Using water-mimic organic compounds to activate guest inclusion by initially dry beta-cyclodextrin

Abstract

Optimal conditions were found enabling anhydrous beta-cyclodextrin (bCD) to include target guests using small monofunctional organic compounds instead of water. Structural criteria were specified for organic substances with such water-mimic behavior. For this, a thermodynamic description of guest and water inclusion by initially dry bCD in binary systems was given using experimentally determined vapor sorption isotherms. These data perform a cooperative inclusion of each water-mimic guest and water with phase transition, and give the values of inclusion and hydration Gibbs energy, respectively. The observed inclusion cooperativity in binary systems with bCD defines a specific size-exclusion effect banning monofunctional organic compounds from entering the dry bCD phase if they exceed a threshold value of molecular size parameter near that of acetone. For larger guests, this threshold was shown to be removed in ternary systems by simultaneous inclusion with water-mimic guests or by solid-phase exchange of such guests. As well as water, water-mimic organic compounds activate the inclusion of target guests by initially dry bCD just by forming ternary clathrates, thus making all other hypotheses on the role of water and its mimics in this inclusion process excessive. These procedures may be useful for practical purposes when the presence of water does not give good results in clathrate preparation with bCD. This may provide a way for developing new techniques for the preparation of beta-cyclodextrin clathrates with various organic compounds.

---

Introduction

Cyclodextrins constitute one of the best studied and applied classes of receptors,¹ which need water to perform their inclusion properties. The role of water is a key element of available theoretical models used to rationalize preparation methods for cyclodextrin complexes.^2–4 These models are based on experimental data for water solutions,^3,5,6 while for practical applications the complexation products are often prepared in the solid state from solid cyclodextrins.^7,8 The corresponding solid-phase inclusion process has several specific features, which have analogies in properties of other hydrophilic receptors, like proteins^9,10 and cross-linked hydrophilic polymer¹¹ under low-water conditions. These are a cooperative hydration effect, favorable for inclusion of hydrophobic compounds^9–11 and an inclusion ability activated by organic “water-mimic” solvents.^9,11 Proteins perform also a binding cooperativity.¹² All these features need to be studied for cyclodextrins to provide an adequate picture of their inclusion process and a guideline for development of clathrate preparation methods.

The most important property of a solid hydrophilic receptor, a cooperative hydration effect was observed for initially dried beta-cyclodextrin (bCD),^13–16 which corresponds to an impossible process 1 but efficient process 2 in Fig. 1. For comparison, hydration induces enzymatic activity^17,18 and inclusion capacity of initially dry proteins^9,10 above a threshold value of humidity. To find, whether water is a unique solvent in this relation, and, if yes, what is an extent of its special properties, one needs to compare hydration effect with effect of water-mimic solvents. For enzymes, an ability to function in such nonaqueous media became a basis of technological breakthrough to an easy and efficient synthesis of enantiopure medical drugs.^19,20 For anhydrous beta-cyclodextrin, its inclusion capacity for D-limonene was found to be activated when suspended in this guest solution in liquid ethanol at an optimal ethanol/bCD ratio.¹⁴ But cyclodextrins dissolved in organic solvents form in many cases less stable complexes than in water.⁵

Fig. 1 Scheme of the bCD clathrate preparation procedures.

To meet these ends, water-mimic compounds should be found as capable to activate inclusion properties of dry bCD without its dissolution. Besides, the optimal conditions for this activation should be specified. This study was made in the present work both for consecutive inclusion of water-mimic and large/hydrophobic molecules in processes 3 + 4 or in simultaneous process 5, Fig. 1.

As well as for proteins,^9,10 a chosen criterion of water-mimic property is an ability of compound to be included by bCD in absence of water, process 3, Fig. 1. In the present work, the structure–property relationship was investigated, which defines the range of guests included by bCD in such binary systems. For these relationships, the values of Gibbs energy for guest inclusion by dry bCD and bCD hydration are needed. These data have been not available even for bCD hydrates, which were otherwise extensively studied by a number of methods.^21–27

Determination of Gibbs energy for guest or water inclusion with formation of solid clathrates requires an adequate thermodynamic description of this process. Specific feature of guest inclusion process by dry bCD is its cooperativity.¹⁶ This property may be called biomimetic, since it was observed for dissolved hemoglobin and some other allotropic proteins.¹² Dry cyclodextrins, being crystalline²⁴ unlike dry proteins,²⁸ should perform such cooperativity per se, just according to Gibbs phase rule. Dried bCD has such a behavior in binary systems with ethanol, methanol and acetonitrile.¹⁶ Still, no clear experimental evidence of such cooperativity has been obtained for such a well-studied process as cyclodextrin hydration. Available hydration isotherms for cyclodextrins differ much in shape, especially for bCD as studied by different authors.^24–27 In present work, inclusion Gibbs energies were determined for dried bCD in its binary systems with water and organic guests using formalism offered elsewhere.^29,30

Comparison of bCD hydration process and saturation of this host with water-mimic compounds, as well as comparison of activating effects of these substances on the guest inclusion by initially dry bCD can help to understand the role of water in complexation and inclusion by this host. Hence the problems with the extrapolation of thermodynamic data on complexation in solution^5,31,32 to the properties of solid complexes may be resolved in significant extent. Corresponding conclusions may be used to explain a similar role of water for proteins and cross-linked hydrophilic polymers. Besides, the use of water-mimic guests may give a way for development of new and effective preparation methods for inclusion compounds of cyclodextrins.

Experimental

Materials

β-Cyclodextrin (bCD), ICN, Cat. No. 190053, was dried at 373 K for 8 h in vacuum of 100 Pa before experiments. In dried bCD, thermogravimetry with mass-spectrometry of evolved vapors performs a hydration level less than 0.5% w/w and an absence of volatile guests. Organic guests were purified and dried as described in ref. 33. The purity of guests checked by GC was at least 99.5%.

Sample preparation

bCD hydrates and clathrates were prepared by saturation of dried bCD or anhydrous bCD clathrate with vapors of water or organic guest in sealed 15 mL vial. In each case, liquid sorbate had not any direct phase contact with solid host. For determination of sorption isotherms in binary systems with organic guests, dried bCD samples, 0.100 g, were equilibrated for 72 h at 298 K separately with vapors of varied amounts of guest liquid as described elsewhere,¹⁶ so that the guest was completely evaporated at equilibration. For determination of water sorption isotherm, dried bCD samples, 15–20 mg, were equilibrated with vapors of water solutions in polyethylene glycol (PEG-400) having a known humidity.³⁴ These solutions were taken in large excess, 400 μL, to avoid significant change of aqueous vapor pressure in a host saturation process.

Static method of GC headspace analysis

Static method of GC headspace analysis (HSGC) was used for determination of vapor sorption isotherms in systems with organic guests as described elsewhere.³⁵ Using this method, a relative vapor pressure (thermodynamic activity), P/P₀, of organic guest in the studied systems was determined, where P is partial vapor pressure of guest and is P₀ its saturated vapor pressure. The guest uptake A (mol of guest per 1 mol of dry bCD) was determined as a difference between initial amount of guest added and its contents in vapor phase calculated from a value of and vapor volume. The error of P/P₀ determination is 5%. Guest uptake A was determined with an error of 5% but no less than 0.1 mol per 1 mol of bCD. Each isotherm was determined at least twice with fresh samples of bCD.

Simultaneous thermogravimetry, differential scanning calorimetry with mass-spectrometry

Simultaneous thermogravimetry, differential scanning calorimetry with mass-spectrometry of evolved vapors (TG/DSC/MS) was used for the determination of water sorption isotherm and to check the guest contents in samples of in their inclusion compounds with bCD. This experiment was performed as described elsewhere using STA 449 C Jupiter (Netzsch) device coupled with quadrupole mass-spectrometer QMS 403 C Aeolos.³⁶ In thermal analysis, the sample temperature was first scanned with a rate of 10 K min⁻¹ up to 250 °C, which followed by isothermal mode of heating at this temperature for 20 min and further heating with the same rate to 285 °C. In this experiment, a continuous purge with argon of 75 mL min⁻¹ was used.

For ternary clathrates with strongly overlapping peaks on ion curves or for clathrates with two organic guests, an additional MS-calibration was used. In this calibration, a guest liquid mixture was sampled directly to TG/DSC/MS device in the isothermal mode at 120 °C. A ratio of guest peaks on ion curves was used for calculation as the ratio of MS sensitivity to the guests studied. The bCD hydration and contents of organic guest A were estimated with an error of 0.2 mol per 1 mol of bCD.

X-ray powder diffractogram

X-ray powder diffractograms were determined by a Rigaku MiniFlex 600 diffractometer equipped with a D/teX Ultra detector. In this experiment, Cu Kα radiation (30 kV, 10 mA) was used, Kβ radiation was eliminated with Ni filter. The diffractograms were determined at room temperature in the reflection mode, with scanning speed 5° min⁻¹. Clathrate samples were loaded into a glass holder. Patterns were recorded in the 2θ range between 3° and 60° without sample rotation. The most diffractograms were determined also with addition of standard silicon powder SRM 640d, and corresponding corrections were applied to 2θ values, ESI.†

Results and discussion

The choice of water-mimic guests by study of binary systems with dried bCD

To find guests capable of inclusion by dried bCD without its dissolution under water-free conditions, vapor sorption isotherms of acetone, nitromethane and 1,1,1,2,2,2-hexafluoroisopropanol (HFIP) by dried bCD were determined using HSGC, Fig. 2. For these guests and for pyridine, dichloromethane and 2-butanone, TG/MS method was used to determine the guest contents in products of bCD saturation with their vapors at P/P₀ = 1, ESI.† Vapor sorption isotherm of water was determined by TG/DSC/MS for samples of initially dried bCD equilibrated with vapors of aqueous PEG-400 solutions having known humidity P/P₀ values,³⁴ Fig. 2a. Detailed TG/DSC/MS data for each point of this isotherm are given in ESI.† For comparison, vapor sorption isotherms of methanol, ethanol and acetonitrile, which were determined under the same conditions,^16,37 are also shown in Fig. 2b–d.

Fig. 2 Vapor sorption isotherms by anhydrous bCD of (a) water, (b) methanol,¹⁶ (c) ethanol,³⁷ (d) acetonitrile,¹⁶ (e) nitromethane, (f) acetone, (g) hexafluoro-2-propanol (HFIP) determined using TG/MS (water) and HSGC (organic guests) methods at 298 K. Solid lines are fitting curves calculated by eqn (2). Dashed lines correspond to capillary condensation of guest in bCD powder and formation of liquid solution with HFIP. Filled points are TG/MS data on organic guest contents in products of bCD saturation at P/P₀ = 1 from Table 2.

The obtained vapor sorption isotherms for all studied guests except HFIP have one or two steps of guest inclusion by its thermodynamic activity (relative vapor pressure) P/P₀ corresponding to phase transition from initially dried bCD or intermediate inclusion compound to a stable clathrate, Fig. 2. Isotherms of acetone and nitromethane by dried bCD, Fig. 2e and f, and earlier determined³⁷ isotherm of ethanol, Fig. 2c, have a sigmoidal shape with one inclusion step.

Two steps are observed for isotherm of bCD hydration, Fig. 2a, similar to earlier obtained sorption isotherms of methanol and acetonitrile,¹⁶ Fig. 2b and d. Hydration value S = 9.2 mol of water per 1 mol of host at the first step, Table 1, corresponds to the composition bCD·9.35H₂O of “dry” hydrate from single crystal X-ray diffraction data.³⁸ This step has a clearer threshold than that of hydration isotherms determined earlier for bCD by different authors.^24–27 The second saturation step of bCD with resulting water content of S = 11.5 gives the same hydrate as bCD·11 ± 0.5H₂O crystallizing from aqueous solution.^16,39

Table 1 Parameters of sorption isotherms and clathrates prepared in binary systems with guest vapors and dried bCD at 298 K

Guest	MRD cm^3 mol^−1	S mol per mol bCD	a0.5S	N	δ^a	ΔGc^b/kJ mol^−1	STG/MS
a δ is a standard deviation of the fitting for the shortest distances between experimental points and the line calculated by eqn (2) in normalized coordinates, ESI.b In brackets, ΔGc values for separate inclusion steps are given.c For the 2nd hydration step, parameters a0.5S, S and N were taken arbitrarily to obtain their minimal values in the fitting procedure.d Parameters of separate inclusion steps.e In brackets, the values of guest contents in the saturated intermediate clathrate (hydrate) are given.f Parameters for the isotherm from ref. 37.g TG/MS data on guest contents in a saturated clathrate were used in isotherm fitting because of a too short saturation part of an isotherm.
H2O	3.7	11.5 (9.2)^c^,^e	0.26; 0.97^c^,^d	3.5; 91^c^,^d	0.01	−2.7 (−3.3; −0.1)	11.2
MeOH	8.3	4.2 (0.6)^e	0.01; 0.19^d	3.3; 4.2^d	0.02	−5.2 (−11.4; −4.2)	4.1
EtOH^f	12.9	2.9	0.31	3.5	0.04	−2.9	2.6
MeCN^g	11.1	2.30 (0.93)^e	0.2; 0.53^d	1.10; 13.7^d	0.01	−2.5 (−4.0; −1.5)	2.1
MeNO2^g	12.5	2.09	0.34	3.2	0.02	−2.7	2.0
Me2CO	16.1	1.15	0.38	5.3	0.02	−2.4	1.0

---

Two-step inclusion of water by initially dried bCD, Fig. 2a, Table 1, correlates with the two-step bCD dehydration in thermal analysis performed using simultaneous TG/DSC/MS method,¹⁶ ESI.†

For example, saturated hydrate bCD·11.2H₂O formed in two steps also loses its water in two steps,¹⁶ while clathrate bCD·9H₂O formed in the first hydration step, Fig. 2a, Table 1, has only one step of dehydration.¹⁶ The same was observed for hydrates with lower water contents, ESI.†

One-step isotherms were fitted using Hill eqn (1) adapted for vapor sorption:^29,30

A = SC(P/P₀)^N/[1 + C(P/P₀)^N] (1)

where A is guest uptake, S – guest contents in saturated clathrate in mol of guest per 1 mol of host, N – cooperativity parameter and C – sorption constant. Isotherms with two inclusion steps were fitted by a sum of two eqn (1) as described elsewhere.⁴⁰ This equation is a version of a more general mathematical expression used to describe cooperative processes in biological and other systems.⁴¹

Integration of obtained isotherms by host saturation extent Y = A/S gives inclusion Gibbs energy:^29,30,40

The value of ΔG_c corresponds to a transfer Gibbs energy of 1 mol guest from a standard state of its pure liquid to a saturated inclusion compound. Eqn (2) does not depend on the model used to describe the guest sorption but requires formation of stable inclusion compound with a saturation part of sorption isotherm.

To calculate Gibbs energy ΔG_c directly from fitting parameters of eqn (2) the next equation may be used, which is a consequence of eqn (1) and (2):

ΔG_c = RT ln a_0.5S = −RT(ln C)/N (3)

where a_0.5S is guest activity at 50% saturation extent of a host. For 2-step isotherms ΔG_c value is equal to a weighted average of inclusion Gibbs energies for separate inclusion steps.⁴⁰ Eqn (3) defines a physical meaning of parameters C and N in eqn (2) with N value linked also to a slope of isotherm at its inflection point. Approximation parameters of vapor sorption isotherms including guest contents S, guest threshold activity a_0.5S, and inclusion Gibbs energies ΔG_c are given in Table 1 for saturated and stable intermediate clathrates.

The used experimental HSGC procedure does not allow investigating sorption process just below guest activity P/P₀ = 1. More complicated experiment with TG/MS analysis of bCD hydrates helps to study bCD hydration at high values of this parameter. The second hydration step observed in this experiment at high humidity has a hydration Gibbs energy of only ΔG_c = −0.1 kJ mol⁻¹. This slight change in Gibbs energy for water molecules included above 9.2 mol per 1 mol of bCD implies a very small change in host packing and state of water compared with its pure liquid, which correlates with no phase transition observed earlier in structural studies of bCD dehydration from 12 to 10.5 waters per host.⁴²

The results obtained show the ability of dried bCD to include only small molecules in vapor–solid inclusion process without formation of liquid solution. Such inclusion was observed only for water and 5 monofunctional organic guests: methanol, ethanol, acetonitrile, acetone and nitromethane having the values of molar refraction MR_D ≤ 16.1 cm³ mol⁻¹, Fig. 2, Table 1. Molar refraction MR_D calculated by Lorentz–Lorenz equation from molar volume, density and refraction index of guest liquid is a good molecular size parameter for correlations with parameters of inclusion compounds.^29,30 Bigger molecules, like dichloromethane and 2-butanone, which MR_D values are equal to 16.4 and 20.7 cm³ mol⁻¹, respectively, are not included to the level above 0.3 mol of guest per 1 mol bCD according to TG/MS data given in ESI.† The same negligible level of inclusion was shown previously for propanols, 1-butanol, tert-butanol, propionitrile, chloroform, and C₆–C₇ hydrocarbons¹⁶ with MR_D > 17 cm³ mol⁻¹.

The observed size exclusion effect for those volatile guests, which can be included by dried bCD in absence of liquid phase, can be seen in correlation of guest contents in saturated binary clathrates S vs. guest molar refraction MR_D, Fig. 3. This correlation may be presented as two descending curves separately for hydroxylic and non-hydroxylic compounds with inclusion limit near MR_D = 16.1, Fig. 3. Bigger guests probably cannot find a place inside available or potential empty spaces of bCD packing in absence of water. Besides they are not strong enough H-donors or/and H-acceptors to break host–host hydrogen bonds of bCD. Propionitrile with MR_D = 16.0 cm³ mol⁻¹ is in the threshold region of molecular sizes where subtler factors may be relevant making this guest excluded from dried bCD matrix. Such factors, for example, were discussed for calixarenes, which size exclusion effect has a step-wise shape.^29,30

Fig. 3 Correlation between the guest contents S in saturated clathrates of anhydrous bCD with water and alcohols (○), aprotic guests (◇) and guest molar refraction MR_D. Solid lines are drawn to guide the eye. Dotted line is the inclusion threshold by guest molecular size parameter MR_D.

A cause of the observed larger S values of hydroxylic guests, Fig. 3, may be in a compromise between strength of their H-bonds with bCD and a packing efficiency of inclusion compound, while part of guest molecules is included to fill the space formed at transition to optimal H-bond geometry. Nearly the same model was offered to describe properties of hydrated proteins.⁴³ This explains the ability of bCD hydrated above a threshold water contents to include relatively large and hydrophobic compounds, which are not included by dried bCD.¹⁶ At this, a total volume of such guests and water included is in many cases larger than that of water in saturated bCD hydrate. Such compromise probably exists also in H-bonded matrix of dried bCD being able to include non-hydroxylic acetone, nitromethane and acetonitrile.

The exceptions from the observed size exclusion effect for dried bCD are HFIP and pyridine with MR_D values of 18.2 and 24.2 cm³ mol⁻¹. Unlike other studied guests, vapors of HFIP and pyridine are not included but dissolve dried bCD due to its high solubility in these solvents.^44,45 Partial dissolution of dried bCD in these vapors gives its specific texturing in form of crystal-like aggregates, Fig. 2, ESI.† When dried from excessive liquid solvent, the resulting products have composition of bCD·(∼1.7HFIP) and bCD·6.5C₅H₅N, according to TG/MS data, ESI.† Their X-ray powder diffractograms, ESI,† coincide by position of reflections with that of dried bCD but with much lower average intensity of peaks. The composition of complex with pyridine is close to that of crystals formed from anhydrous gel of bCD in this solvent.⁴⁴

Vapor sorption isotherm of HFIP by dried bCD, Fig. 2g, is typical for an equilibrium of a solid with its saturated solution in a volatile liquid. This isotherm is a vertical straight line crossing the activity axis at P/P₀ = 0.49. No guest uptake is observed below this activity. Saturation of dried bCD with pyridine vapor shows the same behavior. So, the corresponding saturation products can be regarded rather as complexes that cannot be formed without dissolution of dried bCD. Hence, HFIP and pyridine do not fit to the given above definition of water-mimic compounds.

Additional information on the size exclusion effect, which imposes restrictions on the number of possible water-mimic guests for bCD, can be found in structure–property relationships for inclusion Gibbs energies ΔG_c calculated from threshold activity a_0.5S, eqn (1–3). The observed values of ΔG_c (average values for two-step guest inclusion) are very close for all studied guests and water except methanol, being in the range from −2.2 to −2.7 kJ mol⁻¹, Table 1. This parameter for methanol is more negative both as an average value and a value for the second inclusion step, being equal to −5.2 and −4.2 kJ mol⁻¹, respectively. Probably, Gibbs energy ΔG_c = −2.4 ± 0.3 kJ mol⁻¹ corresponds to some optimal host/guest affinity, which, being exceeded by absolute value, creates a possibility to include additional guest molecules. Hence, a large variation in inclusion capacity S is observed for saturated clathrates of dried bCD with different guests, Table 1.

Nearly equal and small average ΔG_c values observed for ethanol, acetonitrile, acetone, nitromethane and water mean that their molecular interactions and a number of H-bonds do not differ much in the initial state of guest pure liquid and in saturated inclusion compound. More negative inclusion Gibbs energies ΔG_c for both inclusion steps of methanol (−11.4 and −4.2 kJ mol⁻¹) and the first inclusion step of acetonitrile imply either stronger host–guest H-bonding or inclusion in bCD molecular cavities without a change of its packing compared with that of the initial dried state. The last case was checked in the present work using XRPD method for saturated anhydrous clathrates of bCD with studied guests and for intermediate clathrate samples with composition of bCD·0.6MeOH and bCD·0.6MeCN formed at corresponding inflection points of sorption isotherms, Fig. 2b and d. Obtained powder diffractograms are given in Fig. 4.

Fig. 4 X-ray powder diffractograms for bCD and its clathrates formed by equilibration of dried bCD in binary or ternary systems with guest vapors at various activities P/P₀: (a) dried bCD; (b) bCD·0.6MeOH at P/P₀ = 0.07; (c) bCD·0.6MeCN at P/P₀ = 0.33; (d) bCD·1.0Me₂CO at P/P₀ = 1; (e) bCD·3.8MeOH at P/P₀ = 0.29; (f) bCD·2.0MeNO₂ at P/P₀ = 1; (g) bCD·1.9MeCN·0.1C₆H₆ at P/P₀ = 0.48 (MeCN) and P/P₀ = 0.06 (C₆H₆); (h) bCD·2.1MeCN at P/P₀ = 1; (i) bCD·2.6EtOH·0.2C₆H₆ at P/P₀ = 0.69 (EtOH) and P/P₀ = 0.04 (C₆H₆); (j) bCD·2.6EtOH at P/P₀ = 1. Clathrate compositions at P/P₀ = 1 are from TG/MS data, ESI.†

Analysis of diffractograms obtained, Fig. 4, confirms supposal on guest inclusion inside available host cavities without phase transition on the first steps of methanol and acetonitrile isotherms, Fig. 2b and d. Comparison of XRPD data for water-free bCD clathrates formed in binary host–guest systems in absence of liquid phase, Fig. 4, shows significant changes in packing are observed in those cases where guest inclusion significantly exceeds 1 : 1 stoichiometry. Formation of intermediate clathrates bCD·0.6MeCN and bCD·0.6MeOH and saturated inclusion compound with acetone bCD·1.15Me₂CO leaves diffractogram of initially dried bCD practically intact, Fig. 4a–d, while inclusion to much higher extent of methanol, nitromethane, acetonitrile and ethanol, Fig. 4e–h, as well as of ethanol¹⁶ and water,²⁴ gives essentially different packing.

Hence, two types of packing are observed for anhydrous binary clathrates: (1) of dried bCD, its saturated clathrate with acetone and intermediate clathrate bCD·0.6MeCN with main peaks at 2θ values of 6.7°, 11.1°, 13.0°, 13.5° and 18.3°, Fig. 4a–d; (2) of nitromethane, acetonitrile and ethanol clathrates with main peaks at 5.6°, 11.2°, 12.8°, 14.6° and 18.2°, Fig. 4f, h and j. Diffractogram of nearly saturated clathrate with methanol bCD·3.8MeOH prepared MeOH activity of P/P₀ = 0.29 has close pattern to the second type of diffractograms in the 2θ range 4.6–13.0° with peaks at 4.6–11° shifted on 0.2–0.3° to the lower 2θ values. Diffractograms of the first type are close to that of dry bCD reported elsewhere.⁴⁶ The XRPD patterns of the second do not have any analogies among the common cage, channel or layer packings observed for inclusion compounds of bCD crystallized from water.^46–48

The obtained results for binary anhydrous guest-bCD clathrates show that few volatile monofunctional organic compounds may be regarded as water-mimic guests, which are capable of inclusion in dry bCD matrix without its dissolution. These water-mimic guests may change or not the packing of initially dried bCD depending on the inclusion stoichiometry.

Effect of the water-mimic compounds on inclusion of hydrophobic guest

The most important property expected from water-mimic guests is their ability to activate the inclusion of more hydrophobic and/or larger molecules in the initially dried bCD, as it has been found earlier for water.^13,14,16 To study this property, vapor sorption isotherms were determined, Fig. 5, using HSGC method for ternary systems bCD–ethanol–benzene and bCD–acetonitrile–benzene with simultaneous sorption of two organic guests according to process 5 in Fig. 1. Benzene was chosen as having too large and too hydrophobic molecules to be included by dried bCD in binary system,¹⁶ Fig. 5c. In the studied systems, the benzene/ethanol and benzene/acetonitrile molar ratio was taken constant along each sorption isotherm and was 1 : 14 and 1 : 15 to simplify the 2D presentation of results and to minimize the possible effect of benzene on the uptake of water-mimic guests.

Fig. 5 Sorption isotherms of benzene vapor in ternary systems with initially dried bCD and simultaneously added vapors of the 3rd component: (a and e) ethanol, (b and f) acetonitrile, and (d and g) water. (c) Vapor sorption isotherm of benzene in binary system with dried bCD.¹⁶ Benzene/ethanol, benzene/acetonitrile and benzene/water molar ratios in ternary systems are constant and equal to 1 : 14, 1 : 15 and 1 : 16, respectively. For the ternary system with water, isotherm from ref. 16 is shown with humidity values P/P₀ being calculated from hydration isotherm in Fig. 2a. Solid lines are drawn to guide the eye. T = 298 K.

The obtained vapor sorption isotherms are presented in two forms: (1) as a normal plot of guest uptake A vs. its activity P/P₀, Fig. 5a and b; (2) as a plot of guest uptake derived to the unity of its activity A/(P/P₀) vs. activity of the water-mimic guest P/P₀, Fig. 5e and f. These two presentations give a complete display of guest uptake A dependence on activities of both guests in the studied systems. For comparison, earlier obtained isotherms for benzene in binary system with dried bCD and its simultaneous sorption with water at benzene/water molar ratio 1 : ​16 by this initially anhydrous host¹⁶ are also given in Fig. 5c and d.

The presence of ethanol and acetonitrile in the studied ternary systems increases much the uptake of benzene by dried bCD, Fig. 5a and b, as compared to practically no inclusion of benzene in its binary system with this host, Fig. 5c. The activating effect of ethanol is significantly higher than that of acetonitrile and even of water, which may be caused by the alcohol proton-donor ability and the presence of alkyl group in its molecule. Besides, the observed threshold of this effect by ethanol and acetonitrile activity is much lower than for water, Fig. 5e–f. In this relation, bCD is similar to proteins, which need a threshold humidity near 50% (P/P₀ = 0.50) or corresponding hydration of 0.1 g g⁻¹ to activate their enzymatic kinetics¹² and sorption of hydrophobic compounds,^9,10 while in the presence of ethanol and acetonitrile the sorption of benzene and dioxane by dry protein human serum albumin is activated without a threshold.⁹ The same was found for cross-linked poly(N-6-aminohexyl acrylamide)¹¹ modeling proteins in this respect. So, the effect of these two water-mimic guests on the inclusion properties of dried bCD may be regarded as biomimetic.

Unlike water, ethanol and acetonitrile much less compete with benzene for the inclusion in bCD. This can be seen by monotonous increase of benzene uptake A and affinity A/(P/P₀) for bCD at the increasing activity P/P₀ of these water-mimic guests, Fig. 5a, b, e and f. The effect of bCD hydration on its affinity for benzene A/(P/P₀) performs maximum near humidity of P/P₀ = 0.65, Fig. 5g. The values of humidity for ternary bCD + benzene + water systems were calculated from hydration isotherm, Fig. 2a, for binary system assuming low influence of benzene on water sorption by initially dried bCD.

Proteins do not have such maximum for volatile hydrophobic molecules.^9,10 The same was observed for cross-linked poly(N-6-aminohexyl acrylamide).¹¹

The assumption of low benzene effect on water sorption may be only partially valid because according to Gibbs phase rule the reduction of guest inclusion threshold by the second guest should work in both ways: if this was found for benzene, then an inclusion threshold a_0.5S of water or water-mimic component in such ternary system should also be lower than in their binary systems with bCD. This benzene effect was studied in the same sorption experiments, which results are shown in Fig. 5. Vapor sorption isotherms of acetonitrile and ethanol determined for the studied ternary systems are shown in Fig. 6 together with isotherms of these guests in binary systems with dried bCD from ref. 16 and 37.

Fig. 6 Vapor sorption isotherms of (a) ethanol at simultaneous sorption of ethanol and benzene at constant benzene/ethanol molar ratio 1 : 14, (b) ethanol,³⁷ (c) acetonitrile at simultaneous sorption of acetonitrile and benzene at constant benzene/acetonitrile molar ratio 1 : 15, (d) acetonitrile¹⁶ by anhydrous bCD determined using HSGC method at 298 K. Solid lines are fitting curves calculated by eqn (2). Dashed lines correspond to capillary condensation of guest in bCD powder.

The comparison of ethanol sorption isotherms by dried bCD in the presence and absence of benzene, Fig. 6a and b, shows that for ethanol the effect of benzene additive is not significant except for the ethanol activity range from 0.20 to 0.27. In this range, benzene shifts ethanol activity to lower P/P₀ values by small increment of 0.07, Fig. 5a. On the contrary, much lower benzene uptake less than 0.1 mol mol⁻¹ gives much more dramatic change of acetonitrile isotherm decreasing a number of inclusion steps from two to one, Fig. 6c and d. Such an effect of the third component taken in small amount was observed for tert-butylcalix[4]arene.⁴⁹

Necessary precondition of such inclusion behavior, formation of one ternary crystalline phase in each studied system,⁵⁰ is met according to XRPD data, Fig. 4. Ternary clathrates bCD·2.6EtOH·0.2C₆H₆ and bCD·1.9MeCN·0.1C₆H₆, Fig. 4g and i, have nearly the same packing as corresponding binary clathrates of ethanol and acetonitrile with dried bCD, Fig. 4h and j.

The cause of the different effect of benzene on the sorption of ethanol and acetonitrile may be, respectively, the lower and higher competition for binding sites in bCD matrix. This can be seen also from relatively large uptake of benzene with the same uptake of ethanol at P/P₀ = 0.8, Fig. 5a, as in its binary system, Fig. 6a.

For dried human serum albumin, the same small amount of benzene doubles the uptake of ethanol and acetonitrile,⁹ while for cross-linked poly(N-6-aminohexyl acrylamide), which can be plasticized by vapors of these water-mimic solvents, no such effect was found.⁵ The used thermodynamic description does not relate to a specific molecular structure of bCD. So it may be applied to any solid hydrophilic matrix with the same key features: flexible packing and strong size exclusion effect in dry state. Such features are intrinsic for proteins.^9,51 Respectively, the observed bCD behavior can be extrapolated to the role of water and water mimic solvents in protein–substrate interaction.

Preparation of bCD clathrates by solid-phase exchange of water-mimic guest

The ability of water-mimic guests to activate the inclusion of compound that cannot be bound in absence of water may be exploited most efficiently in solid-phase exchange procedure 4, Fig. 1. In the present work, this exchange was performed in the systems, where activity P/P₀ of target and water-mimic guests is near unity and zero, respectively. Under these conditions, an inclusion of a larger guest amount by dried bCD may be expected than at the simultaneous sorption of two guests.

The initial clathrates for guest exchange were bCD·2.6EtOH and bCD·2.1MeCN prepared by procedure 3, Fig. 1, in binary system with dried bCD and saturated vapors of water-mimic guest. These clathrates were equilibrated with saturated vapors of volatile organic compounds of different classes and water. Composition and thermal properties of the resulting exchange products were studied using TG/MS method. The results are given in Fig. 7, ESI† and Tables 2 and 3, including total mass loss Δm and its values for separate guest elimination steps, temperatures of MS peaks T_max for each guest released under heating. In those cases, where guest elimination is extended much over temperature scale and no clear MS peak is observed, an average value between onset and endset points of this process on MS curve is given in Tables 2 and 3.

Fig. 7 Curves of TG/MS analysis for products of solid-phase guest exchange in bCD·2.6EtOH (a and b) or bCD·2.0MeCN (c and d) clathrates: (a) bCD·1.9EtCN; (b) bCD·0.4C₆H₆·1.7EtOH; (c) bCD·2.2i-PrOH·0.5MeCN; (d) bCD·2.0n-PrOH·0.7MeCN.

Table 2 Data of TG/MS analysis for bCD clathrates with ethanol and products of ethanol exchange for guest 2 at 25 °C

Guest 2	Clathrate	Δm^a/%	Tmax/°C
			Guest 2	EtOH
a In brackets, mass loss is given on the second step of clathrate decomposition.b Initial clathrate used in guest exchange with data from ref. 16.c An average value between MS peak onset and endset points, (Tonset + Tendset)/2.
—	bCD·2.6EtOH^b	9.6 (4.3)	—	104; 202
H2O	bCD·7.4H2O·0.3EtOH	11.3	103	>250
MeOH	bCD·3.5MeOH·0.1EtOH	9.4	90	190
n-PrOH	bCD·1.9n-PrOH·0.1EtOH	9.3 (3.4)	105; 257	109; 257
i-PrOH	bCD·1.7i-PrOH	8.2 (3.3)	116; 255	120; 246
n-BuOH	bCD·1.3n-BuOH·0.1EtOH	8.2 (4.5)	211^c	141; 227
MeCN	bCD·2.0MeCN	6.7	114	—
EtCN	bCD·1.6EtCN	7.4 (4.0)	109; 242	—
THF	bCD·1.3THF·0.5EtOH	9.5 (6.4)	255	123; >250
Benzene	bCD·0.5C6H6·1.7EtOH	9.2 (3.8)	102; 255	111; 242
n-Hexane	bCD·2.4EtOH	9.0 (4.2)	—	121; 192
Cyclohexane	bCD·0.1c-C6H12·2.2EtOH	8.5 (3.5)	>250	102; 234

---

Table 3 Data of TG/MS analysis for bCD clathrates with acetonitrile and products of acetonitrile exchange for guest 2 at 25 °C

Guest 2	Clathrate	Δm^a/%	Tmax/°C
			Guest 2	MeCN
a In brackets, mass loss is given on the second step of clathrate decomposition.b Initial clathrate used in guest exchange with data from ref. 16.c Onset temperature of bCD decomposition is near 250 °C.d Mass loss above the 1st step.e An average value between MS peak onset and endset points, (Tonset + Tendset)/2.
—	bCD·2.1MeCN^b	7.1	—	103
H2O	bCD·7.9H2O·0.4MeCN	12.3 (1.2)	102	212
MeOH	bCD·4.5MeOH^c	11.3	88	—
EtOH	bCD·3.2EtOH^c	11.6 (3.4)	108; 157	—
n-PrOH	bCD·1.6n-PrOH·1.1MeCN	11.2 (5.9)	115; 253	115; 249
i-PrOH	bCD·2.0i-PrOH·0.8MeCN	12.0 (8.6^d)	84; 115; 164; 256	80; 114; 253
n-BuOH	bCD·1.3n-BuOH·0.2MeCN	8.5 (4.0)	126; 248	117
EtCN	bCD·1.8EtCN·0.1MeCN^c	8.3 (6.1)	105; 232	105
THF	bCD·1.0THF·0.1MeCN	6.4 (3.9)	107; 258	105
Benzene	bCD·1.4C6H6·0.6MeCN	10.6 (6.9)	173^e; 276	102; 189
n-Hexane	bCD·0.1n-C6H14·2.1MeCN^c	8.0	111; 243	114
Cyclohexane	bCD·2.0MeCN^c	6.8	150^c	110

---

The compositions of the clathrates prepared by solid-phase exchange of water-mimic acetonitrile and ethanol in bCD indicate a significant shift of inclusion selectivity compared with corresponding water-free exchange of benzene and interaction between saturated guest vapors and hydrate bCD·11.2H₂O, Fig. 8. While water does not activate inclusion of guests with an intermediate hydrophobicity, log P,⁵² like propanols and propionitrile, these guests may be included to a level near S = 2 by bCD through an exchange of EtOH and MeCN. For propanols with their size parameter being equal to MR_D = 17.5 cm³ mol⁻¹, this S value is close to that expected from the dependence of this parameter on guest molar refraction MR_D for bCD binary inclusion compounds with hydroxylic substances at MR_D, Fig. 3. This is quite comparable with an exchange capacity of bCD·0.9C₆H₆ (ref. 16) and bCD·THF·H₂O (ref. 53) clathrates for propanols and propionitrile. For more hydrophilic guests, the exchange capacity of bCD·2.1MeCN and bCD·2.6EtOH is comparable to that of saturated bCD hydrate, bCD·0.9C₆H₆ (ref. 16) and bCD·THF·H₂O,⁵³ while for studied aliphatic hydrocarbons, activating ability of ethanol and acetonitrile is less effective than that of water, Fig. 8. To make this comparison complete, some experiments on THF exchange were made in this work, ESI.†

Fig. 8 Contents of guest 2 in bCD clathrates prepared by solid-phase exchange of various leaving guests (guest 1) at 25 °C vs. hydrophobicity parameter log P of guest 2. Data for exchange of water and benzene are from ref. 16, and for exchange of THF – from ref. 53, ESI.†

Rather specific is an exchange behavior of water (guest 2) replacing ethanol and acetonitrile (guest 1) in this work as well as earlier studied benzene¹⁶ and THF,⁵³ Tables 2 and 3, Fig. 8. In all these cases, final hydration level is 7.4–7.9 mol of H₂O per 1 mol of bCD, which is much lower than in saturated hydrate bCD·11.2H₂O formed in binary system.¹⁶ Residual contents of leaving guest (guest 1), 0.1–0.4 mol per 1 mol of host in these four cases, is too small to explain this hydration difference by occupying sites of water in hydrate crystals. But these contents may be enough to shift the second inclusion threshold by relative vapor pressure of water to the value above unity, a_0.5S > 1, making impossible an inclusion of the last 3 water molecules above the first hydration step observed for binary bCD–water system, Fig. 2, Table 1. Such increase of guest inclusion threshold a_0.5S in the presence of small amounts of the second guest was observed for tert-butylcalix[4]arene.⁴⁹

The extent of water-mimic nature of ethanol and acetonitrile can be seen from a comparison of their T_max values for exchange products prepared with these, Tables 2 and 3, and other leaving guests and water.^16,53 In those cases, where acetonitrile exchange is not complete, this guest leaves clathrate in one step with MS peak, T_max, at 103–115 °C, Table 3. The fraction of acetonitrile remained encapsulated in bCD after the first step of guest release is negligibly small, which is quite similar to thermal behavior of water in clathrates prepared from saturated bCD hydrate and organic vapors.¹⁶

Water leaves these clathrates in one step with MS peak at 86–119 °C.¹⁶ The thermal behavior of residual ethanol in products of its exchange in water-free bCD is essentially different. In most cases, ethanol evolves from such products in two nearly equal steps with MS peaks at 102–141 °C and 192–257 °C, Table 2, Fig. 7a and b, ESI.† By this property, ethanol is an intermediate as a leaving guest between water and more hydrophobic THF and benzene. The last two guests evolve from the products of their exchange in one high-temperature step with MS peak at 192–252 °C (ref. 53) and 215–236 °C,¹⁶ respectively.

So, acetonitrile is comparable with water as an agent activating the inclusion capacity of bCD for other guests and being able to evolve from a ternary inclusion compound before a target guest. However, the presence of residual acetonitrile, even on the trace level, decreases thermal stability of bCD in a greater or lesser extent. Clathrates prepared by exchange of MeCN exhibit decomposition of bCD at 250 °C. This decomposition is significant for products of its exchange for methanol, ethanol, propionitrile and n-hexane with corresponding increase of water ion curve (m/z = 18) level. For comparison, in thermal analysis of clathrates prepared by exchange of ethanol, bCD remains stable up to its normal decomposition point of 270 °C,⁴ Fig. 7, ESI.†

Lower thermal stability of bCD in the presence of acetonitrile is probably a cause of the higher release temperatures for most guests included through its exchange, Table 3, Fig. 7, ESI.† Almost every second studied guest included by this procedure is retained partially up to 270 °C, where the host begins to decompose. This is not observed for exchange products of ethanol, Table 2, Fig. 7, ESI,† benzene¹⁶ and THF,⁵³ which were used instead of water to activate inclusion by initially dry bCD.

Advantages of using solid-phase exchange of water-mimic guest, procedure 4, Fig. 1, can be seen in comparison of its results with those of procedure 5 with simultaneous inclusion of water-mimic and hydrophobic guests. In the simultaneous inclusion process, target and water-mimic guests compete with each other for their sites in bCD packing. If the target guest is benzene, this competition is stronger for acetonitrile as the excess of this water-mimic guest gives much lower benzene uptake, Fig. 5b, than the same excess of ethanol, Fig. 5a. In exchange procedure, where benzene activity P/P₀ is much higher and equal to unity and that of water-mimic guest is near zero, this inclusion selectivity is inverted. In exchange of acetonitrile, benzene reaches almost the same inclusion level in its clathrate bCD·1.4C₆H₆·0.6MeCN as the value A/(P/P₀) = 1.4 for the highest point of the determined sorption isotherm, Fig. 5f, while exchange of ethanol gives almost 3 times lower benzene uptake with resulting composition of bCD·0.5C₆H₆·1.7EtOH.

Since the observed dependence of A/(P/P₀) value for benzene on its activity and activity of acetonitrile is monotonous, Fig. 5f, an uptake of benzene in solid phase exchange of MeCN should be larger than in any system where dry bCD includes these two guests simultaneously, for example, when suspended in a liquid benzene/acetonitrile mixture. For benzene inclusion in simultaneous process, procedure 5, Fig. 1, ethanol is preferred as a water-mimic guest over acetonitrile. Thus, one may choose between these two water-mimic guests for inclusion of benzene depending on its thermodynamic activity.

Conclusions

The role of water in bCD inclusion properties is made explicit by its comparison with the effect of water-mimic organic compounds. For this, thermodynamic description of bCD interaction with water and organic compounds in binary systems is given, which helped to obtain the first data on Gibbs energy of bCD hydration and of formation of saturated binary clathrates with water-mimic guests in absence of water. In this description, guest (or water) is included by bCD cooperatively with phase transition and corresponding sigmoidal vapor sorption isotherm. The difference in interaction of water and small hydrophilic organic molecules with initially dry bCD is only in inclusion capacity of bCD for these compounds.

Analysis of guest contents in bCD clathrates formed in binary host–guest systems and corresponding inclusion Gibbs energies helped to find the range of water-mimic compounds that can be included by bCD without water. The guest contents in such clathrates fall according to the observed size exclusion effect with increase of guest molecular size at nearly constant inclusion Gibbs energy. For the studied homological series of monofunctional guests, this structure–property relationship has a step: beginning from a certain threshold value of guest size parameter, inclusion Gibbs energy becomes positive making impossible the guest inclusion in binary system.

Present work reveals how these water-mimic guests function in ternary systems, where a target guest is to be included by dry bCD but no inclusion is observed in absence of water. The inclusion threshold by guest activity, being too high in binary systems because the guest is too large/hydrophobic, decreases below unity in the presence of the water-mimic component so that a desired clathrate can be formed.

This mechanism activating inclusion ability of initially dry bCD is the same for organic water-mimics and water itself, and requires formation of a common ternary solid phase consisted of all three independent components to have Gibbs phase rule working in a needed way. As a result, no such explanation as a release of “high energy” water remains necessary. So, a concept of water-mimic guest or solvent becomes more comprehensive for bCD and can be extrapolated to any solid hydrophilic matrix with flexible packing, e.g. protein, which performs a strong size exclusion effect in binary host–guest systems without water.

Obtained results have also significant practical implications. Once compounds that can be included by dried bCD without water are found and a thermodynamic description of their host–guest interaction in solid phase is given, new preparation methods may be developed using these compounds instead of water as activating agents for inclusion of practically important substances. This may be necessary when ordinary procedures with water fail to give good results. In this case, a solid-phase exchange of previously included water-mimic guest may give an effective inclusion of target guest by bCD minimizing competition for its inclusion sites.

Acknowledgements

The work was supported by Russian Government Program of Competitive Growth of Kazan Federal University and by RFBR, grant No. 14-03-01007. The equipment of Federal Center of Shared Equipment of Kazan Federal University was used in this work.

Notes and references

1. G. Crini, Chem. Rev., 2014, 114(21), 10940–10975.
2. E. M. M. Del Valle, Process Biochem., 2004, 39, 1033–1046.
3. L. Liu and Q.-X. Guo, J. Inclusion Phenom. Macrocyclic Chem., 2002, 42, 1–14.
4. H. M. C. Marques, Flavour Fragrance J., 2010, 25, 313–326.
5. K. A. Connors, The Stability of Cyclodextrin Complexes in Solution, Chem. Rev., 1997, 97, 1325–1357.
6. G. Astray, C. Gonzalez-Barreiro, J. C. Mejuto, R. Rial-Otero and J. Simal-Gandara, Food Hydrocolloids, 2009, 23, 1631–1640.
7. A. R. Hedges, Chem. Rev., 1998, 98, 2035–2044.
8. J. Szejtli, J. Mater. Chem., 1997, 7, 575–587.
9. V. V. Gorbatchuk, M. A. Ziganshin and B. N. Solomonov, Biophys. Chem., 1999, 81, 107–123.
10. N. A. Mironov, V. V. Breus, V. V. Gorbatchuk, B. N. Solomonov and T. Haertle, J. Agric. Food Chem., 2003, 51, 2665–2673.
11. V. V. Gorbatchuk, N. A. Mironov, B. N. Solomonov and W. D. Habicher, Biomacromolecules, 2004, 5, 1615–1623.
12. M. F. Perutz, A. J. Wilkinson, M. Paoli and G. G. Dodson, Annu. Rev. Biophys. Biomol. Struct., 1998, 27, 1–34.
13. H. Yoshii, T. Furuta, A. Yasunishi and H. Hirano, J. Biochem., 1994, 115, 1035–1037.
14. H. Yoshii, T. Kometani, T. Furuta, Y. Watanabe, Y.-Y. Linko and P. Linko, Biosci., Biotechnol., Biochem., 1998, 62, 2166–2170.
15. T. Furuta, H. Yoshii, A. Miyamoto, A. Yasunishi and H. Hirano, Supramol. Chem., 1993, 1, 321–325.
16. V. V. Gorbatchuk, A. K. Gatiatulin, M. A. Ziganshin, A. T. Gubaidullin and L. S. Yakimova, J. Phys. Chem. B, 2013, 117, 14544–14556.
17. F. Yang and A. J. Russell, Biotechnol. Bioeng., 1996, 49, 709–716.
18. G. Bell, A. E. M. Janssen and P. J. Halling, Enzyme Microb. Technol., 1997, 20, 471–477.
19. H. W. Blanch and D. S. Clark, Applied Biocatalysis, Marcel Dekker, New York, 1991, vol. 1.
20. G. Carrea and S. Riva, Organic Synthesis With Enzymes in Non-Aqueous Media, Wiley-VCH, Weinheim, 2008.
21. S. Belica, M. Sadowska, A. Stępniak, A. Graca and B. Pałecz, J. Chem. Thermodyn., 2014, 69, 112–117.
22. M. Bastos, S. Milheiras and G. Bai, Thermochim. Acta, 2004, 420, 111–117.
23. M. Bilal, C. de Brauer, P. Claudy, P. Germain and J. M. Létoffé, Thermochim. Acta, 1995, 249(1–2), 63–73.
24. Y. Nakai, Y. Keiji, T. Katsuhide, K. Atsushi and S. Isao, Chem. Pharm. Bull., 1986, 34(5), 2178–2182.
25. C. Dos Santos, M. P. Buera and M. F. Mazzobre, Food Chem., 2012, 132(4), 2030–2036.
26. G. S. Pande and R. F. Shangraw, Int. J. Pharm., 1995, 124, 231–239.
27. S. Tanada, T. Nakamura, N. Kawasaki, S. Kitayama and Y. Takebe, J. Colloid Interface Sci., 1997, 186, 180–184.
28. J. J. Hill, E. Y. Shalaev and G. Zografi, J. Pharm. Sci., 2005, 94, 1636–1667.
29. V. V. Gorbatchuk, A. G. Tsifarkin, I. S. Antipin, B. N. Solomonov and A. I. Konovalov, Mendeleev Commun., 1999, 9, 11–13.
30. V. V. Gorbatchuk, A. G. Tsifarkin, I. S. Antipin, B. N. Solomonov, A. I. Konovalov, P. Lhotak and I. Stibor, J. Phys. Chem. B, 2002, 106, 5845–5851.
31. M. V. Rekharsky and Y. Inoue, Chem. Rev., 1998, 98, 1875–1918.
32. N. Taulier and T. V. Chalikian, J. Phys. Chem. B, 2006, 110(25), 12222–12224.
33. W. L. F. Armarego and C. L. L. Chai, Purification of laboratory chemicals, Butterworth-Heinemann, Burlington, MA, 6th edn, 2009.
34. L. Ninni, M. Camargo and A. J. Meirelles, Thermochim. Acta, 1999, 328, 169–176.
35. V. V. Gorbatchuk, A. G. Tsifarkin, I. S. Antipin, B. N. Solomonov, A. I. Konovalov, J. Seidel and F. Baitalov, J. Chem. Soc., Perkin Trans. 2, 2000, 2287–2294.
36. L. S. Yakimova, M. A. Ziganshin, V. A. Sidorov, V. V. Kovalev, E. A. Shokova, V. A. Tafeenko and V. V. Gorbatchuk, J. Phys. Chem. B, 2008, 112, 15569–15575.
37. V. V. Gorbatchuk, M. A. Ziganshin, L. S. Savelyeva, N. A. Mironov and W. D. Habicher, Macromol. Symp., 2004, 210, 263–270.
38. T. Steiner and G. Koellner, J. Am. Chem. Soc., 1994, 116, 5122–5128.
39. C. Betzel, W. Saenger, B. E. Hingerty and G. M. Brown, J. Am. Chem. Soc., 1984, 106, 7545–7557.
40. M. A. Ziganshin, A. V. Yakimov, G. D. Safina, S. E. Solovieva, I. S. Antipin and V. V. Gorbatchuk, Org. Biomol. Chem., 2007, 5, 1472–1478.
41. M. Abundo, L. Accardi, A. Finazzi Agrò, G. Mei and N. Rosato, Biophys. Chem., 1996, 58(3), 313–323.
42. J. A. Ripmeester, Supramol. Chem., 1993, 2(2–3), 89–91.
43. R. B. Gregory, Protein Hydration and Glass Transition Behavior, in Protein–Solvent Interactions, ed. R. B. Gregory, Marcel Dekker, New York, 1995, pp. 191–264.
44. C. De Rango, et al., J. Am. Chem. Soc., 1992, 114, 5475–5476.
45. T. Kida, S. Sato, H. Yoshida, A. Teragaki and M. Akashi, Chem. Commun., 2014, 50, 14245–14248.
46. Z. Dang, X. Song, Q. Guo, Y. Du and J. Yang, Curr. Org. Chem., 2011, 15, 848–861.
47. M. D. Veiga and M. Merino, J. Pharm. Biomed. Anal., 2002, 28(5), 973–982.
48. Y.-M. Zhang, Z.-X. Yang, Y. Chen, F. Ding and Y. Liu, Cryst. Growth Des., 2012, 12, 1370–1377.
49. V. V. Gorbatchuk, A. G Tsifarkin, I. S. Antipin, B. N. Solomonov and A. I. Konovalov, Mendeleev Commun., 1997, 7, 215–217.
50. Y. Furusho and T. Aida, Chem. Commun., 1997, 22, 2205–2206.
51. V. V. Gorbatchuk, M. A. Ziganshin, B. N. Solomonov and M. D. Borisover, J. Phys. Org. Chem., 1997, 10, 901–907.
52. J. Sangster, J. Phys. Chem. Ref. Data, 1989, 18(3), 1111–1227.
53. A. K. Gatiatulin, M. A. Ziganshin and V. V. Gorbatchuk, J. Therm. Anal. Calorim., 2014, 118(2), 987–992.

---

Footnote

† Electronic supplementary information (ESI) available: Equation used for calculation of inclusion Gibbs energy for clathrates formed in two steps; description of fitting procedure for sorption isotherms in binary systems; TG/MS curves and data for bCD hydrates prepared by at various humidities P/P₀, for bCD clathrates prepared by solid-phase exchange of ethanol, acetonitrile and THF, for bCD binary clathrates with various organic guests, for bCD clathrates prepared by simultaneous sorption of water-mimic guest and benzene; pictures of initial dried bCD and its samples equilibrated with saturated vapors of ethanol, HFIP and pyridine; complete X-ray powder diffractograms. See DOI: 10.1039/c6ra11378h

---