of magnetophotoluminescence curves in Fig. 1 can give the values for the key parameters, as shown in Table I. At a low photoexcitation of 50 mW/cm2 , the parameters are determined to be A1 = 0.276, B1 = 5.56 mT, A2 = 0.730, B2 = 45.1 mT. At a high photoexcitation of 1000 mW/cm2 , the parameters are then given to be A1 = 0.186, B1 = 5.57 mT, A2 = 0.824, B2 = 32.4 mT. These curve fitting results indicate that increasing photoexcitation intensity from 50 to 1000 mW/cm2 can considerably decrease the spin-exchange interaction (B2 ) from 45.1 to 32.4 mT, while the hyperfine interaction (B1 ) remains almost unchanged (5.56 versus
N , N0 + N
(2)
where N refers to the number of intermolecular CT states with a long electron-hole separation distance, which are magnetically sensitive, N0 denotes the number of intermolecular CT states with a short electron-hole separation distance, which are magnetically insensitive. A higher photoexcitation can increase the density of intermolecular excited states, which can cause a reduction in the average electron-hole separation distance in the liquid DMA:pyrene system. When the average electron-hole separation distance decreases, N and N0 will decrease and increase, respectively, leading to the reduction of magnetophotoluminescence amplitude [Eq. (2)]. The effects of interactions between CT states on the line shape of magnetophotoluminescence can be further studied by changing the molecular concentration in the DMA:pyrene:DMF liquid solution. Figure 2 shows normalized magnetophotoluminescence curves at different molar
155304-3
LEI HE, MINGXING LI, AUGUSTINE URBAS, AND BIN HU
PHYSICAL REVIEW B 89, 155304 (2014)
FIG. 2. (Color online) Normalized magnetophotoluminescence curves for pyrene (x mmol):DMA (12.6 mmol):DMF (0.8 ml) liquid solution under a photoexcitation intensity of 50 mW/cm2 . The amount of pyrene (x mmol) was varied from 0.15 to 2.4 mmol. The solid symbols are experimental data. The solid lines are fitted curves from Eq. (1). The inset shows magnetophotoluminescence curves before normalization.
FIG. 3. (Color online) Normalized magnetophotoluminescence curves for pyrene (0.6 mmol):DMA (12.6 mmol):DMF (x ml) liquid solution with different volumes of DMF solvent under a photoexcitation intensity of 50 mW/cm2 . The volume (x ml) of DMF solvent was varied from 0 to 3.2 ml. The inset shows magnetophotoluminescence curves before normalization.
concentrations of pyrene molecules at a given photoexcitation intensity of 50 mW/cm2 . Because the quantity of DMA is largely in excess over pyrene, increasing the molar concentrations of pyrene molecules essentially increases the density of intermolecular CT states at a given photoexcitation intensity. We can see from Fig. 2 that increasing the molar concentration of pyrene molecules can lead to a line-shape narrowing that is very similar to the phenomenon induced by increasing photoexcitation intensity (shown in Fig. 1). The line-shape narrowing caused by increasing molar concentration confirms that the interactions between intermolecular CT states can decrease the spin-exchange interaction within intermolecular CT states, with the consequence of increasing ) the changing rate [ ∂(S/T ] of singlet/triplet ratio in the ∂B development of magnetic field effects. Fitting the experimental magnetophotoluminescence data in Fig. 2 by using Eq. (1) can give the values for critical parameters (B1 , B2 , A1 , and A2 ), as shown in Table I. At a low photoexcitation intensity of 50 mW/cm2 , the spin-exchange interaction (B2 ) decreases from 54.8 to 45.1 and 37.1 mT when the pyrene concentration increases from 0.15 to 0.6 and 2.4 mmol, leading to a decreasing rate of about 4 mT/mmol for the spin-exchange interaction. The hyperfine interaction (B1 ) remains almost unchanged with the pyrene concentration (Table I). At a high photoexcitation intensity of 1000 mW/cm2 , the spin-exchange interaction (B2 ) decreases from 33.8 to 32.4 and 30.4 mT when the pyrene concentration is increased, giving a decreasing rate of about 0.8 mT/mmol for the spin-exchange interaction. Clearly, increasing the pyrene molar concentration can lead to large and small decreasing rates for the spin-exchange interaction at low and high photoexcitation intensities, respectively. This is because, at a high photoexcitation intensity with an already high density of intermolecular CT states, increasing the pyrene concentration can no longer largely enhance the interactions between intermolecular CT states, thus leading to a small change in the spin-exchange interaction within intermolecular CT states. We further demonstrate the effects of interactions between CT states on the line shape of magnetophotoluminescence by changing the density of intermolecular excited states
through solvent dilution effects. Basically, increasing the volume of DMF solvent in the DMA:pyrene:DMF liquid solution can generate two outcomes: (i) decreasing the density of intermolecular CT states by reducing the [DMA:pyrene] concentration (i.e., density effects) and (ii) increasing the electron-hole separation distance in [DMA+ :pyrene− ] pairs by increasing the DMA-pyrene separation distance (i.e., distance effects). It should be noted that in the density effect, decreasing the density of intermolecular CT states reduces the interactions between intermolecular CT states, which should generate a broadening line shape with an increase in the amplitude of magnetophotoluminescence, whereas in the distance effect, increasing the electron-hole separation distance can directly decrease the spin-exchange interaction within intermolecular CT states, which should cause a line-shape narrowing with a decrease in the amplitude of magnetophotoluminescence. Figure 3 shows the magnetophotoluminescence curves with increasing volumes of DMF solvent. We can see that increasing the solvent volume in a low-dilution regime from 0.4 to 1.6 ml leads to a line-shape broadening with an increase in the magnetophotoluminescence amplitude. Further increasing the solvent volume in a high-dilution regime from 1.6 to 3.2 ml generates an opposite phenomenon: line-shape narrowing with a decrease in the magnetophotoluminescence amplitude. This experimental observation suggests that the density and distance effects dominate at low and high dilution regimes, respectively, when the density of intermolecular CT states is changed by solvent dilution. In particular, the solvent dilution results further confirm that the interactions between intermolecular CT states can decrease the spin-exchange interaction within individual CT states, which can then lead to a line-shape narrowing in magnetophotoluminescence. In general, the interactions between intermolecular CT states can occur through three possible mechanisms: longrange Coulomb interaction, midrange spin-orbital interaction, and short-range spin interaction. The long-range Coulomb interaction between CT states comes from the electrical dipole-dipole coupling when each intermolecular CT state is treated as an electrical dipole. It can generate two outcomes: (i) weakening the electron-hole Coulomb attraction within
155304-4
MAGNETOPHOTOLUMINESCENCE LINE-SHAPE NARROWING . . .
Decrease of spinexchange interaction
Changing rate B S ratio T Interactions between CT states hyperfine spin-exchange interaction interaction
spin-exchange hyperf ine interaction interaction
FIG. 4. (Color online) Schematic illustration to show the effects of interactions between intermolecular CT states on the development ) ] represents the changing rate of the of magnetic field effects. [ ∂(S/T ∂B singlet/triplet ratio in intermolecular CT states.
individual CT states due to the dipole field redistribution and (ii) increasing the electron-hole separation distance due to Coulomb shielding effect. These two outcomes can increase the electron-hole separation distance within individual CT states, leading to a decrease in the spin-exchange interaction and a line-shape narrowing in magnetophotoluminescence. The midrange spin-orbital interaction can occur between two CT states when an electron spin in a CT state enters into the orbital field from an adjacent CT state. This is equivalent to intermolecular spin-orbital coupling between two molecules [28,29] and can essentially increase the effective spin-orbital coupling of each CT state, which can generate a line-shape broadening in magnetophotoluminescence. The short-range spin interaction can occur between two adjacent CT states with different spin configurations: (e1 ↑-e2 ↓), (e1 ↑-e2 ↑), (h1 ↑-h2 ↓), (h1 ↑-h2 ↑), (e1 ↑-h2 ↓), (e1 ↓-h2 ↑), where subscripts “1” and “2” refer to two different CT states and e and h are “electron” and “hole,” respectively. The spin interaction between CT states can weaken the spin contribution to the spin-orbital coupling and spin-exchange interaction within individual CT states due to the spin dipole field redistribution. This can be shown as a line-shape narrowing in magnetophotoluminescence. In summary, the interactions between intermolecular CT states can change the line shape of magnetophotoluminescence through three different mechanisms: long-range Coulomb interaction, midrange spin-orbital coupling, and short-range spin interaction. In this work, our magnetophotoluminescence studies indicate that the long-range Coulomb interaction between intermolecular CT states decreases the spin-exchange interaction within individual CT states and causes a line-shape narrowing in magnetophotoluminescence. Figure 4 schematically summarizes the effects of interactions between intermolecular CT states on the line shape of magnetophotoluminescence by changing the
[1] B. Hu, L. Yan, and M. Shao, Adv. Mater. 21, 1500 (2009). [2] F. Ito, T. Ikoma, K. Akiyama, A. Watanabe, and S. Tero-Kubota, J. Phys. Chem. B 109, 8707 (2005).
PHYSICAL REVIEW B 89, 155304 (2014)
spin-exchange interaction within individual CT states. Within the intermolecular CT states, the spin-exchange interaction and internal magnetic interaction (hyperfine interaction for the DMA:pyrene system) are accountable for spin-conserving and spin-dephasing processes, respectively, which mutually compete with each other, leading to a dynamic equilibrium in the singlet-triplet intersystem crossing with a certain singlet/triplet ratio. At the field above hyperfine interaction, an applied magnetic field can compete with the spin-exchange interaction and consequently perturb the singlet-triplet intersystem crossing, leading to a magnetic field–dependent singlet/triplet ratio and a magnetophotoluminescence phenomenon in the spin-exchange field regime. The changing ) ], reflected by the line rate of the singlet/triplet ratio [ ∂(S/T ∂B shape of magnetophotoluminescence, is controlled by the spin-exchange interaction. Our experimental studies show that the long-range Coulomb interactions between intermolecular CT states can decrease the spin-exchange interaction within individual CT states, which consequently leads to a line-shape narrowing in magnetophotoluminescence. IV. CONCLUSIONS
We experimentally studied the effects of interactions between intermolecular CT states on the development of magnetic field effects in a well-controlled light-emitting DMA:pyrene liquid system based on magnetophotoluminescence measurements. We found that increasing the interactions between intermolecular CT states can lead to a line-shape narrowing in magnetophotoluminescence. The line-shape narrowing indicates that the interactions between intermolecular CT states can decrease the spin-exchange interaction within individual intermolecular CT states in the development of magnetic field effects. Our results suggest that the long-range Coulomb interactions between intermolecular CT states can decrease the spin-exchange interaction within individual CT states. Our experimental studies provide an understanding of the development of magnetic field effects in organic semiconducting materials by considering the effects of interactions between intermolecular CT states on the spin-exchange interaction within intermolecular CT states. ACKNOWLEDGMENTS
The authors acknowledge financial support from the Air Force Office of Scientific Research (AFOSR: Grant No. FA9550-11-1-0082). The authors are also thankful for support from NSF (Grant No. ECCS-0644945). This research was partially conducted at the Center for Nanophase Materials Sciences based on user projects (CNMS2012-106 and CNMS2012-107), which is sponsored at Oak Ridge National Laboratory by the Division of Scientific User Facilities, US Department of Energy.
[3] J. Kalinowski, M. Cocchi, D. Virgili, P. Di Marco, and V. Fattori, Chem. Phys. Lett. 380, 710 (2003). [4] J. P. Wang, A. Chepelianskii, F. Gao, and N. C. Greenham, Nat. Commun. 3, 1191 (2012).
155304-5
LEI HE, MINGXING LI, AUGUSTINE URBAS, AND BIN HU
PHYSICAL REVIEW B 89, 155304 (2014)
¨ [5] T. L. Francis, O. Mermer, G. Veeraraghavan, and M. Wohlgenannt, New J. Phys. 6, 185 (2004). [6] P. A. Bobbert, T. D. Nguyen, F. W. A. van Oost, B. Koopmans, and M. Wohlgenannt, Phys. Rev. Lett. 99, 216801 (2007). [7] B. Hu and Y. Wu, Nat. Mater. 6, 985 (2007). [8] P. Janssen, M. Cox, S. H. W. Wouters, M. Kemerink, M. M. Wienk, and B. Koopmans, Nat. Commun. 4, 2286 (2013). [9] E. L. Frankevich, A. A. Lymarev, I. Sokolik, F. E. Karasz, S. Blumstengel, R. H. Baughman, and H. H. H¨orhold, Phys. Rev. B 46, 9320 (1992). [10] P. Shakya, P. Desai, T. Kreouzis, W. P. Gillin, S. M. Tuladhar, A. M. Ballantyne, and J. Nelson, J. Phys.: Condens. Matter 20, 452203 (2008). [11] Z. H. Xu and B. Hu, Adv. Funct. Mater. 18, 2611 (2008). [12] H. Tajima, M. Miyakawa, H. Isozaki, M. Yasui, N. Suzuki, and M. Matsuda, Synth. Met. 160, 256 (2010). [13] B. Brocklehurst, R. S. Dixon, E. M. Gardy, V. J. Lopata, M. J. Quinn, A. Singh, and F. P. Sargent, Chem. Phys. Lett. 28, 361 (1974). [14] K. Schulten, H. Staerk, A. Weller, H. J. Werner, and B. Nickel, Z. Phys. Chem. 101, 371 (1976). [15] N. J. Turro and B. Kraeutler, Acc. Chem. Res. 13, 369 (1980). [16] H. Staerk, W. Kuhnle, R. Treichel, and A. Weller, Chem. Phys. Lett. 118, 19 (1985). ¨ Mermer, G. Veeraraghavan, T. L. Francis, Y. Sheng, D. T. [17] O. Nguyen, M. Wohlgenannt, A. K¨ohler, M. K. Al-Suti, and M. S. Khan, Phys. Rev. B 72, 205202 (2005).
¨ Mermer, [18] Y. Sheng, D. T. Nguyen, G. Veeraraghavan, O. M. Wohlgenannt, S. Qiu, and U. Scherf, Phys. Rev. B 74, 045213 (2006). ¨ Mermer, and [19] Y. Sheng, T. D. Nguyen, G. Veeraraghavan, O. M. Wohlgenannt, Phys. Rev. B 75, 035202 (2007). [20] P. Shakya, P. Desai, M. Somerton, G. Gannaway, T. Kreouzis, and W. P. Gillin, J. Appl. Phys. 103, 103715 (2008). [21] P. K. Bera, D. Nath, and M. Chowdhury, J. Phys. Chem. A 101, 384 (1997). [22] A. J. Schellekens, W. Wagemans, S. P. Kersten, P. A. Bobbert, and B. Koopmans, Phys. Rev. B 84, 075204 (2011). [23] J. B. Birks, Organic Molecular Photophysics (Wiley, London, 1975). [24] A. Weller, F. Nolting, and H. Staerk, Chem. Phys. Lett. 96, 24 (1983). [25] T. D. Nguyen, G. Hukic-Markosian, F. J. Wang, L. Wojcik, X. G. Li, E. Ehrenfreund, and Z. V. Vardeny, Nat. Mater. 9, 345 (2010). [26] A. Kadashchuk, A. Vakhnin, I. Blonski, D. Beljonne, Z. Shuai, J. L. Br´edas, V. I. Arkhipov, P. Heremans, E. V. Emelianova, and H. B¨assler, Phys. Rev. Lett. 93, 066803 (2004). [27] Given that DMA molecules are treated as spheres, the radius of DMA can be calculated from Vsphere = 4π R 3 /3, where Vsphere is determined by the concentration of DMA. [28] Z. S. Romanova, K. Deshayes, and P. Piotrowiak, J. Am. Chem. Soc. 123, 2444 (2001). [29] Y. Wu, Z. Xu, B. Hu, and J. Howe, Phys. Rev. B 75, 035214 (2007).
155304-6
