ESR and ENDOR investigations on various Wurster's radical cations in solution. Experimental results, theoretical ab initio, and DFT calculations

Article abstract of DOI:10.1007/s00706-004-0224-4

ESR and ENDOR spectra are reported of several symmetrical substituted N,N,N′,N′-tetraalkyl-p-phenylenediamine radical cations in solution. Different N,N′-alkyl substituted paraphenylenediamines, like the ethyl, n-propyl, and iso-propyl derivative are compared with the parent N,N,N′,N′-tetramethyl-p-phenylenediamine (Wurster's Blue Cation). N,N,N′,N′-Tetrabenzyl-p-phenylenediamine, 1,4-dipyrrolidinylbenzene, and N,N′-bis[4-(dimethylamino)phenyl]piperazine are additionally investigated. Experimental and calculated hyperfine coupling constants are compared. Characteristic UV-VIS data and redox potentials in acetonitrile are reported, together with the syntheses of the compounds. Springer-Verlag 2005.

Full text of DOI:10.1007/s00706-004-0224-4

Monatshefte f€ur Chemie 136, 519–536 (2005)
DOI 10.1007/s00706-004-0224-4
ESR and ENDOR Investigations
on Various Wurster’s Radical Cations
in Solution. Experimental Results,
Theoretical ab initio,
and DFT Calculations
Gu¨nter Grampp1;ꢀ, Anne-Marie Kelterer1, Stephan Landgraf1,
Michael Sacher1, Dominique Niethammer2, Joa˜o P. Telo3,
Rui M. B. Dias3, and Abel J. S. C. Vieira4
1
Graz University of Technology, Institute of Physical and Theoretical Chemistry,
A-8010 Graz, Austria
2
Free University of Berlin, Institute of Organic Chemistry, D-14195 Berlin, Germany
´
Instituto Superior Tecnico, Quımica Organica, P-1096 Lisbon, Portugal
3
4
´
Universidade de Nova Lisboa, Departmento de Quımica, P-2825-114
Monte de Caparica, Portugal
Received February 6, 2004; accepted (revised) May 23, 2004.
Published online January 31, 2005 # Springer-Verlag 2005
Summary. ESR and ENDOR spectra are reported of several symmetrical substituted N,N,N0,N0-
tetraalkyl-p-phenylenediamine radical cations in solution. Different N,N0-alkyl substituted para-
phenylenediamines, like the ethyl, n-propyl, and iso-propyl derivative are compared with the parent
N,N,N0,N0-tetramethyl-p-phenylenediamine (Wurster’s Blue Cation). N,N,N0,N0-Tetrabenzyl-p-pheny-
lenediamine, 1,4-dipyrrolidinylbenzene, and N,N0-bis[4-(dimethylamino)phenyl]piperazine are addi-
tionally investigated. Experimental and calculated hyperfine coupling constants are compared.
Characteristic UV-VIS data and redox potentials in acetonitrile are reported, together with the synth-
eses of the compounds.
Keywords. ESR; ENDOR; Wurster’s radical cations; Hyperfine coupling constants; Ab initio; AM1;
DFT; B3LYP; Spin density calculations.
Introduction
More than 100 years ago the persistent one-electron oxidation products of p-
phenylenediamines and their derivatives have been discovered by C. Wurster [1].
ꢀ
Corresponding author. E-mail: grampp@ptc.tugraz.at

520
G. Grampp et al.
The radical character of these compounds has been recognized by use of magnetic
balance measurements much later [2], the coloured radicals have been named in the
literature according to their discoverer as Wurster’s cations, especially the highly
stable semiquinone radical cation of N,N,N0,N0-tetramethyl-p-phenylenediamine
(Wurster’s Blue), and the radical cation of N,N-dimethyl-p-phenylenediamine
(Wurster’s Red).
p-Phenylenediamines found a wide range of applications from industrial, com-
mercial, and scientific points of view. One of the important industrial applications
is their utilization as film developers, especially as coupling components in the
developing process of color films and slides [3]. Weissman et al. [4–6] published
the first papers on ESR investigations of Wurster’s Blue and Wurster’s Red radical
cations. Other early reports were given by Hausser [7, 8]. The first detailed in-
terpretation of the complex ESR-spectrum of N,N,N0,N0-tetramethyl-p-phenyl-
enediamine was reported by Bolton et al. [9] and later in the textbook of Wertz
and Bolton [10, 11], who used it as an ESR-standard for magnetic field calibration
and for g-factor measurements. The number of papers dealing with ESR data of the
radical cation of N,N-dimethyl-p-phenylenediamine (Wurster’s Red Cation) is rela-
tively small [12–14]. Slightly different results are reported on the hyperfine cou-
pling constants of this radical cation [15, 16]. Grampp and Stiegler [12] gave a
detailed analysis of the corresponding ESR spectrum based on specific deuteration.
A degeneration by chance is reported for the nitrogen and the methyl proton
coupling constants ðaN ¼ 3aMH eÞ of the 2,3,5,6-tetramethyl-p-phenylenediamine
radical cation [12, 17, 18]. To our knowledge, the ESR spectrum of the
N,N,N0,N0-tetraethyl-p-phenylenediamine radical cation is reported only in three
publications [12, 19, 20]. The published ESR coupling constants in Refs. [12] and
[19] correlate well with our ESR and ENDOR results, whereas in Ref. [20] identical
coupling constants are given for different alkyl substituents (R ¼ ethyl, n-propyl,
n-butyl). This is probably due to poor spectral resolution.
From the higher derivatives of the sym-N,N,N0,N0-tetraalkyl substituted
Wurster’s cations only very few ESR-reports exist [20–23]. Up to now no detailed
ENDOR-investigations have been reported. The aim of this paper is to present
detailed ESR and ENDOR data of higher substituted tetraalkyl-p-phenylenedia-
mines together with UV-VIS data, redox properties, and detailed quantum chemical
spin density calculations. We report on the radical cations of N,N,N0,N0-tetraethyl-
PPD (TEPPD), N,N,N0,N0-tetra-n-propyl-PPD (TPPPD), N,N,N0,N0-tetra-iso-propyl-
PPD (TiPPPD), N,N,N0,N0-tetrabenzyl-PPD (TBzPPD), 1,4-dipyrrolidinylbenzene
(DPB), and for comparison on N,N,N0,N0-tetramethyl-p-phenylenediamine
(TMPPD). Additionally we report on the N,N0-bis[4-(dimethylamino)phenyl]piper-
azine BDMAP radical cation, a compound recently described by Nelsen et al. [24].
Structures are shown in Scheme 1.
The alkyl substituted compounds have, to the best of our knowledge, not yet
been investigated by computational methods. Therefore, various quantum chemical
investigations have been performed at appropriate levels. The complete conforma-
tional analyses done with the semiempirical AM1 method include TEPPD and
TPPPD with respect to the conformers which are populated under our experimen-
tal conditions at room temperature. For TBzPPD and DPB the global minimum of

Various Wurster’s Radical Cations
521
Scheme 1
TEPPD was the starting point in our study. Final conformational analyses done by
ꢀ
the density functional approach using the B3LYP functional with the 6-31G basis
set are presented, leading to calculated spin densities which are correlated with the
experimental hyperfine coupling constants.
Results and Discussion
Figures 1–5 show the experimental ESR and ENDOR spectra of the radical cations
from TEPPD, TPPPD, TiPPPD, TBzPPD, DPB, and N,N0-bis[4-(dimethyl-
amino)phenyl]piperazine (BDMAP), together with the corresponding simulations.
All ESR hyperfine coupling constants obtained are listed in Tables 1–7. As ex-
pected, only slight changes in the hyperfine coupling constants occur compared to
Wurster’s Blue cation of TMPPD, depending on the nature of the substituents. The
ESR nitrogen coupling aN decreases from 0.705 mT (TMPPD, R ¼ Me, Table 1) to
0.696 mT for the ethyl substituent (Table 2) and then increases again to 0.715 mT
for both, the n-propyl (Table 3) and the i-propyl (Table 4) compound. The coupling
constants of the tetrabenzyl substituted radical show values similar to the other
alkylsubstituted compounds. An interesting change in the relation of the nitrogen
coupling constant, aN, and the adjacent hydrogen coupling is observed in the DPB
radical cation compared to the other radical cations. Unhindered rotation of the
methyl groups in TMPPD gave a ratio of R ¼ 1.04 for aN=aMH e, indicating a planar
geometry with freely rotating CH3-groups. Changing from methyl to methylene
groups, as per TEPPD, TPPPD, TBzPPD, this ratio increases to a mean value of
R ¼ 1.85. The strong steric hindrance in TiPPPD increases R to 7.15, as also
shown by Bock et al. [28].

522
G. Grampp et al.
Fig. 1. a) Experimental CW-ESR spectrum of the N,N,N0,N0-tetraethyl-p-phenylenediamine radical
cation in H2O at T ¼ 298 K; b) simulated spectrum, inserts: enlargements of 0.3 mT; c) ENDOR-
spectrum in glycerol=H2O (3:1 v=v) at T ¼ 298 K
This ratio changes to values lower than one for the less flexible radical cations
of DPB and BDMAP, where different ring systems adjacent to the nitrogen atoms
are present. R ¼ 0.70 and 0.94 are the aN=aꢁHCH ꢁ-ratios, respectively.
2
The coupling constants of the aromatic protons remain nearly constant.
These values cover a small range from 0.189 to 0.199 mT. According to the
strong out-of-plane arrangement of the isopropyl groups, two types of ꢀ-protons
with slightly different couplings are found, both experimentally and theoreti-
cally, see Table 4. aH1 ꢀ ¼ 0:100 mT and aH2 ꢀ ¼ 0:075 mT are the corresponding

Various Wurster’s Radical Cations
523
Fig. 2. a) Experimental CW-ESR spectrum of the N,N,N0,N0-tetra-n-propyl-p-phenylenediamine
radical cation in 2-propanol=DMSO (1:1 v=v) at T ¼ 280 K; b) simulated spectrum; c) ENDOR-
spectrum in 2-propanol=DMSO (1:1 v=v) at T ¼ 280 K
proton coupling constants. Values found for the ꢁ-protons are extremely small,
as normally reported for such type of protons [29, 30]. The assignment of the
hfc-constant of the N,N0-bis[4-(dimethylamino)phenyl]piperazine radical cation
is given in Table 7. The hyperfine pattern shows two nearly equivalent nitro-
gen couplings, four equivalent aromatic protons, and ten nearly equivalent
protons from two methyl and two methylene groups. These values clearly
indicate that the unpaired electron is localized on one p-phenylenediamine
subunit.

524
G. Grampp et al.
Fig. 3. a) Experimental CW-ESR spectrum of the N,N,N0,N0-tetra-iso-propyl-p-phenylenediamine
radical cation in H2O at T ¼ 298 K; b) simulated spectrum; c) ENDOR-spectrum in glycerol=H2O
(3:1 v=v) at T ¼ 298 K
The cyclic voltamogramms of all compounds clearly show two separated rever-
sible oxidation steps. Redox potentials of the first and second ðo1xÞ idation steps are
given in Table 10. Whereas the first oxidation potentials, Eox , are similar for
TMPPD and TiPPPD and for TEPPD and TPPPD, the second oxidation steps
for the quinonediimine are nearly identical for these four compounds and indepen-
dent of the substituents on N. The reason for this may be that these radical cations
as well as the corresponding quinonediimines are completely planar, as indicated

Various Wurster’s Radical Cations
525
Fig. 4. a) Experimental CW-ESR spectrum of the N,N,N0,N0-tetrabenzyl-p-phenylenediamine radical
cation in 2-propanol=DMSO (1:1 v=v) at T ¼ 300 K; b) simulated ESR-spectrum; c) ENDOR-spec-
trum in 2-propanol=DMSO (1:1 v=v) at T ¼ 300 K
by both ESR results and theoretical calculations. The oxidation potentials decrease
from TMPPD, TEPPD to TPPPD, revealing the increasing electron releasing effect
in the homologous series of substituents: methyl, ethyl, and propyl. A Hammett
plot with these three compounds is linear. The redox potential of TiPPPD does not
fall on such a Hammett plot, its potential increases to ꢁ295 mV, indicating the
lower stability of this radical cation. Again its steric hindrance may be the reason
for such a behaviour. An X-ray structure for the diprotonated TiPPPD H22þ is
reported in the literature [28], the two isopropyl groups attached to each nitrogen

526
G. Grampp et al.
Fig. 5. a) Experimental CW-ESR spectrum of the 1,4-dipyrrolidinylbenzene radical cation in H2O at
T ¼ 298 K; b) simulated spectrum, inserts: enlargements of 0.3 mT region
Table 1. Experimental and theoretical hfc-constants [mT] of the TMPPDþꢃ ; geometries of the structure
ꢀ
in Ref. [25] were reoptimized with the B3LYP=6-31G method; the Fermi contact analysis is made by
the methods listed
TMPPDþꢃ
aN
ESR
UB3LYP=6-31G
UB3LYP=ESR-II
ꢀ
0.705a
0.7051b
0.652
0.552
aHarom
0.199a
0.1989b
ꢁ0.19
ꢁ0.185
aHCH
0.677a
0.6773b
0.731
0.781
3
a
b
In abs. methanol at T ¼ 293 K, Ref. [12]; in abs. ethanol at T ¼ 296 K, Ref. [10]
atom are twisted by 56 and 77ꢂ out of the plane of the benzene ring. A different
behaviour is also found for the potentials of TBzPPD and DPB. DPD is a some-
what different system since the alkyl groups are fixed there in a five-membered
ring. The first oxidation step of DPB is identical to that of TPPPD, whereas the
second oxidation is performed easier compared with TPPPD. TBzPPD shows the
highest Eoð1xÞ-value, ꢁ120 mV vs. Fc=Fcþ, and also the highest Eoð2xÞ, 525 mV vs.
Fc=Fcþ, for forming the corresponding quinonediimine dication.

Various Wurster’s Radical Cations
527
Table 2. Experimental and theoretical hfc-constants [mT] of TEPPDþꢃ ; the calculated hfc-constants
are the result of Boltzmann distribution at room temperature including the DFT data from Table 8;
Fermi contact analysis is made with the methods listed
TEPPDþꢃ
aN
ESR
ENDOR
–
UB3LYP=6-31G
UB3LYP=ESR-II
ꢀ
0.688a
0.696b
0.665
0.565
aHarom
0.185a
0.198b
0.188c
0.369a
0.372b
0.376c
0.1967d
0.197b
ꢁ0.188
0.371
ꢁ0.184
aHꢀ
0.3841d
0.379b
0.308
0.475
aHꢁ
0.010a
0.010b
0.0096c
0.0105d
0.010b
ꢁ0.008
ꢁ0.005
In abs. methanol at T ¼ 293 K, Ref. [12]; b in 2-propanol=DMSO mixture at T ¼ 280 K; c in H2O at
a
d
T ¼ 298 K; in glycerol=H2O at T ¼ 298 K
Table 3. Experimental and calculated hfc-constants [mT] of the TPPPDþꢃ ; the calculated hfc-
constants are the result of Boltzmann distribution at room temperature including the DFT data from
Table 9; Fermi contact analysis is made with the method listed
TPPPDþꢃ
aN
ESR
ENDOR
0.695a
UB3LYP=6-31G
ꢀ
0.715a
0.692b
0.697c
0.664
aHarom:
0.199a
0.181b
0.194c
0.194a
ꢁ0.185
aHꢀ
aHꢁ
0.377a, c
0.380b
0.027a
0.028b
0.0203c
0.378a
0.022a
0.374
ꢁ0.036
aHꢂ
–
–
0.022
a
b
c
In 2-propanol=DMSO at T ¼ 290 K; in methanol at T ¼ 265 K; in H2O at T ¼ 298 K
Theoretical Investigations of Spin Densities
Methods
The isotropic hyperfine coupling constants depend strongly on the geometry and
especially the bond lengths of the radicals. Based on a paper of Gescheidt et al.
[31] that compares semiempirical Hartree-Fock and density functional methods for

528
G. Grampp et al.
Table 4. Experimental and calculated hfc-constants [mT] of the TiPPPDþꢃ ; the Fermi contact analysis
is done for the all-t-conformer with the method listed
TiPPPDþꢃ
aN
ESR
ENDOR
0.714a
UB3LYP=6-31G
ꢀ
0.715a
0.715b
0.700
aHarom:
aH1 ꢀ
0.189a
0.177b
0.100a
0.101b
0.187a
0.187c
0.100a
0.104c
ꢁ0.158
0.163
aH2 ꢀ
0.075a
0.075a
0.085c
ꢁ0.024
aHꢁ
0.016a
0.017a, c
–
a
b
c
In 2-propanol=DMSO at T ¼ 280 K; in methanol at T ¼ 267 K; in glycerol=H2O at T ¼ 298 K
Table 5. Experimental hfc-constants [mT] of the TBzPPDþꢃ
TBzPPDþꢃ
aN
ESR
ENDOR
–
0.672b
0.662c
aHarom:
aHꢀ
aHphenyl
0.193b
0.197c
0.388b
0.389c
0.188a
0.390a
0.014a
0.007b
a
b
c
In 2-propanol=DMSO at T ¼ 300 K; in 2-propanol=DMSO at T ¼ 280 K; in methanol at
T ¼ 265 K
Table 6. Experimental and calculated hfc-constants [mT] of the DPBþꢃ ; the Fermi contact analysis is
done for the global minimum with the method listed
DPBþꢃ
aN
ESR
ENDOR
0.686a
UB3LYP=6-31G
ꢀ
0.682a
0.687b
0.648
aHarom:
aHꢀ
aHꢁ
a
0.201a
0.201b
0.975a
0.959b
0.017a
0.014b
0.200a
0.966a
0.015a
–
–
–
b
In 2-propanol=DMSO at T ¼ 270 K; in H2O at T ¼ 298 K

Various Wurster’s Radical Cations
529
Table 7. Experimental hfc-constants [mT] of the N,N0-bis[4-(dimethylamino)phenyl]piperazine
radical cation
BDMAPþꢃ
ESR
aN
0.703a
0.750a
0.189a
aHCH ;ꢁCH ꢁ
3
2
aHarom:
a
In methanol=ethanol mixture (1:4 v=v) at T ¼ 268 K
ꢀ
(mT) of the TEPPD radical cation; the DFT conformers were used for the Boltzmann distribution; the
Table 8. Relative AM1 and B3LYP=6-31G energies (in kJ=mol) and hyperfine coupling constants
ꢀ
B3LYP=6-31G absolute energy of the global minimum is ꢁ657.2386734 a.U.; the dihedral angles of
the ethyl unit are gauche, g ꢄ 60 degree, g0 ꢄ ꢁ60 degree and trans, t ꢄ 180 degree for the rotation
around the ꢁCH2ꢁCH3 bond
Structure
ggg0g0
gggg
gggg0
ggg0g0
gg0gg0
degeneration
Erel(AM1)
Erel(DFT)
4
2
8
4
4
0.0
0.0
0.12
1.63
4.60
17%
ꢁ0.187
3.13
3.26
0.10
25.1%
ꢁ0.186
8.90
9.36
Boltzmann distribution 50.6%
6.1%
ꢁ0.187
1.2%
ꢁ0.187
a(Har)
a(Hꢃ)
ꢁ0.189
0.294 (4H)
0.307 (4H)
0.303 (3H)
0.522 (4H)
0.339 (4H)
0.448 (4H)
0.432 (4H)
0.443 (3H)
0.576 (1H)
0.173 (1H)
0.662
0.222 (4H)
0.224 (4H)
a(N)
0.663
0.664
0.675
0.678
0.669
a number of organic radicals, two different methods were used for our molecules
for optimization of the structures.
The semiempirical AM1 method can be used for a fast calculation of the
potential energy surface. Density functional theory (DFT) can be employed with
low computational costs and high accuracy for the geometry reoptimization of
radicals with more than five heavy atoms [31]. Becke’s Three Parameter Hybrid
method including the Lee, Yang, Perdew Correlation Functional (B3LYP) and the
ꢀ
larger contracted Gaussian basis set 6-31G were used in this work.
TMPPD and DPB were optimized with the two methods described. The ge-
ometries of TEPPD and TPPPD have been calculated by using the optimized
structure of TMPPD so that the ꢃ-carbons were fixed in the plane of the aromatic
ring based on the results of full geometry optimizations of TMPPD and DPB. The
full potential energy surfaces have been investigated using the semiempirical AM1
method for TEPPD and TPPPD. Reoptimization of the five and nine lowest local
minima by the DFT method results then in five and eight minima, which have been
included in this study. Fermi contact analyses were done for all these minima. For

530
G. Grampp et al.
TiPPPD, the global minimum of TEPPD was used as initial structure with an ꢃ-
hydrogen changed into CH3. Published optimized geometry of TMPPD served as a
reference [25, 26].
AM1 calculations were done with the program VAMP [32] for TEPPD and
TPPPD. All other calculations were made with the GAUSSIAN 94 program pack-
age [33].
Fermi contacts are obtained in atomic units by GAUSSIAN 94 via Eq. (1) [34].
ꢀ
(mT) of the TPPPD radical cation; the DFT conformers were used for the Boltzmann distribution; the
Table 9. Relative AM1 and B3LYP=6-31G energies (in kJ=mol) and hyperfine coupling constants
ꢀ
B3LYP=6-31G absolute energy of the global minimum is ꢁ814.4968627 a.U.; the dihedral angles
A-B of the propyl unit are gauche g ꢄ 87 degree, g0 ꢄ ꢁ87 degree and trans t ꢄ 180 degree for A (the
rotation around the ꢁCH2ꢁCH2 bond) and gauche, g ꢄ 60 degree, g0 ꢄ ꢁ60 degree and trans, t ꢄ 180
degree for B (the rotation around the ꢁCH2ꢁCH3 bond)
Structure
gt,gt,gt,gt
g0t,g0t,gt,gt
gg,gt,gt,gt
g0g0,g0t,gt,gt
g0t,gt,gt,gt
degeneration
Erel(AM1)
Erel(DFT)
Boltzmann
distribution
a(Har)
4
2
2
2
2
0.10
0.0
0.0
1.42
4.20
6.5%
1.34
4.32
6.2%
1.50
4.39
6.0%
4.40
6.0%
70.6%
ꢁ0.185
ꢁ0.177
ꢁ0.184
ꢁ0.186
ꢁ0.178
ꢁ0.193
ꢁ0.193
a(Hꢃ)
0.430 (4H)
0.316 (4H)
0.491 (4H)
0.270 (4H)
0.364 (1H)
0.357 (1H)
0.485 (1H)
0.289 (1H)
0.439 (2H)
0.313 (2H)
0.665
0.467 (1H)
0.303 (1H)
0.347 (1H)
0.377 (1H)
0.445 (2H)
0.300 (2H)
0.663
0.488 (1H)
0.256 (1H)
0.561 (1H)
0.194 (1H)
0.436 (2H)
0.318 (2H)
0.659
a(N)
0.664
0.659
0.668
0.664
0.664
0.668
Structure
g0g0,g0g0,gt,gt g0g0,g0g0,g0t,g0t g0g0,g0t,gg,gt
gt,gg,gg,gt
degeneration
Erel(AM1)
Erel(DFT)
Boltzmann
distribution
a(Har)
2
2
2
2
2.60
8.35
1.2%
2.66
8.19
1.3%
2.69
8.57
1.1%
2.74
8.44
1.1%
ꢁ0.177
ꢁ0.193
ꢁ0.177
ꢁ0.191
ꢁ0.185
ꢁ0.185
a(Hꢃ)
0.488 (1H)
0.556 (2H)
0.438 (2H)
0.483 (2H)
0.292 (2H)
0.370 (2H)
0.353 (2H)
0.255 (1H)
0.561 (1H)
0.194 (1H)
0.439 (2H)
0.313 (2H)
0.659
0.215 (2H)
0.439 (2H)
0.314 (2H)
0.292 (2H)
0.336 (2H)
0.390 (2H)
a(N)
0.662
0.664
0.664
0.644
0.668

Various Wurster’s Radical Cations
531
Table 10. Redox potentials of the first and second oxidation steps of various p-phenylenediamines in
CH3CN; supporting electrolyte: 0.1 M TBAP; potentials in mV vs. Fc=Fcþ (ferrocene=ferrocenium);
UV-VIS data of the semiquinone radical cations in methanol
þ
þ
þ
Eoð1xÞðR=Sꢃ Þ
Eoð2xÞðSꢃ =T2þÞ
ꢆmaxðSꢃ Þ
mV
mV
nm
TMPPDa
TEPPDa
TPPPD
TiPPPD
TBzPPD
DPB
ꢁ280
ꢁ360
ꢁ375
ꢁ295
ꢁ120
ꢁ375
þ295
þ295
þ300
þ300
þ525
þ235
614, 565, 527 (sh)
611, 561, 526 (sh)
617, 574, 528 (sh)
617, 574, 528 (sh)
619, 570, 528 (sh)
616, 565, 524
a
From Ref. [42]
8ꢄ
ai ¼ 0:28025 ꢅ geꢀegNꢀN ꢅ ꢅðriÞ
3h
ð1Þ
2
ꢅ(ri) is the spin density on the nucleus i, which is just ꢅðriÞ ¼ jCðrÞj , where C(r)
is the molecular orbital containing the unpaired electron. ge and gN are the electron
and nuclear g-factors, ꢀB and ꢀN are the Bohr and nuclear magnetons, respectively.
Conversion factors from atomic units to mT are 1595 for hydrogen, 404.3 for
nitrogen, and 115.4 for carbon, respectively.
ꢀ
Fermi contacts are also computed with the B3LYP=6-31G method and the
more reliable basis set for magnetic properties B3LYP=EPR-II. Boltzmann distri-
butions were calculated for TEPPD and TPPPD after degeneration of the local
minima to receive the mean computed hyperfine coupling constant for comparison
with the experimental values at room temperature. Tables 8 and 9 show the com-
puted hyperfine coupling constants for the local minima of TEPPD and TPPPD.
The TMPPD radical cation has been reoptimized with the method described in the
Methods section, based on published results [25, 26]. The methyl carbon atoms lie
in the plane of the aromatic ring. The calculated hyperfine coupling constants are
compared with experimental data [12, 35] in Table 1. Although the EPR-II basis set
is referred specially for hyperfine coupling constants, the combination of the den-
ꢀ
sity functional method using the B3LYP functional and the 6-31G basis set gives
hyperfine coupling constants closer to the experimental ESR values. The nitrogen
hyperfine coupling constant is calculated around 8% too small, whereas the methyl
value is calculated around 5% too large. The aromatic hydrogen couplings constant
agree very well with the experimental ones.
For the radical cation of TEPPD, 64 local minima are found using the AM1
method, some of them are degenerated. The five energetically lowest confor-
mations with relative energies below 3.26 kJ=mol were then reoptimized with
DFT, and lead to a range of relative energies up to 9.36 kJ=mol. After Fermi
contact analyses the calculated Boltzmann weighted coupling-constants are listed
in Table 3. Again, the EPR-II coupling constant for nitrogen is ꢄ20% smaller than
ꢀ
the experimental one, whereas the 6-31G hfc-constant for nitrogen is in good
agreement with the experimental value. The hyperfine coupling constants for the
aromatic and the ꢃ- and the ꢀ-hydrogen atoms are in good agreement with the
experimental values. The total spin densities for the global minimum are 0.286 on

532
G. Grampp et al.
the nitrogen atoms and 0.093 on the connecting aromatic carbon atoms for all five
conformations included.
For the radical cation of TPPPD the nine energetically lowest AM1 minima,
have relative energies lower than 2.74 kJ=mol (see Table 9). DFT optimization of
these nine geometries lead to nine conformations with relative energies lower than
8.57 kJ=mol. Boltzmann weighed hfc-constants are listed in Table 3. Because the
EPR-II basis set gives hyperfine coupling constants which are around 20% too
ꢀ
small for the nitrogen atoms in TMPPD and TEPPD, only the 6-31G basis set
was used for further investigations. Again the nitrogen hyperfine couplings are
around 5% smaller than the experimental values, which show a slight solvent
dependence. The hydrogen coupling constants are in good agreement with the
experimental data.
For the radical cation of TiPPPD the global minimum of TEPPD was used to
create the starting configuration, which is optimized to an all-trans conformation of
all ꢃ-hydrogen atoms. The computed Fermi contacts are shown in Table 4 and
deviate from the experimental ꢃ-proton-constants. This leads to the conclusion that
our choice of the all-t structure is not the most stable conformer in solution.
The radical cations of TBzPPD and BDMAP have not been included in the
computational study. The measured coupling constants are listed in Tables 5 and 7,
respectively.
ꢀ
The radical cation of DPB was optimized using the B3LYP=6-31G method
and it shows the ꢃ-carbon atoms of the five-ring in the plane of the benzene ring,
whereas the ꢀ-carbon atoms lie 5 degree out of plane. The nitrogen Fermi contact
again gives nitrogen couplings 5% smaller than the experimental ones measured in
different solvents (see Table 6).
Conclusion
EPR and ENDOR spectra of the radical cations of six alkyl- and benzyl-substi-
tuted N,N,N0,N0-p-phenylenediamines (TMPPD-DPB), together with N,N0-
bis[4-dimethylamino)phenyl]piperazine (BDMAP) have been investigated experi-
mentally. Both, EPR and ENDOR spectra are well resolved, not only for the
ꢃ-hydrogen couplings, but also for the ꢀ-hydrogen couplings in TEPPD-DPB
and even for the ꢁ-hydrogen couplings in TiPPPD.
Density functional analyses of the hyperfine coupling constants have been
performed for the compounds TMPPD, TEPPD, TPPPD, TiPPPD and DPB,
including Boltzmann-weighted conformers for TEPPD and TPPPD and global or
local minima for all other compounds. The computed hyperfine coupling constants
are always too small and show deviations from the experimental ones of around
0.01 mT for the ꢃ-hydrogen atoms, 0.015 mT for the ꢀ-hydrogen atoms (ꢄ5%),
and around 0.05 mT (ꢄ8%) for the nitrogen atoms. Such deviations have also been
reported for the calculations of other one-electron magnetic properties [35]. The
EPR-II basis set, which is reported explicitly for the calculation of hyperfine cou-
pling constants with the B3LYP functional, shows a deviation of more than 20%
ꢀ
from the experimental values of the nitrogen atoms using our B3LYP=6-31G
optimised structures of TMPPD and TEPPD and was therefore not applied for
the larger molecules.

Various Wurster’s Radical Cations
533
Experimental
1H and 13C NMR spectra were recorded on a Bruker MSL-300 NMR spectrometer. UV-VIS spectra of
the radical cations (Sþꢃ ) were obtained by iodine oxidation of the corresponding p-phenylenediamine
in methanol. The UV-VIS spectra of the neutral compounds (R) were measured in cyclohexane. UV-VIS
spectra were recorded either with a Shimadzu-spectrometer UV-3101PC or a Bruins Instruments Omega-
€
10 UV=VIS spectrometer. Melting points were determined using a Buchi melting point apparatus. The
reported melting points are uncorrected. A JEOL-ESR X-band spectrometer, type PE-3X (100kHz field
modulation) with microwave preamplification [37] and a temperature control unit JEOL-VT-10 was used
to record the CW-ESR spectra (Graz group). The amplified signals were stored via an A=D-interface
(BMC, Typ PC20) in a PC. Coupling constants were extracted from the CW-ESR spectra by using the
autocorrelation-function procedure [38, 39]. ESR and ENDOR spectra were recorded on a Bruker ER
220D ESR spectrometer equipped with a Bruker ENDOR cavity ER200ENB (Berlin group). For ESR
measurements typical experimental conditions were 1.24 or 2 mW microwave power level and 0.01 mT
field modulation (12.5kHz). For ESR saturation and ENDOR measurement a microwave power level of
20mW was used. The RF power was between 120 and 230 W. FM modulation amplitude was set
between ꢆ30 kHz and ꢆ50 kHz, modulation frequency 10kHz. ESR spectral simulations were per-
formed on a local workstation (SGI Indigo) or on an Pentium PC by means of the programs EPRFT
or HFFIT [39]. The Lisbon group used a Bruker ESR spectrometer, type ESP 300 E equipped with an
EN 810 ENDOR unit. A conventional three-electrode arrangement was used to get the redox poten-
tials by cyclic voltammetry in CH3CN. Tetrabutylammonium perchlorate (0.1M) served as supporting
electrolyte. Redox potentials and UV-VIS data are listed in Table 10.
N,N,N0,N0-Tetramethyl-p-phenylenediamine (TMPPD)
The compound was purchased from Aldrich and sublimated before use. Completely white crystals
have been obtained after several careful sublimations.
N,N,N0,N0-Tetraethyl-p-phenylenediamine (TEPPD)
Freshly distilled N,N-diethyl-p-phenylenediamine (5.0g, 30mmol, Merck, 98%) was dissolved in
75mmol of a 50% potassium hydroxide solution at elevated temperature. Under nitrogen atmosphere
9.3 g (60 mmol) of distilled diethylsulfate (Fluka, purum, >99%) were added then in drops. After
refluxing for another one hour the solution was cooled for two hours. The precipitated product was
filtered off, washed with cold argon saturated water, and dried. Additionally the filtrate was extracted
with diethyl ether. After drying with anhydrous sodium sulfate (Merck, p.a.) the solvent was removed
and the residue was purified together with the precipated product by vacuum sublimation to yield
1
bright white crystals. Mp 50ꢂC; H NMR (C6D6): ꢂ ¼ 0.99 (t, 12H, JH–H ¼ 7.0 Hz), 3.05 (q, 8H,
JH–H ¼ 7.1 Hz), 6.80 (s, 4H) ppm; 13C NMR (C6D6): ꢂ ¼ 13.3, 46.4, 117.7, 141.9 ppm; UV-VIS
(cyclohexane): ꢆmax. (log ") ¼ 269 (4.161), 321 (shoulder, 3.279) nm.
N,N,N0,N0-Tetra-n-propyl-p-phenylenediamine (TPPPD) [39]
Freshly sublimated p-phenylenediamine (4.0g, 37mmol, Merck, 99%) and 85.0g (500mmol) of
distilled n-propyliodide (Fluka, purum, >98%) were dissolved in 25 cm3 of benzene. Then 7.0 g
(180 mmol) of sodium amide (Fluka, pract.) were added carefully by small amounts to this solution
over a 1 hour period under argon atmosphere. The suspension was stirred for another 2 hours at room
temperature and subsequently refluxed for 24 hours. At this stage the course of the reaction could be
controlled by applying a simple oxidation reaction to a small sample of the reaction mixture: A few
drops of a methanolic iodine solution added to a diluted solution of the sample in methanol should
generate the stable radical cation of the product which is easily to be identified by its characteristic

534
G. Grampp et al.
UV-VIS-spectrum (ꢆmax. (methanol) ¼ 569, 619 nm). After passing this test positively the solvent and
the excess of n-propyliodide was removed completely. The dried residue was extracted in a Soxhlet
extractor with 300 ml of peroxide free diethylether for 4 hours. The solution obtained was washed with
argon-saturated water and dried with anhydrous sodium sulfate (p.a.). Thereafter the solution was
filtered through a layer of neutral aluminum oxide (Fluka, chromatography type 507 C neutral,
100–125 mesh) to remove the coloured side products. Repeated packed column rectification of
the crude product in vacuo under nitrogen atmosphere yielded a slightly yellowish viscous oil. Bp
1
150ꢂC (1mbar); H NMR (C6D6): ꢂ ¼ 0.82 (t, 12H, JH–H ¼ 7.5Hz), 1.50 (hex, 8H), 3.02 (t, 8H,
JH–H ¼ 7.1 Hz), 6.80 (s, 4H) ppm; 13C NMR (C6D6): ꢂ ¼ 12.1, 21.5, 55.2, 117.4, 142.3 ppm; UV-
VIS (cyclohexane): ꢆmax. (log ") ¼ 269 (4.190), 321 (shoulder, 3.267) nm.
N,N,N0,N0-Tetra-isopropyl-p-phenylenediamine (TiPPPD)
Freshly sublimated p-phenylenediamine (4.0g, 37mmol, Merck, 99%) and 85.0 cm3 (500 mmol) of
redistilled 2-iodopropane (Aldrich, 99%) were dissolved in 25cm3 of benzene. Then 7.0 g (180 mmol)
of sodium amide (Fluka, pract.) were added carefully and by small amounts to this solution over a 1
hour period under nitrogen atmosphere. The suspension was stirred for 2 hours at room temperature
and subsequently refluxed for 48 hours performing the oxidation test from time to time. Realizing that
even after such a long time the alkylation process was not completed but came to an end the reaction
was stopped. The solvent and the excess of 2-iodopropane were removed completely. The dried residue
was extracted in a Soxhlet extractor with 300 cm3 of peroxide free diethyl ether for 4 hours. The
solution obtained was washed with argon-saturated water and dried with anhydrous sodium sulfate.
The solution was filtered through a layer of neutral aluminum oxide to remove the solvent as well as
coloured side products. Packed column rectification of the crude product in vacuo under nitrogene
atmosphere followed by chromatography on basic aluminum oxide (Fluka, chromatography, type 5016
A basic) using benzene as eluent and controlling the fraction composition by UV-VIS-spectroscopy
of the oxidized semiquinonoid radical cation (ꢆmax. (methanol)¼ 574, 617nm) finally resulted in
slightly yellowish crystals. Mp 34.5ꢂC; 1H NMR (C6D6): ꢂ ¼ 1.06 (d, 24H, JH–H ¼ 6.4 Hz), 3.50 (sept.,
4H, JH–H ¼ 6.4 Hz), 6.96 (s, 4H) ppm; 13C NMR (C6D6): ꢂ ¼ 22.1, 48.9, 125.2, 142.4 ppm; UV-VIS
(cyclohexane): ꢆmax. (log ") ¼ 270 (4.093), 314 (shoulder, 3.068) nm.
N,N,N0,N0-Tetrabenzyl-p-phenylenediamine (TBzPPD)
A mixture of 3.0 g (28 mmol) of freshly sublimated p-phenylenediamine (Merck, 99%), 12.0g
(113 mmol) of dried anhydrous sodium carbonate (Fluka, puriss. p.a., >99.5%) and 88.0g (695 mmol)
of benzylchloride (Fluka, >99%) was refluxed for 12 hours in a nitrogen atmosphere. An oxidation
test (s.a.) was performed to indicate the end of reaction (ꢆmax (methanol) ¼ 570, 619 nm). Excessive
benzylchloride was removed completely under reduced pressure and the remaining highly viscous
residue was extracted in a Soxhlet extractor with 300 cm3 of tetrahydrofurane. The extract was con-
centrated and cooled down. The yellow needles precipitated were filtered off and washed carefully with
little portions of ice-cold tetrahydrofurane. Purification could be achieved by recrystallization
1
from benzene. Mp 149ꢂC; H NMR (C6D6): ꢂ ¼ 4.27 (s, 8H), 6.59 (s, 4H), 7.09 (m, 20H) ppm;
13C NMR (CCl4=D2O): ꢂ ¼ 56.0, 116.0, 127.8, 129.3, 140.0, 142.8ppm; UV-VIS (cyclohexane):
ꢆmax. (log ") ¼ 265 (4.262), 339 (3.415) nm.
1,4-Dipyrrolidinobenzene (DPB) [36]
Freshly sublimated p-phenylenediamine (6.1g, 56mmol, Merck, 99%) and 35.0g (113mmol) of
freshly distilled 1,4-diiodobutane were dissolved in 75cm3 of tetrahydrofurane. 24.4 g (230 mmol)
of dried anhydrous sodium carbonate (p.a.) were added then and the mixture was refluxed for 1 hour

Various Wurster’s Radical Cations
535
under nitrogen atmosphere. After passing the oxidation test successfully (ꢆmax. (methanol) ¼ 524, 565,
616 nm) the solvent was removed completely and the dried residue was dissolved in argon-saturated
semi-concentrated hydrochloric acid. The hydrochloric acid solution was neutralized carefully with
argon-saturated ammonium hydroxide and the precipitated crude product was collected, washed with
cold water, and dried. Subsequently the solid substance was dissolved again in tetrahydrofurane
refluxed together with a little amount of activated charcoal (Merck, p.a.) under a nitrogen atmosphere
and finally hot filtered through a layer of basic aluminum oxide (s.a.) to remove coloured side products.
After removal of the solvent under reduced pressure the obtained product was purified by vacuum
sublimation to yield bright yellowish needles. Mp 151ꢂC (dec.); 1H NMR (C6D6): ꢂ ¼ 1.61 (quin., 8H,
JH–H ¼ 3.3 Hz), 3.08 (m, 8H), 6.74 (s, 4H) ppm; 13C NMR (C6D6): ꢂ ¼ 25.8, 48.9, 114.3, 141.5 ppm;
UV-VIS (cyclohexane): ꢆmax. (log ") ¼ 270 (4.520), 342 (3.699) nm.
N,N0-Bis[4-(dimethylamino)phenyl]piperazine (BDMAP) [24]
The synthesis of this compound followed the procedure recently described by Nelsen et al. [24].
Radical Generation
For CW-ESR and ENDOR measurements concentrations in the range of 10ꢁ3–10ꢁ4 M for the various
diamines in a 2-propanol=DMSO (dimethylsulfoxide) mixture (1:1v) were used after argon had been
bubbled through them for 15 minutes. After addition of a trace of dry silver perchlorate the cation
radical emerges spontaneously (Berlin group). The Graz group used methanolic iodine solution
(5ꢅ 10ꢁ5 M) to oxidize the corresponding p-phenylenediamines (1 ꢅ 10ꢁ4 M) in methanol. Oxidation
with iodine in a 3:1 v mixture of glycerol-water was performed by the Lisbon group for ENDOR-
spectra. The CW-ESR spectra were measured in water, there.
It is well established since the early work of Michaelis and Granick [40], that there exist two well
defined oxidation steps for p-phenylenediamines. The neutral tetra-substituted p-phenylenediamines
(R) are oxidized by one-electron steps to the corresponding quinoid radical cations (Sþꢃ ) and further-
more to the doubly charged diamagnetic þquinonediimines (T2þ). The synproportionation equilibrium
shown in Eq. (2) is strongly shifted to (Sꢃ ) in aprotic solvents.
2Sþꢃ Ð R þ T2þ
This reaction scheme was later extended to Weitz and Wurster type radicals by H€unig et al. [41].
ð2Þ
Acknowledgement
€
G. Grampp and SL would like to thank the Austrian Academic Exchange Service (OAD), the
Volkswagen-Foundation (Germany), and the Austrian Federal Ministery of Education, Science, and
Culture for financial support. ERASMUS-SOCRATES scholarships from the EC are gratefully
acknowledged by JPT and SL. JPT and AJSCV thank for financial support from FCT through its
´
ꢀ
Centro Processos Quımicos da Universidade Tecnica de Lisboa. SL would also like to thank the
DFG (Deutsche Forschungsgemeinschaft) for financial support. AK would gratefully like to thank
Matthias Hennemann at the Computer-Chemie-Centrum, University of Erlangen (Germany) for cal-
culating the AM1 geometries of TEPPD and TPPPD. The authors wish to thank G. Gescheidt for
helpful discussions.
References
[1] Wurster C, Sendtner R (1879) Ber Dtsch Chem Ges 12: 1803
[2] Weitz E, Fischer K (1925) Angew Chem 38: 1100
[3] James TH (ed) (1977) The Theory of the Photographic Process. New York

536
G. Grampp et al.: Various Wurster’s Radical Cations
[4] Weissman SI, Townsend J, Paul DE, Pake GE (1953) J Chem Phys 21: 2227
[5] Chu TL, Pake GE, Paul DE, Townsend J, Weissman SI (1953) J Phys Chem 57: 504
[6] Lipkis D, Paul DE, Townsend J, Weissman SI (1953) Science 117: 534
[7] Hausser KH (1960) Naturwiss 47: 152
[8] Hausser KH (1963) Mol Phys 7: 195
[9] Bolton JR, Carrington A, dos Santos-Veiga J (1962) Mol Phys 5: 615
[10] Weil JA, Bolton JR, Wertz JE (1994) Electron Paramagnetic Resonance. Wiley, New York, p 512
[11] Knolle WR (1970) PhD Thesis, University of Minnesota, Minneapolis, MN
[12] Grampp G, Stiegler G (1984) Z Phys Chem NF 141: 185
[13] Weissman SI (1954) J Chem Phys 22: 1135
[14] McLachlan AD (1958) Mol Phys 1: 233
[15] Latta BM, Taft RW (1967) J Am Chem Soc 89: 5172
[16] Yao T, Musha S (1974) Chem Lett 939
[17] Allendoerfer RD, Rieger PH (1966) J Am Chem Soc 88: 3711
[18] Bullock AT, Howard CB (1977) J Magn Res 25: 47
[19] Horsman G (1967) Coll Ampere XIV: 578
[20] Zhang F, Zhonli G, Liu Chem Y (1988) ꢂJ Chinese Univ 4: 24
´
[21] Telo JP, Dias RMB, Vieira AJSC (1995) 2 Congresso de radicais livres em Quımica, Biologia e
Medicina. Porto, Portugal
[22] Truchin MI (1976) Leningradskij institut jadernoj fizihi, B.P. Kostantinova (translated by
Technical Information Libery, Hannover=Germany) 273: 2; Truchin MI (1976) Academy of
Sciences of USSR, Leningrad, Institute of Nuclear Physics 273: 2
[23] Marken F, Webster RD, Bull SD, Davies SG (1997) J Electroanal Chem 437: 209
[24] Nelsen SF, Yunta MJR (1994) J Phys Org Chem 7: 55
[25] Brouwers AM (1997) J Phys Chem A 101: 3626
[26] Rauhut G, Clark T (1993) J Am Chem Soc 115: 9127
[27] McKinney TM, Geske DH (1967) J Am Chem Soc 89: 2806
€
[28] Bock H, Vanpel T, Nather C, Ruppert K, Havlas Z (1992) Angew Chem 104: 348
[29] Kurreck H, Kirste B, Lubitz W (1984) Angew Chem 96: 171
[30] Kurreck H, Kirste B, Lubitz W (1988) Electron Nuclear Double Resonance Spectroscopy of
Radicals in Solution. VCH, Weinheim
[31] Batra R, Giese B, Spichty M, Gescheidt G, Houk KN (1996) J Phys Chem 100: 18371
[32] Clark T, Alex A, Beck B, Chandrasekhar J, Gedeck P, Horn A, Hutter M, Rauhut G, Sauer W,
Steinke T (1998) F90VAMP, Version 7.0 alpha, Erlangen
[33] Frisch ME, Trucks GW, Schlegel HB, Gill PMW, Johnson BG, Robb MA, Cheeseman JR,
Keith T, Petersson GA, Montgomery JA, Raghavachari K, Al-Laham MA, Zakrzewski VG,
Ortiz JV, Foresman JB, Cioslowski J, Stefanov BB, Nanayakkara A, Challacombe M, Peng CY,
Ayala PY, Chen W, Wong MW, Andres JL, Binkley JS, Defrees DJ, Baker J, Stewart JP,
Head-Gordon M, Gonzalez C, Pople JA (1995) Gaussian 94, Revision E.1, Gaussian, Inc.,
Pittburgh PA
[34] Schlegel HB (1995) Exploring Chemistry with Electronic Structure Methods. Gaussian, Inc.,
Pittsburgh PA
€
[35] Kaupp M, University of Wurzburg, Germany (personal communication)
[36] Vieira AJSC, Universidade de Nova Lisboa, Portugal (personal communication)
[37] Grampp G (1985) Rev Sci Instrum 56: 2050
[38] Grampp G, Schiller CA (1981) Anal Chem 53: 560
[39] Kirste B (1992) Anal Chim Acta 365: 191
[40] Michaelis L, Granick S (1943) J Am Chem Soc 65: 1747
[41] Deuchert K, H€unig S (1978) Angew Chem 90: 927
[42] Grampp G, Pluschke P (1987) Coll Czech Chem Commun 52: 819