Selective photoredox decarboxylation of α-ketoacids to allylic ketones and 1,4-dicarbonyl compounds dependent on cobaloxime catalysis

Hong Zhang a, Qian Xiao a, Xu-Kuan Qi a, Xue-Wang Gao b, Qing-Xiao Tong a and Jian-Ji Zhong *a
aDepartment of Chemistry and Key Laboratory for Preparation and Application of Ordered Structural Materials of Guangdong Province, Shantou University, and Chemistry and Chemical Engineering Laboratory of Guangdong Province, Guangdong 515063, P. R. China. E-mail: jjzhong@stu.edu.cn
bKey Laboratory of Photochemical Conversion and Optoelectronic Materials, Technical Institute of Physics and Chemistry, Chinese Academy of Sciences, Beijing 100190, P. R. China

Received 17th August 2020 , Accepted 10th September 2020

First published on 11th September 2020


A photoredox/cobaloxime co-catalyzed coupling reaction of α-ketoacids and methacrylates to obtain allylic ketones is described. Without the cobaloxime catalyst, 1,4-dicarbonyl compounds are generated. The cobaloxime catalyst enables dehydrogenation to generate the formation of new olefins. The generality, good substrate scope and mild conditions are good features in the photoredox/cobaloxime catalysis protocol, and this method will provide new opportunities for the functionalization of more olefins.


The addition of carbon radicals to alkenes has been acknowledged as a powerful and straightforward strategy for constructing C–C bonds.1 In this respect, photoredox catalysis,2 serving as an environmental and efficient approach to initiate the generation of radicals, has seen dramatic developments over the last decade.3 Conventionally, the radical that is generated in situ can be trapped by olefins to give the new alkyl radical species (Scheme 1). Afterwards, two main reaction pathways can occur: hydrogen abstraction of the new alkyl radical, or termination with reductants will result in the generation of alkane products (Scheme 1a). For the formation of new olefins, sacrificial oxidants are usually required (Scheme 1b). However, the use of external oxidants will result in stoichiometric amounts of chemical waste. Recently, Wu and co-workers4 developed a photoredox/cobaloxime co-catalyzed strategy for hydrogen evolution cross-coupling reactions.5 Cobaloximes as mimics of coenzyme B12 have been extensively investigated.6 Therein, cobaloximes were used as catalysts to capture the electrons and protons eliminated from the substrates to generate H2 gas as the only by-product.7 This represents an ideal atom-economic strategy for C–C bond formation from C–H/C–H bonds. Inspired by this, we believe that if cobaloximes were employed in path b of Scheme 1, the new alkyl radical could be captured by the cobaloximes to generate a transient Co(III)–C intermediate, and subsequent β-H elimination would result in the formation of new olefins.8
image file: d0cc05580h-s1.tif
Scheme 1 Reaction types after radical addition.

Allylic ketones are versatile and widely applicable building blocks for the synthesis of complex pharmaceutical molecules and natural products. In light of their importance, methodologies for their synthesis have attracted much attention, involving the reactions of various allylic organometallics with acyl halides,9 or transition-metal catalyzed reactions of pre-functionalized allylic precursors and acyl precursors.10 Despite the considerable progress in this field, the pre-functionalization of substrates has limited the application of these methodologies. In our design, one simple and straightforward approach to accessing allylic ketones is the direct addition of acyl radicals to allylic substrates. In recent years, the decarboxylation of α-ketoacids by photoredox catalysis to generate acyl radicals for the synthesis of alkyl ketones has been extensively investigated.11 However, the photoredox acylation of olefins to unsaturated ketones without the addition of any external oxidants has rarely been reported. In this work, we report a photoredox/cobaloxime co-catalyzed strategy for the synthesis of allylic ketones by direct benzoyl radical addition to methacrylate derivatives (Scheme 2a). In the case where the cobaloxime catalyst is not present in the same reaction, corresponding alkyl ketone products of the 1,4-dicarbonyl compounds were generated (Scheme 2b).


image file: d0cc05580h-s2.tif
Scheme 2 Cross-coupling of α-ketoacids and methacrylate derivatives (this work).

Our initial study focused on the cross-coupling of phenylglyoxylic acid (a1) and ethyl methacrylate (b1). As shown in Table 1, after examining a series of reaction parameters, the standard reaction conditions were identified. The treatment of a1 with b1 in the presence of 5 mol% 4-CzIPN, 20 mol% Co(dmgBF2)2(H2O)2 and 1.5 equivalents of 2,6-lutidine in deaerated CH3CN under 450 nm LED irradiation for 16 h at room temperature selectively gave the allylic ketone c1 in 85% yield (Table 1, entry 1). Other photocatalysts such as Ru(bpy)3(PF6)2 and Acr+-MesClO4 were also investigated, and they were found to be ineffective in this reaction (Table 1, entries 2 and 3). The solvent test proved that CH3CN was the best choice (Table 1, entries 4–6). There were two main factors that we speculated to be critical to the solvent effect: (1) the solubility of the substrate and of the cobaloxime catalyst; (2) the coordination ability of the solvent. No base or other base addition resulted in a less effective reaction than that with 2,6-lutidine (Table 1, entries 7–9). The reaction yield decreased to 54% when Co(dmgH)2PyCl instead of Co(dmgBF2)2(H2O)2 was used (Table 1, entry 10). Control experiments indicated that the photocatalyst, cobaloxime catalyst and base were essential for the synthesis of allylic ketones (Table S1, ESI). Further studies found that when the cobaloxime catalyst was removed in the reaction, the alkyl ketone 1,4-dicarbonyl compound d1 was formed, but only in a 17% yield (Table 1, entry 11). When t-BuOK replaced 2,6-lutidine in the reaction, this made the transformation more feasible, resulting in an 89% yield (Table 1, entry 12).

Table 1 Optimization of the reaction conditionsa

image file: d0cc05580h-u1.tif

Entry Variation from the “standard reaction conditions” Yield (%)b
c1 d1
a Reaction conditions for allylic ketone synthesis: a1 (0.2 mmol), b1 (0.6 mmol), 4-CzIPN (5 mol%), Co(dmgBF2)2(H2O)2 (20 mol%), 2,6-lutidine (0.3 mmol), CH3CN (5 mL), under N2, room temperature, 450 nm LED irradiation for 16–24 h. b Isolated yield. c Reaction conditions for alkyl ketone synthesis: a1 (0.2 mmol), b1 (0.4 mmol), 4-CzIPN (5 mol%), t-BuOK (0.2 mmol), CH3CN (5 mL), under N2, room temperature, 450 nm LED irradiation for 8–12 h.
1 None 85
2 Ru(bpy)3(PF6)2 as a photocatalyst 0
3 Acr+-MesClO4 as a photocatalyst 0
4 CH2Cl2 as a solvent 0
5 DMF as a solvent 0
6 CH3OH as a solvent Trace
7 No 2,6-lutidine 37
8 t-BuOK as a base 47 Trace
9 4-Methylpyridine as a base 51
10 Co(dmgH)2PyCl as a catalyst 54
11 No cobaloxime catalyst 17
12c No cobaloxime catalyst, t-BuOK as a base 89


Next, we investigated the generality and limitations of this approach. Various α-ketoacids were examined, and their corresponding allylic ketones were produced in moderate to good yields (Table 2, c1–c20). The aromatic α-ketoacids with electron-donating groups or electron-withdrawing groups at different positions on the phenyl moiety (ortho, meta and para positions) were found to be compatible substrates in this reaction. Meanwhile, when strong electron-withdrawing groups such as –CF3 and –NO2 were introduced, the reaction efficiency decreased sharply (Table 2, c18 and c21). However, the steric effect cannot be forgotten about, since the reaction efficiency decreased when the ortho-position of the substrates was substituted (Tables 2, c3 62%, c5 55%, c6 45%, c12 46% and c17 42%). It is worth noting that halo-substituted substrates were tolerated well, providing opportunities for the products to undergo further transformations (Table 2, c10–c17). Naphthyl and heterocyclic α-ketoacids could also undergo this reaction to give moderate yields of 50% and 42%, respectively (Table 2, c19 and c20). Unfortunately, no reaction was observed when the aliphatic α-ketoacids were used (Table 2, c22 and c23). Notably, a 10 mmol scale reaction could be carried out smoothly to give c1 in 65% yield (ESI, Scheme S1). For alkyl ketone synthesis, these substrates were also examined for the production of their corresponding 1,4-dicarbonyl compounds, in moderate to good yields, in the absence of the cobaloxime catalyst (Table 2, d1–d21). Nevertheless, some substrates, e.g.d5, d12, d17, d18 and d21, were ineffective in this transformation process, probably due to the steric hindrance effect or electronic effect present.

Table 2 Scope of the α-ketoacidsa
a Reaction conditions for allylic ketone synthesis c: a (0.2 mmol), b1 (0.6 mmol), 4-CzIPN (5 mol%), Co(dmgBF2)2(H2O)2 (20 mol%), 2,6-lutidine (0.3 mmol), CH3CN (5 mL), under N2, room temperature, 450 nm LED irradiation for 16∼24 h. For alkyl ketone synthesis d: a (0.2 mmol), b1 (0.4 mmol), 4-CzIPN (5 mol%), t-BuOK (0.2 mmol), CH3CN (5 mL), under N2, room temperature, 450 nm LED irradiation for 8–12 h. b Isolated yield.
image file: d0cc05580h-u2.tif


We further explored the variety of the methacrylate derivatives. To our delight, the methacrylate derivatives bearing different kinds of groups were tolerated well in this transformation, including various linear or branched aliphatic chain groups, or benzyl groups with electron-donating or electron-withdrawing substituents (Table 3, c24–c39). Furan-2-ylmethyl methacrylate also proceeded smoothly in the reaction to give the corresponding allylic ketones in 53% yield (Table 3, c38). It is worth noting that when allyl methacrylate was used in this reaction, the selective addition of a benzoyl radical to the double bond of a methylallyl group, rather than to an allyl group, led to the formation of the corresponding allylic ketone (Table 3, c39). Other alkenes were also investigated in the reaction, however we found that the scope of the alkenes was limited to methacrylate derivatives (Table S2, ESI, b40–b45). Furthermore, the substrate scope of alkyl ketone synthesis was also investigated without the use of the cobaloxime catalyst, where good to excellent yields could be achieved, as shown in Table 3 (d24–d39).

Table 3 Scope of the methacrylate derivativesa
a Reaction conditions for allylic ketone synthesis c: a1 (0.2 mmol), b (0.6 mmol), 4-CzIPN (5 mol%), Co(dmgBF2)2(H2O)2 (20 mol%), 2,6-lutidine (0.3 mmol), CH3CN (5 mL), under N2, room temperature, 450 nm LEDs irradiation for 16–24 h. For alkyl ketones synthesis d: a1 (0.2 mmol), b (0.4 mmol), 4-CzIPN (5 mol%), t-BuOK (0.2 mmol), CH3CN (5 mL), under N2, room temperature, 450 nm LEDs irradiation for 8–12 h. b Isolated yield.
image file: d0cc05580h-u3.tif


In order to gain insights into the reaction mechanism, more experiments were carried out. Quenching experiments showed that a1 only slightly quenched the emission of 4-CzIPN. However, the addition of a1 to the base 2,6-lutidine could sharply decrease the luminescence (Fig. S1, ESI), revealing that this photoinduced electron transfer from the anion of a1 to 4-CzIPN is more feasible than with a1 alone. Furthermore, we found that a radical inhibitor (2,2,6,6-tetramethyl-1-piperidinyloxy (TEMPO)) completely inhibited the reaction and the TEMPO-trapped benzoyl product was generated in 69% yield, indicating that a benzoyl radical is involved in this reaction (Scheme 3a). We further investigated the kinetic isotope effect (KIE) of benzyl methacrylate (b34), where the intermolecular KIE value of 1.56 indicated that the methyl C–H elimination process was fast and was not involved in the rate-determining step in this reaction (Scheme 3b). Due to the fact that for some substrates, which were not sufficient enough such as a5 (c5, 55%) and a6 (c6, 45%), corresponding aldehydes were observed (Scheme S3, ESI), we speculated that acyl radical addition to alkenes might be involved in the rate determining step.


image file: d0cc05580h-s3.tif
Scheme 3 Reaction mechanism investigations.

Based on these above results and on previous literature reports, a plausible reaction pathway is proposed for the synthesis of allylic ketones (Scheme 4). The excited *4-CzIPN is reductively quenched by the anion of a1, leading to a benzoyl radical with CO2 extrusion, which was detected by gas chromatography, and the reductive species 4-CzIPN. The redox potentials from previous reports12 also support the electron transfer from the anion of a1 (0.98 V vs. SCE) to the excited 4-CzIPN compound (1.35 V vs. SCE). Benzoyl radical addition to the double bond of ethyl methacrylate generates a new alkyl radical species A, which could be captured by Co(II) to form a transient Co(III)–C intermediate. The cleavage of Co–C and subsequent β-H elimination result in the final allylic ketone c1 product and Co(III)–H. According to previous reports,4–6 a plausible pathway for Co(III)–H reacting with a proton or another Co(III)–H species to generate H2 and Co(III) might occur. Indeed, detection of H2 by GC-TCD confirmed this possibility. However, only 23% H2 was detected, revealing that, herein, other pathways to generate Co(III) could also occur. Due to this, the excess of methacrylate is necessary in the reaction, and we speculated that b1 might compete with the proton to react with Co(III)–H to generate a reductive by-product and Co(III). Subsequent single electron transfer from reductive 4-CzIPN to Co(III) regenerates both the photoredox catalytic cycle and the cobaloxime catalytic cycle. A plausible mechanism for the synthesis of alkyl ketones is also described in Scheme S4 (ESI). Without the cobaloxime catalyst, the new alkyl radical species A is directly reduced by the reductive 4-CzIPN to the alkyl anion species C. Protonation of C furnishes the alkyl ketones.


image file: d0cc05580h-s4.tif
Scheme 4 Proposed reaction pathway for the synthesis of the allylic ketones.

In summary, we have described a photoredox/cobaloxime co-catalyzed dehydrogenative cross-coupling reaction of α-ketoacids and methacrylate derivatives for the synthesis of allylic ketones. Interestingly, the products could be selectively achieved by controlling the catalytic system. When the cobaloxime catalyst was removed, the corresponding alkyl ketones, rather than the allylic ketones, were produced via photoredox catalysis. Herein, the cobaloxime catalyst enabled the dehydrogenation process in order to achieve the formation of new olefins. The generality, good substrate scope and mild reaction conditions are features of the dual photoredox/cobaloxime catalysis protocol. We believe that this protocol will provide opportunities for more radical-involved olefin functionalization reactions, which are being carried out in our laboratory.

The authors gratefully acknowledge financial support from the National Natural Science Foundation of China (21801163 and 21702213), the STU Scientific Research Foundation for Talents (NTF18003) and the Guangdong Province Universities and Colleges Pearl River Scholar Funded Scheme 2019 (GDUPS2019).

Conflicts of interest

There are no conflicts to declare.

Notes and references

  1. C. P. Jasperse, D. P. Curran and T. L. Fevig, Chem. Rev., 1991, 91, 1237 Search PubMed.
  2. (a) K. Zeitler, Angew. Chem., Int. Ed., 2009, 48, 9785 Search PubMed; (b) T. P. Yoon, M. A. Ischay and J. Du, Nat. Chem., 2010, 2, 527 Search PubMed; (c) J. Xuan and W. J. Xiao, Angew. Chem., Int. Ed., 2012, 51, 6828 Search PubMed; (d) C. K. Prier, D. A. Rankic and D. W. C. MacMillan, Chem. Rev., 2013, 113, 5322 Search PubMed; (e) N. A. Romero and D. A. Nicewicz, Chem. Rev., 2016, 116, 10075 Search PubMed; (f) J. P. Goddard, C. Ollivier and L. Fensterbank, Acc. Chem. Res., 2016, 49, 1924 Search PubMed; (g) D. Staveness, I. Bosque and C. R. J. Stephenson, Acc. Chem. Res., 2016, 49, 2295 Search PubMed; (h) Q. Liu and L. Z. Wu, Natl. Sci. Rev., 2017, 4, 359 Search PubMed; (i) B. G. Cai, J. Xuan and W. J. Xiao, Sci. Bull., 2019, 64, 337 Search PubMed; (j) J. Xuan, X. K. He and W. J. Xiao, Chem. Soc. Rev., 2020, 49, 2546 Search PubMed.
  3. (a) Y. Chen, L. Q. Lu, D. G. Yu, C. J. Zhu and W. J. Xiao, Sci. China: Chem., 2018, 62, 24 Search PubMed; (b) J. J. Zhong, Q. Y. Meng, G. X. Wang, Q. Liu, B. Chen, K. Feng, C. H. Tung and L. Z. Wu, Chem. – Eur. J., 2013, 19, 6443 Search PubMed; (c) J. J. Zhong, C. Yang, X. Y. Chang, C. Zou, W. Lu and C. M. Che, Chem. Commun., 2017, 53, 8948 Search PubMed; (d) X. K. Qi, H. Zhang, Z. T. Pan, R. B. Liang, C. M. Zhu, J. H. Li, Q. X. Tong, X. W. Gao, L. Z. Wu and J. J. Zhong, Chem. Commun., 2019, 55, 10848 Search PubMed; (e) J. J. Zhong, W. P. To, Y. Liu, W. Lu and C. M. Che, Chem. Sci., 2019, 10, 4883 Search PubMed; (f) S. Cai, Y. Xu, D. Chen, L. Li, Q. Chen, M. Huang and W. Weng, Org. Lett., 2016, 18, 2990 Search PubMed.
  4. (a) B. Chen, L. Z. Wu and C. H. Tung, Acc. Chem. Res., 2018, 51, 2512 Search PubMed; (b) J. J. Zhong, Q. Y. Meng, B. Chen, C. H. Tung and L. Z. Wu, Acta Chim. Sin., 2017, 75, 34 Search PubMed; (c) J. J. Zhong, Q. Y. Meng, B. Liu, X. B. Li, X. W. Gao, T. Lei, C. J. Wu, Z. J. Li, C. H. Tung and L. Z. Wu, Org. Lett., 2014, 16, 1988 Search PubMed; (d) Q. Y. Meng, J. J. Zhong, Q. Liu, X. W. Gao, H. H. Zhang, T. Lei, Z. J. Li, K. Feng, B. Chen, C. H. Tung and L. Z. Wu, J. Am. Chem. Soc., 2013, 135, 19052 Search PubMed; (e) X. W. Gao, Q. Y. Meng, J. X. Li, J. J. Zhong, T. Lei, X. B. Li, C. H. Tung and L. Z. Wu, ACS Catal., 2015, 5, 2391 Search PubMed; (f) G. Zhang, C. Liu, H. Yi, Q. Meng, C. Bian, H. Chen, J. X. Jian, L. Z. Wu and A. Lei, J. Am. Chem. Soc., 2015, 137, 9273 Search PubMed; (g) Y. W. Zheng, B. Chen, P. Ye, K. Feng, W. Wang, Q. Y. Meng, L. Z. Wu and C. H. Tung, J. Am. Chem. Soc., 2016, 138, 10080 Search PubMed; (h) W. Q. Liu, T. Lei, S. Zhou, X. L. Yang, J. Li, B. Chen, J. Sivaguru, C. H. Tung and L. Z. Wu, J. Am. Chem. Soc., 2019, 141, 13941 Search PubMed.
  5. (a) H. Wang, X. Gao, Z. Lv, T. Abdelilah and A. Lei, Chem. Rev., 2019, 119, 6769 Search PubMed; (b) G. Zhang, L. Zhang, H. Yi, Y. Luo, X. Qi, C. H. Tung, L. Z. Wu and A. Lei, Chem. Commun., 2016, 52, 10407 Search PubMed; (c) S. Tang, L. Zeng and A. Lei, J. Am. Chem. Soc., 2018, 140, 13128 Search PubMed; (d) K. H. He and Y. Li, ChemSusChem, 2014, 7, 2788 Search PubMed; (e) K. H. He, F. F. Tan, C. Z. Zhou, G. J. Zhou, X. L. Yang and Y. Li, Angew. Chem., Int. Ed., 2017, 56, 3080 Search PubMed; (f) H. Cao, H. M. Jiang, H. Y. Feng, J. M. C. Kwan, X. G. Liu and J. Wu, J. Am. Chem. Soc., 2018, 140, 16360 Search PubMed; (g) X. Sun, J. T. Chen and T. Ritter, Nat. Chem., 2018, 10, 1229 Search PubMed; (h) W. L. Yu, Y. C. Luo, L. Yan, D. Liu, Z. Y. Wang and P. F. Xu, Angew. Chem., Int. Ed., 2019, 58, 10941 Search PubMed.
  6. (a) G. N. Schrauzer, Acc. Chem. Res., 1968, 1, 97 Search PubMed; (b) G. Pattenden, Chem. Soc. Rev., 1988, 17, 361 Search PubMed . And references therein.
  7. (a) J. L. Dempsey, B. S. Brunschwig, J. R. Winkler and H. B. Gray, Acc. Chem. Res., 2009, 42, 1995 Search PubMed; (b) T. Lazarides, T. McCormick, P. W. Du, G. G. Luo, B. Lindley and R. Eisenberg, J. Am. Chem. Soc., 2009, 131, 9192 Search PubMed; (c) V. Artero, C. M. Kerlidou and M. Fontecave, Angew. Chem., Int. Ed., 2011, 50, 7238 Search PubMed; (d) J. C. Hu, S. Sun, M. D. Li, W. Xia, J. Wu, H. Liu and F. Wang, Chem. Commun., 2019, 55, 14490 Search PubMed.
  8. (a) M. E. Weiss, L. M. Kreis, A. Lauber and E. M. Carreira, Angew. Chem., Int. Ed., 2011, 50, 11125 Search PubMed; (b) M. E. Weiss and E. M. Carreira, Angew. Chem., Int. Ed., 2011, 50, 11501 Search PubMed.
  9. (a) L. Zhang, Y. Huang, H. Jiang, D. Jun and Y. Liao, J. Org. Chem., 1992, 57, 774 Search PubMed; (b) R. C. Larock and Y. D. Lu, J. Org. Chem., 1993, 58, 2846 Search PubMed.
  10. (a) Y. Obora, Y. Ogawa, Y. Imai, T. Kawamura and Y. Tsuji, J. Am. Chem. Soc., 2001, 123, 10489 Search PubMed; (b) J. Huang, E. Bunel and M. M. Faul, Org. Lett., 2007, 9, 4343 Search PubMed; (c) L. Cui, H. Chen, C. Liu and C. Li, Org. Lett., 2016, 18, 2188 Search PubMed.
  11. (a) F. Penteado, E. F. Lopes, D. Alves, G. Perin, R. G. Jacob and E. J. Lenardão, Chem. Rev., 2019, 119, 7113 Search PubMed; (b) J. Xuan, Z. G. Zhang and W. J. Xiao, Angew. Chem., Int. Ed., 2015, 54, 15632 Search PubMed; (c) H. Tan, H. Li, W. Ji and L. Wang, Angew. Chem., Int. Ed., 2015, 54, 8374 Search PubMed; (d) G. Bergonzini, C. Cassani and C. J. Wallentin, Angew. Chem., Int. Ed., 2015, 54, 14066 Search PubMed; (e) M. Zhang, R. Ruzi, J. Xi, N. Li, Z. K. Wu, W. Li, S. Yu and C. Zhu, Org. Lett., 2017, 19, 3430 Search PubMed; (f) L. Capaldo, R. Riccardi, D. Ravelli and M. Fagnoni, ACS Catal., 2018, 8, 304 Search PubMed; (g) J. Genovino, Y. J. Lian, Y. Zhang, T. O. Hope, A. Juneau, Y. Gagné, G. Ingle and M. Frenette, Org. Lett., 2018, 20, 3229 Search PubMed; (h) H. X. Zou, Y. Li, Y. Yang, J. H. Li and J. Xiang, Adv. Synth. Catal., 2018, 360, 1439 Search PubMed; (i) M. J. Luo, T. T. Zhang, F. J. Cai, J. H. Li and D. L. He, Chem. Commun., 2019, 55, 7251 Search PubMed; (j) J. J. Zhao, H. H. Zhang, X. Shen and S. Yu, Org. Lett., 2019, 21, 913 Search PubMed; (k) J. Lu, X. K. He, X. Cheng, A. J. Zhang, G. Y. Xu and J. Xuan, Adv. Synth. Catal., 2020, 362, 2178 Search PubMed.
  12. (a) H. Uoyama, K. Goushi, K. Shizu, H. Nomura and C. Adachi, Nature, 2012, 492, 234 Search PubMed; (b) J. Liu, Q. Liu, H. Yi, C. Qin, R. Bai, X. Qi, Y. Lan and A. Lei, Angew. Chem., Int. Ed., 2014, 53, 502 Search PubMed.

Footnote

Electronic supplementary information (ESI) available. See DOI: 10.1039/d0cc05580h

This journal is © The Royal Society of Chemistry 2020