1 Introduction

In the field of heterogeneous catalysis, the aim of several studies has been the development of solid catalysts with acidic properties, and metal oxides have been the basis of numerous research works. In this context, zirconium oxide or zirconia (ZrO2) is notable for its high thermal stability, as well as for its redox and acid–base properties. Furthermore, the addition of oxoanions (SO4 2−, WO4 2−, and MoO4 2−) leads to increase in its acid strength [1]. Sulfated zirconia (SO4 2-/ZrO2) has strong acid sites and crystalline properties that allow it to catalyze the isomerization of light paraffins and other reactions promoted by acidity [2–5]. The preparation method selected as well as the synthesis variables involved have a marked influence on the catalytic activity of sulfated zirconia, and this has generated abundant literature about the effect of different preparation parameters on the final properties of this material [5–8]. Sulfuric acid is used as sulfating agent to obtain SO4 2−/ZrO2; nevertheless, the modification of zirconia with other mineral acids (phosphoric or boric) produces materials with acidic properties suitable for catalytic applications, although these have lower acid strength compared to its sulfated counterpart [9–11]. On this matter, we can refer to the modification of ZrO2 with boric acid in aqueous solution, using the incipient wetness technique. It should be noted that the support was previously calcined at 500 °C, and it was heat-treated at 350 °C after impregnation. This catalyst was highly active and selective for the synthesis of ε-caprolactam, being superior in comparison to other supports treated with boric acid (Al2O3, TiO2, SiO2, MgO, and HZSM-5) [12]. In a later work, these researchers suggested that a better method for the preparation of the catalyst may be a direct modification of the hydroxide precursor with boric acid, followed by calcination at a temperature as high as 600 °C [13]. Afterwards, some researchers have generally obtained catalysts from an amorphous precursor of zirconia, impregnated with an aqueous solution of boric acid and, subsequently, heat-treated at different temperatures. Boron-modified zirconia (B2O3/ZrO2) is active in various industrially important organic transformations such as benzoylation of anisole [10], methylation of phenol [14], transesterification of β-ketoesters [15], and condensation of Knoevenagel [16, 17], among others. In those studies, hydroxide precursor was prepared from zirconium salt precipitation; the powder obtained was added to an aqueous solution of boric acid, stirring constantly until slurry was obtained. The calcination temperature of the solid, obtained after drying, was within the range of 500–750 °C [10, 14–16]. On the other hand, the synthesis of support by micellar method was reported, varying the zirconium concentration in the starting solution, as well as the boron concentration in the solution used for impregnation. In contrast to the above-mentioned studies, these catalysts were calcined at a relatively low temperature (320 °C); however, they showed activity in the acetylation of 2-phenoxyethanol [18].

In particular, we considered it feasible to continue exploring the influence of different synthesis parameters on the physicochemical properties of boron-modified zirconia. According to literature, ultrasonic waves can enhance and promote the active species dispersion on the support surface, thereby improving the catalyst performance [19, 20]. Therefore, the aim of this work is to compare and discern whether there are differences between the materials prepared by conventional impregnation and those obtained by ultrasonic impregnation, applying the vibration generated in an ultrasound bath.

2 Experimental

2.1 Chemicals

Zirconium(IV) butoxide solution (Zr[O(CH2)3CH3]4; 80 wt% in 1-butanol, Sigma-Aldrich), 1-butanol (CH3(CH2)3OH; 99.8% anhydrous, Sigma-Aldrich), distilled water (Quimicrón), and boric acid (H3BO3, 99.5%, Técnica Química) were used for the preparation of the catalysts.

2.2 Synthesis

Zirconium hydroxide was prepared via sol-gel using molar ratios established in alcohol/alkoxide = 12 and water/alkoxide = 8. Zirconium(IV) butoxide was dissolved in 1-butanol by constant stirring at 70 °C. The homogenized solution was hydrolyzed by slow drip with distilled water and the gel produced was aged 72 h at room temperature. Subsequently, the excess of solvent was evaporated at 100 °C in a muffle furnace. The powder obtained was added to an aqueous solution of boric acid to deposit 3.7% theoretical weight of boron on the support, and this suspension was stirred for 1 or 4 h. Similarly, the impregnation was performed using the vibration generated in an ultrasound bath Crest Ultrasonics model CP1200D (45 kHz, sonic power: step 3). The modified zirconium hydroxides were dried at 110 °C, and then they were calcined for 3 h at 600 °C, at a heating rate of 5 °C/min under air dynamic atmosphere. The nomenclature identifies the synthesized catalysts according to the method (C = conventional or U = ultrasonic) and impregnation time: ZB1hC, ZB4hC, ZB1hU, and ZB4hU. In addition, a portion of unmodified zirconium hydroxide was calcined in the same way to obtain the pure zirconia, ZrO2.

2.3 Characterization

Precursors of materials (zirconium hydroxides: pure and boron-modified) were thermally analyzed in airflow (30 mL/min) at a heating rate of 5 °C/min from room temperature to 900 °C (TA Instruments SDT 2960 Simultaneous DSC-TGA). The X-ray diffraction (XRD) patterns of the oxides were obtained in the 2 θ angular range from 20 to 80°, using Cu Kα radiation (λ = 0.15406 nm) and a scanning rate of 0.2 rad/s (Bruker D8 Advanced). Infrared spectra were collected using pellets of the sample in KBr, and the measurements were performed in the range of 400–4000 cm−1 with 16 scans and a resolution of 4 cm−1 (Perkin Elmer Spectrum 100). The textural properties of the solids were determined from the adsorption-desorption isotherms of nitrogen; the samples were degassed previously at 250 °C (Quantachrome Autosorb-1). The surface morphology of the materials was observed by means of electronic images obtained using a scanning microscope with an accelerating voltage of 2.0 kV (Jeol JSM 5800LV). The acidic properties were evaluated by the potentiometric titration technique with n-butylamine and by the ethanol decomposition reaction. After 3 h of constant stirring, the solid suspended in acetonitrile was titrated using a solution of n-butylamine 0.025 N and the variation of the electrode potential was recorded until the equilibrium was reached (Chemcadet Cole Parmer). The ethanol decomposition was carried out at 300 °C with 100 mg of catalyst placed in a fixed-bed tubular reactor. The microplant of catalytic activity operates at atmospheric pressure and continuous flow. The reaction products were analyzed on-line by a gas chromatograph (Varian 3400 FID).

3 Results and discussion

3.1 Thermal analysis (DTA-TGA)

The thermal behavior of the synthesized hydroxides is shown in Figs. 1 and 2, analyzing from room temperature to 900 °C. Boron-modified solids showed less weight loss compared to their pure counterpart, consequence of the boron presence. All DTA profiles presented a similar behavior at low temperature (50–100 °C), which corresponds to the removal of physically adsorbed moisture [18, 21]. After that endothermic process, in the range of 100–900 °C, the modified materials had similar percentages of weight loss: 10.8, 10.9, 10.1, and 10.0, respectively, for the final oxides ZB1hC, ZB4hC, ZB1hU, and ZB4hU. Therefore, the observed differences in the values of total weight loss are not related to time or impregnation method, rather it is a result of the moisture present in each sample. Melada et al. [6] analyzed TGA profiles of sulfated zirconia xerogels and indicated that the weight loss recorded in the 150–350 °C range was due to dehydration and dehydroxylation processes, as well as the elimination of carbonaceous species from the surface. Accordingly, the maximums located at 269 and 303 °C in the DTA curve of pure precursor (Fig. 2) are attributed to the combustion of residual organic groups, since the zirconium hydroxide was obtained from the hydrolysis of an alkoxide [6, 21]. Precursors of the boron-modified oxides did not present those peaks, but they exhibited a gradual weight loss in the range of 100–500 °C (Fig. 1), which would include the changes reported by Melada et al. [6]. Finally, the exothermic signal observed at 405 °C is assigned to the transformation of an amorphous state to a crystalline zirconia [18, 21]; nevertheless, this change occurred at a higher temperature (682–685 °C) by effect of the addition of boron, so it follows that the boron presence retards the crystallization process of the material. Ivanov et al. [22] observed similar behavior due to the introduction of sulfate ions, which favors the stabilization of the amorphous zirconia and delays the crystallization process. For their part, Osiglio et al. [18] indicated that the crystallization of zirconia, obtained by the micellar method, showed a maximum at 440 °C in the DTA diagram; however, this signal was presented with peaks at 630, 732, and 697 °C, depending on the content boron in the samples. The materials analyzed in this study were modified with the same theoretical content of boron (3.7 wt%); thus, the signal position attributed to the crystallization is practically the same for all solids, indistinctly of the method or impregnation time. Above 500 °C, there were no significant weight changes in the TGA profiles of the modified materials; hence, it is important to note that there is no evidence of boron species evacuation. This last observation allows us to infer a strong interaction between the boron and zirconia framework, and may represent a convenient feature compared to its sulfated counterpart. It is well known, in the particular case of sulfated zirconia xerogels, that sulfate ions begin to decompose at temperatures above 600 °C, generating both an endothermic signal in the DTA profile and a marked weight loss in the TGA curve [6, 22]. Moreover, the loss of sulfate groups also occurs above 600 °C even in sulfated zirconia previously calcined [2], which is a disadvantage for regeneration processes at high temperatures. Conversely, boric acid was thermally analyzed under the same conditions as the catalytic precursors. The endothermic changes at 118 and 161 °C are associated with the dehydration of boric acid (DTA profile not shown), thereby generating the formation of boron oxide (B2O3) [23, 24].

Fig. 1
figure 1

TGA profiles of the catalytic precursors (zirconium hydroxides: pure and boron-modified): a pure, b HB1hC, c HB4hC, d HB1hU, and e HB4hU

Full size image
Fig. 2
figure 2

DTA profiles of the catalytic precursors (zirconium hydroxides: pure and boron-modified): a pure, b HB1hC, c HB4hC, d HB1hU, and e HB4hU

Full size image

3.2 X-ray diffraction

The XRD pattern of ZrO2 powder calcined at 600 °C (Fig. 3) mainly presented diffraction lines assigned to the tetragonal phase at 2θ = 30.210°, 34.570°, 35.248°, 42.966°, 50.201°, 50.705°, 59.266°, 60.165°, 62.823°, 72.919°, and 74.536° (PDF 01-072-7115). The presence of the monoclinic phase was identified in low proportion, showing very weak signals at 2θ = 28.182° and 31.472° (PDF 01-081-1314). In accordance with the DTA profile corresponding to the precursor ZrO2 (shown in Fig. 2), the transition from amorphous phase to a crystalline structure occurs at 405 °C; therefore, heating at 600 °C generated well-crystallized zirconia. On the other hand, the XRD patterns of the modified solids (Fig. 4) confirmed the delay in the crystallization of these materials due to the boron incorporation, also according to the DTA curves (Fig. 2) that exhibited this structural change with a maximum around 680 °C. The intensity of the diffraction signals decreased significantly, showing an amorphous tendency; however, the broad peak centered on 30° in 2θ scale suggests the formation of an incipient crystalline phase (tetragonal). Conversely, Zaleska et al. [25] synthesized titanium dioxide (TiO2) of anatase phase, using the sol-gel method, and they reported similar behavior. The addition of H3BO3 inhibited crystallite growth and transformation from the amorphous state to the anatase structure. The XRD pattern of TiO2 with 0.5 wt% of boron presented only a very weak signal attributed to the anatase phase, and the material was amorphous when the boron content increased to 10 wt% [25]. In this work, in order to identify the presence of diffraction signals corresponding to boron oxide in the modified materials, a sample of boric acid was calcined at 600 °C and then it was analyzed by XRD. As noted above, B2O3 is formed from the dehydration of H3BO3. In the XRD pattern of the boron oxide (not shown) a narrow and intense peak at 2θ = 27.6° was observed [26]. It is important to note that this signal is not seen in the diffractograms of the modified solids (Fig. 4), either due to good dispersion of B2O3 in amorphous state or because the crystallites are so small that they cannot be detected by this technique [18].

Fig. 3
figure 3

XRD pattern of the material ZrO2

Full size image
Fig. 4
figure 4

XRD patterns of modified materials: a ZB1hC, b ZB4hC, c ZB1hU, and d ZB4hU

Full size image

3.3 Infrared spectroscopy (FT-IR)

The infrared spectrum of the solid ZrO2 is shown in Fig. 5. In the region assigned to the stretching frequency of O–H bonds (3800–3000 cm−1) only a minuscule signal centered at 3400 cm−1 was presented; this slightly pronounced band indicates the removal of surface water because of the high calcination temperature [27, 28]. Moreover, the spectrum also showed an intense band in the region of 800–400 cm−1, with signals around 580 and 495 cm−1 that characterize the crystalline zirconia, agreeing with thermal analysis and XRD [27]. The spectra of the boron-modified oxides exhibited the same signals, indistinctly of the method or impregnation time (Fig. 6). The bands observed in the region of 1500–900 cm−1 corroborated the presence of boron, since they are attributed to stretching of B–O bonds [29]. Boron may be present in two structural units: trigonal and tetrahedral. The minimums accentuated around 1370 and 1010 cm−1, respectively, correspond to BO3 and BO4 species [30]. Furthermore, because of the interaction of boron with zirconia, the band assigned to stretching Zr–O bonds showed a shift of the signals to 685 and 505 cm−1. These spectra also presented a signal at 1630 cm−1, which is attributed to bending vibrations of -(H–O–H)- bonds [27, 28] and a band with minimums in the region of 3400–3200 cm−1, due to stretching of O–H bonds. On the other hand, once thermal analysis was completed (from room temperature to 900 °C), the remaining materials were analyzed by infrared spectroscopy. In Fig. 7, the similarity between the spectra of the samples ZB1hC-R and ZB1hU-R and the spectrum of boric acid can be seen, where absorptions close to 1460 and 1190 cm−1 represent BO3 units [12]. This reveals that the boron species are present in the synthesized materials even at high temperatures of heat treatment.

Fig. 5
figure 5

FT-IR spectrum of the material ZrO2

Full size image
Fig. 6
figure 6

FT-IR spectra of the modified materials: a ZB1hC, b ZB4hC, c ZB1hU, and d ZB4hU

Full size image
Fig. 7
figure 7

FT-IR spectra of a H3BO3 and recovered materials after the thermal analysis: b ZB1hU-R and c ZB1hC-R

Full size image

3.4 Textural properties (N2 physisorption)

In this work, the boron incorporation promoted an increase in the specific area of the materials in comparison to the pure ZrO2, reported in Table 1; therefore, boron had a stabilizing effect that prevented the oxides sintering during the calcination step. In correlation to the XRD analysis, only the pure material developed a well-crystallized tetragonal phase, while the modified solids had a lower degree of crystallinity (being mainly amorphous); thus, it is consistent that the materials ZB1hC, ZB4hC, ZB1hU, and ZB4hU have specific area values of the order of 130–170 m2/g. In general, the solids showed similar textural properties, independently of the method or modification time with boric acid. ZrO2 presented a type III isotherm (Fig. 8), which is convex toward the abscissa axis for the whole range of relative pressure [27]. This type of isotherm is characteristic of nonporous solids, or possibly macroporous, that have a low energy of adsorption. This is according to the low volume of nitrogen adsorbed by this material. Conversely, the nitrogen physisorption analysis performed to samples of the modified solids revealed isotherms of hybrid form of type I-IV with a narrow hysteresis, appreciable in the materials ZB1hU and ZB4hU (Fig. 9), that indicates the existence of a combination of micro-mesoporosity, confirmed by the pore size distribution (Fig. 10). Sinhamahapatra et al. [31, 32] observed similar behavior in the isotherms of adsorption-desorption of N2 developed by sulfated and borated zirconia samples, showing average pore size distributions located in the mesoporous region, but on the border with the microporous region (2–2.7 nm). Similarly, Osiglio et al. [18] obtained borated zirconia catalysts with pores in a like range.

Table 1 Textural and acidic properties of the synthesized materials
Full size table
Fig. 8
figure 8

N2 adsorption-desorption isotherm of the material ZrO2

Full size image
Fig. 9
figure 9

N2 adsorption-desorption isotherms of the modified materials: a ZB1hC, b ZB4hC, c ZB1hU, and d ZB4hU

Full size image
Fig. 10
figure 10

Pore size distributions of the modified materials: a ZB1hC, b ZB4hC, c ZB1hU, and d ZB4hU, estimated by the method of Barret–Joyner–Halenda from the desorption branch

Full size image

3.5 Scanning electron microscopy (SEM)

The images in Fig. 11 exhibit the morphology of zirconia samples, pure and boron-modified. ZrO2 showed spherical particles with heterogeneous size (up to 1 μ) and tendency to aggregate formation. The micrographs of the material ZB1hC also showed particles with similar morphology, conserving spherical tendency, and formation of agglomerates; the particle size is in a range of 0.5–1 μ and there are also small aggregates of particles on the surface of the consolidated aggregates. Finally, the images corresponding to the ultrasonically modified material (ZB1hU) showed similarity to the material ZB1hC, although agglomerates were losing their spheroidal shape becoming elongated particles (up to 1 μ); in addition, the growth of particles finer than those observed in the material ZB1hC was noted.

Fig. 11
figure 11

SEM images of the materials: a, b ZrO2, c, d ZB1hC and e, f ZB1hU

Full size image

3.6 Acidic properties

3.6.1 Potentiometric titration with n-butylamine

The surface acidity of the synthesized materials was determined by potentiometric titration with n-butylamine solution in acetonitrile. The basis of this technique establishes that the acid environment around the electrode membrane causes the potential difference; therefore, the measured potential is an indicator of the acidic properties of the dispersed solid particles. The criterion adopted for the interpretation of the results indicates that the maximum acid strength (MAS) of the sites is determined by the initial potential of the electrode (Ei). The acid strength of these sites is classified according to the following scale: E > 100 mV (very strong sites), 0 < E < 100 mV (strong sites), −100 < E < 0 mV (weak sites), and E< −100 mV (very weak sites) [33, 34]. In addition, the total number of acid sites corresponds to the value of meq amine/g of solid, which is determined when there are no more changes in the potential. Table 1 shows that all MAS values of the boron-modified samples were > 200 mV in the following order: ZB1hU > ZB1hC > ZB4hC > ZB4hU, while the ZrO2 only reached −28 mV. The total number of n-butylamine milliequivalents, required for neutralization, also confirmed a higher concentration of acid sites in these materials compared to its pure counterpart. Figure 12 presents the titration curves for each solid. ZrO2 exhibited only weak acid sites, while the impregnation of the catalytic precursors using boric acid generated strong and very strong acid sites in all samples. Neutralization profiles of the catalysts ZB1hC and ZB4hC had similar behavior for potential values > 50 mV (Fig. 12a). The curve of the sample ZB1hC showed stability at values close to 0 mV, so it indicates that this material presented only strong and very strong acid sites, while the material ZB4hC furthermore showed weak sites. On the other hand, Fig. 12b exhibits the neutralization profiles obtained for the samples ZB1hU and ZB4hU, which showed a similar tendency with a distribution of acid sites from very strong to weak. However, between these two materials it is evident that the catalyst ZB1hU encompasses a greater total concentration of acid sites, being even superior to its modified counterpart (ZB1hC).

Fig. 12
figure 12

Potentiometric titration curves for synthesized catalysts: Black Diamond Suit ZrO2, a Black Circle ZB1hC and Black Square ZB4hC, b Black Circle ZB1hU and Black Square ZB4hU. E is the electrode potential (mV)

Full size image

3.6.2 Ethanol decomposition

Alcohols decomposition is commonly used to study the acid/basic nature of the surface of metal oxides; the formation of aldehydes and ketones occurs via dehydrogenation on basic catalysts, while alcohol dehydration leads to olefins and ethers when acid sites are present [35]. In particular, the products of the ethanol dehydration are ethylene (intramolecular reaction) and diethyl ether (intermolecular reaction) [36–39]. The ethanol decomposition results, conducted at 300 °C, are presented in Fig. 13. In all cases, dehydration products were obtained (Fig. 14), confirming the predominantly acid nature of the synthesized solids. When the reaction was carried out using ZrO2, 2% conversion was achieved, but a remarkable increase was observed when the boron-modified materials were evaluated. For example, with the solids ZB1hC and ZB4hC, the ethanol conversion increased more than 30 times, oscillating in the range of 60–70%. Conversely, the materials ZB1hU and ZB4hU had a better performance in reaction, obtaining up to 100% of conversion with the catalyst that was ultrasonically modified for 1 h. The latter is consistent with the results of potentiometric titration, since the most active solid has a higher concentration of acid sites and superior MAS. However, it should be noted that the catalyst ZB4hU presented conversions between 70 and 80%, although it was slightly less acidic than the solids ZB1hC and ZB4hC. Therefore, it is possible to suggest a synergistic effect between boron incorporation and its dispersion on the support influenced by the impregnation method, thereby leading to more active catalysts compared to those modified by a conventional method. Moreover, as previously mentioned, all catalysts favored the alcohol decomposition via dehydration, due to an increment in acid strength of the materials by the presence of boron. Even ZrO2 was selective only to ethylene obtention; nevertheless, its weak acidity led to a very low conversion level. Furthermore, the distribution of the reaction products also indicates influence due to the impregnation method. The catalysts ZB1hC and ZB4hC were preferably selective to the formation of diethyl ether (60–70%), while the solids ZB1hU and ZB4hU increased the generation of ethylene. In fact, it can be noted that the products distribution is inverse for the catalysts impregnated during 1 h, suggesting that the surface of the solids favors either an elimination mechanism to produce the olefin (ZB1hU) or a substitution mechanism that leads to the formation of ether (ZB1hC).

Fig. 13
figure 13

Ethanol conversion at 300 °C using the synthesized catalysts: Black Diamond Suit ZrO2, Black Circle ZB1hC, White Circle ZB4hC, Black Square ZB1hU, and White Square ZB4hU

Full size image
Fig. 14
figure 14

Products distribution of ethanol dehydration at 300 °C using the synthesized catalysts: Black Circle ZB1hC, White Circle ZB4hC, Black Square ZB1hU, and White Square ZB4hU

Full size image

In Scheme 1 is shown a simplified representation of the ethanol dehydration. Elimination reactions (olefin formation) and substitution reactions (ether formation) are carried out simultaneously and compete with each other. These reactions are superficial and require different types of active sites: acid (electrophilic) and basic (nucleophilic). In general, the alcohol chemisorption on the acid site polarizes the C–O bond, converting the hydroxyl into a better leaving group. Moreover, the alcohol chemisorption on a basic site increases the oxygen nucleophilic character of this molecule. This may cause a nucleophilic displacement on the carbon atom of the adsorbed alcohol on the acid site, which results in the formation of the ether. In an alternative pathway, the adsorbed alcohol molecule on the acid site may lose a β proton to form the olefin [40]. In the literature, it is suggested that dehydration of ethanol to ethylene occurs through a concerted elimination mechanism, E2, involving an acid site and a basic site [41].

Scheme 1
scheme 1

Ethanol dehydration via elimination reaction (E2) and nucleophilic substitution reaction (NS). A and B represent acidic and basic sites, respectively

Full size image

A desired feature in solid acid catalysts is their reuse in reaction. The literature regarding the recyclability of boron-modified zirconia is scarce, and such studies involve the use of this catalyst in reactions carried out in liquid phase and at low temperatures (90–150 °C). In general, after at least three reaction cycles, only a slight decrease in the yield of the desired product was observed in the reactions of benzoylation of anisole [10], transesterification of β-ketoesters [15], condensation of Knoevenagel [16], and acetylation of 2-phenoxyethanol [18]. Therefore, in our work we decided to test the reusability of the catalyst ZB1hU, which showed a better performance in the ethanol decomposition. The results are presented in Fig. 15. In this case, the reaction was conducted at 250 °C, so it was also possible to observe changes in the products distribution. The conversion was stable around 28%, both with fresh catalyst and with the used material. On the other hand, the selectivity was mainly oriented toward the formation of diethyl ether, as a consequence of a lower reaction temperature compared to the results shown in Fig. 14. It is well known that low temperatures favor the ether obtention, and when the reaction temperature increases the production of the olefin also increases [36–39]. As shown in Fig. 15, the products distribution did not exhibit significant changes when the catalyst was reused. It should be noted that, after finishing the first test, the material was exposed to heat treatment at 270 °C in the presence of a nitrogen stream to evacuate moisture and any impurities.

Fig. 15
figure 15

Ethanol conversion at 250 °C and products distribution using the catalyst ZB1hU: fresh (open symbol) and used (closed symbol)

Full size image

4 Conclusions

By means of this research work, we can conclude that the method and impregnation time influence the acidic properties of boron-modified zirconia. Although the materials showed similarity in several of their properties, in general the boron presence delayed the crystallization process and the solids presented an amorphous tendency. In addition, the specific surface of the modified materials enhanced because the boron incorporation prevented sintering. Nevertheless, beyond the thermal, textural, and structural properties, the method and impregnation time significantly affected the catalytic performance of the solids. Even though all the boron-modified materials showed strong and very strong acid sites, the most active catalysts in the ethanol decomposition were those obtained by the ultrasonic method. Therefore, we can suppose that the catalysts synthesized by ultrasonic impregnation possess more accessible acid sites to the alcohol molecule, which preferably dehydrated to ethylene when the reaction was carried out at 300 °C.