Alkanes to carboxylic acids in aqueous medium: Metal-free and metal-promoted highly efficient and mild conversions

Article abstract of DOI:10.1039/b900853e

A convenient and clean route to transform, in aqueous medium, various alkanes to carboxylic acids via single-pot carboxylation with CO and water, under mild conditions, has been achieved, proceeding efficiently and selectively even without any metal catalyst and any acid additive, at low temperatures; the relevant hydroxylating role of H₂O and radical mechanisms are disclosed by radical-trap, H₂¹⁸O and DFT studies. The Royal Society of Chemistry 2009.

Full text of DOI:10.1039/b900853e

Supplementary Material (ESI) for Chemical Communications
This journal is (c) The Royal Society of Chemistry 2009
Electronic Supplementary Information (ESI)
Alkanes to Carboxylic Acids in Aqueous Medium: Metal-free
and Metal-promoted Highly Efficient and Mild Conversions
Marina V. Kirillova, Alexander M. Kirillov, Maxim L. Kuznetsov, José A. L. Silva, João J. R.
Fraústo da Silva, and Armando J. L. Pombeiro*
Centro de Química Estrutural, Complexo I, Instituto Superior Técnico, TU Lisbon, Av. Rovisco Pais, 1049-001, Lisbon,
Portugal. Fax: +351 218464455; Tel: +351 218419237; E-mail: pombeiro@ist.utl.pt
ESI contains Full Experimental Details including general methods, experiments with 18O-labeled
water and full computational details; Supporting Discussion regarding the reaction mechanism;
Supporting References; Supporting Figures S1–S5 with effects of different reaction parameters
(Figs. S1, S2) on the yields of carboxylic acids, the ESI-MS– spectra (Fig. S3) of the reaction
18
mixtures (including experiments with O-labeled water), as well as graphical material from DFT
calculations with equlibrium structures of the transition states (Fig. S4), and energy profile for the
carboxylation of ethane (Fig. S5); Supplementary Tables with comparison of different systems for
carboxylation of alkanes in aqueous medium (Table S1), full version (including all experimental
conditions) of Table 1 (Table S2), with results on carboxylation of cyclohexane promoted by
various copper(II) compounds (Table S3), and with total energies and enthalpies of the structures
from theoretical calculations (Table S4).

2
1. Full Experimental Details
General methods. All chemicals were obtained from commercial sources and used as
received. Compound 1 was prepared according to the known12 method. GC analyses were
performed on a Fisons Instruments GC 8000 series gas chromatograph with a DB WAX (J&W)
capillary column (P/N 122-7032; 30 m × 0.25 mm × 0.25 μm; helium was the carrier gas) and the
Jasco-Borwin v.1.50 software. In some cases, carboxylation products were also identified by GC-
MS, 1H and 13C-{1H} NMR techniques using a Trio 2000 Fisons spectrometer with a coupled Carlo
Erba (Auto/HRGC/MS) gas chromatograph, and a Varian UNITY 300 NMR spectrometer,
respectively.
Alkane Carboxylations. In a typical experiment the reaction mixtures were prepared as
follows: to 2.0−32.0 µmol (typically 8.0 µmol) of metal promoter (optional) contained in a 13.0 mL
stainless steel autoclave, equipped with a Teflon-coated magnetic stirring bar, were added 1.00–
2.00 mmol (typically 1.50 mmol) of K2S2O8, 2.0−3.0 mL of H2O, 2.0–4.0 mL of MeCN (typical
total solvent volume of 6.0 mL) and 1.00–1.50 mmol (typically 1.00 mmol) of alkane (if liquid).
Then the autoclave was closed and flushed with N2 three times for removing the air and pressurized
with 1−10 atm of alkane (if gaseous) and 0−40 atm (typically 20 atm) of CO. The reaction mixture
was stirred for 1−20 h (typically 6 h) at 25–60 °C (typically 50 or 60°C) using a magnetic stirrer
and an oil bath, whereupon it was cooled in an ice bath, degassed, opened and transferred to a
Schlenk flask. Diethyl ether (9.0–11.0 mL) and 90 μL of cycloheptanone (GC internal standard)
were added. The obtained mixture was stirred and the organic layer was typically analyzed by GC.
Additional experiments were performed under typical reaction conditions in the presence of
the carbon-centered radical trap CBrCl3 (used in an equimolar amount relatively to the alkane),
revealing the full suppression of carboxylic acid formation, either in the presence or in the absence
of the metal promoter.
Blank tests indicate that the metal-free carboxylation reactions proceed with similar efficacy
in a Teflon reactor, thus confirming that reactions are not promoted by the reactor metal wall.
Experiments with 18O-labeled water. Additional experiments (cyclohexane carboxylation)
18
18
have been performed using H2 O (97% O, Cambridge Isotope Laboratories, Inc.) as solvent in
order to check the involvement of water in the formation of the COOH group of the carboxylic acid
product. The experiments were carried out similarly as described in Methods Summary (in a 13.0
18
mL stainless steel autoclave), reacting 1.00 mmol of K2S2O8, 1.0 g of H2 O, 2.0 mL of MeCN, 1.00
mmol of cyclohexane, 20 atm of CO and 2.0 μmol of metal promoter 1 (or in the absence of 1 in a
separate batch) for 5 h at 50ºC. After the completion the reaction and addition of 3.0 mL of Et2O,
the reaction mixture (organic phase) was analyzed by gas chromatography (using a Fisons

3
Instruments GC 8000 series gas chromatograph) and ESI-MS– spectrometry (using a 500-MS LC
Ion Trap instrument; Varian Inc, Alto Palo, CA, USA). The ESI–MS– spectra were run using (if
necessary) further diluted MeCN/Et2O solutions (with ca. 10–3 M of cyclohexanecarboxylic acid
product) of the reaction mixtures. The ESI–MS– spectra (Fig. S3) reveal the formation of
C6H11CO18OH as the main product [the relative peak intensities of C6H11CO18O– (m/z = 128.9) and
C6H11COO– (m/z = 126.9) are 4:1 or 3:2 for the experiments with cyclohexane in the presence or in
the absence of the copper promoter 1, respectively].
Computational details. Calculations were performed for reactions of ethane at the B3LYP
level using the Gaussian-98 program.S1 For the non-metal atoms, the 6-311+G** basis sets were
applied. For the copper atoms, a relativistic Stuttgart pseudopotential and the appropriate basis set
(8s7p6d)/[6s5p3d] were used.S2 Solvent effects were treated using the CPCM model with water as
solvent. Model aqua-copper-complexes were considered instead of the tetranuclear CuII complex for
simplicity and because Cu(NO3)2 also shows a promoting effect.
The Hessian matrix was calculated analytically for the optimized structures in order to prove
the location of correct minima (no “imaginary” frequencies) or saddle points (only one negative
eigenvalue). Thermodynamic parameters were calculated at 60°C and pressure of 10 atm, i.e at
typical experimental conditions. The solvent effects have been estimated at the single-point
calculations on the basis of the gas-phase equilibrium geometries. The enthalpies in solution (Hs)
were calculated by addition of the solvation energy ΔGsolv to gas-phase enthalpies (Hg).
The solvent effects cannot be treated accurately if the total number of the charged species
with the same charges is different before and after reaction. In order to preserve this number, one
chloride ion atom was added to the complexes [Cu(H2O)4]n+. First, the full geometry optimization
of [Cu(H2O)4]n+ has been carried out, then the Cl– ion, as a counterion, was placed along the axis
orthogonal to the {Cu(O)4} plane at the distance of 4 Å from the Cu atom, and the gas-phase and
CPCM single-point calculations have been performed on the basis of the equilibrium geometries of
[Cu(H2O)4]n+. Zero point energies and translation, rotation and vibrational contributions to the total
energies were taken from the equilibrium geometries of the [Cu(H2O)4]n+ species.
The geometry optimization of the reduced species [Cu(H2O)4]+ led to elimination of two
water molecules from the coordination sphere of the metal.
2. Supporting Discussion
Other less favourable mechanisms involving the hydroxyl radical. Alternatively, but
–
with an higher barrier of activation, SO4 could react with water [H2O + SO4 → HO▪ + HSO4 ]
giving the HO▪ radical (ΔHs≠ = 17.9 kcal/mol) which on reaction with the acyl radical RCO▪ would

4
yield the carboxylic acid. RCO• could also directly react with water giving RCOOH and H• (ΔHs≠ =
2–
16.8 kcal/mol), but this is also less favourable than the oxidation by S2O8 or Mn. In accord, the
selectivity observed when using linear alkanes (see text) also suggests that the reactions do not
proceed via the highly reactive HO• radical.S3 In fact, the linear derivative (i.e. butyric acid) has
been detected in a significant amount (ca. 4−5% yield) only in the case of propane carboxylation
(Table 1, entry 2). For all the other n-alkanes, the yields of the linear carboxylic acids have not
exceeded 1%.
The carboxylation reaction does not proceed in the presence of O2 as the oxidant, but when
O2 is used together with persulfate, a significant yield drop (ca. 10 times) of carboxylic acid is
observed. Therefore, O2 acts also as a radical trap, reacting with the alkyl radical to give the peroxo
radical ROO▪, thus leading to higher yields (ca. tenfold increase) of alcohol and ketone, what is in
accord with the previously reportedS4,S5 radical mechanism for the oxidation of alkanes (RH) via the
formation of R• radicals. Nevertheless, in our systems the overall formation of the alcohols and
ketones (e.g. cyclohexanol and cyclohexanone in the case of cyclohexane) does not exceed 1–3% of
the total yield based on alkane, and is detected preferably at lower CO pressures, when the
carboxylation is less favourable.
As in the TFA containing systems,10] the carboxylation in H2O/MeCN medium does not
appear to proceed via formation of free CO2 or free alcohols, since CO2 does not promote the
carboxylation of alkanes, and the alcohols are not converted into carboxylic acids in the presence of
CO in our systems.
3. Supporting References
(S1) M. J. Frisch et. al., Gaussian 98, revision A.9, Gaussian, Inc.: Pittsburgh PA, 1998.
(S2) M. Dolg, U. Wedig, H. Stoll, H. Preuss, J. Chem. Phys. 1987, 86, 866.
(S3) M. Costas, K. Chen, L. Que, Jr. Coord. Chem. Rev. 2000, 200–202, 517.
(S4) G. B. Shul'pin, J. Mol. Cat. A. Chem. 2002, 189, 39.
(S5) A. E. Shilov, G. B. Shul'pin, Chem. Rev. 1997, 97, 2879.
(S6) A. M. Kirillov, M. N. Kopylovich, M. V. Kirillova, E. Y. Karabach, M. Haukka, M. F. C. G.
da Silva, A. J. L. Pombeiro, Adv. Synth. Catal. 2006, 348, 159.

5
4. Supporting Figures
Full experimental conditions for Fig. 1. General reaction conditions: p(C3H8) = 5 atm (1.33
mmol) or cyclohexane (1.00 mmol), compound 1 (8.0 μmol, optional), in an autoclave (13.0 mL-
capacity), 6 h. Reaction conditions in (a): p(CO) = 30 or 20 atm for C3H8 or C6H12 carboxylation,
respectively; K2S2O8 = 1.5 mmol; total H2O/MeCN mixed solvent volume (6 mL); 60 or 50°C for
C3H8 or C6H12 carboxylation, respectively. Reaction conditions in (b): p(CO) = 20 atm; K2S2O8 =
1.5 or 1.0 mmol for C3H8 or C6H12 carboxylation, respectively; H2O (3.0 mL)/MeCN (3.0 mL);
30−100°C.
80
70
60
50
40
30
20
10
0
0
2
4
6
8
10
12
14
16
Metal promoter amount, μmol
Figure S1. Effect of the amount of metal promoter 1 on the yield of cyclohexanecarboxylic acid.
Reaction conditions: cyclohexane (1.00 mmol), p(CO) = 20 atm (5.32 mmol), K2S2O8 (1.50 mmol),
H2O (3.0 mL)/MeCN (3.0 mL), 50 °C, 6 h in an autoclave (13.0 mL capacity).

6
(a)
(b)
Figure S2. Effects of time (a) and CO pressure (b) on the total yields (% relatively to alkane) of
carboxylic acids. Orange and blue curves correspond to carboxylation of propane and cyclohexane,
respectively, promoted by 1 or metal-free (curves with asterisks). General reaction conditions:
p(C3H8) = 5 atm (1.33 mmol) or cyclohexane (1.00 mmol), compound 1 (8.0 μmol, optional), in an
autoclave (13.0 mL-capacity), 6 h. Reaction conditions in (a): p(CO) = 30 or 20 atm for C3H8 or
C6H12 carboxylation, respectively; K2S2O8 = 1.5 or 1.0 mmol for C3H8 or C6H12 carboxylation,
respectively; H2O (3.0 mL)/MeCN (3.0 mL); 50 or 80°C for C3H8 or C6H12 carboxylation,
respectively; 0−20 h. Reaction conditions in (b): p(C3H8) = 5 atm; p(CO) = 0−40 atm; K2S2O8 = 1.5
mmol; H2O (3.0 mL)/MeCN (3.0 mL); 60 or 50°C for C3H8 or C6H12 carboxylation, respectively; 6
h.

Supplementary Material (ESI) for Chemical Communications
This journal is (c) The Royal Society of Chemistry 2009
Figure S4. Equilibrium structures of the transition states for ethane carboxylation.

9
ΔHs
kcal/mol
TS1
5
0
(1)
•
–
(2)
TS2
SO4
+
•
C2H5
–10
–20
–30
–40
–50
–60
C2H6
TS3'
(3')
C2H5CO•
(5)
TS5
(6)
C2H5CO+
(6a)
C2H5C(O)OSO2OH
TS4'
–70
–80
–90
(4')
(9)
C2H5CO(H)OH+
TS6
(7)
C2H5C(O)OSO2OH
C2H5COOH•••SO3
C2H5COOH
C2H5COOH
Figure S5. Energy profile for the carboxylation of ethane to propionic acid.

Supplementary Material (ESI) for Chemical Communications
This journal is (c) The Royal Society of Chemistry 2009
Table S2. Full version (including all experimental details) of Table 1. Metal-free and metal-promoted
direct carboxylation, in H2O/MeCN medium, of Cn alkanes to Cn+1 carboxylic acids.a
Total yield (%)
Run
Alkane
Carboxylic acid product
Metal-promoted
(by 1)
Metal-free
9.9 b
1
2
9.4 c
COOH
COOH
23.2 d,e
38.0 c,f
29.6 g
COOH
3
6.2
COOH
COOH
4
5
6
17.6 f
8.2 g,h
9.0 f
23.2 f
22.2 c,g,h
44.6 f
COOH
COOH
COOH
COOH
COOH
7
12.3 g
72.3 g,i
a Selected results; typical (unless otherwise stated) reaction conditions: p(gaseous alkane) = 10, 3 or 1.5 atm (2.66,
0.78 or 0.40 mmol) for C2H6, C3H8 or n-C4H10, respectively, liquid alkane (1.00 mmol), p(CO) = 20 atm (5.32 mmol),
K2S2O8 (1.50 mmol), H2O (3.0 mL)/MeCN (3.0 mL), in the presence (optional) of compound 1 (8.0 μmol), 60 °C, 6 h in an
autoclave (13.0 mL capacity). [Product yield % (moles of carboxylic acid products / 100 moles of alkane) determined by
GC analysis; % yields based on oxidant, if required, can be estimated as [(% yield based on alkane)×n(mmol alkane)/1.5],
e.g. 8.2 and 48.2% for entry 7. Major product is indicated first; molar ratios between the two carboxylic acids for metal-free
and metal-promoted (values in brackets) systems are: 6.5:1 (9.0:1), 3.2:1 (2.7:1) or 1.0:1 (1.1:1) for C3H8, n-C5H12 or n-
C6H14 carboxylation, respectively; the formation of minor amounts (<1%) of linear carboxylic acids is not considered
c
e
herein. b K2S2O8 (2.00 mmol). p(CO) = 30 atm. d 70 °C. p(C3H8) = 5 atm (1.33 mmol). f H2O (2.0 mL)/MeCN (4.0 mL).
g 50 °C. h C5H10 (1.50 mmol). i Metal promoter (4.0 μmol).

12
Table S3. Carboxylation of cyclohexane to cyclohexanecarboxylic acid, in H2O/MeCN medium,
promoted by copper(II) compounds.a
Metal
Entry
Metal promoter
promoter
Yield (%)
amount (μmol)
1
2
3
4
Cu(NO3)2·2.5H2O
[Cu(H2tea)(N3)]b
[Cu2(H2tea)2{C6H4(COO)2-1,4}]n·2nH2Ob
32
16
31.6
32.5
38.8
72.3
8.0
4.0
[O⊂Cu4{N(CH2CH2O)3}4(BOH)4][BF4]2 (1)
a Selected results; typical (unless otherwise stated) reaction conditions: cyclohexane (1.00 mmol), p(CO) = 20 atm (5.32
mmol), K2S2O8 (1.50 mmol), H2O (3.0 mL)/MeCN (3.0 mL), 50 °C, 6 h in an autoclave (13.0 mL capacity). Product yield
% (moles of cyclohexanecarboxylic acid / 100 moles of cyclohexane). H2tea = monodeprotonated form of triethanolamine.
b See refs. 11 and S6 for the syntheses of these compounds.

13
Table S4. Total energies for gas phase and for solution (Eg and Es) and enthalpies (H) of the structures
calculated at the B3LYP/6-311+G** level of theory (in Hartree).
Structure
Eg
Es
H
–
HS2O8
–1398.911453
–699.174833
–699.840384
–700.342619
–79.857573
–79.186023
–113.350544
–192.558831
–192.305151
–892.367288
–268.483713
–268.805144
–892.373626
–623.859515
–75.763068
–76.459175
–76.731723
–0.502156
–1398.983700 –1398.854742
–•
SO4
–699.266116
–699.946412
–700.368983
–79.854516
–79.186193
–113.349498
–192.561275
–192.398868
–892.376104
–268.500856
–268.917258
–892.376469
–623.859028
–75.771250
–76.469648
–76.901184
–0.499790 [a]
–699.154177
–699.806624
–700.296325
–79.777690
–79.120906
–113.341594
–192.479502
–192.224536
–892.248910
–268.385190
–268.693667
–892.254917
–623.842190
–75.750707
–76.433414
–76.692790
–0.499360
–
HSO4
H2SO4
C2H6
•
C2H5
CO
C2H5CO•
C2H5CO+
C2H5C(O)OSO2OH
C2H5COOH
C2H5CO(H)OH+
C2H5COOH···SO3
SO3
OH•
H2O
H3O+
H•
{[Cu(H2O)4]Cl}+
[Cu(H2O)2](H2O)2Cl
[Cu(H2O)4]2+
–963.193436 [b] –963.342604 [b]
–963.463504 [b] –963.545060 [b]
–502.596129
–503.052478
–779.011071
–192.532472
–1591.465185
–968.801036
–268.725353
–892.362069
–502.939167
–503.140830
–779.105302
–192.531920
–502.479940
–502.940458
–778.914931
–192.456921
+
[Cu(H2O)2](H2O)2
TS1
TS2
TS3'
TS4'
TS5
TS6
–1591.539337 –1591.328272
–968.656418
–268.855249
–892.370841
–968.828791
–268.620461
–892.248117
a For H•, the Bondi’s atomic radius was used. b For details, see Computational details section.