Synthesis and kinetics of electronically altered photochromic dithizonatophenylmercury(II) complexes

Article abstract of DOI:10.1016/j.jphotochem.2012.11.009

A series of phenyl-substituted dithizones were synthesized, and preparation of the corresponding series of photochromic phenylmercury(II) complexes is for the first time reported. Adaption of previous methods enhanced synthesis convenience and product yields. A single crystal X-ray data collection of the ortho-S-methyl nitroformazan reaction intermediate was done. ADF computed molecular orbitals of the title compound show the HOMO and LUMO orbitals stretching along the entire ligand. The spontaneous back reaction kinetics of dithizonatophenylmercury(II) was studied at varied concentrations, temperatures and in different solvents. An exponential correlation was found between the rate of reverse isomerization and temperature, while increased solvent polarity and decreased molar mass facilitate higher return rates. The kinetic study of the series of twelve electronically altered complexes yielded a lowest rate of 0.0002 s^-1 for the ortho-methyl derivative, while the highest rate of 0.0106 s^-1 was measured for the meta-methoxy derivative.

Full text of DOI:10.1016/j.jphotochem.2012.11.009

Journal of Photochemistry and Photobiology A: Chemistry 252 (2013) 159–166
Contents lists available at SciVerse ScienceDirect
Journal of Photochemistry and Photobiology A:
Chemistry
journal homepage: www.elsevier.com/locate/jphotochem
Synthesis and kinetics of electronically altered photochromic
dithizonatophenylmercury(II) complexes
Karel G. von Eschwege∗
Department of Chemistry, PO Box 339, University of the Free State, 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:
A series of phenyl-substituted dithizones were synthesized, and preparation of the corresponding series of
photochromic phenylmercury(II) complexes is for the first time reported. Adaption of previous methods
enhanced synthesis convenience and product yields. A single crystal X-ray data collection of the ortho-
S-methyl nitroformazan reaction intermediate was done. ADF computed molecular orbitals of the title
compound show the HOMO and LUMO orbitals stretching along the entire ligand. The spontaneous back
reaction kinetics of dithizonatophenylmercury(II) was studied at varied concentrations, temperatures
and in different solvents. An exponential correlation was found between the rate of reverse isomerization
and temperature, while increased solvent polarity and decreased molar mass facilitate higher return
rates. The kinetic study of the series of twelve electronically altered complexes yielded a lowest rate of
0.0002 s−1 for the ortho-methyl derivative, while the highest rate of 0.0106 s−1 was measured for the
meta-methoxy derivative.
Received 22 August 2012
Accepted 20 November 2012
Available online 5 December 2012
Keywords:
Photochromism
Nitroformazan
Crystal structure
Ammonium disulphide
Dithizone
UV–vis spectrocopy
ADF
© 2012 Elsevier B.V. All rights reserved.
1. Introduction
product, no data was provided on the success of the reaction
in terms of yield. In the current study the reaction conditions
In 1878 already, Emil Fischer noticed formation of the ␤-
phenyldithiocarbazic acid phenylhydrazine salt, 2, during mixing
of carbon disulphide and phenylhydrazine in organic solvents (see
Scheme 1) [1]. After loss of hydrogen sulphide during heating, the
thiocarbazide (H4Dz), 6, was obtained. Addition of dilute alkali
yielded the deep orange-red anion, HDz−, with the final black-green
dithizone product (H2Dz), 7, precipitating out on addition of dilute
acid [2].
of a series of eleven electronically altered phenyl-substituted
dithizones and corresponding phenylmercury(II) complexes were
examined, and characterization data and yields are reported.
Apart from dithizone being a very important colour devel-
opment reagent in trace metal analyses, as may be noted from
numerous related reports published annually, it also forms the basis
of a marked photochromic reaction in some metal complexes (see
Scheme 2). Photo-excitation by blue-green light results in colour
Billman and Cleland [3] introduced solvent free conditions based
on the Fischer method, achieving an overall dithizone yield of
52–66%.
changing isomerization around the
taneous thermal back-reaction is unaffected by light.
C
N
bond, while the spon-
Around 1950, Irving [9] and Webb [10] independently observed
photochromic behaviour in the bis-dithizonatomercury(II) com-
plex, Hg(HDz)2. Fifteen years later Meriwether et al. did a thorough
investigation which led to the discovery of more photochromic
metal dithizonates; Pt, Pd, Cd, Zn, Ag, Pb and Bi [11]. Failure to
observe spectral changes in other metal dithizonates was ascribed
to the possibility of very fast return rates. A kinetic study in ben-
zene involving the back reaction of Hg(HDz)2, which was observed
to have the longest half-life of all metal dithizonates, probed the
effects of deuteration and water presence [12]. N-deuteration of
the complex produced a threefold decrease in return rate, while
the presence of water increased the return rate.
In 1943, Hubbard and Scott [4] published a method similar to
an earlier method reported by Bamberger [5]. This method does
away with initial preparation of 2, but instead starts from the aryl-
diazonium chloride, 4, the latter being prepared by diazotization
of the arylamine with sodium nitrite. Addition of nitromethane
precipitates the deep red nitroformazan, 5. Bubbling ammonia and
hydrogen sulphide gases through the nitroformazan solution yields
the unstable dirty white thiocarbazide, 6. Pelkis et al. [6] used a
similar method to that of Bamberger [5] and Tarbell [7], while
Mirkhalaf [8] employed ammonium sulphide instead of NH3 and
H2S gases during conversion of 5 to 6. Although he obtained a pure
The initial photochromic reaction of dithizonatophenylmer-
cury(II), DPM, in solution was recently investigated by femtosecond
transient absorption spectroscopy [13]. Ultrafast excitation takes
place within less than100 fs, followed by 180◦ photo-isomerization
∗
Tel.: +27 051 4012923; fax: +27 051 4446384.
E-mail address: veschwkg@ufs.ac.za
1010-6030/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.jphotochem.2012.11.009

160
K.G. von Eschwege / Journal of Photochemistry and Photobiology A: Chemistry 252 (2013) 159–166
H
N
H
H
H
N
H
N
S
H
N
H
N
H
N
H
S
2
+
CS2
2
6
1
-H2S
Fischer
Billman
H
N
S
H
H
N
S
H
N
N
oxidation
N
H
N
H
N
N
7
Tarbell
NH3 & H2S
or (NH4)2S
Bamberger
Hubbard
Pelkis
Cl N
2
NH2
N
Mirkhalaf
N
N
N
O
O
H
CH3NO2
HCl
NaNO2
N
2
N
5
3
4
Scheme 1. Different approaches for the synthesis of dithizone, 7, as reported by various authors.
with a time constant of 1.5 ps. Excited-state absorption was
observed for the orthogonally (90◦) twisted intermediate state,
from where bifurcation along pathways towards the ground states
of the orange cis and blue trans configurations occurs below the
funnel of the conical intersection. Faster than may be seen pho-
tochromism of DPM in a polar solvent like methanol was also for
the first time observed by employing ultrafast laser spectroscopy.
The structure of the DPM orange resting state was solved by
means of X-ray crystallography [14]. DFT calculations convincingly
resolved ambiguity about the structure of the photo-induced blue
form of DPM [15]. Whereas intra-molecular proton transfer was
previously anticipated [16], geometry optimizations established
the imine proton to stay intact during the photochromic reaction,
with no geometry changes apart from mere isomerization around
the
C
N
bond (Scheme 2).
Comprehensive cyclic voltammetry studies of the redox path-
ways of both the free ligand [17] and the phenylmercury complex
[18] were recently reported. The latter also revealed electrochromic
behaviour, while dithizone itself is now known to be both solva-
tochromic and concentratochromic under certain conditions [19].
2. Experimental methods
2.1. General
Unless otherwise stated, dithizone, 7, and all synthesis reagents
and solvents (Sigma–Aldrich, Merck) were used without further
purification. Reaction solutions were cooled on a KP250 cooling
stirrer during syntheses. Water was doubly distilled. 1H NMR spec-
tra at 20 ◦C were recorded on Bruker Avance DPX 300 and Bruker
Avance II 600 NMR spectrometers at 300 and 600 MHz respectively,
with chemical shifts presented as ı values referenced to SiMe4
at 0.00 ppm. C6D6 and CDCl3 solvents were used. CDCl3 was rid
of acid by passing through basic alumina immediately before use.
Melting points were determined on a Reichert Thermopan micro-
scope with a Koffler hot-stage, and are uncorrected. UV–visible
measurements were made on a Shimadzu UV-2550 spectrometer
fitted with a Shimadzu CPS temperature controller, using an opti-
cal glass cuvette. Photo-excitation of the photochromic compounds
was accomplished with a 400 W mercury–halide lamp.
ADF (Amsterdam Density Functional) calculations were carried
out using DFT (Density Functional Theory) [20] with the PW91
(Perdew-Wang, 1991) exchange and correlation functional [21].
Optimized geometries were calculated for the neutral species and
reported elsewhere [15]. Molecular orbital representations were
Scheme 2. Photochromism of the 4-SCH3 derivative of dithizonatophenylmer-
cury(II), PhHg(4-SCH3)HDz.

K.G. von Eschwege / Journal of Photochemistry and Photobiology A: Chemistry 252 (2013) 159–166
161
1H NMR (C6D6, 300 MHz) (ppm): 2.12 (s, 6H, 2 × CH3), 6.72–6.99 (m,
6H, 2 × C6H4), 8.09–8.16 (d, 2H, 2 × C6H4), 13.28 (s, 2H, 2 × NH).
processed and visualized with the Chemcraft software package,
applied to the former output files [22].
2.2.2.5. 3-Methyldithizone. Nitroformazan yield: 53%, dithizone
yield: 23%, mp. 149–151 ◦C, ꢀmax (nm) (dichloromethane) 448
and 613, 1H NMR (CDCl3, 300 MHz) (ppm): 2.46, 2.48 (2 × s, 6H,
2 × CH3), 7.37–7.49 (m, 4H, 2 × C6H4), 7.70–7.85 (dd, 4H, 2 × C6H4).
2.2. Synthesis
2.2.1. General procedure
Formazans: The phenyl-substituted aniline (30 mmol) is added
to a mixture of concentrated hydrochloric acid (15 mL) and water
(25 mL) in a 100 mL beaker at 0 ◦C on a cooling stirrer. Sodium
nitrite (3.1 g, 45 mmol) is slowly added while stirring, reacting with
the aniline to form a clear solution (ca 30 min). The diazo solution
is added to a mixture of sodium acetate (60 g), glacial acetic acid
(35 mL) and water (20 mL) in a 500 mL beaker (stir the dry salt
vigorously while adding the acid–water mixture). Nitromethane
(8.6 mL, 150 mmol) is added after 10 min. After stirring for 2 h at
room temperature the volume is increased with water to ca 300 mL.
The red formazyl precipitate is filtered in a large Büchner funnel and
washed with copious amounts of water. The nitroformazan may be
dried overnight at 50 ◦C.
2.2.2.6. 4-Methyldithizone. Nitroformazan yield: 88%, dithizone
yield: 30%, mp. 98–103 ◦C, ꢀmax (nm) (dichloromethane) 452 and
622, 1H NMR (CDCl3, 300 MHz) (ppm): 2.43 (s, 6H, 2 × CH3),
7.32–7.92 (m, 8H, 2 × C6H4).
2.2.2.7. 3,4-Dimethyldithizone. Nitroformazan yield: 78%, dithi-
zone yield: 13%, mp. 129–131 ◦C, ꢀmax (nm) (dichloromethane) 453
and 623, 1H NMR (C6D6, 300 MHz) ı 1.89–1.91 (2 × s, 12H, 4 × CH3),
6.76 – 6.79 (d, 2H, 2 × C6H3), 7.11–7.19 (2 × s, 4H, 2 × C6H3), 13.04
(s, 2H, 2 × NH).
Dithizones: The nitroformazan (3 mmol) is suspended in abso-
lute ethanol (60 mL) in a 150 mL lightly stoppered Erlen Meyer
flask at 35 ◦C. While stirring, ammonium sulphide (20% aq, 6 mL,
17.6 mmol) is added. After 20 min (red colour fades, or precipitate
forms) the mixture is poured onto crushed ice (100 mL) and a lit-
tle water added to promote precipitation. The mostly dirty-white
thiocarbazide is immediately filtered off in a sintered glass funnel
and washed with water. Methanolic potassium hydroxide (2%) is
intermittently poured onto the precipitate in the funnel (no suc-
tion), oxidizing it to the orange-red potassium salt, dissolving and
washing it into a 250 mL flask. The solution is acidified with dilute
HCl (2%) until full precipitation of the black dithizone product is
effected, which then is gravity filtered through a filter paper and
washed with copious amounts of water. For purification purposes
the base–acid treatment is repeated.
2.2.2.8. 2-Fluorodithizone. Nitroformazan yield: 32%, dithizone
yield: 43%, mp. 114 ◦C, ꢀmax (nm) (dichloromethane) 450, 621,
1H NMR (CDCl3, 300 MHz) ı 7.41–7.29 and 8.02–8.10 (2 × m, 8H,
C6H4F).
2.2.2.9. 3-Fluorodithizone. Nitroformazan yield: 74%, dithizone
yield: 25%, mp. 112 ◦C, ꢀmax (nm) (dichloromethane) 447, 611,
1H NMR (CDCl3, 300 MHz) ı 7.03–7.16 and 7.35–7.63 (2 × m, 8H,
C6H4F).
2.2.2.10. 4-Fluorodithizone. Nitroformazan yield: 80%, dithizone
yield: 48%, mp. 174 ◦C, ꢀmax (nm) (dichloromethane) 446, 610, 1
H
NMR (CDCl3, 300 MHz) ı 7.24 (m, 4H, o-C6H4F), 7.73 (m, 4H, m-
C6H4F).
Dithizonatophenylmercury(II) complexes: Triethylamine (0.2 mL,
1.4 mol) is added to
a solution of the dithizone (0.8 mmol)
2.2.2.11. 2-S-methyldithizone. Nitroformazan yield: 93%, dithizone
yield: 76%, mp. 118 ◦C, ꢀmax (nm) (dichloromethane) 477 and
643, 1H NMR (CDCl3, 300 MHz) ı 2.45–2.61 (2 × s, 6H, 2 × SCH3),
6.86–8.07 (m, 8H, 2 × C6H4), 13.37 (s, 2H, 2 × NH).
and phenylmercury(II) chloride (0.2694 g, 0.82 mmol) in
dichloromethane (50 mL). After stirring (15 min) and com-
plete colour change from dark green to bright orange-red the
solvent is removed under reduced pressure. The product is again
dissolved in minimum dichloromethane and filtered through a
small filter paper into a 25 mL Erlen Meyer flask. Ethanol (10 mL)
is added to the filtrate and the product allowed to crystallize. (The
2-OCH3, 2-SCH3 and 4-SCH3 derivative products required addition
of methanol instead of ethanol, as well as a few drops of water to
aid crystallization.)
2.2.2.12. 4-S-methyldithizone. Nitroformazan yield: 87%, dithizone
yield: 60%, mp. 85 ◦C, ꢀmax (nm) (dichloromethane) 486 and 667, 1H
NMR (CDCl3, 300 MHz) ı 2.54–2.58 (2 × s, 6H, 2 × SCH3), 7.29–7.38
(m, 4H, 2 × C6H4), 7.61–7.70 (m, 4H, 2 × C6H4), 12.53 (s, 2H, 2 × NH).
2.2.3. Dithizonatophenylmercury complexes
2.2.3.1. Dithizonatophenylmercury(II). Yield: 81%, mp. 166 ◦C, ꢀmax
(nm) (dichloromethane) 470, 1H NMR (CDCl3, 300 MHz) ı 7.60–8.00
(3 × m, 15 H, C6H5).
2.2.2. Dithizone derivatives
2.2.2.1. 2-Methoxydithizone. Nitroformazan yield: 95%, dithizone
yield: 76%, mp. 178–179 ◦C, ꢀmax (nm) (dichloromethane) 474 and
641, 1H NMR (CDCl3, 300 MHz) ı 4.00 (s, 6H, OCH3), 6.97–7.17 (m,
6H, 2 × C6H4), 7.95–8.02 (d, 2H, 2 × C6H4), 12.92 (s, 2H, 2 × NH).
2.2.3.2. 2-Methoxydithizonatophenylmercury(II). Yield: 84%, mp.
212–213 ◦C, ꢀmax (nm) (dichloromethane) 498, 1H NMR (CDCl3,
300 MHz) ı 3.68 and 4.03 (2 × s, 6H, 2 × CH3), 6.57–7.89 (m, 13H,
2 × C6H4, 1 × C6H5), 9.75 (s, 1H, 1 × NH).
2.2.2.2. 3-Methoxydithizone. Nitroformazan yield: 73%, dithizone
yield: 46%, mp. 148–149 ◦C, ꢀmax (nm) (dichloromethane) 453 and
625, 1H NMR (C6D6, 300 MHz) ı 3.17 (s, 6H, 2 × OCH3), 6.60–6.92
(m, 8H, 2 × C6H4), 13.03 (s, 2H, 2 × NH).
2.2.3.3. 3-Methoxydithizonatophenylmercury(II). Yield: 63%, mp.
137–139 ◦C, ꢀmax (nm) (dichloromethane) 480, 1H NMR (C6D6,
300 MHz) ı 3.21 and 3.34 (2 × s, 6H, 2 × CH3), 6.67–7.87 (m, 13H,
2 × C6H4, 1 × C6H5), 9.52 (s, 1H, 1 × NH).
2.2.2.3. 4-Methoxydithizone. Nitroformazan yield: 59%. Dithizone
synthesis failed, as the product decomposed during the last syn-
thesis step.
2.2.3.4. 2-Methyldithizonatophenylmercury(II). Yield: 63%, mp.
120–122 ◦C, ꢀmax (nm) (dichloromethane) 464, 1H NMR (CDCl3,
300 MHz) ı 2.46 and 2.54 (2 × s, 6H, 2 × CH3), 6.98–7.82 (m, 13H,
2 × C6H4, 1 × C6H5), 9.20 (s, 1H, 1 × NH).
2.2.2.4. 2-Methyldithizone. Nitroformazan yield: 75%, dithizone
yield: 74%, mp. 162.8 ◦C, ꢀmax (nm) (dichloromethane) 457 and 623,

162
K.G. von Eschwege / Journal of Photochemistry and Photobiology A: Chemistry 252 (2013) 159–166
2.2.3.5. 3-Methyldithizonatophenylmercury(II). Yield: 75%, mp.
170–173 ◦C, ꢀmax (nm) (dichloromethane) 476, 1H NMR (C6D6,
300 MHz) ı 2.08 and 2.19 (2 × s, 6H, 2 × CH3), 6.76–8.04 (m, 13 H,
2 × C6H4, 1 × C6H5), 9.41 (s, 1H, 1 × NH).
improvements, the 4-methoxy and 2,4,6-trimethyl derivatives
could not be synthesized. Of the 2,4,6-trimethyl species only small
amounts of the typical brick-red formazan product formed, and
only at elevated temperatures of up to 35 ◦C. Heating to 40 ◦C
caused blackening of the product. Although the 4-methoxy deriva-
tive gave a 59% nitroformazan yield, both this and the trimethyl
compounds decomposed during the last step where the thiocar-
bazide (6) is oxidized to the orange-red K+HDz− salt. During the
last step the otherwise typical intense orange-red colour disap-
peared. This finding indicates the boundary at which even stronger
electron donating substituents may no longer be employed. At
the opposite end, i.e. the ‘electron-withdrawing boundary’, it was
found that the transition from one to more fluoro-substituents
on each phenyl ring results in gradual loss of product colour,
and consequently also its application in photochromic reactions
[23].
Complexation was accomplished by a convenient procedure
whereby triethylamine was employed to deprotonate dithizone
before reacting with the mercury salt [15]. Reactions with all
the ligand derivatives take place instantaneously. Recrystallization
from dichloromethane and ethanol allows the product to pre-
cipitate out, while keeping the small excess of PhHgCl and the
(CH3CH2)3NH+Cl− byproduct dissolved in the remaining ethanol.
Although the photochromic dithizonatomercury complex was
discovered more than 60 years ago [9], derivatization of this
class of photochromic metal complexes is for the first time
explored.
2.2.3.6. 4-Methyldithizonatophenylmercury(II). Yield: 87%, mp.
146–148 ◦C, ꢀmax (nm) (dichloromethane) 482, 1H NMR (3 C6D6,
300 MHz) ı 2.02 and 2.18 (2 × s, 6H, 2 × CH3), 6.87–8.12 (m, 13H,
2 × C6H4, 1 × C6H5), 9.44 (s, 1H, 1 × NH).
2.2.3.7. 3,4-Dimethyldithizonatophenylmercury(II). Yield: 65%, mp.
180 ◦C, ꢀmax (nm) (dichloromethane) 485, 1H NMR (CDCl3,
300 MHz) ı 2.21, 2.2988, 2.294 and 2.33 (4 × s, 12H, 4 × CH3),
7.12–7.80 (m, 11H, 1 × C6H5, 2 × C6H3), 9.15 (s, 1H, 1 × NH).
2.2.3.8. 2-Fluorodithizonatophenylmercury(II). Yield: 76%, mp.
163 ◦C, ꢀmax (nm) (dichloromethane) 468, 1H NMR (CDCl3,
300 MHz) ı 6.99–7.88 (m, 13 H, 2 × C6H4F and 1 × C6H5).
2.2.3.9. 3-Fluorodithizonatophenylmercury(II). Yield: 64%, mp.
143 ◦C, ꢀmax (nm) (dichloromethane) 475, 1H NMR (CDCl3,
300 MHz) ı 6.75–7.79 (m, 13 H, 2 × C6H4F and 1 × C6H5)
2.2.3.10. 4-Fluorodithizonatophenylmercury(II). Yield: 74%, mp.
208 ◦C, ꢀmax (nm) (dichloromethane) 471, 1H NMR (CDCl3,
300 MHz) ı 7.06–7.99 (m, 13 H, 2 × C6H4F and 1 × C6H5)
2.2.3.11. 2-S-methyldithizonatophenylmercury(II). Yield: 73%, mp.
130 ◦C, ꢀmax (nm) (dichloromethane) 498, 1H NMR (CDCl3,
300 MHz) ı 2.31 and 2.52(2 × s, 3H, 1 × SCH3), 6.99–7.84 (m, 13H,
1 × C6H5 and 2 × C6H4), 10.29 (s, 1H, 1 × NH).
3.2. Structure
Towards supplementary characterization, attempts were made
to grow crystals of all the derivatives of nitroformazan, dithizone
and the dithizonatophenylmercury complexes. Sizeable nitrofor-
mazan crystals were readily produced, while dithizone crystals
large enough for single crystal X-ray crystallography were not
obtained. The relative ease by which crystals of especially the ortho
nitroformazan derivatives are produced is not only illustrated by
the recent publication of crystal structures of the ortho-methyl
[24], ortho-methoxy [25] and ortho-S-methyl [26] nitroformazans,
but also by the fact that two different polymorphs of the latter
compound were obtained from different solvent systems. Ace-
tone overlaid with hexane yielded the latter monoclinic crystal
system in the P21/c space group (Z = 4), while the structure that
is reported here,1 was found to be a triclinic crystal in the P−1
space group (Z = 2), grown from a tetrahydrofuran-methanol solu-
tion (see Fig. 1). Performing the data collection at both 298 K and
200 K without observing any significant change eliminated the
possibility of the observed polymorphism being a temperature
effect.
2.2.3.12. 4-S-methyldithizonatophenylmercury(II). Yield: 72%, mp.
176 ◦C, ꢀmax (nm) (dichloromethane) 509, 1H NMR (CDCl3,
300 MHz) ı 2.53 (s, 6H, 2 × SCH3), 7.10–7.95 (m, 13H, 1 × C6H5 and
2 × C6H4), 9.20 (1H, s, 1 × NH).
3. Results and discussion
3.1. Synthesis
Syntheses of the dithizone ligands were done according to
an adapted method by Mirkhalaf, Whittaker and Schiffrin [8].
This method conveniently allows the use of a series of phenyl-
substituted anilines as starting reagents, from which the intended
electronically altered series of photochromic mercury complexes
could for the first time be synthesized.
Low yields of some intermediary nitroformazan and dithi-
zone derivatives necessitated revising certain reaction conditions.
It was consequently found that increasing the sodium nitrite,
nitromethane and ammonium sulphide excess quantities signifi-
cantly improved product yields, see Scheme 1 (bottom path) and
the general synthesis description in Section 2.2. A slightly higher
solution temperature (35 ◦C) during addition of (NH4)2S also gave
increased reaction rates and higher thiocarbazide yields. An addi-
tional increase in the reaction rate at even higher temperatures
(>40 ◦C) was observed, but it simultaneously lowered the yields of
the relatively unstable thiocarbazides. The main advantage of using
(NH4)2S lies in the abolition of the inconvenient use of noxious NH3
and H2S gasses, traditionally employed in the substitution reaction
of NO2 with S.
The similar conformation of all reported nitroformazan struc-
tures is ascribed to the intramolecular H-bond interaction between
the single imine proton, (N11)H, and N13 which is distanced at
˚
˚
1.976 A, and to a lessor degree N14, which is situated 2.566 A from
the hydrogen atom. The bent conformation differs from the linear
conformation of dithizone, where the two imine protons are each
bound to a nitrogen bordering a phenyl ring, and thus sterically pre-
venting a similar bent conformation (see Scheme 1 (5 and 7) [19]).
The nitroformazan molecule is otherwise largely planar, with the
NO2 and S(12) groups being coplanar to the molecular formazan
Highest nitroformazan yields, in excess of 90%, were achieved
for the two electron donating species, OCH3 and SCH3, both
being in the ortho positions. This is reflected in corresponding
highest dithizone yields of 76% each. The lowest yield of 13%
was found for 3,4-dimethyldithizone. Regardless implemented
1
Crystallographic data for the structure reported in this article had been
deposited with the Cambridge Crystallographic Data Centre with deposition num-
ber, CCDC 896693, and may be obtained free of charge from the Director, CCDC, 12
Union Road, Cambridge CB21EZ, UK [fax: +44 1223 336 033; deposit@ccdc.cam.ac.uk
or www.ccd.cam.ac.uk].

K.G. von Eschwege / Journal of Photochemistry and Photobiology A: Chemistry 252 (2013) 159–166
163
Fig. 3. The photochromic back-reaction of the blue-grey photo-induced isomer, 9,
of a 6 × 10−5 M solution of PhHg(4-SCH3)HDz in DCM, reverting to the red ground
state, 8, with isosbestic points at 563, 429, 372 and 281 nm. Spectra were recorded
at 0.1, 1, 3, 5, 7, 10, 15, 30 and 60 min after exposure to white light from a 400 W
mercury-halide lamp.
3.3. Kinetics
Since discovery of the photochromic behaviour of Hg(HDz)2
in 1950 [9] no comprehensive kinetic study of the radiationless
thermal back reaction of metal dithizonates had been reported.
The photochromic property, as measured by the return rate in
solution, is found to be very sensitive to temperature, the metal,
impurities, acids and bases, and the polarity of solvents. Addition
of a few drops of ethanol, acetic acid or triethylamine reduces
the return reaction half-life profoundly. Hydroxylic solvents and
organic acids and bases are generally the poorest media for observ-
ing photochromism. The strongest photochromic effect is observed
in dry non-polar solvents, such as the halogenated solvents and
toluene.
In an attempt to quantify these observations, the rates of
return were measured at different concentrations, temperatures
and in different solvents. In addition, return rates of a series of
twelve electronically altered complexes in dichloromethane were
also measured. In view of the many factors possibly affecting
the rate of return, where applicable, care was taken to maintain
similar conditions during all measurements, e.g. similar wash-
ing procedures, solvents from similar batches, and the same
cuvette and instrument were used right through. Complex sta-
bility in dichloromethane proved to be of no concern, as e.g. a
sealed solution of 2-methoxydithizonatophenylmercury(II), hav-
ing been stored in the dark for two years, showed undiminished
and repeated photochromic behaviour in sunlight.
Fig. 1. The molecular structure of the ortho-S-methylnitroformazan, i.e. (E)-1-[2-
(methylsulfanyl)phenyl]-2-({(E)-2-[2-(methylsulfanyl)phenyl]hydrazinylidene}
(nitro)methyl)diazene, with anisotropic displacement ellipsoids drawn at 50%
probability level. Intramolecular H-bonds are depicted by dotted lines.
backbone. The second, S(11) CH3, substituent is twisted out of the
plane by 58.73◦.
ADF computed molecular orbitals of the title compound reveals
orbitals extending over the entire ligand, including both phenyl
rings and the sulphur atom (only HOMO) (see Fig. 2). This computed
result corresponds to the experimental observation that the pho-
tochromic property is characteristic of dithizonates with phenyl
end groups, as opposed to aliphatic groups. The large degree of elec-
tron delocalization resulting from ␲-orbital overlapping is typical
of sp2 hybridization of nitrogen atoms along the linear and planar
backbone of the molecule. The
C N double bond character in the
HOMO is made manifest by the higher depicted electron density
between the carbon and nitrogen atoms, as opposed to decreased
electron density in the LUMO (see arrows). Photo-induced exci-
tation of the DPM HOMO results in the electron promoted to the
Scheme 2 represents the photochromic reaction of PhHg(4-
SCH3)HDz, where the absorption of green light results in a
reversible colour change from red, 8, to blue-grey, 9. By the over-
lay of consecutive spectra, Fig. 3 illustrates the corresponding
spontaneous radiationless back reaction in dichloromethane. The
absorbance maxima at 617, 407 and 291 nm disappear over a
LUMO, where the
C N double bond is reduced to a single bond,
which is a prerequisite for subsequent rotation around this axis.
Apart from the blue isomer and its orbitals having a bent geometry
(not shown), respective orbitals of the two isomers are essentially
alike.
Fig. 2. (a) ADF calculated HOMO and (b) LUMO of the orange isomer of DPM. Bonding and anti-bonding orbitals are differentiated by colour. Arrows indicate axis of rotation.

164
K.G. von Eschwege / Journal of Photochemistry and Photobiology A: Chemistry 252 (2013) 159–166
Table 1
Absorbance maxima of the dithizone free ligand (ꢀ1 and ꢀ2) and corresponding dithi-
zonatophenylmercury(II) complexes (ꢀ – resting state isomer, ꢀ* – photo-induced
isomer) in DCM, arranged in increasing order of complex ꢀ.
Substituent
Ligand
Complex
ꢀ1
ꢀ2
ꢀ
ꢀ*
2-CH3
2-F
H
4-F
3-F
3-CH3
3-OCH3
4-CH3
3,4-(CH3)2
2-SCH3
2-OCH3
4-SCH3
457
450
443
446
447
448
453
452
453
477
474
486
623
621
609
620
610
613
625
622
623
643
641
667
464
468
470
471
475
476
480
482
485
488
498
509
580
590
590
590
591
591
597
597
602
594
610
617
60 min time period, while the resting state peaks appear at 509, 388
and 280 nm. The small peak observed at ca 407 nm is unique to the
two S-methyl derivatives. In the other complexes this peak appears
as a blue-shifted shoulder (ca 360 nm) to the right of the high-
energy absorption band (ca 280 nm). The S-methyl substituent is
thus observed to have a significant influence on the two lower
energy absorption bands, but leaves the high-energy absorption
at ca 280 nm relatively unaffected.
In the current series the longest wavelength absorbance
maxima are observed for the former PhHg(4-SCH3)HDz deriva-
tive (ꢀ = 509 nm and ꢀ* = 617 nm) (see Table 1). The ortho-
methyl derivative yielded shortest ꢀmax values (ꢀ = 464 nm and
ꢀ* = 580 nm). These values correspond to overall wavelength shifts
of 45 and 37 nm respectively. Considering the earlier mentioned
boundaries within which the dithizone ligand may be synthesized
(Section 3.1) the above wavelengths give the maximum shifts in
absorbance maxima, to date, that may be achieved by electronically
altering the ligand.
In general, complexes associated with electron withdrawing flu-
orine substituents absorb at higher energies, while the electron
donating substituents have the opposite effect. This trend is roughly
reflected in corresponding free ligand data (Table 1). Molar absorp-
tivity coefficients at ꢀmax of the resting state complex isomers vary
from ca 14 000 to 20 000 dm3 mol−1 cm−1 in DCM, with the excep-
tion of the meta-fluoro derivative being only 4800 dm3 mol−1 cm−1
.
No particular trend was observed.
During all the kinetic studies reported here the series of dithi-
zonatophenylmercury(II) complexes were photo-excited under
illumination with white light from a 400 W metal-halide lamp, and
the spontaneous back reaction spectrophotometrically followed at
ꢀmax of the photo-induced isomers. Application of first order kinetic
models [27] to spectroscopic data established the regeneration of
the ground states to all be first order processes.
Fig. 4. Top: Disappearance of the photo-induced blue isomer of photochromic DPM
(3 × 10−5 M) in DCM, at 0 (᭹), 10 (ꢀ), 20 (ꢁ) and 30 ◦C (ꢂ), measured at 590 nm.
Middle: First order kinetics, illustrated by linear plots of ln[A0/At] versus time. Bottom:
A trend line through the plot of rate constant versus temperature fits an exponential
correlation with R2 > 0.98.
At the outset, a concentration study done in dichloromethane
solutions at 20 ◦C, with concentrations varying from 1.5 to
7.5 × 10−5 M, yielded no significant trend or difference in rate
constants. The result was expected, as the reaction is merely an
isomerization which is not dependent on reactant concentrations
per se. Although much increased proximity of analyte molecules
may in fact alter the environment and possibly the reaction rate,
higher concentrations were not investigated spectroscopically due
to expected deviations from Beer’s law. Furthermore, photochem-
ical conversion of the bulk of high concentration solutions to the
blue isomeric form becomes increasingly difficult as applied light
is blocked out by outer dye molecules.
for the blue photo-induced isomer of unsubstituted DPM. The
exponential increase in rate constants (Fig. 4, middle and bot-
tom) confirms early claims of a thermal radiationless back reaction
[11,16]. The relative small overall temperature increase on going
from 0 ◦C (k = 0.0005 s−1) to 30 ◦C (k = 0.0052 s−1) resulted in a
roughly ten times increase in reaction rate. Similarly, a decrease in
temperature significantly slows down the reaction, with the con-
sequence that DPM may be frozen almost indefinitely in the blue
photo-induced state at very low temperatures, an observation that
was experimentally observed. Under similar conditions extrapo-
lation of the exponential curve (y = 0.0004e0.0814x) predicts a blue
isomer half life [where A(t ) = 0.11] of 1 s at 91 ◦C, while for t to be
½
½
Secondly, a temperature study was performed by following the
back reaction at 0, 10, 20 and 30 ◦C in dichloromethane. Fig. 4
(top) illustrates the exponential decline curves (R2 > 0.99) obtained
equal to 1 day the temperature has to be lowered to −48 ◦C. Accord-
ingly, Meriwether et al. observed photochromism of Cd and Pb
dithizonate complexes only as low as −80 ◦C, while the longer lived

K.G. von Eschwege / Journal of Photochemistry and Photobiology A: Chemistry 252 (2013) 159–166
165
Table 2
Rates of the spontaneous reverse reaction of DPM in five solvents at 10 ◦C, arranged
in ascending order, related to solvent molar mass (MM), dielectric constant (ε) and
ꢀ* of the photo-induced blue isomer.
Solvent
Rate (s−1
)
MM (g mol−1
)
ε
ꢀ* (nm)
DCM
Toluene
THF
Ether
Acetone
0.0008
0.0083
0.0250
0.0309
0.0939
84.93
92.14
72.11
74.12
58.08
8.9
2.4
7.6
4.3
590
603
600
598
602
20.7
Table 3
Reaction rates of the spontaneous reverse reaction of DPM and eleven electronically
altered derivatives (3 × 10−5 M) in DCM at 20 ◦C, arranged in increasing order.
Substituent
Rate
2-CH3
4-F
2-OCH3
2-SCH3
2-F
4-CH3
3,4-(CH3)2
4-SCH3
3-CH3
H
0.0002
0.0003
0.0003
0.0004
0.0006
0.0012
0.0013
0.0017
0.0020
0.0024
0.0038
0.0106
3-F
3-OCH3
Trend line fits gave R2 > 0.98.
having the smallest ꢀmax values (see Table 1, 2-CH3 – ꢀ and ꢀ*),
no other correlations between reaction rate and absorbance max-
ima are seen. Also, no correlation is found with the well-known
Hammett constants [28] and relative group electronegativities [29].
Together with altered electronic influences obtained by the system-
atic phenyl substitution pattern in the current photochromic series,
steric effects, hydrogen bonding and non-bonding interactions, etc.,
are all expected to influence the back-isomerization of the photo-
induced state. Thus, no simple explanation may be given for the
observed trend. Nevertheless, a general internal trend is observed
which may be seen in the ortho-substituted photo-induced iso-
mers being most stable (with the exception of 4-F), as is reflected
by corresponding lowest reaction rates. These are followed by the
para-substituted species, while the meta derivatives (including the
unsubstituted DPM, H) gave the highest rates. Only the meta-fluoro
and meta-methoxy compounds have faster rates than the parent
DPM compound, while all other substituents slowed down the
reverse reaction.
Fig. 5. Top: First order kinetics, illustrated by linear plots of ln[A0/At] versus time,
representing disappearance of the blue photo-induced isomer of photochromic DPM
(3 × 10−5 M) in different solvents, measured at ꢀmax in each solvent, at 10 ◦C. Bottom:
Rate dependence on solvent molar mass (᭹) and dielectric constant (ꢁ) as indicated
by trend lines.
complexes of Hg, Pt, Pd, Ag and Bi were seen to be photochromic at
room temperature [11].
Rate measurements of the return reaction of DPM were made
in the five solvents; dichloromethane, toluene, tetrahydrofuran,
diethyl ether and acetone (see Fig. 5 (top)). In order to make
standard UV/visible measurements, i.e. without the use of special-
ized equipment, fast reactions in the more polar solvents had to be
slowed down by cooling all solutions to 10 ◦C.
A rough correlation was found between the observed rates
and corresponding solvent dielectric constants and molar masses
(Fig. 5, bottom). This observation corresponds to earlier results
from measurements done in benzene and tetrahydrofuran [12].
As return of the photochromic reaction merely involves free rota-
tion of the NN(H)Ph moiety of DPM in solution, resistance against
this motion is expected to increase as the molar mass of the sur-
rounding solvent increases, i.e. a reflection of inertia in solvent
molecules. Accordingly it was found that the highest return rate
occurs in acetone (k = 0.0939 s−1), which has the smallest molar
mass and highest dielectric constant. The lowest rates are conse-
quently observed in dichloromethane (k = 0.0008 s−1) and toluene
(k = 0.0083 s−1) (see Table 2).
Lastly, a kinetic study involving the series of twelve elec-
tronically altered dithizonatophenylmercury(II) complexes was
done. Observed rates for the spontaneous back reactions in
dichloromethane at 20 ◦C are listed in Table 3. The highest rate
in the series was found for the meta-methoxy compound, with k
being 0.0106 s−1, which is 53 times faster than the lowest rate of
0.0002 s−1 of the ortho-methyl compound. Apart from the ortho-
methyl complex also absorbing the highest energy photons, i.e.
4. Conclusion
Modification of earlier methods simplified the synthetic proce-
dure and brought about increased yields for the series of twelve
photochromic dithizonatophenylmercury(II) complexes, via the
nitroformazan intermediate and dithizone free ligand. Computed
molecular orbitals correspond to the observed isomerization reac-
tion by revealing a single C N bond in the LUMO, which allows
free rotation around this axis, as opposed to the HOMO C N double
bond. A comprehensive kinetic study of the spontaneous radiation-
less thermal back reaction of the photochromic mercury complexes
illustrates rate constants dependent on temperature, solvent and
phenyl substituent.
Acknowledgements
The author acknowledges the Central Research Fund of the Uni-
versity of the Free State for financial assistance, and Tania N. Hill
for the X-ray data collection.

166
K.G. von Eschwege / Journal of Photochemistry and Photobiology A: Chemistry 252 (2013) 159–166
References
(c) K.G. von Eschwege, F. Muller, A. Muller, Acta Crystallographica Section E E67
(2011) m1858.
[15] K.G. von Eschwege, J. Conradie, J.C. Swarts, Journal of Physical Chemistry A 112
(2008) 2211.
[16] H.M.N.H. Irving, Dithizone, Analytical Sciences Monographs, The Chemical
Society, London, 1977, p. 39.
[17] K.G. von Eschwege, J.C. Swarts, Polyhedron 29 (2010) 1727.
[18] K.G. von Eschwege, L. van As, J.C. Swarts, Electrochimica Acta 56 (2011)
10064.
[19] K.G. von Eschwege, J. Conradie, A. Kuhn, Journal of Physical Chemistry A 115
(2011) 14637.
[20] G.T. Velde, F.M. Bickelhaupt, E.J. Baerends, C.F. Guerra, S.J.A. van Gisbergen, J.G.
Snijders, T.J.J. Ziegler, Computers and Chemistry 22 (2001) 931.
[21] J.P. Perdew, J.A. Chevary, S.H. Vosko, K.A. Jackson, M.R. Perderson, D.J. Singh, C.
Fioihais, Physical Review B46 (1992) 6671.
[22] G.A. Zhurko, D.A. Zhurko, CHEM CRAFT, Version 1.6 (Build 342), 2009.
[23] K.G. von Eschwege, Polyhedron 39 (2012) 99.
[24] K.G. von Eschwege, E.C. Hosten, A. Muller, Acta Crystallographica Section E E68
(2012) o425.
[25] K.G. von Eschwege, F. Muller, T.N. Hill, Acta Crystallographica Section E E68
(2012) o609.
[26] K.G. von Eschwege, F. Muller, E.C. Hosten, Acta Crystallographica Section E E68
(2012) o199.
[1] E. Fischer, Annalen 190 (1878) 67.
[2] E. Fischer, E. Besthom, Annalen 212 (1882) 316.
[3] J.H. Billman, E.S. Cleland, Journal of the American Chemical Society 65 (1943)
1300.
[4] D.M. Hubbard, E.W. Scott, Journal of the American Chemical Society 65 (1943)
2390.
[5] E. Bamberger, R. Padova, E. Omerod, Liebigs Annalen 260 (1926) 307.
[6] P.S. Pelkis, R.G. Dubenko, L.S. Pupko, Journal of Organic Chemistry of the USSR
27 (1957) 2190.
[7] D.S. Tarbell, C.W. Todd, M.C. Paulson, E.G. Lindstrom, V.P. Wystrach, Journal of
the American Chemical Society 70 (1948) 1381.
[8] F. Mirkhalaf, D. Whittaker, D.J. Schiffrin, Journal of Electroanalytical Chemistry
452 (1998) 203.
[9] H. Irving, G. Andrew, E.J. Risdon, Journal of the Chemical Society (1949) 541.
[10] J.L.A. Webb, I.S. Bhatia, A.H. Corwin, A.G. Sharp, Journal of the American Chem-
ical Society 72 (1950) 91.
[11] L.S. Meriwether, E.C. Breitner, C.L. Sloan, Journal of the American Chemical
Society 87 (1965) 4441.
[12] L.S. Meriwether, E.C. Breitner, N.B. Colthup, Journal of the American Chemical
Society 87 (1965) 4448.
[13] H. Schwoerer, K.G. von Eschwege, G. Bosman, P. Krok, J. Conradie,
ChemPhysChem (2011) 2653.
[14] (a) A.T. Hutton, H.M.N.H. Irving, L.R. Nassimbeni, Acta Crystallographica B36
(1980) 2064;
[27] J.C. Kotz, K.F. Purcell, Chemistry and Chemical Reactivity, Saunders College
Publishing, Philadelphia, 1987, 482 pp.
[28] C. Hansch, A. Leo, R.W. Taft, Chemical Reviews 97 (1991) 165.
[29] P.R. Wells, Progress in Physical Organic Chemistry 6 (1968) 111.
(b) K.G. von Eschwege, F. Muller, A. Muller, Acta Crystallographica Section E
E67 (2011) m1804;