Catalytic Conversion of Sugars into Lactic Acid via a RuO_x/MoS₂ Catalyst

Abstract

The catalytic transformation of sugars into lactic acid has shown great potential for the scalable utilization of renewable biomass. Herein, RuO_x/MoS₂ catalysts were synthesized with the assistance of CaO for the one-pot conversion of glucose to lactic acid. Under the reaction conditions of 120 °C and 1 MPa O₂, a 96.6% glucose conversion and a 54.3% lactic acid selectivity were realized in the one-pot catalytic reaction, with relatively high stability after four successive cycles. This catalytic system was also effective for the conversion of many other carbohydrate substrates, such as fructose, xylose and cellulose (selectivity 68.9%, 78.2% and 50.6%, respectively). According to catalyst characterizations and conditional experiments, the highly dispersed RuO_x species on the surface of MoS₂, together with OH⁻, promoted isomerization, retro-aldol condensation, dehydration and hydration reactions, resulting in a relatively high lactic acid yield for sugar conversions.

1. Introduction

As the only abundant renewable carbon resource, the efficient conversion of biomass into fuel and chemicals has attracted significant attention [1,2]. Due to their abundance and wide availability, cellulose and sugars have been used in central research on biomass conversion and converted into a range of oxygen-containing compounds [3,4,5,6]. Among these attractive chemicals, lactic acid (LA) has emerged as one of the most promising due to its fast-growing market [7,8,9]. Currently, LA is widely used in the food and beverage sector as well as the cosmetics and pharmaceutics industries. Additionally, LA is also an important raw material for the production of bioplastics, polylactide in particular, which is regarded as one of the top ten industrially important chemicals by the United States Department of Energy [10,11].

At present, the fermentation of starch is the most common method for LA production, providing more than 90% of LA in market [12,13]. However, the bacterial fermentation method has the disadvantages of multiple treatment steps and low productivity. Alternatively, the catalytic conversion method of LA production from biomass has attracted increasing research interest during recent years thanks to its versatile feedstocks and high reaction efficiency [14]. To alleviate the condensation or decomposition reactions, the catalytic reaction is usually conducted under alcohol medium to produce esters or in water solution with metal salts to stabilize the final LA. Typically, Sn(IV)-beta or Sn(IV) grafted onto mesoporous silica support catalyzed the conversion of carbohydrates in an alcohol medium into a ca. 60–75% yield of alkyl lactate [15,16,17,18,19]. Alternatively, homogeneous metal salts were applied to crack sugars and cellulosic biomass into LA under hydrothermal conditions and yielded > 60% LA from the lignocellulose-based biomass [20,21,22,23,24]. The carbohydrates, including glucose, cellulose and starch, are feasibly transformed into LA in the presence of NaOH and/or Ca(OH)₂ but suffer from low LA selectivity and fierce reaction conditions [25]. Therefore, a number of works have investigated heterogeneous catalysts to promote the conversion of sugars into LA. For instance, Murzin et al. developed a Sn(salen)/octylmethyl imidazolium bromide catalyst and achieved a 58% yield of LA with 68% glucose conversion at 473 K [26]. Zhang et al. reported that the introduction of indium and tin into Beta zeolites is able to promote the transformation of glucose to LA, affording a LA yield of 53% at 443 K [27]. Nakajima et al. synthesized a new type of solid acid-base bifunctional catalyst YNbO₄ through a simple co-precipitation method, which obtained a 36% LA yield at 413 K for 5 h [28]. As for the catalysts with a base assistance, Choudhary et al. prepared a CTAB-capped Cu/MgO catalyst for the conversion of biomass-derived sugars in the presence of NaOH and obtained a 70% yield of LA at a mild temperature of 393 K [6]. Wang et al. employed the polymerizates of imidazole and epichlorohydrin ([IMEP]Cl) as catalysts to convert glucose to LA in water in the presence of 50 mM NaOH and yielded 63% LA with 99% glucose conversion at 373 K [29]. In spite of these impressive results, developing novel catalysts for the conversion of sugars to LA is still a hotly researched topic.

Ru-based catalysts have been widely used in the conversion of biomass-derived chemicals to organic acids due to the unique chemical properties of Ru/RuO_x [30,31]. Additionally, MoS₂ is a promising catalyst or stable support used in hydrothermal conditions [32,33]. Herein, a series of RuO_x/MoS₂ catalysts were synthesized for the conversion of sugars into LA with the assistance of CaO in aqueous solutions. Different catalysts and supports were first screened to yield the best RuO_x/MoS₂ catalyst. The influence of pivotal reaction parameters on the LA yield was investigated. Then, the optimized catalysts were systematically characterized to unveil the role of Ru species in the reaction and the reaction networks was proposed according to conditional experiments.

2. Results and Discussion

2.1. Catalytic Conversion of Glucose via Different Catalysts

In view of the versatile acidic and metallic properties of Ru-based catalysts in biomass conversions [30,34], a variety of 1 wt% Ru catalysts supported on different carriers were screened in glucose conversion for LA production. As shown in

Table 1. Product distribution for glucose conversion via different Ru-based catalysts.

Entry a	Catalysts	Con. (%)	Sel. (%)
			L-Lactic Acid	Glycolic Acid	Glyceric Acid	Gluconic Acid	Fructose
1	1RuOx/AC	98.3	9.3	5.1	6.3	0	32.8
2	1RuOx/TS-1	72.1	17.4	0	0	16.2	50.0
3	1RuOx/MWW	94.5	17.0	5.6	0	25.8	6.7
4	1RuOx/HPA	30.1	12.6	0	0	29.3	0
5	1RuOx/TiO2	94.6	35.7	4.5	0	2.4	30.2
6	1RuOx/BN	73.6	30.4	2.1	0	0	15.5
7	1RuOx/MoS2	64.9	46.3	5.5	14.0	2.3	0

^a Reaction conditions: 20 mL 1% glucose, 0.10 g catalyst, 120 °C, 1.0 MPa O₂, 2 h; con., sel., MWW and BN are abbreviations for conversion, selectivity, zeolite with MWW structure and boron nitride, respectively.

, under a mild condition of 2 h reaction at 120 °C under the pressure of 1.0 MPa O₂, the glucose conversion ranged from 30% to 98%, and the LA selectivity varied greatly with the carbon balance changed within a large range of 40–85%, according to the HPLC, GC-MS and HPLC-MS analysis of the products. Via the activated carbon (AC)- and TS-1-supported RuO_x catalysts (Table 1, Entries 1–2), fructose was the major product, with 30–50% selectivity and high conversions (>72%) of glucose, but LA selectivity was merely ca. 10%. This indicates that these two catalysts have notable activity in glucose isomerization but are much less active for retro-aldol reaction and C-C bond breakage [35]. On the other hand, besides the low yields of LA, the 1RuO_x/MWW and 1RuO_x/HPA catalysts (Table 1, Entries 3–4) exhibited distinguished selectivity (ca. 25%) for gluconic acid formation, reflecting the typical activity for catalytic oxidation usually observed via a noble metal catalyst [36,37]. With respect to the LA formation, the most remarkable performance was observed in the catalysts of 1RuO_x/TiO₂, 1RuO_x/BN and 1RuO_x/MoS₂ (Table 1, Entries 5–7), which provided 30–46% LA selectivity and 65–95% glucose conversion. The carbon balance in the reaction via these RuO_x catalysts changed within the large range of 40–85%, suggesting that certain side reactions often took place. The reactant solution was clear and colorless after the reaction. The HPLC analysis showed that no 5-hydroxylmethylfurfural (a product of fructose dehydration, often serving as a precursor for polymerization to form humins) was detected in the products, indicating the formation of heavier products was not the main cause of the carbon loss. The analysis of the gas phase product showed that CO₂ was generated in the reaction. We speculated that the CO₂ production largely accounted for the loss of carbon balance in the reaction. According to the highest LA selectivity and the high thermal stability of the MoS₂-supported metal catalysts previously observed [38], the RuO_x/MoS₂ system was chosen for further investigation.

Alkaline environments greatly contribute to the catalytic conversion of bio-based compounds to LA [39]. Therefore, two types of frequently used base, i.e., NaOH and CaO, with different alkalinities, were added into the reaction system for glucose conversion. As shown in

Table 2. Effects of different components via catalysts on the glucose conversions.

Entry a.	Catalyst	Con. (%)	Sel. (%)
			L-Lactic Acid	Glycolic Acid	Glyceric Acid	Gluconic Acid	Fructose
1 b	1RuOx/MoS2+ NaOH	75.5	37.9	1.9	2.1	27.7	3.1
2 c	1RuOx/MoS2+ CaO	98.9	50.5	7.5	6.7	6.3	0
3	CaO	100	29.5	20.0	7.1	15.2	2.7
4	MoS2	70.1	9.2	0.6	0	38.3	0
5	MoS2+CaO	100	25.0	16.1	10.7	31.3	4.2
6	1RuOx/CaO	97.4	30.5	7.2	0	0	29.3
7	0.5RuOx/CaO	80.7	50.5	6.7	9.7	15.2	0
8	2RuOx/CaO	96.6	54.3	5.4	4.6	6.3	0
9	5RuOx/CaO	99.2	49.3	10.6	1.1	1.7	30.2

^a Reaction conditions: 20 mL 1% glucose, 0.10 g catalyst, 120 °C, 1.0 MPa O₂, 2 h; con. and sel. are abbreviations for conversion and selectivity, respectively. ^b 0.10 mol/L NaOH. ^c 0.10 g CaO.

, Entries 1 and 2, the addition of the base obviously improved the conversion of glucose, while the selectivity to LA was only increased under the assistance of CaO. According to the reaction results and previous studies [40,41,42], we speculated that CaO provided alkalinity and Ca²⁺ cations for the reaction. The glucose isomerization and the followed retro-aldol reaction were catalyzed by the suitable base, forming C3 intermediates as the precursor of LA. At the same time, the soluble Ca²⁺ combined with part of the generated LA to form more stable calcium lactate to a certain extent, which promoted the forward reaction. Via the 1RuO_x/MoS₂+CaO catalyst, the LA selectivity was improved to 50.5% with a 98.9% glucose conversion. To investigate the role of different sectors in the catalytic conversion, CaO, MoS₂ and MoS₂+CaO were evaluated in the catalytic system, respectively. As shown in Table 2, Entries 3–5, the presence of MoS₂ merely showed a remarkable selectivity to gluconic acid but not to the LA product (38.8% vs. 9.2%). On the other hand, CaO afforded 29.5% yield of LA along with 15.2% gluconic acid formation. Combining the MoS₂+CaO in the reaction simply obtained a compromised performance but without the synergistic effect on LA or gluconic acid production. In another case, Ru was loaded on CaO to detect the collaboration between them. Over the 1RuO_x/CaO catalyst (Table 2, Entry 6), LA selectivity was very similar to that of CaO, but the fructose selectivity was increased by more than 25%, similar to that via 1RuO_x/TiO₂ (Table 1, entry 5). Thus, from these conditional experiments, it can be learned that merely through the RuO_x/MoS₂+CaO catalyst, synergistic effects were observed, which realized a higher yield of LA and glucose conversion simultaneously. Note that the carrier CaO was partially hydrolyzed during the reaction process, which may induce the supported RuO_x species to being dissolved in aqueous solution. In contrast, the MoS₂ is very stable under hydrothermal conditions, providing a potential stable support to load Ru species [43]. Thus, the RuO_x/MoS₂ catalyst was optimized for conversions under different conditions.

The effect of Ru loading on catalytic performance was finally investigated. As shown in Table 2, Entries 7–9, the LA yield was maximized to 52.4% via the 2RuO_x/MoS₂ catalyst. At lower Ru loadings (0.5% and 1%), the main by-product was C₆ acids, indicating the weak activity of catalysts in retro-aldol condensation reactions. In contrast, at a high Ru loading of 5%, glyceric acid and fructose were formed largely due to the hydration of dihydroxyacetone and/or glyceraldehyde and depressed retro-aldol condensation. Thus, a reasonable Ru content, i.e., 2%, should be loaded on MoS₂ to balance the retro-aldol condensation and hydration reactions for high LA yields.

For the optimized 2RuO_x/CaO catalyst, the reaction parameters, including reaction temperature, time, CaO dosage and catalyst stability, were investigated for the conversion of glucose to LA. As shown in Figure 1a, the selectivity to LA displayed a volcano curve as a function of temperature and was maximized to 54.3% at 120 °C with the conversion of glucose maintained at 95% at temperatures between 100 °C and 160 °C. When the reaction temperature was higher than the optimal point of 120 °C, the yield of LA decreased dramatically with a concomitantly increased yield of acetic acid. The reason for this should be attributed to the random decomposition of glucose and the oxidative decarbonylation of LA due to the metastable of glucose and LA at high temperatures. The influence of reaction time is shown in Figure 1b. At the 0.5 h reaction, the glucose conversion reached 93.1% with ca. 30% LA selectivity. Prolonging the reaction time to 2 h led to the increased LA yield, mainly due to the cascade conversion of the key intermediate of fructose to the downstream product. It is worth noting that even further extending the reaction time, the product compositions remained almost unchanged, confirming the good stability of LA in the present reaction system. The reaction atmosphere was also tested, as seen in Figure S1. Although the conversion of glucose to LA did not need the participation of O₂, the presence of O₂ herein encouraged the synthesis of LA in comparison to that in the N₂ atmosphere. The enhanced performance should be related to the catalyst state under the oxidation conditions, e.g., RuO_x species.

As the key catalytic component in this catalytic system, the effect of CaO dosage was investigated and is shown in Figure 1c. As expected, LA formation is very dependent on the amount of CaO, reaching a maximum with a 0.10 g addition of CaO. Concurrently, the undesired glucose oxidation route was greatly depressed by the competition of base-catalyzed glucose isomerization and retro-aldol condensation reaction. Excessive CaO dosage played negative roles in LA production, which should be attributed to the low saturated solubility of Ca(OH)₂ (~0.025 M at room temperature) and uncontrollable basicity. The reusability of RuO_x/MoS₂ catalyst was evaluated, and results are shown in Figure 1d. After four cycles, the glucose conversion slightly decreased from 96% to 76% with a relatively stable LA selectivity of ca. 45%, indicating the relatively high stability of the catalyst. After the second cycle, the key intermediate of fructose emerged, which can be attributed to the coverage of the Ru species with a strongly interacting Ca species.

2.2. Active Sites for Glucose Conversions

To verify the stability of support and the optimized catalyst under oxygen-containing hydrothermal reaction conditions, TGA was carried out to evaluate the decomposition temperature of the MoS₂ and the supported 2RuO_x/MoS₂ in air. As shown in Figure 2a, the first weight loss stage from room temperature to 54 °C corresponds to the desorption of physically adsorbed water. The second and third weight losses distributed at around 390 °C and 720 °C could be attributed to the decomposition of the surficial and complete oxidation of MoS₂, respectively [44]. Figure 2b illustrates that the introduction of the Ru species has no discernible impact on the thermal stability of the sample. Thus, at low reaction temperatures between 100 °C and 160 °C, the catalyst structure will not suffer from oxidative deterioration, which provides solid guarantee for the high stability of catalysts in the reaction.

The XRD patterns of the fresh and used catalysts are displayed in Figure 3a. For the fresh 2RuO_x/MoS₂ catalyst, the diffraction peaks at 14.4°, 32.2°, 39.6°, 49.8°, 58.8° (labeled with clubs), and 37.4° and 53.8° (labeled with spades) were attributed to MoS₂ and CaO, respectively. No characteristic peaks corresponding to RuO_x were detected, indicating the high dispersion of the Ru species. For the used catalyst, new peaks attributed to the CaCO₃ phase appeared, which may be caused by the reaction of CaO with CO₂ generated in the reaction. Raman spectra in Figure 3b further reveal the structure of catalysts. Both MoS₂ and 2RuO_x/MoS₂ showed two characteristic peaks at 409 cm⁻¹ and 383 cm⁻¹, corresponding to the E¹_2g and A_1g characteristic vibration modes of MoS₂-2H phase, respectively [45]. Similar to the XRD patterns, the RuO_x species were not detected, confirming the high dispersion of the Ru species.

To disclose the metal dispersion and morphology of catalysts, the SEM and TEM images of the optimized catalyst are compiled in Figure 4. The carrier MoS₂ showed a multilayer smooth sheet morphology with a thickness of about 100 nm. The highly dispersed of RuO_x species are clearly shown in the high-resolution TEM images of Figure 4c,d. Even up to 2 wt% of the metal loading onto the MoS₂ with a low BET surface area of < 10 m²/g (Figure S2), the average size of RuO_x particles was centered at 0.53 nm. Meanwhile, the elemental mapping in Figure 4e–j reveals that the Ru particles were uniformly distributed on the catalyst surface. This is highly consistent with the results obtained via the Pt/MoS₂ catalysts reported previously, owing to the uniquely strong interaction between MoS₂ and noble metals [38]. The used catalyst was also characterized and is shown in Figure S3. According to the SEM image, a large number of amorphous particles attributed to Ca species were randomly distributed on the layered MoS₂. In addition, the TEM and EDS analyses verified the remarkable structural stability of 2RuO_x/MoS₂. The intrinsic structures of MoS₂ and RuO_x were completely maintained through the catalytic reaction, similar to that of the fresh catalyst. The ICP-OES measurements in

Table 3. Elemental analysis of catalysts and product solution by ICP-OES.

Catalysts	Ru Loading a /%	Mo Amount /%	Leached Amount/% b
			Ru	Mo
2RuOx/MoS2	2.3	60.0	2.4 × 10−5	9.5 × 10−4
2RuOx/MoS2 -used	2.0	56.2	2.9 × 10−6	5.5 × 10−6

^a The Ru loading of the catalysts. ^b The mass ratio of elements leached into product solution.

further highlighted the fact that Ru leaching rarely occurred under hydrothermal reaction conditions. Above all, despite Ru sites being partially covered by Ca species, the 2RuO_x/MoS₂ catalyst was still able to produce a reasonable yield of LA in the subsequent cycling tests thanks to its excellent stability.

The chemical states of Mo and Ru species on the catalysts were investigated by XPS (the whole spectrum is shown in Figure S4). As shown in Figure 5a, for the catalyst of 2RuO_x/MoS₂, three valence states of Mo species were identified, including the dominant Mo⁴⁺ (229.09 eV and 232.28 eV) and a small amount of Mo⁵⁺ (233.45 eV) and Mo⁶⁺ (235.56 eV) [46,47]. Since no MoO₃ related signals were observed by Raman characterizations, it was speculated that the high-valence Mo species was distributed in the bulk phase of MoS₂. Through the catalytic reaction, the proportion of Mo⁶⁺ increased by 8.95% after reaction, which may be caused by the partial oxidation of the carrier under the oxidizing atmosphere. For the Ru spectra in Figure 5b, the signal of Ru 3d overlaps with that of C ls. By means of deconvolution analysis, the Ru⁴⁺ species in the form of RuO₂ (280.93 eV and 286.08 eV) were observed [48]. After the reaction, characteristic peaks assigned to RuO₂ vanished, which should be attributed to the tendentiously deposition of CaCO₃ onto the metal sites of the 2RuO_x/MoS₂ catalyst, agreeing with our hypothesis that the Ru species were covered to decrease the activity, rather than the deactivation of 2RuO_x/MoS₂ catalysts.

2.3. Reaction Networks

Various C5 and C6 sugars, polyols and polyhydroxy acids were employed as feedstocks to evaluate the possible reaction networks via this catalytic system. As shown in

Table 4. Catalytic performance of RuO_x/MoS₂ catalyst in the presence of CaO with different feedstocks ^a.

Entry	Reagent	Con. /%	Sel./%
			L-Lactic Acid	Glycolic Acid	Glyceric Acid	Gluconic Acid	Fructose
1	Glucose	96.6	54.3	5.4	4.6	6.3	0
2	Fructose	97.3	68.9	6.2	5.0	0	0
3	Xylose	100	78.2	17.6	0	0	0
4	Sorbitol	51.8	35.5	27.5	0	0	0
5	Mannitol	100	9.3	8.4	0	0	0
6	Gluconic acid	58.1	12.9	3.2	4.8	0	0
7 b	Cellulose	20.5	50.6	4.8	2.1	0	0
8 c	Glucose	76.2	43.5	2.6	1.9	0	0

^a Reaction conditions: 20 mL 1% reagent, 0.10 g catalyst, 0.10 g CaO, 120 °C, 1.0 MPa O₂, 2 h; ^b 20 mL H₂O, 0.16g reagent, 160 °C, 8 h; ^c 20 mL 5 wt% reagent, 0.50 g catalyst, 0.50 g CaO; con. and sel. are abbreviations for conversion and selectivity, respectively.

, Entries 1–3, the RuO_x/MoS₂ catalyst with the assistance of CaO showed a good catalytic performance in the conversion of sugars, including glucose, fructose and xylose. As expected, because the obstacle of glucose isomerization was eliminated, the feedstock of fructose provided a higher selectivity to LA than glucose (68.9% vs. 54.3%). Interestingly, xylose exhibited the best performance for LA production, possibly due to the fewer side reactions occurring after the C5 reactant being degraded into C3 and C2 moieties. Moreover, it was difficult to effectively convert both polyols and gluconic acid into LA. It is noteworthy that the selectivity of LA reached 50.6% in the 8 h reaction at 160 °C when cellulose was used as the reactant (Table 4, Entry 7). This suggests that the RuO_x/MoS₂+ CaO catalyst has a potential to treat recalcitrant feedstock.

In addition, the activity of the catalyst was tested in glucose conversion at a high concentration (5 wt%). As shown in Table 4, Entry 8, the selectivity of LA was slightly decreased to 43.5%, providing LA productivity of 8.3 g/L/h, which is notably higher than that of the standard of LA production via industrial fermentation (5 g/L/h) [49].

The above results demonstrate again that the base-catalyzed and RuO_x/MoS₂-catalyzed retro-aldol condensation route together contributed to the LA formation. CaO dissolved in water to release OH⁻ and Ca²⁺. OH^- catalyzed the isomerization of glucose to fructose, followed by degradation into dihydroxyacetone and glyceraldehyde through retro-aldol condensation. Afterward, glyceraldehyde was further catalytically dehydrated by the elimination reaction under the actions of OH⁻ to generate the pyruvaldehyde intermediate [40,42]. Finally, obtained pyruvaldehyde can form metal chelates with Ca²⁺ and transform into LA and/or lactate via the 1,2-hydride shift [41]. On the other hand, the Ru species on the RuO_x/MoS₂ in an oxidation state played the crucial role in promoting retro-aldol condensation reactions.

Finally, the reaction pathway is briefly proposed in Scheme 1. Under the mild conditions of 120 °C and 1MPa O₂, the CaO and Ru species catalyzed the isomerization of glucose into fructose. Afterward, the retro-aldol condensation occurred via the RuO_x species with dihydroxyacetone and glyceraldehyde production. Finally, the obtained glyceraldehyde was transformed into LA via dehydration and hydration reactions via the Ru species. Simultaneously, side reactions, such as glucose direct oxidation and the direct conversion of dihydroxyacetone, occurred, resulting in the by-products of gluconic acid, glyceric acid, glycolic acid and acetic acid.

3. Materials and Methods

3.1. Materials

All reagents were purchased commercially without further purification. Glucose (99%) and calcium oxide were purchased from Tianjin Damao Chemical Reagent Co., Ltd. (Tianjin, China). Sodium hydroxide was purchased from Tianjin Yongda Chemical Reagent Co. Ltd. Activated carbon (AC) was purchased from Norit Corporation. Calcium carbonate was purchased from Xilong Science Co., Ltd. (Shantou, China). Standard sample of lactic acid (LA) with 95% purity was purchased from Sigma-Aldrich Co., Ltd. (St. Louis, MO, USA). Ruthenium chloride hydrate, molybdenum disulfide, hydroxyapatite (HPA), titanium dioxide, titanium silicalite-1 (TS-1), Boron nitride (BN) and all other chemicals were analytical reagents and purchased from Shanghai Aladdin Biochemical Technology Co., Ltd. (Shanghai, China).

3.2. Catalysts Preparation

RuO_x/support catalysts were prepared using the incipient wetness impregnation method with MoS₂, AC, TiO₂, TS-1, HPA and BN as supports. Typically, 2 g of support was mixed with 0.26–2.6mL aqueous solutions of RuCl₃·3H₂O (3.792% g/mL) and maintained at 25 °C for 12 h. Then, the sample was dried at 60 °C overnight and calcined at 350 °C for 4 h in air. The theoretical contents of Ru were controlled to remain at 0.5–5 wt%. The cost of preparing 1 g of 1RuO_x/MoS₂ catalyst was approximately CNY 1.1 (USD 0.16).

3.3. Catalysts Characterizations

X-ray powder diffraction (XRD, PANalytical B. V., Almelo, The Netherlands) patterns of catalysts were collected on a PANalytical X’pert Pro-1 equipped with Cu-Kα monochromatic radiation (0.1541 nm), operating at 40 kV and 40 mA. A continuous mode was used in the 2θ range from 10° to 80° at a scanning speed of 5°/min.

Thermogravimetric (TG, TA Instruments, Newcastle, DE, USA) analysis was conducted with SDT Q600 equipment under an air flow of 100 mL/min from 30 to 900 °C at a heating rate of 10 °C/min.

Scanning electron microscopy (SEM, HITACHI, Tokyo, Japan) images were obtained by using a Carl Zeiss Merlin Compact scanning electron microscope.

High-angle annular dark-field scanning-transmission electron microscopy (HAADF-STEM, JEOL, Tokyo, Japan) images and energy-dispersive X-ray spectroscopy (EDS, JEOL, Tokyo, Japan) elemental mapping were obtained by using a JEM-2100F microscope, operated at a voltage of 200 kV.

The contents of Ru and Mo in the synthesized catalysts and the leaching amount in the reaction solution were determined by an inductively coupled plasma optical emission spectrometer (ICP-OES, PerkinElmer, Waltham, MA, USA) on a Perkin Elmer ICP-OES 7300DV.

X-ray photoelectron spectra (XPS Thermo Fisher Scientific, Waltham, MA, USA) were recorded on a Thermo Scientific K-Alpha+ spectrometer with a monochromatic Al-Kα X-ray source as the excitation source. The binding energies were corrected by using the C1s peak at 284.4 eV as a reference.

3.4. Catalytic Conversion of Glucose

The catalytic conversion of glucose was carried out in a high-pressure batch reactor. Typically, 20 mL of 1% glucose aqueous solutions, 0.10 g of catalyst and 0.10 g of CaO were placed into the reactor. Then, the reactor was charged in O₂ five times to remove the air. Finally, the oxygen pressure in the reactor was increased to 1.0 MPa and sealed for the reaction at 120 °C for 2 h.

After reaction, the product was cooled to room temperature and analyzed with a high-performance liquid chromatograph HPLC, Agilent 1200, (Agilent Technologies, Palo Alto, CA, USA) equipped with a Rezex ROA-Orgenic Acid H⁺ column and a differential refractive index detector with 5 mM sulfuric acid aqueous as the mobile phase. In detail, the GC-MS (Agilent Technologies, Palo Alto, CA, USA) and HPLC-MS methods (Agilent Technologies, Palo Alto, CA, USA)were used to identify the products in reaction solutions or identify the products after methyl esterification treatment [31]. Then, the standard samples were used to further confirm and quantitatively analyze the products on the HPLC instrument.

Glucose conversion and product selectivity were calculated using following equations:

$$\mathrm{Conversion}(\%)=(1-\frac{\mathrm{m}\mathrm{o}\mathrm{l}\mathrm{e}\mathrm{s}\mathrm{o}\mathrm{f}\mathrm{u}\mathrm{n}\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{t}\mathrm{e}\mathrm{d}\mathrm{g}\mathrm{l}\mathrm{u}\mathrm{c}\mathrm{o}\mathrm{s}\mathrm{e}}{\mathrm{m}\mathrm{o}\mathrm{l}\mathrm{e}\mathrm{s}\mathrm{o}\mathrm{f}\mathrm{i}\mathrm{n}\mathrm{i}\mathrm{t}\mathrm{i}\mathrm{a}\mathrm{l}\mathrm{g}\mathrm{l}\mathrm{u}\mathrm{c}\mathrm{o}\mathrm{s}\mathrm{e}})\times 100\%$$

(1)

$$\mathrm{Selectivity}(\%)=\frac{\mathrm{m}\mathrm{o}\mathrm{l}\mathrm{e}\mathrm{s}\mathrm{o}\mathrm{f}\mathrm{p}\mathrm{r}\mathrm{o}\mathrm{d}\mathrm{u}\mathrm{c}\mathrm{t}}{\mathrm{t}\mathrm{h}\mathrm{e}\mathrm{o}\mathrm{r}\mathrm{e}\mathrm{t}\mathrm{i}\mathrm{c}\mathrm{a}\mathrm{l}\mathrm{m}\mathrm{o}\mathrm{l}\mathrm{e}\mathrm{s}\mathrm{o}\mathrm{f}\mathrm{p}\mathrm{r}\mathrm{o}\mathrm{d}\mathrm{u}\mathrm{c}\mathrm{t}}\times 100\%$$

(2)

4. Conclusions

A series of RuO_x/MoS₂ catalysts were synthesized for the conversion of sugars into LA with the assistance of CaO. Unlike the conventional base-catalyst system, the 2RuO_x/MoS₂ catalyst exhibited enhanced catalytic activity and stability. Under the low reaction temperature of 120 °C with low alkalinity, the LA selectivity reached 54.3% at a 96.6% glucose conversion, even in the recycle runs. The RuO_x species promoted the isomerization of glucose to fructose, and then the fructose was cracked into the C3 compound of glyceraldehyde via the RuO_x species via the retro-aldol condensation reaction. Finally, the C3 intermediate was further converted into LA via dehydration and hydration reactions. Although the finally yield of LA still needs to be improved, this work still provides valuable information for the development of the novel heterogeneous catalyst for carbohydrates conversions. The future study may focus on how to further enhance the LA selectivity and the reaction efficiency, particularly in the cellulosic feedstock conversion, which involves the catalyst development as well as a kinetic reaction investigation.