Machine learning-assisted screening of SA-FLP dual-active-site catalysts for the production of methanol from methane and water

doi:10.1016/S1872-2067(24)60225-1

Abstract

Abstract:

One-step direct production of methanol from methane and water (PMMW) under mild conditions is challenging in heterogeneous catalysis owing to the absence of highly effective catalysts. Herein, we designed a series of “Single-Atom” - “Frustrated Lewis Pair” (SA-FLP) dual active sites for the direct PMMW via density functional theory (DFT) calculations combined with a machine learning (ML) approach. The results indicate that the nine designed SA-FLP catalysts are capable of efficiently activate CH₄ and H₂O and facilitate the coupling of OH* and CH₃* into methanol. The DFT-based microkinetic simulation (MKM) results indicate that CH₃OH production on Co₁-FLP and Pt₁-FLP catalysts can reach the turnover frequencies (TOFs) of 1.01 × 10⁻³ s^-1 and 8.80 × 10⁻⁴ s^-1, respectively, which exceed the experimentally reported values by three orders of magnitude. ML results unveil that the gradient boosted regression model with 13 simple features could give satisfactory predictions for the TOFs of CH₃OH production with RMSE and R² of 0.009 s^-1 and 1.00, respectively. The ML-predicted MKM results indicate that four catalysts including V₁-, Fe₁-, Ti₁-, and Mn₁-FLP exhibit higher TOFs of CH₃OH production than the value that the most relevant experiments reported, indicating that the four catalysts are also promising catalysts for the PMMW. This study not only develops a simple and efficient approach for design and screening SA-FLP catalysts but also provides mechanistic insights into the direct PMMW.

Key words: Single-atom catalyst, Frustrated Lewis pair, Machine learning, Dual active sites, Methanol synthesis

Tao Ban, Jian-Wei Wang, Xi-Yang Yu, Hai-Kuo Tian, Xin Gao, Zheng-Qing Huang, Chun-Ran Chang. Machine learning-assisted screening of SA-FLP dual-active-site catalysts for the production of methanol from methane and water[J]. Chinese Journal of Catalysis, 2025, 70: 311-321.

×
share this article

Add to citation manager EndNote|Ris|BibTeX

URL: https://www.cjcatal.com/EN/10.1016/S1872-2067(24)60225-1

https://www.cjcatal.com/EN/Y2025/V70/I3/311

Figures/Tables 10

Fig. 1. (a) The optimized CeO2(110). (b) The optimized M1@CeO2. (c) The optimized M1@CeO2-Vo. In (a), the black circle indicates the position where a Ce atom is substituted by a transition metal atom, while in (b), the blue circle marks the site where an O atom has been removed.

Fig. 2. (a) Eads of H2O at SA and FLP sites. (b) Typical adsorption configurations of H2O at SA and FLP sites.

Fig. 3. (a) Eads of CH4 on SA site. (b) Typical adsorption configurations of CH4 adsorption and dissociation with pre-adsorbed H2O. (c) Reaction and activation energy profiles for the adsorption and activation of CH4.

Fig. 4. (a) The potential energy profile for the PMMW on the nine SA-FLP catalysts. The zero-energy reference is defined as the sum of the energies of SA-FLP surface, H2O(g), and CH4(g). (b) Representative images of the PMMW.

Fig. 5. TOFs for CH3OH production varying with pressure and temperature for Co1-FLP (a), Ni1-FLP (b), Cu1-FLP (c), Rh1-FLP (d), Pd1-FLP (e), Ag1-FLP (f), Ir1-FLP (g), Pt1-FLP (h), and Au1-FLP (i), with the initial CH4/H2O ratio set to 1.

Fig. 6. Pearson correlation analysis corresponds to the 17 descriptors used in this study. Values below 0.01 are displayed as zero.

Table 1 The RMSE and R2 for all the evaluated ML models for both the test set and cross-validation, respectively.

Model	Test set		Cross-validation
Model	R²	RMSE (s^-1)	R²	RMSE (s^-1)
SVM	0.999	0.072	0.999	0.072
RFR	1.00	0.030	1.00	0.028
GBR	1.00	0.010	1.00	0.011
DTR	1.00	0.036	1.00	0.027
ETR	1.00	0.037	1.00	0.033
KNN	1.00	0.037	1.00	0.028
LASSO	0.877	1.051	0.875	1.052
OLR	0.988	0.329	0.988	0.326
PLS	0.988	0.329	0.988	0.326

Fig. 7. MKM-simulated CH3OH TOF vs ML-predicted CH3OH TOF on the LASSO model (a), PLS model (b), OLR model (c), SVM model (d), ETR model (e), RFR model (f), KNN model (g), DTR model (h), and GBR model (i).

Fig. 8. (a) Permutation feature importance score in the GBR model. (b) MKM-simulated CH3OH TOF vs. ML-predicted CH3OH TOF on the GBR model with 10-fold cross-validation.

Fig. 9. The ML-predicted CH3OH TOFs varying with pressure and temperature for Ti1-FLP (a), V1-FLP (b), Mn1-FLP (c), Fe1-FLP (d), Zn1-FLP (e), Zr1-FLP (f), Nb1-FLP (g), Mo1-FLP (h), Tc1-FLP (i), Ru1-FLP (j), Cd1-FLP (k), Hf1-FLP (l), Ta1-FLP (m), W1-FLP (n), Re1-FLP (o), Os1-FLP (p), with the initial CH4/H2O ratio set to 1.

References

[1]	Liu. Y, Deng. D, Bao. X, Chem, 2020, 6, 2497-2514.
[2]	Y. Tsuji, K. Yoshizawa, J. Phys. Chem. C, 2018, 122, 15359-15381.
[3]	H. Cruchade, I. C. Medeiros-Costa, N. Nesterenko, J.-P. Gilson, L. Pinard, A. Beuque, S. Mintova, ACS Catal., 2022, 12, 14533-14558.
[4]	P. Zhai, Y. Li, M. Wang, J. Liu, Z. Cao, J. Zhang, Y. Xu, X. Liu, Y.-W. Li, Q. Zhu, D. Xiao, X.-D. Wen, D. Ma, Chem, 2021, 7, 3027-3051.
[5]	R. Vakili, R. Gholami, C. E. Stere, S. Chansai, H. Chen, S. M. Holmes, Y. Jiao, C. Hardacre, X. Fan, Appl. Catal. B, 2020, 260, 118195.
[6]	R. Palkovits, M. Antonietti, P. Kuhn, A. Thomas, F. Schüth, Angew. Chem. Int. Ed., 2009, 48, 6909-6912.
[7]	Z. Jin, L. Wang, E. Zuidema, K. Mondal, M. Zhang, J. Zhang, C. Wang, X. Meng, H. Yang, C. Mesters, F.-S. Xiao, Science, 2020, 367, 193-197.
[8]	J. Xie, R. Jin, A. Li, Y. Bi, Q. Ruan, Y. Deng, Y. Zhang, S. Yao, G. Sankar, D. Ma, J. Tang, Nat. Catal., 2018, 1, 889-896.
[9]	Y. Kim, T. Y. Kim, H. Lee, J. Yi, Chem. Commun., 2017, 53, 4116-4119.
[10]	K. T. Dinh, M. M. Sullivan, K. Narsimhan, P. Serna, R. J. Meyer, M. Dinca, Y. Roman-Leshkov, J. Am. Chem. Soc., 2019, 141, 11641-11650.
[11]	K. Narsimhan, K. Iyoki, K. Dinh, Y. Roman-Leshkov, ACS Cent. Sci., 2016, 2, 424-429.
[12]	V. L. Sushkevich, D. Palagin, M. Ranocchiari, J. A. van Bokhoven, Science, 2017, 356, 523-527.
[13]	Y. R. Jeong, H. Jung, J. Kang, J. W. Han, E. D. Park, ACS Catal., 2021, 11, 1065-1070.
[14]	A. Koishybay, D. F. Shantz, J. Am. Chem. Soc., 2020, 142, 11962-11966.
[15]	W. Bi, Y. Tang, X. Li, C. Dai, C. Song, X. Guo, X. Ma, Commun. Chem., 2022, 5, 124.
[16]	H. Zhang, X. Zhang, X. Shi, X. Liu, J. Liang, S. Li, S. Ma, X. Xiao, J. Guo, L. Zhang, M. Tan, Appl. Catal., B, 2025, 361, 124648.
[17]	Z. Liu, E. Huang, I. Orozco, W. Liao, R. M. Palomino, N. Rui, T. Duchon, S. Nemsak, D. C. Grinter, M. Mahapatra, P. Liu, J. A. Rodriguez, S. D. Senanayake, Science, 2020, 368, 513-517.
[18]	Y.-Q. Su, J.-X. Liu, I. A. Filot, L. Zhang, E. J. M. Hensen, ACS Catal., 2018, 8, 6552-6559.
[19]	S. Zhao, D. Kang, Y. Liu, Y. Wen, X. Xie, H. Yi, X. Tang, ACS Catal., 2020, 10, 11739-11750.
[20]	T. Ban, X.-Y. Yu, H.-Z. Kang, H.-X. Zhang, X. Gao, Z.-Q. Huang, C.-R. Chang, J. Catal., 2022, 408, 206-215.
[21]	Z.-Q. Huang, T.-Y. Zhang, C.-R. Chang, J. Li, ACS Catal., 2019, 9, 5523-5536.
[22]	Y. Dong, K. K. Ghuman, R. Popescu, P. N. Duchesne, W. Zhou, J. Y. Y. Loh, A. A. Jelle, J. Jia, D. Wang, X. Mu, C. Kubel, L. Wang, L. He, M. Ghoussoub, Q. Wang, T. E. Wood, L. M. Reyes, P. Zhang, N. P. Kherani, C. V. Singh, G. A. Ozin, Adv. Sci., 2018, 5, 1700732.
[23]	S. Zhang, Z. Xia, Y. Zou, F. Cao, Y. Liu, Y. Ma, Y. Qu, J. Am. Chem. Soc., 2019, 141, 11353-11357.
[24]	S. Zhang, Y. Liu, M. Zhang, Y. Ma, J. Hu, Y. Qu, Nat. Commun., 2022, 13, 5527.
[25]	S. K. Kim, H. W. Kim, S. J. Han, S. W. Lee, J. Shin, Y. T. Kim, Commun. Chem., 2020, 3, 58.
[26]	M. Deimel, K. Reuter, M. Andersen, ACS Catal., 2020, 10, 13729-13736.
[27]	M. Sun, A. W. Dougherty, B. Huang, Y. Li, C. H. Yan, Adv. Energy Mater., 2020, 10, 1903949.
[28]	G. H. Gu, J. Noh, S. Kim, S. Back, Z. Ulissi, Y. Jung, J. Phys. Chem. Lett., 2020, 11, 3185-3191.
[29]	J. M. Rickman, G. Balasubramanian, C. J. Marvel, H. M. Chan, M. T. Burton, J. Appl. Phys., 2020, 128, 221101.
[30]	T. Toyao, K. Suzuki, S. Kikuchi, S. Takakusagi, K.-I. Shimizu, I. Takigawa, J. Phys. Chem. C, 2018, 122, 8315-8326.
[31]	T. S. Choksi, L. T. Roling, V. Streibel, F. Abild-Pedersen, J. Phys. Chem. Lett., 2019, 10, 1852-1859.
[32]	Q. Zhu, Y. Gu, X. Liang, X. Wang, J. Ma, ACS Catal., 2022, 12, 12336-12348.
[33]	M. Xing, Y. Zhang, S. Li, H. He, S. Sun, J. Phys. Chem. C, 2022, 126, 17025-17035.
[34]	A. Kumar, J. Iyer, F. Jalid, M. Ramteke, T. S. Khan, M. A. Haider, ChemCatChem, 2021, 14, e202101481
[35]	W. Liao, P. Liu, Catal. Sci. Technol., 2022, 12, 3836-3845.
[36]	X. Liu, P. Xu, J. Zhao, W. Lu, M. Li, G. Wang, J. Alloys Compd., 2022, 921, 165984.
[37]	G. Kresse, J. Furthmüller, Comput. Mater. Sci., 1996, 6, 15-50.
[38]	G. Kresse, J. Hafner, Phys. Rev. B, 1994, 49, 14251-14269.
[39]	G. Kresse, D. Joubert, Phys. Rev. B, 1999, 59, 1758-1775.
[40]	J. P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett., 1996, 77, 3865-3868.
[41]	Z.-Q. Huang, L.-P. Liu, S. Qi, S. Zhang, Y. Qu, C.-R. Chang, ACS Catal., 2018, 8, 546-554.
[42]	S. Grimme, J. Antony, S. Ehrlich, H. Krieg, J. Chem. Phys., 2010, 132, 154104.
[43]	S. Smidstrup, A. Pedersen, K. Stokbro, H. Jonsson, J. Chem. Phys., 2014, 140, 214106.
[44]	A. Heyden, A. T. Bell, F. J. Keil, J. Chem. Phys., 2005, 123, 224101.
[45]	A. J. Medford, C. Shi, M. J. Hoffmann, A. C. Lausche, S. R. Fitzgibbon, T. Bligaard, J. K. Norskov, Catal. Lett., 2015, 145, 794-807.
[46]	F. Pedregosa, G. Varoquaux, A. Gramfort, V. Michel, B. Thirion, O. Grisel, M. Bloder, P. Prettenhofer, R. Weiss, V. Dubourg, J. Vanderplas, A. Passos, D. Cournapeau, M. Brucher, M. Perrot, E. Duchesnay, J. Mach. Learn. Res., 2011, 12, 2825-2830.
[47]	H.-J. Lu, N. Zou, R. Jacobs, B. Afflerbach, X.-G. Lu, D. Morgan, Comput. Mater. Sci., 2019, 169, 109075.
[48]	L. Breiman, Machine Learning, 2001, 45, 5-32.
[49]	J. F. Magee, Harv. Bus. Rev., 1964, 42, 126-138.
[50]	J. H. Friedman, Comput. Stat. Data Anal., 2002, 38, 367-378.
[51]	J. M. Pisane, L. Wehenkel, J. Verly, IEEE Radar Conf., 2010, 1454-1458.
[52]	A. Seko, T. Maekawa, K. Tsuda, I. Tanaka, Phys. Rev. B, 2014, 89, 054303.
[53]	E. A. Patrick, F. P. Patrick, Inf. Control., 1970, 16, 128-152.
[54]	G. M. D'Angelo, D. Rao, C. C. Gu, BMC Proc., 2009, 3, S62.
[55]	T. Ban, X.-Y. Yu, H.-Z. Kang, Z.-Q. Huang, J. Li, C.-R. Chang, ACS Catal., 2023, 13, 1299-1309.