Gold Sols as Catalysts for Glycerol Oxidation: The Role of Stabilizer



Gold nanoparticles (Au NPs), stabilized by polyvinylalcohol (PVA), tetrakishydroxypropylphosphonium chloride (THPC), and citrate, have been synthesized and tested in the liquid-phase oxidation of glycerol. THPC-stabilized Au NPs have been found to exhibit remarkable catalytic activity (TOF=2478 h−1) in comparison to PVA- and citrate-stabilized nanoparticles (TOF=715 and 160 h−1, respectively). The catalytic results can be explained in terms of particle size but aging treatment also revealed the influence of the protective agent. PVA as stabilizer produced small and quite stable gold nanoparticles differently from the THPC system. However PVA also limits the accessibility of reactant to the active site.


In recent years, the conversion of glycerol into higher-value chemicals has attracted much attention.1, 2 With the decline of petrochemical resources accompanied by a continuing increase in oil prices, research has been intensively focused on biomass.3, 4 Glycerol is an abundant renewable resource obtained in large quantities from biomass as a by-product of the transesterification of triglycerides or in alternative from the fermentation of glucose. Moreover, glycerol is a highly functionalized molecule and a large number of useful compounds can be obtained through its oxidation, hydrogenolysis, dehydration, esterification, and/or polymerization.1, 2 Many studies have recently been devoted to the selective oxidation of glycerol, using gold as the catalyst. In particular, it has been demonstrated that, under the right circumstances, Au can be more active and selective towards glyceric acid than Pd or Pt.5, 6 The gold particle size is crucially important in determining the activity and selectivity of the gold catalyst. Small particles (3–5 nm) are known to be more active than larger particles (10–30 nm), whereas larger particles give higher selectivity towards glyceric acid.711 Although supported Au nanoparticles have received much attention, only a few studies have been dedicated to the catalytic activity of Au sols for liquid-phase reactions. Rossi and co-workers were the first to claim that gold nanoparticles can be used as a successful catalyst for the aerobic oxidation of glucose despite their instability under the operating conditions.13 In fact, the gold clusters showed a very short lifetime prior to aggregation, which limits their application as reusable catalysts. Mertens et al. reported the good performance of PVA-stabilized nanoparticles in the oxidation of diols and also demonstrated that the sols can be efficiently recovered by membrane filtration.13 Furthermore PVP-stabilized Au nanoparticles have been successfully used for the oxidation of benzyl alcohol in water. In particular Tsunoyama et al. focused their attention on the particle size effect.14, 15 Studies on ethylene hydrogenation, Suzuki coupling, and CO oxidation indicated that the nature of the protective agent affects the catalytic activity of Pt, Pd, and Rh NPs.1618 Herein, we report on the aerobic oxidation of glycerol catalyzed by Au NPs with three different stabilizers, polyvinylalcohol (PVA), tetrakishydroxypropylphosphonium chloride (THPC) and citrate, with the aim to study the influence of these protective agents on the activity and stability of the Au NPs in glycerol oxidation. By using PVA and THPC, we were able to obtain small nanoparticles with comparable sizes (ca. 2 nm) but different catalytic behaviors: the THPC-stabilized system resulted in drastically more active catalyst than the PVA-stabilized system but with similar selectivity towards glyceric acid. However, only PVA was able to efficiently stabilize the nanoparticles against coalescence and to maintain the initial particle size, even after reaction, representing a quite stable system.

Results and Discussion

PVA-, THPC-, and citrate-stabilized Au nanoparticles (labeled as AuPVA NPs, AuTHPC NPs and Aucitrate NPs, respectively) were synthesized and tested in the liquid phase oxidation of glycerol. During the study, we paid particular attention to the stability of the different colloidal systems on aging and under the reaction conditions. AuPVA NPs were synthesized by NaBH4 reduction of the AuIII salt in aqueous solution in the presence of PVA, which acts as steric stabilizer.19 AuTHPC NPs were prepared by reducing the aqueous AuIII solution with partially hydrolyzed THPC, whereby THPC also acts as an electrostatic stabilizer.21 Finally, Aucitrate NPs were produced by a slight modification of the procedure by Turkevich et al.,20 with citrate as an electrostatic stabilizer and reducing agent. All the colloidal systems were characterized by high-resolution transmission electron microscopy (HRTEM) when just formed (Figure 1) and after an aging period of one week (Figure 2).

Figure 1.

TEM images of the freshly prepared stabilized gold nanoparticles: a) AuPVA; b) AuTHPC; c) Aucitrate.

Figure 2.

TEM images of the stabilized gold nanoparticles after aging for one week: a) AuPVA; b) AuTHPC; c) Aucitrate.

By using the three different methods for the reduction of AuIII, Au nanoparticles with different diameters were obtained, increasing in the order THPC/NaOH<PVA/NaBH4<citrate. The mean Au particle size was, respectively, 2.02 nm for the THPC/NaOH system, 2.25 nm for the PVA/NaBH4 system, and 9.76 nm for the citrate system (Table 1). Furthermore, AuPVA and Aucitrate nanoparticles both had a narrow size distribution (Figure 1 a and c), whereas for AuTHPC, the particle-size distribution showed a tail of bigger particles (10–13 nm) alongside majority of smaller particles (Figure 1 b and 3 c).

Table 1. Size of Au nanoparticles freshly prepared, after reaction, and after aging for one week.
CatalystAverage diameter [nm]Standard deviation [σ]
  1. [a] Bimodal distribution.

after reaction3.891.97
after reaction11.053.02
aged4.28, 10.15[a]0.32, 0.23[a]
after reaction12.033.35

To study the stability of the three sols, they were kept in the absence of light for one week. AuPVA and Aucitrate showed a good stability against coagulation, whereas for AuTHPC a drastic increase in particles size was observed. For AuPVA NPs in particular, a negligible aging effect was revealed (average sizes of 2.45 nm in the fresh sample and 2.50 nm after aging; Table 1, Figure 3 a and b). Moreover the increase in size of AuTHPC NPs was probably due to particle coalescence (Table 1); in fact, a bimodal distribution appeared (Figure 3 c and d), which indicates that PVA, a steric stabilizer, is a more effective protective agent for small particles than THPC, an electrostatic stabilizer. On the other hand, for larger particles, citrate, also a purely electrostatic stabilizer, can efficiently prevent the aggregation or coalescence of nanoparticles.

Figure 3.

Particle size distribution for the stabilized gold nanoparticles: a) fresh AuPVA; b) aged AuPVA; c) fresh AuTHPC; d) aged AuTHPC; e) fresh Aucitrate; f) aged Aucitrate.

Au NPs were then tested as catalysts for the liquid-phase oxidation of glycerol. The activity was expressed in terms of turnover frequencies (TOF) based on the total metal loading. In good agreement with previous studies,10 the catalytic activity increases with decreasing particle size: AuTHPC (2478 h−1)>AuPVA (715 h−1)>Aucitrate (160 h−1). However the small difference in the particle sizes of AuTHPC and AuPVA (2.02 nm vs. 2.45 nm) cannot be the only reason for such a great difference in the catalytic activity (2478 h−1 vs. 715 h−1). A possible explanation could lie in the different accessibility of the reactant through the protective layer constituted by the different stabilizing agents.

The selectivity follows the expected trend: larger gold nanoparticles showed a higher selectivity toward glycerate than smaller ones. Aucitrate, in fact showed a selectivity, at 90 % conversion, to glyceric acid of 70 %, whereas AuTHPC and AuPVA showed selectivities of 62 % and 64 %, respectively.7 For the latter two systems, there was a partial overoxidation of glyceric acid to tartronic acid. The selectivity to glyceric acid of unsupported Au NPs is slightly lower than the value reported for supported Au nanoparticles, in agreement with our recent results on a direct comparison between unsupported AuPVA NPs and activated carbon-supported AuPVA NPs.22 The difference in behavior can be correlated to the higher stability of the produced H2O2 in the presence of unsupported Au nanoparticles in comparison to carbon-supported ones. H2O2, formed during the oxidation process, promotes the C[BOND]C bond cleavage, thus decreasing the selectivity to glyceric acid

We also tested the sols, under similar conditions, after an aging treatment of one week. The aged sols all showed only a slightly lower activity than the fresh samples; the activity decreased, from 715 h−1 to 640 h−1 for AuPVA, freshly prepared and after aging, respectively, and from 160 h−1 to 145 h−1 for Aucitrate, freshly prepared and after aging, respectively (Table 2). The selectivity to glyceric acid was also almost maintained after the aging treatment.

Table 2. Activity and selectivity of Au sols in the glycerol oxidation.
CatalystSelectivity at 90 % conversion [%]TOF [h−1][a]
 Glyceric acidGlycolic acidTartronic acidOxalic acidFormic acid 
  1. [a] TOF calculated after 15 min based on total metal loading. [b] Selectivity calculated at 15 % conversion.

Freshly prepared      
After aging treatment      

These catalytic data are in agreement with the slight change in particle dimensions revealed by TEM between the freshly prepared and aged samples of the AuPVA and the Aucitrate sols (Table 1, Figure 2 and 3). However, although the AuTHPC sol does not show a large decrease in activity after aging (from 2478 h−1 to 2158 h−1), a considerable increase in the particle size was observed (Figure 3 and Table 1). The selectivity towards glyceric acid increased from 63 % to 72 %, as expected for bigger particles. A possible explanation of this counterintuitive behavior is the bimodal distribution of particle sizes, which was revealed by TEM of the aged AuTHPC NPs. The presence of small-sized particles (ca. 4 nm) would contribute to maintaining a high activity, comparable to the freshly prepared NPs, whereas the presence of the bigger particles (ca. 10 nm) would contribute to the enhancement in selectivity towards glyceric acid, from 63 % for the freshly prepared sol to 72 % after the aging treatment.

However, the fact that the AuTHPC sol after aging remains considerably more active (bimodal distribution, ca. 4 nm and ca. 10 nm; TOF=2158 h−1) than both the freshly prepared (2.45 nm; TOF=715 h−1) or aged AuPVA (2.50 nm; TOF=640 h−1) systems, both of which are smaller in size, leads us to conclude that the activity is not merely related to the particle size.

We could then suppose that the protective agent plays a fundamental role not only in determining the stability of the colloidal systems but also in tuning the catalytic activity. The more efficient stabilization of small gold nanoparticles by the steric effect of PVA in comparison to the electrostatic effect of THPC was also revealed by TEM observation of the colloidal systems after reaction. For AuTHPC NPs, there was a drastic a change in the particle size, from 2 to 11 nm, whereas for AuPVA NPs the growth was less marked, from 2.45 to 3.88 nm (Table 1). Conversely, bulky molecules, such as PVA, surrounding the Au NPs seem to limit their activity, with AuPVA showing a drastically lower TOF than AuTHPC (715 h−1 vs. 2478 h−1).

Thus we can conclude that the difference in activities between AuPVA and AuTHPC systems can be accounted for by the different types of stabilization offered by the two protective agents. Polyvinyl alcohol stabilizes the gold nanoparticles by a steric effect, coordinating to gold by van der Waals interactions. The polymer is soluble in water only at temperatures higher than 60 °C. Thus, we assume that, at the reaction temperature (50 °C), the polyvinyl alcohol is still able to stabilize the gold nanoparticles. The adsorption of the reactant on the active site is possible due to the high porosity of the polymer. The combination of these two aspects is explains why PVA-stabilized Au NPs form a stable and active catalyst. Conversely, THPC is an electrostatic stabilizer, so the negatively charged Au NPs coordinate with the positive part of the THPC molecule. During the formation of the metallic sol and the storage time, a huge excess of NaOH is present. THPC/NaOH, acts as the reducing agent by the formation of formaldehyde, a well-known reducing agent for gold under basic conditions,23 following the mechanism:24


By storing the solution in the presence of NaOH, THPC could progressively decompose, thus reducing its ability to stabilize the gold nanoparticles. This phenomenon is more evident during the oxidation of glycerol in the presence of NaOH, during which a drastic increase in gold particle size took place (Table 1).


Gold nanoparticles were synthesized in the presence of three different stabilizers, namely polyvinyl alcohol (PVA), tetrakishydroxypropylphosphonium chloride (THPC) and citrate. The three resultant systems were tested as catalysts for the liquid-phase oxidation of glycerol. AuTHPC showed a high activity (TOF=2478 h−1) in comparison to the other two systems (TOF=715 and 160 h−1 for AuPVA and Aucitrate, respectively). The activities and selectivities could be explained in terms of particle size. However, the catalytic results using sols that had been aged for a week revealed that the protective agent played a fundamental role not only in stabilizing the Au NPs but also in determining their activity. Thus the scale of activity was not merely a matter of the particle size but the result of a compromise between the particle size and the protective layer. TEM observations showed that PVA and citrate were able to maintain the initial particle size, whereas for the THPC-stabilized system, the particles grew due probably to coalescence of the small particles into bigger ones. Gold nanoparticles stabilized by steric effects formed stable unsupported small, representing a suitable compromise between stability and catalytic activity. AuPVA NPs were more stable yet less active than AuTHPC NPs, due to limited accessibility of the reactants through the pores of the polymer.

Experimental Section

Catalyst preparation

PVA-stabilized gold sol: Solid NaAuCl4⋅2 H2O (17 mg, 0.043 mmol) and aqueous 2 wt % PVA solution (410 μL; 1:1 PVA/Au w/w) were dissolved in water (130 mL). After 3 min, 0.1 M aqueous NaBH4 solution (1.3 mL) was added to the yellow solution under vigorous stirring. A ruby-red Au0 sol was immediately formed.17 A UV/Vis spectrum of the gold sol was recorded, showing complete reduction of the AuIII species.

Citrate-stabilized gold sol: Sol generated in the presence of citrate was prepared by modifying the procedure reported elsewhere.20 An aqueous 0.296 M sodium citrate solution (1.63 mL; 50:1 citrate/Au w/w) was added to an aqueous 2.36×10−4M HAuCl4 solution (300 mL) at 65 °C. After a reaction time of 1 h, the yellow solution turned violet in color, changing to ruby red after 3 h of reaction. A UV/Vis spectrum of the gold sol was recorded, showing complete reduction of the AuIII species.

THPC-stabilized gold sol: Sol generated in the presence of the THPC/NaOH system was prepared as reported elsewhere.21 A freshly prepared 0.05 M aqueous solution of THPC was added (THPC/Au (w/w)=0.95) to a 10−3M NaOH solution (2 mL). After five minutes, 10−3M aqueous HAuCl4 (1.0 mL) was added dropwise, forming a brown metallic sol. A UV/Vis spectrum of the gold sol was recorded, showing complete reduction of the AuIII species.

The resultant sols were analyzed and tested immediately after synthesis and again after aging for one week in the absence of light.

Catalytic tests

Reactions were carried out in a 30 mL glass reactor equipped with a thermostat and an electronically controlled magnetic stirrer connected to a 5000 mL reservoir charged with oxygen (300 kPa). The oxygen uptake was followed by a mass-flow controller connected to a PC through an A/D board, plotting a flow time diagram. A solution of glycerol (0.316 g) in water (4 mL) was mixed with the desired amount of NaOH and the gold colloidal solution (final concentration of glycerol=0.3 M, glycerol/metal molar ratio=1000:1, NaOH/glycerol molar ratio=4:1). The reactor was pressurized at 300 kPa of oxygen and set to 50 °C. Once this temperature was reached, the gas supply was switched to oxygen and the monitoring of the reaction started. The reaction was initiated by stirring. Samples were removed periodically and analyzed by high-performance liquid chromatography (HPLC) using a column (Alltech OA-10308, 300 mm×7.8 mm) with UV and refractive index (RI) detection to analyze the mixture of the samples. Aqueous H3PO4 solution (0.1 wt %) was used as the eluent. Products were identified by comparison with the original samples.


The morphologies and microstructures of the catalysts were characterized by TEM. A few drops of the sols were mounted onto copper grids covered with holey carbon. A Philips CM200 FEG electron microscope, operating at 200 kV and equipped with a Gatan imaging filter, GIF Tridiem, was used for TEM observation. EDX analysis was performed in the same microscope using a DX4 analyzer system (EDAX).


Vigoni Program, AIT-DADD and Fondazione Cariplo are gratefully acknowledged for financial support.