Understanding the solvent effect on the catalytic oxidation of 1,4-butanediol in methanol over Au/TiO2 catalyst: NMR diffusion and relaxation studies

Article abstract of DOI:10.1002/chem.201201922

The effect of water on the catalytic oxidation of 1,4-butanediol in methanol over Au/TiO₂ has been investigated by catalytic reaction studies and NMR diffusion and relaxation studies. The addition of water to the dry catalytic system led to a decrease of both conversion and selectivity towards dimethyl succinate. Pulsed-field gradient (PFG)-NMR spectroscopy was used to assess the effect of water addition on the effective self-diffusivity of the reactant within the catalyst. NMR relaxation studies were also carried out to probe the strength of surface interaction of the reactant in the absence and presence of water. PFG-NMR studies revealed that the addition of water to the initial system, although increasing the dilution of the system, leads to a significant decrease of effective diffusion rate of the reactant within the catalyst. From T₁ and T₂ relaxation measurements it was possible to infer the strength of surface interaction of the reactant with the catalyst surface. The addition of water was found to inhibit the adsorption of the reactant over the catalyst surface, with the T₁/T₂ ratio of 1,4-butanediol decreasing significantly when water was added. The results overall suggest that both the decrease of diffusion rate and adsorption strength of the reactant within the catalyst, due to water addition, limits the access of reactant molecules to the catalytic sites, which results in a decrease of reaction rate and conversion. Copyright

Full text of DOI:10.1002/chem.201201922

FULL PAPER
DOI: 10.1002/chem.201201922
Understanding the Solvent Effect on the Catalytic Oxidation of 1,4-
Butanediol in Methanol over Au/TiO2 Catalyst: NMR Diffusion and
Relaxation Studies
Carmine DꢀAgostino,[a] Gemma L. Brett,[b] Peter J. Miedziak,[b] David W. Knight,[b]
Graham J. Hutchings,[b] Lynn F. Gladden,[a] and Mick D. Mantle*[a]
Abstract: The effect of water on the
catalytic oxidation of 1,4-butanediol in
methanol over Au/TiO2 has been inves-
tigated by catalytic reaction studies and
NMR diffusion and relaxation studies.
The addition of water to the dry cata-
lytic system led to a decrease of both
conversion and selectivity towards di-
methyl succinate. Pulsed-field gradient
(PFG)-NMR spectroscopy was used to
assess the effect of water addition on
the effective self-diffusivity of the reac-
tant within the catalyst. NMR relaxa-
tion studies were also carried out to
probe the strength of surface interac-
tion of the reactant in the absence and
presence of water. PFG-NMR studies
revealed that the addition of water to
the initial system, although increasing
the dilution of the system, leads to a
significant decrease of effective diffu-
sion rate of the reactant within the cat-
alyst. From T1 and T2 relaxation meas-
urements it was possible to infer the
strength of surface interaction of the
reactant with the catalyst surface. The
addition of water was found to inhibit
the adsorption of the reactant over the
catalyst surface, with the T1/T2 ratio of
1,4-butanediol decreasing significantly
when water was added. The results
overall suggest that both the decrease
of diffusion rate and adsorption
strength of the reactant within the cata-
lyst, due to water addition, limits the
access of reactant molecules to the cat-
alytic sites, which results in a decrease
of reaction rate and conversion.
Keywords: adsorption · diffusion ·
heterogeneous catalysis
·
NMR
spectroscopy · solvent effects
Introduction
to be a successful route for the selective liquid-phase oxida-
tion of a variety of chemicals. Taarning and co-workers stud-
ied the gold catalyzed oxidative esterifications of glycerol,
1,2-propanediol and 1,3-propanediol using methanol as the
solvent and found these to be potentially important routes
for the production of fine chemicals such as methyl lactate
and methyl acrylate.[9] Selective oxidation of 1,2-propanediol
to lactate was also carried out in aqueous solvent by the
Hutchings group.[10] They found that the addition of palladi-
um to gold, supported on TiO2 and carbon, significantly en-
hanced the activity, while retaining a high selectivity to lac-
tate. A comparison of the activity of Au–Pd/TiO2 for the ox-
idation of different C3 alcohols was also carried out, with
glycerol showing the highest conversion, followed by 1,2-
propanediol, 1,3-propanediol and both propanols. Au/C and
Au/oxide (g-Al2O3 and TiO2) catalysts have been compared
with regard to the liquid-phase oxidation of ethylene glycol
to glycolate.[11] The activity of gold supported on the oxides
was found to increase with decreasing particle size, whereas
the gold supported on carbon did not show a clear trend. A
support effect was also found when comparing the results of
different catalysts with similar sizes of gold nanoparticles.
Alumina-supported catalysts were found to be more active
than those supported on TiO2. The authors pointed out that
the presence of the water solvent can markedly modify the
reaction pathway through interaction with the catalytically
active surface, but the paper did not discuss details or give
Heterogeneous gold catalysis has attracted much attention
in recent years since the discovery of gold catalytic proper-
ties in the early 80s by Hutchings[1] and Haruta.[2] Several re-
views on the subject can be found in the literature.[3–5] An
increasing interest in heterogeneous gold catalysis has been
turned to the use of gold nanoparticles supported on metal
oxides as effective catalysts for aerobic oxidation reactions
of renewable feedstocks.[6–8] Unlike conventional organic
synthesis, which uses large amounts of hazardous oxidants
such as nitric or chromic acid, the use of dioxygen has enor-
mous advantages in terms of safety issues and environmen-
tal benefits. The use of supported gold catalysts has proven
[a] Dr. C. DꢀAgostino, Prof. L. F. Gladden, Dr. M. D. Mantle
Department of Chemical Engineering & Biotechnology
University of Cambridge, Pembroke Street
Cambridge CB2 3RA (UK)
E-mail: mdm20@cam.ac.uk
[b] Dr. G. L. Brett, Dr. P. J. Miedziak, Prof. D. W. Knight,
Prof. G. J. Hutchings
Cardiff Catalysis Institute, School of Chemistry
Cardiff University, Main Building
Park Place, Cardiff CF10 3AT (UK)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201201922.
Chem. Eur. J. 2012, 00, 0 – 0
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
1
&
ÞÞ
These are not the final page numbers!

any evidence of how water may affect such reactions. The
effect of water is indeed widely reported in heterogeneous
catalysis. It has been reported to act as a poison in zirconia
catalysts used in propane dehydrogenation.[12] Catalytic stud-
ies of 2-butanone hydrogenation over Ru/SiO2 catalyst
showed that treating the catalyst with water led to a com-
plete deactivation after six hours, due to the formation of
Ru(OH)x on the catalyst surface.[13] The effects of solvents
on heterogeneous catalytic processes have been widely re-
ported.[13–16] For example, Lignier and co-workers[16] studied
the aerobic epoxidation of stilbene over Au/TiO2 catalyst
and found that the choice of the solvent is critical. Among
the selected solvents, only alkyl-substituted cyclohexanes led
to a high yield of epoxide. Other solvents, such as the more
polar ones, were associated with significantly lower yields.
An attempt to explain the solvent effect was made by corre-
lating catalytic activity in each solvent with the dielectric
constant and the solubility of oxygen in the solvent. Howev-
er, the solvent efficiency in terms of catalytic activity did
not follow any trend in terms of dielectric constant (associ-
ated with polar effects, that is, polarity of the solvent mole-
cules) and oxygen solubility. The role of the solvent in heter-
ogeneous catalysis is still unclear and remains a topic of
debate. A better understanding of the physical chemistry in-
volved in liquid-phase heterogeneous catalysis is hence im-
portant for the understanding of the solvent effect and the
design of novel catalytic processes.
NMR relaxometry was also used to probe how the strength
of surface interaction of 1,4-butanediol with the catalyst sur-
face changes when water is added. This represents the first
paper that combines catalytic reaction studies with NMR
diffusion and relaxation techniques, aiming at providing an
understanding of the liquid-phase behaviour of the catalytic
process at a molecular level.
Results and Discussion
Reaction studies: The activity of the 1.5%Au/TiO2 catalyst
was assessed in methanol and a 10% by volume water/meth-
anol mixture. The effect of the reaction time on this oxida-
tion is shown in Figure 1. It is clear that the addition of
In this context, NMR spectroscopy represents a technique
that may provide insight into heterogeneous catalytic reac-
tions because it is able to probe molecular diffusion and ad-
sorption in porous media. These phenomena lie at the heart
of heterogeneous catalytic processes. An excellent review on
diffusion and adsorption in porous materials probed by
NMR techniques has been recently published by Gladden
and Mitchell,[17] which covers a wide range of applications of
magnetic resonance techniques to porous media. Diffusion
of liquids in porous media can be probed by using the
pulsed-field gradient (PFG)-NMR technique,[18] which is
non-invasive and chemically selective. Liquid adsorption
phenomena in porous materials can be also probed and this
relies on the measurements of NMR relaxation time con-
stants of liquids imbibed in porous media.[19,20] We have re-
cently published an article that combined catalytic and
PFG-NMR studies of alcohols in supported gold catalysts.[21]
The results revealed that the inhibition effect observed from
traditional catalytic activity studies for the oxidation of 2-oc-
tanol most likely results from competitive adsorption of the
ketone product.
In the present work, we have studied the catalytic oxida-
tion of 1,4-butanediol to g-butyrolactone in methanol over a
Au/TiO2 catalyst. The effect of water addition on catalytic
activity was assessed. NMR diffusion and adsorption studies
were carried out to elucidate the possible mechanisms re-
sponsible for the catalytic inhibition due to water addition.
Effective self-diffusion coefficients of 1,4-butanediol in pure
methanol and methanol/water mixtures within the Au/TiO2
catalyst were measured using PFG-NMR spectroscopy.
Figure 1. The effect of water addition on conversion and selectivity to g-
butyrolactone and dimethyl succinate. Conversion of 1,4-butanediol in
&
methanol ( ), in 10% water in methanol (&); selectivity to g-butyrolac-
^
^
tone in methanol ( ) and 10% water in methanol ( ); and selectivity to
!
!
dimethyl succinate in methanol ( ) and 10% water in methanol ( ).
water leads to a decrease in the conversion at all points of
the reaction profile. The initial activity of the catalyst is sig-
nificantly affected, as can been seen by the turnover fre-
quencies (TOFs), with the TOF when methanol is used as a
solvent being 403 molmolꢀ1 hꢀ1 compared with 200 mol
molꢀ1 hꢀ1 for the water/methanol mixture. From Figure 1, it
appears that the addition of water to the methanol solvent
also has an effect on the selectivity to g-butyrolactone. How-
ever, when the selectivity is considered in terms of the con-
version (Figure 2), it is clear that the reactions under the dif-
ferent solvent conditions are proceeding through the same
pathway and the selectivity to g-butyrolactone is unaffected
by the addition of water to the methanol.
To gain an understanding of the difference of rate ob-
served in the different solvent systems, we have carried out
detailed PFG-NMR and NMR relaxometry studies and
these are discussed in the subsequent section.
&
2
&
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Catalytic Oxidation of 1,4-Butanediol in Methanol over Au/TiO2 Catalyst
FULL PAPER
&
Figure 2. Selectivity to g-butyrolactone in methanol ( ), in 10% water in
methanol (&).
1
NMR diffusion and adsorption studies: The H NMR spec-
trum of a 5% mole fraction 1,4-butanediol/methanol mix-
ture imbibed in Au/TiO2 catalyst grains is shown in Figure 3.
The signal of the two -CH2- groups of 1,4-butanediol is
clearly visible at approximately d=1.4 ppm and this makes
Figure 4. PFG-NMR log attenuation plots for 1,4-butanediol in methanol
and methanol/water mixtures: a) bulk liquid mixtures and b) liquid mix-
tures imbibed in Au/TiO2 catalyst. Solid lines represent fittings obtained
using a single exponential decay.[18]
of 1,4-butanediol in methanol and 1,4-butanediol in a 1:1
methanol/water mixture. The log attenuation plots for bulk
liquid mixtures show a linear behaviour on a semi-logarith-
mic scale, which is characteristic for free unrestricted diffu-
sion. The log attenuation plots of the two 1,4-butanediol
mixtures within the Au/TiO2 catalyst are depicted in Fig-
ure 4b. Similar plots were obtained for the mixtures in the
presence of base (see the Supporting Information S1).
Linear behaviour in porous media is referred to as quasi-ho-
mogeneous and is typical of liquids imbibed in porous
media with mesopores.[22] In the presence of mesopores
(pores with a diameter between 2 and 50 nm according to
the IUPAC[23]), the distance travelled by molecules in the
typical observation times probed in a PFG-NMR experi-
ment is much greater than the typical pore dimension, that
is, the root mean square displacement (R.M.S.D.) of mole-
cules, which is of the order of several micrometers, is much
larger than the average pore size of the porous Au/TiO2 cat-
alyst,[21] which was found to be 13 nanometers with a unimo-
Figure 3. 1H NMR spectrum of 5% mole fraction 1,4-butanediol in meth-
anol imbibed in Au/TiO2 catalyst. The aliphatic resonance of 1,4-butane-
diol (-CH2-) is clearly distinguishable at approximately d=1.4 ppm. The
peaks at approximately d=3.5 and 5.3 ppm are the methoxy and hydrox-
yl resonances of the methanol solvent, respectively. The alkoxy and hy-
droxyl resonances of 1,4-butanediol overlap with those of methanol and
cannot be distinguished.
it possible to measure unambiguously its diffusion coeffi-
cient and relaxation times. This is not always possible for
liquid mixtures in porous catalysts as the line broadening
can sometimes be quite significant so that different resonan-
ces are not distinguishable.
PFG-NMR diffusion log attenuation plots of 1,4-butane-
diol mixtures in methanol and methanol/water are shown in
Figure 4. Figure 4a reports the plots for bulk liquid mixtures
Chem. Eur. J. 2012, 00, 0 – 0
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
3
&
ÞÞ
These are not the final page numbers!

M. D. Mantle et al.
dal pore size distribution (from BET and BJH analysis of ni-
trogen adsorption). As a consequence, molecules will expe-
rience many collisions with the pore walls and the measured
diffusivity, Deff, will be that of the liquid confined in the het-
erogeneous medium.
By comparing the case of 1,4-butanediol in methanol with
1,4-butanediol in methanol/water, it is clearly visible that
the log plot for this latter mixture shows a significantly
lower slope, that is, lower self-diffusivity values. This is ob-
served for both bulk liquids and liquids in catalyst particles.
Numerical values of the diffusion coefficients are reported
in Table 1. The addition of water leads to a decrease of
aTHz =aidealꢀaexperimental) deviated significantly from the value
of zero, that is, non-interacting ideal mixing. Over the entire
composition range, the positive value of aTHz indicated a de-
crease of absorption upon mixing in all alcohol/water mix-
tures (negative excess terahertz absorptions), which suggest-
ed that the introduction of a second species into an initially
pure liquid destroys the original hydrogen-bonded network
of the pure liquid, forming a new extended hydrogen-
bonded alcohol/water network, which is characterized by
slower dynamics than the original bulk liquid structure.
Moreover, it is also known that the presence of water modi-
fies the nature of the hydrogen bonds with the diol due to
its ability to act as a “bridge” between solute mole-
cules that are able to form hydrogen bonds,[29,30]
leading to the formation of larger aggregates, hence
Table 1. Numerical values of self-diffusivity and PFG interaction parameter, x, from
PFG-NMR measurements. The tortuosity of the catalyst, t, is also reported.
a slower mobility. Indeed, despite water and metha-
nol in their pure liquid state having similar diffusion
coefficients, as previously discussed, the self-diffu-
sivity of 5% mole fraction 1,4-butanediol in metha-
nol was found to be 9.2ꢂ10ꢀ10 m2 sꢀ1, whereas for
the same concentration in pure water the self-diffu-
sivity of the diol was found to be lower, with a
value of approximately 4.1ꢂ10ꢀ10 m2 sꢀ1.
Dꢂ109 [m2 sꢀ1
]
Deff ꢂ109 [m2 sꢀ1
]
x [–]
Bulk liquid
Liquid in Au/TiO2
1,4-butanediol in MeOH
1,4-butanediol in MeOH/H2O
0.92ꢁ0.01
0.48ꢁ0.01
0.54ꢁ0.01
0.28ꢁ0.01
1.7ꢁ0.04
1.7ꢁ0.04
1,4-butanediol in MeOH/base
1,4-butanediol in MeOH/H2O/base 0.41ꢁ0.01
0.71ꢁ0.02
0.27ꢁ0.01
0.15ꢁ0.01
2.6ꢁ0.07
2.7ꢁ0.07
It is therefore clear that the decrease of 1,4-buta-
nediol self-diffusivity in 1,4-butanediol/methanol
mixtures, caused by addition of water and reported
in this work, is a consequence of the changes in the
t [–]
1.9ꢁ0.05
cyclohexane
1.35ꢁ0.01
0.71ꢁ0.01
almost 50% in the diffusion rate of 1,4-butanediol within
the catalyst. The addition of base further reduces the diffu-
sion rate of 1,4-butanediol in all cases. This is expected since
the addition of the sodium methoxide salt increases the vis-
cosity of the solution. The self-diffusion coefficients of the
three components as single-component bulk liquid are: 1,4-
butanediol (1.6ꢂ10ꢀ11 m2 sꢀ1); methanol (2.2ꢂ10ꢀ9 m2 sꢀ1)
and water (2.0ꢂ10ꢀ9 m2 sꢀ1). The addition of water, diluting
the system, would intuitively lead to the conclusion that an
increase of the molecular motion of 1,4-butanediol should
be observed. Experimental results, however, clearly show a
net decrease of molecular mobility. The explanation of this
lies in the non-ideality of water/methanol mixtures, widely
reported in the literature[24–27] and in the nature of the hy-
drogen bonds involved in the different solvents.
The mutual and tracer diffusion profiles of water/metha-
nol mixtures as a function of composition show a marked
convex profile, with a minimum value,[25] which means that
when methanol is mixed with water, the diffusion of the
system is slower when compared with the pure species and
this is attributed to a higher degree of structuring in the
mixture caused by hydrogen bonding. A recent study has in-
vestigated the mixing mechanism of water/alcohol mix-
tures.[28] In this study, terahertz time domain spectroscopy
(THz-TDS) and PFG-NMR spectroscopy were used to
study the dynamics of 2-propanol/water mixtures across the
entire composition range. Self-diffusivity profiles showed a
convex profile, with a minimum in self-diffusivity. Absorp-
tion measurements in the THz frequency range showed that
the relative terahertz absorption coefficient, aTHz, (i.e.,
nature of hydrogen bonding due to water addition to the
original mixture. Values of the PFG interaction parameter,
x, that is, the ratio of the bulk liquid self-diffusivity to pore
self-diffusivity,[21] for 1,4-butanediol in both base-free mix-
tures and mixtures with base, are also reported in Table 1.
For the base-free mixtures the x-value of 1,4-butanediol was
found to be approximately 1.7 in both methanol and metha-
nol/water. This value was found to be statistically lower than
the PFG interaction parameter value of 1.9, obtained using
cyclohexane as guest molecule, which represents a good esti-
mate of the porous catalyst tortuosity, t.[31] The use of liquid
alkanes as the “optimal choice” to probe tortuosity in
porous catalysts has been validated in our recent work,
which showed that alkanes are suitable molecules to probe
structural properties of mesoporous solids, whereas the use
of functionalized molecules may yield very different values
that depend on the interactions within the porous matrix.[31]
This means that relative to the bulk liquid, 1,4-butanediol in
porous Au/TiO2 catalyst has a self-diffusion coefficient
higher than the theoretical value, the latter being the 1,4-bu-
tanediol bulk liquid self-diffusivity divided by the catalyst
tortuosity, t. This phenomenon seems to be widely observed
for viscous molecules, such as diols and glycerol, which are
capable of forming hydrogen bonds and has been reported
and discussed in our previous work.[21,31] The reason for this
enhanced diffusivity relative to the theoretical value is
thought to originate from a disruption of the extensive hy-
drogen bonding network of such molecules in mesoporous
materials,[31] although the mechanism of this is still unclear.
It is interesting to note that in the presence of the sodium
&
4
&
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Catalytic Oxidation of 1,4-Butanediol in Methanol over Au/TiO2 Catalyst
FULL PAPER
methoxide base the PFG interaction parameter of 1,4-buta-
nediol has a value much higher than for the base-free mix-
tures, which is also higher than the tortuosity factor. This is
probably due to changes in physical interactions within the
catalyst pores due to the presence of Na+ and CH3Oꢀions
interacting with the bulk liquid and the surface of the cata-
lyst. This also suggests that the mechanism of hydrogen
bonding network disruption[31] is related to changes in sur-
face properties of the catalyst. However, the important
point to note is that both the 1,4-butanediol/methanol and
1,4-butanediol/methanol/water mixtures are affected in the
same way, which implies that despite the fact that the base
changes the absolute values of self-diffusivity, it does not
modify the relative values in the two solvent systems.
The adsorption of 1,4-butanediol onto the catalytic sur-
face was also assessed by using NMR relaxometry techni-
ques.[19] The importance of adsorption of reactants over het-
erogeneous catalysts is often considered a key step of the
whole oxidation reaction. Several oxidation mechanisms
have been proposed to explain the oxidation of alcohols
over heterogeneous catalysts.[32] According to the classical
and widely accepted dehydrogenation mechanism,[33] the ox-
idation of alcoholic groups over catalytic surfaces proceeds
according to the following scheme [Eq. (1)]:
RCH2OHðsolÞ $ RCH2OHads ! RCHOads þ 2Hads
ð1Þ
The adsorption of the reactant in solution onto the cata-
lyst surface occurs at equilibrium and is the first step of the
overall oxidation process; adsorbed oxygen is necessary to
oxidize the co-product hydrogen, thus shifting the equilibri-
um towards the product. The adsorption of the reactant
onto the catalyst surface, although being not the only proc-
ess involved in the oxidation reaction, is however a funda-
mental step for the reaction to occur. Information on the ad-
sorption strength of the reactant within the Au/TiO2 catalyst
in pure methanol and the methanol/water mixture, was ob-
tained by measuring the T1 and T2 NMR relaxation times,
the ratio (T1/T2) of which relates to the strength of surface
interaction of a probe molecule.[19,34] NMR relaxation time
plots are reported in Figure 5 for the base-free mixtures.
Similar plots were obtained for the mixtures in the presence
of base (see the Supporting Information S2). The T1 inver-
sion recovery plot is reported in Figure 5a, whereas the T2
plot is reported in Figure 5b. All the experimental data were
fitted by using a mono-exponential model, giving a single
value of T1 and T2. This finding, together with Brunauer–
Emmett–Teller (BET) and Barrett–Joyner–Halenda (BJH)
analysis and linearity of log attenuation plots, indicates that
the pore structure of the catalyst is homogeneous, with mol-
ecules experiencing a single relaxation environment for both
T1 and T2. The relaxation time results of mixtures within the
catalyst are summarized in Table 2. A comparison with the
cyclohexane values is reported and the reason for this com-
parison is subsequently explained. Relaxation measurements
on 1,4-butanediol within the mixtures were also carried out
in bulk liquid mixtures (i.e., in the absence of the catalyst).
Figure 5. Experimental results from NMR relaxation measurements:
a) T1 relaxation and b) T2 relaxation of liquid mixtures imbibed in Au/
TiO2 catalyst. Solid lines represent fittings obtained using single exponen-
tial: a) inversion recovery[39] and b) CPMG decay.[40]
Table 2. Numerical values from T1 and T2 NMR relaxation measure-
ments in Au/TiO2 catalyst. The T1/T2 value is a measure of the strength
of surface interaction of the molecules with the solid surface of the cata-
lyst.
T1 [ms]
T2 [ms]
T1/T2 [–]
1,4-butanediol in MeOH
1,4-butanediol in MeOH/H2O
1100ꢁ11
670ꢁ7
6ꢁ0.2
183ꢁ7
48ꢁ2
14ꢁ0.6
1,4-butanediol in MeOH/base
1,4-butanediol in MeOH/H2O/base
926ꢁ9
765ꢁ8
16ꢁ0.6
49ꢁ1.9
58ꢁ2
16ꢁ0.6
cyclohexane
2200ꢁ22
134ꢁ5
16ꢁ0.6
As expected for bulk liquids, the measurements gave a T1/T2
ꢂ1 in all cases, which is in excellent agreement with the
theory of relaxation of liquids.[35]
Inspection of Table 2 shows that in general the T1/T2 ratio
of cyclohexane is much lower than that of 1,4-butanediol in
the mixtures. This is expected, because cyclohexane is a
non-polar molecule with no functional groups, so interac-
Chem. Eur. J. 2012, 00, 0 – 0
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
5
&
ÞÞ
These are not the final page numbers!

M. D. Mantle et al.
tions with the catalyst surface are expected to be much
weaker when compared with molecules such as alcohols and
other hydroxyl group-bearing molecules. This is well in
agreement with the work carried out by Lanin et al.,[36] who
have used gas chromatography to study the adsorption of
several classes of organic compounds over titanium dioxide
and have shown that alkanes exhibit lower adsorption ener-
gies than other molecules such as alkenes and oxygenated
polar molecules. For this reason, cyclohexane was used as a
reference compound for both diffusion and relaxation meas-
urements. Focusing the attention on the T1/T2 ratio of 1,4-
butanediol in the two different mixtures, it is clear that the
addition of water to a 1,4-butanediol/methanol mixture
leads to a dramatic drop in the T1/T2 ratio of 1,4-butanediol
in the Au/TiO2 catalyst. This means that the surface interac-
tion of 1,4-butanediol with the catalyst surface is significant-
ly inhibited when water is added. The addition of base to
both mixtures leads to a drop of the T1/T2 ratio. It is impor-
tant to note that also in this case, as for the PFG interaction
parameter, the T1/T2 ratio of 1,4-butanediol in methanol and
in methanol/water is scaled down in exactly the same pro-
portion when the base is added; hence the addition of base
changes the absolute values but not the relative values be-
tween the two solvent systems. The decrease of T1/T2 of 1,4-
butanediol in the presence of base in both solvent systems is
likely to be due to the interaction of the ionic species in the
base with the catalyst surface.
work, will also affect the reaction pathway. However, this
would influence the oxidation reaction in case oxygen is rate
limiting, which is not the case as these reactions are run in a
large excess of oxygen. The NMR results explain extremely
well the drop of catalytic activity observed when water is
added to the system and give new insights into the solvent
effect in heterogeneous catalysis. We are currently extending
our work to other diol oxidations over Au/TiO2 catalysts in
solvent systems similar to those used in this work. The initial
results strongly support the finding of this work and will be
the subject of future publications.
Conclusion
The catalytic oxidation of 1,4-butanediol in methanol over
an Au/TiO2 catalyst has been studied with both catalytic re-
action and NMR diffusion and adsorption studies. The
effect of water on catalytic activity was assessed and found
to be detrimental to both conversion and selectivity to di-
methyl succinate. PFG-NMR spectroscopy was used to
study mass transport phenomena involved in the catalytic
system and combined with NMR relaxation measurements,
which gives an indication of the strength of adsorption of
liquid molecules in porous catalysts. The NMR spectroscopy
results revealed that the addition of water decreases the dif-
fusion rate of the reactant in the porous Au/TiO2 catalyst by
almost 50%, due to the non-ideal mixing of water/methanol
mixtures. At the same time, water addition was found to in-
hibit significantly the adsorption of the reactant on the cata-
lyst surface. This, in turn, limits the access of reactant mole-
cules to the catalytic sites and is responsible for the drop of
catalytic activity observed when water is added to the initial
system. The work is a very good example of how combining
catalytic reaction studies with NMR techniques can give
great insights into the understanding of heterogeneous cata-
lytic processes and elucidate the role of the solvent in heter-
ogeneous catalysis.
A summary, containing both diffusion and T1/T2 results
for the base-free system, is provided in Figure 6. From the
experimental results, it is clear that the addition of water is
detrimental both in terms of diffusion rate and the strength
Experimental Section
Reaction studies: A HEL autoclave reactor with a nominal volume of
75 mL and a maximum working pressure of 1000 bar was charged with
1,4-butanediol (4 mmol, 0.72 g), sodium methoxide (Na+CH3Oꢀ,
0.8 mmol, 0.17 g, 25 wt% solution in methanol), methanol (27.73 g) or
10% water in methanol by volume (28.44 g) and catalyst (51.2 mg). The
catalyst was a 1.5%Au/TiO2 (World Gold Council) and the substrate to
metal ratio was equal to 1000. The average pore size of the porous cata-
lyst was equal to 13 nm with a unimodal pore-size distribution. The reac-
tion vessel was purged 3 times with oxygen; the oxygen inlet remained
open throughout the reaction so that oxygen was replenished as it was
consumed. The oxygen pressure during reaction was set to 10 bar. The re-
action temperature was 1008C and the stirrer speed was 1000 rpm. Sam-
ples were taken periodically and analyzed by gas chromatography. GC
analysis was carried out using a Varian 3800 chromatograph equipped
with a CP 8400 autosampler and CP-wax 52 column. Products were iden-
tified by comparison with authentic samples and quantifications were es-
tablished using an external calibration method. The main oxidation prod-
Figure 6. Summary reporting the effect of water on catalytic activity, self-
diffusion coefficient and T1/T2 ratio of 1,4-butanediol in Au/TiO2 catalyst.
of surface interaction. Adding water to a 1,4-butanediol/
methanol mixture significantly decreases the diffusion rate
of the reactant within the catalyst and at same time inhibits
adsorption onto the catalyst surface. Those two factors will
therefore limit the access of reactant molecules to the cata-
lytic sites, explaining the drop of catalytic activity observed
when water is added to the initial system. It has to be point-
ed out that the effect in the gas phase, which cannot be
probed with the techniques and conditions used in this
&
6
&
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Catalytic Oxidation of 1,4-Butanediol in Methanol over Au/TiO2 Catalyst
FULL PAPER
uct of the reaction was g-butyrolactone, whereas methyl-4-hydroxybuty-
rate and dimethyl succinate were the secondary oxidation products.
quality of life. For more information see www.innovateuk.org. We also
wish to thank the CASTech consortium for supporting this work (EP/
G011397/1).
NMR experiments: All the NMR experiments were performed in a
Bruker Biospin DMX 300 operating at a 1H frequency of 300.13 MHz.
The spectrum of tetramethylsilane (TMS) was used as the chemical shift
reference. The 1H PFG-NMR experiments were carried out using a
Bruker Biospin Diff-30 diffusion probe capable of producing magnetic
field gradient pulses up to 11.6 Tmꢀ1. Diffusion measurements of pure
bulk liquids were performed using the PGSTE[37] pulse sequence, where-
as the APGSTE[38] sequence was used when studying diffusion of liquids
within catalyst particles to minimize the effects of background magnetic
field gradients. The measurements were carried out holding the gradient
pulse duration, d, constant and ramping the magnetic field gradient
strength, g, through sixteen values, from 0.05 to 8 Tmꢀ1. The observation
time, D, was set to 50 ms for pure bulk liquid samples and 100 ms for liq-
uids in catalyst particles. The diffusion coefficient was obtained by taking
the negative slope on a semi-log scale of the signal attenuation as a func-
[1] G. J. Hutchings, J. Catal. 1985, 96, 292–295.
[2] M. Haruta, T. Kobayashi, H. Sano, N. Yamada, Chem. Lett. 1987,
405–408.
[3] L. McEwan, M. Julius, S. Roberts, J. C. Q. Fletcher, Gold Bull. 2010,
43, 298–306.
[4] G. J. Hutchings, M. Haruta, Appl. Catal. A 2005, 291, 2–5.
[5] A. I. Kozlov, A. P. Kozlova, H. C. Liu, Y. Iwasawa, Appl. Catal. A
1999, 182, 9–28.
[6] D. I. Enache, J. K. Edwards, P. Landon, B. Solsona-Espriu, A. F.
Carley, A. A. Herzing, M. Watanabe, C. J. Kiely, D. W. Knight, G. J.
Hutchings, Science 2006, 311, 362–365.
[7] T. Hayashi, T. Inagaki, N. Itayama, H. Baba, Catal. Today 2006, 117,
210–213.
tion of the b-factor,[18] g2g2d
ACHTUNGTRENNUNG
ratio of the nuclei being studied (i.e., 1H in our case). The PFG interac-
tion parameter,[21] x, was defined as the ratio of the bulk liquid self-diffu-
sivity to that of the liquid within the porous catalyst. For cyclohexane,
and alkanes in general, this ratio corresponds to the tortuosity, t, of the
porous matrix.[21] The gradient pulse duration, d, was set to 1 ms. NMR
nuclear spin relaxation times T1 and T2 were measured using standard in-
version recovery[39] and CPMG[40] techniques, respectively. NMR samples
were prepared using 5 mm NMR tubes. For bulk liquid measurements,
the tube was filled with liquid to a height of approximately 10 mm. Solid
samples were prepared by soaking catalyst grains (average diameter of
2–3 mm) in the liquid for at least 24 h. The catalyst was then dried on
pre-soaked filter paper to remove any excess liquid on the external sur-
face and finally transferred into 5 mm NMR tubes. To ensure a saturated
atmosphere into the NMR tube, hence minimizing errors due to evapora-
tion of volatile liquids, a small amount of pure liquid was placed onto ab-
sorbent filter paper, which was then placed under the cap of the NMR
tube. All the NMR measurements were carried out at atmospheric pres-
sure and 20ꢁ0.18C.
[8] I. S. Nielsen, E. Taarning, K. Egeblad, R. Madsen, C. H. Christen-
sen, Catal. Lett. 2007, 116, 35–40.
[9] E. Taarning, A. T. Madsen, J. M. Marchetti, K. Egeblad, C. H. Chris-
tensen, Green Chem. 2008, 10, 408–414.
[10] N. Dimitratos, J. A. Lopez-Sanchez, S. Meenakshisundaram, J. M.
Anthonykutty, G. Brett, A. F. Carley, S. H. Taylor, D. W. Knight,
G. J. Hutchings, Green Chem. 2009, 11, 1209–1216.
[11] C. Bianchi, F. Porta, L. Prati, M. Rossi, Top. Catal. 2000, 13, 231–
236.
[12] S. De Rossi, G. Ferraris, S. Fremiotti, A. Cimino, V. Indovina, Appl.
Catal. A 1992, 81, 113–132.
[13] H. G. Manyar, D. Weber, H. Daly, J. M. Thompson, D. W. Rooney,
L. F. Gladden, E. H. Stitt, J. J. Delgado, S. Bernal, C. Hardacre, J.
Catal. 2009, 265, 80–88.
[14] H. Takagi, T. Isoda, K. Kusakabe, S. Morooka, Energy Fuels 1999,
13, 1191–1196.
[15] S. Mukherjee, M. A. Vannice, J. Catal. 2006, 243, 108–130.
[16] P. Lignier, S. Mangematin, F. Morfin, J. L. Rousset, V. Caps, Catal.
Today 2008, 138, 50–54.
Mixtures of 5% mole fraction 1,4-butanediol in methanol without and
with the addition of water (water/methanol molar ratio of 1:1) were used.
Lower compositions did not allow a clear identification of the aliphatic
-CH2- peak of 1,4-butanediol from the other resonances of the mixture.
Mixtures were prepared without and with base (i.e., sodium methoxide)
because in general, aerobic diol oxidations over gold catalysts are carried
out in the presence of base.[10] The experiments were designed such that
they were carried out under equilibrium conditions (i.e., when no reac-
tion occurs as we are looking at initial reaction mixtures). Therefore, the
initial set of measurements was done in the absence of base to avoid any
unwanted reaction. However, it is noted that at the temperature and
pressure conditions used in the NMR measurements, the reaction rate,
even in the presence of base, is expected to be negligible. This was
indeed confirmed by inspection of the NMR spectra and by catalytic re-
action studies in those conditions. Therefore, a second set of experiments
was done adding the base (molar ratio reactant/base being the same as
that used in the catalytic reactions). Since the two systems studied here
differ only in the nature of solvent (all other conditions are identical), it
is reasonable to assume that the addition of the same type and amount of
base will affect diffusion and relaxation properties of the reactant in a
similar way, hence the same trend in terms of solvent effect should be ob-
served, with or without base. The measurements in the presence of base
confirmed the validity of this assumption.
[17] L. F. Gladden, J. Mitchell, New J. Phys. 2011, 13, 035001.
[18] P. T. Callaghan, Principles of Nuclear Magnetic Resonance Microsco-
py, Oxford University Press, Oxford, 1993.
[19] D. Weber, J. Mitchell, J. McGregor, L. F. Gladden, J. Phys. Chem. C
2009, 113, 6610–6615.
[20] J. Mitchell, S. C. Stark, J. H. Strange, J. Phys. D 2005, 38, 1950–
1958.
[21] M. D. Mantle, D. I. Enache, E. Nowicka, S. P. Davies, J. K. Edwards,
C. D’Agostino, D. P. Mascarenhas, L. Durham, M. Sankar, D. W.
Knight, L. F. Gladden, S. H. Taylor, G. J. Hutchings, J. Phys. Chem.
C 2011, 115, 1073–1079.
[22] M. Dvoyashkin, R. Valiullin, J. Kꢃrger, Phys. Rev. 2007, 75, 041202.
[23] J. Rouquerol, D. Avnir, C. W. Fairbridge, D. H. Everett, J. M.
Haynes, N. Pernicone, J. D. F. Ramsay, K. S. W. Sing, K. K. Unger,
Pure Appl. Chem. 1994, 66, 1739–1758.
[24] S. Dixit, W. C. K. Poon, J. Crain, J. Phys. Condens. Matter 2000, 12,
L323–L328.
[25] Z. J. Derlacki, A. J. Easteal, A. V. J. Edge, L. A. Woolf, Z. Roksan-
dic, J. Phys. Chem. 1985, 89, 5318–5322.
[26] S. Dixit, J. Crain, W. C. K. Poon, J. L. Finney, A. K. Soper, Nature
2002, 416, 829–832.
[27] A. Wakisaka, S. Komatsu, Y. Usui, 26th International Conference
on Solution Chemistry (26 ICSC) 1999, 175–184.
[28] R. Li, C. DꢀAgostino, J. McGregor, J. H. P. Collins, E. P. J. Parrot,
M. D. Mantle, A. Zeitler, L. F. Gladden, 35th International Confer-
ence on Infrared, Millimeter and THz Waves 2010.
[29] A. Menjoge, J. Dixon, J. F. Brennecke, E. J. Maginn, S. Vasenkov, J.
Phys. Chem. B 2009, 113, 6353–6359.
[30] D. Xu, C. J. Tsai, R. Nussinov, Protein Eng. 1997, 10, 999–1012.
[31] C. DꢀAgostino, J. Mitchell, L. F. Gladden, M. D. Mantle, J. Phys.
Chem. C 2012, 116, 8975–8982.
Acknowledgements
The authors would like to acknowledge the Technology Strategy Board
for funding this work (Grant No.: TP/7/ZEE/6/I/N0262B). The Technolo-
gy Strategy Board is a government body, which promotes and supports
the development and exploitation of technology and innovation for the
benefit of UK business, to increase economic prosperity and improve
[32] T. Mallat, A. Baiker, Chem. Rev. 2004, 104, 3037–3058.
Chem. Eur. J. 2012, 00, 0 – 0
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
7
&
ÞÞ
These are not the final page numbers!

M. D. Mantle et al.
[33] M. Besson, P. Gallezot, Catal. Today 2000, 57, 127–141.
[34] K. M. Song, J. Mitchell, H. Jaffel, L. F. Gladden, J. Mater. Sci. 2010,
45, 5282–5290.
[35] N. Bloembergen, E. M. Purcell, R. V. Pound, Phys. Rev. 1948, 73,
679–746.
[38] R. M. Cotts, M. J. R. Hoch, T. Sun, J. T. Markert, J. Magn. Reson.
1989, 83, 252–266.
[39] E. Fukushima, S. W. Roeder, Experimental Pulse NMR, Addison-
Wesley, Reading, 1981.
[40] H. Y. Carr, E. M. Purcell, Phys. Rev. 1954, 94, 630–638.
[36] S. N. Lanin, E. V. Vlasenko, N. V. Kovaleva, F. M. Zung, Russ. J.
Phys. Chem. A 2008, 82, 2152–2155.
Received: May 31, 2012
[37] E. O. Stejskal, J. E. Tanner, J. Chem. Phys. 1965, 42, 288–292.
Published online: && &&, 0000
&
8
&
www.chemeurj.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Catalytic Oxidation of 1,4-Butanediol in Methanol over Au/TiO2 Catalyst
FULL PAPER
The solvent effect on the catalytic oxi-
dation of 1,4-butanediol over gold
nanoparticles supported on titania
(Au/TiO2) has been assessed and a det-
rimental effect on the reaction rate
upon water addition to the initial
methanol solvent was observed (see
figure). NMR diffusion and relaxation
studies revealed that water addition
leads to a significant decrease of diffu-
sion coefficient and adsorption
Solvent Effects
C. DꢀAgostino, G. L. Brett,
P. J. Miedziak, D. W. Knight,
G. J. Hutchings, L. F. Gladden,
M. D. Mantle* ................ &&&& — &&&&
Understanding the Solvent Effect on
the Catalytic Oxidation of 1,4-Butane-
diol in Methanol over Au/TiO2 Cata-
lyst: NMR Diffusion and Relaxation
Studies
strength of the reactant over the cata-
lyst surface, which in turn affects the
reaction rate of the process.
Chem. Eur. J. 2012, 00, 0 – 0
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
9
&
ÞÞ
These are not the final page numbers!