A chromium-based metal organic framework as a potential high performance adsorbent for anaesthetic vapours

Abstract

In this work, a chromium-based metal organic framework (Cr-MOF) was synthesized, characterized and tested for the adsorption of a model highly ozone-depleting anaesthetic (sevoflurane). Adsorption isotherms were measured at different temperatures e.g., 283, 298, 313 and 328 K on both Cr-MOF and a conventionally used reference adsorbent. At the temperatures used in this study, the Cr-based MOF showed a significantly higher sevoflurane (selected anaesthetic) equilibrium adsorption capacity compared to the reference sample, although adsorption on the selected MOF did not take place on all active sites (i.e., it did not expose its coordinatively unsaturated sites). Moreover, sevoflurane adsorption on Cr-MOF was found to be fully reversible in the 283–328 K temperature range, and the adsorbent was fully regenerated by vacuum treatment at ambient temperature. The semiempirical Sips model was successfully used to fit sevoflurane adsorption data, substantially confirming the phenomenological aspects of the process inferable from the experimental results.

---

1. Introduction

Since the work published by Fox et al. in 1975,¹ more public attention has been directed towards addressing the potential damage that the release of halogenated general anaesthetic gases constitutes for the global environment.^2–8 All the species that are currently employed as volatile anaesthetics (namely, desflurane, enflurane, halothane, isoflurane and sevoflurane) are halogenated organic compounds potentially noxious to the ozone layer. As modern anaesthesia becomes more and more available to the world population, the global usage of volatile anaesthetics is quickly growing. Anaesthetic vapours are also extensively employed in dentistry, veterinary medicine and research activities involving animals for in vivo experiments. What distinguishes these anaesthetic vapours from other medical drugs is that very small amounts are metabolized with the excess being emitted into the outside environment. Whereas the installation of scavenging systems has significantly decreased spillage of general anaesthetics into operating rooms,⁹ they are still freely vented into the environment. Moreover, scarce attention has been paid to the ecotoxicological properties of gaseous general anaesthetics.

From a chemical point of view, halogenated volatile anaesthetics are in the same category of chlorofluorocarbons (CFCs), that constitute the class of the most aggressive ozone depleting agents. Moreover, the global warming potential (GWP) of halogenated anaesthetics is three orders of magnitude higher than that of carbon dioxide.⁸ The employment of volatile anaesthetics could be reduced by 80–90% if closed circuit anaesthesia were routinely adopted in clinical practice. However, despite the well-trained anaesthesiologists' aptitude and the availability of modern anaesthesia systems, rigorous gas monitoring is required in closed-circuit anaesthesia in order to avoid inadequate oxygenation or volatile anaesthetic concentration.¹⁰

Among the available technologies for the treatment of emitted volatile anaesthetics, those based on adsorption processes seems to be the most promising for achieving a significant reduction of their emissions. Adsorption processes are usually carried out employing pressure swing adsorption (PSA) or vacuum swing adsorption (VSA) techniques. The typical adsorption materials are microporous adsorbents, that are characterized by pores of sizes below 2 nm. Among these, both traditional adsorbing materials (such as silica-gel, activated alumina, activated carbon) and alumino-silicates of the class of zeolites have been considered. In particular, the well-known tunability of the physical and chemical properties of zeolites, including their adsorptive properties, often allows to achieve the selectivity required for a specific separation application.¹¹ As an example, a technique for capturing and storing halogenated anaesthetic vapours by means of a filter containing Y zeolite pellets has been proposed by Perhag et al.:¹² such filter exploits the molecular sieving features of the zeolitic adsorbent to capture the volatile anaesthetics from exhaust streams. Isoflurane emissions were found in bench scale tests to be reduced by about 80%.¹² More recently, Doyle et al. proved that an adsorption column packed with 750 g of a purely siliceous zeolite was able to completely remove isoflurane (whose concentration was 1% in exhaled gases) in a vent line for a period of 8 hours.¹³

During the last 15 years, a new class of nanoporous adsorbents called metal organic frameworks (MOFs) have been developed. MOFs are crystalline hybrid porous solids consisting of metal clusters connected by organic linkers to form tridimensional structures. When an appropriate choice of metal groups and/or organic linkers is made, many different chemical structures can be obtained, sometimes with very large pores and surface areas, thereby increasing the number of accessible adsorption sites and improving the diffusion of a wide variety of substrates to be adsorbed.^14–16 Despite the plethora of papers about the use of MOFs for the capture/storage of a number of industrially and/or environmentally relevant gaseous species,^11,17–19 at the best of our knowledge, no one of them deals with the adsorption of volatile halogenated anaesthetics. For this reason, the present work presents a preliminary study on using MOFs for mitigating the emissions of volatile halogenated anaesthetics into the atmosphere. The MOF adsorbent selected for this study was a Cr-based MOF firstly synthesized by Férey et al.²⁰ in 2005: its zeotype cubic structure has a “giant” cell volume (702 nm³) and larger mesopores (3.0 to 3.4 nm). Furthermore, this Cr-based MOF is known as being stable at air and not destroyed or altered when treated with various organic solvents at ambient temperature or in solvothermal conditions.²⁰ These properties were already successfully used for performing investigations on the adsorption of various gases.^16,21,22 In this work, samples of the aforementioned Cr-based MOF were synthesized, characterized and tested for the adsorption of sevoflurane (Scheme 1).

Scheme 1 Sevoflurane molecule (grey: carbon atoms; white: hydrogen atoms; green: fluorine atoms; red: oxygen atom).

Adsorption isotherms were collected at 298 K for both Cr-MOF and reference adsorbent,¹³ in order to compare the performances of these materials. Moreover, sevoflurane adsorption isotherms on the Cr-based MOF were also collected at 283, 313 and 328 K. These data, together with those obtained at ambient temperature, were analysed using the Sips isotherm model to determine the isosteric heat of adsorption as well as other important parameters, such as sevoflurane affinity and the heterogeneity of the adsorbent.

2. Results and discussion

The X-ray diffraction (XRD) pattern of the synthesized Cr-based MOF (Cr-MOF) shown in Fig. 1 is consistent with other previously reported patterns.^16,20

Fig. 1 X-ray diffraction (XRD) pattern of Cr-MOF.

Fig. 2 shows field emission scanning electron microscopy (FE-SEM) micrographs of the same adsorbent. Sub-micron-sized crystallites are clearly visible.

Fig. 2 Field emission scanning electron microscopy (FE-SEM) images of Cr-MOF crystals obtained using a conventional secondary electron (SE2) detector (a) and an in-lens detector (b).

Inspection of Fig. 2 reveals well-defined octahedral crystals, that are consistent with other SEM investigations performed on the same material.^16,23

The N₂ adsorption isotherm at 77 K (Fig. 3) also conforms to the results reported in the literature.¹⁶ Indeed, the total pore volume, as estimated from N₂ adsorbed amounts at p/p₀ = 0.99, was found to be about 1.5 cm³ g⁻¹, while the specific surface area, as estimated by applying the BET method, was about 3000 m² g⁻¹. In summary, the combined characterization studies using XRD, SEM, and N₂ adsorption analyses clearly demonstrate the successful synthesis of the Cr-based MOF.

Fig. 3 N₂ adsorption isotherm on Cr-MOF measured by a Micromeritics ASAP 2020 apparatus at 77 K.

The sevoflurane adsorption isotherms at 298 K (Fig. 4) clearly reveal that the MOF adsorbent performs much better than the reference material in terms of adsorption capacity for the whole pressure range considered in this study. In particular, if the working conditions reported in literature are considered (i.e., the anaesthetic bulk pressure is approximately 1 kPa),¹³ the sevoflurane amount adsorbed by the reference at the equilibrium is about 2 mol kg⁻¹. This is quite comparable with the isoflurane captured amount that can be extrapolated for the same substrate when saturation under dynamic adsorption conditions is reached.¹³ Considering the actual working conditions of a scavenging line instead (i.e., an anaesthetic partial pressure of about 2–3 kPa), it can be clearly noted from Fig. 4 that the sevoflurane amount adsorbed on Cr-MOF is over 7.5 mol kg⁻¹. This is about three times higher than the corresponding adsorption capacity of the reference adsorbent. Keeping in mind that MOF samples used in this work are in powder form, and even considering a post-shaping decrease of the adsorption capacity by an estimated 30%,²⁴ the performance of Cr-MOF will still be superior than the reference adsorbent. These results can be regarded as even more favourable for the Cr-based MOF considering that the reference adsorbent was activated at a relatively high temperature, e.g., 393 K compared to ambient temperature used for Cr-MOF activation. It was earlier reported¹⁶ that the synthesized Cr-based MOF can be activated under vacuum at 423 K, achieving sevoflurane adsorption performances that were also better than those reported in Fig. 4 (data not shown). However, after a second adsorption–desorption cycle, such process was found to be irreversible, i.e., it was not possible to regenerate the adsorbent. For this reason, two cycles with regeneration under vacuum and 298 K have been performed on a fresh Cr-MOF sample. It can be clearly noted from Fig. 5 that the cubic B-splines interpolations of the two data sets are almost superimposed, thus showing that Cr-MOF maintains its sevoflurane adsorption capabilities (i.e., it shows a complete regeneration ability) when re-activated under vacuum at ambient temperature.

Fig. 4 Sevoflurane adsorption isotherms at 298 K on Cr-MOF (squares) and reference adsorbent (circles). Continuous lines: interpolating cubic B-splines.

Fig. 5 Sevoflurane adsorption isotherms at 298 K on Cr-MOF. Squares: first adsorption run. Circles: second adsorption run after re-activation under vacuum at 298 K. Continuous lines: interpolating cubic B-splines.

This favourable Cr-MOF behaviour exhibited at adsorption of sevoflurane is believed to be caused by the presence of coordinatively unsaturated sites (CUSs) on the pore surface of this MOF. Indeed, it is known that, unlike zeolites and inorganic mesoporous molecular sieves, some MOFs can develop CUSs in the pore channels during the synthesis by the 3D covalent connection of inorganic and organic parts in their structure (Scheme 2).²³

Scheme 2 Structures of the purely siliceous, zeolite-type reference adsorbent (a) and of Cr-MOF (b) (yellow: silicon atoms; red: oxygen atoms; grey: carbon atoms; magenta: chromium atoms).

Furthermore, trimeric chromium(III) octahedral clusters are coordinated with terminal water molecules of Cr-MOF in the 298–473 K temperature range,²⁰ thus supplying CUSs that can perform as active sites for surface functionalization²³ or selective adsorption.¹⁶ Presumably, when sevoflurane molecules adsorb on Cr-MOF surface, without the CUSs being exposed, the weak nature of this interaction is considered adequate to ensure both a significant adsorption capacity and a full regeneration of the adsorbent. On the contrary, when CUSs are exposed, their specific interaction with sevoflurane molecules looks similar to that of water vapour with the CUSs of Cu-BTC MOF, that leads to an irreversible damage of the porous substrate in terms of the adsorption capabilities.^25–27

Using the same experimental procedure that produced the data reported in Fig. 5, sevoflurane adsorption isotherms on Cr-MOF at 283, 313 and 328 K were also collected, as shown in Fig. 6.

Fig. 6 Sevoflurane adsorption isotherms on Cr-MOF at T = 283 K (circles), 298 K (squares), 313 K (triangles) and 348 K (diamonds). Continuous lines: best fitting Sips theoretical isotherms.

Moreover, to obtain more insight in the adsorption phenomena examined, the whole set of equilibrium data reported in Fig. 6 was analysed by using the semiempirical three parameter Sips isotherm model.²⁸ According to this model, the pressure dependence of the adsorbed amount takes the following form:

where q_max, b and n are model parameters, where q_max represents the maximum adsorption capacity, b is the affinity constant and n is the heterogeneity coefficient (for n = 1 the Sips isotherm reduces to the Langmuir isotherm, which applies to homogeneous adsorbent–adsorbate systems). Sips parameters are generally temperature dependent, as reported by Do;²⁹ however, q_max and n are assumed to be independent of temperature. Based on that, description of sevoflurane adsorption on Cr-MOF was performed by coupling eqn (1) with the following expression to relate the dependence of the affinity coefficient b on temperature:²⁹

In eqn (2), b₀ is the value of b at a reference temperature T₀ and Q is the value of the isosteric heat of adsorption when the adsorbent fractional coverage is equal to 0.5. The experimental data of sevoflurane adsorption on Cr-MOF were analysed by a non-linear regression method using the MATLAB Surface Fitting Toolbox to simultaneously calculate the optimal values of the parameters that appear in eqn (1) and (2), i.e., q_max, b₀, Q and n. The calculated values of the parameters, obtained using T₀ = 283 K as the reference temperature, are reported in Table 1, and the comparisons between model and experimental results are reported in Fig. 6, in which the symbols refer to experimental data and the solid curves refer to the best fitting of the Sips theoretical isotherms.

Table 1 Sips parameters for sevoflurane adsorption on Cr-MOF

Parameter	Best fitting value ± standard deviation
qmax (mol kg^−1)	8.99 ± 0.15
b0 (kPa^−1)	7.79 ± 0.78
Q (kJ mol^−1)	52.78 ± 1.74
n	1.29 ± 0.06
Regression coefficient R^2 = 0.984

---

Inspection of Fig. 6 clearly indicates a good correlation between model curves and experimental results, which is confirmed by the regression coefficient R² value as reported in Table 1.

The results of the modelling process contain further insights on the way sevoflurane molecules interact with Cr-MOF. Indeed, applying the “Gurvitch rule”,³⁰ it was possible to re-calculate the total pore volume of the metal organic substrate multiplying q_max by the liquid molar density of sevoflurane, which gave a value of about 1.2 cm³ g⁻¹ as compared to the value of 1.5 cm³ g⁻¹ obtained from microporosimetric analysis. This discrepancy most probably arises from the different degassing conditions between sevoflurane and nitrogen adsorption tests (298 K for 1 h versus 373 K overnight), at least partially accounting for the pore space occupied by coordinated water molecules in samples prepared for anaesthetic capture.

Regarding the affinity coefficient b for sevoflurane adsorption on Cr-MOF at ambient temperature, its value has been found to be 2.66 kPa⁻¹, that is quite similar to the value determined for the adsorption of other condensable vapours, such as H₂O¹⁵ on MOFs, and is significantly higher than the corresponding value for H₂S adsorption on the same Cr-based MOF substrate.¹⁶ Besides, the good affinity of Cr-MOF towards sevoflurane is evident from the experimental data shown in Fig. 4–6, where the adsorbent, at ambient temperature, clearly exploits more than 2/3 of its total adsorption capacity for pressures as low as 1 to 3 kPa, corresponding to the partial pressures of halogenated anaesthetics in typical anaesthetic applications.¹³

Further analysis involved the heterogeneity parameter n, which can be considered as an important indication of the adsorbent–adsorbate interaction. In the case of sevoflurane adsorption on Cr-MOF, the value of n is lower than obtained for adsorption processes involving the exposure of CUSs on the surface of the same substrate,¹⁶ and predictably is even lower than values reported in literature for chemisorption-type phenomena.²⁵ These observations can be used as an indirect indication that ambient temperature-activated Cr-MOF adsorbs sevoflurane molecules just by means of weak interactions, thus resulting in easy regeneration of the adsorbent as is clearly shown in Fig. 5 and 6.

Finally, the value of the isosteric heat of sevoflurane adsorption on Cr-MOF (52.78 kJ mol⁻¹, as reported in Table 1) can be considered independent of the adsorbent fractional coverage because the expression for isosteric heat of adsorption derived from the Sips model reduces to the constant Q when the hypothesis of the temperature independence of the heterogeneity parameter n is taken into account.²⁹ When compared to that of similar substrates reported in literature,^14–16 the calculated value of Q is relatively high. It is known that, on plant scale, fixed-bed adsorption is an essentially adiabatic operation, so the isosteric heat of adsorption is responsible for the temperature rise during the process and, since adsorption is an exothermic process, an increase in temperature leads to a decrease in adsorption capacity. Based on this, a high value of the isosteric heat for sevoflurane adsorption on Cr-MOF could adversely affect the performance of this adsorbent in the real operation. Moreover, it is known that other species, such as water vapour, can compete with the target compounds for the adsorption sites, possibly forcing the implementation of an additional guard column that would act as a moisture damper under practical use. With the aim of better clarifying this issues, further studies involving packed (and possibly hydrophobized)²⁷ Cr-MOF pellets will be conducted.

3. Conclusions

The results reported in this work, though preliminary, provide a useful insight into the potential use of a chromium-based metal organic framework as a higher performance replacement of adsorbent materials that are currently used for the capture of halogenated anaesthetics. From the thermodynamic point of view, the chosen Cr-based MOF provides, at ambient temperature, a significantly higher sevoflurane adsorption capacity compared to a reference substrate, although the selected MOF does not appear to exploit all its adsorption active sites (i.e., it does not expose its coordinatively unsaturated sites) in the process. Moreover, sevoflurane adsorption on Cr-MOF was found to be reversible in the 283–328 K temperature range for sevoflurane pressures up to about 15 kPa, thus the adsorbent was easily regenerated by vacuum treatment at ambient temperature. Finally, the semiempirical Sips model was successfully implemented to better describe the obtained sevoflurane adsorption data, confirming the phenomenological aspects of the process from the analysis of experimental data. The obtained results warrant further investigations, i.e., dynamic adsorption tests using columns packed with shaped MOF samples.

4. Experimental

Cr-MOF samples were synthesized as powders using a procedure described by Férey et al.²⁰ 4.00 g of Cr(NO₃)₃·9H₂O (Baker) was dissolved in 43.20 g of ultra-purified water which was produced by a TKA Smart2Pure device; 1.64 g of terephthalic acid (Aldrich) and 5.0 ml of a 4 wt% hydrofluoric acid solution, obtained by diluting a 37 wt% pristine solution (Carlo Erba), were then added. These chemicals were mixed and kept in a stirrer for about 5 min. The resulting suspension was then transferred to a Teflon-lined autoclave bomb and kept in an oven at 493 K for 8 h. After equilibration at ambient temperature, the large terephthalic acid crystals present in the batch were eliminated by filtration using a large pore fritted glass filter (no. 2); the water suspension of Cr-MOF powders which passed through the filter was then filtered again on Whatman ashless grade 42 filtration paper. The retained green product was finally dried at 333 K overnight.

XRD patterns of finely grinded Cr-MOF samples were collected using a Philips X'Pert PRO apparatus with CuKα radiation: the scanning range was 2–15° in 2θ, the scanning step size was 0.013°, and the scan speed was 0.072° s⁻¹. FE-SEM micrographs were collected with a Zeiss Ultra Plus instrument using both a conventional secondary electron detector (SE2, operating voltage: 10 kV) and an in-lens detector (operating voltage: 8 kV). Textural characterization was carried out by N₂ adsorption at 77 K: the specific surface area was evaluated by means of the Brunauer–Emmett–Teller (BET) method, while the total pore volume was estimated from the N₂ adsorbed amount at p/p₀ = 0.99; a Micromeritics ASAP 2020 volumetric instrument was used for this purpose. The as-synthesized samples were degassed at 373 K overnight prior to characterization.

The sevoflurane adsorption isotherms on Cr-MOF samples at four different temperatures (283, 298, 313 and 328 K), together with adsorption data for the same adsorbate on the reference sample at ambient temperature were obtained using a gravimetric technique based on a McBain-type balance.¹⁶ The device was equipped with a quartz spring (Ruska Instrument Co., Houston, Texas), with a sensitivity of 5 mm mg⁻¹, and a small quartz pan, containing between 10 and 15 mg of the adsorbent sample, hooked to the spring. The amount of adsorbate was evaluated by measuring the spring elongation with the help of a cathetometer, which enabled the reading of the spring deflection down to 0.05 mm. Gas pressure in the adsorption chamber was electronically measured by means of a capacitive pressure transducer (Edwards Datametrics 1500). A Heto thermostating unit allowed temperature control of the gas in the adsorption chamber within a range of ±0.1 K. Before measurement, Cr-MOF samples were degassed in situ at 298 K under high vacuum (p < 10⁻³ Pa) by means of an Edwards turbomolecular pump for 1 h. As explained in detail in the “Results and discussion” section, contrary to what reported even for thermally unstable adsorbents,^31–33 the activation of the Cr-based MOF for sevoflurane adsorption was chosen to be carried out at the ambient temperature. On the other hand, the reference adsorbent was degassed at 393 K using the same high vacuum line employed for activating Cr-MOF.

Notes and references

1. J. W. Fox, E. J. Fox and R. Villanueva, Lancet, 1975, 305, 864.
2. A. C. Brown, C. E. Canosa-Mas, A. D. Parr, J. M. Pierce and R. P. Wayne, Nature, 1989, 341, 635–637.
3. M. Logan and J. G. Farmer, Br. J. Anaesth., 1989, 63, 645–647.
4. P. Hutton, J. A. Kerr, J. M. T. Pierce and S. P. K. Linter, Lancet, 1989, 333, 1011–1012.
5. K. J. Hopkins, R. Albanese, B. D. Joyner, R. C. Rodgers and J. A. S. Ross, Lancet, 1989, 333, 1209–1210.
6. T. Langbein, H. Sonntag, D. Trapp, A. Hoffmann, W. Malms, E. P. Roth, V. Mors and R. Zellner, Br. J. Anaesth., 1999, 82, 66–73.
7. R. Byrick and D. J. Doyle, Br. J. Anaesth., 2001, 86, 595–596.
8. Y. Ishizawa, Anesth. Analg., 2011, 112, 213–217.
9. R. J. Gardner, Ann. Occup. Hyg., 1989, 33, 159–173.
10. P. Schober and S. A. Loer, Eur. J. Anaesthesiol., 2006, 23, 914–920.
11. N. Gargiulo, F. Pepe and D. Caputo, J. Nanosci. Nanotechnol., 2014, 14, 1811–1822.
12. L. Perhag, P. Reinstrup, R. Thomasson and O. Werner, Br. J. Anaesth., 2000, 85, 482–486.
13. D. J. Doyle, R. Byrick, D. Filipovic and F. Cashin, Can. J. Anaesth., 2002, 49, 799–804.
14. P. Aprea, D. Caputo, N. Gargiulo, F. Iucolano and F. Pepe, J. Chem. Eng. Data, 2010, 55, 3655–3661.
15. N. Gargiulo, M. Imperatore, P. Aprea and D. Caputo, Microporous Mesoporous Mater., 2011, 145, 74–79.
16. A. Peluso, N. Gargiulo, P. Aprea, F. Pepe and D. Caputo, Sci. Adv. Mater., 2014, 6, 164–170.
17. J.-R. Li, Y. Ma, M. C. McCarthy, J. Sculley, J. Yu, H.-K. Jeong, P. B. Balbuena and H.-C. Zhou, Coord. Chem. Rev., 2011, 255, 1791–1823.
18. Y. Sun, L. Wang, W. A. Amer, H. Yu, J. Ji, L. Huang, J. Shan and R. Tong, J. Inorg. Organomet. Polym., 2013, 23, 270–285.
19. K. Konstas, T. Osl, Y. Yang, M. Batten, N. Burke, A. J. Hill and M. R. Hill, J. Mater. Chem., 2012, 22, 16698–16708.
20. G. Férey, C. Mellot-Draznieks, C. Serre, F. Millange, J. Dutour, S. Surblé and I. Margiolaki, Science, 2005, 309, 2040–2042.
21. P. L. Llewellyn, S. Bourrelly, C. Serre, A. Vimont, M. Daturi, L. Hamon, G. De Weireld, J.-S. Chang, D.-Y. Hong, Y. K. Hwang, S. H. Jhung and G. Férey, Langmuir, 2008, 24, 7245–7250.
22. M. Latroche, S. Surblé, C. Serre, C. Mellot-Draznieks, P. L. Llewellyn, J.-H. Lee, J.-S. Chang, H. J. Sung and G. Férey, Angew. Chem., Int. Ed., 2006, 45, 8227–8231.
23. D.-Y. Hong, Y. K. Hwang, C. Serre, G. Férey and J.-S. Chang, Adv. Funct. Mater., 2009, 19, 1537–1552.
24. V. Finsy, L. Ma, L. Alaerts, D. E. De Vos, G. V. Baron and J. F. M. Denayer, Microporous Mesoporous Mater., 2009, 120, 221–227.
25. L. Hamon, E. Jolimaître and G. D. Pirngruber, Ind. Eng. Chem. Res., 2010, 49, 7497–7503.
26. S. K. Henninger, G. Munz, K.-F. Ratzsch and P. Schossig, Renewable Energy, 2011, 36, 3043–3049.
27. J. B. Decoste, G. W. Peterson, M. W. Smith, C. A. Stone and C. R. Willis, J. Am. Chem. Soc., 2012, 134, 1486–1489.
28. R. Sips, J. Chem. Phys., 1948, 16, 490–495.
29. D. D. Do, Adsorption Analysis: Equilibria and Kinetics, Imperial College Press, London, 1998.
30. S. J. Gregg and K. S. W. Sing, Adsorption, Surface Area and Porosity, Academic Press, London, 1982.
31. N. Gargiulo, D. Caputo and C. Colella, Stud. Surf. Sci. Catal., 2007, 170, 1938–1943.
32. N. Gargiulo, F. Pepe and D. Caputo, J. Colloid Interface Sci., 2012, 367, 348–354.
33. N. Gargiulo, A. Peluso, P. Aprea, F. Pepe and D. Caputo, J. Chem. Eng. Data, 2014, 59, 896–902.

---