Effect of copper salts on hydrothermal oxidative decarboxylation: A study of phenylacetic acid

Article abstract of DOI:10.1039/c9cc09825a

Decarboxylation of carboxylic acids is favored under hydrothermal conditions, and can be influenced by dissolved metals. Here, we use phenylacetic acid as a model compound to study its hydrothermal decarboxylation in the presence of copper(ii) salts but no O₂. Our results showed a strong oxidizing role of copper in facilitating oxidative decarboxylation.

Full text of DOI:10.1039/c9cc09825a

ChemComm
View Article Online
View Journal
COMMUNICATION
Effect of copper salts on hydrothermal oxidative
decarboxylation: a study of phenylacetic acid†
Cite this:DOI: 10.1039/c9cc09825a
a
a
b
a
a
Xuan Fu, Megan Jamison, Aaron M. Jubb, Yiju Liao, Alexandria Aspin,
a
c
a
Received 18th December 2019,
Accepted 24th January 2020
Kyle Hayes, Christopher R. Glein and Ziming Yang
*
DOI: 10.1039/c9cc09825a
rsc.li/chemcomm
1
4–17
Decarboxylation of carboxylic acids is favored under hydrothermal redox reactions have been observed.
For example, reduction of
15
conditions, and can be influenced by dissolved metals. Here, we use ketones to alcohols can be facilitated by iron sulfide minerals,
phenylacetic acid as a model compound to study its hydrothermal oxidation of alcohols to aldehydes can be efficient in the presence
. Our of copper(II), and selective reduction of alkenes to alkanes is
results showed a strong oxidizing role of copper in facilitating feasible using Fe/Ni under hydrothermal conditions.
1
4
decarboxylation in the presence of copper(II) salts but no O
2
16
oxidative decarboxylation.
Organic–inorganic hydrothermal reactions are also of growing
interest to green chemistry. The concept of ‘‘geomimicry’’
Organic compound transformations in hydrothermal systems are has been proposed, which applies relatively benign and Earth-
of great importance to a wide range of chemical and geological abundant materials (e.g., iron, copper, zinc) to organic synthesis,
1
4,16
processes, including degradation of sedimentary organic matter, rather than the use of toxic or rare metal catalysts.
For
metabolism in the subsurface biosphere, and the deep carbon example, copper salts have been used as effective catalysts for
1
–4
18,19
cycle. Unlike at ambient conditions, water at elevated tempera- organic oxidation in green chemistry,
tures and pressures assumes key properties that promote organic usually require O to be present as the oxidant.
reactions. For example, hydrothermal water has a considerably study, we found that under hydrothermal conditions copper(II)
lower dielectric constant but a higher dissociation constant (K chloride was an efficient oxidant for alcohols, even in the
than ambient water, making it not only a good solvent for organic absence of O
molecules but also providing Brønsted acid and base catalysts in redox pathways for organic compounds in both natural and
but the reactions
2
0,21
2
In a previous
w
)
1
4
2
.
The oxidizing role of copper could create novel
+
À 5,6
the form of increased concentrations of H and OH . The human-made hydrothermal systems. For example, deep-sea
chemistry of a variety of organic functional groups has been mining indicates that copper-bearing minerals were deposited
examined under hydrothermal conditions, and their reaction at the seafloor from hydrothermal fluids,
2
2,23
and these minerals
7–12
pathways and mechanisms have been elucidated.
In particular, could release dissolved copper species that interact with organic
reversible and irreversible interconversions between organic func- carbon. Here we investigate the scope of hydrothermal organic
1
tional groups have been proposed in sedimentary basins, and oxidation using a suite of copper(II) salts, and we describe results
experimental data were collected showing the interconversions that compare effects of different cupric salts in oxidative dec-
between alkanes, alkenes, alcohols, aldehydes/ketones, and car- arboxylation of carboxylic acids. Carboxylic acids are known to
7,13
12,24,25
boxylic acids.
However, the hydrothermal chemistry of organic undergo decarboxylation under hydrothermal conditions,
compounds can be influenced by the surrounding rocks and but studies of hydrothermal interactions between carboxylic acids
minerals, and water-rock interactions can release dissolved metal and inorganic species are few, and in particular, the effect of
species that potentially dominate hydrothermal organic reactivity. dissolved metal ions on acid decarboxylation is largely unknown.
Transformations of organic molecules could be promoted by the In this study, we use phenylacetic acid (PAA) as a model
presence of minerals, and indeed some remarkable hydrothermal compound to examine the effects of cupric salts, based on its
known hydrothermal chemistry and predictable decarboxylation
a
12,14
Department of Chemistry, Oakland University, Rochester, MI 48309, USA.
E-mail: zimingyang@oakland.edu
pathways.
Results from this work may provide new insights
into the roles of dissolved metal species in hydrothermal organic
transformations.
Hydrothermal experiments were performed in fused silica
glass tubes using a previously developed method. Experiments
b
c
U.S. Geological Survey, 12201 Sunrise Valley Drive, Reston, VA 20192, USA
Space Science and Engineering Division, Southwest Research Institute, San
Antonio, TX 78228, USA
2
6
†
Electronic supplementary information (ESI) available: Detailed experimental
set ups and experimental materials. See DOI: 10.1039/c9cc09825a
were conducted under a hydrothermal condition of 200 1C and
This journal is © The Royal Society of Chemistry 2020
Chem. Commun.

View Article Online
Communication
ChemComm
promote acid decarboxylation and the following oxidation of
primary products. UV-Vis spectroscopy analysis showed that the
dissolved copper(II) was mostly consumed after the reaction
(Fig. S3, ESI†), which verifies the oxidative role of copper(II)
in this reaction scheme. Benzaldehyde and benzoic acid were
formed as the major products of PAA, but their distributions
were dependent on the type of copper(II) salts. As shown in
Fig. 2, all copper experiments showed trends of increasing
concentrations of benzaldehyde and benzoic acid over time,
except for copper nitrate which showed a decreasing concen-
tration of benzaldehyde that implies its further oxidation to
benzoic acid after 0.5 hours. To compare the oxidizing effects
between the copper salts, the ratio of benzoic acid to benzalde-
hyde was calculated. Fig. 2c shows that the acid:aldehyde ratios
were between 0.3 and 2.3 in the presence of copper chloride/
Fig. 1 Hydrothermal decarboxylation of phenylacetic acid (PAA) in the sulfate/acetate, whereas the ratio in the copper nitrate experi-
presence of copper or sodium salts at 200 1C and 15 bar (Psat) for up to 6 hours.
ment substantially increased to a value of 47.6 after 6 hours.
This high acid:aldehyde ratio indicates that copper(II) nitrate
was a much stronger oxidant than the other copper salts in
hydrothermal transformation of carboxylic acids.
2
7
1
5 bar (Psat, calculated using SUPCRT92) in the absence of O
and the effects of four copper(II) salts including CuCl , copper
sulfate (CuSO ), copper acetate [Cu(OAc) ], and copper nitrate
Cu(NO ] were compared. At 200 1C, PAA reacted very slowly in
pure hydrothermal water, with a o2% decomposition after
hours. In the presence of the copper salts, however, more than
0% of the acid was degraded after 6 hours (Fig. 1). Time-series
2
,
2
Because sodium salts did not show any oxidizing effects under
the same hydrothermal condition (Fig. 1), we hypothesize that the
main driver of oxidative decarboxylation could be complexes
between copper(II) and the anions, rather than freely solvated
copper(II). Indeed, thermodynamic calculations on copper speciation
under similar hydrothermal conditions (e.g., 250 1C and 40 bar)
suggested that the dominant forms of aqueous copper species
were copper complexes (e.g., with chloride or acetate) instead of
4
2
[
3 2
)
6
4
experiments showed that the decomposition of PAA mostly
occurred in the first 2 hours and then reached a steady state
Fig. 1). Among the four copper(II) salts with the same concen-
(
1
4
uncomplexed copper(II). To investigate the effect of the nitrate
complex, other divalent but non-oxidizing metal ions including
tration of copper (i.e., 0.2 molal), the trend in decomposing PAA
followed as copper nitrate * copper chloride E copper acetate 4
copper sulfate, with copper nitrate exhibiting the most significant
effect (e.g., B70% acid decomposition after 2 hours). Experiments
with a lower molar ratio of 1 : 4 between PAA and copper(II) were
also conducted, in which even higher decomposition of PAA was
observed when larger amounts of copper(II) were used (Table S1,
ESI†). In contrast, hydrothermal experiments with sodium salts
with the same counterions showed negligible PAA decomposition
2
+
2+
2+
magnesium (Mg ), zinc (Zn ), and nickel (Ni ) were studied in
the form of nitrate salts to react with PAA at 200 1C for 4 hours,
respectively. Compared to sodium nitrate, magnesium nitrate,
zinc nitrate, and nickel nitrate all increased the PAA conversion
by 10–15%, with benzaldehyde and benzoic acid formed as the
major products (Fig. 3). Since these metals are non-oxidizing,
forming the oxidation products implies a potential oxidizing role
of divalent metal–nitrate complexes in the hydrothermal system.
Interestingly, all three metal nitrates yielded more benzoic acid
than benzaldehyde (acid/aldehyde ratio between 3.6 and 4.5),
indicating the oxidation step from benzaldehyde to benzoic
acid was promoted by the presence of nitrate. However, the
PAA decomposition with non-oxidizing metal nitrates was not
comparable to that with copper nitrate (Fig. 3), suggesting that
the oxidative decarboxylation was mainly driven by the copper(II)
complexes, whereas the oxidation of benzaldehyde was facili-
tated by the nitrate (Scheme 1).
(
all below 2% after 4 hours) (Fig. 1), suggesting the reactions were
not driven by the anions or nitrate alone. These results demon-
strate different effects of copper salts on carboxylic acid decom-
position, which could be attributed to the formation of different
copper(II) complexes under the hydrothermal condition (see below).
Copper(II)-facilitated PAA hydrothermal transformation
was proposed to undergo a series of oxidative steps starting
with oxidative decarboxylation to form benzyl alcohol, then
benzaldehyde, and benzoic acid (Scheme 1). In these steps,
copper ions were proposed to serve as a reactive oxidant to
To further study the copper complexation, Raman spectro-
scopy analysis was also performed for hydrothermal experiments
2
with PAA and with PAA + CuCl , respectively. Raman spectra
were collected every 30 minutes during a 2 hour hydrothermal
reaction of PAA at 200 1C, and the peaks were assigned according
2
8
to Lin-Vien et al. (1991). No obvious changes were observed in
the PAA alone experiment (Fig. 4a), which is consistent with the
low reactivity of PAA under the hydrothermal condition. In the
Scheme 1 Proposed hydrothermal oxidative decarboxylation pathway
from phenylacetic acid (PAA) to benzyl alcohol, benzaldehyde, and ben-
zoic acid in the presence of copper(II).
2
presence of CuCl , however, several significant changes are
Chem. Commun.
This journal is © The Royal Society of Chemistry 2020

View Article Online
ChemComm
Communication
Fig. 2 Concentrations (mmolal) of major product benzaldehyde (a) and benzoic acid (b), and ratios of benzoic acid/benzaldehyde (c) vs. time in
hydrothermal experiments of PAA with copper chloride (blue circles), copper sulfate (orange squares), copper acetate (green diamonds), and copper
nitrate (red triangles) at 200 1C and 15 bar (Psat).
Fig. 3 Hydrothermal decarboxylation of PAA (a) and its product distribu-
tion (b) in the presence of magnesium nitrate, zinc nitrate, nickel nitrate,
and copper nitrate at 200 1C and 15 bar (Psat) after 4 hours.
observed in the spectra (Fig. 4b). The CQO group became more
abundant through the hydrothermal process, which is consistent
with the production of benzaldehyde and benzoic acid from PAA.
The intensity of a Raman response is equal to the product of the
input laser energy, the number of oscillators, and the Raman
28
cross-section of the vibration; the Raman cross-section is propor-
tional to the change in polarizability induced during the vibration.
The increased CQO and CQC (from the benzene ring) intensities
in Fig. 4b could be explained by a higher cross-section of the
vibration in benzaldehyde and benzoic acid than that in PAA, since
an increased proximity of the aromatic ring to the CQO moiety
may induce a greater change in polarizability during vibration.
More interestingly, an increasing fluorescence was detected over
the reaction time, which may suggest the formation of copper- Fig. 4 Time-series Raman spectroscopy analysis (every 30 minutes) on
2
9–31
coordinated complexes in the system.
These Raman results hydrothermal decarboxylation of PAA (a) and PAA with copper chloride
(
b) at 200 1C and 15 bar (Psat) for up to 2 hours. Dashed lines indicate major
provided additional evidence for the presence of copper complexes
in the PAA experiments.
Copper(II) nitrate is a readily available cupric salt and has
emerged as a versatile reagent for organic synthesis. It has been
peaks with assignments taken from Lin-Vien et al. (1991). Aro = aromatic,
sym = symmetric.
32
used as an oxidizing agent and also a Lewis acid catalyst. For experimental conditions at 200 1C, however, we found the conver-
example, an organic synthesis study showed that copper nitrate sions of PAA after 1 and 2 hours were 4.4 Æ 0.2% and 14.6 Æ 0.3%,
3 2
allows facile oxidative decarboxylation of benzoyl formic acid to respectively, which were much lower than those in the Cu(NO )
33
form benzonitrile in the presence of aqueous ammonium and O
2
,
experiments (Fig. 1). In contrast, experiment of PAA with a mixture
although the reaction conditions were quite different from the of Cu(OTf) and NaNO yielded similar conversion to that with
2
3
current work. To investigate whether copper nitrate played a role of Cu(NO ) (Table S1, ESI†). These results further suggest that the
3
2
Lewis acid in this decarboxylation, we chose another common complexation, instead of Lewis acidity, is more important to
Lewis acid – copper(II) triflate [Cu(OTf) ] to study. The triflate ion is enhance the oxidative ability of copper(II).
weakly nucleophilic, which makes the cupric ion more cationic and
2
Copper nitrate experiments with substituted phenylacetic acids,
At the same including p-tolylacetic acid and p-fluorophenylacetic acid, were also
34
Cu(OTf)
2
a stronger Lewis acid than Cu(NO
3
)
2
.
This journal is © The Royal Society of Chemistry 2020
Chem. Commun.

View Article Online
Communication
ChemComm
Conflicts of interest
There are no conflicts to declare.
Notes and references
1
2
3
4
J. S. Seewald, Nature, 2003, 426, 327–333.
I. M. Head, D. M. Jones and S. R. Larter, Nature, 2003, 426, 344–352.
T. M. McCollom and J. S. Seewald, Chem. Rev., 2007, 107, 382–401.
D. A. Sverjensky, V. Stagno and F. Huang, Nat. Geosci., 2014, 7,
Fig. 5 Hydrothermal transformation of benzyl alcohol (a) and its product
distribution (b) in the presence of the copper salts at 200 1C and 15 bar
(Psat) after 4 hours.
909–913.
5
6
N. Akiya and P. E. Savage, Chem. Rev., 2002, 102, 2725–2750.
M. Watanabe, T. Sato, H. Inomata, R. L. Smith, K. Arai, A. Kruse and
E. Dinjus, Chem. Rev., 2004, 104, 5803–5822.
7
8
9
Z. Yang, I. R. Gould, L. B. Williams, H. E. Hartnett and E. L. Shock,
Geochim. Cosmochim. Acta, 2012, 98, 48–65.
C. Bockisch, E. D. Lorance, H. E. Hartnett, E. L. Shock and
I. R. Gould, ACS Earth Space Chem., 2018, 2, 821–832.
K. M. Fecteau, I. R. Gould, C. R. Glein, L. B. Williams, H. E. Hartnett
and E. L. Shock, ACS Earth Space Chem., 2019, 3, 170–191.
conducted to further clarify the decarboxylation mechanism. At
2
1
00 1C, the acid conversions after 1 hour were 42.1 Æ 0.3% and
3.6 Æ 0.2% for p-tolylacetic acid and p-fluorophenylacetic acid,
respectively, while the conversion for parent PAA was 36.1 Æ 0.1%.
This conversion trend ( p-tolylacetic acid 4 PAA 4 p-fluorophenyl-
acetic acid) thus suggests a carbocation intermediate (in Scheme 1),
10 K. J. Robinson, I. R. Gould, K. M. Fecteau, H. E. Hartnett, L. B.
Williams and E. L. Shock, Geochim. Cosmochim. Acta, 2019, 244,
113–128.
3
since the electron-donating substituent (–CH ) on the benzene ring
11 Z. Yang, E. D. Lorance, C. Bockisch, L. B. Williams, H. E. Hartnett,
accelerated the reaction, whereas the electron-withdrawing substi-
E. L. Shock and I. R. Gould, J. Org. Chem., 2014, 79, 7861–7871.
tuent (–F) slowed down the reaction. In addition, experiments 12 C. R. Glein, I. R. Gould, E. D. Lorance, H. E. Hartnett and
E. L. Shock, Geochim. Cosmochim. Acta, 2020, 269, 597–621.
starting with benzyl alcohol were also performed, and results
showed that benzaldehyde and benzoic acids were formed as the
major products (Fig. 5), which further support the reaction steps
1
3 E. L. Shock, P. Canovas, Z. Yang, G. Boyer, K. Johnson, K. Robinson,
K. Fecteau, T. Windman and A. Cox, Rev. Mineral. Geochem., 2013,
76, 311–350.
1
1
1
1
1
1
2
2
4 Z. Yang, H. E. Hartnett, E. L. Shock and I. R. Gould, J. Org. Chem.,
proposed in Scheme 1. Specifically, in the presence of CuCl
CuSO , and Cu(OAc) , the benzyl alcohol conversions were below
0% with an acid:aldehyde ratio o0.1, while in the presence of
Cu(NO the conversion was nearly 100%, with a ratio increased to
5 : 1. These observations are also consistent with the PAA experi-
2
,
2015, 80, 12159–12165.
4
2
5 Z. Yang, I. R. Gould, L. B. Williams, H. E. Hartnett and E. L. Shock,
Geochim. Cosmochim. Acta, 2018, 223, 107–126.
6 C. Bockisch, E. D. Lorance, G. Shaver, L. B. Williams, H. E. Hartnett,
E. L. Shock and I. R. Gould, Green Chem., 2019, 21, 4159–4168.
7 J. A. Shipp, I. R. Gould, E. L. Shock, L. B. Williams and
H. E. Hartnett, Proc. Natl. Acad. Sci. U. S. A., 2014, 111, 11642–11645.
8 P. Gamez, P. G. Aubel, W. L. Driessen and J. Reedijk, Chem. Soc. Rev.,
4
3 2
)
1
ments, in which copper nitrate behaved as the strongest oxidant.
In conclusion, the fact that the organic reactivity and product
distribution were dependent on the type of copper salts highlights a
selective role of metal ions/complexes in hydrothermal organic
redox transformations. The unique properties of water at high
temperatures and pressures not only allow organic reactions to
2001, 30, 376–385.
9 J. Liu, X. Zhang, H. Yi, C. Liu, R. Liu, H. Zhang, K. Zhuo and A. Lei,
Angew. Chem., Int. Ed., 2015, 54, 1261–1265.
0 S. E. Allen, R. R. Walvoord, R. Padilla-Salinas and M. C. Kozlowski,
Chem. Rev., 2013, 113, 6234–6458.
1 S. D. McCann and S. S. Stahl, Acc. Chem. Res., 2015, 48, 1756–1766.
occur in aqueous medium, but also influence the speciation and 22 J. R. Hein, K. Mizell, A. Koschinsky and T. A. Conrad, Ore Geol. Rev.,
2
013, 51, 1–14.
complexation of dissolved metals, which can selectively control
organic redox transformations in hydrothermal environments.
2
3 K. A. Miller, K. F. Thompson, P. Johnston and D. Santillo, Frontiers
in Marine Science, 2018, 4, 418.
While the majority of hydrothermal organic geochemistry research 24 J. B. Dunn, M. L. Burns, S. E. Hunter and P. E. Savage, J. Supercrit.
Fluids, 2003, 27, 263–274.
have been focusing on organic-mineral interactions, this work
provides new insights into the hydrothermal interactions between
2
5 T. M. McCollom and J. S. Seewald, Geochim. Cosmochim. Acta, 2003,
7, 3625–3644.
organic and dissolved metal species, which are important to further 26 Z. Yang and X. Fu, JoVE, 2018, e58230.
6
2
7 J. W. Johnson, E. H. Oelkers and H. C. Helgeson, Comput. Geosci.,
992, 18, 899–947.
our understanding of hydrothermal systems, as well as to discover
1
14,16
more ‘‘geomimicry’’ approaches in green chemistry.
2
8 D. Lin-Vien, N. B. Colthup, W. G. Fateley and J. G. Grasselli, The
handbook of infrared and Raman characteristic frequencies of organic
molecules, Academic Press, 1991.
9 M. H. Lim and S. J. Lippard, J. Am. Chem. Soc., 2005, 127, 12170–12171.
0 R. Kr ¨a mer, Angew. Chem., Int. Ed., 1998, 37, 772–773.
ZY acknowledges the support from the Michigan Space
Grant Consortium and startup fund from Oakland University.
AMJ was supported by the USGS Energy Resources Program.
2
3
The authors thank Bill Orem (USGS) for his review that 31 X.-M. Chen and M.-L. Tong, Acc. Chem. Res., 2007, 40, 162–170.
3
2 M. Gao, R. Ye, W. Shen and B. Xu, Org. Biomol. Chem., 2018, 16,
602–2618.
3 P. K. Arora and L. M. Sayre, Tetrahedron Lett., 1991, 32, 1007–1010.
improved this manuscript. Use of trade, product, or firm names
is for descriptive purposes only and does not imply endorse-
ment by the U.S. Government.
2
3
34 C. Hertweck, J. Prakt. Chem., 2000, 342, 316–321.
Chem. Commun.
This journal is © The Royal Society of Chemistry 2020