Spectral Dynamics of Nitro Derivatives of Xanthione in Solutions

Article abstract of DOI:10.1021/acs.jpca.8b11146

Nitro derivatives of xanthione, 2,7-dinitro-9H-xanthene-9-thione and 2,4,7-trinitro-9H-xanthene-9-thione, have been first synthesized and their stationary and transient spectra have been measured. The stationary spectra show that the attachment of the nit

Full text of DOI:10.1021/acs.jpca.8b11146

Subscriber access provided by Macquarie University
A: Spectroscopy, Molecular Structure, and Quantum Chemistry
Spectral Dynamics of Nitro-Derivatives of Xanthione in Solutions
Stanislav L. Bondarev, Sergei A. Tikhomirov, Oleg V. Buganov, Valery N. Knyukshto,
Nikolaii A. Galinovskii, Roman G. Fedunov, Svetlana S. Khohlova, and Anatoly I. Ivanov
J. Phys. Chem. A, Just Accepted Manuscript • DOI: 10.1021/acs.jpca.8b11146 • Publication Date (Web): 07 Feb 2019
Downloaded from http://pubs.acs.org on February 12, 2019
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 35
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
9
Spectral Dynamics of Nitro-Derivatives of
Xanthione in Solutions
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
Stanislav L. Bondarev,† Sergei A. Tikhomirov,† Oleg V. Buganov,† Valeri N.
Knyukshto,† Nikolaii A. Galinovskii,‡ Roman G. Fedunov,¶
Svetlana S. Khokhlova,¶ and Anatoly I. Ivanov∗,¶
†B. I. Stepanov Institute of Physics, National Academy of Sciences of Belarus, Prospect
Nezavisimosti 68, Minsk BY-220072, Republic of Belarus
‡Institute of Chemistry of New Materials, National Academy of Sciences of Belarus, 36,
Fr. Skaryna Street, Minsk BY-220141, Republic of Belarus
¶Volgograd State University, University Avenue 100, Volgograd 400062, Russia
E-mail: Anatoly.Ivanov@volsu.ru
1
ACS Paragon Plus Environment

The Journal of Physical Chemistry
Page 2 of 35
1
2
3
4
Abstract
5
6
7
8
Nitro-derivatives of xanthione, 2,7-dinitro-9H-xanthene-9-thione and 2,4,7-trinitro-
9H-xanthene-9-thione, have been first synthesized and their stationary and transient
spectra have been measured. The stationary spectra show that the attachment of the
nitro-groups to the xanthione scaffold leads to strong quenching of S2 → S0 fluores-
cence and the decrease of the oscillator strength of the S2 ← S0 electronic transition.
Analysis of the transient absorption spectra uncovers the ultrafast stimulated emission
quenching from the second excited state, S2, in the both derivatives. A kinetic scheme
has been suggested to rationalized the complex spectral dynamics of transient absorp-
tion signal. The kinetic scheme is deduced from the analysis of the transient spectra
and supported by the quantum-chemical calculations which predict existence of a dark
state and S2 state splitting into two close levels. The ultrafast transitions between
S2 state sublevels and transition into the dark state play crucial role in spectral dy-
namics. These new features discovered in the nitro-derivatives of xanthione distinguish
essentially their spectral dynamics from that observed in xanthione.
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
Considerable interest of researchers in thioketones during past decades is associated with
their remarkable spectral and photophysical properties.1–6 Spectral properties are closely
related to the electronic structure of their highest excited electronic states.3 For the thioke-
tones in solutions experimentalists can distinguish separate S1, S2, and S3 absorption bands,
as well as direct absorption T1 ← S0, thermally activated delayed fluorescence from S1 state,
fluorescence and phosphorescence from S2 and T2, correspondingly.7 The essential feature of
thioketones is strong fluorescence from the second excited state and negligible one from the
first excited state that is a bright example of the Kasha’s rule violation.2,5,8,9 The latter is
caused by the large energy gap between S2 and S1 states10 and appears in S2 long lifetimes
of thioketones τ ∼ 10−9 − 10−11 s.1,4,7
2
ACS Paragon Plus Environment

Page 3 of 35
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
9
The second peculiarity of thioketones is the dependence of the phosphorescence quantum
yield on the excitation wavelength. The triplets quantum yield, φT , for xanthione (XT)
was shown to be close to unity upon the excitation into the first excited state, S1, while
it is smaller than unity (φT = 0.5 − 0.6) in the case of excitation into S2 state.11 It is
a manifestation of the breakdown of the classical Vavilov-Kasha’s law9,12 stating that the
quantum yield of luminescence is independent of excitation wavelength. Besides, Vavilov-
Kasha’s law violation was observed for XT derivative, 4H-1-benzopyran-4-thione,13 for which
the S2 → S0 fluorescence yield, φfl, under excitation into the absorption band S4 ← S0
decreases by 25% comparing with that at the excitation of S2 ← S0 band.
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
Input of large excess of vibrational energy in highly excited singlet states is often accom-
panied by photochemical processes of electron and proton transfer or isomerization, which
do not always result in the irreversible formation of products. In the case of ineffective
transitions to the triplet states, the electron excitation energy non-radiatively deactivates
into the environment. Such non-radiative processes are named “induced internal conversion”
Sn →S0 (n = 1, 2, 3, ...),14,15 and also known as transition to the “dark state”.16 For exam-
ple, the quantum yield of S2 → S0 fluorescence in XT is very small (φfl ∼ 0.05)7 and the
decrease of the quantum yield φT upon excitation in the S2 state comparing to excitation
into S1 state is explained by the appearance of an effective S2 → S0 internal conversion,3,11
while in DNA nucleobases (adenine, thiamine, guanine, etc.) practically the entire electron
excitation energy deactivates through the internal conversion S1 → S0 with rate constant
kic ∼ 1012 s−1.14
Experimental and theoretical studies of the dynamics of highly excited singlet states
of xanthone16,17 and its derivatives18,19 have shown that upon the excitation of xanthone
molecule to the singlet S2 (π, π∗)-state the direct or indirect intersystem crossings to the
lower T1-state occur with a very large probability due to high rate kISC ∼ 1011 s−1. Such a
high efficiency of intersystem crossings in xanthone is caused by close arrangement of singlet
S1 (n, π∗), S2 (π, π∗) and triplet T1 (n, π∗), T2 (π, π∗) states, which can change their mutual
3
ACS Paragon Plus Environment

The Journal of Physical Chemistry
Page 4 of 35
1
2
3
4
5
6
7
8
9
arrangement depending on the polarity of the solvent.16,17 In particular a strong dependence
of the fluorescence yield on the polarity of the solvent is observed for thioxanthone.18 Thus,
in cyclohexane fluorescence yield, φfl, equals 0.0002, and in 2,2,2-trifluoroethanol the yield is
0.46, and the fluorescence spectrum exhibits noticeable solvatochromism.18 At present there
are two conceptions of the energy deactivation mechanism of the S2 state in thioxanthone.
The first one presents the mechanism of S2 deactivation as the process S2(π, π∗) → T1(π, π∗)
which lasts 4.0 ps. The second conception supposes that the S2 quenching is the two-
stage process with lifetime constants τ1 = 0.4 ps [(S2(π, π∗) → S1(n, π∗)] and τ2 = 4.0 ps
[S1(n, π∗) → T1(π, π∗)].18
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
A strong effect on the probabilities of non-radiative processes in another xanthone deriva-
tive, 2,7-diaminoxanthone,19 is rendered by the nature of the substituent (the electron-
donating properties of amino-group) as well as the proton-donating properties of solvents,
such as alcohols, which may form complexes with hydrogen bonds interacting with the amino-
groups of the 2,7-diaminoxanthone molecule, followed by quenching of its fluorescence.
The relatively long-lived second excited state in thioketones (S2 and T1) makes these
systems promising as elements of molecular electronic devices and, in particular, as light
switches. The existence of two or more long-lived states which can be selectively populated
with a photoexcitation pulse is necessary for the operation of such light switches. A pos-
sibility to design optical switches operating in the frequency range up to tens of GHz and
providing a high yield (up to 90%) of the charge separation process requires experimental-
ists and theorists to consider derivatives of thioketones with strong electron donor groups.20
Besides, it requires detailed studies of the mechanisms of pico- and femtosecond deactivation
processes of highly excited singlet and triplet states of heterocyclic compounds.6,20–22
A number of preliminary results obtained for XT and its derivatives6,19,21 allows outlining
the subsequent course of experimental and theoretical studies to construct a molecular model
with the desired frequency and charge parameters. In particular, for 1,3-dimethoxy-9H-
xanthene-9-thione (DXT)6 significant charge transfer (about 0.1 of the electron charge) was
4
ACS Paragon Plus Environment

Page 5 of 35
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
9
observed from the methoxy group to the thiocarbonyl group, and the fitting of transient
absorption spectra of XT and DXT showed an increase in the intersystem crossing probability
of S1 → T1 upon attaching electron-donor methoxy groups to the xanthione molecule.
Considering the results of the studies,6,19 we synthesized two nitro-derivatives of XT
with pronounced electron-acceptor properties: 2,7-dinitro-9H-xanthene-9-thione (DNXT)
and 2,4,7-trinitro-9H-xanthene-9-thione (TNXT). We provided complex experimental (sta-
tionary luminescent and transient absorption femtosecond measurements) and theoretical
(non-stationary density functional theory) studies of their photodynamics and the spectral-
kinetic characteristics of the new charge-transfer states arising for them. The main attention
was paid to clarifying all the channels of non-radiative deactivation of the electron excitation
energy in the XT molecule and its nitro-derivatives, on the basis of which the energy scheme
of the electronic transitions of the proposed optical switch can be constructed. In the article
we discuss in detail only the DNXT properties while a major part of information on TNXT
is presented in the Supporting Information.
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 aims of this work are: (i) synthesis of DNXT and TNXT, (ii) study of its spectral
properties including the spectral dynamics, (iii) identification of the kinetic scheme of elec-
tronic transitions to fit the theoretical spectral dynamics to that measured for DNXT, (iv)
estimation of the rate constants of electronic transitions which are triggered by an excitation
of the fluorophore to the second excited state.
Experimental and Theoretical Methods
Xanthione and its nitro-derivatives, DNXT and TNXT, have been transformed from cor-
responding ketones by Lawesson’s reagent using methods described in ref 23. Obtained
thioxanthones have been purified using standard techniques: HPLC and thin-layer chro-
1
matography. Their purity have been measured by IR, H, 13C NMR spectroscopy and gas
chromatography. The detailed description of DNXT and TNXT synthesis and their purity
5
ACS Paragon Plus Environment

The Journal of Physical Chemistry
Page 6 of 35
1
2
3
4
checking are provided below.
5
6
7
8
Xanthone (1.57 g, 8.00 mmol) and Lawesson’s reagent (1.63 g, 4.04 mmol) have been
dissolved in dry benzene (15 mL) and boiled for 2 h with stirring. The reaction has been
monitored by thin layer chromatography (eluent petroleum ether/diethylether 19:1). The
pre-dried silica gel (10 g) has been added to the reaction mixture. The benzene has been
evaporated under reduced pressure. The dry residue has been chromatographed (eluent
petroleum ether/diethyl ether 98:2). Finally 9H-xanthene-9-thione has been obtained with
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
◦
◦
1
the yield 1.25 g (74%). M.p. 155 C (lit. 157–178 C). H NMR (DMSO-d6, 500 MHz), δ:
7.48–7.51 (2H, m, H5, H7), 7.69 (2H, d, J 6.8 Hz, H1, H8), 7.92–7.95 (2H, m, H4, H5), 8.59
(2H, d, J 6.8 Hz, H3, H6). 13C NMR (DMSO-d6, 125 MHz) δ: 116.2, 125.9, 128.7, 129.6,
136.6, 150.6, 204.5 (C=S). EIMS (m/z): 212.03 (M+).
2,7-Dinitro-9H-xanthen-9-one (0.1 g, 0.3 mmol) and Lawesson’s reagent (0.1 g, 0.3 mmol)
have been dissolved in dry toluene (1 mL) and boiled for 2 h with stirring. The reaction has
been monitored by thin layer chromatography (eluent petroleum ether/diethylether 19:1).
The pre-dried silica gel (1 g) has been added to the reaction mixture. The toluene was evapo-
rated under reduced pressure. The dry residue has been chromatographed (eluent petroleum
ether/diethyl ether 98:2). Finally 2,7-dinitro-9H-xanthene-9-thione has been obtained with
◦
yield 0.074 g (82%). M.p. 193–197 C. IR (KBr, ν cm−1): stretch symmetric and asym-
metric frequencies of NO2 groups are at 1344 and 1540, stretch frequencies of C=S group at
1
1252–1182. H NMR (CDCl3, 500 MHz), δ: 7.68 (2H, d, J 9 Hz, H4, H5), 8.51 (2H, dd, J1 9
Hz, J2 3 Hz, H3, H6), 9.47 (2H, d, J 3 Hz, H1; H8). 13C NMR (CDCl3, 125 MHz) δ: 120.0,
123.0, 127.0, 129.9, 140.4, 156.6, 204.8 (C=S). EIMS (m/z): 302.00 (M+).
2,4,7-trinitro-9H-xanthen-9-one (0.1 g, 0.3 mmol)) and Lawesson’s reagent (0.1 g, 0.3
mmol) have been dissolved in dry toluene (1 mL) and boiled for 2 h with stirring. The reac-
tion has been monitored by thin layer chromatography (eluent petroleum ether/diethylether
19:1). The pre-dried silica gel (1 g) has been added to the reaction mixture. The toluene
was evaporated under reduced pressure. The dry residue has been chromatographed (eluent
6
ACS Paragon Plus Environment

Page 7 of 35
The Journal of Physical Chemistry
1
2
3
4
petroleum ether/diethyl ether 98:2). Finally 2,4,7-trinitro-9H-xanthene-9-thione has been
5
6
◦
obtained with yield 0.1 g (96%). M.p. 201–202 C. IR (KBr, ν cm−1): stretch symmetric
7
8
and asymmetric frequencies of NO2 groups are at 1340 and 1542, stretch frequencies of C=S
9
1
group at 1252–1184. H NMR (CDCl3, 500 MHz), δ: 7.79 (1H, d, J 9 Hz, H5), 8.67 (1H,
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
dd, J1 9 Hz, J2 3 Hz, H6), 9.17 (1H, d, J 3 Hz, H8), 9.40 (1H, d, J 3 Hz, H1), 9.64 (1H, d,
J 3 Hz, H3); 13C NMR (CDCl3, 125 MHz) δ: 120.0, 123.4, 127.0, 129.1, 129.9, 130.8, 136.9,
140.4, 141.3, 150.3, 156.6, 204.8 (C=S). EIMS (m/z): 346.98 (M+).
Cary-500 spectrophotometer (USA) has been utilized to register absorption spectra of
XT and DNXT in various solvents. The solvents (n-hexane, acetonitrile) were produced
by Merck (Sigma-Aldrich). Determination of the molar extinction coefficients has been
conducted by weighting of three samples of thioketone which have been dissolved later in a
certain volume of acetonitrile. The oscillator strengths have been estimated in accordance
with ref 19. A large-aperture setup has been used to detect the luminescence spectra in
a wide spectral range from 200 nm to 1000 nm. The details of a stationary and pump-
probe measurement setups have been described elsewhere.6,19,24–26 A pump signal has been
generated by the second harmonic of Al2O3:Ti3+ laser (395 nm) with the energy in a range
10 ∼ 30 µJ. Sample has been prepared in 2 mm thick cell with dye concentration in a range
of 10−3 and 10−5 M. At the dye concentrations of 10−3 and 10−5 M the absorption and
fluorescence spectra are close to each other. The probe pulse being transmitted through the
water achieves white continuum ranging in 400 – 1100 nm from the fundamental laser beam
with λ = 790 nm. A polychromator with Hamamatsu S7031 – 1007S back-thinned type
CCD area image sensor had recorded the signal of probe and reference pulses which then
has been processed with special software on PC. The sensor has improved sensitivity in UV
region and high signal/noise ratio. The time-dependent signal of absorbance ∆D(λ, ∆t) is
ꢀ
ꢁ
0
0
˜ ˜
˜
˜
determined as ∆D(λ, ∆t) = lg T0/T where T = Eprob/Eref and T0 = Eprob/Eref are the
energy ratios of the probe and reference pulses passed through the sample with and without
preceding pump, respectively. The correction of transient spectra accounts for the group
7
ACS Paragon Plus Environment

The Journal of Physical Chemistry
Page 8 of 35
1
2
3
4
velocity dispersion of the probe pulse.
5
6
7
8
The GAMESS-US27,28 software package has been used for quantum chemical simulations.
The optimized structures of DNXT and TNXT have been calculated in the approximation
of an isolated molecule in the gas phase at DFT level. The density functional B3LYP (key-
word DFTTYP=B3LYP in $CONTROL) in the 6-31G∗+ basis set (keywords GBASIS=N31
NGAUSS=6 NDFUNC=1 DIFSPP=.TRUE. in $BASIS) has been used. The initial guess for
wavefunction and type of calculation has been chosen to be RHF (keyword SCFTYPE=RHF
in $CONTROL). The initial geometry has been C1 (without symmetry restrictions). The
geometry optimization in the ground and excited states has been carried out by the gradi-
ent method. The convergence criteria to end the optimization process has been set to 10−4
(keyword OPTTOL=0.0001 in $STARTPT). The solvent effect on the structure and spectra
has been accounted for within polarized continuum model (PCM). The effect of following
solvents have been studied: n-hexane (the static dielectric permittivity d = 1.890, the re-
fractive index n2 = 1.375, and the solvent radius, Rsolv = 1.5, keywords SOLVENT=input
EPS=1.890 EPSINF=1.375 RSOLV=1.5 in $PCM) and acetonitrile (d = 36.6, n2 = 1.807,
Rsolv = 1.5, keywords SOLVENT=input EPS=36.6 EPSINF=1.807 RSOLV=1.5 in $PCM).
The time-dependent density functional theory (TDDFT) has been applied to calculate the en-
ergies of electronic transitions (keywords TDDFT=excite in $CONTROL and NSTATE=30
IROOT=i SG1=.TRUE. in $TDDFT, where i denotes corresponding excited state). Because
of a very slow convergence of the geometry optimization cycles for excited state in solvents,
the calculations have been stopped at the tolerance parameter 5 · 10−4 (n-hexane) and 10−3
(acetonitrile).
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
Results and Discussion
The calculation results of energetic and electronic properties of 2,7-dinitro-9H-xanthene-9-
thione (DNXT) are listed in Tables 1 and 2. It should be noted that the solvent influence
8
ACS Paragon Plus Environment

Page 9 of 35
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
9
on the solute are calculated within the electrostatic interaction approximation (PCM). The
energies of the solute-solvent interaction in the excited states are estimated by using the
equilibrium electronic structure of the ground electronic state. The geometry and electronic
structure of DNXT in the ground state is presented in Figure 1, the information on TNXT
is presented in Supporting Information. The bond lengths, bond angles, and atomic charges,
indicated in the Figure 1, are calculated using the DFT/B3LYP method for the equilibrium
geometry of the ground state in the gas phase. The lengths of characteristic bonds in the
ground state of DNXT in two solvents, listed in the Table 1, were optimized using the same
method, and the TDDFT/B3LYP method for excited states. The DNXT geometry structure
in the ground state obtained in n-hexane and acetonitrile matches its gas phase geometry.
In the first excited electronic state, the DNXT geometry slightly differs from that listed in
Figure 1 for the pyran and benzene rings. Alterations in the bond lengths and bond angles
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
◦
˚
are ±0.01 A and ±2 , respectively. The bond length RS1−C2 has only a valuable deformation
upon the excitation.
Table 1: The energetic characteristics of DNXT in the ground and excited
electronic states obtained in the equilibrium geometry. Si, Max.grad. (in
Hartree/Bohr), Etoti, EEI, EHOMO, ELUMO (in eV), µ (in Db), RS1−C2, RC2−C3
,
˚
RC3−C4, RC4−C5, RC14−N16 (in A) are the i-th electronic state, the maximum gradi-
ent achieved during optimization, the total energy, the energy of the electrostatic
interaction of a solute with solvent, the energy of the highest occupied molec-
ular orbital, the energy of the lowest unoccupied molecular orbital, the dipole
moment, and the bond lengths, respectively.
No Si Max.grad.
Solvent
vacuum
Etoti
-37608.35
EEI EHOMO ELUMO
µ
RS1−C2 RC2−C3 RC3−C4 RC4−O7 RC14−N16
1
2
3
4
5
6
7
8
9
S0
S0
S1
1.0 · 10−4
1.0 · 10−4
1.0 · 10−4
-
-6.76
-3.66
-3.55
-3.73
-3.73
-3.73
-3.74
-3.45
-3.62
-3.63
6.64
7.87
5.07
9.82
8.86
8.86
9.97
6.86
6.07
1.66
1.66
1.74
1.70
1.74
1.75
1.66
1.74
1.70
1.70
1.70
1.47
1.47
1.43
1.44
1.43
1.43
1.47
1.43
1.44
1.44
1.44
1.41
1.41
1.42
1.42
1.43
1.43
1.41
1.42
1.42
1.42
1.42
1.36
1.36
1.36
1.33
1.34
1.35
1.36
1.36
1.35
1.34
1.36
1.47
1.47
1.47
1.45
1.46
1.45
1.47
1.46
1.43
1.43
1.44
n-hexane -37608.33 -0.24 -6.68
n-hexane -37606.68 -0.25 -6.51
n-hexane -37606.06 -0.26 -6.50
n-hexane -37605.32 -0.26 -6.49
n-hexane -37605.23 -0.26 -6.47
Ds1 1.0 · 10−4
S2
S3
S0
S1
5.5 · 10−4
8.9 · 10−3
1.0 · 10−4 acetonitrile -37608.20 -0.66 -6.60
1.0 · 10−4 acetonitrile -37606.19 -0.72 -6.47
Ds1 1.0 · 10−4 acetonitrile -37605.61 -0.75 -6.49
10 S2
2.7 · 10−4 acetonitrile -37605.65 -0.74 -6.49
-3.66 11.02
-3.60 5.72
11 Ds2 1.0 · 10−4 acetonitrile -37605.27 -0.72 -6.49
Here we use S1, S2, and S3 to denote the local excited states while the excited state
with zero or close to zero oscillator strength (Ds) are called ”dark states”. The dark states
9
ACS Paragon Plus Environment

The Journal of Physical Chemistry
Page 10 of 35
1
2
3
4
5
6
Table 2: The charges (in the units of the elementary charge) located on the
atoms of DNXT molecule. The numbers in the first column correspond to the
numbering given in the first column of Table 1.
7
8
9
No
1
2
3
4
5
6
7
8
S1
C2
C3
C4
C5
C6
O7
C8
C11 C12 C15 N17 O18 O19 H23
-0.26 0.12 1.83 -0.65 1.80 -0.59 -0.39 -0.57 0.12 -0.33 -0.51 -0.34 0.00 -0.01 0.27
-0.26 0.12 1.83 -0.65 1.80 -0.59 -0.39 -0.57 0.12 -0.33 -0.51 -0.34 0.00 -0.01 0.27
0.06 -0.22 0.21 0.33 0.21 0.33 -0.45 -0.23 -0.23 -0.38 0.13 0.32 -0.32 -0.26 0.25
0.23 -0.15 0.18 0.35 0.19 0.29 -0.41 -0.23 -0.23 -0.36 0.15 0.24 -0.45 -0.38 0.25
-0.32 0.22 1.77 -0.62 1.78 -0.51 -0.36 -0.56 0.20 -0.36 -0.63 -0.37 -0.01 -0.02 0.28
-0.31 0.25 1.78 -0.59 1.77 -0.53 -0.36 -0.58 0.18 -0.34 -0.64 -0.33 -0.02 -0.03 0.28
-0.35 0.15 1.96 -0.65 1.92 -0.59 -0.36 -0.58 0.18 -0.30 -0.70 -0.27 -0.07 -0.06 0.27
0.07 -0.23 0.22 0.33 0.22 0.33 -0.43 -0.23 -0.22 -0.38 0.14 0.34 -0.35 -0.30 0.25
0.28 -0.16 0.21 0.33 0.21 0.33 -0.40 -0.21 -0.22 -0.35 0.15 0.29 -0.38 -0.44 0.26
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
9
10 -0.38 0.24 1.88 -0.61 1.87 -0.53 -0.35 -0.58 0.26 -0.35 -0.73 -0.34 -0.06 -0.06 0.27
11 0.27 -0.16 0.21 0.33 0.21 0.33 -0.43 -0.22 -0.22 -0.34 0.15 0.29 -0.37 -0.43 0.26
appearance opens new channels of the electronic transitions such as internal conversion from
higher excited states. To simplify further analysis, transitions to two dark states are modeled
by a transition to one dark state. The analysis of the Mulliken charges on DNXT atoms (see
Table 2) has shown that molecular bonds in polar solvent (acetonitrile) is more polarized in
comparison with that calculated in the gas phase or apolar solvent (n-hexane). However, the
dipole moment direction in the first excited state matches its direction in the ground state.
Note that excitation of DNXT results in a considerable variation of the atomic charges on
the thiocarbonyl group (qS1, qC2). In the dark state the electronic structure of DNXT and
TNXT significantly differs from that in the locally excited state in both polar and apolar
solvents, which is associated with considerable charge transfer between the sulfur and both
nitrogen atoms. This conclusion is approved by the analysis of the ”lambda diagnostic”
procedure (see details in SI).29 It should be noted that quantum chemical calculations do
not predict any dark states in XT lying between S2 and S1 states.
The electronic absorption and luminescence spectra of XT in apolar and polar solvents:
n-hexane and acetonitrile are shown in Figure 2 A and B, correspondingly.
The absorption spectra are characterized by three vibronic bands in the UV region (280–
320 nm, 320–360 nm, and 360–410 nm), which according to the literature1–3 corresponds to
electronic transitions S4 ← S0, S3 ← S0 and S2 ← S0. The transitions S4 ← S0 and S2 ← S0
10
ACS Paragon Plus Environment

Page 11 of 35
The Journal of Physical Chemistry
1
2
3
4
-0.26
S1
5
6
7
8
0.27
0.27
0.00
0.00
H22
H23
O20
O19
1.66
1.08
1.08
121
123
-0.35
-0.41
C13
0.12
-0.34
N16
-0.33
125
O21
N17
117
C2
119
C12
122
119
114
-0.48
-0.01117
1.83
1.80
C14
-0.51
-0.01
120
C3
120
C5
1.41
122
C15
O18
9
117
122
117
122
119
1.4
1.41
1.4
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
119
122
0.12
-0.59
C6
-0.65
0.10
C11
C4
C10
119
119
120
-0.54 115
115
0.24
H27
-0.57
C9
0.24
H26
-0.39
C8
O7
119
121
1.08
1.08
0.22
0.22
H24
H25
Figure 1: The equilibrium structure of the DNXT molecule in the ground electronic state
calculated in the gas phase. All the parameters (charges, bond lengths, and angles) have
been also calculated in the gas phase.
are allowed (π, π∗)-transitions and their molar extinction coefficients are ε ≥ 104 M−1cm−1.
The low-intensity transition S0 ← S3 is forbidden in symmetry and it becomes allowed, for
example, in 1,3-dimethoxy-9H-xanthene-9-thione, in which the symmetry is lower than that
in the XT.6 The S0 ← S1 transition in the XT and its derivatives is strongly forbidden
and its molar extinction coefficient is estimated to be ε ' 20 M−1cm−1.2–5 Therefore, it
practically appears neither in the absorption spectra nor in the fluorescence spectra (Figure 2
A). Such an unusual fact for the organic compounds spectroscopy is due to the presence of
heteroatom of sulfur in the XT molecule. Its unshared pair of electrons participates in the
S0 ← S1 electron transition and, accordingly, this transition is (n, π∗)-transition. For the
same reason, phosphorescence is observed at room temperature, which spectrum in n-hexane
is characterized by the resolved vibronic structure (Figure 2 A). The distances between the
vibrational maxima in the fluorescence, S0 ← S2, and phosphorescence, S0 ← T1, spectra
are approximately the same and amount 1200 ± 40 cm−1. This frequency corresponds to
the frequency of the stretching vibration of the C = S bond,30 which indicates the main
role of the S atom in the formation of the vibrational structure of the electron spectra of
the XT. There is no visible vibronic structure in the XT spectra in acetonitrile, and a wide
band appears at λmax = 686 nm instead of pronounced structure with four maxima as it is
in n-hexane (see insert in Figure 2 B).
11
ACS Paragon Plus Environment

The Journal of Physical Chemistry
Page 12 of 35
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: The absorption (blue) and luminescence (red) spectra of XT molecule in n-hexane
(frame A) and acetonitrile (frame B) measured experimentally (solid lines) at 293 K and
calculated theoretically (histogram). Inserts in the figure show phosphorescence spectra of
XT in the region 650 – 900 nm.
12
ACS Paragon Plus Environment

Page 13 of 35
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
9
Table 3: The spectral characteristics of XT, DNXT and TNXT in n-hexane.
max
abs
Here λ
maxima, λ
and λmfl ax are the wavelengths of the absorption and fluorescence band
max
Phi
is the wavelength of the i-th vibronic phosphorescence band max-
Ph
0,0
imum, E0fl,0 and E are the energy gaps for the S2 ← S0 and T1 ← S0 electronic
transition between the 0-th vibronic sublevels, correspondingly, ∆EPh is the dif-
ference between the energy of the triplet levels in DNXT and XT, ∆E(S2 − T1)
is the energy gap between the S2 and T1 electronic levels.
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
max
max
max
max
Ph
λ
,
λmfl ax
nm
450
444
426
,
λ
,
λ
,
λ
,
E0fl,0
,
E
,
0,0
∆EPh
,
∆E(S2 − T1),
cm−1
abs
Ph1
Ph2
Ph3
No Molecule
nm
402
394
390
nm
673
711
732
nm
733
777
806
nm
cm−1 cm−1 cm−1
1
2
3
XT
DNXT
TNXT
798 22220 14860
854 22520 14060
885 23470 13660 1200
-
7360
8460
9810
800
Table 4: The spectral characteristics of XT, DNXT and TNXT in acetonitrile.
Here fabs and ffl are the absorption and fluorescence band oscillator strengths,
respectively, ε is the extinction coefficient, Eflmax and E
of the fluorescence and phosphorescence band maxima, respectively. The values
of the absorption band oscillator strengths, fabs, are measured while ffl are cal-
culated using the well known Einstein’s relations. The definitions of the other
observables are given in the caption of Table 3.
max
Ph
are the wavenumbers
max
abs
max
max
λ
,
λmfl ax
nm
,
ε,
λ
,
Eflmax
cm−1 cm−1 cm−1
0.54 0.48 686 21980 14580
,
E
,
Ph
∆EPh
,
∆E(S2 − T1),
cm−1
Ph
No Molecule
fabs
ffl
nm
406
402
402
M
−1cm−1
nm
1
2
3
XT
DNXT
TNXT
455
459
469
17000
8500
8700
-
7400
8370
8820
0.31 0.27 745 21790 13420 1160
0.32 0.27 800 21320 12500 2080
When nitro-groups are attached to the XT scaffold, the phosphorescence spectra in both
solvents are shifted to the long-wavelength region. However, fluorescence spectra demon-
strate a hypsochromic shift in n-hexane and bathochromic shift in acetonitrile (see Tables 3
and 4). Such changes in the spectral-luminescent characteristics of the thioketones, de-
pending on their structure and the polarity of the medium, testify that the energy gap
∆E(S2 − T1) increases as the redox potential of the thioketone molecule increases with the
number of nitro-groups attached. Changes in the spectral properties due to attachment of a
next nitro-group to the XT scaffold evidences on quenching of the luminescence that appar-
ently is caused by the interaction of nitro-groups with the conjugated thioketone system (or
in other words, caused by the charge transfer). The effect of charge transfer on the spectral
properties of DNXT and TNXT also manifests itself in a significant decrease in their molar
13
ACS Paragon Plus Environment

The Journal of Physical Chemistry
Page 14 of 35
1
2
3
4
5
6
7
8
9
extinction coefficients (about two times) and the oscillator strengths (1.7 times) as compared
to XT (See Table 4). This experimental fact apparently reflects the electron density trans-
fer from the conjugated system of the XT scaffold to electron-acceptor nitro-groups and a
corresponding decrease in its absorption capacity. The shift of the triplet level of the DNXT
and TNXT with respect to the XT, given in the column ∆EPh of Table 3 is of particular
interest. A large bathochromic shift of the T1 level has been observed in acetonitrile for the
TNXT (see Supporting Information), which value amounts 2080 cm−1. Quantum-chemical
calculation of the energy levels of the TNXT in the gas phase supports this observation (see
energy diagram S7 in the Supporting Information). Earlier we published data on the effect
of an electron-donor methoxy-group on the position of the T1-level in the XT molecule.6 In
n-hexane solution the T1-level of DXT is 850 cm−1 higher than the corresponding level in the
XT molecule. This result was interpreted in ref 6 as a convergence of the S1 and T1 levels in
the DXT molecule in comparison with the XT with a corresponding increase in the quantum
yield of the intersystem crossing. In the case of XT nitro-derivatives, the opposite effect
is observed, a decrease of the energy of the T1 level, which amounts 2080 cm−1 for TNTX
in acetonitrile. In order to elucidate the mechanism of non-radiative deactivation of the
electron excitation energy in this case, we carried out experimental and theoretical studies
of the femtosecond spectral dynamics of the XT and its nitro-derivatives in acetonitrile.
The absorption and fluorescence spectra of DNXT in n-hexane and acetonitrile measured
experimentally (solid curves) and calculated theoretically (vertical lines) are demonstrated
in Figure 3. The number of observed experimental bands in the spectral window considered
are less than the number of excited states. The latter results in some problems in labeling
of the excited states. To maximize the correspondence we labeled the lowest singlet state as
S1 which associates with the first weak but observable absorption band of XT and its nitro
derivatives: 570 nm (XT, f = 0.0009), 582 nm (DNXT, f = 0.0012), and 602 nm (TNXT,
f = 0.0013) in acetonitrile (not shown in Figure 3). Next two excited states having even
lesser oscillator strengths, hence, invisible in absorption spectrum are labeled as Ds1 and
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
14
ACS Paragon Plus Environment

Page 15 of 35
The Journal of Physical Chemistry
1
2
3
4
5
6
Ds2. The second absorption band having much larger oscillator strength associates with two
excited states S2 and S3.
7
8
9
In the absorption spectra calculated for both the solvents there are three well-defined
”doublets” (pairs of closely spaced lines of comparable intensity) which correspond to the
main experimental bands. In n-hexane they occupy the regions around 380 (S2 and S3 levels
in Table 1), 320, and 285 nm. The energy level scheme containing the singlet and triplet
states in the equilibrium geometry of the ground state in vacuum is demonstrated in Figure 4.
The transitions to the singlet excited states with nonzero oscillator strength are shown by
vertical arrows. The state with the energy about 2 eV is labelled as S1 in Figure 4 while the
states located about 3.0 eV are called the ”dark” states. In Figure 4 the oscillator strengths
shown in brackets and wavelengths of the transitions are listed above the arrows.
The transition energies calculated in the apolar (Figure 3 A) and polar (Figure 3 B)
solvents together with its oscillator strengths and those obtained experimentally for both the
absorption and fluorescence spectra of DNXT are in good agreement. As seen from Figure 3,
three main absorption bands have a doublet structure. Similar results were obtained earlier
for methoxy derivatives of XT that manifest itself in the broadening of the absorption band,
S2 ← S0.6
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 experimental and simulated transient absorption spectra (TA) of DNXT are pre-
sented in Figure 5. The spectra of TNXT are shown in Supporting Information. The
measurements of DNXT and TNXT spectra in various solvents at room temperature were
carried out in the wavelength domain from λ = 400 to 740 nm for the time window from 1
ps to 100 ps (See Supporting Information). Since there is an overlap of the bleach band and
pump pulse at 395 nm, the part of the TA spectrum including part of bleach band is not
shown in figures. Nevertheless, TA signal kinetics at the red edge of the bleach are given in
Figures 7 and S9 (Supporting Information). The all measurements of DNXT has revealed an
excited state absorption band with the maximum near to λmax ' 635 nm and gain spectra
in the region 420 − 520 nm. The positions of these bands are close to the band positions
15
ACS Paragon Plus Environment

The Journal of Physical Chemistry
Page 16 of 35
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 3: The absorption and luminescence spectra of DNXT in n-hexane (frame A) and
acetonitrile (frame B). The solid lines correspond to experimental results for absorption
(blue) and luminescence (red) measured at 293 K. The vertical lines show the oscillator
strengths obtained from the quantum-chemical calculations. The red lines at the frame A
corresponds to the transitions S2 → S0 and S3 → S0.
16
ACS Paragon Plus Environment

Page 17 of 35
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
4
3
2
1
0
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 4: The diagram of energies of the stationary states and transitions between them
obtained at DFT level for DNXT calculated in the gas phase. The arrows indicate the
maxima of the transition bands with nonzero oscillator strengths (shown next to the arrows
in parentheses). Their wavelengths are shown next to the arrows. The lifetime constants
returned from the fitting to the spectral dynamics are also presented.
observed for the transient spectra of XT in the n-hexane (see Figure 2 in ref 6). However,
there is an important difference between the transient spectra of DNXT measured in this
work and of XT measured earlier. The stimulated emission is strongly suppressed in DNXT
in the solvents of strong polarity (See Supporting Information). Apparently, the difference
is due to the appearance of a new absorption band of the excited state overlapping with
the stimulated emission band. The kinetics of the transient absorption signal of DNXT at
500 nm consist of two components which are similar to those in XT: the short component
with a lifetime about 20 ps and long-lived one exceeding 100 ps. Apparently, the first short
lifetime of 20 ps associates with the excited S2 state decay and long time constant, proba-
bly, characterizes rising the triplet-triplet Tk ← T1 absorption. Up to the moment of the
first measuring with a delay of 0.4 ps a considerable part of S3 state population deactivates
resulting in suppression of the stimulated emission signal.
The simulated transient absorption spectra of DNXT in different solvents have been
simulated on the approach elaborated in ref 31. This theory has already been approved for
17
ACS Paragon Plus Environment

The Journal of Physical Chemistry
Page 18 of 35
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 5: The experimental (dotted lines) and calculated (solid lines) transient absorption
spectra of DNXT in acetonitrile at 293 K. The simulated spectra are obtained from the
theoretical approach (7) – (11). All the energetic parameters are determined from the fitting
and listed in Table 5. Arrows indicate the maxima of the corresponding bands after relaxation
(see Table 5).
18
ACS Paragon Plus Environment

Page 19 of 35
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
xanthione6 and xanthone19 derivatives as well as for pyrylium phenolate.32 The theory allows
quantitatively and qualitatively reproducing of experimental transient absorption spectra.
Analysis of the experimental transient absorption spectra indicates that the total absorp-
9
tion signal
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
X
∆D(νe, νp, t) =
∆Di(νe, νp, t)
(1)
i
in the case of DNXT molecule can be decomposed on bleach, ∆DBL, stimulated emission,
∆DSE, k excited state absorption bands, ∆DESAk, and triplet-triplet absorption, ∆DTTA
Here ∆Di are the values of the optical transition probabilities:
.
Vp2i
m
e−ξ ξi
i
X
πτp
∆Di(νe, νp, t) = νpfi
Pn
(2)
(3)
P σi(t) n,m
m!
"
#
(δνmi − Qni(t))2
2σi2(τ)
× exp −
P(t)
where the index i may take one of the values: {ESAk, SE, BL, TTA}. P(t) is the population
of one electronic states: {S3, S2, S1, Ds1, T1}, σi is the time-dependent half-width of the
corresponding band:
σi2(t) = τp−2 + 2ErikBT − (2ErikBTX(t)/σe)2
and Qni is the coordinate of center-of-gravity of the corresponding band:
Qni(t) = 2EriX(t)(1 − δi,BL + δνnekBT/σe2)
(4)
(5)
Here we used the following notions: σe2 = τe−2 + 2ErSEkBT, δνne = νe + ∆GSE − ErSE
−
nΩhf, δνmESAk = νp + ∆GESAk − ErESAk − mΩhf, δνmSE = νp + ∆GSE + ErSE + mΩhf,
δνmBL = νp + ∆GSE − ErSE − mΩhf, δνmTTA = νp + ∆GTTA − ErTTA − mΩhf are the solvent
relaxation functions, where ∆Gi is the change of free energy for corresponding transition,
Eri and Erhi are the reorganization energies of solvent and high-frequency mode for i-th
19
ACS Paragon Plus Environment

The Journal of Physical Chemistry
Page 20 of 35
1
2
3
4
5
6
7
8
transition, respectively, excluding ESA band, Ωhf is the frequency of the high-frequency
mode for corresponding transition. δi,BL is the Kroneker’s symbol. Here all the quantities
Ω, ν, and δν are measured in the energy units.
9
The probability of the transition generating n vibrational quanta at the pump stage can
be represented in the following form:
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
"
#
n
πVe2τe e−ξ ξSE
(δνne)2
2σe2
SE
Pn =
exp −
(6)
σe
n!
P
where P =
Pn, ξi = Erhi/Ωhf is the Huang-Rhys factor, where τe, τp, νe, and νp are
the durations and the carrier frequencies of the pump and probe pulses, Ve, and Vp are
proportional to the matrix elements of the dipole moment of interaction with electromagnetic
field. kB and T are the Boltzmann constant and the temperature.
Since the intramolecular vibrational relaxation of high-frequency modes proceeds in the
femtosecond timescale,33 we assume that it is instant in comparison with the solvent vibra-
tional relaxation. The solvent relaxation function is
X
X
X(t) =
xi exp {−t/τi} ,
xi = 1
i
i
where the solvent relaxation is modeled in the terms of a few Debye modes whose weights
xi and relaxation times τi was reported in ref 33.
At first, the simplest scheme of transitions elaborated earlier for the description of tran-
sient spectra in XT6 (see Figure 6 showing the experimental and simulated transient spectra
of XT in acetonitrile), which does not involve the dark and S3 states, was used. A comparison
of the ratios of stimulated emission to ESA signals at a delay of 0.4 ps shows that the ratio
in DNXT is about 6 times larger than that in XT. Since the experimental data show that
the fluorescence oscillator strength in DNXT is 1.8 times smaller while the ESA signal is
only 1.14 times larger than that in XT, so that it is impossible to fit the transient absorption
spectra without an ultrafast S2 state depopulation. This fact may indicate the important
20
ACS Paragon Plus Environment

Page 21 of 35
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
9
role of the dark state (Ds1) and the doublet structure of the second excited state (S3, S2),
which is predicted by quantum-chemical calculations (see Figure 4). Given the similarity of
the structures of S2 and S1 states in XT and DNXT, one can expect that the rate constants
of the internal conversion S2 → S1 are also close to each other. The quantum-chemical
calculations (see Table S2 in the Supporting Information) confirm the conclusion that in
XT, DNXT and TNXT the oscillator strengths of the transitions from S1 to higher excited
states do not noticeably differ. The simultaneous ultrafast decay of both S2 and S3 states in
DNXT would lead to vanishing of ESA band in the transient absorption spectra (see Figure
5). Thus, only the transitions from one state of the doublet can proceed in ultrafast mode.
Since the ultrafast decrease of ESA is much smaller than the decrease of SE, we need to
assume that there are at least two ultrafast channels S3 → S2 and S3 → Ds1. Adding these
two transitions together with the internal conversion channel S2 → S0 to the simplest scheme
gives a new kinetic scheme which is used in the fit to the transition spectra in DNXT.
The kinetic scheme presented in Figure 4 describes the population state dynamics and it
can be written in the differential form:
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
dS3(t)
dt
S3(t) S3(t)
= −
−
(7)
τIC32
τIC3D
dS2(t)
dt
S2(t) S2(t) S2(t) S3(t)
= −
−
−
+
(8)
(9)
τIC21
τIC2D
τIC20
τIC32
dS1(t)
S1(t) S2(t)
= −
+
dt
τISC1
τIC21
dT1(t)
dt
S1(t)
=
(10)
(11)
τISC1
dDs1(t)
dt
S2(t) S3(t)
+
=
τIC2D τIC3D
where Si(t), T1(t), and Ds1(t) stand here for the populations of corresponding states, τIC32
,
τIC21, τIC3D, τIC2D, τIC20 and τISC1 are the time constants of the internal conversion and
intersystem crossing, respectively. The initial population of all states are zero except for
21
ACS Paragon Plus Environment

The Journal of Physical Chemistry
Page 22 of 35
1
2
3
4
5
6
7
8
S3(0) = 0.4 and S2(0) = 0.6. The latter are determined by the oscillator strengths of the
optical transitions and the frequency of the excited pulse since the splitting of the states S3
and S2 is pretty small and both states can be populated by single pulse.
9
The set of the parameters utilized in the calculations corresponds to the experimental
setup: kBT = 0.25 eV, Ωhf = 0.15 eV, ωe = 3.1 eV, τe = τp = 100 fs, Ve = 0.005 eV,
Vp = 0.015 eV. The vibrational frequencies have been obtained from the DNXT fluorescence
band structure. The relaxation time constants and weights of the solvent modes conform to
acetonitrile: τ1 = 100 fs, τ2 = 600 fs, x1 = 0.6, x2 = 0.4.
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
Table 5: The model parameters returned from fitting TA spectra of DNXT in
acetonitrile (see eq 2). fi, Eri, Erhi, ∆Gi, and Emax are the weight, the reorgani-
zation energy of the medium, the reorganization energy of intramolecular high
frequency vibrational modes, free energy change, and the energy of the maxi-
mum of the corresponding band after relaxation, respectively. All the energies
are given in eV.
No
1
2
3
4
i
fi
transition
P(t)
T1(t)
Ds1(t)
S1(t)
Eri Erhi ∆Gi Emax
0.12 0.09 -2.36 2.57
0.14 0.09 -2.28 2.51
0.12 0.05 -1.86 2.03
TTA 1.3 T1 → Tk
DSA 0.85 Ds1 → Sj
ESA2 0.9
S1 → Sj
ESA1 0.85 S2,3 → Sj S3(t) + S2(t) 0.11 0.05 -1.85 2.01
5
6
SE
BL
0.73 S0 ← S2 S3(t) + S2(t) 0.10 0.40 -2.88 2.38
0.4 1.0 − Ds1 0.10 0.40 -2.88 3.38
S2 → S0
Analysis of the simulation results pictured in Figure 5 shows that the negative TA signal
in the area around 500 nm predominantly consists of the stimulated emission S2 → S0. An
overlapping of the dark state absorption with stimulated emission bands is considerable at
all time delays from 0.4 to 100 ps. The triplet – triplet absorption rises and leads to a
positive signal in this region at the time delays from 10 to 100 ps. The TA dynamics is
governed by several processes: the internal conversion between singlet states, the singlet
– triplet intersystem crossing, and the transitions from the local singlet excited states (S2
and S3) to the singlet dark state. Figure 4 shows the minimal scheme including these
processes along with the time constants returned from the fitting. The dark state absorption
22
ACS Paragon Plus Environment

Page 23 of 35
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 6: The experimental transient absorption spectra of XT in acetonitrile measured at
293 K.
23
ACS Paragon Plus Environment

The Journal of Physical Chemistry
Page 24 of 35
1
2
3
4
5
6
7
8
9
band overlapping the stimulated emission at early time delays plays important role in the
suppression of stimulated emission at 0.4 ps. Due to the different initial population of S3 and
S2 states and fast transitions between the S3 state and both S2 and Ds1 states the fitting at
early time delays reproduces the experimental signal. The measurements performed at other
TA experiment including wavelength region from 300 to 550 nm reveal the bleach dynamics.
It is seen the bleach signal is unchanged at the time delays from 1 to 100 ps. So the dark
state does not decay at the time under consideration (100 ps) and decreasing of the negative
TA signal can be caused by increasing dark state absorption.
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 positive TA signal in the wavelength region from 580 to 700 nm is the excited state
absorption band including the absorption both the S3 and S2 states. The positive signal
at the red wing can be explained to the S3 and S2 absorption with the low energies of
transitions. The decreasing of the positive signal reflects the depopulation of the S2 state
with τIC21 = 17 ps time constant to be close to the same time constant in the XT. The minor
shift of the band maximum from 632 nm to 625 nm is caused by increasing of the S1 state
absorption. The irreversible intersystem crossing between the S1 and Ti states results in
decreasing the positive signal up to zero after the time delay exceeding 40 ps. The fitting
gives the intersystem transition time constant τISC1 = 26 ps. But the nearly white noise in
the region from 550 to 750 nm can distort the estimation of these time constants.
The kinetic scheme of the transitions is shown in Figure 4. It includes the dark state
and describes a lot of aspects of the TA dynamics in both the time and energetic domains.
The change in the initial population of the states S3 and S2 at the pumping stage allows
controlling the dynamics of the population of both the dark and triplet states. The excluding
of the dark state gives the kinetic scheme used to reproduce the TA signal of XT in n-hexane
where the intersystem time constant is smaller then presented here.
To demonstrate the effect of nitro-group on the excited state dynamics, let us compare
the spectral dynamics of DNXT and XT shown in Figures 5 and 6. The spectra in Figure 6
are close to that reported in ref 3. The XT transient spectra consist of three main parts:
24
ACS Paragon Plus Environment

Page 25 of 35
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 7: The kinetics of XT (Frame A) and DNXT (Frame B) measured experimentally
(points) and their best fit (lines) in acetonitrile.
25
ACS Paragon Plus Environment

The Journal of Physical Chemistry
Page 26 of 35
1
2
3
4
5
6
7
8
9
the negative band with a decay time about 14 ps (in the range 410 − 510 nm), the positive
absorption band with a decay time about 20 ps (in the range 540 − 740 nm), and the
positive long-lived triplet-triplet absorption band (in the range 410 − 520 nm). The decay
of the band in the range 410 − 510 nm well agrees with the time constant of the internal
conversion S2 → S1, τIC21 = 14 ps, obtained from the fit within the simplest scheme. The
time constant of intersystem crossing S1 → T1 is τISC1 = 18 ps, so there is a considerable
population of S1 state within the time interval 10 – 40 ps. Since the pikes of the ESA bands
from S2 and S1 are close each other, the decay of the band in the range 540 − 740 nm is
slower than the decay of SE.
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 addition of the third nitro-group to the DNXT molecule leads to the C2v symme-
try breaking, which results in significant changes in the TA spectra (see Figure S8 in SI).
First, the TA spectrum shifts to the red, so that the bleach signal becomes observable in
the experimental window. This agrees well with changes in stationary absorption spectra.
Second, the ESA spectra from the first and second singlet and triplet excited states are much
broader than the TA spectra in DNXT, apparently, because the symmetry breaking allows
for a series of new transitions. The general features of the spectral dynamics in TNXT are
similar to those observed in XT and DNXT. At the same time, the time constants of signal
decay at the characteristic wavelengths in TNXT are two or three times shorter than that
in XT and DNXT.
Figure 7 shows the kinetics of the transient absorption of XT and DNXT in acetonitrile
at the room temperature at a few characteristic wavelengths. One can see that the kinetics
at λ = 425 nm consists of two components with the time constants 2.4 and 14.9 ps in
DNXT (frame B in Figure 7), whereas in XT (Frame A) at λ = 410 nm it can be described
perfectly by single exponential function with time constant τ = 4.2 ps. There is also the third
long-lived component caused by triplet-triplet absorption. This wavelength corresponds to
the long-wave edge of the steady-state S2 ← S0 absorption band. The complex dynamics
at 425 nm results from the overlapping of the bleach with triplet-triplet absorption bands,
26
ACS Paragon Plus Environment

Page 27 of 35
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
9
depopulation of S2 state, and solvent relaxation. This is the reason of the difference between
the time constants obtained from fitting of signal dynamics at a specific wavelength and the
time constants returned from fitting of spectral dynamics in the total spectral range. The
approach adopted here is based on fitting the spectral dynamics in the entire window of the
wavelengths, which returns from the fitting the true time constants of decay and population
of electronic states involved in the photochemical process. Just these time constants are in
use in chemical kinetics, while the time constants obtained from fitting the signal dynamics
at a certain wavelength are indirectly related to the kinetic characteristics of the electronic
transition.
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
Conclusions
The current research has uncovered ultrafast fluorescence quenching of the second excited
state in DNXT occurring on shorter timescale than that of the time resolution of the setup
exploited. This follows from a comparison of the stimulated emission signals of DNXT and
XT at early times. The stimulated emission signal of DNXT is about 3.5 times less than
that in XT at the delay of 0.4 ps. Since the fluorescence oscillator strength of the stimulated
emission in DNXT is only 1.8 times less than that in XT, the difference can not be explained
by a decrease of the oscillator strength. The discrepancy is rationalized in terms of ultrafast
depopulation of a new state, S3, located slightly higher than the second excited state. The
decay of this state results in ultrafast population of both the lower component of the doublet,
S2, and the dark state, Ds1, with comparable rates. The decay of the lower component of
the doublet S2 leads to a slower decay of the stimulated emission with a time constant close
to that of the decay of ESA which is about 20 ps.
On the base of the doublet structure of the second excited state including the S2 and
S3 states along with a dark state, revealed by quantum chemical calculations, as well as S1,
and T1 states, a scheme of energy levels and transitions between them has been constructed,
27
ACS Paragon Plus Environment

The Journal of Physical Chemistry
Page 28 of 35
1
2
3
4
5
6
7
8
9
which is used to fit the TA spectra. Due to the fact that the doublet, S2 and S3, of DNXT
is populated by excitation pulse that is followed by a rapid transfer of population from S3
to S2 and the dark state, it is possible to achieve a slow decrease of the intense ESA band
in the vicinity of 630 nm. In the scheme, the higher level of the doublet is non-radiatively
coupled with S2 and the dark state, while S2 is connected with S1 and Ds1 states. The
ultrafast decay channel, S3 → Ds1, and slower channels, S2 → S1 and S2 → Ds1, explain the
two stage nature of the fluorescence decay. At the same time, the transition S2 → S1 occurs
at a rate close to that in XT. In the scheme, the dark state decay is neglected, since Ds1,
in contrast to S2 and T1 states, is characterized by a considerable charge transfer from XT
scaffold to nitro-groups that can be the reason why the dark state is weakly coupled to S2
and T1 states.
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 doublet structure of the second excited state has been predicted by quantum-chemical
calculations not only for DNXT but also for 2,7-diaminoxanthone.19 Here a fairly complex
scheme of electronic transitions in excited DNXT has been suggested. The scheme involves
two relatively long-lived excited states, the dark singlet and lower triplet states. The popu-
lations of these states can be controlled by changing the carrier frequency of the pump pulse,
since the change leads to a change in the initial populations of the S3 and S2 states, which,
in turn, leads to a change in Ds1 and T1 populations. Supposing the scheme is true, XT
derivatives like DNXT can be promising molecules for design of optical molecular switches.
For further verification of the proposed electronic transition scheme, additional experimental
studies are needed, for example, the study of the effect of variation of carrier frequency of
the exciting pulse within the absorption band on the spectral dynamics. Such a variation
should change the populations of the states of the doublet, which will lead to significant
changes in the kinetics of transitions and TA dynamics.
28
ACS Paragon Plus Environment

Page 29 of 35
The Journal of Physical Chemistry
1
2
3
4
Supporting Information
5
6
7
8
9
Transient absorption spectra of DNXT in various solvents (polar and apolar), the results of
measurements and calculations of TNXT stationary absorption and fluorescence (n-hexane)
and transient absorption (acetonitrile) spectra, the structure of HOMO and LUMO of TNXT,
the details of ”lambda diagnostic”, the diagram of TNXT energy levels calculated in the gas
phase and the time-dependent signal kinetics of TNXT in acetonitrile obtained at certain
wavelengths, a comparison of the calculated oscillator strengths of the transitions for three
considered compounds (XT, DNXT and TNXT) with their energies of electronic transitions.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Acknowledgement
The study was performed by a grant from the Russian Science Foundation (Grant No. 16-
13-10122). This work was partially supported by the Belarusian Republican Foundation for
Fundamental Research (Grant No. P18R-202).
29
ACS Paragon Plus Environment