Advancements in Heterogeneous Catalysis for Biodiesel ... - CiteSeerX

1 downloads 0 Views 400KB Size Report
Apr 14, 2010 - strong base catalyzed transesterification is shown in Fig. 1. [17–19]. Lee et al. .... and the isomerization of safrole, dimethyl butene and vi- nylbicyclo ..... carbonate, hydroxide, and epoxide which cover the basic sites and ...

Top Catal (2010) 53:721–736 DOI 10.1007/s11244-010-9460-5

ORIGINAL PAPER

Advancements in Heterogeneous Catalysis for Biodiesel Synthesis Shuli Yan • Craig DiMaggio • Siddharth Mohan • Manhoe Kim • Steven O. Salley • K. Y. Simon Ng

Published online: 14 April 2010 Ó Springer Science+Business Media, LLC 2010

Abstract Heterogeneous catalysts are promising for the transesterification reaction of vegetable oils to produce biodiesel and have been studied intensively over the last decade. Unlike the homogeneous catalysts, heterogeneous catalysts can be easily separated from reaction mixture and reused for many times. They are environmentally benign and could be easily operated in continuous processes. This review classifies the solid catalysts into two categories based on their catalytic temperature, i.e. high temperature catalysts and low temperature catalysts. The nature of the catalysts can be specified into solid bases and solid acids. Three aspects, catalyst activity, catalyst life and oil flexibility, will be reviewed. Two kinds of heterogeneous catalysts, reported by IFP Inc. and by WSU, respectively, show a high catalytic activity, long catalyst life and low leaching of catalyst components. These two catalysts also show ability to simultaneously catalyze esterification and transesterification, and can be used in half-refined or crude oil system which provide a potential for greatly decrease the feedstock cost. Keywords Biodiesel  Heterogeneous catalyst  Transesterification  Esterification  Solid base  Solid acid

increasingly important to search for sustainable alternative fuels. Among the many possible sources, biodiesel derived from vegetable oil attracted early attention as a promising fuel for substitution or blending with petroleum based diesel fuel, because biodiesel and petroleum diesel share similar physical and chemical properties [1]. Thus, pure biodiesel or biodiesel blends can be used in conventional compression-ignition engines without the need for engine modifications [2]. Furthermore, certain properties of biodiesel, such as flash point, cetane number, ultralow sulfur content, lubricity, biodegradability, and smaller carbon footprint were all superior to petroleum diesel [3, 4]. 1.1 Biodiesel Chemical Background Biodiesel is an industry adopted term for a mixture of fatty acid alkyl esters, that are a product of the transesterification reaction of triglycerides with methanol. In that reaction, three alkyl esters are produced from one triglyceride molecule. A second product, glycerol, is also produced in a molar ratio of 1:1 glycerol:triglyceride, further adding to the value of processing oils. These reactions are represented by:

1 Introduction Due to growing worldwide demand for energy and its resulting impact on the environment, it is becoming

S. Yan  C. DiMaggio  S. Mohan  M. Kim  S. O. Salley  K. Y. S. Ng (&) Department of Chemical Engineering and Material Science, Wayne State University, Detroit, MI 48202, USA e-mail: [email protected]

This well established process, introduced in the nineteenth century, has been used to exploit the fatty acid and triglyceride content of a variety of natural oils for biodiesel

123

722 Fig. 1 Reaction mechanism of base catalyzed transesterification [17, 19]

Top Catal (2010) 53:721–736

(1)

ROH

+

B

R O

-

+B

+

H

OR

O

(2)

O

R2

+

O

OH

O

R1

O

R O

O

R2

-

R1

O O

O

R3

R3 O

O OR

(3)

R2

O

OH

O O

O

R1

-

O

R2

+

O

O

R3

R3 O

O

O

(4)

O -

O

R2

+

O O

+

B H

R2

OH

+

O

O

fuel over the years. Among these renewable oils and fats are soybean oil [5], rapeseed oil [6], cotton seed oil [7], sunflower seed oil [8], jojoba oil [9], waste cooking oil [10], chicken fat [11], lard [12], and beef tallow [3]. Typical alcohols used as the reactant in this biodiesel process include methanol [5], ethanol [13] and butanol [14], with methanol being the most widely used because of its lower cost. Some of the first industrial processes to create biodiesel relied on either strong base or strong acid homogeneous catalysts for this transesterification reaction. Examples of the base catalyst are potassium hydroxide [9, 13] and sodium hydroxide [14, 15], while sulfuric acid has been used as an acid catalyst [16]. The reaction mechanism for strong base catalyzed transesterification is shown in Fig. 1 [17–19]. Lee et al. [19] described the first step in the synthesis of alkyl esters as the formation of an alkoxide ion (RO–) through proton transfer from the alcohol using the base catalyst; the alkoxide ion then attacks a carbonyl carbon on the triglyceride molecule and forms a tetrahedral intermediate ion (step 2). This ion rearranges to generate a diglyceride ion and alkyl ester molecule (step 3). The diglyceride ion reacts with the protonated base catalyst,

B

O R3

123

OR

R1

O

O

R3 O

which generates a diglyceride molecule and returns the base catalyst to its initial state (step 4). The resulting diglyceride is then ready to react with another alcohol molecule, thereby starting the next catalytic cycle until all the glyceride molecules have been converted to alkyl esters (biodiesel). Strong acid catalyzed transesterification is illustrated in Fig. 2 [20]. This process can simply be described as the protonated carbonyl group nucleophilicly attacks the alcohol, forming a tetrahedral intermediate; the proton then migrates, and the intermediate decomposes forming a new ester. This process can be extended to di- and mono-glycerides as well. Additionally, research shows that heterogeneous catalysis, both base and acid, also follow the above mentioned mechanism for alkyl ester production [19]. 1.2 First Generation Homogeneous Catalysts and Their Limitations Traditional or first generation homogeneous catalysts enjoy certain advantages over other catalysts including costeffectiveness, high activity, and easily attained reaction conditions (25–130 °C, atmospheric pressure). However,

Top Catal (2010) 53:721–736 Fig. 2 Reaction mechanism of acid catalyzed transesterification [17, 20]

723 +

O

OH

O

(1)

R2

R1

O

O H

+

R1

R2

O

O

O O

O R3

R3

O

O +

OH O

(2)

R2

R1

HO

O

O

O

O

+

R4

R2

R4 OH

O

O

H

+

R1

O R3

O

O

R3 O

HO

O

+

O

(3)

R2

H O R4

O

R1

OH

R2

O

O

+

O O

R4 R1

O

+

+

H

O R3

R3 O

O

these same homogeneous catalysts, by virtue of the associated production process, face a variety of technical hurdles that limit their use for biodiesel production and eventually may cause the demise of the early biodiesel producers. Homogeneous catalysts are normally limited to batch-mode processing [21]. In addition, other steps in the biodiesel production process also require time consuming and costly processing. These steps include oil pretreatment, catalytic transesterification, separation of fatty acid/methyl ester (FAME) from crude glycerin, neutralization of waste homogeneous catalyst, distillation of accessory methanol, water washing of the FAME phase, and vacuum drying of the desired products [22]. Each of these steps introduce additional processing time and cost. As an example, separation of the products from the spent waste catalyst require a post treatment with large volumes of water to neutralize the used catalyst in the product mixture. This creates an additional process burden by generating waste water that must be treated before release into the environment [22]. Other difficulties with using homogeneous catalysts center on their sensitivity to free fatty acid (FFA) and water in the source oil. FFAs react with basic catalysts (NaOH, KOH) to form soaps when the FFA and water content are above 0.50 and 0.06%, respectively [5, 23]. This soap formation complicates the glycerol separation, and reduces

the FAME yield. Water in the feedstock results in the hydrolysis of FAME in the presence of strong basic or acidic catalyst. Thus, some inexpensive oils, such as crude vegetable oils, waste cooking oil, and rendered animal fats, which generally contain a high content of FFA and water, cannot be directly utilized in existing biodiesel facilities with homogeneous catalysts. Likewise, the cost of oil feedstock in 2006 accounted for up to 80% of biodiesel production cost [24, 25]. So when petroleum diesel prices fell in 2008, the relatively expensive soybean derived biodiesel could not ompete, forcing many biodiesel facilities to close. Therefore, part of the current solution is to develop a second generation technology based on heterogeneous catalysts that are capable of effectively processing less costly feedstocks high in FFAs and water content with a simpler less costly processing method. 1.3 Second Generation Heterogeneous Catalyst Advancements Recent developments in heterogeneous catalysis for biodiesel production has the potential to offer some relief to the biodiesel producers by improving their ability to process alternative cheaper feedstocks, and to use a shortened and less expensive manufacturing process. Whereas homogeneous based process required batch mode operation,

123

724

heterogeneous processes can be run in either batch or continuous mode giving the producers the option to continue with their current batch reactors or retrofit their operations with a packed bed continuous flow reactor operation. Heterogeneous catalysts in either mode are in a separate phase from the reaction products, thereby removing costly and time consuming water washing and neutralization steps to separate and recover the spent catalyst. Additionally, contaminated water from that process is greatly reduced and the need for waste water treatment minimized. The greatest advantage of the heterogeneous approach over the homogeneous method is the prolonged lifetime of the heterogeneous catalysts for FAME production. This attribute is generally related to the stability of the microcrystal structure of the catalyst surface. Poisoning and leaching of catalyst components can change the bulk and surface structure of the catalyst and cause catalyst deactivation quickly if the catalyst is not formulated properly. The above three factors—catalytic activity, catalyst life and oil flexibility—have a tremendous impact on the cost of biodiesel. Because of this, we have undertaken a review of the past and current heterogeneous catalyst technology with these aspects as the focus. In this paper, the reported heterogeneous catalysts are separated into two categories based on their operation temperature. For reaction temperatures lower than the flash point of biodiesel (130 °C), we refer to this type of catalyst as a low temperature catalyst. For reaction temperatures greater than 130 °C, we classified these catalysts as high temperature catalysts. These catalysts are characterized by the need for additional safety considerations and require more energy intensive operations. To further reduce the field of catalysts for this review, only heterogeneous catalyst technologies which provide the potential for decreasing the biodiesel process cost and feedstock cost will be discussed.

2 Low Temperature Catalysts As previously stated, studies of solid base catalysts began burgeoning in the 1970 s. Most dealt with common single metal oxides such as alkaline oxides and rare earth metal oxides. Subsequently, studies were expanded to include alkali metal exchanged zeolites, alkali metal ion-supported catalysts, and clay minerals such as hydrotalcites. 2.1 Solid Base Catalysts 2.1.1 Alkaline Metal Salts on Porous Supports Alkali metals are the most common source of super basicity and are frequently selected as the active species for biodiesel production. The loading of many kinds of

123

Top Catal (2010) 53:721–736

alkaline salts on supports have been reported as a way to prepare basic catalysts, such as NaOH [2, 26–29], KOH [30, 31], K2CO3 [32], KI [33, 34], KNO3 [35, 36], KF [37– 40], and LiNO3 [41, 42]. The supports for these catalysts include Al2O3 [33, 39], Zeolite [43], ZnO [41] and SiO2 [34]. An example of a commercialized super base catalyst is Na/NaOH/Al2O3. It is used for the alkylation of cumene, and the isomerization of safrole, dimethyl butene and vinylbicyclo heptene [44]. Kim et al. [29] also tested its activity for soybean oil transesterification with methanol and found almost the same activity as homogeneous NaOH catalyst under optimized reaction conditions (FAME yield was 94% with a reaction temperature of 60 °C, reaction time of 2 h, stirring speed 300 rpm, co-solvent n-hexane 10 mL, amount of catalyst 1 g). The basicity is believed to be associated with the Lewis base concept according to the O 1s XPS results presented. Table 1 shows the effect of preparation on catalyst basicity. The oxygen 1s binding energy shifts downward as the Na and NaOH impregnation onto the c-Al2O3 support progresses, indicating that the basicity increases together with the degree of impregnation. Consistent with O 1s binding energy, the catalyst’s FAME activity also proportionally increases. Xie et al. [35] and Vyas et al. [36] investigated the activity of KNO3/Al2O3. Xie et al. [35] pointed out that the active phase was K2O derived from KNO3 at high temperature, and the surface Al–O–K groups were the main active sites. Cui et al. [39] prepared KF/c-Al2O3 and found there were two types of basic sites on the catalyst. The strong basic sites (super basic) promote the transesterification reaction at low temperature (65 °C), while the basic sites with medium strength require a higher temperature to process the reaction. Later, Boz et al. [40] prepared KF catalysts loaded on nano-c-Al2O3 and Wang et al. [45] loaded KF on malodorous CaO–MgO. They both found that the catalyst’s FAME activity is closely related to the basic nature of the catalyst and also to the high surface to volume ratio and porosity of the catalyst.

Table 1 XPS analysis of O 1s orbital of four catalysts Catalyst

Binding energy of O 1s (eV)

Biodiesel yield (%)

c-Al2O3

538.8

5

NaOH/ c-Al2O3

538.5

60

Na/ c-Al2O3

537.5

70

Na/NaOH/ c-Al2O3

535.5

78

Reaction conditions: Methanol/oil molar ratio is 6:1, reaction temperature is 60 °C, and stirring speed is 300 rpm. All data are taken from literature [29]

Top Catal (2010) 53:721–736

However, in spite of the high activity of the supported alkaline catalysts, they have important limitations. First, these catalysts like their homogeneous alkaline hydroxide counterparts have a low tolerance to FFA and water in raw materials. At this time, there is no report of using this kind of catalyst for directly processing crude oils which have a high total acid number (TAN). Only refined oils can be used with these catalyst systems. Secondly, some researchers have observed lixiviation of catalyst components into reaction mixtures. Arzamendi et al. [46] found that 55% of K2CO3, 20% of Na2CO3 and 15% of Na3PO4 dissolved into the reaction mixtures and catalyzed the transesterification reaction. Also, Xie et al. [35] found KNO3/Al2O3 catalysts have a high solubility in water and were therefore unstable in the transesterification system. However, Noiroj et al. [47] found that the type of support strongly affected the activity and leaching of the active species of the catalyst. In this case, the amount of leached potassium of the KOH/Al2O3 was higher than that of the KOH/NaY catalyst. And they found that the interaction between active phase and support affected the leaching results. Additionally, Ramos et al. [48] prepared sodium hydroxide on a zeolite support and hypothesized the presence of a homogeneous-like mechanism where the alkali methoxide species were leached out. 2.1.2 Alkaline Earth Metal Oxide Catalysts Much attention has been paid to alkaline earth metal oxides since they have shown less solubility in reaction mixtures and less corrosion in comparison to supported alkaline catalysts. As solid super base can be synthesized from alkaline metal oxides, researchers started from pure alkaline metal oxides. In fact, alkaline metal oxides have already been used as base catalysts in many organic reactions. For example CaO is widely used for as the isomerization of 5-vinylbicyclo [2.2.1] hept-2-ene (VBH) to 5-ethylidenebicyclo [2.2.1] hept-2-ene (ENB) [49, 50], synthesis of 1, 3-dialkylurea from ethylene carbonate and amine [51] and the synthesis of monoglyceride [52]. With respect to biodiesel production, the basicity of this type of metal oxide catalyst has been shown to have an influence on its activity for FAME generation. The basic strength of the Group II metal oxides follows the order: MgO \ CaO \ SrO \ BaO. Corresponding research has demonstrated the catalyst’s activity for transesterification of oil with methanol follows the same order [53–55]. But, compared to a homogeneous NaOH catalyst, the above alkaline earth metal oxides show a relatively low transesterification activity. In particular, MgO exhibits almost no activity in transesterification of vegetable oils into biodiesel. Pure CaO reacts at a slow rate and requires about 6–24 h to reach a state of reaction equilibrium [56–58]. BaO is not

725

suitable for biodiesel production because it dissolves in methanol and forms some noxious species [59, 60]. Conversely, SrO, has a high activity and is insoluble in methanol, but will react strongly with CO2 and water in the air to form unreactive SrCO3 and Sr(OH)2. Furthermore, these strontium compounds are difficult to regenerate by calcining, requiring temperatures above 1200 °C [17]. As a partial solution to these limitations, recent work has focused on using mixed metal oxides to enhance the basicity of CaO or MgO-based catalysts and elevate their respective selectivity for FAME. 2.1.2.1 Supported CaO Catalysts Although earlier work showed weak FAME activity for pure CaO catalysts, more recent research has shown that the smaller particle size of CaO catalysts can increase the total amount of base sites and base strength, which leads to an improved activity in the oil transesterification reaction. Reddy et al. [61] tested the activity of nanocrystalline CaO and found it active even at room temperature. However, as Gryglewics [6, 62] and Martyanov and Sayari [63] pointed out, pure CaO converted to a form of suspensoid due to its poor mechanical strength, which would lead to difficulties in separating the waste catalyst from biodiesel and glycerol products after transesterification. Since these findings could have a potential impact on industrial applications, many researchers have tried to solve this problem by applying CaO on different metal oxide supports. In particular, CaO has been combined with ZnO [64], MgO [46, 65], Al2O3 [65, 66], zeolite [48, 65], SiO2 [65, 67], and La2O3 [68] with improved base characteristics, activity and catalytic life. Rubio-Caballero et al. [64] used the calcined calcium zincate as a solid catalyst for the methanolysis of sunflower oil to FAME resulting in yields higher than 90% after 45 min of reaction. The reaction conditions of the heterogeneous process (60 °C, methanol: sunflower oil molar ratio of 12, 3 wt% catalyst) were very similar to those observed under homogeneous conditions (KOH dissolved in methanol). Yan et al. [65] investigated the effects of a second metal oxide by impregnating CaO on basic oxides such as MgO, neutral oxides such as SiO2, and acidic oxides such as Al2O3 and zeolite HY. In this work, the best results were obtained for a catalyst with 16.5% of CaO loading on MgO. The same catalyst also possessed the strongest base strength and largest number of base sites. The conversion of rapeseed oil using this catalyst reached 92% at 64.5 °C. Further work by Yan et al. revealed the active centers of the CaO/MgO catalyst. CO2TPD profiles of CaO/MgO showed that there were two types of basic sites each with a different strength. The desorption peaks of CO2 at *600 °C were attributed to the strong basic sites corresponding to unbounded O2- anions, while CO2 desorption peaks at low temperature (*350 °C)

123

726

were attributed to the weak basic sites related to oxygen in both Ca2?–O2- and Mg2?–O2- pairs. Yan et al. [17, 65, 68] also quantified the effects of base property on oil transesterification. They found that the activity of CaO/MgO linearly increased with base amount, and base amount linearly increased with CaO loading when the CaO loading was lower than 16.5%. Later, Yan et al. [70] combined CaO with the basic oxide La2O3, and used the Hammett method to determine the base properties of catalysts. They found that the binary metal oxides had a higher base strength and a wider base site distribution than pure CaO and La2O3 catalysts individually, and they also had a higher reactivity than CaO and La2O3 individually. Using this binary metal oxide catalyst with a 3:1 of molar ratio of Ca to La, they found the FAME yield reached 94.3% within 60 min at just 58 °C. This suggested a reaction rate much closer to that of a homogeneous NaOH catalyzed processes and that heterogeneous catalysts were capable of attaining the same activities as homogeneous catalysts. Equally important here, this work indicated that CaO type heterogeneous catalysts also showed a high tolerance to water and FFA which is present in unrefined raw oil feedstocks. This implies these heterogeneous catalysts have a potential for biodiesel production. Until now, many CaObased catalysts were reported to be more tolerant than the supported alkaline catalysts. Yan et al. [65] reported that conversion of rapeseed oil by CaO/MgO reached as high as 98% when the water content of the raw oil was in the range of 0–2 wt% and the total acid number was below 7.4 mg KOH/g (FFA content around 3.7%). Later, Yan et al. [70] reported that CaO–La2O3 was active when the oil contained 10% of water and when the FFA content was lower than 3.5%. They then tested the basicities of the catalysts which had adsorbed small amounts of FFA and water, respectively, naming the catalysts Ca3La1–FFA and Ca3La1–water. Yan and co-workers found that the basic properties of the Ca3La1–water catalyst are much closed to the fresh CaO–La2O3 catalyst; therefore, it can be assumed that CaO–La2O3 shows a high tolerance to small amounts of water, while the basic property of the Ca3La1–FFA catalyst notably decreased both the base strength and sites, indicating CaO–La2O3 has a low tolerance for FFA in oils. Further characterization results indicated that there are Lewis base sites and Bronsted base sites on the surface of fresh CaO–La2O3 catalysts and both of these base sites are active centers for oil transesterification with methanol. In fact, the addition of small quantities of water can change the Lewis base sites into Bronsted base sites suggesting Ca3La1–water is still active in transesterification. Conversely, FFA will react with and bind to both base sites resulting in poisoning of the catalyst. Therefore, Ca3La1– FFA shows a low activity for FAME production. Further work by Yan et al. [70] using CaO–La2O3 for processing

123

Top Catal (2010) 53:721–736

crude soybean oil, crude palm oil and waste cooking oil, which satisfy the limitation of water content lower than 10% and FFA lower than 3.5%, showed a FAME yield in excess of 95% within 3 h. For the oils with a high FFA content, dilution with refined oil will lower the FFA content of the mixture again allowing the use of a CaO–La2O3 catalyst to produce a high FAME yield. Equally important to the usefulness of a catalyst is its lifetime. Reddy et al. [61] found that nanocrystalline calcium oxide particles deactivated after eight cycles with soybean oil and after only three cycles with higher FFA content poultry fat. Similarly, Kawashima et al. [57, 58] found evidence of decreased FAME activity with CaTiO3 and CaZrO3 catalysts after as few as three cycles. To overcome these limitations, therefore, it is important to understand the mechanism of catalyst deactivation. In general, there are three pathways by which catalyst deactivation can occur: poisoning, blocking of reactant fragments and lixiviation. Many studies have already paid close attention to poisoning due to high levels of either FFA or water present in raw materials [65, 70], so they will not be discussed here. But little research has been performed that focuses on the effects of exposing a stored catalyst to CO2, moisture, and O2 present in the air. As stated by Busca [69], base catalysts can easily react with these components of ambient air to form very stable surface species like carbonate, hydroxide, and epoxide which cover the basic sites and deactivate the base catalysts. Yan et al. [70] found that when comparing the activity of a fresh CaO–La2O3 catalyst to one exposed to air for 12 h, the yield of FAME sharply decreased from 96.8 to 34.5%, and the total base amount decreased from 14.0 to 1.5 mmol CO2/g. They later attributed this decrease in activity to moisture and CO2 in air restructuring the surface of CaO–La2O3 catalyst from metal oxides to hydroxide and carbonate. Other research groups have found that the reused base catalysts have a lower basic strength and a lower activity than fresh the catalysts [71, 72]. This was explained by the blockage of active surface sites on the catalyst by strongly adsorbed intermediates or product species. In particular, Martyanov and Sayari [63] studied reused catalysts (CaO, Ca(OCH3)2) and found that surface adsorbed butyric acids were the most likely species responsible for the catalyst deactivation in their experiments. Catalyst lixiviation or leaching is another frequently encountered pathway for catalyst deactivation. Many researchers have studied the leaching of catalyst components into reaction mixtures. Kouzu et al. using a pure CaO catalyst [71] found that the calcium concentration would be as high as 3065 ppm in the FAME when waste cooking oil was used as the source oil. Later Kouzu et al. [72] showed calcium in the products of transesterified refined soybean oil. In that case, he found that the calcium content in

Top Catal (2010) 53:721–736

glycerol was about 2000 ppm and calcium content in FAME was around 10–100 ppm. Similarly, Granados et al. [73, 74] studied the leaching of species from solid CaO and the role of these species in the catalytic reaction. He pointed out that CaO can react with glycerol to form Ca diglyceroxide which is more soluble than CaO and active in oil transesterification. He found that the solubility of CaO in alcohol is around 0.6 mg/mL and when CaO was less than 1 wt% the major reaction mechanism is homogeneous. When the catalyst loading is greater than 1 wt% CaO, the total homogeneous contribution is much smaller than that arising from the heterogeneous sites [74]. Thus, it is very apparent that determining the homogeneous contribution to the FAME yield of a CaO-based catalyst system, and other like catalysts, is as important as quantifying the heterogeneous component. 2.1.2.2 Supported MgO Catalysts One of the drawbacks of using a CaO-based catalyst is the low BET surface area associated with the catalyst. Because activity is closely related to surface area for many catalyst systems, loss of active surface area through deactivation can have a proportionally larger effect on the product yield. Therefore, one solution would be simply to use more catalyst in your industrial application design. However, this introduces additional cost into the plant design and material usage. To avoid these concerns some research groups have turned to metal oxide catalysts from hydrotalcites which are welldispersed, have a high surface area, and are characterized by a strong base property. One such example is a MgO based catalyst. Cantrell et al. [55] reported on a supported MgO–Al2O3 catalyst which was active in transesterification of glyceryl tributyrate using methanol. He found that the BET surface area was as high as 166 m2/g. Using the same catalyst, Xie et al. [75] found that the Mg/Al molar ratio also had an effect on FAME activity. Using a Mg/Al ratio of 3, 773 K calcination temperature, 15:1 M ratio of soybean oil: methanol, and a catalyst dosage of 7.5 wt%, oil conversion was found to be 67% after 9 h. Other work by Li et al. [76], using mixed oxides from Mg–Co–Al–La hydroxide, found those catalysts maintained its activity for 7 recycles in a batch reactor. Additional work on the MgO– Al2O3 catalyst, by Fraile et al. [77], found that the reaction mechanism relied on the residual alkaline ions as the main source of strong basicity and catalytic activity in the transesterification of sunflower oil with methanol.

727

various commercial resin catalysts. They found that the anion-exchange resins, such as Diaion PA308, PA306s, HPA25 (Mitsubishi Chemical C., Ltd, Tokyo, Japan), exhibited much higher catalytic activity than the cationexchange resins like PK208 (Mitsubishi Chemical C., Ltd, Tokyo, Japan). The anion-exchange resins characterized by a lower cross-linking density and a smaller particle size produced both a high reaction rate and conversion. The best catalytic performance was obtained on Diaion PA306s resin, which yielded over 80% conversion of soybean oil to ethyloleate after 3 h reaction at 323 K. However, other groups have found less satisfactory results using base resin catalysts. In particular, Aracil and coworkers [81] used an anion-exchange resin in the transesterification of sunflower oil to biodiesel and found the conversion was less than 1% after 8 h at a typical reaction temperature of 333 K. Kim et al. [82] found that trace amounts of CH3ONa, functioning as a homogeneous catalyst, exhibited a synergetic effect with the resin catalyst for conversion. 2.1.4 Biont Shell Based Catalysts Recently, catalysts derived from renewable materials, such as shrimp shell [83, 84], turtle shell [83], crab shell [83], oyster shell [85] and egg shell [86] have been employed for conversion of oils to FAME. Previously, these catalysts were generally considered as waste. The major components of these biont shells are chitin, protein and CaCO3. Normally, disposal of these waste materials from seafood processing are an economic or environmental problem for entrepreneurs and local governments. However, biodiesel production catalysts prepared from these ‘‘wastes’’ are a promising ‘‘green’’ technology. Xie et al. [83] first reported preparing biont shell supported KF catalysts for biodiesel production and found methylester yields from rapeseed oil as high as 97.5% within 3 h under optimal reaction conditions. Yang et al. [84] reported, a three step preparation procedure for a shrimp shell catalyst which included incomplete carbonization of the shrimp shells, loading KF onto the altered shells, and activation. In a different approach, Nakatani et al. [85] and Wei et al. [86] prepared CaO catalysts by simple calcination of oyster shells and egg shells, high in CaCO3 (95%), at 700–1000 °C. Although these catalysts proved active for biodiesel synthesis, further work is required to limit the amount of Ca leaching to improve the catalyst life and tolerance to water and FFA in oil feedstock.

2.1.3 Base Resin Catalysts

2.2 Solid Acid Catalysts

Not only inorganic bases, but also some organic bases were also tested for biodiesel production [78, 79], especially for base resin catalysts. Shibasaki-Kitakawa et al. [80] investigated the transesterification of triolein with ethanol using

Even though many heterogeneous base catalysts have been reported as highly active for biodiesel synthesis, they still cannot tolerate acidic oils with FFA content [3.5%, such as yellow and brown grease. However, sulfur based acidic

123

728

homogeneous catalysts such as H2SO4 show a much higher tolerance to FFA and water than the basic homogeneous NaOH and KOH catalysts, suggesting these catalysts may be more suited for processing acid oils. Using this line of reasoning, some research turned to investigating sulfur based heterogeneous acid catalysts for converting acidic oils into biodiesel. 2.2.1 Sulfated Zirconia Based Catalysts Sulfated metal oxides show superacid properties because of the interaction between the sulfate group and the metal oxide centers. These kinds of catalysts, including sulfated zirconia [87–89] and sulfated tin oxides [90] have been widely used in esterification and transesterification reactions under mild conditions. Kiss et al. [87] studied several solid acid catalysts (zeolites, ion-exchange resins, and mixed metal oxides) as catalysts for the esterification of dodecanoic acid with 2-ethylhexanol, 1-propanol, and methanol. That work revealed that sulfated zirconia was the most active for esterification. Later, Garcı´a et al. [91] investigated the activity of sulfated zirconia for soybean oil transesterification. That group found that the catalyst preparation method had a significant effect on the resulting catalyst activity. Under optimized conditions (120 °C, 1 h and 5 wt% of catalyst) and using sulfated zirconia prepared by a solvent-free method, the methanolysis of soybean oil was 98.6% and ethanolysis was 92.0%. The sulfated zirconia prepared by standard methods [88] was poor for soybean oil methanolysis (conversion of 8.5%) and conventional zirconia even less so. Similarly, Suwannakarn et al. [89] studied the activity and stability of a commercial sulfated zirconia catalyst for transesterification of tricaprylin with a series of aliphatic alcohols at 120 °C. He found that the catalytic activity decreased as the number of carbons in the alkyl chain of the alcohol increased. In addition, the sulfated zirconia catalyst exhibited significant activity loss with subsequent reaction cycles. Characterization of the recycled catalysts showed that the concentration of the SO42- moieties in the sulfated zirconia had permanently decreased. Essentially, the SO42- species were leached out. As explained by Yadav and coworkers [92, 93], the sulfate groups leached out as H2SO4 and HSO4-, which in turn gave rise to a homogeneous acid catalysis which interfered with activity measurements of the intended heterogeneous catalyst. 2.2.2 Heteropolyacid Catalysts A series of heteropolyacid (HPAs) catalysts have also attracted much attention due to their high activity in

123

Top Catal (2010) 53:721–736

biodiesel formation reactions, both transesterification and esterification. Alsalme et al. [94] studied some HPA catalysts and compared them with some homogeneous and heterogeneous catalysts such as H2SO4, Amberlyst-15, and zeolites HY and H-Beta. The intrinsic catalytic activity, expressed as turnover frequency (TOF), of the HPA catalyst is significantly higher than that of the conventional acid catalysts in these reactions. They also tested the catalytic activity and acid strength of several kinds of HPA catalysts. The TOF values decreased with decreasing catalyst acid strength in the order: H3PW12O40 & Cs2.5H0.5PW12O40 [ H4SiW12 O40 [ 15%H3PW12O40/Nb2O5, 15%H3PW12O40/ZrO2, 15% H3PW12O40/TiO2 [ H2SO4 [ HY, H-Beta [ Amberlyst-15. They found that Cs2.5H0.5PW12O40 exhibits high catalytic activity as well as high resistance to leaching. The other types of supported HPA catalysts suffered from leaching and exhibited a significant homogeneous component to the catalyst’s activity caused by the leached HPA. Pesaresi et al. [95] studied the catalytic mechanisms of CsxH4-x SiW12O40 (x = 0.8–4) in the transesterification of C4 and C8 triacyglycerides and esterification of a C16 FFA. The catalyst material, loading C1.3 Cs per Keggin, provided an insoluble, heterogeneous catalyst active for both transesterification and esterification, with reactivity correlating with the number of accessible H? sites residing within the mesopore structure. For loadings B0.8 Cs per Keggin, transesterification activity arises from the homogeneous contribution. Narasimharao et al. [96] investigated structure related activity for CsxH3xPW12O40 (x = 0.9–3). Materials with the Cs content in the range x = 2.0–2.7 were well dispersed, having a high surface areas *100 m2/g-1 and high Bronsted acid strength. CsxH3-xPW12O40 was active in both esterification of palmitic acid and transesterification of tributyrin. Further work showed an optimum performance occurs for Cs loadings of x = 2.0–2.3, correlating with the accessible surface acid site density. These catalysts were recovered for three times and leaching of soluble heteropolytungstate wasn’t observed. Other HPAs were also reported. Katada et al. [97] found that H4PNbW11O40, H3PW12O40 and the heteropolyacidderived solid acid catalyst, H4PNbW11O40/WO3–Nb2O5, were highly active for the transesterification of triolein with ethanol. But, H4PNbW11O40 and H3PW12O40 dissolved into the reaction mixture; H4PNbW11O40/WO3–Nb2O5 was insoluble to the reaction mixture. Further study showed that the activity of H4PNbW11O40/WO3–Nb2O5 was sensitive to calcination temperature, and calcination around 773 K provided a highly active catalyst. The activity was observed in the co-presence of water (3.9 wt%) and oleic acid (5 wt%). In a fixed-bed continuous-flow reaction, it maintained the yield of FAME around 25–40% for 4 days.

Top Catal (2010) 53:721–736

2.2.3 Organically-Functionalized Acid Catalysts The purpose of preparing organically-functionalized acid catalysts is to overcome the shortcomings of other acid catalysts, such as leaching and low surface area. Some attempts have been made with the sulfonic acid ionicexchange resins, such as Poly (DVB) resin sulfonated with H2SO4 [98], Amberlyst-35(Rohm & Haas) [98], Amberlyst-15 (Rohm & Haas) [94, 99], and Nafion SAC-13[100]. Rezende et al. [98] prepared different polymer supports based on styrene and divinylbenzene which were conveniently functionalized with sulfonic acid. In order to obtain an appropriated triglyceride conversion at low temperature (65 °C), it was necessary to use a high ratio of methanol to oil (50:1–300:1) and high catalyst dosage (25–50%). Under the optimal conditions FAME yields reached a maximum value over 90% using a sulfonated poly (DVB) ionexchange resin which had 442 m2/g of specific surface area and 3.4 meqH? g-1 of acid capacity. Some polymer based catalysts were claimed to be active for both oil transesterification and fatty acid esterification reaction in unrefined oil systems. As an example, Soldi et al. [101] prepared sulfonated polystyrene compounds where sulfonation was between 5.0 and 6.2 mmol SO3H/g of dry polymer. That work showed conversion of beef tallow, with a 53 mg KOH/g acid number, reached 70% within 18 h. 2.2.4 Natural Based Catalysts A novel type of renewable catalyst has been prepared from various carbohydrates such as D-glucose, sucrose, cellulose and starch [102–104]. These catalysts were made by incomplete carbonization of carbohydrates followed by sulfonation. The incomplete carbonization of D-glucose leads to a rigid carbon material consisting of small polycyclic aromatic carbon sheets in a three dimensional sp3bonded structure [103]. Sulfonation has been demonstrated to provide a highly stable solid with a high density of active SO3H sites. This type of catalyst was found to be physically robust without leaching of SO3H groups during use. This resulted in remarkable catalytic performance for FAME formation reactions for both transesterification and esterification [103, 105, 106]. In studies by Lou et al. [106], carbohydrate-derived acid catalysts had been successfully applied to biodiesel production with higher fatty acid oils, such as waste oils with high acid values. Variables such as starting material, carbonization temperature and time, and sulfonation temperature and time for catalyst preparation all had a significant impact on the catalytic and textural properties of the prepared solid acids. Under optimal reaction conditions (80 °C, 20: 1 of molar ratio of methanol to oil, 10 wt% of catalyst loading, over a starch-derived sulfonic acid catalyst), the FAME yield was measured at

729

about 92% after 8 h’s reaction. Furthermore, the starchderived solid acid catalyst proved exceptionally stable under reaction conditions.

3 High Temperature Catalysts Even though solid acid catalysts exhibit improved activity for converting acid oils into FAME, most of them show a relative low reaction rate and deactivate quickly in comparison to solid base catalysts. Intensifying reaction conditions, by increasing reaction temperature and pressure, has been shown to effectively accelerate the reaction rate and prolong the catalyst lifetime. Some of the more successful examples of these catalysts that function at higher temperature and pressure, including both solid base and acid catalysts, were tested in subcritical or supercritical methanol flow conditions (240 °C, 8 MPa) and reported below. 3.1 Solid Base Catalysts Several strong base catalysts have been tested at high reaction temperatures. One of which, CaO investigated by Demirbas [107], was operated under supercritical methanol conditions. When the temperature was 252 °C, transesterification was completed within 6 min with 3 wt% CaO and 41:1 methanol/oil molar ratio. In other work on Ca based catalysts, Suppes et al. [108] found that CaCO3 was active when the temperature was greater than 200 °C and required about 18 min to essentially convert all the oil; FFA in the oil was esterified by CaCO3 and did not appear to inhibit the catalyst; Also, no decrease in activity of the calcium carbonate was observed after weeks of utilization suggesting little leaching or deactivation. Separately, Barakos et al. [109] used both non-calcined and calcined Mg–Al–CO3 hydrotalcite catalysts in refined cottonseed oil, acidic cottonseed oil, and crude animal fat feedstock. Mg–Al–CO3 hydrotalcite catalysts were active in both transesterification and esterification. The activity of the calcined catalyst was lower than the non-calcined catalyst. But, the non-calcined catalyst showed evidence of deactivation when recycled. However, additional information regarding catalyst leaching and stability was not presented in this work. 3.2 Solid Acid Catalysts 3.2.1 Sulfate Salts As with the low temperature catalysts, the activity of the sulfated salt family of catalysts is based on the presence

123

730

of sulfonic acid sites, which can be considered as the heterogeneous counterpart of sulfuric acid. Also, the acid strength of the catalyst still has an important role in the transesterification reaction. Some factors, such as the preparation technology and suitable selection of support, greatly influenced the acid site distribution on the catalyst. For instance, Jothiramalingam and Wang [21] reported that catalysts prepared from a stronger acid precursor containing benzene sulfonic acid groups had a higher acid strength and were more active than those containing only propylsulfonic acid groups. Chen et al. [110] presented evidence that good catalytic performance of the sulfated silica-zirconia material was attributed to an improved preparation process which resulted in a higher dispersion of zirconia, thus creating a higher acid site density. The carriers for sulfonic acid include not only some inorganic metal oxides (zirconia oxide [21, 110], tin oxide [111], stannia [112]), but also some mesostructured silica [113] and carbon materials such as multi-wall carbon nanotubes [114, 115] and asphalt [115]. In their work, Jitputti et al. [112] evaluated the activities of sulfated zirconia and stannia for crude palm kernel oil and coconut oil conversion to biodiesel. They showed minor activity at 200 °C, which subsequently decreased with additional recycling. They associated the decrease in activity to both sulfate leaching and active site poisoning. In other work, Melero et al. [113] prepared propylsulfonic acid SBA-15 material and found it highly active for the conversion of refined and crude palm oil and soybean oil. The catalytic performance was attributed to the large surface area and pore diameter of the mesoporous support. However, they also found a slight decrease of activity in recycled catalyst testing. To remedy this, they pointed out that further work would be performed to enhance the strength of acid sites and control the surface properties of the silica support in order to enhance the durability of these sulfonated mesostructure catalysts. Shu et al. [115] prepared sulfonation of carbonized vegetable oil asphalt and sulfonated multi-walled carbon nanotubes (s-MWCNTs). They found that the asphalt-based catalyst showed higher activity than the s-MWCNTs for the production of biodiesel and that this behavior might be correlated to the high acid site density of asphalt catalysts resulting from its loose irregular network and large pores. Using the asphalt based catalyst, the conversion of cottonseed oil achieved 89.93% when the methanol/cottonseed oil molar ratio was 18.2, reaction temperature 260 °C, reaction time 3.0 h, and a catalyst/cottonseed oil mass ratio of 0.2%. Also, the asphalt based catalyst can be re-used. Shu et al. [114, 115] thought that the sulfonated polycyclic aromatic hydrocarbons provided an electron-withdrawing function to keep the acid sites stable.

123

Top Catal (2010) 53:721–736

3.2.2 Heteropolyacid Catalysts Sunita et al. [116] compared the activities of zirconia supported isopoly and heteropolytungstate catalysts. Zirconia-supported heteropolytungstate possessed a high total acidity and showed superior catalytic performance compared to zirconia-supported heteropolytungstate catalysts. Under their reaction conditions of 200 °C, methanol/oil molar ratio 15, and 15% WO3/ZrO2 calcined at 750 °C the ZrO2 catalyst achieved 97% conversion of oil. Kulkarni et al. [117] impregnated tungstophosphoric acid on four different supports such as hydrous zirconia, silica, alumina and activated carbon, and used them for converting low quality canola oil containing about 20 wt% FFA to biodiesel. The hydrous zirconia supported tungstophosphoric acid was found to be the most active. At 200 °C, 1:9 oil to alcohol molar ratio, and 3 wt% catalyst loading a maximum ester yield of 90 wt% was observed. 3.2.3 Other Catalysts Except for the above two types of catalysts, other weaker solid acid catalysts were also tested for activity in biodiesel formation reactions. These included many types of zeolite and phosphate based catalyst systems. For instance, Brito et al. [118] studied the activity of several commercialized Y-type zeolites in a continuous tubular reactor at atmospheric pressure and within a temperature range of 200– 476 °C. The results showed that higher temperature accelerated transesterification reaction rate. With used fry oil, they found that the optimal reaction conditions for transesterification with methanol could be achieved with zeolite Y530 at 466 °C, 12.35 min residence time, and a methanol/oil molar ratio of only 6. In other work, some phosphate salts have also been reported active for biodiesel formation. One such example is by Li and Xie [119] who prepared Fe3?-vanadyl phosphate. When the transesterification reaction was performed at a molar ratio of methanol to oil of 30:1, reaction temperature of 473 K, reaction time of 3 h, and a catalyst loading of 5 wt%, the maximum conversion of soybean oil was found to be 61.3%. The importance of this work is that it showed the activity of this catalyst was not significantly affected by the presence of free fatty acids and water in the reactant mixture. In addition, the catalyst also exhibited catalytic activity for the esterification of free fatty acids with methanol. Unfortunately, the same catalyst slowly deactivated after 5 recycles. Serio et al. [120] suggested that the deactivation of vanadyl phosphate catalyst was strongly affected by reaction temperature. In effect, higher reaction temperatures accelerated the deactivation process. However, catalyst leaching at higher temperatures was not the root cause of deactivation. Instead, surface characterization work showed

Top Catal (2010) 53:721–736

that deactivation was primarily due to the progressive reduction of surface vanadium from V5? to V3? by methanol where V5? and intermediate V4? species were active and V3? was inactive. Further work revealed that the deactivation was reversible and catalyst activity could be restored by simple oxidation. Yet another catalyst technology based on Fe–Zn doublemetal cyanide complexes was tested by Sreeprasanth et al. [121]. The catalysts were hydrophobic, and contained only Lewis acidic sites. Bronsted acid sites as well as basic sites were absent. These catalysts proved active for both the transesterification and esterification of unrefined and waste cooking oils. Thus Lewis acidic sites were found to be active centers for both the transesterification and esterification reactions and the surface hydrophobicity of these catalysts improved their tolerance to water. 3.3 Amphoteric Metal Oxide Some amphoteric metal oxides, such as PbO, PbO2, and ZnO, have attracted attention from researchers because of their adjustable basic and acid properties. Singh and Fernando [122] found that the FAME yield reached 89% using amphoteric PbO and PbO2. However, further testing showed Pb content in the glycerol and biodiesel products was as high as 2000 ppm which implied some dissolution of the catalyst. With respect to ZnO based catalyst systems, two candidates appear to offer the best next generation catalyst solution for improved biodiesel synthesis. One is a catalyst composed of zinc aluminum oxides from Institut Francais Du Petrole (Vernaison, France); the other is a catalyst of zinc lanthanum oxides from Wayne State University (MI, USA). 3.3.1 Zinc–Aluminum Catalyst for the Esterfif-HTM Process The Esterfif-HTM process was developed by the French Institute of Petroleum (IFP) and commercialized by Axens. It was first used in an industrial context in 2006, by Sofiproteol in Se`te [123]. This process uses a heterogeneous catalyst, a spinel mixed oxide of zinc and alumina metals. The use of heterogeneous catalysts eliminates the need for catalyst recovery and washing steps—and associated waste streams—required by processes using homogeneous catalysts such as sodium hydroxide or sodium methylate. The process chart is shown in literature [22]. The catalyst section includes two fixed bed reactors (CSTR), fed with vegetable oil and methanol at a given ratio. Excess methanol is removed after each reactor by partial evaporation. Then, esters and glycerol are separated in a settler. The remaining glycerol is collected and the residual methanol removed by evaporation.

731

A patent [124] awarded to IFP described an acid catalyst with a formula of ZnAl2O4, xZnO, yAl2O3 (with x and y being in the range of 0–2) which originated from a hydrotalcite precursor. The BET surface area is between 50 and 200 m2/g and a pore volume is greater than 0.3 cm3/g. In the continuous process, the reaction temperature is required to be between 210 and 250 °C, pressure between 30 and 50 bar, and VVH (volume of injected oil/volume of catalyst/hour) from 0.3 to 3. Using these conditions, they showed a FAME yield of 91% based on a catalyst with a surface area of 65 m2/g, pore volume of 0.63 cm3/g under the conditions of 240 °C, 50 bar, 160 min of contact time. This patent also addressed the catalyst lifetime. They found that on the 14th day, the FAME yield decreases to 30.9% at 240 °C, 50 bar, 1 VVH. The structure of the deactivated catalyst is not reported in this patent. A subsequent IFP patent [125] states that this catalyst is quite sensitive to water. In fact they maintained the water concentration below 1000 ppm, which implies that the oil feedstocks used in Esterfif-HTM process must be well refined. 3.3.2 Zinc Lanthanum Catalysts A series of zinc and lanthanum containing catalysts were developed by Yan et al. [68, 126, 127]. This type of catalyst exhibits weak basic properties and is composed of ZnO, La2CO3 and LaOOH. Chief among its attributes are activity, longevity, FFA and water tolerance, and oil flexibility. The ZnLa catalysts demonstrate high catalytic activity. When using refined soybean oil, FAME yields as high as 95% under reaction conditions of 60 min, 200 °C, 500 psi, 36:1 M ratio of methanol to oil, and 2.3 wt% catalyst dosage in a stirred batch reactor. Yan et al. [127] reported the Zn3La1 catalyst which had a 3:1 M ratio of zinc to lanthanum was recycled 17 times in a batch reactor without loss of activity and maintained a high FAME yield (*92.3%) for 70 days in a continuous tubular flow reactor (Fig. 3a, b). Additional testing showed the catalyst still active after more than 100 days of use [128]. At this point, there is no report of any other heterogeneous catalyst with a longer catalyst life than this Zn3La1 catalyst for biodiesel production. As part of the above work, food-grade soybean oils combined with 5.20%, 10.13%, 15.21% and 30.56% of oleic acid and 1.03%, 3.12% and 5.07% water were tested. The results (Tables 2 and 3) show that all the oils were converted within 150 min to FAME (*96%) even with FFA content as high as 30.56% or water content as high as 5.07%. This suggests that by properly controlling the temperature of the reaction, hydrolysis reactions in the presence of water can be minimized, which will allow for higher FAME yields from higher FFA and water content

123

732

a

Top Catal (2010) 53:721–736 Table 3 Yield of FAME in the presence of different water addition

100

Water addition (%)

Yield of FAME at different reaction time (%) 20 min

60 min

90 min

150 min

70

0

52

88

93

94

60

1.03

50

80

90

96

3.12

22

74

89

93

5.07

13

57

78

90

90

Yield of FAME %

80

50 40

Reaction conditions: catalyst amount of Zn3La1 is 2.3 wt%, molar ratio of methanol to oil is 36:1, reaction temperature is 200 °C. All data are taken from literature [68]

30 20 10 0

100 0

2

4

6

8

10

12

14

16

18

b

FAME content %

Recycle times 100

Yield of FAME (%)

80

80

60

40

20 60

0 0

20

40

60

80

100

120

140

160

180

200

Reaction time min

40

Crude coconut oil Waste cooking oil Crude soybean oil Crude palm oil Crude algae oil Crude corn oil from DDGs Food grade soybean oil NaOH H2SO4

20

Food grade soybean oil with 3 % water and 5 % FFA addition

0 0

10

20

30

40

50

60

70

80

Time (day)

Fig. 3 FAME yield (a) in the batch reactor using the recycled Zn3La1 catalyst. Note the catalyst was reused for 17 times. The average yield of soybean methyl esters is 93.7% (b) in the continuous reactor. Note that this catalyst has run for 70 days, and the average yield of FAME during stable stage is 92.3% [127]

Table 2 Yield of FAME in the presence of different FFA addition FFA addition (%)

Yield of FAME at different reaction time (%) 5 min

10 min

20 min

40 min

0

10

16

52

83

88

93

5.20

38

72

83



95

92

10.13

73

83

88

92



92

15.21

89



95



98

92

30.56

75

85

93

95



01

60 min

90 min

Reaction conditions: catalyst amount of Zn3La1 is 2.3 wt%, molar ratio of methanol to oil is 36:1, reaction temperature is 200 °C. All data are taken from literature [68]

123

Fig. 4 FAME yield of crude corn oil from DDGs, crude algae oil, crude coconut oil, crude palm oil, crude soybean oil, waste cooking oil, food grade soybean oil and food grade soybean oil with 3% water and 5% oleic acid addition. Reaction conditions: 126 g of oil, 180 g of methanol, 3 g of catalyst, 200 °C, 500 psi, in the batch stir reactor. Note that all of these oils was converted into FAME within 3 h [127]

feedstocks. Furthermore, this catalyst was found to be relatively insensitive to species in air such as CO2, moisture and O2 in air, which poison base catalysts. Zn3La1 was used to process multiple unrefined and waste oils, i.e. crude corn oil from DDGs, crude algae oil, crude coconut oil, crude palm oil, crude soybean oil, waste cooking oil, food grade soybean oil with 3% water and 5% FFA addition. Figure 4 shows that all the oils were completely converted to FAME within 3 h in a batch reactor. This is an impressive result considering the fatty acid composition and total acid number (TAN) of these oils (Table 4). Here, it should be noted that crude corn oil from DDGs contains 93% triglycerides and has a TAN as high as 25.19 mg KOH/g. Similarly, crude algae oil contains only 80% triglycerides, and has a TAN of 26.38 mg KOH/g, while crude coconut oil has a TAN of 8.48 mg KOH/g. After reaction, the TAN of all the oils is significantly

Top Catal (2010) 53:721–736

733

Table 4 Fatty acid composition and TAN of food grade soybean oil, crude soybean oil, crude palm oil, waste cooking oil, crude corn oil from DDGs, crude algae oil and crude coconut oil [127] Crude soybean oil (%)

Crude palm oil (%)

Waste cooking oil (%)

Crude corn oil from DDGs (%)

Crude algae oil (%)

Crude coconut oil (%)

C 12:0

0

0

0

0

0

0

49.13

C 14: 0

0

0.27

0.21

0

0

2.72

19.63

13.05 0.39

41.92 0.23

20.91 10.62

10.12 1.79

C 16: 0 C 16: 1

11.07 0.09

11.58 0.18

11.50 0

C 18: 0

3.62

4.17

3.85

4.26

4.77

6.95

2.83

C 18: 1

20.26

22.75

42.44

24.84

26.26

33.33

7.59

C 18: 2

57.60

52.78

11.30

53.55

56.20

18.45

2.75

C 18: 3

7.36

6.59

0.04

5.60

1.27

1.16

0.15

Others

0

0

0

0

0

6.86

6.01

TAN

0.03

6.62

0.48

7.56

26.38

8.48

a

100

600

Yield of FAME %

80

500 400

Yield of FAME La content in FAME phase Zn content in FAME phase

60

300

40

200 100

20

0 0 0

10

20

30

40

50

60

70

80

Time (day)

b

100

1800

80

1400

1600

Yield of FAME %

reduced. For instance, the TAN of algae oil after reaction is 0.94 mg KOH/g and that of corn oil from DDGs is 1.32 mg KOH/g. This implies that during the reaction process, esterification of FFA with methanol is simultaneously performed with the transesterification of triglycerides with methanol. Where traditional homogeneous catalysts would have been deactivated using these high TAN/FFA feedstocks, the Zn3La1 shows a remarkably high activity for biodiesel formation reactions with a variety of feedstocks. Unlike previously reported solid catalysts, the Zn3La1 catalyst is very stable. Yan et al. pointed out that the Zn and La contents in the FAME product are only 6 and 2 ppm. The Zn and La contents in the glycerine phase were measured at only 8 and 4 ppm after a short induction period when the yield of FAME stabilized (Fig. 5a, b). The low level of Zn and La in FAME and glycerin products suggests that Zn3La1 is a true heterogeneous catalyst with a very stable crystal structure that does not deactivate under reaction conditions. Figure 6 illustrates the reaction pathway for biodiesel formation. Where other catalysts have failed, this catalyst can successful handle all the components of a crude feedstocks. In particular, there are three major components to these inexpensive oils: triglyceride, FFA, and water. Thus, there are four major reactions: transesterification of triglyceride with methanol, which results in the formation of FAME; esterification of FFA with methanol, which results in FAME; hydrolysis of FAME, which consumes FAME; and hydrolysis of triglycerides, which results in FFA. The transesterification and esterification reactions will lead to higher yields of FAME. However, hydrolysis reactions will lead to lower FAME yields. With appropriate control of the reaction temperature, Yan et al. [68] was able to maximize the transesterification and esterification reactions while de emphasizing the oil and biodiesel hydrolysis reactions.

25.19

Zn and La content (ppm)

Food grade soybean oil (%)

1200

Yield of FAME La content in glycerine phase Zn content in glycerine phase

60

1000 800 600

40

400 200

20

Zn and La content (ppm)

Fatty acid components

0 -200

0 0

10

20

30

40

50

60

70

80

Time (day)

Fig. 5 Zn and La contents in a FAME product and b glycerine product. Note that after 3 days metal contents in FAME product reached a low level; after 7 days metal contents in glycerine product reached a low level [127]

4 Summary and Future Opportunity The use of heterogeneous catalysts for biodiesel production is an emerging research field which has been quickly growing over the last 10 years. Most of the reported

123

734

Top Catal (2010) 53:721–736

Fig. 6 Reactions pathways of transesterification, esterification, and hydrolysis of unrefined and waste oils [68]

Unrefined or Waste Oils

Triglyceride

Transesterification

Water

Hydrolysis

FFA

Esterification

Hydrolysis

Fatty Acid Methyl Esters

catalysts can be divided into two types, solid base and solid acid catalysts, according to their active center. Both Lewis acid–base sites and Bronsted acid–base sites have the ability to catalyze the oil transesterification reaction. Therefore, catalyst activity is closely related to the acid/ base strength. Other texture properties of the catalyst also impact the catalyst’s activity, such as specific surface area, pore size, pore volume and active site concentration. Modification of reaction conditions, such as increasing reaction temperature (130–250 °C), pressure (100– 1000 psi), catalyst quantity (3–10 wt%), and methanol/oil molar ratio (10:1–42:1) is effective for obtaining high FAME yields. Reported catalysts, operating at high temperature, exhibit a low base or acid strength which many research groups have demonstrated is the basis for improved activity, improved tolerance to FFA and water, and extended catalyst lifetimes. Within these high temperature catalysts, two catalysts stand out as leading technologies: one is the ZnO–Al2O3 catalyst from IFP and the other is the ZnO–La2O3 catalyst developed at Wayne State University. Additional work is still required to find ways to rejuvenate or re-active catalysts that have failed because of poisoning, leaching, or loss of surface area. Finally, the most recognized drawbacks of heterogeneous catalysts are their slow reaction rates in comparison to homogeneous catalysts. Perhaps, by enhancing the number and type of active sites and intensifying reaction conditions, to minimize mass transfer limitations, these catalysts may be able to overcome this limitation as well.

References 1. Clark SJ, Wagner L, Schrock MD (1984) J Am Chem Soc 61:1632 2. Ma F, Hanna MA (1999) Bioresour Technol 70:1

123

3. Muniyappa PR, Brammer SC, Noureddini H (1996) Bioresour Technol 6:19 4. Dorado MP, Ballesteros E, Arnal JM, Go´mez J, Lo´pez FJ (2003) Fuel 82:1311 5. Freedman B, Pryde EH, Mounts TL (1984) J Am Oil Chem Soc 61:1638 6. Gryglewicz S (1999) Bioresour Technol 70:249 7. Zaher FA (2003) Energy Sour 25:819 8. Labeckas G, Slavinskas S (2006) Renew Energy 31:849 9. Bouaid A, Diaz Y, Martinez M, Aracil J (2005) Catal Today 106:193 10. Zhang Y, Dube MA, Mclean DD, Kates M (2003) Bioresour Technol 90:229 11. Goodrum JW, Geller DP, Adams TT (2002) J Am Oil Chem Soc 79:961 12. Lee KT, Foglia TA, Chang KS (2002) J Am Oil Chem Soc 79:191 13. Encinar JM, Gonzalez JF, Rodriguez JJ, Tejedor A (2002) Energy Fuels 16:443 14. Warabi Y, Kusdiana D, Saka S (2004) Bioresour Technol 91:283 15. Kim HJ, Kang BS, Park M (2004) J Catal Today 93:283 16. Al-Widyan MK, Al-Shyoukh AO (2005) Bioresour Technol 85:253 17. Yan S (2007) School of Chemical Engineering Sichuan University Chengdu 51 18. Schuchardta U, Serchelia R, Vargas RM (1998) J Braz Chem Soc 13:199 19. Lee DW, Park YM, Lee KY (2009) Catal Surv Asia 13:63 20. Stoffel W, Chu F, Ahrens EH (1959) Anal Chem 31:307 21. Jothiramalingam R, Wang MK (2009) Ind Eng Chem Res 48:6162 22. Bournay L, Casanave D, Delfort B, Hillion G, Chodorge JA (2005) Catal Today 106:190 23. Ma F, Clements LD, Hanna MA (1998) Trans ASAE 41:1261 24. Canakci M, Gerpen JV (1999) Trans ASAE 42:1203 25. Johnston M, Holloway T (2007) Environ Sci Technol 41:7967 26. Freedman B, Butterfield RO, Pryde EH (1986) J Am Oil Chem Soc 63:1375 27. Schwab AW, Bagby MO, Freedman B (1987) Fuel 66:1372 28. Aksoy HA, Becerik I, Karaosmanogly F, Yamaz HC, Civelekoglu H (1990) Fuel 69:600 29. Kim HJ, Kang BS, Kim MJ, Park YM, Kim DK, Lee JS, Lee KY (2004) Catal Today 93:315

Top Catal (2010) 53:721–736 30. Ma H, Li S, Wang B, Wang R, Tian S (2008) J Am Oil Chem Soc 85:263 31. Ilgen O, Akin AN (2009) Energy Fuels 23:1786 32. Lukic´ I, Krstic´ J, Jovanovic´ D, Skala D (2009) Bioresour Technol 100:4690 33. Xie W, Li H (2006) J Mol Catal A 255:1 34. Samart C, Sreetongkittikul P, Sookman C (2009) Fuel Process Technol 90:922 35. Xie W, Peng H, Chena L (2006) Appl Catal A 300:67 36. Vyas AP, Subrahmanyam N, Patel PA (2009) Fuel 88:625 37. Xie W, Huang X (2006) Catal Lett 107:53 38. Hameed BH, Lai LF, Chin LH (2009) Fuel Process Technol 90:606 39. Cui L, Xiao G, Xu B, Teng G (2007) Energy Fuels 21:3740 40. Boz N, Degirmenbasi N, Kalyon DM (2009) Appl Catal B Environ 89:590 41. Xie W, Yang Z, Chun H (2007) Ind Eng Chem Res 46:7942 42. Alonsoa DM, Mariscal R, Granados ML, Maireles-Torres P (2009) Catal Today 143:167 43. Lo´pez DE, Goodwin JG Jr, Bruce DA, Lotero E (2005) Appl Catal A Gen 295:97 44. Hattori H (2004) J Jpn Petro Inst 47:67 45. Wang Y, Hu SY, Guan YP, Wen LB, Han HY (2009) Catal Lett 131:574 46. Arzamendi G, Arguin˜arena E, Campo I, Zabala S, Gandı´a LM (2008) Catal Today 133:305 47. Noiroj K, Intarapong P, Luengnaruemitchai A, Jai-In S (2009) Renew Energy 34:1145 ´ (2008) 48. Ramos MJ, Casas A, Rodrı´guez L, Romero R, Pe´rez A Appl Catal A 346:79 49. Baba T, Endou T, Handa H, Ono Y (1993) Appl Catal A 97:L19 50. Kabashima H, Tsuji H, Hattori H (1996) React Kinet Catal Lett 58:255 51. Fujita S-i, Bhanage BM, Kanamaru H, Arai M (2005) J Mol Catal A 230:43 52. Bancquart S, Vanhove C, Pouilloux Y, Barrault J (2001) Appl Catal A 218:1 53. Tsuji H, Yagi F, Hattori H, Kita H (1994) J Catal 148:759 54. Seki T, Kabashima H, Akutsu K (2001) J Catal 204:393 55. Cantrell DG, Gillie LJ, Lee AF, Wilson K (2005) Appl Catal A 287:183 56. Peterson GR, Scarrah WP (1984) J Am Oil Chem Soc 10:1593 57. Kabashima H, Hattori H (1999) Catal Today 44:277 58. Kabashima H, Tsuji H, Hattori H (1997) Appl Catal A 165:319 59. Lin Y, Xiexian G, Yide X (1997) Appl Catal A 164:47 60. Gayko G, Wolf D, Kondratenko J (1998) J Catal 178:441 61. Reddy CRV, Oshel R, Verkade JG (2006) Energy Fuels 30:1310 62. Gryglewicz S (2000) Appl Catal A 192:279 63. Martyanov IN, Sayari A (2008) Appl Catal A 339:45 64. Rubio-Caballero UM, Santamarı´a-Gonza´lez J, Me´rida-Robles J, Moreno-Tost R, Jime´nez-Lo´pez A, Maireles-Torres P (2009) Appl Catal B 91:339 65. Yan S, Lu H, Liang B (2008) Energy Fuels 22:646 66. Albuquerque MCG, Santamarı´a-Gonza´lez J, Me´rida-Robles JM, Moreno-Tost R, Rodrı´guez-Castello´n E, Jime´nez-Lo´pez A, Azevedo DCS, Maireles-Torres CLCP Jr (2008) Appl Catal A 347:162 67. Albuquerque MCG, Jime´nez-Urbistondo I, Santamarı´a-Gonza´lez J, Me´rida-Robles JM, Moreno-Tost R, Rodrı´guez-Castello´n E, Jime´nez-Lo´pez A, Azevedo DCS, Torres CLC, Jr PM (2008) Appl Catal A 334:35 68. Yan S, Salley SO, Ng KYS (2009) Appl Catal A 353:203 69. Busca G (2009) Ind Eng Chem Res 48:6486 70. Yan S, Kim M, Salley SO, Ng KYS (2009) Appl Catal A 360:163

735 71. Kouzu M, Kasuno T, Tajika M, Sugimoto Y, Yamanaka S, Hidaka J (2008) Fuel 87:2798 72. Kouzu M, Yamanaka SY, Hidaka JS, Tsunomori M (2009) Appl Catal A 355:94 73. Granados ML, Poves MDZ, Alonso DM, Mariscal R, Galisteo FC, Moreno-Tost R, Santamarı´a J, Fierro LG (2007) Appl Catal B 73:317 74. Granados ML, Alonso DM, Sa´daba I, Mariscal R, Oco´n P (2009) Appl Catal B 89:265 75. Xie W, Peng H, Chen L (2006) J Mol Catal A 246:24 76. Li E, Xu ZP, Rudolph V (2009) Appl Catal A 88:42 77. Fraile JM, Garcı´a Nuria, Mayoral JA, Pires E, Rolda´n L (2009) Appl Catal A 364:87 78. Schuchardt U, Vargas RM, Gelbard G (1995) J Mol Catal A 99:65 79. Schuchardt U, Vargas RM, Gelbard G (1996) J Mol Catal A 109:37 80. Shibasaki-Kitakawa N, Honda H, Kuribayashi H, Toda T, Fukumura T, Yonemoto T (2007) Bioresour Technol 98:416 81. Vicente G, Coteron A, Martinez M, Aracil J (1988) Ind Corps Prod 8:29 82. Kim M, Salley S, Ng KYS (2008) Energy Fuels 22:3594 83. Xie J, Zheng X, Dong A, Xiao Z, Zhang J (2008) Green Chem 11:355 84. Yang L, Zhang A, Zheng X (2009) Energy Fuels 23:3859 85. Nakatani N, Takamori H, Takeda K, Sakugawa H (2009) Bioresour Technol 100:1510 86. Wei Z, Xu C, Li B (2009) Bioresour Technol 100:2883 87. Kiss AA, Dimian AC, Rothenberg G (2006) Adv Synth Catal 348:75 88. Gore RB, Thomson WJ (1998) Appl Catal A 168:23 89. Suwannakarna K, Loteroa E, Lu JGGC Jr (2008) J Catal 255:279 90. Du Y, Liu S, Ji Y, Zhang Y, Wei S, Liu F, Xiao FS (2008) Catal Lett 124:133 ´ lvarez M, Aguilar DH, Quintan P (2008) J 91. Garcı´a J, Lo´pez T, A Non-Cryst Solids 354:729 92. Yadav GD, Nair JJ (1999) Microporous Mesoporous Mater 33:1 93. Yadav GD, Murkute AD (2004) J Catal 224:218 94. Alsalme A, Kozhevnikova EF, Kozhevnikov IV (2008) Appl Catal A 349:170 95. Pesaresi L, Brown DR, Lee AF, Montero JM, Williams H, Wilson K (2009) Appl Catal A 360:50 96. Narasimharao K, Brown DR, Lee AF, Newman AD, Siril PF, Tavener SJ, Wilson K (2007) J Catal 248:226 97. Katada N, Hatanaka T, Ota M, Yamada K, Okumura K, Niwa M (2009) Appl Catal A 362:164 98. Rezende SMD, Castro MD, Garcia M, Coutinho FMB, Silva PL Jr, da Silva San Gil RA, Lachter ER (2008) Appl Catal A 349:198 99. Lotero E, Liu Y, Lopez DE, Suwannakarn K, Bruce DA, Goodwin JG Jr (2005) Ind Eng Chem 44:5353 100. Lo´pez DE, Goodwin JG Jr, Bruce DA, Furuta S (2008) Appl Catal A 339:76 101. Soldi RA, Oliveira ARS, Ramos LP, Ce´sar-Oliveira MAF (2009) Appl Catal A: Gen 361:42 102. Hara M, Yoshida T, Takagaki A, Takata T, Kondo JN, Hayashi S, Domen K (2004) Angew Chem Int 43:2955 103. Toda M, Takagaki A, Okamura M, Kondo JN, Hayashi S, Domen K, Hara M (2005) Nature 438:178 104. Takagaki A, Toda M, Okamura M, Kondo JN, Hayashi S, Domen K, Hara M (2006) Catal Today 116:157 105. Zong MH, Duan ZQ, Lou WY, Smith TJ, Wu H (2007) Green Chem 9:434 106. Lou W-Y, Zong M-H, Duan Z-Q (2008) Bioresour Technol 99:8752

123

736 107. Demirbas A (2007) Energy Conv Manag 48:937 108. Suppes GJ, Bockwinkel K, Lucas S, Botts JB, Mason MH, Heppert JA (2001) J Am Oil Chem Sci 78:139 109. Barakos N, Pasias S, Papayannakos N (2008) Bioresour Technol 99:5037 110. Chen XR, Ju YH, Mou CY (2007) J Phys Chem 111:18731 111. Furuta S, Matsuhashi H, Arata K (2004) Catal Commun 5:721 112. Jitputti J, Kitiyanan B, Rangsunvigit P, Bunyakiat K, Attanatho L, Jenvanitpanjakul P (2006) Chem Eng J 116:61 113. Melero JA, Bautista LF, Morales G, Iglesias J, Briones D (2009) Energy Fuels 23:539 114. Shu Q, Zhang Q, Xu G, Wang J (2009) Food Bioprod Process 87:164 115. Shu Q, Zhang Q, Xu G, Nawaz Z, Wang D, Wang J (2009) Fuel Process Technol 90:1002 116. Sunita G, Devassy BM, Vinu A, Sawant DP, Balasubramanian VV, Halligudi SB (2008) Catal Commun 9:696 117. Kulkarni MG, Gopinath R, Meher LC, Dalai AK (2006) Green Chem 8:1056

123

Top Catal (2010) 53:721–736 118. Brito A, Borges ME, Otero N (2007) Energy Fuels 21:3280 119. Li H, Xie W (2008) J Am Chem Soc 85:655 120. Serio MD, Cozzolino M, Tesser R, Patrono P, Pinzari F, Bonelli B, Santacesaria E (2007) Appl Catal A 320:1 121. Sreeprasanth PS, Srivastava R, Srinivas D, Ratnasamy P (2006) Appl Catal A 314:148 122. Singh AK, Fernando SD (2008) Energy Fuels 22:2067 123. http://www.ifp.com/axes-de-recherche/carburants-diversifies/ biocarburants-de-1ere-generation 124. Stern R, Hillion G, Rouxel J-J, Leporq S (1999) US Patent 5908946 125. Bournay L, Hillion G, Boucot P, Chodorge J-A, Bronner C, Forestiere A (2005) US Patent 6878837 126. Yan S, Salley SO, Ng KYS (2009) US Patent 20100010246 127. Yan S, Kim M, Mohan S, DiMaggio C, Salley SO, Ng KYS (2009) Submitted 128. Yan S, Salley SO, Ng KYS (2009) US Patent 20090307966

Suggest Documents