Synthesis and characterization of reduced heme and heme/copper carbonmonoxy species

Article abstract of DOI:10.1021/ic020521i

Carbon monoxide readily binds to heme and copper proteins, acting as a competitive inhibitor of dioxygen. As such, CO serves as a probe of protein metal active sites. In our ongoing efforts to mimic the active site of cytochrome c oxidase, reactivity toward carbon monoxide offers a unique opportunity to gain insight into the binding and spectroscopic characteristics of synthetic model compounds. In this paper, we report the synthesis and characterization of CO-adducts of (^5/6L)Fe^II, [^(5/6L)Fe^II...Cu^I](B (C₆F₅)₄), and [(TMPA)Cu^I(CH₃CN)](B(C₆ F₅)₄), where TMPA = tris(2-pyridylmethyl)amine and ^5/6L = a tetraarylporphyrinate tethered in either the 5-position (⁵L) or 6-position (⁶L) to a TMPA copper binding moiety, Reaction of (^5/6L)Fe^II {in THF (293 K): UV-vis 424 (Soret), 543-544 nm; ¹H NMR δ_pyrrole 52-59 ppm (4 peaks); ²H NMR (from (⁵L-d₈)Fe^II) δ_pyrroole 53.3, 54.5, 55.8, 56.4 ppm} with CO in solution at RT yielded (^5/6L)Fe^II-CO {in THF (293 K): UV-vis 413-414 (Soret), 532-533 nm; IR v(CO)_Fe 1976-1978 cm^-1; ¹H NMR δ_pyrrole 8.8 ppm; ²H NMR (from (⁵L-d₈)Fe_II-CO) δ_pyrrole 8.9 ppm; ¹³C NMR δ_(CO)Fe 206.8-207.1 ppm (2 peaks)}. Experiments repeated in acetonitrile, acetone, toluene, and dichloromethane showed similar spectroscopic data. Binding of CO resulted in a change from five-coordinate, high-spin Fe(II) to six-coordinate, low-spin Fe(II), as evidenced by the upfield shift of the pyrrole resonances to the diamagnetic region (¹H and ²H NMR spectra), Addition of CO to [(^5/6L)Fe^II...Cu^I](B (C₆F₅)₄) {in THF (293 K): UV-vis (⁶L only) 424 (Soret), 546 nm; ¹H NMR δ_pyrrole 54-59 ppm (multiple peaks); ²H NMR (from [(⁵L-d₈)Fe^II...Cu^I (B(C₆F₅)₄)) δ_pyrrole 53.4 ppm (br)} gave the bis-carbonyl adduct [(^5/6L)Fe^IICO...Cu^I-CO](B (C₆F₅)₄) {in THF (293 K): UV-vis (⁶L only) 413 (Soret), 532 nm; IR v(CO)Fe 1971-1973 cm^-1, v(CO)_Cu 2091-2093 cm^-1, ～2070(sh) cm^-1; ¹H NMR δ_pyrole 8.7-8.9 ppm; ²H NMR (from [(⁵L-d₈)Fe^II·· ·Cu^ICO](B(C₆F₅)₄)) δ_pyrole 8.9 ppm; ¹³C NMR δ_(CO)Fe 206.8-208.1 ppm (2 peaks), δ_CO)Cu 172.4 (⁵L), 178,2 (⁶L) ppm}. Experiments in acetonitrile, acetone, and toluene exhibited spectral features similar to those reported, The [(^5/6L)Fe^II-CO··· Cu^ICO](B(C₆F₅)₄) compounds yielded (CO)_Fe spectra analogous to those seen for (^5/6L)Fe^II-CO and (CO)_Cu, spectra similar to those seen for [(TMPA)Cu^ICO](B(C₆F₅)₄) {in THF (293 K): IR v(CO)_Cu, 2091 cm^-1, ～2070(sh) cm^-1; ¹³C NMR δ_(CO)Cu 180.3 ppm}. Additional IR studies were performed in which the [⁵L)Fe^II-CO···Cu^I-CO] (B(C₆F₅)₄) in solution was bubbled with argon in an attempt to generate the iron-only mono-carbonyl [⁵L)Fe^II-CO···Cu] (B(C₆F₅)₄) species; in coordinating solvent or with axial base present, decreases in characteristic IR-band intensities revealed complete loss of CO from copper and variable loss of CO from the heme.

Full text of DOI:10.1021/ic020521i

Inorg. Chem. 2003, 42, 3016−3025
Synthesis and Characterization of Reduced Heme and Heme/Copper
Carbonmonoxy Species
Ryan M. Kretzer, Reza A. Ghiladi, Estelle L. Lebeau, Hong-Chang Liang, and Kenneth D. Karlin*
Department of Chemistry, The Johns Hopkins UniVersity, Charles and 34th Streets,
Baltimore, Maryland 21218
Received August 22, 2002
Carbon monoxide readily binds to heme and copper proteins, acting as a competitive inhibitor of dioxygen. As
such, CO serves as a probe of protein metal active sites. In our ongoing efforts to mimic the active site of cytochrome
c oxidase, reactivity toward carbon monoxide offers a unique opportunity to gain insight into the binding and
spectroscopic characteristics of synthetic model compounds. In this paper, we report the synthesis and characterization
II
II
I
I
of CO-adducts of (5/6L)Fe , [(5/6L)Fe ‚‚‚Cu](B(C F ) ), and [(TMPA)Cu(CH CN)](B(C F ) ), where TMPA ) tris(2-
6
5 4
3
6 5 4
pyridylmethyl)amine and 5/6L ) a tetraarylporphyrinate tethered in either the 5-position (5L) or 6-position (6L) to a
TMPA copper binding moiety. Reaction of (5/6L)Fe {in THF (293 K): UV−vis 424 (Soret), 543−544 nm; 1H NMR
II
II
δ
pyrrole 52−59 ppm (4 peaks); 2H NMR (from (5L-d8)Fe ) δpyrrole 53.3, 54.5, 55.8, 56.4 ppm} with CO in solution at
RT yielded (5/6L)Fe −CO {in THF (293 K): UV−vis 413−414 (Soret), 532−533 nm; IR ν(CO)Fe 1976−1978 cm-1;
II
2
II
1H NMR δpyrrole 8.8 ppm; H NMR (from (5L-d8)Fe −CO) δpyrrole 8.9 ppm; 13C NMR δ(CO)Fe 206.8−207.1 ppm (2
peaks)}. Experiments repeated in acetonitrile, acetone, toluene, and dichloromethane showed similar spectroscopic
data. Binding of CO resulted in a change from five-coordinate, high-spin Fe(II) to six-coordinate, low-spin Fe(II), as
evidenced by the upfield shift of the pyrrole resonances to the diamagnetic region (1H and 2H NMR spectra).
Addition of CO to [(5/6L)Fe ‚‚‚Cu](B(C F ) ) {in THF (293 K): UV−vis (6L only) 424 (Soret), 546 nm; H NMR
II
I
1
6
5 4
δpyrrole 54−59 ppm (multiple peaks); H NMR (from [(5L-d8)Fe ‚‚‚Cu](B(C F ) )) δpyrrole 53.4 ppm (br)} gave the
2
II
I
6
5 4
bis-carbonyl adduct [(5/6L)Fe −CO‚‚‚Cu−CO](B(C F ) ) {in THF (293 K): UV−vis (6L only) 413 (Soret), 532 nm;
II
I
6
5 4
IR ν(CO)Fe 1971−1973 cm-1, ν(CO)Cu 2091−2093 cm-1, ∼2070(sh) cm-1; 1H NMR δpyrrole 8.7−8.9 ppm; 2H NMR
(from [(5L-d8)Fe −CO‚‚‚Cu−CO](B(C F ) )) δpyrrole 8.9 ppm; 13C NMR δ(CO)Fe 206.8−208.1 ppm (2 peaks), δ(CO)Cu
II
I
6
5 4
172.4 (5L), 178.2 (6L) ppm}. Experiments in acetonitrile, acetone, and toluene exhibited spectral features similar to
those reported. The [(5/6L)Fe −CO‚‚‚Cu−CO](B(C F ) ) compounds yielded (CO)Fe spectra analogous to those
II
I
6
5 4
seen for (5/6L)Fe −CO and (CO)Cu spectra similar to those seen for [(TMPA)Cu−CO](B(C F ) ) {in THF (293 K):
II
I
6
5 4
IR ν(CO)Cu 2091 cm-1, ∼2070(sh) cm-1; 13C NMR δ(CO)Cu 180.3 ppm}. Additional IR studies were performed in
which the [(5L)Fe −CO‚‚‚Cu−CO](B(C F ) ) in solution was bubbled with argon in an attempt to generate the
II
I
6
5 4
I
iron-only mono-carbonyl [(5L)Fe −CO‚‚‚Cu](B(C F ) ) species; in coordinating solvent or with axial base present,
II
6
5 4
decreases in characteristic IR-band intensities revealed complete loss of CO from copper and variable loss of CO
from the heme.
Introduction
through fast time scale investigations of the photochemical
ejection and subsequent rebinding of CO.3-8 Other examples
Carbon monoxide binding to hemes and hemoproteins has
always been a subject of great interest, especially as a probe
or surrogate for dioxygen (O2) binding to iron-porphyrin
centers.1-3 Thus, especially for the O2-carriers hemoglobin
and myoglobin, CO has been employed to obtain structural
insights and used as a dynamic probe of the active site
(1) Ghosh, A. In The Porphyrin Handbook; Kadish, K. M., Smith, K.
M., Guilard, R., Eds.; Academic Press: San Diego, CA, 2000; Vol.
7.
(2) Spiro, T. G.; Zgierski, M. Z.; Kozlowski, P. M. Coord. Chem. ReV.
2001, 219-221, 923-936.
(3) Spiro, T. G.; Kozlowski, P. M. Acc. Chem. Res. 2001, 34, 137-144.
(4) Uchida, T. I., H.; Ishimori, K.; Morishima, I.; Nakajima, H.; Aono,
S.; Mizutani, Y.; Kitagawa, T. Biochemistry 2000, 39, 12747-12752.
* To whom correspondence should be addressed. E-mail: karlin@jhu.edu.
3016 Inorganic Chemistry, Vol. 42, No. 9, 2003
10.1021/ic020521i CCC: $25.00 © 2003 American Chemical Society
Published on Web 03/29/2003

Reduced Heme and Heme/Copper Carbonmonoxy Species
delineating the importance of carbon monoxide can be found
in several biologically important systems or processes. The
enzyme heme-oxygenase (HO) breaks down heme forming
biliverdin and CO, thereby providing the means for removing
damaged (i.e., toxic) heme.9-12 The heme-protein sensor
CooA mediates transcriptional and regulatory events upon
binding of CO to its unique heme prosthetic group; this and
other heme sensors of small molecules (i.e., CO, O2, and
NO) have elicited excitement and further investigation into
newer types of heme proteins.13,14 Further, there is also the
suggestion that CO itself can act as a physiological
effector.15-18
Figure 1. Crystal structure of the carbon monoxide adduct of cytochrome
c oxidase. Coordinates (1OCO) were taken from the Protein Data Bank
(Brookhaven) and displayed using the program Rasmol.
Our own interest in carbon monoxide chemistry arises
from our general program in synthetic modeling of the active
site of cytochrome c oxidase (CcO),19-25 the terminal
respiratory enzyme for aerobic organisms responsible for the
4e-/4H+ reduction of O2 to water occurring at a heme-copper
heterobinuclear active site; protein X-ray structures are now
available.26,27 Carbon monoxide bound to the heme of the
heme-Cu active site has been a very important protein
derivative in the elucidation of the CcO O2-binding and
reduction mechanism.28-32 Flash photolysis of the photolabile
Chart 1
(5) Chu, K. V. J.; McMahon, B. H.; Sweet, R. M.; Berendzen, J.;
Schlichting, I. Nature 2000, 403, 921-923.
(6) Taube, D.; Traylor, T. G.; Magde, D.; Walda, K.; Luo, J. J. Am. Chem.
Soc. 1992, 114, 9182-9188.
CO-adduct has been used as an active site probe, in fact
showing that the CO switches transiently to the nearby
(∼ 4.5 Å away) CuB.33-35 A crystal structure of the heme-
CO adduct of cytochrome c oxidase has also been reported
(Figure 1).26
Considerable efforts in our modeling studies have focused
on the binucleating ligands 5L and 6L, wherein a tetradentate
derivatized TMPA moiety {TMPA ) tris(2-pyridylmethyl)-
amine} is tethered to the periphery of a tetraarylporphyrinate
through one pyridyl group of TMPA, either in the 5-position
(5L) or 6-position (6L) (Chart 1).24,25,36-38 Utilizing these
ligands, we described µ-oxo complexes [(5/6L)FeIII-O-
CuII]+, formed from O2-reactions with reduced precursors
[(5/6L)FeII‚‚‚CuI]+;24,25 analogous chemistry occurs with
separate (untethered) heme and copper complexes, forming
[(F8TPP)FeIII-O-CuII(TMPA)]+ {F8TPP ) tetrakis(2,6-
difluorophenyl)porphyrinate(2-)}.19 In fact, recent work has
(7) Greenwood, C.; Gibson, Q. H. J. Biol. Chem. 1967, 242, 1782-
1787.
(8) Greenwood, C.; Gibson, Q. H. Biochem. J. 1963, 86, 541-554.
(9) Sono, M.; Roach, M. P.; Coulter, E. D.; Dawson, J. H. Chem. ReV.
1996, 96, 2841-2887.
(10) Ortiz de Montellano, P. R.; Wilks, A. AdV. Inorg. Chem. 2001, 51,
359-407.
(11) Ortiz de Montellano, P. R. Curr. Opin. Chem. Biol. 2000, 4, 221-
227.
(12) Fujii, H.; Zhang, Y.; Tomita, T.; Ikeda-Saito, M.; Yoshida, T. J. Am.
Chem. Soc. 2001, 123, 6475-6484.
(13) Jain, R.; Chan, M. K. J. Biol. Inorg. Chem., JBIC 2003, 8, 1-11.
(14) Rodgers, K. R. Curr. Opin. Chem. Biol. 1999, 3, 158-167.
(15) Verma, A.; Hirsch, D. J.; Glatt, C. E.; Ronnett, G. V.; Snyder, S. H.
Science 1993, 259, 381-384.
(16) Snyder, S. H.; Jaffrey, S. R.; Zakhary, R. Brain Res. ReV. 1998, 26,
167-175.
(17) Dore, S.; Goto, S.; Sampei, K.; Blackshaw, S.; Hester, L. D.; Ingi,
T.; Sawa, A.; Traystman, R. J.; Koehler, R. C.; Snyder, S. H.
Neuroscience 2000, 99, 587-592.
(18) Cary, S. P. L.; Marletta, M. A. J. Clin. InVest. 2001, 107, 1071-
1073.
(19) Ghiladi, R. A.; Hatwell, K. R.; Karlin, K. D.; Huang, H.-w.; Moe¨nne-
Loccoz, P.; Krebs, C.; Huynh, B. J.; Marzilli, L. A.; Cotter, R. J.;
Kaderli, S.; Zuberbu¨hler, A. D. J. Am. Chem. Soc. 2001, 123, 6183-
6184.
(20) Ghiladi, R. A.; Kretzer, R. M.; Guzei, I.; Rheingold, A. L.; Neuhold,
Y.-M.; Hatwell, K. R.; Zuberbu¨hler, A. D.; Karlin, K. D. Inorg. Chem.
2001, 40, 5754-5767.
(21) Kopf, M.-A.; Karlin, K. D. In Biomimetic Oxidations; Meunier, B.,
Ed.; Imperial College Press: London, 2000; Chapter 7, pp 309-362.
(22) Kopf, M.-A.; Neuhold, Y.-M.; Zuberbu¨hler, A. D.; Karlin, K. D. Inorg.
Chem. 1999, 38, 3093-3102.
(23) Kopf, M.-A.; Karlin, K. D. Inorg. Chem. 1999, 38, 4922-4923.
(24) Ghiladi, R. A.; Ju, T. D.; Lee, D.-H.; Moe¨nne-Loccoz, P.; Kaderli,
S.; Neuhold, Y.-M.; Zuberbu¨hler, A. D.; Woods, A. S.; Cotter, R. J.;
Karlin, K. D. J. Am. Chem. Soc. 1999, 121, 9885-9886.
(25) 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.
(28) Babcock, G. T. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 12971-12973.
(29) Kitagawa, T.; Ogura, T. Prog. Inorg. Chem. 1997, 45, 431-479.
(30) Szundi, I.; Liao, G.-L.; Einarsdottir, O. Biochemistry 2001, 40, 2332-
2339.
(31) Han, S.; Takahashi, S.; Rousseau, D. L. J. Biol. Chem. 2000, 275,
1910-1919.
(32) Giuffre, A.; Forte, E.; Antonini, G.; D’Itri, E.; Brunori, M.; Soulimane,
T.; Buse, G. Biochemistry 1999, 38, 1057-1065.
(33) Woodruff, W. H. J. Bioenerg. Biomembr. 1993, 25, 177-188.
(34) Einarsdottir, O.; Dyer, R. L.; Lemon, D. D.; Killough, P. M.; Hubig,
S. M.; Atherton, S. J.; Lopez-Garriga, J. J.; Palmer, G.; Woodruff,
W. H. Biochemistry 1993, 32, 12013-12024.
(35) Koutsoupakis, K.; Stafrakis, S.; Pinakoulaki, E.; Soulimane, T.;
Varotsis, C. J. Biol. Chem. 2002, 277, 32860-32866.
(36) 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.
(37) Ju, T. D.; Woods, A. S.; Cotter, R. J.; Moe¨nne-Loccoz, P.; Karlin, K.
D. Inorg. Chim. Acta 2000, 297, 362-372.
(38) Moe¨nne-Loccoz, P.; Richter, O.-M. H.; Huang, H.-w.; Wasser, I. M.;
Ghiladi, R. A.; Karlin, K. D.; de Vries, S. J. Am. Chem. Soc. 2000,
122, 9344-9345.
(26) 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.
(27) Michel, H.; Behr, J.; Harrenga, A.; Kannt, A. Annu. ReV. Biophys.
Biomol. Struct. 1998, 27, 329-356.
Inorganic Chemistry, Vol. 42, No. 9, 2003 3017

Kretzer et al.
6
293 K) 8.76 (br, 3 H), 7.86 (m, 3 H), 7.50 (br, 6 H), 4.29 (v br, 6
H), 2.05 (s, CH3CN). Anal. Calcd for [(TMPA)CuI(CH3CN)]-
(B(C6F5)4 C44H21BCuF20N5: C, 49.21; H, 1.97; N, 6.52. Found:
C, 48.90; H, 1.97; N, 6.42.
(5L-d8)FeII. This complex was synthesized in an analogous
procedure to that previously reported for (6L-d8)FeII.40 UV-vis (nm,
THF, 293 K) 424 (Soret), 543; 2H NMR (61.3 MHz, THF, 293 K)
53.3, 54.5, 55.8, 56.4 ppm (pyrr H).
shown that O2-adducts of the L or untethered systems can
best be described as peroxo species, [(6L)FeIII-(O22-)-CuII]+
24 or [(F8TPP)FeIII-(O22-)-CuII(TMPA)]+,19 respectively. To
expand and broaden the scope of this work, we describe here
the CO-binding chemistry of heme(empty-tether) (5/6L)FeII
and heme-copper [(5/6L)FeII‚‚‚CuI](B(C6F5)4) complexes (see
Chart 1). Bis-carbonyl (i.e., Fe-CO‚‚‚Cu-CO) or mono-
carbonyl (heme-CO only) adducts were formed and charac-
terized using a variety of probes, including UV-vis, infrared
(IR), and multinuclear (i.e., 1H, 2H and 13C) magnetic
resonance spectroscopies. The importance of this area is
reflected by the recently published complementary study by
Collman and co-workers39 on carbon monoxide binding to
synthetic trisimidazole-copper picket metalloporphyrins.
[(5L)FeII‚‚‚CuI](B(C6F5)4). In the glovebox, in a 50 mL Schlenk
flask were dissolved (5L)FeII (120 mg, 0.11 mmol) and [CuI(CH3-
CN)4](B(C6F5)4) (100 mg, 0.11 mmol, 1 equiv) in ∼2-3 mL
deaerated THF. After allowing the deep red solution to stir for 15
min, 10-15 mL deaerated heptane (5:1 v/v) was added, thereby
forming an opaque precipitate. The mother liquor was removed by
decantation and the solid washed with fresh, deaerated heptane.
After the residual solvent was removed under vacuum for 45 min
on the benchtop, the flask with resulting solid was reintroduced
into the glovebox. [(5L)FeII‚‚‚CuI](B(C6F5)4) was thus obtained in
high yield as a deep red/purple microcrystalline solid. The purity
was judged to be suitable based on 1H NMR spectroscopy (lack of
FeIII-OH impurities at δ ) 80-82 ppm) and mass spectrometry.
UV-vis (nm, toluene, 293 K) 413, 432(sh) (Soret), 530, 556(sh);
1H NMR (400 MHz, THF-d8, 293 K) 58.8 (s, sh, pyrr H), 57.7 (s,
sh, pyrr H), 54.7 (s, sh, pyrr H), 54.0 (s, sh, pyrr H), ca. 9.0 (s, br),
8.31 (s, sh), 7.4-8.0 (mult peaks, br, pyridyl H), 7.25 (s, sh), 6.80
(s, sh), 6.41 (s, sh), 5.77 (d, J ) 56.8 Hz), 5.10 (s, sh), 4.73 (s, sh),
4.28 (s, sh), ca. 2.6 (s, br). ESI-MS Found: 1157.5 {m/z [M -
(B(C6F5)4)-]+}. Calcd: 1158.4.
[(5L-d8)FeII‚‚‚CuI](B(C6F5)4). In a glovebox, (5L-d8)FeII (20 mg,
0.018 mmol) and [CuI(CH3CN)4](B(C6F5)4) (17 mg, 0.019 mmol)
were dissolved in 1 mL of THF and allowed to stir for 5 min,
thereby forming [(5L-d8)FeII‚‚‚CuI](B(C6F5)4) in situ. This deep red
solution was then transferred to an airtight, screw-cap NMR tube.
2H NMR (61.3 MHz, THF, 293 K) 53.4 ppm (s, br, pyrr H).
UV-Vis Spectroscopy. Absorption spectra were obtained at
room temperature using either a Shimadzu UV 160U or a Hewlett-
Packard model 8453 diode array spectrophotometer operating HP
Chemstation software. Each sample was prepared in an air-free
cuvette assembly in which a 10 mm quartz cuvette was fused to a
glass tube that was equipped with a high-vacuum stopcock and a
14/20 ground glass joint as described elsewhere.43,44 In a drybox,
solid (5/6L)FeII or [(5/6L)FeII‚‚‚CuI] (B(C6F5)4) was dissolved in one
of the organic solvents chosen for these studies. On the benchtop,
carbonylation was effected by puncturing the rubber septum of the
cuvette assembly with a long needle, followed by direct bubbling
of CO through the sample solution for ∼10 s. For each solvent,
two samples were prepared in separate cuvettes, the first at a
concentration at which the Soret band fell between 0.5 and 1
absorbance units and the second at a concentration where the R
band fell in this absorbance range. Extinction coefficients (M-1
cm-1) were determined as the average of at least two trials. All
wavelengths are reported in nanometers.
Experimental Section
Materials and Methods. All reagents and solvents were
purchased from commercial sources and were of reagent quality
unless otherwise stated. Air-sensitive compounds were handled
under argon atmosphere using standard Schlenk techniques, or in
an MBraun Labmaster 130 inert atmosphere (<1 ppm O2, <1 ppm
H2O) glovebox filled with nitrogen. Acetonitrile (CH3CN), meth-
ylene chloride (CH2Cl2), and heptane were distilled from calcium
hydride, tetrahydrofuran and toluene from sodium/benzophenone,
and acetone from Drierite (97% CaSO4, 3% CoCl2) all under argon.
Deoxygenation of these solvents was achieved by bubbling with
argon for 30 min followed by 3 freeze/pump/thaw cycles prior to
introduction into the glovebox. NMR solvents were distilled as
described here and deoxygenated through 5 freeze/pump/thaw
cycles. Addition of carbon monoxide (Matheson Gas Products) for
UV-vis, IR, H NMR, and H NMR spectroscopic studies was
effected by direct bubbling through an R & D Separations oxygen/
moisture trap model OT3-4 into the sample solution via a syringe
needle. Electrospray ionization (ESI-MS) mass spectra were
recorded on a Finnigan LCQ ion trap mass spectrometer (San Jose,
CA) equipped with a standard electrospray source. 13CO was
purchased in 100 mL breakseals from ICON Services, Inc., Marion,
NY.
1
2
Synthesis. (F8TPP)FeII,20,22 (F8TPP- d8)FeII,20,22 (5L)FeII,25 (6L)-
FeII,25,40 (6L-d8)FeII,40 [(6L)FeII‚‚‚CuI](B(C6F5)4),40 [(6L-d8)FeII‚‚‚CuI]-
(B(C6F5)4),40 and [CuI(CH3CN)4](B(C6F5)4)41 were synthesized
according to published procedures.
[(TMPA)CuI(CH3CN)](B(C6F5)4). Under an Ar atmosphere in
a standard 100 mL Schlenk flask was added the off-white solid
TMPA42 (110 mg, 0.380 mmol) and 25 mL degassed diethyl ether.
To the resulting light yellow solution was added 1 equiv of
[CuI(CH3CN)4](B(C6F5)4) (343 mg, 0.378 mmol) under Ar. The
now light orange solution was stirred for 15 min, followed by the
slow addition of a large excess (∼50 mL) of degassed pentane over
the course of 10 min. The flask was cooled to -20 °C for 2 h,
after which an oily, yellow precipitate had formed. After decanta-
tion, the yellow precipitate was dried in vacuo overnight to yield
318 mg (81.3%) of yellow solid: 1H NMR (400 MHz, THF-d8,
Infrared Spectroscopy. IR spectra were obtained at room
temperature using a Mattson Galaxy 4030 series FT-IR spectro-
photometer. IR samples were prepared in a drybox by dissolving
5-10 mg of (5/6L)FeII, [(5/6L)FeII‚‚‚CuI](B(C6F5)4), or [(TMPA)-
CuI(CH3CN)](B(C6F5)4) in ∼1 mL solvent in a vial. In samples in
which the exogenous axial ligand base 1,5-dicyclohexylimidazole
(39) Collman, J. P.; Sunderland, C. J.; Boulatov, R. Inorg. Chem. 2002,
41, 2282-2291.
(40) Ghiladi, R. A.; Lee, D.-H.; Huang, H.-w.; Moe¨nne-Loccoz, P.; Kaderli,
S.; Zuberbu¨hler, A. D.; Woods, A. S.; Cotter, R. J.; Karlin, K. D.
Manuscript in preparation.
(41) Liang, H.-C.; Kim, E.; Incarvito, C. D.; Rheingold, A. L.; Karlin, K.
D. Inorg. Chem. 2002, 41, 2209-2212.
(42) Karlin, K. D.; Hayes, J. C.; Shi, J.; Hutchinson, J. P.; Zubieta, J. Inorg.
Chem. 1982, 21, 4106-4108.
(43) Karlin, K. D.; Cruse, R. W.; Gultneh, Y.; Farooq, A.; Hayes, J. C.;
Zubieta, J. J. Am. Chem. Soc. 1987, 109, 2668-2679.
(44) Karlin, K. D.; Haka, M. S.; Cruse, R. W.; Meyer, G. J.; Farooq, A.;
Gultneh, Y.; Hayes, J. C.; Zubieta, J. J. Am. Chem. Soc. 1988, 110,
1196-1207.
3018 Inorganic Chemistry, Vol. 42, No. 9, 2003

Reduced Heme and Heme/Copper Carbonmonoxy Species
Table 1. UV-Vis Data (nm) for Fe/Empty-Tether Compounds at Room Temperature
solvent
(5L)FeII a
(5L)FeII-COa
(6L)FeII b
(6L)FeII-COb
THF (Soret)
424
414
424
413
(2.32 ( 0.14) × 105
(3.2 ( 0.2) × 105
(2.40 ( 0.014) × 105
(3.16 ( 0.16) × 105
R
544
533
543
532
(1.18 ( 0.03) × 104
424
(1.50 ( 0.06) × 104
(1.0 ( 0.2) × 104
426
(1.4 ( 0.08) × 104
acetonitrile (Soret)
415
540
414
534
416
535
416
536
415
539
413
534
419
537
419
538
R
530, 555(sh)
429
535(sh), 557
416, 431
532, 558(sh)
416(sh), 431
536(sh), 557
530, 555(sh)
429
acetone (Soret)
R
536(sh), 556
417(sh), 433
534, 556(sh)
416(sh), 434
538(sh), 557
toluene (Soret)
R
CH2Cl2 (Soret)
R
a Extinction coefficients were determined based on three trials. b Extinction coefficients were determined based on two trials.
(DCHIm) was added, 1 equiv was weighed out in the drybox and
added to the solution prior to carbonylation. The vial was then
sealed with a 14/20 rubber septum prior to removal to the benchtop.
After taking an IR spectrum using a SPECAC solution cell of the
13CO. Chemical shifts are reported as δ values internally referenced
to solvent peaks.
Results and Discussion
(
5/6L)FeII, [(5/6L)FeII‚‚‚CuI](B(C6F5)4), or [(TMPA)CuI(CH3CN)]-
Synthesis and Characterization. The Fe/empty-tether
species, (5/6L)FeII and (6L-d8)FeII (Chart 1), and the Fe/Cu
complexes [(5/6L)FeII‚‚‚CuI](BArF20) and [(6L-d8)FeII‚‚‚CuI]-
(B(C6F5)4), have been synthesized and characterized as
described in the preceding section, or elsewhere.25,40 The
deuterated compounds (5L-d8)FeII and [(5L-d8)FeII‚‚‚CuI]-
(B(C6F5)4) were synthesized in an analogous fashion to that
(B(C6F5)4) sample as a control, the stock solution in the vial was
bubbled with CO for 5-10 s. An IR spectrum was then taken of
the carbonylated species in order to observe the specific (CO)Fe
and/or (CO)Cu stretching frequency.
In an attempt to generate the [(5/6L)FeII-CO‚‚‚CuI](B(C6F5)4)
species, the [(5/6L)FeII‚‚‚CuI](B(C6F5)4) stock solution was first
bubbled with CO and then followed by Ar bubbling for 5-10 s.
After transfer to the solution cell, an IR spectrum was taken as
previously described.
A complex formulated as [(PY1)CuI-CO](B(C6F5)4) {PY1 )
bis(2-picolyl)amine (Richmond Chemicals)} was generated in situ
by dissolving 22 mg PY1 and 100 mg [CuI(CH3CN)4](B(C6F5)4)
in ∼5 mL CH3CN, followed after 45 min of stirring by the addition
of CO.
1H and 2H NMR Spectroscopy. 1H and 2H NMR spectra were
recorded at room temperature on a Varian Unity 400 NMR
spectrometer (1H at 400 MHz, 2H at 61.3 MHz). 1H NMR samples
were prepared in a glovebox by dissolving 20-30 mg of (5/6L)FeII,
[(5/6L)FeII‚‚‚CuI](B(C6F5)4), (F8TPP)FeII, or [(TMPA)CuI(CH3CN)]-
(B(C6F5)4) in 1 mL of deuterated NMR solvent in a 5 mm, airtight,
screw-cap NMR tube (Wilmad Glass). After a spectrum of the
reduced sample was obtained, carbonylation was effected by
bubbling CO through the sample via syringe needle for 5-10 s.
2H NMR spectra were collected using a tunable broadband probe
to enhance 2H detection. Samples were prepared by dissolving 20-
30 mg of (5L-d8)FeII in 1 mL of solvent (undeuterated) in a screw-
cap NMR tube, followed by carbonylation as already described.
2H NMR spectra of [(5L-d8)FeII‚‚‚CuI](B(C6F5)4) were prepared in
situ as previously described. Chemical shifts are reported as δ values
internally referenced to solvent.
6
reported for the deuterated L ligand.40
Five organic solvents were chosen for study of the Fe/
empty-tether species with carbon monoxide. These were
acetonitrile, tetrahydrofuran, acetone, toluene, and methylene
chloride, ranging from strongly coordinating to noncoordi-
nating, in their ability to interact with iron(II).20 These
solvents have been shown to have a marked effect on the
iron(II) spin-state and have also yielded variations in
dioxygen reactivity, in the electronically similar (F8TPP)-
FeII system.20,45 For this reason, their effect on carbonyl
chemistry is also of interest. For the Fe/Cu species, all but
methylene chloride were used due to its reaction with CuI-
(TMPA) species, yielding CuII-Cl.46
UV-Vis Spectroscopy. In general, addition of CO caused
an immediate color change in the dilute samples (Soret band,
with absorption ∼0.5-1.0) from yellow to light tan. In the
more concentrated (∼10×) samples used to measure absorp-
tion bands in the R region, the color changed from a deep
orange/brown to red. Overall, for both the Fe/empty-tether
and Fe/Cu species, carbonylation of the sample resulted in
a blue shift for both the Soret and R bands and loss of any
double-featured characteristics in the Soret or R region.
Addition of CO was marked by a sharpening of the Soret
feature in each case and an increase in intensity in both the
Soret and R absorptions, as determined by measurements
of the extinction coefficients (ꢀ) in THF for (5/6L)FeII and
(5/6L)FeII-CO (Table 1). Such changes upon carbonylation
have been observed in related porphyrin systems.47 In fact,
there is almost no variation in λmax values from solvent to
13C NMR Spectroscopy. 13C NMR spectra were also collected
on a Varian Unity 400 NMR spectrometer at 100 MHz using a
tunable broadband probe to enhance 13C detection. Samples of
(
5/6L)FeII, [(5L)FeII‚‚‚CuI](B(C6F5)4), (F8TPP)FeII, and [(TMPA)-
CuI(CH3CN)](B(C6F5)4) were prepared in an analogous fashion to
1
those already described for the H NMR spectroscopic studies.
However, carbonylation was effected by using a 3-way T-valve
syringe assembly to bubble each sample with 2 × 5 mL isotopically
labeled 13CO stored in 100 mL breakseals (ICON Services, Inc.).
For (5/6L)FeII-CO and (F8TPP)FeII-CO, spectra were taken in
THF-d8 and CD2Cl2 using 13CO 67% enriched in 13C, while spectra
for the [(5/6L)FeII-CO‚‚‚CuI-CO](B(C6F5)4) and [(TMPA)CuI-
CO](B(C6F5)4) samples were taken in THF-d8 using 99% enriched
(45) Ghiladi, R. A.; Karlin, K. D. Inorg. Chem. 2002, 41, 2400-2407.
(46) Jacobson, R. R.; Tyekla´r, Z.; Karlin, K. D. Inorg. Chim. Acta 1991,
181, 111-118.
(47) Traylor, T. G.; Chang, C. K.; Geibel, J.; Berzinis, A.; Mincey, T.;
Cannon, J. J. Am. Chem. Soc. 1979, 101, 6716-6730.
Inorganic Chemistry, Vol. 42, No. 9, 2003 3019

Kretzer et al.
Chart 2
Table 2. ν(CO)Fe (cm-1) IR Data at Room Temperature
[(5L)FeII-
[(6L)FeII-
CO‚‚‚CuI-CO]- CO‚‚‚CuI-CO]-
solvent
THF
(5L)FeII-CO (6L)FeII-CO
(B(C6F5)4)
(B(C6F5)4)
Figure 2. Overlayed UV-vis spectra of (5L)FeII (--) and (5L)FeII-CO
(- -) in THF at room temperature. A more concentrated spectrum of the R
region for each sample is included.
1978
1980
1968
1975
1969
1976
1978
1969
1975
1976
1973
1979
1968
1973
1971
1977
1968
1970
acetonitrile
acetone
toluene
CH2Cl2
solvent (Table 1), and use of either constitutional isomer
6
(5L vs L) did not significantly alter the observed optical
spectra.
‚CuI-CO](B(C6F5)4) samples yielded similar features to their
(5/6L)FeII and (5/6L)FeII-CO counterparts; thus, the presence
(5/6L)FeII Carbonylation. UV-vis spectra of (5/6L)FeII
before carbonylation revealed a marked effect of solvent on
both peak shape and position (Table 1). It is useful to note
that in weakly coordinating solvents, e.g., CH2Cl2, toluene,
and possibly acetone, a pyridine arm from the tether may
coordinate to the heme (see diagram).45
5
6
of copper(I) in the L or L tether seemed to have little or
no effect on electronic absorption spectra in the solvents
tested, since the heme dominates in UV-vis spectra and
copper(I) complexes (d10) generally do not show charge-
transfer or d-d absorptions in the visible region. Two
examples in support of this conclusion follow: In toluene,
[(5L)FeII‚‚‚CuI](B(C6F5)4) exhibited multiple Soret and R
bands as in the Fe/empty tether system,45 with absorptions
at 413, 432(sh) (Soret), and 530, 556(sh) nm. Upon carbo-
nylation, the peak positions changed to 418 (Soret) and 535
nm. Similarly, [(6L)FeII‚‚‚CuI](B(C6F5)4) in toluene produced
415, 435(sh) (Soret), and 530, 556(sh) nm absorptions,
followed by 417 (Soret) and 533 nm peaks after addition of
CO. In THF, [(6L)FeII‚‚‚CuI](B(C6F5)4) showed single-
featured Soret and R bands at 424 and 546 nm, as in
(5/6L)FeII.45 Upon carbonylation, these bands shifted to 413
(Soret) and 532 nm, identical to the absorptions in (6L)FeII-
CO.
IR Spectroscopy. The same solvents used for the UV-
vis spectroscopic studies were employed in order to deter-
mine their relative ability to act as an axial base ligand for
the iron in the presumed (solvent)FeII-CO six-coordinate
complex. As is well-known for binding of transition metals
to carbon monoxide, and predicted for the present Fe-CO
and/or Fe-CO/Cu-CO systems, both ν(CO)Fe and ν(CO)Cu
values were found at lower energy than the 2143 cm-1
expected from free CO (vide infra). Table 2 provides ν(CO)Fe
IR data for (5/6L)FeII-CO and [(5/6L)FeII-CO‚‚‚CuI-CO]-
(B(C6F5)4) in various solvents.
Here we highlight the UV-vis changes for (5L)FeII in THF
{UV-vis: 424 (Soret), 544 nm}. After carbonylation, the
(5L)FeII-CO spectral features shifted to 414 (Soret) and 533
nm (Figure 2); similar spectral changes are observed for the
formation of (6L)FeII-CO (Table 1). The results show clear
UV-vis changes that are analogous to those of other heme
complexes and proteins.48,49 We conclude this addition of
CO causes the species to become six-coordinate, low-spin
(see also NMR spectroscopic evidence, vide infra), with
either a solvent molecule or a pyridine arm occupying the
sixth coordination site on the iron(II) (Chart 2, where B )
solvent axial base). The CO molecule may be on the “tether
side” of the porphryin plane or bound to the open face, as
also suggested for O2-coordination.45
(5/6L)FeII Carbonylation and Effect of Added Base. In
general, ν(CO)Fe values for (5/6L)FeII-CO were found from
1968 to 1980 cm-1, showing a modest dependence on the
solvent used; an infrared spectrum of a typical carbonylation
reaction is given in Figure 3 for (5L)FeII. These stretching
frequencies agree favorably with those observed for the
“parent” reference compound (F8TPP)FeII-CO, with ν(CO)Fe
[(5/6L)FeII‚‚‚CuI](B(C6F5)4) Carbonylation. UV-vis spec-
tra for the [(5/6L)FeII‚‚‚CuI](B(C6F5)4) and [(5/6L)FeII-CO‚‚
(48) Tsuchida, E.; Komatsu, T.; Arai, K.; Nishide, H. J. Chem. Soc., Dalton
Trans. 1993, 2465-2469.
(49) Collman, J. P.; Brauman, J. I.; Doxsee, K. M.; Halbert, T. R.;
Bunnenberg, E.; Linder, R. E.; La Mar, G. N.; Del Gaudio, J.; Lang,
G.; Spartalian, K. J. Am. Chem. Soc. 1980, 102, 4182-4192.
3020 Inorganic Chemistry, Vol. 42, No. 9, 2003

Reduced Heme and Heme/Copper Carbonmonoxy Species
Figure 3. Overlayed IR spectra of (5L)FeII (--) and (5L)FeII-CO (- -) in
acetone at room temperature. A difference spectrum appears as an inset.
Figure 4. Overlayed IR spectra of [(5L)FeII‚‚‚CuI](B(C6F5)4) (--) and
[(5L)FeII-CO‚‚‚CuI-CO](B(C6F5)4) (- -) in acetone at room temperature.
A difference spectrum appears as an inset.
values of 1976 cm-1 (acetone), 1979 cm-1 (THF), and 1982
cm-1 (acetonitrile).50 For other related porphyrin macro-
cycles, including “capped” and “pocket” porphyrins, similar
ν(CO)Fe values are observed.48,51-53 Note the similar values
for the hemoproteins myoglobin (ν(CO)Fe ) 1945 cm-1) and
hemoglobin (ν(CO)Fe ) 1951 cm-1).52
In order to determine the effect of a nonsolvent derived
strongly donating axial base ligand, 1 equiv of 1,5-dicyclo-
hexylimidazole was added to a sample of (5L)FeII in toluene
prior to carbonylation. A ν(CO)Fe absorption at 1984 cm-1
was observed, representing a shift of 9 cm-1 from the 1975
cm-1 stretch found for (5L)FeII-CO in toluene. Note that
among the coordinating solvents (THF, MeCN, and acetone),
both THF and MeCN provide (5/6L)FeII-CO complexes with
the greatest ν(CO)Fe stretching frequencies (Table 2), con-
sistent with our previously20,45 derived notions of their relative
strength of coordination to the heme (i.e., THF > MeCN >
acetone).
Table 3. ν(CO)Cu (cm-1) IR Data at Room Temperature
[(5L)FeII-
[(6L)FeII-
[(TMPA)-
CO‚‚‚ CuI-CO]- CO‚‚‚CuI-CO]- CuI-CO]- [(PY1)CuI-CO]-
solvent
THF
acetonitrile
acetone
(B(C6F5)4)
(B(C6F5)4)
(B(C6F5)4)
(B(C6F5)4)
2091
2094
2092
2093
2093
2092
2093
2093
2091
2092
2092
2091
2091
toluene
Chart 3
[(5/6L)FeII‚‚‚CuI](B(C6F5)4) Carbonylation. For the [(5/6L)-
FeII-CO‚‚‚CuI-CO](B(C6F5)4) species, ν(CO)Fe values analo-
gous to those in empty-tether (5/6L)FeII-CO species were
observed, ranging from 1968 to 1979 cm-1 (Table 2 and
Figure 4).
IR spectra of the CuI-CO moiety in [(5/6L)FeII-CO‚‚‚
CuI-CO](B(C6F5)4) complexes, obtained by directly bub-
bling carbon monoxide through solutions of the reduced
samples under an inert atmosphere, exhibited a strong
absorption in the range 2091-2094 cm-1 (Table 3 and Figure
4). These values are typical for copper complexes although
somewhat higher in energy than for Cu-CO moieties in
metalloproteins.54,55
The CuI-CO infrared region for [(5/6L)FeII-CO‚‚‚CuI-
CO](B(C6F5)4) also exhibited a weaker shoulder at lower
energy, ∼2070 cm-1 (see Figure 4 inset), which varied in
intensity with the solvent used. However, in toluene the
shoulder was more pronounced, appearing as a separate peak
at 2068 cm-1 in [(5L)FeII-CO‚‚‚CuI-CO](B(C6F5)4) and at
2037 cm-1 in [(6L)FeII-CO‚‚‚CuI-CO](B(C6F5)4). In order
to gain insight into the origin of the ν(CO)Cu absorptions,
including this weak but persistent shoulder, control experi-
ments were performed utilizing [(TMPA)CuI(CH3CN)]-
(B(C6F5)4) and [(PY1)CuI(CH3CN)](B(C6F5)4) (Chart 3).
Upon addition of CO to each compound in solution, a color
change was seen from yellow to colorless. IR experiments
using [(TMPA)CuI-CO](B(C6F5)4) yielded spectra with
ν(CO)Cu values from 2091 to 2092 cm-1, but also exhibit-
ing a shoulder in the 2068-2071 cm-1 range that varied
in intensity with solvent. Thus, the Cu moiety in the
[(5/6L)FeII‚‚‚CuI]+ complexes binds CO in a manner that
appears to be independent of the heme.
(50) Thompson, D. W.; Kretzer, R. M.; Lebeau, E. L.; Scaltrio, D.; Ghiladi,
R. A.; Lam, K.-C.; Rheingold, A. L.; Meyer, G. J.; Karlin, K. D.
Manuscript in preparation.
(51) Linard, J. E.; Ellis, P. E., Jr.; Budge, J. R.; Jones, R. D.; Basolo, F. J.
Am. Chem. Soc. 1980, 102, 1896-1904.
(52) Collman, J. P.; Brauman, J. I.; Collins, T. J.; Iverson, B. L.; Lang,
G.; Pettman, R. B.; Sessler, J. L.; Walters, M. A. J. Am. Chem. Soc.
1983, 105, 3038-3052.
(53) Collman, J. P.; Reed, C. A. J. Am. Chem. Soc. 1973, 95, 2048-2049.
(54) Rondelez, Y.; Se´neque, O.; Rager, M.-N.; Duprat, A. F.; Reinaud, O.
Chem. Eur. J. 2000, 6, 4218-4226.
(55) Hirota, S.; Iwamoto, T.; Tanizawa, K.; Aduchi, O.; Yamauchi, O.
Biochemistry 1999, 38, 14256-14263.
Inorganic Chemistry, Vol. 42, No. 9, 2003 3021

Kretzer et al.
Scheme 1
Figure 5. Overlayed IR difference spectra of [(PY1)CuI-CO](B(C6F5)4)
- [(PY1)CuI(CH3CN)](B(C6F5)4) (--) and [(TMPA)CuI-CO](B(C6F5)4)
- [(TMPA)CuI(CH3CN)](B(C6F5)4) (- -) in CH3CN at room temperature.
The spectra have been normalized to account for differences in absorbance
due to variations in concentrations.
An experiment with [(PY1)CuI(CH3CN)](B(C6F5)4) (Chart
3) bubbled with CO in CH3CN gave more insight into the
likely origin of the IR shoulder. [(PY1)CuI(CH3CN)]-
(B(C6F5)4) differs from [(TMPA)CuI(CH3CN)](B(C6F5)4) in
that only two pyridine moieties are present to bind Cu, as
opposed to three in TMPA. Upon addition of CO to the
[(PY1)CuI(CH3CN)](B(C6F5)4) species in CH3CN (see Ex-
perimental Section), an IR stretch was observed at 2091
cm-1, identical to the major peak seen at 2092 cm-1 for
[(TMPA)CuI-CO](B(C6F5)4) in CH3CN. A difference spec-
trum was generated in which the IR spectrum of [(PY1)-
CuI(CH3CN)](B(C6F5)4) was subtracted from that of [(PY1)-
CuI-CO](B(C6F5)4), showing only the 2091 cm-1 stretch
with no shoulder at ∼2070 cm-1 (Figure 5).
carbonyl species are possible for the [(5/6L)FeII-CO‚‚‚CuI-
CO](B(C6F5)4) complexes (Scheme 1). With respect to CuI-
CO adduct formation, the relative intensities of the two
(CO)Cu IR stretches point to the “major” product of the
carbonylation reaction containing a “dangling” pyridine arm
which is not coordinated to the copper(I). By analogy to the
geometry of [(PY1)CuI-CO](B(C6F5)4), this accounts for the
IR stretch at 2091-2094 cm-1 (Table 3). The noncoordinated
pyridyl moiety is assumed to be that which is directly
tethered to the heme via an ether linkage, as the previously
described crystal structure of [(6L)FeIII-O-CuII](B(C6F5)4)
showed this Cu-N bond to be elongated by ∼0.4 Å relative
to the other two.25 This allows the tethered (TMPA)CuI to
rotate away from the porphyrin ring plane. For this (TMPA)-
CuI conformation, two possible isomers exist for CO binding
to the ferrous heme, one in which the CO is on the same
side as the tethered (TMPA)CuI, and one in which it is on
the opposite side. In order to account for the (CO)Cu IR
stretch at ∼2070 cm-1, it is likely that the “minor” product
(Scheme 1) involves weak coordination of the third pyridyl
arm to the copper(I), which could potentially localize the
(TMPA)CuI over the heme to form a sterically hindered
binding pocket. However, due to free rotation about the either
linkage of the tether, it is also possible that the TMPA(CuI)
rotates from 0° to 180° relative to the porphyrin. This
conformation in which the copper(I) has three coordinated
pyridyl arms also has two possible isomers for heme-CO
These observations provide good evidence that the pre-
dominate species in solutions contains a single uncoordinated
pyridyl arm of the TMPA ligand, both for [(TMPA)CuI-
CO](B(C6F5)4) and [(5/6L)FeII-CO‚‚‚CuI-CO](B(C6F5)4).
We therefore propose that the shoulder at ∼2070 cm-1 could
then correspond to a species in which all three pyridyl arms
of the TMPA moiety are coordinated to the copper(I) ion
complex (Chart 3). This hypothesis seems logical as the
coordination of three pyridyl arms to the copper(I) would
increase the electron density around this metal, thereby
decreasing the strength, and in turn the stretching frequency,
of the C-O bond. This assumption seems to be further
supported by the presence of an even more prominent peak
at ∼2070 cm-1 in toluene upon CO addition, as toluene is a
noncoordinating solvent, making it more likely that all three
pyridyl donors coordinate in the CO-adduct. Furthermore,
previous experiments in which PPh3 was added to [(TMPA)-
CuI](PF6) have shown that a pyridyl arm can “dangle”,
leaving tridentate coordination from the potentially tetraden-
tate TMPA chelated.56
Possible Solution-State Structures of [(5/6L)FeII-
CO‚‚‚CuI-CO](B(C6F5)4). On the basis of the IR experi-
ments comparing the CO binding of [(PY1)CuI-CO]-
(B(C6F5)4) and [(TMPA)CuI-CO](B(C6F5)4), four isomeric
(56) Tyekla´r, Z.; Jacobson, R. R.; Wei, N.; Murthy, N. N.; Zubieta, J.;
Karlin, K. D. J. Am. Chem. Soc. 1993, 115, 2677-2689.
3022 Inorganic Chemistry, Vol. 42, No. 9, 2003

Reduced Heme and Heme/Copper Carbonmonoxy Species
Scheme 2
binding, as shown. In all possible geometric isomers, the
sixth ligand-coordination site can be occupied by a solvent
molecule such as THF (B ) axial ligand base, Scheme 1).
Further evidence for these formulations comes from the
observed invariability of the ν(CO)Fe in both the Fe/empty-
tether and the Fe/Cu systems. If the (TMPA)CuI moiety was
primarily localized in close proximity to the heme, forming
an Fe/Cu binding pocket, then one might expect a decrease
in ν(CO)Fe in the [(5/6L)FeII-CO‚‚‚CuI-CO](B(C6F5)4) spe-
cies relative to (5/6L)FeII-CO due to interaction of the
positively charged copper(I) with (CO)Fe. Collman and co-
workers39 have elaborated on this phenomenon, noting that
(i) such an interaction would serve to stabilize the Fe-dπ-
to-CO-π* back-bonding interaction, thereby weakening the
CO bond, (ii) citing biological precendence, where in mutants
of cytochrome bo ubiquinol oxidase (from E. coli), and other
terminal oxidases in which the ability to bind copper(I) was
removed, there is an increase in ν(CO)Fe relative to the wild
type,57-59 and (iii) in general, Fe-C-O prefers a linear
geometry in order to maximize the back-bonding stabiliza-
tion, and steric effects that alter this geometry from linearity
have been associated with changes in ν(CO)Fe values.3,60-65
However, no significant difference in ν(CO)Fe was observed
in comparing [(5/6L)FeII-CO‚‚‚CuI-CO](B(C6F5)4) and (5/6L)-
FeII-CO species; thus, it is likely that the copper(I) ion in
these Fe/Cu systems is not close enough to significantly
interact with the Fe-CO unit.
Figure 6. Overlayed IR difference spectra of ([(5L)FeII-CO‚‚‚CuI-CO]+
+ DCHIm) - ([(5L)FeII‚‚‚CuI]+ + DCHIm) (--) and ([(5L)FeII-CO‚‚‚
CuI]+ + DCHIm) - ([(5L)FeII‚‚‚CuI]+ + DCHIm) (- -) in THF at room
temperature. DCHIm (1 equiv) was added to the solution prior to CO
bubbling, and the [(5L)FeII-CO‚‚‚CuI]+ species was generated by bubbling
with Ar.
Table 4. 1H NMR Data (δ, ppm) for Fe/Empty-Tether Compounds at
Room Temperature
solvent
THF-d8
(5L)FeII
(5L)FeII-CO
(6L)FeII
(6L)FeII-CO
55.4, 55.7,
56.9, 57.5
8.8
52.9, 54.2,
57.8. 58.6
41.1, 42.8,
43.6, 44.5
48.2, 50.1,
51.8, 52.5
31.8, 32.8,
34.5
8.8
acetonitrile-d3 30.0, 30.9,
8.8
8.8
8.7
8.7
8.8
8.8
8.8
8.9
31.2, 31.6
47.3, 48.6,
49.8, 50.4
acetone-d6
toluene-d8
CD2Cl2
31-37 (v br)
43.6 (br)
44.2, 45.7,
47.3, 48.1
tively displace the CO bound to the copper(I) and generate
the [(5/6L)FeII-CO‚‚‚CuI](B(C6F5)4) species (Scheme 2).
This resulted in the nearly complete loss of CO from the
Cu(I), as proven by a disappearance of the stretch at ∼2090
cm-1 in IR spectra. Loss of intensity of the 1970 cm-1 ν-
(CO)Fe band also occurred, but the extent was solvent
dependent, following the general trend acetone (∼60% loss)
> MeCN (∼35% loss) > THF > THF with added 1,5-
dicyclohexylimidazole (DCHIm, 1 equiv; ∼40% loss, Figure
6). Thus, these experiments are consistent with solvent or
axial base ligand (DCHIm) binding to the heme and
stabilizing the Fe-CO interaction. We previously observed
a very similar solvent coordination trend for just the heme-
Fe(II).20,45
Selective Binding to Fe; Generation of Base-FeII-
CO‚‚‚CuI. Using the same stock solution of [(5/6L)FeII‚‚‚CuI]-
(B(C6F5)4) that was used to generate the bis-carbonyl (i.e.,
Fe-CO/Cu-CO) species, samples first bubbled with CO
were now followed by Ar for 5-10 s, in an effort to selec-
(57) Calhoun, M. W.; Hill, J. J.; Lemieux, L. J.; Ingledew, W. J.; Alben,
J. O.; Gennis, R. B. Biochemistry 1993, 32, 11524-11529.
(58) Uno, T.; Mogi, T.; Tsubaki, M.; Nishimura, Y.; Anraku, Y. J. Biol.
Chem. 1994, 269, 11912-11920.
(59) Hosler, J. P.; Kim, Y.; Shapleigh, J.; Gennis, R. B.; Alben, J. O.;
Ferguson-Miller, S.; Babcock, G. T. J. Am. Chem. Soc. 1994, 116,
5515-5516.
(60) Park, K. D.; Guo, K.; Adebodun, F.; Chiu, M. L.; Sligar, S. G.;
Oldfield, E. Biochemistry 1991, 30,
1H and H NMR Spectroscopy. (5/6L)FeII, (5L-d8)FeII,
2
1
and CO-Adducts. H NMR spectra of reduced (5/6L)FeII
prior to carbonylation yielded from one to four distinct
pyrrole resonances in the (δ) 25-60 ppm range depending
on the solvent used, as previously published.20,45 Coordinating
solvents, such as THF-d8, acetonitrile-d3, and acetone-d6,
yielded four well-defined pyrrole peaks for both the (5/6L)FeII
species (Table 4). However, for (5L)FeII in toluene-d8 and
dichloromethane-d2, the downfield shifted pyrrole signature
was broad and poorly defined. This was not the case for the
(6L)FeII compound. For acetonitrile-d3, the pyrrole resonance
positions are indicative of a thermal equilibrium between
five-coordinate (mono-MeCN) high-spin and six-coordinate
(61) Ray, G. B.; Li, X.-Y.; Ibers, J. A.; Sessler, J. L.; Spiro, T. G. J. Am.
Chem. Soc. 1994, 116, 162-176.
(62) Tetreau, C.; Lavalette, D.; Momenteau, M.; Fischer, J.; Weiss, R. J.
Am. Chem. Soc. 1994, 116, 11840-11848.
(63) Slebodnick, C.; Duval, M. L.; Ibers, J. A. Inorg. Chem. 1996, 35,
3607-3613.
(64) Vogel, K. M.; Kozlowski, P. M.; Zgierski, M. Z.; Spiro, T. G. Inorg.
Chim. Acta 2000, 297, 11-17.
(65) Kalodimos, C. G.; Gerothanassis, I. P.; Pierattelli, R.; Troganis, A. J.
Inorg. Biochem. 2000, 79, 371-380.
Inorganic Chemistry, Vol. 42, No. 9, 2003 3023

Kretzer et al.
Table 5. 1H NMR Data (δ, ppm) for Fe/Cu Compounds at Room Temperature
[(5L)FeII‚‚‚CuI]-
[(5L)FeII-CO‚‚‚CuI-CO]-
[(6L)FeII‚‚‚CuI]-
[(6L)FeII-CO‚‚‚CuI-CO]-
solvent
THF-d8
acetonitrile-d3
acetone-d6
(B(C6F5)4)
(B(C6F5)4)
(B(C6F5)4)
(B(C6F5)4)
54.0, 54.7, 57.7, 58.8
27.6 (sh), 28.7, 29.2
8.74, 8.82
8.77
54.2, 56.7
28.8 (sh), 29.8
45.9, 46.9, 48.2, 48.7
8.75, 8.81, 8.86
8.74, 8.78, 8.81
8.87
Figure 8. 2H NMR spectra of (a) [(5L-d8)FeII‚‚‚CuI](B(C6F5)4) and (b)
[(5L-d8)FeII-CO‚‚‚CuI-CO](B(C6F5)4) in THF at room temperature. THF
solvent peaks appear at 1.73 and 3.59 ppm in each spectrum.
Figure 7. 2H NMR spectra of (a) (5L-d8)FeII and (b) (5L-d8)FeII-CO in
THF at room temperature. THF solvent peaks appear at 1.73 and 3.58 ppm
in each spectrum.
(bis-MeCN) low-spin ferrous hemes, whereas in noncoor-
dinating solvents (toluene-d8, dichloromethane-d2), a pyridyl
group from the TMPA tether ligates the heme, vide supra.20,45
Carbonylation in all cases yields an upfield shift in the
pyrrole resonances to the diamagnetic region (8.7-8.9 ppm),
analogous to what occurs for other five-coordinate, high-
spin iron(II) heme compounds when CO or O2 is added
(Table 4).66
Figure 9. 13C NMR spectra of (5L)FeII
+
13CO (67% enriched in 13C) in
THF-d8 at room temperature. The inset shows two resonance peaks at 206.8
and 207.0 ppm, both of which are assigned to (13CO)Fe. The two very intense
upfield peaks (67 and 25 ppm) are due to the 13C (at natural abundance) of
the THF solvent, while the strong peak near 0 ppm is due to grease.
peak at the correct chemical shift but lacking the finer
Further evidence in support of the pyrrole resonance
1
resolution found in H NMR spectroscopy.67,68 Thus, the
2
assignments was aided by collection of H NMR spectra
results of the 1,2H NMR spectroscopic studies indicate that
the presence of the CuI-CO adduct in [(5/6L)FeII-CO‚‚‚
CuI-CO](B(C6F5)4) complexes does not significantly
affect the spin-state or coordination of the heme-Fe-CO
species.
obtained in THF for (5L-d8)FeII and (5L-d8)FeII-CO. This
experiment yielded a group of four pyrrole peaks at 53.3,
54.5, 55.8, and 56.4 ppm for (5L-d8)FeII, and a single pyrrole
resonance at 8.9 ppm for (5L-d8)FeII-CO (Figure 7).
[(5/6L)FeII‚‚‚CuI](B(C6F5)4), [(5L-d8)FeII‚‚‚CuI](B(C6F5)4),
13C NMR Spectroscopy of 13CO-Adducts. In light of the
interest in 13C NMR spectroscopic data for (13CO)-adducts
of heme proteins and model complexes,52,60,65,69 we wished
to obtain corresponding data not only for our heme and
heme-Cu compounds but also if possible for CuI-13CO
species. We are unaware of any 13C NMR spectral data in
the literature pertaining to a copper(I)-bound carbonyl group.
(5/6L)FeII Carbonylation. For the (5L)FeII-CO sample in
THF-d8, two peaks were found that correspond to (13CO)Fe.
1
and CO-Adducts. H NMR spectra for [(5/6L)FeII‚‚‚CuI]-
(B(C6F5)4) were very similar to those found for the Fe/empty-
tether compounds. However, in some cases it was possible
to resolve more than one pyrrole peak after addition of CO
in the 8.7-8.9 ppm range (Table 5).
2H NMR spectra were also obtained in THF for [(5L-d8)-
FeII‚‚‚CuI](B(C6F5)4) and [(5L-d8)FeII-CO‚‚‚CuI-CO]-
(B(C6F5)4). For [(5L-d8)FeII‚‚‚CuI](B(C6F5)4), one broad
pyrrole signature was found at 53.4 ppm, which shifted to
8.9 ppm upon addition of CO to yield [(5L-d8)FeII-CO‚‚‚
(67) Walker, F. A. In The Porphyrin Handbook; Kadish, K. M., Smith K.
M., Guilard, R., Eds.; Academic Press: San Diego, CA, 2000; Vol. 5
(NMR and EPR), pp 81-184.
2
CuI-CO](B(C6F5)4) (Figure 8). H NMR of paramagnetic
species, due to the quadrapolar relaxation of deuterons, often
causes deuterium resonances to coalesce into a single, broad
(68) Walker, F. A.; Simonis, U. In Biological Magnetic Resonance; Berliner,
L. J., Reuben, J., Eds.; Plenum Press: New York, 1993; Vol. 12 (NMR
of Paramagnetic Molecules), pp 133-274.
(69) Kalodimos, C. G.; Gerothanassis, I. P.; Pierattelli, R.; Ancian, B. Inorg.
Chem. 1999, 38, 4283-4293.
(66) Collman, J. P.; Fu, L. Acc. Chem. Res. 1999, 32, 455-463.
3024 Inorganic Chemistry, Vol. 42, No. 9, 2003

Reduced Heme and Heme/Copper Carbonmonoxy Species
Table 6. 13C NMR Data (δ, ppm) for (13CO)Fe at Room Temperaturea
[(5L)FeII-CO‚‚‚
[(6L)FeII-CO‚‚‚
solvent
(F8TPP)FeII-CO
(5L)FeII-CO
(6L)FeII-CO
CuI-CO]-(B(C6F5)4)
CuI-CO]-(B(C6F5)4)
THF-d8
CD2Cl2
206.5
206.1
206.8 (minor), 207.0
204.5
207.0, 207.1 (∼equal)
206.8 (minor), 207.4
207.0 (minor), 208.1
202.8, 203.6 (minor)
a Characterization of [(5/6L)FeII-CO‚‚‚CuI]+ by 13C NMR was not attempted.
Table 7. 13C NMR Data (δ, ppm) for (13CO)Cu at Room Temperature
in THF
[(5L)FeII-CO‚‚‚
CuI-CO]
(B(C6F5)4)
[(6L)FeII-CO‚‚‚
CuI-CO]
(B(C6F5)4)
[(TMPA)CuI-CO]
(B(C6F5)4)
172.4
178.2
180.3
at 172.4 ppm that was not seen in the (5L)FeII-CO spectrum.
This peak was significantly stronger in intensity than the
numerous (unlabeled) ligand peaks (85-160 ppm), and so,
it was attributed to a CuI-13CO resonance (Figure 10). This
was confirmed by recording a 13C NMR spectrum of
[(TMPA)CuI(CH3CN)](B(C6F5)4) in THF-d8 after bubbling
with 13CO, which yielded a single, downfield shifted peak
at 180.3 ppm for (13CO)Cu, similar in shape and position to
that found in the Fe/Cu systems (e.g., Figure 10). See Table
7 for a list of (13CO)Cu 13C NMR resonances.
Figure 10. 13C NMR spectra of [(5L)FeII‚‚‚CuI](B(C6F5)4) + 13CO (99%
enriched in 13C) in THF-d8 at room temperature. The inset shows two
resonance peaks at 206.8 and 207.4 ppm, which are assigned to (13CO)Fe,
and a peak at 172.4 ppm, which is assigned to (13CO)Cu. The two very
intense upfield peaks (67 and 25 ppm) are due to the 13C (at natural
abundance) of the THF solvent.
Conclusions
A predominant resonance was resolved at 207.0 ppm,
accompanied by a minor peak at 206.8 ppm (Figure 9). A
similar spectrum was obtained for (6L)FeII-CO in THF-d8,
with the two (13CO)Fe peaks falling at 207.0 and 207.1 ppm.
However, these appear with approximately equal intensity.
The appearance of two resonances that correspond to FeII-
13CO can be attributed to two different, isomeric carbonyl
species in the (5/6L)FeII systems. We therefore assign these
two peaks to isomers in which the CO is heme-bound on
either the same or opposite side of the tethered TMPA, with
the sixth ligand coordination site thus occupied by a THF
solvent molecule or TMPA pyridyl group, respectively, as
discussed previously (Chart 2). Differences in (13CO)Fe peak
ratios for the (5L)FeII-CO and (6L)FeII-CO samples might
reflect the differences in heme accessibility of a solvent
molecule from the same side as the TMPA group, caused
by linkage at the 5-position (5L) versus the 6-position (6L).
Table 6 lists (13CO)Fe 13C NMR resonances.
(5/6L)FeII carbonylation was repeated in methylene chloride-
d2 in hopes that a noncoordinating solvent would force the
CO to bind below the porphyrin ring due to pyridine arm
coordination above it (Chart 2). This appeared to be the case
for (5L)FeII in which a single, broad peak was found at 204.5
ppm. Although the (6L)FeII sample yielded a similar broad
peak at 202.8 ppm, it was accompanied by a smaller, sharp
resonance at 203.6 ppm.
Reduced (5/6L)FeII and [(5/6L)FeII‚‚‚CuI](B(C6F5)4) species
have been isolated, characterized, and shown to have
reactivity toward CO. Upon carbonylation, spectral changes
have been observed in UV-vis, IR, and 1,2H NMR spec-
troscopies for the (5/6L)FeII-CO and [(5/6L)FeII-CO‚‚‚CuI-
CO](B(C6F5)4) species. Unique 13C NMR spectra have also
been obtained for the carbonylated compounds. Results for
these tethered systems have shown close agreement to those
obtained for the untethered, carbonylated parent compounds
(F8TPP)FeII-CO and [(TMPA)CuI-CO](B(C6F5)4) as well.
In addition, IR studies have shown that it is possible to
displace CO from the Cu of the tethered Fe/Cu compounds
by addition/bubbling with Ar, yielding compounds in solution
of the form [(5/6L)FeII-CO‚‚‚CuI](B(C6F5)4). These mono-
carbonylated compounds are of special interest for future
photochemical experimentation. Laser-induced flash pho-
tolysis of the heme-CO may cause intermediate binding of
the photoejected CO to the copper(I), with subsequent return
to the iron(II) upon relaxation. In this way, the kinetics of
CO binding to the iron(II) and the copper(I) may be studied,
by analogy to investigations with cytochrome c oxidase.
Preliminary experiments with the (5/6L)FeII-CO compounds
are currently underway.
Acknowledgment. We thank Dr. Lisa A. Marzilli and
Prof. Robert J. Cotter (The Johns Hopkins University School
of Medicine) for their assistance in obtaining ESI-MS data.
We are grateful to the National Institutes of Health (K.D.K.,
Grant GM28962) for support of this research. Ryan M.
Kretzer also acknowledges a Howard Hughes Summer
Fellowship through the Johns Hopkins University.
[(5/6L)FeII‚‚‚CuI](B(C6F5)4) Carbonylation. 13C NMR
spectra were obtained for the Fe/Cu systems in THF-d8 by
bubbling [(5/6L)FeII‚‚‚CuI](B(C6F5)4) with isotopically labeled
13CO. This yielded two (13CO)Fe peaks at 206.8 and 207.4
ppm for [(5L)FeII-CO‚‚‚CuI-CO](B(C6F5)4), in nearly iden-
tical ratios to those found in the (5L)FeII-CO sample in THF-
d8. However, a new and comparatively broad peak emerged
IC020521I
Inorganic Chemistry, Vol. 42, No. 9, 2003 3025