Selective Hydrogenation of 5-Acetoxymethylfurfural over Cu-Based Catalysts in a Flow Reactor: Effect of Cu-Al Layered Double Hydroxides Synthesis Conditions on Catalytic Properties

Abstract

Cu-containing layered double hydroxides (LDHs) were synthesized by a co-precipitation method at different reaction conditions, such as aging time, pH, precipitation rate and synthesis temperature. The effect of these parameters on the structure and chemical composition of the catalysts were investigated using a set of physical methods, including thermogravimetric analysis (TGA), X-ray diffraction (XRD), H₂-TPR and in situ X-ray photoelectron spectroscopy (XPS). It allowed for checking of the reducibility of the samples. 5-Acetoxymethylfurfural was catalytically hydrogenated to 5-(acetoxymethyl)-2-furanmethanol (AMFM) over Cu-containing catalysts synthesized from layered double hydroxides so as to investigate its catalytic properties in flow reaction. It was shown that synthesis pH decreasing from 10 to 8 resulted in rise of AMF conversion that coincided with the higher surface Cu/Al ratio obtained by XPS. Preferable aging time of LDH materials for obtaining the most active catalyst was 2 h, an amount of time that favored the production of the catalyst with high surface Cu/Al ratio up to 0.38. Under optimized reaction conditions, the AMFM yield was 98%. Finally, a synthesis strategy for the preparation of highly efficient Cu-based hydrogenation catalyst with optimized characteristics is suggested.

1. Introduction

5-Hydroxymethylfurfural (HMF), derived from available non-edible lignocellulosic biomass, is an attractive starting material for the synthesis of various chemicals [1,2]. However, HMF’s low stability leads to problems in its production and storage [3,4]. 5-Acetoxymethylfurfural (AMF), which can be prepared from carbohydrates without isolating HMF, is less reactive, more stable and easily extracted from the aqueous reaction mixture. In this connection, AMF is an appropriate substitute for HMF [5,6,7,8]. In recent years, increased attention has been paid to the development of none-noble metal catalysts for the synthesis of chemicals [9,10,11,12,13,14,15]. Among them, Cu-based catalysts have shown promising properties in hydrogenation reactions, such as methanol synthesis [16,17], nitroarene reduction [18,19,20], reductive amination of aldehydes or ketones [20,21], selective hydrogenation of formyl group in furanic aldehydes [22,23,24,25] and so on. Layered double hydroxides (LDH) were disclosed as the promising catalyst precursors, that give dispersed metal oxide or metal nanoparticles after calcination or after calcination and reduction, respectively [11,23,24,25,26,27,28]. LDHs are a series of inorganic lamellar compounds, that are usually prepared by co-precipitation of Me²⁺/Me³⁺ nitrate salts with the base solution, containing sodium carbonate [23,24,25,26,27,28]. Change in precipitation (Me²⁺/Me³⁺ and Na₂CO₃/NaOH ratios, pH, precipitation and aging temperature) and consequent calcination and reduction conditions influence the physico-chemical and catalytic properties of Cu-based catalysts produced from LDH [23,24,25,26,27,28,29,30]. It was shown that the use of Na₂CO₃ alone as a precipitating agent prevents complete removal of sodium ions from Cu-Al-LDH by washing, so Na₂CO₃/NaOH mixture was more preferable for LDH formation [29,30]. The calcination temperature of 650 °C should be used to remove carbonate ions from precipitates, as the presence of carbonate ions in the Cu-Al-LDH structure retards the reduction of Cu precursor [25,31].

Previously, the effects of Cu²⁺/Al³⁺ ratio and calcination temperature were studied on the performance of CuAl-LDH in selective flow hydrogenation of AMF to 5-(acetoxymethyl)-2-furanmethanol (AMFM) [25]. CuAl-LDH samples were precipitated with NaOH/Na₂CO₃ solution at 70 °C and pH 9.0 ± 0.1, with subsequent aging during 4 h, washing, and drying at 110 °C. It was shown that decrease of the Cu²⁺/Al³⁺ ratio from 2 to 1 and increase of the calcination temperature of CuAl-LDH from 450 to 650 °C improved the conversion of AMF, providing AMFM selectivity of 98% and complete AMF conversion at 90 °C and 10 bar.

The main goal of the presented work was the comparison of catalytic properties of a series of Cu-Al-LDH samples, differing in preparation conditions, in AMF hydrogenation in order to develop a synthesis strategy for the preparation of a highly efficient catalyst with optimized characteristics. To achieve the goal, the effect of synthesis condition variations, such as pH, precipitation and aging temperatures, which were kept the same, and aging time and precipitation rate, on the formation of LDH material with the Cu²⁺/Al³⁺ ratio of 1.0 was studied. The aim of our work was to define the lowest value of aging time at which the formation of a well-crystalline hydrotalcite phase of LDH material prepared at pH 9 was possible. Besides, this was the first attempt to study the effect of preparation conditions of CuAl-LDH regarding their activity in AMF hydrogenation.

2. Results and Discussion

2.1. Synthesis of Cu Layered Double Hydroxides and Mixed Oxides Based on the LDHs

The co-precipitation method of nitrate solution (Cu(NO₃)₂ and Al(NO₃)₃) by base solution (NaOH and Na₂CO₃) has been described elsewhere [30,31] and was used for the synthesis of Cu layered double hydroxides with the Cu²⁺/Al³⁺ molar ratio of 1:1, differing in terms of synthesis conditions, such as pH, precipitation and aging temperature (T, °C), precipitation rate of nitrate solution (v, mL/min) and aging time (τ, h). The list of CuAl-LDH materials is presented in

Table 1. The list of CuAl catalysts and CuAl-LDH materials prepared at different synthesis conditions.

Samples	Precursor	Synthesis Conditions				Chemical Composition 4, wt.%
		pH	T, °C 1	v, mL/min 2	τ, h 3	Cu	Al
CuAl	CuAl-LDH	9	70	5	4	47.2	17.3
CuAl-10 mL	CuAl-LDH-10 mL	9	70	10	4	47.7	17.4
CuAl-pH8	CuAl-LDH-pH8	8	70	5	4	47.3	17.2
CuAl pH10	CuAl-LDH-pH10	10	70	5	4	46.7	17.5
CuAl-T50	CuAl-LDH-T50	9	50	5	4	48.1	17.3
CuAl-1 h	CuAl-LDH-1 h	9	70	5	1	46.5	17.3
CuAl-2 h	CuAl-LDH-2 h	9	70	5	2	46.5	17.4

¹ Precipitation and aging temperature. ² Precipitation rate of nitrate salts mixture.³ Aging time. ⁴ Chemical composition, determined by ICP-OES for the samples, calcined at 650 °C.

.

The reference sample was CuAl-LDH, which was prepared at constant pH 9.0 ± 0.1, temperature of 70 °C, precipitation rate of nitrate salts mixture of 5 mL/min, and aging time of precipitate of 4 h [25]. Precipitation pH 9 was used as the base of the review [32] where it was shown that lower pH results in formation of a brucite phase instead of hydrotalcite. Using pH 9–11 allows the obtaining of a pure hydrotalcite phase, but keeping higher pH requires using a higher concentration of base solutions. Regarding precipitation, an aging temperature higher than 70 °C leads to faster water evaporation during the aging of the obtained suspension. A decrease of aging temperature leads to decrease of both crystallinity and crystallite size [32]. The lower aging time causes a decrease in crystallinity degree [32], while using higher aging time can lead to the performance of energy-consuming experiments. All other samples were prepared by changing one of the synthesis parameters that is reflected in the sample designations (Table 1). The CuAl-LDH-pH8 and CuAl-LDH-pH10 materials were prepared at constant pH 8 ± 0.1 and 10.0 ± 0.1, respectively, temperature of 70 °C and precipitation rate of nitrate salts mixture of 5 mL/min. The CuAl-LDH-T50 sample was synthesized at constant pH 9.0 ± 0.1, temperature of 50 °C and precipitation rate of nitrate salts mixture of 5 mL/min. The CuAl-LDH-10mL material was prepared at constant pH 9.0 ± 0.1, temperature of 70 °C and precipitation rate of nitrate salts mixture of 10 mL/min. All above mentioned samples were aged for 4 h. Two samples, denoted as CuAl-LDH-1 h and CuAl-LDH-2 h, were synthesized at the same conditions as CuAl-LDH (pH 9, 70 °C, 5 mL/min) and aged for 1 and 2 h, respectively. After aging, the precipitates were filtered and washed with hot water (5 L) to remove the sodium ions. The precipitates were dried at 110 °C for 14 h and calcined at 650 °C for 4 h. The calcined samples were denoted as CuAl, CuAl-pH8, CuAl-pH10, CuAl-T50, CuAl-10 mL, CuAl-1 h, CuAl-2 h, respectively. The concentrations of the main components (Cu and Al) obtained by ICP-OES in the samples calcined at 650 °C were nearly the same, regardless of synthesis conditions, excluding CuAl-1 h and CuAl-2 h, probably due to incomplete precipitation during synthesis (Table 1).

2.2. Characterization of the Cu-Al Samples

Three air-dried samples CuAl-LDH, CuAl-LDH-1 h and CuAl-LDH-2 h prepared at different aging times were subjected to thermal analysis to investigate the effect of aging time on the phase transformation temperature of the LDHs. Non-isothermal temperature-programmed treatment of the samples in the temperature range was accompanied by endo- and exo-effects on DSC curves (Figure 1).

The endo-effects at the temperatures of 145–166 °C were assigned to removing of interlamellar water molecules. The second step at 195–225 °C was caused by dehydroxylation and decarbonization of interlayer carbonate. It should be noted that the endo-effect temperatures were shifted to higher values for the CuAl-LDH sample in comparison to that of the samples aged for 1 and 2 h, meaning that the carbonate ions were more strongly bound to the LDH structure. Additionally, the small exothermic peak at 612–625 °C could be related to the strongly bonded carbonates, which was confirmed by CO₂ evolving at 616 °C from the Cu-Zn-Al LDH catalyst precursor during thermal analysis coupled with mass spectrometer [33]. Thus, the calcination temperature of 650 °C was selected to exclude the presence of carbonate in the sample structure. All other samples aged for 4 h showed similar behavior as the CuAl-LDH sample during thermal analysis (the TA-DSC curves are not present in the paper).

XRD was done for all samples dried at 110 °C and calcined at 650 °C. The characteristic peaks corresponding to the (00l) crystallographic planes of the copper aluminum hydrotalcite (PDF 046-0099) were observed for the sample dried at 110 °C. The typical diffractogram was present for the CuAl-LDH sample (Figure 2a). The comparison of LDHs aged for 1, 2 and 4 h is presented on Figure 2b (low angle region). It can be seen that absolute intensities of hydrotalcite reflexes of the CuAl-LDH sample were approximately 4 times higher than those of the CuAl-LDH-1 h and CuAl-LDH-2 h samples.

This phenomenon could be caused by the presence of amorphous states in these two samples, confirmed by background enhancement. Besides, the increase of aging time resulted in formation of larger particle sizes of hydrotalcite phase, that was confirmed by increase of coherent scattering region (CSR). The average coherent scattering regions were calculated by Scherrer equation using the most intensive reflexes. It was seen that CSR values of the hydrotalcite phase in the dried samples varied widely (130–890 Å), depending on the preparation conditions (

Table 2. Phase composition and coherent scattering region of the samples.

Sample	Phase Composition	CSR, Å	Phase Composition	CSR, Å
	110 °C		650 °C
CuAl-LDH	Hydrotalcite	890	CuO	70
CuAl-LDH-2 h		590	CuO	70
CuAl-LDH-1 h		420	CuO	75
CuAl-LDH-pH8		250	CuO CuAl2O4	90 –
CuAl-LDH-pH10		250	CuO	70
CuAl-LDH-10 mL		680	CuO	60
CuAl-LDH-T50		130	CuO CuAl2O4	80 –

).

Changing pH to lower or higher values from pH 9 gave lower particle sizes of the hydrotalcite phase. The CuAl-LDH-pH8 and CuAl-LDH-pH10 samples had identical CSR values of 250 Å, while CSR of the CuAl-LDH sample was much higher (Table 2). Probably, higher aging time could result in further crystal growth or aggregation of primary particles to larger ones by aging. Using a lower temperature for synthesis of LDH promoted formation of the hydrotalcite phase having smaller particle sizes. This could be explained by the formula dL/dt = C × exp(−1/RT) where C is constant, L is the crystal size, and T corresponds to aging temperature [34]. It meant that rate of crystal growth was proportional to the aging temperature. The higher temperature, the larger the particles that were formed. Co-precipitation temperature higher than 70 °C leads to evaporation of the mother liquor containing the LDH precursor, which can affect the formation of hydrotalcite phase. Thus, the LDH synthesis was not performed at temperatures higher than 70 °C because of possible evaporation of the water solutions. The effect of precipitation rate on the formation of LDH was also investigated. The CuAl-LDH-10mL sample had slightly smaller particle size than the CuAl-LDH sample synthesized at a lower precipitation rate. In the case of obtaining the same catalytic properties of CuAl-LDH-10mL as that of CuAl-LDH, the high precipitation rate could be used to decrease the synthesis time of LDH material.

The samples’ heat-treatment at 650 °C resulted in collapse of the layered structure and formation of copper-containing phases: tenorite (CuO) and spinel (CuAl₂O₄). The CSR values of tenorite phase were similar for all samples and lay in the range of 60–90 Å (Table 2) meaning that the CSR values of the starting hydrotalcite phase had no effect on the dispersion of the final phase.

TEM images would be the best option to get information about dispersion of metal Cu particles, but it was shown [30] that formation of copper oxide occurs even after contact with air at ambient conditions for a short time. To get a deeper insight into the regularities of the reduction of CuAl samples (after calcination at 650 °C) an attempt was made to use several available methods, avoiding contact of the reduced samples with air, by means of H₂-TPR, XRD and XPS measurements.

In situ XRD measurements in H₂/He flow were performed for the samples aged for 1 and 2 h for evaluation of the size of Cu particles obtained after reduction (Figure 3). For starting materials, the broad peaks corresponding to CuO phase (PDF 05-0661) were observed. After reduction at 300 °C these reflexes disappeared while the peaks at 2θ of 43.3 and 50.5°, corresponding to metal Cu phase (PDF 04-0836), were present on the XRD patterns.

There was an additional low-intense broad peak at 2θ of 37.3°, which could be associated with the most intensive 311 reflex of CuAl₂O₄ spinel or 111 reflex of Cu₂O. It was difficult to distinguish between the two phases, since the peak positions overlapped. After cooling down the CuAl-LDH-1 h and CuAl-LDH-2 h samples to room temperature, the same phases remained on diffractograms. The CSR values for Cu metal phase were 90 and 85 Å for the CuAl-LDH-1 h and CuAl-LDH-2 h samples, respectively. Thus, the particle sizes of Cu metal were similar in these two samples, meaning that aging time had no effect on the formation of Cu particles.

H₂-TPR measurements of the samples were performed to investigate the reducibility of copper oxides (Figure 4,

Table 3. H₂-TPR, XPS and coherent scattering region of the reduced samples.

Sample	H2 Consumption, Mole × 104		Cu/Al (XPS)		CSR of Cu Phase, Å
	TPR Data	Calculated	300 °C	120 °C
CuAl	7.6	7.5	0.22	0.33	70
CuAl-10 mL	7.5	7.5	0.22	0.35	70
CuAl-pH8	7.6	7.5	0.22	0.34	70
CuAl-pH10	7.5	7.4	0.20	0.30	75
CuAl-T50	7.7	7.6	0.19	0.30	85
CuAl-1 h	7.0	7.3	0.19	0.33	80
CuAl-2 h	7.0	7.3	0.19	0.38	65

). The reduction profile of TPR experiment of the CuAl catalyst aged for 4 h had a different shape in comparison to that of the samples aged for 1 and 2 h (Figure 4a). The TPR curve of the CuAl sample showed one reduction peak at 245 °C corresponding to the reduction of CuO to Cu⁰.

There were two reduction peaks at 235 and 303 °C for the CuAl-1 h and CuAl-2 h samples, indicating the presence of two types of Cu²⁺ species in the samples. The first peak at 235 °C could be associated with the reduction of surface highly dispersed CuO while the second peak, at 303 °C, could be attributed to the reduction of bulk-like CuO phases or CuAl₂O₄ [35,36]. It is worth mentioning that increase of aging time resulted in formation of well-distributed highly dispersed CuO particles in the calcined sample, and, as a result, only one peak is present in Figure 4a.

The effect of precipitation pH on reducibility of the samples was also investigated (Figure 4b). As can be seen, there was no difference in the profiles of the TPR curves for the CuAl and CuAl-pH10 samples, meaning that pH increasing from 9 to 10 had no effect on the reducibility of the samples. There was only one peak at 242 °C, corresponding to reduction of highly dispersed CuO to Cu⁰. On the other hand, the TPR curve of the CuAl-pH8 sample prepared at pH 8 showed slightly different reduction behavior. The main peak at 235 °C, with the shoulders at 214 and 279 °C, were present on the TPR curve.

The high temperature reduction peak at 390 °C was observed (Figure 4c) for the CuAl-T50 sample prepared at lower temperature than the reference CuAl sample (50 vs. 70 °C). Besides, the presence of bulk-like CuO phase could not be excluded, due to the presence of the shoulder at 280 °C. Using the higher precipitation rate during LDH synthesis had no significant effect on reducibility of the CuAl and CuAl-10 mL samples: the reduction temperature shifted from 245 to 255 °C (Figure 4c).

The quantitative evaluation of TPR measurements is presented in Table 3. Using chemical analysis data (Table 1) the calculated CuO amount in the samples were varied in the range of 58.2–60.2 wt.%. According to the equation $CuO+H_2\to C{u}^0+H_{2}O$, the amount of hydrogen consumed for the reduction should be 7.3–7.6 × 10⁻⁴ mole. The obtained results were in agreement with the calculated values from chemical analysis for all samples, besides CuAl-1 h and CuAl-2 h (Table 3). The CuAl-1 h and CuAl-2 h samples showed lower amounts of hydrogen consumed, probably due to incomplete precipitation during synthesis.

According to the obtained results, the CuAl, CuAl-pH8, CuAl-pH10 and CuAl-10 mL samples could be reduced completely at temperature 300 °C in the 10% H₂/He flow, while higher reduction temperatures were needed for the complete reduction of the CuAl-1 h, CuAl-2 h and CuAl-T50 samples. H₂-TPR could not be performed in pure hydrogen flow. Therefore, the reduction of the samples in pure hydrogen was investigated using XPS measurements.

Before XPS measurements the samples were reduced in a special high-pressure cell (HPC) at 120 °C and 300 °C in 500 mbar H₂. The reduction temperature of 300 °C was chosen due to the data obtained by the H₂-TPR method. The reduction temperature of 120 °C was selected since the samples were reduced at 120 °C in H₂ flow prior to testing in the catalytic set-up. It was important to study the copper state in the samples reduced at those conditions.

The Cu2p XP spectra and CuLMM Auger spectra measured for the CuAl sample calcined at 650 °C (fresh) and then reduced at 120 or 300 °C, presented in Figure 5. It should be mentioned that for all other samples the same changes of XP and Auger spectra were observed. Analysis of Cu2p lines (shape and position) measured for the sample calcined at 650 °C showed that copper presented only in one state with the binding energy of 934.8 eV, which was attributed to Cu²⁺ in Cu(OH)₂ [27,30]. The Auger spectrum proved the suggestion about Cu(OH)₂ formation on the sample surface, since the kinetic energy value was 916.8 [30].

One can see that reduction of the sample in hydrogen led to changing of the XP and Auger spectra (Figure 5). According to the TPR results, reduction of Cu²⁺ ions occurred completely at 300 °C, which was confirmed by the XP spectra of the CuAl sample reduced at this temperature. However, according to XP data (Figure 5), the complete reduction of copper had already occurred at temperature of 120 °C in hydrogen. For both reduction temperatures, the copper had only one state with the binding energy of 932.6 eV, which corresponded to Cu⁰ [30,31,37]. CuLMM Auger spectra also shifted after reduction (in both cases 120 and 300 °C) to the kinetic energy value of 918.6 eV, indicating the reduction of Cu²⁺ to Cu⁰ [38]. It should be mentioned that reduction of Cu²⁺ depends on heating rate as well as H₂ concentration in feed mixture. It was shown in [33] that the higher the heating rate the higher the reduction temperature of Cu²⁺ ions are. Lower reduction temperature in the course of H₂-TPR experiments could be achieved by adjusting the heating rate lower. Besides, pure H₂ was used for reduction of the samples in a special HPC for XPS measurements that allowed reduction of the temperature of Cu samples to 120 °C.

The Cu/Al atomic ratios on the sample surface, which were relative characteristic of the surface concentration of metallic Cu, are presented in Table 3. It can be seen that all samples reduced at 300 °C had similar surface Cu/Al atomic ratio (0.19–0.21). The Cu/Al ratios rose with decrease of reduction temperature to 120°C and lay in the range of 0.30–0.38 for different samples. Thus, the lower reduction temperature promoted increase of surface concentration of copper on one hand. On the other hand, Cu/Al ratios were varied for different samples depending on synthesis parameters that could influence catalytic properties of the Cu-Al samples.

Additional XRD measurements were performed for the CuAl samples reduced by hydrogen (30 mL/min) mixed with isopropanol at a temperature of 120 °C for 2 h [39]. Figure 6 shows the XRD patterns of the reduced CuAl-pH8 and CuAl-10 mL samples (the diffractograms of other reduced samples are similar). Peaks at 43.3, 50.5 and 74.3°, corresponding to 111, 200 and 220 reflexes of Cu⁰, were observed. The CSR values of the copper phase for all catalysts were in the range of 65–85 Å (Table 3). The CuAl-2 h sample had the lowest CSR value, while the CuAl-T50 and CuAl-1 h contained the largest copper crystallites. This result correlated with the surface Cu/Al atomic ratios obtained from the XPS data.

2.3. Catalytic Test of the Cu-Al Samples

Continuous-flow AMF hydrogenation at the temperature of 60 °C and hydrogen pressure of 10 bar was performed for investigation of the catalytic properties of Cu-Al oxides. Prior to catalytic experiment, the catalysts containing CuO were reduced in hydrogen flow at 120 °C to obtain Cu⁰ metal particles [20,25,39]. It was observed that AMF hydrogenation over the CuAl catalyst gave 5-(acetoxymethyl)-2-furanmethanol (AMFM) with selectivity of 98% (

Table 4. AMF hydrogenation over CuAlO_x catalysts in a flow reactor ¹.

Entry	Catalyst	Conversion, %	Selectivity, %
1 2	CuAl	85.0	99
2	CuAl	93.8	98
3 3	CuAl	94.0	98
4	CuAl-pH10	88.8	98
5	CuAl-T50	86.8	98
6	CuAl-pH8	97.2	98
7	CuAl-10 mL	93.9	98
8	CuAl-1 h	87.0	98
9	CuAl-2 h	98.3	98
10 4	CuAl-2 h	>99.9	98

¹ AMF (0.05 M), isopropanol, catalyst (165 mg), 60 °C, 10 bar H₂, liquid flow rate of 0.5 mL min⁻¹, H₂ flow rate of 30 mL min⁻¹, reaction time 30–35 min. Before catalytic run, the catalyst was reduced by H₂ at 120 °C for 2h (H₂ flow rate of 30 mL min⁻¹); ² Reduction time was 1 h; ³ Reduction time was 3 h; ⁴ T = 80 °C.

). In addition to AMFM, 5-methylfurfural and 5-methylfurfuryl alcohol were formed in small amounts.

Prior to the reaction test, the reference CuAl sample was reduced at 120 °C for 1, 2 and 3 h. It is seen (Table 4, Entry 1–3) that increase of reduction time led to enhancement of AMF conversion from 85.0 to 94.0. However, the difference in AMF conversion was negligible for the CuAl catalyst reduced for 2 and 3 h. Thus, the reduction time of 2 h was chosen for all other experiments.

The comparison of AMF conversion (Table 4) with the surface Cu/Al ratio obtained by XPS analysis (Table 3) showed that the higher the surface Cu/Al ratio, the higher the AMF conversion. Aging time during synthesis of LDHs had effects on catalytic properties of the obtained Cu mixed metal oxides. Among CuAl-1 h, CuAl-2 h and CuAl aged for 1, 2 and 4 h, respectively, the highest AMF conversion of 98.3% was obtained for the CuAl-2 h catalyst with the Cu/Al ratio of 0.38 (Table 4, Entry 9). The CuAl-1 h and CuAl samples showed lower AMF conversion of 87.0 and 93.8%, respectively (Table 4, Entries 8 and 2). This could be explained by lower surface Cu/Al ratio of 0.33 for both these samples (Table 3). In this case, an additional parameter, which could affect catalytic properties, was reducibility of the samples, which was determined by H₂-TPR measurements (Table 3). The hydrogen amount of 0.7 mmole consumed for the CuAl-1 h sample was lower than the value calculated from Cu content (0.73 mmole) in the sample, meaning that some Cu²⁺ ions were not reduced. This assumption was confirmed by the presence of CuAl₂O₄/Cu₂O phase on the XRD pattern after reduction of the CuAl-1 h sample at 300°C (Figure 3a). On the other hand, the CuAl sample was reduced completely. Thus, all copper ions introduced in the CuAl-LDH material on the precipitation stage were reduced to Cu⁰, which acted as the active species in AMF hydrogenation.

Lower AMF conversions of 88.8 and 86.8% (Table 4, Entries 4 and 5) were obtained for the CuAl-pH10 and CuAl-T50, respectively, due to a lower surface Cu/Al ratio of 0.30. Decreasing synthesis pH resulted in increase of AMF conversion to 93.8 and 97.2% for CuAl prepared at pH 9 and CuAl-pH8 prepared at pH 8, respectively. According to TPR measurements (Table 3), the reducibility of the sample was similar. At the same time, surface Cu/Al ratio rose with decrease in synthesis pH from 0.30 for CuAl-pH10 to 0.34 for CuAl-pH8, meaning higher surface Cu concentration of CuAl-pH 8. The sample CuAl-10 mL (Table 4, Entry 7) prepared using a higher precipitation rate showed the same AMF conversion of 93.9 as the CuAl sample. Thus, LDH materials could be prepared using a high precipitation rate without losing any catalytic activity.

It should be noted that the Cu/Al ratio of CuAl-pH8 and CuAl-10 mL was similar. However, these samples showed different AMF conversions (Table 4, Entries 6 and 7). Thus, there was another parameter which could affect the catalytic properties of CuAl mixed metal oxides on the base of LDH. The broad peaks at 2θ = 31.3, 36.9, 60.0 and 65.6° relating to formation of CuAl₂O₄ oxide were present in XRD patterns of the CuAl-pH8 and CuAl-10 mL samples (Figure 6). Lattice parameter (8.047(3) and 8.029(4) for CuAl-10 mL and CuAl-pH8, respectively) of the obtained compound in both samples were higher than the value for pure Al₂O₃ (7.910 Å, PDF 02-1420) and lower than that for CuAl₂O₄ (8.075 Å, PDF 33-0448). Therefore, formation of a solid Cu_1-xAl_2+xO₄ solution could be assumed. A rough estimation of the phase composition by the Rietveld method showed that the CuAl-10 mL and CuAl-pH 8 samples contained 33 wt.% and 40 wt.% of Cu metal, that could cause higher catalytic activity of the CuAl-pH 8 sample. During the Rietveld simulation, the ratio between Cu and CuAl₂O₄, lattice parameters and profile parameters were refined. The occupancy of Cu and Al in tetragonal and octahedral position in the spinel phase, as well as their thermal factors, were fixed, R_w factors were 13–16%. Changing the occupancy of Cu and Al in spinel oxide did not improve refinement. However, the obtained difference in the XRD data was very subtle. Thus, comparing the catalytic activity of the obtained samples, not only XRD measurements but also other methods of its characterization should be taken into account.

Thus, it can be concluded that using lower synthesis pH (9 vs. 10), higher synthesis temperature (50 vs. 70 °C) and shorter aging time (2 h vs. 4 h) results in the obtaining of catalysts with higher activity. The best hydrogenation activity was obtained over the CuAl-2 h sample (prepared at constant temperature of 70 °C, constant pH 9 with aging time of 2 h) which provided 98% yield of AMFM at 80 °C (Table 4, Entry 10).

3. Materials and Methods

3.1. Characterization Techniques

The metal content (Cu and Al) was determined by ICP-OES on an Optima 4300 DV instrument (Perkin Elmer, Waltham, MA, USA).

Thermogravimetric analysis (TGA) of the Cu-based materials was performed in an air atmosphere in the temperature range of 40–700 °C on a Netzsch STA 449 thermogravimeter (Netzsch, Selb, Germany) with a heating rate of 10 K min⁻¹ (the sample weight was 50 mg).

X-ray diffraction analysis of the samples was carried out on a Bruker D8 Advanced diffractometer (Bruker, Billerica, MA, USA) with CuKα-radiation (2θ = 5–70° with step of 0.05°, the accumulation time is 3 s). The signals were detected by the multichannel LynxEye detector. Phase analysis was done by comparison of interlayer distances d_i and intensities I_i of corresponding reflexes with theoretical values from ICDD PDF-2 database.

Temperature programmed reduction of the samples by hydrogen (H₂-TPR) was done in the temperature range of 100–600 °C in a quartz reactor in gas mixture of 10% H₂ in Ar with a heating rate of 6 °C/min. Prior to reduction, the samples (0.10 g) were preliminary treated at 200 °C in Ar for 1.5 h. H₂ consumption was determined by thermal conductivity detector (TCD).

The X-ray photoelectron spectroscopy (XPS) measurements were performed on the photoelectron spectrometer built by SPECS (SPECS GmbH, Berlin, Germany), equipped with a PHOIBOS-150 hemispherical energy analyzer and AlK_α irradiation (hν = 1486.6 eV, 150 W). The binding energy (BE) scale was pre-calibrated using the positions of the photoelectron of Au4f_7/2 (BE = 84.0 eV) and Cu2p_3/2 (BE = 932.67 eV) core level peaks. Residual gas pressure was better than 8 × 10⁻⁹. The Al2p peak at 74.5 eV of the Al₂O₃ support was used as an internal standard. For the measurements, the samples were supported onto the double-sided conducting copper scotch tape (Scotch 3M©, St. Paul, MN, USA). Spectra analysis and peak fitting were performed with XPSPeak 4.1 software. Integrated line intensities were measured from area of the corresponding narrow regions (Al2p, C1s, O1s, Cu2p and Na1s). The relative amount of the elements on the sample surface and the ratio of their atomic concentrations were determined from the integrated intensities of the lines corrected by their respective atomic sensitivity factors [38]. Additionally, the samples were measured after reduction in hydrogen at different temperatures. Reduction was performed in the chamber of a photoelectron spectrometer equipped with a special HPC, which allowed pretreatments of the samples under different gases at a pressure up to 1 bar and temperature range from 50 to 450 °C. The samples were rubbed into a stainless-steel mesh spot welded on the standard flag-style sample holder. Reduction was performed for 2 h at hydrogen pressure ~500 mbar and different temperatures—120 and 300 °C, then the samples were cooled down to RT and hydrogen was pumped out from the cell (to UHV conditions). Finally, the samples were transferred to an analyzer chamber after reduction without contact with air.

3.2. Catalytic Test

Studying of the catalytic properties was carried out in a continuous flow set-up H-Cube Pro™ (Thalesnano, Budapest, Hungary) equipped with stainless-steel CatCart^®30 reactors [20,25,39]. Prior to the catalytic run, the catalyst (0.165 g, 250–500 μm) was reduced by hydrogen (30 mL/min) mixed with isopropanol (0.5 mL/min) at a temperature of 120 °C and pressure of 10 bar for 2 h. When the desired parameters were achieved, the input was switched to the flask with 0.05 M solution of AMF (the starting point of the catalytic test). Catalytic tests were carried out at temperatures of 60 and 80 °C, pressure of 10 bar, liquid and hydrogen flow rates of 0.5 and 30 mL/min, respectively (inlet H₂/substrate molar ratio was 54). The AMF flow rate of 0.5 mL/min was the optimal flow rate for liquids. The lower flow rate could lead to unstable work of the setup while the higher flow rate resulted in bad mixing of liquid with hydrogen flow. The catalyst performance was determined by analysis of the samples taken after 30–35 min of the experiment’s start.

The reaction mixtures were analyzed by gas chromatography (Agilent 6890 N Instrument equipped with a capillary HP-1MS column (30 m × 320 μm × 1.00 μm) using n-decane as the internal standard. Identification of the products was carried out by GC–MS (Agilent 7000B Triple Quad System).

The conversion of AMF, X (%), and the selectivity, S (%) were calculated respectively by using Equations (1) and (2):

$$X=\left(1-\frac{C_{AMF}}{C_{AMF}^0}\right)\times 100\%$$

(1)

$$S=\frac{C_{AMFM}}{{\displaystyle \sum }{C}_i}\times 100\%$$

(2)

where C⁰_AMF and C_AMF are the inlet and outlet concentrations of AMF, respectively; C_AMFM is the outlet concentration of AMFM; ΣC_i is the total concentration of reaction products.

4. Conclusions

The effect of synthesis conditions, such as pH, temperature, aging time and precipitation rate, of Cu-Al layered double hydroxides on the formation of Cu-based materials, as well as catalytic properties of the Cu-Al mixed oxides in flow hydrogenation of 5-acetoxymethylfurfural (AMF) to 5-(acetoxymethyl)-2-furanmethanol (AMFM), was studied. As was shown, the longer aging time resulted in formation of larger particles, due to further crystal growth or aggregation of primary particles. However, calcination of LDH at 650 °C led to formation of the tenorite phase of similar particle size. According to XPS measurements, the reduction of the Cu-Al samples at 300 °C led to sintering of Cu particles, leading to the similar surface Cu/Al ratio of all the samples, while decrease in reduction temperature to 120 °C allowed the obtaining of a different surface concentration of copper. The higher Cu/Al ratio was obtained for the CuAl-LDH sample synthesized at constant pH 9, temperature of 70 °C, precipitation rate of 5 mL/min and aging time of 2 h. This was confirmed by catalytic data obtained for the CuAl-2 h in AMF hydrogenation. The higher AMF conversion of 98.3% was obtained over this catalyst. The lower and higher aging time (1 h and 4 h) resulted in decrease of AMF conversion due to lower amounts of metallic Cu on the surface. The effect of synthesis pH (8, 9 or 10) on catalytic properties for the catalysts aged for 4 h was also determined. It was shown that pH 8 was preferable for obtaining higher AMF conversion, due to higher Cu/Al ratio in the sample. Using a higher precipitation rate (5 vs. 10 mL/min) did not affect catalytic activity of the sample and allowed synthesis of LDH materials in a shorter time. Under optimized reaction conditions, the CuAl-2 h catalyst, which was prepared at constant temperature of 70 °C, constant pH 9 with aging time of 2 h, provided 98% yield of AMFM. Thus, the synthesis strategy for the preparation of a highly efficient catalyst for hydrogenation of AMF to AMFM with optimized characteristics is suggested.