4666 J. Am. Chem. Soc., Vol. 122, No. 19, 2000
EVans and Reed
Br- < OH- ∼ Cl-. Some variation of the OH- field strength
might be expected as a function of H-bonding, likely to be
present in both Fe(OH)(tetramethylchiroporphyrin)36 and Fe-
(OH)(TPP).
More generally, the prevalence of σ-bonding ligands and high-
spin states in metalloenzymes and metallobiomolecules means
that the magnetochemical series may become preferred over the
spectrochemical series for correlating ligand field strengths
throughout bioinorganic chemistry.
The origin of the H2O < OH- field strength reversal must
lie in the different electronic structures of the reporter metals.
The spectrochemical series is derived primarily from the
electronic spectroscopy of low-spin octahedral d6 complexes.
Experimental Section
All manipulations of air-sensitive materials were carried out in a
Vacuum Atmospheres drybox (O2, H2O < 1 ppm) or on a dual manifold
vacuum line using Schlenk techniques and flame-dried glassware. Arene
solvents and hexanes were dried by distillation from Na/benzophenone
outside the drybox and again from Na/K alloy inside the drybox
immediately prior to use. NMR solvents were dried over molecular
sieves. NMR spectra were recorded on Bruker Aspect (200 and 360
6
The t2g configuration fills the dπ orbitals making their energy
levels particularly sensitive to π effects from the ligands. On
the other hand, high- and intermediate-spin complexes of iron-
(III) have only partially filled dπ orbitals. More importantly,
2
the dz orbital, whose σ-antibonding lobe lies along the axial
1
ligand direction in FeX(TPP) complexes,21 is occupied in all
high- and intermediate-spin derivatives. The consequence is long
Fe-X axial bonds and a significant attenuation of π-bonding
opportunities. As a result, the ligand field effects become more
a reflection of σ bonding and electrostatic effects. This explana-
tion readily rationalizes the weaker ligand field of H2O (Fe-O
) 2.04 Å) relative to OH- (Fe-O estimated ∼ 1.84 Å based
on the X-ray structure of Fe(OCH3)(mesoP)40). That all oxya-
nions, even perchlorate, have stronger ligand fields than H2O
toward Fe(TPP)+ can be rationalized in terms of charge. Indeed,
the portion of the magnetochemical series based on high- and
intermediate-spin states can be viewed as a field strength ranking
MHz) or Bruker AMX (500 MHz) systems. H NMR spectra were
referenced internally to TMS or to residual proto-solvent. 11B NMR
spectra were referenced externally to 1.0 M BBr3 in hexanes (40 ppm)
and 19F to external C6F6 (-162 ppm) both as inserts in a tube containing
acetone-d6. Mo¨ssbauer spectra were recorded using instrumentation
assembled by Tom Kent of Web Research. Samples (10-50 mg) were
embedded in melted paraffin wax and referenced to iron foil at room
temperature. SQUID magnetic susceptibility data were recorded at 2
and 10 kG on ground, packed samples on Quantum Design instrumen-
tation at the California Institute of Technology. Diamagnetic corrections
were applied using values of -600 × 10-6 cgs units for TPP, and Pascal
constants, for the remainder of the atoms. Fe(TPP)44 and [H(mesitylene)]-
[F20-BPh4]29 were prepared using published procedures. Water titrations
were done using microsyringe techniques in the drybox.
more closely reflecting the expectations of crystal field theory
Fe(TPP)(CB11H6X6), 1. AgCB11H6Br6 (0.4221 g, 0.583 mmol) was
added to a Schlenk flask, A, equipped with a stir bar and placed on a
double-manifold vacuum line where it was thoroughly dried by
evacuation/ argon refill cycling. A separate Schlenk flask, B, equipped
with a stir bar was charged with vacuum-dried tris-p-bromophenyl-
amine, (BrC6H4)3N (0.2827 g; 0.587 mmol), and iodine (0.0783 g;
0.3085 mmol). The flask was flushed with argon, cooled to liquid-N2
temperature, evacuated to 100 mTorr, removed from the N2-bath, and
back-filled with argon. This was repeated five times, taking care to
prevent loss of iodine. Dichloromethane (ca. 25 mL) was vacuum
transferred into this flask and stirred until dissolution of the solids was
achieved. The contents of B were then transferred to A using a Schlenk
transfer tube, and the resultant solution in flask A was stirred vigorously
for 15 min. Flask A was then transferred to the drybox where the AgI
precipitate was removed by filtration through both medium- and fine-
porosity frits. The solvent was removed under reduced pressure. The
navy blue microcrystalline material was repeatedly washed with hexanes
to remove excess iodine to give [(BrC6H4)3N][CB11H6Br6] (0.429 g,
67%). The corresponding hexachlorocarborane compound was made
in a simliar manner. Fe(TPP) (0.1 g, 0.149 mmol) was dissolved in
benzene (25 mL) at gentle reflux, and [(BrC6H4)3N][CB11H6X6] (25
mL of a 6 mM solution in o-dichlorobenzene) was added gradually.
After 30 min of gentle reflux, the solvent volume was reduced by half
and hexanes (150 mL) were added. The brown microcrystalline product
was collected on a fine frit, washed with hexanes, and dried under
vacuum (138 mg, 91%). Anal. Calcd (X ) Cl) for C45H34B11N4Cl6Fe:
C, 53.08; H, 3.37; N, 5.50. Found: C, 52.36; H, 3.31; N, 5.15.
Mo¨ssbauer (X ) Cl) at 25 °C: δ ) 0.33, ∆Eq ) 3.68 mm/s (Figure
S5). µeff (X ) Cl) ) 4.0 µB at 25°C (Figure S6). 1H NMR (X ) Br) 20
mM in benzene-d6 at 25 °C: -62 (8H, â-pyrrole, s, br), 2.1 (1H,
carborane, s, br), 3.4 (5H, carborane, s, br), 8.1 (4H, para, s, br), 9.1
(8H, meta, s, br), and 10.4 (8H, ortho, s, br) (Figures 1A, S9A).
[Fe(H2O)(TPP)][CB11H6X6], 2. A 1-equiv amount of water was
added to a refluxing solution of 1 in benzene. After 30 min, a 10-fold
addition of hexanes was used to precipitate the brown microcrystalline
product, which was collected on a fine frit and allowed to dry (85%).
Anal. Calcd (X ) Cl) for C45H36B11N4OCl6Fe: C, 52.16; H, 3.50; N,
-
than ligand field theory. That portion is Ag(CB11H12)2
<
CB11H12- < SbF6- < Co(C2B9H11)2- < AsF6- < H2O < ClO4
-
-
-
-
-
-
< C(CN)3 < CF3SO3 < BF4 < ONC(CN)2 < ReO4
<
OTeF5 < I- < Br- < OH- ∼ Cl- < NCS- < OAc- ∼ N3
< OPh(p-NO2)- < OPh- < F- < RS-. The order is largely a
reflection of increasing charge proximity to the metal and/or
increasing covalency in the metal-ligand σ bond. The effects
of π bonding appear to be rather minimal. An abbreviated
-
-
-
-
-
-
ordering such as SbF6 < H2O < ClO4 < CF3SO3 < BF4
< ReO4- < I- < Br- < OH- ∼ Cl- < OAc- < OPh(p-NO2)-
< OPh- < F- < RS- might usefully accompany the introduc-
tion of crystal field theory that typically precedes ligand field
theory in textbooks of inorganic chemistry.
Minor reversals of spectrochemical ligand field rankings are
not uncommon in going from one metal to another. A germane
example is the reversal of H2O and OH- observed some time
ago in Co(CN)5(H2O)2- (λmax ) 380 nm) and Co(CN)5(OH)3-
(λmax ) 375 nm).41 Unlike a typical reporter complex such as
the pentammine [Co(NH3)5X]n+, the CoIII(CN)5Xn- complex has
very strong π back-bonding ligands. Possibly the pentacyano
moiety renders the sixth coordination site more sensitive σ
bonding effects. The pentacyanocobaltate system is an interest-
ing reporter of ligand field strengths because dichloromethane
is not a much weaker field ligand than chloride; the λmax
difference is only 13 nm.42
It is also interesting to note that the H2O < OH- ordering of
ligand field strengths has long been implied by the magnetic
data on aqua-methemoglobin. At high pH, high-spin met(H2O)-
Hb is in equilibrium with a low-spin species logically assigned
to met(OH)Hb.43 Yet, to our knowledge, this inconsistency with
the spectrochemical series has never been discussed in the
chemical literature. Its explanation is now evident.
(40) Hoard, J. L.; Hamor, M. J.; Hamor, T. A.; Caughey, W. S. J. Am.
Chem. Soc. 1965, 87, 2312-2319.
5.41. Found: C, 50.96; H, 3.47; N, 4.98. IR (X ) Br): 3427 w (νO-H
)
cm-1. Mo¨ssbauer (X ) Cl) at 25 °C: δ ) 0.30, ∆Eq ) 3.24 mm‚s-1
(Figure S7). µeff (X ) Cl) ) 4.2 µB at 25 °C (Figure S8). 1H NMR (X
) Br) 20 mM in benzene-d6 at 25 °C: -43 (8H, â-pyrrole, s, br), 7.9
(4H, para, s, br), 9.5 (8H, meta, s, br), and 10.2 (8H, ortho, s, br)
(Figures 1B, S9B, S10).
(41) Wrighton, M.; Bredesen, D. Inorg. Chem. 1973, 12, 1707-1709.
(42) Milder, S. J.; Gray, H. B.; Miskowski, V. M. J. Am. Chem. Soc.
1984, 106, 3764-3767.
(43) Perutz, M. F. Annu. ReV. Biochem. 1979, 48, 327-386.
(44) Landrum, J. T.; Hatano, K.; Scheidt, W. R.; Reed, C. A. J. Am.
Chem. Soc. 1980, 102, 6729-6735.