Application of group V polyoxometalate as an efficient base catalyst: a case study of decaniobate clusters

Abstract

The base catalytic activity of the decaniobate cluster (TMA)₆[Nb₁₀O₂₈]·6H₂O (TMA⁺ = tetramethylammonium cation) was studied theoretically and experimentally. Density functional theory calculations showed that the oxygen atoms in the cluster are highly negatively charged and suggested the possibility that the cluster can act as a base catalyst. We demonstrated for the first time that [Nb₁₀O₂₈]⁶⁻ actually exhibits base catalytic activity for aldol-type condensation reactions including Knoevenagel and Claisen–Schmidt condensation reactions. The catalytic reactions proceeded under ambient conditions, suggesting that [Nb₁₀O₂₈]⁶⁻ holds promise as a practical strong base catalyst.

---

1. Introduction

A polyoxometalate (POM),¹ [M_xO_y]ⁿ⁻, is an anionic metal-oxide cluster that consists of a given number of [MO₆] octahedral units by sharing edges or corners. POMs have been widely used as catalysts, especially for acid-catalyzed, oxidation, and photocatalytic reactions.^2–4 Recently, Mizuno and coworkers reported that [WO₄]²⁻ and [γ-HGeW₁₀O₃₆]⁷⁻ acted as base catalysts for CO₂ fixation^5–7 and Knoevenagel condensation reaction,^8,9 respectively. The base catalysis has been ascribed to the negativity of the oxygen atoms, mainly due to the high charge densities of the cluster.⁵ These studies have motivated us to develop stronger base catalysts based on POMs.

Our simple working hypothesis is that POM having more electronegative oxygen will exhibit stronger base catalysis. The −n/y value of [M_xO_y]ⁿ⁻ represents a negative charge averaged over the oxygen atoms without assuming electronic charge transfer from metal to oxygen and thus provides a lower limit of the negativity of the oxygen atom. On the assumption that the −n/y value gives a measure of the basicity of the POMs, group V POMs (M = V, Nb, Ta) are more promising than group VI POM (M = Mo, W) for base catalysts. For example, polyoxoniobates (PONbs)^10,11 such as hexaniobate ([Nb₆O₁₉]⁸⁻) and decaniobate ([Nb₁₀O₂₈]⁶⁻) cluster have more negative −n/y values (−0.42 and −0.21, respectively) than polyoxotungstates (POWs) such as [HGeW₁₀O₃₆]⁷⁻ (−0.19), [W₁₀O₃₂]⁴⁻ (−0.13) and [W₆O₁₉]²⁻ (−0.11); although, [WO₄]²⁻ has a much more negative value of −0.50.

In order to test the hypothesis, we studied in this work homogeneous base catalysis of PONb. As an initial target, we focused on [Nb₁₀O₂₈]⁶⁻ rather than [Nb₆O₁₉]⁸⁻ although the latter has more negative −n/y value. This is simply because the crystal structure of tetramethylammonium (TMA) salt, that can be dispersed in organic solvent, have been reported only for [Nb₁₀O₂₈]⁶⁻.¹² In this communication, we theoretically compared negative charges of oxygen atoms of [Nb₁₀O₂₈]⁶⁻ with those of POWs. Then, we applied (TMA)₆[Nb₁₀O₂₈]·6H₂O as homogeneous catalyst for aldol-type condensation reactions such as Knoevenagel and Claisen–Schmidt condensation reactions and revealed for the first time that it acted as a Brønsted base catalyst whose strength is comparable to superbases. This result will open up a new possibility of application of PONbs since their catalytic application has been limited to electrocatalysts¹³ and photocatalysts so far.^14–17

2. Results and discussion

The structure of [Nb₁₀O₂₈]⁶⁻ was optimized by density functional theory (DFT) calculations. The structure obtained by optimization (Table S1†)¹⁸ reproduced that determined experimentally by single-crystal X-ray diffraction (XRD)¹² except that bond lengths were elongated by <3% relative to the experimental values. The localized charge of the oxygen atoms obtained by natural bond orbital (NBO) analysis is shown in Fig. 1; this gives a measure of the basicity of the individual oxygen atoms. The NBO charges of the oxygen atoms are much more negative than −0.21, which is the lower limit of the negativity without assuming electronic charge transfer between Nb and O. This result indicates electronic charge transfer from Nb to O atoms. Fig. 1 suggests that the four edge-sharing oxygen atoms of [Nb₁₀O₂₈]⁶⁻ are the most active sites for base catalytic reaction.

Fig. 1 Optimized structure and NBO charges of oxygen atoms in [Nb₁₀O₂₈]⁶⁻ (D_2h).

The electronic charge of the most negative oxygen atom in [Nb₁₀O₂₈]⁶⁻ (−0.873) was compared with those of other typical polyoxotungstates such as [W₁₀O₃₂]⁴⁻, [W₆O₁₉]²⁻, and [WO₄]²⁻ (ref. 5) obtained from the same level of calculation (Fig. 2). The charge of the most negative oxygen atom in [Nb₁₀O₂₈]⁶⁻ (−0.873) was much more negative than those of [W₁₀O₃₂]⁴⁻ (−0.753) and [W₆O₁₉]²⁻ (−0.721), but was slightly less negative than that of [WO₄]²⁻ (−0.934). This comparison suggests that [Nb₁₀O₂₈]⁶⁻ will show base catalysis as in the case of POWs reported so far.^5–9

Fig. 2 NBO charges of the most negative oxygen atoms.

To test the base catalytic activity of [Nb₁₀O₂₈]⁶⁻ cluster dispersed in organic solvent, we synthesized the TMA salt, (TMA)₆[Nb₁₀O₂₈]·6H₂O, according to the reported method.¹² The synthesized product was characterized by powder XRD, negative-ion electrospray ionization mass spectrometry (ESI-MS), and elemental analysis. The powder XRD pattern of the product agreed well with that of (TMA)₆[Nb₁₀O₂₈]·6H₂O reported previously (Fig. 3).¹² The ESI-MS of the product in water–MeOH (1 : 1) exhibited a series of mass peaks assigned to [TMA_xH_(5−x)Nb₁₀O₂₈]⁻ (x = 2–5) (Fig. 4). The elemental analysis of C, H, N, and Nb agreed with the calculated values for [(CH₃)₄N]₆[Nb₁₀O₂₈]·6H₂O to an accuracy of <0.5% (calcd: C, 14.94; H, 4.36; N, 4.36; Nb, 48.2. Found: C, 14.75; H, 4.38; N, 4.27; Nb, 48.7). These results show that (TMA)₆[Nb₁₀O₂₈]·6H₂O was successfully synthesized.

Fig. 3 Powder XRD pattern of the product and simulated pattern from reported crystal structure.¹²

Fig. 4 Negative-ion ESI-MS of (TMA)₆[Nb₁₀O₂₈]·6H₂O.

We applied (TMA)₆[Nb₁₀O₂₈]·6H₂O as a homogeneous catalyst for the Knoevenagel condensation reaction. This is a fundamental coupling reaction to form carbon–carbon double bonds between active methylene compounds such as nitriles (donors) and carbonyl compounds (acceptors). The key step in the reaction is proton abstraction from donors by base catalysts. We studied whether (TMA)₆[Nb₁₀O₂₈]·6H₂O catalyzed the coupling reaction between benzaldehyde (1) and various nitriles having different pK_a values. The results of the catalytic reactions are summarized in Table 1. [Nb₁₀O₂₈]⁶⁻ cluster afforded in good yields coupling products with nitriles 2a and 2b (entries 1 and 2) whose pK_a values are 12.3 and 16.0 even at mild temperature T = 313 K (entries 1-1 and 2-1). Negative-ion ESI-MS of the catalyst after the reaction (entry 2-1) exhibited a series of mass peaks of [TMA_xH_(5−x)Nb₁₀O₂₈]⁻ (x = 1–3) (Fig. S1†),¹⁸ suggesting that the catalysts did not decompose. Turnover frequencies for these reactions (entries 1-2 and 2-2) were calculated to be 66 and 28 h⁻¹ at 343 K, respectively, from the kinetic data (Fig. S2†).¹⁸ [Nb₁₀O₂₈]⁶⁻ cluster showed base catalytic activity for nitriles 2c (pK_a = 19.4, entry 3), 2d (pK_a = 19.5, entry 4), 2e (pK_a = 21.9, entry 5) and 2f (pK_a = 23.8, entry 6), although the yields gradually decreased with an increase in the pK_a values of the nitriles. However, [Nb₁₀O₂₈]⁶⁻ cluster did not show any activity for the reaction of 2g (pK_a = 32.5, entry 7).

Table 1 Knoevenagel condensation catalyzed by (TMA)₆[Nb₁₀O₂₈]·6H₂O^a

Entry	Donor	pKa	Temp. (K)	Yield (%)
a Reaction conditions: catalyst (1 mol% with respect to 1), 1 (1.0 mmol), 2 (1.0 mmol), MeOH 1 mL, 343 K, 24 h.
1-1		12.3	313	61
1-2			343	88
2-1		16.0	313	77
2-2			343	88
3		19.4	343	61
4		19.5	343	39
5		21.9	343	35
6		23.8	343	8
7		32.5	343	0

---

Catalytic activity of (TMA)₆[Nb₁₀O₂₈]·6H₂O for Claisen–Schmidt condensation reaction between 1 and acetophenone (2h, pK_a = 24.7) was also studied. As shown in Scheme 1, [Nb₁₀O₂₈]⁶⁻ also acted as a homogeneous base catalyst for Claisen–Schmidt condensation.

Scheme 1 Claisen–Schmidt condensation catalyzed by (TMA)₆[Nb₁₀O₂₈]·6H₂O. ^aReaction conditions: catalyst (1 mol% with respect to 1), 1 (1.0 mmol), 2 (1.0 mmol), MeOH 1 mL, 343 K, 24 h.

Although the yields of the reactions with 2f (pK_a = 23.8) and 2h (pK_a = 24.7) were low compared to typical base catalysts such as hydroxide and alkoxide,^19,20 the fact that [Nb₁₀O₂₈]⁶⁻ cluster could activate less acidic nitriles indicated its strong basicity. To the best of our knowledge, [Nb₁₀O₂₈]⁶⁻ cluster is the first example of a strong base catalyst that can activate 2f and 2h among various POMs. The base strength of [Nb₁₀O₂₈]⁶⁻ is as high as 24.7 if we assume that the base strength of a catalyst is comparable to the pK_a values of methylene groups from which protons can be abstracted. This value is comparable to those of superbases, whose base strengths are defined to be larger than 26. However, those superbases are generally very sensitive to water and CO₂ and exhibit strong basicities only under rigorously controlled conditions.²¹ In contrast, [Nb₁₀O₂₈]⁶⁻ cluster showed base catalysis even when it was used in non-hydrated solvent under ambient conditions. This novel feature suggests that the PONbs hold promise as a practical strong base catalyst.

3. Conclusions

We have reported for the first time that decaniobate cluster [Nb₁₀O₂₈]⁶⁻ acts as a strong and homogeneous base catalyst for Knoevenagel condensation reaction of benzaldehyde and p-methoxyphenylacetonitrile (pK_a = 23.8) and Claisen–Schmidt condensation reaction of benzaldehyde and acetophenone (pK_a = 24.7). This high catalytic activity is associated with the large negative charge on the oxygen sites (−0.79 to −0.87) originating mainly from the large total negative charge on the cluster.

4. Experiment and theory

4.1. Density functional theory calculations

DFT calculations were conducted using the Gaussian 09 program.²² The structure of [Nb₁₀O₂₈]⁶⁻ was optimized at the level of B3LYP/6-31G++(d). The absence of imaginary frequencies was confirmed from vibrational analysis. Natural bond orbital (NBO) analysis was conducted to obtain localized charge on the oxygen atoms.

4.2. Chemicals

Niobic acid was supplied from CBMM (Nb₂O₅·nH₂O, water 20% w/w). Tetramethylammonium hydroxide pentahydrate (TCI, 97%), benzaldehyde (Wako, 98%), 4-nitrobenzyl cyanide (TCI, 98%), 4-cyanophenylacetonitrile (Aldrich, 97%), 3-bromobenzyl cyanide (TCI, 98%), m-chlorobenzyl cyanide (Wako, 98%), phenylacetonitrile (Wako, 98%), 4-methoxyphenylacetonitrile (Aldrich, 97%), propionitrile (Wako, 98%), and acetophenone (Wako, 98.5%) were used without further purification. Deionized water (Milli-Q, >18 MΩ cm) was used in all experiments.

4.3. Synthesis

(TMA)₆[Nb₁₀O₂₈]·6H₂O was synthesized according to the literature procedure.¹² In an autoclave, niobic acid (Nb₂O₅·nH₂O, water 20% w/w, 3.00 g, 9.02 mmol) was mixed with 30 mL of EtOH solution of tetramethylammonium hydroxide (3.60 g, 19.9 mmol). The autoclave was heated at 120 °C for 18 h. After the reaction, the orange solution was removed and a white precipitate was collected by centrifugation. The precipitate was washed with EtOH (150 mL) and acetone (150 mL) and dried well. The crude product was dissolved into 50 mL of MeOH and centrifuged to remove unreacted niobic acid. The solution was concentrated into 15 mL by evaporation and separated into three vials. Recrystallization was conducted by adding 20 mL of acetone slowly to each solution to form a second layer. Typical yield was 2.01 g (1.04 mmol, 58% based on Nb).

4.4. Characterization

The powder XRD pattern was collected with a Rigaku SmartLab 3 using nickel-filtered Cu-Kα radiation. Negative-ion ESI-MS was measured with a JEOL JMS-T100LP AccuTOF LC-plus. (TMA)₆[Nb₁₀O₂₈]·6H₂O was dissolved into water–MeOH (1 : 1, 1 mg mL⁻¹) and electrosprayed at a bias voltage of −2 kV. The mass spectrum was calibrated by using Cs_nI_(n+1)⁻ clusters generated from CsI solution as a reference. Elemental analysis for C, H, and N was conducted with an Elementar vario MICRO cube. ICP-AES analysis for Nb was performed with a Thermo Scientific iCAP DUO-6300 at a wavelength of 309.4 nm. GC analysis was conducted with a Shimadzu GC-2014 with a Restek Rtx-1 (internal diameter = 0.53 mm, length = 30 m) capillary column, and GC-MS analysis was conducted with a Shimadzu GCMS-QP2010 Ultra with a Restek Rtx-1 (internal diameter = 0.32 mm, length = 30 m) capillary column at an ionization voltage of 70 eV.

4.5. Catalytic test

Benzaldehyde (1, 1.0 mmol), nitrile or acetophenone (2, 1.0 mmol), and biphenyl (internal standard, ca. 0.1 mmol) were dissolved into MeOH, and (TMA)₆[Nb₁₀O₂₈]·6H₂O (10 μmol, 1 mol%) was added to start the reaction. The reaction mixture was stirred for 24 h and then analysed by GC and GC-MS.

Acknowledgements

We thank Dr Naoko Nonose (National Institute of Advanced Industrial Science and Technology; AIST) for preliminary ICP analysis and Dr Noritaka Mizuno, Dr Kazuya Yamaguchi, and Dr Kosuke Suzuki (The University of Tokyo) for fruitful discussions. This research was financially supported by the Elements Strategy Initiative for Catalysis and Batteries (ESICB) and by a Grant-in-Aid for Scientific Research (No. 24750210) from the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan.

Notes and references

1. D. Long, R. Tsunashima and L. Cronin, Angew. Chem., Int. Ed., 2010, 49, 1736.
2. C. L. Hill and C. M. Prosser-McCartha, Coord. Chem. Rev., 1995, 143, 407.
3. I. V. Kozhevnikov, Chem. Rev., 1998, 98, 171.
4. N. Mizuno and M. Misono, Chem. Rev., 1998, 98, 199.
5. T. Kimura, K. Kamata and N. Mizuno, Angew. Chem., Int. Ed., 2012, 51, 6700.
6. T. Kimura, H. Sunaba, K. Kamata and N. Mizuno, Inorg. Chem., 2012, 51, 13001.
7. K. Kamata, T. Kimura, H. Sunaba and N. Mizuno, Catal. Today, 2014, 226, 160.
8. K. Sugahara, T. Kimura, K. Kamata, K. Yamaguchi and N. Mizuno, Chem. Commun., 2012, 48, 8422.
9. K. Sugahara, N. Satake, K. Kamata, T. Nakajima and N. Mizuno, Angew. Chem., Int. Ed., 2014, 53, 13248.
10. M. Nyman, Dalton Trans., 2011, 40, 8049.
11. H. Wu, Z. Zhang, Y. Li, X. Wang and E. Wang, CrystEngComm, 2015, 17, 6261.
12. C. A. Ohlin, E. M. Villa and W. H. Casey, Inorg. Chim. Acta, 2009, 362, 1391.
13. Y. Ye, C. Chen, H. Feng, J. Zhou, J. Ma, J. Chen, J. Yuan, L. Kong and Z. Qian, Open J. Inorg. Chem., 2013, 3, 59.
14. L. Shen, Y. Xu, Y. Gao, F. Cui and C. Hu, J. Mol. Struct., 2009, 934, 37.
15. Z. Zhang, Q. Lin, D. Kurunthu, T. Wu, F. Zuo, S. Zheng, C. J. Bardeen, X. Bu and P. Feng, J. Am. Chem. Soc., 2011, 133, 6934.
16. Z. Wang, H. Tan, W. Chen, Y. Li and E. Wang, Dalton Trans., 2012, 41, 9882.
17. P. Huang, C. Qin, Z. Su, Y. Xing, X. Wang, K. Shao, Y. Lan and E. Wang, J. Am. Chem. Soc., 2012, 134, 14004.
18. See ESI†.
19. B. A. Bhat, K. L. Dhar, S. C. Puri, A. K. Saxena, M. Shanmugavel and G. N. Qazi, Bioorg. Med. Chem. Lett., 2005, 15, 3177.
20. A. Ramírez-Rodríguez, J. M. Méndez, C. C. Jiménez, F. León and A. Vazquez, Synthesis, 2012, 44, 3321.
21. L. Chen, J. Zhao, S. Yin and C. Au, RSC Adv., 2013, 3, 3799.
22. M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, et al., Gaussian 09, Revision D.01, Gaussian, Inc., Wallingford CT, 2013.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra00338a

---