Electronic structures and population dynamics of excited states of xanthione and its derivatives

Article abstract of DOI:10.1016/j.chemphys.2017.07.007

A new compound, 1,3-dimethoxy xanthione (DXT), has been synthesized and its absorption (stationary and transient) and luminescence spectra have been measured in n-hexane and compared with xanthione (XT) spectra. The pronounced broadening of xanthione vibr

Full text of DOI:10.1016/j.chemphys.2017.07.007

Accepted Manuscript
Electronic structures and population dynamics of excited states of xanthione and
its derivatives
Roman G. Fedunov, Marina V. Rogozina, Svetlana S Khokhlova, Anatoly I.
Ivanov, Sergei A. Tikhomirov, Stanislav L. Bondarev, Tamara F. Raichenok,
Oleg V. Buganov, Vyacheslav K. Olkhovik, Dmitrii A. Vasilevskii
PII:
S0301-0104(17)30337-3
DOI:
http://dx.doi.org/10.1016/j.chemphys.2017.07.007
CHEMPH 9827
Reference:
Chemical Physics
To appear in:
Please cite this article as: R.G. Fedunov, M.V. Rogozina, S.S. Khokhlova, A.I. Ivanov, S.A. Tikhomirov, S.L.
Bondarev, T.F. Raichenok, O.V. Buganov, V.K. Olkhovik, D.A. Vasilevskii, Electronic structures and population
dynamics of excited states of xanthione and its derivatives, Chemical Physics (2017), doi: http://dx.doi.org/10.1016/
j.chemphys.2017.07.007
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

1
2
Electronic structures and population dynamics of
excited states of xanthione and its derivatives
3
4
5
6
Roman G. Fedunova, Marina V. Rogozinaa, Svetlana S Khokhlovaa,
Anatoly I. Ivanova,∗, Sergei A. Tikhomirovb, Stanislav L. Bondarevb,
Tamara F. Raichenokb, Oleg V. Buganovb, Vyacheslav K. Olkhovikc,
Dmitrii A. Vasilevskiic
7
8
aVolgograd State University, University Avenue 100, Volgograd 400062, Russia
bB.I. Stepanov Institute of Physics, National Academy of Sciences of Belarus, 68
Prospekt Nezavisimosti, Minsk BY-220072, Republic of Belarus
9
10
11
cInstitute of Chemistry of New Materials, National Academy of Sciences of Belarus, 36,
Fr. Skaryna Street, Minsk BY-220141, Republic of Belarus
12 Abstract
13
A new compound, 1,3-dimethoxy xanthione (DXT), has been synthesized
14 and its absorption (stationary and transient) and luminescence spectra have
15 been measured in n-hexane and compared with xanthione (XT) spectra. The
16 pronounced broadening of xanthione vibronic absorption band related to the
17 electronic transition to the second singlet excited state has been observed.
18 Distinctions between the spectra of xanthione and its methoxy derivatives
19 are discussed. Quantum chemical calculations of these compounds in the
20 ground and excited electronic states have been accomplished to clarify the
21 nature of electronic spectra changes due to modification of xanthione by
22 methoxy groups. Appearance of a new absorption band of DXT caused by
23 symmetry changes has been discussed. Calculations of the second excited
24 state structure of xanthione and its methoxy derivatives confirm noticeable
25 charge transfer (about 0.1 of the charge of an electron) from the methoxy
26 group to thiocarbonyl group. Fitting of the transient spectra of XT and DXT
27 has been fulfilled and the time constants of internal conversion S2 → S1
28 and intersystem crossing S1 → T1 have been determined. A considerable
29 difference between the time constants of internal conversion S2 → S1 in XT
30 and DXT is uncovered.
31 Keywords: higher excited states, molecular electronic devices, charge
32 transfer
∗
Preprint submitted to Chemical Physics
July 18, 2017
Corresponding author. Tel. +7-8442-464894. E-mail: Anatoly.Ivanov@volsu.ru

33 1. INTRODUCTION
Development of molecular scale electronic devices is one of the most
34
35 promising contemporary technologies in microelectronics [1, 2]. The organic
36 molecules can be used to construct such key electronic devices as spatial
37 light modulators, thin-film transistors, light-emitting diodes, sensors, solar
38 cells, optical switches and others [3]. Among these molecular size electronic
39 devices the optical switches are one of the most challenging. The nanoscale
40 electronic switches can open the new generation of tiny electronic devises far
41 beyond the modern silicon-based technology [4].
42
The optical switches are the particular devices capable to switch between
43 two stable and selectively chosen states with different physical properties
44 which can be flipped by the direct optical impulse. Molecular triad of type
45 acceptorL–donor–acceptorR (AL − D − AR) is one of the possible implemen-
46 tations of optical switches [5]. In such molecular triads there are two locally
47 excited electronic states of donor enabling the electron transfer to selectively
48 switch between “right” (AR) or “left” (AL) acceptor addressing to the wave-
49 length of excitation pulse. The molecular systems in which the photoreac-
50 tions involving some excited states may proceed are intensively investigated
51 during past decade [6, 7, 8, 9, 10, 11, 12, 13, 5, 14, 15, 16, 17, 18, 19, 20]. The
52 latest researches [5, 14, 15] were focused on molecular switches based on sys-
53 tems where zinc-porphyrin (ZnP) and imide play the role of electron donor
54 and electron acceptor, respectively. The principal challenge being interfered
55 the development of optical switches containing ZnP-imide systems is a prob-
56 lem of low ions quantum yield of photoinduced charge transfer (PCT) from
57 the second excited state S2 of ZnP to the covalently-bounded acceptor. The
58 results of computer simulations of PCT in such systems [14, 15, 18] show
59 that the low efficiency of charge separation is determined by the ultrafast
60 recombination to the first excited state of ZnP during the “hot” stage when
61 nuclear subsystems of reactants and solvent are far from equilibrium. All
62 known studies of PCT demonstrate low quantum yield of thermalized ion
63 products (about 20%) from the second excited state despite the fact that
64 the rate of electron transfer significantly exceeds the rate of non-radiative
65 relaxation [5]. Thus, the yield of ionic products is limited by the ultrafast
66 ion recombination to the first excited state S1. The conclusion is confirmed
67 by the experimental data showing that ultrafast fluorescence decay of second
68 excited state being accompanied by the competing fluorescence decay from
69 the first excited state.
2

70
Since the lifetime of the second excited state of the ZnP derivatives is not
71 strongly shorter than the characteristic time of internal conversion (approx-
72 imately 2 ps) the quantum yield of charge separation states is considerably
73 less than unity. The strong coupling between the local excited S2 and ionic
74 states is necessary for effective competition between charge transfer and inter-
75 nal conversion. That causes in the effectiveness of hot charge recombination
76 to the S1 state [14, 15]. The latest results attest that the high quantum yield
77 of thermalized ionic state can be only achieved in molecular systems with
78 lifetimes of S2 state significantly greater than 2 ps.
79
Molecular systems based on azulene or thioketone derivatives may be
80 used as an alternative to the systems containing ZnP-imide. According to
81 Refs. [8] research of mechanisms of luminescence quenching in azulene faces
82 even greater complexity. Despite the longer lifetime of S2 state (approx-
83 imately few nanoseconds) the definition of the states where the electron
84 passes to is complicated due to the interference of ion absorption spectra
85 with absorption of azulene from S2 [21]. The measurement of the kinetics
86 of S1 population is also problematic in consequence of its short lifetime (less
87 than 1 ps). It was ascertained that the internal conversion to the ground
88 state is the primary channel of the S1-state decay. Hence at the moment
89 the thioketones looks more promising as a compounds of the optical molec-
90 ular switches. Their S2 lifetimes can be as large as hundred picoseconds
91 [21] therefore the S2-state quenching, that is, photoinduced charge separa-
92 tion, can be much more effective than that in porphyrins. For example, the
93 lifetime of S1-state of xanthione (XT) molecule is short but it decays into
94 the triplet state with quantum yield close to 1. In literature there is some
95 information on the electron structure and optical spectra of XT [22, 23, 24]
96 whereas donor-acceptor dyads and triads containing XT are still not stud-
97 ied. A rather effective bimolecular electron transfer quenching of xanthione
98 S2 state by series of electron donor in acetonitrile solution was observed [8].
99
Thus, to develop effective high-speed optical molecular switches the next
100 steps would be essential: 1) theoretical study of molecular structures to se-
101 lect potentially perspective heterocyclic compounds with electron-donor or
102 electron-acceptor substituents; 2) synthesis and following experimental and
103 theoretical research of physical and chemical properties of the compounds
104 suitable for constructing molecular electronic devices.
105
The aims of this paper are: (i) measurements and analysis of station-
106 ary absorption and luminescence spectra of XT and its newly synthesized
107 methoxy derivatives, (ii) quantum chemical calculations of energetic struc-
3

108 ture and geometry of XT and a few of its methoxy derivatives, (iii) measure-
109 ments and analysis of transient spectra of XT and 1,3-dimethoxy xanthione
110 (DXT), (iv) fitting of the transient spectra of XT and DXT and calculation
111 of the second excited state lifetime, (v) determination of the time constants
112 of internal conversion S2 → S1 and intersystem crossing S1 → T1.
113 2. METHODOLOGY
114 2.1. Computational methods
115
The simulations were performed using MOPAC 2016 [25, 26, 27, 28]. A
116 semiempirical Parametric Method Number 3 (PM3) [25, 29] was used to
117 provide full geometry optimization and compute electronic structures of XT
118 molecule and its methoxy derivatives. The choice of PM3 is justified by the
119 fact that the method was parameterized for enthalpy and geometry charac-
120 teristics calculations of the molecules similar to benzene [27]. The systems
121 considered here belong to the same class. The calculations of the heat of for-
122 mation (∆f H◦) were carried out in the approximation of isolated molecule in
123 the gas phase at 298 K. COSMO [30] solvent model was employed to take it
124 implicitly into the account. Hexane with dielectric permittivity EPS = 1.89
125 was chosen as a solvent.
126
Approach based on the many-electron configuration interaction MECI
127 [31] was used to calculate the vertical energies of transitions to the ground
128 electronic state. To specify the number of occupied energy levels in MECI
129 calculation the MOPAC keyword value for configuration interaction calcula-
130 tions, C.I., is chosen to be: 4, 6, 8 and 9. The approach allows the excited
131 states energy calculation accounting for solvent within the COSMO solvent
132 model. Additionally it takes into account the quadratic influence of the
133 refractive index on the dielectric energy of the excited states. In the calcu-
134 lations we used following value of keyword accounting for squared refractive
135 index: N ∗ ∗ 2 = 1.89. The rest of the parameters was chosen to be default.
136 2.2. Experimental methods
137
Xanthion was synthesized from xanthone being acted by the Lawessons
138 reagent. The xanthone dimethoxy derivative (1,3-dihydroxy-xanthone) was
139 synthesized in the reaction of phloroglucinol with salic acid in presence of
140 fresh melted zinc chloride and phosphate chloride. Obtained 1,3-dihydroxy
141 xanthone acting with excess of dimethyl sulfate in presence of potassium
142 carbonate produces 1,3-dimethoxy xanthone. The latter smoothly turns into
4

143 DXT being acted by the Lawassons reagent. Registration of the electronic
144 absorption and luminescence spectra of compounds was carried out in 1 cm
145 quartz cuvettes by spectrophotometer Cary-500 (USA) and spectrofluorime-
146 ter (Solar, Belarus). n-hexane (Aldrich, USA) was used as a solvent.
In the literature there is large scattering of the values of molar extinc-
tion coefficient [32]. This is the reason why this quantity was measured in
this work. Molar extinction coefficients were determined by weighing three
samples for each thioketone and subsequent dissolving in certain volume of n-
hexane. The electronic spectra were recorded for obtained solution of XT and
DXT using spectrometer Cary-500 (USA). The oscillator strengths shown in
Table 3 of the manuscript were calculated using the following equation [33]:
Z
9n
(n2 + 2)2
f = 1.3 · 10−8
ε(ν)dν
147 where n is the refractive index of the solvent (for n-hexane n = 1.375), ε(ν)
148 is the molar extinction coefficient, and ν is the wave number in cm−1.
149
Pump-probe measurement setup was exploited as described elsewhere [34,
150 35]. The laser pulses of 140 fs duration of the second harmonic of Al2O3 : Ti3+
151 laser (395 nm) with the energy in a range 10 – 30 µJ was used us a pump.
152 A 2 mm thick cell filled with a solvent at a dye concentration of 10−4
M
153 was utilized as a sample. The probe was a white continuum ranging from
154 400 nm to 1100 nm generated in water by a fraction of the fundamental
155 laser beam at 790 nm. The spectra of the signal probe and reference pulses
156 were recorded by a polychromator equipped with silicon CCD matrix and
157 transmitted to a computer. The differential absorbance signal ∆D(λ, ∆t)
e e
e
e
158 is obtained as ∆D(λ, ∆t) = lg(T0/T), where T = Eprob/Eref and T0 =
159 Ep0rob/Er0ef are the energy ratios of the probe and reference pulses passed
160 through the sample with and without preceding pump, respectively. The
161 pump and the probe beam diameters of 1 mm and 0.7 mm, respectively,
162 with nearly Gauss intensity distribution have been controlled by a CCD
163 camera. To show the stability of samples the stationary absorption spectra
164 were measured before and after pump-probe measurements. No changes
165 were detected. The transient spectra were corrected for the group velocity
166 dispersion of the probe pulse. The group velocity dispersion function has
167 been measured, a software was elaborated to automatically account for this
168 correction.
169
For stationary measurements the concentrations of molecular oxygen,
170 XT and DXT dissolved in n-hexane are CO2 = 1.56 · 10−2 mol/L, CXT
=
5

171 CDXT = 2 · 10−5 mol/L, correspondingly. For non-stationary measurements
172 the concentrations of XT and DXT are CXT = CDXT = 10−4 mol/L. The
173 excitation wavelength is λexc = 340 nm.
174 3. RESULTS AND DISCUSSION
175 3.1. Energetic and electronic properties of xanthione derivatives
Table 1: Physical characteristics of methoxy derivatives of XT in the ground state. Here
∆f H◦ (in kcal/mol), Etot, IP, EHOMO, ELUMO (in eV), µ (in Db) are the formation
heat, total energy, ionization potential, energy of the highest occupied molecular orbital,
energy of the lowest unoccupied molecular orbital, dipole moment, respectively.
No Comp. R1
R2
H
H
H
H3CO
A1
A2
∆f H◦ (∆f Hg◦as
52.60 (54.19)
13.16(15.22)
13.05(15.12)
)
Etot
µ
IP EHOMO ELUMO
1
2
3
4
5
6
7
8
9
I
II
II
III
III
IV
IV
IV
IV
H
119.8 119.7
114.0 119.7
117.9 119.7
-2140.964 4.11 9.00 -9.009
-2583.905 3.66 8.95 -8.955
-2583.910 5.54 8.94 -8.943
-1.973
-1.914
-1.908
-1.714
-1.713
-1.946
-1.922
-1.946
-1.916
H3CO(cis)
H3CO(trans)
H3CO(cis)
H3CO(trans) H3CO
H3CO(cis)
H3CO(cis)
113.9 122.8 -17.20(-14.52) -3026.452 4.45 8.80 -8.805
125.1 122.9 -17.46(-14.85) -3026.464 5.05 8.80 -8.806
125.1 113.6 -24.78(-22.43) -3026.781 2.87 8.92 -8.927
H3CO(cis)
H3CO(trans) 125.2 125.1 -23.97(-21.47) -3026.746 4.34 8.91 -8.919
H3CO(trans) H3CO(cis)
H3CO(trans) H3CO(trans) 125.2 125.1 -24.09(-21.57) -3026.751 6.47 8.90 -8.906
125.3 113.7 -24.92(-22.56) -3026.781 2.87 8.92 -8.927
Table 2: The atomic charges of the XT skeleton. The charges are presented in the units
of the elementary charge. The numbers in the first column correspond to the numeration
of conformers given in Table 1.
No C1
C2
C3
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
O
S
1
2
3
4
5
6
7
8
9
-0.1 -0.04 -0.03 0.07 -0.11 -0.05 0.11 0.07 -0.03 -0.04 -0.10 -0.05 -0.11 -0.08 -0.42
-0.14 -0.03 -0.10 0.10 -0.19 0.14 0.10 0.06 -0.01 -0.002 -0.10 -0.04 -0.11 -0.07 -0.40
-0.14 -0.06 -0.07 0.11 -0.17 0.10 0.10 0.10 -0.02 -0.02 -0.09 -0.04 -0.12 -0.07 -0.41
-0.22 0.17 -0.09 0.10 -0.22 0.15 0.10 0.04 -0.02 -0.03 -0.10 -0.06 -0.10 -0.08 -0.33
-0.18 0.15 -0.08 0.10 -0.25 0.15 0.10 0.04 -0.02 -0.03 -0.10 -0.06 -0.10 -0.09 -0.32
-0.17 -0.06 -0.07 0.11 -0.13 0.10 0.10 0.07 -0.01 -0.09 0.08 -0.05 -0.09 -0.07 -0.41
-0.17 -0.06 -0.08 0.11 -0.13 0.10 0.10 0.07 -0.01 -0.05 0.08 -0.09 -0.09 -0.07 -0.41
-0.14 -0.06 -0.06 0.11 -0.16 0.09 0.10 0.07 -0.01 -0.10 0.08 -0.05 -0.09 -0.07 -0.40
-0.14 -0.06 -0.06 0.11 -0.16 0.09 0.10 0.07 -0.01 -0.06 0.08 -0.09 -0.08 -0.07 -0.40
176
The results of energetic and electronic properties calculations of following
177 molecules: (I) 9H-xanthene-9-thione, (II) 3-methoxy-9H-9-thione, (III) 1,3-
178 dimethoxy-9H-xanthene-9-thione, (IV) 2,6-dimethoxy-9H-xanthene-9-thione
179 and their possible conformers are listed in Table 1 and 2. Geometry struc-
180 tures of the molecules I, II and IV are shown in Fig. 1. The geometry
6

181 characteristics of plain XT skeleton are almost stay the same with attach-
182 ment of methoxy groups in cis- or trans-configurations. In Fig. 1 the pa-
183 rameters are shown only for compound I (R1 = R2 = H). For compounds II
184 (R1 = H, R2 = OCH3) and IV (R1 = OCH3, R2 = OCH3) the bond lengths
◦
˚
185 and angles differ from that listed in Fig. 1 for values ±0.02 A and ±2 , re-
186 spectively. Figure 2 demonstrates the geometry structure of compound III
187 (R1 = OCH3, R2 = OCH3) characterized by considerable deformation of the
188 XT skeleton. There is a noticeable deviation from plain structure. The latter
189 leads a bond angle C2 − C7 − C10 to vary from value 172◦ in the compounds
190 I, II and IV up to 155◦ in the compound III. It should be also noted a sig-
191 nificant deviation of sulfur atom from the XT skeleton plane that leads to
192 variation of the bond angle O − C7 − S from 180◦ in the compounds I, II, IV
193 to 159◦ in compound III. Such a deformation of the XT plane may result in
194 appearance of the enantiomer properties for compound III.
195
Conformers of xanthione methoxy derivatives are defined by different ar-
196 rangement of the methyl fragments of methoxy groups relatively the XT
197 skeleton. Configuration of C1 − C6 − O − CH3 may be in cis- or trans-
198 positions. The angles A1 and A2 are shown in Figures 1 and 2 since they
199 vary more than 2◦ for cis- and trans-configurations. Their values are listed
200 in Table 1.
201
Values of formation heat, ∆f H◦, calculated for the compounds I–IV are
202 in good agreement with experimental values in liquid, ∆f Hl◦iq, at standard
203 conditions for benzene and its methoxy derivatives. According to NIST
204 database[36], the quantity of ∆f Hl◦iq for benzene is positive and equals 49
205 kJ/mol[37]. Attachment of a methoxy group to benzene results in consider-
206 able decreasing the quantity ∆f Hl◦iq and for anisole it equals −120 kJ/mol[38].
207 Attachment of the second methoxy group to anisole further reduces the
208 value of ∆f H◦ and according to NIST for molecule 1,2-dimethoxy-benzene
209 it achieves −l2iq90 kJ/mol [39]. The values of formation heat of xanthione
210 methoxy derivatives shown in Table 1 manifest similar decreasing tendency
211 with growing the number of attached methoxy groups. The formation heat
212 in gas phase, ∆f Hg◦as, at the standard conditions evaluated by NIST for the
213 thioxanthone molecule which geometry structure close to compounds under
214 consideration is positive and equals 95 kJ/mol [40].
215
Additional calculation of the thioxanthone molecule with full geometry
216 optimization in gas phase gives the value of the heat of formation 115.5
217 kJ/mol which is pretty close to the experimental data. Thus we deduce that
218 selected method reproduces the values of the heat of formation for xanthione-
7

219 like molecules with precision ±20 kJ/mol. The heats of formations in gas and
220 liquid phase for compounds I–IV are also shown in Table 1. The differences
221 of ∆f H◦ between the gas and liquid phases for considered compounds are
222 about 2 − 3 kcal/mol. It seems that in COSMO model the hexane solvent
223 influences negligibly on the geometry structure of the compounds.
224
The correlation of the ionization potential, IP, the energy of the highest
225 occupied molecular orbital (HOMO), EHOMO, and the energy of the lowest
226 unoccupied molecular orbital (LUMO), ELUMO, with NIST data is also pro-
227 nounced. The ionization potentials of methoxy derivatives measured in gas
228 phase decrease with attaching a methoxy group to the benzene: from 9.24 eV
229 (benzene) to 8.20 eV (anisole). However the attachment of the next methoxy
230 group to anisole influences negligibly on IP: 8.20 eV (anisole) → 8.14 eV (1,3-
231 dimethoxy-benzene) whereas attachment of a group in the neighbor position
232 leads to the significant decrease of IP: 7.8 eV (1,2-dimethoxy-benzene). The
233 very similar picture one obtains with IP of the XT methoxy derivatives (Ta-
234 ble 1).
235
Analysis of the Mulliken charges at atoms of XT skeleton (Table 2) shows
236 a strong polarization of the compounds that is confirmed by high magnitudes
237 of the dipole moments presented in Table 1. The high magnitudes of the
238 dipole moments are also observed for all the conformers. According to the
239 Table 2, the negative charges at the carbon atoms C6 and C11 in the com-
240 pound I increase by 0.2 units and become positive after attaching methoxy
241 groups. At the same time at the oxygen atoms of the methoxy groups neg-
242 ative charge reaches −0.19. The latter leads to change both in magnitude
243 and in direction of the dipole moment (see Figures 1 and 2). The influence
244 of the methoxy group attachment on charges of the other atoms of the com-
245 pounds II and IV is negligible. However in the compound III the methoxy
246 groups attachment to the carbons C2, C6 affects the charges of all carbon
247 atoms of the benzene fragment (from C1 to C6). The fragment is strongly
248 polarized resulting in its distortion in comparison with the opposite benzene
249 fragment (C8-C13). A comparison of the results of both the formation heat
250 and the dipole moment of the compounds II and IV shows the conformers
251 with smaller dipole moment to be energetically favorable whereas compound
252 III to demonstrate the opposite situation.
8

253 3.2. Xanthione and its methoxy derivatives absorption and fluorescence spec-
254
tra
255
The calculations for several versions of the configuration interactions with
256 different number of molecular orbitals were carried out: C.I. = 4, 6, 8, 9.
257 Figure 3 shows the absorption and fluorescence spectra of the XT only for
258 C.I. = 9 (the rest results are presented in supporting information in Figure
259 S1). The absorption spectra obtained for C.I. = 9 are characterized by four
260 well-defined lines corresponding to the main experimental bands. The energy
261 diagram of XT is shown in Figure 4. The transitions between S0 and S2, S4, S9
262 states of XT molecule have considerably lower oscillator strengths comparing
263 with the transition S15 ← S0. Due to the symmetry of the molecule the
264 oscillator strength of S1 ← S0 transition equals to zero. The fluorescence
265 bands are calculated for optimized geometry of the S2 and S3 excited states.
266
The absorption and fluorescence spectra as well as energy diagram of two
267 conformers of DXT are shown in Figures 5 and 6. The intensity of a transi-
268 tion to the first excited state does not equal to zero. That can be explained
269 by broken symmetry of the compound due to the attachment of a methoxy
270 group to a XT molecule. The transitions S3 ← S0 and S15 ← S0 have os-
271 cillator strengths of the same order. For both conformers an additional line
272 appears in the absorption spectra. Comparing the MECI calculations of the
273 absorption band at 380–400 nm which is mainly due to S2 ← S0 and S3 ← S0
274 transitions for DXT with the experimental spectra we can conclude that in
275 experimental spectrum both transition bands seem to have similar intensi-
276 ties. Therefore the mirror symmetry of absorption and fluorescence bands
277 is broken. It should be noted that conformational changes influence slightly
278 on the spectrum of the compound III. The latter can be explained by the
279 fact that the conformational changes in the compounds III do not distort the
280 XT skeleton in the ground electronic state. Comparing fluorescence spectra
281 presented in Figures 3 and 5, one can see that the vibrational structure is
282 smoothed in DXT although it is seen in the XT spectrum. Appearance of
283 the additional line in the calculated absorption and fluorescence spectra can
284 explain the reason of vibrational structure smoothing due to the overlap of
285 two transitions.
286
In the case of XT there are minor shifts of the calculated lines relatively to
287 that observed in experiments. For the compound III the simulated spectrum
288 well reproduces a shift of the S2 ← S0 transition line to the blue in accordance
289 with the experiment. The experimentally observed broadening of the band
9

Table 3: The experimental and calculated (in parentheses) oscillator strengths of elec-
max
abs
tronic transitions in XT and DXT. ε
is the extinction coefficient.
Compound εmabasx, M−1cm−1 S1 ← S0
S2 ← S0
0.31 (4.52) 0.03 (1.13) 0.46 (5.27) 1.34 (26.41)
0(10−5) 0.45 (0.44) 0.45 (5.51) 0.58 (2.25) 1.59 (11.02)
S3 ← S0
S4 ← S0
S5(15) ← S0
XT
10700
10000
0 (0)
DXT
290 in the vicinity of 360 nm for the compound III relatively to the XT band can
291 be associated with the appearance of additional transition S3 ← S0.
292
The luminescence spectra of XT and DXT shown in Figures 3 and 5
293 are consisted of two parts: fluorescence, S2 → S0, being in the spectral re-
294 gion 400 – 600 nm and phosphorescence, T1 → S0, (600–800 nm) [22]. The
295 luminescence spectrum of DXT as well as its absorption spectrum reveals
296 the hypsochromic shift relative to XT luminescence one. However, the shift
297 ∆νfl ' 300 cm−1 evaluated as difference between S2 → S0 fluorescence max-
298 ima of XT and DXT is 6 times smaller than ∆νabs ' 1800 cm−1 for S2 ← S0
299 absorption maxima. This fact can be explained by considerable variation
300 of the normal vibrational frequencies of XT accompanying electronic exci-
301 tation S2 ← S0. It is also interesting to note that there is a difference of
302 full experimental widths at the half maximum of vibronic bands between XT
303 (∆νh/2 = 2950 cm−1) and DXT (∆νh/2 = 4730 cm−1). At the same time the
304 difference for the fluorescence bands (4200 cm−1 and 4800 cm−1) is drastically
305 smaller. The reason is the superposition of the two absorption bands S2 ← S0
306 and S3 ← S0. Increasing absorption band width in DXT is in agreement with
307 quantum chemical calculations. The prediction of nonzero oscillator strength
308 for S3 ← S0 transition that overlaps with S2 ← S0 transition explains the
309 broadening of the band. At the same time fluorescence of DXT occurs from
310 a single state that results in practically identical fluorescence band widths in
311 XT and DXT. This is also supported by small hypsochromic shift of S2 → S0
312 maximum from 448 nm to 442 nm and much larger shift (27 nm) for S2 ← S0.
313 The parameters of stationary absorption and luminescence bands are listed
314 in Table 4. Here the shifts ∆νband are determined as the differences between
315 corresponding experimental band maxima of DXT and XT, pictured in Fig-
316 ures 5 and 3.
317
The phosphorescence, T1 → S0, in XT and DXT, also exhibits a hyp-
10

Table 4: The characteristics of the stationary absorption and luminescence bands of the
band
max
XT and DXT. The energy (E
) and the wavelength (λmax) of the corresponding band
maximum. The shifts of the band maxima DXT relative to XT, ∆νband
.
XT
λmax
DXT
λmax,
DXT vs XT
band
max
band
E
,
,
E
,
max
band
∆νband, cm−1
eV
1.851
nm
670
632
450
400
307
255
229
eV
nm
630
−
ph
abs1
fl2
abs2
abs3
abs4
abs5
1.968
b
950
−
1.962a
2.755
3.100
4.039
4.863
5.415
−
2.805
3.221
4.052
4.769
5.322
442
385
306
260
233
406
975
105
−758
−750
athe value was taken from ref. 41
bthe band was not observed
318 sochromic shift of the vibronic maximum from 670 nm in XT to 630 nm in
319 DXT that amounts to ∆νPh ≈ 950 cm−1. That shift is 2.4 times larger than
320 ∆νf ≈ 400 cm−1 for fluorescence maxima of XT and DXT.
321
Comparing the allowed S2 → S0 and T1 → S0 transitions of π π∗ type
322 with the forbidden S1 → S0 transition of n π∗ type, it can be assumed that
323 attached methoxy groups weakly influence on S1 → S0 transition. The quan-
324 tum chemical calculations confirm the conclusion (see Figures 4 and 6). Con-
325 sequently, the energy of S1 state in DXT is close to that of XT. Since the
326 locations of S1 n π∗ and T1 π π∗ states in XT[22] are 15970 cm−1 and 15130
327 cm−1, respectively, the energies of these two states coincide with that in
328 DXT. Therefore due to the intersystem crossing the total excitation energy
329 of S1 state deactivates through T1 state.
330
The bands of absorption spectra for S1 ← S0 transition in Figures 3
331 and 5 are not shown since they are very weak. For example for XT the
332 molar extinction coefficient 20 M−1cm−1 and an oscillator strength of 0.0003
333 [42]. There is some difference between calculated and experimental results.
334 The positions of the calculated absorption lines are shifted to the long-wave
335 region, 740 nm, whereas in the experimental spectra such maxima S1 ← S0
336 are observed in the vicinity of 630 nm for XT [41].
337
To understand the nature of the spectra of the considered compounds,
11

Table 5: The characteristics of the XT and DXT compounds in the electronic states: S0,
S2, and S3. Here q is a cha∗rge of corresponding atom. The dimensions of the quantities
are the same as in Table 1. The results presented in row 6 are obtained at the geometry
∗∗
termination optimization criteria: GNORM = 3.
Results obtained at GNORM = 17.
No Comp. R1
R2
H
H
State ∆f H◦
Etot
µ
qC3
qC7
,
qC9
qS
qO(R1) qO(R2)
1
2
I
I
H
H
S0
S2
S3
S0
S2
S3
S0
S2
S3
34.18 -2141.762 3.875 -0.16 0.05 -0.16 -0.16
102.54 -2138.798 5.917 -0.1 -0.03 -0.1 -0.31
113.25 -2138.334 6.198 -0.11 -0.19 -0.11 -0.28
-33.13 -3027.143 4.251 -0.22 0.07 -0.16 -0.17 -0.19 -0.15
34.48 -3024.211 5.451 -0.12 -0.02 -0.12 -0.29 -0.17 -0.12
41.75 -3023.895 6.233 -0.14 -0.16 -0.08 -0.25 -0.19 -0.08
-33.15 3027.144 4.925 -0.22 0.07 -0.16 -0.17 -0.19 -0.15
-
-
-
-
-
-
3
I
H
H
4
5
6∗
7
8
9∗∗
III
III
III
III
III
III
H3CO(cis)
H3CO(cis)
H3CO(cis)
H3CO(trans) H3CO
H3CO(trans) H3CO
H3CO(trans) H3CO
H3CO
H3CO
H3CO
42.13 -3023.879 6.888 -0.15 -0.11 -0.12 -0.25 -0.2
41.98 -3023.886 7.269 -0.14 -0.17 -0.09 -0.26 -0.2
-0.05
-0.05
338 calculations of energies of the excited states S2 and S3 were performed with
339 full geometry optimization. The results are listed in Table 5. According to
340 the calculations, the excitation of S2 state results in considerable variation
341 of the sulfur charge and increase of the bond length between C7 and the
˚
342 sulfur by 0.2 A. The charges on C3 and C9 also slightly vary. Since the area
343 where a variation of the atomic charges is not bulky the alteration of the
344 dipole moment is not strong. In XT derivatives containing a methoxy group
345 the oxygen charge may noticeably vary but only for some their conformers.
346 Thus, methoxy group demonstrates its electron-donor origin.
347 3.3. Transient spectra of XT and DXT
348
The transient time-resolved spectra of XT and DXT do not have great
349 differences. Our femtosecond measurements of XT and DXT in n-hexane
350 at room temperature allowed to reveal the transient absorption spectra at
351 λmax ' 650 nm and gain spectra at λmax ' 480 nm. The kinetics of TA
352 signal at 475 nm consist of two components: the short component with a
353 lifetime about 20 ps and very long-lived one exceeding 100 ps. Apparently,
354 the first short lifetime of 20 ps associates with the excited S2 state decay and
355 long time constant probably characterizes rising the triplet triplet Tk ← T1
356 absorption.
357
The femtosecond measurements of XT and DXT in n-hexan at room
358 temperature in the wavelength domain from λ = 400 to 740 nm in the time
359 window from 1 ps to 100 ps were carried out. The transient time-resolved
360 spectra of XT and DXT are presented in Figures 7.
12

361
The complex dynamics of experimental transient spectra shown in Fig. 7
362 leads to the supposition that there are several constituents. Among them
363 the most pronounced are: two excited-state absorption bands, S2 → Sn and
364 S1 → Sm, stimulated emission, S2 → S0, triplet-triplet absorption, T1 → Tk
365 and the bleach (the bleach is not shown because it was out of the experimental
366 spectral window). The transient spectra of DXT pictured in Fig. 7 are similar
367 to XT spectra but there are differences. DXT spectra consist at least of
368 six constituents: two excited-state absorption bands, S2 → Sn, the excited
369 state absorption S1 → Sm, stimulated emission, S2 → S0 and triplet-triplet
370 absorption, T1 → Tk and the bleach.
371
For quantitative description of spectral evolution the theory reported in
372 ref [43] has been adopted. The transitions S2 → S1 → T1 are considered to
373 be exponential. The simulations confirm the adopted schema of electronic
374 transitions in excited XT and DXT: the initially excited S2 states decay due
375 to internal conversion to S1 state with time constant τIC that subsequently
376 decays due to intersystem crossing to populate the triplet state T1 with time
377 constant τISC
.
378
The transient absorption signal ∆A in the pump-probe technique can be
379 represented as follows [43]
X
∆A(ωe, ωp, τ) =
∆Ai(ωe, ωp, τ)Pj(τ)
(1)
i
where the subscript i runs through the values i = {ESAk, TTA, SE, BL} ac-
counting for the k-th excited-state absorption, ∆AESAk, the triplet-triplet
absorption, ∆ATTA, stimulated emission, ∆ASE, and bleach, ∆ABL, the in-
dex j can take any value from the following list of the electronic states
{S2, S1, T1, S0}:
Vp2
m
e−ξ ξi
X
i
πτp
P σi(τ)
∆Ai(ωe, ωp, τ) = ωpwi
Pn
×
m!
n,m
380
ꢀ
ꢁ
(δωmi − Qni(τ))2
× exp −
.
(2)
2σi2(τ)
381 Here wi is the weight of corresponding band, σi(τ) and Qni(τ) are the values
382 of the time-dependent half-width and the coordinate of the center-of-gravity
383 of the corresponding band:
σi2(τ) = τp−2 + 2ErikBT − (2ErikBTX(τ)/σe2)2,
(3)
13

Qni(τ) = 2EriX(τ)(1 − δi,BL + δωnekBT/σe2),
(4)
384 where Eri is the reorganization energy of the solvent for i-th transition,
385 ξi = Erhi/Ωhf is the Huang-Rhys factor, Erhi and Ωhf are the reorganization
386 energy and frequency of the high-frequency mode for a corresponding transi-
387 tion. The rest of parameters is δωne = ωe +∆GSE −ErSE −nΩhf , δωmESAk
=
388 ωp + ∆GESAk − ErESAk − mΩhf , δωmTTA = ωp + ∆GTTA − ErTTA − mΩhf ,
389 δωmSE = ωp + ∆GSE + ErSE + mΩhf , δωmBL = ωp + ∆GSE − ErSE − mΩhf ,
P
390 where ∆Gi are the free energies for ESAk, TTA, SE, and BL. P =
391 where Pn is the probability of the transition generating n vibrational quanta
392 during the pump stage:
Pn,
ꢀ
ꢁ
n
πVe2τe e−ξ ξSE
(δωne)2
2σe2
SE
Pn =
exp −
,
(5)
σe
n!
393 σe2 = τe−2 + 2ErSEkBT. Here τp and ωp are the duration and frequency of the
394 probe pulse, τe and ωe are the duration and carrier frequency of the pump
395 pulse, δi,BL is the Kronecker delta, kB and T – the Boltzmann constant and
396 the temperature, respectively, the Planck constant is set to be equal to unity,
397 ~ = 1. Ve and Vp are the matrix elements of the dipole interaction with the
398 electromagnetic wave field strength for the pump and probe pulses. The
399 spectral dynamics is supposed to be much slower than the intramolecular
400 vibrational redistribution and relaxation. This is the reason why the vibra-
P
2
i
401 tional relaxation is considered to be instant. X(t) =
i=1 xie−t/τ is the
402 relaxation function of the solvent. The populations kinetics of the excited
403 states S2, S1, and T1 are
PS2(t) = e−t/τ
,
(6)
(7)
IC
404
405
(e−t/τ − e−t/τ )τIC
IC
ISC
PS1(t) =
,
τIC − τISC
e−t/τ τIC − e−t/τ τISC
IC
ISC
PT1(t) = 1 −
,
(8)
τIC − τISC
406 where τIC and τISC are the internal conversion and the intersystem crossing
407 time constants, respectively. The nonradiative channel S2 ← S0 is too weak
408 to be include in consideration. The model does not take into account the
409 ground state recovery because the main channel of decay of S2 is S2 → S1 →
410 T1 and the lifetime of T1 is much larger than the timescale of the transient
14

411 absorption experiment. That is the reason why the ground state does not
412 recovery, hence, its population is constant, PS0 = 1.
413
The parameters fixed for all calculations are following: kBT = 0.25 eV,
414 Ωhf = 0.15 eV, ωe = 3.1 eV, τe = τp = 100 fs, Vp = 0.005 eV. The parameters
415 are set to satisfy the experimental setup. The vibrational frequency is de-
416 termined from the structure of the XT fluorescence band. The dynamics of
417 the experimental transient spectra is not very fast and the solvent relaxation
418 does not manifest itself. This means that the time scale of the solvent relax-
419 ation is much shorter than that of electronic transitions. Since the dielectric
420 solvation in nonpolar solvents is insignificant, the response of a solvent on
421 the alteration of volume of a molecule accompanying electronic transition is
422 an important part of the reorganization of such solvents. The relaxation of
423 such a response should correlate with its viscosity [44]. The dynamics pa-
424 rameters of n-hexane are not known. Since n-hexane has rather low viscosity
425 its relaxation is expected to be fast: τ1 = 100 fs, τ2 = 600 fs, x1 = 0.6,
426 x2 = 0.4. The parameters obtained in the fitting are listed in Tables 6 and
427 7. A similar approach to transient absorption spectra modelling was used in
428 recent publication [45].
Table 6: The model parameters are fitted from TA spectra of XT. Emax is the energy of
the maximum of the corresponding band. All the energies are in eV.
No
1
2
i
wi
j
Eri Erhi ∆Gi Emax
ESA2 1.4 S2 0.1 0.12 -1.82 2.04
ESA1 1.4 S1 0.1 0.12 -1.88 2.10
3
4
5
SE
TTA 0.8 T1 0.1 0.16 -2.57 2.83
BL 0.2 S0 0.1 0.37 -2.97 3.44
0.8 S2 0.1 0.37 -2.97 2.50
Table 7: The model parameters are fitted from TA spectra of DXT. Emax is the energy
of the maximum of the corresponding band. All the energies are in eV.
No
1
2
i
wi
j
Eri Erhi ∆Gi Emax
ESA3 0.6 S2 0.12 0.18 -1.70 2.00
ESA2 1.15 S2 0.12 0.12 -2.72 2.96
ESA1 1.15 S1 0.12 0.12 -2.72 2.96
3
4
5
6
SE
TTA 0.3 T1 0.12 0.14 -2.46 2.72
BL 0.5 S0 0.12 0.37 -2.97 3.46
0.5 S2 0.12 0.38 -2.92 2.34
429
For both XT and DXT the negative signal in the area around 480 nm
15

430 predominantly consists of the stimulate emission S2 → S0 but there is a con-
431 siderable overlapping with ESAk and TTA bands. Triplet-triplet absorption
432 leads to a positive signal in the area of 450 nm when time exceeds 10 and 40 ps
433 in XT and DXT, correspondingly. For XT the wavelength of triplet-triplet
434 absorption is in good agreement with data from reference [32]. Overlap-
435 ping of the bands requires modelling of transient spectra for determining the
436 time constants of the internal conversion S2 → S1 and intersystem crossing
437 S1 → T1. Analysis of decay kinetics of the signal at 499 gives τIC = 20.0
438 and at 523 nm gives τIC = 28.0 ps. The fitting by using Eq. 1 allows de-
439 termination of both the internal conversion and intersystem crossing time
440 constants. They are τIC = 14.0 and τISC = 17.0 ps for XT and τIC = 26.0
441 and τISC = 20.0 ps for DXT. Larger difference in estimations of τIC for XT is
442 caused by larger overlapping of the stimulated emission S2 → S0 and excited
443 state absorption bands from the S1 and S2 states in the red wing.
444
To fit the transient spectra of XT and DXT a few ESA bands are ex-
445 ploited. The energies of ESAk maxima obtained from the fitting are listed
446 in Tables 6 and 7. Their values well correlate with data obtained from sta-
447 tionary absorption spectra. For example, ESA from the second excited state
448 has a maximum at 2.04 eV that is close to the difference of the energies
abs4
fl2
449
E
− E
= 4.86 − 2.75 = 2.11 eV (see Table 4). ESA from the first
max
max
450 excited state has a maximum at 2.10 eV that is close to the difference of the
451 energies E
452 obtained for DXT if one supposes that the S1 energy level is higher than the
453 T1 level by a few kBT.
abs3
max
ph
max
− E
= 4.04 − 1.85 = 2.19 eV. Similar correlations can be
454
The kinetic curves for several selected wavelengths are shown in Figures
455 8 and 9 for XT and DXT. The figures also demonstrate the contributions of
456 the constituents to the total signal depicted by colored curves. The necessity
457 of accounting for S1 → Sn absorption in XT is well seen from the signal
458 kinetics at 599 and 635 nm pictured in Figure 8. S2 → Sn absorption in
459 DXT is also seen at 419 and 433 nm in Figure 9.
460
Despite the simplicity of the kinetic model Eqs. 6 – 8 applied to each sep-
461 arate constituent, the total signal at selected wavelengths is well reproduced
462 in whole. The negligible deviations of the simulated curves from experimen-
463 tal data are apparently caused by weak excited state absorption bands that
464 are not included in the model.
16

465 4. CONCLUSIONS
466
The quantum chemical calculations of methoxy derivatives of the XT
467 molecule in the ground electronic state reveal an influence of methoxy groups
468 on the XT geometry and electronic structure. It is ascertained that only
469 the compound III demonstrates remarkable geometry changes. Therefore
470 the deformation of the XT skeleton plane may result in the appearance of
471 enantiomer properties.
472
A comparison of MECI calculations of the pure XT and its methoxy
473 derivative shows a shift and pronounced broadening of absorption spectra
474 band of newly synthesized compound III relatively to the band of the com-
475 pound I. The experimentally observed broadening for the compound III can
476 apparently be associated with the manifestation of transition S3 ← S0 that
477 is forbidden in XT. The calculations of the second excited state of XT and
478 its methoxy derivative confirm noticeable charge transfer (about 0.1 of the
479 charge of an electron) from the methoxy group to the thiocarbonyl group.
480
The fitting of the transient spectra measured in n-hexane has shown that
481 the decay time constants of XT and DXT S2 state population considerably
482 differ amounting 14.0 and 26.0 ps, respectively. That presumably evidences
483 that the charge transfer in DXT is not sufficiently large to effectively quench
484 the second excited electronic state by electron transfer. The primary channel
485 of the second excited state deactivation is the internal conversion in the both
486 XT and DXT with subsequent intersystem crossing to T1. The time constants
487 of the intersystem crossing S1 → T1 for XT and DXT amount to 17.0 and 20
488 ps, correspondingly and is comparable with the lifetime of S2 state.
489
Noticeable enlarging the S2 state lifetime in DXT relatively to XT is prob-
490 ably caused by the increase of the energy gap between the first and second
491 excited states and partial charge transfer from the methoxy to thiocarbonyl
492 groups. Thus, methoxi derivatives of XT may be perspective molecules for
493 molecular devices exploiting the second excited state.
494 ACKNOWLEDGMENTS
495
This work was supported by the Russian Foundation for Basic Research
496 (Grants No. 16-53-00189) and the Belarusian Republican Foundation for
497 Fundamental Research (Grants No. C16R-104).
17

498
499
500
[1] E. Greenbaum, Biomolecular electronics and applications, in: R. Mey-
ers (Ed.), Molecular biology and biotechnology: A comprehensive desk
reference, Wiley-VCH, Inc., Toronto, 1995, pp. 98–106.
501
502
503
[2] S.-S. Sun, L. R. Dalton (Eds.), Introduction to organic electronic and
optoelectronic, Taylor & Francis Group, LLC, Boca Raton, FL, 2008, p.
936.
504
505
506
[3] O. Ostroverkhova, Organic optoelectronic materials: mechanisms and
applications, Chem. Rev. 116 (2016) 13279–13412. doi:10.1021/acs.
chemrev.6b00127.
507
508
509
[4] M. Sauer, P. Tinnefeld, Molecular optical switches and waveguides. new
concepts of tiny electronic devices, Optic & Photonik 2 (2007) 45–48.
doi:10.1002/opph.201190251.
510
511
512
513
514
515
[5] S. Wallin, C. Monnereau, E. Blart, J.-R. Gankou, F. Odobel, L. Ham-
marstro¨m, State-selective electron transfer in an unsymmetric acceptor-
Zn(II)porphyrin-acceptor triad: toward a controlled directionality of
electron transfer from porphyrin S2 and S1 states as a basis from
molecular switch, J. Phys. Chem. A 114 (2010) 1709–1721. doi:
10.1021/jp907824d.
516
517
518
519
520
[6] M. Andersson, J. Davidsson, L. Hammarstr¨om, J. Korppi-Tommola,
T. Peltola, Photoinduced electron transfer reactions in a porphyrin- vi-
ologen complex: observation of S2 to S1 relaxation and electron trans-
fer from S2 state, J. Phys. Chem. B 103 (1999) 3258–3262. doi:
10.1021/jp9829795.
521
522
523
524
[7] N. Mataga, Y. Shibata, H. Chosrowjan, N. Yoshida, A. Osuka, Inter-
nal conversion and vibronic relaxation from higher excited state of por-
phyrins: femtosecond fluorescence dynamics studies, J. Phys. Chem. B
104 (2000) 4001–4004. doi:10.1021/jp9941256.
525
526
527
528
[8] P.-A. Muller, E. Vauthey, Charge recombination dynamics of geminate
ion pairs formed by electron transfer quenching of molecules in an upper
excited state, J. Phys. Chem. A 105 (2001) 5994–6000. doi:0.1021/
jp010015z.
18

529
530
531
[9] A. Morandeira, L. Engeli, E. Vauthey, Ultrafast charge recombination of
photogenerated ion pairs to an electronic excited state, J. Phys. Chem.
A 106 (2002) 4833–4837. doi:10.1021/jp014393f.
532 [10] N. Mataga, S. Tanigushi, H. Chosrowjan, A. Osuka, N. Yoshida, Ul-
533
534
535
536
trafast charge transfer and radiationless relaxations from higher excited
state (S2) of directly linked zinc-porphyrin (ZP)-acceptor dyads: inves-
tigations into fundamental problems of exciplex chemistry, Chem. Phys.
295 (2003) 215–228. doi:10.1016/j.chemphys.2003.09.005.
537 [11] M. Kubo, Y. Mori, M. Otani, M. Murakami, Y. Ishibashi, M. Ya-
538
539
540
suda, K. Hosomuzu, H. Imahori, S. Nakashima, Ultrafast photoinduced
electron transfer in directly linked porphyrin-ferrocene dyads, J. Phys.
Chem. A 111 (2007) 5136–5143. doi:10.1021/jp071546b.
541 [12] M. Fujitsuka, D. W. Cho, S. Tojo, A. Inoue, T. Shiragami, M. Yasuda,
542
543
544
T. Majima, Electron transfer from axial ligand to S1- and S2-excited
phosphorus tetraphenylporphyrin, J. Phys. Chem. A 111 (2007) 10574–
10579. doi:10.1021/jp076303y.
545 [13] V. N. Ionkin, A. I. Ivanov, Numerical simulations of ultrafast charge
546
547
548
549
separation dynamics from second excited state of directly linked zinc-
porphyrin-imide dyads and ensuring hot charge recombination into the
first excited state, J. Phys. Chem. A 113 (2009) 103–107. doi:10.1021/
jp806589m.
550 [14] M. V. Rogozina, V. N. Ionkin, A. I. Ivanov, What factors control product
551
552
553
yield in charge separation reaction from second excited state in zinc-
porphyrin derivatives?, J. Phys. Chem. A 116 (4) (2012) 1159–1167.
doi:10.1021/jp210988k.
554 [15] M. V. Rogozina, V. N. Ionkin, A. I. Ivanov, Dynamics of charge sepa-
555
556
557
ration from second excited state and following charge recombination in
zinc-porphyrin-acceptor dyads, J. Phys. Chem. A 117 (2013) 4564–4573.
doi:10.1021/jp402734h.
558 [16] B. Robotham, K. A. Lastman, S. J. Langford, K. P. Ghiggino, Ultra-
559
560
561
fast electron transfer in a porphyrin-amino naphthalene diimide dyad,
J. Photochem. Photobiol. A 251 (2013) 167–174. doi:10.1016/j.
jphotochem.2012.11.002.
19

562 [17] S. V. Feskov, A. I. Ivanov, Efficiency of intramolecular charge separation
563
564
565
from the second excited state: suppression of the hot charge recombina-
tion by electron transfer to the secondary acceptor, J. Phys. Chem. A
117 (45) (2013) 11479–11489. doi:10.1021/jp408516q.
566 [18] M. V. Rogozina, V. N. Ionkin, A. I. Ivanov, Kinetics of charge separated
567
568
569
state population produced by intramolecular electron transfer quenching
of second excited state, J. Photochem. Photobiol. A 301 (2015) 55–61.
doi:10.1016/j.jphotochem.2015.01.002.
570 [19] S. V. Feskov, A. I. Ivanov, Effect of geometrical parameters of dyad d−a
571
572
573
and triad d − a1 − a2 on the efficiency of ultrafast intramolecular charge
separation from second excited state, Chem. Phys. 478 (2016) 164–172.
doi:10.1016/j.chemphys.2016.03.013.
574 [20] S. V. Feskov, V. A. Mikhailova, A. I. Ivanov, Non-equilibrium effects in
575
ultafast photoinduced charge transfer kinetics, J. Photochem. Photobiol.
576
C 29 (2016) 48–72. doi:10.1016/j.jphotochemrev.2016.11.001.
577 [21] V. Ramamurthy, K. S. Shanze (Eds.), Organic photochemistry and pho-
578
tophysics, Vol. 14, Taylor & Francis Group, LLC, Boca Raton, FL, 2006,
p. 300.
579
580 [22] A. Maciejewski, R. P. Steer, The photophysics, physical photochemistry,
581
and related spectroscopy of thiocarbonyls, Chem. Rev. 93 (1993) 67–98.
582
doi:10.1021/cr00017a005.
583 [23] N. O. Obi-Egbedi, I. B. Obot, M. I. El-Khaiary, Quantum chemical in-
584
585
586
587
vestigation and statistical analysis of the relationship between corrosion
inhibition efficiency and molecular structure of xanthene and its deriva-
tives on mild steel in sulphuric acid, J. Mol. Struct. 1002 (1–3) (2011)
86–96. doi:10.1016/j.molstruc.2011.07.003.
588 [24] A. Maciejewski, R. P. Steer, Photophysics of the second excited singlet
589
states of xanthione and related thiones in perfluoroalkane solvents, J.
590
Am. Chem. Soc. 105 (1983) 6738–6740. doi:10.1021/ja00360a046.
591 [25] J. J. P. Stewart, Optimization of parameters for semiempirical methods
592
I. method., J. Comput. Chem. 10 (1989) 209–220. doi:10.1002/jcc.
540100208.
593
20

594 [26] J. J. P. Stewart, Comparison of the accuracy of semiempirical and some
595
dft methods for predicting heats of formation, J. Mol. Model 10 (2004)
596
6–12. doi:10.1007/s00894-003-0157-6.
597 [27] J. J. P. Stewart, Use of semiempirical methods for detecting anomalies in
598
reported enthalpies of formation of organic compounds, J. Phys. Chem.
599
Ref. Data 33 (2004) 713–724. doi:10.1063/1.1643403.
600 [28] G. B. Rocha, R. O. Freire, A. M. Simas, J. J. P. Stewart, RM1: A
601
reparameterization of AM1 for H, C, N, O, P, S, F, Cl, Br, and I, J.
602
Comput. Chem. 27 (2006) 1101–1111. doi:10.1002/jcc.20425.
603 [29] J. J. P. Stewart, Optimization of parameters for semiempirical methods
604
II. applications., J. Comput. Chem. 10 (1989) 221–264. doi:10.1002/
jcc.540100209.
605
606 [30] A. Klamt, G. Scu¨u¨mann, COSMO. a new approach to dielectric screen-
607
608
609
ing solvents with explicit expressions for the screening energy and
its gradient, J. Chem. Soc., Perkin Trans. 2
(5) (1993) 799–805.
doi:10.1039/P29930000799.
610 [31] D. R. Armstrong, R. Fortune, P. G. Perkins, Molecular orbital theory for
611
the excited state of transition metal complexes, J. Chem. Soc., Faraday
612
Trans. 2 68 (1972) 1839–1846. doi:10.1039/F29726801839.
613 [32] M. Szymanski, M. Balicki, M. Binkowski, J. Kubicki, A. Maciewski,
614
615
616
617
E. Pawlowska, T. Wrozowa, Effects of solute-solvent interactions on ra-
diationless decay of thioketones excited to their s-2-states and t-1-states,
Acta Physica Polonica A 89 (1996) 527–544. doi:10.12693/APhysPolA.
89.527.
618 [33] N. G. Bakhshiev, M. I. Knyazhanskii, V. I. Minkin, O. A. Osipov, G. V.
619
620
621
Saidov, Experimental determination of the dipole moments of organic
molecules in excited electronic states, Russ. Chem. Rev. 38 (9) (1969)
740–754. doi:10.1070/RC1969v038n09ABEH001831.
622 [34] N. A. Borisevich, O. V. Bouganov, S. A. Tikhomirov, G. B. Tol-
623
624
625
626
storozhev, G. L. Shkred, Passive mode locking of a femtosecond
ti:sapphire laser with pulsed synchronous pumping by a finite train of
picosecond pulses, Quant. Electron. 29 (1999) 780–786. doi:10.1070/
QE1999v029n09ABEH001572.
21

627 [35] V. I. Stsiapura, S. A. Kurhuzenkau, V. A. Kuzmitsky, O. V. Bouganov,
628
629
630
S. A. Tikhomirov, Solvent polarity effect on nonradiative decay rate
of thioflavin T, J. Phys. Chem. A 120 (28) (2016) 5481–5496. doi:
10.1021/acs.jpca.6b02577.
631 [36] H. Y. Afeefy, J. F. Liebman, S. E. Stein, Neutral thermochemical data,
632
633
634
635
636
in: P. J. Linstrom, W. G. Mallard (Eds.), NIST Chemistry WebBook.
NIST Standard Reference Database Number 69, National Institute of
Standards and Technology, Gaithersburg, MD, 20899, (retrieved March
10, 2017).
URL http://webbook.nist.gov
637 [37] M. V. Roux, M. Tompardo, J. C. Chickos, Y. Nagano, Critacally
638
639
640
evaluated thermodynamical properties of polycyclic aromatic hydrocar-
bons, J. Phys. Chem. Ref. Data 37 (2008) 1855–1996. doi:10.1063/1.
2955570.
641 [38] J. O. Fenwick, D. Harrop, A. J. Head, Thermodynamic properties of
642
643
644
organic oxigen compounds 41. enthalpies of formation of eight ethers,
J. Chem. Thermodyn. 7 (1975) 943–954. doi:10.1016/0021-9614(75)
90158-5.
645 [39] J. D. Cox, G. Plicher, Thermochemistry of organic and organomethallic
compounds, Academic Press, New York, 1970.
646
647 [40] R. Sabbah, L. E. Watik, Etude therodynamique de l’acridone et de la
thioxanthone, Can. J. Chem. 70 (1992) 24–28. doi:10.1139/v92-005.
648
649 [41] J. R. Huber, M. Mahaney, S2 → S0 fluorescence in an aromatic
650
651
thiokethone, xanthione, Chem. Phys. Lett. 30 (1975) 410–412. doi:
10.1016/0009-2614(75)80019-4.
652 [42] D. A. Capitanio, H. J. Pownall, J. R. Huber, Spectroscopic and photo-
653
chemical properties of aromatic thioketones: xanthione, J. Photochem.
654
3 (1) (1974) 225–236. doi:10.1016/0047-2670(74)80022-5.
655 [43] R. G. Fedunov, A. I. Ivanov, Influence of spectral characteristics of the
656
657
658
pump pulse on the transient absorption of donoracceptor complexes in
polar solvents, J. Russ. Laser Res. 33 (2) (2012) 152–165. doi:10.1007/
s10946-012-9269-8.
22

659 [44] A. I. Ivanov, Hydrodynamic model of complex volume fluctuations in-
660
teractions with medium and its influence on the rate of outer-sphere
electron transfer, Russian J. Chem. Phys. 9 (1990) 1276–1282.
661
662 [45] R. G. Fedunov, A. Plotnikova, A. I. Ivanov, E. Vauthey, Simulations of
663
664
665
the ultrafast transient absorption dynamics of a donor-acceptor biaryl
in solution, J. Phys. Chem. A 121 (2) (2017) 471–481. doi:10.1021/
acs.jpca.6b11581.
23

1
3
6
2
9
7,8
S
R2
C2
C7
C10
121
C1
C3
C9
C11
A2
118
122
121
121
124
C6
C4
C8
C12
119
118
114
114
C5
C13
O
R1
A1
Figure 1: The geometry structure of XT derivatives in the ground state for compounds:
I, II, and IV. The arrows show the dipole moment directions for a number of conformers.
The conformer numeration is given in Table 1.
24

R1
C2
S
4,5
A2
126.7
C7
C10
C1
C6
C3
C4
C9
C8
C11
116.1
118.6
123.4
119.9
A1
123.5
118.1
117.3
115.9
C12
R2
C5
O
C13
Figure 2: The geometry structure of XT derivatives in the ground state for compound
III. The arrows show the dipole moment directions.
25

Figure 3: The absorption and luminescence spectra of XT in n-hexane solvent. The solid
lines correspond to experimental results for absorption (blue) and luminescence (red). The
vertical lines show the results of calculation. The numbers of C.I. are listed in Figure.
26

Figure 4: The energy level diagram of XT for C.I. = 9. The arrows show only the
transitions with considerable oscillator strength. The numbers correspond to the values
of the non-normalized oscillator strength as MOPAC 2016 produces.
27

Figure 5: The absorption and luminescence spectra of DXT in n-hexane solvent. The
solid lines correspond to experimental results for absorption (blue) and luminescence (red).
The vertical lines show the results of calculation for C.I. = 9.
28

Figure 6:
The energy level diagram of the DXT for C.I. = 9. The left and right
columns correspond to the cis- and trans-conformers. The arrows show only the tran-
sitions with considerable oscillator strength. The numbers correspond to the values of the
non-normalized oscillator strength as MOPAC 2016 produces.
29