356
V.R. Ruiz et al. / Journal of Catalysis 271 (2010) 351–357
lysts have been summarized in Table S3 and it can be seen the ben-
eficial role of gold compared to other catalysts.
5. Experimental section
5.1. Catalysts
3. Acetalization of n-heptanal with glycerol/water mixtures
using homogeneous acid Brønsted and Lewis catalysts
Beta-1 zeolite in the acidic form was supplied by PQ corpora-
tion. Zeolite Beta samples (Beta(F)-15, Beta(F)-30, Beta(F)-50, Be-
ta(F)-100) were synthesized in our laboratory according to
previous procedures [14], and their characteristics are given in Ta-
ble S1. Acidity measurements were taken by adsorption–desorp-
tion of pyridine by IR spectroscopy. The infrared spectra were
recorded in a Nicolet 710 FTIR using self-supported wafers of
10 mg cmÀ2. The calcined samples were outgassed overnight at
673 K and 10À3 Pa dynamic vacuum. Then, pyridine was admitted
into the cell at room temperature, and after saturation, the samples
were outgassed at 523 K for 1 h under vacuum, cooled to room
temperature, and the spectra recorded.
As it was mentioned before, obtaining an extractable hydropho-
bic acetal of glycerol would be of interest [10]. Thus, we have stud-
ied the acetalization reaction of glycerol with n-heptanal in
dioxane (Table 2). This aldehyde is a product of biomass, obtained
by thermal cracking of the castor oil methyl ester and can be used
to extract glycerol from glycerol–water mixtures [9].
As it can be seen in Table 2, a high yield of acetal was obtained
when AuCl3 was used as catalyst at room temperature, leading to
an acetal mixture of heptanal glyceryl acetal (HGA) with a ratio
of 34:66 (dioxane:dioxolane, entry 1). The addition of different
amounts of water (25–40 wt.% respect to glycerol, entries 2 and
3) and the use of NaAuCl4 (both without and with water, entries
4 and 5) required an increase in temperature in order to achieve
similar conversions. LaOTf gave a similar yield than NaAuCl4, with
or without water, although longer reaction times were required
(entries 6–7). The use of a Brønsted acid as HCl gave again a lower
yield than AuCl3 (75 instead of 93%) even when 25% of HCl was
used as catalyst (entries 8–9).
The AuPPh3NTf2 complex was obtained from Aldrich as a dimer-
toluene adduct [15]. AuCl3, NaAuCl4, trioxane, formalin, glycerol,
dioxane, and toluene (purity >99%) were purchased from Aldrich
and were used without further purification.
5.2. Reaction procedure for acetalization of formaldehyde and glycerol
with solid catalysts
We have seen up to now that gold catalysts give good results for
the formation of acetals of glycerol with both aldehydes, formalde-
hyde and n-heptanal. However, the price of gold can limit its use.
Therefore, we have studied the possibility of using recoverable
gold catalysts. Firstly, gold supported on TiO2 [11], CeO2 [12],
and Fe2O3 [11] were tested and, unfortunately, it was found that
they were not active as catalyst when reacting with trioxane and
Activation of the solid catalysts was performed in situ in closed
conic glass batch reactors by heating the solid under vacuum
(1 Torr) at 523 K (at 373 K in the case of Amberlyst-36) for 2 h.
After this time, the system was left at room temperature and then,
glycerol (625 mg, 6.6 mmol) and the formaldehyde source
(200 mg, 6.6 mmol) were added at atmospheric pressure. The sys-
tem was closed and heated up at the desired temperature with a
silicone oil bath equipped with an automated temperature control
system. After reaction, the system was cooled at room tempera-
ture, filtered under vacuum, and the catalyst was washed with
methanol. For kinetics, individual reactions were run for each time,
filtering the mixture after cooling at 0 °C and washing the catalyst
with methanol. Nitrobenzene was added to the combined filtrates
as external standard, and the solution was then analyzed. For reus-
ing, the catalyst was reactivated by calcination as indicated earlier.
Reaction products were analyzed by gas chromatography (GC)
(Varian 3900 equipped with an split–splitless injector, flame ioni-
glycerol. Alternatively, we used
a cationic gold complex,
AuPPh3NTf2, which takes advantage of being soluble in the reaction
mixture and can precipitate when n-hexane is added at the end of
the reaction, allowing in this way its recovery and recycling [13].
Indeed, when reacting with n-heptanal and glycerin, ꢀ90% yield
of acetal (dioxane to dioxolane ratio 44:56) was obtained (Fig. 7).
The complex was reused up to seven times without any observable
decrease in the acetal yield. ICP of the extracted products indicates
that the amount of gold leached at the end of the seven cycles was
less than 0.6% of the original amount.
The evolution of the gold(I) complex was followed during the
reuses by 31P NMR spectroscopy (Fig. S6). It was found that a
new signal at 45 ppm appears, corresponding to the diphosphine
complex [Au(PPh3)2]NTf2. This complex was prepared (see experi-
mental) and showed no catalytic activity for the reaction, indicat-
ing that all the activity comes from the remaining original
AuPPh3NTf2 complex.
zation detector, and a 30 m  0.25 mm  0.25
lm fused silica cap-
illary column (SUPELCO Equity 5)), GC–MS, and 1H NMR (300 MHz
Bruker). In all experiments, the recovered material was superior to
95 wt.%.
5.3. General procedure for acetalization reactions with Lewis acid
catalysts
4. Conclusion
Catalytic reactions were carried out in a closed glass reactor
(2.5 mL) equipped with a micro-syringe to extract samples for
analysis. To an appropriate catalytic amount of catalyst, a mixture
of the aldehyde (1 mmol) and glycerol (1 mmol) in 1 mL of dioxane
was added. The reaction mixture was stirred at 80–100 °C and
monitored by gas chromatography. Nitrobenzene and n-dodecane
were used as internal standard for reactions with formaldehyde
and n-heptanal, respectively. When the reaction was completed,
the solvent was evaporated under vacuum.
For the reaction performed with the recoverable AuPPh3NTf2
catalyst, the following procedure was applied: 78.4 mg (0.1 mmol)
of AuPPh3NTf2 and 184 mg (2 mmol) of glycerol were placed in an,
oven-dried, 10-mL round-bottomed flask. Air evacuation-nitrogen
pumping cycles were carried out, and a rubber septum was rapidly
fitted after the last nitrogen refilling. The system was kept under
N2 atmosphere. Then, n-heptanal (228 mg, 2 mmol) and anhydrous
Valorization of pure glycerol and glycerol in water has been
done by reaction with aldehydes to form valuable acetals. Solid
Brønsted acids such as zeolites (Beta) and resins (Amberlyst) were
used and compared with soluble acids such as p-toluenesulfonic
acid (PTSA). It has been found that in absence of water, PTSA and
resins are much more active than Beta zeolite. However, when
water is present (up to 31 wt.% H2O), a hydrophobic Beta zeolite
gives better results than either PTSA or Amberlyst. Nevertheless,
the best results with the aforementioned catalysts never exceeded
80%. However, a soft Lewis acid such as AuCl3 allows achieving up
to 94% yield of acetal under very mild reaction conditions. The gold
catalyst has been made recoverable by preparing a cationic Au(I)
complex that can be separated and reused up to seven times with
>90% yield of acetal.