Efficient Catalytic Conversion of Glucose into Lactic Acid over Y-β and Yb-β Zeolites.

In this work, a new type of modified β zeolites with rare earth elements (ree) was discovered for producing lactic acid from glucose and achieved a good catalytic effect. At first, the catalytic performances of ree-β zeolites, ree oxides, and single-transition-metal-β zeolites were compared, and the result showed that Y-β and Yb-β zeolites had the best catalytic activity under the same reaction conditions. Under the best reaction conditions, the maximum yields of lactic acid with Y-β and Yb-β catalysts were 45.3 and 43.6%, respectively. The acid characterization showed that Y/Yb-β zeolites had a similar number of Lewis acid sites as Sn-β zeolites, and they were also more than other transition-metal-β zeolites. Thus, Y-β and Yb-β zeolites had a higher lactic acid yield than those catalysts. It is interesting to note that Y-β and Yb-β zeolites owned more Brønsted acids but produced fewer byproducts. Combining the decomposition experiment of 5-hydroxymethyl furfural, fewer byproducts were produced with Y-β and Yb-β zeolites because the low amount of Brønsted acid contained could hinder the decomposition of 5-hydroxymethyl furfural, thereby slowing down the side reaction.

Introduction

With the continuous consumption of fossil energy,[1] the world’s resources are becoming increasingly scarce and the environment is gradually deteriorating.[2,3] At the same time, the development of clean and renewable energy has become more and more of concern because of the shortage of fossil energy.[4,5] Biomass energy is regarded as the most potential alternative energy because of its wide distribution, low price, environmental friendliness, and sustainable use.[6,7] Among the chemicals obtained from the biomass resources, lactic acid, as an important platform molecular compound,[8,9] has a wide range of uses in the chemical, pharmaceutical, and food industries.[10,11] Especially, it is often used as a synthetic material for biodegradable plastics.[12,13] The traditional methods of producing lactic acid are mainly microbial fermentation methods, and the main raw materials for microbial fermentation are starch-rich crops such as rice, corn, wheat, and potatoes. The production process of the fermentation method is complicated, and the production cycle is long.[14,15] Compared with the biological fermentation method of the production of lactic acid, the catalytic method has a higher large-scale production potential and higher production efficiency, and especially, it uses biomass instead of food as a substrate. Thus, the catalytic method seems to provide a more acceptable way for the preparation of lactic acid.[16]

According to the mixing state of the catalyst and the reaction substrate during the catalytic reaction of biomass, the catalysts can be divided into homogeneous catalysts and heterogeneous catalysts. However, the homogeneous catalyst has the disadvantages of difficult separation and inconvenient recycling.[17,18] Therefore, heterogeneous catalysts without the above shortcomings have attracted greater attention. The β molecular sieve, as a heterogeneous catalyst, was synthesized for the first time in 1967. Because of the unique pore structure and high silicon–aluminum ratio, the β molecular sieve has good acid resistance, thermal stability, and hydrophobicity.[19] It is even widely used in petroleum refining and petrochemical processes such as isomerization, hydrocracking, alkylation, dewaxing, etc. In particular, adding a small amount of β molecular sieve or modified β molecular sieve to the cracking reactions can significantly improve the hydrothermal stability, coking resistance, and abrasion resistance.[20,21] In addition, the β molecular sieve has also been found to have unique effects in the fields of biomass catalysis.[22−27] Therefore, it is very worthwhile to look forward to the application of the β molecular sieve or the modified β molecular sieve in biomass catalytic reactions. Taarning et al. reported in 2010 that Sn-β zeolites prepared by the hydrothermal synthesis method catalyzed sugars to produce lactic acid or lactate in the aqueous and organic phases.[28] As a heterogeneous catalyst, the excellent catalytic performance and structural stability of modified β zeolites immediately triggered a research boom in its catalytic utilization of biomass.[29−31] In 2016, Dong et al. reported the chemical conversion of sugars using Zn-Sn-β zeolites. The glucose conversion rate exceeded 99% and the lactic acid production rate reached 54% within 2 h under ambient pressure at 190 °C.[32]

The Sn element is the dominant element of the modified β zeolite catalyst commonly used for converting biomass into lactic acid. Is it possible to find a new type of metal element that can replace Sn? Rare earth elements have attracted our attention due to their excellent catalytic properties and abundant reserves in the chemical industry.[33,34] Currently, rare earth catalytic materials play an important role in petrochemical catalysis, vehicle exhaust catalysis, fuel cell catalysis, catalytic combustion, photocatalysis, and so on.[35−37] However, there is very little research in the field of lactic acid production from biomass using rare earth catalytic materials. Herein, we investigated the effects of different rare earth elements on the production of lactic acid from glucose and the various influencing factors such as reaction time, temperature, and metal loading. The result showed that ree-β zeolites developed in this study could produce more lactic acid and fewer byproducts than those prior catalysts. Therefore, this work broke the situation that most modified β zeolites were dominated by Sn[38,39] and applied rare earth elements with excellent catalytic efficiency in the field of catalytic biomass production of lactic acid, which was of great significance for the industrial production of lactic acid.

Experimental Section

Materials and Experimental Procedure

d-(+)-glucose (99.5%), d-(−)-fructose (99%), ethanol (≥99.5%), indium acetate, zinc acetate, copper acetate, lanthanum acetate, cerium acetate, ytterbium acetate, yttrium acetate, lanthanum oxide, cerium oxide, yttrium oxide, and ytterbium oxide were purchased from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). Tin(II) acetate was obtained from Sigma-Aldrich. Formic acid (≥98%), acetic acid (99.8%), and levulinic acid (99%) were purchased from Wako Co., Ltd. Lactic acid (1.0 M) was purchased from Alfa Isa (China) Chemical Co., Ltd. 5-Hydroxymethyl furfural (98%) was purchased from Johnson & Johnson Technology Co., Ltd. The β molecular sieve was purchased from Nankai University Catalyst Factory Co., Ltd.

Synthesis of Materials

The catalysts were prepared following the previously published protocol.[40] The dealumination of β zeolite (deAl-β) was prepared as follows: 25 g of commercial β zeolite (Catalyst Plant of Nankai University) with a Si/Al ratio of 25 was put into 0.5 L of HNO3 solution (65–68%, w/w) at 103 °C for 20 h. The deAl-β zeolite was washed until the supernatant was neutral; then, the solid was collected and dried overnight at 100 °C.

Catalyst Preparation

Solid-state ion exchange was performed by grinding the deAl-β and the metallic acetate for 40 min. Finally, the samples were calcined in static air at 550 °C with a ramp of 2 °C/min for 6 h.

Catalyst Recycling

The catalysts were recovered by centrifugation, washed extensively with absolute ethanol, dried at 90 °C overnight, and then roasted at 550 °C for 6 h.

Catalytic Reactions

The catalytic reactions were performed in a closed Teflon vessel (25 mL) in a stainless-steel autoclave. In a typical procedure, the vessel was charged with glucose, catalyst, and water (10 mL) followed by heating to the desired reaction temperature in a rotating oven (20 rpm). After hours of stirring, the autoclave was cooled, and the reaction mixture was analyzed. All of the experiments were replicated at least three times, and the mean values are reported.

Products Analysis

The concentrations of sugars, organic acids (i.e., lactic acid, formic acid, and levulinic acid), and 5-hydroxymethy furfural were determined on an Agilent 1260 series HPLC equipped with an RI detector and a UV detector (210 nm) using a column (Bio-Rad HPX-87H). The mobile phase was a 5 mM H2SO4 aqueous solution at a flow rate of 0.4 mL/min, and the temperatures of the column and the RI detector were set at 55 and 45 °C, respectively. The concentration of each respective component was determined based on the external standard calibration curves.

Results and Discussion

Effect of Different Catalysts on the Reaction

Before the experiments, we checked the literature and learned the specific yields of lactic acid catalyzed from glucose by some single-transition-metal-β zeolites,[41,42] and repeated the experiments. To investigate the effect of rare earth elements on the production of lactic acid from glucose, we loaded the rare earth elements (Y, Yb, Ce, and La) on β molecular sieves according to the above method. The catalytic performances of converting glucose into lactic acid by ree oxides, ree-β zeolites, and single-transition-metal-β zeolites are shown in Figure . During the catalyst preparation process, the metal loading of the catalyst was 0.42 mmol/g, and the hydrothermal reaction conditions were 190 °C, 2 h, catalyst 160 mg, glucose 225 mg, H2O 10 mL.

Figure 1

Yield of lactic acid from glucose over different catalysts.

It could be seen in Figure that the lactic acid yield of the blank experimental group was less than 4%, indicating that it was difficult to convert glucose into lactic acid under purely hydrothermal conditions without adding a catalyst. After adding β or deAl-β zeolite, the lactic acid yields were 5.7 and 4.4%, respectively. Compared with the experimental group without the catalyst, the lactic acid yields did not increase significantly, showing that β or deAl-β zeolite could hardly produce lactic acid from glucose. The yield of lactic acid was greatly improved when using single-transition-metal-β zeolites as catalysts, and their catalytic capacity for producing lactic acid was as follows: Sn-β > Zn-β > In-β > Cu-β. Among them, Sn-β zeolite showed the best catalytic effect, and the lactic acid yield reached 22.4%. As for the four rare earth element-supported catalysts, Y-β and Yb-β zeolites showed excellent catalytic efficiency and obtained lactic acid yields of 27.5 and 25.7%, respectively. This might be due to the unique properties of Y and Yb as heavy rare earth elements. The catalytic effects of Y-β and Yb-β zeolites were even better than those of Sn-β zeolites. By the way, the catalytic effect of the rare earth oxide was very poor, and only a small amount of lactic acid (<5%) was produced. In summary, the efficiency sequence of modified β zeolites to catalyze glucose into lactic acid was as follows: Y-β > Yb-β > Sn-β > Zn-β > La-β > In-β > Cu-β > Ce-β. In the above work, we found that the zeolites loaded with rare earth metal Y and Yb exhibited the best performance for the formation of lactic acid from glucose, so we selected these two rare earth elements for exploration in the follow-up research.

Catalyst Characterization

BET Characterization of Catalysts

Figure a shows the nitrogen adsorption–absorption curve of different β zeolites. It shows fine hysteresis loops at the relative pressure of 0.4 < P/P0 < 0.9 in all catalysts, which is a typical type I isotherm. This meant that micropores existed in all catalysts and the modification did not change the overall framework structure. The hysteresis loop of β zeolites became larger after dealumination treatment. Compared with deAl-β zeolites, the modified Y/Yb-β zeolites showed a smaller hysteresis loop, indicating that the metal had been successfully loaded and the channel was partially filled. In addition, it could be seen from Figure b that the hysteresis loop decreased with the increase of metal loading, indicating that the number of internal holes of the catalyst decreased with the increase of metal loading.

Figure 2

(a) Nitrogen adsorption/desorption isotherms of different catalysts. (b) Nitrogen adsorption/desorption isotherms of catalysts with different loadings.

XRD Characterization of Catalysts

The structure and morphology characteristics of various catalysts and rare earth oxides were also characterized, and the XRD patterns of each catalyst are shown in Figure . All β zeolites had typical topological characteristics of BEA, with a typical diffraction peak at 2θ = 22.38°, confirming that the structure of the commercial β zeolite was not notably destroyed during the process of dealumination and modification, and all catalysts maintained the integrity of the phase structure. As the metal (Y/Yb) loading increased, the peak width of the modified catalyst decreased at 2θ = 22.38°. It was worth mentioning that hardly ree oxides were observed for the peaks that occurred at 29.2 and 29.84°, respectively, indicating that the introduced metal elements did not form a nonframework oxide structure on the surface, but might have entered the framework structure of the molecular sieve.[41]

Figure 3

X-ray diffraction patterns of various catalysts.

Acid–Base Characterization of Catalyst

The excellent catalytic activity of the Y-β and Yb-β zeolites might be explained by its acid–base properties. NH3-TPD and CO2-TPD characterization is a more general and simpler method for determining the acid–base distribution of catalysts.[32] The acid and base properties of the β-type zeolites were examined by NH3-TPD and CO2-TPD characterizations in the study.

Generally speaking, according to the different desorption temperatures, acid sites can be divided into weak acid sites, medium-strong acid sites, and strong acid sites. The NH3-TPD characterization results of each catalyst are shown in Figure . It could be seen that deAl-β after the dealumination had no acidic sites, while both modified catalysts showed weak acidic sites at 100–200 °C. Y-β exhibited certain medium acidic sites at 300–400 °C, and Yb-β exhibited strong acidic sites at 400–600 °C. Table lists the distribution of specific acidic sites of the catalysts. On the whole, the total number of acidic sites of the two catalysts was basically the same, which might be the reason for the similar lactic acid yields obtained under the two catalytic systems.

Figure 4

(a) NH3-TPD curves of various catalysts. (b) Acid strength distribution of Y-β. (c) Acid strength distribution of Yb-β.

Table 1

Acidity of Different Catalysts

catalyst	weak acidity (%)	medium acidity (%)	strong acidity (%)	B150 (mmol·g–1)	L150 (mmol·g–1)
Y-β	61.3	38.7		0.028	0.032
Yb-β	53.4		46.6	0.025	0.030

The distribution of acidic sites between the two catalysts is shown in Table . They were both mainly based on weak acid sites, and Y-β zeolites had more weak acid sites and less medium/strong acids comparatively.

Pyridine-probed FT-IR (Py-IR) analysis could identify and quantify the Brønsted and Lewis acid sites readily. Figure presents the infrared spectra of pyridine adsorbed on the catalysts at 150 °C. The characteristic peaks of Lewis acid sites were mainly at 1450 and 1610 cm–1, and the characteristic peaks of Brønsted acid sites were mainly at 1544 and 1633 cm–1. It could be seen from the figure that there were two types of acidic sites at the same time in the catalysts. The content of Lewis acid also increased with the increase in metal loading, and the content of Brønsted acid did not change significantly.

Figure 5

FT-IR spectra following the adsorption of pyridine on various catalysts.

The base properties of the catalysts were analyzed, as shown in Figure . It could be seen that the modified catalysts (Y/Yb-β zeolites) exhibited a new strong desorption peak located at about 300 °C, associated with the moderate base sites at 300–600 °C, and the deAl-β showed no basic sites. These phenomena confirmed the incorporation of ree metal (Y/Yb) and the resultant basicity increase.

Figure 6

CO2-TPD curves of various catalysts.

Effect of Metal Loading on the Reaction

The effect of the metal contents of Y and Yb in the β zeolite employed in the catalytic reaction on the formation of lactic acid from glucose was also studied in this section.

ICP characterization of catalysts with different metal loadings was also carried out (as shown in Table S1). ICP analysis showed that the amount of metal was identical to the desired value, confirming that the modification method selected in this study successfully loaded the metal to β zeolites. As shown in Figure , the lactic acid yields of Y-β and Yb-β zeolites increased gradually with the increase in metal loading from 0.2 to 1.2 mmol/g. At a metal loading of 1.2 mmol/g, the maximum lactic acid yields of Y-β and Yb-β zeolites were 45.2 and 43.6%, respectively. When the metal loading continued to increase to 1.4 and 1.6 mmol/g, the lactic acid yield decreased instead. It is speculated that the decrease in lactic acid yield might be due to the congestion inside the catalysts. In addition, it could be also seen from Figure b that the hysteresis loop decreased with the increase in metal loading from 0.4 to 1.2 mmol/g, indicating that the number of internal holes of the catalyst decreased with the increase of metal loading. In summary, the metal loading of the catalyst selected in the subsequent experiments was 1.2 mmol/g.

Figure 7

Effect of metal loading on the yield of lactic acid. Reaction condition: 225 mg of glucose, 160 mg of catalyst, 10 mL of H2O, 190 °C, 3 h.

Effect of Temperature and Time on the Reaction

To optimize the yield of lactic acid, we investigated the effects of different reaction temperatures and reaction times on the production of lactic acid from glucose under the Y-β and Yb-β catalytic systems (as shown in Figure ).

Figure 8

Effect of reaction temperature and reaction time on lactic acid yield. Reaction condition: 225 mg of glucose, 160 mg of catalyst, 10 mL of H2O.

It could be seen in Figure that in the two catalytic systems of Y-β and Yb-β zeolites, when the reaction temperature was 130 °C, the formation of lactic acid was almost undetectable in the first 2 h. After 5 h of reaction, only a small amount of lactic acid was produced, indicating that the catalytic reaction did not proceed to a high degree under this temperature condition. This might be because fructose in the intermediate reaction step generates dihydroxyacetone or glyceraldehyde through retro aldol, which required higher energy to open the C–C bond. Therefore, in general experiments, the reaction temperature involved in this step was above 140 °C.[43]

As the reaction temperature increased from 130 to 230 °C, the lactic acid yield and the catalytic rate obviously increased in the first 1 h. The yield of lactic acid reached the highest at 3 h when the temperatures were 190, 210, and 230 °C. As the temperature rose to 210 and 230 °C from 190 °C, the yield of lactic acid decreased in the end of the reaction. It has been reported that lactic acid could be converted to formic acid and acetic acid in high-temperature water.[44,45] Thus, the reason for this phenomenon (decreased yield of lactic acid) in the preliminary analysis might be that the decomposition rate of lactic acid is higher than the production rate at 210 and 230 °C.

When the temperature was 190 °C, with the extension of the reaction time, the glucose conversion rate gradually increased, and the maximum yield of lactic acid (45.2% for Y-β, 43.2% for Yb-β) was obtained at 3 h. Then, the lactic acid yield remained unchanged without a noticeable trend of increasing or decreasing, indicating that the produced lactic acid was relatively stable. In summary, we ensured the optimal reaction conditions for converting glucose into lactic acid as 190 °C, 3 h.

Catalyst Stability Research (Recycling Research)

The recycling ability of the Y-β and Yb-β zeolites for the conversion of glucose was investigated, and the results are shown in Figure . After each reaction, the solid–liquid mixture in the reactor was taken out, centrifuged, and then washed with absolute ethanol. Then, the separated solids were dried overnight and roasted at a temperature of 550 °C for 6 h. As shown in Figure , the catalytic systems of Y-β and Yb-β zeolites achieved lactic acid yields of 45.24 and 42.05% in the first recycle reaction, respectively. Then, a decrease in the yield of lactic acid was observed within three cycles, and the leaching of metals during the reaction might be the cause.[46] ICP characterization was supplemented for the catalysts with three cycles of recycling (as shown in Table S2). As could be seen from the data in the table, the measured metal content in the catalyst decreased with the increase in recycling times, confirming the possibility of metal leaching in the recycling process.

Figure 9

Recyclability tests of catalysts on production of lactic acid. Reaction condition: 225 mg of glucose, 160 mg of catalyst, 10 mL of H2O, 190 °C, 3 h.

However, the yields of lactic acid could still reach more than 35% after three cycles, which were still higher than those of Sn-β zeolites and single-metal-β zeolites. Therefore, Y-β and Yb-β zeolites were potential catalysts for future application in the conversion of biomass to lactic acid.

Product Analysis of the Catalytic System

We further examined the glucose conversion and product distributions using the Y-β and Yb-β zeolites as a function of reaction time (Figure ). It could be seen that the conversion processes of the products using the two catalysts were very similar. As the reaction time increased from 0.5 to 3.0 h, the conversion of glucose increased to about 95%, and the yield of lactic acid with Y-β and Yb-β zeolites increased from about 7 to 45.2% and 43.2%, respectively. For a longer reaction time, the yield of lactic acid did not further change.

Figure 10

Conversion of glucose and yields of lactic acid, formic acid, levulinic acid, and 5-hydroxymethyl furfural (HMF) over the catalyst. Reaction condition: 225 mg of glucose, 160 mg of catalyst, 10 mL of H2O, 190 °C.

In the early stage of the reaction, fructose was mainly produced by the isomerization of glucose. As the reaction time went by, fructose disappeared, and lactic acid began to form rapidly. Soluble byproducts such as formic acid, levulinic acid, and 5-hydroxymethyl furfural gradually increased with the time of reaction. The yield of the main byproduct (5-hydroxymethyl furfural) eventually stabilized at around 10%. According to previous reports, the current mainstream Sn-β catalytic system could finally obtain a lactic acid yield of about 20% and a 5-hydroxymethyl furfural yield of about 15% in the reaction of catalyzing glucose into lactic acid.[28,32] Compared with Sn-β zeolites, the catalysts found in this study produced more lactic acid and less 5-hydroxymethyl furfural, showing a better catalytic efficiency.

Acidity Comparison of Catalysts

The quantitative data of Brønsted acid and Lewis acid for different catalysts at the desorption temperature of 150 °C are shown in Table . It could be seen from the table that with the same metal loading, Y-β and Yb-β zeolites had more Lewis acid quantification than other single-transition-metal-β zeolites except for Sn-β zeolite. This might be the reason why Y-β and Yb-β zeolites could obtain higher lactic acid yields in the preliminary screening experiments. As the metal loading increased, so did the Lewis acid contents of Y-β and Yb-β zeolites. However, Y-β and Yb-β zeolites had less Lewis acid quantification than those double-metal-β zeolites, so they produced less lactic acid than those double-metal-β zeolites. The experimental results in this study confirmed that the lactic acid yield was directly related to the acid–base distribution of the catalyst, which was consistent with previous reports (Kong ammoniated Sn-β to increase the basicity;[47] Xia supported various metals on β to adjust the acidity and alkalinity of the catalyst[40,41]). Therefore, the acidic sites with appropriate strength and type played a key role in the catalytic reaction.

Table 2

Acidity of Different Catalysts

entry	catalyst	B150 (mmol·g–1)	L150 (mmol·g–1)	B/L
1	β	0.127	0.100	1.270
2	deAl-β	0.003	0.015	0.020
3	Pb-β (0.4 mmol/g)	0.004	0.016	0.250
4	Cu-β (0.4 mmol/g)	0.003	0.020	0.150
5	Zn-β (0.4 mmol/g)	0.003	0.022	0.130
6	Sn-β (0.4 mmol/g)	0.004	0.026	0.154
7	Y-β (0.4 mmol/g)	0.021	0.027	0.778
8	Yb-β (0.4 mmol/g)	0.019	0.025	0.760
9	Y-β (1.2 mmol/g)	0.028	0.032	0.875
10	Yb-β (1.2 mmol/g)	0.025	0.030	0.833
11	Zn-Sn-β (0.4 mmol/g)	0.007	0.058	0.121
12	Pb-Sn-β (0.4 mmol/g)	0.005	0.056	0.089
13	Cr-Sn-β (0.4 mmol/g)	0.018	0.055	0.327

It could also be seen from Table that Y-β and Yb-β zeolites had much more Brønsted acids than other single-transition-metal-β zeolites. As is known, Brønsted acid could dehydrate fructose to produce 5-hydroxymethyl furfural,[48] and the occurrence of side reactions would inevitably reduce the production of lactic acid, the target product. Thus, it is interesting that more lactic acid and less 5-hydroxymethyl furfural were produced in the catalytic systems of this study.

To find out why Y-β and Yb-β zeolites obtained fewer byproducts when they had more Brønsted acid sites, we carried out the 5-hydroxymethyl furfural decomposition experiment under the same hydrothermal conditions as the previous experiment, and the experimental results are shown in Table .

Table 3

Reaction Behaviors of 5-Hydroxymethyl Furfural under Water or Lactic Acid Solutiona

no.	catalyst	conversion (%)
1	Y-β	66
2	Yb-β	68
3	Sn-β	80
4	no catalyst	50
5	lactic acid	78

Nos. 1–4: 50 mg of 5-hydroxymethyl furfural, 10 mL of H2O, 160 mg of catalyst, 190 °C, 3 h; No. 5: 50 mg of 5-hydroxymethyl furfural, 10 mL of lactic acid (5 mg/mL), 190 °C, 3 h.

Comparing the experimental data in Table , it could be seen that 5-hydroxymethyl furfural was only decomposed by 50% under the hydrothermal condition without a catalyst, and the decomposition efficiency was improved after the catalyst was added. The Y-β and Yb-β zeolites showed a poor decomposition effect on 5-hydroxymethyl furfural compared with Sn-β zeolites, whose conversion rate of 5-hydroxymethyl furfural reached 80%. Therefore, in the process of catalyzing glucose to produce lactic acid, accumulation of the byproduct (5-hydroxymethyl furfural) would hinder the side reaction, resulting in more lactic acid and fewer byproducts.

The lactic acid was used as the catalyst to decompose 5-hydroxymethyl furfural (Table , No. 5). A conversion rate of 78% was obtained in this system, indicating that lactic acid as Brønsted acid promoted the decomposition of 5-hydroxymethyl furfural. Under the same reaction conditions, the decompositions effects of Y-β and Yb-β zeolites on 5-hydroxymethyl furfural were 66 and 68%, respectively, and weaker than that of the lactic acid. Compared with the Brønsted acid brought by lactic acid, Brønsted acid in Y-β and Yb-β zeolites had little effect on the experiment. In conclusion, we speculated that during the conversion of glucose to lactic acid, a larger amount of Brønsted acids such as lactic acid produced would promote the progress of the side reaction (dehydration of fructose to 5-hydroxymethyl furfural) and a smaller amount of Brønsted acids such as the catalysts found in the study would hinder the hydrolysis of 5-hydroxymethyl furfural. Based on the reaction pathways proposed in previous reports[32,46] and the reaction effect of Y/Yb-β zeolites in this study, the reaction pathways in the conversion of glucose to lactic acid were speculated as shown in Figure .

Figure 11

Proposed reaction pathways in the conversion of glucose to lactic acid.

In the catalytic reaction, the acidic strength and type of acid sites of the catalyst might affect the reaction steps. In this study for example, the more Lewis acids shown in Y-β and Yb-β zeolites promoted the main reaction to produce more lactic acid, and their unique Brønsted acidity could hinder the decomposition of 5-hydroxymethyl furfural and thus slow down the progression of side reactions.

Conclusions

This study developed a new modified β zeolite (Y-β and Yb-β) for catalyzing glucose to produce lactic acid, and the effects of reaction parameters were mainly investigated. Results showed that the Y-β and Yb-β zeolites obtained the maximum lactic acid yields of 45.2 and 43.2%, respectively, under optimized conditions (metal loading 1.2 mmol/g, 190 °C, 3 h). The reusability of the two catalysts was explored, and the lactic acid yield of more than 35% was maintained after reusing three times. Compared with the Sn-β zeolite or other single-transition-metal-β zeolites, Y-β and Yb-β zeolites showed more Lewis acid sites, resulting in a higher yield of lactic acid. Combining the decomposition experiment of 5-hydroxymethyl furfural with each catalyst and the unique Brønsted acidity in catalysts revealed by characterization, we speculated that a smaller amount of Brønsted acids would hinder the hydrolysis of 5-hydroxymethyl furfural, thereby slowing down the side reaction. Therefore, Y-β and Yb-β could produce more lactic acid and fewer byproducts than those prior catalysts. This study had a certain guiding significance for the industrial production of lactic acid.