One-Pot Production of Lactic Acid from Acetol over Dealuminated Sn-Beta Supported Gold Catalyst

Article abstract of DOI:10.1021/acscatal.7b01499

Staged biomass pyrolysis or condensation of the pyrolysis vapor generates less complex and concentrated liquid product streams, which provide a very promising route to produce value-added chemicals. Acetol is one of the major light oxygenates in fractionated bio-oil. Here we report an efficient one-pot conversion of acetol to lactic acid, a key platform chemical and an important monomer of biodegradable polymer, in aqueous phase over a bifunctional dealuminated Sn-Beta supported gold catalyst. Under relatively mild conditions (393 K, 1.9 MPa N₂ + 0.1 MPa O₂), 93.4% acetol conversion and 73.2% selectivity toward lactic acid were achieved. On the basis of measured rate constants, the reaction pathway likely proceeds via kinetically relevant oxidation of acetol to form pyruvaldehyde, followed by hydration and 1,2-hydride shift (internal Meerwein-Ponndorf-Verley reduction and Oppenauer oxidation) on Sn-Lewis sites to produce lactic acid.

Full text of DOI:10.1021/acscatal.7b01499

Subscriber access provided by BOSTON UNIV
Article
One-pot Production of Lactic Acid from Acetol over
Dealuminated Tin Beta Supported Gold Catalyst
Yan Wan, Mengqi Zhuang, Shaopeng Chen, Wenda Hu,
Jie Sun, Jingdong Lin, Shaolong Wan, and Yong Wang
ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.7b01499 • Publication Date (Web): 24 Jul 2017
Downloaded from http://pubs.acs.org on July 28, 2017
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 31
ACS Catalysis
1
2
3
4
5
6
7
8
9
1
Oneꢀpot Production of Lactic Acid from Acetol over
Dealuminated Tin Beta Supported Gold Catalyst
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
2
†
†
†
†
†
†,‡
3
4
Yan Wan , Mengqi Zhuang , Shaopeng Chen , Wenda Hu , Jie Sun , Jingdong Lin , Shaolong
*
,†,
‖
,
‡
*,
†
,
‖
,
,
‡ §
Wan
, Yong Wang
†
5
College of Chemistry and Chemical Engineering, Xiamen University, Xiamen 361005, China
‖
6
7
Collaborative Innovation Center of Chemistry for Energy Materials, Xiamen University,
Xiamen 361005, China
‡
8
9
National Engineering Laboratory for Green Chemical Productions of AlcoholsꢀEthersꢀEsters,
Xiamen University, Xiamen 361005, China
§
10
Voiland School of Chemical Engineering and Bioengineering, Washington State University,
Pullman, WA 99164, United States
1
1
2
1
KEYWORDS: acetol, bifunctional catalyst, biomass, lactic acid, tin Beta
13
14
ABSTRACT
1
5
6
Staged biomass pyrolysis or condensation of the pyrolysis vapor generates less complex and
concentrated liquid product streams, which provides a very promising route to produce
1
ACS Paragon Plus Environment
1

ACS Catalysis
Page 2 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
2
3
4
5
6
7
8
9
valueꢀadded chemicals. Acetol is one of major light oxygenates in those fractionated bioꢀoil.
Here we report an efficient oneꢀpot conversion of acetol to lactic acid, a key platform chemical
and an important monomer to biodegradable polymer, in aqueous phase over a bifunctional
dealuminated tin Beta supported gold catalyst. Under relatively mild conditions (393 K, 1.9 MPa
N + 0.1 MPa O ), 93.4% acetol conversion and 73.2% selectivity towards lactic acid (LA) were
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
2
achieved. Based on measured rate constants, reaction pathway likely proceeds via kinetically
relevant oxidation of acetol to form pyruvaldehyde (PyA), followed by hydration and
1,2ꢀhydride shift (internal MeerweinꢀPonndorfꢀVerley reduction and Oppenauer oxidation) on
SnꢀLewis sites to produce lactic acid.
1
1
0
1
INTRODUCTION
12
Lignocellulose biomass is being studied worldwide as a feedstock for biofuels or valueꢀadded
1ꢀ3
1
1
1
1
3
4
5
6
chemicals due to its abundance and renewability. Among the different biomass conversion
4
ꢀ7
processes, pyrolysis coupled with catalytic upgrading is deemed as one of the most attractive.
However, the liquid pyrolysis oil consists of numerous different oxygenated molecules. This
exerts major obstacles for its catalytic upgrading to transportation fuels, in which oxygen needs
8
17
to be effectively removed without sacrificing C and H efficiency. Alternatively, a feasible
1
8
9
stratagem is to selectively convert the light oxygenates, which mainly accounts for high oxygen
content in pyrolysis oil, to valueꢀadded chemicals. Such strategy is especially appealing when
advanced bifunctional catalysts are coupled with a relatively simple cut of products that can be
readily obtained from abundant lignocellulosic biomass. Santhanaraj et al. reported a novel
1
20
2
1
2
9
2
method to produce gluconic acid from levoglucosan, which can be obtained in a high
ACS Paragon Plus Environment
2

Page 3 of 31
ACS Catalysis
1
2
3
1
4
10
concentration from staged condensation of biomass pyrolysis. Similarly, acetol in a
5
2
3
4
5
concentrated solution can be readily produced through staged pyrolysis of biomass that is
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
1
conducted at relatively low temperature (473 K–573 K). However, acetol is too reactive and is
a culprit of poor stability of bioꢀoil. Thus it’s highly desired to selectively convert acetol to stable
and valueꢀadded compounds such as lactic acid.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
6
7
8
Scheme 1. Acetol, readily produced from staged condensation or pyrolysis of raw biomass
directly, can be converted to lactic acid, a valuable chemical.
9
Lactic acid is widely used in the food, pharmaceutical, cosmetic and chemical industries.
Particular interest has been gained for use as a monomer in the production of biodegradable
10
1
2ꢀ14
1
1
2
plastics.
It is also considered as an important buildingꢀblock chemical, from which various
1
bulk chemicals such as alkyl lactate, pyruvic acid, acrylic acid and propanediol can be
1
2
13
produced. Currently lactic acid is primarily produced by the fermentation of glucose, which is
limited by the efficiency of biological process and generates a large amount of waste. Owing to
the increasing demand, much attention has been recently paid to the development of
thermochemical conversion processes using renewable feedstock.
1
1
1
1
4
5
6
7
To date, much efforts have been focused on the catalytic conversion of different carbohydrates
as feedstock. Esposito et al. found that barium hydroxide can readily convert glucose to LA with
18
1
4
1
9
0
up to 57% yield under hydrothermal conditions. Wang et al. used ballꢀmilled cellulose as a feed
1
5
2
and achieved a 68% yield of LA using Pb(Ⅱ) ion in water. For these processes, the recovery of
ACS Paragon Plus Environment
3

ACS Catalysis
Page 4 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2+
2+
1
2
3
4
5
6
7
8
9
0
homogeneous catalysts and the toxicity of those metal ions (e.g. Ba , Pb ) may prevent their
practical application. Holm et al. first reported the conversion of sucrose to methyl lactate with
1
6
up to 68% yield using SnꢀBeta zeolite. Alkali ions were found to have a promoting effect over
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
17
the selectivity of methyl lactate from sugars on stannosilicates catalyst. Wang et al. reported
that lanthanide triflates were efficient catalysts for the hydrothermal conversion of cellulose to
lactic acid, where 89.6% yield of lactic acid was obtained upon using Er(OTf) catalyst at 513
3
1
8
K. Orazov et al. combined MoO (a distinct retroꢀaldol catalyst) with SnꢀMFI (a 1,2ꢀhydride
3
1
9
shift catalyst), which enabled alkyl lactate formation from fructose at around 373 K. However,
most of these works are based on C6 sugar monomers, dimer like sucrose or polysaccharide like
cellulose, which require costly preꢀtreatment processes for raw biomass feedstocks.
1
11
Another widely studied route is the conversion of trioses such as dihydroxyacetone (DHA) and
glyceraldehyde (GLA), which can be generated from fermentation and catalytic oxidation of
glycerol respectively. Heeres et al. applied metal salts on the conversion of trioses to LA in water
1
1
1
2
3
4
2
0
under mild condition, with Al(Ⅲ) counterpart identified as the most promising. Solid acid
1
1
1
5
6
7
catalysts like commercially available zeolites have also been investigated. The selectivity to
lactic acid was 22% and 71% over HꢀZSMꢀ5 (Si/Al = 25) and HꢀUSY (Si/Al = 6) respectively,
21ꢀ22
when applied in aqueous GLA and DHA solutions.
ZSMꢀ5 zeolites desilicated in solutions of
18
alkali metal hydroxides can further enhance the aqueousꢀphase isomerization of DHA with >90%
2
3
1
9
0
selectivity to LA. Cho et al. developed a baseꢀfree oneꢀpot reaction to produce LA in high yield
from glycerol over a bifunctional Pt/SnꢀMFI catalyst. In this cascade reaction route, Pt
nanoparticles first catalyze the oxidation of glycerol to GLA, which is then converted to the final
2
21
24
2
2
3
product LA via SnꢀMFI with strong Lewis acidity and medium/weak Brønsted counterpart. It is
2
generally believed that both Brønsted and Lewis acid sites are critical to these trioses’
ACS Paragon Plus Environment
4

Page 5 of 31
ACS Catalysis
1
2
3
1
4
transformation toward lactic acid. Nevertheless, due to the limited amount and uncertainty in the
supply currently of trioses, this route is considered as an approach to better utilize glycerol
byproduct in bioꢀdiesel production rather than a replacement of lactic acid production currently
practiced on a large scale.
5
2
3
4
5
6
7
8
9
0
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Here, we try to develop a novel catalytic approach to produce LA from acetol, which can be
readily obtained from staged condensation or pyrolysis of abundant lignocellulosic biomass. Not
much work has yet been reported on this challenging reaction. Albuquerque et al. reported that
60
% yield of lactic acid was achieved from acetol using Pt/Al O catalyst under strong alkaline
2
3
1
conditions (pH = 10), which resulted in the salt of lactate and would need further treatment to
2
5
11
obtain LA, making the process less attractive. Although MgꢀAl layered double hydroxides
12
13
14
15
(LDH) and their derived 3D mixed oxides were exploited as basic solid catalyst to replace the
2
6
strong base additive in their following research, extensive leaching of magnesium from the
base catalyst occurred inevitably due to the chemical interaction with the formed lactic acid. The
favored reaction pathway of acetol under base condition is postulated to mainly proceed with
oxidation (to pyruvaldehyde) followed by intramolecular Cannizzaro mechanism to form lactic
acid, which is inevitably to react with the base or the base support, causing the issues mentioned
above. Therefore, the development of stable heterogeneous catalysts for the conversion of acetol
to lactic acid under baseꢀfree conditions is highly desired.
16
1
7
8
1
19
20
21
22
23
Considering the proposed catalytic conversion, an alternative reaction route under base free
can be applied through the integration of a selective oxidation catalyst coupled with a Lewis acid
catalyst for the 1,2ꢀhydride shift, as illustrated in Scheme 2. The hydrothermally synthesized
SnꢀBeta, featured as excellent water/acidꢀtolerant Lewis acid catalyst, can efficiently catalyze the
ACS Paragon Plus Environment
5

ACS Catalysis
Page 6 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
27ꢀ28
1
2
3
4
5
6
7
8
9
0
1,2ꢀhydride shift,
one of the key steps mentioned above. Recent studies have shown that
SnꢀBeta prepared via postꢀtreatment of dealuminated Beta zeolite also exhibits good catalytic
29ꢀ31
performance over this same reaction in aqueous solution.
Au is known to be an efficient
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
catalyst for aqueous oxidation and is more resistant to oxygen poisoning compared to Ptꢀgroup
3
2ꢀ36
metals.
Herein we report that acetol can be efficiently converted to LA in aqueous phase over
bifunctional dealuminated tin Beta supported Au nanoparticles catalyst, which combines
selective oxidation function with Lewis acidity for the 1,2ꢀhydride shift. The possible reaction
pathway under baseꢀfree condition is also proposed.
EXPERIMENT SECTION
1
2.1 Chemicals
11
Acetol (95% purity), pyruvic acid (98% purity), pyruvaldehyde (80% purity) and lactic acid
(85
1
1
1
2
3
4
% purity) were purchased from Alfa Aesar. Lactaldehyde (1 M in H O), poly (vinyialcohol) and
2
tin (Ⅱ) acetate (98% purity) were purchased from Sigma Aldrich. HꢀBeta zeolite (Si/Al = 12.5,
1
1
1
5
6
7
acetic acid (98% purity) and tetrachloroauric acid (HAuCl , 99.9%) were purchased from
4
Sinopec Co.
2.2 Catalyst preparation
3
1
18
Dealuminated tin Beta was prepared using the method previously reported. Briefly,
1
2
2
9
0
1
commercial HꢀBeta zeolite was first dealuminated by treatment in 13 M HNO solution at 373 K
3
ꢀ1
for 20 h (20ml g (zeolite)). The solid was recovered by filtration and washed repeatedly, then
dried overnight. Solidꢀstate ionꢀexchange was performed by grinding 0.604 g tin(Ⅱ)acetate with
2
2
2.7 g dealuminated Beta zeolite for 1 h, followed by calcination at 823 K in flowing air for 10 h.
Next, two successive steps of sodium ion exchanges were performed by stirring the obtained
23
ACS Paragon Plus Environment
6

Page 7 of 31
ACS Catalysis
1
2
3
1
4
dealuminated tin Beta in 0.2 M NaNO solution. The obtained sample was denoted as
3
5
2
3
4
5
6
7
8
9
0
DeAlꢀSnꢀBeta. In addition, the dealuminated Beta with two successive steps of sodium ion
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
exchanges sample was denoted as DeAlꢀBeta. SnO /DeAlꢀBeta was prepared via mechanical
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
mixing of DeAlꢀBeta with SnO , followed by calcination at 823 K for 12 h.
2
3
7ꢀ38
Au nanoparticles were loaded on the support by using solꢀimmobilization method.
In brief,
Au NPs stabilized with poly (vinyialcohol) (PVA) were first synthesized by adding of 1.31 ml
2wt
% PVA (MW = 10000) aqueous solution into 0.71 g HAuCl solution (diluted in 65 ml H O); a
4
2
freshly prepared 0.52 ml NaBH solution (0.1 M) was then added dropwise into the mixture
4
1
under vigorous stirring. After 10 minutes, the pH of the generated sol was adjusted to 1.5 with
diluted H SO . Maintaining the suspension under vigorous stirring, 0.97 g prepared
11
2
4
12
13
14
15
DeAlꢀSnꢀBeta support was added to make the sol immobilized. After 2 h, the solid was
recovered by filtration and washed with hot distilled water to remove PVA stabilizer. The
resultant product was dried overnight. The final obtained catalyst was denoted as
Au/DeAlꢀSnꢀBeta.
16
2.3 Characterization
1
7
8
Xꢀray diffraction (XRD) patterns of samples were recorded on Rigaku UItima IV Xꢀray
diffractometer with Cu Kα radiation (35 kv and 15 mA) from 5° to 70° with a scan speed of 2θ =
10.0 °/min. The chemical compositions were determined by inductively coupled plasmaꢀoptical
emission spectrometry (ICP–OES) on a Thermo IRIS Intrepid II XSP atomic emission
spectrometer. Sn content was determined by energy dispersive Xꢀray spectroscopy (EDS)
analysis, operated on a Hitachi Sꢀ4800 SEM microscopy with a Bruker Quantax 400 detector.
The surface areas and pore volumes of the calcined samples were measured by means of nitrogen
1
19
20
21
22
23
ACS Paragon Plus Environment
7

ACS Catalysis
Page 8 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
2
3
4
5
6
7
8
9
0
physisorption on a Micromeritics ASAP 2020M instrument. Before the adsorption, samples were
evacuated at 573 K for 3 h. The total surface area was calculated via the
Brunauer−Emmett−Teller (BET) equation, and the microporous pore volume was determined
using the tꢀplot method. Transmission electron microscopy (TEM) measurements were
performed on a Phillips Analytical FEI Tecnai 30 electron microscope operated at an
acceleration voltage of 300 kV. Diffuse reflectance ultraviolet−visible (UV−vis) spectra of
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
dehydrated samples were recorded using BaSO as a reference in the region of 200−700 nm on a
4
Varian Cary 5000 UV−vis spectrophotometer. Samples were measured after drying overnight at
473 K. Xꢀray photoelectron spectra (XPS) were recorded with a Quantum 2000 Scanning ESCA
1
Microprob instrument (Physical Electronics) using Al K radiation. The binding energy was
α
11
calibrated by using the C photoelectron peak at 284.4 eV as a reference. Fourier transform
1
s
12
13
14
15
infrared (FTIR) spectra were collected on the Nicolet 6700 instrument equipped with an MCT
ꢀ1
detector at a spectral resolution of 4 cm . A selfꢀsupporting pellet made of the sample was
placed in the flow cell and evacuated under reduced pressure at 673 K for 2 h. After cooling to
423 K, the samples were saturated with pyridine vapor and then evacuated. Subsequently spectra
−
1
16
were recorded at the same temperature in the 4000−650 cm range by using coaddition of 32
1
7
8
scans. NH temperatureꢀprogrammed desorption (NH –TPD) was carried out on a homemade
3
3
1
instrument. Typically, the sample was loaded and pretreated in a quartz reactor with highꢀpurity
He (99.999%) flow at 573 K for 2 h, followed by cooling down to 373 K. The adsorption of NH3
was performed at 373 K in an NH –He mixture (5 vol% NH ). Then, the temperature was raised
19
20
21
22
23
3
3
ꢀ1
ꢀ1
to 1173 K at a rate of 10 K min . The desorbed NH was absorbed by 0.01 mol L HCl solution,
3
ꢀ1
and quantified via a counter titration with 0.01 mol L NaOH solution in the presence of mixed
indicator (0.1% bromocresol green and 0.2% methyl red). The number of framework Sn titrated
ACS Paragon Plus Environment
8

Page 9 of 31
ACS Catalysis
1
2
3
1
4
by ammonia molecular was estimated from values using the following equation: Sn(Framework) sites
5
ꢀ1
2
density (ꢁmol g ) = n
⁄m = (c V − c V )⁄m (Eq. 1)
(NH3) 1 2 2
1
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
ꢀ1
3
4
where c is the concentration of HCl solution (mmol L ), V is the volume of HCl solution (ml);
1 1
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ꢀ1
c is the concentration of NaOH solution (mmol L ), V is the volume of NaOH solution (ml)
2
2
5
and m is the mass of catalyst (g).
6
2.4 Catalytic test and products analysis
7
The conversion of acetol and possible reaction intermediates was carried out in a batchꢀtype
Teflonꢀlined stainlessꢀsteel autoclave reactor with an inner volume of 60 ml. Typically, acetol
was dissolved in 20 ml deionized water, in which the catalyst was added subsequently. Then the
8
9
10
autoclave was sealed, purged with N to remove the remaining air in the reaction system,
2
1
1
2
followed by the introduction of O to achieve a certain pressure. The system was then further
2
1
pressurized with N and the final pressure was adjusted to 2 MPa. Then it was heated up to 373
2
13
K−413 K. After continuous magnetic stirring for a given time, autoclave was cooled to room
temperature in cold water, and the catalyst was separated by filtration. The liquid products were
analyzed by HPLC (Angilent 1260series) equipped with a refractive index (RI) detector and
ultraviolet (UV) detector. An Aminex HPX−87H column (BioꢀRad Laboratories Inc.) was
1
1
1
1
1
1
2
4
5
6
7
8
9
0
ꢀ1
employed, where 0.005 M H SO aqueous solution with a flow rate of 0.6 ml min was used as
2
4
mobile phase. The conversion of the substrate and the selectivity to a specific product were
defined as the molar percentage and calculated on a carbon basis. Recycling tests were
performed to examine the catalyst stability. After each run of reaction, the residual catalyst
powder was washed by five cycles of successive addition of deionized water, centrifugation and
removal of the liquid, and then dried overnight in the oven at 378 K.
21
2
2
2
3
2.5 Modeling of the reaction kinetics
ACS Paragon Plus Environment
9

ACS Catalysis
Page 10 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
2
3
A simple pseudoꢀhomogeneous kinetic model kinetic model was developed based on the
pathway proposed in Scheme 2 of Section 3.4. The detailed equations are listed as follows,
considering firstꢀorder reactions for all steps involved:
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
V dc(acetol)
4
5
6
7
= − k1cꢀacetolꢁ − k c
+ k c
(Eq. 2)
3 ꢀacetolꢁꢂ
4 ꢀLaAꢁ
m
dt
V dc(PyA)
= k1cꢀacetolꢁ − k c
(Eq. 3)
2 ꢀPyAꢁ
m
dt
V dc(LaA)
= k3cꢀacetolꢁ − k c
− k c
(Eq. 4)
4 ꢀLaAꢁ
5 ꢀLaAꢁ
m
dt
V dc(LA)
= k2cꢀPyAꢁ + k c
dt
− k c
(Eq. 5)
5 ꢀLaAꢁ
6 ꢀLAꢁ
m
8
9
0
where V is the reaction mixture volume (L), m is the mass of catalyst (g), and c is the molar
concentration of the reactive species. The concentrations of substrate, intermediates and products
over the time course were used to simultaneously fit the experimental results. The fitting was
solved by nonlinear leastꢀsquare regression analysis of the differences between the experimental
values and the kinetic model calculated values. Several constraints were used to ensure that the
fitting resulted in the global minimum and values with reasonable physical meanings. For
example, the values for the initial guesses towards rate constants (k –k ) in the fitting routine
1
11
1
2
3
1
14
1
6
1
1
5
6
were obtained by using all the intermediate compounds as the feed, other than just acetol.
3.1 Catalyst characterization
1
7
ACS Paragon Plus Environment
10

Page 11 of 31
ACS Catalysis
1
2
3
1
4
Figure 1. XRD patterns of the samples: (a) Beta; (b) DeAlꢀBeta; (c) DeAlꢀSnꢀBeta; (d) 0.28wt%
5
2
Au/DeAlꢀSnꢀBeta.
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
3
4
5
6
7
8
The XRD results in Figure 1 confirmed that the topology of Beta zeolite was well preserved
for the prepared DeAlꢀSnꢀBeta, which indicated no destruction of zeolite framework during
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
9, 31, 39
dealumination, incorporation of Sn and calcination, as consistent with earlier reports.
The
further loading of Au nanoparticles on DeAlꢀSnꢀBeta does not influence the crystalline structure.
A small amount of SnO species exists, evidenced by occurrence of the reflection peak at 26.6°
2
，33.9°and 51.8°, assigned to crystalline SnO . Table 1 summarizes the physicochemical
2
9
properties of HꢀBeta, DeAlꢀBeta and DeAlꢀSnꢀBeta. The BET surface area and micropore
10
volume of DeAlꢀSnꢀBeta zeolite was slightly decreased after delumination and incorporation of
Sn.
1
1
2
a
1
Table 1. Physicochemical properties of the materials
SBET
Vmicro
b
c
b
Catalyst
SiO /Al O
3
Sn (wt%)
Na (wt%)
2
2
2
ꢀ1
ꢀ3 ꢀ1
(m g ) (cm g )
Beta
476
456
0.13
0.11
0.10
25
ꢀ
ꢀ
DeAlꢀBeta
990
990
ꢀ
0.19
0.21
DeAlꢀSnꢀBeta 429
9.6
a
b
1
1
3
4
SBET = Brunauer−Emmet−Teller surface area, Vmicro = micropore volume. Determined by
c
ICPꢀOES. Determined by SEMꢀEDX
1
1
1
1
5
6
7
8
Pyridine probed FTIR analysis could readily distinguish the Brønsted and Lewis acid sites,
ꢀ1
ꢀ1
characterized by bands in the FTIR spectrum from 1512 to 1567 cm and 1423 to 1472 cm ,
respectively. As shown in Figure 2, almost all the acid sites were removed for DeAlꢀBeta,
ascribed to dealumination with concentrated HNO . The subsequent incorporation of Sn
3
ACS Paragon Plus Environment
11

ACS Catalysis
Page 12 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
2
3
4
5
6
7
8
9
significantly enhances the Lewis acidity of DeAlꢀSnꢀBeta, as confirmed by a strong absorption
ꢀ1
24, 29, 40ꢀ41
band at 1450 cm due to the pyridine bound with the Lewis acid Sn sites.
Diffuse reflectance UV−vis analysis was employed to detect the coordination states of
incorporated Sn species. Figure 3 illustrated the UV−vis spectra of the dehydrated DeAlꢀBeta,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
DeAlꢀSnꢀBeta and SnO /DeAlꢀBeta. For DeAlꢀSnꢀBeta, the absorption band centered at ca.206
2
2
ꢀ
4+
nm, arising from the ligandꢀtoꢀmetal charge transfer from O to Sn , is generally assigned to
isolated Sn species in the framework with unsaturated tetracoordinated coordination exhibiting
3
9ꢀ40
excellent waterꢀtolerant Lewis acidity.
This result indicates that Sn species has been
successfully incorporated into the framework of Beta via postꢀtreatment. The other band centered
10
1
1
Figure 2. Pyridine probed FTIR spectra for (a) DeAlꢀBeta; (b) DeAlꢀSnꢀBeta.
1
2
ACS Paragon Plus Environment
12

Page 13 of 31
ACS Catalysis
1
2
3
1
4
Figure 3. UV−vis spectra of (a) DeAlꢀBeta, (b) DeAlꢀSnꢀBeta and (c) SnO /DeAlꢀBeta.
2
5
6
39ꢀ40
2
at ca.270 nm has previously been assigned to hexacoordinated tin（Ⅳ）oxide,
in line with
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
24, 39ꢀ42
3
4
5
6
7
8
SnO /DeAlꢀBeta sample, which are not expected to contribute to the Lewis acidity.
As
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
shown in Figure 4, the parent Beta zeolite exhibits the presence of some external silanols,
ꢀ1
ꢀ1
featured at the adsorption at 3740 cm . However, a vibrational signal at 3733 cm is strikingly
enhanced in the spectra of DeAlꢀBeta, which is attributed to the generation of more silanol
groups (forming so called silanol nests) as more defect sites created in the framework through
the acid treatment. Solidꢀstate ionꢀexchange with tin(Ⅱ) acetate leads to an obvious decrease in
ꢀ1
9
0
1
the intensity of OH band at 3733 cm on DeAlꢀSnꢀBeta, suggesting that tin has been
1
1
incorporated into the dealuminated framework, resulting in the removal of those defect sites and
30, 39
the decrease of attached silanol groups.
1
1
1
2
3
4
Figure 4. FTIR spectra of (a) Beta; (b) DeAlꢀBeta; and (c) DeAlꢀSnꢀBeta.
The amounts of framework Sn species and extraꢀframework Sn species were further
determined quantitatively. Based on the ammonia TPD and titration results, the total acid
15
ꢀ1
1
6
concentration of DeAlꢀSnꢀBeta is 505 ꢁmol g . Since there is negligible Brønsted acid sites on
ACS Paragon Plus Environment
13

ACS Catalysis
Page 14 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
2
3
4
5
6
7
8
the DeAlꢀSnꢀBeta as indicated from the pyridine probed FTIR spectra, we can assume that the
total acidity measured can be attributed exclusively to Lewis acid sites. From previous reports,
3
0, 42
extraꢀframework Sn species (SnO phase) are generally considered as nonꢀLewis acid.
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Assuming equimolar adsorption of ammonia per Sn site, the measured amount of acidity is
equivalent to the framework Sn of 6.0wt%. Since total amount of Sn species determined by
SEMꢀEDX is 9.6wt% as shown in Table 2, the amount of extraꢀframework Sn species should be
around 3.6wt%.
TEM characterizations in Figure 5 confirm that Au nanoparticles are well dispersed on the
(a)
(b)
9
0
1
Figure 5. (a), (b) TEM images for 0.28wt% Au/DeAlꢀSnꢀBeta; (c) Au particle size distribution.
11
surface of the support. The particle size distribution mainly lies between 3 nm and 8 nm with an
average of 5.2 nm, based on which the dispersion of Au can be estimated to be around 22%. Au
1
2
3
1
nanoparticles loaded on DeAlꢀBeta and SnO /DeAlꢀBeta were also synthesized using the same
2
14
method, resulting in a similar Au particle size (Figure S1). The Au content determined by ICP–
1
1
5
6
OES was 0.28wt%. Both Au/DeAlꢀBeta and Au/SnO /DeAlꢀBeta also had similar content of Au
2
(0.29wt% and 0.28%, respectively).
17
XPS was used to determine the oxidation state of the Au particles. The binding energy
0
1
8
measured indicates that Au is present as Au (Figure S2a).
ACS Paragon Plus Environment
14

Page 15 of 31
ACS Catalysis
1
2
3
1
4
3.2 Catalyst activity testing
5
2
6
Initial catalytic tests were carried out to investigate the catalyst activities affected by reaction
temperature, oxygen partial pressure, acetol concentration and loading amount of Au. Figure 6
shows the conversion of acetol and yields of major products as a function of temperature. Acetol
conversion increases with temperature over the studied range of 373 K to 413 K. However, the
yield of LA first increases and then decreases with temperature, reaching a maximum of 68.4%
at 393 K. Higher temperatures result in loss of LA yield, accompanied with more undesired
acetic acid formation, which shows a monotonous increase. Another byproduct, pyruvic acid
(PA) also exhibited a similar trend of yield as lactic acid, i.e., reaching the highest at 403 K and
then decreasing notably. Other two products, pyruvaldehyde and lactaldehyde, remain at a low
level over the entire temperature range studied.
7
3
4
5
6
7
8
9
0
1
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
1
1
2
3
1
Figure 6. Effect of reaction temperature on the conversion of acetol and yield of various
products. Reaction conditions: 2 mmol acetol; 0.3 g 0.28wt% Au/DeAlꢀSnꢀBeta catalyst; 0.1
MPa O + 1.9 MPa N ; 20 ml H O; 3 h.
14
1
5
6
2
2
2
1
Experiments over various O partial pressures at 393 K were then investigated. As shown in
2
17
Figure 7, the catalyst shows a low activity without O , e.g., only reaching 10.9% acetol
2
ACS Paragon Plus Environment
15

ACS Catalysis
Page 16 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
2
3
4
5
6
7
8
conversion with merely 1.4% LA yield at 393 K. The yield of LA increases significantly as O2
partial pressure increases to 0.05 MPa, and reaches a maximum of 68.4% at 0.1 MPa. Further
increasing the oxidant pressure results in a notable drop of LA yield, probably due to
overꢀoxidation and decarboxylation that lead to the formation of acetic acid (AA). These results
suggest that aerobic atmosphere is indispensable for the efficient conversion of acetol towards
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
LA. A minor amount of LA (1.4% yield) produced in the absence of O is likely due to the
2
residual O in autoclave. Again, only small amount of pyruvaldehyde and lactaldehyde were
2
observed as the oxygen partial pressure increases up to 0.5 MPa.
9
1
1
1
0
1
2
Figure 7. Effect of O partial pressure on the conversion of acetol and yield of various products.
2
Reaction conditions: 2 mmol acetol; 0.3 g 0.28wt% Au/DeAlꢀSnꢀBeta catalyst; p(O + N ) =2
2
2
MPa; 393 K; 20 ml H O; 3 h.
2
ACS Paragon Plus Environment
16

Page 17 of 31
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
1
2
Figure 8. Effect of acetol concentration on conversion of acetol and yields of various products.
3
4
Reaction conditions: 0.3 g 0.28wt% Au/DeAlꢀSnꢀBeta catalyst; 20 ml H O; 393 K; 0.1 MPa O +
2
2
1.9 MPa N ; 3 h.
2
5
6
7
8
9
0
The experiments with various acetol concentrations were performed next, as shown in Figure
8. When the acetol concentration was increased from 0.1 M to 0.5 M, LA yield decreased from
68.4% to 40.3%, probably due to the formation of unknown polymeric products and humins
induced by the higher acetol concentration. The volumetric productivity of LA reached a
ꢀ1
ꢀ1
maximum of 6.6 g L h under the condition of 0.4 M acetol, which is higher than that in the
ꢀ1
ꢀ1
1
conversion of glucose and sucrose by fermentation (0.3–5 g L h ) and comparable with the
ꢀ1
ꢀ1 18
11
other sugar based catalysis process (6.3 g L h ). Moreover, if the catalyst amount was
ꢀ1 ꢀ1
1
1
1
2
3
4
doubled, the volumetric productivity of LA can be further increased to 7.7 g L h , apparently
attributing to the more active sites to favor the desired products, which in turn suppresses the
competing side reaction towards heavy ones.
ACS Paragon Plus Environment
17

ACS Catalysis
Page 18 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
Figure 9. Effect of loading amount of Au on the conversion of acetol and yield of major
products. Reaction conditions: 2 mmol acetol; 0.3 g Au/DeAlꢀSnꢀBeta catalyst; 20 ml H O; 0.1
2
MPa O + 1.9 MPa N ; 393 K; 3 h.
2
2
5
6
7
8
9
0
The performances of Au/DeAlꢀSnꢀBeta catalysts with various gold loadings were also studied
and results are shown in Figure 9. Without Au nanoparticles, only 31.5% acetol is converted.
The presence of Au nanoparticles with 0.28wt% loading remarkably promotes the acetol
conversion to 93.4% with LA yield up to 68.4%. With the further increase in gold loading, acetol
conversion remains relatively constant but LA selectivity decreases and levels off at about 57%.
Since the highest yield of lactic acid can be achieved in the presence of catalyst with 0.28wt%
1
11
Au loading in 0.1 M acetol at 393 K under 0.1 MPa O , these reaction conditions were applied to
2
1
1
1
2
3
4
all the tests below.
3.3 Catalytic behavior
a
Table 2. Conversion of acetol and selectivities of products on the catalysts prepared in the study
Selectivity (%)
Conv.
Entry Catalyst
(%)
LaA
1.3
PyA
0.2
LA
AA
8.2
PA
9.0
1
Au/DeAlꢀSnꢀBeta
93.4
73.2
ACS Paragon Plus Environment
18

Page 19 of 31
ACS Catalysis
1
2
3
4
5
6
7
8
9
2
3
DeAlꢀSnꢀBeta
Au/DeAlꢀBeta
Au/DeAlꢀBeta +
DeAlꢀSnꢀBeta
31.5
94.9
4.1
1.7
0.6
3.5
41.3
9.7
29.5
33.9
0.5
11.9
4
5
93.4
2.1
0.0
0.1
4.9
57.5
7.6
10.5
16.9
11.1
27.8
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
Au/SnO /DeAlꢀBeta 91.1
2
a
1
2
3
Reaction conditions: 2 mmol acetol; 0.3 g catalyst; 20 ml H O; 393 K; 0.1 MPa O + 1.9 MPa
2 2
N ; 3 h. LaA: Lactaldehyde; PyA: Pyruvaldehyde; LA: Lactic acid; PA: Pyruvic acid; AA:
Acetic acid.
2
4
5
6
7
8
9
0
1
To provide fundamental understanding on acetol conversion to LA, various catalysts prepared
in this work were systematically evaluated under the relatively mild and baseꢀfree conditions,
and the results are summarized in Table 2. The data listed in entry 1 clearly demonstrates the
benefits of bifunctional Au/DeAlꢀSnꢀBeta catalyst, exhibiting high acetol conversion and
superior selectivity to lactic acid, reaching 93.4% and 73.2% respectively. The pH changes
before and after the reaction are also monitored, as summarized in Table S1 with a direct
comparison with those literature reports under base condition. The role of the presence of Au
nanoparticles on the support is manifested by entry 2, where Auꢀfree DeAlꢀSnꢀBeta leads to a
striking drop of conversion and selectivity, only 31.5% and 41.3% respectively. Such a low
activity suggests that the oxidation step is mainly catalyzed by the gold nanoparticles. For
Snꢀfree Au/DeAlꢀBeta, although the conversion shown in entry 3 is comparable to the
bifunctional catalyst, the selectivity towards LA decreases dramatically to only 9.7%, and that to
acetic acid increases to 33.9% instead. Meanwhile, unknown polymeric products and humins are
formed, which accounts for the missing portion of overall mass balance in Table 2. The poor
selectivity from the Snꢀfree catalyst suggests that the Lewis acidity of the catalyst is
indispensible to achieving high selectivity to the desired product, LA. In addition, on the
mechanical mixture of Au/DeAlꢀBeta and DeAlꢀSnꢀBeta (entry 4), LA selectivity reduces from
1
1
12
1
3
4
1
15
1
1
1
1
6
7
8
9
20
ACS Paragon Plus Environment
19

ACS Catalysis
Page 20 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
2
3
4
5
73.2% to 57.5%, and more byꢀproducts are formed. This implies that the combination of Lewis
acid Sn sites and Au sites in those proximity favors the conversion of acetol into LA but
suppresses the side reactions. For Au/SnO /DeAlꢀBeta in entry 5, LA selectivity is merely 7.6%,
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
which is comparable to that for Au/DeAlꢀBeta in entry 3. This indicates that the
extraꢀframework, namely hexacoordinated tin（Ⅳ）oxide, is not catalytically active to convert
6
7
acetol to LA. Instead, the Lewis acidity of the catalyst, arising from the Sn sites located in the
zeolite framework, is responsible for 1,2ꢀhydride shift, which is consistent with previous
2
8, 39ꢀ40
+
8
reports.
It should be noted that all those catalysts in Table 2 have undergone Na ꢀexchange
9
process, because the Au/DeAlꢀSnꢀBeta with such exchange achieved a better selectivity toward
+
10
LA compared with the counterpart without Na ꢀexchange. It is generally agreed that the
+
1
1
1
1
1
1
1
1
2
3
4
5
6
7
Na ꢀexchange can diminish the medium/weak Brønsted acid sites which may facilitate the side
reactions of acetol such as aldoꢀcondensation, resulting in formation of heavy products and loss
of carbon efficiency.
3.4 Reaction pathways
Two reactions are associated with the conversion of acetol to lactic acid under baseꢀfree
conditions, namely 1,2ꢀhydride shift and oxidation. Possible reaction pathways for the
conversion of acetol to lactic acid over 0.28wt% Au/DeAlꢀSnꢀBeta is proposed in Scheme 2. In
the first path, acetol is selectively oxidized to PyA catalyzed by Au nanoparticles under aerobic
atmosphere, followed by hydration and 1,2ꢀhydride shfit on Lewis acid sites to form LA. In the
second path, acetol is first converted to LaA via 1,2ꢀhydride shift, followed by selective
oxidization to LA.
18
1
9
0
2
21
ACS Paragon Plus Environment
20

Page 21 of 31
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
1
2
3
Scheme 2. Proposed tandem reaction pathways for the conversion of acetol to lactic acid over
0.28wt% Au/DeAlꢀSnꢀBeta.
4
5
6
7
8
9
0
1
To shed light on the reaction pathway over 0.28wt% Au/DeAlꢀSnꢀBeta catalyst, a series of
experiments were conducted to provide the detailed product distributions over timeꢀonꢀsteam. As
shown in Figure 10a, the reaction of acetol proceeds quite fast initially, resulting in a conversion
of 40.8% in the first 20 minutes, followed by a steady increase with timeꢀonꢀstream. Lactic acid
is a predominant product, and the yield of lactic acid over timeꢀonꢀstream exhibits a similar trend
to acetol conversion. Formation of pyruvic acid and acetic acid is negligible in the first 20
minutes, and then gradually and monotonously increases with timeꢀonꢀstream. Pyruvaldehyde
also forms but remains at a relatively consistent and low level (about 0.3% yield). The yield of
lactaldehyde is also quite low (less than 3% yield).
1
1
12
1
3
4
It should be noted that the selectivity towards lactic acid is slightly higher at a short
timeꢀonꢀstream (e.g., 20 min) and remains at above 70% over entire time course (3 hours) as
illustrated in Figure 10b. In particular, the yield to lactic acid is significantly higher than other
products at any
1
15
1
6
7
1
timeꢀonꢀstream, while the selectivity to LaA or PyA remains very low even in the initial reaction
ACS Paragon Plus Environment
21

ACS Catalysis
Page 22 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
Figure 10. Time course for (a) the conversion of acetol and yields of various products and (b)
the conversion of acetol and selectivity of lactic acid over 0.28wt% Au/DeAlꢀSnꢀBeta. Reaction
conditions: acetol: 2 mmol; 0.28wt% Au/DeAlꢀSnꢀBeta catalyst: 0.3 g; temperature: 393 K; H O:
2
20 ml; atmosphere: 0.1 MPa O + 1.9 MPa N .
2
2
6
7
stage. However, as shown in Scheme 2, it is expected that either LaA or PyA (or both) would be
the primary intermediates, through which lactic acid can be formed. A plausible explanation is
the
8
9
conversion of these intermediates to LA is not a kinetically relevant step. Instead, the formation
of the intermediates could be a rateꢀlimiting step, leading to low yield of LaA or PyA. This is
consistent with the previous literature reports on GLY and DHA, which has been extensively
10
1
1
2
2
1ꢀ23, 40
1
studied to produce lactic acid or lactate for the last decade.
In these studies, it is generally
13
accepted that formation of PyA by the dehydration of GLA is likely the rateꢀdetermining step,
1
4
5
while the subsequent steps including hydrate formation with addition of H O and isomerization
2
1
to LA through 1,2ꢀhydride shift are not kinetically relevant. Similar statements could be also
applied to LaA, which may act as another possible main intermediate in this work. To further
16
ACS Paragon Plus Environment
22

Page 23 of 31
ACS Catalysis
1
2
3
1
4
elucidate which route (PyA or LaA) in Scheme 2 is the dominant one, the experiments with
different substrates including PyA, and LaA as well as LA as substrates were further conducted.
5
2
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
3
4
5
6
7
8
9
0
1
2
3
4
As shown in Table 3, LA conversion on 0.28wt% Au/DeAlꢀSnꢀBeta catalyst only reaches
9.3%, with 51.6% selectivity to PA. This suggests that LA is relatively stable under the reaction
conditions studied, and the low yield of PA is probably due to the further oxidation of LA (entry
3). However, on the same catalyst, both PyA and LaA could be readily converted with nearly
100% conversion in 3 hours with 77.8% and 57.7% selectivity towards LA, respectively. These
results seem to indicate that both reaction pathways might contribute to the production of LA.
Additional experiment using LaA as the substrate showed the selectivity to acetol reached 79.8%
after 20 min timeꢀonꢀstream, and then dropped to 31.0% after 3 h timeꢀonꢀstream (Figure S3).
This suggests that the direct oxidation of LaA to LA may be not a favorite route. Instead, LaA is
probably isomerized to form acetol first via 1,2ꢀhydride shift, and is subsequently converted to
lactic acid through PyA intermediate.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
1
1
1
1
a
Table 3. Conversion of different substrates catalyzed by 0.28wt% Au/DeAlꢀSnꢀBeta
Selectivity (%)
Conv.
Entry Reactant
(
%)
LaA
0.0
ꢀ
PyA
ꢀ
LA
77.8
57.7
ꢀ
AA
PA
Acetol
0.0
1
2
3
PyA
LaA
LA
99.8
97.5
9.3
11.9 5.4
0.2
0.0
4.3
0.4
0.4
31.1
0.0
0.0
51.6
a
1
5
Reaction conditions: 2 mmol substrate; 0.3 g 0.28wt% Au/DeAlꢀSnꢀBeta catalyst; 20 ml H O;
2
16
393 K; 0.1 MPa O + 1.9 MPa N ; 3 h. LaA: Lactaldehyde; PyA: Pyruvaldehyde; LA: Lactic
2 2
1
7
acid; PA: Pyruvic acid; AA: Acetic acid.
1
8
Based on the experimental results, a simplified kinetic model was established to provide a
1
9
comparison over the relative rates of each step involved in Scheme 2. The detailed equations are
ACS Paragon Plus Environment
23

ACS Catalysis
Page 24 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
2
3
4
listed in section 2.5, where all steps involved are considered as firstꢀorder reactions. As seen in
Figure 11, a reasonably good agreement is achieved between the calculated values based on the
model and results obtained experimentally. The calculated rate constants for all the elemental
steps are listed in Table 4.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
6
7
Figure 11. Experimental (circles) and modelled (solid lines) concentration of acetol and different
products in the conversion of acetol to lactic acid over 0.28wt% Au/DeAlꢀSnꢀBeta.
ꢀ1
ꢀ1
8
Table 4. The rate constants (L g h ) calculated for the kinetic model in Scheme 2
k1
k2
k3
k4
k5
k6
0.05273 3.8620 0.27288 2.8753 0.27602 0.01244
9
Based on the calculated rate constants, the fastest step of the overall process is the conversion
10
of PyA to LA (k ), which is consistent with the reactivity trend observed with PyA as substrates
2
1
1
2
and the fact that LA was the predominant product in the reaction starting from both acetol and
1
PyA (Figures 10a and S4). The rate of LaA to acetol (k ) is much higher than that of reverse
4
13
reaction (k ), also in line with the previous experimental results. This again supports that acetol
3
1
4
5
will probably remain as the dominant component under the equilibrium with LaA through the
reversible 1,2ꢀhydride shift. The mechanism should be analogous to the isomerization between
1
ACS Paragon Plus Environment
24

Page 25 of 31
ACS Catalysis
1
2
3
1
4
glyceraldehyde and dihydroxyacetone, where the rate of glyceraldehyde isomerization to
5
4
3
2
3
4
5
6
7
8
9
0
dihydroxyacetone is one order of magnitude faster than that of the reverse reaction. With the
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
help of these kinetic insights, the dominant pathway towards LA in Scheme 2 can be determined.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Although the oxidation rate of acetol to PyA (k ) is relatively low, the PyA intermediate can be
1
rapidly converted to LA (k ) once it is formed. The unfavored equilibrium with acetol leaves
2
LaA at only a low concentration during the reaction, which then slowly converts to LA (k5).
Overall, we speculate that the formation of lactic acid likely proceeds via the route with PyA as
an intermediate, which first forms by oxidation of acetol on the active gold sites, followed by
hydration and 1,2ꢀhydride shift at Lewis acid Sn sites.
1
We performed the recycle tests with short reaction time (20 min) for each run, where the
reaction rate calculated at this very initial stage with lower conversion could be used to estimate
the extent of catalyst deactivation. As shown in Figure 12, no notable drop occurred over the
selectivity of LA during 3 runs, and the turnover frequency (TOF) listed in Table 5 also remains
11
1
1
1
1
2
3
4
5
ꢀ1
essentially the same at 0.71ꢀ0.72 s . Since the oxidation catalyzed by Au nanoparticles is
rateꢀdetermining step as concluded earlier, the TOF calculated here is referred as the moles of
acetol converted per unit time per mole of surface Au atoms, which is deduced from the
dispersion of Au estimated according to the sizes from the TEM measurement. The XPS result in
16
1
7
8
0
1
Figure S2b also reveals that Au surface in the spent catalysts remains as the metallic Au after 3
19
reaction runs, which is consistent with the TOF numbers obtained above. The reusability of the
Au/DeAlꢀSnꢀBeta catalyst was also investigated under 3h’s reaction, the spent catalyst was
regenerated by washing with deionized water and dried overnight. No significant change over the
activity and selectivity was observed for the recycled catalyst in 3 runs (Figure S5). It should be
2
2
2
2
0
1
2
3
24
noted that the catalyst stability is better than that of Pt/SnꢀMFI reported in the literature, which
ACS Paragon Plus Environment
25

ACS Catalysis
Page 26 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
2
3
4
5
6
7
8
9
may be ascribed to hydrophobic innerꢀsurface of SnꢀBeta, and more oxygen poisoning resistance
3
2ꢀ36
of Au in aqueous solution.
Moreover, we also synthesized the Au/DeAlꢀSnꢀMFI catalyst to
directly compare with Au/DeAlꢀ
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
SnꢀBeta in the acetol conversion under the same conditions. As shown in Table S2,
Au/DeAlꢀSnꢀMFI resulted in a yield of 60% LA and 13.7% AA respectively at 97.8%
conversion, whereas the use of Au/DeAlꢀSnꢀBeta achieved a yield of 68.4% LA and 7.7% AA
with 93.4% conversion. The main difference between MFI and Beta (beside their structure) is the
2
4
presence of more medium/weak Brønsted acid sites in the MFI, which may cause the relatively
lower selectivity towards LA and higher propensity to AA compared to Beta.
10
11
Figure 12. Recycling test of Au/DeAlꢀSnꢀBeta catalyst. Reaction conditions: 2mmol acetol; 20
1
2
3
ml H O; 0.3 g 0.28wt% Au/DeAlꢀSnꢀBeta catalyst; 393 K; 20 min; 0.1 MPa O + 1.9 MPa N .
2
2
2
1
Table 5. Meanꢀsize and dispersion of Au nanoparticles and TOF
a
Recycle Meanꢀsize of
number Au NPs (nm) (%)
Au dispersion
TOF
(s )
a
ꢀ1
1
2
3
5.3
5.5
5.6
22.5
21.3
20.9
0.71
0.72
0.71
ACS Paragon Plus Environment
26

Page 27 of 31
ACS Catalysis
1
2
3
1
4
a
Estimated from the meanꢀsize of Au NPs.
5
2
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
3
4
5
6
7
8
9
0
1
2
3
CONCLUSIONS
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Acetol can be efficiently converted to lactic acid in aqueous phase using bifunctional 0.28wt%
Au/DeAlꢀSnꢀBeta catalyst. 93.4% conversion and 73.2% selectivity towards LA were achieved
under mild conditions (393 K and 0.1 MPa O + 1.9 MPa N ). It was found that both Lewis acid
2
2
Sn site and Au nanoparticles are essential to achieving such high selectivity and activity. Based
on the experiment results and measured rate constants, reaction pathway likely proceeds via
acetol being first oxidized to pyruvaldehyde, a kinetically relevant step catalysed by Au
nanoparticles, followed by hydration and 1,2ꢀhydride shift at SnꢀBeta Lewis sites to form lactic
acid.
1
1
1
1
ASSOCIATED CONTENT
1
4
Supporting Information. TEM images and Au particle size distributions; XPS spectra;
Comparison of our work under baseꢀfree conditions with literature reports using base; Time
course for conversions of lactaldehyde and pyruvaldehyde; Recycling tests; The comparison of
Au/Beta, Au/DeAlꢀSnꢀBeta and Au/DeAlꢀSnꢀMFI in conversion of acetol to lactic acid.
15
1
6
7
1
18
1
9
AUTHOR INFORMATION
20
21
Corresponding Author
*Eꢀmail for Shaolong Wan: swan@xmu.edu.cn
ACS Paragon Plus Environment
27

ACS Catalysis
Page 28 of 31
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
*Eꢀmail for Yong Wang: Wang42@xmu.edu.cn
2
3
4
5
6
7
ORCID
Yan Wan: 0000ꢀ0002ꢀ4186ꢀ1064
Yong Wang: 0000ꢀ0002ꢀ8460ꢀ7410
Notes
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The authors declare no competing financial interest.
8
ACKNOWLEDGMENT
9
0
1
2
This work was supported by the Major Research Plan of National Natural Science Foundation
of China (No. 21576227, No. 91545114, and No. 91545203), the National Thousand Talents
Program of P.R. China, and the 985 Program of the Chemistry and Chemical Engineering
disciplines of Xiamen University.
1
1
1
1
1
3
4
REFERENCES
15
(1) Huber, G. W.; Iborra, S.; Corma, A., Chem. Rev. 2006, 106, 4044–4098.
(2) Corma, A.; Iborra, S.; Velty, A., Chem. Rev. 2007, 107, 2411–2502.
(3) Antunes, M. M.; Lima, S.; Neves, P.; Magalhães, A. L.; Fazio, E.; Fernandes, A.; Neri, F.;
Silva, C. M.; Rocha, S. M.; Ribeiro, M. F., J. Catal. 2015, 329, 522–537.
(4) Czernik, S.; Bridgwater, A. V., Energy Fuels 2004, 18, 590–598.
1
1
1
1
6
7
8
9
20
(5) Mohan, D.; Pittman, C. U.; Steele, P. H., Energy Fuels 2006, 20, 848–889.
ACS Paragon Plus Environment
28

Page 29 of 31
ACS Catalysis
1
2
3
1
4
(6) Vispute, T. P.; Zhang, H.; Sanna, A.; Xiao, R.; Huber, G. W., Science 2010, 330, 1222–
1227.
5
2
3
4
5
6
7
8
9
0
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(7) Anex, R. P.; Aden, A.; Kazi, F. K.; Fortman, J.; Swanson, R. M.; Wright, M. M.; Satrio, J.
A.; Brown, R. C.; Daugaard, D. E.; Platon, A., Fuel 2010, 89, S29–S35.
(8) Wang, H.; Male, J.; Wang, Y., ACS Catal. 2013, 3, 1047–1070.
(9) Santhanaraj, D.; Rover, M. R.; Resasco, D. E.; Brown, R. C.; Crossley, S., ChemSusChem
2014, 7, 3132–3137.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(10) Pollard, A. S.; Rover, M. R.; Brown, R. C., J. Anal. Appl. Pyrolysis 2012, 93, 129–138.
(11) Resasco, D. E.; Crossley, S. P., Catal. Today 2014, 257, 185–199.
(12) MäkiꢀArvela, P.; Simakova, I. L.; Salmi, T.; Murzin, D. Y., Chem. Rev. 2014, 114, 1909–
1971.
1
11
12
13
14
15
(13) Dusselier, M.; Wouwe, P. V.; Dewaele, A.; Makshina, E.; Sels, B. F., Energy Environ. Sci.
2013, 6, 1415–1442.
(14) Esposito, D.; Antonietti, M., ChemSusChem 2013, 6, 989–992.
(15) Wang, Y.; Deng, W.; Wang, B.; Zhang, Q.; Wan, X.; Tang, Z.; Wang, Y.; Zhu, C.; Cao,
Z.; Wang, G., Nat. Commun. 2013, 4, 2141.
16
1
7
8
(16) Holm, M. S.; Saravanamurugan, S.; Taarning, E., Science 2010, 328, 602–605.
(17) Tolborg, S.; Sádaba, I.; Osmundsen, C. M.; Fristrup, P.; Holm, M. S.; Taarning, E.,
ChemSusChem 2015, 8, 613–617.
1
19
20
21
22
23
(18) Wang, F. F.; Liu, C. L.; Dong, W. S., Green Chem. 2013, 15, 2091–2095.
(19) Orazov, M.; Davis, M. E., Proc. Natl. Acad. Sci. U.S.A. 2015, 112, 11777–11782.
(20) Rasrendra, C. B.; Fachri, B. A.; Makertihartha, I. G.; Adisasmito, S.; Heeres, H. J.,
ChemSusChem 2011, 4, 768–777.
ACS Paragon Plus Environment
29