Control of retinal isomerization in bacteriorhodopsinin the high-intensity regimeAndrei C. Floreana, David Cardozab, James L. Whiteb, J. K. Lanyic, Roseanne J. Sensiona,1, and Philip H. Bucksbaumb,1

aDepartment of Physics, University of Michigan, Ann Arbor, MI 48109; bPULSE Institute and Department of Physics, Stanford University, Stanford, CA 94305;and cSchool of Medicine, University of California, Irvine, CA 92697

Contributed by Philip H. Bucksbaum, May 20, 2009 (sent for review October 13, 2008)

A learning algorithm was used to manipulate optical pulse shapesand optimize retinal isomerization in bacteriorhodopsin, for exci-tation levels up to 1.8 � 1016 photons per square centimeter. Below1/3 the maximum excitation level, the yield was not sensitive topulse shape. Above this level the learning algorithm found that aFourier-transform-limited (TL) pulse maximized the 13-cis popula-tion. For this optimal pulse the yield increases linearly with inten-sity well beyond the saturation of the first excited state. Tounderstand these results we performed systematic searches vary-ing the chirp and energy of the pump pulses while monitoring theisomerization yield. The results are interpreted including the in-fluence of 1-photon and multiphoton transitions. The populationdynamics in each intermediate conformation and the final branch-ing ratio between the all-trans and 13-cis isomers are modified bychanges in the pulse energy and duration.

coherent control � photoisomerization � ultrafast science

Bacteriorhodopsin (bR) is a photosynthetic protein found inthe purple membrane of Halobacterium salinarum and ca-

pable of conversion of solar energy into chemical energy. Thisenergy conversion is efficient (1–3) and has several possibleapplications (4–12). A retinal chromophore is responsible forphoton absorption. After photoexcitation retinal undergoesultrafast isomerization from the all-trans to a 13-cis configura-tion, accompanied by additional changes in the conformation ofbR (3, 8). The initial steps of the bR photocycle (see Fig. 1) havebeen studied intensively (11, 13–33), but there are still unan-swered questions regarding the electronic potential energy sur-faces (PES) of retinal, the interaction with its surroundings in theprotein, and related ultrafast vibrational coupling. A number ofmodels have been proposed, each explaining parts of the largenumber of experiments (4, 13, 14, 16, 34–40). Attempts havebeen made to reconcile the differences between these models(4).

We aim to understand how the optical pulse shape andintensity affect the all-trans 3 13-cis yield and to explorepotential pathways for producing high photoproduct yields on anultrafast time scale. This is relevant for energy storage usingbio-molecular machines (9, 32). Recently, Prokhorenko et al.showed that the isomerization yield of retinal in bR could bemanipulated in a low intensity, biologically relevant regimethrough the use of phase and amplitude shaped optical fields(25). Modifications of as much as �20% were observed com-pared with unshaped pulses capable of exciting an equal numberof molecules. Yet the ultimate yields remain small as photon fluxwas restricted to excite �0.3% of the chromophores in theexcitation volume. In a different experiment, Vogt et al. usedmuch higher intensity, shorter wavelength pump pulses to excitebR and a shaped 800-nm dump pulse to study the evolution ofthe molecule on the excited state PES (30). They found that theexcited population is transferred most effectively back to theall-trans state by means of a near-infrared ‘‘dump’’ pulse with adelay of 200 fs, and with the minimum bandwidth for its pulseduration (i.e., a TL pulse). This is consistent with the dynamicsof the I460 intermediate state. We investigate the pulse shape

dependence of the isomerization yield at pulse excitation levelsup to 2 orders of magnitude higher than previous studies (25),so that pump-dump mechanisms or multiphoton excitation canhelp to control the molecular dynamics. These studies provideinformation on excited state dynamical mechanisms that pro-duce increased yields of photoproduct. We find that short,intense pulses increase the isomerization yield by �50% over thesame energy delivered in long, low intensity pulses. The yieldincreases approximately linearly well beyond the saturation ofthe initial 1-photon transition. To investigate the mechanism weperform a number of systematic scans while monitoring signalsproportional to the population of the initial conformationsinvolved in the isomerization reaction. We find that higherexcited states (Sn) play a significant role in enhancing the yield.We present a model describing the multipathway isomerizationprocess.

ResultsWe report on 5 types of transient absorption experiments. Thefirst type is white-light continuum absorption spectroscopy from

Author contributions: R.J.S. and P.H.B. designed research; A.C.F., D.M.C., and J.L.W. per-formed research; J.K.L. contributed new reagents/analytic tools; A.C.F., D.M.C., and J.L.W.analyzed data; and A.C.F., R.J.S., and P.H.B. wrote the paper.

The authors declare no conflict of interest.

1To whom correspondence may be addressed. E-mail: rsension@umich.edu or phb@slac.stanford.edu.

This article contains supporting information online at www.pnas.org/cgi/content/full/0904589106/DCSupplemental.

Fig. 1. Scheme of the isomerization reaction of retinal in bR. H � theconformation excited vertically from the all-trans (bR570) ground state. Sn �the higher excited state reached through absorption from the H state. I460 isthe conformer corresponding to the shallow bottom of the first excitedelectronic potential energy surface. The red arrow depicts the transition fromI460 to the ground all-trans conformation. The J625 conformer is tentativelyassigned to the conical intersection region (see discussion below). K590 desig-nates the 13-cis isomer.

10896–10900 � PNAS � July 7, 2009 � vol. 106 � no. 27 www.pnas.org�cgi�doi�10.1073�pnas.0904589106

Dow

nloa

ded

by g

uest

on

Dec

embe

r 18

, 202

0

460 to 870 nm. Here, the data demonstrate that the photoprod-uct observed at long times (38–40 ps) is consistent with theformation of the K intermediate. The remaining 4 experimentsdetect narrow-band time-delayed spectral absorption to deter-mine how intermediate state formation is controlled by theshape of the excitation pulse (see Fig. 2), A learning algorithmwas used to allow multiparameter modification of the excitationpulses. Searches were performed to identify optimal pulses toenhance and minimize photoproduct formation probed at 650nm 30–40 ps after excitation. Photoproduct formation was alsoanalyzed using 1 parameter searches varying linear chirp andpulse intensity. Finally transient absorption kinetic traces wereobtained at a range of pulse energies and chirps to probe the timeevolution of the various photoproduct signals. These experi-ments are discussed in more detail below.

Learning Control Experiments. Optical control experiments werecarried out to find the optimal phase profile of the pump pulseto maximize the yield of the 13-cis photoproduct. A geneticalgorithm (GA) search (41, 42) was used to find the optimalpulses that maximize the absorption signal at 650 nm at a timedelay of 40 ps after excitation (see Fig. 2). This is the wavelengthwith largest differential absorption between the 13-cis andall-trans state. The time delay was sufficient to assure that all ofthe population had relaxed to the all-trans or 13-cis conforma-tion. At pump energies �30 nJ, the GA consistently convergedto a transform-limited (TL) pulse solution. Fig. 3 Inset showsfitness vs. generation for a typical GA optimization. The blackupper curve shows the best pulse shape in each generation of theevolutionary search whereas the gray curve shows the averageperformance of all pulse shapes in that generation. Both reachasymptotic values after �20 generations. The dashed line showsthe 13-cis absorption for a TL pulse, and the optimizationexperiment converges on this value. At pump energies �30 nJthe search algorithm fails to improve the isomerization yield forany pulse shape. We also tried to minimize K production usingthe GA. The pulses retrieved from these minimizations were alllong pulses (� 500 fs) and had no distinguishing amplitudefeatures in the time domain and no distinguishing spectral phasefeatures. This is a common result for minimizations performedby the GA, particularly when optimization is based on multipho-ton absorption and peak intensity is the controlling factor.

Intensity Scans. To explore further the results of these optimiza-tion experiments we performed intensity scans in which thebR570, I460, J625, and K590 populations were monitored as a

function of pump energy for TL excitation pulses at specificpump-probe wavelengths and delays (see Fig. 2). The results areplotted in Fig. 3. The green curves show the decrease (bleach)of the bR570 absorption 100 fs and 200 fs after photoexcitation.This signal is proportional to the population removed from theground state and has a sublinear dependence on the excitationenergy. Previous studies reported that the population is initiallyexcited to the H conformation and a fluorescence Stokes shift isobserved as the population moves to I460 in �200 fs (16, 23, 43).This time-delay in the growth of the I460 signal is also observedin the present experiments (see section 2.4 below). The intensity-dependent difference between the bR570 bleach at 200 fs (dashedgreen line) and 100 fs (solid green line) as pump energies exceed30 nJ reflects the appearance of an absorbing species or deple-tion of an emissive species on an ultrafast time scale. This mayindicate that a fraction of the initially excited population returnsrapidly to the all-trans ground state. Alternatively the absorptionmay arise from the formation of vibrationally hot I460 or anotherintermediate excited state species after multiphoton excitationto higher electronic states. The intensity dependence of the I460population was monitored via stimulated emission at 850 nm andexcited state absorption at 487 nm. Both I460 signals saturatemore strongly than the all-trans bleach, demonstrating that someof the bleached population is excited from H to higher states,avoiding the I460 state altogether.

Fig. 2. bR absorption spectrum (red line), pump spectrum (green line) andprobe spectra (black line) for the 4 monitored wavelengths. The transitionlabels follow the nomenclature of Fig. 1. The J625 and K590 signals are moni-tored at the same wavelength (650 nm) at early (�2 ps) and late (�30 ps)pump-probe delays.

Fig. 3. Summary of absorption data. (Inset) Fitness as a function of gener-ation for a typical GA run. The black line represents the fitness (intensity of theK absorption) of the best individual pulse in each generation, the gray linerepresents the average fitness for each generation and the thin dashed linerepresents the fitness of the worst pulse in each generation. The red linerepresents the fitness of a TL pulse. The GA almost doubles the intensity of theK absorption over that of random stretched pulses in the early generations,converging on the K absorption intensity characteristic of transform-limitedpulses. (Main Figure) I460, bR570, J625 and K590 signals versus pump energy fortransform-limited excitation pulses. The horizontal scale represents the esti-mated energy entering the focal volume where the pump and probe pulsesoverlap. The corresponding excitation level is given along the top axis. Thearrow indicates the estimated linear saturation level of 4.8 � 1015 photons persquare centimeter (25). The data are scaled to give similar linear behavior atlow pump intensities, highlighting the differences in saturation at higherpump intensities. The I460 signal is given by the stimulated emission (SE) at 850nm (red curve) and excited state absorption (ESA) at 487 nm (black curve). Theinitial bR570 bleach is measured at 570 nm 100 fs (solid green) and 200 fs(dashed green) after the coherent spike (zero pump-probe delay). The J625 andK590 signals are given by the transient absorption at 650 nm, at 2 ps (light blue)and 40 ps (dark blue) respectively. The K and J absorption signals continue toincrease with pump-intensity despite the saturation of the bleaching signaland the I460 population.

Florean et al. PNAS � July 7, 2009 � vol. 106 � no. 27 � 10897

BIO

CHEM

ISTR

YA

PPLI

EDPH

YSIC

AL

SCIE

NCE

S

Dow

nloa

ded

by g

uest

on

Dec

embe

r 18

, 202

0

The photoproduct populations at 650 nm exhibit a verydifferent intensity dependence. The 13-cis population, K590(dark blue), observed 40 ps after excitation exhibits a nearlylinear dependence on intensity. The J625 absorption monitoredat 2 ps (light blue) displays a similar dependence on the pumpenergy, although there may be a slight energy dependent mis-match between the J625 and K590 curves. The pulse energydependence of the bR570 bleach at 570 nm 30 ps after excitationis consistent with the K590 and J625 signals as expected fordepletion of the reactant trans-retinal and production of thephotoproduct. The near linear dependence of the ultimatephotoproduct yield is in distinct contrast to the sublinear be-havior ofthe initial bR570, bleach and the I460 emission. Theincrease in the K590 yield can be maintained only if the branchingefficiency toward this conformation increases with intensity.

Linear Chirp Scans. The influence of pulse shape on isomerizationyield was explored further by monitoring the different photo-product signals while systematically varying the second orderspectral phase (linear chirp) of the excitation pulses. Linear chirpcan be a useful tool to probe the influence of pulse shape onexcited state population or wavepacket dynamics in absorptionto 1-photon allowed excited states and to probe the spectralresonances in resonant 2-photon absorption (44). The magnitudeof the K590 absorption signal depends on linear chirp only atenergies �30 nJ whereas at low energies the chirp scans are flatwithin the � 5% RMS noise of the measurement. Fig. 4compares the K590 (red squares) and J625 (blue diamonds)absorption signals as a function of linear chirp for 80 nJexcitation pulses. Both signals are maximized by a pulse with asmall negative chirp, near the limit of the temporal resolution.These data suggest that the higher excited state responsible forthe enhanced production of K is primarily accessed from theFranck–Condon region of the initially excited H state (Fig. 1).Evolution of the sample away from this region reduces theoverall signal from this channel as evinced in the rapid drop-offwith chirp. We also observe a small asymmetry for pulses ofopposing linear chirp. This asymmetry, which is seen as a steeperdrop in photoproduct formation with positive chirp, could be dueto the slightly asymmetric excitation spectrum. It could alsoreflect the influence of a pump-dump pathway reducing theexcited state population. Excited state depletion is often ob-served with positively chirped excitation. The difference be-tween the early (J625) and late (K590) signals reflects an influence

of excitation pathway (1-photon vs. 2-photon) on the photo-product relaxation dynamics.

Transient Scans. The above analysis maps the multipathway evo-lution of the retinal chromophore but does not explain themechanism through which the branching ratio is controllable. Toaddress this question we performed transient scans with severaldifferent excitation pulses while monitoring the dynamics of theconformations involved in the isomerization process. Fig. 5 Topshows transient scans taken at 570 nm, the peak of the bR570ground state absorption. The absorption at 570 nm has contri-butions from the all-trans conformer and intermediates I460, J625and K590, and this complicates the interpretation of the bleachrecovery. Nevertheless, the bR570 bleach dominates the signal atthis wavelength, particularly at early times, before J625 and K590come into play. The decay of the bleach has 3 notable features.The first feature is an intensity dependent absorption signal atearly times that is altered by the pump phase and energy. Thisfeature is responsible for the difference in the intensity depen-dence 100 fs and 200 fs after excitation (see Fig. 3). At low pumpexcitation levels, there is an increase in the bleaching signal atearly times followed by a decay, whereas at higher energies thereis an enhanced bleach at early times followed by a somewhatslower decay leading to a plateau between �100 and 200 fs. Thesecond feature, whose onset is marked by the vertical arrow atT � T0 � 250 fs, can be modeled as an exponential decay witha time constant T1 � 600 fs. This is responsible for most of thedecay of the bleach corresponding to repopulation of the bRground state or production of the J625-K590 states, which alsoabsorb weakly at this wavelength. The third feature has anintensity dependent time constant T2 � 5–15 ps correspondingto relaxation of the photoproducts. The relative magnitude of theoverall bleach recovery is dependent on pump energy.

Fig. 5 Middle shows the intensity-dependent dynamics of thetransient absorption signal at 650 nm, where the J625 3 K590transition is monitored. At early times this signal also contains

Fig. 4. Probe absorption change vs. chirp of the pump pulse. (Upper) The J625

(blue diamonds) and K590 (red squares) signals versus linear chirp at 80 nJ pulseenergy. (Lower) K590 signal versus linear chirp at 12 nJ pulse energy (greentriangles). All signals are measured at 650 nm. The J625 signal is measured at 2ps and the K590 signal is measured at 40 ps.

Fig. 5. Transient scans taken at 570 nm (Top), 650 nm (Middle) and 850 nm(Bottom). The monitored signals are, in order: the bR570 (all-trans) groundstate bleach the J625 absorption and the I460 stimulated emission. The arrow inTop indicates the onset of the exponential recovery of the bleach. The neg-ative chirp pulse has a linear chirp rate of �0.35 � 105 fs2. In Middle, the traceshave been normalized by the amplitude of the corresponding I460 trace,generated in identical excitation conditions.

10898 � www.pnas.org�cgi�doi�10.1073�pnas.0904589106 Florean et al.

Dow

nloa

ded

by g

uest

on

Dec

embe

r 18

, 202

0

contributions due to stimulated emission as the fluorescenceundergoes a dynamic stokes shift (43). At low excitation intensitythis stimulated emission feature is clearly observed, decaying ona �150 fs time scale as the fluorescence spectrum shifts to thered. The observed rise times of the absorption due to the J625intermediate is within 10% of 400 fs for all pump energies used.The ensuing decay of the 650-nm signal is caused by therelaxation toward the K590 conformer. For excitation with 80 nJpulses the relaxation is completed within 30–40 ps. For lowerenergies the relaxation is completed within 10–20 ps. Transientscans with lengths up to 150 ps show that no changes areobserved after 40 ps for all traces. Transient absorption spectraat 38–40 ps were consistent with the formation of J-K with noother species contributing to the difference spectrum between460 nm and 870 nm. The 3 traces shown in Fig. 5 are normalizedusing the amplitude of the corresponding I460 signal generatedunder identical excitation conditions. The relative intensity ofthe J625 signal obtained with the highest intensity excitation issignificantly larger than that obtained with low excitation inten-sity. This supports the conclusion that some of the excitedpopulation reaches the photoproduct channel without passingthrough I460, in agreement with the saturation of the I460 curvein Fig. 3.

Fig. 5 Bottom shows the time dynamics of the stimulatedemission of the I460 state, measured at 850 nm. Unlike themeasurements performed at other probe wavelengths, which areaffected by spectral congestion from multiple intermediates, the850-nm optical signal observed at low excitation intensity can beassigned to I460 stimulated emission alone. As the intensity isincreased a fast negative transient appears in the data where thepump and probe pulses overlap in time. This transient could arisefrom a Raman contribution to the signal or a short-livedstimulated emission from the state excited after 2-photon ab-sorption. The fast transient is followed by the rise and decay ofthe stimulated emission signal. The stimulated emission signalattributed to I460 is not well described by a sum of exponentials.The signal reaches a maximum 200 fs after T0 with a rise that isdistinctly nonexponential. The decay of the signal is describedmoderately well by an exponential decay with a 1/e time from thepeak of the signal of �600 fs with TL 80 nJ excitation, 680 fs witha chirped 80 nJ excitation (linear chirp rate of �0.35 � 105 fs2)and 750 fs with TL 37 nJ excitation, although the overall signaldecay at this wavelength is better described as multiexponential.The change in the decay of I460 may represent changes in therelative amplitudes of different components that are not welldefined in the short 2-ps time window available. The I460 decayand bR570 bleach recovery occur on similar time scales confirm-ing that the I460 population is the dominant source for the bR570bleach recovery under all excitation conditions.

These data lead to the question of how intensity influences theI460 population dynamics. Studies have shown that I460 can bealso reached through the all-trans3H3 Sn3 I460 pathway andthe corresponding I460 population has a faster decay (17, 30, 43).Transient absorption scans taken at 487 nm showed features atT � T0 � 200 fs that did not match well the cross correlationmeasurements in the neat buffer solution. The mismatch couldbe due to fast relaxation of some of the Sn population to I460. Thiscomponent of the I460 population will correspond to hottervibrational levels and will easily bypass the small barrier to theright (see Fig. 1). It is not clear from the available data whetherthe influence of pulse parameters on the I460 population is simplydescribed by modifications in the peak intensity of the pulse andthe relative importance of 1-photon and multiphoton pathwaysor whether more complicated pathways could play a role. Thisis a subject worth careful investigation. An alternative way tocreate a hot I460 population in a controlled fashion is throughstimulated Raman scattering between Sn and H. This mechanism

is not mentioned in the literature, but it is consistent with theintensity-dependent effects observed here.

Discussion and ConclusionsThe model emerging from the data (see Fig. 1) demonstratesthe presence of at least 2 pathways for the formation of the13-cis K590 conformation. With low intensity excitation andphase-only control the GA was unable to identify pulse shapescapable of modifying the formation of the photoproduct. Thisis in direct contrast to the control observed in the study in ref.25. The contrast in controllability may relate to the differencesin available bandwidth or peak wavelength, but highlights theneed for additional work to characterize the mechanisms forcontrol.

In the present study high-intensity excitation opens pathwaysfor one or more additional excited states Sn, reached throughmultiphoton excitation. The GA consistently found TL pulses tooptimize the photoproduct formation whereas anti-optimizationminimizing photoproduct formation led to long pulses withoutany outstanding features. It is interesting that the GA fails toidentify pathways characterized by more complex interaction atthe highest intensities where multiple resonant multiphotonprocesses could be accessed. From the higher excited states someportion of the population couples back to I460. Potential couplingmechanisms include nonradiative internal conversion and stim-ulated Raman scattering between H and Sn. A common conse-quence of these 2 pathways is that the population will bevibrationally hot when reaching I460, accounting for the morerapid decay of this state at high intensity. It is also apparent thatmuch of the Sn population bypasses the I460 region altogether, assuggested particularly by the intensity scans (Fig. 3) and by thenormalized J625 transient scans (Fig. 5 Middle). This populationcould convert to the 13-cis conformation through the sameconical intersection region accessed after linear excitation, orthrough another conical intersection. The isomerization due tothis process must be exceedingly efficient, because it competesfavorably with the already efficient process that proceedsthrough the initial 1-photon excitation channel. The dominantmultiphoton pathway results in the formation of the 13-cisphotoproduct with unit or near unit efficiency.

Quantum coherence does not appear to play a significant rolein the experiments described in this work. Weak vibrationalcoherences with a period of a few hundred wave numbers areobserved in some of the transient scans for chirped pulseexcitation, but not for the optimal, TL pulse. The related analysiswill be presented elsewhere. Refs. 15, 21, and 45 report highfrequency oscillations that persist for �1 ps. The resolutionneeded to observe them is beyond the capabilities of ourexperimental setup. The model presented above does not requirecoherent wave packet dynamics.

The experiments presented in this work report control of theisomerization efficiency of retinal in bR in an intensity regime,which is not found in nature. The optimal solution is simple: thepulse of highest intensity/shortest duration maximizes 13-cisisomerization yield (37). However, the postexcitation moleculardynamics are complex and clearly suggest that the population ofeach conformer along the multipath trajectory is influenced by thepulse energy and duration. At lower intensities, the yield does notdepend on phase in our measurements. The findings presented inthis work show that bacteriorhodopsin carries out its functionefficiently over a range of light level conditions far beyond thosefound in nature. Solar radiation provides broad spectral excitationvery different from the ideal characteristics of the laser radiationused in these control experiments. Yet the isomerization channelidentified here may have biological relevance. Biological systemsare vulnerable to photodamage after UV excitation. The isomer-ization channel observed in these experiments rapidly and effi-ciently channels high energy excitation into a biologically useful

Florean et al. PNAS � July 7, 2009 � vol. 106 � no. 27 � 10899

BIO

CHEM

ISTR

YA

PPLI

EDPH

YSIC

AL

SCIE

NCE

S

Dow

nloa

ded

by g

uest

on

Dec

embe

r 18

, 202

0

channel. It would be interesting to compare isomerization of retinalin different environments after multiphoton or direct UV excita-tion to determine if the protein environment tunes or controls thisisomerization channel.

Materials and MethodsTunable noncollinear optical parametric amplifiers and programmableacousto-optic pulse shapers were used in this work, together with standardwhite light ultrafast probe techniques (46, 47). The details are discussed in SI

Text. The bR preparation was also similar to previous photoabsorption exper-iments on this molecule (25, 30, 45). Care was taken to limit the total bRexposure to laser radiation to avoid the effects of permanent photo-induceddamage, although the intensity-dependent nature of this study necessitatedhigher photon doses than some previous studies. The integrated photon dosewas kept to �5 or fewer absorbed photons per molecule.

ACKNOWLEDGMENTS. This work was supported by National Science Foun-dation Grants CHE 0718219 and PHY 0649578 and the FOCUS Center at theUniversity of Michigan.

1. Rohr M, Gartner W, Schweitzer G, Holzwarth AR, Braslavsky SE (1992) Quantum yieldsof the photochromic equilibrium between bacteriorhodopsin and its bathointerme-diate-k—femtosecond and nanosecond optoacoustic spectroscopy. J Phys Chem96:6055–6061.

2. Dioumaev AK, Savransky VV, Tkachenko NV, Chukharev VI (1989) Quantum yield andextinction measurements in strongly overlapping reactant and photoproduct absorp-tion-bands. 2. Bathointermediate formation in bacteriorhodopsin photocycle at room-temperature. J Photochem Photobiol B 3:397–410.

3. Balashov SP, Imasheva ES, Govindjee R, Ebrey TG (1991) Quantum yield ratio of theforward and back light reactions of bacteriorhodopsin at low-temperature and pho-tosteady-state concentration of the bathoproduct-k. Photochem Photobiol 54:955–961.

4. Abramczyk H (2004) Femtosecond primary events in bacteriorhodopsin and its retinalmodified analogs: Revision of commonly accepted interpretation of electronic spectraof transient intermediates in the bacteriorhodopsin photocycle. J Chem Phys120:11120–11132.

5. Fleming GR, vanGrondelle R (1997) Femtosecond spectroscopy of photosyntheticlight-harvesting systems. Curr Opin Struct Biol 7:738–748.

6. Gai F, Hasson KC, McDonald JC, Anfinrud PA (1998) Chemical dynamics in proteins: Thephotoisomerization of retinal in bacteriorhodopsin. Sci 279:1886–1891.

7. Haupts U, Tittor J, Oesterhelt D (1999) Closing in on bacteriorhodopsin: Progress inunderstanding the molecule. Annu Rev Biophys Biomol Struct 28:367–399.

8. Kuhlbrandt W (2000) Bacteriorhodopsin—the movie. Nature 406:569–570.9. Lanyi JK (2006) Proton transfers in the bacteriorhodopsin photocycle. Biochim Biophys

Acta Bioenerg 1757:1012–1018.10. Lanyi JK (2007) Studies of the bacteriorhodopsin photocycle without the use of light:

Clues to proton transfer coupled reactions. J Mol Microbiol Biotechnol 12(3–4):210–217.

11. Schenkl S, van Mourik F, van der Zwan G, Haacke S, Chergui M (2005) Probing theultrafast charge translocation of photoexcited retinal in bacteriorhodopsin. Sci309:917–920.

12. Vos MH, Martin JL (1999) Femtosecond processes in proteins. Biochim Biophys Acta-Bioenerg 1411:1–20.

13. Abe M, Ohtsuki Y, Fujimura Y, Domcke W (2005) Optimal control of ultrafast cis-transphotoisomerization of retinal in rhodopsin via a conical intersection. J Chem Phys123:10.

14. Atkinson GH, Ujj L, Zhou YD (2000) Vibrational spectrum of the j-625 intermediate inthe room temperature bacteriorhodopsin photocycle. J Phys Chem A 104:4130–4139.

15. Bardeen CJ, Wang Q, Shank CV (1998) Femtosecond chirped pulse excitation ofvibrational wave packets in ld690 and bacteriorhodopsin. J Phys Chem A 102:2759–2766.

16. Dexheimer SL, et al. (1992) Femtosecond impulsive excitation of nonstationary vibra-tional-states in bacteriorhodopsin. Chem Phys Lett 188(1–2):61–66.

17. Haran G, et al. (1996) Excited state dynamics of bacteriorhodopsin revealed by tran-sient stimulated emission spectra. Chem Phys Lett 261(4–5):389–395.

18. Herbst J, Heyne K, Diller R (2002) Femtosecond infrared spectroscopy of bacteriorho-dopsin chromophore isomerization. Sci 297:822–825.

19. Hou BX, Friedman N, Ottolenghi M, Sheves M, Ruhman S (2003) Comparing photoin-duced vibrational coherences in bacteriorhodopsin and in native and locked retinalprotonated schiff bases. Chem Phys Lett 381(5–6):549–555.

20. Kennis JTM, et al. (2002) Ultrafast protein dynamics of bacteriorhodopsin probed byphoton echo and transient absorption spectroscopy. J Phys Chem B 106:6067–6080.

21. Kobayashi T, Saito T, Ohtani H (2001) Real-time spectroscopy of transition states inbacteriorhodopsin during retinal isomerization. Nature 414:531–534.

22. Logunov SL, Volkov VV, Braun M, El-Sayed MA (2001) The relaxation dynamics of theexcited electronic states of retinal in bacteriorhodopsin by two-pump-probe femto-second studies. Proc Natl Acad Sci USA 98:8475–8479.

23. Mathies RA, Cruz CHB, Pollard WT, Shank CV (1988) Direct observation of the femto-second excited-state cis-trans isomerization in bacteriorhodopsin. Sci 240:777–779.

24. Myers AB, Harris RA, Mathies RA (1983) Resonance raman excitation profiles ofbacteriorhodopsin. J Chem Phys 79:603–613.

25. Prokhorenko VI, et al. (2006) Coherent control of retinal isomerization in bacteriorho-dopsin. Sci 313:1257–1261.

26. Ruhman S, Hou BX, Friedman N, Ottolenghi M, Sheves M (2002) Following evolution ofbacteriorhodopsin in its reactive excited state via stimulated emission pumping. J AmChem Soc 124:8854–8858.

27. Schenkl S, et al. (2006) Insights into excited-state and isomerization dynamics ofbacteriorhodopsin from ultrafast transient uv absorption. Proc Natl Acad Sci USA103:4101–4106.

28. Terentis AC, Ujj L, Abramczyk H, Atkinson GH (2005) Primary events in the bacterio-rhodopsin photocycle: Torsional vibrational dephasing in the first excited electronicstate. Chem Phys 313(1–3):51–62.

29. Terpugov EL, Degtyareva OV (2001) Infrared emission from photoexcited bacteriorho-dopsin: Studies by fourier transform infrared spectroscopy. J Mol Struct 565:287–292.

30. Vogt G, Nuernberger P, Brixner T, Gerber G (2006) Femtosecond pump-shaped-dumpquantum control of retinal isomerization in bacteriorhodopsin. Chem Phys Lett 433(1–3):211–215.

31. Wang JP, El-Sayed MA (2002) Time-resolved long-lived infrared emission from bacte-riorhodopsin during its photocycle. Biophys J 83:1589–1594.

32. Xu J, et al. (2003) Direct measurement of the photoelectric response time of bacterio-rhodopsin via electro-optic sampling. Biophys J 85:1128–1134.

33. Ye T, et al. (1999) Resolving the primary dynamics of bacteriorhodopsin, and of a‘‘c-13�c-14 locked’’ analog, in the reactive excited state. Chem Phys Lett 314(5–6):429–434.

34. Lee HM, Kim J, Kim CJ, Kim KS (2002) Ab initio study of the isomerization of retinalchromophore and its derivatives. J Chem Phys 116:6549–6559.

35. Flores SC, Batista VS (2004) Model study of coherent-control of the femtosecondprimary event of vision. J Phys Chem B 108:6745–6749.

36. Herek JL, Wohlleben W, Cogdell RJ, Zeidler D, Motzkus M (2002) Quantum control ofenergy flow in light harvesting. Nature 417:533–535.

37. Hoki K, Brumer P (2005) Mechanisms in adaptive feedback control: Photoisomerizationin a liquid. Phys Rev Lett 95:4.

38. Nuernberger P, Vogt G, Brixner T, Gerber G (2007) Femtosecond quantum control ofmolecular dynamics in the condensed phase. PCCP 9:2470–2497.

39. Ohtsuki Y, Ohara K, Abe M, Nakagami K, Fujimura Y (2003) New quantum controlpathway for a coupled-potential system. Chem Phys Lett 369(5–6):525–533.

40. Akiyama R, Kakitani T, Imamoto Y, Shichida Y, Hatano Y (1995) Temperature-dependence of femtosecond excited-state dynamics of bacteriorhodopsin analyzed bythe fourier-transform of optical-absorption spectra. J Phys Chem 99:7147–7153.

41. Pearson BJ, White JL, Weinacht TC, Bucksbaum PH (2001) Coherent control usingadaptive learning algorithms. PhRvA 6306:12.

42. Judson RS, Rabitz H (1992) Teaching lasers to control molecules. Phys Rev Lett 68:1500–1503.

43. Schmidt B, et al. (2005) Excited-state dynamics of bacteriorhodopsin probed by broad-band femtosecond fluorescence spectroscopy. Biochim Biophys Acta-Bioenerg1706:165–173.

44. Carroll EC, Florean AC, Bucksbaum PH, Spears KG, Sension RJ (2008) Phase control ofthe competition between electronic transitions in a solvated laser dye. Chem Phys350:75–86.

45. Kahan A, Nahmias O, Friedman N, Sheves M, Ruhman S (2007) Following photoinduceddynamics in bacteriorhodopsin with 7-fs impulsive vibrational spectroscopy. J AmChem Soc 129:537–546.

46. Florean AC, Carroll EC, Spears KG, Sension RJ, Bucksbaum PH (2006) Optical control ofexcited-state vibrational coherences of a molecule in solution: The influence of theexcitation pulse spectrum and phase in ld690. J Phys Chem B 110:20023–20031.

47. Carroll EC, Pearson BJ, Florean AC, Bucksbaum PH, Sension RJ (2006) Spectral phaseeffects on nonlinear resonant photochemistry of 1,3-cyclohexadiene in solution.J Chem Phys 124:114506.

10900 � www.pnas.org�cgi�doi�10.1073�pnas.0904589106 Florean et al.

Dow

nloa

ded

by g

uest

on

Dec

embe

r 18

, 202

0