Nitric Oxide Catalysis of Diazene E / Z Isomerization

Article abstract of DOI:10.1021/acs.inorgchem.5b00476

Nitric oxide is an efficient catalyst for the cis-trans (E/Z) isomerization of diazenes. We compare the effect of room temperature solutions bearing low concentrations of nitric oxide, nitrogen dioxide, or oxygen on the rate of cis-trans isomerization, CTI, of the alkene bond in stilbene and on the azo double bond in azobenzene, as well as in four azo derivatives as measured by UV-vis spectroscopy. These rate enhancements can be as large as 3 orders of magnitude for azobenzene in solution. A mechanism is proposed where catalysis is promoted by the interaction of the nitric oxide with the diazene nitrogen lone pairs. Density functional theory, B3LYP/6-311++g suggests that the binding of NO to the diazene should be weak and reversible but that its NO adduct has an E/Z isomerization barrier of 7.5 kcal/mol.

Full text of DOI:10.1021/acs.inorgchem.5b00476

Article
pubs.acs.org/IC
Nitric Oxide Catalysis of Diazene E/Z Isomerization
D. Scott Bohle* and Kristopher A. Rosadiuk
Department of Chemistry, McGill University, 801 Sherbrooke Street W., Montreal H3A 0B8, Canada
S
* Supporting Information
ABSTRACT: Nitric oxide is an efficient catalyst for the cis-trans (E/
Z) isomerization of diazenes. We compare the effect of room
temperature solutions bearing low concentrations of nitric oxide,
nitrogen dioxide, or oxygen on the rate of cis-trans isomerization,
CTI, of the alkene bond in stilbene and on the azo double bond in
azobenzene, as well as in four azo derivatives as measured by UV−vis
spectroscopy. These rate enhancements can be as large as 3 orders of
magnitude for azobenzene in solution. A mechanism is proposed
where catalysis is promoted by the interaction of the nitric oxide with
the diazene nitrogen lone pairs. Density functional theory, B3LYP/6-
311++g** suggests that the binding of NO to the diazene should be
weak and reversible but that its NO adduct has an E/Z isomerization
barrier of 7.5 kcal/mol.
Double-bond isomerization in general is an excellent
candidate for spin catalysis; as the double bond rotates out of
plane, the potential energy surface of the T1 state drops below
that of the ground state (at 90°, the π electrons cannot be in
the same orbital, and this separation is more easily achieved for
two electrons of the same spin, i.e., the triplet state), Figure S1.
Low concentrations of nitric oxide are common in biological
systems, so it is important to clarify whether or not it can act as
a spin catalyst. As part of this investigation, we studied the
interaction between nitric oxide and azo bonds (-NN-) and
compared them to the interactions with the olefin analogue
(stilbene) under tightly controlled conditions. The key
differences between the two systems are the energy of
isomerization (azo bonds are more easily isomerizable), the
presence of lone pairs on nitrogen in place of hydrogen
(potentially allowing association through this lone pair), the
mechanism of isomerization (azo bonds isomerize either by
rotation or by inversion, where olefinic bonds can only rotate),
and the geometry of the cis conformer (cis-stilbene is planar
and somewhat more sterically hindered, whereas cis-azobenzene
twists into a three-dimensional “palm-to-palm” formation
(Figure 1)).13 Differences in how these two systems behave
when exposed to NO can therefore help illuminate the mode of
interaction.
he isomerization of fatty acids by “oxides of nitrogen” was
first reported by J.J.E. Poutet in 1819. Subsequent
1
T
investigations have shown that nitric oxide,2,3 nitrogen
dioxide,4,5 and even oxygen itself6 can catalyze the cis/trans
isomerization (CTI) of olefins. The adventitious oxygenation of
NO to NO2 has led to some confusion as to which is the more
active cis−trans catalyst, and it is noteworthy that for those
working in carefully controlled, low-concentration experiments,
NO is reported as the catalyst for gas-phase experiments, while
researchers working in solution usually report NO2 as the
́
catalyst. De Mare et al. have found that photosensitized NO, in
4
its Π excited state, will catalyze CTI in 2-butene but that
ground state NO on its own is not catalytically active.7,8
Both NO and NO2 are stable radicals, so it is possible that
they catalyze CTI by virtue of a spin exchange,9,10 which allows
access to triplet pathways, Supporting Information, Figure
S1.11,12 What makes these pathways normally inaccessible is the
inability of electrons to spontaneously change spins. In most
cases, intersystem crossing proceeds by spin orbital coupling, in
which the interaction between the magnetic moment of an
electron in a high-energy molecular orbital and the magnetic
spin of that electron is sufficiently strong to effect a change in
spin. This can be considered a form of “spin catalysis”.9
Another form, less well-known, involves an open-shell radical
species associating with a system (such as a transition state) and
exchanging its unpaired electron for an electron of opposite
spin from this associated system, thus transforming a singlet to
a triplet, or vice versa. Since a singlet-to-triplet transfer is
usually energetically uphill, it only occurs in conformations
where the triplet energy has dropped or is otherwise
energetically available and the spin change remains as the
sole barrier. Since the original radical is released unchanged, it
qualifies as a catalyst.
Furthermore, the isomerization of diazo bonds in the
presence of NO has not been previously described. Herein
we test NO as a CTI catalyst in solution at room temperature
and at low concentrations, on stilbene, azobenzene, and four
derivatives of azobenzene (two of which are soluble and two
insoluble in water) (Figure 2). The results are then compared
Received: September 2, 2014
© XXXX American Chemical Society
A
DOI: 10.1021/acs.inorgchem.5b00476
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
reducing to) the minimum water and adding ethanol or isopropanol
will precipitate the inorganic salt. Two further solvent recrystallizations
with water/THF or water/ethanol can separate the desired dye from
the more brightly colored side product. A telltale test of purity is to
leave a very dilute solution of the compound in a quartz vessel under
365 nm UV light for ∼20 min: trace impurities in this reaction will
begin to visibly fluoresce (if left under UV for >1 h, photochemistry
will also produce these fluorescent impurities). UV−vis: 320 (major),
1
433 nm. NMR: H: 7.883 (d), 7.826 (d).
3,3′-Bis(sulfonato)-4,4′-bis(acetamido)azobenzene (BBZ). Diami-
nobenzenesulfonic acid is dissolved in distilled acetic acid with 1 equiv
of acetic anhydride. Acetamido product precipitates and is removed by
filtration. Intermediate product is treated as in the procedure for ADS
to create the rich orange azo-acetamido product. UV−vis: 354
(major), 245, 445 nm.
4-Hydroxy-4′-methoxyazobenzene (HMA). The method used by
Amoros et al.16 was modified. 4-Anisidine is dissolved in 1 M HCl at 0
°C (as well as possible), and equimolar NaNO2 is added as an aqueous
solution. This mixture is allowed to stir for 30 min. Meanwhile,
equimolar phenol is dissolved in 1 M NaOH solution (phenol to
NaOH molar ratio should be 1:2), and stirred. After 30 min, it is added
dropwise to the first solution, and the two are stirred at 0 °C for 1 h.
Solution is neutralized with 5% HCl, and product is filtered out
(adding more water can help precipitate it). Desired product can be
separated with column chromatography, with dichloromethane eluent.
UV−vis: 337, 355, 372 (major peak, vibronic coupling), 440 nm
(shoulder).
Figure 1. Isomerization of trans-stilbene (A) and azobenzene (B).
to DFT calculations to formulate a possible model of NO CTI
catalytic behavior.
4,4′-Dinitroazobenzene. The dinitroazobenzene is synthesized by
dissolving 4-nitroaniline in ethanol, cooling to 0 °C, then adding 10%
NaOCl solution dropwise and stirring for 2 h. It is separated by
column chromatography. UV−vis: 336 (major), 446 nm.
2. Isomerization Protocol. All of the diazo species prepared
undergo rapid isomerization from their thermodynamically favored
trans form to the higher energy cis-form when exposed to low-
intensity UVA light. We found that 5−10 min of 365 nm light with a 7
nm half-width (emitted by an 18.4 W handheld UV lamp) was enough
to achieve a steady state of ∼70% cis (measured by NMR and
confirmed by comparison to the UV spectrum of pure cis) for
azobenzene, ADS, DNazo, and BBZ, and nearly 100% for HMA and
stilbene. HMA, unlike the others, was found to have a profound
solvent dependence; it isomerizes very slowly under 365 nm UV light
in toluene, while in some polar solvents, such as ethanol, the cis-to-
trans thermal return is on the millisecond scale. Ethyl acetate or
tetrahydrofuran (THF) were both found to allow easy isomerization
while maintaining a slow, observable thermal return rate.
Figure 2. Diazo derivatives used in the study. (left) Stilbene,
azobenzene, 4,4′-dinitroazobenzene (DNazo). (right) 4,4′-Azobenze-
nedisulfonate (ADS), 3,3′-bis-sulfanato-4′,4-bis-acetamidoazobenzene
(BBZ), 4-hydroxy-4′-methoxyazobenzene (HMA).
EXPERIMENTAL SECTION
■
The five diazo compounds were first (1) prepared and dissolved in
appropriate solutions, (2) then isomerized under UV light. (3) The
test gases were prepared by two different methods, and (4) the
solutions were treated with these test gases. Finally (5) the rate of the
isomerization back to the original configuration was determined by
UV−vis spectroscopy.
3. Nitric Oxide Preparation. Nitric oxide is delivered to the
solution by either one of two sources.
Method A. A dipropylammonium salt of dipropylamine diazenium
diolate (DPA-NONO, which breaks apart in neutral conditions into 2
equiv of nitric oxide and two of dipropylamine), is prepared according
to literature procedure.17 Stock solutions of 10 mmol/L DPA-NONO
are prepared fresh in 0.01 mol/L NaOH solution on ice and mixed as
needed into a phosphate buffer (0.0029 mol/L K2HPO4 + 0.0073
mol/L KH2PO4 in water, pH ≈ 6.8) 6 min prior to use. Aliquots of
this solution are then injected via syringe, thus giving controlled and
relatively high concentrations of NO in solution. Since the
concentration of NO in solution depends on the partial pressure of
NO in the headspace, the level of dissolved nitric oxide will slowly
drop as the gas diffuses out of solution into the headspace. Thus,
measurements are made soon after injection, with a minimum of
stirring or shaking.
Method B. A commercial sample of nitric oxide is injected with a
gastight syringe into a sealed and purged cuvette (filled with solvent)
and agitated to dissolve the gas. NO2 was purged from the gas stream
by a packed column of potassium hydroxide; the gas was examined by
IR spectroscopy for the presence of N2O, but this was determined to
be present in only trace amounts. This method has the advantage of
saturating the headspace with NO and thus maintaining a
concentration over the long-term, but because the compressibility of
gas allows for large differences when measuring small volumes and
1. Diazo Preparation. The five diazo compounds used (Figure. 2)
were all prepared in house. cis/trans-Stilbene and those reagents used
to prepare the diazo substrates were purchased from Sigma-Aldrich.
Azobenzene. Method used by Hien et al. was modified.14 Aniline
and 2 equiv of sodium perborate are dissolved in acetic acid (2 g of
aniline to 50 mL of acid) at 64 °C. Solution quickly darkens to black.
After 30 min, adding excess hexane extracts a red-orange product.
Removing hexane and residual acid from product, then redissolving in
hexane and recrystallizing, gives a purified azobenzene (vivid orange,
35% yield). UV−vis (in hexane): 314 (major), 441 nm.
Azobenzenedisulfonate. Method used by Zhang et al. was
modified.15 Sulfanilic acid is dissolved in warm water with 2 equiv of
Na2CO3 (1 g of sulfanilic acid to 12 mL of water works well). It is
cooled in an ice−salt bath to below 0 °C. A 15% NaOCl solution is
added dropwise, about one every 3 s. An immediate color change to
deep orange should be seen. When addition is complete, the reaction
vessel is left in a 4 °C refrigerator for 12 h, then filtered: the filtrate will
be rich in the kinetic product (cis-ADS, which must be protected from
ambient light if it is to be kept in this state), while the precipitate will
be predominantly trans-ADS and side products. Dissolving in (or
B
DOI: 10.1021/acs.inorgchem.5b00476
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(RHF)21 and density functional theory (DFT) levels. DFT
calculations used the hybrid B3LYP functional and triple-ζ 6-
311++G** basis sets.22,23 The calculated molecular geometries
were fully optimized and correspond to minima on the
potential energy surface as confirmed by the absence of
imaginary vibrational frequencies. All transition states were
confirmed by reaction path (IRC) following calculations.
reproducible results are therefore more difficult to obtain, it was used
for qualitative comparisons with experiments completed by Method A,
and for the nonaqueous experiments (HMA and stilbene).
Control experiments for NO2 and N2O were also performed:
Nitrous oxide: The gas was obtained from a Praxair cylinder.
The solution to be tested was saturated with the gas.
Nitrogen dioxide: To obtain NO2 with a minimum of NO,
lead(II) nitrate is heated over a flame to produce the gas.
Calculations are adjusted to account for the 0.5 equiv of oxygen
produced per 2 equiv of NO2.18
EXPERIMENTAL RESULTS
■
All -NN- bonded molecules investigated showed increases in
rate of isomerization when in the presence of NO, except for
DNAzo. Stilbene did not respond under these conditions
(Table 1). Control tests were performed with oxygen,
4. Solution Treatment. All solutions were purged of oxygen with
vigorous bubbling under argon on ice (15 min for bulk (∼10 mL
solutions), 5 min for small cuvette samples (<2 mL); initial tests were
performed by freeze−thaw-degassing the solutions, but this yielded the
same results). Since acid is also known to catalyze CTI, all aqueous
solutions are adjusted to pH 6.8 with the phosphate buffer previously
described; ADS and BBZ solutions are prepared in buffer only, while
azobenzene is prepared in 50:50 acetone/buffer, HMA in ethyl acetate,
DNAzo in 50:50 ethanol/buffer, and stilbene is dissolved in hexane.
Dye solutions are placed in a gastight quartz cuvette in a dark box and
irradiated for 5 min with 365 nm light. DPA-NONO is mixed with a
buffer and left to incubate for 6 min (just over 4.3 half-lives, until 95%
has decayed), and the desired amount of this NO solution is injected
into the dye cuvette once it has been transferred to the spectrometer.
The cuvette is mixed, and the rate of isomerization is measured.
5. Measurement. The sample is placed into an HP 8453 Diode-
Array UV−vis spectrometer; it is protected from ambient light
throughout the measurement time, and measurements are evenly
spaced every 60 at 0.5 s exposures over 30 min, to minimize the effects
of the measurement on the rate of isomerization. The possibility of a
back reaction due to photolysis during the UV detection was found to
be insignificant (the difference in isomerization rate between a sample
measured 30 times and one measured twice was found to be less than
the calculated measurement error). All controls were subjected to the
same measurement procedure. The changing absorption at the λmax is
plotted against time, and the rate constant is determined by a least-
squares fit to a first-order curve (catalyzed isomerization will be
pseudo-first-order). Error is reported as the standard deviation of a
minimum of three repeated measurements, or as the error in slope
derived from the variance between the data and the first-order model,
whichever error is larger.
NO2 and N2O control experiments were performed only in dry
hexane solutions of azobenzene and stilbene, as well as ethyl acetate
solutions of HMA. The gas is injected via syringe, and this is compared
to equimolar injections of NO gas.
EPR measurements of the solutions after mixing were taken using a
Bruker ElexSYS 580 X-band spectrometer, in an effort to detect any
NO−azobenzene adducts; low-temperature measurements were
performed in a Wilmad low-temperature dewar insert filled with
liquid nitrogen.
To gauge the turnover capacity of NO, a 1.72 mM solution of
azobenzene in toluene was purged, sealed, saturated with the test gas
by injecting with 1 mL of pure gas at atmospheric pressure, and
irradiated with 365 nm UV light for 17 h. This should be sufficient to
completely isomerize the azobenzene from trans to cis 41.9 times,
based on a rate of 1.4 × 10−7 mol/min in a 1.72 mM toluene solution.
(This rate was estimated from the response of a fresh trans-azobenzene
solution to 1 min of UV exposure, making use of Halpern’s UV−vis
extinction coefficients19 of the two isomers to gauge the concentration
of each). This number of turnovers corresponds to 3.5 times the
amount of NO that was injected, if one molecule of NO was
consumed per CTI. The sample was removed from the UV light, and
the rate of its CTI was measured by UV−vis spectrometry and
compared to both the baseline thermal rate and the rate prior to the 17
h of radiation.
Table 1. Pseudo-First-Order Rate of cis to trans
Isomerization of Various Species under Different Gas
Exposures
conditions of thermal cis to trans
a
species
isomerization
k of CTI, s−1
stilbene (28 μM) with oxygen
2.05(8) × 10−5
2.05(9) × 10−5
1.50(4) × 10−5
3.0(7) × 10−5
∼2(1) × 10−5
1.69(6) × 10−5
no oxygen
dipropylamine
nitric oxide (excess)
nitrogen dioxide (excess)
with oxygen
b
azobenzene
(760 μM)
no oxygen
1.72(2) × 10−5
1.7(2) × 10−5
1.81(7) × 10−2
4.23(1) × 10−4
1.72(2) × 10−5
1.9(2) × 10−5
1.68(1) × 10−4
4.5(8) × 10−4
1.35(1) × 10−3
dipropylamine
nitric oxide (NO/dye 1:2)
nitrogen dioxide (NO2/dye 1:2)
nitrous oxide (large excess)
no oxygen
ADS (50 μM)
no oxygen, in 50% ethanol
nitric oxide (NO/dye 1:2)
nitric oxide (NO/dye 1:2) in 50%
ethanol
BBZ (50 μM)
HMA (50 μM)
no oxygen
8.3(2) × 10−5
2.4(9) × 10−4
3.94(2) × 10−4
1.16(6) × 10−2
3.29(3) × 10−2
nitric oxide (NO/dye 1:2)
in ethyl acetate, no oxygen
in ethyl acetate, (NO/dye 1:2)
in ethyl acetate, with NO2
(NO2/dye 1:2)
c
in ethyl acetate, with ethanol
(EtOH/dye 1:2)
1.58(8) × 10−3
3.9(2) × 10−4
in ethyl acetate, with N2O (large
excess)
DNAzo (50 μM) no oxygen, 50:50 ethanol/water
1.04(1) × 10−4
9.6(3) × 10−5
1.7(4) × 10−5
nitric oxide (NO/dye 1:2)
nitrogen dioxide (1:2 ratio), in ethyl
c
acetate
a
T = 25 °C. Number in parentheses indicates the error in the last
b
reported digit. NO2 injection confounds accurate measurement of
stilbene isomerization via UV, due to overlap and possible side
reaction, but the observed change is within the limits of error for the
c
thermal rate. The solubility of nitrogen dioxide in ethyl acetate was
set to be that of NO in ethyl acetate, as the gases have similar
solubilities in polar solvents.
dipropylamine, and nitrogen dioxide. Nitrous oxide was
included in the tests with azobenzene and HMA, when it was
observed that solvent polarity had a measurable effect (dipole
moment of NO is 0.1595 D, vs 0.166 D for N2O and 0.289 D
for NO2).24−26
THEORETICAL METHODS
■
All of the calculations were performed using Gaussian 03.20
Computations were performed at the restricted Hartree−Fock
C
DOI: 10.1021/acs.inorgchem.5b00476
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Unsubstituted azobenzene responds most dramatically to the
introduction of NO and, to a lesser extent, NO2. The
substituted azo dyes appear at first glance to have had a
meager response, but at slightly higher dose levels, they clearly
show the same behavior, DNAzo excepted. Taking ADS as
representative, we see (Figure 3) the spectrum of a 50 μM
Figure 5. ADS dye response to NO exposure.
back reaction k−1 is small enough that it can be neglected, as
evidenced by the final steady-state ratio of cis/trans, which is
essentially 100% trans. The curves were fitted to this first-order
model using a least-squares approach, and k was calculated at
different dose levels. Plotting the logarithmic factor k versus
time gives a linear dose/response curve (Figure 6).
Figure 3. UV−vis spectrum of 50 μM ADS in water, before UV
exposure (solid line, trans isomer) and after (dashed line, enriched in
cis isomer).
sample at equilibrium (effectively 100% trans) and after 5 min
of exposure to UV (69% cis), (Figure 4) the same cis-enriched
ADS exposed to 100 μM NO, and (Figure 5) a kinetic trace of
the peak at 320 nm when exposed to a range of solvated NO
gas concentrations.
Figure 6. Dependence of first-order rate constants for cis−trans
isomerization (CTI) of ADS as a function of [NO]. Error bars
represent one standard deviation.
Figure 4. UV−vis of cis-enriched ADS after exposure to 100 μM NO.
Measured over 15 min, every 60 s. Lowest signature at 300 nm is at
time 0.
Performing this operation for the three compounds in which
a significant change was found, the rate constant (k) for the
relation d[cis-Dye]/dt = −k[NO][cis-Dye] was calculated:
The normal CTI thermal rate for ADS is a typical first-order
reaction, with k = 1.9 × 10−5 s−1. The initial rates for the
samples exposed to NO follow first-order kinetics; the overall
reaction is thus pseudo-first-order. The generalized standard
formula for reversible first order reactions (eq 1) was used:
azobenzene: kazo = 47( 4) L·mol−1·s−1
ADS: kADS = 18.1( 0.1) L·mol−1·s−1
BBZ: kBBZ = 16.0( 0.6) L·mol−1·s−1
ln((At − A∞)/(A0 − A∞)) = −(k1 + k−1)t
HMA is a special case in that its isomerization can be
provoked by the presence of certain solvents (ethanol, pyridine,
dichloromethane, and water). Ethanol has one of the most
potent effects, so it was chosen for comparison with the gases;
NO2 and NO can be seen to have a greater effect than that of
ethanol, but the effect due to electronic interaction with the
Equation 1: Integrated reversible first-order rate law, where
At is the absorbance at time t, A0 is the absorbance initially, and
A∞ is the absorbance at equilibrium. Ao and A∞ were
determined independently from the kinetics runs, as they
were found to vary slightly with extended UV exposure. The
D
DOI: 10.1021/acs.inorgchem.5b00476
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
solvent cannot be distinguished from spin catalysis, and further
kinetic determinations were not performed for HMA.
relationships are diagramed in Figure 7, and information about
the key metric parameters is collected in Table 3.
When samples were irradiated for 17 h to gauge catalytic
turnover, no difference was found in their rate of CTI after the
long-term irradiation (i.e., after ∼41.9 turnovers). Azobenzene
itself degrades with extensive UV exposure, showing the
development of a broad undefined absorption band between its
major (π−π*) and minor (n−π*) peaks, but a control sample
demonstrated that these photoproducts do not interfere with
the normal rate of CTI and, presumably, do not interfere with
NO-mediated CTI either.
Electron paramagnetic resonance (EPR) measurements of
azobenzene/nitric oxide solutions show a weak anisotropic
signal centered at 3493 G when frozen at −196 °C, extending
from 3460 to 3515 G. This shifts to an isotropic triplet at room
temperature. This signal is attributable to trace amounts of NO2
trapped in solution. No other EPR signals are detectable.
DISCUSSION
■
That nitric oxide catalytically converts cis-azobenzene to trans-
azobenzene is unequivocal. The conversion of cis to trans
occurs rapidly on exposure to NO, and further UV irradiation
will not produce a population of cis-azobenzene, even after 17 h
of continuous UV light. Purging a solution with argon to
remove NO restores normal functionality, and cis-azobenzene
can develop normally. Though catalysts are typically used in
much lower concentrations than their substrates, practical
considerations of measurement and handling with nitric oxide
made this difficult; nevertheless, it fulfills the definition of a
catalyst in that it accelerates reactions without itself being
consumed.
NO cis−trans catalysis functions most readily when a
nitrogen lone pair is available. It does not proceed at all in
stilbene and is also completely ineffective when strong electron-
withdrawing groups (-NO2) are present, which would pull
electron density from the lone pair into the arene ring. More
THEORETICAL RESULTS
■
To model the possible mechanisms for the observed spin
catalysis, DFT calculations (UB3LYP/6-311++g**) were
performed for the most likely ground and transition states for
nitric oxide interacting with azobenzene. The absolute gas-
phase energies for the stationary points are collected in Table 2
−
moderate electron-withdrawing groups (−SO3 ) allow some
catalysis. Weak or moderate electron-donating groups also
impede the action of NO, however, and it may be that the
greater electron density partly occupies the orbital through
which the NO interacts (such as a π*). Furthermore, nitric
oxide is 2 orders of magnitude more catalytically active than
NO2 for azobenzene. NO has a more available unpaired
electron (based on an ionization potential of 9.25 eV, compared
to 11.7 eV for NO227); this electron resides in the π* orbital
and hence is delocalized over the molecule. NO2, by contrast,
has its unpaired e− localized on the nitrogen in a nonbonding
orbital. This suggests that the ability of this electron to be
transferred determines the rate of catalysis. The reduced effect
of NO2 may also be ascribed to steric bulk inhibiting close
association with the N atom beside an arene ring.
Table 2. Key Ground- and Transition-State Energies
(B3LYP/6-311++G**) for Azobenzene and its NO adducts
a
cis
trans
Δ
GS + NO
adducts
A
C
B
−702.815283
−702.816978
−702.810431
−702.8397537
−702.8188314
−702.815107
G
E
F
15.36
1.16
T state to
adducts
T state
between
adducts
−702.80503
D
a
Energies are given in Hartrees except for the differences Δ, which are
given kcal/mol. Capital letters correspond to geometries shown in
Figure 7.
Careful examination of the system by UV−vis and NMR
before and after short-term UV and NO exposure reveals no
side products, either from organic decay or the formation of
nitrites, nitrates, etc. EPR analysis of the mixtures showed no
signal (which is consistent with free nitric oxide, where a bound
azo-NO adduct should be visible). IR of the headspace gas
with letters A to G given to the successive points in the cis-to-
trans isomerization. The relative energies, in kcal/mol, and their
Figure 7. Relative gas-phase energies for the ground states and transition state present in the NO-catalyzed isomerization of azobenzene.
E
DOI: 10.1021/acs.inorgchem.5b00476
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
detects no change in composition (production of N2O3 or
NO2, for example). Litchfield et al. suggested that the
spontaneous formation of NO2 in NO injections is the active
principle for CTI catalysis, but in our case, the greatly reduced
response of azo bonds to injected NO2 rules this out.5
Furthermore, NO2 has a distinctive spectrum under UV, and
this was not seen. We are thus confident in ruling out an
interaction between NO and NO2 as being responsible for the
catalysis, and attribute it solely to NO.
should spin-couple to this system in the way proposed for the
outer-sphere mechanism of diazene−NO catalysis. The fact that
NO appears to have no effect on stilbene under these
conditions suggests that the mechanism of interaction is
fundamentally different and that either gas phase conditions,
high temperatures, or high concentrations are needed to
promote CTI catalysis. There is evidence41 that at higher
temperatures, NO catalyzes olefin isomerization by formal
electron-transfer mechanisms, but this has not been reported at
room temperature.
All of this suggests that nitric oxide is associating with, but
not permanently bonding to, the NN moiety through the
lone pairs. This may be modeled in two different ways: spin
catalysis may proceed by outer-sphere associations, in a
mechanism akin to solvent interactions or Eyring’s sticky
collisions,28 in which transient interactions allow spin exchange
to occur without the formation of chemically reactive transition
states. This spin exchange allows the molecule to cross over
from the S0 to the T1 state, which becomes the lower-energy
pathway as the molecule twists. Alternatively, an inner-sphere
mechanism may operate, without necessarily invoking spin
catalysis, where discrete but weak bonds reversibly form
between the nitric oxide and the diazene. This mechanism
would be similar to how thiyl radicals are proposed to catalyze
E/Z isomerizations in lipids.29 Either mechanism must account
for the dearth of side products or intermediates. Thus, any NO
+ diazene adduct equilibrium should favor the free NO form, be
reversible, and be rapidly established. The inner-sphere
mechanism is well-poised for modeling by modern density
functional ab initio theory, and the results of this modeling is
shown in Figure 6 (above) as well as in Table 2 and Supporting
Information, Table S1.
Theoretically, both cis- and trans-azobenzenes form weakly
stabilized adducts with NO, structures C and E. Both adducts
offer only ∼1 kcal/mol of electronic stabilization, and the
barriers that separate them from cis- and trans-azobenzene,
namely, B and F, are very low (2−3 kcal/mol). The crucial
features in this scheme are the final loss of NO from the trans
adduct E, which favors the free diazene by 13.6 kcal/mol, and
the highest barrier D in this system, 7.5 kcal/mol, is the critical
Z to E isomerization in the adduct. Although these calculations
are for the gas phase and are not zero-point corrected, they
suggest that an inner-sphere mechanism is possible.
CONCLUSION
■
Although nitric oxide has complicated reactivity patterns that
reflect its accessible redox potentials as well as its electro-
philicity,28,42 we have shown that it is a surprisingly robust cis−
trans isomerization catalyst. Previous research into its catalytic
quality has focused, understandably, on the gas phase, where
nitric oxide is most commonly found in the laboratory and in
industry. But this study has demonstrated that it maintains this
behavior in solution, even in water, and can act to isomerize
double bonds to a significant degree at low concentrations.
Furthermore, it does so quite well with azo bonds, a previously
unreported effect. Though the integrity of azo bonds
specifically is not a concern in much of biochemistry, there
are numerous and growing applications for diazenes in
biomedical probes and materials. Any application of these
materials to physiological conditions with even a minor nitric
oxide flux needs to consider these findings.
Indeed, there are many other applications of diazenes, and
questions about the stability of these materials in the presence
of NO (as an atmospheric pollutant for example) must be
considered as well. Diazenes are used extensively in organic
chemistry as stereospecific electrophiles and oxidants.43
Furthermore, the unique facile E/Z structural photoisomerism
of diazenes has led to many creative uses, for example, in fuel
cells,44 fuel additives,45 liquid crystals,46 nonlinear optical
materials,47 anti-retroviral (HIV) drugs,48 antibiotics,49 and
antifungal agents.50 Diazene interactions with the GST family
of enzymes51 has led to their incorporation into other drugs,
such as chemotherapy treatments.52−55 In all of these situations,
potential spin-catalyzed isomerization by low concentrations of
nitric oxide will interfere or modulate their function.
A complete description of this system would require an
understanding of the mechanism of azobenzene isomerization;
unfortunately, there is still controversy around this issue.30−37
Like olefins they may isomerize by rotation (in which the
double bond is broken), but may also do so by inversion (in
which the double bond remains intact but the nitrogen
rehybridizes to an sp form, putting its lone pair in a p orbital
and becoming linear). It is likely that the molecule can proceed
by either mechanism and uses whichever is favored by the
current environment; for example, polar solvents may promote
rotation by stabilizing charged intermediates, whereas inversion
is preferred in a nonpolar solvent.38 NO-catalyzed CTI
proceeds most quickly in less polar solvents (a threefold
increase in the rate for ADS is seen in a 50% ethanol/water mix
compared to pure water), far more than could be explained by
changes in NO solubility. This suggests that the isomerization
may proceed by inversion or at least by an uncharged
intermediate.
A particularly intriguing biochemical consequence of these
findings is that nitric oxide may qualify as a cofactor for
enzymatic activity. Enzymatic transformations that generate
triplet intermediates or progress through triplet excited states
would also be facilitated by this type of “spin cofactor”.
Although applied magnetic fields are able to alter the kinetics of
these types of processes,56,57 albeit to only a modest degree,
nitric oxide’s reversible Lewis acidity allows it to introduce spin
into enzymatic active sites to a much higher degree. When this
feature is combined with its nearly universal permeability to all
organelles and cells, along with its being a compact diatomic
molecule, it suggests there may be a large array of possible
enzymes in which it could be such a cofactor.
ASSOCIATED CONTENT
■
S
* Supporting Information
Synthetic details, discussion of use of diazenium diolates as NO
source, energy profile for CTI of double bonds, calculated
structure for ground and transition states, calculated metrical
parameters, and plots indicating dose response to NO. The
It has been recognized for some time39 that olefinic bonds
can isomerize through either a singlet or triplet state, and
Plachkevytch40 et al. have calculated that, in theory, nitric oxide
F
DOI: 10.1021/acs.inorgchem.5b00476
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(25) Liu, Y.; Guo, Y.; Lin, J.; Huang, G.; Duan, C.; Li, F. Mol. Phys.
2001, 99, 1457−1461.
Supporting Information is available free of charge on the ACS
Publications website at DOI: 10.1021/acs.inorgchem.5b00476.
(26) Reinartz, J. M. L. J.; Meerts, W. L.; Dymanus, A. Chem. Phys.
1978, 31, 19−29.
AUTHOR INFORMATION
■
(27) Lowe, J. P. J. Am. Chem. Soc. 1977, 99, 5557−5570.
(28) Bohle, D. S. In Stable Radicals; Hicks, R. G., Ed.; Wiley: New
York, 2010; p 147−171.
(29) Renaud, P.; Sibi, M. P. Radicals in Organic Synthesis; VCH-
Wiley: Weinheim, Germany, 2001.
Corresponding Author
*E-mail: Scott.Bohle@McGill.CA.
Notes
The authors declare no competing financial interest.
(30) Crecca, C. R.; Roitberg, A. J. Phys. Chem. A 2006, 110, 8188−
8203.
(31) Blevins, A. A.; Blanchard, G. J. J. Phys. Chem. B 2004, 108,
4962−4968.
ACKNOWLEDGMENTS
■
The authors gratefully acknowledge NSERC and the CRC for
support of this research.
(32) Gagliardi, L.; Orlandi, G.; Bernardi, F.; Cembran, A.; Garavelli,
M. Theor. Chem. Acc. 2004, 111, 363−372.
(33) Cembran, A.; Bernardi, F.; Garavelli, M.; Gagliardi, L.; Orlandi,
G. J. Am. Chem. Soc. 2004, 126, 3234−3243.
REFERENCES
■
(34) Rau, H.; Luddecke, E. J. Am. Chem. Soc. 1982, 104, 1616−1617.
̈
(1) Griffiths, H. N.; Hilditch, T. P. J. Chem. Soc. 1932, 2315−2324.
(2) Egger, K. W.; Benson, S. W. J. Phys. Chem. 1967, 71, 1933−1936.
(3) Flowers, M. C.; Jonathan, N. J. J. Chem. Phys. 1969, 50, 2805−
2812.
(4) Khan, N. A. J. Chem. Phys. 1955, 23, 2447−2448.
(5) Litchfield, C.; Harlow, R. D.; Isbell, A. F.; Reiser, R. J. Am. Oil
Chem. Soc. 1965, 42, 73−78.
(35) Dokic, J.; Gothe, M.; Wirth, J.; Peters, M. V.; Schwarz, J.; Hecht,
S.; Saalfrank, P. J. Phys. Chem. A 2009, 113, 6763−6773.
(36) Asano, T.; Okada, T. J. Org. Chem. 1984, 49, 4387−4391.
(37) Gabor, G.; Fischer, E. J. Phys. Chem. 1971, 75, 581−583.
(38) King, N. R.; Whale, E. A.; Davis, F. J.; Gilbert, A.; Mitchell, G. R.
J. Mater. Chem. 1997, 7, 625−630.
(39) Magee, J. L.; Shand, W. J.; Eyring, H. J. Am. Chem. Soc. 1941, 63,
677−688.
(40) Plachkevytch, O.; Minaev, B.; Agren, H. J. Phys. Chem. 1996,
100, 8308−8315.
(41) Rabinovitch, B. S.; Looney, F. S. J. Chem. Phys. 1955, 23, 2439.
(42) Arulsamy, N.; Bohle, D. S. J. Org. Chem. 2006, 71, 572−581.
(43) Kosmrlj, J.; Kocevar, M.; Polanc, S. Synlett 2009, 2009, 2217−
2235.
(44) Uchiyama, S.; Matsuura, H. JP2013202430A, 2013.
(45) Fernandes, J. B. US20050229479A1, 2005.
(46) Podruczna, M.; Hofmanska, A.; Niezgoda, I.; Pociecha, D.;
Galewski, Z. Liq. Cryst. 2014, 41, 113−125.
(47) Raposo, M. M. M.; Fonseca, A. M. C.; Castro, M. C. R.; Belsley,
M.; Cardoso, M. F. S.; Carvalho, L. M.; Coelho, P. J. Dyes Pigm. 2011,
91, 62−73.
(48) Sanchez, T. W.; Debnath, B.; Christ, F.; Otake, H.; Debyser, Z.;
Neamati, N. Bioorg. Med. Chem. 2013, 21, 957−963.
(49) Khalaf, A. I.; Anthony, N.; Breen, D.; Donoghue, G.; MacKay, S.
P.; Scott, F. J.; Suckling, C. J. Eur. J. Med. Chem. 2011, 46, 5343−5355.
(50) Sareen, V.; Khatri, V.; Shinde, D.; Sareen, S. Indian J. Chem.
2011, 50B, 937−940.
(51) Polanc, S. J. Heterocycl. Chem. 2005, 42, 401−412.
(52) Cimbora, T.; Bombek, S.; Polanc, S.; Osmak, M. Toxicol. In
Vitro 2003, 17, 159−164.
(53) Cimbora-Zovko, T.; Bombek, S.; Kosmrlj, J.; Kovacic, L.;
Polanc, S.; Katalinic, A.; Osmak, M. Drug Dev. Res. 2004, 61, 95−100.
(54) Jakopec, S.; Dubravcic, K.; Brozovic, A.; Polanc, S.; Osmak, M.
Cell Biol. Toxicol. 2006, 22, 61−71.
(55) Martin-Kleiner, I.; Bombek, S.; Kosmrlj, J.; Cupic, B.; Cimbora-
Zovko, T.; Jakopec, S.; Polanc, S.; Osmak, M.; Gabrilovac, J. Toxicol. In
Vitro 2007, 21, 1453−1459.
(56) Afanasyeva, M. S.; Taraban, M. B.; Purtov, P. A.; Leshina, T. V.;
Grissom, C. B. J. Am. Chem. Soc. 2006, 128, 8651−8658.
(57) Anderson, M. A.; Xu, Y.; Grissom, C. B. J. Am. Chem. Soc. 2001,
123, 6720−6721.
(6) Srorch, H. H. J. Am. Chem. Soc. 1934, 56, 374−378.
(7) De Mare, G. R.; Lantonnois van Rode, Y. J. Photochem. Photobiol.,
A 1992, 63, 131−138.
(8) De Mare,
19, 131−145.
(9) Buchachenko, A. L.; Berdinsky, V. L. Chem. Rev. 2002, 102, 603−
612.
́
G.; Lantonnois van Rode, Y. Res. Chem. Intermed. 1993,
(10) Buchachenko, A. L.; Berdinsky, V. L. Russ. Chem. Rev. 2004, 73,
1033−1039.
(11) Plachkevytch, O.; Minaev, B.; Agren, H. J. Phys. Chem. 1996,
100, 8308−8315.
(12) Minaev, B. F.; Agren, H. Int. J. Quantum Chem. 1996, 57, 519−
532.
(13) Yager, K. G.; Barrett, C. In Polymeric Nanostructures and Their
Applications; Nalwa, H. S., Ed.; American Scientific Publishers: Los
Angeles, CA, 2007; Vol. 2, pp 243−280.
(14) Hien, L. T.; Schierling, B.; Ryazanova, A. Y.; Zatsepin, T. S.;
Volkov, E. M.; Kubareva, E. A.; Velichko, T. I.; Pingoud, A.; Oretskaya,
T. S. Russ. J. Bioorg. Chem. 2009, 35, 549−555.
(15) Zhang, Z.; Burns, D. C.; Kumita, J. R.; Smart, O. S.; Woolley, G.
A. Bioconjugate Chem. 2003, 14, 824−829.
(16) Garcia-Amoros, J.; Sanchez-Ferrer, A.; Massad, W. A.; Nonell,
S.; Velasco, D. Phys. Chem. Chem. Phys. 2010, 12, 13238−13242.
(17) Bohle, D. S.; Smith, K. N. Inorg. Chem. 2008, 47, 3925−3927.
(18) Nicol, A. Compt. Rend. 1948, 226, 253−255.
(19) Halpern, J. Ph.D. Thesis, McGill University, 1948.
(20) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A.;
Stratmann, R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels, A.
D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.; Barone, V.;
Cossi, M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford,
S.; Ochterski, J.; Petersson, J. A.; Ayala, P. Y.; Cui, Q.; Morokuma, K.;
Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.;
Cioslowski, J.; Ortiz, J. V.; Stefanov, B. B.; Liu, G.; Liashenko, A.;
Piskorz, P.; Kamaromi, I.; Gomperts, R.; Martin, R. L.; Fox, D. J.;
Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Gonzalez,
C.; Challacombe, M.; Gill, P. M. W.; Johnson, B. G.; Chen, W.; Wong,
M. W.; Andres, J. L.; Head-Gordon, M.; Replogle, E. S.; Pople, J. A.;
Gaussian 98; Gaussian Inc.: Pittsburgh, PA, 1998.
(21) Roothaan, C. C. Rev. Mod. Phys. 1951, 23, 69−89.
(22) Lee, C.; Yang, W.; Parr, R. G. J. Mol. Struct.: THEOCHEM 1988,
40, 305−313.
(23) Becke, A. D. J. Chem. Phys. 1993, 98, 1372−1377.
(24) Heitz, S.; Lampka, R.; Weidauer, D.; Hese, A. J. Chem. Phys.
1991, 94, 2532−2535.
G
DOI: 10.1021/acs.inorgchem.5b00476
Inorg. Chem. XXXX, XXX, XXX−XXX