Mechanistic insights into transition metal-catalysed oxidation of a hydroxamic acid with in situ Diels-Alder trapping of the acyl nitroso derivative

Article abstract of DOI:10.1002/adsc.200700568

New insights into the mechanism for the transition metal-mediated oxidation of hydroxamic acids to give intermediate acyl nitroso species, with subsequent hetero-Diels-Alder trapping are presented. The activation of triphenylphosphine-ligated ruthenium-salen complexes is examined, and evidence is presented for the ruthenium-oxo species which are involved in the oxidative process of the hydroxamic acid. The observation of the lack of asymmetric induction involved in the intermolecular cycloaddition process involving the intermediate acyl nitrsoso species is explained, with the aid of comparing the ruthenium-salen-based systems with nitrosotoluene, and copper(I)/copper(II) BINAP-based catalysis of nitrosopyridine complexes. This study demonstrates the importance of secondary coordination to achieve asymmetric induction in nitroso-Diels-Alder reactions.

Full text of DOI:10.1002/adsc.200700568

FULL PAPERS
DOI: 10.1002/adsc.200700568
Mechanistic Insights into Transition Metal-Catalysed Oxidation of
a Hydroxamic Acid with in situ Diels–Alder Trapping of the Acyl
Nitroso Derivative
Judith A. K. Howard,a Gennadiy Ilyashenko,a Hazel A. Sparkes,a
Andrew Whiting,a,* and Allen R. Wrightb
a
Department of Chemistry, Science Laboratories, Durham University, South Road, Durham DH1 3 LE, U.K.
Fax : (+44)-(0)191-386-1127; e-mail: andy.whiting@durham.ac.uk
School of Chemical Engineering and Advanced Materials, Merz Court, University of Newcastle, Newcastle upon Tyne
b
NE1 7AR, U.K.
Received: December 4, 2007; Published online: March 25, 2008
Abstract: New insights into the mechanism for the nitrsoso species is explained, with the aid of compar-
transition metal-mediated oxidation of hydroxamic ing the ruthenium-salen-based systems with nitroso-
acids to give intermediate acyl nitroso species, with toluene, and copper(I)/copper(II) BINAP-based cat-
subsequent hetero-Diels–Alder trapping are present- alysis of nitrosopyridine complexes. This study dem-
ed. The activation of triphenylphosphine-ligated onstrates the importance of secondary coordination
ruthenium-salen complexes is examined, and evi- to achieve asymmetric induction in nitroso-Diels–
dence is presented for the ruthenium-oxo species Alder reactions.
which are involved in the oxidative process of the
hydroxamic acid. The observation of the lack of
asymmetric induction involved in the intermolecular Keywords: hetero-Diels–Alder reaction; oxidation;
cycloaddition process involving the intermediate acyl oxo ligands; ruthenium
Introduction
mic. The opposite is also true: if the dissociation of
the oxidised species or re-oxidation is slower than the
The use of nitroso compounds as efficient hetero di- DA reaction, then the product should show some
enophiles in [4+2]cycloaddition reactions with conju- level of enantiomeric induction. It is possible that in-
gated dienes to produce 3,6-dihydro-1,2-oxazines has tramolecular DA reactions form products with good
been studied for over half a century.[1] These types of levels of asymmetric induction, while the correspond-
hetero-Diels–Alder (DA) reactions have been used as ing intermolecular reactions do not, because their re-
powerful synthetic tools in the formation of natural actions are much faster. Although, this hypothesis has
products such as polyhydroxylated alkaloids and their been postulated previously[8] no robust mechanistic
derivatives.[2,3] In addition to well established stoichio- studies had been carried out. Our long-term interest
metric oxidative processes,[4] there are a few transition is to develop systems for the oxidation of hydroxamic
metal-based catalysts which are capable of oxidising acids with asymmetric in situ DA trapping of the acyl
hydroxamic acids to their corresponding nitroso spe- nitroso derivative,[5,7] to this end detailed mechanistic
cies.[5,6] Despite numerous attempts to carry out an studies of the reactions have been carried out and
asymmetric version of this transformation only limit- herein we report our results.
ed success (low ee) was achieved for intermolecular
cycloaddition,[7] however, much improved asymmetric
induction has been achieved in intramolecular ni- Results and Discussion
troso-Diels–Alder reactions.[8] The low ees in intermo-
lecular cases and good levels of asymmetric induction Mechanism Studies on in situ Generated Acyl
in intramolecular cases poses a question about the in- Nitroso Compounds
volvement of the catalyst in these DA reactions. If the
dissociation of an oxidised species from the catalyst, In order to detect possible intermediates in the ruthe-
or re-oxidation of the catalyst is faster than the subse- nium-salen-mediated oxidation of a hydroxamic acid
quent DA reaction, then the product would be race- a combination of techniques was employed; time-re-
Adv. Synth. Catal. 2008, 350, 869 – 882
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
869

Judith A. K. Howard et al.
FULL PAPERS
solved infrared spectroscopy using ReactIR, mass
spectrometry and concentration profiles being ob-
tained from 1H NMR spectroscopy. The results of
these experiments were utilised to construct a reac-
tion network in order to postulate possible reaction
mechanism(s). The initial model reaction studied was
the Ru(salen) 4 catalysed oxidation of Z-protected
G
hydroxamic acid 1 [Eq. (1)] with t-BuOOH and the
subsequent trapping of acyl nitroso species with a
diene 2. The imine shifts of the ruthenium complex
were monitored by IR which it was thought would
provide a direct probe for the coordination environ-
ment of ruthenium during the reaction as well as pro-
viding structural information on any important cata-
lytically active species. Mass spectrometry was also
employed to try and confirm the identity of short-
lived intermediates, while time-resolved 1H NMR
spectroscopy was used to measure quantitatively the
concentrations of 1, 2 and 3 during the reaction and
enable kinetic analysis.
Initially, the reaction shown in Eq. (1) was studied
using time-resolved IR spectroscopy. Stoichiometric
reactions between each component of the reaction
mixture (1, 2 and oxidant) and the catalyst 4, were set
up to investigate the interactions between each of the
reactant species and the catalyst. No changes were
observed in the IR spectrum when either hydroxamic
acid 1 or the diene 2 were mixed with 4, in both cases
this indicates that there are no interactions between
either 1 or 2 and 4. However, upon addition of the ox-
idant to 4, a new absorption peak corresponding to
Figure 1. Structure of proposed active catalyst 5a or 5b and
ReactIR snapshots were generated using Excel: top, expand-
ed imine region; bottom, expanded Ru=O region.
the formation of O=PPh3 appeared at 1250–1300 cmÀ1 the initial oxidation of the Ru
(salen) complex 4. The
A
(see Figure 1, top). More interestingly was the simul- oxidation of 4 to give 5b can occur via two possible
taneous appearance of a weak absorption peak routes, as shown in Scheme 1. In pathway 1, an equi-
around 830 cmÀ1 (see Figure 1, bottom), which can be librium is established between 4 and a species in
assigned to an Ru=O stretch.[9] The existence of spe- which PPh3 has dissociated from the metal centre.
cies 5a and 5b in the reaction mixture was confirmed Upon the addition of oxidant both PPh3 and
by mass spectrometry with the appearance of ions at
N
N
m/z=529 for 5a and 791 for 5b. Based upon these ob- oxide and 5b, respectively, hence, the equilibrium
servations and by drawing analogies with the shifts to the right until none of complex 4 remains.
chromium(V)oxo
A
we propose that the active catalyst is Ru
N
N
5a. It is most likely that 5a is formed by the dissocia- tached to the metal centre, resulting in oxidation of
tion of PPh3 from complex 5b which had arisen from PPh3 and dissociation from the metal centre. Which-
870
asc.wiley-vch.de
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 869 – 882

Mechanistic Insights into Transition Metal-Catalysed Oxidation of a Hydroxamic Acid
FULL PAPERS
Scheme 1. Two possible pathways towards the oxidation of Ru
N
N
ever pathway is followed, complex 5b forms initially these results remain qualitative due to the complexity
and is then converted to the active catalyst 5a. of the spectra and changes in the base line over time.
The ReactIR studies carried out on the reaction It was also not possible to monitor the formation of
outlined in Eq. (1), also indicated that catalase-like the acyl nitroso species using ReactIR as the absorp-
(dioxygen evolution) activity of the ruthenium com- tions corresponding to compounds 1 and 3 overlap
plex 4 (or its derivative 5a and 5b) was negligible. In and were further masked by the water absorption.
fact, it is only when hydroxamic acid 1 is added to the However, it is possible to draw the following conclu-
reaction mixture containing all of the components, in- sions: firstly, there were no detectable interactions be-
cluding oxidant that the reaction is initiated. Time-re- tween 1 or 2 and 4 which might have indicated an in-
solved IR monitoring of the oxidation process clearly termediate species; secondly, the catalase activity of
showed the formation of water in the reaction mix- complex 4 is negligible; thirdly, the oxidation of hy-
ture (see Figure 2), the absorption peak at approxi- droxamic acid 1 is fast and is most likely to result
mately 1700 cmÀ1 was assigned to hydroxamic acid 1 from oxidation by an Ru=O complex, either 5a or 5b,
while the broad peak around 1650 cmÀ1 corresponds formed in situ in the manner described above.
1
to the formation of water during the oxidation pro-
The H NMR spectroscopic studies were carried on
cess. Despite extensive efforts to quantify the data, each of the reactions described in Eq. (1) and Eq. (2),
Figure 2. Relevant part of the ReactIR spectrum obtained during catalytic reaction described in Eq. (1).
Adv. Synth. Catal. 2008, 350, 869 – 882
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
871

Judith A. K. Howard et al.
FULL PAPERS
mic acid had continued, it would have resulted in side
products arising from decomposition of the nitroso in-
termediate; these were not identified. Attempts were
made to monitor these reactions using an increased
catalyst loading of 5 mol%, however, this resulted in
the reactions occurring at such a fast rate that it was
impossible to obtain meaningful NMR spectroscopic
data. In accordance with previous findings, the use of
enantiomerically pure catalyst 4 in these intermolecu-
lar reactions resulted in the formation of racemic
using anisole (100 mol%) as internal standard. Reac- products.[5]
tions were followed over time to determine the rate
In conjunction with the intermolecular study, the
of conversion of starting materials to products and intramolecular reaction [Eq. (3)] of Shea et al.[8] was
the results are shown in Figure 3.
also investigated using the same approach. This reac-
tion has been shown to proceed with good levels of
asymmetric induction when an enantiomerically pure
catalyst 4 is employed.[8] As with the intermolecular
reactions, the concentration of starting materials and
1
products could be followed over time using H NMR
spectroscopy (Figure 4).
Figure 3. 1H NMR-spectroscopy derived concentration pro-
files for the reactions described in Eqs. (1) and (2) using:
top, cyclohexadiene 2; bottom, 2,3-dimethybutadiene 6.
Figure 4. 1H NMR spectroscopy-derived concentration pro-
file for the intramolecular reaction described in Eq. (3)
using hydroxamic acid 8.
The concentration profiles obtained from the
1H NMR spectroscopy measurements (Figure 3) sug-
gest that the rate of each of the reactions described
by Eq. (1) and Eq, (2) is controlled by the catalytic
oxidation of 1 by 4, that is, the simultaneous disap-
Unsurprisingly, Figure 4 shows a similar linear pro-
pearance of both starting materials and concomitant file to those shown in Figure 3, however, the non-
appearance of the product occur at the same rate quantitative conversion of starting material to product
(within experimental error). This indicates that as was not expected, i.e., there is a clear mass imbalance
soon as the nitroso intermediate is produced through shown in Figure 4. After approximately 24 min, most
the oxidation, it is essentially consumed instantly by of the starting hydroxamic acid had been consumed,
the diene. If the diene is used as the limiting reagent, but only approximately 2/3 of the expected product
then it is consumed completely to form the cycload- had been produced. It has been postulated that the
duct in quantitative yield, if oxidation of the hydroxa- observed enantiomeric excess can be explained by the
872
asc.wiley-vch.de
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 869 – 882

Mechanistic Insights into Transition Metal-Catalysed Oxidation of a Hydroxamic Acid
FULL PAPERS
assertion that the intramolecular Diels–Alder reaction using BatchCADꢁ[11] to provide the results shown in
is faster than dissociation of the nitroso species from Figure 5.
the ruthenium catalyst.[8] Indeed, if Diels–Alder cycli-
From the data presented in Figure 5, it can be seen
sation is faster than the dissociation process, the start- that the reaction shown in Eq. (4) follows simple
ing material would be expected to be converted quan-
titatively into product. Since this is not the case, as
demonstrated graphically by Figure 4, the implication
is that the acyl nitroso species does dissociate from
the ruthenium-salen catalyst species faster than intra-
molecular cyclisation can occur, and although cyclisa-
tion does occur, it is in competition with decomposi-
tion of the intermediate nitroso species. In addition,
when product 9 [Eq. (3)] was isolated it was found to
be racemic. Further experiments were undertaken in
an attempt to prepare the cycloadduct 9 in an enan-
tiomerically enriched form, including carrying out the
reaction at lower temperatures, however, these failed
to produce any sign of asymmetric induction.
Figure 5. Concentration profile for the thermal reaction de-
1
scribed in Eq. (4), experimental (as determined by H NMR
Mechanism Sudies on ortho-Nitrosotoluene
spectroscopy) and calculated (modelled using Batch-
CADꢁ).
The results obtained thus far all seemed to indicate
that the oxidation of each of the hydroxamic acids
employed in Eqs. (1)–(3) is the controlling step in second-order kinetics, with k=2.2”10À3 sÀ1 MÀ1
each of the reactions and, therefore, it necessary to {rate=k*[cyclohexadiene][nitrosotoluene]} (obtained
investigate the kinetics of nitroso-Diels–Alder cyclo- by UV/vis spectroscopy). The concentration profiles
1
addition independently of the oxidation process. Un- obtained experimentally by H NMR methods were
11
fortunately, the instability of the intermediate acyl ni- similar to those obtained using BatchCADꢁ
troso species involved in Eqs. (1)–(3) means that (Figure 5). In addition, the calculated rate constant
these types of reactive species could not be used for (k=2.7”10À3 sÀ1 MÀ1) showed good agreement with
the desired kinetic experiments and so attention was that obtained experimentally by UV/vis spectroscopy.
turned to more stable nitroso derivatives, starting In order to compare the kinetics of the thermal cyclo-
with ortho-nitrosotoluene 10. In the first instance, ki- addition reactions with the corresponding catalysed
netic investigations into the thermal nitroso-cycload- reactions, the effect of including catalysts 4 and 12
dition reaction were undertaken between 10 and cy- into the reaction [Eq. (4)] was investigated. The re-
clohexadiene 2 [Eq. (4)]. This reaction was followed sults of these experiments are summarised in the reac-
1
over time by H NMR as described above, and model- tion profiles shown in Figure 6.
ling of the reactant/product profile was carried out
The results shown in Figure 6 clearly show that cat-
alyst 12 has no effect, whereas catalyst 4 actually re-
duces the rate of product formation to a small degree.
Closer examination of reaction profile revealed that,
in all cases, nitrosotoluene 10 was consumed at ap-
proximately the same rate, whereas cyclodiene 2 was
consumed slightly slower in the presence of Ru(salen)
A
4. It can, therefore, be postulated that the negative
effect in the rate of cycloaddition can be attributed to
the competing side reaction of the oxidation of triphe-
nylphosphine by nitrosotoluene after it has been re-
leased from the catalyst. In fact, the reaction of free
triphenylphosphine (1 equiv.) with nitrosotoluene
(1 equiv.) at room temperature is rapid (complete in
ca. 5 min) and indeed produces triphenylphosphine
oxide and 2-methylaniline in almost quantitative
yields.
We have previously shown that complexes 12, 13
and 15 were active catalysts in the oxidation of hy-
Adv. Synth. Catal. 2008, 350, 869 – 882
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
873

Judith A. K. Howard et al.
FULL PAPERS
those obtained for thermal reaction and no asymmet-
ric induction was observed which confirms that ni-
troso species are poor coordinating ligands.
Mechanism Studies on 6-Methyl-2-nitrosopyridine
As illustrated above, the nitroso group is a poor coor-
dinating ligand for ruthenium, which means that the
most likely explanation for the lack of the catalytic
activity of the ruthenium-based catalysts towards the
nitroso-Diels–Alder reactions is the absence of a suit-
able binding function to enable the catalyst and sub-
strate to interact. In order to overcome this problem,
one needs to have an extra binding motif on the sub-
strate to provide further coordination to the metal
centre. Indeed, Yamamoto has elegantly illustrated
this by use 6-methyl-2-nitrosopyridine 21 which reacts
with diene 2 in the presence of Cu(I)-BINAP complex
to give an enantiomerically enriched cycloadduct 22
[Eq. (5)].[13]
Following previously described protocols,[13,16] the
synthesis of nitrosopyridine 21 was attempted in
order to compare the kinetics of its reaction with
those discussed above. The synthesis of nitrosopyri-
dine 21 involves the mCPBA-mediated oxidation of
the corresponding ylide 24, which was formed from
aminopyridine 23 as outlined in Scheme 2, however,
instead of isolating the expected nitroso compound
21, the azo mono-oxide 26 was isolated as the major
component, the structure of which was confirmed by
single crystal X-ray crystallography (Figure 7). Fur-
ther investigation showed that the presence of un-
reacted starting material 23 triggers the disproportina-
tion of dimeric azo di-oxy species 25 to give 26. The
separation of 24 and 23 is not possible at an earlier
stage of the reaction, and the oxidation of 24 occurs
with predominant formation of the by-product 26.
This can be largely circumvented by carrying out the
oxidation of 24 at lower temperature, and once the
solvent is removed after the oxidation step, the result-
ing mixture can be loaded on to a column and puri-
fied allowing separation of 21 and 23. Although the
decomposition of 21 or 25 to 26 still occurs, it is at
Figure 6. 1H NMR derived study showing the effect of the
catalysts 4 (10 mol%) and 12 (2 mol%) on the nitroso-
Diels–Alder reaction shown in Eq. (4). top: consumption of
nitrosotoluene 10; middle: consumption of diene 2; bottom:
formation of cycloadduct 11.
droxamic acid to nitroso species, albeit the subsequent much reduced rate and both compounds can be ob-
nitroso-Diels–Alder reaction proceeded with no tained as reasonably stable solids. It is worth noting
asymmetric induction.[5c] Since then we have expand- that 26 can seemingly not be re-oxidised to 25 or 21
ed our study to include complexes 14[12] and 16. under the reaction conditions and is completely inert
Prompted by Yamamotoꢂs report that Cu(I)-BINAP in the nitroso-Diels–Alder reaction.
complexes produced high ee in nitroso-Diels–Alder
Since the synthesis of 21 could be capricious and
reactions[13] we have also decided to include in this low yielding, we attempted to eliminate the isolation
study complexes 18 and 19 with analogous of 21 and instead carry out the oxidation of 24 to 21
A
N
containing a ligand with chiral phosphorus.[15] All of in situ using complexes 4 and 12–16 as potential cata-
these complexes 12–20 were screened in the reaction lysts [Eq. (6)].
and compared to the catalysts 4 at room temperature
Of the various combinations of the catalysts and
and at À788C. In all cases the yields were similar to oxidants examined [Table 1, Eq. (6)], most proved to
874
asc.wiley-vch.de
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 869 – 882

Mechanistic Insights into Transition Metal-Catalysed Oxidation of a Hydroxamic Acid
FULL PAPERS
be largely unsuccessful, with the notable exception of
the iron-based catalyst 15 (entries 5 and 7) which
when used in conjunction with H2O2 at room temper-
ature or peracetic acid at À788C in acetone gave the
nitroso cycloadduct 22 in 42% and 47% yields, re-
spectively (both racemic). To the best of our knowl-
edge, this is the first example of the in situ generation
of the nitroso species 21, and considering the intrica-
Scheme 2. Decomposition of nitrosopyridine 21 to 26.
Adv. Synth. Catal. 2008, 350, 869 – 882
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
875

Judith A. K. Howard et al.
FULL PAPERS
of UV/Vis spectroscopy to follow the reaction was
also not possible. Therefore, it was decided to screen
available catalysts shown above in this reaction,
under the conditions described by Yamomoto,[13] in
order to use asymmetric induction to probe the cyclo-
addition [Eq. (5)] further, the results are presented in
Table 2.
A combination of Cu(I) with (R)-BINAP (entry 3)
yielded product 22 in an 88% yield with a 65% ee
which although lower than the ees reported by Yamo-
moto,[13] was essentially as expected. In contrast, em-
ploying complexes 4, 12–15, 17 and 20 (entries 4–8, 10
and 12) gave identical results to the thermal reaction
(entry 2) with 100% conversion and 0% ee On the
other hand, when Cu(II) complex 16 was used
(entry 9), 22 was produced with an 84% yield and
13% ee This indicates that Cu(II) complexes may be
used as catalysts in these types of cycloaddition reac-
tions, and it raises the question of whether they may
in fact be preferable to Cu(I) complexes, which might
be sensitive to the potentially oxidising conditions
created by the presence of nitrosopyridine 21. In
order to probe this mechanistic point further, the use
Figure 7. X-ray crystal structure of azo-oxide 26, ellipsoids
are depicted at the 50% probability level. Hydrogen atoms
are shown as spheres of arbitrary radius. For clarity only
one position is shown for O1 (labelled O1a).
cies of the formation of 21 as a pure reagent and its of Cu(I) in the original protocol[12] was replaced by
lack of commercial availability, this may be a useful Cu(II) in the same reaction (entry 12). When the re-
finding applicable in other reactions, such as the ni- action was complete, 22 was isolated in an 89% yield
troso-aldol and related reactions.[17]
with a 50% ee Further investigation into the Cu(II)-
A
The thermal cycloaddition between nitrosopyridine (R)-BINAP-catalysed reaction of 2 with nitrosopyri-
21 and diene 2 under the conditions identical to those dine 21 at À858C was carried out, monitoring the ee
used for nitrosotoluene proceeded much faster (ca. of product 22 over time (entry 13). Aliquots were col-
<5 min versus 100 min, respectively) and it was there- lected every hour for 7 h and analysed using HPLC,
1
fore not possible to employ H NMR spectroscopy to however, the enantiomeric excess remained constant
obtain a starting material/product reaction profile. at approximately 10% throughout the reaction. Upon
The nitrosopyridine 21 exists in organic solvents in an increasing of the temperature to À208C the ee im-
equilibrium between the azo-monomer and di-oxide proved dramatically to 45% and further investigations
dimer, since their UV chromophores overlap, the use revealed that at or below À708C both the catalysed
Table 1. Screening of various conditions for the reaction described in the Eq. (6).
Entry
Catalyst [1 mol%]
Oxidant
Solvent
Temperature [oC]
Time [h]
Yield[a] [%]
ee [%]
1
2
-
-
-
H2O2
Acetone
MeCN
DCM
Acetone
Acetone
MeCN
Acetone
DCM
DCM
DCM
Acetone
Acetone
DCM
22
22
22
22
22
22
-78
-40
22
-40
22
22
-40
22
16
2
2
1
16
2
0
-
CH3CO3H[b]
mCPBA
H2O2
38
33
0
42
15
47
0
0
0
0
0
0
0
-
0
0
0
-
-
-
-
-
3
4c
5
FeCl3
15
15
15
15
4
H2O2
6[c]
7
CH3CO3H[b]
CH3CO3H[b]
mCPBA
t-BuOOH
mCPBA
H2O2
8
8
8
9
16
16
16
16
8
10
11
12
13
14
4
12
13
14
16
H2O2
mCPBA
0
10
-
0
CH3CO3H[b]
MeCN
16
[a]
Isolated yields.
[b]
[c]
Peracetic acid was prepared by stirring acetic acid and H2O2 over Amberlite 400 resin to give 25% w/v solution.
Disproportionation of oxidant was observed.
876
asc.wiley-vch.de
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 869 – 882

Mechanistic Insights into Transition Metal-Catalysed Oxidation of a Hydroxamic Acid
FULL PAPERS
Table 2. Screening of catalysts in asymmetric nitroso-Diels–Alder reaction as shown in Eq. (5).
Entry
Catalyst [10 mol%]
Temperature [oC]
Conversion [%][a]
Yield [%][b]
ee [%][c]
1
2
3
4
5
6
7
8
-
-
18
4
12
13
14
15
16
17
20
19
19
19
25
100
100
100
100
100
100
100
100
100
100
100
100
<10
100
88
91
88
-
-
-
-
-
84
-
-
0
0
65
0
0
0
0
0
13
0
À85 to À20[d]
À85 to À20[d]
À85 to À20[d]
À85 to À20[d]
À85 to À20[d]
À85 to À20[d]
À85 to À20[d]
À85 to À20[d]
À85 to À20[d]
À85 to À20[d]
À85 to À20[d]
À85
9
10
11
12
13
14
0
89
-
90
50
10
60
À70 to À55[e]
[a]
Defined as% of starting material consumed.
Only isolated yields are quoted unless otherwise stated.
Obtained from HPLC.
Reaction initiated at À858C, held at that temperature for 1 hour and then allowed to warm up to À208C over 5 h period
at which point it was held there for additional hour before allowing to reach room temperature.
The temperature was maintained in this window using cryostat.
[b]
[c]
[d]
[e]
and thermal reactions were negligible, whereas above portionation product 26. The voltammogram, there-
À558C, the thermal reaction contributed significantly fore, has the characteristics of an ECE process. Final-
to the product formation. When the reaction tempera- ly, a series of tests to check for interactions between
ture was maintained between À70 and À558C and al- nitrosopyridine and Cu(I) were carried out (Figure 8,
lowed to proceed for 24 h, the final product 22 was bottom). As expected, no interactions were observed
obtained in 90% yield and 60% ee (Entry 14), i.e., es- between the Cu(I) salt and nitrosopyridine 21 in
sentially matching the results reported above for the DCM, which is most likely to be due to the poor solu-
Cu(I)-based system.
bility of the metal salts in this solvent. Upon the addi-
In order to gain further insight into the reaction tion of BINAP to this suspension, the concentration
mechanisms involved in the Cu(I) versus Cu(II)-cata- of nitrosopyridine 21 dramatically decreased, while
lysed cycloaddition reactions [Eq. (5)], it was decided the concentration of the disproportionation product
to study the electrochemistry of nitrosopyridine 21, 26 appeared to increase. Nitroso species are known to
together with the Cu(I)- and Cu(II)-(BINAP) com- be good oxidants[14] and, under the experimental con-
plexes, and compound 26. These experiments were ditions, nitrosopyridine 21 appears to be similar to
carried out in dichloromethane (DCM) using all Pt molecular oxygen in its ability to carry out oxidations,
electrodes, with Bu4NPF6 as the electrolyte and ferro- which implies that Cu(I) is oxidised to Cu(II) under
cene as an internal reference. The results of these ex- these reaction conditions. Furthermore, the natural
periments are shown in Figure 8.
tendency of nitrosopyridine 21 to disproportionate
The cyclic voltammogram of the free BINAP ligand can be further enhanced by the presence of Cu(I).
showed two irreversible oxidation processes at 1.19 V Thus, these results seem to demonstrate that the
and 1.54 V (Figure 8, top). Upon coordination with Cu(I)-(BINAP) complex is oxidised to Cu(II)-
copper, these oxidation events shifted to 1.25 V and (BINAP) in the presence of nitrosopyridine 21, and,
1.61 V, respectively, likely a consequence of the in- that it is actually the Cu(II)-(BINAP) complex that
creased coulombic charge on the complex; under the catalyses the reaction described in Eq. (5).
conditions employed, the Cu(I)/Cu(II) couple was not
observed. Cyclic voltammetry of nitrosopyridine 21
showed two reversible waves with different peak cur- Proposed Mechanism
rents, the first DEp =À0.69 V and the second DEp =
À1.16 V (Figure 8, middle). Initially, it was believed When all of the above is taken into consideration, a
that these two potentials were due to monomer-dimer network of reactions can be proposed for the
equilibration, however, it was later confirmed that the
A
G
N
second wave (DEp =À1.16 V) was due to the dispro- mic acid such as 1, with in situ Diels–Alder trapping
Adv. Synth. Catal. 2008, 350, 869 – 882
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
877

Judith A. K. Howard et al.
FULL PAPERS
itself. It is therefore postulated that dissociation of
the phosphine ligand must occur first to give
A
N
then oxidised to the corresponding oxo-species 5a or
5b, and triphenylphosphine oxide side product, re-
spectively. Since the Ru
trophilic oxidant than t-BuOOH, it is also most likely
that the Ru(oxo) species is involved in the oxidation
of free phosphine. Once formed, the Ru(salen)-oxo
(oxo) species is a more elec-
AHCTREUNG
AHCTREUNG
complex 5a or 5b proceeds to catalyse the oxidation
of hydroxamic acid 1 to the corresponding acyl ni-
troso 27. This oxidation can occur in two ways: either
as a concerted or stepwise process. In the concerted
pathway, deprotonation of hydroxamic acid by Ru-
A
from the hydroxamic acid 1 to the metal centre result-
ing in an acyl nitroso species 27 and metal hydride
complex 28. The Ru
ly loses water and is subsequently reoxidised to Ru-
(oxo)(salen) 5a by t-BuOOH. Alternatively, the oxi-
(salen) hydride complex 28 quick-
A
ACHTREUNG
dation could proceed in a stepwise manner very simi-
lar to the well documented reactions between hydrox-
ylamines and metals.[18] Following initial deprotona-
tion by the metal-oxo species, hydroxamic acid 1 can
coordinate to the metal centre through the terminal
oxygen, with the resulting intermediate 29 then col-
lapsing to expel the acyl nitroso 27 and water from
the coordination sphere of the catalyst. After the col-
lapse, the complex is oxidised back to the Ru(oxo) 5a
H
and returns to the beginning of the catalytic cycle. Al-
though both pathways are possible, we believe that
stepwise process is more likely, since it is analogous
to the well established oxidation of alcohols with
PCC/TPAP[19] and does not require a hydride transfer.
Whichever route the oxidation takes, it results in a
free, very reactive, acyl nitroso species which rapidly
reacts with dienes via a conventional cycloaddition
Figure 8. Cyclic voltammetry measurements for: top: rever- process to give adducts such as 3. Since this seems to
sible reduction of BINAP; middle: reduction of nitrosopyri-
dine 21 and compound 26; bottom: change in CV upon addi-
occur outside the coordination sphere of the catalyst,
the reaction generally proceeds without selective
asymmetric induction, at least in the intermolecular
cycloaddition process.
tion of Cu(I)(OTf) and BINAP.
G
The fact that there is no acceleration of the reac-
tion or asymmetric induction in the product 11 during
of the acyl nitroso derivative (Scheme 3). The overall the cycloadditions using nitrosotoluene 10 [Eq. (4)],
reaction can be broken down into three parts: oxida- further confirms our hypothesis that acyl nitroso is
tion of the metal catalyst into the active oxo-species, not bound to the catalyst during the DA step. Fur-
which subsequently oxidises hydroxamic acid to the thermore, Ru
A
nitroso species, and finally, a hetero-Diels–Alder cy- be inactive as a catalyst in nitrosopyridine cycloaddi-
A
tions indicating that ruthenium metal is a poor cata-
Although the oxidation of catalyst 4 into the active lyst for such Diels–Alder processes. Again, this con-
Ru
(Scheme 1), further comments are warranted. hydroxamic acids are excellent for the oxidation step,
Ru(salen)
(PPh3)2 4 is an electron-rich 18eÀ complex, however they are not involved in the subsequent
hence it is unlikely that conversion to the Ru(oxo) hetero-Diels–Alder step.
A
G
A
N
ACHTREUNG
AHCTREUNG
species would occur via nucleophilic attack of the oxi-
dant on either the phosphine ligand or the metal
878
asc.wiley-vch.de
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 869 – 882

Mechanistic Insights into Transition Metal-Catalysed Oxidation of a Hydroxamic Acid
FULL PAPERS
Scheme 3. Proposed mechanism for oxidation of hydroxamic acids with in situ Diels–Alder trapping of the acyl nitroso deriv-
ative as catalysed by Ru
N
N
Conclusions
dine interactions if we are to design a catalyst with is
capable of catalysing both the oxidation and Diels–
We have postulated on the mechanism of the oxida- Alder reactions. The work to develop such asymmet-
tion of hydroxamic acids and explained why the sub- ric catalysts is currently being undertaken in our labo-
sequent cycloaddition occurs without asymmetric in- ratory and the results will be presented in the due
duction in the product. In previous work, we and course.
others have shown that complexes 4, 12–16 are good
catalysts for the oxidation step,[7,8] in view of the re-
sults obtained from the screening of nitrosopyridine
we need to review catalytic systems in order to carry
out the oxidation and cycloaddition on the same
metal centre. We know that Cu(II)-(BINAP) can cat-
alyse a cycloaddition reaction between nitrosopyri-
dine 21 and diene 2 with good levels of enantioselec-
tivity, and that a combination of Cu(II) and pyridine
moiety is essential at this point, however we need to
Experimental Section
General Methods
Where applicable, glassware was oven dried (1308C) before
use and cooled under a positive pressure of argon. All dry
solvents were dried using a commercial drying system. All
other materials were purchased directly from standard
chemical suppliers and used without further purification
understand the intrinsic nature of the Cu/nitrosopyri- unless otherwise stated. TLC was performed on plastic-
Adv. Synth. Catal. 2008, 350, 869 – 882
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
879

Judith A. K. Howard et al.
FULL PAPERS
backed silica gel plates PET backed plates with visualisation
achieved using a UV lamp or staining with K2MnO4 solu-
tion. Drying was carried out over anhydrous MgSO4, fol-
lowed by filtration. Evaporations were carried out at 20
mmHg using a rotary evaporator and water bath, followed
by evaporation to dryness under vacuum (<2 mmHg). Pu-
rification by medium pressure column chromatography was
performed using silica gel 35–70 mm. Melting points are un-
corrected. All 1H and 13C NMR were recorded on either
400 MHz or 500 MHz Bruker spectrometers. Chemical shifts
are expressed as parts per million (ppm) downfield from the
internal standard TMS. Asymmetric induction was deter-
mined using an HPLC system and UV/vis detector. Time-re-
solved IR was recorded on ReactIR 4000 and data manipu-
lated using Mettler-Toledo ReactIR 3.0 software. CV was
recorded on Autolab PGSTAT-30 in DCM using n-Bu4NF as
electrolyte and Cp* as internal reference.
mixture was stirred for one hour at 08C and then for an ad-
ditional hour at room temperature. Sodium methoxide
(25 w/v% in MeOH, 4 mL, 18 mmol) was then added and
resultant suspension was stirred for 10 min after which
water (15 mL) was added and final mixture was stirred for
an additional hour. The organic layer was separated and the
aqueous layer was extracted with DCM (50 mL). The com-
bined organic phases were washed with brine (100 mL),
dried over MgSO4 and filtered. The filtrate was concentrat-
ed under reduced pressure to give a yellow residue which
was used in the next step without purification. 1H NMR
spectroscopy indicated a 1:4 ratio (starting material:prod-
uct).
A solution of sulfilimine from the first step (1.0 g,
5.9 mmol) in DCM (10 mL) was slowly added to an ice-cold
solution of mCPBA (77% max, 1.76 g, 7.13 mmol) in DCM
(30 mL). Upon addition the reaction mixture turned green.
The reaction mixture was then stirred for 1 hour at 08C
during which it became brown-red. Dimethyl sulfide
(0.22 mL, 2.95 mmol) was added to the solution and the re-
action was then stirred for an additional hour at room tem-
perature. A saturated solution of Na2CO3 (30 mL) was
added in and the layers were separated. The organic phase
was washed with water (2”30 mL), dried over MgSO4 and
filtered. The solvent was removed under reduced pressure
to give a brown solid. The product was obtained from dieth-
yl ether as a brown-red crystalline solid (0.45, 66%) g, and
the structure was confirmed by single crystal X-ray crystal-
lography; mp 85–878C. 1H NMR: (400 MHz, CDCl3): d=
2.59 (s, 3H), 2.65 (s, 3H), 7.12 (d, J=7.6 Hz, 1H), 7.37 (d,
J=7.6 Hz, 1H), 7.12 (t, J=7.8 Hz, 1H), 7.80 (t, J=8.0 Hz,
1H), 8.12 (d, J=8.0 Hz, 1H), 8.17 (d, J=7.8 Hz, 1H);
13C NMR (400 MHz, CDCl3): d=24.2, 24.4, 224.7, 115.4,
123.5, 126.8, 138.0, 139.1, 155.7, 157.0, 158.6, 158.7; IR (NaCl
film): nmax =3058 (w), 1606 (m), 1591 (m), 1560 (m), 1477
(s), 1449 (s), 1375 (m) 1340 (m), 1272 (m), 1207 (m), 1167
(m), 1095 (w), 1082 (w), 1037 (w), 998 (m), 922 (w), 853
(m), 815 (m), 796 (m), 728 (m), 699 (w), 670 (w), 624 (w),
600 (w), 552 cmÀ1 (w); HR-MS (ES): m/z=229.1086 [M+
H]+ (C12H12N4O requires 229.1084); anal. found (%): C
62.67, H 5.34, N 23.60 (C12H12N4O requires C 63.15, H 5.30,
Synthesis of Nitrosopyridine 21
A solution of N-chlorosuccimide (1.36 g, 10.2 mmol) in
DCM (40 mL) was added dropwise to solution of 6-amino-
2-picoline (1.00 g, 9.20 mmol) and dimethyl sulfide (0.75 mL,
10.2 mmol) in DCM (15 mL) at À108C over a 1.5 hour
period. The reaction mixture was stirred for one hour at
À108C and then for an additional hour at room tempera-
ture. Sodium methoxide (25 w/v% in MeOH, 4 mL,
18 mmol) was then added and resultant suspension was
stirred for 10 min after which water (15 mL) was added and
final mixture was stirred for an additional 1.5 h. The organic
layer was separated and the aqueous layer was extracted
with DCM (50 mL). The combined organic phases were
washed with brine (100 mL), dried over MgSO4 and filtered.
The filtrate was concentrated under reduced pressure to
give yellow residue (1H NMR spectroscopy indicated 1:7.5
ratio of starting material to product). This residue was then
redissolved in DCM (10 mL) and added in one portion to
an ice-cold solution of mCPBA (77% max, 2.33 g, 10 mmol)
in DCM (40 mL). The reaction mixture was stirred for
10 min until a green colour became dominant and then di-
methyl sulfide (0.34 mL, 4.61 mmol) was injected into the
solution and the reaction mixture was stirred for a further
5 min. The green solution was then poured into saturated
solution of Na2CO3 (30 mL) and the layers were separated.
The aqueous phase was washed with additional portion of
DCM (25 mL). The organic phases were combined, dried
over MgSO4 and filtered. The solvent was removed under
reduced pressure on the rotary evaporator (water bath of
which was kept at 58C) to give a brown solid. The product
was purified by silica gel chromatography (ethyl acetate:
hexane, 1:2 as eluent). The blue-green fractions were com-
bined and upon solvent removal the colour turned brown-
yellow. The solid (yield: 0.39 g, 35%) could be stored in a
fridge at 48C for up to 2 months. All spectroscopic and ana-
lytical data was identical to those reported.[13]
1
N 24.55) (N.B. H NMR shows DMSO solvent impurity at 1/
50 level).
Preparation of m-Alkoxo-(1-[(2-hydroxy-3,5-di-tert-
butylbenzylidene)-amino]-indan-2-ol)copper(II)
Complex16
Cu(MeCN)4ClO4 (89 mg, 0.27 mmol) was added in one por-
R
tion to a stirred solution of (1S,2R)-1-[(2-hydroxy-3,5-di-tert-
butylbenzylidene)amino]indan-2-ol (100 mg, 0.27 mmol) in
DCM (5 mL). The reaction mixture was stirred for 1 hour
after which it was washed with brine (10 mL). The green or-
ganic phase was collected, dried (MgSO4), filtered, and
concentrated under reduced pressure to give product 11
as a blue-green solid; yield: 74 mg (63%); mp 98–1008C.
Synthesis of N,N’-Bis(6-methyl-pyridin-2-yl)-diazene
N-Oxide 26
Anal.
found
(%):C
66.49,
H
7.48
N
2.49
A solution of N-chlorosuccimide (1.48 g, 11.1 mmol) in
DCM (50 mL) was added dropwise to an ice-cold solution
of 6-amino-2-picoline (1.00 g, 9.20 mmol) and dimethyl sul-
fide (0.82 mL, 11.1 mmol) in DCM (15 mL). The reaction
(C48H58Cu2N2O4·hexane·0.5DCM requires C 66.61, H 7.49,
N 2.85); IR: nmax =2958 (s), 1647(vs), 1623 (vs), 1528 (m),
1459 (m), 1432 (m), 1385 (m), 1362 (m), 1322 (m), 1271 (m),
1255 (m), 1168 (m); 1053 (m), 749 (m), 535 cmÀ1 (m); HR-
880
asc.wiley-vch.de
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 869 – 882

Mechanistic Insights into Transition Metal-Catalysed Oxidation of a Hydroxamic Acid
FULL PAPERS
MS (ES): m/z=853.3055 [M+H]+ [C48H58Cu2N2O4 requires
Reaction between PPh3 and Nitrosotoluene 10
853.3061 (63Cu)].
Triphenylphosphine (43 mg, 0.16 mmol, 1.0 equiv.) was
added to a solution of 2-nitrosotoluene (20 mg, 0.16 mmol,
1.0 equiv.) in DCM-d2 (0.75 mL) in an NMR tube at room
temperature. The reaction was analysed by 31P NMR. After
5 min, the peak corresponding to free phosphine disap-
peared and peak corresponding to phosphine oxide ap-
peared. The reaction mixture was filtered through a pad of
silica and solvent was removed under vacuum to give triphe-
nylphosphine oxide (43 mg, 93%) as an off-white solid. 2-
Typical Procedure: NMR-Based Oxidation of
Hydroxamic Acid 1 and Nitroso-Diels–Alder
Reaction
Hydroxamic acid 1 (15 mg, 0.088 mmol) and catalyst 4
(0.7 mg, 0.88”10À3 mmol) were dissolved in DCM-d2
(0.75 mL). Diene 2 (17 mL, 0.19 mmol) and anisole (9.8 mL,
0.088 mmol) were added into the mixture. At this point a
1H NMR spectrum was taken thus allowing for better shim-
ming. A solution of t-BuOOH in hexane (5.5M, 23 mL,
0.13 mmol) was added instantly and the progress of the reac-
tion was monitored by 1H NMR spectroscopy. The same
procedure was used to follow the reactions with other
dienes and the intramolecular case. The product can be iso-
lated a following previously reported procedure.[5b] The en-
antiomers could be separated by chiral HPLC (Chiracel
OD, 10% EtOH in hexane, flow 1 mLminÀ1, Rt1 =11 min,
Rt2 =13 min or 10% IPA in hexane, flow 1 mLminÀ1, Rt1 =
19 min, Rt2 =21 min).
1
Methylanisole was also detected by H NMR spectroscopy,
however it was not isolated.
Typical Procedure for the Oxidation of Sulfilimine 24
and in situ Nitroso-Diels–Alder
A solution of catalyst 15 (5.4 mg, 1 mol%), cyclohexadiene
(62 mL, 0.65 mmol, 1.1 equiv.) and sulfilimine 24 (100 mg,
0.59 mmol, 1.0 equiv.) in acetone (2 mL) at room tempera-
ture was treated with an aqueous solution of 33% w/v H2O2
(72 mL, 0.71 mmol, 1.2 equiv.). The reaction mixture was
stirred for 16 h and then was diluted with DCM (10 mL).
The organic phase was washed with brine, dried over
MgSO4 and concentrated under vacuum. The residue was
purified by silica gel chromatography (petroleum ether:ethyl
acetate, 4:1 as eluent) and product isolated as a white solid
(yield: 49 mg, 42%) with analytical data matching that pre-
viously reported.[13]
Typical Procedure: NMR-Based Diels–Alder
Reaction of Nitrosotoluene
Nitrosotoluene 10 (15 mg, 0.12 mmol) and anisole (13.5 mL,
0.12 mmol) were dissolved in DCM-d2 (0.75 mL). At this
point a 1H NMR spectrum was taken thus allowing for
better shimming. Diene (12 mL, 0.12 mmol) was added in-
stantly and the progress of the reaction was monitored by
1H NMR spectroscopy. The same procedure was used to
follow the reactions with ruthenium complexes in it. Enan-
tiomers could be separated by chiral HPLC (Chiracel OK,
10% IPA in hexane, flow 1 mLminÀ1, Rt1 =18 min, Rt2 =
24 min or Chiracel OJ, 10% IPA in hexane, flow 1 mLminÀ1,
Rt1 =19 min, Rt2 =29 min)
X-Ray Crystallography
An orange single crystal of 26 (0.15”0.15”0.20 mm) was an-
alysed at 100(2) K on a Nonius Kappa CCDdiffractometer
using graphite monochromated Mo-Ka radiation (l=
0.71073 Š). Structure solution was carried out using
SHELXS-97[21] and refined by full matrix least squares in
SHELXL-97.[22] All non-hydrogen atoms were refined with
anisotropic displacement parameters. Hydrogen atoms were
positioned geometrically with isotropic displacement param-
Typical Procedure for the Cycloaddition of
Nitrosopyridine 21 to Cyclohexadiene 2
À
eters fixed to ride on the parent atom [aromatic C H
À
0.94 Š, Uiso =1.2”Ueq(C); methyl C H 0.98 Š, Uiso =1.5”
A Schlenk tube was charged with Cu(II)(OTf)2 (4.4 mg,
G
Ueq(C)]. The asymmetric unit consisted of one molecule.
The oxygen atom, O1, is modelled over two sites (N2 or
N3) with a ratio of 35:65; the NMR data supports this asser-
tion, indicating that there is only one oxygen atom per mol-
ecule. Selected bond lengths and angles are listed in Table 3.
Crystal data for 26: M=228.26, monoclinic space group P21,
0.012 mmol) and R,R-BINAP (11.1 mg, 0.012 mmol). DCM
(2 mL) was added to the reaction vessel and the mixture
was stirred at room temperature until all solid dissolved. (In
cases where catalysts were pre-made, they were just dis-
solved in DCM). The reaction mixture was then cooled to
À708C using a cryostat and a solution of nitrosopyridine 1
(15 mg, 0.12 mmol) in DCM (0.5 mL) was added dropwise
to the reaction mixture. Upon addition the colour of the re-
action changed to dark blue then to brown (also depends on
the metal centre). The reaction mixture was stirred for addi-
tional 10 min and then solution of cyclohexadiene
(0.012 mL, 0.12 mmol) in DCM (0.5 mL) was added drop-
wise over a period of 1 hour. The reaction was maintained
between À70 and À558C for the period of 24 h during
which samples were collected, filtered through a pad of
silica and analysed by HPLC (Chiracel OJ, 10% IPA in
hexane, flow 1 mLminÀ1, Rt1 =27 min, Rt2 =39 min). In cases
where product was purified by silica gel chromatography
(hexane:ethyl acetate, 3:1 as eluent; Rf =0.7) the spectro-
scopic data matched those previously reported.[13]
a=6.2863(1),
112.623(1)8, V=561.77(2) Š3, Z=2, 1cald =1.349 Mg/m3, m=
0.0091 mmÀ1,
(000)=240, 19627 reflections collected
b=12.3951(3),
c=7.8106(2) Š,
b=
F
ACHTREUNG
(2.83ꢀqꢀ30.518), 1778 independent reflections (Rint
=
0.0510) were used for structure refinement, final R1 =0.0412
Table 3. Selected bond angles and lengths for structure 26.
Bond Length [Š]
Angle [8]
À
À
À
C(1) N(2)
1.459(3)
C(1) N(2) N(3)
115.3(2)
115.1(2)
129.7(2)
À
À
À
C(7) N(3)
1.453(3)
1.280(2)
1.329(3)
C(7) N(3) N(2)
À
À
À
N(2) N(3)
N(2) N(3) O(1 A)
ACHTREUNG
À
N(3) O
A
Adv. Synth. Catal. 2008, 350, 869 – 882
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
881

Judith A. K. Howard et al.
FULL PAPERS
[F2 >2s(F)], wR2 =0.1043 [F2 > 2s(F)], R1 =0.0492 (all
data), wR2 =0.1088 (all data), GOF (F2)=1.036, largest
peak, hole=0.310, À0.203 eŠÀ3. PLATON[23] suggested the
possible presence of a c-glide, however upon examining the
reflection file significant intensities were found for the h0L
reflections with (h+L) odd, which would be absent if the c-
glide were real. CCDC reference number CCDC 663249.
[9] a) M. E. Marmion, K. J. Takeuchi, J. Am. Chem. Soc.
1988, 110, 1472–1480; b) C. Che, W. Tang, W. Wong, T.
Lai, J. Am. Chem. Soc. 1989, 111, 9048–9056.
[10] a) T. S. Siddall, N. Miyaura, J. C. Huffman, J. Kochi, J.
Chem. Soc. Chem. Commun. 1983, 1185–1186; b) E. G.
Samsel, K. Srinivasan, J. K. Kochi, J. Am. Chem. Soc.
1985, 107, 7606–7617.
[11] BatchCADv8, http://www.aspentech.com/.
[12] S. E. Schaus, J. Brnalt, E. N. Jacobsen, J. Org. Chem.
1998, 63, 403–405.
[13] Y. Yamamoto, H. Yamamoto, J. Am. Chem. Soc. 2004,
126, 4128–4129.
[14] K. Mashima, T. Hino and H. Takaya, J. Chem. Soc.
Dalton Trans. 1992, 2099–2107.
[15] S. Juge, J. P. Genet, S. Mallari, PCT Int. Appl. WO
Acknowledgements
We thank Professor Ken Shea (University of California,
Irvine) for providing compound 8 and for helpful discus-
sions, Dr Paul Low (Durham University) for assistance with
cyclic voltammetry, and the EPSRC, Swansea Mass Spec-
trometry service.
9102588A1 19910307, 1991.
[16] a) E. C. Taylor, C. Tseng, J. B. Rampal, J. Org. Chem.
1982, 47, 552–555; b) S. Huang, D. Swern, J. Org.
Chem. 1979, 44, 2510–2513.
References
[17] J. Lee, L. Chen, A. H. West, G. B. Richter-Addo,
Chem. Rev. 2002, 102, 1019–1065.
[1] G. W. Kirby, Chem. Soc. Rev. 1977, 6, 1–24.
[2] a) O. Wichterle, Collect. Czech. Chem. Commun. 1947,
##43##12, 292; b) Y. A. Arbuzov, Dokl. Akad. Nauk
SSSR 1948, 60, 993–996; c) J. Hammer, A. Ahmad,
R. E. Holliday, J. Org. Chem. 1963, 28, 3034–3036;
d) G. Kresze, J. Firl, H. Zimmer, U. Wollnik, Tetrahe-
dron 1964, 20, 1605–1611.
[3] a) H. Yamamoto, N. Momiyama, Chem. Commun.
2005, 3514–3525; b) Y. Yamamoto, H. Yamamoto, Eur.
J. Org. Chem. 2006, 2031–2043.
[18] a) D. B. Sams, R. J. Doedens, Inorg. Chem. 1979, 18,
153–156; b) S. Clamp, N. G. Connelly, J. A. K. Howard,
I. Manners, J. D. Payne and W. E. Geiger, J. Chem. Soc.
Dalton Trans. 1984, 1659–1665.
[19] a) W. P. Griffith, S. V. Ley, G. P. Whitcombe, A. D.
White, J. Chem. Soc. Chem. Commun. 1987, 1625–
1627; b) I. E. Markó, P. R. Giles, M. Tsukazaki, I.
ChellØ-Regnaut, C. J. Urch, S. M. Brown, J. Am. Chem.
Soc. 1997, 119, 12661–12662.
[20] a) N. Momiyama, H. Yamamoto, Org. Lett. 2002, 4,
3579; b) Y. Hayashi, J. Yamaguchi, T. Sumiya, K.
Hibino, M. Shoji, J. Org. Chem. 2004, 69, 5966–5973;
c) G. Zhong, Chem. Commun. 2004, 606–607; d) G.
Zhong, Angew. Chem. 2003, 115, 4379–4382; Angew.
Chem. Int. Ed. 2003, 42, 4247–4250; e) Y. Yamamoto,
N. Momiyama, H. Yamamoto, J. Am. Chem. Soc. 2004,
126, 5962–5963; f) N. Momiyama, H. Yamamoto, J.
Am. Chem. Soc. 2003, 125, 6038–6039.
[4] a) L. H. Dao, J. M. Dusk, D. Mackay, K. N. Watson,
Can. J. Chem. 1979, 57, 1712–1719; b) N. E. Jenkins,
R. W. Ware, Jr., R. N. Atkinson, S. B. King, Synth.
Commun. 2000, 30, 947–953.
[5] a) K. R. Flower, A. P. Lightfoot, H. Wan, A. Whiting,
Chem. Commun. 2001, 1812–1813; b) A. P. Lightfoot,
R. G. Pritchard, H. Wan, J. E. Warren, A. Whiting,
Chem. Commun. 2002, 2072–2073; c) J. A. K. Howard,
G. Ilyashenko, H. A. Sparkes, A. Whiting, Dalton
Trans. 2007, 2108–2111.
[21] G. M. Sheldrick, Acta Crystallogr. Sect. A 1990, 46,
467–473.
[6] S. Iwasa, K. Tajima, S. Tsushima, H. Nishiyama, Tetra-
[22] G. M. Sheldrick, Computer Program for Crystal Struc-
ture Refinement, University of Gçttingen.
[23] A. L. Speck PLATON, A Multipurpose Crystallograph-
ic Tool, Utrecht University, Utrecht, The Netherlands,
1998.
hedron Lett. 2001, 42, 5897–5899.
[7] K. R. Flower, A. P. Lightfoot, H. Wan, A. Whiting, J.
Chem. Soc. Perkin Trans. 1 2002, 2058–2064.
[8] C. P. Chow, K. J. Shea, J. Am. Chem. Soc. 2005, 127,
3678–3679.
882
asc.wiley-vch.de
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2008, 350, 869 – 882