Conversion of Cellulose to Lactic Acid by Using ZrO₂–Al₂O₃ Catalysts

Abstract

Lactic acid has a wide range of applications in many industries, both as an ingredient and as an intermediate. Here, we investigated the catalytic conversion of cellulose to lactic acid by using heterogeneous mixed-oxide catalysts containing ZrO₂. Although pure ZrO₂ has catalytic activity for the conversion of cellulose to lactic acid, the yield of lactic acid obtained is not satisfactory. In contrast, a series of ZrO₂–Al₂O₃ catalysts containing various percentages of ZrO₂ provided higher yields of lactic acid. The ZrO₂–Al₂O₃ catalysts had more Lewis acid sites and far fewer base sites than ZrO₂. This suggests that the Lewis acid sites on ZrO₂–Al₂O₃ catalysts are more important than the base sites for the conversion of cellulose to lactic acid.

1. Introduction

Biomass is a renewable organic resource that contains components that are essential for the production of many important chemicals and fuels [1,2,3]. Inedible lignocellulosic biomass is plant biomass that is mainly composed of cellulose, hemicellulose, and lignin. Currently, cellulose is the most promising component for biomass utilization because cellulose is the major component of lignocellulosic biomass [4,5,6,7].

Lactic acid is an organic compound that is used in large quantities as the food, cosmetic, pharmaceutical, and chemical productions [8,9]. Moreover, lactic acid is an important intermediate for conversion to other products such as propylene glycol, acrylic acid, and polylactic acid [8,10]. Conventionally, lactic acid is produced via the fermentation of sugars from starch [11]; however, the rate of production using this method is low, and the pH of the solution during fermentation must be kept in the range 5–7 to provide an environment suitable for lactic acid-fermenting bacteria. Thus, catalytic conversion of inedible cellulose is an attractive alternative to fermentation for the production of lactic acid.

Recently, several homogeneous catalysts (e.g., PbCl₂ and ErCl₃) have been investigated for their suitability for the conversion of cellulose to lactic acid [12,13]. These catalysts provide a high yield of lactic acid; however, their use in industries remains unrealized because of difficult separation of the products from the catalyst as well as the poor stability and recyclability of the catalysts. Therefore, heterogeneous catalysts for the conversion of cellulose to lactic acid are desired because products can be easily separated from this type of catalyst. Several heterogeneous catalysts have been reported to have catalytic activity for the conversion of cellulose to lactic acid, such as LaCoO₃, NbF₅–AlF₃, and AlW, which provided yields of lactic acid of 24%, 27%, and 28%, respectively [14,15,16]. However, these solid catalysts have a disadvantage that the metal species leach into solution during the reaction—2.4% of the Co and 1.5% of the La from LaCoO₃ [14] and 1.5% of the W from AlW [16] leached into solution. Additionally, these heterogeneous catalysts are relatively expensive, making their industrial use difficult. Previously, we reported a simple ZrO₂ catalyst for the conversion of cellulose to lactic acid (yield, 21.2%) that did not leach metal species into the reaction solution [17]. In addition, the ZrO₂ catalyst was stable in hot water. However, the yield of lactic acid obtained with this catalyst was less than those reported for the heterogeneous catalysts. Thus, the aim of the present study was to examine how to increase the yield of lactic acid obtainable with ZrO₂-based catalysts. Dispersion of active species on supports such as metal oxides is a common means of increasing the catalytic activity of heterogeneous catalysts. Therefore, here we compared the catalytic activity of a series of mixed-oxide catalysts containing various amounts of ZrO₂ (5%, 10%, or 20%) on an Al₂O₃ support with that of pure ZrO₂. We found that the 10%ZrO₂–Al₂O₃ catalyst showed activity for the production of lactic acid from cellulose with a product yield of 25.3%, which was cheaper than the reported heterogeneous catalysts.

2. Results and Discussion

2.1. Effect of Ball-Milling Treatment on Cellulose Conversion

In this paper, pulverization of cellulose by ball-milling has been applied for the increase of cellulose conversion [18,19,20]. The conversion of cellulose was only 25% in the case of un-milled cellulose (473 K, 6 h, without catalysts) and increased to 86% by the ball-milling treatment of cellulose. Figure 1 shows XRD patterns of un-milled and milled cellulose. The cellulose crystallinity was decreased by the ball-milling treatment as revealed by the peak broadening at 2θ = 22.6 degree (crystalline plane 002) in XRD patterns (Figure 1) [19,21], indicating that the cellulose conversion increased with decreasing cellulose crystallinity. In the reaction pathway, cellulose was firstly hydrolyzed to soluble glucose; thus, cellulose with high crystallinity was hydrolyzed very slowly at 473 K from the result of the low conversion of un-milled cellulose.

2.2. Catalytic Conversion of Cellulose to Lactic Acid by Using ZrO₂–Al₂O₃

Previously, we investigated the catalytic conversion of cellulose to lactic acid by using various transition metal oxides (i.e., ZrO₂, Al₂O₃, TiO₂, Fe₃O₄, V₂O₅, CeO₂, Y₂O₃, Tm₂O₃, HfO₂, Ga₂O₃, MgO, La₂O₃, Nb₂O₅, and Ta₂O₅) [17]. Of these catalysts, ZrO₂ had the highest catalytic activity (lactic acid yield, 21.2%; reaction temperature, 473 K). Therefore, in the present study, we examined the use of an Al₂O₃ support to increase the catalytic activity of ZrO₂ for the conversion of cellulose to lactic acid. The yields of lactic acid obtained with ZrO₂–Al₂O₃ mixed-oxide catalysts containing different percentages of ZrO₂ at 473 K were higher than those obtained by using the individual metal oxides, indicating that the dispersion of active ZrO₂ species on the Al₂O₃ support increased the catalytic activity of ZrO₂ (Figure 2). The maximum yield of lactic acid (25.3%) was obtained when 10%ZrO₂–Al₂O₃ was used as the catalyst. The yield of lactic acid increased slightly as the amount of ZrO₂ on the Al₂O₃ support was increased from 5% to 10%; however, the yield decreased as the amount of ZrO₂ was increased to 20%, indicating that the yield of lactic acid may decrease further as the percentage of ZrO₂ on the support increases.

2.3. Reaction Conditions

Next, we optimized the reaction conditions to obtain the highest possible yields of lactic acid with 10%ZrO₂–Al₂O₃ used as the catalyst. Figure 3 shows the product yields obtained as a function of reaction time. The yield of lactic acid increased for up to 6 h, after which it decreased due to decomposition or polymerization of lactic acid.

Figure 4 shows the product yields obtained as a function of reaction temperature. The maximum yield of lactic acid (25.3%) was obtained at 473 K. At the lowest temperature examined (i.e., 453 K), the cellulose conversion reaction proceeded very slowly. At the highest temperature examined (i.e., 493 K), the yield of lactic acid was less than that at the optimum temperature because the lactic acid was likely decomposed at this temperature.

Finally, we investigated the reusability of the 10%ZrO₂–Al₂O₃ catalyst for repeated conversion of cellulose to lactic acid. After the reaction with the optimized conditions (473 K, 6 h), the recovered solid, which contained the used 10%ZrO₂–Al₂O₃ catalyst and any unreacted cellulose, was heated at 673 K for 15 h under an air atmosphere to remove carbon-based material accumulated on the surface of the catalyst. The yield of lactic acid obtained by using the reused 10%ZrO₂–Al₂O₃ catalyst at 473 K for 6 h was 31.7%, which was higher than the yield obtained from the first reaction. One possible reason is that the calcination of the 10%ZrO₂–Al₂O₃ catalyst after the reaction would eliminate contamination on the surface, resulting in the higher lactic acid yield. This result suggests that the 10%ZrO₂–Al₂O₃ catalyst is recyclable.

2.4. Characterization

Next, we characterized the ZrO₂–Al₂O₃ catalysts to further understand their catalytic activity. The X-ray diffraction (XRD) patterns of the catalysts are shown in Figure 5. In the XRD patterns of the three ZrO₂–Al₂O₃ catalysts, peaks at 46.0° and 66.5° were attributed to γ-Al₂O₃ [22]. In the XRD patterns of 10%ZrO₂–Al₂O₃ and 20%ZrO₂–Al₂O₃, peaks at 30.3° and 50.7° were attributed to tetragonal ZrO₂ [23]; however, these peaks were not seen in the XRD pattern of 5%ZrO₂–Al₂O₃, indicating that ZrO₂ was either highly dispersed or in the amorphous phase in this catalyst. The XRD pattern for pure ZrO₂ revealed that it was in the monoclinic phase (Figure 5d).

In our previous paper, we hypothesized that the properties of the acid and base sites on ZrO₂ play an important role in the catalytic activity of this oxide for the conversion of cellulose to lactic acid [17]. Therefore, we examined the acid and base properties of the ZrO₂–Al₂O₃ catalysts by means of temperature-programmed desorption of ammonia (NH₃-TPD) and temperature-programmed desorption of carbon dioxide (CO₂-TPD), respectively. Figure 6 shows NH₃-TPD profiles of the ZrO₂–Al₂O₃ and ZrO₂ catalysts. The NH₃ desorption temperatures of the three ZrO₂–Al₂O₃ catalysts were all 470 K, which was similar to that of ZrO₂, indicating that the acid sites on the ZrO₂–Al₂O₃ and ZrO₂ catalysts were weakly acidic Lewis acid sites [24] (Figure 6). The shoulder peaks at 550 K were observed in the case of ZrO₂–Al₂O₃ catalysts, which were ascribed to slightly stronger acid sites than weakly acid sites (peak at 470 K). The slightly stronger acid sites might be located on the boundary between ZrO₂ and Al₂O₃. Although the XRD analysis revealed that the crystal phase of ZrO₂ differed depending on whether it was in pure or mixed-oxide form, the NH₃ desorption temperatures revealed that the acid strength was almost the same irrespective of form.

The desorption temperatures of CO₂ from the ZrO₂–Al₂O₃ and ZrO₂ catalysts were in the range 340–450 K (Figure 7), indicating that the base sites on the ZrO₂ catalysts were weakly basic [25]. The amount of CO₂ desorbed from the ZrO₂–Al₂O₃ catalysts was smaller than that desorbed from the ZrO₂ catalyst.

Table 1. Properties of the ZrO₂–Al₂O₃ and ZrO₂ catalysts.

Catalyst	Surface Area 1/m2 g−1	Crystal Phase of ZrO2 2	Amount of Acid Sites 3/mmol g−1	Amount of Base Sites 4/mmol g−1	Yield of Lactic Acid 5/%
5%ZrO2–Al2O3	135	-	0.109	0.093	25.0
10%ZrO2–Al2O3	126	Tetragonal	0.131	0.057	25.3
20%ZrO2–Al2O3	130	Tetragonal	0.180	0.098	21.9
ZrO2 6	101	Monoclinic	0.118	0.485	21.2

¹ Reported by manufacturer; ² Determined by XRD; ³ Determined by temperature-programmed desorption of ammonia (NH₃-TPD); ⁴ Determined by temperature-programmed desorption of carbon dioxide (CO₂-TPD); ⁵ Reaction conditions: ball-milled cellulose, 0.5 g; catalyst, 1.0 g; water, 50 g; reaction temperature, 473 K; reaction time, 6 h; ⁶ Presented in reference [17].

shows the amounts of acid and base sites on the ZrO₂–Al₂O₃ and ZrO₂ catalysts. The ZrO₂–Al₂O₃ catalysts had more acid sites but far fewer base sites than the ZrO₂ catalyst. In our previous paper, we suggested that the retro-aldol reaction (the conversion of fructose to glyceraldehyde and dihydroxyacetone)—the key step in the conversion of cellulose to lactic acid—involved both the acid and base sites on the ZrO₂ catalyst [17]. However, the ZrO₂–Al₂O₃ catalysts had more acid sites and far fewer base sites than did ZrO₂. We discuss the reaction pathway in the next section.

2.5. Reaction Pathway

The conversion of cellulose to lactic acid is a multi-step reaction (Scheme 1). First, cellulose is converted to glucose by hydrolysis, which can be catalyzed by acid catalysts [26,27]. Then, glucose is isomerized to fructose by a Lewis acid or basic catalysts [28]. The key step in the conversion pathway from cellulose to lactic acid is the conversion of fructose to glyceraldehyde and dihydroxyacetone via a retro-aldol reaction involving C–C bond cleavage, which can be enhanced by Lewis acids [29,30]. In a previous paper, we hypothesized that the combination of acid and base sites on ZrO₂ enhanced the conversion of fructose to glyceraldehyde and dihydroxyacetone [17]. However, in the present study, the yield of lactic acid obtained from cellulose by using the ZrO₂–Al₂O₃ catalysts did not appear to depend on the relative amount of acid and base sites. That is, the ZrO₂–Al₂O₃ catalysts provided a higher yield of lactic acid than did the ZrO₂ catalyst, even though they had more acid sites and far fewer base sites than did ZrO₂ (Table 1). This indicates that the number of base sites is not important for the catalytic conversion of cellulose to lactic acid with ZrO₂–Al₂O₃ catalysts. Furthermore, although the Lewis acid sites on the ZrO₂–Al₂O₃ catalysts played an important role in the conversion of cellulose to lactic acid, the yield of lactic acid obtained was not proportional to the number of acid sites. This suggests that dispersion of ZrO₂ on a Al₂O₃ support and accessibility to the acid sites on the catalyst are the most important factors for the conversion of cellulose to lactic acid when using ZrO₂–Al₂O₃ catalysts. Further studies are required to elucidate the details of the relationship between the active sites and the product yield and also the reaction mechanism.

In the final step of the reaction pathway, glyceraldehyde and dihydroxyacetone can be converted to lactic acid at 463 K without a catalyst [17]. However, the conversion of pyruvaldehyde to lactic acid is reported to require a catalyst with Lewis acid sites [31,32], which a ZrO₂-based catalyst could provide.

Previously, we reported that we used a ZrO₂ catalyst to convert cellulose to lactic acid with a yield of 21.2% [17]. However, this yield of lactic acid was less than that obtained with other reported heterogeneous catalysts (24% yield of lactic acid using LaCoO₃ [14], 28% using AlW [16], and 27% using NbF₅–AlF₃ [15]). In the present study, we found that by using the 10%ZrO₂–Al₂O₃ catalyst we could obtain a yield of lactic acid of 25.3%, which was comparable with the yields using the reported catalysts. One advantage of the 10%ZrO₂–Al₂O₃ catalyst is that it can be obtained in lower cost than the reported heterogeneous catalysts. Additionally, a few percentages of the metal species in the reported catalysts leached into solution during the reaction. Conversely, the Zr species was not leached out from ZrO₂ into water during the reaction [17].

3. Materials and Methods

3.1. Materials

ZrO₂ (ZRO-7; reference catalyst from Catalysis Society of Japan) and mixed oxides containing ZrO₂ and Al₂O₃ (5%ZrO₂–Al₂O₃, 10%ZrO₂–Al₂O₃, and 20%ZrO₂–Al₂O₃) were obtained from Daiichi Kigenso Kagaku Kogyo Co., Ltd. (Osaka, Japan). Aluminum oxide (Al₂O₃) was obtained from Sigma–Aldrich Co., LLC (St. Louis, MO, USA). These materials were used without pretreatment.

3.2. Catalytic Reaction

Cellulose (microcrystalline cellulose; Merck KGaA, Darmstadt, Germany) was pulverized with a ball mill at 60 rpm. The conversion of cellulose to lactic acid was carried out in a stainless steel batch reactor with an inner volume of 100 cm³ (MMJ-100; OM Lab-Tech, Tochigi, Japan), as described in a previous paper [17]. Briefly, the reactor was loaded with ball-milled cellulose (0.5 g), catalyst (1.0 g), and water (50 g); purged with nitrogen gas (0.1 MPa); and then heated to 453–493 K for 3–9 h with screw stirring. After the reaction was allowed to run, the resulting mixture was filtered to separate the liquid from the solid. The quantitative analyses of lactic acid, levulinic acid, 5-hydroxymethylfurfural, and furfural in the liquid fractions were carried out with a gas chromatograph (GC-2014; Shimadzu, Kyoto, Japan) equipped with a flame ionization detector and an InertCap capillary column (GL Sciences Inc., Tokyo, Japan). The chemicals in the liquid fraction (e.g., glucose) were analyzed by using a high-performance liquid chromatograph (Shimadzu, Kyoto, Japan) equipped with a refractive index detector (RID-10A; Shimadzu, Kyoto, Japan), an ultraviolet/ visible detector (SPD-20AV; Shimadzu, Kyoto, Japan), and a Rezex RPM-Monosaccharide Pb+2 column (Phenomenex Inc., Torrance, CA, USA). The product yields were calculated based on moles of carbon as follows:

$$\text{Product yield} (\%)=\frac{(\text{Moles of carbon atoms in each product})}{(\text{Moles of carbon atoms in cellulose used})}\times 100$$

(1)

$$\text{Conversion} (\%)=\left(1-\frac{(\text{Weight of solid residue})-(\text{Weight of solid catalyst})}{(\text{weight of cellulose used})}\right)\times 100$$

(2)

3.3. Catalyst Characterization

XRD patterns of the catalysts were determined by using a Rigaku SmartLab XRD system (Rigaku, Tokyo, Japan) with Cu Kα radiation in the 2θ range of 5–90°.

NH₃-TPD profiles were determined with a TPD-1-AT instrument (Bel Japan, Inc., Osaka, Japan) with an online quadrupole mass spectrometer. Samples (ca. 0.05 g) were pretreated at 773 K in flowing helium for 1 h, saturated in flowing 5% ammonia diluted with helium (0.5 cm³ s⁻¹) at 373 K for 30 min, and then treated at 373 K in flowing helium (0.83 cm³ s⁻¹) for 1 h. The samples were then heated at a constant rate of 10 K min⁻¹ from 373 to 953 K while the amount of NH₃ desorbed was detected.

CO₂-TPD profiles were determined with a Micromeritics 3FLEX 3500 chemisorption analyzer (Micromeritics, Norcross, GA, USA) with an online quadrupole mass spectrometer. Samples (ca. 0.2 g) were pretreated at 773 K in flowing helium for 1 h, saturated in flowing CO₂ (0.5 cm³ s⁻¹) at 323 K for 30 min, and then treated at 323 K in flowing helium (0.83 cm³ s⁻¹) for 1 h. The samples were then heated at a constant rate of 10 K min⁻¹ from 323 to 953 K while the amount of CO₂ desorbed was detected.

4. Conclusions

Compared with using pure ZrO₂ as the catalyst (yield, 21.2%), a greater yield of lactic acid was obtained from cellulose by using a 10%ZrO₂–Al₂O₃ catalyst (yield, 25.3%) and optimized reaction conditions (reaction temperature, 473 K; reaction time, 6 h). XRD analysis revealed that the 10%ZrO₂–Al₂O₃ catalyst contained tetragonal ZrO₂ and γ-Al₂O₃. NH₃-TPD and CO₂-TPD analyses revealed that all three ZrO₂–Al₂O₃ catalysts had more Lewis acid sites and far fewer base sites than did ZrO₂. This suggests that the number of Lewis acid sites on the ZrO₂–Al₂O₃ catalysts was more important than the number of base sites for the conversion of cellulose to lactic acid.