Low-Temperature Dehydrogenation of Ethanol on Atomically

Nov 24, 2015 - Atomically dispersed gold supported on nanoscale ZnZrOx composite oxides was prepared and investigated in this work as a catalyst for t...
0 downloads 12 Views 3MB Size
Subscriber access provided by The University of Liverpool

Article

Low-Temperature Dehydrogenation of Ethanol on Atomically Dispersed Gold Supported on ZnZrO x

Chongyang Wang, Gabriella Garbarino, Lawrence F. Allard, Faith Wilson, Guido Busca, and Maria Flytzani-Stephanopoulos ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.5b01593 • Publication Date (Web): 24 Nov 2015 Downloaded from http://pubs.acs.org on November 29, 2015

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Catalysis is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 20

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

Low-Temperature Dehydrogenation of Ethanol on Atomically Dispersed Gold Supported on ZnZrOx Chongyang Wang†, Gabriella Garbarino†, ‡, Lawrence F. Allard§, Faith Wilson†, Guido Busca‡, Maria Flytzani-Stephanopoulos†* †

Tufts University, Department of Chemical and Biological Engineering, 4 Colby Street, Medford, MA 02155, USA



University of Genoa, Department of Civil, Chemical and Environmental Engineering (DICCA), Piazzale Kennedy 1, I-16129 Genoa, Italy §

Oak Ridge National Laboratory, Materials Science and Technology Division, Oak Ridge, TN 37831, USA

ABSTRACT: Atomically dispersed gold supported on nanoscale ZnZrOx composite oxides was prepared and investigated in this work as a catalyst for the low-temperature ethanol dehydrogenation reactions. The composite ZnZrOx support disperses gold atomically and stabilizes it against growth much better than either of the neat oxides. Sequential ethanol conversion reactions to acetaldehyde and acetone take place on the Au/ZnZrOx catalysts within well-separated temperature windows over the range of tested temperatures (30-400˚C). ZnO modulates the acidity of the ZrO2 surface, and the extent of this was followed by isopropanol temperature-programmed desorption with online mass spectrometry (IPA-TPD/MS). Catalyst activity and selectivity were tested by temperature-programmed surface reaction (TPSR) and under steady-state reaction conditions. The work has demonstrated that ZnZrOx with optimized ZnO distribution preserves the active AuOx-surface species under reaction conditions and suppresses undesired dehydration reactions. Addition of gold on the bare zirconia support passivates the acid sites catalyzing ethanol dehydration, and introduces desired dehydrogenation sites at low temperatures (~200˚C).

KEYWORDS: Ethanol dehydrogenation; acetone; acetaldehyde; gold catalysts; zirconia; zinc oxide

INTRODUCTION With the advances in cellulosic biomass processing, bio-derived ethanol is becoming more abundant, and its use as a feedstock for “green” chemicals production may be realized in the near future1-2. During the past decade, the production of hydrogen from ethanol has been intensively studied, especially for application to fuel cells3-4. The reaction most investigated is ethanol steam reforming reaction (ESR: C2H5OH + 3H2O →2CO2 +6H2). The selectivity to H2 production and catalyst stability under steam reforming conditions has been addressed using various catalysts5, especially Ni-6-9 and Co-based catalysts10-12, with the latter aimed at suppressing coking and the byproduct methane by ethanol decomposition (C2H5OH → CH4 + CO + H2). Thermodynamic equilibria in the temperature range of 100-1000˚C predict the main products of ESR to be the ethanol cracking products like CO, CH4, CO2 and H2. However, products of ethanol dehydration and dehydrogenation are commonly detected in practical reactions in this temperature range, suggesting that these reactions are kinetically controlled and occur faster before steam reforming light off13. Hence, another application of bio-ethanol would be as a feedstock to value-added chemicals14, e.g. the production of acetaldehyde on Au-based bimetallic catalysts15-18, or the production of acetone19-23 and isobutene24-26 on metal oxide

catalysts. Catalytic conversion of ethanol involves a cascade of elementary reactions and often a thorough understanding of the reaction mechanisms is lacking. Fundamental work by Mavrikakis and Barteau27 has discussed the thermal stability and configuration of C1 and C2 alcohol adsorption on group VIII (Ni, Pd, Pt, Ru) and group IB metals (Cu, Ag) as: Alcohol→ Alkoxide → η1(O) or η2(C,O)-surface-bonded aldehyde → Acyl Formation of the aldehyde species requires rapid transfer of hydroxyl group hydrogen and α-hydrogen to adjacent basic oxygen28. The adsorption configuration of the aldehyde species depends primarily on the electronegativity of the metal surface, i.e. η1(O)-configuration occurs via carbonyl compound bonding with the surface through an oxygen lone pair orbital, acting as a Lewis base, while the η2(C,O)-configuration occurs with backdonation of electrons from the metal to carbon in the carbonyl compound. Stability of the η2-adsorption promotes decomposition of the adsorbates more than desorption in η1-adsorption. At low (150-250˚C) and intermediate temperatures (250-350˚C), catalytic conversion of ethanol to value-added chemicals, like acetaldehyde and acetone, includes molecular adsorption of ethanol on the catalyst surface; scission of O-H bond to form ethoxy species, followed by dehydrogenation of the ethoxy species to acetaldehyde (Rxn 1) and oxidation of acetaldehyde to acetone by mobile surface oxygen species (Rxn 2).

ACS Paragon Plus Environment

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

CH3CH2-OH



CH3CHO

+

H2

(1) 2CH3CHO + O*→ CH3COCH3 +CO2 + H2

(2)

Page 2 of 20

interface of metal and support, CuO/CeO2. Tanabe and coworkers20-21 reported that on basic metal oxides, like ZnO and CaO, acetate condensation is the reaction pathway of ethanol conversion to acetone. Appel et al.23 have proposed that both aldol condensation from acetaldehyde as well as acetate condensation can take place on Cu/ZrO2 catalysts.

Deviations depend on whether acetaldehyde or acetate is the precursor to acetone synthesis (Sch. S1). Nishiguchi et al.22 have argued the formation of acetaldehyde on CuO itself, but the acetate condensation on the Table 1. Physical and chemical properties of ZnO, ZrO2 and ZnZrOx and Au/ZrO2 nano structures Sample

Au/ ZrO2

ZrO2

Zn1Zr10Ox

Zn1Zr5Ox

Zn1Zr1Ox

ZnO

BET surf. area (m /g)

141

146

136

116

105

31

Bulk composition (ICP)

NA

NA

Zn1Zr7.8Ox

Zn1Zr3.7Ox

Zn1Zr0.8Ox

NA

Surf. composition (XPS)

NA

NA

Zn1Zr13.7Ox

Zn1Zr6.4Ox

Zn1Zr1.3Ox

NA

t-ZrO2

t-ZrO2

t-ZrO2

t-ZrO2

t-ZrO2, h-ZnO

hZnO

4

4

4

4

4(ZrO2), 11(ZnO)

31

2

Crystal phase (XRD) Crystal size (XRD) (nm) IPA-TPD (˚C)

b

193

161, 181

168, 183

174, 196

193

NA

c

0

145, 8

95, 97

27, 92

58

0

d

5.8

0

0

0

0

0

Propene (mmol/g) Acetone (mmol/g)

a

a

b

: Crystal phase of the nano structures: t-ZrO2 is tetragonal zirconia, h-ZnO is hexagonal ZnO. : Propene or acetone producc d tion peak temperature from IPA-TPD, heating from 30-400˚C in 4h. : Propene and : acetone amount produced per gram of catalyst by integrating the deconvoluted TPD-peak areas.

In this paper, we report the dehydrogenation of ethanol at low and intermediate temperatures on ZnZrOxsupported Au-Ox- species. The nanoscale ZnZrOx composite oxide incorporates ZnO into ZrO2 with modulated overall acidity as a function of the content of ZnO29, and provides many more binding sites for atomic gold than either of its constituent oxides. We have previously reported that highly dispersed Au-Ox–ZnO species are the active sites for the dehydrogenation and self-coupling of methanol30-31 to methyl formate, which is an important intermediate in the steam reforming reaction of methanol to CO2 and H2 without any CO production over a wide temperature window, from 150-275˚C. The methyl formate pathway is in agreement with the reports of partial oxidation of methanol on unsupported nanoporous gold32 and atomic O- activated single crystal surfaces of Au[111]3334 . The atomically dispersed gold would be interesting to test as catalyst for similar, but more complex C2-alcohol reactions. In this work, we have investigated the ethanol dehydrogenation reaction and were able to identify and delineate the reactions that occur exclusively at low temperature on the gold species from those on the support at higher temperatures, which is important for practical applications of this type catalyst.

EXPERIMENTAL Nanoscale ZnO, ZrO2 and ZnZrOx supports were prepared by adaptation of a carbon hard template method reported recently24. The desired amounts of zirconyl nitrate hydrate (Sigma-Aldrich, ZrO(NO3)2• xH2O) and zinc nitrate hydrate (Sigma-Aldrich, Zn(NO3)2•6H2O) were

dissolved in DI water to designed Zn:Zr atomic ratios of 1:10, 1:5 and 1:1. The solution of the precursor salts was impregnated onto carbon black (Cabot, BP2000) to incipient wetness endpoint. The use of a porous carbon template ensures that precursors are well mixed, and homogeneous distribution of Zn/Zr components is preserved in the bulk structure of binary metal oxides. Two-step calcination was conducted at 400˚C for 4h and 550˚C for 20h in air sequentially. Gold (Alfa-Aesar, HAuCl4•3H2O) was added on the supports by the anion-adsorption method35 to reach 1wt.% loading, as detailed in previous work29, 35. The gold-containing samples were calcined at 300˚C in 2% O2/He for 2h after drying at room temperature in vacuum for 12h. NaCN-leaching was conducted by dissolving 0.5g of the parent gold catalysts in 0.05 wt.% NaCN aqueous solution for 3 min, followed by DI water washing and drying at room temperature. IR spectra were recorded with a Thermo Nicolet FT-IR 380 spectrometer using pressed disks of 0.01g sample powder diluted in 1g of KBr. Diffuse reflectance UVVis-NIR spectra were collected using a Jasco V570 instrument. Aberration-corrected high-angle annular dark-field scanning transmission electron microscopy (ac-HAADFSTEM) was conducted using a JEOL 2200FS instrument. Titration of surface acidity/basicity with isopropanol was carried out by temperature-programmed desorption (TPD) with online mass spectrometry (RGA, MKS model RS-1) monitoring the effluent gases. Selectivity and activity tests were conducted in both temperatureprogrammed surface reaction (TPSR) mode and steady-

ACS Paragon Plus Environment

Page 3 of 20

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

state mode at atmospheric pressure, both with online mass spectrometry. The reactants, either pure ethanol or ethanol/water mixtures were injected into He carrier gas through a syringe pump and vaporized with trace heating of the tubing. Details of the test procedure are given in the Supporting Information (SI).

RESULTS AND DISCUSSION Physical and chemical properties of ZrO2, ZnO and ZnZrOx nanostructures are shown in Table 1. X-ray diffractograms (Fig. S1) show highly crystallized structures for all the samples synthesized by the carbon black hard-template method. Pure ZrO2 and ZnZrOx with low Zn concentration (Zn/Zr = 1:5, 1:10) contain solely the tetragonal zirconia phase (P42/nmc). Zn1Zr1Ox, with high concentration of Zn in the composite oxide, shows both the tetragonal zirconia phase (P42/nmc) and hexagonal zinc oxide phase (P63/mc) with prominent characteristic peaks corresponding to ZnO. Nanoparticles of ZnO of 11 nm on average are seen in this sample. The small crystallite size (4 nm, Table 1) of ZrO2 in all samples and the lack of the ZnO phase in all but Zn1Zr1Ox indicates that ZrO2 may incorporate small amounts of ZnO in its lattice. There appears a slight peak shift of the tetragonal ZrO2 phase towards higher theta in the ZnZrOx samples compared to pure ZrO2, which indicates lattice size decrease due to the solubility of Zn2+ into the ZrO2 phase36. In contrast, the peaks of ZnO in Zn1Zr1Ox do not shift compared to pure ZnO, showing no Zr4 entering the ZnO phase. The bulk Zn:Zr ratios of the ZnZrOx composite oxides were measured by ICP in triplicates and are shown in Table 1. Surface compositions of Zn:Zr were calculated based on the peak fittings of high-resolution XP spectra of Zn 2p and Zr 3d spectral regions. The surface composition of the ZnZrOx composite oxides shows considerable enrichment of Zr. Thorough mixing of Zn and Zr precursors is achieved by using a carbon black template (Cabot Black Pearl 2000) with large pore volume (3ml/g). The homogeneous dispersion of ZnO in the crystalline ZrO2 in sample Zn1Zr10Ox is shown by scanning transmission electron microscope images of the Au/ZnZrOx samples with EDS elemental mapping (Fig. S2-d). Even in Zn1Zr1Ox with extra ZnO nanoparticles, the latter remain dispersed in the composite (Fig. S2-c). Therefore, this method serves to prepare well-dispersed, small ZnO particles in ZrO2. Beyond the homogeneous morphology of the ZnZrOx composite oxides, there clearly has formed chemical bonding between the ZnO and ZrO2 phases. Highresolution XP spectra of Zn 2p and Zr 3d of ZnO, ZrO2 and ZnZrOx are compared in Figure 1. All the spectra were referenced to the elemental carbon at binding energy (BE) of 284.8eV. XP spectra of Zn 2p show a distinct trend of binding energy shift towards higher values on ZnZrOx samples as the amount of ZnO is decreased indicating the strong interaction between ZnO clusters and ZrO2. As the Zn amount increases, the properties of pure ZnO start to dominate. The opposite trend was observed in the Zr 3d spectra. The Zr binding energy of ZnZrOx samples with

higher Zn concentration tends to shift toward lower values, where more ZrO2 is in close contact with ZnO. The energy shift trends of Zn and Zr provide clear evidence of the strong association between ZnO and ZrO2 phases, with a likely substitution of Zn2+ into the ZrO2 lattice36. The surface properties of the composite oxides are thus expected to differ from those of pure ZrO2.

Figure 1. XP spectra of Zn 2p and Zr 3d regions of ZnO, ZrO2 and ZnZrOx nano structures

Skeletal IR spectra of the ZnO, ZrO2 and ZnZrOx in the wavenumber range of 400-1400 cm-1 are shown in Figure S3. Addition of Zn to the domain of ZrO2 produces a shift of the complex band toward lower wavenumbers, which can be a further indication of the substitution of Zn2+ into the ZrO2 lattice as deduced also from the XRD profiles (Fig. S1). UV-Vis spectra in the wavelength range of 230-600 nm are shown in Figure S4. The spectrum of ZrO2 with a weak component at 235nm indicates the presence of a small amount of monoclinic zirconia, which was not detected by XRD. Features of Zn are detectable even in Zn1Zr10Ox; the shape of the absorption curve indicates the presence of isolated Zn2+ ions. Only in the Znrich samples (ZnO and Zn1Zr1Ox) the full absorption in the region of 250-370 nm is observed, indicating the presence of ZnO nanoparticles as demonstrated by the XRD results. Noticeable sulfate and silicate impurity was detected by both the skeletal IR and UV-Vis spectra (SI)37, which is attributed to the carbon black material employed. Formation of ZnZrOx composite oxide does not only modify the bulk properties; it also alters the surface chemistry drastically. To probe the modification of surface acidity/basicity by addition of ZnO to ZrO2, isopropanol (IPA)-TPD was conducted on various ZnZrOx composite oxides. Isopropanol is a good probe molecule for surface acidity/basicity evaluation, since dehydration to propene occurs on strong acidic sites and weak basic sites38, and dehydrogenation to acetone involves moderate acid and strong basic sites39-40. Despite some disagreement on assigning the strength of basic sites by IPATPD tests, it is well accepted that the strength of acidic sites is reflected by the apparent activation energy41 and temperature dependence42 of isopropanol to propene

ACS Paragon Plus Environment

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

conversion, and the number of sites determined by the amount of propene production43. In our study, IPA-TPD was performed on all the bare supports and on Au-added ZrO2. The results (Table 1 and Fig. S5) revealed that propene is the only product on all four Zr-containing oxides. On bulk ZnO, due to its low surface area, no obvious adsorption/desorption of isopropanol and its derivatives was observed. The amount of propene produced, shown in Table 1, indicates that the number of total acid sites slightly increases in Zn1Zr10Ox, and starts to decrease with the ZnO concentration in the composite oxides. The most affected is the low-temperature (165˚C peak) propene produced which totally disappears in Zn1Zr1Ox. Addition of ZnO thus alters the surface acidity of ZrO2 by passivating the strong Lewis acidic sites that are active in the temperature range of 150-250˚C. Addition of gold to the ZrO2 support, however, had the most pronounced effect, as it produced acetone exclusively over the temperature range of 100-200˚C, with only trace amounts of propene measured at 200-250˚C. Thus, addition of gold not only introduces a new set of mild acidic sites (single site Auδ+ centers) and basic sites (surrounding O- species) producing acetone exclusively, but also passivates the strong acid sites of ZrO2 within the whole low-temperature range. To identify the Lewis and Brønsted acid sites and the strength of each type of acidic sites, IR of pyridine (Py) adsorption was collected (Details of peak assignments described in SI). Figure 2 shows pyridine adsorption on the single and composite oxide samples with different pyridine desorption temperatures upon outgassing. Absorption bands of both pyridine and pyridinium ions at room temperature indicate the occurrence of both Lewis and Brønsted acid sites on ZrO2 (Fig. 2a). Disappearance of the pyridinium ion bands (~1540 cm-1) at elevated temperatures indicates no strong Brønsted acid sites, compared to the presence of pyridine absorption bands (~1450 and ~1610 cm-1), which indicates the occurrence of strong Lewis acid sites on ZrO2. In Fig. 2b, consistent presence of pyridine and pyridinium ion bands from room temperature to 500˚C suggests both strong Lewis acid sites and Brønsted acid sites on Zn1Zr10Ox. Compared to pure ZrO2, a slight shift up of the pyridine absorption bands was observed on Zn1Zr10Ox suggesting that Zn2+ centers are present over this catalyst, characterized by stronger Lewis acid strength than Zr4+ cations. With higher Zn loading the band position of pyridine adsorbed to Lewis acid sites remains the same at ~1612cm-1 for Zn1Zr5Ox (Fig. 2c), Zn1Zr1Ox (Fig. 2d) and ZnO (Fig. 2e); on the other hand, bands at ~1540cm-1, which indicates pyridinium ions adsorbed to Brønsted acid sites, are not present. The IR spectroscopy of adsorbed pyridine confirms the TPD of isopropanol showing the predominant acidic character of the ZrO2-containing samples, which changes to a more basic one by increasing the Zn content. It is worth noticing that sulfate impurities44 likely introduced by the carbon template method cause pyridine absorption bands at below 1400 cm-1 (SI).

Page 4 of 20

Fig. 2f. Together with the features due to pyridine adsorbed on Lewis acid sites (1606, 1579 and 1445 cm-1) no bands due to Brønsted acidity were detected (~1540 cm-1) in this case. This indicates that gold strongly modifies the surface acidity of the ZrO2 support, eliminating the Brønsted sites. It is likely that Au cations exchange the protons thus neutralizing them. This is in good agreement with the IPA-TPD suggesting that stronger acidic sites and in particular Brønsted acid sites are passivated by gold addition. No information can be extrapolated for the Lewis acidic sites coming from cationic Au, since the important bands coming the Lewis acidity of the support might hide the contribution due to those sites.

Figure 2. IR-Pyridine adsorption analysis on a) ZrO2, b) Zn1Zr10Ox, c) Zn1Zr5Ox, d) Zn1Zr1Ox f) 1%Au/ZrO2 with desorption temperature from RT to 500˚C and e) ZnO with desorption temperature from RT to 200˚C

Ethanol conversion on the single ZrO2, ZnO and on the ZnZrOx composite oxides was studied under temperature-programmed surface reaction mode (TPSR) to investigate the effect of surface acidity on the reaction pathways and the product distribution. Other than ethanol dehydrogenation, competing reactions include ethanol dehydration and C-C bond scission at intermediate temperatures: C2H5OH → C2H4 + H2O

(3)

C2H4O → CH4 + CO

(4)

The effect of the addition of gold to zirconia was also investigated by pyridine adsorption, as reported in

ACS Paragon Plus Environment

Page 5 of 20

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis that basic sites are the necessary sites for the further conversion of acetaldehyde to acetone through acetaldehyde coupling and aldol decarboxylation: 2C2H4O



CH3COCH2CH2OH

(5) Decrease of ethylene production and improved selectivity to acetaldehyde and acetone was observed with the suppression of surface acidity in the EtOH-TPSR tests. The same relationship between the surface acidity and ethanol conversion pathway has been previously reported on metal oxide supported copper catalysts22-23. Based on the IR findings (Fig. 2) weakening of the Lewis acidity correlates with the increase of ZnO loading. The Brønsted acidity, manifested by pyridinium ion bands at ~1540 cm-1, peaks on Zn1Zr10Ox and decreases at higher ZnO loadings.

Figure 3. 8%EtOH/He-TPSR (2%EtOH/He for Zn1Zr1Ox), total flow rate=50mL/min, T increases linearly 30-500˚C in 4h. 10%O2/He-TPO on samples post EtOH-TPSR, total flow rate= 50mL/min, T increases linearly 30-600˚C in 4.85h. Catalyst loading=100mg.

Ethanol-TPSR/MS data are shown in Figure 3. Temperature-programmed oxidation (TPO) on samples post EtOH-TPSR was conducted up to 600˚C with linear temperature ramping to investigate potential carbon or CHx fragments deposition during EtOH-TPSR. To eliminate the effects of molecularly adsorbed hydrocarbons after TPSR, mild heating treatment in He at 200˚C was performed prior to TPO. In EtOH-TPSR, ZrO2 and Zn1Zr10Ox both produced ethylene from ethanol dehydration (Rxn 3) as the predominant product, lighting off at ~300˚C. Full conversion of ethanol was reached at ~350˚C on both catalysts. Ethylene production starting from 350˚C carried on sustainably on ZrO2, while it decreased significantly on Zn1Zr10Ox. The decrease of ethylene may be attributed to the activation of ethanol dehydrogenation on Zn1Zr10Ox at temperatures ≥350˚C and a slightly larger amount of CHx deposition on Zn1Zr10Ox indicated by the production of H2O and CO2 by TPO of spent Zn1Zr10Ox. With increased Zn loading, acetaldehyde, the product of ethanol dehydrogenation, increased to comparable amounts to ethylene, lighting off at 300-350˚C on Zn1Zr5Ox and Zn1Zr1Ox. On ZnO neat, acetaldehyde was the dominant product, in agreement with Tanabe20-21 and Halawy’s45 findings. Acetone was detected on samples with higher basicity, such as Zn1Zr5Ox and ZnO, indicating

Figure 4. 2%EtOH+2%H2O-TPSR, total flow rate=50ml/min, temperature increases linearly from 30-500˚C in 4h. 10%O2/He-TPO on samples post TPSR, total flow rate=50mL/min, temperature increases linearly from 30500˚C in 4h. Catalyst loading=100mg.

Addition of steam to ethanol serves several purposes, one as an oxidant to facilitate the acetaldehyde to acetone conversion24; another role is to suppress the formation of ethylene by suppressing the ethanol dehydration reaction. Carbon and CHx fragments deposition on the catalysts may also be moderated due to the mild oxidizing atmosphere caused by adding steam. Carbon deposition from the Boudouart reaction 2CO ↔ CO2 + C was

ACS Paragon Plus Environment

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 20

analyzed by García and Laborde46 under various steam to ethanol (S/C) ratios, and it was calculated that carbon will form at any temperature without steam addition; while no carbon deposition will take place at temperatures ≥600K with H2O:EtOH ≥2:1. To investigate the effect of steam addition under dynamic conditions, a gas mixture of 2mol% ethanol and 2mol% water in He was used for TPSR followed by TPO of the spent catalyst with continuous MS monitoring (Fig. 4). Mild heat treatment in He at 200˚C was performed prior to TPO. Ethylene was still found to be the predominant product of EtOH+H2OTPSR on ZrO2 and Zn1Zr10Ox, lighting off at ~300˚C. Peaked at ~400˚C on both of these catalysts, ethylene started to decrease as temperature increased, and CO2 started to be produced at ~450˚C on Zn1Zr10Ox via the ethanol steam reforming reaction. Selectivity to acetaldehyde increased with ZnO content. On Zn1Zr5Ox and ZnO, the amount of acetaldehyde, peaking at 0.9% and 1.7% respectively, exceeded that of ethylene (400˚C, and more CO2 was produced via the ethanol steam reforming reaction. Another benefit of adding equal moles of steam to ethanol is the suppression of carbon and CHx formation, indicated by the TPO post EtOH+H2O-TPSR in Fig. 4. CHx species were oxidized to CO2 beginning at ~300˚C, which is the same temperature range for CO2 production on the spent sample from pure ethanol-TPSR. ZrO2 carried the highest amount of carbonaceous species, while all three ZnZrOx mixtures show the same amount of CO2 production from TPO (Fig. 4). Addition of steam in ethanol, therefore, plays a more prominent role in suppressing carbon and CHx deposition on the Brønsted acid sites.

To investigate the reaction pathway of ethanol on gold, we repeated the EtOH-TPSR on gold catalysts (Fig. 5) supported on the neat and composite oxides (Zn:Zr=1:10, 1:1). On the bare oxides, ethanol lights off at moderately high temperatures in the range of 300-350˚C, as shown in Figure 3. Addition of gold at 1wt.% loading remarkably lowers the light-off temperature to 200-250˚C on Au/ZrO2 (Fig. 5a) and Au/ZnZrOx (Fig. 5b, c), and to 300˚C on the Au/ZnO (Fig. 5d) with acetaldehyde and hydrogen as the products. Production of ethylene was activated at much higher temperatures (400-450˚C) on all samples, considerably higher than on the bare oxides. A wide temperature window (from 180- 350˚C) is now possible, over which to exclusively produce acetaldehyde and hydrogen from ethanol. A similar investigation of the effect of surface acidity on the product yields from ethanol steam reforming was conducted by Verykios and coworkers51-52 on Ni-based catalysts on a series of La2O3, La2O3/Al2O3 and Al2O3 supports, which have a wide distribution of surface acidity/basicity. EtOH+H2O-TPSR (H2O:EtOH=2%:1%) on 20%Ni-Al2O3 found that C-C bond scission is the predominant reaction at all temperatures in the range of 100-700˚C producing CO, CH4 and H2 as the main products. Ni on the basic support La2O3, however, favored the formation of acetaldehyde in a wider temperature window (170-300˚C), but C1 species and H2 were still the predominant products from 200˚C to 700˚C. Thus, the chemistry of gold identified here is specific to gold without catalyzing C-C bond scission and with a remarkable 100% selectivity to acetaldehyde achieved below 300˚C. Analogous to the production of formaldehyde from methanol dehydrogenation53, acetaldehyde is expected to adsorb linearly on the Au sites in η1configuration. Weak binding between the adsorbate and Au promotes desorption faster than decomposition, which explains why acetaldehyde is the predominant product, and why no C-C bond scission takes place to produce CO and CH4 (Rxn 4).

Although promising activity was shown, the selectivity to acetaldehyde and/or acetone of the ZnZrOx composite oxides as shown in Fig. 3-4 is not adequate for practical application. Indeed, there is an overlapping temperature range where both products are found, and more so the dehydration to ethylene complicates the picture. Clearly, one would have to add a minor amount of a metal to shift to lower temperatures the window over which acetaldehyde alone is produced. We recently employed this strategy successfully in the case of methanol reactions by adding small amounts of gold on the ZnZrOx surfaces29. We have explored this here for the ethanol reactions. As mentioned briefly in the Introduction, the choice of gold is a good one as long as atomic dispersions of gold are achieved on supports to boost activity in cases where single gold atoms (cations) are the active sites. This also maximizes the atomic efficiency of the gold and reduces cost. Catalysis of alcohol-coupling reactions has

From the ethanol-TPSR data of Fig. 5, gold catalysts supported on the composite oxides outperform the pure supports in terms of activity in the low-temperature region and ethylene suppression in the high-temperature region. As discussed earlier, addition of gold modifies the surface acidity of ZrO2 (Table1) and the selectivity to acetaldehyde shown in Fig. 5a was anticipated. The use of the composite ZnZrOx supports is advantageous in that it increases the activity of gold (Fig. 5b-c) by preserving a higher number of gold active sites than ZrO2 or ZnO neat. The extended temperature window (200-330˚C) for the exclusive production of acetaldehyde and hydrogen offered by the Au/ZnZrOx catalysts is promising for the practical application of these catalysts. Another reason for investigating gold preparation on the composite oxide supports is the atomic dispersion and stability of gold on ZnZrOx supports under reaction conditions. Gold catalysts synthesized by the anion adsorption method on

been demonstrated on atomic oxygen-dosed Au single crystal surfaces34, 47-48, and on single Au-Ox species on various supports29-31, 49-50; the latter exclusively for methanol.

ACS Paragon Plus Environment

Page 7 of 20

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

ZnZrOx with low and high Zn concentration (Zn:Zr=1:10, 1:1) are both able to carry and disperse 1 wt.% Au. ICP tests of the anion adsorption filtrate showed no gold precursor remaining in the solution, i.e., the designed 1wt.% loading was successfully achieved on the ZnZrOx composite oxides. The morphology of the as prepared (calcined in 2%O2/He at 300˚C) Au/ZnZrOx was imaged by aberration-corrected high-angle annular dark-field ac-HAADFSTEM (Fig. 6a-b). The as prepared gold catalysts, with atomically dispersed gold species on both Zn1Zr1Ox and Zn1Zr10Ox, also contain Au clusters and nanoparticles, complicating the identification of the active sites of gold. Therefore, NaCN-leaching treatment was performed on the as prepared gold catalysts (referred as “parent” samples) which effectively removes spectator gold species from oxide supports54-57. ac-HAADF-STEM images of the leached Au/ZnZrOx samples are shown in Fig. 6c-d, presenting only gold atoms after NaCN-leaching (see also Fig. S6). Mild reduction of Au catalysts had a negligible effect on the product distribution of the reaction. Au catalysts supported on Zn1Zr5Ox (Zn loading in between Zn1Zr10Ox and Zn1Zr1Ox, Table1) and pure ZnO were first calcined in 2%O2/He at 300˚C for 2h, then tested in ethanol-TPSR up to 500˚C; the other portion of calcined Au catalysts were reduced in 5%H2/He at 250˚C for 2h, then tested in ethanol-TPSR up to 500˚C. Same product distribution was observed on the two groups of samples (Fig. S8) Activity tests under steady-state conditions were conducted on the two selected gold catalysts, 1wt.% Au/Zn1Zr1Ox and 1wt.% Au/Zn1Zr10Ox, before and after NaCN-leaching (L-). A moderate amount of steam (EtOH:H2O =1:1) was added to the reaction gas mixture, and steady-state ethanol conversion tests were conducted in the temperature range of 275-400˚C with 75 min isothermal holding at each set temperature with online MS monitoring of the effluent gases. As shown in Fig. 7a, ≥93% conversion of ethanol was reached on all catalysts except L-Au/Zn1Zr1Ox (84%) at 400˚C. Acetaldehyde and acetone (above 320˚C) were the main products for all catalysts up to 400˚C through Rxn 1 and 2 (Fig. 7c-d), accompanied by the production of H2 (Fig. 7b). The yield of acetaldehyde increased on Au/Zn1Zr10Ox until 350˚C, and started to decrease at 400˚C, while it was maintained on the parent Au/Zn1Zr1Ox and continued to increase on LAu/Zn1Zr1Ox up to 400˚C, but from a much lower initial value. The yield of acetone on all four samples started to increase at ≥325˚C, and was higher on the sample with low Zn concentration. Acetaldehyde and acetone closed

the carbon balance to ~90% at 400˚C compared to the ethanol consumption. Production of H2 was sustained up to 400˚C and matches the production of acetaldehyde and acetone for each of the samples. The conversion rate of ethanol and production rate of acetaldehyde at 350˚C per catalyst weight are shown in Table 2.

Figure 5. 2%EtOH/He-TPSR on a) Au/ZrO2, b) Au/Zn1Zr10Ox, c) Au/ Zn1Zr1Ox and d) Au/ZnO, total flow rate=50mL/min, temperature increases linearly 30-500˚C in 4h. Catalyst loading=100mg, WHSV=1.11 gEtOH/(gCat.•h).

Table 2. Reaction rates and carbon deposition amount of parent and leached samples Sample EtOH conv. rate

a

CH3CHO prod. rate Residual Au % c

CHx amt.

b

L-Au/Zn1Zr1Ox

L-Au/Zn1Zr10Ox

Au/Zn1Zr1Ox

Au/Zn1Zr10Ox

3.04

6.04

5.53

6.52

2.28

5.28

5.19

5.79

54

58

-

-

1.53

3.60

1.52

2.49

ACS Paragon Plus Environment

ACS Catalysis a

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

-1

-1 b

Page 8 of 20 -1

-1

: Ethanol conversion rate at 350˚C in units of µmol•s •gcat ; : Acetaldehyde production rate at 350˚C in units of µmol•s •gcat ;

c

: CHx deposition amount on a carbon basis of samples post steady-state ethanol conversion up to 400˚C in units of mmol•gcat

-1

Au/Zn1Zr1Ox probably destabilizes the residual gold, more so in Zn1Zr1Ox than in Zn1Zr10Ox, as indicated by gold aggregation only on the spent L-Au/Zn1Zr1Ox but not on the spent L-Au/Zn1Zr10Ox after steady-state reaction up to 400˚C for 75min (Fig. 6e,f). TPO profiles of all the spent samples are shown in Fig. S7 with CHx deposition amount shown in Table 2. Similar CO2 production from TPO was found between the two sets of parent and leached samples, indicating the active gold sites for ethanol conversion are most likely also the sites for carbonaceous species deposition. Simultaneous production of CO2 and H2O centered at ~350˚C during TPO indicates CHx deposition as the carbon source, and excludes the formation of graphitic carbon as the carbon source, as this is oxidized at much higher temperatures (≥550˚C)58.

Figure 6. ac-HAADF-STEM of fresh parent samples a) Au/Zn1Zr10Ox, b) Au/Zn1Zr1Ox; fresh leached samples c) Au/Zn1Zr10Ox, d) Au/Zn1Zr1Ox; and spent leached samples e) Au/Zn1Zr10Ox, f) Au/Zn1Zr1Ox.

From the comparison of the parent and leached catalysts in Fig. 7, we see that catalyst activity and selectivity are maintained on the leached Au/Zn1Zr10Ox catalyst compared to the parent sample, which indicates that the residual strongly bound gold atoms on the leached sample are the active sites. On the other hand, considerably decreased was the activity of the leached Au/Zn1Zr1Ox. ICP tests of the Au/ZnZrOx leachates indicate ~50% of the original gold and 22% and 12% of Zn was leached away from 1wt.% Au/Zn1Zr10Ox and 1wt.% Au/Zn1Zr1Ox, respectively, while no dissolution of Zr took place for either sample (Table 2). Considering the high bulk Zn concentration in Zn1Zr1Ox, ~5 times more Zn was leached from 1wt.% Au/Zn1Zr1Ox than from 1wt.% Au/Zn1Zr10Ox. The larger amount of Zn dissolution from the parent Au/Zn1Zr1Ox sample very likely contributes to the catalyst activity loss after leaching. Isolated ZnO nanoparticles in Zn1Zr1Ox (11 nm avg. size, Table1) dissolve promptly in the NaCN solution, resulting in the loss of active gold sites bound to them. The dissolution of ZnO particles from

Figure 7. Steady-state 2%EtOH+2%H2O/He on leached and parent Au/ZnZrOx. Total flow rate=50mL/min, catalyst loading=100mg, WHSV=1.1 gEtOH/(gcat.•h). a) ethanol conversion, b) H2 production, c) acetaldehyde and d) acetone yield. — Au/Zn1Zr1Ox, ---Au/Zn1Zr10Ox, solid symbol: parent, open symbol: leached.

Stability testing of the Au/ZnZrOx catalysts at 325˚C was carried out for 11h on-stream (Fig. 8). The acetaldehyde yield was closely matched to the ethanol conversion, showing that no other side reactions occurred up to 325˚C on both catalysts. The inclusion of water up to 325˚C does not further convert ethanol to oxygenated products, such as acetic acid and ethyl acetate, which typically decrease the yield of acetaldehyde when ethanol and O2 are co-fed on supported gold catalysts16-17, 59. Moreover, this new catalyst may have much better longterm stability than commercial Cu/SiO2 catalysts for the dry ethanol dehydrogenation, which rapidly deactivate due to copper metal sintering.

ACS Paragon Plus Environment

Page 9 of 20

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis University for their assistance with sample characterization. G.G. acknowledges the University of Genoa for financial support to conduct research at Tufts University. Aberrationcorrected electron microscopy research at Oak Ridge National Laboratory was sponsored by the U. S. Department of Energy, Office of Energy Efficiency and Renewable Energy, Vehicle Technologies Office, Propulsion Materials Program.

REFERENCES

Figure 8. Stability tests (2%EtOH+2%H2O/He) on Au/Zn1Zr10Ox and Au/Zn1Zr1Ox, total flow rate=50mL/min, T=325˚C for 11h. Catalyst loading=100mg, WHSV=1.11 gEtOH/(gCat.•h).

CONCLUSIONS Ethanol conversion on various ZnZrOx supports occurs through different reaction pathways depending primarily on the surface acidity/basicity of the supports. Modification of ZrO2 by ZnO increases the selectivity to the ethanol dehydrogenation products: acetaldehyde and acetone. Moreover, ZnO incorporation into ZrO2 allows for many more anchoring points for metal addition than uncombined ZnO or ZrO2 nanoparticles. Addition of 1wt.% gold to ZrO2 or the composite supports has a significant effect of eliminating the Brønsted acid sites and forming acetaldehyde and hydrogen with 100% selectivity at temperatures below 300˚C. A wider temperature window down to 200˚C is realized by dispersing gold on the low-content Zn1Zr10Ox support due to a much better dispersion of gold on the latter. NaCN-leaching of the 1 wt.% gold catalysts has revealed that the optimal loading capacity of this ZnZrOx is ~0.5wt.% Au to achieve atomic gold dispersion. The stability of the new catalysts is very good at 325˚C, at which temperature steam reforming reaction products are still not present. Ample scope, therefore, exists for the design of practical gold catalysts for the low-temperature ethanol dehydrogenation exclusively to acetaldehyde and hydrogen, with potentially much better stability than the commercial Cu/SiO2 catalysts.60-62

ASSOCIATED CONTENT Supporting Information Experimental details, Figure S1-S8, Scheme S1. This material is available free of charge via the Internet at http://pubs.acs.org.

AUTHOR INFORMATION Corresponding Author Email: [email protected]

ACKNOWLEDGMENT We acknowledge the financial support of this work by the U.S. Department of Energy under Grant No. DE-FG0205ER15730. C.W. thanks Dr. Y. Zhang and Dr. C. Settens at the Center for Material Science and Engineering of MIT; and Dr. H. Lin at the Center for Nanoscale Systems of Harvard

(1) Cséfalvay, E.; Akien, G. R.; Qi, L.; Horváth, I. T., Catal. Today 2015, 239, 50-55. (2) Alex Marvin, W.; Schmidt, L. D.; Benjaafar, S.; Tiffany, D. G.; Daoutidis, P., Chem. Eng. Sci. 2012, 67, 68-79. (3) Deluga, G. A.; Salge, J. R.; Schmidt, L. D.; Verykios, X. E., Science 2004, 303, 993-997. (4) Huber, G. W.; Shabaker, J. W.; Dumesic, J. A., Science 2003, 300, 2075-2077. (5) Mattos, L. V.; Jacobs, G.; Davis, B. H.; Noronha, F. B., Chem. Rev. 2012, 112, 4094-4123. (6) Rossetti, I.; Biffi, C.; Bianchi, C. L.; Nichele, V.; Signoretto, M.; Menegazzo, F.; Finocchio, E.; Ramis, G.; Di Michele, A., Appl. Catal. B-ENVIRON 2012, 117-118, 384-396. (7) Garbarino, G.; Finocchio, E.; Lagazzo, A.; Valsamakis, I.; Riani, P.; Escribano, V. S.; Busca, G., Appl. Catal. B-ENVIRON 2014, 147, 813-826. (8) Wu, C.; Williams, P. T., Environ. Sci. Technol. 44, 59935998. (9) Garbarino, G.; Wang, C.; Valsamakis, I.; Chitsazan, S.; Riani, P.; Finocchio, E.; Flytzani-Stephanopoulos, M.; Busca, G., Appl. Catal. B-ENVIRON 2015, 174–175, 21-34. (10) Song, H.; Zhang, L.; Ozkan, U., Top. Catal. 2012, 55, 13241331. (11) Casanovas, A.; Roig, M.; de Leitenburg, C.; Trovarelli, A.; Llorca, J., Int. J. Hydrogen Energy 2010, 35, 7690-7698. (12) Garbarino, G.; Riani, P.; Lucchini, M. A.; Canepa, F.; Kawale, S.; Busca, G., Int. J. Hydrogen Energy 2013, 38, 82-91. (13) Rabenstein, G.; Hacker, V., J. Power Sources 2008, 185, 1293-1304. (14) Sun, J.; Wang, Y., ACS Catal. 2014, 4, 1078-1090. (15) Redina, E. A.; Greish, A. A.; Mishin, I. V.; Kapustin, G. I.; Tkachenko, O. P.; Kirichenko, O. A.; Kustov, L. M., Catal. Today 2015, 241, Part B, 246-254. (16) Guan, Y.; Hensen, E. J. M., J. Catal. 2013, 305, 135-145. (17) Guan, Y.; Hensen, E. J. M., Appl. Catal. A-GEN 2009, 361, 49-56. (18) Gazsi, A.; Koós, A.; Bánsági, T.; Solymosi, F., Catal. Today 2011, 160, 70-78. (19) Murthy, R. S.; Patnaik, P.; Sidheswaran, P.; Jayamani, M., J. Catal. 1988, 109, 298-302. (20) Nakajima, T.; Nameta, H.; Mishima, S.; Matsuzaki, I.; Tanabe, K., J. Mater. Chem. 1994, 4, 853-858. (21) Nakajima, T.; Tanabe, K.; Yamaguchi, T.; Matsuzaki, I.; Mishima, S., Appl. Catal. 1989, 52, 237-248. (22) Nishiguchi, T.; Matsumoto, T.; Kanai, H.; Utani, K.; Matsumura, Y.; Shen, W.-J.; Imamura, S., Appl. Catal. A-GEN 2005, 279, 273-277. (23) Rodrigues, C. P.; Zonetti, P. C.; Silva, C. G.; Gaspar, A. B.; Appel, L. G., Appl. Catal. A-GEN 2013, 458, 111-118. (24) Sun, J.; Zhu, K.; Gao, F.; Wang, C.; Liu, J.; Peden, C. H. F.; Wang, Y., J. Am. Chem. Soc. 2011, 133, 11096-11099. (25) Tago, T.; Konno, H.; Ikeda, S.; Yamazaki, S.; Ninomiya, W.; Nakasaka, Y.; Masuda, T., Catal. Today 2011, 164, 158-162. (26) Tago, T.; Konno, H.; Sakamoto, M.; Nakasaka, Y.; Masuda, T., Appl. Catal. A-GEN 2011, 403, 183-191. (27) Mavrikakis, M.; Barteau, M. A., J. Mol. Catal. A: Chem. 1998, 131, 135-147.

ACS Paragon Plus Environment

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(28) Wachs, I. E.; Madix, R. J., Surf. Sci. 1978, 76, 531-558. (29) Wang, C.; Boucher, M.; Yang, M.; Saltsburg, H.; FlytzaniStephanopoulos, M., Appl. Catal. B-ENVIRON 2014, 154-155, 142152. (30) Yi, N.; Boucher, M. B.; Gittleson, F.; Zugic, B.; Saltsburg, H.; Flytzani-Stephanopoulos, M., J. Phys. Chem. C 2011, 115, 12611268. (31) Boucher, M. B.; Goergen, S.; Yi, N.; FlytzaniStephanopoulos, M., PCCP 2011, 13, 2517-2527. (32) Wittstock, A.; Zielasek, V.; Biener, J.; Friend, C. M.; Bäumer, M., Science 2010, 327, 319-322. (33) Xu, B.; Haubrich, J.; Freyschlag, C. G.; Madix, R. J.; Friend, C. M., Chem. Sci. 2010, 1, 310-314. (34) Xu, B.; Haubrich, J.; Baker, T. A.; Kaxiras, E.; Friend, C. M., J. Phys. Chem. C 2011, 115, 3703-3708. (35) Lessard, J. D.; Valsamakis, I.; Flytzani-Stephanopoulos, M., Chem. Commun. 2012, 48, 4857-4859. (36) Dutkiewicz, J., JPE 1992, 13, 430-433. (37) Gillespie, R. J.; Robinson, E. A., Can. J. Chem. 1964, 42, 2496-2503. (38) Bezen, M. C. I.; Breitkopf, C.; El Kolli, N.; Krafft, J.-M.; Louis, C.; Lercher, J. A., Chem. Eur. J. 2011, 17, 7095-7104. (39) Gervasini, A.; Fenyvesi, J.; Auroux, A., Catal. Lett. 1997, 43, 219-228. (40) Manriquez, M. E.; Lopez, T.; Gomez, R.; Navarrete, J., J. Mol. Catal. A: Chem. 2004, 220, 229-237. (41) Gervasini, A.; Auroux, A., J. Catal. 1991, 131, 190-198. (42) Waugh, K. C.; Bowker, M.; Petts, R. W.; Vandervell, H. D.; O'Malley, J., Appl. Catal. 1986, 25, 121-128. (43) Iglesia, E.; Barton, D. G.; Biscardi, J. A.; Gines, M. J. L.; Soled, S. L., Catal. Today 1997, 38, 339-360. (44) Morterra, C.; Cerrato, G.; Pinna, F.; Meligrana, G., Top. Catal. 2001, 15, 53-61. (45) Halawy, S. A.; Mohamed, M. A., J. Mol. Catal. A: Chem. 1995, 98, L63-L68. (46) García, E. Y.; Laborde, M. A., Int. J. Hydrogen Energy 1991, 16, 307-312. (47) Xu, B.; Madix, R. J.; Friend, C. M., Acc. Chem. Res. 2014, 47, 761-772. (48) Kosuda, K. M.; Wittstock, A.; Friend, C. M.; Bäumer, M., Angew. Chem. Int. Ed. 2012, 51, 1698-1701. (49) Yi, N.; Si, R.; Saltsburg, H.; Flytzani-Stephanopoulos, M., Appl. Catal. B-ENVIRON 2010, 95, 87-92. (50) Yi, N.; Si, R.; Saltsburg, H.; Flytzani-Stephanopoulos, M., Energy Environ. Sci. 2010, 3, 831-837. (51) Fatsikostas, A. N.; Kondarides, D. I.; Verykios, X. E., Chem. Commun. 2001, 9, 851-852. (52) Fatsikostas, A. N.; Verykios, X. E., J. Catal. 2004, 225, 439452. (53) Boucher, M. B.; Marcinkowski, M. D.; Liriano, M. L.; Murphy, C. J.; Lewis, E. A.; Jewell, A. D.; Mattera, M. F. G.; Kyriakou, G.; Flytzani-Stephanopoulos, M.; Sykes, E. C. H., ACS Nano 2013, 7, 6181-6187. (54) Yang, M.; Li, S.; Wang, Y.; Herron, J. A.; Xu, Y.; Allard, L. F.; Lee, S.; Huang, J.; Mavrikakis, M.; Flytzani-Stephanopoulos, M., Science 2014, 346, 1498-1501. (55) Yang, M.; Allard, L. F.; Flytzani-Stephanopoulos, M., J. Am. Chem. Soc. 2013, 135, 3768-3771. (56) Fu, Q.; Saltsburg, H.; Flytzani-Stephanopoulos, M., Science 2003, 301, 935-938. (57) Fu, Q.; Deng, W.; Saltsburg, H.; Flytzani-Stephanopoulos, M., Appl. Catal. B-ENVIRON 2005, 56, 57-68. (58) Liu, B. S.; Au, C. T., Appl. Catal. A-GEN 2003, 244, 181-195. (59) Chen, H.; Jia, X.; Li, Y.; Liu, C.; Yang, Y., Catal. Today 2015, 256, Part 1, 153-160. (60) Chang, F.-W.; Kuo, W.-Y.; Lee, K.-C., Appl. Catal. A-GEN 2003, 246, 253-264.

Page 10 of 20

(61) Chang, F.-W.; Yang, H.-C.; Roselin, L. S.; Kuo, W.-Y., Appl. Catal. A-GEN 2006, 304, 30-39. (62) Tu, Y.-J.; Chen, Y.-W., Ind & Eng Chem Res 2001, 40, 5889-5893.

ACS Paragon Plus Environment

Page 11 of 20

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

Insert Table of Contents artwork here

ACS Paragon Plus Environment

11

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 1. XP spectra of Zn 2p and Zr 3d regions of ZnO, ZrO2 and ZnZrOx nano structures 84x60mm (300 x 300 DPI)

ACS Paragon Plus Environment

Page 12 of 20

Page 13 of 20

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

Figure 2. IR-Pyridine adsorption analysis on a) ZrO2, b) Zn1Zr10Ox, c) Zn1Zr5Ox, d) Zn1Zr1Ox f) 1%Au/ZrO2 with desorption temperature from RT to 500˚C and e) ZnO with desorption temperature from RT to 200˚C 105x131mm (300 x 300 DPI)

ACS Paragon Plus Environment

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 3. 8%EtOH/He-TPSR (2%EtOH/He for Zn1Zr1Ox), total flow rate=50mL/min, T increases linearly 30500˚C in 4h. 10%O2/He-TPO on samples post EtOH-TPSR, total flow rate= 50mL/min, T in-creases linearly 30-600˚C in 4.85h. Catalyst loading=100mg. 115x156mm (600 x 600 DPI)

ACS Paragon Plus Environment

Page 14 of 20

Page 15 of 20

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

Figure 4. 2%EtOH+2%H2O-TPSR, total flow rate=50ml/min, temperature increases linearly from 30-500˚C in 4h. 10%O2/He-TPO on samples post TPSR, total flow rate=50mL/min, temperature increases linearly from 30-500˚C in 4h. Catalyst loading=100mg. 122x177mm (600 x 600 DPI)

ACS Paragon Plus Environment

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 5. 2%EtOH/He-TPSR on a) Au/ZrO2, b) Au/Zn1Zr10Ox, c) Au/Zn1Zr1Ox and d) Au/ZnO, total flow rate=50mL/min, temperature increases linearly 30-500˚C in 4h. Catalyst loading=100mg, WHSV=1.11 gEtOH/(gCat.•h). 84x166mm (300 x 300 DPI)

ACS Paragon Plus Environment

Page 16 of 20

Page 17 of 20

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

Figure 6. ac-HAADF-STEM of fresh parent samples a) Au/Zn1Zr10Ox, b) Au/Zn1Zr1Ox; fresh leached samples c) Au/Zn1Zr10Ox, d) Au/Zn1Zr1Ox; and spent leached samples e) Au/Zn1Zr10Ox, f) Au/Zn1Zr1Ox. 126x190mm (300 x 300 DPI)

ACS Paragon Plus Environment

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 7. Steady-state 2%EtOH+2%H2O/He on leached and parent Au/ZnZrOx. Total flow rate=50mL/min, catalyst loading=100mg, WHSV=1.1 gEtOH/(gcat.•h). a) ethanol conversion, b) H2 production, c) acetaldehyde and d) acetone yield. —Au/Zn1Zr1Ox, ---Au/Zn1Zr10Ox, solid symbol: parent, open symbol: leached. 77x70mm (300 x 300 DPI)

ACS Paragon Plus Environment

Page 18 of 20

Page 19 of 20

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

Figure 8. Stability tests (2%EtOH+2%H2O/He) on Au/Zn1Zr10Ox and Au/Zn1Zr1Ox, total flow rate=50mL/min, T=325˚C for 11h. Catalyst loading=100mg, WHSV=1.11 gEtOH/(gCat.•h). 43x21mm (600 x 600 DPI)

ACS Paragon Plus Environment

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

table of content artwork 143x108mm (150 x 150 DPI)

ACS Paragon Plus Environment

Page 20 of 20