Synthesis and chemical oxidation of 3-ferrocenylpyrrole and ferrocenyl-substituted triazoles: Iron versus ligand based oxidation

Article abstract of DOI:10.1016/j.jorganchem.2009.06.011

Copper catalyzed [3+2] cycloaddition reactions between ethynylferrocene and benzylazides yields 1-benzyl-4-ferrocenyl-1,2,3-triazoles (2-5). Reaction between phenylacetylene and azidoferrocene yields 1-ferrocenyl-4-phenyl-1,2,3-triazole (6). Anodic electrochemistry of 2-6 suggests reversible oxidation at potentials more positive than ferrocene. Chemical oxidation of 2 and 3-ferrocenylpyrrole (1) with dichlorodicyanoquinone (DDQ) yields the salts [2^{+{radical dot}}] [DDQ^{{radical dot}-}] and [1^{+{radical dot}}] [DDQ^{{radical dot}-}], respectively. ⁵⁷Fe M?ssbauer spectroscopy reveals the presence of low-spin Fe^II in [1^{+{radical dot}}][DDQ^{{radical dot}-}] while Fe^II is oxidized to low-spin Fe^III in [2^{+{radical dot}}][DDQ^{{radical dot}-}]. Magnetization measurements indicate that [1^{+{radical dot}}][DDQ^{{radical dot}-}] is paramagnetic and cannot be viewed as a simple neutral charge transfer complex reminiscent of the mixed stack diamagnetic [ferrocene]⁰[TCNE]⁰.

Full text of DOI:10.1016/j.jorganchem.2009.06.011

Journal of Organometallic Chemistry 694 (2009) 3262–3269
Contents lists available at ScienceDirect
Journal of Organometallic Chemistry
journal homepage: www.elsevier.com/locate/jorganchem
Synthesis and chemical oxidation of 3-ferrocenylpyrrole
and ferrocenyl-substituted triazoles: Iron versus ligand based oxidation
Michael Verschoor-Kirss a, Joszef Kreisz a, William Feighery b, William M. Reiff a,
,
*
Christoph M. Frommen a, Rein U. Kirss a,
*
a Department of Chemistry and Chemical Biology, Northeastern University, Boston, MA 02115, United States
b Department of Chemistry, Indiana University-South Bend, South Bend, IN 46634, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
Copper catalyzed [3+2] cycloaddition reactions between ethynylferrocene and benzylazides yields 1-ben-
zyl-4-ferrocenyl-1,2,3-triazoles (2–5). Reaction between phenylacetylene and azidoferrocene yields 1-
ferrocenyl-4-phenyl-1,2,3-triazole (6). Anodic electrochemistry of 2–6 suggests reversible oxidation at
potentials more positive than ferrocene. Chemical oxidation of 2 and 3-ferrocenylpyrrole (1) with dichlo-
rodicyanoquinone (DDQ) yields the salts [2+Å] [DDQÅꢀ] and [1+Å] [DDQÅꢀ], respectively. 57Fe Mössbauer
spectroscopy reveals the presence of low-spin FeII in [1+Å][DDQÅꢀ] while FeII is oxidized to low-spin FeIII
in [2+Å][DDQÅꢀ]. Magnetization measurements indicate that [1+Å][DDQÅꢀ] is paramagnetic and cannot be
viewed as a simple neutral charge transfer complex reminiscent of the mixed stack diamagnetic
[ferrocene]0[TCNE]0.
Received 6 January 2009
Received in revised form 8 June 2009
Accepted 10 June 2009
Available online 13 June 2009
Keywords:
Copper catalyzed [3+2] cycloaddition
reactions
Ferrocenyl-1,2,3-triazoles
3-Ferrocenylpyrrole
Ó 2009 Elsevier B.V. All rights reserved.
57Fe Mössbauer spectroscopy
Oxidation of ferrocenes
1. Introduction
the iron(II) center. For example, electrochemical oxidation of
1,10-bis(diphenylphosphino) ferrocene (dppf) and 1,10-bis(dii-
Copper catalyzed [3+2] cycloaddition (‘‘click”) reactions have
been used to functionalize dendrimers [1], modify graphite
surfaces [2] and self-assembled monolayers at electrodes [3] with
4-ferrocenyl-1,2,3-triazoles. The functionalized electrodes allow
for studies of electron transfer rates in a variety of redox processes.
A fundamental assumption in the latter studies has been that
oxidation occurs at the FeII of the ferrocene moiety. Evidence sup-
porting this assumption is found in the spectroelectrochemistry of
sopropylphosphino) ferrocene (dippf) leads to the dimerization of
phosphorus-centered radicals [7]. Similarly, the anodic electro-
chemistry of [CpFe(C5H4)]3P, [CpFe(C5H4)]2PPh, [CpFe(C5H4)]2P@
O(Ph) and [CpFe(C5H4)]3P@Se, is consistent with loss of an electron
from a largely ferrocenyl-based HOMO with significant phosphine-
or phosphine chalcogenide character [8]. In these ferrocenylphos-
phines, electrochemical oxidation also appears to lead to products
derived from dimerization of phosphorus-centered radicals [9].
Anodic oxidation of 1,10-bis(di-tert-butylpropylphosphino)ferro-
cene (dtbpf), however, is reversible and does not lead to dimeriza-
tion, reflecting the larger substituents on phosphorus [10].
Chemical oxidation of [CpFe(C5H4)]3P and [CpFe(C5H4)]2PPh
with DDQ yields products containing the DDQÅꢀ anion and low-
spin FeII [11]. The possibility that oxidation of ferrocenylpyrrole
and related ferrocene-substituted nitrogen heterocycles might lead
to loss of an electron from a HOMO with significant nitrogen char-
acter rather than iron, led to the present study of the chemical oxi-
dation of 1 and 1-benzyl-4-ferrocenyl-1,2,3-triazole (2). In this
paper, we report on the synthesis and characterization of five
1,10-bis(4-pyridyl) ferrocene where oxidation leads to
a UV
spectrum consistent with formation of a ferrocenium ion [4]. On
the other hand, the nitrogen containing ferrocene derivative
3-ferrocenylpyrrole (1) undergoes a reversible, electrochemical
oxidation at 0.25 V versus SCE in acetonitrile solution, however,
electrochemistry alone could not distinguish between oxidation
at the FeII center and oxidation of the pyrrole substituent [5].
In related work, the premise that oxidation of ferrocenes with
phosphorus-containing substituents occurs at iron has recently
been explored by several groups [6–9]. Studies of the chemical
and electrochemical oxidation of ferrocenyl phosphines indicate
that oxidation and/or intramolecular electron transfer results in
the apparent loss of electrons from the phosphorus rather than
ferrocene-substituted
triazoles,
1-benzyl-4-ferrocenyl-1,2,3-
triazole (2), 1-(p-methylbenzyl)-4-ferrocenyl-1,2,3-triazole (3),
1-(m-bromobenzyl)-4-ferrocenyl-1,2,3-triazole (4), 1-(p-bromo-
benzyl)-4-ferrocenyl-1,2,3-triazole (5) and 1-ferrocenyl-4-phenyl-
1,2,3-triazole (6). We also describe the differences in the chemical
oxidation products of 1 and 2.
* Corresponding author. Tel.: +1 617 373 4513; fax: +1 617 373 8795.
E-mail address: rkirss@neu.edu (R.U. Kirss).
0022-328X/$ - see front matter Ó 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2009.06.011

M. Verschoor-Kirss et al. / Journal of Organometallic Chemistry 694 (2009) 3262–3269
3263
residues that remain at the top of the column could not be success-
fully eluted with more polar solvents such as chloroform or meth-
anol. A slightly improved yield (34%) was obtained from reactions
at ambient temperature for 3 days. m.p. 157–158 °C.
1H NMR (CDCl3): d 4.09 s (5H, C5H5), 4.33 m (2H, C5H4), 4.74 m
(2H, C5H4), 5.52 s (2H, CH2), 7.25–7.39 m (6H, C6H5, CH).
(C6D6): d 3.99 s (5H, C5H5), 4.10 m (2H, C5H4), 4.74 m (2H, C5H4),
4.85 s (2H, CH2), 6.72–7.3 m (6H, C6H5, CH).
C6H5
H
R
N
N
N
N
N
N
N
Fe
Fe
Fe
6
13C NMR (CDCl3): singlets at d 54.01, 66.83, 68.94, 69.92, 75.99,
118.77, 127.83, 128.63, 129.06, 134.87, 147.22.
2 R = C6H5CH2
Anal. Calc. for C19H17N3Fe: C, 66.50; H, 4.99, N, 12.24. Found: C,
66.30; H, 4.92; N, 12.01%.
3 R = 4-CH3C6H5CH2
4 R = 3-BrC6H4CH2
5 R = 4-BrC6H4CH2
1
2.2. Ruthenium-catalyzed reaction of ethynylferrocene with benzyl
azide
2. Experimental
All compounds described in this work were handled using
Schlenk techniques or in a M.I. Braun glove-box under purified ar-
gon or nitrogen atmospheres. Ethynylferrocene [12], 3-ferrocenyl-
pyrrole (1) [5] and benzyl azides [13] were prepared by literature
methods. Solvents were purified by refluxing over Na/benzophe-
none (petroleum ether) or P2O5 (dichloromethane) and distilled
prior to use. Elemental analyses (C, H, N) were performed by
Columbia Analytical Services, Inc.
A 5-mm screw capped NMR tube was charged with a solution of
26 mg (0.12 mmol), 19 mg (0.14 mmol) benzyl azide and 3 mg
Cp*Ru(PPh3)2Cl in 0.5 mL C6D6. The tube was heated to 80 °C for
2 h and the composition of the mixture evaluated by 1H NMR. Res-
onances assigned to the starting materials disappeared and were
replaced by resonances assigned to 1-benzyl-5-ferrocenyl-1,2,3-
triazole.
1H NMR (C6D6): d 4.01 s (5H, C5H5), 3.84 m (2H, C5H4), 4.34 m
(2H, C5H4), 5.27 s (2H, CH2), 6.85–7.80 m (6H, C6H5, CH).
NMR spectra were recorded at 300 MHz for 1H, and 75.4 MHz
for 13C{1H} on a Varian Mercury VX300 spectrometer. Proton
chemical shifts are reported relative to residual protons in the sol-
vent (CHCl3 at d 7.24 ppm) referenced to TMS at 0.00 ppm. Carbon
chemical shifts are reported relative to solvent (CHCl3 t at d
77.0 ppm) referenced to TMS at 0.00 ppm.
Electrochemical measurements were made under nitrogen on a
BAS 100 B/W electrochemical workstation at 22 °C using
1 ꢁ 10ꢀ3 M solutions in dry CH2Cl2, 0.1 M nBu4NPF6 as supporting
electrolyte at a scan rate of 100 mV/s. The working electrode was
a 3-mm Pt disk with a Pt wire as auxiliary electrode. A silver wire
was used as a pseudo-reference electrode with ferrocene added as
an internal standard with all potentials referenced to ferrocene
(E1/2 = 0.00 V).
2.3. Preparation of 1-(p-methylbenzyl)-4-ferrocenyl-1,2,3-triazole (3)
Following the procedure described above for 2, reaction of
250 mg (1.2 mmol) ethynylferrocene, 263 mg (1.8 mmol) p-meth-
ylbenzylazide, 30 mg (0.12 mmol) CuSO4ꢂ5H2O and 60 mg
(0.30 mmol) sodium ascorbate at 60 °C for 4 days yielded 139 mg
(33% yield) 1-(p-methylbenzyl)-4-ferrocenyl-1,2,3-triazole (3) as
a yellow-orange solid after chromatography and crystallization
from CH2Cl2/petroleum ether. m.p. 157–159 °C.
1H NMR (CDCl3): d 2.35 s (3H, CH3) 4.21 s (5H, C5H5), 4.48 m
(2H, C5H4), 4.88 m (2H, C5H4), 5.44 s (2H, CH2), 7.00–7.39 m (5H,
C6H4, CH).
Variable temperature Mössbauer spectra were recorded on a
standard constant acceleration apparatus [14] operated in the
13C NMR (CDCl3): singlets at d 21.13, 53.81, 66.67. 68.72, 69.83,
75.72, 118.62, 127.90, 129.71, 131.81, 138.55, 147.10.
Anal. Calc. for C20H19N3Fe: C, 67.24; H, 5.36; N, 11.76. Found: C,
67.14; H, 5.04; N, 11.75%.
mode using a c
-ray source of 25 mCi 57Co in a rhodium metal ma-
trix. Temperature control was achieved using an uncalibrated sili-
con diode coupled to a Lake Shore Cryotronics Model DT-500 set
point controller. Temperature measurements were made using a
2.4. Preparation of 1-(m-bromobenzyl)-4-ferrocenyl-1,2,3-triazole (4)
calibrated silicon diode driven by a 10 lA constant current source
in conjunction with a digital voltmeter. Mössbauer spectra were
fitted using a least-squares-Lorentizian progam [15]. Magnetiza-
tion measurements were performed on a Lakeshore AC susceptom-
eter and a Quantum Design MPMS SQUID device.
Following the procedure described above for 2, reaction of
250 mg (1.2 mmol) ethynylferrocene, 378 mg (1.8 mmol) m-bro-
mobenzylazide, 30 mg (0.12 mmol) CuSO4ꢂ5H2O and 60 mg
(0.30 mmol) sodium ascorbate for 4 days at 60 °C yielded 185 mg
(36% yield) 1-(m-methylbenzyl)-4-ferrocenyl-1,2,3-triazole (4) as
a yellow-orange solid after chromatography and crystallization
from CH2Cl2/petroleum ether. m.p. 158–160 °C.
2.1. Preparation of 1-benzyl-4-ferrocenyl-1,2,3-triazole (2)
Two-hundred fifty milligram (1.2 mmol) ethynylferrocene was
added to a solution of 236 mg (1.8 mmol) benzylazide in 25 mL
50% aqueous tbutanol (v/v) containing 30 mg (0.12 mmol) Cu-
SO4ꢂ5H2O and 60 mg (0.30 mmol) sodium ascorbate. Heating the
mixture at 60 °C for 48 h yielded a yellow-orange slurry. The reac-
tion mixture was poured into 25 mL of water and extracted
3 ꢁ 25 mL dichloromethane. The organic layers were combined,
dried over magnesium sulfate and filtered. Evaporation of solvent
from the filtrate yielded an orange solid that was chromatographed
on alumina using dichloromethane as the eluent. A single orange
fraction was eluted and evaporated to dryness yielding 120 mg
(29% yield) of 1-benzyl-4-ferrocenyl-1,2,3-triazole (2) as a bright
orange powder. Re-crystallization from dichloromethane/petro-
leum ether was used to obtain analytically pure samples. Brown
1H NMR (CDCl3): d 4.07 s (5H, C5H5), 4.31 m (2H, C5H4), 4.72 m
(2H, C5H4), 5.48 s (2H, CH2), 7.15–7.50 m (5H, C6H4, CH).
13C NMR (CDCl3): singlets at d 53.16, 66.56, 68.66, 69.50, 75.04,
118.69, 123.03, 126.26, 130.62, 130.68, 131.75, 137.10, 147.50.
Anal. Calc. for C19H16BrN3Fe: C, 54.07; H, 3.82; N, 9.95. Found: C,
54.03; H, 3.46; N, 9.82%.
2.5. Preparation of 1-(p-bromobenzyl)-4-ferrocenyl-1,2,3-triazole (5)
Following the procedure described above for 2, reaction of
157 mg (0.76 mmol) ethynylferrocene, 297 mg (1.4 mmol) p-bro-
mobenzylazide, 32 mg (0.13 mmol) CuSO4ꢂ5H2O and 44 mg
(0.23 mmol) sodium ascorbate for 4 days at ambient temperature
yielded 79 mg (25% yield) 1-(p-bromobenzyl)-4-ferrocenyl-1,2,3-

3264
M. Verschoor-Kirss et al. / Journal of Organometallic Chemistry 694 (2009) 3262–3269
triazole (5) as a yellow-orange solid after chromatography and
crystallization from CH2Cl2/petroleum ether. m.p. 185–186 °C.
1H NMR (CDCl3): d 4.04 s (5H, C5H5), 4.27 m (2H, C5H4), 4.67 m
(2H, C5H4), 5.48 s (2H, CH2), 7.14 d (J = 8.1 Hz, 2H, C6H4). 7.35 s (1H,
CH), 7.51 d (J = 8.1 Hz, 2H, C6H4).
XXXIX40
8.00E-06
6.00E-06
4.00E-06
2.00E-06
13C NMR (CDCl3): singlets at d 53.31, 66.57, 68.66, 69.58, 72.34,
118.57, 122.79, 129.45, 132.24, 133.86, 147.46.
Anal. Calc. for C19H16BrN3Fe: C, 54.07; H, 3.82; N, 9.95. Found: C,
54.32; H, 3.63; N, 9.85%.
0.00E+00
600
400
200
0
-200
-400
-2.00E-06
2.6. Preparation of 1-(ferrocenyl)-4-phenyl-1,2,3-triazole (6)
-4.00E-06
-6.00E-06
-8.00E-06
A solution of ferrocenylazide [16] was prepared in situ by stir-
ring 114 mg (1.8 mmol) NaN3 and 466 mg (1.8 mmol) bromoferro-
cene in 25 mL 50% aqueous tbutanol (v/v) at ambient temperature
-1.00E-05
for 2 days. Subsequently, 400
lL (3.4 mmol) phenylacetylene,
Voltage (mV)
30 mg (0.12 mmol) CuSO4ꢂ5H2O and 60 mg (0.30 mmol) sodium
ascorbate were added and the reaction mixture stirred at ambient
temperature for 4 days. Following the workup described for 2,
60 mg (10% yield) of 1-ferrocenyl-4-phenyl-1,2,3-triazole (6) was
isolated as a yellow-orange solid after chromatography and crys-
tallization from CH2Cl2/petroleum ether. Compound 6 decomposes
above 215 °C without melting.
Fig. 1. Cyclic voltammogram of 2 at 22 °C (1 ꢁ 10ꢀ3 M solutions in dry CH2Cl2,
0.1 M nBu4NPF6) at a scan rate of 100 mV/s. The working electrode was a 3 mm Pt
disk with a Pt wire as auxiliary electrode. A silver wire was used as a pseudo-
reference electrode with ferrocene added as an internal standard with all potentials
referenced to ferrocene (E1/2 = 0.00 V).
1H NMR (CDCl3): 4.26 s (5H, C5H5), 4.32 m (2H, C5H4), 4.91 m
(2H, C5H4), 7.36 d (J = 7.5 Hz, 1H, phenyl), 7.45 t (J = 7.5 Hz, 2H,
phenyl), 7.88 d (J = 7.5 Hz, 2H, phenyl), 7.96 s (1H, CH).
Anal. Calc. for C18H15N3Fe: C, 65.68; H, 4.59; N, 12.77. Found: C,
65.74; H, 4.25, N, 12.28%.
and its dark color prevents the observation of an accurate decom-
position temperature.
Anal. Calc. for C27H17Cl2FeN5O2: C, 56.87; H, 3.00; N, 12.28.
Found: C, 56.02; H, 3.08; N, 11.99%.
3. Results
2.7. Reaction of 3-ferrocenylpyrrole (1) with DDQ
3.1. Synthesis of 1,4-triazoles 2–5
An orange solution of 90 mg (0.40 mmol) DDQ in 12 mL THF
was added dropwise to an orange solution of 100 mg (0.40 mmol)
1 in 30 mL THF at ambient temperature where 1, as prepared in
Ref. [5], had the expected elemental a Analysis:, calculated, C,
66.96; H, 5.22; N, 5.58. Found: C, 67.225; H, 5.14; N, 5.63%. The
resulting deep purple solution was evaporated to dryness, yielding
170 mg (90% yield) [1Å+][DDQÅꢀ] as a purple solid. The compound
does not melt and its dark color prevents the observation of an
accurate decomposition temperature. No visible color changes
were observed upon exposure of [1Å+][DDQÅꢀ] to air. The effects
of prolonged exposure to atmospheric oxygen and water on the
Mössbauer spectra of the compound were not investigated.
Anal. Calc. for C22H12Cl2FeN3O2: C, 55.27; H, 2.74; N, 8.79.
Found: C, 55.68; H, 3.31; N, 8.11%.
Copper catalyzed [3+2] cycloaddition of azides to alkynes [17]
provides a versatile route for the synthesis of 1,2,3-triazoles. Thus
reaction between ethynylferrocene and a series of benzyl azides
yields 1,4-triazoles (2–5) as shown in Eq. (1). Compounds 2–5
N3
[Cu]
R
Fe
+
tBuOH/H2O
ð1Þ
In
a
separate experiment, an orange solution of 180 mg
N
Fe
R
(0.80 mmol) DDQ in 10 mL THF was added dropwise to an orange
solution of 100 mg (0.40 mmol) 1 in 20 mL THF at ambient temper-
ature. The purple solution was stirred for 66 h, precipitating a dark
solid. The Mössbauer spectrum was essentially identical to that of
the preceding 1:1 preparation save that there was no evidence of
FeIII impurity (see arrow in Fig. 2 for the 1:1 preparation).
N
N
2
3
4
5
R = H
R = p-CH3
R = m-Br
R = p-Br
are isolated as orange, air and thermally stable solids soluble in
dichloromethane, THF, acetonitrile and acetone, sparingly soluble
in aromatic hydrocarbons with minimal solubility in saturated
hydrocarbon solvents. The yields obtained from reactions in aque-
ous tBuOH (50% v/v) at ambient temperature and at 60 °C are mod-
erate, falling in a range from 29% to 36%. Little effort was spent on
optimizing yields but yields seem to be somewhat independent of
temperature. A high yield synthesis of 2 by Eq. (1) in ionic liquids
was reported [18] however the spectra and other characterization
data were not provided. The yields from 1,3-dipolar addition of
ethynylferrocene to surface bound azidothiols appear to be
quantitative as well. During the course of our work, additional
ferrocenyl-1,2,3-triazoles were reported, including 1-p-CH3-
OC6H4CH2-4-ferrocenyl-1,2,3-triazole in 68% yield [19]. Finally,
2.8. Reaction of 1-benzyl-4-ferrocenyl-1,2,3-triazole (2) with DDQ
An orange solution of 34 mg (0.15 mmol) DDQ in 20 mL CH2Cl2
was added dropwise to an orange solution of 50 mg (0.15 mmol) 2
in 10 mL CH2Cl2 at ambient temperature over one hour. A dark pre-
cipitate forms rapidly upon mixing. After an additional 30 min of
reaction, the nearly colorless supernatant was decanted and the
dark product dried under vacuum, yielding 56 mg (67% yield)
[2Å+][DDQÅꢀ] as a bright purple solid. No visible color changes were
observed upon exposure of [2Å+][DDQÅꢀ] to air for periods <15 min.
The effect of prolonged exposure to atmospheric oxygen and water
were not investigated. Once again, the compound does not melt

M. Verschoor-Kirss et al. / Journal of Organometallic Chemistry 694 (2009) 3262–3269
3265
Fig. 2. 57Fe Mössbauer spectrum of 1 at 293 K, top left, [1Å+][DDQÅꢀ] at 293, 77.5, 44.5 and 4.2 K. The arrow in the 293 K spectrum (bottom left) indicates the presence of a FeIII
impurity in the sample. Spectra at 293 and 4.5 K were obtained on different samples of [1Å+][DDQÅꢀ] than spectra at 77.5 and 44.5 K.
1-pentyl-4-ferrocenyl-1,2,3-triazole was isolated from the copper
catalyzed reaction of ethynyl ferrocene with azidopentane in aque-
ous THF in an unspecified yield [1] suggesting that further modifi-
cations in reaction conditions (solvent, temperature and catalyst
concentration) should increase the yields of 2–5.
five new compounds analyzed correctly for carbon, hydrogen and
nitrogen.
N
3
Ph
In principle, the reaction in Eq. (1) can lead to two isomers of
the triazole: a 1,4- and a 1,5-isomer. The literature on copper cat-
alyzed [3+2] cycloaddition reactions between alkynes and azides
suggests a selectivity for the 1,4-isomer while the 1,5-isomer is ob-
served when Cp*Ru(PPh3)3Cl is used as the catalyst [20]. Compar-
ing the 1H NMR spectra of 1-benzyl-4-phenyl-1,2,3-triazole with
1-benzyl-5-phenyl-1,2,3-triazole in benzene, the chemical shift of
the methylene proton of the former is ꢃ0.4 ppm upfield of the cor-
responding resonance for the 1,5-isomer. A similar difference is ob-
served when comparing the CH2 resonance in 2 (d 4.85) with the
product of the Cp*Ru(PPh3)2Cl catalyzed reaction between ethynyl-
ferrocene and benzyl azide (d 5.27) suggesting that 2–5 are indeed
the 1,4-isomers of 1,2,3-triazoles in Eq. (1).
Fe
[Cu]/tBuOH/H2O
ð2Þ
N
Fe
N
6
N
3.2. Electrochemistry of 2–6
One additional compound, 1-ferrocenyl-4-phenyl-1,2,3-triazole
(6), was prepared by reaction of azidoferrocene with phenylacety-
lene as shown in Eq. (2). Compound 6 is also an air-stable, diamag-
netic, orange compound that differs from 2 to 5 in having the
ferrocene bonded to one of the nitrogens in the triazole ring. All
The anodic electrochemistry of 2–6 was investigated by cyclic
voltammetry in CH2Cl2 solution with [nBu4N][PF6] as the support-
ing electrolyte. The results are summarized in Table 1 and a repre-
sentative cyclic voltammetric scan (CV) for 2 shown in Fig. 1.

3266
M. Verschoor-Kirss et al. / Journal of Organometallic Chemistry 694 (2009) 3262–3269
Table 1
ature range 293–4.2 K. These spectra, shown in Fig. 2, are virtually
indistinguishable from that of the neutral 1 and are fully consistent
with oxidation at the pyrrole ring as opposed to the iron(II) centers.
The isomer shift and the quadrupole splittings (Table 2 and Fig. 2)
are characteristic of low-spin FeII in ferrocenes i.e. d (with respect
to iron metal) ꢄ0.4–0.6 mm/s, d = 2.20–2.40 mm/s) [24].
Electrochemical data for 3-ferrocenylpyrrole (1) and ferrocenyltriazoles 2–6.a
0=þ
Compound
E
D
Ec
iap=icp
1=2
1b
2
3
4
5
6
ꢀ150
37
31
32
22
160
60
68
68
64
59
76
1.04
0.93
0.93
0.97
1.04
1.01
1.00
Given the similarities in the cyclic voltammetry of 2–6, only
compound 2 was selected for chemical oxidation with DDQ
[25a]. Reaction between 2 (Fig. 3, top) and one equivalent of
193
0
Ferrocene
DDQ also yields
a
dark purple solid, [2+Å][DDQÅꢀ]. The 57Fe
a
Experimental conditions: 1 mM solution in CH2Cl2 containing 0.1 M
[nBu4N][PF6] as the supporting electrolyte. Potentials in mV relative to FcH/FcH+ at a
scan rate 100 mV/s.
Mössbauer spectra (Fig. 3, middle and bottom) of [2+Å][DDQÅꢀ] at
both ambient (293 K) and (77.5 K) are characteristic of the low-
spin FeIII of typical ferrocenium ions, i.e. the isomer shift is little if
at all changed on oxidation from low-spin FeII to low-spin FeIII
while the electric field gradient tensor largely collapses resulting
in a decrease in quadrupole splittings from generally >2 mm/s for
the low-spin FeII to <ꢄ0.5 mm/s for the low-spin FeIII ferrocenium
centers. The latter feature has been studied in the work of Collins
and Travis [25b] and molecular orbital studies in references
therein.
b
In CH3CN with LiClO4 as supporting electrolyte relative to FcH/FcH+. See Ref. [5].
c
D
E = Epa ꢀ Epc
.
Reversible waves (DE = 60–68 mV, ipc/ipa = 0.93–1.04) are observed
for 2–5 at slightly more positive potentials relative to ferrocene
(22–37 mV versus FcH/FcH+). These values are similar to those re-
ported for 1-pentyl-4-ferrocenyl-1,2,3-triazole (E1/2 = 8 mV in
CH2Cl2,
DE = 60 mV, ipc/ipa = 1.0) [1]. Ferrocenyl substituted pyri-
The zero field cooled isothermal magnetization of [2+Å][DDQÅꢀ
]
dines and pyrimidines [4] are oxidized at potentials ranging from
620 to 790 mV in THF versus SCE. When referenced to SCE, the oxi-
dation potentials of 2–5 ranged from 422 to 437 mV, making them
slightly easier to oxidize than ferrocenylpyridines and pyrimidines
[21]. Relatively small electronic effects of the substituents on the
benzyl group on E1/2 are observed. An electron donating methyl
group in 3 only shifts the potential closer to that of ferrocene by
6 mV. Curiously, the electron withdrawing bromide substituents
in 4 and 5 actually decreases the oxidation potential between 5
(4) and 15 mV (5). The triazoles, however, are more difficult to oxi-
dize than ferrocenylpyrrole (1) by 170–200 mV. Even after
accounting for the differences in solvent and supporting electrolyte
for the electrochemical measurements, the oxidation potential for
1 appears to be significantly lower than for 2–5 [5].
at 2 K is shown in Fig. 4 wherein it is apparent that M at 5 T
(ꢄ9300 emu/mol) is significantly below the saturation value
(11, 165 emu/mol) expected for spin only behavior of an S = 1 per
mole species (i.e. two unpaired electrons). This is consistent with
measurable antiferromagnetic interactions between the compo-
nents of [1+Å][DDQÅꢀ] (presumably the pyrrol radical cation and
the DDQÅꢀ radical anion and for [2+Å][DDQÅꢀ] between the ferroce-
nium center and the DDQÅꢀ radical anion.
The oxidation of N-bonded ferrocene in 6 is also reversible but
occurs at much more positive potentials than 2–5. Comparisons of
the electrochemistry of 6 with other N-substituted ferrocenyl het-
erocycles is hampered by differences in solvent and supporting
electrolyte both of which can effect the oxidation potential. Never-
theless, the oxidation potential of 6 (193 mV versus FcH) is about
ꢃ150 mV more positive than the potential for N-ferrocenylpyrrole
(50 mV) [22]. The E1/2 for 6 is also considerably more positive than
E1/2 for 1,10-diaminoferrocenes [23]. For example, E1/2 for 1,10-
bis(diphenylamino)ferrocene in acetonitrile is ꢀ400 mV relative
to FcH/FcH+, close to 600 mV easier to oxidize than 6.
3.3. Chemical oxidation of 1 and 2
The electrochemical potentials for 1–6 suggest that only a
strong oxidant like dichlorodicyanoquinone (DDQ) will oxidize
these compounds. Reaction of 1 with one equivalent of DDQ in
THF produces a purple salt, [1+Å][DDQÅꢀ], in high yield. 57Fe
Mössbauer spectra of [1+Å][DDQÅꢀ] were recorded over the temper-
Table 2
57Fe Mössbauer spectral parametersa for 1, [1+Å][DDQÅꢀ] and [2+Å][DDQÅꢀ].
Compound
Temperature (K)
Isomer shiftb (d)
Quadrupole splitting (D)
1
293
293
77.5
44.5
4.5
77.5
293
77.5
0.46
0.47
0.54
0.59
0.57
0.57
0.48
0.59
2.38
2.35
2.36
2.34
2.34
2.35
0.25
0.24
[1+Å][DDQÅꢀ
]
Fig. 3. 57Fe Mössbauer spectra of 2 at 77.5 K (top). The asymmetry of the spectrum
is likely primarily the result of texture in the context of the direction of the principal
2
[2+Å][DDQÅꢀ
]
component of the electric field gradient tensor relative to the direction of E (the
c
high nitrogen content of 2 suggests that grinding of the crystals to minimize texture
a
may be hazardous and was not done), [2Å+][DDQÅꢀ
[2Å+][DDQÅꢀ] at 77.5 K (bottom).
] at 293 K (middle), and
mm/s.
b
Relative natural iron foil (6.4 lm).

M. Verschoor-Kirss et al. / Journal of Organometallic Chemistry 694 (2009) 3262–3269
3267
Fig. 4. DC magnetization of [2+Å][DDQÅꢀ] vs. H at 2 K.
Magnetization measurements via SQUID magnetometry lead to
the effective moment temperature dependencies (2–300 K) for
[1+Å][DDQÅꢀ] and [2+Å][DDQÅꢀ] shown in Figs. 5 and 6, respectively,
and that are qualitatively remarkably similar. At ambient temper-
ature the average moments are ꢄ2.1–2.2 Bohr magnetons/spin car-
rying unit superficially approximating ferrocene donor–acceptor
complex salts containing two unpaired electrons that do not inter-
Fig. 6. Effective moment per mol of [2Å+][DDQÅꢀ] vs. T at 0.05 T (top) and the
corresponding molar susceptibility (bottom).
act strongly [26]. Recall that the average spin-only l/spin unit for
4. Discussion
non-interacting spins (neglecting orbital contributions) is ꢄ1.73
P
BM i.e. <
l
> = [( (i) li
)
2 1/2]/(N)1/2 where N is the number of inde-
The anodic electrochemistry of 1–6 reveals similar cyclic voltam-
mograms but distinctly different oxidation potentials. The electro-
chemistry of 3-ferrocenylpyrrole (1) reveals a single oxidation at
pendent metals or radicals or metals and radicals. However, with
decreasing temperature magnetic exchange interaction generally
becomes stronger relative to kBT and for the present complexes
leads to a gradual decrease of the effective moments to limiting
Ea = ꢀ150 mV versus SCE in CH3CN (LiClO4 as the supporting elec-
p
values of ꢄ0.15 BM for [1+Å][DDQÅꢀ] and 1.5 BM for [2+Å][DDQÅꢀ
]
trolyte), a potential more negative than ferrocene itself [5]. The sep-
aration between the anodic and cathodic peaks in the cyclic
at 2 K. This gradual decrease is clearly attests to primarily antifer-
romagnetic interactions as opposed to single ion zero field splitting
effects. In event, the larger overall decrease of leff for the former
suggests stronger antiferromagnetic exchange in this material than
the latter. However, any further comment on the detailed magne-
tism of these systems is clearly speculative in the absence of defin-
itive structure determinations save to say that there is no evidence
of cooperative long-range order at the lowest temperatures that
we can conveniently reach. See for example the low field suscepti-
bility of [2+Å][DDQÅꢀ] (bottom of Fig. 6) in this context.
voltammogram,
D
Ep = Ea ꢀ Ec = 160 mV and Ea ꢀ Ep=2 = 90 mV sug-
p
p
p
gests that the oxidation is not entirely reversible. The ratio
iap=icp = 1.04, however, suggests a reversible process. These data were
interpreted using the scheme in Eq. (3) [5] where initial oxidation
occurs at the pyrrole ligand and is followed by intramolecular elec-
tron transfer from iron yielding a ferricenium ion. However, the
Mössbauer spectroscopy study of the present investigation of
[1Å+][DDQÅꢀ] clearly demonstrates that in the solid state, the product
of chemical oxidation of 1 contains low-spin iron(II) not iron(III).
H
N
H
N
H
N
• +
- e-
Fe
1
Fe
Fe
a
b
1•+
ð3Þ
H
N
H
N
•
Fe+
Fe
Fig. 5. Effective moment per mol of [1Å+][DDQÅꢀ] vs. T at 1T.
c
d

3268
M. Verschoor-Kirss et al. / Journal of Organometallic Chemistry 694 (2009) 3262–3269
The susceptibility data (both leff versus T and M versus H) con-
5. Conclusions
firm that electron transfer has indeed occurred and that the prod-
uct of the reaction between 1 and DDQ is not a (diamagnetic)
neutral charge transfer complex similar to the mixed stack tetracy-
anoethylene complex of ferrocene [27], or the iodine adduct of
FcCH@NC6H4@NCHFc, [FcCH@NC6H4@NCHFc][I4.5][28]. The tem-
perature dependence of the Mössbauer spectra of [FcCH@NC6H4@
NCHFc][I4.5], a mixed iron(II) and iron(III) complex at ambient T)
reveals [28] an increase in the fraction of low-spin FeII relative to
FeIII at lower temperatures concurrent with a decrease in the effec-
tive moment. These results are consistent with a unique retro-elec-
tron transfer process from the poly-iodide anion to FeIII and that is
fully reversible. No such behavior is observed for either of
[1+Å][DDQÅꢀ] or [2+Å][DDQÅꢀ].
The electrochemistry of 2 alone provides few clues to the nature
of the oxidized product. Like the electrochemistry of 1, the anodic
electrochemistry of 2 suggests a single-electron, reversible process
in solution. The oxidation potential of 2 is only slightly more posi-
tive than that of ferrocene but 2 is, apparently, considerably more
difficult to oxidize than 1. Again, electron transfer from an electron
donor, 2, to the electron acceptor (DDQ) is confirmed by the mag-
netic susceptibility measurements. In the solid state, dramatically
different behavior is observed in the Mössbauer spectra of
[2Å+][DDQÅꢀ]. The chemical oxidation of 2 clearly favors a product
where iron has been oxidized from FeII to FeIII. Compound
[2Å+][DDQÅꢀ] contains essentially all low-spin FeIII at 293 and
77.5 K. This is the exact opposite of the observations for
[1Å+][DDQÅꢀ].
While the anodic electrochemistry of 1 and 2 appears to be con-
sistent with oxidation of ferrocene FeII to ferrocenium FeIII, the
products of chemical oxidation with the single-electron oxidant
DDQ lead to surprisingly different results. The product of chemical
oxidation of 1 suggests loss of an electron from the ligand, delocal-
ization of the odd electron into the p-system of the pyrrole ligand
or initial oxidation at iron followed by intramolecular electron
transfer from the ligand to iron. The structure of [1]Å+ is mostly
likely to be either a, b, c, or a hybrid thereof, in Eq. (3) rather than
d. The strongest evidence for this conclusion comes from the
Mössbauer spectra, a technique that has proven to be accurate in
distinguishing low-spin FeII from FeIII in the solid state. Low-spin
FeIII, however, is clearly seen in the Mössbauer spectrum of the
product of chemical oxidation of 2. The contrasting behavior of
two seemingly similar ferrocene-substituted nitrogen heterocycles
suggests that further studies of the chemical oxidation of a wider
range of compounds are warranted. Finally, the present results
may in part ultimately reflect the fundamental differences be-
tween solution species as obtained via classical electrochemical
oxidations versus direct chemical reactions accompanied in some
cases by rapid crystallization of the solid and for which imponder-
able lattice effects can be important in the context of the species
that is actually stabilized.
References
The limited number of chemically oxidized ferrocenyl-substi-
tuted nitrogen heterocycles studied to date makes it difficult to
[1] C. Ornelas, J.R. Aranzaes, E. Cloutet, S. Alves, D. Astruc, Angew. Chem., Int. Ed.
Engl. 46 (2007) 872.
[2] A. Devadoss, C.E.D. Chidsey, J. Am. Chem. Soc. 129 (2007) 5370.
[3] (a) N.K. Devaraj, R.A. Decreau, W. Ebina, J.P. Collman, C.E.D. Chidsey, J. Phys.
Chem. B 110 (2006) 15955;
fully explain the origin of the differences between [1Å+][DDQÅꢀ
]
and [2Å+][DDQÅꢀ]. The observation that the oxidation potential for
1 is lower than pyrrole itself suggests that oxidation occurs initially
at iron followed by intramolecular electron transfer as proposed in
Eq. (3) [5]. A similar explanation has been proposed for the anodic
electrochemistry of nitrogen containing ferrocenophanes (Eq. (4))
and aminomethylferrocenes (Eq. (5)) where equilibria between
nitrogen-centered cation radicals and ferrocenium radical cations
are established [29].
(b) J.P. Collman, N.K. Devaraj, T.P.A. Eberspacher, C.E.D. Chidsey, Langmuir 22
(2006) 2457;
(c) N.K. Devaraj, P.H. Dinolfo, C.E.D. Chidsey, J.P. Collman, J. Am. Chem. Soc.
128 (2006) 1794;
(d) J.P. Collman, N.K. Devaraj, C.E.D. Chidsey, Langmuir 20 (2004) 1051.
[4] D. Paolucci, M. Marcaccio, C. Bruno, D. Braga, M. Polito, F. Paolucci,
Organometallics 24 (2005) 1198.
[5] T.L. Rose, A.B. Kon, Inorg. Chem. 32 (1993) 781.
[6] K. Sutoh, S. Sasaki, M. Yoshifuji, Inorg. Chem. 45 (2006) 992.
[7] (a) G. Pilloni, B. Longato, B. Corain, J. Organomet. Chem. 420 (1991) 57;
(b) C. Nataro, A.N. Campbell, M.A. Ferguson, C.D. Incarvito, A.L. Rheingold, J.
Organomet. Chem. 673 (2003) 47;
- e-
Fe+
Fe
N
R
N
R
(c) J.H.L. Ong, C. Nataro, J.A. Golen, A.L. Rheingold, Organometallics 22 (2003)
5027;
(d) B.L. Ghent, L.A. Sites, A.L. Rheingold, C. Nataro, Organometallics 24 (2005)
4788.
[8] F. Barriere, R.U. Kirss, W.E. Geiger, Organometallics 24 (2005) 48.
[9] B.D. Swartz, C. Nataro, Organometallics 24 (2005) 2447.
[10] F.N. Blanco, L.E. Hagopian, W.R. McNamara, J.A. Golen, A.L. Rheingold, C.
Nataro, Organometallics 25 (2006) 4292.
R = H, NH2, OH
ð4Þ
•+
Fe
N
R
[11] D.A. Durfey, R.U. Kirss, C. Frommen, W.M. Reiff, Inorg. Chim. Acta 357 (2004)
311.
[12] G. Doisneau, G. Balavoine, T. Fillebeen-Khan, J. Organometal. Chem. 425 (1992)
113.
[13] S.G. Alvarez, M.T. Alvarez, Synthesis (1997) 413.
[14] C. Cheng, W.M. Reiff, Inorg. Chem. 16 (1977) 2097.
[15] M. Bancroft, A.G. Maddock, W.K. Ong, R.H. Prince, A.J. Stone, J. Chem. Soc.
(1967) 1966.
H
N
H
N
Ar
Ar
- e-
Fe•+
Fe
[16] A.N. Nesmeyanov, V.N. Drozd, V.A. Sazanova Doklady, Akad. Nauk SSSR 150
(1963) 321.
[17] (a) J.E. Moses, A.D. Moorhouse, Chem. Soc. Rev. 36 (2007) 1249;
(b) V.D. Bock, H. Hiemstra, J.H. van Maarseveen, Eur. J. Org. Chem. (2006) 51;
(c) H.C. Kolb, M.G. Finn, K.B. Sharpless, Angew. Chem., Int. Ed. Engl. 40 (2001)
2004.
[18] Y.-B. Zhao, Z.-Y. Yan, Y.-M. Liang, Tetrahedron Lett. 47 (2006) 1545.
[19] S. Badèche, J-C. Daran, J. Ruiz, D. Astruc, Inorg. Chem. 47 (2008) 4903.
[20] (a) L. Zhang, X. Chen, P. Xue, H.H.Y. Sun, I.D. Williams, K.B. Sharpless, V.V.
Fokin, G. Jia, J. Am. Chem. Soc. 127 (2005) 15998;
ð5Þ
Ar = C6H4Me
H
N
Ar
Fe
One interpretation of the apparent oxidation of the pyrrole li-
gand in [1Å+][DDQÅꢀ] and the observation of Fe3+ in [2Å+][DDQÅꢀ] is
in terms of Eqs. (4) and (5). For [1Å+][DDQÅꢀ], the equilibrium
may favor a ligand centered radical cation while in [2Å+][DDQÅꢀ],
the equilibrium likely lies on the side of the ferricenium ion.
(b) M.M. Majireck, S.M. Weinreb, J. Org. Chem. 71 (2006) 8680.
[21] The oxidation potential for ferrocene in CH2Cl2 versus SCE is 0.40 V.
[22] A. Merz, R. Baumann, A. Haimerl, Makromol. Chem. Macromol. Symp. 8 (1987)
61.
[23] N.F. Blank, D.S. Glueck, L.N. Zakharov, A.L. Rheingold, M.D. Saybolt, B.L. Ghent,
C. Nataro, Organometallics 24 (2005) 5184.

M. Verschoor-Kirss et al. / Journal of Organometallic Chemistry 694 (2009) 3262–3269
3269
[24] N.N. Greenwood, T.C. Gibbs, Mössbauer Spectroscopy, Chapman and Hall, Ltd.,
London, UK, 1971.
[28] S.K. Pal, K. Alagesan, A.G. Samuelson, J. Pebler, J. Organometal. Chem. 575
(1999) 108.
[25] (a) Other things being equal, brominated versions of compounds 4–5 are not
good candidates for high S/N Mössbauer spectroscopy as bromine has a large
[29] (a) T. Sakano, M. Horie, K. Osakada, H. Nakao, Eur. J. Inorg. Chem. (2005)
644;
non-resonant absorption coefficient for the 14.4 kev gamma ray of Fe57
(b) R.L. Collins, J. Travis, J. Chem. Phys. 42 (1964) 1072.
;
(b) I. Yamaguchi, T. Sakano, H. Ishii, K. Osakada, T. Yamamoto, J. Organomet.
Chem. 584 (1999) 213;
[26] E. Adman, M. Rosenblum, S. Sullivan, T.N. Margulis, J. Am. Chem. Soc. 89 (1967)
4540.
(c) T. Sakano, H. Ishii, I. Yamaguchi, K. Osakada, T. Yamamoto, Inorg. Chim.
Acta 296 (1999) 176;
[27] R.L. Brandon, J.H. Osiecki, A. Ottenberg, J. Org. Chem. 31 (1966) 1214.
(d) T. Sakano, K. Osakada, Macromol. Chem. Phys. 202 (2001) 1829.