Single rhodium atoms anchored in micropores for efficient transformation of methane under mild conditions

Catalytic transformation of CH4 under a mild condition is significant for efficient utilization of shale gas under the circumstance of switching raw materials of chemical industries to shale gas. Here, we report the transformation of CH4 to acetic acid and methanol through coupling of CH4, CO and O2 on single-site Rh1O5 anchored in microporous aluminosilicates in solution at ≤150 °C. The activity of these singly dispersed precious metal sites for production of organic oxygenates can reach about 0.10 acetic acid molecules on a Rh1O5 site per second at 150 °C with a selectivity of ~70% for production of acetic acid. It is higher than the activity of free Rh cations by >1000 times. Computational studies suggest that the first C–H bond of CH4 is activated by Rh1O5 anchored on the wall of micropores of ZSM-5; the formed CH3 then couples with CO and OH, to produce acetic acid over a low activation barrier.


ICP-AES measurements of concentrations of Rh in catalysts.
ICP-AES was used in the measurements of Rh concentration in catalysts before and after catalysis. Four standard solutions with different concentration of Rh 3+ (0.1ppm, 1ppm, 5ppm, 10ppm) were prepared by dissolving Rh(NO3)3 into de-ionized water. The volume of each solution is 40 ml. The standard curve was built through measuring the four solutions under the exactly same setup and parameters of the ICP-AES (mode: JY 2000 2 manufacture: HORIBA) and then plotting the known concentrations of the four solutions as a function of the optical emission spectrometry intensity. Supplementary Figure  11 is the plot of the known concentration of the solution as a function of the atomic emission spectrometry intensity. It is the standard curve for the measurements of concentration of Rh in the fresh and used catalysts.
To prepare a test solution of ICP-AES analysis, certain amount of fresh or used catalyst (0.10wt%Rh/ZSM-5) was dissolved in NaOH solution through simply mixing the accurately weighed catalyst into 10mL 1M NaOH solution and then sonicating the mixture for about 1 hr. Then, aqua regia (mixture of nitric acid and hydrochloric acid) was added to the solution until the pH was less than 5. The transparent solution was diluted by adding DI water to make the volume of the diluted solution to 30 mL. All test solution was tested under the under the exactly same setup and parameters of the same ICP-AES.

Isotope-labelled experiments using 13 CH 3 OH.
To test whether acetic acid could form from carbonylaiton of CH3OH on our catalyst 0.10wt%Rh/ZSM-5, 1.0 mmol isotope-labeled 13 CH3OH (99 atom% 13 C, Aldrich) was added to 10 ml deionized H2O before introduction of 10 bar CH4, 5 bar CO, and 4 bar O2 to the Parr reactor. The purpose of adding isotope-labeled 13 CH3OH to H2O before catalysis is to test whether 13 CH3OH could act as an intermediate to react with CO on the catalyst to form isotope-labelled acetic acid, 13 CH3COOH. In reaction pathway β, acetic acid is formed through carboxylation of methanol ( Figure 5a); methanol should be involved definitely. The reaction pathway α in Figure 5a does not involve CH3OH. Figs. 5b presents the three sets of possible products if 13 CH3OH was added as a probe agent. Figure 5c and 5d are a regular 1 H NMR spectrum of solution obtained from 10 bar CH4, 5 bar CO and 4 bar O2 and the 1 H NMR spectrum of solution obtained from 10 bar CH4, 5 bar CO and 4 bar O2 with added 1 mmol 13 CH3OH, respectively.
Calculations of TOR of 0.10wt%Rh/ZSM-5. Catalytic activities of these catalysts reported in literatures 1,2 and the Rh1O5@ZSM-5 of this work were calculated in terms of the number of product molecules per Rh site per second. They are listed in Table 1 of the main text. The following paragraphs will describe how they were calculated.
For our catalyst, Rh1O5@ZMS-5 (0.10wt%Rh/ZSM-5), as shown in Figure 2, 846 μmol of acetic acid was formed from 28 mg of catalyst at 150 o C for 12 hrs under the mixture of mixture of 50 bar CH4, 10 bar CO, and 8 bar O2; the concentration of rhodium in the catalyst is 0.10wt%. The amount of all Rh atoms is × × . % = 2.8 × 10 . By assuming all Rh atoms anchored to ZSM-5 participate into this catalysis, TOR for production of acetic acid can be calculated as the following: With the same calculation method of TOR, TORs of acetic acid and organic oxygentates under catalysis condition 1 (mixture of 50 bar CH4, 10 bar CO and 8 bar O2 for 2 hrs) were calculated with the yields of acetic acid and organic oxygentates presented in Figure 2; these TORs were listed in entry 1 of Table 1.
Calculations of TOR of Rh cations without any support in aqueous solution. A similar experiment was performed. 5 ml of 0.01 mol/l Rh(NO3)3 was added in Parr reactor and then 50 bar CH4, 10 bar CO and 8 bar O2 was introduced the Parr reactor. The reaction was performed at 150 o C for about 90 hrs. With the same calculation method, the TORs were calculated. The TORs of all organic products (CH3COOH, CH3OH and HCOOH) and TOR of acetic acid are 2.4×10 -5 organic molecules and 6.3×10 -6 acetic acid molecules per Rh site per second generated from homogeneous catalyst Rh(NO3)3 without any promoter. They are listed in entry 3 of Table 1.
Methods of DFT calculations. The periodic density functional theory (DFT) calculations were performed using the Vienna ab initio Simulation Package (VASP). 3,4 The Perdew-Burke-Ernzerhof (PBE) 5 functional of generalized-gradient approximation (GGA) was used for the electron exchange and correlation. The D3 method for van der Waals correction by Grimme is used. 6 The electron-core interaction was described using the projector-augmented wave method (PAW). 7,8 The kinetic energy cutoff was set to 450 eV for the plane wave basis set, and the Brillouin zone was sampled using the gamma point only. A section of a relaxed ZSM-5 framework was used in a cluster model and the dangling bonds capped with hydrogen. The ZSM-5 cluster was placed in a 18×18×22 Å 3 box. The Rh site and first neighbors were allowed to relax during the subsequent calculations with the rest of the cluster fixed. The adsorption energies were calculated using Eads = Ecluster+adsorbate -(Ecluster + Eadsorbate), where the energy of the adsorbate Eadsorbate was computed by placing the adsorbate in a 15 Å wide cubic cell. Transition states were found using the climbing image nudged elastic band method implemented in VASP, using eight images and a force convergence criterion of 0.05 eV Å −1 . 9

Supplementary Note
Ready separation of products from solvent by using dodecane. Transformation of CH4, CO and O2 to organic oxygenates was performed at 150 o C for 2 hrs on 28 mg of catalyst while solvent dodecane was used. The yields of acetic acid and formic acid in dodecane under a mixture of 30 bar CH4, 10 bar CO and 5 bar O2 at 150 o C for 4 hrs are 225 μmol and 82 μmol, respectively (Supplementary Figure 8). The advantage of using dodecane is the ready separation of hydrophilic product molecules from hydrophobic solvent molecules.

Supplementary Discussion
Does acetic acid form from reaction of CO with formic acid? To test whether acetic acid could form through reaction between formic acid and CO, we performed three experiments by adding 20 mg 0.10wt%Rh/ZSM-5 into 10 ml H2O, dispersing HCOOH into 10 ml DI H2O and introducing 0 bar CO, 5 bar CO or 10 bar CO and then heating the solution to 150 o C and remaining it at 150 o C for 3 hrs. As shown in Supplementary Figure 7, there was no any acetic acid formed in the experiments. Thus, formation of acetic acid from coupling between formic acid and CO is not a possible pathway for synthesis of acetic acid from CH4, CO and O2.
Does acetic acid form from dry reforming of CH 4 ? Direct reforming CH4 with CO2 to produce acetic acid at a temperature ≥250 o C was reported in literatures. [10][11][12][13][14] Presumably, one potential reaction pathway for the formation of acetic acid on our catalyst is that CO could be first oxidized by O2 to form CO2 and then CO2 could couple with CH4 to form acetic acid. To check this possibility, 30 bar CH4 and 30 bar CO2 were introduced to the Parr reactor and the reaction was performed under the same catalytic condition on 28 mg 0.10wt%Rh/ZSM-5 (at 150 o C for 5 hrs). As shown in Supplementary Figure 6d, no acetic acid, formic acid or methanol was formed. Thus, the pathway consisting of CO oxidation to from CO2 and then reforming CH4 with CO2 to form acetic acid was excluded.

Preservation of Rh cations in micropores after catalysis.
One concern is whether Rh cations were still in the micropores after catalysis. Solution after catalysis consisted of solvent and products (in liquid) and solid catalyst. As most zeolite particles deposited to the bottom, they were readily separated after centrifugation. Notably, small particles couldn't be precipitated; thus, we used filter paper to filter these small catalyst particles from the solution after majority catalyst particles were deposited through centrifugation. In this way, the most solid catalyst particles were collected for ICP analysis.
The collected catalyst (after catalysis) was dissolved in solution for ICP test. The details of preparation solution were described in the section entitled "ICP-AES measurements of concentrations of Rh in catalysts" of Supplementary Methods. ICP-AES tests showed that the Rh atoms in the collected catalyst was 0.098wt%Rh, very close to the original weight ratio of Rh, 0.10wt%Rh. It suggested that there was little leaching of Rh from ZSM-5. From this point of view, Rh cations remained in the micropores during catalysis. Why selectivity for producing formic acid is higher at a shorter reaction time? The catalytic performances in Figure 2 obtained at 10 bar CH4, 10 CO and 8 bar O2 for 2 hrs and 50 bar CH4, 10 CO and 8 bar O2 for 2 hrs in Figure 2 and data in Figure 3 were catalysis data collected after 1.5 or 2 hrs. The selectivity for formation of formic acid is higher than that for acetic acid. However, a longer reaction time such as the data under the catalytic conditions (in the mixture of 10 bar CH4 with 10 bra CO and 8 bar O2 for 12 hrs or the mixture of in 50 bar CH4 with 10 bar CO and 8 bar O2 for 12 hrs in Figure 2 gave selectivity for formation of acetic acid higher than formic acid.

Does Rh cations chemically bond to O atoms in micropores?
The high selectivity for formic acid (the low selectivity for acetic acid) is relevant to the large portion of incubation heating of catalyst from 25 o C to 150 o C among a whole heating when the formal heating time at 150 o C is short. Here the whole heating of catalyst includes the incubation heating from 25 o C to ideal temperature (typically 150 o C) and formal heating at the ideal temperature (typically 150 o C); the time reported for heating is only the time of reactor remaining at ideal temperature (typically 150 o C); the time used for heating the catalyst from 25 o Cto 150 o C (called incubation heating) is about 1 hr. If the formal heating time at ideal temperature is only 2 hrs or even 1 hr, the incubation heating is be an important portion of the whole heating. If the formal heating time at ideal temperature is 12 hrs, the incubation heating is a minor portion of the overall heating.
Since the selectivity for formation of formic acid in heating of short time is higher than that of heating of long time, it suggests that a relatively low temperature of incubation heating from 25 o C-150 o C favors the formation of formic acid. When the heating time is only 1 or 2 hrs, the incubation heating from 25 o C to 150 o C probably mainly forms formic acid and thus results in a relatively high selectivity for the formation of formic acid. This interpretation is consistent with the proposed reaction pathway by DFT calculation. As shown in the energy profile Figure 6b, to form acetic acid, the barrier to across the transition state (c7 in Figure 6c) from c6 to c8 in Figure  6c to form the first acetic acid is quite high. This high barrier makes the formation of the first acetic acid at low temperature not kinetically favorable. Alternatively, the intermediate (c6 in Figure 6c), a formate (HCOO) adsorbed in Rh could readily couple with one H to form formic acid at low temperature to desorb from the site, instead of crossing the high barrier of the transition state (c7 in Figure 6) to form acetic acid.
To further check whether this interpretation is correct or not, we performed time-dependent study of the yields of formic acid and acetic acid, respectively. The parallel studies were done for formal heating at 150 o C for 0.5 hrs, 2 hrs, 3hrs, 5, hrs and 12 hrs under the same condition (50 bar CH4 10 bar CO and 8 bar O2); as plotted in Supplementary Figure 9, the selectivity for formation of acetic acid increases as a function of time. This is consistent with the kinetically favorable formation of formic acid at low temperature since an experiment with short formal heating time at 150 o C has a large portion of heating at low temperature (25 o C-150 o C). Thus, our interpretation of the high selectivity for producing formic acid is supported by the experiments in Supplementary  Figure 9.
Understand the CO pressure-dependent catalytic activity through computation. Our experimental studies found that CO at high pressure (Figure 3b) in fact decreased the selectivity for producing acetic acid and finally poisoned the active sites. To understand this observation, we evaluated the CO adsorption on the Rh1 atom in DFT calculations. We found that the first adsorbed CO molecule binds strongly to the Rh1 site, with an adsorption energy of -1.92 eV ( Supplementary  Figure 10a), and -0.69 eV for the second CO (Supplementary Figure 10b). It suggests that the Rh1 site could adsorb two CO molecules.
We also explored the C-H activation of methane by Rh1 atom when the Rh1 has already adsorbed a CO molecule (Supplementary Figure 10c). Supplementary Figure 10c is the transition state in activation of the first C-H of CH4 on Rh1 with one pre-adsorbed CO molecule. With a preadsorbed CO molecule on Rh1O5, the barrier for activating the first C-H of CH4 is only 0.34 eV.
Unfortunately, the activation barrier for activating CH4 on a Rh1 atom with two pre-adsorbed CO molecules is increased to 1.36 eV. The large increase of barrier for activating CH4 suggested by DFT calculation rationalized the poison of CO to Rh1O5 sites in the formation of acetic acid when CO pressure is higher than 10 bar, observed in Figure 11 This difference shows that some Brønsted sites of H-ZSM-5 lost due to the replacement by Rh cations in the ion exchange (IWI process). Peak at 1.7 ppm was assigned to the H-O-Si-(on external frame of ZSM-5) based on the literature. 11 . The peak at 6.1 ppm is contributed from adsorbed H2O molecules based on the literature 11 and from the second type of Brønsted acid sites (BAS) based on literature 16 . The references of these assignments were presented in Table S1. To remove the contribution of H2O to 1 H spectra, catalyst was loaded to a stainless tube reactor and then were annealed to 400 o C and remained for 5 hrs while they were being pumped by a vacuum pump. After 5 hrs annealing at 400 o C in vacuum to remove H2O and other impurity, the valves at the two sides of the stainless tube were closed. The stainless steel tube containing the dry catalyst was transferred to a glove box. A similar treatment for removal of water molecules adsorbed in ZSM-5 was used in literature. 15 Then, the valve of the stainless steel tube was open to transfer the sample to a NMR test tube. The NMR test tube was then sealed. The sealed NMR test tube was immediately used for NMR studies.  Table 3. Energies and activation barriers for the simulated pathway leading to the formation of acetic acid as shown in Figure 6 of the main text.

Reaction
Step