Dendritic iron porphyrins with tethered axial ligands: New model compounds for cytochromes

Full text of DOI:10.1002/(SICI)1521-3773(19991102)38:21<3215::AID-ANIE3215>3.0.CO;2-S

COMMUNICATIONS
protein shell[8] have not been intensively investigated using
model compounds and are less well understood.
[13] Crystal data for 15: Mr ˆ 1001.44, monoclinic, space group P21, a ˆ
1073.7(1), b ˆ 2317.0(2), c ˆ 1756.3(1) Š, b ˆ 97.03(1)8, Vˆ
3
4336.4(6) Š3, Z ˆ 4, 1calcd ˆ 1.534 Mgm
,
MoKa radiation (l ˆ
We have already reported the use of dendritic iron
porphyrins[9, 10] as model compounds for cytochromes in
which the protein shell around the buried electroactive core
is mimicked by the dendritic superstructure. These investiga-
tions revealed a strong correlation between the redox
potential and the degree of dendritic branching. In these
early systems, however, the nature of the axial ligation to the
iron center, which is known to have a very strong influence on
the redox properties[3] was not controlled. Therefore, the
observed shifts in redox potential caused by the dendritic shell
could not be quantified independently from axial ligation
effects and no general conclusions concerning the effects of
the dendritic superstructure could be drawn.
We now present a new series of dendritic cytochrome
mimics, which contain a defined and stable axial ligation
pattern. This allows, for the first time, a quantitative evalua-
tion of the effect of an insulating dendritic shell on the
redox properties of the embedded iron porphyrin core. The
three novel dendrimers of generation zero ([1 ´ Fe]Cl), one
([2 ´ Fe]Cl), and two ([3 ´ Fe]Cl) feature controlled axial
ligation at the iron center by two imidazoles tethered to the
porphyrin core. This stable ligation pattern, which is kineti-
cally inert towards coordinating solvents, is found in the
cytochrome b5 family of electron transfer proteins.[11] The
optimal length of the alkyl tethers between the iron-coordi-
0.71073 Š), m ˆ 0.847 mm 1. Data were collected on a STOE IPDS
system at 193 K. The structure was solved by direct methods and
refined on F20 by full-matrix least-squares methods (SHELXS-97,
SHELXL-97, SHELDRICK, 1997). All non-hydrogen atoms were
refined anisotropically. wR2 ˆ 0.0893 (all unique data), R1 ˆ 0.0367
for data with I > 2s(I). Crystallographic data (excluding structure
factors) for the structure reported in this paper have been deposited
with the Cambridge Crystallographic Data Centre as supplementary
publication no. CCDC-134965. Copies of the data can be obtained free
of charge on application to CCDC, 12 Union Road, Cambridge
CD21EZ, UK (fax: (‡44)1223-336-033; e-mail: deposit@ccdc.cam.
ac.uk).
[14] Selected 31P NMR data (81 MHz, CDCl3) of the phosphane ligands 2
and their complexes with RhI: 2a: d ˆ 16.7 (d, JP,P ˆ 19.1 Hz), 23.2
(d, JP,P ˆ 19.1 Hz); [Rh(nbd)(2a)]BF4: d ˆ 23.1 (dd, JP,P ˆ 22.0, JP,Rh
129.7 Hz), 8.6 (dd, JP,P ˆ 22.0, JP,Rh ˆ 117.6 Hz); 2b: d ˆ 17.1 (d, JP,P
ˆ
ˆ
20.3 Hz), 22.4 (d, JP,P ˆ 20.3 Hz); [Rh(nbd)(2b)]BF4: d ˆ 24.4 (dd,
JP,P ˆ 30.5, JP,Rh ˆ 155.1 Hz), 5.1 (dd, JP,P ˆ 30.5, JP,Rh ˆ 153.2 Hz); 2c:
d ˆ 12.9
(d,
JP,P ˆ 18.4 Hz),
22.4
(d,
JP,P ˆ 18.4 Hz);
[Rh(nbd)(2c)]BF4: d ˆ 17.9 (dd, JP,P ˆ 23.5, JP,Rh ˆ 101.9 Hz), 16.0
(dd, JP,P ˆ 23.5, JP,Rh ˆ 101.9 Hz); 2d: d ˆ 17.8 (d, JP,P ˆ 26.7 Hz),
22.8 (d, JP,P ˆ 26.7 Hz); [Rh(nbd)(2d)]BF4: d ˆ 24.8 (dd, JP,P ˆ 31.1,
JP,Rh ˆ 155.1 Hz), 12.3 (dd, JP,P ˆ 31.1, JP,Rh ˆ 155.1 Hz).
RHNOC
CONHR
[1·Fe]Cl R =
O
O O
Dendritic Iron Porphyrins with Tethered
Axial Ligands: New Model Compounds for
Cytochromes**
N
O
O
O
O
N
O
O
O
O
O
N Fe N
N
N
O
O
N
[2·Fe]Cl R =
O
O
Philipp Weyermann, Jean-Paul Gisselbrecht, Corinne
Boudon, François Diederich,* and Maurice Gross
O
O
O
O
N
O
CONHR
RHNOC
O
O
O
O
O
O
Cl
O
Dedicated to Professor Jean-Marie Lehn
on the occasion of his 60th birthday
O
O
O
O
O
O
O
O
O
O
O
O
The most intriguing characteristics of the cytochrome
family of electron transfer proteins is the very broad range
of redox potentials featured by the FeIII/FeII couple at the
electroactive heme core.[1] A variety of model studies have
identified a dependency of this potential from the nature of
substituents at the porphyrin ring,[2] axial ligation to the iron
center,[3] hydrogen bonding to the axial ligands,[4] and ruffling
of the porphyrin macrocycle.[5] In contrast, the influence of
environmental effects such as heme solvation,[6] polarity of the
heme microenvironment,[7] and the nature of the surrounding
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
HN
NH
O
O
O
HN
NH
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O O
O
O O
HN
O
NH
O
HN
NH
O
O
O
O
O
O
N
O
O
O
O
O
O
O
O
O
O
O
O
O
O
N
N Fe N
N
N
N
O
O
O
O
O
O
O
N
O
O
O
O
O
O
O
O
O
O
O
O
O
HN
O
[*] Prof. Dr. F. Diederich, Dipl.-Chem. P. Weyermann
Laboratorium für Organische Chemie, ETH-Zentrum
Universitätstrasse 16, CH-8092 Zürich (Switzerland)
Fax: (‡41)1-632-1109
NH
O
O
HN
NH
O
O
O
O O O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
Cl
NH
HN
O
O
O
O
HN
NH
O
O
O
O
O
O
O
O
O
O
O
O
E-mail: diederich@org.chem.ethz.ch
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
Dr. J. P. Gisselbrecht, Dr. C. Boudon, Prof. Dr. M. Gross
Laboratoire dꢁElectrochimie et de Chimie Physique du Corps Solide
Faculte de Chimie, Universite Louis Pasteur and CNRS, UMR
no. 7512
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
4, rue Blaise Pascal, F-67000 Strasbourg (France)
O
O
[3·Fe]Cl
[**] This work was supported by the ETH Research Council.
Angew. Chem. Int. Ed. 1999, 38, No. 21
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
1433-7851/99/3821-3215 $ 17.50+.50/0
3215

COMMUNICATIONS
give meso-dibromoporphyrin 8 ´ H2, partial reduction with
nBu3SnH to give monobromoporphyrin 5 ´ H2, and metal-
ation. Phenylboronic ester 6 was prepared from 4-bromo-2,6-
dimethoxytoluene 9[14] by cleavage of the methyl ethers (conc.
HI; !10) reprotection of the phenolic hydroxy groups with
chloromethyl methyl ether (MOM-Cl; !11), metalation
(nBuLi) and boronic ester formation (B(OMe)3, then pinacol;
!12), cleavage of the MOM protecting groups (HCl; !13),
and alkylation with 1-(6-bromohexyl)imidazole (14).[15] The
FeIII complex [4´Fe]Cl was obtained from 4 ´ Zn by acid-
induced demetalation (!4´H2) and subsequent insertion of
FeII (FeCl2), followed by air oxidation.
Comparison of the FeIII complex [4 ´ Fe]Cl with the six-
coordinate FeIII complex [7 ´ Fe(N-MeIm)2]Cl[16] (N-MeIm ˆ
1-methylimidazole) unambiguously showed complete intra-
molecular axial coordination of the two tethered imidazoles
under formation of a paramagnetic low-spin complex. The
UV/Vis[17] and EPR spectra[18] as well as the magnetic
moments,[19] determined by the Evans ± Scheffold method[20]
of both compounds compared very well with those of other
bis-imidazole-coordinated FeIII porphyrins. Reduction of [4 ´
Fe]Cl, similar to that of [7 ´ Fe(N-MeIm)2]Cl, with Na2S2O4
produced a diamagnetic low-spin complex with a hemo-
chrome absorption spectrum which was readily identified as
that of a six-coordinate FeII species.[21]
nating imidazoles and the phenyl ring at the porphyrin core
was designed with the help of molecular modeling.[12] The
second-generation compound [3 ´ Fe]Cl (11719 Da) is com-
parable in mass to typical single-heme cytochromes such as
cytochrome c (tuna, 11384 Da),[13] or cytochrome b5 (bovine,
15198 Da).[11] With their triethyleneglycol monomethyl ether
surface groups, the three dendritic mimics are soluble in
solvents of widely differing polarity.
The key intermediate in the synthesis of the dendritic
porphyrins was the bis-imidazole-appended zinc porphyrin 4 ´
Zn, which was obtained in high yield by a Suzuki coupling
between the brominated zinc porphyrin 5 ´ Zn and the bis-
imidazole-appended phenylboronic ester 6 (Scheme 1). The
meso-bromoporphyrin 5 ´ Zn was obtained from precursor 7 ´
[9c]
H2 by dibromination with N-bromosuccinimide (NBS) to
For the preparation of the dendritic porphyrins of gener-
ation zero (1 ´ Zn), one (2 ´ Zn), and two (3 ´ Zn), the core
tetraacid 15 ´ Zn, obtained by hydrolysis of 4 ´ Zn, was coupled
to the corresponding dendritic wedges 16 ± 18,[22] respectively,
using HATU as coupling reagent (Scheme 2). They were
purified by preparative gel permeation chromatography
(GPC, Biorad Biobeads SX-1, CH2Cl2) and fully characterized
by standard spectroscopic methods (Table 1). In the ZnII
porphyrin derivatives, one imidazole is complexed to the
metal ion forming a dynamic five-coordinate species. Deme-
talation to the highly air- and light-sensitive free-base
porphyrins and iron insertion finally yielded the target
compounds [1 ´ Fe]Cl, [2 ´ Fe]Cl, and [3 ´ Fe]Cl which were
purified by GPC (Biorad Biobeads SX-3, CH2Cl2) and shown
by MALDI-TOF mass spectrometry to be free of any
structural defects. According to UV/Vis and EPR spectro-
scopy, they are six-coordinate low-spin complexes with double
axial imidazole ligation (Table 1). Reduction with Na2S2O4 in
various solvents led instantaneously to typical low-spin FeII
porphyrin absorption spectra, regardless of the size of the
dendrimer.
The redox properties of [1 ´ Fe]Cl ± [3 ´ Fe]Cl were first
investigated in the rather nonpolar solvent CH2Cl2 using
cyclic (CV) and steady-state voltammetry (SSV). All three
compounds showed a reversible one-electron reduction step
(Figure 1) which was clearly assigned to the FeIII/FeII couple
by spectroelectrochemical methods: the UV/Vis spectra of
the electrochemically reduced species were identical to those
obtained by chemical reduction, and well defined isosbestic
points evolved during controlled potential electrolysis. The
generation zero complex [1 ´ Fe]Cl exhibited a redox potential
of 0.21 V (vs. SCE; Table 2). This is in the expected range
for a bis-imidazole-ligated iron porphyrin complex.[23] In
the higher generation compounds [2 ´ Fe]Cl (‡0.08 V) and
Scheme 1. Synthesis of the bis-imidazole-ligated FeIII porphyrin core
[4 ´ Fe]Cl. a) NBS (2.0 equiv), CHCl3, RT, 30 min, 91%; b) nBu3SnH
(1.5 equiv), AIBN (0.1 equiv), PhH, reflux, 4 h, 45%; c) Zn(OAc)2
(10.0 equiv), CHCl3/MeOH, RT, 4 h, 98%; d) conc. HI (excess), AcOH,
reflux, 6 h, 95%; e) MOM-Cl (4.0 equiv), K2CO3 (8.0 equiv), MeCN, 08C,
30 min, 98%; f) nBuLi (1.6 equiv), TMEDA (1.6 equiv), THF, 788C,
60 min, then B(OMe)3 (5.0 equiv), RT, 2 h, then pinacol (10.0 equiv), PhH,
reflux, 12 h, 96%; g) conc. HCl (excess), THF/MeOH, RT, 3 d, 66%; h) 14
(5.0 equiv), Cs2CO3 (10.0 equiv), DMF, RT, 4 h, 49%; i) 6 (2.0 equiv),
[Pd(PPh3)4] (0.1 equiv), Cs2CO3 (8.0 equiv), PhMe, 908C, 4 h, 82%;
j) CF3COOH (excess), CHCl3, 08C, 5 min, 78%; k) FeCl2 (10.0 equiv),
2,6-lutidine (10.0 equiv), THF, reflux, 2 h, then 1% HCl in CHCl3 (excess),
RT, 5 min, then ªproton spongeº (excess), THF, RT, 15 min, 66%. AIBN ˆ
aa'-azobis(isobutyronitrile);
MOM ˆ methoxymethyl;
TMEDA ˆ
N,N,N',N'-tetramethylethylenediamine; DMF ˆ N,N-dimethylformamide;
ªproton spongeº ˆ 1,8-bis(dimethylamino)naphthalene.
3216
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
1433-7851/99/3821-3216 $ 17.50+.50/0
Angew. Chem. Int. Ed. 1999, 38, No. 21

COMMUNICATIONS
generation one dendrimer showing a very
weak spread-out reduction step and the
second-generation compound being electro-
chemically silent.[10a] Therefore, we changed
to chemical methods to determine the redox
potential for the FeIII/FeII couple in this
solvent. UV/Vis spectroscopic investigations
showed that equilibria with solutions of
suitable reducing agents were rapidly estab-
lished after mixing with the porphyrin
dendrimers. This allowed an accurate, highly
reproducible determination of the redox
potentials from equilibrium measure-
O
O
16
N
H2N
O
b)
1·Zn
[1·Fe]Cl
N
c)
c)
c)
RO2C
CO2R
O
O
O
O
O
H2N
O
O
17
O
O
O
b)
3
N Zn N
2·Zn
[2·Fe]Cl
N
N
O
O
H2N
O
O
O
18
3
O
N
O
O
3
RO2C
CO2R
H
O
3·Zn
[3·Fe]Cl
O
b)
4·Zn R = Et
15·Zn R = H
N
a)
N
Scheme 2. Synthesis of the dendritic cytochrome mimics [1 ´ Fe]Cl ± [3 ´ Fe]Cl. a) NaOH
(excess), dioxane/H2O, RT, 3 d; b) 16, 17, or 18 (12.0 equiv), HATU (6.0 equiv), Et3N
(24.0 equiv), DMF, 08C, 24 h, 85% (1 ´ Zn); 3 d, 65% (2 ´ Zn); 7 d, 42% (3 ´ Zn); all yields
starting from 4 ´ Zn; c) TFA (excess), CHCl3, 08C, 5 min, then FeCl2 (10.0 equiv), 2,6-lutidine
(10.0 equiv), THF, reflux, 4 h, then 1% HCl in CHCl3 (excess), RT, 5 min, then ªproton
spongeº (excess), THF, RT, 15 min, 68% ([1 ´ Fe]Cl); 73% ([2 ´ Fe]Cl); 78% ([3 ´ Fe]Cl).
HATU ˆ O-(7-azabenzotriazol-1-yl)-N,N,N',N'-tetramethyluronium hexafluorophosphate.
4
[ 24a, b]
ments using [Fe(ox)3] /[Fe(ox)3]
or
4
3 [24c, d]
[Fe(CN)6] /[Fe(CN)6]
(Figure 2), as it
is commonly done with proteins.[25] The
potentials measured for [1 ´ Fe]Cl by chemical
and electrochemical methods were in excel-
lent agreement. In contrast to the results
obtained in the organic solvents, the potential
of the FeIII/FeII couple hardly shifts upon
Table 1. Selected physical and spectroscopic data of 3 ´ Zn and [3 ´ Fe]Cl.[a]
3 ´ Zn: Viscous purple oil. Rf ˆ 0.48 (SiO2, CH2Cl2/MeOH 90:10); UV/Vis
(CHCl3): lmax(e) ˆ 595 (5100), 559 (20700), 520 (4200), 427 (544900), 408
(sh, 63600), 311 nm (22500); IR (CHCl3): nÄ ˆ 2880m, 1735s, 1670m, 1580w,
1520m, 1460m, 1380w, 1355w, 1325w, 1245m, 1185s, 1110s, 1030w, 990w,
1
950w, 855m, 810w, 620w cm
;
1H NMR (500 MHz, [D5]pyridine, 300 K):
d ˆ 10.20 (s, 1H), 9.47 (d, J ˆ 4.0 Hz, 2H), 9.28 ± 9.32 (m, 4H), 9.23 (d, J ˆ
4.0 Hz, 2H), 8.05 (br. s, 2H), 8.02 (t, J ˆ 7.9 Hz, 2H), 7.52 (s, 2H), 7.49 (d,
J ˆ 7.9 Hz, 4H), 7.40 (br, s, 2H), 7.31 (br, s, 2H), 6.30 (br, s, 16H), 4.48 ±
4.52 (m, 4H), 4.41 ± 4.45 (m, 4H), 4.32 ± 4.40 (m, 72H), 4.10 ± 4.22 (m, 8H),
3.47 ± 4.05 (m, 576H), 3.29 (s, 108H), 2.69 (t, J ˆ 6.2 Hz, 72H), 2.67 (s, 3H),
1.55 ± 1.73 (m, 24H), 1.42 ± 1.54 (m, 4H), 1.21 ± 1.32 (m, 4H); 13C NMR
(125 MHz, [D5]pyridine, 300 K): d ˆ 172.4, 171.7, 171.4, 160.6, 155.9, 151.3,
151.2, 150.3, 150.2, 142.8, 136.7, 132.0, 132.0, 131.9, 131.4, 130.7, 128.3, 122.1,
120.6, 120.3, 113.2, 113.1, 113.0, 106.2, 105.3, 72.3, 70.8, 70.8, 70.7, 69.7, 69.7,
69.3, 68.7, 68.4, 68.2, 67.3, 64.0, 60.6, 60.6, 58.7, 47.4, 37.3, 35.2, 32.1, 31.1,
29.4, 26.4, 25.9, 24.9, 9.2; MALDI-TOF-MS (2-(4'-hydroxyphenylazo)ben-
‡
zoic acid): m/z (%): 11750.8 (100, [M ‡ Na] , calcd for C533H918N24O250Zn ´
‡
‡
‡
Na : 11750.9),[b] 11729.5 (37, M , calcd for C533H918N24O250Zn : 11727.9).[b]
[3 ´ Fe]Cl: Viscous brown oil. Rf ˆ 0.32 (SiO2, CH2Cl2/MeOH 90:10); UV/
Vis (CHCl3): lmax(e) ˆ 543 (10200), 415 (128000), 313 nm (25400); IR
(CHCl3): nÄ ˆ 2915s, 2880s, 1735s, 1670m, 1580w, 1515m, 1460m, 1380w,
1
1350w, 1320w, 1240m, 1185s, 1110s, 1030w, 995w, 950w, 850w, 620w cm
;
EPR (X-Band, CHCl3, 77 K): gx ˆ 1.558, gy ˆ 2.321, gz ˆ 2.890; MALDI-
TOF-MS (2-(4'-hydroxyphenylazo)benzoic acid): m/z (%): 11719.2 (100,
‡
‡
[M Cl] , calcd for C533H918N24O250Fe : 11718.9).[b]
[a] All new compounds were fully characterized by 1H and 13C NMR
(except Fe complexes), IR, UV/Vis and FAB or MALDI-TOF mass spectra
as well as by elemental analysis or high-resolution mass spectra. For all Fe
complexes, EPR spectra were recorded and in some cases magnetic
moments were calculated by the method of Evans and Scheffold.
[b] Calculated highest peak in the isotope pattern.
Figure 1. Cyclic voltammograms showing the first reduction step of
[1 ´ Fe]Cl (a, d), [2 ´ Fe]Cl (b, e), and [3 ´ Fe]Cl (c, f) in CH2Cl2 (left) and
MeCN (right). For the conditions, see Table 2.
[3 ´ Fe]Cl (‡0.10 V), the FeIII !FeII reduction becomes great-
ly facilitated, with the largest change in potential occurring
between generation zero and one.
CV measurements in MeCN, a solvent of intermediate
polarity, showed a similar trend: upon changing from gen-
eration zero ( 0.24 V) to two (‡0.09 V), reduction of FeIII
becomes increasingly favored (by 330 mV) with the largest
change in potential again occurring at the stage of the
generation one dendrimer ( 0.01 V).
changing from generation zero ( 0.29 V) to one ( 0.25 V)
but a dramatic increase is observed upon changing to the
generation two dendrimer [3 ´ Fe]Cl (‡0.09 V) (see Table 2).
For comparison, the FeIII/FeII redox potential of cyto-
chrome b5 in H2O was determined as ‡0.24 V vs. SCE.[11b]
The results of the redox studies in the three solvents are
summarized in Figure 3: 1) In all solvents, the redox potential
of the FeIII/FeII couple becomes more positive with increasing
dendritic generation. Remarkably, the potential of the
In H2O, the redox process could only be monitored
electrochemically for the generation zero complex, with the
Angew. Chem. Int. Ed. 1999, 38, No. 21
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
1433-7851/99/3821-3217 $ 17.50+.50/0
3217

COMMUNICATIONS
Table 2. Redox potentials (in V vs. SCE) of the FeIII/FeII couple of
[1 ´ Fe]Cl ± [3 ´ Fe]Cl in different solvents.
Porphyrin
E(FeIII/FeII)
MeCN[a]
[a]
CH2Cl2
H2O
[1 ´ Fe]Cl
[2 ´ Fe]Cl
[3 ´ Fe]Cl
0.21
‡ 0.08
‡ 0.10
0.24
0.01
‡ 0.09
0.29[b]
0.25[b]
‡ 0.09[c]
[a] Values from CV approximated as E1/2 ˆ (Epa Epc)/2; supporting
electrolyte 0.1m Bu4NPF6; glassy carbon working electrode, Ag/AgCl
reference electrode, platinum wire counter electrode; Tˆ 298 K; scan
rate ˆ 0.1 Vs 1; typical concentration 5  10 4 m; ferrocene was used as an
internal standard, and the redox potentials referenced against the standard
‡
calomel electrode (SCE) using published values for the Fc/Fc couple in
CH2Cl2 (0.46 V vs. SCE; N. G. Connelly, W. E. Geiger, Chem. Rev. 1996, 96,
877 ± 910) and MeCN (0.45 V vs. SCE; W. G. Barrette, Jr., H W. John-
son, Jr., D. T. Sawyer, Anal. Chem. 1984, 56, 1890 ± 1898). [b] Values from
3
equilibrium measurements with [Fe(ox)3] 4/[Fe(ox)3] (ox ˆ oxalate) as
reference compound.[26] [c] Values from equilibrium measurements with
[Fe(CN)6] 4/[Fe(CN)6] 3 as reference compound.[26]
Figure 3. Plot of the redox potentials (in V vs. SCE) of the FeIII/FeII couple
in [1 ´ Fe]Cl ± [3 ´ Fe]Cl in CH2Cl2, MeCN and H2O vs. dendrimer gener-
ation G.
shift to more positive potential occurs upon changing from the
generation zero to the generation one complex. Clearly, the
special microenvironment is already largely created by the
first generation branching in these solvents. 3) In sharp
contrast, the redox potential in H2O does not vary much
upon passing from the generation zero to the generation one
complex. Solvation effects are much more pronounced in
H2O, and, as we had previously suggested,[9b,c] the relatively
open dendritic branches in [2 ´ Fe]Cl do not impede access of
bulk solvent to the central core for stabilization of the FeIII
state. However in [3 ´ Fe]Cl, the dendritic superstructure is
sufficiently dense to prevent the contact between the por-
phyrin and the external bulk solvent, thereby creating the
same, unique core microenvironment as in the organic
solvents.
This model study confirms the large contributions of the
densely packed protein shell to the strong positive shifts of the
FeIII/FeII potential in cytochromes[7a,b] and firmly establishes
well-designed dendrimers as powerful mimics for globular
proteins.
3
Figure 2. Redox equilibria in H2O between [Fe(ox)3] 4/[Fe(ox)3] and
3
[1 ´ Fe]Cl (a) and [2 ´ Fe]Cl (b) and between [Fe(CN)6] 4/[Fe(CN)6] and
[3 ´ Fe]Cl (c). Spectral evolution of the Soret band region in the UV/Vis
spectrum during reductive titration (left) and nonlinear least-squares fit of
the titration data to the Nernst equation, yielding the redox potential
(right). For the conditions, see Table 2. cox/cred is the concentration ratio of
oxidized and reduced porphyrin.
Received: June 1, 1999 [Z13494IE]
German version: Angew. Chem. 1999, 111, 3400 ± 3404
Keywords: dendrimers ´ electrochemistry ´ porphyrinoids ´
redox chemistry ´ voltammetry
second-generation complex [3 ´ Fe]Cl is, within experimental
error, identical in all three solvents (of extremely different
polarity), which clearly demonstrates that the dendritic
branching creates a unique local microenvironment around
the isolated electroactive core. Therefore, the dendritic shell
fully mimics the protecting peptide shell which modulates the
redox potential of the FeIII/FeII couple in a similar way in
cytochromes.[7a,b] 2) In the two organic solvents, the largest
[1] a) G. R. Moore, G. W. Pettigrew, Cytochromes-c: Evolutionary,
Structural and Physicochemical Aspects, Springer, Berlin, 1990;
b) F. S. Mathews, Prog. Biophys. Mol. Biol. 1985, 45, 1 ± 56.
[2] a) K. M. Kadish, M. M. Morrison, L. A. Constant, L. Dickens, D. G.
Davis, J. Am. Chem. Soc. 1976, 98, 8387 ± 8390; b) A. Giraudeau, H. J.
Callot, J. Jordan, I. Ezhar, M. Gross, J. Am. Chem. Soc. 1979, 101,
3857 ± 3862.
3218
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
1433-7851/99/3821-3218 $ 17.50+.50/0
Angew. Chem. Int. Ed. 1999, 38, No. 21

COMMUNICATIONS
[3] a) H. A. Harbury, J. R. Cronin, M. W. Fanger, T. P. Hettinger, A. J.
Murphy, Y. P. Myer, S. N. Vinogradov, Proc. Natl. Acad. Sci. USA
1965, 54, 1658 ± 1664; b) P. K. Warme, L. P. Hager, Biochemistry 1970,
9, 1606 ± 1614; c) J. C. Marchon, T. Mashiko, C. A. Reed, Electron
Transport and Oxygen Utilization (Ed.: C. Ho), Elsevier, North
Holland, 1982, 67 ± 72; d) S. G. Sligar, K. D. Egeberg, J. T. Sage, D.
Morikis, P. M. Champion, J. Am. Chem. Soc. 1987, 109, 7896 ± 7897;
e) A. L. Raphael, H. B. Gray, J. Am. Chem. Soc. 1991, 113, 1038 ±
1040.
[22] T. Habicher, F. Diederich, V. Gramlich, Helv. Chim. Acta, 1999, 82,
1066 ± 1095.
[23] a) L. A. Constant, D. G. Davis, Anal. Chem. 1975, 47, 2253 ± 2260;
b) K. M. Kadish, L. A. Bottomley, Inorg. Chem. 1980, 19, 832 ± 836;
c) M. J. M. Nesset, N. V. Shokhirev, P. D. Enemark, S. E. Jacobsen,
F. A. Walker, Inorg. Chem. 1996, 35, 5188 ± 5200.
[24] a) J. J. Lingane, J. Am. Chem. Soc. 1946, 68, 2448 ± 2453; b) W. B.
Schaap, H. A. Laitinen, J. C. Bailar Jr., J. Am. Chem. Soc. 1954, 76,
5868 ± 5872; c) R. C. Murray Jr., P. A. Rock, Electrochim. Acta 1968,
13, 969 ± 975; d) G. I. H. Hanania, D. H. Irvine, W. A. Eaton, P.
George, J. Phys. Chem. 1967, 71, 2022 ± 2030.
[4] a) R. Quinn, J. Mercer-Smith, J. N. Burstyn, J. S. Valentine, J. Am.
Chem. Soc. 1984, 106, 4136 ± 4144; b) P. OꢁBrien, D. A. Sweigart,
Inorg. Chem. 1985, 24, 1405 ± 1409.
[25] L. P. Dutton, Methods Enzymol. 1978, 54, 411 ± 435.
[5] a) K. M. Barkigia, L. Chantranupong, K. M. Smith, J. Fajer, J. Am.
Chem. Soc. 1988, 110, 7566 ± 7567; b) K. M. Kadish, E. van Caemel-
becke, P. Boulas, F. DꢁSouza, E. Vogel, M. Kisters, C. J. Medforth,
K. M. Smith, Inorg. Chem. 1993, 32, 4177 ± 4178.
[6] a) E. Stellwagen, Nature 1978, 275, 73 ± 74; b) K. M. Kadish, L. A.
Bottomley, S. Kelly, D. Schaeper, L. R. Shiue, Biolelectrochem.
Bioenerg. 1981, 8, 213 ± 222; c) L. A. Bottomley, K. M. Kadish, Inorg.
Chem. 1981, 20, 1348 ± 1357.
[26] A stock solution of fully oxidized porphyrin (typical concentration
5 Â 10 6 m) was titrated with a stock solution of fully reduced reference
compound (typical concentration 2.5 Â 10 4 m), and the concentrations
of all species in solution calculated from the spectrophotometrically
determined ratio of oxidized to reduced porphyrin and known total
concentrations; Tˆ 298 K; E values versus the normal hydrogen
electrode (NHE) were obtained from a nonlinear least-squares fit of
the series of equilibria produced by the titration to the Nernst
equation and referenced for comparison against SCE (0.24 V vs.
NHE; D. T. Sawyer, A. Sobkowiak, J. L. Roberts, Electrochemistry for
Chemists, 2nd ed., Wiley, New York, 1995, Chap. 5, pp. 170 ± 248).
Titrations were reproduced with excellent agreement in triplicate
runs.
[7] a) R. J. Kassner, Proc. Natl. Acad. Sci. USA 1972, 69, 2263 ± 2267;
b) R. J. Kassner, J. Am. Chem. Soc. 1973, 95, 2674 ± 2677; c) A. K.
Churg, A. Warshel, Biochemistry 1986, 25, 1675 ± 1681; d) D. Lexa, M.
Â
Momenteau, P. Rentien, G. Rytz, J. M. Saveant, F. Xu, J. Am. Chem.
Soc. 1984, 106, 4755 ± 4765; e) C. Gueutin, D. Lexa, M. Momenteau,
Â
J. M. Saveant, F. Xu, Inorg. Chem. 1986, 25, 4294 ± 4307.
[8] a) M. S. Caffrey, M. A. Cusanovich, Arch. Biochem. Biophys. 1991,
285, 227 ± 230; b) K. K. Rodgers, S. G. Sligar, J. Am. Chem. Soc. 1991,
113, 9419 ± 9421.
[9] a) P. J. Dandliker, F. Diederich, M. Gross, C. B. Knobler, A. Louati,
E. M. Sanford, Angew. Chem. 1994, 106, 1821 ± 1824; Angew. Chem.
Int. Ed. Engl. 1994, 33, 1739 ± 1742; b) P. J. Dandliker, F. Diederich,
J. P. Gisselbrecht, A. Louati, M. Gross, Angew. Chem. 1995, 107,
2906 ± 2909; Angew. Chem. Int. Ed. Engl. 1995, 34, 2725 ± 2728; c) P. J.
Dandliker, F. Diederich, A. Zingg, J. P. Gisselbrecht, M. Gross, A.
Louati, E. Sanford, Helv. Chim. Acta 1997, 80, 1773 ± 1801.
Novel U-Shaped Systems Containing an Imide-
Functionalized Cleft for the Study of Solvent-
Mediated Electron Transfer and Energy
Transfer: Synthesis and Binding Studies**
[10] For other dendritic porphyrins, see: a) K. W. Pollak, J. W. Leon,
Nicholas J. Head, Anna M. Oliver, Kai Look, Nigel R.
Lokan, Garth A. Jones, and Michael N. Paddon-Row*
Â
Ä
J. M. J. Frechet, M. Maskus, H. D. Abruna, Chem. Mater. 1998, 10, 30 ±
38; b) D. L. Jiang, T. Aida, J. Am. Chem. Soc. 1998, 120, 10895 ±
10901; c) P. Bhyrappa, G. Vaijayanthimala, K. S. Suslick, J. Am. Chem.
Soc. 1999, 121, 262 ± 263; d) S. A. Vinogradov, L. W. Lo, D. F. Wilson,
Chem. Eur. J. 1999, 5, 1338 ± 1347; e) U. Puapaiboon, R. T. Taylor,
Rapid Commun. Mass Spectrom. 1999, 13, 508 ± 515.
One of the most intensively investigated issues of long-
range electron transfer (ET) and energy transfer (EnT) is how
the electronic coupling for these processes depends on the
nature of the intervening medium between a pair of chromo-
phores.[1] Using structurally well-defined systems, consider-
able progress has been made in delineating the characteristics
of electronic coupling involving saturated hydrocarbon
bridges,[1a±c] protein-like pathways,[1d] and the p stacks of base
pairs in DNA molecules.[1e]
By contrast, solvent-mediated electronic coupling remains
a vexing issue, owing to the dynamic, jostling nature of the
solvent molecules which blur the electronic coupling path-
ways. Progress has recently been made by using novel rigid
U-shaped multichromophoric systems, represented by Class I
in Figure 1,[2] in which the terminal chromophores face each
other across a ªrigidº cavity within which a certain number of
solvent molecules are present (in a dynamic sense). However,
because these systems still do not address the problem of
[11] a) F. S. Mathews, M. Levine, P. Argos, Nature 1971, 233, 15 ± 16; b) S. F.
Velick, P. Strittmatter, J. Biol. Chem. 1955, 221, 265 ± 275.
[12] Molecular modeling was performed on a simple FeII tris(meso-
phenyl)porphyrin, using the UFF force field implemented in Cerius2,
Version 3.0, BIOSYM Technologies, Inc., San Diego, CA, 1997.
[13] R. Swanson, B. L. Trus, N. Mandel, G. Mandel, O. B. Kallai, R. E.
Dickerson, J. Biol. Chem. 1977, 252, 759 ± 775.
[14] F. P. Doyle, J. H. C. Nayler, H. R. J. Waddington, J. C. Hanson, G. R.
Thomas, J. Chem. Soc. 1963, 497 ± 506.
Á
[15] F. Montanari, M. Penso, S. Quici, P. Vigano, J. Org. Chem. 1985, 50,
4888 ± 4893.
[16] Directly prepared by dissolving [7 ´ Fe]Cl (obtained from 7 ´ H2 by iron
insertion with FeCl2) in the appropriate solvent, containing 0.1m N-
MeIm.
[17] a) H. Kobayashi, T. Higuchi, Y. Kaizu, H. Osada, M. Aoki, Bull.
Chem. Soc. Jpn. 1975, 48, 3137 ± 3141; b) F. A. Walker, M. W. Lo, M. T.
Ree, J. Am. Chem. Soc. 1976, 98, 5552 ± 5560.
[18] a) J. Peisach, W. E. Blumberg, A. Adler, Ann. N.Y. Acad. Sci. 1973,
206, 310 ± 327; b) G. N. La Mar, F. A. Walker, J. Am. Chem. Soc. 1973,
95, 1782 ± 1790; c) R. Quinn, M. Nappa, J. S. Valentine, J. Am. Chem.
Soc. 1982, 104, 2588 ± 2595.
[19] a) L. M. Epstein, D. K. Straub, C. Maricondi, Inorg. Chem. 1967, 6,
1720 ± 1724; b) K. M. Adams, P. G. Rasmussen, W. R. Scheidt, K.
Hatano, Inorg. Chem. 1979, 18, 1892 ± 1899.
[*] Prof. M. N. Paddon-Row, Dr. N. J. Head, Dr. A. M. Oliver,
Dr. K. Look, N. R. Lokan, G. A. Jones
School of Chemistry
University of New South Wales
Sydney, N.S.W., 2052 (Australia)
Fax: (‡61)2-9385-6141
[20] a) D. F. Evans, J. Chem. Soc. 1959, 2003 ± 2005; b) J. Löliger, R.
Scheffold, J. Chem. Educ. 1972, 49, 646 ± 647.
[21] a) H. Kobayashi, M. Shimizu, I. Fujita, Bull. Chem. Soc. Jpn. 1970, 43,
2335 ± 2341; b) H. Kobayashi, Y. Yanagawa, Bull. Chem. Soc. Jpn.
1972, 45, 450 ± 456.
E-mail: m.paddonrow@unsw.edu.au
[**] We thank the Australian Research Council for support and for the
award of a Senior Research Fellowship (to MNP-R).
Angew. Chem. Int. Ed. 1999, 38, No. 21
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
1433-7851/99/3821-3219 $ 17.50+.50/0
3219