Applied Catalysis A, General 652 (2023) 119040 2due to kinetics [25,27]. In light of what has preceded in the literature, Scheme 1 illustrates a generalized reaction pathway for thetert-butylation of phenol. The carbocation, formed from either TBA or isobutylene (IBE) [6], initially alkylates at the oxygen to form TBPE; the ether then undergoes trans- alkylation to produce 2-TBP, which further isomerizes to form 4-TBP [15,17,25,27]. Generation of water from the dehydration of TBA may foment stabilization of the charged alkylated intermediate and promote transalkylation of ether to carbon-alkylated product [25]. Higher tem- peratures and highly acidic conditions can promote dealkylation and side reactions resulting from IBE formation, such as oligomerization and cracking [16,28–30]. Numerous research groups have carried out this reaction in both vapor [16–18,21,31], liquid [15,26,29,32–36], and mixed [6] phases and in batch [15,26,29,32–36] and continuous flow [16–18,21,31] systems, though few have managed to performtert-butylation of phenol under conditions relevant to industrial applications, i.e. low tempera- ture and pressure and under solvent-free conditions.Tert-butylation of phenol via heterogeneous catalysis has seen applications of several mesoporous materials [18,21,37], rice husk ash derivatives [2,15,38], metal chlorides on silica gel [1], cation exchange resin Amberlyst®15 [6,32,37], supported heteropolyacids (HPAs) [19,20], ionic liquids [3, 29,35,39–41], and zeolites [5,30,36,37,42,43]. Metal oxides have been reported in literature for decades to exhibit enhanced acidic properties and high activity for an array of reactions [44]. Sulfated metal oxides– tin oxide, titania, and zirconia in particular–are known to display enhanced acidic properties and thus hold great potential for alkylation reactions. The latent acidity of these materials can differ significantly with varying synthesis procedures and reaction conditions, thereby substantially affecting the inherent activities of these catalysts. As such, comparison of the intrinsic reactivities of metal oxides and sulfated metal oxides across variant studies in the literature is an arduous task. The preeminent objective of this work was thus to, firstly, characterize and, secondly, investigate the efficacy of such catalysts at solvent-free, low temperature and pressure conditions. Promising reactivity was noted in our preliminary screening [33] of a series of sulfated metal and mixed metal oxides (SO4 2-/SnO2, SO4 2-/ZrO2, and SO4 2-/SnO2/ZrO2) that set the foundation of the present study. Given that the effect of acidity on catalytic behavior remains elusive [6,30,34,42,43,45], efforts have been placed to investigate potential acidity-activity relationships for this reaction. 2.Experimental 2.1.Materials Titanium butoxide (TBOT) was purchased from Sigma-Aldrich (re- agent grade, 97%) as the precursor for TiO2and SO4 2-/TiO2. Synthesis of ZrO2and SO4 2-/ZrO2was carried out using Zirconyl chloride octahydrate (ZrOCl2•8 H2O, Sigma-Aldrich, reagent grade, 98%). Tin (IV) chloride pentahydrate (SnCl4•8 H2O, Sigma-Aldrich, 98%) was used as the precursor for SnO2and SO4 2-/SnO2. Aqueous ammonia solution (Supelco, 28–30%) was utilized in the synthesis of both zirconia and tin oxide. Sulfuric acid solution (Fluka, 5 M H2SO4) was used in the sulfation of the metal oxide materials. Pyridine (Fisher chemical, certified ACS,≥99%) and 2,6-dimethylpyridine (Sigma-Aldrich,≥99%) were used for the temperature-programmed desorption (TPD) experiments. All chemicals were used as received from the supplier without further purification. 2.2.Synthesis of the supports Titania was prepared via a sol gel method reported elsewhere [46]. In a typical synthesis, TBOT was precipitated in ethanol (95%) and deionizedwaterina7:20:10ratio,respectively,byvolumewith vigorous mixing for 1 h. The resulting precipitant was then left to age for 6 h, filtered under vacuum and washed with deionized water, and then left to dry overnight at 110◦C. Thereafter, the dried powder was calcined at 500◦C for 6 h to produce titania (TiO2). A 0.1 M solution of either Zirconyl chloride octahydrate or Tin (IV) chloride pentahydrate in deionized water was prepared. Precipitation was then carried out by dropwise addition of 28–30% aqueous ammonia solution under vigorous stirring until reaching pH 8 [47]. The solution was then left to age for 1 h, filtered under vacuum and washed with deionized water, and then left to dry overnight at 110◦C. Thereafter, the dried powder was calcined at 500◦C for 6 h to produce either zirconia (ZrO2) or tin oxide (SnO2). 2.3.Synthesis of sulfated metal oxide catalysts To produce the sulfated catalysts, after vacuum filtration and drying overnight, dried but uncalcined samples were suspended for 1 h under vigorous stirring in a 0.1 M, 0.5 M or 1.0 M H2SO4solution, such that 15 mL of H2SO4was utilized per gram of powder. The resulting sample was then filtered under vacuum with deionized water washing and thereafter dried overnight at 110◦C. The powder was then calcined at Scheme 1.A generalized reaction network for thetert-butylation of phenol. The oligomerization to form C8 and C12, noted in red, was not observed under the reaction conditions in this study. A. Zuber and G. Tsilomelekis
Applied Catalysis A, General 652 (2023) 119040 3500◦C for 6 h to produce either sulfated titania (SO4 2-/TiO2), sulfated zirconia (SO4 2-/ZrO2), or sulfated tin oxide catalyst (SO4 2-/SnO2). 2.4.Physicochemical characterization 2.4.1.X-ray diffraction X-ray diffraction (XRD) patterns were collected with a PANalytical Philips XPERT powder diffractometer to determine crystallinity and phase composition changes. The XRD instrument was equipped with a CuKαsource at 40 kV and 40 mA and angular incidence 2θbetween 10◦ and90◦with0.03◦stepand2.00 s/step.Phasecompositionwas analyzed by whole pattern fitting (WPF) refinement 2-phase analysis with relative error R% targeted below 16%. Silicon was used as an external standard reference to determine possible peak shifts. 2.4.2.DRIFTS DiffusereflectanceinfraredFouriertransformspectroscopy (DRIFTS) was employed for in situ characterization of the structure of anchored species by utilization of a Thermo Scientific Nicolet™iS50 FTIR Spectrometer equipped with a Harrick Scientific Praying Mantis diffusereflectanceaccessory.Sampleswereloadedintothehigh- temperature reaction chamber equipped with a Praying Mantis™ dome with two ZnSe windows and one glass observation window. Temperature was ramped at 5◦C/min in the presence of air flow (40 sccm), with spectra taken at 25◦C and then in 100◦C intervals between 100 and 500◦C; spectra were likewise taken upon cooling in the same intervals. Spectra presented herein are the average of 64 scans between 720 and 4000 cm-1with a resolution of 8.0 cm-1and optical velocity of 0.4747 using a DTGS KBr detector. To establish a background, KBr powder (Alfa Aesar, FTIR grade) was loaded into the reaction cell and heated from 25◦to 100◦C at 2◦C/min in the presence of inert flow (either nitrogen or argon) (20 sccm); after dwelling at 100◦C for 30 min, the heating was shut off, and a background spectrum was taken at 25◦C. 2.4.3.Pyridine adsorption via DRIFTS Pyridine adsorption via DRIFTS was carried out with the same experimental setup. Samples were loaded into the Harrick cell with relevant accessories. The chamber was purged with inert (40 sccm of Ar) for 1 h, and then pyridine was bubbled under inert flow (40 sccm of Ar) for 1 h at 25◦C. The chamber was then flushed with inert (40 sccm of Ar) for 1 h, and spectra were taken at 25◦C. Pyridine TPD-DRIFTS was carried out by heating the reaction cell from 25◦to 750◦C at 5◦C/min with 40 sccm Ar flow; in addition to the spectrum taken at 25◦C, spectra were taken every 50◦C from 50◦to 750◦C. 2.4.4.Pyridine and 2,6-dimethylpyridine TPD Temperature-programmeddesorption(TPD)wascarriedoutin home-made U-shaped quartz tube (Fig. S1), which was heated by an insulated furnace (Carbolite Gero, Serial No. 21–702721, MTF 12/38/ 250 110–120 V 1PH), with the reactor outlet analyzed by an online mass spectrometer (Cirrus™3-XD Atmospheric Gas Analyzer, Quadrupole MS). In a typical measurement, approximately 0.05 g of sample was loaded into the quartz tube, which was subsequently purged with inert gas (Ar, 30 sccm). Afterwards, adsorbate (pyridine or 2,6-dimethylpyr- idine)wasbubbledintothequartztubeat25◦Cuntilreaching maximum capacity of adsorbate, as monitored by MS. The system was then purged once more with inert to remove physisorbed species, as monitored by MS. Thereafter, desorption was monitored while heating the chamber at 5◦C/min from 25◦to 1000◦C under inert flow (Ar, 30 sccm). 2.4.5.BET BET surface area measurements were conducted with a Micro- meritics TriStar 3000 system. Typically, around 30 mg of catalyst was loaded into a BET tube and degassed at 150◦C for 3 h prior to BET analysis in order to remove chemisorbed water from the sample surface. 79 points were collected within 0.01–1.0 P/P0with a 0.02 increment. Silica-alumina (Micromeritics, 004–16821–00, 99.7–99.9% aluminum oxide, 0.1–0.2% silicon dioxide, 0.1% ferric oxide) was used as the standard. 2.5.Phenol alkylation with tert-Butyl Alcohol 15 mL pressure vessels from Ace Glass Incorporated (Product # 8648–164), equipped with a thermowell to accommodate a thermo- couple for precise control of reaction temperature, were used fortert- butylation of phenol in batch mode. Reactant and liquid product con- centrations were determined by gas chromatography (Agilent Technol- ogies 7890B GC System) equipped with a flame ionization detector (FID) and capillary column (HP-5, 30 m×0.320 mm×0.25μm). Toluene (Sigma-Aldrich,≥99.5%) was used as a dilution solvent for the GC analysis. In a typical reaction, a 10 mL mixture oftert-butyl alcohol (TBA) (Sigma-Aldrich, ACS reagent,≥99.0%) and phenol (Sigma-Aldrich, ≥96.0%) in a 1:10 molar ratio was added to the reaction vessel with an appropriate stir bar. 0.313 g of catalyst, corresponding to 3 wt% catalyst loading (with respect to the total reactant mixture), was carefully added to the vessel, which was subsequently sealed tightly with the provided bushing and O-ring. A thermocouple was placed in the thermowell with silicon oil; the vessel was submerged into an oil bath on top of a heating plate. The reaction was then carried out with 1050 rpm mixing at the desired temperature. Time was recorded from the point at which the reaction mixture reached the set temperature. Initial samples were taken prior to addition of catalyst and analyzed by GC to confirm initial con- centration. After reaction, the mixture underwent centrifugation to separate the liquid products from solid catalyst particles; final samples were subsequently taken to confirm concentration and product distri- bution by GC. Product distribution in the liquid fraction was attributed to generation of alkylated products, i.e. TBPE (Sigma-Aldrich, 96%), 2- TBP(Sigma-Aldrich,99%),4-TBP(Sigma-Aldrich,99%),2,4-TBP (Sigma-Aldrich, 99%), 2,6-TBP (Sigma-Aldrich, 99%), and 2,4,6-TBP (Sigma-Aldrich, 98%). Generation of IBE was not analyzed via GC, but given the lack of observation of oligomerization products, the remaining carbon balance from the conversion of TBA was attributed to IBE. TBA conversion and product selectivity, respectively, were calculated as follows: XTBA=CTBA,converted CTBA,initial •100% Si=Ci CTBA,converted •100% Selectivity relative to the alkylated products was calculated in the following manner: Si=Ci CTBPE+C2TBP+C4TBP+C2,4TBP+C2,6TBP+C2,4,6TBP •100% The rate of the reaction for catalyst i was calculated as follows: Ratei=(molesconvertedTBA)i (time)(molesBrønstedacidsites)i 3.Results and discussion 3.1.Physicochemical characterization Fig. 1 displays the XRD patterns for the selected sulfated metal oxide catalysts and their supports. Analysis of the crystalline phases and relevant crystallite sizes is summarized in Table 1; amorphous domains were also observed. Smaller crystallite size with increasing sulfation was generally observed across all samples, consistent with other literature reports [47–50]. Sulfation has been reported to hinder particle sintering, A. Zuber and G. Tsilomelekis
Applied Catalysis A, General 652 (2023) 119040 4thus resulting in reduced particle size [33,50,51]. Tin oxide samples retained tetragonal morphology even upon sulfation [47,52], while all titania samples maintained an approximate 4:1 ratio of anatase to brookite [34,53]. Zirconia samples however showed a transition from primarilymonoclinicphaseinthesupporttomainlytetragonal conformation upon sulfation, as has been shown elsewhere [6,31,33,47, 54]. The relative conformity of unit cell sizes between the analyzed crystalline phases and those of standard crystalline materials is indica- tive of the lack of heteroatom substitution, denoting the presence of surface anchored sulfate species. Relevant measurements of the specific surface area, pore volume, and pore diameter are also summarized in Table 1. While all sulfated catalysts were synthesized under similar conditions, their properties varysignificantly. Sulfated metaloxides aregenerally reported to exhibit higher surface area than their non-sulfated analogues. It is believed that loading sulfate species hinders sintering of metal oxide particles, thereby reducing particle size and increasing surface area [33, 50]. 3.2.Spectroscopic characterization of select sulfated metal oxide catalysts Plausible molecular structures of bidentate (chelating and bridging), tridentate, and polymeric sulfate species that may be present on the surface of the various oxides [51,55–65] are presented in Fig. S2. It is expected that the electron-withdrawing effect of the sulfate group can enhance the Lewis acidity at the connected metal site [61,62,64], which can alter the adsorption properties of local sites and in turn affect catalyst activity; moreover, while the sulfate groups may not directly participate in the reaction, they may enhance Brønsted acidity. As sug- gested in the literature, several structures for acidic centers on sulfated metal oxides are shown in Fig. 2 [51,60,62–66]. The metal oxide support itself could exhibit both Brønsted and Lewis acidity due to surface hy- droxylgroups(M-OH)andthecoordinativelyunsaturatedcations (Mn+), respectively. In sulfated materials, depending on the sulfation content, different local sulfate structures may be present that can induce different Lewis/Brønsted acidities; besides, enhanced Lewis acid sites could foment Brønsted acidity by binding to a water molecule. Likewise, the neighboring oxygen could be protonated. Alternatively, protonation could occur on the sulfate group itself. DRIFTS is used as a mean to investigate the possible molecular configurations of sulfate species on the various oxides, which, in turn, could affect the acid centers that promotetert-butylation of phenol. Since the catalysts have not been dehydrated prior to reaction, spec- troscopic studies are compared herein for both ambient (25◦C) and dehydrated (500◦C) conditions. 3.2.1.In-situ DRIFTS: evolution of anchored species with temperature Figs. 3–5 show the in-situ DRIFTS results for the 1.0 M sulfated metal oxides, providing information on sulfate configurations. The complete set of DRIFTS data for the remaining materials are presented in the SI (Figs. S3-S11). Both physisorbed and chemisorbed water was observed at room temperature as is noted by the broad peak in the 3000–3800 cm- Fig. 1.XRD patterns for the sulfated metal oxide catalysts and their supports: (a) ZrO2, (b) TiO2, and (c) SnO2. Table 1 Analysis of BET and XRD results. The following nomenclature is used: A=anatase; B=brookite; M=monoclinic; T=tetragonal. Crystallite sizes are listed in order as per their respective phase. CatalystSulfation (M)Surface Area (m2/g)Pore volume (cc/g)Pore diameter (nm)Crystalline Phase (s)Crystallite Size (Å) ZrO2–63.80.1819.5M (95.9%) T (4.1%) 126 123 SO4 2-/ZrO20.1121.70.0813.4M (8.4%) T (91.6%) 35 109 SO4 2-/ZrO20.5130.10.0893.2M (0.9%) T (99.1%) >1000 101 SO4 2-/ZrO21.0117.50.0723.4M (1.2%) T (98.8%) 66 96 TiO2–100.40.1755.7A (78.2%) B (21.8%) 118 111 SO4 2-/TiO20.1100.10.1926.5A (81.3%) B (18.7%) 115 118 SO4 2-/TiO20.594.60.1776.3A (78.9%) B (21.1%) 114 95 SO4 2-/TiO21.099.90.1976.8A (81.1%) B (18.9%) 112 91 SnO2–48.30.0775.5T91 SO4 2-/SnO20.156.30.1056.5T91 SO4 2-/SnO20.556.10.0945.8T85 SO4 2-/SnO21.075.20.0603.7T77 A. Zuber and G. Tsilomelekis
Applied Catalysis A, General 652 (2023) 119040 51spectral range [59,67,68] as well as by the band at 1600 cm-1that corresponds to the bending mode of water [56,61,69–73]. Upon heating of these materials a peak above 3600 cm-1revealed that corresponds to free hydroxyl species [67], which remain even at high temperature. In thecaseofsulfatedzirconiaandtinoxides(Figs.3(a)and5(a), respectively), this peak shifts to higher frequency upon heating, whereas in the case of sulfated titania (Fig. 4(a)), this peak shifts to lower fre- quency upon heating and then back to higher frequency upon cooling. When comparing the sulfated catalysts with their unsulfated counter- parts, it can be seen that upon sulfation the free hydroxyl group appear at higher wavenumbers. As such, it can be said that there are hydroxyl species that are strongly bound to the catalytic support, while sulfation induces new OH groups that can account for the improved Brønsted acidity noted in these materials. Given that the frequency of the hy- droxyl stretching is observed at lower wavenumbers in sulfated tin oxide and sulfated zirconia, it is expected that these materials may show stronger Brønsted acidity as compared to sulfated titania. While similar trends are noted across the materials in the hydroxyl stretching region, significant differences could be seen in the sulfate region of the spectra. In the case of sulfated zirconia (Fig. 3(b)) at room temperature, the two broad peaks at approximately 1100 and 1250 cm-1 are ascribed to the asymmetric S-O-H stretch in hydrated bidentate species [51,56,59,73] and the symmetric O––S––O stretch in bidentate species [59], respectively. The latter shifts to 1295 cm-1upon dehy- dration. Some [33,74,75] have attributed the 1295 cm-1peak to poly- meric species (S2O7 2-), though there are no further peaks to directly Fig. 2.Potential acidic sites on sulfated metal oxides, where M represents a metal site. Note that anchored sulfate structures are not necessarily limited to bidentate configurations. Fig. 3.In situ DRIFTS for SO4 2-/ZrO2(1.0 M) in two regions of interest: (a) hydroxyl region and (b) the sulfate region. A. Zuber and G. Tsilomelekis
Applied Catalysis A, General 652 (2023) 119040 6support the presence of polymeric species. The peak at ~1400 cm-1has been associated with the S––O stretch of polymeric species [51,55], but the development of this peak upon heating is incongruent with that of the 1295 band, which decreases with temperature. Similar can be said regarding the 1295 cm-1band in the case of sulfated titania and tin oxides (Figs. 4(b) and 5(b), respectively). In the case of sulfated zirconia, two peaks at 1364 and 1383 cm-1 were seen to develop upon heating, evidence of the symmetric and asymmetric S––O stretches of bidentate di-oxo species, respectively [51, 74,76]. These two peaks shift to even higher wavenumbers due to dehydration, followed by the development of an asymmetric peak at approximately 1400 cm-1, indicative of the S––O stretch of tridentate mono-oxo sulfate species [51,74,76]. The asymmetric character of this band is indicative of the possible presence of both bidentate and Fig. 4.In situ DRIFTS for SO4 2-/TiO2(1.0 M) in two regions of interest: (a) hydroxyl region and (b) the sulfate region. Fig. 5.In situ DRIFTS for SO4 2-/SnO2(1.0 M) in two regions of interest: (a) hydroxyl region and (b) the sulfate region. A. Zuber and G. Tsilomelekis
Applied Catalysis A, General 652 (2023) 119040 7tridentate configurations, with the latter to be the dominant. Appear- ance of a peak at 1040 cm-1, corresponding to the asymmetric S-O stretchof tridentatespecies [51,55], wasfurther evidence ofthis development of mono-oxo tridentate; the peak at 1217 cm-1, however, pointed to the presence of some bidentate species [51,58], even at elevated temperature. An analogous trend was further observed in the case of sulfated titania (Fig. 4(b)), though to a lesser extent, indicating that mono-oxo tridentate species are more probable to develop in sulfated zirconia than in sulfated titania upon dehydration; no devel- opment of tridentate species was noted in sulfated tin oxide (Fig. 5(b)). The tendency of these materials to develop mono-oxo tridentate con- figurations is expected to affect acidity, and thereby their activity for tert-butylation of phenol. Contrary to sulfated zirconia, sulfated titania displayed both hy- drated bidentate and tridentate configurations at room temperature, with the former evidenced by a peak at 1215 cm-1[51,58] as well as an asymmetric S-OH stretch of bidentate species at 1129 cm-1[51]; a broad peak at 928 cm-1was indicative of tridentate species [51,55]. Sulfated tin oxide, likewise, exhibited tridentate configurations at room tem- perature via a peak at 956 cm-1[51,55], with bidentate moieties indi- cated by peaks at 1020, 1148, and 1285 cm-1, corresponding to the S-O stretch of protonated bidentate [51,56,59], the asymmetric S-O stretch of protonated bidentate [51,58], and the symmetric O––S––O stretch of bidentate [33,47,59,74,77], respectively. As discussed previously, the presence of protonated sulfate configurations on these materials at room temperature are expected to contribute to an apparent enhancement of Brønsted acidity. 3.2.2.DRIFTS: pyridine adsorption Pyridine adsorption via DRIFTS is a complementary technique to identify the nature of acidic sites on these materials and was thus carried out for the sulfated metal oxides and their supports (Fig. 6). For these experiments, the region between 1400 and 1700 wavenumbers provides information regarding theνCC(N)modes of chemisorbed pyridine. In titania and zirconia samples, the peaks observed at approximately 1445 and 1580 cm-1correspond to the combined C-C stretching and in-plane CH bending modes at a Lewis acid site [61,66,78–80]. The former mode was observed in tin oxide at 1451 cm-1, which may be attributed to the weaker Lewis acidity in tin oxide samples. The 1490 cm-1peak, shared across all materials, corresponds to adsorbed pyridine on both Brønsted and Lewis acid sites [61,66,80]. Peaks at ~1540 and 1640 cm-1appear in most of the materials, which is ascribed to the C-C stretching and in-plane CH and NH bending modes of the pyridinium ion, corre- sponding to Brønsted acid sites [34,61,66,78–80]. Weak shoulders were observed at approximately 1597 and 1610 cm-1in titania and zirconia samples, corresponding to hydrogen bonding with surface hydroxyl groups (i.e. Brønsted acidity) [78] and the stretching characteristic of pyridine bonded at a Lewis acid site, respectively [61,66,79,80]. In general, the relative intensities of these peaks (Fig. S12) indicate that the overall Lewis and Brønsted acidities in SO4 2-/ZrO2and SO4 2-/TiO2should be greater than that of SO4 2-/SnO2. This observation is also consistent with the TPD measurements presented in the next section. 3.3.Acid loading Quantitative analysis of acid sites was conducted via temperature- programmed desorption (TPD) of pyridine and 2,6-dimethylpyridine; as seen in the previous section, pyridine serves as a probe molecule for Brønsted and Lewis acid sites, while 2,6-dimethylpyridine selectively probes for only Brønsted acid sites [81]. Figs. S13-S15 of the supporting information show the full desorption profiles. All acidity measurements are summarized in Table 2. In reference to the distribution of acid sites, sulfated zirconia exhibits augmented overall (Lewis&Brønsted) acidity and total Brønsted acidity as compared to the other two sulfated metal oxides at any given level of sulfation. Generally, in accordance with the pyridine adsorption study via DRIFTS, the quantity of Lewis and Brønsted acid sites in sulfated titaniaandsulfatedzirconiaexceedsthatofsulfatedtinoxide. Furthermore, while sulfation was generally observed to enhance overall Brønsted acidity in both sulfated zirconia and sulfated tin oxide, sulfa- tion of titania resulted in less overall Brønsted acidity as compared to the bare titania support. Comparing the 2,6-dimethylpyridine TPD of titania and that of the varying sulfated titania samples (Fig. S14(b),(d),(f), and (h)), sulfation of the support was seen to afford a significant increase in moderate to strong Brønsted acidity. However, strong Brønsted acidity was observed to decrease with increasing sulfation, until there were few, if any, strong Brønsted acid sites at 1.0 M sulfation; similar trends have been reported elsewhere [66]. In the case of zirconia (Fig. S13(b),(d),(f), and (h)), sulfation resulted in significant augmentation of moderate Brønsted acidity between 250 and 450◦C, as well as enhanced strong Brønsted acidity. However, strong Brønsted acidity was observed to decrease with Fig. 6.DRIFTS spectra of (a) ZrO2, (b) TiO2, and (c) SnO2supports along with their respective sulfated counterparts after pyridine adsorption at 25◦C. A. Zuber and G. Tsilomelekis
Applied Catalysis A, General 652 (2023) 119040 8increasing sulfation, until two distinct peaks were seen at ~510 and 590◦C for sulfated zirconia at 1.0 M sulfation (Fig. S13(h)). For tin oxide and its sulfated conjugates (Fig. S15(b),(d),(f), and (h)), no development of moderate Brønsted acidity was noticed, but a strong Brønsted acid peak evolved between 520 and 560◦C; as before, this peak decreased in strength with increasing sulfation, though strong Brønsted acidity was still distinguishable at approximately 520◦C at 1.0 M sul- fation (Fig. S15(h)). The observation of strong Brønsted acidity in 1.0 M sulfated tin oxide and sulfated zirconia as compared to sulfated titania correlates well with the findings of the DRIFTS study. Next, we discuss the catalytic performance of all materials evaluated in this work in response of their acidic properties that highlighted in the preceding sections. 3.4.Catalytic evaluation of sulfated metal oxides for tert-butylation of phenol As has been reported in the literature, under the reaction conditions evaluated in this study,tert-butylation of phenol will compete with the dehydration oftert-butyl alcohol that produces isobutylene. The latter reaction can directly be influenced by the amount as well as strength of Brønsted acids, while the former, as mentioned above, can be affected by both Brønsted and Lewis acidity. In addition, phenol alkylation can proceed with isobutylene as the alkylating agent under the same reac- tion conditions, thus highlighting the complex nature of the reaction network. Consequently, controlling the selectivity of alkylated products remains a challenge. The catalytic evaluation of the bare and sulfated metal oxides was conducted to investigate the effect of sulfation on activity and relevant results are shown in Fig. 7. The temperature was set to 120◦C and comparison is made at fixed reaction time of 4 h. Significant differences in TBA conversion and product selectivity were noted between SO4 2-/ ZrO2(Fig. 7a), SO4 2-/TiO2(Fig. 7b) and SO4 2-/SnO2(Fig. 7c). In the case of zirconia samples, upon sulfation at 0.1 M and 0.5 M, a 30% increase in the conversion of TBA was observed; a gradual improvement on the total yield of mono-alkylated products was also observed at the expense of the TBPE. Further increase of sulfation (1.0 M) resulted in a small increase in conversion as compared to the bare support. The observed behavior in the ZrO2samples follows the overall trend of the total Brønsted acid concentration reported in Table 2. However, it should be mentioned that the dominant crystalline phase in bare ZrO2was the monoclinic phase while all of its sulfated counterparts were principally in the tetragonal phase, and as such, catalytic changes due to crystallinity cannot be fully excluded. Contrary to the ZrO2samples, the effect of sulfation on the catalytic performance was highly pronounced in sulfated titania as compared to bare titania (Fig. 7b). The conversion of TBA on the TiO2support was 30%, producing only the ether product. Although the total acidity of TiO2was found to be similar to some of the sulfated ZrO2materials, the strength of Brønsted acid sites in bare TiO2is very weak, as was remarked by the desorption peak between 150 and 250◦C shown in Fig. S14(b). Upon sulfation, an increase to 65% conversion was observed followed by a significant increase of mono-alkylated products. It is important to highlight that although the overall conversion and product distribution of mono-alkylated products is similar among the SO4 2-/TiO2 samples, an increase ~20% in carbon yield was observed for the ma- terial with highest sulfation, i.e. 1.0 M. This increase was accompanied by curtailed TBPE production. Considering that the conversion of all SO4 2-/TiO2catalysts was similar, this increase in the carbon yield of mono-alkylated products may arise from direct alkylation of phenol by isobutylene (formed via TBA dehydration). The bare SnO2support showed high conversion (~70%) with good selectivity to mono-alkylated products. TBPE was observed with a selectivity of 10% over the alkylated products. This activity of the unsulfated SnO2may arise from the large amount of surface Sn-OH sites, as shown earlier in the FTIR results. Upon sulfation with 0.1 M and 0.5 M, interestingly, the conversion dropped to 40% and the product selectivity shifted primarily to the ether product. Further increase in the sulfation to 1.0 M resulted in 90% TBA conversion with almost exclusive formation of mono-alkylated products. Minor formation of 2,4-TBP was observed, which is indicative of excessive alkylation. The highly sulfated SnO2catalystshowedthehighestcarbonyieldtomono-alkylated products. The changes in catalytic behavior between sulfated SnO2 samples can partially be attributed to the differences between the observed strength of acid sites (see Fig. S15). It was observed that with low to moderate sulfation, larger amount of stronger acid sites was Table 2 Acid site concentrations for each sulfated material and their respective supports. CatalystSulfation (M) Total Acid Sites (mmol/g) Brønsted Acid Sites (mmol/ g) Lewis Acid Sites (mmol/g) L:B Ratio ZrO2–0.2040.1190.0850.714 SO4 2-/ ZrO2 0.10.5920.3050.2870.941 SO4 2-/ ZrO2 0.50.7200.3490.3711.06 SO4 2-/ ZrO2 1.00.4320.2330.1990.854 TiO2–0.4400.2040.2361.16 SO4 2-/ TiO2 0.10.5370.2180.3191.46 SO4 2-/ TiO2 0.50.4830.1920.2911.52 SO4 2-/ TiO2 1.00.3610.1390.2221.60 SnO2–0.2840.1010.1831.81 SO4 2-/ SnO2 0.10.1650.1090.0560.514 SO4 2-/ SnO2 0.50.2010.1240.0770.621 SO4 2-/ SnO2 1.00.2390.1580.0810.513 Fig. 7.Initial screening of metal oxide and sulfated metal oxide catalysts fortert-butylation of phenol with product selectivity relative to only alkylated products for (a) SO4 2-/ZrO2, (b) SO4 2-/TiO2and (c) SO4 2-/SnO2. Reaction conditions: t=4 h, T=120◦C, catalyst loading=3.0 wt%, TBA:phenol molar ratio=1:10. A. Zuber and G. Tsilomelekis
Applied Catalysis A, General 652 (2023) 119040 9found by comparing the TPD area between 500 and 650◦C. It has been previously shown that large amount of strong Brønsted acid may cause dealkylation, thus driving the reaction to the opposite direction. How- ever, by comparing the FTIR spectra of the sulfated SnO2samples in Fig.5,S10,andS11,acleartransitionisobservedbetweenthe 1357 cm1and 1388 cm1peaks (stretching vibration of S––O) from low (0.1 M) to high (1 M) sulfation, indicative of changes in the distri- bution of bidentate and tridentate species, thereby influencing the na- ture of acidic centers. Such changes in the local structure of active sites can affect catalytic reactivity by altering the distribution of Lewis and Brønsted acids as well as the adsorption and desorption behavior of the reactants. It is, however, also important to mention that TBA does not only convert to alkylated products. A mole balance on the consumed TBA reveals that a significant portion of overall product selectivity is not accounted for by alkylated products alone. This is attributed to the production of undesired IBE gas from the dehydration of TBA, as illus- trated in Fig. S16; high yields of IBE are expected due to thermodynamic equilibrium from the dehydration of TBA [82]. Althoughtert-butylation of phenol with TBA is catalyzed by Brønsted acidity, when comparing the TPD data with the reactivity trends above, overall Brønsted acidity alone does not fully justify activity. In addition, acid site density, as normalized by the surface area, cannot account for the enhanced reactivity observed in some sulfated materials. The nature as well as distribution of Lewis and Brønsted acidity has however been reported to affect catalytic performance [6,34,66]. Fig. 8 illustrates the rate of TBA. consumption normalized by the total Brønsted acidity (from TPD measurements) as a function of the Lewis-to-Brønsted acid (L/B) ratio. It was found that an initial increase in the L/B ratio hinders alkylation at Brønsted acid sites. However, further increase of Lewis acid sites relative to overall Brønsted acidity results in an increase in activity. To the best of our knowledge, this type of inverse volcano reactivity trend has not been reported in the past for this reaction. Considering the excess of phenol under the selected reaction conditions (solvent-free), one can assume that the catalytic surface will be saturated by phenol. Phenol has been previously shown to be adsorbed on both Brønsted and Lewis acids [83–85]. In addition, it has been suggested that Fe3+Lewis acid centers may strengthen adjacent Brønsted acid sites [86]. Characteristicdesorptiontemperaturesforeachcatalystwere determined by temperature-programmed desorption of pyridine via DRIFTS (Figs. S17-S19). The molecular structure of pyridine is similar to phenol, and thus its adsorption behavior is expected to be akin to that of phenol [87–89]. Interpolation of the behavior of the band at approxi- mately 1450 wavenumbers (corresponding to Lewis acidity) resulted in an approximate desorption temperature representative for each mate- rial. Thereafter, it was observed that correlation of pyridine desorption temperature with the L/B ratio produced a volcano trend (Fig. 8(b)). ThishighlightsthatoptimalpyridineadsorptionoccursatL/Bof approximately 1.2, whereas pyridine adsorption is weaker in catalysts at low and high L/B ratios. In general, phenol adsorption can occur via three types of bonding: (1) hydrogen bonding at a Brønsted site, (2) interaction between the phenolic hydroxyl and a Lewis site, and (3) interaction between the catalyst and the aromatic ring electrons. In consideration of these re- sults, it may be hypothesized that enhanced phenol adsorption occurs at L/B of ~1.2 by way of these modes of adsorption, whereas some of these interactions may not be possible at high and low L/B; the strength of these adsorption modes may also be enhanced for L/B close to 1. Regardless, given that the trend in Fig. 8(b) is inverse of that in Fig. 8(a), Fig. 8.(a) Reaction rate (normalized by Brønsted acidity) as a function of the Lewis-to-Brønsted acid site ratio in each material. Rate=(total converted TBA) / [ (total Brønsted acidity) (time)]. (b) Pyridine desorption temperature (as determined by interpolation from pyridine TPD-DRIFTS) as a function of Lewis-to-Brønsted acid site ratio in each material. Reaction rate as a function of pyridine desorption temperature in (c) tin oxide materials, (d) titania materials and, (e) zirco- nia materials. A. Zuber and G. Tsilomelekis
Applied Catalysis A, General 652 (2023) 119040 10it can be proposed that there is an inherent relationship between reac- tionrateanddesorptiontemperature(Fig.8(c)-(e)).Theevident monotonic decrease of reaction rate with increasing desorption tem- perature implies that weak phenol adsorption is necessary to achieve high reaction rates for the tert-butylation of phenol under these condi- tions. These analyses also suggest that Lewis acid sites have a significant role in phenol adsorption, thus resolving the apparent correlation be- tween reaction rate and L/B ratio. Of note, Zhang et al. [16] observed a positive correlation between pyridine desorption temperature in highly acidic zeolitic materials and phenol conversion in thetert-butylation of phenol (though under significantly variant reaction conditions to those of this study) to highlight that strong acid sites are necessary to achieve high phenol conversion and 2,4-TBP selectivity. We believe that the observed trends in Fig. 8 highlight a potential synergistic effect of a) phenol adsorption on Lewis acids sites with b) strength of Brønsted acid centers. However, to fully disentangle and quantitate such effects requires future research endeavors that aim to couple thorough kinetic analysis with adsorption studies via spectro- scopic techniques. Such efforts can unravel possible competing reaction pathways and ultimately shed light on the underlying mechanism of phenol alkylation with TBA. 4.Conclusion Several solid superacid metal oxide catalysts were used in the solvent-freetert-butylation of phenol to determine their effectiveness in industrially relevant conditions. Thetert-butylation of phenol at 120◦C was carried out with these materials at varying levels of sulfation, with significant production of the mono-alkylated isomers (2- and 4-TBP). Amongst the catalysts used herein, sulfated TiO2and SnO2exhibited enhanced reactivity. Spectroscopic analysis of the sulfate species bound on the support surface revealed that both tridentate and protonated bidentate configurations may be anchored while their relative distri- bution is support specific. This distribution in the molecular structure of the surface species is hypothesized that can alter the distribution of Lewis and Brønsted acid centers. Analysis of the catalytic performance data along with results from the TPD of pyridine and 2,6-dimethylpyri- dine showed a correlation between overall activity and the relative ratio of Lewis and Brønsted acids. CRediT authorship contribution statement Adam Zuber:Conceptualization, Methodology, Experimentation, Validation, Formal analysis, Investigation, Writing–original draft, Writing–review&editing, Project administration.George Tsilome- lekis:Conceptualization, Resources, Writing–review&editing, Su- pervision, Project administration, Funding acquisition. Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Data Availability Data will be made available on request. Acknowledgements This work was supported by funding provided by Rutgers, The State University of New Jersey and the American Chemical Society Petroleum Research Fund (Doctoral New Investigator Award 58542-DNI5). The authors would further like to thank Aditya Khandare for contributing to development of the methodology for gas chromatography and initial kinetic screening, as well as Thu Nguyen and Chenfeng Huang for aiding in design and construct of the TPD system. Recognition is also due to Dr. Tewodros Asefa and Maricely Ramírez-Hern ́andez for providing the BET measurements. Appendix A.Supporting information Supplementary data associated with this article can be found in the online version at doi:10.1016/j.apcata.2023.119040. References [1]F.M. Moghaddam, et al., A new eco-friendly and efficient mesoporous solid acid catalyst for the alkylation of phenols and naphthols under microwave irradiation and solvent-free conditions, Sci. Iran. Trans. C., Chem., Chem. Eng. 16 (2) (2009) 81. [2]F. Adam, C.-W. Kueh, Phenyl-amino sulfonic solid acid-MCM-41 complex: a highly active and selective catalyst for the synthesis of mono-alkylated products in the solvent free tert-butylation of phenol, J. Taiwan Inst. Chem. Eng. 45 (2) (2014) 713–723. [3]J. Gui, et al., Selective alkylation of phenol with tert-butyl alcohol catalyzed by Br ̈onsted acidic imidazolium salts, J. Mol. Catal. A, Chem. 225 (1) (2005) 27–31. [4]S. Wu, et al., Synthesis, characterization, and catalytic performance of mesoporous Al-SBA-15 for tert-butylation of phenol, Chin. J. Catal. 27 (1) (2006) 9–14. [5]V. Krishnan, K. Ojha, N.C. Pradhan, Alkylation of phenol with tertiary butyl alcohol over zeolites, Org. Process Res. Dev. 6 (2) (2002) 132–137. [6]E. Modrogan, M. Valkenberg, W. Hoelderich, Phenol alkylation with isobutene— influence of heterogeneous Lewis and/or Brønsted acid sites, J. Catal. 261 (2) (2009) 177–187. [7]A. Sakthivel, S.K. Badamali, P. Selvam, para-Selective t-butylation of phenol over mesoporous H-AlMCM-41, Microporous Mesoporous Mater. 39 (3) (2000) 457–463. [8]F.A. Carey (ed.). Organic chemistry, seventh ed., McGraw-Hill, New York, 2008. [9]Alkylation process. 1963, Google Patents. [10]Ecke, G.G. and A.J. Kolka,Phenol alkylation process. 1958, Google Patents. [11]Kyrides, L.P.,Alkyl phenols and method of manufacturing the same. 1940, Google Patents. [12]Laufer, R.J. and M.D. Kulik,Ortho-tertiary butylation of phenol. 1968, Google Patents. [13]Perkins, R.P.,Method of making tertiary alkyl phenols. 1934, Google Patents. [14]Van Winkle, J.L.,Alkylation process. 1965, Google Patents. [15]F. Adam, K.M. Hello, T.H. Ali, Solvent free liquid-phase alkylation of phenol over solid sulfanilic acid catalyst, Appl. Catal. A Gen. 399 (1–2) (2011) 42–49. [16]K. Zhang, et al., Alkylation of phenol with tert-butyl alcohol catalyzed by large pore zeolites, Appl. Catal. A, Gen. 207 (1) (2001) 183–190. [17]S. Subramanian, et al., Para-selective butylation of phenol over silicoaluminophosphate molecular sieve SAPO-11 catalyst, Appl. Catal. A, Gen. 159 (1) (1997) 229–240. [18]S.K. Badamali, A. Sakthivel, P. Selvam, Tertiary butylation of phenol over mesoporous H–FeMCM-41, Catal. Lett. 65 (1) (2000) 153–157. [19]N. Bhatt, et al., Supported 12-tungstophosphoricacid: an efficient and selective solid acid catalyst for tert-butylation of phenol and cresols, Catal. Commun. 9 (7) (2008) 1545–1550. [20]N. Bhatt, A. Patel, Supported 12-tungstophosphoricacid: a recoverable solid acid catalyst for liquid phase Friedel–Crafts alkylation of phenol, J. Taiwan Inst. Chem. Eng. 42 (2) (2011) 356–362. [21]S.E. Dapurkar, P. Selvam, Mesoporous H-GaMCM-48: a remarkable solid acid catalyst for tertiary butylation of phenol, J. Catal. 224 (1) (2004) 178–186. [22]R.H.H.K. Abdol, Zirconium phosphate nanoparticles as a remarkable solid acid catalyst for selective solvent-free alkylation of phenol, Chin. J. Catal. 35 (7) (2014) 1136–1147. [23]Q. Ma, et al., Alkylation of phenol: a mechanistic view. the journal of physical chemistry. a, molecules, spectroscopy, kinetics, Environ., Gen. Theory 110 (6) (2006) 2246–2252. [24]X. Nie, et al., Reaction mechanism of tert-butylation of phenol with tert-butyl alcohol over H-βzeolite: An ONIOM study, Catal. Today 165 (1) (2011) 120–128. [25]T. Patra, F. Parveen, S. Upadhyayula, Mechanistic insights into solvent induced alkylation of p-cresol with tert-butyl alcohol using Br ̈onsted acidic ionic liquids, Mol. Catal. 433 (2017) 175–184. [26]A.B. Shinde, N.B. Shrigadi, S.D. Samant, tert-Butylation of phenols using tert-butyl alcohol in the presence of FeCl3-modified montmorillonite K10, Appl. Catal. A: Gen. 276 (1) (2004) 5–8. [27]X. Nie, et al., Brønsted acid-catalyzed tert-butylation of phenol, o-cresol and catechol: A comparative computational study, J. Mol. Catal. A: Chem. 332 (1) (2010) 145–151. [28]F. Jamil, et al., Evaluation of synthesized green carbon catalyst from waste date pits for tertiary butylation of phenol, Arab. J. Chem. 13 (1) (2020) 298–307. [29]T. Patra, S. Ahamad, S. Upadhyayula, Highly efficient alkylation of phenol with tert-butyl alcohol using environmentally benign Bronsted acidic ionic liquids, Appl. Catal. A: Gen. 506 (2015) 228–236. [30]R. Anand, et al., Tertiary butylation of phenol over HY and dealuminated HY zeolites, J. Mol. Catal. A Chem. 193 (1) (2003) 251–257. [31]A. Sakthivel, N. Saritha, P. Selvam, Vapour phase tertiary butylation of phenol over sulfated zirconia catalyst, Catal. Lett. 72 (3) (2001) 225–228. A. Zuber and G. Tsilomelekis