In Situ-Generated Supported Potassium Lactate: Stable Catalysis for

May 3, 2019 - Nor the follow-up of the liquid-phase catalytic dehydration of LA to AA has .... SiO2-Al2O3(23), SiO2-Al2O3(36), SiO2-Al2O3(77), and SiO...
1 downloads 0 Views 9MB Size
Download PDF
This is an open access article published under an ACS AuthorChoice License, which permits copying and redistribution of the article or any adaptations for non-commercial purposes.

Article Cite This: ACS Omega 2019, 4, 8146−8166

http://pubs.acs.org/journal/acsodf

In Situ-Generated Supported Potassium Lactate: Stable Catalysis for Vapor-Phase Dehydration of Lactic Acid to Acrylic Acid Lin Huang,* De Sheng Theng, Lili Zhang, Luwei Chen, Chuan Wang,*,† and Armando Borgna Heterogeneous Catalysis, Institute of Chemical and Engineering Sciences, Agency for Science, Technology and Research, 1 Pesek Road, Jurong Island, 627833 Singapore

Downloaded via 79.133.106.106 on May 3, 2019 at 18:02:26 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

S Supporting Information *

ABSTRACT: We have studied unsupported, silica gel- and amorphous silica−alumina-supported catalysts derived from K salts for the vapor-phase dehydration of lactic acid (LA) to acrylic acid (AA). A catalytic study shows that the supported catalysts improve the activity and selectivity for the production of AA and decrease the selectivity for the production of propionic acid (PA). The silica−aluminasupported catalysts remain fairly stable in the catalytic performance during 90 h of reaction. The IR spectroscopic characterization combined with the catalytic study demonstrates that potassium lactate (C3H5KO3) in situ generated from LA and a K salt is an important reaction intermediate for the production of AA and the catalytic stability is associated with the chemical stability of C3H5KO3 and the activity for the regeneration of C3H5KO3 in the catalytic cycle. On silica−alumina, C3H5KO3 is well stabilized and smoothly regenerated during the reaction, leading to the good catalytic stability. This work suggests for the first time that lactate salt acts as the true catalytic active species for the dehydration of LA to AA. We also propose a predominant reaction pathway for the vapor-phase dehydration of LA to AA with K salt catalyst systems. of CH3CHO and CO as the solution acidity increased.4 In the presence of H2SO4 as a precatalyst, the conversion of LA was 57−100% with the production of CH3CHO and CO/CO2 as the predominant products. With the presence of NaOH instead of the acid as a precatalyst, the conversion of LA decreased to 23−41% with an increase in the yields of AA and propionic acid (PA). The yields of AA and PA reached a maximum (13 and 27%, respectively) when a small amount of NaOH was added.4 When Na2HPO4 and NaOH were used as precatalysts with an initial LA concentration of 0.4 M and residence time of ∼70 s, the yields of AA were found to be 15 and 6.8%, respectively.5 While the poor catalytic performance of these liquid-phase catalyst systems for the dehydration of LA to AA was shown, the nature of the catalytic active species was not discussed in the earlier studies. Nor the follow-up of the liquid-phase catalytic dehydration of LA to AA has been reported. The first patent application on the vapor-phase catalytic dehydration of LA and alkyl lactates to AA and acrylates was filed by Holmen et al. in 1958.6 Higher AA yields of 42−68% were achieved from aqueous 10−50% solutions of LA over sulfate, phosphate, and mixed inorganic salts at 400−425 °C and atmospheric pressure. Evidently, the vapor-phase catalytic reaction displays an incomparable advantage over the liquidphase catalytic reaction in the production of AA. Since then,

1. INTRODUCTION Acrylic acid (AA) and its derivatives are very versatile monomers for many industrial and commodity chemicals such as absorbents, detergents, dispersants, and flocculants.1,2 Due to a complicated propene oxidative process currently used in the industry, which is nonrenewable, the development of new processes producing sustainable AA from biomass derivatives has attracted much research interest. One of the promising biomass derivatives is lactic acid (LA), which is inexpensively available from biomass (e.g., starch) fermentation. AA can be produced from either liquid- or vapor-phase catalytic dehydration of LA: C2H4(OH)COOH → C2H3COOH + H 2O

(1)

The liquid-phase catalytic formation of AA from LA was first reported by Odell et al. in 1985.3 At 150−250 °C and atmospheric pressure or below in the presence of group VIII metal complexes such as [PtH(PEt3)3]+, PtCl2(PPh3)2, IrH(CO)(PPh3)3, and Ni(CO)2(PPh3)2 in aqueous solutions, hydrothermal reactions of LA produced β-hydroxypropanoic acid, pyruvic acid, acetic acid (HAc), acetaldehyde (CH3CHO), acetone, and ethanol (EtOH). Under optimal conditions, an AA yield of only 3.4% was obtained. Later on, the liquid-phase catalytic dehydration of LA to AA was investigated by Mok et al.4 and further by Lira and McCrackin.5 The reaction was reportedly operated in supercritical water at 320−400 °C and 310−340 bar. With an initial LA concentration of 0.1 M and residence time of ∼30 s, the conversion of LA generally increased together with the yields © 2019 American Chemical Society

Received: March 18, 2019 Accepted: April 23, 2019 Published: May 3, 2019 8146

DOI: 10.1021/acsomega.9b00745 ACS Omega 2019, 4, 8146−8166

ACS Omega

Article

AA.7,15,17 This contribution includes some basic catalytic data that suggest a predominant reaction pathway with a lactate salt intermediate for the production of AA. To obtain insights into the reaction mechanism and catalytic stability, IR spectroscopic monitoring is shown for the reactivities of LA and reaction intermediates including potassium lactate (C3H5KO3), potassium acrylate (C3H3KO2), and C3H5KO2 on different support surfaces. We establish the association of the catalytic stability with the chemical stability of C3H5KO3 and the ability of a Kcontaining catalyst system to regenerate C3H5KO3 under the reaction conditions. By combining catalytic and IR spectroscopic studies, we demonstrate that C3H5KO3 is not only an important reaction intermediate but also the true catalyst in the vapor-phase dehydration of LA to AA.

the catalytic process of vapor-phase dehydration of LA to AA has been attracting considerable interest of catalytic investigators for the sakes of not only high efficiency but also green chemical manufactural process.1,2 In this area, the selective catalyst systems reported for the production of AA include sulfate salts, phosphate salts, mixed inorganic salts, phosphate salt-containing materials, and modified zeolites.1,2,7−11 Alkaliand alkali-earth-metal phosphates and alkali-metal-modified zeolites were reported to be prominent catalyst systems for dehydration of LA to AA.1,2,7,12−17 Likely, the catalytic performance for the vapor-phase LA conversion promoted investigators to shed light on the LA reaction mechanism and catalytic active species on various catalyst systems.12,16,18−25 Miller and co-workers earlier studied the vapor-phase LA conversion mechanism on catalyst surfaces by means of IR, NMR, and GC/MS analyses of surface species apart from catalytic testing and kinetics.18−22,26 They discovered that low temperatures ( KNO3/SiO2 > KNO3. As indicated in Figures 13 and 16 and Figure S4, the order of the chemical stability of the C3H5KO3 systems at 350 °C is ranked as C3H5KO3/SiO2-Al2O3(36) > C3H5KO3/SiO2 > C3H5KO3. This order is in accordance with the order of the IR band intensity of C3H5KO3 on the spent KNO3, KNO3/ SiO2, and KNO3/SiO2-Al2O3(36) catalyst samples in Figures 8 and 10. The AA yield and catalytic stability over the KNO3 systems are indeed associated with the chemical stability of C3H5KO3 in these systems. A good chemical stability of C3H5KO3 as the possible catalytic active species can ensure a high and stable conversion of LA to AA. At this stage, we hypothesize that C3H5KO3 would play the roles of important intermediate and catalytic active species in the vapor-phase dehydration of LA to AA. As stated above, the transition of C3H5KO3 to AA does not seem to proceed independently without the attendance of LA during the reaction process. Although initial C3H5KO3 as the true catalyst is easily formed from LA and a starting K salt, the regeneration of C3H5KO3 poses a challenge in the catalytic cycle. The regeneration of C3H5KO3 usually more or less weakens as the catalytic reaction proceeds. The ability to regenerate C3H5KO3 from LA is related to the actual activity of a catalyst system and certainly will influence the reaction rate and LA conversion. The approaches to regenerating C3H5KO3 are likely to involve the reactions of LA with C3H3KO2 following the C3H5KO3 dehydration step and C3H5KO2 following the C3H3KO2 hydrogenation step. This certainly will affect the AA and PA yields. To acquire further insights into the reaction pathway and catalyst deactivation issue with C3H5KO3, we next 8160

DOI: 10.1021/acsomega.9b00745 ACS Omega 2019, 4, 8146−8166

ACS Omega

Article

Figure 18. IR spectra after 0.5 h of reaction of LA with the KBr-supported species shown in Figure 17f under flowing N2 at (a) 22 °C, (b) 100 °C, (c) 200 °C, (d) 300 °C, and (e) 350 °C.

Figure 19. IR spectra of the following treatments on KBr-supported spent KNO3 in 28 h of the vapor-phase dehydration of LA at 350 °C: (a) under vacuum at 22 °C for 1 h, (b) under vacuum at 350 °C for 0.5 h, (c) with LA under flowing N2 at 22 °C for 0.5 h, (d) with LA under flowing N2 at 100 °C for 0.5 h, and (e) with LA under flowing N2 at 200 °C for 0.5 h.

in the production of AA during the reaction. At 200 °C, C3H3KO2 almost vanished in favor of the formation of C3H5KO3 (Figure 17d). The results demonstrate that the displacement of unsupported C3H3KO2 with LA susceptibly takes place to generate C3H5KO3 and release AA. When the temperature was further increased, the 1590 cm−1 band shifted toward 1568 cm−1 with the concurrent reappearance of the features of C3H3KO2. This accounts for that the C3H5KO3 produced from LA and C3H3KO2 is prone to dehydrating to C3H3KO2 at higher temperatures (≥300 °C), in agreement with the IR results on the spent unsupported K salt catalyst samples in Figure 8 and the thermal reactivity of C3H5KO3 on KBr (see the Supporting Information). Following the above reaction at 350 °C, the KBr-supported product wafer was allowed to further react with LA for testing the ability to further regenerate C3H5KO3. The IR results are presented in Figure 18. After LA was added onto the wafer followed by treatment under flowing N2 at 22 °C, the 1568

examined the reactivities of LA with C3H3KO2 and C3H5KO2 on KBr and SiO2-Al2O3(36). In Figure 17, the surface IR spectra observed in the 2500− 900 cm−1 region during the thermal reaction of LA and C3H3KO2 on KBr under flowing N2 are shown. A KBr wafer was impregnated with a drop of aqueous 10% C3H3KO2 solution in air followed by removal of water under vacuum at 22 °C for 0.5 h. The resultant C3H3KO2/KBr was impregnated with a drop of aqueous 10% LA solution and subsequently treated under flowing N2 at 22 °C for 3 min. At this point, the signature shoulder band of C3H5KO3 at 1591 cm−1 appeared together with a band of AA at 983 cm−1, indicating the easy formation of C3H5KO3 and AA from LA and C3H3KO2, as shown in Figure 17a. As the reaction was continued, the signature bands of C3H3KO2 at 1636 and 1559 cm−1 and LA at 1726 cm−1 decreased in intensity in favor of the increase of the band intensity of C3H5KO3. Meanwhile, the shift of the 1726 cm−1 band to 1708 cm−1 implies an increase 8161

DOI: 10.1021/acsomega.9b00745 ACS Omega 2019, 4, 8146−8166

ACS Omega

Article

Figure 20. IR spectra after the reaction of LA and C3H5KO2 on KBr under flowing N2 at (a) 22 °C for 5 min, (b) 22 °C for 0.5 h, (c) 100 °C for 0.5 h, (d) 200 °C for 0.5 h, and (e) 350 °C for 0.5 h.

Figure 21. IR spectra after 0.5 h of reaction of LA and C3H3KO2 on SiO2-Al2O3(36) under flowing N2 at (a) 22 °C, (b) 100 °C, (c) 200 °C, and (d) 350 °C.

cm−1 band shifted to 1585 cm−1. This indicates that C3H5KO3 is regenerated only partly from the supported product. Increasing the temperature was unable to produce more C3H5KO3. The dehydration of C3H5KO3 to C3H3KO2 was carried out again at higher temperatures (≥300 °C). When a wafer of the KBr-supported spent KNO3 catalyst sample in 28 h of the vapor-phase dehydration of LA at 350 °C was subjected to reacting with LA, the signature band of C3H3KO2 at 1566 cm−1 was able to shift to 1575 cm−1 only, as shown in Figure 19. This hints that C3H5KO3 is regenerated even more difficultly from the spent KNO3 catalyst sample. This is also evidence that, with unsupported K salt catalyst systems, the ability to regenerate C3H5KO3 weakens obviously as the catalytic cycle increases. In Figure 20, the surface IR spectra recorded in the 2500− 900 cm−1 region during the thermal reaction of LA and C3H5KO2 on KBr under flowing N2 are shown. A KBr wafer was first impregnated with a drop of aqueous 10% C3H5KO2 solution in air followed by removal of water under vacuum at 22 °C for 0.5 h. Upon impregnation of the resultant

C3H5KO2/KBr with a drop of aqueous 10% LA solution, the signature band of C3H5KO2 at 1562 cm−1 nearly vanished, while the signature band of C3H5KO3 at 1590 cm−1 appeared, as seen in Figure 20a. The result suggests the easy reaction of LA and C3H5KO2 to produce C3H5KO3 and PA: C2H4(OH)COOH + C2H5COOK → C2H5COOH + C2H4(OH)COOK

(10)

The absence of bands of PA may be due to the difficult adsorption of PA on the surface. The reaction was markedly accelerated when heated to 100 °C, the 1590 cm−1 band of C3H5KO3 growing at the expense of the 1716 cm−1 band of LA and the 1562 and 1370 cm−1 bands of C3H5KO2, as shown in Figure 20c. When heated to 200 °C, a nearly pure spectrum of C3H5KO3 was present. The simultaneous presence of a tiny band at 1760 cm−1 is ascribed to the production of poly(LA) in trace amounts. As expected, raising the temperature to 350 °C led to the dehydration of C3H5KO3 to C3H3KO2, as shown by the spectral change between Figure 20d,e. It follows that the 8162

DOI: 10.1021/acsomega.9b00745 ACS Omega 2019, 4, 8146−8166

ACS Omega

Article

Figure 22. IR spectra after the reaction of LA and C3H5KO2 on SiO2-Al2O3(36) under flowing N2 at (a) 22 °C for 3 min, (a) 22 °C for 0.5 h, (c) 100 °C for 0.5 h, (d) 200 °C for 0.5 h, and (e) 350 °C for 0.5 h.

Scheme 1. Proposed Reaction Pathway for the Vapor-Phase Dehydration of LA to AA with a K Salt Catalyst System

vacuum at 22 °C for 0.5 h. The resultant C3H3KO2/SiO2Al2O3(36) was impregnated with a drop of aqueous 10% LA solution and subsequently treated under flowing N2 at 22 °C for 0.5 h. The obtained spectrum contained surface molecular H2O and C3H3KO2 at 1639(m), 1553(s), and 1364(m) cm−1. From this spectrum, LA has rather low reactivity with C3H3KO2/SiO2-Al2O3(36) at 22 °C. While the temperature was raised toward 350 °C, the bands of C3H3KO2 at 1639 and

reactivity of LA and C3H5KO2 is similar to that between Figure 20d,e. It follows that the reactivity of LA and C3H5KO2 is similar to that of LA and C3H3KO2. In Figure 21, the surface IR spectra observed in the 2500− 1100 cm−1 region during the thermal reaction of LA and C3H3KO2 on SiO2-Al2O3(36) under flowing N2 are shown. A SiO2-Al2O3(36) wafer was impregnated with a drop of aqueous 10% C3H3KO2 solution followed by removal of water under 8163

DOI: 10.1021/acsomega.9b00745 ACS Omega 2019, 4, 8146−8166

ACS Omega

Article

1364 cm−1 gradually fell in intensity, whereas the bands of LA and C3H3KO2 at 1725 and 1553 cm−1 shifted to 1702 and 1612 cm−1, respectively. We thus reason that the reactivity of LA with C3H3KO2 and the surface Na+ to form C3H5KO3, C3H5NaO3, and AA increases with increasing temperature. After the temperature had reached 350 °C, the bands of both LA and C3H3KO2 tremendously declined in intensity in favor of the emergence of bands of C3H5KO3 and C3H5NaO3 at 1612(s), 1455(m), and 1410(m) cm−1. This hints that the displacement of LA with C3H3KO2 to produce C3H5KO3 and AA is favorable at 350 °C. The upward shift of the signature band of C3H5KO3 and C3H5NaO3 to 1612 cm−1 may be due to the interaction between the lactate salt and AA on SiO2Al2O3(36). The fact that C3H5KO3 and C3H5NaO3 are stably retained on SiO2-Al2O3(36) at 350 °C shows that C3H5KO3 and C3H5NaO3 dehydrate to C3H3KO2 and C3H3NaO2 on SiO2-Al2O3(36) at 350 °C harder. In Figure 22, the surface IR spectra observed in the 2500− 1100 cm−1 region during the thermal reaction of LA and C3H5KO2 on SiO2-Al2O3(36) under flowing N2 are shown. A SiO2-Al2O3(36) wafer was impregnated with a drop of aqueous 10% C3H5KO2 solution followed by removal of water under vacuum at 22 °C for 0.5 h. The resultant C3H5KO2/SiO2Al2O3(36) was impregnated with a drop of aqueous 10% LA solution and displayed the features of LA at 1725(w) cm−1 and lactate salts around 1615(sh) cm−1 (subsequently treated under flowing N2 at different temperatures). After 0.5 h of reaction at 22 °C, the upward shift of the signature band of C3H5KO2 at 1561 cm−1 already occurred, suggesting that LA and C3H5KO2 already starts to transform to C3H5KO3 and PA on SiO2-Al2O3(36) at 22 °C. With increasing temperature, the band shift kept increasing up to 1604 cm−1 at 350 °C. In the meantime, the signature band of LA at 1723 cm−1 shifted to 1715 cm−1, hinting the formation of small amounts of AA on the surface. The observations implicate that LA would catalytically react with C3H5KO2 and the surface Na+ to form C3H5KO3, C3H5NaO3, PA, and AA on SiO2-Al2O3(36) during this process. The existence of the signature band of C3H5KO3 and C3H5NaO3 at 1604 cm−1 after 0.5 h of heating at 350 °C confirms the good chemical stability of C3H5KO3 and C3H5NaO3 on SiO2-Al2O3(36) at 350 °C. The lack of bands of PA may likewise be due to the difficult adsorption of PA on the surface. The presence of the 1604 cm−1 band further assumes that the interaction between the lactate salt and AA on SiO2-Al2O3(36) would cause an upward shift of the main band of the lactate salt. It seems that C3H5KO2 is more reactive with LA than C3H3KO2 on SiO2-Al2O3(36). The good thermochemical stability of C3H5KO3 issued from LA and C3H3KO2 or C3H5KO2 on SiO2-Al2O3(36) stated above virtually reflects the strong ability to regenerate C3H5KO3 from the supported K salt catalyst systems as the catalytic reaction proceeds. This matches with the IR results of the spent SiO2-Al2O3(36)-supported K salt catalyst samples in Figure 10. The combined catalytic and IR studies presented above have suggestively established the relationship among LA, C3H5KO3, C3H3KO2, AA, C3H5KO2, and PA in the dehydration process of LA to AA and PA over the K salt (nitrate, hydroxide, and carbonate) catalyst systems. We propose a general reaction pathway for the vapor-phase dehydration of LA to AA with a K salt catalyst system, as depicted in Scheme 1. In a reaction from LA over a K salt catalyst system, on principle, the AA yield is dependent on the catalytic selectivities for both C3H5KO3

dehydration and regeneration steps. Given a catalytic selectivity for the conversion of LA to AA, the catalytic stability is related to the ability of the catalyst system to regenerate C3H5KO3 from LA and C3H3KO2 or C3H5KO2. In this work, the contrast in the catalytic stability over KNO3, KNO3/SiO2, and KNO3/SiO2-Al2O3(36) is shown to reflect their distinct abilities to regenerate C3H5KO3 from LA and C3H3KO2 or C3H5KO2 based on the IR results of the spent catalyst samples and the reactivity between LA and C3H3KO2 or C3H5KO2. The three KNO3 catalyst systems give rise to LA conversion of 67−97% and selectivity to AA of 35−43% at an initial stage (5 h). However, only the KNO3/SiO2-Al2O3(36) system behaves well in the catalytic stability during the reaction on stream because of its strong ability to regenerate C3H5KO3 from LA and C3H3KO2 or C3H5KO2. The cause leading to the difference in the activity of C 3 H 5 KO 3 regeneration from LA and C3H3KO2 or C3H5KO2 is not clear for the time being. The relevant factors may presumably involve the catalyst acid−base property, interaction between K salt and support, dispersion of K salt and catalyst coking, etc. An in-depth investigation is in progress. Since the production of AA requires the presence of C3H5KO3, the ability of a K salt catalyst system to regenerate C3H5KO3 from LA and C3H3KO2 or C3H5KO2 and the chemical stability of C3H5KO3 under the catalytic conditions determine the catalytic performance stability of this system. The poor regeneration ability of C3H5KO3 affects the reaction rate, LA conversion, and AA yield. Both easy decomposition of C 3H5KO3 and the consumption of LA via other pathways are responsible for low AA yields. In the case with an unsupported or SiO2-supported K salt as a precatalyst, the step of the C3H3KO2-to-AA conversion with the regeneration of C3H5KO3 via the displacement of C3H3KO2 with LA proceeds more slowly than the C3H5KO3to-C3H3KO2 conversion step via dehydration at 350 °C, as demonstrated by the IR observations (Figures 8 and 17−19). C3H5KO2 could be more readily formed via C3H3KO2 hydrogenation to give PA eventually. The steps of formation of AA and PA from C3H3KO2 and C3H5KO2, respectively, are believed to be rate-limiting. In the case with a SiO2(36)-Al2O3supported K salt as a precatalyst, the step of the C3H5KO3-toC3H3KO2 conversion proceeds more slowly than the step of the C3H3KO2-to-AA conversion with the regeneration of C3H5KO3 at 350 °C, as shown by the IR results (Figures 10 and 21). The displacement of C3H3KO2 with LA could go more rapidly, and the formation of C3H5KO2 and PA could get suppressed. The C3H5KO3-to-C3H3KO2 conversion step is assumed to be rate-limiting. In both cases, C3H5KO3 is suggested to be the true catalytic active species and an important intermediate for the vapor-phase dehydration of LA to AA.

4. CONCLUSIONS K or Na salt catalyst systems (nitrate, hydroxide, and carbonate) enable the vapor-phase dehydration of LA to AA to proceed predominantly via a pathway with the lactate salt intermediate. The unsupported K salt systems give rise to LA conversion and selectivity to AA of ∼70 and ∼35%, respectively. The SiO2-supported K salt systems exhibit increased LA conversion and decreased selectivity to PA and similar selectivity to AA. Nevertheless, these two types of systems are catalytically unstable and deactivate seriously in the reaction on stream. The SiO2-Al2O3(36)-supported K salt 8164

DOI: 10.1021/acsomega.9b00745 ACS Omega 2019, 4, 8146−8166

ACS Omega

Article

(2) Bonnotte, T.; Paul, S.; Araque, M.; Wojcieszak, R.; Dumeignil, F.; Katryniok, B. Dehydration of lactic acid: the state of the art. ChemBioEng Rev. 2018, 5, 34−56. (3) Odell, B.; Earlam, G.; Cole-Hamilton, D. J. Hydrothermal reactions of lactic acid catalysed by group VIII metal complexes. J. Organomet. Chem. 1985, 290, 241−248. (4) Mok, W. S. L.; Antal, M. J., Jr.; Jones, M., Jr. Formation of acrylic acid from lactic acid in supercritical water. J. Org. Chem. 1989, 54, 4596−4602. (5) Lira, C. T.; McCrackin, P. J. Conversion of lactic acid to acrylic acid in near-critical water. Ind. Eng. Chem. Res. 1993, 32, 2608−2613. (6) Holmen, R. E.; Township, W. B.; County, R. Production of acrylates by catalytic dehydration of lactic acid and alkyl lactates. U.S. Patent 2,859,240, 1958. (7) Yan, B.; Tao, L.-Z.; Mahmood, A.; Liang, Y.; Xu, B.-Q. Potassium-ion-exchanged zeolites for sustainable production of acrylic acid by gas-phase dehydration of lactic acid. ACS Catal. 2017, 7, 538− 550. (8) Zhang, X.; Lin, L.; Zhang, T.; Liu, H.; Zhang, X. Catalytic dehydration of lactic acid to acrylic acid over modified ZSM-5 catalysts. Chem. Eng. J. 2016, 284, 934−941. (9) Li, X.; Chen, Z.; Cao, P.; Pu, W.; Zou, W.; Tang, C.; Dong, L. Ammonia promoted barium sulfate catalyst for dehydration of lactic acid to acrylic acid. RSC Adv. 2017, 7, 54696−54705. (10) Nagaraju, N.; Kumar, V. P.; Srikanth, A.; Rajan, N. P.; Chary, K. V. R. Vapor-phase catalytic dehydration of lactic acid to acrylic acid over nano-crystalline cerium phosphate catalysts. Appl. Petrochem. Res. 2016, 6, 367−377. (11) Lyu, S.; Wang, T. Efficient production of acrylic acid by dehydration of lactic acid over BaSO4 with crystal defects. RSC Adv. 2017, 7, 10278−10286. (12) Zhang, J.; Zhao, Y.; Pan, M.; Feng, X.; Ji, W.; Au, C.-T. Efficient acrylic acid production through bio lactic acid dehydration over NaY zeolite modified by alkali phosphates. ACS Catal. 2011, 1, 32−41. (13) Ghantani, V. C.; Lomate, S. T.; Dongare, M. K.; Umbarkar, S. B. Catalytic dehydration of lactic acid to acrylic acid using calcium hydroxyapatite catalysts. Green Chem. 2013, 15, 1211−1217. (14) Yan, B.; Tao, L.-Z.; Liang, Y.; Xu, B.-Q. Sustainable production of acrylic acid: catalytic performance of hydroxyapatites for gas-phase dehydration of lactic acid. ACS Catal. 2014, 4, 1931−1943. (15) Yan, B.; Tao, L.-Z.; Liang, Y.; Xu, B.-Q. Sustainable production of acrylic acid: alkali-ion exchanged beta zeolite for gas-phase dehydration of lactic acid. ChemSusChem 2014, 7, 1568−1578. (16) Guo, Z.; Theng, D. S.; Tang, K. Y.; Zhang, L.; Huang, L.; Borgna, A.; Wang, C. Dehydration of lactic acid to acrylic acid over lanthanum phosphate catalysts: the role of Lewis acid sites. Phys. Chem. Chem. Phys. 2016, 18, 23746−23754. (17) Zhang, L.; Theng, D. S.; Du, Y.; Xi, S.; Huang, L.; Gao, F.; Wang, C.; Chen, L.; Borgna, A. Selective conversion of lactic acid to acrylic acid over alkali and alkaline-earth metal co-modified NaY zeolites. Catal. Sci. Technol. 2017, 7, 6101−6111. (18) Gunter, G. C.; Miller, D. J.; Jackson, J. E. Formation of 2,3pentanedione from lactic acid over supported phosphate catalysts. J. Catal. 1994, 148, 252−260. (19) Gunter, G. C.; Craciun, R.; Tam, M. S.; Jackson, J. E.; Miller, D. J. FTIR and 31P-NMR spectroscopic analyses of surface species in phosphate-catalyzed lactic acid conversion. J. Catal. 1996, 164, 207− 219. (20) Wadley, D. C.; Tam, M. S.; Kokitkar, P. B.; Jackson, J. E.; Miller, D. J. Lactic acid conversion to 2,3-pentanedione and acrylic acid over silica-supported sodium nitrate: reaction optimization and identification of sodium lactate as the active catalyst. J. Catal. 1997, 165, 162−171. (21) Tam, M. S.; Gunter, G. C.; Craciun, R.; Miller, D. J.; Jackson, J. E. Reaction and spectroscopic studies of sodium salt catalysts for lactic acid conversion. Ind. Eng. Chem. Res. 1997, 36, 3505−3512. (22) Tam, M. S.; Craciun, R.; Miller, D. J.; Jackson, J. E. Reaction and kinetic studies of lactic acid conversion over alkali-metal salts. Ind. Eng. Chem. Res. 1998, 37, 2360−2366.

systems display the fairly stable catalytic performance in 90 h of reaction, in which the LA conversion and selectivity to AA reach more than 75 and 40%, respectively, and the selectivity to PA remains