Nonequimolar Mixture of Organic Acids and Bases ... - ACS Publications

Jul 11, 2016 - Nonequimolar Mixture of Organic Acids and Bases: An Exception to the Rule of Thumb for Salt or Cocrystal. Saied Md Pratik and Ayan Datt...
0 downloads 0 Views 2MB Size
Download PDF
Article pubs.acs.org/JPCB

Nonequimolar Mixture of Organic Acids and Bases: An Exception to the Rule of Thumb for Salt or Cocrystal Saied Md Pratik and Ayan Datta* Department of Spectroscopy, Indian Association for the Cultivation of Science, Jadavpur-700032, West Bengal, India S Supporting Information *

ABSTRACT: Formation of salt and/or cocrystal from organic acid−base mixtures has significant consequences in the pharmaceutical industry and its related intellectual property rights (IPR). On the basis of calculations using periodic dispersion corrected DFT (DFT-D2) on formic acid−pyridine adduct, we have demonstrated that an equimolar stoichiometric ratio (1:1) exists as a neutral cocrystal. On the other hand, the nonequimolar stoichiometry (4:1) readily forms an ionic salt. While the former result is in agreement with the ΔpKa rule between the base and the acid, the latter is not. Calculations reveal that, within the equimolar manifold (n:n; n = 1−4), the mixture exists as a hydrogen bonded complex in a cocrystal-like environment. However, the nonequimolar mixture in a ratio of 5:1 and above readily forms salt-like structures. Because of the cooperative nature of hydrogen bonding, the strength of the O−H···N hydrogen bond increases and eventually transforms into O−···H−N+ (complete proton transfer) as the ratio of formic acid increases and forms salt as experimentally observed. Clearly, an enhanced polarization of formic acid on aggregation increases its acidity and, hence, facilitates its transfer to pyridine. Motion of the proton from formic acid to pyridine is shown to follow a relay mechanism wherein the proton that is far away from pyridine is ionized and is subsequently transferred to pyridine via hopping across the neutral formic acid molecules (Grotthuss type pathway). The dynamic nature of protons in the condensed phase is also evident for cocrystals as the barrier of intramolecular proton migration in formic acid (leading to tautomerism), ΔH⧧tautomer = 17.1 kcal/mol in the presence of pyridine is half of that in free formic acid (cf. ΔH⧧tautomer = 34.2 kcal/mol). We show that an acid−base reaction can be altered in the solid state to selectively form a cocrystal or salt depending on the strength and nature of aggregation.

■

INTRODUCTION Noncovalent interactions (NCI) such as hydrogen bonding, cation···π interactions, π···π interactions, dipole···dipole interactions, electrostatic interactions, and van der Waals interactions have received paramount attention, as they grant access to engineer the molecular arrangement in the solid state.1−7 Many such synthetically engineered materials like polymorphs, hydrates, salts, and cocrystals have gained attention due to their potential applications in photoconductivity, nonlinear optics, liquid crystalline materials, host−guest compounds, and metallo-organic frameworks (MOF).8−15 The formation of salts or cocrystals as a consequence of acid−base reaction in the solid state has gained significant attention in the pharmaceutical industry.16−20 The performance of the drugs with respect to solubility, pharmacokinetics, or physicochemical properties depends critically on its state of aggregation, namely, salt or cocrystal.21−24 In contrast to salts, which are composed of © 2016 American Chemical Society

ionizable multicomponents, cocrystals are generally composed of two or more neutral/nonionizable building blocks and can be identified (within the resolution limit of X-ray diffraction) by the absence of proton transfer from acid to base (Scheme 1). Nevertheless, both salt and cocrystal can be represented as two interlinked phases (states) that are connected through Scheme 1. 1:1 Adduct of Formic Acid and Pyridine Illustrating (a) Neutral Cocrystal and (b) Ionized Salt

Received: June 9, 2016 Revised: July 11, 2016 Published: July 11, 2016 7606

DOI: 10.1021/acs.jpcb.6b05830 J. Phys. Chem. B 2016, 120, 7606−7613

Article

The Journal of Physical Chemistry B

crystals of formic acid−pyridine, we have deciphered the origin of cocrystal → salt transformation in the latter case. Following a bottom up approach starting from 1:1 to 8:1 composition of formic acid and pyridine calculations reveal an increasing propensity of proton transfer at higher formic acid ratios >5:1, a direct consequence of cooperative enhancement of acidity in the solid state. The ΔpKa rule has been computationally substantiated for equimolar binary mixtures of formic acid− pyridine, formic acid−4-aminopyridine, and trifluoroacetic acid−4-aminopyridine. More importantly, we show a violation of the ΔpKa rule in the case of a nonequimolar mixture of formic acid−pyridine at higher stoichiometry. We propose that the selective formation of salt or cocrystal can be realized by changing the stoichiometry of the acids and bases notwithstanding their ΔpKa. The minimum energy pathway for motion of a proton from formic acid to pyridine in the 4:1 asymmetric unit of the salt is found to follow a relay mechanism wherein the proton hops from an initial formic acid to other neutral formic acid molecules before being transferred to pyridine. This results in a formate−pyridium salt complex which is H-bonded by three nondissociated formic acids within the crystal. Calculations reveal that, even within the 1:1 cocrystal form, the O−H···N hydrogen bonds are not static and tautomerization in formic acid is assisted by the presence of pyridine. Clearly, the nature of protons within crystals of organic acid− base mixtures is shown to be highly dynamic and additional care must be taken before classifying them into cocrystals or salts particularly when they are nonequimolar.

transition/migration of at least one proton. The formation of the cocrystal or salt is governed by the nature of synthons, categorized as homosynthons and heterosynthons depending on the type of molecular constituents.25 Using chemically tuned heterosynthons, one might control cocrystallization, and hence, their active pharmaceutical ingredient (API) properties can be improved.26 For the selective isolation of salts or cocrystals, an appropriate counterion is chosen on the basis of the ΔpKa [pKa(base) − pKa(acid)] criterion that has emerged as a useful guideline for the drug industry.27,28 It has been observed that, when the ΔpKa is large enough, salt formation is common, hence a “rule of thumb” stating that salts would be formed if ΔpKa [pKa(base) − pKa(acid)] ≥ 2−3 during a particular acid−base solid-state reaction.24 Nangia and co-workers have studied a series of salts and cocrystals formed between the pyridines and carboxylic acids and found that, when ΔpKa < 0, a cocrystal stabilized by O−H···N hydrogen bonding interaction is formed and, when ΔpKa > 3.75, salt formation is more likely.29 Although ΔpKa is found to be a reliable indicator to distinguish between salt and cocrystal, ambiguity exists in the intermediate region, 0 < ΔpKa > 3.75, for which prediction is complicated due to partial polarization of the O− H···N hydrogen bond.30 For such cases, crystallization might lead to salt, cocrystal, or disordered solid arising out of incomplete proton transfer depending upon the crystalline environment. Price and co-workers have studied a series of binary crystals of pyridine and derivatives with maleic, fumaric, phthalic, isophthalic, or terephthalic acids and successfully classified them into salts or cocrystals depending upon the ΔpKa.31 Analyzing ∼6465 crystalline ionized and nonionized acid−base complexes, Cruz-Cabeza has proposed a modified ΔpKa rule where the cocrystals are accessible when ΔpKa < −1 and salts are for ΔpKa > 4. Again, by contrast, in the region of −1 ≤ ΔpKa ≥ 4, salt and cocrystal coexist but the probability of salt formation increases as the ΔpKa increases.32 The binary mixture of formic acid and pyridine has been well-studied in the context of salt−cocrystal formation.33−35 As expected from the ΔpKa rule, Mootz et al. have reported that a 1:1 mixture of formic acid and pyridine interacts via a sufficiently strong O−H···N bond, resulting in cocrystal.36 On increasing the ratio of formic acid to pyridine to 3:1, this complex still remains as a neutral cocrystal assembly with a relatively stronger O−H···N hydrogen bond. Interestingly, however, on further increase in the ratio of formic acid to pyridine to 4:1, there is a sharp O−H···N → O−···H−N+ proton transition which leads to formation of the salt phase. It consists of one pyridinium cation, one formate anion along with three additional neutral formic acid molecules.36,37 Previously, in the context of inorganic acid−base reactions between NH3 and HX (X = F, Cl, Br, NO3, HCO3, and HSO4), we have also shown that the proton transfer from hydrogen bonded complex (NH3 ···HX) to ionic salt (NH 4 +X −) is facilitated via cooperativity by increasing their stoichiometric ratio.38 Hence, while the 2:1 mixture of NH3 and H2SO4 is a H-bonded complex (NH3···H2SO4···NH3, cocrystal-like), its higher analogue (4:2 mixture) is a salt, (NH4)2SO4. Therefore, the preference for a salt structure for the 4:1 formic acid−pyridine mixture might be a consequence of subtle cooperativity arising out of increasing formic acid stoichiometry rather than just the difference of the pKa of the acid and base.39,40 In this Article, on the basis of plane wave dispersioncorrected DFT (DFT-D2) calculations on the 1:1 and 4:1

■

COMPUTATIONAL SECTION Periodic DFT calculations for 1:1 cocrystal and 4:1 salt were performed using the PWSCF (plane-wave self consistent field) code with generalized gradient approximation (GGA) and Perdew−Burke−Ernzerhof (PBE) functional as implemented in the QUANTUM ESPRESSO Package.41,42 The ionic cores were described by an ultrasoft pseudopotential.43 The crystals were fully relaxed until all the forces became less than 10−3 a.u. using a 6 × 6 × 6 k-point mesh. The molecular entities, namely, pyridine, formic acid, pyridinium, and formate ions, were optimized within a 15 × 15 × 15 cubic box to create a gasphase-like environment with a 2 × 2 × 2 k-point mesh. A kinetic energy cutoff of 30 Ry was used for both of the situations. The dispersion interactions were taken into account using Grimme’s DFT-D2 empirical formalism.44 The formation energies of the cocrystal were calculated as ΔEcocrystal formation = Ecocrystal − m × Eformic acid − n × Epyridine and salt as ΔEsalt formation = Esalt − m × Eformic acid − n(Epyridinium + Eformate ion), where m and n represent the number of formic acids and number of pyridine/ pyridinium/formate, respectively, within the unit cell of the crystal. The presence of un-ionized formic acids within the 4:1 crystal is considered while computing the formation energy of the salt, ΔEsalt formation. All of the quantum chemical calculations using a localized Gaussian basis set were performed using the G09 suite of programs (version D.01).45 Harmonic frequencies were calculated for all minimized structures to ensure the absence of any first order saddle points. Truhlar’s hybrid metaGGA functional, M06-2X, which is well calibrated to account for middle range electron correlations is utilized.46 The M062X functional has been shown to be suitable for describing intramolecular dispersion forces like nonbonded interactions between closed shell atoms relevant for H-bonding interactions and also provides a reasonable balance between accuracy and computation resources. The Pople’s 6-31+G(d,p) split basis set 7607

DOI: 10.1021/acs.jpcb.6b05830 J. Phys. Chem. B 2016, 120, 7606−7613

Article

The Journal of Physical Chemistry B

Figure 1. Optimized crystal structures of (a) 1:1 cocrystal and (b) 4:1 salt. The right panel shows the molecular structures of their corresponding asymmetric units.

Figure 2. Optimized cocrystal-like H-bonded complexes of (a) 1:1, (b) 2:2, (c) 3:3, and (d) 4:4 formic acid−pyridine calculated at the M06-2X/631+G(d,p) level. The significant O−H···N and C−H···O hydrogen bonding distances are shown.

was used.47 Intrinsic reaction coordinate (IRC) calculations were performed to ensure correct transition state structures for the tautomers in formic acid−pyridine H-bonded structures. Basis set superposition error (BSSE) corrected binding energies of the formic acid−pyridine complexes and their respective salt forms using counterpoise (CP) correction.48 We have considered each molecule (formic acid and pyridine separately) as fragments within the H-bonded or proton transferred complexes for the CP corrections.

Files (CIFs) of the experimental structures.36 In both cases, the lattice parameters get contracted by only ∼3% upon optimization with respect to the experimental crystal structure, indicating robustness of our computations. The unit cell of the cocrystal consists of 4 pyridines and 4 formic acids in a ratio of 1:1, while the same in the case of salt consists of 4 pyridiniums, 4 formates, along with 12 un-ionized formic acids resulting in an overall 4:1 acid−base ratio. For the 1:1 case, formic acid interacts with pyridine via O−H···N hydrogen bond (dO−H···N = 1.55 Å) which leads to a neutral cocrystal of self-assembled pyridine and formic acid (Figure 1a). On the other hand, for the 4:1 salt crystal, one proton is transferred from formic acid to pyridine and leads to a pyridinium−formate pair along with three un-ionized formic acids. In the 4:1 salt, the formate anion

■

RESULTS AND DISCUSSION In Figure 1, the optimized unit cells of 1:1 cocrystal and 4:1 salt are shown along with their asymmetric units. The initial structures are retrieved from the Crystallographic Information 7608

DOI: 10.1021/acs.jpcb.6b05830 J. Phys. Chem. B 2016, 120, 7606−7613

Article

The Journal of Physical Chemistry B

O−H···O and C−H···O in the range 1.72−2.35 Å. Interestingly, though the 4:1 stoichiometry of formic acid to pyridine forms salt in the crystalline phase, the H-bonded complex is more stable (by 3.8 kcal/mol) than the proton-transferred complex, pyridinium···formate···3 HCOOH. However, both of these structures remain stable in their respective state of aggregation (with no imaginary frequencies), indicating the slow onset of the salt-like phase on and after 4:1 stoichiometry. This is in marked contrast to the lower stoichiometries for which any attempts to minimize the salt-like structures readily revert back to the H-bonded complexes. Further increasing the stoichiometry of formic acid from 5:1 to 8:1, the proton gets transferred from one of the formic acids to pyridine to finally form an ionized mixture of pyridinium cation and formate anion with the additional formic acids as observed in the crystal of the salt form. As shown in Figure 3d−g, the formate anion is associated via strong O−H···O hydrogen bonds (dO−H···O ∼ 1.40−1.65 Å) to surrounding formic acids. We have calculated the salt formation energy, ΔEsalt (defined as the energy difference between the salt-like form and the Hbonded complex at a particular stoichiometry), for the formic acid−pyridine complexes. ΔEsalt for 1:1, 2:1, 3:1, and 4:1 are found to be 19.3, 14.1, 11.0, and 3.8 kcal/mol, respectively, which indicates the instability of these systems toward proton transfer. Nevertheless, this instability decreases from 1:1 to 4:1 and the H-bonded and proton transferred complex become almost degenerate in the case of 4:1. This explains independent existence of the salt form as a local minimal structure discussed above. Further increase in formic acid ratio to 5:1, 6:1, 7:1, and 8:1 leads to crossover to a stable phase of salt with ΔEsalt of −11.0, −18.2, −19.0, and −18.4 kcal/mol, respectively. The variation ΔEsalt as a function of the number of formic acids (n) is shown in Figure 4a. There is a clear phase transition as evidenced from ΔEsalt as it switches its sign from positive (unstable) to negative (stable) on changing the stoichiometry from 4:1 to 5:1, and thus, it can be concluded that at the molecular level at least 5:1 stoichiometry is required to form salt-like structures. As reported in Table 2, the BSSE corrected energies (ΔE2) of the complexes gradually increase with the increase in the stoichiometry of formic acid. However, there is a sharp gain in ΔE2 from 4:1 to 5:1 stoichiometry as a consequence of salt formation. As clearly observed from Figure 2a and Figure 3a−c, the N··· H−O hydrogen bonds become stronger as the stoichiometry increases from 1:1 to 4:1 which is expressed in terms of the sequential shortening of the N···H distance. The dN...H decreases as 1.66 Å (1:1) → 1.59 Å (2:1) → 1.55 Å (3:1) → 1.50 Å (4:1) and eventually to ∼1.03 Å for 5:1 and onward, indicating complete proton transfer (as shown in Figure 4b). The increased propensity of proton transfer at higher formic acid stoichiometry is a direct consequence of the cooperative nature of the molecules in aggregates. The specific number and nature of noncovalent interactions becomes an important factor in dictating the salt or cocrystal phase for the hydrogen bonded complexes in the nonequimolar mixture of formic acid and pyridine and hence eventually for 4:1 crystal. ΔpKa between pyridine and formic acid is 1.46 and should lead to a cocrystal as indeed observed for the 1:1 stoichiometry. For other 1:1 mixtures like formic acid−4-aminopyridine and trifluoroacetic acid−4-aminopyridine which have ΔpKa of 5.42 and 8.65, respectively, we have also computationally studied their states of aggregation. Both of them are stabilized in a salt form as expected from the ΔpKa rule (see Figure S2 in the

is connected with three formic acids through three strong O− H···O hydrogen bonds, dO−H···O = 1.48, 1.51, 1.53, and 1.86 Å (right panel in Figure 1b). The formation energy (ΔEformation) per asymmetric unit of 1:1 cocrystal is found to be −33.8 kcal/ mol, while that for the 4:1 salt is −141.2 kcal/mol (−87.7 kcal/ mol based on comparison of the formation energy with respect to neutral constituents) which explains the stability of the cocrystal and salt in their respective state of aggregation. From a crystallographic viewpoint, at least a 4:1 stoichiometry of formic acid to pyridine is necessary to form salt. However, to understand the molecular mechanism of the aggregation induced proton transfer, a bottom-up approach is necessary. We have optimized structures starting from the simplest molecular stoichiometry of formic acid to pyridine from 1:1 to 4:4 in the equimolar mixture and from 2:1 to 8:1 for the nonequimolar mixture. In each case, we have considered several conformations/orientations and those reported here are the lowest energy ones.49 As shown in Figure 2, they remain as a hydrogen bonded complex in the cocrystal molecular motif via N···H−O and C−H···O hydrogen bonds. For the 1:1 Hbonded complex, dO−H···N = 1.66 Å and dC−H···O = 2.31 Å with a BSSE corrected binding energy of −13.1 kcal/mol which is in good agreement with the previously reported dO−H···N and dC−H···O of 1.71 and 2.46 Å, respectively, with a binding energy of −11.1 kcal/mol calculated at B3LYP/6-311++G(d,p).49 The formation of this hydrogen bonded complex is consistent with the large proton transfer energy (ΔEPT = +121.7 kcal/mol) calculated from the energetics of the process: formic acid + pyridine → pyridinium + formate ion. The small H-bonding stabilizing for the 1:1 complex does not bestow sufficient stability to a salt-like structure. Increasing the stoichiometry up to 4:4 does not trigger proton transfer. The average N···H−O and C−H···O H-bond distances are ∼1.60 and ∼2.40 Å, respectively, with the BSSE corrected binding energy being −33.7, −55.4, and −77.4 kcal/mol for 2:2, 3:3, and 4:4 complexes, respectively, as reported in Table 1. Thus, the Table 1. BSSE Corrected Binding Energy in Formic Acid− Pyridine Complexes at Various Stoichiometries Calculated at the M06-2X/6-31+G(d,p) Level of Theory stoichiometry of formic acid to pyridine

ΔE1 (in kcal/mol)

1:1 2:2 3:3 4:4

−13.1 −33.7 −55.4 −77.4

binding energy per pair of formic acid−pyridine complex increases from −13.1 kcal/mol → −16.9 kcal/mol → −18.5 kcal/mol → −19.3 kcal/mol for 1:1 → 2:2 → 3:3 → 4:4, respectively. Notwithstanding, the appreciable gain in binding energies on aggregation from cooperativity, the small binding energies per pair of the formic acid−pyridine in an equimolar mixture explains their inability to overcome ΔEPT even at the higher stoichiometry. Hence, the equimolar mixture of formic acid−pyridine remains as a hydrogen bonded complex as observed in experimental cocrystal form. On increasing the stoichiometry of formic acid to 2:1, 3:1, and 4:1, the nonequimolar mixture of the formic acid to pyridine still exists as the hydrogen bonded complex with a N··· H−O distance of 1.59, 1.55, and 1.50 Å, respectively, in the corresponding structures. Apart from the N···H−O hydrogen bond, they can have several other weak hydrogen bonds like 7609

DOI: 10.1021/acs.jpcb.6b05830 J. Phys. Chem. B 2016, 120, 7606−7613

Article

The Journal of Physical Chemistry B

Figure 3. Optimized geometries of cocrystal-like structures in (a) 2:1, (b) 3:1, and (c) 4:1 and salt-like structures in (d) 5:1, (e) 6:1, (f) 7:1, and (g) 8:1 stoichiometric ratios of formic acid−pyridine at the M06-2X/6-31+G(d,p) level of theory. The important H-bonding distances are also shown.

Figure 4. Variation of (a) salt formation energies (ΔEsalt) and (b) N−H bond distances (dN−H) as a function of the number of formic acids (n). For n < 5, dN−H corresponds to the O−H···N hydrogen bond distances, while, for n ≥ 5, it corresponds to the N−H covalent bond distance in pyridinium cation.

level both of these complexes require at least a 3:1 stoichiometry of acid to base for the formation of salt-like structures (see Figure S3 in the Supporting Information). The ΔEsalt values for formic acid−4-aminopyridine are 18.0, 12.7, and −2.1 kcal/mol in the case of 1:1, 2:1, and 3:1 ratios, respectively. For trifluoroacetic acid−4-aminopyridine, ΔEsalt = 10.3, 5.1, and −35.8 kcal/mol in 1:1, 2:1, and 3:1 stoichiometry, respectively. Hence, it can be concluded that, for lower stoichiometry, ΔpKa is sufficiently robust to explain formation of salt or cocrystal. Because of the weakly dissociating nature of formic acid, it should also exist as a cocrystal even at a 4:1 ratio. Therefore, the existence of the salt phase at this ratio is an anomaly. Hence, at increased ratio of formic acids, the “rule of thumb” is violated and ΔpKa does not act as an indicator for proton transfer. The additional stabilization of the N−H···O interaction arising out of cooperativity dominates over the individual pKa’s of formic acid and pyridine.

Table 2. BSSE Corrected Binding Energy (ΔE2) for 2:1 to 8:1 Formic Acid−Pyridine Adduct Calculated at the M062X/6-31+G(d,p) Level of theory stoichiometry of formic acid to pyridine

ΔE2 (kcal/mol)

2:1 3:1 4:1 5:1 6:1 7:1 8:1

−18.0 −27.1 −38.1 −177.1 −192.2 −190.7 −204.8

Supporting Information for their computed crystal structures). Any attempt to optimize the unit cells in cocrystal-like Hbonded structures immediately reverts back to the stable salt forms. Again it is important to note here that at the molecular 7610

DOI: 10.1021/acs.jpcb.6b05830 J. Phys. Chem. B 2016, 120, 7606−7613

Article

The Journal of Physical Chemistry B

Figure 5. Proton transfer pathway via relay mechanism in the 4:1 asymmetric unit of the salt phase of formic acid−pyridine crystal. The important hydrogen bonded distances are shown. The atoms directly involved in proton migration are indicated.

A critical examination of the asymmetric unit as shown Figure 1b shows that, for the 4:1 formic acid−pyridine salt, proton transfer from formic acid to pyridine is not vicinal which in fact is separated by one un-ionized formic acid. To get the insight of the proton transfer pathway (see Figure 5), the asymmetric unit of the salt as well as the 4:1 H-bonded complex is chosen for computations. In both cases, proton transfer requires the assistance of three neutral formic acids. During the course of proton transfer, two oxygen atoms of the formate anion are connected to Hb and Hc of the two nearest neighbor formic acids on either side with d(O1−Hb) = 1.12 Å, d(O2−Hb) = 1.29 Å, d(O3−Hc) = 1.17 Å, and d(O4−Hc) = 1.24 Å, respectively (Figure 5b). Following this, Hb and Hc are simultaneously transferred to O4 and O2, respectively (Figure 5c). Finally, Ha from the pyridinium ion is transferred to O4 with the assistance of the remaining formic acids to form a cocrystal-like molecular motif (Figure 5d). In the 4:1 H-bonded complex, the initial loss of a proton by formic acid due to proton abstraction by pyridine is sequentially compensated by their neighboring formic acids to ultimately form the pyridinium−formate pair as anticipated from the cooperative behavior of the molecules within the condensed phase. Clearly, proton transfer proceeds stepwise via a relay mechanism, though the direction of proton transfer is expectedly reversed in the H-bonded complex vis-à-vis the crystalline phase. Apart from the mechanism of intermolecular proton transfer between formic acid and pyridine, we have also studied the intramolecular hydrogen transfer in formic acid (tautomerization in formic acid) in the presence and absence of the H-bond acceptors like pyridine and 4-aminopyridine. Tautomerism in formic acid alone is unlikely as the barrier height is 37.2 kcal/ mol (ΔH⧧ = 34.2 kcal/mol) which is far too high to be achieved at room temperature. This is in agreement with previous reports.50 However, in the case of a 1:1 complex of formic acid−pyridine, the barrier height is reduced to 18.3 kcal/ mol (ΔH⧧ = 17.1 kcal/mol). In the presence of a stronger Hbond acceptor like 4-amino-pyridine, this barrier can be further reduced to 17.3 kcal/mol (ΔH⧧ = 16.2 kcal/mol). As shown in Figure 6, for both cases, the intramolecular proton transfer proceeds via a symmetric transition state. Therefore, while the intermolecular proton transfer leading to a salt-like structure for the 1:1 complex is not feasible yet, an intramolecular proton transfer dynamics resulting in tautomerization of formic acid is certainly possible. Such a large (∼50%) reduction in the activation energies for tautomerism in the presence of a favorable H-bond acceptor should result in sufficient mobility of the protons in formic acid even at lower stoichiometry. An experimental outcome of such an enhanced rate of tautomerism

Figure 6. Potential energy surface for intramolecular proton transfer in pyridine−formic acid complex calculated at the M06-2X/6-31+G(d,p) level of theory (black solid line: R1 = H, R2 = H; green dashed line: R1 = H, R2 = NH2).

in formic acid would be an elevated kH/kD isotope effect, since a lighter H atom (compared to D) would benefit additionally from quantum mechanical tunneling.51,52

■

CONCLUSION On the basis of periodic DFT calculations, we have revealed that a 1:1 mixture of formic acid and pyridine is stabilized as cocrystal while formic acid−4-aminopyridine and trifluoroacetic acid−4-aminopyridine prefer salt formation in harmony with the ΔpKa rule. For the 4:1 mixture, which forms a pair of pyridinium−formate salt capped by three un-ionized formic acid molecules in the asymmetric unit, cooperative enhancement of acidity of formic acid plays a decisive role. Such an unusual influence of polarization for higher stoichiometric mixtures of organic acid−organic base clearly cannot be predicted by just the difference of pKa’s between them. In fact, in a very recent report, Steed and co-workers reported salt formation in a 2:1 mixture of formic acid and pyridine under pressure.53 A molecular picture of the cocrystal → salt transformation shows that the strength of the O−H···N Hbond increases with an increase in the number of formic acids and a crossover to the salt form (O−H···N → O···H−N) occurs for a formic acid to pyridine ratio after 4:1. The proton migration pathway within the 4:1 crystal follows a relay mechanism wherein the proton hops from the donor (acid) to the acceptor (base) assisted cooperatively by the neutral acids. This work indicates that by means of cooperativity the outcome of an acid−base reaction in the solid state can be tuned to exclusively isolate a cocrystal or salt. This should have 7611

DOI: 10.1021/acs.jpcb.6b05830 J. Phys. Chem. B 2016, 120, 7606−7613

Article

The Journal of Physical Chemistry B

(11) Ohi, H.; Tachi, Y.; Itoh, S. Supramolecular and Coordination Polymer Complexes Supported by a Tripodal Tripyridine Ligand Containing a 1,3,5-Triethylbenzene Spacer. Inorg. Chem. 2004, 43, 4561−4563. (12) Pratik, S. M.; Nijamudheen, A.; Datta, A. Topochemical Transformations of CaX2 (X = C, Si, Ge) to Form Free-Standing TwoDimensional Materials. Chem. - Eur. J. 2015, 21, 18454−18460. (13) Huang, J.; Kertesz, M. Spin Crossover of Spiro-Biphenalenyl Neutral Radical Molecular Conductors. J. Am. Chem. Soc. 2003, 125, 13334−13335. (14) Shattock, T. R.; Arora, K. K.; Vishweshwar, P.; Zaworotko, M. J. Hierarchy of Supramolecular Synthons: Persistent Carboxylic Acid··· Pyridine Hydrogen Bonds in Cocrystals That Also Contain a Hydroxyl Moiety. Cryst. Growth Des. 2008, 8, 4533−4545. (15) Lemmerer, A.; Adsmond, D. a.; Esterhuysen, C.; Bernstein, J. Polymorphic Co-Crystals from Polymorphic Co-Crystal Formers: Competition between Carboxylic Acid···pyridine and Phenol··· pyridine Hydrogen Bonds. Cryst. Growth Des. 2013, 13, 3935−3952. (16) Berge, S. M.; Bighley, L. D.; Monkhouse, D. C. Pharmaceutical Salts. J. Pharm. Sci. 1977, 66, 1−19. (17) Shan, N.; Zaworotko, M. J. The Role of Cocrystals in Pharmaceutical Science. Drug Discovery Today 2008, 13, 440−446. (18) Vishweshwar, P.; McMahon, J. A.; Bis, J. A.; Zaworotko, M. J. Pharmaceutical Co-Crystals. J. Pharm. Sci. 2006, 95, 499−516. (19) Duggirala, N. K.; Perry, M. L.; Almarsson, O.; Zaworotko, M. J. Pharmaceutical Cocrystals: Along the Path to Improved Medicines. Chem. Commun. 2016, 52, 640−655. (20) Almarsson, O.; Zaworotko, M. J. Crystal Engineering of the Composition of Pharmaceutical Phases. Do Pharmaceutical CoCrystals Represent a New Path to Improved Medicines? Chem. Commun. 2004, 1889−1896. (21) Schultheiss, N.; Newman, A. Pharmaceutical Cocrystals and Their Physicochemical Properties. Cryst. Growth Des. 2009, 9, 2950− 2967. (22) Serajuddin, A. T. M. Salt Formation to Improve Drug Solubility. Adv. Drug Delivery Rev. 2007, 59, 603−616. (23) Good, D. J.; Rodríguez-Hornedo, N. Solubility Advantage of Pharmaceutical Cocrystals. Cryst. Growth Des. 2009, 9, 2252−2264. (24) Haynes, D. A.; Jones, W.; Motherwell, W. D. S. Cocrystallisation of Succinic and Fumaric Acids with Lutidines: A Systematic Study. CrystEngComm 2006, 8, 830−840. (25) Mukherjee, A.; Desiraju, G. R. Combinatorial Exploration of the Structural Landscape of Acid-Pyridine Cocrystals. Cryst. Growth Des. 2014, 14, 1375−1385. (26) Sudhakar, P.; Srivijaya, R.; Sreekanth, B. R.; Jayanthi, P. K.; Vishweshwar, P.; Babu, M. J.; Vyas, K.; Iqbal, J. Carboxylic AcidPyridine Supramolecular Heterocatemer in a Co-Crystal. J. Mol. Struct. 2008, 885, 45−49. (27) Handbook of pharmaceutical salts: properties, selection, and use; Stahl, P. H., Wermuth, C. G., Ed.; International Union of Pure and Applied Chemistry, VHCA; Wiley-VCH: Weinheim, New York, 2002. (28) Childs, S. L.; Stahly, G. P.; Park, A. The Salt−Cocrystal Continuum: The Influence of Crystal Structure on Ionization State. Mol. Pharmaceutics 2007, 4, 323−338. (29) Bhogala, B. R.; Basavoju, S.; Nangia, A. Tape and Layer Structures in Cocrystals of Some Di- and Tricarboxylic Acids with 4,4′Bipyridines and Isonicotinamide. From Binary to Ternary Cocrystals. CrystEngComm 2005, 7, 551. (30) Mohamed, S.; Tocher, D. A.; Price, S. L. Computational Prediction of Salt and Cocrystal structuresDoes a Proton Position Matter? Int. J. Pharm. 2011, 418, 187−198. (31) Mohamed, S.; Tocher, D. a.; Vickers, M.; Karamertzanis, P. G.; Price, S. L. Salt or Cocrystal? A New Series of Crystal Structures Formed from Simple Pyridines and Carboxylic Acids. Cryst. Growth Des. 2009, 9, 2881−2889. (32) Cruz-Cabeza, A. J. Acid−base Crystalline Complexes and the pKa Rule. CrystEngComm 2012, 14, 6362. (33) Shattock, T. R.; Arora, K. K.; Vishweshwar, P.; Zaworotko, M. J. Hierarchy of Supramolecular Synthons: Persistent Carboxylic Acid???

important consequences in the drug industry, as one can design and predict the API properties of small molecules a priori. Also, from a fundamental aspect, we suggest that the protons in acid−base mixture within crystals can be rather dynamic, and hence, sufficient attention (more so nonequimolar cases) must be paid to aggregation induced structural transformation before classifying them into cocrystals or salts.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jpcb.6b05830. Optimized crystal structure at DFT-D2, Cartesian coordinates and harmonic frequencies for the optimized structures, and complete ref 45 (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Phone: +91-33-24734971. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS S.M.P. thanks CSIR India for SRF. A.D. thanks INSA, DST, and BRNS for partial funding.

■

REFERENCES

(1) Desiraju, G. R. Crystal Engineering: From Molecule to Crystal. J. Am. Chem. Soc. 2013, 135, 9952−9967. (2) Caira, M. R. In Design of Organic Solids; Weber, E., Aoyama, Y., Caira, M. R., Desiraju, G. R., Glusker, J. P., Hamilton, A. D., Meléndez, R. E., Nangia, A., Eds.; Springer: Berlin, Heidelberg, 1998; pp 163− 208. (3) Desiraju, G. R. Supramolecular Synthons in Crystal EngineeringA New Organic Synthesis. Angew. Chem., Int. Ed. Engl. 1995, 34, 2311−2327. (4) Pratik, S. M.; Nijamudheen, A.; Datta, A. Janus All-Cis-1,2,3,4,5,6Hexafluorocyclohexane: A Molecular Motif for Aggregation-Induced Enhanced Polarization. ChemPhysChem 2016, DOI: 10.1002/ cphc.201600262. (5) Nangia, A.; Desiraju, G. R. In Supramolecular Synthons and Pattern Recognition BT - Design of Organic Solids; Weber, E., Aoyama, Y., Caira, M. R., Desiraju, G. R., Glusker, J. P., Hamilton, A. D., Meléndez, R. E., Nangia, A., Eds.; Springer: Berlin, Heidelberg, 1998; pp 57−95. (6) Du, M.; Zhang, Z.-H.; Zhao, X.-J. Cocrystallization of Bent Dipyridyl Type Compounds with Aromatic Dicarboxylic Acids: Effect of the Geometries of Building Blocks on Hydrogen-Bonding Supramolecular Patterns. Cryst. Growth Des. 2005, 5, 1199−1208. (7) Wang, W.-H.; Xi, P.-H.; Su, X.-Y.; Lan, J.-B.; Mao, Z.-H.; You, J.S.; Xie, R.-G. Supramolecular Assemblies of Multifunctional Diimidazole and Dicarboxylic Acids via Various Hydrogen Bonds and X···π (X = π, CH) Interactions. Cryst. Growth Des. 2007, 7, 741− 746. (8) Atwood, J. L.; Barbour, L. J.; Jerga, A. A New Type of Material for the Recovery of Hydrogen from Gas Mixtures. Angew. Chem., Int. Ed. 2004, 43, 2948−2950. (9) Pratik, S. M.; Nijamudheen, A.; Bhattacharya, S.; Datta, A. Color Polymorphism: Understanding the Diverse Solid-State Packing and Color in Dimethyl-3,6-Dichloro-2,5-Dihydroxyterephthalate. Chem. Eur. J. 2014, 20, 3218−3224. (10) Zelichenok, A.; Burtman, V.; Zenou, N.; Yitzchaik, S.; Di Bella, S.; Meshulam, G.; Kotler, Z. Quinolinium-Derived Acentric Crystals for Second-Order NLO Applications with Transparency in the Blue. J. Phys. Chem. B 1999, 103, 8702−8705. 7612

DOI: 10.1021/acs.jpcb.6b05830 J. Phys. Chem. B 2016, 120, 7606−7613

Article

The Journal of Physical Chemistry B pyridine Hydrogen Bonds in Cocrystals That Also Contain a Hydroxyl Moiety. Cryst. Growth Des. 2008, 8, 4533−4545. (34) Singh, D.; Bhattacharyya, P. K.; Baruah, J. B. Structural Studies on Solvates of Cyclic Imide Tethered Carboxylic Acids with Pyridine and Quinoline. Cryst. Growth Des. 2010, 10, 348−356. (35) Liang, P. A Cocrystal of Pyridine-2,4-Dicarboxylic Acid and Serine. Acta Crystallogr., Sect. E: Struct. Rep. Online 2008, 64, o43. (36) Wiechert, D.; Mootz, D. Molecular beside Ionic: Crystal Structures of a 1/1 and a 1/4 Adduct of Pyridine and Formic Acid. Angew. Chem., Int. Ed. 1999, 38, 1974−1976. (37) Aakeröy, C. B.; Fasulo, M. E.; Desper, J. Cocrystal or Salt: Does It Really Matter? Mol. Pharmaceutics 2007, 4, 317−322. (38) Pratik, S. M.; Chakraborty, S.; Mandal, S.; Datta, A. Cooperativity in a New Role: Stabilization of the Ammonium Salts in the Solid State over Their H-Bonded Complexes in the Gas Phase. J. Phys. Chem. C 2015, 119, 926−933. (39) Datta, A.; Pati, S. K. Dipolar Interactions and Hydrogen Bonding in Supramolecular Aggregates: Understanding Cooperative Phenomena for 1st Hyperpolarizability. Chem. Soc. Rev. 2006, 35, 1305−1323. (40) Pratik, S. M.; Datta, A. 1,4-DithiinePuckered in the Gas Phase but Planar in Crystals: Role of Cooperativity. J. Phys. Chem. C 2015, 119, 15770−15776. (41) Perdew, J. P.; Burke, K.; Ernzerhof, M. Generalized Gradient Approximation Made Simple. Phys. Rev. Lett. 1996, 77, 3865−3868. (42) Giannozzi, P.; Baroni, S.; Bonini, N.; Calandra, M.; Car, R.; Cavazzoni, C.; Ceresoli, D.; Chiarotti, G. L.; Cococcioni, M.; Dabo, I.; et al. QUANTUM ESPRESSO: A Modular and Open-Source Software Project for Quantum Simulations of Materials. J. Phys.: Condens. Matter 2009, 21, 395502. (43) Vanderbilt, D. Soft Self-Consistent Pseudopotentials in a Generalized Eigenvalue Formalism. Phys. Rev. B: Condens. Matter Mater. Phys. 1990, 41, 7892−7895. (44) Grimme, S. Semiempirical GGA-Type Density Functional Constructed with a Long-Range Dispersion Correction. J. Comput. Chem. 2006, 27, 1787−1799. (45) Frisch, M. J. Gaussian 09, rev A.01; Gaussian, Inc.: Wallingford, CT, 2009. (46) Zhao, Y.; Truhlar, D. G. The M06 Suite of Density Functionals for Main Group Thermochemistry, Thermochemical Kinetics, Noncovalent Interactions, Excited States, and Transition Elements: Two New Functionals and Systematic Testing of Four M06-Class Functionals and 12 Other Function. Theor. Chem. Acc. 2008, 120, 215−241. (47) Hehre, W. J.; Ditchfield, R.; Pople, J. A. SelfConsistent Molecular Orbital Methods. XII. Further Extensions of Gaussian Type Basis Sets for Use in Molecular Orbital Studies of Organic Molecules. J. Chem. Phys. 1972, 56, 2257−2261. (48) Simon, S.; Duran, M.; Dannenberg, J. J. How Does Basis Set Superposition Error Change the Potential Surfaces for HydrogenBonded Dimers? J. Chem. Phys. 1996, 105, 11024. (49) Fernandez-Berridi, M. J.; Iruin, J. J.; Irusta, L.; Mercero, J. M.; Ugalde, J. M. Hydrogen-Bonding Interactions between Formic Acid and Pyridine. J. Phys. Chem. A 2002, 106, 4187−4191. (50) Jemmis, E. D.; Giju, K. T.; Leszczynski, J. Tautomeric Rearrangements in Mono- and Dichalcogenide Analogs of Formic Acid, HC(X)YH (X, Y = O, S, Se, Te): A Theoretical Study. J. Phys. Chem. A 1997, 101, 7389−7395. (51) Karmakar, S.; Datta, A. Tunneling Assists the 1,2-Hydrogen Shift in N-Heterocyclic Carbenes. Angew. Chem., Int. Ed. 2014, 53, 9587−9591. (52) Jose, D.; Datta, A. Tunneling Governs Intramolecular Proton Transfer in Thiotropolone at Room Temperature. Angew. Chem., Int. Ed. 2012, 51, 9389−9392. (53) Lee, R.; Firbank, A. J.; Probert, M. R.; Steed, J. W. Expanding the Pyridine−Formic Acid Cocrystal Landscape under Extreme Conditions. Cryst. Growth Des. 2016, 16, 4005−4011.

7613

DOI: 10.1021/acs.jpcb.6b05830 J. Phys. Chem. B 2016, 120, 7606−7613