Design and synthesis of photodissociable ligands based on azoimidazoles for light-driven coordination-induced spin state switching in homogeneous solution

Article abstract of DOI:10.1021/acs.joc.5b02817

Light-switchable azoimidazoles were rationally designed and synthesized, and their performance was investigated as photodissociable ligands (PDL) and for spin state switching of Ni porphyrins. The rationally designed ligands exhibit a high photochemical conversion rate (trans a? cis > 98%) and no measurable fatigue over a large number of switching cycles at room temperature under air. As compared to the known phenylazopyridines, the phenylazoimidazoles exhibit a much stronger affinity as axial ligands to Ni porphyrin in the binding trans configuration and a low affinity in their cis form. This affinity switching was used to control the coordination number of Ni²⁺. Concomitant with the change in coordination number is the change of the spin state from triplet (high spin) to singlet state (low spin). We report on phenylazoimidazole-based PDLs that switch the paramagnetic ratio of the investigated nickel species by up to 70%. Consequently, azoimidazoles exhibit considerably higher switching efficiencies than previously described pyridine-based PDLs.

Full text of DOI:10.1021/acs.joc.5b02817

Article
pubs.acs.org/joc
Design and Synthesis of Photodissociable Ligands Based on
Azoimidazoles for Light-Driven Coordination-Induced Spin State
Switching in Homogeneous Solution
Christian Schutt, Gernot Heitmann, Thore Wendler, Bahne Krahwinkel, and Rainer Herges*
̈
Otto Diels-Institute for Organic Chemistry, University of Kiel, Otto-Hahn-Platz 4, Kiel D-24119, Germany
S
* Supporting Information
ABSTRACT: Light-switchable azoimidazoles were rationally
designed and synthesized, and their performance was
investigated as photodissociable ligands (PDL) and for spin
state switching of Ni porphyrins. The rationally designed
ligands exhibit a high photochemical conversion rate (trans →
cis > 98%) and no measurable fatigue over a large number of
switching cycles at room temperature under air. As compared
to the known phenylazopyridines, the phenylazoimidazoles
exhibit a much stronger affinity as axial ligands to Ni porphyrin in the binding trans configuration and a low affinity in their cis
form. This affinity switching was used to control the coordination number of Ni2+. Concomitant with the change in coordination
number is the change of the spin state from triplet (high spin) to singlet state (low spin). We report on phenylazoimidazole-
based PDLs that switch the paramagnetic ratio of the investigated nickel species by up to 70%. Consequently, azoimidazoles
exhibit considerably higher switching efficiencies than previously described pyridine-based PDLs.
by adding and removing axial ligands from a square planar Ni
porphyrin (Figure 1).20
INTRODUCTION
■
Switchable ligands for photocontrolled binding are of general
interest in a large number of fields. Practical applications range
from the reversible switching of the activity of drugs, inhibitors,
ion channels, and chemical optogenetics.1−3 Particularly
important as well is the light control of optical, electric, and
magnetic properties of materials (such as the orientation of the
magnetization or the magnetic susceptibility). Magnetic
bistability at room temperature is a typical materials property.
Cooperative effects between a large number of spin centers in the
solid state stabilize the magnetic states and lead to hysteresis.4
Notably, spin crossover and the light-induced excited spin state
trapping (LIESST) phenomenon have been used to switch the
spin state in crystals of transition metal complexes.5 Until
recently, however, there was little work on spin switching in the
solution phase6,7 or on single molecules at surfaces,8,9 even
though this would open a number of interesting applications,
such as data storage or their use as switchable contrast agents in
MRI. For achieving bistability in isolated molecules, the solid
state cooperative effects have to be replaced by molecule inherent
mechanisms. We have recently demonstrated that the spin state
of isolated molecules can be switched between low spin
(diamagnetic) and high spin (paramagnetic) using light.10−13
Our approach is based on the fact that a number of transition
metal ions (e.g., Fe2+, Fe3+, Co2+, Ni2+) change the spin state
upon changing their coordination number. We chose Ni2+
because it is stable in its oxidation state and the spin state
change is reliable and predictable. Square planar (4-coordinate)
Ni2+ is low spin (S = 0), and square pyramidal (5-coordinate) or
square bipyramidal (6-coordinate) Ni2+ are high spin (S =
1).14−19 Coordination change, and thus spin change, is achieved
Figure 1. Coordination-induced spin state switch (CISSS) in Ni(II)
complexes.
Light-induced axial ligand association and dissociation is
operated either by tethering the switchable ligands to the square
planar Ni porphyrin (record player design)11,13,21 or by properly
designed photodissociable ligands (PDLs).10,12 Both designs
have pros and cons. “Record player” complexes that move the
coordinating nitrogen up and down like a tone arm the needle
exhibit a large switching efficiency; however, at large
concentrations, they undergo intermolecular coordination.
PDLs, if properly designed, are sterically hindered in one of
the two photoisomers, which prevents inter- as well as
intramolecular coordination. Moreover, new types of PDLs
Received: December 11, 2015
© XXXX American Chemical Society
A
DOI: 10.1021/acs.joc.5b02817
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
might also find applications as switchable drugs, inhibitors, or
other biochemical probes. The most frequently used PDLs are
probably azopyridines.22−25 They combine the photochromism
of azobenzene and the coordination properties of pyridine in one
small molecule and are easy to synthesize. Other heterocyclic azo
compounds are scarce.26 Azo heterocycles with a higher basicity,
nucleophilicity, and coordination power than azopyridines would
be particularly valuable for the switching of functions, such as
spin state switching, or the control of binding properties, e.g., in
active sites of enzymes. We chose imidazole as the heterocycle,
replacing one phenyl group in azobenzene. The binding constant
of N-methylimidiazole to Ni porphyrins is more than an order of
magnitude higher compared to pyridine and three times the value
of 4-methoxypyridine.27 There are three regioisomers of phenyl
azoimidazoles (azo group in 2, 4, and 5 position if tautomerism is
prevented by substitution at N-1). However, imidazoles that bear
substituents adjacent to the coordinating nitrogen (in 2- and 4-
position) do not coordinate to metal porphyrins for steric
reasons (Figure 2).28,29 Only 5-phenylazoimidazoles are
potential candidates for coordination switching (see also
coordination of histidine to iron in hemoglobin as an example
in nature).
the two switching states. In a two-component system (metal
complex/photodissociable ligand or receptor/photoswitchable
inhibitor), it seems obvious that the maximum switching
efficiency (SEmax) depends on the concentration of both
components. Surprisingly, this is not the case. There is an
optimal ligand concentration ([L]opt) that leads to the maximum
switching efficiency (SEmax) independent of the metal complex
(receptor) concentration.12 The maximum switching efficiency
and the optimal ligand concentration are system parameters. The
most important conclusion that we draw from this finding is that
it is more important to optimize the conversion to the non-
binding isomer and to reduce its coordination by steric hindrance
than optimizing the binding state.
RESULTS AND DISCUSSION
■
Steric Design at the Imidazole. The starting point of our
design was the introduction of bulky substituents in such a way
that the binding of the cis isomer would be impaired while leaving
the coordination of the trans form unaffected. To achieve this
goal, one has to consider that both isomers have two
conformations. To prevent the cis isomer from binding, both
conformations (cis-α and cis-β, Figure 2) have to be included in
the design process. In contrast to the cis-α conformer, the
coordination of the cis-β conformer to the porphyrin is not
sterically hindered because the phenyl group points away from
the porphyrin (Figure 2, right). Thus, it is obvious that steric
constraints have to be introduced at the N-1 position (Figure 2,
right, blue arcs) to disfavor the formation of the cis-β
conformation. For predicting which substituent would be needed
to sufficiently disfavor the cis-β conformer, quantum chemical
calculations were performed using the program TURBO-
MOLE31 6.3 at the PBE/SVP level of DFT. The relative energies
show that in the parent system (R1 = H, 1) both conformations
cis-α and cis-β are almost equal in energy (Figure 3). Upon
introducing a methyl group (R1 = Me, 2) in the 1-position of the
imidazole, the formation of the cis-β conformer is already
disfavored by more than 5 kcal mol−1.
Further increasing the steric repulsion between the 1-position
and the phenyl ring leads to a larger energy difference of the two
Figure 2. Schematic representation of the steric implications upon
coordination of different phenylazoimidazoles to metal porphyrins.
Imidazoles with substituents next to the coordinating nitrogen (2- and 4-
phenylazoimidazole) do not bind to the metal for steric reasons (red
arcs). Only 5-phenylazoimidazole provides a basis for the development
of PDLs. The trans and the cis isomers each can adopt two
conformations (α and β). Both trans conformations are binding. One
of the cis conformations (α) exhibits sterical problems upon
coordination (red arcs); however, the β conformation does not. The
cis-β conformer would still bind, and photoisomerization of the ligand
would not lead to photodissociation. However, formation of the cis-β
conformation could be hampered by steric interaction of the substituent
R at N-1 and the phenyl ring (blue arcs).
We recently published a convenient synthesis for the hitherto
unknown 5-phenylazoimidazoles using a novel type of
azocoupling of lithiated imidazole with benzene diazonium
salts.30 We now report on the rational design of substituted 5-
phenylazoimidazoles, which are structurally optimized with
respect to their switching efficiency (SE). The switching
efficiency is defined as the difference of a target parameter in
Figure 3. Bar diagram of the energy difference (Eβ − Eα) of the α and β
conformation of several N-1-substituted azoimidazoles at the PBE/SVP
level of theory. Energies are given in kcal mol−1. Intramolecular steric
repulsion is indicated by blue arcs.
B
DOI: 10.1021/acs.joc.5b02817
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
cis conformers. The introduction of an isopropyl group in the 1-
position (3, ΔE = 7.04 kcal mol−1) disfavors the formation of the
cis-β conformer roughly by an additional 1.5 kcal mol−1, whereas
substitution with a trimethylsilyl- (4, ΔE = 9.65 kcal mol−1) or
tert-butyl-group (5, ΔE = 11.32 kcal mol−1) leads to a
significantly larger energy difference compared to phenyl-
azoimidazole 2 (see Figure 3). As a result, the non-binding cis-
α conformer should be the predominant species in solution in all
cases except R = H. To gain information on how substitution
(R1) at N-1 would affect the coordination of the trans and the cis
ligands to Ni porphyrins, and more specifically, if disfavoring the
binding cis-β conformation would be sufficient to obtain an
efficient photodissociable ligand, we calculated the binding
energy of the four species (trans-α, trans-β, cis-α, and cis-β) of
azoimidazoles 1−5 to the Ni porphyrin NiTPPF20 at the B3LYP/
def2TZVP//PBE/SVP level of density functional theory. This
level has proven to yield reliable results for the formation of
complexes of 1-methylimidazole (see Supporting Information)
with NiTPPF20 (Table 1). Because the α and β conformations are
Table 1. Calculated Complex Formation Energies (ΔEf, kcal
mol−1) of 1−5 (see Figure 4) with NiTPPF20 at the B3LYP/
def2TZVP//PBE/SVP Level of Theory
Figure 4. Calculated (B3LYP/def2TZVP//PBE/SVP) complex for-
mation energies (ΔEf in kcal mol−1) of phenylazoimidazole derivatives
(1−5) with NiTPPF20. The definition of the complexation energy (ΔEf)
is given on top. Intermolecular and intramolecular steric hindrance is
indicated with red and blue arcs. The energy differences between the
strongest binding trans and cis conformations of each ligand are
indicated by arrows. The N-methyl-substituted ligand (2) exhibits the
largest change in binding energy (ΔΔEf = 2.88 kcal mol−1) upon trans-
cis isomerization.
a
R1
trans-α
trans-β
cis-α
cis-β
ΔΔEf,calc
1
2
3
4
5
H
−4.35
−7.70
−8.54
−7.42
−8.24
−7.70
−8.08
−7.07
−6.79
−4.26
−5.18
−5.20
−6.66
−6.86
−6.44
−6.67
−0.65
−0.79
3.63
1.03
2.88
1.88
0.57
1.80
Me
iPr
SiMe3
tBu
3.44
a
Energy difference of ΔEf of the strongest binding trans and cis
conformer.
zole (2a−c, Figure 5). Electron-donating and -withdrawing
groups were introduced at the phenyl group. Substitution with
in fast equilibrium, the complex formation energies listed in
Table 1 are relative to the most stable conformation. The
calculated binding energies in Table 1 and Figure 4 clearly
predict that the parent system (R = H) would bind strongly in all
configurations and conformations, and therefore, it would be a
very inefficient photodissociable ligand. Introduction of a methyl
group at N-1, as expected, prevents the formation of the complex
of the cis-β species with NiTPPF20 almost completely (ΔEf =
t
−0.65 kcal mol−1). For the larger substituents SiMe3 and Bu,
complex formation of the cis-β species is predicted to be
endothermic (3.63 and 3.44 kcal mol−1).
Even though introduction of substituents such as Me, Pr,
i
SiMe3, and tBu at the imidazole nitrogen efficiently prevent the
formation of the complexes with the β-cis ligand, the α-cis ligands
still bind with ΔEf < −5 kcal mol−1. Obviously, steric interaction
of the phenyl group with the porphyrin (red arcs in Figure 4) is
not sufficient to disfavor complex formation. Therefore, the steric
hindrance of the phenyl group has to be increased by substitution
of the phenyl ring with bulky substituents.
Photophysical Design of the Ligand. The photophysical
properties of the parent azoimidazole switch, 1-methyl-5-
phenylazoimidazole (2), have been recently investigated.30 The
light-induced conversion (365 nm) of the more stable trans to cis
state is extremely efficient (>99%), and the thermal back reaction
is very slow (t1/2 = 22 days at 25 °C). Unfortunately, the light-
induced isomerization of the cis isomer back to the trans form
with visible light (455 nm) is incomplete (40%) because there is a
considerable overlap of the π−π* and n−π* bands in the cis
configuration. To improve the photochemical back reaction, we
synthesized several derivatives of 1-methyl-5-phenyl azoimida-
Figure 5. Photostationary states of 4′-substituted 1-methyl-5-phenyl-
azoimidazoles determined by UV spectroscopy. Irradiation times: 20
min.
fluorine in the para position (2a) improves the photostationary
state (PSS) at 525 nm from 40 to 54% trans, whereas the trans to
cis isomerization (PSS 365 nm) is almost unchanged. The
introduction of electron-donating groups in the para position
further improves the cis to trans conversion (PSS 525 nm) with
increasing donor strength from 85% (OMe, 2b) to 95% (NMe2,
2c). However, although the trans to cis conversion is almost
unchanged for 2b (97%), the photo-induced switching to the cis
isomer in 2c decreases to 59% (Figure 5). TD-DFT calculations
(see Supporting Information) provide a rational understanding
of the experimental results. They predict the largest separation of
the π−π* and n−π* bands for 2b.
C
DOI: 10.1021/acs.joc.5b02817
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
The results of the experimental and theoretical investigations
lead to the conclusion that a methoxy group in the para position
of the phenyl ring is the most effective substituent to improve the
switching efficiency of 5-phenyl-azoimidazoles. Therefore, 2b is a
promising starting point for the further design of PDLs.
Steric Design at the Phenyl Group. As stated before, the
PDL design has to include both conformations (α and β) of the
cis isomer. For preventing the cis isomer from binding, the
formation of the β conformation was suppressed by introduction
of a methyl group at the imidazole ring (Figure 3). However, as
predicted by theoretical calculations, steric hindrance in the α
conformation obviously is not sufficient to prevent the parent
system 2 from binding (Table 1, Figure 4). Therefore, we have to
increase the steric demand of the phenyl group by the
introduction of bulky substituents (Figure 6).
Figure 7. Calculated (B3LYP/def2TZVP//PBE/SVP) complex for-
mation energies. The energy difference between the trans-β conformer
and strongest binding cis configuration of the considered ligands are
indicated by arrows. Ligand 2e exhibits the largest change in binding
energy (ΔΔEf = 4.17 kcal mol−1) upon trans-cis isomerization. Energies
are given in kcal mol−1. Intermolecular and intramolecular steric
repulsion is indicated with red and blue arcs.
Figure 6. Binding modes of trans- and cis-1-methylphenylazoimidazole
derivatives. Intermolecular steric hindrance of the cis isomer with
NiTPPF20 is indicated with red curved lines. Intramolecular hindrance is
indicated by blue arcs.
Table 2. Calculated Complex Formation Energies (ΔEf, kcal
mol−1) of 2, 2b, 2d, and 2e (see Figure 7) with NiTPPF20
(B3LYP/def2TZVP//PBE/SVP)
We calculated the complex formation energies of the
photophysically optimized methoxy derivative 2b and the
sterically more demanding systems 2d and 2e with NiTPPF20
at the B3LYP/def2TZVP//PBE/SVP level of theory to quantify
the effect of the steric bulk (Figure 7 and Table 2).
The calculations predict that two t-butyl groups in the meta
position are most effective in preventing the α conformation
from binding to the Ni porphyrin. Methyl substituents in the
meta position or the methoxy group in the para position are
considerably less effective.
a
trans-α
trans-β
cis-α
cis-β
ΔΔEf,calc
2
−7.70
−7.86
−7.15
−8.20
−8.08
−8.16
−8.45
−8.43
−5.20
−6.40
−6.94
−4.26
−0.65
−2.98
−2.07
−0.79
2.88
1.76
1.51
4.17
2b
2d
2e
a
Energy difference of ΔEf of the strongest binding trans and cis
conformer.
Thus, 2e is the most promising candidate for an effective PDL.
The methoxy group in the para position improves the
photophysical properties; the methyl group at the imidazole
ring suppresses the formation of the β conformation, and the two
t-butyl substituents prevent the α conformation from binding.
Synthesis and Properties of the Improved PDL.
Azoimidazole 2e was synthesized analogously to a previously
published strategy.30 The key step is the directed lithiation of a
doubly protected imidazole and the reaction with a phenyl-
diazonium salt (prepared as tetrafluoroborate in five steps out of
3,5-di-tert-butyl-4-hydroxybenzoic acid; overall yield of 31%),
forming the azo compound by a polar C−N bond formation
(Scheme 1). Both protecting groups at the imidazole are then
removed. A trityl (Tr) group is regioselectively introduced at the
sterically less hindered nitrogen. Subsequent methylation of the
second nitrogen leads to the imidazolium ion, which easily
cleaves the Tr group leading to the desired 1,5 substitution
pattern at the imidazole ring. The overall yield over nine steps
(starting from sulfamoyl imidazole) is 37%. The preparation is
quite convenient because only two intermediate compounds
have to be isolated in the three pot reaction.
We determined the association constants (K) and enthalpies
of complex formation (ΔH) of the cis and trans isomer of 2e with
NiTPPF20 using a previously developed NMR titration method
(for details, see the Supporting Information).10 For comparison
with the theoretically predicted energies of complex formation
and to quantify the improvement of the switching efficiency by
our molecular design, we also included the parent system 2 and
2b in our experimental studies. Table 3 summarizes the
experimental values for K1S and ΔH1S (formation of the 1:1
complex) for 2, 2b, and 2c in cis and trans configurations. Our
molecular design strategy aims at increasing the difference in
binding energy of the cis and trans isomer (ΔΔH1S). Introduction
of the two tert-butyl substituents in 2e indeed decreases the
binding energy of the cis isomer substantially while having little
effect on the association of the trans form. The ratio between the
trans and cis association constants (K1S,trans/K1S,cis) in the parent
system (2) is 3.60 and increases to 43.8 in the sterically hindered
system 2e (Table 3). It is noteworthy that the experimentally
derived binding enthalpy differences are in good agreement with
the calculated values (Table 3).
D
DOI: 10.1021/acs.joc.5b02817
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
a
binding of the trans, and residual binding of the cis isomer. Using
a mathematical analysis, we have previously shown that there is
an optimum ligand (PDL) concentration [L]opt that leads to a
maximum switching efficiency (SEmax).10 A plot of the switching
efficiency as a function of the ligand (PDL) concentration of 2e is
given in Figure 8.
Scheme 1. Synthesis of Azoimidazole 2e
a
Reagents and conditions. (a) 1. MeI, KOH, acetone, 60 °C, 18 h; 2.
LiOH, THF, H2O, 60 °C, 18 h, 71%; (b) DPPA, Et3N, tBuOH, 85 °C,
18 h, 72%; (c) TFA, CH2Cl2, rt, overnight, 92%; (d) isopentyl nitrite,
HBF4 (aq), EtOH, rt, 15 min, 65%; (e) 1. n-BuLi, THF, −78 °C, 30
min; 2. dimethylthexylchlorosilane, THF, −78 °C to rt, overnight; (f)
n-BuLi, THF, −78 °C, 30 min; (g) 1. THF, −78 °C to rt, overnight; 2.
TBAF trihydrate, CsF, THF, rt, overnight, 58% (5 steps, 1 pot); (h)
HCl, EtOH, 50 °C, 3 h; (i) trityl chloride, Et3N, CH2Cl2, rt, overnight,
77% (2 steps, 1 pot); (j) 1. methyl triflate, CH2Cl2, rt, overnight; (k)
H2O, acetone, CH2Cl2, rt, 3 h, 82%.
Figure 8. Switching efficiency (SE) (black squares) as a function of the
ligand concentration of 2e calculated from the experimentally
determined association constants.10 The maximum switching efficiency
(SEmax = 70%) is predicted at a ligand concentration ([L]opt) of 64.3
mM. The gray bar indicates the concentration used in our experiments
(67.9 mM), which is close to the optimum ligand concentration.
The maximum switching efficiency SEmax for the system
NiTPPF20/2e is obtained at the optimum ligand concentration
of 2e ([L]opt = 64.3 mM). The optimum ligand concentrations
for 2 and 2b are 35.1 and 21.9 mM. Surprisingly, a rigorous
mathematical treatment proves that [L]opt is independent of the
receptor (NiTPPF20) concentration over a concentration range
from 0 to 4 mM. (Figure 9).
Table 3. Experimentally Determined Association Constants
(K1S) and Binding Enthalpies (ΔH1S) for Coordination of the
trans and cis Isomers of 2, 2b, and 2e to NiTPPF20
2
2b
2e
a
K1S,trans
38.31
−6.84
10.76
−4.69
2.15
59.14
−6.63
15.83
−4.97
1.66
60.03
−7.10
1.37
The switching experiments including 2, 2b, and 2e were
performed with concentrations close to the optimum ligand
b
ΔH1S,trans
a
K1S,cis
b
ΔH1S,cis
−3.01
4.09
c
ΔΔH1S
d
ΔΔE1S
2.88
1.75
4.18
b
a
At 298 K for 2 and 2b, 300 K for 2e, toluene-d8, [L mol−1]. In [kcal
c
mol−1]. Experimental binding enthalpy differences to NiTPPF20
d
(ΔΔH1S = ΔH1S,cis − ΔH1S,trans) [kcal mol−1]. Calculated binding
energy differences (B3LYP/def2TZVP//PBE/SVP) to NiTPPF20
(ΔΔE1S = ΔE1S,cis − ΔE1S,trans) [kcal mol−1].
Switching Experiments. On the basis of the experimentally
determined association constants (K1s, K2, for data, see the
Supporting Information) of the trans and cis configurations of 2,
2b, and 2e, the conditions for achieving the maximum switching
efficiency SEmax can be theoretically calculated. The switching
efficiency is defined as the difference of the properties of a system
in the two switching states. The properties of interest are the
amount of cis isomer, or in our case, even more interesting is the
amount of paramagnetic species in solution in both photosta-
tionary states. A perfect system would exhibit 0% paramagnetic
Figure 9. Maximum switching efficiency (SEmax) and optimal ligand
concentration ([L]opt) of 2e as a function of the concentration of
NiTPPF20 (concentration range: 0−4 mM). In this region, SEmax and
Lopt are independent from the concentration of NiTPPF20. The gray bar
indicates the concentration of 2e used in our experiments (67.9 mM).
species in one switching state and 100% in the other state. SEmax
,
thus, would be 100%. Parameters that lower SEmax are incomplete
photochemical conversions between cis and trans, incomplete
E
DOI: 10.1021/acs.joc.5b02817
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Table 4. Important Photochemical and Magnetic Switching Properties of Three Azoimidazoles (2, 2b, and 2e) and the to Date
Most Efficient Azopyridine PDL 610 at Photostationary States PSS-365 and PSS-vis
a
f
% cis isomer
% para. Ni2+
b
c
c
d
d
e
f
PSS-365 PSS-vis t1/2 [h] K1S‑trans [L mol−1
]
K1S‑cis [L mol−1
]
[L]opt [mM] SEmax [%] [L]exptl [mM] PSS365 nm PSSvis SEexptl [%]
g
2
98
95
98
82
67
528
37.5
38.3
59.1
60.0
10.8
15.8
1.37
0.30
35.1
21.9
64.9
91.9
16
26
70
50
34.9
19.4
67.9
69.2
43
36
19
21
64
66
89
68
21
30
70
48
h
2b
2e
6
22
49
h
g
158
262
10
9.99
a
Amount of cis isomer in both photostationary states (PSS-365 and PSS-vis), 298 K, toluene-d8, in the presence of NiTPPF20 (1.94 mM for 2, 2b,
b
c
and 2e, 0.106 mM for 6). Thermal half-life of cis isomer, 298 K, toluene-d8. Association constants (K1S) for cis and trans isomers determined from
1H NMR titration experiments (for a full list of association constants, see the Supporting Information), 298 K for 2, 2b, and 6, 300 K for 2e, toluene-
d8, NiTPPF20 (1.94 mM for 2, 2b, and 2e, 0.106 mM for 6). Theoretically calculated optimal ligand concentrations ([L]opt) and maximum
d
e
switching efficiencies (SEmax) on the basis of experimentally determined association constants. Ligand concentration used in the switching
f
experiment. Ratios of paramagnetic Ni2+ (% para. Ni2+) in both photostationary states; the experimental switching efficiencies (SEexptl) are
g
h
determined on the basis of the porphyrin pyrrole shifts. Irradiation with 455 nm. Irradiation with 525 nm. Irradiation times were 20 min.
concentrations ([L]opt). The experimental maximum switching
efficiencies (SEexptl) under these conditions are in good
agreement with the theoretically predicted values (SEmax),
supporting our mathematical model. Table 4 summarizes
important properties of 2, 2b, and 2e, including the parameters
of azopyridine 6, which has been the most efficient system so
far.10 Figure 10 shows a comparison of 6 with compound 2e. The
light-induced switching process is fully reversible without any
sign of fatigue for both ligands over a large number of cycles.
To gain further insight into the compositions and concen-
trations of the complexes that are actually present in the
solutions, we calculated the concentrations of the five possible
complex species as a function of the ligand concentration
(speciation plots, Figure 11). At the experimentally used ligand
concentration ([2e] = [L]exptl ≅ [L]opt), the prevalent complex at
PSS-365 is the bare NiTPPF20 (81%, Figure 11a), and at PSS-
525, the predominant species is the six-coordinate complex of
NiTPPF20 with two axial trans-2e ligands (55%, Figure 11b).
CONCLUSIONS
■
We present the most efficient photodissociable ligand (PDL) to
date for light-driven, coordination-induced spin state switching
(LD-CISSS). The coordination properties of imidazole were
combined with the photochromic properties of azo compounds
in a 5-phenylazoimidazole. The starting point of our design was
the already published parent system,30 which however exhibits an
unsatisfying switching efficiency. We extensively used quantum
chemical calculations to improve the properties by suitable
substitution. The most promising candidates were synthesized
and investigated. Azoimidazole 2e turned out to be superior to
previously used azopyridines (e.g., 6)10,12 in most aspects. The
coordination power of the binding isomer is substantially higher,
and smaller concentrations of the ligand are necessary for
optimal switching. The photochemical conversion to the non-
binding cis form is almost quantitative (98% as compared to 82%
in azopyridine). Consequently, the switching efficiency was
improved to 70% (as compared to 48% in azopyridine).
Azoimidazoles are promising candidates to expand the
coordination-induced spin state switching concept to other
transition metal ions, such as Fe(II), Fe(III), and Mn(III). They
are also suitable to be covalently tethered to the porphyrin core, a
Figure 10. Reversible spin state switching of NiTPPF20 with the PDL 2e
(azoimidazole, black curve) and 610 (azopyridine, green curve). The
proportion of paramagnetic nickel species is plotted as a function of the
switching cycles. Switching was performed by irradiation with 365 nm
(trans → cis) and 525 nm (2e) or 455 nm (6) (cis → trans). The
azoimidazole ([2e] = 67.9 mM, [NiTPPF20] = 1.94 mM, toluene-d8)
exhibits a switching efficiency (SE) of 70%, whereas the azopyridine ([6]
= 69.2 mM, [NiTPPF20] = 0.106 mM, toluene-d8) is less efficient with
SE = 48%. The predominant species at both PSSs are shown as
simplified structures.
strategy that has been successfully used with azopyridines.11,13
This so-called “record player” design has already been used to
switch MRI contrasts.21 Contrast agents based on the CISSS
concept have a high potential for functional imaging in MRI and
interventional radiology (minimal invasive, catheter-based
F
DOI: 10.1021/acs.joc.5b02817
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
layers were treated with TBAF trihydrate (1.5 equiv) and a spatula tip of
cesium fluoride and were stirred at room temperature overnight. Diluted
aq NaHCO3 was added; layers were separated, and the aqueous layer
was extracted with DCM (3×). The combined organic layers were dried
over MgSO4 and evaporated. Purification was achieved via column
chromatography on silica gel.
General Synthetic Procedure C. Synthesis of 1-trityl-4-phenyl-
azoimidazoles.30 The sulfamoyl-protected azo-imidazole was dissolved
in EtOH/HCl (4:1) (25 mL/mmol), and the mixture was stirred at 50
°C for 3 h. After cooling to 0 °C (ice bath), the mixture was basified with
40% KOH (aq) (pH ∼10). The solution was extracted with CHCl3 (3 ×
25 mL); the combined organic layers were dried over MgSO4 and
evaporated to dryness. The residue was taken up in DCM (15 mL/
mmol), treated with trityl chloride (1.1 equiv) and triethylamine (1.3
equiv), and stirred at room temperature overnight. The organic layer
was washed with aq NaHCO3 (2×), and the combined aq layers were
extracted with DCM (2×). The combined organic layers were dried over
MgSO4, and the solvent was removed. Purification was achieved via
column chromatography on silica gel.
General Synthetic Procedure D. Synthesis of 1-methyl-5-phenyl-
azoimidazoles.30 1-Trityl-5-phenyl-azoimidazole was dissolved under
nitrogen atmosphere in dry DCM (7.5 mL/mmol), and methyl triflate
(1.2 equiv) was added via a syringe at room temperature. The reaction
mixture was stirred at room temperature overnight; acetone/water (1:1,
15 mL/mmol) was added, and stirring was continued for an additional 4
h. Then, saturated NaHCO3 was added, and the layers were separated.
The aq layer was extracted with DCM (3×); the combined organic
layers were dried over MgSO4, and the solvent was removed.
Purification was achieved via column chromatography on silica gel.
Synthesis of 1-Methyl-5-(phenylazo)imidazole (2). 1-Methyl-
5-(phenylazo)imidazole (2) was synthesized and characterized as
previously reported.30
Figure 11. (a) Speciation plot of the six species in a solution of
NiTPPF20 (1.94 mM) in the presence of the cis/trans mixture of the
ligand 2e at PSS-365 (98% cis, 2% trans) and (b) speciation plot of the
same solution at PSS-525 (48% cis, 52% trans). Both are given as a
function of the total concentration of 2e ([2e]0 = [trans-2e] + [cis-2e]).
The proportion of the five complexes and free porphyrin at the ligand
concentration used in the experiment [2e]0 = [L]exptl = 67.9 mM, are
given on the right. Note that only the free porphyrin is diamagnetic. All
other complexes are paramagnetic.
Synthesis of 1-Methyl-5-(4′-fluorophenylazo)imidazole (2a). Step
1: 4-Fluorodiazoniumbenzene tetrafluoroborate. 4-Fluoroaniline (4.30
g, 45.0 mmol) was converted to the desired diazonium compound
following general procedure A. The product was obtained as a white
solid (7.31, 34.8 mmol, 77%). 1H NMR (200 MHz, 300 K, CD3CN): δ
8.72−8.50 (m, 2 H, 2,6-H), 7.78−7.55 (m, 2 H, 3,5-H) ppm. Step 2: 1-
Dimethylsulfamoyl-5-(4′-fluorophenylazo)imidazole. Starting from 1-
N,N-dimethyl-sulfamoylimidazole (2.00 g, 11.4 mmol), general
procedure B was applied. Purification via column chromatography on
silica gel (CH2Cl2/EtOAc (9:1), Rf = 0.51) gave the desired product as
an orange-yellow solid (2.01 g, 6.75 mmol, 59%). Mp: 98 °C. FT-IR
(layer) ν: 3142 (w), 2931 (w), 1592 (m), 1518 (w), 1501 (m), 1457
(m), 1388 (m), 1336 (m), 1226 (m), 1168 (s), 1143 (m), 1115 (m),
1091 (s), 1055 (m), 969 (s), 850 (s), 820 (s), 748 (m), 723 (s), 649 (m),
surgery).32−34 Azoimidazoles could improve and expand
applications toward this end.
EXPERIMENTAL SECTION
■
General Experimental Methods. THF was dried and distilled
from sodium/benzophenone. Dichloromethane was dried and distilled
from CaH2. N,N-Dimethylsulfamoylimidazole was synthesized analo-
gously to the procedure of Chadwick et al.35
General Synthetic Procedure A. Synthesis of benzenediazonium
tetrafluoroborates. If not stated otherwise, diazonium salts were
prepared analogously to the procedure of Dunker et al.36 Aniline
(10.0 mmol) was suspended in tetrafluoroboric acid (20.0 mL, 50 wt %
in water), and water was added until a clear solution was obtained. It was
cooled to 0 °C, and sodium nitrite (10.0 mmol in 1.50 mL of water) was
added dropwise. The precipitate was filtered off and washed
subsequently with water, ethanol, and diethyl ether before being dried
in vacuo. Because of the low thermal stability, we solely performed 1H
NMR spectroscopy on the diazonium salts and used them in the
following step immediately after preparation.
General Synthetic Procedure B. Synthesis of 1-N,N-dimethylsulfa-
moyl-5-phenylazoimidazoles.30 1-N,N-Dimethylsulfamoylimidazole
was dissolved under a nitrogen atmosphere in dry THF (6 mL per
mmol) and cooled to −78 °C. Then, n-BuLi (2.5 M in hexane, 1.0 equiv)
was added slowly over 15 min, and the solution was stirred at −78 °C for
30 min. Dimethylthexylchlorosilane (1.0 equiv) was added; the cooling
bath was removed, and stirring was continued at room temperature
overnight. The solution was again cooled to −78 °C; n-BuLi (2.5 M in
hexane, 1.1 equiv) was added slowly over 10 min, and the mixture was
stirred at −78 °C for 30 min. Then, diazonium tetrafluoroborate (1.0
equiv) was added, and the reaction mixture immediately turned from
light yellow to deep red. The cooling bath was removed, and the mixture
was stirred at room temperature overnight. The reaction mixture was
quenched with diluted aq NaHCO3, layers were separated, and the
aqueous layer was extracted once with THF. The combined organic
1
629 (m), 584 (s), 518 (s) cm−1. H NMR (500 MHz, 300 K, CDCl3,
TMS): δ 8.13 (s, 1 H, 2-H), 7.88−7.83 (m, 2 H, 2′,6′-H), 7.50 (s, 1 H, 4-
H), 7.22−7.17 (m, 2 H, 3′,5′-H), 2.97 (s, 6 H, -N(CH3)2) ppm. 13C
NMR (125 MHz, 300 K, CDCl3, TMS): δ 164.8 (C-4′), 149.6 (C-1′),
145.1 (C-5), 141.1 (C-2), 125.2 (C-2′,6′), 119.0 (C-4), 116.6 (C-3′,5′),
38.5 (-N(CH3)2) ppm. 19F-NMR (470 MHz, 300 K, CDCl3): δ −107.52
(s, 4′-F) ppm. MS (EI, 70 eV) m/z (%): 297 (42) [M]+, 190 (44) [M −
DMS]+, 134 (100), 108 (40) [SO2NMe2]+. MS (CI, isobutane) m/z
(%): 298 (100) [M + H]+, 123 (11). UV−vis (toluene) λmax (lg ε): 357
nm (4.214). Anal. Calcd for C11H12N5O2FS: C, 44.44; H, 4.07; N, 23.56;
S, 10.79. Found: C, 44.79; H, 4.04; N, 23.42; S, 10.71. Step 3: 1-Trityl-4-
(4′-fluorophenylazo)imidazole. Starting from 1-dimethylsulfamoyl-5-
(4′-fluoro-phenylazo)imidazole (1.00 g, 3.36 mmol), general procedure
C was applied. Purification via column chromatography on silica gel
(CH2Cl2/EtOAc (1:1), Rf = 0.80) gave the desired product as a yellow
solid (884 mg, 2.04 mmol, 61%). Mp: 234 °C. FT-IR (layer) ν: 3059
(w), 1593 (m), 1492 (m), 1444 (m), 1322 (w), 1299 (m), 1222 (m),
1124 (m), 1090 (w), 1038 (w), 1001 (w), 846 (m), 749 (s), 702 (s), 658
(s), 639 (m), 528 (s) cm−1. 1H NMR (500 MHz, 300 K, CDCl3, TMS):
δ 7.92−7.88 (m, 2 H, 2′,6′-H), 7.54 (s, 1 H, 5-H), 7.52 (s, 1 H, 2-H),
7.39−7.36 (m, 9 H, m-Tr-H, p-Tr-H), 7.22−7.18 (m, 6 H, o-Tr-H),
7.16−7.10 (m, 2 H, 3′,5′-H) ppm. 13C NMR (125 MHz, 300 K, CDCl3,
TMS): δ 164.1 (C-4′), 153.2 (C-4), 149.7 (C-1′), 141.9 (-CPh3), 139.5
(C-2), 129.9 (C-o-Tr), 128.6 (C-m-Tr), 128.5 (C-p-Tr), 124.7 (C-
G
DOI: 10.1021/acs.joc.5b02817
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
2′,6′), 122.5 (C-5), 116.0 (C-3′,5′) 76.4 (C-i-Tr) ppm. 19F-NMR (470
MHz, 300 K, CDCl3): δ −110.40 (s, 4′-F) ppm. MS (EI, 70 eV) m/z
(%): 432 (1) [M]+, 243 (100) [Tr]+, 165 (42) [Tr-Ph]+. MS (CI,
isobutane) m/z (%): 433 (1) [M + H]+, 243 (100) [Tr]+, 191 (20) [M
− Tr + H]+, 167 (31) [Tr − Ph + 2H]+. UV−vis (toluene) λmax (lg ε):
336 nm (4.246). Anal. Calcd for C28H21N4F: C, 77.76; H, 4.89; N, 12.95.
Found: C, 77.80; H, 5.21; N, 12.96. Step 4: 1-Methyl-5-(4′-
fluorophenylazo)imidazole (2a). Starting from 1-trityl-4-(4′-fluorophe-
nylazo)-imidazole (400 mg, 925 μmol), general procedure D was
applied. Purification via column chromatography on silica gel (EtOAc, Rf
= 0.30) gave the desired product 2a as a yellow solid (150 mg, 735 μmol,
79%). Mp: 113 °C. FT-IR (layer) ν: 3089 (w), 1588 (m), 1496 (m),
1429 (m), 1402 (w), 1345 (m), 1284 (m), 1232 (s), 1207 (s), 1170 (m),
1119 (s), 1096 (m), 1034 (m), 894 (w), 842 (s), 750 (m), 717 (m), 682
(m), 663 (s), 646 (s), 552 (m), 514 (s) cm−1. 1H NMR (500 MHz, 300
K, CDCl3, TMS): δ 7.87−7.80 (m, 2 H, 2′,6′-H), 7.69 (s, 1 H, 2-H), 7.57
(s, 1 H, 4-H), 7.20−7.12 (m, 2 H, 3′,5′-H), 3.96 (s, 3 H, N-CH3) ppm.
13C NMR (125 MHz, 300 K, CDCl3, TMS): δ 164.3 (C-4′), 149.7 (C-
1′), 145.2 (C-5), 140.4 (C-2), 124.4 (C-2′,6′), 122.8 (C-4), 116.2 (C-
3′,5′), 32.7 (N-CH3) ppm. 19F-NMR (470 MHz, 300 K, CDCl3): δ
−109.48 (s, br, 4′-F) ppm. MS (EI, 70 eV) m/z (%): 204 (100) [M]+,
124 (15), 109 (39). MS (CI, isobutane) m/z (%): 205 (100) [M + H]+.
HR-MS (EI, 70 eV): m/z [M]+ calcd for C10H9N4F, 204.0811; found
204.0811. UV−vis (toluene) λmax (lg ε): 362 nm (4.230).
Synthesis of 1-Methyl-5-(4′-methoxyphenylazo)imidazole (2b).
Step 1: 4-Methoxydiazoniumbenzene tetrafluoroborate. 4-Anisidine
(5.00 g, 40.6 mmol) was converted to the desired diazonium compound
following general procedure A. The product was obtained as light purple
solid (6.70, 30.2 mmol, 74%). 1H NMR (200 MHz, 300 K, CD3CN): δ
8.43 (s, 2 H, 2,6-H), 7.36 (s, 2 H, 3,5-H), 4.06 (s, 3 H, -OCH3) ppm.
Step 2: 1-Dimethylsulfamoyl-5-(4′-methoxyphenylazo)imidazole. Start-
ing from 1-N,N-dimethyl-sulfamoylimidazole (2.00 g, 11.4 mmol),
general procedure B was applied. Purification via column chromatog-
raphy on silica gel (CH2Cl2/EtOAc (9:1), Rf = 0.49) gave the desired
product as an orange-red solid (2.35 g, 7.60 mmol, 67%). Mp: 125 °C.
FT-IR (layer) ν: 3134 (w), 2943 (w), 2838 (w), 1602 (m), 1582 (m),
1419 (m), 1454 (m), 1385 (s), 1251 (s), 1166 (s), 1138 (s), 1090 (s),
1023 (s), 972 (s), 890 (m), 841 (m), 808 (m), 723 (s), 651 (m), 634
(m), 584 (s), 551 (s), 529 (s) cm−1. 1H NMR (500 MHz, 300 K, CDCl3,
TMS): δ 8.09 (d, 4J = 0.8 Hz, 1 H, 2-H), 7.85−7.82 (m, 2 H, 2′,6′-H),
7.45 (d, 4J = 0.8 Hz, 1 H, 4-H), 7.02−6.98 (m, 2 H, 3′,5′-H), 3.90 (s, 3 H,
-OCH3), 2.97 (s, 6 H, -N(CH3)2) ppm. 13C NMR (125 MHz, 300 K,
CDCl3, TMS): δ 162.7 (C-4′), 147.3 (C-1′), 145.1 (C-5), 140.3 (C-2),
125.0 (C-2′,6′), 118.1 (C-4), 114.5 (C-3′,5′), 55.7 (-OCH3), 38.4
(-N(CH3)2) ppm. MS (EI, 70 eV) m/z (%): 309 (72) [M]+, 202 (27)
[M − DMS + H]+. MS (CI, isobutane) m/z (%): 310 (100) [M + H]+.
UV−vis (toluene) λmax (lg ε): 371 nm (4.314). Step 3: 1-Trityl-4-(4′-
methoxyphenylazo)imidazole. Starting from 1-dimethylsulfamoyl-5-
(4′-meth-oxyphenylazo)imidazole (1.00 g, 3.23 mmol), general
procedure C was applied. Purification via column chromatography on
silica gel (EtOAc, Rf = 0.74) gave the desired product as a yellow solid
(1.30 g, 2.92 mmol, 90%). Mp: 190 °C. FT-IR (layer) ν: 3057 (w), 2838
(w), 1597 (m), 1578 (m), 1491 (m), 1442 (m), 1298 (w), 1244 (s),
1184 (m), 1143 (m), 1119 (m), 1029 (m), 991 (m), 839 (s), 808 (m),
747 (s), 700 (s), 655 (s), 536 (m) cm−1. 1H NMR (500 MHz, 300 K,
CDCl3, TMS): δ 7.91−7.86 (m, 2 H, 2′,6′-H), 7.50 (d, 4J = 1.4 Hz, 1 H,
5-H), 7.49 (d, 4J = 1.3 Hz, 1 H, 2-H), 7.39−7.36 (m, 9 H, m-Tr-H, p-Tr-
H), 7.22−7.18 (m, 6 H, o-Tr-H), 6.98−6.94 (m, 2 H, 3′,5′-H), 3.86 (s, 3
H, -OCH3) ppm. 13C NMR (125 MHz, 300 K, CDCl3, TMS): δ 161.5
(C-1′), 153.4 (C-4), 147.4 (C-3′,5′), 141.9 (C-i-Tr), 139.1 (C-2), 129.8
(C-o-Tr), 128.3 (C- p-Tr), 128.3 (C-m-Tr), 124.4 (C-2′,6′), 121.0 (C-
5), 114.1 (C-3′,5′), 76.1 (-CPh3), 55.5 (-OCH3) ppm. MS (EI, 70 eV)
m/z (%): 444 (1) [M]+, 243 (100) [Tr]+, 202 (15) [M − Tr + H]+, 165
(42) [Tr − Ph]+. MS (CI, isobutane) m/z (%): 445 (1) [M + H]+, 243
(100) [Tr]+, 203 (20) [M − Tr + 2H]+. UV−vis (toluene) λmax (lg ε):
360 nm (4.349). Anal. Calcd for C29H24N4O: C, 78.36; H, 5.44; N,
12.60. Found: C, 78.15; H, 5.77; N, 12.36. Step 4: 1-Methyl-5-(4′-
methoxyphenylazo)imidazole (2b). Starting from 1-trityl-4-(4′-methox-
yphenyl-azo)imidazole (400 mg, 900 μmol), general procedure D was
applied. Purification via column chromatography on silica gel (EtOAc, Rf
= 0.32) gave desired product 2b as an orange-yellow solid (179 mg, 828
μmol, 92%). Mp: 103 °C. FT-IR (layer) ν: 2916 (w), 1599 (m), 1581
(m), 1525 (w), 1495 (m), 1444 (m), 1343 (m), 1240 (s), 1143 (s), 1107
(m), 909 (m), 844 (s), 812 (s), 758 (m), 699 (m), 640 (s), 559 (s), 524
(s) cm−1. 1H NMR (500 MHz, 300 K, CDCl3, TMS): δ 7.85−7.80 (m, 2
H, 2′,6′-H), 7.77 (s, 1 H, 2-H), 7.52 (s, 1 H, 4-H), 7.01−6.97 (m, 2 H,
3′,5′-H), 3.98 (s, 3 H, N-C H3), 3.89 (s, 3 H, -OCH3) ppm. 13C NMR
(125 MHz, 300 K, CDCl3, TMS): δ 162.2 (C-1′), 147.5 (C-4′), 145.4
(C-5), 139.5 (C-2), 124.5 (C-2′,6′), 120.5 (C-4), 114.5 (C-3′,5′), 55.8
(-OCH3), 32.6 (N-CH3) ppm. MS (EI, 70 eV) m/z (%): 216 (100)
[M]+, 122 (15), 109 (16), 107 (60) [C7H7O]+. MS (CI, isobutane) m/z
(%): 217 (100) [M + H]+. HR-MS (EI, 70 eV): m/z [M]+ calcd for
C11H12N4O, 216.1011; found, 216.1020. UV−vis (toluene) λmax (lg ε):
372 nm (4.284).
Synthesis of 1-Methyl-5-(4′-N,N-dimethylaminophenylazo)-
imidazole (2c). Step 1: 4-N,N-Dimethylaminodiazoniumbenzene
tetrafluoroborate. 4-N,N-Dimethylaminoaniline (3.70 g, 27.2 mmol)
was converted into the desired diazonium compound following general
procedure A. The product was obtained as a white solid (3.26 g, 13.9
mmol, 51%). 1H NMR (200 MHz, 300 K, CD3CN): δ 8.08−7.94 (m, 2
H, 2,6-H), 7.00−6.89 (m, 2 H, 3,5-H), 3.27 (s, 6 H, -N(CH3)2) ppm.
Step 2: 1-Dimethylsulfamoyl-5-(4′-N,N-dimethylaminophenylazo)-
imidazole. Starting from 1-N,N-dimethylsulfamoylimidazole (1.50 g,
8.57 mmol), general procedure B was applied. Purification via column
chromatography on silica gel (CH2Cl2/EtOAc (9:1), Rf = 0.34) gave the
desired product as a red solid (1.30 g, 4.03 mmol, 47%). Mp: 157 °C.
FT-IR (layer) ν: 3116 (w), 2901 (w), 1598 (m), 1514 (m), 1460 (w),
1432 (m), 1417 (m), 1388 (m), 1353 (m), 1272 (m), 1231 (m), 1170
(m), 1087 (s), 959 (s), 886 (m), 844 (m), 823 (s), 725 (s), 702 (m), 631
(m), 594 (s), 524 (s) cm−1. 1H NMR (600 MHz, 300 K, CDCl3, TMS):
δ 8.04 (d, 4J = 0.7 Hz, 1 H, 2-H), 7.80−7.77 (m, 2 H, 2′,6′-H), 7.36 (d, 4J
= 0.7 Hz, 1 H, 4-H), 6.75−6.71 (m, 2 H, 3′,5′-H), 3.11 (s, 6 H, C4′-
N(CH3)2), 2.96 (s, 6 H, -SO2N(CH3)2) ppm. 13C NMR (125 MHz, 300
K, CDCl3, TMS): δ 152.8 (C-1′), 145.6 (C-5), 144.0 (C-4′), 139.2 (C-
2), 125.4 (C-2′,6′), 116.9 (C-4), 111.6 (C-3′,5′), 40.3 (C4′-N(CH3)2),
38.4 (-SO2N(CH3)2) ppm. MS (EI, 70 eV) m/z (%): 322 (100) [M]+,
214 (20) [M-SO2NMe2]+. MS (CI, isobutane) m/z (%): 323 (100) [M
+ H]+, 216 (13) [M − SO2NMe2 + H]+. UV−vis (toluene) λmax (lg ε):
455 nm (4.395). Anal. Calcd for C13H18N6O2S: C, 48.43; H, 5.63; N,
26.07; S, 9.95. Found: C, 48.85; H, 5.46; N, 25.79; S, 10.22. Step 3: 4(5)-
(4′-N,N-Dimethylaminophenylazo)imidazole. 1-Dimethylsulfamoyl-5-
(4′-N,N-dimethylaminophenylazo)imidazole (100 mg, 310 μmol) was
dissolved in THF/HCl (4:1) (16 mL), and the mixture was heated to
reflux for 1 h. The mixture was basified with 40% KOH (aq) (pH ∼10)
and extracted with CH2Cl2 (3 × 50 mL). The combined organic layers
were dried over MgSO4 and evaporated to dryness. Purification via
column chromatography on silica gel (EtOAc/acetone, 1:1) gave a red
solid (33.0 mg, 153 μmol, 49%). 1H NMR (500 MHz, 300 K, CDCl3): δ
7.81−7.77 (m, 2 H, 2′,6′-H), 7.67 (d, 4J = 0.8 Hz, 1 H, 5(4)-H), 7.64 (d,
4J = 0.8 Hz, 1 H, 2-H), 6.75−6.71 (m, 2 H, 3′,5′-H), 3.06 (s, 6 H,
-N(CH3)2) ppm. MS (EI, 70 eV) m/z (%): 215 (100) [M]+, 149 (18)
[M − Im + H]+, 120 (19) [C8H10N]+. MS (CI, isobutane) m/z (%): 216
(10) [M + H]+. Step 4: 1-Trityl-4-(4′-N,N-dimethylaminophenylazo)-
imidazole. 4(5)-(4′-N,N-Dimethylaminophenylazo)imidazole (186 mg,
767 μmol) was dissolved in CH2Cl2 (10 mL). The solution was treated
with trityl chloride (235 mg, 844 μmol) and triethylamine (140 μL, 997
μmol) and stirred at room temperature overnight. The organic layer was
washed twice with water, and the combined aq layers were extracted
with DCM (2 × 20 mL). The combined organic layers were dried over
MgSO4, and the solvent was removed. Purification via column
chromatography on silica gel (CH2Cl2/EtOAc (1:1), Rf = 0.67) gave
the desired product as a yellow solid (230 mg, 503 μmol, 66%). Mp: 263
°C. FT-IR (layer) ν: 2906 (w), 1595 (s), 1511 (m), 1488 (m), 1442 (m),
1406 (w), 1361 (s), 1224 (m), 1147 (s), 1113 (m), 1035 (w), 986 (m),
942 (m), 865 (w), 823 (s), 751 (s), 704 (s), 655 (s), 526 (s) cm−1. 1H
NMR (600 MHz, 300 K, CDCl3, TMS): δ 7.86−7.84 (m, 2 H, 2′,6′-H),
7.46 (d, 4J = 1.4 Hz, 1 H, 2-H), 7.41 (d, 4J = 1.5 Hz, 1 H, 5-H), 7.37−7.34
(m, 9 H, m-Tr-H, p-Tr-H), 7.22−7.19 (m, 6 H, o-Tr-H), 6.73−6.70 (m,
2 H, 3′,5′-H), 3.03 (s, 6 H, -N(CH3)2) ppm. 13C NMR (150 MHz, 300
K, CDCl3, TMS): δ 153.6 (C-4), 152.0 (C-4′), 143.9 (C-1′), 142.0 (C-i-
H
DOI: 10.1021/acs.joc.5b02817
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Tr), 138.8 (C-2), 129.8 (C-o-Tr), 128.2 (C-m-Tr), 127.9 (C-p-Tr),
124.6 (C-2′,6′), 119.2 (C-5), 111.5 (C-3′,5′), 76.0 (-CPh3), 40.3
(-N(CH3)2) ppm. MS (EI, 70 eV) m/z (%): 457 (5) [M]+, 243 (100)
[Tr]+, 215 (10) [M-Tr]+, 165 (41) [Tr-Ph]+. MS (CI, isobutane) m/z
(%): 458 (1) [M + H]+, 243 (100) [Tr]+, 216 (10) [M − Tr + 2H]+, 167
(18) [Tr − Ph + 2H]+. UV−vis (toluene) λmax (lg ε): 402 nm (4.146).
Step 5: 1-Methyl-5-(4′-N,N-dimethylaminophenylazo)imidazole (2c).
Starting from 1-trityl-4-(4′-N,N-dimethylaminophenylazo)-imidazole
(220 mg, 481 μmol), general procedure D was applied. Purification
via column chromatography on silica gel (acetone, Rf = 0.52) gave
desired product 2c as a red solid (60.0 mg, 262 μmol, 54%). Mp: 126 °C.
FT-IR (layer) ν: 3120 (w), 2917 (m), 2820 (w), 2652 (w), 1600 (s),
1557 (m), 1506 (m), 1445 (m), 1367 (s), 1311 (m), 1278 (m), 1231
(m), 1144 (s), 1108 (s), 1069 (s), 942 (m), 892 (m), 821 (s), 796 (s),
711 (m), 669 (s), 642 (s), 529 (s) cm−1. 1H NMR (500 MHz, 300 K,
CDCl3, TMS): δ 7.80−7.76 (m, 2 H, 2′,6′-H), 7.54 (s, 1 H, 2-H), 7.42
(d, 4J = 0.9 Hz, 1 H, 4-H), 6.76−6.72 (m, 2 H, 3′,5′-H), 3.92 (s, 3 H,
-CH3), 3.08 (s, 6 H, C4′-N(CH3)2) ppm. 13C NMR (125 MHz, 300 K,
CDCl3, TMS): δ 152.3 (C-1′), 145.8 (C-5), 144.2 (C-4′), 139.0 (C-2),
124.5 (C-2′,6′), 120.1 (C-4), 111.7 (C-3′,5′), 40.5 (C4′-N(CH3)2), 32.2
(−CH3) ppm. MS (EI, 70 eV) m/z (%): 229 (100) [M]+, 105 (10)
[C6H5N2]+. MS (CI, isobutane) m/z (%): 230 (100) [M + H]+. HR-MS
(EI, 70 eV): m/z [M]+ calcd for C12H15N5, 229.1328; found, 229.1323.
UV−vis (toluene) λmax (lg ε): 440 nm (4.339).
added dropwise and stirred for 15 min at room temperature. Then, 50%
HBF4 (5.0 mL) and water (30.0 mL) were added. The precipitate was
filtered off and washed with little EtOH. The desired product was
obtained as a light purple solid (2.07 g, 9.57 mmol, 65%), which was
rather unstable and therefore was immediately used after preparation.
1H NMR (200 MHz, 300 K, CD3CN): δ 8.48 (s, 2 H, 2,6-H), 3.47 (s, 3
H, -OCH3), 1.53 (s, 18 H, 2x -C(CH3)3) ppm. Step 5: 1-
Dimethylsulfamoyl-5-(3′,5′-di-tert-butyl-4′-methoxyphenylazo)-
imidazole. Starting from 1-N,N-dimethylsulfamoylimidazole (1.00 g,
5.71 mmol), general procedure B was applied. Purification via column
chromatography on silica gel (CH2Cl2/EtOAc (19:1), Rf = 0.43) gave
the desired product as an orange-red solid (1.40 g, 3.32 mmol, 58%).
Mp: 114 °C. FT-IR (layer) ν: 3142 (w), 2956 (m), 1455 (m), 1387 (s),
1223 (m), 1162 (s), 1093 (s), 1003 (m), 972 (m), 889 (w), 834 (w), 750
(w), 726 (s), 638 (m), 591 (s), 537 (m), 513 (s) cm−1. 1H NMR (500
MHz, 300 K, CDCl3, TMS): δ 8.12 (s, 1 H, 2-H), 7.81 (s, 2 H, 2′,6′-H),
7.46 (s, 1 H, 4-H), 3.75 (s, 3 H, -OCH3), 3.01 (s, 6 H, -SO2N(CH3)2),
1.47 (s, 18 H, 2x -C(CH3)3) ppm. 13C NMR (125 MHz, 300 K, CDCl3,
TMS): δ 163.3 (C-4′), 148.2 (C-1′), 145.2 (C-3′,5′), 145.1 (C-5), 140.3
(C-2), 121.9 (C-2′,6′), 117.8 (C-4), 64.5 (-OCH3), 38.4 (-SO2N-
(CH3)2), 36.0 (2x -C(CH3)3), 31.9 (2x -C(CH3)3) ppm. MS (EI, 70 eV)
m/z (%): 421 (100) [M]+, 314 (30) [M − SO2NMe2 + H]+, 257 (35)
[M − SO2NMe2 − tBu + H]+, 234 (10). MS (CI, isobutane) m/z (%):
422 (100) [M + H]+. UV−vis (toluene) λmax (lg ε): 368 nm (4.204).
Step 6: 1-Trityl-4-(3′,5′-di-tert-butyl-4′-methoxyphenylazo)imidazole.
Starting from 1-dimethyl-sulfamoyl-5-(3′,5′-di-tert-butyl-4′-
methoxyphenylazo)imidazole (1.10 g, 2.61 mmol), general procedure
C was applied. Purification via column chromatography on silica gel
(CHCl3, Rf = 0.75) gave the desired product as an orange-yellow solid
(1.12 g, 2.01 mmol, 77%). Mp: 208 °C. FT-IR (layer) ν: 2956 (w), 1491
(w), 1444 (m), 1407 (w), 1282 (w), 1218 (m), 1157 (w), 1119 (m),
1009 (m), 893 (w), 859 (w), 754 (s), 700 (s), 660 (m), 639 (m), 504
(m) cm−1. 1H NMR (500 MHz, 300 K, CDCl3, TMS): δ 7.85 (s, 2 H,
2′,6′-H), 7.80 (d, 4J = 1.6 Hz, 1 H, 5-H), 7.52 (d, 4J = 1.4 Hz, 1 H, 2-H),
7.42−7.38 (m, 9 H, m-Tr-H, p-Tr-H), 7.21−7.17 (m, 6 H, o-Tr-H), 3.72
(s, 3 H, -OCH3), 1.44 (s, 18 H, 2x -C(CH3)3) ppm. 13C NMR (125
MHz, 300 K, CDCl3, TMS): δ 162.1 (C-4′), 153.3 (C-4), 148.4 (C-1′),
144.5 (C-3′,5′), 142.0 (C-i-Tr), 139.0 (C-2), 129.9 (C-o-Tr), 128.4 (C-
m-Tr), 128.0 (C-p-Tr), 121.6 (C-2′,6′), 121.3 (C-5), 76.3 (-CPh3), 64.5
(-OCH3), 36.2 (2x -C(CH3)3), 32.1 (2x -C(CH3)3) ppm. MS (EI, 70
eV) m/z (%): 260 (34), 243 (100) [Tr]+, 183 (100), 165 (31) [Tr-Ph]+,
105 (92). MS (CI, isobutane) m/z (%): 557 (4) [M + H]+, 315 (14) [M
− Tr + H]+, 243 (100) [Tr]+, 236 (17). UV−vis (toluene) λmax (lg ε):
344 nm (4.152). Step 7: 1-Methyl-5-(3′,5′-di-tert-butyl-4′-
methoxyphenylazo)imidazole (2e). Starting from 1-trityl-4-(3′,5′-di-
tert-butyl-4′-methoxyphenylazo)imidazole (800 mg, 1.44 mmol),
general procedure D was applied. Purification via column chromatog-
raphy on silica gel (EtOAc, Rf = 0.55) gave desired product 2e as an
orange-yellow solid (388 mg, 1.18 mmol, 82%). Mp: 107 °C. FT-IR
(layer) ν: 3102 (w), 2956 (m), 1503 (m), 1405 (m), 1340 (m), 1246
(w), 1219 (s), 1169 (m), 1114 (s), 996 (m), 893 (m), 809 (m), 643 (m),
Synthesis of 1-Methyl-5-(3′ , 5′-di-tert-butyl-4′-
methoxyphenylazo)imidazole (2e). Step 1: 3,5-Di-tert-butyl-4-me-
thoxybenzoic acid. 3,5-Di-tert-butyl-4-hydroxybenzoic acid (10.0 g, 39.9
mmol) was dissolved in acetone (200 mL) and treated with potassium
hydroxide (5.60 g, 100 mmol) and iodomethane (7.40 mL, 120 mmol).
The reaction mixture was heated to 60° and stirred under reflux for 18 h.
The mixture was partially evaporated and treated with ethyl acetate and
water. Layers were separated, and the aqueous layer was extracted three
times with ethyl acetate. The combined organic layers were dried over
magnesium sulfate and evaporated to dryness. The residue was dissolved
in THF/water (1:1, 100 mL) and treated with lithium hydroxide
monohydrate (5.00 g, 119 mmol). The reaction mixture was stirred at 60
°C for 18 h. The solution was partially evaporated and acidified with
concd HCl. The precipitate was filtered off and recrystallized from
CH2Cl2/n-hexane (1:1). The desired product was obtained as a
colorless solid (7.49 g, 26.7 mmol, 71%). 1H NMR (200 MHz, 300 K,
CDCl3, TMS): δ 8.04 (d, 4J = 0.4 Hz, 2 H, 2-H), 3.74 (s, 3 H, -OCH3),
1.47 (s, 18 H, 2x -C(CH3)3) ppm. Step 2: 3,5-Di-tert-butyl-4-
methoxyphenylcarbamic tert-butylester. 3,5-Di-tert-butyl-4-methoxy-
benzoic acid (3.00 g, 11.4 mmol) was treated with triethylamine (1.66
mL, 11.4 mmol) and DPPA (2.60 mL, 11.4 mmol) in tert-butanol (50.0
mL). The mixture was stirred at 85 °C overnight, and after cooling, the
precipitate was filtered off multiple times and washed with little EtOH to
yield 3,5-di-tert-butyl-4-methoxyphenylcarbamic tert-butylester as a
1
white solid (2.75 g, 8.2 mmol, 72%). H NMR (500 MHz, 300 K,
CDCl3, TMS): δ 7.21 (s, 2 H, 2,6-H), 6.32 (br s, 1 H, NH), 3.66 (s, 3 H,
-OCH3), 1.51 (s, 9 H, -OC(CH3)3), 1.42 (s, 18 H, 2x Ar-C(CH3)3) ppm.
13C NMR (125 MHz, 300 K, CDCl3, TMS): δ 223.7 (-N-CO-O-), 155.3
(C-4), 144.1 (C-3,5), 123.9 (C-1), 117.5 (C-2,6), 80.0 (O-C(CH3)3),
64.2 (-OCH3), 35.9 (2x Ar-C(CH3)3), 32.0 (2x Ar-C(CH3)3), 28.4 (O-
C(CH3)3) ppm. MS (EI, 70 eV) m/z (%): 335 (17) [M]+, 279 (100) [M
− tBu + H]+, 220 (21) [C15H24O]+. MS (CI, isobutane) m/z (%): 336
(18) [M + H]+, 280 (100) [M − tBu + 2H]+. Step 3:3,5-Di-tert-butyl-4-
methoxyaniline. 3,5-Di-tert-butyl-4-methoxyphenylcarbamic tert-buty-
lester (335 mg, 10.0 mmol) was dissolved in DCM (6.0 mL) and treated
with trifluoracetic acid (1.0 mL). After stirring overnight at room
temperature, the mixture was treated with sodium hydroxide (1.00 g) in
water (6.0 mL) and extracted with DCM (3 × 30 mL). The organic layer
was separated and dried over magnesium sulfate, and the solvent was
evaporated in vacuo to give 3,5-di-tert-butyl-4-methoxy-aniline as a
1
597 (m), 535 (m) cm−1. H NMR (500 MHz, 300 K, toluene-d8,
toluene-d8): δ 7.98 (s, 2 H, 2′,6′-H), 7.84 (s, br, 1 H, 4-H), 6.94 (s, br, 1
H, 2-H), 3.40 (s, 3 H, -OCH3), 3.09 (s, 3 H, N-CH3), 1.44 (s, 18 H, 2x
-C(CH3)3) ppm. 13C NMR (125 MHz, 300 K, toluene-d8, toluene-d8): δ
162.3 (C-4′), 149.3 (C-1′), 145.8 (C-5), 144.8 (C-3′,5′), 140.5 (C-2),
124.0 (C-4), 121.5 (C-2′,6′), 64.3 (-OCH3), 36.2 (2x -C(CH3)3), 32.1
(2x -C(CH3)3), 31.4 (N-CH3) ppm. MS (EI, 70 eV) m/z (%): 328 (74)
[M]+. MS (CI, isobutane) m/z (%): 329 (33) [M + H]+. HR-MS (EI, 70
eV): m/z [M]+ calcd for C19H28N4O, 328.2263; found, 328.2261. UV−
vis (toluene) λmax (lg ε): 369 nm (4.156). Anal. Calcd for C19H28N4O:
C, 69.48; H, 8.59; N, 17.06. Found: C, 69.38; H, 8.61; N, 16.88.
ASSOCIATED CONTENT
* Supporting Information
■
1
white solid (217 mg, 923 μmol, 92%). H NMR (200 MHz, 300 K,
S
CDCl3, TMS): δ 6.55 (s, 2 H, 2,6-H), 3.57 (s, 3 H, -OCH3), 3.41 (s, br, 2
H, NH2), 1.33 (s, 18 H, 2x -C(CH3)3) ppm. Step 4: 3,5-Di-tert-butyl-4-
methoxydiazoniumbenzene tetrafluoroborate. 3,5-Di-tert-butyl-4-me-
thoxyaniline (2.25 g, 9.57 mmol) was dissolved in EtOH (22.5 mL)
and 50% HBF4 (2.00 mL). Isopentyl nitrite (1.35 mL, 10.1 mmol) was
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.joc.5b02817.
Details of computational studies, 1H NMR, 13C NMR, and
UV−vis spectra for compounds 2a−c and 2e, details of 1H
I
DOI: 10.1021/acs.joc.5b02817
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
NMR titration experiments, and details of 1H NMR LD-
CISSS experiments (PDF)
(27) Tabata, M.; Nishimoto, J. Equilibrium data of porphyrins and
metalloporphyrins. Porphyrin Handb. 2000, 221−419.
(28) Walker, F. A.; Lo, M.-W.; Ree, M. T. J. Am. Chem. Soc. 1976, 98,
5552−5560.
AUTHOR INFORMATION
Corresponding Author
*E-mail: rherges@oc.uni-kiel.de.
■
(29) Portela, C. F.; Magde, D.; Traylor, T. G. Inorg. Chem. 1993, 32,
1313−1320.
(30) Wendler, T.; Schutt, C.; Nather, C.; Herges, R. J. Org. Chem. 2012,
̈
̈
77, 3284−3287.
Notes
(31) TURBOMOLE, V6.3, 2011; a development of University of
Karlsruhe and Forschungszentrum Karlsruhe GmbH, 1989−2007,
TURBOMOLE GmbH, since 2007; available from http://www.
turbomole.com.
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We gratefully acknowledge financial support by the Deutsche
Forschungsgemeinschaft via the collaborative research center
SFB 677 “Function by Switching”.
(32) Herges, R.; Jansen, O.; Tuczek, F.; Venkatamarani, S. Photo-
sensitive metal porphyrin complexes with pendant photoisomerizable
chelate arm as photochromic molecular switches undergoing photo-
induced spin transition. Patent DE 102010034496 A1, Feb 16, 2012.
(33) Herges, R.; Jansen, O.; Tuczek, F.; Venkatamarani, S. Transition
metal complexes with photosensitive tethered ligands as visible light-
induced spin-crossover magnetic molecular switches. Patent WO
2012022299 A1, Feb 23, 2012.
REFERENCES
■
(1) Szymanski, W.; Beierle, J. M.; Kistemaker, H. A. V.; Velema, W. A.;
Feringa, B. L. Chem. Rev. 2013, 113, 6114−6178.
(2) Hilf, R. J. C.; Bertozzi, C.; Zimmermann, I.; Reiter, A.; Trauner, D.;
Dutzler, R. Nat. Struct. Mol. Biol. 2010, 17, 1330−1336.
(3) Polosukhina, A.; Litt, J.; Tochitsky, I.; Nemargut, J.; Sychev, Y.; De
Kouchkovsky, I.; Huang, T.; Borges, K.; Trauner, D.; Van Gelder, R.;
Kramer, R. Neuron 2012, 75, 271−282.
(34) Dommaschk, M.; Gutzeit, F.; Boretius, S.; Haag, R.; Herges, R.
Chem. Commun. 2014, 50, 12476−12478.
(35) Chadwick, D. J.; Ngochindo, R. I. J. Chem. Soc., Perkin Trans. 1
1984, 481−486.
(36) Dunker, M. F. W.; Starkey, E. B.; Jenkins, G. L. J. Am. Chem. Soc.
1936, 58, 2308−2309.
(4) Kahn, O. Molecular Magnetism; VCH, 1993; p 380.
(5) Decurtins, S.; Gutlich, P.; Kohler, C. P.; Spiering, H.; Hauser, A.
̈
̈
Chem. Phys. Lett. 1984, 105, 1−4.
(6) Gutlich, P.; Garcia, Y.; Goodwin, H. A. Chem. Soc. Rev. 2000, 29,
̈
419−427.
(7) Gutlich, P.; Goodwin, H. A. Spin Crossover in Transition Metal
̈
Compounds I; Springer, 2004.
(8) Gopakumar, T. G.; Matino, F.; Naggert, H.; Bannwarth, A.;
Tuczek, F.; Berndt, R. Angew. Chem. 2012, 124, 6367−6371.
(9) Gopakumar, T. G.; Matino, F.; Naggert, H.; Bannwarth, A.;
Tuczek, F.; Berndt, R. Angew. Chem., Int. Ed. 2012, 51, 6262−6266.
(10) Thies, S.; Sell, H.; Schutt, C.; Bornholdt, C.; Nather, C.; Tuczek,
̈
̈
F.; Herges, R. J. Am. Chem. Soc. 2011, 133, 16243−16250.
(11) Venkataramani, S.; Jana, U.; Dommaschk, M.; Sonnichsen, F. D.;
̈
Tuczek, F.; Herges, R. Science 2011, 331, 445−448.
(12) Thies, S.; Sell, H.; Bornholdt, C.; Schutt, C.; Kohler, F.; Tuczek,
F.; Herges, R. Chem. - Eur. J. 2012, 18, 16358−16368.
(13) Dommaschk, M.; Schutt, C.; Venkataramani, S.; Jana, U.; Nather,
̈
̈
̈
̈
C.; Sonnichsen, F. D.; Herges, R. Dalton Trans. 2014, 43, 17395−17405.
̈
(14) Caughey, W. S.; Deal, R. M.; McLees, B. D.; Alben, J. O. J. Am.
Chem. Soc. 1962, 84, 1735−1736.
(15) Caughey, W. S.; Fujimoto, W. Y.; Johnson, B. P. Biochemistry
1966, 5, 3830−3843.
(16) McLees, B. D.; Caughey, W. S. Biochemistry 1968, 7, 642−652.
(17) Cole, S. J.; Curthoys, G. C.; Magnusson, E. A.; Phillips, J. N. Inorg.
Chem. 1972, 11, 1024−1028.
(18) Kim, D.; Su, Y. O.; Spiro, T. G. Inorg. Chem. 1986, 25, 3988−3993.
(19) Song, Y.; Haddad, R. E.; Jia, S.-L.; Hok, S.; Olmstead, M. M.;
Nurco, D. J.; Schore, N. E.; Zhang, J.; Ma, J.-G.; Smith, K. M.; Gazeau, S.;
Pec
Soc. 2005, 127, 1179−1192.
(20) Thies, S.; Bornholdt, C.; Kohler, F.; Sonnichsen, F. D.; Nather, C.;
́
aut, J.; Marchon, J.-C.; Medforth, C. J.; Shelnutt, J. A. J. Am. Chem.
̈
̈
̈
Tuczek, F.; Herges, R. Chem. - Eur. J. 2010, 16, 10074−10083.
(21) Dommaschk, M.; Peters, M.; Gutzeit, F.; Schutt, C.; Nather, C.;
Sonnichsen, F. D.; Tiwari, S.; Riedel, C.; Boretius, S.; Herges, R. J. Am.
̈
̈
̈
Chem. Soc. 2015, 137, 7552−7555.
(22) Otsuki, J.; Narutaki, K. Bull. Chem. Soc. Jpn. 2004, 77, 1537−1544.
(23) Otsuki, J.; Narutaki, K.; Bakke, J. M. Chem. Lett. 2004, 33, 356−
357.
(24) Suwa, K.; Otsuki, J.; Goto, K. Tetrahedron Lett. 2009, 50, 2106−
2108.
(25) Suwa, K.; Otsuki, J.; Goto, K. J. Phys. Chem. A 2010, 114, 884−
890.
(26) Weston, C. E.; Richardson, R. D.; Haycock, P. R.; White, A. J. P.;
Fuchter, M. J. J. Am. Chem. Soc. 2014, 136, 11878−11881.
J
DOI: 10.1021/acs.joc.5b02817
J. Org. Chem. XXXX, XXX, XXX−XXX