Synthesis of biobased succinimide from glutamic acid via silver-catalyzed decarboxylation

Abstract

Glutamic acid was transformed into succinimide in a two step procedure involving a dehydration in water to pyroglutamic acid followed by an oxidative decarboxylation using a silver catalyst.

---

With the accelerated depletion of oil resources, the chemical industry is in a transition from a petrochemical based industry toward a biobased one. Large-scale production of biobased chemicals is becoming a reality because biomass is the most abundant renewable carbon resource, including agricultural and forest residues, municipal solid and liquid waste.^1–3 Amino acids, which are on account of a high content of low-value products from biofuel production such as sugar beet vinasse, soybean mean meal, corn starch etc., become interesting starting materials for the synthesis of bio-based chemicals. For example, the production of glutamic acid in 2020 is expected to reach 20 million tons in the world.⁴ With amino, carboxyl functional groups and small backbone, amino acids are good industrial intermediates. In particular, it does not require re-introducing the nitrogen atom compared with traditional petrochemical methods in the production of nitrogenous organic compounds.³

On one hand, glutamic acid, with an amino and two carboxyl groups, is an economically viable platform molecular that can be utilized to produce numerous nitrogen containing chemicals, such as acrylonitrile,⁴ aminobutyricacid (GABA),⁵ N-methylpyrrolidone (NMP),⁶ N-vinylpyrrolidone (NVP),⁶ succinonitrile,⁷ 3-cyanopropanoic acid⁸ and adiponitrile.⁹ On the other hand, because of its versatile functions and benign properties, succinimide is considered as one of the most promising reproducible molecules that can be converted into organic intermediates for the production of pesticides and pharmaceuticals such as N-chlorosuccinimide (NCS), N-bromosuccinimide (NBS) and N-iodosuccinimide (NIS). Furthermore, succinimide has notable commercial applications in silver plating processes and assay techniques. Conventionally, succinic acid and nitrogen sources, including ammonia and urea, are needed for succinimide synthesis (Scheme 1).¹⁰ Therefore, the development of a highly efficient catalytic system, which can be applied in mild condition for the production of succinimide from glutamic acid without additional nitrogen sources is imminent.

Scheme 1 Comparison chart of succinimide synthesis from petroleum and biomass.

Herein we report the synthesis of succinimide, without the use of additional nitrogen, by using glutamic acid as feedstock in a two-step procedure (Scheme 1). The procedure involves the heating dehydration cyclization of glutamic acid to pyroglutamic acid,¹¹ followed by oxidative decarboxylation using silver as catalyst and persulfate as oxidant. Using a catalytic amount of silver, this reaction achieves a sustainable industrial route to produce succinimide from biomass feedstock in mild condition and much higher yield by simple steps compared with some previous methods.

In our experimental program, the second step is critical. The method using iodosobenzene to oxidative decarboxylation has been reported, but the yield was low and the reaction time was too long to be applied in industry.¹² Recently, Chaozhong Li et al. reported the use of silver salts as catalyst for decarboxylative fluorination¹³ and chlorination¹⁴ of various aliphatic carboxylic acids in high yields under mild condition. Moreover, under the above condition, Chaozhong Li et al. realized the decarboxylation alkynylation with K₂S₂O₈ as the oxidant in aqueous solution using AgNO₃ as catalyst.¹⁵ In addition, Lu Lin et al. reported oxidative decarboxylation of levulinic acid to 2-butanone by AgNO₃ with persulfate.¹⁶ These approaches are considered to be commercially viable, high efficient and chemselective methods under mild condition for decarboxylation.

Inspired by the above works, we initially investigated the influence of the amount and types of persulfate on the yields. The initial experiments were performed at room temperature in water in the presence of 1 mol% of the AgNO₃ catalyst. After 10 h, for 1 equivalent of K₂S₂O₈, we observed that the succinimide yield was 36% (Table 1, entry 1). However, when the amount of K₂S₂O₈ increased to 2.2 eq., the yield increased to 97% (Table 1, entry 2). It was foreseeable that without the catalyst, the yield decreased to 7% (Table 1, entry 7).

Table 1 Optimisation of the oxidative decarboxylation of pyroglutamic acids using different kinds and amount of the oxidizing reagents^a

Entry	Cat. (mol%)	Oxidant (eq.)	Time/h	Conv.^b/%	Yield^b/%
a Reaction conditions: pyroglutamic acid (130 mg, 1 mmol), AgNO3(1.7 mg, 0.01 mmol), water (3 mL), persulfate (2.2 mmol), room temperature.b Determined by HPLC, yield of isolated product is in parenthesis.
1	AgNO3 (1%)	K2S2O8 (1.0)	10	71	36
2	AgNO3 (1%)	K2S2O8 (2.2)	10	100	97
3	AgNO3 (1%)	K2S2O8 (2.2)	2	100	71
4	AgNO3 (1%)	Na2S2O8 (2.2)	2	100	98
5	AgNO3 (1%)	(NH4)2S2O8 (2.2)	2	100	>99 (96)
6	AgNO3 (1%)	0	10	27	0
7	0	K2S2O8 (2.2)	10	19	7
8	0	(NH4)2S2O8 (2.2)	10	18	6

---

Such experimental results suggested the catalytic mechanism of silver as shown in Scheme 2. Firstly, one mole of Ag(I) ions are oxidized to Ag(II) ions by 0.5 mole of the persulfate ions, the generated Ag(II) ions convert free carboxyl group in the pyroglutamic acid into a carboxyl radical, and Ag(II) ions return to Ag(I) ions. Next, the unstable carboxyl radical decarboxylates into carbon radical, which is affected by the adjacent nitrogen atom's higher electronegativity and oxidized to enamine by 0.5 mole of the persulfate ions. Then, by an addition of water, the unsaturated enamine is converted into hydramine. Finally, by another one mole of the persulfate ions, the hydramine is oxidated into an amide. Throughout the course of this reaction, the process to succinimide consumes a total of two mole of the persulfate ions.¹⁷ Silver, as a catalyst, plays a crucial role in the reaction: the reduction potential of Ag(II) is about 1.98 V and the reduction potential of S₂O₈²⁻ is about 2.01 V, at this point, silver is selected as a bridge connecting pyroglutamic acid with oxidant S₂O₈²⁻ and leads pyroglutamic acid to a radical decarboxylation way which requires only mild conditions based on the literatures where oxidation of carboxylate anions by Ag(II) complexes was mentioned.

Scheme 2 The proposed catalytic mechanism.

Next, we reduced the reaction time to 2 h with all other conditions unchanged, the yield reduced to 71% (Table 1, entry 3). However, when the oxidant was changed to Na₂S₂O₈, the yield increased to 98% and with the oxidant of (NH₄)₂S₂O₈, it showed the highest yield more than 99% (Table 1, entry 4 and 5). The results demonstrate the effect of different persulfates on yields and the difference may be due to their different solubility. Under the conditions described in the experimental section (water 3 mL, persulfate 2.2 mmol), the concentration of persulfate theoretically was about 0.7 mol L⁻¹. However, at room temperature, the concentration of persulfate was less than 0.2 mol L⁻¹. As a result, K₂S₂O₈ can not be completely dissolved by calculation so that the ratio theoretically required of pyroglutamic acid and persulfate can not be achieved. Whereas, the solubility of Na₂S₂O₈ or (NH₄)₂S₂O₈ can be far from meeting the ratio to obtain a high yield.¹⁸

With the optimized oxidant (NH₄)₂S₂O₈, we then explored the effect of different silver salts on the yield and the results were summarized in Table 2. To investigate the catalytic efficiency of different silver salts, we used AgOAc (1 mol%), Ag₂SO₄ (0.5 mol%), Ag₂CO₃ (0.5 mol%), Ag₃PO₄ (0.3 mol%) and AgCl (1 mol%) instead of AgNO₃ (1 mol%) with all other conditions unchanged. When the Ag(I) ion was added with an amount of 1 mol%, most of the silver catalyst efficiency remained at a high level (Table 2, entries 1–4) except AgCl. The yield of AgCl was markedly reduced to 34% (Table 2, entry 5). Based on these results, taking into account the major differenence of these catalysts on solubility, we predicted that the solubility of the catalyst would largely affect the catalytic efficiency when the amount of dissolved catalyst was insufficient to achieve the required amount of catalytic theory. After calculation, based on the amount of catalyst added and their respective K_sp,¹⁹ the actual concentration of Ag(I) from AgCl dissolved in the aqueous solution was only 1.34 × 10⁻⁵ mmol mL⁻¹. For Ag₂CO₃ (0.5 mol%) and Ag₃PO₄ (0.3 mol%), whose yields remained at more than 99%, the actual amount of water-soluble salt was only about 2.96 × 10⁻⁴ mmol mL⁻¹ and 1.43 × 10⁻⁴ mmol mL⁻¹. Seen in this light, the amount of catalyst required in the process of oxidative decarboxylation using silver as catalyst and persulfate as oxidant is minimal (high TON up to 10⁴), which is crucial for practical applications.

Table 2 The effect of different silver salts on the yield of succinimide^a

Entry	Cat. (mol%)	Oxidant (eq.)	Time/h	Conv.^b/%	Yield^b/%
a Reaction conditions: apyroglutamic acid (130 mg, 1 mmol), Ag(I) ion (0.01 mmol), water (3 mL), (NH4)2S2O8 (0.50 g, 2.2 mmol), room temperature.b Determined by HPLC, yield of isolated product is in parenthesis.c AgOAc (1.7 mg, 0.01 mmol).d Ag2SO4 (1.7 mg, 0.005 mmol).e Ag2CO3 (1.4 mg, 0.005 mmol).f Ag3PO4 (1.4 mg, 0.003 mmol).g AgCl (1.5 mg, 0.01 mmol).
1	AgOAc (1%)^c	(NH4)2S2O8 (2.2)	2	100	>99 (96)
2	Ag2SO4 (0.5%)^d	(NH4)2S2O8 (2.2)	2	100	96 (92)
3	Ag2CO3 (0.5%)^e	(NH4)2S2O8 (2.2)	2	100	99 (93)
4	Ag3PO4 (0.3%)^f	(NH4)2S2O8 (2.2)	2	100	>99 (94)
5	AgCl (1%)^g	(NH4)2S2O8 (2.2)	2	48	34

---

The above silver-catalyzed oxidative decarboxylation was then extended to other transition metal such as copper and nickel. Nevertheless, these compounds, such as Cu(NO₃)₂ (1 mol%), CuSO₄ (1 mol%), Cu(OAc)₂ (1 mol%), Ni(NO₃)₂ (1 mol%) and Ni(OAc)₂ (1 mol%), did not exhibit good catalytic effect (Table 3, entries 1–5), even though they have a good solubility in water without considering the factor that the dissolved amount is too small to exert its effect. In addition, we selected Al(NO₃)₃ as a typical example of short-period mental elements as a catalyst, but an unsatisfactory result was also presented with the yield of 3% (Table 3, entry 6). These results all support the involvement redox cycling reaction of silver mechanism in the oxidation by S₂O₈²⁻. We know that the reduction potential of S₂O₈²⁻ is about 2.01 V. At this point, silver was selected as a bridge connecting pyroglutamic acid and oxidant S₂O₈²⁻ thanks to the reduction potential of 1.98 V to promote the oxidative decarboxylation based on the literatures where oxidation of carboxylate anions by Ag(II) complexes was mentioned.²⁰ Until now, we have chosen silver as the best catalyst for the oxidative pyroglutamic of pyroglutamic acid.

Table 3 The effect of different metals on the yield of succinimide^a

Entry	Cat. (mol%)	Oxidant (eq.)	Time/h	Conv.^b/%	Yield^b/%
a Reaction conditions: apyroglutamic acid (130 mg, 1 mmol), Ag(I) ion (0.01 mmol), water (3 mL), (NH4)2S2O8 (0.50 g, 2.2 mmol), room temperature.b Determined by HPLC.c Cu(NO3)2 (1.9 mg, 0.01 mmol).d CuSO4·5H2O (2.5 mg, 0.01 mmol).e Cu(OAc)2 (1.8 mg, 0.01 mmol).f Ni(NO3)2·6H2O (2.9 mg, 0.01 mmol).g Ni(OAc)2·4H2O (2.5 mg, 0.01 mmol).h Al(NO3)3·9H2O (3.7 mg, 0.01 mmol).
1	Cu(NO3)2 (1%)^c	(NH4)2S2O8 (2.2)	2	39	6
2	CuSO4 (1%)^d	(NH4)2S2O8 (2.2)	2	40	7
3	Cu(OAc)2 (1%)^e	(NH4)2S2O8 (2.2)	2	36	6
4	Ni(NO3)2 (1%)^f	(NH4)2S2O8 (2.2)	2	38	5
5	Ni(OAc)2 (1%)^g	(NH4)2S2O8 (2.2)	2	36	3
6	Al(NO3)3 (1%)^h	(NH4)2S2O8 (2.2)	2	38	3

---

Conclusions

In summary, we reported a sustainable succinimide preparation method without the use of additional nitrogen. Using silver as the catalyst, we achieved the conversion from a biobased chemical (glumatic acid) to target product (succinimide) in excellent yields under mild conditions. We observed the yield up to 96% in aqueous solution at room temperature after only 2 h with the minimal amount of catalyst. At the same time, this process was simple and green with only two steps and water as the solvent instead of some organic solvents which were toxic or expensive. In this way, it opened a green way for succinimide production from biomass and pushed the skylight to see the broad application prospects of biomass resources.

Experimental section

Preparation of pyroglutamic acid from glutamic acid

In a typical experiment, a beaker was filled with glutamic acid (100 g) and placed in an oil bath (145–150 °C). The dehydration reaction incubated for 45 minutes and the solution gradually turned brown. After the dehydration reaction, the solution was poured into boiling water (350 mL) and dissolved in the water. When the temperature droped to 40–50 °C, the solution was bleached by activated carbon twice and became colorless. The colorless solution was concentrated by directly heating until the volume was reduced to about half, then the solution was placed in a water both. When the volume of solution continued to decline to 1/3, the heat source was removed. The solution cooled and crystallized slowly. After 10–20 h, a kind of colorless prismatic crystal (pyroglutamic acid) was prepared.

Preparation of succinimide

In a typical experiment, a reaction tube was filled with pyroglutamic acid (0.13 g, 1 mmol), AgNO₃ (1.7 mg, 0.01 mmol), (NH₄)₂S₂O₈ (0.50 g, 2.2 mmol) and water (3 mL).Then, the solution was stirred at room temperature for 2 h. After sufficient reaction, the solvent was removed under reduced pressure. The product was analyzed by ¹HNMR spectrum and HPLC to determine the pyroglutamic acid yield (>99%). HPLC apparatus: Hitachi L2000 HPLC; column: Alltima C18(5μ × 250 mm × 4.6 mm); mobile phase: methanol–water = 10 : 90 (v/v, containing 0.1% phosphoric acid); flow rate: 1.0 mL min⁻¹; column temperature: 30 °C; detector: diode array; detection wavelength: 204 nm.

Under the same conditions described above, with AgNO₃ as the catalyst, the kinds and amount of the oxidizing agents were replaced, the reaction time (stirring time) and the present or absence of catalyst were described in Table 1. According to the method as above, the product was tested by HPLC. The conversion rate of pyroglutamic acid and yield of pyroglutamic acid were listed in Table 1.

Under the same conditions described above, with (NH₄)₂S₂O₈ as the oxidant, the catalyst was replaced as described in Tables 2 and 3. According to the method as above, the product was tested by HPLC. The conversion rate of pyroglutamic acid and yield of pyroglutamic acid were listed in Tables 2 and 3 respectively.

Acknowledgements

This work was supported by the 973 Program (2012CB215305, 2013CB228103), NSFC (21325208, 21172209), CAS (KJCX2-EW-J02), FRFCU (WK2060190025, WK2060190033), SRFDP (20123402130008) and Fok Ying Tung Education Foundation.

Notes and references

1. A. Corma, S. Iborra and A. Velty, Chem. Rev., 2007, 107, 2411.
2. D. M. Alonso, J. Q. Bond and J. A. Dumesic, Green Chem., 2010, 12, 1493.
3. C. O. Tuck, E. Pérez, I. T. Horvath, R. A. Sheldon and M. Poliakoff, Science, 2012, 337, 695.
4. J. L. Nôtre, E. L. Scott, M. C. R. Franssen and J. P. M. Sanders, Green Chem., 2011, 13, 807.
5. T. M. Lammens, D. D. Biase, M. C. R. Franssen, E. L. Scott and J. P. M. Sanders, Green Chem., 2009, 11, 1562.
6. T. M. Lammens, M. C. R. Franssen, E. L. Scott and J. P. M. Sanders, Green Chem., 2010, 12, 1430.
7. T. M. Lammens, J. L. Nôtre, M. C. R. Franssen, E. L. Scott and J. P. M. Sanders, ChemSusChem, 2011, 4, 785.
8. A. But, J. L. Nôtre, E. L. Scott, R. Wever and J. P. M. Sanders, ChemSusChem, 2012, 5, 1199.
9. J. J. Dai, Y. B. Huang, C. Fang, Q. X. Guo and Y. Fu, ChemSusChem, 2012, 5, 617.
10. H. T. Clarke and L. D. Behr, Org. Synth., 1936, 15, 76.
11. J. R. Lennox, S. C. Turner and H. Rapoport, J. Org. Chem., 2001, 66, 7078.
12. M. Ochiai, M. Inenaga, Y. Nagao, R. M. Moriarty, R. K. Vaid and M. P. Duncan, Tetrahedron Lett., 1988, 29, 6917.
13. F. Yin, Z. T. Wang, Z. D. Li and C. Z. Li, J. Am. Chem. Soc., 2012, 134, 10401.
14. Z. T. Wang, L. Zhu, F. Yin, Z. Q. Su, Z. D. Li and C. Z. Li, J. Am. Chem. Soc., 2012, 134, 4258.
15. X. S. Liu, Z. T. Wang, X. M. Cheng and C. Z. Li, J. Am. Chem. Soc., 2012, 134, 14330.
16. Y. Gong and L. Lin, Molecules, 2011, 16, 2714.
17. W. H. Huang, M. L. Wang and H. Yue, Synthesis, 2008, 9, 1342.
18. χ[K₂S₂O₈] = 0.196 mol L⁻¹ at 20 °C, χ[Na₂S₂O₈] = 2.31 mol L⁻¹ at 20 °C and χ[(NH₄)₂S₂O₈] = 3.51 mol L⁻¹ at 25 °C.
19. K_sp[Ag₂SO₄] = 6 × 10⁻⁵, K_sp[Ag₂CO₃] = 1.3 × 10⁻¹¹, K_sp[Ag₃PO₄] = 1.4 × 10⁻¹⁶ and K_sp[AgCl] = 1.8 × 10⁻¹⁰, L. G. Sillen and A. E. Martell in Lange's Handbook: Stability Constants of Metal-Ion Complexes, The Chemical Society, London, 1964, pp. 8-6 to 8-11.
20. P. Połczyński, R. Jurczakowski and W. Grochala, Chem. Commun., 2013, 49, 7480.

---