Acceptor-substituted ferrocenium salts as strong, single-electron oxidants: Synthesis, electrochemistry, theoretical investigations, and initial synthetic application

Article abstract of DOI:10.1002/chem.201201499

A series of mono- and 1,1'-diheteroatom-substituted ferrocene derivatives as well as acylated ferrocenes was prepared efficiently by a unified strategy that consists of selective mono- and 1,1'-dilithiation reactions and subsequent coupling with carbon, phosphorus, sulfur and halogen electrophiles. Chemical oxidation of the ferrocene derivatives by benzoquinone, 2,3-dichloro-5,6- dicyanobenzoquinone, AgPF₆, or 2,2,6,6-tetramethyl-1-oxopiperidinium hexafluorophosphate provided the corresponding ferrocenium salts. The redox potentials of the synthesized ferrocenes were determined by cyclic voltammetry, and it was observed that all new ferrocenium salts have stronger oxidizing properties than standard ferrocenium hexafluorophosphate. An initial application of selected derivatives in an oxidative bicyclization revealed that they mediate the transformation under considerably milder conditions than ferrocenium hexafluorophosphate. Quantum chemical calculations of the reduction potentials of the substituted ferrocenium ions were carried out by using a standard thermodynamic cycle that involved the gas-phase energetics and solvation energies of the contributing species. A remarkable agreement between theory and experiment was found: the mean average deviation amounted to only 0.030-V and the maximum deviation to 0.1-V. This enabled the analysis of various physical contributions to the computed reduction potentials of these ferrocene derivatives, thereby providing insight into their electronic structure and physicochemical properties. Copyright

Full text of DOI:10.1002/chem.201201499

FULL PAPER
DOI: 10.1002/chem.201201499
Acceptor-Substituted Ferrocenium Salts as Strong, Single-Electron Oxidants:
Synthesis, Electrochemistry, Theoretical Investigations, and Initial Synthetic
Application
Dushant A. Khobragade,[a] Shraddha G. Mahamulkar,[a] Lubomꢀr Pospꢀsˇil,[a, b]
Ivana Cꢀsarˇovꢁ,[c] Lubomꢀr Rulꢀsˇek,[a] and Ullrich Jahn*[a]
Dedicated to the memory of Prof. Dr. Werner Schroth
Abstract: A series of mono- and 1,1’-
diheteroatom-substituted ferrocene de-
rivatives as well as acylated ferrocenes
was prepared efficiently by a unified
strategy that consists of selective
mono- and 1,1’-dilithiation reactions
and subsequent coupling with carbon,
phosphorus, sulfur and halogen electro-
philes. Chemical oxidation of the ferro-
cene derivatives by benzoquinone, 2,3-
dichloro-5,6-dicyanobenzoquinone,
AgPF6, or 2,2,6,6-tetramethyl-1-oxopi-
peridinium hexafluorophosphate pro-
vided the corresponding ferrocenium
salts. The redox potentials of the syn-
thesized ferrocenes were determined
by cyclic voltammetry, and it was ob-
served that all new ferrocenium salts
have stronger oxidizing properties than
standard ferrocenium hexafluorophos-
phate. An initial application of selected
derivatives in an oxidative bicyclization
revealed that they mediate the trans-
formation under considerably milder
conditions than ferrocenium hexafluor-
ophosphate. Quantum chemical calcu-
lations of the reduction potentials of
the substituted ferrocenium ions were
carried out by using a standard thermo-
dynamic cycle that involved the gas-
phase energetics and solvation energies
of the contributing species. A remark-
AHCTUNGTRENNaGUN ble agreement between theory and ex-
periment was found: the mean average
deviation amounted to only 0.030 V
and the maximum deviation to 0.1 V.
This enabled the analysis of various
physical contributions to the computed
reduction potentials of these ferrocene
derivatives, thereby providing insight
into their electronic structure and phys-
icochemical properties.
Keywords: cyclic voltammetry · lac-
tones · metallocenes · oxidation ·
redox chemistry
Introduction
switch the redox state of complexes.[1,2] In contrast, applica-
tions of ferrocenium salts in organic chemistry are rare and
mostly limited to easily accessible ferrocenium hexafluoro-
phosphate A (Scheme 1).[3,4] A promising, but not well-ex-
plored field are oxidative tandem reactions, in which starting
molecules M are transformed to products P through multi-
ple intermediates I of different redox state. High molecular
complexity can be generated from very simple precursors in
a single operation.[3d,f,h] Few applications of A in the total
synthesis of natural products exist.[3a,b,5] Salt A is well suited
to oxidize anionic intermediates to radicals, but has only
limited capacity to oxidize radicals to carbocations.
Ferrocene derivatives are extensively used in many areas of
chemistry, such as catalysis, material science, and bioorgano-
metallic chemistry.[1] One of the reasons for their attractive-
ness is the fact that their redox state can be easily varied,
which allows for a number of their applications in analytical
and supramolecular chemistry.[1] Ferrocenium salts have
been used extensively as single-electron transfer (SET) re-
agents in organometallic and coordination chemistry to
ˇ
[a] Dr. D. A. Khobragade, S. G. Mahamulkar, Dr. L. Pospꢀsil,
ˇ
Dr. L. Rulꢀsek, Dr. U. Jahn
In comparison to other metallic and non-metallic SET oxi-
dants, ferrocenium salts are highly attractive, since their
redox potential can be tuned easily by the introduction of
substituents. Electron-donating substituents decrease their
oxidation power, whereas acceptor substituents increase it,
thus a wide potential area that ranges from very weak oxi-
dants, such as the decamethylferrocenium ion, to very
strong oxidants, such as the 1,1’-diacetylferrocenium ion, is
accessible in principle.[1,2] However, a major limitation that
prevents the design of oxidative SET-mediated (tandem)
transformations and other applications is the restricted
availability of substituted and fine-tuned (with respect to
their oxidation potential) ferrocenium salts. Generally,
Institute of Organic Chemistry and Biochemistry
Academy of Sciences of the Czech Republic
Flemingovo nꢁmesti 2, 16610 Prague 6 (Czech Republic)
E-mail: jahn@uochb.cas.cz
ˇ
´
[b] Dr. L. Pospꢀsil
J. Heyrovsky Institute of Physical Chemistry
Academy of Sciences of the Czech Republic
Dolejskova 3, 18223 Prague 8 (Czech Republic)
ˇ
ˇ
[c] Dr. I. Cꢀsarovꢁ
Department of Inorganic Chemistry
Charles University Prague, Albertov 6
12843 Prague 2 (Czech Republic)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201201499.
Chem. Eur. J. 2012, 00, 0 – 0
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
1
&
ÞÞ
These are not the final page numbers!

pared ferrocenes and ferrocenium salts, a series of quantum
chemical calculations was performed. Recent developments
in computational chemistry including modern implicit solva-
tion models, for example, the conductor-like screening
model: real solvent (COSMO-RS),[6] and accumulated expe-
rience[7] nowadays enable us to make direct calculations of
reduction potentials often with an accuracy better than
0.1 V (ꢀ10 kJmolꢁ1 in the overall DGox/red value). Quantum
chemical calculations thus represent an indispensable and
complementary tool that not only reproduces the experi-
mental data and makes qualified predictions for new ferro-
cenes, but also provides a genuine link between the elec-
tronic structure of the target molecules and their electro-
chemical properties.
Results and Discussion
Synthesis of mono- and 1,1’-disubstituted ferrocene deriva-
tives: Two unified, generally applicable approaches based on
ferrocene 1 were developed to cover the desired substituent
spectrum in mono- and 1,1’-disubstituted ferrocenes 2 and 3,
respectively. All monosubstituted ferrocene derivatives 2a–h
were synthesized in high yield and selectivity by adapting a
procedure developed originally by Sanders and Mueller-
Westerhoff,[8a] which was later used by Leclercq et al. for the
synthesis of diethyl ferrocenephosphonate (2d),[8b] whereas
a modified approach based on the one that was reported by
Alley and Henderson for 3d[9] was applied for the synthesis
of all 1,1’-disubstituted ferrocenes (Table 1).
Scheme 1. Ferrocenium salts as oxidants: status and perspective.
mono- and 1,1’-disubstituted ferrocene derivatives have
been prepared by reactions of lithiated ferrocenes with elec-
trophiles, Friedel–Crafts reactions, or by means of ferroce-
nylmercury derivatives.[1] However, most of these methods
display only a limited substrate spectrum.
In this study, we describe the selective synthesis of a
series of mono- and 1,1’-disubstituted ferrocene derivatives
by a unified approach and reliable and convenient access to
their ferrocenium salts B and C. The study focuses primarily
on heteroatom-substituted ferrocenes, since they can be con-
sidered the chemically most
Thus, the reaction of 1 using the Schlosser–Lochmann
base (tBuLi/KOtBu) in THF at ꢁ788C followed by quench-
inert, whereas the very few
Table 1. Synthesis of mono and 1,1’-disubstituted ferrocenes 2 and 3 from ferrocene 1.[a]
known acylated ferrocenes,
such as acetylferrocenes, are
often not compatible with orga-
nometallic reaction conditions
due to their acidity and electro-
philicity. However, one mode of
access to acylated ferrocenium
salts that are less prone to de-
protonation or nucleophilic
attack is also presented. The
applicability of selected ferroce-
nium salts B and C in an oxida-
tive tandem cyclization, which
does not proceed well with A
under the very mild conditions,
is reported. The reduced ferro-
cenes can be recovered, thus
guaranteeing repeated use.
Entry
Condition
Electrophile
t [h]
E
2 [%][b]
3 [%][b]
ꢁ
1
2
3
4
5
6
7
8
A
B
A
B
A
B
A
B
A
B[c]
A
B
A
B
ClCO2iPr
ClCO2iPr
ClCOPh
ClCOPh
ClPPh2
2
0.75
4
2.5
3
3
4
3
2
68
2
2.5
1
5
CO2iPr
CO2iPr
COPh
COPh
a 80
a 18
b 86
b 16
c 75
c –
d 89
d 22
e 76
e –
f 72
f 15
g 77[e]
g –
h 71
h –
a 3
a 58
b –
b 65
c –
c 73
d –
ꢁ
ꢁ
ꢁ
ꢁ
PPh2
ꢁ
ClPPh2
PPh2
PO(OEt)2
(OEt)2
ꢁ
ClPO
ClPO
U
R
ꢁ
PO
R
d 60
e –
ꢁ
9
PhSSPh
PhSSPh
TsCl
SPh
SPh
ꢁ
10
11
12
13
14
15
16
e 70
f 7
ꢁ
Cl
TsCl
Cl
f 64[d]
g –
ꢁ
ꢁ
BrCH2CH2Br
CBrCl2CBrCl2
I2
Br
ꢁ
Br
g 89
h –
h 60
ꢁ
A
B
1
2.5
I
The oxidation power of salts
B and C is determined by cyclic
voltammetry. To gain deeper in-
sight into the observed physico-
ꢁ
I2
I
[a] Condition A: 2.0 equiv tBuLi, 0.12 equiv KOtBu, THF, ꢁ788C to RT, then electrophile. Condition B:
2.2 equiv nBuLi, 2.4 equiv TMEDA, hexane, ꢁ788C to RT, then electrophile, THF, ꢁ788C to RT. [b] Isolated.
[c] 2.0 equiv of each nBuLi and TMEDA. [d] Compounds 2 f and 3 f not separable; yield based on 1H NMR
chemical properties of the pre- spectroscopy. [e] Not separable from 1 by chromatography, separation by sublimation of 1.
&
2
&
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Single-Electron Oxidants
FULL PAPER
ing with isopropyl chloroformate furnished isopropyl ferro-
cenecarboxylate (2a)[10] in 80% yield accompanied by 3%
of diisopropyl ferrocene-1,1’-dicarboxylate (3a; Table 1,
entry 1). Benzoylferrocene (2b),[11] (diphenylphosphanyl)fer-
rocene (2c),[12] diethyl ferrocenephosphonate (2d),[8] and
(phenylthio)ferrocene (2e)[13] were also obtained in good
yields (Table 1, entries 3, 5, 7, 9). The reaction of ferrocenyl-
lithium with tosyl chloride did not afford the corresponding
sulfonylferrocene, but provided a good yield of chloroferro-
cene (2 f; Table 1, entry 11).[14,15] N-Chlorosuccinimide gave
a poor yield in contrast. Bromoferrocene (2g)[15] and iodo-
ferrocene (2h)[15,16] were obtained by bromination using di-
bromoethane and iodination with iodine, respec-
Scheme 3. Oxidation of phosphanyl and sulfenyl ferrocenes.
tively (Table 1, entries 13, 15).
Table 2. Oxidation of (diphenylphosphanyl)ferrocene (2c), 1,1’-bis(diphenylphospha-
nyl)ferrocene (3c), (phenylthio)ferrocene (2e), and 1,1’-bis(phenylthio)ferrocene (3e).
By contrast, lithiation of 1 with nBuLi/N,N,N’,N’-
tetramethylethylenediamine (TMEDA) at ꢁ78 to
ꢁ408C furnished insoluble 1,1’-dilithioferrocene,
which gave 3a by reaction with an excess amount
of isopropyl chloroformate in 58% yield along with
18% of 2a (Table 1, entry 2). 1,1’-Dibenzoylferro-
cene (3b),[11b] 1,1’-bis(diphenylphosphanyl)ferro-
cene (3c),[17] tetraethyl ferrocene-1,1’-diphospho-
nate (3d),[9] and 1,1’-bis(phenylthio)ferrocene
(3e)[13e,18] were similarly synthesized in moderate to
Entry
Substrate
Oxidant [equiv]
Product [%]
1
2
3
4
5
6
2c
3c
2e
2e
3e
3e
H2O2 (excess amount)
H2O2 (excess amount)
mCPBA (1.2)
2k 82
3k 82
2l 88
2m 84
3l 85
mCPBA (3.0)
mCPBA (2.4)
mCPBA (5.0)
3m 82
good yields (Table 1, entries 4, 6, 8, 10). The 1,1’-dihaloferro-
cenes 3 f–h were synthesized by quenching 1,1’-FcLi2 with
the corresponding electrophilic halogen sources as men-
tioned before in good yield (Table 1, entries 12, 14, 16). 1,1’-
Dichloroferrocene (3 f) was purified for analytical purposes
by repeated crystallization from methanol/water.
trated H2SO4 and KPF6 in 84% yield as a dark blue pow-
der.[3g] These conditions are not applicable to other ferro-
cene derivatives, since they often lead to decomposition.
Therefore, different oxidants were employed (Scheme 4,
Table 3).
Benzylferrocene (2i) and 1,1’-dibenzylferrocene (3i) were
obtained by reduction of benzoylferrocene (2b) and 1,1’-di-
benzoylferrocene (3b), respectively, using the borane–di-
methyl sulfide complex (Scheme 2).[11a]
Scheme 2. Reduction of benzoylferrocenes 2b, 3b to benzylferrocenes 2i,
3i, respectively.
Scheme 4. Oxidation of ferrocenes 2 and 3 to ferrocenium salts 4 and 5.
Oxidation of ferrocenylphosphanes and (phenylthio)ferro-
cenes: Mono- and 1,1’-disubstituted diphenylphosphanyl fer-
rocenes 2c and 3c were oxidized to the corresponding ferro-
cenyl phosphane oxides 2k and 3k by using 30% aqueous
hydrogen peroxide in good yield (Scheme 3, Table 2, en-
tries 1 and 2). The ferrocenyl sulfoxides 2l and 3l as well as
the corresponding ferrocenyl sulfones 2m and 3m were ob-
tained by oxidation of the phenylthioferrocenes 2e or 3e
with meta-chloroperoxybenzoic acid (mCPBA) in very good
yields (Table 2, entries 3–6).[19]
The most generally applicable oxidants for these more
electron-deficient ferrocenes proved to be benzoquinones
6a,b (Table 3, entries 1–5, 8–16). The stronger oxidant 2,3-
dichloro-5,6-dicyanobenzoquinone (DDQ; 6b) was more
useful to oxidize 1,1’-diacceptor-substituted ferrocenes 3
(Table 3, entries 2 and 4). The strongest oxidizing ferroceni-
um salts 5 (R=SOPh, SO2Ph, POPh2, POACTHUNGRTNEUNG(OEt)2) were also
generated by using 6b as indicated by the formation of deep
green or blue solutions in dichloromethane. They were, how-
ever, very sensitive to reduction to the starting ferrocenes 3
even by weak donor solvents such as diethyl ether. This pre-
cluded their isolation and characterization so far. Oxidation
with AgPF6 in dichloromethane was also efficient (Table 3,
Synthesis of ferrocenium salts: Ferrocenium hexafluoro-
phosphate is commonly prepared from 1 by using concen-
Chem. Eur. J. 2012, 00, 0 – 0
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
3
&
ÞÞ
These are not the final page numbers!

U. Jahn et al.
Yield [%]
Table 3. Synthesis of substituted ferrocenium hexafluorophosphates 4 and 5.
Entry
2/3
R
Oxidant
Yield [%]
Entry
2/3
R
Oxidant
1
2
3
4
5
6
7
8
2a
3a
2b
3b
2k
2d
2d
2d
CO2iPr
CO2iPr
COPh
COPh
POPh2
6a
6b
6a
6b
6a
4a 90
5a 81
4b 82
5b 76
4k 76
4d 66
4d 64
4d 97
9
2e
2l
2m
2 f
2g
3g
2h
3h
SPh
SOPh
SO2Ph
Cl
Br
Br
6a
6a
6b
6a
6a
6a
6a
6a
4e 91
4l 70
10
11
12
13
14
15
16
4m 72
4 f 70
4g 74
5g 72
4h 80
5h 81
PO
PO
PO
G
7
E
AgPF6
6a
I
I
G
entry 7). It can also be used for the oxidation of other ferro-
cenes (not shown), but its price likely prohibits large-scale
applications. Easily accessible oxoiminium hexafluorophos-
phate 7[20] also proved to be an efficient oxidant for the fer-
rocenes (Table 3, entry 6). Most of the synthesized ferroceni-
um salts 4 and 5 are blue, blue-green, or green powders,
which are stable in the solid state. They are stable and solu-
ble in acetonitrile and dichloromethane, but slowly reduced
back to 2 or 3, however, by alcoholic or ethereal solvents,
DMF or DMSO.
Structural characterization of ferrocenes: The structures of
most ferrocenes was confirmed by comparison of their
NMR spectroscopic data.[21] Those of ferrocenes 2a, 3a, 2m,
and 3m were established by X-ray crystallography (Figures 1
Figure 2. X-ray crystallographic view of ferrocenyl phenyl sulfone (2m).
The displacement ellipsoids are drawn at a 50% probability level.
Figure 1. X-ray crystal structure of isopropyl ferrocenecarboxylate (2a).
The displacement ellipsoids are drawn at a 50% probability level.
The equilibrium geometries of 2a, 2m, and 3m, obtained
by quantum chemical methods (all of the calculated equili-
brium structures are deposited in the Supporting Informa-
tion), are in excellent agreement with the solid-state struc-
tures. The comparison in 3a revealed an interesting phe-
nomenon: the conformer with the antiperiplanar arrange-
ment of the ester groups was energetically marginally less
stable in the gas phase (0.5–2.5 kcalmolꢁ1 depending on the
basis set used; the lower value was obtained using the larger
def2-TZVP basis set). However, the D(EZPVEꢁRTlnQ) term
(“gas-phase thermochemistry”; ZPVE=zero-point vibra-
tional energy) and the difference in solvation energies re-
versed this order in the overall free-energy values, that is,
the antiperiplanar conformation was more stable in solution
by 1.0 and 0.5 kcalmolꢁ1 for the reduced and oxidized form,
to 4).[22] The cyclopentadiene rings in 2a and 2m are in a
synclinal arrangement (Figure 1, Figure 2). The ester group
of 2a is almost coplanar to the cyclopentadiene ring and
thus in full conjugation.
The difference between structures of 3a and 3m is more
pronounced. A staggered arrangement of the cyclopenta-
dienyl rings was found in 3a. The ester units in 3a are ar-
ranged in a gauche conformation, with both being conjugat-
ed to the cyclopentadiene rings (Figure 3). By contrast, the
rings are almost synclinal in 3m, but an almost perfect anti-
periplanar arrangement of the sulfonyl groups is favored in
the solid state (Figure 4).
&
4
&
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Single-Electron Oxidants
FULL PAPER
respectively. One has to admit, though, that these values are
probably within the error bar of the methods. The calculated
energy difference between the two forms translates to less
than 0.02 V in the reduction potentials and should not signif-
icantly hamper the calculated values (vide infra). It does,
however, suggest that the (free) energy differences of the
major conformers of the studied ferrocenes are quite small.
An initial application of selected ferrocenium salts in a
double oxidative cyclization: Oxidative tandem cyclizations
of less reactive 2-(4-pentenyl)malonates, such as the 5,5-di-
phenyl derivative 8, proceed by using lithium bis(trimethyl-
silyl)amide (LiHMDS) or lithium diisopropylamide (LDA)
as a base and ferrocenium hexafluorophosphate A as the
prototypical oxidant in good yield only at temperatures of
08C or higher (Scheme 5, Table 4, entry 1).[3g] Product 10
Figure 3. X-ray crystallographic representation of ferrocene 3a. The dis-
placement ellipsoids are drawn at a 50% probability level. The molecule
is positioned on a twofold axis of space group Pbcn; symmetry code (i):
2ꢁx, y, 0.5ꢁz.
Scheme 5. Oxidative cyclization of diethyl 2-(5,5-diphenylpent-4-en-1-yl)-
malonate (8). DME=1,2-dimethoxyethane.
forms by transfer of an ethyl group from the bicyclic carbo-
cationic intermediate, which leads to 9 to the enolate of sub-
strate 8;[3g] this is an indication of the slow oxidation of the
latter. The use of BuLi as the base at ꢁ788C leads to an
even more complex reaction and gives products 9, 11, and
12 as a partly separable mixture (Table 4, entry 2). This
result can also be traced to a slow SET oxidation of the eno-
late of 8.
To improve the yield and selectivity, the oxidative cycliza-
tion of 8 was performed by using selected ferrocenium salts
4m, 4a, 5a, and 5b at ꢁ788C. Bicyclic lactone 9 was ob-
tained in 48–77% yield (Table 4, entries 3–6). No product 10
that resulted from the transfer of an ethyl group to the eno-
late of 8 was detected in any of the cyclizations, thus indicat-
Figure 4. Structure of 3m in the crystal. The displacement ellipsoids are
drawn at a 50% probability level.
Table 4. Results of the oxidative cyclizations of 8.
Entry
Base
Oxidant
9 [%]
10 [%]
11 [%]
12 [%]
8 [%][a]
1, 2, or 3 [%][b]
1[c]
2
3
4
5
LDA
nBuLi
nBuLi
LiHMDS
nBuLi
nBuLi
A
A
4m
4a
5a
5b
62
28
61
20
–
–
–
–
2
24
–
–
–
–
13
15
–
12
–
2
22
26
3
31
20
66[d]
58[d]
94
48
93
53[e]
77
64
6
–
–
72[f]
[a] Recovered 8. [b] Recovered 2 or 3. [c] Conditions exactly as in Ref. [3g] at 08C. [d] The recovery of ferrocene was lower on account of incomplete ox-
1
idation. The remainder of the mass balance is unchanged A. [e] Product 9 and reduced ferrocene 3a are not separable; yield based on the H NMR spec-
trum. [f] Benzoylcyclopentadiene dimers isolated: 7%.
Chem. Eur. J. 2012, 00, 0 – 0
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
5
&
ÞÞ
These are not the final page numbers!

U. Jahn et al.
ing two fast SET oxidation steps. However, in cyclizations in
which BuLi was applied as the base, small amounts of acy-
clic products 12 that resulted from the addition of BuLi to 8
were formed (Table 4, entries 2, 3, 5). The yield and mass
balance of the reaction was reduced to 48% when LiHMDS
was used as the base, but lactone 9 remained the major
product (Table 4, entry 4). The electrophilic carbonyl groups
of the ferrocenium derivatives 4a, 5a, and 5b were not at-
tacked by the nucleophilic enolate of 8, thus demonstrating
the high facility of SET oxidation of both the enolate and
the cyclized radical, although the recovered amount of the
reduced ferrocene 5a was somewhat lower (Table 4,
entry 5). In the reaction that used 5b, 7% of benzoylcyclo-
pentadiene dimers were isolated that apparently resulted
from some competing decomplexation under the reaction
conditions (Table 4, entry 6).
nate substituents led to a significant increase in the redox
potential over the diphenylphosphanyl substituent (Table 5,
entries 6 and 7 versus 5), although the effect of the phospha-
noyl group on the redox potential is stronger than that of
the phosphonate unit (Table 5, entry 6 versus 7). For sulfur
substituents, the oxidation potential increases considerably
as the degree of oxygenation increased (Table 5, entries 8–
10). On the other hand, the redox potentials of the halogen-
substituted ferrocenes are quite similar to each other
(Table 5, entries 11–13). Incorporation of a second benzyl
substituent has almost no effect on the redox potential
(Table 5, entries 2 versus 14). The redox potential increases,
however, with further acceptor substitution, though the
effect of the second group is less pronounced (Table 5, en-
tries 14–16). Similar trends to those for the monosubstituted
derivatives 2c, 2k, 2d, or 2e, 2l, 2m were found for the cor-
responding diphosphorus and disulfur-substituted ferrocenes
3c, 3k, 3d, or 3e, 3l, 3m (Table 5, entries 17–19 and 20–22).
The bis(phenylsulfonyl)ferrocenium ion is the strongest oxi-
dant of the whole series (Table 5, entry 22). The introduc-
tion of a second halogen substituent increases the oxidation
potential by only 0.1–0.17 V (Table 5, entries 23–25).
Electrochemistry: The redox behavior of ferrocene deriva-
tives 2a–m and 3a–m was studied by cyclic voltammetry in
0.1m Bu4NPF6 solution in acetonitrile (Table 5). Ferrocene 1
was studied for comparison (Table 5, entry 1). Both cyclic
voltammetry and impedance spectroscopy confirmed a re-
versible one-electron transfer reaction. For a given substitu-
ent atom, the anticipated direct relationship between the
electron-withdrawing character of the substituents and its
oxidation potential was observed. Therefore, upon going
from benzylferrocene to isopropyl ferrocenecarboxylate, the
oxidation increased in power from E0 =0.463 to 0.704 V
(Table 5, entries 2–4). The phosphane oxide and phospho-
Theoretical investigation of acceptor-substituted 1- and 1,1’-
disubstituted ferrocenes and ferrocenium salts: Quantum
chemical calculations of the reduction potentials by using
the DFT method (Perdew–Burke–Ernzerhof (PBE) func-
tional) and one of the presumably most accurate implicit
solvation models, COSMO-RS, yielded values of the reduc-
Table 5. Experimental and calculated reduction potentials of ferrocenes 2 and 3 with respect to the Ag/AgCl 1m LiCl electrode.[a]
[d]
Entry
Compound
E0exp [V]
E0calcd [V]
IE [eV][b]
D(EZPVEꢁRTlnQ)[c] [eV]
DDGsolv [eV]
1
2
3
4
5
6
7
8
1
0.480
0.463
0.693
0.704
0.560
0.713
0.695
0.610
0.783
0.847
0.624
0.630
0.635
0.458
0.917
0.914
0.653
0.946
0.903
0.681
0.983
1.190
0.771
0.797
0.741
0.525
–
6.75
–
ꢁ0.020
ꢁ1.709
–
2i (CH2Ph)
2b (COPh)
2a (CO2iPr)
2c (PPh2)
2k (POPh2)
2d (PO3Et2)
2e (SPh)
2l (SOPh)
2m (SO2Ph)
2 f (Cl)
–
0.712
0.702
0.502
0.700
0.708
0.567
0.735
0.823
0.633
0.668
0.640
0.501
0.952
0.933
0.554
1.000
0.917
0.671
0.994
1.165
0.765
0.828
0.768
6.69
6.80
6.26
6.57
6.65
6.40
6.75
6.84
6.82
6.83
6.77
6.38
6.77
6.88
6.07
6.50
6.53
6.28
6.79
6.94
6.92
6.93
6.85
ꢁ0.028
ꢁ0.025
ꢁ0.003
ꢁ0.049
ꢁ0.004
0.003
ꢁ0.016
ꢁ0.008
ꢁ0.008
ꢁ0.001
ꢁ0.003
0.012
ꢁ1.454
ꢁ1.576
ꢁ1.259
ꢁ1.318
ꢁ1.436
ꢁ1.340
ꢁ1.495
ꢁ1.507
ꢁ1.679
ꢁ1.657
ꢁ1.626
ꢁ1.387
ꢁ1.274
ꢁ1.399
ꢁ1.036
ꢁ1.001
ꢁ1.095
ꢁ1.157
ꢁ1.290
ꢁ1.266
ꢁ1.661
ꢁ1.620
ꢁ1.574
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
2g (Br)
2h (I)
3i (CH2Ph)2
3b (COPh)2
3a (CO2iPr)2
3c (PPh2)2
3k (POPh2)2
3d (PO3Et2)2
3e (SPh)2
3l (SOPh)2
3m (SO2Ph)2
3 f (Cl)2
ꢁ0.038
ꢁ0.024
0.016
0.001
ꢁ0.017
0.050
ꢁ0.005
ꢁ0.007
0.003
0.016
ꢁ0.009
3g (Br)2
3h (I)2
[a] The electrode potential is 0.21 V versus standard hydrogen electrode.[23] [b] In vacuo ionization energy: IE=E
N
ACHTUNGERTN(NUNG reduced), calculated at
C
CHTUNGTRENNUNG
&
6
&
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Single-Electron Oxidants
FULL PAPER
tion potentials that are in excellent agreement with the ex-
perimental results (Table 5). The mean unsigned deviation
(MUD) is approximately 0.03 V, which is well below the an-
ticipated error bar of theoretical methods (ꢀ0.1 V). It can
be concluded that the agreement is almost quantitative, thus
suggesting a high predictive potential of theoretical calcula-
tions for this series of compounds. On the contrary, the cor-
relation between the calculated ionization energies (IE) and
the experimental reduction potentials is only qualitative.
This qualitative agreement, though, is in line with the ex-
pected change of oxidative power induced by the ferrocene
substituents as discussed above. Essentially, the reduction
potential is dependent on two terms, IE and DDGsolv, where-
as the effects of zero-point vibrational energy and entropic
contributions are only marginal (<0.05 V). The importance
of both IE and DDGsolv can be demonstrated for compounds
2 f and 3k with IE values of 6.82 and 6.50 eV, respectively.
The difference in the solvation energies of their oxidized
and reduced species corresponds, however, to 0.678 eV and
reverses the order of their reduction potentials. It is there-
fore a delicate balance between the two quantities that de-
termines the oxidative power of the ferrocenes. The obser-
vation of a dramatic influence of solvation therefore pre-
cludes any semiquantitative analysis on the basis of molecu-
lar orbital pictures (such as the stabilization of HOMO/
LUMO orbitals) to be generally applicable to a series of
chemically distinct species. On the other hand, within a
given series of similar substituents, the trends in their reduc-
tion potentials are already included in their ionization ener-
gies, which is in line with the qualitative discussion in the
previous section. In other words, the contribution that arises
from the difference in solvation energies of the reduced and
oxidized species remains approximately constant for similar
substituents.
Figure 5. Electron spin density in bis(diphenylphosphanyl)ferrocenium
ion (5c) as obtained by density functional theory calculations using a
Mulliken population analysis.
The analysis of the resulting wave functions of the oxi-
dized complexes revealed—not unexpectedly—that the un-
paired electron resides mostly on the iron atom. However,
variations can be observed for ferrocenium ions that possess
lone pairs at the heteroatom substituents. The bis(phospha-
nyl)ferrocenium ion 5c carries significant spin density at one
of the phosphorus atoms according to a Mulliken population
analysis (Figure 5), whereas it is negligible in the corre-
sponding bis(phenylthio)ferrocenium ion 5e (Figure 6).
Figure 6. Electron spin density in bis(phenylthio)ferrocenium ion (5e) as
obtained by the density functional theory calculations using a Mulliken
population analysis.
medium potentials in the series. Selected mono- and 1,1’-ac-
ceptor-substituted ferrocenium salts were advantageously
applied to an oxidative tandem cyclization under very mild
conditions, which ferrocenium hexafluorophosphate A is not
able to promote well.
It is demonstrated that the synthetic effort can be assisted
by computational methods. The electrochemical properties
of ferrocenes were reliably and almost quantitatively repro-
duced, and computations can be therefore easily used in the
predictive design of new ferrocenium salts with fine-tuned
oxidative power. The results clearly show the delicate bal-
ance between the solvation energies of the oxidized and re-
duced species and their ionization energies. Therefore, any
Conclusion
A series of mono- and 1,1’-disubstituted ferrocene deriva-
tives and their ferrocenium salts was prepared. The new fer-
rocenium salts 4 and 5 are stronger oxidants than parent fer-
rocenium hexafluorophosphate. The redox potentials of the
prepared acceptor-substituted ferrocene derivatives were de-
termined by using cyclic voltammetry. 1,1’-Disubstituted fer-
rocenes with electron-withdrawing groups (SO2Ph, SOPh,
POACHTUNGTRENNUNG(OEt)2, POPh2, COPh, and CO2iPr) have higher oxida-
tion potentials than their monosubstituted counterparts.
Halogen, thioether, and phosphane derivatives have
Chem. Eur. J. 2012, 00, 0 – 0
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
7
&
ÞÞ
These are not the final page numbers!

U. Jahn et al.
solution at ꢁ788C. The solution was slowly warmed to room temperature
until the reaction was complete as indicated by TLC. The reaction was
quenched after the given time by the addition of water (Table 1). The re-
action mixture was diluted with an organic solvent (Table 1, EtOAc for
entry 2, and CH2Cl2 for entries 4, 6, 8, 10, 12, 14, 16, 18). The layers were
separated and the aqueous was extracted thoroughly. The combined or-
ganic fractions were dried over Na2SO4 and evaporated to give the crude
product 3, which was purified by flash column chromatography. Com-
pounds 3b,[11] 3c,[17] 3d,[9] 3e,[13e,18] 3g,[14] and 3h[24] are known and their
analytical data correspond to those in the cited references.
quantitative predictions for ferrocenes with different types
of substituents based solely on in vacuo computational data
will lead to erroneous results. The theoretical treatment sig-
nificantly extends the understanding of the electrochemistry
of substituted ferrocenes. In addition to that, the quantum
chemical calculations provide insights into the electronic
structure of ferrocenium salts, such as their spin densities
and the contribution of ligand-centered radical cations. With
tunable oxidation potentials, these ferrocenium salts can be
expected to find wide applications in chemoselective oxida-
tive reactions in organic chemistry. The investigation of the
reactivity of these ferrocenium salts as SET oxidants in or-
ganic chemistry is underway in these laboratories. Moreover,
they will be certainly useful in organometallic and inorganic
chemistry. Applications in material chemistry as dopants
and as units in electron-transfer active materials can also be
foreseen.
Diisopropyl 1,1’-ferrocenedicarboxylate (3a): Prepared with ClCO2iPr
(21.49 mL, 21.50 mmol, 1.0m solution in toluene). Flash chromatography
hexane gradient to hexane/EtOAc: 19:1. Yield: 1.11 g (58%) as a red-
dish-orange solid. M.p. 518C; Rf (hexane/EtOAc 9:1)=0.48; 1H NMR
(400 MHz): d=5.16 (sept, J=6.3 Hz, 2H; CO2CH
2.0 Hz, 4H; Ha), 4.37 (t, J=1.9 Hz, 4H; Hb), 1.34 ppm (d, J=6.3 Hz,
12H; CO2CH (CH3)2),
(CH3)2); 13C NMR (100 MHz): d=170.1 (s; CO2CH
73.6 (s; ipsoCCp), 73.0 (d; Cb), 71.5 (d; Ca), 67.8 (d; CO2CH(CH3)2),
22.2 ppm (q; CO2CH(CH3)2); IR: n˜ =2978, 1706, 1459, 1373, 1273, 1144,
ACHTUGNTRENN(UNG CH3)2), 4.79 (t, J=
A
ACHTUNGTRENNUNG
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
1104, 1027, 932, 830, 775 cmꢁ1; UV/Vis (CH3CN): lmax (loge)=451 (2.45),
350 (2.53), 300 (sh) (3.10), 258 (3.55), 225 nm (sh) (3.39); MS (EI): m/z
(%): 358/356 (40/1) [M+], 274/272 (100/2) [M+ꢁ2CH2=CH2CH3]; HRMS
(EI): m/z calcd for C18H2256FeO4: 358.0867; found: 358.0855; elemental
analysis calcd (%) for C18H22FeO4 (358.21): C 60.35, H 6.19, Fe 15.59;
found: C 60.51, H 6.35, Fe 15.50.
Experimental Section
(Diphenylphosphanoyl)ferrocenes 2k, 3k (general procedure): A solu-
tion of 30% aqueous H2O2 (5 mL for 2c, or 10 mL for 3c) was added to
a solution of 2c (1.0 g, 2.70 mmol) or 3c (1.0 g, 1.80 mmol) in CH2Cl2
(30 mL). The reaction mixture was stirred at room temperature for 2 h.
The layers were separated and the aqueous was extracted with CH2Cl2,
dried over Na2SO4, and evaporated. The crude products 2k or 3k were
purified by flash column chromatography. Compound 2k is known and
its analytical data correspond to those reported.[25]
For the general experimental methods, see the Supporting Information.
Monosubstituted ferrocenes
2 (general procedure): KOtBu (72 mg,
0.65 mmol) was added to a solution of ferrocene (1.0 g, 5.38 mmol) in
THF (10 mL) at room temperature. The solution was cooled to ꢁ788C
and tBuLi (6.32 mL, 10.76 mmol, 1.7m in pentane) was added dropwise.
The reaction mixture was stirred at ꢁ788C for 1 h. The electrophile (see
at the individual compounds and in the Supporting Information) was
added to the orange solution at ꢁ788C. The solution was slowly warmed
to room temperature until the reaction was complete as indicated by
TLC. The reaction was quenched after the given time by the addition of
water (Table 1). The reaction mixture was diluted by an organic solvent
(Table 1, CH2Cl2 for entries 1, 3, 5, 7, 9, 11, 15, and diethyl ether for
entry 13 ). The layers were separated and the aqueous layer was extract-
ed three times. The combined organic fractions were dried over Na2SO4
and evaporated. The crude product 2 was purified by flash column chro-
matography. Compounds 2b,[11] 2c,[12] and 2e[13] are known and their ana-
lytical data are in full agreement with those reported in the cited refer-
Sulfinyl and sulfonyl ferrocenes 2l, 3l, and 2m, 3m (general procedure):
Ferrocene 2e (1.0 g, 3.40 mmol) or 3e (1.0 g, 2.48 mmol) was dissolved in
dry CH2Cl2 (20 mL). After cooling the reaction mixture to 08C, mCPBA
(amounts provided for the individual compounds) in CH2Cl2 (10 mL) was
added at 08C. The reaction mixture was stirred at room temperature for
1 h and quenched by addition of water. The layers were separated and
the aqueous was extracted with CH2Cl2. The combined organic fractions
were dried over Na2SO4 and evaporated to give crude products 2l, 3l,
2m, or 3m, which were purified by flash column chromatography. Com-
pounds 2l[13c,d] and 2m[19,21] are known and their analytical data corre-
spond to those found in the literature.
ACHTUNGTRENNUNG
ences. Compounds 2d[8] and 2 f–h[14–16] were previously partially charac-
terized; their full data are provided in the Supporting Information.
meso/dl-1,1’-Bis(phenylsulfinyl)ferrocene (3l): Prepared with mCPBA
(1.28 g, 7.0 mmol). Flash chromatography hexane/EtOAc (9:1) gradient
to CH2Cl2/iPrOH: 10:1. Yield: 0.90 g (84%) as an unassigned 1.2:1 meso/
Isopropyl ferrocenecarboxylate (2a): Prepared with ClCO2iPr (10.75 mL,
10.76 mmol, 1.0m solution in toluene). Flash chromatography hexane gra-
dient to hexane/EtOAc: 4:1. Yield: 1.17 g (80%) as an orange solid. M.p.
1
dl mixture as
a yellow solid. M.p. 164–1668C; Rf (EtOAc)=0.53;
318C; Rf (hexane/EtOAc 9:1)=0.60; H NMR (400 MHz): d=5.17 (sept,
1H NMR (400 MHz): d=7.61 (m, 4H; o-HAr*), 7.59 (m, 4H; o-HAr),
7.45–7.43 (m, 12H; m-HAr*, p-HAr*, m-HAr, p-HAr), 4.80 (m, 1H; Ha),
4.78 (m, 1H; Ha*), 4.72 (m, 1H; Ha’*), 4.70 (m, 1H; Ha’), 4.66 (m, 1H;
Hb), 4.65 (m, 2H; Hb*, Hb’*), 4.58 ppm (m, 1H; Hb’); 13C NMR
(100 MHz): d=146.1 (s; ipsoCAr*), 146.0 (s; ipsoCAr), 130.9 (d; p-CAr*),
130.8 (d; p-CAr), 129.2 (d; m-CAr, m-CAr*), 124.2 (d; o-CAr), 124.1 (d; o-
CAr*), 96.8 (s; ipsoCCp, ipsoCCp*), 72.2 (d; Cb), 72.0 (d; Cb*), 71.9 (d;
Cb’*), 71.8 (d; Cb’), 70.2 (d; Ca), 69.9 (d; Ca*), 66.4 (d; Ca’*), 66.3 ppm
(d; Ca’); IR: n˜ =3085, 1674, 1476, 1443, 1415, 1306, 1164, 1106, 1044, 828,
750 cmꢁ1; UV/Vis (CH2Cl2): lmax (loge)=436 (2.42), 282 (sh) (3.57),
248 nm (3.61); MS (ESI): m/z (%): 459/457/455 (11/90/5) [M+Na+], 437/
435/433 (12/100/7) [M+H+]; elemental analysis calcd (%) for
C22H18FeO2S2 (434.35): C 60.83, H 4.18; found: C 60.93, H 4.26.
J=6.3 Hz, 1H; CO2CH
1.9 Hz, 2H; Hb), 4.19 (s, 5H; CpH), 1.33 ppm (d, J=6.2 Hz, 6H;
CO2CH (CH3)2), 72.1
(CH3)2); 13C NMR (100 MHz): d=171.2 (s; CO2CH
(s; ipsoCCp), 71.2 (d; Cb), 70.2 (d; Ca), 69.8 (d; CpC), 67.4 (d; CO2CH-
(CH3)2), 22.2 ppm (q; CO2CH(CH3)2); IR: n˜ =2978, 2934, 1703, 1656,
ACHTUNGTRENNUNG(CH3)2), 4.80 (t, J=1.9 Hz, 2H; Ha), 4.37 (t, J=
G
ACHTUNGTRENNUNG
E
ACHTUNGTRENNUNG
1458, 1374, 1273, 1141, 1104, 1024, 1001, 932, 820, 774 cmꢁ1; UV/Vis
(CH3CN): lmax (loge)=445 (2.31), 335 (2.50), 307 (2.97), 260 (3.53),
221 nm (sh) (3.37); MS (EI): m/z (%): 272/270 (32/5) [M+], 230/228 (100/
4) [M+ꢁH2C=CHCH3]; HRMS (EI): m/z calcd for C14H1656FeO2:
272.0500; found: 272.0501; elemental analysis calcd (%) for C14H16FeO2
(272.12): C 61.79, H 5.93; found: C 61.71, H 5.99.
Disubstituted ferrocenes
3 (general procedure): TMEDA (1.95 mL,
12.9 mmol) and nBuLi (7.39 mL, 11.84 mmol, 1.6m in hexane) were
added dropwise to a stirred solution of (C5H5)2Fe (1.0 g, 5.38 mmol) in
hexane (10 mL) under a dry N2 atmosphere at room temperature. The
solution was stirred at room temperature overnight. The orange slurry
was allowed to settle and the hexane layer was removed with a syringe.
The remaining orange powder was washed with dry hexane (5 mL) and
dissolved in dry THF (10 mL). The electrophile (see at the individual
compounds and in the Supporting Information) was added to the orange
1,1’-Bis(phenylsulfonyl)ferrocene (3m): Prepared with mCPBA (2.13 g,
12.40 mmol). Crystallized from CH2Cl2. Yield: 0.95 g (82%) as orange
crystals. M.p. 2868C (decomp); Rf (EtOAc)=0.85; 1H NMR (400 MHz):
d=7.85 (m, 4H; o-HAr), 7.55 (m, 2H; p-HAr), 7.47 (m, 4H; m-HAr), 4.93
(t, J=2.0 Hz, 4H; Ha), 4.89 ppm (t, J=2.0 Hz, 4H; Hb); 13C NMR
(100 MHz): d=142.4 (s; ipsoCAr), 133.2 (d; p-CAr), 129.3 (d; m-CAr),
127.0 (d; o-CAr), 92.4 (s; ipsoCCp), 74.6 (d; Cb), 71.6 ppm (d; Ca); IR:
&
8
&
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Single-Electron Oxidants
FULL PAPER
n˜ =3096, 1446, 1316, 1306, 1144, 1085, 829, 756 cmꢁ1; UV/Vis (CH2Cl2):
lmax (loge)=437 (2.54), 306 (3.37), 280 (sh) (3.56), 246 nm (3.60); MS
(EI): m/z (%): 491/489/487 (12/100/6) [M+Na+], 469/467/465 (10/68/3)
[M+H+]; elemental analysis calcd (%) for C22H18FeO4S2 (466.35): C
56.66, H 3.89; found: C 56.54, H 3.74.
3116, 1653, 1596, 1314, 826, 556 cmꢁ1; UV/Vis (CH2Cl2): lmax (loge)=655
(2.54), 290 (sh) (4.26), 259 nm (4.46); MS (ESI): m/z (%): 417 (100)
[M+Na+ꢁPF6]; HRMS (ESI): m/z calcd for C24H1856FeO2Na+: 417.0548;
found: 417.0548; elemental analysis calcd (%) for C24H18F6FeO2P
(539.21): C 53.46, H 3.36, F 21.14, Fe 10.36, P 5.74; found C 53.86, H
3.37, F 20.86, Fe 10.41, P 5.68.
Ferrocenium hexafluorophosphates 4a, 4b, 4d–h, 4k–m, and 5a, 5b, 5g,
5h: Benzoquinone or DDQ (0.5 mmol) (see Table 3) was added to ferro-
cenes 2a, 2b, 2d–h, 2k–m, or 3a, 3b, 3g, 3h (1.0 mmol) in diethyl ether
(20 mL) at room temperature. The reaction mixture was stirred for
10 min. Then HPF6 (65 wt% solution in H2O, 2.0 mmol) was added at
08C. A greenish blue solid was formed, which was filtered through a sin-
tered glass funnel and washed with ice cold ether until the ether solution
became colorless to give products 4a, 4b, 4d–h, 4k–m, or 5a, 5b, 5g, 5h.
The analytical data of representative compounds 4a, 4b, 4g, 4h, 4m, and
5a, 5b, 5g, 5h are presented below. The full analytical characterization
of all prepared compounds can be found in the Supporting Information.
1,1’-Dibromoferrocenium hexafluorophosphate (5g): Yield: 1.02 g (72%)
as greenish blue solid. M.p. 177–1798C (decomp); IR: n˜ =3121, 1416,
827, 557 cmꢁ1; UV/Vis (CH2Cl2): lmax (loge)=716 (2.35), 400 (2.94),
282 nm (3.98); MS (ESI): m/z (%): 346/344/342/340 (55/100/75/5) [M+
ꢁPF6], 300/298/296 (13/12/1), 256/254/252 (11/18/1); HRMS (ESI): m/z
calcd for C10H879Br256Fe+: 341.8337; found: 341.8339; elemental analysis
calcd (%) for C10H8Br2F6FeP (488.79): C 24.57, H 1.65; found: C 24.65, H
1.59.
1,1’-Diiodoferrocenium hexafluorophosphate (5h): Yield: 1.07 g (81%)
as a blue solid. M.p. 133–1358C (melting with decomp); IR: n˜ =3116,
1409, 826, 556 cmꢁ1; UV/Vis (CH2Cl2): lmax (loge)=725 (2.47), 410 (3.23),
320 (sh) (3.73), 287 (sh) (3.93), 259 (3.92), 236 nm (sh) (3.86); MS (ESI):
m/z (%): 438/436 (100/15) [M+ꢁPF6], 312/310 (55/5) [M+H+ꢁPF6ꢁI];
HRMS (ESI): m/z calcd for C10H8I256Fe+: 437.8059; found: 437.8061; ele-
mental analysis calcd (%) for C10H8F6I2FeP (582.79): C 20.61, H 1.38, Fe
9.58; found: C 20.81, H 1.43, Fe 9.94.
(Isopropyloxycarbonyl)ferrocenium hexafluorophosphate (4a): Yield:
1.37 g (90%) as a blue solid. M.p. 131–1328C (melting with decomp); IR:
n˜ =3010, 1716, 1284, 1109, 824, 556 cmꢁ1; UV/Vis (CH2Cl2): lmax (loge)=
628 (2.31), 360 (3.17), 282 (sh) (3.94), 259 (4.06), 236 nm (sh) (3.95); MS
(ESI): m/z (%): 272/270 (48/5) [M+ꢁPF6], 231/229 (82/7) [M+H+
ꢁPF6ꢁH2C=CHCH3], 230/228 (82/7) [M+ꢁPF6ꢁH2C=CHCH3], 213/211
(100/12); HRMS (ESI): m/z calcd for C14H1656FeO2+: 272.0494; found:
272.0493; elemental analysis calcd (%) for C14H16F6FeO2P (417.08): C
40.32, H 3.87, F 27.33, Fe 13.39; found: C 40.25, H 3.81, F 27.04, Fe
13.68.
Oxidative cyclization of diethyl 2-(5,5-diphenylpent-4-en-1-yl)malonate 8
(general procedure): nBuLi (0.17 mL, 0.27 mmol, 1.6m in hexane) was
added dropwise to a stirred solution of 8 (0.100 g, 0.26 mmol) in 1,2-di-
methoxyethane (10 mL) under a dry N2 atmosphere at ꢁ788C. The reac-
tion mixture was stirred at ꢁ788C for 15 min. The ferrocenium hexafluor-
ophosphates A, 4m, 5a, or 5b (0.65 mmol, 2.5 equiv) were added in one
portion with vigorous stirring at ꢁ788C. TLC indicated complete conver-
sion after completion of the addition. The reaction was quenched after
15 min by the addition of a few drops of water. The solution was diluted
with ether and filtered through a pad of silica gel. The solvent was evapo-
rated under reduced pressure, the nonhomogeneous residue was pread-
sorbed on silica gel, and purified by flash column chromatography
(hexane, gradient to hexane/EtOAc 9:1, 9 eluted last). The analytical
data of 9 and 10 are in agreement with those published.[3g] The analytical
data of compounds 11 and 12 are found in the Supporting Information.
Benzoylferrocenium hexafluorophosphate (4b): Yield: 1.22 g (82%) as a
green solid. M.p. 118–1208C (melting with decomp); IR: n˜ =3120, 1661,
1597, 1280, 834, 557 cmꢁ1; UV/Vis (CH2Cl2): lmax (loge)=635 (2.56), 290
(sh) (4.07), 258 nm (4.29); MS (ESI): m/z (%): 313/311 (27/2) [M+Na+
ꢁPF6], 290/288 (100/20) [M+ꢁPF6]; HRMS (ESI): m/z calcd for
C17H1456FeO+: 290.0389; found: 290.0387; elemental analysis calcd (%)
for C17H14F6FeOP (435.10): C 46.93, H 3.24, Fe 12.83, P 7.12; found: C
46.99, H 3.25, Fe 13.17, P 7.39.
Bromoferrocenium hexafluorophosphate (4g): Yield: 1.14 g (74%) as a
green solid. M.p. 233–2358C (decomp); IR: n˜ =3121, 1420, 826, 731,
556 cmꢁ1; UV/Vis (CH2Cl2): lmax (loge)=684 (2.31), 565 (2.12), 272 (sh)
(4.18), 263 nm (4.19); MS (ESI): m/z (%): 266/264/262 (85/100/5) [M+
ꢁPF6], 186/184 (66/6) [M+H+ꢁPF6ꢁBr]; HRMS (ESI): m/z calcd for
C10H979Br56Fe+: 263.9232; found: 263.9231; elemental analysis calcd (%)
for C10H9BrF6FeP (409.89): C 29.30, H 2.21, Br 19.49, F 27.81, Fe 13.62, P
7.56; found: C 29.45, H 2.10, Br 19.56, F 27.89, Fe 13.38, P 7.42.
Computational methods: The quantum chemical calculations were per-
formed using the TURBOMOLE 6.3 program. The geometry optimiza-
tions were carried out at the DFT level, using the PBE functional.[26] The
DFT (PBE) calculations were expedited by expanding the Coulomb inte-
grals in an auxiliary basis set, the resolution-of-identity (RI-J) approxi-
mation.[27] For the geometry optimization, the def2-SVP basis set was em-
ployed on all of the atoms,[28] whereas the def2-TZVP basis set was used
for the final single-point calculations to obtain presumably accurate mo-
lecular energies.[29]
Iodoferrocenium hexafluorophosphate (4h): Yield: 1.17 g (80%) as a
blue solid. M.p. 233–2358C (melting with decomp); IR: n˜ =3116, 1413,
829, 557 cmꢁ1; UV/Vis (CH2Cl2): lmax (loge)=718 (2.68), 278 (sh) (3.97),
259 nm (4.03). MS (ESI): m/z (%): 312/310 (100/6) [M+ꢁPF6]. HRMS
(ESI): m/z calcd for C10H9I56Fe+: 311.9093; found: 311.9092; elemental
analysis calcd (%) for C10H9F6IFeP (456.89): C 26.29, H 1.99, I 27.78, Fe
12.22, P 6.78; found: C 26.28, H 2.00, I 27.40, Fe 12.64, P 7.25.
The solvation effects were taken into account by using the COSMO-RS
method[6,30] using TURBOMOLE 6.3 for the COSMO calculation[31] with
er =1 (ideal screening) and the COSMOtherm[32] program for the subse-
quent COSMO-RS calculation. The recommended protocol involving the
Becke–Perdew (B-P) functional[33] for the in vacuo and the er =1 calcu-
lations together with the def-TZVP basis set was used.
(Phenylsulfonyl)ferrocenium hexafluorophosphate (4m): Yield: 1.03 g
(72%) as a green solid. M.p. 118–1208C; IR: n˜ =3108, 1447, 1304, 1141,
840, 724, 558 cmꢁ1; UV/Vis (CH2Cl2): lmax (loge)=635 (2.39), 287 (sh)
(4.35), 259 (4.51), 230 nm (4.47); MS (EI): m/z (%): 328/326/324 (8/100/
18) [M+ꢁPF6]; HRMS (ESI): m/z calcd for C16H1456FeO232S+: 326.0064;
found: 326.0060; elemental analysis calcd (%) for C16H14F6FeO2PS
(471.16): C 40.79, H 3.00, Fe 11.85, S 6.81; found: C 40.27, H 3.01, Fe
12.21, S 7.21.
The Gibbs free energy was then calculated as the sum of the following
contributions [Eq. (1)]:
G ¼ Eel þ Gsolv þ EZPVEꢁRT lnðqtransqrotqvib
Þ
ð1Þ
1,1’-Bis(isopropyloxycarbonyl)ferrocenium hexafluorophosphate (5a):
in which Eel is the in vacuo energy of the system (at the RI-PBE/def2-
TZVP//RI-PBE/def2-SVP level) and Gsolv is the solvation free energy
(calculated using the RI-BP/def-TZVP(COSMO-RS, e=1, e=1)
method as described above). EZPVE is the zero-point energy, and
ꢁRTln(qtransqrotqvib) accounts for thermal corrections to the enthalpy and
entropic terms. EZPVE and ꢁRTln(qtransqrotqvib) were obtained from a fre-
quency calculation by using the same method and software as for the ge-
ometry optimization (RI-PBE/def2-SVP level), 298 K, and 1 atm using
the ideal-gas approximation.
Yield: 1.14 g (81%) as
a blue solid. M.p. 90–928C (melting with
decomp); IR: n˜ =2987, 1722, 1471, 1376, 1280, 1107, 829, 557 cmꢁ1; UV/
Vis (CH2Cl2): lmax (loge)=641 (2.41), 287 (sh) (3.84), 262 (4.00), 229 nm
(4.08); MS (ESI): m/z (%): 381/379 (100/10) [M+Na+ꢁPF6]; HRMS
(ESI): m/z calcd for C18H2256FeO4Na+: 381.0760; found: 381.0759; ele-
mental analysis calcd (%) for C18H22F6FeO4P (503.17): C 42.97, H 4.41, F
22.65, Fe 11.10, P 6.16; found: C 42.47, H 4.33, F 22.28, Fe 11.13, P 6.18.
1,1’-Dibenzoylferrocenium hexafluorophosphate (5b): Yield: 1.03 g
(76%) as a green solid. M.p. 99–1018C (melting with decomp); IR: n˜ =
The reduction potentials were then calculated according to Equation (2):
Chem. Eur. J. 2012, 00, 0 – 0
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
9
&
ÞÞ
These are not the final page numbers!

U. Jahn et al.
E0 ½Vꢂ ¼ 27:21ðGox ½a:u:ꢂꢁGred ½a:u:ꢂÞꢁE0absðNHEÞ ½Vꢂ
[9] S. R. Alley, W. Henderson, J. Organomet. Chem. 2001, 637–639,
216.
ð2Þ
[10] For applications of 2a, see: a) H. Naka, M. Uchiyama, Y. Matsumo-
to, A. E. H. Wheatley, M. McPartlin, J. V. Morey, Y. Kondo, J. Am.
Chem. Soc. 2007, 129, 1921; b) G. Diehl, A. Liesener, U. Karst, Ana-
lyst 2001, 126, 288.
[11] Alternative preparation: a) M. A. Carroll, A. J. P. White, D. A. Wid-
dowson, D. J. Williams, J. Chem. Soc. Perkin Trans. 1 2000, 1551;
b) M. Rausch, M. Vogel, H. Rosenberg, J. Org. Chem. 1957, 22, 903;
c) N. Weliky, E. S. Gould, J. Am. Chem. Soc. 1957, 79, 2742.
[12] Alternative preparation: a) G. P. Sollott, H. E. Mertwoy, S. Portnoy,
J. L. Snead, J. Org. Chem. 1963, 28, 1090; b) A. K. Diallo, J. Ruiz, D.
Astruc, Org. Lett. 2009, 11, 2635.
in which Gox and Gred are the free energies calculated according to Equa-
tion (1) and E0
absA(NHE) is the absolute potential of the normal hydrogen
electrode (NHE). In the literature, the values for E0
absA(NHE) range be-
tween 4.24 and 4.5 V;[34] the most recent value of E0
absA(NHE)=4.281 V
was advocated by Isse and Gennaro.[35] Solely for the sake of conven-
ience, the value of E0
absA(NHE)=4.291 V was used, which yielded the best
CHTUNGTRENNUNG
CTHUNGTRENNUNG
CTHUNGTRENNUNG
CHTUNGTRENNUNG
agreement between the calculated and experimental data, that is, the
mean signed error (MSE) equal to zero. In Table 5, the calculated data
are also referenced to the AgjAgCl 1m LiCl electrode, that is, 0.21 V
[23]
was subtracted from Ec0alcd
.
[13] Alternative preparation: a) M. D. Rausch, J. Org. Chem. 1961, 26,
3579; b) C. Pichon, B. Odell, J. M. Brown, Chem. Commun. 2004,
598; c) P. Diter, S. Taudien, O. Samuel, H. B. Kagan, J. Org. Chem.
1994, 59, 370; d) P. Diter, S. Taudien, O. Samuel, H. B. Kagan, Tetra-
hedron: Asymmetry 1994, 5, 549; e) J. A. Adeleke, Y.-W. Chen, L.-
K. Liu, Organometallics 1992, 11, 2543.
[14] Similar results were obtained with TsBr: T.-Y. Dong, L.-L. Lai, J.
Organomet. Chem. 1996, 509, 131.
[15] For alternative preparations, see: a) A. S. Romanov, J. M. Mulroy,
V. N. Khrustalev, M. Y. Antipin, T. V. Timofeev, Acta Crystallogr. C
2009, 65, m426; b) R. W. Fish, M. Rosenblum, J. Org. Chem. 1965,
30, 1253.
Acknowledgements
Generous funding by the Institute of Organic Chemistry and Biochemis-
try, the Academy of Sciences of the Czech Republic (RVO: 61388963),
and the Grant Agency of the Czech Republic (grant no. 203/09/0705) is
gratefully acknowledged. I.C. thanks the Ministry of Education, Youth,
and Sports of the Czech Republic (MSM0021620857) for financial sup-
port.
[1] For reviews, see: a) Ferrocenes: Ligands, Materials and Biomolecules
(Ed.: P. Stepnicka), Wiley, Hoboken, 2008; b) Ferrocenes (Eds.: A.
Togni, T. Hayashi), Wiley-VCH, Weinheim, 1995; c) A. Togni, N.
Bieler, U. Burckhardt, C. Kçllner, G. Pioda, R. Schneider, A.
Schnyder, Pure Appl. Chem. 1999, 71, 1531; d) R. Gꢃmez Arrayꢁs, J.
Adrio, J. C. Carretero, Angew. Chem. 2006, 118, 7836; Angew.
Chem. Int. Ed. 2006, 45, 7674; e) N. J. Long, Angew. Chem. 1995,
107, 37; Angew. Chem. Int. Ed. Engl. 1995, 34, 21; Angew. Chem.
1995, 107, 37; f) D. R. van Staveren, N. Metzler-Nolte, Chem. Rev.
2004, 104, 5931.
[2] N. G. Connelly, W. E. Geiger, Chem. Rev. 1996, 96, 877.
[3] a) U. Jahn, E. Dinca, J. Org. Chem. 2010, 75, 4480; b) U. Jahn, E.
Dinca, Chem. Eur. J. 2009, 15, 58; c) U. Jahn, F. Kafka, R. Pohl,
P. G. Jones, Tetrahedron 2009, 65, 10917; d) U. Jahn, P. Hartmann, E.
Kaasalainen, Org. Lett. 2004, 6, 257; e) U. Jahn, P. Hartmann, I. Dix,
P. G. Jones, Eur. J. Org. Chem. 2002, 718; f) U. Jahn, Chem.
Commun. 2001, 1600; g) U. Jahn, P. Hartmann, I. Dix, P. G. Jones,
Eur. J. Org. Chem. 2001, 3333; h) U. Jahn, M. Mꢄller, S. Aussieker,
J. Am. Chem. Soc. 2000, 122, 5212; i) U. Jahn, J. Org. Chem. 1998,
63, 7130.
[4] a) E. S. Krygowski, K. Murphy-Benenato, M. D. Shair, Angew.
Chem. 2008, 120, 1704; Angew. Chem. Int. Ed. 2008, 47, 1680;
Angew. Chem. 2008, 120, 1704; b) J. M. Richter, B. W. Whitefield,
T. J. Maimone, D. W. Lin, M. P. Castroviejo, P. S. Baran, J. Am.
Chem. Soc. 2007, 129, 12857; c) J. P. Goddard, C. Gomez, F. Brebion,
S. Beauviere, L. Fensterbank, M. Malacria, Chem. Commun. 2007,
2929; d) M. P. Sibi, M. Hasegawa, J. Am. Chem. Soc. 2007, 129,
4124; e) F. C. Pigge, J. J. Coniglio, N. P. Rath, J. Org. Chem. 2004, 69,
1161; f) P. Q. Nguyen, H.-J. Schꢅfer, Org. Lett. 2001, 3, 2993; g) T.
Langer, M. Illich, G. Helmchen, Synlett 1996, 1137.
[16] For alternative preparation, see: a) J. C. Goeltz, C. P. Kubiak, Orga-
nometallics 2011, 30, 3908; b) G. Laus, K. Wurst, W. Stolz, H. Schot-
tenberger, Z. Kristallogr. New Cryst. Struct. 2005, 220, 229; c) I. R.
Butler, S. B. Wilkes, S. J. McDonald, L. J. Hobson, A. Taralp, C. P.
Wilde, Polyhedron 1993, 12, 129.
[17] For a related preparation, see: a) J. J. Bishop, A. Davison, M. L.
Katcher, D. W. Lichtenberg, R. E. Merrill, J. C. Smart, J. Organomet.
Chem. 1971, 27, 241. For electrochemical oxidations of dppf, see:
b) G. Pilloni, B. Longato, B. Corain, J. Organomet. Chem. 1991, 420,
57; c) D. Brett, B. D. Swartz, C. Nataro, Organometallics 2005, 24,
2447.
[18] Alternative preparation: B. McCulloch, D. L. Ward, J. D. Woollins,
C. H. Brubaker Jr., Organometallics 1985, 4, 1425.
[19] Alternative preparation and electrochemical investigation of 2k–m:
A. Gref, P. Diter, D. Guillaneux, H. B. Kagan, New J. Chem. 1997,
21, 1353.
[20] Prepared by using HPF6 in analogy to: M. Shibuya, M. Tomizawa,
Y. Iwabuchi, J. Org. Chem. 2008, 73, 4750.
[21] T. E. Pickett, C. J. Richards, Tetrahedron Lett. 1999, 40, 5251.
[22] CCDC-877550 (2a), CCDC-877552 (3a), CCDC-877551 (2m), and
CCDC-877549 (3m) contain the supplementary crystallographic
data for this paper. These data can be obtained free of charge from
The
www.ccdc.cam.ac.uk/data_request/cif.
[23] a) R. Sokolovꢁ, I. Degano, M. Hromadovꢁ, J. Bulꢀckovꢁ, M. Gꢁl, M.
Cambridge
Crystallographic
Data
Centre
via
ˇ
ˇ
Valꢁsek, Collect. Czech. Chem. Commun. 2010, 75, 1097. For an
overview, see: b) D. J. G. Ives, G. J. Janz, Reference Electrodes, Aca-
demic Press, New York, 1961.
[24] M. D. Rausch, D. J. Ciappenelli, J. Organomet. Chem. 1967, 10, 127.
[25] M. Sawamura, A. Yamauchi, T. Takegawa, Y. Ito, J. Chem. Soc.
Chem. Commun. 1991, 874.
[26] J. P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett. 1996, 77, 3865.
[27] K. Eichkorn, O. Treutler, H. ꢆhm, M. Hꢅser, R. Ahlrichs, Chem.
Phys. Lett. 1995, 240, 283.
[28] F. Weigend, R. Ahlrichs, Phys. Chem. Chem. Phys. 2005, 7, 3297.
[29] A. Schꢅfer, C. Huber, R. Ahlrichs, J. Chem. Phys. 1994, 100, 5829.
[30] A. Klamt, J. Phys. Chem. 1995, 99, 2224.
[31] a) A. Klamt, G. Schuurmann, J. Chem. Soc. Perkin Trans. 2 1993,
799; b) A. Schꢅfer, A. Klamt, D. Sattel, J. C. W. Lohrenz, F. Eckert,
Phys. Chem. Chem. Phys. 2000, 2, 2187.
[5] a) H. G. Lee, J. Y. Ahn, A. S. Lee, M. D. Shair, Chem. Eur. J. 2010,
16, 13058 and cited ref; b) J. Xu, E. J. E. Caro-Diaz, L. Trzoss, E. A.
Theodorakis, J. Am. Chem. Soc. 2012, 134, 5072–5075.
[6] A. Klamt, V. Jonas, T. Buerger, J. C. W. Lohrenz, J. Phys. Chem. A
1998, 102, 5074.
[7] a) P. Winget, C. J. Cramer, D. G. Truhlar, Theor. Chem. Acc. 2004,
´
112, 217; b) M. Srnec, J. Chalupsky, M. Fojta, L. Zendlovꢁ, L.
´
ˇ
Havran, M. Hocek, M. Kyvala, L. Rulꢀsek, J. Am. Chem. Soc. 2008,
ˇ
130, 10947; c) C. K. Chua, M. Pumera, L. Rulꢀsek, J. Phys. Chem. C,
2012, 116, 4243–4251; d) M. Namazian, C. Y. Lin, M. L. Coote, J.
Chem. Theory Comput. 2010, 6, 2721.
[8] a) R. Sanders, U. T. Mueller-Westerhoff, J. Organomet. Chem. 1996,
512, 219; b) O. Oms, F. Maurel, F. Carre, J. Le Bideau, A. Vioux, D.
Leclercq, J. Organomet. Chem. 2004, 689, 2654.
[32] COSMOtherm, version C2.1, release 01.10. COSMOlogic GmbH &
Co KG, Leverkusen, Germany.
&
10
&
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Single-Electron Oxidants
FULL PAPER
[33] a) A. D. Becke, Phys. Rev. A 1988, 38, 3098; b) S. H. Vosko, L. Wilk,
M. Nusair, Can. J. Phys. 1980, 58, 1200; c) J. P. Perdew, Phys. Rev. B
1986, 33, 8822.
[35] A. A. Isse, A. Gennaro, J. Phys. Chem. B 2010, 114, 7894.
[34] a) J. Llano, L. A. Eriksson, Phys. Chem. Chem. Phys. 2004, 6, 4707;
b) S. Trasatti, Pure Appl. Chem. 1986, 58, 955; c) C. P. Kelly, C. J.
Cramer, D. G. Truhlar, J. Phys. Chem. B 2007, 111, 408.
Received: April 30, 2012
Published online: && &&, 0000
Chem. Eur. J. 2012, 00, 0 – 0
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
11
&
ÞÞ
These are not the final page numbers!

U. Jahn et al.
Expanding the range: A combined
Ferrocenium Salts
synthetic, electrochemical, and theoret-
ical approach lays the foundation for
the application of new ferrocenium
salts with fine-tuned oxidative power,
as exemplified by a double oxidative
cyclization to bicyclic lactones under
very mild conditions (see scheme).
D. A. Khobragade, S. G. Mahamulkar,
ˇ
ˇ
ˇ
L. Pospꢀsil, I. Cꢀsarovꢁ, L. Rulꢀsek,
U. Jahn* ....................... &&&& — &&&&
Acceptor-Substituted Ferrocenium
Salts as Strong, Single-Electron
Oxidants: Synthesis, Electrochemistry,
Theoretical Investigations, and Initial
Synthetic Application
&
12
&
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!