Direct immobilization of an atomically dispersed Pt catalyst by suppressing heterogeneous nucleation at −40 °C

Kai Huang ab, Ruyue Wang ab, Hongbo Wu c, Hao Wang ad, Xian He a, Hehe Wei b, Shanpeng Wang b, Ru Zhang a, Ming Lei *a, Wei Guo *c, Binghui Ge *ef and Hui Wu *b
aState Key Laboratory of Information Photonics and Optical Communications, School of Science, Beijing University of Posts and Telecommunications, Beijing 100876, China. E-mail: mlei@bupt.edu.cn
bState Key Laboratory of New Ceramics and Fine Processing, School of Materials Science and Engineering, Tsinghua University, Beijing 100084, China. E-mail: huiwu@tsinghua.edu.cn
cKey Lab of Advanced Optoelectronic Quantum Architecture and Measurement (MOE), Beijing Key Lab of Nanophotonics & Ultrafine Optoelectronic Systems, School of Physics, Beijing Institute of Technology, Beijing 100081, China. E-mail: weiguo7@bit.edu.cn
dMaterials Science and Engineering Program, Texas Materials Institute, The University of Texas at Austin, Austin, TX 78712, USA
eInstitutes of Physical Science and Information Technology, Anhui University, Hefei 230601, China. E-mail: bhge@ahu.edu.cn
fKey Laboratory of Structure and Functional Regulation of Hybrid Materials, Anhui University, Ministry of Education, Hefei, 230601, P. R. China

Received 11th July 2019 , Accepted 8th September 2019

First published on 9th September 2019


Direct deposition of isolated metal atoms onto substrates has been recognized as a simple route to obtain high performance supported atomically dispersed metals (SACs), however, the agglomeration driven by high surface energy is difficult to avoid. Herein, we demonstrate a one-pot solution synthesis to obtain atomically dispersed platinum (Pt) supported on nitrogen (N)-doped mesoporous carbon (NMC) substrates (Pt/NMC-LT) by conducting the whole synthesis at −40 °C, owing to the sluggish nucleation kinetics. We obtained the Pt/NMC-LT catalyst with superior electrochemical hydrogen evolution reaction (HER) activity and stability, in comparison with the NMC supported dominant Pt sub-nanometer cluster catalyst from solution synthesis at RT ∼ 25 °C (Pt/NMC-RT) and commercial carbon supported Pt nanoparticle catalysts (Pt/C). Lower over-potential values (only 17.0 and 49.8 mV) are needed for high HER current densities (10 and 100 mA cm−2, respectively), and no obvious degradation is observed after an accelerated durability test (ADT) for 5000 CV cycles.


Supported atomically dispersed metal catalysts (SACs), which are also referred to as single-atom catalysts, have been actively studied in the past several years and regarded as a new frontier in heterogeneous catalysis.1–3 By virtue of size reduction, we can realize the ultimate goal of manipulating individual atoms for function (i.e. 100% atom utilization efficiency), thus bridging the gap between homogeneous and heterogeneous catalysis.4–7 Considering the feasibility of the fabrication process and potential industrial applications, wet-chemistry approaches relative to other methods (such as atomic layer deposition, combustion synthesis, high-temperature vapor transport, metal leaching and mass-selected soft-landing) can be employed to synthesize supported atomically dispersed metal catalysts, since metal precursors already contain atomic metal species in solution.7,8 However, the high surface energy of isolated metal atoms will spontaneously force them to form clusters and nanoparticles by nucleation and growth, especially under harsh preparation or catalytic conditions.9,10 Therefore, we need to anchor atomically dispersed metal atoms onto suitable supports and avoid their aggregation during the syntheses and post-treatment processes.

Notably, a typical one-pot solution synthesis can be directly adopted to prepare SACs, only if the precursor monomers undergo a direct transformation on the surface of substrates without any further nucleation. Two possible reaction pathways will happen to facilitate the synthesis of SACs, including the reduction of atomic species in solution followed by anchoring to the supports, and/or the adsorption of the metal precursor followed by the surface reduction. However, undesired reaction pathways still exist due to the competition between homogeneous and heterogeneous nucleation in one-pot solution syntheses, yielding metal clusters or nanoparticles instead of atomically selective deposition on the supports, for example, the formation of small clusters in solution through homogeneous nucleation, followed by their attachment and deposition onto the supports and a combination of the above-mentioned pathways.11,12 Therefore, it is of great significance to understand and regulate the deposition of isolated metal atoms on the surface of substrates in solution synthesis. Although the mechanism is still unclear and controversial,13,14 previous theoretical and experimental results have shown that the necessary free energy for heterogeneous nucleation cannot physically surpass the energy barrier in the homogeneous nucleation process due to the presence of active centers.11–15 Consequently, the reaction driving force must be maintained at a lower level to program one-pot solution synthesis of SACs and avoid homogeneous nucleation in solution, which requires a lower concentration of monomers or reaction temperature.

Herein, we develop an ultralow-temperature (−40 °C) one-pot solution-phase synthesis process to obtain nitrogen (N)-doped mesoporous carbon (NMC) supported atomically dispersed platinum (Pt) catalysts (Pt/NMC-LT) for efficient electro-catalytic hydrogen evolution reaction (HER). As a contrast, a one-pot solution synthesis process at room temperature (RT) yields dominant Pt clusters supported on NMC substrates (Pt/NMC-RT) because of the lack of thermodynamics and kinetics regulations for nuclei formation by low reaction temperature. Similarly, we also demonstrate the successful preparation of supported atomically dispersed iridium (Ir) and rhodium (Rh). Moreover, the electrochemical experimental results demonstrated a superior HER activity and long-term operational stability of the Pt/NMC-LT catalyst in acid electrolyte, originating from the maximum noble metal atom efficiency and most exposed active sites, with respect to Pt/NMC-RT and commercial Pt/C catalysts.

As illustrated in Fig. 1, we proposed a typical temperature-controlled one-pot solution synthesis method using a circulating cooling system. Taking the chemical reduction of the H2PtCl6 precursor as an example, we employed ethanol as a relatively weak reductant, which aims at performing a slow reduction kinetics to exclude the possible involvement of homogeneous nucleation (see ESI for Experimental details). The presence of the mesoporous structure and nitrogen dopants in NMC substrates favors the trapping and anchoring of atomic Pt species in solution.16 Nevertheless, the formation of Pt clusters on the substrates still exists in the solution-phase reduction at RT, resulting from the incomplete control of homogeneous nucleation in solution. In contrast, we can successfully obtain supported atomically dispersed Pt species, just by decreasing the reaction temperature down to −40 °C. According to previous results,11,12 the nucleation barriers for both homogeneous and heterogeneous nuclei formation increase while the reduction and nucleation rates decrease, when running a reaction at lower temperatures. Therefore, the reduction pathway will switch from solution to the surface of supports due to the low-temperature regulations in both thermodynamics and kinetics.


image file: c9ta07469d-f1.tif
Fig. 1 Schematic illustration of the one-pot solution synthesis of atomically dispersed Pt on NMC substrates. Running at low-temperature (−40 °C, top) and room temperature (RT ∼ 25 °C, bottom).

The aberration-corrected high-angle annular dark field-scanning transmission electron microscopy (HAADF-STEM) images at different magnifications are shown in Fig. 2; only isolated Pt atoms are anchored on the surface of NMC substrates in Pt/NMC-LT samples, compared with the appearance of Pt sub-nanometer clusters prepared at RT. We also observed similar comparisons between clusters and atomically dispersed moieties for the solution synthesis of NMC supported Ir- and Rh-based catalysts and mesoporous carbon (MC) or commercial XC-72R supported Pt based catalysts (Fig. S1–S3). Notably, this result matches well with our schematic illustration and design principles. Moreover, we further performed ab initio molecular dynamics (AIMD) simulations at −40 °C and RT (see ESI for Computational details), to understand the mechanism of low-temperature induced sluggish heterogeneous nucleation and the formation of atomically dispersed Pt atoms rather than Pt clusters. Depending on local reaction surroundings, the simulations indicated that Pt atoms might react with ethanol molecules or deposit onto NMC substrates before nucleation. For Pt atoms in direct interaction with ethanol molecules, we observed that one nearby ethanol molecule decomposed with a hydrogen atom transferred from hydroxymethyl or methyl to the Pt atom, thus forming the three-coordinate (Pt3c) or four-coordinate (Pt4c) atomic configurations (Fig. S4). The Pt–ethanol interaction proceeded with C–H bond breaking and Pt–O and Pt–C bond formation, due to the orbital hybridization between the d orbitals of the Pt atom and the p orbitals of C and O atoms. In this case, the Pt4c structure has been identified with better thermodynamic stability and larger charge transfer according to the calculated binding energy, Bader charge and bond-length (Table S1). Moreover, Pt atoms would deposit at the defective sites with durable stability for the direct diffusion of Pt atoms onto NMC substrates, due to their large binding strength on the defective NMC substrates. Obviously, as shown in Fig. 2c, there may exist two possible Pt–Pt nucleation pathways in the reaction mixture: Path-1, two isolated Pt atoms are both in ethanol; Path-2, one Pt atom is in ethanol and the other Pt atom is located on the NMC substrate.


image file: c9ta07469d-f2.tif
Fig. 2 HAADF-STEM images of (a) Pt/NMC-LT and (b) Pt/NMC-RT. (c) Schematic illustration of the two types of nucleation pathways of Pt atoms in the ethanol and NMC substrate reaction mixture. (d) Energy diagram of Path-2 for Pt–Pt dimer formation at −40 °C (up panel) and RT (bottom panel). The initial state (IS), transition state (TS) and final state (FS) are shown from left to right. The ethanol molecules bound with Pt atoms and the NMC substrate are highlighted by spheres. C, N, Pt, O and H atoms are depicted by gray, blue, light blue, red and white spheres, respectively.

Due to the structural effect and the contribution from the ethanol molecular shell effect, the nucleation barriers for Pt–Pt in the Path-1 mode at an ultralow temperature are larger than that at RT as we had confirmed previously,17 leading to an extremely lower nucleation rate at low temperature. Furthermore, in this work, we explored the atomic structure of atomically dispersed Pt atoms in ethanol solvent by performing long time AIMD simulations. As shown in Fig. S5, the two Pt atoms remain in the non-bonding state for at least 30 ps with distances ranging from 3.8 Å to 5.4 Å at −40 °C, indicating that the atomically dispersed Pt atoms can be successfully synthesized under ultralow temperature conditions, whereas for simulations at RT, the two separated Pt atoms quickly get close within 3 ps and form a dimer configuration (Pt–Pt ∼2.6 Å), owing to the rather faster nucleation kinetics. The final dimer state is stable and can be transferred onto the NMC substrate easily due to the thermodynamic preference.

Moreover, we adopted the constrained minimization technique for the calculation of the Path-2 mode,18,19 demonstrating the nucleation barrier for the initial state of one Pt atom in the typical Pt3c configuration in ethanol and the other Pt atom at the defective site of NMC. As shown in Fig. 2d, our calculations indicated that the nucleation barrier at −40 °C (∼0.425 eV) was slightly larger than that at RT (∼0.288 eV). Notably, only this small barrier difference (∼0.137 eV) will result in a four orders of magnitude faster nucleation rate constant at RT than that at −40 °C (Fig. S6). Thus, the low reaction temperature can effectively suppress the Path-2 nucleation process of Pt atoms in the reaction mixture. Similarly, the two Pt atoms also remain in the non-bonding state for at least 30 ps at −40 °C as shown in Fig. S7. However, the distance between the two separated Pt atoms quickly decreases forming a dimer configuration (Pt–Pt ∼2.5 Å) within 3 ps for simulations at RT owing to the higher nucleation rate constant compared to that at −40 °C. Again, the two Pt atoms would stay in the dimer state on the NMC substrate due to the energetical preference. Although the dispersed Pt atoms in the reaction mixture may have various complex configurations, the underlying mechanism is similar and well revealed by our calculations. Induced by the ultralow temperature, a larger nucleation barrier and significantly decreased nucleation kinetics happened at −40 °C. In good agreement with the experimental results, the products should be atomically dispersed Pt atoms at the low temperature rather than dimers or clusters at RT. Thus, the formation of atomically dispersed Pt atoms in the low-temperature one-pot solution synthesis can be attributed to the ensemble of suppressed heterogeneous nucleation on the surface of NMC substrates when homogeneous nucleation in solution is excluded. Moreover, our calculated results confirm that the temperature-control strategy is an effective solution for synthesizing an atomically dispersed Pt single-atom catalyst, and can be widely generalized to other single-atom syntheses.

Further characterization was performed to determine the chemical states and coordination environment of the as-prepared samples in Fig. 3, by X-ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS) and X-ray absorption fine structure (XAFS). Identical to the pure NMC substrate, no obvious Pt phase can be indexed in the patterns of Pt/NMC-LT and Pt/NMC-RT, which indicates the absence of grown Pt nanocrystals as shown in Fig. 2 and the selected area electron diffraction (SEAD) patterns in Fig. S8. Due to the strong metal–support interaction between atomically dispersed Pt atoms and the NMC substrate as reported,20–22 the XPS binding energy of Pt/NMC-LT and Pt/NMC-RT in Pt 4f7/2 spectra shifted to a higher energy of 72.7 eV with respect to that of metallic Pt (71.2 eV), which corresponds to a Ptδ+ (between +2 and +4) state. In addition, a majority of pyridinic-N and graphitic-N (centered at the binding energies of 400.8 and 398.5 eV, respectively) can serve as effective binding sites to enhance interactions between supports and metal atoms (Fig. S9).23 To reflect the local atomic and electronic structure of Pt elements, the normalized X-ray absorption near edge structure (XANES) spectra are shown in Fig. 3c, where the increased white-line intensity of isolated Pt atoms relative to Pt clusters and foil further confirms the strong interactions in the Pt/NMC-LT sample.24 Furthermore, EXAFS fitting results of Pt/NMC-LT showed no appreciable Pt–Pt coordination and a dominant peak at 2.13 Å corresponding to the possible Pt–C/N/O coordination (Table S2). In clear contrast, Pt foil exhibited an obvious peak at 2.76 Å corresponding to Pt–Pt coordination, while Pt/NMC-RT showed an intermediate performance indicating that both Pt–Pt coordination and Pt–C/N/O coordination exist. Thus, we can conclude that the atomically dispersed Pt species in Pt/NMC-LT have an oxidation state of Ptδ+ depending on their interaction with the support as reported in earlier literature.20–22


image file: c9ta07469d-f3.tif
Fig. 3 (a) XRD results of Pt/NMC-LT, Pt/NMC-RT and NMC. (b) Pt 4f XPS spectra of Pt/NMC-LT and Pt/NMC-RT. (c) Normalized X-ray absorption near edge structure (XANES) spectra at the Pt L3-edge and (d) extended X-ray absorption fine structure (EXAFS) spectra of the Pt foil, Pt/NMC-LT, and Pt/NMC-RT, respectively. (e) Schematic diagrams of structures and distributions for atomically dispersed Pt atoms and Pt–Pt dimers on the sites of NMC without vacancy defects (WVD), single point defects (SPD), double point defects (DPD) and hole vacancy defects (HVD). C, N and Pt atoms are represented as brown, blue, and light blue spheres, respectively.

By using DFT calculations (see ESI for details), the stabilities of isolated Pt atoms located at various sites of the NMC substrate are carefully examined. Under low temperature conditions, most of the product Pt atoms in the reaction mixture directly bind to the NMC defective sites due to the large binding strength, and these Pt atoms can keep the atomically dispersed configurations on the NMC substrate. Meanwhile, depending on the Pt–ethanol interaction environment, some of the Pt atoms may exist in the Pt3c or Pt4c configurations in the reaction mixture. Our calculations reveal that, for various NMC structures, the Pt atoms in both Pt3c and Pt4c configurations would finally break away from the specific configurations and bind to nearby defective sites of the NMC substrate, and then the dissociated ethanol molecules prefer to recover upon Pt atoms adsorbed on the NMC substrate (Fig. S10 and S11). The NMC substrates can enhance this Pt transfer process, and are also beneficial to inhibiting the Pt nucleation process in the reaction mixture. Specifically, the Pt@NMC with molecular ethanol structures is about 2.731/5.260 eV and 1.809/5.268 eV energetically more favorable than the Pt3c/Pt4c with NMC substrates (Fig. S10 and S11). Based on our DFT results, all isolated Pt atoms finally tend to deposit on the NMC substrates due to the thermodynamic driving force, and this is in good agreement with the experimental observations.

To strengthen our understanding of the STEM observation that a large number of atomically dispersed Pt atoms are anchored on the NMC substrate, furthermore, we have checked whether the Pt atoms on the NMC substrate tend to form dimers or clusters. The defective sites on the NMC structure with a low coordination number bind Pt atoms more strongly and charge is transferred from Pt atoms to those target sites for enhanced stability. Notably, the larger the binding strength of Pt adsorption on the substrate, the greater the charge transfer from Pt to C and N atoms of the NMC substrate. As shown in Fig. 3e, our model calculations reveal that the atomically dispersed Pt configurations are energetically more stable (SPD, DPD and HVD sites, 0.809, 4.038 and 1.558 eV) than Pt atoms in dimer configurations at the corresponding defective points or edge sites. Such results clearly support the experimental observation that the produced Pt atoms on the NMC substrate preferentially remain in the isolated state rather than in dimer or cluster configurations.

As shown in Fig. 4, we explored the hydrogen evolution reaction (HER) electro-catalysis as a fundamentally important model reaction to evaluate size effects in catalysts less than one nanometer. Significantly, the Pt/NMC-LT catalyst with atomically dispersed Pt active sites exhibited a superior HER activity compared with Pt/NMC-RT and commercial Pt/C catalysts, with the same loading of 10 μg cmPt−2. The onset potential of the Pt/NMC-LT catalyst was much close to the standard hydrogen electrode potential, while Pt/NMC-RT and Pt/C catalysts exhibited quite larger overpotentials (22.0 and 41.3 mV, respectively) to drive hydrogen production. What's more, the overpotentials of the Pt/NMC-LT catalyst to achieve higher current densities (10 and 100 mA cm−2) were 17.0 and 49.8 mV, respectively. In contrast, higher values for Pt/NMC-RT (49.1 and 83.8 mV) and commercial Pt/C (75.3 and 104.2 mV) catalysts indicated relatively sluggish kinetics for potential applications due to a larger charge transfer resistance confirmed by the electrochemical impedance spectroscopy (EIS) analysis (Fig. S12 and Table S3).This result can be further demonstrated by Tafel plots derived from HER polarization curves in Fig. 3c. A smaller Tafel slope of 26.2 mV dec−1 for Pt/NMC-LT than values of 30.3 mV dec−1 (Pt/NMC-RT) and 33.7 mV dec−1 (Pt/NMC-RT) has been verified to follow the Volmer–Tafel mechanism. In addition, the Pt/NMC-LT catalyst also has competitive advantages in terms of catalytic activity and Tafel kinetics over related literature (Table S4). Moreover, the excellent catalytic stability of Pt/NMC-LT was further convinced by an accelerated durability test (ADT) with negligible degradation after 5000 CV cycles (Fig. 3d) due to the structural stability of atomically dispersed active sites in Pt/NMC-LT SACs under HER conditions (Fig. S13), compared with deactivated Pt/NMC-RT and commercial Pt/C catalysts (Fig. 3e, S14 and S15). As discussed above, all results suggested the advanced HER performance of the Pt/NMC-LT catalyst by low-temperature one-pot solution synthesis. To gain further insight into the superior electro-catalytic performance of the Pt/NMC hybrid catalyst, we calculated the Gibbs free energy (ΔGH) of the adsorbed H atom on various defective NMC configurations (Fig. S16), as well as on the Pt(111) catalytic facet for comparison. It is well accepted that the closer the binding free energy of the intermediate H* is to zero, the higher the activity is observed.25,26 At the atomic scale, each dispersed Pt single atom serves as an active site. As shown in Fig. 3f, for various Pt/NMC configurations, our calculated ΔGH values were −0.011 eV (olive), 0.044 eV (red), 0.061 eV (pink), 0.106 eV (blue) and −0.039 (purple), compared to −0.094 eV of Pt(111). In general, Pt/NMC hybrid catalysts possess better ΔGH (closer to zero) than that of the Pt(111) surface. On the other hand, for the Pt single atom is located at NMC-SPD sites, H adsorption on this type of site possesses large binding strength (Fig. S17), thus meaning that the hydrogen production reaction on the Pt/NMC system may partly follow the Tafel mechanism. For this case, our calculated H–H association activation barriers are 0.2–0.3 eV lower than that on the Pt(111) catalytic surface (Fig. S18),27,28 which indicates that the HER reaction on this site possesses faster hydrogen evolution dynamics than that on the Pt(111) surface. Overall, our theoretical calculations further confirm the enhanced HER activity of the atomically dispersed Pt atoms on the NMC substrate.


image file: c9ta07469d-f4.tif
Fig. 4 Electro-catalytic performance of different catalysts in 0.5 M H2SO4. (a) Polarization curves of Pt/NMC-LT, Pt/NMC-RT, Pt/C and NMC. (b) The overpotential of Pt/NMC-LT, Pt/NMC-RT and Pt/C at j10 and j100. (c) Tafel plots of Pt/NMC-LT, Pt/NMC-RT and Pt/C. Polarization curves of (d) Pt/NMC-LT and (e) Pt/NMC-RT (collected before and after 5000 CV cycles). (f) The free energy diagram of hydrogen evolution on various Pt/NMC systems. The DFT optimized structures are shown as insets, and ΔGH of hydrogen evolution on the Pt (111) surface is shown for comparison. The C, N, Pt and H atoms are represented as brown, blue, light blue and white spheres, respectively.

Conclusions

In conclusion, we demonstrated that atomically dispersed metals can be directly anchored onto NMC substrates forming highly efficient SACs, only by decreasing the reaction temperature from RT down to −40 °C with a one-shot injection process for metal precursor solutions. Combining HAADF-STEM, XRD, XPS and XAFS techniques with DFT calculations, we confirmed that the dispersion of Pt species can be finely tuned from sub-nanometer to atomic scale, due to the effective inhibition of possible nuclei formation and thus the shift from homogeneous to less energy-driven heterogeneous nucleation. Moreover, the superior performance advantages for atomically dispersed Pt species, compared with Pt clusters (<1 nm) as well as Pt nanoparticles (2–3 nm) have also been examined and verified towards typical electro-catalytic HER in terms of both activity and stability. Considering the heterogeneous catalyst configuration, we believe that this one-pot low-temperature solution synthesis will further promote the large-scale preparation of functional atomic materials for potential applications due to the technology simplicity.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This study is supported by the National Natural Science Foundations of China (Grant No. 51902027, 51788104, 51661135025, 51706117 and U1564205), National Basic Research of China (Grants No. 2015CB932500, 2016YFE0102200 and 2018YFB0104404) and Fundamental Research Funds for the Central Universities (Grant No. 2019RC20 and 2017CX10007). Hongbo Wu also thanks the support from the Graduate Technological Innovation Project of Beijing Institute of Technology (Grant 2018CX10028). We thank the XAFS station (BL14W1) of the Shanghai Synchrotron Radiation Facility.

References

  1. Y. J. Chen, S. F. Ji, C. Chen, Q. Peng, D. S. Wang and Y. D. Li, Joule, 2018, 2, 1242–1264 CrossRef CAS.
  2. J. W. Su, R. X. Ge, Y. Dong, F. Hao and L. Chen, J. Mater. Chem. A, 2018, 6, 14025 RSC.
  3. H. Zhang, G. Liu, L. Shi and J. Ye, Adv. Energy Mater., 2018, 8, 1701343 CrossRef.
  4. L. Liu and A. Corma, Chem. Rev., 2018, 118, 4981–5079 CrossRef CAS PubMed.
  5. K. Huang, J. Liu, L. Wang, G. Chang, R. Wang, M. Lei, Y. Wang and Y. He, Appl. Surf. Sci., 2019, 487, 1145–1151 CrossRef CAS.
  6. W. Zhong, S. Shen, M. He, D. Wang, Z. Wang, Z. Lin, W. Tu and J. Yu, Appl. Catal. B Environ., 2019, 258, 117967 CrossRef CAS.
  7. Z. W. Chen, L. X. Chen, C. C. Yang and Q. Jiang, J. Mater. Chem. A, 2019, 7, 3492–3515 RSC.
  8. L. Zhang, K. Doyle-Davis and X. L. Sun, Energy Environ. Sci., 2019, 12, 492–517 RSC.
  9. J. Wang, Z. Li, Y. Wu and Y. Li, Adv. Mater., 2018, 30, 1801649 CrossRef PubMed.
  10. K. Huang, L. Zhang, T. Xu, H. Wei, R. Zhang, X. Zhang, B. Ge, M. Lei, J. Y. Ma, L. M. Liu and H. Wu, Nat. Commun., 2019, 10, 606 CrossRef PubMed.
  11. N. T. K. Thanh, N. Maclean and S. Mahiddine, Chem. Rev., 2014, 114, 7610–7630 CrossRef CAS PubMed.
  12. Y. Xia, K. D. Gilroy, H. C. Peng and X. Xia, Angew. Chem., Int. Ed., 2017, 56, 60–95 CrossRef CAS PubMed.
  13. E. D. Bøjesen and B. B. Iversen, CrystEngComm, 2016, 18, 8332–8353 RSC.
  14. J. Polte, CrystEngComm, 2015, 17, 6809–6830 RSC.
  15. Y. Sun, Natl. Sci. Rev., 2015, 2, 329–348 CrossRef CAS.
  16. Z. Zhang, Y. Chen, L. Zhou, C. Chen, Z. Han, B. Zhang, Q. Wu, L. Yang, L. Du, Y. Bu, P. Wang, X. Wang, H. Yang and Z. Hu, Nat. Commun., 2019, 10, 1657 CrossRef PubMed.
  17. H. Wei, H. Wu, K. Huang, B. Ge, J. Ma, J. Lang, D. Zu, M. Lei, Y. Yao, W. Guo and H. Wu, Chem. Sci., 2019, 10, 2830–2836 RSC.
  18. S. R. Bahn and K. W. Jacobsen, Comput. Sci. Eng., 2002, 4, 56–66 CrossRef CAS.
  19. A. H. Larsen, J. J. Mortensen, J. Blomqvist, I. E. Castelli, R. Christensen, M. Dułak, J. Friis, M. N. Groves, B. Hammer, C. Hargus, E. D. Hermes, P. C. Jennings, P. B. Jensen, J. Kermode, J. R. Kitchin, E. L. Kolsbjerg, J. Kubal, K. Kaasbjerg, S. Lysgaard, J. B. Maronsson, T. Maxson, T. Olsen, L. Pastewka, A. Peterson, C. Rostgaard, J. Schiøtz, O. Schütt, M. Strange, K. S. Thygesen, T. Vegge, L. Vilhelmsen, M. Walter, Z. Zeng and K. W. Jacobsen, J. Phys.: Condens. Matter, 2017, 29, 273002 CrossRef PubMed.
  20. H. Wei, K. Huang, D. Wang, R. Zhang, B. Ge, J. Ma, B. Wen, S. Zhang, Q. Li, M. Lei, C. Zhang, J. Irawan, L. Liu and H. Wu, Nat. Commun., 2017, 8, 1490 CrossRef PubMed.
  21. B. Qiao, A. Wang, X. Yang, L. F. Allard, Z. Jiang, Y. Cui, J. Liu, J. Li and T. Zhang, Nat. Chem., 2011, 3, 634–641 CrossRef CAS PubMed.
  22. L. DeRita, J. Resasco, S. Dai, A. Boubnov, H. V. Thang, A. S. Hoffman, I. Ro, G. W. Graham, S. R. Bare, G. Pacchioni, X. Q. Pan and P. Christopher, Nat. Mater., 2019, 18, 746–751 CrossRef CAS PubMed.
  23. X. He, S. Z. Luan, L. Wang, R. Y. Wang, P. Du, Y. Y. Xu, H. J. Yang, Y. G. Wang, K. Huang and M. Lei, Mater. Lett., 2019, 244, 78–82 CrossRef CAS.
  24. C. H. Choi, M. Kim, H. C. Kwon, S. J. Cho, S. Yun, H. T. Kim, K. J. J. Mayrhofer, H. Kim and M. Choi, Nat. Commun., 2016, 7, 10922 CrossRef CAS PubMed.
  25. J. K. Nørskov, T. Bligaard, J. Rossmeisl and C. H. Christensen, Nat. Chem., 2009, 1, 37–46 CrossRef PubMed.
  26. B. Hinnemann, P. G. Moses, J. Bonde, K. P. Jørgensen, J. H. Nielsen, S. Horch, I. Chorkendorff and J. K. Nørskov, J. Am. Chem. Soc., 2005, 127, 5308–5309 CrossRef CAS PubMed.
  27. G. F. Wei and Z. P. Liu, Chem. Sci., 2015, 6, 1485–1490 RSC.
  28. E. Skúlason, V. Tripkovic, M. E. Björketun, S. Gudmundsdóttir, G. Karlberg, J. Rossmeisl, T. Bligaard, H. Jónsson and J. K. Nørskov, J. Phys. Chem. C, 2010, 114, 18182–18197 CrossRef.

Footnotes

Electronic supplementary information (ESI) available. See DOI: 10.1039/c9ta07469d
These authors contributed equally.

This journal is © The Royal Society of Chemistry 2019