Seven-coordinate iron complex as a ditopic receptor for lithium salts: Study of host-guest interactions and substitution behavior

Article abstract of DOI:10.1021/ic7005056

Interactions between the seven-coordinate tweezerlike [Fe(dapsox)(H ₂O)₂]ClO₄ complex (H₂dapsox = 2,6-diacetylpyridine-bis(semioxamazide)) with different lithium salts (LiOTf, LiClO₄, LiBF₄, and LiPF₆) in CH₃CN have been investigated by electrochemical, spectrophotometric, ⁷Li and ¹⁹F NMR, kinetic, and DFT methods. It has been demonstrated that this complex acts as ditopic receptor, showing spectral and electrochemical ion-pair-sensing capability for different lithium salts. In general, the apparent binding constants for lithium salts increase in the order LiOTf < LiClO₄ < LiBF₄. From the electrochemical measurements, the apparent lithium salt binding constants for the Fe(III) and Fe(II) forms of the complex have been obtained, suggesting a stronger host-guest interaction with the reduced form of the complex. In the presence of LiPF₆, the solution chemistry is more complex because of the hydrolysis of PF ₆^-. The kinetics of the complexation of [Fe(dapsox)(CH₃CN)₂]⁺ by thiocyanate at -15°C in acetonitrile in the presence of 0.2 M NBu₄OTf shows two steps with the following rate constants and activation parameters: k₁ = 411 ± 14 M^-1 s^-1; ΔH₁ ^≠ = 9 ± 2 kJ mol^-1; ΔS₁ ^≠ = -159 ± 6 J K^-1 mol^-1; k ₂ = 52 ± 1 M^-1 s^-1; ΔH ₂^≠ = 4 ± 1 kJ mol^-1; ΔS ₂^≠ = -195 ± 3 J K^-1 mol^-1. The very negative ΔS^≠ values are consistent with an associative (A) mechanism. Under the same conditions but with 0.2 M LiOTf, k_1Li and k_2Li are 1605 ± 51 and 106 ± 2 M^-1 s^-1 respectively. The increased rate constants for the {[Fe(dapsox)(CH₃CN)₂]-LiOTf}⁺ adduct are in agreement with an associative mechanism. Kinetic and spectrophotometric titration measurements show stronger interaction between the lithium salt and the anion-substituted forms, [Fe(dapsox)(CH₃CN)(NCS)] and [Fe(dapsox)(NCS)₂]^-, of the complex. These experiments demonstrate that in nonaqueous media lithium salts cannot be simply used as supporting electrolytes, since they can affect the kinetic behavior of the studied complex. DFT calculations revealed that the negatively charged α-oxyazine oxygen atoms are responsible for cation binding (electrostatic interactions), whereas the two terminal amide groups bind the anion via hydrogen bonding.

Full text of DOI:10.1021/ic7005056

Inorg. Chem. 2007, 46, 7848−7860
Seven-Coordinate Iron Complex as a Ditopic Receptor for Lithium
Salts: Study of Host Guest Interactions and Substitution Behavior
−
David Sarauli,† Vesselina Popova,‡ Achim Zahl,† Ralph Puchta,† and Ivana Ivanovi c´ -Burmazovi c´ *,†
Institute for Inorganic Chemistry, UniVersity of Erlangen-N u¨ rnberg, Egerlandstrasse 1,
D-91058 Erlangen, Germany, Max-Planck-Institute for Iron Research GmbH,
Max-Planck-Strasse 1, D-40237 D u¨ sseldorf, Germany
Received March 15, 2007
Interactions between the seven-coordinate tweezerlike [Fe(dapsox)(H
diacetylpyridine-bis(semioxamazide)) with different lithium salts (LiOTf, LiClO
been investigated by electrochemical, spectrophotometric, Li and F NMR, kinetic, and DFT methods. It has been
demonstrated that this complex acts as ditopic receptor, showing spectral and electrochemical ion-pair-sensing
capability for different lithium salts. In general, the apparent binding constants for lithium salts increase in the order
2
O)
2
]ClO
, LiBF
4
complex (H 2,6-
2
dapsox
)
4
4
, and LiPF ) in CH CN have
6
3
7
19
LiOTf < LiClO
the Fe(III) and Fe(II) forms of the complex have been obtained, suggesting a stronger host
the reduced form of the complex. In the presence of LiPF , the solution chemistry is more complex because of the
hydrolysis of PF CN) 15 C in acetonitrile
-. The kinetics of the complexation of [Fe(dapsox)(CH ]+ by thiocyanate at
4
< LiBF
4
. From the electrochemical measurements, the apparent lithium salt binding constants for
−guest interaction with
6
6
3
2
−
°
in the presence of 0.2 M NBu
4
*
OTf shows two steps with the following rate constants and activation parameters:
14 M s-1;
-
1
∆
H
)
±
9
±
2 kJ mol 1;
-
∆
S
*
) −159
±
∆
6 J K- mol-1; k
1
)
52
±
1 M- s-1;
1
∆
H
*
k
)
1
)
411
±
1
1
2
2
1 kJ mol-1;
∆S
*
) −195
3 J K mol- . The very negative
-1
1
S* values are consistent with an associative
4
±
2
2 M- s-1,
1
(
A) mechanism. Under the same conditions but with 0.2 M LiOTf, k1Li and k2Li are 1605
respectively. The increased rate constants for the [Fe(dapsox)(CH CN)
LiOTf +
an associative mechanism. Kinetic and spectrophotometric titration measurements show stronger interaction between
the lithium salt and the anion-substituted forms, [Fe(dapsox)(CH CN)(NCS)] and [Fe(dapsox)(NCS)
]-, of the complex.
±
51 and 106
±
{
3
2
]
‚
} adduct are in agreement with
3
2
These experiments demonstrate that in nonaqueous media lithium salts cannot be simply used as supporting
electrolytes, since they can affect the kinetic behavior of the studied complex. DFT calculations revealed that the
negatively charged
R-oxyazine oxygen atoms are responsible for cation binding (electrostatic interactions), whereas
the two terminal amide groups bind the anion via hydrogen bonding.
Introduction
desirable because of their higher kinetic and thermodynamic
stability than manganese complexes.
2
The solution properties and mechanistic behavior of seven-
coordinate 3d metal complexes are still unrevealed research
areas. These complexes are especially interesting as excellent
To gain an understanding of the solution chemistry of the
iron seven-coordinate complexes, the substitution behavior
and proton-coupled electron-transfer reactions of the pen-
•
-
low-molecular-weight catalysts for superoxide (O
2
) dis-
tagonal-bypiramidal (PBP) Fe(III) complex, [Fe(dapsox)-
mutation (SOD), providing important therapeutic applica-
+
(H
2
O)
2
] (H
2
dapsox ) 2,6-diacetylpyridine-bis(semioxam-
1
3
tions. Although the best SOD mimetics known to date are
azide)), with an anionic planar pentadentate ligand in the
seven-coordinate complexes of Mn(II) with macrocyclic
pentadentate ligands, iron SOD mimetics would also be
(1) (a) Aston, K.; Rath, N.; Naik, A.; Slomczynska, U.; Schall, O. F.;
Riley, D. P. Inorg. Chem. 2001, 40, 1779-1789. (b) Salvemini, D.;
Riley, D. P.; Cuzzocrea, S. Nat. ReV. Drug DiscoVery 2002, 1, 367-
374. (c) Muscoli, C.; Cuzzocrea, S.; Riley, D. P.; Zweier, J. L.;
Thiemermann, C.; Wang, Z.-Q.; Salvemini, D. Br. J. Pharmacol. 2003,
140, 445-460.
*
To whom correspondence should be addressed. E-mail:
Ivana.Ivanovic@chemie.uni-erlangen.de.
†
Institute for Inorganic Chemistry.
Max-Planck-Institute for Iron Research GmbH.
‡
(2) Riley, D. P. Chem. ReV. 1999, 99, 2573-2587.
7848 Inorganic Chemistry, Vol. 46, No. 19, 2007
10.1021/ic7005056 CCC: $37.00
© 2007 American Chemical Society
Published on Web 08/23/2007

Iron Complex as a Ditopic Receptor for Lithium Salts
Scheme 1
tions and hydrogen-bonding also play an important role in
the host-guest chemistry, in designing specific receptors for
small ions, and in generating a dynamic combinatorial library
(
DCL) as a new strategy for identifying receptors of high
8
affinity for the respective target. Amide-group-containing
receptors are very often used for anion recognition, whereas
9
hydrazone chemistry was involved in creating DCLs and the
10
discovery of receptors for cations, like ammonium, cobalt-
1
1
+ 12
+
(
II), and Li . Receptors for Li are of special interest
+
since Li is the smallest metal cation and has clinical,
pharmacological, and biochemical importance.
1
3
In light of the above-mentioned aspects, it was of
importance to investigate whether the nature of added
electrolyte can exert a significant effect on substitution
equatorial plane (Scheme 1), were systematically studied.4
Mechanistic information on the two-step displacement of the
trans-coordinated solvent molecules by the monodentate
monoanionic thiocyanate ligand were obtained in the case
kinetics of our seven-coordinate Fe(III) complex in CH
3
CN
solution, especially due to the fact that our dapsox2 ligand
possesses two amide and two hydrazone functional groups
at the same time. Thus, from a structural point of view, our
seven-coordinate iron complex can potentially be a ditopic
receptor, capable of binding anions and small cations
simultaneously.
-
4
b,c
2
of H O, MeOH, EtOH, and DMSO as axial ligands. The
latter solvent was selected because kinetic experiments with
the superoxide radical anion, especially with more concen-
trated solutions (in the millimolar range), require the use of
aprotic solvents and the mixing of potassium superoxide
solutions with aqueous or alcoholic solutions of the studied
complexes for the kinetic measurements at ambient and low
7
Herein we report the results from electrochemical, Li and
1
9
F NMR, and DFT studies of the host-guest interactions
+
temperatures, respectively. In the DMSO/CH
3
CN (29:71)
3 2
between the seven-coordinate [Fe(dapsox)(CH CN) ] com-
mixture, it is possible to perform low-temperature kinetic
measurements down to -53 °C. In this respect, it was
important to understand the substitution behavior of [Fe-
plex and different lithium salts in CH
3
CN. The kinetic and
5
spectrophotometric studies on the substitution behavior of
this PBP iron(III) complex in the presence and absence of
lithium ions in acetonitrile solutions have also been reported.
Since the seven-coordinate pentagonal-bipyramidal 3d metal
complexes with planar pentadentate ligands are known to
(
dapsox)(H
In general, kinetic measurements should be performed at
constant ionic strength, requiring the use of so-called
innocent” salts as supporting electrolytes, which do not
2 2 4 3
O) ]ClO in the CH CN solution as well.
1
“
be synthetic SOD enzymes, simulation of an enzymatic
coordinate to the metal center in solution. Because of the
good solubility, tetraalkylammonium and some lithium salts
are suitable as supporting electrolytes in aprotic solvents like
active site, which could bind small metal cations and in that
way enhance binding of superoxide, would be of importance.
Experimental Section
3
DMSO and CH CN. However, added salts can exert strong
effects on the rates of reactions between charged species in
solution, affecting solvation of inorganic complexes. The
Materials. The solvents acetonitrile (Roth, dry, g99.9%, e0.001%
6
H
2
O), DMSO (Acros Organics, g99.9%, e0.1% H
2
O), and
interaction via hydrogen bonds should also be taken into
consideration, especially in aprotic solvents where these
interactions are more prominent. The electrostatic interac-
methanol (Roth, g99.9%, e0.005% H O), as well as the salts
2
NaSCN (Aldrich, 96%), LiOTf (Aldrich, 96%), LiClO
4
(Aldrich,
7
98%), LiBF (Aldrich, 98%), LiPF (Acros Organics, 98%), NaClO
4
6
4
(
3) Andjelkovic, K.; Bacchi, A.; Pelizzi, G.; Jeremic, D.; Ivanovi c´ -
Burmazovi c´ , I. J. Coord. Chem. 2002, 55, 1385-1392.
(8) Corbett, P. T.; Leclaire, J.; Vial, L.; West, K. R.; Wietor, J.-L.; Sanders,
J. K. M.; Otto, S. Chem. ReV. 2006, 106, 3652-3711.
(
4) (a) Sarauli, D.; Meier, R.; Liu, G.-F.; Ivanovi c´ -Burmazovi c´ , I.; van
Eldik, R. Inorg. Chem. 2005, 44, 7624-7633. (b) Ivanovi c´ -Burma-
zovi c´ , I.; Hamza, M. S. A.; van Eldik, R. Inorg. Chem. 2002, 41,
(9) (a) Evans, A. J.; Beer, P. D. Dalton Trans. 2003, 4451-4456. (b)
Mahoney, J. M.; Beatty, A. M.; Smith, B. D. Inorg. Chem. 2004, 43,
7617-7621. (c) Mahoney, J. M.; Stucker, K. A.; Jiang, H.; Carmichael,
I.; Brinkmann, N. R.; Beatty, A. M.; Noll, B. C.; Smith, B. D. J. Am.
Chem. Soc. 2005, 127, 2922-2928. (d) Lankshear, M. D.; Cowley,
A. R.; Beer, P. D. Chem. Comm. 2006, 612-614.
5
150-5161. (c) Ivanovi c´ -Burmazovi c´ , I.; Hamza, M. S. A.; van Eldik,
R. Inorg. Chem. 2006, 45, 1575-1584.
(
(
(
5) (a) McClune, G. J. Iron catalyzed superoxide dismutation; University
of Michigan: Ann Arbor, MI, 1979, From: Diss. Abstr. Int. B 40,
(10) Cousins, G. R. L.; Furlan, R. L. E.; Ng, Y.-F.; Redman, J. E.; Sanders,
J. K. M. Angew. Chem., Int. Ed. 2001, 40, 423-428.
2175. (b) D u¨ rr, K.; Macpherson, B. P.; Warratz, R.; Hampel, F.;
Tuczek, F.; Helmreich, M.; Jux, N.; Ivanovi c´ -Burmazovi c´ , I. J. Am.
Chem. Soc. 2007, 129, 4217-4228.
(11) Guo, D.; Han, G.; Duan, C.-y.; Pang, K.-l.; Meng, Q.-j. Chem. Comm.
2002, 1096-1097.
6) (a) Grigoriev, V. A.; Cheng, D.; Hill, C. L.; Weinstock, I. A. J. Am.
Chem. Soc. 2001, 123, 5292-5307. (b) Curtis, J. C.; Inagaki, M.;
Chun, S. J.; Eskandari, V.; Luo, X.; Pan, Z. N.; Sankararaman, U.;
Pengra, G. E.; Zhou, J.; Hailey, P.; Laurent, J.; Utalan, D. Chem. Phys.
006, 326, 43-53.
7) (a) Beer, P. D.; Gale, P. A. Angew. Chem., Int. Ed. 2001, 40, 486-
16. (b) Bu, J.; Lilienthal, N. D.; Woods, J. E.; Nohrden, C. E.; Hoang,
K. T.; Truong, D.; Smith, D. K. J. Am. Chem. Soc. 2005, 127, 6423-
429. (c) Prins, L. J.; Reinhoudt, D. N.; Timmerman, P. Angew. Chem.,
(12) Furlan, R. L. E.; Ng, Y.-F.; Otto, S.; Sanders, J. K. M. J. Am. Chem.
Soc. 2001, 123, 8876-8877.
(13) (a) Christian, G. D. Sensors 2002, 2, 432-435. (b) Christian, G. D. J.
Pharm. Biomed. Anal. 1996, 14, 899-908. (c) Srinivasan, C.; Toon,
J.; Amari, L.; Abukhdeir, A. M.; Hamm, H.; Geraldes, C. F. G. C.;
Ho, Y.-K.; Mota de Freitas, D. J. Inorg. Biochem. 2004, 98, 691-
701. (d) Layden, B. T.; Abukhdeir, A. M.; Malarkey, C.; Oriti, L. A.;
Salah, W.; Stigler, C.; Geraldes, C. F. G. C.; Mota de Freitas, D.
Biochim. Biophys. Acta 2005, 1741, 339-349. (e) Sastre, E.; Nicolay,
A.; Bruguerolle, B.; Portugal, H. Life Sci. 2005, 77, 758-767. (f)
Parker, M. S.; Parker, S. L.; Kane, J. K. Regul. Pept. 2004, 118, 67-
74.
2
5
6
Int. Ed. 2001, 40, 2382-2426. (d) Rudner, M. S.; Jeremic, S.;
Petterson, K. A.; Kent, D. R.; Brown, K. A.; Drake, M. D.; Goddard,
W. A., III; Roberts, J. D. J. Phys. Chem. A 2005, 109, 9076-9082.
Inorganic Chemistry, Vol. 46, No. 19, 2007 7849

Sarauli et al.
(
(
Acros Organics, 99+%), and all tetrabutylammonium (TBA) salts
Fluka, g99%, puriss electrochemical grade) were used as received
versus a Ag/AgCl electrode. The use of ferrocene as internal
standard is not recommended, since its redox potential depends very
without further purification. The water content in solutions was
much on the concentration of lithium salts present in the CH
solution (∆E ) 30 mV in the presence of 16 mM LiBF and ∆E
) 300 mV in the presence of 16 mM LiPF , Figure S1).
Nuclear Magnetic Resonance Spectrometry. The Li NMR
measurements were conducted on a Bruker Avance DRX 400WB
spectrometer using a 5 mm broadband probe. 19F NMR spectra were
recorded on a Bruker Avance DPX-300 spectrometer at 282.4 Hz.
The solutions to be studied were placed in NMR tubes, and the
3
CN
determined by Karl Fischer titrations. [Fe(dapsox)(H
2
O)
2
]ClO
4
was
4
prepared and characterized previously.3 The [Fe(dapsox)(NCS)
,14
-
2
]
6
7
species was prepared in situ by adding a minimum 100-fold excess
of NaSCN to the acetonitrile solution of [Fe(dapsox)(H O) ]ClO
2
2
4
.
Kinetic Measurements. All data were obtained by recording
time-resolved UV/vis spectra using a modified Bio-Logic stopped-
flow module µSFM-20 combined with a Huber CC90 thermostat
(
(0.1°) and equipped with a J & M TIDAS high-speed diode-
array spectrometer with combined deuterium and tungsten lamps
200-1015 nm wavelength range). Data were analyzed using the
2 2 3
solution of external standard (CF Br in CD CN) in a concentric
capillary was used for the field/frequency deuterium lock. All
(
measurements were run at room temperature. The chemical shifts
7
integrated Bio-Kine software, version 4.23, and the Specfit/32
program. At least 10 kinetic runs were recorded under all conditions,
and the reported rate constants represent the mean values. All kinetic
measurements were carried out under pseudo-first-order conditions,
i.e., the substituent concentration was in a large excess (concentra-
were related to CF
2
Br
2
(δ
F
) 7 ppm vs CFCl
3
). Li NMR spectra
were recorded at 155.5 MHz with a spectral width of 620 Hz and
an acquisition time of 13 s. All systems were referenced to the
+
free Li salt (LiOTf) in DMF (external standard). Measurements
were run at room temperature.
7Li T1 relaxation measurements were performed using the
-5
tions of iron complex were 5 × 10 M). The reactions were studied
*
*
at a constant ionic strength (0.2 M). Values of ∆H and ∆S were
calculated from the slopes and intercepts of plots of ln(k/T) vs 1/T,
respectively.
inversion recovery pulse sequence (D , 180°; τ, 90°). At least 15 τ
1
values were used for each spin-lattice relaxation time (T1)
determination, and an interpulse delay of at least 10 times the T1
Spectrophotometric Titrations. UV/vis spectra were recorded
on a Hewlett-Packard 8542A spectrophotometer at 25 °C. Solutions
value was used before repeating the pulse sequence. The observed
+
T
1
values at different Li concentrations were used to calculate the
+
-
-5
III
of [Fe(dapsox)(CH
1
3
CN)
× 10 M) in acetonitrile were placed in a 1.0 cm path length
cuvette in the thermostated cell block of a spectrophotometer for
0-15 min. These solutions were titrated by the addition of small
aliquots of concentrated stock solutions of substrates (NaSCN,
LiOTf, LiBF , and LiPF ) in CH CN. The spectra were recorded
after equilibration for 10 min.
2
] and [Fe(dapsox)(NCS)
2
] (5 × 10 and
apparent binding constants K to the seven-coordinate [Fe(dapsox)-
(CH
-
4
+
3
CN)
2
]
complex. The relaxation measurements were ac-
companied by a 2-10% uncertainty.
1
Quantum Chemical Methods. We performed B3LYP/
LANL2DZp hybrid density functional calculations, i.e., with
pseudo-potentials on the heavy elements and the valence basis set
augmented with polarization functions.15 The resulting structures
were characterized as minima by computation of vibrational
frequencies, and the wave functions were tested for stability. The
GAUSSIAN suite of programs was used.16
4
6
3
Electrochemical Measurements. Cyclic (CV) and differential
pulse (DPV) voltammetry measurements were performed with an
Autolab PGSTAT 30 device (Eco Chemie). All measurements were
2
conducted under N in a jacketed, one-compartment cell with the
2
Results and Discussion
Au disk working electrode (geometric area: 0.07 cm ) (Metrohm),
a platinum wire counter electrode (Metrohm), and a Ag wire
pseudo-reference electrode. The working electrode surface was
polished with 0.05-µm alumina, sonicated in water, and air-dried
Electrochemical Studies. Cyclic voltammetry measure-
ments of the seven-coordinate [Fe(dapsox)(CH
+
3
CN)
2
] com-
-
-
-
immediately before use. Various NBu
4
6 4 4
A (A ) PF , BF , ClO ,
(
15) For recent papers that include the same type of calculations done,
-
OTf ) salts (0.1 M) were used as electrolytes. The sweep rate was
see: (a) Dees, A.; Zahl, A.; Puchta, R.; van Eikema Hommes, N. J.
R.; Heinemann, F. W.; Ivanovi c´ -Burmazovi c´ , I. Inorg. Chem. 2007,
46, 2459-2470. (b) Summa, N.; Maigut, J.; Puchta, R.; van Eldik, R.
Inorg. Chem. 2007, 46, 2094-2104. (c) Summa, N.; Schiessl, W.;
Puchta, R.; van Eikema Hommes, N.; van Eldik, R. Inorg. Chem. 2006,
5
0 mV/s. No IR compensation was applied. The stoichiometry of
each electron-transfer was established by wave-height comparisons
with known one-electron redox processes. O was removed from
the electrolyte solution by bubbling N through the solvent for
several minutes prior to making the measurement. A N atmosphere
2
2
4
5, 2948-2959. (d) Puchta, R.; van Eikema Hommes, N.; Meier, R.;
2
van Eldik, R. Dalton Trans. 2006, 3392-3395. (e) Puchta, R.; Meier,
R.; van Eikema Hommes, N. J. R.; van Eldik, R. Eur. J. Inorg. Chem.
was continuously maintained above the solution while the experi-
ments were in progress. Temperature was controlled by using a
circulating water bath to run 25 °C water through the outer cell
jacket.
2
006, 4063-4067. (f) Illner, P.; Zahl, A.; Puchta, R.; van Eikema
Hommes, N.; Wasserscheid, P.; van Eldik, R. J. Organomet. Chem.
2005, 690, 3567-3576 and references therein.
(
16) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin,
K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone,
V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G.
A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.;
Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai,
H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.;
Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R.
E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J.
W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.;
Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.;
Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari,
K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.;
Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.;
Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.;
Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.;
Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A.
Gaussian 03, revision C.02; Gaussian, Inc.: Wallingford, CT, 2004.
In order to improve solubility of the complex, solutions were
prepared in dry CH
To determine the equilibrium associated with the redox-dependent
binding, titration measurements were run in 0.1 M NBu
3
CN containing a maximum of 0.1% DMSO.
4
A
electrolyte solutions, using the seven-coordinate Fe(III) complex
as the receptor and different lithium and/or tetrabutylammonium
salts as the substrates. Experimentally, aliquots of concentrated
lithium/tetrabutylammonium salt solutions were added incrementally
to a 20 mL sample of the 1 mM iron complex, and cyclic
voltammograms were run after each addition. Potentials are cited
(
14) Ivanovi c´ -Burmazovi c´ , I.; Andjelkovic, K. AdV. Inorg. Chem. 2004,
55, 315-360.
7850 Inorganic Chemistry, Vol. 46, No. 19, 2007

Iron Complex as a Ditopic Receptor for Lithium Salts
Scheme 2
plex in the presence of different tetrabutylammonium
+
(NBu
4
3
) and lithium salts in CH CN afford a simple and
effective way for estimating the communication between the
redox-active iron center and binding sites for studied cation-
anion pairs as potential guests. If present, an interaction
between the seven-coordinate iron complex in both its
III
II
oxidation states (Fe and Fe ) and the ions in its substrate
binding sites would result in a shift of the [Fe(dapsox)-
0
/+
(CH
3
CN)
2
]
couple redox potential upon increase of the
17
salt concentration. For the electrochemical guest binding
titrations, the equilibria can be represented by a redox square
(Scheme 2), and the interaction can be quantified by
determining the apparent guest binding constant according
to eq 1,6
a,17,18
where ∆E is a shift in the host redox potential
II
III
as a function of guest concentration and K and K
correspond to the 1:1 host/guest complex in the divalent and
trivalent forms of the receptor, respectively.
II
RT
F
1 + K [guest]
∆
E )
ln
(1)
[
III
]
1
+ K [guest]
However, no increase or decrease in the half-wave
potential (E1/2) of the [Fe(dapsox)(CH
0
/+
3
CN)
2
]
couple (1
+
-
-
mM) was observed when concentrations of NBu
4
A (A
) were varied from 1 to 50
CN solution (Figure S2). Although in the
crystal structure of [Fe(dapsox)(H O) ]ClO the charge-
assisted hydrogen-bonding interactions between the twee-
-
-
-
-
)
PF
6 4 4 3 3
, BF , ClO , CF SO
+
-
-
Figure 1. Plot of E1/2 vs lithium salt concentration [Li A ] (A ) OTf ,
mM in the CH
3
-
-
+
ClO4 , BF4 ); [Fe(dapsox)(CH3CN)2] ) 1 mM, 25 °C, E1/2 vs Ag/AgCl
2
2
4
+
-
electrode. The corresponding NBu4 A salts (0.1 M) were used as
supporting electrolytes. Solid curves represent nonlinear least-squares fitting
of the data to eq 1.
+
-
2 2 4
zerlike [Fe(dapsox)(H O) ] cation and the ClO anion have
3
0/+
been observed, this interaction, based on the performed
electrochemical titrations, seems to be not strong enough to
be observed in solution.
(CH CN) ] in CH CN. The addition of LiOTf, LiClO ,
3
2
3
4
or LiBF to the electrolyte solution of [Fe(dapsox)(CH -
4
3
+
CN) ] (ionic strength of 0.1 M was kept constant) led to a
progressive positive shift in the E1/2 (one-wave behavior).
2
Lithium salts differ from tetrabutylammonium salts in that
they increase the solubility of [Fe(dapsox)(H
significantly shift the half-wave potential of [Fe(dapsox)-
2
O)
2
]ClO
4
and
The maximal shift in the half-wave potential is reached in
the presence of an excess of guest species (Figure 1, Table
S1). Cyclic and differential pulse voltammograms obtained
(17) Butler, J. N. Ionic Equilibrium: Solubility and pH Calculations;
Wiley: New York, 1998; p 474.
during the electrochemical titrations of [Fe(dapsox)(CH
3
-
+
(18) Bernhardt, P. V.; Moore, E. G. Aust. J. Chem. 2003, 56, 239-258.
CN)
2
] (Figure 2) show the presence of a single reversible
Inorganic Chemistry, Vol. 46, No. 19, 2007 7851

Sarauli et al.
Figure 3. Differential pulse voltammograms representing electrochemical
response of [Fe(dapsox)(CH3CN)2]0 (1 mM) upon addition of increasing
/+
+
amounts of LiPF6. The wave (a) is assigned to the [Fe(dapsox)(CH3CN)2]
in the absence of lithium salt. The waves (b) and (c) appear in acetonitrile
solution of complex after addition of LiPF6 (up to 0.0019 M). The wave
(d) (inset) appears after further addition of LiPF6 (0.0019-0.019 M).
Figure 2. Typical electrochemical response of [Fe(dapsox)(CH3CN)2]0/+
1 mM) upon addition of increasing amounts of (a) LiOTf (differential pulse
(
voltammograms) and (b) LiBF4 (cyclic voltammograms) using correspond-
ing NBu4A salts (0.1 M) as electrolytes.
redox couple that shifts anodically with added lithium salt,
indicating a weak binding.18,19 The difference in diffusion
+
coefficients of [Fe(dapsox)(CH
3
CN)
2
] and its lithium salt
adduct causes the difference in the current intensity observed
in Figure 2.
Figure 4. Typical 19F NMR spectra of the CH3CN solution, containing
II
4
The values of K are quite high (∼10 ) and are more than
-
0
.2 M NBu4PF6 and 0.2 M LiPF6. Signals (a) are assigned to PF6 . Signals
III
-
1
order of magnitude greater than K , which is consistent
(b) and (c) belong to PO2F2 and HF, respectively.
with the observed positive shift of E1/2, confirming that the
reduced state of the receptor is stabilized upon complexation
Consequently, a stronger anion binding results in the smaller
shift of E1/2, since, opposite from the Li coordination, it
increases the electron density around the metal center.
We have also performed the electrochemical titration with
+
with lithium salts.1
9-20
The K values are less sensitive to
II
III
the nature of the counteranion than the K values, suggesting
that the more prominent interaction with the anion exists in
the case of the Fe(III) form of the complex. The somewhat
NaClO
4 3 4
in CH CN and with LiBF in MeOH in order to
test the cation and solvent variation effects (Table 1). In both
higher apparent binding constants are observed for LiBF
where the strongest hydrogen-binding interactions between
the anion and amide -NH groups are to be expected.
4
,
III
III
-1
-1
cases, the K value was too small (K < 10 M ) to be
determined,
1
9b,c
showing that the interactions between the
2
Fe(III) form of the complex and guest salts are very weak
under those conditions. Comparison of the apparent binding
(
19) (a) Beer, P. D.; Gale, P. A.; Chen, G. Z. J. Chem. Soc., Dalton Trans.
II
-1
constants K for LiClO
(
4
(4004 ( 402 M ) and NaClO
4
1999, 1897-1910. (b) Bernhardt, P. V.; Hayes, E. J. Inorg. Chem.
2002, 41, 2892-2902. (c) Bernhardt, P. V.; Hayes, E. J. Inorg. Chem.
2003, 42, 1371-1377.
-
1
2690 ( 400 M ) in CH CN (Table 1) leads to the
conclusion that the complex exhibits modest Li selectivity.
The similar complexation selectivity was observed for several
3
+
(20) Moutet, J. C.; Saint-Aman, E.; Ungureanu, E. M.; Arion, V.; Gerbeleu,
N.; Revenco, M. Electrochim. Acta 2001, 46, 2733-2740.
7852 Inorganic Chemistry, Vol. 46, No. 19, 2007

Iron Complex as a Ditopic Receptor for Lithium Salts
1
9
Figure 5. F NMR spectra of the CH3CN solutions at 25 °C, containing constant iron complex concentration (0.5 mM) and increasing concentrations of
-
LiPF6. Signals (a) and (b) are assigned to PF6 and [Fe(dapsox)(CH3CN)(PO2F2)], respectively.
+
-
II
+
Table 1. Apparent Binding Constants for the Complexation of Li A Salts in CH3CN by [Fe(dapsox)(CH3CN)2] (K ) and [Fe(dapsox)(CH3CN)2]
III
(K )
solvent
CH3CN
CH3OH
apparent binding constants, M-
1
Li CF3SO3
+
-
Li ClO4
+
-
Li BF4
+
-
Na ClO4
+
-
Li BF4
+
-
II
a
K (electrochem)
3754 ( 418
90 ( 11
4004 ( 402
309 ( 28
6616 ( 1812
458 ( 102
2690 ( 400
<0.1b
2529 ( 301
III
a
b
K
K
K
K
(electrochem)
<0.1
III
III
III
7
c
( Li NMR)
203 ( 32
49 ( 10
d
(kinetics)
c
(UV-vis spectra)
165 ( 15
a
At 25 °C in 0.1 M NBu4A. b Too small to determine. c At 25 °C with no supporting electrolyte. d At -15 °C in 0.2 M NBu4A.
2
1
redox-active ferrocene crown ethers. Polarity of the solvent
also plays an important role.7 Therefore, in more polar
constant J(P,F) ) 707 Hz, is observed (Figure 4a), which is
a,22
- 23
19
typical for PF
CH CN consists of two doublets with chemical shifts (a) δ
) -69.8 [J(P,F) ) 707 Hz] and (b) δ ) -81.6 [J(P,F) )
959 Hz], as well as the singlet (c) with δ
6
.
6
However, the F spectrum of LiPF in
MeOH, the interaction of the complex with LiBF
than in CH CN, resulting in the smaller apparent binding
constants K and K (Table 1).
Solution Behavior in the Presence of LiPF
4
is weaker
3
F
3
F
II
III
F
) -148.6 (Figure
-
6
. The
4b). The major low-field doublet (a) is assigned to PF
6
,
electrochemical titration was also performed with LiPF
6
.
whereas the doublet (b) is unambiguously assigned to the
-
-
However, lithium-catalyzed hydrolysis of PF
6
causes the
PO
2
-
F
2
ion, which is the product of partial hydrolysis of
+
appearance of additional fluorine-containing anionic species
PF
6
in CH
3
CN, accelerated by the presence of Li ions in
23
23
and HF, which makes the solution chemistry more complex.
solution. The singlet (c) is characteristic for HF, which is
also consistent with the hydrolysis processes observed
in the presence of Li . When the complex (1 mM) was
Immediately upon addition of a small aliquot of 0.4 M LiPF
6
+
+ 23
in CH
3
CN (complex/Li ) 4:1 molar ratio), two new redox
waves were observed (Figures 3 (DPV) and S3 (CV)): the
intense one (b) with negatively shifted E1/2 (∆E1/2 ) -130
mV) and the lower intensity one (c) with positively shifted
added to the solutions containing 0.1 M NBu
different concentrations of LiPF , the downfield shift of the
doublet (b) δ ) -78.6 [J(P,F) ) 946 Hz] and the
appearance of a new one (d), δ ) -72.1 [J(P,F) ) 920
Hz] could be observed (Figure 5). The new doublet (d)
4 6
PF and
6
F
E1/2 (∆E1/2 ) +132 mV) relative to the original redox couple.
F
The appearance of two redox waves and the negative shift
in redox potential upon addition of lithium salt are unique
observations. By increasing the lithium concentration, the
intensity of wave (b) decreases and that of (c) increases
2
-
23
corresponds to the PO
3
F
anion.
-
-
2-
The observed anionic species (A ) PO
can act as the complexing agents, resulting in the negatively
shifted redox wave (b) (Figure 3). The positively shifted
2 2 3
F and PO F )
2
4
(Figures 3 and S3).
To obtain more information about the species present in
(
24) (a) Akbayeva, D. N.; Di Vaira, M.; Seniori Costantini, S.; Peruzzini,
M.; Stoppioni, P. Dalton Trans. 2006, 389-395. (b) Fernandez-Galan,
R.; Manzano, B. R.; Otero, A.; Lanfranchi, M.; Pellinghelli, M. A.
Inorg. Chem. 1994, 33, 2309-2312. (c) Fernandez-Galan, R.; Man-
zano, B. R.; Otero, A. J. Organomet. Chem. 1999, 577, 271-282. (d)
White, C.; Thompson, S. J.; Maitlis, P. M. J. Organomet. Chem. 1977,
134, 319-325. (e) Wimmer, F. L.; Snow, M. R. Aust. J. Chem. 1978,
31, 267-278. (f) Mildvan, A. S.; Leigh, J. S.; Cohn, M. Biochemistry
1967, 6, 1805-1818. (g) Nowak, T. Arch. Biochem. Biophys. 1978,
186, 343-350.
19
solution, the F NMR measurements were performed. In
1
9
the F NMR spectrum of NBu
4 6
PF in acetonitrile, only one
) -69.8, with the 3 P- F coupling
1
19
1
:1 doublet at δ
F
(
(
(
21) Plenio, H.; Diodone, R. Inorg. Chem. 1995, 34, 3964-3972.
22) Severin, K. Coord. Chem. ReV. 2003, 245, 3-10.
23) Plakhotnyk, A. V.; Ernst, L.; Schmutzler, R. J. Fluorine Chem. 2005,
126, 27-31.
Inorganic Chemistry, Vol. 46, No. 19, 2007 7853

Sarauli et al.
wave (c) corresponds to the lithium salt adduct of [Fe(dapsox)-
CH CN)(A)]. This two-wave behavior can be interpreted
(
3
in terms of the high stability constant of the lithium salt
adduct.25 The coordination of negatively charged A , af-
fecting the overall charge of the complex, facilitates stronger
-
+
Li binding. Analogous to Scheme 2 and according to eq
1
8,19a
II III
2,
the K /K quotient (the binding enhancement factor
19a
or the reaction coupling efficiency ) is 52 600, suggesting
the strong coupling between the lithium binding and iron
reduction. It also shows that in the Fe(II) form the negatively
charged [Fe(dapsox)(CH
efficiently.
-
+
3
CN)(A)] adduct binds Li very
II
RT
F
K
∆
E )
ln
(2)
III
K
At higher lithium concentrations, wave (b) vanishes
whereas wave (c) becomes broader and starts to anodically
shift further (wave (d) in Figures 3 and S3). These additional
Figure 6. 7Li NMR Spectra of LiOTf (0.2 M) solution in CH3CN at
25 °C in the presence of (a) 0.5, (b) 1, and (c) 1.5 mM [Fe(dapsox)(CH -
3
+
CN)2] .
processes are most probably related to the further hydrolysis
-
of PF
6
-related anionic species and increase in the HF
concentration, which might result in protonation of the
complex.
7Li NMR Titrations. In order to better visualize the
lithium ion binding to the seven-coordinate iron(III) complex,
7
Li NMR measurements were performed. Different from the
electrochemical measurements, where the effects of host-
guest interactions on the redox active iron center can be
7
observed, Li NMR titrations can demonstrate the complex-
+
ation effect on the Li cation itself. The NMR chemical shift
titration has been widely used as a method for the determi-
6
a,26
nation of binding constants.
The easily observed and
7
sensitive Li isotope has nuclear spin I ) 3/2 and high
abundance (95%), making it an attractive nucleus for NMR
27
titrations. In the case of host-guest complexation reactions,
7
either the concentration of guest ( Li nucleus) or of the host
is varied and the resulting change in the chemical shift is
followed. If the reaction is fast on the NMR time scale, only
one shift will be observed, representing a concentration-
weighted average of the shifts of free and bound species.
Figure 7. 7Li NMR Spectra of (a) 5, (b) 15, (c) 25, (d) 35, and (e) 50
mM LiOTf in the presence of constant iron complex concentration (0.5
mM) in CH3CN. Measurements were run at room temperature.
+
In the first set of measurements, the concentration of Li
(
200 mM LiOTf) was kept constant. Three different probes
around the lithium cation is perturbed upon addition of the
with the iron complex concentration of 0.5, 1.0, and 1.5 mM
were measured. It can be seen in Figure 6 that the lithium
resonance peak broadens and shifts downfield (relative to
the chemical shift of free Li in CH
3 2
Figure 7), with an increase in [Fe(dapsox)(CH CN) ]
3
B
paramagnetic (µeff ) 5.26 µ ) receptor complex.
Another set of data was obtained by keeping the concen-
tration of the iron complex constant (0.5 mM) and varying
+
3
CN with δ ) 5.7, see
+
the concentration of Li between 5 and 50 mM (Figure 7).
+
Increase in the lithium salt concentration causes an upfield
concentration. The downfield resonance shift is indicative
of the complexation, implying that the solvation environment
7
+
shift of the Li resonance toward the position of free Li .
This also provides qualitative evidence for the host-guest
interaction.
(
25) Miller, S. R.; Gustowski, D. A.; Chen, Z. H.; Gokel, G. W.; Echegoyen,
L.; Kaifer, A. E. Anal. Chem. 1988, 60, 2021-2024.
In an attempt to quantitatively determine the apparent
(26) (a) Masiker, M. C.; Mayne, C. L.; Eyring, E. M. Magn. Reson. Chem.
7
lithium binding constant, Li NMR T
1
relaxation measure-
2
006, 44, 220-229. (b) Popov, K.; Lajunen, L. H. J.; Popov, A.;
Ronkkomaki, H.; Hannu-Kuure, M.; Vendilo, A. Inorg. Chem. Comm.
ments were performed. The concentration of the seven-
coordinate iron complex was held constant (0.5 mM),
whereas the concentration of the lithium salts (LiOTf,
2
2
002, 5, 223-225. (c) Shamsipur, M.; Madrakian, T. J. Coord. Chem.
000, 52, 139-149.
(27) (a) Suzuki, M.; Koyama, H.; Noyori, R. Tetrahedron 2004, 60, 1571-
1
579. (b) Mueller, A.; Rehder, D.; Haupt, E. T. K.; Merca, A.; Boegge,
H.; Schmidtmann, M.; Heinze-Brueckner, G. Angew. Chem., Int. Ed.
004, 43, 4466-4470. (c) Reich, H. J.; Borst, J. P.; Dykstra, R. R.;
LiClO
, and LiBF
) was varied between 5 and 50 mM (10-
4
4
+
1
00-fold excess). The concentration ([Li ] ) 500 mM) was
2
Green, P. D. J. Am. Chem. Soc. 1993, 115, 8728-8741.
used in all cases to observe relaxation time at saturation
7854 Inorganic Chemistry, Vol. 46, No. 19, 2007

Iron Complex as a Ditopic Receptor for Lithium Salts
constant was found to be 203 ( 68 M- (Table 1), which is
somewhat higher than the corresponding value obtained from
the electrochemical titration, general in agreement with the
higher ionic strength of the solutions used in electrochemical
experiments.
1
III
Spectrophotometric Determination of K . In an attempt
III
to spectrophotometrically determine the K value, the [Fe-
+
(
dapsox)(CH
3
CN)
2
]
complex was titrated with LiOTf at
25 °C. The absorbances at 460 nm vs [LiOTf] (inset in Figure
S5) were fitted by applying eq 4
III
(
A - A )K [LiOTf]
∞ 0
A ) A +
(4)
x
0
III
1
+ K [LiOTf]
The values of A
100% formation of complex, respectively, and A represents
x
the absorbance at any given lithium triflate concentration.
The observed spectral changes upon addition of LiOTf are
not so prominent (Figure S5); however, the resulting apparent
0
and A
∞
represent absorbance at 0% and
Figure 8. Representative James-Noggle plots for [Fe(dapsox)(CH3CN)2]+
complex in the presence of (9) LiOTf, (red circles) LiBF4, and (green
triangles) LiClO4. The x axis corresponds to the lithium salt concentrations
used in the titrations of seven-coordinate iron complex, and the y axis
represents 1/(Robs - Rfree), where Robs and Rfree are the relaxation rates (or
+
the reciprocals of the relaxation times) for the Li nuclei of the observed
III
-1
binding constant K ) 165 ( 15 M is in excellent
+
sample (T1(obs)) and with saturating Li concentrations (T1(free)), respectively
III
7
agreement with the corresponding K value obtained by -
Li NMR measurements (Table 1), where in both experiments
no additional supporting electrolyte was present in solution
(eq 3).
+
(
T
1(free)). Only one type of Li binding site was assumed to
(experiments under the same ionic strength).
+
be present, as well as monoexponential relaxation of the Li
-
Reaction with SCN in CH
3
CN. The kinetics of the
+
nucleus (Figure S4) and fast exchange of Li between free
+
reaction between [Fe(dapsox)(CH
3 2
CN) ] complex and thio-
and bound states.28 According to these assumptions, the
-
cyanate (SCN ) in acetonitrile was studied under pseudo-
III
apparent binding constants K were calculated from James-
-
first-order conditions (a large excess of SCN ) at -15 °C
and at constant ionic strength (I ) 0.2 M), adjusted with
NBu OTf and LiOTf (see below), respectively.
4
Spectral changes and corresponding kinetic traces (Figure
-
1
+
Noggle plots of ∆R against [Li ] according to eq 3, using
7
13d,28,29
at least five Li T
1
values,
under the experimental
+
conditions where [Li ] . [Fe]
9
) clearly showed two subsequent reaction steps over the
R- ) (Robs - Rfree)-1 ) (K ) {[Fe](R
1
III -1
-
∆
-
bound
whole SCN concentration range. Similar to what we found
in aqueous, methanol, and ethanol solutions, the two reaction
steps represent the substitution of two axially coordinated
Rfree)}- + [Li ]{[Fe](R
1
+
- Rfree)}-1 (3)
bound
[
Li ] and [Fe] are the concentrations of Li+ and iron
+
solvent molecules by SCN . Kinetic traces were fitted to
- 4b,c
complex, respectively. Robs ) 1/T1(obs), Rfree ) 1/T1(free), and
a double-exponential function that resulted in kobs values for
the first and second reaction steps. Plots of kobs vs [SCN ]
(Figure 10) for both reaction steps at -15 °C in the presence
-
R
bound ) 1/T1(bound) are the relaxation rates (or the reciprocals
+
of the relaxation times) for the Li nuclei of the observed
sample (Robs), with saturating Li concentrations (Rfree(LiOTf)
+
of NBu
4
OTf, as well as LiOTf, show a linear increase in
-
1
-1
) ) 7.4 s-1),
-
)
7.6 s , Rfree(LiClO
4
) ) 5.6 s , Rfree(LiBF
4
k
obs with increasing [SCN ] over the whole concentration
+
or with Li bound (Rbound), respectively (Table S2). This
range, with more prominent intercepts in the case of the
-
approach has the inherent approximation of using a binding
first reaction step. The dependence of kobs on [SCN ] for
+
model with a single type of Li binding site, with a single
1 2
both reactions can be expressed by eq 5, where k and k
III
K
and Rbound value.
represent the rate constants for the forward reactions
(from the slopes of the solid lines) and k-1 and k-2 represent
the rate constants for the backward (from the intercepts)
-
1
+
In all three cases, good linear plots of ∆R vs [Li ] were
obtained (Figure 8). By dividing the slopes of the linear fits
solid lines) in Figure 8 by the corresponding intercepts and
multiplying the quotients by 1000, the values of the apparent
(
4
reactions in the presence of NBu OTf. For the reaction
performed in the presence of LiOTf, an expression analogous
to eq 5 can be applied for the first and second substitution,
with the corresponding k1Li, k2Li, k-1Li, and k-2Li rate
constants.
III
-1
binding constant K in M could be derived (Table 1).
However, the values obtained for the intercepts in the case
4 4
of LiClO and LiBF are rather small with the high
uncertainties and, consequently, could not be used for the
quantification of K . In the case of LiOTf, the K binding
-
III
III
kobs(1,2) ) k [SCN ] + k
(5)
1,2
-1,-2
SCN
(
28) Connors, K. A. Binding Constants: The Measurements of Molecular
Complex Stability; Wiley: New York, 1987; p 350.
From these values, the equilibrium constants K
) k
/k-2 for the binding of the first and second
OTf and LiOTf can be
1
/
1
SCN
k
-1 and K
SCN in the presence of NBu
2
) k
2
(29) James, T. L.; Noggle, J. H. Proc. Natl. Acad. Sci. U.S.A. 1969, 62,
-
644-649.
4
Inorganic Chemistry, Vol. 46, No. 19, 2007 7855

Sarauli et al.
Figure 9. Spectral changes during the reaction of [Fe(dapsox)(CH3CN)2]+
with NaSCN in the presence of (a) NBu4OTf and (b) LiOTf ([Fe(III)] ) 5
-
Figure 10. Plots of kobs vs [SCN ] for the first (a) and second (b) reaction
steps in the absence (9) and in the presence (red circles) of LiOTf.
-5
-
-5
×
10 M, [SCN ] ) 0.005 M, I ) 0.2 M, -15.0 °C). Inset: corresponding
Experimental conditions: ([Fe(III)] ) 5 × 10 M, I ) 0.2 M, -15.0 °C).
kinetic traces at 448 nm and double-exponential fits (solid lines).
Scheme 3
calculated. The reaction scheme and corresponding rates and
equilibrium constants are summarized in Scheme 3 and Table
2
, respectively.
-
Equilibrium Studies for Binding of SCN . The equi-
SCN
SCN
librium constants K
spectrophotometric titrations of [Fe(dapsox)(CH
NaSCN in CH
1
and K
2
were determined by
+
3
CN)
2
] with
3
CN solution at 25 °C. The spectral changes
for the reactions with and without LiOTf (I ) 0.2 M) are
presented in Figure S6. For determination of equilibrium
-
constants, the absorbances at 448 nm vs [SCN ] (insets in
Figure S6) were fitted, in both cases, for the first and the
second equilibrium together to eq 6.
SCN
-
SCN SCN
1
- 2
A + A K
[SCN ] + A K
K2 [SCN ]
0
1
1
2
Ax )
(6)
SCN
-
SCN SCN
- 2
1
+ K1 [SCN ] + K1 K2 [SCN ]
The values of A
of complex, A
dapsox)(CH
respectively, and A
thiocyanate concentration [SCN ]. The average values of
0
represent absorbance at 0% formation
1
and A
2
ascribe absorbances of the [Fe-
-
(
3
CN)(NCS)] and [Fe(dapsox)(NCS)
2
] species,
nm according to eq 6, as well as independently over the 200-
700 nm wavelength range, with Specfit/32 global analysis
are summarized in Table 2.
x
represents the absorbance at any given
-
SCN
SCN
K
1
and K
2
obtained by fitting titration curves at 448
7856 Inorganic Chemistry, Vol. 46, No. 19, 2007

Iron Complex as a Ditopic Receptor for Lithium Salts
Table 2. Kinetic Results for the Reactions of NaSCN with [Fe(dapsox)(CH3CN)2]+ at -15 °C and 0.2 M Ionic Strength in CH3CN
0.2 M (NBu4OTf)
0.2 M (NBu4OTf/LiOTf)
k1, M- s-1
k-1, s
1
411 ( 14
k1Li, M-1 s-1
k-1Li, s
1605 ( 51
2.8 ( 0.3a
574 ( 64
727 ( 17b
-
1
-1
0.79 ( 0.08
K1SCN, M-1
520 ( 56
a
b
2
37 ( 53
*
-1
∆
H1 , kJ mol
9 ( 2
∆
S1 , J mol K-1
*
-1
-159 ( 6
52 ( 1
-
1
-1
k2Li, M-1 s-1
k2, M
s
106 ( 2
-
1
-1
k-2, s
0.033 ( 0.007
k-2Li, s
0.056 ( 0.004
1893 ( 140a
1900 ( 210b
K2 , M-1
SCN
1575 ( 336a
1
326 ( 468b
*
-1
∆
H2 , kJ mol
4 ( 1
-195 ( 3
∆
S2 , J mol K-1
*
-1
III
-1
49 ( 10c
K , M
III
, M-1
c
K
152 ( 24
SCN
III
, M-1
775 ( 54d
K2
(SCN)
a
Calculated as K1 ) k1/k-1, K2 ) k2/k-2. b Results from nonlinear fitting of the spectrophotometric titration curves of [Fe(dapsox)(CH3CN)2] and
SCN
SCN
c
{
1
[Fe(dapsox)(CH3CN)2]‚LiOTf}with NaSCN at 25 °C (eq 6, Figure S6). Determined by nonlinear fitting of saturation kinetic curves (eqs 7 and 8, Figure
1). Results from nonlinear fitting of the spectrophotometric titration curves of [Fe(dapsox)(NCS)2] with LiOTf at 25 °C (eq 4, Figure S8).
d
-
Temperature-Dependent Kinetic Measurements. The
electrochemical and NMR titrations of [Fe(dapsox)(CH
3
-
+
CN)
2
] with LiOTf (see above) have shown that there is a
host-guest interaction between the complex and lithium salt
and that in the presence of 0.2 M LiOTf the predominant
+
3 2
species in solution is the {[Fe(dapsox)(CH CN) ]‚LiOTf}
adduct. At the same time, our kinetic and thermodynamic
-
data for the reactions with SCN show (Figure 10 and Table
2) that in the presence of 0.2 M LiOTf both reaction steps
proceed with higher rate constants (k1Li and k2Li) and the
SCNand K
SCN
)
corresponding apparent binding constants (K
1
2
are also somewhat higher than in the case when instead of
LiOTf, 0.2 M NBu
suggests that the {[Fe(dapsox)(CH
has a higher reactivity toward the SCN anion and that its
mono- and bis(thiocyanato) species have higher thermody-
namic stability. To clarify mechanistically this phenomenon,
4
OTf was present in solution. This clearly
+
3
CN)
2
]‚LiOTf} adduct
-
Figure 11. Dependences of kobs(1) and kobs(2) on the lithium concentration.
-
5
-
Experimental conditions: ([Fe(III)] ) 5 × 10 M, [SCN ] ) 0.005 M, I
we have performed the temperature-dependent kinetic mea-
)
0.2 M, -15.0 °C). Solid lines represent nonlinear fitting curves according
+
surements of the reaction of [Fe(dapsox)(CH
3
CN)
2
] with
to eqs 7 and 8.
-
SCN and 0.2 M NBu
4
OTf as supporting electrolyte. Plots
-
of kobs vs [SCN ] for the first and second reaction steps as
a function of temperature, and the corresponding Eyring plots
are presented in Figures S7. The obtained activation param-
eters ∆H and ∆S are also included in Table 2. The very
negative ∆S values suggest the associative nature of the
-
of SCN . The coordinated lithium cation close to the iron
center decreases the electron density around the metal center
and consequently increases its electrophility.
*
*
*
Lithium-Dependent Kinetic Measurements. In order to
better visualize the effect of the lithium cation on the kinetic
behavior of our complex, we performed another set of
rate-determining process for both substitution steps. In the
+
case of previously studied reactions of [Fe(dapsox)(S)
2
] (S
b,c
-
)
H
2
O, MeOH, EtOH, DMSO)4 with SCN in aqueous,
+
measurements. The reaction was followed varying [Li ] and
alcoholic, and DMSO solutions, respectively, an associative
-
keeping [SCN ] and ionic strength constant. Dependences
*
interchange (I
a
) mechanism (negative ∆S values) has been
of kobs(1) and kobs(2) on the lithium concentration are depicted
in Figure 11, showing an increase of the observed rate
constants for both reaction steps with increasing lithium
concentration. Significant curvatures are in agreement with
observed for the first substitution step. The second substitu-
4
b,c
tion step had the general dissociative character (D or I
d
).
+
However, [Fe(dapsox)(CH
3
CN)
2
] shows a stronger associa-
tive character for both substitution steps, as well as the higher
reactivity, with k and k (extrapolated to 25 °C) 2 and 1
order of magnitude higher than the corresponding values for
the equilibrium between two different reactive species,
1
2
+
[
Fe(dapsox)(CH
3
CN)
+
2
] (“lithium-free”) and {[Fe(dapsox)-
+
4b,c
(CH CN) ]‚LiOTf} (“lithium-bound”) for the first, and
the other studied [Fe(dapsox)(S)
2
] complexes, respectively.
3
2
[Fe(dapsox)(NCS)] (“lithium-free”) and {[Fe(dapsox)(NCS)]‚
Increased reactivity and product stability in the case of the
+
{
[Fe(dapsox)(CH
3
CN)
2
]‚LiOTf} adduct can also be ex-
LiOTf} (“lithium-bound”) for the second reaction step.
plained in terms of an associative mechanism for the binding
According to Scheme 3, the first and second reaction
Inorganic Chemistry, Vol. 46, No. 19, 2007 7857

Sarauli et al.
Figure 12. Spectra of the [Fe(dapsox)(NCS)2]- solutions (5 × 10-5 M) measured after addition of different lithium salts (0.2 M) in CH3CN and the change
in color of corresponding product solutions.
3 2
CN)
]+ complex.
the LiOTf binding to the [Fe(dapsox)(CH
By analogy, kobs(2) is the measured rate constant for the
30
second reaction step, k
2
, k2Li, and k-2Li are the corresponding
rate constants for the “lithium-free” and “lithium-bound”
III
SCN
forms as reactive species (Scheme 3), and K
is the
equilibrium constant for the LiOTf binding to the mono-
III
(
thiocyanato) [Fe(dapsox)(NCS)] adduct. Although the K
III
and K
values are not very high, suggesting that the
SCN
binding of LiOTf to the iron complex in acetonitrile is not
very efficient, the interaction is strong enough to affect the
III
kinetic behavior. The K equilibrium constant corresponds
to the constants obtained in electrochemical, and NMR
experiments, respectively. The differences in the obtained
values (Table 1) arise from the differences in the experi-
mental conditions (temperature and ionic strength).
Spectrophotometric Observations. Different from
+
Figure 13. Spectral changes upon addition of LiBF4 to the 5 × 10-5
the [Fe(dapsox)(CH
3
-
CN)
2
]
complex its bis(thiocyanato)
M
solution of [Fe(dapsox)(NCS)2]. Inset: the change in absorbance at 474
nm plotted vs [LiBF4]; solid line represents nonlinear fitting curve (according
to eq 4).
2
[Fe(dapsox)(NCS) ] adduct with the strong charge-transfer
band at 440 nm shows more prominent spectral changes upon
addition of LiOTf, with a shift of the absorption maximum
toward longer wavelengths (∼460 nm) as the LiOTf con-
centration is increased (Figure S8). This allowed us to
determine the apparent lithium binding constant to the
steps can be expressed using the somewhat simplified rate
laws:30
-
III
2(SCN)
III
+
[
Fe(dapsox)(NCS)
2
]
species (K
in Table 2). The
k1 + k1LiK [Li ]
-
kobs(1)
)
[SCN ] + k
-1Li
(7)
(8)
trend of K < K
III
III
< K
III
(Table 2) is consistent with
III
+
SCN 2(SCN)
1
+ K [Li ]
-
the expectation that coordinated SCN will increase the
nucleophilicity of the complex.
III
SCN
+
k2 + k2LiK
[Li ]
-
kobs(2)
)
[SCN ] + k
-2Li
Although upon addition of LiOTf to the solution of
III
+
1
+ K [Li ]
SCN
-
[Fe(dapsox)(NCS)
2
]
a shift in absorption maximum is
observed, it does not cause a significant color change of the
solution. Similar spectral changes are observed upon addition
k
k
obs(1) is the measured rate constant for the first reaction step,
30
1
, k1Li, and k-1Li are the corresponding rate constants for
of LiClO
4
. However, addition of LiBF
4
and LiPF
6
causes a
the “lithium-free” and “lithium-bound” forms as reactive
species (Scheme 3), and K is the equilibrium constant for
-
2
significant change in the spectrum of [Fe(dapsox)(NCS) ]
III
and, consequently, in the color of its solution. The spectral
changes and the colors of the solutions, resulting from the
addition of the same amount of different lithium salts (0.2
(30) Since the reverse rate constants are more prominent in the case of the
III
SCN
III
2(SCN)
“
lithium bound” forms and the corresponding K
and K
values are relatively high, the terms for the overall reverse rate
constants for the first and second reaction steps in eqs 7 and 8 can be
reduced to k-1Li and k-2Li, respectively.
-
-5
M) to [Fe(dapsox)(NCS)
2
] (5 × 10 M) in CH
3
CN, are
-
presented in Figure 12. In addition, [Fe(dapsox)(NCS)
2
] was
7858 Inorganic Chemistry, Vol. 46, No. 19, 2007

Iron Complex as a Ditopic Receptor for Lithium Salts
{
[Fe(dapsox)(CH
3
CN)
2
]‚LiBF
4
}+ adduct, with the strongest
interaction between the iron complex and lithium salt (Table
-
1
) and its bis(thiocyanato) {[Fe(dapsox)(NCS)
2
4
]‚LiBF }
form (Figure 13). Instead of acetonitrile, hydrogen cyanide
was used as a working model for the calculations. The
obtained structures are shown in Figure 14.
-
In the calculated structures, the BF
4
anion interacts with
the seven-coordinate iron complex by forming hydrogen
bonds between the fluoride atoms and two hydrogen atoms
2
-
of the pendant amide groups of the dapsox ligand. The
two amide parts of the ligand are held in a preorganized
arrangement by coordination of the R-oxyazine oxygen atoms
to the iron center and, together with the bound anion, define
+
+
a pocket of suitable size to host the Li cation. The Li cation
binds to the negatively charged R-oxyazine oxygen atoms,
-
solvent molecule, and the two fluoride atoms of BF
4
which
closes the tweezerlike pocket.
The Fe-N bond lengths in the equatorial plane are in
excellent agreement with that found in the crystal structure
3
+
of [Fe(dapsox)(H
2
O)
2
]ClO
4
. As expected, upon Li coor-
-
dination, the Fe-O bonds become longer (2.08 Å, Figure
1
4a) than those observed in the crystal structure of [Fe-
(
dapsox)(H O) ]ClO (2.046(3) and 2.066(3) Å) due to the
2
2
4
partial neutralization of the negative charge of the R-oxyazine
oxygen donor atoms. Negative charge of the bridged
R-oxyazine oxygen atoms results in somewhat shorter Li-
-
O bonds (2.01 Å, Figure 14a) than the ones calculated for
+
32
[
Li(H
2
O)
5
] (average d(Li-O) ) 2.07 Å).
-
The coordination of two SCN to the iron center causes
elongation of the equatorial Fe-ligand bonds, whereas the
elongation of Fe-O bonds is more prominent than that of
Figure 14. Calculated structures (B3LYP/LANL2DZp) of (a) {[Fe-
(
-
dapsox)(CH3CN)2]‚LiBF4}+ and (b) {[Fe(dapsox)(NCS)2]‚LiBF4} .
-
the Fe-N bonds. At the same time, an increase in the
-
electron density upon SCN coordination results in somewhat
titrated with LiBF
4
(Figure 13) and the corresponding
III
-1
shorter Li-O bonds (Figure 14b). This is in agreement with
K2
constant was determined to be 2828 ( 236 M ,
(SCN)
the higher apparent LiOTf binding constant for [Fe(dapsox)-
showing a more prominent binding than in the case of LiOTf
-
III
2SCN
+
III
(NCS)
2
] , K
,
than for [Fe(dapsox)(CH
3
CN)
2
] , K
(Table 2).
(Table 2). The N-H‚‚‚F hydrogen bonds are not affected
4
The product solutions in the case of LiOTf, LiClO , and
by the changes in the coordination sphere of the more distant
iron center.
LiBF
4
are stable, whereas the solution with LiPF
6
becomes
] (up to
.004 M) causes similar spectral changes as those observed
colorless within 15 min. Interestingly, lower [LiPF
0
6
Conclusion
4
with 0.2 M LiBF (Figures 12 and S9), but further addition
6
of LiPF leads to the absorbance decrease and shift of the
In the present paper, we have shown by electrochemical,
maximum to 570 nm (Figure S9). The resulting violet
solution bleaches in a short time. The violet color and the
absorbance at 570 nm are typical for the Fe(II)-thiocyanato
NMR, kinetic, thermodynamic, spectrophotometric, and DFT
studies that the seven-coordinate [Fe(dapsox)(CH CN) ]
3 2
complex behaves as a ditopic receptor for lithium salts in
acetonitrile. The observed apparent binding constants increase
+
species, which is formed in the redox reaction between Fe-
-
(
III) and SCN in acetonitrile, and dissociates to Fe(II) and
in the order LiOTf < LiClO
4
4 6
< LiBF . In the case of LiPF ,
31
2
thiocyanogen (SCN) , resulting in a colorless solution. Such
the solution chemistry is more complex due to the lithium-
complex chemistry suggests that in the presence of LiPF
6
-
catalyzed PF
6
hydrolysis processes. However, the electro-
show that the coordina-
-
and high [SCN ], dechelation of the iron(III) [Fe(dapsox)-
chemical measurements with LiPF
tion of anionic species (in that case PO
6
-
(NCS)
2
]
complex takes place, followed by the above-
-
2-
F
2 2
3
or PO F ) to
mentioned redox processes.
the iron center strongly increases the interactions between
the complex and lithium salt and the coupling between the
lithium binding and iron reduction. From the kinetic and
thermodynamic studies of the reaction between [Fe(dapsox)-
DFT Calculations. Since the attempts to crystallize the
seven-coordinate Fe(III) complex with bound lithium salt
were not successful, we performed DFT calculations to
provide insight into the structural characteristics of the
(
32) Puchta, R.; Galle, M.; van Eikema Hommes, N.; Pasgreta, E.; van
(31) Kratochvil, B.; Long, R. Anal. Chem. 1970, 42, 43-46.
Eldik, R. Inorg. Chem. 2004, 43, 8227-8229.
Inorganic Chemistry, Vol. 46, No. 19, 2007 7859

Sarauli et al.
(
CH
3
CN)
2
]+ and SCN- in the presence of NBu
4
OTf and
synthetic variations, which our ligand system offers, make
these types of mononuclear complexes promising structural
motifs for designing efficient lithium salt receptors that can
meet the requirements in terms of different applications.
LiOTf, respectively, it has also been concluded that the
increase in the electron density around the iron center upon
anion coordination facilitates lithium binding. These experi-
ments show that in nonaqueous media lithium salts cannot
be simply used as supporting electrolytes, since they can
affect the kinetic behavior of the studied complex. Mecha-
3 2
nistic studies have revealed that the [Fe(dapsox)(CH CN) ]
complex and its mono(thiocyanato) adduct undergo substitu-
tion processes according to an associative (A) mechanism.
This can be interpreted in terms of the poor donor ability of
Acknowledgment. The authors gratefully acknowledge
fruitful discussions with Prof. Rudi van Eldik, Prof. Klaus
Koch (University of Stellenbosch, South Africa), and Joachim
Maigut and the financial support from the Deutsche Fors-
chungsgemeinschaft through SFB 583 “Redox-active Metal
Complexes”.
+
CH
3
CN, where the Fe-NCCH
3
bonds are so labilized that
-
7
the rate-determining step has almost only Fe-NCS bond
making character. The higher reactivity of the lithium adduct
of the complex also speaks in favor of an associative
character of the rate-determining step in the substitution
mechanism.
Supporting Information Available: Electrochemical data; -
+
Li NMR T
in CH
potentials of [Fe(dapsox)(CH
tions; cyclic voltammograms of [Fe(dapsox)(CH
presence of LiPF ; a monoexponential relaxation of the Li nucleus,
1
and R values; cyclic voltammograms of Fc/Fc couple
3
CN in the presence of different lithium salts; the half-wave
0/+
3 2
CN) ]
vs different anion concentra-
0
/+
3
CN)
2
]
+
in the
In conclusion, the seven-coordinate [Fe(dapsox)(CH
3
-
6
+
obtained by using the inversion recovery pulse sequence; different
spectrophotometric titration data; and temperature-dependent kinetic
measurements. This material is available free of charge via the
Internet at http://pubs.acs.org.
CN)
2
] complex, in its Fe(III), Fe(II), and anion-substituted
forms, having a shape of “charged tweezers”, is able to
intimately associate with the lithium salts by means of shape
recognition, hydrogen-bond complementarity, and charge
assistance. This, together with the possibility for simple
IC7005056
7860 Inorganic Chemistry, Vol. 46, No. 19, 2007