Sulfated zirconia as a robust superacid catalyst for multiproduct fatty acid esterification

Abstract

Sulfated zirconia catalysts obtained by employing chlorosulfuric acid show significantly higher activity in the esterification of fatty acids with different alcohols compared with catalysts made using sulfuric acid. The superior performance results from higher sulfur content, larger pores and stronger acid sites. These catalysts are robust and do not leach out sulfonic groups. Catalyst performance depends strongly on the sulfation reagent and the calcination conditions of the intermediate zirconium hydroxide. A series of kinetic experiments was carried out with lauric acid and various alcohols (methanol, 2-ethylhexanol, propanols and butanols). The new catalysts are ca. five times faster when using primary alcohols independent of the alcohol chain length. When using secondary and tertiary alcohols the reaction rate drops considerably. This is explained by a linear free energy relationship of substituent reactivity. The kinetic investigation shows that chlorosulfated zirconia is suitable as a multiproduct catalyst for manufacturing fatty esters, by employing a catalytic reactive distillation process.

---

1. Introduction

Fatty acid esters are valuable fine chemicals. They are used in the production of food additives, cosmetics and pharmaceuticals.¹ The important industrial components are lauric (C12 : 0), myristic (C14 : 0), palmitic (C16 : 0) and oleic (C18 : 1) acids, often combined with methanol, propanol and iso-propanol as alcohols. The current industrial manufacturing relies on batch processes and homogeneous acid catalysts, such as sulfuric and p-toluenesulfonic acids, as well as homogeneous alkali metal hydroxides and methoxides.² These batch processes are flexible, and can be tailored to quality specifications of different products. However, their productivity is low and their energy consumption (for recovering the excess of alcohol) is high.

One attractive alternative for overcoming these drawbacks is Catalytic Reactive Distillation (CRD). This efficient continuous manufacturing method combines reaction and in situ separation and is energy saving. In 2003, we first showed the feasibility of using CRD for manufacturing fatty esters.³ This work was then extended to light alcohols forming azeotropes with water, such as n- and iso-propanols, by using entrainer-enhanced reactive distillation.⁴ CRD is also highly beneficial for manufacturing biodiesel from fatty-acid-rich (so-called high-FFA) feedstock and light alcohols, as it saves energy in the separation process.^5–7

Despite the strong scientific interest, the number of actual industrial applications of CRD processes is low. This reflects the practical difficulty of finding active yet robust catalysts capable of working at the temperatures and concentrations required for maintaining the liquid–vapour equilibrium. Additionally, fine-chemical processes require multi-product catalysts. This means that a variety of products should be manufactured using the same hardware and catalyst by changing only the feed, pressure and temperature. Another key problem when applying CRD to esterification and other condensation reactions is inefficient water removal from the vapour phase. This has two goals: shifting the chemical equilibrium to the products side, and minimizing catalyst deactivation.⁸

Previously we showed that sulfated zirconia (SZr) is a good catalyst for fatty acid esterification with normal alcohols at high temperatures.^5,9 This catalyst is stable up to 170 °C, with practically no leaching of sulfonic groups. Subsequently, Yadav et al. suggested that the sulfur content, and consequently the activity of sulfated zirconia, would increase notably by using chlorosulfuric acid as the sulfonating agent.^10,11 This hypothesis was tested in alkylation and acylation reactions,¹² but not in fatty acid esterification. A recent review emphasised that sulfated zirconia based catalysts are suited to be used in heterogeneous liquid phase reactions, offering new opportunities for developing environmentally benign and friendly processes. However, ensuring high activity, reproducibility and robustness remains a challenge.^12,13

A close inspection of the literature reveals that the activity of sulfated zirconia catalysts depends subtly on a large number of factors. These include catalyst pre-treatment, preparation, and calcination. Farcasiu and Li stated that in the preparation of sulfated zirconia by impregnation, the catalytic activity is influenced by the Zr(OH)₄ precursor, the precipitation pH, and the temperature and duration of the calcination after sulfation.^14,15 A tetragonal phase was found necessary for superacidity, which would justify that using a colloidal sol–gel technique would lead to much higher sulphur content than via the impregnation method.¹⁶ Lercher's group developed these aspects in more detail.^16–19 Thus, highly active sulfated zirconia materials were prepared by the sulfation of crystalline zirconia with gaseous SO₃, circumventing the final calcination step.¹⁷ The aging procedure leads to different activities. Calcination parameters, such as atmosphere, duration and catalyst bed depth, can also influence the catalytic performance.^18,19 Here we report the synthesis, characterization and application of a sulfated zirconia catalyst made by chlorosulfonation. This robust solid acid catalyses the esterification of fatty acids with various alcohols (Scheme 1). We research the influence of the preparation conditions on the reaction rate, and study the possibility of multiproduct catalysis. Finally, we present a simplified practical kinetic model for the case study of the catalytic esterification of lauric acid with n-propanol and iso-propanol. This model can be used in the preliminary conceptual design of a CRD process.

Scheme 1 Overview of the catalytic esterification carried out with lauric acid 1 and various alcohols 2.

2. Experimental

2.1 Materials and instrumentation

Powder X-ray diffraction data were collected on a Miniflex Bench Top XRD with a Cu source. Surface area measurements were done by the BET and BJH methods using N₂ at 77 K on a Thermo Scientific Surfer instrument. All samples were dried under vacuum for 24 h at 300 °C prior to each measurement. Acid site distribution in the catalyst was measured by NH₃-TPD using a thermal conductivity detector (TCD). The catalyst sample (0.3 g) was heated up to 600 °C using He (30 ml min⁻¹) to remove adsorbed components. Then, the sample was cooled at room temperature and saturated for 2 h with 100 ml min⁻¹ of 8200 ppm NH₃ in He as carrier gas. Subsequently, the system was flushed with He at a flow rate of 30 ml min⁻¹ for 2 h. The temperature was ramped up to 600 °C at a rate of 10 °C min⁻¹. The sulfated zirconia catalyst samples were also characterised by atomic emission spectroscopy with inductively coupled plasma atomisation (ICP-AES). A reference sample of commercial (uncalcined) sulfated zirconia was purchased from the Mel Company, England. Esterification reaction samples were analyzed by gas chromatography (InterScience 8000) using a Restek Stabilwax-DA column.

2.2 Procedure for kinetic measurements

Kinetics experiments were performed in stainless steel autoclaves provided with heating, stirring and sampling devices and a reaction cell of 25 ml. This set-up can be employed up to 50 bars and 250 °C, and is suitable when using low boiling alcohols, such as methanol and propanols. The vessel filled initially with fatty acid and catalyst was heated up close to the reaction temperature, when preheated alcohol was added under pressure and continuous stirring. This moment was considered time zero. Reaching steady state took 5–10 min. Samples were periodically drawn off for GC analysis. In the case of higher boiling alcohols, such as 2-ethylhexanol, the experiments were done in a six-parallel glass reactor (Omni-Reactor Station 6100), with each reactor fitted with a magnetic stirrer and reflux condenser. In this set-up the reaction temperature can be raised up to 180 °C under ambient pressure.

General procedure for catalyst preparation. Zirconium hydroxide was precipitated at pH = 9 by adding 25 wt% of aqueous ammonia to an aqueous solution of ZrOCl₂·8H₂O. The resulting zirconium hydroxide was thoroughly washed with an excess of distilled water, removing all chloride ions from the solution. The residue was dried at 140 °C for 16 h, and then impregnated with 0.013 mol of sulfuric or chlorosulfuric acid and calcined in air for 4 h at 650 °C. Table S1 in the ESI† summarises the synthesis conditions for the four catalysts.

3. Results and discussion

3.1 Catalyst synthesis and characterization

We prepared four sulfated zirconia catalysts by impregnating Zr(OH)₄ with either sulfuric or chlorosulfuric acid (noted as SSZr and ClZr, respectively). The catalysts were dried and then calcined in air (see the Experimental section for detailed procedures). A commercial sulfated zirconia catalyst was also tested for comparison. Table 1 presents the surface parameters and sulfur content for all catalysts. We see that the chlorosulfated catalysts have larger pore diameters and volumes than those prepared using sulfuric acid. The median pore size of ClSZr-1 is about 16 nm, almost double that for SSZr-1 and three times larger than the commercial catalyst. Table 1 illustrates the difference in pore size by means of pore volume distribution. This is broader and shifted to larger pore sizes for the chlorosulfated catalyst. The larger pores, higher sulfur content and strong acid sites (vide infra) are all important for better performance in fatty acid esterification.^16,20

Table 1 Physical characterisation of catalysts by surface, pore and sulfur content

Entry	Catalyst	Surface area/m^2 g^−1	Pore volume/cm^2 g^−1	Pore diameter	Sulfate content [%]
				Median/max (nm)
1	SSZr-1	78	0.165	7.5/7.8	3.2
2	CISZr-1	52	0.248	16.5/9.4	7.2
3	CISZr-2	97	0.347	14.5/7.6	7.5
4	CISZr-3	118	0.289	10.4/5.9	7.4
5	SZr-comm	54	0.100	5.8/3.6	1.6

---

A measure of the acid site strength is obtained from the total acidity measured by temperature programmed desorption (TPD) analysis. We see that in the SSZr-1 catalyst most of the acid sites are located at low temperatures (150–220 °C) as well as around 400 °C. In contrast, for the ClSZr-1 catalyst there is a large peak at higher temperature, namely around 530–540 °C. Note that acid sites located above 400 °C are considered as “super-acidic” (the TPD results are included in the ESI†).²¹ TGA analysis (see Fig. S7 in the ESI†) indicates that catalyst dehydration occurs up to 200 °C. However, the high-temperature peaks were ascribed to the desorption of ammonia bound to strong Brönsted and Lewis acidic sites.²¹

3.2 Catalyst testing: activity and robustness

Our first results showed that the catalysts made by chlorosulfonation exhibited much higher activity than those prepared using sulfuric acid. Fig. 1 presents comparative activity measurements for three catalysts, ClSZr-1, ClSZr-2 and SSZr-1. Here we measured the kinetics of the esterification of lauric acid (LA) with 2-ethylhexanol (2-EH) in the molar ratio 1 : 1 at 170 °C with a catalyst amount of 10% w/w with respect to fatty acid. Note that this high temperature was chosen as relevant for testing the industrial capability of sulfated zirconia catalysts, as opposed to the much lower working temperature of commercial resin catalysts. The reactions with ClSZr-1 and ClSZr-2 are considerably faster compared with SSZr-1 (the ratios of the slopes on the quasi-linear zone of the reaction profile are 2.1 and 1.6, respectively). Moreover, the results obtained with ClSZr-3 are close to those using ClSZr-2. This behavior was confirmed by a robustness test: after two hours of reaction the catalyst was filtered off, dried at 120 °C for one hour, and re-used. The activity remained practically unchanged after three runs.

Fig. 1 Lauric acid esterification with 2-ethyl hexanol catalysed by sulfated zirconias obtained by chlorosulfonic vs. sulfuric acid. All reactions were performed in triplicate, agreeing to within ±0.5% between triplicate runs (each point on the graph is an average value of three runs). Reaction conditions: lauric acid : 2-ethylhexanol molar ratio 1 : 1, temperature 170 °C, 10 wt% catalyst. ClSZr-1 (●), ClSZr-2 (○), SSZr-1 (■).

3.3 Comparing catalytic performance using different alcohols

Fig. 2 compares the activity of different sulfated zirconia catalysts in the esterification of lauric acid with n-propanol (NPA) and iso-propanol (IPA), two industrially important alcohols. The chlorosulfated catalysts are much faster than their sulfuric acid analogues. In the case of iso-propanol, a reaction rate of ∼1% min⁻¹ can be achieved with ClSZr-1, which is three times faster than SSZr-1, and seven times faster than the commercial catalyst.

Fig. 2 Activity of various sulfated zirconia catalysts in the esterification of lauric acid with n-propanol (left) and iso-propanol (right). Reaction conditions: acid : alcohol molar ratio 1 : 5; 5 wt% catalyst; temperature 133 °C. ClSZr-1 (●), ClSZr-2 (○), ClSZr-3 (■), SSZr-1 (□), commercial SZr (▲).

Further insight into the effect of chain length and branching of various alcohols was obtained by measuring the kinetics of the esterification of lauric acid using homogeneous catalysis. Fig. 3 compares the reaction profiles using 2 wt% H₂SO₄ at 80 °C for n-propanol, n-butanol, i-propanol, 2-butanol, and tert-butanol, respectively. The alcohol activity follows 1° > 2° > 3°. Moreover, the curves for 1-propanol and 1-butanol are practically identical. The observed reaction rates confirm the prediction based on the linear free energy relationship of the Taft equation:²²

log(k/k_CH3) = ρ^*σ^* + δE_s (1)

Here, the term k/k_CH3 represents the ratio of the reaction rate of a given substituent with respect to a methyl group. The right-hand terms account for the polar and steric effects of the substituent. In addition, there are two correction factors related to the sensitivity of the two terms to specific reaction conditions, respectively ρ* and δ.

Fig. 3 Kinetics of fatty acid esterification by primary, secondary and tertiary alcohols. Reaction conditions: 2 wt% H₂SO₄ catalyst, acid : alcohol molar ratio 1 : 5, temperature 80 °C. NPA (●), IPA (○), 2-butyl (■), 1-butyl (□), tert-butyl (▲).

Table 2 gives some values of parameters compiled from the literature.²³ From these figures, the steric effect for secondary and tertiary alcohols should dominate over the electronic effect. Accordingly, iso-propanol should react about 4.5 times slower than the methanol, while the tert-butyl alcohol about 70 times slower. This prediction is verified here qualitatively for both homogeneous and heterogeneous catalysts. Conversely, we see only a small kinetic effect of the substituent for the primary alcohols. One possible explanation is that for fast reactions, as in this case for primary alcohols, the discrimination of finer kinetic effects is more difficult, as they are masked by the effect of the catalyst amount.

Table 2 Taft polar and steric substituent parameter values

Entry	Parameters	CH3	C2H5	2-C3H7	tert-butyl
1	E s	0.000	−0.070	−0.470	−1.540
2	σ*	0.000	−0.100	−0.190	−0.300
3	E s + σ*	0.000	−0.170	−0.660	−1.840
4	k/kCH3	1.000	0.676	0.219	0.014

---

Murzin et al.²⁴ investigated the esterification of different acids over heterogeneous and homogeneous catalysts. They found that substituent effects followed the Taft relationship for a series of alcohols (methanol, ethanol, n-propanol, n-butanol, i-propanol), but not for the acids (acetic, propanoic, pentanoic). However, the reported values for substituents (σ* = 0, −0.1, −0.12, −0.13 and −0.20 for CH₃, C₂H₅, n-C₃H₇, n-C₄H₉ and i-C₃H₇) account only for polar effects, and not for steric ones. Conversely, Goodwin et al. found that the Taft correlation was applicable to model steric kinetic effects by the esterification of the aliphatic acids with methanol, up to C₄ residues.²⁵

The above results indicate that the reaction mechanism for fatty acid esterification catalyzed by sulfated zirconia should follow the same path as by homogeneous catalysis. Since the effect of alcohol substituents is similar, the rate-limiting step should be the surface chemical reaction.²⁶ However, three important aspects specific to the heterogeneous process should be taken into account:

(1) The occupation of active acid sites is a competition between fatty acid molecules and water. Thus, the presence of water could hinder the reaction at high conversions.

(2) The surface hydrophobicity of the sulfated zirconia catalyst plays a key role. A sufficiently hydrophobic surface would favor the water desorption. Otherwise, this desorption becomes the rate-limiting step.⁵

(3) The diffusion kinetics are sensitive to steric effects of both long-chain fatty acids and iso-alcohols. Larger catalyst pores are thus beneficial. This characteristic should be mainly responsible for the higher activity of chlorosulfated zirconia catalysts, as well as for differences among their samples as given in Table 1.

In the following sections, we focus on the most active catalyst, ClSZr-1, as a multi-substrate esterification catalyst using lauric acid combined with methanol, n-propanol, iso-propanol, and 2-ethylhexanol.

3.4 Esterification of lauric acid with MeOH, NPA, IPA, and 2-ethylhexanol

This reaction with methanol was done in an autoclave because of the low boiling point of methanol. Fig. 4 shows the effect of the acid : alcohol molar ratio, as well as that of temperature at equimolar ratios. Equilibrium is reached after 30 min. Using an excess of alcohol increases conversion, while the excess of acid decreases it. The ClSZr-1 catalyst is very fast, giving an initial reaction rate of ∼6% min⁻¹. This fast reaction rate makes the quantification of short-term temperature effects difficult. The equilibrium conversion increases slightly with the temperature, indicating a small positive heat of reaction.

Fig. 4 Esterification of lauric acid with methanol catalyzed by ClSZr-1; (left) the effect of the acid : alcohol molar ratio: 2 : 1 (■), 1 : 1 (●), 1 : 2 (○), temperature 150 °C and (right) the effect of temperature: T = 133 °C (●), T = 140 °C (○), T = 150 °C (■). Reaction conditions: acid : alcohol molar ratio 1 : 1,5 wt% catalyst.

We then studied the esterification of lauric acid with n-propanol (NPA) and iso-propanol (IPA, Fig. 5). These reactions were run at 133 °C in the presence of 5 wt% ClSZr-1 catalyst. Lower acid : alcohol ratios gave higher conversions (>95% and 80% with NPA IPA, respectively). However, the initial reaction rate also depends on the ratio of reactants. An excess of alcohol can even slow down the reaction, probably due to the alcohol adsorption on the active sites.

Fig. 5 Reaction profiles for the esterification of lauric acid with n-propanol (left) and iso-propanol (right), catalyzed by CSZr-1. Reaction conditions: acid : alcohol molar ratios 1 : 1 (○), 1 : 3 (●), 1 : 5 (□); 5 wt% catalyst; temperature 133 °C.

Subsequently, we ran an analogous experiment using 2-ethylhexanol. Fig. 6 shows the effect of the acid : alcohol ratio (1 : 2, 1 : 1 and 2 : 1) at 160 °C (similar results were obtained for 140 °C, see ESI†). A 1 : 2 acid : alcohol molar ratio gives almost complete conversion, while a 2 : 1 ratio gives a fatty acid conversion close to the stoichiometric limit of 0.5. Note that the contribution of the thermal autocatalysis accounts for over 60% conversion at 160 °C.

Fig. 6 Esterification of lauric acid with 2-ethylhexanol at 160 °C using different acid : alcohol molar ratios: 1 : 2 (■), 1 : 1 (○), and 2 : 1 (□). Note that the catalyst concentration used here is 2% w/w catalyst/acid, allowing a better visualization of the influence of alcohol concentration on the initial part of the reaction profiles. For comparison, the thermal background reaction for a 1 : 1 ratio without any catalyst is also shown (●).

By comparing the reaction profiles of MeOH, NPA, IPA and 2-ethylhexanol (Fig. 4–6), one sees a sharp difference between the primary and secondary alcohols. The difference in the reaction rate between NPA and IPA, estimated by the slope of the conversion-time plot from Fig. 5, is about 6 : 1. Moreover, the initial reaction rate is practically independent of the carbon chain length for primary alcohols, in agreement with the experiments with homogeneous H₂SO₄ as a catalyst.

3.5 Simplified kinetic model for fatty acid esterification with propanols

After validating the catalytic performance of ClSZr-1 in the esterification of lauric acid, we examined the possibilities of using it in a single reactor with multi-product capabilities. This has important industrial implications. Consider, for example, the four following esters: NPA-laurate, IPA-laurate, NPA-myristate and IPA-myristate. All four are important fine chemicals, but their individual demand does not justify a dedicated process. In such cases, using a single CRD column that can be switched between different conditions is advantageous. Indeed, we see that the esterification of myristic acid with IPA gives very similar results to lauric acid (Fig. 7). This shows that ClSZr-1 can be a multiproduct multi-substrate catalyst.

Fig. 7 Esterification of myristic acid (●) and lauric acid (○) with IPA. Reaction conditions: acid : alcohol molar ratio 1 : 1, temperature 150 °C, 5 wt% catalyst.

To apply this approach, however, one must first understand the factors that govern the reaction kinetics. Esterification is an equilibrium reaction. Several papers discuss the kinetics of esterification with propanols.^27–35 From these, we note two important points. First, although in theory the equilibrium constant, K_eq, should depend only on temperature, its value depends in practice on the reactants' molar ratio. This is due to non-ideal behavior, including the possibility of azeotropes and liquid–liquid splits. For this reason, activities rather than concentrations should be used.³⁴ However, the lack of experimental data on the liquid–liquid equilibrium (LLE) and the uncertainty in calculating the activity coefficients by predictive methods such as UNIFAC make this approach difficult.

Second, using heterogeneous catalysis complicates things further, since adsorption and/or desorption steps might be rate-limiting. This is especially true for the adsorption of fatty acids and the desorption of water. For practical reasons, we opted for a pseudo-homogeneous model. This is a robust kinetic model that, when simplified, can nevertheless also describe the system well enough for preliminary design purposes.^3,4,8 More complex (albeit more realistic) models, e.g. Langmuir–Hinshelwood or Elay–Rideal, are discussed elsewhere.^27–29,34

By comparing the kinetic behavior of NPA and IPA at acid : alcohol molar ratios of 1 : 1 and 1 : 3 (plots included in the ESI†), we see that the reaction rates differ. The slopes of the reaction profiles are 5.4 and 6.6, respectively. Note that de Jong et al. reported a ratio of 3.8 for the esterification of myristic acid by NPA and IPA with homogeneous p-toluenesulfonic acid.³¹ The plateau value of conversion at a ratio of 1 : 1 is 80% for NPA and 50% for IPA. When the acid : alcohol ratio is 1 : 3, these values rise to 97% and 80%.

With these results, we propose a set of kinetic equations for preliminary design purposes. The differential equation describing the conversion-time dependency in a batch reactor is

In the above equation X_A is the conversion of the reference reactant (here the fatty acid) with the initial molar concentration c_A0, W the amount of catalyst in %weight with respect to the fatty acid amount, M = c_B0/c_A0 the initial molar ratio of reactants, while K_c the equilibrium constant based on concentrations. If K_c is assumed or known from independent equilibrium measurements, the forward kinetic constant k_f can be obtained by simple linear regression of the following equation:

X ₁ and X₂ are the roots of the equilibrium equation:

(1 − 1/K_c)X² − (M +1)X + M = 0 (4)

respectively,

X₁ = X_eq = (M + 1) − a₂/2a₁

X₂ = (M + 1) + a₂/2a₁ (5)

in which

Note also that the initial concentration of the reference reactant depends on the initial molar ratio of reactants. Assuming additive volumes gives:

with (M_A, M_B) and (ρ_A, ρ_B) the molecular weights and the densities of the reactants. The densities (g l⁻¹) of lauric acid, NPA and IPA are 0.809, 0.693 and 0.660 at 133 °C and 0.796, 0.672 and 0.635 at 150 °C as calculated by the simulation package Aspen Plus™. For the molar ratios 1, 3, and 5 the initial molar concentrations of the fatty acid are 2.918, 1.920 and 1.422 using IPA and 2.920, 1.923 and 1.437 using NPA.

Table 3 presents the Arrhenius parameters for the esterification of lauric acid with iso-propanol and n-propanol in which the equilibrium constant is optimized by minimizing the squared residues of the linear regression plot. The pseudo-equilibrium constants K_c giving the best fit are 3.0 for IPA and 10.5 for NPA, in excellent agreement with the values found by the esterification of myristic acid catalyzed by p-toluenesulfonic acid, as 3.3 for IPA and 11 for NPA.³⁶ For the molar ratios 1, 3, and 5 the normalized sum of residuals is 0.91, 0.89, 0.91 in the case of IPA and 0.97, 0.76, 0.94 when using n-propanol. It can be seen that the data for NPA give a better fit than those for IPA. The larger errors at the molar ratio of three, as well as the shape of the curves in Fig. 5, are due to the limitations of the pseudo-homogeneous model, which does not account for the adsorption of components, namely of the alcohol.

Table 3 Arrhenius parameters

Entry	Alcohols	A (1/mol/s/% cat)	E a/kJ mol^−1
1	IPA	0.3533	34.7
2	NPA	1.3412	33.7

---

Table 3 reports also the activation energies, both ca. 35 kJ mol⁻¹. Similar values were reported in the manufacture of isopropyl palmitate by a heterogeneous catalyst.^30,35,37

4. Conclusions

Sulfated zirconia catalysts prepared using chlorosulfonic acid are 2–7 times more active in the esterification of fatty acids with alcohols compared to the ones prepared using sulfuric acid. The chlorosulfonic catalysts contain significantly more sulfur, and have larger pores, which could account for their superior activity. Moreover, chlorosulfated zirconia catalysts are robust, showing no leaching of sulfonic groups. However, the reproducibility of the activity is sensitive to the synthesis conditions, more specifically to the amount of the sulfonation agent and the drying conditions of the intermediate Zr(OH)₄.

Chlorosulfated zirconias can be used as multi-product multi-substrate esterification catalysts. Employing 5–10% w/w catalyst/fatty acid at 130–150 °C gives reaction rates sufficient for driving a catalytic distillation process. The key advantage of such a process is that by removing the water continuously, one shifts the chemical equilibrium to completion, obtaining high purity products.

The reaction kinetics are independent of the alcohol hydrocarbon chain length when employing primary alcohols (e.g. methanol, n-propanol and 2-ethyl-hexanol). But, when using secondary and tertiary alcohols the reaction slows down considerably. Typically, i-propanol reacts 4–5 times slower than n-propanol, regardless of the catalyst used. This difference in reactivity can be explained by both polar and steric effects. This behaviour suggests that the rate-limiting step is the surface chemical reaction between acid and alcohol, when using the chlorosulfonic catalyst for the esterification with various alcohols of fatty acids up to C12–C14.

Acknowledgements

Most of this work was accomplished in the frame of a consortium supported by the Dutch Technology Foundation (NWO/STW project no. 700.54.653), the University of Amsterdam, TU Eindhoven and the companies Cognis, Oleon, Sulzer, Engelhard (now BASF) and Uniquema. We thank Dr M. C. Mittelmeijer-Hazeleger and Dr J. Beckers for technical support.

References

1. Y. C. Sharma, B. Singh and S. N. Upadhyay, Fuel, 2008, 87, 2355–2373.
2. L. C. Meher, D. Vidya Sagar and S. N. Naik, Renewable Sustainable Energy Rev., 2006, 10, 248–268.
3. F. Omota, A. C. Dimian and A. Bliek, Chem. Eng. Sci., 2003, 58, 3175–3185.
4. A. C. Dimian, F. Omota and A. Bliek, Chem. Eng. Process., 2004, 43, 411–420.
5. A. A. Kiss, A. C. Dimian and G. Rothenberg, Adv. Synth. Catal., 2006, 348, 75–81.
6. A. A. Kiss, A. C. Dimian and G. Rothenberg, Energy Fuels, 2008, 22, 598–604.
7. A. A. Kiss, F. Omota, A. C. Dimian and G. Rothenberg, Top. Catal., 2006, 40, 141–150.
8. A. C. Dimian, C. S. Bildea, F. Omota and A. A. Kiss, Comput. Chem. Eng., 2009, 33, 743–750.
9. F. Omota, A. C. Dimian and A. Bliek, Chem. Eng. Sci., 2003, 58, 3159–3174.
10. G. D. Yadav and G. S. Pathre, Appl. Catal., A, 2006, 297, 237–246.
11. G. D. Yadav and M. Siddiqui, Ind. Eng. Chem. Res., 2009, 48, 10803–10809.
12. H. Matsuhashi, H. Nakamura, T. Ishihara, S. Iwamoto, Y. Kamiya, J. Kobayashi, Y. Kubota, T. Yamada, T. Matsuda, K. Matsushita, K. Nakai, H. Nishiguchi, M. Ogura, N. Okazaki, S. Sato, K. Shimizu, T. Shishido, S. Yamazoe, T. Takeguchi, K. Tomishige, H. Yamashita, M. Niwa and N. Katada, Appl. Catal., A, 2009, 360, 89–97.
13. K. Arata, Green Chem., 2009, 11, 1719–1728.
14. D. Farcasiu and J. Q. Li, Appl. Catal., A, 1995, 128, 97–105.
15. D. Farcasiu and J. Q. Li, Appl. Catal., A, 1998, 175, 1–9.
16. V. Parvulescu, S. Coman, P. Grange and V. I. Parvelescu, Appl. Catal., A, 1999, 176, 27–43.
17. X. B. Li, K. Nagaoka, R. Olindo and J. A. Lercher, J. Catal., 2006, 238, 39–45.
18. X. B. Li, K. Nagaoka, L. J. Simon, J. A. Lercher, S. Wrabetz, F. C. Jentoft, C. Breitkopf, S. Matysik and H. Papp, J. Catal., 2005, 230, 214–225.
19. X. B. Li, K. Nagaoka, L. J. Simon, R. Olindo and J. A. Lercher, Catal. Lett., 2007, 113, 34–40.
20. D. Farcasiu, J. Q. Li and S. Cameron, Appl. Catal., A, 1997, 154, 173–184.
21. A. Corma, V. Fornes, M. I. Juanrajadell and J. M. L. Nieto, Appl. Catal., A, 1994, 116, 151–163.
22. R. Taft, J. Am. Chem. Soc., 1952, 74, 2729.
23. D. A. D. E. Ansyln, Modern physical organic chemistry, University Science, United States of America, 2006.
24. J. Lilja, D. Y. Murzin, T. Salmi, J. Aumo, P. M. Arvela and M. Sundell, J. Mol. Catal. A: Chem., 2002, 182–183, 555–563.
25. Y. J. Liu, E. Lotero and J. G. Goodwin, J. Catal., 2006, 243, 221–228.
26. K. Wilson, D. J. Adams, G. Rothenberg and J. H. Clark, J. Mol. Catal. A: Chem., 2000, 159, 309–314.
27. R. Aafaqi, A. R. Mohamed and S. Bhatia, J. Chem. Technol. Biotechnol., 2004, 79, 1127–1134.
28. S. H. Ali and S. Q. Merchant, Int. J. Chem. Kinet., 2006, 38, 593–612.
29. S. H. Ali, A. Tarakmah, S. Q. Merchant and T. Al-Sahhaf, Chem. Eng. Sci., 2007, 62, 3197–3217.
30. S. Y. Chin, A. L. Ahmad, A. R. Mohamed and S. Bhatia, Appl. Catal., A, 2006, 297, 8–17.
31. M. C. de Jong, R. Feijt, E. Zondervan, T. A. Nijhuis and A. B. de Haan, Appl. Catal., A, 2009, 365, 141–147.
32. N. Gokulakrishnan, A. Pandurangan and P. K. Sinha, J. Mol. Catal. A: Chem., 2007, 263, 55–61.
33. S. R. Kirumakki, N. Nagaraju and K. V. Chary, Appl. Catal., A, 2006, 299, 185–192.
34. M. T. Sanz and J. Gmehling, Chem.–Eur. J., 2006, 123, 1–8.
35. T. Yalcinyuva, H. Deligoz, S. Boz and M. A. Guerkaynak, Int. J. Chem. Kinet., 2008, 40, 136–144.
36. S. Bhatia, A. L. Ahmad, A. R. Mohamed and S. Y. Chin, Chem. Eng. Sci., 2006, 61, 7436–7447.
37. S. Bhatia, A. R. Mohamed, A. L. Ahmad and S. Y. Chin, Comput. Chem. Eng., 2007, 31, 1187–1198.

---

Footnote

† Electronic supplementary information (ESI) available: Detailed experimental procedures. See DOI: 10.1039/c2cy00432a

---