Metal-Nitrate-Catalyzed Levulinic Acid Esterification with Alkyl Alcohols: A Simple Route to Produce Bioadditives

Abstract

This work developed an efficient route to produce fuel bioadditive alkyl levulinates. Special attention was paid to butyl levulinate, which is a bioadditive with an adequate carbon chain size to be blended with liquid fuels such as diesel or gasoline. In this process, levulinic acid was esterified with butyl alcohol using cheap and commercially affordable metal nitrates as catalysts, producing bioadditives at more competitive costs. Iron (III) nitrate was the most active and selective catalyst toward butyl levulinate among the salts evaluated. In solvent-free conditions, with a low molar ratio and catalyst load (1:6 acid to alcohol, 3 mol% of Fe (NO₃)₃), conversion and selectivity greater than 90% after an 8 h reaction was achieved. A comparison of the iron (III) nitrate with other metal salts demonstrated that its superior performance can be assigned to the highest Lewis acidity of Fe³⁺ cations. Measurements of pH allow the conclusion that a cation with high Lewis acidity led to a greater H⁺ release, which results in a higher conversion. Butyl levulinate and pseudobuty levulinate were always the primary and secondary products, respectively. The consecutive character of reactions between butyl alcohol and levulinic acid (formation of the pseudobutyl levulinate and its conversion to butyl levulinate) was verified by assessing the reactions at different temperatures and conversion rates. A variation in Fe(NO₃)₃ catalyst load impacted the conversion much more than reaction selectivity. The same effect was verified when the reactions were carried out at different temperatures. The reactivity of alcohols with different structures depended more on steric hindrance on the hydroxyl group than the size of the carbon chain. A positive aspect of this work is the use of a commercial iron nitrate salt as the catalyst, which has advantages over traditional mineral acids such as sulfuric and hydrochloric acids. This solid catalyst is not corrosive and avoids neutralization steps after reactions, minimizing the generation of residues and effluents.

1. Introduction

Biomass is an inexhaustible source of compounds that are potential candidates to replace fossil-origin fuels or chemicals [1,2]. Oleaginous biomass has been utilized for the production of biodiesel, a sustainable fuel that can be utilized in diesel engines without costly modifications, resulting in low emissions of particulate matter and greenhouse gases. Vegetal oil or animal fats contain triglycerides or free fatty acids (FFA), which can be converted to methyl or ethyl fatty esters with similar properties to diesel hydrocarbons [3]. Solid catalysts are preferentially used in these reactions due to environmental reasons [4].

Nowadays, our planet is impacted by the greenhouse effect triggered by the burning of fossil fuels; therefore, biomass conversion to biofuels has assumed key importance, including the production of renewable fuels as the priority of several countries. In this sense, reactions of esterification, etherification, or the condensation of biomass derivatives produce esters, ethers, ketals, or acetals, which can also be blended with liquid fuels, reducing dependence on petrol fuels [5,6]. Particularly, levulinic acid (LA) has been produced from acidic hydrolyses of carbohydrates, resulting in the formation of lactones or levulinate esters that, in addition to being used as green solvents, are also potential fuel bioadditives [7,8]. Levulinic acid is one of the top twelve building blocks, which can be converted either to chemicals or biofuels [9].

Ethyl levulinates derived from cellulose can be blended into diesel or biodiesel, enhancing chemical and physical properties, such as spray evaporation, cleanliness, oxidative stability, lubricity, and cold flow, while reducing corrosiveness, sulfur, hydrocarbon, and carbon monoxide emissions [10,11,12]. Among the various derivatives of levulinic acid, butyl levulinate is particularly attractive, because it falls within the range of carbon chain sizes found in gasoline hydrocarbons [13].

There are various ways to produce levulinate alkyl esters synthetically. They can be obtained directly from biomass by converting it in an alcoholic medium to esters through multistep processes. These processes involve pretreating the raw material, hydrolyzing saccharides, and separating/purifying the esters [14,15]. However, controlling the selectivity in this process remains a challenge due to the variety of products that can be obtained. Furthermore, the costs of levulinate esters are increased by the separation and purification steps. Another approach involves converting furfural or furfuryl alcohol to alkyl levulinates through hydrogenation/hydrolysis reactions [16].

The most selective route is to obtain pure levulinic acid and esterify it with alkyl alcohols. Despite the costs of LA, this process selectively produces esters, saving solvent and isolation/purification time/multisteps. Traditionally, esterification reactions have been carried out in the presence of liquid Brønsted acids, such as sulfuric or hydrochloric. However, although inexpensive, these catalysts are highly corrosive and must be neutralized at the end of the reaction, leading to the generation of large amounts of residue and effluents that need to be disposed of into the environment [17]. In addition, the recovery and reuse of these liquid catalysts are feasible on an industrial scale.

Heterogeneous catalysts such as acid resins, zeolites, metal oxides, carbonaceous materials, and mesoporous molecular sieves can be used as an alternative to traditional liquid acids in esterification reactions of levulinic acid [18,19,20,21]. Levulinic acid esterification can also be conducted using solid-supported catalysts, which usually consist of a high surface area solid support doped with Lewis or Brønsted acid [22,23]. The use of heterogeneous catalysts has advantages such as the possibility of reuse and the lower generation of effluents and residues of neutralization. These catalysts have been used in alkyl levulinate production reactions directly from carbohydrates; however, their use in such reactions requires high temperatures, which favors the formation of humins and side products.

A scarcely explored option is using nitrate metal salts as catalysts in acid-catalyzed reactions [24,25,26]. Although metal salts have been used in different reactions, such as in the esterification of glycerol, terpene alcohols, and condensation of aldehydes with ketones, the use of nitrate metal salts in levulinic acid esterification remains little explored [27,28,29,30,31]. Notwithstanding, other metal salts such as sulfate, chloride, and metal tungstates have been successfully used in LA esterification reactions [32,33,34].

In this study, we utilized readily available and cost-effective metal nitrates as catalysts to transform levulinic acid and butyl alcohol into butyl levulinate, which is a biofuel additive. We investigated various reaction parameters such as the type of metal nitrate, catalyst load, temperature, and time. Additionally, we examined how the strength of the Lewis acidity of cations affected their catalytic performance by comparing salts containing cations with different oxidation numbers and anions. Furthermore, we evaluated the effectiveness of iron (III) nitrate catalyst in esterification reactions of levulinic acid with different short-chain alcohols.

2. Materials and Methods

2.1. Chemicals

All the chemicals were used as received. The metal nitrate catalysts Co(NO₃)₂·6 H₂O, Cu(NO₃)₂·3 H₂O, and NaNO₃ (all > 98 wt.%) were acquired from Dinamica Ltd. (São Paulo, Brazil). Fe(NO₃)₃·9 H₂O, FeCl₃·6 H₂O; FeSO₄·7 H₂O; Fe₂(SO₄)₃·5 H₂O, Zn(NO₃)₂·6 H₂O (all 99 wt.%) were acquired from Sigma-Aldrich (San Luis, CA, USA)). Levulinic acid, methanol, ethanol, propanol, butanol, 2-butanol, 2-propanol, and tert-butanol (99 wt.%) were purchased from Sigma-Aldrich.

2.2. Catalytic Runs

Catalytic tests were carried out in a glass reactor (50 mL) equipped with a sampling septum and reflux condenser in a heated oil batch while being stirred magnetically. Typically, levulinic acid (15.3 mmol) was dissolved in alkyl alcohol (10 mL, 91.8 mmol). The solution was stirred and heated to reaction temperature. Afterward, the metal nitrate catalyst (3.0 mol%) was added to initiate the reaction, which was allowed to proceed for 8 h. The reactions were conducted under these conditions regardless of the type of alcohol used. The catalytic activity of all the salts, as well as the influence of other reaction parameters, was tested using only butyl alcohol. The temperature’s impact was assessed within the range of 298 to 338 K. The effect of Fe(NO₃)₃ catalyst concentration was studied using concentrations ranging from 0.5 to 4.0 mol%.

2.3. Reaction Monitoring and Identification of Main Reaction Products

The reaction progress was monitored using GC analyses of aliquots taken at regular intervals. A Shimadzu GC 2010 (Kyoto, Japan) instrument with FID, fitted with a Carbowax 20M capillary column (30 m length, 0.25 mm i.d., 0.25 mm film thickness) was utilized in this procedure. Toluene was the internal standard. By comparing the GC peak areas of products and substrates with the corresponding calibrating curves, the reaction conversions were determined. Gas chromatographic conditions were as follows: 80 °C (3 min); temperature rate: 10 °C/min; final temperature: 260 °C; injector temperature: 250 °C; detector temperature: 280 °C.

The reaction products were identified in a Shimadzu GC-2010 gas chromatographer coupled with an MS-QP 2010 mass spectrometer (Kyoto, Japan) (electronic impact 70 eV) using He as the gas carrier, scanning range of m/z 50–450, and co-injection of authentic samples. The chromatographic conditions were the same as those used in GC analyses.

3. Results

3.1. Catalytic Tests

3.1.1. Effect of Metal Nitrate Catalyst

The effect of the cation on the activity of nitrate catalysts was first evaluated by testing the performance of salts containing different metals. The reaction conditions were chosen based on previous research, and the kinetic curves are depicted in Figure 1 [30].

Even with an excess of butyl alcohol, the reaction did not proceed well without a catalyst. Conversely, when a metal nitrate catalyst was used, the reaction between butyl alcohol and levulinic acid was much more successful. The nonahydrate iron (III) nitrate proved to be the most effective catalyst among those tested, resulting in a 78% conversion after an 8 h reaction. The efficiency of the catalysts can be ordered as follows: Fe(NO₃)₃ 9 H₂O > Cu(NO₃)₂ 3 H₂O > Co(NO₃)₂ 6 H₂O > Zn(NO₃)₂ 6 H₂O > NaNO₃.

To comprehend the differences in catalytic activity among various metal nitrates in this reaction, it is crucial to analyze how the Lewis acidity of the metal cations affects the key steps of the process. The Lewis acidity of the metal cations determines their interaction with water molecules. Strong Lewis acidity in the cations leads to a strong interaction with water molecules, which act as Lewis bases [26].

The metal cation in nitrate salt can react with water molecules from the salt, the solvent, or those produced in the esterification reaction. This reaction results in the formation of the H₃O⁺ hydronium cation or H⁺ ion (see Equations (1) and (2) in Scheme 1).

M^X+ + n H₂O ⇌ [M(H₂O)n]^X+

(1)

[M(H₂O)n]^X+ + H₂O ⇌ [M(H₂O)_n(OH)]^−1+X + H₃O⁺

(2)

In this scenario, the carbonyl group of levulinic acid can be protonated by H⁺ ions, and its carbonyl carbon becomes more susceptible to attack by the alcohol’s hydroxyl group (see Scheme 1) [28,30].

Afterward, the protonate intermediate undergoes a prototropism step, resulting in the release of water and another protonated intermediate. The latter has a protonated carbonyl group that releases an H⁺ cation, regenerating the catalyst to participate in another catalytic cycle and producing butyl levulinate ester.

Notwithstanding, the M⁺ cation can react with the carbonyl group of levulinic acid. In this case, the dissolution of metal nitrate facilitated by the water and/or butyl alcohol leads to the formation of M^x+ cations (Equation (3), Scheme 2) [28]. Instead of reacting with water, as depicted in Equations (1) and (2) (Scheme 1), these M^x+ cations can coordinate with the oxygen atom of the carbonyl group of levulinic acid, making its carbonylic carbon more electrophilic and increasing its ability to interact with nucleophilic groups, such as the hydroxyl group of alcohol [28,30]. Scheme 2 depicts these and other steps involved in the esterification reaction.

$${\mathrm{M}\left({\mathrm{N}\mathrm{O}}_3\right)}_{\mathrm{X}}\stackrel{\mathrm{b}\mathrm{u}\mathrm{t}\mathrm{y}\mathrm{l} \mathrm{a}\mathrm{l}\mathrm{c}\mathrm{o}\mathrm{h}\mathrm{o}\mathrm{l}}{\to }{\mathrm{M}}^{\mathrm{X}+}+{\mathrm{N}\mathrm{O}}_3^{-}$$

(3)

When a metal cation with a strong affinity for oxygen is present, it can bind to the oxygen atom of the carbonyl group in levulinic acid. This caused the carbon atom bonded to the oxygen to develop a positive charge, making it more electrophilic. As a result, just like what happens when H⁺ ions act as catalysts, the reaction can proceed with a nucleophilic attack of the hydroxyl group of butyl alcohol over the carbonylic carbon, followed by the loss of the M⁺ cation and a prototropism step. This sequence is then followed by the release of a water molecule, resulting in the formation of butyl levulinate ester (Scheme 2).

Based on the findings, it can be concluded that in the M(NO₃)_x-catalyzed esterification of levulinic acid with butyl alcohol, there are two possible reaction pathways.

To determine if metal cations undergo hydrolysis and release H⁺ cations in solution (see Equations (1) and (2), Scheme 1), the pH of metal nitrate alcoholic solutions was measured and compared with the conversion of the reactions (refer to

Table 1. Measurements of pH obtained from metal nitrate solutions in butyl alcohol and conversion achieved in the esterification reactions of levulinic acid.

Run	Catalyst	pH	Conversion/%
1	-	6.0	5.0
2	Fe(NO3)3 9 H2O	0.4	83.0
3	Cu(NO3)2 3 H2O	1.7	40.0
4	Co(NO3)2 6 H2O	2.3	24.0
5	Zn(NO3)2 6 H2O	2.7	26.0
6	NaNO3	3.6	18.0

for details).

Notably, the addition of Fe(NO₃)₃ salt resulted in the most significant decrease in pH in the butyl alcohol solution, indicating the highest production of H⁺ ions. This observation may explain why the Fe(NO₃)₃-catalyzed reaction achieved the highest conversion.

The HASB theory (Hard Acid Soft Base) is useful for explaining the behavior of metal cations. A metal cation with a higher positive charge and a small ionic radium is considered a hard acid. In this case, its interaction with a hard base, such as the oxygen atoms of the water molecules, will be strong. According to the literature, Pearson’s absolute electronegativity is one convincing criterion for predicting the catalytic activity of these salts [35]. If the metal salt has a cation with high electronegativity, this should have a strong Lewis acidity and exhibit good catalytic performance.

In the presence of Fe³⁺ cations, more H⁺ ions are produced, which efficiently catalyze the esterification of levulinic acid. Similarly, the release of Cu²⁺ cations in solution caused a significant decrease in the pH value (from 6.0 to 1.7), and the Cu(NO₃)₂-catalyzed reaction achieved the second-highest conversion. Conversely, the dissolution of salt containing Na⁺, the weakest Lewis acid, and the softest metal cation had a smaller impact on the medium’s pH (from 6.0 to 3.6). Consequently, the reaction in the presence of this catalyst led to the lowest conversion (Table 1). These aspects corroborate that esterification reactions involve mainly Brønsted acid catalysis.

3.1.2. Effect of Iron Catalyst Nature

The type of the metal cation also influenced the reaction’s selectivity (Figure 2). Regardless of metal nitrate, the esterification of levulinic acid resulted in two products: butyl levulinate and pseudobutyl levulinate (5-methyl-5-butoxy-γ-butyrolactone) (Scheme 3). However, the ratio of these two products varied depending on the type of metal cation (Figure 2).

In the presence of Lewis acid metal cations with empty or partially filled “d” orbitals, butyl levulinate is the major product. On the other hand, pseudobutyl levulinate is preferentially obtained without a catalyst or in the presence of sodium nitrate. This means that even a weak acidity, such as that of levulinic acid itself, is sufficient to promote the reaction leading to pseudolevulinate, although with low conversion.

The importance of the anion and the iron oxidation number was evaluated by performing reactions with FeSO₄ and Fe₂(SO₄)₃ (Figure 3). While Fe³⁺ salts had a similar performance, the FeSO₄-catalyzed reaction achieved a poor conversion. The reaction selectivity was also affected by the charge of iron cations. Figure 4 shows the conversion and selectivity of reactions, as well as the pH values of alcoholic solutions of iron catalysts.

The Fe₂(SO₄)₃ and Fe(NO₃)₃ salts were the most active catalysts and achieved the highest selectivity toward the butyl levulinate (>99%). On the other hand, although the FeSO_4-catalyzed reaction also produced butyl levulinate as the main product, its selectivity was lower (70%), while the selectivity of the pseudobutyl levulinate was increased to 30%. Once again, the catalysts that released more H⁺ ions in solution (see pH values, Figure 4) were the most efficient.

Zhao et al. proposed a mechanism in which levulinic acid is converted to pseudobutyl levulinate and then into butyl levulinate [36]. The authors suggested that initially there was a keto–enol tautomeric equilibrium, generating an intermediate that releases a water molecule, forming angelica lactone (Scheme 4).

It is important to note that these steps do not necessarily require the presence of metal or cations H⁺ in large amounts. Then, the addition of H⁺ ions to the double bond results in a stable tertiary carbocation, which then undergoes a nucleophilic attack of butyl alcohol, resulting in pseudobutyl levulinate. Although not included in Scheme 4, the authors suggested that the pseudobutyl levulinate can itself suffer successive protonation steps, generating intermediates that after the release of H⁺ ions, butyl alcohol, and water molecules also give butyl levulinate [35].

To verify the consecutive character of reactions between butyl alcohol and levulinic acid (formation of the pseudobutyl levulinate and its conversion to butyl levulinate), selectivity versus time plots for the reactions conducted with different catalysts and temperatures were built and are displayed in Figure 5.

Initially, two reactions in the presence of Fe(NO₃)₃ catalyst were studied at different conversions and temperatures. At the lowest temperature (298 K, Figure 5a), both selectivities for pseudobutyl levulinate and butyl levulinate achieved 50% within the first hour of the reaction. Moreover, the conversion maximum was also achieved in this time interval (see Section 3.1.4). Therefore, it seems reasonable to assume that after this period, all the butyl levulinate ester formed came from the pseudobutyl levulinate, as indicated by the selectivity curves shown in Figure 5a. This supports the hypothesis presented in Scheme 4.

Secondly, at the highest temperature (338 K, Figure 5b), the butyl levulinate selectivity achieved 90% within the first-hour reaction. The pseudobutyl levulinate selectivity also had its maximum within the first-hour reaction, being gradually posteriorly diminished over time. The difference between the two reactions (Figure 5a,b)) is that in the first, the conversion of levulinic acid stayed almost constant after the first reaction hour (≈40%), whereas, in the reaction at the highest temperature, the conversion was gradually increased over time, rising from 46% to 92%. It is clear that in this case, the most butyl ester came from levulinic acid.

In Figure 5c, the Cu(NO₃)₂-catalyzed reaction was chosen. During the first hour of the reaction, the selectivity of pseudobutyl levulinate achieved its maximum; afterward, it was gradually diminished during the reaction. However, the conversion of the reaction also increased, from 23% to 42%. Therefore, it can be concluded that both pseudobutyl levulinate and levulinic acid were simultaneously converted to butyl levulinate ester.

3.1.3. Effect of Fe(NO₃)₃ Catalyst Concentration

The impact of catalyst concentration on the esterification of levulinic acid was evaluated, and the kinetic curves are shown in Figure 6. An increase in catalyst concentration had a beneficial effect on both the initial reaction rate and conversion of the reactions. This effect can be attributed to the greater amount of Fe³⁺ cations, which results in a higher production of H⁺ ions.

It is worth noting that even when Fe(NO₃)₃ was used at the lowest concentration (0.5 mol%) (Figure 5), the reaction reached a conversion higher than that achieved in the presence of NaNO₃, the worst catalyst (Figure 1). The reaction selectivity was also affected by the change in catalyst concentration (Figure 7).

Regardless of catalyst load, butyl levulinate was always the main product. However, a reduction in catalyst concentration resulted in lower butyl levulinate formation. Surprisingly, even in the absence of a metal catalyst, pseudobutyl levulinate was the main product (see Figure 2). This suggests that its formation is less dependent on the H⁺/M⁺ presence than butyl levulinate ones.

3.1.4. Effect of Reaction Temperature

The impact of temperature on the levulinic acid esterification was assessed the kinetic curves obtained are shown in Figure 8.

When the temperature increased, the reactions occurred more quickly. However, at 298 and 308 K, after 2 h of reaction, almost no progress occurred, and the conversions remained almost constant. On the other hand, at temperatures of 318 K or higher, the reaction conversions gradually increased, albeit at a slower rate than initially due to a decrease in substrate concentration. As the reaction temperature increases, the number of effective collisions also increases, resulting in a faster reaction.

The reaction selectivity was also affected by the increase in reaction temperature (Figure 8). An increase in reaction temperature favored the preferential formation of butyl levulinate, as demonstrated unequivocally in Figure 9. It is evidence of the endothermic character of this reaction. Moreover, in the formation of butyl levulinate, two water molecules are lost, while for the pseudobutyl levulinate, only one is released. Certainly, an increase in temperature enhances this step.

The selectivity of the reaction was also influenced by the rise in reaction temperature (see Figure 8). An increase in temperature led to greater production of butyl levulinate, as depicted in Figure 9. This indicates that the reaction is endothermic. Additionally, the formation of butyl levulinate involves the loss of two water molecules, while only one water molecule is lost in the formation of pseudobutyl levulinate. Therefore, increasing the temperature enhances this particular step.

3.1.5. Effect of Alcohol in Levulinic Acid Esterification

The effectiveness of the Fe(NO₃)₃ catalyst was tested in levulinic acid esterification reactions using short-chain alkyl alcohols. The kinetic curves are shown in Figure 9. The reactivity of alcohols in esterification reactions follows the pattern primary > secondary > tertiary. This is due to steric hindrance caused by the methyl groups on the hydroxyl group. As shown in Figure 10, this trend was observed in the evaluated reactions, with secondary (2-propanol and 2-butanol) and tertiary (tert-butanol) alcohols being less reactive. On the other hand, the reactions with primary alcohols proceeded more quickly and achieved higher conversions than the others.

The reactivity of primary alcohols was as follows: methanol > ethanol > propanol ≈ butanol. These results agree with the literature [30,37]. The highest reactivity of methanol is attributed to the electron-donating effect induced by the methyl group on the hydroxyl group, which makes it a more efficient nucleophile.

The tert-butyl alcohol had the lowest conversion rate, and no product was detected through GC analyses. It is possible that due to the acidic medium, it underwent dehydration and converted into olefins (butenes) that were not identified. Conversely, in all the other reactions, regardless of the alcohol, alkyl levulinate was consistently the primary product, and pseudolevulinate was the secondary one (see Figure 11).

4. Conclusions

The effectiveness of metal nitrate catalysts was evaluated in levulinic acid esterification reactions using butyl alcohol. The Fe(NO₃)₃ salt was found to be the most active and selective catalyst for producing butyl levulinate. This high activity was attributed to the strong Lewis acidity of the Fe³⁺ cation, which increases its ability to react with water and release H⁺ cations in solution, which are efficient Brønsted acid catalysts. Regardless of the metal nitrate used, butyl levulinate and pseudobutyl levulinate were always the primary and secondary products, respectively. The impact of various reaction parameters was studied. It was observed that increasing the catalyst load or reaction temperature had a positive effect, mainly on the conversion rates of the reaction, as well as on the selectivity of butyl levulinate. The reaction mechanism was also investigated. Both Mⁿ⁺ and H⁺ cations efficiently activated the carbonyl group of levulinic acid, making it more susceptible to attack by the alcohol hydroxyl group, resulting in esterification products. It was confirmed that in reactions with low conversion (40%), there is a consecutive conversion of levulinic acid to pseudobutyl levulinate and posteriorly to butyl levulinate ester. By assessing the influence of the type of alcohol, it was verified that primary alcohols were the most reactive, while secondary and tertiary alcohols led to poor conversion and ester selectivity due to steric hindrance caused by the methyl groups on the hydroxyl group.