In order to prove the accurate CMOF synthesis powder XRD and SEM images data were obtained. The comparison with [32] has shown the triangular form of the crystals in both cases (Figs. 1). Additionally, five peaks at 6.9, 8.0, 9.8, 12.2 and 14.4 deg were found (Fig. 2). It corresponds to XRD data from[32]. So, SEM and XRD data have supported the CMOF accurate synthesis.
In order to estimate the temperature limit for measurements, TGA analysis was performed. The data on weight losses are presented in Fig. 3 (differential curve) and Fig. 1S (integral curve). Up to 150°C, the mass loss was insignificant (less than 3%). This allows the determination of a temperature range for IGC measurements up to 150°C. Two maxima are observed on the curve (Fig. 3). The first one appears at 250°C, and the second one at 400°C. The first maximum may be attributed to the loss of crystal water, while the second could result from partial decomposition of MPMD. The weight loss at temperatures above 600°C is likely due to the complete destruction of the framework.
A difference in the adsorption of limonene enantiomers was observed on SU-MB (Fig. 4, Figs. 3S–5S). At 130 and 150°C, the apparent difference in limonene adsorption was maximal. As with other samples exhibiting supramolecular chirality [47, 53, 54], enantioselectivity increased with adsorption values. To assess the reliability of the observed chiral recognition, a t-test was employed. The p-values are listed in Table 1 and Tables 1S–4S. At 110°C, no chiral recognition was observed up to 400 Pa (Table 1). However, at higher pressures, a statistically significant difference in adsorption values was detected. At other temperatures, p < α at all pressures, indicating enantioselectivity even in cases where no visible difference was apparent in the figures. Enantioselectivity in a wide range of pressures is observed only for porous chiral crystals [39, 45]. In pores, a smaller amount of substance is required to form an enantiomer layer, enabling chiral recognition even at low enantiomer concentrations.
For limonenes, no dependence of the enantioselectivity coefficient (α) on temperature was observed. The α values ranged from 1.06 at 110°C to 1.07 at 150°C. This is unusual, as enantioselectivity generally decreases with temperature[55]. The achieved enantioselectivity was similar to that reported for the [{Cu12I(trz)8}4Cl 8H2O]n MOF sample [39]. Thus, for limonenes, the studied sample exhibited a typical enantioselectivity level.
All isotherms in Figs. 4 and 3S–5S can be classified as BET type III. Such isotherms are well approximated by the Freundlich equation. The results of the approximation are presented in Table 2. All isotherms were fitted with high correlation coefficients (R ≥ 0.99). The Freundlich constants differed beyond the margin of error at 130, 140, and 150°C, corresponding to the isotherms with the highest enantioselectivity. Generally, adsorption isotherm constants derived from experimental data are less sensitive to enantioselectivity, and at low enantioselectivity levels, they rarely differ significantly.
For α-pinene, the difference in enantiomer vapor adsorption was higher than for limonenes (Figs. 5–6, Figs. 6S–10S). At 90 and 110°C, the first adsorption isotherm measurements for α-pinenes exhibited BET type I behavior (Fig. 5, Fig. 6S). Differences in enantiomer adsorption were observed at both low and high pressures (Table 3, Table 4S). However, at certain pressures (up to 292.3 Pa, 113–226.1 Pa, and 146.1 Pa at 90°C; 153.8 and 217.5 Pa at 110°C), no significant difference was detected.
Subsequent measurements at 120°C still showed BET type I isotherms (Fig. 9S). At this temperature, the adsorption of α-pinene enantiomers was statistically different at all partial pressures. However, after heating at 120°C, the isotherm shape at 90–110°C changed to BET type III (Fig. 6, Figs. 7S–8S). This shift from type I to type III indicates weakened adsorbate-adsorbent interactions, likely due to the collapse of the smallest pores at 120°C. Statistical analysis of enantiomer adsorption differences in subsequent measurements revealed a broader enantioselectivity range. For instance, at 110°C, adsorption values differed significantly at all pressures (Table 4), as was the case at other temperatures (Tables 5S–6S).
In all measurements at 120 and 130°C, the isotherms had a BET type I shape. The change in isotherm shape upon reaching 120°C suggests an uneven weakening of adsorbate-adsorbate and adsorbate-adsorbent interactions with increasing temperature. Since adsorbate-adsorbate interactions weaken faster than adsorbate-adsorbent interactions, the latter eventually dominate, altering the isotherm shape.
Table 1
P-values for pairs of points of limonenes vapors adsorption at 110°C (α = 0.05, n = 5)
Р, Pa | р | Р, Pa | р | Р, Pa | р |
|---|
227 | 0.0906 | 245 | 0.1699 | 263 | 0.2861 |
335 | 0.9337 | 400 | 0.004 | 417 | 0.0045 |
435 | 0.0049 | 453 | 0.005 | 471 | 0.005 |
489 | 0.0046 | 508 | 0.0043 | 526 | 0.0035 |
Table 2
Approximation parameters of limonenes adsorption isotherms by Freundlich equation
| | S-(-)-limonene | R-(+)-limonene |
|---|
Kf· 104 | n | R | Kf· 104 | n | R |
|---|
110 | 80 ± 10 | 1.34 ± 0.02 | 0.998 | 100 ± 10 | 1.28 ± 0.02 | 0.9993 |
130 | 43 ± 6 | 1.30 ± 0.02 | 0.9981 | 65 ± 7 | 1.23 ± 0.02 | 0.9982 |
140 | 51 ± 6 | 1.23 ± 0.02 | 0.9987 | 69 ± 6 | 1.18 ± 0.01 | 0.9993 |
150 | 59 ± 5 | 1.16 ± 0.01 | 0.9992 | 79 ± 4 | 1.11 ± 0.01 | 0.9998 |
Table 3
P-values for pairs of points of α-pinene vapors adsorption at 110°C (α = 0.05, n = 5, first measurement)
Р, Pa | р | Р, Pa | р | Р, Pa | Р |
|---|
64 | 0.0451 | 154 | 0.23 | 218 | 0.057 |
281 | 0.0227 | 371 | 0.0296 | 435 | 0.0122 |
499 | 0.0058 | 589 | 0.0045 | 653 | 0.0028 |
This phenomenon has been previously reported by our group for both non-porous and porous chiral crystals [45, 46]. At 120°C, the adsorption isotherms of α-pinene enantiomers differed at all partial pressures, whereas at 130°C, the adsorption values coincided. The decrease of chiral recognition at higher temperatures can be attributed to the proximity to α-pinene’s boiling point. A surface with supramolecular chirality cannot recognize enantiomers in the gaseous state
Table 4
P-values for pairs of points of α-pinene vapors adsorption at 110°C (α = 0.05, n = 5, second and any further measurements)
Р, Pa | р | Р,kPa | р | Р, Pa | р |
|---|
108 | 0.00001 | 328 | 0.00001 | 392 | 0.00004 |
431 | 0.0001 | 504 | 0.0001 | 537 | 0.0001 |
574 | 0.0003 | 593 | 0.0003 | 610 | 0.0005 |
Table 5
Approximation parameters of α-pinenes adsorption isotherms by the Freundlich equation
T | (-)-α-pinene | (+)-α- pinene |
|---|
Kf·104 | n | R | Kf·104 | n | R |
|---|
90 | 272 ± 2 | 1.10 ± 0.01 | 0.9999 | 220 ± 10 | 1.15 ± 0.01 | 0.9996 |
100 | 243 ± 5 | 1.07 ± 0.01 | 0.9999 | 180 ± 8 | 1.14 ± 0.01 | 0.9997 |
110 | 189 ± 4 | 1.06 ± 0.01 | 0.9999 | 136 ± 6 | 1.13 ± 0.01 | 0.9998 |
Table 6
Approximation parameters of α-pinenes adsorption isotherms by the Langmuir equation
T | (-)-α-pinene | (+)-α- pinene |
|---|
KL·102 | am | R | KL·102 | am | R |
|---|
90 | 64 ± 2 | 34 ± 3 | 0.9996 | 57 ± 2 | 39 ± 3 | 0.9995 |
110 | 11 ± 1 | 35 ± 3 | 0.9998 | 11 ± 1 | 38 ± 3 | 0.9996 |
120 | 2.3 ± 0.5 | 42 ± 6 | 0.9822 | 0.19 ± 0.02 | 49 ± 5 | 0.9901 |
130 | 0.17 ± 0.02 | 108 ± 9 | 0.9955 | 0.17 ± 0.02 | 112 ± 9 | 0.9961 |
because the adsorbed layer behaves as a two-dimensional van der Waals gas, lacking short- or long-range order [56].Comparison of limonene and α-pinene isotherms reveals that for limonenes, chiral recognition persisted even at 150°C (27°C below their boiling point), particularly at high partial pressures. In contrast, for α-pinenes at 130°C (25°C below boiling), no chiral recognition was observed. This suggests that the adsorption layer of the bulkier α-pinene molecule is less stable than that of limonene.
Adsorption isotherms at 90–110°C from subsequent measurements were approximated using the Freundlich equation (Table 5). The approximation quality was high (R ≥ 0.9996), and adsorption constants differed at all temperatures, decreasing with temperature for both enantiomers. For BET type I isotherms (90–130°C), R values were also high. The monolayer capacity (am) remained constant at 90 and 110°C but increased at higher temperatures (Table 6). At all temperatures, am values for both enantiomers
Table 7
Enantioselectivity coefficients for limonenes and α-pinenes
T | α |
|---|
limonenes | α-pinenes |
|---|
90 | - | 1.12 |
100 | - | 1.09 |
110 | 1.06 | 1.15 |
120 | - | 1.06 |
130 | 1.07 | 1.00 |
140 | 1.05 | - |
150 | 1.07 | - |
160 | 1.05 | - |
coincided, as expected. However, Langmuir constants at 90 and 120°C were not equal, supporting the conclusion of enantioselectivity.
When comparing enantioselectivity of limonenes and α-pinenes, α-pinenes exhibited a maximum α of 1.15 at 110°C, whereas limonenes reached only 1.07 (Table 7). Data at 130°C could not be compared due to the loss of α-pinene enantioselectivity near its boiling point. The higher enantioselectivity for α-pinenes may result from a greater difference in the isosteric heats of adsorption (Qst) between enantiomers (Fig. 7). For limonenes, Qst increased with adsorption, but the difference was notable only at low coverages. In contrast, for α-pinenes, Qst exhibited a more pronounced divergence, suggesting a different adsorption mechanism.
For limonene enantiomers, Qst values at high coverages approached the heat of liquefaction (L), consistent with typical adsorption behavior. At coverages > 1, the second and further adsorption layers form, for which Qst≈L. A similar trend was observed for (−)-α-pinene, but not for (+)-α-pinene, indicating that the latter did not form a complete monolayer. This difference in monolayer formation likely contributed to the additional enantioselectivity observed for α-pinenes.