Energetics of carboxylic acid–pyridine heterosynthon revisited: A computational study of intermolecular hydrogen bond domination on phenylacetic acid–nicotinamide cocrystals

Aris Perdana Kusuma; Sundani Nurono Soewandhi; Rachmat Mauludin; Veinardi Suendo; Fransiska Kurniawan; Gawang Pamungkas; Yuda Prasetya Nugraha

doi:10.1515/chem-2022-0207

Article Open Access

Energetics of carboxylic acid–pyridine heterosynthon revisited: A computational study of intermolecular hydrogen bond domination on phenylacetic acid–nicotinamide cocrystals

Aris Perdana Kusuma , Sundani Nurono Soewandhi , Rachmat Mauludin , Veinardi Suendo , Fransiska Kurniawan , Gawang Pamungkas and Yuda Prasetya Nugraha

Published/Copyright: September 28, 2022

Published by

Become an author with De Gruyter Brill

Submit Manuscript Author Information Explore this Subject

From the journal Open Chemistry Volume 20 Issue 1

Abstract

Carboxylic acid–pyridine heterosynthon (CPHS) is one of the most common synthons found in cocrystal packing. Phenylacetic acid (PYC)–nicotinamide (NIC) (PYCNIC) cocrystals were used as a computational model to assess the most important factor in the emergence of the synthon. Geometry optimization was carried out on every possible two molecules of PYC–NIC conformation based on B3LYP-D3BJ/6-311G (d,p). Various energetic parameters, including total energy, interaction energy, and hydrogen bond energy, were used to compare the existing conformation to the putative conformation. The conformation with CPHS has −53.87 kJ mol⁻¹ of single intermolecular hydrogen bond energy (EHB), which is the strongest of all. It turns out that there is no other parameter better than EHB to describe the superiority of CPHS in PYCNIC.

Graphical abstract

Keywords: cocrystal; synthon; hydrogen bond; phenylacetic acid; B3LYP

1 Introduction

Synthons are somehow special in the discussion of cocrystal structures [1]. Such discussion is even more overwhelming compared to the best practice of total energy-based crystal prediction [2]. In the pharmaceutical field, the manufacture of cocrystals by utilizing synthon matching is used to improve the physical properties of drugs [3,4]. Quantitative insight into the dominant energy descriptors in synthon formation will help cocrystal design, complementing qualitative knowledge of synthon terms.

One of the most interesting synthons is the carboxylic acid–pyridine heterosynthon (CPHS). CPHS is one of the strongest synthons and is found abundant in cocrystal structures. This synthon is composed of a carboxyl group which is hydrogen bonded to nitrogen of pyridine. This synthon is so strong and so commonly found that people tend to start or predict cocrystal synthesis using this approach [5].

With such frequency and significance of its occurrence, it would be unusual to consider this synthon as a coincidence in cocrystal packing formation involving energy minimization. It is necessary to recognize the possibility of thought that the synthon could have more roles, such as being the precursor or seed of such cocrystal formation [6]. This indicates that the synthon has probably existed since the beginning and will always appear until a final crystal packing is formed. Many studies have been looking for the most determining factors behind the strength of the synthon, but there is no fixed consensus as to which factor plays the most crucial role.

In addition, it is necessary to observe the conformation of two molecules because a synthon is formed in that dimer structure [7]. An intact cocrystal packing, however, will be initially formed from molecules joining one after another, either with the same molecule or with other molecules, thus forming a heteromolecular conformation. In an isolated chamber, there will only be three possibilities of the early process: A–A, B–B, and A–B. Hence, it is important to look further into the initial structure, which is the conformation of two molecules, from now on, referred to as conformation. This conformation is where the synthon of interest (CPHS) exists. Thus, it is necessary to conduct an in-depth study of this conformation as a whole, not only on the synthon section [7]. Even if the conformation containing CPHS were special, it would be appropriate to observe other existing conformations. In fact, these conformations exist in the packing instead of other non-existing putative conformations. It is assumed that the existing conformation in cocrystal packing, either the whole or only one with the synthon of interest, has unique characteristics that are more prominent or stronger than putative conformations, thus enabling them to defeat those conformations and then exist in the final cocrystal packing.

Several studies have been conducted to reveal the reasons for conformational superiority, especially those containing the synthon. Some of them state that the superiority is because of the interaction energy (Ei). Ei is obtained from the energy difference of the dimer of the molecule with its constituents. A previous study examined a series of homomolecular pyrazine-carboxylic acid structures, some of which form CPHS [8]. The synthon, based on Ei, was said to be more energetically favored than carboxyl–carboxyl and pyridine–pyridine. It also stated that both O‒H⋯N (7–8 kcal mol⁻¹) and C‒H⋯O (1–2 kcal mol⁻¹) contribute to the formation of CPHS.

Other studies on cocrystals and acid–pyridine salts suggested that the formation of CPHS is also based on the strength of Ei [9]. It was reported that the synthon strength of acid–amide > acid–pyridine. Sarma and Saikia [3], who conducted experiments on several theophylline cocrystals, stated that Ei in the CPHS (−14.1777 kcal mol⁻¹) and O‒H⋯O – N‒H⋯O (−17.444 kcal mol⁻¹) was stronger than the single point synthon O‒H⋯N (−6.446 kcal mol⁻¹) and correlate with the high occurrence of the synthon in real structures. A more recent study stated that the formation of theophylline–acetyl salicylic acid (TEO–ASA) cocrystal could be rationalized because Ei of TEO–ASA (−71.2 kJ mol⁻¹) > ASA dimer (−66.9 kJ mol⁻¹) and TEO dimer (−46.9 kJ mol⁻¹) [10].

However, the Ei approach is sometimes inappropriate because it turns out that some strong Ei synthons are not formed in cocrystals. Thus, we should start over by sifting through more precise assessment parameters to identify the reasons behind the superiority of the existing cocrystal conformations and the conformation containing CPHS in particular. To the best of our knowledge, only very few studies compare two-molecule conformations to discover the advantages of the existing over the putative one. Furthermore, they have not explicitly compared the chance of occurrence among various conformations of two interacting molecules. Therefore, in this case, the cocrystals of phenylacetic acid (PYC)–nicotinamide (NIC) (PYCNIC) are used as the computational model.

PYCNIC is a CPHS-based cocrystal like other carboxylic acids such as salicylic acid–nicotinamide [11]. PYCNIC cocrystal packing comprises several conformations of two molecules linked by hydrogen bonds. The synthon of interest (CPHS) in PYCNIC is in a conformation (PN1) consisting of one PYC molecule and one NIC molecule linked by a strong hydrogen bond O‒H⋯N and one weaker hydrogen bond, C‒H⋯O. There are two other conformations, namely NIC–PYC (PN2) and NIC–NIC (PN3) (Figure 1 and Figure S1). In addition, several other putative conformations do not exist there, such as the homomolecular conformations of PYC and NIC per se. Hence, it is interesting to investigate why the cocrystal conformation exists instead of the homomolecular one.

Figure 1

Main conformation of PYCNIC cocrystal. Molecular dimers are denoted by P (PYC) and N (NIC). Capital letters represent the existing cocrystal conformation, while lowercase letters represent the putative conformation.

A thorough discussion of the energetic aspects of the conformation will complement previous discussions, which have focused more on synthons as a partial component. Although the results may not be generalized for all synthons, they can at least be used as a specific reference for this CPHS.

2 Methods

The crystal structure of PYCNIC (Figure S1) was obtained from the Cambridge Crystallographic Data Center (CCDC Number 1519693) [12]. Initial geometry dimer conformations with intermolecular hydrogen bonds were composed according to Figure 1 and Figures S2 and S3, covering all possible PYCNIC conformations of both the existing and putative ones (Scheme 1). Geometry optimization followed by frequency analysis and basis set superposition error adjustment [13] were carried out at the B3LYP-D3BJ/6-311G (d,p) computational level using Orca 4.2.1 [14,15]. Topological analysis of this optimized geometry was performed in MultiWFN 3.6 [16] employing the quantum theory of atoms in molecules (QTAIM) by finding the critical point position of the intermolecular hydrogen bonds.

Natural bond orbital (NBO) analysis was performed in NBO7 by mapping the intra- and intermolecular orbitals as well as using the second-order perturbation theory [17]. Meanwhile, Avogadro 1.2.0 [18] and VMD 1.9.3 [19] were used to visualize the conformational geometries.

Scheme 1

Workflow of PYCNIC cocrystal conformational analysis.

3 Results and discussion

3.1 Geometry optimization

The optimization of PYC geometry on PYCNIC cocrystal produced a bent shape with an angle of about 110°. Meanwhile, the NIC geometry tended to be planar, although the amide group was oblique. This is in line with their homomolecular structure, where PYC and NIC also have the same shape [20,21].

The single-molecule optimized geometry was used for the two-molecule conformation optimization. The conformation with CPHS, namely PN1, was formed by the PYC–NIC heteromolecular structure. There is only one type of this conformation in a real PYCNIC cocrystal. However, by pairing the molecules of PYC and NIC, some variations in the conformation were obtained (Figure S2). The differences were in the bent direction of the PYC and the orientation and oblique shape of the NIC. All variations had different total energy, Ei, and EHB. Only the conformational variation with the strongest energy value was included regarding this difference.

CPHS in PYCNIC is formed by O‒H⋯N bond. This also applies to PYC cocrystals in combination with acridine, caffeine, and isonicotinamide, where the bond always appears in one of the conformations [22]. The cocrystals of PYCNIC, PYC–isonicotinamide, and PYC–acridine have a CPHS structure, while in caffeine, the bond is formed in imidazole. Here, it can be seen that the typical O–H⋯N interaction between one molecule with a carboxyl group and another molecule with an N atom in its ring is a specialty of CPHS and other similar synthons.

The obvious difference between the conformation of the two molecules and their crystalline packing lies in the presence or absence of weak co-bonds, such as C‒H⋯O, next to the main O‒H⋯N bond. The weak C‒H⋯O bond of CPHS always exists in the two-molecule conformation, although this is not the case in the final crystal packing. The absence of C‒H⋯O in the final crystal packing is probably due to the interaction of the synthon-forming molecule with another adjacent molecule, causing this additional stabilizer bond to break.

In addition to the conformation with CPHS, an optimization was also carried out on other existing conformational geometries in PYCNIC cocrystal, namely PN2 (NIC–PYC) and PN3 (NIC–NIC). Furthermore, putative conformations which did not exist in the cocrystal were also optimized, be it heteromolecular or homomolecular PYC and NIC (Figure S3).

Typical results were obtained in the geometry optimization, where strong EHB and Ei will have shorter bond distances and straighter angles. For example, the main hydrogen bond of PN1f, one of the existing conformations, has a distance and angle of 1.73 Å and 179.33°, while pp, its non-existing counterpart, was even better with 1.64 Å and 178.92°. The distance and angle superiority are not always observable in the existing conformations, given that the values are not always better than the non-existing ones (Table S1). Thus, the geometry parameter cannot be used as a benchmark for the existence of conformations in crystal packing.

3.2 Total energy

The total energy is the best practice parameter to accurately predict crystal packing structure, although it involves an expensive and lengthy computational process. Total energy can only capture the whole final structure without being able to see the step-by-step or the beginning of the formation process [2]. However, the total energy parameter can still be utilized to optimize the conformational geometry of two molecules. This parameter can accurately represent the best local minima geometry for a two-molecule conformation. The theory level used, B3LYP-D3BJ/6-311G(d,p), was sufficient for such purpose considering the performance and computational cost [23]. It has been widely used in molecular optimization and computation of other energetic parameters, such as the optimization of acid–pyridine and acid–amide series of cocrystal and catechin–epicatechin conformer [7,9,24].

In the PYCNIC cocrystal packing, CPHS was formed by the conformation PN1a or PN1f, of which the NIC part rotated by about 90^o. In the packing, the C‒O in the PYC carboxyl group did not bind to the H on the NIC pyridine ring as in the two-molecule conformation but bound instead to one of the H atoms in the amide group of the other NIC molecule. This may be influenced by its interaction with other adjacent molecules, which will change shape to achieve energy minimization throughout the cocrystal packing.

In an isolated space, if each PYC and NIC molecule interacts, it is possible to obtain a conformation with the lowest total energy, namely PN1c, or even its global minima, namely pn. The latter is a PYCNIC heteromolecular conformation, having the lowest energy with O‒H⋯O and N‒H⋯O bonds, that does not even exist in the cocrystal packing.

The homomolecular conformation of PYC–PYC certainly has the lowest total energy compared to PYC–NIC and NIC–NIC because the energy of PYC is lower than that of NIC (Table 1). This means that the total energy parameter cannot directly indicate the superiority of the existing cocrystal conformation in general, nor the conformation with CPHS in particular, over other non-existing putative conformations. Thus, it is necessary to look for other parameters, or probably their derivatives, which may influence the existence of a conformation in cocrystal packing.

Table 1

Total energy (E), interaction energy (Ei), bond energy, and Laplacian bond order (LBO) of PYCNIC

Structure	Bond type	E (kJ mol⁻¹)	Ei (kJ mol⁻¹)	EiHB₁ (kJ mol⁻¹)	EiHB₂ (kJ mol⁻¹)	LBO1	LBO2
Existing
PN1a	PYC–NIC	−2302406.827	−59.228	−53.482	−5.746	0.00041243	0.00004431
PN1c	PYC–NIC	−2302407.375	−59.660	−53.872	−5.788	0.00041137	0.00004420
PN1e	PYC–NIC	−2302399.513	−52.589	−48.997	−3.592	0.00040092	0.00002939
PN1f	PYC–NIC	−2302399.358	−52.422	−48.856	−3.567	0.00039970	0.00002918
PN1g	PYC–NIC	−2302399.699	−52.688	−48.922	−3.766	0.00039437	0.00003036
PN1h	PYC–NIC	−2302399.951	−52.967	−49.453	−3.513	0.00039426	0.00002801
PN2	PYC–NIC	−2302390.406	−43.681	−41.689	−1.993	0.00010208	0.00000488
PN3	NIC–NIC	−2189158.521	−62.860	−31.773	−31.087	0.00028387	0.00027774
Putative
pn	PYC–NIC	−2302418.834	−70.015	−48.207	−21.808	0.00063132	0.00028560
pp	PYC–PYC	−2415676.969	−75.199	−37.634	−37.566	0.00064995	0.00064877
nn1	NIC–NIC	−2189129.349	−37.365	−35.748	−0.890	0.00006986	0.00000174
nn2	NIC–NIC	−2189153.904	−58.078	−29.203	−28.875	0.00006428	0.00006356
nn3	NIC–NIC	−2189112.753	−20.616	−15.919	−4.696	0.00001044	0.00000308
nn4	NIC–NIC	−2189147.781	−52.511	−25.756	−21.051	0.00007220	0.00005901
nn5	NIC–NIC	−2189135.214	−41.359	−34.785	−6.575	0.00020073	0.00003794

3.3 Chemical reactivity

Chemical reactivity is mainly expressed as the highest occupied/lowest unoccupied molecular orbital (HOMO–LUMO), both of which form the frontier molecular orbital. HOMO, associated with the ionization potential, reflects the ability of the species to donate electrons. On the other hand, in terms of electron affinity, LUMO reflects the ability to accept electrons. Both are considered significant to determine the reactivity and stability of the molecule calculated by the energy gap descriptor. In addition, there are various other supporting descriptors, such as hardness (η), softness (S), chemical potential (μ), electronegativity (χ), and electrophilicity index (ω). The hardness and softness descriptors are closely related to the energy gap. A stable molecule will have a large energy gap and hardness with low softness. On the contrary, a less stable molecule tends to have higher softness with a small energy gap and hardness [25]. Nevertheless, those chemical reactivity descriptors of PYCNIC do not specifically demonstrate the superiority of the existing conformation over the putative one (Figure 2).

Figure 2

HOMO and LUMO view of existing PYCNIC conformation.

The pp conformation had the best energy gap, hardness, and softness with −6.1655, 3.0828, and 0.1622 eV, respectively, and this was much better than the existing conformation. There is no linear relationship to describe the advantages of being an existing conformation. Therefore, the stability according to those descriptors cannot be used to show the superiority of the existing conformation. Even Table S3 shows that the putative conformations are more stable than the existing ones. Chemical reactivity descriptors may only represent the overall stability of a specific structure, such as the base, salt, and cocrystal forms [25,26]. However, these descriptors cannot explain the preference for PYCNIC cocrystal formation to their homomolecular constituent.

3.4 Electrostatic potential surface (EPS) analysis

EPS describes a molecule’s positive and negative charges (Figure 3). The charge distribution determines the interactions between molecules. Some colors are used to distinguish the charges. The blue areas represent negative charges, while the red areas are positively charged.

Figure 3

EPS of PN1: (i) front view and (ii) opposite view. The blue areas represent negative charges, while the red areas are positively charged.

The intermolecular hydrogen bonds in PYCNIC cocrystals are dominated by opposite charges that attract each other. The hydrogen bond donor has a positive charge, while the acceptor has a negative one. This also applies to the synthon of interest, CPHS, in which the positively charged O–H portion of PYC will bind to the negatively charged N atom of NIC. Similarly, it works for the weaker bond C–H⋯O, where C–H will interact with the O atom of PYC, which has the opposite charge. This EPS approach is more of an effort to confirm and harmonize with general knowledge of bonds and their constituent charges.

3.5 Interaction energy (Ei)

Ei comes from the interaction of two molecules in a conformation. In CPHS of PYCNIC, Ei is mainly composed of two hydrogen bonds, namely O‒H⋯N and C‒H⋯O. Ei is considered one of the main factors affecting the existence of conformations in packing [3,10]. However, based on Table 1, the strongest Ei is in pp (−75.199 kJ mol⁻¹), a homomolecular conformation that does not exist in cocrystal packing. The conformation only exists in the PYC single molecule packing [20]. In this case, the Ei of pp originates primarily in the combined interaction of two nearly identical hydrogen bonds O‒H⋯O. It is stronger than PN3 (−62,860 kJ mol⁻¹), one existing conformation, and even much stronger than the CPHS variant, PN1c (−59,660 kJ mol⁻¹).

Hence, Ei alone is not the main factor in the existence of a conformation in packing. If Ei is a determining factor, the combination of NIC and PYC will not form a cocrystal because the homomolecular conformation (pp) has stronger Ei than the heteromolecular (PN). Alternatively, the cocrystal formed will have a pp conformation with the O‒H⋯O bond. In reality, however, the PYCNIC cocrystal does not contain this conformation.

In addition, in the CPHS of PYCNIC final crystal packing, the accompanying C‒H⋯O bond does not originate in the same NIC molecule as the O‒H⋯N bond. However, those two bonds are on the same NIC molecule in other cocrystals, such as salicylic acid–nicotinamide and salicylic acid–isonicotinamide [11]. This fact further weakens the opinion that Ei is the main element forming CPHS, considering that, in the conformation of the two molecules, the bond is always present beside the main O–H⋯N bond. In PYCNIC, Ei represents a molecular interaction consisting mainly of two hydrogen bonds, both of which, if the interaction is so important, will then be intact during the formation of crystal packing.

3.6 Bond energy

The bond energy here is devoted to a single EHB in the intermolecular conformation. In CPHS, there are usually two intermolecular hydrogen bonds, namely O‒H⋯N and much weaker C‒H⋯O. Meanwhile, in a double carboxylic acid conformation like pp, almost identical EHB is often obtained from two O‒H⋯O. There are various methods to obtain EHB. However, different parameters will result in different values. In this project, we examined the bond energy based on QTAIM, Ei, and NBO with the discussion as follows.

3.7 QTAIM-based EHB

QTAIM [27,28] can be used in a topological analysis of optimized geometry to obtain the bond critical point (BCP) on covalent bonds, intramolecular, and intermolecular hydrogen bonds. The location of BCP is particular for each conformational geometry from which various properties can be obtained, such as electron density (ρ _BCP), Laplacian of electron density (∇² ρ _BCP), and potential energy density (V _BCP). Among these parameters, the potential energy density directly correlates with the QTAIM-based bond energy (EBCP = ½V _BCP) [29]. Therefore, it is used to compare the strength of intermolecular hydrogen bonds between conformations.

Apparently, EBCP can measure single EHB at identical bonds very well. It is proven in Table S2, in which pp with its two O‒H⋯O bonds has identical EBCP, each −66.786 and −66.704 kJ mol⁻¹. This is indeed in line with the initial assumption that identical bonds should also have identical strength of each bond. However, this parameter could not give a satisfactory answer when we compared it with other conformations, primarily based on the assumption that the existing conformation must be superior to the putative ones. Many EBCPs from the existing conformation were considered weaker by this parameter, such as O‒H⋯N of PN1c, which only had −58.881 kJ mol⁻¹. Therefore, it is concluded that the validity of this parameter in assessing EHB is only limited to comparing it with the same conformation but not with different conformations.

3.8 Ei-based EHB

The Ei-based EHB (EiHB) is obtained from Ei, which is divided proportionally based on the Laplacian bond order (LBO) of each intermolecular hydrogen bond. The LBO is a bond order based on ∇² ρ _BCP which has a direct correlation with the bond dissociation energy [30]. Surprisingly, EiHB can actually show the superiority of the existing conformation containing CPHS over other conformations (Figure 4). This parameter also successfully made the EHB of O–H⋯N in CPHS (PN1c: −53.872 kJ mol⁻¹) significantly superior to the strong bond O‒H⋯O from pp, which was only −37.634 kJ mol⁻¹.

Figure 4

Intermolecular interaction profile based on QTAIM: (i) homomolecular PYC; (ii) homomolecular NIC, and (iii) cocrystal PYCNIC. The written value is EiHB in kJ mol⁻¹.

Thus, we now obtain a parameter which is able to make one of the existing conformations superior to the putative one. This advantage lies in the single hydrogen bond, considering that synthons are generally formed from two hydrogen bonds, both identical and non-identical ones. In addition, this advantage is only seen in one existing conformation containing CPHS but not in the other two existing conformations.

Thus, the superiority of CPHS is evident and may determine the existence of its conformation in the final crystal packing. Given that CPHS or similar synthon almost always exists in cocrystals, the formation of the synthon conformation is not a mere coincidence. A coincidence means the molecules will arrange themselves so that the free energy is minimal, and a conformation with CPHS will coincidentally form in the final packing. It will certainly be unusual to think of a phenomenon as a mere coincidence when it almost always occurs. It makes more sense to assume that the conformation with very strong CPHS has been established at the outset. The strong EHB of O‒H⋯N makes the bond unbreakable, although the shape and orientation of rotation change in the process. The next PYC and NIC molecules will follow the conformational template, which acts as the foundation for subsequent conformations [6].

Despite its advantages, however, this parameter has limitations. Ei sometimes consists of more than two intermolecular hydrogen bonds or can also be composed of phi–phi interactions. Therefore, EiHB cannot be accurately applied to such conditions.

3.9 NBO-based EHB

NBO-based EHB (ENBO) also gives similar results to EiHB. Table S4 shows that O‒H⋯N of PN1c has the strongest ENBO, which is 47.949 kJ mol⁻¹. It is also stronger than O‒H⋯O of pp, the non-existing conformation, which only has 31.798 kJ mol⁻¹. It indicates that the conformation with strong ENBO will have more opportunities to be reserved in the packing. Although the value of ENBO is slightly different from EiHB, they both place the hydrogen bond in CPHS as the strongest. The strong hydrogen bond in CPHS can defend itself in such a way that even though other molecules hit it from different sides during the packing formation, it can adjust itself by rotating in any directions without breaking the hydrogen bond [22].

Besides appearing in heteromolecular cocrystals, this phenomenon is also seen in homomolecular crystals [20,31,32]. For example, PYC homomolecular crystals always have a pp conformation with relatively strong ENBO [20]. However, the opposite occurs in homomolecular NIC crystals with numerous conformational variation polymorphs [21,33,34]. Different from the strong EHB in CPHS, the EHB in NIC is only half of it, with approximately 30 kJ mol⁻¹, thus making the bond less powerful than that in CPHS. Therefore, not only PN3 with the strongest EHB but also other NIC–NIC conformations frequently exist in those polymorphs.

Additional parameters are required to compare the strengths between conformations. In this case, the margin percentage is used to identify the extent of the strength differences. Through the abovementioned explanation, it can be seen that only EiHB and ENBO align with the initial assumptions and logic regarding the strength of the synthon, making these two parameters usable for margin comparison. The margin compares the strongest cocrystal heteromolecular conformation with the strongest homomolecular conformation, which in this case is either PYC or NIC. The pp conformation is stronger than nn, so pp is used as a comparator. Therefore, we find that the margin is the difference in EHB strength between PN1c and pp, which are 43.15 and 50.79%, according to EiHB and ENBO, respectively. Both make the heteromolecular conformation of cocrystal superior to its homomolecular one.

4 Conclusion

In the context of synthon, looking at the PYCNIC, it seems that the constituent molecules form a cocrystal packing structure with the lowest possible energy as long as the synthon, CPHS, exists therein. Hence, at the beginning of the mix-and-match cocrystal formation, a very strong PN1c conformation seemingly formed and became the foundation for forming the following crystal structures. The formation of other conformations will then follow this earlier template [6].

Furthermore, looking at the margin of EHB, being much stronger is relative since more samples are needed to determine the cut-off of such margin. There is an opportunity to use this parameter as a predictive tool for cocrystal formation. However, according to this data alone, we can at least conclude that one of the special features of CPHS, especially for PYCNIC cocrystal, is the strength of the single EHB.

sundani@itb.ac.id

Acknowledgments

The first author thank Institut Teknologi Bandung (ITB) and Universitas Islam Indonesia (UII) ßor providing facilities during this research.

Funding information: This is part of the dissertation research of Aris Perdana Kusuma funded by Lembaga Pengelola Dana Pendidikan (LPDP-BUDI DN) Indonesia.
Author contributions: A.P.K. – formal analysis, funding acquisition, investigation, project administration, writing – original draft, writing – review and editing; S.N.S. – conceptualization, supervision; R.M. – conceptualization, supervision; V.S. – conceptualization, supervision; F.K. – methodology, validation; G.P. – methodology, validation; Y.P.N. – supervision, validation.
Conflict of interest: There are no conflict of interest to declare.
Ethical approval: The conducted research is not related to either human or animal use.
Data availability statement: All data generated or analyzed during this study are included in this published article and its supplementary information files.

References

[1] Desiraju GR. Crystal engineering: from molecule to crystal. J Am Chem Soc. 2013 Jul;135(27):9952–67.10.1021/ja403264cSearch in Google Scholar PubMed

[2] Reilly AM, Cooper RI, Adjiman CS, Bhattacharya S, Boese AD, Brandenburg JG, et al. Report on the sixth blind test of organic crystal structure prediction methods. Acta Crystallogr B Struct Sci Cryst Eng Mater. 2016 Aug;72(Pt 4):439–59.10.1107/S2052520616007447Search in Google Scholar PubMed PubMed Central

[3] Sarma B, Saikia B. Hydrogen bond synthon competition in the stabilization of theophylline cocrystals. Cryst Eng Comm. 2014;16(22):4753–65.10.1039/C3CE42332HSearch in Google Scholar

[4] Zheng Q, Rood SL, Unruh DK, Hutchins KM. Co-crystallization of anti-inflammatory pharmaceutical contaminants and rare carboxylic acid–pyridine supramolecular synthon breakdown. Cryst Eng Comm. 2018;20(41):6377–81.10.1039/C8CE01492BSearch in Google Scholar

[5] Shattock TR, Arora KK, Vishweshwar P, Zaworotko MJ. Hierarchy of supramolecular synthons: persistent carboxylic acid·pyridine hydrogen bonds in cocrystals that also contain a hydroxyl moiety. Cryst Growth Des. 2008;8(12):4533–45.10.1021/cg800565aSearch in Google Scholar

[6] Mukherjee A. Building upon supramolecular synthons: some aspects of crystal engineering. Cryst Growth Des. 2015;15(6):3076–85.10.1021/acs.cgd.5b00242Search in Google Scholar

[7] Dudek MK, Day GM. Explaining crystallization preferences of two polyphenolic diastereoisomers by crystal structure prediction. CrystEngComm. 2019;21(13):2067–79.10.1039/C8CE01783BSearch in Google Scholar

[8] Vishweshwar P, Nangia A, Lynch VM. Recurrence of carboxylic acid–pyridine supramolecular synthon in the crystal structures of some pyrazinecarboxylic acids. J Org Chem. 2002 Jan;67(2):556–65.10.1021/jo0162484Search in Google Scholar PubMed

[9] Vener MV, Levina EO, Koloskov OA, Rykounov AA, Voronin AP, Tsirelson VG. Evaluation of the lattice energy of the two-component molecular crystals using solid-state density functional theory. Cryst Growth Des. 2014;14(10):4997–5003.10.1021/cg5005243Search in Google Scholar

[10] Darwish S, Zeglinski J, Krishna GR, Shaikh R, Khraisheh M, Walker GM, et al. A new 1:1 drug–drug cocrystal of theophylline and aspirin: discovery, characterization, and construction of ternary phase diagrams. Cryst Growth Des. 2018;18(12):7526–32.10.1021/acs.cgd.8b01330Search in Google Scholar

[11] Berry DJ, Seaton CC, Clegg W, Harrington RW, Coles SJ, Horton PN, et al. Applying hot-stage microscopy to co-crystal screening: a study of nicotinamide with seven active pharmaceutical ingredients. Cryst Growth Des. 2008;8(5):1697–712.10.1021/cg800035wSearch in Google Scholar

[12] Amombo Noa FM, Jacobs A. CCDC 1519693: experimental crystal structure determination; 2017.Search in Google Scholar

[13] Jian FF, Zhao PS, Li YF, Wang X, Yu Q. A combined computational and experimental approach for investigating a hydrogen-bonded supermolecular compound comprising benzimidazole and malonic acid. Int J Quantum Chem. 2008;108(3):521–31.10.1002/qua.21511Search in Google Scholar

[14] Neese F. The ORCA program system. Wiley Interdiscip Rev Comput Mol Sci. 2012;2(1):73–8.10.1002/wcms.81Search in Google Scholar

[15] Neese F. Software update: the ORCA program system, version 4.0. Wiley Interdiscip Rev Comput Mol Sci. 2017;8:e1327.10.1002/wcms.1327Search in Google Scholar

[16] Lu T, Chen F. Multiwfn: a multifunctional wavefunction analyzer. J Comput Chem. 2012 Feb;33(5):580–92.10.1002/jcc.22885Search in Google Scholar

[17] Glendening ED, Badenhoop JK, Reed AE, Carpenter JE, Bohmann JA, Morales CM, et al. NBO 7.0. Madison: Theoretical Chemistry Institute, University of Wisconsin; 2018.Search in Google Scholar

[18] Hanwell MD, Curtis DE, Lonie DC, Vandermeersch T, Zurek E, Hutchison GR. Avogadro: an advanced semantic chemical editor, visualization, and analysis platform. J Cheminform. 2012 Aug;4(1):17.10.1186/1758-2946-4-17Search in Google Scholar

[19] Humphrey W, Dalke A, Schulten K. VMD: visual molecular dynamics. J Mol Graph. 1996 Feb;14(1):33–8.10.1016/0263-7855(96)00018-5Search in Google Scholar

[20] Hodgson DJ, Asplund RO. Phenylacetic acid. Acta Crystallogr C. 1991;47(9):1986–7.10.1107/S0108270191002275Search in Google Scholar

[21] Miwa Y, Mizuno T, Tsuchida K, Taga T, Iwata Y. Experimental charge density and electrostatic potential in nicotinamide. Acta Crystallogr B. 1999 Feb;55(Pt 1):78–84.10.1107/S0108768198007848Search in Google Scholar

[22] Amombo Noa FM, Jacobs A. Phenylacetic acid co-crystals with acridine, caffeine, isonicotinamide and nicotinamide: crystal structures, thermal analysis, FTIR spectroscopy and Hirshfeld surface analysis. J Mol Struct. 2017;1139:60–6.10.1016/j.molstruc.2017.02.066Search in Google Scholar

[23] Mardirossian N, Head-Gordon M. Thirty years of density functional theory in computational chemistry: an overview and extensive assessment of 200 density functionals. Mol Phys. 2017;115(19):2315–72.10.1080/00268976.2017.1333644Search in Google Scholar

[24] Jarzembska KN, Hoser AA, Varughese S, Kaminski R, Malinska M, Stachowicz M, et al. Structural and energetic analysis of molecular assemblies in a series of nicotinamide and pyrazinamide cocrystals with dihydroxybenzoic acids. Cryst Growth Des. 2017;17(9):4918–31.10.1021/acs.cgd.7b00868Search in Google Scholar

[25] Srivastava K, Shukla A, Karthick T, Velaga SP, Tandon P, Sinha K, et al. Molecular structure, spectroscopic signature and reactivity analyses of paracetamol hydrochloride monohydrate salt using density functional theory calculations. CrystEngComm. 2019;21(5):857–65.10.1039/C8CE01761ASearch in Google Scholar

[26] Srivastava K, Shimpi MR, Srivastava A, Tandon P, Sinha K, Velaga SP. Vibrational analysis and chemical activity of paracetamol–oxalic acid cocrystal based on monomer and dimer calculations: DFT and AIM approach. RSC Adv. 2016;6(12):10024–37.10.1039/C5RA24402ASearch in Google Scholar

[27] Kumar PS, Raghavendra V, Subramanian V. Bader’s theory of atoms in molecules (AIM) and its applications to chemical bonding. J Chem Sci. 2016;128(10):1527–36.10.1007/s12039-016-1172-3Search in Google Scholar

[28] Bader RF. A quantum theory of molecular structure and its applications. Chem Rev. 1991;91(5):893–928.10.1021/cr00005a013Search in Google Scholar

[29] Espinosa E, Molins E, Lecomte C. Hydrogen bond strengths revealed by topological analyses of experimentally observed electron densities. Chem Phys Lett. 1998;285(3–4):170–3.10.1016/S0009-2614(98)00036-0Search in Google Scholar

[30] Lu T, Chen F. Bond order analysis based on the Laplacian of electron density in fuzzy overlap space. J Phys Chem A. 2013 Apr;117(14):3100–8.10.1021/jp4010345Search in Google Scholar PubMed

[31] Long S, Mao T, Chen P, Liu M, Parkin S, Zhang M, et al. Strong hydrogen bond leads to a fifth crystalline form and polymorphism of clonixin. ChemistrySelect. 2017;2(17):4942–50.10.1002/slct.201700947Search in Google Scholar

[32] Sundaralingam M, Jensen LH. Refinement of the structure of salicylic acid. Acta Crystallogr. 1965 Jun;18(6):1053–8.10.1107/S0365110X65002517Search in Google Scholar

[33] Li J, Bourne SA, Caira MR. New polymorphs of isonicotinamide and nicotinamide. Chem Commun (Camb). 2011 Feb;47(5):1530–2.10.1039/C0CC04117CSearch in Google Scholar

[34] Fábián L, Hamill N, Eccles KS, Moynihan HA, Maguire AR, McCausland L, et al. Cocrystals of fenamic acids with nicotinamide. Cryst Growth Des. 2011;11(8):3522–8.10.1021/cg200429jSearch in Google Scholar

Received: 2022-07-09

Revised: 2022-08-17

Accepted: 2022-08-24

Published Online: 2022-09-28

This work is licensed under the Creative Commons Attribution 4.0 International License.

Supplementary material

Articles in the same Issue

https://doi.org/10.1515/chem-2022-0207

Keywords for this article

cocrystal; synthon; hydrogen bond; phenylacetic acid; B3LYP

Creative Commons

BY 4.0

Energetics of carboxylic acid–pyridine heterosynthon revisited: A computational study of intermolecular hydrogen bond domination on phenylacetic acid–nicotinamide cocrystals

Article

Abstract

Graphical abstract

1 Introduction

2 Methods

3 Results and discussion

3.1 Geometry optimization

3.2 Total energy

3.3 Chemical reactivity

3.4 Electrostatic potential surface (EPS) analysis

3.5 Interaction energy (Ei)

3.6 Bond energy

3.7 QTAIM-based EHB

3.8 Ei-based EHB

3.9 NBO-based EHB

4 Conclusion

Acknowledgments

References

Supplementary Material

Articles in the same Issue

Articles in the same Issue

Articles in the same Issue