pH-Dependent Crystallization of 2-, 4-, 5-, and 6-Hydroxynicotinic Acids in Aqueous Media

Abstract

2-, 4-, 5-, and 6-hydroxynicotinic acids were crystallized in a pH-dependent manner using only water as the preferred solvent. The crystallization outcome was quite diverse: individual crystals of different sizes and shapes, microcrystalline powders, crystalline aggregates, and almost amorphous solids. Such a variety of solid forms demonstrates the relevance of pH control during crystallization processes.

1. Introduction

Crystallization is extensively used for product separation and purification in production sectors such as food, agriculture, electronics, and pharmaceuticals [1]. There is still, however, a considerable lack of knowledge about the molecular mechanisms behind the formation of crystals. Solubility and crystallization studies on families of compounds are useful in understanding how systematic changes in a molecular structure can impact the crystallization outcome. Extending previous studies on the hydroxynicotinic acid family [2], particularly for their behavior in strictly aqueous media, in this work, the effect of pH on the crystallization of 2-, 4-, 5-, and 6-hydroxynicotinic acid isomers (2, 4, 5, and 6HNA, Figure 1a) was investigated. In aqueous media the pH affects the solubility and the outcome of crystallization processes. Voges et al. examined the solubility of amino acids by investigating the effects of varying pH values and additives [3]. By utilizing the acid-base behaviors (pKa values) of the individual amino acids, Fuchs et al. achieved accurate solubility predictions in aqueous electrolyte solutions, demonstrating strong agreement with the experimental data [4]. An approach proposed by Daldrup et al. allowed the simultaneous determination of solubility, species distributions, and pH in mixed amino acid solutions [5]. The findings of Sun et al. also indicated that the pH value strongly influenced the solubility, and when the pH values increased, the solubility of the cephradine form I solution initially decreased, then increased, reaching a minimum value at the isoelectric point [6]. Any assessment of the solubility variances between polymorphic forms of compounds of interest should consider the pH [7]. The pH influences not only the solubility but also the solids obtained. Indeed, pH-dependent crystallization of spinal steroid injections has been reported [8], attesting to the relevance of knowing well the acid-base behavior of the compounds in use and how it impacts their solubility and bioavailability. Bearing this in mind, hydroxynicotinic acids (HNAs) were crystallized at pH values ranging from 0 to 14. Each sample was made at roughly the saturation level of the compound in water. As the pH varied, the hydrogen bond acceptors and donors differed, as well as the counterions (Cl⁻ or Na⁺, as HCl or NaOH were used to adjust the pH) around the diverse protonated species of HNAs, with clear repercussions on the microcrystalline solids and their X-ray powder diffraction (PXRD) patterns. Using the species distribution diagrams, calculated from the protonation constants of each compound, it was possible to establish a link between the HNAs’ expected structure in the solution and the one observed in the solid state.

2. Materials and Methods

The 2-, 4-, 5-, and 6-hydroxynicotinic acids used in this work were purified and characterized as described before [2]. Shortly, the HNAs were purchased from commercial sources and were purified by either sublimation or recrystallisation. 2 and 6HNA were purified by two sequential recrystallizations at 363 K from distilled water. 4 and 5HNA were purified by sublimation at 448 K and 1.3 Pa. PXRD diffractograms were acquired using a Philips X’Pert PRO apparatus fitted with an X’Celerator detector using automatic data acquisition (X’Pert Data Collector, v2.0b, software) and the vertical goniometer PW 3050/60. The source of Kα radiation was copper. The tube current intensity was 30 mA, and the potential difference was 40 kV. The diffractograms were acquired at 293 ± 2 K in the range of 7 to 35° (2θ) using the continuous mode with a step size of 0.017° (2θ) and an accumulation time of 20 s per step. A silicon sample holder was used to mount the samples. The crystal structure of PA at 150 ± 2 K was solved from single-crystal X-ray diffraction data. A colorless and small prismatic 4HNA crystal obtained at pH ≈ 0 yielded X-ray diffraction (SCXRD) data. See Table S1 for the crystal data and refinement parameters. Data were collected using a Bruker AXS-KAPPA APEX II area detector diffractometer. The crystal was placed in Paratone-N oil and mounted on a Kaptan loop. A graphite-monochromated MoK_α (λ = 0.71073 Å) radiation source operating at 50 kV and 30 mA was used. Scanning electron microscopy (SEM) images of Au-sputtered samples (15 nm of gold, the sample fixed on a carbon tape) were acquired in high vacuum, using a FEI XL30 ESEM (FEI is now Thermo Fisher Scientific, Waltham, MA, USA) with a resolution of 50 nm. The electron beam voltage was 20 kV. The pH measurements were made at 293.2 K, with a TIM900 pH meter fitted with a Mettler Toledo InLab Routine pH electrode.

3. Results and Discussion

2-, 4-, 5-, and 6-hydroxynicotinic acids were crystallized in a pH-dependent manner by using water as the preferred solvent. The crystallization outcome was characterized by PXRD and either optical or electronic (SEM) microscopies. Species distribution diagrams calculated from known protonation constants were used to know which HNA species to expect at a given pH and to help interpret the different crystallization outcomes.

The PXRD pattern (Figure 2, top left, 3rd spectrum labeled pH ≈ 0) obtained for the 2HNA sample crystallized through the slow evaporation of its saturated solution at pH ≈ 0 might be attributed to the LH₃⁺ species, where the N atom was protonated and a Cl⁻ counterion was expected. The calculated speciation (Figure 2, right) shows that at this pH circa 5% of LH₂ species should be present. The referred 2HNA PXRD pattern displays some similarities to the one obtained for the recrystallized starting material (Figure 2, left, labeled “Recryst.” PXRD pattern), particularly up to 2θ = 21°, which is also analogous to the simulated pattern obtained for form VIII (see Table S2 for hydroxynicotinic acid polymorphs) [9]. The number VIII was given to this published 2HNA form in a chronological order, as other authors obtained 2HNA forms previously (Table S2). The same line of thinking was used for the other isomers, and in Figure 2, Figure 3, Figure 4 and Figure 5, the forms closer to the ones obtained in this study were put near them. The pattern from the sample at pH 3.3 shows more differences from the referred patterns, especially for 2θ values higher than 24°. The speciation indicates that, at this sample pH, a mixture of almost 50/50 LH₃⁺ and LH₂ species should be observed. The crystals obtained at pH ≈ 0 are larger and have less defects than those found at pH 3.3 (Figure 2). SCXRD was attempted with these samples; however, the data quality was insufficient for publication. Above this pH value, the other crystalline solids had no reason to attempt SCXRD (Figure 2). Then, the pattern from the sample at pH 6.0 contains many distinctive reflections all through 2θ. According to the species distribution diagram of this sample pH, a mixture of 50/50 LH₂ and LH⁻ species should be expected. The LH⁻ species is expected to bear a Na⁺ counterion. This, and the mixture of phases, must contribute to the highly populated patterns. The pattern from the sample at pH 8.2 has less diffraction peaks but shows some similarities to the previous, which might correlate well with the LH⁻ species. The pattern acquired for the sample at pH 10.0 might be consistent with a mixture of LH⁻ and L²⁻ species, and the one from the pH 12.0 sample is quite different and cleaner, indicating the presence of the L²⁻ species only.

In the case of 4HNA, the starting material was sublimed and represented between the patterns simulated from forms I and IV (Figure 3), as it resembled a mixture of both forms. Then, the PXRD obtained for the sample of 4HNA at pH ≈ 0 was represented above the aforementioned and was quite different from all the other 4HNA patterns, with an unrepeatable preferential orientation at 2θ = 27°. At this pH, the presence of the LH₃⁺ species (with a Cl⁻ counterion) was expected; however, it was not featured in the species distribution diagram, as the first protonation constant was too low to be accurately determined. Indeed, the crystal structure shown in Figure 1 provides validation for this species. This is relevant to prove that, although one cannot determine a very low protonation constant due to experimental constraints, it does not mean that such protonation is absent. Regarding the pattern observed for the sample of 4HNA at pH 6.1, it is similar to that of the hydrate H-I [11] and possibly consistent with the LH₂ species. It is noteworthy that the crystals from the sample at pH ≈ 0 and from the sample at pH 6.1 are indeed very dissimilar (Figure 3, bottom left images, where the crystals obtained at pH ≈ 0 are prismatic and have a tendency to aggregate, while the ones at pH 6.1 are long needles). The patterns for the samples at pH 9.4 and 11.1 are analogous: the first, probably consistent with the LH⁻ species and the second possibly consistent with a mixture of LH⁻ and L²⁻ species. Finally, the sample for pH 14.0 has a slightly different pattern, which correlates well with the fact that at this pH the L²⁻ species is expected.

For 5HNA, the starting material used was also obtained from sublimation; however, 5HNA was also recrystallized from its saturated aqueous solution (pH ≈ 4) for comparison purposes. The recrystallized sample has a pattern that is in good agreement with that simulated from the hydrate H-II (Figure 4). Then, the sample obtained at pH ≈ 0 shares a few diffraction peaks with the recrystallized sample, but it is substantially different. This corroborates the possible existence of the LH₃⁺ species. At pH 3.3, the pattern mutates, reflecting what might be a new species, LH₂, with the diffraction peak at 2θ = 31.7°, which is unique among all the other patterns. The pattern obtained for the sample at pH 4.5 is plotted above the one obtained for the sublimed, as some similar diffraction peaks are observable. Then, at pH 6.6, the pattern obtained might mirror the LH⁻ species. At pH 9.0, a mixture of LH⁻ and L²⁻ species is expected, and at pH 11.2 and at pH 14.0, the predominance of the L²⁻ species seems possible.

In the case of 6HNA, the PXRD pattern from the starting material is similar to that simulated from form II (see Figure 5; although a preferential orientation is observed at 2θ = 28° if the pattern is zoomed in on (Figure S27 of Ref. [2]), the correspondences to form II are evident). Then, the pattern acquired for the sample at pH 2.0 is similar to the simulated one from form II, which is consistent with the possible predominance of LH₂ species. The sample at pH 4.1 starts to show differences, particularly the appearance of diffraction peaks such as those found at 2θ = 9° and 31.7°, correlating well with a possible mixture of LH₂ and LH⁻ species. At pH 7.0, the pattern is again mutated in good agreement with the probable predominance of LH⁻ species. Then, the pattern obtained for the pH 11.0 sample changes once more and might correspond to a mixture of LH⁻ and L²⁻. Finally, the pattern for the sample at pH 13.0 is similar to the previous pattern, somewhat clearer, and concordant with the possible existence of L²⁻ species. The dominant species for all the HNAs and their calculated pH range are summarized in Table 1.

Crystals of different sizes and shapes were obtained and crystalline aggregates observed. Amorphous solids were also formed. The registered variety of solid forms attests to the relevance of pH control during crystallization. For the crystals of HNAs obtained at low pH values, SEM was used, and the results are shown in Figure 6. The starting materials of 2 and 6HNA essentially contained crystalline needles or blades, respectively. These morphologies are somewhat preserved in the samples recrystallized at a low pH. The 4 and 5HNA starting materials were aggregates that resulted in thick layers, whereas the recrystallized 4 and 5HNA displayed diverse prismatic shapes. A remark concerning the diffraction of the crystals obtained for 6HNA should be made here, as numerous unfruitful attempts were undertaken to achieve single-crystal structures by X-ray diffraction. The poor results were connected to the poor diffraction of the crystals.

4. Conclusions

The pH-dependent crystallization of 2-, 4-, 5-, and 6-hydroxynicotinic acids was investigated in aqueous media, as studies regarding families of related compounds are useful in understanding how systematic changes in a molecular structure can impact the crystallization outcome. A plethora of solid forms were obtained for each HNA at similar pH ranges, showing how such similar molecules can undergo different inter- and intramolecular interactions with repercussions in their solution and solid-state structures. Indeed, the crystals of different sizes and shapes and the crystalline aggregates obtained demonstrate well the impact that pH control can have during crystallization.

Author Contributions

Conceptualization, C.V.E.; methodology, C.V.E. and A.V.J.; investigation, C.V.E. and A.V.J.; resources, M.F.M.P.; data curation, C.V.E.; writing—original draft preparation, C.V.E.; writing—review and editing, C.V.E.; supervision, C.V.E.; project administration, C.V.E. and M.F.M.P.; and funding acquisition, M.F.M.P. All authors have read and agreed to the published version of the manuscript.

Acknowledgments

The authors thank A. Mourato (CQE-FCUL, Portugal) for her help with PXRD; F. Emmerling and M. Heilman (BAM, Berlin) for the SEM images; C. E. S. Bernardes, R. G. Simões, C. S. D. Lopes, and I. O. Feliciano for their assistance at the lab whenever needed; and M. E. Minas da Piedade for the resources and helpful discussions.

Conflicts of Interest

The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of the data; in the writing of the manuscript; or in the decision to publish the results.

Footnotes

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Figures and Table

Figure 1. Molecular structure of hydroxynicotinic acids (a), and a view of the 4HNA crystal structure acquired in this work (HCl was used to lower the pH) (b).

Figure 2. Bottom: Images of the 2HNA analyzed solids and microscopy images of the crystals obtained at pH ≈ 0 (scale bar reads 100 µm)and 3.3 (scale bar reads 500 µm). Top left: Comparison of the X-ray diffraction patterns acquired at 293 ± 2 K, the diffraction pattern of the 2HNA starting material (figure label reads “Recryst.”, as 2HNA was purified by recrystallization at pH ≈ 4, the pH observed for the saturated aqueous solutions), and the simulated pattern obtained from a published crystal structure [9] (all the diffractograms were normalized to the peak of highest intensity—In). Top right: Species distribution diagram for 2HNA in an aqueous solution calculated with protonation constants from Ref. [10] (C2HNA = 1.0 × 10−3 moldm−3).

Figure 3. Bottom: Images of the 4HNA analyzed solids and microscopy images of the crystals obtained at pH ≈ 0 (scale bar reads 500 µm) and 6.1 (scale bar reads 0.6 mm). Top left: Comparison of X-ray diffraction patterns acquired at 293 ± 2 K, the diffraction pattern of the 4HNA starting material (figure label reads “Sublim.”, as 4HNA was purified by sublimation), and the simulated patterns from crystal structures previously published [9,11,12] (all the diffractograms were normalized to the peak of highest intensity—In). Top right: Species distribution diagram for 4HNA in aqueous solution calculated with protonation constants from Ref. [13] (C4HNA = 1.0 × 10−3 moldm−3).

Figure 4. Bottom: Images of the 5HNA analyzed solids and microscopy images of the crystals obtained at pH ≈ 0 (scale bar reads 400 µm) and 14.0 (scale bar reads 0.8 mm). Top left: Comparison of X-ray diffraction patterns acquired at 293 ± 2 K, the diffraction pattern of the 5HNA starting material (figure label reads “Sublim.”, as 5HNA was purified by sublimation), and the simulated pattern obtained from a crystal structure published by our group [14] (all the diffractograms were normalized to the peak of highest intensity—In). Top right: Species distribution diagram for 5HNA in an aqueous solution calculated with the values from Ref. [15] (C5HNA = 1.0 × 10−3 moldm−3).

Figure 5. Bottom: Images of the 6HNA analyzed solids. Top left: Comparison of X-ray diffraction patterns acquired at 293 ± 2 K, the diffraction pattern of the 6HNA starting material (figure label reads “Recryst.”, as 6HNA was purified by recrystallization at pH ≈ 4, the pH observed for the saturated aqueous solutions), and the simulated pattern obtained from a crystal structure published by our group [9] (all the diffractograms were normalized to the peak of highest intensity—In). Top right: Species distribution diagram for 6HNA in aqueous solution calculated with the values from Ref. [16] (C6HNA = 1.0 × 10−3 moldm−3).

Figure 6. SEM images of the hydroxynicotinic acids used in this work. Left column: starting materials that were purified by (i) recrystallization from water (2HNA, scale bar 200 µm, and 6HNA, scale bar 100 µm at pH ≈ 4, which was the pH observed for the saturated solutions) and by (ii) sublimation (4HNA, scale bar 20 µm, and 5HNA, scale bar 20 µm). Right column: selected samples from the pH-dependent crystallization study obtained at low pH values (2HNA, scale bar 500 µm, 4HNA, scale bar 200 µm, 5HNA, scale bar 100 µm, and 6HNA scale bar, 100 µm.

Table 1

Table summarizing the dominant species and their pH range.

Compound	pH Range	Dominant Species
2HNA	~0–1.5	LH3+
	1.5–6.3	LH2
	6.3–9.6	LH−
	9.6–12	L2−
4HNA	~0–6.2	LH2
	6.2–10.8	LH−
	10.8–12	L2−
5HNA	~0–1.9	LH3+
	1.9–4.7	LH2
	4.7–8.6	LH−
	8.6–12	L2−
6HNA	~0–3.8	LH2
	3.8–10.8	LH−
	10.8–12	L2−

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/2073-4352/13/7/1062/s1, Table S1: Crystal data and structure refinement parameters for 4HNA HCl. Table S2: Hydroxynicotinic acid polymorphs. References [17,18,19,20,21,22] are cited in supplementary materials.

References

1. Mullin, J.W. Crystallization; 4th ed. Butterworth-Heinemann: Oxford, UK, 2001.

2. Esteves, C.V. Hydroxynicotinic Acid Crystallisation and Solubility Systematic Studies. New J. Chem.; 2022; 46, pp. 21124-21135. [DOI: https://dx.doi.org/10.1039/D2NJ00627H]

3. Voges, M.; Prikhodko, I.V.; Prill, S.; Hübner, M.; Sadowski, G.; Held, C. Influence of pH Value and Ionic Liquids on the Solubility of L-Alanine and L-Glutamic Acid in Aqueous Solutions at 30 °C. J. Chem. Eng. Data; 2017; 62, pp. 52-61. [DOI: https://dx.doi.org/10.1021/acs.jced.6b00367]

4. Fuchs, D.; Fischer, J.; Tumakaka, F.; Sadowski, G. Solubility of Amino Acids:  Influence of the pH Value and the Addition of Alcoholic Cosolvents on Aqueous Solubility. Ind. Eng. Chem. Res.; 2006; 45, pp. 6578-6584. [DOI: https://dx.doi.org/10.1021/ie0602097]

5. Daldrup, J.-B.G.; Held, C.; Sadowski, G.; Schembecker, G. Modeling pH and Solubilities in Aqueous Multisolute Amino Acid Solutions. Ind. Eng. Chem. Res.; 2011; 50, pp. 3503-3509. [DOI: https://dx.doi.org/10.1021/ie1010367]

6. Sun, H.; Jiang, C.; Liu, B.; Zhang, J. Determination and Correlation of Solubility of Cephradine and Cefprozil Monohydrate in Water as a Function of pH. J. Chem. Eng. Data; 2017; 62, pp. 3423-3430. [DOI: https://dx.doi.org/10.1021/acs.jced.7b00446]

7. Repin, I.A.; Loebenberg, R.; DiBella, J.; Conceição, A.C.L.; Minas da Piedade, M.E.; Ferraz, H.G.; Issa, M.G.; Bou-Chacra, N.A.; Ermida, C.F.M.; de Araujo, G.L.B. Exploratory Study on Lercanidipine Hydrochloride Polymorphism: pH-Dependent Solubility Behavior and Simulation of its Impact on Pharmacokinetics. AAPS PharmSciTech; 2021; 22, 54. [DOI: https://dx.doi.org/10.1208/s12249-021-01923-0] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/33475891]

8. Watkins, T.W.; Dupre, S.; Coucher, J.R. Ropivacaine and dexamethasone: A Potentially Dangerous Combination for Therapeutic Pain Injections. J. Med. Imaging Radiat. Oncol.; 2015; 59, pp. 571-577. [DOI: https://dx.doi.org/10.1111/1754-9485.12333] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/26076198]

9. Santos, R.C.; Figueira, R.M.; Piedade, M.F.M.; Diogo, H.P.; Minas da Piedade, M.E. Energetics and Structure of Hydroxynicotinic Acids. Crystal Structures of 2-, 4-, 6-Hydroxynicotinic and 5-Chloro-6-Hydroxynicotinic Acids. J. Phys. Chem. B; 2009; 113, pp. 14291-14309. [DOI: https://dx.doi.org/10.1021/jp906908n] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/19785458]

10. Di Marco, V.B.; Tapparo, A.; Dolmella, A.; Bombi, G.G. Complexation of 2-Hydroxynicotinic and 3-Hydroxypicolinic Acids with Zinc(II). Solution State Study and Crystal Structure of trans-Diaqua-bis-(3-Hydroxypicolinato) Zinc(II). Inorg. Chim. Acta; 2004; 357, pp. 135-142. [DOI: https://dx.doi.org/10.1016/S0020-1693(03)00500-0]

11. Matias, E.P.; Bernardes, C.E.S.; Piedade, M.F.M.; Minas da Piedade, M.E. A Robust Yet Metastable New Hemihydrate of 4-Hydroxynicotinic Acid. Cryst. Growth Des.; 2011; 11, pp. 2803-2810. [DOI: https://dx.doi.org/10.1021/cg1015687]

12. Long, S.H.; Zhang, M.T.; Zhou, P.P.; Yu, F.Q.; Parkin, S.; Li, T.L. Tautomeric Polymorphism of 4-Hydroxynicotinic Acid. Cryst. Growth Des.; 2016; 16, pp. 2573-2580. [DOI: https://dx.doi.org/10.1021/acs.cgd.5b01639]

13. Ross, W.C. The Preparation of Some 4-Substituted Nicotinic Acids and Nicotinamides. J. Chem. Soc. Perkin Trans. 1; 1966; 20, pp. 1816-1821. [DOI: https://dx.doi.org/10.1039/j39660001816] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/4223908]

14. Joseph, A.; Alves, J.S.R.; Bernardes, C.E.S.; Piedade, M.F.M.; Minas da Piedade, M.E. Tautomer Selection through Solvate Formation: The Case of 5-Hydroxynicotinic Acid. CrystEngComm; 2019; 21, pp. 2220-2233. [DOI: https://dx.doi.org/10.1039/C8CE02108B]

15. Lezina, V.P.; Stepanyants, A.U.; Golovkina, N.I.; Smirnov, L.D. Investigation of the Acid-Base Transformations in 3-Hydroxypyridine, 3-Hydroxypyridine N-Oxide, and β-Hydroxypyridinecarboxylic Acids by the NMR Method. Bull. Acad. Sci. USSR Div. Chem. Sci.; 1980; 29, pp. 85-92. [DOI: https://dx.doi.org/10.1007/BF00951884]

16. Roduit, J.-P. 6-Hydroxynicotinic Acid. Encyclopedia of Reagents for Organic Synthesis; Wiley: Hoboken, NJ, USA, 2001; Available online: https://onlinelibrary.wiley.com/doi/abs/10.1002/047084289X.rh062 (accessed on 1 June 2023).

17. Long, S.; Siegler, M.; Li, T. 2-Oxo-1,2-Dihydropyridine-3-Carboxylic Acid. Acta Cryst.; 2006; 62, pp. O5664-O5665. [DOI: https://dx.doi.org/10.1107/S1600536806048768]

18. Djurdjevic, S.; Leigh, D.; Parsons, S. CCDC 660787: Experimental Crystal Structure Determination. 2008; Available online: https://www.ccdc.cam.ac.uk/structures/Search?Ccdcid=CCDC%20660787&DatabaseToSearch=Published (accessed on 1 June 2023).

19. Kukovec, B.-M.; Popović, Z.; Pavlović, G.; Linarić, M.R. Synthesis and Structure of Cobalt(II) Complexes with Hydroxyl Derivatives of Pyridinecarboxylic Acids: Conformation Analysis of Ligands in the Solid State. J. Mol. Struct.; 2008; 882, pp. 47-55. [DOI: https://dx.doi.org/10.1016/j.molstruc.2007.09.011]

20. Miklovič, J.; Segľa, P.; Mikloš, D.; Titiš, J.; Herchel, R.; Melník, M. CCDC 670679: Experimental Crystal Structure Determination. 2010; Available online: https://www.ccdc.cam.ac.uk/structures/Search?Ccdcid=CCDC%20670679&DatabaseToSearch=Published (accessed on 1 June 2023).

21. Miklovič, J.; Segľa, P.; Mikloš, D.; Titiš, J.; Herchel, R.; Melník, M. Copper(II) and Cobalt(II) Hydroxypyridinecarboxylates: Synthesis, Crystal structures, Spectral and Magnetic Properties. Chem. Papers; 2008; 62, pp. 464-471. [DOI: https://dx.doi.org/10.2478/s11696-008-0056-5]

22. Long, S.H.; Zhou, P.P.; Theiss, K.L.; Siegler, M.A.; Li, T.L. Solid-State Identity of 2-Hydroxynicotinic Acid and Its Polymorphism. CrystEngComm; 2015; 17, pp. 5195-5205. [DOI: https://dx.doi.org/10.1039/C4CE02290D]