Oxo- and hydroxo-bridged heme-copper assemblies formed from acid-base or metal-dioxygen chemistry

Article abstract of DOI:10.1021/ic981431+

The iron-copper dinuclear active site in heme-copper oxidases (e.g., cyctochrome c oxidase) has spurred the development of the inorganic chemistry of bridged heme-copper complexes, including species possessing (porphyrinate)Fe^III-O(H)-Cu^II

Full text of DOI:10.1021/ic981431+

Inorg. Chem. 1999, 38, 3093-3102
3093
Oxo- and Hydroxo-Bridged Heme-Copper Assemblies Formed from Acid-Base or
Metal-Dioxygen Chemistry
Marie-Aude Kopf,† Yorck-Michael Neuhold,‡ Andreas D. Zuberbu1hler,‡ and
Kenneth D. Karlin*,†
Department of Chemistry, The Johns Hopkins University, Baltimore, Maryland 21218, and Institut fu¨r
Anorganische Chemie, University of Basel, CH-4056 Basel, Switzerland
ReceiVed December 15, 1998
The iron-copper dinuclear active site in heme-copper oxidases (e.g., cyctochrome c oxidase) has spurred the
development of the inorganic chemistry of bridged heme-copper complexes, including species possessing
(porphyrinate)FeIII-O(H)-CuII-L moieties. We describe here the synthesis, by two routes, of [(F8TPP)FeIII-
O-CuII(MePY2)]+ (5) {F8TPP ) tetrakis(2,6-difluorophenyl)porphyrinate; MePY2 ) N,N-bis[2-(2-pyridyl)ethyl]-
methylamine}. First, 5-(CF3SO3) was generated by reaction of [(MePY2)CuII](CF3SO3)2 (3-(CF3SO3)2) and
[(F8TPP)FeIII-OH] (4) with triethylamine in THF or CH3CN in 65-70% yield. The complex was also prepared
by reduction of O2 by a 1:1 mixture of copper(I) and iron(II) complexes, [(MePY2)CuI(CH3CN)](BArF) (1-
(BArF)) (BArF ) tetrakis(3,5-bis-trifluoromethylphenyl)borate) and (F8TPP)FeII (2) in O2-saturated THF or acetone,
at -80 °C with subsequent warming to room temperature. Preliminary stopped-flow kinetics on the O2 reaction
with the 1:1 mixture show the formation of at least two intermediates (i.e., a superoxo species (F8TPP)Fe-O2
first, and then a presumed peroxo-bridged Fe-O2-Cu species) prior to the formation of the final µ-oxo complex
[(F8TPP)FeIII-O-CuII(MePY2)]+ (5-(BArF)). The 1H NMR spectrum of 5-(CF3SO3) in CD2Cl2 exhibits a pyrrole
2
peak at 67.7 ppm (corroborated by H NMR), while downfield (23.4 and 18.9 ppm) and dramatically upfield-
shifted resonances (-87.7, -155.4 and -189.4) have been assigned to hydrogens of the MePY2 moiety, by
specific deuteration. The µ-hydroxo complex [(F8TPP)Fe-(OH)-Cu(MePY2)](OTf)2 (6-(CF3SO3)2) was
synthesized by addition of 3-(CF3SO3)2 to 4 in CH3CN, or by protonation of 5-(CF3SO3) with CF3SO3H. In a 1H
NMR-spectroscopic protonation titration (CF3SO3H), the pyrrole 67.7 ppm resonance for 5-(CF3SO3) progressively
converts to 70.3 ppm, diagnostic of 6-(CF3SO3)2. The protonation is slow on the NMR time scale. The 1H NMR
spectral properties are consistent with antiferromagnetically coupled high-spin iron(III) and Cu(II) ions (S ) 2
spin state), and the interaction is weaker in 6-(CF3SO3)2 (5-(CF3SO3), µeff ) 5.05 µB; 6-(CF3SO3)2, µeff ) 5.60
µB; Evans NMR method). By titration using a series of organic acids, the pKa for 6-(CF3SO3)2 has been estimated
to be 16.7 < pKa < 17.6 (CH3CN solvent), or 9.6 ( 2 (aqueous). Plots of δ vs 1/T for both µ-oxo and µ-hydroxo
complexes 5-(CF3SO3) and 6-(CF3SO3)2 have been obtained, showing linear Curie (for downfield resonances) or
anti-Curie (for upfield peaks) behavior.
Introduction
(CN-),6-10 carboxylato (RCO2-),11 oxo (O2-), and hydroxo
(OH-)12-17 bridging ligands.
The characterization of (P)FeIII-X-CuII(ligand) species (X
) bridging ligand; P ) porphyrinate) has generated considerable
interest.1,2 Elucidating synthetic strategies and determining
physical and chemical properties of such bridged heme-copper
assemblies is in large part inspired by our current knowledge
of the active site structure and chemistry of heme-copper
oxidases such as cytochrome c oxidase.3 Here, the active site
where O2 binding and reduction occurs consists of a high-spin
heme group (with proximal histidine), which lies at ∼4.5 Å4,5
from CuB, which possesses three histidine nitrogen ligands.
Adjacent or bridged heme-copper centers have thus been
targeted for biomimetic modeling studies, with particular recent
attention concentrated on compounds with the X being cyano
In particular, µ-oxo and µ-hydroxo heme-copper complexes
have drawn our attention. While proposed18,19 quite early as a
(4) Yoshikawa, S.; Shinzawa-Itoh, K.; Nakashima, R.; Yaono, R.;
Yamashita, E.; Inoue, N.; Yao, M.; Jei-Fei, M.; Libeu, C. P.;
Mizushima, T.; Yamaguchi, H.; Tomizaki, T.; Tsukihara, T. Science
1998, 280, 1723-1729.
(5) Ostermeier, C.; Harrenga, A.; Ermler, U.; Michel, H. Proc. Natl. Acad.
Sci. U.S.A. 1997, 94, 10547-10553.
(6) Scott, M. J.; Holm, R. H. J. Am. Chem. Soc. 1994, 116, 11357-11367.
(7) Holm, R. H. Pure Appl. Chem. 1995, 67, 217-224.
(8) Kauffmann, K. E.; Goddard, C. A.; Zang, Y.; Holm, R. H.; Mu¨nck,
E. Inorg. Chem. 1997, 36, 985-993.
(9) Lim, B. S.; Holm, R. H. Inorg. Chem. 1998, 37, 4898-4908.
(10) Corsi, D. M.; Murthy, N. N.; Young, V. G., Jr.; Karlin, K. D. Inorg.
Chem. 1999, 38, 848-858.
* Author for correspondence.
(11) Scott, M. J.; Goddard, C. A.; Holm, R. H. Inorg. Chem. 1996, 35,
2558-2567.
† Johns Hopkins University.
‡ University of Basel.
(12) Karlin, K. D.; Nanthakumar, A.; Fox, S.; Murthy, N. N.; Ravi, N.;
Huynh, B. H.; Orosz, R. D.; Day, E. P. J. Am. Chem. Soc. 1994, 116,
4753-4763.
(1) Kopf, M.-A.; Karlin, K. D. In Biomimetic Oxidations; Meunier, B.,
Ed.; Imperial College Press: London, 1999; Chapter 7, in press.
(2) Kitajima, N. AdV. Inorg. Chem. 1992, 39, 1-77.
(3) (a) Ferguson-Miller, S.; Babcock, G. T. Chem. ReV. 1996, 96, 2889-
2907. (b) Michel, H.; Behr, J.; Harrenga, A.; Kannt, A. Annu. ReV.
Biophys. Biomol. Struct. 1998, 27, 329-356.
(13) Fox, S.; Nanthakumar, A.; Wikstro¨m, M.; Karlin, K. D.; Blackburn,
N. J. J. Am. Chem. Soc. 1996, 118, 24-34.
(14) Nanthakumar, A.; Fox, S.; Murthy, N. N.; Karlin, K. D. J. Am. Chem.
Soc. 1997, 119, 3898-3906.
10.1021/ic981431+ CCC: $18.00 © 1999 American Chemical Society
Published on Web 06/05/1999

3094 Inorganic Chemistry, Vol. 38, No. 13, 1999
Kopf et al.
possible bridging ligand in the extensively studied oxidized
resting state enzyme form, FeIII-(µ-O(H))-CuII complexes have
now been generated. Their existence is perhaps somewhat
surprising in light of the known thermodynamic stability of
µ-oxo dinuclear complexes P-FeIII-O-FeIII-P. Yet, P-FeIII-
O-CuII(ligand) compounds can be generated by acid-base
assembly procedures13,15-17 and/or by dioxygen reactivity with
reduced P-FeII/CuI(ligand) precursors.1,12,15b,20 Compounds
studied in our own laboratories have focused on oxo and hy-
droxo complexes with the tripodal tetradentate TMPA (tris(2-
pyridylmethyl)amine) copper ligand, in either self-assembled12-14
or tethered heterobinucleating15 ligand (L) systems (see dia-
gram). Oxo compounds with linear or bent configurations have
These copper(I) complexes with tridentate ligands react with
O2 to give completely different dioxygen adducts, compared to
TMPA or its analogues. The tridentate ligands give binuclear
µ-η2:η2-peroxo-dicopper(II) and/or bis-µ-oxo-dicopper(III)
products,23 while TMPA and its analogues form end-on bound
CuII-O-O-CuII species.25 Furthermore, the CuB at the Fe-
Cu active site of heme-copper oxidases has N3 (from histidine
imidazoles) tridentate coordination.4,5 Thus, MePY2-copper is
worthy of study in association with heme chemistry,15c to
compare and contrast with TMPA or other systems, with respect
to FeII-CuI/O2 reactivity, and formation and properties of
possible FeIII-(µ-O(H))-CuII products. Furthermore, the tri-
dentate nitrogen ligand in MePY2 matches the number and kind
of donor atoms observed at the heme-copper oxidase active site,
as mentioned above.
Here, we describe new µ-oxo and µ-hydroxo complexes with
the MePY2 ligand on copper, and F8TPP (tetrakis(2,6-difluo-
rophenyl)porphyrinate) as the iron ligand. As for the TMPA-
containing system, we have been successful in generating the
µ-oxo product [(F8TPP)FeIII-O-CuII(MePY2)]+ (5), either by
O2 reactivity with [(MePY2)CuI(MeCN)]+ (1) and (F8TPP)FeII
(2) or by acid-base self-assembly procedures. The synthetic
details are presented, and some insights into the oxygenation
reaction are provided by low-temperature UV-vis studies,
including a preliminary stopped-flow kinetic investigation. The
conjugate acid, µ-hydroxo complex [(F8TPP)FeIII-OH-CuII-
(MePY2)]2+ (6), can be obtained by protonation of 5, and an
estimate of the pKa of 6 has been determined. In addition to the
acid-base interconversion chemistry, the 1H NMR and magnetic
data of 5 and 6 are described.
been characterized. These compounds have novel and interesting
magnetic and 1H NMR properties, and the µ-oxo ligand is very
basic, but its protonation behavior depends on the detailed nature
of the ligand environment.15 Such acid-base relationships and
proton-transfer chemistry interconverting µ-oxo and µ-hydroxo
FeIII-(µ-O(H))-CuII complexes is relevant to enzyme mech-
anism since heme-copper oxidases function as transmembrane
proton pumps. It has been suggested that FeIII-(µ-O(H))-CuII
species could represent the resting state enzyme oxidized forms,
or possibly could even occur as enzyme turnover intermediates.13
In this report, we focus on chemistry with a different copper
ligand, N,N-bis[2-(2-pyridyl)ethyl]methylamine (MePY2). One
rationale for previously using TMPA as the copper ligand in
heme-copper oxidase model systems was that we have exten-
sively characterized its copper(I)-dioxygen chemistry, preparing
us for future studies in PFeII‚‚‚CuI/O2 reactivity. We have also
described in some detail the dioxygen reactivity of [(MePY2)-
CuI(MeCN)]+ (1), analogues with N-substituents other then
methyl,21 and binucleating analogues.22-24
Results and Discussion
Synthesis of [(F8TPP)FeIII-O-CuII(MePY2)](CF3SO3) (5-
(CF3SO3)). Initially, the µ-oxo complex was synthesized by an
acid-base self-assembly method, as a triflate salt. Thus, when
[(MePY2)CuII](CF3SO3)2 (3-(CF3SO3)2) is reacted with an
equimolar quantity of the iron-hydroxo complex [(F8TPP)FeIII-
OH] (4) in the presence of Et3N, a microcrystalline black
precipitate 5-(CF3SO3) can be isolated in high yield (Scheme
1). This preparation is directly analogous to that carried out for
the TMPA analogue [(F8TPP)FeIII-O-CuII(TMPA)]+.12,13 Com-
pound 5-(CF3SO3) is very moisture sensitive, but otherwise
stable, and is soluble in a variety of solvents including
tetrahydrofuran (THF), acetonitrile, dichloromethane, and ac-
etone. The triflate anion is probably uncoordinated, which is
consistent with the observed solution conductivity of the
(15) (a) Obias, H. V.; van Strijdonck, G. P. F.; Lee, D.-H.; Ralle, M.;
Blackburn, N. J.; Karlin, K. D. J. Am. Chem. Soc. 1998, 120, 9696-
9697. (b) Ju, T. D.; Ghiladi, R. A.; Lee, D.-H.; van Strijdonck, G. P.
F.; Woods, A. S.; Cotter, R. J.; Young, J. V. G.; Karlin, K. D. Inorg.
Chem. 1999, 38, 2244-2245. (c) Kopf, M.-A.; Karlin, K. D., submitted
for publication.
(16) Lee, S. C.; Holm, R. H. J. Am. Chem. Soc. 1993, 115, 5833-5834.
(17) Scott, M. J.; Zhang, H. H.; Lee, S. C.; Hedman, B.; Hodgson, K. O.;
Holm, R. H. J. Am. Chem. Soc. 1995, 117, 568-569.
(18) Blumberg, W. E.; Peisach, J. In Cytochrome c oxidase; King, T. e. a.,
Ed.; Elsevier/North-Holland Biomedical Press: Amsterdam, The
Netherlands, 1979; pp 153-159.
complex (molar conductivity, Λm (CH3CN) ) 154 Ω cm2 mol-1
,
(19) Reed, C. A.; Landrum, J. T. FEBS Lett. 1979, 106, 265-267.
(20) Karlin, K. D.; Lee, D.-H.; Obias, H. V.; Humphreys, K. J. Pure Appl.
Chem. 1998, 70, 855-862.
1:1 electrolyte,26 Experimental Section) and the fact that
5-(CF3SO3) has properties essentially identical to those of the
BArF analogue (BArF ) tetrakis(3,5-bis-trifluoromethylphenyl)-
borate) [(F8TPP)FeIII-O-CuII(MePY2)](BArF) (5-(BArF)) (vide
infra).
(21) Sanyal, I.; Mahroof-Tahir, M.; Nasir, S.; Ghosh, P.; Cohen, B. I.;
Gultneh, Y.; Cruse, R.; Farooq, A.; Karlin, K. D.; Liu, S.; Zubieta, J.
Inorg. Chem. 1992, 31, 4322-4332.
(22) Karlin, K. D.; Tyekla´r, Z.; Farooq, A.; Haka, M. S.; Ghosh, P.; Cruse,
R. W.; Gultneh, Y.; Hayes, J. C.; Toscano, P. J.; Zubieta, J. Inorg.
Chem. 1992, 31, 1436-1451.
In addition to elemental analysis (Experimental Section),
5-(CF3SO3) possesses characteristics that clearly identify it as
(23) Obias, H. V.; Lin, Y.; Murthy, N. N.; Pidcock, E.; Solomon, E. I.;
Ralle, M.; Blackburn, N. J.; Neuhold, Y.-M.; Zuberbu¨hler, A. D.;
Karlin, K. D. J. Am. Chem. Soc. 1998, 120, 12960-12961.
(24) Pidcock, E.; Obias, H. V.; Abe, M.; Liang, H.-C.; Karlin, K. D.;
Solomon, E. I. J. Am. Chem. Soc. 1999, 121, 1299-1308.
(25) Tyekla´r, Z.; Jacobson, R. R.; Wei, N.; Murthy, N. N.; Zubieta, J.;
Karlin, K. D. J. Am. Chem. Soc. 1993, 115, 2677-2689.
(26) Geary, W. J. Coord. Chem. ReV. 1971, 7, 81-122.

Oxo- and Hydroxo-Bridged Heme-Copper Assemblies
Inorganic Chemistry, Vol. 38, No. 13, 1999 3095
Figure 1. 1H NMR spectrum (297 K) and assignments for 5-(CF3SO3) in CD2Cl2 solvent at 300 MHz. * ) solvent peak.
Scheme 1
observed for 5-(CF3SO3), which is not seen for the starting
copper or iron complexes, 3-(CF3SO3)2 and [(F8TPP)FeIII-OH]
(4). Tentatively, that IR absorption is assigned to the Fe-O-
Cu asymmetric stretch, as previously discussed for [(F8TPP)-
FeIII-O-CuII(TMPA)]+ and TMPA-tethered analogues [(L)-
FeIII-O-CuII]+ 15 (see Introduction).
NMR Spectroscopy of 5-(CF3SO3). The room-temperature
1H NMR spectrum of 5-(CF3SO3) is shown in Figure 1. The
most downfield shifted peak at 67.7 ppm is assigned to the
pyrrole resonances of the F8TPP portion of the molecule. This
assignment was confirmed by synthesis of [(F8TPP-d8)FeIII-
O-CuII(MePY2)]+ (5-d8) (Experimental Section) and examina-
tion using 2H NMR spectroscopy (Table 1). This pyrrole
chemical shift is consistent with high-spin Fe(III), somewhat
upfield from the typical value of ∼80 ppm seen for (TPP)FeIII-
X.28 The pyrrole resonance in [(F8TPP)FeIII-O-CuII(TMPA)]+
is in a similar position (Table 1), and as explained for that
system,14 this shift is a result of the antiferromagnetic coupling
5
1
between the high-spin Fe(III) (S ) /2) and Cu(II) (S ) /2) to
give an S ) 2 spin system. Room-temperature magnetic moment
measurements on 5-(CF3SO3), both in solution (µeff ) 5.05 µB;
Evans method) and in the solid state (µeff ) 5.15 µB), are
consistent with a similar S ) 2 assignment for 5-(CF3SO3);
these values are considerably reduced compared to what would
be expected for an uncoupled system.13 The F8TPP meta (m)
and para (p) phenyl hydrogens appear at their usual place, with
two peaks observed for the meta protons (9.90 and 9.24 ppm),
while the para hydrogens resonate at 7.95 ppm. The split in the
m-phenyl protons is indicative of different magnetic environ-
ments (on the NMR scale, at room temperature), with one set
of hydrogens being “up”, on the same side of the µ-oxo and
copper moieties, while the others are “down”. This asymmetry
is also observed with the o-fluorine atoms with a 19F NMR
spectrum of 5-(CF3SO3) showing a split signal (δ ) -95.87
and -96.70 ppm).
containing the µ-oxo FeIII-O-CuII moiety. First, 5-(CF3SO3)
has a distinctive red-shifted Soret band (446 nm), which is quite
different from high-spin µ-oxo porphyrin-iron(III) dimers such
as [(F8TPP)Fe]2O (λmax 400 nm) and (F8TPP)Fe-OH (λmax 408
nm).12 It has been shown that, in the case of Zn(TPP)L
complexes, an increase in negative charge correlates with a
decrease in Soret transition energies.27 Thus, in µ-oxo FeIII-
O-CuII species, the Cu(II) ion is a poorer π-acceptor and has
less positive charge than the Fe(III) in corresponding FeIII-
O-FeIII complexes. Second, the 1H NMR properties of
5-(CF3SO3), discussed in detail below, reveal that it has upfield
paramagnetically shifted copper-ligand resonances and down-
field-shifted pyrrole absorptions, as seen in [(F8TPP)FeIII-O-
CuII(TMPA)]+.12,14 Third, a new infrared peak at 837 cm-1 is
By synthesis (Experimental Section) of MePY2-d7 and
1
2
MePY2-d11 (see diagram) and by a combination of H and H
NMR spectroscopy, the signals at 23.4, 18.9, -87.7, -155.4,
and -189.4 ppm have also been unambiguously assigned, Figure
1 and Table 1. The proton signals of a given methylene group
are diastereotopic, thus inequivalent, and all individual hydrogen
resonances appear. The remaining four signals (-37.2, -23.5,
(28) Goff, H. M. In Iron Porphyrins; Lever, A. B. P., Gray, H. B., Eds.;
(27) Nappa, M.; Valentine, J. S. J. Am. Chem. Soc. 1978, 100, 5075.
Addison-Wesley: Reading, MA, 1983; Part 1, Chapter 4.

3096 Inorganic Chemistry, Vol. 38, No. 13, 1999
Kopf et al.
1
2
Table 1. Comparison of Chemical Shifts (ppm) in H and H NMR Spectra of 5, 6, and Relevant Compounds
1H NMR Spectra
peak position (ppm)
5
67.7
81.9
81.1
65
9.9
7.95
8.09
7.73
7.6
23.4
18.9
29.8
23.50
32.8
24.8
-87.7
-155.4
-88.2
-209.1
-189.4
-232.5
9.24
10.7
9.75
12.3
11.47
9.6
5 (243 K)
6 (243 K)
-120.6, -168.4, -184
[(F8TPP)FeIII-O-CuII)(TMPA)]+ a
-104
N/A
CH3
9.2
assignment
pyrrole
m-phenyl
p-phenyl
CH2-PY
CH2-N
2H NMR Spectra
peak position (ppm)
[(d8-F8TPP)FeIII-O-CuII(MeN)(CH2CH2PY)2]+
assignment
66.36
pyrrole
[(F8TPP)FeIII-O-CuII(CD3N)(CH2CD2PY)2]+
23.75
19.61
-182.76
assignment
CD2-PY
CD3
[(F8TPP)FeIII-O-CuII)(CD3N)(CD2CD2PY)2]+
24.26
20.20
-75.78
-149
-182.76
assignment
CD2-PY
CD2-N
CD3
a Reference 13.
(III) chlorin π-cation radical complex with overall S ) 2 has
5
also been reported (where the high-spin S )
/ Fe(III) is
2
1
antiferromagnetically coupled to a porphyrin S ) /2 radical
cation).33 The behavior of the present S ) 2 FeIII-O-CuII
complexes with large chemical shift range is unprecedented.
Formation of 5 from Fe/Cu/O2 Reaction. The µ-oxo
complex 5-(BArF) has also been generated directly by reaction
of O2 with the copper(I) and iron(II) complexes. The equimolar
mixture of [(MePY2)CuI(CH3CN)](BArF) (1-(BArF)) and iron-
(II) species 2 in THF or acetone, when saturated with dry O2 at
-80 °C, gives, upon warming to room temperature, a red
solution of the µ-oxo product (Scheme 1). Its 1H NMR spectrum
is nearly identical to that observed for the acid-base syntheti-
cally generated complex, 5-(CF3SO3) (see Experimental Sec-
tion).
-2.7, and 1.2 ppm) must be due to pyridyl ring hydrogen
resonances of the MePY2 moiety. Note that the -CH2PY
methylene hydrogens are downfield shifted, opposite to the
direction of the pyridyl ring hydrogens; it is not uncommon for
a methylene pair bound to a pyridine ring to have a chemical
shift similar in magnitude but opposite in sign to the aromatic
proton in the same position.29,30 A final point is that the pyridyl
hydrogen resonance at 1.2 ppm, the least shifted from an
expected diamagnetic position, is most likely assigned to the
pyridyl 4-hydrogens of the MePY2 copper ligand, since it is
the furthest in distance and the highest number of bonds away
from the paramagnetic copper ion.
Since monomeric Cu(II) complexes usually do not exhibit
well-resolved proton NMR signals, the presence of relatively
sharp, upfield-shifted peaks indicates an antiferromagnetic
interaction between the iron(III) and copper(II) ions, as men-
tioned above. The same phenomenon is observed in [(F8TPP)-
FeIII-O-CuII(TMPA)]+, [(F8TPP)FeIII-OH-CuII(TMPA)]2+
and tethered complexes [(5L)FeIII-O-CuII]+,15 and we have
previously presented a detailed discussion of the theoretical
framework from Bertini and Luchinat and co-workers,31 used
to interpret these NMR spectra. The upfield and downfield peak
signature is seen in cobalt(II)-substituted Cu-Zn superoxide
dismutase31,32 and certain forms of iron ferrodoxins.31 An iron-
Some insight into the dioxygen reactivity involved has been
obtained from low-temperature UV-vis experiments (Scheme
2 and Figure 2). When 2 (λmax ) 422 nm) is reacted with O2 at
-70 °C in THF, there is an immediate change in the UV-vis
spectrum, giving a low-temperature stable species with λmax
)
416 (Soret) and 536 nm, Figure 2. This has independently been
characterized as an Fe/O2 1:1 adduct, hereafter designated as
(F8TPP)Fe-O2, formally a superoxo-iron(III) complex, with
THF acting as an axial ligand.34 After application of a vacuum
to remove excess O2 {(F8TPP)Fe-O2 is stable under these
conditions}, 1 equiv of 1-(BArF) was added. The first UV-
vis spectrum recorded after ca. 30 s following addition of
1-BArF and mixing showed that the 416 Soret band shifted to
422 nm (but with much lower intensity than that of pure 2)
while a broad R absorption is observed. (These time-dependent
spectral changes are shown in a figure given in the Supporting
(29) Murthy, N. N.; Karlin, K. D.; Bertini, I.; Luchinat, C. J. Am. Chem.
Soc. 1997, 119, 2156-2162.
(30) McConnell, H. M. J. Chem. Phys. 1956, 24, 632.
(31) Luchinat, C.; Ciurli, S. In Biological Magnetic Resonance; Berliner,
J., Reuben, J., Eds.; Plenum: New York, 1993; Vol. 12, pp 357-
420.
(33) Ozowa, S.; Watanabe, Y.; Morishima, I. J. Am. Chem. Soc. 1994,
116, 5832-5838.
(34) Ghiladi, R. A.; Dahan, M.; Neuhold, Y.-M.; Zuberbu¨hler, A. D.; Karlin,
K. D. Manuscript in preparation. Characterization of (F8TPP)Fe-O2
comes from UV-vis and NMR spectroscopic properties (diamagnetic,
i.e., with THF axial base ligand), in addition to O2-uptake (Fe:O2 )
1:1) measurements.
(32) Bertini, I.; Luchinat, C.; Piccioli, M. Progress in NMR Spectroscopy;
London, 1994.

Oxo- and Hydroxo-Bridged Heme-Copper Assemblies
Inorganic Chemistry, Vol. 38, No. 13, 1999 3097
Figure 3. Experimental spectra (stopped-flow kinetic, 182 s total, -90
°C, acetone) for the reaction of O2 with a 1:1 mixture of 1-(BArF)
and 2. See text for further explanation.
Stopped-flow kinetics support this conclusion and provide
additional insight. UV-vis spectral changes (R region only) are
shown in Figure 3. At -90 °C in acetone, addition of O2 to a
1:1 mixture of 1-(BArF) and 2 results in the formation of a
new species with λmax ) 535 nm, within the mixing time of the
experiment (1 ms). This species corresponds to (F8TPP)Fe-
O2, as described above (Scheme 2); this very fast formation is
also observed when (F8TPP)FeII (2) independently is reacted
with O2 in a similarly weakly coordinating solvent, THF.34 By
contrast, the kinetics of oxygenation of 1-(BArF) by itself in
acetone show that its reaction with O2 does not occur within
the stopped-flow mixing time; the formation of the binuclear
adduct, [{(MePY2)Cu}2(O2)]2+, is relatively slow, with ∆Hq)
-0.7 ( 1 kJ M-1, ∆Sq) -164 ( 4 J K-1 M-1, and no Cu/O2
) 1:1 adduct is observed.23 Thus, in mixtures of reactive iron-
(II) and copper(I) complexes, 1-(BArF) and 2, iron interacts
with O2 first.
Figure 2. Spectra (Soret and R regions) for O2 chemistry, leading to
formation of the µ-oxo product 5-(BArF); 446 (Soret) and 558 nm.
Addition of O2 to 2 (422, 542 nm) gives (F8TPP)Fe-O2 (416, 536
nm). After removal of excess O2 and addition of 1 equiv of 1-(BArF),
5-(BArF) is obtained. Also, see Scheme 2 and Supporting Information.
Scheme 2
Further, the low-temperature stopped-flow kinetic data reveal
that, within 1 s, (F8TPP)Fe-O2 is transformed into a new species
with a broad R-band absorption and with λmax ∼560 nm, perhaps
similar to that obtained in the benchtop experiment for addition
of O2 to a 1:1 mixture of 1-(BArF) and 2 in THF or acetone
(vide supra). This further fast reaction shows two isosbestic
points (518 and 548 nm; Figure 3); the corresponding second-
order rate constant was determined to be 3.4 ( 0.2 × 104 M-1
s-1. An absorbance vs time plot is given in the Supporting
Information. While further studies are clearly necessary, this
reaction must represent an interaction of (F8TPP)Fe-O2 with
1-(BArF), perhaps producing a peroxo-bridged species.
The fast-forming product with broad λmax ∼560 nm absorp-
tion subsequently decays in a complicated manner (at least two
relaxations) to the final product with λmax ) 552 nm, Figure 3.
This product corresponds to our µ-oxo product complex,
5-(BArF). In the stopped-flow experiments, these latter reactions
occur rather rapidly; 5-(BArF) is formed to a significant extent
within 3-5 min at -90 °C, which contrasts with our observa-
tions from the benchtop monitoring experiments described
above, i.e., formation in hours. The explanation for these
different reaction time scales appears to be the involvement of
photochemistry, which is probably occurring in the stopped-
flow experiments with a diode-array (unfiltered) light source.
We have corroborated this assumption by carrying out our
benchtop oxygenation experiments in the presence of additional
light sources. In fact, when carrying out the reaction of a
1-(BArF)/2/O2 mixture (∼-75 °C) and recording UV-vis
spectra having an open shutter in a diode-array instrument, or
by exposing the same mixture in a dewar cuvette assembly to
Information). Then, over a period of several hours (still at -70
°C), the complex 5-(BArF), with characteristic red-shifted 446
nm Soret and 558 nm R band, forms cleanly.
In acetone, when a 1:1 mixture of 1-(BArF) and 2 was
bubbled with O2 at -75 °C, a species with very similar spectral
characteristics (422 nm Soret band of analogous intensity, broad
R band) is observed. These observations suggest that in the FeII/
CuI mixture of 1-(BArF) and 2, O2 binding to iron(II) occurs
first (further evidence is given below), but the presence of Cu
subsequently affects the chemistry.

3098 Inorganic Chemistry, Vol. 38, No. 13, 1999
Kopf et al.
a lightbulb, the rate of formation of 5-(BArF) was significantly
increased. We have previously noted strong photochemistry
occurring with copper-dioxygen complexes employing ligands
similar to MePY2,35,36 while photolytic activity has been
observed for cytochrome c oxidase O2 intermediates37 as well
as O2 adducts of other metal complexes.38,39
Further detailed stopped-flow kinetic studies or further
investigations of the photochemistry which may be occurring
here were not carried out. While the O2 chemistry appears to
be very rich, detailed analysis would be thwarted by the ever-
present problem of not having exactly 1:1 ratios of iron(II) and
copper(I) starting materials. Future inquiries may be directed
at Fe‚‚‚Cu complexes possessing binucleating ligands, such as
L (see Introduction),15 or analogues with tridentate tethers.15c
Protonation Titration. Formation of [(F8TPP)FeIII-(OH)-
CuII(MePY2)](CF3SO3)2 (6-(CF3SO3)2). As shown in Scheme
1, the corresponding dicationic µ-hydroxo complex, 6-(CF3SO3)2,
can also be prepared by two routes (Experimental Section). First,
Et3N is left out when reacting 3 and 4. The second method is
via the direct protonation of 5-(CF3SO3). Figure 3 shows the
transformation of µ-oxo complex to 6-(CF3SO3)2 via addition
of increasing quantities of triflic acid, CF3SO3H. The 67.7 ppm
pyrrole absorption (vide supra), converts to a new peak at 70.3
ppm. This further downfield shift suggests a weaker coupling
in 6-(CF3SO3)2 than in 5-(CF3SO3). This is also confirmed by
comparison of the room-temperature magnetic moment of
6-(CF3SO3)2 (µeff ) 5.60 µB; Evans NMR method). We can
rule out that (F8TPP)FeIII-triflate is a product of the protonation,
since it is known to have a pyrrole resonance at ∼53 ppm.13
In the diamagnetic region, the “doublet” at 9.9 and 9.24 ppm,
assigned to the m-phenyl porphyrinate resonances, also shifts
to a new “doublet” at 11.54 and 10.73 ppm. By contrast, no
significant shift of the 7.95 ppm p-phenyl absorption occurs
upon conversion of 5-(CF3SO3) to 6-(CF3SO3)2. This behavior
observed for the meso-aryl groups was also seen in the
protonation of [(F8TPP)Fe-O-Cu(TMPA)]+, which gave the
1
Figure 4. Addition of triflic acid to 5-(CF3SO3) as followed by H
NMR spectroscopy (300 MHz, 297 K, CD2Cl2): (a) spectrum of
5-(CF3SO3), (b) after adding 0.33 equiv of CF3SO3H, (c) after adding
0.66 equiv of CF3SO3H, and (d) after adding 1 equiv of CF3SO3H,
producing 6-(CF3SO3)2.
plexes.41 Protonation usually results in bending of the previously
near-linear M-O-M′ unit; the required rehybridization around
the µ-oxo atom to give a M-(OH)-M′ product and concomitant
structural rearrangements are generally seen to be the causes
for the relatively slow proton-transfer reactions.40,41
The structure of 6-(CF3SO3)2 has not been determined. On
the basis of the apparent decrease in magnetic coupling (vide
supra) of this complex compared to the parent µ-oxo complex
5-(CF3SO3), we propose that 6-(CF3SO3)2 has a bent Fe-
(OH)-Cu structure, as observed in [(F8TPP)Fe-OH-Cu-
(TMPA)]2+ ( Fe-OH-Cu ∼ 157°).13 We observe 6-(CF3SO3)2
to be rather unstable and particularly moisture sensitive; for
example, repeated attempts to obtain X-ray quality crystals
invariably led to decomposition reactions yielding isolated
species such as the µ-oxo diiron(III) complex [(F8TPP)FeIII-
O-FeIII(F8TPP)]+ and/or the bis-aquo complex [CuII(MePY2)-
(OH2)2](CF3SO3)2.
µ-OH- complex [(F8TPP)Fe-OH-Cu(TMPA)]2+ 13
. The other
peaks, which are associated with the protons of the MePY2
moiety, disappear during the protonation experiment (with 1
equiv of CF3SO3H) carried out at room temperature. We suggest
that the relatively weaker M-(OH) bonds could result in partial
dissociation of the iron and copper complex moieties within
6-(CF3SO3)2, in equilibrium with (F8TPP)Fe-OH and 3-(CF3-
SO3)2. This supposition is consistent with the observation (vide
infra) that the characteristic upfield-shifted peaks for the coupled
S ) 2 µ-hydroxo complex 6-(CF3SO3)2 do appear when the
temperature is lowered.
The protonation process is slow on the NMR scale, since
two distinctive sets of resonances in mixtures of 5-(CF3SO3)
and 6-(CF3SO3)2 are observed (Figure 4), rather than a single
averaged peak which would be present if the exchange process
were fast. A similar slow protonation was observed in the
conversion of [(F8TPP)Fe-O-Cu(TMPA)]+ to [(F8TPP)Fe-
Estimation of the pKa of the Complex 6-(CF3SO3)2. The
protonation of complexes like 5-(CF3SO3) is of general interest
in the study of µ-oxo metal compounds13,41,42 and is of possible
importance in heme-copper oxidase function, since protons
either are taken up to produce water from O2 (scalar protons)
or are translocated (vectorial protons).3,13 In order to assess the
pKa of 6-(CF3SO3)2, we titrated 5-(CF3SO3) with a variety of
acids having known pKa values in acetonitrile solvent (Table
2)43 and monitored the formation of conjugate acid 6-(CF3SO3)2
using 1H NMR spectroscopy. A range of acids, from anilinium
triflate (pKa ) 10.56) to trimethylammonium chloride (pKa )
17.6 in CH3CN), were used.
OH-Cu(TMPA)]2+ 13
. Slow protonation reactions seem to be a
common feature for oxo-bridged metal ion species, e.g., bis-
(µ-oxo)-dimanganese(III)40 and (µ-oxo)-diiron(III) com-
(35) Karlin, K. D.; Nasir, M. S.; Cohen, B. I.; Cruse, R. W.; Kaderli, S.;
Zuberbu¨hler, A. D. J. Am. Chem. Soc. 1994, 116, 1324-1336.
(36) Karlin, K. D.; Wei, N.; Jung, B.; Kaderli, S.; Niklaus, P.; Zuberbu¨hler,
A. D. J. Am. Chem. Soc. 1993, 115, 9506-9514.
(37) Varotsis, C. A.; Babcock, G. T. J. Am. Chem. Soc. 1995, 117, 11260-
11269.
(40) Carroll, J. M.; Norton, J. R. J. Am. Chem. Soc. 1992, 114, 8744-
8745.
(38) MacArthur, R.; Sucheta, A.; Chong, F. F. S.; Einarsdo´ttir, O. Proc.
Natl. Acad. Sci. U.S.A. 1995, 92, 8105-8109.
(39) Maeder, M.; Ma¨cke, H. R. Inorg. Chem. 1994, 33, 3135-3140.
(41) Kramarz, K. W.; Norton, J. R. Prog. Inorg. Chem. 1994, 42, 1-65.
(42) Evans, D. R.; Mathur, R. S.; Heerwegh, K.; Reed, C. A.; Xie, Z.
Angew. Chem., Int. Ed. Engl. 1997, 36, 1335-1337.

Oxo- and Hydroxo-Bridged Heme-Copper Assemblies
Inorganic Chemistry, Vol. 38, No. 13, 1999 3099
Table 2. Protonation of 5 by Different Acids
Addition of 1 equiv of trimethylammonium triflate does not
cause protonation of 5-(CF3SO3). However, a complete and
clean conversion of 5-(CF3SO3) to 6-(CF3SO3)2 occurs when
N-methylmorpholinium triflate and 2,4-lutidinium triflate are
used. The protonation of the µ-oxo complex has also been
observed with benzylammonium triflate and the stronger protic
acid anilinium triflate, Table 2. In these latter cases, a second
set of peaks, typical of a low-spin Fe(III) porphyrin, was
observed at 23 and -20 ppm. This behavior indicates that these
sterically nonhindered ammonium salts first protonated the
µ-oxo bridge and the hydroxo complex formed reacted im-
mediately with the liberated amine (conjugate base) to give the
low-spin species product. We can conclude that the pKa of
6-(CF3SO3)2 is 16.7 < pKa < 17.6. It has been suggested by
Pecoraro et al.43a that the corresponding aqueous pKa would be
7.5 ( 1 pKa units lower than in CH3CN, meaning we can
estimate a pKa for 6-(CF3SO3)2 in water of 9.6 ( 2. Clearly,
the oxo moiety in 5-(CF3SO3) is quite basic and the pKa is very
comparable to that observed in our study of µ-oxo complex [(F8-
TPP)Fe-O-Cu(TMPA)]+ (pKa (H2O), 8 ( 2.5). In contrast,
other known oxo-bridged Fe, Mn, and Ru complexes have pKa
values below this range.43a,44,45 For instance, the pKa value for
[(HBpz3)2Fe2(OH)(O2CCH3)2]+ [HBpz3 ) hydrotris(1-pyra-
zolyl)borate] and [(L′)Ru(OH)(O2CCH3)2Fe(L′′)]3+ (L′ ) 1,4,7-
triazacyclononane, L′′ ) 1,4,7-trimethyl-1,4,7-triazacyclononane)
are estimated to be around 3.5.44 Oxo-bridged manganese
Figure 5. Variable-temperature 1H NMR spectra (300 MHz, CD2Cl2)
of 6-(CF3SO3)2. Curie plots are given in the Supporting Information.
antiferromagnetic coupling as T approaches infinity; theoreti-
cally, the system uncouples to behave as a Cu(II) mononuclear
species.
Figure 5 shows the effect of lowering the temperature on 1H
NMR spectra of the µ-hydroxo-bridged complex 6-(CF3SO3)2.
Resonances, which might be assignable to the MePY2 ligand,
are not seen at room temperature. However, upon cooling, these
peaks do appear, and they exhibit a profile rather similar to
that observed for the µ-oxo analogue 5-(CF3SO3). Along with
the observed magnetic moment of 6-(CF3SO3)2 (vide supra),
the behavior is what is expected for the postulated S ) 2
antiferromagnetically coupled system. The behavior (i.e., the
lack of signals at room temperature) must be ascribed to the
presence of the µ-OH- rather than µ-oxo ligand. The same
phenomenon was also observed for [(F8TPP)FeIII-OH-CuII-
(TMPA)]2+, in comparison to [(F8TPP)FeIII-O-CuII(TMPA)]+.
As proposed and discussed for this TMPA analogue set of
compounds,14 the weaker M-(OH) bonds allow for relatively
rapid ligand exchange phenomena near room temperature,
leading to line broadening.
complexes, [MnIV(µ-O)n(µ-OH)]2 (n ) 143a or 245), have
+
aqueous pKa values around -2 < pKa (H2O) < 7. The lower
pKa value in these FeIII-O-CuII complexes may be due to the
presence of copper(II), thus a metal with a lower oxidation state
(and therefore decreased Lewis acidity) compared to the metals
(Fe, Ru, Mn) in the complexes mentioned above.
Variable-Temperature (VT) NMR Studies. As expected
for a paramagnetic system, the proton resonances in 5-(CF3SO3)
exhibit a strong temperature dependence. Variable-temperature
1H NMR data have been obtained, and these, along with plots
of chemical shift versus 1/T (Curie plots) for the temperature
range 293-213 K, are given in the Supporting Information. No
dramatic changes in the diamagnetic region are observed when
the temperature is lowered. However, the upfield-shifted signals
move further upfield while the downfield signals move further
downfield, all with broadening. The δ vs 1/T plots show linear
Curie and anti-Curie behavior, consistent with a strong antifer-
romagnetic coupling dominated by a single spin state in the
temperature range studied.14 The MePY2 proton upfield-shifted
signals, all except one, have associated positive (i.e., downfield)
extrapolated intercepts, consistent with the breaking of the strong
Summary and Conclusion
The following are the principal results and conclusions of
this investigation:
The oxo-bridged heme-copper assembly [(F8TPP)Fe-O-
Cu(MePY2)]+ (5) is readily prepared in good yield by acid-
base self-assembly, using [(MePY2)CuII](CF3SO3)2 (3-(CF3SO3)2)
and the hydroxo species [(F8TPP)FeIII-OH] (4) in the presence
of a suitable base. This µ-oxo FeIII-O-CuII complex has been
characterized by techniques including UV-vis, multinuclear
NMR, and IR spectroscopies, magnetism, and solution conduc-
1
tivity. Its H NMR spectrum has been assigned by specific
deuteration of the porphyrinate moiety as well as the MePY2
ligand and reveals the presence of an antiferromagnetic coupling
(43) (a) Baldwin, M. J.; Gelasco, A.; Pecoraro, V. L. Photosynth. Res. 1993,
38, 303-308 and references cited therein. (b) Izutsu, K. Acid-Base
Dissociation Constants in Dipolar Aprotic SolVents; Blackwell
Scientific Publications: Boston, 1990.
(44) Turowski, P. N.; Armstrong, W. H.; Liu, S.; Brown, S. N.; Lippard,
S. J. Inorg. Chem. 1994, 33, 636-645 and references cited therein.
(45) Hage, R.; Krijnen, B.; Warnaar, J. B.; Hartl, F.; Stufkens, D. J.; Snoeck,
T. L. Inorg. Chem. 1995, 34, 4973-4978.
5
1
between the high-spin Fe(III) (S ) /2) and Cu(II) (S ) /2)
ions, to give an S ) 2 spin system. Room-temperature magnetic
moment measurements corroborate the S ) 2 assignment for
5-(CF3SO3).
We have shown that 5-(BArf) could also be prepared by O2
reaction with [(MePY2)CuI(CH3CN)](BArF) (1-(BArF)) and

3100 Inorganic Chemistry, Vol. 38, No. 13, 1999
Kopf et al.
(F8TPP)FeII (2). Low-temperature stopped-flow kinetics reveals
the presence of rapidly forming intermediates, including a 1:1
adduct (F8TPP)Fe-O2, which forms first, even in the presence
of 1; the species observed subsequently requires the presence
of Cu and is tentatively assigned as a peroxo-bridged species.
The µ-oxo final product 5-(BArF) forms in a slower process.
We have also established the acid-base interconversion of
[(F8TPP)Fe-O-Cu(MePY2)](CF3SO3) (5-(CF3SO3)) to [(F8-
TPP)Fe-(OH)-Cu(MePY2)](CF3SO3)2 (6-(CF3SO3)2) by syn-
thetic chemistry, and through protonation of the oxo ligand as
NMR instrument at 300 MHz or on a Varian NMR instrument at 400
MHz. Chemical shifts are reported as δ values, downfield from an
internal standard (Me4Si) (1H), as δ values referenced to solvent (2H),
or as δ values referenced to an external standard of R,R,R-trifluoro-
toluene (19F). EI/CI and FAB-MS spectra were recorded at the
University of Illinois, Urbana-Champaign.
F2dp, (2,6-Difluorophenyl)dipyrromethane. A solution of 2,6-
difluorobenzaldehyde (5 g, 35.2 mmol) in 100 mL of freshly distilled
pyrrole was stirred and degassed for 15 min with a stream of argon.
Then, 0.5 mL of trifluoroacetic acid was added dropwise anaerobically.
The resulting red mixture was stirred for an additional 1 h. Then 100
mL of ACS-grade dichloromethane was added. The organic layer was
washed three times with 75 mL of 0.1 M NaOH and dried over
anhydrous MgSO4. Dichloromethane was distilled off using a rotary
evaporator while excess pyrrole was distilled off under high vacuum.
A greenish paste was obtained and was chromatographed over silica
using 4:1 hexane/diethyl ether containing 1% triethylamine. A red (third,
over Br2, Rf ) 0.3) fraction was collected. This fraction was crystallized
in a mixture of hexane/ether (10:1) to yield white needles (4.1 g, 45%).
1H NMR (CDCl3): 8.17 (s, br, 2H, N-H pyrrole), 7.16-7.26 (m, 1H,
para phenyl), 6.87-6.93 (t, 2H, meta phenyl), 6.68-6.70 (m, 2H, γ
pyrrole), 6.13-6.15 (q, 2H, R pyrrole), 6.01 (s, 2H, â pyrrole), 5.91
(s, 1H, CH).
1
followed in a H NMR titration. The physical properties of
5-(CF3SO3) and 6-(CF3SO3)2 have been compared. The oxo
group in 5-(CF3SO3) is more basic than most of the known
oxo-bridged metallic systems, but is similar to the previously
studied µ-oxo complex [(F8TPP)FeIII-O-CuII(TMPA)]+. The
aqueous pKa of 6-(CF3SO3)2 is estimated to be 9.6 ( 2.
The observed room-temperature and variable-temperature 1H
NMR chemical shifts of 5-(CF3SO3) and 6-(CF3SO3)2 are
consistent with the presence of a predominant σ contact shift
mechanism. Linear Curie plots of the respective pyrrole and
MePY2 chemical shifts are consistent with a pure spin state, in
each case S ) 2. The lack of signals at room temperature for
the MePY2 moiety in 6-(CF3SO3)2 must be ascribed to the
presence of the µ-OH- rather than µ-oxo ligand; the weaker
M-(OH) bonds allow for relatively rapid ligand exchange
phenomena near room temperature, leading to line broadening.
Complexes 5 and 6 are comparable in composition and
physical properties (especially, magnetic and paramagnetically
shifted NMR spectra) to the previously studied TMPA ana-
logues, [(F8TPP)FeIII-O-CuII(TMPA)]+ and [(F8TPP)FeIII-
F8TPPH2, Tetrakis(2,6-difluorophenyl)porphyrin. In a 1 L two-
necked round-bottom flask, 4 g (15.5 mmol) of 2,6-difluorodipyr-
romethane and 2.2 g (15.5 mmol, 1 equiv) of 2,6-difluorobenzaldehyde
were mixed with 1 mL of boron trifluoride-diethyl etherate (BF3‚Et2O)
in 750 mL of freshly distilled dichloromethane. The orange solution
was stirred for 2 h. During this time, the color changed from initial
yellow orange through red to purple. Formation of porphyrinogen (λmax
) 500 nm) was monitored using optical spectra. Then, 6.15 g (27.1
mmol, 3.5 equiv) of 2,3-dichloro-5,6-dicyanobenzoquinone (DDQ) was
introduced to oxidize the solution. The mixture was stirred for 8 h.
The volume was reduced and the reaction mixture further purified first
by flash chromatography over basic Al2O3 eluted with CH2Cl2 and then
by column chromatography (silica, 2% methanol in CH2Cl2) to afford
OH-CuII(TMPA)]2+ 13,14
While detailed structural insights for
.
5 and 6 are currently unavailable, one can anticipate very similar
near-linear µ-oxo and bent µ-hydroxo cores, respectively, to be
present. Thus, as previously discussed,13 5 and, especially, 6
may represent structural models for the oxidized FeIII-X-CuII
resting state, or turnover intermediates, in heme-copper oxidases.
1
3 g of F8TPPH2. Rf (alumina, CH2Cl2) ) 0.85. H NMR (CDCl3, 300
MHz): δ 8.8 (m, 8H, pyrrole H), 7.74 (d, 4H, para phenyl H), 7.31
(m, 8H, meta phenyl H).
F8TPPH2-d8. The pyrrole-deuterated F8TPPH2-d8 tetrakis(2,6-difluo-
rophenyl)porphyrin was prepared by modification of the standard
benzaldehyde-pyrrole condensation in propionic acid/nitrobenzene
reflux, as previously reported in the literature.46 To a 1 L two-necked
round-bottom flask were added 500 mL of propionic anhydride (97%)
and 72 mL of D2O (99.9 atom % D) under argon, and the solution was
brought to reflux for 1 h. Under argon, 6.0 g pyrrole was added to this
deuterated propionic acid, and the reaction mixture was refluxed for
an additional 1 h. In air, 125 mL of nitrobenzene and 12.5 g of 2,6-
difluorobenzaldehyde were added. Reflux continued for 1 h, after which
the reaction mixture was cooled to room temperature and placed in a
freezer (-20 °C) for 2 days. The resulting purple solid was collected
via filtration over Celite, washed with water (1 L), dried over air, and
removed from Celite with CH2Cl2 (1 L). The solution was dried over
MgSO4, filtered, and concentrated to dryness via rotary evaporation.
The crude material was first purified over an alumina column using
methylene chloride (Rf ) 0.85) eluent, and then purified further over
a silica column with 50:50 CH2Cl2/hexane to afford 1.3 g of purple
Experimental Section
Reagents and solvents used were of commercially available analytical
reagent quality, and all chemicals were obtained from Aldrich Chemical
Co. unless otherwise noted. Diethyl ether (Et2O), tetrahydrofuran (THF),
and heptane were distilled over sodium/benzophenone ketyl under an
argon atmosphere. Acetonitrile (CH3CN) was distilled from CaH2.
Acetone (CH3C(O)CH3) was shaken with Drierite for several hours
before it was decanted and distilled from fresh Drierite under argon.
Dichloromethane (CH2Cl2) was stirred with concentrated sulfuric acid
for several days and washed with water, sodium carbonate (10%)
solution, and water. It was then dried over anhydrous magnesium
sulfate, filtered, and predried over CaH2 before a final distillation over
fresh CaH2 under an argon atmosphere. Triethylamine (Et3N) was first
dried with KOH and then distilled from CaH2. Preparation and handling
of air-sensitive materials were carried out under an argon atmosphere
with use of standard Schlenk techniques. Deoxygenation of solvents
and solutions was effected either by repeated vacuum/purge cycles using
argon (freeze-pump-thaw) or through bubbling of Ar directly through
the solutions. Solid samples were stored and transferred, and samples
for IR, UV-vis, and NMR spectra were prepared in Vacuum
Atmospheres Co. and/or MBraun Co. dryboxes filled with prepurified
nitrogen. Elemental analyses were performed by Desert Analytics,
Tucson, AZ. Infrared spectra were recorded as Nujol mulls on a Mattson
Galaxy 4030 FT-IR spectrometer driven by a Dell Dimension P133
computer using software written by Mattson. Melting points were
determined on a Mel Temp II, Laboratory Devices. UV-visible spectra
were recorded on a Shimadzu 160 spectrophotometer or a Hewlett-
Packard 8452A diode-array spectrophotometer driven by a Gateway
2000 P75 computer using software written by On-Line Instruments
Systems, Inc. 1H, 2H, and 19F NMR spectra were recorded on a Bruker
1
solid. H NMR (CDCl3, 300 MHz): δ 7.75 (d, 4H, para phenyl H),
7.30 (m, 8H, meta phenyl H).
(F8TPP)FeIIIOH (4) and (F8TPP-d8)FeIIIOH were synthesized
according to published procedures.12
(F8TPP)FeIIIOH. 1H NMR (CDCl3): δ 81.1 (v br, 8H, pyrrole H),
11.5, 10.5 (d, 8H, meta phenyl), 7.3 (s, 4H, para phenyl). UV-vis
(CH2Cl2): 408 (Soret), 572 nm.
(F8TPP)FeII (2). In a 100 mL Schlenk flask, 500 mg (0.6 mmol) of
(F8TPP)FeIIIOH was dissolved in Ar-saturated CH2Cl2 (ca. 50 mL), and
(46) (a) Adler, A. D.; Longo, F. R.; Kampas, F.; Kim, J. J. Inorg. Nucl.
Chem. 1970, 32. (b) Adler, A. D.; Longo, F. R.; Finarelli, J.;
Goldmacher, J.; Assour, J.; Korsakoff, L. J. Org. Chem. 1967, 32,
476.

Oxo- and Hydroxo-Bridged Heme-Copper Assemblies
Inorganic Chemistry, Vol. 38, No. 13, 1999 3101
to it was added a 1 M aqueous solution (ca. 30 mL) of sodium
hydrosulfite (Ar-saturated) under argon. The two solutions were mixed
vigorously for about 30 min. The color of the reaction mixture changed
from brown to bright red. The two layers were separated, and the
organic layer was dried over MgSO4 (under Ar). The solvent was
concentrated in vacuo, and addition of deoxygenated heptane (ca. 50
mL) precipitated 2 as a purplish solid (350 mg, 72%). Anal. Calcd for
C44H20N4F8Fe‚H2O: C, 63.63; H, 2.67; N, 6.75; F, 18.3. Found: C,
layered with 100 mL of anhydrous heptane. A pale yellow solid was
collected after several hours, washed several times with heptane, and
dried in vacuo to afford 425 mg of yellow solid (yield 85%). Anal.
Calcd for (C47H31N3F31BCu)‚CH3CN: C, 48.67; H, 2.83; N, 4.63.
Found: C, 48.77; H, 2.72; N, 4.62. The presence of acetonitrile has
been confirmed by H NMR. H NMR (CD3NO2, 300 MHz): δ 8.65
(d, 2H, o-PY), 7.88 (t, 2H, p-PY), 7.85 (s br, 8H, BArF-H2), 7.66 (s,
4H, BArF-H4), 7.41 (m, 4H, m-PY), 3.06 (br, 8H, -CH2-PY), 2.50
(br, 3H, N-CH3), 2.10 (s, 3H, CH3CN). IR (Nujol, cm-1): 1611 (s,
br, CdC), 1573 (w, CdC) 1115 (s, br, BArF).
[(F8TPP)Fe-O-Cu(MePY2)](CF3SO3), 5-(CF3SO3). In a 100 mL
Schlenk flask equipped with a stir bar were charged 215 mg (0.259
mmol) of (F8TPP)FeIIIOH and 156 mg (0.259 mmol) of [(MePY2)-
CuII](CF3SO3)2. These were stirred for 30 min to effect intimate mixing.
Then deaerated THF or CH3CN (10 mL) was introduced. A brownish-
red solution was generated, to which was promptly added Et3N (36
µL, 1 equiv), via a syringe. The resulting red solution was stirred for
an 1 h and then layered with diethyl ether. The precipitate was filtered
off on a coarse frit and dried in vacuo for 24 h to give 220 mg of
black microcrystalline product (66% yield). Anal. Calcd for (C60H39N7F11-
SO4CuFe)‚2H2O: C, 54.64; H, 3.26; N, 7.438; S, 2.42. Found: C,
54.23; H, 3.02; N, 7.34; S, 2.36. The presence of H2O was confirmed
by examination of the 1H NMR spectrum. 1H NMR (CD2Cl2): δ 67.7
(s, 8H, pyrrole), 23.4-18.9 (d, 4H, -CH2-PY), 9.90-9.24 (d, 8H,
m-phenyl), 7.95 (s, 4H, p-phenyl), 1.2, -2.7, -23.5, and -37.2 (4 s,
4 2H, pyridyl hydrogens), -87.7 and -155.4 (2 s, 2 2H, N-CH2),
-189.4 (brs, 3H, N-CH3). 19F NMR (CH2Cl2): δ -78.74 (s, CF3SO3-),
-95.87 and -96.71 (d, fluorine F8TPP). UV-visible (CH2Cl2): 443,
555 nm.
1
1
1
63.78; H, 2.35; N, 6.84; F, 18.9. H NMR (acetone-d6, 300 MHz): δ
48.71 (s, 8H, pyrrole H), 7.72 (s, 4H, para phenyl H), 7.20-7.38 (d,
8H, meta phenyl H). UV-vis (acetone): 421 nm, 545 nm.
Vinylpyridine-d2 (C4H4NCHdCD2). To a 100 mL Schlenk flask
was introduced (methyl-d3)triphenylphosphonium iodide (4.5 g, 11.05
mmol). The solid was suspended with freshly distilled THF (30 mL)
and cooled to 0 °C. n-BuLi (2.5 M in hexane, 4.65 mL, 1.04 equiv)
was added to the suspension at 0 °C, via a syringe. The resulting deep
orange mixture was stirred for 15 min at 0 °C and allowed to warm to
room temperature. To this solution was added a solution of 2-pyridine-
carboxaldehyde (1.18 g, 1 equiv) in 50 mL of THF under argon. After
stirring at room temperature for 18 h, the solution was filtered and the
solvent was evaporated under reduced pressure. The resulting reddish
oil was purified by column chromatography (silica, 40% ethyl acetate
in hexane) to afford 350 mg of product (30% yield). 1H NMR (CDCl3,
300 MHz): δ 8.51 (d, 1H, pyridyl H6), 7.58 (td, 1H, pyridyl H4),
7.28 (d, 1H, pyridyl H3), 7.05 (m, 1H, pyridyl H5), 6.7 (brs, 1H, CHd).
2H NMR (CDCl3, 400 MHz): δ 6.10 (D cis), 5.4 (D trans).
N-(Methyl-d3)-N-[2-(2-pyridinyl)(ethyl-d2)]-2-pyridine(ethane-d2)-
amine, CD3N(CH2CD2PY)2, MePY2-d7. 2-Vinylpyridine (483 mg, 4.6
mmol) and methylamine-d5 deuteriochloride (112 mg, 1.53 mmol) were
placed in a 10 mL round-bottom flask. The reactants were stirred under
reflux in 3 mL of D2O and 200 mL of CD3OD for 16 h. After cooling,
the methanol was removed and the resulting mixture was washed with
1 N NaOH solution (ca. 10 mL) and extracted with CH2Cl2. The organic
layer was dried over MgSO4 and purified by column chromatography
(alumina, 2% MeOH in ethyl acetate) to afford 175 mg of MePY2-d7
(yield 45%). 1H NMR (CDCl3, 300 MHz): δ 8.33 (d, 2H, pyridyl H6),
7.38 (t, 2H, pyridyl H4), 6.95 (d, 2H, pyridyl H3), 6.91 (t, 1H, pyridyl
H5), 2.67 (s, 4H, CHd). 2H NMR (CDCl3, 400 MHz): δ 3.165 (br s,
-CD2-PY), 2.5 (br s, -CD3).
The same procedure was used to prepare the following complexes.
[(F8TPP-d8)Fe-O-Cu(MePY2)](CF3SO3). 2H NMR (CH2Cl2): δ
66.36 (s, D-pyrrole).
[(F8TPP)Fe-O-Cu(CD3N)(CH2CD2PY)2](CF3SO3). 2H NMR (CH2-
Cl2): δ 23.75-19.61 (d, -CD2-PY), -182.76 (s, -CD3).
[(F8TPP)Fe-O-Cu(CD3N)(CD2CD2PY)2](CF3SO3). 2H NMR (CH2-
Cl2): δ 24.26-20.20 (d, -CD2-PY), -75.78 and - 149 (-CD2-N),
-182.76 (s, -CD3).
[(F8TPP)Fe-O-Cu(MePY2)](BArF), 5-(BArF). In a 100 mL
Schlenk flask equipped with a stir bar were placed, in the drybox, 210
mg (0.2 mmol) of (F8TPP)FeII and 302 mg (0.2 mmol, 1 equiv) of
[(MePY2)CuI(CH3CN)](BArF), to which was added air-free freshly
distilled THF (20 mL). The reaction mixture was cooled to - 78 °C
(dry ice-acetone bath) and stirred for 30 min. The solution was
subjected to three cycles of vacuum/O2 purging (O2; UHP, Aldrich).
The solution was finally allowed to warm slowly at room temperature
and layered with 80 mL of deoxygenated heptane. After 5 h, the solution
was filtered and the black microcrystalline solid dried in vacuo (330
mg, 65% yield). Anal. Calcd for (C91H51N7F31BOCuFe)‚H2O: C, 54.77;
N-(Methyl-d3)-N-[2-(2-pyridinyl)(ethyl-d4)]-2-pyridine(ethane-d4)-
amine, CD3N(CD2CD2PY)2, MePY2-d11. The same procedure as the
one described for MePY2-d7 was used, starting with the deuterated
d2-2-vinylpyridine. 1H NMR (CDCl3, 300 MHz): δ 8.45 (d, 2H, pyri-
dyl H6), 7.50 (t, 2H, pyridyl H4), 7.07 (d, 2H, pyridyl H3), 7.05 (t,
2
1H, pyridyl H5). H NMR (CDCl3, 400 MHz): δ 2.995 (m, -CD2-
CD2-), 2.518 (brs, CD3).
[(MePY2)CuII](CF3SO3)2, 3-(CF3SO3)2. To a solution of CuII(CF3-
SO3)2 (2 g, 5.5 mmol) in MeOH (20 mL) under a stream of argon was
added a solution of deoxygenated MePY2 (1.33 g, 5.5 mmol, 1 equiv)
dropwise, and the resulting blue mixture was stirred for 1 h. The solvent
was then evaporated, and freshly distilled air-free diethyl ether was
added. The resulting precipitate was filtered off, washed several times
with diethyl ether, and dried in vacuo to give [(MePY2)CuII](CF3SO3)2
as a blue solid (2.9 g, 87%). Anal. Calcd for C17H19N3CuF6S2O6: C,
33.86; H, 3.18; N, 6.97; S, 10.63. Found: C, 33.90; H, 3.07; N, 6.76;
S, 10.25.
1
H, 2.68; N, 4.91. Found: C, 54.39; H, 2.4; N, 4.93. H NMR (CD2-
Cl2): δ 67.6 (s, 8H, pyrrole), 24.01-20.53 (d, 4H, -CH2-PY), 9.8-
9.14 (d, 8H, m-phenyl), 7.67-7.51 (m, 16H, 8H BArF + p-phenyl),
-84.4 and -147.4 (2 s, 2 2H, -CH2-N), -177.1 (brs, 3H, -CH3).
[(F8TPP)Fe-OH-Cu(MePY2)](CF3SO3)2, 6-(CF3SO3)2. In a 100
mL Schlenk flask equipped with a stir bar were charged 200 mg (0.241
mmol) of (F8TPP)FeIIIOH and 145 mg (0.241 mmol) of [(MePY2)-
CuII](CF3SO3)2. These were stirred for 30 min to effect intimate mixing.
Then deaerated CH3CN (7.5 mL) was introduced. A brownish-red
solution was generated and stirred at room temperature for 1 h. The
solution was then layered with toluene and allowed to stay at -20 °C
for several hours. The black precipitate was filtered off on a coarse frit
and dried in vacuo for 24 h to give 275 mg of black microcrystalline
product (80% yield). Anal. Calcd for (C61H40N7F14S2O7CuFe)‚2H2O:
C, 49.89; H, 3.02; N, 6.68. Found: C, 49.34; H, 3.16; N, 6.32. The
presence of H2O was confirmed by examination of the 1H NMR
spectrum. 1H NMR (CD2Cl2, 300 MHz, 297 K): δ 70.3 (s, 8H, pyrrole),
11.54-10.73 (d, 8H, m-phenyl), 7.95 (s, 4H, p-phenyl).
[(MePY2)Cu(I)(CH3CN)](BArF), 1-(BArF). To a flame-dried 100
mL Schlenk flask were transferred anaerobically 0.5 g (0.56 mmol) of
sodium tetrakis[(3,5)-bis(trifluoromethyl)phenyl]borate (NaBArF), 280
mg (1.5 equiv, 0.85 mmol) of copper(I) tetrakisacetonitrile perchlorate
([Cu(CH3CN)4](ClO4)), and an oven-dried spin bar. Then 30 mL of
degassed freshly distilled dry diethyl ether was added. The beige mixture
was stirred for 30 min and then filtered into another Schlenk flask to
remove the excess of [Cu(CH3CN)4](ClO4). After reduction of the
solvent volume (in vacuo), addition of deoxygenated dry heptane gave
450 mg of a white microcrystalline material (73%). Then 100 mg (0.415
mmol) of vacuum-dried MePY2 was dissolved in 25 mL of freshly
distilled diethyl ether. The yellow solution was introduced anaerobically
into the Schlenk flask containing the [Cu(CH3CN)4](BArF). The bright
yellow mixture was stirred for 30 min and then filtered into a 250 mL
Schlenk flask. The yellow solution was concentrated before being
NMR Titration. Protonation of 5-(CF3SO3). In the glovebox, a
solution of 5-(CF3SO3) (30 mg, 0.0234 mmol) in CD2Cl2 was prepared
and transferred to an NMR tube, and the 1H NMR spectrum was
recorded. CF3SO3H (0.4 µL, 4.68 × 10-6 mol, 0.2 equiv), from a freshly
opened vial from Aldrich, was then introduced via a 1 µL syringe, and

3102 Inorganic Chemistry, Vol. 38, No. 13, 1999
Kopf et al.
1
the H NMR spectrum was recorded. This process of adding aliquots
+ (ø0Mw) - øD, where ∆ν is the difference in shift of the reference
signal in hertz, ν is the spectrometer frequency, c is the concentration
of the complex in moles per liter, ø0 is the solvent susceptibility, Mw
is the molecular weight of the complex, and øD is the diamagnetic
contribution to susceptibility. The latter was calculated using tabulated
Pascal’s constants.48 The magnetic moment µeff is derived from the
following formula: µeff ) 2.84x(ø′mT)µB.
of acid and recording spectra was performed several times until the
titration “end-point” was reached. The same procedure was used for
the pKa determination, using the appropriate acid (see below).
pKa Determination. General Preparation of the Protonated
Amines. In a 25 mL round-bottom flask containing 10 mL of absolute
ethanol was placed 5 mmol of the corresponding amine (i.e., aniline,
2,4-lutidine, methylmorpholine, and so on). To the solution was added
1 equiv of triflic acid, and the mixture was stirred for an additionnal
30 min. The volume of ethanol was reduced to 1 mL, and ∼15-20
mL of diethyl ether was added. After 1 or 2 h, a white crystalline solid
Electrical Conductivity. Electrical conductivity measurements were
carried out in CH3CN with a Barnstead model PM-70CB conductivity
bridge and YSI model 3403 conductivity cell. The cell constant was
determined with a standard aqueous solution of KCl (0.1 M). The molar
conductivity, Λm, of a solution of 5-(CF3SO3) was determined from
the expression Λm ) 1000K/Cm, where K is the cell constant divided
by the measured resistance and Cm is the molar concentration of the
solute (∼10-3 M). Λm (CH3CN) ) 154 Ω cm2 mol-1, corresponding
to a 1:1 electrolyte.26
1
was filtered on a coarse frit and dried in vacuo for 12 h. H NMR
(CDCl3 + 1 drop of d6-DMSO).
Methylmorpholinium triflate: δ 9.56 (s br, 1H, NH+), 3.95 (d,
2H, -CH2-), 3.75 (t, 2H, -CH2-), 3.44 (d, 2H, -CH2-), 3.01 (t,
2H, -CH2-), and 2.9 (s, 3H, -CH3). Mp: 92.5 °C.
Stopped-Flow Kinetics. The instrumentation and methods for
Anilinium triflate: δ 9.8 (v br, 3H, NH3+), 7.37 (m, 4H, H aromatic).
Mp: 268 °C.
analysis were as previously described.35,36 Here, seven runs at -90 °C
in acetone (Uvasol, Fluka) were carried out, with 2.75 × 10-4
1-(BArF) and 2.73 × 10-4 M 2; [O2] ) 5.92 × 10-3 M.
M
2,4-Lutidinium triflate: δ 15.13 (s br, 1H, NH+), 8.6 (d, 1H, H
ortho), 7.714 (d, 1H, meta) 7.65 (s, 1H, meta), 2.58 (s, 3H, -CH3),
and 2.47 (s, 3H, -CH3). Mp: 62.5 °C.
Acknowledgment. We are grateful (K.D.K., National In-
stitutes of Health; A.D.Z., Swiss National Science Foundation)
for research support. We are also thankful to Dr. Honorio V.
Obias for initial observations.
Benzylammonium triflate: δ 6.96 (s br, 3H, NH3+), 6.31-6.22 (m,
5H, H aromatic), and 2.899 (s, 2H, -CH2). Mp: 175.5 °C.
Magnetic Susceptibility Measurements. Room-temperature solid
magnetic moments were determined with the use of a Johnson Matthey
magnetic susceptibility balance calibrated against Hg[Co(SCN)4)].
Solution magnetic moment measurements were conducted using the
Evans method47 on a Bruker NMR instrument at 300 MHz. To this
end, samples with known concentration were made up in CD2Cl2. From
the downfield shift of the TMS signal, with respect to that from the
capillary reference tube, the paramagnetic susceptibility ø′m was
calculated using the following formula: ø′m ) (-3/4π)(∆ν/ν)(1000/c)
Supporting Information Available: UV-vis spectra during the
oxygenation of a 1:1 mixture of 1-(BArF) and 2 (Figure S1), absorbance
vs time in the stopped-flow kinetics experiment (Figure S2), and
variable-temperature 1H NMR information (spectra and δ vs 1/T plots)
for 5-(CF3SO3) and 6-(CF3SO3)2 (Figures S3 and S4). This material
is available free of charge via the Internet at http://pubs.acs.org.
IC981431+
(48) Handbook of Chemistry & Physics, 74th ed.; Lide, D. R., Ed.; CRC
(47) Evans, D. F. J. Chem. Soc. 1959, 2003.
Press: Boca Raton, FL, 1993.