Experimental and Theoretical Study of Selectivity in Mechanochemical Cocrystallization of Nicotinamide with Anthranilic and Salicylic Acid

Article abstract of DOI:10.1021/acs.cgd.7b01512

Selectivity in mechanochemical cocrystal formation between nicotinamide and anthranilic acid or salicylic acid was studied using tandem in situ reaction monitoring by powder X-ray diffraction (PXRD) and Raman spectroscopy. Selectivity was probed by offering a competing cocrystal coformer to a previously prepared cocrystal or under competitive reaction conditions where all cocrystal coformers, in different stoichiometric ratios, were introduced together in the starting reaction mixture. Reaction paths were dependent on the starting mixture composition, and we find that the formation of intermediates and the final product can be predicted from solid-state ab initio calculations of relative energies of possible reaction mixtures. In some cases, quantitative assessment revealed different reaction profiles derived from PXRD and Raman monitoring, directly indicating, for the first time, different mechanochemical reactivity on the molecular and the bulk crystalline level of the reaction mixture.

Full text of DOI:10.1021/acs.cgd.7b01512

Article
pubs.acs.org/crystal
Cite This: Cryst. Growth Des. XXXX, XXX, XXX−XXX
Experimental and Theoretical Study of Selectivity in
Mechanochemical Cocrystallization of Nicotinamide with Anthranilic
and Salicylic Acid
Published as part of a Crystal Growth and Design virtual special issue Honoring Prof. William Jones and His
Contributions to Organic Solid-State Chemistry
Stipe Lukin,† Ivor Loncaric,† Martina Tireli,† Tomislav Stolar,† Maria V. Blanco,‡ Predrag Lazic,†
̌
́
́
Krunoslav Uzarevic,† and Ivan Halasz*,†
̌
́
†Ruđer Bosk
̌ ́ ̌
ovic Institute, Bijenicka c. 54, 10000 Zagreb, Croatia
‡ESRF - the European Synchrotron, 71 Avenue des Martyrs, 38000 Grenoble, France
S
* Supporting Information
ABSTRACT: Selectivity in mechanochemical cocrystal for-
mation between nicotinamide and anthranilic acid or salicylic
acid was studied using tandem in situ reaction monitoring by
powder X-ray diffraction (PXRD) and Raman spectroscopy.
Selectivity was probed by offering a competing cocrystal
coformer to a previously prepared cocrystal or under
competitive reaction conditions where all cocrystal coformers,
in different stoichiometric ratios, were introduced together in
the starting reaction mixture. Reaction paths were dependent
on the starting mixture composition, and we find that the
formation of intermediates and the final product can be predicted from solid-state ab initio calculations of relative energies of
possible reaction mixtures. In some cases, quantitative assessment revealed different reaction profiles derived from PXRD and
Raman monitoring, directly indicating, for the first time, different mechanochemical reactivity on the molecular and the bulk
crystalline level of the reaction mixture.
conducted in the solid state. Moreover, we hypothesize that
many principles derived for solution reactions may also be valid
for mechanochemical reactions. In principle, we believe that a
surface of potential energy could be constructed where each
stable solid phase would lie in its minimum and where energy
barriers could be identified for their potential transforma-
tions.30−32 So far, such an energy landscape is evidenced by the
observation of the Ostwald’s rule of stages,29,33 the equilibra-
tion of solids,34−36 and a stronger than expected influence of
temperature on mechanochemical reactions.37 Also, while the
catalytic effect of liquid additives is well-known,38,39 direct
observation of lowering of an energy barrier and hence reaction
acceleration by different liquid additives in a liquid-assisted
grinding (LAG) reaction was recently observed.40 In creating
parallels with solution reactions, one also needs to consider
solvation of solid particles,41 which seems to have enabled
control over selective preparation of three polymorphs of
nicotinamide/benzoic acid cocrystals.29
INTRODUCTION
■
Mechanochemical reactions are becoming increasingly attrac-
tive since they offer a means to perform selective chemical
synthesis with excellent atom- and energy-economy and with
little or, in some cases, no waste generation.1−3 Being
previously limited largely to processing of minerals, metals,
alloys, and inorganic materials,4 chemical transformations
achievable by using mechanochemistry now encompass almost
every aspect of chemical synthesis, including synthesis of
organic,5−8 organometallic,9−11 metal−organic com-
pounds12−14 as well as multicomponent solids such as
cocrystals15,16 and the preparation of nanoparticles,17 molecular
nanostructures,18 pharmaceuticals,19 advanced peptide syn-
thesis,20,21 supramolecular recognition,22 and recently, enzy-
matic catalysis.23,24 The widespread use of mechanochemistry
has been since recently accompanied by methodology to probe
the course and mechanisms of milling reactions by real-time in
situ monitoring techniques based on X-ray diffraction,25,26
Raman spectroscopy,9,27 or their simultaneous use in
tandem.28,29
Specific energy relations between solids in an energy
landscape would need to result in selectivity in mechanochem-
These techniques have revealed surprisingly fast mechano-
chemical reactions and metastable and short-lived intermedi-
ates, indicating that mechanochemical reactions may bear a
strong resemblance to solution chemistry, despite them being
Received: October 30, 2017
Revised: January 5, 2018
© XXXX American Chemical Society
A
DOI: 10.1021/acs.cgd.7b01512
Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design
Article
ical reactions. Indeed, in a pioneering study, Etter has shown
specific pairing of nucleobases in a grinding reaction in the
presence of a competing nucleobase.42 Caira and co-workers
have also shown selective cocrystal preparation starting from a
mixture of cocrystal coformers,43 but they have taken their
study a step further in showing that an already formed cocrystal
will react if a more selective coformer is offered. In a more
recent example, we have shown strong selectivity in supra-
molecular anion binding with the ability of a host molecule to
discriminate even among six dicarboxylic acids.22 Abourahma
and co-workers have studied the robustness of a particular
theophylline cocrystal while grinding with other potential
coformers and offered a qualitative interpretation of coformer
replacements based on hydrogen bonding preferences.44
Emmerling and co-workers have described a selectivity study
between several active pharmaceutical ingredients and
anthranilic acid as a cocrystal coformer and suggested that
energetic and kinetic factors may need to be taken into account
as well.45
used in these experiments is designated as polymorph I (ana-I)
and is zwitterionic. More precisely, of the two molecules in the
asymmetric unit of polymorph I, one is a “normal” non-
zwitterionic molecule while the other is zwitterionic. Before
studying selectivity in multicomponent reaction mixtures, we
have attempted to prepare naana and nasal cocrystals
separately. In the 1:1 na:ana reaction mixture, the naana
forms readily and directly from reactants with no detectable
intermediates. Formation of naana begins immediately upon
milling, and the reaction was fully complete within 20 min
milling. Crystal structure of the naana cocrystal was thus far
unknown, so we have solved it from powder diffraction data
(Figures 1c and S6). The crystal structure resembles that of the
known nasal cocrystal46 (Figure 1b), but it is stabilized by a
larger number of hydrogen bonds since the amino group of ana
acts both as an acceptor and a donor of hydrogen bonds. In the
naana cocrystal, ana is not zwitterionic. The 1:1 cocrystal
formation between sal and na also proceeds directly from
reactants, and the reaction is complete within 20 min milling.
Final product is the known nasal cocrystal.46 Even though both
reactions are complete within ca. 20 min milling, in situ
monitoring revealed naana formation to be somewhat faster
than nasal formation, suggesting a lower energy barrier for
naana formation (Figure 2). For both cocrystal synthesis
reactions, quantitative assessments based on either Raman and
PXRD monitoring have revealed similar reaction profiles,
indicating that the reaction mixture is crystalline and contains
little or no amorphous component.
Herein, we describe selectivity in mechanochemical cocrystal
formation between nicotinamide (na) and two benzenecarbox-
ylic acids differing in the functionality on the ortho position:
anthranilic acid (ana) and salicylic acid (sal) (Figure 1). Via
Cocrystal Selectivity and Replacement Reactions.
Following cocrystal synthesis, we have attempted to predict
the selectivity in formation of each cocrystal by using ab initio
calculations in the solid state to estimate the relative energies of
cocrystals and cocrystal coformers. Ab initio calculations are
likely to provide sufficiently precise energies, which can be used
to assess possibilities of cocrystal formation48 as well as
polymorphic transitions.49 In particular, density functional
theory (DFT) with newly developed functionals that include
van der Waals interactions achieve good accuracy in modeling
molecular crystals.50 Of several van der Waals implementations
in DFT, here we choose the nonemipirical vdW-DF-cx
functional51,52 that proved its accuracy in similar systems.49,53,54
Since the evaluation of (temperature dependent) Gibbs free
energies from DFT comes with a very high computational cost,
here we only report the relative enthalphies. It should be noted
that this could be, in addition to the exchange-correlation
functional, a source of imprecision in reported predictions.
While comparing solids with the same composition is
straightforward, cocrystals with different compositions can
also be compared, provided that the overall composition is kept
fixed by considering also the energy of the competing coformer
(Scheme 1). Such considerations should predict which
cocrystal will be stable if offered a competing coformer and
which cocrystals will react to form a more stable cocrystal.
Our calculations predict all cocrystals to be more stable than
pure components (Scheme 1). The naana cocrystal (see later)
is only 0.5 kJ mol−1 more stable than the mixture of na and ana-
I. The salana cocrystal is by 4.0 kJ mol−1 more stable than pure
sal and ana-I. With 13.8 kJ mol−1 difference, the nasal cocrystal
is most stable of all three cocrystals relative to pure
components. The mixture nasal and ana-I, in the 1:1 molar
ratio, is the most stable combination.47 Therefore, sal should be
able to replace ana from the naana cocrystal, while ana should
not replace sal from the nasal cocrystal.
Figure 1. (a) Mechanochemical preparation of cocrystals of na with sal
and ana. Hydrogen bonding in the (b) crystal structure of the nasal
cocrystal46 and (c) crystal structure of the naana cocrystal.
quantitative assessment by tandem in situ powder X-ray
diffraction and Raman monitoring,29 we have observed varied
and intricate reaction paths in competitive cocrystal formation
and cocrystal replacement reactions, involving formation of
metastable intermediates and multistep solid-state proton
transfer. Accompanied by theoretical calculations of stabilities
of participating solid phases, we find that the observed
selectivity can be readily predicted from relative stabilities of
cocrystals and coformers, while the reaction mechanisms may
nevertheless involve various, less stable intermediate phases. In
the course of this study we have solved, from powder diffraction
data, the crystal structure of the nicotinamide/anthranilic acid
(naana) cocrystal (Figure 1c) and revised the crystal structure
of the salicylic acid/anthranilic acid (salana) cocrystal.
RESULTS AND DISCUSSION
■
Cocrystal Synthesis. Anthranilic acid is known to form
three polymorphs.47 The one we obtained commercially and
B
DOI: 10.1021/acs.cgd.7b01512
Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design
Article
Figure 2. PXRD and Raman monitoring quantitative reaction profiles for (a, b) naana and (c, d) nasal formation.
of ana to the mechanochemically prepared nasal cocrystal. As
theory predicted, we found that ana is not able to replace sal
from the nasal cocrystal, but that ana is slowly becoming
amorphous during milling, evidenced by a gradual decrease in
the scale factor of ana derived from Rietveld refinement
(Figures S1 and S2). On the other hand, sal readily replaced
ana from the naana cocrystal in a reaction mechanism that
involved intermediate formation of salana (Figures 3 and S3).
The replaced ana from the naana cocrystal crystallized
concomitantly as a pure phase, but in a different polymorphic
form than the starting ana. The in situ formed ana belongs to
the monoclinic polymorph III of ana (ana-III), which no longer
has zwitterionic molecules.47 Formation of ana-III is surprising,
since it was previously found to transform to ana-I upon
grinding47 and should be less stable than ana-I (Scheme 1).
Formation of ana-III provides further evidence that unusual
and metastable solid species may be formed in a milled reaction
environment. While a full investigation as to why the less stable
polymorph ana-III was formed here is outside the scope of this
work, we may offer possible explanations to its unexpected
formation. Nucleation and growth of ana-III during milling may
be in accordance with the Ostwald’s rule of stages, but we have
not observed its transformation to the stable ana-I in the period
Scheme 1. Theoretical Prediction of the Order of Stability of
Cocrystals and Competing Coformers in the System of na,
sal, and ana
a
a
Two polymorphs of ana are included, ana-I and ana polymorph III
(ana-III).47
For replacement experiments, a competitive cocrystal
coformer was added to the mechanochemically prepared
naana or nasal cocrystals. Thus, we have added an equimolar
amount of sal to the reaction vessel containing the naana
cocrystal and, in the second experiment, an equimolar amount
C
DOI: 10.1021/acs.cgd.7b01512
Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design
Article
become zwitterionic. Since determination of hydrogen atoms
from powder diffraction data may be unreliable, we have aided
ourselves with ab initio calculations to rank in energy these two
crystal structure candidates and found the second, zwitterionic
cocrystal option, to be slightly more stable. Final Rietveld
refinement was performed against high-resolution powder X-
ray diffraction data, collected at the powder diffraction beamline
11-BM at the Advanced Photon Source, to reveal a chemically
reasonable network of hydrogen bonds in the zwitterionic
salana cocrystal (Figures 4 and S7).
Figure 4. Hydrogen bonding in the crystal structure of the zwitterionic
salana cocrystal. Hydrogen bonds are denoted with orange dashed
lines. (a) Three hydrogen bonds surrounding the protonated amino
group of ana and (b) the hydrogen bond between the carboxylic group
of sal and the carboxylate group of ana.
With crystal structures of all participating phases now known,
Rietveld refinement enabled extraction of reaction profiles
which reveal immediate and concomitant formation of nasal
and salana cocrystal at approximately the same rates in the
beginning of the reaction, while sal is present in the reaction
mixture (Figure 3). Later, formation of salana gradually slows
down and stops just at the time pure sal is fully consumed when
it becomes the source of salicylic acid needed for further
formation of the nasal cocrystal. The crystallization of pure ana
then also accelerates, but, as mentioned above, it crystallizes as
polymorph III, rather than polymorph I. In situ Raman
monitoring revealed essentially the same reaction course but
suggests a significantly longer presence of naana and also salana,
which could possibly be amorphous. Worth mentioning, in this
reaction sequence, the amino group of ana underwent several
proton transfers: from the starting zwitterionic ana-I, to the
nonzwitterionic form in the naana cocrystal, followed again by
protonation in the salana cocrystal and final crystallization of a
nonzwitterionic ana-III (Scheme 2).
Competitive Experiments. Next to replacement reactions,
we have performed direct competitive experiments in mixtures
of na, ana-I, and sal (Figure 5). Milling equimolar amounts of
na, sal, and ana-I resulted in the final formation of the nasal
cocrystal. The reaction course, however, included intermediate
formation of the naana cocrystal and formation of a small
amount of the salana cocrystal which remained in the final
reaction mixture. Rietveld analysis of the reaction course
exhibits an initial decline in the weight fraction of ana-I
consistent with the formation of a small amount of the naana
intermediaate. As naana starts to diminish after 2 min milling,
the weight fraction of ana-I increases, as the released ana
crystallizes into the starting polymorph I, before it is again
consumed for the formation of salana. The rate of na
consumption is slightly faster than the rate of consumption of
sal, consistent with parallel formation of nasal and naana.
Raman monitoring revealed the same reaction mechanism with
Figure 3. Reaction profiles for the mechanochemical reaction of the
naana cocrystal with sal, derived from time-resolved tandem in situ
monitoring by (a) PXRD and (b) Raman spectroscopy. Error bars
correspond to standard deviations as obtained from the corresponding
refinement procedures.
of 25 min milling. Formation of ana-III could be stochastic, as
was e.g., the case with the formation of the katsenite metal−
organic framework,33 but this is impossible to judge since the
experiment was repeated only once. Finally, a likely explanation
for the occurrence of ana-III involves stabilization of its
crystallite particles via surface interaction with the species
present in the reaction mixture, as was demonstrated by
Belenguer et al.41 and by us29 when using liquid additives.
The crystal structure of the salana cocrystal, observed here as
an intermediate, was recently solved as a cocrystal of sal and ana
molecules.45 However, we have revised here the crystal
structure of salana with the amino group of ana protonated
in order to form three N−H···O hydrogen bonds. The
protonated amino group leaves two options for the position
of one hydrogen atom: either on the carboxylic group of sal or
on the carboxylic group of ana. In the case of sal deprotonation,
a salt would be formed, while in the case the carboxylic group
of ana would be deprotonated, a zwitterionic cocrystal would
form where sal would remain nonzwitterionic and ana would
D
DOI: 10.1021/acs.cgd.7b01512
Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design
Article
a
Scheme 2. Protonation−Deprotonation Sequence of ana in the (na + ana) + sal Reaction Sequence
a
Polymorph designation is given according to Etter.47
Figure 5. Competitive mechanochemical reaction of na, sal, and ana in the molar ratio 1:1:1. Tandem in situ monitoring via (a, b) PXRD and (c, d)
Raman spectroscopy. (a) Two-dimensional time-resolved PXRD and (b) the corresponding reaction profile. (c) Two-dimensional time-resolved
Raman spectra and (d) the corresponding reaction profile. Error bars correspond to standard deviations as obtained from the corresponding
refinement procedures.
E
DOI: 10.1021/acs.cgd.7b01512
Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design
Article
Figure 6. Competitive mechanochemical reaction of (a) na, sal, and ana in the molar ratio 1:1/2:1/2 monitored by tandem in situ (a, b) PXRD and
(c, d) Raman spectroscopy. (a) Two-dimensional time-resolved PXRD and (b) the corresponding reaction profile. (c) Two-dimensional time-
resolved Raman spectra and (d) the corresponding reaction profile. The reaction profiles in (b) and (d) exhibit a slight increase in the content of
ana-I which is likely due to the reaction being non-homogeneous immediately after initiation of milling and thus would not correspond to an actual
increase in the content of ana-I in the reaction mixture. Error bars correspond to standard deviations as obtained from the corresponding refinement
procedures.
naana formation and a small amount of salana in the final
mixture. Consumption of na and sal and the formation of nasal
followed the same rates and trends in both PXRD and Raman
monitoring indicating that nasal is highly crystalline. Formation
of salana starts just after 3 min milling according to both Raman
and PXRD. naana is slightly more persistent according to
Raman than according to PXRD indicating the presence of
amorphous or nanocrystalline naana undetectable by PXRD.
A limited study of LAG reactions was here performed where
we have used methanol or propanol as liquid additives in the
na:sal:ana = 1:1:1 reaction mixtures (Figures S4 and S5).
Alcohols accelerated the reactions and nasal formation was
finished within 5 min milling. With methanol, we observed
intermediate formation of naana and the formation of a small
amount of salana, which remained upon further milling. Using
propanol as the liquid additive resulted in a similar reaction
mechanism, but without the formation of the intermediate
naana. These results strongly indicate the sensitivity of
mechanochemical reactions to additives and the possibility to
alter relative energies of solid species and barriers between
them and thus support the concept of the energy landscape in a
mechanochemical reaction environment.
Milling na, sal, and ana but now in the ratio of 1:1/2:1/2
resulted in surprisingly different reactivity as compared to the
1:1:1 mixture. The reaction mixture in this experiment
contained sufficient na to fully form both nasal and naana
cocrystals. Comparing PXRD and Raman monitoring (Figure
6), there are several significant similarities and differences: (i)
Formation profiles of nasal are similar in PXRD and Raman
monitoring, while the formation profiles of naana are
F
DOI: 10.1021/acs.cgd.7b01512
Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design
Article
significantly different. (ii) According to PXRD monitoring, the
formation of naana did not commence until crystalline sal was
fully consumed and nasal formation was complete. On the
other hand, Raman monitoring revealed naana formation soon
after nasal started to form. (iii) Formation curve of naana
derived from Raman monitoring is erratic, while it is smooth
according to PXRD. (iv) On the basis of the initial
stoichiometric ratio of reactants and the similarity of molecular
masses of nasal and naana cocrystals, final weight fractions of
nasal and naana should be similar and around 50%. However,
PXRD revealed weight fraction of nasal to be around 55% and
weight fraction of naana around 40% (the remaining 5% belong
to minor amounts of reactants). On the other hand, Raman
monitoring revealed, as expected, equal amounts of nasal and
naana at the end of milling.
There are several possibilities to the observed discrepancy
between in situ PXRD and Raman monitoring. It may be due to
partial amorphization of naana which would result in apparent
lowering of its weight fraction in Rietveld analysis, but it does
not explain why naana formation is not observable by PXRD
until all sal is consumed. Another possibility involves formation
of a solid solution of naana in the nasal cocrystal which would
increase weight fraction of nasal at the expense of naana, but
this explanation assumes that naana in a solid solution in nasal
would have the same Raman spectrum as pure ana, which may
be an unlikely assumption. It is also likely that naana crystal
growth is hindered as long as sal is present in the reaction
mixture, consistent with the replacement reaction of naana with
sal. Since Raman monitoring indicates presence of naana, it can
only be amorphous or nanocrystalline and thus not detectable
by X-ray diffraction. This is supported by a small decrease in
ana weight fraction before the onset of naana crystallization.
Only when sal is fully consumed, growth of microcrystalline
naana can proceed when it can be detected by PXRD. This last
possibility, however, does not explain the discrepancy in final
weight fractions of nasal and naana.
detect a reaction occurring in an amorphous component of the
reaction mixture. We therefore expect Raman monitoring to
become an indispensable tool in understanding the dynamics of
the mechanochemical reaction environment.
EXPERIMENTAL SECTION
■
Mechanochemical reactions were carried out as described previously29
using a Retsch MM301 (Germany) mixer mill operating at 30 Hz.
Translucent and amorphous reaction vessels made from polymethyl-
metacrylate (PMMA) had the internal volume of 14 mL and were
purchased from InSolido Technologies (Croatia). Two halves of the
vessel snapped upon closure to form a leak-proof seal. Liquids in
liquid-assisted grinding reactions were added using a Gilson automated
micropipet. Tandem in situ monitoring experiments were conducted at
the ESRF beamline ID31 as described previously.29 Briefly, the
experimental hutch was air-conditioned to 20 °C, the X-ray beam and
the Raman laser focus were positioned to approximately coincide on
the same portion of the reaction mixture, X-ray radiation wavelength
of 0.195 Å was selected using a multilayer monochromator, diffraction
data were recorded on a Dectris Pilatus CdTe 2 M detector positioned
1067 mm from the sample, radial integration of the raw diffraction
images was performed using PyFAI,59 and exposure time for each
pattern was 5.0 s. Time resolution between consecutive diffraction
patterns was ca. 6.5 s, and time resolution of Raman spectra was
typically 10 s. As milling media, two 7 mm stainless steel balls were
used, each weighing 1.4 g.
Crystal structure of naana was solved from powder X-ray diffraction
data collected on a laboratory Panalytical instrument in the Bragg−
Brentano geometry. Structure was solved by simulated annealing in
direct space using known molecular fragments of na and ana, which
were treated as rigid bodies. Crystal structure solution was recognized
when a meaningful hydrogen bond was assembled between molecular
fragments of na and ana and with no close contacts between
nonbonded atoms. The structure model was finally refined treating
molecules of na and ana as rigid bodies. The crystal structure of salana
was previously reported as a cocrystal of nonzwitterionic molecules of
sal and ana45 and is here revised as a zwitterionic cocrystal where the
ana molecule in the cocrystal is zwitterionic. The new structure model
of salana cocrystal was refined against high-resolution synchrotron
powder diffraction data, collected at the 11-BM beamline of the
Advanced Photon Source, using restraints on bond distances and
angles as well as planarity restraints. All calculations were performed
using the program Topas. Crystal structures of salana and naana are
deposited with the Cambridge Crystallographic Data Center (CCDC)
under deposition numbers 1581782 and 1581783 and can be retrieved
from CCDC upon request.
CONCLUSION
■
We have described mechanochemical transformations and
selectivity in a system of nicotinamide cocrystals with two
similar aromatic carboxylic acids: salicylic acid and anthranilic
acid. Cocrystals show a distinct order of selectivity and a less
preferred one is readily converted to a preferred cocrystal when
the competing coformer is offered. Among various approaches
to explaining the order of selectivity, we find theoretical energy
determination to be advantageous. With the advance in
computational techniques and the availability of high-perform-
ance computers, we believe it is no longer necessary to resort to
qualitative approaches, such as Etter’s rules,55,56 which in this
case, would suggest opposite reactivity. Theoretical calculations
will likely provide an order of stability with a high predictive
power, but we should nevertheless remain aware that in milling
the real sample need not conform to perfect crystals,57,58 which
are taken into theoretical predictions, but have defects, a
distribution of particle sizes, and could easily contain an
amorphous component.
Quantitative Rietveld refinement60 was performed on a series of in
situ collected powder diffraction patterns in an automated fashion
using batch files and the command-line version of Topas, either always
starting from the same input file or in a sequential manner.26,29
Rietveld refinements included refinement of the parameters for the
shifted Chebyshev polynomial used to describe the background
parameters, parameters contributing the peak position and shape
(contribution to the Lorenzian and Gaussian full widths at half-
maximum, zero shift, unit cell parameters). No instrument
contribution to peak shape was assumed. Upon convergence of
Rietveld refinement for each pattern, the obtained relative crystalline
phase weight fractions were output to a separate file and used for
further plotting. All plots were created using the program Mathematica
with the help of the SciDraw package.61 For two-dimensional time
-resolved plots, the background of each diffraction pattern was
subtracted prior to plotting using the Sonneveld-Visser62 algorithm
implemented in Mathematica. Spectral range of 316−1713 cm−1 of
Raman spectra were taken for analysis. Reaction vessel subtraction and
baseline subtraction in Raman spectra was performed as previously
described.29 Raman spectra of pure phases of na, sal, ana-I, ana-III,
nasal, naana, and salana were collected in the same experimental
conditions. ana-III was prepared as described elsewhere.47 Raman
This work has provided a clear example of different reaction
profiles derived from PXRD and Raman monitoring putting an
emphasis on the complementarity of the two techniques for in
situ and real-time monitoring of mechanochemical reactions.
While PXRD will detect bulk crystalline species, Raman
monitoring, which is more sensitive to the molecular structure
and the immediate molecular surroundings, will nevertheless
G
DOI: 10.1021/acs.cgd.7b01512
Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design
Article
reaction profiles were derived using nonnegative classic least-squares
approach implemented in MATLAB.63
(5) Stolle, A.; Szuppa, T.; Leonhardt, S. E. S.; Ondruschka, B. Ball
milling in organic synthesis: solutions and challenges. Chem. Soc. Rev.
2011, 40, 2317−2329.
DFT calculations were performed using a plane-wave basis set code
Quantum Espresso64 with GBRV pseudopotentials65 and vdW-DF-cx
exchange-correlation functional.51,52 The plane-wave basis set cutoff
was set to 820 eV, and the first Brillouin zone was sampled by a 3 × 3
× 3 Monkhorst−Pack k-point mesh. In each calculation, crystal lattice
and atom positions were relaxed until the change in the total energy
was <0.5 meV, all the forces were smaller than 0.01 eV/Å, and
pressure was <0.5 kbar.
(6) Kaupp, G. Organic solid-state reactions with 100% yield. Top.
Curr. Chem. 2005, 254, 95−183.
(7) Wang, G.-W. Mechanochemical organic synthesis. Chem. Soc. Rev.
2013, 42, 7668−7700.
́
(8) Hernandez, J. G.; Bolm, C. Altering Product Selectivity by
Mechanochemistry. J. Org. Chem. 2017, 82, 4007−4019.
́
̌
̌ ́ ́ ́
(9) Juribasic, M.; Uzarevic, K.; Gracin, D.; Curic, M. Mechanochem-
ical C-H bond activation: rapid and regioselective double cyclo-
palladation monitored by in situ Raman spectroscopy. Chem. Commun.
2014, 50, 10287−10290.
ASSOCIATED CONTENT
■
S
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.cgd.7b01512.
́
(10) Hernandez, J. G. C−H Bond Functionalization by Mechano-
chemistry. Chem. - Eur. J. 2017, 23, 17157−17165.
(11) Hermann, G. N.; Becker, P.; Bolm, C. Mechanochemical
Rhodium(III)-Catalyzed C−H Bond Functionalization of Acetanilides
under Solventless Conditions in a Ball Mill. Angew. Chem., Int. Ed.
2015, 54, 7414−7417.
PXRD and Raman spectra; Rietveld fit data; crystallo-
graphic ad refinement data (PDF)
(12) Pichon, A.; Lazuen-Garay, A.; James, S. L. Solvent-free synthesis
of a microporous metal-organic framework. CrystEngComm 2006, 8,
211−214.
Accession Codes
CCDC 1581782−1581783 contain the supplementary crystal-
lographic data for this paper. These data can be obtained free of
charge via www.ccdc.cam.ac.uk/data_request/cif, or by email-
ing data_request@ccdc.cam.ac.uk, or by contacting The
Cambridge Crystallographic Data Centre, 12 Union Road,
Cambridge CB2 1EZ, UK; fax: +44 1223 336033.
̌
(13) Stolar, T.; Batzdorf, L.; Lukin, S.; Zilic,
́ ̌ ̌ ́
D.; Motillo, C.; Friscic,
T.; Emmerling, F.; Halasz, I.; Uzarevic, K. In Situ Monitoring of the
́
̌
Mechanosynthesis of the Archetypal Metal−Organic Framework
HKUST-1: Effect of Liquid Additives on the Milling Reactivity.
Inorg. Chem. 2017, 56, 6599−6608.
(14) Uzarevic,
T.; Farha, O. K.; Frisci
̌ ̌ ́
́
K.; Wang, T. C.; Moon, S.-Y.; Fidelli, A. M.; Hupp, J.
c, T. Mechanochemical and solvent-free
̌
AUTHOR INFORMATION
■
assembly of zirconium-based metal-organic frameworks. Chem.
Commun. 2016, 52, 2133−2136.
Corresponding Author
*E-mail: ivan.halasz@irb.hr.
ORCID
(15) Braga, D.; Maini, L.; Grepioni, F. Mechanochemical preparation
of co-crystals. Chem. Soc. Rev. 2013, 42, 7638−7648.
(16) Hasa, D.; Schneider Rauber, G.; Voinovich, D.; Jones, W.
Cocrystal Formation through Mechanochemistry: from Neat and
Liquid-Assisted Grinding to Polymer-Assisted Grinding. Angew. Chem.,
Int. Ed. 2015, 54, 7371−7375.
Tomislav Stolar: 0000-0002-9824-4462
Krunoslav Uzarevic: 0000-0002-7513-6485
̌
́
Ivan Halasz: 0000-0002-5248-4217
Notes
(17) Xu, C.; De, S.; Balu, A. M.; Ojeda, M.; Luque, R.
Mechanochemical synthesis of advanced nanomaterials for catalytic
applications. Chem. Commun. 2015, 51, 6698−6713.
The authors declare no competing financial interest.
(18) Icļ i, B.; Christinat, N.; Tonnemann, J.; Schuttler, C.; Scopelliti,
̈
ACKNOWLEDGMENTS
■
R.; Severin, K. Synthesis of Molecular Nanostructures by Multi-
component Condensation Reactions in a Ball Mill. J. Am. Chem. Soc.
2009, 131, 3154−3155.
Financial support from the Croatian Science Foundation
(Grant No. UIP-2014-09-4744) is gratefully acknowledged.
I.H. is grateful to the Adris foundation for supporting this work.
S.L. is supported by the Croatian Science Foundation. I.L. and
P.L. are supported by the Unity Through Knowledge Fund,
Contract No. 22/15 and the H2020 CSA Twinning Project No.
692194, RBI-T-WINNING. We are grateful to the team at the
̌ ̌ ́
(19) Tan, D.; Loots, L.; Friscic, T. Towards medicinal mecha-
nochemistry: evolution of milling from pharmaceutical solid form
screening to the synthesis of active pharmaceutical ingredients (APIs).
Chem. Commun. 2016, 52, 7760−7781.
(20) Bonnamour, J.; Metro, T.-X.; Martinez, J.; Lamaty, F.
Environmentally benign peptide synthesis using liquid-assisted ball-
milling: application to the synthesis of Leu-enkephalin. Green Chem.
2013, 15, 1116−1120.
̌ ́
fine-mechanics workshop of the Ruđer Boskovic Institute for
their continuous support and the staff of the 11-BM beamline
for powder data collection of the salana cocrystal. Use of the
Advanced Photon Source at Argonne National Laboratory was
supported by the U.S. Department of Energy, Office of Science,
Office of Basic Energy Sciences, under Contract No. DE-AC02-
06CH11357.
(21) Konnert, L.; Gauliard, A.; Lamaty, F.; Martinez, J.; Colacino, E.
Solventless Synthesis of N-Protected Amino Acids in a Ball Mill. ACS
Sustainable Chem. Eng. 2013, 1, 1186−1191.
(22) Uzarevic,
́
K.; Halasz, I.; Đilovic,
́ ́ ̌ ́
I.; Bregovic, N.; Rubcic, M.;
̌
̌
Matkovic-
́
Calogovic, D.; Tomisic, V. Dynamic Molecular Recognition
́
̌
́
in Solid State for Separating Mixtures of Isomeric Dicarboxylic Acids.
Angew. Chem., Int. Ed. 2013, 52, 5504−5508.
REFERENCES
■
(23) Hernandez, J. G.; Ardila-Fierro, K. J.; Crawford, D.; James, S. L.;
Bolm, C. Mechanoenzymatic peptide and amide bond formation.
Green Chem. 2017, 19, 2620−2625.
(1) James, S. L.; et al. Mechanochemistry: opportunities for new and
cleaner synthesis. Chem. Soc. Rev. 2012, 41, 413−447.
(2) Kaupp, G.; Schmeyers, J.; Boy, J. Waste-free solid-state syntheses
with quantitative yield. Chemosphere 2001, 43, 55−61.
́
(24) Perez-Venegas, M.; Reyes-Rangel, G.; Neri, A.; Escalante, J.;
Juaristi, E. Mechanochemical enzymatic resolution of N-benzylated-β3-
amino esters. Beilstein J. Org. Chem. 2017, 13, 1728−1734.
̌ ̌ ́
(3) Do, J.-L.; Friscic, T. Chemistry 2.0: Developing a New, Solvent-
Free System of Chemical Synthesis Based on Mechanochemistry.
Synlett 2017, 28, 2066−2092.
(25) Frisc
̌
i
̌
c,
́
T.; Halasz, I.; Beldon, P. A.; Belenguer, A. M.; Adams,
F.; Kimber, S. A. J.; Honkimaki, V.; Dinnebier, R. E. Nat. Chem. 2013,
(4) Takacs, L. The historical development of mechanochemistry.
Chem. Soc. Rev. 2013, 42, 7649−7659.
̈
5, 66−73.
H
DOI: 10.1021/acs.cgd.7b01512
Cryst. Growth Des. XXXX, XXX, XXX−XXX

Crystal Growth & Design
Article
(26) Halasz, I.; Frisc
Mottillo, C.; Julien, P.; Strukil, V.; Honkimaki, V.; Dinnebier, R. E.
Quantitative in situ and real-time monitoring of mechanochemical
̌
i
̌
c,
́
T.; Kimber, S. A. J.; Uzarevic,
́
K.; Pusk
̌
aric,
́
A.;
(46) Berry, D. J.; Seaton, C. C.; Clegg, W.; Harrington, R. W.; Coles,
S. J.; Horton, P. N.; Hursthouse, M. B.; Storey, R.; Jones, W.; Friscic,
̌ ̌ ́
̌
̌
̈
T.; Blagden, N. Applying Hot-Stage Microscopy to Co-Crystal
Screening: A Study of Nicotinamide with Seven Active Pharmaceutical
Ingredients. Cryst. Growth Des. 2008, 8, 1697−1712.
(47) Ojala, W. H.; Etter, M. C. Polymorphism in anthranilic acid: a
reexamination of the phase transitions. J. Am. Chem. Soc. 1992, 114,
10288−10293.
reactions. Faraday Discuss. 2014, 170, 203−221.
̌
(27) Gracin, D.; Strukil, V.; Frisc
̌
i
̌
c,
́
T.; Halasz, I.; Uzarevic,
́
K.
̌
Laboratory Real-Time and In Situ Monitoring of Mechanochemical
Milling Reactions by Raman Spectroscopy. Angew. Chem., Int. Ed.
2014, 53, 6193−6197.
̌
(48) Bucar, D.-K.; Day, G. M.; Halasz, I.; Zhang, G. G. Z.; Sander, J.
(28) Batzdorf, L.; Fischer, F.; Wilke, M.; Wenzel, K.-J. r.; Emmerling,
F. Direct In Situ Investigation of Milling Reactions Using Combined
X-ray Diffraction and Raman Spectroscopy. Angew. Chem., Int. Ed.
2015, 54, 1799−1802.
R. G.; Reid, D. G.; MacGillivray, L. R.; Duer, M. J.; Jones, W. The
curious case of (caffeine)·(benzoic acid): how heteronuclear seeding
allowed the formation of an elusive cocrystal. Chem. Sci. 2013, 4,
4417−4425.
(29) Lukin, S.; Stolar, T.; Tireli, M.; Blanco, M. V.; Babic,
́ ̌ ̌ ́
D.; Friscic,
(49) Lonca
̌
ric,
́ ́ ̌ ́
I.; Popovic, J.; Despoja, V.; Burazer, S.; Grgicevic, I.;
T.; Uzarevic, K.; Halasz, I. Tandem In Situ Monitoring for
́
̌
̌
Popovic,
́
D.; Skoko, Z. Reversible Thermosalient Effect of N′-2-
Quantitative Assessment of Mechanochemical Reactions Involving
Structurally Unknown Phases. Chem. - Eur. J. 2017, 23, 13941−13949.
(30) Jansen, M.; Doll, K.; Schoen, J. C. Addressing chemical diversity
by employing the energy landscape concept. Acta Crystallogr., Sect. A:
Found. Crystallogr. 2010, 66, 518−534.
Propylidene-4-hydroxybenzohydrazide Accompanied by an Immense
Negative Compressibility: Structural and Theoretical Arguments
Aiming toward the Elucidation of Jumping Phenomenon. Cryst.
Growth Des. 2017, 17, 4445−4453.
(50) Reilly, A. M.; et al. Report on the sixth blind test of organic
crystal structure prediction methods. Acta Crystallogr., Sect. B: Struct.
Sci., Cryst. Eng. Mater. 2016, 72, 439−459.
(31) Jansen, M. The energy landscape concept and its implications
for synthesis planning. Pure Appl. Chem. 2014, 86, 883−898.
(32) Andersen, J. M.; Mack, J. Decoupling the Arrhenius equation via
mechanochemistry. Chem. Sci. 2017, 8, 5447−5453.
(51) Berland, K.; Hyldgaard, P. Exchange functional that tests the
robustness of the plasmon description of the van der Waals density
functional. Phys. Rev. B: Condens. Matter Mater. Phys. 2014, 89, 035412.
(52) Berland, K.; Arter, C. A.; Cooper, V. R.; Lee, K.; Lundqvist, B. I.;
Schroder, E.; Thonhauser, T.; Hyldgaard, P. van der Waals density
functionals built upon the electron-gas tradition: Facing the challenge
of competing interactions. J. Chem. Phys. 2014, 140, 18A539.
(53) Brown-Altvater, F.; Rangel, T.; Neaton, J. B. Ab initio phonon
dispersion in crystalline naphthalene using van der Waals density
functionals. Phys. Rev. B: Condens. Matter Mater. Phys. 2016, 93,
195206.
̌
(33) Katsenis, A. D.; Puskaric,
A.; Uzarevic, M. H.; Pham, K.; Do, T. O.; Kimber, S. A. J.; Lazic,
Magdysyuk, O.; Dinnebier, R. E.; Halasz, I.; Frisc c, T. Nat. Commun.
2015, 6, 6662.
̌ ̌ ́
(34) Belenguer, A. M.; Friscic, T.; Day, G. M.; Sanders, J. K. M.
́
A.; Strukil, V.; Mottillo, C.; Julien, P.
́
́
P.;
̌
̌ ̌ ́
i
Solid-state dynamic combinatorial chemistry: reversibility and
thermodynamic product selection in covalent mechanosynthesis.
Chem. Sci. 2011, 2, 696−700.
(35) Bygrave, P. J.; Case, D. H.; Day, G. M. Is the equilibrium
composition of mechanochemical reactions predictable using
computational chemistry? Faraday Discuss. 2014, 170, 41−57.
(36) Belenguer, A. M.; Lampronti, G. I.; Wales, D. J.; Sanders, J. K.
M. Direct Observation of Intermediates in a Thermodynamically
Controlled Solid-State Dynamic Covalent Reaction. J. Am. Chem. Soc.
2014, 136, 16156−16166.
(54) Rangel, T.; Berland, K.; Sharifzadeh, S.; Brown-Altvater, F.; Lee,
K.; Hyldgaard, P.; Kronik, L.; Neaton, J. B. Structural and excited-state
properties of oligoacene crystals from first principles. Phys. Rev. B:
Condens. Matter Mater. Phys. 2016, 93, 115206.
(55) Taylor, R.; Kennard, O. Hydrogen-bond geometry in organic
crystals. Acc. Chem. Res. 1984, 17, 320−326.
(56) Etter, M. C. Encoding and decoding hydrogen-bond patterns of
organic compounds. Acc. Chem. Res. 1990, 23, 120−126.
(57) Halasz, I. Single-Crystal-to-Single-Crystal Reactivity: Gray,
Rather than Black or White. Cryst. Growth Des. 2010, 10, 2817−2823.
(58) Coquerel, G. The structural purity of molecular solids−An
elusive concept? Chem. Eng. Process. 2006, 45, 857−862.
(59) Kieffer, J.; Karkoulis, D. PyFAI, a versatile library for azimuthal
regrouping. J. Phys.: Conf. Ser. 2013, 425, 202012.
(60) Hill, R. J.; Howard, C. J. Quantitative phase analysis from
neutron powder diffraction data using the Rietveld method. J. Appl.
Crystallogr. 1987, 20, 467−474.
(61) Caprio, M. LevelScheme: A level scheme drawing and scientific
figure preparation system for Mathematica. Comput. Phys. Commun.
2005, 171, 107−118.
(62) Sonneveld, E. J.; Visser, J. W. Automatic collection of powder
data from photographs. J. Appl. Crystallogr. 1975, 8, 1−7.
(63) MATLAB, version R2016a; The MathWorks Inc.: Natick, MA,
2016.
(64) Giannozzi, P.; et al. QUANTUM ESPRESSO: a modular and
open-source software project for quantum simulations of materials. J.
Phys.: Condens. Matter 2009, 21, 395502.
̌
(37) Uzarevic,
̌ ̌ ́
́
K.; Strukil, V.; Mottillo, C.; Julien, P. A.; Pusk
̌ ́
aric, A.;
̌
Friscic, T.; Halasz, I. Exploring the Effect of Temperature on a
Mechanochemical Reaction by in Situ Synchrotron Powder X-ray
Diffraction. Cryst. Growth Des. 2016, 16, 2342−2347.
(38) Shan, N.; Toda, F.; Jones, W. Mechanochemistry and co-crystal
formation: effect of solvent on reaction kinetics. Chem. Commun. 2002,
2372−2373.
(39) Halasz, I.; Pusk
̌
aric,
́
A.; Kimber, S. A. J.; Beldon, P. J.; Belenguer,
A. M.; Adams, F.; Honkimaki, V.; Dinnebier, R. E.; Patel, B.; Jones,
̈
̌
W.; Strukil, V.; Frisc
̌
i
̌
c,
́
T. Angew. Chem., Int. Ed. 2013, 52, 11538−
11541.
(40) Tireli, M.; Juribasi
Biliskov, N.; Borovina, M.; Curic,
Commun. 2015, 51, 8058−8061.
̌ ́
c Kulcsar, M.; Cindro, N.; Gracin, D.;
́
̌
̌
́ ́
M.; Halasz, I.; Uzarevic, K. Chem.
(41) Belenguer, A. M.; Lampronti, G. I.; Cruz-Cabeza, A. J.; Hunter,
C. A.; Sanders, J. K. M. Solvation and surface effects on polymorph
stabilities at the nanoscale. Chem. Sci. 2016, 7, 6617−6627.
(42) Etter, M. C.; Reutzel, S. M.; Choo, C. G. Self-organization of
adenine and thymine in the solid state. J. Am. Chem. Soc. 1993, 115,
4411−4412.
(43) Caira, M. R.; Nassimbeni, L. R.; Wildervanck, A. F. Selective
formation of hydrogen bonded cocrystals between a sulfonamide and
aromatic carboxylic acids in the solid state. J. Chem. Soc., Perkin Trans.
2 1995, 2, 2213−2216.
(44) Abourahma, H.; Urban, J. M.; Morozowich, N.; Chan, B.
Examining the robustness of a theophylline cocrystal during grinding
with additives. CrystEngComm 2012, 14, 6163−6169.
(45) Fischer, F.; Joester, M.; Rademann, K.; Emmerling, F. Survival
of the Fittest: Competitive Co-crystal Reactions in the Ball Mill. Chem.
- Eur. J. 2015, 21, 14969−14974.
(65) Garrity, K. F.; Bennett, J. W.; Rabe, K. M.; Vanderbilt, D.
Pseudopotentials for high-throughput DFT calculations. Comput.
Mater. Sci. 2014, 81, 446−452.
I
DOI: 10.1021/acs.cgd.7b01512
Cryst. Growth Des. XXXX, XXX, XXX−XXX