Vapour-Phase Selective O-Methylation of Catechol with Methanol Over Metal Phosphate Catalysts

catalysts

Article Vapour-Phase Selective O-Methylation of Catechol with Methanol over Metal Phosphate Catalysts

Haifang Mao 1, Xiaolei Li 1, Fen Xu 1, Zuobing Xiao 2, Wenxiang Zhang 3 and Tao Meng 1,*

1 School of Chemical and Environmental Engineering, Shanghai Institute of Technology, Shanghai 201418, China; [email protected] (H.M.); [email protected] (X.L.); [email protected] (F.X.) 2 School of Perfume and Aroma Technology, Shanghai Institute of Technology, Shanghai 201418, China; [email protected] 3 College of Chemistry, Jilin University, Changchun 130023, China; [email protected] * Correspondence: [email protected]; Tel.: +86-21-6087-7223

Abstract: Guaiacol (2-methoxyphenol), an important ﬁne chemical intermediate, is conventionally synthesized by liquid-phase processes with expensive, corrosive and toxic methylating agents such as dimethyl sulphate and dimethyl iodide. Recently, vapour-phase alkylation of catechol (1, 2- dihydroxybenzene) with methanol for the synthesis of guaiacol in the presence of heterogeneous catalysts has received more attention, as the route is economical and environmentally friendly. However, most of the investigated catalysts exhibited unsatisfactory catalytic performance for industrial applications. In this study, ﬁve metal phosphates M-P-O (M = La, Ce, Mg, Al, Zn) catalysts were synthesized and tested in the selective O-methylation of catechol with methanol. Among these catalysts, cerium phosphate (CP) showed the highest catalytic activity and guaiacol yield. Lanthanum

phosphate (LP) was the second most active, which was still obviously better than magnesium phosphate (MP) and aluminium phosphate (ALP). Zinc phosphate (ZP) was not active in the reaction.

Citation: Mao, H.; Li, X.; Xu, F.; Xiao, Relevant samples were characterized by X-ray diffraction (XRD), Fourier transform infrared (FT- Z.; Zhang, W.; Meng, T. Vapour-Phase IR), scanning electron microscopy (SEM), N2 adsorption-desorption, temperature programmed Selective O-Methylation of Catechol desorption of NH3 or CO2 (NH3-TPD, CO2-TPD). The suitable acid-base properties contribute to the with Methanol over Metal Phosphate superior catalytic performance of CP. Long-term stability and regeneration tests were also studied. Catalysts. Catalysts 2021, 11, 531. https://doi.org/10.3390/catal11050531 Keywords: metal phosphates; catechol; methanol; O-methylation; guaiacol

Academic Editor: Mickael Capron

Received: 25 March 2021 1. Introduction Accepted: 16 April 2021 Guaiacol (2-methoxyphenol) is a significant intermediate in fine chemical production, Published: 21 April 2021 especially for the synthesis of vanillin (4-hydroxy-3-methoxybenzaldehyde) [1,2]. Tradi- tionally, guaiacol is prepared in liquid-phase by using dimethyl sulphate or alkyl halides as Publisher’s Note: MDPI stays neutral methylating agents and sodium hydroxide as homogeneous catalyst, however the methy- with regard to jurisdictional claims in lating agents are expensive, corrosive, and toxic [3]. Currently, vapour-phase methylation published maps and institutional affil- iations. of catechol (1, 2-dihydroxybenzene) with non-toxic methanol or dimethyl carbonate (DMC) has attracted more attention because of its economical and environmentally friendly [4,5]. From the point of practical application, methanol is more suitable as methylating agent than DMC due to its lower cost. Various heterogeneous catalysts have been investigated in vapour-phase methylation Copyright: © 2021 by the authors. of catechol with methanol, including metal oxides [6,7], sulphates [2], metal Licensee MDPI, Basel, Switzerland. phosphates [5,8,9], and supported heteropoly acids [10]. In general, the catalytic perfor- This article is an open access article mance of these catalysts is greatly affected by the operation conditions, catalyst structures, distributed under the terms and conditions of the Creative Commons acid and/or base properties. However, a common problem is that most of the investigated Attribution (CC BY) license (https:// catalysts exhibited unsatisfactory catalytic performance. creativecommons.org/licenses/by/ Metal phosphates are a family of metal salts with high thermal stability and acid- 4.0/). base properties, which can be used as catalysts directly for oxidative dehydrogenation of

Catalysts 2021, 11, 531. https://doi.org/10.3390/catal11050531 https://www.mdpi.com/journal/catalysts Catalysts 2021, 11, x FOR PEER REVIEW 2 of 11

Catalysts 2021, 11, 531 2 of 12 Metal phosphates are a family of metal salts with high thermal stability and acid-base properties, which can be used as catalysts directly for oxidative dehydrogenation of isobutane (2-methylpropane) [11] or ethylbenzene [12], hydrolysis [13,14] or decomposition isobutane[15,16] of fluoride, (2-methylpropane) vapor-phase [11 O]-alkylation or ethylbenzene of phenol [12], [17], hydrolysis and selective [13,14] catalytic or decomposi- reduc- tiontion [of15 ,NO16] ofwith fluoride, NH3 [18]. vapor-phase Alternatively,O-alkylation metal phosphat of phenoles can [17], also and be selective used to catalytic prepare reductionsupported of catalysts NO with such NH3 as[18 Ru/CePO]. Alternatively,4 for the metal aerobic phosphates oxidation can of alcohols also be used [19], to Pd/AlPO prepare4 supportedfor hydrogenation catalysts suchof naphthalene as Ru/CePO [20],4 for metal the aerobic phosphate-supported oxidation of alcohols Au or [19 Pt], for Pd/AlPO CO oxi-4 fordation hydrogenation [21,22], metal of naphthalenephosphate-supported [20], metal RuO phosphate-supportedx catalysts for N2O Audecomposition or Pt for CO [23]. oxidation [21Zhang,22], metaland co-workers phosphate-supported [9] found that RuO thex catalysts acid-base for properties N2O decomposition of aluminium [23]. phos- phateZhang with andlower co-workers Ti content [9 were] found weaker, that the favouring acid-base guaiacol properties formation of aluminium and catalytic phosphate sta- withbility. lower In addition, Ti content the were catalytic weaker, performance favouring of guaiacol aluminophosphates formation and could catalytic be enhanced stability. Inby addition,adjusting the P/Al catalytic ratio [24] performance or coating ofmicrop aluminophosphatesorous aluminophosphate could be enhanced on mesoporous by adjusting SBA- P/Al15 [25], ratio further [24] confirming or coating microporousthat suitable acid-base aluminophosphate properties onwere mesoporous crucial to vapour-phase SBA-15 [25], furthermethylation confirming of catechol that suitable with methanol. acid-base However, properties Jafari were et crucial al. [5] tofound vapour-phase that catechol methy- con- lationversion of and catechol product with distribution methanol. However,depended Jafarihighly et on al. acidity [5] found of Ti that or catecholCs modified conversion lantha- andnum product phosphate. distribution Thus, the depended reaction mechanism highly on is acidity still controversial. of Ti or Cs modified lanthanum phosphate.Previous Thus, studies the reaction mainly mechanismfocused on isa stillparticular controversial. metal phosphate. The objective of the currentPrevious work studies is to mainly systematically focused onstudy a particular the catalytic metal performance phosphate. of The different objective metal of thephosphates current workM-P-O is (M to systematically= La, Ce, Mg, Al, study Zn) in the vapour-phase catalytic performance methylation of of different catechol metal with phosphates M-P-O (M = La, Ce, Mg, Al, Zn) in vapour-phase methylation of catechol with methanol. Among these catalysts, cerous phosphate (CP) is the most active one. Reasons methanol. Among these catalysts, cerous phosphate (CP) is the most active one. Reasons for the observations were elucidated through detailed characterization. for the observations were elucidated through detailed characterization.

2.2. ResultsResults 2.1.2.1. BasicBasic PhysicochemicalPhysicochemical PropertiesProperties ofof MetalMetal PhosphatesPhosphates SamplesSamples XRDXRD patternspatterns ofof thethe as-synthesizedas-synthesized productsproducts areare shownshown inin FigureFigure1 .1. According According to to the the literaturesliteratures [[26–29],26–29], thethe characteristiccharacteristic diffractiondiffraction peakspeaks are obtained for crystallized CP, LP, MPMP andand ZPZP samples.samples. ALPALP showshow aa broadbroad peakpeak betweenbetween 2020°◦ and and 3030°◦ correspondingcorresponding toto thethe characteristiccharacteristic ofof amorphous amorphous phase, phase, and and two two small small sharp sharp peaks peaks (around (around 20 ◦20°)) corresponding correspond- toing crystalline to crystalline aluminium aluminium phosphate phosphate [30 [30].]. FT-IR FT-IR test test was was carried carried out out to to further further identify identify thethe products.products.

ALP Intensity / a.u.

20 40 60 80 2 Theta / º FigureFigure 1.1. XRDXRD patternspatterns ofof differentdifferent metalmetal phosphatephosphate catalysts.catalysts.

FigureFigure2 2 shows shows the the FT-IR FT-IR spectra spectra of of different different M-P-O M-P-O samples. samples. As As shown shown for for all all metal metal − − phosphates,phosphates, thethe absorptionabsorption bandband atat 34303430 cmcm−11 andand 1628 1628 cm cm−1 are1 are ascribed ascribed to toO–H O–H stretching stretching vibration and O–H bending vibration of adsorbed water molecules, respectively [31,32]. The band at ca. 1046 cm−1 and 615 cm−1 are attributed to P–O asymmetric stretch vibration and O=P–O bending vibration, and the bands at ca. 538 cm−1 are due to O–P–O bending

CatalystsCatalysts 20212021,, 1111,, xx FORFOR PEERPEER REVIEWREVIEW 33 ofof 1111

Catalysts 2021, 11, 531 3 of 12 vibrationvibration andand O–HO–H bendingbending vibrationvibration ofof adsorbedadsorbed waterwater molecules,molecules, respectivelyrespectively [31,32].[31,32]. TheThe bandband atat ca.ca. 10461046 cmcm−−11 andand 615615 cmcm−−11 areare attributedattributed toto P–OP–O asymmetricasymmetric stretchstretch vibrationvibration andand O=P–OO=P–O bendingbending vibration,vibration, andand thethe bandsbands atat ca.ca. 538538 cmcm−−11 areare duedue toto O–P–OO–P–O bendingbending 3− vibration typical for PO 3− [32,33]. These observations conﬁrm the formation of metal vibrationvibration typicaltypical forfor POPO4443− [32,33].[32,33]. TheseThese observationsobservations confirmconfirm thethe formationformation ofof metalmetal phosphates,phosphates, thusthus supportingsupporting thethe XRDXRD results.results.

CPCP

LPLP

MPMP

ALPALP Absorbance / a.u. Absorbance / a.u. ZPZP

16281628 34303430 615615 538538 10461046

500500 1000 1000 1500 1500 2000 2000 2500 2500 3000 3000 3500 3500 4000 4000 -1-1 WavenumberWavenumber // cmcm FigureFigure 2. 2. FT-IRFT-IR spectra spectra of of different different metalmetal phosphatephosphate catalysts.catalysts.

TheThe morphologymorphology of ofof M-P-O M-P-OM-P-O samples samplessamples was waswas investigated inveinvestigatedstigated with withwith SEM SEMSEM analysis. analysis.analysis. As AsAs shown shownshown in inFigurein FiguresFigures3 and 33 Figureandand S1,S1, S1, metalmetal metal phosphatesphosphates phosphates havehave have irregularirregular irregular particlesparticles particles exceptexcept except LP.LP. LP. LPLP isisis mainlymainlymainly composedcomposed ofofof smallsmallsmall cubic cubiccubic column columncolumn with withwith a aa length lengthlength of ofof about aboutabout 5 µ 55m. μμm.m. The TheThe particle particleparticle size sizesize distribution distribu-distribu- tionoftion CP ofof and CPCP ALPandand ALPALP are mainly areare mainlymainly in range inin rangerange of 1–5 ofof µ1–51–5m, μ whileμm,m, whilewhile that that ofthat ZP ofof is ZPZP in isis range inin rangerange of 10–50 ofof 10–5010–50µm. μTheμm.m. averageTheThe averageaverage particles particlesparticles sizes of sizessizes MP of areof MPMP about areare 2 aboutaboutµm. The 22 μμ degreem.m. TheThe of degreedegree agglomeration ofof agglomerationagglomeration is increased isis increasedfromincreased CP to fromfrom LP, andCPCP toto further LP,LP, andand to MP.furtherfurther However, toto MP.MP. ALPHowever,However, and ZP ALPALP are andand highly ZPZP dispersed.areare highlyhighly dispersed.dispersed.

FigureFigureFigure 3.3. 3. SEMSEMSEM imagesimages images ofof of (( (aaa))) CP,CP, CP, (( bb)) LP,LP, ( (cc)) MP, MP, (((dd))) ALP,ALP,ALP, and andand ( (e(ee))) ZP. ZP.ZP.

Catalysts 2021, 11, x FOR PEER REVIEW 4 of 11

Table 1 summarizes the textual properties of M-P-O samples obtained by the N2 adsorption-desorption data. The BET surface areas of catalysts show the sequence of LP > CP > MP > ALP > ZP. As shown in Figure 3, an increase in particle size of ZP may lead to a significant decrease in the BET surface area. The pore volumes of LP and MP are much larger than the others, which is probably owing to secondary pores formed by obvious aggregation.

Table 1. Textural properties of different metal phosphate catalysts.

Catalysts 2021, 11, 531 Catalyst BET Surface Area (m2/g) Pore Volume4 of 12 (cm3/g) CP 68 0.09 LP 88 0.20 Table1 summarizesMP the textual properties of M-P-O64 samples obtained by the N0.352 adsorption-desorption data. The BET surface areas of catalysts show the sequence of LP > CP > MPALP > ALP > ZP. As shown in Figure3, an29 increase in particle size of ZP may0.11 lead to a signiﬁcantZP decrease in the BET surface area.7 The pore volumes of LP and MP0.02 are much larger than the others, which is probably owing to secondary pores formed by 2.2.obvious Catalytic aggregation. Performance TableThe 1. Textural catalytic properties activity of different and metal selectivity phosphate for catalysts. the selective O-methylation of catechol had

been investigatedCatalyst over the M-P-OBET Surface catalysts. Area (m As2/g) presented Pore in Volume Figure (cm 34a,/g) for the CP sample, an initial catechol conversion of 80.3% and falls to 65.3% after 11 h on stream. The catechol CP 68 0.09 conversion LPobserved for the LP catalyst 88 keeps constant at about 0.20 40.4% during 11 h on stream. However,MP the initial catechol 64 conversion on MP (10.5%) 0.35 and ALP (7.2%) are obvi- ALP 29 0.11 ous lower thanZP CP and LP, and deactivation 7 is rather fast. ZP 0.02shows almost no conversion of catechol, indicating it is inactive in selective O-methylation of catechol. Apparently, the catechol2.2. Catalytic conversion Performance on catalysts follows the sequence of CP > LP > MP > ALP > ZP. TheThe catalytic effects activityof time and on selectivity stream on for the the selectiveselectivityO-methylation of guaiacol of for catechol each had catalyst is shown inbeen Figure investigated 4b. As over shown, the M-P-O the initial catalysts. guaiacol As presented selectivity in Figure of4 CP,a, for LP, the CPMP, sample, and ALP are 92.6%, 88.4%,an initial 85.7%, catechol 97.4%, conversion respectively. of 80.3% and However, falls to 65.3% ZP after shows 11 h onno stream. selectivity. The catechol The guaiacol selec- conversion observed for the LP catalyst keeps constant at about 40.4% during 11 h on stream. tivityHowever, of CP, the initial LP, and catechol ALP conversion increased on slightly MP (10.5%) with and time. ALP (7.2%) After are 11 obvious h on stream, lower the guaiacol selectivitythan CP and of LP, ALP and deactivation and CP was is rather as high fast. ZPas shows99.3% almost and 97.8%, no conversion respectively. of catechol, In contrast, the guaiacolindicating selectivity it is inactive of in selectiveMP decreasedO-methylation with oftime. catechol. The Apparently,guaiacol selectivity the catechol follows the se- quenceconversion of onALP catalysts > CP follows> LP > theMP sequence > ZP. of CP > LP > MP > ALP > ZP.

100 CP LP MP ALP ZP a 90

30 Conversion of catechol / % of Conversion 20

0 34567891011 Time on stream / h

Figure 4. Cont.

CatalystsCatalysts 2021, 202111, x, 11 FOR, 531 PEER REVIEW 5 of 12 5 of 11

CP LP MP ALP ZP 100 b

30 Selectivity of guaiacol / % 20

0 34567891011 Time on stream / h FigureFigure 4.4.Effect Effect of of time time on stream on stream on catechol on catechol conversion conversion (a) and guaiacol (a) and selectivity guaiacol (b) over selectivity different (b) over dif- ferentmetal phosphatemetal phosphate catalysts. Reactioncatalysts. conditions: Reaction atmospheric conditions: pressure; atmospheric T = 270 ◦C; pressu WHSVre; = 0.4T = h −2701. °C; WHSV = 0.4 h−1. The effects of time on stream on the selectivity of guaiacol for each catalyst is shown in Figure4b. As shown, the initial guaiacol selectivity of CP, LP, MP, and ALP are 92.6%, 88.4%, 2.3.85.7%, TPD 97.4%, Characterization respectively. However, ZP shows no selectivity. The guaiacol selectivity of CP, LP,For and selective ALP increased O-methylation slightly with of time. catechol After 11 and h on some stream, other the guaiacol alkylation selectivity reactions, the cat- of ALP and CP was as high as 99.3% and 97.8%, respectively. In contrast, the guaiacol alyticselectivity activity of MP is decreasedcorrelated with to time.the acidity The guaiacol and basicity selectivity of follows catalysts the sequence[25,32,34]. of ALPThe > CP acid > LP >strength MP > ZP. and the amounts of acid sites on M-P-O catalysts were characterized by NH3-TPD (Figure 5, Table 2). As shown, M-P-O catalysts show two types of peaks 2.3. TPD Characterization below and above 300 °C, corresponding to NH3 desorbed from the weak and strong acid sites,For respectively selective O -methylation[35]. The temperature of catechol and corresponding some other alkylation to the reactions,desorption the peak of weak catalytic activity is correlated to the acidity and basicity of catalysts [25,32,34]. acid sitesThe acid follows strength the and sequence the amounts of ZP of < acidMP sites< CP on < M-P-OALP < catalystsLP, consistent were charac- with the sequence ofterized weak by acid NH 3strength.-TPD (Figure The5, area Table 2of). NH As shown,3 desorption M-P-O peak catalysts follows show the two sequence types of CP > LP ◦ >of ALP peaks >MP below > ZP, and aboveindicating 300 C, that corresponding the amount to NHof weak3 desorbed acid from sites the of weakCP is and the highest. For strongstrong acid sites,sites, respectively the desorption [35]. Thetemperature temperature follows corresponding the sequence to the desorptionof CP = ALP < ZP < MP, peak of weak acid sites follows the sequence of ZP < MP < CP < ALP < LP, consistent and the area of NH3 desorption peak follows the sequence of MP ≈ ZP > ALP > CP. LP with the sequence of weak acid strength. The area of NH3 desorption peak follows the havesequence no desorption of CP > LP > of ALP NH >MP3 above > ZP, indicating300 °C, indicating that the amount the lack of weak of strong acid sites acid of sites. CP is the highest. For strong acid sites, the desorption temperature follows the sequence of CP = ALP < ZP < MP, and the area of NH3 desorption peak follows the sequence of ≈ ◦ MP ZPCP > ALP > CP. LP have no desorption of NH3 above 300 C, indicating the lack of strong acid sites.

MP Intensity / a.u.

ALP

100 200 300 400 500 600 Temperature / ℃

Figure 5. NH3-TPD curves of different metal phosphate catalysts.

Catalysts 2021, 11, x FOR PEER REVIEW 5 of 11

CP LP MP ALP ZP 100 b

30 Selectivity of guaiacol / % 20

0 34567891011 Time on stream / h Figure 4. Effect of time on stream on catechol conversion (a) and guaiacol selectivity (b) over different metal phosphate catalysts. Reaction conditions: atmospheric pressure; T = 270 °C; WHSV = 0.4 h−1.

2.3. TPD Characterization For selective O-methylation of catechol and some other alkylation reactions, the catalytic activity is correlated to the acidity and basicity of catalysts [25,32,34]. The acid strength and the amounts of acid sites on M-P-O catalysts were characterized by NH3-TPD (Figure 5, Table 2). As shown, M-P-O catalysts show two types of peaks below and above 300 °C, corresponding to NH3 desorbed from the weak and strong acid sites, respectively [35]. The temperature corresponding to the desorption peak of weak acid sites follows the sequence of ZP < MP < CP < ALP < LP, consistent with the sequence of weak acid strength. The area of NH3 desorption peak follows the sequence of CP > LP > ALP >MP > ZP, indicating that the amount of weak acid sites of CP is the highest. For strong acid sites, the desorption temperature follows the sequence of CP = ALP < ZP < MP, Catalysts 2021, 11, 531 6 of 12 and the area of NH3 desorption peak follows the sequence of MP ≈ ZP > ALP > CP. LP have no desorption of NH3 above 300 °C, indicating the lack of strong acid sites.

MP Intensity / a.u.

ALP

100 200 300 400 500 600 Temperature / ℃

FigureFigure 5. 5.NH NH3-TPD3-TPD curves curves of different of different metal phosphate metal phosphate catalysts. catalysts.

Table 2. Textural properties of different metal phosphate catalysts.

◦ Catalyst Peak Temperature ( C) Intensity (a.u.) 215 391.24 CP 426 42.66 233 381.51 LP —— 199 128.91 MP 525 740.59 224 351.23 ALP 426 408.90 179 35.40 ZP 450 734.60

The basicity of the catalysts is evaluated by CO2-TPD (Figure6, Table3). The strength and the amounts of base sites were related to the temperature and area of CO2 desorption peak, respectively. Figure6 shows CP, LP and ALP have obvious peaks corresponding to CO2 desorption, while MP have very limited desorption of CO2. The CO2 desorption temperature of CP, MP and ALP are very close, and higher than that of LP, indicating LP has relatively weak basicity. On the other hand, ZP shows no CO2 desorption peak, indicating the absence of base sites on these catalysts. The amounts of base sites of these catalysts follow the sequence of MP < CP < LP < ALP (Table3). The interpretation of the TPD data will be presented in Section4. Catalysts 2021, 11, x FOR PEER REVIEW 6 of 11

Table 2. Textural properties of different metal phosphate catalysts.

Catalyst Peak Temperature (°C) Intensity (a.u.) 215 391.24 CP 426 42.66 233 381.51 LP — — 199 128.91 MP 525 740.59 224 351.23 ALP 426 408.90 179 35.40 ZP 450 734.60

The basicity of the catalysts is evaluated by CO2-TPD (Figure 6, Table 3). The strength and the amounts of base sites were related to the temperature and area of CO2 desorption peak, respectively. Figure 6 shows CP, LP and ALP have obvious peaks corresponding to CO2 desorption, while MP have very limited desorption of CO2. The CO2 desorption temperature of CP, MP and ALP are very close, and higher than that of LP, indicating LP has relatively weak basicity. On the other hand, ZP shows no CO2 desorption peak, indicating Catalysts 2021, 11, 531 the absence of base sites on these catalysts. The amounts of base sites of these catalysts7 of 12 follow the sequence of MP < CP < LP < ALP (Table 3). The interpretation of the TPD data will be presented in Section 4.

MP ×6 Intensity Intensity / a.u. ALP

50 100 150 200 250 300 350 400 Temperature / ℃

Figure 6. CO2-TPD-TPD curves curves of of different different metal phosphate catalysts.

Table 3. CO2-TPD results of different metal phosphate catalysts. Table 3. CO2-TPD results of different metal phosphate catalysts.

Catalyst DesorptionDesorption Temperature Temperature (°C) Intensity (a.u.) Catalyst Intensity (a.u.) CP (◦150C) 3.71 CPLP 15094 3.714.24 MPLP 94147 4.240.50 ALPMP 147154 0.505.46 ALPZP 154- 5.46 - ZP - -

2.4. Catalytic Stability and Catalyst Regeneration Figure7 shows the catalytic performance of the most active CP catalyst in two-cycle experiments. In the first cycle, the catechol conversion of CP decreased from 71.7% to 43.3% and gradually stabilized after 72 h on stream (Figure7a). Then the catalyst was regenerated, and its catechol conversion recovered from 43.3% to 66.5%. The catechol conversion of regenerated CP in the second cycle decreased from 66.5% to 37.8% with the time on stream, consistent with the trend seen in the first cycle. As shown in Figure7b, the initial guaiacol selectivity of the regenerated CP was slightly lower than that of fresh CP, but increased with the reaction time, and finally close to the guaiacol selectivity of fresh CP (97.6%). Catalysts 2021, 11, x FOR PEER REVIEW 7 of 11

2.4. Catalytic Stability and Catalyst Regeneration Figure 7 shows the catalytic performance of the most active CP catalyst in two-cycle experiments. In the first cycle, the catechol conversion of CP decreased from 71.7% to 43.3% and gradually stabilized after 72 h on stream (Figure 7a). Then the catalyst was regenerated, and its catechol conversion recovered from 43.3% to 66.5%. The catechol conversion of regenerated CP in the second cycle decreased from 66.5% to 37.8% with the time on stream, consistent with the trend seen in the first cycle. Catalysts 2021, 11, 531 As shown in Figure 7b, the initial guaiacol selectivity of the regenerated CP 8was of 12 slightly lower than that of fresh CP, but increased with the reaction time, and finally close to the guaiacol selectivity of fresh CP (97.6%).

100 a

40 Conversion of catechol / % catechol of Conversion

0 10 20 30 40 50 60 70 80 Time on stream / h 100 b

40 Selectivity of guaiacol / %

0 10 20 30 40 50 60 70 80 Time on stream / h FigureFigure 7. 7. CatecholCatechol conversion conversion (a) ( aand) and guaiacol guaiacol selectivity selectivity (b) (ofb) CP of CPcatalyst catalyst in cycle in cycle experiment, experiment, ■ (()) the the first ﬁrst cycle, cycle, ( (●•) the second cycle.

3.3. Discussion Discussion InIn this this work, work, metal phosphates werewere preparedprepared by by precipitation precipitation method. method. According According to toXRD XRD data data (Figure (Figure1) 1) and and FT-IR FT-IR spectra spectra (Figure (Figure2), 2), all all as-synthesized as-synthesized samples samples were were metal metal phosphatesphosphates in nature.nature. AmongAmong them, them, CP, CP, LP, LP, MP andMP ZPand were ZP inwere crystallized in crystallized phase, whereasphase, ALP was in amorphous phase. We found that the initial catalytic activities (conversion of catechol) of metal phosphates follow the sequence of CP > LP > MP > ALP > ZP (Figure4a), while the initial guaiacol selectivity follows the sequence of ALP > CP > LP > MP > ZP (Figure4b), thus resulting in the initial guaiacol yield follows the sequence of CP (74.4%) > LP (39.4%) > MP (9.0%) > ALP (6.4%) > ZP (0%). The selective O-methylation of catechol with methanol is commonly regarded as an acid-base catalytic reaction for the production of guaiacol [4,9,10,24,25]. Therefore, the different acid-base properties of metal phosphates may have an effect on their catalytic performance. From NH3-TPD data (Figure5), it can be seen that there are two kinds of acid sites in metal phosphates samples, i.e., weak and strong acid sites. Generally, coke deposition easily occurs on strong acid sites [4,9]. The amount of strong acid sites on Catalysts 2021, 11, 531 9 of 12

MP, ALP, ZP was much higher than CP and LP (Table2), thus coke rapidly formed and blocked acid sites of MP, ALP and ZP, further causing their low initial activities and quick deactivation in the reaction. Additionally, MP, ALP and ZP samples were black in colour after 11 h reaction, indicating the fast coke formation. The catalytic activity in O-methylation of catechol is correlated with the weak acid properties of metal phosphates [8,24]. The weaker strength and larger amount of the weak acid sites (Figure5) might be responsible for the higher catechol conversion of CP. However, due to the presence of small amount of strong acid sites on CP, the catechol conversion decreased by 15.0% after 11 h reaction. In contrast, the catechol conversion of LP kept constant because of the absence of strong acid sites. The weak base sites on the catalyst are beneficial for forming O-methylated products [2,6,24]. By comparing ALP, CP, and MP with similar base strength, it is noteworthy that the guaiacol selectivity is associated with the amount of base sites, i.e., the selectivity increases as the amount of base sites increases (Figure4b or Figure6). Though CP and LP possess comparable amounts of base sites, the guaiacol selectivity of CP is higher because of its suitable base strength (Figure1). ZP shows no guaiacol selectivity, probably due to its strong acid sites and the absence of weak basic sites. From the above results, it can be concluded that the high catalytic activity and selectivity of CP own to its suitable weak acid-base sites. Moreover, it further confirms the acid-base co-operation mechanism in selective O-methylation of catechol with methanol [2,9]. The most active CP is compared with other acid and/or basic catalysts. As shown in Figures S2–S4, SAPO-34 and HAP samples exhibits poor catalytic performance, which is related to their high amount of strong acid sites and the presence of strong basic sites ◦ (CO2 desorbed above 250 C). Similar to the strong acid sites, coke is easily formed on strong basic sites [9]. Besides, the blockage of SAPO-34 pores also results in the decrease of catalytic performance [4]. Therefore, irregular pores created by the agglomeration of CP and MP may be blocked by oligomeric products and coke (Figure3, Table1), this is probably another reason for deactivation of CP and MP. In the cycle experiment (Figure7), although the catalytic activity of the fresh and regenerated CP both decrease with the time on stream, most of the activity of the used CP can be recovered by regeneration, indicating that coke deposition is the main reason for the loss of catalytic activity of CP. It is noteworthy that the guaiacol selectivity of fresh and regenerated CP keep constant at 97.6% as the reaction proceeded, which was significantly higher than that in the literature [2,5,25]. High guaiacol selectivity facilitates the separation and purification of the products.

4. Materials and Methods 4.1. Materials The chemical reagents used included catechol, methanol, aqueous ammonia, phosphoric acid, La(NO3)3·6H2O (lanthanum nitrate hydrate), Ce(NO3)3·6H2O (cerium (III) nitrate hexahydrate), Mg(NO3)2 (magnesium nitrate), Al(NO3)3·9H2O (aluminum nitrate nonahydrate), and Zn(NO3)2·6H2O (zinc nitrate hexahydrate). All the above reagents of analytical grade were obtained from Adamas and used as received.

4.2. Sample Preparation Metal phosphates M-P-O (M = La, Ce, Mg, Al, Zn) catalysts were prepared by precipitation method. In a typical synthesis, 8.702 g La(NO3)3·6H2O was dissolved in 1000 mL deionized water, then 4.980 g phosphoric acid solution was added dropwise with vigorous stirring. The pH of the gel was adjusted to 8.0 by adding concentrated aqueous ammonia. After stirring continuously at room temperature for 24 h, the product was separated by ﬁltration and washed with water, dried at 100 ◦C overnight, and calcinated at 400 ◦C for 4 h. The obtained sample is denoted as LP. Catalysts 2021, 11, 531 10 of 12

Other metal phosphates M-P-O (M = Ce, Mg, Al, Zn) samples were prepared following the same synthesis method as LP. M-P-O prepared from Ce(NO3)3·6H2O, Mg(NO3)2, Al(NO3)3·9H2O, and Zn(NO3)2·6H2O are denoted as CP, MP, ALP, and ZP, respectively.

4.3. Catalyst Characterization X-ray diffraction (XRD) patterns were recorded on a Rigaku Ultima IV diffractometer (Rigaku, Tokyo, Japan) with Cu Kα radiation, operated at 40 kV and 40 mA. Fourier transform infrared (FT-IR) spectra were recorded on a Nicolet iS5 spectrometer (Thermo Scientific, Baltimore, MD, USA). A sample was mixed with dry KBr and com- pressed at 10 MPa for 5 min. Measurements were conducted at a resolution of 0.8 cm−1, using the spectrum of pure KBr tablet as the background. Scanning electron microscopy (SEM) experiments were conducted on a Hitachi S-4000 instrument (Hitachi, Tokyo, Japan). Textural properties were determined with a Micrometrics ASAP 2020 M+C adsorption ◦ apparatus (Micromeritics, Norcross, GA, USA) at liquid N2 temperature (−196 C), using ◦ N2 as adsorbate. Before measurement, all the samples were degassed at 200 C for 10 h. Temperature programmed desorption of NH3 or CO2 was performed on a conven- tional flow apparatus. Prior to adsorption, 0.15 mg sample (40–60 mesh) was pretreated ◦ ◦ in flowing N2 (30 mL/min) at 300 C for 2 h, cooled down to 30 C. Then the sample was exposed to 30 mL/min of 10% NH3/N2 or 10% CO2/N2 flow for 30 min and swept with N2 (30 mL/min) to remove physically adsorbed NH3 or CO2. NH3 or CO2 desorption was conducted in flowing N2 (30 mL/min) by raising the temperature at a heating rate of ◦ 10 C/min. TCD or MS ()(Pfeiffer Omnistar, Germany) was used to detect desorbed NH3 and CO2, respectively.

4.4. Catalytic Evaluation Vapour-phase selective O-methylation of catechol with methanol was carried out in a continuous-flow, fixed-bed microreactor under atmospheric pressure. A 1.5 g catalyst (40–60 mesh) was placed in a quartz tube (i.d. = 10 mm), and pretreated in flowing N2 (50 mL/min) at 400 ◦C for 2 h. After the catalyst cooled down to 270 ◦C, the reactant mixture (catechol: methanol = 1:6, molar ratio) was pumped through a piston pump (WHSV of reactant mixture was 0.4 h−1). The effluent products were condensed using a cold trap, and then analysed via an FLGC9720 gas chromatograph (Fuli, Zhejiang, China) with a flame ionization detector (FID). The catechol conversion was calculated as ([catechol]in − [catechol]out)/[catechol]in × 100%, where [catechol]in is the catechol content in the reactant mixture (no reaction occurs), and [catechol]out is the catechol content in the effluent products. Long-term stability and catalyst regeneration tests were conducted, they were mea- sured as function of time on stream at 270 ◦C. After the first cycle of stability test, the catalyst was regenerated in flowing air (40 mL/min) at 400 ◦C for 6 h, then cooled down to 270 ◦C, the reactant mixture was fed again (denoted as the second cycle).

5. Conclusions Different metal phosphates M-P-O (M = La, Ce, Mg, Al, Zn) catalysts were prepared by precipitation method. MP, ALP and ZP with high amount of strong acid sites showed low catalytic activity in the selective O-methylation of catechol with methanol. The stable catalytic activity of LP can be ascribed to its absence of strong acid sites. The higher activity of CP can be related to the weaker strength and higher amount of the weak acid sites. In addition, the suitable base strength and larger amount of weak base sites contribute to the higher guaiacol selectivity of CP. Most of the catalytic activity of used CP can be recovered by regeneration. Although here we only studied the selective O-methylation of catechol with methanol, we believe that potential uses of metal phosphates can be explored in other alkylation reactions in the future. Catalysts 2021, 11, 531 11 of 12

Supplementary Materials: The following are available online at https://www.mdpi.com/article/ 10.3390/catal11050531/s1, Figure S1: SEM images of (a) CP, (b) LP, (c) MP, (d) ALP, and (e) ZP. (a magnification of 2000), Figure S2: Effect of time on stream on catechol conversion (a) and guaiacol selectivity (b) over CP, SAPO-34, HAP catalysts. Reaction conditions: atmospheric pressure; T = 270 ◦C; −1 WHSV = 0.4 h , Figure S3: NH3-TPD curves of CP, SAPO-34, HAP catalysts, Figure S4: CO2-TPD curves of CP, SAPO-34, HAP catalysts. Author Contributions: Investigation, H.M., X.L., F.X., Z.X., W.Z. and T.M.; supervision, H.M. and T.M.; writing—original draft, X.L.; writing—review and editing, H.M. and T.M. All authors have read and agreed to the published version of the manuscript. Funding: This research received no external funding. Data Availability Statement: Not applicable. Conflicts of Interest: The authors have no conflict of interest to declare.

References 1. Dorothea, G.; Barbara, E. Ullmann´s Encyclopaedia of Industrial Chemistry; Verlagsgesellschaft: Weinheim, Germany, 1991; Volume A19. 2. Vishwanathan, V.; Balakrishna, G.; Rajesh, B.; Jayasri, V.; Sikhwivhilu, L.M.; Coville, N.J. Alkylation of catechol with methanol to give guaiacol over sulphate-modified zirconia solid acid catalysts: The influence of structural modification of zirconia on catalytic performance. Catal. Commun. 2008, 9, 2422–2427. [CrossRef] 3. Stoochnoff, B.A.; Benoiton, N.L. The methylation of some phenols and alcohols with sodium hydride / methyl iodide in tetrahydrofuran at room temperature. Tetrahedron Lett. 1973, 14, 21–24. [CrossRef] 4. Luque, R.; Campelo, J.M.; Conesa, T.D.; Luna, D.; Marinas, J.M.; Romero, A.A. Catechol O-methylation with dimethyl carbonate over different acid-base catalysts. New J. Chem. 2006, 30, 1228–1234. [CrossRef] 5. Jafari, A.A.; Khodadadi, A.; Mortazavi, Y. Vapor-phase selective O-alkylation of catechol with methanol over lanthanum phosphate and its modified catalysts with Ti and Cs. J. Mol. Catal. A 2013, 372, 79–83. [CrossRef] 6. Vishwanathan, V.; Ndou, S.; Sikhwivhilu, L.; Plint, N.; Raghavan, K.V.; Coville, N.J. Evidence for weak base site participation in the vapour phase methylation of catechol over solid base catalysts. Chem. Commun. 2001, 10, 893–894. [CrossRef] 7. Fu, Z.H.; Yu, Y.; Yin, D.L.; Xu, Y.Z.; Liu, H.P.; Liao, H.Y.; Xu, Q.; Tan, F.Q.; Wang, J. Vapor-phase highly selective O-methylation of catechol with methanol over ZnCl2 modified gamma-Al2O3 catalysts. J. Mol. Catal. A 2005, 232, 69–75. [CrossRef] 8. Li, X.M.; Zhang, W.X.; Liu, G.; Jiang, L.; Zhu, X.M.; Pan, C.L.; Jiang, D.Z.; Tang, A.Q. Effect of the P/Al ratio of Al-P-O on the catalytic activity of O-methylation of catechol with methanol. React. Kinet. Catal. Lett. 2003, 79, 365–371. [CrossRef] 9. Zhu, X.M.; Li, X.M.; Jia, M.J.; Liu, G.; Zhang, W.X.; Jiang, D.Z. Vapour-phase selective O-methylation of catechol with methanol over Ti-containing aluminium phosphate catalysts. Appl. Catal. A 2005, 282, 155–161. [CrossRef] 10. Zhu, X.M.; Li, X.M.; Zou, X.J.; Wang, Y.L.; Jia, M.J.; Zhang, W.X. Supported ammonium metatungstate as highly efficient catalysts for the vapour-phase O-methylation of catechol with methanol. Catal. Commun. 2006, 7, 579–582. [CrossRef] 11. Takita, Y.; Qing, X.; Takami, A.; Nishiguchi, H.; Nagaoka, K. Oxidative dehydrogenation of isobutane to isobutene III reaction mechanism over CePO4 catalyst. Appl. Catal. A 2005, 296, 63–69. [CrossRef] 12. Bautista, F.M.; Campelo, J.M.; Luna, D.; Marinas, J.M.; Quiros, R.A.; Romero, A.A. Screening of amorphous metal-phosphate catalysts for the oxidative dehydrogenation of ethylbenzene to styrene. Appl. Catal. B 2007, 70, 611–620. [CrossRef] 13. Takita, Y.; Moriyama, J.; Yoshinaga, Y.; Nishiguchi, H.; Ishihara, T.; Yasuda, S.; Ueda, Y.; Kubo, M.; Miyamoto, A. Adsorption of water vapor on the AlPO4-based catalysts and reaction mechanism for CFCs decomposition. Appl. Catal. A 2004, 271, 55–60. [CrossRef] 14. Takubo, T.; Hirose, Y.; Kashiwagi, D.; Inoue, T.; Yamada, H.; Nagaoka, K.; Takita, Y. Metal phosphate and fluoride catalysts active for hydrolysis of NF3. Catal. Commun. 2009, 11, 147–150. [CrossRef] 15. Takita, Y.; Tanabe, T.; Ito, M.; Ogura, M.; Muraya, T.; Yasuda, S.; Nishiguchi, H.; Ishihara, T. Decomposition of CH2FCF3 (134a) over metal phosphate catalysts. Ind. Eng. Chem. Res. 2002, 41, 2585–2590. [CrossRef] 16. Kashiwagi, D.; Takai, A.; Takubo, T.; Nagaoka, K.; Inoue, T.; Takita, Y. Metal phosphate catalysts effective for degradation of sulfur hexafluoride. Ind. Eng. Chem. Res. 2009, 48, 632–640. [CrossRef] 17. Devi, G.S.; Giridhar, D.; Reddy, B.M. Vapour phase O-alkylation of phenol over alkali promoted rare earth metal phosphates. J. Mol. Catal. A 2002, 181, 173–178. [CrossRef] 18. Li, F.; Zhang, Y.B.; Xiao, D.H.; Wang, D.Q.; Pan, X.Q.; Yang, X.G. Hydrothermal method prepared Ce-P-O catalyst for the selective catalytic reduction of NO with NH3 in a broad temperature range. ChemCatChem 2010, 2, 1416–1419. [CrossRef] 19. Zhang, Y.J.; Wang, J.H.; Zhang, T. Novel Ca-doped CePO4 supported ruthenium catalyst with superior catalytic performance for aerobic oxidation of alcohols. Chem. Commun. 2011, 47, 5307–5309. [CrossRef] 20. Danjo, Y.; Kikuchi, I.; Ino, Y.; Ohshima, M.; Kurokawa, H.; Miura, H. Support effect of Pd/AlPO4 catalyst in hydrogen storage of organic hydride method in the presence of CO. React. Kinet. Mech. Cat. 2012, 105, 381–389. [CrossRef] Catalysts 2021, 11, 531 12 of 12

21. Ma, Z.; Yin, H.; Overbury, S.H.; Dai, S. Metal phosphates as a new class of supports for gold nanocatalysts. Catal. Lett. 2008, 126, 20–30. [CrossRef] 22. Qian, X.S.; Qin, H.M.; Meng, T.; Lin, Y.; Ma, Z. Metal phosphate-supported Pt catalysts for CO oxidation. Materials 2014, 7, 8105–8130. [CrossRef] 23. Cui, Y.W.; Liu, H.; Lin, Y.; Ma, Z. Metal phosphate-supported RuOx catalysts for N2O decomposition. J. Taiwan Inst. Chem. E. 2016, 67, 254–262. [CrossRef] 24. Liu, G.; Wang, Z.; Jia, M.; Zou, X.; Zhu, X.; Zhang, W.; Jiang, D. Thermally stable amorphous mesoporous aluminophosphates with controllable P/Al ratio: Synthesis, characterization, and catalytic performance for selective O-methylation of catechol. J. Phys. Chem. B 2006, 110, 16953–16960. [CrossRef] 25. Liao, X.Z.; Zhou, Z.; Wang, Z.L.; Zou, X.J.; Liu, G.; Jia, M.J.; Zhang, W.X. Preformed precursor of microporous aluminophosphate coating on mesoporous SBA-15: Synthesis, characterization, and catalytic property for selective O-methylation of catechol. J. Colloid Interface Sci. 2007, 308, 176–181. [CrossRef] 26. Rajesh, K.; Mukundan, P.; Pillai, P.K.; Nair, V.R.; Warrier, K.G.K. High-surface-area nanocrystalline cerium phosphate through aqueous sol-gel route. Chem. Mater. 2004, 16, 2700–2705. [CrossRef] 27. Rajesh, K.; Shajesh, P.; Seidel, O.; Mukundan, P.; Warrier, K.G.K. A facile sol-gel strategy for the synthesis of rod-shaped, nanocrystalline, high-surface-area lanthanum phosphate powders and nanocoatings. Adv. Funct. Mater. 2007, 17, 1682–1690. [CrossRef] 28. Qi, C.; Zhu, Y.J.; Wu, C.T.; Sun, T.W.; Chen, F.; Wu, J. Magnesium phosphate pentahydrate nanosheets: Microwave-hydrothermal rapid synthesis using creatine phosphate as an organic phosphorus source and application in protein adsorption. J. Colloid Interface Sci. 2016, 462, 297–306. [CrossRef] 29. Zhou, X.M.; Du, H.J.; Ma, H.; Sun, L.N.; Cao, R.; Li, H.S.; Zhang, P.X. Facile preparation and characterization of zinc phosphate with self-assembled ﬂower-like micro-nanostructures. J. Phys. Chem. Solids 2015, 78, 1–7. [CrossRef] 30. Campelo, J.M.; Jaraba, M.; Luna, D.; Luque, R.; Marinas, J.M.; Romero, A.A.; Navio, J.A.; Macias, M. Effect of phosphate precursor and organic additives on the structural and catalytic properties of amorphous mesoporous AlPO4 materials. Chem. Mater. 2003, 15, 3352–3364. [CrossRef] 31. Srilakshmi, C.; Lingaiah, N.; Nagaraju, P.; Prasad, P.S.S.; Narayana, K.V.; Martin, A.; Luecke, B. Studies on bulk metal phosphate catalysts for the ammoxidation of 2-methylpyrazine. Appl. Catal. A 2006, 309, 247–253. [CrossRef] 32. Chada, R.R.; Ketike, T.; Boilla, A.R.; Gangadharam, S.D.; Kamaraju, S.R.R.; Burri, D.R. Preparation and characterization of lanthanum phosphate catalysts for O-methylation of phenol to anisole in gas phase. Mol. Catal. 2017, 438, 224–229. [CrossRef] 33. Lin, Y.; Meng, T.; Ma, Z. Catalytic decomposition of N2O over RhOx supported on metal phosphates. J. Ind. Eng. Chem. 2015, 28, 138–146. [CrossRef] 34. Lee, H.; Lee, S.; Ryoo, R.; Choi, M. Revisiting side-chain alkylation of toluene to styrene: Critical role of microporous structures in catalysts. J. Catal. 2019, 373, 25–36. [CrossRef] 35. Nguyen, T.T.N.; Ruaux, V.; Massin, L.; Lorentz, C.; Afanasiev, P.; Mauge, F.; Belliere-Baca, V.; Rey, P.; Millet, J.M.M. Synthesis, characterization and study of lanthanum phosphates as light alcohols dehydration catalysts. Appl. Catal. B 2015, 166, 432–444. [CrossRef]