Decarboxylation of a Wide Range of Amino Acids with Electrogenerated Hypobromite

Article abstract of DOI:10.1002/ejoc.201403112

Bromide-assisted electrochemical decarboxylation efficiently produces valuable nitriles in high yields from a wide range of naturally occurring amino acids in a single step. Bromide salts are used as both redox mediators and supporting electrolytes in a simple one-compartment setup. As demonstrated for lysine, the selectivity of the decarboxylation can be tuned towards nitriles, amines or amides. An electrochemical system is developed that allows the selective decarboxylation of a wide range of amino acids. Valuable nitriles are obtained in high yields in a single step by using bromide salts as both redox mediators and supporting electrolytes. The product selectivity of lysine can be tuned towards nitriles, amines, or amides.

Full text of DOI:10.1002/ejoc.201403112

SHORT COMMUNICATION
DOI: 10.1002/ejoc.201403112
Decarboxylation of a Wide Range of Amino Acids with Electrogenerated
Hypobromite
Roman Matthessen,[a] Laurens Claes,[a] Jan Fransaer,[b] Koen Binnemans,[c] and
Dirk E. De Vos*[a]
Keywords: Electrochemistry / Redox chemistry / Amino acids / Decarboxylation / Nitriles
Bromide-assisted electrochemical decarboxylation efficiently
produces valuable nitriles in high yields from a wide range
of naturally occurring amino acids in a single step. Bromide
salts are used as both redox mediators and supporting elec-
trolytes in a simple one-compartment setup. As demon-
strated for lysine, the selectivity of the decarboxylation can
be tuned towards nitriles, amines or amides.
have been developed, but these have to cope with the high
substrate specificity of enzymes, each with specific and
strict conditions for optimal activity.[12–14] Whereas amino
acids obtained by protein hydrolysis can be separated easily
by electrodialysis into acidic, basic, and neutral fractions,[15]
further separation is very demanding. Therefore, a low-
waste decarboxylation method that is applicable to a mix-
ture of amino acids, as encountered in each fraction, would
be highly desirable.
Introduction
The chemical industry is in a transition from a petro-
chemical industry towards a more biobased industry.[1]
A
significant amount of the available biomass consists of pro-
tein-rich waste streams that originate mainly from the agro-
industry and from the increasing production of biofuels
such as bioethanol.[2] This protein-rich waste is often un-
suitable for human consumption and demands efficient val-
orization. As a source of nitrogen, amino acids can be inter-
esting starting materials for a variety of nitrogen-containing
bulk chemicals after selective removal of the carboxylic
group on the α-carbon atom.[3] This nitrogen recycling avo-
ids the repeated, energy-intensive functionalization of hy-
drocarbons with ammonia.[1]
Routes for chemical or enzymatic decarboxylation of
amino acids have been studied in the past.[4–14] Chemical
methods use stoichiometric oxidants such as sodium hypo-
halites,[4–6] N-bromosuccinimide,[7,10] 1-chlorobenzotri-
azole,[9] trichloroisocyanuric acid,[10] and 2-iodoxybenzoic
acid.[12] These oxidants supply the halonium species that
effect the oxidative decarboxylation on the α-carbon atom.
Depending on the pH of the system, a mixture of aldehydes
and nitriles can be formed in varying ratios.[8] The used oxi-
dants are sometimes toxic, may hamper product purifica-
tion, and generate significant amounts of waste. On the
other hand, greener, enzymatic decarboxylation methods
In this paper, we describe the bromide-assisted electrocat-
alytic decarboxylation of amino acids (Figure 1). Only lim-
ited information is available on this subject; most reports
focus on the electrochemical characteristics of nonmediated
amino acid oxidation at Pt,[16–18] Ni,[19] and Ag anodes,
without describing in detail the products formed.[20] Up to
now, no electrochemical system has been disclosed that
would allow a wide range of amino acids to be converted
in near-quantitative yields. By analogy with the aforemen-
tioned chemical methods, a halide-mediated electrochemi-
cal system appears very promising. Through in situ forma-
tion and regeneration of hypohalite species, the carboxylic
group on the α-carbon atom is selectively removed,[6] which
resembles a non-Kolbe electrolysis.[21] So far, only one such
case has been reported, for which glutamic acid was con-
verted into a 3-cyanopropionate salt in a three-step pro-
cedure via the methyl ester.[22] In this reported procedure,
an extra carbon atom is first incorporated, after which two
carbon atoms are removed; this makes the atom economy
inadequate. Furthermore, acceptable decarboxylation yields
can only be obtained upon working at 0 °C. Finally, the
proposed electrolysis conditions are not suited for various
other amino acids; in our hands extremely poor yields for
proline, lysine, and arginine were obtained. Herein, we show
that under well-defined conditions (solvent, salt, electrode
material, current density, acidity) it is possible to convert a
range of amino acids into the corresponding nitriles at
[a] Centre for Surface Chemistry and Catalysis, KU Leuven,
Kasteelpark Arenberg 23, 3001 Leuven, Belgium
E-mail: Dirk.DeVos@biw.kuleuven.be
http://www.biw.kuleuven.be/m2s/cok/
[b] Department of Metallurgy and Materials Engineering, KU
Leuven,
Kasteelpark Arenberg 44, 3001 Leuven, Belgium
[c] Department of Chemistry, KU Leuven,
Celestijnenlaan 200F, 3001 Leuven, Belgium
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejoc.201403112.
Eur. J. Org. Chem. 2014, 6649–6652
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
6649

R. Matthessen, L. Claes, J. Fransaer, K. Binnemans, D. E. De Vos
SHORT COMMUNICATION
room temperature, in a single step, with sound atom econ- Table 1. Electrochemical decarboxylation of phenylalanine (1a) to
benzyl cyanide (2a, entries 1–8) and of isoleucine (1b) to 2-methyl-
omy and with very good to excellent yields.
butyronitrile (2b, entries 9–15): influence of solvent, redox media-
tor, electrode material, and current density.[a]
Figure 1. Schematic representation of the bromide-mediated anodic
decarboxylation of amino acids.
Entry
Solvent
(CH3OH/H2O)
Salt
j[b]
Yield
[%]
[mAm–2]
1[c]
2[c]
3[c]
4[c]
0:1
1:1
4:1
13:1
1:0
13:1
13:1
13:1
13:1
13:1
13:1
13:1
13:1
13:1
13:1
NH4Br
NH4Br
NH4Br
NH4Br
NH4Br
NH4I
10
10
10
10
10
10
10
10
10
10
10
60
30
15
5
48
74
84
93
81
0
Results and Discussion
Amino acids with a neutral, nonpolar side chain such as
phenylalanine and isoleucine were chosen as model sub- 5[c]
strates to optimize the decarboxylation process. The choice
of solvent and redox mediator is of great importance for the
6[c]
7[c]
NH4Cl
NaBr
4
8[c]
76
32
86
78
38
64
83
85
reaction outcome, as shown in Table 1 for phenylalanine.
In CH3OH/H2O mixtures, the solubility of phenylalanine
increases with increasing water content, as is the case for
most amino acids; however the efficiency of phenylalanine
decarboxylation reaches an optimum at a CH3OH/H2O vol-
umetric ratio of 13:1 (Table 1, entries 1–5). One reason is
a decreased rate of anodic hypobromite formation in H2O
relative to that in CH3OH owing to competitive H2O oxi-
dation. Fortunately, oxidation of bromide, which is only a
two-electron oxidation, is faster than four-electron water
oxidation. The reaction gives almost no product if iodide or
chloride salts are used (Table 1, entries 6 and 7): hypoiodite
oxidants are too weak, whereas hypochlorite is almost not
9[d,e]
10[d]
11[d,f]
12[d]
13[d]
14[d]
15[d]
NH4Br
NH4Br
NH4Br
NH4Br
NH4Br
NH4Br
NH4Br
[a] Anode: Pt; cathode: Ni; [amino acid]: 0.15 molL–1; [salt]/[amino
acid]: 1; amount of charge: 6 Fmol–1. [b] Current density.
[c] Amino acid: phenylalanine. [d] Amino acid: isoleucine. [e] Con-
ditions as in a, but anode: graphite. [f] Conditions as in a, but cath-
ode: Pt.
formed under the applied conditions as a result of competi- hypobromite reduction peaks were higher at platinum than
tive oxidation of H2O. The latter also demonstrates the dif- at graphite (Figure S2). At the cathode, on the other hand,
ficulty of a direct anodic process. The cation type also plays platinum could easily be replaced by cheaper nickel, which
a role in the decarboxylation reaction, and ammonium gives even slightly better yields (Table 1, entries 10 and 11).
bromide gave significantly better results than sodium brom- Another important parameter is the current density, with
ide (Table 1, entries 4 and 8). The more acidic nature of an optimal performance at 10 mAcm–2 (Table 1, entries 10,
ammonium salts relative to that of sodium salts favors the 12–15). At values higher than 15 mAcm–2, a significant
anodic oxidation of bromide by suppressing the oxidation amount of charge is lost to undesired reactions such as
of water. Moreover, increased alkalinity would promote al- solvent oxidation and reduction.
dehyde formation,[8] and the abstraction of the α-hydrogen
On the basis of the results of Table 1, the electrochemical
atom by the electrophilic bromine species occurs more read- decarboxylation was tested for a wide range of naturally
ily under slightly acidic conditions.[6] As the decarboxyl- occurring amino acids (Table 2). For amino acid reactant 1
ation of amino acids to nitriles is a four-electron oxidation with general formula RCH(NH2)COOH, the typical prod-
(Figure 1), the theoretically required charge is 4 Fmol–1. ucts were nitriles RCN 2, or in certain cases amines
Passing various amounts of charge through the cell gives an RCH2NH2 3 or amides RC(=O)NH2 4. Under the same
optimal compromise between yield and current efficiency at optimal conditions as those found for phenylalanine (1a)
6 Fmol–1, which results in acceptable faradaic yields (Fig- and isoleucine (1b), very good yields of the nitrile product
ure S1 in the Supporting Information).
were obtained for alanine (1c), valine (1d), and leucine (1e;
The choice of electrode material has a large influence on Table 2, entries 1–5). The reactions of serine (1f) and threo-
the reaction efficiency, as illustrated in Table 1 for the de- nine (1g) showed partial conversion of the nitrile product
carboxylation of isoleucine. Bromide oxidation is much to the corresponding amine at a nickel cathode. However,
faster on a platinum surface than on a graphite surface, and the selectivity could be shifted completely towards the
this leads to a significant difference in product yield nitrile as the sole product by using a platinum cathode
(Table 1, entries 9 and 10). This effect was proven with a (Table 2, entries 6–9). For both isoleucine and threonine, no
cyclic voltammetry experiment; the bromide oxidation and racemization occurred during the reaction, and the S and
6650
www.eurjoc.org
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2014, 6649–6652

Electrochemical Decarboxylation of Amino Acids
Table 2. Electrochemical decarboxylation of amino acids 1 to corre- R configurations, respectively, were preserved in the side
sponding nitriles 2, amines 3, or amides 4.[a]
chain.
Glutamine (1h) reacted smoothly to 3-cyanopropion-
amide (2h) in both a 13:1 mixture of CH3OH/H2O and in
pure water as solvent (Table 2, entries 10 and 11). Aspara-
gine (1i) on the other hand is insoluble in methanol and
precipitated immediately, whereas in pure water it was con-
verted into acetamide, obtained by decomposition of the
nitrile or one of the intermediates (Table 2, entry 12). The
acid fraction, comprising glutamic acid and aspartic acid,
is poorly soluble in both methanol and water. The corre-
sponding sodium salts, 1j and 1k, easily dissolve in water,
which led to high conversions to the nitrile product
(Table 2, entries 13–16). For 2k, minor decomposition into
sodium acetate was observed, a process which is similar to
the aforementioned acetamide formation. Therefore, yields
of 2k were slightly lower than those for more stable 2j. If
the amino acid solubility was much higher in water than
that in methanol, as for 1i, 1j, and 1k, higher yields were
obtained in water despite slower hypobromite formation.
Sodium bromide appears to be the most suited electrolyte
if using water as the solvent. If the solubility in both meth-
anol and water is relatively low and/or comparable, the ef-
fect of the bromide oxidation rate is much more important,
and better results are obtained in methanol-enriched solu-
tions. Protein hydrolysis, as would be performed on waste
streams, will convert glutamine and asparagine into glu-
tamic and aspartic acid, respectively. This way, they make
up a large part of the total protein fraction, for example,
more than 60 wt.-% of the protein byproduct of bioethanol
production from sugarbeets.[2] As shown in Table 2, en-
try 16, the electrocatalytic method allows this easily isolated
acid fraction to be valorized efficiently.
Proline (1l) is another amino acid that is highly soluble in
water. Upon subjecting it to the oxidative decarboxylation,
remarkably 2-pyrrolidone was obtained in 91% yield, likely
through hydration of an intermediate iminium species
(Table 2, entry 17). 2-Pyrrolidone is a high-value precursor
for compounds such as the monomer N-vinylpyrrolidone or
the solvent N-methylpyrrolidone.[2] Decreasing the bromide
concentration to 0.25 equiv. relative to that of the amino
acid is useful to prevent unwanted ring bromination. Lysine
and arginine, on the other hand, are amino acids with a
basic side chain and can only be used if neutralized with
one equivalent of HCl, which results again in a slightly
acidic environment. Lysine monohydrochloride (1m) has a
high solubility in water, whereas it is poorly soluble in meth-
anol. The acidic nature of the ammonium group on the
alkyl side chain facilitates reduction of the nitrile product.
Depending on the cathode material the selectivity can be
shifted towards nitrile 2m or amine 3m, for which nickel
[a] Anode: Pt; cathode: Ni; [amino acid]: 0.15 molL–1; [salt]/[amino
acid]: 1; current density: 10 mAcm–2; amount of charge: 6 Fmol–1;
reaction time: 2 h 40 min. [b] Solvent: CH3OH/H2O = 13:1; salt:
NH4Br. [c] As in [a], but cathode: Pt. [d] Solvent: H2O; salt: NaBr.
[e] As in [a], but reaction performed for 15 h at ΔV = 2.5 V. [f] As
in [a], but [salt]/[amino acid]: 0.25. [g] Extension with 2nd electroly-
sis step under same conditions after 2-fold dilution of reaction mix-
ture with 2-propanol.
promotes nitrile hydrogenation (Table 2, entries 18 and 19).
The amine selectivity is further increased in a second elec-
trolysis step by adding 2-propanol (Table 2, entry 20). If the
current supply is maintained after full conversion of the
amino acid, 2m and 3m are almost quantitatively trans-
formed into 5-aminopentanamide monohydrochloride 4m
(Table 2, entry 21). This amide is a possible precursor to a
Eur. J. Org. Chem. 2014, 6649–6652
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
6651

R. Matthessen, L. Claes, J. Fransaer, K. Binnemans, D. E. De Vos
SHORT COMMUNICATION
lactam and, hence, could be used for production of nylon-
type polymers. The stronger basicity of guanidine relative
to that of an amine group makes arginine monohydrochlor-
ide (1n) less acidic than 1m, which results in the formation
of nitrile product 2m at a Ni cathode in excellent yields
(Table 2, entries 22–24). Owing to its solubility properties,
the reaction rate of 1n is higher in the 13:1 CH3OH/H2O
mixture, although excellent yields can be obtained in pure
water by supplying larger amounts of charge. Concerning
the remaining natural amino acids, tyrosine, tryptophan,
and histidine show higher reactivity for bromination of the
substituted aromatic ring than for decarboxylation. Sulfur-
containing methionine and cysteine, on the other hand, are
converted into sulfoxides and sulfones. Fortunately, the lat-
ter amino acids contribute less than 10 wt.-% to the total
protein fraction in relevant protein sources such as maize,
wheat, and sugarcane waste.[2]
Acknowledgments
Acknowledgments are made to the Industrieel Onderzoeksfonds
KU Leuven (project IKP/10/005) for funding this research and to
KU Leuven for support through the Methusalem grant CASAS.
[1] E. Scott, F. Peter, J. Sanders, Appl. Microbiol. Biotechnol. 2007,
75, 751–762.
[2] T. M. Lammens, M. C. R. Franssen, E. L. Scott, J. P. M. San-
ders, Biomass Bioenergy 2012, 44, 168–181.
[3] J. P. M. Sanders, E. L. Scott, R. Weusthuis, H. Mooibroek,
Macromol. Biosci. 2007, 7, 105–117.
[4] T. M. Lammens, J. Le Notre, M. C. R. Franssen, E. L. Scott,
J. P. M. Sanders, ChemSusChem 2011, 4, 785–791.
[5] J. Le Nôtre, E. L. Scott, M. C. R. Franssen, J. P. M. Sanders,
Green Chem. 2011, 13, 807–809.
[6] A. H. Friedman, S. Morgulis, J. Am. Chem. Soc. 1936, 58, 909–
913.
[7] G. W. Stevenson, J. M. Luck, J. Biol. Chem. 1961, 236, 715–
717.
[8] G. Gopalakrishnan, J. L. Hogg, J. Org. Chem. 1985, 50, 1206–
1212.
[9] R. C. Hiremath, S. M. Mayanna, N. Venkatasubramanian, J.
Chem. Soc. Perkin Trans. 2 1987, 1569–1573.
[10] L. De Luca, G. Giacomelli, Synlett 2004, 2180–2184.
[11] E. V. Bellale, S. N. Huddar, U. S. Mahajan, K. G. Akamanchi,
Pure Appl. Chem. 2011, 83, 607–612.
Conclusions
[12] T. M. Lammens, D. De Biase, M. C. R. Franssen, E. L. Scott,
J. P. M. Sanders, Green Chem. 2009, 11, 1562–1567.
[13] A. But, J. Le Nôtre, E. L. Scott, R. Wever, J. P. M. Sanders,
ChemSusChem 2012, 5, 1199–1202.
[14] M. Nieder, L. Hager, Arch. Biochem. Biophys. 1985, 240, 121–
127.
[15] J. Sandeaux, R. Sandeaux, C. Gavach, H. Grib, T. Sadat, D.
Belhocine, N. Mameri, J. Chem. Technol. Biotechnol. 1998, 71,
267–273.
[16] D. G. Marangoni, I. G. N. Wylie, S. G. Roscoe, Bioelectrochem.
Bioenerg. 1991, 25, 269–284.
In conclusion, a thorough optimization of bromide-me-
diated electrochemical decarboxylation allows most natu-
rally occurring amino acids to be converted into valuable
products in very good to excellent yields. A CH3OH/H2O
solvent mixture in combination with ammonium bromide
appears most suited for decarboxylation of neutral, less-
polar amino acids. The alkaline and acid protein fractions
are most efficiently valorized in pure water with sodium
bromide as the mediator and electrolyte. Tunable selectivity
is observed for the decarboxylation of lysine towards
nitriles, amines, or amides depending on the cathode mate-
rial, acidity, and the amount of charge supplied to the cell.
[17] D. G. Maragoni, I. C. Smith, S. G. Roscoe, Can. J. Chem. 1989,
67, 921–926.
[18] S. M. MacDonald, S. G. Roscoe, Electrochim. Acta 1997, 42,
1189–1200.
[19] W. Huang, J. Zheng, Z. Li, J. Phys. Chem. C 2007, 111, 16902–
16908.
[20] N. A. Hampson, J. B. Lee, K. I. MacDonald, M. J. Shaw, J.
Chem. Soc. B 1970, 1766–1769.
[21] E. Klocke, A. Matzeit, M. Gockeln, H. J. Schäfer, Chem. Ber.
1993, 126, 1623–1630.
[22] J. J. Dai, Y. B. Huang, C. Fang, Q. X. Guo, Y. Fu, ChemSus-
Chem 2012, 5, 617–620
Experimental Section
Supporting Information (see footnote on the first page of this arti-
cle): Detailed description of the electrochemical procedures and
product analysis.
Received: August 19, 2014
Published Online: September 16, 2014
6652
www.eurjoc.org
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2014, 6649–6652