W.-Q. Feng, L.-Y. Wang, J. Gao et al.
Journal of Molecular Structure 1234 (2021) 130166
interactions in the crystal. They were produced by the Crystal Ex-
plorer software, using the calculated crystal structure parameters
[29]. dnorm, di, and de surface were mapped over a fixed color
˚
˚
˚
scale of −0.552−1.359 A, 1.000−2.838 A, and 0.773−2.809 A, re-
2.2.4. Powder X-ray diffraction (XRPD)
PXRD was performed using a Bruker D8 Advance diffractometer,
˚
using a Cu-Kα (λ = 1.54178 A) source at 40 kV and 400 mA. Each
sample was collected from 5 ° to 50 ° (2θ) at ambient temperature
and scan speeds of 2 ° min−1, And the results are analyzed by Jade
5.0 software.
2.2.5. Fourier transform infrared spectroscopy (FT-IR)
Fig. 2. ORTEP view of the cocrystal with the atom displacement ellispoids drawn
at a 30% probability level, showing the atom numbering. H atoms are drawn as
spheres of arbitrary radii.
The FT-IR spectra of samples were recorded using a Nicolet
model Impact 470 Fourier transform infrared spectrometer and KBr
as pellets. The KBr pellet was used to obtain background spectra.
The range was set from 4000 to 4000 cm−1
.
2.3. Results and discussion
2.2.6. Differential scanning calorimetry (DSC) and thermogravimetric
analysis (TGA)
2.3.1. SCXRD analysis
Single crystal X-ray diffraction reveals DFA-MET crystallized
in the monoclinic P21/c space group. As the Fig. 2 shows, one
molecule of DFA and one molecule of MET constitute the asym-
metric unit connected by N − H•••O hydrogen bond involving an
amidogen of the MET and a carboxylate Ion of the DFA contain-
ing an intramolecular hydrogen bond. The details of the hydrogen
MET are interconnected by N − H•••O hydrogen bonds to form
the centrosymmetric dimer described as R22(8) in graph set no-
tation. And each of these connects an adjacent molecule of DFA
through the R22(8) system and is connected to another a molecule
of DFA N − H•••O hydrogen bond, which form a one-dimensional
(1D) chain. The 1D chains are further connected through R24(8)
(Fig. 3) to form a two-dimensional (2D) layer (Fig. 4). Finally, a lot
of 2D layers repeatedly pile up to form a three-dimensional (3D)
supramolecular structure along a axis by various weak interactions
DSC and TGA measurement of DFA-MET were carried out em-
ploying a Mettler Toledo TGA/SDTA 851e module and a Mettler
Toledo DSC 822e module, respectively. The samples tested were
heated from 50 to 400 °C at a scanning rate of 5 °C min−1 under
a N2 flow of 20 mL min−1
.
2.2.7. Solubility experiments
The solubility of DFA and DFA-MET in the pH 1.2 pH, pH4.0
and pH6.8 buffer solutions at 37.0
0.5 °C was measured by the
shake-flask method. An excess of each sample was placed in an Ep-
pendorf tube, 6 mL of the solvent followed was added and stirred.
After 48 h, the suspension was filtered through a 0.22-μm nylon
filter, and the filtrate was later determined by a Cary 300 spec-
trophotometer at a detection wavelength using a validated analyt-
ical method [31]. The experimental methods were repeated three
times.
2.2.8. Intrinsic dissolution rate (IDR) experiments
Intrinsic dissolution rate (IDR) measurements were carried out
on a RC-6 dissolution tester by the rotating disk method. About
100 mg of pure DFA and DFA-MET were compressed for 1 min at
2.5 ton per inch2 to form the 8-mm-diameter of a nonporous disk
covered by paraffin wax and leaving a flat surface to measure the
dissolution. Then these disks were rotated at 100 rpm in 400 mL
2.4. Hirshfeld surface analysis
The Hirshfeld surface and their 2D fingerprint plots anal-
ysis are effective tools for studying crystal packing. The sin-
gle molecule of DFA in crystal is used as the input for calcu-
lations. Through analysis of this results, four red spots, where
close contacts are formed, are found on the dnorm surfaces of
the single molecule of DFA, which means that there are not the
strong charge-assisted hydrogen bonds between each molecule
of DFA. As shown in Fig. 6, the red spots on the dnorm sur-
medium preheated to 37
0.5 °C at different pH values, respec-
tively. 5 mL of dissolution medium was withdrawn at regular inter-
vals, the same volume of the corresponding buffer solution added
to the original medium. Finally, the concentration of all the sam-
ples filtered through a 0.22-μm nylon filter were measured by UV
analysis and repeated three times.
face correspond to the close contacts due to four N − H•••
O
hydrogen-bonding interactions [29]. As shown in Fig. 7, the dom-
inate surface contacts for DFA in DFA-MET can be split five
ways: H•••H (37.3%), H•••C/C•••H(18.8%), H•••Cl/Cl•••H(17.4%),
H•••O/O•••H(14.3%) and H•••N/N•••H(3.4%). The most of the con-
tributions over the total Hirshfeld surface are H•••H (37.3%) con-
tacts, suggesting that the molecular surface is composed of a sea
of H atoms. A large amount of H atoms over the surface sup-
ported H•••C contacts to be the second one (14.8%). In addition,
the third largest population of contacts is attributed to the H•••Cl
contacts, which shows that the weak hydrogen-bonding interac-
tions are also the major driving forces between 3D supramolecular
structure. In contrast, H•••O strong hydrogen-bonding interactions
account for only 14.3% of total Hirshfeld surface, showing that the
strong hydrogen-bonding interactions are not the main intermolec-
ular forces for the crystal packing. Finally, H•••N contacts play only
a small part in the crystal packing.
2.2.9. Permeability experiments
Permeability experiments of DFA and DFA-MET were measured
by the modified Franz diffusion cell apparatus through a cellulose
nitrate membrane (0.45 mm, Sartorius, Germany). The membrane
was placed in between the donor compartment and the recipient
compartment to which 5 mL of buffer medium (pH 6.8) was added.
After the buffer medium was kept at 37 0.5 °C and rotated at
50 5 rpm. About 20 mg powder samples were placed on the
membrane. The sample solution (0.5 mL) was extracted from the
receptor chamber every 1 h and replaced with the same volume of
buffer medium every 1 h over 8 h. Finally, the concentration of DFA
and DFA-MET were measured by UV analysis from the respective
calibration plots. The experimental methods were repeated three
times.
3