Ultrafast Excited-State Dynamics of all-trans ... - ACS Publications

Article Cite This: J. Phys. Chem. A XXXX, XXX, XXX-XXX

pubs.acs.org/JPCA

Ultrafast Excited-State Dynamics of all-trans-Capsanthin in Organic Solvents Mirko Scholz, Oliver Flender, Thomas Lenzer, and Kawon Oum* Physikalische Chemie, Universität Siegen, Adolf-Reichwein-Str. 2, 57076 Siegen, Germany S Supporting Information *

ABSTRACT: The C40 carotenoid capsanthin is of photophysical interest because it belongs to the family of terminally carbonyl-substituted apocarotenes. These have the potential to exhibit intramolecular charge transfer (ICT) character in the excited state. We studied its ultrafast dynamics in different solvents using broadband transient absorption spectroscopy in the 260−1600 nm range. Photoexcitation initially populated the S2(11Bu+) state which decayed to the S1(21Ag−) state in 120−150 fs. The lifetime of the S1 state decreased from 10.3 ps (isooctane) to 9.1 ps (methanol). Together with the absence of stimulated emission this suggested a weak ICT character of S1. A steric influence of the five-membered ring in capsanthin was identified based on a comparison with the sister apocarotenone citranaxanthin which features methyl substitution at the keto group. While both apocarotenones exhibit the same S1 lifetime in isooctane, the decrease in lifetime in polar solvents is weaker for capsanthin because presumably the five-membered ring sterically perturbs stabilization of ICT character by the solvent. For the S1 state of capsanthin we further observed intramolecular vibrational redistribution and collisional energy transfer to the solvent with time constants of 340−420 fs and 8.5−9.1 ps, respectively. No evidence for excited-state photoisomerization of capsanthin was found.

1. INTRODUCTION The C40 carotenoid all-trans-capsanthin (Figure 1) contributes to the intense red color emerging during the ripening process

character of peridinin and fucoxanthin which constitute the key carbonyl carotenoid pigments for harvesting light in the bluegreen spectral region inside peridinin−chlorophyll−protein (PCP) and fucoxanthin−chlorophyll−protein (FCP) complexes of marine algae.16−18 Capsanthin features 11 conjugated double bonds and a fivemembered ring (denoted as κ) at the keto group. This enables us to investigate the importance of ICT character in a longerchain apocarotenone and the possible impact of steric congestion at the carbonyl group. For comparison, we provide selected results on the photoinduced dynamics of the related apocarotenone citranaxanthin which features a methyl group in the 6′-position instead of the five-membered ring and lacks the OH group at the 3-position.10 To disentangle the excited-state dynamics on a femto- to picosecond time scale, we employ pump-supercontinuum probe (PSCP) spectroscopy with wide spectral coverage (260−1600 nm) and perform a detailed modeling using global kinetic analysis.

Figure 1. Chemical structure of the C40 carotenoid all-trans-capsanthin (all-E(3R,3′S,5′R)-3,3′-dihydroxy-β,κ-caroten-6′-one).

of pepper and chili fruits. This is induced by expression of the capsanthin−capsorubin synthase gene (Ccs) initiating capsanthin biosynthesis from the carotenoid antheraxanthin.1,2 In addition, this xanthophyll is of photophysical interest because of the terminal conjugated keto substituent attached to the polyene chain, rendering it a member of the β-apo-6′carotenone family. Such molecules with asymmetric terminal carbonyl substitution at their backbone often exhibit intramolecular charge transfer (ICT) character and a polarityinduced reduction in lifetime of the first electronically excited state, in contrast to conventional C40 carotenoids such as βcarotene derivatives.3−15 Understanding the photoinduced excited-state processes of such carbonyl apocarotenoids might also eventually provide a better understanding of the ICT © XXXX American Chemical Society

2. EXPERIMENTAL SECTION all-trans-Capsanthin (Toronto Research Chemicals, purity >97%), all-trans-citranaxanthin (BASF SE, purity >97%), isooctane (Merck, Uvasol, >99.8%), acetonitrile, and methanol Received: August 18, 2017 Revised: October 12, 2017

A

DOI: 10.1021/acs.jpca.7b08252 J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A (both from Fisher, HPLC gradient grade, >99.99%) were used as received. Steady-state absorption spectra were recorded at T = 296 K over the spectral range 200−800 nm using Varian Cary 5000 and Varian Cary 5E spectrophotometers. The sample solutions were bubbled with nitrogen prior to the experiments. Ultrafast broadband transient absorption spectra for capsanthin in the UV−vis and NIR regions were recorded using two setups for PSCP spectroscopy. The experimental implementation19,20 and the principle of the method21 were described in detail previously. Briefly, capsanthin was excited either at 400 nm (second harmonic of the amplifier system) or 500 nm (NOPA). Supercontinua were generated using either frequency-doubled (400 nm) or fundamental pulses (800 nm) of the amplifier system resulting in a spectral coverage of 260− 700 and 850−1600 nm, respectively. Sample and reference spectra were recorded using two grating spectrographs employing 512-element Si or InGaAs photodiode arrays allowing for a shot-to-shot correction of the white-light intensity fluctuations in the transient spectra. In addition, selected kinetics for citranaxanthin in isooctane and acetonitrile were recorded using the single wavelength pump−probe setup (λpump = 390 nm, λprobe = 475 nm) described earlier.11 In each case, 10−15 mL of a nitrogensaturated capsanthin or citranaxanthin solution was circulated through a flow cell with quartz windows (path length 1 mm). The concentration of capsanthin was about 10 μM, and typical laser pump pulse fluences were about 100−150 μJ cm−2, as determined by a combination of a CCD camera and a calibrated photodiode detector. The relative polarization of the pump and supercontinuum (or single-wavelength) probe pulses was set at the magic angle (54.7°). The time-zero correction was performed using the coherent response of the respective pure organic solvent. A cross-correlation and a time-zero accuracy of ca. 80 and 10 fs, respectively, were obtained using this procedure. PSCP data sets were analyzed by global kinetic modeling as described previously14,15,20,22,23 using the kinetic scheme discussed in section 3.3. Gaussian functions were employed to parametrize the species-associated spectra (SAS) of the different electronic species. During the optimization procedure, which also considered the experimental time resolution, the parameters of the Gaussian functions as well as the lifetimes of the different electronic species were varied.

Figure 2. Steady-state absorption spectra of all-trans-capsanthin in isooctane (black solid line), acetonitrile (red solid line), and methanol (green dashed line). Inset: magnified view near the absorption maximum.

different hydrogen-bond motifs with capsanthin. The effect is absent for nonpolar C40 carotenoids such as β-carotene derivatives.14,15 3.2. Excited-State Dynamics of Capsanthin. UV−vis− NIR broadband transient absorption spectra of capsanthin in acetonitrile for pumping at 500 nm are presented in Figure 3.

3. RESULTS AND DISCUSSION 3.1. Steady-State Absorption Spectra. Figure 2 shows steady-state absorption spectra of capsanthin. The S0 → S2 band in isooctane is structured and peaks at 470 nm. The vibronic progression with a spacing of about 1250 cm−1 is typically assigned to a combination of two symmetric vibrational stretching modes C−C (1150 cm−1) and CC (1600 cm−1).24,25 In acetonitrile this vibrational structure is considerably washed out, and the band is asymmetrically broadened toward larger wavelengths (see the inset in Figure 2). A similar effect has been observed for other asymmetrically carbonyl-substituted carotenoids.6−9,11,26 It is attributed to the dipolar character of S0 leading to a broad distribution of solvation structures with different degree of stabilization of the negative partial charge on the carbonyl oxygen and consequently inhomogeneous broadening. In methanol, the spectral structure has even completely disappeared, and the red wing extends to even larger wavelengths. This might indicate enhanced inhomogeneous broadening due to formation of

Figure 3. UV−vis−NIR transient absorption spectra of all-transcapsanthin in acetonitrile at 296 K. Different panels highlight the distinct excited-state dynamics for time windows of particular interest. The inverted steady-state absorption spectrum is shown in the lowest panel as a dashed orange line.

The corresponding contour plot and selected kinetic traces are shown in the Supporting Information (Figure S1). Around zero time delay (top panel) the negative signal centered at 480 nm is assigned to ground-state bleach (GSB). It appears together with NIR excited-state absorption (ESA) above 900 nm, which arises from the initially populated S2 state. Already at very short times we observe the formation of a more prominent ESA band (peak at 575 nm) at the expense of the NIR S2 → Sn ESA band. It is assigned to ESA of the S1 state which is formed from S2 by B


Article

The Journal of Physical Chemistry A internal conversion (IC) with a time constant of τ2 = 116 fs. Another much weaker S1 ESA band is observed in the NIR region around 1500 nm and most easily seen in the contour representation of Figure S1.27 By ca. 0.3 ps the S2 ESA has disappeared (second panel). The S1 → Sn ESA band at 575 nm rises further and at the same time narrows due to intramolecular vibrational redistribution (IVR) and collisional energy transfer (CET) to solvent molecules with time constants of τIVR,1 = 340 fs and τCET,1 = 8.6 ps, respectively (for the extraction of the individual time constants see the global analysis in section 3.3). In the bottom panel, the IC process S1 → S0 is observed, resulting in the decay of the S1 → Sn ESA band and the GSB. The associated time constant is τ1 = 9.9 ps. There is no residual bleach which suggests that there is no measurable excited-state photoisomerization to cis-isomers: Any formation of cis-isomers should manifest itself as a persistent GSB because they possess substantially lower peak absorption coefficients than the all-trans species.28 This is however not observed. Because of the large number of vibrational modes of C40 carotenoids, the redistribution of excess energy into the critical torsional isomerization coordinate cannot compete with fast intramolecular electronic decay to the ground electronic state, in contrast to the situation in shorter carotenoids, such as all-trans-retinal where photoisomerization is detected.3 We also note that the UV part of the transient spectra is spectrally “silent”: Except for weak GSB features, which correlate well with the steady-state absorption spectrum (Figure 2), no substantial ESA features are observed down to 260 nm. Next we investigated the dependence of the excited-state dynamics on initial excitation energy. Figure 4 shows a comparison of two contour plots of transient absorption spectra in the visible range for all-trans-capsanthin in acetonitrile after excitation at 500 nm (A) and 400 nm (B). Panel C compares representative kinetics at 475 nm (S0 → S2

GSB) and 575 nm (S1 → Sn ESA) for these two conditions with 5000 cm−1 difference in initial excess energy. A clear difference between the contour plots is seen in the S1 → Sn ESA band in the first 2 ps. The absorption on the red edge of the ESA band around 660 nm is considerably larger and decays slower when capsanthin is excited at higher excess energy. After IC from S2 the initially formed internally nonequilibrated molecules in S1 obviously exhibit increased absorption. Differences are also observed in the kinetics: For instance, at the top of the ESA band (575 nm, panel C), a difference in curvature is observed for the transients up to ca. 2 ps. Global kinetic analysis (section 3.3) shows that internal equilibration takes longer (τIVR,1 = 420 vs 340 fs) if the molecules enter S1 with higher excess energy. In contrast, the kinetics in the GSB (475 nm) is virtually the same, showing that the IC time constant is largely unaffected. Figure 5 compares PSCP transient broadband absorption spectra of capsanthin in acetonitrile (A) and isooctane (B) after

Figure 5. Comparison of transient absorption spectra of all-transcapsanthin in (A) acetonitrile and (B) isooctane upon photoexcitation at 400 nm (T = 296 K). The different panels highlight the excited-state dynamics in time windows of particular interest. The inverted steadystate absorption spectrum for each case is shown in the lowest panel as a dashed orange line.

excitation at 400 nm. The dynamics in both solvents is similar, the main difference being the more structured shape of the GSB and ESA bands in the nonpolar solvent. This behavior is consistent with the loss in structure found for the ground-state steady-state absorption spectra (Figure 2). In addition, global kinetic analysis identifies a slightly slower S1 → S0 IC process in the nonpolar solvent (τ1 = 10.3 vs 9.7 ps). A corresponding comparison for acetonitrile and methanol at the pump wavelength 500 nm is shown in the Supporting Information (Figure S2). In this case, the transient spectra in the two solvents are structureless and very similar in shape, the only difference being the slightly faster decay (9.1 ps) to the ground electronic state in the case of methanol. 3.3. Global Kinetic Analysis. The global analysis of the UV−vis−NIR transient absorption data for all-trans-capsanthin was carried out using the kinetic model presented in Figure 6. It contains the minimum set of species and time constants required to successfully describe all details of the transient spectral behavior: The initially prepared S2 species decays by IC

Figure 4. (A) Contour plot of transient absorption spectra of all-transcapsanthin in acetonitrile for excitation at 500 nm. (B) Same, but for excitation at 400 nm. (C) Comparison of the kinetic time traces at the probe wavelengths 475 and 575 nm. Different colors of the lines indicate the different pump and probe wavelength conditions. C


Article

The Journal of Physical Chemistry A

Figure 6. Kinetic model for the relaxation dynamics of all-transcapsanthin in organic solvents.

with the time constant τ2 to an internally nonequilibrated hot species denoted as S1**.29,30 Subsequent IVR of S1** with the time constant τIVR,1 populates a vibrationally hot species S1* with a close to statistical energy distribution in its vibrational modes. This internally largely equilibrated but vibrationally hot species relaxes by CET with solvent molecules to a thermally (296 K) equilibrated species S1 (time constant τCET,1). The S1**, S1*, and S1 species decay to S0 with the same time constant τ1. In our analysis we found that this already led to a perfect description of the experimental data sets. Therefore, no further attempts were made to assign individual IC time constants for the different S1 species. Because of the very weak S1 absorption signal across the entire NIR range, a simplified consecutive model S2 → S1 → S0 (time constants τ2 and τ1) was applied in that spectral region, with both time constants being shared during the fitting procedure over the whole UV−vis−NIR range. The optimized SAS of S2, S1**, S1*, and S1 as well as the fixed spectrum for S0 are depicted in Figure 7, with the absolute absorption coefficients referenced to the value for S0 in n-hexane provided by Polgár and Zechmeister (115 000 M−1 cm−1, also employed in Figure 2).28 Contributions of the individual species at four selected time delays are highlighted in the Supporting Information (Figure S3) for capsanthin in acetonitrile. A summary of the optimized time constants is provided in Table 1. The S2 lifetime of capsanthin (conjugation length N = 11) in the two polar solvents is 116 fs and independent of excitation energy. In isooctane it increases slightly to 154 fs. These lifetimes fit well into the trend previously observed by us for a series of β-apocarotenals in n-hexane, where values of 200, 180, and 105 fs were found for N = 14, 12, and 8.4 The SAS of the S1** species is clearly broader than the SAS of the S1* species; see e.g. capsanthin in isooctane in Figure 7D. This change in shape and the subpicosecond time scale of the process are consistent with previous findings for IVR-induced spectral narrowing in transient absorption spectra of large organic molecules, such as trans-stilbene.31 Similar dynamics and time scales have been later on reported for lycopene, βcarotene, and zeaxanthin29,30 and afterward e.g. also found in other systems, such as 13,13′-diphenyl-β-carotene14 and β-apo12′-carotenal.4 IVR also appears to depend on excess energy: In acetonitrile the time constant τIVR,1 for intramolecular vibrational redistribution increases from 340 fs (500 nm excitation) to 420 fs (400 nm excitation). Obviously, it takes slightly longer to equilibrate a higher internal excess energy among the different vibrational degrees of freedom in the first electronically excited state. The further spectral narrowing of the electronic band on a 10 ps time scale is consistent with the findings of transient absorption experiments on CET between vibrationally hot species and solvent molecules performed over wide pressure ranges.4,15,31−34 The shape of the SAS of S1* and S1 in

Figure 7. Species-associated spectra from the global kinetic analysis of the transient absorption spectra of all-trans-capsanthin: (A) acetonitrile, 500 nm excitation; (B) acetonitrile, 400 nm excitation; (C) methanol, 500 nm excitation; (D) isooctane, 400 nm excitation. Different colors of the solid lines indicate the different species in the global kinetic model.

isooctane in Figure 7D is representative for such a CET process. The time constant τCET,1 for collisional energy transfer of S1* is almost solvent-independent and lies in the range 8.5− 9.1 ps. Such a value is fully consistent with previous findings for CET time constants of other systems, such as β-apo-12′carotenal in S1 (8.3 ps),4 longer conjugated β-apo-n′-carotenals and β-carotene derivatives in S0 (9−12 ps),4,15,35 and azulene in S0.34 The polarity dependence of the S1 lifetime τ1 is rather weak. IC accelerates from isooctane (10.3 ps) to methanol (9.1 ps) by only about a picosecond. This already suggests a rather weak influence of ICT character on the nonradiative decay rate of the first electronically excited singlet state. The S1 decay of capsanthin in methanol is slightly faster than in acetonitrile although both solvents have the same polarity (0.71 on the Δf scale).8 Therefore, hydrogen-bond interactions might provide additional stabilization of the ICT state, resulting in a reduction of the lifetime τ1. Representative fits of the all-trans-capsanthin kinetics in acetonitrile upon excitation at 500 nm are presented in Figures 8 and 9 for short and long time scales, respectively. The kinetics in the NIR at 900 nm is mainly sensitive to the S2 lifetime, resulting in a very accurate τ2 value. Further in the NIR range (1400 nm) the amplitude of the absorption transient is much weaker, and the dynamics is dominated by the decay of the S1 state. The contributions of the IVR, CET, and IC processes to the dynamics of the S1 state are clearly demonstrated in the transient at 565 nm. The delayed rise of the signal is due to the D


Article

The Journal of Physical Chemistry A Table 1. Time Constants for Capsanthin from the Global Kinetic Analysis Including Statistical Errors (±2σ) solvent

λpumpa (nm)

τCCb (fs)

isooctane MeCN MeCN MeOH

400 400 500 500

90 90 110 110

τ2c (fs) 154 115 116 116

± ± ± ±

3 2 2 7

τIVR,1d (fs) 340 420 340 340

± ± ± ±

10 10 10 10

τCET,1e (ps) 9.1 8.5 8.6 8.6

± ± ± ±

0.1 0.2 0.4 0.4

τ1f (ps)

τS1**g (fs)

τS1*h (ps)

± ± ± ±

330 400 330 330

4.6 4.5 4.6 4.4

10.3 9.7 9.9 9.1

0.1 0.1 0.1 0.1

a

Pump wavelength of the PSCP experiments. bCross-correlation time of the PSCP experiments. cTime constant for internal conversion from S2 to internally not equilibrated S1 molecules with vibrational excess energy (S1**). dTime constant for intramolecular vibrational redistribution of S1** resulting in the formation of internally equilibrated S1 molecules with vibrational excess energy (S1*). eTime constant for deactivation of vibrationally hot S1* molecules by collisional energy transfer to the solvent. This results in molecules with a thermal (296 K) energy distribution (S1). fTime constant for internal conversion of the different S1 species to S0, i.e., the same time constant τ1 was employed for S1**, S1*, and S1. gLifetime of the S1** species calculated as τS1** = (τIVR,1−1 + τ1−1)−1. hLifetime of the S1* species calculated as τS1* = (τCET,1−1 + τ1−1)−1.

transient is dominated by the nonradiative decay to the ground electronic state S0 with superimposed contributions of CET. We note that up to about 6 ps the IC process to S0 predominantly occurs from the vibrationally hot S1* species (blue line), whereas at longer times it is governed by IC of the cooled S1 species (magenta line). The absence of any stimulated emission (SE) features in the kinetics together with the only weak dependence of the lifetime τ1 on solvent polarity suggests that all-trans-capsanthin exhibits only weak intramolecular charge transfer (ICT) character in S1. 3.4. Comparison with Other Carbonyl-Based Apocarotenoids. Figure 10 shows a comparison of the lifetimes τ1

Figure 8. Contributions of the individual species S0, S1, S1*, S1**, and S2 to the kinetic traces at early times for all-trans-capsanthin in acetonitrile upon 500 nm photoexcitation.

Figure 10. Dependence of the lifetime τ1 of the first excited electronic state on the number of conjugated double bonds N for all-transapocarotenals (black), all-trans-apocarotenoic acids (red), and the apocarotenone all-trans-capsanthin (blue). Filled symbols: n-hexane or isooctane. Open symbols: methanol. The value for all-trans-retinoic acid in methanol was obtained from a preliminary kinetic analysis of PSCP experiments. Solid and dashed lines are only intended as a guide for the eye. Note the semilogarithmic representation.

determined previously for a range of terminally carbonylsubstituted apocarotenoids with different conjugation lengths featuring aldehyde, carboxyl, and keto end groups.3−9,36 In nonpolar solvents (n-hexane or isooctane), the S1 lifetime increases strongly with decreasing conjugation length, e.g., from ca. 1 to 200 ps in the apocarotenal series torularhodinaldehyde (N = 14) up to β-apo-12′-carotenal (N = 8); see the filled black

Figure 9. Same as in Figure 8, but for longer time scales up to 40 ps.

formation of S1** by IC from S2. The curvature around 0.4 ps arises from IVR of S1** to form S1*. On longer time scales the E


Article


character. After nonradiative decay from 11Bu+ to 21Ag− via a conical intersection a repopulation of 11Bu+ is assumed to occur induced by torsional motion.12 We add that solvent relaxation likely contributes to the dynamics of the “S1/ICT” spectral bands, making the interpretation of the time constants even more complex.5,46 In any case, on the basis of the studies of Wagner et al. and Di Donato et al., it appears reasonable that the S2(11Bu+) state is the main contributor to the ICT character in carbonyl carotenoids.12,43 The shorter-chain derivatives all-trans-retinal and all-transretinoic acid appear to be special cases.3 While retinoic acid exhibits acceleration of the IC process in polar solvents, the lifetime is in general much shorter than expected from the EGL correlation. In all-trans-retinal, the nonradiative decay is slower in polar solvents. This is again connected to the special influence of the 1nπ* state, where the S1 → 1nπ* transition is slowed down substantially with increasing solvent polarity.3 In addition, all-trans-retinal so far has been the only compound of the apocarotenal family where separate S1 and ICT species were clearly identified in transient absorption experiments based on their distinctly different kinetic behavior (black circles and squares in Figure 10).3 As discussed above, the separate ICT state3 may be indeed identified with the relaxed S2(11Bu+) state.12,43 3.5. Steric Congestion at the Terminal Keto Group of Capsanthin. We finally investigated the effect of a different substituent at the keto group and its impact on the excited-state dynamics. For that we carried out selected single-wavelength pump−probe transient absorption experiments of citranaxanthin in isooctane, acetonitrile, and methanol. Citranaxanthin possesses the same β-apo-6′-carotenone-based polyene backbone but features a smaller methyl substituent at the keto group instead of the bulky five-membered ring substituent of capsanthin.10 In addition, citranaxanthin lacks the hydroxy substituent at the 3-position (β-ionone ring).10 This will however have no effect on the excited-state dynamics because this OH substituent is not attached to the polyene backbone. Figure 11 compares single-wavelength kinetics of capsanthin (red) and citranaxanthin (black) in the GSB region at 475 nm for the solvents isooctane (A) and acetonitrile (B). Panel C shows the very similar steady-state absorption spectra of both apocarotenones in the two solvents. The kinetic traces for isooctane in panel A are virtually identical, and one obtains the values τ1 = 10.3 and 10.2 ps for capsanthin and citranaxanthin, respectively. The identical S1 lifetime in the nonpolar solvent suggests that the effective conjugation length of both apocarotenones is the same. In panel B corresponding results are shown for polar acetonitrile. Here the S1 lifetime of citranaxanthin is considerably shorter (8.1 ps) than for capsanthin (9.7 ps). Because both apocarotenones feature the same conjugation length, the effect must be due to the stronger ICT character in the case of citranaxanthin which leads to an acceleration of the IC process to S0. Obviously there is a better solvent stabilization of the negative partial charge on the keto group of citranaxanthin than in the case of capsanthin. This suggests that the bulky five-membered ring of capsanthin sterically perturbs an optimal solvation of the excited-state dipole. The importance of such local solvent−solute interactions in stabilizing the ICT character was also pointed out previously. For instance, local charge-dipole interactions of inorganic cations in alcoholic salt solutions11 or organic cations of ionic

circles and solid black line.4 As demonstrated previously,9 this dependence of τ1 on the conjugation length may be described by the energy-gap-law (EGL) expression τ1 = a exp(bΔEgap), where a and b are constants and ΔEgap is the S1−S0 energy gap. The smaller S1−S0 energy gap in long-chain apocarotenoids therefore results in an acceleration of the nonradiative IC process.9,37,38 We note that this trend is independent of the type of carbonyl group attached to the polyene backbone; see e.g. the time constants for the apocarotenoic acid series (filled red circles and red solid line) which are identical to the values for the respective aldehydes. Also, the present value for the β-apo6′-carotenone derivative capsanthin (N = 11, filled blue circle) fits very well into this correlation. We note that the correlation breaks down for very short conjugation lengths, e.g., in the case of all-trans-retinal (= βapo-15-carotenal) and all-trans-retinoic acid (= β-apo-15carotenoic acid), both having N = 6.3 In the case of all-transretinal the presence of a 1nπ* state close to S2 and S1 opens up an ultrafast pathway to the T1 triplet state which reduces the lifetime dramatically.3 This 1nπ* channel is switched off in retinoic acid,3 but still the lifetime is much shorter than predicted by the simple EGL expression. A decrease in lifetime is observed for apocarotenals in the polar solvent methanol (open black circles and dashed black line), with the most dramatic effect in the case of β-apo-12′carotenal.5,11 The polarity effect becomes much weaker upon polyene chain extension; see e.g. β-apo-4′-carotenal (N = 12).4,9 It also becomes weaker upon changing the type of carbonyl group in the sequence aldehyde > ketone > carboxyl. The small (1 ps) decrease of τ1 for capsanthin (open blue circle) is therefore due to a combination of a fairly long conjugation length and rather weak ICT character induced by the keto group. The polarity-induced lifetime reduction of terminally carbonyl-substituted apocarotenoids is usually assigned to the presence of ICT character in the S1 state. This state is typically denoted as “S1/ICT”. The acceleration of the nonradiative IC process may be tentatively explained by a polarity-induced decrease of the S1/ICT−S0 energy gap. We note that experimental evidence for separate S1 and ICT species has not yet been found for the systems down to N = 8. The presence of ICT character was also observed for the carbonyl carotenoid peridinin and closely related compounds.18,39−43 The nature of the “S1/ICT” state is still under debate: The model proposed by Frank, Birge, and co-workers for the carbonyl-carotenoid peridinin assumes that the electronic properties of the ICT state are generated by polar-solventinduced mixing of the 11Bu+ and the 21Ag− states, resulting in an ICT state with predominantly 1Bu+ character and large oscillator strength.43 They suggest that in most solvents two populations exist in equilibrium: one with a lowest-lying ICT ionic state and a second with a lowest-lying 21Ag− covalent state. The two populations are separated by a small barrier.43 Such a small barrier and a presumably fast equilibrium are actually compatible with the aforementioned observation that in transient absorption experiments for carbonyl apocarotenoids down to N = 8 the decay of the “S1/ICT” ESA is monoexponential. Another similar model was proposed for βapo-8′-carotenal by Di Donato et al. based on transient IR and Raman experiments as well as computations.12,44,45 Here the energetic difference between 11Bu+ and the 21Ag− states is small, and again the former one mainly contributes to the ICT F



■

Article

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jpca.7b08252. Contour plot and kinetics for all-trans-capsanthin in acetonitrile, comparison of transient absorption spectra in acetonitrile and methanol, contributions of different electronic species to the PSCP spectra (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*E-mail [email protected]; Tel +49-271-740-2803 (K.O.). ORCID

Mirko Scholz: 0000-0002-5648-7925 Thomas Lenzer: 0000-0002-0766-709X Kawon Oum: 0000-0001-6137-2236 Notes

The authors declare no competing financial interest.

■

liquids47 with the carbonyl group resulted in an acceleration of the IC process.

ACKNOWLEDGMENTS Highly purified all-trans-citranaxanthin samples were kindly provided by Hansgeorg Ernst (BASF SE). We thank Duncan A. Wild for recording the kinetic transients at 475 nm of citranaxanthin in isooctane and acetonitrile as well as AnnaCamille Pokorra for her participation in this capsanthin project during her school lab course. We are also grateful to Nikolaus P. Ernsting (Humboldt University Berlin, Germany), José Luis Pérez Lustres (University of Heidelberg, Germany), Jürgen Troe, Klaus Luther, Jörg Schroeder, Dirk Schwarzer, and Alec M. Wodtke (Georg August University Göttingen, Germany) for their continuous support and advice.

4. CONCLUSIONS Our investigation of the photoinduced excited-state dynamics of all-trans-capsanthin in solution provided detailed information regarding the photophysics of this β-apo-6′-carotenone derivative. The lifetime of the initially populated S2 state was about 120−150 fs and only weakly dependent on solvent polarity. Internal conversion populated an internally not yet equilibrated S1 state which exhibited intramolecular vibrational redistribution (340−420 fs) and collisional energy transfer to the solvent (8.5−9.1 ps). The lifetime of S1 dropped mildly from 10.3 ps in isooctane to 9.1 ps in methanol. This only weak polarity-induced decrease of the S1 lifetime and the absence of any S1 stimulated emission suggested that only weak ICT character was present in the first excited singlet state of capsanthin. A systematic comparison with previously studied carbonyl apocarotenoids featuring terminal CHO, CRO, or COOH substituents identified the fairly long conjugation length and the type of carbonyl (here keto) substitution as the main reasons for the weak ICT character. Both properties determine how strongly the S1−S0 energy gap is varied upon polarity increase. A comparison with the apocarotenone citranaxanthin having the same conjugated polyene system but a methyl substituent instead of a five-membered ring attached to the keto group showed that steric congestion induced by the bulky ring hampers efficient dipolar solvation of the carbonyl group in polar solvents. This explains the even weaker expression of ICT character for capsanthin’s S1 state.

(1) van den Berg, H.; Faulks, R.; Fernando Granado, H.; Hirschberg, J.; Olmedilla, B.; Sandmann, G.; Southon, S.; Stahl, W. The Potential for the Improvement of Carotenoid Levels in Foods and the Likely Systemic Effects. J. Sci. Food Agric. 2000, 80, 880−912. (2) Lefebvre, V.; Kuntz, M.; Camara, B.; Palloix, A. The CapsanthinCapsorubin Synthase Gene: A Candidate Gene for the y Locus Controlling the Red Fruit Colour in Pepper. Plant Mol. Biol. 1998, 36, 785−789. (3) Flender, O.; Scholz, M.; Hölzer, J.; Oum, K.; Lenzer, T. A Comprehensive Picture of the Ultrafast Excited-State Dynamics of Retinal. Phys. Chem. Chem. Phys. 2016, 18, 14941−14948. (4) Ehlers, F.; Scholz, M.; Schimpfhauser, J.; Bienert, J.; Oum, K.; Lenzer, T. Collisional Relaxation of Apocarotenals: Identifying the S* State with Vibrationally Excited Molecules in the Ground Electronic State S0*. Phys. Chem. Chem. Phys. 2015, 17, 10478−10488. (5) Oum, K.; Lohse, P. W.; Ehlers, F.; Scholz, M.; Kopczynski, M.; Lenzer, T. 12′-Apo-β-caroten-12′-al: An Ultrafast “Spy” Molecule for Probing Local Interactions in Ionic Liquids. Angew. Chem., Int. Ed. 2010, 49, 2230−2232. (6) Ehlers, F.; Lenzer, T.; Oum, K. Excited-State Dynamics of 12′Apo-β-caroten-12′-al and 8′-Apo-β-caroten-8′-al in Supercritical CO2, N2O, and CF3H. J. Phys. Chem. B 2008, 112, 16690−16700. (7) Stalke, S.; Wild, D. A.; Lenzer, T.; Kopczynski, M.; Lohse, P. W.; Oum, K. Solvent-Dependent Ultrafast Internal Conversion Dynamics of n′-Apo-β-carotenoic-n′-acids (n = 8, 10, 12). Phys. Chem. Chem. Phys. 2008, 10, 2180−2188. (8) Kopczynski, M.; Ehlers, F.; Lenzer, T.; Oum, K. Evidence for an Intramolecular Charge Transfer (ICT) State in 12′-Apo-β-caroten-12′al and 8′-Apo-β-caroten-8′-al: Influence of Solvent Polarity and Temperature. J. Phys. Chem. A 2007, 111, 5370−5381.

Figure 11. Comparison of normalized transient absorption decays of all-trans-capsanthin (red) and all-trans-citranaxanthin (black) in the GSB region at 475 nm in (A) isooctane and (B) acetonitrile. Excitation was performed at 400 nm in the case of capsanthin and at 390 nm in the case of citranaxanthin. Fit results for the lifetime τ1 of the S1 state are indicated. In (C) the steady-state absorption spectra of capsanthin (red) and citranaxanthin (black) are shown. Solid and dashed lines are for the solvents isooctane and acetonitrile, respectively.

■

G

REFERENCES


Article


(3R,3′R,6′R)-Lutein, Echinenone, Canthaxanthin, and Astaxanthin. Phys. Chem. Chem. Phys. 2005, 7, 2793−2803. (26) Ekimova, M.; Fröhlich, D.; Stalke, S.; Lenzer, T.; Oum, K. Probing the Local Polarity of Alkylammonium Formate Ionic Liquids and Their Mixtures with Water by Using a Carbonyl Carotenoid. ChemPhysChem 2012, 13, 1854−1859. (27) Polívka, T.; Herek, J. L.; Zigmantas, D.; Akerlund, H.; Sundström, V. Direct Observation of the (Forbidden) S1 State in Carotenoids. Proc. Natl. Acad. Sci. U. S. A. 1999, 96, 4914−4917. (28) Polgár, A.; Zechmeister, L. A Spectroscopic Study in the Stereoisomeric Capsanthin Set. cis-Peak Effect and Configuration. J. Am. Chem. Soc. 1944, 66, 186−190. (29) Billsten, H. H.; Zigmantas, D.; Sundström, V.; Polívka, T. Dynamics of Vibrational Relaxation in the S1 state of Carotenoids Having 11 Conjugated CC Bonds. Chem. Phys. Lett. 2002, 355, 465−470. (30) de Weerd, F. L.; van Stokkum, I. H. M.; van Grondelle, R. Subpicosecond Dynamics in the Excited State Absorption of all-transβ-Carotene. Chem. Phys. Lett. 2002, 354, 38−43. (31) Kovalenko, S. A.; Schanz, R.; Hennig, H.; Ernsting, N. P. Cooling Dynamics of an Optically Excited Molecular Probe in Solution from Femtosecond Broadband Absorption Spectroscopy. J. Chem. Phys. 2001, 115, 3256−3273. (32) Hippler, H.; Troe, J.; Wendelken, H. J. UV Absorption Spectra of Vibrationally Highly Excited Toluene Molecules. J. Chem. Phys. 1983, 78, 5351−5357. (33) Brouwer, L.; Hippler, H.; Lindemann, L.; Troe, J. Measurement of Internal Energies by Hot Ultraviolet Absorption Spectroscopy: Spectra of Excited Azulene Molecules. J. Phys. Chem. 1985, 89, 4608− 4612. (34) Schwarzer, D.; Troe, J.; Votsmeier, M.; Zerezke, M. Collisional Deactivation of Vibrationally Highly Excited Azulene in Compressed Liquids and Supercritical Fluids. J. Chem. Phys. 1996, 105, 3121−3131. (35) Ehlers, F.; Scholz, M.; Lenzer, T.; Oum, K. Excited State Dynamics of Selected All-trans C40 Xanthophyll Carotenoids. Z. Phys. Chem. 2015, 229, 1815−1830. (36) He, Z.; Gosztola, D.; Deng, Y.; Gao, G.; Wasielewski, M. R.; Kispert, L. D. Effect of Terminal Groups, Polyene Chain Length, and Solvent on the First Excited Singlet States of Carotenoids. J. Phys. Chem. B 2000, 104, 6668−6673. (37) Chynwat, V.; Frank, H. A. The Application of the Energy Gap Law to the S1 Energies and Dynamics of Carotenoids. Chem. Phys. 1995, 194, 237−244. (38) Englman, R.; Jortner, J. The Energy Gap Law for Radiationless Transitions in Large Molecules. Mol. Phys. 1970, 18, 145−164. (39) Frank, H. A.; Bautista, J. A.; Josue, J.; Pendon, Z.; Hiller, R. G.; Sharples, F. P.; Gosztola, D.; Wasielewski, M. R. Effect of the Solvent Environment on the Spectroscopic Properties and Dynamics of the Lowest Excited States of Carotenoids. J. Phys. Chem. B 2000, 104, 4569−4577. (40) Bautista, J. A.; Connors, R. E.; Raju, B. B.; Hiller, R. G.; Sharples, F. P.; Gosztola, D.; Wasielewski, M. R.; Frank, H. A. Excited State Properties of Peridinin: Observation of a Solvent Dependence of the Lowest Excited Singlet State Lifetime and Spectral Behavior Unique among Carotenoids. J. Phys. Chem. B 1999, 103, 8751−8758. (41) Zigmantas, D.; Hiller, R. G.; Sharples, F. P.; Frank, H. A.; Sundström, V.; Polívka, T. Effect of a Conjugated Carbonyl Group on the Photophysical Properties of Carotenoids. Phys. Chem. Chem. Phys. 2004, 6, 3009−3016. (42) Fuciman, M.; Enriquez, M. M.; Kaligotla, S.; Niedzwiedzki, D. M.; Kajikawa, T.; Aoki, K.; Katsumura, S.; Frank, H. A. Singlet and Triplet State Spectra and Dynamics of Structurally Modified Peridinins. J. Phys. Chem. B 2011, 115, 4436−4445. (43) Wagner, N. L.; Greco, J. A.; Enriquez, M. M.; Frank, H. A.; Birge, R. R. The Nature of the Intramolecular Charge Transfer State in Peridinin. Biophys. J. 2013, 104, 1314−1325. (44) Di Donato, M.; Ragnoni, E.; Lapini, A.; Kardaś, T. M.; RatajskaGadomska, B.; Foggi, P.; Righini, R. Identification of the Excited-State CC and CO Modes of trans-β-Apo-8′-carotenal with Transient

(9) Ehlers, F.; Wild, D. A.; Lenzer, T.; Oum, K. Investigation of the S1/ICT ® S0 Internal Conversion Lifetime of 4′-Apo-β-caroten-4′-al and 8′-Apo-β-caroten-8′-al: Dependence on Conjugation Length and Solvent Polarity. J. Phys. Chem. A 2007, 111, 2257−2265. (10) Lenzer, T.; Oum, K.; Seehusen, J.; Seidel, M. T. Transient Lens Spectroscopy of Ultrafast Internal Conversion Processes in Citranaxanthin. J. Phys. Chem. A 2006, 110, 3159−3164. (11) Wild, D. A.; Winkler, K.; Stalke, S.; Oum, K.; Lenzer, T. Extremely Strong Solvent Dependence of the S1 → S0 Internal Conversion Lifetime of 12′-Apo-β-caroten-12′-al. Phys. Chem. Chem. Phys. 2006, 8, 2499−2505. (12) Di Donato, M.; Centellas, M. S.; Lapini, A.; Lima, M.; Avila, F.; Santoro, F.; Cappelli, C.; Righini, R. Combination of Transient 2D-IR Experiments and Ab Initio Computations Sheds Light on the Formation of the Charge-Transfer State in Photoexcited Carbonyl Carotenoids. J. Phys. Chem. B 2014, 118, 9613−9630. (13) Polívka, T.; Kaligotla, S.; Chábera, P.; Frank, H. A. An Intramolecular Charge Transfer State of Carbonyl Carotenoids: Implications for Excited State Dynamics of Apo-carotenals and Retinal. Phys. Chem. Chem. Phys. 2011, 13, 10787−10796. (14) Golibrzuch, K.; Ehlers, F.; Scholz, M.; Oswald, R.; Lenzer, T.; Oum, K.; Kim, H.; Koo, S. Ultrafast Excited State Dynamics and Spectroscopy of 13,13′-Diphenyl-β-carotene. Phys. Chem. Chem. Phys. 2011, 13, 6340−6351. (15) Lenzer, T.; Ehlers, F.; Scholz, M.; Oswald, R.; Oum, K. Assignment of Carotene S* State Features to the Vibrationally Hot Ground Electronic State. Phys. Chem. Chem. Phys. 2010, 12, 8832− 8839. (16) Ilagan, R. P.; Koscielecki, J. F.; Hiller, R. G.; Sharples, F. P.; Gibson, G. N.; Birge, R. R.; Frank, H. A. Femtosecond Time-Resolved Absorption Spectroscopy of Main-Form and High-Salt PeridininChlorophyll a-Proteins at Low Temperatures. Biochemistry 2006, 45, 14052−14063. (17) Gelzinis, A.; Butkus, V.; Songaila, E.; Augulis, R.; Gall, A.; Büchel, C.; Robert, B.; Abramavicius, D.; Zigmantas, D.; Valkunas, L. Mapping Energy Transfer Channels in Fucoxanthin-Chlorophyll Protein Complex. Biochim. Biophys. Acta, Bioenerg. 2015, 1847, 241− 247. (18) Kosumi, D.; Kajikawa, T.; Sakaguchi, K.; Katsumura, S.; Hashimoto, H. Excited State Properties of β-Carotene Analogs Incorporating a Lactone Ring. Phys. Chem. Chem. Phys. 2017, 19, 3000−3009. (19) Oum, K.; Lenzer, T.; Scholz, M.; Jung, D. Y.; Sul, O.; Cho, B. J.; Lange, J.; Müller, A. Observation of Ultrafast Carrier Dynamics and Phonon Relaxation of Graphene from the Deep-Ultraviolet to the Visible Region. J. Phys. Chem. C 2014, 118, 6454−6461. (20) Flender, O.; Scholz, M.; Klein, J. R.; Oum, K.; Lenzer, T. Excited-State Relaxation of the Solar Cell Dye D49 in Organic Solvents and on Mesoporous Al2O3 and TiO2 Thin Films. Phys. Chem. Chem. Phys. 2016, 18, 26010−26019. (21) Dobryakov, A. L.; Kovalenko, S. A.; Weigel, A.; Pérez Lustres, J. L.; Lange, J.; Müller, A.; Ernsting, N. P. Femtosecond Pump/ Supercontinuum-Probe Spectroscopy: Optimized Setup and Signal Analysis for Single-Shot Spectral Referencing. Rev. Sci. Instrum. 2010, 81, 113106. (22) Scholz, M.; Flender, O.; Boschloo, G.; Oum, K.; Lenzer, T. Ultrafast Electron and Hole Transfer Dynamics of a Solar Cell Dye Containing Hole Acceptors on Mesoporous TiO2 and Al2O3. Phys. Chem. Chem. Phys. 2017, 19, 7158−7166. (23) Oum, K.; Lohse, P. W.; Flender, O.; Klein, J. R.; Scholz, M.; Lenzer, T.; Du, J.; Oekermann, T. Ultrafast Dynamics of the Indoline Dye D149 on Electrodeposited ZnO and Sintered ZrO2 and TiO2 Thin Films. Phys. Chem. Chem. Phys. 2012, 14, 15429−15437. (24) Polívka, T.; Sundström, V. Ultrafast Dynamics of Carotenoid Excited States - From Solution to Natural and Artificial Systems. Chem. Rev. 2004, 104, 2021−2071. (25) Kopczynski, M.; Lenzer, T.; Oum, K.; Seehusen, J.; Seidel, M. T.; Ushakov, V. G. Ultrafast Transient Lens Spectroscopy of Various C40 Carotenoids: Lycopene, β-Carotene, (3R,3′R)-Zeaxanthin, H


Article

The Journal of Physical Chemistry A 2D-IR-EXSY and Femtosecond Stimulated Raman Spectroscopy. J. Phys. Chem. Lett. 2015, 6, 1592−1598. (45) Ragnoni, E.; Di Donato, M.; Iagatti, A.; Lapini, A.; Righini, R. Mechanism of the Intramolecular Charge Transfer State Formation in all-trans-β-Apo-8′-carotenal: Influence of Solvent Polarity and Polarizability. J. Phys. Chem. B 2015, 119, 420−432. (46) Lohse, P. W.; Ehlers, F.; Oum, K.; Scholz, M.; Lenzer, T. Ultrafast Solvation Dynamics of 12′-Apo-β-carotenoic-12′-acid in [C6mim]+[Tf2N]−. Chem. Phys. 2010, 373, 45−49. (47) Lohse, P. W.; Bürsing, R.; Lenzer, T.; Oum, K. Exploring 12′Apo-β-carotenoic-12′-acid as an Ultrafast Polarity Probe for Ionic Liquids. J. Phys. Chem. B 2008, 112, 3048−3057.

I


Ultrafast Excited-State Dynamics of all-trans ... - ACS Publications

Recommend Documents