|
1. |
Carotenoid mediated B800–B850 coupling in LH2 |
|
PhysChemComm,
Volume 2,
Issue 8,
1999,
Page 34-40
Brent P. Krueger,
Preview
|
|
摘要:
Carotenoid mediated B800–B850 coupling in LH2 Brent P. Krueger,a Gregory D. Scholes,a Ian R. Gouldband Graham R. Fleming*a a Department of Chemistry, University of California, and Physical Biosciences Division, Lawrence Berkeley National Laboratory, Berkeley, CA 94720 USA b Department of Chemistry, Imperial College of Science, Technology and Medicine, London, UK SW7 2AY Received 21st April 1999, Accepted 17th June 1999, Published 20th July 1999 Ab initio calculations of the excited state properties of a portion of the LH2 complex consisting of a supermolecule of two B850 bacteriochlorophyll (BChl), one B800 BChl, and one rhodopin glucoside carotenoid show significant mixing between the BChl and carotenoid states. Calculation of the B800–B850 coulombic coupling, through the transition density cube method, both with and without the carotenoid reveal that the carotenoid affects an increase in the B800–B850 coupling via an indirect (superexchange) mechanism.This carotenoid-mediated coupling explains a portion of the difference between the B800–B850 energy transfer rate determined experimentally and through traditional (Förster) calculations. (a) Fig. 1 Four protomer units from the LH2 ring of Rps. acidophila.1 The phytyl chains of the BChl have been removed for clarity. (a) In the 2D image, the molecules are colored according to group: supermolecule 1 = green, supermolecule 2 = red, neither = blue. (b) In the 3D model, clicking on an atom reveals its group and chain labels (in Netscape Navigator’s information bar).The group label identifies the pigment (ABC = aB850, BBC = bB850, BC8 = B800, and RG1 = RG) and whether it belongs to supermolecule 1, 2, or neither (0). The chain label identifies the protomer unit (A, B, C, or D) to which the pigment belongs. I. Introduction The light harvesting complexes of purple bacteria have been heavily studied, particularly since the publication of the high resolution crystal structures for the peripheral light harvesting complexes (LH2) from two species.1,2 A primary thrust of recent experimental3–8 and computational7,9–13 work has been to understand the mechanisms for, and role of, electronic delocalization (excitonic interactions) between B850 bacteriochlorophyll pigments in LH2.14–16 In the present work, we utilize ab initio calculations to investigate the effect of electronic interactions between bacteriochlorophyll (BChl) and carotenoid molecules.Fig. 1 gives the structures and relative orientations of the relevant pigment molecules. The presence of the carotenoid modifies the electronic excited states and thus the transition densities of the BChl transitions, which has important consequences for electronic energy transfer (EET) in LH2. The B800–B850 EET timescale has been measured by various techniques to be 650–700 fs.7,17–20 Careful calculations by several different groups fail to reproduce this timescale, generally arriving at values of ~2 ps.7,17,21–24 One likely failure of previous calculations is the neglect of spectral overlap of the B800 with an upper excitonic component of the B850 ring that lies near 800 nm.7,9,12,14,25,26 However, this spectral overlap issue does not appear to account for the entire difference between calculations and experiments.27 We suggest an additional factor: perhaps the carotenoid, not considered in any of these previous calculations, plays a role in mediating the B800–B850 electronic coupling and, therefore, EET.This superexchange, or indirect coupling, picture has been briefly mentioned previously,7,21,28 but will be investigated in detail in the present work. Here, we describe delocalized electronic states calculated ab initio for a supermolecule consisting of one B800 BChl, one carotenoid, and two B850 BChl. The presence of the carotenoid significantly affects the B800 and B850 transition densities, and therefore, their coulombic coupling.This work focuses on the transition densities of the supermolecule compared to its unperturbed constituents PhysChemComm, 1999, 8 (b)(monomers) and the respective B800–B850 coulombic couplings. Future work will extend discussion to the B800– B850 energy transfer rate, by including the coulombic couplings calculated here along with spectral overlap calculations,27 which include the excitonic nature of the B850 ring and disorder. These full calculations reproduce both the magnitude and temperature dependence of the B800–B850 energy transfer rate.29 II. Electronic coupling II.A Coulombic contribution to the coupling The electronic coupling between the transitions of two molecules is critical in determining the rate of energy transfer between those molecules.Within the weak coupling limit,30 the rate of EET is given by |V | JDA 2 DA kDA 2c 1) ( d (1)a(2) (1)a V r 12 -1 ) rA eg ( A h = where V is the electronic coupling (cm–1) and J is the spectral overlap (cm). The spectral overlap is an energy conservation term, which, following Förster,30 is given by the overlap of the donor emission spectrum with the acceptor absorption spectrum. The total electronic coupling may be partitioned into coulombic and orbital overlap dependent (or short range) contributions. While the short-range interactions are significant at close separations, they are generally still small compared to the coulombic coupling term for strongly allowed singlet transitions.It has been shown recently that for the closely spaced B850 BChl, the orbital overlap terms comprise ~20% of the total B850–B850 interaction.9 When molecules are further apart (and not in van der Waals contact), as in the B800–B850 interaction, orbital overlap coupling mechanisms are negligible. Thus, in the present work direct orbital overlap coupling mechanisms are not explicitly considered between the B800 and B850 BChl. However, the effects of orbital overlap between the BChl and carotenoid are considered within the supermolecule calculations and, as will be shown, do significantly influence the calculated coulombic coupling.At all separations beyond the direct orbital overlap region, the total coupling is well described by a coulombic interaction which, when electron correlation is neglected, takes the form coul d 2 = coul = V ò 0 ¢ ò º 2 (d¢d | aa 12 Meg (r D D 4 e ¢ in which r-1 is the coulombic repulsion between electron 1 of the donor, which is de-excited from d to d, and electron 2 of the acceptor, which is simultaneously excited from a to a’. This coulombic coupling can also be expressed as the interaction between the transition densities, M, of the donor and acceptor ) M r | A | r D - where the transition density (as a function of position, r) is given by the product of ground and excited state wavefunctions (1) ) dr dr ¢(2) d t (2) (3) A D eg ds (4) Ng *Ne MN N SE V D ( ) r = ò integrated over spin.(See also McWeeney31 eqn. (14.6.8) and our previous work.21) The recently introduced transition density cube (TDC) method21,32 interacts full transition densities as in eqn. (3) to give the coulombic coupling for all donor–acceptor separations. In contrast, the standard transition dipole– transition dipole method reduces the transition densities to transition dipoles (the ideal dipole approximation) and is valid only for reasonably symmetric transition densities at moderate or large separations. For the interactions discussed in this work—unsymmetric transition densities at small separations—the TDC method is necessary for accurate evaluation of the coulombic coupling.While transition dipoles, along with the dipole–dipole orientation factor k, will be utilized for convenient description of the calculated transition densities, the TDC method is employed for all coulombic coupling calculations. II.B Superexchange Recently,21 we briefly examined the possible indirect role of the carotenoid, rhodopin glucoside (RG), in mediating B800–B850 electronic coupling through superexchange. This is usually treated in a perturbative approach in which the coupling of the mediator to the donor (VMD) and the mediator to the acceptor (VMA) perturbs the coupling of the donor and the acceptor. The superexchange component of the coupling is then - = VMD VMA (5) EM - E where ED is the energy of the donor and EM the energy of the mediator.In our previous work,21 we used eqn. (5) to estimate VSE for the B800–RG–B850 system. We found that a small additional coupling of 2–3 cm–1 between B800A and the aB850C of a next-to-nearest-neighbor protomer unit was mediated by the RG. (Throughout this work, we use A, B, C... subscripts to denote the protomer unit to which a chromophore belongs, as shown in Fig. 1). However, this perturbative type of superexchange calculation accounted for the presence of the RG carotenoid in only the simplest sense as the VMD and VMA couplings we employed did not include orbital overlap effects. II.C Multimer calculations In the present work, we show that this superexchange coupling plays a significant role in the B800–B850 interaction.By calculating the excited state properties of a supermolecule of B800A, RGB, bB850B, and aB850C, the presence of the carotenoid is included directly into the excited state calculation. Note that the bB850B–aB850C interaction is strong, so it will be more intuitive to consider the constituent B850 chromophores as a dimer with a single ground state (designated Dim) and two excited states: upper (+bBaC, or simply +) and lower (–bBaC, or simply –). This set of four pigments (see Fig. 1) embodies the largest carotenoid–BChl interactions, and we expect the strongest orbital overlap effects are captured by the supermolecule calculation. However, the single strongest bB800–aB850 interaction is not within this supermolecule, but is between two neighboring supermolecules.The strongest B800– B850 interaction is the coulombic coupling between the b B850B–aB850C dimer states and the B800B monomer. By interacting the multimer +bBaC and –bBaC transitiondensities with the multimer B800B transition density of the neighboring supermolecule, the most important B800– B850 coulombic coupling interactions are determined. Importantly, all coupling mechanisms between the B800 RGB, bB850B, and aB850C, and between the B800B, RGC, bB850C, and aB850D have been included in the B800– B850 coulombic interaction through the multimer transition densities of the supermolecule calculation. In other words, all of the strong interactions that distort the electronic structure of the B800 and B850 pigments (aside from the coupling between the B800B and the bB850B–aB850C dimer) have been included in the supermolecule transition densities prior to calculating the desired coulombic couplings between the B800B and the bB850B–aB850C dimer.j For example, let the unperturbed wavefunctions be given by yA, yRG, and yDim for the constituents of the first (B800A–RGB–bB850B–aB850C) supermolecule and by jB, RG, jDim for the constituents of the second (B800B–RGC–b C–aB850D). The ground states of the two B850 supermolecules are given by 0 Dim RG A (6) = 0 j yDim yRG y B = ¢B = Dim RG j¢ jj The multimer B800B excited state (denoted with a prime) is a linear combination of all unperturbed excited states RG RG jB j g B RG B + EB + + C.T. EV + - E B b» C.T.= l RG jj + a( ) j j j + b j j¢ ¢ ( j) ( ) B d(j j j¢) RG coul B coul - B E V - - E B RG + - + - B m j j j B Dim ~ + ~B-+ + l m j j j j RG-B + - ~B + l - m j j j j j -+-- -B » d From coupling calculations with the B850 dimer (data not shown) and previous work21,25 we roughly estimate that b » 170/–7 000 = –0.025, g» 2/–320 = –0.006, and d» – 5/740 = –0.007 (the energy differences are taken from spectroscopic data). The C.T. denotes charge transfer interactions. - B B-+ ~B-- g»-RG +B+B RG RG +B (8) -+ Dim RG (9) jj+++- - where the coefficients a, b, g, and d denote coulombic interactions with a» 1, and coul - B V - - E RG RG j ( ) ( ) ( ) ( ) ( ) ( ) The l and m are determined by orbital overlap mediated couplings in the same manner as the coulombic coefficients. Note that the direct orbital overlap couplings between the B800 and the dimer excited states are zero because of their large spatial separation, but that through superexchange (indicated by the tilde), the carotenoid can mediate non-zero coefficients for the B800-dimer charge transfer coefficients. Because the orbital overlap coupling is generally much smaller than the coulombic coupling, we estimate that the largest charge transfer coefficients are £20% of the comparable coulombic coefficients.A, Dim (7) j B The transition density for the multimer (mixed) B800 transition, is | (10) 0 B MB | = M+ | = + ¢ Dim Dim RG d + c + y yñá ¢ Similarly, the multimer +bBaC transition density is 0ñá ¢ | (11) A A + with the excited state wavefunction given by RG RG + RG A - ) ¢ ¢ RG AA ¢ V ¢ g (d ( (13a) Taking the additional approximations that exchange terms, which have been shown previously to be small,28,33,34 as well as all two coefficient terms [e.g. a = ( [ 2 A A | B BRG G R ¢ ¢ V y¢ = a ( ) y¢ y y + b ( ) y y¢ y ( ) A(y y y¢) (12) + C.T. where c » 1, and the other coefficients are analogous to the B800B state such that a = – g» 0.006, b » 100/–6680 = – 0.015, and d = 0.Interacting the transition densities as in eqn. (3), with d = 0 and the approximations a» 1 and c » 1, and using the notation of eqn. (2) (with the wavefunctions referred to simply by their subscripts) gives multi ) a = ( [ 2 ) ( ( + + | B B + + | Va + RG + G ) ¢| ¢ ¢ ¢ ¢ ¢ G R + a + G ) ) ¢g(+¢ A A | B BA A | A A | + A A | - - a + | B B RG| +RGRRGRG RGR RGRG )+ )+ )+ )+ b + )+ ¢ b(RG RG + + | + + + | - - ¢ ¢ ¢ ¢ b +G R+ b | - - b + C.T. + Ex. ¢) d(G R)+ ¢ ¢ ¢ )]¢) ] are )] ¢a b ( RGRG )+ ¢ ¢¢ | B B + + | + ¢¢ + + | b (g ( d ( ) ( b + )+ + ¢( + + | B B + + | - - ¢ ¢ ¢ ¢ ¢ ¢ ¢d( b ( zero gives multi + unp b( + C.T.(13b) Thus, the coulombic coupling calculated between the multimer transition densities includes effects from neighboring chromophores, primarily through the carotenoid mediated terms b(B’B½RGRG) and b(RGRG½++') and analogous C.T. terms. The neighboring chromophore effects can be quantified by comparing the above coupling with the coupling between the two unperturbed transitions, which is simply+ + | B B = 2 g( (14) ( ) ¢ ¢ Vunp is the third (and dominant) term of eqn. (13). Therefore, all remaining terms of eqn. (13) represent indirect, superexchange interactions that modify the B800– B850 coulombic coupling. We have calculated Vmulti and Vunp for the interaction of B800B with +bBaC and with –bBaC and will show that Vmulti and Vunp are considerably different. Therefore, the presence of the neighboring molecules, especially the carotenoid, represents a significant modification to the coulombic coupling that drives energy transfer.Note that we calculate the total Vmulti and Vunp, not the individual terms that makeup these couplings, which were presented above simply for descriptive purposes. III. Calculation methods Nuclear geometries for the molecules were taken from the crystal structure of the LH2 of Rps. acidophila1 with hydrogen atoms added using the GAUSSVIEW utility. It is possible that the reported (heavy atom) coordinates from the X-ray diffraction data do not correspond to the minimum of the free energy curve for this molecule, which in turn may not correspond to the minimum energy geometry that would be obtained from a geometry optimization of LH2.However, we are concerned here with excitation energies and, owing to the small reorganization energy of BChl in LH2,5 these should not be significantly affected by small uncertainties in the geometry. As discussed above, calculations were carried out on both an unperturbed (i.e. monomer) representation and a multimer representation. The unperturbed representation includes three different BChl grouped into two chromophores: the B800B monomer and the bB850B–a B850C interprotomer dimer, which gave unperturbed B800B, unperturbed +bBaC, and unperturbed –bBaC transition density cubes (TDCs). The multimer calculations were carried out on the B800A–RGB–bB850B–aB850C supermolecule, which yielded multimer B800A, multimer +bBaC, and multimer –bBaC TDCs (and a multimer RGB TDC).The multimer B800B TDC was generated by shifting the multimer B800A TDC to the B800B position. One important consideration within the TDC method is the size of the volume elements that make up the TDCs. Thorough analysis of the dependence of coupling strength accuracy on TDC element volume size32 has shown that large element volumes (coarsely grained TDCs) can lead to significant errors in the resulting coupling calculations. For the most sensitive interaction in LH2, the carotenoid–BChl interaction, element volumes 0.5 Å3 yield <5% error in the resulting coupling strengths, while element volumes of 0.25 Å3 give <2% error.32 In the present work, we have utilized element volumes of 0.42 Å3.The TDCs were generated using the GAUSSIAN 9435 suite of programs, though some modification was necessary to permit the large number of basis functions (over 2000) that were needed. Excited state calculations utilized the ab initio CI-Singles method36 with the 3-21G* basis set.37 Use of larger basis sets and more accurate excited state methodologies (such as CAS-SCF38–41) is desirable, but impractical given the size of system under consideration. The CI-Singles method is known to yield accurate ground and excited state molecular geometries,42 thus the shape of the CI-Singles wavefunctions must be qualitatively correct. Absolute energies and transition moments are more sensitive to electron correlation, thus the CI-Singles method overestimates the transition energies and the magnitudes of the transition moments.43–45 We correct approximately for the overestimation of transition moment magnitudes by scaling the charge density in each TDC by the factor mexp/ mcalc as was done previously.21 In this way, a transition dipole calculated from a properly scaled TDC matches an experimentally determined value.The unperturbed B800 TDC is easily scaled to an experimental transition dipole magnitude of 6.13 D. For the unperturbed +bBaC and –bBaC transitions, we use the scaling factor k determined from 2 2 2 + (15) m m k ÷= 2 exp C - B + B C m1 öø where mexp is, again, chosen to be that of monomeric BChl, 6.13 D.Scaling the multimer TDCs is substantially more complicated because the mixing of orbitals and transitions make an appropriate mexp difficult to determine. Examining the CI for the multimer +bBaC and –bBaC transitions reveals that they are composed of HOMO–LUMO configurations that are predominantly of B850 character. In addition, the calculated transition dipole magnitudes from the multimer +bBaC and –bBaC transitions differ by <3% from those of the unperturbed dimer. Therefore, the scaling factor for the multimer +bBaC and –bBaC transitions is determined as in eqn. (15) for the unperturbed case. The multimer B800 transition contains three CI components of significant magnitude.Two of the CI components correspond to the unperturbed B800 transition, while the third component is from the RG CI. The unperturbed B800 CI components make up 0.896 of the total multimer B800 CI, and the RG CI component makes up the remaining 0.104. Therefore, we have scaled the multimer B800 transition using a mixed scaling factor æ 2 çè Mscaled calc 2 = ( ) 0.896 k + 0.104 k M (16) where k2 is the same scaling factor found from the monomer RG transition (scaled to 13.0 D) and k1 is determined from 0.896 k1 mcalc = mexp = 6.13. While this scaling procedure may seem dubious given the complexities of the multimer states, it does connect the TDC coupling calculations to experimental transition magnitudes. Even in this complex system, where scaling factors are difficult to determine, the scaled TDCs provide reasonable approximations to the magnitudes of the multimer transition moments.It is reassuring to note that the ratio of the scaled multimer B800 and scaled multimer –bBaC transition dipoles is 0.819. When squared to give the ratio of oscillator strengths (0.8192 = 0.67), this reproduces the observed B800/B850 absorption ratio of 0.65.46 B, multimer –bBaC, and multimerBB IV. Discussion The multimer B800 +bBaC transition densities are shown in Fig. 2–4. Most striking is the delocalization of the multimer B800 and multimer +bBaC transitions, which have significant transition density extending along the carotenoid and even onto the other BChl. In contrast, the –bBaC transition, which carries most of the dimer oscillator strength, is effectively localized on the B850 dimer; its appearance is nearly identical to that of the unperturbed –bBaC transition.Trends in the changes of the properties of the transition densities are summarized in Table 1. The extensive delocalization of the +bBaC and B800B transitions is evidenced by large changes in dipole moment directions and in the centers of charge of the transition densities. In addition, an increase in oscillator strength of the B800 transition owing to mixing with the RG transition is also shown. Also of consequence is the small but significant change in transition dipole direction of the –bBaC transition, which provides a physical basis for the finding of Koolhaas et al.25 that simulations of the CD spectra of LH2 are improved by slightly tilting the B850 transition dipoles relative to directions expected from the crystallographic data.Fig.4 Transition density for the multimer +bBaC transition. These molecules correspond to supermolecule 1 from Fig. 1. Click here for a 3D representation. Fig. 2 Transition density for the multimer B800B transition. These molecules correspond to supermolecule 2 from Fig. 1. Click here for a 3D representation. The changes in the coupling strengths are substantial: a 43% increase in the B800B– –bBaC coupling, a 110% increase in the B800B– +bBaC coupling. It is significant that both B800–B850 interactions have increased considerably. The energy transfer timescale governed by these interactions has been poorly predicted by previous calculations as ~2 ps7,17,21–24 compared to 650–700 fs from experiment.7,17–20 The large increases in B800–B850 coupling strengths shown in Table 2 suggest that the calculations presented here have identified one source of discrepancy between previous calculations and experiment—the indirect coupling of the B800 and B850 BChl by the carotenoid.In principle, the increased B800– B850 couplings, as modified by the carotenoid, will provide an increase in the predicted B800–B850 energy transfer rate. In addition, recent work in this laboratory27 has suggested that excitonic interactions within the B850 ring as well as the site disorder in the electronic transition energies also play key roles in this energy transfer process.A more complete description of the B800–B850 energy transfer process, including the above effects with the modified couplings given here will be reported elsewhere.29 Fig. 3 Transition density for the multimer –bBaC transition. These molecules correspond to supermolecule 1 from Fig. 1. Click here for a 3D representation. The extended shape of the B800B and +bBaC transition densities suggests that they should be poorly described by transition dipoles, which is confirmed by the large difference (>25%) between VTDC and Vd–d in Table 2. While the symmetric –bBaC transition density should be well described by a transition dipole, it is surprising that VTDC and Vd–d agree so completely (<3% difference) for the B800B–bBaC interaction.Previous work on coupling strengths in LH221,32 suggests that the interaction between one symmetric and one unsymmetric transition density should be poorly described by Vd–d at these separations. Thus, it may be simply fortuitous that Vd–d accurately describes the B800B–bBaC interaction. These changes in the properties of the transition densities produce significant changes in the intermolecular interactions, which are quantified by determining the coulombic coupling between transition densities. The results of TDC calculations are given in Table 2. For reference, the dipole–dipole coupling (Vd–d),the dipole– dipole orientation factor [ k = uD·uA –3(uD·rDA)(uA·rDA)], and the center to center separation (R) of the centers of charge are also given.Table 1 Changes in transition dipole directions and the centers of charge between the unperturbed and multimer representations, along with the transition dipole magnitudes (scaled) of the three transitions in both the unperturbed and multimer representations Transition D center/Å D angle (º) 0.52 2.51 4.27 1.43 25.36 15.13 –bBaC +bBaC B800B Coupling/cm–1 Vd–d VTDC multi unp multi unp Acceptor 34 –29 28 –15 33 –40 23 –19 –bBaC +bBaC The values of the k factor and R suggest that the increase in the B800–B850 interactions between the unperturbed and multimer representations is mainly the result of the decreased separation between the transition densities.Fig. 2 and 4 show that because the carotenoid lies between the B800 and B850, its interaction with the BChl tends to draw each transition density toward the other, resulting in the decrease in separation. Table 2 Values of the TDC coupling, dipole–dipole coupling, orientation factor, and separation between the B800B donor transition and the given acceptor transition for both the unperturbed (unp) and multimer (multi) representations. Separation is the distance between the centers of charge of the transition densities. k is the dipole– dipole orientation factor defined in the text m 8.28 2.58 6.13 methods is necessary to quantitatively determine coupling strengths. We also note that, because the CI-Singles method has been employed in these calculations, the carotenoid S1 state, composed primarily of double excitations, has been ignored.Energetically, this state (roughly 13 000 cm–1, though this is not well-known) likely lies slightly above the B800 and B850 states (12 540 and 12 000 cm–1) and well below the S2 state (18 940 cm–1). Including the S1 state may increase the influence of the carotenoid on the BChl states, though without proper calculations we cannot exclude the possibility that there may be cancellation between the effects of the S1 and S2 states. Remarkably, despite large changes in the transition dipole directions for two of the three transitions, the k values change very little between the unperturbed and multimer representations. For the B800–B850 interactions, this must be, to some extent, fortuitous as the orientation factor is a complicated function of both the transition dipole directions and the relative center-to-center positions, which are all perturbed significantly.However, it may be the case that the carotenoid shifts the B800B and +bBaC transitions in a parallel fashion such that, despite large changes in the transition dipole directions of both, their relative orientation is effectively unchanged. Because the B800B transition changes significantly and the –bBaC transition changes very little, it is not surprising that the change in relative orientation is most significant for this pair of transitions. The electronic couplings given here are based on certain approximations discussed in detail in ref.9. It is difficult to estimate absolute errors in these approximations; however, because we can predict absorption and CD spectra that closely resemble those measured by experiment27 using these couplings, together with lineshape functions determined from analysis of three pulse echo data,6 we believe that our estimates are reasonable. However, we emphasize that because limitations of the 3-21G* basis set cause the charge transfer terms of eqn. (13) to be underestimated, and use of the CIS method causes the coulombic terms of eqn. (13) to be overestimated, a study involving more complete basis sets and computational D multi D unp Orientation, k multi multi unp m 8.30 2.50 6.80 Separation, R /Å unp 17.1 17.0 15.2 13.8 0.41 –0.91 0.54 –0.91 V.Conclusions We have reported the results of ab initio calculations of the excited state properties of a supermolecule made up of the B800A, RGB, bB850B, and aB850C pigments from the LH2 of Rps. acidophila. The resulting transition densities show significant perturbation due to orbital mixing compared to calculations on the unperturbed B800 and bB850B–aB850C dimer. These effects are clearly demonstrated by the multimer B800B (identical to the multimer B800A) and +bBaC transition densities, which show delocalization extending along the carotenoid and onto the other BChl. This orbital mixing effect tends to draw the transition densities closer together, increasing the B800–B850 electronic coupling.Therefore, the presence of the carotenoid mediates the B800–B850 coupling through an indirect, superexchange interaction. This indirect interaction explains a portion of the difference between previous 2 ps estimates of the B800–B850 energy transfer timescale7,17,21–24 and the experimentally observed 650–700 fs.7,17–20 Acknowledgements We thank L. Butler and C. Hsu for helpful comments. This work was supported by a grant from the NSF.References 1 G. McDermott, S. M. Prince, A. A. Freer, A. M. Hawthornthwaite-Lawless, M. Z. Papiz, R. J. Cogdell and N. W. Isaacs, Nature, 1995, 374, 517–521. 2 J. Koepke, X. Hu, C. Muenke, K. Schulten and H. Michel, Structure, 1996, 4, 581–597.3 J. T. M. Kennis, A. M. Streltsov, H. Permentier, T. J. Aartsma and J. Amesz, J. Phys. Chem. B, 1997, 101, 8369–8374. 4 R. Monshouwer, M. Abrahamsson, F. van Mourik and R. van Grondelle, J. Phys. Chem. B, 1997, 101, 7241–7248. 5 R. Jimenez, F. van Mourik, J. Y. Yu and G. R. Fleming, J. Phys. Chem. B, 1997, 101, 7350–7359. 6 J.-Y. Yu, Y. Nagasawa, R. van Grondelle and G. R. Fleming, Chem. Phys. Lett., 1997, 280, 404–410. 7 T. Pullerits, S. Hess, J. L. Herek and V. Sundström, J. Phys. Chem. B, 1997, 101, 10560–10567. 8 T. Pullerits, M. Chachisvillis and V. Sundström, J. Phys. Chem., 1996, 100, 10787–10792. 9 G. D. Scholes, I. R. Gould, R. J. Cogdell and G. R. Fleming, J. Phys. Chem. B, 1999, 103, 2543–2553. 10 M. G. Cory, M. C. Zerner, X.Hu and K. Schulten, J. Phys. Chem. B, 1998, 102, 7640–7650. 11 K. Sauer, R. J. Cogdell, S. M. Prince, A. Freer, N. W. Isaacs and H. Scheer, Photochem. Photobiol., 1996, 64, 564–576. 12 O. Kühn and V. Sundström, J. Phys. Chem. B, 1997, 101, 3432–3440. 13 J. A. Leegwater, J. Phys. Chem., 1996, 100, 14403–14409. 14 V. Sundström, T. Pullerits and R. van Grondelle, J. Phys. Chem. B, 1999, 103, 2327–2346. 15 O. Kühn, T. Renger, V. May, J. Voigt, T. Pullerits and V. Sundström, Trends in Photochem. Photobiol., 1997, 4, 213. 16 G. R. Fleming and R. van Grondelle, Curr. Opin. Struct. Biol., 1997, 7, 738–748. 17 R. Jimenez, S. N. Dikshit, S. E. Bradforth and G. R. Fleming, J. Phys. Chem., 1996, 100, 6825–6834. 18 A. P. Shreve, J. K. Trautman, H.A. Frank, T. G. Owens and A. C. Albrecht, Biochim. Biophys. Acta, 1991, 1058, 280–288. 19 S. Hess, F. Feldchtein, A. Rabin, I. Nurgaleev, T. Pullerits, A. Sergeev and V. Sundström, Chem. Phys. Lett., 1993, 216, 247– 257. 20 T. Joo, Y. Jia, J.-Y. Yu,D. M. Jonas and G. R. Fleming, J. Phys. Chem., 1996, 100, 2399–2409. 21 B. P. Krueger, G. D. Scholes and G. R. Fleming, J. Phys. Chem. B., 1998, 102, 5378–5386. 22 S. Hess, Visscher, K. J. T. Pullerits, V. Sundström, G. J. S. Fowler and C. N. Hunter, Biochemistry, 1994, 33, 8300–8305. 23 G. J. S. Fowler, S. Hess, T. Pullerits, V. Sundström and C. N. Hunter, Biochemistry, 1997, 36, 11282–11291. 24 S. V. Kolaczkowski, J. M. Hayes and G. J. Small, J. Phys. Chem., 1994, 98, 13418–13425.25 M. H. C. Koolhaas, R. N. Frese, G. J. S. Fowler, T. S. Bibby, S. Georgakopoulou, G. van der Zwan, C. N. Hunter and R. van Grondelle, Biochemistry, 1998, 37, 4693–4698. 26 H. M. Wu, S. Savikhin N. R. S. Reddy, R. Jankowiak, R. J. Cogdell, W. S. Struve and G. J. Small, J. Phys. Chem., 1996, 100, 12022–12033. 27 G. D. Scholes and G. R. Fleming, manuscript in preparation. 28 G. D. Scholes, R. D. Harcourt and G. R. Fleming, J. Phys. Chem. B, 1997, 101, 7302–7312. 29 G. D. Scholes, B. P. Krueger, I. R. Gould and G. R. Fleming, manuscript in preparation. 30 T. Förster, 'Delocalized Excitation and Excitation Transfer', in Modern Quantum Chemistry, ed. O. Sinanoglu, Academic Press, Inc., New York, 1965, vol. III, pp. 93–137. 31 R. McWeeny, Methods of Molecular Quantum Mechanics, 2nd edn., Academic Press, London, 1992. 32 B. P. Krueger, PhD Thesis, The University of Chicago, Chicago, 1999. 33 H. Nagae, T. Kakitani, T. Katoh and M. Mimuro, J. Chem. Phys., 1993, 98, 8012–8023. 34 A. Damjanovic, T. Ritz and K. Schulten, Phys. Rev. E, 1999, 59, 3293–3311. 35 M. J. Frisch, G. W. Trucks, H. B. Schlegel, P. M. W. Gill, B. G. Johnson, M. A. Robb, J. R. Cheeseman, T. Keith, G. A. 36 J. B. Foresman, M. Head-Gordon, J. A. Pople and M. J. Frisch, Petersson, J. A. Montgomery, K. Raghavachari, M. A. Al- Laham, V. G. Zakrzewski, J. V. Ortiz, J. B. Foresman, J. Cioslowski, B. B. Stefanov, A. Nanayakkara, M. Challacombe, C. Y. Peng, P. Y. Ayala, W. Chen, M. W. Wong, J. L. Andres, E. S. Replogle, R. Gomperts, R. L. Martin, D. J. Fox, J. S. Binkley, D. J. Defrees, J. Baker, J. P. Stewart, M. Head- Gordon, C. Gonzalez and J. A. Pople, GAUSSIAN 94, revision D.4, Gaussian Inc., Pittsburgh, PA, 1995. J. Phys. Chem., 1992, 96, 135. 37 P. C. Hariharan and J. A. Pople, Theor. Chim. Acta, 1973, 28, 213. 38 D. Hegarty and M. A. Robb, Mol. Phys., 1979, 38, 1795. 39 R. H. E. Eade and M. A. Robb, Chem. Phys. Lett, 1981, 83, 362. 40 P. E. M. Siegbahn, J. Amlöf, J. Heiberg and B. O. Roos, J. Chem. Phys., 1981, 74, 2384. 41 B. O. Roos, P. R. Taylor and P. E. M. Siegbahn, Chem. Phys., 1980, 48, 157. 42 U. Lommatzsch and B. Brutschy, Chem. Phys., 1998, 234, 35– 57. 43 J. B. Foresman and Æ. Frisch, Exploring Chemistry with Electronic Structure Methods, 2nd edn., Gaussian Inc., 44 J. M. O. Matos, B. O. Roos and P.-A. Malmqvist, J. Chem. 45 J. M. O. Matos and B. O. Roos, Theor. Chim. Acta, 1988, 74, 46 R. K. Clayton and B. J. Clayton, Proc. Natl. Acad. Sci. USA, Pittsburgh, PA, 1996. Phys., 1987, 86, 1458. 363. 1981, 78, 5583–5587. 9/03172C PhysChemComm © The Royal Society of Chemistry 1999
ISSN:1460-2733
DOI:10.1039/a903172c
出版商:RSC
年代:1999
数据来源: RSC
|
|