3.1 Primarily Characterization
Differential scanning calorimetry (DSC) is the most important technique that gives information about the phenomenon of crystallization, recrystallization, phase transformation and desolvation etc. Figure 1(A) gives the DSC thermograms of pure LZ form II, form IV, SY, FA, INC and eutectic polymorphs. As discussed, LZ used in the present study is form II which is a low melting metastable polymorph. The DSC thermogram of LZ very well showed that it melts and recrystalizes into a high melting form IV. Preliminary investigation of the coground binary mixture of linezolid and syringic acid (LS) showed two endotherms (129.7oC & 141.5oC) which are separated by an exotherm (132.5oC) indicating towards the formation of cocrystal through eutectic formation. It is very well documented in the literature [35–39] that presence of two endothermic peaks separated by an exotherm are due to the existence of metastable eutectic melting followed by an exothermic peaks attributed to cocrystal formation which is followed by an endothermic peak associated with cocrystal melting. Herein, both the endothermic events in LS lie below the individual melting endotherms of LZ (155.5oC, 180.3oC) and SY (209.6oC). Therefore, to explore the class of molecular product to whom LS belong to, Fourier transform infrared spectroscopy (FTIR) and powder X-ray diffraction (PXRD) studies were performed. The characteristic FTIR absorption bands corresponding to pure LZ form II (amide N-H stretch at 3363cm− 1; lactone C = O and amide C = O stretch at 1747cm− 1 and 1674cm− 1 respectively) and SY (like aliphatic C-H stretch and acidic C = O stretch at 2971cm− 1 and 1699cm− 1 respectively) were present in the spectra of LS. Neither any new absorption frequency nor shifting of any absorption frequencies was observed in the spectrum [Fig. 2(A)]. Similarly in diffractogram of LS, the unique diffraction peaks corresponding to LZ form II (at 2θ 7.1 o, 9.5 o, 16.8o and 22.4o) and SY (at 2θ 11.5o, 12.5o and 13.7o) remains intact but with weaker intensities, suggesting a low degree of crystallinity than of the individual substances due to the shearing force applied during grinding [Fig. 3(A)]. This indicates that there is no interaction in solid state between LZ form II and SY. This verifies that the molecular product with a melt at 129.7oC corresponds to a eutectic mixture.
Now to confirm to which molecular state the product get transformed after melting, the coground mixture was heated upto 133.0oC in an oven. The resulting sample was again subjected to DSC analysis [Fig. 1(B)] and only one endothermic peal was observed. The heated product was further subjected to FTIR [Fig. 2(B) and Table S1] and PXRD analysis. Interestingly, all the absorption frequencies corresponding to SY (like aliphatic C-H stretch and acidic C = O stretch at 2971cm− 1 and 1699cm− 1 respectively) are same and the absorption bands for LZ matches with the characteristic frequencies of form IV of LZ (like amide N-H stretch at 3339cm− 1; lactone C = O and amide C = O stretch at 1742cm− 1 and 1662cm− 1 respectively) and are also intact. Here also, neither any new band frequency nor shifting of any absorption frequencies were observed. In addition, the typical diffraction peaks appearing at 2θ 7.3o, 20.9o, 25.3o and at 11.5o, 12.5o, 13.7o in the diffractogram [Fig. 3(B) and Table 1] corresponds to form IV of LZ and SY and are intact confirming its crystalline nature also.
This unequivocally confirms that the heat mediated transformation of the eutectic does not lead to the cocrystal formation but again eutectic. Most probably it shows the existence of polymorphism in eutectic. Polymorphism which is very well documented in cocrystals and salts is first time reported in eutectics. The similar behavior was shown by eutectic mixtures of LZ with FA (LF) and LZ with INC (LN). Once it is established that the present eutectic mixtures shows the prevalence of polymorphism in eutectics, it was decided to ascertain their stoichiometries.
Table 1
Characteristic powder diffraction peaks of LZ form II, form IV, eutectic polymorphs and coformers used.
Compound | Peak Position |
LZ form II | 7.1o, 9.5o, 13.5o, 14.2o, 16.8o, 22.4o, 25.3o, 27.1o |
LZ form IV | 7.3o, 9.3o, 13.4o, 16.8o, 17.9o, 18.5o, 20.9o, 22.1o, 25.3o, 28.2o |
SY | 11.5o, 12.5o, 13.7o, 22.8o, 24.5o, 34.2o |
LS(A) | 7.1o, 9.5o, 11.5o, 12.5o, 13.5o, 13.7o, 14.2 o, 16.8o, 22.4o, 22.8o, 24.5o, 25.3o, 27.1o, 34.3o |
LS(B) | 7.3o, 9.3o, 11.5o, 12.5o, 13.4o, 13.7o, 16.8o, 17.9o, 18.5o, 20.9o, 22.1o, 22.8o, 24.5o, 25.3o, 28.2o, 34.2o |
INC | 17.7o, 19.3o, 20.7o, 23.3o, 25.7o, 30.8o, 38.7o |
LN(A) | 7.1o, 9.5o, 13.5o, 14.2o, 16.8o, 17.7o, 19.3o, 20.6o, 22.4o, 23.2o, 25.3o, 27.1o, 30.8o, 38.2o |
LN(B) | 7.3o, 8.6o, 9.3o, 13.4o, 16.8o, 17.9o, 18.5o, 19.4o, 20.9o, 22.1o, 23.2o, 25.3o, 28.2o, 30.7o, 38.5o |
FA | 22.8o, 24.8o, 25.7o, 26.8o, 28.3o, 29.4o, 38.1o, 38.5o, 38.7o, 39.9o, 42.2o, 42.9o |
LF(A) | 7.1o, 9.5o, 13.5o, 14.2o, 16.8o, 22.4o,22.8o, 24.8o, 25.3o, 26.8o, 27.1o, 28.3o, 29.4o, 38.1o, 38.4o, 38.7o, 39.8o, 42.1o, 42.8o |
LF(B) | 7.3o, 9.3o, 13.4o, 16.8o, 17.9o, 18.5o, 20.9o, 22.1o, 22.8o, 24.8o, 25.3o, 26.8o, 28.2o, 29.4o, 38.1o, 38.4o, 38.7o, 39.8o, 42.1o, 42.8o |
3.2 Establishment of stoichiometries of eutectic mixtures
To get clearer picture of eutectic composition of the LS, binary phase diagram was constructed by plotting temperature versus mole fraction. For this, various molar concentrations of LZ and various coformers were prepared and have been subjected to DSC analysis (Fig. 4). In case of LS, as the molar concentration of LZ increased from 0:100 to 20:80, multiple peaks appeared with two endotherms at around 129.7oC and 143.5oC along with a small exotherm (133.8oC) in between the two. Another endotherm is also visible at around 186.5oC of unreacted SY. As the ratio is further increased to 40:60, the position of endo-exo-endo event in thermogram remains the same. However, the endotherm for unreacted SY further decreased to 159.2oC. At 50:50, only endo-exo-endo trio event was appeared without any change in its position. On further increasing the ratio from 60:40 to 90:10, the incident of endo-exo-endo was found to be invariant but a new higher melting endotherm has appeared at 156.5oC, 161.4oC, 165.1oC and 172.4oC at molar concentration of 60:40, 70:30, 80:20 and 90:10 respectively. This peak is due to excess concentration of LZ and it gets broadened due to inherent tendency of form II of LZ to convert into form IV together with presence of eutectic as an impurity. At this point it is appropriate to define two polymorphic forms of LS, namely, LS(A) with invariant solidus melting at 129oC and another LS (B) with solidus melting at 143oC.
Though it was difficult to draw phase diagram (Fig. 5) due to complexity of the DSC endothermic peaks but the overall shape of phase diagram is more or less similar to “V” shape which is characteristic of eutectics. It is because the pure LZ form II is converted into form IV at transition temperature of 158oC, and excess of drug in all the binary mixture where ratio of drug is more than 50:50, the LZ which is initially present in form II is now available in form IV. Phase diagram was constructed by taking solidus point at 143oC. The left arm of V-shaped curve showed liquidus points due excess of coformer and right arm comprised of liquidus points due to excess of LZ form IV. Similarly, phase diagrams for binary mixtures (LF and LN) were constructed. Interestingly, both the systems showed ‘V’ shaped curve characteristic of a eutectic system.
To get more clear understanding of the polymorphic transition occurring in these eutectic systems in real time visuals, hot stage microscopy (HSM) experiment was performed. Microcrystalline powder samples chosen from the batch of LZ and eutectic systems were heated from 20 to 200oC. In case of microcrystals of LZ and LS, quite remarkable changes were observed in crystal habit and surface characteristics, around 158-160oC and 132-134oC respectively. Similarly, LF and LN microcrystal’s also showed morphological changes at different transition temperature range i.e. 141-143oC and 122-124oC respectively. In LF, the changes were observed at 141–143°C, whereas LN showed changes at 122–124°C. This observation are in good agreement with the DSC data and confirms the fact that the small exotherm in the thermograms of LZ, LS, LF and LN are indeed corresponds to the polymorphic transition from form II to form IV in LZ and likewise eutectic to eutectic transformation in LZ eutectic systems. The visual changes such as in morphology, transparency/opaqueness, etc., observed in the samples during heating were recorded and are shown in Fig. 6.
3.3 QPA analysis of eutectic polymorphs
PXRD is a powerful tool for the determination of phase composition, as it is able to detect periodic molecular structures in a given material owing to constructive X-ray beam diffraction. Therefore, crystalline structures lead to distinct peak intensities in the diffractograms. This feature has been used for the quantification of different crystalline phases within the prepared eutectic polymorphs. For this, Rietveld method has been used, which consists in refining parameters of a structural model, to supply a calculated PXRD pattern similar to the experimental pattern. Based on a structural model, the net result of the comparison between the experimental and simulated PXRD pattern, is termed as residue, which is kept at minimum. In the present study, the crystal structures of LZ form II and form IV are quite different, as displayed in Fig. 7, in which form II crystallizes in orthorhombic crystal system with space group P212121, and form IV in triclinic crystal system with space group P1. Consequently, their PXRD diffraction patterns also possess significant variations.
To adequately identify and quantify the amount of LZ (either form II or IV) together with coformers (SY, FA and INC) present in the 50:50 molar concentration eutectic mixtures of LS(A), LF(A) and LN(A) with invariant solidus melting’s at 129oC, 135oC and 119oC were subjected to QPA analysis (Fig. 8). Visual inspection of the diffraction pattern of sample LS(A) showed great similarity with SY and orthorhombic form II of LZ. QPA analysis for LS(A) resulted in 62% LZ form II and 38% SY percentage weights and there was no trace of form IV is present. Similarly in case of LF(A) and LN(A), the percentage weights were found to be 73% and 71% for LZ form II along with 27% and 29% for FA and INC respectively. Here also not any trace of LZ form IV was present. These calculated percentage weights compositions are reasonable with the experimentally prepared weight amounts. In the same way, the heat treated samples of eutectic mixtures i.e. LS(B), LF(B) and LN(B) with invariant solidus melting at 143oC, 154oC and 128oC were also subjected to QPA analysis (Fig. 9). The percentage weights for LS(B) is 61% LZ form IV and 38% SY while in case of LF(B) and LN(B), the percentage weights were found to be 72% and 71% for LZ form IV along with 26% and 28% for FA and INC respectively.
The final phase compositions and reliability factors are shown in Table 2. Illustrations of the diffraction patterns displaying the refined fits along with percentage weight compositions appear in Figs. 8 & 9. From Table 2, the results for LS(A), LF(A), LN(A), LS(B), LF(B) and LN(B) indicates that the refinement converged adequately well, with R-factors and goodness-of-fit indicators being reasonably good. The observed Rwp index for all the eutectic mixtures ranged from approximately 7.0–11.0%, indicating very good agreement between the experimental and simulated PXRD patterns. It was evident from the QPA analysis that pure polymorphic form II is present in 50:50 binary mixtures of LS(A), LF(A) and LN(A) without any contamination of polymorphic form IV of LZ. In case of heat treated samples of binary eutectic mixtures, there is pure polymorphic form IV with no contamination of polymorphic form II of LZ. This implies that QPA analysis is fair enough; in identifying polymorphism and quantifying the polymorphs of LZ within the eutectic mixtures. The results are also consistent in respect of the calculated and experimental percentage weights of different phases within all the eutectic samples. Finally, this investigation unambiguously validate the existence of heat mediated polymorphism in eutectic mixtures with the presence of orthorhombic form II of LZ and respective coformers (SY, FA and INC) in eutectic samples together with triclinic form IV of LZ and coformers (SY, FA and INC) in heat treated eutectic samples. To our knowledge no quantification of the phase composition of a binary eutectic mixture involving a polymorphic transformation has been carried out.
Table 2
The refined percentage phase compositions and reliability factors for various LZ eutectic polymorphs.
LZ Eutectic polymorph | Form II of LZ (%wt) | Form IV of LZ (%wt) | FA (%wt) | INC (%wt) | SY (%wt) | Rp (%) | Rwp (%) | χ2 |
LS(A) | 62 73 71 - - - | - - - 61 74 70 | - 25 - - 25 - | - - 27 - - 27 | 36 - - 36 - - | 8.8 | 9.2 | 1.6 |
LF(A) | 8.9 | 9.4 | 1.8 |
LN(A) | 10.2 | 10.9 | 1.9 |
LS(B) | 8.6 | 10.1 | 1.9 |
LF(B) | 10.1 | 9.3 | 1.7 |
LN(B) | 9.6 | 10.8 | 1.8 |
3.4 Investigating the morphological changes of eutectic polymorphs
Figure 10 depict the scanning electron microscope (SEM) images of LZ form II and its eutectic mixtures LS(A), LF(A), LN(A) together with heat treated polymorphic form IV of LZ and polymorphic eutectic samples LS(B), LF(B), LN(B). It is evident that the LZ form II appears as large plate like structures in the size range of 10–40 µm while its heat treated polymorphic form IV appears as small plate shaped ministructures with smaller size of c.a. 10–20 µm. The eutectic mixtures LS(A), LF(A) and LN(A) seem to be like rough clusters of form II of LZ and respective coformers along with ruptured surfaces. Moreover, the size of the particles was reduced to 10 µm, as expected. This is due to the applied grinding shear force during preparation of eutectic mixtures. In case of eutectic polymorphs LS(B), LF(B) and LF(B), interesting changes in surface texture was observed, the particles have seen to get integrated from small clusters into big bouquet. This might have occurred due to the polymorphic transformation of eutectic mixtures on heat treatment.
3.5 Solubility and Intrinsic dissolution rate study
Aqueous solubility has long been recognized as a key factor in controlling drug efficacy, design of parentral and ophthalmic formulation as well as important in controlling taste. Before an orally administered drug can become available to its receptor, it first must dissolve in the gastrointestinal fluid. Both the dissolution rate and the maximum amount of drug that can be dissolved are governed by the solubility of the drug in the medium. The lack of sufficient aqueous solubility often causes a drug to appear inactive or less active. Eutectics are shown to have high solubility and faster dissolution because of their high free energy, greater molecular mobility, and weaker intermolecular interactions. As observe in the Table 3, the solubility of low melting LZ form II and eutectic mixtures LS(A), LF(A), LN(A) is higher than that of high melting LZ form IV and eutectic polymorphs i.e. LS(B), LF(B), LN(B). As noted in the literature [40], the solubility of a compound in water is strongly correlated with its melting point. Generally, a compound with low melting point has high solubility and vice-versa. The solubility results for eutectic mixtures and their polymorphs (Fig. 11) are consistent with the fact of melting point-solubility correlation. This correlation is quite system dependent and not universal in nature. The FTIR examination (Fig. S1 in supplementary information) of the residue after solubility experiment of all eutectic forms showed the presence of characteristic peaks of LZ form II and corresponding coformers. This suggests that the LZ form IV present in the solid form of heat treated eutectic polymorphs LS(B), LF(B), LN(B) converts to form II of LZ in water.
Table 3
Solubility, intrinsic dissolution rate of linezolid eutectic polymorphs along with their melting point.
Compounds | Melting endotherm in DSC (oC) | Polymorphic transformation exotherm in DSC (oC) | Coformer m.p. (oC) | Solubility (mg L− 1) ± SD | Intrinsic dissolution rate (mg cm− 2 min− 1) | Coformer solubility (mg L− 1) |
LZ form II | 155 | 159 | - | 2968 ± 10 | 0.52 | - |
LS(A) | 129 | 132 | 205–209 | 4573 ± 12 | 0.63 | 5780 |
LF(A) | 136 | 139 | 286–288 | 6921 ± 10 | 0.64 | 6300 |
LN(A) | 119 | 122 | 155–157 | 101621 ± 32 | 9.4 | 191700 |
LZ form IV | 180 | - | - | 2454 ± 8 | 0.42 | - |
LS(B) | 141 | - | 205–209 | 4181 ± 13 | 0.56 | 5780 |
LF(B) | 155 | - | 286–288 | 5932 ± 10 | 0.58 | 6300 |
LN(B) | 128 | - | 155–157 | 86274 ± 24 | 7.1 | 191700 |
IDR is the method of choice for those drugs that undergo phase transformation (e.g. polymorph, hydrate) or dissociation (cocrystal, eutectic) during slurry solubility measurements. So this is the best way to understand the dissolution behaviors of eutectic polymorphs. IDR experiments on LZ form II, form IV and eutectic polymorphs were performed in purified water (pH 6.3) for 4 hours by the rotating disk intrinsic dissolution rate method at 37oC. The IDR values are given in Table 3. Like solubility, the eutectic mixtures LS(A), LF(A), LN(A) are having high dissolution rate than their corresponding high melting eutectic polymorphs LS(B), LF(B), LN(B). The order of dissolution rate is LN(A) > LN(B) > LF(A) > LF(B) > LS(A) > LS(B) in eutectic polymorphs. Intrinsic dissolution rate curves are displayed in Fig. 12.
The solubility of a binary compound can be affected by the factors like particle morphology and coformer solubility. In the first case, particle morphology may play a role in improving the dissolution of eutectics. From SEM images it is clear that the eutectic mixtures are having smaller size with large surface area that is appropriate for the wetting of particles in eutectic mixtures which ultimately improves the solubility and dissolution rate. On the other hand, in case of polymorphic eutectic forms, there are rosettes or bouquets of particle clusters which are bigger than their corresponding low melting eutectic forms. This leads to small surface area of particles for wetting and further decreases the solubility and dissolution rate. Moreover, the close proximity of two non isomorphous substances as well as accommodation of differently size molecule in their respective individual lattice arrangements in eutectics are accountable in the same manner as they are for amorphous APIs and solid dispersions. Besides, in binary mixtures, the solubility of coformer is very important and the results showed a linear relationship with the coformer solubility. To summarize, the solubility of LZ is increased between 2 and 5 times in linezolid eutectic polymorphs.
3.6 13C NMR spectroscopy
Since, polymorphism is the ability of a solid material to exist in more than one form or crystal structure, it should be characterized in the solid state. Polymorphism can be characterized by various solid state techniques while the solution NMR of different polymorphs of a drug remains identical. For instance, 13C NMR data of LZ (irrespective of form II and form IV) were already reported in literature but carbon peak assignment was done by Maccaroni et al. and they also showed the values of the 13C chemical shifts for the two distinct polymorphic forms, form II and form IV, derived from the solid-state 13C spectra. This quality is used here to validate the polymorphic behavior of prepared eutectic mixtures. Examination of the 13C NMR spectra (Supplementary file) of LZ form II, form IV, ground eutectic mixtures (LS(A), LF(A) and LN(A)) and their heat treated eutectic polymorphs (LS(B), LF(B) and LN(B)) showed the presence of a unique spectrum, which is resulting from the simple blend of LZ form II with respective coformers (FA, INC and SY). This suggests the conversion of LZ form IV to form II in solution phase which results in the identical chemical shifts. Table S2, shows the carbon atom resonances for 13C NMR spectra of LS(A), LF(A), LN(A), LS(B), LF(B) and LN(B) along with the carbon resonances in LZ form II, FA, INC, SY obtained in DMSO- deuterated solution which were already reported in literature (NMR spectra are given in supplementary information).
3.7 Accelerated stability study
The physical stability of prepared eutectics polymorphs of LZ with selected coformers was established at accelerated conditions of 40°C/75% RH for three months. The PXRD patterns of the exposed samples were analyzed to ascertain any affect of accelerated conditions on the physical integrity of prepared eutectics polymorphs. Observation of no significant changes in pattern of PXRD (Fig. 13) when compared with that of the unexposed sample suggests that the prepared LZ eutectic polymorphs to be stable with retained physical stability under the accelerated conditions.