1 Introduction

Due to the depletion of the fossil resources and the growing search for less environmentally harmful chemicals, researchers have been looking for alternatives to supply this energy demand. An interesting alternative to fossil fuels is biodiesel that can be produced from a transesterification reaction using renewable sources such as vegetable oils, animal fats, algae, and fungi oil [1]. However, the increasing production of biodiesel also produces an excess of glycerol as the main byproduct [2]; each transesterified triglyceride molecule generates three ester molecules and one glycerol [3]. Glycerol is a versatile molecule, and it is a precursor to several compounds like glycidol, propanediol, glyceric acid, and glycerol carbonate [4, 5].

One of the alternatives for this glycerol surplus is its use for the production of glycerol carbonate (GC), which is an innovative product, being a sustainable solvent, due to its low toxicity, good interaction with polar and nonpolar solvents, high boiling point, and water solubility. Furthermore, it can be used as polymer monomer [6] and battery carrier [7]. Usually, GC has been synthesized by reaction between glycerol and phosgene, but due to phosgene’s high toxicity, alternative routes have been explored [5]. Another option is the use of urea as reactant, to avoid the use of phosgene, but the reaction produces 2 mol of ammonium as byproduct, which is undesirable in a green route [5, 6]. The use of CO2, as carbonyl source, is attractive with high atom utilization efficiency. The main drawback of this reaction is the low conversion due to severe thermodynamic limitations [8, 9]. Several works in the literature present dimethyl carbonate (DMC) as reactant for GC production, using hydrotalcites, trisodium phosphate, LiCl/CaO, MgCe, among others, as catalysts [10,11,12,13]. Reaction produces 2 mol of methanol for each GC mole produced [14]. It is considered promising since reactions can be performed at milder conditions, and the methanol formed as byproduct can be easily separated [15].

Similar to dimethyl carbonate, diethyl carbonate (DEC) can be used as a source of alkyl carbonates for glycerol transesterification. Glycerol transesterification with DEC for GC synthesis is a green and promising route with nontoxic ethanol as the main byproduct and potential for high GC yield. Àlvarez et al. [16] has studied the synthesis of glycerol carbonate and decarbonate from transesterification of glycerol with diethyl carbonate catalyzed by MgAl hydrotalcites supported on α- and γ-Al2O3. Reactions were performed using continuous reactor and dimethyl sulfoxide (DMSO) as solvent. The authors presented that the type of basic site has a significant influence on the activity, with catalysts that presented Brönsted basicity showing better performance than ones which presented mainly Lewis basic sites. They also found that the stronger adsorption at the Brönsted sites favors the consecutive transesterification of glycerol carbonate. Wu et al. [17] prepared a series of Ce–NiO catalysts for glycerol transesterification with DEC to produce glycerol carbonate in a solvent-free system. The highest activity was observed for the Ce–NiO catalyst with Ce/Ni atomic ratio of 0.2 and calcined at 400 °C (94.14% of glycerol conversion and 90.95% selectivity to GC under the optimum conditions—5 wt% of catalyst, 85 °C, 8 h, glycerol/DEC ratio of 1:3). This catalyst can be reused up to 3 cycles with a slight decrease in its catalytic activity. It was found the well-dispersed NiO particles and the strong basic sites derived from the substitution of Ce4+ by Ni2+ in the CeO2 lattice synergistically promoted the glycerol conversion. The same research group published another work where they prepared a series of Ce1−xCdxO-mixed oxide catalysts for synthesis of glycerol carbonate from glycerol and DEC. The Ce0.7Cd0.3O catalyst annealed at 500 °C presented the highest activity (96.84% of glycerol conversion and 100% selectivity to GC under the optimized conditions—5 wt% catalyst of glycerol, 90 °C, 180 min, glycerol/DEC molar ratio of 1:3). Authors concluded that the conversion to glycerol carbonate was synergistically accelerated by the moderate and strong basic sites [18].

Glycerol transesterification using DEC is presented in Scheme 1 [16]. An equimolar reaction produces one molecule of glycerol carbonate and 2 molecules of ethanol.

Scheme 1
scheme 1

Glycerol transesterification using DEC

Full size image

Glycidol, a possible byproduct of glycerol transesterification, comes from decarboxylation of glycerol carbonate (Scheme 2), as proposed by Liu et al. [19].

Scheme 2
scheme 2

Glycerol carbonate decarboxylation to glycidol

Full size image

Hydrocalumites (HC) are anionic clays that belong to the family of layered double hydroxides (LDH) and differ from hydrotalcites since its main layer has well-ordered Ca/Al distribution due to the large cationic radius of Ca2+ in comparison to Mg2+, while the water and anions are placed in the interlayer spaces. In hydrocalumite crystal structure, the divalent Ca2+ cations are placed in the center of edge-sharing octahedra and partially substituted by trivalent Al3+, resulting in a net positive charge of the layers that has to be compensated by intercalation of gallery anions [20]. HC-type compounds are frequently used as catalysts due to their basic properties. These materials are used in catalysis in their layered form or, more frequently, after calcination, when the layered structure is lost and a Ca/Al-mixed oxide is obtained [21, 22].

To the best of our knowledge, hydrocalumite-based materials have never been studied in glycerol transesterification using diethyl carbonate (DEC) as source of alkyl carbonate. In this context, an exploratory study was performed to evaluate the potential association of Ca/Al-mixed oxides derived from hydrocalumite and DEC as innovative process for glycerol carbonate synthesis. This evaluation took into account the influence of catalyst calcination temperature, catalyst amount in the reaction media, reaction temperature, glycerol to DEC ratio, and different solvents to establish the best conditions for this reaction.

2 Experimental

2.1 Chemicals

DEC anhydrous (≥ 99%), NaOH (≥ 98%), and CaCl2 anhydrous (93%) were purchased from Sigma-Aldrich; glycerol (99.5%) was purchased from ISOFAR; and DMSO, DMF (99.9%), and AlCl3·6H2O (99.5%) were purchased from VETEC.

2.2 Catalyst preparation

The HC was synthesized by coprecipitation method, based on the methodology described elsewhere [21, 23, 24], with Ca/Al atomic ratio of 2. Three solutions were prepared in volumetric flasks: solution A of metal chlorides with concentration 1 mol L−1, solution B containing sodium hydroxide 2 mol L−1, and solution C with water and ethanol in a 2:3 volume ratio. The precipitation was performed using a peristaltic pump where 100 mL of solutions A and B was dripped into 100 mL of solution C with continuous flow rate of 1 mL min−1 in a Teflon reactor equipped with mechanical stirring. After the pumping process, the formed gel was stirred for 1 h to finish the precipitation process. The suspension was aged for 18 h at 95 °C, vacuum filtered, and washed with deionized water at 90 °C to neutralize the pH. The formed hydrocalumite was dried at 100 °C for 18 h and macerated after drying. Ca/Al-mixed oxides were obtained by calcination with 60 mL min−1 air flow at 600, 700, 800, and 900 °C for 3 h using a heating rate of 10 °C min−1. Samples were named as follow: HC600, HC700, HC800, and HC900 for catalysts calcined at 600 °C, 700 °C, 800 °C, and 900 °C, respectively. The non-calcined sample was labeled as HC.

2.3 Characterization methods

X-ray fluorescence (XRF) was employed to determine the chemical composition of hydrocalumite using a Rigaku Primini device.

X-ray diffraction (XRD) was carried out in Rigaku Miniflex II equipment with graphite monochromator using CuKα radiation (30 kV and 15 mA). The analysis was performed from 5 to 90° in 0.05° increment using 2s time count for each point. The phases were identified based on database of JCPDS (Joint Committee on Powder Diffraction Standards, Swarthmore, USA).

Textural characteristics such as specific area, pore volume, and pore size were determined by nitrogen physisorption analysis performed at − 196 °C using Micromeritics TriStar II 3020 device. As a pretreatment, non-calcined and calcined samples were outgassed for 24 h at 150 and 300 °C, respectively.

Temperature-programmed desorption of CO2 (CO2-TPD) was performed in a device coupled to Prisma Plus (Pfeiffer) mass spectrometer in order to identify and quantify the basic sites of the catalysts. The non-calcined sample was pretreated at 150 °C for 1 h under He flow (30 mL min−1) while calcined samples were pretreated at their calcination temperature. CO2 adsorption was performed under a 40-mL min−1 flow of 10% CO2/He gas at room temperature for 30 min. Then, the samples were treated with He flow for 1 h in order to remove CO2 physically adsorbed. At last, samples were heated from room temperature to 1000 °C in a rate of 20 °C min−1 under He flow rate of 40 mL min−1. Relations m/e = 2, 12, 15, 16, 18, 28, 32, and 44 were monitored and m/e = 44 was used for CO2 quantification.

Fourier transform infrared spectroscopy (FTIR) of catalysts was performed with IR Prestige-21 spectrophotometer from Shimadzu with the wavenumber range of 400–4000 cm−1. The dilution used was 3% in KBr, so the mixture was homogenized in gral and pistil, and then, the pellet was prepared to perform the analysis.

Scanning electron microscopy (SEM) was performed on a Hitachi TM-3030 device to evaluate the catalyst morphology before and after the reaction. Energy-dispersive X-ray spectroscopy (EDS) technique was employed to qualitatively and quantitatively determine the elements on the surface of the catalyst, using a TM-1000 Hitachi microscope, coupled with an EDS detector.

2.4 Catalytic tests

In a typical reaction, 19.5 mmol glycerol, different glycerol to DEC molar ratios (1:1, 1:2, 1:3, and 1:4), and 128 mmol DMSO (used as solvent due to immiscibility of reagents) were mixed in a 50-mL two-neck round bottom flask coupled with a reflux condenser on a stirring and heating plate. Reactions were performed at different temperatures (90 °C, 115 °C, and 130 °C) and with different catalyst amounts (5, 10, and 15 wt%). Aliquots were withdrawn during the reaction and then analyzed by Shimadzu HPLC (model LC-20A Prominence) equipped with an Aminex HPX-87H (Biorad) and a refractive index detector (RID-10A). The mobile phase used was 10 mM H2SO4 aqueous solution with a flow rate of 0.6 mL min−1 at 30 °C. To evaluate the glycerol conversion and selectivity to glycerol carbonate, calibration curves were performed with standards of all the reactants and products present in the reaction media. The glycerol conversion and selectivity to GC were determined using Eqs. 1 and 2, respectively.

$$ \mathrm{conversion}\ \left(\%\right)=\frac{\mathrm{no}.\kern0.5em \mathrm{moles}\ \mathrm{of}\ \mathrm{glycerol}\ \left(\mathrm{initial}-\mathrm{final}\right)}{\mathrm{no}.\kern0.5em \mathrm{moles}\ \mathrm{of}\ \mathrm{glycerol}\ \mathrm{initial}}\times 100 $$
(1)
$$ \mathrm{selectivity}\ \left(\%\right)=\frac{\mathrm{no}.\kern0.5em \mathrm{moles}\ \mathrm{of}\ \mathrm{glycerol}\ \mathrm{carbonate}}{\mathrm{no}.\kern0.5em \mathrm{moles}\ \mathrm{of}\ \mathrm{glycerol}\ \left(\mathrm{initial}-\mathrm{final}\right)}\times 100 $$
(2)

3 Results and discussion

3.1 Catalyst characterization

From XRF results, the molar percentages of calcium, aluminum, and chlorine were 57.7, 26.9, and 15.4%, respectively, in synthesized hydrocalumite. Using these values, it was possible to calculate the Ca/Al molar fraction (2.14), which is close to the theoretical value (2.0).

XRD patterns of non-calcined HC (Fig. 1) showed diffraction peaks at (2θ) 11.3°, 22.8°, 23.6°, 31.3°, 39.0°, and 42.7°,which refers to (002), (004), (112), (020), (\( \overline{3}16 \)), and (\( \overline{2}08 \)) planes of a well-crystallized hydrocalumite with molecular formula Ca4Al2O6Cl2·10H2O and monoclinic structure (JCPDS 31–0245) [1, 21, 25].

Fig. 1
figure 1

XRD patterns of the non-calcined, calcined, and used hydrocalumite-based catalysts

Full size image

The synthesized HC was calcined at different temperatures (Fig. 1). HC structure was completely destroyed after calcination at 600 °C, and a broad diffraction peak was formed close to 30° possibly due to amorphous CaO, in accordance with the results presented by Zheng et al. [1] and Rossi et al. [21]. After calcination at 700, 800, and 900 °C, the amorphous phase disappeared and a crystalline structure was formed with a mixture of compounds, Ca12Al14O33 (mayenite, JCPDS 09–0413), CaO (JCPDS 37-1497), and CaClOH (JCPDS 36–0983) [1, 20, 21]. It is important to point out that when the catalysts were calcined at temperatures above 600 °C, the structure of CaClOH did not disappear and peaks related to mayenite became narrower. Pérez-Barrado et al. [24] compared the peaks of mayenite for samples of hydrocalumite calcined at different temperatures. They observed that mayenite peaks were broader for sample calcined at 650 °C and mayenite was completely crystallized at 800 °C.

Mayenite acts as an inert binder and prevents the sintering of CaO [20]. CaO is the active phase responsible for basicity of the catalysts and, consequently, for glycerol transesterification to glycerol carbonate. In this way, the crystallite size of CaO was determined for HC700, HC800, and HC900, from the diffraction peak at 32.2°, which refers to (111) plane of the CaO crystals. HC 700 presented a CaO crystal size of 67.7 (± 8.0) nm; for HC 800, it was 81.6 (± 6.5) nm, and for HC 900, it was 85.0 (± 11.8) nm. It was possible to observe a sintering effect with temperature increasing, and this sintering was more pronounced when calcination temperature was increased from 700 to 800 °C. This is probably due to the fact that mayenite is completely crystallized just at 800 °C [24] and mayenite binder effect is more effective after its crystallization. It is also possible to observe some changes in the XRD profile of the catalyst after reaction at 130 °C. Peaks related to the structure of mayenite remained, but peaks of calcium oxide almost disappeared. It is probably associated with the leaching of calcium oxide from the catalyst. Zheng et al. [1] observed in the XRD analysis of used catalyst (glycerol transesterification with DMC at 70 °C for 3 h using a Ca2Al hydrocalumite calcined at 800 °C) that the crystalline structure of Ca12Al14O33 was similar to that of the fresh calcined sample, but the CaO crystalline phase had disappeared. After ICP analysis, authors confirmed the decrease in Ca2+ content from 12.2 to 9.64 mmol g−1. Granado-Reyes et al. [26] also observed a decrease in the crystallinity of the mayenite phase and the disappearance of CaO peak for the reused hydrocalumite catalyst calcined at 750 °C after glycerol transesterification in DMC and at 90 °C for 3 h.

FTIR spectrum of the as-synthesized hydrocalumite (Fig. 2) is in agreement with reports from Pérez-Barrado et al. [27] and Linares et al. [28]. Two main peaks at 3637 and 3475 cm−1 were identified as OH stretching in metals (AlO–H and CaO–H) [27, 28], as well as another two bands at lower wavenumbers (789 and 586 cm−1) [29]. A weak band at 1623 cm−1 refers to a deformation due to water vibrations in interlamellar region [27, 30]. The peaks at 1470 cm−1 and 879 cm−1 are assigned to C–O stretching and bending, respectively [31].

Fig. 2
figure 2

FTIR of the as-synthetized hydrocalumite

Full size image

The textural properties (specific area, volume, and diameter of pore) of non-calcined and calcined HC are presented in Table 1. It was observed a decrease in specific area and pore volume with increasing calcination temperature. This can be related to the high crystallinity of compounds formed after calcination (mayenite and CaO). The results of textural properties of HC catalysts are in accordance with the literature. Prado et al. [32] prepared a hydrocalumite with specific area of 10.7 m2g−1 before calcination; after calcination at 750 °C, this value dropped to 6.0 m2g−1. Zheng et al. [1] prepared a series of Ca–Al hydrocalumites and observed a decrease in specific area from 13.7 m2g−1 for sample calcined at 400 °C to 6.4 and 5.5 m2g−1 for samples calcined at 600 °C and 800 °C, respectively.

Table 1 Textural properties of the catalysts non-calcined and calcined at different temperatures
Full size table

CO2-TPD analysis (Fig. 3) was performed to investigate the basicity of non-calcined and calcined catalysts. It is possible to observe desorption peaks at different temperatures, which is related to the strength of the basic sites. In this work, CO2 desorption peaks were determined as follows: between 100 and 250 °C as weak basic sites and between 250 and 500 °C as moderate basic sites, and desorption above 500 °C was assigned to strong basic sites. It is known from the literature that the presence of OH− groups, M2+–O2− and Al3+–O2− pairs, and O2− anions of low coordination on the surface is responsible for the weak basicity. The isolated O2− anions and O2− anions adjacent to OH− are related to the presence of moderate basicity, and strong sites are due to the O2− anions different from the others [21, 33, 34]. The x-axis in Fig. 3a and b was adapted for better visualization of CO2 desorption peaks of lower intensity.

Fig. 3
figure 3

CO2 temperature-programmed desorption profiles (CO2-TPD) of HC non-calcined (a) and calcined at different temperatures: (b) 600 °C, 700 °C (c), 800 °C (d), and 900 °C (e)

Full size image

The amount of basic sites and the distribution of basic strength (Table 2) were obtained by peak deconvolution using the multi-peak Gaussian fit. Non-calcined HC presented the highest total basicity, mostly composed by strong basic sites that are related to O2− anions. In calcined samples, there are other phases (CaO and mayenite) that are responsible for total basicity and different basicity profiles [21]. Sample calcined at 600 °C showed lower amount of total basic sites when compared with the sample calcined at 700 °C, which could be explained by the presence of amorphous CaO in HC600, as presented in XRD. It is possible to observe a predominance of strong basic sites in all samples, and the sample calcined at 700 °C presented the highest basicity among the calcined samples.

Table 2 Amount and distribution of basic sites calculated from CO2-TPD experiments for HC non-calcined and calcined at 600, 700, 800, and 900 °C
Full size table

Roelofs et al. [35] evaluated hydrotalcites of different and well-defined platelet sizes and proposed that the active sites are located at the edge of the platelets. Thus, the number of accessible edge active sites is limited by the lateral size of the HT crystallites and increases with decreasing particle size. Hydrotalcites and hydrocalumites belong to the family of layered double hydroxides (LDHs) that after calcination suffer thermal decomposition creating mixed oxides that show good metal dispersion. Thus, in this way, and as previously presented in the XRD analysis, CaO sintering is probably responsible for this basicity profile of the catalysts; the decrease of total basicity with increasing calcination temperature is due to the enlargement of the CaO particle size.

Scanning electron microscopy (SEM) analysis of HC700 catalyst before and after reaction (5 h at 130 °C) is presented in Fig. 4. It is possible to observe the agglomerated particles of the catalyst (Fig. 4a and b), which can be correlated with the low surface area of this material [1]. SEM images of the used HC700 catalyst show that there was a structure modification. Catalyst particles became more agglutinated, and it could hinder the catalyst active sites.

Fig. 4
figure 4

SEM images of a HC700 (× 1000), b HC700 (× 5000), c HC700 used (× 1000), d HC700 used (× 5000)

Full size image

From EDS results, it was possible to know that the content of Ca2+ in catalyst was 36.8% before use and, after reaction, it dropped to 33.9%. It agrees with XRD results (Fig. 1) that show a sharp decrease in CaO peak intensity for the used catalyst, which confirms that calcium was partially leached after use.

3.2 Catalytic reactions

Aiming to evaluate the influence of catalyst loading and reaction temperature in glycerol conversion and selectivity to glycerol carbonate, HC700 catalyst was used, employing 5, 10, and 15 wt% of catalyst (based on the weight of glycerol) at 90 °C, 115 °C, and 130 °C. Figure 5 presents the results of reactions performed at 90 °C, with different catalyst amounts. An increase in catalyst mass from 5 to 10 wt% promoted an increase in conversion from 20 to 40%, while reaction carried out with 15 wt% of catalyst presented intermediate conversion (28%). Concerning GC selectivity, reactions performed at 90 °C with 10 and 15 wt% of catalyst presented similar values (around 65%), while 81% of GC selectivity was reached when 5 wt% of HC700 was used in the reaction media. The same trend is observed with increasing temperature to 115 °C (Fig. 6) and 130 °C (Fig. 7).

Fig. 5
figure 5

Glycerol conversion (a) and GC selectivity (b) of reaction using HC700 as catalyst, 1:3 glycerol to DEC at 90 °C

Full size image
Fig. 6
figure 6

Glycerol conversion (a) and GC selectivity (b) of reactions using HC700 as catalyst, 1:3 glycerol to DEC at 115 °C

Full size image
Fig. 7
figure 7

Glycerol conversion (a) and GC selectivity (b) of reactions using HC700 as catalyst, 1:3 glycerol to DEC at 130 °C

Full size image

Wu et al. [17] performed GC synthesis using DEC, 0.2CeNiO-400 as catalyst at 85 °C, for 6 h, with catalyst loading from 1 to 9 wt%. Authors found that while selectivity to GC remained unchanged, glycerol conversion increased from 73.5% (1 wt%) to 81.6% (5 wt%), and then dropped to 78.2% (9 wt%). In another study, the same research group promoted glycerol transesterification with DEC over Ce0.7Cd0.3O-500 catalyst for 3 h varying catalyst amount from 1 to 9 wt% and observed that conversion increased from 60.3% (1 wt%) to 96.8% (5 wt%); then, it was maintained constantly [18]. Manikandan et al. [36], using DMC and Mg3−xAl1Cux (x = 0.6) as catalyst at 90 °C for 90 min, observed that glycerol conversion increased with increasing catalyst concentration from 5 to 25 wt% (related to glycerol). However, conversion and selectivity declined at higher catalyst concentration, which was ascribed to inadequate catalyst dispersion in the reaction medium.

According to temperature evaluation tests, it can be observed that glycerol conversion increased with increasing temperature, in all catalyst amounts. The selectivity was also affected, with the production of GC being favored by the temperature increase. Wu et al. [17], using DEC and Ce–NiO catalyst, also observed that glycerol conversion increases with temperature from 70 to 95 °C. However, for this catalyst, GC selectivity dropped from 100% at 85 °C to 62.5% at 95 °C. It is worth mentioning that besides GC, the authors also observed the presence of glycidol as byproduct. Since promising results were found with 10 wt% of catalyst at 130 °C, these conditions were chosen to evaluate catalyst calcination temperatures, glycerol to DEC ratios, and different solvents.

In a deep kinetic study performed by Steban et al. [37] where the authors promoted glycerol transesterification with DMC using a homogeneous catalyst CH3OK, under mild conditions, they observed that temperature (from 50 to 70 °C) and the amount of catalyst (from 1000 to 2500 ppm) have a positive effect on the reaction kinetics.

Since different calcination temperatures affect phase composition and basicity profile, as presented in XRD and CO2-TPD results, reactions were performed with hydrocalumite as-synthetized and calcined at different temperatures (600, 700, 800, and 900 °C) (Fig. 8). Reactions were carried out using 10 wt% of catalyst at 130 °C. Evaluating glycerol conversions after 5 h (Fig. 8a), all catalysts presented similar values (around 40%), except HC700 (90%).

Fig. 8
figure 8

Glycerol conversion (a), GC selectivity (b), and proportion of basic sites × conversion (c). Reaction conditions: hydrocalumite (non-calcined and calcined at 600, 700, 800, and 900 °C), 1:3 of glycerol to DEC, 5 h, 130 °C, and 10 wt% of catalyst

Full size image

Comparing the glycerol conversions with the basicity of the samples (Table 2), it is possible to observe that glycerol transesterification is not dependent on the total basicity. Total basicity of the non-calcined HC is by far the highest one (Table 2); however, this catalyst did not show the highest conversion. This finding suggests that CaO, formed after hydrocalumite calcination, is more active for glycerol transesterification than HC itself. It is supported by the work of Ochoa-Gòmez et al. [15] that evaluated the catalytic performance of different acid and basic heterogeneous and homogeneous catalysts for the transesterification of glycerol with DMC. Authors found that CaO was the most active, less expensive, and less toxic active phase for this reaction. As previously discussed, in HC600, CaO was not in its crystalline form. It explains its lower activity if compared to HC800 that presented lower total basicity but higher activity.

The total basicity of calcined catalysts is different when comparing them and it affected glycerol conversion. HC600, 800, and 900 presented lower amounts of basic sites compared to HC700 that was the best catalyst in this work (~ 90% of glycerol conversion). It was discussed above that CaO crystallites are sintered with the temperature increasing above 700 °C; this could explain the lower total basicity and consequently lower activity of HC800 and HC900. Figure 8c shows that, among calcined catalysts, HC700 presented higher total basicity, but the profiles of the basic site distribution are very similar, with predominance of strong basic sites in all of them. This signalizes that when the same active species are being evaluated, in this case CaO, the higher the total basicity, the higher the activity. These findings are supported by some results published in the literature. Liu et al. [38] also demonstrated that the activity of the hydrotalcite-like layered double hydroxides (LDHx) as catalysts for glycerol transesterification with DMC is proportional to the surface density of basic sites. Zhang et al. [39] evaluated a series of mesoporous CaO–ZrO2 catalysts with high surface area in a continuous system using DMC, and they found the catalyst 0.6CaO–ZrO2 reached the highest glycerol conversion (97%) at 90 °C. The authors associated the higher basicity of this catalyst to its better performance. In another hand, Alvarez et al. [40] evaluated glycerol carbonate syntheses via transesterification reaction using DEC and with different hydrotalcite-like compounds that were activated by calcination and followed by rehydration under ultrasound. The authors found that the catalytic activity was not proportional to the number of basic sites present in the catalyst, associating the highest activity of the rehydrated hydrotalcite with basic Brønsted sites. Our work differs from Alvarez et al. [40] because for hydrocalumite-derived catalysts; the highest activity was associated with an increase in the amount of total basic sites (HC700).

Concerning GC selectivity, the catalysts presented some instability during reactions, except for HC700 and 800. The highest GC selectivity (~ 92%) was obtained with HC700 that presented a stable performance during 5 h of reaction.

Zheng et al. [1], using HC calcined at different temperatures as catalysts for glycerol transesterification with DMC, observed that higher conversion was obtained with HC calcined at 800 °C, which presented the highest total basicity (85.3 μmol g−1) and the maximum proportion of strong basic sites. Granados-Reyes et al. [22] evaluated several HC materials, in as-synthesized and calcined (450 and 750 °C) forms, for glycerol transesterification with DMC. The catalyst calcined at 750 °C was the most active, with 91% of conversion and 80% of GC selectivity, at 90 °C. Authors associated this higher activity to the higher basicity strength (evaluated by Hammett indicators).

Wu et al. [18] stated that the glycerol conversion in transesterification with DEC was synergistically accelerated by the moderate and strong basic sites, using CeO2–CdO catalyst. Zhang et al. [39] evaluated a series of mesoporous CaO–ZrO2 in glycerol transesterification with DMC, using a continuous system at 80 °C, and observed good correlation between the catalytic activity and the total amount of surface basic sites. Bai et al. [41] tested a series of NaOH/Al2O3 (from 20 to 80% of NaOH) catalysts for glycerol transesterification with DMC using mild conditions. With 80% NaOH/Al2O3 catalyst, 98% of conversion and 93.8% of selectivity were reached. Authors concluded that the catalytic properties of the catalysts were influenced by the basic strength presented in the solid; the strong basic strength is not beneficial to the synthesis of GC.

Glycerol transesterification for the production of glycerol carbonate with DEC is strongly influenced by the concentration of reagents. In this way, the effect of the glycerol/DEC molar ratio (1:1, 1:2, 1:3, and 1:4) was evaluated, and the results are presented in Fig. 9. It is possible to observe that at low reactant molar ratio (1:1), glycerol conversion was around 45% after 5 h; it increases with the increase in DEC concentration, until 90% with 1:3 glycerol to DEC molar ratio. Further increase in glycerol to DEC molar ratio to 1:4 did not lead to higher glycerol conversion. This may be related to the increase in the viscosity of the reaction media, hampering the interaction between the reactants in the catalyst active sites. Another possible reason is supported by the results of Wu et al. [18] that performed the synthesis of glycerol carbonate from glycerol and diethyl carbonate (DEC) with 5 wt% of Ce0.7Cd0.3O-500 catalyst for 3 h at 90 °C. They observed the same behavior and attributed this to the low catalyst concentration in the reaction media, when a surplus of DEC was added, bringing on the drop of contact rate between reactants and catalyst active sites. GC selectivity is almost constant with time for all glycerol to DEC ratios. However, after 3 h, reactions with lower DEC concentration presented a drop in GC selectivity, which may happen due to the high catalyst concentration at the low reactant molar ratio, which may favor the GC decarbonylation to glycidol.

Fig. 9
figure 9

Glycerol conversion (a) and GC selectivity (b) of reactions using HC700 with different glycerol to DEC molar ratios, at 130 °C

Full size image

An important point to be evaluated in a reaction is the solvent. Under the evaluated conditions, glycerol and diethyl carbonate are immiscible, making the use of a solvent fundamental for this work. Besides, there is a lack of results in literature that explore the influence of different solvents in glycerol transesterification. It is an important contribution for possible reactions in continuous system [17, 18]. We have noticed that several works in literature, using DEC and DMC [26, 42, 43] as alkyl carbonate sources, performed glycerol transesterification to GC in batch system, without the use of a solvent. Alvarez et al. [40, 44, 45], studying the performance of different hydrotalcite-like compounds, in three different studies, carried out the reaction at 130 °C in solvent-free media, but they worked in a great excess of DEC (1:17 of glycerol/DEC molar ratio), very long reaction times (~ 10 h), and GC yields below 75%. Wu et al. [17, 18], in two distinct works, also performed glycerol transesterification at 80 °C without solvent. In both works, the authors presented that the system was stirred variously, which does not contribute to the understanding of the experimental procedures. The choice of the solvent plays a key role in the solubilization of the reactants, promoting a more effective contact between them, and may influence positively or negatively the conversion and selectivity. In glycerol transesterification, glycerol is strongly polar and diethyl carbonate is aprotic. Thus, in this way, four different solvents were tested: water, ethanol (polar and protic), DMSO, and DMF (polar and aprotic). When water and ethanol were used as solvents, no glycerol conversion was observed. Thus, it is possible to conclude that glycerol conversion to GC in the presence of DEC is dependent on the synergistic combination of an aprotic and polar solvent. Lari et al. [46] observed a strong reduction of the reaction rate with the protic liquids, which was expected since glycerol activation on basic solids occurs via deprotonation of a hydroxyl group. The authors also stated that protic solvents compete with the triol for the active centers, hindering them for the desired transformation.

Comparing the results of DMSO and DMF as solvents (Fig. 10), it is clear that DMSO is a better solvent for this reaction. With DMSO, 90% of glycerol conversion and 91.7% of GC selectivity were observed after 5 h of reaction. These values are higher than that presented by reaction performed with DMF: 76.7% of glycerol conversion and 82.2% of selectivity. Phadtare et al. [47], using DMF as solvent, DMC and a crystalline LaCoO3 perovskite as catalyst at 120 °C for 1.5 h, obtained 98% of conversion and 77% of GC selectivity. The better performance of DMSO when compared with DMF, since both are aprotic and polar, may be associated with its higher polarity (7.2 against 6.4), which favors better product desorption [16].

Fig. 10
figure 10

Glycerol conversion (a) and GC selectivity (b) of reactions with different solvents, using HC700, glycerol to DEC molar ratio of 1:3, at 130 °C

Full size image

To the best of our knowledge, the investigation of HC as catalyst for the synthesis of GC using DEC has not been reported in the literature yet. Table 3 shows a comparison between the results of this work and the literature for glycerol transesterification to produce GC. For the sake of comparison, reactions using DMC are also presented. It is possible to observe that reactions performed with DMC are able to produce GC with higher conversion and selectivity (or yield) in shorter reaction times and under mild conditions. However, besides glycidol, which is a possible byproduct from glycerol transesterification using both DEC or DMC, reactions carried out using DMC also generate methanol. On the other hand, reactions performed using DEC generate ethanol that is less harmful to health and environment.

Table 3 Comparison between the results obtained in this work and results presented in the literature for glycerol transesterification to glycerol carbonate
Full size table

There are some studies in the literature that address the use of hydrocalumite-type materials for the production of glycerol carbonate; however, in these studies, DMC was used as alkyl carbonate [1, 22, 41, 42]. Comparing these results with ours, it can be seen that, except for the results presented by Zheng et al. [1], which obtained higher conversion and selectivity, our catalytic system stands out, especially taking into account the selectivity. Comparing reactions performed with DEC and DMSO, both at 130 °C, the reaction using HDT/α-Al2O3 [16] presented higher conversion (99.7%) but lower selectivity (76.4%), after 8 h, than the reaction presented in this work, using HC700 and after 5 h. Higher conversion and selectivity were reached when CeO2–CdO was used as catalyst in the absence of solvent, at 90 °C and after 3 h [18]. The drawback of this approach is Cd toxicity and high market prices. The same research group, in another work, presented the use of 0.2NiCe-400 catalyst and DEC in the absence of solvent at 85 °C [17]. They achieved conversion (94.1%) and GC selectivity (90.9%) similar to this work; nevertheless, the reaction lasted 8 h to reach these results.

4 Conclusions

In the present work, hydrocalumite was evaluated as a potential catalyst for glycerol transesterification to glycerol carbonate (GC) using diethyl carbonate. Encouraging results were found using catalyst Ca/Al-mixed oxides calcined at 700 °C at 130 °C and with 1:3 glycerol/DEC molar ratio: 90% of glycerol conversion and 92% of GC selectivity, after 5 h of reaction. It was found that, for calcined catalysts, the active phase is crystalline CaO, obtained after calcination at 700 °C. This catalyst presented higher concentration of basic sites that is a key factor to achieve high catalyst activity. Different solvents were tested, which demonstrated that the ideal one must be polar and aprotic; thus, in this way, DMSO proved to be a good option.