Photoinduced and thermal denitrogenation of bulky triazoline crystals: Insights into solid-to-solid transformation

Article abstract of DOI:10.1021/ja401577p

The photoinduced and thermal denitrogenation of crystalline triazolines with bulky substituents leads to the quantitative formation of aziridines in clean solid-to-solid reactions despite very large structural changes in the transition from reactant to pr

Full text of DOI:10.1021/ja401577p

Article
pubs.acs.org/JACS
Photoinduced and Thermal Denitrogenation of Bulky Triazoline
Crystals: Insights into Solid-to-Solid Transformation
Denisse de Loera, Antoine Stopin, and Miguel A. Garcia-Garibay*
Department of Chemistry and Biochemistry, University of California, Los Angeles, California 90095-1569, United States
S
* Supporting Information
ABSTRACT: The photoinduced and thermal denitrogenation
of crystalline triazolines with bulky substituents leads to the
quantitative formation of aziridines in clean solid-to-solid
reactions despite very large structural changes in the transition
from reactant to product. Analysis of the reaction progress by
powder X-ray diffraction, solid-state 13C CPMAS NMR, solid-
state FTIR spectroscopy, and thermal analysis has revealed
that solid-to-solid reactions proceed either through metastable
phases susceptible to amorphization or by mechanisms that
involve a reconstructive phase transition that culminates in the formation of the stable phase of the product. While the key for a
solid-to-solid transformation is that the reaction occurs below the eutectic temperature of the reactant and product two-
component system, experimental evidence suggests that those reactions will undergo a reconstructive phase transition when they
take place above the glass transition temperature.
crystal engineering. However, it is impossible to design
chemical reactions without acknowledging the importance of
energetics. With that in mind, over the past few years we have
suggested a general formulation of the topochemical postulate
based on an energetic viewpoint: “Reactions in crystals may be
accomplished with reactants that possess or transiently acquire
large amounts of potential energy.”11,12 Structures that possess
or acquire a high energy content may be capable of breaking
and making bonds without having to experience rotations,
collisions, and large-amplitude motions. At the same time, as
suggested by the Hammond postulate, species with high
potential energy tend to have early transition states and low
activation barriers. Reactant-like transition states are ideal from
a topochemical perspective because they require modest
structural changes early on in their reaction coordinates.
We posit that there are a large number of reactants and
reactions that meet these requirements.8 In a recent test of this
hypothesis, we began exploring the synthesis of ring-strained
aziridines by the photoinduced denitrogenation of crystalline
triazolines.13 To our satisfaction, we confirmed that dry
powdered crystals exposed to UV light with λ ≥ 290 nm
react without melting to give the expected aziridines 2 in
excellent yields within 30−90 min. The reactions proceeded
efficiently for several substituted triazolines (Scheme 1),
suggesting that the transformation may be general and
potentially useful for synthetic applications. It is known that
the reaction proceeds through the formation of a 1,3-biradical
by the extrusion of molecular nitrogen, and that the
recombination of this intermediate generates the corresponding
INTRODUCTION
■
Reports on reactions in organic crystals date back to 1834,
when Trommsdorff described the yellowing and shattering that
occurs when crystals of α-santonin are exposed to sunlight.1−3
In fact, many reactions in crystals were documented through
the 19th century.4 An intuitive perspective based on size and
shape confinement considerations, known as the topochemical
postulate, was developed by Kohlshutter in the early 1920s5 and
put on a solid footing with structural data from single-crystal X-
ray diffraction (XRD) in the 1960s by Schmidt and co-
workers.6,7 Until recently, the only guidelines to engineer
reactions in crystals were based on the topochemical postulate,
indicating that “reactions in crystals can only occur with a
minimum of atomic and molecular motion”. Accordingly, the
feasibility of a given reaction in a crystalline solid depends on
whether the geometric parameters that define the reactant(s)
are preorganized6 to form the product(s), and whether the
reaction cavity8 defined by the close-packing environment is
preserved as a function of reaction progress. Topotactic control
has been studied also in thermal9 and photochemical10
polymerization reactions which are governed by crystal packing.
Although both thermal and photochemical reactions can occur
under topochemical control, thermal activation requires
excitation of normal vibrational modes and lattice phonons
under thermal equilibrium, such that bond-breaking and bond-
making events in the crystal must compete with melting.
Conversely, photochemical activation generates localized
excited states, which are not in thermal equilibrium with the
rest of the lattice and are able to find reactive pathways before
their excess energy is dissipated through the environment.7
Under these guidelines, the planning and execution of
chemical reactivity in the solid state became a challenge in
Received: February 12, 2013
Published: April 2, 2013
© 2013 American Chemical Society
6626
dx.doi.org/10.1021/ja401577p | J. Am. Chem. Soc. 2013, 135, 6626−6632

Journal of the American Chemical Society
Article
the photoinduced solid-to-solid reaction of dinitroanthracene
to anthraquinone occurs by a reconstructive process where
molecular changes precede macroscopic changes in the
crystal.20 While the detailed mechanisms of these phase
transitions and observed macroscopic changes are unknown,
these reactions are only observed when the temperature of the
reaction is below the eutectic point of the two-component
phase diagram; otherwise, melting is observed. It is likely that
these reactions proceed through small, steady-state amounts of
amorphous phases, so that nucleation and growth of the
product phase may only occur at temperatures above the
corresponding glass transition temperature where molecular
reorganization is permitted. There are reports on the nucleation
and growth processes in the solid-state photoreaction of α-
trans-cinnamic acid.21 Finally, as indicated by Figure 1, path
(d), there is an alternative where the starting crystal and
subsequent mixed crystal phase are destroyed by melting or
amorphization, depending on the reaction temperature in
relation to the melting point and glass transition temperature of
the product.
Scheme 1
aziridine.14 The release of gas molecules during reactions in
crystal and its repercussion on the observed reactivity have
been reported by McBride and co-workers with crystalline
diacyl peroxides.15 It was shown that molecules near a reaction
site tend to rearrange to attenuate the local stress created by the
formation of CO2. It is reasonable to expect that similar stress
will be observed upon release of N2 in crystals of triazolines.
While those structural changes are expected to affect the
kinetics of elementary reactions as a function of conversion,
they have no effect on the nature of the product formed in the
case of the triazoline crystals.
Given the efficiency of the chemical transformations in
Scheme 1, it was of interest to determine the nature of the
physical changes accompanying the reaction: whether or not
they are topochemical in nature. Based on our previous analysis
of the phase changes that occur in solid-state reactions,16 it is
expected that all topochemical reactions must begin by
formation of a solid solution of the product(s) in the lattice
of the reactant. The most appealing and most demanding kind
of transformations, commonly referred to as “topotactic” (or
single crystal-to-single crystal),16,17 occur in the same phase and
retain single crystallinity as the reaction goes from starting
material to the product [Figure 1, path (a)]. Other reactions are
We recognized that the large changes in molecular size and
shape that take place during the photoinduced denitrogenation
of triazolines represent a challenge to the most widely accepted
notions of topochemistry. With that in mind, we decided to
investigate the solid-state reaction of triazolines containing very
large bulky groups that could act as anchors, at either one or
both ends of the molecule.
The compounds chosen to test the limits of topochemical
restrictions in molecular crystals are illustrated in Scheme 2.
Scheme 2
Figure 1. Potential phase changes accompanying solid-state reaction
(for a detailed explanation please see main text).
able to continue in the crystal phase of the reactant, even if it is
very different from the stable crystal phase of the final product
[Figure 1, path (b)]. This was observed in our previous work,
where powder X-ray diffraction (PXRD) after 100% conversion
to aziridines 2 (Scheme 1) showed broadened peaks with a
pattern that matched that of unreacted triazoline crystals.13
These reactions have a tendency to lose long-range order and
may not be suitable for single-crystal XRD analysis, but may be
followed by PXRD. Another relatively common type of solid-
to-solid reaction involves a reconstructive phase transition from
the crystal of the reactant to the crystal of the product [Figure
1, path (c)].18 Naumov and co-workers have reported this type
of solid-state reaction in the thermochromism of dinitrobenzyl
derivatives, finding that the phase transition is reversible.19 In
another example, Brillante and co-workers recently showed that
Compounds 3 and 5 are isomers that differ in the location of
the bulky tritylacetylene, either on the phenyl group of the
succinimide fragment or on the phenyl group of the triazoline.
Compound 7 has bulky groups at each of the two positions
(Scheme 2). A representation of the reaction cavity,8 with the
boundary formed by near neighbors in the crystal shown with a
dotted line, in the structures in Scheme 2 highlights the fact
that the size and shape of the reactant and the product are very
different. The main question that arises when the structures of
compounds 3, 5, and 7 are considered in the context of
6627
dx.doi.org/10.1021/ja401577p | J. Am. Chem. Soc. 2013, 135, 6626−6632

Journal of the American Chemical Society
Article
topochemistry is whether their reactions may proceed, and
what type of phase changes may occur as a result of such large
structural changes. From the point of view of the structure of
the chemically evolving system, it is of interest to find out
whether the reacting solid may retain some of its original
integrity as determined by its PXRD pattern. In this Article, we
describe the synthesis of triazolines 3, 5, and 7 and confirm
their efficient solid-state reactivity. We discovered that solid-to-
solid photochemical reactions at ambient temperature proceed
through metastable phases that eventually become amorphous,
while a solid-to-solid thermal reaction carried out at 160 °C
proceeds by a reconstructive phase transition mechanism.
These and other observations suggested that amorphous phases
are formed when there is a large mismatch between reactants
and products, and that they can be trapped at low temperatures,
presumably below the glass transition temperature. In contrast,
analogous reactions occurring at higher temperatures take
advantage of molecular motion that allows for the stable crystal
phase of the product to nucleate and crystallize.
N-phenylmaleimides were prepared according to published
procedures22,23 from aniline and compound 9. Scheme 4b,c
shows the synthesis of the azides 10 and 11 and maleimides 12
and 13, respectively.
Triazolines 3, 5, and 7 were purified by column
chromatography, and their molecular structures were charac-
1
terized by FTIR, H and 13C NMR, and mass spectrometry.
1
The main difference in the H NMR spectra of the three
triazolines is the displacement of the signal from the aromatic
protons in the phenyl rings attached to the succinimide and
triazoline moieties. Triazoline 3 presents the signals corre-
sponding to the phenylene group linked to the succinimide
moiety as an AA′BB′ system at ca. 7.33 and 7.60 ppm, and a set
of signals at ca. 7.19, 7.42−7.45, and 7.71−7.72 ppm for the
phenyl group that is attached to the triazoline. On the other
hand, the triazoline 5 presents the corresponding succinimide
N-phenyl signals at ca. 7.28, 7.43, and 7.48 ppm and the
phenylene AA′BB′ pattern at ca. 7.56 and 7.67 ppm.
Accordingly, triazoline 7 shows two AA′BB′ patterns, one at
ca. 7.32 and 7.60 ppm that corresponds to the succinimide
phenylene, and the other at ca. 7.56 and 7.66 ppm that
corresponds to the phenylene group attached to the triazoline.
Aziridine Preparation and Characterization. Pure
samples of aziridines 4, 6, and 8 were obtained in quantitative
yields by heating polycrystalline samples of the corresponding
triazolines for 30 min in an oil bath at 160 °C. Reactions of
triazolines 3 and 7 yielded oily products that could be
precipitated from dichloromethane. The thermal reaction of
triazoline 5 carried out under these conditions occurred
without melting in a relatively unusual solid-to-solid thermal
reaction.
RESULTS AND DISCUSSION
■
Triazoline Synthesis and Characterization. Synthesis of
compounds 3, 5, and 7 was carried out according to procedures
reported in our previous work by dipolar cycloaddition of
phenyl-substituted azides and maleimides (Scheme 3).13
Scheme 3
The identity of the aziridine products was confirmed by
FTIR, 1H and 13C NMR, and mass spectrometry. The aziridine
IR spectra showed the characteristic frequencies of the N-
substituted aziridines,24 which include ring stretching signals in
the range of 1260−1280 cm−1, ring CH wagging at 1170−1200
cm−1, and ring CH twisting at 1320−1360 cm−1, as well as the
absence of the NN stretching vibration at 1480−1490 cm−1
observed in the starting materials. The 1H NMR spectra
revealed the appearance of a broad singlet at ca. 3.87 ppm,
To prepare the starting materials bearing the bulky
substituent, we synthesized 4-(3,3,3-triphenylpropyn-1-yl)-
aniline 9 by Sonogashira coupling between 1,1,1-triphenyl-
prop-2-yne and 4-bromoaniline (Scheme 4a). Phenylazides and
Scheme 4
6628
dx.doi.org/10.1021/ja401577p | J. Am. Chem. Soc. 2013, 135, 6626−6632

Journal of the American Chemical Society
Article
corresponding to the equivalent protons of the aziridine ring.
The 13C NMR spectra also showed only one signal
corresponding to the equivalent bridgehead carbons in the
aziridine at ca. 40 ppm.
Attempts to crystallize each of the three aziridines from
different solvent mixtures yielded no crystals that were suitable
for characterization by XRD, although analysis of a poorly
diffracting twinned crystal of aziridine 8 grown from ethyl
acetate led to a structural model that supports the expected
connectivity (see Supporting Information). The molecular
structure obtained in this manner is in agreement with a
structure obtained by computational analysis of an aziridine
analogue, N-(4-cyanophenyl)azirido-[2,3-c]-N-phenylmalei-
mide, with the B3LYP/6-31G* level of theory, which indicates
that the endo configuration is ca. 1.0 kcal/mol more stable than
the one where the N-phenylaziridine adopts the exo
configuration.25 The melting points of aziridines 4, 6, and 8
were 178−181, 260−262, and 142−148 °C, respectively.
Thermal Analyses and Thermal Denitrogenation. In
order to investigate the thermal loss of nitrogen from triazolines
3, 5, and 7 in more detail, we carried out thermogravimetric
analysis (TGA) and differential scanning calorimetry (DSC)
measurements in the temperature range of 50−300 °C. The
occurrence of the thermal reaction is readily identified by an
exothermic peak in the DSC trace at a temperature that
matches the loss of mass corresponding to the nitrogen
molecule observed in the TGA. In the case of triazoline 3, an
endothermic transition at 138 °C was followed by a broad
exothermic peak from 140 to 165 °C, indicating that melting
occurs prior to the loss of nitrogen (Figure 2, top). For
triazoline 5, the DSC trace showed an exothermic peak at 159
°C corresponding to the loss of nitrogen followed by an
endothermic peak at 258 °C, a temperature that is almost 100
°C higher (Figure 2, middle). It was subsequently shown that
the second transition corresponds to the melting of aziridine 6,
thus confirming a solid-to-solid thermal reaction with a
reconstructive phase transition mechanism, as indicated by
path (c) in Figure 1. Finally, the DSC trace of triazoline 7
displayed only one broad exothermic peak at ca. 150−160 °C,
which was matched by the loss of nitrogen in the TGA (not
shown). As the melting point of the product is significantly
lower (142−148 °C), the reaction proceeded in the liquid
phase.
Figure 2. Differential scanning calorimetry traces of (top) triazoline 3
and (middle) 5, illustrating melting and the loss of nitrogen. The
bottom frame shows a thermogram of the amorphous phase of
aziridine 6 obtained from photolysis of crystalline 5 at 298 K (melting
not shown occurs at ca. 260 °C).
recrystallized aziridines 4, 6, and 8 in the range 2θ = 5−50°.
While triazolines 3 and 5 presented characteristic sets of Bragg
reflections typical of crystalline materials, samples of triazoline
7 showed no Bragg reflections, confirming that the correspond-
ing solid is amorphous. By contrast, PXRD measurements of
aziridines solids 4, 6, and 8 obtained by slow solvent
evaporation revealed that all three samples are crystalline.
With the stable end-points of the three reactions characterized,
we carried out analogous measurements as a function of
conversion, with samples taken from the photoreaction every
10 min. Optical microscopy was also used to monitor changes
in the birefringence of a crystal of compound 5 as a function of
reaction progress (see below). In order to determine whether
the nitrogen gas generated during the reaction remains trapped
in the crystals or escapes rapidly into the atmosphere, a small
amount of microcrystalline 5 was irradiated in a capillary tube
with a drop of mineral oil deposited with a syringe needle above
the solid in order to trap a plug of air. An increase in the
amount of gas as a function of increased irradiation time was
detected as a displacement of the mineral oil, indicating that
nitrogen escapes from the crystal during the time scale of the
reaction. An estimate of the gas released during the experiment
could be used to determine the extent of reaction, which was
shown to be complete after ca. 40 min. We were able to
confirm that all the nitrogen formed in the reaction had
escaped from the crystals during the time of reaction with the
help of TGA analyses of irradiated samples, which showed that
there was no further loss of mass (N2) when the sample was
heated from ambient temperature to 115 °C. (For details of the
experimental conditions and results, please see the Supporting
Photochemical Reactions and PXRD Analysis as
Function of Reaction Progress. When dry powders of
triazolines 3, 5, and 7 were exposed to UV light with λ ≥ 290
nm from a medium-pressure 450-W Hg Hanovia lamp passed
through a Pyrex filter at 298 K, smooth reactions yielded
aziridines 4, 6, and 8 as the only products. The starting
compound was always a microcrystalline powder, except for the
compound 7, which was an amorphous powder. Solid-to-solid
transformations occurred with full conversion within 40−50
min under the conditions of the experiment described in the
Supporting Information. Reaction progress was determined
1
every 10 min by H NMR by measuring the disappearance of
the triazoline five-membered ring doublets at 4.98 and 5.90
ppm and the appearance of the aziridine singlets at 3.87 ppm.
Several attempts to obtain suitable single crystals of either the
starting materials or products for crystallography structure
determination by X-ray had been unsuccessful. Therefore, to
determine the nature of the photoinduced solid-to-solid
reactions, we set out to measure the PXRD of the solid phases
of starting triazolines 3, 5, and 7, as well as those of the solvent-
6629
dx.doi.org/10.1021/ja401577p | J. Am. Chem. Soc. 2013, 135, 6626−6632

Journal of the American Chemical Society
Article
Figure 3. Comparison of the PXRD of triazolines 3 (left column) and 5 (right column) before reaction (b), and after different photochemical
conversion values (c,d), with the amount of product indicated. Samples of aziridines 4 and 6 were obtained by crystallization of the amorphous phase
by heating above Tg (e), by a solid-to-solid thermal reaction above the glass transition temperature and below melting in the case of 6 (a, right), or by
recrystallization from solution of 4 (a, left). PXRD similar to that for 6 (a, right) was observed upon recrystallization from solution.
Information.) The evolution of gas may also be used to
determine the conversion of the reaction by estimating the
quantity of nitrogen released.
at T ≈ 110 °C (Figure 2, bottom). Notably, DSC analysis of
amorphous aziridine samples obtained by solid-state photo-
reaction of triazoline 5 does not show a clear glass transition
but rather a very broad endothermic signal between ca. 110 and
160 °C (Figure 2, bottom) that is consistent with crystallization
of the amorphous material into the polycrystalline aziridine
phase.
The photodenitrogenation of compound 5 was also studied
by optical microscopy with a small crystal exposed to UV light
for eight periods of 5 min (Figure 4). While no macroscopic
changes were observed and the crystal appeared intact when
viewed under reflected white light (Figure 4, left column), a
Starting with triazoline 5, we discovered that the relatively
sharp PXRD pattern of the starting material (Figure 3b, right)
undergoes relatively minor changes at low conversion values
(not shown), suggesting the formation of a solid solution that
retains the packing arrangement of the starting material. When
product accumulation reached values on the order of 60%
(Figure 3c, right), the percentage of conversion was calculated
on the basis of the NMR signals of the triazoline and aziridine
bridgehead protons. Diffraction peaks broaden greatly until the
pattern loses most of its signals when the reaction goes beyond
ca. 85% (Figure 3d, right). Since none of the peaks of the stable
aziridine are observed as the photochemical reaction proceeds
at 298 K, one can safely conclude that there is not a
reconstructive phase transition and that the photoreaction
occurs by amorphization of the sample [Figure 1, path (d)].
13C CPMAS NMR showed a gradual increase in the signals line
width, in agreement with the amorphization of the sample
observed by PXRD (Supporting Information). Notably, when
the amorphous photoproduct was heated neat to 160 °C, well
below the melting point of aziridine 6 (260−262 °C), we were
able to obtain a PXRD pattern essentially identical to that of
the solvent-recrystallized aziridine (Figure 3e, right). Further-
more, the same powder pattern was obtained when crystals of
triazoline 5 were heated to 160 °C (Figure 3a, right), indicating
a clean solid-to-solid reaction by a reconstructive phase
transition mechanism.26 Taken together, these observations
suggest that the difference between the low-temperature
amorphization and the high-temperature reconstructive phase
transition should be related to the rigidity of the sample,
suggesting that crystallization from amorphous phases is likely
to occur above the corresponding glass transition temperature
Figure 4. Optical microscopy pictures of a crystal of triazoline 5 before
and after irradiation (time indicated in minutes). Images obtained with
cross polarizers (right column) illustrate the loss of crystallinity by the
loss of birefringence. Images obtained with normal illumination (left
column) show no apparent changes.
6630
dx.doi.org/10.1021/ja401577p | J. Am. Chem. Soc. 2013, 135, 6626−6632

Journal of the American Chemical Society
Article
progressive loss of birefringence was observed as a function of
increased exposure (Figure 4, right column). The crystal lost
most of its birefringence after 40 min, and it was later
confirmed that full conversion to aziridine 6 had taken place.
This experiment clearly revealed that the solid becomes
amorphous with increased conversion values, suggesting a
process that occurs along path (d) in Figure 1. By comparison,
the formation of anthraquinone from dinitroanthracene
recently reported by Brillante and co-workers proceeds along
path (c) in Figure 1,20 with the crystal showing notable changes
of shape without amorphization. Even though reaction progress
modifies the crystal (amorphization or restructuration) in these
and other cases, the reaction still occurs in the solid state.
The results obtained upon photolysis and PXRD analysis of
triazoline 3 at 298 K were analogous to those described above
(Figure 3, left). The powder pattern remained relatively
unchanged up to ca. 75% conversion, when it started
broadening in a significant manner. Irradiation to full
conversion transformed the sample into a largely amorphous
solid with a PXRD that has only small residual signals,
suggesting a very small degree of crystallinity or a very small
fraction of a crystalline phase. The fully converted solid was
shown to have a glass transition process at 130 °C, and after
softening the sample crystallized into the stable crystalline
phase of aziridine 4. The reaction of triazoline 7 to aziridine 8
was not investigated in as much detail, considering that the
sample starts as an amorphous solid and the end point of the
photoreaction at ambient temperature is also an amorphous
solid.
start in the crystalline phase of the reactant {R}R, proceed
through the reactant-like mixed crystalline phase {R+P}R until
the solubility limit of the product is reached, and then
transform into a metastable reactant-like phase, or become
amorphous, until the composition of the product-like crystal {R
+P}P is reached and the sample can recrystallize into the new
product phase {P}P. In our experience, these reactions are
common and can be documented as occurring by a
reconstructive phase transition mechanism. Finally, we propose
that reactions taking place at temperatures that are well below
the glass transition temperature of the two-component system
(zone C) do not allow molecules to undergo the motions
required for nucleation and crystallization and therefore must
proceed in a metastable reactant-like phase or, when the
structural changes on going from reactant to product are too
drastic, become amorphous. We propose that the solid-state
photochemical reactions of triazolines 3 and 5 take place in
zone C, as seen through the loss of birefringence of the solid
(Figure 4), with reactions proceeding along metastable and
amorphous phases that can be recrystallized by heating the
solid product above the glass transition temperature. By
contrast, we deduce that the thermolysis of triazoline 5 occurs
above the glass transition temperature (Tg), and well below the
eutectic temperature (Teut) so that liquid phases are not
encountered, as shown by calorimetric analysis (Supporting
Information). Under these conditions, the thermal reaction
proceeds by a solid-to-solid reconstructive phase transition
mechanism along zone B. On the other hand, calorimetric
analysis of crystalline triazoline 3 shows that it melts and then
rapidly loses nitrogen, such that the product is formed in the
liquid phase along zone A. Finally, triazoline 7 starts reacting in
an amorphous phase, and the thermal reaction occurs when the
sample softens. As suggested in Figure 5, it seems reasonable
that the types of phases involved in these reactions, whether
crystalline, amorphous, or liquid, will depend on the physical
properties of the reactant, product, and eutectic mixtures, as
well as on their ability to form mixed crystalline phases.
On the basis of these results, we propose a qualitative phase
diagram as a function of reaction progress that accounts for the
reaction modes indicated in Figure 1. Shown in Figure 5, the
CONCLUSIONS
■
Using the photoinduced denitrogenation of crystalline triazo-
lines with bulky substituents as a test model to challenge the
topochemical postulate, we were able to document several
types of solid-to-solid reactions that follow different phase
transformation mechanisms. Despite very large differences in
size and shape between the triazoline reactants and the aziridine
products, photoinduced solid-to-solid reactions proceeded to
completion within 50 min to yield the corresponding aziridines
in amorphous phases and ca. 100% yields. By contrast,
depending on the substituents, the thermal reactions were
shown to occur upon melting, after softening, or by a solid-to-
solid reaction that occurs by a reconstructive phase transition
mechanism. It was also shown that amorphous phases obtained
by photochemical irradiation could be recrystallized by heating,
presumably above the glass transition temperature. The results
obtained in this study increase our understanding of phase
transformation mechanisms involved in solid-to-solid reactions
and thus may help in the development of novel green chemistry
strategies and materials science applications.
Figure 5. Schematic phase diagram of the triazoline denitrogenation
reaction progress (for a detailed explanation please see main text).
phase diagram has three temperature zones that are expected to
display different phase behavior when crystalline R is the
starting material. Reactions occurring in the higher temperature
zone, indicated with label A, will start in the solid state in a
reactant-like phase {R+P}R until its composition reaches the
solid−liquid line and the reaction begins to melt. If the melting
point of P is higher than that of R, as illustrated in the figure,
the product mixture may recrystallize in a product-like phase {R
+P}P when its composition reaches the values of the solid−
liquid line near the product. If the melting point of R is much
higher than that of P and the reaction occurs at a temperature
that is above the melting point of the product, the reaction
would be completed in a liquid phase. Reactions carried out at
temperatures between the eutectic point and the glass
transition temperatures of the two-component system are
labeled as occurring in zone B. These reactions are expected to
ASSOCIATED CONTENT
■
S
* Supporting Information
Synthesis and characterization of all compounds; H and 13C
1
NMR and mass spectra; DSC, PXRD, 13C CPMAS, NQS, and
6631
dx.doi.org/10.1021/ja401577p | J. Am. Chem. Soc. 2013, 135, 6626−6632

Journal of the American Chemical Society
Article
V., Schanze, K., Eds.; Marcel Dekker: New York, 1998; Vol. 2, pp
195−248.
solid-state FTIR studies. This material is available free of charge
via the Internet at http://pubs.acs.org.
(17) (a) Friscic, T.; MacGillivray, L. R. Z. Kristallogr. 2005, 220, 351.
(b) Barbour, L. J. Aust. J. Chem. 2006, 59, 595. (c) Halasz, I. Cryst.
Growth Des. 2010, 10, 2817. (d) Nieuwendaal, R. C.; Mattler, S. J.;
Bertmer, M.; Hayes, S. E. J. Phys. Chem. B 2011, 115, 5785.
(18) Choi, T.; Cizmeciyan, D.; Kahn, S. I.; Garcia-Garibay, M. A. J.
Am. Chem. Soc. 1995, 117, 12893.
AUTHOR INFORMATION
■
Corresponding Author
mgg@chem.ucla.edu
Notes
(19) (a) Naumov, P.; Sakurai, K. Cryst. Growth Des. 2005, 5, 1699.
(b) Lee, J. H.; Naumov, P.; Chung, I. H.; Lee, S. C. J. Phys. Chem. A
2011, 115, 10087.
The authors declare no competing financial interest.
(20) Salzillo, T.; Bilotti, I.; Della Valle, R. G.; Venuti, E.; Brillante, A.
J. Am. Chem. Soc. 2012, 134, 17671.
(21) (a) Bertmer, M.; Nieuwendaal, R. C.; Barnes, A. B.; Hayes, S. E.
J. Phys. Chem. B 2006, 110, 6270. (b) Nieuwendaal, R. C.; Bertmer,
M.; Hayes, S. E. J. Phys. Chem. B 2008, 112, 12920.
(22) Braish, T. F.; Fox, D. E. Synlett 1992, 12, 979.
ACKNOWLEDGMENTS
■
This research was supported by NSF grants DMR1101934 and
CHE-0844455. We thank UC-Mexus for a postdoctoral
fellowship to D.d.L. We also thank B. Rodriguez-Molina for
13C CPMAS NMR measurements.
(23) Leyva, E.; de Loera, D.; Jimen
2010, 51, 3978.
́
ez-Catano, R. Tetrahedron Lett.
̃
REFERENCES
(1) Trommsdorff, S. Ann. Chem. Pharm. 1834, 11.
(2) Matsuura, T.; Sata, Y.; Katsuyuki, O. Tetrahedron Lett. 1968, 44,
4627.
(3) Natarajan, A.; Tsai, C. K.; Khan, S. I.; McCarren, P.; Houk, K. N.;
Garcia-Garibay, M. A. J. Am. Chem. Soc. 2007, 129, 9846.
(4) Ramamurthy, V.; Venkatesan, K. Chem. Rev. 1987, 87, 433.
(5) (a) Kohlshutter, V. Anorg. Allg. Chem. 1918, 105, 121.
(b) Kohlshutter, V. Kolloid-. 1927, 42, 254. (c) Kohlshutter, V. Z.
Anorg. Allg. Chem. 1920, 111, 193.
■
(24) Spell, H. L. Anal. Chem. 1967, 39, 185.
(25) Computational analysis using B3LYP/6-31G* for N-(4-
cyanophenyl)azirido-[2,3-c]-N-phenylmaleimide: exo configuration,
−969.0976 hartree; endo configuration, −969.0993 hartree; difference,
1.038 kcal/mol.
(26) Similar observations were documented by solid-state 13C
CPMAS NMR and solid-state FTIR analysis. Please see the
Supporting Information.
(6) (a) Cohen, M. D.; Schmidt, G. M. J. J. Chem. Soc. 1964, 1996.
(b) Cohen, M. D.; Schmidt, G. M. J.; Sonntag, F. I. J. Chem. Soc. 1964,
2000. (c) Schmidt, G. M. J. J. Chem. Soc. 1964, 2014.
(7) Examples of topochemical control reactions: (a) Ramamurthy,
V.; Venkatesan, K. Chem. Rev. 1987, 87, 433. (b) MacGilvray, L.;
Papaefstathiou, G. S. Solid-state reactivity/topochemistry. In Encyclo-
pedia of supramolecular chemistry; Atwood, J. L., Steed, J. W., Eds;,
Taylor & Francis: New York, 2004; pp 1316−1321. (c) Mandel, S. M.;
Singh, P. N. D.; Muthukrishnan, S.; Chang, M.; Krause, J. A.;
Gudmundsdottir, A. D. Org. Lett. 2006, 8, 4207. (d) Sankaranarayanan,
J.; Bort, L. N.; Mandel, S. M.; Chen, P.; Krause, J. A.; Brooks, E. E.;
Tsang, P.; Gudmundsdottir, A. D. Org. Lett. 2008, 10, 937.
(8) Cohen, M. D. Angew. Chem., Int. Ed. 1975, 14, 386.
(9) Xu, Y.; Smith, M. D.; Geer, M. F.; Pellechia, P. J.; Brown, J. C.;
Wibowo, A. C.; Shimizu, L. S. J. Am. Chem. Soc. 2010, 132, 5334.
(10) (a) Gnanaguru, K.; Ramasubbu, N.; Venkatesan, K.;
Ramamurthy, V. J. Org. Chem. 1985, 50, 2337. (b) Nomura, S.; Itoh,
T.; Nakasho, H.; Uno, T.; Kubo, M.; Sada, K.; Inoue, K.; Miyata, M. J.
Am. Chem. Soc. 2004, 126, 2035.
(11) (a) Garcia-Garibay, M. A. Acc. Chem. Res. 2003, 36, 491.
(b) Garcia-Garibay, M. A.; Shin, S.; Sanrame, C. Tetrahedron 2000, 56,
6729. (c) Campos, L. M.; Garcia-Garibay, M. A. In Reactive
Intermediates; Platz, M. S., Jones, M., Moss, R., Eds.; Wiley: Hoboken,
NJ, 2007.
(12) Alternative mechanisms could involve reactions that occur by
quantum mechanical tunneling from molecules that are structurally
predisposed to react from their lower vibrational states: (a) Johnson,
B. A.; Garcia-Garibay, M. A. The Spectrum 1998, 11, 1. (b) Garcia-
Garibay, M. A.; Gamarnik, A.; Bise, R.; Pang, L.; Jenks, W. S. J. Am.
Chem. Soc. 1995, 117, 10264.
(13) de Loera, D.; Garcia-Garibay, M. A. Org. Lett. 2012, 14, 3874.
(14) (a) Sheiner, P. J. Am. Chem. Soc. 1968, 90, 988. (b) Meier, H.;
Zeller, K. P. Angew. Chem., Int. Ed. Engl. 1977, 16, 835. (c) Schultz, A.
G.; Sha, C.-K. J. Org. Chem. 1980, 45, 2040. (d) Tukada, H. Chem.
Commun. 2000, 63. (e) Chen, Y.-J.; Hung, H.-C.; Sha, C.-K.; Chung,
W.-S. Tetrahedron 2010, 66, 176.
(15) (a) McBride, J. M. Acc. Chem. Res. 1983, 16, 304. (b) McBride,
J. M.; Segmuller, B. E.; Hollingsworth, M. D.; Mills, D. E.; Weber, B.
A. Science 1986, 234, 830.
(16) Keating, A. E.; Garcia-Garibay, M. A. Photochemical Solid-To-
Solid Reactions. In Organic and Inorganic Photochemistry; Ramamurthy,
6632
dx.doi.org/10.1021/ja401577p | J. Am. Chem. Soc. 2013, 135, 6626−6632