Intermetal Oxygen Atom Transfer Reactions of Titanium Porphyrins: Complete vs Incomplete Atom Transfer. X-ray Structure of (μ-Oxo)bis[(meso-tetra-p-tolylporphyrinato)titaitium(III)]

Article abstract of DOI:10.1021/ic960863m

Treatment of (meso-tetra-p-tolylporphyrinato)titanium(IV) oxide, (TTP)Ti=O, with (octaethylporphyrinato)titanium-(III) chloride, (OEP)Ti-Cl, in toluene-d₈ results in the reversible exchange of axial ligands to form (TTP)Ti-Cl and (OEP)Ti=O. The net result is a formal one-electron redox process. This occurs with a second-order rate constant of (2.4 ± 0.2) × 10² M^-1 s^-1 to form an equilibrium mixture with K = 1.7 ± 0.4 at 20 °C (ΔH^? = 10.8 ± 0.4 kcal/mol, ΔS^? = -10.7 ± 1.2 cal/mol·K). Use of pivalate in place of chloride on the Ti(III) complex causes no significant change in the rate of this one-electron redox process. In addition, excess chloride only partially inhibits the rate of oxygen transfer with chlorotitanium(III) complexes. A complete kinetic analysis indicates that these redox processes proceed through two parallel pathways, both involving an inner sphere μ-oxo intermediate. Competing associative and dissociative (chloride or pivalate loss) mechanisms are in effect. Reversible oxo/chloro exchange also occurs between (TTP)TiCl₂ and (OEP)Ti=O in CHC1₃ to produce (TTP)-Ti=O and (OEP)TiCl₂with K = 47 ± 13 and k_f = 5 ± 1 M^-1 s^-1 at 20 °C (ΔH^? = 11 ± 1 kcal/mol, ΔS^? = 11 ± 3 cal/mol·K, ΔH° = -3.7 ± 0.8 kcal/mol, and ΔS° = -5 ± 2 cal/mol·K). In contrast, when (TTP)Ti=O is treated with (TPP)Ti(η²-PhC≡CPh), incomplete oxygen atom transfer occurs to produce the μ-oxo complex, [(TTP)Ti]₂O. X-ray quality crystals of this complex can be prepared by the slow diffusion of O₂ into a solution of (TTP)Ti(4-picoline)₂. The structure of the μ-oxo complex, C₁₁₄H₉₀N₈OTi₂, was determined by single-crystal X-ray diffraction (crystal data: monoclinic, C2/c, a = 32.044(8) A?, b = 16.908(5) A?, c = 18.525(5) A, β= 117.91(2)°, V = 8869(4) A?³, Z = 4, R = 5.88%, and A_w = 15.38%). Key metrical parameters include a Ti-O distance of 1.7941(8) A?, average Ti-N distances of 2.113(3) A?, and a Ti-O-Ti angle of 170.8(2)°. The Ti atom is displaced 0.66 A? out of the mean porphyrin plane toward the oxo bridge.

Full text of DOI:10.1021/ic960863m

Inorg. Chem. 1996, 35, 7601-7607
7601
Intermetal Oxygen Atom Transfer Reactions of Titanium Porphyrins: Complete vs
Incomplete Atom Transfer. X-ray Structure of
(µ-Oxo)bis[(meso-tetra-p-tolylporphyrinato)titanium(III)]
J. Alan Hays, Catherine L. Day, Victor G. Young, Jr., and L. Keith Woo,*,1
Department of Chemistry, Iowa State University, Ames, Iowa 50011-3111
ReceiVed July 19, 1996X
Treatment of (meso-tetra-p-tolylporphyrinato)titanium(IV) oxide, (TTP)TidO, with (octaethylporphyrinato)titanium-
(III) chloride, (OEP)TisCl, in toluene-d8 results in the reversible exchange of axial ligands to form (TTP)TisCl
and (OEP)TidO. The net result is a formal one-electron redox process. This occurs with a second-order rate
constant of (2.4 ( 0.2) × 102 M-1 s-1 to form an equilibrium mixture with K ) 1.7 ( 0.4 at 20 °C (∆Hq ) 10.8
( 0.4 kcal/mol, ∆Sq ) -10.7 ( 1.2 cal/mol‚K). Use of pivalate in place of chloride on the Ti(III) complex
causes no significant change in the rate of this one-electron redox process. In addition, excess chloride only
partially inhibits the rate of oxygen transfer with chlorotitanium(III) complexes. A complete kinetic analysis
indicates that these redox processes proceed through two parallel pathways, both involving an inner sphere µ-oxo
intermediate. Competing associative and dissociative (chloride or pivalate loss) mechanisms are in effect.
Reversible oxo/chloro exchange also occurs between (TTP)TiCl2 and (OEP)TidO in CHCl3 to produce (TTP)-
TidO and (OEP)TiCl2 with K ) 47 ( 13 and kf ) 5 ( 1 M-1 s-1 at 20 °C (∆Hq ) 11 ( 1 kcal/mol, ∆Sq ) 11
( 3 cal/mol‚K, ∆H° ) -3.7 ( 0.8 kcal/mol, and ∆S° ) -5 ( 2 cal/mol‚K). In contrast, when (TTP)TidO is
treated with (TPP)Ti(η2-PhCtCPh), incomplete oxygen atom transfer occurs to produce the µ-oxo complex,
[(TTP)Ti]2O. X-ray quality crystals of this complex can be prepared by the slow diffusion of O2 into a solution
of (TTP)Ti(4-picoline)2. The structure of the µ-oxo complex, C114H90N8OTi2, was determined by single-crystal
X-ray diffraction (crystal data: monoclinic, C2/c, a ) 32.044(8) Å, b ) 16.908(5) Å, c ) 18.525(5) Å, â )
117.91(2)°, V ) 8869(4) Å3, Z ) 4, R ) 5.88%, and Rw ) 15.38%). Key metrical parameters include a Ti-O
distance of 1.7941(8) Å, average Ti-N distances of 2.113(3) Å, and a Ti-O-Ti angle of 170.8(2)°. The Ti
atom is displaced 0.66 Å out of the mean porphyrin plane toward the oxo bridge.
Introduction
species.7 Incomplete oxygen atom transfer typically involves
a net one electron change. However, complete oxygen atom
Although atom transfer reactions are very common and have
been known for over 40 years, an understanding of oxygen atom
transfer processes that occur between two metal complexes is
still emerging.2 A large portion of the documented oxygen atom
transfer reactions take place between a metal oxo donor and
non-metal oxo acceptor, such as phosphines and alkenes.
Examples of intermetal oxygen-transfer reactions are less
common. In most cases, reaction of a metal oxo donor with
metal acceptor results in the formation of a µ-oxo species, eq
1. Holm has termed this type of reaction as incomplete atom
transfer can mediate both one- or two-electron redox processes.
Thus an understanding of the factors that influence the electron
stoichiometry involved in an oxygen atom transfer is also not
well developed. Our approach toward studying these important
issues has focused on systems which approach the self-exchange
category (∆G° ) 0). This allows an assessment of the relative
stability of the incipient µ-oxo species relative to those of the
oxo donor and acceptor complexes without complications arising
from a thermodynamic driving force. The energetics of this
bridged M-O-M unit clearly dictates the type of reaction
(complete vs incomplete oxygen atom transfer) and determines
the rate of atom transfer.
We have found that electronic factors are extremely important
in intermetal atom transfer reactions of metalloporphyrins.2 In
the work presented here, we describe a more detailed account
of oxygen atom transfer reaction between titanium porphyrin
complexes.
LnMdO + LnM′ f LnMsOsM′Ln
(1)
transfer.3 In contrast, complete atom transfer specifies the case
when the oxo ligand is totally cleaved from the donor complex,
(eq 2). This case has been documented for molybdenum,3
LnMdO + M′Ln f LnM + OdM′Ln
(2)
tungsten,3,4 and vanadium5 complexes, and implicated in the
oxoruthenium porphyrin-catalyzed aerobic epoxidation of alk-
enes.6
The fundamental factors which dictate the occurrence of
complete vs incomplete oxygen atom transfer are not well
understood. However, it is clear that steric factors can promote
complete oxygen atom transfer by destabilizing the µ-oxo
Experimental Section
General Data. All manipulations were carried out under an
atmosphere of nitrogen using either a Vacuum Atmosphere glovebox
equipped with a Model MO40H Dri-Train gas purifier or on a high
vacuum line using standard Schlenk techniques. All solvents were
(4) (a) Templeton, J. E.; Ward, B. C.; Chen, G. J.-J.; McDonald, J. W.;
Newton, W. E. Inorg. Chem. 1981, 20, 1248. (b) Yamashita, S.;
Imamura, T.; Sasaki, Y. Chem. Lett. 1995, 417-418.
X Abstract published in AdVance ACS Abstracts, December 1, 1996.
(1) 1990-1995 Presidential Young Investigator, 1993-1998 Dreyfus
Teacher-Scholar.
(5) Zhang, Y.; Holm, R. H. Inorg. Chem. 1990, 29, 911.
(2) Woo, L. K. Chem. ReV. 1993, 93, 1125.
(3) Holm, R. H. Chem. ReV. 1987, 87, 1401.
(6) Groves, J. T.; Quinn, R. J. J. Am. Chem. Soc. 1985, 107, 5790.
(7) Holm, R. H.; Berg, J. M. Pure Appl. Chem. 1984, 56, 1645.
S0020-1669(96)00863-4 CCC: $12.00 © 1996 American Chemical Society

7602 Inorganic Chemistry, Vol. 35, No. 26, 1996
Table 1. Molar Absorptivities for Metalloporphyrins
Hays et al.
to verify that decomposition of the chloride complex did not occur.
Kinetic runs were also done in the presence of excess chloride in the
form of bis(triphenylphosphoranylidene)ammonium chloride, (PPN)-
Cl. Typical initial concentrations ranged from 7.37 × 10-6 to 4.1 ×
10-5 M. Rate constants were obtained using an integrated law for
second order equilibrium reactions derived by King.13
ꢀ/(104 M-1 cm-1
)
λ ) 572 nma
λ ) 550 nmb
(TTP)TidO
(TTP)TisCl
(TTP)TiCl2
(OEP)TidO
(OEP)TisCl
(OEP)TiCl2
0.313
0.900
2.41
1.33
0.484
Rate data for the (POR)TidO/(POR)TiCl2 systems were obtained
on either a UV-vis spectrophotometer with a thermally regulated cell
holder (for runs at 20, 30, and 50 °C) or on a thermally regulated Varian
VXR 300 MHz NMR spectrometer. For runs on the UV-vis
spectrophotometer, CHCl3 solutions of the oxo complex and the
appropriate dichloride were loaded into a 1-cm quartz cuvette fitted
with a stopcock. The cuvette was placed in the cell holder and allowed
time to equilibrate to the proper temperature, and the run was monitored
at 550 nm. Molar absorptivities of the porphyrin complexes in CHCl3
are given in Table 1. A final spectrum from 500 to 700 nm was
obtained to verify that the dichloride complexes had not decomposed.
Typical concentrations of oxo and dichloride complexes were ap-
proximately 4.5 × 10-5 M. For kinetic runs at 0 °C, solutions of the
oxo complex, the appropriate dichloride complex, and a internal
standard, triphenylmethane, were transferred into a 5-mm NMR tube
fitted with a ground glass joint. The solvents were removed under
reduced pressure and a high vacuum line adapter was attached.
Deuterated solvent was vacuum transferred into the tube on a high
vacuum line, the tube was backfilled with N2 to approximately 550
mm Hg, and the NMR tube was flame sealed. The tube was kept in
liquid N2 prior to insertion into the instrument, and then it was thawed
slightly and placed into the magnet and allowed to equilibrate to the
proper temperature. Runs were monitored with either the meso protons
of the OEP complexes or the pyrrole â protons of the TPP complexes
and integrated with respect to the methine proton of the internal
standard. Comparison of these signals is used to verify that mass
balance was maintained during the run. In both cases, rate constants
were obtained using an integrated rate law for second-order equilibrium
reactions derived by King.13
3.02
1.06
0.479
a In toluene. b In CHCl3.
rigorously dried and degassed prior to use. Benzene-d6, toluene,
tetrahydrofuran, and hexane were vacuum distilled from purple solutions
of sodium-benzophenone. Methanol was distilled from sodium under
nitrogen over to 4 Å molecular sieves and then vacuum distilled a
second time prior to use. CH2Cl2, CDCl3 and CD2Cl2 were stored over
either P2O5 or CaH2 and vacuum distilled prior to use. Triphenyl-
methane was sublimed under reduced pressure at 100 °C prior to use.
Sodium pivalate was obtained from Aldrich Chemical Co., recrystallized
from methanol/CH2Cl2, and dried in vacuo at 125 °C. Commercially
obtained (PPN)Cl was dried in vacuo at 150 °C for 24 h. (TTP)TidO,8,9
11
(OEP)TidO,8 (TTP)TiCl,10 (OEP)TiCl,10 and (TTP)Ti(4-picoline)2
were prepared using literature procedures. (OEP)Ti(O2CC(CH3)3) was
prepared using a modification of a previously developed procedure.12
All reagents used for kinetic and equilibrium studies were recrystallized
twice prior to use. Concentrations of stock solutions were checked
1
before use either by UV-vis or H NMR spectroscopies. 1H NMR
spectra were recorded on either a Nicolet 300-MHz or a Varian VXR
300-MHz Fourier transform spectrometer. UV-vis data were recorded
on a HP8452A diode array spectrophotometer. IR spectra were
recorded on a either a IBM IR98 or a Digilab Fourier transform
spectrometer.
Equilibrium Measurements. Samples for equilibrium determina-
tions for the titanium oxo-chloride exchange and the oxo-dichloride
system were prepared by adding specific volumes of known concentra-
tions of an oxo complex, the opposing chloride or dichloride complex,
and an internal standard, triphenylmethane, into a 5-mm NMR tube
attached to a ground glass joint. The solvent was removed under
reduced pressure in a glovebox. After adding an NMR solvent, either
toluene-d8 (oxo-chloride exchange) or CDCl3 (oxo-dichloride ex-
change), the tube was sealed with a ground glass stopper. The tube
was frozen with liquid nitrogen and flamed sealed. For the oxo-
chloride system, the equilibrium constant was determined by integrating
the meso proton signal of the (OEP)TidO and the pyrrole â proton
signal of the (TTP)TidO. These signals were compared to the signal
of the methine proton of the triphenylmethane to verify that mass
balance was maintained and to determine concentrations. The tubes
were monitored in a temperature controlled NMR probe until no further
change in the spectrum was observed. For the oxo-dichloride system,
all species were diamagnetic, so meso protons and pyrrole â protons
were integrated with respect to the triphenylmethane protons to assure
that mass balance was maintained.
Synthesis of [(TPP)Ti]2O. A solution of (TTP)TidO (18 mg, 24
µmol) and (TTP)Ti(η2-PhCtCPh) (22 mg, 25 µmol) in 10 mL of
toluene was stirred under N2. Over 3 h, the bright ruby red solution
turned brick red with formation of a precipitate. After the mixture
was cooled at -20 °C for 16 h, a brick red product (15.6 mg, 48%)
was isolated by filtering and washing with neat toluene and neat hexane
and was dried in vacuo at 100 °C for 5 h. 1H NMR (CDCl3): δ 2.68
(br, s, CH3). UV-vis (toluene): 424 (Soret), 550 nm.
X-ray Structure Determination of [(TTP)Ti]2O. Crystals of
[(TTP)Ti]2O were grown from slowly diffusing O2 into a solution of
(TTP)Ti(4-picoline)2 in benzene/octane at ambient temperature.
A
brown prismatic crystal (0.35 × 0.35 × 0.15 mm) was attached to the
tip of a glass fiber and mounted on a Siemens P4/RA diffractometer
for data collection at -80 ( 1 °C using Cu KR radiation (λ ) 1.541 78
Å). Cell constants were determined from a list of reflections found by
a rotation photograph. Lorentz and polarization corrections were
applied. A correction based on a nonlinear decay in the standard
reflections of 5.8% was applied to the data. An absorption correction
based on a series of azimuthal scans using the semiempirical method
was applied. The agreement factor for the averaging of observed
reflections was 2.1% (based on F).
Kinetic Measurements. Rate data for the (POR)TidO/(POR)TisX
(X ) chloride, pivalate anion) systems were obtained on a UV-vis
spectrometer equipped with a thermally regulated cell holder. Solutions
of the oxo and the appropriate titanium(III) complexes were loaded
into a 1-cm quartz cuvette under nitrogen atmosphere and capped with
a septum. The cuvette was placed in the cell holder, and the run was
monitored at 572 nm. Molar absorptivities of the porphyrin complexes
are given in Table 1. A final spectrum from 500 to 700 nm was taken
The centric space group C2/c, was indicated by systematic absences
and intensity statistics.14 The positions of all non-hydrogen atoms were
determined by Fourier techniques. All non-hydrogen atoms were
refined with anisotropic thermal parameters. After the least-squares
refinement converged, all hydrogen atoms were placed at calculated
positions 0.95 Å from the attached carbon with isotropic temperature
factors set at a default value of 0.08 Å2. The hydrogens on the
porphyrin ring were later refined isotropically. The asymmetric unit
contains 1.5 benzene molecules. The hydrogens of the benzene
molecules were refined with isotropic temperature factors.
(8) Fournari, C.; Guilard, R.; Fontesse, M.; Latour, J.-M.; Marchon, J.-
C. J. Organomet. Chem. 1976, 110, 205.
(9) Abbreviations: TTP is the dianion of meso-tetra-p-tolylporphyrin, TPP
is the dianion of meso-tetraphenylporphyrin, and OEP is the dianion
of octaethylporphyrin.
(10) Berreau, L. M.; Hays, J. A.; Young, V. G., Jr.; Woo, L. K. Inorg.
Chem. 1994, 33, 105.
(11) Hays, J. A.; Young, V. G., Jr.; Day, C. L.; Caron, C.; D’Souza, F.;
Kadish, K. M.; Woo, L. K. Inorg. Chem. 1993, 32, 4186.
(12) Woo, L. K.; Goll, J. G.; Czapla, D. J.; Hays, J. A. J. Am. Chem. Soc.
1991, 113, 8478.
(13) King, E. L. Int. J. Chem. Kinet. 1982, 14, 1285. For a second-order
reversible reaction, A + B h C + D, ∆ is the displacement of any
species from its equilibrium value. ∆ ) [A] - [A]∞ ) [B] - [B]∞
)
[C]∞ - [C] ) [D]∞ - [D] and R ) [A]∞ + [B]∞ + ([C]∞ + [D]∞)/K.
(14) SHELXTL PLUS. Siemens Industrial Automation, Inc., Madison, WI.

Reactions of Titanium Porphyrins
Inorganic Chemistry, Vol. 35, No. 26, 1996 7603
Table 2. Rate Constants for Oxo Transfer
a
b
T/(°C)
K3
k3/102 M-1 s-1 a
K6
k6/M-1 s-1 b
58 ( 7
50
40
30
20
10
0
26 ( 10
8.1 ( 1.1
4.6 ( 0.2
2.4 ( 0.2
1.4 ( 0.3
0.56 ( 0.09
1.4 ( 0.4
1.7 ( 0.4
1.4 ( 0.4
1.3 ( 0.3
31 ( 8
47 ( 13
7 ( 2
5 ( 1
71 ( 11
0.18 ( 0.08
a In toluene. b In CHCl3.
The X-ray data collection was carried out at the Iowa State University
Molecular Structure Laboratory. Refinement calculations were per-
formed on a SGI INDY R4400-SC or a Pentium computer using the
SHELXTL version 5.0 programs.
Results
Reduction of Oxotitanium(IV) Porphyrin with Titanium-
(III) Porphyrin. Treatment of (TTP)TidO with (OEP)TisCl
in toluene-d8 results in spectral changes which are consistent
with the transfer of a terminally bound oxygen ligand between
Figure 1. Representative absorption vs time plot for reaction 3. [(TTP)-
TidO]o ) 1.95 × 10-5 M and [(OEP)TiCl]o ) 1.94 × 10-5 M, in
toluene at 20 °C. Inset shows plot of ln[∆/(R + ∆(1 + 1/K))] vs t, kf
) 240 ( 20 M-1 s-1
.
1
two metal complexes shown in eq 3. The H NMR spectrum
Effect of Solvent. (POR)Ti-X complexes (X is an anionic
ligand) are known to coordinate neutral donor ligands on the
back side of the metal,15,16 so the forward rates of eq 3 in THF
were examined. In all runs the data were found to follow
second-order reversible rate law for more than 3 half-lives. Rates
at 20 °C for eq 3 were found to exhibit a moderate dependence
(TTP)TidO + (OEP)TisCl {kk
}
3
-3
(TTP)TisCl + (OEP)TidO (3)
on solvent used (kTHF ) 70 M-1 s-1, ktoluene ) 240 M-1 s-1
,
of the resulting mixture contains a new resonance at 10.54 ppm
indicating the formation of (OEP)TidO. The resonance for
(TTP)TidO at 9.24 ppm does not completely disappear indicat-
ing that eq 3 is an equilibrium process. Since the titanium(III)
complexes are d1 and paramagnetic, the signals for these species
are extremely broadened and often not observed. The revers-
ibility of eq 3 can be confirmed by the complementary
experiment in which (OEP)TidO is treated with (TTP)TisCl.
This generates a final spectrum with peak positions identical to
that observed for the forward process.
and kCH Cl ) 460 M-1 s-1). A similar solvent effect was
2
2
observed for the system where chromium was used instead of
titanium.17
Effect of Axial Ligand. Axial ligand effects were studied
in these systems by changing the monoanionic ligand of the
starting titanium(III) porphyrin complexes as shown in eq 5. A
(TTP)TidO + (OEP)TisX {kk
}
5
-5
Equilibrium constants for eq 3 were established by using 1H
NMR to monitor the pyrrole â proton resonance of (TTP)TidO
(9.24 ppm), the meso proton signal of (OEP)TidO (10.54 ppm),
and the methine proton of the internal standard, triphenyl-
methane (5.34 ppm). Flamed-sealed NMR tubes containing
(TTP)TidO, (OEP)TisCl, and triphenylmethane in toluene-d8
were prepared. During the course of the experiment, mass
balance was maintained in terms of total oxo complex concen-
tration. The equilibrium constants for eq 3, determined over a
40-deg range, were all very close to K3 ) 1.5 ( 0.4 and are
listed in Table 2. Small errors in integration of proton
resonances lead to a relatively wide range of standard deviations
in calculated equilibrium constants. Thus, a reliable temperature
dependence could not be obtained.
The forward rates of eq 3 in toluene were determined
spectrophotometrically by following the absorbance changes
at 572 nm. A summary of rate constants is listed in Table 2.
All of the compounds involved, (TTP)TidO, (TTP)TisCl,
(OEPTidO, and (OEP)TisCl, obey Beer’s law over the con-
centration range used in these experiments. This indicates self-
association does not complicate the mechanistic analysis. In
all kinetic runs, the data were found to obey an integrated rate
law for reversible second-order reactions shown in eq 4.13 Plots
(TTP)TisX + (OEP)TidO (5)
comparison of two different axial ligands, pivalate and chloride,
was made at 20 °C in THF. Kinetic runs for the pivalate system
were found to follow an integrated rate law for second-order
reversible reactions. Rates for eq 5 were found to have little
dependence on the nature of the axial ligand (at 20 °C: kpiv
)
100 ( 10 M-1 s-1 and kchloride ) 70 ( 10 M-1 s-1).
Effect of Added Axial Ligand. The rate of oxygen atom
transfer was studied as a function of added chloride ion. Bis-
(triphenylphosphoranylidene)ammonium chloride, (PPN)Cl, was
used as the source of chloride. Due to the low solubility of
(PPN)Cl in toluene, kinetic runs were performed in CH2Cl2
solutions. The dissociation constant for (PPN)Cl at 20 °C in
CH2Cl2 is 5.10 × 10-4 18
.
The rate data for all runs in CH2Cl2 were found to follow a
second-order reversible rate law at 20 °C. In the absence of
any added chloride, the rate constant for the eq 3 in CH2Cl2 at
20 °C is (4.6 ( 0.4) × 102 M-1 s-1. A 23-fold excess of Cl-,
with respect to (OEP)TiCl, decreases the rate constant to (4.3
( 0.3) × 102 M-1 s-1; a 54-fold excess of [Cl-] resulted in a
rate constant of (3.9 ( 0.3) × 102 M-1 s-1. Using a 427-fold
excess, the rate constant was only reduced to (2.6 ( 0.3) × 102
M-1 s-1
.
∆
(15) Latour, J.-M.; Marchon, J.-C.; Nakajima, M. J. Am. Chem. Soc. 1979,
101, 3974.
(16) Marchon, J.-C.; Latour, J.-M.; Boreham, C. J. J. Mol. Catal. 1980, 7,
227.
(17) Woo, L. K.; Goll, J. G.; Berreau, L. B.; Weaving, R. J. Am. Chem.
Soc. 1992, 114, 7411.
(18) Algra, G. P.; Balt, S. Inorg. Chem. 1981, 20, 1102.
ln
) -Rkt + constant
(4)
[
]
R + ∆(1 + 1/K)
of ln[∆/(R + ∆(1 - 1/K))] vs time are linear for at least 3 half-
lives. An absorbance vs time curve of a typical kinetic run
and a log plot of the data are shown in Figure 1.

7604 Inorganic Chemistry, Vol. 35, No. 26, 1996
Hays et al.
Figure 2. Absorption vs time plot of reaction 6. [(TTP)TiCl2]o ) 4.44
× 10-5 M and [(OEP)TidO]o ) 4.41 × 10-5 M, in CHCl3 at 20 °C.
Figure 3. Molecular structure for the asymmetric unit of [(TTP)Ti]2O.
Thermal ellipsoids are drawn at the 30% probability level.
Inset shows plot of ln[∆/(R + ∆(1 + 1/K))] vs t, kf ) 5 ( 1 M-1 s-1
.
Oxygen Atom Transfer between (TTP)TiCl2 and (OEP)-
TidO. Treatment of (TTP)TiCl2 with (OEP)TidO in CDCl3
results in spectral changes consistent with the transfer of a
terminally bound oxo ligand between two metal centers (eq 6).
of all of the 1H NMR signals for the starting materials is
observed. The final NMR spectrum contains peaks due to free
diphenyl acetylene and a broad resonance at 2.68 ppm from a
new paramagnetic µ-oxo product, [(TTP)Ti]2O (eq 7). The Ti-
(TTP)TiCl2 + (OEP)TidO { k } (TTP)TidO + (OEP)TiCl2
6
(TTP)TidO + (TTP)Ti(PhCtCPh) a
[(TTP)Ti]2O + PhCtCPh (7)
k-6
(6)
All species are diamagnetic making it easy to verify that transfer
has occurred. The 1H NMR spectrum of the resulting mixture
contains two new signals at 10.48 and 9.12 ppm indicating the
formation of (OEP)TiCl2 and (TTP)TidO, respectively. The
signals for (TTP)TidO (8.99 ppm) and (OEP)TidO (10.54
ppm) do not entirely disappear indicating that eq 6 is an
equilibrium process. The equilibrium (eq 6) can be verified
by the reverse process in which (TTP)TidO is treated with
(OEP)TiCl2. This generates a final spectrum identical to that
observed for the forward process.
Equilibrium constants for eq 6 are once again conveniently
established by 1H by monitoring all meso proton and pyrrole â
proton signals, since all species are diamagnetic. Flamed-sealed
NMR tubes were prepared containing (TTP)TiCl2 and (OEP)-
TidO plus an internal standard, triphenylmethane in CDCl3.
Initial concentrations for the porphyrin complexes range from
0.5 to 2.0 mM. During the course of the experiment, mass
balance was maintained. The equilibrium constants to eq 6 were
determined over a 50-deg range and are listed in Table 2. The
thermodynamic parameters, ∆H6° ) -3.7 ( 0.8 kcal/mol and
∆S6° ) -5.1 ( 0.7 cal(mol‚K)-1, were determined from this
temperature dependence.
O-Ti dimer can be cleanly converted to (TTP)TidO on
treatment with O2. Despite repeated attempts, it has not been
possible to prepare pure samples of [(TTP)Ti]2O which are free
of the terminal oxo complex, (TTP)TidO. Nonetheless, X-ray
quality single crystals of [(TTP)Ti]2O were produced by slow
diffusion of O2 into a benzene/octane solution of the Ti(II)
complex, (TTP)Ti(4-picoline)2.
X-ray Structure of [(TPP)Ti]2O. The molecular structure
of [(TTP)Ti]2O was determined by single-crystal X-ray diffrac-
tion. Thermal ellipsoids and the atom numbering scheme of
the asymmetric unit are shown in Figure 3. Crystallographic
data for the structure determination are listed in Table 3 and
fractional coordinates for the non-hydrogen atoms are in Table
4. Selected bond distances and angles are given in Table 5.
The structure determination of [(TTP)Ti]2O confirms its
formulation as a neutral µ-oxo-bridged Ti(III) dimer. The Ti
atom resides in a square pyramidal coordination sphere formed
by four basal pyrrole nitrogens and an apical oxygen atom with
the Ti displaced from the N4 plane 0.66 Å toward the bridging
oxygen. The average metrical parameters for the porphyrin
skeleton are N-CR ) 1.397(5) Å, CR-Câ ) 1.436(6) Å, Câ1-
Câ2 ) 1.347(6) Å, CR-Cmeso ) 1.401(6) Å, and Cmeso-Cipso
)
The forward rates of eq 6 in CHCl3 (Table 2) were examined
both by spectrophotometric methods, monitoring the absorbance
1.494(6) Å. These bond lengths are comparable to those found
in other porphyrin complexes.19 In addition, the porphyrin
ligand has an S4-type ruffling in which the meso carbons are
alternately displaced 0.08 Å above and below the mean
porphyrin plane. The Ti-O distance of 1.794 (1) Å is shorter
than that expected for a Ti-O single bond (ca. 2.0 Å)20 but is
very similar to the Ti-O distance (1.77(1) Å) found in
1
changes at 550 nm, and by H NMR, observing the spectral
changes of the meso and pyrrole â protons. For the UV-vis
runs, all compounds listed in Table 1 obey Beer’s law over the
concentration range of the experiments. This indicates self-
association does not complicate the mechanistic analysis. In
all kinetic runs, the data are found to obey an integrated rate
law for reversible second-order reactions as shown in eq 4. Plots
of ln[∆/(R + ∆(1 - 1/K))] vs time are linear for at least 3 half
lives. A typical kinetic curve and the corresponding log plot
are illustrated in Figure 2.
21
(TPP)Ti-OCH3 and falls in the range observed for other
Ti-O-Ti fragments.22 The two mean porphyrin planes in
(19) (a) Scheidt, W. R. In The Porphyrins, Dolphin, D., Ed.; Academic
Press: New York, 1978; Vol. III, Chapter 10. (b) Scheidt, W. R.;
Lee, Y. J. Struct. Bonding 1987, 64, 1.
(20) Peterson, J. L. Inorg. Chem. 1981, 20, 181.
(21) Boreham, C. J.; Buisso, G.; Duee, E.; Jordanov, J.; Latour, J.-C.;
Marchon, J.-C. Inorg. Chim. Acta 1983, 70, 77.
Reduction of (TTP)TidO with (TTP)Ti(η2-PhCtCPh).
When (TTP)TidO is treated anaerobically with the Ti(II)
complex, (TTP)Ti(η2-PhCtCPh), in a sealed NMR tube, loss

Reactions of Titanium Porphyrins
Inorganic Chemistry, Vol. 35, No. 26, 1996 7605
Table 3. Crystallographic Parameters for [(TTP)Ti]2O‚3C6H6
Table 4. Atomic Coordinates (×104) and Equivalent Isotropic
Displacement Coefficients (Å2 × 103) for [(TTP)Ti]2O‚3C6H6
empirical formula: C114H90N8O1Ti2
fw ) 1683.78
cryst syst: monoclinic
space group C2/c
unit cell dimens
x
y
z
U(eq)
SOF
Ti
O
N(1)
N(2)
N(3)
N(4)
C(1)
C(2)
4945(1)
5000
1537(1)
1451(2)
2196(2)
2570(2)
929(2)
6494(1)
7500
29(1)
41(1)
33(1)
33(1)
36(1)
33(1)
37(1)
46(1)
44(1)
36(1)
35(1)
37(1)
82(2)
88(2)
53(1)
87(2)
76(2)
77(2)
34(1)
39(1)
40(1)
35(1)
38(1)
40(1)
42(1)
46(1)
51(1)
56(1)
50(1)
70(1)
40(1)
55(1)
55(1)
38(1)
36(1)
36(1)
92(2)
90(2)
50(1)
118(3)
103(2)
79(2)
35(1)
46(1)
47(1)
38(1)
37(1)
39(1)
49(1)
54(1)
51(1)
51(1)
45(1)
74(2)
75(1)
77(1)
92(2)
95(2)
75(1)
72(1)
164(5)
171(5)
173(4)
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
4324(1)
5310(1)
5523(1)
4543(1)
3869(1)
3546(1)
3800(1)
4288(1)
4664(1)
4553(1)
4592(2)
4491(2)
4355(1)
4332(2)
4430(2)
4244(2)
5140(1)
5522(1)
5920(1)
5789(1)
6101(1)
6605(1)
6889(1)
7355(1)
7557(1)
7276(1)
6809(1)
8063(2)
5972(1)
6290(1)
6040(1)
5560(1)
5189(1)
5318(1)
5292(3)
5439(2)
5616(1)
5652(3)
5502(3)
5782(2)
4716(1)
4333(1)
3932(1)
4060(1)
3742(1)
3228(1)
2933(1)
2461(1)
2271(1)
2564(1)
3034(1)
1757(1)
2557(2)
2274(2)
1824(2)
1653(2)
1937(2)
2389(2)
2579(6)
2130(5)
2064(4)
5811(2)
6464(2)
6492(2)
5851(2)
5431(2)
5169(2)
5392(2)
5793(2)
6073(2)
5984(2)
5396(3)
5298(3)
5781(3)
6370(4)
6475(3)
5649(4)
6388(2)
6633(2)
6837(2)
6732(2)
6851(2)
7118(2)
7878(2)
8112(3)
7612(3)
6864(3)
6618(3)
7854(3)
6734(2)
6880(3)
6729(3)
6483(2)
6266(2)
6370(2)
5753(3)
5862(3)
6585(3)
7213(3)
7103(3)
6706(3)
5961(2)
5700(2)
5431(2)
5516(2)
5301(2)
4879(2)
5216(2)
4822(3)
4096(3)
3765(3)
4145(2)
3684(3)
1558(3)
1337(3)
1232(4)
1342(4)
1557(3)
1674(3)
4983(8)
4563(8)
4595(6)
a ) 32.044(8) Å
b ) 16.908(5) Å
c ) 18.525(5) Å
543(2)
â ) 117.91(2)°
1913(2)
2564(2)
3232(2)
3011(2)
3547(2)
4415(2)
4851(3)
5652(3)
6045(2)
5620(3)
4814(3)
6919(3)
3335(2)
3883(2)
3461(2)
2639(2)
2016(2)
2210(2)
2559(2)
2743(2)
2586(2)
2226(3)
2045(2)
2805(3)
1216(2)
575(2)
V ) 8869(4) Å3
Z ) 4
d(calcd) ) 1.261 g/cm3
C(3)
C(4)
C(5)
abs coeff: 1.975 mm-1
F(000) ) 3528
C(51)
C(52)
C(53)
C(54)
C(55)
C(56)
C(57)
C(6)
radiation: Cu KR (λ ) 1.54178 Å)
T ) 193(2) K
θ range: 3.04-57.15°
monochromator: highly oriented graphite cryst
no. of reflcns collcd: 11 953
no. of independent reflcns: 5983 (Rint ) 4.07%)
no. of observed reflcns: 4570 (F > 6.0σ(F))
max and min transm: 1.000 and 0.6090
no. of params refined: 568
final R indices (obs data): R ) 5.88%, Rw ) 15.38%
goodness-of-fit on F2: 1.060
data-to-param ratio: 8.0:1
largest difference peak: +0.458 e Å-3
largest difference hole: -0.444 e Å-3
C(7)
C(8)
C(9)
C(10)
C(101)
C(102)
C(103)
C(104)
C(105)
C(106)
C(107)
C(11)
C(12)
C(13)
C(14)
C(15)
C(151)
C(152)
C(153)
C(154)
C(155)
C(156)
C(157)
C(16)
C(17)
C(18)
C(19)
C(20)
C(201)
C(202)
C(203)
C(204)
C(205)
C(206)
C(207)
C(301)
C(302)
C(303)
C(304)
C(305)
C(306)
C(307)
C(308)
C(309)
[(TTP)Ti]2O deviate from a parallel arrangement by only 5.7°.
In addition, the two porphyrin ligands are staggered 26.4° about
the Tia-Tib axis from the eclipsed geometry. The Ti-O-Ti
alignment is nearly linear with a bond angle of 170.9(2)°. Thus,
[(TTP)Ti]2O is essentially isostructural with [(TPP)Fe]2O,23 but
notably different from the scandium analog, [(TTP)Sc]2O, where
Sc-O-Sc angles are 109(3) and 111(3)°.24
-104(2)
115(2)
-423(2)
-1281(2)
-1733(3)
-2508(3)
-2862(2)
-2406(3)
-1624(3)
-3713(3)
-218(2)
-773(2)
-347(2)
476(2)
1106(2)
903(2)
1088(2)
868(3)
457(2)
277(2)
501(2)
218(3)
128(3)
776(3)
721(3)
16(3)
-640(3)
-587(3)
8300(7)
7981(8)
7200(7)
Discussion
We have found that complete oxygen atom transfer resulting
in either one- or zero-electron exchange between two metal-
loporphyrins can be achieved. It is possible to observe this
process by using different porphyrin ligands as UV-vis and
1H NMR spectroscopic labels. An additional benefit derived
from the use of metalloporphyrins arises from the structural
integrity maintained by these complexes throughout the reaction,
eliminating complications due to ligand loss. Furthermore,
stereochemical rearrangements found in other systems that
undergo complete atom transfer are not possible here.3,25 Rate
constants for eq 3 have been measured in the forward direction
over a 40-deg temperature span in toluene. These rate constants
range between 60 and 810 M-1 s-1. The temperature depen-
dence of the forward rate yields activation parameters of ∆Hq
3
) 10.8 ( 0.4 kcal/mol and ∆Sq3 ) -10.7 ( 1.2 cal(mol‚K)-1
.
Rate constants for eq 6 have been measured in the forward
direction over a 50-deg temperature range in CH2Cl2. These
rate constants range between 0.18 and 58 M-1 s-1
.
The
temperature dependence of the forward rate constants yield
activation parameters of ∆Hq6 ) 11 ( 1 kcal/mol and ∆Sq
)
6
11 ( 3 cal(mol‚K)-1. Although eq 3 proceeds, in all probability,
by an inner sphere process, mechanistic aspects of this reaction
are complicated by the bridging abilities of both the chloride
and oxo ligands. In order to address this issue, we have
compared the reduction of (TTP)TidO with that of both (OEP)-
TisCl and (OEP)TisO2CC(CH3)3. The rate constants for these
two reaction are similar, regardless of whether chloride (kCl
70 M-1 s-1) or pivalate (kpiv ) 100 M-1 s-1) is the axial ligand
on titanium(III). Because pivalate is a much bulkier ligand,
this rate comparison serves to rule out a binuclear intermediate
in which the monoanionic ligand is involved in bridging the
(22) Orpen, A. G.; Brammer, L.; Allen, F. H.; Kennard, O.; Watson, D.
G.; Taylor, R. J. Chem. Soc., Dalton Trans. 1989, S1.
(23) Hoffman, A. B.; Collins, D. M.; Day, V. W.; Fleischer, E. B.;
Srivastava, T. S.; Hoard, J. L. J. Am. Chem. Soc. 1972, 94, 3620.
(24) Sewchok, M. G.; Haushalter, R. C.; Merola, J. S. Inorg. Chim. Acta
1988, 144, 47.
)
(25) Holm, R. H. Coord. Chem. ReV. 1990, 100, 183.

7606 Inorganic Chemistry, Vol. 35, No. 26, 1996
Hays et al.
Table 5. Selected Bond Lengths (Å) and Angles (deg) for
[(TTP)Ti]2O‚3C6H6
Ti-O
1.7941 (8)
2.110 (3)
2.119 (3)
Ti-N(3)
2.118 (3)
2.114 (3)
1.376 (5)
Ti-N(1)
Ti-N(2)
O-Ti-N(1)
Ti-N(4)
N(1)-C(1)
O-Ti-N(4)
105.7(1)
105.0(1)
O-Ti-N(2)
107.8(1)
85.8(1)
107.5(1)
146.7(1)
84.6(1)
N(1)-Ti-N(4)
N(2)-Ti-N(4)
N(3)-Ti-N(4)
Ti-O-Ti #1
85.8(1)
147.1(1)
85.2(1)
N(1)-Ti-N(2)
O-Ti-N(3)
N(1)-Ti-N(3)
N(2)-Ti-N(3)
170.8(2)
Scheme 1
two metals. Thus the µ-chloro-bridged adduct, A, and the
doubly bridged species, B, must not be important intermediates
along the reaction pathway.
Figure 4. Building up of the orbitals of [(NH2-)4M-O-M(NH2-)4]4-
.
From left to right: the orbitals of the pyramidal N4M in which the M
is 0.5 Å out of the N4 plane; the orbitals of two N4M units with an
M-M separation of 3.526 Å; the orbitals of the composite oxo complex.
It is likely that the redox reaction between (TTP)TidO and
(OEP)TisCl involves a µ-oxo intermediate. The single-crystal
X-ray structure of (TTP)TisOsTi(TTP) serves as a model for
this bridged intermediate. Formation of the µ-oxo intermediate
can occur either by an SN1-type or an SN2-type mechanism.
These two processes differ primarily by the timing in which
Cl- dissociatessprior to or subsequent to formation of the
binuclear µ-oxo species. Note that added Cl- weakly inhibits
the oxo transfer in eq 3. Formation of a six-coordinate anion,
[(OEP)TiCl2]-, which slows the formation of a µ-oxo interme-
diate is unlikely since related porphyrin complexes show little
affinity for an anionic ligand.26 The important steps to consider
in unravelling the mechanism are shown in Scheme 1. The
integrated rate law derived from Scheme 1 is given in eq 8.
reductant, the result is reversible oxygen atom transfer. In
the case of oxotitanium(IV) porphyrins, we observe similar
results. Mechanistic evidence for the TidO/TisX reaction (eq
3) is consistent with a pathway which involves a µ-oxo
intermediate, [TiIV-O-TiIII]5+. Single-electron transfer within
this activated complex always produces one metal in a +IV
oxidation state. The instability of [TiIVsOsTiIII]4+ relative to
the structurally characterized [TiIII-O-TiIII]4+ complex may
simply reflect an inherent tendency for the Ti(IV) d0 center
and the oxo ligand to form a strong metal-oxygen double
bond rather than for the oxo ligand to form two M-O single
bonds. Thus, a key difference between the stability of the
two µ-oxo species, [TiIV-O-TiIII]5+ and [TiIII-O-TiIII]4+
,
kS 1K + k 2[Cl-]
SN
d[products]
N
solely may be a matter of oxidation state and the d-electron
configuration. In order to form OdTiIV, the III-III complex
must disproportionate while the III-IV intermediate simply
dissociates.
)
[TisCl][TidO]
(8)
{
}
K + [Cl-]
dt
The apparent rate constant, kapp, and observed chloride ion
dependencies in CH2Cl2 were found to fit eq 8 with values of
kS 1 ) (4.8 ( 0.3) × 102 and kS 2 ) (2.3 ( 0.9) × 102 M-1
An examination of the electronic structure of µ-oxometal-
loporphyrin dimers provides an alternative rationale for the
oxygen atom transfer chemistry of titanium porphyrins. The
d-orbital energy level for [Fe(TTP)]2O shown in Figure 4 was
derived by Tatsumi and Hoffmann28 using extended Hu¨ckel
calculations on [(NH2-)4Fe-O-Fe(NH2-)4]4- as a computa-
tional model. To a first approximation, this energy level
diagram should apply to the [(POR)M-O-M(POR)]0,+ (M )
Cr, Ti) system. In the chromium case, when chromium(II) is
the reductant, the neutral (POR)Cr-O-Cr(POR) dimer has six
d-electrons. This leads to a (e2)4(e3)2 ground state which is
electronically stable with respect to complete oxygen atom
N
N
s-1. Thus both SN1- and SN2-type pathways occur as competing
processes. This is consistent with the weak chloride inhibition
results. In addition, the small rate reduction in a coordinating
solvent such as THF is presumably due to inhibition of both
SN1 and SN2 pathways.
The reduction products of oxotitanium porphyrins are strongly
dependent on the nature of the reducing agent. This behavior
is analogous to the inner sphere reduction of oxochromium-
(IV) porphyrin complexes. When chromium(II) porphyrin is
the reducing agent, the reaction halts at the µ-oxo dimer.27
However, when a chromium(III) porphyrin is used as the
(27) (a) Liston, D. J.; Murray, K. S.; West, B. O. J. Chem. Soc., Chem.
Commun. 1982, 1109. (b) Liston, D. J.; West, B. O. Inorg. Chem.
1985, 24, 1568.
(28) (a) Tatsumi, K.; Hoffmann, R.; Whangbo, M.-H. J. Chem. Soc., Chem.
Commun. 1980, 509. (b) Tatsumi, K.; Hoffmann, R. J. Am. Chem.
Soc. 1981, 103, 3328.
(26) (a) Bottomley, L. A.; Kadish, K. M. Inorg. Chem. 1983, 22, 342-
349. (b) Summerville, D. A.; Jones, R. D.; Hoffmann, B. M.;
Basolo, F. J. Am. Chem. Soc. 1977, 99, 8195. (c) Basolo, F.;
Jones, R. D.; Summerville, D. A. Acta Chem. Scand., Ser. A 1978,
A32, 771.

Reactions of Titanium Porphyrins
Inorganic Chemistry, Vol. 35, No. 26, 1996 7607
Concluding Remarks
transfer. However, when chromium(III) is the reductant, the
cationic d5 µ-oxo species must have either a low-spin (e2)4(e3)1
or high-spin (e2)2(e3)2(e1)1 ground state. Either state is elec-
tronically degenerate and Jahn-Teller destabilized, leading
to cleavage of the µ-oxo species. In similar manner, we can
describe the titanium system. If titanium(II) is used as the reduc-
tant, the neutral µ-oxo species has a (e2)2 ground state which
is electronically stable. When titanium(III) is used as the
reductant, the cationic d1 µ-oxo intermediate has a (e2)1 ground
state and is electronically degenerate, leading to cleavage of
the µ-oxo intermediate. Hence, the OdTiIV/ClsTiIII reaction
is electronically disposed toward complete oxygen atom trans-
fer while the OdTiIV/TiII system produces a stable µ-oxo
product.
The OdTi/TiCl2 exchange, eq 6, is not predicted by a simple
MO analysis. In this case, the putative [TiIV-O-TiIV]6+
intermediate is d0 and should be electronically stable. Further-
more, a neutral binuclear complex, Cl(TTP)Ti-O-Ti(TTP)Cl
would appear to be a reasonable product. However, the
formation of a strong TidO double bond again may be an
important driving force. Mechanistically, the pseudooctahedrdal
(TTP)TiCl2 complex is unlikely to undergo an SN2-type attack
by the oxo complex. Prior dissociation of one of the chloride
ligands is necessary to open a vacant coordination site. The
positive ∆Sq for eq 6 suggests that Cl- loss is the rate-
determining step.
A number of significant results have evolved from this study.
Of particular interest is the observation that complete intermetal
oxygen atom transfer is possible and facile for titanium
porphyrins. Oxygen atom transfer in titanium porphyrin
complexes can mediate both a one- and zero-electron process.
The reversible processes reported here are the first such
examples involving titanium porphyrins. It is clear that the oxo
ligand is the preferred bridging species despite the presence
of other good bridging ligands (e.g. chlorides and carboxylates).
Finally, a simplified molecular orbital analysis suggests that
the complete oxo transfer between (POR)TidO and (POR)-
TiCl is promoted by electronic factors and the strong tendency
for metal-oxygen double bond formation in a d° Ti(IV)
complex.
Acknowledgment. Support for this work was provided by
National Science Foundation Grant CHE-9057752 and the
donors of the Petroleum Research Fund, administered by the
American Chemical Foundation.
Supporting Information Available: Tables of crystallographic
data, bond lengths and angles, anisotropic displacement factors, torsion
angles, and hydrogen atom coordinates (17 pages). See any current
masthead page for ordering information and Internet access instructions.
IC960863M