PELDOR measurements on a nitroxide-labeled Cu(II) porphyrin: Orientation selection, spin-density distribution, and conformational flexibility

Article abstract of DOI:10.1021/jp710504k

Metal ions are functionally or structurally important centers in metalloproteins or RNAs, which makes them interesting targets for spectroscopic investigations. In combination with site-directed spin labeling, pulsed electron-electron double resonance (PELDOR or DEER) could be a well-suited method to characterize and localize them. Here, we report on the synthesis, full characterization, and PELDOR study of a copper(II) porphyrin/nitroxide model system. The X-band PELDOR time traces contain besides the distance information a convolution of orientational selectivity, conformational flexibility, exchange coupling, and spin density distribution, which can be deconvoluted by experiments with different frequency offsets and simulations. The simulations are based on the known experimental and spin Hamiltonian parameters and make use of a geometric model as employed for structurally similar bis-nitroxides and spin density parameters as obtained from density functional theory calculations. It is found that orientation selection with respect to dipolar angles is only weakly resolvable at X-band frequencies due to the large nitrogen hyperfine coupling of the copper porphyrin. On the other hand, the PELDOR time traces reveal a much faster oscillation damping than observed for structurally similar bis-nitroxides, which is mainly assigned to a small distribution in exchange couplings J. Taking the effects of orientation selectivity, distribution in J, and spin density distribution into account leads finally to a narrow distance distribution caused solely by the flexibility of the structure, which is in agreement with distributions from known bis-nitroxides of similar structure. Thus, X-band PELDOR measurements at different frequency offsets in combination with explicit time trace simulations allow for distinguishing between structural models and quantitative interpretation of copper-nitroxide PELDOR data gives access to localization of copper(II) ions.

Full text of DOI:10.1021/jp710504k

5064
J. Phys. Chem. A 2008, 112, 5064–5073
PELDOR Measurements on a Nitroxide-Labeled Cu(II) Porphyrin: Orientation Selection,
Spin-Density Distribution, and Conformational Flexibility
Bela E. Bode, Jörn Plackmeyer, Thomas F. Prisner, and Olav Schiemann*,†
Institute of Physical and Theoretical Chemistry and Center for Biomolecular Magnetic Resonance,
Max-Von-Laue-Strasse 7, J. W. Goethe-UniVersity, 60438 Frankfurt am Main, Germany
ReceiVed: October 31, 2007; ReVised Manuscript ReceiVed: March 7, 2008
Metal ions are functionally or structurally important centers in metalloproteins or RNAs, which makes them
interesting targets for spectroscopic investigations. In combination with site-directed spin labeling, pulsed
electron-electron double resonance (PELDOR or DEER) could be a well-suited method to characterize and
localize them. Here, we report on the synthesis, full characterization, and PELDOR study of a copper(II)
porphyrin/nitroxide model system. The X-band PELDOR time traces contain besides the distance information
a convolution of orientational selectivity, conformational flexibility, exchange coupling, and spin density
distribution, which can be deconvoluted by experiments with different frequency offsets and simulations.
The simulations are based on the known experimental and spin Hamiltonian parameters and make use of a
geometric model as employed for structurally similar bis-nitroxides and spin density parameters as obtained
from density functional theory calculations. It is found that orientation selection with respect to dipolar angles
is only weakly resolvable at X-band frequencies due to the large nitrogen hyperfine coupling of the copper
porphyrin. On the other hand, the PELDOR time traces reveal a much faster oscillation damping than observed
for structurally similar bis-nitroxides, which is mainly assigned to a small distribution in exchange couplings
J. Taking the effects of orientation selectivity, distribution in J, and spin density distribution into account
leads finally to a narrow distance distribution caused solely by the flexibility of the structure, which is in
agreement with distributions from known bis-nitroxides of similar structure. Thus, X-band PELDOR
measurements at different frequency offsets in combination with explicit time trace simulations allow for
distinguishing between structural models and quantitative interpretation of copper-nitroxide PELDOR data
gives access to localization of copper(II) ions.
Introduction
centers10 and paramagnetic organic cofactors,11 while exceeding
the CW EPR limit (<2 nm) by up to 6 nm, is called pulsed
electron-electron double resonance (PELDOR or DEER).12
Another method for measuring dipolar couplings, which is less
frequently applied, is double quantum coherence (DQC) EPR.13
Yet, PELDOR distance measurements between a copper(II) ion
and a nitroxide have only been reported once for an in situ
generated model compound;14 that PELDOR measurements
involving Cu2+ are feasible has been shown in three reports
concerning copper-copper distances.15 Nonetheless, general
questions remain with respect to the influence of orientation
selection,15c,16 conformational flexibility,15c,17 and spin density
distribution.10b,c
Here, we introduce a newly synthesized and fully character-
ized asymmetric copper(II)porphyrin-nitroxide model system
and show that a detailed analysis of the PELDOR time domain
data yields, even in the presence of strong spin delocalization,
the precise copper-nitroxide distance and the conformational
flexibility of the molecule. In addition, we discuss the influence
of orientation selection and exchange coupling based on
simulations of the experimental data.
Metal ions are crucial for the structure and function of
catalytically active metalloproteins1 and RNAs.2 To characterize
the role of individual metal ions, biophysical methods probing
the local structure of these ions and their position within the
global fold of the biopolymer in solution are needed. Nuclear
magnetic resonance (NMR) spectroscopy is frequently applied
for such purposes, but reaches its limits in cases of large
biomacromolecular complexes or in the case of paramagnetic
metal centers like copper(II), manganese(II), molybdenum(V),
or iron(III).3 However, paramagnetic centers or the exchange
of diamagnetic ions by paramagnetic ones enables the applica-
tion of continuous wave (CW) and pulsed electron paramagnetic
resonance (EPR) techniques,4 which were proven to yield
detailed and precise information about local structures of metal
ion binding pockets in proteins5 and RNAs.6 EPR methods have
also been shown to yield reliable results with respect to locating
ions in relation to other metal centers or organic cofactors.7 In
a similar approach, paramagnetic metal ions may also be
localized via distance measurements to artificially and site-
directedly attached nitroxides; attempts based on relaxation
measurements have been reported.8 A pulsed EPR method that
has been demonstrated to be suitable and straightforward for
measuring nanometer distances between nitroxides,9 metal
Methods and Materials
Synthesis. All reactions were performed with exclusion of
air under argon employing standard Schlenk techniques. Reagent-
grade solvents and chemicals were used without further
purification, except where stated otherwise. Dry solvents were
purchased from Fluka or ACROS and thoroughly degassed prior
* To whom correspondence should be addressed. E-mail: os11@
st-andrews.ac.uk.
† Current address: Center of Biomolecular Science, School of Biology,
University of St. Andrews, St. Andrews, Fife, KY16 9UA, Scotland.
10.1021/jp710504k CCC: $40.75
2008 American Chemical Society
Published on Web 05/20/2008

PELDOR Measurements on Nitroxide-Labeled Cu(II) Porphyrin
J. Phys. Chem. A, Vol. 112, No. 23, 2008 5065
to use. Triethylamine was freshly distilled from CaH2. Solvents
and reagents for Sonogashira cross-couplings were degassed by
freeze-pump-thaw cycles. Octaethylporphyrin (H2(OEP)) and
copper(II)-octaethylporphyrin (Cu(OEP)) were purchased from
Aldrich and 4-hydroxy-4′-iodobiphenyl from ABCR. 1-Oxyl-
2,2,5,5-tetramethylpyrrolin-3-carboxylic acid (TPC) 1 was
prepared from 2,2,6,6-tetramethyl-4-oxopiperidine by modified
procedures of Hideg et al.18 TLC: SiO2 60 F254 and Al2O3 60
F254 neutral both from Merck. Column chromatography: SiO2
60 (70-230 mesh) from Aldrich or Al2O3 MP Alumina
N-Super I from MP Biomedicals.
Figure 1. Pulse sequences for 4-pulse PELDOR.
Analytic observations relied on elemental analysis and mass
spectrometry such as electron ionization (EI), electrospray
ionization (ESI), and MALDI. EI mass spectra were recorded
on a CH7A spectrometer from MAT, ESI mass spectra were
recorded on a LCQ Classic spectrometer from Thermo Electron,
and MALDI mass spectra were recorded on a Voyager DE-Pro
or STR spectrometer, both from Applied Biosystems. Proton
NMR spectra were acquired of diamagnetic molecules at 250
MHz on a Bruker AM-250 spectrometer and calibrated using
residual nondeuterated solvents as internal standard (δ CHCl3
) 7.240). Elementary analysis was performed on a Foss-Heraeus
CHN-O-Rapid, and UV-vis spectra were recorded on an
Agilent 8453 Spectrophotometer. Graphical representations of
the crystal structure were produced using the WinGX software.19
The Ni(OEP)20 and Ni(meso-formyl-OEP) derivatives of
H2(OEP) were prepared according to literature procedures.21
Ni(meso-ethynyl-OEP) and Cu(meso-ethynyl-OEP) 2 were
prepared according to ref 22 from a mixture of the corresponding
meso-E/Z-2-chloroethenyl-OEP-compounds instead of the bro-
moethenyl derivatives (see Supporting Information) and used
as obtained due to the products susceptibility to spontaneous
oxidative dimerization.23
the solvents were removed in vacuo, and the solid residue was
redissolved in dichloromethane. The organic phase was washed
three times with water and dried with sodium sulfate. After the
solvent was removed in vacuo, the residue was redissolved in
dicholoromethane/hexane (2:1) and purified via column chro-
matography (3 × 20 cm, Al2O3, 4% H2O, dicholoromethane/
hexanes (2:1)). A dark green band was eluted first, followed by
a red-brown fraction containing unreacted 3 and the desired
product. This main fraction was stripped from solvent, and the
residue was recrystallized from chloroform layered with metha-
nol. The product was obtained as red-violet leaflets and needles
after filtration, washing with methanol, and drying in vacuo.
Yield: 35 mg (0.037 mmol, 31%). Single crystals suitable for
X-ray diffraction were obtained from a benzene solution layered
with petroleum ether. Anal. Calcd for C59H64N5O3Cu: C, 74.22;
H, 6.76; N, 7.34. Found: C, 74.07; H, 6.64; N, 7.28. UV-vis
(benzene) λmax (log ε): 426 nm (5.448), 554 (4.289), 594 (4.284).
MALDI-TOF-MS (matrix, DHB) m/z calcd for C59H64N5O3Cu:
954.735. Found: 954.9 (M+, 100%), 894.0 ((M - 4 × CH3)+,
33%).
PELDOR Theory. All following PELDOR experiments were
performed with the 4-pulse sequence shown in Figure 1.
The pulses at the detection frequency νA create an echo from
the spins on resonance, named in the following A spins (the
unpaired electron centered at the Cu(II) porphyrin in this study).
Introduction of an inversion pulse at the pump frequency νB
flips spins resonant with this second frequency, here defined as
B spins (here the electron spin centered at the nitroxide). The
dipole-dipole splitting ωdd, between A and B spins causes a
shift in the Larmor frequency of the A spins by (ωdd. Therefore,
after pumping the B spins the A spins accumulate a phase shift
ωddt, where t defines the time delay of the pump pulse. The
resulting PELDOR signal V(t) is a product of two contributions
(eq 1)
(1-Oxyl-2,2,5,5-tetramethyl-pyrrolin-3-carboxylic acid 4′-
iodobiphenyl-4-yl-ester) (3). 1,3-dicyclohexylcarbodiimide (DCC)
(1.671 g, 8.09 mmol) was added in one portion to a solution of
TPC 1 (746 mg, 4.05 mmol), dimethylaminopyridine (DMAP)
(900 mg, 8.10 mmol), and 4-hydroxy-4′-iodobiphenyl (800 mg,
2.70 mmol) in 20 mL of dry tetrahydrofuran (THF). After a
yellow precipitate had formed, 10–12 mL of dry THF was added
to facilitate stirring. The reaction was allowed to continue at
room temperature over 3 d in the dark. All solids were removed
by filtration and washed with a small amount of THF and then
thoroughly washed with diethyl ether (150 mL). The yellow
filtrate was evaporated to dryness, and the residue redissolved
in diethyl ether. The ether solution was extracted with water
and brine. After drying over sodium sulfate, the diethyl ether
was removed and the yellow residue redissolved in diethyl ether/
hexane 2:1 and was subjected to column chromatography
(Al2O3, 4% H2O, 4 × 18 cm). A bright yellow fraction was
eluted first, directly followed by the broad yellow main band
of the product. After removal of the solvent, a yellow solid was
obtained (1.004 g, 2.17 mmol, 80%). Anal. Calcd for
C21H21INO3: C, 54.56; H, 4.58; N, 3.03. Found: C, 54.61; H,
4.63; N, 2.87. ESI-MS (+) m/z calcd for C21H21INO3, 462.3;
found, 462.8 (M+, 100%).
V(t) ) Vintra(t)Vinter(t)
(1)
Vintra(t) describes the intramolecular contribution, whereas
Vinter(t) takes into account the signal decay caused by a
distribution of spin centers in the sample. In case of a
homogeneous distribution Vinter(t) is a monoexponential decay.24
For a single orientation (so) of the magnetic field B0, the
intramolecular contribution Vintra(t)so can be described by eq 217b
Vintra(t)so ) V0(1 - λ[1 - cos(ωddt)])
(2)
In eq 2, it is assumed that the spin-orientations are not
changed due to spin diffusion or spin-lattice relaxation and that
the exchange coupling J is negligible versus ωdd. V0 is the A
spin echo intensity at t ) 0, and λ is the fraction of B spins,
which are coupled to the detected A spin and inverted by the
pump pulse. Both, V0 and λ are dependent on the polar angles
of the magnetic field vector ꢀ and φ, and λ also depends on the
set of Euler angles Ω describing the orientation of spin B in
the axis system of spin A. In the point-dipole approximation
(Cu(OEP)-TPC-ester) (4). To a solution of 3 (55 mg, 0.12
mmol) and PdCl2(PPh3)2 (12.6 mg, 0.018 mmol) in 9 mL of
NEt3 and 5 mL of THF was added first CuI (12 mg, 0.06 mmol)
and then a mixture of 74 mg (0.12 mmol) of Cu(meso-ethynyl-
OEP) 2 and 12 mg (0.05 mmol) of PPh3 in 4 mL of NEt3 and
3 mL of THF. The resulting red-brown solution was stirred at
room temperature for 16 h with exclusion of light and finally
stirred at 50 °C for 1.5 h. After cooling to room temperature,

5066 J. Phys. Chem. A, Vol. 112, No. 23, 2008
Bode et al.
r2mnδij - 3rmni mnj
µ0µ2
Bg g
r
and for small g-anisotropies, such that the spin states are
quantized parallel to the external field, and in the weak coupling
regime, ωdd is given by eq 3
dij ) -
F F
m n
(8)
B∑∑
A
r5mn
4πp
m
n
where δij is the Kronecker delta, m and n are the spin bearing
atoms at centers A and B, respectively, F is their respective
spin density, rmn is the respective interatomic distances, and rmni
and rmnj are the i and j components of these interatomic distance
vectors in the A spin molecular frame. The sums extend over
all point-charges that constitute the interacting spins. In this
logical extension of the point-dipole the g-tensors are still
approximated to be isotropic.
PELDOR Simulations. Simulations were performed based
on eqs 2–4 and 6, with detection and pumping efficiencies
explicitly calculated for each orientation, each hyperfine state
of the nitroxide nitrogen, and each hyperfine state of the copper
nucleus. The nitrogen hyperfine coupling of the porphyrin ring
was assumed to be isotropic.29 The respective resonance
positions are computed based on the g- and hyperfine tensors
of the free nitroxide30 and Cu(OEP)29,31 and the experimental
values for pulse lengths and microwave frequencies. A confor-
mational ensemble typically containing 1000 different conform-
µ0µ2B gAgB
ωdd ) -
3(3 cos2 θ - 1)
(3)
4πp
|rAB
|
In eq 3, gA and gB are the g-values of the spins A and B,
respectively, p is the Planck constant divided by 2π, rAB is the
distance vector connecting the spins, θ is the angle between
rAB and the external magnetic field, and µ0 is the vacuum
permeability.
For disordered powder samples, the observed echo intensity
is obtained by integration over all magnetic field orientations:
Vintra(t) ) ∫∫V0(1 - λ[1 - cos(ωddt)])sin(φ)dφdꢀ (4)
The relative orientation of spin A, spin B and the vector rAB
will all be random, if angular correlations can be neglected. In
this case, the integration over the Euler angles effectively
averages the magnetic field orientation dependence of the
modulation depth parameter λ,25 which then allows converting
the double integration over different magnetic field orientations
in eq 4 into an integration over all dipolar angles θ
ers was generated, each characterized by a distance vector rAB
,
with polar angles (ψ, η) in the axis system of spin A and Euler
angles Ω describing the mutual orientation of spin center B with
respect to A. Full rotational freedom about the acetylene and
ester linkers together with a single, harmonic backbone bending
mode with a normal distribution of (15° (standard deviation)
was assumed. For each of these conformers, the resonance
positions of spins A and B were calculated for all orientations
of the magnetic field vector B0 in the molecular axis frame of
spin A (typically 20 000 orientations). Hyperfine and g-tensor
axes were considered to be collinear to the molecular axis
system. The resonance frequencies of spin B can be calculated
after describing hyperfine and g-tensor of spin B in the
coordinate system of spin A. Additionally, an inhomogeneous
line width has been taken into account by a Gaussian distribution
to calculate the final resonance frequencies for spin A and B,
respectively. Lorentzian excitation profiles were calculated from
the experimental pulse lengths, which empirically describe the
excitation profiles of rectangular pulses with a sinusoidal B1
distribution over the cavity. This allows to calculate the
excitation functions of spin A and B, V0 and λ, and thereafter
the dipolar distribution function P(θ) for each conformer. The
dipolar coupling constant was computed for each orientation
of the magnetic field according to eq 3, determining gA and gB
for the given field orientation from the copper and nitroxide
g-tensors respectively. If the field strength (B1) of the detection
and pump pulses (approximated via their pulse lengths tPA and
tPB, respectively) is not much larger than the coupling strength,
λ will become a function of ωdd, which can be approximated
by eq 9.32
π⁄2
Vintra(t) ) V 1 - λ 1 - cos(ω t)sin θdθ
(5)
∫
0
dd
(
)
[
]
0
Within these boundaries, the coupling between spins can be
deduced parameter-free from the time domain signal by division
of V(t) by Vinter(t) and cosine Fourier transformation or inversion
of time or frequency domain data to the distance domain.
Because these inversions is generally ill-posed usually regular-
ization methods are employed such as Tikhonov regularization.26
If angular correlations between the spins exist, this simplifica-
tion is not valid.12b,27 The pump efficiency λ will depend on
the mutual orientation of the two radicals A and B. Additionally,
the orientation selectivity of the excitation of spin A described
by V0 in combination with the orientation selectivity of the pump
pulse will lead to a distribution function P(θ) of dipolar angles
that differs from the sin(θ) distribution of uncorrelated spin
centers. The echo signal intensity Vintra(t) for a given pump and
probe frequency can in such general case be described by12b,25
π⁄2
Vintra(t) ) V0 1 -
P(θ)[1 - cos(ω t)]dθ
(6)
∫
dd
(
)
[
]
0
with
P(θ) ) λ(θ)sin θ
(7)
In these cases data inversion based on eq 5 might lead to
erroneous results and inversions based on eq 6 have not yet
been reported. Spin centers exhibiting large g- or hyperfine
anisotropies and thus broad spectra, such as copper, might
pronounce this effect due to high spectral selectivity of the
pulses. We have therefore chosen to simulate the experimental
Cu2+/nitroxide PELDOR time traces from a conformational
ensemble of molecules created by a simple geometrical model.
In case of extensive spin density delocalization the point-
dipole approximation breaks down, and the dipolar coupling
has to be calculated treating all spin-bearing atoms. Each point-
dipole is broken up into point charges to reflect the delocaliza-
tion of the spins. In this model, elements of the interaction tensor
are given by eq 828
λ(ωdd) ) λ(θ)exp(-4ω2dd · tP2A)exp(-4ωd2d · tP2B) (9)
The final PELDOR signal for a given frequency offset ∆νAB
between detection and pumping frequency is the sum over all
conformers and all magnetic field orientations (typically 1000
× 20 000).
Accounting for delocalization, the dipolar tensor was calcu-
lated according to eq 8, which was then multiplied with the
dipole moments of the two spins. To take g-tensor anisotropy
into account the dipole moments were calculated by multiplica-
tion of the respective g-tensors with the electron spin angular
momentum using the high field approximation. The dipolar

PELDOR Measurements on Nitroxide-Labeled Cu(II) Porphyrin
J. Phys. Chem. A, Vol. 112, No. 23, 2008 5067
coupling was then computed from this as the energy difference
between the two manifolds of the B spin.
SCHEME 1: Synthesis of Model Compound 4
All simulations were performed with a home-written
MATLAB program. More details regarding the program and
structure generation procedure are described elsewhere.17b
Tikhonov regularizations were performed with the program of
G. Jeschke32 and an optimized regularization parameter of 1.
PELDOR Measurements. Samples of all three molecules,
Cu(OEP), nitroxide 1 and biradical 4 were prepared as d8-toluene
(200 µM, 80 µL) solutions. Five equivalents of 1-methylimi-
dazole were added to both Cu2+-porphyrin samples to prevent
aggregation. The samples were degassed by several freeze-pump
cycles and frozen in liquid nitrogen prior to sealing the sample
tube. All PELDOR spectra were recorded on a Bruker ELEX-
SYS E580 pulsed X-band EPR spectrometer with a standard
flex line probe head housing a dielectric ring resonator (MD5
W1) equipped with a continuous flow helium cryostat (CF935)
and temperature control system (ITC 502), both from Oxford
instruments. The second microwave frequency was coupled into
the microwave bridge by a commercially available setup (E580-
400U) from Bruker. All pulses were amplified via a pulsed
traveling wave tube (TWT) amplifier (117X) from Applied
Systems Engineering. The resonator was over-coupled to a
quality factor Q of about 50. PELDOR experiments were
performed with the pulse sequence π/2(νA)-τ1-π(νA)-(τ1 + t)-
π(νB)-(τ2 - t)-π(νA)-τ2-echo. The detection pulses (νA) were
set to 32 ns for both π and π/2 pulses and applied at a frequency
226 MHz higher than the resonance frequency of the resonator.
The amplitude was chosen to optimize the refocused echo. The
π/2-pulse was phase-cycled to eliminate receiver offsets. To
achieve a frequency offset (∆νAB) of 226 MHz, the pump pulse
(νB) was set to 12 ns at the resonance frequency of the resonator.
For ∆νAB ) 603 MHz, νB was set to a frequency of 377 MHz
lower than the resonance frequency of the resonator and the
pump pulse length was set to 32 ns. The field was adjusted in
such a way that in either case the pump pulse is applied to the
maximum of the nitroxide spectrum, where it selects the central
mI ) 0 transition of Azz together with the mI ) 0, (1 transitions
of Axx and Ayy. The pulse amplitude was in both cases optimized
to maximize inversion of a Hahn-echo at the pump frequency.
All spectra were recorded at 20 K with an experiment repetition
time of 3 ms, a video amplifier bandwidth of 25 MHz and an
amplifier gain of 63 dB. τ1 was set to 400 ns and τ2 to 1200 ns.
Usually 150 scans were accumulated with 180 data points and
time increments ∆t of 8 ns giving an approximate measurement
time of 3.5 hours. Typically 500 scans were accumulated for
spectra with ∆νAB ) 603 MHz giving an approximate measure-
ment time of 10 hours. Proton modulation was suppressed by
addition of 8 spectra of variable τ1 with a ∆τ1 of 8 ns.33 For
comparison with simulations and data inversion the time traces
were divided by a mono-exponential decay and normalized to
the point t ) 0.
An ultrafine integration grid (58410 integration points per atom)
and standard SCF)Tight convergence criteria were applied. For
these calculations, a decontracted FII-A basis set of Arbuznikov
et al.43 (16s,13p,10d)/[11s,10p,10d] was used for Cu and the
Huzinaga-Kutzelnigg-type IGLO-II basis sets44 were used for
all other atoms: (5s,1p)/[3s,1p] for H and (9s,5p,1d)/[5s,4p,1d]
for C, O, and N.
Results and Discussion
Synthesis. Molecule 4 was chosen to be a suitable model
system for studying the effects of spin delocalization, orientation
selection and conformational flexibility on Cu2+/nitroxide
PELDOR measurements. The Cu(OEP) moiety resembles bind-
ing motives found in proteins, ensures considerable amount of
spin density distribution into the coordinated ligand, and the
interconnecting bridge allows for some degree of structural
flexibility. In addition, the synthesis allowed for the covalent
attachment of only one nitroxide group to the copper complex,
in contrast to the previously reported in situ-synthesis of a
copper-bis(terpyridyl) derivative.14 The two step synthesis of
the model compound is summarized in Scheme 1. In the initial
step, the nitroxide-carboxylic acid 1 was coupled to the precursor
bridge 4-hydroxy-4′-iodobiphenyl via esterification using DMAP
and DCC as accelerators. In the second step, the resulting 4′-
iodobiphenyl-ester 3 was attached to Cu(meso-ethynyl-OEP) 2
by means of Sonogashira cross-coupling yielding the desired
nitroxide-metallo-porphyrine 4 in an overall yield of 33%.
X-ray Structure. Crystals of 4 suitable for X-ray diffraction
could be grown from benzene.45 A graphical representation of
the molecular structure is shown in Figure 2 and selected bond
lenghts and angles are given in tables S1 to S3 in the Supporting
Information. The compound forms a triclinic crystalline system
and exhibits significant deviations of the porphyrinic ring from
Density Functional Theory (DFT) Calculations. The cal-
culations for 1 and 2 were performed using unrestricted Kohn-
Sham34 DFT methods as implemented in GAUSSIAN 0335 and
TURBOMOLE.36 All structure optimizations were carried out
with TURBOMOLE employing Becke’s exchange functional
B37 and Perdew’s P8638 correlation functional, together with
the TZVP basis set39 for all atoms and accelerated with the RI
approximation using the standard TZVP auxiliary basis set from
TURBOMOLE.40 For the computation of the Mulliken spin-
densities with GAUSSIAN 03, a combination was chosen of
Becke’s three-parameter hybrid exchange functional B341
together with the Perdew/Wang correlation functional PW91.42

5068 J. Phys. Chem. A, Vol. 112, No. 23, 2008
Bode et al.
Figure 2. Molecular structure of the nitroxide substituted Cu2+
porphyrin 4.
-
planarity (Supporting Information, Figure S2), as commonly
found in crystal structures of substituted porphyrins as for
example in the homo- and heterobimetallic ethene-linked
bisporphyrins reported by Smith et al.46 With respect to the
PELDOR simulations, the most interesting feature is the
intramolecular Cu2+-nitroxide distance, which amounts to
21.254(3) and 20.185(3) Å for Cu2+-O and Cu2+-N, respec-
tively. Because the spin density is roughly equally distributed
between the nitrogen and oxygen atom a mean distance of 20.7
Å can be extracted. The crystal structure also reveals that the
porphyrin moieties stack onto each other, forming a nearly
coplanar arrangement such that the next neighbour porphyrin
ring is 3.621 Å apart (Supporting Information, Figure S2). This
aggregation leads to the formation of “head-to-tail” structures
with respect to the meso-substituent, where the unsubstituted
meso-C(25) of one porphyrin is in close contact of 3.189 Å to
the Cu2+ ion of the next porphyrin ring. The shortest intermo-
lecular Cu-Cu distance amounts to 4.760 Å, whereas the next
intermolecular Cu-(nitroxide-O) distance is 8.050 Å.45 To avoid
such intermolecular interactions, which have been reported for
toluene solutions of Cu(OEP),31a 1-methylimidazole was added
in excess to solutions used for PELDOR studies.47
PELDOR Measurements. A 2-pulse field swept spectrum
of model compound 4 is depicted in Figure 3a. The PELDOR
experiments were performed applying the pump pulse on the
central line of the nitroxide (position A), to achieve large
modulation depths, whereas the detection pulses were applied
at higher frequency (positions B and C), to solely select spectral
contributions from the copper(II) ion. The PELDOR time-trace
recorded at detection position B (corresponding to a ∆νAB of
226 MHz) is depicted in Figure 3b together with the time trace
of a reference sample composed of an equimolar mixture of 1
and Cu(OEP). The PELDOR time trace of 4 exhibits three
clearly resolved periods of modulation, whereas the reference
measurement does not, proving that the modulation is only
caused by intramolecular spin-spin coupling and not arising
from intermolecular interactions or experimental artifacts.
The modulation depth λ of 0.40 is comparable to those found
in time traces of bis-nitroxide systems using identical conditions
for the pump pulse.27 However, the damping of the modulation
is much faster compared to structurally analogous bis-nitroxide
model compounds (see Supporting Information),17a,48 which can
be due to different reasons. First, the spin density distribution
in the porphyrin moiety might lead to a pronounced rhombicity
in the spin-spin interaction. Second, the spectral selection on
copper might correspond to a selection of distance vector
orientations that leads to rapidly interfering frequencies. Third,
the g-anisotropy of the copper-spin might cause a significant
orientation dependence of gA in eq 3, which leads to a
dependence of the coupling strength not only on the orientation
Figure 3. (a) Hahn-echo field swept spectrum of 4. The following
PELDOR experiments were performed with pumping at position A and
detecting at B or C for a ∆νAB of 226 or 603 MHz, respectively. (b)
Experimental 4-Pulse PELDOR time trace of 4 with ∆νAB ) 226 MHz
(black line) together with a time trace recorded under the same
conditions of a reference mixture of 1 and Cu(OEP) (1:1), shown in
gray. (c) The 4-Pulse PELDOR time trace with ∆νAB ) 603 MHz (black
line) together with a time trace recorded under the same conditions of
the same reference mixture, shown in gray.
of the distance vector but also on the orientation of the copper-
porphyrin moiety with respect to the external magnetic field.
Fourth, a high conformational flexibility of the porphyrin moiety.
A high conformational flexibility of the biphenylene linker and
the nitroxide spin-label in 4 is excluded, because this bridge is
the same as in structural analogous bis-nitroxide systems. Fifth,
an exchange coupling contribution J, which according to CW
EPR data and simulations, is smaller than 0.2 MHz in average
(see Supporting Information).
To investigate the effect of spectral selectivity on the
PELDOR time traces, spectra of 4 were recorded with different
frequency offsets ∆νAB; the one with the maximum offset of
∆νAB ) 603 MHz (position C) is shown in Figure 3c, together

PELDOR Measurements on Nitroxide-Labeled Cu(II) Porphyrin
J. Phys. Chem. A, Vol. 112, No. 23, 2008 5069
with the measurement on the reference sample under the same
conditions.49 The time trace shows three well-resolved modula-
tion periods that are not present in the reference sample.
Interestingly, slight changes in the shape of the modulation
pattern, especially in the first minimum, can be seen compared
to the measurement with ∆νAB ) 226 MHz, indicating different
orientation selections. The smaller λ of 0.23 compared to 0.40
at 226 MHz is due to the more selective π-pulse of 32 ns
compared to the 12 ns pulse length used at 226 MHz. The reason
for the longer pump pulse is the limited resonator bandwidth
and thus the limited B1 which does not allow shorter π-pulses
at this large frequency offset. Time traces recorded with an offset
of 226 MHz and 32 ns pump pulse length show similar
modulation depths (see Supporting Information).
To investigate the influence of the spectral selection on the
PELDOR spectra in more detail, the excitation efficiencies on
the copper ion and on the nitroxide have been calculated for
the pumping and detection frequencies corresponding to the
frequency offsets ∆νAB ) 226 MHz and ∆νAB ) 603 MHz.
The excitation profiles of the pulses were assumed to be
Lorentizan. Figure 4 shows the resulting normalized excitation
probabilities in dependence of the magnetic field orientation in
the principle axis system of the respective molecule. Whereas
all orientations of the nitroxide are pumped independent of the
pump pulse length of either 12 or 32 ns, the selection on copper
is more specific. Detecting at position B (corresponding to ∆νAB
) 226 MHz), mainly Cu(OEP)-moieties with the molecular x-
and y-axes (axes within the porphyrin plane) parallel to the
magnetic field axis are selected, whereas those with the z-axis
parallel to the field are not excited. In contrast, detecting at
position C (∆νAB ) 603 MHz), the probability of Cu(OEP)-
rings with the molecular z-axis parallel to the field increases,
whereas x and y are deselected. Because the g-tensor is collinear
to the molecular axis system, detecting at Β mainly selects gxx
and gyy, whereas detecting at C selects mainly noncanonical
orientations.
To estimate the effect of the different selections on the
PELDOR time-traces we calculated the form factors P(θ)
(Figure 5), which describe within the point-dipole approximation
the probabilities of excitation in dependence of the dipolar angle.
The calculation has been performed by connecting the axis
systems of the two spins and thus the two excitation profiles
via the molecular structure, which is known from X-ray
diffraction. To account for freezing-out a range of different
conformations adopted by 4 at room temperature, a geometric/
dynamic model was built taking into account the mean distance
of 20.7 Å between the copper and the nitroxide, a single bending
motion of 15° centered at the mid-point of the biphenyl bridge
and free rotation of the nitroxide group around the phenolic
bond on a cone of 31.4°. On the basis of this model, which has
been proven to be useful for structurally analogous bis-
nitroxides,17 a conformational ensemble typically containing
1000 different conformers was generated. The form factors,
calculated for these ensembles, show in either case significant
deviation from the sin(θ) probability distribution function of
uncorrelated spin centers. For ∆νAB ) 226 MHz, distance
vectors parallel to B0 (θ ) 0°) become more probable and those
perpendicular to B0 (θ ) 90°) less probable than expected for
a sin(θ) distribution. In case of ∆νAB ) 603 MHz the intensities
of both singularities are smaller than compared with the sin(θ)
function, whereas intermediate angles increase in probability.
Thus, both form factors differ significantly from each other close
to θ ) 0°. It is interesting to note that even though the form
factors are quite different, this is not translated into a strong
Figure 4. Orientation selection on copper with (a) ∆νAB ) 226 MHz,
tPA ) 32 ns and (c) ∆νAB ) 603 MHz, tPA ) 32 ns. The orientation
selection on the nitroxide is shown in (b) for ∆νAB ) 226 MHz, tPB
)
12 ns and in (d) for ∆νAB ) 603 MHz, tPB ) 32 ns. The molecular
axis systems are depicted in (e).
Figure 5. Calculated form factors P(θ) for the different frequency
offsets of 226 MHz (black curve) and 603 MHz (gray curve). A sin(θ)
distribution for uncorrelated centers is shown for comparison (dotted
line).
change in the experimental PELDOR time traces. It can,
however, be visualized by Fourier transformation (see Support-
ing Information). The weak effect may be attributed to both

5070 J. Phys. Chem. A, Vol. 112, No. 23, 2008
Bode et al.
Figure 6. Simulations of the PELDOR time domain data of 4 assuming
a Gaussian distribution of the copper nucleus in the porphyrin plane.
The experimental time trace with ∆νAB ) 226 MHz is shown in black
and the one with ∆νAB ) 603 MHz in red. Their simulations are overlaid
as dotted lines.
form factors having maxima at approximately 45° and still
strong intensity at θ ) 90°.
With the assumption that the damping is due to the flexibility
of the copper atom in the porphyrin moiety, the time traces have
been simulated employing the form factors as calculated above,
and the explicit distance vector has been extracted from the
geometric model. All calculated spectra are based on the
identical geometric models and spin Hamiltonian parameters.
Pulse lengths, frequencies, and magnetic field values were taken
from the corresponding experiments. The flexibility of the
copper was assumed to be described by a normal distribution
in the porphyrin plane. If the width of this function was set to
1 Å (standard deviation) the simulation could be made to fit
the experimental data (Figure 6). However, this flexibility model
implies a large deformation of the porphyrin moiety with a
standard deviation of ∼30% of the porphyrin radius (3.5 Å).
This and especially the implied significant elongation of the
minimum energy structure is physically unreasonable.
Spin-density distribution in the Cu(OEP)-moiety may be an
additional reason for the damping of the dipolar oscillation. To
estimate the significance of the spin-density distribution, DFT
calculations have been performed of the smaller but already
asymmetric molecule 2 and Mulliken atomic spin densities have
been computed. The resulting atomic spin densities are to 56%
localized on the copper atom, 42% are equally distributed among
the four nitrogen atoms (10.6% each), and 2% are delocalized
within the rest of the porphyrin ring. (for details see Supporting
Information). These values are in very good agreement with
other calculations.50 Experimental estimates from ENDOR data
show slightly smaller spin densities on the porphyrin nitrogens
(∼8%).51 Similar calculations for 1 show that its spin density
is localized to 50% on the oxygen and to 45% on the nitrogen
of the nitroxide group (Supporting Information).
To judge the effect the form factor has together with the spin
density distribution on the PELDOR time traces, they have been
simulated calculating the dipolar interaction tensor as specified
in eq 8, but having the Cu nucleus fixed in the center of the
porphyrin. Figure 7a demonstrates that despite the rather large
spin density distribution, its effect on the simulated PELDOR
time trace is only marginal.
Figure 7. Simulations of the PELDOR time domain data of 4 treating
the spin density distribution explicitly (a), assuming an ensemble of
asymmetrically distorted porphyrins (b), and considering a distribution
in exchange couplings of 0 ( 0.8 MHz (c). The experimental time
trace with ∆νAB ) 226 MHz is shown in black and the one with ∆νAB
) 603 MHz in red. The respective simulations are overlaid as dotted
lines.
remaining two N and 50% on copper. In addition, the copper
nucleus was set 20 pm closer to Cmeso. This polarization of the
spin density leads to a slightly faster modulation period.
Applying the inverse polarization and averaging those two
signals to mimic an ensemble of asymmetrically distorted
porphyrins leads to an increased damping but still does not
reproduce the experiment (Figure 7b). Increasing the flexibility
in the geometric model also leads to pronounced damping, but
the conformational distributions needed to match the experiment
contradict the findings on structurally analogous bis-nitroxides
and make no physical sense (see Supporting Information).
However, the delocalization of spin density into the porphy-
rin-bridge system could lead to a small exchange coupling
constant J30,52 for which simulations of room temperature CW
EPR spectra set an upper limit of ∼0.2 MHz. Thus, freezing
out a conformational ensemble might lead to a distribution in
On the other hand distortions of the porphyrin lead to different
spin density distributions, especially the meso-carbon nuclei
exhibit up to 5% spin density each in saddled porphyrins.50
Possible asymmetric deformations were simulated by assigning
5% spin density on Cmeso, 12% to the vicinal N, 9% to the

PELDOR Measurements on Nitroxide-Labeled Cu(II) Porphyrin
J. Phys. Chem. A, Vol. 112, No. 23, 2008 5071
exchange couplings, which is small on average. Good agreement
of the data with the experiment was found for a normal
distribution in J53 centered at 0 with a standard deviation of
0.8 MHz (Figure 7c). In the case of the bisnitroxide systems,
the localization of the spin density on the two NO groups and
the two ester linkages prevent an exchange coupling. These
simulations indicate that the damping of the modulation induced
by the conformational flexibility alone is not sufficient to
reproduce the experimentally observed damping, but that the
time domain signal is a convolution of distributions in distances,
spin densities, and exchange couplings induced by conforma-
tional heterogeneity. In addition, it is important to note, that
not only the damping and modulation pattern is reproduced but
that also the modulation depth is met for both pump pulse
lengths (12 and 32 ns) using the same geometric model for both
frequency offsets and thus indicating that the calculated form
factors are reasonable.
Further simulations indicated that the shape of the first
minimum depends strongly on mutual orientation of the spin-
label with respect to the porphyrin ring (data not shown). Even
though the assumption of full rotational freedom of the bridging
acetylene and phenolic bonds works considerably well, devia-
tions from this model may be the reason for the observed
imperfection.
A convolution of the resonance frequencies of the two spins
with the spin-spin coupling, to treat the different pulse
selectivity in coupled systems, did not show significant effects
on the simulations. Also the influences of the g-anisotropy of
the Cu-center and from g-strains up to 1% on the dipolar
coupling have been examined by simulations but were found
to be negligible.
Finally, decreasing the frequency offset ∆νAB in the simula-
tions down to 100 MHz does not lead to a more pronounced
appearance of the parallel component, as also observed experi-
mentally (data not shown). The reason for the latter effect is
probably due to “orientational smearing” caused by the isotropic
nitrogen hyperfine coupling (∼45 MHz). Thus, spectral selectiv-
ity on copper contributes to PELDOR at X-band but the effects
are shallow.
Comparison with Data Inversion by Tikhonov Regular-
ization. For the extraction of distance distributions (P(rAB)) in
the limit of uncorrelated spin centers time domain data inversion
by Tikhonov regularization based on eq 5 is the method of
choice.26 Even though this limit is not met in our case, the results
of this method were investigated. Figure 8 shows the simulations
based on the distance distributions obtained by inversion of both
time traces of 4. The main peak at 20.6 Å in both distance
distributions matches nicely the distance inferred from the crystal
structure and the fit of the time domain data is exceptional.
However, additional distances appear at higher and smaller
distance, which have no structural reasoning. The reason for
these artifacts are the deviations of the form factors from sin θ,
which is compensated by taking additional distances into account
(the distance distribution is a best fit of P(rAB) with P(θ) ∼
sin θ). When mutual orientations are fixed as in protein
cofactors11 and spectral selection is strong,16 this approach, as
implemented in recent PELDOR simulation programs, fails.
Additionally, the width of these peaks is commonly interpreted
to represent the conformational distribution of a biradical. Yet
in the present case, this distribution does not represent the
conformational flexibility of the molecule but a convolution of
different spin delocalizations, a distribution in exchange cou-
plings, and mobility. The simulating approach used here allows
deconvoluting of the contributions and indeed, the width of
Figure 8. Results from Tikhonov regularizations: (a) experimental time
traces with ∆νAB ) 226 MHz (black curve) and ∆νAB ) 603 MHz
(red curve) and their simulations (dotted lines). In (b) the regularized
distance distributions with ∆νAB ) 226 MHz (black curve) and with
∆νAB ) 603 MHz (red curve) are shown together with the distribution
of spin-spin distances obtained from a Gaussian distribution of the
Cu nucleus in the porphyrin plane (blue curve) and the distribution of
copper-nitroxide distances from the same model without the Gaussian
distribution (green curve). (c) Comparison of this narrow distance
distribution (green) with the distance distribution obtained from data
inversion on a structurally analogous bis-nitroxide (black).
P(rAB) from the structural dynamics of the model (Figure 8b,
green curve) is in good agreement with data from Tikhonov
regularization obtained on a structurally analogous bis-nitroxide
(Figure 8c and Supporting Information).
Conclusion
We showed on a newly synthesized, isolated, and fully
characterized copper/nitroxide model system that orientation
selectivity on copper effects PELDOR time traces at X-band
frequencies although the experimentally observable differences
are shallow. The chosen simulation approach also allowed for
the deconvolution of effects contributing to the shape of
PELDOR time traces: from the angular correlations the form

5072 J. Phys. Chem. A, Vol. 112, No. 23, 2008
Bode et al.
Cammack, R.; Evans, R. W. J. Am. Chem. Soc. 2007, 129, 4868. (c) Yang,
Z.; Becker, J.; Saxena, S. J. Magn. Reson. 2007, 188, 337.
(16) (a) Denysenkov, V. P.; Prisner, T. F.; Stubbe, J.; Bennati, M. Proc.
Natl. Acad. Sci. U.S.A. 2006, 103, 13386. (b) Polyhach, Y.; Godt, A.; Bauer,
C.; Jeschke, G. J. Magn. Reson. 2007, 185, 118. (c) Savitsky, A.; Dubinskii,
A. A.; Flores, M.; Lubitz, W.; Möbius, K. J. Phys. Chem. B 2007, 111,
6245.
(17) (a) Godt, A.; Schulte, M.; Zimmermann, H.; Jeschke, G. Angew.
Chem., Int. Ed. 2006, 45, 7560. (b) Margraf, D.; Bode, B. E.; Marko, A.;
Schiemann, O.; Prisner, T. F. Mol. Phys. 2007, 105, 2153.
(18) (a) Kálai, T.; Balog, M.; Jekö, J.; Hideg, K. Synthesis 1999, 973.
(b) Schiemann, O.; Piton, N.; Plackmeyer, J.; Bode, B. E.; Prisner, T. F.;
Engels, J. W. Nat. Protoc. 2007, 2, 904.
factors can be calculated disentangling distance distributions
from spectral selectivity, spin density distribution, and exchange
coupling. Thus, if the agreement between experimental data and
simulation is good and if the experimental and spin Hamiltonian
parameters are known, more information than mere distances
can be extracted from PELDOR measurements, and structural
models can easily be verified or disproved.
Acknowledgment. We thank Dr. Ute Bahr, J. W. Goethe-
University, for recording mass spectra. This work has been
supported by the Deutsche Forschungsgemeinschaft (SFB 579
(“RNA-ligand interactions”) and the Center for Biomolecular
Magnetic Resonance (BMRZ) Frankfurt. We gratefully ac-
knowledge support by the Frankfurt Center for Scientific
Computing. We thank Professor Dr. Gunnar Jeschke for helpful
discussions.
(19) Farrugia, L. J. J Appl. Cryst. 1999, 32, 837.
(20) Adler, A. D.; Longo, F. R.; Kampas, F.; Kim, J. J. Inorg. Nucl.
Chem. 1970, 32, 2443.
(21) (a) Arnold, D. P.; Nitschinsk, L. J. Tetrahedron 1992, 48, 8781.
(b) Inhoffen, H. H.; Fuhrhop, J.-H.; Voigt, H.; Brockmann, H. Justus Liebigs
Ann. Chem. 1966, 695, 133. (c) Johnson, A. W.; Oldfield, D. J. Chem. Soc.
C 1966, 794. (d) Watanabe, E.; Nishimura, S.; Ogoshi, H.; Yoshida, Z.
Tetrahedron 1975, 31, 1385. (e) Yeh, C.-Y.; Miller, S. E.; Carpenter, S. D.;
Nocera, D. G. Inorg. Chem. 2001, 40, 3643.
(22) Hayashi, N.; Murayama, M.; Mori, K.; Matsuda, A.; Chikamatsu,
E.; Tani, K.; Miyabayashi, K.; Miyake, M.; Higuchi, H. Tetrahedron 2004,
60, 6363.
(23) (a) Arnold, D. P.; James, D. A. J. Org. Chem. 1997, 62, 3460. (b)
Arnold, D. P.; Hartnell, R. D. Tetrahedron 2001, 57, 1335.
(24) Mims, W. B. In Electron Paramagnetic Resonance; Geschwind,
S., Ed.; Plenum: New York, 1972; pp 263.
Supporting Information Available: Synthesis of 2, X-ray
data, UV-vis and cw EPR spectra of 4 together with DFT
calculations as well as further PELDOR data, further simula-
tions, and Fourier transformations.
References and Notes
(25) Milov, A. D.; Maryasov, A. G.; Tsvetkov, Y. D. Appl. Magn. Reson.
1998, 15, 107.
(1) (a) Handbook of Metalloproteins; Messerschmidt, A., Huber, R.,
Poulos, T., Wieghardt, K., Eds.; John Wiley & Sons, Ltd.: Chichester,
England, 2001; Vol. 1, p 2. (b) DeGrado, W. F.; Summa, C. M.; Pavone,
V.; Nastri, F.; Lombardi, A. Ann. ReV. Biochem. 1999, 68, 779.
(2) (a) Eckstein, F.; Lilley, D. M. J. Nucleic Acids and Molecular
Biology, Vol. 10: Catalytic RNA; Springer: Berlin, 1995; (b) Gesteland,
R. F.; Cech, T. R.; Atkins, J. F. The RNA World, 3rd ed.; Cold Spring
Harbor: New York, 2006.
(3) (a) Fragai, M.; Luchinat, C.; Parigi, G. Acc. Chem. Res. 2006, 39,
909. (b) Arnesano, F.; Banci, L.; Piccioli, M. Q. ReV. Biophys. 2005, 38,
167–219.
(4) (a) Calle, C.; Sreekanth, A.; Fedin, M. V.; Forrer, J.; Garcia-Rubio,
I.; Gromov, I. A.; Hinderberger, D.; Kasumaj, B.; Léger, P.; Mancosu, B.;
Mitrikas, G.; Santangelo, M. G.; Stoll, S.; Schweiger, A.; Tschaggelar, R.;
Harmer, J. HelV. Chim. Acta 2006, 89, 2495. (b) Bennati, M.; Prisner, T. F.
Rep. Prog. Phys. 2005, 68, 411.
(5) (a) AdVanced EPR Applications in Biology and Biochemistry; Hoff,
A. J., Ed.; Elsevier: Amsterdam, 1989. (b) Paramagnetic Resonance of
Metallobiomolecules; Telser, J., Ed.; ACS Symposium Series, Vol. 858;
Oxford University Press: Oxford, 2003.
(6) (a) DeRose, V. J. Chem. Biol. 2002, 9, 961. (b) Kisseleva, N.; Kraut,
S.; Jäschke, A.; Schiemann, O. HFSP J. 2007, 1, 127. (c) Kisseleva, N.;
Khvorova, A.; Westhof, E.; Schiemann, O. RNA 2005, 11, 1. (d) Schiemann,
O.; Fritscher, J.; Kisseleva, N.; Sigurdsson, S. T.; Prisner, T. F. ChemBio-
Chem 2003, 4, 1057.
(7) Schiemann, O.; Prisner, T. F. Q. ReV. Biophys. 2007, 40, 1.
(8) (a) Voss, J.; Salwinski, L.; Kaback, H. R.; Hubbell, W. L. Proc.
Nat. Acad. Sci. U.S.A. 1995, 92, 12295. (b) Distance measurements in
biological systems by EPR. Biological Magnetic Resonance; Berliner, L. J.,
Eaton, S. S., Eaton, G. R., Eds.; Kluwer Academic/Plenum Publishers: New
York, 2000; vol. 19.
(9) (a) Jeschke, G.;.; Polyhach, Y. Phys. Chem. Chem. Phys 2007, 9,
1895. (b) Schiemann, O.; Piton, N.; Mu, Y.; Stock, G.; Engels, J. W.; Prisner,
T. F. J. Am. Chem. Soc. 2004, 126, 5722.
(10) (a) Hara, H.; Kawamori, A.; Astashkin, A. V.; Ono, T.-a. Biochim.
Biophys. Acta 1996, 1276, 140. (b) Elsässer, C.; Brecht, M.; Bittl, R. J. Am.
Chem. Soc. 2002, 124, 12606. (c) Astashkin, A. V.; Seravalli, J.; Man-
soorabadi, S. O.; Reed, G. H.; Ragsdale, S. W. J. Am. Chem. Soc 2006,
128, 3888.
(11) (a) Bennati, M.; Robblee, J. H.; Mugnaini, V.; Stubbe, J.; Freed,
J. H.; Borbat, P. J. Am. Chem. Soc. 2005, 127, 15014. (b) Kay, C. W. M.;
Elsässer, C.; Bittl, R.; Farell, S. R.; Thorpe, C. J. Am. Chem. Soc. 2006,
128, 76.
(12) (a) Milov, A. D.; Salikov, K. M.; Shirov, M. D. Fiz. TVerd. Tela
1981, 23, 975. (b) Larsen, R. G.; Singel, D. J. J. Chem. Phys. 1993, 98,
5134.
(13) (a) Borbat, P. P.; Freed, J. H. Chem. Phys. Lett. 1999, 313, 145.
(b) Becker, J. S.; Saxena, S. Chem. Phys. Lett. 2005, 414, 248.
(14) Narr, E.; Godt, A.; Jeschke, G. Angew. Chem., Int. Ed. 2002, 41,
3907.
(26) (a) Jeschke, G.; Panek, G.; Godt, A.; Bender, A.; Paulsen, H. Appl.
Magn. Reson. 2004, 26, 223. (b) Bowman, M. K.; Maryasov, A. G.; Kim,
N.; DeRose, V. J. Appl. Magn. Reson 2004, 26, 23. (c) Chiang, Y.-W.;
Borbat, P. P.; Freed, J. H. J. Magn. Reson. 2005, 172, 279.
(27) Bode, B. E.; Margraf, D.; Plackmeyer, J.; Dürner, G.; Prisner, T. F.;
Schiemann, O. J. Am. Chem. Soc. 2007, 129, 6736.
(28) Mansoorabadi, S. O.; Reed, G. H. In Paramagnetic Resonance of
Metallobiomolecules; Telser, J.,. Ed.; American Chemical Society: Wash-
ington, DC, 2003; pp 82.
(29) Cunningham, K. L.; McNett, K. M.; Pierce, R. A.; Davis, K. A.;
Harris, H. H.; Falck, D. M.; McMillin, D. R. Inorg. Chem. 1997, 36, 608.
(30) Fritscher, J.; Beyer, M.; Schiemann, O. Chem. Phys. Lett. 2002,
364, 393.
(31) (a) Chikira, M.; Kon, H.; Hawley, R. A.; Smith, K. M. J. Chem.
Soc., Dalton Trans 1979, 245. (b) Iwaizumi, M.; Ohba, Y.; Iida, H.;
Hirayama, M. Inorg. Chim. Acta 1984, 82, 47.
(32) Jeschke, G.; Chechik, V.; Ionita, P.; Godt, A.; Zimmermann, H.;
Banham, J.; Timmel, C. R.; Hilger, D.; Jung, H. Appl. Mag. Reson. 2006,
30, 473.
(33) Jeschke, G.; Bender, A.; Paulsen, H.; Zimmermann, H.; Godt, A.
J. Magn. Reson. 2004, 169, 1.
(34) Kohn, W.; Sham, L. J. Phys. ReV. 1965, 140, A1133.
(35) Ahlrichs, R.; Bär, M.; Baron, H.-P.; Bauernschmitt, R.; Böcker,
S.; Crawford, N.; Deglmann, P.; Ehrig, M.; Eichkorn, K.; Elliott, S.; Furche,
F.; Haase, F.; Häser, M.; Hättig, C.; Hellweg, A.; Horn, H.; Huber, C.;
Huniar, U.; Kattannek, M.; Köhn, A.; Kölmel, C.; Kollwitz, M.; May, K.;
Nava, P.; Ochsenfeld, C.; Öhm, H.; Patzelt, H.; Rappoport, D.; Rubner,
O.; Schäfer, A.; Schneider, U.; Sierka, M.; Treutler, O.; Unterreiner, B.;
Arnim, M. v.; Weigend, F.; Weis, P.; Weiss, H. TURBOMOLE, version
5.6; Quantum Chemistry Group, University of Karlsruhe: Karlsruhe,
Germany, 2002.
(36) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R. Montgomery, J.; Vreven, T.; Kudin, K. N.; Burant,
J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.;
Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada,
M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima,
T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.;
Hratchian, H. P.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.;
Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.;
Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.;
Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels,
A. D.; Strain, M. C.; Farkas, O.; Malick, K.; Rabuck, A. D.; Raghavachari,
K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.;
Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.;
Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng,
C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.;
Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03, revision
C.02; Gaussian Inc: Pittsburgh, PA, 2004.
(37) Becke, A. D. Phys. ReV. A 1988, 38, 3098.
(38) (a) Perdew, J. P. Phys. ReV. B 1986, 33, 8822. (b) Perdew, J. P.
Phys. ReV. B 1986, 34, 7406.
(15) (a) Amsterdam, I. M. C. v.;.; Ubbink, M.; Canters, G. W.; Huber,
M. Angew. Chem., Int. Ed. 2003, 42, 62. (b) Kay, C. W. M.; Mkami, H. E.;

PELDOR Measurements on Nitroxide-Labeled Cu(II) Porphyrin
J. Phys. Chem. A, Vol. 112, No. 23, 2008 5073
(39) Schafer, A.; Huber, C.; Ahlrichs, R. J. Chem. Phys. 1994, 100,
(47) CW X-band EPR studies of 4 without 1-methylimidazole showed
strong half-field signals at g ) 4 indicative of close spin-spin contacts.
These signals disappeared completely upon adding 1-methylimidazole.
(48) Weber, A.; Schiemann, O.; Bode, B.; Prisner, T. F. J. Magn. Reson.
2002, 157, 277.
(49) Offsets larger than 603 MHz are not feasible due to limited B1 and
signal-to-noise ratio.
5829.
(40) (a) Eichkorn, K.; Treutler, O.; Öhm, H.; Haser, M.; Ahlrichs, R.
Chem. Phys. Lett. 1995, 242, 652. (b) Eichkorn, K.; Weigend, F.; Treutler,
O.; Ahlrichs, R. Theor. Chem. Acc 1997, 97, 119.
(41) Becke, A. D. J. Chem. Phys. 1993, 98, 5648.
(42) (a) Perdew, J. P.; Wang, Y. Phys. ReV. B 1992, 45, 13244. (b)
(50) Shao, J.; Steene, E.; Hoffman, B. M.; Ghosh, A. Eur. J. Inorg.
Chem. 2005, 2005, 1609.
Perdew, J. P. Physica B 1992, 172, 1.
(43) Arbuznikov, A. V.; Vaara, J.; Kaupp, M. J. Chem. Phys. 2004,
120, 2127.
(44) Kutzelnigg, W.; Fleischer, U.; Schindler, M. In NMR Basic
Principles and Progress; Diehl, P., Fluck, E., Günther, H., Kosferld, R.,
Seelig, J., Eds.;Springer-Verlag: Berlin/Heidelberg, Germany, 1990; Vol.
23, p 165–262.
(51) (a) Iwaizumi, M.; Kudo, T.; Kita, S. Inorg. Chem. 1986, 25, 1546.
(b) Astashkin, A. V.; Raitsimring, A. M.; Walker, F. A.; Rensing, C.;
McEvoy, M. M. J. Biol. Inorg. Chem. 2005, 10, 221.
(52) Elschenbroich, C.; Wolf, M.; Schiemann, O.; Harms, K.; Burghaus,
O.; Pebler, J. Organometallics 2002, 21, 5810.
(53) In case of a small exchange coupling, the spin-spin coupling ωAB
(45) Bolte, M.; Plackmeyer, J. Acta Cryst. 2006, E62, e22.
ˆ
is defined as ωdd + J with the exchange coupling Hamiltonian given by HJ
(46) (a) Senge, M. O.; Vicente, M. G. H.; Gerzevske, K. R.; Forsyth,
T. P.; Smith, K. M. Inorg. Chem. 1994, 33, 5625. (b) Clement, T. E.; Nurco,
D. J.; Smith, K. M. Inorg. Chem. 1998, 37, 1150.
ˆ
ˆ
) J SASB.
JP710504K