⇦Chinese Physics Letters, 2017, Vol. 34, No. 4, Article code 043101 Two-Photon Excited State Dynamics of Dark Rydberg and Bright Valence States in Furan * Jin-You Long(龙金友), Chun-Long Hu(胡春龙), Bing Zhang(张冰)** Affiliations State Key Laboratory of Magnetic Resonance and Atomic and Molecular Physics, Wuhan Institute of Physics and Mathematics, Chinese Academy of Sciences, Wuhan 430071 Received 29 November 2016 *Supported by the National Natural Science Foundation of China under Grant Nos 21303255, 21273274 and 91121006.
**Corresponding author. Email: bzhang@wipm.ac.cn Citation Text: Long J Y, Hu C L and Zhang B 2017 Chin. Phys. Lett. 34 043101 Abstract Two-photon absorption in systems with parity permits access to states that cannot be directly prepared by one-photon absorption. Here we investigate ultrafast internal conversion (IC) dynamics of furan by using this strategy in combination with femtosecond time-resolved photoelectron imaging. The dark Rydberg $S_{1}$ and bright valence $S_{2}$ states are simultaneously excited by two photons of 405 nm, and then ionized by two photons of 800 nm. The IC from $S_{2}$ to $S_{1}$ is clearly observed and extracted from the time dependence of the higher photoelectron kinetic energy (PKE) component. More importantly, the internal conversions to hot $S_{0}$ from directly-prepared $S_{1}$ and secondarily-populated $S_{1}$ are unambiguously identified by the time-dependence of the lower PKE component. The average lifetime of the $S_{2}$ and $S_{1}$ states is measured to be 29 fs. The internal conversions of $S_{2}$ to $S_{1}$, $S_{1}$ to hot $S_{0 }$ occur on estimated timescales of 15.4 fs and 38 fs, respectively. DOI:10.1088/0256-307X/34/4/043101 PACS:31.70.Hq, 33.15.Vb, 33.50.-j © 2017 Chinese Physics Society Article Text In systems with parity, nonresonant two-photon excitation is the easiest way to prepare dark excited states that are inaccessible via one photon absorption. In combination with the femtosecond time-resolved photoelectron imaging (TR-PEI) technique, one can utilize this strategy to investigate the nonadiabatic dynamics of dark excited states as well as bright excited states, especially deciphering the role of the dark states in the relaxation dynamics. TR-PEI could measure both the energy and angular distributions of the photoelectrons as well as their correlation as a function of time.[1] Such nonadiabatic dynamics in polyatomic molecules are essential in many photophysical or photobiological processes such as the photostability of DNA and photoisomerization in vision.[2] Furan has served as an excellent prototype system for studying nonadiabatic dynamics involving the nature of bright and dark excited states. In light of the importance of furan and its derivatives relevant for atmospheric degradation[3] and combustion chemistry,[4] numerous experimental[5-13] and theoretical[14-22] studies have been performed on furan, largely focusing on assignments and characterizations of its first vacuum ultraviolet (VUV) absorption spectrum. Four low-lying singlet excited states, i.e., $^{1}A_{2}(\pi3s)$, $^{1}B_{2}(\pi \pi*)$, $^{1}A_{1}(\pi \pi*)$ and $^{1}B_{1}(\pi3p_{y})$ have been identified and assigned in the energy range of 5.7–6.5 eV. Among them, the bright $S_{2}[^{1}B_{2}(\pi \pi*)]$ valence state has rather large optical oscillator strength, whereas the $S_{1}[^{1}A_{2}(\pi3s)]$ Rydberg state is a one-photon forbidden dark state. The ongoing interest in the excited state potential energy surfaces and photodynamics of excited furan is reflected by several recent theoretical[23-29] and experimental[30-32] works. Gromov et al.[23,24] predicted the timescale for the ultrafast IC from the $S_{2}$ state to the $S_{1}$ state to be $\sim$25 fs. However, the IC to the $S_{0}$ state has not been considered in their model. Fuji et al.[30,31] have performed the first experimental observation of the one-photon excited state dynamics in furan and proposed that the IC from $S_{2}$ to $S_{0}$ rather than $S_{1}$ occurs within 29 fs. Recently, Liu et al.[32] have carried out photo-fragmentation spectroscopy of furan and proposed the possibility of formation of $\alpha$- and $\beta$-carbenes during the deactivation of the $S_{1}$ state. Therefore, the full relaxation pathways for the deactivation of electronically excited furan are still not fully concluded. In this work, we apply the TR-PEI technique to investigate the excited state dynamics of furan. The full relaxation pathways and the role of the dark $S_{1}$ state in the relaxation dynamics have been directly interrogated experimentally. The timescales for the relaxation processes have also been determined. Our femtosecond laser system[33] and experimental setup[34] employed in the present work have been described elsewhere. The second harmonic pulse was generated in a 0.5-mm-thick BBO crystal and the central wavelength was measured to be 405 nm with a bandwidth of $\sim$6 nm. In the experiments, the pump pulse (405 nm) energy was attenuated to be less than 2 μJ and the optimal probe pulse (800 nm) energy was controlled to be near 35 μJ. The liquid furan sample, seeded in helium at a background pressure of 2 atm, was expanded to generate a pulsed molecular beam. The beam was skimmed and intersected perpendicularly with the linear polarized pump and probe laser beams. The generated photoelectrons were extracted and accelerated by the electrostatic immersion lens and then projected onto a detector. Each image was accumulated over 40000 laser shots. Three-dimensional distribution reconstructions were performed by the basis-set expansion (BASEX) forward convolution method.[35] As shown in Fig. 1(a), a typical time-of-flight (TOF) mass spectrum of furan, recorded with the two-photon 405 nm excitation and two-photon 800 nm ionization at zero delay time, yields only one peak at 9.1 μs, corresponding to the parent ion C$_{4}$H$_{4}$O$^{+}$ with the $m/e$ ratio of 68/+. It is reliable that the contributions to the total photoelectron signal from the fragment ions such as C$_{4}$H$_{4}^{+}$ or O$^{ +}$ can be completely ignored. The ion yields provide a measure of the lifetime of the excited states. The time-dependent ion signal of C$_{4}$H$_{4}$O$^{+}$ is presented in Fig. 1(b). The decay profile is found to be well reproduced only by a single exponential function convoluted with a Gaussian that describes the instrument response function (IRF). A lifetime of 29 fs is obtained and its fitting error is reasonably within $\pm$2 fs. The unsatisfactory fittings with two or even three exponential functions might be due to the extremely short lifetimes of $S_{2}$ and $S_{1}$, which are largely restricted by our IRF of 165 fs. In consideration of Fuji's pump-probe scheme,[30] the lifetime of 29 fs obtained in our experiment is ascribed to the average lifetime of the $S_{2}$ and $S_{1}$ states.
cpl-34-4-043101-fig1.png
Fig. 1. (a) Typical TOF mass spectrum of furan recorded with the two-photon excitation at 405 nm and two-photon ionization at 800 nm at the zero delay time. (b) Time-resolved total ion signals of C$_{4}$H$_{4}$O$^{+}$ as a function of the delay time between the pump and probe pulses.
In Fig. 2, we show the time-dependent photoelectron kinetic energy distributions (PKEDs) extracted from the corresponding images measured as a function of the pump-probe time delay. The adiabatic ionization potential (AIP) of furan is 8.88 eV,[13] therefore the available energy ($D_{0}=h\nu_{\rm pump}+h\nu_{\rm probe}-{\rm AIP})$ in the continuum state can be determined to be 0.34 eV for the two-photon 800 nm ionization to the zero vibrational level of the cationic ground state. Two featured peaks centered at 0.11 and 0.24 eV are extracted from the continuous energy distribution (0.05–0.32 eV), and are marked by two vertical dashed-dotted lines labeled by $E_{1}$ and $E_{2}$, respectively. The $E_{2}$ band shows the largest intensity at $\Delta t=5$ fs, and then sharply decreases. At the delay time of 40 fs, the whole 'bump' of the $E_{2}$ band disappears. Meanwhile, the intense $E_{1}$ band gradually goes down with a much slower rate as compared with the $E_{2}$ band. Based on the previous spectral data,[5-22] these two peaks centered at 0.11 and 0.24 eV are assigned as the central peak (i.e., $S_{1}$ band and $S_{2}$ band) resulting from the ionization of the $S_{1}$ and $S_{2}$ states to the cationic ground state by two photons of 800 nm, respectively.
cpl-34-4-043101-fig2.png
Fig. 2. Time-resolved photoelectron kinetic energy distributions extracted from the corresponding images. A BASEX-inverted image at the delay time of 5 fs is shown in the inset.
The energies and geometries of $S_{1}$, $S_{2}$ and the cationic ground state of the furan molecule indicate that these states have roughly similar geometries as that of the $S_{0}$ state within the Franck–Condon area.[5-22] Thus the conservation of vibrational energy to the ions upon ionization is favored. If the $S_{2}$ state rapidly internally converts to the lower-lying $S_{1}$ state, the electronic energy will transform into vibrational energy and thus the vibrational energy acquired for the vibrationally hot $S_{1}$ state could be calculated to be 0.21 eV (0.21 eV$=h\nu _{\rm pump}(2\times3.06)-E_{\rm origin~of~S1} (5.91)$). The observed PKE of 0.11 eV is smaller by 0.23 eV than the available energy of 0.34 eV, indicating that the cationic ground state carries the vibrational energy of 0.23 eV when the vibrationally hot $S_{1}$ state is ionized. The same situation holds for the $S_{2}$ state, in which the observed vibrational energy (0.10 eV) is in agreement with the calculated vibrational energy of 0.08 eV (0.08 eV$=h\nu_{\rm pump}(2\times3.06)-E_{\rm origin~of~S2}(6.04)$). For a further analysis of the PKEDs associated with the correlated relaxation dynamics of the $S_{1}$ and $S_{2}$ states, we extract the spectral components by the Levenberg–Marquardt method.[36] The measured PKED at each delay time is fitted by the sum of two Voigt functions and a polynomial. The Voigt function profile (i.e., a convolution of Gaussian and Lorentzian functions) is preferentially selected to reproduce the $S_{1}$ and $S_{2}$ component spectra by assuming the component peak centers to be fixed at 0.11 and 0.24 eV, respectively. In addition, a polynomial is unavoidably added to match the residual background. Although the background and the real signal also overlap each other to a small extent in the low energy region, the characteristics of the photoelectron signal of $S_{2}$ and $S_{1}$ are superimposed on the background and not seriously dampened. As an example, Fig. 3(a) shows the fitting of PKED at the delay time of 5 fs, and the fitting residue is given in Fig. 3(b). As seen in Fig. 3(a), the $S_{1}$ component carries more intensity than that of the $S_{2}$ component although they overlap each other to some extent.
cpl-34-4-043101-fig3.png
Fig. 3. (a) The nonlinear least squares fitting of the photoelectron kinetic energy distribution at the delay time of 5 fs by the Levenberg–Marquardt method.[36] (b) The residue for the fitting in (a).
cpl-34-4-043101-fig4.png
Fig. 4. (a) Time-dependent intensity for the $S_{2}$ component. The solid circles are fitted by a single exponential decay function. (b) Time-dependent intensity for the $S_{1}$ component. The solid diamonds are fitted by a sum of an exponential decay function and an $S$-shaped growth function.
The time-dependent intensity of the $S_{2}$ component is shown in Fig. 4(a) by integrating the area of the $S_{2}$ component at different delay times. The intensity of the $S_{2}$ component exhibits rapid decay due to the IC to the $S_{1}$ state and is well fitted by a single exponential decay function. The time constant for the IC is obtained to be 15.4$\pm$1.5 fs, which is remarkably in agreement with the IC time constant of 10 fs simulated by Fuji et al.[30] and 15 fs obtained by Liu et al.[32] Likewise, the time-dependent intensity of the $S_{1}$ component is shown in Fig. 4(b). The time profile is well explained by two components, i.e., a rapid decay and an accompanying growth. The growth time is set to be on the same timescale as in the case of Fig. 4(a), in support of the decay of the $S_{2}$ state and the corresponding growth of the $S_{1}$ state by IC from the $S_{2}$ state. Thus the growth time and the decay time are roughly estimated to be $\sim$15 fs and $\sim$38 fs, respectively.
cpl-34-4-043101-fig5.png
Fig. 5. Schematic deactivation mechanism of the electronically excited furan.
Finally, the deactivation mechanism for the electronically excited furan is concluded on the basis of our results and previous work.[23-32] As shown in Fig. 5, following the two-photon excitation into the $S_{2}$ and $S_{1}$ states, the $S_{2}$ state rapidly internally converts to the lower-lying $S_{1}$ state. The timescale for the IC from $S_{2}$ to $S_{1}$ is 15.4 fs. Meanwhile, the internal conversions from the directly-prepared $S_{1}$ and secondarily-populated $S_{1}$ to $S_{0}$ occur within 38 fs. Such ultrafast internal conversions could be realized via the conical intersections of the potential energy surfaces. The conical intersection between the $S_{2}$ and $S_{1}$ states was calculated to be just 0.06 eV above the origin of the $S_{2}$ state,[23] which is very close to our excitation energy. The deactivation of the $S_{1}$ state might initially continue on the $S_{1}$ potential energy surface and finally mainly internally converts to the lower-lying $S_{0}$ state through the crossing of $S_{1}/S_{0}$. Owing to a shortage of tracking the whole $S_{1}$ deactivation path via the two-photon ionization at 800 nm, we could not exclude the possibility of furan ring-opening or formation of the $\alpha $- and $\beta $-carbenes during the deactivation of the vibrationally hot $S_{0}$ state. However, these deactivation channels were found to be with very low probability.[30] It is also noted that the triplet states differ in energy as the prepared singlet states in furan. There are no molecular features[37,38] which would drive an ultrafast intersystem crossing in the observed time window, neither by an El Sayed mechanism nor by a heavy-atom effect. In summary, we have used femtosecond TR-PEI to observe the full relaxation pathways of excited furan. The time-dependent intensities of the two components in the photoelectron kinetic energy spectra exhibit rapid decay, corresponding to the decay of the $S_{2}$ state by IC to the $S_{1}$ state and the decay of the $S_{1}$ state by IC to the hot $S_{0}$ state. The ultrafast IC from $S_{2}$ to $S_{1}$ takes place within an extremely short time of 15.4 fs. At the same time, the deactivation of $S_{1}$ to hot $S_{0}$ occurs on an estimated time scale of 38 fs. References FEMTOSECOND TIME-RESOLVED PHOTOELECTRON IMAGINGF EMTOSECOND T IME -R ESOLVED P HOTOELECTRON S PECTROSCOPY OF P OLYATOMIC M OLECULESAtmospheric degradation of 3-methylfuran: kinetic and products studyProduction of dimethylfuran for liquid fuels from biomass-derived carbohydratesA Vibrational Analysis of the Absorption Spectrum of Furan in the Schumann RegionThe Vacuum Ultraviolet Absorption Spectra of Cyclic Compounds. II. Tetrahydrofuran, Tetrahydropyran, 1,4-Dioxane and Furan 1Absorption and Photoionization Coefficients of Furan VaporTriplet states of furan, thiophene, and pyrroleElectron impact investigation of electronic excitations in furan, thiophene, and pyrroleTransmission of 0–15 eV monoenergetic electrons through thin-film molecular solidsResonantly enhanced multiphoton ionization of pyrrole, N ‐methyl pyrrole, and furanNear UV spectra of furan and its derivativesThe electronic states of furan studied by VUV absorption, near-threshold electron energy-loss spectroscopy and ab initio multi-reference configuration interaction calculationsConfiguration interaction study of the electronic spectrum of furanCluster expansion of the wave function. Valence and Rydberg excitations and ionizations of pyrrole, furan, and cyclopentadieneTheoretical study of the electronic spectra of cyclopentadiene, pyrrole, and furanTheoretical study of the excitation spectra of five‐membered ring compounds: Cyclopentadiene, furan, and pyrrolePolarization propagator study of electronic excitation in key heterocyclic molecules II. FuranIntegral-direct coupled cluster calculations of frequency-dependent polarizabilities, transition probabilities and excited-state propertiesThe Electronic Spectrum of FuranElectronic excitation spectra of furan and pyrrole: Revisited by the symmetry adapted cluster–configuration interaction methodStudy of excited states of furan and pyrrole by time-dependent density functional theoryTheoretical study of the low-lying excited singlet states of furanTheoretical study of excitations in furan: Spectra and molecular dynamicsMulti-mode–multi-state quantum dynamics of key five-membered heterocycles: spectroscopy and ultrafast internal conversionDeactivation via ring opening: A quantum chemical study of the excited states of furan and comparison to thiopheneTheoretical study of photoinduced ring-opening in furanTime-resolved photoelectron imaging spectra from non-adiabatic molecular dynamics simulationsReaction surface approach to multimode vibronic coupling problems: General framework and application to furanUltrafast photodynamics of furanExcited-state dynamics of furan studied by sub-20-fs time-resolved photoelectron imaging using 159-nm pulsesTracking ultrafast relaxation dynamics of furan by femtosecond photoelectron imagingProbing ultrafast internal conversion of o-xylene via femtosecond time-resolved photoelectron imagingReconstruction of Abel-transformable images: The Gaussian basis-set expansion Abel transform methodFityk : a general-purpose peak fitting programPhotochemistry of aryl halides: Photodissociation dynamicsVisualizing competing intersystem crossing and internal conversion with a complementary measurement
[1] Suzuki T 2006 Annu. Rev. Phys. Chem. 57 555
[2] Stolow A 2003 Annu. Rev. Phys. Chem. 54 89
[3] Tapia A, Villanueva F, Salgado M S et al 2011 Atmos. Chem. Phys. 11 3227
[4] Román-Leshkov Y, Barrett C J, Liu Y et al 2007 Nature 447 982
[5] Picket L W 1940 J. Chem. Phys. 8 293
[6] Pickett L W, Hoeflich N J, Liu T et al 1951 J. Am. Chem. Soc. 73 4865
[7] Watanabe K and Nakayama T 1958 J. Chem. Phys. 29 48
[8] Flicker W M, Mosher O A and Kuppermann A 1976 Chem. Phys. Lett. 38 489
[9] Flicker W M, Mosher O A and Kuppermann A 1976 J. Chem. Phys. 64 1315
[10] Sanche L 1979 J. Chem. Phys. 71 4860
[11] Cooper C D, Williamson A D, Miller J C et al 1980 J. Chem. Phys. 73 1527
[12] Nyulaszi L 1992 J. Mol. Struct. 273 133
[13] Palmer M H, Walker I C, Ballard C C et al 1995 Chem. Phys. 192 111
[14] Thunemann K, Buenker R J and Butscher W 1980 Chem. Phys. 47 313
[15] Nakatsuji H, Kitao O and Yonezawa T 1985 J. Chem. Phys. 83 723
[16] Serrano-Andres L, Merchan M and Nebot-Gil I 1993 J. Am. Chem. Soc. 115 6184
[17] Nakano H, Tsuneda T and Hashimoto T 1996 J. Chem. Phys. 104 2312
[18] Trofimov A B and Schirmer J 1997 Chem. Phys. 224 175
[19] Christiansen O, Halkier A, Koch H et al 1998 J. Chem. Phys. 108 2801
[20] Christiansen O and Jørgensen P 1998 J. Am. Chem. Soc. 120 3423
[21] Wan J, Meller J, Hada M et al 2000 J. Chem. Phys. 113 7853
[22] Burcl R, Amos R D and Handy N C 2002 Chem. Phys. Lett. 355 8
[23] Gromov E V, Trofimov A B, Vitkovskaya N M et al 2003 J. Chem. Phys. 119 737
[24] Gromov E V, Trofimov A B, Vitkovskaya N M et al 2004 J. Chem. Phys. 121 4585
[25] Köppel H, Gromov E V and Trofimov A B 2004 Chem. Phys. 304 35
[26] Gavrilov N, Salzmann S and Marian C M 2008 Chem. Phys. 349 269
[27] Gromov E V, Trofimov A B, Gatti F et al 2010 J. Chem. Phys. 133 164309
[28] Humeniuk A, Wohlgemuth M, Suzuki T et al 2013 J. Chem. Phys. 139 134104
[29] Gromov E V, Reddy V S, Gatti F et al 2013 J. Chem. Phys. 139 234306
[30] Fuji T, Suzuki Y, Horio T et al 2010 J. Chem. Phys. 133 234303
[31] Spesyvtsev R, Horio T, Suzuki Y et al 2015 J. Chem. Phys. 143 014302
[32] Liu Y, Knopp G, Qin C et al 2015 Chem. Phys. 446 142
[33]Liu Y, Qin C and Zhang S 2011 Acta Phys. -Chim. Sin. 27 965
[34] Liu Y, Tang B, Shen H et al 2010 Opt. Express 18 5791
[35] Dribinski V, Ossadtchi A, Mandelshtam V A et al 2002 Rev. Sci. Instrum. 73 2634
[36] Wojdyr M 2010 J. Appl. Crystallogr. 43 1126
[37] Han K L and He G Z 2007 J. Photoch. Photobio. C 8 55
[38] Liu Y, Gerber T, Qin C et al 2016 J. Chem. Phys. 144 084201