Direct Synthesis of Calcium Lactate through the Reaction of Glycerol with Calcium Hydroxide Catalyzed by Bimetallic AuCu/SiO₂ Nanocatalysts

Abstract

Bimetallic AuCu/SiO₂ nanosized catalysts were prepared using the wet chemical reduction technique. From among Au_0.1–1.5Cu₁₀/SiO₂ catalysts, the Au_0.5Cu₁₀/SiO₂ catalyst gave the highest yield of calcium lactate of 87% at a glycerol conversion of 96% when the reaction of glycerol with calcium hydroxide at a mole ratio of calcium hydroxide to glycerol of 0.8:1 was conducted under an anaerobic atmosphere at 200 °C for 2 h. The interactions between metallic Au⁰ and Cu⁰ nanoparticles facilitate calcium lactate formation. The simulation of glycerol consumption rate with an empirical power-function reaction kinetics equation yielded a reaction activation energy of 44.3 kJ∙mol⁻¹, revealing that the catalytic reaction of glycerol with calcium hydroxide to calcium lactate can be conducted by overcoming a mild energy barrier. The synthesis of calcium lactate through the catalytic reaction of glycerol with calcium hydroxide on a bimetallic AuCu/SiO₂ nanosized catalyst under a safe anaerobic atmosphere is an alternative to the conventional calcium lactate production technique through the reaction of expensive lactic acid with calcium hydroxide.

1. Introduction

With the depletion of fossil fuels and the deterioration of the environment by CO₂ emissions, the utilization of green and sustainable biomass and its related high value-added fine chemicals has gained much more attention worldwide. Glycerol, a byproduct in biodiesel production, has been produced at a large scale; around 1/10 of total biodiesel production capacity [1]. In addition to this, glycerol is also available in surfactant production from vegetable oils and animal fats. The worldwide glycerol market was estimated at about USD 615 million in 2022 and is expected to reach USD 954 million by 2029 [2]. The high glycerol production capacity makes glycerol conversion to high-value-added chemicals practical and beneficial.

Glycerol molecules of a polyhydroxy nature are multifunctional and can undergo catalytic reactions, such as hydrogenolysis, oxidation, and dehydration, to produce propanediol [3], glyceric acid, glycolic acid [4], dihydroxyacetone [5], glycerol carbonate [6], lactic acid [7,8,9], and calcium lactate [10]. From among them, calcium lactate can be used as a calcium supplement, antimicrobial agent, and in a variety of food additives in food, medicine, and feed fields.

Calcium lactate is commercially produced through the neutralization reaction of lactic acid with calcium hydroxide. For the raw material lactic acid, numerous studies have disclosed that noble metal catalysts, such as Au [11,12], Pt [13], and AuPt [14], can catalyze glycerol oxidation with gaseous O₂ to lactic acid at reaction temperatures in a range of 60–100 °C, yielding lactic acid selectivities of 16%, 92.3%, 69.3%, and 83% at glycerol conversions of 15%, 82.4%, 100%, and 100%. Although noble metal catalysts can catalyze glycerol oxidation to lactic acid at mild reaction temperatures of around 95 °C, lower lactic acid selectivity at higher glycerol conversion rates and a higher noble metal catalyst cost limit their practical applications in lactic acid production.

Researchers have recently reported that non-precious metals exhibit impressively catalytic activities in hydrothermal glycerol conversion to lactic acid in alkaline aqueous solutions under an anaerobic atmosphere. It was reported that the glycerol conversion and lactic acid selectivity on the Cu/ZIF-8 catalyst were 97.5% and 84.2% when the glycerol conversion was conducted in sodium hydroxide aqueous solutions under an anaerobic N₂ atmosphere at 230 °C for 8 h [15]. When the synthesis of lactic acid through hydrothermal glycerol conversion in sodium hydroxide aqueous solutions was conducted on a 20CuHT catalyst bed at 240 °C, the glycerol conversion and lactic acid yield were around 96.5% and 64%, respectively [16]. It was also reported that when the hydrothermal conversion of glycerol to lactate was conducted in calcium hydroxide aqueous solutions at 230 °C, the lactate selectivity was above 91% at a glycerol conversion of around 93% [10].

Cobalt-derived catalysts were also utilized in the conversion of glycerol to lactic acid in sodium hydroxide aqueous solutions at a higher reaction temperature of 250 °C [17]. At the comparable glycerol conversion of around 53.5%, the selectivities of lactic acid reached 90% for Co₃O₄/CeO₂, 78% for Co₃O₄/ZrO₂, and 68% for Co₃O₄/TiO₂. Furthermore, Ye et al. [18,19] found that CaO and CaO/CuO catalysts could effectively catalyze the conversion of glycerol to lactic acid with a yield of 40.8% at a glycerol conversion of 97.8% and after esterification with methanol, the methyl lactate yield reached 57% at a glycerol conversion of 94%.

Active components and surface acidic/basic performance play important roles in catalytic glycerol conversion to lactic acid [10,15,16,17]. Considering that noble metal catalysts can catalyze glycerol conversion at a mild reaction temperature, whereas copper-based catalysts exhibit higher lactic acid selectivity at a higher temperature, we suggest that combining non-precious Cu with noble Au could produce good catalytic performance in glycerol conversion to lactate in an alkaline aqueous solution under anaerobic atmosphere.

In our present work, SiO₂-nanoparticle-supported bimetallic AuCu nanocatalysts were prepared using a wet chemical reduction technique. The catalytic activity of the bimetallic AuCu/SiO₂ nanocatalyst in the synthesis of calcium lactate through the reaction of glycerol with calcium hydroxide under an anaerobic atmosphere was assessed. And a power-function type reaction kinetics equation for glycerol consumption rate was simulated to assess the impacts of reaction parameters. Our present work reveals that calcium lactate can be effectively synthesized through the direct reaction between biomass glycerol and Ca(OH)₂ catalyzed by the bimetallic AuCu/SiO₂ catalyst, which is a sustainable methodology for the synthesis of calcium lactate instead of the conventional production technique through the reaction of expensive lactic acid with calcium hydroxide.

2. Results and Discussion

2.1. XRD Analysis

The X-ray diffractograms of Au₁/SiO₂, Cu₁₀/SiO₂, and Au_0.1–1.5Cu₁₀/SiO₂ catalysts are illustrated in Figure 1. For all catalysts, wide XRD peaks centering at 2θ = 22.3° are ascribed to that of amorphous SiO₂ nanoparticle support.

The XRD peaks of the Cu₁₀/SiO₂ catalyst centered at 2θ = 43.3, 50.5, and 74.3° are assigned to the (1 1 1), (2 0 0), and (2 2 0) crystal planes of metallic Cu⁰ (JCPDS 04-0836), respectively. For Au_0.1–1.5Cu₁₀/SiO₂ catalysts, their XRD peaks at 2θ = 43.3, 50.5, and 74.2° are also assigned to the (1 1 1), (2 0 0), and (2 2 0) crystal planes of metallic Cu⁰ (JCPDS 04-0836). Calculated with the utilization of the Scherrer equation, crystallite sizes of metallic Cu (1 1 1) in the Au_0.1–1.5Cu₁₀/SiO₂ catalysts decreased from 18.4 to 12.1 nm with the increase in Au content from 0% to 1.5 wt%. The presence of the Au component inhibited the crystal growth of the Cu⁰ component.

The XRD peaks of the Au_1.0/SiO₂ catalyst centered at 2θ = 38.1, 44.4, 64.6, and 77.5° are ascribed to (1 1 1), (2 0 0), (2 2 0), and (3 1 1) crystal planes of face-centered cubic Au⁰ (JCPDS 99-0056), respectively. For the Au_0.1–1.5Cu₁₀/SiO₂ catalysts, at lower Au contents of less than 0.8 wt%, there were no XRD peaks for metallic Au⁰ detected, whereas with larger Au contents of 0.8 wt%, 1.0 wt%, and 1.5 wt%, only XRD peaks ascribed to the (1 1 1) crystal plane of metallic Au⁰ were detected, and XRD peaks centered at 38.3, 38.2, and 38.2°, respectively. As compared to the Au_1.0/SiO₂ catalyst, the positive shift in XRD peaks (1 1 1) of AuCu/SiO₂ catalysts should be caused by interactions between Cu and Au components.

According to the Scherrer equation, the crystallite sizes of Au (1 1 1) of Au_0.8Cu₁₀/SiO₂, Au_1.0Cu₁₀/SiO₂, Au_1.5Cu₁₀/SiO₂, and Au_1.0/SiO₂ catalysts were 6.7, 9.9, 12.2, and 18.1 nm, respectively. The results reveal that Au components in catalysts are at a nanosized scale. For the same Au content, the crystallite size of the Au component in the Au_1.0Cu₁₀/SiO₂ catalyst is obviously less than that in the Au_1.0/SiO₂ catalyst, which could be interpreted as the interactions between Cu and Au components inhibiting the crystal growth of the metallic Au⁰ nanoparticles.

2.2. TEM and HRTEM Analyses

TEM images show that particle sizes of SiO₂ support are in a range of 9–24 nm and center at 13 nm (Figure 2a–d). Average particle sizes and particle size distributions of metallic Cu⁰ and Au⁰ nanoparticles of the Cu₁₀/SiO₂ and Au₁/SiO₂ catalysts are 8.2, 5.3–12.3 and 3.4, 1.6–6.6 nm, respectively (Figure 2a,b). Average particle sizes and particle size distributions of metallic Cu⁰ and Au⁰ nanoparticles of the Au_0.1Cu₁₀/SiO₂ and Au_0.5Cu₁₀/SiO₂ are 8.2, 5.4–11.4; 2.9, 1.4–4.6 and 7.5, 5.1–10.7; and 2.8, 1.1–4.7 nm (Figure 2c,d). TEM and HRTEM images clearly disclose that Cu and Au components are present at a nanosized scale in catalysts.

The lattice fringes of Au₁/SiO₂ and Au_0.1Cu₁₀/SiO₂ catalysts show that gold and copper components are present in metallic states (Figure 2e,f). The small-sized metallic Cu and Au nanoparticles were anchored and dispersed on the surfaces of SiO₂ nanoparticles. The adjacent Cu and Au nanoparticles may have interactions.

2.3. XPS Analysis

Figure 3a shows that the binding energies of Si2p of SiO₂ support, Au_1.0/SiO₂, Cu₁₀/SiO₂, Au_0.1Cu₁₀/SiO₂, and Au_0.5Cu₁₀/SiO₂ catalysts decrease from 103.7 to 103.2 eV with the introduction of Cu and Au components. The obvious shift of Si2p with the introduction of Cu and Au components reveals that there are interactions between SiO₂ support and metallic Cu⁰ and Au⁰ components. Considering that the electronegativity of Si⁴⁺ (>14.15) is much larger than that of Au atoms (2.54) and Cu atoms (1.9) [20,21], the negative shift in Si2p indicates that electrons of Au and Cu components probably transfer to silicon (Si⁴⁺) components.

For the Cu₁₀/SiO₂, Au_0.1Cu₁₀/SiO₂, and Au_0.5Cu₁₀/SiO₂ catalysts, there were no characteristic satellite Cu2p peaks around 944 eV for Cu²⁺ observed (Figure 3b), which indicates that copper components in the above-mentioned catalysts are present at metallic Cu⁰ and/or Cu⁺ states [22]. Binding energies of Cu2p_1/2 and Cu2p_3/2 of the Cu₁₀/SiO₂, Au_0.1Cu₁₀/SiO₂, and Au_0.5Cu₁₀/SiO₂ catalysts were 952.7, 932.6; 952.4, 932.5; 952.2, and 932.2 eV. The addition of the Au component makes the Cu2p peaks shift negatively.

Binding energies of Au4f_5/2 and Au4f_7/2 of the Au_1.0/SiO₂, Au_0.1Cu₁₀/SiO₂, and Au_0.5Cu₁₀/SiO₂ catalysts were 87.2, 83.6; 88.2, 84.1; and 87.6, 83.8 eV (Figure 3c). The presence of metallic Cu⁰ in catalysts makes the Au4f shift positively. To reveal surface chemical valences of Au⁰ nanoparticles in Au_1.0/SiO₂, Au_0.1Cu₁₀/SiO₂, and Au_0.5Cu₁₀/SiO₂ catalysts, their XPS peaks were deconvoluted by utilizing the Gaussian–Lorentzian bands with peak positions at ca.87.2 and 83.6 eV for Au4f_5/2 and Au4f_7/2 of Au⁰ and ca. 88.2 and 84.2 eV for Au4f_5/2 and Au4f_7/2 of Au^δ+ [23,24]. The deconvoluting results show that the area ratios of Au^δ+/Au⁰ for Au4f_5/2 and Au4f_7/2 of Au_1.0/SiO₂, Au_0.1Cu₁₀/SiO₂, and Au_0.5Cu₁₀/SiO₂ catalysts were 0.03:1, 0.05:1; 9.41:1, 4.61:1; and 0.62:1, 0.55:1, respectively. The shifts in Au and Cu peaks reveal that there are interactions between Au⁰ and Cu⁰ nanoparticles in AuCu/SiO₂ catalysts through the electron transfer from the Au component to the Cu component, which is consistent with that observed by XRD analysis. Interestingly, it was found that the shift extent of the Au4f of Au_0.1Cu₁₀/SiO₂ was larger than that of Au_0.5Cu₁₀/SiO₂. It could be explained because with a relatively lower Au content, Au⁰ nanoparticles could efficiently interact with Cu⁰ nanoparticles, whereas with a higher Au content, some of the Au⁰ nanoparticles are probably apart from the surfaces of the Cu⁰ nanoparticles.

To reveal the surface chemical valences of Cu⁰ nanoparticles in Cu₁₀/SiO₂, Au_0.1Cu₁₀/SiO₂, and Au_0.5Cu₁₀/SiO₂ catalysts, their XAES peaks were deconvoluted by utilizing the Gaussian–Lorentzian bands with peak positions at 916 eV for Cu⁺ and 918.7 eV for metallic Cu⁰, and 914.3 eV for other orbital electrons [25]. The deconvoluting results show that the area ratios of Cu⁰/Cu⁺ of Cu₁₀/SiO₂, Au_0.1Cu₁₀/SiO₂, and Au_0.5Cu₁₀/SiO₂ catalysts were 0.16:1, 0.26:1, and 0.67:1, respectively (Figure 3d). The electronegativities of Cu²⁺, Cu⁺, and Au⁰ were 10.28, 5.04, and 2.54, respectively [20,21]. The results disclose that the Au⁰ nanoparticles present in AuCu/SiO₂ catalysts could transfer electrons to Cu²⁺ and even Cu⁺ components, which prevents the oxidation of Cu⁰ nanoparticles to some extent.

On the other hand, the atomic ratios of Au/Cu in Au_0.1Cu₁₀/SiO₂ and Au_0.5Cu₁₀/SiO₂ catalysts were 0.32:100 and 1.61:100, respectively. The XPS analysis showed that the atomic ratios of Au/Cu for Au_0.1Cu₁₀/SiO₂ and Au_0.5Cu₁₀/SiO₂ catalysts were 0.36:100 and 8.64:100. These results disclose that the surface compositions of the Au component are larger than the bulk ones. It is reasonable to suggest that metallic Au⁰ nanoparticles could anchor at surfaces of metallic Cu⁰ nanoparticles.

2.4. N₂ Physisorption

As illustrated in Figure 4, nitrogen adsorption/desorption curves of representative samples show that the N₂ adsorption/desorption isotherms belong to the III type. The SiO₂ nanoparticle support has a large surface area of up to 221.2 m²∙g⁻¹ (

Table 1. Structural properties of the SiO₂, Cu/SiO₂, Au/SiO₂, and AuCu/SiO₂ catalysts.

Catalysts	SBET (m2∙g−1)	Vpore (cm3∙g−1)	Dpore (nm)
SiO2	221.4	0.57	11.10
Cu3/SiO2	151.3	1.14	28.34
Cu5/SiO2	141.0	1.07	26.67
Cu10/SiO2	132.4	0.71	21.06
Au1/SiO2	104.2	0.75	28.16
Au0.1Cu10/SiO2	151.9	0.73	18.83
Au0.3Cu10/SiO2	212.2	0.73	15.05
Au0.5Cu10/SiO2	177.8	0.65	14.07
Au0.8Cu10/SiO2	160.5	0.60	12.76
Au1.0Cu10/SiO2	105.0	0.62	11.96
Au1.5Cu10/SiO2	104.6	0.60	11.91

). For the Cu_3–10/SiO₂ catalysts, their surface areas decreased from 151.3 to 132.4 m²∙g⁻¹ with the increase in Cu content from 3 wt% to 10 wt%. Their pore volumes and pore sizes were larger than those of the SiO₂ support, respectively.

When the second metal, Au, was introduced, the surface areas of the Au_0.1–1.5Cu₁₀/SiO₂ catalysts increased to 212.2 m²∙g⁻¹ with the increase in Au content to 0.3 wt% and then decreased to 104.6 m²∙g⁻¹ with the further increase in the Au content to 1.5 wt%. Both pore volume and pore size decreased with the increase in Au content in Au_0.1–1.5Cu₁₀/SiO₂ catalysts. The small-sized Au nanoparticles easily choked catalyst pores, which lead to a decrease in the specific surface area of the Au₁/SiO₂ catalyst. When Cu nanoparticles were present in bimetallic CuAu/SiO₂ catalysts, large-sized Cu nanoparticles reduced the choking to some extent, causing an increase in the surface area. However, the choking effect by Au nanoparticles increased with a higher content of Au nanoparticles, which decreased the specific surface area of the catalyst.

The Cu_3–10/SiO₂, Au₁/SiO₂, and Au_0.1–1.5Cu₁₀/SiO₂ catalysts have large surface areas and pore sizes, which could supply a large amount of catalytic active sites and diminish mass transfer resistance in pores.

2.5. Catalytically Hydrothermal Conversion of Glycerol with Calcium Hydroxide to Calcium Lactate

2.5.1. Impact of Catalyst Composition

When the reaction of glycerol with calcium hydroxide at an initial mole ratio of calcium hydroxide to glycerol of 0.8:1 was catalyzed on Cu_3–10/SiO₂, Au₁/SiO₂, and Au_0.1–1.5Cu₁₀/SiO₂ catalysts in aqueous solutions at 200 °C for 2 h, for the Cu_3–10/SiO₂ catalysts, the glycerol conversion, calcium lactate yield, and carbon balance increased with the increase in Cu content. From among Cu_3–10/SiO₂ catalysts, the Cu₁₀/SiO₂ catalyst gave the maximum glycerol conversion of 67% and a calcium lactate yield of 58% (

Table 2. Catalytic performance of the SiO₂, Cu/SiO₂, Au/SiO₂, and AuCu/SiO₂ catalysts.

Catalysts	Glycerol Conversion (%)	Product Selectivity (%)				Yield (%)	Carbon Balance (%)
		OA	FA	LA	AA	LA
Cu3/SiO2	22	0.2	5	60	8	13	73.2
Cu5/SiO2	44	0.1	1	77	2	34	80.1
Cu10/SiO2	67	0.2	2	87	1	58	90.2
Au1/SiO2	24	0.3	0.2	4	1	1	5.5
Au0.1Cu10/SiO2	81	0.2	2	94	3	76	99.2
Au0.3Cu10/SiO2	90	0.2	2	93	3	84	98.2
Au0.5Cu10/SiO2	96	0.2	2	91	3	87	96.2
Au0.8Cu10/SiO2	97	0.2	2	89	4	86	95.2
Au1.0Cu10/SiO2	99	0.2	1	87	3	86	91.2
Au1.5Cu10/SiO2	100	0.2	2	84	4	84	90.2

Glycerol aqueous solution, 2.0 mol∙L⁻¹; initial mole ratio of calcium hydroxide to glycerol at 0.8:1; catalyst, 9.2 g∙L⁻¹; reaction time, 2.0 h; reaction temperature, 200 °C. OA, oxalate; FA, formate; LA, lactate; AA, acetate.

). Larger metallic Cu content facilitates calcium lactate formation and inhibits byproduct formation.

The monometallic Au₁/SiO₂ catalyst exhibits poor catalytic performance in the reaction of glycerol with calcium hydroxide to calcium lactate, which leads to a much lower carbon balance value of 6%. This phenomenon is worthy of further investigation when metallic Au is solely used as the active component.

When metallic Au⁰ was introduced into the Cu₁₀/SiO₂ catalyst, glycerol conversions increased with the increase in Au content in bimetallic Au_0.1–1.5Cu₁₀/SiO₂ catalysts. When Au contents were 1 wt% and 1.5 wt%, the glycerol conversions were above 99%. At a Au content of 0.5–1.0 wt%, the calcium lactate yields were above 86% with glycerol conversions ranging from 96% to 99%. As compared to the catalytic activity of the Cu_3–10/SiO₂ catalysts, the bimetallic AuCu/SiO₂ catalyst exhibited high catalytic performance for calcium lactate synthesis through the hydrothermal reaction, starting from glycerol and calcium hydroxide at the reaction temperature of 200 °C.

It has been reported that small-sized CuO particles facilitate the conversion of glycerol to lactic acid [26]. However, when Cu-ZnO@C catalysts with metallic Cu particle sizes of 36.9, 67.8, and 93.6 nm catalyzed glycerol conversion to lactic acid, the Cu-ZnO@C catalyst with a metallic Cu particle size of 67.8 nm gave the highest lactic acid yield [27]. As compared to the Cu(16)/ZrO₂ catalyst with a metallic Cu crystallite size of 44.5 nm, the Cu(16)HAP catalyst with a metallic Cu crystallite size of 33.8 nm exhibited a high glycerol conversion of 99% and a high lactic acid selectivity of 87% after reacting at 230 °C for 8 h [28]. Liu et al. [29] reported that the metallic Cu nanoparticle with a mean particle size of 11 nm exhibited the highest catalytic activity when Cu/SiO₂ catalysts with mean metallic Cu particle sizes of 4–38 nm catalyzed xylitol hydrogenolysis. Generally, both the metallic Cu particle size and support type co-affect catalytic reactions.

In our present research, although the Cu₁₀/SiO₂, Au_0.1Cu₁₀/SiO₂, and Au_0.5Cu₁₀/SiO₂ catalysts had comparable metallic Cu nanoparticle sizes (around 8.0 nm), glycerol conversions on the Au_0.1Cu₁₀/SiO₂ and Au_0.5Cu₁₀/SiO₂ catalysts were 1.21 and 1.43 times those on the Cu₁₀/SiO₂ catalyst. The result clearly indicates that the interaction between Au⁰ and Cu⁰ nanoparticles improves the catalytic activity.

2.5.2. Impact of Reaction Temperature

Catalytic activities of the Au_0.1–1.5Cu₁₀/SiO₂ catalysts in the reaction of glycerol with calcium hydroxide at an initial mole ratio of calcium hydroxide to glycerol of 0.8:1 to calcium lactate were investigated at 180, 200, and 230 °C for 2 h (Figure 5). When the Au content in AuCu/SiO₂ catalysts was above 0.3 wt%, the glycerol conversion was more than 95% at the temperature of 200 °C. As the temperature was raised to 230 °C, on all the Au_0.1–1.5Cu₁₀/SiO₂ catalysts, glycerol was almost completely converted (Figure 5a).

When bimetallic AuCu/SiO₂ catalysts with Au content of 0.1–0.5 wt% catalyzed the reaction at 180 and 200 °C, the calcium lactate selectivity was above 90% (Figure 5b). When bimetallic AuCu/SiO₂ catalysts with Au content of 0.8–1.5 wt% catalyzed the reaction at 180–230 °C, the calcium lactate selectivity was lower than 90%. The Au_0.5Cu₁₀/SiO₂ catalyst gave the highest calcium yield at the reaction temperature of 200 °C.

At the reaction temperatures of 180–230 °C on the Au_0.1–1.5Cu₁₀/SiO₂ catalysts, the selectivities of formate, acetate, and oxalate were below 2%, 6%, and 1%, respectively (Figure 5c–e). The Au_mCu₁₀/SiO₂ catalysts with a larger Au content gave slightly higher byproduct selectivity than those with a smaller Au content.

The catalysis results show that at a Au content of 0.5 wt%, the Au_0.5Cu₁₀/SiO₂ catalyst is more profitable for calcium lactate formation through the catalytic reaction of glycerol with calcium hydroxide. As certified by the XPS analysis, at a Au content of 0.5 wt% or less, Au nanoparticles could efficiently interact with Cu nanoparticles, which improves the reaction of glycerol with calcium hydroxide to calcium lactate. At a larger Au content, isolated Au nanoparticles favorably lead to byproduct formation.

2.5.3. Impact of Calcium Hydroxide Content

When the reaction of glycerol with calcium hydroxide at various mole ratios was catalyzed by the Au_0.5Cu₁₀/SiO₂ catalyst at 200 °C for 0.5–4 h, the glycerol conversion remarkably increased, whereas the calcium lactate selectivity slightly decreased with the increase in the initial mole ratio of calcium hydroxide to glycerol and the extension of the reaction time (Figure 6a,b). When the initial mole ratios of calcium hydroxide to glycerol were 0.6:1 and 0.8:1 at the reaction time of 2 h, glycerol conversions were 94% and 96%, respectively (Figure 6a). Selectivities of calcium lactate arrived at 91% (Figure 6b). Selectivities of calcium formate, acetate, and oxalate slightly increased with the increase in the initial mole ratio of calcium hydroxide to glycerol, but slightly decreased with the extension of the reaction time (Figure 6c–e).

Reaction routes in the catalytic reaction of glycerol with calcium hydroxide are illustrated in Scheme 1 according to the experimental results. The bimetallic Au_0.5Cu₁₀/SiO₂ catalyst catalyzed the dehydrogenation of glycerol to glyceraldehyde [10,30,31], which is a slower reaction process because glyceraldehyde could not be identified under our experimental conditions. Glyceraldehyde could be catalytically dehydrated to form pyruvaldehyde by OH⁻ anions and the resultant pyruvaldehyde can be converted to calcium lactate [10,31,32]. The intermediates, glyceraldehyde and pyruvaldehyde, could also not be identified in our experiments, which indicates that the dehydration of glyceraldehyde and conversion of pyruvaldehyde are rapid reactions. The formation of oxalate, formate, and acetate could be through the decomposition of the intermediate and lactate catalyzed by OH⁻ anions [10]. The decrease in contents of oxalate, formate, and acetate with an extended reaction time is probably caused by their decomposition in an alkaline atmosphere.

2.5.4. Impacts of Glycerol Concentration and Catalyst Content

At initial glycerol concentrations of 0.5–2.0 mol∙L⁻¹ and a mole ratio of calcium hydroxide to glycerol of 0.6:1, after reacting at 200 °C for 2 h on the Au_0.5Cu/SiO₂ catalyst, the glycerol conversions were above 94% at a calcium lactate selectivity of larger than 91% (Figure 7). When the glycerol concentration was further increased to 3.0 mol∙L⁻¹, the glycerol conversion and calcium lactate selectivity remarkably declined to 70% and 63%, respectively. We suggest that the optimal concentration of glycerol is 2 mol∙L⁻¹ for obtaining a higher calcium lactate yield under our present experimental conditions.

When the reaction of glycerol (2 mol∙L⁻¹) with calcium hydroxide at an initial mole ratio of calcium hydroxide to glycerol of 0.6:1 was catalyzed by the Au_0.5Cu₁₀/SiO₂ catalyst with catalyst contents of 1.84–9.2 g∙L⁻¹ at 200 °C for 2 h, glycerol conversion and calcium lactate selectivity were obviously improved with the increase in catalyst content (Figure 8). With a catalyst content of 9.2 g∙L⁻¹, the glycerol conversion reached 94% at a calcium lactate selectivity of 92%. Obviously, a high catalyst content is profitable for the catalytic reaction of glycerol with calcium hydroxide to calcium lactate.

2.6. Reaction Kinetics Simulation

To ascertain the impacts of the temperature, glycerol concentration, calcium hydroxide concentration, and catalyst content on the glycerol conversion rate, the reaction rate of glycerol consumption was simulated by utilizing an empirical power-function type reaction kinetics equation.

$$-r_A=-\frac{{dn}_A}{Vdt}=k{C}_A^a{C}_B^b{C}_{cat}^c$$

(1)

$$k=Aexp(-E_a/(RT\left)\right)$$

(2)

where k is the rate constant. a, b, and c are the reaction orders for glycerol concentration (C_A, mol∙L⁻¹), calcium hydroxide concentration (C_B, g∙L⁻¹), and catalyst concentration (C_cat, g∙L⁻¹), respectively.

Although it is difficult to dissolve Ca(OH)₂ in water, the produced lactate anion could rapidly react with Ca²⁺ to form easily soluble calcium lactate. The concentration of Ca(OH)₂ is the whole mass concentration because the amount of Ca(OH)₂ significantly affects the reaction. The catalyst concentration is the whole mass concentration.

Combing Equations (1) and (2) and taking the natural logarithm on both sides of the resultant equation yields Equation (3) in linear form.

$$\ln \left(-r_A\right)=lnA-Ea/RT+aln{C}_A+bln{C}_B+cln{C}_{cat}$$

(3)

Initial conversion rates of glycerol for the first 0.5 h under various reaction conditions were used for the reaction kinetics simulation by the linear regression method (Table S1). The simulated reaction kinetics equation with a correlation coefficient of 0.97 is listed as Equation (4). The higher correlation coefficient indicates that the reaction kinetic equation is well simulated using the reaction data under various reaction conditions.

$${-r}_A=-{dn}_A/Vdt=247.87\exp (-\frac{\mathrm{44,305.3}}{RT})C_A^{0.23}{C}_B^{1.04}{C}_{cat}^{0.7}$$

(4)

The simulated reaction activation energy (Ea) is 44.3 kJ∙mol⁻¹, which implies that glycerol conversion can be conducted by overcoming a mild energy barrier. According to the reaction orders, the impacts of concentrations of glycerol, calcium hydroxide, and catalyst on glycerol conversion rate are in the order of Ca(OH)₂ > catalyst > glycerol.

Researchers reported that when the oxidation of glycerol to lactic acid was catalyzed by a CuO/C catalyst in a NaOH aqueous solution, the activation energy was 134.39 kJ·mol⁻¹ [26]. The reaction activation energies for glycerol conversion to lactic acid on bimetallic CuAu₁, CuAu₂, CuAu₃, and CuAu₄ catalysts were 64.0, 53.4, 46.8, and 36.9 kJ∙mol⁻¹, respectively [33]. Furthermore, Hausoul et al. [34] reported that in the hydrogenolysis of glycerol on the Pt/C catalyst, alkaline earth metal hydroxides improved the reaction. Therefore, we suggest that bimetallic CuAu and Ca(OH)₂ facilitate glycerol conversion, which leads to a lower activation energy.

3. Experimental

3.1. Materials

Glycerol (99%), acetic acid (99.5%), oxalic acid (99.5%), lactic acid (85%), Cu(NO)₃∙3H₂O (99%), HAuCl₄∙4H₂O (Au, 47.8 wt%), hydrazine hydrate (85%), sodium hydroxide (NaOH, 96%), phosphoric acid (H₃PO₄, 85%), concentrated hydrochloric acid (HCl, 37%), disodium hydrogen phosphate (99%), n-butanol (99.5%), and methanol (HPLC, 99.8%) were supplied by the Dewei Chem. Reagent Co., Zhenjiang, China. Formic acid (88%) was supplied by the Nanjing Chem. Reagent Co., Nanjing, China. Silicon dioxide (average particle size, 13 nm) was supplied by the Hailong Nuclear Mater. Technol. Co., Zhenjiang, China.

3.2. Preparation of AuCu/SiO₂ Nanocatalysts

Bimetallic AuCu/SiO₂ nanocatalysts were prepared using the wet chemical reduction technique. First, 9.0 g of silicon dioxide was added to 40 mL of Cu(NO)₃∙3H₂O (0.4 mol∙L⁻¹) aqueous solution. After the suspension was stirred at 40 °C for 1 h, a NaOH (4 wt%) aqueous solution was added dropwise to adjust the pH to 11.5. When the above-mentioned suspension was continuously stirred for 1 h, a prescribed volume of HAuCl₄∙4H₂O (1 wt%) aqueous solution was slowly dropped under stirring for 1 h and the pH value was maintained at 11.5. To reduce active components, hydrazine hydrate with a hydrazine hydrate/total metal ion molar ratio of 10:1 was added dropwise at 0.5 h. After reducing for 4 h, the resultant AuCu/SiO₂ catalyst samples were separated by centrifugation at a rotation rate of 10,000 rpm for 15 min and washed with pure water to eliminate sodium salts. The residual concentrations of gold and copper components in the water phase were negligibly detected by ICP measurement, which indicates that gold and copper components could be completely loaded onto SiO₂ support. The as-prepared AuCu/SiO₂ nanocatalysts were kept in anhydrous ethanol to prevent oxidation by oxygen from the atmosphere. The catalysts were denoted as Au_mCu_n/SiO₂ and m and n were the mass percentages of Au and Cu components in catalysts.

3.3. Characterization

Crystal structures and morphologies of the as-prepared AuCu/SiO₂ nanocatalysts were examined on a powder X-ray diffractometer (D8 super speed Bruker-AEX Co. (Billerica, MA, USA), Cu Kα radiation at λ of 1.54056 Å) and a transmission electron microscope (JEM-2100, JEOL Ltd., Tokyo, Japan, 200 kV). Particle sizes of metallic Au⁰ and Cu⁰ nanoparticles on silicon dioxide nanoparticle supports were measured by combining both TEM and HRTEM analyses. The particle sizes were directly measured from TEM images of at least 50 individual particles.

X-ray photoelectron spectra of the AuCu/SiO₂ nanocatalysts were acquired on an ESCALAB 250 spectrometer (Thermo Fisher Scientific, Waltham, MA, USA). Binding energies of Au, Cu, and Si components in catalysts were calibrated by referencing the C1s peak at 284.8 eV of contaminated carbon.

To determine surface areas and pore sizes of the AuCu/SiO₂ nanocatalysts, their N₂ adsorption–desorption isotherms were acquired by utilizing a NOVA 2000e physical adsorption apparatus at −196 °C (Quantachrome, Boynton Beach, FL, USA).

3.4. Catalysis

The synthesis of calcium lactate was catalyzed by the as-prepared AuCu/SiO₂ nanocatalysts through a glycerol dehydrogenation reaction in a calcium hydroxide aqueous solution under a nitrogen atmosphere. The prescribed quantities of glycerol, calcium hydroxide, and catalysts were added to deionized water to obtain 100 mL of reaction mixture in a 1 L steel stainless autoclave. After a pure nitrogen (99.999%) stream was introduced into the reactor to replace air inside for at least 0.25 h, reaction temperatures were raised to prescribed values by heating the autoclave with the use of a 1200 W electric oven outside. As the reaction temperatures arrived at prescribed values, stirring started at 500 rpm, and the timing of the reaction was started.

After reacting for a prescribed time, the spent AuCu/SiO₂ catalyst was filtrated from the reaction mixture at 80 °C. The filtrate was acidified with concentrated hydrochloric acid to a pH of around 1.0 and analyzed using the high-performance liquid chromatography(Agilent 1290 Infinity II, Agilent Technologies, Santa Clara, CA, USA), utilizing a reversed-phase C18 column (Wonda Cr act ODS-2, 5 μm, 4.6 × 250 nm) at 35 °C. H₃PO₄/Na₂HPO₄ (20 mmol∙L⁻¹ Na₂HPO₄, acidified by H₃PO₄ to pH = 2.5) aqueous solution and methanol (v:v = 90:10) were mixed and utilized as the mobile phase with a volumetric flowing rate at 0.7 mL∙min⁻¹. Products were identified by comparison with the retention times of standard samples, and their compositions were quantified by an external standard method. The quantity of unreacted glycerol was quantitatively analyzed by gas chromatography (Agilent Technologies 7890A, PEG-20M, Agilent Technologies, Santa Clara, CA, USA) with an FID detector using n-butanol as an internal standard. The calculation of product selectivity was based on the carbon number of the product molecules.

4. Conclusions

SiO₂ nanoparticle-supported bimetallic AuCu (AuCu/SiO₂) nanocatalysts were prepared using the wet chemical reduction technique. The metallic Au⁰ and Cu⁰ nanoparticles in AuCu/SiO₂ catalysts experienced interactions by electron transfer from metallic Au to metallic Cu nanoparticles.

From among the Au_0.1–1.5Cu₁₀/SiO₂ catalysts, the Au_0.5Cu₁₀/SiO₂ and Au_0.8Cu₁₀/SiO₂ catalysts gave high calcium lactate yields of 87% and 86% at glycerol conversion rates of 96% and 97% when the reaction of glycerol with calcium hydroxide at an initial mole ratio of calcium hydroxide to glycerol of 0.8:1 was conducted under anaerobic atmosphere at 200 °C for 2 h. The interaction between Cu and Au nanoparticles improved the synthesis of calcium lactate through the direct reaction between glycerol and Ca(OH)₂ under an anaerobic condition.

The consumption rate of glycerol was well simulated with the empirical power-function type reaction kinetics, which gave the reaction activation energy (Ea) of 44.3 kJ∙mol⁻¹ and the reaction orders of 0.23 for glycerol concentration, 1.04 for Ca(OH)₂ concentration, and 0.7 for the Au_0.5Cu₁₀/SiO₂ catalyst. The simulated reaction kinetics show that the synthesis of calcium lactate through the catalytic reaction of glycerol with calcium hydroxide could be efficiently carried out by overcoming a mild energy barrier.

The present research shows that the synthesis of calcium lactate can be efficiently and directly conducted through the catalytic reaction of glycerol with calcium hydroxide on a bimetallic AuCu/SiO₂ nanosized catalyst under a safe anaerobic atmosphere. This new methodology could replace the conventional calcium lactate production technique through the neutralization reaction between expensive lactic acid and Ca(OH)₂.