Solvatochromism of new substituted 4-[(E)-(4-nitrophenyl)diazenyl]phenolate dyes

Article abstract of DOI:10.1016/j.molliq.2019.112330

Studies on the reverse solvatochromism of dyes in pure solvents are of interest to gain a better understanding of the nature of this phenomenon. In this paper, a series of nine compounds, exhibiting phenol and 4-nitrophenyl groups linked with an azo conjugated bridge, was synthesized and characterized. These compounds were deprotonated to generate solvatochromic phenolate dyes. The azo bridge makes the phenol more acidic, in comparison with the related compounds exhibiting a CH[dbnd]CH or CH[dbnd]N bridge in their molecular structure, indicating the electron withdrawing character of the azo bridge. The solvatochromism of the dyes was investigated in 29 solvents and their intense colors in solution were attributed to π-π* electronic transitions, with an intramolecular charge transfer from the phenolate towards the 4-nitroaryl group. The probes exhibited a reversal in their solvatochromism, in a similar fashion to the behavior verified with related dyes presenting CH[dbnd]CH or CH[dbnd]N as the conjugated bridge; however, the azo dyes are much less sensitive to solvent polarity changes. Multiparametric analysis showed that the azo dyes are very sensitive to the acidity of the medium, but still to a lesser degree than the corresponding imine and stilbene dyes, due to the electronegativity of the azo group. The reverse solvatochromism of the dyes is dictated by the electron-donor phenolate and by the electron-acceptor 4-nitroaryl groups, independently of the conjugated bridge, of the additional nitro group and of the nature of substituents in the 2,6-positions of the phenolate groups. Thus, the solvatochromic reversal verified represents a committed step involving the interaction of the nitro and phenolate groups in the probe with the molecules of structurally different solvents.

Full text of DOI:10.1016/j.molliq.2019.112330

Journal Pre-proof
Solvatochromism
of
new
substituted
4-[(E)-
(4-nitrophenyl)diazenyl]phenolate dyes
Carlos E.A. de Melo, Celso R. Nicoleti, Leandro G. Nandi, Felipe
S.S. Schneider, Robson da S. Oliboni, Giovanni F. Caramori,
Vanderlei G. Machado
PII:
S0167-7322(19)35798-8
DOI:
https://doi.org/10.1016/j.molliq.2019.112330
MOLLIQ 112330
Reference:
To appear in:
Journal of Molecular Liquids
Received date:
Revised date:
Accepted date:
18 October 2019
12 December 2019
15 December 2019
Please cite this article as: C.E.A. de Melo, C.R. Nicoleti, L.G. Nandi, et al.,
Solvatochromism of new substituted 4-[(E)-(4-nitrophenyl)diazenyl]phenolate dyes,
Journal of Molecular Liquids(2020), https://doi.org/10.1016/j.molliq.2019.112330
This is a PDF file of an article that has undergone enhancements after acceptance, such
as the addition of a cover page and metadata, and formatting for readability, but it is
not yet the definitive version of record. This version will undergo additional copyediting,
typesetting and review before it is published in its final form, but we are providing this
version to give early visibility of the article. 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.
© 2020 Published by Elsevier.

Journal Pre-proof
1
Solvatochromism of New Substituted 4-[(E)-(4-
Nitrophenyl)diazenyl]phenolate Dyes
Carlos E. A. de Melo,[a] Celso R. Nicoleti,[a] Leandro G. Nandi,[a] Felipe S. S.
Schneider,[a] Robson da S. Oliboni,[b] Giovanni F. Caramori[a] and Vanderlei G.
Machado*[a]
[a]Departamento de Química, Universidade Federal de Santa Catarina, UFSC, CP 476,
Florianópolis, SC, 88040-900, Brazil
[b] Centro de Ciências Químicas, Farmacêuticas e de Alimentos, Universidade Federal
de Pelotas, UFPEL, Pelotas, RS, 96010-900 Brazil.
*E-mail: vanderlei.machado@ufsc.br

Journal Pre-proof
2
ABSTRACT
Studies on the reverse solvatochromism of dyes in pure solvents are of interest to gain a
better understanding of the nature of this phenomenon. In this paper, a series of nine
compounds, exhibiting phenol and 4-nitrophenyl groups linked with an azo conjugated
bridge, was synthesized and characterized. These compounds were deprotonated to
generate solvatochromic phenolate dyes. The azo bridge makes the phenol more acidic,
in comparison with the related compounds exhibiting a CH=CH or CH=N bridge in
their molecular structure, indicating the electron withdrawing character of the azo
bridge. The solvatochromism of the dyes was investigated in 29 solvents and their
intense colors in solution were attributed to * electronic transitions, with an
intramolecular charge transfer from the phenolate towards the 4-nitroaryl group. The
probes exhibited a reversal in their solvatochromism, in a similar fashion to the
behavior verified with related dyes presenting CH=CH or CH=N as the conjugated
bridge; however, the azo dyes are much less sensitive to solvent polarity changes.
Multiparametric analysis showed that the azo dyes are very sensitive to the acidity of
the medium, but still to a lesser degree than the corresponding imine and stilbene dyes,
due to the electronegativity of the azo group. The reverse solvatochromism of the dyes
is dictated by the electron-donor phenolate and by the electron-acceptor 4-nitroaryl
groups, independently of the conjugated bridge, of the additional nitro group and of the
nature of substituents in the 2,6-positions of the phenolate groups. Thus, the
solvatochromic reversal verified represents a committed step involving the interaction
of the nitro and phenolate groups in the probe with the molecules of structurally
different solvents.
Keywords: solvatochromic dyes; reverse solvatochromism; solvatochromism;
solute/solvent interactions; azo dyes; molecular dynamics

Journal Pre-proof
3
1. Introduction
Organic solvents are commonly used in many chemical procedures, such as in
the preparation of solutions used as media in chemical reactions, and in different
processes for the separation and purification of compounds. The medium influences the
rate and position of equilibrium of many chemical processes and is responsible for
important changes in the spectra for many chemical species.[1-4] This influence is
commonly attributed to the polarity of the medium, a term that considers all specific and
non-specific interactions involved in the overall solvation of a given solute by the
solvent molecules.[1-6] To simplify our understanding of the complex phenomenon of
solvation, the solvatochromic approach has been used in recent decades.[1-8] This
strategy involves the use of solvatochromic dyes, which are probes that are sensitive to
changes in the polarity of the medium. In other words, by varying the polarity of the
solvent the position and/or maximum intensity of the absorption (or emission) band of
the probe is modified.[2, 5-11] The most popular solvatochromic probe is (2,6-diphenyl-
4-(2,4,6-triphenylpyridinium-1-yl) phenolate) (1; Figure 1), known as Reichardt‘s
dye.[12] This pyridinium N-phenolate betaine dye is well known for its very
pronounced negative solvatochromic behavior, i.e., the vis solvatochromic band of the
probe is hypsochromically shifted with an increase in the polarity of the solvent. This
compound is the basis for the ET(30) empirical solvent polarity scale,[2, 8, 9, 12, 13]
defined by the expression: ET(30) = 28590/max, where max is the maximum
wavelength given in nm for the vis solvatochromic band for the dye in the solvent and
ET(30) represents the mole transition energy, in kcal mol-1, for this electronic transition.
Compound 2, Effenberger‘s dye,[14, 15] is a known example of a probe which exhibits

Journal Pre-proof
4
a positive solvatochromism, with the vis solvatochromic band of the probe undergoing a
bathochromic shift when the polarity of the solvent is increased.
Figure 1. Molecular structure of compounds 1-3.
Other classical example of a solvatochromic dye is Brooker‘s merocyanine
(3).[16] Solvatochromic studies involving this dye have revealed an intriguing behavior:
a reversal in the solvatochromism occurs in the region comprising low–polarity
solvents. This phenomenon has been the subject of debate for the past sixty years, with
many experimental and theoretical investigations, because a reversal of the
solvatochromism occurs in solvents in which the dye is poorly soluble, with the
possibility of self–aggregation in solution.[16-22]
Several theories have been proposed to explain the solvatochromic behavior of
merocyanines and other related dyes. In these compounds a solvent-dependent internal
charge transfer (CT) from an electron–donor to an electron–acceptor moieties (D and A)
has been verified, which are eventually connected to an unsaturated conjugated bridge.
Based on the classical cyanine–limit model [23], the solvatochromism in
merocyanines and related dyes can be described considering the contributions of a
neutral (N) and a zwitterionic (Z) resonance structure to the resonance hybrid of the
probe. The weight of the contribution of these resonance structures is influenced by the

Journal Pre-proof
5
type of D, A, and connector used, which is also responsible for changes in the D–A
strength of the end groups (SD–A) of the solvatochromic dye and consequently to
changes in the transition energy of the solvatochromic band of the probe.[23, 24]
According to this theory, two classes of compounds, I and II, can be identified.[24-26]
Class I dyes have charged D and A groups in their ground state, being predominantly
zwitterionic, while Class II dyes have neutral D and A moieties in their ground state.
Pyridinium N-phenolate betaine 1 and compound 2 are examples of Class I and Class II
dyes, respectively.
A more recent model, the generalized reversal model,[24] which complements
the cyanine–limit model, describes all types of solvatochromic behavior in
merocyanines (and related dyes) as particular cases of a general reversal behavior. It
addresses how modifications in the molecular structure of the probe displace the
position of the solvatochromic reversal along an extended solvent–polarity scale.
According to this theory, experimental solvatochromic plots can be fitted to a
hyperbolic function of a solvent–polarity parameter, allowing a direct comparison
between experimental and theoretical results.[24]
Thus, investigations on the systems which exhibit a clear experimental inverted
solvatochromism, in solvents of intermediate polarity, represent important contributions
to an in-depth analysis of these theories. These probes should be easily soluble in such
media, in order to inhibit the possibility of self-aggregation.[27-31]
In recent years we have investigated the solvatochromism of a family of dyes
comprised of a charged phenolate electron–donor and neutral nitroaryl electron–
acceptor moieties, which are connected through an unsaturated ethylene or
methyleneimino bridge.[27-29, 31, 32] Two examples of these compounds are the 4–
(4–nitrobenzylideneamino)phenolate 4b and the 4-(4-nitrostyryl)phenolate 5b (Figure

Journal Pre-proof
6
2), which are generated in solution by deprotonation of their corresponding phenols 4a
and 5a. These dyes exhibit an experimentally well-defined reverse solvatochromism in
N
solvents with medium polarity, such as DMSO (ET(30) = 45.1 kcal mol-1; ET = 0.444)
N
and DMA (ET(30) = 42.9 kcal mol-1; ET = 0.377). Their solvatochromism has been
investigated in many solvents of different polarity, from water to n–hexane, allowing
the influence of various structural features on the reverse solvatochromism of
merocyanines and other related solvatochromic dyes to be investigated. Thus, these
compounds can be considered ―hybrid cyanine‖ solvatochromic probes and are neither
Class I nor Class II dyes, although they exhibit structural features from both classes of
dyes.
Figure 2. Molecular structures of phenols 4a and 5a and ―hybrid cyanine‖ dyes 4b and
5b.
In the study reported herein, compounds 6a-14a (Figure 3) were synthesized
and characterized. These compounds were deprotonated to generate solvatochromic
dyes 6b-14b, and their solvatochromism was investigated in 29 solvents. The behavior
verified was compared with that exhibited by the families of related dyes represented by
dyes 4b and 5b, to verify the influence of the conjugated bridge (CH=CH, CH=N and
N=N) on the solvatochromism of the dyes.

Journal Pre-proof
7
Figure 3. Molecular structure of compounds 6a,b–14a,b.
2. Experimental
2.1. Materials and methods
All reagents used in the synthesis of the compounds were analytically pure and
were purchased from commercial sources (Sigma–Aldrich and Vetec). They were
employed without any further purification. All solvents were HPLC grade and were
purified following a methodology described in the literature.[33, 34] Deionized water
was used in all measurements, and it was boiled and bubbled with nitrogen and kept in a
nitrogen atmosphere to avoid the presence of carbon dioxide. The aqueous Britton-
Robinson buffer was prepared through the mixing of 1.2 g of boric acid, 1.2 mL of
glacial acetic acid, 1.1 mL of phosphoric acid and 497.7 mL of deionized water.[35]
The melting points were determined through DSC analysis using a Perkin
Elmer Jade-DSC and a Shimadzu DSC-50. The IR spectra were obtained with KBr
pellets on a FT Varian 3100 spectrometer. The NMR spectra were recorded in DMSO-d6
or acetone-d6 using 200–MHz Bruker AC–200F and 400–MHz Bruker Avance 400
spectrometers. Chemical shifts were recorded in ppm with the solvent resonance as the
internal standard and data are reported as follows: chemical shift, multiplicity (s =

Journal Pre-proof
8
singlet, d = doublet, t = triplet, dd = double doublet, m = multiplet), coupling constants
(Hz), and integration. High-resolution mass spectra were obtained with a Bruker
OTOF–Q II 10243 electrospray ionization-quadrupole time-of-flight mass spectrometer
(HR ESI-MS QTOF). UV–vis spectra were obtained with an Agilent Technologies Cary
60 spectrometer. The pKa values were determined using a benchtop pH meter Kasvi
K39-1014B and the pH measurements were recorded at 25.0 ºC.
2.2. Synthesis of the compounds 6a-9a and 11a-14a
Compounds 6a-9a and 11a-14a were synthesized according to a methodology
adapted from Keil et al.[36] In an Erlenmeyer flask, 0.4 mmol of the corresponding
amine (4-nitroaniline or 2,4-dinitroaniline), 7.2 mL of glacial acetic acid and 720 μL of
concentrated HCl were mixed in an ice bath (0 ºC). In parallel, in another Erlenmeyer
flask, also in an ice bath, 0.4 mmol of the corresponding phenol and 5 mL of dry
methanol were mixed. An aqueous solution of NaNO2 (0.4 mol L-1, 1 mL) was then
added to a solution of the amine. The reaction mixture was stirred for 30 min and then
poured into the mixture containing the phenol. In the next step, 50 mL of deionized
water was added to the resulting reaction solution to precipitate the product that was
collected by filtration. Compounds 6a-9a and 11a-14a were recrystallized from a
water:ethanol mixture (1:1 v/v).
4-[(E)-(4-Nitrophenyl)diazenyl]phenol (6a). Brown solid (yield: 40%). M.p.: 212.49 ºC.
IR (KBr,
̅
/cm–1): 3406 (OH); 1603 and 1584 (aromatic C=C); 1505 and 1336
(NO2); 1458 (N=N); 1138 (C-O). 1H NMR (200 MHz, DMSO-d6) δ/ppm: 10.59 (s, 1H);
8.41 (d, 2H, J = 9.0 Hz); 8.00 (d, 2H, J = 9.0 Hz); 7.89 (d, 2H, J = 8.8 Hz); 6.99 (d, 2H,
J = 8.8 Hz). 13C NMR (50 MHz, DMSO-d6) δ/ppm: 162.31; 155.51; 147.71; 145.36;

Journal Pre-proof
9
125.75; 124.95; 122.94; 116.19. HRMS m/z: 242.0556 [M-H]-, calculated for
-
C12H8N3O3 , 242.0560.
2,6-Dimethyl-4-[(E)-(4-nitrophenyl)diazenyl]phenol (7a). Orange solid (yield: 82%).
M.p.: 188.04 ºC. IR (KBr,
̅
/cm–1): 3512 (OH); 1591 (aromatic C=C); 1519 and
1
1340 (NO2); 1474 (N=N); 1340 (CH3); 1119 (C-O). H NMR (200 MHz, DMSO-d6)
δ/ppm: 9.43 (s, 1H); 8.40 (d, 2H, J = 9.0 Hz); 7.99 (d, 2H, J = 9.0 Hz); 7.63 (s, 2H);
13
2.26 (s, 6H). C NMR (50 MHz, DMSO-d6) δ/ppm: 158.44; 147.60; 145.01; 128.09;
124.98; 124.15; 122.80; 16.54. HRMS m/z: 270.08726 [M-H]-, calculated for
-
C14H12N3O3 , 270.08732.
2,6-Dichloro-4-[(E)-(4-nitrophenyl)diazenyl]phenol (8a). Orange solid (yield: 55%).
M.p.: 199.15 ºC. IR (KBr,
̅
/cm–1): 3416 (OH); 1603 and 1584 (aromatic C=C);
1
1523 and 1342 (NO2); 1454 (N=N); 1136 (C-O); 801 (C-Cl). H NMR (200 MHz,
13
DMSO-d6) δ/ppm: 8.32 (d, 2H, J = 8.6 Hz); 7.87 (d, 2H, J = 8.6 Hz); 7.83 (s, 2H). C
NMR (50 MHz, acetone-d6) δ/ppm: 156.69; 154.21; 150.39; 146.81; 126.22; 125.22;
-
124.86; 124.18. HRMS m/z: 309.9785 [M-H]-, calculated for C12H6Cl2N3O3 , 309.9781.
2,6-Dibromo-4-[(E)-(4-nitrophenyl)diazenyl]phenol (9a). Dark yellow solid (yield:
40%). M.p.: 198.31 ºC. IR (KBr,
̅
/cm–1): 3418 (OH); 1607 and 1582 (aromatic
1
C=C); 1521 and 1344 (NO2); 1466 (N=N); 1136 (C-O); 740 (C-Br). H NMR (200
MHz, DMSO-d6) δ/ppm: 8.44 (d, 2H, J = 8.8 Hz); 8.17 (s, 2H); 8.04 (d, 2H, J = 8.8
Hz). 13C NMR (50 MHz, DMSO-d6) δ/ppm: 155.66; 154.49; 148.15; 145.18; 127.64;
-
125.05; 123.10. HRMS m/z: 399.8751 [M-H]-, calculated for C12H6Br2N3O3 , 399.8750.

Journal Pre-proof
10
4-[(E)-(2,4-Dinitrophenyl)diazenyl]phenol (11a). Red solid (yield: 55%). M.p.: 191.79
/cm–1): 3471 (OH); 1603 and 1584 (aromatic C=C); 1519 and 1342
ºC. IR (KBr,
̅
(NO2); 1460 (N=N); 1142 (C-O). 1H NMR (200 MHz, DMSO-d6) δ/ppm: 10.87 (s, 1H);
8.92 (d, 1H, J1 = 2.3 Hz); 8.61 (dd, 1H, J1 = 2.3 Hz and J2 = 8.8 Hz); 7.90 (t, 3H, J2 =
8.8 Hz and J3 = 8.8 Hz); 7.01 (d, 2H, J3 = 8.8 Hz). 13C NMR (100 MHz, DMSO-d6)
δ/ppm: 163.54; 147.58; 146.98; 146.28; 145.60; 128.27; 126.67; 129.13; 119.98;
-
116.55. HRMS m/z: 287.0415 [M-H]-, calculated for C12H7N4O5 , 287.0411.
2,6-Dimethyl-4-[(E)-(2,4-dinitrophenyl)diazenyl]phenol (12a). Yellow solid (yield:
45%). M.p.: 224.71 ºC. IR (KBr,
̅
/cm–1): 3286 (OH); 1611 and 1595 (aromatic
1
C=C); 1511 and 1348 (NO2); 1325 (CH3); 1427 (N=N); 1136 (C-O). H NMR (200
MHz, DMSO-d6) δ/ppm: 8.90 (d, 1H, J1 = 2.4 Hz); 8.60 (dd, 1H, J1 = 2.4 Hz and J2 =
9.0 Hz); 7.87 (d, 1H, J2 = 9.0 Hz); 7.57 (s, 2H); 2.24 (s, 6H). 13C NMR (100 MHz,
DMSO-d6) δ/ppm: 147.57; 146.43; 145.42; 145.13; 128.31; 126.24; 125.07; 120.21;
120.04; 16.63. HRMS m/z: 315.0723 [M-H]-, calculated for C14H11N4O5, 315.0724.
2,6-Dichloro-4-[(E)-(2,4-dinitrophenyl)diazenyl]phenol (13a). Brown solid (yield:
43%). M.p.: 177.31 ºC. IR (KBr,
̅
/cm–1): 3404 (OH); 1599 and 1562 (aromatic
1
C=C); 1527 and 1348 (NO2); 1325 (CH3); 1474 (N=N); 1148 (C-O); 799 (C-Cl). H
NMR (200 MHz, DMSO-d6) δ/ppm: 8.81 (d, 1H, J1 = 2.3 Hz); 8.52 (dd, 1H, J1 = 2.3 Hz
and J2 = 8.9 Hz); 8.44 (d, 1H, J1 = 2.4 Hz); 8.19 (dd, 1H, J1 = 2.4 Hz and J2 = 8.9 Hz);
13
8.01 (s, 2H); 7.96 (s, 2Hz); 7.80 (d, 1H, J2 = 8.9 Hz); 7.77 (d, 1H, J2 = 8.9 Hz). C
NMR (50 MHz, CDCl3 and acetone-d6) δ/ppm: 143.87; 143.64; 129.83; 129.29; 128.04;
127.66; 127.16; 125.89; 124.42; 124.11; 123.84; 123.49; 122.68; 122.62; 122.42;

Journal Pre-proof
11
119.94; 119.04; 118.15; 117.19. HRMS m/z: 354.9729 [M-H]-, calculated for
-
C12H5Cl2N4O5 , 354.9632.
2,6-Dibromo-4-[(E)-(2,4-dinitrophenyl)diazenyl]phenol (14a). Orange solid (yield:
52%). M.p.: 210.70 ºC. IR (KBr,
̅
/cm–1): 3390 (OH); 1597 (aromatic C=C); 1527
and 1344 (NO2); 1462 (N=N); 1142 (C-O); 728 (C-Br). 1H NMR (200 MHz, DMSO-d6)
δ/ppm: 8.94 (d, 1H, J1 = 2.4 Hz); 8.63 (dd, 1H, J1 = 2.4 Hz and J2 = 8.8 Hz); 8.10 (s,
2H); 7.87 (d, 1H. J2 = 8.8 Hz). 13C NMR (50 MHz, DMSO-d6) δ/ppm: 149.78; 135.11;
129.29; 128.68; 128.33; 123.36; 120.24; 119.83; 119.75; 113.55. HRMS m/z: 444.8604
-
[M-H]-, calculated for C12H5Cl2N4O5 , 444.8601.
2.2. Synthesis of compound 10a
Compound 10a was synthesized according to a methodology adapted from
Rajagopal and Buncel.[37] In an Erlenmeyer flask, 0.4 mmol of 4-nitroaniline and 4 mL
of 4 mol L-1 HCl were mixed in an ice bath (0 ºC). Simultaneously, in another
Erlenmeyer flask, also in an ice bath (0 ºC), 0.4 mmol of 2,6-diphenylphenol, 10 mL of
dry ethanol, 0.3 mmol of sodium tetraborate (borax) and 5 mL of an aqueous NaOH
solution (1.7 mol L-1) were mixed. An aqueous NaNO2 solution (0.15 mol L-1; 2 mL)
was then added to the solution of the amine and the reaction mixture was stirred for 1
min and then poured into the mixture containing the phenol. The medium was acidified
until pH 6 and the solid product formed was collected by filtration. Compound 10a was
purified by silica gel column chromatography using an n-hexane:ethyl acetate mixture
(9:1 v/v) as the eluent.
2,6-Diphenyl-4-[(E)-(4-nitrophenyl)diazenyl]phenol (10a): Red-orange solid (yield:
20%). M.p.: 187.98 ºC. IR (KBr,
̅
/cm–1): 3520 (OH); 1588 (aromatic C=C); 1519

Journal Pre-proof
12
1
and 1331 (NO2); 1462 (N=N); 1138 (C-O). H NMR (200 MHz, acetone-d6) δ/ppm:
8.45 (d, 2H, J = 9.1 Hz); 8.12 (d, 2H, J = 9.1 Hz); 7.95 (s, 2H); 7.69 (m, 3H); 7.53 (m,
3H); 7.46 (m, 6H). 13C NMR (50 MHz, acetone-d6) δ/ppm: 138.87; 131.36; 130.99;
130.56; 129.79; 129.36; 129.22; 128.51; 126.65; 126.24; 124.61. HRMS m/z: 394.1187
[M-H]-, calculated for C24H16N3O3, 394.1186.
2.3. UV-vis measurements
The following procedure was typical for all measurements performed. A
7.0- mol L-1 stock solution of each compound was prepared in acetone. From this
stock solution, 3-12 L were transferred to twenty-nine 5 mL volumetric flasks. After
evaporation of the acetone, the probe was dissolved in the pure solvent, resulting in a
solution presenting a final dye concentration of 1.010-5 to 4.010-5 mol L-1. In order to
generate the deprotonated compounds, 5 L of an aqueous tetra-n-butylammonium
hydroxide solution (0.1 mol L-1) was added to each flask. The addition of this very
small amount of water did not change the position of the UV-vis band of the dye. The
bulky tetra-n-butylammonium ion has no influence on the UV-vis spectrum of the
anionic dye. The UV-vis spectra were recorded at 25.0 C, using a 1-cm square cuvette.
The maxima of the UV-vis spectra were calculated from the first derivative of the
absorption spectrum, with a precision of 0.5 nm, and the reproducibility of the max
values was verified through the determination of five spectra for each dye in each pure
solvent. The max values thus obtained were transformed into ET(dye) values, according
to the expression ET(dye) = 28590/max,[2, 3, 9] given in kcal mol-1 with a precision of
±0.1 kcal mol-1.

Journal Pre-proof
13
2.4. Determination of pKa
Stock solutions of compounds 6a-14a were prepared in acetone with a
concentration of 7.0- mol L-1. The aqueous Britton-Robinson buffer was placed in a
beaker in the amount appropriate to perform the study and the pH meter electrode was
then immersed in it. Aliquots of an aqueous NaOH solution (0.1 mol L-1) were added
under magnetic stirring, in order to adjust the pH. When the desired pH was reached,
1.5 mL of the solution was transferred to a cuvette, 9 L of the stock solution of the dye
was added to give a concentration of 4.0×10-5 mol L-1 and the UV-vis spectrum was
then collected. This procedure was carried out for all compounds at different pH values
and the absorbances were collected at the max values corresponding to the deprotonated
species. Plots of pH as a function of absorbance were constructed (Figure S47) and the
pKa values were determined from a sigmoidal fit of the curve using the Boltzmann
equation.
3. Calculation Methods
The multiparametric analysis was performed from non-linear regressions using
ORIGIN 8.5 software. Optimizations were carried out using the ORCA quantum
chemistry package, version 4.1.1,[38] at the ωB97X-D3/ma-def2-TZVP/SMD [39-44]
level of theory, with both water and n-hexane as implicit solvents. Resolution of identity
was used throughout the calculations.[45, 46] Vibrationally-resolved spectra were
obtained using the ESD module of ORCA.[47] Molecular dynamics simulations were
performed using the GROMACS package,[48] with the OPLS–AA force field [49] and
the SPC/E model [50] for water. Simulation boxes containing the solvents (water, DMF
and n-hexane) were thermalized at the NVT ensemble (298 K, τ = 0.1 ps) for 500 ps and
equilibrated at the NPT ensemble (1 bar, τ = 0.5 ps) for 2 ns. Atomic charges for

Journal Pre-proof
14
compounds 4b-6b were obtained with the Ante RED software,[51] with the GAMESS
package [52] at the HF/6-31G** level of theory.[53, 54] Relevant molecular mechanics
(MM) parameters are described in the Supporting Information (Tables S6 and S7).
4. Results and discussion
The synthesis of compounds 6a-14a was performed in two steps according to
procedures adapted from the literature [36, 37], as shown in Scheme 1. Firstly, the
nitroanilines were reacted with sodium nitrite in acidic medium, in an ice bath, to form
the corresponding diazonium salts. The coupling reaction products were obtained with
the addition of the corresponding phenols in dried methanol (compounds 6a-9a and
11a-14a) or in a basic ethanolic borax solution (compound 10a). Compounds 6a-9a and
11a-14a were recrystallized from a water:ethanol mixture (1:1 v/v) and compound 10a
was purified by silica gel column chromatography using a n-hexane:ethyl acetate
mixture (9:1 v/v) as the eluent. The products of the reactions were obtained in yields of
1
13
20-82%. Compounds 6a-14a were characterized using IR, H NMR, C NMR, HRMS,
and DSC techniques.
Scheme 1. Preparation of compounds 6a-14a.

Journal Pre-proof
15
o
Table 1 gives the pKa values for compounds 6a–14a at 25.0 C. Electron
withdrawing groups in the phenol moiety reduce the pKa while methyl and phenyl
groups reduce the acidity of the compounds. The data also show that, in general, the
presence of a second nitro group in the molecule reduces the pKa value. However, for
compounds 7a and 12a the presence of an additional nitro group leads to 12a 1.1 pKa
unit less acidic than 7a. These compounds are more acidic than 2,6-dimethylphenol
(pKa = 10.59 [55]), which is expected from the electron withdrawing substituents in the
para position of 7a and 12a. Thus, the fact that compound 12a is less acidic than 7a
suggests a stereoelectronic influence of the ortho nitro group on the acidity of 12a.
Table 1. The pKa values for compounds 6a–14a at 25.0 oC.
b
Dyea
6a
pKa
8.01 ± 0.01
8.83 ± 0.01
5.63 ± 0.04
5.68 ± 0.03
11.43 ± 0.03
7.63 ± 0.03
9.90 ± 0.01
5.28 ± 0.02
5.25 ± 0.03
7a
8a
9a
10a
11a
12a
13a
14a
a c (dye) = 4×10-5 mol L-1.
b Calculated by spectrophotometric method using aqueous Britton-Robinson buffer.
Buncel and Rajagopal verified that the azo bridge represents a powerful
acceptor group in merocyanine dyes.[56] The influence of the conjugated bridge on the
acidity of the nitrophenols can be observed by comparing the pKa of 6a (8.01± 0.01),
with those of the iso--electronic compounds 4a (10.57±0.01)[57] and 5a (10.79 ±

Journal Pre-proof
16
0.02).[58] These results demonstrate that the 4-(nitrophenyl)diazenyl group makes the
phenol more than 2.5 pKa units more acidic, which illustrates the electron withdrawing
character of the azo bridge.
4.1. Solvatochromism of probes 6b-14b
Figure 4A shows the solutions of dye 9b in four selected solvents, while
Figure 4B provides the corresponding UV-vis spectra, confirming that the compound is
solvatochromic. In methanol, the absorption band for this dye has a maximum at 486
nm, which is displaced to max = 519 nm in toluene, while in dimethoxyethane and
acetophenone the band maximum is shifted to 550 and 571 nm, respectively. The other
azo dyes studied exhibited a similar behavior (Figures S48 and S49), their intense
colors being a result of * electronic transitions, with an intramolecular CT from the
electron-donor phenolate towards the 4-nitrophenyl (or 2,4-dinitrophenyl) electron-
acceptor group.
Figure 4. (A) Solutions and (B) UV-vis spectra of dye 9b in (a) methanol, (b) toluene,
(c) dimethoxyethane, and (d) acetophenone.

Journal Pre-proof
17
Experiments were performed, according to a previously described
methodology[59], to discard the possibility of cis–trans isomerization as a source of the
observed spectral changes. No significant changes in the form and position of the
solvatochromic bands were verified (Figure S50). In addition, the possibility of dye
aggregation within the range of absorbance readings collected was also discarded as a
cause of the observed solvatochromic reversal. Plots of the absorbance reading as a
function of the dye concentration in ethyl acetate and methanol, two solvents of widely
different polarities, were all linear (Figures S51-S59).
Compounds 6b-14b are salts, which can form, in principle, ion pairs in
solution, especially in low polarity solvents, which would influence the absorption
spectra. Although studies performed previously with related dyes have demonstrated
that the possibility of the formation of ion pairs can be discarded,[27] some experiments
to verify this were conducted. Thus, the influence of the addition of increasing amounts
of tetra-n-butylammonium iodide on the UV-vis absorption spectrum of dye 13b in
toluene was investigated. The addition of this salt at a concentration of up to 1.9×10-3
mol L-1 did not cause a shift or changes in the form of the solvatochromic band of the
probe (Figure S60). In another experiment, compound 6a in toluene was deprotonated
to form dye 6b through the addition of KOH (Figure S61). The UV-vis spectrum was
obtained and the solvatochromic band of 6b had a max value of 424 nm, which is much
lower than the value obtained when tetra-n-butylammonium hydroxide is used (max
=
516 nm), due to the electrostatic interaction of K+ with the phenolate moiety of the dye.
It was verified that the addition of 18-crown-6, which is highly selective for K+, leads to
the spectrum obtained being exactly the same as that measured in the presence of tetra-
n-butylammonium hydroxide. Thus, the experiments allow us to conclude that the

Journal Pre-proof
18
presence of the tetra-n-butylammonium cation does not influence the solvatochromism
of the dyes.
The UV-vis spectra for each probe in the 29 solvents were used to obtain the
maximum wavelength (max) values for the solvatochromic bands, which were required
for the calculation of the corresponding transition energies ET = 28590/max, in kcal
mol-1 (Table 2).
Table 2. ET(30) and ET(dye) values for compounds 6b-14b in 29 solvents.
ET(30)a
ET(6b)b ET(7b)b ET(8b)b ET(9b)b
ET(10b)b
ET(11b)b ET(12b)b ET(13b)b
ET(14b)b
Solvent
30.9
31.0
33.9
34.5
37.4
38.1
38.2
39.1
40.6
40.7
41.3
42.2
42.9
43.2
43.3
45.1
45.6
47.1
47.7
48.1
48.4
49.1
49.7
50.4
50.7
51.9
55.4
56.3
63.1
58.1
58.3
55.4
53.3
49.3
50.2
49.9
52.3
49.4
50.7
50.4
49.5
47.4
47.7
51.5
47.8
50.6
52.8
55.5
54.5
53.8
55.2
55.6
55.5
55.6
56.9
59.2
58.4
59.7
52.5
52.9
49.7
49.7
45.6
46.5
45.8
47.0
45.2
46.4
46.3
46.1
45.3
45.5
47.7
45.2
46.8
49.0
50.2
49.7
46.5
50.1
50.4
49.7
50.3
51.3
53.5
53.2
55.3
58.0
58.1
56.3
54.6
52.1
53.1
52.3
54.1
51.4
53.5
53.0
52.1
50.5
51.3
53.6
51.1
53.3
54.0
55.6
55.4
54.7
55.5
55.8
55.4
56.1
56.7
58.9
58.9
60.4
57.5
58.2
55.1
53.7
51.3
52.6
52.0
53.1
50.1
51.9
51.8
51.5
50.6
51.1
53.8
51.1
52.9
53.9
55.4
55.5
54.8
56.0
55.7
55.5
56.4
56.7
58.8
58.8
61.0
51.7
52.0
49.4
48.4
45.8
46.9
46.3
47.6
44.8
46.6
46.3
46.1
45.2
45.5
48.0
45.5
47.2
48.6
49.6
49.7
49.1
49.8
49.9
49.5
50.2
50.9
52.8
52.4
55.1
51.8
52.1
50.2
48.5
47.1
47.5
47.2
47.5
46.4
47.2
46.9
47.5
47.0
47.1
48.1
46.9
48.0
48.4
50.1
50.2
49.1
49.7
50.0
49.9
50.4
51.2
54.0
53.2
55.7
47.1
47.3
46.3
46.9
46.3
47.3
46.4
45.9
45.4
45.7
45.7
46.6
46.0
46.1
46.0
45.7
46.8
46.0
46.4
46.4
46.4
46.6
46.7
46.0
46.8
47.3
48.5
47.9
49.6
52.4
52.9
51.8
50.6
50.7
48.5
48.7
49.1
47.0
47.9
47.9
48.4
47.8
48.2
49.5
48.0
49.3
50.3
51.2
51.2
50.1
51.0
51.2
51.6
50.7
52.7
54.4
54.5
56.8
52.0
52.4
50.4
49.5
47.5
48.7
48.1
48.6
46.9
47.8
47.5
48.5
48.1
48.2
49.0
48.1
49.1
49.4
50.4
50.4
50.0
50.6
50.9
50.5
51.2
52.0
54.3
54.7
57.0
Cyclohexane
n-Hexane
Toluene
Ethyl ether
THF
Ethyl acetate
Dimethoxyethane
Trichloromethane
Acetophenone
Dichloromethane
1,2-Dichloroethane
Acetone
DMA
DMF
2-Methylpropan-2-ol
DMSO
Acetonitrile
Butan-2-ol
Decan-1-ol
Octan-1-ol
Propan-2-ol
Pentan-1-ol
Butan-1-ol
Benzyl alcohol
Propan-1-ol
Ethanol
Methanol
Ethane-1,2-diol
Water
a Values obtained from ref. [2].
b In kcal mol-1.
Figure 5 shows a plot of the ET values as a function of the ET(30) solvent
polarity values for dye 10b, these being similar to those obtained for the other probes

Journal Pre-proof
19
(Figure S62). In all cases, an inverted behavior was verified, as previously described for
analogous dyes.[27-29, 31] As a typical example, for dye 10b, the solvatochromic band
has a maximum at 519 nm [ET(10b) = 52.0 kcal mol-1] in n-hexane (the least polar
solvent studied), while in acetophenone max = 638 nm [ET(10b) = 44.8 kcal mol-1]. This
reduction in the ET(10b) value follows an increase in the medium polarity, leading to a
bathochromic shift of max = +119 nm, characteristic of a positive solvatochromism. A
reversal in the solvatochromism starts to occur in the region of the plot obtained with
acetophenone as the solvent. A further increase in the solvent polarity leads to an
increase in the ET(10b) value, a negative solvatochromism being verified. The vis band
of 10b in water has a maximum at 550 nm [ET(10b) = 55.1 kcal mol-1], which,
compared with the same dye in acetophenone, corresponds to max = -88 nm.
56
54
52
50
48
46
44
30
35
40
45
50
55
60
65
ET(30) [kcal mol-1]
Figure 5. ET(30) value as a function of ET(dye) for dye 10b.
Table 3 summarizes the inverted behavior of dyes 6b-14b, in the form of a list
of their maximum λmax variations in the least polar, intermediate polarity, and most polar
solvents, as well as the Δλmax values for each dye, considering the regions of positive
and negative solvatochromic effects.

Journal Pre-proof
20
A comparison of dyes 6b-14b with the standard stilbene probe 5b (Table S1
and Figure S63) could be made with the aid of Eqn. (1), where m measures the average
shift of the solvatochromic max value for a given dye with respect to dye 5b and n
compares the solvatochromic sensitivity of the dye to polarity changes with that of 5b.
ET(dye) = m + n ET(5b)
(1)
Table 3. Values of λmax for compounds 6b–14b in the most polar, least polar, and
intermediate polarity solvents, as well as the Δλmax values obtained considering the
regions of positive and negative solvatochromism.
Most polar solvent
(H2O)
Intermediate
polarity solvent
λmax (nm)
Least polar
solvent (n-hexane)
λmax (nm)
a
b
Δλmax Δλmax
Dye
(nm)
(nm)
λmax (nm)
479
517
474
469
550
513
577
503
502
602 (DMA)
490
540
492
491
519
548
604
540
546
+112
+93
+74
+80
+119
+68
+26
+68
+64
-123
-116
-92
6b
7b
633 (Acetophenone)
566 (DMA)
8b
571 (Acetophenone)
638 (Acetophenone)
616 (Acetophenone)
630 (Acetophenone)
608 (Acetophenone)
610 (Acetophenone)
-102
-88
9b
10b
11b
12b
13b
14b
-103
-53
-105
-108
a Δλmax = λmax (intermediate solvent) – λmax (n-hexane).
b Δλmax = λmax (H2O) – λmax (intermediate solvent).
Table 4 summarizes these values for dyes 6b-14b. The linear correlations are better for
the 4-nitrophenyl in comparison with the 2,4-dinitrophenyl derivatives. Positive values
of m were obtained for all dyes, which indicates that the solvatochromic band of the
probe is hypsochromically shifted in comparison with the vis band of the standard probe

Journal Pre-proof
21
5b. The presence of substituents on the phenolate moiety makes the m values more
positive and the same occurs if the second nitro group is present in the molecular
structure of the dye. The inspection of the n values shows that all azo dyes studied
herein are much less sensitive to solvent polarity changes in comparison to the standard
dye 5b. The differences verified for the m and n values in comparison with structurally
related dyes,[29, 31] but with other different conjugated bridges (CH=CH and CH=N),
suggest that the azo bridge has plays an important role in the solvatochromism of these
systems, making them less solvatochromic, as suggested by Fabian and Hartmann.[60]
Table 4. Values for the slope n and intercept m on linear plots of ET(dye) as a function
of ET(5b) values, for compounds 6b–14b.
Dye
6b
m
n
r2
9.77 ± 2.51
0.74 ± 0.04
0.91
15.95 ± 2.76
24.29 ± 1.67
22.26 ± 2.53
19.02 ± 2.34
22.93 ± 2.43
38.65 ± 1.47
25.13 ± 2.49
24.70 ± 2.70
0.55 ± 0.04
0.51 ± 0.02
0.54 ± 0.04
0.50 ± 0.04
0.44 ± 0.04
0.13 ± 0.02
0.43 ± 0.04
0.43 ± 0.04
0.83
0.92
0.85
0.85
0.80
0.50
0.78
0.75
7b
8b
9b
10b
11b
12b
13b
14b
4.2. Multiparametric studies on the solvatochromism of 6b-14b
The experimental ET(dye) values for compounds 6b-14b in pure solvents were
analyzed using the Kamlet-Abboud-Taft (KAT) [61, 62] and Catalán [63]
multiparameter equations. The KAT approach uses Eqn. (2), where , and * are
parameters that represent the solvent hydrogen-bond donor (HBD) acidity, hydrogen-

Journal Pre-proof
22
bond acceptor (HBA) basicity, and solvent dipolarity/polarizability, respectively, while
 is a polarizability correction term for the solvent. The Catalán approach [63] requires
the use of Eqn. (3), where specific solvent parameters SA and SB represent the solvent
HBD acidity and HBA basicity, respectively, while SP and SdP are the nonspecific
solvent parameters polarizability and dipolarity, respectively. In these equations,
ET(dye)0 represents the ET(dye) value related to an inert solvent.
ET(dye) = ET(dye)0 + aα + bβ + s(π* + dδ)
(2)
ET(dye) = ET(dye)0 + aSA+ bSB + cSP + dSdP
(3)
Tables 5, S2, S3 and S4 show the contributions of the solvent properties to the
ET(dye) values determined for dyes 6b-14b, obtained from the multiple square
correlation analysis with the use of Eqns. (2) and (3). Similarly to the analysis
performed with related systems that exhibit reverse solvatochromism,[27-29, 31], the
application of Eqn. (3) to compounds 6b-14b provided more coherent results in
comparison with those obtained from the KAT analysis: (a) the r values are > 0.93 for
all compounds; (b) the contribution of the HBA basicity of the solvent is small (small b
coefficients); (c) in general, the most important contribution was that of the HBD
acidity of the solvent; and (d) the coefficients related to the solvent parameters
polarizability and dipolarity represent important contributions to the solvatochromic
behavior of the systems studied. For all dyes, with the exception of dye 6b, the
contribution of the polarizability term is more important than that of the dipolarity term.
Table 6 shows data from the Catalán multiparametric analysis for dyes 1, 2,
4b, 5b, and 6b to allow a comparison of the results for these compounds. Reichardt‘s
dye (1) is very sensitive to the acidity of the medium and quite sensitive to the
dipolarity of the medium. Effenberger‘s dye (2) shows almost no sensitivity to the

Journal Pre-proof
23
acidity and basicity of the medium, but is sensitive to the polarizability of the medium
and also, although to a lesser extent, to the dipolarity of the medium. It is important to
observe that all coefficients are positive for dye 1, which exhibits a well-known
negative solvatochromism, while both the polarizability and dipolarity coefficients are
negative for dye 2, which represents a classical example of a positively solvatochromic
probe. Regarding this aspect, as a common feature for the series of dyes 4b, 5b, and 6b,
a positive sign is verified for the acidity coefficient, while the other three coefficients
are negative. This common trend is related to the reverse solvatochromic behavior
exhibited by these dyes, reflecting the strong interaction with hydrogen-bond donating
solvents in the negative solvatochromism region and mainly the non-specific
interactions with the less polar solvents in the positive solvatochromism region. The
data show that although dyes 4b-6b are very sensitive to the acidity, the azophenolate
6b is the least sensitive, which could reflect the previously discussed action of the azo
groups as electron acceptors, reducing the electronic availability of the oxygen
phenolate moiety.
Table 5. Correlation coefficients a, b, c, and d obtained from the Catalán
multiparametric (Eqn. 2) analysis through the treatment of ET(dye) values for
compounds 6b–14b in various solvents.
Dye
6b
ET(dye)0
62.25
57.40
62.11
62.35
56.40
55.30
50.91
a
b
c
d
Na
29
29
29
29
29
29
29
r b
S.D. c
1.16
0.80
0.66
0.75
0.60
0.74
0.33
12.93
10.27
9.04
10.25
9.42
8.71
2.57
-1.73
-1.47
-2.16
-1.43
-1.49
-1.78
-1.12
-6.77
-8.26
-6.38
-7.90
-7.73
-6.07
-5.79
-8.62
-5.87
-5.29
-5.40
-4.94
-3.71
-0.19
0.96
0.96
0.97
0.97
0.97
0.96
0.93
7b
8b
9b
10b
11b
12b

Journal Pre-proof
24
55.98
55.95
8.77
8.66
-1.54
-1.79
-5.52
-6.77
-3.87
-2.79
29
29
0.95
0.96
0.77
0.71
13b
14b
a Number of solvents.
b Linear correlation.
c Standard deviation.
Table 6. Correlation coefficients a, b, c, and d obtained from the Catalán
multiparametric analysis through the treatment of ET(dye) values for different dyes in
various solvents.
N a
29
18
23
28
29
R b
S.D. c
1.62
0.32
2.22
1.60
1.16
Dye
1d,e
2e
ET(dye)0
29.47
70.09
69.14
71.46
62.25
a
b
c
d
22.64
0.01
4.96
0.06
-3.77
-3.33
-1.73
2.12
9.20
-7.32
-6.41
-9.79
-8.62
0.98
0.99
0.90
0.95
0.96
-14.66
-12.52
-10.54
-6.77
4be
5bf
6b
14.42
15.48
12.93
a Number of solvents.
b Linear correlation.
c Standard deviation.
d Using ET(30) values from Table 1.
e Calculated by the authors with data from the literature.[26]
f From the literature.[28, 29]
4.3. Theoretical Calculations
Table 7 shows the calculated absorption wavelengths for compounds 4b-6b. In
all cases the values are blue-shifted in water when compared with n-hexane. For dye 6b,
the calculated values are around 60 nm lower than the experimental values, although the
differences are in agreement with the empirical results. The calculated results for 6b
show a difference of 9 nm between n-hexane and water, while the experimental
difference was 11 nm. This difference is smaller than those observed for dyes 4b and
5b, which are considerably more solvatochromic. The excitations associated with the
values in Table 7 can be seen in Table 8. In all cases, the most significant transitions

Journal Pre-proof
25
are associated with polarizations among the highest occupied molecular orbitals of the
phenolate moiety during excitation. Figure 6 shows the calculated spectra for 6b in n-
hexane and water, along with the most significant transition.
Table 7. Experimental and calculated λmax values for 4b-6b in water and n-hexane.
max
(nm)
max
(nm)
H2O
n-hexane
H2O
n-hexane
Dye
λmaxtheor (nm) λmaxtheor (nm)
λmaxexp (nm) λmaxexp (nm)
430.6 a
428.0 b
479.0
486.2 a
450.2 b
490.0
-96.9
-76.6
-8.8
-55.6
-22.2
-11.0
401.3
390.5
498.2
467.1
425.7
4b
5b
6b
416.9
a Data from ref. 26.
b Data from ref. 27.
Figure 6. Calculated spectra for 6b in water and n-hexane. The most significant
transitions (HOMO-2 and HOMO-1 to HOMO, respectively) are shown in the insert.

Journal Pre-proof
26
Table 8. Most prominent states associated with transitions of Table 7. Density cut-offs
are 0.03 a.u. The strongest delocalization for each structure is shown in bold text.
Dye
4b
Most polar solvent (H2O)
Least polar solvent (n-hexane)
Donor
Acceptor
ΔE (eV) / %
Donor
Acceptor
ΔE (eV) / %
8.18 (< 1%)
9.90 (< 1%)
7.81 (1%)
HOMO-3 (60a) HOMO (63a)
HOMO-3 (60a) HOMO (63a)
HOMO-3 (60a) LUMO (64a)
9.57 (< 1%)
HOMO-3 (60a)
LUMO (64a)
7.89 (< 1%)
6.85 (< 1%)
HOMO-2 (61a) HOMO (63a)
HOMO-2 (61a) HOMO (63a)
HOMO-2 (61a)
LUMO (64a) 9.61 (< 1%) HOMO-2 (61a) LUMO (64a) 8.61 (< 1%)
5.98 (72%)
7.70 (16%)
8.16 (8%)
5.11 (83%)
6.87 (9%)
7.89 (5%)
HOMO-1 (62a) HOMO (63a)
HOMO-1 (62a) HOMO (63a)
HOMO-1 (62a)
LUMO (64a)
HOMO-1 (62a) LUMO (64a)
5b
HOMO-3 (60a) HOMO (63a)
HOMO-3 (60a) HOMO (63a)
HOMO-3 (60a) LUMO (64a)
9.92 (1%)
9.67 (< 1%)
7.06 (< 1%)
HOMO-3 (60a)
LUMO (64a)
8.04 (< 1%)
HOMO-2 (61a)
HOMO-2 (61a) HOMO (63a)
HOMO (63a)
HOMO-2 (61a)
LUMO (64a) 9.81 (< 1%) HOMO-2 (61a) LUMO (64a) 8.84 (< 1%)
6.01 (77%)
5.29 (83%)
HOMO-1 (62a) HOMO (63a)
HOMO-1 (62a) HOMO (63a)
HOMO-1 (62a)
LUMO (64a)
7.78 (10%)
HOMO-1 (62a) LUMO (64a)
7.07 (6%)
6b
8.10 (3%)
9.37 (2%)
7.67 (19%)
9.01 (9%)
HOMO-3 (60a) HOMO (63a)
HOMO (63a)
LUMO (64a)
HOMO-3 (60a)
HOMO-3 (60a)
HOMO-3 (60a)
LUMO (64a)
7.50 (11%)
8.84 (6%)
7.88 (47%)
HOMO-2 (61a)
HOMO-2 (61a) HOMO (63a)
HOMO-2 (61a) LUMO (64a)
HOMO (63a)
HOMO-2 (61a)
LUMO (64a)
9.14 (30%)

Journal Pre-proof
27
6.20 (7%)
5.76 (43%)
HOMO-1 (62a) HOMO (63a)
HOMO-1 (62a) LUMO (64a)
HOMO-1 (62a) HOMO (63a)
HOMO-1 (62a) LUMO (64a)
7.47 (2%)
7.09 (6%)
In order to investigate the solvation pattern of the dyes, molecular dynamics
simulations of compounds 4b-6b were performed with water, DMF and n-hexane as
solvents. These dyes and solvents were chosen as prototypes to analyze different
solvation behaviors that can arise depending on the bridging group and on the solvent
polarity. Radial distribution functions (RDFs) were calculated from equilibrated
trajectories obtained via molecular dynamics simulations. The analysis was focused on
bridging atoms (nitrogen or carbon of CH=CH, CH=N and N=N bridges) and on the
phenolate and nitro substituents of the donor and acceptor groups, respectively.
All three dyes yielded very similar RDFs around the phenolate oxygen (Figure
7A) and nitro oxygens (Figure 7C) with water as the solvent. In the phenolate moiety, a
first solvation shell centered at 1.9 Å with three water molecules and a second more
diffuse shell centered at 3.3 Å with 16-17 water molecules were verified (Figure 8).
The nitro oxygen of 4b-6b also formed hydrogen bonds with the solvent, as shown by
the RDFs of Figure 7C, giving rise to a broader solvation shell centered at 2.9 Å with
seven water molecules. The solvation shell around the nitro moiety is less intense
compared to the donor group. A solvation shell consisting of 21 molecules is observed
for the bridging groups of compounds 4b-6b. However, different patterns were
observed depending on the bridging moiety. A broad and unique shell centered at ~2.5
Å was observed for the ethylene bridge of 4b. In comparison, two well-defined
solvation shells were verified at ~2 Å and 3 Å with 2 and 19 water molecules,
respectively, for compounds 5b and 6b. This difference in the behavior observed for the
solvation pattern is due to the presence of nitrogen atoms in the bridging group, and it

Journal Pre-proof
28
may be associated with the differences in the solvatochromism observed for compounds
4b-6b as the bridging atoms are varied.
A similar pattern for the solvation around the phenolate and nitro groups was
observed with DMF as the solvent for the three dyes, as evidenced in Figures 7B and
7D. All compounds showed a diffuse first solvation layer at 4.4 Å with 4-6 solvent
molecules for the phenolate group and the peak centered at 4.4-4.6 Å with four DMF
molecules for the nitro moiety. As expected, interactions of 4b-6b with DMF are
specific, but less intense compared to water. Differences in the solvation structure of the
bridging moieties are also observed for DMF. Figure 7F demonstrates a more compact
solvation structure for the azo group of 6b compared to the CH=CH group of 4b, as
observed from the peak centered at ~ 6 Å for 4b with 12 molecules and the peak
centered at ~ 5.4 Å with 9 molecules for 6b. Similarly to the case where water was used
as the solvent, the behavior of compound 5b was intermediate, between those of 4b and
6b, because of the CH=N group, with two peaks at 4.5 Å and 6 Å and with nine
molecules in the solvation shell.
The solvation patterns of the three dyes in n–hexane are very similar (see Figure
S66). The hydrogens of the solvent interact more strongly with the electronegative
atoms of the probes, but the interactions are, as expected, much weaker than those in the
intense and strongly packed first solvation shell verified with water. Similar results have
been obtained recently in a study on a related family of imine solvatochromic dyes.[30]

Journal Pre-proof
29
Water
DMF
Figure 7. Radial distribution functions (RDFs) for the solvation of (A)-(B) phenolate
oxygen, (C)-(D) nitro oxygen, and (E)-(F) bridging atom of 4b-6b in water (left) and
DMF (right). Atom types (HW) and (OW) denote the hydrogen and oxygen atoms of
water, respectively, (OH) the oxygen of the phenolate group, (ON) the oxygen of the
acceptor nitro group, (CM) the sp2 carbon of the CH=CH bridge and (NC) the nitrogen
of the azo and methyleneimino moieties, while (N) and (O) represent heteroatoms of
DMF.