Boosting the utilization efficiency of glucose via a favored C–C coupling reaction

Yaxuan Jinga, Yayun Zhangb, Qing Lva, Yong Guoa, Xiaohui Liua and Yanqin Wang*a
aKey Laboratory for Advanced Materials and Joint International Research Laboratory of Precision Chemistry and Molecular Engineering, Feringa Nobel Prize Scientist Joint Research Center, Research Institute of Industrial Catalysis, School of Chemistry and Molecular Engineering, East China University of Science and Technology, Shanghai, 200237, China. E-mail:
bKey Laboratory of Chemical Engineering, East China University of Science and Technology, Shanghai 200237, China

Received 24th August 2019 , Accepted 23rd October 2019

First published on 31st October 2019

The efficient utilization of renewable glucose has enormous potential to reduce the excessive dependence on fossil resources, but remains challenging owing to the inevitable formation of undesirable humins and multistep processes. Reported here is a new 2,4-pentanedione/H2O–NaCl biphasic system for the direct one-step conversion of glucose to high-quality fuel precursors with a high utilization efficiency of glucose (90%). The formation of humins is markedly restrained with an entirely new pathway, i.e. glucose directly reacts with 2,4-pentanedione to afford fuel precursors via several chemical transformations, which is totally different from the traditional glucose-to-fructose-to-HMF way. Control experiments and DFT calculations confirm that this new pathway is thermodynamically favorable compared with the traditional way. This simple system is also capable of converting starch to fuel precursors, paving a new way for the efficient valorization of glucose-based natural sugars.


Carbohydrates are the most abundant form (ca. 75%) of renewable biomass and are produced through natural photosynthesis.1 Glucose, as the most widespread and important monosaccharide in nature, represents the major building block of carbohydrates and its conversion to a wide variety of chemicals, fuels and carbon-based functional materials has attracted significant attention.1b,c,2

Starting from glucose, a wide array of chemicals, such as various alcohols, 5-hydroxymethylfurfural (HMF), lactic acid, gluconic acid, 2,5-dimethylfuran, pentanoic acid esters, and levulinic acid, can be produced via state-of-the-art catalytic systems and reaction networks.1b,3 Among these, the production of HMF has been intensively explored, since a large number of value-added chemicals and fuels can be synthesized through using HMF as the substrate.3a,4 However, during the conversion of glucose, especially the glucose-to-HMF process, large amounts of undesirable humins are inevitably generated, leading to a low utilization ratio of glucose.5 Hence, avoiding the generation of humins to improve the utilization efficiency of glucose is still an important scientific challenge.

Previous studies reveal that humins are mainly formed through self-polymerization of HMF and cross-polymerization of HMF with glucose or fructose.5b,6 To prevent the formation of humins and enhance the utilization efficiency of glucose, designing a new reaction pathway to overcome thermodynamically humin generation is highly desirable. Glucose and its intermediates generated during conversion contain aldehyde groups and can directly react with active ketones to obtain long-chain oxygenates via aldol condensation. After total hydrodeoxygenation, the prepared long-chain oxygenates can be converted to alkanes and directly used as transportation fuels or additives.7 2,4-Pentanedione, as a typical β-dicarbonyl compound, can be produced from renewable biomass by chemical conversion or fermentation treatment and its corresponding enol is readily formed due to its resonance stabilization, making it a good carbon nucleophile to attack active electrophilic aldehydes.8 Inspired by the above understanding, we believe that the highly active catcher may rapidly attack the aldehyde groups of glucose or the intermediates to avoid the traditional glucose-to-fructose-to-HMF process, resulting in the significant suppression of humin generation.

Herein, a simple system for the direct chemical transformation of glucose into fuel precursors in a 2,4-pentanedione /H2O–NaCl biphasic system is designed in the absence of any external-added acid catalysts under mild conditions (Scheme 1). The generation of humins is markedly prevented owing to a clearly novel reaction pathway. Control experiments and DFT calculations are conducted to rationalize this new route. We also demonstrate that this new system enables the direct conversion of starch into fuel precursors with a high carbon yield.

image file: c9gc02987g-s1.tif
Scheme 1 A comparison of the previous strategy and the new strategy for the conversion of glucose.

Results and discussion

Investigation of the reaction process

A previous study found that Cl ions can catalyze the conversion of glucose and cellulose to HMF.9 Here, the simple and environmentally-friendly reaction system was also conducted in the solution of NaCl, and the results are collected in Table 1. When using glucose as the substrate, the condensation products, denoted as C11-glu (main product) and C16-glu (Fig. 1) were obtained with an exceptionally high total carbon yield (84.6%). Detailed analysis shows that it really passes through a new reaction pathway i.e. glucose or its reaction intermediate is directly reacted with 2,4-pentanedione via a series of reactions including aldol condensation, dehydration, enol–keto tautomerism, and ring closing to obtain the C11 condensation product. In addition, a small number of deep condensation products, namely C16-glu, were also detected. According to our previous study, C16-glu were produced through the Michael addition of 2,4-pentanedione with C11-glu, and the proposed reaction routes are presented in Scheme S1.[thin space (1/6-em)]7b To further confirm this new reaction pathway, fructose and HMF were also used as substrates, the same condensation products, denoted as C11-HMF (main product) and C16-HMF, were obtained with a total yield of 50.8% and 78.8%, respectively, which are all lower than that when using glucose as the substrate. Based on the structural analysis (Fig. S2–S8) and the same products obtained from HMF and fructose, we can conclude that fructose first undergoes dehydration to HMF, and then HMF converts to C11-HMF by aldol condensation with 2,4-pentanedione. It must be noted that the carbon yield from glucose is even high than that from HMF, indicating the absolute advantages of this reaction pathway.
image file: c9gc02987g-f1.tif
Fig. 1 Optimization of reaction temperature for the conversion of glucose and starch into C11-glu and C16-glu. (a) 0.1 g glucose, 1 g H2O, 0.3 g NaCl, 6 g 2,4-pentanedione, 0.5 MPa N2, 3 h; (b) 0.1 g starch, 1 g H2O, 0.1 g NaCl, 6 g 2,4-pentanedione, 0.5 MPa N2, 24 h.
Table 1 Reaction results in a 2,4-pentanedione/H2O–NaCl biphasic system using various substrates
Substrates Products Conv. (%) Yield (%)
C11-glu/C11-HMF C16-glu/C16-HMF C11–16-glu/C11–16-HMF
Reaction conditions: 0.1 g substrate, 1 g H2O saturated with NaCl (0.3 g), 6 g 2,4-pentanedione, 0.5 MPa N2, 150 °C, 3 h.
Glucose C11-glu, C16-glu 99.8 70.7 13.9 84.6
Fructose C11-HMF, C16-HMF 93.1 47.8 3.0 50.8
HMF C11-HMF, C16-HMF >99.9 71.4 7.4 78.8

The influence of reaction conditions on the yield of C11–16-glu was investigated and the results (Fig. 1 and Fig. S9) showed that a reaction at 140 °C for 4 h is optimal, yielding 90% of C11–16-glu. Interestingly, when the concentration of NaCl was reduced from 30% to 10%, the total carbon yield and product distribution remained unchanged, also when the concentration of NaCl was further decreased to 3%, it still remained at 80.1% (Fig. S10). In addition, when the ratio of 2,4-pentanedione/H2O was reduced from 6/1 to 3/1, the carbon yield reduced slightly (from 90% to 88.3%) (Fig. S11). All these results indicate that the extraction effect is not important for the highly efficient utilization of glucose in this system, rather it is reaction pathway predominated. In addition, C11-glu can be upgraded into more valuable C16-glu via deep condensation and the detailed discussion is provided in the ESI (Fig. S12)

Starting from glucose to produce HMF, almost all of the previous reports could not avoid the generation of plenty of humins, giving a low utilization ratio of carbon in glucose (generally less than 60%).6a,10 We also verify the traditional process in a THF/H2O–NaCl biphasic system and found that the highest yield of HMF was less than 60% (Fig. S13), which is in well accordance with the results reported in most literature studies.6a,9a,10b,11 Clearly, our system successfully prevents the generation of a large amount of humins and breaks the bottleneck of the utilization efficiency of carbon in glucose. In addition, the milder reaction temperature (140 °C) compared with that (190 °C) of the THF/H2O–NaCl biphasic system also confirms the entirely different reaction pathway (Table S1).

Our previous studies fully proved that Pd-supported Nb-based catalysts have excellent activity and stability for the total hydrodeoxygenation.7b,g Here, we also presented an example of hydrodeoxygenation of C11-glu over Pd/NbOPO4. As expected, a large amount of C11 branched alkanes was achieved (Fig. S14 and S15). In addition, a small amount of C–C cleavage products, namely C9 alkanes, was generated by the hydrocracking of the branched C2 alkyl groups, which is consistent with previous reports.7b,d,e

Reaction pathway and theoretical studies

On the basis of the above results and discussion, we propose a new reaction pathway (1–10, Pathway 3), see Fig. 2a. This is of obvious dominance compared with the traditional glucose–fructose–HMF process (Pathway 1) in the 2,4-pentanedione/H2O–NaCl biphasic system. It can be predicted that 10 was formed from glucose through several possible reaction pathways (Scheme S2) and two typical ones (Pathway 2 and Pathway 3) are elucidated here. One is that glucose is directly reacted with 2,4-pentanedione to give 6 through aldol condensation, which then undergoes a series of reactions including dehydration, enol–keto tautomerism, ring closing and re-dehydration to form 10 (Pathway 2). Another is that glucose is first converted to 3 via dehydration and enol–keto tautomerism, which then undergoes dehydration and ring closing to form 5, finally obtaining 10 through aldol condensation. Obviously, the competition among the initial routes 1–2, 1–6 and 1–11 is the key for the direction of the following reaction. To explore and compare the possibility of each reaction pathway, thermodynamic analysis is conducted using DFT in detail (see the ESI for computational details). Profiles of Gibbs free energy (ΔGa) changes along each reaction pathway are presented in Fig. 2a. It can be predicted that the direct aldol condensation between glucose and 2,4-pentanedione in pathway 2 is not supported thermodynamically due to the positive ΔGa (1.6 kJ mol−1). In contrast, a slight decrease in ΔGa, that is −3.6 kJ mol−1, occurs in the enol–keto tautomerism of glucose conversion to 11, indicating that the reaction can take place under the given reaction conditions. Interestingly, the dehydration between C3–OH and C2–H of glucose is thermodynamically favored with a large negative ΔGa of −62.6 kJ mol−1. Thus, the initial reaction 1–2 in pathway 3 is predicted to be the dominating step compared with the other reactions considered, meanwhile, revealing that pathway 3 including the initial dehydration is more likely to take place in terms of thermodynamic analysis.
image file: c9gc02987g-f2.tif
Fig. 2 (a) Profiles of Gibbs free energy changes along each reaction pathway; (b) the relative free energy along with the glucose conversion to C11-glu under aqueous conditions with H+ and Cl. White, red, grey, yellow, and green represent H, O, C, H+, and Cl, respectively.

The ΔGa changes along the whole reaction pathways are also depicted. Based on previous studies, fructose is indicated to be a thermally stable intermediate during glucose conversion to HMF.1a,4c,6a,12 The ΔGa of the reactant (glucose) and the final product (fructose) in pathway 1 is −31.6 kJ mol−1, which is much higher than that from glucose to C11-glu (−138.9 kJ mol−1), indicating that the formation of C11-glu is much more favored under the current reaction conditions theoretically. Although the subsequent dehydration of 13 to HMF is thermodynamically supported, the first formation of 13 which is the rate-determining step hinders the whole glucose conversion to HMF. This is in well agreement with the high yield of C11-glu obtained simultaneously without any HMF and C11-HMF in the corresponding experiment result. Looking into the reactions constituting pathway 3, the ΔGa of each step is found to be negative, indicating a smooth reaction process from glucose to C11-glu, further confirming that the reasonability of the proposed pathway is thermodynamically supported. It should be noted that the ΔGa of aldol condensation from 5 to C11-glu is −12.3 kJ mol−1, which is over three times than that from HMF to C11-HMF (−3.8 kJ mol−1), indicating that aldol condensation between 5 and 2,4-pentanedione is more favorable, which is also consistent with the higher yield of C11-glu compared with that of C11-HMF given in the Investigation of the reaction process section.

Detailed pathway 3 and roles of catalysts

It has been reported that Cl favors the formation of HMF from glucose.9 Here, the function of Cl was also investigated and showed that Cl really plays a key role in the 2,4-pentanedione/H2O–NaCl biphasic system (Table S2, Fig. S16–S21); the detailed discussion is provided in the ESI.

Generally, a small amount of H3O+ ions formed in situ in hot water (>200 °C) can work as Brønsted acids and promote the conversion of sugars.9a,12c However, the amount of H3O+ ions would be very low at 140 °C in our system, so the role of H3O+ ions from acetic acid formed by the hydrolysis of 2,4-pentanedione was studied and the results showed that acetic acid does have catalytic activity for the reaction (Fig. S22). In addition, we also confirmed that NaCl cannot promote the hydrolysis of 2,4-pentanedione through controlled experiments (Table S3). In short, both Cl and acetic acid are responsible for the conversion of glucose into C11-glu in the 2,4-pentanedione/H2O–NaCl biphasic system.

The relative free energy landscape of the aforementioned pathway is also explored to conduct a kinetic analysis in the presence and absence of a catalyst. In the bare H2O system, reactions related to glucose are difficult to occur due to extremely high energy barriers, especially for those dehydrations (Fig. S23). This is ascribed to the high Bond Dissociation Energy (BDE) of the C–O bond, resulting in direct bond-breaking being tough. The highest free energy barrier (ΔGb), as high as 201.6 kJ mol−1, is observed in the reaction from 3 to 4, which is therefore the rate-determining step of the pathway. Although the proposed reaction route is thermodynamically supported, the relatively high energy barriers prevent the involved reactions to proceed under mild conditions, leaving the glucose intact in this reaction system. With the assistance of the acetic acid generated from the hydrolysis of 2,4-pentanedione, dehydrations of glucose are accelerated through initial protonation and subsequent deprotonation (Fig. 2b). The free energy barriers decrease from 187.4 and 201.6 to 139.7 and 134.7 kJ mol−1 for the beginning sequential dehydrations, respectively, forming intermediate 4. Followed by the aldol condensation and the third dehydration, which is also promoted via a decrease in ΔGb (from 180.3 to 170.2 kJ mol−1), the final product C11-glu is generated. Interestingly, the addition of Cl ions into the system could induce hydrogen bonds between the –OHs of glucose or intermediates and Cl ions, stabilizing the active intermediate structures during glucose conversion processes. Therefore, the ΔGb can be further reduced to 80.1, 123.0, and 110.2 kJ mol−1 in the corresponding involved dehydrations, dramatically stimulating the reaction steps of glucose conversion to C11-glu, which is in good agreement with the enhanced product yield and selectivity with acetic acid and Cl ions. These results also in turn recommend the reasonability of the newly proposed reaction mechanism associated with the current 2,4-pentanedione/H2O–NaCl biphasic system.

Conversion of starch

We also attempted the direct conversion of starch, which is a natural polysaccharide, into C11-glu and C16-glu in the 2,4-pentanedione/H2O–NaCl biphasic system (Fig. 1b). The results showed that starch was efficiently converted into C11-glu and C16-glu at 150 °C after 24 h, and a high yield of C11–16-glu (79.3%) was obtained, indicating that this new system enables an efficient valorization of starch and the direct transformation of starch into branched biofuel precursors.


In summary, we have presented a single-step process for the conversion of glucose into branched fuel precursors in a 2,4-pentanedione/H2O–NaCl biphasic system. This system is significantly successful in preventing the formation of humins because of a completely new reaction pathway, resulting in a high utilization ratio of glucose. Controlled experiments and DFT calculations reveal that the 2,4-pentanedione/H2O–NaCl biphasic system favors the direct conversion of glucose to fuel precursors rather than the traditional multistep processes. We also demonstrate that this 2,4-pentanedione/H2O–NaCl biphasic system is capable of converting abundant starch into branched fuel precursors. This work provides a new way for renewable glucose-based sugar valorization with high utilization efficiency and opens a novel strategy for the direct production of branched fuel precursors from sustainable sources.

Based on the above results and discussion, we find that this system has four obvious advantages. First and foremost, a state-of-the-art system is designed to prevent the formation of a large amount of humins, resulting in high utilization efficiency of carbon in glucose. Secondly, this system is implemented in the absence of any external-added acid catalysts with a high yield of fuel precursor under very mild reaction conditions. Thirdly, the thus-prepared fuel precursor lies in the range of diesel and jet fuel precursors and contains a branched chain, making it a promising substitute for fossil-based fuels. Finally, this system enables the direct conversion of natural starch to fuel precursors without any pre-treatment.

Conflicts of interest

There are no conflicts to declare.


This work was supported financially by the NSFC of China (No. 21832002, 21872050, and 21808063), the Shanghai Municipal Science and Technology Major Project (No. 2018SHZDZX03), the Fundamental Research Funds for the Central Universities (222201718003), the Science and Technology Commission of Shanghai Municipality (18ZR1408500) and the “Zhang Jiangshu” excellent PhD scheme of ECUST.

Notes and references

  1. (a) A. M. Ruppert, K. Weinberg and R. Palkovits, Angew. Chem., Int. Ed., 2012, 51, 2564–2601 CrossRef CAS PubMed; (b) L. T. Mika, E. Csefalvay and A. Nemeth, Chem. Rev., 2018, 118, 505–613 CrossRef CAS; (c) Y. Liu, G. Li, Y. Hu, A. Wang, F. Lu, J.-J. Zou, Y. Cong, N. Li and T. Zhang, Joule, 2019, 3, 1028–1036 CrossRef CAS.
  2. (a) Z. Zhang and G. W. Huber, Chem. Soc. Rev., 2018, 47, 1351–1390 RSC; (b) H. B. Zhao, J. E. Holladay, H. Brown and Z. C. Zhang, Science, 2007, 316, 1597–1600 CrossRef CAS PubMed; (c) D. Klemm, B. Heublein, H. P. Fink and A. Bohn, Angew. Chem., Int. Ed., 2005, 44, 3358–3393 CrossRef CAS PubMed; (d) Y. Jing, Y. Guo, Q. Xia, X. Liu and Y. Wang, Chem, 2019, 5, 2520–2546 CrossRef CAS; (e) G. Qiang, R. Limin, K. Prashant, J. C. Viktor, K. A. Mkhoyan, E. D. Mark and T. Michael, Angew. Chem., Int. Ed., 2018, 57, 4926–4930 CrossRef PubMed.
  3. (a) C. Shuo, W. Robert, D. Franck, M. Eric and R. Sébastien, Chem. Rev., 2018, 118, 11023–11117 CrossRef PubMed; (b) B. Saha and M. M. Abu-Omar, Green Chem., 2014, 16, 24–38 RSC; (c) V. Choudhary, S. H. Mushrif, C. Ho, A. Anderko, V. Nikolakis, N. S. Marinkovic, A. I. Frenkel, S. I. Sandler and D. G. Vlachos, J. Am. Chem. Soc., 2013, 135, 3997–4006 CrossRef CAS PubMed; (d) J. Wang, J. Xi, Q. Xia, X. Liu and Y. Wang, Sci. China: Chem., 2017, 60, 870–886 CrossRef CAS; (e) G. Yong, Y. Zhang and J. Y. Ying, Angew. Chem., Int. Ed., 2008, 47, 9345–9348 CrossRef CAS PubMed.
  4. (a) Y. Roman-Leshkov, J. N. Chheda and J. A. Dumesic, Science, 2006, 312, 1933–1937 CrossRef CAS PubMed; (b) Q. Hou, M. Zhen, L. Liu, Y. Chen, F. Huang, S. Zhang, W. Li and M. Ju, Appl. Catal., B, 2018, 224, 183–193 CrossRef CAS; (c) J. Tang, L. Zhu, X. Fu, J. Dai, X. Guo and C. Hu, ACS Catal., 2017, 7, 256–266 CrossRef CAS.
  5. (a) V. V. Ordomsky, J. van der Schaaf, J. C. Schouten and T. A. Nijhuis, ChemSusChem, 2013, 6, 1697–1707 CrossRef CAS PubMed; (b) S. K. R. Patil and C. R. F. Lund, Energy Fuels, 2011, 25, 4745–4755 CrossRef CAS; (c) X. Fu, J. Dai, X. Guo, J. Tang, L. Zhu and C. Hu, Green Chem., 2017, 19, 3334–3343 RSC.
  6. (a) R.-J. van Putten, J. C. van der Waal, E. de Jong, C. B. Rasrendra, H. J. Heeres and J. G. de Vries, Chem. Rev., 2013, 113, 1499–1597 CrossRef CAS PubMed; (b) S. K. R. Patil, J. Heltzel and C. R. F. Lund, Energy Fuels, 2012, 26, 5281–5293 CrossRef CAS; (c) T. M. C. Hoang, E. R. H. van Eck, W. P. Bula, J. G. E. Gardeniers, L. Lefferts and K. Seshan, Green Chem., 2015, 17, 959–972 RSC; (d) I. van Zandvoort, Y. H. Wang, C. B. Rasrendra, E. R. H. van Eck, P. C. A. Bruijnincx, H. J. Heeres and B. M. Weckhuysen, ChemSusChem, 2013, 6, 1745–1758 CrossRef CAS; (e) G. Tsilomelekis, M. J. Orella, Z. X. Lin, Z. W. Cheng, W. Q. Zheng, V. Nikolakis and D. G. Vlachos, Green Chem., 2016, 18, 1983–1993 RSC.
  7. (a) G. W. Huber, J. N. Chheda, C. J. Barrett and J. A. Dumesic, Science, 2005, 308, 1446–1450 CrossRef CAS PubMed; (b) Y. Jing, Q. Xia, X. Liu and Y. Wang, ChemSusChem, 2017, 10, 4817–4823 CrossRef CAS; (c) Y. Jing, Q. Xia, J. Xie, X. Liu, Y. Guo, J. Zou and Y. Wang, ACS Catal., 2018, 8, 3280–3285 CrossRef CAS; (d) S. Liu, S. Dutta, W. Zheng, N. S. Gould, Z. Cheng, B. Xu, B. Saha and D. G. Vlachos, ChemSusChem, 2017, 10, 3225–3234 CrossRef CAS PubMed; (e) J. Xu, N. Li, X. Yang, G. Li, A. Wang, Y. Cong, X. Wang and T. Zhang, ACS Catal., 2017, 7, 5880–5886 CrossRef; (f) S. Dutta and B. Saha, ACS Catal., 2017, 7, 5491–5499 CrossRef CAS; (g) Q. N. Xia, Q. Cuan, X. H. Liu, X. Q. Gong, G. Z. Lu and Y. Q. Wang, Angew. Chem., Int. Ed., 2014, 53, 9755–9760 CrossRef CAS PubMed; (h) Y. Jing, Y. Xin, Y. Guo, X. Liu and Y. Wang, Chin. J. Catal., 2019, 40, 1168–1177 CrossRef CAS; (i) Y. Xin, Y. Jing, L. Dong, X. Liu, Y. Guo and Y. Wang, Chem. Commun., 2019, 55, 9391–9394 RSC.
  8. (a) M. Chia, T. J. Schwartz, B. H. Shanks and J. A. Dumesic, Green Chem., 2012, 14, 1850–1853 RSC; (b) B. H. Shanks and P. L. Keeling, Green Chem., 2017, 19, 3177–3185 RSC; (c) K. Tarade, S. Shinde, S. Sakate and C. Rode, Catal. Commun., 2019, 124, 81–85 CrossRef CAS.
  9. (a) X. Li, Y. Zhang, Q. Xia, X. Liu, K. Peng, S. Yang and Y. Wang, Ind. Eng. Chem. Res., 2018, 57, 3545–3553 CrossRef CAS; (b) Z. Jiang, J. Yi, J. Li, T. He and C. Hu, ChemSusChem, 2015, 8, 1901–1907 CrossRef CAS PubMed.
  10. (a) I. K. M. Yu and D. C. W. Tsang, Bioresour. Technol., 2017, 238, 716–732 CrossRef CAS PubMed; (b) Z. M. Xue, M. G. Ma, Z. H. Li and T. C. Mu, RSC Adv., 2016, 6, 98874–98892 RSC.
  11. X. Li, K. Peng, X. Liu, Q. Xia and Y. Wang, ChemCatChem, 2017, 9, 2739–2746 CrossRef CAS.
  12. (a) Y. J. Pagan-Torres, T. F. Wang, J. M. R. Gallo, B. H. Shanks and J. A. Dumesic, ACS Catal., 2012, 2, 930–934 CrossRef CAS; (b) L.-K. Ren, L.-F. Zhu, T. Qi, J.-Q. Tang, H.-Q. Yang and C.-W. Hu, ACS Catal., 2017, 7, 2199–2212 CrossRef CAS; (c) C. Luo, S. Wang and H. Liu, Angew. Chem., Int. Ed., 2007, 46, 7636–7639 CrossRef CAS PubMed.


Electronic supplementary information (ESI) available. See DOI: 10.1039/c9gc02987g
These authors contributed equally.

This journal is © The Royal Society of Chemistry 2019