J. Phys. Chem. 1993,97, 8531-8534
8531
Energy Gap Dependence of the Photocarrier Generation Efficiency in Layered Organic Photoreceptors Minoru Umeda,' Tomoyuki Shimada, Tamotsu Aruga, Tatsuya Niimi, and Masaomi Sasaki Chemical Products Division, Ricoh Co., Ltd., 146- I Nishisawada, Numazu, Shizuoka 41 0. Japan Received: April 29, I993
The photocarrier generation mechanism was studied in layered-type organic photoreceptors containing triphenylamine-based trisazo pigment in the carrier generation layer (CGL). The photocarrier generation efficiency strongly depended upon a series of N,N-diphenyl-4-biphenylamine derivatives utilized in the carrier transport layer (CTL). The dependence was successfully elucidated by means of the energy gap law using the energy gap of the electrochemical oxidation-potential difference between the azo pigment and the N,N-diphenyl4-biphenylamine derivatives. This confirms that photocarrier generation occurs at the CGL/CTL interface and is based on photoinduced electron transfer. The reorganization energy originated in the solid system and the extent of adiabaticity of the reaction are discussed.
Introduction Recently, the high efficiency of light-to-electrical energy conversion in layered-typeorganic photoreceptors has attracted interest from theviewpoint of physical chemistry,' since the energy conversion involves a photochemical process. For some highly sensitive layered organic photoreceptorsfor electrophotography, photocarrier generation has been found to occur not in the carrier generation layer (CGL) where photons are absorbed but at the interface between the CGL and the carrier transport layer (CTL)." This is based on observations in which the photosensitivity of the same single CGLs is quite low. Consequently, this implies that the CTL extrinsically sensitizes the CGL. Thus far, the following mechanisms are conventionally known to bring about extrinsic photocarrier generation in organic materials: exciton dissociation by electrostatic fields which arise between phthalocyanine and adsorbed OZl5and between phthalocyanine and poly(viny1idene fluoride);6 field-assisted dissociation of geminate pairs which are produced from a higherexcited charge-transfer state in poly(N-vinylcarbazole) doped with tetracyanobenzene;' and the same kind of dissociation from a nonrelaxed exciplex state in poly(N-vinylcarbazole) doped with dimethyl terephthalate.* For the extrinsic photocarrier generation at the CGL/CTL interface, we previously developed a mechanism from the energeticsof the carrier generation material (CGM) and carrier transport material (CTM).c9 The mechanism involves a photoinduced electron transfer (ET) at ionizationpotential levels from the CTM in the ground state to the photoexcited CGM. There are many methods to identify an ET;l0J1we have found it effective to measure reaction rate constants which depend upon the energy gap between the initial and final states, based on the Marcus theory.12 This relation is called the energy gap law. In this paper, we will focus on a layered photoreceptor in which photocarrier generation has been proven to occur at the CGL/ CTL interface.4 We will also investigate the mechanism by means of the energy gap dependenceof photocarrier generation efficiency using a series of N,N-diphenyl-4-biphenylaminederivatives13 in the CTL.
Experimental Section Layered photoreceptors used in this study were prepared as follows. A cyclohexanonedispersion containing polyvinyl butyral and a triphenylamine-basedtrisazo pigment (CGM),14 shown in Figure 1, was prepared in a weight ratio of 4:lO. The dispersion was applied to the surface of an aluminized polyester film, and 0022-3654/93/2097-853 1$04.00/0
Figure 1. Chemical structure of the carrier generation material used in this study.
we used a blade to form a CGL with a 0.24 pm thickness. A tetrahydrofuran solution containing bisphenol-A-polycarbonate and CTM, shown in Table Ill3was applied in a weight ratio of 1:l to the surface of the CGL to form a CTL of about 20 pm thickness. Thus, 18 layered photoreceptors having different CTMs and the same CGL were prepared. Photocarrier generation efficiency was evaluated using xerographic techniquesI15which enabled measurement under illumination of the surface potential on the photoreceptor that had been activated by a negative corona charge. The apparatus for the measurement was detailed elsewherea3s4 The quantum efficiency, 4, of the photoreceptor is defined as the number of surface charges removed by one absorbed photon. The value of C$ is given by16
where Cis the capacitance of the photoreceptor per unit area, e is the electronic charge, I is the incident light intensity in photons per second and per unit area, V, is the surface potential, and F is the electric field. Since the maximum absorbance of the CGL occurs at 700 nm, a monochromatic light of 700 nm was applied to the CGL through the optically transparent CTL (see inset of Figure 2). During this measurement, light power for illumination was reduced so that the photoinduced discharge of the surface potential could be governed by an emission-limited conditionI7in order to eliminate photocarrier transport. Thus, measured C$ solely reflects the photocarrier generation efficiency. The first oxidation half-wave potentials (Eox) of the CTMs were measured by means of cyclic voltammetry in separate 0 1993 American Chemical Society
Umeda et al.
8532 The Journal of Physical Chemistry, Vol. 97, No. 32, 1993
TABLE I: Carrier Transport Materials Used in This Study 2j&Rl R3N @ 3 R 24
CTM 1 2 3 4 5 6 7
R3
8 9
Ri
CTM 10 11 12 13 14 15 16 17 18
R2 3-CH3 3-CH3 3-CH3 -H 3-CH3 -H 3-CH3 -H -H
3-CH3 3-CH3 3-CH3 -H 3-CH3 4-CH3 3-CH3 -H -H
Exothermic
1oo 100
A
j
VO V
0
1 os
1 o6
1o7 1o8 E l e c t r i c Field / V.rn-’ Figure 2. The electric field dependence of the quantum efficiency. For layered photoreceptors, the CTMs used in the CTLs were 1 (A), 5 (A), 6(0),7(.),9(0), 14(w),and 16(v)(seeTableI). TheCGLthickness was 0.24 pm, and the CTL thickness was about 20 pm. For the single CGL (0),the thickness was also 0.24 pm. The illumination intensity was 1.2 X 1016photonsdm-2. The inset shows a schematicconfiguration of the layered photoreceptor with photoinduced discharge.
acetonitrile solutions containing 0.1 moladm” tetraethylammonium perchlorate as a supportingelectrolyteand 5 X 10-3molmdm-3 CTM. A platinum disc, platinum foil, and SCE were used as the working, counter, and reference electrodes, respectively. The lifetime of the CGL photoluminescence was measured with a streak camera and a laser-diode pulsar (670 nm, 60 ns duration, 4.2 pJ/pulse) for the CTL excitation. All of the measurements were carried out at 25 f 2 OC. Results and Discussion Figure 2 shows the relationships between the electric field and the quantum efficiency of 7 layered photoreceptors of varying CTM in the CTL, representative of the above-mentioned 1 8 photoreceptors, and that of a single CGL (without CTL). In Figure 2, the quantum efficiency, 4, obtained under the emissionlimited condition, depends on the electric field. Nevertheless, it also depends strongly upon CTM, and its order of magnitude due to CTM is not inverted by the electric field. This implies that the CTM strongly affects the photocarrier generation efficiency. Moreover, in Figure 2, the quantum efficiencies of the layered photoreceptorsare apparently larger than that of the single CGL. This reveals that the CTL sensitizesphotocarrier generation, and photocarriers are generated at the CGL/CTL interface where the CGM exists with the CTM.3.4 Since the CTMs used are of the hole-transport type, the photocarrier generation will involve a hole injection process.
I
CTM
,CGM
0.4
0.6 0.8 1.0 of CTM I V vs. SCE Figure 3. The CTM-oxidation-potential dependence of the quantum efficiency at 4 X lo7 Vm-l. The numbers in this figure refer to the CTMs used in the CTLs and listed in Table I. The curves are fits in accordance with eqs 2,3, and 4 (dashed curve) and eqs 2,4, and 7 (solid curve) using the parameters described in the text. The inset shows an energy scheme of the photocarrier generation proccss at the CGL/CTL interface. €Ox
Therefore, the photocarrier generation process at the CGL/CTL interface is hypothesized, as schematically shown in the inset of Figure 3, to consist of photoinduced ET from the CTM to the CGM which competes with the deactivation of the photoexcited CGM. In general, geminate pairs produced by photoinduced ET (charge separation) cause geminate recombination (charge re“bination).18 However, it can be assumed that thegeneration efficiency of free carriers (holes and electrons) from thegeminate pairs equals unity at higher electric fields.19 Under this condition, 9 is expressed as
4 = k , / ( k , + kL) (2) Here, k, is the rate constant of ET in the photocarrier generation process and kL is the sum of the rate constants of other processes originating in the photoexcited CGM. In the present layered system, the value for kL used in eq 2 is evaluated to be 3.1 X 1Olo s-l as the reciprocal photoluminescence lifetime20of the photoexcited single CGL. The Marcus expression for ET is given by12
[
k, = k, exp -
(x221
(3)
where ko is the preexponential factor, Xis the total reorganization energy, and AE is the energy gap which is defined as the E O X
Layered Organic Photoreceptors
The Journal of Physical Chemistry, Vol. 97, No. 32, 1993 8533
difference between the CGM and CTM: (4) Cyclic voltammograms of the first oxidation for all CTMs exhibit a typical reversible wave; hence, E O X is easily defined by anodic and cathodic peak potentials.21 In Figure 3 , the quantum efficiency of the photoreceptor measured at a higher electric field of 4 X lo7 V.m-l is plotted versus the E O X of the CTM used in the CTL. The dashed curve in Figure 3 is calculated by means of eqs 2, 3 , and 4 that employ the following fitting parameters: ko = 3.8 X 1Olos-l, X = 0.19 eV, and EO'CGM = 0.90 V versus SCE. The total reorganization energy in eq 3 is given by X = A, A,. The term A, is the high-frequency (so-called inner sphere) reorganization energy of the molecules involving changes in the lengths and angles of the bonds.22 The X , term is the low-frequency reorganizationenergy of the surroundingmedium which is usually described according to a dielectric continuum model:12
+
e' A, = 4 m o 2r,
-(
1
+ 2r
1
1
")( 7 -J
- rDA
-
(5)
in which n is the refractive index, t, is the static dielectric constant of the surrounding medium, rD and rA are the ionic radii for donor and acceptor, respectively, and rDA is the interior distance. In the solid system of which CGM and CTM are dispersed in binder resins, we think it is important to know the role of the binder resin on ET between the CGM and CTM (CGM* + CTM CGW- + CTM'+). We have observed that the magnitude of 4 of the photoreceptor shifts when binder resin used in the CGL was ~hanged.~3 A similar effect due to binder resin on carrier transport in a CTL (CTM'+ CTM CTM CTM*+) is reported and interpreted based on the Marcus theory.24 Thus, we presume that the electicallyinert resin used in the photoreceptor works as a rigid solvent for ET but that it is not as rigid as X , = 0. For the photoreceptor, n and e, are assessed to be 1.60-1.64 and 2.8 by means of the PRETTI method25 and corona-charging characteristics,26respectively. The measured values are almost the same as those of nonpolar solvents27 so that the third term of eq 5 approaches not 0 but the magnitude of the nonpolar solvents. It is, therefore, believed that the photoreceptor is a kind of nonpolar-rigid-solvent system. In the above simulation, E o X c is~ expected ~ to be 0.90 V versus SCE. We measured Eox of the dispersed CGM (solid state) according to the manner described in ref 28 by means of cyclic ~oltammetry.~g The obtained value is 0.86 V versus SCE, which is in good accord with but slightly negative to the expected value. The E O X of the CTM was measured in CTM-dissolved acetonitrile,whereas the measured CGM was not dissolved. Thus, the difference in measured and expected E o X c is~ considered ~ to be due to the solvation energy of the CGM. In the nonadiabatic limit of eq 3, ko is expressed as30
-
+
-
+
where Vis the electronic coupling matrix element, then ko = 3.8 X 1010 s-1 derives V = 8.0 cm-1. According to the multiphonon theory of electron transfer, the value for k, is expressed by31
(jhv, - AE + A,)' 4X,kBT (7)
S = &/hv, where hv, is the average energy of the active vibrational mode. The solid curve in Figure 3 is calculated using eqs 2, 4, and 7,
employing the following fitting parameters: X, = 0.18 eV, A, = 0.15 eV, V = 10.5 cm-l, hu, = 1400 cm-l, and E O X c o = ~ 0.91 V versus SCE. The evaluated values for Vboth using eqs 3 and 6 and using eq 7 are very small ( V