Femtosecond laser spectroscopy and DFT studies of photochromic dithizonatomercury complexes

Article abstract of DOI:10.1021/jp410186y

The ultrafast dynamics of the photochromic reaction of dithizonatophenylmercury(II) was recently reported. For purpose of investigating the effect of electronically different substituents (X = o-F, m-F, p-F, p-Cl, o-CH₃, m-CH₃, p-CH₃, m,p-diCH₃, p-OCH₃, o-SCH₃, and p-SCH₃) on this reaction, a series of phenyl-substituted dithizones were synthesized and complexed with phenylmercury(II). A variation of more than 3 ps in ground state repopulation times was observed, with the o-methyl derivative absorbing both at shortest wavelength and having the fastest repopulation time, while the p-S-methyl derivative lies at the opposite extremity. An increase in both decay times and λ_max values is generally reflected by an increase in electron density in the chromophore. Ultrafast rates also proved to be dependent on solvent polarity, while a profound solvatochromic effect was observed in the transition state absorbance. Density functional theory realistically simulated isomer stabilities, electronic spectra and molecular orbitals. Increased electron density enhances stability in the photoexcited blue isomer relative to the orange resting state, as seen from a comparison between orange and blue isomer total bonding energies. A linear trend between computed HOMO energies and experimental λ_max of related aliphatic substituted derivatives was found.

Full text of DOI:10.1021/jp410186y

Article
pubs.acs.org/JPCA
Femtosecond Laser Spectroscopy and DFT Studies of Photochromic
Dithizonatomercury Complexes
Karel G. von Eschwege,†,* Gurthwin Bosman,‡ Jeanet Conradie,† and Heinrich Schwoerer‡
†Department of Chemistry, University of the Free State, PO Box 339, Bloemfontein, 9300, South Africa
‡Laser Research Institute, Stellenbosch University, Private Bag X1, Matieland, 7602, South Africa
S
* Supporting Information
ABSTRACT: The ultrafast dynamics of the photochromic
reaction of dithizonatophenylmercury(II) was recently re-
ported. For purpose of investigating the effect of electronically
different substituents (X = o-F, m-F, p-F, p-Cl, o-CH3, m-CH3,
p-CH3, m,p-diCH3, p-OCH3, o-SCH3, and p-SCH3) on this
reaction, a series of phenyl-substituted dithizones were
synthesized and complexed with phenylmercury(II). A
variation of more than 3 ps in ground state repopulation
times was observed, with the o-methyl derivative absorbing
both at shortest wavelength and having the fastest
repopulation time, while the p-S-methyl derivative lies at the
opposite extremity. An increase in both decay times and λmax
values is generally reflected by an increase in electron density
in the chromophore. Ultrafast rates also proved to be dependent on solvent polarity, while a profound solvatochromic effect was
observed in the transition state absorbance. Density functional theory realistically simulated isomer stabilities, electronic spectra
and molecular orbitals. Increased electron density enhances stability in the photoexcited blue isomer relative to the orange resting
state, as seen from a comparison between orange and blue isomer total bonding energies. A linear trend between computed
HOMO energies and experimental λmax of related aliphatic substituted derivatives was found.
1. INTRODUCTION
Photochromism in mercury dithizonates was discovered as far
back as 1945,1 with the first systematic studies undertaken 20
years later by Meriwether et al.2 Of the 24 metal dithizonates
prepared, only 9 were found to be visibly photochromic.
Photoexcitation by visible light promotes isomerization to a
different color complex, which spontaneously reverts to the
original resting state, see Figure 1. The reaction is typically
dependent on solvent, temperature, the metal center,
alterations on the ligand and the presence of impurities.3
Also, N-deuteration of bis(dithizonato)mercury(II), Hg-
(HDz)2, was found to result in a 3-fold decrease in the return
rate.4 The structures of the orange and photoinduced blue
dithizonatophenylmercury(II) (DPM) isomers were deter-
mined by X-ray crystallography5 and quantum computational
geometry optimization.6
Figure 1. Spontaneous radiationless thermal back-reaction of (o-
methoxy)dithizonatophenylmercury(II) in dichloromethane. The
Whereas the above work focused on the photochromic
reaction at large, as measured in the seconds time domain, we
purple-blue photoexcited state (λmax = 610 nm) reverts back to the
recently embarked on femtosecond laser spectroscopy research
to also establish aspects of the initial photochromic reaction of
DPM.7 Ultrafast excitation by 40 fs laser pulse occurred within
100 fs, resulting in a photoreaction with a time constant of 1.5
ps. Results were indicative of an orthogonally twisted excited
state that bifurcates below the funnel of the conical intersection
toward the orange cis and blue trans configurations. By the use
of ultrafast techniques, photochromism was also for the first
red ground state (λmax = 505 nm, ε = 28 346 dm3 mol−1 cm−1), with
isosbestic points at 405 and 560 nm.3.
Received: October 14, 2013
Revised: January 14, 2014
Published: January 15, 2014
© 2014 American Chemical Society
844
dx.doi.org/10.1021/jp410186y | J. Phys. Chem. A 2014, 118, 844−855

The Journal of Physical Chemistry A
Article
Figure 2. Schematic diagram of the femtosecond transient absorption setup.
time observed in strongly polar methanolic mediuma
preliminary indication that the photoreaction may indeed
occur in many other similar complexes but which have been up
to date not visibly observed.
amounts of water. The nitroformazan solid was dried overnight
in an oven at 70 °C (3.71 g, 95%).
(o-Methoxy)thiocarbazone. The nitroformazan (1.50 g, 4.7
mmol) was dissolved in a mixture of absolute ethanol (90 mL)
and ammonium sulfide solution (10 mL, 20% aq) in a 250 mL
stoppered round-bottom flask and stirred for 1 h. After
reduction was complete, the solution was added to a water/ice
mixture (500 mL). The dirty-white/yellow carbazide was
filtered off on a Buchner funnel containing silica gel (for better
filtration) and washed with water. The unstable thiocarbazide
was immediately oxidized to the dark red thiocarbazone by the
addition of cold methanolic potassium hydroxide (2%, 50 mL).
Stirring was continued until complete dissolution.
(o-Methoxy)dithizone. The dark green dithizone derivative
was precipitated by the addition of the thiocarbazone to dilute
hydrochloric acid (1%, 100 mL). The product was filtered and
again precipitated from an alcoholic alkali solution to which
diluted hydrochloric acid was added. The crude product sample
was dissolved in dichloromethane and passed through a short
silica column using toluene as eluent. A yellow impurity ran
ahead of the green product band, while some impurities stayed
behind at the origin. Alternatively, instead of running a column,
repeating the aforementioned precipitation procedure four
times, yielded pure o-methoxydithizone (1.09 g, 76%), as
confirmed by thin layer chromatography and NMR spectros-
copy. Mp 178−179 °C, UV/vis (dichloromethane): λmax 474
and 641, 1H NMR (300 MHz, CDCl3): δ 4.00 (6 H, s, OCH3),
6.97−7.17 (6 H, m, 2 × C6H4), 7.95−8.02 (2 H, d, 2 × C6H4),
12.92 (2H, s, 2 × NH).
(ortho-methoxy)dithizonatophenylmercury(II). Triethyl-
amine (0.05 g, 0.50 mmol) was added to a solution of o-
methoxydithizone (0.10 g, 0.32 mmol) and phenylmercury(II)
chloride (0.11 g, 0.35 mmol) in dichloromethane (30 mL), and
stirred for 15 min. The solution was overlaid with absolute
ethanol (15 mL) and evaporation of the DCM component
liberated 0.16 g (84%) of pure crystalline product. Mp: 212−
213 °C, λmax/nm (dichloromethane): 505. 1H NMR (300 MHz,
CDCl3): δ 3.68, 4.03 (6 H, 2 × s, 2 × OCH3), 6.57−7.89 (13
H, m, 2 × C6H4, 1 × C6H5), 9.75 (1H, s, 1 × NH).
2.3. Laser Spectroscopy. Femtosecond transient absorp-
tion measurements were performed at the Laser Research
Institute at Stellenbosch University. Figure 2 depicts the most
essential components in the time-resolved absorption arrange-
ment used for these measurements, of which the details are
reported elsewhere.7 However, in view of recent improvements
a brief description is presented here. The time-resolved
In many photochemical studies with potential practical
application, chemical tuning of spectral and rate aspects of the
photoswitching properties becomes of paramount importance.
The chromophore here under consideration lends itself ideally
toward chemical modification with regard to these goals. As
part of the total project, the second stage here was to synthesize
electronically altered dithizonatomercury complexes by sym-
metrically substituting both phenyl rings on the dithizone
ligand with moieties that vary not only in group electro-
negativity, but also in location on the phenyl groups. These
complexes were then studied by means of ultrafast laser
spectroscopy, in polar as well as nonpolar solvents, and
additionally clarified by appropriate DFT studies.
2. EXPERIMENTAL AND COMPUTATIONAL METHODS
2.1. General Data. Synthesis reagents (Sigma-Aldrich) and
solvents (Merck) were used without additional purification. 1H
NMR spectra were recorded on a 300 MHz Bruker Avance
DPX 300 NMR spectrometer, at 298 K. Chemical shifts are
reported relative to SiMe4 at 0 ppm. Ultraviolet and visible
spectra of dilute solutions in quartz cuvettes were recorded,
utilizing a Varian Cary 50 Probe UV/visible spectrophotometer.
Laser experiments were conducted as described in section 2.3,
and quantum computational methods are described in section
2.4.
2.2. Synthesis. The synthesis of o-methoxydithizone and its
corresponding phenylmercury complex, as the model com-
pound, is given. (Characterization data of all complexes were
reported elsewhere.3)
(o-Methoxy)nitroformazan. In a 100 mL beaker, o-
methoxyaniline (3.00 g, 24.4 mmol) was added to a mixture
of concentrated hydrochloric acid (13 mL) and water (25 mL)
and cooled to −10 °C on a cold plate while stirring. Sodium
nitrite (2.53 g, 36.5 mmol) dissolved in water (5 mL) was
slowly added. In a 500 mL beaker, the diazo solution was added
to a mixture of sodium acetate trihydrate (40 g), glacial acetic
acid (25 mL), and water (10 mL) and stirred at room
temperature for 5 min. Nitromethane (7.5 g, 123 mmol) was
added and stirred for 1 h, followed by the addition of water
(400 mL) and stirring for another 30 min. The red precipitate
was filtered through a Buchner funnel and washed with copious
845
dx.doi.org/10.1021/jp410186y | J. Phys. Chem. A 2014, 118, 844−855

The Journal of Physical Chemistry A
Article
spectroscopy system is based on the conventional pump−probe
technique, where the sample to be analyzed is activated by a
pump pulse and monitored at a later time by a probe pulse.
Generally, the pump pulse initiates a photoreaction, where
upon the expected absorption changes are detected in the
spectrum of the probe pulse. In order to generate a kinetic view
of the changes which occur in the sample after pump activation,
the temporal delay between the pump and the probe pulse is
altered sequentially. Therefore, by monitoring the spectrum of
the probe pulse as a function of delay time, a real-time view of
the photoreaction is obtained.
optical density (ΔOD) in each given sample was determined
through the following ratio,
⎛
⎞
T*(τ, λ)
T(λ)
ΔOD(τ, λ) = −log
⎜
⎟
⎠
⎝
Here T* represents the pumped and T the unpumped probe
spectrum at a delay time τ and wavelength λ. The sensitivity of
the change in optical density is estimated as the standard
deviation of the signal before pump−probe overlap which was
determined to be less than 10−3 (see Figure 3). The temporal
A commercially available femtosecond amplifier (Clark-
MXR: CPA 2101) is used as light source which delivers a
bandwidth-limited 150 fs duration 1 mJ pulse at a repetition
rate of 1 kHz centered at 775 nm. For the pump pulse, a
portion of the light (200−300 μJ) is used to drive a
noncollinearly phase-matched parametric amplifier (NOPA).
In the NOPA a chirped pulse centered at 480 nm with energy
in the micro-Joule range is generated. This chirped pulse is
compressed to about 40 fs with a combination of two Brewster
prisms and is then reflected off two mirrors mounted in retro-
reflector geometry on a translation stage. An optical chopper
operated at half the repetition rate of the CPA is introduced,
which ensures that every second pump pulse is effectively
blocked. A reflective type neutral density filter is used in order
to fine-tune the pulse energy in accordance with the
requirements of the experiment. The reflected pump pulse is
sent to one of two simultaneously triggered spectrometers
(Andor SR163 spectrograph), where it is used as an on−off
signal to determine when the sample is activated or not. The
transmitted pump pulse reflects off a mirror mounted on a 2-
axis piezoelectric adjustable mount which allows for the spatial
correction of the pump pulse at the sample position when
monitored by the CCD camera. A half-wave plate is used to set
the polarization of the pump pulse to the magic angle (54.7°)
with respect to the probe pulse. Finally, the pump pulse is
focused onto the sample with a spherical mirror (f = 200 mm).
A whitelight pulse generated in a 3 mm thick CaF2 crystal
was used for the probe pulse. The broad spectrum of this WL
allowed for absorption change measurements in the spectral
range of 340−650 nm. One drawback of using CaF2 is that it
has a very low damage threshold and is therefore continuously
moved as to increase its lifetime. Within this regard a circular
translation motion is chosen as to ensure that the orientation of
the crystal axes relative to the polarization of the fundamental
CPA pulse is constant. After whitelight generation and
collimation with an off-axis parabolic mirror (f = 50 mm) the
probe pulse is focused with a UV lens (f = 100 mm) onto the
liquid sample which is contained in a quartz flow cell (Starna
Scientific Ltd.: path length 200 μm). Also, the sample was
continually circulated with a reservoir using an inline rotary
pump (HNP Mikrosysteme GmbH: MZR 2921) at a speed in
excess of 500 mm/s. This guaranteed that a fresh specimen was
ready for every laser shot, hence avoiding product buildup.
Having passed through the sample the transmitted probe
pulse was imaged onto a fiber coupled to the second
spectrometer which as previously mentioned is on a common
clock with the first spectrometer used to detect the reflected
pump light. These spectrometers were fitted with fast line scan
Figure 3. Kinetic trace of dichloromethane at 600 nm. Inset: Gaussian
fit of the coherent artifact.
resolution of the setup was assumed to be equal to the fwhm of
the second derivative of a Gaussian function when fitted to the
coherent artifact generated with dichloromethane (DCM)
solvent as sample (Figure 3, inset). In doing so a temporal
resolution of 70 5 fs was obtained. An analogous method to
determine the temporal resolution of the setup is to generate
stimulated impulsive Raman scattering (SIRS).8 In this method
the ultrashort pump pulse activates Raman modes in the
solvent. The activation of these modes results in a modification
of the refractive index of the solvent which is expressed as
oscillations in the kinetic trace. These oscillations are clearly
visible in Figure 3 and have a period in the order of 100 fs.
Therefore, the temporal resolution is definitely less than this
period, because these oscillations are observable.
Because of the chirp of the probe white light, the temporal
position of the coherent artifact changes with respect to
wavelength. This temporal offset is corrected for by recording
the maximum position of the coherent artifact for a number of
wavelength values, fitting it with a higher order polynomial and
subtracting the amount of shift per wavelength position from
subsequent transient traces.
The dithizone ligand phenyl-substituted mercury complexes
were all dissolved in DCM to concentrations in the μM range.
These were subsequently introduced into the UTA setup upon
which transient spectra were recorded when activated with a 40
fs pump pulse centered at 480 nm. The conditions (temper-
ature, pump fluence, sample acquisitions, etc.) for the various
samples were all kept identical.
2.4. Quantum Computational Methods. Calculations
were done using the hybrid functional B3LYP9−14 (B3 Becke 3-
parameter exchange and Lee−Yang−Parr correlation) func-
tional for both exchange and correlation as implemented in the
Gaussian program package, version 09.15 Geometries were
optimized in gas phase (if no solvent is indicated) with the
triple-ζ basis set 6-311G(d,p)16−24 on all atoms except Hg,
cameras (1024 pixel photodiode array, Entwicklungsburo
̈
Stresing) which detected spectra with a 1 kHz repetition rate.
Therefore, by using the on−off trigger from the pump
spectrometer, the excited versus the nonexcited change in
846
dx.doi.org/10.1021/jp410186y | J. Phys. Chem. A 2014, 118, 844−855

The Journal of Physical Chemistry A
Article
Scheme 1. Synthesis of Altered Dithizones from Various Aniline Reagents (X = substituent)
where LANL2DZ was used (corresponding to the Los Alamos
effective core potential plus DZ).25−28 Solvation effects with full
geometry optimization were calculated with the solvent model
IEFPCM, using dichloromethane (ε = 8.9) or methanol (ε =
32.6) as solvent. TDDFT calculations, to generate the
calculated UV spectra of the complexes, were subsequently
carried out on the optimized geometries in dichloromethane as
solvent.
stepping from mono- to difluorinated phenyl rings, resulted in
the gradual loss of the typical intense color of dithizone, with
consequent loss in photochromic properties as well.31
Mercury complexation reactions were accomplished quanti-
tatively, but with expected losses due to the final recrystalliza-
tion step.
3.3. Laser Spectroscopy. In this femtosecond spectrosco-
py study, unsubstituted DPM in dichloromethane was used as
reference in order to illustrate the effect that changes in the
substitution pattern have on the initial photochromic reaction.
A previous study involving photochromic DPM reported the
instantaneous rise of excited-state absorption, followed by rapid
decay with a time constant of 300 fs.7 The orthogonally twisted
intermediate bifurcates through a conical intersection to the
two ground states, i.e. the orange and blue states, with a time
constant, τ, of 1.5 ps. Bifurcation at the conical intersection
implies that repopulation of the orange ground state and
formation of the blue photoproduct are inherently coupled.
Therefore, only the repopulation time of the orange isomer will
be considered as the measurement parameter in this report.
This choice was not made arbitrarily, but rather because the
absorption band of the blue photoproduct overlaps with that of
the intermediate species formed at orthogonal geometry
(Figure 4, spectra above 500 nm). Femtosecond photo-
excitation of orange DPM results in the immediate (<100 fs)
depletion of the 477 nm absorbance band (Figure 4, blue
trench) with the simultaneous appearance of two prominent
transition state absorption peaks at ca 390 and 550 nm (Figure
4, yellow and orange peaks). Relaxation of the short-lived
3. RESULTS AND DISCUSSION
3.1. General Information. By systematically replacing
dithizone phenyl ring hydrogens (ortho, meta, and para
positions) with a variety of electron donating and electron
withdrawing substituents, electron density on the delocalized
dithizone backbone may be altered. Hereby the substituent-
induced electronic effect on the ultrafast dynamics of the
photochromic reaction in corresponding phenylmercury
complexes was investigated by means of femtosecond laser
spectroscopy. During a concurrent solvent study, measure-
ments were made on the unsubstituted complex in nonpolar
dichloromethane (ε = 8.93), polar methanol (ε = 32.7) and
deuterated methanol. The latter solvent effects the imine
proton on the rotating arm (see Scheme 2) of the
photochromic moiety being replaced by a deuterium isotope.
Computational chemistry utilizing density function theory was
employed to quantify energies and UV−visible oscillators
associated with the orange (cis) and blue (trans) ground state
isomers.
3.2. Synthesis. Traditionally dithizone has been synthesized
by the diazotization of aniline, followed by the coupling of two
diazo molecules on addition of nitromethane to form
nitroformazan.29 Treatment of the nitroformazan with either
carbon disulfide or a combination of ammonia and hydrogen
sulfide gases eventually yields the intensely colored dithizone
product. The method by Mirkhalaf, however, proposed a more
convenient alternative, namely to use ammonium sulfide
instead of the noxious ammonia and hydrogen sulfide foul
smelling gases, see Scheme 1.30 The latter method was
successfully employed here, toward synthesis of a series of
electronically altered dithizonatophenylmercury(II) complexes.
Dithizone phenyl hydrogens were substituted by F, Cl, CH3,
OCH3, and SCH3, starting reaction procedures with corre-
spondingly functionalized commercially available anilines.
Dithizone yields varied from 13% for 3,4-dimethyldithizone
to 76% for the o-methyl substituted ligand. It was not at this
stage possible to synthesize the p-methoxy derivative, as the
large increase in donated electron density here resulted in the
decomposition of the product at the final dithizone formation
steps, i.e., at the oxidation of the thiocarbazide, H4Dz, to
potassiumdithizonate, K+HDz−, and protonation of K+HDz− to
dithizone, H2Dz.3 At the opposite boundary where attempts
were made to synthesize the most electron deficient dithizone,
Figure 4. Transient change in optical density ΔOD(λ,τ) of DPM in
dichloromethane between 300 and 650 nm after excitation with a 40 fs
short laser pulse at 480 nm.
847
dx.doi.org/10.1021/jp410186y | J. Phys. Chem. A 2014, 118, 844−855

The Journal of Physical Chemistry A
Article
orthogonally twisted intermediate toward the orange and blue
ground states is seen in the partial restoration of the original
absorbance band (absorbance increase at ca. 477 nm, spectra at
>0.5 ps) and decrease in the initial two transition state peaks.
3.3.1. Solvent Effect. Figure 5 shows excited state
absorbance changes effected by dissolving unsubstituted DPM
of the photochromic rotating moiety, added nuclear mass at
this position is indeed expected to slow down rotation, in other
words, the repopulation time. These ultrafast dynamics
involving deuterium agrees with earlier kinetic studies of the
overall photochromic reaction where blue to orange back
isomerization in nonpolar benzene was slowed down almost 3-
fold as a consequence of deuteration.2
The increased time constant in polar protic solvents at the
above ultrashort time scales (where 90° rotational segments are
considered) is in marked opposition to what is observed for the
much slower blue to orange back isomerization reaction (180°
rotation) where the latter reaction in exactly the same medium
is in turn much faster than in DCM, in fact so fast that it may
not be seen with the naked eye. On the contrary, in nonpolar
DCM the long-lived blue isomer has a typical time constant
many orders of magnitude (1014) larger than initial fast
dynamics, i.e., several minutes.3,33 Intermolecular solvent-
analyte van der Waals and hydrogen bonding in polar solvent
medium is expected to bind all species into a tighter framework,
thereby limiting free rotation of the orthogonally twisted
transition state during ultrafast dynamics, and thus slow down
repopulation times. On the other hand, the labile methanol
hydroxy proton with its well-known liquid phase proton
exchange reactions,34 is expected to be involved in the overall
blue to orange back reaction, being much faster than in DCM.
In this solvent a linear inversion pathway for the radiationless
back reaction is proposed, i.e., the CN−N−H moiety (see
Figure 1) losing the blue isomer imine proton to the polar
solvent medium while gaining another proton from an adjacent
methanol hydroxyl group at the opposite side, see Scheme 2
Figure 5. Transient change in optical density ΔOD(λ) of DPM
(circles) in dichloromethane, deuterated methanol (squares) and
methanol (triangles) at 500 fs after excitation with a short laser pulse
(40 fs) at 480 nm. (See Supporting Information, Figure S1 for
corresponding 3D plots in all three solvents.)
in dichloromethane, methanol, and deuterated methanol,
respectively. The solvatochromic effect is especially pro-
nounced in the molar absorptivities, ε, of the transition state
absorption bands at ca 390 and 550 nm, < 1 ps. In DCM the ε
ratio at these wavelengths is 3:10, changing to 13:10 in MeOD
and 18:10 in MeOH, a condition which partially persists in the
blue isomer absorbance, > 1 ps. This change to more intense
absorption at higher energy in MeOH is remarkably paralleled
by what was observed for the free dithizone ligand in similar
solvents.32 Dithizone itself has two visible absorption bands at
ca. 450 and 610 nm (both being 60 nm red-shifted with respect
to the present corresponding mercury complex), with an ε ratio
of 5:10 in DCM, changing to 80:10 in MeOH. By means of
TDDFT calculations, this solvent effect was satisfactorally
explained in terms of intramolecular transfer of one of the
dithizone imine protons to sulfur on dissolution in methanol.
Unfortunately, theory and software developments are not
advanced enough, as yet, for the successful optimization of
excited state complexes containing heavy metals, which
necessarily also have to include relativistic effects. While
being engaged in ongoing attempts toward finding a computa-
tional explanation, we may at this stage at most propose that a
similar phenomenon might be responsible for the observed
solvatochromism in the photochromic DPM intermediate.
Apart from the above spectral differences a change in time
constants in the different solvents were also observed, with
orange repopulation time constants, τ, in ps
Scheme 2. Fragment of DPM Indicating the Photo-Induced
Rotation and Thermal Relaxation Pathways in Non-Polar
a
DCM and Polar Methanol
a
The latter intermediate (bottom) is illustrated as both gaining and
losing a proton from/to the protic solvent medium.
(bottom). A polar solvent like methanol will indeed better
support polarization of bonds during such a transition than a
nonpolar solvent. In further support of this notion is the fact
that the visible photochomic reaction is not observed in the
presence of trace amounts of acids and bases. The color change
reaction in a polar solvent like acetone is nevertheless still
clearly visible. Acetone, on the contrary, does not carry labile
protons. Additionally, the visible photochromic reaction does
take place in very dilute tetrachloromethane or carbon disulfide
solutions, i.e., in solvents that are completely devoid of protons
and thus not involved in a proton transfer mechanism. This
observation is in support of a rotational path in DCM during
the reverse reaction, or alternatively, slow inversion that does
not include proton transfer. Vibrational energy is required to
1.03 0.08 (DCM) < 1.55 0.15 (MeOH)
< 1.61 0.10ps (MeOD)
illustrating a decrease in reaction rate, or increase of τ, of ca.
60% in polar solvents. A very small but detectable deuteration
effect (1.61 vs 1.55 ps) is also observed, supporting the notion
that only the more labile imine proton on the DPM backbone
is replaced by deuterium when dissolved in MeOD. Being part
848
dx.doi.org/10.1021/jp410186y | J. Phys. Chem. A 2014, 118, 844−855

The Journal of Physical Chemistry A
Article
Table 1. Absorbance Maxima in DCM and Selected Time Constants for the Repopulation of the Orange Isomer and Population
of the Blue Isomers of Electronically Altered Dithizonatophenylmercury(II) Complexes, Arranged in Ascending Order of
Orange Isomer λmax Values
orange ground state
excited state
blue ground state
compound
λmax, nm
τ, ps
λmax, nm
λmax at 1 ps, nm
τ, ps
λmax at 5 s, nm
o-CH3
o-F
470
474
476
477
479
483
488
489
493
500
507
519
0.51
1.95
1.42
1.04
1.35
1.36
1.91
1.28
1.53
1.61
1.02
3.79
374
380
384
384
384
387
390
393
394
387
397
426
555
552
555
560
555
559
560
567
570
577
584
613
1.7
2.3
2.5
2.1
1.9
4.4
2.6
2.5
2.5
2.4
1.9
3.6
580
590
591
590
590
591
596
597
602
594
610
617
m-F
H
p-F
m-CH3
p-Cl
p-CH3
m,p-diCH3
o-SCH3
o-OCH3
p-SCH3
overcome the energy barrier that restricts thermal relaxation to
the orange resting state. A recent temperature kinetics study
clearly illustrates the large effect of temperature on the rate of
return, e.g., the blue state half-life may increase to 24 h by
decreasing the temperature to ca −50 °C.3
Lastly, linear inversion during the thermal and radiationless
back reaction would not require degeneration of the CN
double bond axis, which is a prerequisite for rotation to take
place.35 A frequent argument that had been used in favor of
inversion is that activation energies for aromatic imine
isomerizations are approximately 20 kcal/mol36 as opposed to
the 40 kcal/mol that is required by aromatic olefins,37,38 the
latter which can isomerize only by the torsion mechanism. In
fact, some has suggested a mechanistic continuum whereby
isomerization in azomethines is neither taking place via rotation
alone, nor purely through inversion.39
species, and 37 nm (t = 5 s) in the blue isomer. The
observation that introduction of a methyl group at the ortho
position greatly accelerates the rate of isomerization is in
complete agreement with a similar finding by Herkstroeter 40
years ago when studying a series of substituted azomethine
isomerization reactions.39
Experimental data in Figure 6 (squares and triangles) shows
the initial depletion and repopulation of the DPM and (p-
In conclusion, it may thus only be stated that circumstantial
evidence is consistent with the above hypothesis. Exhaustive
computational work would however be required to provide
additional substantiation, which lies outside the scope of the
present study.
3.3.2. Substituent Effect. Selected data obtained by ultrafast
transient absorption spectroscopy measurements for the entire
series of photochromic complexes in DCM are listed in Table 1
(see Supporting Information Table S1 for an extended data
set). The time constants, τ, are indicative of how fast the orange
and blue ground state isomers form, following orange resting
state excitation to the excited state orthogonal geometry. With
the series arranged by increasing λmax values (wavelengths at
absorbance maxima of orange resting state in DCM), it
becomes clear that this trend is closely followed by λmax values
of the excited and blue ground states, with the exception of
only a few. This order is not paralleled by corresponding time
constants, where little overall tendency is observed. However,
as will be seen later, comparison within isolated groups of
closely related compounds do indeed reveal correlating trends.
As for the entire series, it is nevertheless noticed that the three
o-methyl DPM (o-CH3) states absorb light of shortest
wavelength while accompanied by smallest time constants. At
the opposite extremity the p-S-methyl (p-SCH3) compound in
turn absorbs the longest wavelengths while also having largest
time constants. By thus employing the methyl and S-methyl
substituents in the ortho and para phenyl positions respectively,
affects a maximum shift of 49 nm in λmax of the resting state
Figure 6. Transient change in optical density ΔOD(τ) for DPM
(squares) and (p-SCH3)DPM (triangles) at λmax of the GSB. The
numerical fit model is included (red line).
SCH3)DPM ground states. Ground state depletion is
accomplished by means of the 40 fs laser pump pulse centered
at 480 nm. Because of the manner in which the ground state is
depleted, it is expected that the ground state bleach (GSB)
signal only shows an exponential increase in the change of
optical density. Instead, an initial decrease (<1 ps) in the
change of optical density is observed, which is even larger than
the temporal resolution (70 5 fs). This is the result of slight
spectral overlapping of the ESA (positive signature) band and
the GSB (negative signature) band. After the signal drops, it
recovers with a time constant of about 1.03 ps for DPM and
about 3.79 ps for the p-SCH3 derivative. The fit function
plotted in Figure 6 is a two-exponential trend derived from the
kinetic model which is convoluted by the instrument response
function.7,40 The ubiquitous profile which corresponds to the
ESA (GSB and product dynamics is visible in Figure 6) serves
as verification that the various substituted samples follow the
photochemical reaction as described earlier.
849
dx.doi.org/10.1021/jp410186y | J. Phys. Chem. A 2014, 118, 844−855

The Journal of Physical Chemistry A
Article
From the time constants listed in Table 1, it is noted that o-
methylation more than halved τ(orange) of DPM (1.04 ps) to
0.51 ps, where as the same was increased (slowed down) almost
4-fold by the p-SCH3 substituent (3.79 ps). A similar trend is
seen in the blue isomer, however less pronounced. Neither the
well-known Hammett constants41,42 nor group electronegativ-
ities,43 traditionally applied to phenyl ring substituent
correlations, were found to compare well with the data in
Table 1. Indeed, multiple influences in the present system are
expected to affect deviations from the latter quantities. Among
these may certainly be solvent effects, intra- and intermolecular
hydrogen bonding, as well as steric effects. It is furthermore an
established fact that in general, close correlations can only be
obtained among related molecules, i.e. molecules with
substituents on different phenyl positions correlate poorly if
the substituent is not the same right through.44,45
the isomerization reaction obeys an inertial model; i.e., the
heavier the substituent the larger inertia becomes, with
consequent slowing down in the isomerization reaction. Figure
8 illustrates the correlation graphically. Even in this selected
Figure 8. Correlation of substituent molecular weights and
repopulation times of the para-substituted complexes. The added
steric effect of the larger SCH3 substituent contributes toward it being
an outlier in the present para-substituent series, as seen in its
significantly decreased orange repopulation rate.
Figure 7 displays the transient change in optical density
ΔOD(λ) 1 ps after photoactivation for all the para-substituted
group, τ(orange) is not tailing Hammett constants.41,42 The
Hammett equation relates reaction rates and equilibrium
constants with meta and para substituents via only the two
parameters; a substituent constant and a reaction constant.46,47
Electronic influence dominates in the Hammett equation. By
finding no Hammett correlation with τ it is clear that the
photochromic reaction is not foremost or only affected by
changes in the electronic properties of substituents, but also by
inertia and steric effects. The steric influence is illustrated by
the outlier, SCH3 (Figure 8), where τ is double that of Cl, while
the molecular weight is only 34% higher.
Where in the former comparison the phenyl ring position
was kept constant while substituents varied, the fluorinated
compounds on the other hand represents an example of where
the position is changed while the substituent is kept constant.
Here, λmax decreases slightly in going from p-F to m-F, and o-F
(see Table 1). At the same time τ (orange) increases in a
similar order, with (o-F)DPM having the highest value, which is
indicative of the slowest repopulation time. The close proximity
of the o-fluoro substituent to the ligand backbone is expected to
be involved in hydrogen bonding with the adjacent imine
proton, thereby restraining vibrational freedom with conse-
quent slowing down of the back reaction.
3.4. Quantum Computational Chemistry. 3.4.1. Struc-
tural Investigation. Quantum computational techniques are
here for the first time applied to substituted dithizonatomercury
complexes. On the basis of previous successful applications to
the free ligand and corresponding unsubstituted mercury
complexes, the B3LYP functional was again employed.6,32
Contrary to para and unsubstituted complexes, ortho and meta
derivatives may assume four different conformational isomers in
both the orange and blue states, see Figure 9. As for the meta-
substituted derivatives, energy differences are minimal, as meta-
placed groups do not impose steric limitations on conforma-
tion.
Figure 7. Transient change in optical density ΔOD(λ) for DPM, (p-
F)DPM, (p-Cl)DPM, (p-CH3)DPM, and (p-SCH3)DPM when
dissolved in DCM 1 ps after activation.
complexes; p-F, p-Cl, p-CH3, and p-SCH3, in comparison to the
parent DPM compound.
Apart from p-F, relative to DPM all other compounds in this
selection display bathochromic shifts at the absorbance maxima,
λmax, for GSB. This shift can in first approximation be
understood as the lowering of the HOMO−LUMO gap due
to charge donation and delocalization along the dithizonate
backbone. This has been verified by steady state density
functional theory (DFT) calculations, as presented in par. 3.4.2.
An identical trend for the GSB dynamics was observed for p-F,
p-Cl and p-CH3 (Supporting Information). To better elucidate
the trend among extracted parameters, the orange isomer
repopulation time was compared with molecular weights of
substituents, see Table 2. Repopulation times are seen to
increase with substitution by heavier groups. This indicates that
Table 2. Repopulation Times and Molecular Weights of
para-Substituted Hg Complexes
Geometry optimizations for (o-OCH3)dithizonato-
phenylmercury(II) are presented in Figure 9, thus chosen for
comparison purposes as it is the only ortho-substitued complex
for which X-ray crystallographic data is available.48 Optimiza-
tions yielded total bonding energies having an almost regular
stepwise increase of ca. 0.16 eV among the four orange
conformers. The computed most stable geometry was found to
be in agreement with the corresponding X-ray crystal structure
compound
τ(orange), ps
Mr(substituent), g/mol
H
1.03 0.08
1.28 0.10
1.35 0.09
1.91 0.11
3.79 0.07
1
15
19
35
47
p-CH3
p-F
p-Cl
p-SCH3
850
dx.doi.org/10.1021/jp410186y | J. Phys. Chem. A 2014, 118, 844−855

The Journal of Physical Chemistry A
Article
Figure 9. Relative total bonding energies of each of the four isomers of the orange (bottom) and blue (top) states of (o-
methoxy)dithizonatophenylmercury(II), corresponding to Figure 1. Inset: Orange isomer crystal structure.48 (o-OCH3)DPM was chosen for
comparison with available X-ray crystal data of the same.
Figure 10. Molecular orbital renderings of the compounds that lie at the extremities of λmax in the present series (Table 1), i.e., the o-methyl (top)
and p-S−methyl (bottom) derivative complexes. Elongation of the molecular orbital is seen by inclusion of at least one substituent sulfur atom,
resulting in an additional node in the waveform.
(Figure 9, inset). In this structure the methoxy substituent
forces the otherwise coplanar Hg−phenyl fragment out of the
plane of the dithizone ligand by ca 36°, while the C(phenyl)−
Hg−S bond is remarkably linear, having a bond angle of 179.0°.
Both experimental and computed NN and CN (expt,
The longer molecular orbital in itself is not responsible for
the lower energy absorption bands, as the p-F and p-Cl species
reveals similar extensions (see Supporting Information, Figure
S2), considering that (p-SCH3)DPM absorbs at 519 nm, while
(p-F)DPM absorbs at 479 nm, the latter which lies very close to
477 nm of unsubstituted DPM. Only the methylated species do
not accommodate similar orbital extensions. The varying
substituent electronic effect otherwise does not appear to affect
orbital geometry in general.
To a first approximation, photoexcitation of the orange form
can be considered as the excitation of a HOMO electron of the
orange resting state into its LUMO, leading to the proposed
rotation around the C−N bond and subsequent formation of
the blue form. The orange LUMO can be considered as the
HOMO of the photoexcited molecule of which the C−N bond
exhibits single bond character, as opposed to the stronger
double bond seen as a π-bond in the HOMO of the orange
́
́
1.290(6) Å; calcd, 1.321 Å) double bonds are longer than
typical bonds of this type, while backbone single bonds are all
shorter, which indicates a large degree of electron delocalization
along the dithizonato π-conjugated backbone, also including
the end phenyl rings. This observation is further demonstrated
by molecular orbital renderings illustrated in Figure 10, where
both the orange and blue HOMO and LUMO’s are seen to
span the entire extent of the ligand (blue LUMO not shown).
In fact, inclusion of an electron-rich substituent like the S−CH3
group extends the molecular orbital even further, to also
include at least the sulfur atom on the S−CH3 moiety closest to
the metal.
851
dx.doi.org/10.1021/jp410186y | J. Phys. Chem. A 2014, 118, 844−855

The Journal of Physical Chemistry A
Article
Table 3. Computed HOMO and HOMO−LUMO Gap Energies and Experimental and TDDFT Calculated Absorbance Maxima
a
of the Orange and Blue Ground State Isomers of the DPM Series in DCM, Arranged in Decreasing Order of EHOMO
orange
blue
EHOMO
,
EHOMO−LUMO
,
λmax(exp),
λmax(calc),
EHOMO
eV
,
λmax(exp),
λmax(calc),
b
compound
eV
eV
nm
nm
compound
nm
nm
compound ΔE, eV
o-OCH3
m,p-diCH3
p-CH3
m-CH3
p-SCH3
o-CH3
H
−5.216
−5.254
−5.317
−5.401
−5.439
−5.447
−5.476
−5.480
−5.562
−5.573
−5.687
−5.719
2.644
2.741
2.743
2.782
2.692
2.854
2.782
2.727
2.761
2.707
2.723
2.791
507
493
489
483
519
470
477
500
479
474
488
476
497
487
482
478
499
477
483
489
477
485
490
o-OCH3
m,p-diCH3
p-CH3
p-Cl
−4.970
−5.083
−5.142
−5.217
−5.258
−5.283
−5.285
−5.310
−5.370
−5.379
−5.497
−5.513
610
602
597
596
590
617
590
594
591
591
590
580
636
628
620
640
620
634
629
630
634
620
602
618
p-Cl
0.308
0.308
0.306
0.304
0.296
0.295
0.291
0.290
0.290
0.287
0.267
0.212
o-F
m-F
p-F
o-F
o-SCH3
H
p-SCH3
p-F
p-CH3
o-CH3
p-SCH3
m,p-diCH3
o-CH3
o−OCH3
o-SCH3
p-F
o-SCH3
m-F
o-F
m-CH3
H
p-Cl
m-F
477
b
o-CH3
a
ΔE values are similarly arranged in the rightmost column. Total bonding energy of the blue relative to the orange isomer.
isomer (Figure 10, up arrows). Free rotation around the C−N
bond therefore becomes possible after an electron is excited to
the orange-LUMO. The rotation involves the entire substituted
end phenyl ring bordering the nitrogen that carries the imine
proton, i.e. not only does the N−H rotate by 180°, but also the
substituted phenyl ring with it (see Figure 9, lowest E
geometries). Geometry optimization of the four possible blue
conformers pointed to this conclusion, i.e. the conformer that
maintains the substituent in the same orientation as the imine
proton during the photochromic reaction has the lowest
energy, and is therefore most stable. Additionally, it is
concluded that there is a strong π-bond interaction between
(H)N and the phenyl ring, not allowing free rotation around
the (H)N−C(Ph) bond during the photochromic reaction,
which is once again realistically simulated as seen in the LUMO
orbital (Figure 10, down arrow). These results reflect what was
found for all the ortho-derivatives. Since an optimized excited
state geometry is not available, only the character of the orange
LUMO may be evaluated, namely the MO to which
photoexcitation presumably takes place.
Figure 11. GO9/B3LYP-calculated TDDFT electronic spectra (bars)
and experimental UV−visible spectra of the orange (gray lines) and
blue (black lines) isomers of (o-OCH3)DPM in DCM. The shoulder
at ca. 500 nm in the blue UV−visible spectrum represents partial
conversion back to the orange resting state.
3.4.2. Electronic Spectra. Contrary to the pronounced
solvatochromic effect seen in both experimental and computed
UV−visible spectra and oscillators for the dithizone free
ligand,32 as well as in the transition state experimental spectra
of the complexes here reported (see Supporting Information,
Figure S1), the DPM ground state complexes are less affected
by medium. In agreement with this experimental observation,
the TDDFT computed unsubstituted DPM oscillator of lowest
energy that lies at 457 nm when calculated in gas phase (ε = 1),
is red-shifted by only about 20 nm to 475 and 483 nm when
calculated in methanol (ε = 32.6) and dichloromethane (ε =
8.9) respectively.
Experimental and TDDFT calculated maxima of the UV−
visible spectra of the complexes considered in this study are
compared in Table 3. The TDDFT calculated solvent value of
483 nm for unsubstituted DPM in DCM represents a close
simulation of the corresponding experimental value of 477 nm,
as do higher energy values (shorter wavelengths) and oscillators
for all the other substituted DPM’s. TDDFT computed
oscillators for all the blue isomers are found to be higher, i.e.
ca. 610 nm in gas phase and 620 nm in DCM solvent medium,
while the corresponding experimental value is 590 nm. The (o-
OMe)DPM example in Figure 11 is representative of calculated
TDDFT oscillators overlaid with the experimental spectra of
the orange and blue ground states (see related presentations for
the complete series in Supporting Information Figure S4).
Experimental values for the absorbance maxima (λmax
)
obtained at 1 ps (laser spectroscopy) and 5 s (UV−visible
spectroscopy) after photoexcitation differ slightly, see Table 1.
A red-shift for all complexes are seen when comparing
experimental λmax obtained at 1 ps with that measured at 5 s,
varying from 4 nm for the p-SCH3 to 38 nm for the o-F
derivative. This drift may not as yet be explained, also due to
the fact that our instruments are not equipped to investigate the
time period between 500 ps (maximum time window during
laser spectroscopy) and several seconds (manual procedure),
i.e. the time during which this transition completes.
HOMO energies may be viewed as representative of the ease
with which an electron in the HOMO may be accessed during
photoexcitation. A HOMO of higher energy is indicative of a
lower stability, and may thus be accessed more readily, with
consequent easier photoexcitation. Computed HOMO energies
were correlated here, with corresponding experimental data
obtained from ultrafast spectroscopy measurements, as listed in
Table 1, and graphically represented in Figure 12. The
previously mentioned finding that only closely related
852
dx.doi.org/10.1021/jp410186y | J. Phys. Chem. A 2014, 118, 844−855

The Journal of Physical Chemistry A
Article
pattern was observed, while ortho species again did not follow
observed trends.
Table 3 (rightmost column) also lists the energy differences,
ΔE, between the orange and blue isomers of the entire series of
complexes, which serves as an indication of the relative
stabilization effect due to substituent addition. Clearly, the
stronger electron donating species represent the best stabilizers,
resulting in a reduction in ΔE of almost 0.1 eV. According to
these figures the methoxy substituent presents the smallest
difference of 0.212 eV, as opposed to the halogenated series
that lies at the opposite extremity, where ΔE > 0.3 eV.
Figure 12. Computed HOMO energies correlated with energies of the
4. CONCLUSIONS
⧫
●
orange ( ) and blue isomers ( ) of substituted DPM in dichloro-
methane measured by laser spectroscopy (at central wavelengths of
absorption). Halogenated derivatives are encircled, S-methyls are
connected by a dashed line and derivatives containing methyls by a
solid line.
We hereby conclude that substitution on dithizone phenyl rings
significantly alter both wavelength of light absorption and
ultrafast time dynamics during the photochromic reaction of
corresponding phenylmercury complexes. These changes are
not mainly the consequence of electronic alterations, but also
due to steric effects and inertia. Dithizonatophenylmercury(II)
was for the first time observed to be notably solvatochromic in
the excited state. Solvent studies revealed a reversed trend
between ultrafast dynamics and the overall rate during the
photochromic reaction. DFT computational data established
favored geometries in both isomers and also indicates electron
donating substituents to enhance stabilization of the photo-
excited blue state.
molecules traditionally shows good correlation is well illustrated
also when comparing computed HOMO energies with
experimental data (energies related to λmax values calculated
by E = hc/λ). The halogenated compounds represent the
lowest HOMO energies, with the meta-F derivative having the
overall lowest orange isomer calculated HOMO energy of
−5.719 and −5.370 eV in the corresponding blue isomer. An
almost linear correlation is seen for the all methylated species,
which includes the o-OCH3 derivativethe latter consistently
having the highest values. The data pattern of the blue isomer
closely resembles that of the orange isomer.
ASSOCIATED CONTENT
■
S
* Supporting Information
Figure 13 illustrates the general trend found for the
correlation between the calculated HOMO−LUMO gap
Table S1, arbance maxima, and growth and decay times of the
orange, excited and intermediate states, and blue isomers of
electronically altered dithizonatophenylmercury complexes;
Figure S1, transient change in optical density of DPM in
dichloromethane, deuterated methanol, and methanol between
300 and 650 nm, after excitation with a 40 fs short laser pulse at
480 nm; Figure S2, HOMO renderings of selected DPM
derivitaves; Figure S3, TDDFT computed oscillators of the
orange and blue DPM isomers, in gas phase, DCM, and
methanol; and Figure S4, GO9/B3LYP calculated electronic
spectra and experimental spectra of the orange and blue
isomers of the substituted DPM complexes in dichloromethane.
This material is available free of charge via the Internet at
http://pubs.acs.org.
Figure 13. Computed HOMO−LUMO gap energies, ΔEgap
,
correlated with energies of the orange isomers of substituted DPM
in DCM measured by laser spectroscopy (at central wavelengths of
absorption). ortho-Substituted derivatives (o-CH3, o-F, o-OCH3) are
AUTHOR INFORMATION
■
Corresponding Author
*(K.G.v.E.) Telephone: +27-(0)51-4012923. Fax: +27-(0)51-
4446384. E-mail: veschwkg@ufs.ac.za.
●
indicated by circles ( ). The trend among meta and para compounds
■
is indicated by the line through squares ( ).
Notes
energies with experimental λmax values. A better and consistent
overall trend is indeed expected here, as the experimental data
also represents energy differences and not relative energies (of
HOMO’s) as discussed in the former paragraph. The ortho-
substituted species, however, deviates from the otherwise
observed trend. The close proximity of an ortho substituent to
the backbone has a significant influence on the molecule at
large, as also argued with regard to the Hammett series of
constants, i.e. Hammett constants for ortho variants do not
exist.41,42 DFT calculated HOMO and LUMO energies of
dithizone derivatives were previously related to corresponding
oxidation (Epa,1) and reduction potentials (Epc,C).31 A similar
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work is based upon research supported by the South
African Research Chair Initiative of the Department of Science
and Technology and the National Research Foundation. This
work additionally received support from the Norwegian
Supercomputing Program (NOTUR) through a grant of
computer time (Grant No. NN4654K) (J.C.) and the Central
Research Fund of the University of the Free State,
Bloemfontein.
853
dx.doi.org/10.1021/jp410186y | J. Phys. Chem. A 2014, 118, 844−855

The Journal of Physical Chemistry A
Article
(22) Binning, R. C., Jr.; Curtiss, L. A. Compact Contracted Basis Sets
for Third-Row Atoms: Ga−Kr. J. Comput. Chem. 1990, 11, 1206−
1216.
(23) McGrath, M. P.; Radom, L. Extension of Gaussian-1 (G1)
Theory To Bromine-Containing Molecules. J. Chem. Phys. 1991, 94,
511−516.
(24) Curtiss, L. A.; McGrath, M. P.; Blaudeau, J. P.; Davis, N. E.;
Binning, R. C., Jr.; Radom, L. Extension of Gaussian-2 Theory to
Molecules Containing Third-Row Atoms Ga−Kr. J. Chem. Phys. 1995,
103, 6104−6113.
(25) Dunning, T. H.; Hay, P. J. In Modern Theoretical Chemistry;
Schaefer, H.F., III, Ed.; Plenum: New York, 1976; Vol. 3, pp 1−28.
(26) Hay, P. J.; Wadt, W. R. Ab Initio Effective Core Potentials for
Molecular Calculations. Potentials for the Transition Metal Atoms Sc
to Hg. J. Chem. Phys. 1985, 82, 270−283.
(27) Hay, P. J.; Wadt, W. R. Ab Initio Effective Core Potentials for
Molecular Calculations. Potentials for Main Group Elements Na to Bi.
J. Chem. Phys. 1985, 82, 284−298.
(28) Hay, P. J.; Wadt, W. R. Ab Initio Effective Core Potentials for
Molecular Calculations. Potentials for K to Au Including the
Outermost Core Orbitals. J. Chem. Phys. 1985, 82, 299−310.
(29) Pelkis, P. S.; Dubenko, R. G.; Pupko, L. S. Investigations in the
Field of Substituted 1,5-Diphenylthiocarbazone. J. Org. Chem. USSR
1957, 27, 2190−2194.
(30) Mirkhalaf, F.; Whittaker, D.; Schiffrin, D. J. Electrochemistry of
ITO and Gold Electrodes Covalently Modified with Dithizone and the
Effect of Cu(II) and Pb(II) Ions. J. Electroanal. Chem. 1998, 452, 203−
213.
REFERENCES
■
(1) Reith, J. F.; Gerritsma, K. W. Photometric Determination of
Micro-Quantities of Mercury as Mercury Dithizonate. Recl. Trav. Chim.
1945, 64, 41−46.
(2) Meriwether, L. S.; Breitner, E. C.; Sloan, C. L. The
Photochromism of Metal Dithizonates. J. Am. Chem. Soc. 1965, 87,
4441−4447.
(3) Von Eschwege, K. G. Synthesis and Kinetics of Electronically
Altered Photochromic Dithizonatophenylmercury(II) Complexes. J.
Photochem. Photobiol. A. 2012, 252, 159−166.
(4) Meriwether, L. S.; Breitner, E. C.; Colthup, N. B. Kinetic and
Infrared Study of Photochromism of Metal Dithizonates. J. Am. Chem.
Soc. 1965, 87, 4448−4454.
(5) Hutton, A. T.; Irving, H. M. N. H.; Nassimbeni, L. R. Irregular
Three-Coordination in Mercury: Structures of Phenyl- and
Methylmercury(II) Dithizonate. Acta Crystallogr. 1980, B36, 2064−
2070.
(6) 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.
(7) Schwoerer, H.; Von Eschwege, K. G.; Bosman, G.; Krok, P.;
Conradie, J. Ultrafast Photochemistry of Dithizonatophenylmercury-
(II). ChemPhysChem. 2011, 12, 2653−2658.
(8) Megerle, U.; Pugliesi, I.; Schriever, C.; Sailer, C.; Riedle, E. Sub-
50 fs Broadband Absorption Spectroscopy with Tunable Excitation:
Putting the Analysis of Ultrafast Molecular Dynamics on Solid
Ground. Appl. Phys. B: Laser Opt. 2009, 96, 215−231.
(9) Lee, C.; Yang, W.; Parr, R. G. Development of the Colle-Salvetti
Conelation Energy Formula into a Functional of the Electron Density.
Phys. Rev. B 1988, 37, 785−789.
(10) Becke, A. D. Density-Functional Exchange-Energy Approx-
imation with Correct Asymptotic Behavior. Phys. Rev. A 1988, 38,
3098−3100.
(11) Becke, A. D. Density-Functional Thermochemistry. I. The Effect
of the Exchange-Only Gradient Correction. J. Chem. Phys. 1992, 96,
2155−2160.
(12) Becke, A. D. Density-Functional Thermochemistry. II. The
Effect of the Perdew-Wang Generalized-Gradient Correlation
Correction. J. Chem. Phys. 1992, 97, 9173−9177.
(13) Becke, A. D. Density-Functional Thermochemistry. III. The
Role of Exact Exchange. J. Chem. Phys. 1993, 98, 5648−5652.
(14) Stephens, P. J.; Devlin, F. J.; Chabalowski, C. F.; Frisch, M. J. Ab
Initio Calculation of Vibrational Absorption and Circular Dichroism
Spectra Using Density Functional Force Fields. J. Phys. Chem. 1994,
98, 11623−11627.
(15) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Montgomery, Jr., J. A.; Vreven, T.;
Kudin, K. N.; Burant, J. C. et al. Gaussian 09, Revision C.01; Gaussian,
Inc.: Wallingford CT, 2004.
(31) Von Eschwege, K. G. Oxidation Resilient Dithizones
Synthesis, Cyclic Voltammetry and DFT Perspectives. Polyhedron
2012, 39, 99−105.
(32) Von Eschwege, K. G.; Conradie, J.; Kuhn, A. Dithizone and its
Oxidation Productsa DFT, Spectroscopic and X-Ray Structural
Study. J. Phys. Chem. A 2011, 115, 14637−14646.
(33) Von Eschwege, K. G.; Van As, L.; Swarts, J. C. Electrochemistry
and Spectro-Electrochemistry of Dithizonatophenylmercury(II). Elec-
trochim. Acta 2011, 56, 10064−10068.
(34) Skoog, D. A.; Holler, F. J.; Crouch, S. R. Principles of
Instrumental Analysis: NMR, 6th ed.; Thompson-Brooks/Cole:
Stamford, CT, 2007, 520.
̌ ́
́
(35) Bonacic-Koutecky, V.; Michl, J. Photochemical Syn-Anti
Isomerization of a Schiff Base: A Two-Dimensional Description of a
Conical Intersection in Formaldimine. Theor. Chim. Acta 1985, 68,
45−55.
(36) Fischer, E.; Frei, Y. Photoisomerization Equilibria Involving the
CN Double Bond. J. Chem. Phys. 1957, 27, 808−809.
(37) Kistiakowsky, G. B.; Smit, W. R. Kinetics of Thermal Cis-Trans
Isomerization. III. J. Am. Chem. Soc. 1934, 56, 638−642.
(38) Taylor, T. W. J.; Murray, A. R. Isomeric Change in Certain
Stilbenes. J. Chem. Soc. 1938, 2078.
(39) Herkstroeter, W. G. Mechanism of Syn-Anti Isomerization of
Azomethine Dyes. J. Am. Chem. Soc. 1973, 95, 8686−8691.
(40) Bosman, G. Transient Absorption Spectroscopy of Metal
Complexes: Dithizonatophenylmercury(II) and Derivatives. Doctoral
Dissertation, University of Stellenbosch, 2012, http://www.laser-
research.co.za/sarchi/Site/Theses.html
(16) McLean, A. d.; Chandler, G. S. Contracted Gaussian Basis Sets
for Molecular Calculations. I. Second Row Atoms, Z=11−18. J. Chem.
Phys. 1980, 72, 5639.
(17) Raghavachari, K.; Binkley, J. S.; Seeger, R.; Pople, J. A. Self-
Consistent Molecular Orbital Methods. 20. A Basis Set for Correlated
Wave Functions. J. Chem. Phys. 1980, 72, 650−654.
(18) Blaudeau, J. P.; McGrath, M. P.; Curtiss, L. A.; Radom, L.
Extension of Gaussian-2 (G2) Theory to Molecules Containing Third-
Row Atoms K and Ca. J. Chem. Phys. 1997, 107, 5016−5021.
(19) Wachters, A. J. H. Gaussian Basis Set for Molecular Wave-
Functions Containing Third-Row Atoms. J. Chem. Phys. 1970, 52,
1033−1036.
(20) Hay, P. J. Gaussian Basis Sets For Molecular Calculations. The
Representation of 3d Orbitals in Transition-Metal Atoms. J. Chem.
Phys. 1977, 66, 4377−4384.
(21) Raghavachari, K.; Trucks, G. W. Highly Correlated Systems.
Excitation Energies of First Row Transition Metals Sc−Cu. J. Chem.
Phys. 1989, 91, 1062−1065.
(41) Zuman, P. Substituent Effects in Organic Polarography; Plenum:
New York, 1967.
(42) Taft, R. W. Sigma Values from Reactivities. J. Phys. Chem. 1960,
64, 1805−1815.
(43) Mulliken, R. S. A New Electroaffinity Scale; Together with Data
on Valence States and on Valence Ionization Potentials and Electron
Affinities. J. Chem. Phys. 1934, 2, 782−793.
(44) Kuhn, A.; Von Eschwege, K. G.; Conradie, J. Reduction
Potentials of Para-Substituted Nitrobenzenes - a Theoretical
Approach. J. Phys. Org. Chem. 2012, 25, 58−68.
(45) Von Eschwege, K. G.; Conradie, J. Redox Potentials of Ligands
and Complexes − a DFT Approach. SA J. Chem. 2011, 64, 203−209.
854
dx.doi.org/10.1021/jp410186y | J. Phys. Chem. A 2014, 118, 844−855

The Journal of Physical Chemistry A
Article
(46) IUPAC, Compendium of Chemical Terminology, 2nd ed. (the
“Gold Book”), IUPAC: Zurich, Switzerland, 1997. Online corrected
version: 2006 “Hammett equation (Hammett relation)”.
(47) Keenan, S. L.; Peterson, K. P.; Peterson, K.; Jacobson, K.
Determination of Hammett Equation Rho Constant for the Hydrolysis
of p-Nitrophenyl Benzoate Esters. J. Chem. Educ. 2008, 85, 558−560.
(48) Von Eschwege, K. G.; Muller, F.; Muller, A. [1,5-Bis(2-
methoxyphenyl)thiocarbazonato-κ2N5,S]phenylmercury(II). Acta Crys-
tallogr. E 2011, E67, m1804.
855
dx.doi.org/10.1021/jp410186y | J. Phys. Chem. A 2014, 118, 844−855