Au Clusters Supported on Defect-Rich Ni-Ti Oxides Derived from Ultrafine Layered Double Hydroxides (LDHs) for CO Oxidation at Ambient Temperature

Abstract

Ultrafine layered double hydroxides (LDHs) have abundant hydroxy groups at their edge sites, serving as anchor sites for metal NPs. Furthermore, transformation of ultrafine LDHs into mixed metal oxides (MMOs) generates abundant oxygen vacancies, which are advantageous for O₂ activation during Au-catalyzed CO oxidation. We used ultrafine Ni-Ti LDHs with low crystallinity or Ni-Ti MMOs supported on SiO₂ onto which Au NPs were deposited by deposition–precipitation (DP) and DP–urea (DPU). The catalytic activity of the Au catalysts was significantly affected by the preparation method, with the highest activity obtained by depositing Au onto LDH/SiO₂ by DPU, followed by transformation of LDH to MMO (Au/Ni-Ti MMO/SiO₂ (LDH-DPU)). The presence of Au on LDHs affected the transformation of LDHs into MMOs, resulting in LDH-DPU having the greatest number of oxygen vacancies in the TiO₂ domain in MMOs. Consequently, the adsorbed or the lattice oxygen on the surface of LDH-DPU can be easily utilized for CO oxidation at low temperatures. Moreover, the catalytic activity of LDH-DPU increased with water vapor concentration up to 100% relative humidity at room temperature, suggesting the potential of Au/Ni-Ti MMO/SiO₂ as an air purification catalyst.

1. Introduction

The catalytic oxidation of harmful CO to CO₂ in air at ambient temperature is a fundamental and important process for air purification. In 1987, Haruta et al. reported that Au nanoparticles (NPs) supported on reducible metal oxides exhibited outstanding catalytic activity for CO oxidation at −70 °C [1]. This discovery sparked extensive research on Au catalysts because other metal catalysts, such as Pd and Pt, required higher temperatures exceeding 100 °C. Since then, tremendous efforts have been devoted to the development of highly active Au catalysts. Because the catalytic performance of Au significantly depends on the size of Au NPs and the type of supports, numerous preparation methods have been developed to deposit small Au particles onto various supports [2,3,4]. Specifically, Au clusters with diameters less than 2 nm are desirable to exhibit high activity. The reaction mechanism of Au-catalyzed CO oxidation has been discussed both experimentally and theoretically [2,4,5,6,7,8]. It is now widely accepted that the active site of Au-catalyzed CO oxidation is Au–support interface. Au works as a CO adsorption sites, and O₂ is activated at the oxygen vacancies on the metal oxide supports [4,6,7,8]. Therefore, reducible metal oxides, such as NiO, Co₃O₄, CeO₂, and TiO₂, have been frequently used as supports for Au NPs.

Recently, single-Au-atom catalysts have attracted increasing interest in order to minimize the use of precious metals [9]. CeO₂ [10], Co₃O₄ [11], FeO_x [12], and NiO [13] are regarded as suitable supports because single Au atoms can be stabilized at their metal vacancies. However, these single-Au-atom catalysts can be only realized in extremely low loading weights, such as 0.03 wt% of Au [12]_, except for NiO [13]. This means that a large amount of support is required to exhibit demanding catalytic performance, despite the fact that these metal oxides are classified as rare metals except for FeO_x. Therefore, there is a need to minimize the usage of both Au and these rare metal oxides for sustainable development. Although FeO_x is an abundant metal oxide, the T_1/2 value of Au₁/FeO_x, i.e., the temperature of 50% CO conversion, has been reported to be at around 40 °C [12]. Photocatalytic CO oxidation over Au/TiO₂ has been also reported, but the working temperature exceeded 60 °C [14].

Despite the abundance of SiO₂ on Earth, SiO₂ is regarded as an unsuitable support for Au NPs because of the lack of oxygen vacancies and aggregation of Au NPs on SiO₂. Conversely, irreducible metal oxides such as SiO₂ and Al₂O₃ are tolerant of deactivation in CO oxidation under humid air conditions, while Au/TiO₂ experiences a decrease in activity in high humidity [15,16,17]. Thus, there is still room for the development of CO oxidation catalysts that incorporate minimal amounts of Au and rare metals and catalysts capable of functioning at ambient temperature in humid air for applications in air purification.

Layered double hydroxides (LDHs) consist of divalent metal hydroxides with a brucite-layered structure, wherein trivalent metal ions are partially replaced with divalent ones. Many types of LDHs have been synthesized [18]. The chemical formula of LDH is [M²⁺_1−xM³⁺_x(OH)₂][Aⁿ⁻_x/n·yH₂O], where the anion (Aⁿ⁻) and water molecules are intercalated to compensate for the positive surface charge. LDHs have interesting features such as a solid base, anion exchange ability, and transformation into mixed metal oxides (MMOs). Natural LDHs, known as hydrotalcite ([Mg_0.75Al_0.25(OH)₂][(CO₃²⁻)_0.125·0.5H₂O]), have been used as supports for Au. Because the hydroxyl groups at the edge sites of LDHs are more basic, Au tends to be initially deposited at the edge sites on Mg-Al LDHs, resulting in the formation of Au clusters at low Au loadings. Subsequently, Au is deposited on the lateral face of the LDH to produce aggregated Au NPs [19]. The size of the LDH affects the generation of defects in LDHs [20], and metal- and oxygen vacancies after transformation into MMOs [21,22]. Zhang et al. transformed ultrafine Ni-Ti LDH particles (<4 nm) into defect-rich Ni-Ti MMOs with abundant Ni³⁺ and Ti³⁺ sites [21]. Thus, Ni-Ti LDH NPs with abundant edge OH groups and Ni vacancies are beneficial for stabilizing small Au particles. In addition, the rich oxygen vacancies on the TiO_x domain derived from the Ni-Ti LDH NPs are expected to efficiently activate O₂, leading to high catalytic activity for CO oxidation.

In this study, we attempted to form highly dispersed ultrafine Ni-Ti LDH on SiO₂ and deposited Au NPs on the prepared Ni-Ti LDH/SiO₂ and Ni-Ti MMO/SiO₂. The Au particle size and catalytic activity for CO oxidation varied according to the preparation method and support used (LDH/SiO₂ or MMO/SiO₂). The deposition of Au onto LDH/SiO₂ by deposition–precipitation with urea (DPU) followed by the transformation of LDH into MMO gave the most active catalyst, which was active for CO oxidation at room temperature. On the other hand, deposition–precipitation (DP) of Au onto LDH/SiO₂, followed by calcination, resulted in relatively large Au NPs with much lower activity. However, DP is more suitable for the deposition of Au onto MMO/SiO₂ to obtain small Au NPs. The presence of Au during calcination affected the transformation of LDHs into MMOs, and the number of Ni- and oxygen vacancies in the resultant ultrafine Ni-Ti MMO varied based on the preparation method. Au/Ni-Ti MMO/SiO₂, prepared from LDH/SiO₂ followed by calcination, possessed abundant oxygen vacancies in the TiO_x domain, which contributed to its high activity. Consequently, we found that the oxygen vacancies could be generated easily by using ultrafine Ni-Ti LDH. Because Ni-Ti LDHs were highly dispersed on SiO₂, the use of Ni could also be reduced. Furthermore, the LDH-DPU catalyst exhibited enhanced activity in CO oxidation even at 4% water concentration, which is equal to 100% relative humidity (RH) at room temperature. The results suggest that Au on ultrafine Ni-Ti MMO/SiO₂ can serve as efficient air purification catalysts under high humidity conditions.

2. Results and Discussion

2.1. Characterization of Ni-Ti LDH/SiO₂

Ni-Ti LDH/SiO₂ (Ni/Ti = 4/1 atomic ratio) was prepared through hydrothermal synthesis using Ni(NO₃)₂·6H₂O and TiCl₄, following previously reported methods [23,24] with minor modification. Subsequently, Ni-Ti LDH/SiO₂ was converted to Ni-Ti MMO/SiO₂ by calcination in air at 500 °C for 2 h (Scheme 1, upper side). Additionally, bulk Ni-Ti MMO was obtained by preparing pristine Ni-Ti LDH in the absence of SiO₂ in the same manner, followed by calcination at 500 °C for 2 h. The powder X-ray diffraction (XRD) patterns of bulk Ni-Ti LDH, Ni-Ti LDH/SiO₂, and Ni-Ti MMO/SiO₂ (Ni/Ti = 4) are shown in Figure 1a. Zhang et al. prepared Ni-Ti LDH NPs with ~20 nm using a reverse microemulsion method in an organic solvent and reported diffraction peaks at 35 and 60°, which corresponded to the (012) and (110) planes, respectively. Additionally, unassigned peaks were observed at around 25, 38, and 45° [21,24]. For the bulk Ni-Ti LDH, diffraction peaks were observed at 10, 25, 34, and 60°, which are almost in agreement with previous reports [21,24]. The peaks at 10 and 25° were probably attributed to the (003) and (006) planes, respectively, suggesting the typical layered structure of LDHs. The peak at 19° was probably ascribed to a partial formation of β-Ni(OH)₂ (ICDD No. 00-014-0117) [25]. The broad peaks indicate that small Ni-Ti LDH crystallites were obtained using hydrothermal method. The XRD peaks corresponding to Ni-Ti LDH were also observed for Ni-Ti LDH/SiO₂, together with the halo pattern of amorphous SiO₂ at approximately 25°. The broad and weak peaks of Ni-Ti LDH suggest that very small Ni-Ti LDH crystallites were formed on SiO₂ and/or that the Ni-Ti LDH on SiO₂ had low crystallinity.

Thermogravimetric analysis (TGA) was performed to elucidate the transformation of LDHs into MMOs (Figure S1). In the case of bulk Ni-Ti LDH, the weight gradually decreased up to 300 °C, followed by a sharp decline from 300 to 350 °C. These steps coincide with the desorption of the intercalated H₂O and CO₂ and the dehydration of hydroxides, respectively [26]. The transformation of LDH to MMO was completed at 400 °C, with a weight loss of 31%, which is consistent with the previous report [21]. These desorption and dehydration steps were shifted to higher temperatures for Ni-Ti LDH/SiO₂ (Figure S1), and the transformation into MMO was completed at 550 °C.

2.2. Characterization of Ni-Ti MMO/SiO₂

A temperature-dependence study was conducted by varying the calcination temperatures of Ni-Ti LDH/SiO₂ within the range of 500 to 900 °C (Figure S2). No obvious change was observed after calcination of Ni-Ti LDH/SiO₂ at 500 and 600 °C. However, above 700 °C distinct peaks corresponding to NiO peak at 43° and NiTiO₃ peaks at 33 and 35.8° (ICDD No. 00-033-0960) were observed. Calcination below 600 °C produced NiO and TiO₂, while Ni₂TiO₄ and NiTiO₃ were formed above 600 °C [21]. Thus, Ni-Ti LDH/SiO₂ was calcined at 500 °C to convert Ni-Ti MMO/SiO₂ (Figure 1a).

The XRD pattern of the Au/bulk Ni-Ti MMO showed sharp peaks at 25, 37, 43, and 63°, which were attributed to TiO₂ (anatase, ICDD No. 00-021-1272) and NiO (ICDD No. 00-044-1159) (Figure 1b). No obvious change of XRD pattern was observed for Ni-Ti MMO/SiO₂, whereas the (003) diffraction at 10° disappeared. In addition, the peaks at 25 and 37° attributed to TiO₂ and NiO, respectively, became slightly increased. These results suggest that LDH was converted to MMO, although LDH remaining was not fully ruled out. The broad and weak diffraction peaks suggest that the aggregation of Ni-Ti MMO crystallites could be prevented on SiO₂ support during calcination. Scanning electron microscopy (SEM) revealed that Ni and Ti species were highly dispersed on SiO₂ (Figure S3). The Ni and Ti loadings were calculated using inductively coupled plasma–atomic emission spectroscopy (ICP-AES) (

Table 1. Physicochemical properties and catalytic activity of Au/Ni-Ti MMO/SiO₂.

Entry	Catalyst	Loading/wt% 1			Au Size 2/nm	T1/23/°C
		Au	Ni	Ti
1	LDH-DP	0.24	15	2.0	4.0 ± 1.5	115
2	LDH-DPU	0.28	15	2.6	2.3 ± 2.0	10
3	MMO-DP	0.15	15	2.9	1.7 ± 0.5	38
4	MMO-DPU	0.3	15	2.0	3.5 ± 1.8	191
5	Au/Ni-Ti LDH/SiO2	0.28	15	2.6	3.5 ± 1.2	182
6	LDH-DPU-300	0.28	15	2.6	–	87
7	LDH-DPU-400	0.28	15	2.6	–	11
8	Au/Ni-Ti MMO/SiO2 (Ni/Ti = 2/1)	0.25	32	2.0	6.0 ± 2.1 4	65
9	Au/bulk Ni-Ti MMO	0.23	–	–	3.2 ± 1.3	24
10	Au/TiO2	0.96	–	–	4.1 ± 2.2	10–20
11	Au/NiO	0.93	–	–	– 5	57
12	Au/SiO2	1.00	–	–	6.6 ± 2.7	>300
13	Au/Fe2O3	0.81	–	–	1.8 ± 0.7	36

¹ Estimated by ICP-AES. ² Estimated by HAADF-STEM. ³ The temperature at 50% CO conversion. CO oxidation conditions: 0.1 g of catalyst, 1 vol% CO in air (33.3 mL min⁻¹), and space velocity (SV) of 20,000 mL h⁻¹ g_cat⁻¹. ⁴ The mean diameters of Au particles were estimated by TEM before calcination. ⁵ Single Au atoms were mainly deposited on NiO [13].

). The Ni/Ti atomic ratio was estimated to be 4.8 by ICP-AES, which is almost consistent with the initial Ni/Ti ratio (4/1).

2.3. Characterization of Au/Ni-Ti MMO/SiO₂

The deposition of Au NPs to Ni-Ti LDH/SiO₂ and Ni-Ti MMO/SiO₂ was performed by DP and DPU to give four types of Au catalysts, as shown in Scheme 1. First, Au was deposited on Ni-Ti LDH/SiO₂ by DP and DPU, followed by calcination in air at 500 °C to transform LDH to MMO, yielding Au/Ni-Ti MMO/SiO₂; the catalysts were denoted as LDH-DP and LDH-DPU, respectively. Alternatively, Au was also deposited on Ni-Ti MMO/SiO₂ by DP and DPU; this was followed by the reduction of Au(III) species by air calcination at 300 °C and by reduction with NaBH₄, according to the literature [27], giving Au/Ni-Ti MMO/SiO₂, denoted as MMO-DP and MMO-DPU, respectively. The initial Au loading was set to 0.3 wt%, and almost 80% of Au was deposited, except for MMO-DP (Table 1). The relatively low actual Au loading for MMO-DP (Au 0.15 wt%) was ascribed to the negatively charged surface (−29 mV) at pH 7, as determined by zeta potential measurement. This surface charge led to electronic repulsion between the MMO/SiO₂ and negatively charged Au precursors.

The XRD patterns of four Au/Ni-Ti MMO/SiO₂ (Figure 1b) resembled that of Ni-Ti MMO/SiO₂ (Figure 1a). No diffraction peaks corresponding to Au were observed owing to low Au loading (Figure 1b). The size of Au particles was estimated by high-angle annular dark-field transmission electron microscopy (HAADF-STEM), and the size distributions are summarized in Table 1 and Figure 2. In Figure 2, a needle-like structure can be observed, which would be ascribed to the layered structure of MMO derived from LDH. The Au particles were highly dispersed on the Ni-Ti MMO/SiO₂ of all the catalysts. Although LDH-DP and MMO-DPU gave Au NPs with diameters of 3.5–4.0 nm, Au clusters with diameters of 2.3 and 1.7 nm could be deposited on LDH-DPU and MMO-DP. The Au particle size in the Au/bulk Ni-Ti MMO prepared by DP was estimated to be 3.2 nm (Figure S4). From these results, LDH-DPU and MMO-DP are suitable preparation methods and highly dispersed ultrafine LDH and MMO are beneficial for the deposition of Au clusters.

Because the M²⁺/M³⁺ ratio of LDHs can be varied up to 2/1, Ni-Ti LDH with a Ni/Ti ratio of 2/1 was synthesized. Subsequently, Au/Ni-Ti MMO/SiO₂ with the Ni/Ti ratio of 2/1 was prepared from NiTi LDH/SiO₂ by DPU, followed by calcination to form MMO. However, relatively large Au NPs (6.0 nm) were obtained (Figure S5). The Ni/Ti ratio affected the crystallite size of NiO and TiO₂ domains, and a suitable Ni/Ti ratio (4/1) resulted in small Au particles.

2.4. Catalytic Activity for CO Oxidation

The catalytic activity of the Au/Ni-Ti MMO/SiO₂ for CO oxidation was evaluated using a fixed-bed flow reactor, and the results are shown in Table 1 and Figure 3. The order of catalytic activity is LDH-DPU > MMO-DP > LDH-DP > MMO-DPU (entries 1–4). When compared to LDH-DP and LDH-DPU, LDH-DPU stabilized smaller Au particles with higher loading, exhibiting in higher catalytic activity. In contrast, comparing MMO-DPU and MMO-DP, MMO-DP stabilized smaller Au particles, resulting in higher activity. However, the actual Au loading of MMO-DP was lower than LDH-DPU. As a result, LDH-DPU exhibited the highest catalytic activity and recorded T_1/2 of 10 °C.

In order to reveal the factors for the high activity of LDH-DPU, the transformation of LDH into MMO is discussed. Au/Ni-Ti LDH/SiO₂ prepared by DPU followed by the reduction of Au(III) with NaBH₄ (entry 5) showed much higher T_1/2 than LDH-DPU, suggesting that the transformation of LDH into MMO is important for achieving high catalytic activity. The calcination of Au/Ni-Ti LDH/SiO₂ at 300 and 400 °C denoted as LDH-DPU-300 and LDH-DPU-400, respectively, was also performed. Although LDH-DPU-400 showed similar activity to that of LDH-DPU, LDH-DPU-300 showed much lower catalytic activity than that of LDH-DPU. Thus, the incomplete transformation of LDH to MMO at 300 °C gave unsatisfactory results (entries 2, 6, and 7).

Next, the dependence of Ni/Ti ratio of Ni-Ti MMO on the catalytic activity was examined. Compared to Ni/Ti = 4/1, Au/Ni-Ti MMO/SiO₂ (Ni/Ti = 2/1) showed lower activity than LDH-DPU because of its larger Au NPs (entry 8). Moreover, LDH-DPU exhibited higher activity than Au/bulk Ni-Ti MMO (entries 2 and 9). Given that bulk Ni-Ti MMO has large NiO and TiO₂ crystallites, according to XRD (Figure 1), highly dispersed ultrafine NiO and TiO₂ domains would work as an efficient support for CO oxidation.

The catalytic activity of the most active LDH-DPU was compared with those of commercially available Au/TiO₂, Au/NiO, Au/SiO₂, and Au/Fe₂O₃ (entries 10–13). Au/SiO₂ showed very low activity because O₂ could not be activated due to the lack of oxygen vacancies in SiO₂ (entry 12). LDH-DPU exhibited higher catalytic activity than Au/NiO, although single Au atoms are major species in Au/NiO (entry 11). It has been reported that inverse NiO-on-Au supported on SiO₂ catalysts exhibited high catalytic activity for CO oxidation at room temperature and that the reaction rate was recorded to be 1.74 mol_CO g_Au⁻¹ h⁻¹ [28]. The reaction rate of LDH-DPU was calculated to be 2.32 mol_CO g_Au⁻¹ h⁻¹ at 25 °C, which is superior to the previous report. Au/NiO prepared by double impregnation was also reported to show high activity, and the reaction rate was calculated to be 2.59 mol_CO g_Au⁻¹ h⁻¹ [29]. Although the reaction rate of LDH-DPU was slightly lower than this value, the amount of Ni was lower than the reported Au/NiO. The T_1/2 of LDH-DPU was almost the same as that of Au/TiO₂ whereas the Au loading was smaller (0.3 wt%) than that of 1 wt% Au/TiO₂ (entry 10). Moreover, the T_1/2 of Au/Fe₂O₃ was 36 °C, which showed lower activity than LDH-DPU despite the small Au particle size and redox activity of Fe₂O₃ (entry 13). From these results, LDH-DPU exhibited the highest catalytic activity based on the amount of Au used.

2.5. Evaluation of Oxygen Vacancies on Ni-Ti MMO/SiO₂

These CO oxidation results suggest that the catalytic activity significantly varies depending on not only the Au particle size but also the support (LDH or MMO) on which the Au particles are deposited and on the preparation method used. It has been reported that abundant Ni- and O vacancies are formed by the conversion of LDH to MMO, as the size of Ni-Ti LDH decreases [21]. Therefore, the Ni 2p, O 1s, and Ti 2p XPS of the four Au catalysts were measured to evaluate these vacancies, and the results are shown in Figure 4.

Ni 2p_3/2 peaks of LDH-DP, LDH-DPU, MMO-DP, and MMO-DPU were observed at 857.3, 856.7, 858.4, and 857.7 eV, respectively (Figure 4A). The peaks of Ni²⁺, Ni³⁺, Ni oxyhydroxide (NiOOH), and Ni(OH)₂ were observed at 854.0–854.4, 855.5–856.3, 856.8–857.7, and 855.4–856.2 eV, respectively [30,31,32,33,34,35]. When Au was deposited on LDH, the resultant MMO contained Ni²⁺ as the major species, while NiO domains of MMO-DP and MMO-DPU showed a higher energy shift. Although Ni(OH)₂ species remaining cannot be fully ruled out, the higher binding energy shift compared to the typical Ni²⁺ of NiO implies the presence of Ni³⁺ species, i.e., rich Ni vacancies, in MMO/SiO₂. However, the energy shift did not correlate with the size of the Au particles. Therefore, the size of the Au particles was affected by the preparation method. Accordingly, the smallest Au particle size of MMO-DP was probably due to the low actual Au loading, which prevented aggregation during calcination. For Ni-Ti LDH/SiO₂, LDH-DP resulted in larger Au NPs than LDH-DPU, indicating that the aggregation of Au proceeded during calcination at 300 °C before converting LDH into MMO at 500 °C. Compared the samples prepared by DPU, LDH-DPU produced smaller Au particles than MMO-DPU, suggesting that the abundant edge OH groups of ultrafine Ni-Ti LDH were beneficial for stabilizing small Au particles.

The presence of oxygen vacancies was evaluated by O 1s and Ti 2p XPS. As shown in Figure 4B, the major O 1s peak was observed at 532.2–532.8 eV corresponding to the lattice oxygen of SiO₂ and MMO. The peak at higher binding energy was observed in the range of 533.8–534.2 eV, which was attributed to the oxygen vacancy on the MMO surface [36,37]. Percentages of the peak area ascribed to the oxygen vacancy were calculated to be 23.7%, 14.6%, 14.2%, and 12.3% for LDH-DPU, MMO-DP, LDH-DP, and MMO-DPU, respectively. The number of the oxygen vacancies of MMO increased particularly for LDH-DPU, and the order of the number of oxygen vacancies was consistent with the catalytic activity order.

Furthermore, Ti 2p XPS was also investigated to discuss the oxygen vacancies of the TiO_x domain. Ti 2p_3/2 peaks of LDH-DP, MMO-DP, and MMO-DPU were observed at 459.4, 460.4, and 459.8 eV, respectively (Figure 4C), which are almost consistent with those of TiO₂ [38,39]. In contrast, the Ti 2p_3/2 peak of LDH-DPU shifted toward lower binding energies compared to the other three catalysts, at 458.9 eV. The lower energy shift suggests the presence of Ti³⁺ species, indicating an increase in the number of oxygen vacancies in the TiO_x domain. Thus, LDH-DPU, which was rich in oxygen vacancies, exhibited the highest CO oxidation activity. The TGA diagram shows that the transformation of LDH to MMO shifted to a higher temperature in the presence of Au particles on the LDH. This leads to the suppression of MMO crystal growth, resulting in small MMOs with rich oxygen vacancies.

Au 4f XPS analysis was also measured. The Au 4f_7/2 peaks of LDH-DP, LDH-DPU, MMO-DP, and MMO-DPU were observed at 83.4, 83.5, 83.3, and 83.8 eV, respectively. These values are consistent with those of the Au⁰ NPs (Figure S6).

Figure 5 depicts the relationships between the mean diameter of Au particles and the catalytic activity for CO oxidation based on the XRD, XPS, TEM, and CO oxidation reactions. To exhibit high activity, both an Au particle size smaller than about 2 nm and rich oxygen vacancies on MMOs were found to be key factors. As a result, LDH-DPU exhibited the highest activity.

2.6. Temperature-Programmed Reaction of CO

To further investigate the difference in the CO oxidation activity with respect to O₂ activation, a temperature-programmed reaction of CO (CO-TPR) was performed. Figure 6 shows CO-TPR profiles after pretreatment in air at 250 °C. CO was fed into the sample cell, and the amount of CO₂ generated was determined by mass spectrometry. LDH-DPU showed two CO₂ desorption peaks at approximately 0 and 100 °C, whereas the other catalysts showed peaks at much higher temperatures. CO₂ gradually formed at 20 °C for MMO-DP, whereas only at temperatures above 100 °C for MMO-DPU. The order of the CO₂ desorption temperatures was consistent with the T_1/2 values for CO oxidation.

Behm et al. proposed the Au-assisted Mars–van Krevelen mechanism for CO oxidation over Au/TiO₂ [8]. In this mechanism, CO is adsorbed on Au, CO reacts with the surface lattice oxygen of TiO₂, and the oxygen vacancy is formed. The lattice oxygen is replenished again by O₂, and CO₂ finally desorbs. This mechanism is well explained at ambient temperature or above. We reported that the Au-assisted Mars–van Krevelen mechanism for Au/ZnO-catalyzed CO oxidation changes at around −20 °C, and that the activation of weakly adsorbed O₂ molecules is critically important for CO oxidation at low temperature [40]. According to these reports, it is likely that the lattice oxygen was used for CO oxidation above 30 °C, but that the CO₂ desorption peak at 0 °C is related to weakly adsorbed O₂ molecules at the oxygen vacancies. Thus, the CO-TPR results suggest that LDH-DPU possessed more active oxygen species on the MMO surface; O₂ would be adsorbed on the abundant oxygen vacancies in the TiO_x domain of MMOs; and/or the surface lattice oxygen of MMOs would be more labile owing to the disordered small crystallites.

2.7. CO Oxidation in Humid Air

We examined the dependence of water concentration on Au-catalyzed CO oxidation, because high catalytic activity in humid air at ambient temperature is crucial for catalytic application in air purification. We have reported that the catalytic activity of Au/TiO₂ markedly dropped to about half of that in dry conditions, even in the presence of 1% H₂O [15]. However, the catalytic activity of Au/Al₂O₃ and Au/SiO₂ at room temperature was maintained or increased up to 4% H₂O vapor (ca. 100% RH at 20 °C). As shown in Figure 7, the catalytic activity of LDH-DPU increased with increasing water concentration and was 1.5 times higher at 4 vol% H₂O than that under dry conditions. Thus, LDH-DPU was found to be tolerant to high humidity and is a promising candidate for air purification catalysts.

3. Materials and Methods

3.1. Catalyst Preparation

3.1.1. Materials

Au/SiO₂, Au/NiO, Au/TiO₂, and Au/Fe₂O₃ (Au 1 wt%) were purchased from Haruta Gold Inc., Tokyo, Japan (https://www.haruta-gold.com/en/) (accessed on 20 July 2023). Aqueous solution of tetrachloroauric acid (0.497 g/L Au) was purchased from Tanaka Kikinzoku Kogyo K.K. and used as received. SiO₂ (Q-10) was supplied from Fuji silysia Co., Ltd., Aichi, Japan. Nickel (II) nitrate hexahydrate (Ni(NO₃)₂·6H₂O: 98%), super-dehydrated toluene (99.5+%), and sodium tetrahydroborate (NaBH₄: >95.0%) were purchased from FUJIFILM Wako Pure Chemical Corporation. Titanium (IV) chloride (TiCl₄: >99.0%) and urea (CO(NH₂)₂: >99.0%) were purchased from Kanto Chemical Co., Inc., Tokyo, Japan (https://www.kanto.co.jp/english/) (accessed on 20 July 2023). Other reagents were commercially available and used without further purification.

3.1.2. Ni-Ti LDH/SiO₂

Ni-Ti LDH/SiO₂ (Ni/Ti = 4) was prepared using hydrothermal synthesis method according to the literature [23,24] with modifications in the presence of SiO₂. Ni(NO₃)₂ 6H₂O (6 mmol), TiCl₄ (1.5 mmol), and urea (15 mmol) were dissolved in 30 mL of distilled water at room temperature, and the solution was stirred at room temperature for 2 h. SiO₂ (1.2 g) was added to the solution, and the mixture was transferred to a 50 mL Teflon-lined autoclave and aged at 130 °C for 48 h. The resulting solid was centrifuged, washed with water three times, and dried in air at 80 °C to obtain Ni-Ti LDH/SiO₂. The obtained Ni-Ti LDH/SiO₂ was calcined in air at 500 °C for 2 h to obtain Ni-Ti MMO/SiO₂. Bulk Ni-Ti LDH was prepared in the same manner in the absence of SiO₂.

3.1.3. Au/Ni-Ti MMO/SiO₂ (LDH-DP and MMO-DP)

Au/Ni-Ti MMO/SiO₂ (LDH-DP and MMO-DPU) was prepared by DP. Typically, an aqueous solution of HAuCl₄ (10 mM, 0.76 mL) dissolved in distilled water (6.8 mL) was heated to ca. 70 °C, and the pH of the solution was adjusted to 5 by adding 0.1 M NaOH aq. while stirring. A Ni-Ti MMO/SiO₂ support was added, and the pH of the suspension was adjusted to 7 by adding 0.1 M NaOH aq. After aging at 70 °C for 1 h, the precipitate was collected by centrifugation, washed five times with water at 40 °C, and dried in air at 80 °C overnight. The resulting solid was calcined in air at 300 °C for 4 h to obtain MMO-DP. Au/Ni-Ti MMO/SiO₂ (LDH-DP) was prepared in a similar manner using Ni-Ti LDH/SiO₂ as a support, followed by calcination in air at 500 °C for 2 h to convert LDH into MMO, after depositing Au.

3.1.4. Au/Ni-Ti MMO/SiO₂ (LDH-DPU and MMO-DPU)

Au/Ni-Ti MMO/SiO₂ (LDH-DPU and MMO-DPU) was prepared by DP-Urea according to the literature [27] with minor modifications. For LDH-DPU, an aqueous solution of HAuCl₄ (8.7 × 10⁻⁵ mol; 10 mM, 870 μL, Au 0.3 wt% initial loading) and 28% ammonia solution (1.2 mL) were dissolved in distilled water (200 mL), and the solution was stirred at room temperature for 4 h. Then, the Ni-Ti LDH/SiO₂ support (525 mg) and urea (2.1 g, 35 mmol) were added into the solution. After stirring the suspension at 75 °C for 4 h, the suspension was left overnight at room temperature. The precipitate was collected by centrifugation, washed three times with water at room temperature, and dried in air at 80 °C overnight. The resulting solid (436 mg) was dispersed in anhydrous toluene (109 mL), and NaBH₄ (54 mg) was added. The suspension was stirred at room temperature for 2 h. Then, EtOH (33 mL) was added, and the suspension was stirred at room temperature for an additional 6 h. After filtrating and washing with water and ethanol, the solid was dried in air at 80 °C overnight. Subsequently, the solid was calcined in an N₂ flow (50 mL min⁻¹) at 300, 400, and 500 °C for 2 h at a heating rate of 5 °C min⁻¹, to convert LDH to MMO, resulting in LDH-DPU-300, LDH-DPU-400, and LDH-DPU, respectively. Au/Ni-Ti MMO/SiO₂ (MMO-DPU) was also prepared by following a similar manner and using Ni-Ti MMO/SiO₂ as the support but without calcination in N₂ at 500 °C at the final step.

3.1.5. Au/bulk Ni-Ti MMO

Au/bulk Ni-Ti MMO was prepared by DP. The obtained bulk Ni-Ti LDH was converted to bulk Ni-Ti MMO by calcination at 500 °C for 2 h at a ramping rate of 5 °C min⁻¹. Subsequently, Au was deposited on bulk Ni-Ti MMO by DP. An aqueous solution of HAuCl₄ (25 mM, 0.30 mL) dissolved in distilled water (7.3 mL) was heated to 70 °C, and the pH of the solution was adjusted to 5 by adding 0.1 M NaOH aq. while stirring. The bulk Ni-Ti MMO was added, and the pH of the suspension was adjusted to 7 by adding 0.1 M NaOH aq. After aging at 70 °C for 1 h, the precipitate was collected by centrifugation, washed five times with water at 40 °C, and dried in air at 120 °C overnight. The resulting solid was calcined in air at 300 °C for 4 h to yield Au/bulk Ni-Ti MMO.

3.2. Characterization

XRD patterns were obtained using a Rigaku, MiniFlex with Cu Kα₁ radiation (λ = 0.154 0562 nm). The scan speed and step scan were 5° min⁻¹ at 0.02°, respectively. TGA was performed using a HITACHI STA7200. The samples were heated from room temperature to 900 °C at a ramping rate of 10 °C min⁻¹ and maintained at 900 °C for 30 min. The actual Au loadings were estimated by ICP-AES using a SPECTRO, CIROS-120, and SHIMADZU, ICPE-9820. The size of the Au particles was estimated HAADF-STEM using a JEOL, ARM-200F NEOARM operating at 200 kV at Tokyo Metropolitan University and an FEI, Titan G2 operating at 300 kV at Hokkaido University. The Au particles were counted more than 150 particles to calculate the mean diameters except for LDH-DPU. For LDH-DPU, it was hard to observe Au particles, and 70 particles were counted. The electronic states of Au, Ni, and Ti were evaluated by XPS using a JEOL, JPS-9010MX with Mg Kα radiation. The C 1s peak at 284.6 eV was used for charge correction in the XPS spectra. The CO-TPR was performed on a MicrotracBEL, BELCAT II to investigate the oxygen activation ability of the catalysts. The catalyst sample was pretreated in a flow of 20 vol% O₂ in He (30 mL min⁻¹) at 250 °C for 1 h and cooled to −100 °C. Then, 0.5 vol% CO in He (30 mL min⁻¹) was passed through the catalyst sample at a temperature range of −100–200 °C with a ramping rate of 10 °C min⁻¹. The amount of CO₂ generated was estimated by mass spectrometry using a MicrotracBEL, BELMASS.

3.3. CO Oxidation

The CO oxidation was performed in a continuous fixed-bed flow reactor. The catalyst sample (0.1 g) was placed in a U-shaped glass reactor and pre-treated in a flow of 20 vol% O₂ in N₂ (50 mL min⁻¹) at 200 °C for 1 h. After cooling to room temperature, 1 vol% CO in air (33.3 mL min⁻¹, SV 20,000 mL h⁻¹ g_cat⁻¹) was passed through the catalyst bed while heating from room temperature to 300 °C. The CO conversion and CO₂ yield were estimated by GC using an Agilent microGC. CO conversion was calculated using the following equation:

$$\mathrm{CO} \mathrm{conversion} \left(\mathrm{\%}\right)=\frac{\left({\mathrm{CO}}_{\mathrm{in}}-{\mathrm{CO}}_{\mathrm{out}}\right)}{{\mathrm{CO}}_{\mathrm{in}}}\times 100$$

The effect of water concentration on CO oxidation was evaluated using a MicrotracBEL, BELCAT II equipped with an Agilent microGC. The Au/Ni-Ti LDH/SiO₂ catalyst (200 mg) was pre-treated in N₂ (50 mL min⁻¹) at 500 °C for 2 h. After the pretreatment, the catalyst was cooled to room temperature in a flow of N₂ (50 mL min⁻¹), and then 1 vol% CO in air containing H₂O vapor (50 mL min⁻¹) was fed into the sample cell with water vapor at 20 °C.

4. Conclusions

The ultrafine Ni-Ti LDH particles were highly dispersed on SiO₂ and converted into small MMO crystallites. Au was deposited on Ni-Ti LDH/SiO₂ and Ni-Ti MMO/SiO₂ by DP and DPU, and LDHs were finally converted to MMOs. The Au particle size was affected by the preparation method and the support, revealing that the deposition of Au onto LDH by DPU was suitable. The abundant OH groups at the edge sites of LDHs contributed to stabilizing Au clusters. Calcination of LDH/SiO₂ after depositing Au by DPU affected the transformation of LDH into MMO, resulting in rich oxygen vacancies in the TiO_x domain. Thus, LDH-DPU exhibited the highest catalytic activity for CO oxidation among the Au/Ni-Ti MMO/SiO₂ catalysts and commercially available Au catalysts, despite low Au loading (0.3 wt%). Furthermore, the dispersion of ultrafine Ni-Ti MMO on SiO₂ reduced the amount of rare Ni species compared with conventional Au/NiO. In addition, LDH-DPU was active for CO oxidation at room temperature in humid air. These findings will pave the way for the development of highly efficient Au catalysts for air purification.