EPR, ENDOR, and DFT studies on (β-octahalo-mesotetraarylporphyrin) copper complexes: Characterization of the metal(dx2 -y2)-porphyrin(a2u) orbital interaction

Article abstract of DOI:10.1002/ejic.200400549

A series of planar (porphyrin)copper(II) complexes and their β-octahalogenated saddled derivatives have been studied by Electron Paramagnetic Resonance (EPR) spectroscopy, Electron Nuclear DOuble Resonance (ENDOR) spectroscopy, and Density Functional Theoretical (DFT) calculations. Both EPR/ENDOR spectroscopy and DFT calculations indicate a decrease in spin density on the central copper(II) ion and on the nitrogen atoms in the saddled compounds relative to the planar complexes. The EPR/ENDOR measurements show that the hyperfine coupling decreases by 12% on the nitrogen atoms and 9% on the copper ion, in going from planar (5,10,15,20-tetraphenylporphyrin)copper (Cu[TPP]) to saddled (2,3,7,8,12,13,17,18-octabromo-5,10,15,20- tetraphenylporphyrin)copper (Cu[Br₈TPP]). Accordingly, saddling results in a decrease in the spin density on the copper ion and on the nitrogen atoms. DFT calculations on Cu[Br₈TPP] yield spin populations of 42.4% on the copper ion, 9.9% on each nitrogen atom and 4.9% on each meso carbon atom, relative to DFT spin populations of 62, 10.2 and 0.3% on the copper ion, each nitrogen and each meso carbon atom, respectively, for porphinecopper (Cu[P]). These calculations further indicate that the decrease in spin density on the copper ion in the saddled complexes results from a saddling-induced Cu(d_x²_-y²)-porphyrin(a _2u) orbital overlap whereby some of the Cu spin density is delocalized onto the porphyrin ring. The decrease in nitrogen spin population with saddling appears to be a more subtle effect caused by a superposition of two opposing factors. Saddling decreases the overlap between the nitrogen lone pairs and the Cu d_x²_-y² orbital on one hand and enhances the overlap between the copper d_x ²_-y² orbital and the porphyrin a_2u HOMO on the other.

Full text of DOI:10.1002/ejic.200400549

FULL PAPER
EPR, ENDOR, and DFT Studies on (β-Octahalo-meso-
tetraarylporphyrin)copper Complexes: Characterization of the
2
2
Metal(dx ؊y )؊Porphyrin(a2u) Orbital Interaction
Junlong Shao,[a] Erik Steene,[b] Brian M. Hoffman,*[a] and Abhik Ghosh*[b]
Keywords: Copper / Porphyrins / Density functional calculations / ENDOR spectroscopy / EPR spectroscopy
A series of planar (porphyrin)copper(II) complexes and their
β-octahalogenated saddled derivatives have been studied by
Electron Paramagnetic Resonance (EPR) spectroscopy, Elec-
tron Nuclear DOuble Resonance (ENDOR) spectroscopy, and
Density Functional Theoretical (DFT) calculations. Both EPR/
ENDOR spectroscopy and DFT calculations indicate a de-
each nitrogen atom and 4.9% on each meso carbon atom,
relative to DFT spin populations of 62, 10.2 and 0.3% on the
copper ion, each nitrogen and each meso carbon atom, re-
spectively, for porphinecopper (Cu[P]). These calculations
further indicate that the decrease in spin density on the cop-
per ion in the saddled complexes results from a saddling-
2
2
crease in spin density on the central copper(II) ion and on induced Cu(dx −y )−porphyrin(a2u) orbital overlap whereby
the nitrogen atoms in the saddled compounds relative to the
some of the Cu spin density is delocalized onto the porphyrin
planar complexes. The EPR/ENDOR measurements show
ring. The decrease in nitrogen spin population with saddling
that the hyperfine coupling decreases by 12% on the nitro- appears to be a more subtle effect caused by a superposition
gen atoms and 9% on the copper ion, in going from
planar (5,10,15,20-tetraphenylporphyrin)copper (Cu[TPP]) to
of two opposing factors. Saddling decreases the overlap be-
2
2
tween the nitrogen lone pairs and the Cu dx −y orbital on one
2
2
saddled
(2,3,7,8,12,13,17,18-octabromo-5,10,15,20-tetra-
hand and enhances the overlap between the copper dx −y
phenylporphyrin)copper (Cu[Br8TPP]). Accordingly, saddling orbital and the porphyrin a2u HOMO on the other.
results in a decrease in the spin density on the copper ion
and on the nitrogen atoms. DFT calculations on Cu[Br8TPP]
yield spin populations of 42.4% on the copper ion, 9.9% on
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2004)
Introduction
dxy orbital is a ‘‘t2g-type’’ d orbital (in the notation of the
Oh point group). Electronic spin density distributions re-
Nonplanar conformations of porphyrins and related co- flecting this orbital interaction have been extensively stud-
factors occur as conserved features of several metalloprote- ied by NMR spectroscopy for ruffled iron(iii)
ins, and it has been proposed that this nonplanarity is actu- porphyrins.[9Ϫ12,14,15]
Saddling
switches
on
the
ally of biological importance.[1] Nonplanar deformations metal(dx Ϫy )Ϫporphyrin(a2u) orbital interaction (as shown
such as ruffling and saddling (Figure 1) exert a strong influ-
ence on the chemical properties of metalloporphyrins and
related molecules; these properties include electrochemical
half-wave potentials,[2,3] electronic absorption[4,5] and vi-
brational spectra,[6Ϫ8] energetics of different spin states and
electron density distributions as reflected in NMR and EPR
spectra,[9Ϫ17] DFT calculations,[18] and reactivity toward ax-
ial ligands, etc.[9,10,12,13,15,16,18] Many of these effects arise
from specific metal(d)Ϫporphyrin(π) orbital interactions
that are symmetry-forbidden in planar metalloporphyrins,
but which become allowed as a result of nonplanar defor-
mations. For example, ruffling switches on the
metal(dxy)Ϫporphyrin(a2u) orbital interaction, where the
2
2
[a]
Department of Chemistry, Northwestern University,
Evanston, IL 60208, USA
[b]
Institute of Chemistry, Faculty of Science, University of
Tromsø,
9037 Tromsø, Norway
Figure 1. Nonplanar distortions of porphyrins; filled and open
circles represent the displacement of the core atoms above and be-
low the porphyrin mean plane
Eur. J. Inorg. Chem. 2005, 1609Ϫ1615
DOI: 10.1002/ejic.200400549
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1609

J. Shao, E. Steene, B. M. Hoffman, A. Ghosh
FULL PAPER
in Figure 2),[7,19] where the dx Ϫy orbital is a ‘‘eg-type’’ d Cu[Br8T(p-CF3-P)P], and Cu[Br8TPP] (Figure 3), and
2
2
orbital (in the notation of the Oh point group). NMR stud- DFT(PW91/TZP) calculations on Cu[Br8TPP],
ies on saddled iron(iii) porphyrins have also thrown light Cu[Br8TPFPP], Cu[Cl8TPP], and Cu[Cl8TPFPP] (Fig-
on this orbital interaction.[13,16] In this paper, we ure 4).
further[7,19] explore the consequences of the metal-
2
2
(dx Ϫy )Ϫporphyrin(a2u) orbital interaction by EPR and
ENDOR spectroscopic studies and density functional the-
ory (DFT) calculations on saddled CuII porphyrins. The re-
sults provide a clear picture of saddling-induced redistri-
bution of unpaired electron density in (porphyrin)CuII com-
pounds.
Figure 3. (Porphyrin)copper complexes studied by EPR and EN-
DOR spectroscopy: Cu[Cl8T(p-CF3-P)P] X ϭ Cl, Y ϭ CF3;
Cu[Cl8TPP] X ϭ Cl, Y ϭ H; Cu[Br8T(p-CF3-P)P] X ϭ Br, Y ϭ
CF3; Cu[Br8TPP] X ϭ Br, Y ϭ H
Figure 4. (Porphyrin)copper complexes studied by DFT calcu-
lations: Cu[Br8TPP] X ϭ Br, Y ϭ H; Cu[Br8TPFPP] X ϭ Br, Y ϭ
F; Cu[Cl8TPP] X ϭ Cl, Y ϭ H; Cu[Cl8TPFPP] X ϭ Cl, Y ϭ F
Results
The EPR spectra of the (porphyrin)Cu complexes exam-
ined here are qualitatively similar to those of other tetrag-
onal planar copper complexes. The g|| peaks partially over-
lap with the gЌ peaks. The g||, gЌ and A|| values were
obtained by EPR simulations with WINEPR SimFornia
Cu
Figure 2. Two views of the singly occupied b2 (a2u-type) HOMO
of Cu[Br8TPP] (D2d)
Investigations of single crystal Cu[TPP] and Ag[TPP] (version 1.25, Bruker Analytische Messtechnik GmbH) and
1
with 14N, H, and metal ENDOR spectroscopy by Brown are listed in Table 1. In the EPR simulations, the nitrogen
and Hoffman have yielded the complete hyperfine coupling hyperfine tensors were taken from the 14N ENDOR results.
and 14N quadrupolar tensors of the two complexes.[20] As can be seen from Table 1, the g|| (ഠ 2.19) and gЌ (ഠ
Theoretical analyses of the spectra suggested metal spin 2.05) values of all the Cu porphyrin complexes examined
populations of 0.62 for Cu[TPP] and 0.38 for Ag[TPP]. are identical within experimental error. However, the abso-
Cu
DFT(PW91/TZP) calculations on CuII porphine yielded a lute values of A|| differ significantly among the various
Cu spin population of 50.7%.[2] In this work, we present compounds studied and decrease in the following order:
EPR and 14N ENDOR measurements on the saddled Cu Cu[T(p-CF3-P)P] ഠ Cu[TPP] Ͼ Cu[Cl8T(p-CF3-P)P] ഠ
porphyrin derivatives Cu[Cl8T(p-CF3-P)P], Cu[Cl8TPP], Cu[Cl8TPP] Ͼ Cu[Br8T(p-CF3-P)P] ഠ Cu[Br8TPP]. These
1610
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2005, 1609Ϫ1615

EPR, ENDOR, and DFT Studies on (β-Octahalo-meso-tetraarylporphyrin)copper Complexes
FULL PAPER
Table 1. 63,65Cu and 14N hyperfine couplings of (porphyrin)copper complexes[a,b]
Cu[T(p-CF3-P)P]
Cu[TPP]
Cu[Cl8T(p-CF3-P)P]
Cu[Cl8TPP]
Cu[Br8T(p-CF3-P)P]
Cu[Br8TPP]
NA1 [MHz]
NA2 [MHz]
NA3 [MHz]
44.5(2)
42.8(3)
55.0(3)
47.4(2)
Ϫ618(4)
2.04(1)
2.17(1)
44.1(2)
42.8(3)
54.2(3)
47.0(2)
Ϫ615(4)
2.04(1)
2.18(1)
43.2(2)
40.5(5)
53.8(5)
45.8(3)
Ϫ591(8)
2.06(2)
2.18(2)
42.5(2)
40.0(5)
53.0(5)
45.2(3)
Ϫ591(8)
2.06(2)
2.21(2)
40.0(2)
38.5(3)
51.5(3)
43.3(2)
Ϫ588(4)
2.06(1)
2.17(1)
38.7(2)
37.6(3)
50.0(3)
42.1(2)
Ϫ585(4)
2.06(1)
2.18(1)
NAiso
Cu
A [MHz]
||
gЌ
g||
[a]
g|| ϭ 2.19 Ϯ 0.03 and gЌ ϭ 2.05 Ϯ 0.02 for all complexes, within experimental error as obtained by X-band EPR spectroscopy at
77 K. 14N ENDOR were taken at X-band and 4.2 K. [b] Simulations in all cases employed the quadrupole parameters found for Cu[TPP]:
P1 ϭ Ϫ0.3, P2 ϭ Ϫ0.6, and P3 ϭ 0.9.[20]
results are compared with those of other (porphyrin)copper solved, as shown in Figure 6. The well-matched simulation
complexes in Table 2.
results (obtained with GENSIM, a locally developed
software) are also presented in Figure 6. The sharp simu-
lated peaks reveal the positions of the hyperfine couplings
(An/2, n ϭ 1, 2, 3), nuclear Zeeman splitting (2νN, since An/
2 Ͼ νN) and quadrupole splitting (3Pn, n ϭ 1, 2, 3). The
quadrupole splitting was taken from previous work on
Cu[TPP] and are assumed to differ only slightly in other
complexes. Numerous hyperfine coupling values were at-
tempted in the simulation to best match the simulated and
Table 2. Spin-Hamiltonian parameters of (porphyrin)copper(ii)
complexes[21]
N
N
experimental spectra. The AЌ at g|| (or A1) values, unlike
the others, were measured directly from the spectra. Unlike
the single-crystal-like spectra taken at the parallel field posi-
tion, there are two components, NA2 (or NAЌ) and NA3 (or
NA||), at the perpendicular field position. These originate
from the CuϪN vectors perpendicular and parallel to the
magnetic field, respectively. The full 14N ENDOR spectra
[a]
X and Y are functional groups marked in Figure 3. [b] Unit is in
[c]
MHz. Spin density on copper ion calculated with Equation (1).
Figure 5 displays the 14N ENDOR spectra of Cu[TPP]
and Cu[Br8TPP] taken at g|| and CuIϪ3/2, where only single-
crystal-like spectra, derived from the molecules with their
symmetry axis parallel to the static magnetic field, were col-
lected for both samples. A significant hyperfine coupling
shift (Ϫ5.4 MHz or Ϫ2.7 MHz peak-to-peak) between
Cu[TPP] and Cu[Br8TPP] was observed. Unlike the well-
resolved double ENDOR spectra presented by Schweiger in
his review on ENDOR studies of transition metal com-
plexes,[22] the 14N ENDOR spectra of Cu[T(p-CF3-P)P] at
Figure 5. 14N Davies ENDOR spectra of Cu[TPP] and Cu[Br8TPP]
at the parallel field position; the peak center (A1/2) shift is marked
between these two (porphyrin)copper complexes; experimental con-
ditions: temperature, 4.2 K; microwave frequency, 9.715 GHz; mag-
netic field, 0.29 T; microwave pulse length, 32Ϫ16Ϫ32 ns; τ, 752 ns;
RF pulse length, 20 ns; repetition rate, 20 ms; total points of each
both parallel and perpendicular field positions are unre- spectrum, 256; number of shots at each point, 300
Eur. J. Inorg. Chem. 2005, 1609Ϫ1615
www.eurjic.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1611

J. Shao, E. Steene, B. M. Hoffman, A. Ghosh
FULL PAPER
of the various (porphyrin)copper complexes studied at par-
allel and perpendicular field positions, along with the simu-
lated spectra, are shown in Figures 7 and 8, respectively.
The arrows mark the hyperfine splitting peak positions in
the spectra and the 14N hyperfine couplings are listed in
Table 1. The 14N hyperfine couplings decrease in the same
Cu
order as
A , namely Cu[T(p-CF3-P)P] ഠ Cu[TPP] Ͼ
||
Cu[Cl8T(p-CF3-P)P] ഠ Cu[Cl8TPP] Ͼ Cu[Br8T(p-CF3-
P)P] ഠ Cu[Br8TPP]. This sequence is also depicted in the
bar graph of Figure 9, which again shows that hyperfine
couplings decrease on going from the β-unsubstituted to
the β-octachloro and β-octabromo complexes.
Figure 7. 14N Davies ENDOR spectra of all the complexes at paral-
lel field position; the arrows denote the centers of the hyperfine
coupling peaks; the experimental conditions are the same as in Fig-
ure 5
Figure 10 shows the X-band (9 GHz, bottom) and Ka-
band (35 GHz, top) proton and fluorine ENDOR spectra
of Cu[Br8T(p-CF3-P)P]. With the X-band, the 19F split-
tings, arising from the trifluoromethyl group, overlap at the
low-frequency shoulder of the proton hyperfine coupling
profile. With the Ka-band, the hyperfine couplings from the
two components are completely separated. Since there are
no β-protons on the pyrrole rings in this compound, the
proton signals are assigned to phenyl ring protons. This
indicates that proton hyperfine couplings originally as-
signed to pyrrole hydrogen atoms[20] are in fact associated
with phenyl hydrogen atoms. Absent selective deuteration
or chemical substitution, as done here, the distinction be-
Figure 6. 14N Davies ENDOR spectra of Cu[T(p-CF3-P)P] at both
parallel and perpendicular field positions; the solid lines are the
experimental spectra, the dotted lines the simulated spectra, and
the dashed lines the simulated lines with narrow line width to show
the peak positions, which are marked in the graph; the experimen-
tal conditions are the same as in Figure 5
˚
˚
Table 3. Selected optimized geometries [A, °] and z-displacements [A]
Compound
Distances
c
Angles
ξ
z-displacements
a
b
d
m
α
β
δ
γ
zα
zβ
zN
CuII[Br8TPP]
1.378 1.449 1.373 1.411 2.023
108.6 107.2 108.1 121.3 122.6
108.6 107.2 108.0 121.8 122.4
108.7 107.2 107.9 122.7 124.3
108.7 107.2 107.8 123.3 124.3
0.449 1.231
0.459 1.217
0.344 0.939
0.334 0.868
0.066
0.092
0.057
0.084
CuII[Br8TPFPP] 1.376 1.449 1.372 1.410 2.012
CuII[Cl8TPP]
CuII[Cl8TPFPP] 1.378 1.450 1.368 1.405 2.027
1.379 1.450 1.369 1.406 2.032
1612
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2005, 1609Ϫ1615

EPR, ENDOR, and DFT Studies on (β-Octahalo-meso-tetraarylporphyrin)copper Complexes
FULL PAPER
tween these two types of proton at similar distances from
Cu is very challenging.
Table 3, in conjunction with Figure 4, shows selected op-
timized geometry parameters for CuII[Br8TPP], CuII[Br8-
TPFPP], CuII[Cl8TPP], and CuII[Cl8TPFPP]. As shown by
the out-of-plane z-displacement of the β-carbon atoms (zβ),
the β-octabrominated complexes are significantly more
saddled than the corresponding β-octachlorinated ana-
logues. However, analogous TPP and TPFPP derivatives are
comparably saddled. The CuϪN bond lengths are approxi-
˚
mately the same (2.01Ϫ2.03 A) in all four molecules. Table 4
shows the calculated atomic spin populations for these four
saddled porphyrins, and these data are discussed in the
next section.
Table 4. Gross atomic spin populations
Compound
Metal
Nporph Cα
Cβ
Cmeso
CuII[Cl8TPP]
CuII[Br8TPP]
0.4465 0.1015 Ϫ0.0028 0.0039 0.0362
0.4151 0.0983 Ϫ0.0105 0.0034 0.0531
CuII[Cl8TPFPP] 0.4722 0.1025 Ϫ0.0029 0.0043 0.0212
CuII[Br8TPFPP] 0.4388 0.0998 Ϫ0.0076 0.0040 0.0389
Discussion
Isotropic hyperfine coupling arises from the Fermi inter-
action and is proportional to the spin population on the
observed nonmetal nucleus (nitrogen nuclei in this case)
[Equation (1)],23] where ge is the electron g value, g|| the par-
Figure 8. 14N Davies ENDOR spectra of the whole complexes at
the perpendicular field position; the arrows denote the centers of
the hyperfine coupling peaks; the experimental conditions are the
same as in Figure 5
Figure 9. Bar graph of the hyperfine couplings of all (porphyrin)copper complexes
www.eurjic.org
Eur. J. Inorg. Chem. 2005, 1609Ϫ1615
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1613

J. Shao, E. Steene, B. M. Hoffman, A. Ghosh
FULL PAPER
Cu
14
As mentioned above, both the A|| and
N
hyperfine
iso
couplings decrease in the following order: Cu[T(p-CF3-
P)P] ഠ Cu[TPP] Ͼ Cu[Cl8T(p-CF3-P)P] ഠ Cu[Cl8TPP] Ͼ
Cu[Br8T(p-CF3-P)P] ഠ Cu[Br8TPP] (Table 1). In other
Cu
words, Table 1 indicates that the A|| hyperfine couplings
of the saddled complexes are generally considerably lower
than those of their planar counterparts and are comparable
only to those observed for highly electron-deficient meso-
tetrakis(pyridiniumyl)porphyrin derivatives, Cu[TMPyP][26]
(Table 2). This indicates that the spin populations on the
Cu and N atoms decrease with increasing saddling of the
macrocycle. The decrease in the Cu spin population is
understandable as a simple consequence of the saddling-
2
2
induced Cu(dx Ϫy )Ϫporphyrin(a2u) orbital overlap whereby
some of the Cu spin density is delocalized onto the porphy-
rin ring. However, we believe that the decrease in nitrogen
spin population with saddling reflects two opposing factors.
On the one hand, saddling inhibits contact between the ni-
2
2
trogen lone-pairs and the Cu dx Ϫy orbital. On the other
hand, saddling encourages and enhances the overlap be-
2
2
tween the copper dx Ϫy and the porphyrin a2u HOMO. The
combined result of these two factors is a certain decrease
in spin population on the nitrogen atoms.
The decrease in spin density on going from the β-unsub-
stituted to the β-octahalogenated complexes, as measured
Cu
14
by the A|| and
N
hyperfine couplings, is consistent
iso
with DFT calculations showing atomic spin populations for
CuII[Br8TPP]: Cu 42.4, N 9.9 and Cmeso 4.9%. This can be
compared with (porphine)CuII: Cu 50.7, N 10.8 and Cmeso
0.2%. The DFT calculations indicate that the loss of spin
1
Figure 10. H and 19F Mims ENDOR spectra of Cu[Br8T(p-CF3-
P)P] at 9 GHz and 35 GHz at the perpendicular field position; the
arrows denote the center of the 19F hyperfine coupling peaks; the
experimental conditions at 9 GHz are: temperature, 4.2 K; micro-
wave frequency, 9.720 GHz; magnetic field, 0.34 T; microwave pulse density on the copper and nitrogen atoms in the saddled
length, 16 ns; τ, 140 ns; RF pulse length, 10 ns; repetition rate, 30
ms; total points of each spectrum, 256; number of shots at each
point, 300; the experimental conditions at 35 GHz are: tempera-
ture, 2 K; microwave frequency, 35.026 GHz; magnetic field, 1.234
T; microwave pulse length, 48 ns; τ, 300 ns; RF pulse length, 40 ns;
repetition rate, 20 ms; total points of each spectrum, 256; number
of shots at each point, 80
complexes is accompanied by an increase in the spin density
on Cmeso. This effect can potentially be explored by EPR
13
and ENDOR analysis of
C
analogues of the relevant
meso
complexes, but such analyses have not yet been completed.
Conclusion
allel g value, gЌ the perpendicular g value, α12 the odd elec-
tron populations on the copper ion, ignoring the interac-
tions between copper dx Ϫy and nitrogen 2s and 2p orbitals.
In summary, we have studied a number of (β-octahalo-
meso-tetraarylporphyrins)copper(ii) complexes by using
EPR/ENDOR spectroscopy and DFT calculations to better
understand the consequences of the saddling-induced
2
2
Cu
A ϭ CuA1 ϭ P[Ϫκ Ϫ 4/7α12 Ϫ 3/7(ge Ϫ gЌ) Ϫ (ge Ϫ g||)]
(1)
||
2
2
metal(dx Ϫy )Ϫporphyrin(a2u) orbital overlap on the elec-
The parameter P is proportional to the radial expectation tron distribution of (porphyrin)copper complexes. Both
2
2
value of the square of the dx Ϫy orbital amplitude of the EPR/ENDOR spectroscopy and DFT calculations show
metal, and κ is an empirical constant that describes the in- that the spin populations on the central copper ion and on
ner-shell polarization by the d-shell vacancy and quantifies the nitrogen atoms are significantly decreased in these
the isotropic metal hyperfine coupling.[24] By taking P ϭ saddled complexes relative to their planar counterparts. The
0.037 and[25] κ ϭ 0.406 [adjusted value to obtain a spin diminished spin density on the copper ion is directly at-
density of 0.624[20] on the copper ion in Cu[TPP] with tributable to the saddling-induced Cu(dx Ϫy )Ϫporphyrin-
Equation (1)], we can empirically estimate the spin popu- (a2u) orbital overlap whereby Cu spin density is delocalized
lations on all (porphyrin)copper complexes examined. onto the porphyrin ring. The decrease in nitrogen spin
Table 2 summarizes the EPR data for a wide range of (por- population with saddling appears to be a more subtle effect
phyrin)copper complexes, including the empirically esti- caused by a superposition of two opposing factors. Sad-
mated copper spin populations. The main conclusions dling decreases the overlap between the nitrogen lone pairs
2
2
2
2
based on Tables 1 and 2 are as follows.
and the Cu dx Ϫy orbital on one hand and enhances the
1614
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2005, 1609Ϫ1615

EPR, ENDOR, and DFT Studies on (β-Octahalo-meso-tetraarylporphyrin)copper Complexes
FULL PAPER
2
2
E. Van Caemelbecke, K. M. Kadish, Inorg. Chem. 1998, 37,
4567.
A. Ghosh, I. Halvorsen, H. J. Nilsen, E. Steene, T. Wondimag-
egn, R. Lie, E. van Caemelbecke, N. Guo, Z. Ou, K. M. Kad-
ish, J. Phys. Chem. B 2001, 105, 8120.
C. M. Drain, S. Gentemann, J. A. Roberts, N. Y. Nelson, C. J.
Medforth, S. L. Jia, M. C. Simpson, K. M. Smith, J. Fajer, J.
A. Shelnutt, D. Holten, J. Am. Chem. Soc. 1998, 120, 3781.
S. Gentemann, C. J. Medforth, T. P. Forsyth, D. J. Nurco, K.
M. Smith, J. Fajer, D. Holten, J. Am. Chem. Soc. 1994, 116,
7363.
L. D. Sparks, K. K. Anderson, C. J. Medforth, K. M. Smith,
J. A. Shelnutt, Inorg. Chem. 1994, 33, 2297.
J. A. Shelnutt, C. J. Medforth, M. D. Berber, K. M. Barkigia,
K. M. Smith, J. Am. Chem. Soc. 1991, 113, 4077.
I. Halvorsen, E. Steene, A. Ghosh, J. Porphyrins Phthalocyan-
ines 2001, 5, 721.
G. Simonneaux, V. Schuenemann, C. Morice, L. Carel, L. Tou-
pet, H. Winkler, A. X. Trautwein, F. A. Walker, J. Am. Chem.
Soc. 2000, 122, 4366.
F. A. Walker, H. Nasri, I. Turowska-Tyrk, K. Mohanrao, C. T.
Watson, N. V. Shokhirev, P. G. Debrunner, W. R. Scheidt, J.
Am. Chem. Soc. 1996, 118, 12109.
M. R. Cheesman, F. A. Walker, J. Am. Chem. Soc. 1996, 118,
7373.
M. K. Safo, G. P. Gupta, C. T. Watson, U. Simonis, F. A.
Walker, W. R. Scheidt, J. Am. Chem. Soc. 1992, 114, 7066.
H. Ogura, L. Yatsunyk, C. J. Medforth, K. M. Smith, K. M.
Barkigia, M. W. Renner, D. Melamed, F. A. Walker, J. Am.
Chem. Soc. 2001, 123, 6564.
T. Ikeue, Y. Ohgo, T. Saitoh, M. Nakamura, H. Fujii, M.
Yokoyama, J. Am. Chem. Soc. 2000, 122, 4068.
M. Nakamura, T. Ikeue, H. Fujii, T. Yoshimura, J. Am. Chem.
Soc. 1997, 119, 6284.
T. Ikeue, Y. Ohgo, T. Yamaguchi, M. Takahashi, M. Takeda,
M. Nakamura, Angew. Chem. Int. Ed. 2001, 40, 2617.
M. W. Renner, K. M. Barkigia, Y. Zhang, C. J. Medforth, K.
M. Smith, J. Fajer, J. Am. Chem. Soc. 1994, 116, 8582.
A. Ghosh, E. Gonzalez, T. Vangberg, J. Phys. Chem. B 1999,
103, 1363.
overlap between the copper dx Ϫy orbital and the porphyrin
a2u HOMO on the other.
[3]
[4]
[5]
Experimental Section
The β-octabrominated (porphyrin)copper complexes studied here
have been described previously.[3] The β-octachlorinated (porphy-
rin)copper complexes were prepared using a slightly modified ver-
sion of the procedure described by Dolphin and co-workers[27] for
β-octachlorination of nickel porphyrins: [tetrakis(para-X-phenyl)-
porphyrin]CuII (X ϭ H or CF3) (0.3 mmol) and N-chlorosuccinim-
ide (3.6 mmol, 12 equiv.) were dissolved in o-dichlorobenzene and
heated to 140 °C. The reaction was monitored by UV/Vis spec-
troscopy and more N-chlorosuccinimide was added during the co-
urse of the reaction to ensure complete conversion into the β-oc-
tachlorinated product (the Soret absorbance changes from ca. 415
to ca. 436 nm). The total reaction time was approximately 24 h.
The resulting crude product was purified by flash chromatography
on silica gel with dichloromethane/n-hexane (2:1) as eluent and
further purified on preparative (20 ϫ 20 cm) TLC plates (silica,
0.5 mm thick) with dichloromethane/n-hexane (2:1) as eluent. EPR
and ENDOR samples were prepared by dissolving the complexes
in a mixed solvent consisting of 40% chloroform (99.9%, Aldrich)
and 60% toluene (99.8ϩ%, Aldrich) by volume. The samples were
frozen in liquid nitrogen. Powder samples of the complexes were
prepared by mixing each complex with NiTPP (which is EPR-sil-
ent) in a 1:100 ratio in toluene solution with subsequent evapor-
ation of the toluene. In the case of the rather insoluble Cu[Cl8T(p-
CF3-P)P] complex, a solid sample of this complex was diluted with
Ni[Cl8T(p-CF3-P)P] to generate a powder sample suitable for EPR
measurements. In general, the powder samples yielded spectra that
were identical to the frozen solution spectra, except that the former
exhibited slightly higher resolution. EPR spectra were obtained on
a Varian E-4 EPR spectrometer at 77 K. Two home-built X-band
and Q-band pulsed ENDOR spectrometers, which have been de-
scribed elsewhere,[28,29] were used to collect the ENDOR spectra at
4.2 K and 2 K, respectively. DFT calculations on selected saddled
(porphyrin)copper complexes (Figure 4) were carried out using
Slater-type valence triple-ζ with polarization basis sets, the VWN
local functional, PerdewϪWang 1991 gradient corrections, a spin-
unrestricted formalism, a fine mesh for numerical integration of
matrix elements, full geometry optimizations, and the ADF pro-
gram system.[30,31]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
H. S. Song, C. A. Reed, W. R. Scheidt, J. Am. Chem. Soc. 1989,
111, 6865.
[20]
[21]
T. G. Brown, B. M. Hoffman, Mole. Phys. 1980, 39, 1073.
K. L. Cunningham, K. M. McNett, R. A. Pierce, K. A. Davis,
H. H. Harris, D. M. Falck, D. R. McMillin, Inorg. Chem. 1997,
36, 608.
A. Schweiger, Angew. Chem. Int. Ed. Engl. 1991, 30, 265Ϫ292.
A. H. Maki, B. R. McGarvey, J. Chem. Phys. 1958, 29, 31.
B. R. McGarvey, J. Phys. Chem. 1967, 71, 51.
J. M. Assour, J. Chem. Phys. 1965, 43, 2477.
S. P. Greiner, D. L. Rowlands, R. W. Kreilick, J. Phys. Chem.
1992, 96, 9132.
D. Dolphin, T. Nakano, T. K. Kirk, T. P. Wijesekera, R. L.
Farrell, T. E. Malone, United States Patent, Patent Number
4,892,941, January 9, 1990.
B. M. Hoffman, V. J. DeRose, J.-L. Ong, C. E. Davoust, J.
Magn. Reson. Ser. A 1994, 110, 52.
[22]
[23]
[24]
[25]
[26]
[27]
Acknowledgments
[28]
[29]
[30]
This work was supported by the VISTA program of Statoil and the
Norwegian Research Council (E. S. and A. G.) and the NSF (J. S.
and B. M. H.). We thank Prof. Joshua Telser, Roosevelt University,
for helpful comments on the manuscript.
C. E. Davoust, P. E. Doan, B. M. Hoffman, J. Mag. Reson.
1996, 119, 38.
G. T. Velde, F. M. Bickelhaupt, E. J. Baerends, C. F. Guerra,
S. J. A. Van Gisbergen, J. G. Snijders, T. Ziegler, J. Comput.
Chem. 2001, 22, 931.
[1]
[31]
J. A. Shelnutt, X.-Z. Song, J.-G. Ma, S.-L. Jia, W. Jentzen, C.
J. Medforth, Chem. Soc., Rev. 1998, 27, 31.
F. D’Souza, M. E. Zandler, P. Tagliatesta, Z. P. Ou, J. G. Shao,
S. J. A. Van Gisbergen, J. G. Snijders, E. J. Baerends, Comput.
Phys. Commun. 1999, 118, 119.
[2]
Received May 4, 2004
Eur. J. Inorg. Chem. 2005, 1609Ϫ1615
www.eurjic.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1615