M. Hronec, et al.
AppliedCatalysisA,General594(2020)117471
carbonate by the reaction of calcium hydroxide with Na2CO3, which
was then used as the palladium support. XRD revealed that the crystal
structure of the CaCO3 is calcite (PDF-01-072-1651). In the XRD pattern
of the Pd/CaCO3 catalyst, there are diffraction peaks indicating the
for the furfural ring fragment, was absent (Fig. S4). Surprisingly, de-
spite very mild reaction conditions and short reaction times the furan
ring is completely hydrogenated. Therefore, during hydrogenation of
furfural in lower alcohols, using 3%Pd/CaCO3 catalyst, the possibility
of the formation of the furfuryl type compounds (acetals, ethers) as the
main products is excluded. On the other hand, under identical reaction
conditions and using the same palladium catalyst, only 42.5 % of fur-
furyl ethyl ether (FEE) was converted into tetrahydrofurfuryl ethyl
ether in ethanolic solution (Table 1, Entry 5, Figs. S5 and S6b-e).
presence of palladium particles with mean diameter 5.93
0.15 nm.
The peaks at 2θ values of 40.1°, 46.6° and 68.1° are characteristic (111),
(200) and (220) planes of isolated palladium crystalline particles. The
surface area of the supported palladium catalyst was 11 m2 g−1. The
amount of basic sites measured through the titration using bromthymol
blue as the indicator was 0.039 mmol.g−1
.
3.2.1. Nuclear magnetic resonance analysis
3.2. Reductive acetalization and identification of products
For the identification of the unknown products NMR analysis was
used. All measurements were done at 25 °C with approx. 500 μl of
sample was dissolved in deuterated DMSO. The usage of DMSO allowed
the direct observation of OH groups of hemiacetals and geminal diols.
Besides basic 1H and 13C spectra, 1D selective TOCSY also 2D various
correlation NMR experiments (1H-1H COSY, 1H-13C (one-bond corre-
lation) HSQC and 1H-13C (multiple-bond correlation) HMBC were used.
All pulse sequences were provided by the spectrometer vendor. NMR
allowed unambiguous identification of all products. Complete assign-
ment of 1H and 13C resonances as well as correlation signals in 2D
experiments is shown directly in the spectra. The hemiacetal poses 2
chiral centres (carbons C2 and C6) therefore they occur in 2 different
stereoisomers. This is reflected in NMR spectra as doubling of all sig-
nals, which is well visible on signals of OH groups in 1H spectra and
signals of all carbons in 13C spectra. The 1H and 13C NMR spectra of
tetrahydrofurfuryl hemimethylacetal and hemiethylacetals are shown
in Fig. 2 while their 2D correlation spectra can be found in the ESI, Figs.
S6a, S7a-c and S8a-c. According to the NMR results it can be un-
ambiguously confirmed that the main products of furfural hydrogena-
tion in methanol and ethanol in the presence of 3%Pd/CaCO3 catalyst
are corresponding tetrahydrofurfuryl hemialkylacetals.
The selectivity of furfural hydrogenation in excess of ethanol was
studied using 3%Pd/CaCO3 catalyst. The choice for the palladium cat-
alyst was due to the efficiency of this metal to reduce the carbonyl
group and the furan ring. The partial and total hydrogenation of fur-
fural arising from the reduction of the C]O group and the furan ring,
and the products of acetalization usually correlate with surface acid-
base properties of the catalyst support, and lead to different distribution
of reaction products of furfural hydrogenation in alcoholic solutions.
The choice for 3%Pd/CaCO3 as the catalyst was due to its basic prop-
erties and as is seen below for the low efficiency of this metal for the
reduction of the C]O group and the furan ring at very mild reaction
conditions, which eliminate undesired side reactions. As is evident from
the results in Table 1 (Entry 1) under these conditions Pd/CaCO3 cat-
alcohol is almost inactive for hydrogenation of carbonyl group and
furan ring. However, at the same condition furfural hydrogenation in
ethanol (96 %) as the solvent is very rapid and within 15 min 94.8 %
conversion of furfural is achieved. The gas chromatographic analysis of
the reaction mixture has shown that with very high selectivity and the
yield of 89.3 % an “unknown” product is formed (Table 1, Entry 2).
Surprisingly, very high yields and selectivities were achieved in similar
hydrogenation experiments performed in methanol and deuterated
methanol (CD3OD) but also in n-propanol 55.6 % yield of hemiacetal
was obtained (Table 1, Entry 3,4 and 5). The identification of these
used for the hydrogenation, the most abundant fragmentation ion ob-
served in the spectra was always m/z = 71, which is consistent with the
molecular formula C4H7O+, corresponding to fragment of a charged
tetrahydrofurfuryl ring (Figs. S2 and S3). In the spectra the ion typical
3.3. Effect of reaction conditions
The investigation of the effect of hydrogen pressure on the product
distribution has shown (Fig. 3) that using the 3% Pd/CaCO3 catalyst
furfural hydrogenation could be carried out even at very low pressure
(0.15 MPa). Under this pressure, during 20 min of reaction, about 90 %
furfural conversion and 80.2 % yield of desired product THFHA is
achieved. At higher hydrogen pressures the conversion of furfural and
THFHA yield increases. The maximal yield of the product is achieved at
0.3 MPa of hydrogen, and at higher pressure the yield is slightly de-
creasing, probably as a result of undesired competitive hydrogenation
of furfural into tetrahydrofurfuryl alcohol, which is accelerated by
higher pressures.
Table 1
Furfural hydrogenation in the presence of 3 % Pd/CaCO3 catalyst.
Conditions: 0.25 g furfural; 10 ml solvent; 0.07 g catalyst; reaction temperature 60 °C;
As evidenced by Fig. 4, the reaction temperature has a significant
from 40 to 60 °C the yield of THFHA increases from 78.5 % to 89.3 %,
while the yields of both competitive undesired products, FAL and THFA
decrease. It suggests that higher temperature accelerates more the rate
of acetalization and promotes the equilibrium of furfural acetalization
to the reaction intermediates, rather than the rates of competitive un-
desired hydrogenation reactions of furfural. However, increase of re-
action temperature to 75 °C influences negatively the selectivity to
THFHA, and the by-products comprise mainly hydrogenated com-
pounds.
0.3 MPa hydrogen
Entry
Solvent
CPME
React.
time
min
Conversion
%
FAL
THFA
THFHA
1
2
3
4
30
8.1
2.8
1.5
1.3
1.7
–
3.8
2.6
2.7
2.2
–
0
Ethanol, 96 % 15
97.8
95.1
94.7
89.3
Methanol
CD3OD
30
30
88.9
90.8
5
Ethanol, 96 % 30
42.5
–
6
n-Propanol
30
100
5.2
6.7
55.6
FUR = Furfural; FAL = Furfuryl alcohol; THFA = Tetrahydrofurfuryl alcohol;
THFHA = Tetrahydrofurfuryl hemiethylacetal; FEE = Furfuryl ethyl ether;
THFEE = tetrahydrofurfuryl ethyl ether.
The effect of catalyst loading is presented in Fig. 5. The experiments
were performed using ca. 5 wt% solution of furfural in ethanol (96 %).
The data showed that increasing the catalyst loading only from 12.5 to
8.6 (expressed as weight ratio of furfural to catalyst) led to a significant
increase in furfural conversion (from 27.0 to 85.3 %) and the yield of
THFHA, from 22.1 to 73.7 %. The maximal furfural conversion and the
yield of THFHA is achieved at the catalyst loading 6.6. Further increase
of catalyst loading to 3.1 led to a slight decrease of conversion and the
yield of desired product.
a
b
c
d
e
f
Determined by GC analysis.
Determined using response factor of THFEE.
The yield check by NMR spectroscopy.
THFHA.
FEE instead of FUR was used.
Conversion of FEE.
g
Tetrahydrofurfuryl hemipropylacetal.
3