Production of COx-Free Hydrogen and Few-Layer Graphene Nanoplatelets by Catalytic Decomposition of Methane over Ni-Lignin-Derived Nanoparticles

Nickel (Ni)-lignin nanocomposites were synthesized from nickel nitrate and kraft lignin then catalytically graphitized to few-layer graphene-encapsulated nickel nanoparticles (Ni@G). Ni@G nanoparticles were used for catalytic decomposition of methane (CDM) to produce COx-free hydrogen and graphene nanoplatelets. Ni@G showed high catalytic activity for methane decomposition at temperatures of 800 to 900 °C and exhibited long-term stability of 600 min time-on-stream (TOS) without apparent deactivation. The catalytic stability may be attributed to the nickel dispersion in the Ni@G sample. During the CDM reaction process, graphene shells over Ni@G nanoparticles were cracked and peeled off the nickel cores at high temperature. Both the exposed nickel nanoparticles and the cracked graphene shells may participate the CDM reaction, making Ni@G samples highly active for CDM reaction. The vacancy defects and edges in the cracked graphene shells serve as the active sites for methane decomposition. The edges are continuously regenerated by methane molecules through CDM reaction.


Introduction
Hydrogen proton-exchange membrane fuel cells (H 2 -PEMFCs) are promising energy conversion devices for electric vehicles and portable appliances. However, the platinum catalysts in the electrodes of PEMFCs are easily poisoned by trace amounts of carbon monoxide (CO) in the hydrogen. Therefore, only high purity (100 ppm or less of CO) hydrogen is desired for use in H 2 -PEMFC devices [1]. More than 50% of hydrogen is presently produced from steam reforming of natural gas, which simultaneously generates CO and CO 2 (Equation (1)). CH 4 + H 2 O → CO + 3 H 2 , ∆H 298 = 206 kJ/mol (1) COx (CO and CO 2 ) must be removed through a series of purification processes before obtaining high purity hydrogen [2]. First, the water-gas shift reaction (WGSR) is performed to convert CO to CO 2 over a catalyst (usually an iron oxide catalyst) (Equation (2)): CO + H 2 O → CO 2 + H 2 , ∆H 298 = −41 kJ/mol (2) Then, CO 2 is separated from hydrogen using the pressure swing adsorption (PSA) process, where adsorbent materials (zeolites, activated carbon, molecular sieves, etc.) are used to remove CO 2 at high pressure [3]. Significant amounts of energy and materials are consumed in these purification processes.
Powdered kraft lignin (BioChoice, Plymouth, NC, USA) was supplied by Domtar. The ash content in the kraft lignin was 1.65 wt% determined using ASTM D1102 standard. The number average molecular weight (M n ) and weight average molecular weight (M w ) of kraft lignin were determined to be M n = 988 ± 56 g/mol and M w = 6672 ± 315 g/mol, respectively, measured using gel permeation chromatography (GPC). Lignin elemental analysis was performed with an elemental analyzer. Nickel nitrate hexahydrate (Ni(NO 3 ) 2 ·6H 2 O) and tetrahydrofuran (THF, anhydrous, ≥99.9%, inhibitor-free) were purchased from Sigma-Aldrich, Inc. (St. Louis, MO, USA).

Preparation of Ni-Lignin Composites
Ni-lignin composites were prepared using the co-precipitation method. Five solutions of nickel nitrate were prepared by adding 13.0 g, 26.6 g, 56.2 g, 89.2 g, and 126.4 g nickel nitrate hexahydrate to 15 mL, 30 mL, 60 mL, 90 mL, and 120 mL of DI water, respectively, in individual 500 mL glass beakers and stirring for 30 min. Each of these five nickel nitrate solutions was added dropwise to its respective tetrahydrofuran and kraft lignin solution (100 g lignin in 100 mL tetrahydrofuran) and the final mixtures were all stirred for 2 h. The mixtures were naturally dried at room temperature for one week. The prepared Ni-lignin composites were labelled as 2.5% Ni-lignin, 5% Ni-lignin, 10% Ni-lignin, 15% Ni-lignin, and 20% Ni-lignin, respectively.

Pretreatment of Ni-Lignin Composites
Thermal decomposition of Ni-lignin composites has previously been observed as a strongly exothermic process. A significant amount of heat is released during the reaction which may trigger a thermal runaway situation [16]. Therefore, the dried Ni-lignin composites were first thermally decomposed in a muffle furnace before loading into a reactor for the catalytic graphitization process. Nitrogen gas was first introduced into the furnace at a flow rate of 100 mL/min for 30 min. The furnace temperature was increased to 300 • C at a rate of 2.5 • C/min and held at 300 • C for 0.5 h, then naturally cooled to ambient temperature under nitrogen flow. The decomposed sample was loaded into a ball mill machine and ground at 1000 rpm for 10 min.

Catalytic Carbonization of Ni-Lignin Composites to Graphene-Encapsulated Nickel Nanoparticles (GENNs, Labelled as Ni@G)
Fifty grams (50 g) of the decomposed Ni-lignin samples was each packed in the middle of a 1 inch OD ceramic tubular reactor. The carrier gas-argon (99.99% purity) was first introduced into the reactor at a flow rate of 100 mL/min for 30 min. The reactor was heated at a temperature-programmed rate of 10 • C/min to a carbonization temperature (600, 700, 800, or 900 • C) and held at the carbonization temperature for 1 h. The furnace was cooled down by 10 • C/min to room temperature under an argon flow. An on-line Hiden QGA quantitative gas analysis system (Hiden Analytical, Livonia, MI, USA) was used to measure gaseous products during the graphitization process. The signals from the mass spectra of 2, 15, 28, and 44 (m/z) were identified as major contributors from the evolved gases and volatiles, and determined to be H 2 , CH 4 , CO, and CO 2 , respectively. The produced Ni@G samples were labeled as Ni@G-600 • C, Ni@G-700 • C, Ni@G-800 • C, and Ni@G-900 • C.

Temperature-Programmed Catalytic Decomposition of Methane (TPCDM)
Temperature-programmed catalytic decomposition of methane (TPCDM) was performed in a fixed-bed reactor system equipped with on-line mass spectroscopy. The Ni@G sample (10 g) was first purged by flowing with Ar (50 mL/min) at room temperature for 0.5 h. Then, the reactant gas mixture of 100 mL/min methane and 50 mL/min argon were introduced. The reactor was heated from 25 to 900 • C at a heating rate of 10 • C/min, during which the m/z intensities at 2 (H 2 ), 16 (CH 4 ), 28 (CO), and 44 (CO 2 ) in the effluents were recorded by the online mass spectrometer.

Catalytic Testing for Methane Decomposition
The catalytic performance for methane decomposition of each catalyst was tested in a ceramic 1-inch (O.D.) tubular fixed-bed reactor at a temperature range of 700-900 • C. The preferred amount of catalyst (i.e., 2.5, 5, 10, 15, and 20 g) was loaded in the middle of the reactor. Catalytic testing was performed using 100 mL/min CH 4 and 50 mL/min Ar. Outlet gases were analyzed using the online mass spectrometer. The argon in the reaction gas was used as a diluent and as an internal analysis standard. Pressure of the reactor system was recorded by a pressure transducer and compared with time on stream. The differential pressure sensor was connected to the reactor to measure the difference in gas pressure before and after the reactor and this pressure difference (∆P) was monitored and recorded. Catalytic activity was evaluated in terms of methane conversion. We defined methane conversion (M conv ) as where M in represents the total quantity (moles) of methane fed into the reactor and M out represents the quantity (moles) of methane out of the reactor. The solid carbon in the fresh Ni@G and the Ni@G sample after CDM process was measured by TGA. Twenty milligrams of the sample (20 mg) was put on a ceramic sample pan for TPO analysis in a Shimadzu TGA-50H instrument. By heating the sample under high flow of air (100 mL/min) from room temperature to 800 • C using a 10 • C/min heating ramp rate, the weight change was recorded as related to carbon burning. C 1 is defined as the carbon in the Ni@G sample originated from lignin, and C m is the carbon in the sample contributed by methane after CDM reaction.

Characterization
Nickel content (wt%) of the graphitized Ni-lignin samples was measured using an inductively coupled plasma-optical emission spectrometer (ICP-OES, Agilent 5800, Santa Clara, CA, USA). Ni@G samples were mineralized with nitric acid before ICP-OES measurement. Surface area measurement of the Ni@G samples was determined using an automatic adsorption unit (Autosorb-1, Quantachrome, Boynton Beach, FL, USA). The specific surface areas were calculated using the Brunauer-Emmett-Teller (BET) analysis method to the respective N 2 adsorption isotherms. X-ray powder diffraction (XRD) patterns of the samples were obtained using a Rigaku Ultima III X-ray Diffraction System (Rigaku, TheWoodlands, TX, USA). The morphology of the samples was investigated with a JEOL 6500F Field Emission Scanning Electron Microscope (SEM, Peabody, MA, USA). The sample particle sizes were examined with a JEOL JEM-100CX II Transmission Electron Microscope (TEM, Peabody, MA, USA). Raman Spectroscopy measurements were carried out using a Jobin-Yvon microspectrometer (Edison, NJ, USA) equipped with an excitation laser source emitting at 514 nm and an incident power around 1 mW on a thin surface. Twenty spectra were collected for each sample.

Results and Discussion
3.1. Catalytic Carbonization of Ni-Lignin Composite to Few-Layer Graphene Encapsulated Nickel Nanoparticles (Ni@G) Figure 1a shows the trends of vent gas species (CO 2 , CO, CH 4 , and H 2 ) during temperature-programmed catalytic carbonization of the Ni-lignin sample. Significant amounts of CO 2 , CO, CH 4 , and H 2 were released from the Ni-lignin sample between the temperatures of 165 and 300 • C. This implied that kraft lignin was catalytically decomposed by nickel ions. Carboxyl, carbonyl, and ether groups in the phenylpropane side chains will decompose and release as CO 2 , CO, and CH 4 when heating Ni-lignin at the lower temperature. CO 2 released in the low temperature range is attributed to the catalytic decomposition of carboxyl groups in kraft lignin [17] and/or catalytic oxidation of functional groups in lignin by Ni 2+ ions [16]. The CO formation in the low temperature range is mainly contributed by decomposition of carbonyl and/or catalytic oxidation of functional groups in lignin by Ni 2+ ions, lignin char is formed after the decomposition of the most of functional groups in lignin. In the second stage, the formed char was catalytically carbonized to carbon from 300 • C to 1000 • C. The strong CO peak at 555 • C was assigned to ether decomposition, implying that nickel ions promoted the hydrolysis of the ether groups. The weak intensity CO 2 peak at 490 • C was assigned to the carbothermal reduction of nickel oxide. Hydrogen evolution was observed when the temperature was above 490 • C and was assigned to the cracking of -CH x (x = 1-3) groups promoted by nickel metal. In the present work, nickel oxide dissolved in the lignin matrix was first reduced by surface functional groups of the lignin, then the reduced metallic nickel reacted with amorphous carbon to form Ni@C nanostructures (NiO + active functional groups → Ni + CO 2 + CO + H 2 O).  TG (thermogravimetric analysis) and DTG (derivative of TG) curves of the Ni-kraft lignin sample are plotted in Figure 1b. As shown, the catalytic decomposition process could be divided into five stages corresponding to five mass loss steps shown in the TG curves: (1) water evaporation, (2) decomposition of nickel nitrate and de-polymerization of side chain structures, (3) decomposition of aromatic ring structures and formation of lignin char, (4) reduction of nickel oxide by active carbon species in lignin char, and (5) catalytic graphitization of lignin char by nickel particles [18].
The FTIR spectra of kraft lignin is plotted in Figure 1c,d and compared to that of the Ni-lignin composites. A strong peak at 3368 cm −1 and two small peaks at 2937 and 2841 cm −1 are assigned to stretching vibrations of hydroxyl, methyl and methylene groups in lignin, respectively [19,20]. The IR peak at 1710 cm −1 is associate with carbonyl groups. Peaks at 1597, 1511, and 1417 cm −1 are due to stretching vibrations of aromatic rings in lignin [13]. The stretch absorptions of C-C, C-O, and C=O are located at 1265 and 1215 cm −1 , respectively [16,19]. The IR band at 1078 cm −1 is associated with C-O deformation of secondary alcohol of the side chains [21].
FTIR spectra of the fresh Ni-lignin composites is plotted in Figure 1c,d and compared to that of kraft lignin. Due to the formation of bonds between nickel ions and the functional groups in lignin molecules, the intensity and location of IR bands were found to be reduced and/or shifted in the Ni-lignin composite. FTIR spectra of Ni-lignin samples heated at 300 °C, 500 °C, and 800 °C are also presented in Figure 1c,d. Most of the TG (thermogravimetric analysis) and DTG (derivative of TG) curves of the Ni-kraft lignin sample are plotted in Figure 1b. As shown, the catalytic decomposition process could be divided into five stages corresponding to five mass loss steps shown in the TG curves: (1) water evaporation, (2) decomposition of nickel nitrate and de-polymerization of side chain structures, (3) decomposition of aromatic ring structures and formation of lignin char, (4) reduction of nickel oxide by active carbon species in lignin char, and (5) catalytic graphitization of lignin char by nickel particles [18].
The FTIR spectra of kraft lignin is plotted in Figure 1c,d and compared to that of the Ni-lignin composites. A strong peak at 3368 cm −1 and two small peaks at 2937 and 2841 cm −1 are assigned to stretching vibrations of hydroxyl, methyl and methylene groups in lignin, respectively [19,20]. The IR peak at 1710 cm −1 is associate with carbonyl groups. Peaks at 1597, 1511, and 1417 cm −1 are due to stretching vibrations of aromatic rings in lignin [13]. The stretch absorptions of C-C, C-O, and C=O are located at 1265 and 1215 cm −1 , respectively [16,19]. The IR band at 1078 cm −1 is associated with C-O deformation of secondary alcohol of the side chains [21].
FTIR spectra of the fresh Ni-lignin composites is plotted in Figure 1c,d and compared to that of kraft lignin. Due to the formation of bonds between nickel ions and the functional groups in lignin molecules, the intensity and location of IR bands were found to be reduced and/or shifted in the Ni-lignin composite. FTIR spectra of Ni-lignin samples heated at 300 • C, 500 • C, and 800 • C are also presented in Figure 1c,d. Most of the functional groups in kraft lignin were decomposed after heating the sample at 300 • C as most of the FTIR bands disappeared or significantly decreased, with the exception of aromatic ring structures, which were partially broken when the sample was thermally treated at 300 • C.
Almost all the FTIR bands disappeared after thermally treating the samples at 500 • C and 800 • C, indicating the structure of kraft lignin was decayed due to catalytic carbonization. FTIR results in Figure

Characterization of Ni@G Samples
Inductively coupled plasma-optical emission spectrometry (ICP-OES) was used to measure the nickel content in the Ni@G samples and the results are listed in Table 1. The Ni contents of 10% Ni-lignin composites, were 22.8%, 23.6%, 24.9%, and 25.5% for Ni@G samples carbonized at 600 • C, 700 • C, 800 • C, and 900 • C, respectively. The nickel contents of 2.5%, 5%, 10%, 15%, and 20% Ni-lignin composites were 6.9%, 11.8%, 25.5%, 34.2%, and 43.5%, respectively, in Ni@G samples carbonized at 900 • C. The average crystallite size of metallic Ni in the Ni@G products was calculated using the Scherrer equation. Table 1. Catalytic graphitization of Ni-lignin composites to graphene-encapsulated nickel nanoparticles (Ni@G). The "@G" indicates the graphene (carbon)-encapsulation.  Table 1 indicates that BET surface areas were 57.4, 83.7, 117.5, 91.6, and 73.5 m 2 /g for kraft lignin materials with nickel loadings of 2.5%, 5%, 10%, 15%, and 20%, respectively. This indicates that with increased nickel loading, surface areas increased from 2.5% to 10% loading, and decreased with 15% to 20% loading.  (220) planes, indicating the Ni has a polycrystalline structure; the XRD profile clearly shows the nickel phase is majority face-centered cubic structure. At 2θ = 26 • a single peak attributed to (002) graphite is present for the samples carbonized at 700 • C, 800 • C, and 900 • C. X-ray photoelectron spectroscopy (XPS) provides information about the chemical state of elements. XPS spectra of the calcined Ni-lignin and the 10% Ni-lignin carbonized at 800 • C are plotted in Figure S1. Ni2p3/2 peaks for both samples were very weak, possibly because the nickel particles were encapsulated in carbon or graphene.

Ni-Lignin Precursor
Raman spectroscopy can identify the presence of graphite and disordered amorphous carbon in the samples. Figure 2b shows Raman spectra of the carbonized 10% Ni-lignin samples under different temperatures. The A D /A G values of the prepared samples were 1.50, 1.46, 1.39, and 1.29 for Ni-lignin samples carbonized at 600 • C, 700 • C, 800 • C, and 900 • C, respectively; therefore, the degree of graphitization of these four samples was in the order of 600 • C < 700 • C < 800 • C < 900 • C. Raman spectroscopy can identify the presence of graphite and disordered amorphous carbon in the samples. Figure 2b shows Raman spectra of the carbonized 10% Nilignin samples under different temperatures. The AD/AG values of the prepared samples were 1.50, 1.46, 1.39, and 1.29 for Ni-lignin samples carbonized at 600 °C, 700 °C, 800 °C, and 900 °C, respectively; therefore, the degree of graphitization of these four samples was in the order of 600 °C < 700 °C < 800 °C < 900 °C.

High-Resolution Transmission Electron Microscopy
The microstructure and morphology of 10% Ni-lignin samples graphitized at temperatures between 600 °C and 900 °C were investigated by HRTEM. Figure 2c shows dark colored Ni nanoparticles (XRD results in Figure 2a) embedded in the light-colored amorphous lignin char matrix in the Ni-lignin-600 sample. One to two layers of graphitic

High-Resolution Transmission Electron Microscopy
The microstructure and morphology of 10% Ni-lignin samples graphitized at temperatures between 600 • C and 900 • C were investigated by HRTEM. Figure 2c shows dark colored Ni nanoparticles (XRD results in Figure 2a) embedded in the light-colored amorphous lignin char matrix in the Ni-lignin-600 sample. One to two layers of graphitic carbon were observed to form around nickel nanoparticles of the Ni-lignin sample carbonized at 700 • C, indicating the minimum temperature for the formation of Ni@G structure was 700 • C; these Ni@G structures were surrounded by amorphous carbon. Similar Ni@G structures formed for the Ni-lignin sample carbonized at 800 • C, where the graphitic carbon around nickel nanoparticles increased to 2-4 layers and the amorphous carbon converted to the turbostratic stacking structure [22]. When the carbonization temperature increased to 900 • C, the graphitic shell increased to 3-8 layers and almost all the amorphous carbon was converted to the turbostratic stacking structure. The effect of nickel loading on the morphology of the Ni@G samples was investigated using high resolution transmission electron spectroscopy (HRTEM). The images of Ni@G structures from 2.5%, 5%, 10%, 15%, and 20 wt% Ni-lignin carbonized at 900 • C are shown in Figure 3. The nickel particles in Ni@G samples from 2.5%, 5%, and 10 wt% Ni-lignin composites were homogenous and well dispersed in the turbostratic stacking structure; these Ni@G nanoparticles were 2-8 nm in diameter. The morphologies of Ni@G samples from 15% and 20% Ni-lignin composites were significantly different from those observed in 2.5%, 5%, and 10% Ni-lignin composites in terms of particle sizes and sample phases.
increased to 900 °C, the graphitic shell increased to 3-8 layers and almost all the a phous carbon was converted to the turbostratic stacking structure.
The effect of nickel loading on the morphology of the Ni@G samples was investig using high resolution transmission electron spectroscopy (HRTEM). The images of N structures from 2.5%, 5%, 10%, 15%, and 20 wt% Ni-lignin carbonized at 900 °C are sh in Figure 3. The nickel particles in Ni@G samples from 2.5%, 5%, and 10 wt% Ni-l composites were homogenous and well dispersed in the turbostratic stacking struc these Ni@G nanoparticles were 2-8 nm in diameter. The morphologies of Ni@G sam from 15% and 20% Ni-lignin composites were significantly different from those obse in 2.5%, 5%, and 10% Ni-lignin composites in terms of particle sizes and sample pha HRTEM micrographs (Figure 3d,e) show that increasing the nickel loading percenta 15 and 20 wt% resulted in metal agglomeration and lower dispersion of the Ni@G n particles in the carbon matrix. This results in a negative effect on the activity and sta of the catalyst sample.

Temperature-Programmed Catalytic Decomposition of Methane (TPCDM) ove Ni@G Nanoparticles
Temperature programmed reaction experiments were carried out to validate i Ni@G materials were active for methane decomposition. Temperature-programmed C reaction was performed by flowing Ar-CH4 mixture (50 mL/min Ar-100 mL/min  Temperature programmed reaction experiments were carried out to validate if the Ni@G materials were active for methane decomposition. Temperature-programmed CDM reaction was performed by flowing Ar-CH 4 mixture (50 mL/min Ar-100 mL/min CH 4 ) through a fixed-bed reactor loaded with 10 g of 25.5% Ni@G material. The reactor was ramped with a heating rate of 10 • C/min to 900 • C. The flow rates of methane and hydrogen in the vent gas, the conversion rate of methane, and the production rate of hydrogen were plotted in Figure 4. CH 4 decomposition over 25.5% Ni@G was initiated at 584.1 • C (Figure 4). Ni@G exhibited very poor activity at low temperature and only a small amount of methane was consumed at 584.1 • C; hydrogen production was initially detected at 598.5 • C. The methane consumption rate accelerated with increase of heating temperature and reached the maximum at 831.2 • C. through a fixed-bed reactor loaded with 10 g of 25.5% Ni@G material. The reactor was ramped with a heating rate of 10 °C/min to 900 °C. The flow rates of methane and hydrogen in the vent gas, the conversion rate of methane, and the production rate of hydrogen were plotted in Figure 4. CH4 decomposition over 25.5% Ni@G was initiated at 584.1 °C (Figure 4). Ni@G exhibited very poor activity at low temperature and only a small amount of methane was consumed at 584.1 °C; hydrogen production was initially detected at 598.5 °C. The methane consumption rate accelerated with increase of heating temperature and reached the maximum at 831.2 °C. The structures and C/Ni mass ratios of graphene products from Ni@G samples are observed to depend on the structure of Ni@G samples. Ni@G prepared at different carbonization temperatures, using different nickel contents, or different process parameters like CDM reaction temperature, mass of the loaded Ni@G samples, or reaction time, will influence the structure. The effects of the Ni@G structures and CDM reaction process parameters on graphene products were investigated in the CDM process. The structures of Ni@G, CDM reaction process conditions, and the structure of the graphene products are listed in Table 2. The structures and C/Ni mass ratios of graphene products from Ni@G samples are observed to depend on the structure of Ni@G samples. Ni@G prepared at different carbonization temperatures, using different nickel contents, or different process parameters like CDM reaction temperature, mass of the loaded Ni@G samples, or reaction time, will influence the structure. The effects of the Ni@G structures and CDM reaction process parameters on graphene products were investigated in the CDM process. The structures of Ni@G, CDM reaction process conditions, and the structure of the graphene products are listed in Table 2. C l : carbon contributed from lignin; C m : carbon from decomposition of methane; C t = C l + C m .

Effect of Carbonization Temperature
Catalytic activity of Ni@G materials from the Ni-lignin composite carbonized under different temperatures was investigated with the CDM reaction, and the results are shown in Figure 5. Figure 5a exhibits the activity of Ni@G prepared at the carbonization temperature of 600 • C, 700 • C, 800 • C, and 900 • C. The results show the conversion of CH 4 was in the following order: Ni@G-600 • C > Ni@G-700 • C > Ni@G-800 • C > Ni@G-900 • C. The Ni@G-600 • C and Ni@G-700 • C catalysts show higher initial methane conversion, and the conversion rate decreases slowly with time on stream. The Ni@G-900 • C sample demonstrates the lowest activity; however, the methane conversion rate remains steady under the testing conditions.
There are several possible explanations for why the Ni@G-600 • C and Ni@G-700 • C exhibit high initial methane conversion rates. (1) Due to lower carbonization temperatures, nickel nanoparticles in both samples are relatively small: 5.7 nm for Ni@G-600 • C and 8.5 nm for Ni@G-700 • C (Table 1). Furthermore, HRTEM images ( Figure 2) show these nickel nanoparticles as uniformly distributed in the amorphous carbon matrix and no graphene (carbon) shells are observed to encapsulate the nickel nanoparticles; therefore, they are active and ready to decompose methane molecules. (2) Because of lower carbonization temperatures, lignin char in the Ni@G-600 • C and Ni@G-700 • C samples are only partially carbonized, and the carbon-based structures in both samples are mainly amorphous carbon. Amorphous carbon structures have proved to be the most active carbon-based materials for catalytic decomposition of methane [4,23]. (3) Ni@G-600 • C and Ni@G-700 • C samples were prepared at the carbonization temperatures of 600 • C and 700 • C, respectively. Lignin char was not fully carbonized in either sample, therefore, significant amounts of oxygencontaining functional groups remained. At the beginning of the CDM reaction, amorphous carbon structures in Ni@G-600 • C and Ni@G-700 • C samples will continue the carbonization process and produce Cox (CO and CO 2 ) ( Figure S2). Initial hydrogen production rates (Figure 5b) of the Ni@G-600 • C and Ni@G-700 • C samples are lower due to the production of Cox (CO and CO 2 ).
Molecules 2022, 27, x FOR PEER REVIEW Figure 5. CH4 conversion (a) and H2 production (b) with Ni@G catalysts carbonized temperature: 10 g catalyst was loaded in the reactor, testing temperature: 800 °C; gas f mL/min. There are several possible explanations for why the Ni@G-600 °C and N exhibit high initial methane conversion rates. (1) Due to lower carbonizatio tures, nickel nanoparticles in both samples are relatively small: 5.7 nm for N and 8.5 nm for Ni@G-700 °C (Table 1). Furthermore, HRTEM images ( Figure 2) nickel nanoparticles as uniformly distributed in the amorphous carbon matrix phene (carbon) shells are observed to encapsulate the nickel nanoparticles; the are active and ready to decompose methane molecules. (2) Because of lower ca temperatures, lignin char in the Ni@G-600 °C and Ni@G-700 °C samples are on carbonized, and the carbon-based structures in both samples are mainly amo bon. Amorphous carbon structures have proved to be the most active carbon

Effect of the Reaction Temperature
The activity of methane decomposition over 25.5% Ni@G at different reaction temperatures (700 • C, 800 • C, and 900 • C) as a function of time on stream was studied, and the results are displayed in Figure 6. Only hydrogen and unreacted methane were detected in the vent gases. As shown in Figure 6, CH 4 conversion and H 2 production rates increase with increase of the reaction temperature from 700 to 900 • C. An initial methane conversion of 15.8% was observed for the reaction at 700 • C, and the methane conversion rate reached a steady state of 18.3% after 45 min. The initial methane conversion rate of 42.3% was observed at 800 • C, and it increased to a stable level of 45% after 25 min. The methane conversion rate was 88.1% from the beginning at 900 • C and there was no induction time present at the higher reaction temperature.
Molecules 2022, 27, x FOR PEER REVIEW 12 hydrogen production rates (Figure 5b) of the Ni@G-600 °C and Ni@G-700 °C sample lower due to the production of Cox (CO and CO2).

Effect of the Reaction Temperature
The activity of methane decomposition over 25.5% Ni@G at different reaction peratures (700 °C, 800 °C, and 900 °C) as a function of time on stream was studied the results are displayed in Figure 6. Only hydrogen and unreacted methane wer tected in the vent gases. As shown in Figure 6, CH4 conversion and H2 production increase with increase of the reaction temperature from 700 to 900 °C. An initial met conversion of 15.8% was observed for the reaction at 700 °C, and the methane conve rate reached a steady state of 18.3% after 45 min. The initial methane conversion ra 42.3% was observed at 800 °C, and it increased to a stable level of 45% after 25 min methane conversion rate was 88.1% from the beginning at 900 °C and there was no in tion time present at the higher reaction temperature.

Effect of the Amount of Catalyst
The effects of Ni@G sample mass were examined for the CDM reaction. Figure lustrates the methane conversion and hydrogen production rates when the reaction is formed with 5, 10, 15, and 20 g of 25.5% Ni@G at 800 °C. The results revealed that conversion and hydrogen production rate increased with increase of the catalyst mas to longer residence time of the methane when more catalyst is loaded in the reactor. I also noticed that methane conversion and hydrogen production rates remained con and no deactivation of the catalyst was observed under the operating conditions. Figure 6. Effect of the reaction temperature on catalytic decomposition of methane over 25.5% Ni@G nanoparticles. Catalyst used: 10 g; gas flow rate: 100 mL/min; testing temperature: 700 • C, 800 • C, and 900 • C.

Effect of the Amount of Catalyst
The effects of Ni@G sample mass were examined for the CDM reaction. Figure 7 illustrates the methane conversion and hydrogen production rates when the reaction is performed with 5, 10, 15, and 20 g of 25.5% Ni@G at 800 • C. The results revealed that CH 4 conversion and hydrogen production rate increased with increase of the catalyst mass due to longer residence time of the methane when more catalyst is loaded in the reactor. It was also noticed that methane conversion and hydrogen production rates remained constant, and no deactivation of the catalyst was observed under the operating conditions.

Effect of Ni@G Nickel Content
Five Ni@G samples with different nickel contents were obtained by graphitiz Ni-lignin composites with nickel loading of 2.5%, 5%, 10%, 15%, and 20%. After c graphitization at 900 °C for 1 h, the nickel contents in the corresponding Ni@G s were 6.9%, 11.8%, 25.5%, 34.2%, and 43.5%, respectively. The activity of Ni@G for c decomposition of methane at 800 °C for 250 min is plotted in Figure 8. As shown in 8, both the CH4 conversion rate and the H2 production rate increase with the inc nickel contents in Ni@G samples. Ni@G samples with 6.9%, 11.8%, and 25.5% Ni c showed relatively lower initial methane conversion and hydrogen production ra good stability under the reaction conditions. Ni@G samples with 34.2% and 43.5% tents exhibited high initial activity but the methane conversion and hydrogen pro rates decreased gradually with time-on-stream. The decrease in activity may be ca

Effect of Ni@G Nickel Content
Five Ni@G samples with different nickel contents were obtained by graphitization of Ni-lignin composites with nickel loading of 2.5%, 5%, 10%, 15%, and 20%. After catalytic graphitization at 900 • C for 1 h, the nickel contents in the corresponding Ni@G samples were 6.9%, 11.8%, 25.5%, 34.2%, and 43.5%, respectively. The activity of Ni@G for catalytic decomposition of methane at 800 • C for 250 min is plotted in Figure 8. As shown in Figure 8, both the CH 4 conversion rate and the H 2 production rate increase with the increase of nickel contents in Ni@G samples. Ni@G samples with 6.9%, 11.8%, and 25.5% Ni contents showed relatively lower initial methane conversion and hydrogen production rates, but good stability under the reaction conditions. Ni@G samples with 34.2% and 43.5% Ni contents exhibited high initial activity but the methane conversion and hydrogen production rates decreased gradually with time-on-stream. The decrease in activity may be caused by the agglomeration of the Ni particles at high carbonization temperature, resulting in large Ni particle size and a lowering of the surface area of the nickel catalyst. Furthermore, the low carbon content in high Ni loading samples may also contribute to the deactivation of the catalyst during CDM reaction since graphene is another active part of the Ni@G sample in the decomposition of methane.

Stability of the Catalyst
To evaluate the stability of the Ni@G materials, 25.5% Ni@G and 10% Ni/Al compared in the CDM reaction. Both the CH4 conversion rate and the pressure d across the catalyst beds are presented as a function of time-on-stream (TOS) in F The 10% Ni/Al2O3 catalyst exhibited unstable activity. The initial methane conver was 78.1% after 10 min, then increased to 79.5% after 20 min, then kept at ~79.0% min TOS. After 50 min TOS, the catalyst began to show continuous deactivation a

Stability of the Catalyst
To evaluate the stability of the Ni@G materials, 25.5% Ni@G and 10% Ni/Al 2 O 3 were compared in the CDM reaction. Both the CH 4 conversion rate and the pressure drop (∆P) across the catalyst beds are presented as a function of time-on-stream (TOS) in Figure 9.
The 10% Ni/Al 2 O 3 catalyst exhibited unstable activity. The initial methane conversion rate was 78.1% after 10 min, then increased to 79.5% after 20 min, then kept at~79.0% after 40 min TOS. After 50 min TOS, the catalyst began to show continuous deactivation and after 150 min, the conversion rate decayed to 43.7%. The pressure drops across the Ni/Al 2 O 3 catalyst bed also changed with TOS. The initial ∆P of Ni/Al 2 O 3 was 7 mbar at 10 min TOS. It increased to 18 mbar after 50 min TOS. The pressure drops then increased very quickly, reaching 537 mbar at 150 min TOS; the reactor was clogged and the CDM reaction was paused. The higher initial methane conversion of Ni/Al 2 O 3 catalyst could be attributed to the larger number of Ni sites [2,24]. However, because of high CDM activity, significant carbon deposits were formed over the catalyst and resulted in the reactor plugging.
Molecules 2022, 27, x FOR PEER REVIEW 15 of 2 was 5 mbar at 10 min TOS and increased slowly to 20 mbar after 600 min TOS. The cata lytic stability may be attributed to the nickel dispersion in the Ni@G sample. The cracked graphene shells may also participate in the CDM reaction as the vacancy defects and edge in the cracked graphene shells can serve as the active sites for methane decomposition [25][26][27]. The vacancies and edge carbon atoms with free bonds in the cracked graphen shells tend to react with methane molecules by catching the carbon atoms to stabilize thei structures. The active sites in vacancy defects will be consumed over TOS, while the car bon atoms on the edges will continuously be regenerated with new atoms decomposed from methane molecules through CDM reaction. Therefore, CDM activity over the Ni@G sample remained stable during the 600 min of TOS.

Sample Characterization after CDM Reaction
HRTEM images of the Ni@G samples after CDM reaction for 1 h at different reaction temperatures (600, 700, 800, and 900 °C) are showed in Figure 10. There is no obviou change in the Ni@G sample cracked at 600 °C; the Ni@G particles keep their original struc tures, i.e., the nickel particles are captured in 1-5 layers of graphene shells which are no yet cracked (Figure 10a). As the temperature increased to 700 °C, the HRTEM image show The Ni@G catalyst showed relatively lower activity compared to the Ni/Al 2 O 3 catalyst. The CH 4 conversion rate was 60.5% at 10 min TOS and increased to 63.1% after 20 min TOS. Gradually, the methane conversion rate became stable at about 60% during 600 min TOS.
The pressure drop for the Ni@G catalyst was very low. Initially, the ∆P of Ni@G was 5 mbar at 10 min TOS and increased slowly to 20 mbar after 600 min TOS. The catalytic stability may be attributed to the nickel dispersion in the Ni@G sample. The cracked graphene shells may also participate in the CDM reaction as the vacancy defects and edges in the cracked graphene shells can serve as the active sites for methane decomposition [25][26][27]. The vacancies and edge carbon atoms with free bonds in the cracked graphene shells tend to react with methane molecules by catching the carbon atoms to stabilize their structures. The active sites in vacancy defects will be consumed over TOS, while the carbon atoms on the edges will continuously be regenerated with new atoms decomposed from methane molecules through CDM reaction. Therefore, CDM activity over the Ni@G sample remained stable during the 600 min of TOS.

Sample Characterization after CDM Reaction
HRTEM images of the Ni@G samples after CDM reaction for 1 h at different reaction temperatures (600, 700, 800, and 900 • C) are showed in Figure 10. There is no obvious change in the Ni@G sample cracked at 600 • C; the Ni@G particles keep their original structures, i.e., the nickel particles are captured in 1-5 layers of graphene shells which are not yet cracked ( Figure 10a). As the temperature increased to 700 • C, the HRTEM image shows the graphene shells of the Ni@G on the sample surface are cracked and partially peeled off the nickel cores due to the CDM reaction ( Figure 10b). When the temperature increases to 800 • C, the core-shell structures of Ni@G are cracked and the graphene shells are skinned off the nickel cores to form graphene nanoplatelets (Figure 10c). The size of the nickel particles was significantly increased to 10-50 nm due to merging of the naked nickel cores. HRTEM image of the Ni@G samples after CDM reaction for 1 h at 900 • C (Figure 10d) shows more large graphene nanoplatelets are formed and the nickel particles are sintering to even larger particles. off the nickel cores due to the CDM reaction ( Figure 10b). When the temperature increases to 800 °C, the core-shell structures of Ni@G are cracked and the graphene shells are skinned off the nickel cores to form graphene nanoplatelets (Figure 10c). The size of the nickel particles was significantly increased to 10-50 nm due to merging of the naked nickel cores. HRTEM image of the Ni@G samples after CDM reaction for 1 h at 900 °C ( Figure  10d) shows more large graphene nanoplatelets are formed and the nickel particles are sintering to even larger particles.  Figure 11a shows the XRD patterns of the products of Ni@G after CDM reaction for different reaction times (60 min, 120 min, 180 min, and 300 min) at 800 °C. The relative intensity of the graphene diffraction peak (2θ = 26.5°) increases with increased reaction time. A weak and flat diffraction peak at ~26.5° is observed for the graphene sample at 60 min, and the peak becomes strong and sharp with increase of reaction time. This indicates the graphene nanoplatelets grow with improved quality with prolonged reaction time.  Figure 11a shows the XRD patterns of the products of Ni@G after CDM reaction for different reaction times (60 min, 120 min, 180 min, and 300 min) at 800 • C. The relative intensity of the graphene diffraction peak (2θ = 26.5 • ) increases with increased reaction time. A weak and flat diffraction peak at~26.5 • is observed for the graphene sample at 60 min, and the peak becomes strong and sharp with increase of reaction time. This indicates the graphene nanoplatelets grow with improved quality with prolonged reaction time. Quality of the graphene produced from Ni@G through CDM reaction at 800 °C with different reaction times was analyzed by Raman spectroscopy (Figure 11b). Quality of grown graphene materials was assessed by the intensity ratio between bands D and G (ID/IG). The ID/IG ratio decreased with the prolonged reaction time, indicating a decrease in defects present in the graphene materials. It is also noticed that the I2D/IG ratio decreased Figure 11. XRD patterns (a) and Raman spectra (b) after CDM reaction at 800 • C with different reaction time.
Quality of the graphene produced from Ni@G through CDM reaction at 800 • C with different reaction times was analyzed by Raman spectroscopy (Figure 11b). Quality of grown graphene materials was assessed by the intensity ratio between bands D and G (I D /I G ). The I D /I G ratio decreased with the prolonged reaction time, indicating a decrease in defects present in the graphene materials. It is also noticed that the I 2D /I G ratio decreased with increase of reaction time, indicating there are more graphene layers in graphene samples produced with longer reaction time.
Fresh Ni@G and Ni@G samples after CDM reaction with different reaction times were boiled in 5 M HNO 3 solution to remove nickel particles and other inorganic impurities. This was followed by rinsing the sample with deionized water and oven drying at 105 • C overnight. HRTEM image (Figure 12a) of the purified fresh Ni@G sample shows it is mainly composed of empty graphene shells after nickel particles were removed. Almost all the nickel particles are removed from the Ni@G samples from different reaction times after acid purification. After CDM reaction for 30 min, the graphene shells in the Ni@G samples are cracked and peeled off the nickel particles and are randomly distributed in various directions. Methane molecules decompose to carbon atoms and hydrogen over both the surface of the nickel particles and the cracked graphene shells. The carbon atoms deposited over nickel particles will form graphene nanoplatelets while the carbon atoms deposited over the cracked graphene shells may fill vacancy defects in the graphene shells or react along the edges of the graphene shells to joint or merge these graphene shell units together. However, few or no joint reactions occur among them due to the short time process, therefore, fluffy graphene is the main structure in the products. Fluffy graphene is usually made up of 1 to 5 layers of graphene with an in-plane size of 30-50 nm (Figure 12b). HRTEM image (Figure 12c) of the purified graphene sample after 60 min of CDM reaction shows the formation of 3D fluffy graphene structures. This suggests welding or jointing reactions happened between the cracked graphene shells during the CDM process; the welding action mainly involves bonding along the edges of the cracked graphene shell units. HRTEM image (Figure 12d) reveals graphene nanoplates are the main products of the Ni@G after 120 min CDM reaction. The cracked graphene shells are combined along both the edges and the in-plane directions of units to form graphene nanoplates which show a thickness between 1 and 2 nm and an in-plane size of 100-300 nm. When the CDM reaction runs over Ni@G samples for a prolonged time, i.e., 180 min and 300 min (Figure 12e,f, respectively), 3D graphitic nanochips are the main structure of the graphene products. These graphitic structures are produced when more cracked graphene shells are welded along the horizontal edge and the in-plane directions due to prolonged reaction time. The graphitic nanochips show an in-plane size of 0.5-2 microns with a thickness range of 5-10 nm.

Possible Decomposition Mechanism of Methane over Ni@G
Nickel is known to be an excellent active metal component for the decomposition of methane [18]; however, Ni@G exhibited very poor activity in the decomposition of methane at low temperature. This is because most of the nickel particles are encapsuled by several layers of graphene, and it is very difficult for CH4 to diffuse and penetrate through the graphene shell if there are no defective cracks [26]. Fortunately, there are some naked nickel particles or Ni@G with defects in the graphene shell. As the reaction temperature is increased, a small amount of methane will diffuse along the cracks in the graphene shell of Ni@G and adsorb onto the surface of the nickel core, followed by the dissociation reaction of methane to hydrogen and carbon [26,28]. The carbon atoms will deposit on the nickel surface while hydrogen atoms may either combine and desorb as H2 to enter the gas phase or diffuse and dissolve into the nickel core of the Ni@G [26,28]. As the smallest molecule and an excellent decarburization reagent, the hydrogen molecules can easily diffuse and penetrate through the few-layer graphene shell then react with carbon dissolved in the nickel particles to form methane. With increasing of the heating temperature, more methane decomposition reactions occur over Ni@G. Consequently, more hydrogen

Possible Decomposition Mechanism of Methane over Ni@G
Nickel is known to be an excellent active metal component for the decomposition of methane [18]; however, Ni@G exhibited very poor activity in the decomposition of methane at low temperature. This is because most of the nickel particles are encapsuled by several layers of graphene, and it is very difficult for CH 4 to diffuse and penetrate through the graphene shell if there are no defective cracks [26]. Fortunately, there are some naked nickel particles or Ni@G with defects in the graphene shell. As the reaction temperature is increased, a small amount of methane will diffuse along the cracks in the graphene shell of Ni@G and adsorb onto the surface of the nickel core, followed by the dissociation reaction of methane to hydrogen and carbon [26,28]. The carbon atoms will deposit on the nickel surface while hydrogen atoms may either combine and desorb as H 2 to enter the gas phase or diffuse and dissolve into the nickel core of the Ni@G [26,28]. As the smallest molecule and an excellent decarburization reagent, the hydrogen molecules can easily diffuse and penetrate through the few-layer graphene shell then react with carbon dissolved in the nickel particles to form methane. With increasing of the heating temperature, more methane decomposition reactions occur over Ni@G. Consequently, more hydrogen penetrates the graphene shells of Ni@G and decarburizes the dissolved carbon in nickel cores to methane. Due to its larger size compared to hydrogen, methane molecules from the decarburization process cannot diffuse through the few-layer graphene shells and are trapped in the interface between the graphene shell and the nickel core. This results in the build-up of methane pressure inside the Ni@G structure which causes the outer graphene shell to break up. With the graphene shell cracked, the nickel core will be naked and exposed to the gaseous phase and more nickel active sites will be available for decomposition of methane. Simultaneously, the cracked graphene shells will serve as a carbon-based catalyst for decomposition of methane. The carbon atoms from methane decomposition will serve as an atomic glue to join the cracked graphene shells and can also deposit to vacancy or defect sites in the cracked graphene shells to improve the quality of the graphene product. The graphene shell cracking process can explain the required induction period at lower reaction temperatures ( Figure 6) as it will take some time for methane molecules to diffuse and dissociate over nickel core particles. Raman results (Figure 11b) indicate a decrease in defects and more layers of graphene produced in the graphene material samples produced with longer reaction time. HRTEM images (Figure 12) show the cracked graphene shells are combined along both the edges and the in-plane directions to form graphene nanoplates which become larger and thicker after prolonged CDM reaction time.
Overall, two reaction processes may simultaneously occur during the CDM reaction over Ni@G: (1) the cracking of Ni@G by methane molecules at the beginning of the reaction. and (2) after the cracking step, the exposed nickel particles and the cracked graphene shells will serve as the active catalyst components for CDM reaction.

Conclusions
Ni-lignin nanocomposites were prepared from nickel nitrate and kraft lignin. These Nilignin nanocomposites were catalytically graphitized to few-layer graphene-encapsulated nickel nanoparticles (Ni@G). Ni@G nanoparticles were used for catalytic decomposition of methane (CDM) to produce COx-free hydrogen and graphene nanoplatelets. Ni@G catalysts exhibit high activity and long-term stability for use in the CDM reaction.
During the CDM reaction process, the graphene shells of the Ni@G nanoparticles were cracked and peeled off the nickel cores at high temperature. Both the exposed nickel nanoparticles and the cracked graphene shells may participate in the CDM reaction, making Ni@G samples highly active for CDM reaction. The vacancy defects and edges in the cracked graphene shells serve as active sites for methane decomposition and the edges are continuously regenerated through the CDM reaction. Therefore, Ni@G samples remained stable for CDM reaction during 600 min TOS. Graphene nanoplatelets, fluffy graphene, 3D fluffy graphene, and 3D graphitic nanochips were formed as the main solid products in the CDM reaction over Ni@G catalyst.
Supplementary Materials: The following supporting information can be downloaded. Figure S1: The XPS spectra for Ni2p3/2 states of the calcined Ni-lignin at 300 • C and the carbonized Ni-lignin at 800 • C. Figure S2: TPCDM over Ni@G of Ni-lignin carbonized at 600 • C.