Letter pubs.acs.org/JPCL
Ultrafast Time-Resolved Spectroscopy of Diarylethene-Based Photoswitchable Deoxyuridine Nucleosides Tiago Buckup,† Christopher Sarter,‡ Hans-Robert Volpp,† Andres Jas̈ chke,*,‡ and Marcus Motzkus*,† †
Physikalisch Chemisches Institut and ‡Institut für Pharmazie und Molekulare Biotechnologie, Ruprecht-Karls-Universität Heidelberg, D-69120 Heidelberg, Germany S Supporting Information *
ABSTRACT: Photoswitches based on the diarylethene architecture have been attracting considerable attention for the investigation and control of a variety of biological processes. The reversible photoisomerization reaction between their open- and closed-ring forms can be selectively addressed by irradiation with light of two markedly different wavelengths. In this work, the dynamics of the photochromic ring-opening reaction of four novel diarylethene-based photoswitchable deoxyuridine nucleosides is investigated by femtosecond transient absorption. Upon photoexcitation with sub-20 fs pulses in the first absorption band (500 nm), all four photoswitches showed a fast ballistic excited-state deactivation within less than 500 fs toward the S1/S0 conical intersection. Transient data was globally analyzed, and a relaxation kinetic model with a branching between open and closed ring forms without any loss channels was derived. Ring-opening quantum yields, Φr‑o, were found to strongly depend on the substituents (R), ranging from 0.64 (dUPSI: R = 2-naphthyl) to 0.30 (dUPSIV: R = 2pyridyl).
L
adenosine and guanosine, respectively, could be successfully combined with the photochromic properties of diarylethenes.17,19 These photoswitchable nucleosides consist of a 7-deazaadenosine (7-deazaguanosine, respectively) unit as the second aryl functionality which is linked to a thiophene via a 1,2-cyclopentenyl bridge to provide the switchable hexatriene/ cyclohexatriene core for the pericyclic isomerization reactions. As a consequence of this design concept, the nucleobase is for the first time an active part of the photoswitch, as a covalent bond between the nucleobase and the switchable core is formed/broken during the photochemical electrocyclic ringclosure/ring-opening reactions. This general design concept has been further expanded to photoswitchable pyrimidine nucleosides.20 These photoswitchable pyrimidine nucleosides challenge several of the currently well-established general design rules for diarylethene-based photoswitches:5 (i) They consist of two different aryl groups, namely, a five-membered methylthiophene ring and a six-membered ring system of the pyrimidine base. (ii) As depicted in the upper part of Figure 1, they contain only one methyl group (rather than two) at the carbon atoms that form the new covalent bond in the ringclosure reaction. As a consequence, these photoswitches may have interesting photochemical properties on their own beyond the realm of photocontrollable nucleic acids.20 Particularly important in the development of new photoswitches in regard to applications is the efficiency of the ringopening and -closing photochemical reactions. Conventional
ight-responsive molecular systems have been subject to quite extensive research efforts in different fields ranging from functional material science1 to biology.2−4 In particular, the use of irreversible phototriggers as well as reversible photochromic molecular switches has emerged as a powerful strategy for the spatially and temporally resolved investigation and control of a variety of biological processes. Among the variety of artificial photochromic molecules, photoswitches based on the diarylethene architecture have been attracting considerable attention.5,6 Photochromic diarylethene derivatives can undergo a reversible photoisomerization reaction between their open- and closed-ring forms, where the direction of the photoisomerization can be selectively addressed by irradiation with light of two markedly different wavelengths. In general, ultraviolet photoexcitation of the colorless open-ring isomer induces the ring-closure (cyclization) reaction, generating the closed-ring isomer, while excitation of the closed-ring isomer with visible light (typically in the 400−750 nm region)7 induces the ring-opening (cycloreversion) reaction back to the colorless open-ring isomer. Due to reversible switchability between these two thermally stable isomers, and their high fatigue resistance with respect to ring-closure and ring-opening photoreactions,8 diarylethene-based photoswitches9,10 have evolved into promising materials for applications in molecular electronics and ultrahigh-density optical data storage.11 Despite the above-described unique advantages provided by the diarylethene architecture, its application in molecular switches for biomolecules such as proteins, peptides,12−16 and nucleic acids17−20 was only recently explored. A novel class of photoswitchable nucleosides was developed, in which the structural features and molecular recognition properties of © XXXX American Chemical Society
Received: September 4, 2015 Accepted: November 10, 2015
4717
DOI: 10.1021/acs.jpclett.5b01949 J. Phys. Chem. Lett. 2015, 6, 4717−4721
Letter
The Journal of Physical Chemistry Letters
Figure 1. Photochromism of novel diarylethene-based photoswitchable deoxyuridine nucleosides (dUPSI−IV) with different substituents (R: I−IV) at the 5 position of the methylthiophene moiety. The absorption spectrum of dUPSI in methanol is shown in the inset: the black solid line corresponds to the open-ring isomer, and the red solid line corresponds to the photostationary state under irradiation with 366 nm light.
diarylethene-based photoswitches are notoriously known by their photochemical reaction quantum yield Φr‑o for ringopening of only a few percent.9,21 In this letter, we investigate the ultrafast dynamics and the quantum yield of the ringopening reaction of four novel diarylethene-based photoswitchable deoxyuridine nucleosides (denoted in the following as dUPSI−IV). They differ in their respective substituent (R: I− IV) at the 5 position of the methylthiophene moiety (Figure 1). Transient absorption (TA) data following 500 nm excitation of the closed-ring isomers of dUPSI−IV in methanol are dominated by two major features (Figure 2). A strong and
Figure 3. Comparison kinetics for (a) dUPSI, (b) dUPSII, (c) dUPSIII, and (d) dUPSIV. Traces with negative amplitudes were integrated between 485 and 495 nm, while traces with positive amplitude were integrated between 590 and 600 nm. Red lines are fittings of the single transients following the model described in section 3 of the Supporting Information. See Table S4 for the obtained values.
oscillations was 240 fs (corresponding to a vibrational frequency of 140 cm−1) with dephasing time constants of about 500 ± 100 fs. dUPSI shows very weak oscillations (almost 10 times smaller in amplitude than those observed for dUPSII−IV), which could therefore not be fitted with the above function. The very fast dephasing of this vibrational mode and its detection almost exclusively in spectral regions with positive TA signal suggest that this mode originates from the electronic excited states of the closed-isomer forms. The nonoscillatory contribution to the signal was fitted with a double exponential decay at all detection wavelengths (see Table S4). In the bleach region, an additional signal offset was included to fit the nonrecovery of the signal. In general, two groups of time constants were obtained: A very fast decay with a time constant faster than 150 fs together with a second slower decay with time constants between 1.3 and 2.7 ps. Note that the fitting with two exponential decays is a pure numerical fitting and does not imply any molecular kinetic model because contributions (absorptions, stimulated emissions, bleach) of different states can spectrally overlap. The constant negative residual signal, which remains unchanged up to a pump−probe delay time of 500 ps at wavelengths below 550 nm clearly demonstrates that a considerable fraction of the initially excited closed-ring isomers are converted to the transparent open-ring isomers in the photoinduced cycloreversion reaction, indicating ring-opening quantum yields, Φr‑o, well above 0.5 in the case of dUPSI and dUPSII. A direct calculation of Φr‑o values via the ratio of the residual bleach amplitude and the bleach amplitude at early probe delay times is, however, challenging because the bleach spectral range seems to overlap to some extent with the excited-
Figure 2. Selected transient absorption spectra obtained after 500 nm excitation of the open-ring isomers of (a) dUPSI, (b) dUPSII, (c) dUPSIII, and (d) dUPSIV. Spectra selected for different pump−probe delay times: 0.2 ps (black), 0.3 ps (red), 0.4 ps (green), 1 ps (dark blue), 3 ps (light blue), and 10 ps (pink).
broad excited-state absorption appears very early in the dynamics with a maximum between 550 and 600 nm. dUPSI has a maximum for this band at about 600 nm, while the photoswitches dUPSII−IV have a maximum at shorter wavelengths. The amplitude of this band decays to zero within 10 ps in all compounds. The second feature is a negative signal at about 500 nm, which does not completely decay to zero even after pump−probe delay times larger than 500 ps (not shown). Selected transients at the maximum of the absorption band at about 590 nm (Figure 3) clearly reveal a third feature, namely, an oscillatory contribution extending from 525 up to 600 nm. The oscillations observed for dUPSII−IV can be well fitted with a single sinusoidal function damped with a single exponential decay. For the latter three photoswitches, the period of the 4718
DOI: 10.1021/acs.jpclett.5b01949 J. Phys. Chem. Lett. 2015, 6, 4717−4721
Letter
The Journal of Physical Chemistry Letters
Table 1. Time Constants, τij, and Ring-Opening Quantum Yields, Φr‑o, Obtained from the GTA of the dUPSI−IV Data Setsa τ12 dUPSI dUPSII dUPSIII dUPSIV
0.08 0.1 0.11 0.11
± ± ± ±
τ23 0.01 0.01 0.02 0.02
0.45 0.35 0.42 0.34
± ± ± ±
τ34 0.1 0.06 0.08 0.08
3.9 4.1 4.5 2.6
± ± ± ±
τ35 0.2 0.3 0.3 0.3
2.2 3.1 6.8 6.0
± ± ± ±
Φr‑o 0.3 0.4 0.5 0.5
0.64 0.57 0.40 0.30
± ± ± ±
0.04 0.04 0.03 0.05
Φr‑o values were calculated from the respective product formation rate constant values of the ground-state open-ring isomer (ko = 1/τ35) and closedring isomer (kc = 1/τ34) via Φr‑o = ko/(ko+ kc). Uncertainties of the Φr‑o values were determined by simple error propagation. All values are given in picoseconds. a
state absorption particularly at early pump−probe delay times (Figure 2). In order to derive more precise Φr‑o values and to gain insight into the dynamics of the ring-opening reactions, a global target analysis (GTA) was carried out for the experimental data sets of all four photoswitches dUPSI−IV. A model comprising five levels (L1−L5) was employed to fit all four photoswitches. A similar model has been used to describe the kinetics of other diarylethene derivatives.22,23 Other kinetic models involving fewer states and/or loss channels were also tested but could not fit satisfactorily the experimental data obtained in this work. The used model is based on two sequential decays L1 → L2 → L3, followed by a branching from L3 to L4 and L5 (Figure S4). The GTA results depict an ultrafast decay from L1 to the L3 within 500 fs for all photoswitches (Table 1). The initial decay from L1 is faster (∼100 fs) than the decay from L2 into L3 (300−500 fs). Note that the L1 → L2 decay constant involves an uncertainty (10−20%) larger than that of the other decay constants due to the temporal overlap with the coherence artifact of transient absorption. This is followed by the branching L3 → L4 and L3 → L5 with time constants varying from 2.2 to 6.8 ps. The ring-opening quantum yields Φr‑o given in Table 1 can be calculated by the ratio between ko = 1/τ35 and ko+ kc (with kc = 1/τ34). It exhibits a clear dependence on the substituent R: dUPSI shows a quantum yield of a 0.64, followed by dUPSII with 0.57. The other two photoswitches have a smaller quantum yield of about 0.3−0.4. Moreover, a comparison of the obtained species associated spectra (SAS) can further help to identify the origin of each level. In general, transient absorption spectra originating from the same potential surface often have similar spectral shapes because the respective electronic transitions obey the same selection rules. This is particularly true for transitions originating spatially close to the other on the potential surface. Therefore, the comparable SAS obtained for L2 and L3 hint that these absorption spectra are originated in the same electronic potential. On the other hand, the SAS (section 5 in the Supporting Information) obtained for L1, which in this model is directly related to the Franck− Condon region, shows a different spectrum for all photoswitches. Such a difference suggests that this species does not share the same potential surface as L2/L3. In this regard, the levels L1−L5 can be assigned to distinct locations on the potential energy surfaces (S2, S1, S0) that are generally considered in the photochromic cycloreversion process of 6π electrocyclic systems.24−26 In our model (Scheme 1), the time constant τ12 describes the decay of the signal contribution from the Franck−Condon region (L1) on S2 as the initially prepared excited-state wave packet enters the S1-PES via the S2/S1 conical intersection (CI). L2 is assigned to the excited-state closed-ring isomer of the photoswitches (S1-c in Scheme 1). In this location the short-lived coherent lowfrequency oscillations (140 cm−1) are observed in the case of
Scheme 1. Schematic Illustration of the Ultrafast ExcitedState Dynamics of the Cycloreversion Reaction after Optical Excitation at 500 nm (Green Vertical Arrow)a
a
Red arrows denote proposed reaction pathways with their associated time constants (τ12, τ23, and τ34, τ35) for which the values derived in the GTA are given in Table 1. TS1: transition state connecting the closed-ring isomer in the S1 state (S1-c) with the open-ring isomer (S1o) in the S1 state. S1/S0-CI(o): open-isomer conical intersection connecting S1-o with the S0 ground state. TS0: transition state of the ground-state isomerization reaction.
the dUPSII−IV photoswitches. The fast decay of L2 with time constant τ23 (300−500 fs) leads to the population of L3 (assigned to the excited-state open-ring isomer of the photoswitches, denoted as S1-o in Scheme 1), which subsequently decays, through the S1/S0-CI(o), with a time constant of τ-S1/S0 ≈ 1−3 ps into the closed-ring (with τ34) and open-ring (with τ35) ground-state isomer products, which are represented in the GTA model by the levels L4 and L5, respectively. A similar model has been put forward by Ward and Elles to rationalize the results of their pump−probe TA studies of the cycloreversion reaction of the 1,2-bis(2,4-dimethyl-5-phenyl-3thienyl)perfluorocyclopentene (DMPT-PFCP) diarylethene derivative.22,23 However, for the latter molecular switch, in striking contrast to the dUPSI−IV photoswitches studied in the present work, a rather low ring-opening quantum yield of Φr‑o = 0.019 ± 0.002 along with a considerably slower excited-state dynamics (with a τ23 value of ∼3 ps) was obtained after isoenergetic one-photon excitation at 500 nm. In the later DMPT-PFCP study the slow decay of the L2 level, which, as in the present case, was assigned to the excited-state closed-ring isomer (S1-c in Scheme 1), was explained by an activated barrier-crossing process (over TS1 in Scheme 1) from closedring to the open-ring isomer (S1-o in Scheme 1). DMPT-PFCP cycloreversion studies by Irie and co-workers,27 which revealed 4719
DOI: 10.1021/acs.jpclett.5b01949 J. Phys. Chem. Lett. 2015, 6, 4717−4721
Letter
The Journal of Physical Chemistry Letters
dehydrocholesterol (DHC), for which a value of Φr‑o = 0.34 was reported.30 However, as indicated by the Φr‑o values of 0.57 and 0.64 obtained for the phenyl- (dUPSII) and naphthyl-substituted (dUPSI) photoswitches, respectively, the cycloreversion efficiency can be substantially increased beyond this “ultra-fast ballistic limit”. In the case of dUPSI, this efficiency can be increased even well beyond the “statistical limit” for diarylethene-based molecular switches, where the ring-opening is about equal to the ring-closure quantum yield (Φr‑o ≈ Φc‑o).25 A statistical limit of Φr‑o ≈ 0.5 could be expected for diarylethene derivatives with no TS1 barrier at all, if the excess energy is efficiently redistributed throughout the modes of the molecule via excited-state intramolecular vibrational energy redistribution (IVR) in a way which would allow these vibrationally “hot” molecule to access the S1/S0-CI(o) and S1/S0-CI(c) conical intersections with about equal probability. In summary, the whole set of information provided by the present experiment led us to conclude that for all diarylethenebased photoswitchable deoxyuridine nucleosides studied in the present work the ballistic mechanism described above is actually operating. The variation of the product branching and hence Φr‑o is determined by subtle details of the local topography31 of the S1/S0-CI(o) conical intersection through which the cycloreversion reaction proceeds.
that the ring-opening quantum yield can be substantially enhanced from Φr‑o = 0.015 at 620 nm to Φr‑o = 0.032 at 480 nm by increasing the photoexcitation energy (and hence the excess energy of the S1-o isomer), independently confirmed that such an activated excited-state barrier-crossing mechanism is indeed operating. Furthermore, evidence for a correlation between the energy difference ΔE = E(S1-c) − E(S1-o) of the excited-state closed-ring (S1-c) and open-ring (S1-o) isomers of diarylethene-based photoswitches and their respective ringopening quantum yields, Φr‑o, was gained in these experiments.27 The considerable differences in the absolute values of Φr‑o along with the different wavelength dependence of Φr‑o observed in the comparative DMPT-PFCP and 1,2-bis(2methyl-1-benzothiophen-3-yl)perfluorocyclopentene irradiation experiments (see Figure 2 of ref 27) clearly demonstrated that an increase of ΔE from negative to positive values is accompanied by a noticeable reduction of the excited-state cycloreversion barrier (TS1), which in turn leads to noticeably higher Φr‑o values. Taking into consideration the latter findings, it seems reasonable to attribute the fast excited-state dynamics observed in the present work for the dUPSI−IV photoswitches to the presence of a considerably lower excited-state energy barrier. Such a lower barrier allows the excited-state wavepacket to continue its ultrafast decay from the Franck−Condon region (S2−FC) through the TS1 transition-state region toward the open-isomer S1/S0-CI(o) conical intersection (Scheme 1) in a direct ballistic fashion. In this way, the system avoids the crossing seam region near the closed-isomer conical intersection S1/S0-CI(c),25 which would otherwise open an efficient pathway for relaxation back to the ground-state closed-ring isomer. This is exactly the case for DMPT-PFCP with its low ring-open quantum yield value of Φr‑o = 0.019 ± 0.002.23 The short-lived coherent low-frequency oscillations (∼140 cm−1) are consistent with the proposed ballistic excited-state wavepacket dynamics, which starts from the Franck−Condon region (L1) on the S2 PES and continues on the S1 PES until the S1/S0 CI(o) is reached. A similar low-frequency coherent oscillation was observed during the ultrafast ring-opening dynamics of 1,3-cyclohexadiene (CHD) and attributed to a CHD ring deformation vibration on the lowest excited-state 2A-state (S1) along the symmetry-breaking coordinate, which connects the pericyclic minimum with the final 2A/1A (S1/S0) CI.26,28 In the latter work, it was found that this vibration is not excited in the FC region but stimulated by an antisymmetric distortion required for the wavepacket to circumvent the first 1B/2A (S2/S1) CI in passing from the optically excited state 1B-state (S2) to the optically dark 2A-state (S1). Hence, the observation of this CHD-skeleton vibration at location L2 on the S1 surface might be taken as a further indication that a ballistic excited-state mechanism, similar to the one derived for the CHD ring-opening reaction, is indeed operating in the dUPSI−IV cycloreversion reactions. In the framework of such a ballistic mechanism, the dynamics of the wavepacket through the S1/S0-CI(o) conical intersection and its final bifurcation on the S0 ground state determines the product quantum yields. The Φr‑o magnitude of the two pyridyl-substituted photoswitches dUPSIV and dUPSIII derived in the GTA is comparable to the recently confirmed Φr‑o value of 0.3−0.4 for the ultrafast CHD ring-opening reaction.29 This might be regarded as a general “ultra-fast ballistic limit” for the ring-opening quantum yield of CHD derivatives such as, e.g., 7-
■
EXPERIMENTAL SECTION Photoswitchable pyrimidine nucleosides dUPSI−IV in their openring isomeric form were synthesized following the methodology described in ref 20, which is summarized in Figure S1 in the Supporting Information. The results of dUPSI−IV fatigue resistance measurements are also shown in Figure S2 in the Supporting Information. For the pump−probe TA measurements, dUPSI−IV closedring isomer solution samples were prepared (directly before each experiment) by irradiating transparent solutions containing the respective open-ring isomers (in concentrations of 60 μM in methanol) with a Hg (Ar)-UV-lamp (equipped with an UG1-filter (Schott, transmission in the UV range between 300 and 400 nm, with a maximum transmittance at ∼365 nm) until the photostationary state was reached. Samples used in the TA measurements had an absorbance between 0.25 and 0.3 OD in a 0.5 mm thickness. The TA measurements were performed using a homemade noncollinear optical parametric amplifier (nc-OPA) pumped by a Ti:sapphire-Laser system at 1 kHz. The output of the nc-OPA was tuned to around 500 nm for photoexcitation of the four photoswitches dUPSI−IV. Typical duration of the photoexcitation laser pulse was around 16 fs. The instrument response function (IRF) was about 60−80 fs, depending on the detection wavelength. White light (450−750 nm), generated by means of a 2 mm sapphire crystal, was used as probe light in the TA measurement of all photoswitches. Typical pump pulse energies were below 50 nJ with a laser beam spot size in the sample of 50 μm. In order to suppress polarization and photoselection effects, the polarization of the pump beam was set at the magic angle with respect to that of the probe beam. TA measurements were performed by chopping every second pump pulse. The probe pulse was dispersed in a spectrometer (LOT: Spectrograph MS 125) for detection with a homemade photodiode array with 256 pixels. A flow cell was used to refresh the sample in the excitation volume between successive pulses. Raw TA data was corrected regarding white light dispersion using the coherence artifact as 4720
DOI: 10.1021/acs.jpclett.5b01949 J. Phys. Chem. Lett. 2015, 6, 4717−4721
Letter
The Journal of Physical Chemistry Letters
diarylethene-based non-natural amino acids. Tetrahedron 2013, 69, 6170−6175. (15) Reisinger, B.; Kuzmanovic, N.; Loffler, P.; Merkl, R.; Koenig, B.; Sterner, R. Exploiting Protein Symmetry To Design Light-Controllable Enzyme Inhibitors. Angew. Chem., Int. Ed. 2014, 53, 595−598. (16) Babii, O.; Afonin, S.; Berditsch, M.; Reisser, S.; Mykhailiuk, P. K.; Kubyshkin, V. S.; Steinbrecher, T.; Ulrich, A. S.; Komarov, I. V. Controlling Biological Activity with Light: Diarylethene- Containing Cyclic Peptidomimetics. Angew. Chem., Int. Ed. 2014, 53, 3392−3395. (17) Singer, M.; Jaeschke, A. Reversibly Photoswitchable Nucleosides: Synthesis and Photochromic Properties of DiaryletheneFunctionalized 7-Deazaadenosine Derivatives. J. Am. Chem. Soc. 2010, 132, 8372−8377. (18) Barrois, S.; Beyer, C.; Wagenknecht, H. A. Covalent Modification of 2 ’-Deoxyuridine with Two Different Molecular Switches. Synlett 2012, 23, 711−716. (19) Singer, M.; Nierth, A.; Jaeschke, A. Photochromism of Diarylethene-Functionalized 7-Deazaguanosines. Eur. J. Org. Chem. 2013, 14, 2766−2769. (20) Cahova, H.; Jaeschke, A. Nucleoside-Based Diarylethene Photoswitches and Their Facile Incorporation into Photoswitchable DNA. Angew. Chem., Int. Ed. 2013, 52, 3186−3190. (21) Irie, M.; Sakemura, K.; Okinaka, M.; Uchida, K. Photochromism of dithienylethenes with electron-donating substituents. J. Org. Chem. 1995, 60, 8305−8309. (22) Ward, C. L.; Elles, C. G. Controlling the Excited-State Reaction Dynamics of a Photochromic Molecular Switch with Sequential TwoPhoton Excitation. J. Phys. Chem. Lett. 2012, 3, 2995−3000. (23) Ward, C. L.; Elles, C. G. Cycloreversion Dynamics of a Photochromic Molecular Switch via One-Photon and Sequential TwoPhoton Excitation. J. Phys. Chem. A 2014, 118, 10011−10019. (24) Boggio-Pasqua, M.; Ravaglia, M.; Bearpark, M. J.; Garavelli, M.; Robb, M. A. Can diarylethene photochromism be explained by a reaction path alone? A CASSCF study with model MMVB dynamics. J. Phys. Chem. A 2003, 107, 11139−11152. (25) Asano, Y.; Murakami, A.; Kobayashi, T.; Goldberg, A.; Guillaumont, D.; Yabushita, S.; Irie, M.; Nakamura, S. Theoretical study on the photochromic cycloreversion reactions of dithienylethenes; on the role of the conical intersections. J. Am. Chem. Soc. 2004, 126, 12112−12120. (26) Kosma, K.; Trushin, S. A.; Fuss, W.; Schmid, W. E. Cyclohexadiene ring opening observed with 13 fs resolution: coherent oscillations confirm the reaction path. Phys. Chem. Chem. Phys. 2009, 11, 172−181. (27) Sumi, T.; Takagi, Y.; Yagi, A.; Morimoto, M.; Irie, M. Photoirradiation wavelength dependence of cycloreversion quantum yields of diarylethenes. Chem. Commun. 2014, 50, 3928−3930. (28) Garavelli, M.; Page, C. S.; Celani, P.; Olivucci, M.; Schmid, W. E.; Trushin, S. A.; Fuss, W. Reaction path of a sub-200 fs photochemical electrocyclic reaction. J. Phys. Chem. A 2001, 105, 4458−4469. (29) Adachi, S.; Sato, M.; Suzuki, T. Direct Observation of GroundState Product Formation in a 1,3-Cyclohexadiene Ring-Opening Reaction. J. Phys. Chem. Lett. 2015, 6, 343−346. (30) Anderson, N. A.; Shiang, J. J.; Sension, R. J. Subpicosecond ring opening of 7-dehydrocholesterol studied by ultrafast spectroscopy. J. Phys. Chem. A 1999, 103, 10730−10736. (31) Atchity, G. J.; Xantheas, S. S.; Ruedenberg, K. Potential-Energy Surfaces near Intersections. J. Chem. Phys. 1991, 95, 1862−1876. (32) van Wilderen, L. J. G. W.; Lincoln, C. N.; van Thor, J. J. Modelling Multi-Pulse Population Dynamics from Ultrafast Spectroscopy. PLoS One 2011, 6, e17373.
reference for time delay zero. Global target analysis was performed using a toolbox developed by van Wilderen et al. in MatLab.32
■
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jpclett.5b01949. Synthesis and photochromic properties of the dUPSI−IV photoswitches, UV−vis absorption data, fitting of transient absorption traces, global target analysis results, and characterization of samples via high-resolution mass spectrometry and NMR (PDF)
■
AUTHOR INFORMATION
Corresponding Authors
*E-mail:
[email protected]. *E-mail:
[email protected]. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS The authors thank S. Mehlhose and D. Starukhin for their assistance in the early stage of the experiments.
■
REFERENCES
(1) Zhang, J. J.; Zou, Q.; Tian, H. Photochromic Materials: More Than Meets The Eye. Adv. Mater. 2013, 25, 378−399. (2) Goeldner, M.; Givens, R. S. Dynamic Studies in Biology: Phototriggers, Photoswitches and Caged Biomolecules; Wiley-VCH Verlag: Weinheim, 2005. (3) Szymanski, W.; Beierle, J. M.; Kistemaker, H. A. V.; Velema, W. A.; Feringa, B. L. Reversible Photocontrol of Biological Systems by the Incorporation of Molecular Photoswitches. Chem. Rev. 2013, 113, 6114−6178. (4) Brieke, C.; Rohrbach, F.; Gottschalk, A.; Mayer, G.; Heckel, A. Light-Controlled Tools. Angew. Chem., Int. Ed. 2012, 51, 8446−8476. (5) Irie, M. Diarylethenes for memories and switches. Chem. Rev. 2000, 100, 1685−1716. (6) Feringa, B. L.; Browne, W. R. Molecular Switches. Wiley-VCH Verlag: Weinheim, 2011. (7) Sun, F.; Zhang, F. H.; Guo, H. B.; Zhou, X. H.; Wang, R. J.; Zhao, F. Q. Efficient synthesis and photochromic properties of 2,3-position hybrid diarylethene derivatives. Tetrahedron 2003, 59, 7615−7621. (8) Herder, M.; Schmidt, B. M.; Grubert, L.; Patzel, M.; Schwarz, J.; Hecht, S. Improving the Fatigue Resistance of Diarylethene Switches. J. Am. Chem. Soc. 2015, 137, 2738−2747. (9) Irie, M.; Mohri, M. Thermally Irreversible Photochromic Systems - Reversible Photocyclization of Diarylethene Derivatives. J. Org. Chem. 1988, 53, 803−808. (10) Nakamura, S.; Irie, M. Thermally Irreversible Photochromic Systems - a Theoretical-Study. J. Org. Chem. 1988, 53, 6136−6138. (11) Irie, M.; Fukaminato, T.; Matsuda, K.; Kobatake, S. Photochromism of Diarylethene Molecules and Crystals: Memories, Switches, and Actuators. Chem. Rev. 2014, 114, 12174−12277. (12) Vomasta, D.; Hogner, C.; Branda, N. R.; Koenig, B. Regulation of human carbonic anhydrase I (hCAI) activity by using a photochromic inhibitor. Angew. Chem., Int. Ed. 2008, 47, 7644−7647. (13) Fujimoto, K.; Kajino, M.; Sakaguchi, I.; Inouye, M. Photoswitchable, DNA-Binding Helical Peptides Assembled with Two Independently Designed Sequences for Photoregulation and DNA Recognition. Chem. - Eur. J. 2012, 18, 9834−9840. (14) Fujimoto, K.; Maruyama, T.; Okada, Y.; Itou, T.; Inouye, M. Development of a new class of photochromic peptides by using 4721
DOI: 10.1021/acs.jpclett.5b01949 J. Phys. Chem. Lett. 2015, 6, 4717−4721