Volume accessibility of acid sites in modified montmorillonite and triacetin selectivity in acetylation of glycerol

Abstract

Organic acid treatment enhances the acidity and surface characteristics of montmorillonite clay by dealumination. Dealuminated and Al-clays were used for acetylation of glycerol with acetic acid. All clays used had comparable acidity but pore characteristics were different. Though glycerol conversions were similar but triacetin selectivity was different. Al-clay and acid treated clays showed poor and improved selectivity, respectively. This was attributed to increased pore volume around the acid sites, which facilitates the multiple acetylation of glycerol. The generated space around acid centers, termed as ‘volume accessibility’, helps glycerol to interact with acylium ions formed on the acid sites more effectively leading to formation of triacetin. Correlations were made between the changed characteristics of clays and triacetin yield. Among the different correlations, triacetin selectivity correlates well with the volume accessibility. The latter is quite useful in predicting the catalytic performance.

---

1. Introduction

Commercial production of biofuels has escalated in recent years resulting in a large buildup of glycerol, a byproduct of biofuel production, in the market.^1–3 Conversion of surplus glycerol into valuable and useful products has been a matter of interest for researchers in the field. Lately, many reports have appeared on glycerol conversion into value added products such as acetins,⁴ acetals,⁵ carbonates,⁶ ethers,⁷ acrolein,⁸ hydroxy acetones,¹¹ and propane diols^9,10 using various solid acids as catalysts.

Because of the many potential applications of acetylated glycerol derivatives, acetylation of glycerol has been studied by many researchers.⁴ Acetylation of glycerol with acetic acid produces three products mono, di and tri acetylated esters.¹² These have applications in cryogenics and are raw materials for the production of biodegradable polyester and cosmetics. They also find applications as fuel additives due to its inherent properties such as enhanced anti-knocking, anti-freezing and viscosity.^13–15

Traditionally, acetylation reactions in industries are performed in batch reactors with homogeneous liquid phase acid catalysts such as sulfuric acid, hydrochloric acid and orthophosphoric acid.^16–19 There are also reports on to drive the process towards green reaction in the absence of solvents. A few research groups have worked on the acetylation of glycerol using solid acid catalysts such as, SiO₂–Al₂O₃ (0.3%),⁴ SBA-15 (11%),¹ SnCl₂ (<1%),²⁰ zeolite (7%),^12,21 mesoporous silica (5%),²² Amberlyst-15 (12%),²³ ion-exchange resins (4%),^24,25 mont K-10 (5%), KSF (2%),^26,27 niobic acid (1%),⁴ supported heteropoly acids (1%)^28,29 and zirconia (4%).^30,31 All these catalysts show higher glycerol conversions but report a low selectivity of the triacetin ester which is a more preferred additive. The focus is now on achieving enhanced selectivity of the triacetin.^1,4,12,32

We have previously modified the Indian montmorillonite clay from Bhuj area, Gujarat. In order to verify the effect of dealumination on organic reactions, clay from a very different source is used in the present work. The clay sample used in the present study is an American montmorillonite GK 129 for glycerol acetylation after dealumination. Previous studies on ester to ketone rearrangement is found to increase only in the presence of enhanced micropore volume generated by dealumination and in the oxazole synthesis, in addition to enhanced micropore volume, an optimum B/L ratio is found to be necessary.^33–34 Deactivation aspect comparison between zeolites and modified clays for the alkylation of p-cresol^35–37 was also studied. In the present paper, however, the best condition to achieve multiple acetylation of glycerol is observed to be related to generated micropore volume and total acidity.

2. Experimental

2.1 Materials method

American montmorillonite GK-129 provided by Ceramic Technological Institute, Bangalore, is used in the present study. Its composition found by ICP-OES analysis to be 70.16% SiO₂, 12.24% Al₂O₃, 1.76% Fe₂O₃, 4.25% MgO, 3.76% Na₂O, 0.62% K₂O and 1.46% CaO with idealized structural formula Si₆[Al_1.19Fe_0.227Mg_0.474]O₁₀(OH)₂Na_(0.426). Methanesulfonic acid and p-toluenesulfonic acid were procured from SD fine chemicals, phenoldisulfonic acid from Loba chemicals, glycerol from Merck chemicals and acetic acid and n-pentanol from CDH chemicals.

2.2 Catalyst preparation

Clay samples were prepared by following reported procedure.³³ 10 g of montmorillonite clay was dispersed in 100 ml of 1 M solution of methanesulfonic acid; the mixture was refluxed under microwave irradiation for 30 minutes by applying variable power of 200–800 W to maintain 110 °C temperatures. After completion of irradiation, the mixture was cooled to ambient temperature and the clay residue was repeatedly washed by centrifugation to remove the excess of acid from the clay. The washed solid was finally dried at 120 °C and finely ground and sieved through a sieve of mesh size 75 μm.³⁴ The treated clays were designated as 1MSA-clay, 1p-TSA-clay and 1PDSA-clay for 1 M of each MSA, p-TSA and PDSA treatment respectively.

2.3 Characterization

Different acid treated clay samples were characterized using various techniques such as, ²⁷Al-NMR, FT-IR, pyridine FT-IR, powder XRD, CEC and BET surface area.

Acidity of acid treated montmorillonite clay samples were recorded using KBr pellet technique of samples in Shimadzu, IR Affinity-1 instrument with resolution of 4 cm⁻¹, of 40 scans in the wavenumber range of 1600–1400 cm⁻¹. Before recording the spectra, the samples were activated by degassing at 110 °C for 2 hours, then cooled under vacuum and saturated with liquid pyridine. Samples were again heated to 120 °C to remove physisorbed pyridine.³⁵ The amount of pyridine chemisorbed on the acid sites of the clay samples was determined gravimetrically which was taken as a measure of the total acidity. The Brønsted and Lewis acidity was quantified based on the peak areas of the two sites with respect to the total acidity. The acidity was expressed in micromole per gram of clay. The molar extinction coefficient of pyridine adsorbed on the acid sites was found to be 5.8 × 10⁴ kg mol⁻¹ cm⁻¹.^34,38–39

Surface area and pore characteristics of parent and modified clays were characterized using Quanta Chrome Nova-1000 surface analyzer instrument under liquid nitrogen temperature. Adsorption–desorption isotherm measurements were done in order to study the evolution of porosity and textural properties and surface area from BET method, BJH method used to evaluate pore diameter and volume and deBoer t-method for the newly generated micro pore volume measurement.³³

Structural integrity of the catalyst samples was checked by powder XRD. The data were recorded by step scanning at 2θ = 0.020° per second from 3° to 80° on PANalytical X'Pert PRO MPD X-ray diffractometer with graphite monochromatized Cu Kα radiation (λ = 0.15406 nm).^33,34

²⁷Al MAS NMR spectra of Na and treated montmorillonite clays were recorded using Joel-ECX instrument having 8 mm rotor of speed 10k.³⁶

The total acidity of clays was evaluated using NH₃-TPD method. In a typical experiment, 0.1 g of sample was taken in a U-shaped quartz sample tube, and heated to 150 °C for 2 h under the flow of He (30 cm³ min⁻¹). Then the mixture of NH₃ (90%) and He (10%) was passed with a rate 30 cm³ min⁻¹ for 30 min by maintaining the temperature of 60 °C. After this the catalyst sample was flushed with He (30 cm³ min⁻¹) for 60 min. TPD measurements were carried out in the range 100–500 °C with ramping of 10 °C min⁻¹. Ammonia concentration in the effluent was monitored with thermal conductivity detector.

CEC value were obtained by barium saturation method, a test portion, 2.0 g clay was shaken for 1 h with 30 ml of 0.1 M BaCl₂ solution. The clay and liquid phases are separated by centrifugation. This operation was repeated thrice and clay suspension was collected and again shaken for 2 h with 30 ml of 0.02 M HCl. The adsorbed barium exchanges with H⁺ and clay suspension was centrifuged and liquid portion was collected, heated liquid portion was stirred with concentrated H₂SO₄, Ba²⁺ precipitates as BaSO_4. The residual BaSO₄ was measured by gravimetric and used for the CEC calculations.^40,41

The amount of interlayer aluminium in different acid treated clay samples was estimated by the method reported elsewhere.³⁵ It involved treatment of 2 g of the clay sample with 0.1 M HCl for 1 h at 40 °C. After treatment, the mixture was centrifuged and the solid was repeatedly washed with deionized water. The centrifugate subjected to Al estimation spectrophotometrically by the aluminon reagent method using a Chemito UV-2100 spectrophotometer.^34,36

2.4 Catalytic activity tests

Evaluation of the catalytic activity of modified montmorillonite clay catalysts were studied under microwave irradiation using Milestone, “START-S” microwave lab station for synthesis, Italy. The equipment enables the control of reaction mixture temperature with aid of infrared sensor through regulation of microwave power output in such a way that the reaction mixture was exactly in line with infrared sensor that monitors the temperature. Variable power up to 1200 W was applied by microprocessor-controlled single-magnetron system.³⁴

Ten mmol of glycerol and thirty mmol of acetic acid were mixed with 0.5 g catalyst in a 25 ml microwave reactor vessel with a magnetic stirring bar. Reactor vessel was kept in microwave reactor and an initial power of 1100 W was applied for 1 minute to attain the reaction temperature of 120 °C with stirring speed of 500 rpm. The reaction temperature was then maintained for 59 minutes by applying variable power of 300–800 W. The reaction mixture was cooled for 5 minutes, the reactants and products mixture was then extracted by stirring with 4 ml of n-pentanol for 10 minutes and filtered to remove the catalyst. The reaction mixture was analyzed using Chemito GC-1000 gas chromatograph with TR-Wax capillary column with a flame ionization detector using methyl salicylate as internal standard. The products were also confirmed by GC-MS analysis.³⁶

3. Results and discussion

Organic acids are known to remove structural elements present in the clay layers such as Al, Fe and Mg by complexation.³⁵ Table 1 shows the surface characteristics, acidity and CEC of the montmorillonite clay treated with different organic acids along with Al-clay. Na montmorillonite clay has a surface area 35 m² g⁻¹ with very low acidity and having a cation exchange capacity (CEC) of 0.92 meq g⁻¹. Al-exchanged clay demonstrated no change in surface characteristics but the interlayer Al content corresponded with CEC of the parent clay and showed an increase in acidity, consequent to interlayer Al, as measured by pyridine FT-IR. Organic acid treated clays exhibited a clear increase in the surface area and micropore volume due to the removal of structural Al, Fe and Mg ions.^33–35 Slight reduction in CEC is attributed to the removal of small amounts of isomorphously substituted Fe and Mg ions by the organic acids used for treatment.^33–37 A part of the structural Al removed occupied the interlayer corresponding to the CEC of the resulting clay after treatment.

Table 1 CEC, surface area, average pore diameter, pore volume, acidities and amount interlayer aluminum in different acid treated clays^a

Clay samples	CEC (meq g^−1)	Surface area (m^2 g^−1)	Micropore volume (cm^3 g^−1)	Average pore diameter (Å)	Pyridine FT-IR acidity (μmol g^−1)	Interlayer Al (meq g^−1)	Structural Al removed (meq g^−1)	NH3-TPD acidity (μmol g^−1)
a Standard deviation (sd) value for the structural Al removed (sd ± 0.12); interlayer Al (sd ± 0.19); CEC (sd ± 0.21).
Na-clay	0.92	35	0.009	50.7	10	—	—	22
Al-clay	0.92	35	0.009	50.7	150	0.91	—	258
MSA-clay	0.82	204	0.048	34.3	148	0.82	0.876	254
p-TSA-clay	0.86	141	0.030	36.8	144	0.84	0.998	256
PDSA-clay	0.80	276	0.198	28.5	148	0.81	1.152	258

---

Surface area and micropore volume depended on the ability of the acids to remove the structural ions and increased in the order PDSA > MSA > p-TSA. PDSA with two sulfonic acid groups was able to remove more structural elements than the other two mono sulfonic acids. It also dislodged a higher number of structural elements creating more micropore regions with considerable increase in the micropore volume and a proportional decrease in the average pore diameter. MSA clay shows a higher pore volume and surface area than p-TSA clay as it removed more structural elements in comparison with p-TSA clay.^33,34

3.1 Reaction parameter variation

The primary objective of this report was to increase the glycerol conversion with maximum selectivity of triacetin (Scheme 1). Study on the variation of reaction parameters was therefore necessary to achieve the objective. Glycerol acetylation with acetic acid was carried out under varied reaction conditions such as temperature (60–120 °C), time (5–240 min), and catalyst amount (0.1–1.0 g) maintaining the reactant mole ratio 1 : 3 of glycerol to acetic acid. Typical reaction was conducted using 1MSA clay. Each experiment was conducted thrice and the mean value was considered for the calculation of standard deviation.

Scheme 1 Reaction between glycerol and acetic acid.

Effect of reaction time. Effect of reaction time on the acetylation of glycerol with acetic acid was studied in the range of 5–240 min. Fig. 1 shows an increase in conversion of glycerol from 5 to 60 min, and the conversion remained constant (96%) thereafter. In the initial time of 5 minutes, conversion of glycerol resulted mainly in the mono ester formation (monoacetin 18%) and diester formation (diacetins 31%) with a very poor yield of tri ester (triacetin, 2%). Further increase in reaction time leads to the formation of a maximum amount (24%) of monoacetin at a time of 20 min. On further increase in time from 20 to 40 min, an increase in diacetins with concomitant decrease in monoacetin was observed. It is apparently due to acetylation of monoacetin to form the higher esters, mostly diacetin and a very small amount of triacetin. Beyond 40 min, continuous increase in triacetin was observed with decreased mono- and di acetins owing to further acetylation of mono- and diacetins. At the end 240 min, maximum yield of 44% triacetin was observed along with 12% mono- and 33% diacetins.

Fig. 1 Effect of reaction time on glycerol acetylation with acetic acid. Reaction conditions: temperature 120 °C, catalyst (1MSA clay) amount 0.5 g, mole ratio of glycerol to acetic acid, 1 : 3.

Effect of catalyst amount. Catalyst amount shows significant effect on the catalytic acetylation of glycerol with acetic acid. As the catalyst amount increased from 0.1 to 1 g, the conversion of glycerol increased up to 0.5 g, thereafter the conversion stabilized and reached a maximum of 96%. From Fig. 2, it can be seen that at lower amount of catalyst (0.1 g) glycerol conversion is 34%, the conversion distribution was more towards the lower esters mono- and di-acetins (19%) and very small amount of triacetin. Increase in catalyst amount resulted in more of triacetin. Beyond 0.4 g of catalyst, a steep increase in the triacetin with a proportional decrease in the mono and diacetins was observed. A maximum of 58% triacetin was formed when 1 g of the catalyst was used. At higher catalyst amounts, dehydration of acetic acid is accelerated leading to generation of acylium ion which readily attacks the free hydroxyl groups in the glycerol resulting in acetins.^1–5

Fig. 2 Effect of catalyst (1MSA clay) amount on glycerol acetylation with acetic acid. Reaction conditions; temperature 120 °C, time 60 min, mole ratio of glycerol to acetic acid; 1 : 3.

Effect of temperature. Temperature also has significant role in the reaction. Increase in reaction temperature showed increased conversion of glycerol (Fig. 3). At lower temperature (60 °C), 82% conversion was achieved. Further increase in temperature led to an increase in conversion and reached a maximum of 98% at temperature of 120 °C. At lower temperatures up to 60 °C monoacetin was the preferred product (53%) with moderate amount of diacetins (28%) and negligible amount of triacetin. Beyond 80 °C, amount of triacetin increased steeply and reached a maximum of 47% at 120 °C. It could also be observed that increase in triacetin followed a proportionate decrease in the lower esters. It is attributed to the dehydration of acetic acid. At lower temperature dehydration of acetic acid to form acylium ions is less and availability of hydroxyl group of the glycerol for acylation to form lower esters is more probable. Beyond 100 °C, rapid dehydration of acetic acid generates more acylium ions which readily attack hydroxyl group of the mono and diacetins leading to more amount of triacetin.

Fig. 3 Effect of reaction temperature on glycerol acetylation with acetic acid. Reaction conditions; reaction time; 60 min; catalyst (1MSA clay) amount 0.5 g; mole ratio of glycerol to acetic acid; 1 : 3.

It was observed that under different reaction conditions, initial formation of lower esters, mono and diacetins, led to their further acetylation to triacetin. Mass balance was made by considering the number of moles of the total reactants taken and number of moles of total products formed, by the analysis it gives 98.06% of reaction mixture was recovered after the reaction.

3.2 Comparison of different clay catalysts

In order to understand the role of various catalyst parameters to form more triacetin, glycerol acetylation was performed with different acid treated clay samples under optimized reaction conditions. Results obtained with product distribution are shown in Table 2.

Table 2 Comparison of different catalysts for conversion and yields of product^a

Catalyst samples	Glycerol conversion (%)	Monoacetin yield (%)	Diacetins yield (%)	Triacetin yield (%)
a Reaction conditions: temperature; 120 °C, time; 60 min, catalyst amount; 1 g and mole ratio; glycerol : acetic acid is 1 : 3.
Al-clay	60	41	10	9
MSA-clay	94	22	31	41
p-TSA-clay	94	30	35	29
PDSA-clay	96	14	26	56

---

Acetylation is known to be catalyzed in presence of acid catalysts. Al-clay showed significant activity in the glycerol acetylation obviously because of its high acidity. But very low amount of triacetin was formed with Al-clay in comparison with the organic acid treated clay samples the latter, on the other hand, exhibited a significant increase in the selectivity for triacetin. With reference to Table 2 shows that an important feature of acid treated clay is the formation of pores upon acid treatment as shown by their increase in surface characteristics. In contrast, Al-clay has no such pores on the surface as shown by its low surface area and micropore volume. It appears that, apart from the acidity of the catalyst, pores formed on acid treatment play an important role in the formation of triacetin. The acid treated clay catalysts, showed almost similar conversion of glycerol but the selectivity to different acetins was different; PDSA treated clay showing more yield of triacetin compared to others.

The conversion of glycerol and yield of mono, di and triacetins are provided in Table 2. It appears that enhancement of triacetin selectivity is affected by the change in catalyst characteristics after acid treatment. Treatment with organic acids is known to bring about changes in the acidities and surface characteristics (Fig. 4 and 5) of the clay by removing the octahedral cations through the formation of soluble complexes.^33–35 These soluble complexes get hydrolyzed at the interlamellar region and the cationic aquo hydroxyl species of Al replace the interlayer Na exhibiting acid characteristics. Pyridine-FT-IR spectra of acid treated clays are shown in Fig. 4. The pattern showed three different peaks at 1550, 1445 and 1490 cm⁻¹ due to Bronsted, Lewis and both Lewis and Bronsted type respectively. The acidity and the interlayer Al of all the three acid treated clay catalysts were found to be almost the same (Table 1). On the other hand, there was a variation in the amount of structural Al removed by the three acids. Depending on their complexing ability, they remove the Al to different extents. MSA removes more of Al compared to p-TSA thus generating more pores in the structure than p-TSA. PDSA removes more Al than the other two acids. This could be seen in the ²⁷Al MAS NMR spectra as shown in Fig. 6. Removal of structural elements from the octahedral layer results in the formation of pores that might be playing a role in reactions such as, for instance, formation of triacetins in higher amounts in the present study. Similar observations have been made by Venkatesha et al.³⁴ on ester rearrangements with Indian bentonite clay.

Fig. 4 Pyridine adsorbed FT-IR acidity of acid treated clay.

Fig. 5 (A–C) are BJH curves, de-Boer t-plots and adsorption–desorption isotherms of (a) PDSA-clay, (b) MSA-clays and (c) p-TSA-clay.

Fig. 6 ²⁷Al-NMR pattern of all clay samples.

Glycerol acetylation reaction generally is a Bronsted acid catalyzed reaction. Scheme 2 shows the possible reaction mechanism. Carbonyl oxygen of acetic acid gets protonated by the Bronsted acid site and generates an acylium ion which attacks the readily available nucleophilic hydroxyl groups of glycerol, first giving glycerol monoacetin by the elimination of a water molecule. Further, one more acylium of the acetic acid attacks the monoacetin generating 1,3 and 1,2 diacetins. The formation of triacetin however, takes place by further acetylation of the diacetins.

Scheme 2 Plausible reaction mechanism for the triacetin formation.

Enhancement in the formation of triacetin only after the treatment of the clay with organic acid (Table 2) requires further study. Mechanism of triacetin formation perhaps involves the pores created by the dealumination due to acid treatment. MSA treated clay was selected for further study.

3.3 Methanesulfonic acid (MSA) treatment

In order to understand how post treatment changes brought about after treatment with MSA influences the triacetin formation, the clay sample was treated with different concentrations of MSA ranging from 0.1 M to 1 M. Their catalytic activities were evaluated by conducting the reactions at chosen conditions for the maximum selectivity of triacetin, with catalyst amount 1 g, temperature 120 °C, time 60 min and glycerol to acetic acid mole ratio 1 : 3, under microwave irradiation. The primary objective was to understand what factors influence the triacetin selectivity. Results obtained are shown in Table 3 along with surface area, pore volumes and acidity. For comparison, results obtained with Na-clay and Al-exchange clay are also presented. r is the factor correlating triacetin selectivity with different parameters of the catalyst samples.

Table 3 Surface characteristics, acidity and CEC of acid treated clay

Clays	Surface area (m^2 g^−1)	Average pore diameter (Å)	Total pore volume (cm^3 g^−1)	Micropore volume (cm^3 g^−1)	Py FT-IR acidity (μmol g^−1)	CEC (meq g^−1)	Interlayer Al (meq g^−1)	NH3-TPD acidity (μmol g^−1)
Na-clay	35	50.7	0.051	0.009	10	0.92	—	22
Al-clay	33	50.7	0.051	0.009	152	0.92	0.912	258
0.1MSA	58	45.3	0.086	0.016	68	0.88	0.436	107
0.25MSA	94	41.8	0.108	0.022	106	0.84	0.681	186
0.5MSA	158	37.4	0.139	0.031	140	0.82	0.802	228
1MSA	204	34.3	0.178	0.048	148	0.82	0.816	254
Correlation coeff. r	0.978	−0.938	0.955	0.974	0.698	—	—	0.701

---

Na-clay had negligible pore volume and acidity whereas the clay exchanged with Al ions (Al-clay) showed higher acidity but no change in the pore volumes. In contrast, MSA treated clays showed a gradual increase in acidity and pore volume as the concentration of the MSA is increased. Increase in surface area and pore volumes with the increase in concentration of MSA used for treatment is attributed to the increased removal of structural Al. Acidity increased with the increase in concentration of MSA but showed insignificant change beyond 0.5MSA as the interlayer Al reached a maximum value (Table 3). Comparison of different acetylated products for different acid treated catalysts is shown in Table 4. Role of pore volume, acidity and surface area can thus be studied as a function of triacetin selectivity using the results in Tables 3 and 4. Such an attempt has been made by correlating the triacetin formation with the changes in the pore characteristics, surface area and the acidity.

Table 4 Comparison of different catalysts for conversion and yields of product^a,b

Clay samples	VAF	Glycerol conversion (%)	Monoacetin selectivity (%)	Diacetin selectivity (%)	Triacetin selectivity (%)
a Reaction conditions: temperature; 120 °C, time; 60 min, catalyst amount; 1 g and mole ratio; glycerol : acetic acid is 1 : 3.b VAF is the term expressing the product of micropore volume and total acidity (VAF = total acidity × pore volume).
Al-clay	0.09	60	73	18	9
0.1MSA-clay	1.088	33	48	40	12
0.25MSA-clay	2.332	56	38	44	18
0.5MSA-clay	4.321	78	33	43	26
1MSA-clay	7.104	94	23	25	41
1p-TSA-clay	5.041	94	33	38	29

---

3.4 Correlation studies

Correlation studies are shown in Fig. 7–9. To evaluate the correlation coefficient r, mean of the results obtained in triplicate measurements for each experiment was used. When values of micropore volumes of different catalyst samples were plotted against the triacetin yield, a correlation coefficient of 0.987 was obtained (Fig. 7). This shows that enhancement in the micropore volume has a role in triacetin formation. Acidity as determined by FTIR peaks of the pyridine adsorbed showed a poor correlation (r = 0.7) (Fig. 8) with the triacetin formation indicating that acidity alone may not be responsible for the increase in triacetin formation. This could be inferred from the fact that Al-exchange clay shows low selectivity for triacetin though it shows higher acidity (Tables 3 and 4). MSA treated clay samples, however, show increased selectivity for triacetin with the increase in concentration of MSA used for treatment. This could be attributed to both increase in micropore volume and acidity; these two factors together, apparently, play a role in the triacetin selectivity. This was further examined by correlating the two factors with the triacetin selectivity (Fig. 9). Correlation coefficient r obtained showed improved correlation with the product of micropore volume and acidity, referred to as the volume accessibility factor (VAF), than the r values obtained individually with micropore volume and acidity. It appears that pore volume and acidity in the absence of each other are ineffective in the triacetin selectivity as, for instance, in the case of Al-clay. The r values of VAF with triacetin selectivity under different reaction conditions (Fig. 9) were also found to be close to unity with high precision (<0.3%). Acetylation reaction is mainly Bronsted acid catalyzed one, but it can also take place in presence of Lewis acid.³⁴ Acetylation can take place easily, if the catalyst samples have both Bronsted and Lewis acid sites. Correlations were, therefore, made with different combinations such as VAF (Bronsted B, Lewis L, B + L and B/L). Among these, best correlation was obtained for VAF B + L (r = 0.99). This supports the use of VAF calculations to get a better picture of triacetin selectivity in the generated pores (Scheme 2).

Fig. 7 Correlations of micropore volume with yield of triacetin. Reaction conditions: temperature; 120 °C, time; 60 min, catalyst amount; 1 g and mole ratio; glycerol : acetic acid is 1 : 3.

Fig. 8 Correlations of acidity with yield of triacetin. Reaction conditions: temperature; 120 °C, time; 60 min, catalyst amount; 1 g and mole ratio; glycerol : acetic acid is 1 : 3.

Fig. 9 Correlations of VAF with yield of triacetin. Reaction conditions: (A) temperature; 120 °C, time; 60 min, catalyst amount; 1 g and mole ratio glycerol : acetic acid is 1 : 3. (B) Temperature; 120 °C, time; 40 min, catalyst amount; 1 g. (C) Temperature; 120 °C, time; 60 min, catalyst amount; 0.5 g.

3.5 Volume accessibility factor (VAF) for triacetin selectivity

Acetylation involves the generation of acylium ions formed by the protonation on the acid sites. Acetylation of glycerol involves the addition of acylium ions successively to glycerol, initially to form monoacetin and then to diacetins. The latter gets acetylated with acylium ions to form triacetin. Table 4 shows the selectivity of mono, di and triacetins in the acetylation of glycerol with acetic acid using different acid treated clays. It is clear that in the absence of pores, such as in Al-exchange clay, monoacetin formation is favoured. As the concentration of MSA used for treatment increases, micropore volume proportionately increases along with the triacetin formation (Table 4). This could be seen by the fact that as the pore volume increases, triacetin quantity increases and reaches a maximum with 1MSA clay (Table 4). Increase in the micropore volume is thus an indication that the micropores, apparently, are influencing triacetin formation. Volume accessibility factor, VAF, defined as the product of micropore volume and corresponding acidity,^33,34 is a measure of the volume that has access to acid sites to facilitate the formation of triacetin.

It was observed that PDSA, being a stronger acid with two sulphonic acid groups in the molecule, removed more Al enhancing the pore volume to a higher extent (micropore volume = 0.198 cm³ g⁻¹). The VAF of PDSA treated clay was thus very much higher than the other acids. It is not shown in the correlation curve (Fig. 7–9) as it falls out of range suggesting that there is a limitation in using VAF. Apparently, the amount of triacetin formed is determined by the acidity irrespective of the increase in the pore volume beyond certain point.

3.6 Kinetic aspects, molecular and pore dimensions

Rate of the formation of different acetins in presence of (i) Al exchanged clay (low pore volume) and (ii) 1MSA clay (increased pore volume) indicated a considerable rate increase in the case of the latter. The rate constants are tabulated in Table 5.

Table 5 Rate constants for the glycerol conversion and individual products formation^a

Clay	Rate constant (k) min^−1
	Overall	Monoacetin	Diacetins	Triacetin
a Reaction conditions: temperature; 120 °C, catalyst amount, 1 g; mole ratio: glycerol : acetic acid, 1 : 3.
Al-clay	0.0118	0.0066	0.0034	0.002
1MSA-clay	0.0332	0.0193	0.0074	0.0066

---

Results in Table 5 indicate that the increase in rate in the case of 1MSA-clay is clearly because of the pores formed after acid treatment. Involvement of pores in the kinetics of the formation of the acetins is thus evident.

Molecular dimensions of all the products, as calculated by Chem Bio 3D software, are compared with the pore diameter of the catalysts. They are (i) monoacetin – 8.25 × 4.09 Å, (ii) diacetins – 9.59 × 6.56 Å, 8.88 × 6.62 Å and (iii) triacetin – 9.53 × 8.61 Å. These are found to be less than the average micropore diameter as determined by the BJH method of the acid treated clays (18.02 Å). Pore dimensions accordingly are not a hindrance for the formation of acetins. And also internal and external diffusional effects were calculated under the optimized reaction conditions for the clay catalysts using Mear criteria (2.135 × 10⁻⁵) for external diffusion, Weisz–Prater (1.948 × 10⁻⁵) and Koros–Nowak criteria. All the three criteria indicate negligible diffusion limitations.

3.7 Comparison between microwave heating and conventional heating

The glycerol acetylation reactions were carried out in both conventional and microwave heating modes. With both the modes, reaction reached equilibrium at different times. In conventional heating, glycerol conversion reached a maximum of 82% in 48 hours, whereas in microwave heating, conversion of glycerol was found to be 96% in 60 minutes. In the case of microwave heating the reason for the observed rate enhancements is purely thermal/kinetic effect, that is, a consequence of the reaction temperatures that can rapidly be attained when irradiating polar materials in a microwave field.

Reaction in microwave heating reached an equilibrium in 60 min whereas conventional took 48 hours (Table 6). This is attributed to the localized uniform intensive heating of reaction mixture by microwaves. Microwave irradiation is a well-known process which has the ability to reduce reaction times from hours to minutes in the case of polar organic molecular reactions.^42–45 The heating characteristics of a particular reactant, as for instance glycerol, acetic acid and polar acid sites in the catalysts are dependent on their dielectric properties under microwave irradiation conditions. The ability of a specific substance to convert electromagnetic energy into heat energy is determined by dielectric loss factor tan δ. The loss factor tan δ expressed as the ratio of ε′′/ε′, where ε′′ is dielectric loss, which is the efficiency to convert electromagnetic energy into heat energy, and ε′ is the dielectric constant describing the ability of molecules to be polarized by the electric field. A reaction medium with a high tan δ value is required for efficient absorption and for rapid heating (Kappe et al. Angew. Chem. 43, 2004, 6250). The polar acid sites in the catalyst also play a major role. Microwave irradiation, however, has been found to be more effective in bringing the reaction to equilibrium at a faster rate than the thermal radiation^42–45.

Table 6 Catalytic activity of 1MSA clay under conventional mode of heating^a

Time in hours	Glycerol conversion	Monoacetin	Diacetins	Triacetin
a Reaction conditions: temperature; 120 °C, catalyst amount; 1 g and mole ratio; glycerol : acetic acid is 1 : 3.
2	18	15	3	—
6	30	22	6	2
10	43	30	9	4
12	50	33	12	5
24	78	28	32	18
48	82	21	33	28
60	82	18	28	36

---

3.8 Reusability and deactivation

Catalyst samples used in the reaction were analyzed to study the deactivation due to adsorption of reactant and product molecules. The FT-IR spectra Fig. 10 and BET and ICP elemental analysis of samples before and after the reaction did not show any change. Further, the catalyst samples did not deactivate after repeated use for five times showing their reusability. XRD pattern (Fig. 11) peak has been identified with respect to JCPDS software file (13-0135) showed absence of layer swelling indicating the reaction does not occur in the interlayer region indirectly supporting the involvement of pores.

Fig. 10 FT-IR spectra of MSA clay.

Fig. 11 XRD pattern of MSA clay (a) before and (b) after reaction.

The catalytic activity of reused catalyst showed consistent results comparable to fresh catalyst sample as shown in Table 7.

Table 7 Catalytic activity of used catalysts samples^a

Catalyst cycles	% Glycerol conversion	Selectivity triacetin (%)	Selectivity diacetins (%)	Selectivity monoacetin (%)
a Reaction conditions: temperature; 120 °C, time; 60 min, catalyst amount; 1 g and mole ratio; glycerol : acetic acid is 1 : 3.
1^st	94	39	33	22
2^nd	92	38	32	22
3^rd	92	38	33	21
4^th	91	38	32	21
5^th	90	38	31	21

---

4. Conclusion

Treatment with organic sulfonic acids enhances the acidity and pore characteristics of the montmorillonite due to dealumination. Increased pore volume around the acid sites facilitates the multiple acetylation of glycerol. Generated micropores around the acid sites, called volume accessibility, favour triacetin formation. Individual correlations of acidity and pore volumes with triacetin selectivity showed moderate correlation coefficients, while combination of the two factors called volume accessibility factor, VAF gives a better correlation. It is seen that both acidity and pore volume are responsible for the triacetin selectivity.

Acknowledgements

The author (Venkatesha) would like to thank the Rajya vokkalighara sangha Bengaluru, for the facilities provided and L.V. Narendra NMRC IISc, Bengaluru for MAS NMR and Raghavendra RVCE for XRD measurements and finally Dr Parag Vilash Naik ITC for GC-MS analysis.

References

1. J. A. Melero, R. V. Grieken, G. Morales and M. Paniagua, Energy Fuels, 2007, 21, 1782–1791.
2. J. Deutsch, A. Martin and H. Lieske, J. Catal., 2007, 245, 428–435.
3. M. Coy, Chem. Eng. News, 2006, 84, 7–13.
4. I. Kim, J. Kim and D. Lee, Appl. Catal., B, 2014, 148, 295–303.
5. N. J. Venkatesh, Y. S. Bhat and B. S. JaiPrakash, RSC Adv., 2016, 6, 18824–18833.
6. V. S. Marakatti and A. B. Halgeri, RSC Adv., 2015, 5, 14286–14293.
7. M. D. Gonzalez, Y. Cesteros, J. Llorca and P. Salagre, J. Catal., 2012, 290, 202–209.
8. L. Liu, B. Wang, Y. Du and A. Borgna, Appl. Catal., A, 2015, 489, 32–41.
9. S. Garcia, I. Gandarias, J. Requies, M. B. Guemez, S. Bennici, A. Auroux and P. L. Arias, J. Catal., 2015, 323, 65–75.
10. E. S. Vasiliadou, V. L. Yfanti and A. A. Lemonidou, Appl. Catal., B, 2015, 163, 258–266.
11. M. Velasquez, A. Santamaria and C. B. Dupeyrat, Appl. Catal., B, 2015, 160, 606–613.
12. L. Zhou, E. Al-Zaini and A. A. Adesina, Fuel, 2013, 103, 617–625.
13. N. Rahmat, A. Z. Abdullah and A. R. Mohamed, Renewable Sustainable Energy Rev., 2010, 14, 987–1000.
14. A. Brandner, K. Lehnert, A. Bienholz, M. Lucas and P. Claus, Top. Catal., 2009, 52, 278–287.
15. Y. Taguchi, A. Oishi, Y. Ikeda, K. Fujita and T. Masuda, JP Pat., 298099, 2000.
16. A. C. Cope and E. C. Herrick, Org. Synth., 1963, 4, 304–310.
17. R. H. Baker and F. G. Bordwell, Org. Synth., 1955, 3, 141–146.
18. J. Iqbal and R. R. Srivastava, J. Org. Chem., 1992, 57, 2001–2007.
19. M. L. Testa, V. L. Parola, L. F. Liotta and A. M. Venezia, J. Mol. Catal. A: Chem., 2013, 367, 69–76.
20. C. E. Goncalves, L. O. Laier and M. J. Silva, Catal. Lett., 2011, 141, 1111–1117.
21. X. Liao, Y. Zhu, S. Wang and Y. Li, Fuel Process. Technol., 2009, 90, 988–993.
22. S. Shylesh, S. Sharma, S. P. Mirajkar and A. P. Singh, J. Mol. Catal. A: Chem., 2004, 212, 219–228.
23. L. Zhou, T. Nguyen and A. A. Adesina, Fuel Process. Technol., 2012, 104, 310–318.
24. I. Dosuna-Rodríguez and E. M. Gaigneaux, Catal. Today, 2012, 195, 14–21.
25. S. Kale, S. B. Umbarkar, M. K. Dongare, R. Eckelt, U. Armbruster and A. Martin, Appl. Catal., A, 2015, 490, 10–16.
26. S. K. Bhorodwaj and D. K. Dutta, Appl. Clay Sci., 2011, 53, 347–352.
27. V. L. C. Goncalves, B. P. Pinto, J. C. Silva and C. J. A. Mota, Catal. Today, 2008, 133, 673–677.
28. P. Ferreira, I. M. Fonseca, A. M. Ramos, J. Vital and J. E. Castanheiro, Catal. Commun., 2011, 12, 573–576.
29. C. E. Goncalves, L. O. Laier, A. L. Cardoso and M. J. da Silva, Fuel Process. Technol., 2012, 102, 46–52.
30. I. D. Rodriguez, C. Adriany and E. M. Gaigneaux, Catal. Today, 2011, 167, 56–63.
31. P. S. Reddy, P. Sudarsanam, G. Raju and B. M. Reddy, Catal. Commun., 2010, 11, 1224–1228.
32. H. Rastegari, H. S. Ghaziaskar and M. Yalpani, Ind. Eng. Chem. Res., 2015, 54, 3279–3284.
33. N. J. Venkatesha, Y. S. Bhat and B. S. JaiPrakash, Appl. Catal., A, 2015, 496, 51–57.
34. N. J. Venkatesha, B. S. JaiPrakash and Y. S. Bhat, Catal. Sci. Technol., 2015, 5, 1629–1637.
35. S. Ramesh, Y. S. Bhat and B. S. JaiPrakash, Clay Miner., 2012, 47, 231–242.
36. N. J. Venkatesha, Y. S. Bhat and B. S. Jai Prakash, RSC Adv., 2015, 5, 69348–69355.
37. S. Ramesh, B. S. Jai Prakash and Y. S. Bhat, Appl. Catal., A, 2012, 413, 157–162.
38. B. M. Choudary, M. Sateesh, M. L. Kantam, K. V. S. Ranganath and K. V. Raghavan, Catal. Lett., 2001, 76, 231–238.
39. L. Cerveny, K. Mikulcova and J. Cejka, Appl. Catal., A, 2002, 223, 65–72.
40. H. Ciesielski and T. Sterckeman, Agronomie, 1997, 17, 9–16.
41. W. H. Hendershot and M. Duquette, Soil Sci. Soc. Am. J., 1986, 50, 605–609.
42. C. O. Kappe, Angew. Chem., Int. Ed., 2004, 43, 6250–6284.
43. B. M. Chandrashekar, B. S. Jai Prakash and Y. S. Bhat, J. Catal., 2012, 290, 101–107.
44. N. J. Venkatesha, B. M. Chandrasekhar, B. S. Jai Prakash and Y. S. Bhat, J. Mol. Catal. A: Chem., 2014, 392, 181–187.
45. H. R. Prakruthi, B. M. Chandrashekara, B. S. JaiPrakash and Y. S. Bhat, Catal. Sci. Technol., 2015, 5, 3667–3674.

---