Synthesis and kinetics of sterically altered photochromic dithizonatomercury complexes

Article abstract of DOI:10.1021/jp5076324

Following a previous study where 12 electronically altered dithizones were synthesized, here we report on attempts to synthesize 26 dithizones. The purpose was to explore the boundaries within which dithizones may be synthesized, explore spectral tuning possibilities, and investigate steric effects on the photochromic reaction of its mercury complexes. Contrary to expectation, large substituents like phenoxy groups increased the rate of the photochromic back-reaction. In the series H-, 2-CH₃-, 4-CH₃-, 3,4-(CH₃)₂-, 2-OC₆H₅-, and 4-OC₆H₅-dithizonatophenylmercury(II), the lowest rate of 0.0004 s^-1 was measured for the 2-CH₃ complex, while the rate for the 2-OC₆H₅ derivative was 20 times higher. A solvent study revealed a direct relationship between dipole moment and the rate of the back-reaction, while the relationship between temperature and rate is exponential, with t_1/2 = 2 min 8 s for the 4-phenoxy complex. The crystal structures of two dithizone precursors, 2-phenoxy- and 4-phenoxynitroformazan, are reported. (Figure Presented).

Full text of DOI:10.1021/jp5076324

Article
pubs.acs.org/JPCA
Synthesis and Kinetics of Sterically Altered Photochromic
Dithizonatomercury Complexes
Ernestine Alabaraoye, Karel G. von Eschwege,* and Nagarajan Loganathan
Department of Chemistry, University of the Free State, P.O. Box 339, Bloemfontein 9300, South Africa
S
* Supporting Information
ABSTRACT: Following a previous study where 12 electronically
altered dithizones were synthesized, here we report on attempts to
synthesize 26 dithizones. The purpose was to explore the
boundaries within which dithizones may be synthesized, explore
spectral tuning possibilities, and investigate steric effects on the
photochromic reaction of its mercury complexes. Contrary to
expectation, large substituents like phenoxy groups increased the
rate of the photochromic back-reaction. In the series H-, 2-CH3-, 4-CH3-, 3,4-(CH3)2-, 2-OC6H5-, and 4-OC6H5-
dithizonatophenylmercury(II), the lowest rate of 0.0004 s−1 was measured for the 2-CH3 complex, while the rate for the 2-
OC6H5 derivative was 20 times higher. A solvent study revealed a direct relationship between dipole moment and the rate of the
back-reaction, while the relationship between temperature and rate is exponential, with t1/2 = 2 min 8 s for the 4-phenoxy
complex. The crystal structures of two dithizone precursors, 2-phenoxy- and 4-phenoxynitroformazan, are reported.
• broadened syntheses, mainly to sterically alter the ligand,
whereby also the aromatic ring system in the ligated
complex may be expanded;
• the effects of these changes, as well as concentration,
solvent, and temperature, on the kinetics of the
photoinduced color-switching back-reaction; and
• accompanying X-ray structural work involving two
nitroformazan reaction intermediates.
1. INTRODUCTION
Around 1950, Irving1 and Webb2 independently observed
photochromic behavior in the bis-dithizonatomercury(II)
complex, Hg(HDz)2. The reaction entails photoinduced
isomerization around the −CN− bond, switching from the
orange resting state to the blue photoinduced new ground
state3 (see Figure 2). A photon of blue-green light excites the
HOMO electron to the S1 energy state, which is followed by
rotation to the 90° orthogonal geometry energy minimum,
where relaxation through the conical intersection is followed by
bifurcationpartially back to the orange ground state and
partially onward to the blue photoinduced state.4 The thermal
back-reaction is a radiationless process, the rate of which is
determined by the temperature, metal, ligand substituent,
medium, and the presence of impurities. The half-life of the
blue state may vary from fractions of a second to minutes.
In 1965, Meriwether et al. conducted the first kinetic studies
involving this reaction,5,6 which were followed up only recently,
in 2013.7 In the latter study the effects of concentration,
solvent, temperature, and electronically altered substituents on
the return reaction of photochromic dithizonatophenyl-
mercury(II) (DPM) were kinetically investigated. A complete
study of the ultrafast forward reaction involving the same series
was subsequently reported in 2014.8 Interestingly, from these
studies it was found that the 2-CH3 derivative of DPM has the
fastest blue-state population time (τ = 1.7 ps), while at the
Data from all these studies build a platform of basic
understanding on the present class of photoactive compounds,
which have potential applications in new energy conversion
technologies. Together with solvatochromism and concentrat-
chromism observed for the dithizone free ligand,9 electro-
chromism10,11 and photochromism in its metal complexes
provide a mix of interesting properties for ongoing research.
2. EXPERIMENTAL SECTION
2.1. General. Reagents purchased from Sigma-Aldrich and
solvents from Merck were used without further purification.
1
The H NMR spectra were recorded on a 300 MHz Bruker
DPX 300 NMR spectrometer (298 K). UV−visible spectra of
dilute solutions in optical glass cuvettes were recorded on a
Shimadzu UV-2550 spectrometer fitted with a Shimadzu CPS
temperature controller, while photoexcitation of the photo-
chromic mercury complexes was achieved with a 400 W
mercury-halide lamp. A Bruker X8 APEXII 4K Kappa CCD
diffractometer was used for single crystal data collections.
2.2. Synthesis. Syntheses of the DPM series employed
commercially available substituted (R) anilines as starting
same time its spontaneous back-reaction is the slowest (t1/2
=
58 min). The 4-SCH3 derivative conversion to its blue form, in
turn, was the slowest (τ = 3.6 ps), while its back-reaction gave
t1/2 = 7 min.
Received: July 29, 2014
To extend the former kinetic studies, we now report on the
following:
Revised: October 22, 2014
Published: October 23, 2014
© 2014 American Chemical Society
10894
dx.doi.org/10.1021/jp5076324 | J. Phys. Chem. A 2014, 118, 10894−10901

The Journal of Physical Chemistry A
Article
reagents, proceeding via the nitroformazan intermediates (see
Table 1) to the dithizone ligands, and finally complexing with
phenylmercury(II) chloride.7
CDCl3): δ 7.65−8.41(14 H, C10H7, 2×C10H7), 13.60 (2 H, s,
2×NH).
Naphthyldithizonatophenylmercury(II). Yield: 79%, mp
1
230−231 °C, λmax/nm (acetone) 489. H NMR (300 MHz,
CDCl3): δ 6.80−8.70 (19 H, m, 2×C10H7 and 1×C6H5), 10.06
(1 H, s, 1×NH). The product is not photochromic.
Table 1. Yields, Melting Points, and UV−Visible Absorbance
Maxima in Acetone of Products Resulting from the
Nitroformazan Synthesis Step (R = Substituent, R′ = Phenyl
Replacements)
3. RESULTS AND DISCUSSION
3.1. Synthesis. The synthesis of a wide range of chemically
altered dithizones was attempted in this study. Although the
primary aim was to investigate the effect of steric alterations on
the photochromic reaction of its mercury complexes, a
secondary aim was to further explore the dithizone synthesis
boundaries that may limit future research in this field.
The method that was used brought about symmetrically
substituted dithizones, i.e., with substituents similarly placed on
both phenyl rings of the ligand. The 26 anilines listed in Table
1 were employed as starting reagents. This series also included
naphthyl, anthracenyl, and pyrenyl moieties, which did not
serve as substituents but in fact replaced the entire dithizone
phenyl groups.
R
yield (%)
mp (°C)
λmax (nm)
H
68
69
52
79
62
70
68
76
72
77
67
64
73
84
72
72
0
130
210
194
>230
86
445
422
423
470s
378s
418
461
360s
487
607
422s
2-OH
3-OH
4-OH
2-SH
3-SH
108
>230
117
116
126
95
4-SH
2-NO2
3-NO2
4-NO2
2-CF3
3-CF3
2-CH3
4-CH3
oil
144
150
162
127
457
455
447
350
3,4-(CH3)2
2,4,6-(CH3)3
2,4,6-(C(CH3)3)3
2,6-(CH2CH3)2
2-COCH3
68
39
81
79
63
49
oil
oil
The first step nitroformazan synthesis method proved to be
largely successful. Synthesis of the 2,4,6-tri-tert-butyl derivative
was the only outright failure, which may be ascribed to steric
hindrance imposed by the two large ortho-placed tert-butyl
substituents. Yields of all the other products ranged from about
50% to almost 90%, with the 4-methyl derivative giving the
highest yield of 88%, and the 4-NH-phenyl derivative yielding
49%. The exception was with the huge pyrenyl compound,
where the nitroformazan yield was only 6%.
Traditionally, the precipitated orange-red nitroformazans
gave absorbance maxima at ca. 445−470 nm, as also seen for
the above methyl-substituted derivatives, 4-OH and -SH, 3-
NO2, 2- and 4-OC6H5, 1,4-NHC6H5, and the anthracenyl
compound, i.e., widely indicative of the desired products. A
bathochromic shift of about 40 nm is seen for the naphthyl-
nitroformazan, to 508 nm. In contrast, a large hypsochromic
shift was observed for 2,4,6-trimethylnitroformazan, namely to
λmax = 378 nm. As in the case of 2,6-diethylnitroformazan, a
red-brown oil was obtained.
2-OC6H5
169
137
154
121
469
467
439
456
4-OC6H5
1,2-NHCH3
1,4-NHC6H5
R′ = naphthyl
R′ = anthracenyl
R′ = pyrenyl
76
51
6
159
>230
>230
508
450
498
Characterization data of dithizones and its mercury
complexes which were successfully synthesized and not
previously reported7 are given below.
2-Phenoxydithizone. Nitroformazan yield: 81%, mp 169−
170 °C, λmax/nm (acetone) 469. Dithizone yield: 98%, mp
1
173−174 °C, λmax/nm (acetone) 465 and 637. H NMR (300
MHz, CDCl3): δ 6.94−8.08 (18 H, m, 2×C6H4OC6H5), 13.07
From the list of seven main compounds, i.e., the derivatives
that were successfully synthesized right through to its mercury
complexes, the following increasing order in wavelengths at
maximum absorbances is seen:
(2 H, s, 2×NH).
2-Phenoxydithizonatophenylmercury(II). Yield: 82%, mp
1
131−132 °C, λmax/nm (acetone) 482. H NMR (300 MHz,
CDCl3): δ 6.76−7.86 (23 H, m, 2×C6H4OC6H5 and 1×C6H5),
9.60 (1 H, s, 1×NH). Product is photochromic.
Nitroformazans:
4-Phenoxydithizone. Nitroformazan yield: 79%, mp 137−
138 °C, λmax/nm (acetone) 467. Dithizone yield: 60%, mp
H (445 nm) < 3,4‐(CH3)2 (447 nm) < 4‐CH3 (455 nm)
< 2‐CH3 (457 nm) < 4‐OPh (467 nm)
< 2‐OPh (469 nm) < Naph (508 nm)
Dithizones:
1
114−115 °C, λmax/nm (acetone) 455 and 634. H NMR (300
MHz, CDCl3): δ 7.10−7.67 (18 H, m, 2×C6H4OC6H5), 12.53
(2 H, s, 2×NH).
4-Phenoxydithizonatophenylmercury(II). Yield: 86%, mp
1
173−174 °C, λmax/nm (acetone) 486. H NMR (300 MHz,
CDCl3): δ 6.98−7.94 (23 H, m, 2×C6H4OC6H5 and 1×C6H5),
H (613 nm) < 2‐CH3 (618 nm) < 4‐CH3 (621 nm)
< 3,4‐(CH3)2 (622 nm) < 4‐OPh (634 nm)
< 2‐OPh (637 nm) < Naph (670 nm)
9.20 (1 H, s, 1×NH). Product is photochromic.
Naphthyldithizone. Nitroformazan yield: 76%, mp 158−159
°C, λmax/nm (acetone) 508. Dithizone yield: 75%, mp 188−189
1
°C, λmax/nm (acetone) 492 and 670. H NMR (300 MHz,
10895
dx.doi.org/10.1021/jp5076324 | J. Phys. Chem. A 2014, 118, 10894−10901

The Journal of Physical Chemistry A
Article
A mostly similar order is observed for both the nitroformazan
and dithizone series, with some exception among the few
methylated compounds. The naphthyldithizone again, as for its
nitroformazan, shows a large red-shift to 670 nm with respect
to the 613 nm of unsubstituted dithizone.
In the second part of the synthesis procedure, addition of
ammonium sulfide exchanged the nitro group for sulfur while
reducing the compound to the thiocarbazide, H4Dz. By
addition of methanolic potassium hydroxide, H4Dz was then
deprotonated to the orange-red potassium dithizonate salt,
K+HDz−. The dithizone product, H2Dz, was finally precipitated
by addition of dilute hydrochloric acid.
During the latter steps, the yields of more than half the series
dropped, even right down to zero. As for the methylated
compounds, dithizone yields decreased from 2-methyl, 4-
methyl, and 3,4-dimethyl to the 2,4,6-trimethyl derivative by 74,
30, 13, and 0%, respectively. Both the increasing steric
hindrance posed by the increasing number of methyls on
both phenyl rings, and the accompanying subtle increase in
electron density on the molecular backbone, are proposed to
contribute toward the drastic yield decrease that was observed.
In contrast, a very high yield of 98% was obtained for the 2-
phenoxydithizone.
Figure 1. Selected UV−visible spectra of unsubstituted (green), 4-
phenoxydithizone (blue-green), and naphthyldithizone (blue) species
in acetone (9 × 10−5 M).
Of the extended ring series, syntheses of the naphthyl- and
anthracenyldithizone were successful; however, as yet this may
not be said with absolute surety about the latter. Traditionally,
all dithizones haved been intensely deep green, but the
anthracenyl product had an intense deep purple color. During
repeated column purifications, similar different color lines
appeared on the column, indicating an unstable anthracenyl
compound that decomposed during chromatography. Also,
extensive ongoing efforts to grow crystals suitable for X-ray
crystallography have, to date, been unsuccessful.
Figure 2. Spontaneous radiationless thermal back-reaction of (4-
OPh)DPM in toluene. The blue excited state (λmax = 615 nm, ε = 12
000 dm3 mol−1 cm−1) reverts back to the orange ground state (λmax
=
490 nm, ε = 21 833 dm3 mol−1 cm−1), with isosbestic points at 406
and 544 nm. The blue spectrum was recorded 5 s after exposure to
white light from a 400 W mercury-halide lamp.
A number of compounds formed a dark tar/oil, especially in
one of the very last two steps (i.e., the base or acid treatment
steps), or otherwise simply did not yield a product that had the
typical intense color of dithizone. Among these were the 2,4,6-
(CH3)3, 2,6-(C2H6)2, and 2-NO2 compounds, all the OH and
SH derivatives, and the 2-COCH3, 2- and 3-CF3 as well as the
2-NHCH3 and 4-NHC6H5 compounds. After repeated
attempts to isolate the desired dithizone ligands, we concluded
that especially the 2-COCH3, NO2, OH, and SH derivatives can
safely be ruled out as suitable compounds for further
investigation.
Most dithizones exhibit two major absorption maxima: the
primary band which peaks from 613 and 670 nm, and the
secondary band typically from 426 to 492 nm (see Figure 1).
When comparing different derivatives, the relative shift of the
higher energy absorption band (ca. 450 nm) trails the one at
lower energy (>600 nm); i.e., shifts among compounds stay
approximately the same for both bands. The colors of all
derivatives of dithizone are dark green in solution, except
naphthyldithizone, which is dark blue.
Apart from naphthyldithizonatophenylmercury, all success-
fully synthesized mercury dithizonate complexes showed
characteristic photochromism, i.e., changing from a shade of
orange to blue or blue-purple. This photochromic reaction is
best observed in direct African sunlight (at our laboratory), but
for the sake of spectroscopy measurements a 400 W mercury
“daylight” lamp was used. Figure 2 shows the spectral/color
change of the model compound of this investigation, 4-
phenoxydithizonatophenylmercury(II), (4-OPh)DPM, in tol-
uene.
Table 2 lists absorbance maxima for all the photochromic
complexes of this study. A red-shift of 117 nm is seen during
Table 2. UV−Visible Absorbance Maxima of the Orange
Ground States and Blue Photoexcited States of the DPM
Series in Acetone
λmax (nm)
R
orange isomer
blue isomer
H
466
468
478
480
482
486
489
583
584
586
590
606
610
2-CH3
4-CH3
3,4-(CH3)2
2-OPh
4-OPh
naphthyl
the photochromic reaction of the unsubstituted compound, i.e.,
going from 466 (orange ground/resting state) to 583 nm (blue
photoinduced ground state). The corresponding shift is slightly
larger for the two phenoxy complexes, namely 124 nm. As for
the naphthyl complex, no photochromic behavior could visibly
be seen. This, however, does not imply that photoinduced
isomerization does not occur; it may simply be too fast to be
seen, and in the future this will be established by means of
ultrafast laser spectroscopy.
With respect to the unsubstituted complex, absorbance
maxima of the phenoxy complexes in acetone are red-shifted by
about 20 nm to 486 (orange) and 610 nm (blue). This shift is
10896
dx.doi.org/10.1021/jp5076324 | J. Phys. Chem. A 2014, 118, 10894−10901

The Journal of Physical Chemistry A
Article
slightly less profound when compared to Δλ = 43 nm seen in
the previous electronically altered series, where the -SCH3
substituent caused the largest shift, albeit in dichloromethane.7
These figures indicate to what degree of spectral tuning
photochomic DPM complexes may be chemically modified,
which at this stage is not much. In both studies unsubstituted
DPM was observed to absorb at highest energy, i.e., shortest
wavelength, with respect to the entire series.
Absorbance maxima of orange (4-OPh)DPM in six
solventsacetone, dichloromethane, tetrahydrofuran, diethyl
ether, chloroform, and tolueneusing the same concentration
(6 × 10−5 M) and temperature (20 °C), are listed in Table 3.
Table 4. Back-Reaction Rates in Decreasing Order for DPM
Complexes under Varying Conditions
rate, k (s−1
)
a
Substituent
4-OPh
0.0102
0.0071
0.0034
0.0028
0.0020
0.0004
2-OPh
H
3,4-(CH3)2
4-CH3
2-CH3
b
Concentration (M)
6.0 × 10−5
0.0065
0.0046
0.0036
0.0027
Table 3. UV−Visible Absorbance Maxima of the Orange
Ground and Blue Photoinduced States of (4-OPh)DPM in
Different Solvents, Arranged in Increasing Order of
λmax(Orange)
5.0 × 10−5
4.0 × 10−5
3.0 × 10−5
c
λmax (nm)
Solvent
acetone
0.071
0.046
0.034
0.017
0.014
0.009
solvent
acetone
orange isomer
blue isomer
tetrahydrofuran
dichloromethane
diethyl ether
chloroform
toluene
434
486
488
488
489
490
560
604
606
608
602
615
DCM
THF
diethyl ether
chloroform
toluene
d
Temperature (°C)
55
45
35
25
15
5
0.0223
0.0151
0.0110
0.0054
0.0032
0.0019
The clear exception in λmax values are those recorded in
acetone. In both the orange and blue isomers, relatively large
blue-shifts are seen, i.e., from 486 nm in DCM to 434 nm in
acetone, with Δλ(orange) = 52 nm and Δλ(blue) = 44 nm. All
the other solvents gave absorbance maxima narrowly grouped
together. Clearly, the polarity of acetone (ε = 20.7) may be
related to this observation, being more than double that of
DCM (ε = 8.9); however, in the present series no further
relationship is seen between dielectric constants and
absorbance maxima in different solvents. Kinetic studies, on
the contrary, show a direct relationship between the rate of the
photochromic return reaction and solvent polarity, as will be
seen in the next section.
a
6.0 × 10−5 M toluene solutions of the substituted series of DPM
b
complexes at 20 °C. (4-OPh)DPM at different concentrations in
c
toluene at 20°C. 6.0 × 10−5 M (4-OPh)DPM in different solvents at
d
10°C. 6.0 × 10−5 M toluene solutions of (4-OPh)DPM at different
temperatures.
the rate of the dimethyl complex was seen to be larger than
those of the two monomethyl derivatives. The observation that
the compounds which were expected to give slower back-
reactions due to sterical reasons alone, was initially counter-
intuitive. In fact, the phenoxy compound return rates were
more than 20 times faster than those of the smaller 2-methyl
derivative. Additionally, the latter results are also consistent
with rate changes reported in 2013 for studies done in
DCM7an agreement between two separate studies which
further vindicates the reliability of experimental measurements.
Explanation for this observed trend thus necessarily has to
include more aspects than merely steric effects. Electronic
effects are not expected to play a significant role, as was recently
reported for both UV−visible7 and femtosecond laser studies in
this field.8 Second, based on structural data, H-bonding
interactions are also not expected to be of significance here.
We thus propose that since the larger molecules have a
significantly larger number of vibrational degrees of freedom,
the faster reaction rates may largely be attributed to this
property. This hypothesis is consistent with the fact that the
radiationless back-reaction is essentially thermal by nature,
facilitated by intramolecular vibrational modes.
3.2. Kinetics. To date, no study has been conducted to
investigate the effect of steric alterations on the back-reaction of
DPM complexes. Apart from the investigation of steric effects,
different concentrations, solvents, and temperatures were also
involved, and results are reported here.
Care was taken to maintain similar conditions throughout all
measurements, since the rate of the back-reaction is sensitive to
a number of factors, e.g., similar washing procedures, solvents
from similar batches, and use of the same cuvette and
instrument.
3.2.1. Substituent. Only six mercury complexes could be
employed in this particular investigation, i.e., those which were
successfully synthesized and which were visibly photochromic.
These included 6 × 10−5 M toluene solutions of sterically
altered DPM complexes with dithizonate phenyl ring
substituents 4-OC6H5-, 2-OC6H5-, 3,4-(CH3)2-, 4-CH3-, and
2-CH3- as well as the unsubstituted complex. Naphthyl-DPM
was the only other complex that was successfully synthesized,
but for which no visible photochromic reaction was observed.
The rates associated with the plots in Figure 3, as obtained
for the spontaneous back-reactions of the above series, are
listed in Table 4. The surprising result was that the sterically
much larger phenoxy derivatives gave the highest rates. Also,
3.2.2. Concentration. Due to the fact that (4-OPh)DPM
had not previously been synthesized and thus not kinetically
10897
dx.doi.org/10.1021/jp5076324 | J. Phys. Chem. A 2014, 118, 10894−10901

The Journal of Physical Chemistry A
Article
investigated, it was selected as model compound in all follow-
up investigations. As with the rest, kinetic measurements
involving concentration were observed to also follow first-order
rates. The linear correlation between absorbance and
concentration confirmed adherence to the Beer−Lambert law
under all present experimental conditions. Four concentrations
were compared, namely 3.0 × 10−5, 4.0 × 10−5, 5.0 × 10−5, and
6.0 × 10−5 M in toluene, while keeping the temperature
constant at 20 °C (see Figure 4). The systematic increase in the
concentration of (4-OPh)DPM increased the rate of the back-
reaction; i.e., doubling the concentration from 3 × 10−5 to 6 ×
10−5 M more than doubled the rate, from 0.027 to 0.064 s−1.
This is consistent with findings from the earliest kinetic studies
done by Meriwether et al. The explanation for this observation
is not yet apparent.
Figure 3. Linear plots of ln[A0/At] vs time, illustrating first-order
kinetics of the spontaneous back-reactions of 6.0 × 10−5 M toluene
▲
solutions of the sterically altered DPM series, at 20 °C: 4-OC6H5 ( ),
■
◆
●
2-OC6H5 (+), H ( ), 3,4-(CH3)2 ( ), 4-CH3 ( ), and 2-CH3 (×).
3.2.3. Solvent. Solvents are known to have an effect on
chemical compound solubility, stability and reaction rates.
Correct choice of solvent allows for thermodynamic and kinetic
control over chemical reactions. In order to study solvent effect
on the lifetime of the photoproduct of (4-OPh)DPM, six
solvents were employedacetone, tetrahydrofuran, dichloro-
methane, chloroform, diethyl ether and tolueneas represent-
ative of a wide range of polarities. This study was done at a
lower temperature, namely 10 °C, as the rate of the reaction
had to be decreased sufficiently for the faster back-reaction in
polar solvents to be observed during the present procedure.
Figure 4 again shows first-order kinetic plots vs time for the
photoinduced isomer after exposure to the bright light of a 400
W mercury-halide “daylight” lamp. Corresponding rates are
listed in Table 2. Graphical correlations between measured
rates and dipole moments, dielectric constants and the molar
masses of solvents are shown in Figures 7 and 8. Polar solvents
were observed to increase reaction rate, with highest rate in
acetone (0.071 s−1, ε = 20.7), while the lowest rate was
observed in toluene (0.009 s−1, ε = 2.4, see Supporting
Information for additional data). An almost perfect linear
relationship between reaction rate and both dipole moment (R2
= 0.93) and dielectric constant (R2 = 0.99) became apparent,
with an 8 times rate increase in acetone over toluene, see Figure
6. THF was excluded as an outlier in the present series, and as
such may be ascribed to the coordinating nature of this solvent,
or possible H-bonding at the THF oxygen position.
Figure 4. Linear plots of ln[A0/At] vs time, illustrating first-order
kinetics of the spontaneous back-reactions of 6.0 × 10−5 M toluene
solutions of (4-OPh)DPM, T = 20 °C, at different concentrations: 6 ×
−5
−5
10−5 (×), 5 × 10−5
■
◆
▲
( ), 4 × 10 ( ), and 3 × 10 M ( ).
Figure 5. Linear plots of ln[A0/At] vs time, illustrating first-order
kinetics of the spontaneous back-reactions of 6.0 × 10−5 M toluene
solutions of (4-OPh)DPM in different solvents at 10 °C, measured at
A large increase in solvent polarity for this particular
derivative is thus also expected to result in such a high return
rate that the photoinduced isomer (blue solution) is no longer
visible. This is indeed what was found during ultrafast studies of
DPM in polar methanol (ε = 32.7), which confirmed the
■
▲
λmax(blue), for different solvents: acetone (×), THF ( ), DCM ( ),
◆
●
diethyl ether (+), chloroform ( ), and toluene ( ).
Figure 6. Linear plots of ln[A0/At] vs time, illustrating first-order
kinetics of the spontaneous back-reactions of 6.0 × 10−5 M toluene
solutions of (4-OPh)DPM, measured at 615 nm, at different
2
■
◆
●
▲
temperatures: 55 ( ), 45 (+), 35 (×), 25 ( ), 15 ( ), and 5 °C ( ).
◆
Figure 7. Rate dependence on solvent dielectric constant ( , R =
2
◇
■
0.99, excluding THF ( )) and dipole moment ( , R = 0.93) as
indicated by the trend lines.
10898
dx.doi.org/10.1021/jp5076324 | J. Phys. Chem. A 2014, 118, 10894−10901

The Journal of Physical Chemistry A
Article
increase of 30 °C,7 the (4-OPh)DPM photochromic compound
in toluene is thus only about 60% as sensitive to changes in
temperature.
Figure 9 shows reaction rate increasing exponentially with
temperature, i.e., k = 0.0015 e0.0505T. With this expression now
established it becomes possible to determine temperatures that
are required for specific preset half-lives of the photoinduced
blue isomer. For example, at T = 25 °C and corresponding k =
0.0054 (Table 4), it follows that t1/2 = 128 s (t1/2 = ln2/k).
In order to slow the reaction down to the point where t1/2
=
1 day, the temperature has to be lowered to −104 °C, as
calculated by combining the former equations, i.e.,
Figure 8. Rate dependence on solvent molar mass as indicated by the
trend line.
ln 2/t1/2 = 0.0015e0.0505T
Similarly, increasing the temperature to 121 °C will decrease
t1/2 to 1 s.
The above results reflect findings by Meriwether et al., who
observed photochromism in, e.g., lead and cadmium
dithizonato complexes only as low as −80 °C and not at
room temperature.5,6 Also, in the previously reported study
involving electronically altered DPM complexes,7 related
temperatures for unsubstituted DPM were found to be −48
°C (t1/2 = 1 day) and +91 °C (t1/2 = 1 s). Comparing these two
sets of data, namely the fact that the temperature related to t1/2
= 1 s is shifted upward by about 30 °C in the 4-OPh complex as
compared to DPM, and downward by 56 °C (t1/2 = 1 day),
again illustrates the lesser sensitivity of (4-OPh)DPM to
changes in temperature.
An increase in temperature ultimately affects the rate at
which the product returns to the original ground state. The
total bonding energy of the blue isomer ground state is slightly
higher than of the orange ground state.8 Thermal relaxation to
the orange form is inhibited by the large energy barrier in
between the two ground states, i.e., at the 90° twist around the
−CN− axis of rotation.4 Whereas this barrier is in direction
of the forward reaction overcome by vertical excitation onto the
S1 excited-state energy surface, the return path depends on
vibrational energy, i.e., heat. From the above kinetic studies it
becomes evident to what extent the temperature associated
with an increased rate aids the photoinduced ground-state blue
molecule to overcome the barrier at the 90° twist angle during
back-isomerization.
3.3. X-ray Crystallography. During this investigation
attempts were made to grow crystals of all the derivatives of
nitroformazan, dithizone and dithizonatophenylmercury, as
well as of some of the apparent unsuccessful syntheses
products. Although crystalline material was obtained for a
number of species, crystal needles of the 2- and 4-phenoxy-
nitroformazans (NF) were the only crystals large enough for X-
ray data collections. Both crystals were grown from diethyl
ether overlaid with hexanes. The structures are shown in Figure
10. See Supporting Information Tables S2−S5 for crystal data,
refinement parameters, bond and torsion angles, and bond
lengths.
Both compounds yielded monoclinic crystals in the P21/c
space group (z = 4 and 8, respectively). Repeated attempts to
grow better crystals of (2-OPh)NF, also by taking crystals
directly from the crystallization solvent onto the diffractometer,
failed to produce better data for this compound. From the
Fourier difference map the (2-OPh)NF phenoxy group was
detected as being disordered over two oxygen atoms and was
treated as such.
Figure 9. Trend line through the plot of rate constants vs temperature
fits an exponential correlation.
photochromic reaction still occurring, but invisible to the naked
eye.4 The much faster return reaction in polar solvent medium
was proposed to proceed via the inversion mechanism, as
opposed to the otherwise rotation pathway.8
A second factor which intuitively plays a role in the rate of
the return reaction in different solvents is the respective molar
masses of the different solvents, as also recently reported
elsewhere.7 Figure 8 illustrates this relationship, which is less
precise than when compared to ε and DM (Figure 7). Here,
with the exception of chloroform and diethyl ether, remaining
solvents follow the linear trend closely. The consideration of
this correlation to be intuitive is attributed to the higher inertia
of heavier surrounding solvent molecules which inhibit
vibrational free rotation around the −CN− bond. In fact,
embedding this compound in a clear polymer slows the return
reaction to the extent that after illumination the “blue plastic”
takes several hours to return to its original orange ground
resting state. Thus, a more rigid environment decreases the rate
of the back-reaction so that free rotation is significantly
inhibited.
3.2.4. Temperature. In order to study the effect temperature
has on the return reaction of the photochromic compound, (4-
OPh)DPM was dissolved in toluene and rates monitored at six
different temperatures5, 15, 25, 35, 45, and 55 °Cthe
results of which are shown in Figure 6 and Table 4. Higher
temperatures could not be attempted, as the reaction became
too fast to accurately measure by means of the present method.
On the other hand, temperatures lower than 5 °C resulted in
water vapor condensation on the cold cuvette, adversely
affecting measurement accuracy.
As the temperature increases from 5 °C (k = 0.0019 s−1) to
35 °C (k = 0.0110 s−1), a significant increase in the rate
constant is seen, i.e., an almost 6 times increase over a
temperature range of 30 °C. Compared to unsubstituted DPM
in DCM, which showed a 10 times increase over a temperature
10899
dx.doi.org/10.1021/jp5076324 | J. Phys. Chem. A 2014, 118, 10894−10901

The Journal of Physical Chemistry A
Article
patterns, extended guidelines are now, for the first time,
established along which dithizones may in future be
synthesized.
Steric effects do not decrease return reaction rates in
photochromic mercury complexesin the traditional sense
but in fact increase the rate due to the increased number of
degrees of vibrational freedom in these large molecules, helping
overcome the transitional energy barrier. This is consistent with
the exponential rate increase with temperature. Linear relation-
ships were found between the rate of the return reaction and
solvent properties like dipole moment, dielectric constant, and
(roughly) molar mass.
X-ray data of two nitroformazan structures serve as indirect
evidence confirming the expected dithizone and mercury
complex structures. The mere observation of photochromism
in these complexes serves a similar purpose.
Figure 10. ORTEP views of the X-ray crystal structures of 2-
phenoxynitroformazan (left) and 4-phenoxynitroformazan (right).
Thermal ellipsoids are drawn at 50% probability level.
As opposed to the linear backbone geometry of dithizone,12
the nitroformazan precursors adopt a bent geometry in the solid
state. This bent geometry is held in place by the strong
intramolecular hydrogen bond between the (N1)H imine
protons and the adjacent N4 atoms in all nitroformazan
structures. H-bond distances are in the order of 1.72 Å.
Molecular backbones of these molecules are typically flat,
except for the phenyl moieties in the substituent, which are
twisted out of the plane due to steric obstruction. The nitro-
groups at the apex are also slightly twisted with respect to the
planar molecular backbone.
ASSOCIATED CONTENT
■
S
* Supporting Information
Table S1, rates of the spontaneous back-reaction of (4-
OPh)DPM in various solvents at 10 °C; Table S2, crystal data
and refinement parameters of nitroformazan structures; Tables
S3−S5, selected bond lengths (Å), bond angles (°), and torsion
angles (°); Figures S1 and S2, crystal packing diagrams of 2-
and 4-phenoxynitroformazan. This material is available free of
charge via the Internet at http://pubs.acs.org.
As in the case of dithizone, π-electrons along the nitro-
formazan backbone (N1−N2−C1−N3−N4) are also partially
delocalized, with double bonds being longer and single bonds
shorter than typically such bonds are otherwise. These
properties are in good agreement with published NF
structures.13−15 Interestingly, the para-substituted derivative,
(4-OPh)NF, gave similar bond lengths for both the N1−N2
and N3−N4 bonds, which is an indication of the imine proton
being shared by atoms N1 and N4. However, the adjacent very
slightly shorter C1−N2 bond (1.332 Å), compared to C1−N3
(1.348 Å) shows that this proton may mostly reside on the N1
position. This higher degree of symmetry in (4-OPh)NF, as
compared to (2-OPh)NF, is also reflected by its bond angles.
The (2-OPh)NF N1−N2−C1 bond angle is almost 4° larger
than its N4−N3−C1 angle. This again is ascribed to the imine
hydrogen situated on the N1 position. The corresponding angle
difference in (4-OPh)NF, where a larger degree of proton
sharing between N1 and N4 is deduced, is decreased to just
more than 1°.
AUTHOR INFORMATION
■
Corresponding Author
*Telephone: +27-(0)51-4012923. Fax: +27-(0)51-4446384. E-
mail: veschwkg@ufs.ac.za.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work is based upon research supported by the Central
Research Fund of the University of the Free State,
Bloemfontein, South Africa. LN thanks the South African
National Research Foundation (SA-NRF/THRIP), University
of the Free State Materials and Nanoscience Research Cluster,
SASOL and PETLABS pharmaceuticals for financial support.
REFERENCES
■
Interestingly, in both of the ortho-substituted compounds, (2-
OPh)NF and (2-OMe)NF,13 the phenoxy and methoxy
substituents on the two phenyl rings are orientated toward
each other, with this geometry thus of lower energy than the
alternative (when 180° rotated), where the substituents would
have been too close to the nitro group oxygens.
(1) Irving, H.; Andrew, G.; Risdon, E. J. Studies with Dithizone. Part
I. The Determination of Traces of Mercury. J. Chem. Soc. 1949, 541−
547.
(2) Webb, J. L. A.; Bhatia, I. S.; Corwin, A. H.; Sharp, A. G. Reactions
with Heavy Metals and their Bearing on Poisoning and Antidote
Action. J. Am. Chem. Soc. 1950, 72, 91−95.
(3) Von Eschwege, K. G.; Conradie, J.; Swarts, J. C. A DFT
Perspective on the Structures and Electronic Spectra of the Orange
and Blue Isomers of Photochromic Dithizonatophenylmercury(II). J.
Phys. Chem. A 2008, 112, 2211−2218.
See Supporting Information Figures S1 and S2 for
projections of molecular packing diagrams.
4. CONCLUSIONS
(4) Schwoerer, H.; Von Eschwege, K. G.; Bosman, G.; Krok, P.;
Conradie, J. Ultrafast Photochemistry of Dithizonatophenylmercury-
(II). ChemPhysChem 2011, 2653−2658.
(5) Meriwether, L. S.; Breitner, E. C.; Sloan, C. L. Kinetic and
Infrared Study of Photochromism of Metal Dithizonates. J. Am. Chem.
Soc. 1965, 87, 4441−4448.
(6) Meriwether, L. S.; Breitner, E. C.; Colthup, N. B. The
Photochromism of Metal Dithizonates. J. Am. Chem. Soc. 1965, 87,
4448−4454.
Limitations during dithizone synthesis are largely ascribed to
chemically active substituents which alter the outcome.
Substituents on both ortho-positions of each dithizone phenyl
pose significant steric hindrance during synthesis, while
extended aromatic ring systems appear to destabilize the
molecule. Single large groups may be utilized, but at the
expense of obtaining solid products in some cases. Together
with previous studies involving altered electronic substitution
10900
dx.doi.org/10.1021/jp5076324 | J. Phys. Chem. A 2014, 118, 10894−10901

The Journal of Physical Chemistry A
Article
(7) Von Eschwege, K. G. Synthesis and kinetics of Electronically
Altered Photochromic Dithizonatophenylmercury(II) Ccomplexes. J.
Photochem. Photobiol. A: Chem. 2013, 252, 159−166.
(8) Von Eschwege, K. G.; Bosman, G.; Conradie, J.; Schwoerer, H.
Femtosecond Laser Spectroscopy and DFT Studies of Photochromic
Dithizonatomercury Complexes. J. Phys. Chem. A 2014, 118, 844−855.
(9) Von Eschwege, K. G.; Conradie, J. C.; Kuhn, A. Dithizone and its
Oxidation Productsa DFT, Spectroscopic and X-Ray Structural
Study. J. Phys. Chem. A 2011, 115, 14637−14646.
(10) Von Eschwege, K. G.; Van As, L.; Swarts, J. C. Electrochemistry
and spectro-electrochemistry of dithizonatophenylmercury(II). Electro-
chim. Acta 2011, 56, 10064−10068.
(11) Von Eschwege, K. G.; Van As, L.; Joubert, C.; Swarts, J. C.;
Aquino, M. A. S.; Cameron, T. S. Intramolecular Interactions in a New
Tris-Dithizonatocobalt(III) Complex. Electrochim. Acta 2013, 112,
747−755.
(12) Laing, M. Dithizone: Redetermination and Refinement of its
Crystal Structure. J. Chem. Soc, Perkin Trans. 1977, 2, 1248−1252.
(13) Von Eschwege, K. G.; Muller, F.; Hill, T. N. (E)-1-[2-
(Methoxyl)phenyl]-2-({(E)-2-[2-(methoxyl)phenyl]hydrazinylidene}-
(nitro)methyl)diazene. Acta Crystallogr. 2012, E68, o609.
(14) Von Eschwege, K. G.; Hosten, E. C.; Muller, A. (E)-1-[2-
(Methyl)phenyl]-2-({(E)-2-[2-(methyl)phenyl]hydrazinylidene}-
(nitro)methyl)diazene. Acta Crystallogr. 2012, E68, o425.
(15) Von Eschwege, K. G.; Muller, F.; Hosten, E. C. (E)-1-[2-
(Methylsulfanyl)phenyl]-2-({(E)-2-[2-methylsulfanyl)phenyl]-
hydrazinylidene}(nitro)methyl)diazene. Acta Crystallogr. 2012, E68,
o199−o200.
10901
dx.doi.org/10.1021/jp5076324 | J. Phys. Chem. A 2014, 118, 10894−10901