Nickel-Cerium Layered Double Hydroxide as Electrocatalyst for Glycerol Oxidation

A new nickel based layered double hydroxide nanomaterial, Ni0.8Ce0.2-LDH, constituted by about 3 nm large nanoparticles exhibiting enhanced electrocatalytic properties towards oxidation of glycerol in alkaline media, was prepared by incorporation of 20 wt.% of Ce ions into α-nickel hydroxide. This can be oxidized to Ce, as demonstrated by X-ray photoelectron spectroscopy, a much electron withdrawing and oxidizing species responsible for electronic interactions and additional electrocatalytic active sites giving rise to synergic effects. High-resolution transmission electron microscopy images and the X-ray diffractograms indicated a material with a significantly lower size and degree of crystallinity than α-Ni(OH)2, one of the factors contributing to its enhanced electrochemical and electrocatalytic activity, as demonstrated by cyclic voltammetry and electrochemical impedance spectroscopy, suggesting a good potentiality for development of glycerol fuel cells.


Introduction
The increasing energy demand and the limited availability of fossil fuels are pushing the development of higher performance and environmentally friendly energy conversion and storage devices, which are one of the key technological challenges in the 21 st century. 1 Fuels cells, electrochemical devices that convert chemical energy directly into electric energy, are promising alternatives to rechargeable batteries, 2 especially when cheap byproducts such as glycerol is employed.
One type of fuel cell, a direct glycerol fuel cell (DGFC) using glycerol or glycerol solution as fuel and proton source is attractive because its storage is much simple and safe as compared with hydrogen gas. 3 Moreover, glycerol is widely available from biodiesel production, being a resource produced in a renewable, more environmental-friendly, and cost effective manner. 4 In addition, its electrooxidation in DGFCs cogenerate electricity and large amounts of high value chemicals 5 as byproducts that can be utilized in the pharmaceutical, food, cosmetics and tobacco industries. 6 In fact, many investigations have been conducted on glycerol oxidation exploring different types of electrode materials. 7 Therefore, many research groups are focused on the development of electrocatalytic materials for this type of device. One class of such materials are the layered nanomaterials, that have attracted great attention due to the large interplanar distances that facilitate the intercalation of several compounds into the precursor materials, as well as their large surface areas 8 and surface concentrations of active sites, favoring the diffusion of reactants and enhanced catalytic performance.
Among the layered transition metal oxides and hydroxides, Ni(OH) 2 constitutes a class of electrocatalytic and electrochromic materials that have been extensively used in rechargeable batteries, 9 hybrid supercapacitors, 10 sensors 11,12 and fuel cells. 13 In fact, Ni(OH) 2 nanoparticles (NPs) wrapped by poly[Ni(salen)] film (Ni(salen): N,N'-bis(salicylidene)-ethylenediaminonickel(II) complex) was investigated as electrocatalyst for oxidation of methanol, ethanol and glycerol in alkaline electrolyte by Bott-Neto et al. 7 The results show that the Ni II -salen complex can be used as precursor of Ni(OH) 2 based NP catalysts with average diameter of 2.4 nm, uniformly dispersed in the polymeric matrix. In situ Fourier transform infrared (FTIR) spectroscopy studies confirmed that the electrocatalytic oxidation of those alcohols is dependent on the formation of the β-NiOOH species, and the oxidation products are similar to those produced in other nickel-based electrodes. In other words, β-NiOOH species is the electrocatalytic active species responsible for the oxidation of alcohols thus regenerating β-Ni(OH) 2 . However, it is important to highlight that the materials in the alpha phase (α-Ni(OH) 2 ) exhibit a much more disorganized structure but higher electrocatalytic activity and specific charge capacity. 14 Accordingly, the efficiency of alkaline fuel cells and energy storage devices tend to decrease as it is converted to crystalline β-Ni(OH) 2 . 15,16 Several studies have shown that the structural stability and electrochemical performance of modified α-Ni(OH) 2 are better than that of neat α-Ni(OH) 2 . 17 In fact, several studies have been carried out focusing on the isomorphic substitution of Ni II by others M II metal cations such as Co II , 18 Mn II , 19 and Zn II , 9 generating materials denominated double hydroxide salts (DHSs) and described by the  26,27 In addition to increased stability, the DHSs and LDHs materials exhibit enhanced synergistic electrochemical properties. In this sense, Ce III based layered nanomaterials have also attracted much attention due to its high chemical stability, and excellent electric conductivity. [28][29][30] In fact, Xu et al., 31 developed Ce III doped NiFe-LDH/CNT (CNT: carbon nanotube) nanoarrays by in situ self-assembly at room temperature. The ternary NiFeCe-LDH/CNT electrocatalyst obtained in such a way displayed high OER (oxygen evolution reaction) activity and a substantially lower overpotential (227 mV at 10 mA cm −2 ) and Tafel slope (33 mV dec −1 ) in alkaline medium, competing favorably with NiFe-based LDHs, and outperforming commercial Ir/C catalysts. The OER activity of the nanocomposite was significantly enhanced mainly because of the much larger specific surface area, high density of lattice defects, much better electron transport efficiency, and synergetic effects associated with the redox properties and larger coordination number of Ce III incorporated in the layered nanostructure.
Accordingly, Ce-based LDHs are known to be excellent electrocatalytic materials because of its ability to store and transfer an oxygen atom concomitantly with an oxidation equivalent as expected for the reversible Ce IV /Ce III process, its high oxygen carrying capacity and low price. 32,33 In this context, hereon we describe the preparation, characterization and enhanced electrocatalytic properties of a new Ni 0.8 Ce 0.2 -LDH and evidence for a possible direct participation of that rare earth ion in the glycerol electrooxidation reaction, in 1.0 mol L -1 NaOH.

Experimental
Materials and methods C e r i u m ( I I I ) a c e t a t e ( C e ( C H 3 C O O ) 3   The NiCe-LDH materials with increasing concentrations of Ce III ions exhibited similar structures and increasingly larger electrocatalytic activity but that become significantly larger only in Ni 0.8 Ce 0.2 -LDH as can be seen in the Figure S1 (Supplementary Information (SI) section), reason why now on this work will be focused on it. The α-Ni(OH) 2 NPs were prepared by dissolving 4.82 mmol of the nickel acetate in 25.0 mL of glycerol and adding stoichiometric amounts of KOH in n-butanol solution (9.64 mmol dissolved in 18.0 mL of n-butanol was added dropwise) at room temperature, according to a previously described procedure. 34,35 The Ni 0.8 Ce 0.2 -LDH samples were prepared in a similar way by reacting a solution of nickel acetate (3.86 mmol) and cerium acetate (0.96 mmol) in 25.0 mL of glycerol with stoichiometric amounts of KOH in n-butanol solution (10.60 mmol in 18.0 mL of n-butanol). This solution was very quickly added into the former one at room temperature.
Preparation of α-Ni(OH) 2 and Ni 0.8 Ce 0.2 -LDH modified FTO electrodes FTO (fluorine doped tin oxide) glass pieces (1.0 × 2.5 cm) were carefully washed with soap and rinsed with isopropanol and water. Then, a 1 cm 2 area was delimited on the surface with a scotch tape and 50 µL of a nanomaterial suspension was spin-coated, the tape removed and the modified electrode heated at 240 °C, for 30 min, to produce adherent and mechanically stable films.

Characterization
The samples were characterized by X-ray diffractometry (XRD) in a Bruker D2 Phaser equipment (Germany) with a Cu Kα source (λ = 1.5418 Å, 30 kV, 15 mA, step = 0.05º) and air-scatter screens to avoid diffuse scattering at lower angles, in the 2θ range from 5 to 70º. Samples were prepared on glass substrates by depositing 50 µL of a nanoparticle suspension and heating at 240 °C for 30 min.
Chemical surface analyses were carried out by X-ray photoelectron spectroscopy (XPS), using a K-α X-ray photoelectron spectrometer (Thermo Fisher Scientific, UK) equipped with a hemispherical electron analyzer and monochromatic Al Kα (1486.6 eV) radiation source. Survey (full-range) and high-resolution spectra of Ce and Ni were acquired using pass energy of 200 and 50 eV, respectively, and the data analyses carried out using the Thermo Avantage Software (version 5.921). The XPS results presented in this work correspond to the average of at least three independent measurements performed in different regions of each sample.
Transmission electron microscopy (TEM) and highresolution transmission electron microscopy (HRTEM) images were obtained in a JEOL JSM-FEG 7401F equipment (Japan), at an accelerating voltage of 200 kV. Samples were prepared on copper grids (Ted Pella) by dispersing 3 µL of a nanoparticle suspension diluted in deionized water.

Electrochemical measurements
All electrochemical measurements were conducted using a PGSTAT 30 electrochemical workstation (Autolab, Metrohm, Netherlands) and a classical three electrodes configuration cell at room temperature, and a 1.0 mol L -1 NaOH aqueous solution as electrolyte. FTO electrodes modified with α-Ni(OH) 2 and Ni 0.8 Ce 0.2 -LDH were employed as the working electrode ( Figure S2, SI section), while a platinum foil and Ag/AgCl (KCl, 3.0 mol L -1 ) were respectively used as the counter and reference electrodes.
The activity of the nanocatalysts for glycerol electrooxidation was measured by cyclic voltammetry in 1.0 mol L -1 NaOH solution containing 1.9 mmol L -1 of glycerol, at room temperature. The CV curves were recorded in the potential range from 0.10 to 0.55 V, at scan rate of 50 mV s -1 .
The electrochemical impedance spectra were recorded at 0.5 V using the same electrochemical arrangement, modulating the frequency of the sinusoidal potential wave (amplitude = 10 mV) superimposed to the AC potential from 100 kHz to 0.01 Hz.

Results and Discussion
Physicochemical characterization TEM images were used to estimate the α-Ni(OH) 2 and Ni 0.8 Ce 0.2 -LDH nanoparticles size and morphology using samples prepared by depositing them on graphene oxide sheets, as shown in Figure 1. It is interesting to note that Ni 0.8 Ce 0.2 -LDH NPs are much smaller suggesting that Ce III ions probably are introducing a significant local distortion/stress to the lattice generating defective sites. A similar influence of cerium ions has also been observed by Xu et al. 31 in the NiFeCe-LDH/CNT composite, in which 5 wt.% of Ce III ions was shown to be enough to reduce the average NPs size. It is worth noting here the possibility of Ni 0.8 Ce 0.2 -LDH conversion to (NiCe)O x by the electron beam due to the local heating and the high vacuum in the sample chamber, generating oxide NPs. 36 In our work, 20 wt.% of Ce III ions in α-Ni(OH) 2 decreased the nanoparticle size from 6 to ca. 3 nm. Such effect can be ascribed to the contrasting NPs nucleation rates and/or lattice distortion caused by the Ce III ions. In fact, a faster nucleation can favor the growth of smaller crystals, as previously demonstrated for incorporation of lanthanide ions. 37 In addition, the incorporation of much larger Ce III (102 pm as compared to 72 pm of Ni II ) ions into Ni(OH) 2 producing Ni 0.8 Ce 0.2 -LDH tend to generate local stress and lattice distortions that may hamper the growth of layered double hydroxide nanocrystals.
The structural information of Ni(OH) 2 and Ni 0.8 Ce 0.2 -LDH NPs was obtained by XRD measurements, as shown in the diffractograms of the respective samples prepared by depositing the materials on glass substrates ( Figure 2). The Ni(OH) 2 sample presented a peak at 10.50° which can be indexed to the (003) plane characteristic of the material in the α-phase. 38 In fact, this is the most characteristic feature of the significantly more disordered α-phase material, as compared to the crystalline β-phase material exhibiting several sharp peaks in the diffractogram. Similarly, considering an isomorphic substitution of Ni II by Ce III ions, the main diffraction peak of Ni 0.8 Ce 0.2 -LDH at 10.42º was assigned to the respective layered double hydroxide material. The incorporation of Ce III ions (102 pm), much larger than Ni II (72 pm), did not change significantly the interplanar distances, and the basal spacing of α-Ni(OH) 2 (d 003 ca. 0.84 nm) was not significantly increased in Ni 0.8 Ce 0.2 -LDH (d 003 ca. 0.86 nm). Nevertheless, the clearly apparent lower intensity and broadness of that XRD peak is characteristic of a less crystalline material, and/or a material constituted by smaller size particles, as expected. The formation of Ni 0.8 Ce 0.2 -LDH nanoparticles was confirmed by STEM (scanning transmission electron microscopy), where the bright field image ( Figure S4a, transmitted electrons, BF) and dark field ( Figure S4b, scattered in high-angle annular dark field, HAADF) exhibited similar contrast and no clear indication of segregation of the Ni(OH) 2 and Ce(OH) 3 phases. 25 These results indicates that 20 wt.% of Ce III ions was incorporated into the α-Ni(OH) 2 lattice without any segregation or phase separation, generating a completely homogeneous material, demonstrating the suitability of the sol-gel method. In fact, this method is attracting considerable attention because employs mild conditions and allows good and easy control of the structure, composition and the degree of homogeneity. 38 The eventual presence of nickel hydroxide derivative in the more organized β phase was further evaluated by FTIR data as shown in SI section ( Figure S5). The broad band at 650 cm −1 in the α-Ni(OH) 2 spectrum is characteristic of the d(OH), but this same IR-active band appears at 500 cm -1 in β-Ni(OH) 2 , 39,40 confirming the presence of nickel hydroxide in the α-phase. The incorporation of Ce III ions into α-Ni(OH) 2 decreased the intensity of that band and led to appearance of two sharper bands at 605 and 669 cm -1 probably associated with d(OH) bond to cerium and nickel sites next to cerium ions as expected for the NiCe-LDH material. In addition, the two bands at 1558 and 1417 cm -1 has been assigned to ν a (OCO) and ν s (OCO) stretching vibrations of organic acid anions, respectively, 41 and that at 1383 cm -1 must be assigned to d(CH) or ρ r (COO) stretching/bending vibrations of CH and carboxylate groups, 7 respectively. These carboxylatecontaining species were probably produced during the heat treatment of the samples (α-Ni(OH) 2 and NiCe-LDH films) by oxidative breakdown of glycerol molecules, which were used as solvent during the synthetic procedure, thus generating smaller compounds such as carbonate and oxalate. As reported by Rives, 42 the presence of intercalated carbonate and oxalate anions in these class of compounds induce interlayer spacing about 0.80 nm for carbonate and in the range of 0.80-0.90 nm for oxalate anions, which are in accordance with the present work (0.84 and 0.86 nm). However, the presence of interlayered acetate anions (from precursor salts) coordinated perpendicularly to the plane of the layer has been ruled out, since which would increase the interlamellar spacing to ca. 1.30 nm as described by Choy et al. 43 XPS measurements were carried out for determination of the valence states of Ni and Ce in the nanomaterial. The Ni 2p region of α-Ni(OH) 2 and Ni 0.8 Ce 0.2 -LDH (Figure 3a) shows the typical Ni II peaks respectively at 873.7 eV (Ni 2p 1/2 ) and 855.9 eV (Ni 2p 3/2 ), and 873.3 eV (Ni 2p 1/2 ) and 855.7 eV (Ni 2p 3/2 ), with their corresponding satellites, in good agreement with the spectrum of a standard α-Ni(OH) 2 sample exhibiting peaks at 873.3 eV (Ni 2p 1/2 ) and 855.6 eV (Ni 2p 3/2 ). The Ce 3d high resolution spectrum (Figure 3b reference, indicating a significantly lower electron binding energy for the Ce III ions as expected for the presence of less electronegative Ni II ions bond next to them. Also, it is interesting to note that a weak peak characteristic of CeO 2 like sites appeared at 917 eV after heat treatment at 240 ºC for 30 min (Figure 3c) indicating that site is available for redox processes. This is a very interesting feature since Ce IV is an potent oxidizing species that besides creating new catalytic active sites may induce synergic effects enhancing the electrocatalytic activity of the NiCe-LDH material for oxidation reactions, as compared to pure α-Ni(OH) 2 .

Electrochemical properties
The electrochemical performance of α-Ni(OH) 2 and Ni 0.8 Ce 0.2 -LDH NPs modified FTO electrodes were evaluated by cyclic voltammetry (CV) measurements within a potential window from 0.10 to 0.55 V (vs. Ag/AgCl 3.0 mol L -1 ) in aqueous 1.0 mol L -1 NaOH solution. The electrochemical behavior during 100 CV cycles at 50 mV s -1 is shown in Figure 4. Note that both materials present an increase in the current as a function of the number of CV cycles. This behavior was attributed to the progressive diffusion of the electrolyte solution into the nanomaterial activating more and more Ni II sites, as previously reported. 44 The oxidation peak was found at 0.43 V (vs. Hg/HgO, KOH 1 M), a potentially typically assigned to the α-nickel hydroxide materials, while that peak appears shifted to potentials more positive than about 0.5 V (vs. Hg/HgO, KOH 1 M) in the β phase material. 45 It is easy to see that the Ni 0.8 Ce 0.2 -LDH/FTO electrode exhibits much broad waves than the α-Ni(OH) 2 /FTO electrode but with similar peak currents, indicating that the amount of exchanged electrons is much larger, or in other words, that the concentration of redox active sites is much larger. This is a very important feature of an electrocatalyst and the wave broadening suggests that the activated Ni II sites probably are in different chemical environments in the LDH lattice. Such a broadening can also be consequence of the resistance to electrolyte diffusion as observed when the thickness of the film deposited on FTO was increased, as previously reported, 25 but this is not the case here. The anodic peak at around 0.35 V is attributed to the oxidation of α-Ni(OH) 2 to γ-NiOOH, while the cathodic peak at 0.26 V is related to the reverse reaction, according to the following electrochemical process: 46,47 The electrocatalytic activity of α-Ni(OH) 2 and Ni 0.8 Ce 0.2 -LDH modified FTO electrodes for oxidation of glycerol was investigated by cyclic voltammetry in 1.0 mol L -1 NaOH solution (Figures 5a and 5c) since LDH derivatives prepared with lower Ce III content did not exhibit significantly larger electrocatalytic activity as compared to pure α-nickel hydroxide ( Figure S1, SI section). The CV curves in the presence of 1.9 mmol L -1 of glycerol clearly demonstrates the much larger electrocatalytic activity of the LDH material, which presented 1.7 times larger current at 0.50 V.  Accordingly, electrochemical impedance spectroscopy (EIS) measurements were performed to clarify the electrochemical properties at 0.50 V, using an amplitude AC perturbation of 10 mV, in the 0.01 Hz to 100 kHz range, and the respective Nyquist plots are shown in Figures 5b  and 5d. The equivalent circuit used to fit the experimental data is shown in the inset of Figure 5d, in which R s represents the combined resistance of the electrolyte, the intrinsic resistance of FTO, and the contact resistance of the nanomaterials with the FTO substrate; R 1 represents the charge-diffusion resistance in the modified electrodes, and R 2 the heterogeneous charge-transfer resistance for oxidation of glycerol (or water) on the electrode surface. The impedance parameters obtained by fitting the experimental data are listed in Table 1.
The R 1 values are relatively low as expected spanning the 7.1 to 12.0 ohms range, whilst R 2 was significantly different for α-Ni(OH) 2 and Ni 0.8 Ce 0.2 -LDH, in the absence and presence of glycerol. In fact, the heterogeneous charge-transfer resistance decreased from about 367 ohms in the pure nickel hydroxide to 230 ohms in the LDH material, as expected for its higher electrocatalytic activity. In addition, it is relevant to mention the quite significant decrease of such parameter, from 230 to 185 ohms for LDH (and from 367 to 344 ohms for α-Ni(OH) 2 ) upon addition of glycerol, indicating that this compound is more easily oxidized than water. 48,49 Such an improvement in performance suggests that the presence of Ce III ions in the LDH material is beneficial and is increasing the heterogeneous charge-transfer kinetics. The possibility of improvement of the materials conductivity can be ruled out since no significant change of R 1 was observed upon incorporation of 20 wt.% of Ce III in the α-Ni(OH) 2 matrix ( Table 1).
As pointed out previously, Ce III ion is much larger than the Ni II ion and can induce local structural distortions that can enhance the activity of the nickel sites. However, we demonstrated by XPS that the Ce III ions can also be oxidized to Ce IV ions in the NiOOH structure, a species that is wellknown for its oxidizing properties. Accordingly, the glycerol electrooxidation mechanism may include the simultaneous action of both, the nickel and cerium sites. First, both sites are activated by oxidation to their respective higher valence states, then the electrochemically generated Ni III and Ce IV species oxidize glycerol and are reduced back regenerating the starting species, according to equations 2-5. Thus, the enhanced electrocatalytic activity was attributed to lattice distortions and the presence of catalytic active Ce IV sites, that can act in parallel or synergistic way. This last hypothesis is considered more probable given its very high electron withdrawing and electronegativity that also can activate further the electrocatalytic activity of the nickel sites for oxidation reactions.
Ni(OH) 2 + OH -→ NiOOH + H 2 O + e -(2) NiOOH + Glycerol → products + Ni(OH) 2 (3) Ce(OH) 3 + OH -→ CeO 2 + 2H 2 O + e -(4) CeO 2 + Glycerol → products + Ce(OH) 3 (5) To better evaluate the electrocatalytic activity of α-Ni(OH) 2 and NiCe-LDH electrodes toward glycerol oxidation, the results obtained in this work were compared with related works in the literature. 6,7,50,51 As shown in Table 2, the lowest oxidation potential of 0.50 V, that indicates a larger electrocatalytic activity, was achieved with the nanomaterials described in this work. The other important parameter is the current density, but this parameter is highly dependent on the substrate concentration and the actual electrode surface area available for the electrocatalytic process and are not directly comparable. In our case, a thin film with relatively low surface area and a 1.9 mM glycerol solution has been considered, while electrodes modified with thicker films with much larger active surface areas and 100 mM glycerol solution was used in the other works. For example, a [Ni(salen)] derivative was electropolymerized on glassy carbon electrode surface generating a thick poly[Ni(salen)] ATV film exhibiting the highest current density of 31.4 mA cm -2 . Accordingly, just considering that the substrate concentration is 50 times  larger, the expected current density for NiCe-LDH modified electrodes in the same experimental conditions should be larger than 100 mA cm -2 , demonstrating its superior electrocatalytic activity.

Conclusions
The incorporation of Ce III ions into α-Ni(OH) 2 lattice influenced the nanoparticles size, degree of crystallinity and electrocatalytic activity of Ni 0.8 Ce 0.2 -LDH, as demonstrated by TEM, XRD and electrochemical (CV and EIS) measurements. The larger Ce III ions probably are distorting the nickel hydroxide lattice and was shown to be easily oxidized generating Ce IV ion, a highly oxidizing and electron-withdrawing species that can act as additional electrocatalytic active site and induce synergic effects enhancing the electrocatalytic activity of the LDH material for oxidation of organic substrates. In fact, the layered double hydroxide material showed an enhanced electrocatalytic activity 1.7 times larger than α-Ni(OH) 2 for oxidation of glycerol, being promising for application in glycerol-based fuel cells.

Supplementary Information
Supplementary data (cyclic voltammetry, electrochemical impedance, ICP-OES (inductively coupled plasma optical emission spectrometry) and FTIR) are available free of charge at http://jbcs.sbq.org.br as PDF file.