A search for blues brothers: X-ray crystallographic/spectroscopic characterization of the tetraarylbenzidine cation radical as a product of aging of solid magic blue

Article abstract of DOI:10.1039/c6ob00140h

Magic blue (MB^+? SbCl₆^- salt), i.e. tris-4-bromophenylamminium cation radical, is a routinely employed one-electron oxidant that slowly decomposes in the solid state upon storage to form so called 'blues brothers', which often complicate the quantitative analyses of the oxidation processes. Herein, we disclose the identity of the main 'blues brother' as the cation radical and dication of tetrakis-(4-bromophenyl)benzidine (TAB) by a combined DFT and experimental approach, including isolation of TAB^+? SbCl₆^- and its X-ray crystallography characterization. The formation of TAB in aged magic blue samples occurs by a Scholl-type coupling of a pair of MB followed by a loss of molecular bromine. The recognition of this fact led us to the rational design and synthesis of tris(2-bromo-4-tert-butylphenyl)amine, referred to as 'blues cousin' (BC: E_ox1 = 0.78 V vs. Fc/Fc⁺, λ_max(BC^+?) = 805 nm, ε_max = 9930 cm^-1 M^-1), whose oxidative dimerization is significantly hampered by positioning the sterically demanding tert-butyl groups at the para-positions of the aryl rings. A ready two-step synthesis of BC from triphenylamine and the high stability of its cation radical (BC^+?) promise that BC will serve as a ready replacement for MB and an oxidant of choice for mechanistic investigations of one-electron transfer processes in organic, inorganic, and organometallic transformations.

Full text of DOI:10.1039/c6ob00140h

Organic &
Biomolecular Chemistry
View Article Online
View Journal | View Issue
PAPER
A search for blues brothers: X-ray crystallographic/
spectroscopic characterization of the tetraaryl-
benzidine cation radical as a product of aging of
solid magic blue†
Cite this: Org. Biomol. Chem., 2016,
14, 2961
Marat R. Talipov, Mohammad M. Hossain, Anitha Boddeda, Khushabu Thakur and
Rajendra Rathore*
Magic blue (MB+• SbCl6 salt), i.e. tris-4-bromophenylamminium cation radical, is a routinely employed
−
one-electron oxidant that slowly decomposes in the solid state upon storage to form so called ‘blues
brothers’, which often complicate the quantitative analyses of the oxidation processes. Herein, we dis-
close the identity of the main ‘blues brother’ as the cation radical and dication of tetrakis-(4-bromophe-
nyl)benzidine (TAB) by a combined DFT and experimental approach, including isolation of TAB+• SbCl6
−
and its X-ray crystallography characterization. The formation of TAB in aged magic blue samples occurs
by a Scholl-type coupling of a pair of MB followed by a loss of molecular bromine. The recognition of this
fact led us to the rational design and synthesis of tris(2-bromo-4-tert-butylphenyl)amine, referred to as
‘blues cousin’ (BC: Eox1 = 0.78 V vs. Fc/Fc+, λmax(BC+•) = 805 nm, εmax = 9930 cm−1 M−1), whose oxidative
dimerization is significantly hampered by positioning the sterically demanding tert-butyl groups at the
para-positions of the aryl rings. A ready two-step synthesis of BC from triphenylamine and the high stabi-
lity of its cation radical (BC+•) promise that BC will serve as a ready replacement for MB and an oxidant of
choice for mechanistic investigations of one-electron transfer processes in organic, inorganic, and
organometallic transformations.
Received 16th January 2016,
Accepted 3rd February 2016
DOI: 10.1039/c6ob00140h
www.rsc.org/obc
MB+• undergoes slow decomposition to produce impurities
which have been dubbed ‘blues brothers’.2,13 During the
Introduction
−
Magic blue (MB+• SbCl6 salt), i.e. tris-4-bromophenylamm- course of our studies of organic cation radicals and dications,
inium cation radical,1 is widely utilized as an aromatic oxidant we observed that the presence of additional bands in the spec-
for the 1e− oxidation of organic, inorganic, and organometallic trum of MB+• not only interfered with spectroscopic characteri-
donors.2–10 The popularity of the magic blue cation radical as zation of oxidized species, but they also prevented accurate
an oxidant in part arises due to its perceived high stability, quantification of redox processes.14
commercial availability, and reasonable oxidizing power (Ered
=
Herein, we describe identification of the major impurity
0.70 V vs. Fc/Fc+). With an intense and uncluttered visible responsible for the presence of additional absorption bands in
−
absorption (λmax = 728 nm, εmax = 28 200 cm−1 M−1) magic the aged samples of MB+• SbCl6 salt as a tetraarylbenzidine
blue has also found extensive application for (quantitative) cation radical/dication with the aid of electronic absorption
spectroscopic characterization of cation radicals and dications spectroscopy, X-ray crystallography, and detailed compu-
of organic electron donors and oxidized inorganic metal tational studies. Moreover, the fact that the ‘blues brother’ in
complexes and organometallic species.11,12 Although MB+• the aged MB+• sample is produced by a dimerization reaction
−
SbCl6 salt is reasonably stable, it has been long known that led us to the rational synthesis of a new triarylamine derivative
which
prevents
the
dimerization
reaction.
This
triarylamine derivative forms a stable cation radical with
similar spectral properties and somewhat improved oxidizing
power, and it is referred to as blues cousin, which should serve
as a ready replacement for MB as an oxidant of choice for
mechanistic investigations of one-electron transfer processes
Department of Chemistry, Marquette University, P.O. Box 1881, Milwaukee,
WI 53201-1881, USA. E-mail: rajendra.rathore@marquette.edu
†Electronic supplementary information (ESI) available: Details of synthesis and
characterization data for various compounds including cyclic voltammetry, gen-
eration of cation radicals, X-ray crystallography, and computational details.
CCDC 1445500, 1445501 and 1445573. For ESI and crystallographic data in CIF
or other electronic format see DOI: 10.1039/c6ob00140h
in
a variety of organic, inorganic, and organometallic
transformations.
This journal is © The Royal Society of Chemistry 2016
Org. Biomol. Chem., 2016, 14, 2961–2968 | 2961

View Article Online
Paper
Organic & Biomolecular Chemistry
molecular oxidative cyclization of MB to carbazole derivatives
(see Chart 1), we first carried out (TD-)DFT calculations of all
Results and discussion
The initial attempt to identify the species responsible for compounds in Chart 1 to obtain both redox potentials and
additional absorption bands in aged samples of MB+• was electronic absorption spectra of their cation radicals. For this
carried out by its reduction using zinc dust or ferrocene, and purpose, we utilized the B1LYP-40 functional [i.e. B1LYP-40/6-
1
1
the H NMR analysis of the resulting neutral residue. The H 31G(d)+PCM(CH2Cl2)]19,20 because our recent careful bench-
NMR analysis suggested that the sample largely contained marking studies on a number of polyaromatic systems17,20–22
MB0 together with a significant amount of new species and showed that this functional allows an accurate description of
multiple other compounds in minor quantities (see Fig. S1 in the spin/charge (hole) distribution in their cation radicals,
the ESI†). The new species was tentatively identified by MALDI and, in turn, the prediction of the corresponding redox/optical
mass spectrometry to be a dimer of MB, which has lost a mole- properties (Eox1, νD0→D1) with high fidelity to experimental
cule of bromine, and was hypothesized to be a tetrakis-(4-bro- data. The excellent performance of the B1LYP-40 functional
mophenyl)benzidine (Fig. S3 in the ESI†). Unfortunately, full owes to the fact that 40% contribution of the exact
identification of this dimer and other minor species could not (i.e. Hartree–Fock) exchange term alleviates the self-interaction
be achieved because of the failure of repeated attempts of error23,24 that causes artificial charge delocalization.
chromatographic separation.
The calculated equilibrium geometries of the neutral and
In order to establish that additional absorption bands in cation radical forms of all compounds listed in Chart 1
the aged samples of MB+• (Fig. 1A) arise from its decompo- produced their free energies of oxidation ΔGox1 by a simple
sition, we obtained an authentic spectrum of MB+• via redox subtraction of the free energies of cation radicals and the
titration of neutral MB using a bicyclo[2.2.1]heptane-annulated corresponding neutral molecules. A plot of the computed oxi-
hydroquinone ether cation radical THEO+• SbCl6 (Ered1
=
dation energies (ΔGox1) against electrochemical oxidation
−
0.67 V vs. Fc/Fc+, λmax = 518 nm, εmax = 7300 cm−1 M−1 15–17 as potentials (Eox1) of compounds MB, 1, 2, and 3, available from
)
an oxidant. The spectrum of MB+• obtained via redox titration the literature25 (see blue diamonds and correlation line in
according to Fig. 1 was found to be identical to the spectrum Fig. 2), showed an excellent linear correlation that instilled
of the freshly prepared sample of MB+• using MB0 and SbCl5 confidence in our use of the B1LYP-40 functional and allowed
or NO+ SbCl6− as oxidants (Fig. 1B).
A subtraction of the authentic spectrum of MB+• with that
of the aged sample (∼14 months old, Aldrich Chemical Co.)
produced a spectrum containing bands at 480, 805, and
1550 nm (Fig. 1C), and the intensities of these bands increase
with aging of the sample. The fact that these absorption bands
−
are not attributable to pure MB+• SbCl6 together with the
presence of a low-energy transition at 1550 nm, suggested that
the new absorption bands most likely arise from a cation
radical and/or dication of a tetraarylbenzidine derivative.18
Before undertaking the synthesis of authentic samples of
tetrakis(4-bromophenyl)benzidine (TAB) and other products
arising from bromination of either MB or TAB or by intra- Chart 1 Possible candidates for ‘blues brothers’.
−
Fig. 1 (A) Absorption spectrum of the aged MB+• SbCl6 salt at 22 °C in CH2Cl2; (B) spectral changes observed upon the redox titration of 42 μM
THEO+• SbCl6 with an incremental addition of 3.2 mM MB0 in CH2Cl2 at 22 °C; (C) spectrum obtained by the subtraction of aged MB+• (shown in
−
panel A) and freshly prepared MB+• (shown in panel B).
2962 | Org. Biomol. Chem., 2016, 14, 2961–2968
This journal is © The Royal Society of Chemistry 2016

View Article Online
Organic & Biomolecular Chemistry
Paper
6–9 as well as carbazole derivatives 4 and 5 using standard lit-
erature procedures,26–36 and all synthetic details are included
1
in the ESI.† All compounds were characterized by H/13C NMR
spectroscopy as well as by MALDI spectrometry. These com-
pounds were then subjected to the cyclic voltammetric analysis
at 22 °C in CH2Cl2 containing 0.2 M tetra-n-butylammonium
hexafluorophosphate at a platinum electrode. The reversible
voltammograms of TAB and 6–9 showed two oxidation waves
(Fig. S4 in the ESI†) corresponding to the formation of the
cation radical and dication. The electrochemical oxidation
potentials of various molecules in Chart 1 are compiled in
Table 1 (columns 4 and 5), and they were in excellent agree-
ment with the predicted oxidation potentials using DFT calcu-
lations (Table 1, column 3). Having obtained the oxidation
potentials of all compounds in Chart 1, they were included in
Fig. 2, and all the new points (shown as red squares) fell on
the original correlation line.
Fig. 2 A plot of experimental Eox1 of compounds in Chart 1 against
computed ΔGox1. Note that the correlation line was generated using
only MB and 1–3 (shown by blue diamonds) for which the electrochemi-
cal oxidation potentials were available from the literature. The red
squares represent the remaining compounds in Chart 1, while the green
triangles are for ‘blues cousins’ 10 and BC discussed later in the text.
the prediction of the oxidation potentials of various com-
We next generated the cation radicals/dications of TAB and
pounds in Chart 1 which are yet to be synthesized, see Table 1. 6–9 using a hindered naphthalene (i.e. 1,2,3,4,7,8,9,10-octa-
The additional bands in the MB+• spectrum are expected to hydro-1,1,4,4,7,7,10,10-octamethylnaphtacene) cation radical
arise from the oxidized (i.e. cation radical or dication) forms of (Ered = 0.94 V vs. Fc/Fc+, λmax = 672 nm, εmax = 9300 cm−1
the blues brothers, and thus the oxidation potentials of these M−1 37,38
via redox titration under careful exclusion of moist-
)
molecules should be lower or comparable to that of MB0 (i.e. ure; see, e.g. Fig. S5 in the ESI.† The resulting absorption
0.70 V vs. Fc/Fc+). Accordingly, the predicted values of E′ox1 of spectra of the cation radicals and dications of TAB and 6–9 are
various compounds in Chart 1 (Table 1) allow one to easily compared in Fig. 3. Interestingly, the bands contained in the
rule out the possibility of brominated MB analogues (i.e. 1–3) absorption spectra of the cation radical and dication of TAB
and carbazole derivatives 4 and 5 as blues brothers in the spec- (i.e. TAB+•: λmax = 1490 nm, εmax = 38 100 cm−1 M−1, and λmax
trum of aged MB+•. However, the E′ox1 values of TAB0 and its 489 nm, εmax = 35 000 cm−1 M−1; TAB2+: λmax = 807 nm, εmax
=
=
brominated analogues 6–9 were predicted to have lower or 112 600 cm−1
M
−1) closely matched the additional absorption
comparable oxidation potentials to that of MB0 and thus con- bands in the spectrum of the aged sample of MB+•, i.e.
stitute plausible candidates for the blues brothers. In addition, compare Fig. 3 and 1C.
TD-DFT calculations (Tables 1 and S4 in the ESI†) predict
In order to further confirm the presence of both TAB+• and
the existence of low-energy bands in the spectra of TAB+• and TAB2+ in the spectrum of the aged MB+• sample, we carried out
6+•–9+•, which are absent in the cation radicals of brominated a redox titration by an incremental addition of sub-stoichio-
MB analogues 1–3 and carbazole derivatives 4.25
metric amounts of TAB0 to a solution of freshly prepared MB+•
Based on the predictions from the (TD-)DFT calculations we in CH2Cl2 at 22 °C (Fig. 4). Interestingly, the spectrum
undertook the synthesis of TAB and its brominated derivatives obtained after addition of ∼0.15 equivalents of TAB0 to a
Table 1 Calculated [B1LYP-40/6–31G(d) + PCM(CH2Cl2)] free energies of oxidation (ΔGox1 and E’ox1a) and electrochemical (Eox1 and Eox2 vs. Fc/Fc+)
oxidation potentials of various compounds in Chart 1, as well as the optical properties of their cation radicals (CRs) obtained by TD-DFT calculations
(λD0→D1 and fosc) and by spectroscopic redox titration (λmax(CR) and εmax(CR))
Compound
ΔGox1, eV
E′ox1 a, V
Eox1, V
Eox2, V
λD0→D1, nm
fosc
λmax(CR), nm
εmax(CR), M−1 cm−1
MB
1
2
3
4
5
TAB
6
7
8
5.306
5.472
5.612
5.803
5.682
5.651
5.044
5.191
5.220
5.442
5.366
4.853
5.404
0.69
0.84
0.97
1.14
1.03
1.00
0.46
0.59
0.61
0.81
0.75
0.29
0.78
0.7025,b
0.8225
0.9625
1.1425
1.08025
1.030b
0.436b
0.576b
0.600b
0.768b
0.720b
0.290b
0.780b
—
—
—
—
—
—
604
686
721
0.30
0.35
0.24
0.14
0.03
0.27
1.00
0.73
0.56
0.21
0.30
0.23
0.09
728
28 200
—
—
—
—
75725
80525
88025
82025
854, 2500
1490
1735
1750
2150
2128
681
823
1291
1526
1353
1420
1304
1339
1428
575
—
0.636b
0.760b
0.780b
0.856b
0.823b
—
38 100
24 000
24 500
15 000
15 000
31 500
9930
9
10c
BCc
—
785
805
a E′ox1 = 0.8955ΔGox1 − 4.06 (Fig. 2). b Obtained in this work by cyclic voltammetry (Fig. S4 of the ESI). c See the discussion in the text below.
This journal is © The Royal Society of Chemistry 2016
Org. Biomol. Chem., 2016, 14, 2961–2968 | 2963

View Article Online
Paper
Organic & Biomolecular Chemistry
Fig. 4 (A) Spectral changes observed upon the incremental addition of
−
1.52 mM TAB to a freshly prepared solution of MB+• SbCl6 (20 μM) in
CH2Cl2 at 22 °C. Thick blue line shows the absorption spectrum of the
aged sample of MB+• SbCl6–. (B) Molar fractions of MB+• (red circles),
TAB+• (black circles), TAB2+ (blue circles) plotted against the number of
added equivalents of neutral TAB at each addition. The data points in
panel B were fitted by accounting multiple equilibria amongst various
oxidized and neutral species which showed that the second oxidation
potential of TAB is somewhat higher (Eox2 = 0.71 V) as compared to the
electrochemical potentials (0.64 V) owing to the fact that redox titra-
tions are carried out in the absence of an electrolyte.
concentrations of each species for a given titration point. This
spectral deconvolution procedure allowed accurate determi-
nation of the concentrations of MB+•, TAB+•, and TAB2+ at every
titration point in Fig. 4, and these are plotted as molar frac-
tions against the added equivalents of TAB0 in Fig. 4B. The
molar fraction vs. the added TAB0 plot clearly shows that oxi-
dation of TAB0 by MB+•, up to the equimolar concentration of
MB+• and TAB0, produces both the cation radical and dication
of TAB. Indeed, this analysis confirmed the presence of ∼0.11
equivalents of TAB+• and ∼0.04 equivalents of TAB2+, i.e. it
amounts to the presence of ∼0.15 equivalents of TAB in the
sample of MB+• used in this study.
The presence of TAB in the aged samples of MB+• was
further validated by comparison of the 1H NMR spectrum of
the reduced MB0 with the spectrum of authentic TAB0 (Fig. S1
and S2 in the ESI†). At the same time, the 1H NMR spectral
analysis showed the absence of 6–9 in the aged (reduced)
sample of MB0 (Fig. S2 in the ESI†).
Beyond its role as an impurity in the aged sample of MB+•,
TAB+• represents an interesting example of a mixed-valence
compound, where the DFT calculations predict that the hole
was largely (88%) localized on the benzidine fragment and
spread only slightly onto the p-bromophenyl rings (i.e. 3% per
aryl group), as shown by the spin/charge distribution in Fig. 5
(see also Fig. S11 in the ESI†). Importantly, the 1−e− oxidation-
induced bond length in TAB → TAB+• transformation were in
accordance with the disposition of the bonding/antibonding
lobes of HOMO, i.e. bonds with bonding HOMO lobes
undergo elongations, whereas the bonds with antibonding
lobes undergo contraction, see Table S5 in the ESI.† 12,39–43
However, it is noted that HOMO density distribution in TAB0 is
more delocalized as compared with the spin/charge
distribution in TAB+• due to the oxidation-induced structural
Fig. 3 Absorption spectra of cation radicals (left) and dications (right)
of TAB, 6–9 in CH2Cl2 at 22 °C obtained via spectroscopic redox titra-
tions using THEO+• or NAP+• (see Fig. S5 in the ESI† for the corres-
ponding titration figures).
solution of MB+• produced
a spectrum which closely
resembled the spectrum of the aged MB+• sample (shown as
thick blue line). The relative amounts of TAB+• and TAB2+ at
various titration points in Fig. 4 were evaluated by a deconvolu-
tion of the individual spectra using the clean absorption
spectra of MB+•, TAB+•, and TAB2+, which resulted in the
2964 | Org. Biomol. Chem., 2016, 14, 2961–2968
This journal is © The Royal Society of Chemistry 2016

View Article Online
Organic & Biomolecular Chemistry
Paper
Fig. 5 (A) Spin density (0.001 au) distribution plot of TAB+• [B1LYP-40/
6–31G(d) + PCM(CH2Cl2)] showing, with the aid of green ellipsoid, that
88% of spin/charge density (evaluated by the Natural Population Analy-
sis)44 is localized onto the benzidine fragment while the remaining 12%
is evenly distributed over the four p-bromophenyl groups. (B, C)
Showing the spatial distribution of HOMO of TAB0 and the distribution
of the HOMO density [evaluated as qm = Σncmn2 where cmn is the coeffi-
cient of the atomic orbital χmn in HOMO (φHOMO = Σkckχk), m is the
atomic index, and n is the index of the atomic orbital in atom m] calcu-
lated at the equilibrium geometry of neutral TAB0 (B) and at the equili-
brium geometry of its cation radical, i.e. TAB+• (C).
reorganization that involves planarization of the benzidine
fragment (i.e. dihedral angle between central phenylene rings
θ reduces from 35 to 19°).17 Indeed, the HOMO density of
TAB0 at the TAB+• geometry is much closer to the spin/charge
density distribution in TAB+• (i.e. 76% on the benzidine frag-
ment and 6% per each aryl group) than the HOMO density at
the neutral TAB geometry (i.e. 60% on the benzidine fragment
and 10% per each aryl group), see Fig. 5.
In order to confirm this DFT prediction, we obtained single
crystals of TAB+• SbCl6 for crystallographic analysis. In a
−
typical experiment, a chilled (∼0 °C) Schlenk tube was charged
with TAB (30 mg, 0.037 mmol), anhydrous CH2Cl2 (5 mL), and
− 45
triethyloxonium hexachloroantimonate (Et3O+ SbCl6
)
Fig. 6 The ORTEP diagrams (50% probability) of TAB0 (A) and TAB+•
(32 mg, 0.074 mmol) under an argon atmosphere, and the
resulting mixture was stirred for 30 minutes. The resulting
dark-orange solution was carefully layered with dry toluene
(10 mL) and placed in a refrigerator for 12 hours at −10 °C.
After this time, a well-formed array of single crystals of TAB+•
−
SbCl6 (B) as well as the juxtaposition of TAB0 (grey color) and TAB+•
(orange color) showing planarization of the benzidine fragment and
enhanced propeller arrangement of the p-bromophenyl groups around
−
the nitrogen atoms (C). The packing arrangement of TAB+• SbCl6
showing the layers of TAB units with no close contact between the
−
−
SbCl6 was obtained and subsequently analysed by X-ray crys-
benzidine fragments, which are separated by the SbCl6 counter anions
(D). Note that the co-crystallized CH2Cl2 molecules and hydrogens were
omitted for clarity. Also see Fig. S8–S10 and Tables S1–S3 in the ESI.†
tallography (see Fig. S8–S10 and Tables S1–S3 in the ESI†).
The ORTEP diagrams of TAB and TAB+• as well as the crystal
packing diagram of TAB+• are presented in Fig. 6. The TAB+• and
−
SbCl6 counter ions form mixed layers along the ab plane, and
−
Availability of precise X-ray structures of TAB+• SbCl6 as
−
within these layers, the SbCl6 counter ions gravitate toward
well as of neutral TAB allowed a comparison of the bond
length changes in the TAB → TAB+• transformation. Expect-
edly, the bond contractions and elongations were mostly con-
fined to the benzidine fragment of TAB, which was
significantly planarized (i.e., the value of θ decreased from 22°
to 3°) upon 1e− oxidation, while the four p-bromophenyl rings
did not undergo significant bond length changes (Fig. 7).
Moreover, a comparison of the bond length changes in TAB →
TAB+• transformation by DFT calculations and X-ray structural
nitrogen atoms. The benzidine moieties are positioned in the
clefts formed by the brominated aryl rings without any parallel
overlap between the benzene rings, see Fig. 6D. An overlap of
TAB and TAB+• structures clearly shows that the benzidine
fragment of TAB is significantly planarized (i.e., the value of θ
decreased from 22° to 3°) while the p-bromophenyl rings
become much more propeller-shaped (i.e. reduced conjugation
with nitrogen lone pairs as judged by change in the average
dihedral angle from ∼45° to ∼35°) upon 1e− oxidation.
This journal is © The Royal Society of Chemistry 2016
Org. Biomol. Chem., 2016, 14, 2961–2968 | 2965

View Article Online
Paper
Organic & Biomolecular Chemistry
Fig. 8 (A) Cyclic voltammograms of 5 mM BC in CH2Cl2 (22 °C) con-
taining 0.2 M tetra-n-butylammonium hexafluorophosphate at different
scan rates (as denoted). (B) Spectral changes attendant upon incremen-
−
tal addition of 3.0 mM BC0 to a solution of NAP+• SbCl6 (0.15 mM) in
Fig. 7 Comparison of the oxidation-induced bond length changes (in
Å) in TAB obtained by X-ray crystallography (abscissa) and DFT calcu-
lations (ordinate). Note that the bond length changes (i.e. b1–b10)
depicted in the structure are based on the averages of equivalent bonds.
See also Fig. S12 and Table S5 in the ESI.†
CH2Cl2 at 22 °C (see also Fig. S6 in the ESI†), and (C) plot of decrease of
absorbance at 672 nm (due to the disappearance of NAP+•) and increase
of absorbance at 805 nm (due to the formation of BC+•) against the
equivalents of added neutral BC0.
BC0 were incrementally added to a solution of NAP+• (Fig. 8).
The absorption spectrum of BC+• showed a low-energy band at
analysis clearly shows a linear correspondence (Fig. 7), and
thus confirming the validity of the usage of B1LYP-40
functional for DFT calculations.
λmax = 805 nm (εmax = 9930 cm−1
M
−1) which is slightly red-
shifted as compared to the absorption band of MB+• (λmax
728 nm, εmax = 28 200 cm−1 M−1). Note that BC+• SbCl6− can be
=
These results clearly suggest that spectral contamination in
the aged sample of MB+• is largely due to the formation of
TAB+• and TAB2+, which most likely are produced by the oxi- prepared as a crystalline solid either by reaction with SbCl5 or
dative Scholl-type reaction,46 where an eventual loss of a mole-
NO+ SbCl6 similar to the preparation of MB+• SbCl6 (see
−
−
ESI†), and the spectrum of the resulting BC+• was identical to
cular bromine produces TAB, which undergoes oxidation with
MB+•.47 To prevent degradation of MB+•, we propose to substi- the spectrum presented in Fig. 8B.
tute bromine atoms in the para-positions of MB by the steri-
−
The high stability of BC+• SbCl6 will allow it to serve as a
convenient replacement of MB+• SbCl6−, and we believe that it
cally demanding tertiary-butyl groups, and place bromine
atoms at ortho-positions of the aromatic ring (see Scheme 1). will find widespread use for the spectroscopic characterization
The synthesis of the proposed tris(2-bromo-4-tbutylphenyl)
of various oxidized species, and will become a useful chemical
amine (BC) was easily accomplished from readily available
triphenylamine by a facile t-butylation48 using t-butanol and
trifluoroacetic acid followed by bromination using NBS in
DMF (see Scheme S8 in the ESI† for details).
oxidant for various oxidative transformations.
Conclusions
Cyclic voltammetric analysis of tris(4-tbutylphenyl)amine
(10) and its brominated analogue BC (Fig. 8A and S7 in the
ESI†) showed that they undergo reversible one-electron
oxidation at varying scan rates (v = 50 to 300 mV s−1). Interest-
ingly, a placement of bromine atoms at the ortho-positions in
10 increases its oxidation potential from 0.29 to 0.78 V vs. Fc/
Fc+, which suggests that ‘blues cousin’ has slightly higher oxi-
dizing power [i.e. Ered(BC+•) = 0.78 V] as compared to ‘magic
blue’ [i.e. Ered(MB+•) = 0.70 V].
In summary, we have demonstrated that ‘magic blue’ (MB+•)
undergoes slow decomposition in the solid state upon aging thus
giving rise to the additional absorption bands at 480 nm,
805 nm, and 1550 nm. A combined DFT and experimental
(NMR, electrochemistry, optical spectroscopy, and X-ray crystallo-
graphy) study led to the identification of the main decomposition
product, or blues brother, as tetrakis-(4-bromophenyl)benzidine
whose cation radical (λmax = 489, 1490 nm) and dication
(807 nm) are responsible for the additional absorption bands.
This study allowed us to further demonstrate the excellent
performance of the B1LYP-40 functional for accurately predict-
ing the electrochemical oxidation potentials of a variety of
triarylamine derivatives and the optical properties of their
cation radicals by (TD-)DFT calculations which aided in
identification of the blues brother. Moreover, the excellent
performance of the B1LYP-40 functional17,19–22 was further
demonstrated by close correspondence of the calculated
structures of TAB and TAB+• with those obtained by X-ray
crystallography.
The cation radical of BC was first generated by the spectro-
scopic redox titration, where sub-stoichiometric amounts of
Scheme 1 Synthesis of BC.
2966 | Org. Biomol. Chem., 2016, 14, 2961–2968
This journal is © The Royal Society of Chemistry 2016

View Article Online
Organic & Biomolecular Chemistry
Paper
The fact that TAB is formed by dimerization of MB led us to 16 R. Rathore, C. L. Burns and M. I. Deselnicu, Org. Synth.,
design and synthesize its close analogue tris(2-bromo-4-tert-
butylphenyl)amine referred to as ‘blues cousin’ (BC: Eox1
0.78 V vs. Fc/Fc+, λmax(BC+•) = 805 nm, εmax = 9930 cm−1 M−1),
2005, 1–9.
=
17 M. R. Talipov, A. Boddeda, S. V. Lindeman and R. Rathore,
J. Phys. Chem. Lett., 2015, 6, 3373–3378.
in which oxidative dimerization is hampered by positioning 18 K. Sreenath, C. V. Suneesh, V. K. R. Kumar and
the sterically demanding tert-butyl groups at the para-positions K. R. Gopidas, J. Org. Chem., 2008, 73, 3245–3251.
of the aryl rings. The ease of preparation of BC0 and high 19 M. Renz, K. Theilacker, C. Lambert and M. Kaupp, J. Am.
stability of its cation radical (BC+•) suggest that it will become
Chem. Soc., 2009, 131, 16292–16302.
a useful one-electron oxidant for widespread use in organic, 20 M. R. Talipov, A. Boddeda, Q. K. Timerghazin and
organometallic, and inorganic chemistry.
R. Rathore, J. Phys. Chem. C, 2014, 118, 21400–21408.
21 M. R. Talipov, T. S. Navale and R. Rathore, Angew. Chem.,
Int. Ed., 2015, 54, 14468–14472.
22 M. R. Talipov, R. Jasti and R. Rathore, J. Am. Chem. Soc.,
2015, 137, 14999–15006.
Acknowledgements
We thank the NSF (CHE-1508677) and NIH (R01- 23 A. J. Cohen, P. Mori-Sánchez and W. Yang, Science, 2008,
HL112639–04) for financial support and Dr Sergey 321, 792–794.
V. Lindeman for X-ray crystallography. The calculations were 24 A. J. Cohen, P. Mori-Sánchez and W. Yang, Chem. Rev.,
performed on the high-performance computing cluster Père at 2011, 112, 289–320.
Marquette University funded by NSF awards OCI-0923037 and 25 W. S. E. Steckhan and W. Schmidt, Chem. Ber., 1980, 113,
CBET-0521602, and the Extreme Science and Engineering Dis-
covery Environment (XSEDE) funded by NSF (TG-CHE130101).
577–585.
26 H. R. Snyder, C. Weaver and C. D. Marshall, J. Am. Chem.
Soc., 1949, 71, 289–291.
27 R. Belcher, A. J. Nutten and W. I. Stephen, J. Chem. Soc.,
1958, 2336–2338.
28 J. Cornforth, R. H. Cornforth and R. T. Gray, J. Chem. Soc.,
Perkin. Trans. 1, 1982, 2289–2297.
29 H.-M. Kang, Y.-K. Lim, I.-J. Shin, H.-Y. Kim and C.-G. Cho,
Org. Lett., 2006, 8, 2047–2050.
30 N. Berton, I. Fabre-Francke, D. Bourrat, F. Chandezon and
S. Sadki, J. Phys. Chem. B, 2009, 113, 14087–14093.
31 J. Vicente, J. Gil-Rubio, G. Zhou, H. J. Bolink and J. Arias-
Pardilla, J. Polym. Sci., Part A: Polym. Chem., 2010, 48, 3744–
3757.
References
1 F. A. Bell, A. Ledwith and D. C. Sherrington, J. Chem. Soc.
C, 1969, 2719–2720.
2 N. G. Connelly and W. E. Geiger, Chem. Rev., 1996, 96, 877–
910.
3 K. T. Lorenz and N. L. Bauld, J. Am. Chem. Soc., 1987, 109,
1157–1160.
4 T. Kim, G. A. Mirafzal, J. Liu and N. L. Bauld, J. Am. Chem.
Soc., 1993, 115, 7653–7664.
5 A. M. Bond, R. Colton, D. A. Fiedler, J. E. Kevekordes, 32 J. H. Cho, Y.-S. Ryu, S. H. Oh, J. K. Kwon and E. K. Yum,
V. Tedesco and T. F. Mann, Inorg. Chem., 1994, 33, 5761–5766. Bull. Korean Chem. Soc., 2011, 32, 2461.
6 W. Yueh and N. L. Bauld, J. Am. Chem. Soc., 1995, 117, 33 M. Wang, C. Li, A. Lv, Z. Wang, Z. Bo and F. Zhang,
5671–5676. Polymer, 2012, 53, 324–332.
7 N. L. Bauld, J. T. Aplin, W. Yueh, H. Sarker and 34 H.-F. Huang, S.-H. Xu, Y.-B. He, C.-C. Zhu, H.-L. Fan,
D. J. Bellville, Macromolecules, 1996, 29, 3661–3662.
8 K. Heinze and S. Reinhardt, Organometallics, 2007, 26,
5406–5414.
9 H. Wang, X. Wang, M. A. Winnik and I. Manners, J. Am.
Chem. Soc., 2008, 130, 12921–12930.
X.-H. Zhou, X.-C. Gao and Y.-F. Dai, Dyes Pig., 2013, 96,
705–713.
35 B.-Y. Lim, B.-E. Jung and C.-G. Cho, Org. Lett., 2014, 16,
4492–4495.
36 J. Safaei-Ghomi and Z. Akbarzadeh, Ultrason. Sonochem.,
2015, 22, 365–370.
10 C. M. Davis, K. Ohkubo, I.-T. Ho, Z. Zhang, M. Ishida,
Y. Fang, V. M. Lynch, K. M. Kadish, J. L. Sessler and 37 R. Rathore and J. K. Kochi, Acta Chem. Scand., 1998, 52,
S. Fukuzumi, Chem. Commun., 2015, 51, 6757–6760. 114–130.
11 R. Rathore and C. L. Burns, J. Org. Chem., 2003, 68, 4071– 38 R. Rathore, C. L. Burns and M. I. Deselnicu, Org. Lett.,
4074. 2001, 3, 2887–2890.
12 T. S. Navale, L. Zhai, S. V. Lindeman and R. Rathore, Chem. 39 M. Banerjee, V. S. Vyas, S. V. Lindeman and R. Rathore,
Commun., 2009, 2857–2859. Chem. Commun., 2008, 1889–1891.
13 L. Eberson and B. Larsson, Acta Chem. Scand., 1986, 40, 40 R. Shukla, S. H. Wadumethrige, S. V. Lindeman and
210–225. R. Rathore, Org. Lett., 2008, 10, 3587–3590.
14 R. Rathore, S. V. Lindeman, A. S. Kumar and J. K. Kochi, 41 V. J. Chebny, T. S. Navale, R. Shukla, S. V. Lindeman and
J. Am. Chem. Soc., 1998, 120, 6931–6939. R. Rathore, Org. Lett., 2009, 11, 2253–2256.
15 R. Rathore and J. K. Kochi, J. Org. Chem., 1995, 60, 4399– 42 V. J. Chebny, R. Shukla, S. V. Lindeman and R. Rathore,
4411.
Org. Lett., 2009, 11, 1939–1942.
This journal is © The Royal Society of Chemistry 2016
Org. Biomol. Chem., 2016, 14, 2961–2968 | 2967

View Article Online
Paper
Organic & Biomolecular Chemistry
43 T. S. Navale, K. Thakur, V. S. Vyas, S. H. Wadumethrige, 47 Magic blue is rather stable and does not undergo ready
R. Shukla, S. V. Lindeman and R. Rathore, Langmuir, 2012,
28, 71–83.
44 F. Weinhold and C. R. Landis, Valency and bonding: a
natural bond orbital donor–acceptor perspective, Cam-
bridge University Press, Cambridge, New York, UK,
2005.
dimerization, and therefore it is likely that a slow decompo-
sition of MB+• to neutral MB catalyses the formation of TAB.
For example, see: O. Yurchenko, D. Freytag, L. zur Borg,
R. Zentel, J. Heinze and S. Ludwigs, J. Phys. Chem. B, 2012,
116, 30–39 and M. Talipov and R. Rathore, Robust Aromatic
Cation Radicals as Redox Tunable Oxidants, in Organic Redox
Systems: Synthesis, Properties, and Applications, ed. T. Nishi-
naga, John Wiley & Sons, Inc, Hoboken, NJ, 2015, pp. 131–175.
45 R. Rathore, A. S. Kumar, S. V. Lindeman and J. K. Kochi,
J. Org. Chem., 1998, 63, 5847–5856.
46 L. Zhai, R. Shukla, S. H. Wadumethrige and R. Rathore, 48 D. Vak, J. Jo, J. Ghim, C. Chun, B. Lim, A. J. Heeger and
J. Org. Chem., 2010, 75, 4748–4760.
D.-Y. Kim, Macromolecules, 2006, 39, 6433–6439.
2968 | Org. Biomol. Chem., 2016, 14, 2961–2968
This journal is © The Royal Society of Chemistry 2016