Transient IR Spectroscopic Observation of Singlet ... - ACS Publications

Subscriber access provided by UNIV OF NEBRASKA - LINCOLN

Article

Transient IR Spectroscopic Observation of Singlet and Triplet States of 2-Nitrofluorene: Revisiting the Photophysics of Nitroaromatics Martin Alex Bjørn Larsen, Jan Thøgersen, Anne Boutrup Stephansen, Jorge Peón, Theis I. Sølling, and Søren Rud Keiding J. Phys. Chem. A, Just Accepted Manuscript • DOI: 10.1021/acs.jpca.5b09125 • Publication Date (Web): 15 Dec 2015 Downloaded from http://pubs.acs.org on December 15, 2015

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 free 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 accessible to all readers and 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.

The Journal of Physical Chemistry A 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 28

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

Transient IR Spectroscopic Observation of Singlet and Triplet States of 2-Nitrofluorene: Revisiting the Photophysics of Nitroaromatics Martin A.B. Larsen†, Jan Thøgersen‡, Anne B. Stephansen†, Jorge Peon§, Theis I. Sølling† and Søren R. Keiding‡,* †

Department of Chemistry, University of Copenhagen, Universitetsparken 5, DK-2100

Copenhagen Ø, Denmark ‡

Department of Chemistry, Aarhus University, Langelandsgade 140, DK-8000 Aarhus C,

Denmark §

Instituto de Química, Universidad Nacional Autónoma de México, Circuito Exterior, Ciudad

Universitaria, México, 04510, D.F., México AUTHOR INFORMATION Corresponding Author *E-mail: [email protected]

ACS Paragon Plus Environment

1


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 2 of 28

ABSTRACT The dynamics of 2-nitrofluorene (2-NF) in deuterated acetonitrile is studied using UV pump, IR probe femtosecond transient absorption spectroscopy. Upon excitation to the vibrationally excited S1 state, the excited state population of 2-NF branches into two different relaxation pathways. One route leads to intersystem crossing (ISC) to the triplet manifold within a few hundred femtoseconds and the other to internal conversion (IC) to the ground state. The experiments indicate that after relaxation to the energetic minimum on S1, 2-NF undergoes internal conversion to the ground state in about 15 ps. IC within the triplet manifold is also observed as the initially populated triplet state relaxes to T1 in about 6 ps. Rotational anisotropy measurements corroborate the assignment of the transient IR frequencies and indicate a rotational diffusion time of 2-NF in the solvent of about 14 ps. The combined set of results provides a unified picture of the dynamics in photoexcited 2-NF. This to our knowledge the first example using femtosecond vibrational spectroscopy for the study of the fundamental photoinduced

processes

in

nitroaromatic

compounds.


2

Page 3 of 28

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


INTRODUCTION Nitrated polycyclic aromatic hydrocarbons (NPAHs) have been studied extensively as they generally are considered atmospheric pollutants1-2 and pose certain health concerns. Many NPAHs have been shown to be carcinogenic and mutagenic3-7 and 2-nitrofluorene is such a substance which has been investigated for its carcinogenicity3 and mutagenic properties.5-8 Not only 2-NF, but also its possible photodegradation products (for instance NO• and ArO•, among others)9-10 have been shown to be harmful species.11-12 However, the photodegradation of 2-NF is inefficient compared to e.g. 9-nitroanthracene,9-12 and the primary relaxation pathway of 2-NF does not appear to lead to photodegradation. Interestingly, the photophysics of many NPAHs involve a surprisingly fast intersystem crossing (ISC) pathway.13-22 This non-radiative process between electronic states of different multiplicity is formally forbidden in the non-relativistic description of quantum mechanics, and thus ISC is expected to happen much slower than for instance internal conversion (IC); the equivalent nonradiative decay mechanism between states of the same multiplicity. This has been highly challenged by the observation of ultrafast ISC in NPAHs,13-27 and also in systems such as benzene28-29 and enones30 among others, thus raising the question of the driving force for the ultrafast ISC and the involved mechanism. In previous work on NPAHs, twisting of the nitro group out of the aromatic plane has been pointed to as a vital structural parameter associated with the observed dynamics. The CrespoHernández group has for instance investigated 1-nitropyrene,13 1-nitronaphthalene,14,

16

2-

methyl-1-nitronaphthalene16 and 2-nitronaphthalene16 using UV/Vis transient absorption spectroscopy, and based on DFT calculations they suggest that the photodegradation and NO· loss of NPAHs may be related with twisting of the nitro group, i.e. the nitro group on the excited


3


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 4 of 28

state rotate out of the aromatic plane prior to photodegradation. In addition, it is suggested that 1nitropyrene reaches a structurally relaxed energetic minimum on S1 by twisting of the nitro group, a process happening on a 100 fs timescale, and that ISC to the triplet manifold happens subsequently on a 7 ps timescale.13 Peon and co-workers have included two qualitative rules to rationalize the ultrafast ISC in a range of NPAHs: A combination of a small energy gap between the first excited singlet state and an upper triplet receiver state may govern and facilitate the ultrafast transition.18, 22-23, 25-26 The triplet receiver state must have a different orbital angular momentum than the S1 state, thereby making the ISC partly allowed according to El-Sayed’s rule, which states that the transition becomes partly allowed if the spin-flip is compensated for by a simultaneous change in orbital angular momentum. Based on experiments and semi-empirical calculations on p-nitroaniline by Kovalenko et al.,21 Morales-Cueto et al.18 suggest that the fastest fluorescence decay component (50-820 fs) of the investigated NPAHs may be related to rotation of the nitro group. The triplet state dynamics of 2-NF have been investigated previously with UV/Vis absorption spectroscopy.19 The absorption stretches from 400-650 nm, and decays within 45±15 ps.19 Furthermore, 2-NF has been shown to be fluorescent with a very short fluorescence lifetime of less than 200 fs, which is taken to indicate that 2-NF undergoes ISC in less than 200 fs.26 Here, we seek to elucidate whether it is possible to observe the electronically excited states of nitroaromatics, in particular of 2-NF in the IR region and thereby gain information on the dynamics of the involved excited state species. To our knowledge these are the first timeresolved transient absorption IR results on any NPAH. Further, the aim is to utilize rotational


4

Page 5 of 28

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


anisotropy of the transient absorption measurements to corroborate assignment of the observed vibrations and to address the role of the nitro group. METHODS A. COMPUTATIONAL METHODS The interpretation of the experimental spectra are guided by (TD)-DFT calculations on the involved species. Calculations are performed using the Gaussian09 program package.31 The PBE0 DFT functional and Pople 6-311+G(d,p) basis set are chosen and solvent effects are included as well by self-consistent reaction field (SCRF) calculations with the polarizable continuum model (PCM) in the integral equation formalism (IEFPCM). All calculations are performed using the standard parameters for the PCM model of acetonitrile included in the Gaussian09 program package. Ground state geometries are optimized at the PBE0/IEFPCM/6311+G(d,p) level while excited state geometries are optimized at the TD-PBE0/IEFPCM/6311+G(d,p) level. These calculations give key structural information on the ground and excited states and in addition yield UV/Vis excitation dipole moments. The harmonic frequencies are calculated for the optimized structures to yield their IR spectra and also serve to verify that the optimized structures are true minima. Furthermore, the frequency calculations yield excitation dipole moments for the individual vibrational modes. The calculated frequencies are corrected by a scaling factor of 0.9594.32 B. EXPERIMENTAL SETUP The experimental setup has been described in detail previously.33 In short, a beam from a Ti:Sapphire laser system producing 100 fs laser pulses with a center wavelength of 800 nm and a


5


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 28

repetition rate of 1 kHz is split into two equally powerful beams. One of the 800 nm beams is frequency tripled in two consecutive BBO crystals to generate the 266 nm pump pulses used for the photoexcitation. The beam of pump pulses is modulated at 0.5 kHz by a mechanical chopper synchronized to the 1 kHz pulse repetition rate and sent through a scanning delay line and a λ/2 wave plate before it is focused through the sample by a concave mirror. The other part of the 800 nm beam pumps an optical parametric amplifier (OPA). The beam of IR probe pulses is generated by difference frequency mixing the signal and idler pulses from the OPA. The beam of probe pulses is divided into a signal and a reference beam. The signal beam is focused onto the sample by a concave mirror thus probing the sample inside the volume excited by the pump pulse, whereas the reference beam passes through the sample to the side of the pump beam. The signal and reference beams are subsequently sent through a grating spectrometer and measured by a 2 × 32 channels cooled dual-array HgCdTe detector referenced to the 0.5 kHz of the mechanical chopper. The isotropic absorption measurements are taken with the pump beam polarization at 54.7° (magic angle) relative to that of the probe beam. The measurements performed for the calculation of rotational anisotropies are taken with the pump beam polarization parallel and perpendicular to the probe beam polarization, which is achieved by adjusting the λ/2 wave plate before focusing the beam through the sample. The Teflon sample flow-cell is equipped with two CaF2 windows and a Teflon spacer allowing a path length of 1 mm. The sample solution is recycled several times during a measurement and lightly replenished during prolonged measurement periods. We use a 0.1 M solution of 2-NF in deuterated acetonitrile resulting in an optical density of the sample of 0.5 at the pump wavelength.34 The ground state vibrations of special interest are the symmetric and asymmetric –


6

Page 7 of 28

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


NO2 stretches which lie in the 1300-1600 cm-1 region. Deuterated acetonitrile conveniently leaves an open spectral “window” in the 1200-1800 cm-1 range35 (see Figure 1).

Figure 1. Steady-state IR spectrum of 2-NF in deuterated acetonitrile, showing that deuterated acetonitrile conveniently allows measurements on the absorptions of 2-NF in the 1200-1800 cm-1 range. The solvent absorptions are indicated in the figure by stars (*). RESULTS In the following we first present our computed excited state structures and frequencies. The optimized structures are essential to the interpretation of the photoexcited molecular dynamics and the calculated ground state frequencies are supplemental to the steady-state IR spectrum of 2-NF. Next, the experimental magic angle spectra and the observed dynamics are presented. The excited state absorptions as well as ground state bleaches are identified based on the calculated frequencies. The state specific dynamics are extracted from the experimental data. Finally, the rotational anisotropy data are presented, corroborating the assignment based on the calculated frequencies. The loss of anisotropy associated with specific vibrations may be followed by these measurements and a rotational diffusion time of the entire molecule in solution is presented. COMPUTATIONAL RESULTS


7


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 8 of 28

The identification of excited states relies on knowing in which spectral area the states present themselves. This knowledge is achieved by calculating the excited state vibrational frequencies and in order to do so, the ground and excited state minimum geometries are first identified. The calculated geometries are shown in Figure 2. All geometries are planar and importantly have the nitro group in the same plane as the aromatic system. Other significant structural differences are observed from ground state (Figure 2 a) to excited states (Figure 2 b and 2 c) which may serve to indicate observable differences in the transient IR spectrum: The main structural characteristics are given in Figure 2 and show that the main structural changes are associated with the nitro group and the C4a-C4b bond connecting the two aromatic rings. The S1 state (Figure 2 b) has distinct n,pi* character and the observed shortening of the C-N and C4a-C4b bonds as well as the elongation of the N-O bond can be rationalized in terms of the electron density moving from the nitro group to the aromatic ring. This causes the C-N bond to gain double bond character while the N-O bonds lose their double bond character.

a) S0

b) S1

c) T1


8

Page 9 of 28

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 2. S0 (a), S1 (b) and T1 (c) minimum energy structures shown with main structural parameters. The eight frequencies that have the highest oscillator strengths and are within the experimental range are listed in Table 1 for each of the optimized states (see Supporting Information (SI) Table 1 for all IR frequencies). The vibrations encompass –NO2 stretching, aromatic stretching, and CH wagging motions and indicate that observation of the excited state may be expected based on spectral differences. Table 1. Comparison of selected vibrational frequencies

S0 (cm-1)

S0 relative intensities

S1 (cm-1)

1338

100

1216

1368

5

1427

S1 relative intensities

T1 (cm-1)

T1 relative intensities

4

1251

45

1324

7

1268

100

2

1368

5

1311

11

1451

3

1403

100

1327

17

1463

2

1427

16

1349

27

1535

35

1439

40

1506

7

1587

9

1454

19

1535

24

1606

8

1479

68

1601

64

The calculated ground state frequencies are listed in Table 1. These account for all observed absorption lines in the FT-IR spectrum (see Discussion) and are in excellent agreement with previous studies on the ground state IR spectrum of 2-NF.36-37 The largest oscillator strengths are associated with the symmetric and asymmetric –NO2 stretches calculated to appear at 1338 cm-1 and 1535 cm-1, which is also consistent with the FT-IR spectrum and previous investigations.36-37


9


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 28

The absolute values of intensity indicate that the S0 and T1 have similar oscillator strengths whereas the intensities of the absorptions pertaining to the S1 state are much weaker. The excited states of 2-NF are not only investigated by the dynamics of their absorptions in the IR range but also by the rotational anisotropies associated with the excited state vibrations. Such measurements do not only give information on the temporal dynamics of the anisotropy loss, but also the relative orientation between the molecular UV transition dipole moment and the vibrational transition dipole moment. The anisotropy at zero time delay, R , contains this information and has its maximum values when the relative orientations of the transition dipole moments are either parallel or perpendicular. In addition to the frequencies, the calculations also yield transition dipole moments for the vibrations. These may be used in combination with the molecular UV transition dipole moment to predict the initial rotational anisotropy values (R ), calculated according to Eq. 1:

Eq. 1 R β 3 cos β 1, where is the angle between the molecular UV transition dipole moment vector and the dipole moment vector for the vibration.

Figure 3. The angle between the molecular UV transition dipole moment and the transition dipole moments for the symmetric (a) and asymmetric (b) –NO2 stretch shown in 2-NF.


10

Page 11 of 28

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 calculated transition dipole moments for the symmetric (1338 cm-1) and asymmetric –NO2 stretch (1535 cm-1) are almost perpendicular (see Figure 3) and yield greatly differing initial anisotropies as a consequence of their very different angles with respect to the molecular UV excitation dipole moment. Such differing values of R (see Table 2) allow for corroboration of the assignment and serve to distinguish vibrational states. Summing up, the calculations show that the optimized minimum structures have significant structural differences which present themselves as distinguishable spectral differences in the IR spectrum. Furthermore, the calculated transition dipole moments and corresponding R values indicate that the ground state symmetric and asymmetric –NO2 stretching vibrations will show markedly different behavior in the rotational anisotropy measurements. MAGIC ANGLE SPECTRA

Transient absorption measurements may be performed under magic angle conditions by having the polarization of probe at 54.7° relative to the pump beam polarization. This eliminates the contribution from rotational diffusion of the molecules in solution. In addition, it is also possible to reconstruct the magic angle spectrum from the parallel and perpendicular components used in the rotational anisotropy experiments and Figure 4 shows the reconstructed absorption spectra with central wavenumbers of 1330 cm-1 and 1540 cm-1 respectively. The magic angle absorption spectra are constructed using Eq. 2. Eq. 2: ∆ ∆∥ + 2∆! ), where ∆∥ and ∆! are the parallel and perpendicular components, respectively. A number of excited state induced absorptions are visible in addition to the ground state bleaches, which will be elaborated in the following.


11


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 12 of 28

Figure 4. Spectra showing both ground-state bleaches and induced excited state absorption with central wavenumbers of 1330 cm-1 (a) and 1540 cm-1 (b). The ground state bleach associated with the symmetric –NO2 stretch is observed at 1336 cm-1 and is evident in Figure 4a as the large central bleach, but also weaker ground state vibrations at 1396 cm-1, 1422 cm-1, 1450 cm-1 and on the edge of the spectrum at 1473 cm-1 are apparent as bleaches. The strongest bleach at 1336 cm-1 appears within the temporal resolution of the setup and is used to designate time zero. A partial repopulation of the ground state is clearly visible at the symmetric stretch as the bleach signal reduces to about half of its maximum value. Upon closer inspection of the weaker signals this partial repopulation of the ground state is also evident at the other ground state frequencies. A broad and high intensity excited state absorption is visible at about 1260 cm-1. The dynamics of this induced absorption is decidedly different from the dynamics of the ground state frequencies and the absorption quickly rises to a plateau. The frequencies covered by this broad induced absorption coincide with two of the calculated triplet state absorptions.


12

Page 13 of 28

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


Fitting the transient at the symmetric –NO2 ground state stretch (1336 cm-1) with a monoexponential decay yields a relaxation time of 18±4 ps. Fitting the weaker ground state bleaches yields identical relaxation times although the uncertainty is somewhat larger (±10 ps) due to a lower signal to noise ratio. The induced absorption is fitted at two wavelengths (1258 cm-1 and 1280 cm-1) obtained from triplet frequency calculations and both yield rise times of 5±2 ps – see SI Table 2 for a complete overview of fitted transients and SI Figure 1 for examples of ∆A vs. time plots. In Figure 4b the expected ground state bleaches at 1521 cm-1,36-37 1592 cm-1 36-37

36-37

and 1614 cm-1

are also present although the signals are superimposed with induced absorptions from the

excited species. These also agree with the calculated T1 triplet state absorptions. Hence, all main ground state frequencies within the measured frequency range are identified in the transient measurements. The observed signals are generally weaker and hence more difficult to discern from the noise in the region from 1400 cm-1 to 1500 cm-1, but two induced absorptions appear above the noise at 1444 cm-1 and 1484 cm-1 (see SI Figure 2). Although the signals are overlapping with the solvent background signals it is possible to fit transients to these peaks and the fitted decay of the induced absorptions agree with the ground state recovery times of about 18 ps (see Discussion). An overview of the fitted dynamics is presented in the Discussion (Table 4) alongside assignment of the observed species and interpretation of the dynamics as a whole. ROTATIONAL ANISOTROPY

The time-resolved rotational anisotropy measurements allow for an investigation of the orientation dynamics of the transition dipole moments. Thus, one can follow the rotational motion of the molecules as a whole and in addition, one may be able to follow the relative


13


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 28

motion of selected molecular constituents – i.e. the nitro group – if they are moving relative to the general molecular structure. The experimental spectra show that the strongest ground state bleaches are at the symmetric and asymmetric –NO2 stretch. The calculations presented earlier show that the dipole moments of these two vibrations are almost perpendicular. The calculated R0-values indicate that this perpendicularity between the two vibrations should result in distinct differences when the rotational anisotropy is measured. The transients for the perpendicular (blue) and parallel (red) components at 1336 cm-1 are shown in Figure 5.

Figure 5. Transient IR absorption of the perpendicular component (blue) and the parallel component (red) is showing a distinct initial difference (a) and the corresponding rotational anisotropy (b) showing mono-exponential decay with a decay time of 16 ps. The transients show that immediately after the decay of the time zero peak, the perpendicular component reaches a background (“zero”) level, then drops to its minimum value after 10 ps and returns to the background level again. This happens as the polarization of the probe at time zero is perpendicular to the ensemble of molecules excited by the pump pulse. The absorption becomes non-zero as the molecules in the ensemble rotate and after about 50 ps it reaches a


14

Page 15 of 28

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


common value with the parallel polarization as the anisotropy of the excited ensemble is lost. The parallel component, on the other hand, has a very large absorption value at short time delays as all excited molecules are selected parallel to the probe polarization. The absorption quickly decreases as the molecules rotate away from the alignment and the anisotropy is completely lost when the parallel and perpendicular components reach a common level of absorption. The time-dependent rotational anisotropy was analyzed using Eq. 3: Eq. 3: "

∆∥ #$∆% # ∆∥ #& ∆% #

" ',

Rotational anisotropies are calculated for the vibrations at 1336 cm-1, 1521 cm-1, 1258 cm-1 and 1280 cm-1. The rotational anisotropy dynamics are all well represented by a mono-exponential decay. The resulting timeconstants and R0-values are listed in Table 2. Table 2. Rotational anisotropy valuesa Wavenumber (cm-1)

Assignment

Rotational timeconstant (ps)

Experimental R0

1336

vsNO2 + vCN

16±3

0.29±0.03

0.398

1521

vasNO2

30±20

–0.10±0.05

–0.185

1258

vCC + vsNO2

11±4

0.4±0.1

0.385

1280

vCC + vsNO2

11±4

0.4±0.1

0.399

a

Theoretical R0

Note that the line positions are the fitted values from the transient data.

The rotational anisotropies at 1258 cm-1 and 1280cm-1 show a mono-exponential decay with a timescale of 11±4 ps. The rotational anisotropies at 1336 cm-1 and 1521 cm-1 decay on 16±3 ps and 30±20 ps, respectively. The timescales are the same within the experimental uncertainty. Especially the rotational timeconstant at 1521 cm-1 is associated with a large uncertainty due to a


15


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 16 of 28

poor signal to noise ratio. A comparison between the experimental and theoretical values are also listed in Table 2 and overall show good agreement although the values are not quite as close to the theoretical maxima of ~0.4 and ~ –0.2 as predicted by the calculations. It is not possible to fit from the exact time zero where the maximum values are expected due to the time zero peak. The different vibrations show a similar timescale for the loss of rotational anisotropy and are confidently identified on basis of their R0 values. The interpretation of the observed timescale is commented upon in the discussion. DISCUSSION The experimental and calculated ground state frequencies are first compared to see if the computational level of theory is adequate. Next, the correlation between calculated excited state absorptions and magic angle experiments is discussed in order to shed light on the state specific dynamics observed. The rotational anisotropy measurements are then included as support for the assignment based on the calculated frequencies. Further, the evidence pertaining to the role of the nitro group is commented on. Finally, a unified dynamical picture of photoexcited 2-NF is presented based on the combined measurements and calculations. Table 3. Comparison of calculated, FT-IR and observed transient S0 frequencies Calculated (cm-1)

FTIR (cm-1)

Transient experiments (cm-1)

1338

1342

1336

1368

1402

1396

1427

1425

1422

1451

1452

1450

1463

1473

1473

1535

1524

1521


16

Page 17 of 28

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


1587

1594

1592

1606

1616

1614

The calculated S0 IR spectrum (Table 3) is in good agreement with the FT-IR spectrum, previous work36-37 and the transient magic angle experiments. The experimental wavenumbers are essentially identical when considering that the spectral resolution of the time-resolved data is ~10 cm-1 while the resolution of the FT-IR measurement is ~2 cm-1. As mentioned in the results section, all the experimental ground state frequencies are reproduced by the calculations. As predicted by the calculated intensities, the symmetric and asymmetric –NO2 stretches give rise to the largest signals. The calculated S1 and T1 minima structures are significantly different from the S0 structure, and these differences are clearly reflected in the IR spectrum of the excited states, allowing a spectral separation and identification of the transient behavior of the excited states. atters. Furthermore, the computational results seem to indicate a lengthening of the N-O and C-N bonds. Since the initial dephasing of the wavepacket involves those degrees of freedom most of the vibrational energy will be deposited in N-O and C-N stretching motions, such localization of the energy is in line with the observation of ultrafast transition from pi,pi* to n,pi*. Comparison of the calculated T1 triplet state frequencies and the observed experimental excited absorption at 1260 cm-1 supports that the lowest excited triplet state is responsible for the induced absorption. The excited state absorption at 1260 cm-1 is actually two absorptions according to our calculations; one at 1251 cm-1 and one at 1268 cm-1. In the time-resolved data the absorptions are observed at 1258 cm-1 and 1280 cm-1 and the extracted dynamics at these frequencies are listed in Table 4 alongside the dynamics associated with the additional observed


17


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 18 of 28

induced T1 triplet state absorptions. The data indicate that the T1 triplet state is populated in about 6 ps. The absorptions at 1552 cm-1 and 1613 cm-1 are also attributed to the triplet based on a comparison with the calculated frequencies. These higher frequency absorptions are partly superimposed with ground state bleaches which makes it slightly more difficult to extract the timescale of the dynamics. However, the extracted temporal dynamics corroborates that the T1 state is populated in about 6 ps. Once the triplet state is populated there is no sign of relaxation to the ground state within our experimental timespan of 400 ps. This is in agreement with a previously measured phosphorescence lifetime of 2-NF of 130 ms38 and the natural lifetime of similar nitroaromatic compounds being on the order of ns to µs.13-14 The rise time of the first excited triplet state is also in accord with previous transient absorptions measurements on NPAHs showing ultrafast ISC to an excited Tn receiver state and subsequent IC within the triplet manifold.14, 16, 20, 22, 24, 27 For 2-NF the Tn triplet formation (ISC) occurs within the time resolution of the experiment, which is approximately 0.2 ps.26 We note a spectral broadening of the absorption bands associated with all ground state frequencies at short time delays. In addition, there is a blue shift of the center frequency of the triplet absorptions at 1260 cm-1. Both observations indicate vibrational cooling as the excess energy from IC within the singlet or triplet manifold is transferred to the solvent. The blue shift of the triplet absorptions occurs concomitantly with the rise of the absorptions and it is not possible to temporally separate the vibrational cooling from the IC within the triplet manifold. The oscillator strengths indicate that the absorptions of the S1 state are much weaker than the S0 and triplet state absorptions, and thus any potential isosbestic points indicating transition to either of those states are therefore not expected. Two weak induced absorptions in the 1400-1500 cm-1 range may be attributed to the relaxed S1 state although they are partially superimposed on


18

Page 19 of 28

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


nearby ground state bleaches (see Table 4). Extracting the temporal dynamics from the experimental data indicate that the vibrationally relaxed S1 state is populated on a shorter timescale (about 1-3 ps) than that for population of the relaxed T1 state. In addition, the absorptions pertaining to the S1 state at 1444 cm-1 and 1484 cm-1 decay on a 15±10 ps timescale. A partial repopulation of the ground state is observed on a similar 15±5 ps timescale (Table 4) and this must be a non-radiative process, i.e. IC or ISC, as the fluorescence of 2-NF has been shown to decay on a 200 fs timescale with no detectable steady-state fluorescence.26 ISC may be excluded on the grounds of the observed rise time of T1 of about 6 ps and the triplet absorptions showing no sign of decay within the measured time delay. Thus, IC back to the ground state must be responsible for the observed signal and it appears most likely that the process goes via the S1 state to S0. The fitted dynamics associated with the S1 state are connected with large uncertainties but nonetheless favor the interpretation that IC from the S1 state is the cause for repopulation of S0. The reduction of the ground state absorption to about half its maximum value indicates a 50:50 branching ratio between IC within the singlet manifold and ISC to the triplet manifold. Only 50% of the ground state absorption recovers as the other 50% remains in the triplet manifold within our experiment. Further, there are no unaccounted absorptions showing a slow build-up over time to indicate photodegradation of 2-nitrofluorene to e.g. fluorenone or 2nitrofluorenone (both having characteristic absorptions above 1700 cm-1).39 Overall, the appearances of the triplet and singlet excited state absorptions at the predicted frequencies demonstrates state specific identification and the possibility of predicting experimental IR observations of excited state structures from relatively simple frequency calculations. Table 4. Extracted timeconstants and assignmentsb


19


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

Wavenumber

Assignment

(cm-1)

b

Page 20 of 28

Relaxation time

Rise time

Decay time

(ps)

(ps)

(ps)

1336

S0

18±4

-

-

1521

S0

11±3

-

-

1444

S1

-

3±2

26±17

1484

S1

-

3±2

12±8

1258

T1

-

5±2

>1ns

1280

T1

-

5±2

>1ns

1552

T1

-

8±4

>1ns

1613

T1

-

8±4

>1ns

The dynamics have been fitted at the indicated experimental wavenumbers.

The rotational anisotropy measurements validate the assignment of the associated states based on the frequency calculation. The symmetric and asymmetric stretches are unequivocally assigned based on their high intensity and distinct R0 differences. The anisotropy measurements also add some confidence to the assignment of the triplet absorptions around 1260 cm-1 although the two absorptions have not been spectrally resolved. The two triplet absorptions should have basically the same initial rotational anisotropy according to the calculated values and the experimental values support this notion. The final conclusion to be drawn from the rotational anisotropy data is that the initial anisotropy is primarily lost due to rotational diffusion of the entire molecule in solution. The method makes it possible to follow the decay of anisotropy associated with specific


20

Page 21 of 28

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


vibrational states but the timeconstants of 11±4 ps associated with the triplet state absorptions are not significantly different from the 16±3 ps observed for the ground state. Thus, there appears to be no separate rotation associated with the nitro group. Rather the anisotropy decays as the entire molecule moves by diffusion. The observed timescale of this rotational diffusion is consistent with previous measurements based on fluorescence up-conversion experiments.18,

24

Regarding the possibility of observing rotation of the nitro group if it indeed occurred, the changes of the angle between the molecular UV transition dipole moment and the considered vibrational transition dipole moments induced by rotation of the nitro group do not offer a clear cut qualitative argument that the rotation must be visible in the experimental data. A change from a parallel to a perpendicular configuration, as argued in the case of relaxation between the electronic states in 9-nitronaphthalene,18 cannot be assumed to occur. This does not the exclude the chance of observing a change in the rotational anisotropy measurements associated with rotation of the nitro group but it is not possible to reach a certain conclusion about the rotation based on the rotational anisotropy measurements. The interpretation of the results therefore relies on the computational investigation. The calculated geometries of ground and excited states all have a planar nitro group with respect to the aromatic plane, and there is no indication of any minima with a rotated nitro group. Optimizations starting from a rotated T1 or S1 structure results in planar configurations. The only possibility of obtaining a twisted structure is by optimization to a transition state and this results in very high energy structures compared to the planar ground state energy (0.28 eV for a twisted ground state and 3.28 eV for a twisted triplet state). The previously suggested involvement of nitro group rotation in excited state dynamics is indicated to take place on a 50-200 fs timescale for most compounds. Crespo-Hernandez and co-workers have implied that rotation has to be that


21


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 22 of 28

fast to for the rotated S1 minimum energy pathway to be competitive with the ultrafast ISC taking place on a

Transient IR Spectroscopic Observation of Singlet ... - ACS Publications

Recommend Documents