research papers
Unveiling the self-association and
in crystal nucleationaNational Engineering Research Center of Industrial Crystallization Technology, School of Chemical Engineering and Technology, Tianjin university, Tianjin 300072, People's Republic of China, and bCollaborative Innovation Center of Chemical Science and Engineering, Tianjin 300072, People's Republic of China
*Correspondence e-mail: x_huang@tju.edu.cn, hongxunhao@tju.edu.cn
As the first step in the crystallization process, nucleation has been studied by many researchers. In this work, phenacetin (PHEN) was selected as a model compound to investigate the relationship between the solvent and nucleation kinetics. Induction times at different
in six solvents were measured. FTIR and NMR spectroscopy were employed to explore the solvent–solute interactions and the self-association properties in solution. Density functional theory (DFT) was adopted to evaluate the strength of solute–solvent interactions and the molecular conformations in different solvents. Based on these spectroscopy data, molecular simulation and nucleation kinetic results, a comprehensive understanding of the relationship between molecular structure, solution chemistry and nucleation dynamics is discussed. Both the solute–solvent interaction strength and the supramolecular structure formed by the self-association of solute molecules affect the nucleation rate. The findings reported here shed new light on the molecular mechanism of nucleation in solution.Keywords: nucleation kinetics; self-association; desolvation; supramolecular structure; crystal engineering; hydrogen bonding; density functional theory; molecular crystals.
1. Introduction
Crystallization from solution is one of the most widely used unit operations and has been extensively used in the food, chemical, dye, pharmaceutical and agrochemical industries. Nucleation is crucial as the first step of the crystallization process. It can affect the physical properties of a crystal, such as the structure, shape, defects, domain size and polymorphs (Ou et al., 2020).
Although there have been more and more studies on clusters in the nucleation process over the past decade, the molecular mechanism of the process is still unclear. During crystal nucleation, molecular aggregates in solution play an important role, though their characterization is extremely difficult to describe. Recent evidence suggests that solution chemistry can be an effective method to investigate self-assembly in the nucleation process by exploring the solute–solvent interactions and self-association of solute molecules in solution (Davey et al., 2013; Gebauer et al., 2014; Tang et al., 2018). For example, Davey and Trout used FTIR to study the existing form of butynoic acid in different solvents. In chloroform it is a carboxyl hydrogen-bonded dimer, whereas in ethanol, though the solute has the potential to form hydrogen bonds, it does not participate in the formation of dimers. This result also matches the crystallization behaviour. Dimers are easily formed in non-polar solvents and cyclic structures are easily formed in polar solvents (Parveen et al., 2005; Chen & Trout, 2008). Other systems were investigated using the solution chemistry method, such as 2,6-dihydroxybenzoic acid (Davey et al., 2001), p-acetanisidide (Saito et al., 2002), isonicotinamid (Kulkarni et al., 2012; Maggioni et al., 2017), benzoic acid (Tang, Zhang et al., 2017; Burton et al., 2010), tolfenamic acid (Tang, Mo et al., 2017; Mattei & Li, 2012), mandelic acids (Davey et al., 2006), inosine (Chiarella et al., 2007), etc. Recent advances in computational and analytical techniques have also facilitated investigations of larger molecular clusters more efficiently (Sosso et al., 2016). Therefore, several attempts have been made to explain the evolution of the so-called `growth unit' during nucleation (Gavezzotti et al., 1997; Di Tommaso, 2013; Zeglinski et al., 2018). By studying the molecular structure evolution pathways from solute molecules to supramolecular arrangements, the structure correlation between the solution aggregates and crystal syntheses can be explained (Byrn et al., 1976; Bernstein & Hagler, 1978; Habgood, 2012).
From these explorations it was found that not only can the final et al., 2015; Sullivan et al., 2014). For example, Davey et al. studied the influence of solvent on nucleation kinetics and concluded that the process of solute dimerization and is the speed-control step of the entire nucleation process (Sullivan et al., 2014). Similar conclusions have also been obtained in other systems, such as risperidone, where the stronger the solvent–solute interaction is, the slower the nucleation rate will be (Khamar et al., 2014; Mealey et al., 2015). However, other research studies showed contrasting results, for example, a series of structure-related benzoic acids were investigated in four solvents. For a variety of solvents and solutes, this assumption was still valid, but when all the solutes and solvents were considered, it did not hold, and eventually led to the aromatic stacking being assigned the key step in nucleation (Cruz-Cabeza et al., 2017).
and nucleation mechanism be determined by the solvent, but that the nucleation rate can also be affected by the solvent in the crystallization process. Therefore, a number of researchers suggested a link between the solution structure and measured nucleation kinetic data (DaveyIn order to better understand the relationship between molecular conformations, . The nucleation process of PHEN in six solvents [chloroform, acetonitrile, methanol, toluene, N,N-dimethylacetamide (DMA), dimethyl sulfoxide (DMSO)] was investigated through a combination of spectroscopic techniques (FTIR, NMR, NOESY) and computational methods [density functional theory (DFT)]. Crystallization of PHEN was carried out in six solvents, and the induction time under different supersaturations was measured to obtain the nucleation kinetic data. Furthermore, computational chemistry was also employed to aid in consistent interpretation, linking solute–solvent interactions and molecular conformations to nucleation behaviours.
solution chemistry and nucleation kinetics, investigations focused on the relationship between solution chemistry and nucleation kinetics were carried out here using phenacetin (PHEN) as the model compound, as shown in Fig. 12. Experimental and simulation methods
2.1. Materials
Phenacetin (PHEN) was purchased from Shanghai Yuanye Biological Technology Co. Ltd, China, and its d (99.8% D), acetonitrile-d3 (99.8% D), methanol-d4 (99.8% D) and DMSO-d6 (99.8% D) were purchased from SAAN Chemical Technology Co. Ltd of China. All chemicals were used without any further purification.
purity was higher than 98%. All the solvents employed (methanol, chloroform, acetonitrile, toluene, DMA and DMSO) were analytical reagent grade with molar purity higher than 99.5% and were obtained from Tianjin Kewei Chemical Technology Co. Ltd, China. Chloroform-2.2. Induction time measurement
The solubility data for PHEN were collected using the supporting information. For the purpose of solubility measurement, data were employed to give guidance for the concentration of detection in ATR-FTIR (attenuated total reflectance Fourier transform infrared spectroscopy) and prepare a solution with different concentrations to create specific supersaturations. The definition of the ratio is S = x/x*, where x is the actual mole fraction of PHEN and x* is the equilibrium solubility mole fraction.
(equilibrium for 24 h, three repeats) at 25 and 40°C, and are presented in Table S1 of theThe induction time, which is defined as the time when the constant
is established (temperature of solution reaches nucleation temperature) to the moments at which the detectable crystal particles appear. The turbidimeter (Crystal Eyes, DMS-2, HEL Ltd) was used to monitor the formation of nuclei. To begin this process, a round-bottomed jacketed glass batch crystallizer (350 ml) was employed to prepare different concentrations of solution by dissolving appropriate amounts of PHEN in the respective solvents at 45°C. To stir thoroughly, a mechanical stirrer with an agitation speed of 300 rpm was used in the crystallizer. The temperature was controlled by two thermostats (Xianou Laboratory Instrument Works Co. Ltd, Nanjing) connected to two t-branch pipes. The temperature accuracy was ±0.01 K. The temperature was first set at 45°C to dissolve the solid completely, then the t-branch pipes were changed to facilitate shock cooling to 25°C. The point at which the system dropped to 25°C was noted as the start of induction time, and when the turbidmeter indicated a sudden increase this was noted as the end point. In order to reduce the experimental error, six reproducible experiments were performed at each composition.The relationship between the nucleation rate and
can be described well using the classical nucleation theory model:where J is the nucleation rate (m−3s−1), A is nucleation kinetic parameter, S is the degree of B is the nucleation thermodynamic parameter, f0 is the factor independent of C0 is the concentration of nucleation sites, v0 is the volume of the solute molecule (m3), γ is the interface energy (mJ m−2), k is (J K−1) and T is the absolute temperature (K). So there is a linear relationship between ln(J/S) and 1/ln2s. By plotting a linear fit, we can obtain the pre-exponential kinetic factor A from the intercept and the thermodynamic parameter B from the slope.
2.3. Powder X-ray diffraction analysis
Crystals were isolated immediately upon the appearance of solids under different conditions. Powder X-ray diffraction (PXRD) was used to identify samples on a Rigaku D/max-2500 (Rigaku) using Cu Kα radiation (0.15405 nm) in the 2θ range 5–50° and with a scanning speed of 8° min−1 to determine the crystal form nucleated in these experiments. The results showed that all the PHEN obtained in this work was pure form I.
2.4. Single-crystal growth and analysis
The slow solvent evaporation method was employed to obtain single crystals of PHEN form I. Specific amounts of PHEN solid were dissolved in methanol, and then the solution was transferred to an open beaker with parafilm. A few holes were made on the parafilm to ensure the solution evaporated slowly. The whole system was placed into an oven and kept at 293.15 K. Crystals of PHEN and its solvent of appropriate size for single-crystal X-ray diffraction (SCXRD) were obtained after several days. SCXRD measurements were conducted on a Rigaku Saturn 70 CCD diffractometer using Mo Kα radiation (λ = 0.71073 Å) with a graphite monochromator. Integration and scaling of intensity data were accomplished using the program SAINT (Bruker, 2017). The structures were solved using the SHELXS2014 (Sheldrick, 2014) suite of programs, and was conducted using SHELXL2018 (Sheldrick, 2015).
2.5. FTIR spectroscopy
Solid spectroscopy data were collected using Bruker ATR-FTIR (attenuated total reflectance Fourier transform infrared spectroscopy), with a resolution value of 4 cm−1, a scan time of 32 and wavenumber ranging from 400 to 4000 cm−1. An ATR-FTIR spectrometer (ReactIRTM45, Mettler-Toledo) equipped with a Duradisc Dicomp probe was adopted to facilitate solution spectroscopy. For each sample, 32 scans were collected over a spectra range from 650 to 2800 cm−1 at 2 cm−1 resolution to investigate the molecular structure of PHEN at different concentrations in the six solvents tested. The concentration of PHEN solution used in this work was determined by solubilities in different solvents, which varies from unsaturated to supersaturated.
2.6. NMR spectroscopy
Different concentrations of 13C-NMR and 1H-NMR spectra were measured in DMSO, acetonitrile, methanol and chloroform. All the 13C-NMR and 1H-NMR spectra were detected using a 600 MHz liquid NMR spectrometer (Bruker AVANCE III) at 298 K after 32 and 1024 scans. The software Mestrenova (https://mestrelab.com/software/mnova/nmr/) was employed to process and analyze the data. The chemical shifts in the 1H and 13C spectra were determined relative to the internal reference TMS.
2.7. Nuclear Overhauser effect spectroscopy
2D NOESY experiments were carried out for PHEN solution in DMSO, methanol, acetonitrile and chloroform at room temperature using a 600 MHz Bruker AVANCE III NMR spectrometer. 2D NOE spectra were measured with a standard pulse for both F1 and F2 dimensions. The number of F1 increments was 256, each with 65 536 data points in the F2 dimension. The NOE mixing time was optimized to 0.8 s by measuring NOE buildups. The number of scans and dummy scans were set to be 16 and 2, respectively.
2.8. analysis
Hirshfeld surface and 2D fingerprint analyses were employed to quantitatively analyze and compare the intermolecular interactions of PHEN using Crystal Explorer 17 (Turner et al., 2017) software.
2.9. Computational method
2.9.1. surface computation
For the τ1 and τ2 (Fig. 2) was generated by scanning for 18 steps with a step length of 10° for both τ1 and τ2. All calculations were performed in the gas phase and solvent environment at the M06-2X/631 + G(d,p) level of theory with Gaussian09 (Frisch et al., 2016).
surface (PES) scan, the conformer in phenacetin was extracted and its geometry optimized (herein, conformer A). Then the PES of phenacetin conformer A with dihedral angles2.9.2. Electrostatic potential distribution
It is generally believed that electrostatic potential can be used to predict and explain the relative molecular orientation and the strength of combination if a complex is mainly assembled by static electricity (such as a hydrogen bond, dihydrogen bond, halogen bond, etc.). The quantitative molecular surface analysis module of the Multiwfn program (Lu & Chen, 2012a,b) is capable of partitioning the whole van der Waals surface into multiple fragments, allowing us to study the characteristics of electrostatic potential distribution (Lu & Manzetti, 2014). Multiwfn and VMD (Humphrey et al., 1996) were used plot the van der Waals surface electrostatic potential distribution.
2.9.3. Solvent–solute interaction calculation
Density functional theory (DFT) calculations were performed by Gaussian09 to quantify the interactions in (1:1) molecular complexes of PHEN in the six solvents (Frisch et al., 2016). The geometries were optimized by the hybrid M06-2x function and 6–31 + G(d,p) basis set with the Grimme D3 dispersion correction using the SMD implicit solvation model (Grimme et al., 2010; Pratt et al., 2007). The Grimme dispersion correction allows a good description of weak interactions, such as van der Waals interactions. The binding energy (ΔEbind) between two molecules is calculated using the following equation:
where EAB is the energy of the PHEN–solvent complex, and EA and EB are the energies of the isolated monomer PHEN and the solvent, respectively. All the energies have been corrected for the zero-point vibrational energies. BSSE is the basis set superposition error and is calculated to correct the over-estimation of binding energies caused by overlapping of the basis functions (Boys & Bernardi, 1970).
3. Results
3.1. Crystallization outcomes
The solid forms of PHEN in methanol, chloroform, acetonitrile, toluene, DMA and DMSO at different supersaturations and temperatures were studied. In all solutions form I was obtained and the corresponding PXRD patterns of are shown in Fig. 3.
The experimentally measured PXRD patterns of PHEN crystallized from different solvents and the simulated patterns from the Cambridge Structural Database (CSD) are compared in Fig. 3. The main diffraction peaks are consistent, indicating that they are all the same crystal form.
3.2. Nucleation rate in different solvents
The induction time of PHEN in different solvents at various supersaturations was measured and the results are shown in Table S2. Because the volume of solution used to measure the induction time is relatively large (generally ≥ 150 ml), the measured induction time fluctuation does not show the random phenomenon of the induction time usually observed in small volumes. By relating the induction time (t) to (S), it is possible to estimate the nucleation rates and nucleation kinetic parameters of crystals (Zong et al., 2019).
Fig. 4(a) presents the relationship between the nucleation rate J and solution S. The results indicate that PHEN showed the fastest nucleation rate in acetonitrile within the experimental followed by methanol. In the case of low the nucleation rate of PHEN in DMSO is faster than that in toluene, whereas the situation is the opposite for high Among the six selected solvents, the slowest nucleation rate is in DMA and chloroform. In Fig. 4(b), ln(J/S) and ln2S show a good linear relationship in six solvents, indicating that method, range and control of experiment conditions to conduct induction time measurements are suitable for this system. The kinetic parameter A, thermodynamic parameter B, molecular f0C0 and interfacial energy γ can be calculated from the slope and intercept, as shown in Table 1. It can be seen from the data that the interfacial energy followed the order of acetonitrile < methanol < DMSO < toluene < DMA < chloroform, almost the same with that of nucleation rate data. That is, the greater the interface energy, the more difficult it is to nucleate, indicating that nucleation is mainly controlled by thermodynamic processes. In contrast, the order of molecular was acetonitrile < methanol < DMSO < toluene < DMA < chloroform. A higher should lead to a shorter induction time, but this parameter had no obvious relationship with the order of the nucleation rates. It can be explained that the nucleation rate of the same solute molecule in different solvents may be determined by the interface energy, and the interface energy is closely related to the interactions between the solute molecule and the solvent molecule, such as hydrogen bonding and solvation.
|
3.3. analysis
The single-crystal data of PHEN form I are presented in Table 2. According to the SCXRD data, PHEN form I belongs to the monoclinic and the is P21/c. There is one PHEN molecule in the asymmetric unit.
|
The crystal structures of PHEN form I are shown in Fig. 5. The Hirshfeld surface was further used to quantify the different types of interactions and their contributions in crystal packing. The 2D fingerprint plot and the percentage of various contacts are shown in Fig. 6. In Fig. 5(a), form I was arranged alternately through N—H⋯O interactions, which corresponds to O⋯H as the strongest interactions in the 2D fingerprint plot. It can be seen that H⋯H contacts and C—H⋯π interactions contribute most to the Hirshfeld surface in form I, which is in agreement with the analysis in Fig. 5. The H⋯H contacts contributing the most part (56.2%) are likely due to the short contacts between the aromatic rings. The formation of a hydrogen bond between the amine and carboxyl group leads to close H⋯H contacts. On the other hand, the hydrogen atoms on aromatic rings become close when aromatic interactions (C—H⋯π and π⋯π) are formed, which contributes another part of the H⋯H contacts.
3.4. FTIR spectroscopy
The solid spectra of PHEN show strong bands for carbonyl stretching at 1643 and 1655 cm−1, indicating the formation of strong hydrogen bonds, as shown in Fig. 7(a). The in Fig. 5(a) shows the formation of hydrogen bonds between the carbonyl and the amine group, which is consistent with the results of the solid spectra.
Compared with the solid spectra, the carbonyl peak in the IR spectrum of PHEN in solution shows an obvious blue shift, this indicates the weakening of interactions of the carbonyl groups in the solution. The observed displacements of the ν(C=O) modes upon solution results from the reduction in the C=O⋯H—N hydrogen bonding present between the molecules in solution. The different values in different solvents are associated with the involvement of hydrogen bonding. Thus, it is reasonable to assume that the solute molecules were solvated by the solvents in solution. The stronger the interactions, the more the carbonyl band is displaced to lower wavenumbers. This feature can be used to rank the strength of solvent–solute interactions (Khamar et al., 2014; Mealey et al., 2015). Thus, Fig. 7(a) shows that the interaction strength of the PHEN carbonyl with the solvent increases in the order toluene < acetonitrile < DMA < DMSO < chloroform < methanol.
The IR spectra of PHEN in acetonitrile, methanol, DMSO and toluene [Figs. 7(b)–7(e)] show strong bands for carbonyl stretching at 1688, 1670, 1681 and 1700 cm−1, which represents varying degrees of solvation. Acetonitrile and DMSO can be hydrogen acceptors, whereas toluene is neither a hydrogen-bond donor nor a hydrogen-bond acceptor. It can only form weak interactions through C—H. Thus, we expect PHEN to show the highest stretching vibration peak of the carbonyl group in toluene. The IR spectrum of PHEN in toluene shows two peaks in the carbonyl group region, the strong peaks at 1700 cm−1 indicate the existence of non-solvated aggregates in toluene, and the weaker one at 1680 cm−1 suggests a small fraction involve hydrogen bonding. The shoulder peak of low wavenumber in toluene is not obvious due to the too-low solubility of PHEN in toluene. With increasing solute concentration in methanol, the carbonyl peak shows a shoulder at a lower wavenumber about 1655 cm−1 which continually increases with concentration. This phenomenon suggests an increase in strong bound carbonyl species with an increase in the concentration of solute. It is difficult to interpret the solute–solute aggregation information in alcohol solutions, as alcohol can act as both a hydrogen bond donor and a hydrogen bond acceptor, which makes it difficult to differentiate solvent–solute and solute–solute interactions.
The IR spectra of PHEN in chloroform shows that the stretching peak of carbonyl group has a significant supporting information, the spectra of the same concentration of PHEN in chloroform and chloroform-d have the same value for carbonyls, indicating that the hydrogen-bond complex is a solute–solute aggregate formed by C=O⋯H—N, which is consistent with the hydrogen bond formed in the crystal. At the same time, as the concentration increases, the blue shift of the benzene ring stretching peak also shows self-association between solutes. This supports that C=O⋯H—N and π⋯π interactions play an important role in self-association. Furthermore, a similar trend was observed in DMA solution: with increasing concentration the spectra move towards their position in the solid state.
with increasing concentration, indicating that the C=O hydrogen-bond complex is formed. According Fig. S1 of the3.5. NMR spectroscopy
1H NMR and 13C NMR chemical shifts are concentration-dependent in methanol-d4 [Figs. 8(a)–8(b)]. All the protons except H8 and the C=O 13C display downfield changes when the concentration increases, indicating H13 show the largest changes, implying the solvation effect is facilitated by hydrogen bonding formed between the ethoxy on PHEN and the hydroxyl on methanol. As the concentration of PHEN increases, the proportion of solvent decreases and the solvation strength decreases. Chemical shifts of 1H and 13C in chloroform-d were present in Figs. 8(c) and 8(d). All the protons display downfield changes with increasing concentration, which supports and self-association. H23 shows the largest changes, indicating that the carbonyl group plays an important role in the process. H7 exhibits a shielding effect as the concentration increases, these changes can be attributed to self-association through C=O⋯H—N and PHEN–PHEN stacking in the solvent (Tang, Mo et al., 2017), which also corresponds to the deshielding effect of the carbonyl group; these correspond to the of the carbonyl group and the blue shift of the C=C of the benzene ring in the IR spectrum.
In acetonitrile-d3 [Fig. 8(f)], carbonyl 13C of PHEN unveils similar concentration-dependent changes to those seen in chloroform-d. The deshielding effect of NH and carbonyl 13C can be attributed to self-association through C=O⋯H—N and π⋯π interactions. The of 1H changed dramatically in DMSO-d6, and the trend suggests two concentration-dependent events [Figs. 8(g) and 8(h)]. The increased of H7, H8, H23 and NH and the reduced of H13 are associated with the competitive phenomenon of and self-association.
3.6. 2D NOESY spectra
The structural details of these solute–solute and solute–solvent assemblies were further explored by 2D NOESY. As shown in Fig. 9, the NOE cross peak of H7 and H23 appears in chloroform-d solution. Combining the data in Figs. 9(a) and 9(b), the distance between H7 and H23 is 6.695 Å. But when the PHEN molecules assemble by forming a C=O⋯H—N hydrogen bond, the distance between H7 and H23 is 4.271 Å. Additionally, an NOE cross peak appears when the space distance is closer than 5 Å. However, PHEN in acetonitrile-d3, methanol-d4 and DMSO-d6 did not show the NOE cross peak of H7 and H23 (Figs. S2 for acetonitrile, S3 for DMSO and S4 for methanol). Therefore, PHEN shows an obvious self-association effect in chloroform, which is consistent with FTIR and NMR spectroscopy results. The self-assembly is not obvious in acetonitrile, methanol and DMSO, this may be due to the solvation effect which creates an energy barrier for its self-association.
3.7. Molecular conformation
Conformation adjustment is a vital process during nucleation (Derdour & Skliar, 2014; Li et al., 2020). If there is a high rotation barrier between the conformations in solid and solution, conformation adjustment may have an obvious effect on the nucleation process, which could decrease the nucleation rate (Zeglinski et al., 2018). Thus, a PES about τ1 and τ2 was generated as shown in Fig. 10.
From the PES, the results are evident. Regardless of whether it is in solution or solid, the energy change trend with the dihedral angle is the same. Both τ1 and τ2 reach the minimum energy at 179°, which is the most stable conformation. The conformations in the six solvents are essentially the same as in the solid, indicating they easily transform to the solid conformation during nucleation. Thus, it is reasonable to consider that the conformation adjustment does not have an important effect on the nucleation rate of PHEN in these six solvents.
3.8. Molecular interactions
The use of electrostatic potential is a well established approach to predict and explain the relative molecular orientation and the strength of the combination if a complex is mainly assembled by static electricity (such as a hydrogen bond, dihydrogen bond or halogen bond). And the more negative (or positive) the electrostatic potential is, the more electrophilic (nucleophilic) the atom is likely to be. Therefore, the distribution of the van der Waals surface electrostatic potential of molecules can be analyzed and used to predict the most active sites; four sites on PHEN were selected to optimize the 1:1 solute–solvent complexes and calculate the binding energy. As can be seen from Fig. 11, the carbonyl oxygen exhibits the greatest and was selected as site 1. The amino hydrogen showing the most positive surface electrostatic potential was selected as site 2. The other two sites are ethoxy oxygen (site 3) and benzene ring π-electrons (site 4). The distributions of the van der Waals surface electrostatic potential of solvent molecules are shown in Fig. S4. The optimized geometries and binding energy results are presented in Fig. 12.
It can be seen from Fig. 12 that site 1 presents the largest binding energy in acetonitrile, chloroform and methanol, this is because these all have strong hydrogen-bond donors, favouring the formation of heterodimers which can significantly affect the binding energy. Toluene and DMA as hydrogen-bond donors show the largest binding energy at site 3. Only DMSO as a hydrogen-bond acceptor shows the largest binding energy at site 2.
4. Discussion
Given the above data, it is now possible to investigate the role of solvents in the nucleation process from a molecular perspective. First, the relative nucleation rate of PHEN does not correlate with the fundamental properties of solvents, such as boiling point and density. Also, there is no correlation between the nucleation rate and the relevant solubility.
In recent years, a number of works focused on building a link between the nucleation rate and the solvation strength, such as risperidone (Mealey et al., 2015), tolbutamide(Zeglinski et al., 2018), fenoxycarb (Zeglinski et al., 2019), salicylic acid (Khamar et al., 2014) and PABA (Sullivan et al., 2014). A similar approach was also taken in this study. The nucleation rates followed the order of acetonitrile < methanol < DMSO < toluene < DMA < chloroform, which was consistent with the interface energy. IR and NMR spectroscopy show solute aggregation effects in DMA and chloroform, but the nucleation rates in chloroform and DMA are slowest. According to the calculated 1:1 solute–solvent binding energy, the binding energy of site 3 (ethoxy oxygen) and the nucleation rate are basically consistent, except for chloroform, indicating that site 3 plays a major role in the nucleation process.
Conformation adjustment also plays an important role in the nucleation process. Thus, the discrepancy in the chloroform and DMA cases may be caused by a high energy barrier between the conformations in chloroform and DMA. However, the computational results of the τ1 and τ2) in different solvents demonstrated that neither an obvious conformational difference in chloroform and DMA nor a high energy barrier exist in this system. Thus, it is concluded that the conformation adjustment should not be the main reason for an abnormal low nucleation rate.
scan about the torsion angles (It may be that as the concentration increases, PHEN forms a supramolecular structure in these two solvents which differs from the
and then needs to undergo a rearrangement process, so the nucleation rate is slowest. At the same time, the solvation in chloroform is through a special halogen-bond interaction, which may also cause a relatively large energy barrier to the subsequent rearrangement of the supramolecular structure.However, when all the solvent systems are considered together, no clear relationship between the specific site interaction, the solvation strength obtained from IR spectroscopy, the conformation structure similarity, the way of gathering and the nucleation difficulty could be summarized. No single factor could individually describe the actual order of the nucleation difficulty and each factor does play a crucial role in certain situations. Therefore, we suggest that none of the four factors: the similarity of the solute in liquid and solid states, the specific site interaction, the way of gathering and the solvation strength, could be neglected. It is speculated that the rearrangement of the supramolecular structure plays a crucial role in determining the nucleation rate.
5. Conclusions
Investigations on the relationship between solution chemistry and nucleation kinetics have been carried out using PHEN as a model compound. We found that form I of PHEN could be obtained in acetonitrile, methanol, toluene, chloroform, DMA and DMSO. The crystal structures were analysed. FTIR, NMR and NOESY spectroscopies were used to analyze the solute species in solution. The results showed that in chloroform and DMA, PHEN undergoes nucleation through C=O⋯H—N hydrogen bond self-association, whereas in the other solvents PHEN more or less undergoes a
process. However, the nucleation kinetics data showed that chloroform and DMA had the slowest nucleation rates. We speculate that the self-association in these two solutions produces a supramolecular structure which is different from the solid structure. Rearrangement plays a major role in the rate of nucleation. This work confirms the importance of the self-association and processes during crystal nucleation. However, since the nucleation of crystals is complicated and many factors could affect the nucleation process, much more work needs to be carried out to fully understand nucleation phenomena.Supporting information
Supporting information file. DOI: https://doi.org/10.1107/S2052252521003882/lq5039sup1.pdf
Funding information
This research was financially supported by the National Natural Science Foundation of China (grant Nos. 21978201; 21908159). The authors declare no competing financial interest.
References
Bernstein, J. & Hagler, A. T. (1978). J. Am. Chem. Soc. 100, 673–681. CrossRef CAS Web of Science Google Scholar
Boys, S. F. & Bernardi, F. (1970). Mol. Phys. 19, 553–566. CrossRef CAS Web of Science Google Scholar
Bruker (2017). SAINT v.8.38A. Bruker, Madison, Wisconsin, USA. Google Scholar
Burton, R. C., Ferrari, E. S., Davey, R. J., Finney, J. L. & Bowron, D. T. (2010). J. Phys. Chem. B, 114, 8807–8816. Web of Science CrossRef CAS PubMed Google Scholar
Byrn, S. R., Graber, C. W. & Midland, S. L. (1976). J. Org. Chem. 41, 2283–2288. CrossRef CAS PubMed Web of Science Google Scholar
Chen, J. & Trout, B. L. (2008). J. Phys. Chem. B, 112, 7794–7802. CrossRef PubMed CAS Google Scholar
Chiarella, R. A., Gillon, A. L., Burton, R. C., Davey, R. J., Sadiq, G., Auffret, A., Cioffi, M. & Hunter, C. A. (2007). Faraday Discuss. 136, 179–193. Web of Science CrossRef PubMed CAS Google Scholar
Cruz-Cabeza, A. J., Davey, R. J., Sachithananthan, S. S., Smith, R., Tang, S. K., Vetter, T. & Xiao, Y. (2017). Chem. Commun. 53, 7905–7908. CAS Google Scholar
Davey, R. J., Back, K. R. & Sullivan, R. A. (2015). Faraday Discuss. 179, 9–26. Web of Science CrossRef CAS PubMed Google Scholar
Davey, R. J., Blagden, N., Righini, S., Alison, H., Quayle, M. J. & Fuller, S. (2001). Cryst. Growth Des. 1, 59–65. Web of Science CrossRef CAS Google Scholar
Davey, R. J., Dent, G., Mughal, R. K. & Parveen, S. (2006). Cryst. Growth Des. 6, 1788–1796. Web of Science CrossRef CAS Google Scholar
Davey, R. J., Schroeder, S. L. M. & ter Horst, J. H. (2013). Angew. Chem. Int. Ed. 52, 2166–2179. Web of Science CrossRef CAS Google Scholar
Derdour, L. & Skliar, D. (2014). Chem. Eng. Sci. 106, 275–292. Web of Science CrossRef CAS Google Scholar
Di Tommaso, D. (2013). CrystEngComm, 15, 6564–6577. Web of Science CrossRef CAS Google Scholar
Frisch, M. J., Trucks, G. W., Schlegel, H. B., Scuseria, G. E., Robb, M. A., Cheeseman, J. R., Scalmani, G., Barone, V., Petersson, G. A., Nakatsuji, H., Li, X., Caricato, M., Marenich, A. V., Bloino, J., Janesko, B. G., Gomperts, R., Mennucci, B., Hratchian, H. P., Ortiz, J. V., Izmaylov, A. F., Sonnenberg, J. L., Williams, Ding, F., Lipparini, F., Egidi, F., Goings, J., Peng, B., Petrone, A., Henderson, T., Ranasinghe, D., Zakrzewski, V. G., Gao, J., Rega, N., Zheng, G., Liang, W., Hada, M., Ehara, M., Toyota, K., Fukuda, R., Hasegawa, J., Ishida, M., Nakajima, T., Honda, Y., Kitao, O., Nakai, H., Vreven, T., Throssell, K., Montgomery, J. A. Jr, Peralta, J. E., Ogliaro, F., Bearpark, M. J., Heyd, J. J., Brothers, E. N., Kudin, K. N., Staroverov, V. N., Keith, T. A., Kobayashi, R., Normand, J., Raghavachari, K., Rendell, A. P., Burant, J. C., Iyengar, S. S., Tomasi, J., Cossi, M., Millam, J. M., Klene, M., Adamo, C., Cammi, R., Ochterski, J. W., Martin, R. L., Morokuma, K., Farkas, O., Foresman, J. B. & Fox, D. J. (2016). GAUSSIAN09. Gaussian Inc., Pittsburgh, Pennsylvania, USA. Google Scholar
Gavezzotti, A., Filippini, G., Kroon, J., van Eijck, B. P. & Klewinghaus, P. (1997). Chem. Eur. J. 3, 893–899. CrossRef CAS Web of Science Google Scholar
Gebauer, D., Kellermeier, M., Gale, J. D., Bergström, L. & Cölfen, H. (2014). Chem. Soc. Rev. 43, 2348–2371. Web of Science CrossRef CAS PubMed Google Scholar
Grimme, S., Antony, J., Ehrlich, S. & Krieg, H. (2010). J. Chem. Phys. 132, 154104. Web of Science CrossRef PubMed Google Scholar
Habgood, M. (2012). Phys. Chem. Chem. Phys. 14, 9195–9203. Web of Science CrossRef CAS PubMed Google Scholar
Humphrey, W., Dalke, A. & Schulten, K. (1996). J. Mol. Graph. 14, 33–38. Web of Science CrossRef CAS PubMed Google Scholar
Khamar, D., Zeglinski, J., Mealey, D. & Rasmuson, C. (2014). J. Am. Chem. Soc. 136, 11664–11673. CrossRef CAS PubMed Google Scholar
Kulkarni, S. A., McGarrity, E. S., Meekes, H. & ter Horst, J. H. (2012). Chem. Commun. 48, 4983–4985. Web of Science CrossRef CAS Google Scholar
Li, X., Wang, N., Yang, J., Huang, Y., Ji, X., Huang, X., Wang, T., Wang, H. & Hao, H. (2020). IUCrJ, 7, 542–556. Web of Science CrossRef CAS PubMed IUCr Journals Google Scholar
Lu, T. & Chen, F. (2012b). J. Comput. Chem. 33, 580–592. Web of Science CrossRef PubMed Google Scholar
Lu, T. & Chen, F.-W. (2012a). Acta Phys-Chim. Sin. 28, 1–18. CrossRef Google Scholar
Lu, T. & Manzetti, S. (2014). Struct. Chem. 25, 1521–1533. Web of Science CrossRef CAS Google Scholar
Maggioni, G. M., Bezinge, L. & Mazzotti, M. (2017). Cryst. Growth Des. 17, 6703–6711. CrossRef CAS Google Scholar
Mattei, A. & Li, T. L. (2012). Pharm. Res. 29, 460–470. CrossRef CAS PubMed Google Scholar
Mealey, D., Zeglinski, J., Khamar, D. & Rasmuson, A. C. (2015). Faraday Discuss. 179, 309–328. CrossRef CAS PubMed Google Scholar
Ou, Z., Wang, Z., Luo, B., Luijten, E. & Chen, Q. (2020). Nat. Mater. 19, 450–455. CrossRef CAS PubMed Google Scholar
Parveen, S., Davey, R. J., Dent, G. & Pritchard, R. G. (2005). Chem. Commun. 1531–1533. Google Scholar
Pratt, L. M., Truhlar, D. G., Cramer, C. J., Kass, S. R., Thompson, J. D. & Xidos, J. D. (2007). J. Org. Chem. 72, 2962–2966. Web of Science CrossRef PubMed CAS Google Scholar
Saito, A., Igarashi, K., Azuma, M. & Ooshima, H. (2002). J. Chem. Eng. Jpn, 35, 1133–1139. Web of Science CrossRef CAS Google Scholar
Sheldrick, G. (2014). SHELXS2014. Universität of Göttingen, Germany. Google Scholar
Sheldrick, G. M. (2015). Acta Cryst. C71, 3–8. Web of Science CrossRef IUCr Journals Google Scholar
Sosso, G. C., Chen, J., Cox, S. J., Fitzner, M., Pedevilla, P., Zen, A. & Michaelides, A. (2016). Chem. Rev. 116, 7078–7116. CrossRef CAS PubMed Google Scholar
Sullivan, R. A., Davey, R. J., Sadiq, G., Dent, G., Back, K. R., ter Horst, J. H., Toroz, D. & Hammond, R. B. (2014). Cryst. Growth Des. 14, 2689–2696. Web of Science CrossRef CAS Google Scholar
Tang, W., Li, S. & Gong, J. (2018). Chem. Ind. Eng. 35, 2–11. Google Scholar
Tang, W. W., Mo, H. P., Zhang, M. T., Parkin, S., Gong, J. B., Wang, J. K. & Li, T. L. (2017). J. Phys. Chem. B, 121, 10118–10124. CrossRef CAS PubMed Google Scholar
Tang, W. W., Zhang, M. T., Mo, H. P., Gong, J. B., Wang, J. K. & Li, T. L. (2017). Cryst. Growth Des. 17, 5049–5053. CrossRef CAS Google Scholar
Turner, M. J., McKinnon, J. J., Wolff, S. K., Grimwood, D. J., Spackman, P. R., Jayatilaka, D. & Spackman, M. A. (2017). Crystal Explorer17. The University of Western Australia. https://crystalexplorer.scb.uwa.edu.au/. Google Scholar
Zeglinski, J., Kuhs, M., Devi, K. R., Khamar, D., Hegarty, A. C., Thompson, D. & Rasmuson, C. (2019). Cryst. Growth Des. 19, 2037–2049. CrossRef CAS Google Scholar
Zeglinski, J., Kuhs, M., Khamar, D., Hegarty, A. C., Devi, R. K., Rasmuson, A. C. & ÅC, (2018). Chem. Eur. J. 24, 4916–4926. Google Scholar
Zong, S., Wang, J., Wu, H., Liu, q, Huang, X., Wu, D., Zhou, G. & Hao, H. (2019). Acta Cryst. B75. 845–854. Google Scholar
This is an open-access article distributed under the terms of the Creative Commons Attribution (CC-BY) Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original authors and source are cited.