Modelling the site of bromide binding in vanadate-dependent bromoperoxidases

Article abstract of DOI:10.1039/c2dt12287a

Treatment of Boc-protected (S)-serine (Ser) methyl ester with triphenylphosphine bromide Ph₃PBr (intermittently generated from PPh₃ and N-bromosuccinimide) yields Boc-3-bromoalanine (R)-Boc-BrAlaMe and, after deprotection, bromoalani

Full text of DOI:10.1039/c2dt12287a

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
Dalton
Transactions
Cite this: Dalton Trans., 2012, 41, 5225
www.rsc.org/dalton
PAPER
Modelling the site of bromide binding in vanadate-dependent
bromoperoxidases†
Verena Kraehmer and Dieter Rehder*
Received 29th November 2011, Accepted 15th February 2012
DOI: 10.1039/c2dt12287a
Treatment of Boc-protected (S)-serine (Ser) methyl ester with triphenylphosphine bromide Ph PBr
3
(
intermittently generated from PPh and N-bromosuccinimide) yields Boc-3-bromoalanine (R)-Boc–
3
BrAlaMe and, after deprotection, bromoalanine methyl ester (R)-BrAlaMe in the form of its
hydrobromide. Boc–BrAlaMe and BrAlaMe have been structurally characterised. The reaction between
2
+
BrAlaMe, salicylaldehyde (sal) and VO results in the formation of Schiff base complexes of
+
composition [VO(sal–BrAlaMe)solv] (solv = CH OH: 3, THF: 5) and [VO(sal–BrAla)THF] 4. DFT
3
calculations of the structures of 3, 4 and 5, based on the B3LYP functional and employing the triple zeta
basis set 6-311++g(d,p), provide distances Br⋯V = 4.0 ± 0.1 Å, if some distortion of the dihedral angle
∠
N–C–C–Br is allowed (affording a maximum energy of ca. 45 kJ mol−1), and thus model Br⋯V
distances detected by X-ray methods in bromoperoxidases from the marine algae Ascophyllum nodosum
and Corallina pilulifera. The DFT calculations have been validated by comparing calculated and found
V
structures, including the new complex [V O(Amp–sal)OMe(MeOH)] (1, Amp is the aminophenol
moiety) and the known complex [VO(L-Ser–van)H O] (van = vanillin). Additional validation has been
2
undertaken by checking experimental against calculated (BHandHLYP) EPR spectroscopic hyperfine
coupling constants. Complexes containing bromine as a substituent at the phenyl moiety of a Schiff base
ligand do not allow for an appropriate simulation of the Br⋯V distance in peroxidases. The closest
agreement, d(Br⋯V) = 4.87 Å, is achieved with [VO(3Br–salSer)THF] (6), where 3Brsal–Ser is the
dianionic Schiff base formed between 3-Br-5-NO -salicylaldehyde and serine.
2
Introduction
amino acids, including Ser483. Intermolecular contacts include
Br⋯O(–V) = 3.6, Br⋯N(Arg) = 3.0, and (V–)O⋯Ser = 2.7 Å.
In the A. nodosum enzyme, the respective amino acids are
Arg349/480 and Ser416. In the solvent-exposed surface tyro-
sines of the bromoperoxidase from Cor. pilulifera, additional
bromine has been found, which is covalently bonded to the
Vanadate-dependent haloperoxidases are present in marine algae,
terrestrial fungi and lichen. They contain vanadate OvV
−
(
OH) O linked to the protein matrix via the N of a histidine.
2
ε
The enzymes, and their functional mimics (i.e. specifically
designed vanadium coordination compounds) catalyse the halo-
genation of organic substrates by reverting to halide, employing
4
phenyl moiety of tyrosine residues. Mono- and dibrominated
surface tyrosines have also been detected by EXAFS investi-
1
5,6
hydrogen peroxide as an oxidant. Bromoperoxidases employ
gations in the A. nodosum bromoperoxidase.
bromide; the brominating species generated by the oxidant,
While bromination of tyrosine and other aromatic and pseudo-
aromatic residues (such as phenylalanine and histidine) is easily
+
−
−
{
Br }, can be Br , BrO /HOBr or Br . Structure information is
2 3
+
available for the bromoperoxidases from the brown alga Asco-
achieved by electrophilic attack of a {Br } species formed from
2
3
−
phyllum nodosum, and the red algae Corallina officinalis and
Br and H O , the generation of an appropriate organic bromine
2
2
4
Cor. pilulifera (wild-type, and the mutant Arg397Trp). The
species in the absence of peroxide is less straightforward. In pre-
vious studies based on XAS we have shown that in genuinely
bromine-free A. nodosum bromoperoxidase, when soaked with
bromide in the absence of peroxide, EXAFS features arise which
can be fitted to a bromine–carbon distance d(Br–C) = 1.88 Å, a
second sphere distance d(Br⋯C) = 2.67 Å, and a bromine–
structure of the Cor. pilulifera enzyme in particular provides
−
information on the location of bromide: Br resides close to the
vanadate centre and interacts with Arg397 and other surrounding
7
–9
Chemistry Department, University of Hamburg, D-20147 Hamburg,
Germany. E-mail: rehder@chemie.uni-hamburg.de
vanadium spacing d(Br⋯V) = 3.96–4.15 Å.
The range of d
2
(
Br–C) for Br bonded to an aromatic sp carbon spans
that for d(Br–C) involving sp carbons
.92–1.96 Å, extending to 1.87–1.99 Å for carbon in the β-pos-
†
Electronic supplementary information (ESI) available. CCDC 855090
6,9
3
1
1
.88–1.92 Å,
(Boc–BrAlaMe) 855112 (BrAlaMe) and 855451 (1). For ESI and crys-
6
tallographic data in CIF or other electronic format see DOI: 10.1039/
c2dt12287a
7
a
ition of an amide bond. In addition, steric constraints within
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 5225–5234 | 5225

View Article Online
the active site protein pocket are expected to give rise to unprece-
dented bond lengths variations. Although, on first sight, d(Br–C)
work) and [VO(Ser–van)(H O)] 2 (Ser–van is the Schiff base
2
10
formed between serine and vanillin). The suitability of the
B3LYP functional (employed in the current work) for predicting
structure parameters and energies has recently been demonstrated
for a an impressive number of transition metal complexes,
=
1.88 Å more appropriately represents Br bonded to an aromatic
framework, we have favoured – in the light of the unavailability
+
of a {Br } species and an electrophilic substitution path –
11
linkage of bromine to an aliphatic carbon of one of the active
including those of vanadium. Complexes addressed in the
present paper are collated in Schemes 1 and 2.
site amino acids, and proposed serine (Ser416 in the A. nodosum
7–9
−
−
enzyme) as a likely candidate. Exchange of the OH for Br
converts serine to bromoalanine, with Br residing on Cβ.
Here, we provide evidence, based on a rational synthesis of
R)-β-bromoalanine (BrAla), for the enantioselective conversion
Results and discussion
(
of (active site) (S)-serine to BrAla. The oxidovanadium(IV) com-
plexes of the Schiff base ligand formed between salicylaldehyde
and BrAla are employed as model compounds to mimic active
site Br⋯V distances under somewhat constrained conditions. In
a second set of model complexes, Schiff bases obtained by con-
densation of serine or phenylalanine with salicylaldehydes bro-
minated in the ring positions 3 or 6 are introduced. Structurally
Bromoalanine: preparation, characterisation and introduction
into the coordination sphere of vanadium
In view of an unavailable rational synthesis for (R)-β-bromoala-
nine (BrAla) from (S)-serine (Ser) – the likely precursor for a
putative BrAla at the enzyme’s active site – we have devised a
respective and efficient reaction course, Scheme 3. In a first step,
the Boc-protected serine methyl ester Boc–SerMe (Boc = tert-
butyloxycarbonyl) is prepared from serine methyl ester hydro-
chloride (SerMe) by deprotonation of SerMe with triethylamine
(by XRD) characterised Schiff base complexes have been
employed to validate structure data obtained for complexes on
the basis of DFT calculations, exemplified here for the com-
plexes [VO(Amp–sal)(OMe)MeOH] 1 (Amp–sal is the Schiff
base formed between aminophenol and salicylaldehyde; this
12
(
TEA) and concomitant reaction with Boc O. The alcoholic
2
function of serine is then replaced by bromine via treatment with
triphenyl phosphine plus N-bromosuccinimide. In this so-called
13
Mukaiyama redox condensation,
intermittently formed
Ph PBr, containing the oxophilic Ph P moiety, is the actual bro-
3
3
minating agent: nucleophilic attack of bromide to Boc–SerMe
and concomitant formation of Ph PvO yields the bromoalanine
3
derivative Boc–BrAlaMe. Deprotection of Boc–AlaMe is then
achieved with acetyl bromide in methanol. The bromoalanine
methyl ester hydrobromide (BrAlaMe) thus formed can be con-
verted to the neutral N-acetylated bromoalanine methyl ester Ac–
14
BrAlaMe by treatment with acetyl bromide in methanol.
The target compounds exhibit, in the EI-MS, the m/z peaks
7
9
81
Scheme 1 Structurally characterised Schiff base vanadium complexes
chosen for the validation of structure data obtained from DFT calcu-
lations. For structure data of 2, see ref. 10.
typical of the two bromine isotopomers ( Br, Br) at 281/283
+
+
([M] ) for Boc–BrAlaMe, 181/183 ([M − HBr] ) for BrAlaMe,
and 223/225 ([M] ) for Ac–BrAlaMe. The C DEPTQ NMR
+
13
Scheme 2 Vanadium complexes with Schiff bases containing bromoalanine, serine or phenylalanine in the backbone. Counter-ion for the cationic
complexes 3 and 5 is acetate.
5226 | Dalton Trans., 2012, 41, 5225–5234
This journal is © The Royal Society of Chemistry 2012

View Article Online
Scheme 3 The preparative route to bromoalanine derivatives. For details and abbreviations cf. text. Central step in this reaction course is the intermit-
tent formation of Ph PBr, allowing for a conversion of Boc–SerMe to Boc–BrAlaMe.
3
−
Fig. 1 Structures of the ligand precursors. See Scheme 3 for abbreviations. For BrAlaMe (counter-ion Br ≡ Br1 omitted), the calculated structure is
also shown.
(BrAlaMe) and 596 cm− (Ac–BrAlaMe), respectively. Boc–
1
signal for the methylene group carrying the bromo substituent
appears at 30.1 ppm; the ν(C–Br) is at 590 (Boc–BrAlaMe), 591
BrAlaMe and BrAlaMe have been characterised structurally by
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 5225–5234 | 5227

View Article Online
Table 1 Selected bond lengths (Å) and angles (°) for bromoalanine
derivatives
The calculations have been carried out with the B3LYP func-
tional, employing the triple zeta basis set 6-311++g(d,p) (Gaus-
sian 03). To validate these calculations, key data are compared
with those obtained from X-ray diffraction studies of the new
complex [V O(Amp–sal)OMe(MeOH)] (1 in Scheme 1) and
several Schiff base complexes with serine as a constituent, such
Boc–BrAlaMe found BrAlaMe found
BrAlaMe calcd
V
Bond lengths
Bond lengths
Bond lengths
Br1–C4 1.986
N1–C1 1.464
O1–C2 1.335
O2–C2 1.206
Br1–C2 1.944(4)
N1–C9 1.448(6)
O4–C5 1.342(5)
O1–C5 1.213(5)
Br2–C2 1.944(3)
N1–C1 1.486(4)
O2–C3 1.317(4)
O1–C3 1.197(3)
IV
as [V O(Ser–van)H O] (2 in Scheme 1), previously prepared in
2
10
our laboratory. We have further validated the calculated struc-
tures by comparing experimental EPR hyperfine coupling con-
stants with those simulated (employing the program system
BHandHLYP) on the basis of calculated bonding parameters.
The V precursor compound [V O(Amp–sal)THF] for the
Bond angles
Br1–C2–C9 110.0(3) Br2–C2–C1 110.9(2)
N1–C9–C2 113.0(3) N1–C1–C2 111.7(2)
Bond angles
Bond angles
Br1–C4–C1 112.95
N1–C1–C4 111.80
C2–C1–C4 110.37
16a
IV
IV
C5–C9–C2 108.4(3)
C3–C1–C2 113.1(3)
V
V complex 1 was prepared by the reaction between VO(acac)2
acac = acetylacetonate(1–)) and the Schiff base obtained from
Dihedral angles
H14–C9/C2–Br1
Dihedral angles
H4–C1/C2–Br2
−176.1
Dihedral angles
H1–C1/C4–Br1 −54.2
(
5
8.7
C5–C9/C2–Br1
75.1
N1–C9/C5–O4 45.4
the condensation of aminophenol and salicylaldehyde in THF–
IV
C3–C1/C2–Br2 −56.0 C2–C1/C4–Br1 66.3
methanol. Aerial oxidation of [V O(Amp–sal)THF] in metha-
1
nol results in the formation of 1. An ORTEP plot of complex 1
is presented in Fig. 2 (top left). Found and calculated bonding
parameters for 1 and 2 are provided in Table 2. Compound 1
N1–C1/C3–O1 118.1
N1–C1/C2–O1
178.8
−
crystallises in the space group P2 /n. Vanadium constitutes a
1
centre of chirality; the unit cell contains two pairs of the C/A
enantiomers, Fig. 2, bottom left. Bond lengths for 1 are in the
expected range. The rather long d(V–OHMe), 2.363(3) Å, is a
consequence of the trans influence exerted by the oxido ligand,
XRD. Fig. 1 provides an overview of the structures, including
the calculated structure for BrAlaMe, and Table 1 contains
selected structure parameters. For crystal and refinement data see
Experimental. The structure of the unprotected serine methyl
17
and has also been observed in other comparable complexes.
15
ester (not shown in Fig. 1) has been published previously.
Larger deviations between calculated and observed bonding
angles reflect influences impaired by packing effects in the
crystal. As for 1, bonding parameters obtained for the exper-
Bond distances and bond angles for Boc–BrAlaMe and
BrAlaMe are essentially identical, and the experimental bond
distances and angles for BrAlaMe also agree with the calculated
ones. On the other hand, the dihedral angles measured and cal-
culated, respectively, for BrAlaMe clearly differ from each other
in particular in the surroundings of the chiral centre C1: While
the carbonyl group and nitrogen are in the same plane (∠O1C2–
C1N1 = 179°) in the solid state structure, the respective angle in
the calculated gas-phase structure is 118°. Further, the bromine
substituent is in a plane perpendicular to the molecular axis
defined by C1, C2 and O1 in the found structure, while the
respective plane is twisted by 86° relative to this axis in the
undisturbed, DFT-based structure. These differences, which
reflect effects arising from the “salt-like” nature – four cationic
10
imental structure of the serine derivative 2 match those which
have been calculated. Deviations in bond lengths amount to
0
.03 Å at the most. The reliability of calculated bonding par-
ameters in the case of compound 2 is supported by the close
agreement (deviation 1.4%) between the calculated and exper-
imental EPR spectroscopic parallel hyperfine constant: |A |
z
−4
−1
(
calcd) = 167.7, |A |(found) = 170.1 × 10 cm .
z
Complexes with BrAla in the coordination sphere. The com-
plexes 3, 4 and 5 (Scheme 2) have been synthesised in acetate-
buffered solution from salicylaldehyde, BrAlaMe and vanadyl
sulphate as depicted in eqn (1). Depending on the solvent, the
cationic complex 3 (isolated from methanol–acetone) or the
neutral complex 4 (isolated from methanol–THF) is recovered.
In 3, the bromoalanine moiety of the Schiff base ligand remains
methylated and consequently coordinates through the carbonyl
oxygen while, in the course of the formation of 4, partial de-
esterification occurs, allowing for coordination of the terminal
carboxylate. In the ESI(+) and ESI(−) MS of the cationic
−
BrAlaMe plus four Br counter ions in the unit cell – and
packing effects in the crystal, minimise the requirement for
space.
Oxidovanadium Schiff base complexes
Validation of calculated structure parameters. We have not
succeeded in obtaining suitable crystals for the structure determi-
nation of vanadium complexes carrying ligands with bromine
substituents, and resorted to DFT calculations for this reason.
complex 3, the mass peaks 417/419 for [M − CH OH − H +
3
+
−
VO] and 283/285 [M − CH OH − H − VO] , respectively, are
present, both deriving from the molecule (M = 383 g mol ).
3
−1
5228 | Dalton Trans., 2012, 41, 5225–5234
This journal is © The Royal Society of Chemistry 2012

View Article Online
Fig. 2 Top left: Structure of the C enantiomer of [VO(Amp–sal)OMe(MeOH)], 1. Bottom left: One of the two pairs of enantiomers present in the
unit cell. Dashed lines indicate intermolecular contacts: The distance between the oxido ligands of the A and C enantiomers is 6.124 Å, the closest
contact between phenyl carbons (belonging to sal and Amp, respectively, of the two enantiomers) amounts to 3.298 Å. Right: Calculated structure for
2
[VO((S)-Ser–van)H O], 2 (for the X-ray structure see ref. 10).
Table 3 Calculated and experimental EPR spectroscopic hyperfine
Table 2 Found and calculated bond lengths (Å) and angles (°) for the
Schiff base complexes [VO(Amp–sal)OMe(MeOH)] 1 and [VO((S)-
−4
−1
coupling constants (×10 cm ) for the complexes 3, 4 and 5 (cf.
Scheme 2) plus minor components formed in the reaction represented by
eqn (1)
2
Ser–van)H O] 2
a
1
Found/calcd
2
Found /calcd
|
A |
x
|A
y
|
z
|A | calcd/
Complex (equatorial functions)
calcd
70.7
61.7
calcd
69.5
60.8
found
V1–O5 [d
1.587(3)/1.572 V1–O1 [d
(VvO)]
1.596(3)/
1.580
1.972(3)/
1.949
1.894(3)/
1.925
(VvO)]
3
(Ophen, Nimine, Oester
,
176.2/177.2
165.7/168.6
V1–O4 [d(V–
OAmp)]
V1–O2 [d(V–
Osal)]
V1–O3 [d(V–
OMe)]
V1–O1 [d(V–
OHMe)]
1.925(3)/1.885 V1–O4 [d(V–
c
Omethanol
)
Ocarb)]
4
O
5
(Ophen, Nimine, Ocarboxylate
THF
(Ophen, Nimine, Oester, OTHF
,
1.869(3)/1.862 V1–O2 [d(V–
Ovan)]
1.763(3)/1.759
)
)
67.6
75.0
61.7
65.2
74.9
67.7
171.7/171.1
182.2/180.6
177.8/179.2
2
+
[
[
VO(MeOH) ]
VO(THF) ]
4
4
2+
2.363(3)
V1–O3 [d(V–
OH )]
1.979(3)/
2.116
2
V1–N1
2.147(4)/2.225 V1–N1
2.022(3)/
2
.041
7
9
81
The two peaks of about equal intensity represent the Br/ Br
C6–O4 [d(V–
Ophe]
1.347(5)/1.295 C9–O4 [d(C–
1.302(4)/
1.318
109.76(7)/
114.32
108.55(7)/
109.80
107.85(7)/
113.68
b
c
Ocarb)]
isotopomers. The FAB-MS spectrum for the neutral complex 4
exhibits a peak at 329 for [M − Br] . Along with 4, an additional
major product, the cationic complex 5, plus a minor product of
∠
V1–O5/O4–
O4/O1
V1–O5/O2–
O4/O3
V1–O5/O3–
O4/O2
100.474(14)/
107.54
102.294(15)/
106.32
99.080(14)/
107.99
∠V1–O1–O2
∠V1–O1–O3
∠V1–O1–O4
∠O1–V1–N1
+
∠
2+
the likely composition [VO(THF)4] are formed. A correspond-
2+
∠
ing (i.e. solvent-derived) minor product, [VO(MeOH) ] , is also
4
observed along with 3. The composition of all of the complexes
is rationalised on the basis of a comparison of the experimental
with the expected (viz. calculated) EPR spectral parameters;
Table 3.
The bond length d(Br–C) in the calculated structures compare
to those which are commonly found in experimental structures.
∠V1–O5/O4–N1 93.519(15)/
103.72(7)/
105.73
9
3.72
a
b
From ref. 10. “phe” refers to the phenolate oxygen of the
c
aminophenol moiety. “carb” refers to the carboxylate oxygen of the
serine moiety coordinating to vanadium.
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 5225–5234 | 5229

View Article Online
In contrast, the distances d(V⋯Br) and d(O⋯Br) (where O is
the oxido ligand) obtained from the calculated structures of 3, 4
and 5 are longer by 20–25% than the corresponding distances
found in bromoperoxidases treated with bromide; Table 4. In the
calculated equilibrium structure of 4 (Fig. 3, left), the dihedral
angle formed between the fragments Br–C12 and C13–N is
(Scheme 2), containing bromo substituents at the salicylaldehyde
moiety of the ligand, have been considered here in order to
evaluate the possibility of bromine activation by intermittent
linkage of Br to a phenyl moiety, such as provided by Phe397 in
18
the fungal Curvularia inaequalis peroxidase. Following a lit-
19
erature procedure, the complexes were prepared in a one-pot
synthesis from brominated 5-nitrosalicylaldehyde, vanadyl sul-
phate and serine in methanol–THF as shown for compound 6 in
eqn (2). Fig. 4 shows the calculated structure for 6, and Table 6
summarises selected structure parameters involving the bromine
substituent, and the EPR spectroscopic coupling constants.
The complexes display non-distorted tetragonal–pyramidal
geometry with THF in one of the equatorial positions. The devi-
ations between calculated and experimental EPR spectroscopic
coupling constants are less than 2% in all three cases, indicating
that the calculated gas phase structure and the structure of the
−
71.05°, and the distance d(V⋯Br) amounts to 5.02 Å. As this
angle becomes narrower (Fig. 3, right), distances (Table 5) are
realised which reproduce those found experimentally. The corre-
sponding distortions of the dihedral angle in the native enzyme
may be caused by constraints in the active site protein pocket;
the additional energy afforded for such a distortion, e.g. 46.5 kJ
−1
mol to adapt to a distance of 3.9 Å, is low.
Complexes with bromophenyl as constituent of the Schiff base
ligand. Complexes 6, and the complex cations 7 and 8
Table 4 Selected experimental and calculated equilibrium distances in
Å
Table 5 Non-equilibrium distances d(V⋯Br) of 4 and corresponding
energies at constrained conditions
Calcd for the
equilibrium dihedral
angle ∠N–C/C–Br =
−71.05°
Dihedral angle ∠N– d(V⋯Br)
Energy difference relative to
1
C/C–Br (°)
(Å)
equilibrium state (kJ mol−
)
Experimental in
bromoperoxidases
a
a
3
4
5
−71.0
5.0
4.1
3.9
3.7
3.6
0
−
24.0
32.9
46.5
57.6
65.5
d(C–Br)
d(V⋯Br)
d(O⋯Br)
1.87–1.99
3.6–4.1
3.6
1.959
5.009
4.844
1.984
5.020
4.840
1.960
5.008
4.813
−13.0
−2.0
+4.0
a
a
For refs. cf. Introduction.
Equilibrium state, cf. Fig. 3.
Fig. 3 Calculated structure of [VO(sal–BrAla)THF] (4) without (left) and with constraints (right). The dihedral angle Br–C13/C12–N is −71.0 (left)
and −2.0° (right).
5230 | Dalton Trans., 2012, 41, 5225–5234
This journal is © The Royal Society of Chemistry 2012

View Article Online
The strong electropositive potential at the bottom region of
4
the active site cleft may invoke bromide to the extent where
the active site serine is converted to bromoalanine, a transform-
ation which we have realised here through the conversion of
serine to β-bromoalanine with triphenylphosphine + N-bromo-
succinimide. Wever et al. have shown that, in the mutant
Ser402Ala of the chloroperoxidase from the fungus Curvularia
inaequalis, the brominating activity of the mutant enzyme
18
decreases, although it still catalyses the oxidation of bromide.
The active sites of the chloro- and bromoperoxidases are very
similar, the main difference being the substitution of a second
histidine in the bromoperoxidase (His411) by a phenylalanine
(Phe397) in the chloroperoxidase. In the light of the brominating
activity of the Ser402Ala mutant of the Cur. inaequalis peroxi-
dase, i.e. the sustenance of some brominating activity in the
absence of serine suggests that active site residues other than –
or in addition to – serine can become involved in bromide
activation.
Fig. 4 Calculated gas-phase structure of [VO(3Br–salSer)THF] 6, with
the serinate moiety in the S and the vanadium centre in the C configur-
ation. Selected bond lengths: V–O1 1.579, V–O2 1.958, V–O3 2.071,
V–O4 1.941, V–N1 2.069, V1⋯Br1 4.874 (red line), O1⋯Br1 5.409 Å.
Experimental
General, instrumentation, and calculations
Table 6 Data (distances d in Å, EPR spectroscopic coupling constants
−4
−1
(absolute values |A| in 10 cm )) for complexes 6, 7 and 8
Where not mentioned otherwise (see syntheses of ligands and
complexes), starting materials were obtained from commercial
sources.
A
x y
/A
A
z
A /A
x y
A
z
d(C–Br) d(Br⋯V) d(Br⋯O) calcd calcd found found
1
IR spectra: Perkin-Elmer FT-IR spectrometer 1720. H NMR
6
7
8
1.907
1.918
1.908
4.874
7.508
4.872
5.409
7.979
5.432
64.7/ 168.3 62.8
2.8
59.9/ 166.1 57.9
7.9
63.4/ 168.3 61.2
2.1
171.3
168.5
171.3
13
and C NMR spectra: Bruker AVANCE 400 and Varian Gemini-
6
2
00 BB. EPR spectra: X-band (9.35 MHz) Bruker Elexsys E500
51
5
CW at 120–100 K, i.e., in glassy frozen solutions. The V NMR
was obtained on a Bruker AVANCE 400 spectrometer at
6
1
05.2 MHz, and is quoted relative to neat VOCl . FAB-MS: VG
3
Analytical mass spectrometer 70–250S, employing argon for
ionisation and m-nitrobenzyl alcohol as a matrix. ESI-MS: Finni-
gan ThermoQuest spectrometer, model MAT 95XL. MS spectra
of organic compounds were obtained by electron impact ionis-
ation with direct inlet. Polarimetry: Krüss polarimeter P8000.
Crystal structures were performed with single crystals, embedded
compounds in solution are essentially identical. The closest
available distance between the bromine substituent and the
vanadium(IV) centre is 4.87 Å in 6 and 8, i.e. in the complexes
where Br is in the meta position with respect to the enamine
linkage of the Schiff base. This distance exceeds that detected by
X-ray methods of the enzymes soaked with bromide by approxi-
mately 0.9 Å; cf. Introduction for details and references.
(under inert gas atmosphere) in viscous paraffin oil with Mo Kα
irradiation (graphite monochromator) on a Bruker Smart Apex
CCD diffractometer at 153 K. Frames were read with SAINT,
absorption corrections were carried out with SADABS, and
20
refinements with SHELXL 97. All non-hydrogen atoms were
refined with anisotropic temperature factors. Hydrogen atoms
were either calculated into idealised positions, or found and
refined isotropically. Table 7 summarises crystal structure data
and refinement data, and also provides CCDCs.
Conclusion
We have shown that the Br⋯V distance of ca. 4 ± 0.1 Å found
by X-ray methods in bromoperoxidases from the algae Ascophyl-
lum nodosum and Corallina pilulifera can be modelled by oxido-
vanadium(IV) complexes of Schiff base ligands containing
β-bromoalanine as a constituent, if some flexibility under con-
strained conditions is admitted. Schiff bases where the aromatic
salicylaldehyde moiety is brominated are insufficiently flexible
to allow for a satisfactorily close contact between vanadium and
bromide. In the active site pocket, where serine (Ser416 in the A.
nodosum enzyme) is directed towards the vanadium centre by
virtue of the protein folding, a sufficiently close Br⋯V contact
can be provided per se.
DFT calculations have been carried out with the programme
21
22
Gaussian 03 using the hybrid functional B3LYP for geome-
try optimisation, and the basis set 6-311g++d,p (for complexes)
and 6-311g++3d3p (for organic compounds). Based on this
basis set and employing the half-and-half functional BHandH-
LYP, the isotropic EPR spectroscopic hyperfine coupling con-
stants Aiso and the tensor components T , T , T were
x
y
z
16
calculated. The directional components of A were then calcu-
lated according to A = A + T (n = x, y, z), and are quoted as
n
iso
n
absolute values, |A|.
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 5225–5234 | 5231

View Article Online
Table 7 Crystal and refinement data for Boc–BrAlaMe, BrAlaMe, and
calc. for C H BrNO (282.13): C, 38.31; H, 5.72; N, 4.96.
9
16
4
[VO(Amp–sal)OMe(MeOH)] (1)
−1
Found: C, 38.11; H, 5.70; N 5.01%. IR (KBr, cm ): ν(N–H)
390; ν(C–H) 2983; ν(CvO) (ester) 1716; ν(CvO) (carba-
3
[
VO(Amp–sal)
1
Boc–BrAlaMe
BrAlaMe
Br NO
OMe(MeOH)]
mate); ν(C–Br) 619. H NMR (CDCl ) δ: 5.41 (NH, 1H, m);
3
4
.72 (N–CH, 1H, m); 3.70 (O–CH3 and CH , 5H, m); 1.43
2
Empirical
formula
M/g mol
C
9
H
16BrNO
4
C
4
H
9
2
2
C
15
H
16NO
5
V
13
3 3
(CH , 9H, s). C NMR (CDCl ) δ: 169.6 (OvC–O), 154.9 (N–
−
1
CvO), 80.4 (C(CH ) ), 53.8 (CH), 52.8 (O–CH ), 34.0
3 3 3
282.13
Monoclinic,
262.93
Orthorhombic,
P2 (no. 19)
341.23
Monoclinic,
P2 /n (no. 14)
+
Crystal
(CH Br), 28.2 (C(CH ) ). MS, m/z: 281 [M] ; 224 [M −
2
3 3
+ + +
system, space
group
a/pm
b/pm
c/pm
P2
1
(no. 4)
1
2
1
2
1
1
C H O ] ; 209 [M − C H O] ; 181 [M − C H O ] ; 57 [M −
2
3
2
4
9
5 9 2
+
C H BrNO ] . Single crystals were obtained by slow evapor-
5
7
4
9.320(8)
5.102(8)
13.905(10)
108.30(10)
2
627.79(2)
1.493
5.253(8)
7.1797(17)
21.233(5)
19.471(5)
99.801(5)
8
2925.0(12)
1.550
ation of CHCl solutions at room temp. For crystal and refine-
11.825(8)
13.896(10)
—
3
ment data see Table 7.
β/°
(b) (R)-3-Bromoalanine methyl ester hydrobromide
Z
V/Å
4
3
(BrAlaMe): Boc–BrAlaMe (10 g, 35.4 mmol), dissolved in
100 mL of abs. methanol, was treated dropwise during two
hours with freshly distilled acetyl bromide (8.7 g, 70.9 mmol)
and stirred at room temp. for three days. Vacuum removal of the
solvent yielded a residue which was treated with 150 mL of abs.
acetone. The colourless residue was filtered off, washed with
acetone and dried in vacuo. Yield: 4.74 g (10.07 mmol; 51%) of
863.17(2)
2.023
9.33
−
3
ρ
calcd/g cm
−
1
μ/mm
3.27
0.701
F(000)
288
0.40 × 0.10 ×
0.10
504
0.50 × 0.07 ×
0.07
1408
0.50 × 0.07 ×
0.07
Crystal size/
3
mm
θ range/°
Index range
2.30–27.48
−7 < h < 11,
2.30–25.00
−6 < h < 6, −15
1.92–27.00
−9 < h < 8,
−
6 < k < 6,
< k < 15, −18 < l −24 < k < 27,
−1
colourless, amorphous BrAlaMe. IR (KBr, cm ): ν(N–H) 3474;
−18 < l < 17
< 18
10 541
−24 < l < 17
18 756
Reflections
collected
4139
ν(C–H) 2951;ν(CvO) (ester) 1746; ν(C–Br) 591. Anal. calc. for
C H Br NO (262.93): C, 18.27; H, 3.45; N, 5.33. Found: C,
4
9
2
2
Independent
reflections
2637 (0.0306)
1971 (0.0388)
6295 (0.0973)
1
1
8.32; H, 3.34; N, 5.21%. H NMR (CD OD) δ: 8.70 (3H,
3
NH , 3H, br s), 4.86 (HC–N, 1H, m), 3.88 (CH and O–CH ,
5H, m). C NMR (CD OD) δ: 167.2 (OvC–O), 53.4 (C–N),
3
+
(
R
1
int
)
3
2
3
1
3
R
, wR
σ(I )]
, wR
data)
2
2
[I <
0.0393, 0.0921 0.0219, 0.0477
0.0981, 0.1062 0.0241, 0.0477
0.0581, 0.0950
0.1397, 0.1116
+
2
0
5
2.6 (O–CH ), 30.72 (CH ). MS, m/z: 181 [M] , 122 [M −
3
2
R
1
(all
+
C H O ] . [α] (20 °C, CH OH) = +3°. Single crystals were
2
3
2
D
3
obtained by slow evaporation of CH OH solutions at room temp.
Residual
electron
density/e Å
Flack
parameter
CCDC no.
0.678 and
−0.305
0.549 and
−0.286
0.509 and
−0.532
3
For crystal and refinement data see Table 7.
−3
(c) N-Acetyl-(R)-3-bromoalanine methyl ester (Ac–BrAlaMe):
0.0438
855090
−0.0088
—
BrAlaMe (1.1 g, 3.8 mmol) was dissolved in 10 mL of 0.1 M
ammonium hydrogencarbonate (pH = 8) and treated dropwise
with 50 mL of a mixture of acetanhydride–methanol 1 : 3. After
855112
855451
2
hours of stirring at room temp., the solvents were removed
with a rotary evaporator to yield a yellowish liquid of Ac–
BrAlaMe (0.77 g, 3.49 mmol = 93% yield). Anal. calc. for
C H BrNO (224.05): C, 32.16; H, 4.50; N, 6.25. Found: C,
Syntheses of ligands and complexes
6
10
3
Esters of bromoalanine. Confer also Scheme 3. (1) (S)-Boc–
serine methyl ester (Boc–SerMe): (S)-Serine methyl ester hydro-
chloride (SerMe) was prepared by treatment of (S)-serine with
thionyl chloride in absolute methanol. Treatment of a suspension
of SerMe in CH Cl with triethyl amine, followed by bis(tert-
−1
3
2
5
2.30, H, 4.41; N, 6.21%. IR (KBr, cm ): ν(NH) 3436; ν(CH)
972; ν(CvO) (ester) 1741; ν(CvO) (amide) 1673; ν(C–Br)
1
97. H NMR (CDCl ) δ: 8.25 (N–H, 1H, s), 4.86 (C–H, 1H,
3
m), 3.91 (m, CH , 2H, m) 3.72 (O–CH , 3H, s), 1.92 (CH , 3H,
2
3
3
2
2
13
s). C NMR (CDCl ) δ: 169.2 (O–CvO), 164.5 (N–CvO),
3
butyl)dicarbonate (Boc O) yielded a viscous residue, from
2
5
2
9.4 (CH), 54.5 (O–CH ), 33.8 (CH ), 25.2 (CH ). MS, m/z:
3 2 3
which Boc–SerMe was obtained by successive extractions with
ethyl acetate, followed by aqueous 1 M KHSO , 1 M NaHCO3
+
+
+
23 [M] ; 164 [M − C H O ] ; 122 [M − 101] .
2
3 2
4
and saturated aqueous NaCl.
Methanol-methoxido-oxido-{N-2-(salicylideneamino)phenolato}-
vanadium(V), [VO(Amp–sal)OMe(MeOH)] (1). Vanadyl acetyla-
cetonate (1.31 g, 4.72 mmol) and salicylideneaminophenol
1.00 g, 4.72 mmol) were dissolved in 50 mL of hot methanol.
(2) Bromoalanine derivatives: (a) (R)-3-Boc–bromoalanine
methyl ester (Boc–BrAlaMe): Boc–SerMe (17 g, 77.5 mmol)
and triphenylphosphine (25.8 g, 98.5 mmol) were dissolved in
(
2
00 mL of abs. THF. The solution was cooled in an ice bath, and
successively (within one hour) treated with N-bromosuccinimide
17.5 g, 98.5 mmol). The dark brown suspension was stirred
After 24 hours of stirring, the hot solution was filtered through
glass wool to remove all black particles. Slow evaporation of the
solvent yielded 1.33 g (3.91 mmol, 91% yield) of light-brown,
crystalline 1. Anal. calc. for C H NO V (314.24): C, 52.80;
(
overnight at room temp., filtered, and the filtrate evaporated to
dryness. The residue was redissolved in hexane–ethylacetate
15
16
5
H, 4.73; N, 4.0. Found: C, 52.71; H, 4.83; N, 4.02%. IR (KBr,
1
: 1, filtered, and the filtrate purified by passage through a
−1
cm ): ν(CH) 3038, 2917, 2831; ν(CvN) 1615; ν(C–C) 1442;
column of silica gel 60 μm (elutant: hexane–ethylacetate 1 : 1).
After removal of the solvents in vacuo, 19.6 g (89.5% yield) of
light-brown, crystalline Boc–BrAlaMe was recovered. Anal.
1
ν(CO) 1296, 1146; ν(VvO) 982. H NMR (CD OD) δ: 9.23
3
(
(
x, 1H, s), 7.72 (1H, dd), 7.69 (1H, dd), 7.62 (1H, m), 7.32
1H, m), 7.14 (1H, m), 6.99 (1H, d), 6.79 (1H, dd), 3.78
5232 | Dalton Trans., 2012, 41, 5225–5234
This journal is © The Royal Society of Chemistry 2012

View Article Online
−
(
CH OH, 1H, s), 3.48 (CH O , 3H, s), 3.32 (CH OH, 3H, s).
the filtrates to dryness in vacuo. 3, with acetate as the counter-
ion, was recovered as a green amorphous solid with a yield of
90%. Anal. calc. for 3[CH CO ]·H O (461.16): C, 36.46; H,
4.37; N, 3.04. Found: C, 36.26; H, 4.30; N, 3.11%. For com-
pounds 4 and 5, the residue obtained after evaporation was re-
dissolved in 50 mL of abs. THF and stirred under reflux over-
night. After filtration and evaporation of the solvent, a mixture
of 4 and 5 (counter-ion: acetate) was isolated in the form of a
green amorphous solid. IR, EPR and ESI-MS data for 3 and 4
3
3
3
1
3
C NMR (CD OD) δ: 1614 (CvN); 160.5, 148.5, 144.6,
39.3, 134.9, 131.7, 131.1, 128.7, 122.2, 121.0, 118.1, 117.5
3
1
(
(
3
2
2
−
51
12 × aromatic C); 53.6 (CH O ), 52.0 (CH OH). V NMR
3
3
CD OD) δ: −532.4. For crystal and refinement data, see
3
Table 7.
+
[
VO(sal–BrAlaMe)MeOH] (3), [VO(sal–BrAla)THF] (4), and
VO(sal–BrAlaMe)THF] (5). All operations were carried out in
oxygen-free solvents. VOSO ·5H O (3: 0.96 g, 3.80 mmol; 4
and 5: 0.32 g, 1.27 mmol) and sodium acetate trihydrate (3:
.29 g, 9.50 mmol; 4 and 5: 0.43 g, 3.17 mmol) were dissolved
in 50 mL of methanol and treated with salicylaldehyde (3:
+
[
4
2
(the predominant component in the mix of 4 and 5) are collated
in Table 8. For EPR hyperfine coupling constants, see also
Table 3.
1
0
.46 g, 3.80 mmol; 4 and 5: 0.16 g, 1.27 mmol) dissolved in
1
0 mL of THF. Green solutions were thus obtained. Drop-wise
[VO(3Br–salSer)THF] (6), [VO(5Br–salSer)THF] (7), [VO
(3Br–salPhe)THF] (8). General procedure: All operations were
addition of bromoalanine methyl ester hydrochloride, BrAlaMe
3: 1.0 g, 3.80 mmol; 4 and 5: 0.34 g, 1.27 mmol) dissolved in
0 mL of methanol, and stirring overnight at room temp. in the
(
2
carried out in O -free solvents. 2 equivalents of sodium acetate
2
trihydrate and one equivalent of the amino acid were dissolved
in 50 mL of methanol, and swiftly treated with one equivalent of
the respective salicylaldehyde dissolved in 30 mL of THF. To
this solution, one equivalent of VOSO ·5H O dissolved in
dark, yielded green suspensions, which were filtered under inert
gas atmosphere. The compounds were recovered by evaporating
4
2
3
0 mL of methanol was added over a time span of 30 min, and
Table 8 IR (cm−1), EPR parameter (|A| in 10−4 cm−1) and MS data for
the mixture stirred at room temp. for 24 hours. The grimy-green
precipitate was filtered off under inert gas atmosphere and
extracted with THF. The THF was then evaporated in vacuo, the
residue washed three times with 20 mL of diethyl ether, and
dried in vacuo. The complexes were recovered in the form of
green, amorphous solids. Specific details are compiled in
Table 9. Table 10 contains selected characteristic spectroscopic
data.
3
and 4
ν(C–
ν(CvN) ν(VvO) Br)
g
g
iso/gx,y
z
,
A
y
, A
iso/Ax,
z
MS m/z
3
1626
987
618
1.972/
93.1/
61.3,
177.2
ESI(+): 283
1
1
.983,
.952
and 285 [M −
MeOH − 2H
+
−
VO]
ESI(−): 417
and 419 [M −
MeOH − H +
−
VO]
Acknowledgements
a
4
1622
988
612
1.973/
92.1/
61.0,
168.6
FAB: 329 [M
+
1
1
.982,
.951
− Br]
Part of the work was funded by the German Research Foun-
dation (DFG project no. RE 431/20). We thank Dr Eugenio Gar-
ribba, University of Sassari (Italy), for support in the
calculations of EPR spectroscopic parameters.
a
ν(CvO) = 1596 cm−1
.
Table 9 Sample weights of the starting products for the preparation of complexes 6, 7 and 8, and product yields
a
b
Complex
Amino acid g (mol)
Br–sal g (mol)
NaAc·3H
2
O g (mol)
VOSO
4
2
·5H O g (mol)
Yield
6
7
8
0.44 (4.19)
0.53 (5.14)
0.68 (4.12)
1.01 (4.11)
1.02 (5.07)
1.01 (4.11)
1.26 (8.28)
1.55 (10.19)
1.26 (8.28)
0.98 (4.13)
1.21 (5.10)
0.98 (4.13)
1.46 g (89%)
1.51 g (84%)
2.04 g (85%)
a
b
(
S)-Serine in the case of 6 and 7, (S)-phenylalanine in the case of 8. 6 and 8: 3-bromo-5-nitrosalicylaldehyde, 7: 5-bromosalicylaldehyde.
Table 10 IR (cm−1), EPR parameters (|A| in 10−4 cm−1) and FAB-MS data for 6, 7 and 8
ν(CvN)
ν(CvO)
ν(VvO)
ν(CBr)
giso/gx,y, g
z
A
iso/Ax,y, A
z
MS m/z
a
a
+
+
+
6
7
8
1619
1616
1615
1587
1578
1582
982
984
981
1089
1066
1092
1.973/1.982, 1.948
1.974/1.981, 1.949
1.974/1.980, 1.949
98.1/62.8, 171.3
94.8/57.9, 168.5
96.8/61.2, 171.3
398 [M − THF + H]
353 [M − THF + H]
458 [M − THF + H]
a
The ν(NvO) appear at 1544 and 1342 for 6, and at 1561 and 1349 cm−1 for 8.
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 5225–5234 | 5233

View Article Online
1
5 A. Schouten and M. Lutz, Acta Crystallogr., Sect. E: Struct. Rep. Online,
References
2009, 65, o3026.
1
(a) D. Wischang, O. Brücher and J. Hartung, Coord. Chem. Rev., 2011,
55, 2204–2217; (b) D. Rehder, Bioinorganic Vanadium Chemistry, John
Wiley & Sons, Chichester, 2008.
16 (a) G. Micera and E. Garribba, Dalton Trans., 2009, 1914–1918;
(b) G. Micera and E. Garribba, J. Comput. Chem., 2011, 32, 2822–
2835.
2
2
3
4
5
6
7
M. Weyand, H.-J. Hecht, M. Kieß, M.-F. Liaud, H. Vilter and
D. Schomburg, J. Mol. Biol., 1999, 293, 595–611.
M. N. Isupov, A. R. Dalby, A. A. Brindley, Y. Izumi, T. Tanabe, G.
N. Murshudov and J. A. Littlechild, J. Mol. Biol., 2000, 299, 1035–1049.
J. Littlechild, E. G. Rodriguez and M. Isupov, J. Inorg. Biochem., 2009,
17 R. Fulwood, H. Schmidt and D. Rehder, J. Chem. Soc., Chem. Commun.,
1995, 1443.
18 N. Tanaka, Z. Hasan and R. Wever, Inorg. Chim. Acta, 2003, 356, 288–
296.
19 J. Costa Pessoa, I. Cavaco, I. Correia, M. T. Duarte, R. D. Gillard, R.
T. Henriques, F. J. Higes, C. Madeira and I. Tomaz, Inorg. Chim. Acta,
1999, 293, 1–11.
20 G. M. Sheldrick, SHELXL-97, Program for Crystal Structure Refinement,
University of Göttingen, Germany.
21 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb,
J. R. Cheeseman, J. A. Montgomery, Jr., T. Vreven, K. N. Kudin, J.
C. Burant, J. M. Millam, S. S. Iyengar, J. Tomasi, V. Barone,
B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G. A. Petersson,
H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa,
M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li,
J. E. Knox, H. P. Hratchian, J. B. Cross, V. Bakken, C. Adamo,
J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin,
R. Cammi, C. Pomelli, J. Ochterski, P. Y. Ayala, K. Morokuma, G.
A. Voth, P. Salvador, J. J. Dannenberg, V. G. Zakrzewski, S. Dapprich, A.
D. Daniels, M. C. Strain, O. Farkas, D. K. Malick, A. D. Rabuck,
K. Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul,
S. Clifford, J. Cioslowski, B. B. Stefanov, G. Liu, A. Liashenko,
P. Piskorz, I. Komaromi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-
Laham, C. Y. Peng, A. Nanayakkara, M. Challacombe, P. M. W. Gill, B.
G. Johnson, W. Chen, M. W. Wong, C. Gonzalez and J. A. Pople, GAUS-
SIAN 03 (Revision C.02), Gaussian, Inc., Wallingford, CT, 2004.
22 (a) A. D. Becke, Phys. Rev. A: At., Mol., Opt. Phys., 1988, 38, 3098–
3100; (b) C. Lee, W. Yang and R. G. Parr, Phys. Rev. B, 1988, 37, 785–
789.
1
03, 617–621.
M. C. Feiters, C. Leblanc, F. C. Küpper, W. Meyer-Klaucke, G. Michel
and P. Potin, J. Am. Chem. Soc., 2005, 127, 15340–15341.
M. C. Feiters, F. C. Küpper and W. Meyer-Klaucke, J. Synchrotron
Radiat., 2005, 12, 85–93.
(a) H. Dau, J. Dittmer, M. Epple, J. Hanss, E. Kiss, D. Rehder,
C. Schulzke and H. Vilter, FEBS Lett., 1999, 457, 237–240;
(b) D. Rehder, C. Schulzke, H. Dau, C. Meinke, J. Hanss and M. Epple,
J. Inorg. Biochem., 2000, 80, 115–121.
D. Rehder and V. Kraehmer, Medimond Int. Proc., EUROBIC 10, 2010,
8
9
7
3–76.
U. Christmann, H. Dau, M. Haumann, E. Kiss, P. Liebisch, D. Rehder,
G. Santoni and C. Schulzke, Dalton Trans., 2004, 2534–2540.
1
1
0 M. Ebel and D. Rehder, Inorg. Chem., 2006, 45, 7083–7090.
1 (a) G. Micera and E. Garribba, Int. J. Quantum Chem., 2011, 111, DOI:
0.1002/qua.23237 online preprint (b) V. A. Nikolakis, J. T. Tsalavoutis,
1
M. Stylianou, E. Evgeniou, T. Jakusch, A. Melman, M. P. Sigalas,
T. Kiss, A. D. Keramidas and T. A. Kabanos, Inorg. Chem., 2008, 47,
1
1698–11710.
2 O. Keller, W. E. Keller, G. V. Look and G. Versin, Org. Synth., 1990, 7,
0–76.
1
7
1
1
3 T. Mukaiyama, Angew. Chem., Int. Ed., 2004, 43, 5590–5614.
4 G. E. Reid, R. J. Simpson and R. A. J. O’Hair, J. Am. Soc. Mass Spec-
trom., 1998, 9, 945–956.
5234 | Dalton Trans., 2012, 41, 5225–5234
This journal is © The Royal Society of Chemistry 2012