1: \documentclass[prb,twocolumn,showpacs,floatfix,superscriptaddress]{revtex4}
2: %\documentclass[prb,preprint,showpacs,floatfix,superscriptaddress]{revtex4}
3: %\usepackage{eepic}
4: \usepackage{epsfig}
5: \usepackage{amsmath}
6: \usepackage{graphicx}
7:
8: \bibliographystyle{apsrev}
9:
10: \topmargin -0.3in
11: %\newcounter{saveeqn}%
12: %\newcommand{\alpheqn}{\setcounter{saveeqn}{\value{equation}}%
13: %\stepcounter{saveeqn}\setcounter{equation}{0}%
14: %\renewcommand{\theequation}
15: % {\mbox{\arabic{saveeqn}\alph{equation}}}}%
16: %\newcommand{\reseteqn}{\setcounter{equation}{\value{saveeqn}}%
17: %\renewcommand{\theequation}{\arabic{equation}}}%
18:
19: \begin{document}
20: \title{Theory of excited state absorptions in phenylene-based
21: $\pi$-conjugated polymers}
22: \author{Alok Shukla}
23: \affiliation{Physics Department, Indian Institute of Technology, Mumbai 400076,
24: India}
25: \author{Haranath Ghosh}
26: \affiliation{Centre for Advanced Technology, Indore 452013, India}
27: \author{Sumit Mazumdar}
28: \affiliation{Department of Physics, University of Arizona, Tucson, AZ 85721}
29: \date{\today}
30: \begin{abstract}
31: %We report the results of large scale multireference singles-doubles
32: %configuration
33: %interaction (MRSDCI)
34: %calculations of excited state absorptions in oligomers of
35: %poly-(paraphenylene) (PPP) and
36: %poly-(paraphenylenevinylene) (PPV) within a rigid-band correlated electron
37: %model. The goal of these calculations is to
38: %develop understanding of the high energy even-parity two-photon states
39: %that may be observed in ultrafast photoinduced absorption (PA) experiments.
40: Within a rigid-band correlated electron model for oligomers of
41: poly-(paraphenylene) (PPP) and poly-(paraphenylenevinylene) (PPV), we show
42: that there exist two fundamentally different classes of two-photon
43: A$_g$ states in these systems to which photoinduced absorption (PA) can occur.
44: At relatively lower
45: energies there occur A$_g$ states which are superpositions of
46: one electron - one hole (1e--1h) and two electron -- two hole (2e--2h)
47: excitations, that are both comprised of the highest
48: delocalized valence band and the lowest delocalized conduction band states
49: only. The dominant PA is to one specific member of this class
50: of states (the mA$_g$). In addition to the above class of A$_g$ states,
51: PA can also occur to a higher energy
52: kA$_g$ state whose 2e--2h component is {\em different} and has significant
53: contributions from excitations involving both delocalized and localized
54: bands. Our calculated scaled energies of the mA$_g$ and the kA$_g$ agree
55: reasonably well to the experimentally observed low and high energy PAs in PPV.
56: The calculated relative intensities of the two PAs are also in qualitative
57: agreement with experiment. In the case of ladder-type PPP and its oligomers,
58: we predict from our theoretical work a new intense PA at an energy considerably lower than the region where PA have been observed currently.
59: Based on earlier work that showed that efficient charge--carrier generation
60: occurs upon excitation to odd--parity states that involve both delocalized
61: and localized bands, we speculate
62: that it is the characteristic electronic nature of the kA$_g$
63: that leads to charge generation subsequent to excitation to this state, as found
64: experimentally.
65: \end{abstract}
66: \pacs{78.47.+p, 78.20.Bh, 42.65.Re, 78.40.Me, 78.66.Qn}
67: \maketitle
68: \section{Introduction}
69: \label{intro}
70: The applications of $\pi$-conjugated polymers in optical emission devices such
71: as organic light emitting
72: diodes \cite{LED} and laser active media
73: \cite{laser1,laser2,laser3} have
74: led to intensive investigations of photoluminescent materials like PPP and PPV,
75: their derivatives, and structurally related ladder type materials.
76: Nonlinear spectroscopy, including
77: both ultrafast photoinduced absorption (PA) measurements
78: \cite{Yan1,Yan2,Leng,Frolov1,Frolov2,Klimov,Kraabel,Silva,Gadermaier1,Gadermaier2}
79: and third order nonlinear
80: optical measurements like electroabsorption (EA) \cite{Leng,Liess,Martin},
81: third harmonic generation (THG) \cite{THG} and two-photon absorption (TPA)
82: \cite{Frolov2,Baker,Lemmer} have been carried out extensively to probe the
83: even parity excited states that are dark under one--photon excitation but
84: are two--photon allowed.
85: In particular, interest in PA experiments stems
86: from the observation that PA at high energies
87: is followed by charge separation, with the creation of
88: charged polarons on neighboring chains. Derivatives of PPV, for example,
89: have been investigated by Frolov {\em et al.} \cite{Frolov1,Frolov2},
90: who found two distinct PA bands
91: in these systems, ``low energy'' PA1 and ``high energy'' PA2, occurring
92: at $\sim$ 0.8 eV and $\sim$ 1.3 -- 1.4 eV, respectively. Frolov {\em et al.}
93: ascribe PA1 to the excited state absorption from the optical 1B$_u$ state
94: to the so-called mA$_g$ state (where m is a chain length dependent
95: unknown quantum number),
96: whose nature has been discussed
97: extensively by theorists in the context of
98: nonlinear spectroscopy of both luminescent polymers like PPV as well as
99: nonluminescent linear chain polyacetylenes and
100: polydiacetylenes
101: \cite{Dixit,Chandross,McWilliams,Soos1,Abe,Beljonne,Yaron,Race}.
102: PA2 has been ascribed by Frolov {\em et al.}
103: to the excited state absorption to a higher energy kA$_g$ state (where k is
104: again an unknown quantum number),
105: whose counterpart does not exist in the linear chain polymers, according
106: to these authors. Interestingly, the relaxation
107: dynamics of PA1 and PA2 are very different: while the mA$_g$
108: decays back to the optical 1B$_u$ exciton by internal conversion, the kA$_g$
109: undergoes a different relaxation pathway that leads to dissociation
110: into long-lived polaron pairs that are probably interchain.
111: Based on this, the authors have speculated
112: that the electronic character of the kA$_g$ is different from the mA$_g$.
113:
114: Similar behavior have been observed
115: by several different groups
116: \cite{Klimov,Kraabel,Silva,Gadermaier1,Gadermaier2}, who have studied both
117: PPV derivatives and structurally related materials like polyfluorene,
118: ladder-type PPP and oligomers of the latter.
119: %%SM1/16 -- deleting following lines
120: %In all cases
121: %strong low energy PA1 and relatively
122: %weaker high energy PA2 are observed. Irresepective of whether PA2 is a
123: %signature of a
124: %kA$_g$ state alone \cite{Frolov1,Frolov2}, or whether there occur also
125: %contributions
126: %to PA2 from the polaron absorptions themselves
127: %\cite{Kraabel,Gadermaier1,Gadermaier2}, the same mechanism of polaron
128: %generation
129: %upon excited state absorption is suggested by these experiments.
130: %%SM1/16 -- adding new.
131: There exist apparent subtleties in comparing some of these experimental results
132: with those of Frolov {\em et al.} \cite{Frolov1,Frolov2},
133: as the PA measurements in some of these cases were carried out only in the
134: high energy region ($>$ 1.4 eV) \cite{Gadermaier1,Gadermaier2}, and
135: therefore the induced absorption
136: termed PA1 by these latter authors actually corresponds to PA2 of
137: Frolov {\em et al.} \cite{Frolov1,Frolov2} (see section V for details). Apart from this difference
138: in nomenclature, the fundamental observation in all cases appear to be the same,
139: viz., A$_g$ states beyond some threshold energy mediate interchain charge--transfer.
140: As an aside, we remark that it has
141: also been claimed from ultrafast PA measurements in the wavelength region
142: of infrared active vibrational modes that charge carriers are generated
143: directly at the optical threshold, and that the quantum efficiency of charge
144: generation is wavelength independent \cite{Moses,Miranda}.
145: This last conclusion has,
146: however, been challenged by Silva {\em et al.} \cite{Silva},
147: whose demonstrations that
148: ultrafast photogeneration of charge carriers in a polyfluorene derivative
149: is a consequence of sequential absorption to an A$_g$ state that is
150: at nearly twice the energy of the optical exciton (note that this is different
151: from the kA$_g$), and that there occur
152: negligible yield of polarons under
153: continuous wave conditions (where excitation to high energy A$_g$ states does
154: not occur) argue against the direct photogeneration scenario.
155: The occurrence of EA \cite{Liess,Osterbacka} and TPA \cite{Frolov2} at
156: the same energy where PA2 occurs are yet other demonstrations of the existence
157: of a kA$_g$ state
158: that is dipole-coupled to the 1B$_u$.
159: Recent photo-current excitation cross-correlation experiments
160: \cite{Zenz,Muller} also support the sequential absorption picture. We shall
161: therefore make no further comments on the experiments by Moses {\em et al.}
162: \cite{Moses,Miranda}, which
163: in any case is of secondary interest of this work; our primary interest is to
164: understand the different electronic natures of the mA$_g$ and the kA$_g$
165: at a qualitative level.
166:
167: Initial progress in theoretical understanding of possibly different classes of
168: even parity states in
169: phenylene-based conjugated polymers was made by Chakrabarti and Mazumdar
170: \cite{Chakrabarti}, who examined the excited states of biphenyl
171: and triphenyl within the Pariser--Parr--Pople model \cite{PPP1,PPP2}. Although
172: many--body techniques used by these authors were accurate (exact and
173: quadruple--CI, hereafter QCI), the basis sets used were limited
174: and included only the degenerate pairs
175: of highest occupied molecular orbitals (HOMOs) and lowest unoccupied
176: molecular orbitals (LUMOs) of benzene. Each pair of degenerate MOs consists of
177: a delocalized
178: MO and a localized MO, with the $\pi$ electron densities vanishing on the
179: para carbon atoms in the latter \cite{Rice,Cornil,Chandross1,Shimoi}.
180: Chakrabarti and Mazumdar showed that
181: the two-photon states in these molecules were not only superpositions of
182: 1e--1h and 2e--2h
183: excitations involving the
184: delocalized bonding ($d$) and antibonding ($d^*$) MOs,
185: there occurred also strong admixing with 2e--2h
186: excitations involving localized bonding ($l$) and antibonding
187: ($l^*$) MOs. The authors therefore suggested that
188: %%SM1/16 - modifying, to make it shorter and less convoluted
189: %the existence
190: %of the $l$ and $l^*$ MOs made the higher A$_g$ excited states in phenylene-based
191: %conjugated polymers qualitatively different from the relatively lower energy
192: %two-photon states.
193: there occurred multiple kinds of A$_g$ states in phenylene--based conjugated polymers.
194: According to these authors, A$_g$ states upto some
195: threshold could qualitatively be understood within the space of
196: $d$ and $d^*$ MOs only. The mA$_g$ belongs to this class of A$_g$ states.
197: Beyond the threshold, however, there
198: occurs a crossover to A$_g$ states which have strong contributions from
199: 2e--2h excitations involving $l$ and $l^*$ MOs. We shall hereafter write
200: such states as $(d \to l^*)^2$ and $(l \to l^*)^2$ double excitations
201: (as opposed to the $(d \to d^*)^2$ double excitations with lower energies).
202: Note that in all cases we do not make finer distinctions between configurations
203: in which the hole or electron pairs occupy the same or different MOs (thus
204: $(d \to l^*)^2$ includes the 2e--2h excitations $(d \to l^*;l \to d^*)$).
205: According to these authors, the
206: kA$_g$ belonged to this second class of A$_g$ states.
207: No attempt
208: was made to explain the charge separation and the calculations were limited
209: strictly to explaining the possible difference between the mA$_g$ and the
210: kA$_g$.
211:
212: The calculations by Chakrabarti and Mazumdar \cite{Chakrabarti} are
213: suggestive but far from complete. Firstly,
214: all calculated A$_g$ states appeared
215: to have nearly equal admixtures of $(d \to d^*)^2$, $(d \to l^*)^2$,
216: $(l \to d^*)^2$ and $(l \to l^*)^2$, which would appear to contradict the
217: conjecture about different classes of A$_g$ states. It was claimed that the
218: strong mixing between different classes of 2e--2h excitations was a consequence
219: of the severe length restrictions in the previous calculations and that with
220: increasing conjugation length distinct A$_g$ states dominated by different
221: kinds of 2e--2h excitations would emerge.
222: The actual demonstration
223: of this requires going beyond the earlier small oligomer calculations.
224: A second shortcoming of the earlier calculations is that they
225: completely ignored the outer delocalized MOs of each benzene (as well as
226: the MOs due to the vinylene segments in case of PPV).
227: The complete band structure of PPV can be seen, for example, in reference
228: \onlinecite{Chandross1}, where one finds that in addition to the
229: innermost delocalized valence and conduction bands
230: (referred to as simply $d$ and $d^*$ in the above, but hereafter as
231: $d_1$ and $d_1^*$) there occur also outer delocalized bands
232: ($d_2$ and $d_2^*$ ; $d_3$ and $d_3^*$) that lie below (above) the $l$
233: ($l^*$) bands (the difference between PPV and PPP is the absence of the
234: outermost $d_3$ and $d_3^*$ bands in PPP). The neglect of the outer $d$-bands
235: raises the question whether higher energy PA is not to qualitatively different
236: A$_g$ states involving the localized levels,
237: but to states that are dominated by excitations involving the outer $d$-bands.
238: Finally, the previous calculations were not useful for even qualitative
239: comparisons of relative energies and oscillator strengths of theoretical
240: PA1 and PA2, again a consequence of the very severe length restriction.
241:
242: In view of the above, we believe
243: that a thorough and accurate theoretical investigation of the
244: two-photon states of
245: oligomers of PPP and PPV is in order. In the present paper, we present
246: large scale correlated calculations employing the multireference singles
247: and doubles
248: configuration interaction (MRSDCI)
249: method \cite{Buenker,Tavan} on longer oligomers of PPP and PPV within a
250: correlated electron Hamiltonian. Unlike the previous calculations
251: \cite{Chakrabarti}, higher energy MOs or bands are not ignored.
252: We take considerable care that the results we report are accurate, by imposing
253: strict convergence criteria on our MRSDCI calculations.
254: For PPP, calculations are presented
255: for oligomers containing three to five benzene rings. We refer
256: to these as PPP3, PPP4 and PPP5, respectively.
257: In the case of
258: PPV, we consider oligomers that terminate with benzene molecules at both
259: ends, in order to preserve spatial symmetry, and present
260: MRSDCI results
261: for oligomers containing three (PPV3) and four (PPV4) benzene
262: rings. While these oligomers are still relatively short, we believe that it is
263: far more important to incorporate electron correlation effects to high order
264: than to go to longer oligomers.
265:
266: The results of our calculations can be summarized as follows. Firstly,
267: our computed PA spectra
268: %%SM1/16 -- rearrange
269: %in all cases
270: resemble the experimental PA spectra qualitatively:
271: %in
272: %the sense that
273: in all cases we find the theoretical PA1 to be much stronger than PA2.
274: We discuss why the calculated A$_g$ states are expected to
275: occur at energies where PA1
276: and PA2 are observed experimentally.
277: Secondly, our present relatively longer chain calculations do indeed find
278: lower and higher energy two--photon states whose 2e--2h components are
279: significantly different, as would be required to claim that these belong
280: to distinct classes.
281: %%SM1/16 -delete
282: %As we remark later,
283: The different
284: nature of the higher energy two--photon state that we
285: %SM1/16 - change wording: assign
286: believe to be the
287: kA$_g$ may explain the charge separation
288: from this state.
289: While this last statement is a speculation currently, it
290: is supported by earlier photocurrent studies \cite{Kohler}.
291: A third result involves our determination that the mA$_g$ is not the lowest
292: two-photon excited state
293: in any of the systems we have studied (i.e., the
294: quantum number $m >$ 2), even though the 2A$_g$ occurs above
295: the 1B$_u$ in these systems. This particular
296: result is in agreement with other recent theoretical
297: work based on the limited basis of $d_1$ and $d_1^*$ bands
298: \cite{Lavrentiev,Beljonne1}.
299: As we remark later, this may also have
300: experimental significance. Finally, we also find that as observed before
301: from fittings of the linear absorption in PPV films \cite{Chandross2},
302: the standard
303: Ohno parameters \cite{Ohno}
304: may be too large
305: for even qualitative
306: fittings of the PA at high energies,
307: and a previously used phenomenological screened
308: Coulomb parametrization \cite{Chandross2} gives more satisfactory
309: results.
310:
311: In the next section we present our theoretical model
312: and discuss the details of the MRSDCI approach as adopted here.
313: In section III we present a brief review of the theory
314: of linear absorption in the polyphenylenes, to point out that there occur
315: distinct classes of one--excitation in the systems of interest, thereby
316: suggesting the idea of distinct two--excitations. In section IV we present
317: our theoretical results. This is followed
318: by our conclusions and discussions of the scope of future work.
319:
320: \section{The theoretical model and methodology}
321: \label{model}
322:
323: We consider oligomers of PPP and PPV within the Pariser-Parr-Pople model
324: Hamiltonian,
325: \begin{eqnarray}
326: \label{H_PPP}
327: H = - \sum_{\langle ij \rangle, \sigma} t_{ij}
328: (c_{i \sigma}^\dagger c_{j\sigma}+ c_{j \sigma}^\dagger c_{i \sigma}) + \nonumber \\
329: U \sum_i n_{i \uparrow} n_{i \downarrow} +
330: \sum_{i<j} V_{ij} (n_i -1)(n_j -1)
331: \end{eqnarray}
332: \noindent where $\langle ij \rangle$ implies nearest neighbors,
333: $c_{i \sigma}^\dagger$ creates an electron of spin
334: $\sigma$ on the $p_z$
335: orbital of carbon atom $i$, $n_{i \sigma} = c_{i \sigma}^\dagger c_{i \sigma}$
336: is the number of electrons with spin $\sigma$, and
337: $n_i = \sum_\sigma n_{i \sigma}$ is the total number of electrons on atom $i$.
338: The parameters $U$ and $V_{ij}$ are the on--site
339: and long--range Coulomb interactions, respectively, while $t_{ij}$ is the
340: nearest neighbor one-electron hopping matrix element that includes
341: bond alternation and connectivity. The parametrization of the intersite Coulomb
342: interactions is done in a manner similar to the Ohno
343: parametrization \cite{Ohno}
344: \begin{equation}
345: V_{i,j} = U/\kappa (1+0.6117R_{i,j}^2)^{1/2} \; \mbox{,}
346: \label{eq-ohno}
347: \end{equation}
348: where $\kappa$ is a parameter which has been introduced to account for
349: the possible screening of the Coulomb interactions in the
350: system \cite{Chandross2,Castleton}. We have examined both the standard Ohno parameters
351: ($U$ = 11.13 eV, $\kappa$ = 1.0), as well as a
352: particular combination of $U$ and $\kappa$
353: ($U$ = 8.0 eV, $\kappa$ = 2.0) that was shown previously to be
354: satisfactory at a semiquantitative level
355: for explaining the full wavelength dependent ground state
356: absorption
357: spectrum of PPV \cite{Chandross2}.
358: We shall hereafter refer to this second set of parameters as screened Ohno
359: parameters. As far as the hopping matrix elements are
360: concerned, we took $t = -2.4$ eV
361: for the C -- C bond in benzene rings.
362: The hopping corresponding to the interbenzene single bond in the PPP oligomers
363: was taken to be $t=-2.23$ eV. For the vinylene linkage of the PPV oligomers,
364: we chose the hopping elements to be $-2.2$ eV for the single bond, and $-2.6$
365: eV for the double bond.
366: We considered PPP and PPV oligomers in their planar configurations, with
367: the conjugation direction along the $x$ axis. Thus
368: the symmetry group of PPP oligomers is $D_{2h}$, while that
369: of PPV oligomers is $C_{2h}$. Since both symmetry groups have inversion
370: as a symmetry element, the many electron states of these oligomers can be
371: classified according to this symmetry.
372: The two-photon states of both the
373: PPP and PPV oligomers belong to the A$_g$ irreducible representation (irrep)
374: of the respective symmetry group, while
375: the one-photon states belong to the B$_u$ irrep for PPV, and B$_{1u}$
376: ($x$-polarized), and
377: B$_{2u}$ ($y$-polarized) irreps for the oligo-PPPs. Since, here we are
378: concerned mainly with the response of these systems to the
379: $x$-polarized photons,
380: therefore, henceforth, for PPP oligomers also we will refer to the B$_{1u}$
381: states as B$_u$ states.
382: In all cases the calculated PA corresponds to the excited state absorption
383: from the optical 1B$_u$ state.
384: In all the many-body calculations presented
385: in this work, full use of the stated point group symmetries was made.
386:
387: As stated above, the correlated electron calculations were done using the
388: MRSDCI approach, which is a powerful CI technique \cite{Buenker} that
389: has been used previously for linear chain polyenes by Tavan and Schulten
390: \cite{Tavan} as
391: well as others \cite{Beljonne}, and by us to calculate the excited state
392: ordering in polyphenyl- and polydiphenylacetylenes \cite{Ghosh}. As discussed
393: in reference \onlinecite{Ghosh} we use very stringent convergence criterion
394: for all excited states.
395: The methodology behind the MRSDCI calculations is as follows.
396: The calculations are initiated with a restricted Hartree-Fock
397: (RHF) computation of the
398: ground state of the oligomer concerned, followed by
399: a transformation of the Hamiltonian from the site representation
400: (Eq.~\ref{eq-ohno}) to the
401: HF molecular-orbital representation.
402: Subsequently, a singles-doubles CI (SDCI) calculation
403: is performed, the different excited states in A$_g$ and B$_u$ subspaces are
404: examined, and the N$_{ref}$
405: configuration state functions (CSFs) making
406: significant contributions to their many-particle wave functions are
407: identified. The next step
408: is the MRSDCI calculation for which the reference space consists of the
409: N$_{ref}$ CSFs identified in the previous step,
410: and the overall Hamiltonian matrix now includes configurations that are
411: singly and doubly
412: excited with respect to these reference CSFs (thereby including the
413: dominant triply and quadruply excited configurations).
414: The new ground and excited states are now re-examined to identify new CSFs
415: contributing
416: significantly
417: to them so as to augment the reference space for the next set of MRSDCI
418: calculations. This procedure is repeated until
419: satisfactory convergences in the excitation energies of the relavant
420: states are achieved.
421: By the time convergence is achieved
422: typically all CSFs with coefficients of magnitude
423: $0.1$ or more in the corresponding many-particle wave functions have been
424: included in the MRSDCI reference space. Naturally, this leads to very large
425: CI matrices. To give some idea of the highly correlated nature of the
426: wavefunctions of the excited states we have examined and the level of accuracy
427: in our calculations, in Table \ref{tab-nref} we have listed the number of
428: reference
429: functions that were used for each symmetry subspace of PPP3, PPP4 and PPP5,
430: respectively, as well as the overall sizes of the Hamiltonian matrix in each
431: case (note that for PPP3 the method used was QCI rather than MRSDCI). The
432: number of MRSDCI reference functions are larger in the A$_g$ subspaces than in
433: the B$_u$
434: subspace because while only the 1B$_u$ was optimized in the B$_u$ subspace,
435: many different A$_g$ states (all those with significant transition dipole
436: couplings with the 1B$_u$) had to be simultaneously optimized in the
437: A$_g$ subspace. The $N_{ref}$ in Table \ref{tab-nref} should be compared to the few (usually
438: 2 or 3) reference states that are retained in calculations of the lowest A$_g$
439: states \cite{Tavan}. To the best of our knowledge, the present calculations are
440: the most accurate correlated electron calculations that incorporate the full
441: basis set for the polyphenylenes.
442: \section{Theory of ground state absorption and its implication}
443: \label{review}
444:
445: Before we present our calculations of PA,
446: it is useful to recall the results
447: of calculations of the ground state absorption \cite{Rice,Cornil,Shimoi,Chandross1,Chandross2}.
448: This is because multiple classes of final states are relevant also in
449: ground state absorption, and as we indicate below analysis of the ground state
450: absorption strongly suggests that there should occur multiple classes of
451: two-photon states.
452: Within band theory ($U$ = 0) ground state absorption consists of 1e--1h
453: excitations that are low energy $d_1 \to d_1^*$, high energy $l \to l^*$,
454: and intermediate energy $d_1 \to l^*$ and $l \to d_1^*$, with the intermediate
455: energy absorption band occurring exactly half way in between the low and high
456: energy absorption band. Here we have ignored absorptions involving $d_2$,
457: $d_2^*$ etc. bands, as excitations involving these bands lie outside the range
458: of experimental wavelengths.
459: The $d_1 \to d_1^*$ and $l \to l^*$ bands are polarized
460: along the $x$-direction in PPP and predominantly along the $x$-direction in
461: PPV, while the $d_1 \to l^*$ and $l \to d_1^*$ bands are polarized along
462: the $y$-direction and predominantly along the y-direction, respectively.
463: Experimentally in PPV
464: there occur absorptions at $\sim$ 2.4 eV,
465: 3.7 eV, 4.7 eV and $\sim$ 6.0 eV, respectively. These have been explained
466: within a correlated electron picture:
467: absorptions at 2.4 eV
468: and 6.0 eV are due to $d_1 \to d_1^*$ and $l \to l^*$ exciton states,
469: respectively \cite{Rice,Cornil,Chandross2,Shimoi}
470: the absorption at 3.7 eV is to a higher energy $d_1 \to d_1^*$
471: exciton \cite{Chandross2};
472: and finally, the absorption at 4.7 eV is
473: to the ``plus'' linear combination of the excitations
474: $d_1 \to l^*$ + $l \to d_1^*$ \cite{Rice,Cornil,Chandross2,Shimoi}.
475: The corresponding ``minus'' linear combination,
476: $d_1 \to l^*$ -- $l \to d_1^*$, occurs also at about 3.7 eV
477: \cite{Rice,Cornil,Chandross2,Shimoi}, but is forbidden
478: in linear absorption.
479: All of these assignments have been confirmed
480: by polarization studies of absorptions in stretch--oriented samples
481: \cite{Chandross2,Comoretto,Miller}.
482:
483: The relevance of these known results to the present case are as follows.
484: Energetically, the kA$_g$ in PPV is at $\sim$ 3.6 -- 3.8 eV
485: \cite{Frolov1,Frolov2}. Since ground state absorption occurs to a high energy
486: $d_1 \to d_1^*$ excitation of B$_u$ symmetry in this energy region,
487: in principle, the kA$_g$
488: can simply be a similar $d_1 \to d_1^*$ excitation of
489: A$_g$ symmetry. If this were true, the mA$_g$ and
490: the kA$_g$ would be qualitatively similar,
491: and charge carrier
492: creation from the kA$_g$ (but not from the mA$_g$) can only be a
493: consequence of possibly greater e--h separation in the kA$_g$.
494: An alternate possibility is that the kA$_g$
495: is dominated by configurations that are fundamentally different.
496: Recall that
497: the occurrence of the 2A$_g$ below the 1B$_u$ in polyacetylenes and
498: polydiacetylenes is a general many--body phenomenon:
499: A$_g$ eigenstates having strong 2e--2h contributions from specific
500: MOs can be close in energy, or can even occur below B$_u$
501: eigenstates that are dominated by 1e--1h
502: excitations involving the same MOs. Since the lowest energy 1e--1h excitations
503: involving the $d_1 \to l^*$ and $l \to d_1^*$ excitations (the minus linear
504: combinations mentioned above) occur at $\sim$ 3.7 eV,
505: it is to be anticipated that
506: many-body A$_g$ eigenstates that are dominated by $(d_1 \to l^*)^2$ and
507: $(l \to d_1^*)^2$ occur also within the same energy range. In principle then,
508: the kA$_g$ can also be dominated by 2e--2h components involving the $l$ and
509: $l^*$ bands, and be qualitatively different from the mA$_g$. As shown in the
510: following sections, we do indeed find evidence for such A$_g$ states with
511: significant dipole coupling to the 1B$_u$.
512:
513: \section{Results of correlated electron calculations}
514:
515: \subsection{Choice of parameters}
516:
517: As we show below electron correlation effects on different kinds of CSFs are
518: different. This necessitates proper
519: choice of the Coulomb parameters in Eq.~\ref{H_PPP}.
520: Here we show that the bare Ohno parameters are not suitable for high energy
521: states of PPP
522: and PPV. In Fig. 1 we show the calculated excited state
523: absorption from the 1B$_u$ state for
524: PPP3, for the case of Ohno parameters, obtained with the QCI approach.
525: The spectral features I and II are due to the 2A$_g$ and the
526: 4A$_g$ states of the oligomer, respectively,
527: while the strong feature III has contributions from
528: both the 7A$_g$ and the 8A$_g$.
529: The close proximity of
530: spectral feature I
531: to the 1B$_u$ in the theoretical spectrum makes this outside
532: the wavelength range within which the experimental PA features are observed.
533: In principle, it might be possible to observe this
534: 2A$_g$ state in TPA. In reality, Fig.~1 indicates that the
535: strength of the TPA to the 2A$_g$ would be rather weak. This is because the
536: calculated PA in Fig.~1 corresponds to linear absorption from the
537: 1B$_u$ and hence consists of only positive terms; in contrast, the
538: third order susceptibility corresponding to TPA contains both positive and
539: negative terms \cite{FGuo}, and the relative strength of the 2A$_g$ peak in
540: Fig.~1 therefore corresponds to an upper limit for TPA.
541: Thus only calculated
542: PA features II and III
543: in Fig. ~1 should be compared to experiments. The relative
544: oscillator strengths of II and III are exactly opposite to the relative
545: oscillator strengths of PA1 and PA2, which makes the calculated PA
546: inconsistent with experiments \cite{Frolov1,Frolov2}.
547:
548: In order to probe this further, we have examined the final states
549: of all PA features
550: in considerable detail. Each of these wavefunctions is highly correlated, and
551: is a superposition of numerous configurations. In order to obtain broad
552: classifications of the different A$_g$ states
553: we expand the correlated wavefunction as,
554:
555: \begin{eqnarray}
556: \label{wavefunction}
557: |nA_g \rangle = \sum_i a_i |d_1 \to d_1^* \rangle_i + \sum_j b_j |(d_1 \to d_1^*)^2 \rangle_j \nonumber \\
558: + \sum_k c_k |(d_1 \to l^*)^2 \rangle_k + ....
559: \end{eqnarray}
560:
561: In the above nA$_g$ is an arbitrary A$_g$ state and each term on the right hand
562: side contains all CSFs of a given class (for example,
563: $|d_1 \to d_1^* \rangle_i$
564: is the $i$th configuration of the type $d_1 \to d_1$ whose coefficient in
565: nA$_g$ is $a_i$).
566: The right hand
567: side of Eq.~\ref{wavefunction} is obviously not complete and the .... implies
568: the existence of many other types of excitations $d_1 \to d_2^*$,
569: $(d_1 \to d_2^*)^2$, etc. Once terms have been collected in the above manner,
570: it is possible to quantify the overall contribution of excitations of a given
571: kind (for example, $\sum_i |a_i|^2$ is the total contribution by excitations
572: of the type $d_1 \to d_1^*$).
573: In Table II we have
574: given the contributions of each kind of excitation
575: that describe the 2A$_g$, 4A$_g$, 7A$_g$ and 8A$_g$ wavefunctions.
576: The overall sum of
577: the contributions corresponding to each A$_g$ state
578: does not add up to 1, since we have retained only the
579: important excitations with coefficients at least 0.1 in our
580: expansions of the wavefunctions.
581: As seen from Table II, the 2A$_g$ wavefunction is very
582: similar to the 2A$_g$ of linear chain polyenes, in that it is
583: predominantly a superposition
584: of 1e--1h $d_1 \to d_1^*$ and 2e--2h $(d_1 \to d_1^*)^2$ excitations. This
585: supports the earlier application of the effective linear chain model for the
586: description of this state in PPV and PPP \cite{Soos2}.
587: The two states 7A$_g$ and 8A$_g$ also have strong contributions from
588: $(d_1 \to d_1^*)^2$, and weak contributions from $(d_1 \to l^*)^2$ or
589: $(l \to d_1^*)^2$. These states are therefore physically related to
590: the mA$_g$ that has been discussed before in the context of nonlinear
591: spectroscopy.
592: In contrast to the above states,
593: the 4A$_g$ has much stronger contribution
594: from $(d_1 \to l^*)^2$ and $(l \to d_1^*)^2$.
595: The occurrence of a $(d_1 \to l^*)^2$ and $(l \to d_1^*)^2$ type state
596: {\it below} the
597: predominantly $(d_1 \to d_1^*)^2$ states suggests that (a) electron correlation
598: effects are stronger on CSFs involving the
599: localized MOs than on CSFs involving only delocalized MOs, and
600: (b) the bare Ohno Coulomb
601: parameters are too strong to describe PPP or PPV films, since it is the large
602: magnitude of the Ohno correlation parameters that
603: causes the effective crossing of eigenfunctions of different types, which in
604: turn leads to the reversal of the intensities of the
605: absorption bands as a function of energy.
606:
607: Although in the above we have shown the theoretical PA spectrum only
608: for PPP3, we emphasize that identical behavior is seen with the Ohno parameters
609: for PPP4, PPP5, PPV3 and PPV4. In all cases the intensity profile of the
610: excited state absorption is opposite to that observed experimentally, and
611: wavefunction analysis indicates that this is due to the ordering of the states
612: as in the above.
613:
614: That the Ohno parameters are too large for high energy states of
615: PPV, PPP etc. was suggested earlier from calculations of ground state
616: absorption of PPV \cite{Chandross3}.
617: With the
618: Ohno parameters, the calculated $l \to l^*$ absorption (polarized along
619: the $x$-direction) occurs below the
620: calculated $(d_1 \to l^* + l \to d_1^*)$ absorption ($y$-polarized)
621: \cite{Chandross3},
622: in contradiction to polarized absorption experiments
623: \cite{Chandross2,Comoretto,Miller}.
624: As pointed out in reference \onlinecite{Chandross3}, this last
625: theoretical result
626: was obtained even for trans-stilbene in one of the earliest
627: computational works \cite{Beveridge}. Very recently, Castleton and Barford
628: \cite{Castleton}
629: have done careful analysis of
630: a very large number of high energy excited states in benzene, biphenyl and
631: trans--stilbene using full--CI and have reached the same conclusion.
632: Within Eq.~\ref{eq-ohno},
633: the authors suggest $U$ = 7.2 eV and $\kappa$ = 1.36 for the hydrocarbon
634: matrix condensed phases of the above molecules, along with finer modifications
635: of the hopping integrals. These values of $U$ and $\kappa$ are close to the
636: $U$ = 8 eV and $\kappa$ = 2.0 that were suggested in reference
637: \cite{Chandross3}, and that we use in our calculations in the next section.
638: Given that our goal is semiquantitative only
639: (recall that our calculations are for
640: relatively short oligomers while the available experimental results
641: include those for
642: much longer chains) finetuning of the parameters as done by
643: Castleton and Barford would be premature. The necessity to incorporate
644: screening of the bare Ohno Coulomb parameters in order to fit the high energy
645: excited states in condensed phases of conjugated polymers have also been
646: discussed by Moore and Yaron \cite{Moore}, from a different perspective.
647:
648: \subsection{PA spectra with screened Coulomb parameters}
649:
650: In Fig.~2 we have shown the calculated excited state
651: absorptions for (a) PPP4,
652: (b) PPP5, (c) PPV3 and (d) PPV4 for $U$ = 8.0 eV and $\kappa$ = 2.
653: In all cases our abcissa is the energy
654: scaled with respect to E(1B$_u$). As pointed out before \cite{DGuo},
655: convergence in the scaled energies of high energy states
656: with increasing chain length is not expected in this region of relatively
657: short chain lengths. This is because the lowest energy excitations
658: (for example, the 1B$_u$ and the 2A$_g$) converge with increasing chain length
659: much faster than the higher energy states. Taken together with the discrete
660: nature of the excited states, this can indicate an {\it apparent} increase
661: in the scaled energy of the high energy excited state with increasing
662: chain length, even as the actual energy is decreasing. This is exactly what
663: happens between PPP4 and PPP5, and between PPV3 and PPV4 in
664: Fig.~2.
665: Quantitative comparisons of the scaled
666: energies of the theoretical excited state absorption bands and those of the
667: experimental PA bands thus cannot be expected, particularly
668: in view of the fact that the relative location of the A$_g$ state
669: dominated by $(d_1 \to l^*)^2$ excitations is very sensitive to
670: relatively minor changes in the Coulomb correlation parameters
671: (see previous subsection and also below). The energies and the relative
672: intensities of the theoretical induced absorptions are therefore for
673: semiquantitative comparisons only.
674: From Figs. 2(c) and 2(d), it can
675: be concluded that PA1 in PPV derivatives is due to the calculated
676: spectral feature II which peaks slightly below 0.4 $\times$ E(1B$_u$), which
677: would correspond to about 0.9 eV in long chain PPV derivatives (with
678: E(1B$_u$) $\sim$ 2.2 eV).
679: This is quite close
680: to experimental PA1 energy in PPV derivatives \cite{Frolov1,Frolov2}.
681: We remark on the PA energies in PPP derivatives
682: in the next section, where more detailed comparisons to experiments are made.
683: In all cases, the theoretical PA feature I
684: corresponds to the 2A$_g$ (see below), and the close proximity of this state
685: to the 1B$_u$ again suggests that this state is outside the wavelength region
686: within which experimental PA has been observed.
687:
688: \begin{figure}
689: \includegraphics[height=7cm, width=8cm]{fig1-test}
690: \caption{Calculated PA spectrum of PPP3, with the standard
691: Ohno Coulomb parameters. A linewidth of 0.15 eV was assumed.}
692: \label{PPP3-ohno}
693: \end{figure}
694: For detailed understanding,
695: in Table III we have described the wavefunctions of each of the
696: final A$_g$ states corresponding to each band in the calculated PA spectra.
697: The items
698: in Table III are similar to those in Table II, i.e., each entry corresponds to
699: the overall contribution of each kind of an excitation. Remarkably, the
700: wavefunction descriptions of the different spectral features are very
701: similar for all four systems shown in Fig. 2. Broadly
702: speaking, there occur three distinct
703: classes of A$_g$ states in all four systems. These are discussed below.
704:
705: \begin{table}
706: \caption{The number of
707: reference
708: configurations ($N_{ref}$) and the total number of configurations ($N_{total}$)
709: involved in the MRSDCI (or QCI, where indicated) calculations, for different
710: symmetry subspaces of the various oligomers.}
711:
712: \begin{tabular}{lrlrr} \hline \hline
713: Oligomer & \multicolumn{2}{c}{$A_g$} & \multicolumn{2}{c}{$B_u$} \\
714: & $N_{ref}$ & $N_{total}$ & $N_{ref}$ & $N_{total}$ \\ \hline \hline
715: PPP3 & 1$^a$ & 193678 & 1$^a$ & 335545 \\
716: PPP4 & 55 & 284988 & 15 & 76795 \\
717: PPP5 & 48 & 663619 & 7 & 87146 \\
718: PPV3 & 37 & 215898 & 12 & 220905 \\
719: PPV4 & 39 & 981355 & 3 & 225970 \\
720: \hline \hline
721: \end{tabular}
722:
723: {\noindent $^a$ QCI method}
724: \label{tab-nref}
725: \end{table}
726: The first class of states is represented by the 2A$_g$, which in all cases
727: is predominantly a superposition of 1e--1h $d_1 \to d_1^*$ and 2e--2h
728: $(d_1 \to d_1^*)^2$, as with the Ohno parameters.
729: However, compared to the bare Ohno
730: parameters, the relative weight of the 1e--1h excitations here is larger. This
731: is a conseqeuence of the smaller $U$, and is to be expected. Furthermore,
732: the relative weight of the 1e--1h excitations is also larger in the PPP
733: oligomers than in the PPV oligomers. This is also to be expected, based on the
734: larger one--electron gap in PPP. The 2A$_g$ can certainly be descibed within
735: an effective linear chain model with large dimerization that retains only the
736: $d_1$ and $d_1^*$ bands, as suggested before \cite{Soos2}.
737:
738: \begin{table}
739: \caption{Relative weights of the dominant contributions to the excited states
740: of PPP3, computed with the standard Ohno parameters (see text).}
741:
742: \begin{tabular}{lccccc} \hline \hline
743: PA Feature & State & $d_1 \rightarrow d_1^{*}$ &
744: $d_1 \rightarrow d_2^{*}$ & $(d_1 \rightarrow d_1^{*})^2$ &
745: $(d_1 \rightarrow l^{*})^2$ \\ \hline \hline
746: I & $2A_g$ & 0.3585 & 0.0601 & 0.2670 & --- \\
747: II & $4A_g$ & 0.0452 & 0.1521 & 0.1266 & 0.2282 \\
748: III & $7A_g$ & 0.1116 & 0.1280 & 0.1152 & 0.0660 \\
749: & $8A_g$ & 0.0939 & 0.0772 & 0.3155 & 0.0985 \\
750: % IV & $12A_g$ & --- & & & \\
751: % V & $15A_g$ & --- & & & \\
752: \hline \hline
753: \end{tabular}
754: \label{tab-ppp3}
755: \end{table}
756: %
757: %
758: The second class of states are represented in all cases by the different
759: A$_g$ states that form the final states in the
760: second and third bands in the calculated PA spectra. These have
761: strong contributions from 2e--2h $(d_1 \to d_1^*)^2$, and weak but nonzero
762: contributions from $(d_1 \to l^*)^2$
763: and $(l \to d_1^*)^2$.
764: In addition, there occur also 1e--1h contributions of the
765: type $d_1 \to d_2^*$, $d_2 \to d_1^*$, etc., and while there are subtle
766: differences in the relative contributions by different kinds of single
767: excitations involving low and high energy delocalized bands, the overall
768: natures of the excitations that are the final states in absorption bands II
769: and III are similar. The qualitative natures of these eigenstates are very
770: similar to that of the mA$_g$ discussed in the context of nonlinear
771: spectroscopy of polyacetylenes and polydiacetylenes. The wavefunction
772: descriptions make it clear that very qualitatively these wavefunctions
773: can also be described within the effective linear chain model,
774: but with less precision than the 2A$_g$. Our calculated
775: relative intensities of the 2A$_g$ and the mA$_g$ are in qualitative
776: agreement with
777: other recent calculations for polyphenylenes that used a basis space of
778: only the $d_1$ and $d_1^*$ bands \cite{Lavrentiev,Beljonne1}.
779:
780: \begin{figure}
781: \includegraphics[height=7cm, width=8cm]{fig2-test}
782: \caption{Calculated PA spectra of (a) PPP4, (b) PPP5, (c) PPV3, and
783: (d) PPV4, with screened Coulomb parameters. The scales for the
784: intensity are different in different
785: cases. A line
786: width of 0.15 eV was assumed in all cases.}
787: \label{PA-spectra}
788: \end{figure}
789: Above band III in the calculated spectra we always have a state that is
790: qualitatively different from lower energy A$_g$ states.
791: %%SM1/16 -- Alok please check next sentence, -- this is new.
792: %% Is my calling the appropriate feture in PPV4 as
793: %% V correct? There was a funny line (item (iv)) in your email of Jan 14 that I
794: %% did not understand.
795: This corresponds to the
796: spectral feature IV in PPP4, PPP5 and PPV3, and the spectral feature V in PPV4.
797: The relative weights
798: of the two classes of 2e--2h excitations, $(d_1 \to d_1^*)^2$ and
799: $(d_1 \to l^*)^2$, are reversed in this high energy state, with the
800: former now making weak contribution and the latter a strong contribution.
801: This qualitative difference between the A$_g$ states at different energies
802: is in agreement with the earlier conjecture by Chakrabarti and Mazumdar
803: \cite{Chakrabarti}. The clear demarkation between
804: A$_g$ states dominated by either $(d_1 \to d_1^*)^2$ or $(d_1 \to l^*)^2$,
805: rather than nearly equal admixing, is a new result.
806: This is an effect of increased oligomer length, as discussed above. Furthermore,
807: we also notice that the highest energy two--photon states in Fig. 2 have the
808: strongest 1e--1h contribution from excitations of the type $d_1 \to d_2^*$
809: (in the case of PPP oligomers) and $d_1 \to d_3^*$ (in the case of PPV oligomers).
810: This is yet another difference between the mA$_g$ and the kA$_g$.
811: What is also significant is the weak role of $(l \to l^*)^2$ 2e--2h excitations
812: in the A$_g$ states included in Table III, in contrast
813: to their relatively strong role in the high energy A$_g$ states of
814: biphenyl and triphenyl within the previous limited basis calculations
815: \cite{Chakrabarti}. This is also an effect of increased chain length,
816: with the $(l \to l^*)^2$ 2e--2h excitations now perhaps dominating even
817: higher energy distinct two-photon states.
818: The weaker oscillator strength of absorption
819: band IV, relative to the strong oscillator strength of absorption band II
820: is easily understood from Table III: 2e--2h excitations of the type
821: $(d_1 \to l^*)^2$ have small dipole coupling with
822: the 1B$_u$, which is predominantly $d_1 \to d_1^*$. Thus the dipole coupling
823: of the A$_g$ states responsible for absorption band IV originates mostly
824: from the small 1e--1h contribution to this state. These highest energy
825: A$_g$ states are neither expected nor found in the limited basis calculations
826: by Lavrentiev {\em et al.} \cite{Lavrentiev} and Beljonne \cite{Beljonne1}.
827:
828: Taken together, the results of Table III indicate the occurrence of
829: different classes
830: of A$_g$ states, which are distinguished by the dominance of different types
831: of 2e--2h excitations. With hindsight, this is perhaps not entirely
832: surprising, as discussed in section~\ref{review}.
833:
834: \section{Conclusions and Discussions}
835:
836: %%SM116 -- modifying this paragraph, pl. read carefully, particularly last sentence.
837: Even as the concept of different types of two--photon states appear to be
838: correct, it might seem that
839: straightforward assignments of experimental PA1 and PA2 from the
840: calculated PA
841: spectra of Fig.~2 alone are not possible. We first discuss applications
842: of our theory to PPV derivatives, and then to ladder PPP, etc. The strong spectral
843: feature II in the theoretical spectra (the theoretical mA$_g$) is certainly
844: a component of the experimental PA1 in PPV derivatives. Beyond this there are
845: two possibilities, viz., (i) the spectral feature III corresponds to PA2, and
846: the spectral feature IV (V in PPV4)
847: is too high in energy to be observed experimentally;
848: or (ii) the spectral feature III is also a part of PA1 (especially in long
849: chains) and it is the high energy $(d_1 \to l^*)^2$ excitation
850: that corresponds to PA2. We ascribe the $(d_1 \to l^*)^2$ excitation
851: to PA2 based on the following reasons. First, A$_g$ states that
852: give rise to spectral feature III occur also in linear chain polyenes
853: \cite{Chandross}. Their contribution to EA and TPA are vanishingly weak in long
854: chains. In
855: contrast, the contributions of the experimental kA$_g$ to EA and TPA in
856: PPV derivatives are clearly visible \cite{Frolov2,Liess,Osterbacka}. On the
857: other hand, eigenstates ithat are predominantly $(d_1 \to l^*)^2$
858: are clearly absent in
859: polyacetylenes and polydiacetylenes, which possess only delocalized valence
860: and conduction bands, and hence our
861: assignment would naturally explain the absence of the kA$_g$ state in
862: these systems \cite{Liess,Osterbacka}. Second, as we have already remarked,
863: spectral feature III (but not IV) has also been found in the calculations by
864: Beljonne \cite{Beljonne1}, who, however, determined that in the long chain
865: limit the energies of features II and III converge. This once again supports
866: our assignment of the $(d_1 \to l^*)^2$ excitation to PA2 (note also that the
867: 1e--1h contributions of the highest energy two--photon states are different,
868: further justifying the notion that these states belong to a different class from
869: the lower energy 2A$_g$ and the mA$_g$).
870: We have already pointed out that the calculated
871: PA1 energy is reasonably close to the experimental PA1 energy in PPV
872: derivatives.
873: The calculated scaled energy of PA2 from Figs. 2(c) and (d),
874: at 0.7 -- 0.8 $\times$ E(1B$_u$) (again with experimental E(1B$_u$) = 2.2 eV in substituted
875: PPV's) is at 1.5 -- 1.76 eV, which is a reasonably good fit to the experimental PA2 energy
876: of 1.3 -- 1.4 eV, given the very short lengths of our oligomers, and the
877: increased difficulty of fitting of very high energy states.
878:
879: As mentioned in section I, we believe that the recent experiments on a methyl--substituted
880: ladder type PPP ($m$-LPPP) \cite{Gadermaier1}
881: and on a ladder type oligophenyl \cite{Gadermaier2} have probed an energy
882: region that is considerably above the region where the mA$_g$ occurs. The PA features
883: that have been called PA1 in these works occur at 1.5 eV in the polymer and at 1.8 eV in
884: the oligomer, and are therefore too high in energy to be the same as PA1 in PPV derivatives.
885: We therefore believe that the observed lowest energy PA features actually correspond to PA2
886: of Frolov {\em et al.}, and this is why charge separation occurs upon excitation to this energy.
887: This assignment is supported by the observation of EA \cite{Harrison1}
888: as well as TPA \cite{Harrison2} to a different lower
889: energy two--photon state that occurs $\sim$ 0.7 eV above the 1B$_u$ exciton in the
890: polymer. We therefore make the testable prediction that there should occur in
891: these systems a
892: lower energy PA at $\sim$ 0.7 eV that is considerably
893: stronger than the PAs in the 1.5 -- 1.8 eV range.
894: The scaled PA2 energy from our PPP oligomer calculations,
895: 0.6 -- 0.8 $\times$ E(1B$_u$), with experimental E(1B$_u$) = 2.7 eV in
896: $m$-LPPP \cite{Harrison1},
897: corresponds to 1.6 -- 2.1 eV, once again reasonably close to the experimental
898: value.
899:
900: The above then leads to the question why dissociation of the kA$_g$ to
901: polaron pairs is so efficient. We speculate that this is related to the
902: specific structure of the kA$_g$.
903: In photoconductivity measurement on a PPV
904: derivative it has been
905: found that a large jump in the photoconductivity occurs at 4.7 eV, exactly
906: where the $d_1 \to l^* + l \to d_1^*$ one-photon exciton is located
907: \cite{Kohler}.
908: The jump in the photoconductivity is due to a sudden increase in
909: the interchain charge carrier generation subsequent to the excitation to this
910: particular excited state. We speculate that there occurs a similar
911: enhanced charge carrier generation subsequent to the sequential excitation
912: to an A$_g$ state that has strong contributions from $(d \to l^*)^2$
913: (whose energy
914: is however lower and close to the minus combination of the one--excitations).
915: Indeed, similar mechanism has also been suggested by Zenz {\em et al.} \cite{Zenz}.
916: In the original work by K\"ohler {\em et al.}, theoretical calculations suggested
917: that the enhanced dissociation of the $d_1 \to l^* + l \to d_1^*$ state
918: was a consequence of the highly delocalized electron--hole character of
919: this state \cite{Kohler}. The latter necessarily
920: implies that the charge carriers are
921: similarly weakly bound even in the $(d_1 \to l^*)^2$ double excitation.
922: Additional contribution to the
923: enhanced tendency
924: to charge separation in the case of PPVs from excitations
925: involving $l$ and $l^*$ bands may also
926: come from phenyl ring rotations that occur in these excited states.
927: Given that the charge densities on the para carbons are exactly zero for the
928: $l$ and $l^*$ MOs, bond orders involving the para carbon atoms
929: in the kA$_g$ will be particularly small.
930: This might lead to greater phenyl ring
931: rotation in the kA$_g$ than in the 1B$_u$ excitation.
932: This feature of the
933: kA$_g$ can lead to a lifetime that is longer than the mA$_g$, and it is conceivable
934: that the relatively long lifetime and the weak binding between the electrons and holes
935: together contribute to the charge separation.
936: In a recent work, one of us (S.M.)
937: and colleagues have calculated the relative yields of singlet and triplet
938: excitons starting from oppositely charged polarons, in the presence of
939: interchain hopping of electrons and holes \cite{Wohl,Tandon}.
940: The computational technique used to calculate charge recombination in
941: these works
942: can be applied also to the opposite process of photoinduced
943: charge transfer. Photoinduced charge transfer from different
944: classes of A$_g$ states is of interest in the present context and is currently
945: being investigated.
946:
947: Our observation that the 2A$_g$ is not the mA$_g$ may also have experimental
948: significance.
949: An early measurement
950: detected strong two-photon
951: fluoroscence in PPV from a state that is only 0.5 eV above the 1B$_u$
952: \cite{Baker}, and evidence for a second two--photon state slightly higher in energy.
953: Later experiments involving nonlinear absorption have
954: invariably found
955: the mA$_g$ to be at least 0.8 eV above the 1B$_u$. It is conceivable that the low energy
956: two--photon state found by Baker {\em et al.} is the 2A$_g$.
957: A similar suggestion has also been
958: made by Lavrentiev {\em et al.} \cite{Lavrentiev}.
959:
960: \section{Acknowledgements}
961:
962: Work at Arizona was partially supported by NSF DMR--0101659,
963: NSF ECS--0108696, and the ONR. We
964: acknowledge many useful
965: discussions with G. Lanzani and Z.V. Vardeny.
966:
967: \begin{thebibliography}{60}
968: \expandafter\ifx\csname natexlab\endcsname\relax\def\natexlab#1{#1}\fi
969: \expandafter\ifx\csname bibnamefont\endcsname\relax
970: \def\bibnamefont#1{#1}\fi
971: \expandafter\ifx\csname bibfnamefont\endcsname\relax
972: \def\bibfnamefont#1{#1}\fi
973: \expandafter\ifx\csname citenamefont\endcsname\relax
974: \def\citenamefont#1{#1}\fi
975: \expandafter\ifx\csname url\endcsname\relax
976: \def\url#1{\texttt{#1}}\fi
977: \expandafter\ifx\csname urlprefix\endcsname\relax\def\urlprefix{URL }\fi
978: \providecommand{\bibinfo}[2]{#2}
979: \providecommand{\eprint}[2][]{\url{#2}}
980:
981: \bibitem[{\citenamefont{Friend et~al.}(1999)\citenamefont{Friend, Gymer,
982: Holmes, Burroughes, Marks, Taliani, Bradley, Santos, Br{\'e}das, Logdlund
983: et~al.}}]{LED}
984: \bibinfo{author}{\bibfnamefont{R.~H.} \bibnamefont{Friend}},
985: \bibinfo{author}{\bibfnamefont{R.~W.} \bibnamefont{Gymer}},
986: \bibinfo{author}{\bibfnamefont{A.~B.} \bibnamefont{Holmes}},
987: \bibinfo{author}{\bibfnamefont{J.~H.} \bibnamefont{Burroughes}},
988: \bibinfo{author}{\bibfnamefont{R.~N.} \bibnamefont{Marks}},
989: \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Taliani}},
990: \bibinfo{author}{\bibfnamefont{D.~D.~C.} \bibnamefont{Bradley}},
991: \bibinfo{author}{\bibfnamefont{D.~A.~D.} \bibnamefont{Santos}},
992: \bibinfo{author}{\bibfnamefont{J.~L.} \bibnamefont{Br{\'e}das}},
993: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Logdlund}},
994: \bibnamefont{et~al.}, \bibinfo{journal}{Nature}
995: \textbf{\bibinfo{volume}{397}}, \bibinfo{pages}{121} (\bibinfo{year}{1999}).
996:
997: \bibitem[{\citenamefont{Tessler et~al.}(1996)\citenamefont{Tessler, Denton, and
998: Friend}}]{laser1}
999: \bibinfo{author}{\bibfnamefont{N.}~\bibnamefont{Tessler}},
1000: \bibinfo{author}{\bibfnamefont{G.~J.} \bibnamefont{Denton}},
1001: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{R.~H.}
1002: \bibnamefont{Friend}}, \bibinfo{journal}{Nature}
1003: \textbf{\bibinfo{volume}{382}}, \bibinfo{pages}{695} (\bibinfo{year}{1996}).
1004:
1005: \bibitem[{\citenamefont{Hide et~al.}(1996)\citenamefont{Hide, Diaz-Garcia,
1006: Schwartz, Andersson, Pei, and Heeger}}]{laser2}
1007: \bibinfo{author}{\bibfnamefont{F.}~\bibnamefont{Hide}},
1008: \bibinfo{author}{\bibfnamefont{M.~A.} \bibnamefont{Diaz-Garcia}},
1009: \bibinfo{author}{\bibfnamefont{B.~J.} \bibnamefont{Schwartz}},
1010: \bibinfo{author}{\bibfnamefont{M.~R.} \bibnamefont{Andersson}},
1011: \bibinfo{author}{\bibfnamefont{Q.}~\bibnamefont{Pei}}, \bibnamefont{and}
1012: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Heeger}},
1013: \bibinfo{journal}{Science} \textbf{\bibinfo{volume}{273}},
1014: \bibinfo{pages}{1833} (\bibinfo{year}{1996}).
1015:
1016: \bibitem[{\citenamefont{Frolov et~al.}(1997)\citenamefont{Frolov, Gellermann,
1017: Ozaki, Yoshino, and Vardeny}}]{laser3}
1018: \bibinfo{author}{\bibfnamefont{S.~V.} \bibnamefont{Frolov}},
1019: \bibinfo{author}{\bibfnamefont{W.}~\bibnamefont{Gellermann}},
1020: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Ozaki}},
1021: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Yoshino}}, \bibnamefont{and}
1022: \bibinfo{author}{\bibfnamefont{Z.~V.} \bibnamefont{Vardeny}},
1023: \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{78}},
1024: \bibinfo{pages}{729} (\bibinfo{year}{1997}).
1025:
1026: \bibitem[{\citenamefont{Yan et~al.}(1994)\citenamefont{Yan, Rothberg,
1027: Papadimitrakopoulos, Galvin, and Miller}}]{Yan1}
1028: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Yan}},
1029: \bibinfo{author}{\bibfnamefont{L.~J.} \bibnamefont{Rothberg}},
1030: \bibinfo{author}{\bibfnamefont{F.}~\bibnamefont{Papadimitrakopoulos}},
1031: \bibinfo{author}{\bibfnamefont{M.~E.} \bibnamefont{Galvin}},
1032: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{T.~M.}
1033: \bibnamefont{Miller}}, \bibinfo{journal}{Phys. Rev. Lett.}
1034: \textbf{\bibinfo{volume}{72}}, \bibinfo{pages}{1104} (\bibinfo{year}{1994}).
1035:
1036: \bibitem[{\citenamefont{Hsu et~al.}(1994)\citenamefont{Hsu, Yan, Jedju,
1037: Rothberg, and Hsieh}}]{Yan2}
1038: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Hsu}},
1039: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Yan}},
1040: \bibinfo{author}{\bibfnamefont{T.~M.} \bibnamefont{Jedju}},
1041: \bibinfo{author}{\bibfnamefont{L.~J.} \bibnamefont{Rothberg}},
1042: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{B.~R.} \bibnamefont{Hsieh}},
1043: \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{49}},
1044: \bibinfo{pages}{712} (\bibinfo{year}{1994}).
1045:
1046: \bibitem[{\citenamefont{Leng et~al.}(1994)\citenamefont{Leng, Jeglinski, Wei,
1047: Benner, Vardeny, Guo, and Mazumdar}}]{Leng}
1048: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Leng}},
1049: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Jeglinski}},
1050: \bibinfo{author}{\bibfnamefont{X.}~\bibnamefont{Wei}},
1051: \bibinfo{author}{\bibfnamefont{R.~E.} \bibnamefont{Benner}},
1052: \bibinfo{author}{\bibfnamefont{Z.~V.} \bibnamefont{Vardeny}},
1053: \bibinfo{author}{\bibfnamefont{F.}~\bibnamefont{Guo}}, \bibnamefont{and}
1054: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Mazumdar}},
1055: \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{72}},
1056: \bibinfo{pages}{156} (\bibinfo{year}{1994}).
1057:
1058: \bibitem[{\citenamefont{Frolov et~al.}(2000)\citenamefont{Frolov, Bao,
1059: Wohlgenannt, and Vardeny}}]{Frolov1}
1060: \bibinfo{author}{\bibfnamefont{S.~V.} \bibnamefont{Frolov}},
1061: \bibinfo{author}{\bibfnamefont{Z.}~\bibnamefont{Bao}},
1062: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Wohlgenannt}},
1063: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{Z.~V.}
1064: \bibnamefont{Vardeny}}, \bibinfo{journal}{Phys. Rev. Lett.}
1065: \textbf{\bibinfo{volume}{85}}, \bibinfo{pages}{2196} (\bibinfo{year}{2000}).
1066:
1067: \bibitem[{\citenamefont{Frolov et~al.}(2001)\citenamefont{Frolov, Bao,
1068: Wohlgenannt, and Vardeny}}]{Frolov2}
1069: \bibinfo{author}{\bibfnamefont{S.~V.} \bibnamefont{Frolov}},
1070: \bibinfo{author}{\bibfnamefont{Z.}~\bibnamefont{Bao}},
1071: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Wohlgenannt}},
1072: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{Z.~V.}
1073: \bibnamefont{Vardeny}}, \bibinfo{journal}{Phys. Rev. {\bf B}}
1074: \textbf{\bibinfo{volume}{65}}, \bibinfo{pages}{205209}
1075: (\bibinfo{year}{2001}).
1076:
1077: \bibitem[{\citenamefont{Klimov et~al.}(1998)\citenamefont{Klimov, McBranch,
1078: Barashkov, and Ferraris}}]{Klimov}
1079: \bibinfo{author}{\bibfnamefont{V.~I.} \bibnamefont{Klimov}},
1080: \bibinfo{author}{\bibfnamefont{D.~W.} \bibnamefont{McBranch}},
1081: \bibinfo{author}{\bibfnamefont{N.}~\bibnamefont{Barashkov}},
1082: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Ferraris}},
1083: \bibinfo{journal}{Phys. Rev. {\bf B}} \textbf{\bibinfo{volume}{58}},
1084: \bibinfo{pages}{7654} (\bibinfo{year}{1998}).
1085:
1086: \bibitem[{\citenamefont{Kraabel et~al.}(2000)\citenamefont{Kraabel, Klimov,
1087: Kohlman, Xu, Wang, and McBranch}}]{Kraabel}
1088: \bibinfo{author}{\bibfnamefont{B.}~\bibnamefont{Kraabel}},
1089: \bibinfo{author}{\bibfnamefont{V.~I.} \bibnamefont{Klimov}},
1090: \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Kohlman}},
1091: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Xu}},
1092: \bibinfo{author}{\bibfnamefont{H.~L.} \bibnamefont{Wang}}, \bibnamefont{and}
1093: \bibinfo{author}{\bibfnamefont{D.~W.} \bibnamefont{McBranch}},
1094: \bibinfo{journal}{Phys. Rev. {\bf B}} \textbf{\bibinfo{volume}{61}},
1095: \bibinfo{pages}{8501} (\bibinfo{year}{2000}).
1096:
1097: \bibitem[{\citenamefont{Silva et~al.}(2001)\citenamefont{Silva, Dhoot, Russell,
1098: Stevens, Arias, MacKenzie, Greenham, and Friend}}]{Silva}
1099: \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Silva}},
1100: \bibinfo{author}{\bibfnamefont{A.~S.} \bibnamefont{Dhoot}},
1101: \bibinfo{author}{\bibfnamefont{D.~M.} \bibnamefont{Russell}},
1102: \bibinfo{author}{\bibfnamefont{M.~A.} \bibnamefont{Stevens}},
1103: \bibinfo{author}{\bibfnamefont{A.~C.} \bibnamefont{Arias}},
1104: \bibinfo{author}{\bibfnamefont{J.~D.} \bibnamefont{MacKenzie}},
1105: \bibinfo{author}{\bibfnamefont{N.~D.} \bibnamefont{Greenham}},
1106: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{R.~H.}
1107: \bibnamefont{Friend}}, \bibinfo{journal}{Phys. Rev. {\bf B}}
1108: \textbf{\bibinfo{volume}{64}}, \bibinfo{pages}{125211}
1109: (\bibinfo{year}{2001}).
1110:
1111: \bibitem[{\citenamefont{Gadermaier
1112: et~al.}(2002{\natexlab{a}})\citenamefont{Gadermaier, Cerullo, Sansone,
1113: Leising, Scherf, and Lanzani}}]{Gadermaier1}
1114: \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Gadermaier}},
1115: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Cerullo}},
1116: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Sansone}},
1117: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Leising}},
1118: \bibinfo{author}{\bibfnamefont{U.}~\bibnamefont{Scherf}}, \bibnamefont{and}
1119: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Lanzani}},
1120: \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{89}},
1121: \bibinfo{pages}{117402} (\bibinfo{year}{2002}{\natexlab{a}}).
1122:
1123: \bibitem[{\citenamefont{Gadermaier
1124: et~al.}(2002{\natexlab{b}})\citenamefont{Gadermaier, Cerullo, Zavelani-Rossi,
1125: Sansone, Lanzani, Zojer, Pogantsch, Beljonne, Shuai, Br{\'e}das
1126: et~al.}}]{Gadermaier2}
1127: \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Gadermaier}},
1128: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Cerullo}},
1129: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Zavelani-Rossi}},
1130: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Sansone}},
1131: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Lanzani}},
1132: \bibinfo{author}{\bibfnamefont{E.}~\bibnamefont{Zojer}},
1133: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Pogantsch}},
1134: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Beljonne}},
1135: \bibinfo{author}{\bibfnamefont{Z.}~\bibnamefont{Shuai}},
1136: \bibinfo{author}{\bibfnamefont{J.~L.} \bibnamefont{Br{\'e}das}},
1137: \bibnamefont{et~al.}, \bibinfo{journal}{Phys. Rev. {\bf B}}
1138: \textbf{\bibinfo{volume}{66}}, \bibinfo{pages}{125203}
1139: (\bibinfo{year}{2002}{\natexlab{b}}).
1140:
1141: \bibitem[{\citenamefont{Liess et~al.}(1997)\citenamefont{Liess, Jeglinsky,
1142: Vardeny, Ozaki, Yoshino, Ding, and Barton}}]{Liess}
1143: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Liess}},
1144: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Jeglinsky}},
1145: \bibinfo{author}{\bibfnamefont{Z.~V.} \bibnamefont{Vardeny}},
1146: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Ozaki}},
1147: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Yoshino}},
1148: \bibinfo{author}{\bibfnamefont{Y.}~\bibnamefont{Ding}}, \bibnamefont{and}
1149: \bibinfo{author}{\bibfnamefont{T.}~\bibnamefont{Barton}},
1150: \bibinfo{journal}{Phys. Rev. {\bf B}} \textbf{\bibinfo{volume}{56}},
1151: \bibinfo{pages}{15712} (\bibinfo{year}{1997}).
1152:
1153: \bibitem[{\citenamefont{Martin et~al.}(1999)\citenamefont{Martin, Bradley,
1154: Lane, Mellor, and Burn}}]{Martin}
1155: \bibinfo{author}{\bibfnamefont{S.~J.} \bibnamefont{Martin}},
1156: \bibinfo{author}{\bibfnamefont{D.~D.~C.} \bibnamefont{Bradley}},
1157: \bibinfo{author}{\bibfnamefont{P.~A.} \bibnamefont{Lane}},
1158: \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Mellor}}, \bibnamefont{and}
1159: \bibinfo{author}{\bibfnamefont{P.~L.} \bibnamefont{Burn}},
1160: \bibinfo{journal}{Phys. Rev. {\bf B}} \textbf{\bibinfo{volume}{59}},
1161: \bibinfo{pages}{15133} (\bibinfo{year}{1999}).
1162:
1163: \bibitem[{\citenamefont{Mathy et~al.}(1996)\citenamefont{Mathy, Ueberhofen,
1164: Schenk, Gregorius, Garay, M{\"u}llen, and Bubeck}}]{THG}
1165: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Mathy}},
1166: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Ueberhofen}},
1167: \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Schenk}},
1168: \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Gregorius}},
1169: \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Garay}},
1170: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{M{\"u}llen}},
1171: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Bubeck}},
1172: \bibinfo{journal}{Phys. Rev. {\bf B}} \textbf{\bibinfo{volume}{53}},
1173: \bibinfo{pages}{4367} (\bibinfo{year}{1996}).
1174:
1175: \bibitem[{\citenamefont{Baker et~al.}(1993)\citenamefont{Baker, Gelsen, and
1176: Bradley}}]{Baker}
1177: \bibinfo{author}{\bibfnamefont{C.~J.} \bibnamefont{Baker}},
1178: \bibinfo{author}{\bibfnamefont{O.~M.} \bibnamefont{Gelsen}},
1179: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{D.~D.~C.}
1180: \bibnamefont{Bradley}}, \bibinfo{journal}{Chem. Phys. Lett.}
1181: \textbf{\bibinfo{volume}{201}}, \bibinfo{pages}{127} (\bibinfo{year}{1993}).
1182:
1183: \bibitem[{\citenamefont{Lemmer et~al.}(1993)\citenamefont{Lemmer, Fischer,
1184: Feldmann, Mahrt, Yang, Greiner, B{\"a}ssler, G{\"o}bel, Heesel, and
1185: Kurz}}]{Lemmer}
1186: \bibinfo{author}{\bibfnamefont{U.}~\bibnamefont{Lemmer}},
1187: \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Fischer}},
1188: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Feldmann}},
1189: \bibinfo{author}{\bibfnamefont{R.~F.} \bibnamefont{Mahrt}},
1190: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Yang}},
1191: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Greiner}},
1192: \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{B{\"a}ssler}},
1193: \bibinfo{author}{\bibfnamefont{E.}~\bibnamefont{G{\"o}bel}},
1194: \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Heesel}}, \bibnamefont{and}
1195: \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Kurz}},
1196: \bibinfo{journal}{Chem. Phys. Lett.} \textbf{\bibinfo{volume}{203}},
1197: \bibinfo{pages}{28} (\bibinfo{year}{1993}).
1198:
1199: \bibitem[{\citenamefont{Dixit et~al.}(1991)\citenamefont{Dixit, Guo, and
1200: Mazumdar}}]{Dixit}
1201: \bibinfo{author}{\bibfnamefont{S.~N.} \bibnamefont{Dixit}},
1202: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Guo}}, \bibnamefont{and}
1203: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Mazumdar}},
1204: \bibinfo{journal}{Phys. Rev. {\bf B}} \textbf{\bibinfo{volume}{43}},
1205: \bibinfo{pages}{6781} (\bibinfo{year}{1991}).
1206:
1207: \bibitem[{\citenamefont{Chandross et~al.}(1999)\citenamefont{Chandross, Shimoi,
1208: and Mazumdar}}]{Chandross}
1209: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Chandross}},
1210: \bibinfo{author}{\bibfnamefont{Y.}~\bibnamefont{Shimoi}}, \bibnamefont{and}
1211: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Mazumdar}},
1212: \bibinfo{journal}{Phys. Rev. {\bf B}} \textbf{\bibinfo{volume}{59}},
1213: \bibinfo{pages}{4822} (\bibinfo{year}{1999}).
1214:
1215: \bibitem[{\citenamefont{McWilliams et~al.}(1991)\citenamefont{McWilliams,
1216: Hayden, and Soos}}]{McWilliams}
1217: \bibinfo{author}{\bibfnamefont{P.~C.~M.} \bibnamefont{McWilliams}},
1218: \bibinfo{author}{\bibfnamefont{G.~W.} \bibnamefont{Hayden}},
1219: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{Z.~G.} \bibnamefont{Soos}},
1220: \bibinfo{journal}{Phys. Rev. {\bf B}} \textbf{\bibinfo{volume}{43}},
1221: \bibinfo{pages}{9777} (\bibinfo{year}{1991}).
1222:
1223: \bibitem[{\citenamefont{Soos and Kepler}(1991)}]{Soos1}
1224: \bibinfo{author}{\bibfnamefont{Z.~G.} \bibnamefont{Soos}} \bibnamefont{and}
1225: \bibinfo{author}{\bibfnamefont{R.~G.} \bibnamefont{Kepler}},
1226: \bibinfo{journal}{Phys. Rev. {\bf B}} \textbf{\bibinfo{volume}{43}},
1227: \bibinfo{pages}{11908} (\bibinfo{year}{1991}).
1228:
1229: \bibitem[{\citenamefont{Abe et~al.}(1992)\citenamefont{Abe, Schreiber, Su, and
1230: Yu}}]{Abe}
1231: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Abe}},
1232: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Schreiber}},
1233: \bibinfo{author}{\bibfnamefont{W.~P.} \bibnamefont{Su}}, \bibnamefont{and}
1234: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Yu}}, \bibinfo{journal}{Phys.
1235: Rev. {\bf B}} \textbf{\bibinfo{volume}{45}}, \bibinfo{pages}{9432}
1236: (\bibinfo{year}{1992}).
1237:
1238: \bibitem[{\citenamefont{Beljonne et~al.}(1997)\citenamefont{Beljonne, Cornil,
1239: Shuai, Br{\'e}das, Rohlfing, Bradley, Torruellas, Ricci, and
1240: Stegeman}}]{Beljonne}
1241: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Beljonne}},
1242: \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Cornil}},
1243: \bibinfo{author}{\bibfnamefont{Z.}~\bibnamefont{Shuai}},
1244: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Br{\'e}das}},
1245: \bibinfo{author}{\bibfnamefont{F.}~\bibnamefont{Rohlfing}},
1246: \bibinfo{author}{\bibfnamefont{D.~D.~C.} \bibnamefont{Bradley}},
1247: \bibinfo{author}{\bibfnamefont{W.~E.} \bibnamefont{Torruellas}},
1248: \bibinfo{author}{\bibfnamefont{V.}~\bibnamefont{Ricci}}, \bibnamefont{and}
1249: \bibinfo{author}{\bibfnamefont{G.~I.} \bibnamefont{Stegeman}},
1250: \bibinfo{journal}{Phys. Rev. {\bf B}} \textbf{\bibinfo{volume}{55}},
1251: \bibinfo{pages}{1505} (\bibinfo{year}{1997}).
1252:
1253: %\bibitem[{\citenamefont{Dixit et~al.}(1991)\citenamefont{Dixit, Guo, and
1254: % Mazumdar}}]{Dixit}
1255: %\bibinfo{author}{\bibfnamefont{S.~N.} \bibnamefont{Dixit}},
1256: % \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Guo}}, \bibnamefont{and}
1257: % \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Mazumdar}},
1258: % \bibinfo{journal}{Phys. Rev. {\bf B}} \textbf{\bibinfo{volume}{43}},
1259: % \bibinfo{pages}{6781} (\bibinfo{year}{1991}).
1260: %
1261: %\bibitem[{\citenamefont{Chandross et~al.}(1999)\citenamefont{Chandross, Shimoi,
1262: % and Mazumdar}}]{Chandross}
1263: %\bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Chandross}},
1264: % \bibinfo{author}{\bibfnamefont{Y.}~\bibnamefont{Shimoi}}, \bibnamefont{and}
1265: % \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Mazumdar}},
1266: % \bibinfo{journal}{Phys. Rev. {\bf B}} \textbf{\bibinfo{volume}{59}},
1267: % \bibinfo{pages}{4822} (\bibinfo{year}{1999}).
1268:
1269: \bibitem[{\citenamefont{Yaron}(1996)}]{Yaron}
1270: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Yaron}},
1271: \bibinfo{journal}{Phys. Rev. {\bf B}} \textbf{\bibinfo{volume}{54}},
1272: \bibinfo{pages}{4609} (\bibinfo{year}{1996}).
1273:
1274: \bibitem[{\citenamefont{Race et~al.}(2001)\citenamefont{Race, Barford, and
1275: Bursill}}]{Race}
1276: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Race}},
1277: \bibinfo{author}{\bibfnamefont{W.}~\bibnamefont{Barford}}, \bibnamefont{and}
1278: \bibinfo{author}{\bibfnamefont{R.~J.} \bibnamefont{Bursill}},
1279: \bibinfo{journal}{Phys. Rev. {\bf B}} \textbf{\bibinfo{volume}{64}},
1280: \bibinfo{pages}{035208} (\bibinfo{year}{2001}).
1281:
1282: \bibitem[{\citenamefont{Moses et~al.}(2000)\citenamefont{Moses, Dogariu, and
1283: Heeger}}]{Moses}
1284: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Moses}},
1285: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Dogariu}}, \bibnamefont{and}
1286: \bibinfo{author}{\bibfnamefont{A.~J.} \bibnamefont{Heeger}},
1287: \bibinfo{journal}{Phys. Rev. {\bf B}} \textbf{\bibinfo{volume}{61}},
1288: \bibinfo{pages}{9373} (\bibinfo{year}{2000}).
1289:
1290: \bibitem[{\citenamefont{Miranda et~al.}(2001)\citenamefont{Miranda, Moses, and
1291: Heeger}}]{Miranda}
1292: \bibinfo{author}{\bibfnamefont{P.~B.} \bibnamefont{Miranda}},
1293: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Moses}}, \bibnamefont{and}
1294: \bibinfo{author}{\bibfnamefont{A.~J.} \bibnamefont{Heeger}},
1295: \bibinfo{journal}{Phys. Rev. {\bf B}} \textbf{\bibinfo{volume}{64}},
1296: \bibinfo{pages}{081201} (\bibinfo{year}{2001}).
1297:
1298: \bibitem[{\citenamefont{{\"O}sterbacka
1299: et~al.}(2002)\citenamefont{{\"O}sterbacka, Wohlgenannt, Shkunov, Chinn, and
1300: Vardeny}}]{Osterbacka}
1301: \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{{\"O}sterbacka}},
1302: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Wohlgenannt}},
1303: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Shkunov}},
1304: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Chinn}}, \bibnamefont{and}
1305: \bibinfo{author}{\bibfnamefont{Z.~V.} \bibnamefont{Vardeny}},
1306: \bibinfo{journal}{J. Chem. Phys.} \textbf{\bibinfo{volume}{xxx}},
1307: \bibinfo{pages}{xxxx} (\bibinfo{year}{2002}).
1308:
1309: \bibitem[{\citenamefont{Zenz et~al.}(2001)\citenamefont{Zenz, Lanzani, Cerullo,
1310: Graupner, Leising, Scherf, and DeSilvestri}}]{Zenz}
1311: \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Zenz}},
1312: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Lanzani}},
1313: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Cerullo}},
1314: \bibinfo{author}{\bibfnamefont{W.}~\bibnamefont{Graupner}},
1315: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Leising}},
1316: \bibinfo{author}{\bibfnamefont{U.}~\bibnamefont{Scherf}}, \bibnamefont{and}
1317: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{DeSilvestri}},
1318: \bibinfo{journal}{Synth. Metals} \textbf{\bibinfo{volume}{116}},
1319: \bibinfo{pages}{27} (\bibinfo{year}{2001}).
1320:
1321: \bibitem[{\citenamefont{M{\"u}ller et~al.}(2001)\citenamefont{M{\"u}ller,
1322: Scherf, and Lemmer}}]{Muller}
1323: \bibinfo{author}{\bibfnamefont{J.~G.} \bibnamefont{M{\"u}ller}},
1324: \bibinfo{author}{\bibfnamefont{U.}~\bibnamefont{Scherf}}, \bibnamefont{and}
1325: \bibinfo{author}{\bibfnamefont{U.}~\bibnamefont{Lemmer}},
1326: \bibinfo{journal}{Synth. Metals} \textbf{\bibinfo{volume}{119}},
1327: \bibinfo{pages}{395} (\bibinfo{year}{2001}).
1328:
1329: \bibitem[{\citenamefont{Chakrabarti and Mazumdar}(1999)}]{Chakrabarti}
1330: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Chakrabarti}} \bibnamefont{and}
1331: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Mazumdar}},
1332: \bibinfo{journal}{Phys. Rev. {\bf B}} \textbf{\bibinfo{volume}{59}},
1333: \bibinfo{pages}{4839} (\bibinfo{year}{1999}).
1334:
1335: \bibitem[{\citenamefont{Pariser and Parr}(1953)}]{PPP1}
1336: \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Pariser}} \bibnamefont{and}
1337: \bibinfo{author}{\bibfnamefont{R.~G.} \bibnamefont{Parr}},
1338: \bibinfo{journal}{J. Chem. Phys.} \textbf{\bibinfo{volume}{21}},
1339: \bibinfo{pages}{466} (\bibinfo{year}{1953}).
1340:
1341: \bibitem[{\citenamefont{Pople}(1954)}]{PPP2}
1342: \bibinfo{author}{\bibfnamefont{J.~A.} \bibnamefont{Pople}},
1343: \bibinfo{journal}{Trans. Faraday Soc.} \textbf{\bibinfo{volume}{68}},
1344: \bibinfo{pages}{81} (\bibinfo{year}{1954}).
1345:
1346: \bibitem[{\citenamefont{Rice and Gartstein}(1994)}]{Rice}
1347: \bibinfo{author}{\bibfnamefont{M.~J.} \bibnamefont{Rice}} \bibnamefont{and}
1348: \bibinfo{author}{\bibfnamefont{Y.~N.} \bibnamefont{Gartstein}},
1349: \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{73}},
1350: \bibinfo{pages}{2504} (\bibinfo{year}{1994}).
1351:
1352: \bibitem[{\citenamefont{Cornil et~al.}(1994)\citenamefont{Cornil, Beljonne,
1353: Friend, and Br{\'e}das}}]{Cornil}
1354: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Cornil}},
1355: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Beljonne}},
1356: \bibinfo{author}{\bibfnamefont{R.~H.} \bibnamefont{Friend}},
1357: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{J.~L.}
1358: \bibnamefont{Br{\'e}das}}, \bibinfo{journal}{Chem. Phys. Lett.}
1359: \textbf{\bibinfo{volume}{223}}, \bibinfo{pages}{82} (\bibinfo{year}{1994}).
1360:
1361: \bibitem[{\citenamefont{Chandross et~al.}(1994)\citenamefont{Chandross,
1362: Mazumdar, Jeglinski, Wei, Vardeny, Kwock, and Miller}}]{Chandross1}
1363: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Chandross}},
1364: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Mazumdar}},
1365: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Jeglinski}},
1366: \bibinfo{author}{\bibfnamefont{X.}~\bibnamefont{Wei}},
1367: \bibinfo{author}{\bibfnamefont{Z.~V.} \bibnamefont{Vardeny}},
1368: \bibinfo{author}{\bibfnamefont{E.~W.} \bibnamefont{Kwock}}, \bibnamefont{and}
1369: \bibinfo{author}{\bibfnamefont{T.~M.} \bibnamefont{Miller}},
1370: \bibinfo{journal}{Phys. Rev. {\bf B}} \textbf{\bibinfo{volume}{50}},
1371: \bibinfo{pages}{14702} (\bibinfo{year}{1994}).
1372:
1373: \bibitem[{\citenamefont{Shimoi and Abe}(1996)}]{Shimoi}
1374: \bibinfo{author}{\bibfnamefont{Y.}~\bibnamefont{Shimoi}} \bibnamefont{and}
1375: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Abe}},
1376: \bibinfo{journal}{Synth. Metals} \textbf{\bibinfo{volume}{78}},
1377: \bibinfo{pages}{219} (\bibinfo{year}{1996}).
1378:
1379: \bibitem[{\citenamefont{Buenker and Peyerimhoff}(1974)}]{Buenker}
1380: \bibinfo{author}{\bibfnamefont{R.~J.} \bibnamefont{Buenker}} \bibnamefont{and}
1381: \bibinfo{author}{\bibfnamefont{S.~D.} \bibnamefont{Peyerimhoff}},
1382: \bibinfo{journal}{Theor. Chim. Acta} \textbf{\bibinfo{volume}{35}},
1383: \bibinfo{pages}{33} (\bibinfo{year}{1974}).
1384:
1385: \bibitem[{\citenamefont{Tavan and Schulten}(1987)}]{Tavan}
1386: \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Tavan}} \bibnamefont{and}
1387: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Schulten}},
1388: \bibinfo{journal}{Phys. Rev. {\bf B}} \textbf{\bibinfo{volume}{36}},
1389: \bibinfo{pages}{4337} (\bibinfo{year}{1987}).
1390:
1391: \bibitem[{\citenamefont{K{\"o}hler et~al.}(1998)\citenamefont{K{\"o}hler, dos
1392: Santos, Beljonne, Shuai, Br{\'e}das, Holmes, Kraus, M{\"u}llen, and
1393: Friend}}]{Kohler}
1394: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{K{\"o}hler}},
1395: \bibinfo{author}{\bibfnamefont{D.~A.} \bibnamefont{dos Santos}},
1396: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Beljonne}},
1397: \bibinfo{author}{\bibfnamefont{Z.}~\bibnamefont{Shuai}},
1398: \bibinfo{author}{\bibfnamefont{J.~L.} \bibnamefont{Br{\'e}das}},
1399: \bibinfo{author}{\bibfnamefont{A.~B.} \bibnamefont{Holmes}},
1400: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Kraus}},
1401: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{M{\"u}llen}},
1402: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{R.~H.}
1403: \bibnamefont{Friend}}, \bibinfo{journal}{Nature}
1404: \textbf{\bibinfo{volume}{392}}, \bibinfo{pages}{903} (\bibinfo{year}{1998}).
1405:
1406: \bibitem[{\citenamefont{Lavrentiev et~al.}(1999)\citenamefont{Lavrentiev,
1407: Barford, Martin, Daly, and Bursill}}]{Lavrentiev}
1408: \bibinfo{author}{\bibfnamefont{M.~Y.} \bibnamefont{Lavrentiev}},
1409: \bibinfo{author}{\bibfnamefont{W.}~\bibnamefont{Barford}},
1410: \bibinfo{author}{\bibfnamefont{S.~J.} \bibnamefont{Martin}},
1411: \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Daly}}, \bibnamefont{and}
1412: \bibinfo{author}{\bibfnamefont{R.~J.} \bibnamefont{Bursill}},
1413: \bibinfo{journal}{Phys. Rev. {\bf B}} \textbf{\bibinfo{volume}{59}},
1414: \bibinfo{pages}{9987} (\bibinfo{year}{1999}).
1415:
1416: \bibitem[{\citenamefont{Beljonne}(2001)}]{Beljonne1}
1417: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Beljonne}},
1418: \bibinfo{howpublished}{Habilitation Thesis, Universite de Mons--Hainaut,
1419: Belgium} (\bibinfo{year}{2001}).
1420:
1421: \bibitem[{\citenamefont{Chandross et~al.}(1997)\citenamefont{Chandross,
1422: Mazumdar, Liess, Lane, Vardeny, Hamaguchi, and Yoshino}}]{Chandross2}
1423: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Chandross}},
1424: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Mazumdar}},
1425: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Liess}},
1426: \bibinfo{author}{\bibfnamefont{P.~A.} \bibnamefont{Lane}},
1427: \bibinfo{author}{\bibfnamefont{Z.~V.} \bibnamefont{Vardeny}},
1428: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Hamaguchi}},
1429: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Yoshino}},
1430: \bibinfo{journal}{Phys. Rev. {\bf B}} \textbf{\bibinfo{volume}{55}},
1431: \bibinfo{pages}{1486} (\bibinfo{year}{1997}).
1432:
1433: \bibitem[{\citenamefont{Ohno}(1964)}]{Ohno}
1434: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Ohno}},
1435: \bibinfo{journal}{Theor. Chim. Acta} \textbf{\bibinfo{volume}{2}},
1436: \bibinfo{pages}{219} (\bibinfo{year}{1964}).
1437:
1438: \bibitem[{\citenamefont{Castleton and Barford}(2002)}]{Castleton}
1439: \bibinfo{author}{\bibfnamefont{C.~W.~M.} \bibnamefont{Castleton}}
1440: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{W.}~\bibnamefont{Barford}},
1441: \bibinfo{journal}{J. Chem. Phys.} \textbf{\bibinfo{volume}{117}},
1442: \bibinfo{pages}{3570} (\bibinfo{year}{2002}).
1443:
1444: \bibitem[{\citenamefont{Ghosh et~al.}(2000)\citenamefont{Ghosh, Shukla, and
1445: Mazumdar}}]{Ghosh}
1446: \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Ghosh}},
1447: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Shukla}}, \bibnamefont{and}
1448: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Mazumdar}},
1449: \bibinfo{journal}{Phys. Rev. {\bf B}} \textbf{\bibinfo{volume}{62}},
1450: \bibinfo{pages}{12763} (\bibinfo{year}{2000}).
1451:
1452: \bibitem[{\citenamefont{Comoretto et~al.}(2000)\citenamefont{Comoretto,
1453: Dellepiane, Marabelli, Cornil, dos Santos, Br{\'e}das, and
1454: Moses}}]{Comoretto}
1455: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Comoretto}},
1456: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Dellepiane}},
1457: \bibinfo{author}{\bibfnamefont{F.}~\bibnamefont{Marabelli}},
1458: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Cornil}},
1459: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{dos Santos}},
1460: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Br{\'e}das}},
1461: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Moses}},
1462: \bibinfo{journal}{Phys. Rev. {\bf B}} \textbf{\bibinfo{volume}{62}},
1463: \bibinfo{pages}{10173} (\bibinfo{year}{2000}).
1464:
1465: \bibitem[{\citenamefont{Miller et~al.}(1999)\citenamefont{Miller, Yoshida,
1466: Yang, and Heeger}}]{Miller}
1467: \bibinfo{author}{\bibfnamefont{E.~K.} \bibnamefont{Miller}},
1468: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Yoshida}},
1469: \bibinfo{author}{\bibfnamefont{C.~Y.} \bibnamefont{Yang}}, \bibnamefont{and}
1470: \bibinfo{author}{\bibfnamefont{A.~J.} \bibnamefont{Heeger}},
1471: \bibinfo{journal}{Phys. Rev. {\bf B}} \textbf{\bibinfo{volume}{59}},
1472: \bibinfo{pages}{4661} (\bibinfo{year}{1999}).
1473:
1474: \bibitem[{\citenamefont{Guo et~al.}(1994)\citenamefont{Guo, Guo, and
1475: Mazumdar}}]{FGuo}
1476: \bibinfo{author}{\bibfnamefont{F.}~\bibnamefont{Guo}},
1477: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Guo}}, \bibnamefont{and}
1478: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Mazumdar}},
1479: \bibinfo{journal}{Phys. Rev. {\bf B}} \textbf{\bibinfo{volume}{49}},
1480: \bibinfo{pages}{10102} (\bibinfo{year}{1994}).
1481:
1482: \bibitem[{\citenamefont{Soos et~al.}(1992)\citenamefont{Soos, Etemad, Galvao,
1483: and Ramasesha}}]{Soos2}
1484: \bibinfo{author}{\bibfnamefont{Z.~G.} \bibnamefont{Soos}},
1485: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Etemad}},
1486: \bibinfo{author}{\bibfnamefont{D.~S.} \bibnamefont{Galvao}},
1487: \bibnamefont{and}
1488: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Ramasesha}},
1489: \bibinfo{journal}{Chem. Phys. Lett.} \textbf{\bibinfo{volume}{194}},
1490: \bibinfo{pages}{341} (\bibinfo{year}{1992}).
1491:
1492: \bibitem[{\citenamefont{Chandross and Mazumdar}(1997)}]{Chandross3}
1493: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Chandross}} \bibnamefont{and}
1494: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Mazumdar}},
1495: \bibinfo{journal}{Phys. Rev. {\bf B}} \textbf{\bibinfo{volume}{55}},
1496: \bibinfo{pages}{1497} (\bibinfo{year}{1997}).
1497:
1498: \bibitem[{\citenamefont{Beveridge and Jaffe}(1965)}]{Beveridge}
1499: \bibinfo{author}{\bibfnamefont{D.~L.} \bibnamefont{Beveridge}}
1500: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{H.~H.} \bibnamefont{Jaffe}},
1501: \bibinfo{journal}{J. Am. Chem. Soc.} \textbf{\bibinfo{volume}{87}},
1502: \bibinfo{pages}{5340} (\bibinfo{year}{1965}).
1503:
1504: \bibitem[{\citenamefont{Moore and Yaron}(1998)}]{Moore}
1505: \bibinfo{author}{\bibfnamefont{E.~E.} \bibnamefont{Moore}} \bibnamefont{and}
1506: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Yaron}}, \bibinfo{journal}{J.
1507: Chem. Phys.} \textbf{\bibinfo{volume}{109}}, \bibinfo{pages}{6147}
1508: (\bibinfo{year}{1998}).
1509:
1510: \bibitem[{\citenamefont{Mazumdar et~al.}(1992)\citenamefont{Mazumdar, Guo, and
1511: Dixit}}]{DGuo}
1512: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Mazumdar}},
1513: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Guo}}, \bibnamefont{and}
1514: \bibinfo{author}{\bibfnamefont{S.~N.} \bibnamefont{Dixit}},
1515: \bibinfo{journal}{J. Chem. Phys.} \textbf{\bibinfo{volume}{96}},
1516: \bibinfo{pages}{6862} (\bibinfo{year}{1992}).
1517:
1518: \bibitem[{\citenamefont{Harrison
1519: et~al.}(1999{\natexlab{a}})\citenamefont{Harrison, Moller, Weiser, Urbasch,
1520: Mahrt, B{\"a}ssler, and Scherf}}]{Harrison1}
1521: \bibinfo{author}{\bibfnamefont{M.~G.} \bibnamefont{Harrison}},
1522: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Moller}},
1523: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Weiser}},
1524: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Urbasch}},
1525: \bibinfo{author}{\bibfnamefont{R.~F.} \bibnamefont{Mahrt}},
1526: \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{B{\"a}ssler}},
1527: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{U.}~\bibnamefont{Scherf}},
1528: \bibinfo{journal}{Phys. Rev. {\bf B}} \textbf{\bibinfo{volume}{60}},
1529: \bibinfo{pages}{8650} (\bibinfo{year}{1999}{\natexlab{a}}).
1530:
1531: \bibitem[{\citenamefont{Harrison
1532: et~al.}(1999{\natexlab{b}})\citenamefont{Harrison, Urbasch, Mahrt, Giessen,
1533: B{\"a}ssler, and Scherf}}]{Harrison2}
1534: \bibinfo{author}{\bibfnamefont{M.~G.} \bibnamefont{Harrison}},
1535: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Urbasch}},
1536: \bibinfo{author}{\bibfnamefont{R.~F.} \bibnamefont{Mahrt}},
1537: \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Giessen}},
1538: \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{B{\"a}ssler}},
1539: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{U.}~\bibnamefont{Scherf}},
1540: \bibinfo{journal}{Chem. Phys. Lett.} \textbf{\bibinfo{volume}{313}},
1541: \bibinfo{pages}{755} (\bibinfo{year}{1999}{\natexlab{b}}).
1542:
1543: \bibitem[{\citenamefont{Wohlgennant et~al.}(2001)\citenamefont{Wohlgennant,
1544: Tandon, Mazumdar, Ramasesha, and Vardeny}}]{Wohl}
1545: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Wohlgennant}},
1546: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Tandon}},
1547: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Mazumdar}},
1548: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Ramasesha}},
1549: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{Z.~V.}
1550: \bibnamefont{Vardeny}}, \bibinfo{journal}{Nature (London)}
1551: \textbf{\bibinfo{volume}{409}}, \bibinfo{pages}{494} (\bibinfo{year}{2001}).
1552:
1553: \bibitem[{\citenamefont{Tandon et~al.}(2003)\citenamefont{Tandon, Ramasesha,
1554: and Mazumdar}}]{Tandon}
1555: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Tandon}},
1556: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Ramasesha}},
1557: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Mazumdar}},
1558: \bibinfo{journal}{Phys. Rev. {\bf B}} \textbf{\bibinfo{volume}{xx}},
1559: \bibinfo{pages}{xxxxxx} (\bibinfo{year}{2003}).
1560:
1561: \end{thebibliography}
1562:
1563:
1564: \pagebreak
1565:
1566: %
1567: \begin{table}
1568: \caption{Relative weights of the dominant contributions to the excited states of
1569: different oligomers of PPP and PPV, computed with the screened Coulomb
1570: parameters (see text). Note that the $d_3$ band does not occur in PPP}
1571: \begin{tabular}{lccccccc} \hline \hline
1572: Oligmer & PA Feature& State & $d_1 \rightarrow d_1^{*}$ &
1573: $d_{1} \rightarrow d_{2}^*$ &
1574: $d_{1} \rightarrow d_{3}^*$&
1575: $(d_1 \rightarrow d_{1}^{*})^2$ &
1576: $(d_1 \rightarrow l^{*})^2$ \\
1577: \hline \hline
1578: PPP4 & I & $2A_g$ & 0.5546 & 0.0387 &---&0.1121&--- \\
1579: & II & $3A_g$ & 0.0471 & 0.3396 &---&0.2164&---\\
1580: & & $4A_g$ & 0.0779 & 0.1787&---&0.2997 &0.0681 \\
1581: & III & $5A_g$ & 0.0385 & 0.2474 &---&0.1037 & 0.0210 \\
1582: & IV & $8A_g$ & --- & 0.3476 &---&0.1566 &0.0147 \\
1583:
1584: & & $10A_g$ & --- & 0.1848 &---& 0.0894 & 0.3289 \\\hline
1585: PPP5 & I & $2A_g$ & 0.6171 & --- &---&0.0903
1586: & --- \\
1587:
1588: & II & $4A_g$ & 0.1236 & ---&---&0.4688 &0.0695 \\
1589: & III & $7A_g$ & 0.1309 & --- &---&0.4390 & --- \\
1590:
1591: & IV & $9A_g$ & --- & 0.2284 & --- & 0.0543
1592: & 0.2516 \\
1593: \hline
1594: PPV3 & I & $2A_g$& 0.3872 &---&---&0.2847 &--- \\
1595: & II & $4A_g$&0.3042 &0.0488&---&0.3313 &0.0169 \\
1596: & III & $7A_g$ &0.2732&0.1458&--- &--- & --- \\
1597: & IV&10$A_g$&---&0.0338&0.1746&0.0256 &0.3193 \\ \hline
1598: PPV4 & I&2$A_g$& 0.5078&---&---&0.1884& ---\\
1599: &II&3$A_g$& 0.1983&---&---&0.4680 &---\\
1600: &III& 4$A_g$ & 0.4178&---&---&0.2483 &---\\
1601: && 5$A_g$& 0.0149&0.3366&---&0.2859& ---\\
1602: &IV&9$A_g$&0.0602&0.4555&---&0.0318& 0.0398\\
1603: &V&12$A_g$& ---&0.1140 &0.1460&0.02071& 0.2284\\ \hline
1604: \end{tabular}
1605: \label{tab-coeff}
1606: \end{table}
1607: %
1608: %
1609: %\end{document}
1610:
1611:
1612: %
1613: \end{document}
1614: