Design of a highly active silver-exchanged phosphotungstic acid catalyst for glycerol esterification with acetic acid

Design of a highly active silver-exchanged phosphotungstic acid catalyst for glycerol esterification with acetic acid

Journal of Catalysis 306 (2013) 155–163 Contents lists available at SciVerse ScienceDirect Journal of Catalysis journal homepage: www.elsevier.com/l...

820KB Sizes 1 Downloads 274 Views

Journal of Catalysis 306 (2013) 155–163

Contents lists available at SciVerse ScienceDirect

Journal of Catalysis journal homepage: www.elsevier.com/locate/jcat

Design of a highly active silver-exchanged phosphotungstic acid catalyst for glycerol esterification with acetic acid Shanhui Zhu a,b, Xiaoqing Gao c, Fang Dong a,b, Yulei Zhu a,c,⇑, Hongyan Zheng c, Yongwang Li a,c a

State Key Laboratory of Coal Conversion, Institute of Coal Chemistry, Chinese Academy of Sciences, Taiyuan 030001, PR China Graduate University of Chinese Academy of Sciences, Beijing 100039, PR China c Synfuels China Co. Ltd., Taiyuan 030032, PR China b

a r t i c l e

i n f o

Article history: Received 26 April 2013 Revised 24 June 2013 Accepted 25 June 2013 Available online 27 July 2013 Keywords: Glycerol Esterification Biofuels Ag Phosphotungstic acid

a b s t r a c t A series of highly active, selective, and stable silver-exchanged phosphotungstic acid (AgPW) catalysts were prepared, characterized, and evaluated for bio-derived glycerol esterification with acetic acid to produce valuable biofuel additives. The structures, morphologies, acidities, and water tolerance of these samples were determined by FTIR, Raman, XRD, SEM-EDX, FT-IR of pyridine adsorption, and H2O-TPD. Several typical acidic catalysts were also performed for comparison. Among them, partially silverexchanged phosphotungstic acid (Ag1PW) presented exceptionally high activity, with 96.8% conversion within just 15 min of reaction time and remarkable stability, due to the unique Keggin structure, high acidity as well as outstanding water-tolerance property. A plausible reaction mechanism was also proposed. In addition, this Ag1PW catalyst exhibited universal significance for esterification, holding great potential for a wide range of other acid-catalyzed reactions. Ó 2013 Elsevier Inc. All rights reserved.

1. Introduction As a consequence of the diminishing fossil resources, significant efforts have been taken worldwide to convert renewable biomass into fuels and value-added compounds [1,2]. In this context, biodiesel derived from the lipid fraction of biomass is currently attracting considerable attention for its sustainability, reduced toxicity and carbon neutral alternative of fossil fuels [2,3]. Biodiesel is manufactured by the transesterification of triglycerides with methanol or ethanol, which concurrently yields large amounts of glycerol equivalent to about 10 wt% of the total biodiesel production [3]. With the steady growth of biodiesel industry, tremendous surplus of glycerol has been produced, and evidently, a glycerol ‘‘lake’’ is being formed, which makes its commercial value depreciate sharply. Due to its nontoxic, edible, biodegradable properties as well as multifunctional structure, glycerol holds the potential of being an extremely important building block for the biorefinery. Consequently, a variety of catalytic processes have been envisaged for the valorization of glycerol by hydrogenolysis, reforming, etherification, esterification, oxidation, dehydration, and so on [4–9]. As such, one of the most promising potential approaches is catalytic esterification of glycerol with acetic acid to the formation of ⇑ Corresponding author at: State Key Laboratory of Coal Conversion, Institute of Coal Chemistry, Chinese Academy of Sciences, Taiyuan 030001, PR China. Fax: +86 351 7560668. E-mail address: [email protected] (Y. Zhu). 0021-9517/$ - see front matter Ó 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jcat.2013.06.026

monoacylglycerol (MAG), diacylglycerol (DAG), and triacylglycerol (TAG) because of their versatile uses as biofuel additives directly or as precursors in the synthesis of polyesters [10]. The esterification of alcohols with organic acids is a conversion of both environmental and industrial importance, being employed extensively in the synthesis of various esters from bio-derived feedstocks. Among them, the selective esterification of glycerol with acetic acid is probably one of the most studied processes in the literature of esterification, and so can be taken as a model reaction in this type of catalysis [11]. Conventionally, the esterification of glycerol with acetic acid is performed in the presence of mineral acids catalysts. However, the inherent disadvantages related with homogeneous system lie in separation and purification of the products, resulting in the environmental problems and economical inconveniences. In order to tackle these problems, a great number of heterogeneous solid acid catalysts have been developed in recent works, such as hydroxylated magnesium fluorides [12], sulfated activated carbon [13], SO3H-functionalized ionic liquids [14], mesoporous silica with sulfonic acid groups [15], zeolites [16], Amberlyst-15 [17], and heteropolyacids [18–20]. Despite the relative efficiency of these catalysts in glycerol esterification, many of them have low densities of effective acid sites, tedious preparation protocols, and rapid loss of catalytic activity. Furthermore, the unavoidable generation of water as a byproduct of esterification can have a serious impact on the catalytic performance of some solid acid catalysts such as supported heteropolyacids (HPAs)

156

S. Zhu et al. / Journal of Catalysis 306 (2013) 155–163

because of their extremely hydrophilic nature. Consequently, how to design a highly active, inexpensive, robust, and water-tolerant heterogeneous solid acid catalyst that can be employed by simple preparation protocol is still a great challenge. HPAs, possessing characteristics of strong Brønsted acidity and easily tunable acidity, have been demonstrated to display outstanding catalytic performance in a wide range of acid-catalyzed reactions [21–24]. However, the lack of thermal stability and high solubility in polar media of HPAs has limited their further applications in heterogeneous acid-catalyzed reactions. In contrast to grafting HPAs onto porous supports, it is more effective to exchange protons of HPAs with different cations (e.g., K+, Cs+, Ag+) to form insoluble salts, which can tune and amplify HPAs reactivity or even result in the appearance of bifunctional or multifunctional catalysis [25]. Such a conceptual strategy would lead to offer a new class of tunable and recoverable HPA salts catalysts with high efficiency and heterogeneity, providing versatile applications in sustainable chemistry. Recently, Borghèse et al. [26] have developed a series of exceedingly effective and reusable silver-exchanged silicotungstic acid catalysts for the rearrangement of alkynyloxiranes to furans. Compared to H4SiW12O40 and other HPAs such as H3PMo12O40, H3PW12O40 (HPW) presents stronger Brønsted acidity and thermal stability, because of the weak interaction between acidic protons and large Keggin anion. Thereby, in this work, we have focused on designing silver modified HPW catalysts prepared by an ion-exchanged method, exhibiting unprecedented catalytic activity and superior stability for glycerol esterification. To the best of our knowledge, this is the first report on the catalytic performance of heterogeneous silver-exchanged HPA catalysts for the esterification. Accordingly, in the present investigation, Ag-exchanged HPW catalysts with varying Ag contents were prepared, characterized, and evaluated for glycerol esterification with acetic acid. The catalyst features were characterized using various spectroscopy techniques and correlated with the observed catalytic performance of glycerol esterification. The esterification of different alcohols with organic acids was performed to check the scope of this catalyst. 2. Experimental 2.1. Catalyst preparation All the chemicals were obtained commercially and used without any further purification. H3PW12O40xH2O was purchased from Sinopharm Chemical Reagent Co., Ltd. (SCRC). AgNO3 was supplied by Tianjin Damao Chemical Co., Ltd. Prior to the preparation, the water content of H3PW12O40xH2O was checked by TG-MS. The Ag-exchanged HPW catalysts were synthesized by an ion-exchanged method, according to the procedure described previously [26,27]. Firstly, 10.25 g HPW was dissolved in 20 ml deionized water at room temperature under vigorous stirring. Then, the appropriate amount of AgNO3 (0.1 mol/L) aqueous solution was added dropwise to the former solution with continuous stirring. The resultant mixture was aged 2 h at room temperature, and the excess water was evaporated in a rotary evaporator. The remaining powder was dried at 80 °C overnight and then calcined at 250 °C in static air for 4 h. Analogously, the number of Ag atom, i.e., x in AgxH3xPW12O40 (x = 1, 2, 3), can be conveniently controlled by varying the amount of AgNO3 aqueous solution. These as-prepared catalysts are designated as Ag1PW, Ag2PW, and Ag3PW, wherein the number implies the number of Ag ions exchanged. The formation of AgxH3xPW12O40 reaction undergoes based on the following equations:

xAgNO3 þ H3 PW12 O40 ! xHNO3 þ Agx H3x PW12 O40

ðx ¼ 1; 2; 3Þ

Wet ion-exchanged resins Amberlyst-15 (30 nm of average pore diameter, 50 m2 g1) and Amberlyst-30 (30 nm of average pore diameter, 53 m2 g1) were dried overnight at 110 °C prior to the catalytic tests. ZSM-5 (Si/Al = 25, 0.56 nm of average pore diameter, 50 m2 g1) was activated at 500 °C in static air for 4 h before the test. ZrO2 (59.7 m2 g1) supplied from Jiangsu Qianye Co., Ltd was used as the support. MoO3/ZrO2, WO3/ZrO2, Nb2O5/ZrO2, and SO2 4 =ZrO2 were prepared by incipient wetness impregnation method by using (NH4)6Mo7O244H2O (SCRC), (NH4)6W7O246H2O (SCRC), Nb(OH)5 (King-Tan Tantalum Industry Ltd.), H2SO4 (SCRC) as precursors. Specifically, these catalysts were prepared by impregnation of ZrO2 with the calculated amount of aqueous solution of desired precursors and then dried overnight at 110 °C followed by calcination at 600 °C in static air for 4 h. Appropriate amount of oxalic acid dehydrate (SCRC) was added to the solution of Nb(OH)5 to improve the solubility during the preparation of Nb2O5/ZrO2. The nominal loadings of acid components (MoO3, WO3, Nb2O5, and SO2 4 ) in the corresponding catalysts were 15%. Cs-exchanged HPW catalyst (Cs2.5PW) was prepared with the same procedure as the aforementioned Ag1PW, wherein 2.5 protons in one HPW molecule can be replaced by Cs atoms. 2.2. Catalyst characterization Powder X-ray diffraction (XRD) patterns of the catalysts were recorded on D2/max-RA X-ray diffractometer (Bruker, Germany) using Cu Ka radiation at 30 kV and 10 mA. The measurements were obtained in the step of 0.04° with account time of 0.5 s and in the 2h rang of 5–90°. Raman spectroscopy was obtained on a Renishaw–UV–vis Raman System 1000 equipped with a CCD detector at room temperature. The air-cooled frequency doubled Nd–Yag laser operating at 532 nm was employed as the exciting source with a power of 30 MW. Scanning Electron Microscopy (SEM) was conducted on a Quanta 400F microscope. EDX spectra were obtained using 20 kV primary electron voltages to determine the composition of the samples. The IR spectra were measured on a Vertex 70 (Bruker) FT-IR spectrophotometer, equipped with a deuterium triglycine sulfate (DTGS) detector. The powder samples were mixed with KBr (2 wt%) and pressed into translucent disks at room temperature. The spectra were recorded in the range of 400–4000 cm1. IR spectra of adsorbed pyridine (Py-IR) were recorded with the same apparatus as above. The samples were pressed into self-supporting wafers, degassed in a vacuum at 300 °C for 1 h, and subsequently exposed to the pyridine vapor after cooling down to 30 °C. The Py-IR spectra were then recorded at 200 °C after applying vacuum for 30 min. The quantification of acidity was calculated by Lambert–Beer equation,



eW c S

where A is the absorbance (area in cm1), e the extinction coefficient (m2/mol), W the sample weight (kg), c the concentration of acid (mol/kg or mmol/g) and S is the sample disk area (m2). The amount of Brønsted and Lewis acid sites was estimated from the integrated area of the adsorption bands at ca. 1540 and 1450 cm1, respectively, using the extinction coefficient values based on the previous report [28]. H2O-TPD was performed in an Auto Chem.II 2920 equipment (Mircromeritics, USA). Prior to each run, 0.3 g catalyst was first pretreated in flowing He at 250 °C for 1 h and then cooled to 50 °C followed by saturating with water using pulse model until saturation. After being purged with He for 30 min, the catalyst

157

S. Zhu et al. / Journal of Catalysis 306 (2013) 155–163

bed was heated to 500 °C at a rate of 10 °C/min, and the desorbed water was monitored by a TCD detector. ICP optical emission spectroscopy (Optima2100DV, PerkinElmer) was employed to analyze the remaining solution after glycerol esterification with acetic acid. 2.3. Catalytic tests All the catalytic tests were conducted at atmospheric pressure. Esterification of glycerol with acetic acid was carried out in a 100 ml stainless steel with Teflon liner autoclave under vigorous stirring. In a typical run, 0.05 g catalyst, 5.0 g glycerol, and 32.6 g acetic acid were introduced into the reactor. Then, the reaction mixture was stirred and heated at 120 °C for an appropriate reaction time. After reaction, the reactor system was rapidly cooled to room temperature in an ice-water bath. Subsequently, the solid catalyst was separated by centrifugation, and the liquid products were analyzed by a gas chromatography (Ruihong chromatogram analysis Co., Ltd, China) with a flame ionization detector using a DB-WAX capillary column. Concurrently, the assignment of the products was identified by a gas chromatography–mass spectrometry (Agilent, USA) with a DB-WAX capillary column. The conversion of glycerol and selectivity of products were calculated based on the following equations:

Conversion ð%Þ ¼

moles of glycerol ðinÞ  moles of glycerol ðoutÞ moles of glycerol ðinÞ  100

Selectivity ð%Þ ¼

moles of one product  100 moles of all products

3. Results and discussion 3.1. Catalyst characterization To thoroughly recognize the structure transformation induced by silver doping and morphologies of as-prepared Ag-exchanged HPW catalysts, a combination of FTIR, Raman, XRD, and SEM-EDX was employed. Infrared spectra are known to be an informative fingerprint of Keggin structure at molecular level [29]. The Keggin anion of HPW is composed of a tetrahedral PO4 surrounded by 12 octahedral WO6, sharing edges in W3O13 triads and corners with

other triads through bridging oxygens [29]. FTIR spectra of parent HPW and Ag-salts catalysts are illustrated in Fig. 1. In the case of parent HPW, the characteristic bands of typical Keggin anions can be clearly observed as follows: 1080 cm1 (mas PAO), 982 cm1 (mas W@O), 890 cm1 (mas WAOAW inter-octahedral), and 797 cm1 (mas WAOAW intra-octahedral) [30]. All silver-exchanged HPW catalysts exhibited these typical bands at similar frequencies, close to those of parent HPW, indicating the whole retention of Keggin structure on the as-prepared Ag-salts catalysts. Raman spectroscopy is uniquely suited for check Keggin structure of HPAs to verify the presence and integrity of HPAs structure. Fig. 2 shows Raman spectra of parent HPW and its Ag-salts catalysts. It has been identified and attributed Raman bands at 1010 cm1 and a shoulder at 990 cm1 to the symmetric and asymmetric stretching modes of W@O [30]. The bands at 900 and 550 cm1 are ascribed to the bending modes of WAOAW and OAPAO, respectively [30]. These bands were all present and clearly identified in the as-prepared silver-exchanged HPW catalysts, suggesting that the Keggin structure was intact even after exchange of acidic protons with silver. Fig. 3 displays the XRD patterns of parent HPW and Ag-salts catalysts. The parent HPW exhibited all of typical X-ray diffractograms of body-centered cubic secondary structure of Keggin anion, with characteristic diffraction peaks at 10.3°, 25.3°, and 34.6° [30]. Although the as-prepared Ag-salts samples exhibited very similar diffraction patterns as that of HPW, a slight shift toward higher 2h values was observed, and no diffraction peaks of pure HPW were detected, suggesting the presence of only one Ag-salt phase with good crystallinity. A detailed study of the predominant 25.0–27.0° reflections confirmed this effect, showing the continuous shift toward lower lattice parameters with the increase in silver content, as reported previously [25,31]. It is speculated that silver-exchanged HPW catalysts possess the same symmetry as parent HPW but with a contracted unit cell. This contraction of unit cell can be explained by the exchange of protons present in the secondary Keggin structure in the form of dihydronium ions H2 Oþ 2 for hydrated silver cation. The same behavior has been observed by Borgh‘ese et al. in the case of silver-exchanged silicotungstic acids catalysts [26,27]. Accordingly, our XRD patterns revealed that only one phase of Ag-salt with uniform crystalline was formed, whatever the content of Ag+ incorporated. It should be noted that the observed fact from XRD is in support of the results from FTIR and Raman, which has demonstrated the successful incorporation of Ag+ into HPW clusters and the formation of good crystalline with the remaining of Keggin structure.

Ag3PW

Ag3PW

Ag2PW Intensity

Absorbance

Ag2PW

Ag1PW

Ag1PW

HPW

HPW

400

600

800

1000

1200 -1

Wavenumber (cm ) Fig. 1. FTIR spectra of bulk HPW and Ag-salts catalysts.

1400

200

400

600

800

1000 -1

Raman shift (cm ) Fig. 2. Raman spectra of bulk HPW and Ag-salts catalysts.

1200

158

S. Zhu et al. / Journal of Catalysis 306 (2013) 155–163 Table 1 Energy dispersive X-ray analysis data for Ag and W on Ag-salts of HPW.

Ag 3 PW

Intensity

Ag 2 PW

Catalyst

W/Ag nominal ratio

W/Ag determined by EDX

W/Ag average value from EDX

Ag1PW

12

11.66

Ag2PW Ag3PW

6 4

14.8; 11.8; 12.2; 10.2; 9.3 7.3; 5.9; 4.4; 6.5; 5.8 4.6; 3.1; 4.3; 3.0; 3.8

5.98 3.76

Ag 1 PW

B

L

Ag3PW

20

40

60

80

26 o

o

2 ()

2 ()

Absorbance

HPW

Ag2PW Ag1PW

Fig. 3. XRD patterns of bulk HPW and Ag-salts catalysts.

To gain further insights, SEM was conducted to directly image the as-prepared catalysts. A representative selection of SEM images of Ag1PW and Ag3PW samples is presented in Fig. 4. It can be found that these as-synthesized materials displayed well-shaped crystalline particles, particularly for the fully substituted Ag3PW. With the increase in Ag content, the size of crystalline particles evidently grew, resulting in the formation of bigger sizes of Ag3PW than those of Ag2PW and Ag1PW. Furthermore, EDX was examined to determine the elemental compositions of these samples. The values of W/Ag atomic ratios calculated by EDX analysis are listed in Table 1. For all the Ag-exchanged HPW samples, the W/Ag ratios determined by EDX analysis were well consistent with the nominal value. In sum, the above observation from SEM-EDX acted in accordance with the XRD results, revealing the presence of only one crystalline phase for all the Ag-exchanged catalysts and not a mixture of different stoichiometric Ag-salts. For the acid-catalyzed esterification of glycerol, both the surface acid density and the nature of acid sites play the key roles in determining the catalytic performance. Therefore, FT-IR spectra of pyridine absorption were recorded to probe accessible surface acid sites, which is a powerful tool for identifying the nature of acid sites. Additionally, in situ pretreatments can be available to the solid acid catalysts. As shown in Fig. 5, all the samples presented typical bands corresponding to strong Brønsted bound pyridine, at around 1540 and 1639 cm1 [32]. The bands at around 1450 and 1610 cm1 were assigned to the coordinated pyridine adsorbed on Lewis acid sites, while the band at 1489 cm1 was originated

HPW

1400

1450

1500

1550

1600

1650

1700

-1

Wavenumber (cm ) Fig. 5. FTIR spectra of pyridine adsorption of bulk HPW and Ag-salts catalysts.

from the combination of pyridine on both Brønsted and Lewis acid sites [33]. The concentrations of Brønsted acid and Lewis acid sites are obtained from the bands at 1540 and 1450 cm1 based on Lambert–Beer equation, and the calculated results are listed in Table 2. In comparison with pure Brønsted acid of parent HPW, Lewis acid sites emerged as a result of the exchange of Ag with protons of HPW, originating from the coordinately unsaturated Ag species in the catalyst. Among them, the Ag1PW catalyst presented the maximum acidity, even more than the parent HPW, owing to the enhanced Lewis acid sites. In comparison with the fully Ag-exchanged HPW sample, partially Ag substituted HPW catalyst increased available acid sites, due to the presence of residual protons capable of mobility and inducing new strong acid sites [29]. Similar behavior has been reported in the cases of Cs- and Sn-exchanged HPW catalysts [29,34]. It was interesting to find that the fully Ag-exchanged HPW sample presented 1.10 mmol/g Brønsted acids, which may originate from the adsorbed surface OH groups or even remaining trace protons.

Fig. 4. SEM micrographs of Ag1PW (a) and Ag3PW (b) catalysts.

159

S. Zhu et al. / Journal of Catalysis 306 (2013) 155–163 Table 2 Amount of acid sites on bulk HPW and Ag-salts catalysts determined by FT-IR spectra of pyridine absorption. Brønsted acidity (mmol/g)

Lewis acidity (mmol/g)

Total acidity (mmol/g)

HPW Ag1PW Ag2PW Ag3PW

1.61 1.38 1.10 1.02

0.04 0.54 0.53 0.41

1.65 1.92 1.63 1.43

To explore the water-tolerant property of Ag-salts, the H2O-TPD test was performed. As shown in Fig. 6, the Ag-salts catalysts showed good water tolerance as indicated by its only one small low-temperature desorption peak, particularly for Ag3PW wherein the protons were entirely substituted by Ag+. Contrarily, the bulk HPW presented two broad peaks at about 270 °C and 455 °C, respectively, suggesting that water was strongly absorbed on this sample. The hydrophobic property decreased in the following order of Ag3PW > Ag2PW > Ag1PW > HPW, consistent well with the increasing proton numbers. 3.2. Catalytic reaction of Ag-salts catalysts Fig. 7 displayed the profiles of glycerol conversion as a function of reaction time for glycerol esterification with acetic acid. The parent HPW catalyst, which is well known for its strong Brønsted acidity for versatile acid-catalyzed reactions, was employed to make a comparison with Ag-salts catalysts. The glycerol conversion over homogenous HPW catalyst rapidly reached 70.3% even after 15 min and then increased slowly with extended reaction time. Compared to HPW, all the Ag-exchanged catalysts presented superior performance, despite the nature of heterogeneity of these samples. As shown in Fig 7, at all reaction time, the activity of the catalysts decreased in the following order of Ag1PW > Ag2PW > Ag3PW > HPW. The conversion of glycerol was up to 100% within only 45 min on Ag1PW, while the glycerol did not reach complete conversion on Ag2PW and Ag3PW within 45 min. The difference in catalytic activity was mainly attributed to the variant concentrations of Brønsted and Lewis acid sites as well as water tolerance. It should be mentioned that although the initially high rate on Ag2PW and Ag3PW declined and further reaction progress at lower rate, complete conversion of glycerol can be still achieved with prolonged reaction time. The above results indicated

Ag3PW

Intensity

Ag 2 PW

Ag1PW

HPW

100

200

300

400 o

Temperature ( C) Fig. 6. H2O-TPD profiles of bulk HPW and Ag-salts catalysts.

80

Conversion (%)

Catalyst

100

60

HPW Ag 1 PW

40

Ag 2 PW Ag 3 PW

20

0 0

10

20

30

40

50

Time (min) Fig. 7. Glycerol conversion as a function of reaction time over HPW and Ag-salts catalysts. Reaction conditions: 120 °C, glycerol/acetic acid = 1:10 (molar ratio), 1 wt% catalyst loading relative to glycerol.

that at the same reaction conditions, the Ag2PW and Ag3PW catalysts needed longer time to attain equilibrium compared to Ag1PW. This may be related to the concentrations of Brønsted and Lewis acid sites, which were lower on Ag2PW and Ag3PW. Although Ag2PW and Ag3PW possessed better water tolerance, Ag1PW exhibited superior catalytic performance. The role of acid sites seems to be more significant for glycerol esterification on these Ag-salts catalysts. In all the cases, the conversion of glycerol improved rapidly within the initial 10 min of reaction time, suggesting the rapid consume of glycerol in the initial reaction stage. Moreover, the reaction order on glycerol concentration in this stage was probably zero. Nevertheless, the reaction rate declined significantly in the latter reaction stage. The above results revealed that the glycerol concentration (acetic acid was excess) has significant impact on the kinetic behavior of glycerol esterification. Additionally, large amount of by-product water was formed during the reaction, which can inhibit the conversion of glycerol and thus decrease the reaction rate, particularly in the latter stage. Therefore, it can be inferred that the initial high reaction rate was switched off, and then, the reaction rate declined remarkably, which was the combined effect of decreased glycerol concentration and the accumulation of by-product water. Table 3 listed the representative product distributions over HPW and Ag-salts catalysts. The obtained products of glycerol esterification were MAG, DAG, and TAG, while no dehydration products of glycerol were detected in the product mixture, probably because of the relatively mild reaction conditions [24]. Among them, DAG and TAG are the most preferred products owing to their versatile applications as fuel additives, which can be utilized to improve the cold and viscosity properties of biodiesel as well as antiknocking properties of gasoline [10]. It is identified that glycerol esterification with acetic acid is a consecutive reaction including three continuous steps (glycerol + acetic acid ? MAG + water; MAG + acetic acid ? DAG + water; DAG + acetic acid ? TAG + water) [35]. Thereby, to maximize production of the valuable DAG and TAG, it is highly favorable to perform this reaction with extended reaction time. Since the Ag1PW catalyst afforded the best reactivity, the following study would focus on the investigation of catalytic behavior of Ag1PW. As displayed in Fig. 8, Ag1PW catalyst presented exceptional activity even just 15 min and achieved complete conversion within 45 min. The selectivity of MAG is high at the beginning of the reaction. Nevertheless, as the reaction proceeded, the selectivity of DAG and TAG enhanced at the expense

160

S. Zhu et al. / Journal of Catalysis 306 (2013) 155–163

Table 3 Catalytic performance of glycerol esterification with acetic acid over various catalysts.a. Entry

Catalyst

1 2 3 4 5 6 7 8 9 10 11 12

Conversion (%)

HPW Ag1PW Ag2PW Ag3PW Blank ZSM-5 Amberlyst-15 Amberlyst-30 MoO3/ZrO2 WO3/ZrO2 Nb2O5/ZrO2 SO2 4 =ZrO2 Cs2.5PW

13

Selectivity (%) MAG

DAG

TAG

70.3 96.8 82.5 75.7 21.3 22.9 33.8 39.8 25.6 23.1 24.7 25.5

59.3 48.4 52.7 59.7 83.5 83.8 65.8 60.3 70.7 63.6 83.1 83.3

37.7 46.4 43.4 37.1 16.3 14.8 32.0 37.7 28.3 35.5 15.5 15.3

3.0 5.2 3.9 3.2 0.2 1.4 2.2 2.0 1.0 0.9 1.4 1.4

32.4

72.2

27.7

0.1

a

Reaction conditions: 120 °C, glycerol/acetic acid = 1:10 (molar ratio), 1 wt% catalyst loading relative to glycerol, 15 min.

Conversion and selectivity (%)

100

Glycerol Monoacetin Diacetin Triacetin

80

60

40

20

0 0

50

100

150

200

250

the previous report [35], glycerol esterification with acetic acid is equilibrium limited and a consecutive reaction including three continuous steps, which can determine the final product distribution. An effective technique to address the thermodynamic limitation is to remove the generated water from the reaction system by reactive distillation. 3.3. Comparision with other typical catalysts A number of typical solid acid catalysts have been selected and evaluated for glycerol esterification in comparison with the best Ag1PW catalyst at 120 °C within 15 min of reaction time. As shown in Table 3, glycerol conversion was up to 21.3% without addition of any catalyst, suggesting that glycerol esterification with acetic acid was a self-catalysis reaction wherein the acidic protons from acetic acid were capable of catalyzing the reaction itself [35]. ZSM-5 did not present any promotion of this reaction with such low amount and short reaction time, which was likely ascribed to the diffusion limitations caused by the narrow pore structure. Contrarily, Amberlyst-15 and Amberlyst-30 acid resins promoted this reaction mildly in terms of glycerol conversion and the selectivity of DAG and TAG. Our previous report [17] revealed that these acid resins presented better performance of glycerol esterification than ZSM5 due to their large pore diameter. With respect to commonly exploited mixed metal oxides such as MoO3/ZrO2, WO3/ZrO2, and Nb2O5/ZrO2, their reactivity was remained rather low, which might be related to their low acidity. Furthermore, other heterogeneous acid catalysts SO2 4 =ZrO2 and Cs2.5PW also gave low activity in comparison with the present Ag1PW catalyst. It should be noted that our as-prepared Ag-salts catalysts presented an extremely high activity for glycerol esterification, despite with very low catalyst amount (1 wt% relative to glycerol) as well as mild reaction conditions. As far as we know, the Ag1PW catalyst is the most active catalyst reported in the literature (Table S1). Actually, compared with the reported catalysts, the Ag1PW catalyst showed exceptionally high specific rate and turnover number (TON), or even one order of magnitude more active (Table S1). The TON of Ag1PW was up to 2189.9 h1.

Time (min) 3.4. Reusability of Ag1PW catalyst Fig. 8. Effect of reaction time on glycerol esterification with acetic acid over Ag1PW catalyst. Reaction conditions: 120 °C, glycerol/acetic acid = 1:10 (molar ratio), 1 wt% catalyst loading relative to glycerol.

of MAG. The combined selectivity of DAG (58.2%) and TAG (31.9%) reached the highest value of 90.1% with the complete conversion of glycerol within 4 h of reaction time, which is superior or at least comparable to the best catalysts [14,17,18,36]. To investigate the kinetic behavior conveniently, the product yield of glycerol esterification as a function of time over Ag1PW catalyst was illustrated in Fig. S1. As mentioned above, the glycerol reaction rate declined quickly after 10 min of reaction time, revealing the rapid consumption of glycerol and accumulation of byproduct water at the initial stage. Then, the conversion of glycerol was restricted. As can be seen in Fig. S1, it took 35 min to convert the remaining glycerol (<20%) and much more time to attain the product distribution equilibrium. The consecutive reaction mechanism of glycerol esterification can be observed from the product distribution. The yield of MAG increased firstly with the reaction time, while that of MAG passed a maximum value before its decline. Nevertheless, the yield of DAG and TAG increased gradually with the reaction time, and only a relatively low amount of TAG was formed in the initial stage. When much MAG converted to DAG at the initial stage, DAG was yielded rapidly. However, DAG formation rate declined remarkably with the time due to its abundant consumption, while the TAG increased steadily. According to

To assess the reusability of Ag1PW catalyst for glycerol esterification, the spent catalyst was separated from the reaction system by centrifugation after the completion of each run. The spent sample was washed with ethanol, dried at 80 °C overnight prior to its reuse for further reaction cycle under the identical reaction conditions. As illustrated in Fig. 9, Ag1PW catalyst exhibited similar activity for glycerol esterification until five reaction cycles without showing distinct decline in catalytic ability. The above results have demonstrated that the Ag1PW catalyst is rather durable and holds the potential for practical applications. Furthermore, the inherent heterogeneous nature of Ag1PW catalyst can favor to recover the catalyst as well as the reaction products. The XRD and Raman spectroscopy were employed to check the state of spent Ag1PW catalyst. As illustrated in Fig. S2, the spent catalyst displayed similar characteristic bands as the fresh one, revealing that the Keggin structure of Ag1PW catalyst remained intact during this reaction. With respect to Ag2PW and Ag3PW, it was anticipated that both of them presented good reusability without obvious loss in the activity (Fig. S3), due to their better water tolerance than Ag1PW. As indicated by XRD and Raman spectroscopy of spent Ag2PW and Ag3PW in Fig. S2, both of them were rather durable. In order to explore the possible leaching of active component, an additional experiment of glycerol esterification with acetic acid was conducted by using Ag1PW catalyst at 120 °C for 15 min. Subsequently, to avoid the adsorption of possible leached metal ions,

161

S. Zhu et al. / Journal of Catalysis 306 (2013) 155–163

100

Conversion (%)

80

60

40

20

0 1

2

3

4

5

Reaction cycle Fig. 9. Reusability test preformed for Ag1PW catalyst in glycerol esterification with acetic acid. Reaction conditions: 120 °C, glycerol/acetic acid = 1:10 (molar ratio), 1 wt% catalyst loading relative to glycerol, 15 min.

the catalyst was removed from the reaction system at reaction temperature followed by continuing this reaction with the remaining solution for another 60 min [37]. The solution after reaction was checked by ICP, and it was verified that there was no leaching of Ag1PW during glycerol esterification. However, the product distributions did change marginally after 60 min of reaction time, which can be explained by the self-catalyzed characteristic of esterification. Therefore, the above results demonstrated unambiguously the inherent heterogeneous nature of Ag1PW catalyst, in well agreement with the properties of Ag-exchanged silicotungstic acid catalyst described previously [37].

3.5. Influence of carboxylic acids and alcohols for glycerol esterification Finally, the catalytic properties for esterification with various carboxylic acids and alcohols were examined to investigate the universal significance of Ag1PW. As shown in Table 4, the

conversion of glycerol esterification with various carboxylic acids decreased gradually with increasing chain length of carboxylic acids (entries 1–3) in the following order: acetic acid > propionic acid > 1-butanoic acid. Analogously, the conversion of primary alcohol esterification with acetic acid decreased gradually with increasing chain length of primary alcohols (entries 4–8) in the following order: ethanol  propanol > 1-butanol > 1-pentanol > 1hexanol. This similar trend has also been observed in the case of myristic acid esterification with different alcohols over sulfated zirconia and cetyl alcohol esterification with various acids over ZrOCl2/MCM-41, which can be explained by the steric hindrance effects of the larger alkyl chains either of acids or alcohols [38,39]. Additionally, the esterification of acetic acid with various polyols including ethylene glycol, 1,2-propanediol, 1,4-butanediol, and 1,5-pentanediol was also exhibited in the presence of Ag1PW catalyst. The esterification of acetic acid with polyols resulted in corresponding monoacetate and diacetate in good conversion despite the gradual decrease in the reactivity with increasing chain length of polyols. The above results suggested that the Ag1PW catalyst exhibited superior performance in the esterification of various alcohols and acids, being a potential catalyst for the selective esterification of glycerol with fatty derivatives which is an attractive pathway for the transformation of biomass-derived substrates. 3.6. Reaction mechanism of glycerol esterification It is well known that the acidity of catalyst has significant impact on the catalytic cycle and is necessary for optimal performance in glycerol esterification, particularly for Brønsted acid sites [14,15,18,20]. To elucidate the superior performance of Ag1PW catalyst, a plausible reaction mechanism of glycerol esterification with acetic acid was proposed in Scheme 1. Initially, the interaction of acetic acid with a proton site derived from Brønsted acid site results in the protonation of the carbonyl group in the acetic acid molecule possessing a higher positive charge. Subsequently, the carbon atom of the carbonyl can be attacked by the hydroxyl group of glycerol, which results in generating a new CAO bond between the carbon atom of carbonyl and the oxygen atom of hydroxyl group of glycerol. With the elimination of water, a new MAG molecule is formed. Owing to the steric constrains, the

Table 4 Esterification of various alcohols with carboxylic acids over Ag1PW catalyst.a.

a

Entry

Alcohol

Carboxylic acid

Conversion (%)

1

Glycerol

Acetic acid

96.8

2

Glycerol

Propionic acid

70.9

3

Glycerol

1-Butanoic acid

64.3

4 5 6 7 8 9

Ethanol Propanol 1-Butanol 1-Pentanol 1-Hexanol Ethylene glycol

Acetic Acetic Acetic Acetic Acetic Acetic

acid acid acid acid acid acid

99.5 99.1 95.2 93.2 82.4 99.2

10

1,3-Propanediol

Acetic acid

98.9

11

1,4-Butanediol

Acetic acid

98.0

12

1,5-Pentanediol

Acetic acid

81.2

Reaction conditions: 120 °C, alcohol/acid = 1:10 (molar ratio), 1 wt% catalyst loading relative to alcohol, 15 min.

Selectivity Product distribution

(%)

MAG DAG TAG Glyceryl monopropionate Glyceryl dipropionate Glyceryl tripropionate Glyceryl momobutyrate Glyceryl dibutyrate Glyceryl tributyrate Acetic ether Propyl acetate 1-Butyl acetate 1-Pentyl acetate 1-Hexyl acetate Ethylene glycol monoacetate Ethylene glycol diacetate 1,3-Propanediol monoacetate 1,3-Propanediol diacetate 1,4-Butanediol monoacetate 1,4-Butanediol diacetate 1,5-Pentanediol monoacetate 1,5-Pentanediol diacetate

48.4 46.4 5.2 55.0 43.1 1.9 53.5 46.4 0.1 100 100 100 100 100 38.9 61.1 61.7 38.3 71.5 28.5 95.8 4.2

162

S. Zhu et al. / Journal of Catalysis 306 (2013) 155–163 O HO

C CH3 O H3 C O CH3

C H3C

OH

H3 C O

O

OH O

CH3

C

O

O H AgH2PW12040

O

AgH2PW12040

H O

AgH2PW12040 C

O

H

O

HO

O

O

H3C

OH

HO

C

O

HO

O HO CH3

HO

O

HO C O

O

H3C

C

CH3

C

C

O

O

C

O AgH2PW12040

C

CH3

O H3C

C O

O H

AgH2PW12040

OH

O H3 C

C

C O

AgH2PW12040

O

H AgH2PW12040

AgH2PW12040

O

OH CH3

O

HO

H3 C

O O

C O

O

HO

H3C

O

CH3

O

CH3

O C

C C

O

H3 C

H3C

O

O

OH

H2 O

O H

AgH2PW12040

O CH3 H2O

H2O

Scheme 1. The proposed reaction mechanism of glycerol esterification with acetic acid.

secondary hydroxyl group of glycerol proceeds less nucleophilic attack on acetic acid than that of primary hydroxyl group, which leads to form less 2-MAG than 1-MAG, as evidenced by our GC– MS and the literature [14]. Similarly, the formed MAG proceeds nucleophilic attack on acetic acid and leads to the formation of DAG. Finally, DAG undergoes further esterification with acetic acid to yield TAG. On the other hand, Lewis acid sites may also play a role in the catalytic performance, as claimed by Troncea et al. [12]. The Lewis acid sites can act as active catalytic sites involved in the formation of reactive nucleophilic intermediate and interact with the by-product water to generate new Brønsted acid sites [12,40]. Thus, the high activity of Ag1PW catalyst can be related to the most acidic sites in terms of Brønsted acid and Lewis acid sites. It is speculated that both Brønsted and Lewis acid sites can activate carbonyl group of acetic acid and hydroxyl group of glycerol, which might be a rate-determining step. However, there is not a consistent correlation between glycerol activity and total acidity over these HPW and Ag-salts catalysts, as shown in Tables 2 and 3. In comparison with HPW, Ag2PW and Ag3PW possessed less acidic sites, but exhibited superior activity. These results might be related to the excellent water tolerance of Ag-salts catalysts, as indicated by H2O-TPD results. The primary structure (heteropolyanion) of HPW is pretty stable, but the secondary structure (heteropolyanion as well as protons) is fairly mobile, which can interact with polar water molecules; even the water molecules can enter into the bulk crystallite of HPW and form protonated clusters [41]. The good solubility of HPW in water originating from the esterification can affect acid strength and cause deactivation. Contrarily, the insoluble Ag-salts are highly hydrophobic and capable of inhibiting the diffusion of water molecules into the secondary structure, where located the acidic protons and exchanged Ag+ sites. Additionally, the generated water can be released rapidly from the secondary structure, which can remain the integrity of Keggin structure as well as acid strength and impair the reverse reaction. This behavior is characteristic of insoluble HPA salts and has already been reported in other water-sensitive reactions, such as hydrolysis of ester and dehydration [40,42]. On the basis of above discussions, it is concluded that the excellent performance of Ag1PW catalyst can be ascribed to

the unique Keggin structure, high amount of Brønsted and Lewis acid sites, and outstanding water tolerance.

4. Conclusions This study has demonstrated that a partially silver-exchanged HPW salt catalyst (Ag1PW) using simple preparation procedure, presented exceptionally high activity for glycerol esterification with acetic acid. The high Brønsted acid and Lewis acid sites, outstanding water tolerance, strong stability in polar reaction environment, and remaining of unique Keggin structure are responsible for the excellent performance of Ag1PW catalyst in this reaction. As far as we know, Ag1PW catalyst exhibited the best catalytic activity ever reported for glycerol esterification. Concurrently, this catalyst gave high selectivity to the desired DAG and TAG with a combined selectivity of 90.7%. Furthermore, Ag1PW catalyst did not suffer from leaching and deactivation in five consecutive reaction cycles. A plausible reaction mechanism was uncovered for further study and as references. This Ag1PW catalyst derived from HPA could be a promising candidate to substitute conventional mineral acids for the catalytic esterification of glycerol with acetic acid and for the synthesis of valuable biofuel additives. In addition to be suitable for esterification, such an environmentally friendly alternative catalyst holds great potential for a wide range of other acid-catalyzed reactions. Acknowledgments The authors gratefully acknowledge the financial support of the Natural Science Foundation of China (No. 20976185). This work was also supported by the Major State Basic Research Development Program of China (973 Program) (No. 2012CB215305).

Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jcat.2013.06.026.

S. Zhu et al. / Journal of Catalysis 306 (2013) 155–163

References [1] J.N. Chheda, G.W. Huber, J.A. Dumesic, Angew. Chem., Int. Ed. 46 (2007) 7164. [2] B. Smith, H.C. Greenwell, A. Whiting, Energy Environ. Sci. 2 (2009) 262. [3] P. Ferreira, I.M. Fonseca, A.M. Ramos, J. Vital, J.E. Castanheiro, Appl. Catal., B: Environ. 98 (2010) 94. [4] S. Zhu, Y. Zhu, X. Gao, T. Mo, Y. Zhu, Y. Li, Bioresour. Technol. 130 (2013) 45. [5] S. Zhu, X. Gao, Y. Zhu, Y. Zhu, X. Xiang, C. Hu, Y. Li, Appl. Catal., B: Environ. 140– 141 (2013) 60. [6] F. Wang, J.L. Dubois, W. Ueda, J. Catal. 268 (2009) 260. [7] L. Zhang, A.M. Karim, M.H. Engelhard, Z. Wei, D.L. King, Y. Wang, J. Catal. 287 (2012) 37. [8] J. Deutsch, A. Martin, H. Lieske, J. Catal. 245 (2007) 428. [9] S. Zhu, X. Gao, Y. Zhu, Y. Zhu, H. Zheng, Y. Li, J. Catal. 303 (2013) 70. [10] X. Liao, Y. Zhu, S.G. Wang, H. Chen, Y. Li, Appl. Catal., B: Environ. 94 (2010) 64. [11] F. Jérôme, Y. Pouilloux, J. Barrault, ChemSusChem 1 (2008) 586. [12] S.B. Troncea, S. Wuttke, E. Kemnitz, S.M. Coman, V.I. Parvulescu, Appl. Catal., B: Environ. 107 (2011) 260. [13] M.S. Khayoon, B.H. Hameed, Bioresour. Technol. 102 (2011) 9229. [14] X. Liu, H. Ma, Y. Wu, C. Wang, M. Yang, P. Yan, U. Welz-Biermann, Green Chem. 13 (2011) 697. [15] J.A. Melero, R. van Grieken, G. Morales, M. Paniagua, Energy Fuels 21 (2007) 1782. [16] V.L.C. Goncalves, B.P. Pinto, J.C. Silva, C.J.A. Mota, Catal. Today 133 (2008) 673. [17] X. Liao, Y. Zhu, S.-G. Wang, Y. Li, Fuel Process. Technol. 90 (2009) 988. [18] K. Jagadeeswaraiah, M. Balaraju, P.S.S. Prasad, N. Lingaiah, Appl. Catal., A: Gen. 286 (2010) 166. [19] P. Ferreira, I.M. Fonseca, A.M. Ramos, J. Vital, J.E. Castanheiro, Catal. Commun. 10 (2009) 481. [20] M. Balaraju, P. Nikhitha, K. Jagadeeswaraiah, K. Srilatha, P.S.S. Prasad, N. Lingaiah, Fuel Process. Technol. 91 (2010) 249.

[21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42]

163

I.V. Kozhevnikov, J. Mol. Catal. A: Chem. 262 (2007) 86. M. Misono, Catal. Today 144 (2009) 285. I.V. Kozhevnikov, J. Mol. Catal. A: Chem. 305 (2009) 104. S. Zhu, Y. Zhu, S. Hao, H. Zheng, T. Mo, Y. Li, Green Chem. 14 (2012) 2607. J. Haber, K. Pamin, L. Matachowski, B. Napruszewska, J. Połtowicz, J. Catal. 207 (2002) 296. S. Borghèse, B. Louis, A. Blanc, P. Pale, Catal. Sci. Technol. 1 (2011) 981. S. Borghe‘se, A. Blanc, P. Pale, B. Louis, Dalton Trans. 40 (2011) 1220. C.A. Emeis, J. Catal. 141 (1993) 347. K. Narasimharao, D.R. Brown, A.F. Lee, A.D. Newman, P.F. Siril, S.J. Tavener, K. Wilson, J. Catal. 248 (2007) 226. C.F. Oliveira, L.M. Dezaneti, F.A.C. Garcia, J.L. de Macedo, J.A. Dias, S.C.L. Dias, K.S.P. Alvim, Appl. Catal., A: Gen. 372 (2010) 153. A. Zieba, L. Matachowski, J. Gurgul, E. Bielan´ska, A. Drelinkiewicz, J. Mol. Catal. A: Chem. 316 (2010) 30. R. Kourieh, S. Bennici, M. Marzo, A. Gervasini, A. Auroux, Catal. Commun. 19 (2012) 119. V.L. Budarin, J.H. Clark, R. Luque, D.J. Macquarrie, A. Koutinasb, C. Webb, Green Chem. 9 (2007) 992. C.R. Kumar, K. Jagadeeswaraiah, P.S.S. Prasad, N. Lingaiah, ChemCatChem 4 (2012) 1360. L. Zhou, E. Al-Zaini, A.A. Adesina, Fuel 103 (2013) 617. P. Ferreira, I.M. Fonseca, A.M. Ramos, J. Vital, J.E. Castanheiro, Appl. Catal., B: Environ. 91 (2009) 416. K.T.V. Rao, P.S.S. Prasad, N. Lingaiah, Green Chem. 4 (2012) 1507. K. Saravanan, B. Tyagi, H.C. Bajaj, Catal. Sci. Technol. 2 (2012) 2512. K. Mantri, K. Komura, Y. Sugi, Green Chem. 7 (2005) 677. A. Alhanash, E.F. Kozhevnikova, I.V. Kozhevnikov, Appl. Catal., A: Gen. 378 (2010) 11. A. Micek-Ilnicka, J. Mol. Catal. A: Chem. 308 (2009) 1. M. Kimura, T. Nakato, T. Okuhara, Appl. Catal., A: Gen. 165 (1997) 227.