Subscriber access provided by University of Glasgow Library
A: Spectroscopy, Molecular Structure, and Quantum Chemistry
Simulation of Impulsive Vibrational Spectroscopy Federico J. Hernandez, Franco P. Bonafé, Bálint Aradi, Thomas Frauenheim, and Cristian G. Sanchez J. Phys. Chem. A, Just Accepted Manuscript • DOI: 10.1021/acs.jpca.9b00307 • Publication Date (Web): 15 Feb 2019 Downloaded from http://pubs.acs.org on February 15, 2019
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
The Journal of Physical Chemistry
Simulation of Impulsive Vibrational Spectroscopy Federico J. Hern´andez∗ ,†,‡ Franco P. Bonaf´e,†,‡ B´alint Aradi,¶ Thomas Frauenheim,¶ and Cristi´an G. S´anchez∗,†,‡ †Universidad Nacional de C´ordoba. Facultad de Ciencias Qu´ımicas, Departamento de Qu´ımica Te´orica y Computacional. C´ordoba, Argentina ‡Instituto de Investigaciones en Fisicoqu´ımica de C´ordoba, INFIQC (CONICET Universidad Nacional de C´ordoba). C´ordoba, Argentina ¶Bremen Center for Computational Materials Science, Universit¨at Bremen. Bremen, Germany E-mail:
[email protected] 1
ACS Paragon Plus Environment
The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Abstract In the present work we have applied a fully atomistic electron-nuclear real-time propagation protocol to compute the impulsive vibrational spectroscopy of the five DNA/RNA nucleobases in order to study the very first steps (sub-picosecond) of their energy distribution after UV excitation. We have observed that after the pump pulse absorption, the system is prepared in a coherent superposition of the ground and the pumped electronic excited states in the equilibrium geometry of the ground state. Furthermore, for relatively low fluency values of the pump pulse, the dominant contribution to the electronic wavefunction of the coherent state is of the ground state and the mean potential energy surface within the Ehrenfest approximation is similar to that of the ground state. As a consequence, the molecular displacements are better correlated with ground state normal modes. On the other hand, when the pump fluency is increased the excited state contribution to the electronic wavefunction becomes more important and the mean potential energy surface resembles more that of the excited state, producing a better correlation between the molecular displacements and the excited state normal modes. Finally, it has been observed that the impulsive activation of several vibrational modes upon the electronic excitation is triggered by the development of excited state forces which accelerate the nuclei from their equilibrium positions causing a distribution of the absorbed electronic energy on the nuclear degrees of freedom and could be closely related to the driving force of the ultrafast non-radiative deactivation observed in these systems.
1
Introduction
Since the advent of picosecond and then femtosecond light sources, a vast new research field has emerged both for photophysics and for photochemistry allowing the study of the dynamics of molecular systems on ultrashort time scales. In this sense, the development of time-resolved pump-probe spectroscopies has increased tremendously during the last three decades, and the study of the non-linear response of complex systems upon interaction with 2
ACS Paragon Plus Environment
Page 2 of 25
Page 3 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
The Journal of Physical Chemistry
coherent laser light pulses has become a frontier research topic in molecular quantum physics. Particular interest has been focused in the investigation of coherence in molecular processes occurring in the condensed phase, leading to a new kind of spectroscopy, namely vibrational coherence spectroscopy (VCS). 1–7 The principle of VCS is based on the interaction of the molecular system with an ultrashort coherent laser pulse (pump pulse) which is shorter than the period of the molecular nuclear motions and has a spectral width larger than the corresponding vibrational levels spacing. Such laser pulse may produce a vibrational wavepacket (i.e. the preparation of the molecular system in a coherent superposition of vibrational levels) in almost any molecule in a impulsive manner. Moreover, if this ultra-short pulse is resonant with any ground-state electronic absorption band, for example the S0 − S1 , it will produce a coherent superposition of electronic and vibrational quantum states. Hence, a non-stationary population will be produced upon light absorption and the generated wavepacket will be in a coherent superposition between the S0 and S1 states until decoherence occurs. Due to the fact that the quantum superposition of the vibrational states results in the classical oscillation of the vibrational degrees of freedom and this oscillation changes the molecular structure, the absorbance signals of the pumped-molecule oscillates as well. Then, a second laser pulse (probe pulse) may interact with the system at different time delays recording the spectrum of the transient (pump-driven) species. Finally, the Fourier transform of the oscillatory absorption signals reveals the Raman activity of the system providing similar information to that available in the frequency domain. Thus, this sequential pump-probe transient absorption spectroscopy using ultrashort laser pulses, which generates vibrational wavepackets both in the ground and excited states, is known as impulsive vibrational spectroscopy (IVS). The main advantage of IVS over frequency domain Raman spectroscopy is that the full vibrationally coherent evolution of the system is measured; however, it has the drawback that vibrational coherences in both ground and excited states might be present, so disentangling them may constitute a challenge. Nonetheless, some strategies have been developed to
3
ACS Paragon Plus Environment
The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
overcome this issue. 8,9 Following the spectroscopic signature of the oscillatory nuclear motions impulsively imprinted with techniques such as IVS has allowed tracking molecular structural dynamics accompanying ultrafast photoinduced processes in molecular systems. Some systems broadly studied using these techniques are DNA and RNA nucleobases (hereafter nucleobases). In this sense, the photo-protection of nucleobases is still a hot topic in photochemistry and photophysics and has marveled scientists for many years. From the large number of articles that have been published, its importance becomes clear. 10–20 The key point that stands out in the literature is that nucleobases survive through an efficient excited-state deactivation mechanism. This prevents potentially dangerous photochemical reactions via ultrafast nonradiative decay to the electronic ground state. The ultrafast decay constitutes a possible explanation of why nucleobases were selected through evolution as building blocks of life, proposing photostability as the principal criterion for that selection. 20 High level ultrafast time-resolved spectroscopies allowed direct real-time studies on nucleobases, providing insight into their fast photodeactivation and it has been well established that it occurs between 0.5 and 1.5 ps after electronic excitation. 12,19,21,22 Furthermore, recent broadband transient absorption (TA) experiments using ultrashort (sub-10 fs) pulses have concluded that a common mechanism operates on electronic deactivation. 19 These experiments have also provided IVS measurements on the modulation of the transient electronic spectrum due to coupling to different vibrational modes. 19,23,24 For all nucleobases, the impulsive vibrational spectra agree relatively well with UV resonance Raman spectra. 25–29 From the rapid electronic dynamics and impulsive activation of some vibrational modes after photoexcitation, it is suggested that the initial excited state dynamics is crucial in determining nucleobase photophysics. Meanwhile, in the last years great progress has been made in the field of computational time-resolved spectroscopy, by the simulation of transient absorption (TA) spectra using atomistic techniques. 30–34 In this regard, very recently we have reported a protocol capable of computing the real-time TA spectra and the impulsive vibrational (IV) spectra from
4
ACS Paragon Plus Environment
Page 4 of 25
Page 5 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
The Journal of Physical Chemistry
electron-nuclear dynamics in gas phase after an initial perturbation (pump) to the electronic density matrix. This protocol was successfully used to study the ultrafast dynamics of Soretexcited zinc(II)-tetraphenylporphyrin in the sub-picosecond time scale observing quantum beats in the TA caused by impulsively excited molecular vibrations. 35 In the present work we apply the same protocol to compute the IVS in vacuo of the five nucleobases: adenine, guanine, cytosine, thymine and uracil, in order to study the energy distribution after their respective UV excitation.
2
Theoretical Methods
The methodology employed has been outlined in a previous work. 35 Briefly, a real-time fully atomistic quantum dynamics is performed treating the nuclei within the so called Ehrenfest approximation and electrons under the semi-empirical density functional tight-binding (DFTB) formalism as implemented in the DFTB+ package. 36 The DFTB approach is based on a second order expansion of the Kohn-Sham energy functional around a reference density 37 and has been utilized to study the laser-driven dynamics of several molecular, 38–41 metallic 42,43 and semiconductor systems. 44,45 To compute the TA spectra of a system, first a pulse-shaped time-dependent perturbation (pump) is added to the self-consistent Hamiltonian of the system. Numerical integration of the Liouville-von Neumann equation for Ehrenfest dynamics within DFTB is performed through real-time propagation of the one-electron reduced density matrix. 43 The absorption spectrum is computed at successive time delays by running a separate calculation applying a Dirac-delta perturbation (probe) to the instantaneous density matrix. Afterwards, the frequency-dependent polarizability α(ω, τ ) (where τ stands for the delay time or time difference between pump and probe) is calculated within the linear response regime, by Fourier transform of the dipole moment difference as illustrated in Eq. 1. There, µP +pτ is the dipole moment of the system (ˆ µ = −eˆ r) in the time-domain, calculated after the probe at delay time
5
ACS Paragon Plus Environment
The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 6 of 25
τ , and µP is the dipole moment of the system with the pump perturbation only. Finally, the absorption coefficient σ(ω, τ ) can be calculated from the polarizability as shown in Eq. 2. An averaging over four different phases of pump field must be performed in order to suppress the coherence produced by the interference of the pump and probe perturbations. 31
α(ω, τ ) = ∫ dt(µP +pτ (t) − µP (t + τ )) exp(iωt)
σ(ω, τ ) =
4πω Im(α(ω, τ )) c
(1)
(2)
The usual reference for TA spectra that is subtracted from the transient absorption signal is the ground state spectrum of the system. The spectra presented hereafter were calculated by subtracting the absorption spectrum computed immediately after the pump pulse (Eq. 3). This is done to avoid spurious effects in the TA due to the absorption peak shifting during the pump pulse action. This peak-shifting is a reported problem of DFT-based methodologies, attributable to the adiabatic approximation of currently used DFT functionals, and is more noticeable for smaller molecular systems. 46 Nonetheless, the maximal shift we observe is for the case of adenine and is of 0.017 eV, which does not cause a significant effect on post-pulse dynamics. ∆A(t) = A(t) − A(tp )
(3)
To compute the IV spectrum a fast Fourier transform (FFT) of the TA spectra along the delay time axis is performed obtaining a spectral density as in the the Fig. 2c, which relates the absorption energy of the transient species and the energy of the different vibrational motions coupled to the ground-state absorption band that has been pumped. Finally, the spectral density is integrated within the pumped absorption peak region, yielding the IV spectra presented in Fig. 3a-e. It is worth mentioning that a Hanning window function must be applied in FFT in order to avoid spurious frequencies. In order to compute the IVS for these systems, each nucleobase was excited with a sin2 6
ACS Paragon Plus Environment
Page 7 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
The Journal of Physical Chemistry
pulse of pulse time tp = 10 fs tuned to the corresponding lowest energy band and a peak field intensity of 0.05 V ˚ A−1 , yielding an energy fluence of 3.1 mJ cm−2 . The combined nuclear and electronic dynamics then evolves for 725 fs. During the dynamics, 3000 snapshots of the system’s density matrix and geometry are stored, enabling a time resolution of 0.24 fs for the calculation of the corresponding TA spectra. In order to compute each IV spectrum presented below, Dirac-delta probe calculations at 3000 delay times have been performed for every spatial direction (x, y or z ) and for each phase of the pump pulse, giving a total amount of 36000 trajectories. This scheme has significant computational cost and is only feasible (within a reasonable time and modest computational resources) with a semi-empirical and relatively accurate method as TD-DFTB.
3
Results and Discussion
The ground state absorption spectra were calculated for all nucleobases by real-time propagation after a Dirac-delta like perturbation and are presented in the Fig. 1. The lowest-energy active band for each nucleobase corresponds to a π → π ∗ transition. The discrepancies between the spectral transitions calculated at TD-DFTB level and the experimental values are relatively low and to be expected given the approximate nature of the method. The experimental π → π ∗ lowest-excitation wavelength measured for adenine, cytosine, guanine, thymine and uracil are 276.96, 314.12, 294.90, 277.78 and 272.48 nm, respectively, 47 whereas from our calculations such wavelengths are 262.37, 302.16, 270.90, 262.18 and 242.47 nm. These bands were excited with a pump pulse as it was explained in section 2 in order to compute the TA spectra and then the IVS of each nucleobase. Fig. 2b exhibits the TA spectra for adenine as an illustrative case. In all cases calculated TA spectra show bleaching of the illuminated band in agreement with available experimental data. 19 The absence of excited state absorption in the results presented here can be explained from the low population in the excited state (as discussed below). The most important
7
ACS Paragon Plus Environment
The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Figure 1: Calculated ground state absorption spectra for the five DNA/RNA nucleobases. Each colored triangle marks the transition excited by the pump pulse.
8
ACS Paragon Plus Environment
Page 8 of 25
Page 9 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
The Journal of Physical Chemistry
feature in the TA spectra, in the context of what is presented in this work, are oscillatory features with changes in sidebands but not in the peak center. In order to study these oscillatory features, the TA spectra were Fourier transformed along the delay time axis – obtaining the spectral density shown in Fig. 2c. Such spectral density shows several peaks in the area between 200 and 2000 cm−1 . These peaks constitute the fingerprint of vibrational modes that are impulsively activated by the electronic excitation. These modes are responsible for the observed TA spectral modulation in time. It is worth mentioning that this impulsive activation of selected vibrational modes has been observed experimentally for thymine and uracil using sub-10-fs deep UV laser pulses. 23,24
Figure 2: Ground state spectrum of adenine (a); TA spectra calculated by subtracting the absorption spectrum at each time with the spectrum at the time of the pulse (b); Fourier transform of the TA spectra along the delay time in the region between 230 and 320 nm (c). Dashed lines indicate the electronic excitation (pump) wavelength. To get further insight on the modulation of the TA spectrum, the IV spectrum was calculated for each nucleobase by integrating the spectral density as it was mentioned in section 2 (between 250 and 275 nm for case of adenine). These IV spectra are presented in Fig. 3 in comparison to the potential energy distribution (PED) on the nucleobase ground state normal modes obtained from the calculated nuclear dynamics. In order to get the PED, the nuclear displacements are projected on normal mode coordinates as shown in Eq. 4, 48 where vAi are the eigenvector matrix elements, ∆rA (t) are the nuclear displacements, mA is 9
ACS Paragon Plus Environment
The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 10 of 25
the atomic mass and the coefficients Qi are the coordinates in the normal mode basis. The potential energy in each mode i can be calculated as shown in Eq. 5.
Figure 3: Impulsive vibrational spectra (black) and potential energy distribution per mode (red) of adenine (a), guanine (b), thymine (c), uracil (d) and cytosine (e).
Qi (t) = ∑ mA ∆rA (t) ⋅ vAi
(4)
A
Vi (t) = (2πc¯ νi Qi (t))2 10
ACS Paragon Plus Environment
(5)
Page 11 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
The Journal of Physical Chemistry
From Figure 3 good agreements between the IV spectra and the ground state normal mode PEDs are observed, indicating that the modulation of the electronic spectra caused by any mode impulsively activated depends on the fraction of the electronic energy that this mode receives after excitation. To confirm such correspondence a correlation analysis between the two quantities was performed for the five nucleobases observing a linear trend. The linear regression coefficients obtained are reported in Table 1. It is worth mentioning that for the IV spectrum of thymine, the peaks centered at 230 and 370 cm−1 were not taken into account in the correlation analysis since they arise from electronic coherences (see section S1 of the Supporting Information). A remarkable result of this analysis is that the good correlation between PED and the IV spectra is obtained only when the trajectory is projected on the ground state normal modes (GS-PED). This was checked by performing a similar IV-PED correlation analysis but considering normal modes of the pumped excited state geometries (ES-PED), obtained from time-dependent DFTB calculations within linear response theory, as implemented in the DFTB+ package. 49 No correlation between the IV spectra and ES-PED could be found. These results can be understood by modeling the evolution of the system at times right after the pump pulse has acted by taking into account only the ground and pumped excited states. It is important to bare in mind that at this point in time, the system will be in a coherent superposition of ground and excited electronic states in the equilibrium geometry of the ground state. For the intensities described before, the dominant contribution to the electronic wave-function will be that of the ground state (as an example, for adenine, the population of the LUMO is of under one percent). In an adiabatic basis the force operator is diagonal and the electronic contribution to the Ehrenfest force will be:
ˆ ρ] = ⟨g∣F∣g⟩ρ ˆ ˆ Fel = Tr[Fˆ gg + ⟨e∣F∣e⟩ρee
(6)
ˆ ˆ Where ⟨g∣F∣g⟩ and ⟨e∣F∣e⟩ are the ground and excited state force expectation values and
11
ACS Paragon Plus Environment
The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ρgg and ρee the respective populations. Given that, as mentioned, the system is initially at the ground state equilibrium geometry, only the second term contributes. Therefore, the nuclei will be accelerated from their equilibrium positions by the excited state forces, but move in a mean potential energy surface similar to the ground state one, owing to the small fraction of electronic population pumped to the excited state. This simplified picture, which is summarized in Fig. 4, explains the better correlation found by projecting onto the ground state normal modes. Thus, if the fluency of the pump pulse were increased, the fraction of the electronic population pumped to the excited state would also rise and the mean potential energy surface would resemble more to the excited state one, improving the correlation between IVS and the ES-PED. This effect was calculated for the adenine case with a three different pump laser fluencies (see Table 2). At the lower pump fluency (3.1 mJ cm−2 ), the population of the LUMO is around 1 %, the correlation between IVS and GS-PED is 0.86 and there is no correlation using the ES-PED (corresponding to a low-pump regime as shown in Fig. 4, upper panel). As the fluency of the pump pulse is increased, the population of the LUMO increases, the correlation between IVS and GS-PED decreases while the correlation between IVS and ES-PED improves reaching a R value of 0.5 using a pump fluency of 112 mJ cm−2 (as intuitively results from the PES in a high-pump regime in Fig. 4, lower panel), confirming what was stated above. This is an interesting result because, to the best of our knowledge, the IVS dependence on the fluency of the pump laser had not been studied. Furthermore, the modification of pump fluency could be used to improve the ratio between the signal of excited state and ground state normal modes in the IV spectrum. Nonetheless, the values presented in Table 2 must be read from a qualitative point of view, because for higher pump fluencies the mean potential approximation from the Ehrenfest approach is only an approximate description, since the mean potential energy surface stops resembling the ground state one and becomes less realistic due to the neglect of electron-ion correlation that the approximation implies. 50 At this point, the question of why certain vibrational modes are impulsively activated
12
ACS Paragon Plus Environment
Page 12 of 25
Page 13 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
The Journal of Physical Chemistry
Figure 4: Schematic representation of the Ehrenfest mean potential energy surface (PES) approach. Top: using low pump laser fluencies, the mean PES is very similar to the ground state one. Bottom: increasing the laser fluency the contribution of the excited state to the mean PES becomes larger.
13
ACS Paragon Plus Environment
The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 14 of 25
emerges naturally. To answer this question, for each nucleobase the force matrix in the corresponding excited state was calculated in its Franck-Condon geometry and projected in a similar manner as presented in Eq. 4 onto the ground state normal modes basis. A good correlation was observed between the PED and the projected forces for all cases (see Table 1). These PED-Forces correlations indicate that the initial forces, as given by equation 6, are responsible for the activation of certain normal modes, which modulate the electronic spectra during the whole dynamics. Table 1: Linear Regression Coefficients (R) of the Correlation Between the Impulsive Vibrational Spectrum (IVS) Intensity and the Potential Energy Distribution (PED) of Each Ground State Normal Mode, and Between PED and the Excited State Nuclear Forces Projected onto the Ground State Normal Modes Base. Nucleobase Adenine Guanine Thymine Uracil Cytosine
IVS-PED corr. 0.86 0.76 0.97 0.92 0.91
PED-Forces corr. 0.96 0.97 0.96 0.97 0.89
Table 2: Linear Regression Coefficients (R) of the Correlation Between IVS Intensity and PED of Both Ground (GS) and Excited (ES) State Normal Modes for Adenine at Different pump Fluencies. In Addition, the LUMO Populations After the Pump Pulse Absorption for Each Pump Fluency are Showed Pump Fluency (mJ cm−2 ) 3.1 50 112
LUMO Pop. (%). ≈1 9.4 17.5
GS Corr. 0.86 0.8 0.36
ES Corr. 0.3 0.5
Finally, the displacements of the normal modes with highest intensity in the IV spectrum have been analyzed for all the nucleobases and are displayed in the Fig. 5. It might be seen that although many of the normal modes activated are in-plane modes, for the purine nucleobases there are some modes that involve out-of-plane displacements in the NH2 group. Moreover, the already known displacements responsible for conducting the system from the Franck-Condon region to the conical intersection zones are depicted in Fig. 6. In this 14
ACS Paragon Plus Environment
Page 15 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
The Journal of Physical Chemistry
sense, it has been proposed that C2 -H puckering, NH2 and C=O out-of-plane vibration or N9 -H stretching are the main geometry distortions to get purine nucleobases into the conical intersection zones, and C5 -C6 twisting and puckering, C=O out-of-plane vibration and ring opening for pyrimidine nucleobases. 12,13,20 A comparison between the impulsively activated modes depicted in the Fig. 5 with the molecular distortion zones shown in Fig. 6 may suggest that molecular distortions on such groups are generated immediately after the electronic excitation. Therefore, the impulsive activation of certain normal modes after the photoexcitation is likely to trigger the energy distribution that leads to deformation of the molecular geometry from the Franck-Condon zone into the distorted geometries which could evolve towards the conical intersections, causing the subsequent ultrafast internal conversion.
15
ACS Paragon Plus Environment
The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Figure 5: Atomic displacements for the normal modes with highest intensity in the IV spectrum activated upon excitation. The displacements are represented by red arrows (for inplane movements), whose direction marks where the atom is moving and its width and length are proportional to the modulus of the displacement, and colored circles (for out-of-plane movements), whose color intensity is proportional to the modulus of the displacement; red meaning an upward displacement from the sheet plane and blue a downward displacement.
Figure 6: Group of atoms previously identified as responsible for conducting the molecular system from its Franck-Condon Region to the conical intersection regions (circled in red)..
16
ACS Paragon Plus Environment
Page 16 of 25
Page 17 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
The Journal of Physical Chemistry
4
Conclusion
We have studied the impulsive vibrational spectroscopy of the five DNA/RNA nucleobases, calculated from their transient absorption spectra, applying a fully atomistic electron-nuclear real-time propagation protocol to compute the dynamics of each nucleobase upon their electronic excitation. We have analyzed the very first steps of energy redistribution onto the nucleobases vibrational modes upon UV excitation. Furthermore, we have observed that the photon absorption prepares the system in a coherent superposition of the ground and the pumped electronic excited states in the equilibrium geometry of the ground state. The dominant contribution to the electronic wavefunction of this coherent state is of the ground state and therefore, the mean potential energy surface within the Ehrenfest approximation resembles to that of the ground state. As a consequence, the molecular displacements responsible for modulating the electronic spectra are better correlated with ground state normal modes. Nonetheless, by increasing the pump intensity, the excited state contribution to the electronic wavefunction becomes more important, as the mean potential energy surface resembles more to that of the excited state and the correlation between the IVS and the potential energy distribution projected onto the excited state normal modes improves. In addition, we have analyzed the driving force of the impulsive activation of certain modes, observing the a sudden change of the electronic state acting on the ground state nuclear geometry produces a change of nuclear forces. Such forces are the launching mechanism of the geometry deformation and are proportional to the intensity on the activation on each normal mode. Finally, these excited state developed nuclear forces likely play an important role in the very first steps leading to the ultrafast non-radiative deactivation observed in these molecules.
17
ACS Paragon Plus Environment
The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Acknowledgement The authors wish to acknowledge CONICET and DFG-RTG2247 for the the funding and the BCCMS for the computational resources provided.
Supporting Information Available Supporting Information includes details about the IV spectrum of thymine simulated keeping the nuclei clampled in comparison with the IV spectrum simulated moving the nuclei under the Ehrenfest approach.
References (1) Collini, E.; Wong, C. Y.; Wilk, K. E.; Curmi, P. M. G.; Brumer, P.; Scholes, G. D. Coherently wired light-harvesting in photosynthetic marine algae at ambient temperature. Nature 2010, 463, 644–647. (2) Scholes, G. D.; Fleming, G. R.; Olaya-Castro, A.; van Grondelle, R. Lessons from nature about solar light harvesting. Nat. Chem. 2011, 3, 763–774. (3) Huelga, S.; Plenio, M. Vibrations, quanta and biology. Contemp. Phys. 2013, 54, 181– 207. (4) Fuller, F. D.; Pan, J.; Gelzinis, A.; Butkus, V.; Senlik, S. S.; Wilcox, D. E.; Yocum, C. F.; Valkunas, L.; Abramavicius, D.; Ogilvie, J. P. Vibronic coherence in oxygenic photosynthesis. Nat. Chem. 2014, 6, 706–711. (5) Romero, E.; Augulis, R.; Novoderezhkin, V. I.; Ferretti, M.; Thieme, J.; Zigmantas, D.; van Grondelle, R. Quantum coherence in photosynthesis for efficient solar-energy conversion. Nat. Phys. 2014, 10, 676–682.
18
ACS Paragon Plus Environment
Page 18 of 25
Page 19 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
The Journal of Physical Chemistry
(6) Scholes, G. D.; Fleming, G. R.; Chen, L. X.; Aspuru-Guzik, A.; Buchleitner, A.; Coker, D. F.; Engel, G. S.; van Grondelle, R.; Ishizaki, A.; Jonas, D. M. et al. Using coherence to enhance function in chemical and biophysical systems. Nature 2017, 543, 647–656. (7) Gueye, M.; Manathunga, M.; Agathangelou, D.; Orozco, Y.; Paolino, M.; Fusi, S.; Haacke, S.; Olivucci, M.; L´eonard, J. Engineering the vibrational coherence of vision into a synthetic molecular device. Nat. Commun. 2018, 9, 313. (8) Schnedermann, C.; Muders, V.; Ehrenberg, D.; Schlesinger, R.; Kukura, P.; Heberle, J. Vibronic Dynamics of the Ultrafast all-trans to 13-cis Photoisomerization of Retinal in Channelrhodopsin-1. J. Am. Chem. Soc. 2016, 138, 4757–4762. (9) Buckup, T.; L´eonard, J. Multidimensional Vibrational Coherence Spectroscopy. Top. Curr. Chem. 2018, 376, 35. (10) Crespo-Hern´andez, C. E.; Cohen, B.; Hare, P. M.; Kohler, B. Ultrafast Excited-State Dynamics in Nucleic Acids. Chem. Rev. 2004, 104, 1977–2020. (11) Middleton, C. T.; de La Harpe, K.; Su, C.; Law, Y. K.; Crespo-Hern´andez, C. E.; Kohler, B. DNA Excited-State Dynamics: From Single Bases to the Double Helix. Annu. Rev. Phys. Chem. 2009, 60, 217–239. (12) Barbatti, M.; Aquino, A. J. A.; Szymczak, J. J.; Nachtigallova, D.; Hobza, P.; Lischka, H. Relaxation mechanisms of UV-photoexcited DNA and RNA nucleobases. Proc. Natl. Acad. Sci. 2010, 107, 21453–21458. (13) Kleinermanns, K.; Nachtigallov´a, D.; de Vries, M. S. Excited state dynamics of DNA bases. Int. Rev. Phys. Chem. 2013, 32, 308–342. (14) Chen, J.; Kohler, B. Base stacking in adenosine dimers revealed by femtosecond transient absorption spectroscopy. J. Am. Chem. Soc. 2014, 136, 6362–6372. 19
ACS Paragon Plus Environment
The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
(15) Schreier, W. J.; Gilch, P.; Zinth, W. Early Events of DNA Photodamage. Annu. Rev. Phys. Chem. 2015, 66, 497–519. (16) Marchetti, B.; Karsili, T. N.; Ashfold, M. N.; Domcke, W. A ’bottom up’,: Ab initio computational approach to understanding fundamental photophysical processes in nitrogen containing heterocycles, DNA bases and base pairs. Phys. Chem. Chem. Phys. 2016, 18, 20007–20027. (17) Improta, R.; Santoro, F.; Blancafort, L. Quantum Mechanical Studies on the Photophysics and the Photochemistry of Nucleic Acids and Nucleobases. Chem. Rev. 2016, 116, 3540–3593. (18) Skowron, D. J.; Zhang, Y.; Beckstead, A. A.; Remington, J. M.; Strawn, M.; Kohler, B. Subnanosecond Emission Dynamics of AT DNA Oligonucleotides. ChemPhysChem 2016, 17, 3558–3569. (19) Prokhorenko, V. I.; Picchiotti, A.; Pola, M.; Dijkstra, A. G.; Miller, R. J. New Insights into the Photophysics of DNA Nucleobases. J. Phys. Chem. Lett. 2016, 7, 4445–4450. (20) Boldissar, S.; De Vries, M. S. How nature covers its bases. Phys. Chem. Chem. Phys. 2018, 20, 9701–9716. (21) Ullrich, S.; Schultz, T.; Zgierski, M. Z.; Stolow, A. Direct Observation of Electronic Relaxation Dynamics in Adenine via Time-Resolved Photoelectron Spectroscopy. J. Am. Chem. Soc. 2004, 126, 2262–2263. (22) Canuel, C.; Mons, M.; Piuzzi, F.; Tardivel, B.; Dimicoli, I.; Elhanine, M. Excited states dynamics of DNA and RNA bases: Characterization of a stepwise deactivation pathway in the gas phase. J. Chem. Phys. 2005, 122, 074316. (23) Kobayashi, T.; Kida, Y. Ultrafast spectroscopy with sub-10 fs deep-ultraviolet pulses. Phys. Chem. Chem. Phys. 2012, 14, 6200–6210. 20
ACS Paragon Plus Environment
Page 20 of 25
Page 21 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
The Journal of Physical Chemistry
(24) Xue, B.; Yabushita, A.; Kobayashi, T. Ultrafast dynamics of uracil and thymine studied using a sub-10 fs deep ultraviolet laser. Phys. Chem. Chem. Phys. 2016, 18, 17044– 17053. (25) Yarasi, S.; Brost, P.; Loppnow, G. R. Initial excited-state structural dynamics of thymine are coincident with the expected photochemical dynamics. J. Phys. Chem. A 2007, 111, 5130–5135. (26) Yarasi, S.; Ng, S.; Loppnow, G. R. Initial excited-state structural dynamics of uracil from resonance Raman spectroscopy are different from those of thymine (5methyluracil). J. Phys. Chem. B 2009, 113, 14336–14342. (27) El-Yazbi, A. F.; Palech, A.; Loppnow, G. R. Initial Excited-State Structural Dynamics of 2-Deoxyguanosine Determined via UV Resonance Raman Spectroscopy. J. Phys. Chem. A 2011, 115, 10445–10451. (28) Oladepo, S. A.; Loppnow, G. R. Initial Excited-State Structural Dynamics of 9Methyladenine from UV Resonance Raman Spectroscopy. J. Phys. Chem. B 2011, 115, 6149–6156. (29) Teimoory, F.; Loppnow, G. R. Initial excited-state structural dynamics of 6-substituted uracil derivatives: Femtosecond angle and bond lengthening dynamics in pyrimidine nucleobase photochemistry. J. Phys. Chem. A 2014, 118, 12161–12167. (30) De Giovannini, U.; Brunetto, G.; Castro, A.; Walkenhorst, J.; Rubio, A. Simulating pump-probe photoelectron and absorption spectroscopy on the attosecond timescale with time-dependent density functional theory. ChemPhysChem 2013, 14, 1363–1376. (31) Perfetto, E.; Stefanucci, G. Some exact properties of the nonequilibrium response function for transient photoabsorption. Phys. Rev. A. 2015, 91, 1–16.
21
ACS Paragon Plus Environment
The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
(32) Fischer, S. A.; Cramer, C. J.; Govind, N. Excited State Absorption from Real-Time Time-Dependent Density Functional Theory. J. Chem. Theory Comput. 2015, 11, 4294–4303. (33) Nguyen, T. S.; Koh, J. H.; Lefelhocz, S.; Parkhill, J. Black-Box, Real-Time Simulations of Transient Absorption Spectroscopy. J. Phys. Chem. Lett. 2016, 7, 1590–1595. (34) Fedunov, R. G.; Plotnikova, A. V.; Ivanov, A. I.; Vauthey, E. Simulations of the Ultrafast Transient Absorption Dynamics of a Donor–Acceptor Biaryl in Solution. J. Phys. Chem. A 2017, 121, 471–481. (35) Bonaf´e, F. P.; Hern´andez, F. J.; Aradi, B.; Frauenheim, T.; S´anchez, C. G. Fully Atomistic Real-Time Simulations of Transient Absorption Spectroscopy. J. Phys. Chem. Lett. 2018, 9, 4355–4359. (36) Aradi, B.; Hourahine, B.; Frauenheim, T. DFTB+, a sparse matrix-based implementation of the DFTB method. J. Phys. Chem. A 2007, 111, 5678–5684. (37) Elstner, M.; Porezag, D.; Jungnickel, G.; Elsner, J.; Haugk, M.; Frauenheim, T.; Suhai, S.; Seifert, G. Self-consistent-charge density-functional tight-binding method for simulations of complex materials properties. Phys. Rev. B 1998, 58, 7260–7268. (38) Oviedo, M. B.; Negre, C. F. A.; S´anchez, C. G. Dynamical simulation of the optical response of photosynthetic pigments. Phys. Chem. Chem. Phys. 2010, 12, 6706. (39) Medrano, C. R.; Oviedo, M. B.; S´anchez, C. G. Photoinduced charge-transfer dynamics simulations in noncovalently bonded molecular aggregates. Phys. Chem. Chem. Phys. 2016, 18, 14840–14849. (40) Mansilla Wettstein, C.; Bonaf´e, F. P.; Oviedo, M. B.; S´anchez, C. G. Optical properties of graphene nanoflakes: Shape matters. J. Chem. Phys. 2016, 144, 224305.
22
ACS Paragon Plus Environment
Page 22 of 25
Page 23 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
The Journal of Physical Chemistry
(41) Medrano, C. R.; S´anchez, C. G. Trap-Door-Like Irreversible Photoinduced Charge Transfer in a Donor–Acceptor Complex. J. Phys. Chem. Lett. 2018, 9, 3517–3524. (42) Douglas-Gallardo, O. A.; Berdakin, M.; S´anchez, C. G. Atomistic Insights into Chemical Interface Damping of Surface Plasmon Excitations in Silver Nanoclusters. J. Phys. Chem. C 2016, 120, 24389–24399. (43) Bonaf´e, F. P.; Aradi, B.; Guan, M.; Douglas-Gallardo, O. A.; Lian, C.; Meng, S.; Frauenheim, T.; S´anchez, C. G. Plasmon-driven sub-picosecond breathing of metal nanoparticles. Nanoscale 2017, 9, 12391–12397. (44) Fuertes, V. C.; Negre, C. F. A.; Oviedo, M. B.; Bonaf´e, F. P.; Oliva, F. Y.; S´anchez, C. G. A theoretical study of the optical properties of nanostructured TiO2. J. Phys. Condens. Matter 2013, 25, 115304. (45) Mansilla Wettstein, C.; S´anchez, C. G. Characterization of ZnO as substrate for DSSC. Phys. Chem. Chem. Phys. 2018, 20, 21910–21916. (46) Provorse, M. R.; Habenicht, B. F.; Isborn, C. M. Peak-Shifting in Real-Time TimeDependent Density Functional Theory. J. Chem. Theory Comput. 2015, 11, 4791–4802. (47) de Vries, M. S. UV-Excitation from an Experimental Perspective: Frequency Resolved. Top. Curr. Chem. 2014, 355, 33–56. (48) Horiuchi, T.; G¯o, N. Projection of monte carlo and molecular dynamics trajectories onto the normal mode axes: Human lysozyme. Proteins Struct. Funct. Bioinforma. 1991, 10, 106–116. (49) Niehaus, T. A.; Suhai, S.; Sala, F. D.; Lugli, P.; Elstner, M.; Seifert, G.; Frauenheim, T. Tight-binding approach to time-dependent density-functional response theory. Phys. Rev. B 2001, 63, 085108.
23
ACS Paragon Plus Environment
The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
(50) Horsfield, A. P.; Bowler, D. R.; Fisher, A. J.; Todorov, T. N.; S´anchez, C. G. Correlated electron–ion dynamics: the excitation of atomic motion by energetic electrons. Journal of Physics: Condensed Matter 2005, 17, 4793–4812.
24
ACS Paragon Plus Environment
Page 24 of 25
Page 25 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
The Journal of Physical Chemistry
Graphical TOC Entry
25
ACS Paragon Plus Environment