Synthesis, DFT and kinetic studies of chromic S-methyldithizone

Article abstract of DOI:10.1016/j.poly.2020.114386

Dithizone is one of the most well-known trace-metal analysis reagents, however, its S-alkylated derivatives received very little attention up to date. Synthesis, kinetic studies of its photo- and chronochromic reactions, as well as DFT and TDDFT results a

Full text of DOI:10.1016/j.poly.2020.114386

Polyhedron 179 (2020) 114386
Contents lists available at ScienceDirect
Polyhedron
journal homepage: www.elsevier.com/locate/poly
Synthesis, DFT and kinetic studies of chromic S-methyldithizone
⇑
F.M.A. Muller, J. Conradie, K.G. von Eschwege
Department of Chemistry, University of the Free State, PO Box 339, Bloemfontein 9300, South Africa
a r t i c l e i n f o
a b s t r a c t
Article history:
Dithizone is one of the most well-known trace-metal analysis reagents, however, its S-alkylated deriva-
tives received very little attention up to date. Synthesis, kinetic studies of its photo- and chronochromic
reactions, as well as DFT and TDDFT results are presented here. For comparitive purposes the correspond-
ing phenylmercury complex was also synthesized and included in this study. Blue-shifts in kmax of the
S-methylated compounds are in excess of 60 nm as compared to the metal complex. Relative geometry
optimized energies, atomic charge distribution in conjunction with 1H NMR, as well as TDDFTs all point
to the pink syn geometry of S-methyldithizone as the prevalent isomer, with the yellow anti geometry
only of slightly higher (3.8 kJ/mol) energy. B3LYP provides significantly better UV–visible spectrum
approximation than LC-BLYP, OLYP and PW91. At 20 °C the rate of the chronochromic reaction is
0.0073 sÀ1 in chloroform, and that of the photochromic return reaction is 0.0023 sÀ1 in ethanol.
Ó 2020 Elsevier Ltd. All rights reserved.
Received 18 November 2019
Accepted 19 January 2020
Available online 22 January 2020
Keywords:
Photochromic
Chronochromic
Concentratochromic
Isomerization
Computational
1. Introduction
dissolution the colour in chloroform is reported to be pink [5]. Over
time, however, partial isomerization from the pink to the yellow
The (S-CH3)HDz derivative of dithizone, H2Dz, that is methy-
lated on the sulphur position, was originally prepared by Irving
et al [1]. This was accomplished in various ways; most conve-
niently by reacting dithizone in alkaline solution with (CH3)2SO4
(see Scheme 1), otherwise by the reactions of CH3I with Ag(HDz),
or NaSCH3 with 1,5-diphenyl-3-chloroformazan. Prior complexa-
tion with certain metals like Ag, Ni and Pd increases reactivity at
the sulphur position, which then readily reacts with the alkyl
group [2]. Methylation of ClHgHDz was also successful, however
only in polar medium like ethanol, while refluxing for several
hours.
Another more elaborate route by which S-alkylation is accom-
plished is to increase the nucleophilic nature of sulphur by first
oxidizing H2Dz to the cyclic dehydrodithizone compound, Dz
(Scheme 1, bottom left) [3]. Reactions of the zwitter-ionic Dz with
eg. I-CH3 and Cl-CH2COOH, even in non-polar chloroform, form
RDz+XÀ water-soluble salt products. Br-alkyls do not react with
Dz, however, addition of 1,8-diazobicyclo-[5,4,0]undec-7-ene
(DBU) had been shown to successfully catalyze the reaction [4].
Reduction of RDz+XÀ by means of alkaline solutions of dextrose
or ascorbic acid opens the tetrazolium ring again, to give the S-
functionalized dithizone, (S-R)HDz [3].
isomer was observed. Addition of trace amounts of acid or base
to pink chloroform solutions was found to greatly accelerate iso-
merization to the yellow isomer (kmax = 425 nm), while in turn,
exposing yellow solutions to light, reverses the reaction to yield
the original pink form (kmax = 550 nm).
Solid state crystal structures of both the (S-CH3)HDz and (S-
CH2COOH)HDz derivatives were published. The geometry of the
carboxylic acid ligand corresponds to L1 (see Fig. 1 syn (top), and
Table 1) [6]. Interestingly, different isomers for the methylated
derivative were reported in two separate X-ray crystallography
studies; the one geometry corresponding to the pink L1 isomer
(Fig. 1, syn), [7] and the other to the yellow L2 isomer (anti) [8].
Hutton and Irving did 1H, 13C and 15N NMR studies of S-methy-
lated dithizone in solution. Differences in chemical shifts and cou-
pling constants, and equilibria were explained in terms of four
tautomers, where intra-molecular proton transfer between back-
bone nitrogens occur, as well as rotation around the nitrogen-car-
bon bonds, i.e. the carbon that is bonded to sulphur (see various
tautomers in Table 1). Infra-red and resonance Raman studies of
this compound in general agree with these findings [10,11],
namely isomerization between ultimately the pink L1 and yellow
L2 geometries. An equilibrium between symmetrical forms of the
pink L1 isomer was also proposed, where the imine proton on
the first backbone N is intramolecularly transferred to the last N,
via the closed-ring L3 (see Table 1) intermediate.
As for (S-CH3)HDz and its ortho-tolyl (with –CH3 substituents
on the phenyl rings) homologue in solution, immediately after
In view of the natural affinity of the unsubstituted dithizone
parent compound for a large number of metals, and its consequent
⇑
Corresponding author.
E-mail address: veschwkg@ufs.ac.za (K.G. von Eschwege).
https://doi.org/10.1016/j.poly.2020.114386
0277-5387/Ó 2020 Elsevier Ltd. All rights reserved.

2
F.M.A. Muller et al. / Polyhedron 179 (2020) 114386
Scheme 1. Three synthesis pathways (left) of S-methyldithizone, (S-CH3)HDz (centre), from dithizone, H2Dz, which is then complexed with phenylmercury chloride in the
presence of a strong base. Diluted acids demethylate the complex ligand (right).
extensive application in trace metal analyses, Loredo and Grote did
an investigation into phenyl carboxy-substituted S-decyl dithizone
as solvent extractant for precious metals [4]. Where-as the decyl
substituent increased solubility in non-polar solvents, metal
extraction ability was debilitated. From a multitude of analytical
chemistry reports it may be seen that unsubstituted dithizone
often complex a variety of metals without addition of base, while
in other cases the presence of only a weak organic base is sufficient
to singly deprotonate H2Dz to its anionic form. The remaining
backbone imine proton in (S-CH3)HDz is however much less labile
and requires a strong base to for instance complex mercury, as will
be seen in the Experimental section here below.
The above-mentioned studies represent almost all the work
that had up to date been done on this class of S-alkylated com-
pounds. In the present study we resolved to do a proper DFT inves-
tigation of the chromic properties of S-methyldithizone, while also
including its phenylmercury complex. The reaction rates of the
chronochromic reaction, as well as photochromic back reaction
of (S-CH3)HDz at various temperatures was experimentally
determined.
Traditionally the photochromic behaviour of the dithizonato
metal complexes received extensive attention, with the mercury
complex having been at the centre of attention [9,12]. Here, colour
change from orange to blue (see Fig. 1 bottom) occur only as a
result of exposure to bright light, which then is instantaneously
followed by the spontaneous thermal reaction, reverting to the
orange resting state geometry. Of particular interest now is to also
investigate properties of the reaction where the metal is replaced
by an organic moiety, where addition of alkane substituents to
the sulphur atom in the ligand represents the first step in a larger
series of studies that may eventually develop into tri-dentate
ligands, or ligands with more than one photochromic moiety, etc.
2. Experimental methods
2.1. General
Synthesis reagents (Sigma-Aldrich) and solvents (Merck, AR
grade) were used without further purification. UV–visible spectra
and reaction kinetics studied on a Shimadzu UV-2550 spectropho-
tometer fitted with a Shimadzu CPS temperature controller, using
a quartz cuvette. Photochromic studies were performed using a
400 W mercury-halide lamp. 1H NMR spectra were obtained on a
300 MHz Bruker Fourier NMR spectrometer at 25 °C.
Fig. 1. Top: Pink L1 and yellow L2 isomers of S-methyldithizone, (S-CH3)HDz, as
reported in two separate X-ray crystal studies [7,8]. Bottom: Corresponding
photochromic complex blue (left) and orange (right) isomers, where instead of the
CH3 group, a metal is bonded to the ligand sulphur [9].

F.M.A. Muller et al. / Polyhedron 179 (2020) 114386
3
Table 1
Structures and DFT computed energies of the six most probable isomers of (S-CH3)HDz, and the two complex isomers of C6H5Hg(S-Me)Dz. Energies are indicated to be relative to
the lowest energy isomers, L1 or C1.
Isomer
Structure
E (kJ/mol)
B3LYP
0.0
LC-BLYP
0.0
OLYP
0.0
PW91
0.0
L1 pink
L2 yellow
3.8
5.4
3.4
4.6
L3
13.6
15.8
15.0
12.3
L4
L5
24.7
66.7
22.1
55.5
23.5
72.9
28.2
67.3
L6
C1
89.1
0.0
76.6
0.0
87.0
0.0
80.3
0.0
C2
22.3
19.8
19.9
24.3
2.2. Synthesis
and finally the organic extract was dried over sodiumsulfate. The
resulting S-methylated dithizone was purified by chromatography
on 100 g silica, using chloroform as eluent. The dark brown product
band, running ahead, was followed by narrower light brown, light
green, and pink bands. Recrystallization from warm ethanol
(700 mL) gave 3.48 g black S-methylated dithizone needles with
a bronze reflex (66%).
a. S-Methyldithizone (adapted dimethylsulfate method) [5]: A
solution of dithizone (5.0 g, 19.5 mmol) in aqueous sodium
hydroxide (0.5 M, 100 mL) was shaken well with dimethylsulfate
(1.9 mL, 2.6 g, 20.6 mmol) in a glass stoppered flask. The initial
orange colour of the solution rapidly disappeared and a black solid
separated almost immediately. A further portion of the NaOH solu-
tion (15 mL) was added, and the mixture was heated gently on a
water bath for 30 min to destroy any excess dimethylsulfate. On
cooling, the mixture was extracted with chloroform and the
organic extract was washed several times with dilute aqueous
ammonia (to remove unmethylated dithizone) and with water,
Mp 118–120 °C, kmax/nm (dichloromethane) 542, 1H NMR
(300 MHz, CDCl3) (two isomers): d 2.41 & 2.55 (3 H, 2 Â s, CH3),
7.05 – 7.97 (10 H, 3 Â m, C6H5), 9.53 & 10.27 (1 H, 2 Â s, NH).
b. S-Methyldithizone (methyliodide method) [13]: Methylio-
dide (0.05 g, 0.34 mmol) was added to potassium dithizonate
(0.05 g, 0.17 mmol) dissolved in acetone (40 mL). While stirring

4
F.M.A. Muller et al. / Polyhedron 179 (2020) 114386
the solution at room temperature for 1 h, the orange colour chan-
ged to deep purple. The reaction mixture was filtered and the sol-
vent removed under reduced pressure. The residue was dissolved
in dichloromethane (100 mL), leaving potassium iodide and any
unreacted potassium dithizonate behind. After a second filtration
the solvent was removed under reduced pressure, and recrystal-
lization done from ethanol (100 mL). The black S-methylated dithi-
zone crystals were washed with 1 M aqueous ammonia (to remove
unmethylated dithizone), followed by water, and a small amount
of ice cold ethanol to give 0.0269 g (64%) pure product.
Characterization data agreed with that of published method (a)
here above.
optimized geometries in both gas phase and ethanol as solvent. All
structures were confirmed to be true minimum structures by a fre-
quency analysis, i.e. no imaginary frequencies. Input coordinates for
the compounds were constructed using Chemcraft [27]. Optimized
coordinates of the DFT/B3LYP gas phase calculations are provided
in the Supporting Information.
3. Results and discussion
3.1. Synthesis
H2Dz reactions with metals often require the presence of base
to singly deprotonate the ligand backbone. Using the ready depro-
tonated potassium salt, K+HDzÀ instead (see Scheme 1) [28,29], is a
convenient alternative that does not require possibly interfering
bases during such reactions. The negative charge in K+HDzÀ is
located on the sulphur atom, which, during methylation with
methyliodide, makes it ideal for nucleophilic attack on the
d + methyl carbon in CH3I, yielding (S-CH3)HDz. In par. 2.2b above,
this method is reported, illustrating similar yield as in the tradi-
tional reaction done with dimethylsulphate (par. 2.2a), namely ca
65% [5]. As opposed to the newly reported KHDz route, the latter
method however requires additional product purification by col-
umn chromatography.
c. S-Methyldithizonatophenylmercury(II): Phenylmercury(II)
chloride (0.58 g, 1.85 mmol) and S-methyldithizone (0.5 g,
1.85 mmol) was dissolved in dichloromethane (100 mL) in a sepa-
rating funnel. An aqueous solution of sodium hydroxide (0.2 M,
100 mL) was added, and the funnel shaken well. The organic layer
was washed several times with distilled water and dried by filtra-
tion through sodium sulphate. The solvent was removed under
reduced pressure and recrystallization done from dichloromethane
and ethanol, yielding 0.95 g (94%) pure product.
Mp 143 °C, kmax/nm (dichloromethane) 442 & 600, 1H NMR
(300 MHz, CDCl3): d 2.53 (3 H, s, CH3), 7.02 – 7.980 (15 H, m,
3 Â C6H5).
During product characterization clear evidence for solution
equilibrium between the two structural isomers of (S-CH3)HDz
was observed in the 1H NMR spectrum obtained in deuterated
chloroform. Fig. 2 shows two methyl singulets at 2.41 and
2.55 ppm, of which the sum of the integration values add up to
the required total of 3 for the three methyl protons. The same is
true for the imine proton signals at 9.53 and 10.27 ppm, adding
up to 1, and the para-phenyl protons at ca 7.9 and 7.1 ppm, adding
up to 2. The integration ratio’s of all three these sets of protons
being 2 : 1 are indicative of a 2 :1 concentration ratio between
the L1 and L2 isomers in CDCl3 at room temperature in this partic-
ular solution.
Assignment of these signals is supported by DFT-computed
atomic charge distributions, where all the backbone nitrogens in
both isomers are partially negatively charged, as opposed to the
d-positive sulphur moiety. With the imine H in L1 close to the third
backbone N (with which hydrogen bonding is expected) this
2.3. Quantum computational methods
Density functional theory (DFT) calculations were performed uti-
lizing the Gaussian 09 calculation package [14]. Full geometry opti-
mization and TDDFT as implemented in Gaussian 09 were done. The
triple-f basis set 6-311G(d,p) [15,16] in combination with a variety
of functionals were used: B3LYP [17,18,19], BLYP [20,21], OLYP [22],
and PW91 [23]. For calculations involving Hg, the triple-f 6-311G(d,
p) basis set was used for lighter atoms (C, H, N, S), the Stuttgart/Dres-
den (SDD) pseudopotential for describing the Hg metal electronic
core, and the def2-TZVPP basis set for the metal valence electrons
[24]. Calculations were conducted in gas, as well as in the ethanol
solvent phase. For the solvent calculations the solvation model den-
sity (SMD) [25] polarizable continuum model (PCM) was used, with
an integral equation formalism variant (IEF-PCM) [26]. TDDFT calcu-
lations to generate UV–visible spectra were carried out on the
Fig. 2. 1H NMR of (S-Me)HDz in CDCl3. Relative integral values and chemical shifts at ca 2.50 and above 9.50, are indicative of two structural isomers.

F.M.A. Muller et al. / Polyhedron 179 (2020) 114386
5
proton will be significantly more shielded than in the case of the L2
isomer where it is close to the d-positive S atom. The more preva-
lent L1 signal at 9.53 ppm is therefore in perfect agreement with
the tautomer of lowest computed energy being that of the L1 iso-
mer, as will be seen in the next section. The chemical shift between
the methyl signals of the two isomers is as expected less pro-
nounced, as it does not experience such a large change in chemical
environment. A similar chemical shift was reported by Romero
et al; in the E/Z isomerization of a nitro-substitued hydrazone,
where the Z isomer shields the imine proton in the intramolecular
N–HÁÁÁÁN bond [30].
The reaction of (S-CH3)HDz with metals, eg. C6H5HgCl in this
case, requires a strong base like sodium hydroxide to remove the
imine proton, which then allows for complexation with the
(S-CH3)DzÀ anion. The green C6H5Hg(S-CH3)Dz complex yield is
almost quantitative and do not require additional purification. Care
must however be taken, as the labile methyl substituent is readily
removed by the presence of weak acid, giving the orange
C6H5HgHDz complex, see Scheme 1. Even the minute amounts of
HCl often present in deuterated chloroform may re-protonate the
ligand, as signalled by a color change from yellow-green to orange.
Efforts to grow sizeable single crystals of the methylated complex
that are suitable for X-ray crystallography were unsuccesful. It was
consequently decided to include a theoretical structure optimiza-
tion of also the phenylmercury complex in the present study.
Fig. 4. Photochromic reaction of 3.24 Â 10À5 M S-methylated dithizone in ethanol
at 20 °C, in a 1 cm cuvette, during 5 s exposure to bright light. Line colors
correspond to solution colors. (The chronochromic reaction in chloroform is
opposite to the indicated direction.)
As previously reported [5,31], (S-CH3)HDz auto-isomerizes from
the freshly dissolved pink solution to yellow over time, and thus
undergoes a form of chronochromism. The spectra of the two iso-
mers, with kmax at 540 and 410 nm and isosbestic points at 350 and
474 nm are shown in Fig. 4. This chronochromic colour change in
non-polar chloroform (
to-yellow direction as observed for the photochromic reverse reac-
tion of the same compound in ethanol ( = 25.3). The polar reaction
medium where-in photochromism of this purely organic com-
pound is visible is different to metal dithizonates, where pho-
tochromism is visible only in mostly non-polar solvents. In polar
e = 4.8 [32]) proceeds in the same pink-
3.2. Spectroscopy and kinetics
e
The UV–visible electronic spectra of intensely colored dithizone,
its derivatives and complexes, are highly sensitive to chemical
changes, eg. an acetone solution of the H2Dz starting material
has a typical blue-green colour (H2Dz: kmax = 444 & 612 nm), which
changes to orange-red when stripped of one proton (KHDz:
kmax = 502 nm). Methylation changes the colour to purple ((S-
CH3)HDz: kmax = 542 nm), while consequent complexation with
phenylmercury(II) results in a deep green colour (C6H5Hg(S-CH3)
Dz: kmax = 442 & 600 nm). Demethylation of C6H5Hg(S-CH3)Dz,
or direct complexation of Hg with H2Dz, gives the familiar orange
photochromic complex that changes colour to blue in the presence
of bright light (C6H5HgHDz: kmax = 470 M 590 nm, in DCM) [9].
Interestingly, the two absorption bands observed in the visible
spectra of both the methylated complex and unsubstituted dithi-
medium like methanol (e = 33.0) the reaction of metal dithizonates
is so fast that it may be observed only by femtosecond laser spec-
troscopy [33,34]. In non-polar solvents the back reaction of the
mercury complex is slowed down to several minutes in selected
cases [35,36]. This is vastly different to (S-CH3)HDz where polar
solvents instead stabilize the photo-induced pink isomer more so
than non-polar solvents. Even in very polar DMSO (
observed (S-CH3)HDz to be photochromic.
e = 47.2) we
The effect of replacing the Hg metal on the ligand sulphur with
an organic methyl group resulted in a blue-shift of 61 nm in ace-
tone, i.e. from 471 (orange) [35] to 410 nm (yellow), and 63 nm
for the corresponding photo-generated isomers, i.e. from 603
(blue) to 540 nm (pink). This shift is significantly more pronounced
than the 37 nm that is reported for when phenyl substituents on
the ligand are varied from strongly electron donating to strongly
electron withdrawing [34]. Exposure of an ethanolic (S-CH3)HDz
solution to direct sunlight results in a color change from yellow
to pink, which spontaneously reverses in shade, i.e. the peak at
540 nm disappears while the peak at 410 nm re-appears. The
absorbance peak shift of 130 nm between these two ground state
isomers is comparable to the 120 nm corresponding shift seen
between the blue and orange isomers of PhHgHDz.
zone overlaps almost perfectly. The molar absorptivities (
ferent though, where for H2Dz, e(612 e(444 nm), while for
e(442 nm), see Fig. 3 (left- and right-
e) are dif-
>
nm)
C6H5Hg(S-CH3)Dz, e(600
<
nm)
most absorbance peaks of both compounds).
Kinetic data of the photochromic back reaction of (S-CH3)HDz in
ethanol (pink ? yellow) is shown in Fig. 5. Photo-excitation of
dilute solutions in a quartz cuvette was accomplished with a
400 W mercury-halide lamp, which closely simulates daylight.
After apparent full color conversion in the bright light, solution
absorbance was monitored at constant temperature in a UV–visi-
ble spectrometer at the absorbance maximum of the yellow iso-
mer, namely 410 nm. This reaction is, as for the dithizonato
metal complexes, also temperature dependant, see Fig. 5 (bottom).
The plot of ln(k/T) where k is the reation rate, vs 1/T with T being
the absolute temperature, shows the expected almost perfect lin-
ear relationship (R2 = 1.00). From the Eyring equation (see insert
Fig. 3. UV/vis spectra (kmax, from left to right) of C6H5Hg(S-CH3Dz), K+HDz-, (S-CH3)
HDz (L1 pink isomer) and H2Dz, in acetone (concentrations not similar). Line colors
correspond to solution colors.

6
F.M.A. Muller et al. / Polyhedron 179 (2020) 114386
Fig. 6. The chronochromic auto-isomerization reaction of (S-Me)HDz at 20 °C in
chloroform shows first order kinetics, as illustrated by the linear plot of ln[A0/At] vs
time. Insert: Measured absorbance vs time for the appearance of the peak at 417 nm.
preferred geometry, see Table 1. The geometry optimization energy
of the linear anti L2 isomer is only slightly higher, namely 3.8 kJ/-
mol (B3LYP). These lowest and relative close energies are consis-
tent with the two different isomers found in the X-ray single
crystal studies mentioned earlier.
Relative to the energies of the former two isomers, a significant
increase in geometry-optimized energy is seen for the remaining
four isomers, varying from 13.6 kJ/mol for the L3 isomer to 89.1
kJ/mol for L6. The large energy difference between L3 and those
Fig. 5. Top: Linear correlation plots of ln[A0/At] vs time shows first order kinetics in
the spontaneous photochromic back reaction of (S-Me)HDz at different tempera-
tures in ethanol. (X – 10, j – 20, r – 30, ▲ – 40, d– 50) °C. The insert give measured
absorbance vs time for the appearance of the peak at 417 nm at 20 °C. Bottom: The
trend line through the Eyring plot shows a linear correlation with R2 = 1.00. Insert:
Eyring-Polanyi equation.
in Fig. 5, bottom) the enthalpy of activation,
from the slope term, H*/RT, while the entropy,
from the intercept term, S*/R + ln(kB/h). These values are found
D
H*, may be derived
D
D
S*, is obtained
D
to be 27 kJ/mol and À191 J/K/mol respectively.
Kinetic data of the chronochromic reaction (pink ? yellow) of a
freshly prepared pink (S-CH3)HDz solution in CHCl3 is shown in
Fig. 6. The chronochromic reaction is comparable to the above pho-
tochromic return reaction in that both represent pink L1 to yellow
L2 isomerization. At 20 °C a rate of 0.0073 sÀ1 is found for the
chronochromic reaction, which is more than three times faster
than the photochromic return reaction rate (k = 0.0023 sÀ1) at
the same temperature. This difference is attributed to the different
solvents, i.e. chronochromism for this purely organic compound is
observed in mainly non-polar solvents and photochromism in
polar solvents. The faster chronochromic rate in non-polar chloro-
form (pink ? yellow) is therefore consistent with the observation
that the photo-excited pink L1 isomer is better stabilized by polar
solvents.
3.3. DFT studies
For the first time Density Functional Theory was involved in the
study of S-alkylated dithizones. Geometry optimizations using four
different basis sets had been done on the six most likely isomers of
(S-CH3)HDz. Results from the four sets were consistent in estab-
lishing the syn L1 isomer to be of lowest energy, and thus the
Fig. 7. B3LYP TDDFT (ethanol as solvent) computed oscillators (bars) overlaid with
experimental UV–visible spectra (also in ethanol) of the pink L1 isomer (top) and
the yellow L2 isomer of (S-CH3)HDz [31]. The L1 shoulder at 420 nm (top) and L2
shoulder at 540 nm (bottom) are due to partial L1 M L2 isomerization.

F.M.A. Muller et al. / Polyhedron 179 (2020) 114386
7
of the L1 and L2 experimentally found geometries therefore
excludes L3 from room temperature equilibria, which finding is
contrary to a previous proposal in this regard (see Introduction).
The L3 configuration is however similar to the nitroformazan syn-
thesis precursor of dithizone, which consistently crystallizes in the
L3-type geometry [37–39]. Instead of the S moiety in (S-Me)HDz a
NO2 group is bonded to the backbone carbon in nitroformazan. The
latter strongly electron-withdrawing substituent thus promotes
the closed ring structure that is maintained by the N–HÁÁÁN hydro-
gen bond in this configuration.
dependant DFT/B3LYP-computed electronic oscillators give clear
confirmation of the structural isomers responsible for particular
colors in solution. From the (ethanol) TDDFTs of all six isomers
unambiguous assignments could be made: the L1 oscillators corre-
spond almost exactly to the UV–vis spectrum of the pink solution,
with eg. the computed main oscillator being at 549 nm while
experimental kmax = 542 nm, see Fig. 7 (top). The higher energy
oscillators below 400 nm correspond well to the overlap of absor-
bance bands seen in this region. The same is true for the L2 oscil-
lators where the computed maximum is at 425 nm, also only
7 nm higher than the experimental absorbance maximum at
418 nm (Fig. 7, bottom).
None of the four highest energy isomer TDDFTs (L3 – L6) resem-
ble the experimentally observed spectra of the pink or yellow solu-
tions (see Supporting Information). Also, and as opposed to the
above solvent calculations, gas phase did not give good resem-
blances; electronic transitions were blue-shifted, eg. peak maxima
were computed to be at 512 nm for L1 and 402 nm for the L2 iso-
mer. Lastly, TDDFTs computed with the PW91 and OLYP function-
als resulted in large red-shifts, often more than 100 nm higher than
The various isomers of (S-CH3)HDz are different to those of
H2Dz. Apart from isomerization around bonds, H2Dz tautomers
are also formed by intramolecular proton transfer reactions. The
first dithizone proton (that is otherwise readily substituted by a
methyl group in (S-CH3)HDz) is significantly more labile than the
remaining imine proton in S-methylated dithizone, and thus read-
ily transfers to neighboring N or S atoms, resulting in the various
tautomers of H2Dz [31].
Apart from computed relative energies, and 1H NMR in combi-
nation with computed atomic charge distribution, also time
Table 2
B3LYP gas phase molecular orbital (MO) renderings of (S-CH3)HDz L1 & L2 (top), C6H5HgHDz c1 (middle) and C6H5Hg(S-Me)Dz C1 & C2
(bottom). C–N axes of rotation are indicated by red arrows. (MOs obtained with other functionals or in solvent phase are similar to
gasphase MOs.)
Isomer
L1
HOMO
LUMO
L2
c1
C1
C2

8
F.M.A. Muller et al. / Polyhedron 179 (2020) 114386
experimentally observed maxima, while LC-BLYP computed the
same transitions more than 100 nm too low. B3LYP was therefore
the only functional that realistically resembled experimental
spectra.
ends, as may also be seen for unsubstituted H2Dz [42] and its mer-
cury complex, see Table 3. This high degree of unsaturation lies at
the basis of the high molar absorptivities, reflected in the intense
colors of dithizone, its derivatives and complexes.
As for solutions of the metal complex, C6H5Hg(S-CH3)Dz, two
absorbance bands were experimentally observed in the visible
region, peaking at 442 and 600 nm, see Fig. 2. As opposed to unsub-
stituted dithizone, H2Dz, where both visible region absorbance
bands are definitely associated with a single H2Dz isomer (see
spectrum in Fig. 2 and Ref. [40] for H2Dz discussion), it is here pro-
posed that the two bands of the C6H5Hg(S-CH3)Dz complex are
representative of different isomers existing simultaneously in solu-
tion. Also, as opposed to unsubstituted C6H5HgHDz where the
HOMO backbone C@N double bond requires photo-excitation to
the LUMO C–N single bond before rotation may occur (see Ref.
[9] for discussion), free rotation around the C6H5Hg(S-CH3)Dz
HOMO C–N single bond (see Table 2) may readily occur in solution.
TDDFT ethanol medium calculations of the C1 (see Table 1 and
Supporting Information for TDDFTs) C6H5Hg(S-CH3)Dz isomer
reveals longest wavelength electronic oscillators at 435 and
448 nm (shoulder), which corresponds almost perfectly with the
observed main experimental peak at 442 nm. Absence of lower
energy (longer wavelength) oscillators confirms the C1 isomer as
origin of this absorbance band. The TDDFT spectrum of C2 shows
the expected red-shift, namely to 509 nm. Although this shift is
less than seen experimentally (600 nm), it needs to be pointed
out that this much lower intensity band is very wide, which may
be indicative of a torsion angle that is not fixed, but enjoys greater
rotational degrees of freedom. Such would be expected in view of
the C–N bond being single, around which rotation may occur
freely. What is therefore experimentally observed, is a solution
mixture where C1 is the prevalent isomer.
Bond strengths around the Hg atom undergo a complete change
as a result of the S-methylation, with the Hg–S bond distance
increasing by ca 0.5 Å. The concomitant shortening of the Hg–N
bond shows a change from the weaker dative covalent bond in
C6H5HgHDz to a direct covalent bond in the S-methylated complex.
The S–C bond distance of 1.77 Å nevertheless stays largely unaf-
fected whether methylated or not.
Lastly it should be noted that a clear change in N–C–S–CH3 rota-
tional angle between the L1 and L2 isomers are seen in the respec-
tive crystal structures (see Fig. 8 A & C). The L1 ground state isomer
is almost perfectly planar, with the methyl group being rotated by
only 5.72° out of the flat plain formed by the rest of the molecule.
This orientation appears to be sterically favored, but may also be
due to crystal packing effects, since this co-planar orientation is
not supported by DFT single molecule calculations (Fig. 8 B). In
the L2 photo-induced linear structure (C), however, the methyl
Orbital renderings of (S-CH3)HDz clearly show simularity
between the two photochromic compounds, with this ligand and
the traditionally well-known photochromic C6H5HgHDz complex
having similar electron distribution patterns along its photo-active
backbones. In all cases significant electron density is, apart from
along the backbone, also located on the protruding sulphur atom
in the HOMOs. This is however not the case for any of the LUMOs,
which is indicative of limited but directional intramolecular charge
transfer away from the sulphur atom upon excitation of the HOMO.
Bond distances in (S-CH3)HDz are theoretically simulated with
a large degree of accuracy, differing from available experimental
bond lengths by at most 0.04 Å, in some cases as little as
0.004 Å, see Table 3. Where as in general N–N and C–N single bond
distances are ca 1.41 and 1.45 Å and the respective double bonds
are 1.23 and 1.28 Å [41], the bond distances along the N and C
backbone of the (S-CH3)HDz ligand, both free and complexed, all
Fig. 8. Experimental (A & C) and computed (B & D) (S-CH3)HDz geometries of the
pink L1 (top) and linear yellow L2 (bottom) isomers. The backbone imine protons
(H) are indicated.
lie in between these values. This is indicative of p-electron delocal-
ization, which also includes the aromatic phenyl rings at opposite
Table 3
Selected X-ray single crystal and B3LYP-calculated bond lengths (Å) of the two isomers of (S-CH3)HDz and C6H5Hg(S-CH3)Dz. For comparison purposes bond lengths of the
photochromic C6H5HgHDz are also included. Nitrogens are numbered along the ligand backbone, with N1 carrying the imine proton.
X-ray
DFT
(S-CH3)HDz
L1 [7]
H2Dz [28]
(S-CH3)HDz
L1
PhHgHDz
c1
PhHg(S-CH3)Dz
C1
Bond
L2 [8]
L2
c2
C2
Hg-S
–
–
–
–
–
–
–
2.399
2.677
–
1.778
1.316
1.312
1.378
1.261
2.380
2.739
–
1.779
1.317
1.315
1.376
1.265
2.908
2.146
1.833
1.803
1.263
1.383
1.304
1.306
2.950
2.136
1.830
1.793
1.262
1.397
1.295
1.313
Hg-N1
S-CH3
S-C
N1-N2
N2-C
C-N3
N3-N4
1.790
1.758
1.344
1.300
1.391
1.266
1.859
1.762
1.337
1.368
1.328
1.300
1.833
1.786
1.319
1.308
1.389
1.262
1.836
1.808
1.315
1.302
1.387
1.260
1.712
1.295
1.345
1.334
1.299

F.M.A. Muller et al. / Polyhedron 179 (2020) 114386
9
[4] M.G. Sanchez-Loredo, M. Grote, Solvent Extraction and Ion Exchange 18 (2000)
55.
[5] A.T. Hutton, H.M.N.H. Irving, J. Chem. Soc., Perkin Trans. II (1982) 1117.
[6] A.T. Hutton, H.M.N.H. Irving, L.R. Nassimbeni, G. Gafner, Acta Crystallogr., Sect.
B: Struct. Crystallogr. Cryst. Chem. 35 (1979) 1354.
[7] Von J. Preuss, A. Gieren, Acta Cryst. B31 (1975) 1276.
[8] F.H. Herbstein, W. Schwotzer, J. Chem. Soc., Perkin Trans. 2 (1984) 1917.
[9] K.G. von Eschwege, J. Conradie, J.C. Swarts, J. Phys. Chem. A 112 (2008) 2211.
[10] O. Yamada, H. Hiura, T. Igarashi, N. Kaneko, H. Takahashi, Spectrochimica Acta
44A (1988) 1409.
[11] H. Takahashi, O. Yamada, H. Isaka, T. Igarashi, N. Kaneko, J. Raman Spectr. 19
(1988) 301.
[12] L.S. Meriwether, E.C. Breitner, C.L. Sloan, J. Am. Chem. Soc. 87 (1965) 4441.
[13] K.G. von Eschwege, J.C. Swarts, Polyhedron 29 (2010) 1727.
[14] M.J. Frisch, et al, Gaussian 09, Revision D.01, Gaussian, Inc., Wallingford, CT
(2009).
[15] A.D. McLean, G.S. Chandler, J. Chem. Phys. 72 (1980) 5639.
[16] K. Raghavachari, J.S. Binkley, R. Seeger, J.A. Pople, J. Chem. Phys. 72 (1980) 650.
[17] A.D. Becke, Phys. Rev. A 38 (1988) 3098.
[18] C. Lee, W. Yang, R.G. Parr, Phys. Rev. B 37 (1988) 785.
[19] B. Miehlich, A. Savin, H. Stoll, H. Preuss, Chem. Phys. Lett. 157 (1989) 200.
[20] B.G. Johnson, P.M.W. Gill, J.A. Pople, J. Chem. Phys. 98 (1993) 5612.
[21] T.V. Russo, R.L. Martin, P.J. Hay, J. Chem. Phys. 101 (1994) 7729.
[22] N.C. Handy, A.J. Cohen, Mol. Phys. 99 (2001) 403.
[23] J.P. Perdew, J.A. Chevary, S.H. Vosko, K.A. Jackson, M.R. Pederson, D.J. Singh, C.
Fiolhais, Phys. Rev. B 46 (1992) 6671;
group is rotated by 85.21° out of the plain, which is in agreement
with ab initio DFT approximation (D).
4. Conclusion
(S-CH3)HDz was obtained as pure product by reaction of the
KHDz dithizonate salt with CH3I. Instead of metalation at the sul-
phur position, methylation also yields a photochromic product,
however with different solvent behavior; the photochromic back
reaction in the metal complex is slowed down in non-polar sol-
vents, while in the methylated compound it is slowed down in
polar solvents. DFT clarified the geometries of both (S-CH3)HDz
and PhHg(S-CH3)Dz unambigiously, eg. by determining geometries
of isomers in solution and thereby also the origins of its observed
chromisms. The colors of these compounds, as seen in its photo-,
chrono- as well as concentratochromisms, are readily corroborated
by time-dependant DFT calculations, and are very sensitive to syn-
thetic and physical alterations.
Declaration of Competing Interest
J.P. Perdew, J.A. Chevary, S.H. Vosko, K.A. Jackson, M.R. Pederson, D.J. Singh, C.
Fiolhais, Erratum: Phys. Rev. B 48 (1993) 4978.
[24] R.A. Weigend, Phys. Chem. Chem. Phys. 7 (2005) 3297.
[25] A.V. Marenich, C.J. Cramer, D.G. Truhlar, J. Phys. Chem. B 113 (2009) 6378.
[26] R.E. Skyner, J.L. McDonagh, C.R. Groom, T. van Mourik, J.B.O. Mitchell, Phys.
Chem. Chem. Phys. 17 (2015) 6174.
The authors declare that they have no known competing finan-
cial interests or personal relationships that could have appeared
to influence the work reported in this paper.
[27] http://www.chemcraftprog.com/ (2019).
[28] M. Laing, J. Chem. Soc., Perkin Trans. 2 (1977) 1248.
[29] H. Noda, L.M. Engelhardt, J.M. Harrowfield, C. Pakawatchai, J.M. Patrick, A.H.
White, Bull. Chem. Soc. Jpn. 58 (1985) 2385.
Acknowledgments
[30] E.L. Romero, R.F. D’Vries, F. Zuluaga, M.N. Chaur, J. Braz. Chem. Soc. 26 (2015)
1265.
[31] L.L.A. Ntoi, B. Buitendach, K.G. von Eschwege, J. Phys. Chem. A 121 (2017)
9243.
[32] Handbook of Physics and Chemistry (2007) 87th ed. Taylor and Francis Group
LLC.
[33] H. Schwoerer, K.G. von Eschwege, G. Bosman, P. Krok, J. Conradie,
ChemPhysChem (2011) 2653.
This work is based upon research supported by the South Afri-
can National Research Foundation (Grants 113327 and 96111) and
the University of the Free State. The CHPC of South Africa and the
High Performance Computing facility of the UFS are acknowledged
for computer time.
[34] K.G. von Eschwege, G. Bosman, J. Conradie, H. Schwoerer, J. Phys. Chem. A 118
(2014) 844.
Appendix A. Supplementary data
[35] K.G. von Eschwege, J. Photochem & Photobiol. 252 (2013) 159.
[36] E. Alabaraoye, K.G. von Eschwege, N. Loganathan, J. Phys. Chem. A 118 (2014)
10894.
[37] K.G. von Eschwege, F. Muller, E.C. Hosten, Acta Cryst. E68 (2012) o199.
[38] K.G. von Eschwege, E.C. Hosten, A. Muller, Acta Cryst. E68 (2012) o425.
[39] K.G. von Eschwege, F. Muller, T.N. Hill, Acta Cryst. E68 (2012) o609.
[40] K.G. von Eschwege, J. Conradie, A. Kuhn, J. Phys. Chem. A 115 (2011) 14637.
[41] F.H. Allen, O. Kennard, D.G. Watson, L. Brammer, A.G. Orpen, R. Taylor, J. Chem.
Soc. Perkin Trans. I I (1987) S1–S19.
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.poly.2020.114386.
References
[1] H.M.N.H. Irving, C.F. Bell, J. Chem. Soc. (1954) 4253.
[2] H.M.N.H. Irving, A.H. Nabilsi, S.S. Sahota, Anal. Chimica Acta 67 (1973) 135.
[3] J.W. Ogilvie, A.H. Corwin, J. Am. Chem. Soc. 83 (1961) 5023.
[42] K.G. von Eschwege, Polyhedron 39 (2012) 99.