The possible tautomerism of the potential rotary switch 2-(2-(2-Hydroxy-4-nitrophenyl)hydrazono)-1-phenylbutane-1,3-dione

Article abstract of DOI:10.1016/j.dyepig.2017.05.021

The title compound is potentially tautomeric and its tautomerism was studied by means of molecular spectroscopy (¹H and ¹³C NMR and UV–Vis) in DMSO as well as by quantum chemical calculations (M06-2X/TZVP). The detailed assignment of the NMR signals supported by the theoretical calculations clearly shows that the previous interpretation, available in the literature, about the coexistence of two tautomeric forms is not correct. The compound exists as major and minor isomer of a single tautomeric form. In addition, a 2-methoxy derivative (the OH group replaced by a methoxy group) is also investigated and show similar trends.

Full text of DOI:10.1016/j.dyepig.2017.05.021

Accepted Manuscript
The possible tautomerism of the potential rotary switch 2-(2-(2-Hydroxy-4-
nitrophenyl)hydrazono)-1-phenylbutane-1,3-dione
Silvia Hristova, Fadhil S. Kamounah, Nevse Molla, Poul Erik Hansen, Daniela
Nedeltcheva, Liudmil Antonov
PII:
S0143-7208(17)30160-2
10.1016/j.dyepig.2017.05.021
DYPI 5983
DOI:
Reference:
To appear in: Dyes and Pigments
Received Date: 25 January 2017
Revised Date: 7 May 2017
Accepted Date: 8 May 2017
Please cite this article as: Hristova S, Kamounah FS, Molla N, Hansen PE, Nedeltcheva D, Antonov L,
The possible tautomerism of the potential rotary switch 2-(2-(2-Hydroxy-4-nitrophenyl)hydrazono)-1-
phenylbutane-1,3-dione, Dyes and Pigments (2017), doi: 10.1016/j.dyepig.2017.05.021.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT
The possible tautomerism of the potential rotary switch 2-(2-(2-Hydroxy-4-
nitrophenyl)hydrazono)-1-phenylbutane-1,3-dione
Silvia Hristova1, Fadhil S. Kamounah2, Nevse Molla1, Poul Erik Hansen3, Daniela Nedeltcheva1
and Liudmil Antonov1,*
1 Institute of Organic Chemistry with Centre of Phytochemistry, Bulgarian Academy of Sciences,
Acad. G. Bonchev str., bl. 9, 1113 Sofia, Bulgaria
2 Department of Chemistry, University of Copenhagen, Universitetsparken 5, DK-2100
Copenhagen Ø, Denmark.
3 Roskilde University, Department of Science and Environment, DK-4000, Roskilde, Denmark
Abstract:
The title compound is potentially tautomeric and its tautomerism was studied by means of
molecular spectroscopy (1H and 13C NMR and UV-Vis) in DMSO as well as by quantum chemical
calculations (M06-2X/TZVP). The detailed assignment of the NMR signals supported by the
theoretical calculations clearly shows that the previous interpretation, available in the literature,
about the coexistence of two tautomeric forms is not correct. The compound exists as major
and minor isomer of a single tautomeric form. In addition, a 2-methoxy derivative (the OH
group replaced by a methoxy group) is also investigated and show similar trends.
* Corresponding author; http://www.orgchm.bas.bg/~i2mp
1

ACCEPTED MANUSCRIPT
Key words: rotary switch, tautomerism, hydrazone, NMR, UV-Vis spectroscopy, DFT
2

ACCEPTED MANUSCRIPT
Introduction:
The switch is one of the simplest components of an electrical circuit. In molecular
electronics this term is used to assign any molecular "structure" able to reversibly shift between
two or more stable states in response to environmental stimuli such as changes in pH, light,
temperature, electric current, microenvironment, or in the presence of an ion [1–6]. Virtually,
every single molecule changes its behaviour when acted upon by external stimuli, but its use as
a molecular switch is possible only if these changes are reproducible, reversible and can be
explicitly controlled [7]. Most of the existing switches [7–12], are based on primary processes
like photochromism, host-guest interactions, protonation/deprotonation, oxidation/reduction
or a combination of them. Tautomerism could be an additional elemental process, because, on
one side, the change in the tautomeric state is accomplished by a fast proton transfer reaction
between two or more structures, each of them with different properties. The practically
unlimited number of on/off cycles and, hence, high fatigue resistance, can be the obvious
advantage over the other existing primary mechanisms [13]. Real or potentially switchable
tautomeric systems have been classified as tautomeric tweezers, tautomeric cavities, proton
cranes and rotary switches [14]. The analysis of the advantages/disadvantages of the first three
of these classes leads to clear conclusion that a real switching is possible only when a suitable
mediator is structurally implemented in the tautomeric molecule. In such a case the addition of
acid/base/ions in solution or suitable irradiation could lead to indirect change in the position of
the tautomeric equilibrium in a controlled and reproducible manner. The mediating unit (arm,
crane [15,16] or cavity [17,18]) plays the role of an antenna transferring the signal from the
external stimuli to the tautomeric backbone, which responses to the changes by altering its
tautomeric state. Of course, a very careful selection of the tautomeric backbone and thoughtful
design of the antenna are required [14].
3

ACCEPTED MANUSCRIPT
It is known [19] that 1,2,3-tricarbonyl-2-aryl-hydrazones exist in solution as an
equilibrated mixture of intra-molecularly H-bonded E/Z isomers (KE and KZ, Scheme 1). The
for/backward isomerization proceeds through keto-enol tautomerization followed by rotation
around a C–N single bond, giving a name “rotary switches”. Actually this may be considered as a
molecular motor consisting a stator (aromatic unit) and a rotor (β-diketone part) connected by
an axle. The position of the isomerization equilibrium can be altered by catalytic amounts of
acid and base, albeit the process does not proceed to full completion. The similar hydrogen
bond accepting capability of the two carbonyl groups, providing insufficient impetus to drive
the equilibrium to full extent in either direction, makes such design not ideal for molecular
switch related applications[20]. Substantial improvement was achieved by Aprahamian and co-
workers [20–22], who replaced one of the carbonyl groups (R1CO) with a pyridyl group, a much
stronger hydrogen bond acceptor. This modified rotor has led to systems having an appreciable
bias in favour of the KE. The basic nature of the pyridyl nitrogen atom allows its hydrogen bond
accepting ability to be switched off using acid input, which in turn makes the KZ isomer the sole
configuration upon protonation. Modifications of the stator, i.e. replacement of the 1-phenyl/1-
naphthyl ring with an 8-quinolinyl one leads to a tristable switch [21,22], controlled by addition
of acid, base and Zn2+. A very recent application has shown that such molecular rotors can be
used not only for switching, but also as a molecule with robotic arms, which deliver cargo from
one part of the molecule to another [23].
4

ACCEPTED MANUSCRIPT
Scheme 1. Sketch of possible mechanism for the acid-catalyzed tautomerization in E/Z
switching process of 1,2,3-tricarbonyl-2-arylhydrazones (R1≠R2).
The title molecule 1, belonging to a series of arylhydrazones of β-diketones, where an
additional tautomeric functionality is added in the stator, is an interesting example for potential
action as a rotary switch, This system has been studied by Mahmudov et al [24,25] by means of
IR spectroscopy and NMR and the authors have pointed out that a tautomeric mixture between
Z-enol-azo-II and hydrazo-III forms (Scheme 2) exists in methanol and dimethyl sulfoxide
(DMSO) solutions. If so, compound 1 differs substantially from the rotary switches described
above, which exist as a mixture of isomers of the K-tautomeric form, and, therefore, can be
counted as a real tautomeric switching system. On the other side it is not likely that the
implementation of the OH group in the stator could lead to additional tautomerism due to
aromatic reasons [26] and the added group, due to its ability to form hydrogen bonding, could
either facilitate the tautomeric proton transfer in the rotor as an accelerator or to hamper it
like a brake. Therefore, critically analyzing the existing results and providing new experimental
and theoretical data we would like in this communication to explain the effect of the additional
OH group in the stator of 1 on the tautomeric properties of this system. The discussion is
facilitated by using a model compound 2, where the OH group is replaced by a methoxy group.
The aim of the present paper is, based on UV-VIS measurements, a complete assignment of all
1H and 13C NMR resonances and NMR studies, including deuterium isotope effects on 13C
chemical shifts, and finally on DFT calculations, to elucidate the structure, the predominant
tautomeric forms and barriers to rotation around vital bonds. This system is a powerful system
5

ACCEPTED MANUSCRIPT
to elucidate tautomerism and factors determining preferred tautomers as well as the influence
of solvent and specific solvent interactions.
Scheme 2. Sketch of the tautomeric forms of the compounds investigated by Mahmudov at al.
[24]. The other substituents in the aromatic ring are omitted for simplicity. (Reprinted from
Journal of Molecular Liquids, 162, K.T.Mahmudov, R.A.Rahimov, M.B.Babanly, P.Q. Hasanov,
F.G.Pashaev, A.G.Gasanov, M.N.Kopylovich, A.J.L.Pombeiro, Tautomery and acid–base
properties of some azoderivatives of benzoylacetone, 84-88, Copyright (2011), with permission
from Elsevier).
6

ACCEPTED MANUSCRIPT
Results and discussion
It is well-known that the azo-hydrazo tautomerism [26–28] is solvent dependent even
when the proton is transferred through an intramolecular hydrogen bond [29]. Therefore the
change in the absorption spectra, when the solvent is changed, is the first sign of existence of a
tautomeric equilibrium. In the case of 1 the spectra are virtually the same in a number of
solvents (from benzene to acetonitrile, see Figure S1), showing an absorption maximum at 420
nm. As seen from Figure 1, in diluted solutions (Figure S2) of alcohols and especially in strong
proton acceptor solvents like DMSO and dimethyl formamide an additional red shifted band
appears.
Figure 1. Absorption spectra of 1 in acetonitrile (_____), ethanol (-----), DMSO (-.-.-) and DMF
(
_____) with the same concentration.
7

ACCEPTED MANUSCRIPT
The origin of the long-wavelength band can be explained having in mind the
deprotonation of the phenols, especially such with strong electron acceptor substituents, in
solution [30–32]. As seen from Figure 2 the addition of acid in DMSO leads to disappearance of
the band at 540 nm, while the deprotonation with base strengthens it. The same effect is
observed in ethanol (Figure S3).
The spectra of 2 are solvent and concentration independent and no additional band is
seen in the solvents listed in Figure 1. The deprotonation of 2 in DMSO occurs in stronger
basicity of the solution and the deprotonation band is shifted only by ~ 60 nm.
Figure 2. Absorption spectra of 1 in DMSO upon acid and base addition (____ without acid/base
____
addition; - - - - final spectrum upon acid addition (2.16 equivalents H2SO4);
upon base addition (21.6 equivalents NaOH)).
final spectrum
8

ACCEPTED MANUSCRIPT
These UV-Vis spectral data show that there are no evidences for tautomeric equilibrium
in 1 in the used solvents (no spectral changes upon changing solvent) and the compound most
probably exists as a mixture of neutral and deprotonated form depending on the solvent and
concentration.
Both compounds 1 and 2 have been investigated by NMR in a series of solvents. The
most important solvent is DMSO-d6 as this refers directly to the study of Mahmudow et al. [24],
but other solvents have been also used although the compounds are not very soluble in non-
polar solvents.
Yao [33] in the synthesis of the phenyl derivative found only one product and assigned this to a
hydrazone with a the NH forming a hydrogen bond to the CH3C=O group based on the its NH
chemical shift of 14.70 ppm. He made a comparison between the derivative of acetylacetone
(NH chemical shift 14.68 ppm) and that of dibenzoylmethane (NH chemical shift 13.40 ppm) to
1
13
make this assignment. However, from both the H and C NMR spectra of 1 and 2 in DMSO-d6
(Figures S4-S8), it is obvious that two species are present in different ratios in the following
called major and minor form. A comparison of data for 1 and 2 indicates very similar structures.
1
13
The assignments are based on H and C NMR spectra, HMBC spectra (including one-bond
1
13
correlations) and NOESY spectra. H and C chemical shift and HMBC correlations are given in
Tables 1 and 2. Numbers and a few assignments have previously been given in original work of
Mahmudov at al. [24]. The 13C data also not based on a full assignment were given in the same
paper.
The NH and phenolic OH chemical shifts for the major species in DMSO-d6 are found at
14.14 ppm and 11.52 ppm (the assignment of NH and OH resonances are based on isotope
9

ACCEPTED MANUSCRIPT
substitution, see below). These are for the NH resonance rather different from the minor
species (11.70 ppm) and OH resonance of the minor species (11.34 ppm).
Scheme 3. Numbering of the carbon atoms in 1 (R=H) and 2 (R=CH3).
1
Table 1. H and 13C chemical shifts of the major form of 1 and 2 in DMSO-d6.
Compound
Carbons
1
1
1
1
2
2
1H chemical 13C chemical
One bond
correlation
HMBC
1H
13C
shift
shift
correlation chemical chemical
shift
2.51
shift
1
2
2.51
30.67
198.08
135.66
191.95
136.25
145.99
110.60
Obs.
C-3(s), C-2(s)
30.19
-
197.80
135.83
191.45
136.48
146.92
107.11
3
-
4
-
1´
2´
3´
-
-
7.70
Obs.
C-5’(s), C-
7.55
4´(w), C-2´(w)
4´
5´
-
143.78
116.78
143.44
118.03
7.72
Obs.
Obs.
C-3´(s), C-
1´(w), C-4´(w)
7.90
7.26
6´
7.11
114.08
C-1´(w) ,C-
2´(s), C-3´(w),
C-4´(w)
113.59
10

ACCEPTED MANUSCRIPT
1”
2”
3”
-
137.96
137.24
130.33
128.11
7.89
7.54
130.75
128.50
Obs.
C-4”(s)
7.90
7.55
C-1”(s), C-
3”(s)
4”
7.66
133.10
-
C-2”(s)
7.67
132.84
-
NH
14.14
C-6´(s),C-3(s)
and C-2´(w)
14.11
OH
11.52
-
-
-
-
Table 2. 1H and 13C chemical shifts of the minor form of 1 and 2 in DMSO-d6.
Compound
Carbons
1
1
13C
chemical
shift
1
1
2
2
1H chemical
shift
One bond
correlation
L.R.
1H
13C
correlation chemical chemical
shift
shift
1
2
2.56
25.92
196.83
139.28
193.17
136.51
145.50
110.49
143.23
116.83
114.25
Obs.
C-3(s), C-2(s)
2.57
25.47
-
196.53
139.68
192.73
136.86
146.67
106.94
142.80
117.98
113.59
3
-
4
-
1´
2´
3´
4´
5´
6´
-
-
7.70
-
7.87
7.81
7.75
Obs.
Obs.
8.00
7.84
C-2´(s),C-
4´(s)
1”
2”
3”
4”
-
137.56
129.08a
128.93a
133.99
137.24
128.68a
128.49a
133.66
7.71
7.52
7.65
7.72
7.51
7.66
Obs.
C-1”(s)
C-2” (w), C-
11

ACCEPTED MANUSCRIPT
3”(w)
NH
OH
11.70
11.34
-
-
C-1´(s), C-
3(m), C-6´(w)
-
-
a. May be interchanged
It has been possible to isolate an almost pure form of 1 (NMR spectra in non-polar solvents,
Figures S9-S10), which turned out to be the major form. In case of 2 a small amount of almost
pure minor form was isolated. Dissolving 1 or the minor form of 2 in DMSO, acetonitrile-d3 or
tetrahydrofuran-d8 and recording the NMR spectra within 10 min led to an equilibrium mixture
as seen in the NMR spectra and with a higher content of the minor form than seen when
dissolving the compound in a nonpolar solvent (see below). Something similar was seen with
DMF-d7 (Figure S11). However, dissolving the compounds in toluene-d8 or CDCl3 led only to a
very slow conversion (days) into the equilibrium mixture.
Deuteriation is one of the effective tools to study complex tautomeric systems [34]. A brief
summary is given below: In case of an equilibrium, fast on the NMR time scale, only averaged
NMR signals are seen. With respect to deuterium isotope effects these will be averaged too,
but in case of equilibrium large effects may be seen [35]. Deuteriation is achieved by dissolving
the compounds in CH3OD and subsequent evaporation. Deuteriation was repeated with
different content of deuterium to determine the signs of the isotope effects. Deuteriation of 1
and 2 will primarily take place at the OH and NH hydrogens but may also take place at the
CH3C=O group. Deuterium substitution may lead to isotope splittings provided the exchange of
the label is slow on the NMR time scale. This can usually be judged from the broadness of the
XH resonance.
12

ACCEPTED MANUSCRIPT
Isotope effects in 1 both the OH of the major and the minor form and the NH resonance of
the minor form merge into two broad signals at 11.64 and 11.50 ppm upon deuteriation. The
NH resonance of the major form is reduced to 33%. The remaining resonances are reduced to
the same extent. The isotope effects are summarized in Scheme 4. It is seen that effects due to
deuteriation of the OH group of 1 are not seen due to exchange. The observed effects are due
to deuteriation at the NH groups and those effects are similar to effects observed for a
derivative of acetylacetone [36] and also to a couple of fluorosubstituted derivatives [37]. The
effects due to deuteriation at the methyl group is only seen at the nearby C=O carbon (see
Scheme 3). The observed isotope effects are consistent with the major form. For the minor
form only one isotope effect is seen as a broadening of C-1` (Schemes 3-4).
Deuteriation would happen at the enolic OH and this would lead to large deuterium isotope
effects at the C=O carbons [38], if any enol-azo form (Scheme 2) was present, but this is clearly
not seen. OH groups like these are less prone the exchange so this could not be the reason for
not seeing such large isotope effects. The mentioned forms can be excluded from consideration.
0.24
0.02
0.12
0.06
0.05
CH3 (D)
-br
NO2
NO2
br
O
N
O
O
N
O
N
N
H
(D)
H
(D)
HO
(D)
HO
(D)
(D)
H3C
0.04,
0.24
-br
1 (minor form)
1 (major form)
13
Scheme 4. Deuteriium isotope effects on C chemical shifts. Numbers in italics are due to
deuteriation at the methyl group. DMSO molecules are left out for simplicity.
13

ACCEPTED MANUSCRIPT
The apparent contradiction between the UV-Vis data (single neutral compound) and NMR
results (two neutral components) can be explained by the theoretical calculations, performed in
DMSO as media. As seen from Scheme 5 the title compound could exist as seven neutral and
three zwitterionic tautomers. Taking the OH group in the phenyl ring as a starting point,
structures I-IV could be attributed to the enol-like tautomer, while the rest are keto-like ones.
Each of them exists as a large number of possible conformers (see Scheme S1 for an example),
but after the performed lengthy calculations the most stable isomer of each tautomer are
presented in Table 3.
OH
O2N
NH N C(COCH3)COPh
I
OH
O
O2N
N
N CH(COCH3)COPh
O2N
N
NH CH(COCH3)COPh
II
V
O
O
OH
C(OH)Ph
O2N
N
NH C(COCH3)
O2N
N
N C(COCH3)
C(OH)Ph
III
VI
OH
O2N
N
N C(COPh)
O2N
N
NH C(COPh)
C(OH)CH3
C(OH)CH3
IV
VII
O
O2N
NH NH C(COCH3)COPh
VIII
14

ACCEPTED MANUSCRIPT
O
O
C(COCH3)C(OH)Ph
C(COPh)C(OH)CH3
O2N
NH N
O2N
NH N
IX
X
Scheme 5. Sketches of the possible tautomers of 1. In 2 only tautomers I-IV are potentially
possible.
Table 3. The most stable isomers of I-X of 1 in DMSO and their predicted absorption maxima (as
position* and oscillator strength).
Structure
Dipole Long-wavelength
∆E
moment
absorption
[kcal/mol]
[D]
f
λmax
[nm]
357
355
0.056
0.735
1(I)
0.00
6.0
323
322
0.107
0.070
1(II)
13.51
3.0
15

ACCEPTED MANUSCRIPT
1(III)
6.32
6.0
367
0.704
442
354
0.012
0.590
1(IV)
7.42
5.4
415
408
0.023
0.200
1(V)
19.10
8.6
1(VI)
17.23
6.1
446
0.327
16

ACCEPTED MANUSCRIPT
1(VII)
17.62
5.2
445
0.234
1(VIII)
No stable structure found
1(IX)
23.29
13.3
531
0.444
1(X)
24.81
14.9
523
0.532
* taking into account that TD-M06-2X systematically underestimates the band positions (see
the Experimental part) these values should not be considered as absolute values, but as
indication of a spectral shift.
17

ACCEPTED MANUSCRIPT
As seen from Table 3, the structures, related to transfer of the OH proton in 1 (V-VII and
IX-X) are unstable, which is expected because such transfer reduces the aromatic character of
the phenyl ring [26]. The tautomer 1(I) is substantially more stable and the energy difference to
the nearer as stability forms (1(III) and 1(IV)) excludes their existence in measurable amounts in
solution. The stability of 1(I) could be also confirmed from the crystallographic data for
structurally similar compounds, which show that in solid state the hydrazone tautomer only
exists stabilized through H3CC=O…HN intramolecular hydrogen bonding [39]. All these three,
most stable according to the calculations structures, have similar dipole moment and are
stabilized by intramolecular hydrogen bonding which means that the change of the solvent in
the solvent model could not change the relative stability. Even if there are specific interactions
with the solvent, which the solvation model does not take into account, structures 1(I), 1(III)
and 1(IV), as given in Table 3, could be inert to proton acceptor solvents and with similar
possibility to interact with proton-donor solvents. Therefore, assuming that the order of
relative stability is the same in the used solvents and in DMSO one should look not for
tautomers, but for isomers from the group of 1(I). According to the data shown in Scheme S1
(summarized in Figure S12) there are four isomers with relative energy, which suppose that
they are presented in solution. As seen 1(Ib) should be dominating (~75%) with substantial
contribution of 1(Id) (~20%), followed by minor amounts of 1(Ic) and 1(Ia). A careful
consideration allows grouping these structures based on their ability to interact with DMSO. On
one side in 1(Ia) and 1(Ib) the OH proton is involved in a intramolecular hydrogen bonding,
which makes it not accessible for interaction with the solvent. On the other side in the couple
1(Ic)/1(Id) the proton can interact with DMSO, which could lead to further stabilization. The
actual effect of taking the specific solute-solvent interactions into account by adding a DMSO
18

ACCEPTED MANUSCRIPT
molecule in the DMSO solvent field is shown in Figure 3. As a result the Id conformer is strongly
stabilized, followed by Ic (10-15%). The remaining two could not be presented in a DMSO
solution. From the view point of UV-Vis spectroscopy the structures Id and Ic could hardly be
distinguished – the conjugated system is the same and, correspondingly, their spectra should
be very similar (see Table S1 with the predicted absorption in DMSO), which is the reason to
see a single spectral envelop and no measurable changes after dissolution in non-polar solvents,
as observed by NMR. The calculations for 2 confirm the conclusions. On one side tautomers
2(V-X) are impossible, on the other – the structures, which rely to stabilization through
phenolic OH intramolecular hydrogen bonding, are strongly destabilized. As a result – the
isomer types 2(Id) and 2(Ic) are the most stable structures (the second higher in energy by 1.08
kcal/mole in DMSO).
19

ACCEPTED MANUSCRIPT
Ib
Ia
Ic
Id
Figure 3. Relative energies (in kcal/mole units) of the most stable isomers of I in DMSO taking
specific interactions into account. The positions of the double bonds can be seen in Scheme S1.
20

ACCEPTED MANUSCRIPT
The interpretation of the NMR data supports the results from the calculations. The
structures IV and the structurally similar isomer of III (IIIa-IIIb, see Scheme S2) are not possible
as no high frequency resonance (~ 15 ppm) is observed or large isotope effects (see previously).
The observation of a NH to C=N cross-peaks in the HMBC do not suggest existence of III and the
structurally similar isomers of IV (IVc, Scheme S2). Tautomer II is not observed as CH resonance
around 5 ppm is missing.
Figure 4. Schematic diagram for transition from Ib to Id and back in DMSO (the values are in
kcal/mol). The positions of the double bonds can be seen in Scheme S1.
21

ACCEPTED MANUSCRIPT
The structures of the major form of 1 in polar solvents are clearly very similar as judged
from the NH and OH chemical shifts (Table 4). In the non-polar solvents, CDCl3 and toluene-d8
the OH chemical shift (Table 4) indicates either no hydrogen bonding in line with a structure in
which the C=O groups points away from the aromatic ring marked `. This is supported by the H-
6`chemical shift (see previously) (Table 4). For the minor form the situation is much the same
(Table 4).
Table 4. NH, OH, methyl and H-6´ chemical shifts for 1 and 2.
Compound
1 (major form)
1 (minor form)
1 (major form)
1 (minor form)
1 (major form)
1 (minor form)
1 (major form)
1 (minor form)
1 (major form)
1 (minor form)
2 (major form)
2 (minor form)
2 (major form)
2 (minor form)
2 (major form)
2 (minor form)
2 (major form)
2 (minor form)
NH
OH
CH3
2.66
2.59
2.50
2.56
2.58
2.57
H-6`
7.13
~7.84
7.20
7.75
7.38
~7.8
6.47
n.o.
Solvent
14.76
12.42
14.14
11.70
14.31
11.94
14.31
12.30
14.26
11.93
14.40
12.14
14.11
11.58
14.40
11.75
14.32
11.85
8.42
CDCl3
7.00
11.52
DMSO-d6
11.34
11.94
DMF-d7
11.94
6.97
Toluene-d8
Acetonitrile-d3
CDCl3
n.o.
n.o.
2.56
2.59
2.63
2.625
2.57
2.57
2.55
2.58
2.54
2.56
10.58
7.23
n.o.
10.58
-
-
-
-
-
-
-
-
7.40
7.78
7.27
7.86
7.30
n.o.
DMSO-d6
Acetonitrile-d3
Tetrahydrofuran-d8
7.40
7.85
The following structures are possible, namely Ia-Id in 1 and Ic-Id in 2. A major difference
between the major and the minor species is the chemical shift of H-6´. A small difference is also
seen for H-2”. In the HMBC spectrum no cross peaks are seen from the OH resonance neither in
the major or the minor form. The carbonyl resonances next to the methyl groups were
22

ACCEPTED MANUSCRIPT
unambiguously assigned by deuterium isotope effects caused by partial deuteriation of the
methyl groups, but also by a HMBC correlation from the CH3 to CO in the major form.
The very low chemical shift of H-6´ma can only be due to ring current effects. These must
come from the ´-ring (Scheme 3) and can only occur in structure Id (see Fig. 3). From the NOESY
spectrum cross peaks from NHma are seen to H-6´, H-2”and H-3´ and/or H-5´. The OHma
resonance shows a cross peak to H-3´ or H-5´ and as H-5´ is far away this is probably to H-3´
indicating that the OHma group is pointing towards H-3´ (see structure Id). It is obvious from the
NOESY spectrum that both the NHma and the OHma hydrogens exchange with water.
Furthermore, the water signal shows a weak cross peak to H-3´ suggesting that the water is
hydrogen bonded to the DMSO molecule, which is hydrogen bonded to OHma.
As the major compound is assigned to Id, the minor compound could be found by
exclusion. On one side, Id and Ib are rotamers and the barrier to rotation around the single C-N
bond is low. As seen from Figure 4 the barrier of rotation in 1 is 8.14 kcal/mol without to
account the specific interaction with the solvent. If it is accounted, the energy of Ib will rise (see
Figure 3) and the transition state, where the OH proton can interact with DMSO molecule,
would be stabilized, which will even more decrease the rotation barrier. For this reason Id and
Ib would not show too different sets of signals. The question is therefore which of the form Ia
or Ic corresponds to the minor form. The spatial closeness of the CH3 protons and the NH
proton would suggest that a cross peak could be observed in the NOESY spectrum. No cross
peak is observed and therefore the most likely structure for the minor form is Ic in DMSO-d6.
The barrier of transition from Id to Ic, as shown in Figure 5, is sufficient to measure separate
signals. Actually the interpretation for coexistence of Id as major component and Ic as minor
one fully corresponds to the theoretical predictions summarized in Figure 3 and the
23

ACCEPTED MANUSCRIPT
crystallographic data [39] in similar compounds. Of course a very straightway support of this
conclusion comes from similarity of the NMR signals of the major and minor from of 1 and 2. In
2 forms Ia and Ib do not exist, consequently Ic and Id are the only forms available. As a side
results it could be claimed that theoretical calculations predict correctly the real situation in
solution.
The combined assignment of existing forms in solution could answer the question of the
effect of the additional OH group in the stator of the potential rotary switch 1. Obviously, at
least in DMSO, the OH group as presented in 1(Id) and 1(Ic) does not interact with the rotor and
consequently cannot play a direct role in the rotation process. It could be expected to act as an
electron acceptor substituent, influencing together with the nitro group the neighboring
hydrazone nitrogen atom from the axle. However, the effect of the OH group by itself is not
large taking into account that in the case of a compound without it [40] again the Id-like and Ic-
like (as interaction of the MeC=O/PhC=O groups with NH from the axle) isomers are presented
in solution in proportion 85%/15%, as shown by NMR, which is very similar to the predicted by
the theoretical calculations above.
Here is the moment to comment the findings of Mahmudov et al. that the title
compound exists in DMSO as a mixture of 22% Z-enol-azo-II and 78% hydrazo-III forms [24].
Comparing Scheme 2 (R’=Ph and R=CH3) and Table 1, Z-enol-azo-II corresponds to 1(III), while
hydrazo-III is 1(Id). The minor component cannot be 1(III), because it is unstable to exist in
solution according to the theoretical calculations from Table 1 and the NMR results. Even if we
assume that it is available, the barrier of proton transfer between 1(III) and 1(Ib) is less than 1
kcal/mol, which means that they cannot be separated on the NMR scale. As already discussed
24

ACCEPTED MANUSCRIPT
above 1(Id) and 1(Ib) also cannot have separate signals. Obviously a misinterpretation of the
NMR data in the original paper leads to incorrect conclusions.
25

ACCEPTED MANUSCRIPT
Figure 5. Schematic diagram for transition from Id to Ic and back in 1 in DMSO (the values are in kcal/mol). The positions of the double bonds
can be seen in Scheme S1. The energetics in 2 is virtually the same.
26

ACCEPTED MANUSCRIPT
1
2
Two additional factors influence the specific processes in 1 in solution, namely the
concentration and the water content. The spontaneous deprotonation shown in Figures 1 and
S1 is concentration dependent. The increase of the concentration leads to disappearing of the
deprotonated form, which happens at relatively low concentrations. For instance, in DMSO at
~5. 10-4 mol/l only neutral form is observed (see Figure S2). This behaviour suggests that in the
NMR concentration scale no deprotonated form exists. The increase of the temperature causes
a slight rise of the band at 540 nm and correspondingly decreases the intensity of the neutral
form band (Figure S13). Most probably 1(Id) and 1(Ic) aggregate in solution and with dilution
the monomers are deprotonated by the solvent. The OH deprotonated forms exist in the same
ratio (1(Id-) is more stable with 1.2 kcal/mol comparing to 1(Ic-)), their spectra are very similar
(Table S1) and again cannot be distinguished by UV-Vis spectroscopy. It could be expected that
the OH proton, not involved in an intramolecular hydrogen bonding, is responsible for the
aggregation behaviour. On Figure 6 the border cases of the possible dimers of 1(Id) are shown.
Although the sandwich dimer is shown to be almost 8 kcal/mol more stable by the calculations,
this value should be considered with care, because no basis set superposition error correction is
taken into account. The more important fact is that in both cases OH…O=CPh is the preferred
interaction, which defends the OH group from deprotonation.
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Figure 6. Theoretically predicted border cases of dimers of 1: linear (left) and sandwich (right).
27

ACCEPTED MANUSCRIPT
1
2
3
The water content is another important factor influencing the absorption spectra. The
effect is well illustrated on Figure 7. It is clear that the addition of water leads to disappearance
of the deprotonated specie(s). According to the calculations the DMSO molecule forms more
stable complexes with the 1(Id)/1(Ic) compared to the water molecule (stabilization energy 13
kcal/mol against 9 kcal/mol, the complexes are shown in Figure S14). According to the NMR
data, discussed above, the water molecules do not interact with the dye directly, but through
the DMSO molecule. This behaviour is logical taking into account Kamlet-Taft solvatochromic
parameters of both solvents [41] (π*=1.00, α=0, β=0.76 for DMSO; π*=1.09, α=1.17, β=0.4 for
water). DMSO is stronger proton acceptor solvent (β term), which could form more stable
solute solvent complex and deprotonate the OH group easier (see Figure S15). In turn the water
molecule being proton donor returns the deprotonated dye in its neutral state.
4
5
6
7
8
9
10
11
12
13
14
15
Figure 7. Absorption spectra of 1 in: dry DMSO (solid black line), DMSO as purchased (dashes),
DMSO with further addition of water (20% water to the commercial DMSO, solid gray line).
28

ACCEPTED MANUSCRIPT
1
2
Conclusions
3
4
5
6
7
8
2-(2-(2-Hydroxy-4-nitrophenyl)hydrazono)-1-phenylbutane-1,3-dione can be potentially
presented in several tautomeric forms. However, the spectral investigations in DMSO and other
solvents and the quantum chemical calculations strongly indicate that the dye exists as a
mixture of two isomers of a single tautomer. As a result of detailed assignment of the NMR
signals it was shown that the previous interpretation made by Mahmudov et al. is not correct.
The OH group, implemented in the phenyl ring, is not tautomeric for aromaticity reasons.
9
10
11
12
Having in mind that the title compound is a potential rotary switch it could be expected that the
additional OH group could influence the process of rotation or could stabilize one of the
conformers. Instead, it forms hydrogen bonds with the solvent leading to deprotonation in
diluted solutions.
13
14
15
Experimental part
Synthesis:
16
17
18
19
20
21
Reagents and solvents were purchased from Sigma-Aldrich, Acros and Alfa-Aesar, and
used without further purification. Fluka silica gel /TLC-cards 60778 with fluorescence indicator
254 nm were used for TLC chromatography. Merck silica gel 60 (0.040-0.063 mm) and basic
alumina (Brockmann activity, 0.05-0.15 mm) were used for flash chromatography purification
of the products. Melting points were determined on a Gallenkamp apparatus and are
uncorrected. HRMS LC/MS was carried out on a Bruker MicrOTOF-QII-system with ESI-source
29