Importance of the C-H···N weak hydrogen bonding on the coordination structures of manganese(III) porphyrin complexes

Article abstract of DOI:10.1021/ic0206138

The reactions between Mn(Por)Cl and Bu₄N⁺CN^- have been examined in various solvents by UV-vis and ¹H NMR spectroscopy, where Por's are dianions of meso-tetraisopropylporphyrin (TⁱPrP), meso-tetraphenylporphyrin (TPP), meso-tetrakis(p-(trifluoromethyl)phenyl)porphy(in (p-CF₃-TPP), meso-tetramesitylporphyrin (TMP), and meso-tetrakis-(2,6-dichlorophenyl)porphyrin (2,6-Cl₂-TPP). Population ratios of the reaction products, Mn(Por)(CN) and [Mn(Por)(CN)₂]^-, have been sensitively affected by the solvents used. In the case of Mn(TⁱPrP)Cl, the following results are obtained: (i) The bis-adduct is preferentially formed in dipolar aprotic solvents such as DMSO, DMF, and acetonitrile. (ii) Both the mono- and bis-adduct are formed in the less polar solvents such as CH₂Cl₂ and benzene though the complete conversion to the bis-adduct is achieved with much smaller amount of the ligand in benzene solution. (iii) Only the mono-adduct is formed in CHCl₃ solution even in the presence of a large excess of cyanide. (iv) Neither the mono- nor the bis-adduct is obtained in methanol solution. The results mentioned above have been explained in terms of the C-H···N and O-H···N hydrogen bonding in chloroform and methanol solutions, respectively, between the solvent molecules and cyanide ligand; hydrogen bonding weakens the coordination ability of cyanide and reduces the population of the bis-adduct. The importance of the C-H···N weak hydrogen bonding is most explicitly shown in the following fact: while the starting complex is completely converted to the bis-adduct in CH₂Cl₂ solution, the conversion from the mono- to the bis-adduct is not observed even in the presence of 7000 equiv of Bu₄N⁺CN^- in CHCl₃ solution. The effective magnetic moments of the bis-adduct has been determined by the Evans method to be 3.2 μB at 25 °C, suggesting that the complex adopts the usual (d_xy)²(d_xz, d_yz)² electron configuration despite the highly ruffled porphyrin core expected for [Mn(TⁱPrP)(CN)₂]^-. The spin densities of [Mn(TⁱPrP)(CN)₂]^- centered on the π MO have been determined on the basis of the ¹H and ¹³C NMR chemical shifts. Estimated spin densities are as follows: meso-carbon, -0.0014; α-pyrrole carbon, -0.0011; β-pyrrole carbon, +0.0066; pyrrole nitrogen, -0.022. The spin densities at the pyrrole carbon and meso nitrogen atoms are much smaller than those of the corresponding [Mn(TPP)(CN)₂]^-, which is ascribed to the nonplanar porphyrin ring of [Mn(TⁱPrP)(CN)₂]^-. This study has reveled that the C-H···N weak hydrogen bonding is playing an important role in determining the stability of the manganese(III) complexes.

Full text of DOI:10.1021/ic0206138

Inorg. Chem. 2003, 42, 2301−2310
Importance of the C−H‚‚‚N Weak Hydrogen Bonding on the
Coordination Structures of Manganese(III) Porphyrin Complexes
,†,‡
Akira Ikezaki† and Mikio Nakamura*
Department of Chemistry, School of Medicine, Toho UniVersity, Tokyo 143-8540, Japan, and
DiVision of Biomolecular Science, Graduate School of Science, Toho UniVersity,
Funabashi 274-8510, Japan
Received October 14, 2002
+
The reactions between Mn(Por)Cl and Bu4N CN- have been examined in various solvents by UV−vis and 1H NMR
i
spectroscopy, where Por’s are dianions of meso-tetraisopropylporphyrin (TPrP), meso-tetraphenylporphyrin (TPP),
meso-tetrakis(p-(trifluoromethyl)phenyl)porphyrin (p-CF -TPP), meso-tetramesitylporphyrin (TMP), and meso-tetrakis-
3
(2,6-dichlorophenyl)porphyrin (2,6-Cl2-TPP). Population ratios of the reaction products, Mn(Por)(CN) and
i
[Mn(Por)(CN)2]-, have been sensitively affected by the solvents used. In the case of Mn(TPrP)Cl, the following
results are obtained: (i) The bis-adduct is preferentially formed in dipolar aprotic solvents such as DMSO, DMF,
and acetonitrile. (ii) Both the mono- and bis-adduct are formed in the less polar solvents such as CH Cl2 and
2
benzene though the complete conversion to the bis-adduct is achieved with much smaller amount of the ligand in
benzene solution. (iii) Only the mono-adduct is formed in CHCl3 solution even in the presence of a large excess
of cyanide. (iv) Neither the mono- nor the bis-adduct is obtained in methanol solution. The results mentioned
above have been explained in terms of the C−H‚‚‚N and O−H‚‚‚N hydrogen bonding in chloroform and methanol
solutions, respectively, between the solvent molecules and cyanide ligand; hydrogen bonding weakens the coordination
ability of cyanide and reduces the population of the bis-adduct. The importance of the C−H‚‚‚N weak hydrogen
bonding is most explicitly shown in the following fact: while the starting complex is completely converted to the
bis-adduct in CH Cl2 solution, the conversion from the mono- to the bis-adduct is not observed even in the presence
2
of 7000 equiv of Bu N+CN- in CHCl3 solution. The effective magnetic moments of the bis-adduct has been determined
4
by the Evans method to be 3.2 µB at 25 °C, suggesting that the complex adopts the usual (d )2(d , d )2 electron
xy
xz yz
i
configuration despite the highly ruffled porphyrin core expected for [Mn(TPrP)(CN)2]-. The spin densities of
i
[Mn(TPrP)(CN)2]- centered on the π MO have been determined on the basis of the 1H and 13C NMR chemical
shifts. Estimated spin densities are as follows: meso-carbon, −0.0014; R-pyrrole carbon, −0.0011; â-pyrrole carbon,
+0.0066; pyrrole nitrogen, −0.022. The spin densities at the pyrrole carbon and meso nitrogen atoms are much
smaller than those of the corresponding [Mn(TPP)(CN)2]-, which is ascribed to the nonplanar porphyrin ring of
i
[Mn(TPrP)(CN)2]-. This study has reveled that the C−H‚‚‚N weak hydrogen bonding is playing an important role
in determining the stability of the manganese(III) complexes.
Introduction
simple synthetic compounds as well as complicated biomole-
cules.1-3 Some years ago, La Mar and co-workers pointed
out that the electronic state of [Fe(TPP)(CN)2]- is perturbed
by the solvents due to the hydrogen bonding between the
coordinated cyanide and solvent molecules.4 We have
C-H‚‚‚X (X ) O, N, etc.) type weak hydrogen bondings
together with C-H‚‚‚π interaction are the focus of attention
in an interdisciplinary area encompassing organic chemistry,
inorganic chemistry, biochemistry, and material science,
because these weak interactions sometimes play a crucial
role for the determination of physicochemical properties of
(1) Desiraju G. R.; Steiner, T. The Weak Hydrogen Bond. In Structural
Chemistry and Biology; Oxford University Press: New York, 1999.
(2) Nishio, M.; Hirota, M.; Umezawa, Y. The CH/π interaction, eVidence,
nature, and consequences; Wiley-VCH: New York, 1998.
(3) Steiner, T. Chem. Commun. 1997, 727-734.
(4) La Mar, G. N.; Gaudio, J. D.; Frye, J. S. Biochim. Biophys. Acta
1977, 498, 422-435.
* To whom correspondence should be addressed. E-mail: mnakamu@
med.toho-u.ac.jp.
† School of Medicine.
‡ Graduate School of Science.
10.1021/ic0206138 CCC: $25.00 © 2003 American Chemical Society
Published on Web 03/01/2003
Inorganic Chemistry, Vol. 42, No. 7, 2003 2301

Ikezaki and Nakamura
recently reported that the ground-state electron configuration
of low-spin dicyano[meso-tetrakis(2,4,6-triethylphenyl)por-
phyrinato]ferrate(III), [Fe(Et-TPP)(CN)2]-, is sensitively
affected by the solvents used.5 For example, while the com-
plex exhibits the common (dxy)2(dxz, dyz)3 electron configu-
ration in nonpolar and dipolar aprotic solvents such as ben-
zene, toluene, DMSO, and DMF, it shows the less common
(dxz, dyz)4(dxy)1 electron configuration in a protic solvent such
as methanol. The results have been explained in terms of
the O-H‚‚‚N hydrogen bonding between methanol and the
coordinated cyanide ligands; the O-H‚‚‚N hydrogen bonding
stabilizes the cyanide pπ* orbital, which in turn stabilizes
the iron dπ (dxz and dyz) orbitals by the strong iron (dπ)-
Experimental Section
Abbreviations: TiPrP, TPP, p-CF3-TPP, TMP, 2,6-Cl2-TPP, and
OEP, dianions of meso-tetraisopropylporphyrin, meso-tetraphenyl-
porphyrin, meso-tetrakis(p-(trifluoromethyl)phenyl)porphyrin, meso-
tetramesitylporphyrin, meso-tetrakis(2,6-dichlorophenyl)porphyrin,
and 2,3,7,8,12,13,17,18-octaethylporphyrin; Mn(TiPrP)Cl, chloro-
(meso-tetraisopropylporphyrinato)manganese(III); Mn(TiPrP)(CN),
cyano(meso-tetraisopropylporphyrinato)manganese(III); [Mn(TiPrP)-
(CN)2]-Bu4N+, tetrabutylammonium dicyano(meso-tetraisopropyl-
t
porphyrinato)manganate(III); BuNC, tert-butyl isocyanide.
Materials. Bu4N+CN- was purchased from Aldrich. Highly pure
DMSO, DMF, methanol, ethanol, 2-propanol, and CCl4 were
purchased from Wako or Kanto. All the deuterated solvents were
purchased from Merck. CHCl3 and CDCl3 were washed several
times with concentrated sulfuric acid and then with dilute sodium
carbonate solution and water.9 They were then dried over K2CO3
and distilled in an argon atmosphere shortly before use. Acetonitrile,
benzene, and toluene were dried and then distilled.
1
cyanide (pπ*) interactions.5,6 Extensive H and 13C NMR
experiments together with EPR studies have revealed that
even CHCl3 and CH2Cl2 stabilize the (dxz, dyz)4(dxy)1 electron
configuration though the degree of stabilization is much
larger in CHCl3 than in CH2Cl2 solution. The results have
been ascribed to the C-H‚‚‚N weak hydrogen bonding
between these solvents and the coordinated cyanide ligands.1,5
Although few examples have been reported on the physi-
cochemical properties showing some difference between
CHCl3 and CH2Cl2 solutions, such a difference could be a
general matter in various aspects of coordination chemistry.
Scheidt and co-workers reported the formation constants of
the mono- and bis(cyanide) adducts in the reaction between
Mn(TPP)(OH) and KCN in methanol by UV-vis spectros-
copy.7 We have expected that the population ratios of these
products, [mono]/[bis], in various solvents could be a good
probe to elucidate the solvent effects. In this study, we have
examined the reactions between Mn(Por)Cl and Bu4N+CN-
in various solvents such as C6H6, C6H5CH3, CH2Cl2, CHCl3,
CH3CN, DMSO, DMF, and CH3OH, where Por stands for
TiPrP, TPP, and p-CF3-TPP. Herein, we report that the
population ratios are quite different among these solvents,
even between CH2Cl2 and CHCl3. We also report the spin
distribution on the porphyrin carbon and nitrogen atoms in
ruffled [Mn(TiPrP)(CN)2]- and compare them with those of
planar [Mn(TPP)(CN)2]- originally reported by Goff and
Hansen.8
Synthesis. Free-base porphyrins (TPP)H2, (p-CF3-TPP)H2, (TMP)-
H2, and (2,6-Cl2-TPP)H2 were prepared according to the litera-
tures.10-13 Manganese(III) complexes, Mn(Por)Cl, such as Mn-
(TPP)Cl, Mn(p-CF3-TPP)Cl, Mn(TMP)Cl, and Mn(2,6-Cl2-TPP)-
Cl were prepared by reaction of porphyrins (Por’s) with MnCl2 in
refluxing DMF solution.14
Synthesis of Mn(TiPrP)Cl. To a 500 mL three-neck flask with
a septum, reflux condenser, and argon inlet port were charged
(TiPrP)H2 (201 mg, 0.410 mmol),15,16 K2CO3 (568 mg), and CHCl3
(300 mL). The solution was purged with argon for 15 min and
was then refluxed. The methanol solution (20 mL) of MnCl2 (106
mg, 0.842 mmol) was added to the refluxed solution repeatedly
with the interval of 1 h. The total amount of MnCl2 added was 530
mg (4.20 mmol). The solution was then rotary evaporated to
dryness. The solid thus obtained was dissolved in 50 mL of CH2-
Cl2 and washed three times with 1 M HCl solution. The organic
layer was washed with water and dried over anhydrous Na2SO4
and was then rotary evaporated to dryness. Mn(TiPrP)Cl was
purified by column chromatography on alumina with chloroform
eluent. After rotary evaporation of the chloroform solution, a green
solid product Mn(TiPrP)Cl was recovered, which was vacuum-dried
at 60 °C. Yield: 187 mg (77%). UV-vis (CH2Cl2; λmax, nm (log
ꢀ)): 328 (4.4), 378 (4.5), 400 (4.5), 480 (4.9), 545 (3.6), 598 (3.7),
1
636 (4.0). H NMR (CD2Cl2, 25 °C; δ, ppm): -19.4 (8H, Py-H),
9.5 (4H, CH), 3.2 (24H, CH3). 13C NMR (CD2Cl2, 25 °C; δ, ppm):
475 (R-Py), -153 (â-py), 216 (meso), 47.9 (CH), 112.5 (CH3).
Synthesis of [Co(TiPrP)(CN)2]-Bu4N+. Co(TiPrP) was synthe-
sized by the literature method.17 A CH2Cl2 solution containing
Co(TiPrP) (50 mg, 0.091 mmol) and Bu4N+CN- (98 mg, 0.37
mmol) was stirred for 1 min under air. The solution was washed
with water for three times. After the removal of the solvent,
[Co(TiPrP)(CN)2]-Bu4N+ was recrystallized from CH2Cl2 and
(9) Riddick, J. A.; Bunger, W. B. In Organic SolVents, Techniques of
Chemistry; Wiley-Interscience: New York, 1970; Vol. II.
(10) Adler, A. D.; Longo, F. R.; Finarelli, J. D.; Goldmacher, J.; Assour,
J.; Korsakoff, L. J. Org. Chem. 1967, 32, 476.
(11) Eaton, S. S.; Eaton, G. R. Inorg. Chem. 1976, 15, 134-139.
(12) Lindsey, J. S.; Wagner, R. W. J. Org. Chem. 1989, 54, 828-836.
(13) Hill, C. L.; Williamson, M. M. J. Chem. Soc. Chem. Commun. 1985,
1228-1229.
(14) Adler, A. D.; Longo, F. R.; Kampas, F.; Kim, J. J. Inorg. Nucl. Chem.
1970, 32, 2443-2445.
(5) Ikezaki, A.; Nakamura, M. Inorg. Chem. 2002, 41, 2761-2768.
(6) Ikezaki, A.; Ikeue, T.; Nakamura, M. Inorg. Chim. Acta 2002, 335,
91-99.
(15) Ema, T.; Senge, M. O.; Nelson, N. Y.; Ogoshi, H.; Smith, K. M.
Angew. Chem., Int. Ed. Engl. 1994, 33, 1879-1881.
(16) Senge, M. O.; Bischoff, I.; Nelson, N. Y.; Smith, K. M. J. Porphyrins
Phthalocyanines 1999, 3, 99-116.
(7) Scheidt, W. R.; Lee, Y. J.; Luangdilok, W.; Haller, K. J.; Anzai, K.;
Hatano, K. Inorg. Chem. 1983, 22, 1516-1522.
(17) Saitoh, T.; Ikeue, T.; Ohgo, Y.; Nakamura, M. Tetrahedron 1997, 53,
12487-12496.
(8) Hansen, A. P.; Goff, H. M. Inorg. Chem. 1984, 23, 4519-4525.
2302 Inorganic Chemistry, Vol. 42, No. 7, 2003

Manganese(III) Porphyrin Complexes
Table 1. UV-Vis Absorption Maxima (λmax, nm) in Various Solvents
solvents
λ
max/nm
Mn(TiPrP)Cl
C6H6
331
331
328
327
326
325
323
320
373
373
378
377
401
402
400
401
481
481
480
481
478
469
471
470
545
547
545
546
543
538
539
539
600
600
598
597
595
585
586
582
639
640
636
637
634
625
624
622
C6H5CH3
CH2Cl2
CHCl3
CH3CN
DMSO
DMF
375
378
377
411
401, 414
398, 415
CH3OH
Mn(TiPrP)(CN)
C6H6
334
335
333
331
332
-
385
387
386
387
385
-
407
408
408
407
407
-
494
494
493
494
470, 490
-
472, 490
470
558
555
555
556
550
-
656
657
656
653
651
-
C6H5CH3
CH2Cl2
CHCl3
CH3CN
DMSO
DMF
588
-
589
584
331
318
384
379
409
409
547
545
654
636
CH3OH
[Mn(TiPrP)(CN)2]-
440
439
436
-
434
438
438
-
C6H6
339
339
-
414
415
413
-
410
414
415
-
464
465
464
-
461
465
462
-
551
550
549
-
547
549
548
-
585
586
582
-
582
584
585
-
711
710
708
-
708
710
708
-
C6H5CH3
CH2Cl2
CHCl3
CH3CN
DMSO
DMF
-
336
338
338
-
CH3OH
hexane to give 74 mg (97%) of the pure complex. UV-vis (CH2-
Cl2; λmax, nm (log ꢀ)): 371 (4.3), 465 (5.0), 55 4(3.2), 604 (3.7),
effective magnetic moment (µ1eff) of Mn(TiPrP)(CN) was similarly
determined at 25 °C.
1
650 (4.1). H NMR (CD2Cl2, 25 °C; δ, ppm): 9.13 (8H, Py-H),
Results and Discussion
4.80 (4H, CH), 2.14 (24H, CH3). 13C NMR (CD2Cl2, 25 °C; δ,
ppm): 140.4 (R-Py), 131.2 (â-py), 122.3 (meso), 34.7 (CH), 28.9
(CH3).
Reactions of Mn(TiPrP)Cl with Bu4N+CN- in Various
Solvents. Titrations of Mn(TiPrP)Cl with Bu4N+CN- have
been monitored in various solvents both by UV-vis spec-
Spectral Measurements. UV-vis spectra were measured on a
Shimadzu MultiSpec-1500 spectrophotometer at ambient temper-
ature. A solution of Mn(TiPrP)Cl was prepared in a graduated flask
and used as a standard solution. A constant volume of the solution
was taken from the standard solution and poured into several 5
mL graduated flasks containing various amounts of Bu4N+CN-.
The concentration of Mn(TiPrP)Cl for the UV-vis measurement
1
troscopy in a low concentration range, 9.7 µM, and by H
NMR spectroscopy in a high concentration range, 7.5 mM.
Tables 1 and 2 list the absorption maxima (λmax, nm) and
chemical shifts (δ, ppm), respectively, for Mn(TiPrP)Cl, Mn-
(TiPrP)(CN), and [Mn(TiPrP)(CN)2]- determined in various
solvents at 25 °C.
1
was maintained at 9.7 µM. H and 13C NMR spectra were taken
1
on a JEOL LA300 spectrometer operating at 300.4 MHz for H.
(1) UV-Visible Spectroscopy. (i) Benzene Solution.
Figure 1a shows the spectral change observed in benzene
solution when Bu4N+CN- was added up to 4.0 equiv relative
to Mn(TiPrP)Cl. In the absence of Bu4N+CN-, the green
Mn(TiPrP)Cl solution showed the absorption bands at 331,
373, 401, 481, 545, 600, and 639 nm. As the Bu4N+CN-
solution was added, the signals of Mn(TiPrP)Cl decreased
and those of the new complex (A) at 334, 385, 407, 494,
and 656 nm increased in intensity with the isosbestic points
at 373, 445, 489, 531, and 646 nm. After the addition of 4.0
equiv of Bu4N+CN-, the green solution showed only the
signals for A. Figure 1b shows the spectral change observed
when much larger amounts of Bu4N+CN-, 4.0-1000 equiv,
were added to the solution. As the Bu4N+CN- solution was
added, the signal intensities for A weakened and those of
the new complex (B) at 339, 414, 440, 464, 551, 585, and
711 nm strengthened with the isosbestic points at 405, 475,
524, 597, and 692 nm. By the addition of a large excess of
Bu4N+CN-, the signals for A completely disappeared and
the resultant red solution exhibited only the signals for B.
Thus, it is most reasonable to consider that the reaction in
Chemical shifts were referenced to the residual peaks of deuterated
1
solvents, CD3OD (δ ) 3.30 ppm for H, δ ) 49.0 ppm for 13C),
CDCl3 (7.24, 77.0), CD2Cl2 (5.32, 53.8), DMSO-d6 (2.49, 39.5),
DMF-d7 (2.70, 35.2), (CD3)2CO (2.04, 29.8), CD3CN (1.93, 118.2),
and C6D6 (7.15, 128.0). For 1H NMR measurement, a 7.5 mM CD2-
Cl2 standard solution of Mn(TiPrP)Cl was prepared. In each
measurement, a 550 µL CD2Cl2 solution was taken from the
standard solution and poured into NMR sample tube under argon
1
atmosphere. An H NMR spectrum was taken at 25 °C each time
after the addition of certain amount of Bu4N+CN- as a CD2Cl2
solution. The 13C NMR spectrum of [Mn(TiPrP)(CN)2]Bu4N+ was
measured at 25 °C by the addition of 6.0 equiv of Bu4N+CN- to
a C6D6 solution containing 20 mg of Mn(TiPrP)Cl.
Effective Magnetic Moments. The effective magnetic moment
(µ1eff) of [Mn(TiPrP)(CN)2]-Bu4N+ was determined by the Evans
method at 25 °C relative to the well-characterized high-spin Mn-
(TiPrP)Cl complex according to µ1eff ) (∆ν1/∆ν2)1/2µ2eff, where ∆ν1
and ∆ν2 are the difference in chemical shifts of the reference signals
in [Mn(TiPrP)(CN)2]-Bu4N+ and [Mn(TiPrP)Cl], respectively.18 The
(18) Evans, D. F.; James, T. A. J. Chem. Soc., Dalton Trans. 1979, 723-
726.
Inorganic Chemistry, Vol. 42, No. 7, 2003 2303

Ikezaki and Nakamura
Table 2. 1H NMR Chemical Shifts (δ/ppm, 25 °C) of Mn(Por)Cl (S), Mn(Por)(CN) (mono), and [Mn(Por)(CN)2]- (bis)
pyrrole-H
mono
CH or ortho
CH3 or meta
solvents
S
bis
S
mono
bis
S
mono
bis
TiPrP
9.5
CD2Cl2
CDCl3
C6D6
DMSO-d6
CD3OD
-19.4
-19.1
-13.1
-29.3c
-31.6c
-23.8
-18.7d
a
-17.2
-17.7
a
9.3
9.4
a
a
a
-1.3d
a
-1.1
-0.6
a
3.2
3.3
3.0
1.8c
2.1c
3.2
3.3
a
a
a
-1.3d
a
-2.4
-2.4
a
-23.5
9.8
8.9
a
a
a
ca. 4c
b
TPP
CD2Cl2
CDCl3
C6D6
DMSO-d6
CD3OD
-21.9
-22.5
-19.2
-27.4c
-30.8c
-28.1
-28.3
-26.6
a
-14.7
-17.5d
-14.4
-14.8
a
b
b
b
b
b
b
a
4.4
4.6d
4.1
4.2
a
8.3
8.2
8.0
7.9c
7.4c
8.2
8.2
b
a
7.4
5.6
5.9d
5.4
5.6
a
8.6c
9.3c
-26.2
8.5
p-CF3-TPP
CD2Cl2
CDCl3
C6D6
DMSO-d6
CD3OD
-21.9
-22.5
-19.3
-27.4c
-31.0c
a
-14.7
-15.4
-14.5
-14.9
-20.1d
b
b
b
a
b
a
a
b
4.8
4.3
4.3
4.6
6.4d
8.6
8.6
8.2
7.9c
7.7c
a
8.4
a
a
b
5.9
5.7
5.8
6.1
6.9d
-28.1
a
a
b
8.7c
9.5c
a Not formed. b Too broad. c [Mn(Por)(solvent)2]+. d Chemical shifts of the mixture containing some amount of the mono-adduct.
Figure 2. UV-vis spectral change observed when Bu4N+CN- was added
to (a) CHCl3 and (b) DMSO solutions of Mn(TiPrP)Cl at ambient
temperature. The blue line shows the UV-vis spectrum of the starting
complex, Mn(TiPrP)Cl in (a) and [Mn(TiPrP)(DMSO)2]+ in (b). Green and
red lines are the UV-vis spectra of Mn(TiPrP)(CN), and [Mn(TiPrP)(CN)2]-,
respectively.
Figure 1. (a) UV-vis spectral change observed when 0-4 equiv of
Bu4N+CN- was added to a benzene solution of Mn(TiPrP)Cl at ambient
temperature. The blue line is the UV-vis spectrum of Mn(TiPrP)Cl, and
the green line is that of [Mn(TiPrP)(CN)] obtained by the addition of 4
equiv of Bu4N+CN-. (b) UV-vis spectral change observed when 4-1000
equiv of Bu4N+CN- was added to a benzene solution of Mn(TiPrP)Cl at
ambient temperature. The green and red lines correspond to the spectra of
[Mn(TiPrP)(CN)] and [Mn(TiPrP)(CN)2]-, respectively.
to the mono-adduct and from the mono- to the bis-adduct,
respectively.
Figure 1a corresponds to the conversion of Mn(TiPrP)Cl to
the mono-adduct (A), and that in Figure 1b corresponds to
the conversion of A to the bis-adduct (B).
(ii) Dichloromethane Solution. Similar spectral change
was observed in CH2Cl2 solution, although much larger
amounts of Bu4N+CN-, at least 10 and 20000 equiv, were
necessary for the complete conversion from Mn(TiPrP)Cl
(iii) Chloroform Solution. Figure 2a shows the spectral
change observed when Bu4N+CN- was added to the CHCl3
solution of Mn(TiPrP)Cl. Complete conversion from Mn-
(TiPrP)Cl to the mono-adduct was observed when 10 equiv
of Bu4N+CN- was added. No further change was observed,
however, even by the addition of 7000 equiv of Bu4N+CN-.
Thus, the result in CHCl3 solution is quite different from
2304 Inorganic Chemistry, Vol. 42, No. 7, 2003

Manganese(III) Porphyrin Complexes
that in CH2Cl2 solution where 78% of the monoadduct is
converted to the bis-adduct in the presence of 7000 equiv of
Bu4N+CN-.
(iv) Dimethyl Sulfoxide Solution. Figure 2b shows the
spectral change observed when Bu4N+CN- was added to the
DMSO solution of Mn(TiPrP)Cl. In the absence of Bu4N+CN-,
the UV-vis spectrum was quite different from the corre-
sponding spectra taken in other solvents such as benzene,
CH2Cl2, and CHCl3; the intense band at 480 nm in CH2Cl2
solution showed a hypsochromic shift to 469 nm in DMSO
solution. The difference should be ascribed to the formation
of [Mn(TiPrP)(DMSO)2]Cl in DMSO solution. In fact, the
absorption maxima are quite close to those of analogous [Mn-
(TPP)(DMSO)2]Cl reported previously by Hansen and Goff;8
the latter complex was formed when Mn(TPP)Cl was
dissolved in DMSO. By the addition of Bu4N+CN-, the new
signals at 414, 438, and 549 nm ascribed to the bis-adduct
increased in intensity. The result indicates that the conversion
from [Mn(TPP)(DMSO)2]Cl to the bis-adduct took place
without accumulation of the mono-adduct. Existence of a
small amount of the mono-adduct could not be ruled out;
however, since no clear isosbsetic points were observed
during the titration process. Similar spectral change was
observed in DMF and acetonitrile solutions though the
signals for the mono-adduct were observed during the
titration process.
(v) Methanol Solution. The UV-vis spectrum of Mn-
(TiPrP)Cl taken in methanol solution was quite different from
the corresponding spectra taken in benzene, CH2Cl2, and
CHCl3. Thus, Mn(TiPrP)Cl is supposed to be converted to
[Mn(TiPrP)(CH3OH)2]Cl in methanol solution. In fact, the
UV-vis spectrum is quite similar to that of [Mn(TPP)(CH3-
OH)2]Cl reported previously.19 The spectrum showed little
change even by the addition of 1000 equiv of Bu4N+CN-.
Thus, it is concluded that neither the mono- nor the bis-
adduct was formed in methanol solution.
Figure 3. 1H NMR spectral change observed when (a) 0, (b) 0.5, (c) 1.0,
(d) 2.0, (e) 4.0, and (f) 10.0 equiv of Bu4N+CN- was added to a CD2Cl2
solution of Mn(TiPrP)Cl at 25 °C.
the pyrrole signal of the mono-adduct, increased in intensity.
The methine and methyl signals showed little change from
the original positions. When 1.0 equiv of Bu4N+CN- was
added, the signal at -19.4 ppm completely disappeared. The
solution color was maintained green. Further addition of
Bu4N+CN- caused a gradual downfield shift of the pyrrole
signal from -24 to -19 ppm with considerable narrowing.
In contrast, the methyl and methine signals showed upfield
shifts. When 10 equiv of Bu4N+CN- was added, the pyrrole,
methine, and methyl signals were observed at -18.7, -1.3,
and -1.3 ppm, respectively, as relatively sharp signals. The
solution color became red at this point. These signals showed
a little shift on further addition of Bu4N+CN-. Complete
conversion to the bis-addcut was not achieved, however, even
by the addition of 15 equiv of Bu4N+CN- as is revealed
from a rather broad pyrrole signal. Thus, the spectrum in
Figure 3f shows a mixture consisting of bis- and mono-
adduct in 3:1 ratio; the ratio was estimated under the
assumption that chemical shifts for the mono- and bis-adduct
are -23.5 and -17.2 ppm, respectively. The fact that the
only one pyrrole signal was observed during the titration
process suggests that the ligand exchange between the mono-
(2) 1H NMR Spectroscopy. (i) Benzene Solution. Titra-
tion of Mn(TiPrP)Cl with Bu4N+CN- in C6D6 solution was
hampered due to the extremely low solubility of Mn(TiPrP)-
Cl. A clear solution was obtained, however, when 6.0 equiv
of Bu4N+CN- was added. The sharp pyrrole signal was
observed at δ -17.2 ppm (∆ν1/2 ) 13 Hz), together with
the methyl and methine signals at -2.4 (∆ν1/2 ) 16 Hz)
and -1.1 ppm (∆ν1/2 ) 28 Hz), respectively. Further addition
of the ligand caused no appreciable change in the position
and width of each signal. Thus, these signals were ascribed
to the bis-adduct. The red color of the solution also supports
the formation of the bis-adduct. A similar result was obtained
in CCl4 solution though the complex was very unstable in
this solution.
1
and bis-adduct is fast on the H NMR time scale at least at
25 °C.
1
(iii) Chloroform Solution. Figure 4 shows the H NMR
spectral change observed when Bu4N+CN- was added to the
CDCl3 solution of Mn(TiPrP)Cl. As in the case of the CD2-
Cl2 solution mentioned above, a broad signal at δ -19.1
ppm weakened and a new signal at -23.5 ppm increased in
intensity. No further spectral change was observed, however,
even by the addition of 15 equiv of Bu4N+CN-. In addition,
the solution color was maintained green. The results strongly
indicate that, contrary to the reaction in CD2Cl2, the mono-
adduct is the only product in CDCl3 solution.
1
(ii) Dichloromethane Solution. Figure 3 shows the H
NMR spectral change observed when Bu4N+CN- was added
to the CD2Cl2 solution of Mn(TiPrP)Cl. By the addition of
0.5 equiv of Bu4N+CN-, a broad pyrrole signal of Mn-
(TiPrP)Cl at δ -19.4 ppm (∆ν1/2 ) 630 Hz) decreased and
a new signal at -23.8 ppm (∆ν1/2 ) 630 Hz), assigned to
(iv) Dimethyl Sulfoxide Solution. Figure 5 shows the 1H
NMR spectral change observed when Bu4N+CN- was added
to the DMSO-d6 solution of Mn(TiPrP)Cl. The pyrrole,
methine, and methyl signals of Mn(TiPrP)Cl appeared at
-29.3, ca. 4, and 1.8 ppm, respectively, at 25 °C. Thus, the
(19) Hatano, K.; Anzai, K.; Iitaka, Y. Bull. Chem. Soc. Jpn. 1983, 56, 422-
427.
Inorganic Chemistry, Vol. 42, No. 7, 2003 2305

Ikezaki and Nakamura
solution turned red at this point. The spectral change
mentioned above indicates that the major product is the bis-
adduct even by the addition of 1.0 equiv of Bu4N+CN-. The
broad pyrrole signal for the bis-adduct, observed when 1.0-
4.0 equiv of Bu4N+CN- was added, suggests that a small
amount of the mono-adduct also exists in the solution and
involves in the fast ligand exchange with the bis-adduct.
Thus, the spectral change observed by 1H NMR spectroscopy
is totally consistent with that observed by UV-vis spec-
troscopy. Titration experiments in other dipolar aprotic
solvents such as acetonitrile, DMF, and acetone were difficult
due to the low solubility of Mn(TiPrP)Cl in these solvents.
A clear red solution was obtained in each case, however, by
the addition of 6 equiv of Bu4N+CN-, which exhibited the
sole entity of the bis-adduct as revealed from the chemical
shift and the signal width. On the basis of these results, it is
concluded that the bis-adduct is preferentially formed in
dipolar aprotic solvents even when Bu4N+CN- is less than
2.0 equiv.
Figure 4. 1H NMR spectral change observed when (a) 0, (b) 0.5, (c) 1.0,
(d) 1.5, (e) 2.0, and (f) 15.0 equiv of Bu4N+CN- was added to a CDCl3
solution of Mn(TiPrP)Cl at 25 °C.
(v) Methanol Solution. As in the case of the DMSO
solution mentioned above, the chemical shifts in methanol
solution are quite different from those determined in less
polar solvents such as CD2Cl2, CDCl3, and C6D6. Thus,
Mn(TiPrP)Cl is converted to [Mn(TiPrP)(CD3OD)2]Cl in
methanol solution. No appreciable change was observed in
1
the H NMR spectrum even by the addition of 15 equiv of
Bu4N+CN-. The result clearly indicates that neither the
mono- nor the bis-adduct is formed in methanol solution in
the presence of a large excess of Bu4N+CN-.
(3) Formation Constants for the Mono- and Bis-
Adducts. The formation constants K1 and K2 corresponding
to eqs 1 and 2, respectively, were determined by the UV-
vis spectroscopy at 25 °C according to the method of Rossotti
and Rossotti.7,20,21
Mn(TiPrP)Cl + Bu4N+CN- a
Mn(TiPrP)(CN) + Bu4N+Cl- (1)
Figure 5. 1H NMR spectral change observed when (a) 0, (b) 1.0, (c) 2.0,
and (d) 10.0 equiv of Bu4N+CN- was added to a DMSO-d6 solution of
Mn(TiPrP)Cl at 25 °C.
Mn(TiPrP)(CN) + Bu4N+CN- a
[Mn(TiPrP)(CN)2]-Bu4N+ (2)
chemical shifts are quite different from those determined in
other solvents such as CD2Cl2, CDCl3, and C6D6; the pyrrole,
methine, and methyl signals appeared at -19.4, 9.5, and 3.2
ppm, respectively, in CD2Cl2 solution at 25 °C. As mentioned
in the UV-vis section, the difference in chemical shifts
should be ascribed to the formation of [Mn(TiPrP)(DMSO)2]-
Cl in DMSO solution rather than to the solvent effect on
the chemical shifts of Mn(TiPrP)Cl.8 When 1.0 equiv of
Bu4N+CN- was added, the signals for the starting complex
decreased and those for the product increased in intensity;
the product showed a broad pyrrole signal at -18.4 ppm.
When 2.0 equiv of Bu4N+CN- was added, the pyrrole signal
of the starting complex completely disappeared and that of
the product increased in intensity. Further addition of
Bu4N+CN- sharpened each signal without considerable
change in its position; the pyrrole, methine, and methyl
signals were observed at -17.7 (∆ν1/2 ) 26 Hz), -0.6 (∆ν1/2
) 33 Hz), and -2.4 (∆ν1/2 ) 17 Hz) ppm, respectively, when
12 equiv of Bu4N+CN- was added. The originally green
In benzene solution, the K1 and K2 values were determined
as follows on the basis of the spectral change given in Figure
1a,b, respectively: K1 ) 15 and K2 ) 1600 M-1. The
formation constants in CH2Cl2 solution were similarly
determined as K1 ) 46 and K2 ) 23 M-1. In CHCl3 solution,
the K1 value was determined as K1 ) 0.96. Because the bis-
adduct was not observed even by the addition of 7000 equiv
of Bu4N+CN-, we estimated the K2 value to be less than
0.8 M-1 by assuming the 5% conversion from the mono- to
the bis-adduct in the presence of 7000 equiv of Bu4N+CN-.
As mentioned, Mn(TiPrP)Cl is converted to the six-
coordinated complexes, [Mn(TiPrP)(Solv)2]Cl, in polar sol-
vents such as DMSO and CH3OH. Thus, the formation
constants in these solvents were not determined.
(20) Rossotti, F. J. C.; Rossotti, H. In The Determination of Stability
Constants; McGraw-Hill: New York, 1961; p 277.
(21) Neya, S.; Morishima, I.; Yonezawa, T. Biochemistry 1981, 20, 2610-
2614.
2306 Inorganic Chemistry, Vol. 42, No. 7, 2003

Manganese(III) Porphyrin Complexes
tively,22 the results indicate that the strength of the O-H‚‚‚N
hydrogen bonding determines the relative population of the
mono- and bis-adducts. Similarly, 66 and 180 µL of CHBr3
and CHCl3 were necessary, respectively, to obtain the
solution containing 20% bis-adduct and 80% mono-adduct.
Thus, the coordination ability of cyanide increases as the
solvent changes from CH3OH to CHCl3 as given in eq 4.
Since CHBr3 is known to be a much stronger hydrogen donor
in the C-H‚‚‚N intermolecular hydrogen bonding,23 the
results again indicate that the O-H‚‚‚N and C-H‚‚‚N
hydrogen bonding plays an important role to determine the
coordination structure of the products.
CH3OH < CH3CH2OH < (CH3)2CHOH < CHBr3 < CHCl3
(4)
On the basis of the discussion given above, the coordina-
tion ability of cyanide toward Mn(III) can be summarized
as follows. In nonpolar or dipolar aprotic solvents where no
hydrogen bonding occurs between cyanide and solvent
molecules, cyanide behaves as a strong ligand and forms
the bis-adduct. In CHCl3 solution, however, the C-H‚‚‚N
hydrogen bonding weakens the coordination ability of the
cyanide ligand. As a result, cyanide behaves as a much
weaker anionic ligand such as halide to form preferentially
the mono-adduct. In methanol solution, strong O-H‚‚‚N
hydrogen bonding further weakens the coordination ability
of cyanide to such an extent as to make the ligand substitution
difficult to occur. To further confirm the validity of the
above-mentioned speculation, we have examined the coor-
dination ability of tert-butyl isocyanide (tBuNC). Because
Figure 6. Decrease in the population ratios of [Mn(TiPrP)(CN)2]- observed
when the compounds with proton-donating ability were added to the CD2-
Cl2 solution consisting of [Mn(TiPrP)(CN)2]- (76%) and Mn(TiPrP)(CN)
(24%).
Solvent Effects on the Mono/Bis Ratios. The UV-vis
and 1H NMR results described above indicate that the
reaction products are greatly influenced by the solvents
used: (i) The reactions in dipolar aprotic solvents such as
DMSO, DMF, acetone, acetonitrile, and nitromethane yielded
preferentially the bis-adduct. (ii) The reactions in the less
polar solvents such as CH2Cl2 and benzene yielded both the
mono- and bis-adduct depending on the amount of cyanide
added, though the complete conversion to the bis-adduct was
achieved with much smaller amount of the ligand in benzene
solution. (iii) The reaction in CHCl3 formed only the mono-
adduct. (iv) The reaction in methanol gave neither the mono-
nor the bis-adduct. Thus, the affinity of cyanide toward the
Mn(III) ion in Mn(TiPrP)Cl is larger in nonpolar and dipolar
aprotic solvents than in protic solvents. The order of the
ligand affinity in various solvents is ranked as given in
eq 3.
t
the electronic structure of BuNC is considered to be quite
similar to the cyanide with a strong hydrogen bonding,4,5,24
the coordination of tBuNC toward Mn(Por)Cl must be very
weak. As expected, no coordination was observed even by
t
the addition of 20 equiv of BuNC to Mn(TiPrP)Cl in CD2-
Cl2 solution.
Comparison of Porphyrins. We have then examined the
population ratios of the mono- and bis-adduct by using Mn-
(TPP)Cl. The reaction products were analyzed by 1H NMR
spectroscopy. The chemical shifts of the mono- and bis-
adducts, Mn(TPP)(CN) and [Mn(TPP)(CN)2]-, in various
solvents are listed in Table 2b together with those of Mn-
(TPP)Cl. The most characteristic point is that the ratio of
the bis-adduct relative to the corresponding mono-adduct
increases in each solvent on going from TiPrP to TPP
complexes. Thus, while the bis-adduct was not formed in
CDCl3 solution in the case of Mn(TiPrP)Cl even by the
addition of 15 equiv of Bu4N+CN-, 73% of the bis-adduct
was formed in the case of Mn(TPP)Cl under the same
condition. Similarly, while no adduct was formed in CD3-
OD solution in the case of [Mn(TiPrP)(CD3OD)2]Cl even
by the addition of 15 equiv of Bu4N+CN-, mono-adduct was
CH3OH < CHCl3 < CH2Cl2 <
(CH3)2CO, DMSO, DMF, CH3CN, C6H6, CCl4 (3)
The order in eq 3 could be the suggestion that the solvents
with hydrogen donor ability weaken the coordination ability
of cyanide ligand.3 To reveal the effect of hydrogen bonding
on the product ratio, we added weak acidic compounds such
as methanol, ethanol, 2-propanol, CHCl3, and CHBr3 to the
CD2Cl2 solution consisting of 76% bis-adduct and 24%
mono-adduct. Figure 6 shows how the population ratio of
the bis-adduct decreases as the acidic compounds were added.
The degree of decrease was quite different depending on the
compounds. In the case of methanol, addition of only 7 µL
converted the bis-adduct to the mono-adduct to form the
solution consisting of 20% bis- and 80% mono-adduct. In
contrast, much larger amounts, 18 and 40 µL, were necessary
for the same conversion in the case of ethanol and 2-pro-
panol, respectively. Since the pKa values of methanol,
ethanol, and 2-propanol are 15.2, 16.0, and 16.5, respec-
(22) Reeve, W.; Erikson, C. M.; Aluotto, P. F. Can. J. Chem. 1979, 57,
2747-2754.
(23) Allerhand, A.; Schleyer, P. von R. J. Am. Chem. Soc. 1963, 85, 1715-
1723.
(24) Simonneaux, G.; Hindre, F.; Pouzennec, M. L. Inorg. Chem. 1989,
28, 823-825.
Inorganic Chemistry, Vol. 42, No. 7, 2003 2307

Ikezaki and Nakamura
actually formed in the case of [Mn(TPP)(CD3OD)2]Cl under
the same conditions; formation of the bis-adduct was not
observed in this condition. As mentioned in the Introduction,
Sheidt and co-workers reported the formation of the bis-
adduct [Mn(TPP)(CN)2]- in methanol solution on the basis
of UV-vis spectroscopic results.7 The discrepancy should
be ascribed to the difference in the molar ratios of cyanide
relative to the complex; ca. 18 000 equiv of K+CN- was
added relative to Mn(TPP)(OH) in UV-vis measurement,7
while 15 equiv of Bu4N+CN- was added relative to Mn-
1
(TPP)Cl in the H NMR experiment. The difference in
product ratio between CDCl3 and CD2Cl2 solutions was also
clearly shown in Mn(TPP)Cl. While the addition of 2 equiv
of Bu4N+CN- produced more than 90% of the bis-adduct
in CD2Cl2, only 16% of the bis-adduct was formed in CDCl3.
The difference in product ratios between two porphyrin
complexes should be ascribed to the steric and electronic
effects of the meso substituents. Because of the presence
of bulky isopropyl group, the porphyrin ring is supposed to
be strongly ruffled as in the case of the Fe(III) complexes
such as Fe(TiPrP)Cl and [Fe(TiPrP)(THF)2]ClO4.25,26 As a
result, the Mn(III)-N(porphyrin) bonds must be short-
ened, resulting in the strong electron donation of the
porphyrin nitrogen atoms toward the central Mn(III) ion. The
presence of the electron-donating alkyl groups also con-
tributes to the strong coordination of the porphyrin nitro-
gen atoms. The intense equatorial coordination in Mn(TiPrP)-
Cl could, in turn, weaken the axial coordination of cyanide
ligand toward Mn(TiPrP)Cl as compared toMn(TPP)Cl.
The results indicate that the population ratio of the bis-
adduct could further increase in Mn(p-CF3-TPP)Cl due to
the presence of strong electron-withdrawing substituents at
the meso-phenyl groups. In fact, even by the addition of 2.0
equiv of Bu4N+CN-, 78% of the bis-adduct was formed in
CDCl3 solution; the population ratio of the bis-adduct was
only 16% in the case of Mn(TPP)Cl under the same
condition. Table 2c lists the chemical shifts of these
complexes taken in various solvents. Figure 7 shows how
the relative population of the bis-adduct changes among
different porphyrin complexes in CDCl3 solution by the
addition of Bu4N+CN-. The data for Mn(TMP)Cl and Mn-
(2,6-Cl2-TPP)Cl are also added in Figure 7. Clearly, binding
of cyanide toward Mn(III) ions increase in the order given
in eq 5.
Figure 7. Increase in the population ratios of [Mn(TiPrP)(CN)2]- observed
when Bu4N+CN- was added to the CDCl3 solutions of various Mn(Por)Cl,
where Por’s are TiPrP, TMP, TPP, p-CF3-TPP, and 2,6-Cl2-TPP.
the effective magnetic moment of Mn(TiPrP)(CN) by the
Evans method in CHCl3; the pure complex can be obtained
in this solution simply by the addition of an excess amount
of Bu4N+CN- to Mn(TiPrP)Cl. The effective magnetic
moment was 5.0 µB at 25 °C, which indicates that the
complex is in the S ) 2 spin state. Six-coordinated Mn(III)
porphyrin complexes with strong axial ligands such as
[Mn(TPP)(CN)2]-K+ and [Mn(TPP)(Im)2]-K+ (Im: imida-
zolate) are reported to exhibit the low-spin (S ) 1) state.8
We are extremely interested in the spin state of [Mn(TiPrP)-
(CN)2]- because the analogous low-spin iron(III) complex
[Fe(TiPrP)(CN)2]- shows a less common (dxz, dyz)4(dxy)1
electron configuration due to the highly ruffled porphyrin
structure;29-32 recent studies have revealed the general
conditions to determine the electron configurations of the
low-spin Fe(III) porphyrin complexes.33-39 If presumably
ruffled [Mn(TiPrP)(CN)2]- shows the reversal of the energy
levels between the (dxz, dyz) and dxy orbitals as in the case of
the corresponding Fe(III) complex, then the Mn(III) complex
should be diamagnetic with the (dxz, dyz)4(dxy)0 electron
configuration. This expectation was not realized because
(29) Nakamura, M.; Ikeue, T.; Fujii, H.; Yoshimura, T. J. Am. Chem. Soc.
1997, 119, 6284-6291.
(30) Nakamura, M.; Ikeue, T.; Fujii, H.; Yoshimura, T.; Tajima, K. Inorg.
Chem. 1998, 37, 2405-2414.
(31) Ikeue, T.; Ohgo, Y.; Saitoh, T.; Nakamura, M.; Fujii, H.; Yokoyama,
M. J. Am. Chem. Soc. 2000, 122, 4068-4076.
(32) Ikeue, T.; Ohgo, Y.; Saitoh, T.; Yamaguchi, T.; Nakamura, M. Inorg.
Chem. 2001, 40, 3423-3434.
Mn(TiPrP)Cl, Mn(TMP)Cl <
Mn(TPP)Cl, Mn(2,6-Cl2-TPP)Cl < Mn(p-CF3-TPP)Cl (5)
(33) Walker, F. A.; Simonis, U. Proton NMR Spectroscopy of Model
Hemes. In NMR of Paramagnetic Molecules; Berliner, L. J., Reuben,
J., Eds.; Biological Magnetic Resonance Vol. 12; Plenum Press: New
York, 1993; pp 133-274.
Electronic States of Mono- and Bis-Adducts of Man-
ganese Complexes. It is well-known that five-coordinated
manganese(III) porphyrin complexes with anionic ligands
(X) such as halides, hydroxide, and cyanide show the high-
spin (S ) 2) state.27,28 In the present case, we have determined
(34) Walker, F. A. In The Porphyrin Handbook; Kadish, K. M., Smith, K.
M., Guilard, R., Eds.; Academic Press: San Diego, CA, 2000; Vol.
5, Chapter 36, pp 81-183.
(35) Wołowiec, S.; Latos-Graz˘yn´ski, L.; Mazzanti, M.; Marchon, J.-C.
Inorg. Chem. 1997, 36, 5761-5771.
(36) Pilard, M.-A.; Guillemot, M.; Toupet, L.; Jordanov, J.; Simonneaux,
G. Inorg. Chem. 1997, 36, 6307-6314.
(37) Wołowiec, S.; Latos-Graz˘yn´ski, L.; Toronto, D.; Marchon, J.-C. Inorg.
Chem. 1998, 37, 724-732.
(25) Ikeue, T.; Ohgo, Y.; Uchida, A.; Nakamura, M.; Fujii, H.; Yokoyama,
M. Inorg. Chem. 1999, 38, 1276-1281.
(26) Ohgo, Y.; Saitoh, T.; Nakamura, M. Acta Crystallogr. 2001, C57, 233-
234.
(27) Goff, H. M.; Hansen, A. P. Inorg. Chem. 1984, 23, 321-326.
(28) Turner, P.; Gunter, M. J. Inorg. Chem. 1994, 33, 1406-1415.
(38) Nakamura, M.; Ikeue, T.; Ikezaki, A.; Ohgo, Y.; Fujii, H. Inorg. Chem.
1999, 38, 3857-3862.
(39) Simonneaux, G.; Schu¨nemann, V.; Morice, C.; Carel, L.; Toupet, L.;
Winkler, H.; Trautwein, A. X.; Walker, F. A. J. Am. Chem. Soc. 2000,
122, 4366-4377.
2308 Inorganic Chemistry, Vol. 42, No. 7, 2003

Manganese(III) Porphyrin Complexes
Table 3. 13C NMR Chemical Shifts (δ/ppm) of Complexes Determined
at 25 °C
contact shift is determined, the proton hyperfine coupling
constant, AH, can be calculated by
complexes
solvents R-pyrrole â-pyrrole meso -CH -CH3
δcon ) (AH)|γe/γH|S(S + 1)/(3kT)
(7)
[Mn(TiPrP)(CN)2]- C6D6
293
503
484
475
490
517
-53
-137
-157
-153
-223
-238
141 39.7 85.8
211 42.6 119.0
220 44.8 113.4
216 47.9 112.5
102 43.7 90.5
113 41.6 99.0
Mn(TiPrP)(CN)
Mn(TiPrP)Cl
CDCl3
CDCl3
CD2Cl2
DMSO-d6
CD3ODb
Because the proton hyperfine coupling constant is pro-
portional to the spin density Fπ of the carbon atom to which
the proton is attached, it is possible to determine the spin
density at the carbon by
a
a [Mn(TiPrP)(DMSO)2]+. b [Mn(TiPrP)(CD3OD)2]+.
AH/h ) (QHCHFπ)/2S
(8)
[Mn(TiPrP)(CN)2]- showed a paramagnetically shifted pyr-
role signal at -17.2 ppm in benzene solution. In fact, the
effective magnetic moment of [Mn(TiPrP)(CN)2]- deter-
mined by the Evans method in benzene solution was 3.2 µB
at 25 °C, indicating that the complex is in the low-spin (S )
1) state with the (dxy)2(dxz, dyz)2 electron configuration as in
the case of the other low-spin Mn(III) complex. Hansen and
Goff reported the unpaired spin densities centered on the π
MO of the carbon atoms in [Mn(TPP)(CN)2]- on the basis
of the 1H NMR chemical shifts.8 Although both [Mn(TiPrP)-
(CN)2]- and [Mn(TPP)(CN)2]- are in the S ) 1 spin state,
the π spin distribution on the porphyrin carbons in
[Mn(TiPrP)(CN)2]- could be different from that in [Mn(TPP)-
(CN)2]- due to the highly ruffled structure in the former
complex. To reveal the effect of porphyrin deformation on
the π spin densities at the porphyrin carbon atoms, the data
for both the 1H and 13C NMR chemical shifts are necessary.
Thus, we have measured the 13C NMR chemical shifts of
[Mn(TiPrP)(CN)2]- and listed them in Table 3 together with
those of Mn(TiPrP)(CN) and Mn(TiPrP)Cl. To determine the
spin densities, we have separated the 1H NMR isotropic shift
(δiso) into dipolar shift (δdip) and contact shift(δcon) by using
the chemical shifts of the corresponding diamagnetic cobalt-
(III) complex [Co(TiPrP)(CN)2]- as δdia values.40-44 The
metal-centered dipolar shift in the complexes with axial
symmetry is defined by eq 6, where ø values are molecular
susceptibilities and the term (3 cos2 θ - 1)/r3 is referred to
as the axial geometric factor:
where QHCH is a proportional constant.45,46 The π spin density
of the â-pyrrole carbon atoms was estimated as +0.0066 by
the use of QH ) -65.8 MHz in eq 8.44,47
CH
To determine the π spin densities at the R-pyrrole and
meso-carbon as well as pyrrole nitrogen atoms, analysis of
the 13C NMR chemical shifts is necessary. Carbon-13
isotropic shift is presented by
δiso ) δdipMC + δdipLC + δcon
(9)
where δdipLC is a ligand-centered dipolar shift. Carbon contact
shifts originate from unpairing of carbon 1s electrons and
unpairing of the three carbon sp2 bonding pairs. Thus, the
contact shift for the â-pyrrole carbon can be expressed by
the Karplus and Frankel equation,
δ
con(py-â) ) [(Sc + 2QCCC′ + QCCH)Fπ
+
â
QCC′C(Fπâ + Fπ )]FC
R
where FC ) |γe/γC|S(S + 1)/(3kT).
In the equation, the SC term indicates polarization of the
1s orbital. The QC and QCCH terms reflect polarization of
CC′
the three sp2 bonds by π-spin density at the observed carbon
atom. The QC term represents polarization of the C-C
C′C
bond by π-spin densities centered on the neighboring carbon
LC
atoms. The δdip is assumed to be proportional to the spin
density Fπ at the observed carbon atom and is given by
δdipMC ) (1/12π)(ø| - ø )(3 cos2 θ - 1)/r3
(6)
LC
LC
δdip ) DFπ. Thus, the (δdip + δcon) values for the py-R,
py-â, meso, and R carbons, which can easily be obtained by
(δiso - δdipMC), are expressed by eqs 10-13, respectively.
In the case of [Mn(TiPrP)(CN)2]-, we have determined the
dipolar shifts of the pyrrole and isopropyl CH protons by
assuming that the isotropic shift of the isopropyl CH3 protons
is totally ascribed to the dipolar shift. The contact shift of
each proton is then obtained by δcon ) δiso - δdip. Once the
py-R: δcon + δdipLC ) DFπpy-R + [(Sc + 2QC
QCCN)Fπpy-R + QCC′C(Fπpy-â + Fπmeso) + QCNCFπN]FC (10)
+
CC′
py-â: δcon + δdipLC ) DFπpy-â + [(Sc + 2QC
+
CC′
(40) La Mar, G. N.; Walker, F. A. In The Porphyrins; Dolphin, D., Ed.;
Academic Press: New York, 1979; Vol. IV, pp 61-157.
(41) Goff, H. In Iron Porphyrins, Part I; Lever, A. B. P., Gray, H. B.,
Eds.; Physical Bioinorganic Chemistry Series 1; Addison-Wesley:
Reading, MA, 1983; pp 237-281.
(42) Bertini, I.; Luchinat, C. In NMR of Paramagnetic Molecules in
Biological Systems; Lever, A. B. P., Gray, H. B., Eds.; Physical
Bioinorganic Chemistry Series 3; Benjamin/Cummings: Menlo Park,
CA, 1986; pp 165-229.
(43) Mispelter, J.; Momenteau, M.; Lhoste, J.-M. Heteronuclear Magnetic
Resonance. In NMR of Paramagnetic Molecules; Berliner, L. J.,
Reuben, J., Eds.; Biological Magnetic Resonance Vol. 12; Plenum
Press: New York, 1993; pp 299-355.
(44) Bertini, I.; Luchinat, C. In NMR of Paramagnetic Substances; Lever,
A. B. P., Ed.; Coordination Chemistry Reviews 150; Elsevier:
Amsterdam, 1996; pp 29-75.
QCCH)Fπpy-â + QCC′C(Fπpy-â + Fπpy-R)]FC (11)
meso: δcon + δdipLC ) DFπmeso + [(Sc + 3QCCC′)Fπ
+
meso
2QCC′CFπpy-R]FC (12)
CR: δcon + δdipLC ) DFπCR + [(Sc + 3QCCC′)Fπ
+
CR
QCC′CFπmeso]FC (13)
(45) Heller, C.; McConnell, H. M. J. Chem. Phys. 1960, 32, 1535.
(46) McLachlan, A. D. Mol. Phys. 1958, 1, 233-240.
(47) Karplus, M.; Fraenkel, G. K. J. Chem. Phys. 1961, 35, 1312-1323.
Inorganic Chemistry, Vol. 42, No. 7, 2003 2309

Ikezaki and Nakamura
Table 4. Separation of 13C NMR Chemical Shift Terms in [Mn(TiPrP)(CN)2]- in C6D6 at 25 °C Together with π Spin Densitiesa
δMC
δLC
δcon
FC
b
π
carbons
δobs
δdia
δiso
dip
dip
meso
141.0 (11.7)
293.0 (374.5)
-53.0 (-37.0)
39.7 (198.7)
85.8
122.3 (119.1)
140.4 (143.4)
131.2 (133.4)
34.7 (142.4)
28.9
18.7 (-107.4)
152.6 (231.1)
-184.2 (-170.4)
5.0 (56.3)
-32.9 (-23.0)
-52.2 (-33.1)
-19.2 (-12.0)
-10.2 (-7.8)
-5.4
61.1 (179.3)
47.8 (-173.0)
-288.7 (-217.6)
0.0 (0.0)
-9.5 (-263.7)
157.0 (437.2)
123.7 (59.2)
15.2 (64.1)
62.3
-0.0014 (-0.0059)
-0.0011 (0.0056)
0.0066 (0.0071)
R-pyrrole
â-pyrrole
CH(ipso)
CH3
56.9
0.0
a Numbers in the parentheses are those of [Mn(TPP)(CN)2]-. D values for [Mn(TiPrP)(CN)2]- and [Mn(TPP)(CN)2]- are -44 000 and -31 000, respectively.
b Chemical shifts of the corresponding carbons in diamagnetic [Co(TiPrP)(CN)2]- and [Co(TPP)(CN)2]-.
tion of the porphyrin core could not strengthen the Mn (dxy)
and porphyrin (a2u) interaction in contrast to the case of the
corresponding Fe(III) complex [Fe(TiPrP)(CN)2]-.
Since FπCR is supposed to be zero, the contact shift value of
the methine carbon is given by eq 14.
Comparison with the Corresponding Fe(III) Com-
δ
con(CR) ) QCC′CFπmesoFC
(14)
plexes. We have recently reported that the electronic state
of the low-spin iron(III) in [Fe(Et-TPP)(CN)2]- is sensitively
affected by the solvents used. For example, while the
complex adopts the (dxy)2(dxz, dyz)3 electron configuration in
benzene, CCl4, and acetonitrile, it exhibits the (dxz, dyz)4(dxy)1
electron configuration in methanol, chloroform, and dichlo-
romethane. Formation of the (dxz, dyz)4(dxy)1 type complex
in the latter solvents has been explained in terms of the
intermolecular O-H‚‚‚N or C-H‚‚‚N hydrogen bonding
between the solvent molecules and the coordinated cyanide
ligands; the hydrogen bonding enhances the iron(dπ)-CN-
(pπ*) interactions and stabilizes the (dxz, dyz)4(dxy)1 electron
configuration. Since the chemical shifts of the pyrrole protons
and meso carbons correlate with the electron configuration
of the low-spin iron(III) ion, we have concluded that the
(dxy)2(dxz, dyz)3 character increases as the solvent changes from
CD3OD to CCl4 as shown in eq 15.5
The SC and QC values are as follows: SC ) -35.5 MHz;
QC ) +40.3 MHz; QC ) +54.6 MHz; QC ) +40.3
CC′
CH
CN
MHz; QC ) -39.0 MHz; QC ) -39.0 MHz.44,47 By
C′C
NC
putting δcon(CR) ) 15.2 ppm into eq 14, we could determine
Fπmeso ) -0.0014. Solution of the simultaneous eqs 10-13
using Fπmeso ) -0.0014 and Fπpy-â ) 0.0066 yields Fπ , Fπ ,
R
N
and D to be -0.0011, -0.022, and -44000, respectively.
Table 4 lists the separation of chemical shift terms together
with the spin densities of carbon atoms.
The data in Table 4 should be compared with those in
[Mn(Por)(CN)2]- (Por ) TPP and OEP) reported by Hansen
and Goff; the Fπ
and Fπ
values are -0.015 and
meso
py-â
+0.0076, respectively.8 Thus, while the Fπpy-â values of the
two complexes are quite similar, the spin densities at the
meso carbon are greatly different; the Fπ
value in
meso
[Mn(TiPrP)(CN)2]- is only 1/10 of that in [Mn(Por)(CN)2]-
(Por ) OEP). The large negative Fπ
value in [Mn(Por)-
CD3OD < CDCl3 < CD2Cl2 <
meso
(CD3)2CO, DMSO-d6, DMF-d7 <
C6D6, C6D5CD3, CD3CN < CCl4 (15)
(CN)2]- could be the indication that [Mn(Por)(CN)2]- has
much larger positive π spin densities at the R-pyrrole carbon
atoms. To determine the spin densities at the R-pyrrole
carbon and nitrogen atoms of [Mn(TPP)(CN)2]-, we have
measured the 13C NMR chemical shifts in C6D6 solution at
25 °C and listed them in Table 4. As expected, the R-pyrrole
signal appears at 374.5 ppm while the meso signal is
observed at 11.7 ppm; the R-pyrrole and meso signals are
observed at 293 and 141.0 ppm, respectively, in the case of
[Mn(TiPrP)(CN)2]-. The spin densities of [Mn(TPP)(CN)2]-
are similarly determined on the basis of the 1H and 13C NMR
It should be noted that the order given in eq 15 parallels the
order given in eq 3. Thus, while the intermolecular hydrogen
bonding controls the electron configurations of low-spin iron-
(III) complexes [Fe(Por)(CN)2]-, it controls the product ratios
[mono-adduct]/[bis-adduct] in the reaction between Mn(Por)-
Cl and CN-. Especially important is the fact that the product
ratio changes to a great extent even between CH2Cl2 and
CHCl3; Mn(TiPrP)Cl (9.7 µM) is completely converted to
the bis-adduct in CH2Cl2 solution though the conversion from
the mono- to the bis-adduct is not observable even in the
presence of 7000 equiv of Bu4N+CN- in CHCl3 solution.
The result clearly indicates that although CHCl3 and CH2-
Cl2 are regarded as similar solvents in many aspects, they
behave as completely different solvents especially in some
coordination complexes.
chemical shifts, and they are also listed in Table 4. The Fπ
py-â
value is consistent with that reported by Goff and Hansen,
though the Fπ
value is much smaller; the Fπ
value
meso
meso
reported by Goff and Hansen was determined on the basis
1
of the H NMR chemical shift of the analogous [Fe(OEP)-
(CN)2]-. The data in Table 4 indicate that [Mn(TPP)(CN)2]-
has much larger spin densities on the pyrrole carbons than
[Mn(TiPrP)(CN)2]-. The smaller spin densities on the pyrrole
carbons in [Mn(TiPrP)(CN)2]- should be ascribed to the weak
Mn (dπ) and porphyrin (3eg) interactions due to the S4 ruffling
of the porphyrin ring. Since the dxy orbital in [Mn(TiPrP)-
(CN)2]- is fully occupied by the electron pair, the deforma-
Acknowledgment. This work was supported by the Fund
for the Advancement of Science in commemoration of Toho
University’s 60th anniversary. A.I. thanks the Shibata Fund
for Young Scientists in Toho University for financial support.
IC0206138
2310 Inorganic Chemistry, Vol. 42, No. 7, 2003