Iron complexes of new pentadentate ligands containing the 1,4,7-triazacyclononane-1,4-diacetate motif. Spectroscopic, electro-, and photochemical studies

Article abstract of DOI:10.1021/ic062001j

Three new pentadentate, pendent arm macrocycles containing the 1,4,7-triazacyclononane-1,4-diacetate motif have been synthesized, and their coordination chemistry with Fe(III) has been investigated. Eight new octahedral Fe(III) complexes containing chloro, azido, or μ-oxo ligands have been synthesized, five of which have been characterized by X-ray crystallography. Spectroscopic characterization of these octahedral Fe(III) complexes by UV-vis, IR, electrochemistry, EPR, magnetic susceptibility, and zero-field Mossbauer measurements firmly establishes the high-spin state of the iron in all complexes. Electrochemistry studies of the azido-Fe(III) complexes show that they can be reversibly oxidized to the corresponding Fe(IV) species at -20°C, and Fe(II), Fe(III), and Fe(IV) species show characteristic IR and UV-vis profiles. Photolysis of one of the azido complexes was studied as a function of temperature (room temperature vs 77 K) and wavelength (480, 419, and 330 nm). Photoreduction to a high-spin Fe(II) species occurs under all conditions, which is proposed to be the dominant photochemical pathway generally available to high-spin ferric azido complexes.

Full text of DOI:10.1021/ic062001j

Inorg. Chem. 2007, 46, 2208−2219
Iron Complexes of New Pentadentate Ligands Containing the
1,4,7-Triazacyclononane-1,4-diacetate Motif. Spectroscopic, Electro-, and
Photochemical Studies
||
Yu-Fei Song,*,†,‡ John F. Berry,*,†, Eckhard Bill,† Eberhard Bothe,† Thomas Weyhermu
1
ller,† and
Karl Wieghardt†
Max-Planck-Institut fu¨r Bioanorganische Chemie, Stiftstrasse 34-36, D-45470 Mu¨lheim an der
Ruhr, Germany, and Department of Chemistry, UniVersity of Wisconsin, 1101 UniVersity AVenue,
Madison, Wisconsin 53706
Received October 19, 2006
Three new pentadentate, pendent arm macrocycles containing the 1,4,7-triazacyclononane-1,4-diacetate motif have
been synthesized, and their coordination chemistry with Fe(III) has been investigated. Eight new octahedral Fe(III)
complexes containing chloro, azido, or
by X-ray crystallography. Spectroscopic characterization of these octahedral Fe(III) complexes by UV
electrochemistry, EPR, magnetic susceptibility, and zero-field Mo¨ssbauer measurements firmly establishes the high-
spin state of the iron in all complexes. Electrochemistry studies of the azido Fe(III) complexes show that they can
be reversibly oxidized to the corresponding Fe(IV) species at 20 C, and Fe(II), Fe(III), and Fe(IV) species show
characteristic IR and UV vis profiles. Photolysis of one of the azido complexes was studied as a function of
µ-oxo ligands have been synthesized, five of which have been characterized
−vis, IR,
−
−
°
−
temperature (room temperature vs 77 K) and wavelength (480, 419, and 330 nm). Photoreduction to a high-spin
Fe(II) species occurs under all conditions, which is proposed to be the dominant photochemical pathway generally
available to high-spin ferric azido complexes.
Introduction
tetraazacyclotetradecane4 (cyclam) from conjugated macro-
cyclic ligands such as porphyrins5 or corroles.6
1,4,7-Triazacyclononane (TACN) is now a classical ligand
in inorganic chemistry; it and many N-functionalized deriva-
tives have been applied in several different areas of chemistry
for many years.1-3 The TACN ring binds metals strongly to
cover a trigonal face of an octahedrally coordinated transition
metal ion, providing kinetic stability to the complex and
useful control over the remaining environment. In addition,
TACN is a pure σ-donor and is therefore “innocent” in a
spectroscopic sense, as well as a redox sense, meaning that
all redox processes exhibited by TACN complexes may
safely be assumed to be metal-based. This distinguishes
TACN and other saturated macrocycles such as 1,4,8,11-
Pentadentate pendent arm derivatives of TACN are cur-
rently of interest because these macrocycles can be utilized
to construct mononuclear octahedral complexes that possess
a single open coordination site allowing the binding and
activation of small molecules.3,7-9 Such systems are never-
theless rare compared with hexadentate N,N′,N′′-trifunction-
alized TACN ligand systems and their chelate complexes.1
The majority of reported complexes with pentadentate
derivatives of TACN and related macrocycles are complexes
(4) Bosnich, B.; Poon, C. K.; Tobe, M. L. Inorg. Chem. 1965, 4, 1102.
(5) Rutter, R.; Hager, L. P.; Dhonau, H.; Hendrich, M.; Valentine, M.;
Debrunner, P. Biochemistry 1984, 23, 6809.
(6) (a) Nardis, S.; Paolesse, R.; Licoccia, S.; Fronczek, F. R.; Vicente,
M. G. H.; Shokhireva, T. K.; Cai, S.; Walker, F. A. Inorg. Chem.
2005, 44, 7030. (b) Walker, F. A.; Licoccia, S.; Paolesse, R. J. Inorg.
Biochem. 2006, 100, 810. (c) Gross, Z.; Gray, H. B. Comments Inorg.
Chem. 2006, 27, 61.
(7) Warden, A. C.; Spiccia, L.; Hearn, M. T. W.; Boas, J. F.; Pilbrow, J.
R. Dalton Trans. 2005, 1804.
* To whom correpondence should be addressed. E-mail:
ysong@chem.gla.ac.uk (Y.-F.S.); berry@chem.wisc.edu (J.F.B.).
† Max-Planck-Institut fu¨r Bioanorganische Chemie.
‡ Current Address: Department of Chemistry, University of Glasgow,
G12, 8QQ, Glasgow, UK.
|| University of WisconsinsMadison.
(1) Chaudhuri, P.; Wieghardt, K. Prog. Inorg. Chem. 1987, 35, 329.
(2) Bernhardt, P. V.; Lawrance, G. A. Coord. Chem. ReV. 1990, 104, 297.
(3) Lindoy, L. F. The Chemistry of Macrocyclic Ligands; Cambridge
University Press: Cambridge, 1989.
(8) McAuley, A.; Subramanian, S.; Zaworotko, M. J.; Atencio, R. Inorg.
Chem. 1998, 37, 4607.
(9) Fallon, G. D.; McLachlan, G. A.; Moubaraki, B.; Murray, K. S.;
O’Brien, L.; Spiccia, L. J. Chem. Soc., Dalton Trans. 1997, 2765.
2208 Inorganic Chemistry, Vol. 46, No. 6, 2007
10.1021/ic062001j CCC: $37.00
© 2007 American Chemical Society
Published on Web 02/22/2007

Iron Complexes of New Pentadentate Ligands
Scheme 1
of Cu(II),7 Mn(II),9 and Cr(III),9 though Fe(III) complexes
with pentadentate TACN-derived ligands having pendent
pyridyl arms are known.9 Therefore, we have investigated
the coordination chemistry of three new pentadentate ligands
containing the 1,4,7-triazacyclononane-1,4-diacetate motif
and their resulting iron complexes. This work was instigated
by recent advances in the chemistry of iron in oxidation states
higher than +3. In particular, the discovery that ferric azido
cyclam complexes10 and later, azido cyclam-acetate com-
plexes11 undergo photo-oxidation to yield unstable Fe(V)-
nitrido species was important. By changing the macrocyclic
ligand from a mono-pendent arm cyclam-acetate to the bis-
pendent arm TACN-diacetate, two important changes are
made. First, the charge of the ligand is increased from 1-
to 2-, which should better stabilize the positive charge of
higher-valent iron ions. Also, the symmetry of the resulting
iron complexes is dramatically changed: iron cyclam
complexes having two more ligands trans to each other are
tetragonally distorted leading to a splitting in the t2g orbitals
such that the dxy orbital lies significantly lower in energy
than the nearly degenerate dxz, dyz pair. TACN-based
complexes are often trigonally distorted, which causes a
much larger splitting between the dxz and dyz orbitals. This
is desirable because the known tetragonal Fe(V)-nitrido
complexes possess a nearly orbitally degenerate S ) 1/2
ground state due to the presence of one unpaired electron in
the dxz, dyz set.12 In order to judge the effects of orbital
degeneracy of the Fe(V) system on its reactivity, we were
intensely interested in the photolysis products of the trigo-
nally distorted Fe(III) complexes reported herein.
In this paper, three new pentadentate ligands (Scheme 1)
containing the 1,4,7-triazacyclononane-1,4-diacetate motif
and eight new octahedral Fe(III) complexes containing
chloro, azido, or µ-oxo ligands have been prepared and
characterized. The spectroscopic, electro-, and photochemical
studies of these azido-Fe(III) compounds are presented, and
although no evidence for photo-oxidation to Fe(V) has been
obtained for these complexes, some very interesting Fe(IV)
species have been electrochemically generated and studied.
Experimental Section
Caution: Although we haVe not encountered any problems, it
should be kept in mind that perchlorate salts and metal azides are
potentially explosiVe and should be handled carefully only in small
amounts with appropriate precautions.
The solvents diethyl ether, toluene, ethanol, dichloromethane,
chloroform, methanol, acetonitrile, n-hexane, and n-pentane were
commercially available and dried over molecular sieves before use;
triethylamine was dried over Na2SO4 overnight. The chemicals
bromoacetic acid ethyl ester, 2-tert-butoxycarbonyloxyimino-2-
phenylacetonitrile (BOC-ON), trifluoroacetic acid, citric acid, iron
powder, 1-chloromethyl-4-methoxybenzene, sodium metal, NaBH4,
NaOH, KOH, K2CO3, acetone, 2-propanol, and dimethylformamide
(DMF) were obtained from commercial sources and used without
further purification.
TACN was synthesized as described in the literature.13 [Fe(dmf)6]-
(ClO4)3 was synthesized by refluxing Fe(ClO4)3 in dry DMF for
several hours, and the compound was obtained as a yellow
(10) Meyer, K.; Bill, E.; Mienert, B.; Weyhermu¨ller, T.; Wieghardt, K. J.
Am. Chem. Soc. 1999, 121, 4859.
(11) Grapperhaus, C. A.; Mienert, B.; Bill, E.; Weyhermu¨ller, T.; Wieghardt,
K. Inorg. Chem. 2000, 39, 5306.
(12) Aliaga-Alcalde, N.; DeBeer George, S.; Mienert, B.; Bill, E.;
Wieghardt, K.; Neese, F. Angew. Chem., Int. Ed. 2005, 44, 2908.
(13) Wieghardt, K.; Schmidt, W.; Nuber, B.; Weiss, J. Chem. Ber.-Recl.
1979, 112, 2220.
Inorganic Chemistry, Vol. 46, No. 6, 2007 2209

Song et al.
precipitate. This starting material was characterized by satisfactory
C, H, N, Fe, and Cl elemental analysis.
in order to precipitate NaBr. The solution was filtered and the filtrate
was concentrated under reduced pressure to give a yellow oil, which
was extracted with Et2O/n-hexane (1:1, 250 mL). The organic phase
was concentrated under reduced pressure, and the residue was dried
Synthesis of Li2L1 (R ) Benzyl). 1,4,7-Triazacyclononane-1,4-
dicarboxylic acid di-tert-butyl ester (1) was synthesized according
1
1
to the literature.14 Yield: 66%. H NMR (400 MHz, chloroform-
under vacuum to give a yellow oil. Yield: 3.8 g, 65%. H NMR
(400 MHz, chloroform-d): 1.28 (m, 6H, -CH3), 2.91 (m, 12H,
TACN ring), 3.34 (d, 2H, -CH2-COO-, J ) 16.3 Hz), 3.40 (d,
2H, -CH2-COO-, J ) 16.3 Hz), 3.68 (s, 2H, -CH2-), 4.13 (m,
4H, -CH2-CH3), 7.30 (m, 5H, -benzene ring). ESI-MS spec-
trometry (positive mode, MeOH) m/z ) 391 (M+).
d): 1.42 (s, 18H, t-Bu), 2.87 (m, 4H, TACN ring), 3.19 (m, 4H,
TACN ring), 3.40 (m, 4H, TACN ring). ESI-MS spectrometry
(positive mode, MeOH) m/z ) 329 (M+). Anal. Calcd for
C16H31N3O4 (329.4 g/mol): C 58.33, H 9.48, N, 12.76; Found: C
58.62, H 9.86, N 12.24.
4-Carboxymethyl-7-benzyl-1,4,7-triazacyclononan-1-yl Acetic
Acid Lithium Salt (6). To a solution of 5 (5.4 g, 13.8 mmol) in
EtOH (50 mL) was added finely powdered LiOH‚H2O (1.26 g, 30
mmol). The mixture was refluxed at 70 °C for 24 h and then filtered;
the filtrate was concentrated under reduced pressure to give a pale
yellow solid. The solid was washed with cold ethanol, Et2O, and
dried under vacuum. Yield: 3.6 g, 70%. 1H NMR (400 MHz, CD3-
OD-d4): δ 2.27 (m, 2H, TACN ring)), 2.56 (m, 4H, TACN ring),
2.73 (m, 6H, TACN ring), 3.12 (d, 2H, -CH2-, J ) 16.4 Hz),
3.26 (d, 2H, -CH2-, J ) 16.4 Hz), 3.80 (s, 2H, -CH2-), 7.29 (t,
1H J ) 7.2 Hz), 7.37 (t, 2H, J ) 7.4 Hz), 7.45 (d, 2H, J ) 7.2
Hz). 13C NMR (100 MHz, CD3OD-d4): δ 52.83 (-CH2-), 54.16
(-CH2-), 54.54 (-CH2-), 63.34 (-CH2-), 64.07 (-CH2-),
129.08 (-CH-), 129.90 (-CH-), 132.03 (-CH-), 138.52
(-C-), 181.31 (-CO2-). ESI-MS spectrometry (negative mode,
MeOH) m/z ) 334 ([Li2L1 - 2Li + H]-). Anal. Calcd for Li2L1‚
2LiOH‚H2O (C17H27N3O7Li4,, 413.2 g/mol); C 49.42, H 6.59, N
10.17; Found C 49.82, H 6.48, N 9.97. IR (KBr, cm-1): 3412 (b),
1590(b), 1494 (w), 1441 (m), 1411 (s), 1331 (m), 1120 (s), 1101
(m), 1078 (m), 1050 (w), 1003 (m), 928(w), 899 (w), 728 (m),
705 (m), 635 (w), 487 (w).
7-Benzyl-1,4,7-triazacyclononane-1,4-dicarboxylic Acid Di-
tert-butyl Ester (2). KOH (2.63 g, 46.9 mmol) was added to a
solution of compound 1 (4.67 g, 14.2 mmol) in toluene (50 mL) at
room temperature. To this mixture was added benzyl bromide (2.91
g, 17.0 mmol) in toluene (20 mL) within 10 min. The reaction
mixture was kept stirring at 70 °C for 24 h, during which the
reaction was monitored by TLC until the starting material (1)
reacted completely. Then, the reaction mixture was filtered, dried
over MgSO4, and concentrated under reduced pressure to give 2
as yellow oil. Yield: 5.6 g, 95%. 1H NMR (400 MHz, chloroform-
d): 1.43 (s, 9H, t-Bu), 1.49 (s, 9H, t-Bu), 2.34 (s, 2H, -CH2-),
3.14 (m, 4H, TACN ring), 3.21 (m, 4H, TACN ring), 3.51 (m, 4H,
TACN ring), 7.14 (m, 2H, -benzene ring), 7.16 (t, 1H, J ) 7.0
Hz), 7.32 (m, 2H, -benzene). ESI-MS spectrometry (positive mode,
MeOH) m/z ) 420 (M+).
4-Benzyl-1,4,7-triazacyclononane-1,4,7-ium Bis(trifluoroac-
etate) (3). To a solution of 2 (8.0 g, 19.0 mmol) in CH2Cl2 (20
mL) was added 50% CF3CO2H/CH2Cl2 (100 mL) at 0 °C. The
solution was stirred at 0 °C for 2.5 h, during which the reaction
was examined by TLC. Then, the solution was concentrated under
reduced pressure and diethyl ether was added to the obtained
residue. The solvent was removed by rotary evaporation in order
to remove the excess CF3CO2H, and this process was repeated three
times. A light yellow powder was obtained, which was washed
with Et2O/n-hexane (1:1, 200 mL) several times and dried under
vacuum. Yield: 8.7 g, 82%. ESI mass spectrometry (positive mode,
MeOH) indicated CF3CO2H free fragment. m/z ) 220 (M′ + H)+.
1H NMR (400 MHz, CD3OD-d4): 3.05 (br, 4H, TACN ring), 3.46
(br, 4H, TACN ring), 3.60 (s, 2H, -CH2-), 3.80 (br, 4H, TACN
ring), 7.31 (m, 5H, -benzene ring).
1-Benzyl-1,4,7-triazacyclononane (4). To a cold (0 °C) solution
of 3 (8.0 g, 14.2 mmol) in CHCl3 (120 mL) was added Et3N (10
mL). The solution was stirred at 0 °C for 1 h and then washed
with saturated NaCl solution (3 × 300 mL). The organic phase
was dried over MgSO4 and concentrated under reduced pressure
to give 4 as a yellow oil. Yield: 2.8 g, 90%. 1H NMR (400 MHz,
chloroform-d): 2.10, (br, 2H, TACN ring), 2.62 (m, 6H, TACN
ring), 2.75 (m, 4H, TACN ring), 3.68 (s, 2H, -CH2-), 7.22 (m,
5H, -benzene ring). ESI-MS spectrometry (positive mode, MeOH)
m/z ) 220 (MH)+.
4-Benzyl-7-ethoxycarbonylmethyl-1,4,7-triazacyclononane-1-
yl Acetic Acid Ethyl Ester (5). To a solution of 4 (3.6 g, 16.4
mmol) in EtOH (30 mL) was added bromoacetic acid ethyl ester
(5.62 g, 33.8 mmol) at room temperature. Then, the solution was
cooled to 0 °C and to this was added sodium metal (0.78 g, 33.8
mmol) dissolved in EtOH (40 mL). The solution was kept stirring
at 0 °C for 1 h, during which the reaction was examined by TLC
until the starting material 4 reacted completely. Then, the solution
was concentrated under reduced pressure. To the obtained residue
was added 50 mL of toluene, and the mixture was stirred for 2 h
Syntheses of Li2L2 and Li3L3 were carried out in a similar way
to that of Li2L1, and the detailed synthetic procedures can be found
in the Supporting Information.
Preparation of Complexes. 2[L1Fe(µ-O)FeL1]‚NaPF6‚14H2O‚
3MeOH (17). A solution of 6 (120 mg, 0.29 mmol) in MeOH (10
mL) was added to Fe(dmf)6(ClO4)3 (254 mg, 0.29 mmol) in MeOH
(10 mL). The reaction mixture was refluxed for 1.5 h, and a yellow
solution with a yellow precipitate was formed. Then, NaN3 (188.5
mg, 2.9 mmol) in MeOH (10 mL) was added to the solution and a
clear, red solution formed immediately. The solution was allowed
to stir for 1.5 h, and then a saturated aqueous solution of KPF6 (5
mL) was added. The resulting solution was concentrated to 10 mL
and filtered; diffusion of diethyl ether into the filtrate resulted in
the formation of red square-shaped crystals. Yield: 203 mg, 67%.
Anal. Calcd for the complex [L1Fe(µ-O)FeL1]‚0.5NaPF6‚7H2O‚
1.5MeOH (C35.5H61F3Fe2N6Na0.5O17.5P0.5, 1047.59 g/mol) C 40.70,
H 5.82, N 8.02; Found C 41.13, H 5.34, N 8.22. ESI-MS
spectrometry (positive mode, MeOH) m/z ) 833 ([L1-Fe-O-Fe-L1
+ K]+). IR (KBr, cm-1): 3446(s), 1635 (s, CdO), 1496 (w), 1458
(m), 1358 (s), 1314 (w), 1099 (s), 1007 (w), 967 (w), 927 (m), 846
(s), 806 (s), 773 (m), 725 (w), 626 (m), 561 (m). It should be
mentioned here that this compound can also be obtained by using
FeCl3‚6H2O instead of Fe(dmf)6(ClO4)3.
[L1-Fe-Cl]‚0.5EtOH‚0.5H2O (18). Compound 6 (240 mg, 0.58
mmol) was added to FeCl3 (97 mg, 0.58 mmol) in MeOH (20 mL).
The reaction mixture was refluxed for 2 h, and then the resulting
yellow precipitate was collected by filtration and dried at room
temperature. Recrystallization of the compound from EtOH/H2O
(1:1, 30 mL) resulted in the formation of yellow crystals suitable
for X-ray crystallographic analysis after 1 week. Yield: 187 mg,
70%. Anal. Calcd for L1-Fe-Cl‚0.5EtOH‚0.5H2O (C18H27N3O5FeCl,
456.7 g/mol) C 47.34, H 5.96, N 9.20; Found C 47.52, H 6.31, N
(14) Kimura, S.; Bill, E.; Bothe, E.; Weyhermu¨ller, T.; Wieghardt, K. J.
Am. Chem. Soc. 2001, 123, 6025.
2210 Inorganic Chemistry, Vol. 46, No. 6, 2007

Iron Complexes of New Pentadentate Ligands
8.95. ESI-MS spectrometry (positive mode, MeOH) m/z ) 389 ([L1
- Fe]+ 100%), 424 ([L1 - Fe - Cl]+ 60%). IR (KBr, cm-1): 3539
(s), 1665 (s, CdO), 1488 (m), 1461 (m), 1451 (m), 1346 (s), 1328
(s), 1315 (s), 1279 (s), 1251 (w), 1228 (w), 1094 (m), 1079 (m),
1065 (m), 1040 (m), 1022 1010, 967 (s), 922 (s), 839 (m), 812
(m), 762 (s), 729 (m), 704 (s), 652(s), 562 (m), 507 (m), 487 (m),
455 (w), 421(w).
led to the formation of orange crystals after 1 week. Yield: 50%.
Anal. Calcd for the complex L3-Fe-N3 (C18H25N6O5Fe, 461.3
g/mol) C 46.87, H 5.46, N 18.22; Found C 46.62, H 5.13, N 17.75.
ESI-MS spectrometry (positive mode, MeOH) m/z ) 462 ([L3-
Fe-N3]+). IR (KBr, cm-1): 3420 (m), 2064 (s, N3), 1677 (s, CdO),
1514 (m), 1439 (m), 1338 (w), 1271 (m), 1252 (m), 1182 (m),
1086 (w), 1066 (w), 967 (w), 919 (w), 825 (m), 702 (w).
[L1-Fe-N3] (19). Compound 18 (77.5 mg, 0.17 mmol) was
dissolved in MeOH (20 mL), to which NaN3 (11 mg, 0.17 mmol)
was then added. The reaction mixture was stirred at room
temperature for 3 h, and a red solid precipitated. The solution was
filtrated, and the solid was recrystallized from MeCN/MeOH (6:4,
30 mL) in the dark. Orange, needle crystals were formed after 1
week, which were suitable for X-ray crystallography. Yield: 48
mg, 66%. Anal. Calcd for the complex [L1-Fe-N3] (C17H23N6O4-
Fe, 431.2 g/mol) C 47.35, H 5.38, N 19.49; Found C 46.98, H
5.24, N 19.23. ESI-MS spectrometry (positive mode, MeOH) m/z
) 431 ([L1-Fe-N3]+). IR (KBr, cm-1): 3463 (s), 2066 (s, N3), 1676
(s, CdO), 1492 (w), 1320 (m), 1283 (s), 1089 (w), 1064 (w), 914
(s), 812 (m), 771 (m), 712 (s), 559 (w).
[L2-Fe-Cl]‚0.5MeOH (20). A solution of 11 (274 mg, 0.66
mmol) in MeOH (10 mL) was added to FeCl3 (106 mg, 0.66 mmol)
in MeOH (10 mL). The reaction mixture was heated to reflux for
2 h, and the resulting yellow precipitate was filtered and dried under
vacuum. Recrystallization of the compound from MeOH/H2O (7:
3, 50 mL) led to the formation of yellow crystals after 1 week.
Yield: 194 mg, 75%. Anal. Calcd for L2-Fe-Cl‚0.5MeOH
(C13.5H25N3O4.5FeCl, 392.4 g/mol) C 41.32, H 6.37, N 10.70; Found
C 41.54, H 6.82, N 10.43. ESI-MS spectrometry (positive mode,
MeOH) m/z ) 341 ([L2-Fe]+ 100%) and 377 ([L2-Fe-Cl]+ 70%).
IR (KBr, cm-1): 3421 (m), 1671 (s, CdO), 1492 (m), 1459 (m),
1331 (s), 1293 (s), 1093 (m), 1030 (w), 1004 (w), 922 (m), 823
(m), 714(m).
[L2-Fe-N3] (21). This compound was synthesized by reaction
of 1 equiv each of compound 20 and NaN3 in MeOH. The reaction
mixture was stirred for 3 h, and small amount of precipitate was
removed by filtration. The filtrate was allowed to evaporate at room
temperature in the dark. Orange crystals suitable for X-ray
crystallography formed after 1 week. Yield: ∼52%. Anal. Calcd
for the complex L2-Fe-N3 (C13H23N6O4Fe, 383.2 g/mol) C 40.75,
H 6.05, N 21.93; Found C 40.28, H 6.23, N 21.38. ESI-MS
spectrometry (positive mode, MeOH) m/z ) 383 ([L2-Fe-N3]+).
IR (KBr, cm-1): 3431 (m), 2964 (w), 2066 (s, N3), 1674 (s, CdO),
1497 (w), 1460 (w), 1335 (s), 1296 (s), 1278 (s), 1090 (w), 1078
(m), 1027 (m), 924 (s) 811 (m), 720 (m), 620 (w), 562 (w), 491
(w), 470 (m), 415 (w)
[L3-Fe-µ-O-Fe-L3]‚6H2O (24). This was synthesized in a similar
way to 17. Yield: 60%. Anal. Calcd for the complex L3-Fe-µ-O-
Fe-L3‚6H2O (C36H62N6O17Fe2, 962.6 g/mol): C 44.92, H 6.49, N
8.73; Found C 44.83, H 6.62, N 8.70. ESI-MS spectrometry
(positive mode, MeOH) m/z ) 855 ([L3-Fe-O-Fe-L3 + H]+). IR
(KBr, cm-1): 3440 (s), 1686 (s, CdO), 1619 (s, CdO), 1515 (s),
1356 (m), 1338 (m), 1292 (m), 1246 (m), 1184 (m), 1099 (m),
1090 (m), 1028 (m), 966 (m), 801 (s, Fe-O-Fe), 642 (w).
57FeCl3 was prepared according to the method reported in the
literature10 and isotopically enriched azido-Fe(III) complexes were
prepared by using 57FeCl3.
Physical Measurements. Infrared spectra (4000-400 cm-1) of
solid samples were recorded on a Perkin-Elmer 2000 FT-IR/FT-
NIR Spectrometer as KBr disks. ESI-MS spectra were recorded
on a Finnigan MAT 95 mass spectrometer. 1H NMR spectra were
measured on Bruker ARX 250 spectrometer. Cyclic voltammetry
and coulometry were performed using an EG&G potentiostat/
galvanostat model 273A with acetonitrile solutions of samples
containing 0.10 M [N(n-Bu)4]PF6 as supporting electrolyte. Fer-
rocene was used as an internal standard, and all reported potentials
are referenced versus the Fc+/Fc couple. UV-vis spectra of
solutions were measured on a Perkin-Elmer Lambda 19 spectro-
photometer in the range 200-1000 nm. Temperature-dependent
magnetic susceptibilities of powdered samples were measured using
a SQUID magnetometer (Quantum Design) at 1.0 T (2.0-300 K).
Corrections for underlying diamagnetism were made by using
tabulated Pascal constants. X-band EPR spectra were recorded on
a Bruker ESP 300E spectrometer equipped with a helium-flow
cryostat (Oxford Instruments ESR 910). Mo¨ssbauer data were
recorded on an alternating constant-acceleration spectrometer. The
minimum experimental line width was 0.24 mm s-1 (full width at
half-height). The field at the sample is oriented perpendicular to
the γ-beam. The 57Co/Rh source (1.8 GBq) was positioned at room
temperature inside the gap of the magnet system at a zero-field
position. Isomer shifts are quoted relative to iron metal at 300 K.
Spin Hamiltonian simulations of EPR spectra were performed
with a program which was developed from the S ) 5/2 routines of
Gaffney and Silverstone15 and which specifically makes use of the
resonance-search procedure described therein.
[L3-Fe-Cl‚H2O] (22). A solution of 16 (365 mg, 0.66 mmol) in
MeOH (10 mL) was slowly added to a solution of FeCl3 (107.5
mg, 0.66 mmol) in MeOH (10 mL). The reaction mixture was
refluxed for 2 h and filtered. The filtrate was kept at room
temperature, and yellow crystals were obtained after 3 weeks.
Yield: 218 mg, 70%. Anal. Calcd for the complex L3-Fe-Cl‚H2O
(C18H27N3O6FeCl, 472.7 g/mol) C 45.73, H 5.76, N 8.89; Found C
45.35, H 5.28, N 9.22. ESI-MS spectrometry (positive mode,
MeOH) m/z ) 419 ([L3-Fe]+, 100%); 455 ([L3-Fe-Cl]+, 58%). IR
(KBr, cm-1): 3431 (m), 1679 (s, CdO), 1516 (s), 1455 (w), 1346
(m), 1329 (m), 1291 (s), 1280 (s), 1255 (m), 1092 (m), 1030 (m),
961 (s), 923 (s), 827 (s), 718 (w), 559 (w), 512 (w).
X-ray Crystallographic Data Collection and Refinement of
the Structures. Single crystals of 17, 19, and 20 and yellow crystals
of 18 and 21 were coated with perfluoropolyether and picked up
with looped nylon fibers. The crystals were immediately mounted
in the nitrogen cold stream (100 K) of a Nonius Kappa-CCD
diffractometer equipped with a Mo-target rotating-anode X-ray
source and a graphite monochromator (Mo KR, λ ) 0.71073 Å).
Final cell constants were obtained from least-squares fits of all
measured reflections. Crystallographic data of the compounds are
listed in Table 1. The Siemens ShelXTL16 software package was
used for solution and artwork of the structure; ShelXL9717 was
used for the refinement. The structures were readily solved by direct
[L3-Fe-N3] (23). This compound was synthesized by reaction
of 1 equiv each of compound 22 and NaN3 in MeOH. The reaction
mixture was stirred for 3 h at room temperature, and an orange-
reddish precipitate was observed, which was collected by filtration.
Recrystallization of the above precipitate from MeCN in the dark
(15) Gaffney, B. J.; Silverstone, H. J., In Biological Magnetic Resonance;
Berliner, L. J., Reuben, J., Eds.; Plenum Press: New York, 1993; Vol.
13.
(16) SHELXTL, 5.0; Siemens Analytical X-Ray Instruments, Inc.: Madison,
WI, 1994.
Inorganic Chemistry, Vol. 46, No. 6, 2007 2211

Song et al.
Table 1. Crystallographic Data of the Fe(III) Complexes Measured at 100 K
compounds
17
18
19
20
21
chem formula
fw
cryst syst
space group
a, Å
b, Å
c, Å
R, deg
â, deg
γ, deg
V, Å3
C
35.5H61F3Fe2N6Na0.5O17.5P0.5
C18H27ClFeN3O5
456.73
triclinic
C17H23FeN6O4
431.26
orthorhombic
P212121
7.4733(2)
12.9538(4)
18.9587(4)
90
C
13.5H25ClFeN3O4.5
C13H23FeN6O4
383.22
orthorhombic
Pbca
12.2611(6)
15.8967(8)
16.5101(9)
90
90
90
3218.0(3)
8
1047.59
monoclinic
P21/n
16.4915(6)
16.5318(6)
36.239(2)
90
97.229(5)
90
9801.5(7)
8
392.67
monoclinic
P21/c
15.6291(6)
6.9003(2)
15.8036(6)
90
103.783(5)
90
P1h
8.3266(6)
13.0757(9)
19.190(1)
73.25(1)
89.24(1)
88.90(1)
2000.39(2)
4
90
90
1835.35(8)
4
1655.3(1)
4
Z
F
calcd, g/cm
1.420
125 003/26.00
1219/0
0.0980; 0.2135
0.1123; 0.2203
1.517
40 720/31.01
517/1
0.0412; 0.0927
0.0503, 0.0970
1.561
46 198/30.99
253/0
0.0251; 0.0550
0.0287, 0.0565
1.576
24 975/31.03
221/0
0.0262; 0.0611
0.0331, 0.0640
1.582
47 895/28.49
219/0
0.0444; 0.0903
0.0738, 0.1024
reflns collected/θmax
no. of params/restr
R1a; wR2 (I > 2σ(I))b
Ra indices (all data)
a R1 ) 3||Fo| - |Fc||/3|Fo|. b wR2 ) [3[w(Fo - Fc )2]/3[w(Fo )2]]1/2, w ) 1/σ2(Fo ) + (aP)2 + bP, where P ) [max(0 or Fo ) + 2(Fc )]/3.
2
2
2
2
2
2
and Patterson methods and subsequent difference Fourier tech-
niques. All non-hydrogen atoms were refined anisotropically.
Hydrogen atoms were placed at calculated positions and refined
as riding atoms with isotropic displacement parameters, except for
some hydrogen atoms in disordered solvent molecules.
TACN with 2 equiv of BOC-ON leads to good yields and
high purity of the N,N′-protected 1,14 from which R groups
can be conveniently attached to the unprotected amine.
Deprotection with CF3CO2H/CH2Cl2 followed by treatment
with Et3N or NaOH and then coupling with ethyl bromoac-
etate yields the ethyl diester of the desired ligands, which
can be hydrolyzed with LiOH to produce the lithium salts
6, 11, and 16.
Reaction of Li2L (L ) L1 and L3) with [Fe(dmf)6](ClO4)3
in the presence of excess NaN3 affords oxo-bridged di-iron
species, whereas reaction with FeCl3 yields the chloro species
18, 20, and 22, which may be conveniently converted to the
azido complexes 19, 21, and 23 by reaction with 1 equiv of
NaN3. Since NaN3 may be hydrolyzed slowly in the presence
of H2O in the reaction mixture, it is recommended that 1-1.5
equiv of NaN3 should be used in the formation of the
corresponding azido compounds in order to prevent the
formation of oxo-bridged dimeric species.
Crystal Structures. The structures of 17-21 at 100 K
have been determined by single-crystal X-ray crystal-
lography. Compound 17 crystallizes in P21/n with Z ) 8.
The complex is neutral and contains a bridging oxo ligand
that connects two pseudo-octahedrally coordinated Fe(III)
ions. As shown in Figure 1, the bridging oxo-ligand is trans
to one of the N atoms of the TACN ring. The average Fe-N
and Fe-O bond lengths are 2.227(5) and 2.019(9) Å,
respectively. The bridging oxo ligand is at an average
distance of 1.791(9) Å from each Fe atom, and the Fe-O-
Fe bond angle is 171.6(3)°. These parameters are typical for
ferric µ-oxo dimers19 and reflect π-bonding within the Fe-
O-Fe core.20
The crystal structure of compound 17 is a disordered structure
of low quality, but it confirms the atom connectivity and the
discussed structural features without any doubt. The asymmetric
unit in 17 contains two crystallographically independent complex
molecules, a PF6- anion, a five-coordinate sodium cation, 14 water,
and three methanol molecules. The five-coordinate sodium ion in
a trigonal-bipyramidal environment is interesting but not unprec-
edented; there are several examples in the literature.18 Disorder was
found in the structure which might be due to a binding site
competition between water and methanol molecules at the sodium
cation. Four coordination sites are occupied by water molecules,
but the fifth site binds water and methanol in a ∼50:50 ratio.
Therefore, the coordinating oxygen atom O(540) was refined with
full occupation whereas the carbon atom C(541) was refined with
a 50% occupation. The disorder continues throughout greater parts
of the complicated hydrogen-bonding network of water and
methanol molecules. Water molecule positions O(450), O(460),
O(470), O(475), O(480), O(485), O(545), O(550), O(555), and
methanol molecule O(490)-C(491) are affected and were refined
with fractional occupation factors.
A water molecule of crystallization in 18 was found to be
disordered over two sites. The split positions were refined with
occupancies of 87% and 13% and restrained thermal displacement
parameters using the EADP instruction of ShelXL. A methanol
molecule in 20 was found to be disordered at a center of inversion.
An occupation factor of 0.5 was therefore assigned to the O and C
atoms.
Results and Discussion
Figures 2 and 3 show the structures of compounds 18,
19, 20, and 21. Table 1 summarizes the crystal data of the
Preparation of Ligands and Complexes. The synthetic
routes employed are depicted in Scheme 1. Treatment of
(17) Sheldrick, G. M. SHELXL-97. Program for Crystal Structure Refine-
ment; University of Go¨ttingen: Go¨ttingen, Germany, 1997.
(19) (a) Kurtz, D. M. Chem. ReV. 1990, 90, 585. (b) Roelfes, G.; Lubben,
M.; Chen, K.; Ho, R. Y. N.; Meetsma, A.; Genseberger, S.; Hermant,
R. M.; Hage, R.; Mandal, S. K.; Young, V. G.; Zang, Y.; Kooijman,
H.; Spek, A. L.; Que, L.; Feringa, B. L. Inorg. Chem. 1999, 38, 1929.
(c) Musie, G.; Lai, C. H.; Reibenspies, J. H.; Sumner, L. W.;
Darensbourg, M. Y. Inorg. Chem. 1998, 37, 4086. (d) Berry, J. F.;
Bill, E.; Garcia-Serres, R.; Neese, F.; Weyhermu¨ller, T.; Wieghardt,
K. Inorg. Chem. 2006, 45, 2027.
(18) (a) Malkowsky, I. M.; Fro¨hlich, R.; Griesbach, U.; Pu¨tter, H.;
Waldvogel, S. R. Eur. J. Inorg. Chem. 2006, 1690. (b) Hanna, T. A.;
Liu, L. H.; Angeles-Boza, A. M.; Kou, X. D.; Gutsche, C. D.; Ejsmont,
K.; Watson, W. H.; Zakharov, L. N.; Incarvito, C. D.; Rheingold, A.
L. J. Am. Chem. Soc. 2003, 125, 6228. (c) Linti, G.; Coban, S.; Dutta,
D. Z. Anorg. Allg. Chem. 2004, 630, 319. (d) Shen, L.; Jing, Z. M.
Acta Crystallogr. Sect. C-Cryst. Struct. Commun. 2002, 58, M591.
(20) Kappock, T. J.; Caradonna, J. P. Chem. ReV. 1996, 96, 2659.
2212 Inorganic Chemistry, Vol. 46, No. 6, 2007

Iron Complexes of New Pentadentate Ligands
Figure 1. Thermal ellipsoid plot (30% probability) of 17 with atom-
numbering scheme; hydrogen atoms, solvent molecules, and sodium salts
in the asymmetric unit are omitted for clarity.
Figure 3. Perspective view (30% probability) and atom labels of (a) 19
and (b) 21; hydrogen atoms and the solvent molecules are not shown here
for clarity.
Table 2. Selected Bond Lengths (Å) of the Fe(III) Complexes
bond
lengths (Å)
17
18
19
20
1.9391(8) 2.001(2)
1.941(1) 1.9943(9) 1.952(2)
1.972(1)
21
Fe(1)-O(1)
Fe(1)-O(20) 1.972(4) 1.975(1)
Fe(1)-O(24) 2.075(4) 1.945(1)
1.792(4)
-
-
-
-
Fe(1)-N(1)
Fe(1)-N(4)
Fe(1)-N(7)
Fe(1)-Cl(1)
Fe(1)-N(21)
Fe(1)-N(31)
Fe(2)-O(1)
Fe(2)-O(54) 1.986(4)
Fe(2)-O(50) 2.049(4)
Fe(2)-N(31) 2.223(5)
Fe(2)-N(34) 2.204(5)
Fe(2)-N(37) 2.255(5)
2.237(5) 2.214(2)
2.277(5) 2.166(2)
2.168(5) 2.191(2)
2.211(1) 2.208(1)
2.205(1) 2.214(1)
2.161(1) 2.171(1)
2.237(2)
2.172(2)
2.196(2)
-
-
-
2.2762(5)
-
2.2831(3) -
-
-
-
1.981(2)
-
-
-
-
-
-
-
1.975(1)
-
-
-
-
-
-
-
-
1.790(4)
-
-
-
-
-
-
-
-
-
-
-
-
Cl- and N3- anions. All the mononuclear Fe(III) complexes
contain a pentadentate macrocycle with a 1,4,7-triazacy-
clononane-1,4-diacetato fragment. All these structures contain
the ferric ion in an octahedral environment, which is slightly
trigonally distorted, as evidenced by the N-Fe-O and
N-Fe-Xaxial angles larger than 100°. Each nitrogen atom
of the macrocycle is unique in its coordination to the iron;
therefore, the Fe-N(macrocycle) distances in 18-21 are all
slightly different, though all are around 2.20 ( 0.03 Å. The
Figure 2. Thermal ellipsoid plots of (a) 18 and (b) 20 showing 30%
probability displacement ellipsoids and the atom-numbering scheme;
hydrogen atoms and solvent molecules are omitted for clarity.
Fe(III) complexes, and selected bond lengths are listed in
Table 2. The crystal structures of the Fe(III)-chloro and Fe-
(III)-azido complexes appear to be similar except for the
Inorganic Chemistry, Vol. 46, No. 6, 2007 2213

Song et al.
Table 3. Zero-Field Mo¨ssbauer Parameters for Fe(III) Complexes with
L1, L2, and L3 at 80 K
Fe-O distances are all likewise similar, averaging ∼1.97
Å, though the Fe-O bond trans to the N-R groups are
uniformly longer by 0.03 Å when R ) benzyl and 0.05 Å
when R ) isopropyl. The Fe-Cl bond lengths in 18 and 20
(∼2.28 Å) and the Fe-N3 bond lengths in 19 and 21 (∼1.98
Å) are typical for Fe(III) chloro and azido complexes.
Magnetic and Spectroscopic Measurements. Oxo-
bridged di-iron(III) complexes generally exhibit strong
antiferromagnetic coupling of the two iron centers through
the nearly linear µ-oxo unit.19 Compounds 17 and 24 are no
different; their measured effective magnetic moments at room
temperature are 3.02 µB and 2.50 µB, respectively, values
far less than the spin-only value for high-spin Fe(III) species
(S ) 5/2; µB ) 5.92). These values decrease monotonically
to reach plateaus of less than 1 µB at low temperatures,
indicating a diamagnetic ground state in each case. Fitting
of these data using the spin Hamiltonian H ) -2J S1‚S2 +
gâB‚S (where S1 ) S2 ) 5/2, J is the exchange coupling
constant, g is the Lande´ factor, â is the Bohr magneton, and
B is the magnetic field) yields isotropic J values of -80 and
-113 cm-1 for 17 and 24, respectively. The electronic g
values in each case were set to 2, and the magnetic data
were fitted taking into account paramagnetic impurities with
S ) 5/2 of 2.5% and 0.3% for 17 and 24, respectively.
Magnetic susceptibility measurements of powdered samples
of compounds 18, 19, 20, 21, 22, and 23 show nearly
temperature-independent magnetic moments of 5.85 µB, 5.80
µB, 5.83 µB, 5.78 µB, 5.80 µB, and 5.76 µB at 290 K,
respectively, which are clearly indicative of five unpaired
electrons and are in good agreement with the spin-only value
of 5.92 µB for octahedral high-spin Fe(III) ions. Further
investigation of the azido-Fe(III) complexes with various
temperatures (4.2-300 K) and various field strengths (1, 4,
and 7 T; VTVH data) allows the determination of zero-field
splitting parameters. The observed curves of Mmol/Ngâ vs
Hâ/kT for the azido-Fe(III) complexes can be well fit with
the spin Hamiltonian in eq 1:
compounds
δ, mm/s
∆EQ, mm/s
17
18
19
21
23
24
0.44
0.41
0.46
0.46
0.56
0.45
1.52
0.58
0.61
0.82
0.47
1.52
Table 4. Electronic Spectra of Azido-Fe(III) Complexes in CH3CN
Solution
complexes
λ
max, nm (ꢀ, L/mol‚cm)a
19
21
23
250 (1.3 × 104), 315 (sh), 420 (4.4 × 103)
250 (5.3 × 103), 310 (sh), 435 (1.9 × 103)
246 (9.6 × 103), 315 (sh), 419 (3.3 × 103).
a sh ) shoulder
by the susceptibility measurement above). The D value here
is far below the typical range 3-15 cm-1 found for that of
high-spin Fe(III) ions in porphyrins,21 as expected for non-
heme species with lesser tetragonal distortion.22,23 As the
6
zero-field splitting of the A1 ground state in a simplistic
picture24 is due to spin-orbit coupling with the first excited
4
quartet term A2, the low D value and the correspondingly
6
large A1/4A2 reflect a smaller 10Dq splitting of the eg and
t2g levels in the TACN-derived complexes as compared to
iron-porphyrin complexes.
Table 3 lists the results of zero-field Mo¨ssbauer measure-
ments of the Fe(III) complexes. The isomer shifts of these
complexes are in the range of 0.45-0.56 mm/s, typical for
high-spin octahedral Fe(III) complexes.10 The quadrupole
splittings for the mononuclear complexes are small and
between 0.47 and 0.72 mm/s, reflecting the high symmetry
of the high-spin d5 electron configuration. For the µ-oxo
dimers, the quadrupole splittings are around 1.5 mm/s, which
are in the normal range for µ-oxo-bridged dinuclear Fe(III)
compounds.25
UV-vis spectra of the azido-Fe(III) complexes have been
measured in MeCN solution at room temperature, and Table
4 summarizes the data. The spectra contain azide-to-Fe
charge-transfer bands in the range of 419-435 nm which
are so assigned due to their absence in the corresponding
chloro complexes. Similar assignments have been made in
the case of azido Fe(cyclam) complexes.10
Electrochemistry. The electrochemistry of the azido-Fe-
(III) complexes 19, 21, and 23 in MeCN has been studied
by cyclic and square-wave voltammetry. Table 5 summarizes
the redox potentials of the compounds, which are referenced
versus the ferrocene/ferrocenium couple. The chloro com-
plexes were also studied, but only irreversible waves were
observed.
1
3
H ) gâBBBS + D BS2z - S(S + 1) + E/D(BSx2 - BS2y) (1)
[
]
In which D represents the axial zero-field splitting
parameter and E/D parametrizes the rhombicity of the zero-
field splitting tensor. Fitting of the VTVH data (see Figure
S1) with S ) 5/2 yields g ) 2.0 (fixed), E/D ) 0.33 for all
three complexes; D ) 0.48 cm-1, θ ) -0.61 K for 19; D )
0.48 cm-1, θ ) -0.85 K for 21 and D ) 0.44 cm-1, θ )
-0.50 K for 23. All these parameters are in line with the d5
electron configuration of the high-spin Fe(III) ions in their
trigonally distorted environment.
The X-band EPR spectra of compounds 18-23 recorded
in normal (perpendicular) mode at 4 K display rhombic
spectra with a broad signal at g ) 4.28 typical of an
octahedral high-spin iron(III) ion. The EPR spectra of the
azido-Fe(III) complexes were successfully simulated using
the effective g value formalism for the zero-field split
Kramers doublets of a rhombic S ) 5/2 system with E/D )
0.33 and D ) 0.5 ( 0.2 cm-1 (the D values were determined
(21) Debrunner, P. G. In Iron Porphyrins; Lever, A. B. P., Gray, H. B.,
Eds.; VCH: Weinheim, 1989; Vol. 3, p 137.
(22) Bocˇa, R. Coord. Chem. ReV. 2004, 248, 757.
(23) Berry, J. F.; Bill, E.; Bothe, E.; Neese, F.; Wieghardt, K. J. Am. Chem.
Soc. 2006, 128, 13515.
(24) (a) Maltempo, M. M. J. Chem. Phys. 1974, 61, 2540. (b) Maltempo,
M. M.; Moss, T. H. Q. ReV. Biophys. 1976, 9, 181.
(25) Ju¨stel, T.; Muller, M.; Weyhermu¨ller, T.; Kressl, C.; Bill, E.;
Hildebrandt, P.; Lengen, M.; Grodzicki, M.; Trautwein, A. X.; Nuber,
B.; Wieghardt, K. Chem.-Eur. J. 1999, 5, 793.
2214 Inorganic Chemistry, Vol. 46, No. 6, 2007

Iron Complexes of New Pentadentate Ligands
stable 1,4,7-triazacyclononane-1,4-diacetate backbone,14 the
redox processes can be safely assigned to the metal center.
Table 5. Redox Potentials of Azido-Fe(III) Complexes
complexes
E, V vs Fc+/Fca
Three electron-transfer series involving six-coordinate Fe-
(II), Fe(III), and Fe(IV) complexes have been recently
reported, all based on the macrocyclic ligand 1,4,8-trimethyl-
1,4,8,11-tetraazacyclotetradecane-11-acetate (Me3cyclam-
acetate) which enforces a compressed octahedral geometry
in the complexes.23,26 Here, a corresponding series of
trigonally distorted complexes, comprised of 19 and its
reduction product (19red) and oxidation product (19ox) is
presented. A spectroscopic study of this series is presented
so that the highly oxidized complex 19ox may be considered
in the context of the growing number of Fe(IV) complexes
which have recently been reported.23,26-29
19
21
23
-749 mV r, 1112 mV r
-738 mV r, 1157 mVb
-734 mV r, 1116 mVb
a For reversible (r), one-electron processes E1/2 is given (E1/2 ) Ep,ox
+
E
p,red)/2. b The redox potentials of the quasi-reversible peak at -20 °C have
been given. All other measurements have been done at room temperature.
Solvents: MeCN; supporting electrolyte: 0.10 M [N(n-Bu)4]PF6.
The reduction potential of 19 (-749 mV) is the same as
the low-spin azido complex [(cyclam-acetate)FeN3]+,11 de-
spite the difference in spin state and the charge difference
of the two complexes. Most likely, these two effects
fortuitously cancel each other in this case. Likewise, the
oxidation potential of 19 is very close to that of [(Me3cyclam-
acetate)FeN3]+.26 Thus, 19 has a large potential range in
which the Fe(III) species is stable.
Combined UV-vis/electrochemistry measurements of 19
were performed at -20 °C by use of an optically transparent
thin-layer electrochemical cell (OTTLE cell). The electronic
spectra of oxidized and reduced species of 19 are shown in
Figure 5 (21 was also studied, and its spectra are given in
the Supporting Information as Figure S2). Upon reduction,
solutions of 19 bleach as the absorptions at 250, 315, and
420 nm all lose intensity. Upon oxidation, however, the
solution becomes more colored as all three bands undergo a
bathochromic shift and a new band is observed at 780 nm
indicative of the Fe(IV) species 19ox. Similar long-
wavelength bands have been observed in the previously
studied [(Me3cyclam-acetate)FeIVN3]2+ species,26 as well as
in Fe(IV)-oxo species for which they have been assigned
as d-d transitions.30
Oxidation and reduction of 19 (and 21; see Figure S3)
were also monitored by IR spectroscopy since two intense
absorption bands at ∼2066 and ∼1674 cm-1 corresponding
to the azide stretch and the CdO stretch of the pendent
acetate groups appear in a region where the solvent and
Figure 4. Cyclic voltammogram of (a) 19 (1 mM) in MeCN at room
temperature (25 °C, 0.10 M [N(n-Bu)4]PF6 supporting electrolyte; glassy
carbon working electrode; scan rates 0.05, 0.10, 0.20, 0.40 V/s). (b) 21 in
CH3CN solution (0.10 M [N(n-Bu)4]PF6, recorded at a scan rate of 0.20
V/s at a glassy carbon electrode at -20 °C. The small irreversible peak at
+1.1 V corresponds to an impurity of the chloro complex.
(26) Berry, J. F.; Bill, E.; Bothe, E.; Weyhermu¨ller, T.; Wieghardt, K. J.
Am. Chem. Soc. 2005, 127, 11550.
(27) (a) Slep, L. D.; Mijovilovich, A.; Meyer-Klaucke, W.; Weyhermu¨ller,
T.; Bill, E.; Bothe, E.; Neese, F.; Wieghardt, K. J. Am. Chem. Soc.
2003, 125, 15554. (b) Costas, M.; Rohde, J. U.; Stubna, A.; Ho, R.
Y. N.; Quaroni, L.; Mu¨nck, E.; Que, L. J. Am. Chem. Soc. 2001, 123,
12931. (c) Zheng, H.; Yoo, S. J.; Mu¨nck, E.; Que, L. J. Am. Chem.
Soc. 2000, 122, 3789. (d) Dong, Y. H.; Que, L.; Kauffmann, K.;
Mu¨nck, E. J. Am. Chem. Soc. 1995, 117, 11377. (e) Kostka, K. L.;
Fox, B. G.; Hendrich, M. P.; Collins, T. J.; Rickard, C. E. F.; Wright,
L. J.; Mu¨nck, E. J. Am. Chem. Soc. 1993, 115, 6746. (f) Rohde, J. U.;
In, J. H.; Lim, M. H.; Brennessel, W. W.; Bukowski, M. R.; Stubna,
A.; Mu¨nck, E.; Nam, W.; Que, L. Science 2003, 299, 1037.
(28) Collins, T. J.; Kostka, K. L.; Mu¨nck, E.; Uffelman, E. S. J. Am. Chem.
Soc. 1990, 112, 5637.
All three azido complexes show two important electro-
chemical processes: a one-electron reduction between -734
and -749 mV and a one-electron oxidation process above
1100 mV which is fully reversible for 19, but less so for the
other two complexes (see Figure 4). The second process is
irreversible for 21 and 23 at room temperature, most likely
due to an E-C process, but when measured at -20 °C, these
waves become quasi-reversible, indicating a greater stability
of the oxidized forms at lower temperatures. Since the
pentadentate ligands used in this study contain the redox-
(29) Lim, M. H.; Rohde, J. U.; Stubna, A.; Bukowski, M. R.; Costas, M.;
Ho, R. Y. N.; Mu¨nck, E.; Nam, W.; Que, L. Proc. Natl. Acad. Sci.
U.S.A. 2003, 100, 3665.
(30) Decker, A.; Rohde, J. U.; Que, L.; Solomon, E. I. J. Am. Chem. Soc.
2004, 126, 5378.
Inorganic Chemistry, Vol. 46, No. 6, 2007 2215

Song et al.
Figure 5. UV-vis spectra of 19red, 19, and 19ox (2.7 × 10-4 mol/L) in
MeCN at -20 °C (see Supporting Information for 21).
Figure 7. Zero-field Mo¨ssbauer spectra measured at 80 K of (A) electro-
oxidation at -40 °C and (B) electro-reduction at -20 °C of 19 (1 mM) in
CH3CN. (a) 19ox, (b) 19, (c) 19red, (d) 17.
donation of azide to Fe(IV) and Fe(II) relative to the high-
spin Fe(III) ion. The CdO and azide stretches are affected
significantly and characteristically with the different oxida-
tion states of iron as +2, +3, and +4, as has been previously
shown for the analogous Me3cyclam-acetate complexes.26
Similar IR profiles have been observed during electro-
oxidation and reduction of 21 (see Figure S3).
Electrochemically generated solutions of 57Fe isotopically
enriched samples of the oxidation and reduction products of
19 were also examined by Mo¨ssbauer spectroscopy in order
to provide further support of the Fe(IV) assignment in 19ox
and determine the spin states of the new Fe(II) and Fe(IV)
species.
In 19red, the Mo¨ssbauer spectrum (Figure 7B) shows a
strong signal (species c) with quadrupole splitting ∆EQ )
3.013 mm/s and isomer shift δ ) 1.13 mm/s, which allows
unambiguous assignment as a high-spin Fe(II) species which
accounts for 78% of the spectral intensity. The Mo¨ssbauer
parameters of 19red are comparable to those of similar high-
spin Fe(II) compounds found in literature.10,26 The remaining
22% of intensity (species d) is fitted to the L1-Fe(III)-O-
Fe(III)-L1 species (δ ) 0.44 mm/s, ∆EQ ) 1.52 mm/s), most
likely formed via aerial oxidation of the sample.
The Fe(IV) species was electrochemically generated at
-40 °C in MeCN and coulometry was stopped when a
charge corresponding to 95% of 1 F/mol was passed. Here,
the main feature of the Mo¨ssbauer spectrum (Figure 7A) is
a doublet (species a) with large ∆EQ of 2.76 mm/s and a
low δ value of 0.01 mm/s, which accounts for 81% of the
Figure 6. IR spectral change observed during electro-oxidation and electro-
reduction of 19 (1 mM) in MeCN at -20 °C.
electrolyte do not absorb. Both of these bands exhibit
significant changes upon oxidation and reduction and the
spectra (Figure 6) show almost 90%∼100% recovery of the
starting material on re-oxidation of 19red or re-reduction of
19ox, indicating that no ligands are removed during the
coulometry.
During the one electron reduction process of 19, both the
azide stretch and the CdO stretch of the pendent acetate
groups decrease dramatically as new bands at 2055, 2005,
and 1635 cm-1 appear and increase. The CdO stretch shifts
to lower energy as the oxidation state of the metal is
decreased, suggesting a weaker Fe-O bond with decreased
oxidation state, as has been quantified previously.23,31 There
are two new azide bands at around 2000 cm-1, which may
be due to the presence of two different orientations of the
azido ligand (the Fe-N-N linkage is bent, and therefore
the azido group has four different directions in which it may
be bent over the ligand).
For the one-electron oxidation process, the CdO stretch
moves to higher energy (1722 cm-1) while the azide stretch
moves to lower frequency (1996 cm-1). In both cases, the
azide stretch shifts to lower energy, emphasizing greater π
(31) Berry, J. F.; Bill, E.; Bothe, E.; George, S. D.; Mienert, B.; Neese, F.;
Wieghardt, K. Science 2006, 312, 1937.
2216 Inorganic Chemistry, Vol. 46, No. 6, 2007

Iron Complexes of New Pentadentate Ligands
Figure 9. Mo¨ssbauer spectrum of 21 after photolysis at room temperature
in MeCN (irradiation wavelength: 330 nm). The spectra were fit with three
components: species (1) is a high-spin Fe(II) species; species (2) is 21,
and species (3) is L2-57Fe-O-57Fe-L2.
Photolysis of 21. Photolysis of a 10 mM solution of 21
in dry acetonitrile was performed at room temperature with
a Rayonet Photochemical Reactor (RPR-100) equipped with
480, 419, or 330 nm tubes. Since the photolysis products of
azido-manganese complexes have been shown to be wave-
length-dependent,33 we decided to examine the photochem-
istry of 21 using both visible (480, 419 nm) and UV (330
nm) light. The photolysis process was followed optically by
changes in the color of the species from orange to pale beige
and monitored by IR spectroscopy and Mo¨ssbauer spectros-
copy.
When the photolysis was carried out in air at 330 nm at
298 K for 4 h, the solution is bleached, and IR spectra show
that the azide stretch gradually loses intensity. The zero-
field Mo¨ssbauer spectrum (80 K, Figure 9) of the product
solution shows an intense quadrupole doublet, species 1, with
δ ) 1.15 mm/s and ∆EQ ) 2.95 mm/s, which accounts for
79% of the spectral intensity. The isomer shift and quadru-
pole splitting parameters clearly indicate that species 1 is a
high-spin Fe(II) complex resulting from the photoreduction
of the starting material. Unreacted 21 is also present at 9%
relative intensity, species 2 (δ ) 0.46 mm/s and ∆EQ ) 0.82
mm/s). The remaining 12% of the intensity is fit with
parameters similar to those of L2-57Fe-O-57Fe-L2 (species
3, δ ) 0.58 mm/s and ∆EQ ) 1.33 mm/s), most likely the
result of oxidation of species 1 by air. Irradiation at 419 and
480 nm gave similar results.
Although photolysis of 21 in liquid solution proceeds
almost completely via photoreduction, its photochemistry was
further investigated in frozen solution in an attempt to
promote photo-oxidation. Photolysis temperature has been
found to be a very critical experimental parameter governing
the product distribution.10 To date, Fe(V) nitrido species have
only been observed during low-temperature photolysis.10,11
Samples for photolysis at 77 K were prepared via the
previously described method II.10 Photolysis was performed
on a stirred suspension of a fine snow of 21 in acetonitrile
kept in liquid nitrogen with stirring with 419 nm irradiation
Figure 8. Mo¨ssbauer spectra of 19ox measured in applied magnetic fields
(perpendicular to the γ source) of 1.0, 4.0, and 7.0 T all at 4.2 K. These
were simulated assuming fast spin relaxation with a spin Hamiltonian
analysis with the following parameters: S ) 2, δ ) 0.01, ∆EQ ) 2.77, D
) 6.68 cm-1, E/D ) 0, Axx ) -3.1 T, Ayy ) -0.3 T, Azz ) +12.8 T. Due
to the presence of remaining high-spin Fe(III) impurities and the large axial
zero-field splitting value, this fit is not unique as described in the text.
spectral intensity and is more than 0.3 mm/s lower than that
of 19 itself, which indicates that 19ox is a genuine Fe(IV)-
azido complex. The δ value is also comparable to those found
in Fe(IV)-oxo compounds,11,29 though higher than in the
five-coordinate Fe(IV) species [FeIV(MAC*)Cl]- (δ ) -0.03
mm/s, MAC* is a macrocyclic tetra-anionic ligand28). The
∆EQ value of 2.76 mm/s is larger than those reported with
either low-spin ferryl species (S ) 1)11,29 or high-spin five-
coordinate species (S ) 2).28 Similar ∆EQ values have,
however, been observed for low-spin non-oxo Me3cyclam-
acetate Fe(IV) species having axial chloro or fluoro ligands.23
Mo¨ssbauer spectra of 19ox were measured at 4.2 K in
applied magnetic fields of 1, 4, and 7 T (Figure 8). The
spectra show very little magnetic splitting, in accord with a
large axial zero-field splitting. Unfortunately, since the
splitting is so small, the magnetically perturbed spectra may
be fit in a number of different ways.32 Reasonable fits include
a high-spin model (S ) 2) having D ) 6.7 cm-1, and A )
{-3.1, -0.3, +12.8} T, although this fit is no better than
low spin (S ) 1) fits with D ≈ 6-7 cm-1 and A ) {-1 to
-2, -9 to -11, +13 to +17} T. Thus, the spin state cannot
be determined at this time.
(32) Fitting of these spectra is also complicated by the residual high-spin
Fe(III) species.
(33) Meyer, K.; Bendix, J.; Metzler-Nolte, N.; Weyhermu¨ller, T.; Wieghardt,
K. J. Am. Chem. Soc. 1998, 120, 7260.
Inorganic Chemistry, Vol. 46, No. 6, 2007 2217

Song et al.
Chart 1
Figure 10. Mo¨ssbauer spectrum (80 K) of photolysis of 21 at 77 K in
MeCN for 4 h (irradiation wavelength: 330 nm). The fit (solid line) is
comprised of unreacted 21 (species 1) and two high-spin Fe(II) species 2
and species 3 with large quadrupole splitting.
Scheme 2
Scheme 3
in a Rayonet reactor. After photolysis of the azide complex
for 4 h, the color of the snow faded from orange to white. A
portion of the photolyzed snow was packed in a Mo¨ssbauer
cup for analysis.
The features of the resulting zero-field Mo¨ssbauer spec-
trum (Figure 10) are two intense quadrupole doublets, species
2 with δ ) 1.02 mm/s, ∆EQ ) 2.43 mm/s and species 3
with δ ) 1.11 mm/s and ∆EQ ) 2.99 mm/s, respectively,
which account for almost 70% of the spectral intensity. These
isomer shifts and quadrupole splitting parameters can be
assigned to two high-spin Fe(II) species, and species 3 is
very similar to the photoreduction product from the liquid
solution measurement. We suggest therefore that species 3
has MeCN occupying the sixth coordination site, whereas
species 2 may be five-coordinate. The remaining 30% of
the intensity was fit with parameters of the starting material.
Again, irradiation at 330 and 480 nm gave similar results.
It is our observation that photolysis of high-spin Fe(III)-
azido complexes generally yields photoreduced products,
whereas photo-oxidation to Fe(V)-nitrido species has only
been directly observed upon photolysis of low-spin precur-
sors. This observation is supported by the data presented in
this paper, as well as in previously reported systems.10-12,25,31
The most poignant examples are from studies of ferric azido
complexes with supporting cyclam-based ligands. For ex-
ample, the complexes cis- and trans-[(cyclam)Fe(N3)2]+ were
both subjected to photolysis. Whereas photolysis of the trans
isomer, which is low spin, led to the observation of a species
formulated as trans-[(cyclam)FeN(N3)]+ containing the Fe(V)
ion, no such species was observed in the case of the cis
isomer, which is high spin.10
FeN3 and LFeVtN are not expected to be very different from
one another12 and neither are high-spin LFeN3 and LFeII.23
The energies of these states are also estimated in Chart 1,
and four Frank-Condon transitions are drawn between these
states: (1) and (3) correspond to photo-oxidation processes,
and (2) and (4) are photoreductions. It is clear that the
processes most kinetically favored will be those for which
the reorganization energy is minimized, i.e., processes (1)
and (4). Importantly, these are also processes for which the
total spin of the system is conserved, a specific requirement
for photochemical reactions since photons carry no angular
momentum. Thus, photo-oxidation is favored from a low-
spin Fe(III) precursor and photoreduction is favored when
the Fe(III) precursor is high-spin.
It is important to clarify this last statement in view of
dinuclear Fe(IV)-N-Fe(III) complexes which are known
to form during photolysis of high-spin ferric azido com-
plexes.10,25 It is often presumed that these species form via
the reaction sequence given in Scheme 2, implying the
intermediacy of an FeVtN species.
A simple four-state model is proposed to account for the
different reactivity as a function of the spin state of the iron
(see Chart 1).
The four states involved are shown as four different
potential energy curves in Chart 1: (1) low-spin LFeN3
(purple curve), (2) high-spin LFeN3 (orange curve), (3)
•
LFeVtN + N2 (red curve), and (4) LFeII + N3 (blue curve).
Given that the central reaction in this sequence is likely
disfavored, the formation of these mixed-valent FeIII/IV
species may involve reactions given in Scheme 3, which
feature a µ-azido di-Fe(II/III) intermediate.
The minima of these curves are offset in the direction of the
abscissa by their estimated average Fe-ligand bond dis-
tances. In this regard, the pair of states involving low-spin
2218 Inorganic Chemistry, Vol. 46, No. 6, 2007

Iron Complexes of New Pentadentate Ligands
Conclusions
19red, and 19ox show very characteristic IR and UV-vis
profiles. It should be noted that 19ox represents a potential
precursor to a new Fe(VI)-nitrido complex,31 a property we
plan to pursue. Photolysis of 21 at room temperature with
different wavelengths such as 480, 419, and 330 nm shows
that the formation of high-spin Fe(II) species or sometimes
Fe(III)-O-Fe(III) dinuclear compounds, but photolysis
performed at 77 K shows two types of high-spin Fe(II)
species, which are likely five- and six-coordinate compounds,
respectively. A simple four-state model has been put forth
to explain the observation that high-spin Fe(III) azido
complexes undergo primarily photoreduction.
To summarize, three new redox-innocent pentadentate
ligands containing the 1,4,7-triazacyclononane-1,4-diacetate
motif have been synthesized and their coordination chemistry
with Fe(III) has been studied. Eight new Fe(III) complexes
have been synthesized with these ligands. X-ray structures
of five Fe(III) complexes reveal that three nitrogen atoms
in the TACN backbone coordinate facially to Fe(III) ions
with two additional coordination sites occupied by two
pendent acetate groups and a chloride or azide anion or a
bridging oxygen atom as the sixth ligand. Characterization
of the iron complexes by magnetic susceptibility measure-
ments and zero-field Mo¨ssbauer measurements are in good
agreement with the high-spin state of the Fe(III) ions. EPR
spectra of the azido-Fe(III) complexes have effective g
values around 4.3, which can be well fit with E/D ) 0.33
and D ) 0.46 ( 0.2 cm-1. The zero-field splitting parameters
agree well with those obtained by VTVH magnetic suscep-
tibility measurements and indicate a small 10Dq splitting
between the eg and t2g levels. Spectroelectrochemistry of 19
has shown that the corresponding Fe(IV) species 19ox can
be generated coulometrically at -40 °C. Compounds 19,
Acknowledgment. We are grateful to the Fonds der
Chemischen Industrie for financial support. J.F.B. thanks the
Alexander von Humboldt Stiftung for support in the form
of a postdoctoral fellowship.
Supporting Information Available: Figures S1, S2, and S3,
as well as X-ray crystallographic information in CIF format. This
material is available free of charge via the Internet at http://
pubs.acs.org.
IC062001J
Inorganic Chemistry, Vol. 46, No. 6, 2007 2219