Selective activation of secondary C-H bonds by an iron catalyst: Insights into possibilities created by the use of a carboxyl-containing bipyridine ligand

Article abstract of DOI:10.1039/c3nj00656e

In this work, we report the discovery of a carboxyl-containing iron catalyst 1 (Fe^II-DCBPY, DCBPY = 2,2′-bipyridine-4,4′- dicarboxylic acid), which could activate the C-H bonds of cycloalkanes with high secondary (2°) C-H bond selectivity. A turnover number (TN) of 11.8 and a 30% yield (based on the H₂O₂ oxidant) were achieved during the catalytic oxidation of cyclohexane by 1 under irradiation with visible light. For the transformation of cycloalkanes and bicyclic decalins with both 2° and tertiary (3°) C-H bonds, 1 always preferred to oxidise the 2° C-H bonds to the corresponding ketone and alcohol products; the 2°/3° ratio ranged between 78/22 and >99/1 across 7 examples. ¹⁸O isotope labelling experiments, ESR experiments, a PPh₃ method and the catalase method were used to characterize the reaction process during the oxidation. The success of 1 showed that, in addition to using a bulky catalyst, high 2° C-H bond selectivity could also be achieved using a less bulky molecular iron complex as the catalyst.

Full text of DOI:10.1039/c3nj00656e

NJC
View Article Online
View Journal
PAPER
Selective activation of secondary C–H bonds by an iron
catalyst: insights into possibilities created by the use
of a carboxyl-containing bipyridine ligand†
Cite this: DOI: 10.1039/c3nj00656e
Shi Cheng, Jing Li, Xiaoxiao Yu, Chuncheng Chen, Hongwei Ji, Wanhong Ma* and
Jincai Zhao*
In this work, we report the discovery of a carboxyl-containing iron catalyst 1 (FeII-DCBPY, DCBPY =
2,20-bipyridine-4,40-dicarboxylic acid), which could activate the C–H bonds of cycloalkanes with high
secondary (21) C–H bond selectivity. A turnover number (TN) of 11.8 and a 30% yield (based on the
H2O2 oxidant) were achieved during the catalytic oxidation of cyclohexane by 1 under irradiation with
visible light. For the transformation of cycloalkanes and bicyclic decalins with both 21 and tertiary (31)
C–H bonds, 1 always preferred to oxidise the 21 C–H bonds to the corresponding ketone and alcohol
products; the 21/31 ratio ranged between 78/22 and >99/1 across 7 examples. 18O isotope labelling
experiments, ESR experiments, a PPh3 method and the catalase method were used to characterize the
reaction process during the oxidation. The success of 1 showed that, in addition to using a bulky
catalyst, high 21 C–H bond selectivity could also be achieved using a less bulky molecular iron complex
as the catalyst.
Received (in Porto Alegre, Brazil)
18th June 2013,
Accepted 2nd August 2013
DOI: 10.1039/c3nj00656e
www.rsc.org/njc
binding pockets of the enzymes.7 Very recently, a synthetic
Introduction
4ꢀ
bulky polyoxometalate catalyst [g-HPV2W10O40
]
was reported
The functional oxidation of inert paraffin carbon–hydrogen (C–H)
bonds has recently received increasing attention because the
process can generate complex molecules from simpler ones.1
However, the selectivity of the C–H bond activation is very difficult
to control when multiple non-equivalent C–H bonds are present
in the substrate molecule. In particular, the remaining challenge
is the selective cleavage of 21 C–H bonds before 31 C–H bonds
with smaller bond dissociation energies (BDE).2 Despite the wide
application of catalytically regiospecific oxidations of inactive C–H
bonds, the majority of transition metal catalysts preferentially
activate 31 C–H bonds.1–5 This preference facilitates the abstrac-
tion of the hydrogen atom on the 31 C–H site with active oxidants,
such as iron–oxo species, followed by the rebounding of an alkyl
radical to form alcohols as products. However, nature has evolved
many iron-containing enzymes, such as the soluble methane
monooxygenases (s-MMO), which achieve the required 21 C–H
bond oxidation, even in the presence of weaker 31 C–H bonds.6
This selectivity could be attributed to the non-covalent or charged
lock-and-key interactions between the substrates and the intricate
to perform highly regioselective oxidations on the 21 C–H
bonds of cycloalkanes with H2O2.8 Inspired by this, we
attempted to develop a less bulky molecular iron catalyst using
carboxyl-containing ligands (see Scheme 1) and we expected 1
to realise the selective activation of 21 C–H bonds.
In previous disclosures, some iron complexes were used to
accelerate the catalytic oxidation of organic compounds
through irradiation with light.9,10 We therefore found that
using a ligand that was modified to include carboxylic groups
on the 4,40-positions of 2,20-bipyridine significantly improved
the reactivity of iron-complex. For example, in addition to
higher photocatalytic activity, 1 could facilitate the use of more
dioxygen in the oxidation with H2O2 than its parent complex 2
Key Laboratory of Photochemistry, Beijing National Laboratory for Molecular
Sciences, Institute of Chemistry, Chinese Academy of Sciences, 100190 Beijing,
China. E-mail: whma@iccas.ac.cn, jczhao@iccas.ac.cn
† Electronic supplementary information (ESI) available. See DOI: 10.1039/
c3nj00656e
Scheme 1 Three catalysts and their corresponding ligands.
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2013
New J. Chem.

View Article Online
Paper
NJC
(FeII–BPY, BPY = 2,20-bipyridine). Recently, a single crystal 11.8 after 20 hours (Table 1, entry 8), which was 1–2 orders of
structure of 1 was successfully determined by Li et al. (right magnitude greater than FeII (Table 1, entry 1), 2 (Table 1, entry 3)
side of Scheme 1, H atoms have been omitted for clarity);11
1
and 3 (Table 1, entry 5). Similar to the results previously reported
resembles the typical N,N,O-coordinated structures of many for the degradation of organic pollutants in water,9,17 visible
non-heme mononuclear iron enzymes.12 Its Fe–Ocarboxylate bond light irradiation significantly enhanced the TN observed for
lengths, which are 2.111(2) Å and 2.148(1) Å, are longer than the the cyclohexane transformation (Table 1, entries 8 and 9).
previously reported bond lengths between a carbonyl and iron Unfortunately, the A/K ratio generated by 1 was quite low and
(2.00–2.08 Å), as well as an amide carbonyl and iron (2.043– this catalytic oxidation did not remain in the alcohol stage.
2.047 Å).13 The Fe–Nbipyridine lengths (2.186(3) Å and 2.178(2) Å) However, the cyclohexanone formation was not caused by
are also longer than those found in 2, which has Fe–Naromatic
E
cyclohexanol over-oxidation or autoxidation in this case, which
1.98 Å,14 and tetradentate tripodal coordinated FeII–TPA was indicated by our control experiments; cyclohexanol, when
(3, TPA = tris-(2-pyridylmethyl)amine) with Fe–Npyridine lengths of used as the initial reactant, could not transform into the ketone
1.92–1.99 Å.15 The coordination of 1 is loose and is therefore over an identical reaction time. Therefore, cyclohexanone for-
envisioned to cover more space with its stereo-electronic effects mation should be attributed to the involvement of a long-lived
for selective C–H cleavage.
alkyl radical Rꢁ. This species would subsequently capture
dioxygen to form ROOH, which is commonly decomposed into
equal parts of alcohol and ketone. To confirm this pathway, two
independent experiments were carried out with 18O2 and H218O
to detect the origin of the O atom in the formed cyclohexanol.
Incorporation of the O-atom from H2O2 was calculated based
on two other values from H2O and O2. During the 18O labelling
experiments, the oxygen atom in the product alcohol was
primarily from O2 (E67% for entry 8, Table 1), while H2O2
contributed E32%. In addition, Ar purge experiments were
conducted to remove O2 from the system and subsequently the
product yield decreased precipitously; therefore, O2 is a participant
in the transformation of alkanes by 1.
Generally, alkyl peroxides (ROOH) readily decompose in
catalytic systems,18 but occasionally these molecules do not
decompose over time and subsequently accumulate at high
concentrations. In the injector of the G, the peroxides decom-
posed to form mixtures of products and led to incorrect results.
According to well-known methods18 we used Ph3P to treat the
reaction samples, as a test to reveal whether ROOH had
accumulated. Our results showed that the product distribution
was unchanged after treatment with Ph3P, which illustrated
that ROOH (if formed) had already decomposed before GC
analysis. In addition, the catalase method19 was utilised to
detect ROOH accumulation in the reaction that was catalysed
by 1. As expected, ROOH had not accumulated during the
transformation of the substrate (Fig. S4, ESI†). As an excellent
non-heme iron catalyst,15 3 had a TN of 3.2 and a high A/K ratio
of 5 in CH3CN (Table 1, entry 7), whereas its catalytic activity
was quite low in aqueous CH3CN solutions (Table 1, entries 5
and 6). In CH3CN, the oxygen atom of the product alcohol was
only E3% from O2 and E70% from H2O2 during the catalytic
Results and discussion
Catalytic and photocatalytic activity in the oxidative
transformation of cyclohexane
A typical oxidation of cyclohexane by H2O2 was first performed
in an aqueous CH3CN solution at room temperature (Table 1),
in which we evaluated the activity of the three iron catalysts
(Scheme 1) mainly by their turnover numbers (TN) (represented
by the products, including both cyclohexanol (A) and cyclo-
hexanone (K), per catalyst). As determined by GC, no products
were formed without H2O2 in either the absence or presence of
visible light irradiation for all the iron catalysts tested. To
obtain the best TN, both the reactant and mixed solvent ratios
needed to be optimised. With the optimal reactive conditions,
1 exhibited superior reaction properties and provided a TN of
Table 1 Cyclohexane oxidation by various iron catalystsa
O-atom of alcohol from
H2O2/H2O/O2 (%)
Ib
TN
0.5
A/Kc
d
Entry
Catalyst
1e
2e
3
4
5
FeII
FeII
2
2
3
+
0.31
0.31
0.56
0.32
0.52
0.36
5.0
57/0/43
57/0/43
49/0/51
70/0/30
—
ꢀ
0.5
0.7
0.4
0.1
0.1
3.2
11.8
1.4
+
ꢀ
+
6
3
3
1
1
ꢀ
ꢀ
+
ꢀ
—
7f
8g
9
70/27/3
32/1/67
55/1/44
0.71
0.54
a
Reaction conditions: 1.5 mM iron catalyst, 1.5 M alkane substrate and process with 3; this value was consistent with that reported by
Que and Chen.15 These preliminary results implied that 1
differed from 3 in both the process of the substrate activation
60 mM H2O2 in 2.5 mL CH3CN–H2O solution (CH3CN–H2O (v/v) =
b
60/40), 20 h reaction time. I = irradiation. + = visible irradiation
c
(l > 420 nm), ꢀ = without visible irradiation. A = cyclohexanol, K =
d
cyclohexanone. Incorporation of the O-atom from H2O2 was calcu- and the active oxidant species.
lated based on the two other values from H2O and O2, which were
detected through experiments with H216O2/H218O/16O2 and H216O2/
Anomalous 28/38 selectivity for transformations of cycloalkane
e
f
H216O/18O2, respectively. FeII
=
FeII(ClO4)2ꢂ6H2O. 1.5 mM iron
catalyst, 1.5 M alkane substrate and 15 mM H2O2 in 2.5 mL CH3CN When mono- and di-substituted methylcyclohexanes were
with 1000 equiv. H2O relative to 3. The quantum yield (+) based on
g
employed as substrates (see Table 2), 1 exhibited surprisingly
high 21/31 selectivity (78/22–92/8) for all of the oxidative products
(A + K). 1 exhibits a high selectivity for secondary products with
the irradiation at wavelength l = 520 nm was about + = 0.23%, which
was obtained by the previously described method.16 For the detailed
method, see the Experimental section.
c
New J. Chem.
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2013

View Article Online
NJC
Paper
Table 2 Selective oxidation of cycloalkanes by 1 under visible irradiationa
51/49 than the bulky polyoxometalate (21/31 = 15/85),8 which
indicated that the steric effect of adamantane was not the
parameter that controlled the regiospecific oxidation; instead,
the selectivity of 1 for 21 C–H bonds might be determined by
the catalyst structure.
Entry Substrate Products (selectivity)
1
TN 21/31b
6.3 92/8
7.2 78/22
Some classic examples have demonstrated that, in addition
to chemically stoichiometric oxidations with peracid and dioxirane,
transition-metal-catalysed systems without bulky ligands nearly
always exclusively favour the formation of 31 alcohols (see entries
4–11 of Table 3 and Table S2, ESI†).21 Iron systems with no ligands
(a Fenton system, entry 6 of Table 3), which generate ꢁOH radicals
as the main oxidant, display enhanced selectivity toward tertiary
sites, which indicates that the ꢁOH radicals did not afford such
specific 21 selectivity. Visible light irradiation had no effect on the
21/31 selectivity for alkanes; therefore, the selectivity was not
induced by visible light irradiation. Except for the use of the bulky
polyoxometalate catalyst [g-HPV2W10O40]4ꢀ,8 such clear selectivity
for the 21 C–H bond in the presence of weaker and more electron-
rich 31 C–H bonds could not be obtained with transition metal
catalysts until now. Catalyst 1 demonstrated high selectivity, which
2
3
6.5 91/9
4.8 51/49
4
a
Reaction conditions: 1.5 mM iron catalyst, 1.5 M alkane substrate and
60 mM H2O2 in 2.5 mL CH3CN–H2O solution (CH3CN–H2O (v/v) = 60/40),
b
20 h reaction time. 21/31 = (all 21 alcohol + ketone)/all 31 alcohols.
both cis- and trans-DMC (DMC = 1,2-dimethylcyclohexane); could only be obtained using carboxyl-based oxidants, such as
however, its 31 alcohol products could not maintain stereo- peracetic acid or m-CPBA.22
specificity, which was most likely due to the epimerisation of a
The effect of substrates with different C–H bonds on the
28/38 selectivity
long-lived radical Rꢁ or carbonium intermediate (cis/trans = 55/
45 for cis- and 56/44 for trans-) (entries 2 and 3 of Table 2).
Interestingly, some iron systems with enhanced 21 C–H selec- We investigated the photocatalytic oxidation of other cyclo-
tivity demonstrated selectivity for only cis-DMC or trans-DMC.20 alkanes with different 21 and 31 C–H bonds (Table 4). The
For example, the H2O2 based system using non-heme iron electron-donating, alkyl-substituted cycloalkanes (Table 4,
catalyst Fe(S,S-PDP),20a that was recently reported, gave a entries 1 and 2) and the bicyclic decalins (Table 4, entries 3
normal selectivity of 21/31 = 20/80 for cis-DMC but a reverse and 4), which had significantly different steric and stereo-
21/31 ratio of 67/33 for trans-DMC. Several iron systems exhib- electronic effects, afforded, without exception, 31 alcohols
ited
a normal 21/31 selectivity such as 3 (25/75) and and 21 ketones as the minor and major products, respectively;
[FeII(N4Py)]2+ (o1/99) during the oxidation of DMC (Table S2, the remarkably high 21/31 ratios of the products ranged from
ESI†).3,5 During the oxidation of adamantane, which had a 87/13 to >99/1. This specific selectivity for 21 C–H bonds in a wide
special steric effect that led to the oxidation of the 31 C–H range of substrates is very similar to that of the polyoxometalate
bond (entry 4 of Table 2). 1 afforded a higher 21/31 value of catalyst.8 Although these C–H bonds were non-equivalent towards
Table 3 Comparison of regioselectivity and product distribution in the different oxidations of ethylcyclohexane
Selectivitya (%)
A
B
C
D
E
F
G
Others
Entry
System
[21]/[31]
TN
6.3
Ref.
1
2
3
4
5
6
7
8
1–H2O2 (this work)
[(n-C4H9)4N]4[g-HPV2W10O40]–H2O2
Cytochrome P450b
Mn(TPP)–PhIOc
8
11
6
8
29
3
55
46
—
4
22
1
47
4
12
2
—
—
1 f
—
—
4 f
2 f
—
—
—
—
92/8
19
25
36
37
41
53
67
88
92
44
43
5
24
23
26
15
81/19
75/25
64/36
63/37
56/44
47/53
33/67
12/88
8/92
28.8
8
—
23
24
24
21e
25
26
23
27
28
2
17
2
20
3.4
3.0
Mn(TPP)–PhIOd
8
Fe(ClO4)3–H2O2
—
—
523
—
1.0
13
Cu(H3L)(NꢀCS)–TBHPe
RuCl3–IO4
—
4
—
—
—
4
—
—
20
—
—
—
7
—
8
6
—
—
—
9
10
11
CF3COOH
Methyl(trifluoromethyl)dioxirane
CrO2(OAc)2–H5IO6
—
—
>99
—
o1/99
a
b
c
Selectivity (%) = product (mol)/total products (mol) ꢃ 100. Obtained from rat-liver microsomes. Performed under anaerobic conditions.
d
e
f
Performed under aerobic conditions. L = N,N,N0,N0-tetrakis(2-hydroxyethyl)ethylenediamine. Hexahydrobenzylalcohol.
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2013
New J. Chem.

View Article Online
Paper
NJC
Table 4 Selective oxidation of cycloalkanes by 1 under visible light irradiationa
Entry
1
Substrate
Products (selectivity)
TN
4.7
21/31b
[Ketones]/[21 alcohols]
97/3 (86/14)
86/14
2
3
5.2
>99/1 (83/17)
90/10
6.8
7.6
96/4c (80/20)
87/13
93/7
4
87/13d (80/20)
a
Reaction conditions: 1.5 mM iron catalyst, 1.5 M alkane substrate and 60 mM H2O2 in 2.5 mL CH3CN–H2O solution (CH3CN–H2O (v/v) = 60/40),
b
20 h reaction time. 21/31 = (all 21 alcohol + ketone)/all 31 alcohols. Number in parentheses were theoretically calculated 21/31 value as statistic
average for all 21 and 31 C–H bonds. cis-9-Decalol/trans-9-decalol = 20/80. cis-9-Decalol/trans-9-decalol = 28/72.
c
d
Table 5 Oxidation of cis-decalin by 1, 2 and 3 as well as FeII a
oxidation,2,8,20 a simple statistical hypothesis was proposed: if
31 and 21 C–H bonds have equal and random reactivity (see the
statistic average values in parentheses in Table 4), then a
preference for the 21 C–H bond was displayed by 1 and the 31
selectivity was below its average value. Similar to the polyoxo-
metalate catalyst,8 which resulted in at least six 21 products (see
entry 2 of Table 3), 1 generated multiple products for all of the
substituted cyclohexanes. This result illustrated that these
catalysts were predisposed towards the cleavage of 21 C–H
bonds, but they could not distinguish between 21 C–H sites
with different chemo-environments. In addition, a higher
chemoselectivity toward ketones was observed for the four
substituted substrates (A/K = 14/86–7/93) than the unsubstituted
cyclohexanes (A/K = 42/58–35/65, see Table 1), which clearly
implies that the 1-mediated ketone formation is not a simple
over-oxidation or autoxidation.
Catalyst
Substrate
[21]/[31]
1
cis-Decalin
cis-Decalin
cis-Decalin
87/13
40/60
42/58
45/55
21/79
58/42
55/45
2
2b
3
cis-Decalin
cis-Decalin
Ethyl cyclohexane
tert-Butyl cyclohexane
FeII
FeII
FeII
a
Reaction conditions: 1.5 mM iron catalyst, 1.5 M alkane substrate and
b
60 mM H2O2 in 2.5 mL CH3CN–H2O solution. Solutions contains
excess acetic acid (2000 equiv. relative to the iron catalyst).
(conventional Fenton reaction) via the ꢁOH radical pathway
furnished a lower 21/31 selectivity (21/79); this selectivity was
Possible mechanisms for the creation of 28/38 selectivity
White and Chen proposed that an in situ-formed bulky oxidant close to the previously reported value,29 as well as being lower
could cause a reversal of the 21/31 selectivity during the oxida- than both 2 and 3. To confirm this result, two other substrates
tion of trans-DMC.20a However, the substrate-independent (ethyl cyclohexane and tert-butyl cyclohexane) were used to
selectivity of 1 should be attributed to the carboxyl groups that extend this comparison using the iron catalyst without any
are located on the BPY ligand. To elucidate the role of the ligand (Fenton system). The oxidation occurred preferentially at
carboxyl groups on 1, we compared 1 with 2 and 3 (both without the tertiary sites; the 21/31 selectivity was 58/42 and 55/45 for
carboxyl groups in the skeleton of the ligand) during the ethyl cyclohexane and tert-butyl cyclohexane, respectively.
oxidation of cis-decalin (Table 5) under identical reaction con- Furthermore, ESR experiments were conducted to detect the
ꢁ
ditions. As expected, 2 and 3 did not demonstrate an obviously presence of any OH radicals in both the 1-catalysed oxidation
high selectivity towards 21/31 C–H bonds. Excess acetic acid, and the Fenton system under identical conditions. In agree-
which was added to a reaction with 2, did not strongly affect the ment with the previous catalytic results,30 the trapped DMPO–
selectivity of 2, which excluded the possibility that the selectiv- ꢁOH radical adduct easily formed under the Fenton system and
ity of 1 was simply induced by the carboxylic acid. For 3, it is the signal exhibited was strong at room temperature. However,
well known that a high-valent iron–oxo species was generated no obvious DMPO–ꢁOH adduct was observed during the
to induce the homolytic cleavage of the weaker 31 C–H bond via 1-catalysed oxidation (Fig. S5, ESI†), which further indicated
H
atom abstraction and the subsequent rebounding of that 1, at least, did not generate a long-lived free radical
the short-lived alkyl radical to give the alcohol products. derivative of the substrate via an ꢁOH radical. The origin of
Simultaneously, the iron-catalysed oxidation without the ligand the special selectivity remained speculative and there are two
c
New J. Chem.
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2013

View Article Online
NJC
Paper
possible explanations for the origin of the selectivity. One to improve the accuracy of the experiments exhibited in
possibility was that a carbonium intermediate of the substrates Table 5.
might form over the course of the reaction; therefore, the
The quantum yield (+) was detected using a monochromatic
carboxyl groups, which have with a negative charge on the light filter (l = 520 nm). + is defined as the number of product
ligand, might interact with the carbonium intermediate and molecules Nmol (5.7 ꢃ 1014 molecules per s, calculated based on
provide the stereo-electronic effects to make the 21 C–H the enhancement of product yields brought by the photo irradia-
cleavage favourable. Another possibility was that peracid was tion) relative to the number of photons Nphotons absorbed by the
formed in situ by the reaction of the carboxyl groups on the photocatalyst:
ligand of 1 and H2O2; the oxygen atom could subsequently be
reacted with the 21 C–H site through a non-covalent interaction,
Nmolðmolecules per sÞ
Nphotonsðphotons per sÞ
+ ¼
which has also observed during non-catalytic oxidations with
peracetic acid or m-CPBA.21a The pendent carboxyl groups of 1
Nphotons, the amount of photons absorbed or scattered by the
photocatalyst, was calculated following the equation:
play important roles in the selective oxidation. However, both
explanations need more evidence and further studies.
Nphotons = Eabs ꢃ 10ꢀ3 CfS
where Eabs is the light energy scattered by the lamp (62 mW cmꢀ2),
Cf is the reciprocal of the average light energy of each photon
Conclusions
(2.6 ꢃ 1018
W
ꢀ1 sꢀ1) emitted by the lamp at 520 nm and S is the
In summary, the iron catalyst 1, which utilised BPY ligands that
were modified with carboxyl groups, was found to not only
exhibit better catalytic activity during the oxidation of alkanes,
but also achieved a reverse oxidative selectivity for 21 over 31
C–H bonds. Although 1 currently affords only intractable
mixtures of ketones and the TN for 1 is not comparable to
the most active catalyst,8 our results provide an alternative
approach, which is to mimic iron enzymes that contain a
2-His-1-carboxylate coordination environment, and realise the
favourable activation of 21 C–H bonds.
entrance area of the light reactor (1.5 cm2).
The isotope-labelling experiments were carried out as fol-
lows: in experiments with H218O, H218O (97% 18O-enriched) was
used as a solvent rather than H2O (entries 1–6 and 8 and 9 of
Table 1), whereas 1000 equiv. H218O, which was relative to the
catalyst, was added to the solution before the addition of
H2O2 (entry 7 of Table 1). In the experiments with 18O2 (96%
18O-enriched), the reaction mixtures were degassed with five
freeze–vacuum–thaw cycles before the reaction; the reaction
was performed under an atmosphere of 18O2. The 16O and 18O
isotope distribution was determined via the relative abun-
dances of the products. The substrates were purified according
to the reported procedure.31 The cycloalkanols (3-ethylcyclo-
Experimental
The iron catalysts were all prepared as described previously.9,15 hexanol, 3-tert-butylcyclohexanol, and 1-decalinol) were synthesised
Catalyst 1 was prepared by mixing FeII(ClO4)2ꢂ6H2O with 3 by the Ru/Al2O3-catalysed hydrogenation of the corresponding
equiv. of DCBPY ligand in H2O. Catalyst 2 was prepared by phenols or naphthol (3-ethylphenol, 3-tert-butylphenol, and
mixing FeII(ClO4)2ꢂ6H2O with 3 equiv. of BPY ligand in CH3CN. 1-naphthol) with H2 (20–50 atm).32 The cycloalkanones (3-ethyl-
Catalyst 3 was prepared by mixing equimolar amounts of cyclohexanone, 3-tert-butylcyclohexanone, and 2-decalinone) were
FeII(ClO4)2ꢂ6H2O and TPA ligand in CH3CN.
synthesised via the oxidation of the corresponding alcohols
The reaction conditions were as follows: photocatalytic (3-ethylcyclohexanol, 3-tert-butylcyclohexanol, and 2-decalinol) with
experiments were conducted under visible light irradiation K2Cr2O7. cis-9-Decalinol and trans-9-decalinol were synthesised by
through a cut-off filter (l > 420 nm), with a 500 W halogen the stereospecific hydroxylation of cis-decalin and trans-decalin
lamp (Philip) as the visible light source. The light source was with m-CPBA/I2, respectively.33 1-Ethylcyclohexanol was synthesised
positioned inside a cylindrical Pyrex vessel that was surrounded via the LiAlH4-catalysed reduction of 1-ethynyl-1-cyclohexanol
by a Pyrex jacket that contained circulating water to cool with H2.34 cis-DMC and trans-DMC were synthesised through the
the lamp. The reaction mixture comprised the iron catalyst oxidation of DMC, which was catalysed by 3/H2O2, respectively.15
(1.5 mM), alkane substrate (1.5 M) and H2O2 (60 mM) in a The diastereomeric mixtures of the synthesised or commercially
solution of CH3CN–H2O (2.5 mL, CH3CN–H2O (v/v) = 60/40). available reagents were separated by GC (Agilent).
H2O2 was added to a vigorously stirred CH3CN solution over
The DPD (N,N0-dialkyl-p-phenylenediamine) method that
ca. 45–75 s. The reaction time was 20 h. In all cases, bromo- was employed for the peroxide measurements was used for
benzene was added as an internal standard. The products were the detection of both H2O2, as well as any ROOH intermediate
extracted with ether and dried over MgSO4 before GC analysis. that was formed during the oxidation of the alkanes.19a The
The samples were analysed twice, once before and once after catalase enzyme, which was used for detecting the presence of a
the treatment of PPh3. Because the TN for 2, 3 and FeII(ClO4)2ꢂ ROOH intermediate,19b was added before the DPD and
6H2O were too small in solutions that contained too much H2O POD (horseradish peroxidase). The concentration of the total
and the ratio for CH3CN–H2O had no effect on the selectivity, peroxides (which included organoperoxides and H2O2) that
the reaction with 2 and 3 was conducted in the pure CH3CN formed during the irradiation was determined after a reaction
solution rather than an aqueous CH3CN–H2O (60/40) solution time of 5 h. A spectrophotometric DPD method was employed
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2013
New J. Chem.

View Article Online
Paper
NJC
(l = 551 nm, e = 21 000 Mꢀ1 cmꢀ1); the DPD reagent was
oxidised by either H2O2 and/or the organoperoxides and was
based on the peroxidase-catalysed reaction. It is known that the
catalase enzyme can eliminate H2O2 from a mixture of H2O2
and organoperoxides. If the catalase is added to a solution that
contains H2O2 and organoperoxides before the DPD method is
used, then this DPD method assays only the organoperoxides.
We can therefore discriminate between H2O2 and the organo-
peroxides to determine if the alkyl peroxides ROOH were
decomposed over the course of the reaction or accumulated
at a high concentration.
The electron spin resonance (ESR) signals of the radicals
that were trapped by DMPO were detected at ambient tempera-
ture with a Bruker ESR 500 E spectrometer under the same
reaction conditions that were used for the other reactions.
A 30 mL sample of the solution was collected at a specified
time and transferred to a quartz capillary for ESR measure-
ment. The settings for the ESR spectrometer were as follows:
centre field, 3443 G; sweep width, 100 G; microwave frequency,
9.64 GHz; modulation frequency, 100 kHz; power, 10.05 mW.
To minimise experimental errors, the same quartz capillary
tube was used for all EPR measurements. The temperature was
controlled with a standard temperature accessory and was
monitored before and after each measurement.
Mendeleev Commun., 1992, 2, 36; (c) G. B. Shul’pin and
M. M. Kats, React. Kinet. Catal. Lett., 1990, 41, 239;
(d) G. B. Shul’pin, A. N. Druzhinina and L. S. Shul’pina,
Pet. Chem., 1993, 33, 321; (e) G. B. Shul’pin, G. V. Nizova and
Y. N. Kozlov, New J. Chem., 1996, 20, 1243.
11 L. Li, S. Niu, D. Li, J. Jin, Y. Chi and Y. Xing, Inorg. Chem.
Commun., 2011, 14, 993.
12 (a) M. Costas, M. P. Mehn, M. P. Jensen and L. Que, Jr.,
Chem. Rev., 2004, 104, 939; (b) M. D. Wolfe and
J. D. Lipscomb, J. Biol. Chem., 2003, 278, 829.
13 (a) P. D. Oldenburg, C. Y. Ke, A. A. Tipton, A. A. Shteinman and
L. Que, Jr., Angew. Chem., Int. Ed., 2006, 45, 7975; (b) A. Beck,
B. Weibert and N. Burzlaff, Eur. J. Inorg. Chem., 2001, 521;
(c) A. Beck, A. Barth, E. Hubner and N. Burzlaff, Inorg. Chem.,
2003, 42, 7182; (d) P. D. Oldenburg, Y. Feng, I. P. Ray, D. Ness
and L. Que, Jr., J. Am. Chem. Soc., 2010, 132, 17713.
14 J. Heilmann, H. W. Lerner and M. Bolte, Acta Crystallogr.,
2006, 62, 1477.
15 K. Chen and L. Que, Jr., J. Am. Chem. Soc., 2001, 123, 6327.
16 A. Danion, J. Disdier, C. Guillard and N. Jaffrezic-Renault,
J. Photochem. Photobiol., A, 2007, 190, 135.
17 (a) X. Tao, W. Ma, T. Zhang and J. Zhao, Angew. Chem., Int.
Ed., 2001, 40, 3014; (b) Y. Huang, W. Ma, J. Li, M. Cheng and
J. Zhao, J. Phys. Chem. B, 2003, 107, 9409.
18 (a) G. B. Shul’pin, J. Mol. Catal. A: Chem., 2002, 189, 39;
(b) G. B. Shul’pin, Y. N. Kozlov, L. S. Shul’pina,
A. R. Kudinov and D. Mandelli, Inorg. Chem., 2009,
48, 10480; (c) G. B. Shul’pin, Y. N. Kozlov, L. S. Shul’pina
and P. V. Petrovskiy, Appl. Organomet. Chem., 2010, 24, 464.
19 (a) H. Bader, V. Sturzenegger and J. Hoigne, Water Res.,
1988, 22, 1109; (b) Y. Zuo and J. Hoigne, Science, 1993,
260, 71; (c) T. Wu, G. Liu and J. Zhao, J. Phys. Chem. B, 1999,
103, 4862.
Acknowledgements
Financial support from 973 projects (No. 2010CB933503 and
2013CB632405), the NSFC (No. 21137004, 21077110 and
21273445), and CAS is greatly appreciated.
References
20 (a) M. S. Chen and M. C. White, Science, 2010, 39, 566;
(b) Y. He, J. D. Gorden and C. R. Goldsmith, Inorg. Chem.,
1 (a) A. Sobkowiak, A. Qui, X. Liu, A. Liobet and D. T. Sawyer,
J. Am. Chem. Soc., 1993, 115, 609; (b) M. S. Chen and
M. C. White, Science, 2007, 318, 783; (c) S. Kille, F. E. Zilly,
J. P. Acevedo and M. T. Reetz, Nat. Chem., 2011, 3, 738.
2 (a) G. B. Shul’pin, Org. Biomol. Chem., 2010, 8, 4217;
(b) T. Newhouse and P. S. Baran, Angew. Chem., Int. Ed.,
2011, 50, 3362.
3 C. Kim, K. Chen, J. Kim and L. Que, Jr., J. Am. Chem. Soc.,
1997, 119, 5964.
4 (a) G. A. Olah, M. B. Comisaro, C. A. Cupas and
C. U. Pittman, J. Am. Chem. Soc., 1965, 87, 2997;
(b) M. Meot-Ner, J. J. Solomon and F. H. Field, J. Am. Chem.
Soc., 1976, 98, 1025.
´
2011, 50, 12651–12660; (c) I. Prat, L. Gomez, M. Canta,
X. Ribas and M. Costas, Chem.–Eur. J., 2013, 19, 1908.
21 (a) H. J. Schneider and W. J. Muller, J. Org. Chem., 1985,
50, 4609; (b) R. Mello, M. Fiorentino, C. Fusco and R. Curci,
J. Am. Chem. Soc., 1989, 111, 6749; (c) A. R. Groenhof,
A. W. Ehlers and K. Lammertsma, J. Phys. Chem. A, 2008,
112, 12855; (d) D. D. DesMarteau, A. Donadelli, V. A. Petrov
and G. Resnati, J. Am. Chem. Soc., 1993, 115, 4897;
(e) G. B. Shul’pin, C. C. Golfeto, G. Su
¨ss-Fink, L. S. Shul’Pina
and D. ManDelli, Tetrahedron Lett., 2005, 46, 4563; ( f ) J. R. L.
Smith and G. B. Shul’pin, Tetrahedron Lett., 1998, 39, 4909;
(g) K. Nehru, S. J. Kim, I. Y. Kim, M. S. Seo, Y. Kim, S. J. Kim,
J. Kim and W. Nam, Chem. Commun., 2007, 4623.
5 G. Roelfes, M. Lubben, R. Hage, L. Que, Jr. and B. L. Feringa,
Chem.–Eur. J., 2000, 6, 2152.
6 J. Green and H. Dalton, J. Biol. Chem., 1989, 17698.
22 In Table 3 of ref. 8, systems with carboxylate group based
oxidant of peracetic acid and m-CPBA increased the regio-
selectivity for secondary C–H bonds.
´
7 R. M. Balleste and L. Que, Jr., Science, 2006, 312, 1885.
8 K. Kamata, K. Yonehara, Y. Nakagawa, K. Uehara and
N. Mizuno, Nat. Chem., 2010, 2, 478.
23 V. Ullrich, Angew. Chem., Int. Ed. Engl., 1972, 11, 701.
9 X. Chen, W. Ma, J. Li, Z. Wang, C. Chen, H. Ji and J. Zhao, 24 M. Fontercave and D. Mansuy, Tetrahedron, 1984, 40, 4297.
J. Phys. Chem. C, 2011, 115, 4089.
10 (a) G. B. Shul’pin and G. V. Nizova, Mendeleev Commun.,
1995, 5, 143; (b) G. B. Shul’pin and A. N. Druzhinina,
25 A. M. Kirillov, M. V. Kirillova, L. S. Shul’ pina, P. J. Figiel,
K. R. Gruenwald, M. F. C. Silva, M. Haukka, A. J. L. Pombeiro
and G. B. Shul’pin, J. Mol. Catal. A: Chem., 2011, 350, 26.
c
New J. Chem.
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2013

View Article Online
NJC
Paper
26 P. H. J. Carlsen, Synth. Commun., 1987, 17, 19.
27 R. Mello, M. Fiorentino, C. Fusco and R. Cursi, J. Am. Chem.
Soc., 1989, 111, 6749.
31 D. D. Perrin and W. L. F. Armarego, Purification of Laboratory
Chemicals, Elsevier, Oxford, 2009.
´
32 A. Solladie-Cavallo, B. Ahmed, M. Schmitt and F. Garin,
28 S. Lee and P. L. Fuchs, J. Am. Chem. Soc., 2002, 124, 13978.
C. R. Chim., 2005, 8, 1975.
29 T. Briffaud, C. Larpent and H. Patin, J. Chem. Soc., Chem. 33 W. Mu¨ller and H. J. Schneider, Angew. Chem., Int. Ed. Engl.,
Commun., 1990, 1193. 1979, 18, 407.
30 B. Kalyanaraman, C. Mottley and R. Mason, J. Biochem. 34 T. Rajamannar and K. K. Balasubramanian, Tetrahedron
Biophys. Methods, 1984, 9, 27.
Lett., 1986, 27, 3777.
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2013
New J. Chem.