Lewis acid properties of some metal salts for lactic acid formation in water: 31P NMR spectroscopy with trimethylphosphine oxide as a molecular probe

Article abstract of DOI:10.1016/j.cattod.2013.10.071

³¹P NMR spectroscopy was applied to the evaluation of water-tolerant Lewis acid catalysis of various metal triflates and chlorides in water using trimethylphosphine oxide (TMPO) as a probe molecule. Direct interaction of the TMPO molecule with the Lewis acid in water changes the original ³¹P NMR chemical shift and line width. ScCl₃, Sc(OTf)₃ and In(OTf)₃ exhibit larger changes in the ³¹P chemical shift by the formation of a Lewis acid-TMPO complex than other triflates and chlorides. This is due to strong interaction between the Lewis acid and TMPO, which leads to the slow exchange between TMPO and H ₂O in ScCl₃, Sc(OTf)₃ and In(OTf)₃. Lewis acids with large change in the ³¹P chemical shift exhibit high catalytic activities for lactic acid formation from 1,3-dihydroxyacetone (1.3-DHA) and pyruvic aldehyde in water: Strong interaction with TMPO on the Lewis acid is a dominant factor for lactic acid formation.

Full text of DOI:10.1016/j.cattod.2013.10.071

Catalysis Today 226 (2014) 198–203
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Lewis acid properties of some metal salts for lactic acid formation in
31
water: P NMR spectroscopy with trimethylphosphine oxide as a
molecular probe
Yusuke Koitoa,b, Kiyotaka Nakajimaa,c, Ryota Hasegawa , Hisayoshi Kobayashi ,
Masaaki Kitano , Michikazu Hara
a
d
d
a,1
a,e,∗
Materials and Structures Laboratory, Tokyo Institute of Technology, Nagatsuta-cho 4259, Midori-ku, Yokohama 226-8503, Japan
Research Fellow of Japan Society for the Promotion of Science (JSPS), 5-3-1 Chiyoda-ku, Tokyo 102-0083, Japan
Japan Science and Technology Agency, PRESTO, 4-1-8 Honcho, Kawaguchi 332-0012, Japan
Department of Chemistry and Materials Technology, Kyoto Institute of Technology, Matsugasaki, Sakyo-ku, Kyoto 606-8585, Japan
Japan Science and Technology Agency, ALCA, 4-1-8 Honcho, Kawaguchi 332-0012, Japan
b
c
d
e
a r t i c l e i n f o
a b s t r a c t
31P NMR spectroscopy was applied to the evaluation of water-tolerant Lewis acid catalysis of various
metal triflates and chlorides in water using trimethylphosphine oxide (TMPO) as a probe molecule. Direct
Article history:
Received 8 August 2013
Received in revised form 10 October 2013
Accepted 28 October 2013
Available online 26 November 2013
31
interaction of the TMPO molecule with the Lewis acid in water changes the original P NMR chemical
31
shift and line width. ScCl3, Sc(OTf)3 and In(OTf)3 exhibit larger changes in the P chemical shift by
the formation of a Lewis acid–TMPO complex than other triflates and chlorides. This is due to strong
interaction between the Lewis acid and TMPO, which leads to the slow exchange between TMPO and
Keywords:
3
1
H2O in ScCl3, Sc(OTf)3 and In(OTf)3. Lewis acids with large change in the P chemical shift exhibit high
catalytic activities for lactic acid formation from 1,3-dihydroxyacetone (1.3-DHA) and pyruvic aldehyde
in water: Strong interaction with TMPO on the Lewis acid is a dominant factor for lactic acid formation.
© 2013 Elsevier B.V. All rights reserved.
Water-tolerant Lewis acid catalyst
Lactic acid formation
31
P NMR
Trimethylphosphine oxide (TMPO).
1
. Introduction
Lewis acid catalysts such as AlCl , BF3 and TiCl4 have been uti-
acid catalysts therefore requires that a strictly anhydrous organic
solvent is employed.
Some metal trifluoromethanesulfonates (–OTf: –OSO CF ) and
3
2
3
lized in the chemical industry for carbon–carbon bond forming
reactions, represented by the Friedel–Crafts alkylation and acyla-
tion of petroleum-derived aromatic compounds [1,2]. Lewis acids
activate the nucleophilic functional group in a reactant molecule by
the formation of a Lewis acid–nucleophile adduct and subsequently
promotes various carbon–carbon bond forming reactions. The high
stability of the Lewis acid–nucleophile adduct causes deactivation
of the original Lewis acidity [3]. As a result, stoichiometric amounts
of Lewis acid catalysts are usually required for the reaction system.
In addition, most Lewis acids are easily decomposed into the cor-
responding metal hydroxides in the presence of water, which also
results in a loss of the original Lewis acid reactivity. The use of Lewis
chlorides, represented by Sc(OTf) , have been reported to function
3
as water-tolerant Lewis acid catalysts [4]. Water, a green reaction
medium, is desirable touseasasolvent forvariousorganicreactions
to reduce the environmental load. Metal triflates are effective Lewis
acid catalysts for typical carbon–carbon bond forming reactions,
such as the Mukaiyama-aldol reaction, the Mannich-type reaction,
and allylation reaction, in the presence of water [4–7]. The catalytic
activity of metal triflates for various carbon–carbon bond form-
ing reactions is strongly related to the metal species. Kobayashi
et al. proposed that the hydrolysis constants (pK ) and water
h
exchange rate constants (WERC) of metal triflates and chlorides are
dependent on the catalytic performance for the Mukaiyama-aldol
reaction [8]. Metal salts with small pKh are easily decomposed in
water into the corresponding metal hydroxides, and metal salts
with large pKh show less interaction of electrophilic molecules,
including water, and are thus non-catalytsis. On the other hand, a
Lewis acid with a large WERC is likely to exchange water molecules
upon hydration of the Lewis acid with other nucleophilic molecules,
including reactants with carbonyl groups, which exhibit high
∗ Corresponding author. Tel.: +81 45 924 5311; fax: +81 45 924 5381.
E-mail address: mhara@msl.titech.ac.jp (M. Hara).
Materials Research Center for Element Strategy, Tokyo Institute of Technology,
Nagatsuta-cho 4259, Midori-ku, Yokohama 226-8503, Japan.
1
0
920-5861/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.cattod.2013.10.071

Y. Koito et al. / Catalysis Today 226 (2014) 198–203
199
catalytic performance in water. Although the Lewis acid catalysis
of metal triflates can be inclusively discussed with respect to two
factors, the direct interaction of Lewis acid sites with nucleophilic
molecules in water has not been investigated to date.
Catalytic activity was tested by the conversion of 1,3-
dihydroxyacetone (1,3-DHA) into lactic acid in water. The reaction
involves the dehydration of 1,3-DHA into pyruvic aldehyde and the
subsequent isomerization of pyruvic aldehyde into lactic acid [9].
Lewis acids were found to promote the isomerization of pyruvic
aldehyde in water much more effectively than Brønsted acids,
because Lewis acids are capable of forming a stable 5-membered
Lewis acid–pyruvic aldehyde complex [9]. Therefore, the
development of optimal Lewis acid catalysts for coordinating
with pyruvic aldehyde is required for lactic acid synthesis.
In this study, we examine the correlation of the catalytic activ-
ity of some metal triflates and chlorides containing various metal
species (Sc, Y, La, Lu, In, and Zn) with their Lewis acidity in water.
Lewis acidity of metal salts in water was studied using 31P NMR
spectroscopy with trimethylphosphine oxide (TMPO) as a probe
3
1
molecule. P NMR spectroscopy with TMPO has been widely used
for characterization of the type (Brønsted and Lewis) and strength
of acid sites on solid surfaces [10,11]. This is due to the 100% nat-
3
1
ural abundance of P atoms and the large chemical shift range
over 430 ppm. Trimethylphosphine (TMP) has been conventionally
applied to the evaluation of Lewis acidity for solid catalysts [10,11].
However, TMP is only available for the estimation of Lewis acids
with same metal center [12]. In addition, oxidation of TMP into
TMPO with O2 easily occurs over Lewis acid sites, which restrict
the use of TMP as a probe molecule in this study. Yang et al. applied
2. Experimental
2.1. Catalytic reaction
Lewis acid catalysis of metal triflates and chlorides was exam-
ined through lactic acid production from 1,3-DHA and pyruvic
aldehyde in water. Sc(OTf) , In(OTf) , Y(OTf) , Lu(OTf) , La(OTf) ,
3
3
3
3
3
3
1
liquid-state P NMR with TMPO to evaluate the acid strength of
various homogeneous Brønsted acid catalysts [13]. They revealed
and Zn(OTf)2 obtained from Aldrich were used as metal triflates
in this study. Prior to NMR measurements and catalytic reaction
experiments, the metal triflates were dehydrated at 423 K for 3 h
under vacuum to remove physisorbed water. ScCl ·6H O (Strem
3
1
that the P NMR chemical shift of TMPO reflects the Brønsted acid
strength. TMPO is therefore also expected to be an excellent probe
molecule available to evaluate the Lewis acidity of metal salts in
water.
3
2
Chemicals Inc.), YCl (Strem Chemicals Inc.) and LaCl ·7H O (Wako)
3
3
2
were also tested as received.
Fig. 1. Time course for the transformation of 1.3-DHA into lactic acid with (a) In(OTf)3, (b) H2SO4 and (c) H3PO4. Reaction conditions: catalyst 0.1 mmol, 2.0 mL of 0.1 M
.3-DHA aqueous solution, 383 K.
1
Scheme 1. Possible reaction pathway for Lewis acid catalyzed transformation of 1.3-DHA to lactic acid.

2
00
Y. Koito et al. / Catalysis Today 226 (2014) 198–203
The catalytic reaction was performed by heating a sealed Pyrex
and H PO4 only produce pyruvic aldehyde with ca. 95% selectiv-
3
tube containing a mixture of catalyst (0.1 mmol) and 2.0 mL of 0.1 M
aqueous pyruvic aldehyde or 1.3-DHA at 383 K for 1 h. The resulting
solutions were analyzed using high performance liquid chromatog-
raphy (HPLC) equipped with a refractive index (RI) detector.
ity. H SO4 gives 80% 1,3-DHA conversion and pyruvic aldehyde
2
yield after 2 h. H PO4 exhibits lower catalytic performance than
3
H SO , which indicates that the strong Brønsted acid facilitates the
2
4
dehydration of 1,3-DHA. It should be noted that these Brønsted
acids cannot produce lactic acid under the present reaction con-
ditions. The difference in catalysis for the reaction between the
Brønsted and Lewis acids is attributed to the reaction pathway for
the isomerization of pyruvic aldehyde. Scheme 1 shows the pro-
posed reaction pathway for the isomerization of pyruvic aldehyde
into lactic acid over a Lewis acid. Pyruvic aldehyde is firstly pro-
duced from 1,3-DHA by dehydration. 1H NMR analysis confirmed
that pyruvic aldehyde is a highly reactive dicarbonyl compound
and present in water as the original aldehyde, monohydrated, and
dihydrated forms with typical distributions of trace levels, 57% and
43%, respectively (monohydrated pyruvic aldehyde: ␦ 2.20 ppm
(3H), 5.18 ppm (1H), dihydrated pyruvic aldehyde: ␦ 1.27 ppm (3H),
4.73 ppm (1H)) [16,17]. Monohydrated pyruvic aldehyde is subse-
quently converted into lactic acid through hydride transfer over the
Lewis acid. As the reaction proceeds, most of the dihydrated pyruvic
aldehyde is continuously converted into the monohydrated form
to maintain an equilibrium distribution, and the monohydrated
pyruvic aldehyde is finally converted into lactic acid. In the case
of a Brønsted acid, the reaction proceeds through a methyl anion
transfer via a benzilic acid rearrangement [16]. This reaction can-
not be promoted under mild conditions. On the other hand, it has
.2. 31P NMR analysis
2
Direct interaction of Lewis acids with TMPO in water was eval-
3
1
uated using liquid-state P NMR spectroscopy (Bruker, AVANCE
III 400) at resonance frequencies of 400.1 MHz and 162.0 MHz for
1
31
H and P, respectively. Same quantity of metal salt and TMPO
was dissolved in D O (Metal salt and TMPO concentrations, 0.25 M;
2
metal salt/TMPO ratio, 1). The Lewis acidities of the metal salts
were compared according to the chemical shift and line width of
3
1
the TMPO signal in the P NMR spectra. NMR measurements were
1
performed at 303 K under H decoupling to remove the influence
1
31
of H- P spin-spin coupling. The number of scans and the recycle
31
delay were set at 32 times and 5 s, respectively. The P chemical
shift was referenced using 0.25 M TMPO/D O solution at 53.5 ppm
2
as an external standard relative to 85% H PO aq. at 0 ppm.
3
4
3
1
Solid state P MAS NMR experiment was carried out with ECA
00 spectrometer running at resonance frequencies and 399.0 MHz
and 161.5 MHz for H and P, respectively, with a T6 double-tune
probe and 7 mm zirconia rotor. NMR spectra were acquired with
4
1
31
1
high-power H decoupling, a recycle delay of 10 s, a ␲/2 pulse
length of 5.5 us, 128 times scans and a spinning speed at 8 k Hz.
31
8
5% H PO aq. was used for the external reference of P chemical
3 4
shift at 0 ppm.
Adsorption of TMPO on solid Sc(OTf)3 and In(OTf)3 was per-
formed by simple impregnation technique. TMPO was dissolved
in anhydrous CH Cl and prepared 0.1 M TMPO/CH Cl solution.
2
2
2
2
Dehydrated Sc(OTf)3 and In(OTf)3 at 323 K for 1 h under vacuum
were added to TMPO-containing CH Cl solution (0.1 M) under an
2
2
argon atmosphere. After removal of the solvent under vacuum,
TMPO-adsorbed Sc(OTf)3 and In(OTf)3 were obtained. The sample
was packed into the zirconia rotor in glove box.
2.3. Density functional theory (DFT) calculations
Molecular structures for TMPO, mono-hydrated pyruvic alde-
hyde, and 1.3-DHA were optimized with consideration of the
solvation effects (of water). The self-consistent reaction field
method was implemented using Gaussian 03 software [14] with
the polarizable continuum model. The B3LYP hybrid functional and
6
-311G basis sets were used in this case. Contour maps for the high-
est occupied molecular orbitals (HOMOs) were constructed using
GaussView software [15]. The contour surfaces are shown at an
−3
electron density of 0.15 e A˚
.
3
. Results and discussion
3.1. Catalysis using metal triflates and chlorides for lactic acid
production from 1,3-DHA and pyruvic aldehyde in water
The catalytic activity for lactic acid formation from 1,3-DHA in
water was first examined with typical Brønsted and Lewis acid
catalysts. Fig. 1 shows time courses for 1,3-DHA conversion and
product yields in the presence of In(OTf) , H SO and H PO . The
3
2
4
3
4
reaction proceeds through the dehydration of 1,3-DHA into pyru-
vic aldehyde and the subsequent isomerization of pyruvic aldehyde
into lactic acid (Scheme 1). In(OTf) , a typical water-tolerant Lewis
3
acid, consumes almost 1.3-DHA within 2 h and gives a lactic acid
yield of ca. 95%. This means that water-tolerant Lewis acid catal-
Fig. 2. Catalytic activities for the transformation of (a) pyruvic aldehyde and (b)
1.3-DHA on various metal salts. Reaction conditions: catalyst 0.1 mmol, 2.0 mL of
ysis is effective for the consecutive reactions. In contrast, H SO4
0.1 mM pyruvic aldehyde or 1.3-DHA aqueous solution, 383 K, 1 h.
2

Y. Koito et al. / Catalysis Today 226 (2014) 198–203
201
been reported that Lewis acids can effectively catalyze the hydride
transfer of glucose into fructose [18,19] and pyruvic aldehyde into
lactic acid [16] in water through the Meerwein–Ponndorf–Verley
and In(OTf)3 have lactic acid yields of 59%, 50%, and 84%, respec-
tively. These results indicate that the water-tolerant Lewis acids are
effective catalysts for lactic acid formation from 1,3-DHA in water,
which is strongly dependent on the metal center.
(
MPV) reduction mechanism. In the case of In(OTf) , the reac-
3
tion is expected to proceed by the formation of a 5-membered
3
3.2. 31P NMR measurement with TMPO as a basic probe molecule
intermediate with In(OTf) , followed by lactic acid formation via
a 1,2-hydride transfer reaction of the monohydrated pyruvic alde-
hyde, as in the presence of other Lewis acids (Scheme 1).
Catalysis over some metal triflates (metal center: Sc, Y, La, Lu, In,
The Lewis acid properties of metal triflates and chlorides were
elucidated using 31P NMR measurements with TMPO as a probe
molecule. Activation of the reactant by the Lewis acid is closely
related to the energy level of the HOMO of the nucleophile and the
lowest unoccupied molecular orbital (LUMO) of the electrophile
[3]. A small HOMO–LUMO energy gap between the Lewis acid
and nucleophile enables the formation of a Lewis acid–nucleophile
complex. Fig. 3a shows a HOMO and LUMO energy diagram for
Zn), chlorides (metal center: Sc, Y, La) and Brønsted acids (H SO4
2
and H PO ) was also examined by lactic acid formation from pyru-
3
4
vic aldehyde and 1,3-DHA in water, and the results are shown in
Fig. 2. While the Brønsted acids show no catalysis for the reaction,
all the Lewis acids tested can convert pyruvic aldehyde into lactic
acid. Sc(OTf) , ScCl and In(OTf)3 have high catalytic performance
3
3
for the reaction; lactic acid yield and selectivity of these catalysts
reached 87% and 90%, 83% and 85%, and 95% and 96%, respectively as
shown in Fig. 2a. On the other hand, the other triflates and chlorides
have lower lactic acid yields (10%–55%) and selectivities (20%–65%)
TMPO, 1,3-DHA, monohydrated pyruvic aldehyde, and H O on the
basis of density functional theory (DFT) calculations. The higher
HOMO energies of these nucleophilic molecules, including TMPO,
2
than that of H O (−8.1 eV) means that the Lewis acids interact with
2
than Sc(OTf) , ScCl and In(OTf) . Fig. 2b shows the catalytic activi-
these reactant molecules in preference to H O molecules. It should
3
3
3
2
ties of the Lewis acids for the conversion of 1,3-DHA into lactic acid
at 383 K for 1 h. While Y(OTf) , La(OTf) , and Lu(OTf) have high 1,3-
DHA conversion of ca. 70%, the combined yield of pyruvic aldehyde
be noted that TMPO has a similar HOMO energy level (–7.5 eV) to
1,3-DHA (–7.1 eV) and monohydrated pyruvic aldehyde (–7.4 eV).
In addition, the electrons in the HOMO of TMPO are located on the
oxygen atom of the phosphoryl group as well as the HOMOs of
1,3-DHA and monohydrated pyruvic aldehyde, which electrons are
found on the oxygen atom of carbonyl groups (Fig. 3b). Such an elec-
tronic analogy among these nucleophiles indicates that TMPO is a
suitable probe molecule for the evaluation of Lewis acid catalysis for
these reactions. Fig. 4 shows 31P NMR spectra of TMPO in the pres-
ence of equimolecular amounts of metal triflates and chlorides in
D O. TMPO in D O have one resonance at 53.5 ppm that is assigned
3
3
3
and lactic acid for all the tested catalysts, except for Sc(OTf) , ScCl ,
3
3
and In(OTf) , are below 30%. 1,3-DHA and pyruvic aldehyde have
3
reactive carbonyl (C=O) groups, so that complex intermolecular
reactions, such as aldol condensation, by Lewis acid catalysts would
result in the formation of complex polymerized species. Therefore,
in the case of these Lewis acids, the formation of polymerized by-
products proceed preferentially to the dehydration of 1,3-DHA and
hydride transfer of pyruvic aldehyde. On the other hand, ScCl3 and
In(OTf)3 have higher lactic acid yields, although Sc(OTf)3 and ScCl3
exhibit low lactic acid selectivity in contrast to the case of hydride
transfer of pyruvic aldehyde, as shown in Fig. 2a: Sc(OTf) , ScCl ,
2
2
to the interaction of TMPO with water molecules (Scheme 2b). The
addition of Sc(OTf) , ScCl , and In(OTf) to the TMPO/D O solution
3
3
3
2
3
3
results in a large shift of the TMPO signal to the higher frequency
Fig. 3. (a) HOMO and LUMO energy diagrams for TMPO, monohydrated pyruvic aldehyde, 1.3-DHA and water, and (b) schematic structure of the HOMOs for (i) TMPO, (ii)
monohydrated pyruvic aldehyde, and (iii) 1.3-DHA.

2
02
Y. Koito et al. / Catalysis Today 226 (2014) 198–203
exchange rate between the two species; therefore, Sc(OTf) , ScCl3
3
and In(OTf)3 are capable of forming stable TMPO–Lewis acid com-
plexes in water. In contrast, Y(OTf) , La(OTf) , Lu(OTf) , Zn(OTf) ,
3
3
3
2
YCl3 and LaCl3 have a sharp TMPO signal at 53–55 ppm (Fig. 4(c),
31
(
d), (e), (g), (i) and (j)). The small change in the P chemical shift
suggests that the TMPO molecules coordinated with these metal
triflates are readily replaced with H O molecules. As a result, most
2
of the Lewis acid in water is present in the hydrated state. This is
due to the weak interaction of the metal center with TMPO. Typical
Brønsted acids (H SO4 and H PO ) were also studied as reference
2
3
4
3
1
materials, which showed small P chemical shift change under
same experimental conditions.
Therefore, the Lewis acids tested here are categorized into
two groups based on the 31P NMR measurements with TMPO;
(
i) Sc(OTf) , ScCl3 and In(OTf)3 with a large change in 31P chem-
3
ical shift for TMPO and (ii) Y(OTf) , YCl , Lu(OTf) , La(OTf) , LaCl
3
3
3
3
3
31
and Zn(OTf)2 with a small or no change in P chemical shift for
TMPO. The former form Lewis acid–TMPO complexes in water more
preferentially than the latter. Such Lewis acids effectively activate
the reactant molecules by the formation of a Lewis acid–reactant
complex and thereby exhibit high catalytic performance for lactic
acid formation from 1,3-DHA and pyruvic aldehyde in water. Lewis
Fig. 4. 31P NMR spectra for TMPO in TMPO/D2O solution with various Lewis acid
catalysts. The concentrations of Lewis acid (cacid) and TMPO (cTMPO) were set at
acidity of solid Sc(OTf)3 and In(OTf) was examined by adsorp-
3
tion of TMPO using solid-state 31P MAS NMR. Fig. 5 shows
31
P
0
.25 M (Lewis acid/TMPO = 1). (a) TMPO without Lewis acid catalyst, (b) Sc(OTf)3,
MAS NMR spectra for TMPO-adsorbed Sc(OTf)3 and In(OTf)3 in
dried powder form. TMPO resonance on Sc(OTf)3 and In(OTf)3
appears at 67.7 and 65.4 ppm, respectively, which are comparable
(
c) Y(OTf)3, (d) La(OTf)3, (e) Lu(OTf)3, (f) In(OTf)3, (g) Zn(OTf)2, (h) ScCl3, (i) YCl3, (j)
LaCl3, (k) H2SO4 and (l) H3PO4.
to those on dissolved Sc(OTf)3 and In(OTf)3 in D O (Fig. 4). Small
2
region. Fig. 4(b), (f) and (h) show TMPO signals on Sc(OTf) , ScCl ,
3
3
difference of TMPO resonance on dissolved metal triflate in D O
2
and In(OTf)3 appear at 64.5, 63.4, and 61.9 ppm, respectively. This
change can be simply interpreted by the formation of a Lewis
acid–TMPO complex and hydrous TMPO in equilibrium, as shown
in Scheme 2. The single and broad peak in Fig. 4(b), (f), and (h) is
is due to fast exchange between Lewis acid–TMPO and hydrous
TMPO as shown in Scheme 2. No decrease in Lewis acidity of
Sc(OTf)3 and In(OTf)3 in water strongly suggests that the exchange
between water and nucleophilic molecule occurs on these metal
triflate.
In contrast, the latter Lewis acids have weak interaction with
the nucleophilic molecules of TMPO, 1,3-DHA and pyruvic alde-
hyde, which causes lower catalytic performance for the reactions.
This difference is due to strong/weak interaction between the
Lewis acid and the nucleophile, which results in much slower
due to the slow exchange between the Lewis acid and D O on TMPO
2
in solution. In this case, the peak position of TMPO (˝peak) is given
by Eq. (1), where [1]eq, [2]eq and ˝01, ˝ denote the equilibrium
0
2
concentrations and the original chemical shifts of species 1 and 2,
respectively [20].
1]eq˝ + [2]eq˝20
0
[
1
˝peak
=
(1)
exchange between the nucleophile and H O for the former Lewis
2
[
1] + [2]eq
eq
acids than for the latter Lewis acids. Preliminary result on density
A large change in chemical shift indicates that the preference of
most TMPO molecules is to form TMPO–Lewis acid complexes as
opposed to with H O molecules. The line width of the TMPO sig-
2
nal on Sc(OTf) , ScCl and In(OTf)3 is considerably broader than
3
3
those of other Lewis acids. While the change in chemical shift
reflects the affinity between the Lewis acid and TMPO, the line
width of the TMPO resonance provides information on the chem-
ical exchange between a Lewis acid–TMPO complex and hydrous
TMPO (Scheme 2). The broad NMR resonance is indicative of a slow
Scheme 2. Schematic illustration of the chemical equilibrium between the
M(OTf)x–TMPO complex and hydrous TMPO.
Fig. 5. 31P MAS NMR spectra for TMPO loading on (a) Sc(OTf)3 and (b) In(OTf)3.

Y. Koito et al. / Catalysis Today 226 (2014) 198–203
203
functional theory (DFT) calculation suggests that lowest unoccu-
References
pied molecular orbital (LUMO) levels of Sc(OTf)3 and Y(OTf)3 were
estimated to be ca. −2.2 and −1.0 eV, respectively, from crystalline
Sc(OTf) ·9H O [21] and Y(OTf) ·9H O [22] structure. Higher HOMO
[
1] D. Schinzer, Selectivities in Lewis Acid Promoted Reactions, Kluwer Academic
Publishers, Dordrecht, The Netherlands, 1989.
3
2
3
2
[2] H. Yamamoto, Lewis Acid in Organic Synthesis, Wiley, Weinheim, Germany,
2000.
[
[
[
[6] S. Kobayashi, K. Manabe, Acc. Chem. Res. 35 (2002) 209–217.
[
[
levels of pyruvic aldehyde and TMPO than that of H O means that
2
3] A. Corma, H. García, Chem. Rev. 103 (2003) 4307–4365.
4] S. Kobayashi, Eur. J. Org. Chem. 1 (1999) 15–27.
5] S. Kobayashi, I. Hachiya, J. Org. Chem. 59 (1994) 3590–3596.
Lewis acid is bonded to nucleophiles such as 1,3-DHA, pyruvic
aldehyde, and TMPO in preference to H O. In addition, low LUMO
2
level of Sc(OTf)3 suggests that the interaction between Sc(OTf)3
7] S. Kobayashi, C. Ogawa, Chem. Eur. J. 12 (2006) 5954–5960.
8] S. Kobayashi, S. Nagayama, T. Busujima, J. Am. Chem. Soc. 120 (1998)
and these nucleophiles is larger than that of Y(OTf) , due to small
3
HOMO–LUMO energy gap. Strong interaction of reactant molecules
with Sc(OTf)3 would form active Lewis acid–reactant complexes,
resulting in high catalytic performance.
8287–8288.
[9] Y. Hayashi, Y. Sasaki, Chem. Commun. (2005) 2716–2718.
[10] Y. Jiang, J. Huang, W. Dai, M. Hunger, Solid State Nucl. Magn. Reson. 39 (2011)
116–141.
[
[
[
11] A. Zheng, S.J. Huang, S.B. Liu, F. Deng, Phys. Chem. Chem. Phys. 13 (2011)
14889–14901.
12] Y. Chu, Z. Yu, A. Zheng, H. Fang, H. Zhang, S.J. Huang, S.B. Liu, F. Deng, J. Phys.
Chem. C 115 (2011) 7660–7667.
13] C.Y. Yang, C.C. Chang, N. Feng, S.J. Huang, A. Zheng, Y.C. Chang, K.C. Lee, F. Deng,
S.B. Liu, J. Anal. Sci. Tech. 2 (2011) A155–A158.
4
. Conclusion
The changes in the 31P NMR chemical shift and line width
by the formation of a TMPO–Lewis acid complex in water are
effective indicators to evaluate the catalytic performance of homo-
geneous water-tolerant Lewis acids in water. Sc(OTf) , ScCl , and
In(OTf)3 show 61, 51, and 96% yield for lactic acid formation,
respectively. High performance of ScCl , Sc(OTf) , and In(OTf)
was evident from the large chemical shift and broad line width
of the TMPO–Lewis acid complexes observed in the 31P NMR
spectra, which indicate the formation of stable reactant–Lewis
acid complexes in water. Similar HOMO level and electronic state
of TMPO to carbonyl group in reactant (1.3-DHA and pyruvic
aldehyde), in addition to 100% natural abundance of 31P atoms,
make TMPO as an effective probe molecule for the evaluation of
Lewis acid catalysis in water. Efficient activation of 1.3-DHA and
pyruvic aldehyde as well as TMPO in water by the Lewis acids
results in high catalytic performance for Lewis acid-catalyzed reac-
tions.
[14] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheese-
man, J.A. Montgomery, Jr. T. Vreven, K.N. Kudin, J.C. Burant, J.M. Millam, S.S.
Iyengar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G.A.
Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa,
M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J.E. Knox, H.P.
Hratchian, J.B. Cross, V.Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R.E. Strat-
mann, O. Yazyev, A.J. Austin, R. Cammi, C. Pomelli, J.W. Ochterski, P.Y. Ayala, K.
Morokuma, G.A. Voth, P. Salvador, J.J. Dannenberg, V.G. Zakrzewski, S. Dapprich,
A. D. Daniels, M.C. Strain, O. Farkas, D.K. Malick, A.D. Rabuck, K. Raghavachari,
J.B. Foresman, J.V. Ortiz, Q. Cui, A.G. Baboul, S. Clifford, J. Cioslowski, B.B. Ste-
fanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R.L. Martin, D.J. Fox, T. Keith,
M.A. Al-Laham, C.Y. Peng, A. Nanayakkara, M. Challacombe, P.M.W. Gill, B. John-
son, W. Chen, M.W. Wong, C. Gonzalez, J.A. Pople, Gaussian 03, Revision E.01,
Gaussian, Inc., Wallingford, CT, USA, 2004.
3
3
3
3
3
[
15] R. Dennington, T. Keith, J. Millam, Gauss View, Version 5, Semichem Inc.,
Shawnee Mission, KS, 2009.
[16] Y. Koito, K. Nakajima, M. Kitano, M. Hara, Chem. Lett. 42 (2013) 873–875.
[17] I. Nemet, D. Viki c´ -Topi c´ , L. Varga-Defterdarovi c´ , Bioorgan. Chem. 32 (2004)
5
60–570.
18] E.A. Pidko, V. Degirmenci, R.A. van Santen, E.J.M. Hensen, Angew. Chem. Int. Ed.
9 (2010) 2530–2534.
[19] Y. Román-Leshkov, M. Moliner, J.A. Labinger, M.E. Davis, Angew. Chem. Int. Ed.
9 (2010) 8954–8957.
[
4
Acknowledgment
4
[
[
20] M.H. Levitt, Spin dynamics, Wiley, Chichester, UK, 2008, pp. 516–527.
21] A. Abbasi, P. Lindqvist-Reis, L. Eriksson, D. Sandström, S. Lidin, I. Persson, M.
Sandström, Chem. Eur. J. 11 (2005) 4065–4077.
We thank Prof. Takashi Tatsumi and Dr. Toshiyuki Yokoi of
Chemical Resources laboratory in Tokyo Institute of Technology for
solid state NMR measurement.
[
22] D.L. Kepert, J.M. Patrick, A.H. White, Aust. J. Chem. 36 (1983) 483–492.