A comparative study of the ruthenium(VI)dioxocarboxylato salts, [PPh4][RuO2(OCOR)Cl2] (R = CH3, CF3, C6H5, C6F5, C5H11), in the oxidation of alcohols

Article abstract of DOI:10.1016/j.molcata.2007.02.049

The compounds [PPh₄][Ru(O)₂(OCOR)Cl₂] (R = CH₃ 1a, CF₃ 1b, C₆H₅ 1c, C₆F₅ 1d, C₅H₁₁ 1e) were prepared and fully characterised. The fluorinated compounds 1b and 1d were obtained in significantly higher yields than their protonated analogues 1a and 1c and compound 1b was found to be a clearly superior stoichiometric oxidant to compound 1a. The compounds 1a-1e were examined as catalytic oxidants for the oxidation of 1- and 2-hexanol, to hexanal and 2-hexanone respectively, with the co-oxidants H₂O₂, NaOCl, t-BuOOH, N-methylmorpholine-N-oxide, Me₃NO, O₂, C₆H₅IO and Bu₄NIO₄. Compounds 1c and 1d were further studied in the catalytic oxidation of a wide range of alcohols (using N-methylmorpholine-N-oxide and Bu₄NIO₄ as co-oxidants) and found to give the corresponding aldehydes or ketones very selectively, with no attack on sensitive linkages or functional groups and no over-oxidation products. Compounds 1c and 1d were also supported on poly(4-vinylpyridine) to give active catalysts.

Full text of DOI:10.1016/j.molcata.2007.02.049

Journal of Molecular Catalysis A: Chemical 271 (2007) 277–283
A comparative study of the ruthenium(VI)dioxocarboxylato salts,
[PPh4][RuO2(OCOR)Cl2] (R = CH3, CF3, C6H5, C6F5, C5H11),
in the oxidation of alcohols
Holger B. Friedrich∗, Vikash Gokul
School of Chemistry, University of KwaZulu-Natal, Durban 4041, South Africa
Received 1 February 2007; accepted 27 February 2007
Available online 3 March 2007
Abstract
The compounds [PPh4][Ru(O)2(OCOR)Cl2] (R = CH3 1a, CF3 1b, C6H5 1c, C6F5 1d, C5H11 1e) were prepared and fully characterised. The
fluorinated compounds 1b and 1d were obtained in significantly higher yields than their protonated analogues 1a and 1c and compound 1b was found
to be a clearly superior stoichiometric oxidant to compound 1a. The compounds 1a–1e were examined as catalytic oxidants for the oxidation of 1-
and 2-hexanol, to hexanal and 2-hexanone respectively, with the co-oxidants H2O2, NaOCl, t-BuOOH, N-methylmorpholine-N-oxide, Me3NO, O2,
C6H5IO and Bu4NIO4. Compounds 1c and 1d were further studied in the catalytic oxidation of a wide range of alcohols (using N-methylmorpholine-
N-oxide and Bu4NIO4 as co-oxidants) and found to give the corresponding aldehydes or ketones very selectively, with no attack on sensitive linkages
or functional groups and no over-oxidation products. Compounds 1c and 1d were also supported on poly(4-vinylpyridine) to give active catalysts.
© 2007 Elsevier B.V. All rights reserved.
Keywords: Ruthenium; Alcohol; Oxidation; Fluorinated ligands
1. Introduction
compound was reported, and showed a distorted octahedral
structure with the acetato and cis-dioxo ligands in one plane
[5].
A number of reports of ruthenium(II) and ruthenium(III) car-
boxylato compounds can be found in the literature and they
mostly exist as dimers or trimers [1]. Reports on monomeric
ruthenium monocarboxylate compounds are, however, not
common and very few ruthenium oxo carboxylates, where
ruthenium is in a high oxidation state, are known. Perier
et al. prepared the complexes trans-[RuO2(py)2(OCOR)2]
(R = CH3, CH3CH2, CH3CH2CH2, (CH3)2CH2, C6H5) and
reported the crystal structure of the acetato derivative [2,3].
These compounds were shown to be unselective oxidants
of organic substrates. Griffith and Jolliffe have reported
the compounds [Ru(O)2(OCOR)Cl2]− (R = CH3, CH3CH2,
CH3CH2CH2, CHF2) and found these to be effective two elec-
tron oxidants in catalysing the oxidation of alcohols to aldehydes
or ketones, sulphides to suphoxides and triphenylphosphine to
triphenylphosphine oxide, using N-methylmorpholine-N-oxide
(NMO) as co-oxidant [4]. The crystal structure of the acetato
We now report on the synthesis and comparative properties of
the series of compounds [PPh4][Ru(O)2(OCOR)Cl2] (R = CH3
1a, CF3 1b, C6H5 1c, C6F5 1d, C5H11 1e), where compounds
1b–1e are new. We were particularly interested in comparing
the fluorinated compounds 1b and 1d with the protonated com-
pounds 1a and 1c. The oxidation chemistry of these compounds
with respect to alcohols was examined in detail, since high valent
ruthenium compounds have been shown to be very effective for
this kind of chemistry [6–8].
2. Experimental
2.1. Techniques
IR spectra were recorded on
a Nicolet 5DX FT
(4000–400 cm−1) and a Pye Unicam SP3-300 (4000–200 cm−1
)
spectrophotometer as KBr disks. Gas-chromatic analyses were
carried out on either a Pye Unicam GCD, a Fisons GC 8000 or
a Perkin-Elmer Autosystem XL chromatograph, all fitted with
flame ionisation detectors.
∗
Corresponding author. Tel.: +27 31 2603107; fax: +27 31 2603091.
E-mail address: friedric@ukzn.ac.za (H.B. Friedrich).
1381-1169/$ – see front matter © 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2007.02.049

278
H.B. Friedrich, V. Gokul / Journal of Molecular Catalysis A: Chemical 271 (2007) 277–283
All stoichiometric and catalytic oxidations were carried out
dissolved in CH3CN (15 mL) at 0 ◦C whilst stirring. The RuO4
solution (10 mL, 1.4 mmol) was added carefully at this tem-
perature before sealing the flask and stirring overnight in the
ice-bath. The resulting green solution was concentrated under
reduced pressure (using a Teflon membrane vacuum pump) until
precipitation occurred. The precipitate was filtered off, washed
with drops of cold water and dried over silica gel under vac-
uum. Further concentration of the filtrate gave a second crop of
sample, which was worked up as before. After drying, the dark
green products were stored in a dessicator in the fridge. Yields
are reported in Table 1. Elemental analyses, found (calculated):
1a C, 51.31 (51.84) H, 3.70 (3.85); 1c C, 56.24 (56.04) H, 3.64
(3.79); 1d C, 50.39 (49.35) H, 2.72 (2.67).
in Schlenk tubes, under N2 (to prevent any possible oxidation
by air) in the dark (to prevent free radical oxidation reactions
initiated by UV radiation). Reactions were monitored by GC,
using either packed or capillary columns. Internal standards
used were iso-butylmethacrylate (Acros), 2-ethoxyethyl acetate
or n-hexane (Lab Scan Analytical Services) and were chosen to
achieve base-line separation on the GC traces. All conversion
(yield) percentages represent the average of at least three runs.
2.2. Materials
The co-oxidants H2O2 (30% (m/v), Saarchem), NaOCl
(15% (m/v), Associated Chemical Enterprises), t-BuOOH (70%
(m/v), Aldrich), Me3NO (Aldrich), O2, N-methylmorpholine-
N-oxide (Aldrich), and tetrabutylammonium periodate (Acros)
were used as supplied. Iodosyl benzene was prepared accord-
ing to a literature procedure and stored under N2 in the
fridge [9]. Tetraphenylphosphonium chloride (98%, Merck),
glacial acetic acid, trifluoroacetic acid, benzoic acid, pentafluo-
robenzoic acid (all Aldrich, 99%) and hexanoic acid (Acros,
98%) were used as supplied. The substrates 1-hexene, 1-
hexanol (both Acros Organics), 2-hexanol (Aldrich), cinnamyl
alcohol (Acros Organics), cinnamyl chloride (Aldrich), geran-
iol (Acros Organics), crotyl alcohol (Aldrich), cyclohexanol
(BDH Chemicals), furfuryl alcohol (H&W Fine Chemicals)
and 4-nitrobenzylalcohol (Acros) were obtained commercially.
Authentic standards of each product, hexanal (Aldrich), 2-
hexanone (Aldrich), cinnamyl aldehyde (Acros Organics),
croton aldehyde (Acros Organics), cyclohexanone (Kleber
Chemicals), furfuraldehyde (BDH Chemicals), citral (BDH
Chemicals) and 4-nitrobenzaldehyde (Aldrich) were also
obtained commercially.
Compounds 1a, 1b and 1d could also be prepared at room
temperature though in substantially reduced yields. Compounds
1c and 1e did not form at room temperature.
2.3.2. Preparation of compounds 2c and 2d
Poly(4-vinylpyridine) (1 g) was stirred with CH2Cl2 (5 mL)
and compounds 1c or 1d added (100 mg). Stirring was contin-
ued for 5 h, after which the solvent was removed under reduced
pressure and the product dried for 5 h under vacuum. The dark
green supported catalysts were stored under N2. Selected bands
of infrared spectra (cm−1)—2c: 3023(w) ν(C–H)-pyridine ring,
2925(m) ν(CH2)-vinyl chain, 1598(s) ν(C C)-pyridine ring,
1414(s) ␯(C N)-pyridine ring, 822(s) νasym(C H)-vinyl chain,
682(m) ν(C H)-aromatic ring, 1107(m) ν(C O)-carboxylate,
875(w) νsym(Ru O); 2d: 3028(w) ν(C H)-pyridine ring,
2924(m) ␯(CH2)-vinyl chain, 1596(s) ν(C C)-pyridine ring,
1414(s) ν(C N)-pyridine ring, 821(s) νasym(C H)-vinyl chain,
684(m) ν(C H)-aromatic ring, 1104(m) ν(C O)-carboxylate,
879(w) νsym(Ru O).
RuO4 was prepared from RuCl3 (1.54 g, 5.88 mmol) and
sodium metaperiodate (5.50 g, 25.7 mmol) in water (40 mL) and
extracted in CCl4 (40 mL) as reported previously [10,11] and
stored in a separating funnel. A solution of sodium metaperi-
odate (1.0 g) in water (10 mL) was layered above the RuO4 in
CCl4 solution, which allowed this solution be stored for many
weeks without degradation.
2.4. Oxidation reactions
2.4.1. Stoichiometric oxidation of 1- and 2-hexanol with 1a
and 1b
˚
Molecular sieves 4 A (180 mg), CH2Cl2 (6 mL), hexanol
(0.5 mmol), the internal standard (0.5 mmol) and the oxidant
(1a or 1b, 0.5 mmol) were added under N2 to a Schlenk tube.
The solution was stirred under N2 at room temperature and the
reaction monitored by GC.
2.3. Catalyst preparation
2.3.1. Preparation of the compounds 1a–1e
These compounds were prepared by a simplification of the
literature procedure for 1a [4]. Thus, Ph4PCl (2.8 mmol) was
2.4.2. Homogeneous catalytic oxidations
Molecular sieves 4 A (180 mg), CH2Cl2 (6 mL), the alco-
hol (0.5 mmol), the internal standard (0.5 mmol), the co-oxidant
˚
Table 1
Summary of the analytical results for the complexes (1a–1e)
b
Dioxoruthenium(VI) complexes
Yield (%) Melting pointa (◦C) Infrared spectra, selected bands (cm−1
)
[PPh4][RuO2(OCOCH3)Cl2] (1a)
[PPh4][RuO2(OCOCF3)Cl2] (1b)
[PPh4][RuO2(OCOC6H5)Cl2] (1c)
[PPh4][RuO2(OCOC6F5)Cl2] (1d)
[PPh4][RuO2(OCOC5H11Cl2] (1e)
26
51
28
81
43
132
125
104
95
886(w) νsym(Ru O); 864(vs) νasym (Ru O); 1508(s) νasym(O
881(s) νsym(Ru O); 862(m) νasym (Ru O); 1516(m) νasym(O
884(w) νsym(Ru O); 859(s) νasym(Ru O); 1538(m) νasym(O
880(m) νsym(Ru=O); 855(w) νasym(Ru O); 1519(s) νasym(O
882(w) νsym(Ru O); 859(m) νasym(Ru O); 1504(w) νasym(O
C
C
C
C
C
O); 334(s) ν(Ru Cl)
O); 335(s) ν(Ru Cl)
O); 322(s) ν(Ru-Cl)
O); 318(s) ν(Ru Cl)
O); 320(s) ν(Ru Cl)
106
a
Decomposed.
vs: very strong, s: strong, m: medium, and w: weak.
b

H.B. Friedrich, V. Gokul / Journal of Molecular Catalysis A: Chemical 271 (2007) 277–283
279
(0.75 mmol) and the catalyst (20 mg) were added under N2 to a
Schlenk tube. The solution was stirred under N2 at room tem-
perature and the reaction monitored by GC.
Turnover limits were investigated by doubling the alcohol
and co-oxidant amounts (to 1 and 0.5 mmol, respectively, and
halving the amount of catalyst (10 mg).
Table 1 also shows infrared data for the compounds 1.
The characteristic νsym(Ru O) stretch is seen in the range
880–886 cm−1 and the νasym(Ru O) stretch is seen in the range
855–864 cm−1 forallthecompounds. The νsym(Ru O)stretches
in the fluorinated compounds 1b and 1d are relatively stronger
than those of their hydrogenated analogues and the converse is
true for the νasym(Ru O) stretches. Again, this may be related
to the different charge distributions as discussed previously.
Further infrared bands are difficult to discern beneath bands
due to [PPh4]+, however, by comparing the infrared spectrum
of the salt [PPh4]Cl with those of the compounds 1, distinct
peaks can be identified apart from the peaks due to [PPh4]Cl.
Thus, slightly broad peaks between 1500 and 1540 cm−1 can
clearly be assigned to the carboxylate group (i.e. νasym(O–C–O))
[2–5,12,13]. The bands between 315 and 334 cm−1 are assigned
to Ru–Cl stretches. Noticeable is that the ruthenium-oxo and
ruthenium-chloride stretches of the fluorinated compounds 1b
and 1d are at lower wave numbers than their hydrogenated
equivalents.
2.4.3. Oxidation of cyclohexanol with 2c and 2d
The procedure was as in Section 2.4.2, except that 200 mg
of supported catalyst was used. Although no detailed leach tests
were carried out, as these were outside the scope of this investi-
gation, the solvent remained clear and colourless throughout the
reactions. Since the catalysts are dark green in colour, no visual
sign of leaching was thus observed.
2.4.4. Product isolation
These were carried out on scaled up reactions. Thus, e.g. for
˚
cyclohexanol, molecular sieves 4 A (360 mg), CH2Cl2 (6 mL),
cyclohexanol (9.7 mmol), NMO (14.6 mmol) and the catalyst
(389 mg) were added under N2 to a Schlenk tube. The solu-
tion was stirred under N2 at room temperature and the reaction
monitored by GC. After 72 h, the reaction mixture was filtered
to remove the molecular sieves. The green filtrate was passed
through a silica gel column, initially eluting with hexane, and
finally flushing the column with CH2Cl2. The clear fractions
were combined, the product isolated by distillation and charac-
terised by 1H and 13C NMR.
Compounds 1a and 1b were investigated as stoichiomet-
ric oxidants for the oxidation of 1-hexanol to hexanal. Both
compounds showed a fairly rapid initial reaction rate, with 1a
showing a 61% conversion and 1b an 81% conversion after
30 min. Compound 1a had converted 69% of the hexanol to
hexanal after 24 h, whilst 1b had converted 91% in 3 h. Clearly
the fluorinated compound is the better stoichiometric oxidant.
This observation matches expectations in that the IR data implies
weaker Ru O bonds in the fluoridated compound.
The compounds 1a–1e were then investigated as catalysts
for the oxidation of 1-hexanol and 2-hexanol to give hexanal
and 2-hexanone, respectively. Their reaction with 1-hexene was
also investigated. Reactions were monitored over 24 h and a
wide range of co-oxidants, namely H2O2, NaOCl, t-BuOOH,
Me3NO, O2, N-methylmorpholine-N-oxide (NMO), iodosyl
benzene (PhIO) and tetrabutylammonium periodate (Bu4NIO4)
was examined. The results are shown in Tables 2 and 3. Since
a fixed mass of catalyst (20 mg) was used for each reaction, the
catalytic turnovers are included in parentheses to allow molar
comparisons. The reaction of the above co-oxidants with 1-
hexanol, 2-hexanol and 1-hexene was also investigated in the
absence of the catalysts to determine to what degree the co-
oxidants acted as stoichiometric oxidants. These results are also
shown in Tables 2 and 3. Thus, H2O2, Me3NO and O2 did
not oxidise 1-hexanol on their own, NaOCl showed a very low
conversion, t-BuOOH, NMO and PhIO gave approximately the
same conversion of 15, 18 and 20%, respectively, after 24 h,
whilst Bu4NIO4 was found to be rather reactive with a 65%
alcohol conversion. All the co-oxidants, except Me3NO and
Bu4NIO4, showed no oxidative activity with the secondary alco-
hol 2-hexanol. Conversions for the latter two were low.
3. Results and discussion
The compounds 1a–1e were prepared by the addition of a
solution of RuO4 to a solution of PPh4Cl and the respective
carboxylic acid in acetonitrile at 0 ◦C (Eq. (1)):
RuO4 + PPh4Cl + RCOOH → [PPh4] [Ru(O)2(OCOR)Cl2]
(R = CF3 1b, C6H5 1c, C6F5 1d, C5H11 1e)
(1)
It was noticed that the fluorinated compounds, 1b and 1d,
were obtained in a much higher yield (two- to three-fold) than
their respective hydrogenated analogues, 1a and 1c (Table 1).
This may be due to the different relative charge distributions
within the compounds caused by the significant differences in
theelectronegativitiesofthefluorineandhydrogenatoms(which
have approximately the same size) that constitute part of the
ligands in these compounds (Fig. 1). Also apparent is that the flu-
orinated compounds decompose at slightly lower temperatures
than their hydrogenated analogues.
None of the co-oxidants or any of the catalysts 1a–1e showed
any activity with respect to 1-hexene. This suggests that the
catalysts do not attack double bonds.
Apart from O2, all the co-oxidants examined were active
for the above reactions. No specific trend in reactivity could
be observed among the catalysts 1a–1e. The effectiveness of
the co-oxidants appears to be catalyst specific and the rela-
Fig. 1. Relative charge distributions for compounds 1a and 1b.

280
H.B. Friedrich, V. Gokul / Journal of Molecular Catalysis A: Chemical 271 (2007) 277–283
Table 2
Percentage conversion (yield) of 1-hexanol to hexanal by the compounds 1a–1e with various co-oxidants
a
Complex
Reaction time
H2O2
NaOCl
t-BuOOH
NMO
Me3NO
O2
C6H5IO
Bu4NIO4
1a
30 min
24 h
53(8)
56(8)
64(10)
83(13)
59(9)
64(10)
100(15)
100(15)
28(4)
60(9)
0
0
51(8)
67(10)
76(11)
100(15)
1b
30 min
24 h
55(9)
51(9)
74(13)
91(15)
37(6)
79(13)
87(13)
89(13)
47(8)
60(10)
8(1)
10(2)
70(12)
70(12)
96(16)
100(17)
1c
30 min
24 h
60(10)
73(12)
72(12)
77(13)
60(10)
66(11)
100(17)
100(17)
20(3)
81(14)
0
9(2)
90(15)
100(17)
100(17)
100(17)
1d
30 min
24 h
30(6)
36(7)
52(10)
71(14)
62(12)
67(13)
83(16)
97(18)
34(7)
58(11)
0
5(1)
44(8)
62(12)
100(19)
100(19)
1e
30 min
24 h
40(7)
53(9)
75(13)
80(14)
45(8)
49(8)
93(16)
96(16)
26(4)
62(11)
0
52(9)
76(13)
100(17)
100(17)
29(5)
None
30 min
3 h
24 h
0
0
0
0
1
4
0
10
18
0
6
15
0
0
0
0
0
0
0
0
20
54
54
65
a
Turnovers are shown in parentheses.
Table 3
Percentage conversion (yield) of 2-hexanol to 2-hexanone by the compounds 1a–1e with various co-oxidants
a
Complex
Reaction time
H2O2
NaOCl
t-BuOOH
NMO
Me3NO
O2
C6H5IO
Bu4NIO4
1a
30 min
24 h
14(2)
18(3)
19(3)
88(13)
23(4)
25(4)
62(9)
63(10)
0
0
0
58(9)
73(11)
80(12)
100(15)
35(5)
1b
1c
1d
1e
30 min
24 h
4(1)
4(1)
31(5)
100(17)
26(4)
29(5)
73(12)
63(10)
33(6)
37(6)
0
0
44(8)
63(11)
31(5)
55(9)
30 min
24 h
23(4)
25(4)
47(8)
38(7)
19(3)
19(3)
100(17)
100(17)
11(2)
38(7)
0
0
45(8)
78(13)
76(13)
76(13)
30 min
24 h
8(2)
10(2)
36(7)
69(13)
17(3)
33(6)
52(10)
67(13)
26(5)
37(7)
0
0
50(10)
77(15)
43(8)
56(11)
30 min
24 h
3(1)
14(2)
62(11)
72(12)
15(3)
11(2)
69(12)
72(12)
34(6)
45(8)
0
0
88(15)
100(17)
46(8)
100(17)
None
30 min
24 h
0
0
0
0
0
0
0
0
0
14
0
0
0
0
7
16
a
Turnovers are shown in parentheses.
tive order, in terms of performance, of the catalysts appear to
depend on which co-oxidant is used. Catalyst 1a works best
with NMO and Bu4NIO4, 1b works best with NaOCl, NMO and
Bu4NIO4, 1c with NMO, PhIO and Bu4NIO4, 1d with NMO and
Bu4NIO4, and 1e with NMO and Bu4NIO4. Clearly, NMO and
Bu4NIO4 are the best overall general co-oxidants, whilst NaOCl
and PhIO also perform well. Whereas Bu4NIO4 was noted to be
very active as a stoichiometric oxidant for the primary alco-
hol, the reaction rates with the catalysts are significantly higher
and the reactions go to completion, thus strongly suggesting
that Bu4NIO4 functions mainly as a co-oxidant in the reactions
involving compounds 1a–1e. In most cases the catalytic reac-
tions with the primary alcohol were faster than those with the
secondary alcohol, with the exception of 1e with PhIO and 1c
with NMO, the latter appearing to be a particularly effective
combination.
examined with 1-hexanol and NMO and Bu4NIO4 as co-oxidant
(Table 4). In the oxidation of 1-hexanol, all three of the catalysts
showed an initial high rate of conversion with high turnovers
when NMO was used as co-oxidant. The reaction rate levelled
off rapidly after 3 h. Catalyst 1c achieved a maximum of 52
Table 4
Summary of the catalytic reactions involving double quantities of 1-hexanol and
half quantities of catalyst, showing percentage conversion/yield of hexanal
Complex
Reaction time
NMOa
Bu4NIO4
1b
30 min
3 h
24 h
77(52)
85(58)
87(59)
40(28)
93(64)
90(61)
1c
30 min
3 h
24 h
65(44)
70(48)
78(52)
50(36)
80(56)
100(68)
Experiments were carried out to determine the maximum
turnover limits of some of these catalysts. Thus, the quantity
of catalyst used was halved to 10 mg and the quantity of alco-
hol and co-oxidant was doubled. Compounds 1b, 1c and 1e were
1e
30 min
3 h
24 h
75(51)
82(56)
84(56)
52(36)
88(60)
86(59)
a
Turnovers are shown in parentheses.

H.B. Friedrich, V. Gokul / Journal of Molecular Catalysis A: Chemical 271 (2007) 277–283
281
Table 5
unsaturated aromatic chloride (cinnamyl chloride), unsaturated
primary alcohols (geraniol and crotyl alcohol), a saturated
cyclic alcohol (cyclohexanol), a heteroatom containing aro-
matic alcohol (furfuryl alcohol) and finally a nitro-substituted
alcohol (4-nitrobenzylalcohol) (Table 6). As previously, the co-
oxidants themselves were also reacted with the substrates in the
absence of catalyst to determine the degree of the direct reaction
(Table 6).
Both 1c and 1d are efficient oxidants for the oxidation of
cinnamyl alcohol, giving almost total conversion to cinnamyl
aldehyde within 30 min of the start of the reaction. The double
bond is not attacked and no other products are observed. Both co-
oxidants on their own (especially NMO) give high conversions
too, but at a slower rate and the reactions involving 1c and 1d
are thus mostly catalytic.
Summary of the catalytic reactions involving double quantities of 2-hexanol and
half quantities of catalyst, showing percentage conversion (yield) of 2-hexanone
Complex
Reaction time
NMOa
Bu4NIO4
NaOCl
1a
30 min
3 h
24 h
–
–
–
19(11)
38(23)
60(36)
–
–
–
1b
1c
30 min
3 h
24 h
–
–
–
–
–
–
14(10)
21(14)
25(17)
30 min
3 h
24 h
69(47)
100(68)
100(68)
–
–
–
–
–
–
1e
30 min
3 h
24 h
–
–
–
30(20)
35(24)
54(37)
–
–
–
The oxidation of cinnamyl chloride with 1c and 1d gave
turnovers of one or less. These results are equal to or lower
than when the co-oxidants are used alone. This suggests that
the catalysts are inactive towards the oxidation of chloride in
organic substrates. When NMO was used on its own, 30% of cin-
namyl chloride was oxidised to cinnamyl aldehyde. This agrees
with the literature which states that NMO reacts with activated
halides [14,15]. The absence of a substantial yield of aldehyde
in the reactions including the catalysts and NMO suggests that
the catalysts inhibit the direct oxidation of organic halides by
NMO.
Both 1c and 1d catalytically converted crotyl alcohol to
croton aldehyde. The double bond was not attacked, nor was
over-oxidation to the carboxylic acid observed. The co-oxidant
Bu4NIO4 was best for this reaction and 1c the better catalyst.
The co-oxidants alone showed minimum reactivity.
Both 1c and 1d, combined with either NMO or Bu4NIO4,
convert cyclohexanol to cyclohexanone rapidly and essen-
tially quantitatively. No rupture of the cyclohexanol ring was
observed. The co-oxidants alone showed low reactivity (<15%
over 24 h).
Both 1c and 1d gave furfuraldehyde from furfuryl alcohol in
98% yield within 24 h with Bu4NIO4 as co-oxidant, although
1c initially reacted more rapidly. In contrast 1d was more effec-
tive than 1c with NMO, although NMO is the poorer co-oxidant
for this transformation. NMO is also fairly reactive on its own,
unlike Bu4NIO4, giving the aldehyde in 21% yield after 30 min.
No further products were observed for the above reactions,
implying that the oxygen within the structure of the aromatic
ring and the double bonds were not attacked. This suggests that
the catalysts show specificity for the oxidation of the alcohol
group.
a
Turnovers are shown in parentheses.
turnovers in 24 h, which equated to the number of turnovers
achieved by the other two catalysts within 30 min. Compound
1b now functioned as the best catalyst of those examined when
NMO was used as co-oxidant, whilst it had appeared to be the
poorest catalyst for this transformation under the conditions
whose results are shown in Table 2. A possible reason for this
behaviour is that although compound 1b showed a slower rate of
conversion of 1-hexanol to hexanal relative to the other catalysts,
it has the ability to undergo more turnovers compared to the other
compounds. This may be due to the structural stability of the flu-
orinatedcompoundrelativetotheothercompounds, whichcould
increase the lifetime of the fluorinated compound under reaction
conditions and consequently allow for easier re-oxidation of the
reduced catalyst to its active form.
A different trend was observed for the oxidation reactions
involving Bu4NIO4, with the initial reaction rate being slower
than with NMO. Here the reaction involving 1b was initially
the slowest, but it had achieved the highest turnover of the three
compounds studied after 3 h. Over time, compound 1c was the
most effective catalyst, however.
Similar experiments were carried out using the best
catalyst/co-oxidant combinations for the oxidation of 2-hexanol
to 2-hexanone from Table 3. These results are shown in Table 5.
Clearly compound 1c with NMO is the best combination of those
examined under these conditions.
Although no specific trend in activity of the catalysts 1a–1e
could be determined from Tables 2 and 3, it could generally
be concluded that compound 1c functioned most effectively
as a catalyst and that compound 1d was the least effective. It
was consequently decided to examine these two compounds
as catalysts for the oxidation of a range of substrates using
the co-oxidants NMO and Bu4NIO4, the latter being the most
effective general co-oxidants from the range studied. The sub-
strates were chosen to represent a wide spectrum of organic
alcohols and further included an unsaturated chloride to deter-
mineiftheoxidativesystemattackedhalogensbondedtoorganic
compounds. The range of substrates included the unsaturated
aromatic primary alcohol (cinnamyl alcohol), its equivalent
Both catalysts were very effective in the oxidation of geraniol
to citral with either co-oxidant, achieving, effectively, a quanti-
tative conversion with no other oxidation or cleavage products
observed. The co-oxidants on their own show conversions of
28% (NMO) and 10% (Bu4NIO4), respectively.
Finally, both catalysts with either co-oxidant convert 4-
nitrobenzyl alcohol to 4-nitrobenzyl aldehyde in effectively
100% yield. For this alcohol, however, both co-oxidants (espe-
cially NMO) are rather active on their own. The nitro functional
group is not attacked.

282
H.B. Friedrich, V. Gokul / Journal of Molecular Catalysis A: Chemical 271 (2007) 277–283
Table 6
Summary of the percentage conversions (yields) of substrates to their corresponding products with compounds 1c and 1d and the co-oxidants NMO and Bu4NIO4
b
Substrate
Producta
Reaction time
1c
1d
NMOb
Bu4NIO4
NMOc
Bu4NIO4
NMO
Bu4NIO4
Cinnamyl alcohol
Cinnamyl chloride
A
A
30 min
3 h
99(17)
100(17)
100(17)
100(17)
100(19)
100(19)
100(19)
100(19)
82
77
13
25
30 min
3 h
24 h
3(<1)
2(<1)
2(<1)
5(<1)
5(<1)
7(1)
3(<1)
1(<1)
2(<1)
3(<1)
5(1)
9(2)
18
16
30
3
8
30
Crotyl alcohol
A
30 min
3 h
24 h
51(9)
56(10)
57(10)
53(9)
67(11)
100(17)
45(9)
54(10)
57(11)
20(4)
47(9)
74(14)
0
3
3
6
4
3
Cyclohexanol
K
A
30 min
3 h
100(17)
100(17)
70(11)
100(17)
98(19)
100(19)
48(9)
98(19)
16
15
6
13
Furfuryl alcohol
30 min
3 h
24 h
50(9)
55(9)
56(10)
89(15)
93(17)
98(17)
52(10)
64(12)
67(13)
54(10)
77(15)
98(19)
21
23
22
0
5
3
Geraniol
A
A
30 min
3 h
92(16)
97(17)
81(14)
100(17)
98(19)
100(19)
57(11)
100(19)
0
2
4
5
4-Nitrobenzyl alcohol
30 min
3 h
99(17)
100(17)
96(16)
100(17)
95(18)
100(19)
99(19)
100(19)
70
65
25
27
a
A: corresponding aldehyde product; K: corresponding ketone product.
In absence of catalyst.
Turnovers are shown in parentheses.
b
c
It can be concluded from the above study that 1c and 1d are
reducethepossibilityof leaching. Tolueneisalsoknowntocause
swelling of the polymer beads, whilst CH2Cl2 does not [19,22].
The oxidation of cyclohexanol to cyclohexanone using
the supported Ru(VI) catalysts 2c and 2d was investigated
using NMO and Bu4NIO4 as co-oxidants. The same quantity
of ruthenium was added as for the homogeneous reactions.
Cyclohexanol was chosen because compounds 1c and 1d had
previously shown 99–100% conversions with the same co-
oxidants. TheresultsareshowninTable7. Mediocreconversions
are observed in CH2Cl2, especially when compared to conver-
sions obtained by the co-oxidants alone (Table 6). Both catalysts
showed a slow rate of conversion of cyclohexanol to cyclohex-
anone in toluene at the beginning of the reaction, however, 2c
with NMO achieved a 40% yield of cyclohexanone after 24 h.
Conversion with Bu4NIO4 was significantly lower as were the
conversions achieved by 2d with either co-oxidant. However,
since almost no cyclohexanol is oxidised by the co-oxidants on
efficient catalysts for a range of alcohol substrates. No sensi-
tive linkages like double bonds, aromatic and saturated rings are
cleaved, nor are heteroatoms or chlorides attacked. The inertness
to double bonds was confirmed by the observation that these
catalysts are inert to hexene. No significant difference in cat-
alytic activity was noted between the fluorinated and protonated
complexes 1c and 1d.
The yields referred to above are GC yields, isolated yields
are a few percent lower due to product loss on work-up in
our hands. The gravimetric determinations were done on scaled
up reactions and the product was further identified by NMR
spectroscopy (versus authentic samples).
Some preliminary studies were made to see if the catalysts
could be supported on an insoluble support. Much effort has
been expended over many years to combine the advantages
of homogeneous and heterogeneous catalysts in this manner
[16–20]. Compounds 1c and 1d were consequently supported
on poly(4-vinylpyridine) (PVP) in the ratio of 1 mg catalyst to
10 mg PVP to give catalysts 2c and 2d. The IR data for the
supported compounds shows a peak at 1105 cm−1, assigned to
the ν(C O) stretch of the carboxylate group. A weak, though
distinct, νassym(Ru O) stretch is observed at approximately
878 cm−1. Both these stretches are shifted to lower wavenum-
bers when compared to those of the unsupported compounds.
Although the interaction between the supported Ru(VI) com-
pounds and PVP was not studied, it is believed to involve an
ionic interaction [21].
Table 7
Percentage conversion (yield) of cyclohexanol to cyclohexanone by the PVP-
supported catalysts 2c and 2d in dichloromethane and toluene
Solvent
Reaction time 2c
2d
NMOa Bu4NIO4 NMO Bu4NIO4
Dichloromethane 30 min
24 h
14(2)
17(3)
29(5)
30(3)
22(4) 15(3)
22(4) 17(3)
Toluene
30 min
3 h
4(<1)
6(1)
3(<1)
6(1)
4(<1) 3(<1)
8(1)
5(1)
Reactions in two different solvents were examined,
dichloromethane and toluene. Toluene was chosen because the
compounds 1c and 1d do not dissolve in it and this should
24 h
40(7)
13(2)
20(3) 21(4)
a
Turnovers are shown in parentheses.

H.B. Friedrich, V. Gokul / Journal of Molecular Catalysis A: Chemical 271 (2007) 277–283
283
their own in toluene, both 2c and 2d appear to behave catalyt-
Acknowledgements
ically. The observed trend of a low rate of conversion at the
beginning of the oxidations in toluene suggests that a solvent
influence prevails. As the polymer swells active sites become
more accessible and thus the rate of oxidation increases [19].
This trend was not observed for the reactions carried out in
CH2Cl2, adding support to this proposal. Since the mole quan-
tity of supported material responsible for the above oxidations
was not known, direct comparisons to catalysts 1c and 1d are
difficult. The nature of the association between the catalyst and
support was not investigated further, since this was just a pre-
liminary study. It is conceivable that different co-oxidants or
different supports, such as ion exchange resins, could be more
effective.
We thank the NRF and URF for financial support, and NMT
Anodes (Pinetown) and Johnson-Matthey for loans of RuCl3.
References
[1] M. Schro¨der, T. Stephenson, in: G. Wilkinson (Ed.), Comprehensive Coor-
dination Chemistry, Pergamon Press, Oxford, Great Britain, 1987, pp.
425–430.
[2] S. Perrier, T.C. Lau, J.K. Kochi, Inorg. Chem. 29 (1990) 4190.
[3] T.C. Lau, J.K. Kochi, Chem. Commun. (1987) 798.
[4] W.P. Griffith, J.M. Jolliffe, J. Chem. Soc., Dalton Trans. (1992) 3483.
[5] W.P. Griffith, J.M. Jolliffe, S.V. Ley, D.J. Williams, Chem. Commun. (1990)
1219.
[6] H.B. Friedrich, Platinum Met. Rev. 43 (1999) 94.
[7] R.A. Sheldon, I.W.C.E. Arends, A. Dijksman, Catal. Today 57 (2000) 157.
[8] T. Naota, H. Takaya, S.-I. Murahashi, Chem. Rev. 98 (1998) 2599.
[9] H. Saltzman, J.G. Sharefkin, Organic Syntheses, vol. 5, John Wiley and
Sons, New York, USA, 1973, p. 658.
[10] H.B. Friedrich, E. Friedrich, V. Gokul, K.K. Kubheka, N. Rinder, S. Afr.
J. Chem. 53 (2000) 2.
[11] H.B. Friedrich, N. Singh, J. Mol. Catal. A 245 (2006) 266.
[12] D.S. Moore, S.D. Robinson, Inorg. Chem. 18 (1979) 2307.
[13] S.D. Robinson, M.F. Uttley, J. Chem. Soc., Dalton Trans. (1973) 1912.
[14] W.P. Griffith, J.M. Jolliffe, S.V. Ley, K.F. Springhorn, P.D. Tiffin, Synth.
Commun. 22 (1992) 1967.
4. Conclusion
In conclusion the fluoridated compounds can be obtained in
significantly higher yield than their protonated equivalents and
the former also appear to be the superior stoichiometric oxi-
dants, in keeping with expectations from the IR data. All of
these compounds are effective catalysts for the oxidation of 1-
and 2-hexanol to hexanal and 2-hexanone, respectively, but no
significant difference could be seen between the fluorinated and
protonated compounds under catalytic conditions. Compounds
1c and 1d were also found to very effectively and selectively
catalytically oxidise a range of alcohols without attacking other
sensitive functional groups or heteroatoms, but again no signifi-
cant difference is noted between the fluorinated and protonated
compounds under catalytic conditions. It seems likely that the
faster rate of oxidation by the fluorinated compounds is offset
by a slower rate of re-oxidation.
[15] S. Suzuki, T. Onishi, Y. Fujita, H. Misana, J. Otera, Bull. Chem. Soc. Jpn.
59 (1986) 3287.
[16] R. Lenz, S.V. Ley, J. Chem. Soc., Perkin Trans. I (1997) 3291.
[17] D.T. Thomson, Platinum Met. Rev. 43 (1999) 114.
[18] W.A. Herrmann, R.M. Kratzer, J. Blu¨mel, H.B. Friedrich, R.W. Fischer,
D.C. Apperley, J. Mink, O. Berkesi, J. Mol. Catal. A 120 (1997) 197.
[19] H.B. Friedrich, N. Singh, Catal. Lett. 110 (2006) 61.
[20] H.B. Friedrich, N. Singh, J. Mol. Catal. A 160 (2000) 461.
[21] J.M.J. Frechet, P. Darling, M. Farrall, J. Org. Chem. 46 (1981) 1728.
[22] C.P. Nicolaides, N.J. Coville, J. Organomet. Chem. 222 (1981) 285.