Solid state 15N and 13C NMR study of several metal 5,10,15,20- tetraphenylporphyrin complexes

Article abstract of DOI:10.1021/ja970447g

The principal values of both the ¹³C and ¹⁵N chemical shift tensors are reported for the Zn, Ni, and Mg 5,10,15,20- tetraphenylporphyrin(TPP)complexes. The principal values of the ¹⁵N chemical shift tensors were obtained from static powder patterns of ¹⁵N- enriched samples. Due to overlap between the powder patterns of the different carbons, the ¹³C values were obtained using the recently developed magic angle turning (MAT) 2D experiment on unenriched materials. The measured principal values are presented along with theoretical calculations of the chemical shift tensors and a discussion of the effects that the metal bonding has on the chemical shift tensors in these compounds. Both the isotropic chemical shift and the principal values of the ¹⁵N chemical shaft tensor are nearly identical for the Mg and Zn complexes. The ¹⁵N isotropic chemical shift for the NiTPP, however, changes by nearly 80 ppm relative to the Mg and Zn values, with large changes observed in each of the three principal values. Calculations show that the differences between the ¹⁵N chemical shifts are almost entirely determined by the metal-nitrogen separation. In addition, both the experimental data and the calculations show only very minor differences in the ¹³C chemical shift tensor components as the metal is changed.

Full text of DOI:10.1021/ja970447g

7114
J. Am. Chem. Soc. 1997, 119, 7114-7120
15
13
Solid State N and C NMR Study of Several Metal
5,10,15,20-Tetraphenylporphyrin Complexes
Mark Strohmeier, Anita M. Orendt, Julio C. Facelli,† Mark S. Solum,‡
Ronald J. Pugmire,‡ Robert W. Parry, and David M. Grant*
Contribution from the Department of Chemistry, Department of Chemical and Fuels Engineering,
and the Center for High Performance Computing, UniVersity of Utah,
Salt Lake City, Utah 84112-1102
ReceiVed February 10, 1997. ReVised Manuscript ReceiVed April 15, 1997X
Abstract: The principal values of both the 13C and 15N chemical shift tensors are reported for the Zn, Ni, and Mg
5,10,15,20-tetraphenylporphyrin (TPP) complexes. The principal values of the 15N chemical shift tensors were obtained
from static powder patterns of 15N-enriched samples. Due to overlap between the powder patterns of the different
carbons, the 13C values were obtained using the recently developed magic angle turning (MAT) 2D experiment on
unenriched materials. The measured principal values are presented along with theoretical calculations of the chemical
shift tensors and a discussion of the effects that the metal bonding has on the chemical shift tensors in these compounds.
Both the isotropic chemical shift and the principal values of the 15N chemical shift tensor are nearly identical for the
Mg and Zn complexes. The 15N isotropic chemical shift for the NiTPP, however, changes by nearly 80 ppm relative
to the Mg and Zn values, with large changes observed in each of the three principal values. Calculations show that
the differences between the 15N chemical shifts are almost entirely determined by the metal-nitrogen separation. In
addition, both the experimental data and the calculations show only very minor differences in the 13C chemical shift
tensor components as the metal is changed.
Introduction
measured in the free base tetraphenylporphyrin.12 There are
also solution 13C NMR studies of the effect that different metals
have on the isotropic chemical shifts,13-15 as well as solution
and solid state measurements on the metal itself.16,17 Additional
studies have measured the effect the addition of ligands has on
the chemical shifts of metal-porphyrin complexes.18
The measurement of the principal values of the chemical shift
tensor, especially in conjunction with quantum chemical cal-
culations of the complete tensor, provides details about elec-
tronic structure that are not observed when only the isotropic
chemical shift is measured.19 With the advent of the 2D
techniques which separate the individual powder patterns of
inequivalent carbons, especially the magic angle turning (MAT)
method,20 the complexity of systems that can be studied has
greatly increased. In addition, due to advances in both
computational methods and in available computer resources,
quantum calculations are now feasible in these large systems.
These advances have made it possible to obtain the anisotropic
chemical shift data in systems such as the metallotetraphe-
nylporphyrins.
In many natural products porphyrins, metalloporphyrins, and
related macrocycles form the reactive site for important biologi-
cal activity.1 They are also abundant in geological formations
such as oils, coals, soils, etc. Thus porphyrins and related
molecules have numerous applications in all fields of chemistry
and in medicine. Because of this importance and their interest-
ing chemical, physical, and biological properties, they have been
widely investigated for over 100 years.
Since NMR chemical shift data can provide detailed insights
into the chemical structure, NMR methods are often used in
the investigation of these compounds. For example, there are
numerous publications on the dynamics of the hydrogen
1
2
rearrangement in various free base porphyrins, using H, H,
3H, 13C, and 15N solution state NMR,2-7 as well as a number
of 13C8-10 and 15N2,5,11 solid state NMR measurements. The
principal values of the 15N chemical shift tensor have been
† Center for High Performance Computing.
‡ Department of Chemical and Fuels Engineering.
X Abstract published in AdVance ACS Abstracts, July 1, 1997.
(1) Smith, K. M. Porphyrins and Metalloporphyrins; Elsevier: Amster-
dam, 1975.
In this paper the principal values of the 15N and the majority
of the 13C chemical shift tensors present in the Zn, Mg, and Ni
complexes of 5,10,15,20-tetraphenylporphyrin are presented
along with DFT-GIAO calculations of the complete shift tensors
(2) Schlabach, M.; Wehrle, B.; Rumpel, H.; Braun, J.; Scherer, G.;
Limbach, H.-H. Ber. Bunsenges. Phys. Chem. 1992, 96, 821.
(3) Limbach, H.-H.; Hennig, J. J. Am. Chem. Soc. 1984, 106, 292.
(4) Stilbs, P. J. Magn. Reson. 1984, 58, 152.
(5) Braun, J.; Schlabach, M.; Wehrle, B.; Kocher, M.; Vogel, E.;
Limbach, H.-H. J. Am. Chem. Soc. 1994, 116, 6593.
(6) Braun, J.; Limbach, H.-H.; Williams, P. G.; Morimoto, H.; Wemmer,
D. E. J. Am. Chem. Soc. 1996, 118, 7231.
(7) Braun, J.; Schwesinger, R.; Williams, P. G.; Morimoto, H.; Wemmer,
D. E.; Limbach, H.-H. J. Am. Chem. Soc. 1996, 118, 11101.
(8) Frydman, L.; Olivieri, A. C.; Diaz, L. E.; Frydman, B.; Morin, F.
G.; Mayne, C. L.; Grant, D. M.; Adler, A. D. J. Am. Chem. Soc. 1988,
110, 336.
(9) Frydman, L.; Olivieri, A. C.; Diaz, L. E.; Valasinas, A.; Frydman,
B. J. Am. Chem. Soc. 1988, 110, 5651.
(10) Frydman, L.; Olivieri, A. C.; Diaz, L. E.; Frydman, B.; Kustanovich,
I.; Vega, S. J. Am. Chem. Soc. 1989, 111, 7001.
(12) Anderson-Altmann, K. L. Ph.D. Thesis; University of Utah; 1994.
(13) Kawano, K.; Ozaki, Y.; Kyogoku, Y.; Ogoshi, H.; Sugimoto, H.;
Yoshida, Z.-I. J. Chem. Soc., Perkin Trans. 2 1978, 1319.
(14) Abraham, R. J.; Hawkes, G. E.; Smith, K. M. J. Chem. Soc., Perkin
Trans. 2 1974, 627.
(15) Abraham. R. J.; Hawkes, G. E.; Hudson, M. F.; Smith. K. M. J.
Chem. Soc. Perkin Trans. 2 1975, 204.
(16) Jakobsen, H. J.; Ellis, P. D.; Inners, R. R.; Jensen, C. F. J. Am.
Chem. Soc. 1982, 104, 7442.
(17) Dominquez, D. D. Ph.D. Thesis; University of Washington; 1979.
(18) Gust, D.; Neal, D. N. J. Chem. Soc., Chem. Commun. 1978, 681.
(19) Facelli, J. C. In Encyclopedia of Nuclear Magnetic Resonance;
Grant, D. M., Harris, R. K., Eds; John Wiley: London, 1996; p 4327.
(20) Gan, Z. J. Am. Chem. Soc. 1992, 114, 8307.
(11) Limbach, H.-H.; Hennig, J.; Kendricks, R.; Yannoni, C. S. J. Am.
Chem. Soc. 1984, 106, 4059.
S0002-7863(97)00447-2 CCC: $14.00 © 1997 American Chemical Society

NMR Study of Metal-5,10,15,20-Tetraphenylporphyrins
J. Am. Chem. Soc., Vol. 119, No. 30, 1997 7115
para-C), 132.17 (t, â-pyrrole-C), 133.72 (t, ortho-C), 140.97 (q, ipso-
C), 142.71 (q, R-pyrrole-C). CI-MS [m/z (%)]: 575 (M+, 100), 603
(∆ ) 28, +CH2CH3+, 5.86).
in free base porphyrin and its Ca, Zn, Mg, and Ni complexes.
The influence of the metal on the chemical shift values is
discussed.
Magnesium 5,10,15,20-Tetraphenylporphyrin-15N4.1,26 A 110 mg
amount of pure 15N-labeled H2TPP (0.18 mmol) and 1 g of MgClO4
were dissolved in 25 mL of pyridine and refluxed overnight. After
the reaction solution was cooled, the precipitate was filtered off and
washed with diethyl ether until the filtrate was colorless. The magenta
ether/pyridine phase was washed three times with 100 mL of distilled
water, dried over Na2SO4, and rotary evaporated. The violet compound
was then placed under high vacuum at 130 °C for two days to remove
the pyridine coordinated to the MgTPP. Yield: 90 mg (79%). 13C-
NMR (125.64 MHz; CDCl3): δ (ppm) ) 121.65 (q, meso-C), 126.24
(t, meta-C), 127.07 (t, para-C), 131.83 (t, â-pyrrole-C), 134.70 (t, ortho-
C), 143.84 (q, ipso-C), 150.02 (q, R-pyrrole-C). CI-MS [m/z (%)]:
641 (M+, 100), 669 (∆ ) 28, +CH2CH3+, 5.86).
Solid State Spectroscopy. The solid state NMR measurements were
made at room temperature on a Varian VXR-200 spectrometer system,
with a 15N frequency of 20.279 MHz and a 13C frequency of 50.318
MHz. The CP/MAS and the 15N static experiments were performed
using a 7 mm high speed variable temperature probe from Doty
Scientific, Inc. The 15N spectra were recorded at room temperature
using cross polarization with a 15 ms contact time and a proton 90°
pulse of about 7 µs. All 15N spectra were recorded using the 15N-
enriched materials.
The 13C MAT experiments20,27,28 were performed on the natural
abundance nitrogen tetraphenylporphyrin samples, as insufficient quan-
tities of the 15N-enriched material were available so that the MAT
spectra could be obtained in a reasonable amount of time. A home
built low speed spinning large volume (1.5 cm3) probe29 with a feedback
loop to synchronize the pulse sequence to the rotor position was used.28
A triple-echo MAT pulse sequence27 with flip back, to allow shorter
delay times, was used. The following experimental conditions were
used: spinning rate about 16 Hz, spectral width in the acquisition
dimension of 30 kHz, recycle time of 2 s, 90° proton pulse of 4.7 µs,
contact time of 4 to 5 ms, spectral width in the evolution dimension of
10 kHz (5 kHz for NiTPP), 800 scans per increment (1024 for NiTPP),
and 37 complex 2D increments (28 for NiTPP). The 2D spectra were
processed and sheared on a VAXstation 3100. When possible the
powder pattern of each individual isotropic shift was obtained and fitted
using a least squares fitting routine with the POWDER30 approach;
otherwise, the principal values were determined by visual comparison
of the simulated spectral line shape to the experimental pattern. The
CR patterns were fit on a IBM RS6000 computer using a POWDER-
based fitting routine which accounts for the dipolar coupling to the
14N.
The literature 13C solution assignments for ZnTPP were confirmed
by performing a 2D INADEQUATE experiment. The spectrum was
analyzed using the FRED31 software. Since the 13C spectra were similar
for all the metallotetraphenylporphyrins, proton-coupled 13C experiments
were sufficient to make unambiguous assignments for the other
compounds. These experiments were performed on a Varian UnityPlus
500 MHz spectrometer system. As the metallotetraphenylporphyrins
are only slightly soluble, the experiments were performed on a 10 mm
broadband probe with a 24 µs 90° pulse in the carbon channel. The
2D INADEQUATE had a 4.0 kHz spectral width in both the acquisition
and double quantum evolution dimension and was optimized to detect
carbon-carbon bonds with a Jcc of 60 Hz. A total of 40 increments,
with 800 scans per increment, were recorded. The acquisition time
was 1.5 s and the delay time 4.5 s.
Experimental Section
Synthesis. The 15N-labeled pyrrole was received from Cambridge
Isotope Laboratory and used without further purification.21 Thin layer
chromatography (TLC) was performed on silica gel from Whatman
(AL SIL G/UV) with CHCl3/petroleum ether (3:1) as eluent. Larger
amounts (1.2-1.5 g) of unlabeled metalloporphyrins were synthesized
from commercial H2TPP, obtained from Aldrich Chemical Company,
following the identical procedure described below for the labeled
compounds.
5,10,15,20-Tetraphenylporphyrin-15N4.22-25 A 1 mL amount of
15N-labeled pyrrole (14 mmol) and 1.5 mL of benzaldehyde (14 mmol)
were dissolved in 1 L of dry CH2Cl2 and stirred under nitrogen for 15
min at room temperature. Then 0.45 mL of 2.5 M BF3 in diethyl ether
was added to the clear solution while shielding the reaction from
ambient lighting. The solution turned dark red immediately. After
90 min, 2.75 g of p-chloranil was added, and the reaction vessel was
immersed into a 45 °C preheated water bath and refluxed for 80 min.
The solution was concentrated by rotary evaporation to ca. 100 mL,
mixed with 12 g of Florisil (100-200 mesh) and evaporated to dryness.
The dark powder was poured on top of a column of 2.5 cm diameter
dry packed with ca 20 cm of Florisil (100-200 mesh) and eluted with
500 mL of CH2Cl2/petroleum ether (1:1) to give fraction 1. Further
elution with CH2Cl2 gave fraction 2 and 3. After evaporating the
solvent the fractions yielded still impure H2TPP. Fraction 1 yielded
476 mg; fraction 2, 661 mg; and fraction 3, 76 mg. The portion of
H2TPP used in the Zn and Ni metalation reactions was not purified
further. The remaining H2TPP was recrystallized from a methanol/
chloroform solution by slowly evaporating CHCl3 from the H2TPP and
adding methanol. 1H-NMR (399.9 MHz): δ (ppm) ) 8.84 (s, 8 H,
meso-H), 8.22 (q, 8 H, o-phenyl), 7.76 (m, 12 H, m,p-phenyl). 13C-
NMR (100.64 MHz; CDCl3): δ (ppm) )120.20 (q, meso-C); 126.75
(t, meta-C); 127.78 (t, para-C); 131(broad, â-pyrrole-C), 134.63 (t,
ortho-C), 142.22 (q, ipso-C), 145 (broad, R-pyrrole-C). CI-MS [m/z
(%)]: 619 (M+, 74.04), 647 (∆ ) 28, +CH2CH3+, 4.69).
Zinc 5,10,15,20-Tetraphenylporphyrin-15N4.1 A 200 mg amount
of the impure 15N-labeled H2TPP (0.33 mmol) was dissolved in 40
mL of CHCl3, and 1 mL of saturated ZnAc2 in methanol solution was
added. The solution was refluxed for 2 h and monitored by TLC.
During rotary evaporation of the solution, methanol was gradually added
to exchange the CHCl3 with methanol. When all the CHCl3 was
exchanged with methanol, purple crystals were collected to yield 100
mg of 15N-labeled ZnTPP. 1H-NMR (399.9 MHz, CDCl3): δ (ppm)
) 8.95 (s, 8 H, meso-H), 8.22 (q, 8 H, o-phenyl), 7.77 (m, 12 H, m,p-
phenyl). 13C-NMR (125.64 MHz; CDCl3): δ (ppm) ) 121.15 (q, meso-
C), 126.52 (t, meta-C), 127.48 (t, para-C), 131.97 (t, â-pyrrole-C),
134.40 (t, ortho-C), 143.81 (q, ipso-C), 150.25 (q, R-pyrrole-C). CI-
MS [m/z (%)]: 681 (M+, 100), 709 (∆ ) 28, +CH2CH3+, 5.86).
Nickel(II) 5,10,15,20-Tetraphenylporphyrin-15N4.1 A 200 mg
amount of the impure 15N-labeled H2TPP (0.33 mmol) was dissolved
in 40 mL of CHCl3, and 7 mL of saturated NiAc2 in methanol solution
was added. The solution was refluxed for 5 days. The reaction was
monitored by TLC. During rotary evaporation of the CHCl3 solution,
methanol was gradually added. From the methanol solution a blue
powder precipitated. When all CHCl3 had been exchanged with
methanol the blue powder was collected to yield 110 mg of 15N-labeled
NiTPP. 1H-NMR (399.9 MHz): δ (ppm) ) 8.74 (s, 8 H, meso-H),
8.01 (q, 8 H, o-phenyl), 7.68 (m, 12 H, m,p-phenyl). 13C-NMR (125.64
MHz; CDCl3): δ ) 118.99 (q, meso-C), 126.87 (t, meta-C), 127.74 (t,
Quantum Chemical Calculations. DFT (Density Functional
Theory) calculations were performed with the Gaussian 94 computer
(26) Baum, S. J.; Burnham, B. F.; Plane, R. A. Proc. Nat. Acad. Sci.
U.S.A. 1964, 52, 1439.
(27) Hu, J. Z.; Orendt, A. M.; Alderman, D. W.; Pugmire, R. J.; Ye, C.;
Grant, D. M. Solid State NMR 1994, 3, 181.
(28) Hu, J. Z.; Wang, W.; Liu, F.; Solum, M. S.; Alderman, D. W.;
Pugmire, R. J.; Grant, D. M. J. Magn. Reson. Ser. A 1995, 113, 210.
(29) Jiang, Y. J.; Pugmire, R. J., Grant, D. M. Manuscript in preparation.
(30) Solum, M. S.; Alderman, D. W.; Grant, D. M. J. Chem. Phys. 1986,
84, 3717.
(31) CBOND program. Varian, Palo Alto, CA. CCBOND is marketed
by Varian as part of the Full Reduction of Entire Datasets (FRED) software.
(21) The commercial pyrrole was brown but could not be distilled because
of the small amount. By using the impure material, the yield was decreased.
(22) Lindsay, J. S.; Wagner, R. W. J. Org. Chem. 1989, 54, 828.
(23) Wagner, R. W.; Lawrence, D. S.; Lindsay, J. S. Tetrahedron Lett.
1987, 28, 3069.
(24) Lindsay, J. S.; Hsu, H. C.; Schreiman, I. C.; Kearney, P. C.;
Marguerettaz, A. M. J. Org. Chem. 1987, 52, 827.
(25) During work on this synthesis a more recent approach appeared in
the literature. Gradillas, A.; del Campo, C.; Sinisterra, J. V.; Llama, E. F.
J. Chem. Soc., Perkin Trans. 1 1995, 2611.

7116 J. Am. Chem. Soc., Vol. 119, No. 30, 1997
Strohmeier et al.
program.32 All calculations employed the GIAO33 (Gauge Invariant
Atomic Orbitals) method with either the 3-21g Pople basis set34 for all
atoms or the D95 basis set,35 for carbons, nitrogens, and hydrogens
and the Los Alamos ECP plus DZ basis (LanL2DZ)36 for the metals.
Calculations with both basis sets were performed when possible. The
DFT calculations use the BLYP exchange correlation functional37,38
and a coupled perturbative scheme without including the magnetic field
effects in the exchange correlation functional.39 The calculated 15N
chemical shieldings were converted to the shift scale by subtracting
the absolute shielding value of nitromethane, -135.8 ppm.40 The
calculated 13C chemical shieldings were converted to the TMS shift
scale by subtracting the TMS shielding values of 186.7 and 193.6 ppm
for the D95/LanL2DZ and 3-21g basis sets, respectively. These values
were estimated from the corresponding calculated values for methane,
193.7 and 200.6 ppm, respectively, minus the 7 ppm difference between
the chemical shieldings of TMS and methane reported in the literature.41
Geometry optimizations on the metal-TPP complexes using the
D95/LanL2DZ basis set were not feasible with the available computer
resources, and only the ZnTPP has a known molecular structure.42 This
room temperature X-ray structure predicts two very different Zn-N
distances (2.045 and 2.029 Å), which, according to the calculations
(see below), should lead to two very different 15N isotropic chemical
shifts. However, there is no NMR experimental evidence to support
this nonequivalence. The very sharp features observed in the 15N NMR
spectra (both CP/MAS and static pattern) of ZnTPP indicate that any
nonequivalence in both the isotropic and the principal values of the
tensor should be less than 3 ppm. In addition, a 15N CP/MAS spectrum
recorded at about -90 °C shows no splitting, ruling out the possibility
that some type of dynamic averaging is occurring in the room
temperature spectrum. A powder diffraction taken on the ZnTPP
sample used for the NMR measurements indicates that the sample is
highly crystalline; however, the pattern lacks agreement with the
literature X-ray data. One possible explanation for these discrepancies
between the NMR and X-ray results may be that the NMR studies were
done on a different crystal form than the X-ray study. Further studies
are necessary to explore this suggestion. It should be noted that
polymorphism has been previously observed in other porphyrins, and
specifically in the free base TPP.43,44 Fortunately, as will be seen in
the following discussion, the DFT-optimized structures of the metal
porphyrin complexes45 that exist in the literature can be used to obtain
information consistent with the experimental observations. The
experimental ZnTPP geometry is only used in the calculations necessary
to explore the effect of including the phenyl groups in the TPP ring on
the calculated nitrogen chemical shift tensor.
Table 1. Comparison of Calculated 15N Chemical Shift Tensor
Principal Values for the Three Modelsa
model
ZnTPP N1
δ11
10
-47
-19
18
-41
-12
13
-42
-15
-16
δ22
δ33
-181
-178
-180
-179
-177
-178
-179
-178
-179
-124
-398
-419
-409
-392
-416
-404
-399
-420
-410
-393
N2
average of N1 and N2
ZnTMP N1
N2
average of N1 and N2
ZnTHP N1
N2
average of N1 and N2
experimental
a Models used are described in text. Reported results are in ppm
relative to nitromethane, as discussed in text.
of the inclusion of the phenyl rings in the calculations of the 15N
chemical shifts tensors. These calculations were done using the smaller
3-21g basis set for different molecular models constructed as follows.
Model I, designated ZnTPP, was constructed by adding the necessary
H atoms, at standard distances and angles from the positions of the
heavy atoms given by the X-ray structure of ZnTPP. Model II, ZnTMP,
was constructed by replacing the phenyl groups in the ZnTPP by methyl
groups. Standard geometrical parameters were used to determine the
position of and bond lengths within the methyl groups. Model III,
ZnTHP, was built using the ZnTPP structure and replacing the phenyl
groups by hydrogen atoms. By preserving the geometry of the
metalloporphyrin core and changing only the meso substituents a set
of calculations which depend only on the direct electronic substituent
effects was produced.
The results (15N only) of this series of calculations are given in Table
1. In this table it is observed, as was mentioned above, that as a
consequence of the lack of symmetry in the X-ray diffraction structure
the calculations predict two nonequivalent nitrogens. It is also observed
that the average values of the principal components are independent
of the substituent at the meso position, with the substituent inducing
variations of less than 5 ppm in any of the principal components. These
small variations are almost within the experimental error of the
measurements. Therefore, these results justify the use of the unsub-
stituted metalloporphyrins as model compounds to calculate the
chemical shift tensors for comparison with the experimental values
measured in the metal-TPP complexes. The use of the unsubstituted
compounds has two important advantages: it becomes possible to use
a better basis set in the calculations, and it is possible to use the already
published optimized structures of these compounds. The calculations
on the unsubstituted porphyrins have the disadvantage that no com-
parison can be made between the experimental and calculated chemical
shift tensor data for the carbons, as sizable substituent effects are
observed in these atoms. Nonetheless, these effects are nearly
independent of the metal used in the complex, as will be shown in the
results.
A series of preliminary calculations were performed using the
experimental geometry of the ZnTPP, in order to test the importance
(32) Gaussian 94 (Revision A.1), Frisch, M. J.; Trucks, G. W.; Schlegel,
H. B.; Gill, P. M. W.; Johnson, B. G.; Robb, M. A.; Cheeseman, J. R.;
Keith, T. A.; Petersson, G. A.; Montgomery, J. A.; Raghavachari, K.; Al-
Laham, M. A.; Zakrzewski, V. G.; Ortiz, J. V.; Foresman, J. B.; Cioslowski,
J.; Stefanov, B. B.; Nanayakkara, A.; Challacombe, M.; Peng, C. Y.; Ayala,
P. Y.; Chen, W.; Wong, M. W.; Andres, J. L.; Replogle, E. S.; Gomperts,
R.; Martin, R. L.; Fox, D. J.; Binkley, J. S.; Defrees, D. J.; Baker, J.; Stewart,
J. P.; Head-Gordon, M.; Gonzalez, C.; Pople, J. A. Gaussian, Inc.,
Pittsburgh, PA, 1995.
In this work, therefore, the previously published DFT-optimized
structures45 for the parent porphyrin and for the Ca-porphyrin, Zn-
porphyrin, Mg-porphyrin, and Ni-porphyrin complexes are used. We
have included calculations on Ca-porphyrin because of its metal-
nitrogen distance, even though no experimental chemical shift data have
been measured for this compound. The Ca-nitrogen distance is one
of the longest, while the Zn-nitrogen and Mg-nitrogen distances are
intermediate and the Ni-nitrogen is one of the shortest for the metal-
porphyrin compounds. It is interesting to note that the major structural
change that occurs in the porphyrin unit to allow for this shortening of
the metal-nitrogen distance is a decrease in the CR-Cmeso-CR angle.
In all cases the metal lies in the plane of the porphyrin ring.
(33) Ditchfield, R. Mol. Phys. 1974, 27, 789.
(34) Pietro, W. J.; Francl, M. M.; Hehre, W. J.; Defrees, D. J.; Pople, J.
A.; Binkley, J. S. J. Am. Chem. Soc. 1982, 104, 5039.
(35) Dunning, T. H.; Hay, P. J. In Modern Theoretical Chemistry;
Schaefer, H. F., III, Ed.; Plenum: New York, 1977; p 1.
(36) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 270. Ibid. pp 284
and 299.
(37) Lee, C.; Yang, W.; Parr, R. G. Phys. ReV. 1988, B37, 785.
(38) Becke, A. D. Phys. ReV. 1988, A38, 3098.
(39) Cheeseman, J. R.; Trucks, G. W.; Keith, T. A.; Frisch, M. J. J.
Chem. Phys. 1996, 104, 5497.
(40) Jameson, C. J.; Mason, J. In Multinuclear NMR; Mason, J., Ed.;
Plenum: New York, 1987, p 56.
Results and Discussion
(41) Jameson, A. K.; Jameson, C. J. Chem. Phys. Lett. 1987, 134, 461.
(42) Scheidt, W. R.; Mondal, J. U.; Eigenbrot, C. W.; Adler, A.;
Radonovich, L. J.; Hoard, J. L. Inorg. Chem. 1986, 25, 795.
(43) Silvers, S. J.; Tulinsky, A. J. Am. Chem. Soc. 1967, 89, 3331.
(44) Hamor, T. A.; Caughey, W. S.; Hoard, J. L. J. Am. Chem. Soc.
1965, 87, 2305.
The atomic designations used are shown in Figure 1. Figure
2 contains the 15N powder patterns obtained for each of the
three metal-TPP complexes along with their best spectral fits.
The 13C MAS spectra of all three of the metal-TPP complexes
labeled with the assignments are given in Figure 3, and Figure
4 shows the 1D slices obtained from the 13C 2D MAT of NiTPP.
(45) Matsuzawa, N; Ata, M.; Dixon, D. A. J. Phys. Chem. 1995, 99,
7698.

NMR Study of Metal-5,10,15,20-Tetraphenylporphyrins
J. Am. Chem. Soc., Vol. 119, No. 30, 1997 7117
Figure 1. Structure of the metal complexes of 5,10,15,20-tetraphe-
nylporphyrin, showing the labeling of the carbons used in this paper.
Figure 3. 13C CP/MAS spectra for (a) MgTPP, (b) ZnTPP, and (c)
NiTPP. Assignments of the different resonances are indicated.
Figure 2. Experimental 15N powder patterns along with the best fit
for (a) MgTPP, (b) ZnTPP, and (c) NiTPP.
15N Chemical Shifts. The resultant principal values of the
chemical shift tensors are reported in Table 2, along with the
tensor components for the nitrogens in the related species of
pyrrole and the pyrrole anion.46 As may be seen from the table,
the principal values are nearly identical for the Mg and Zn, and
these values are similar to those obtained for the pyrrole anion.
The values obtained for the NiTPP are very different from those
in the Mg- and ZnTPP, but are similar to the values measured
for the nitrogen in the parent pyrrole, with an exception for a
large difference between the δ33 values.
The principal values reflect the change in the bonding present
between the metal and the porphyrin ring. The similarity of
both the isotropic chemical shift and the principal values of the
chemical shift tensor for the Mg and Zn complexes is not too
surprising as the 2+ ions of both have a closed outer shell,with
Figure 4. Slices taken from the 2D MAT spectrum of ZnTPP. The
isotropic chemical shift of each slice is listed.
the Mg2+ having a s2p6 electronic configuration and the Zn2+ a
d10 electronic configuration. Ni2+, on the other hand, has a d8
electronic configuration, which allows for the formation of a
much stronger coordination bond between the TPP nitrogen and
the nickel. In addition, the interaction between the metal and
nitrogen has more covalent nature in NiTPP than in either Zn-
or MgTPP, and in the NiTPP there also exists a substantial
amount of backbonding from the metal to the porphyrin ring.
The large shift observed in the δ33 component in NiTPP is
(46) Solum, M. S.; Anderson-Altmann, K. L.; Strohmeier, M.; Berges,
D. A.; Zhang, Y.; Facelli, J. C.; Pugmire, R. J.; Grant, D. M., J. Am. Chem.
Soc., in press.

7118 J. Am. Chem. Soc., Vol. 119, No. 30, 1997
Strohmeier et al.
Table 4. 13C Chemical Shift Tensor Principal Values for
Table 2. Experimental 15N Chemical Shift Tensor Principal Values
of the Metal-TPP Complexes and Related Compoundsa
Metal-TPP Complexesa
compound
δ11
δ22
δ33
δiso
δCP/MAS
δsoln
carbon
δ11
δ22
δ33
NiTPP
δiso
δCP/MAS
δliq
ZnTPP
MgTPP
-16 -124 -393 -178 -178 -180.5
-10 -118 -392 -173 -179 -177.6
-136 -152 -478 -255 -254 -253.7
CR
Ci
Co
Câ
Cm,p
143.6
139.6
134.0
130.0
126.0
142.7
140.9
133.7
132.2
m: 127.7
p: 126.9
118.9
NiTPP
236
215
217
156
147
141
24
30
24
139
131
127
pyrrole anionb
pyrroleb
-39
-90 -372 -167
-136 -234 -328 -233
a All measurenments were done on 15N enriched samples. The
chemical shifts are reported in ppm relative to nitromethane. b Values
taken from reference 46.
Cmeso
185
231
149
162
33
122
119.7
MgTPP
Table 3. Calculated 15N Chemical Shift Tensor Principal Values
for Porphyrin and Metal-Porphyrin Complexesa
CR
Ci
Co,â
150.4
143.2
132.6
149.9
143.7
o: 134.7
â: 131.8
m: 127.0
p: 126.2
121.6
27
140
R-
(Me-N),b
Cm,p
125.8
123.2
compound
Å
δ11
δ22
δ33
porphyrin, N-H
porphyrin, N
Ca-porphyrin
Mg-porphyrin
Zn-porphyrin
Ni-porphyrinc
-218, -170 (δr) -248, -202 -381, -354
20, 103 (δt) -147, -91 -423, -404
Cmeso
185
150
33
123
ZnTPP
2.173
2.045
2.028
1.953
-61, 31 (δt)
-182,-139 -405, -384
CR
Ci
Co,â
208
233
209
224
188
166
137
144
59
23
48
27
151.4
142.6
133.8
150.2
142.8
o: 134.4
â: 132.0
m: 127.5
p: 126.5
121.1
-90, -13 (δt) -178, -127 -406, -397
141
131
132
-69, 5 (δt)
-156, - (δr)
-178, -127 -404, -391
-162, - -422, -
b
a Values correspond to the 3-21g, D95/LanL2DZ basis set results,
respectively. The components are given in rank order. δ33 is always
perpendicular to the molecular plane and the orientation of δ11 predicted
by the calculations at both basis set levels is given in parentheses.
b Optimized metal-nitrogen distances taken from reference 45. c Cal-
culations using the D95/LanL2DZ basis set did not converge in Ni-
porphyrin; the spatial assignment reported is based on the 3-21g
calculations (see text).
Cm,p
126.6
122.7
Cmeso
186
152
31
123
a All chemical shifts obtained from the slices of 2D MAT spectra
of the natural abundance nitrogen samples, as described in the text,
and are reported in ppm relative to TMS. b There is no basis for making
the assignment between these two carbons.
As discussed previously,46 the δr component is quite insensi-
tive to the metal bonding even though large changes occur in
the δt component. These changes in δt may be rationalized by
an increase in the energy gap between the nitrogen lone pair
(n) and π* orbitals. The symmetry of this pair of orbitals is
such as to dominate the value of δt. When a bond is formed
with the unpaired electrons of the nitrogen atom, the energy
separation is increased appreciably. This behavior is similar
to that observed in a series of nitrogen heterocycles.46 Finally,
it is important to emphasize that the metal-nitrogen separation
is the most relevant factor in determining the 15N chemical shift
tensor. Note, the tensors in Mg-porphyrin and Zn-porphyrin
are quite similar, and their metal-nitrogen distances are also
comparable.
probably a direct consequence of this bonding interaction placing
a higher electron density at the nitrogen. However, before
discussing the effect of the different metals on tensor compo-
nents it becomes necessary to rely on the calculations, in order
to gain the information about the spatial orientation of the
principal values.
The results of ab initio calculations of the chemical shift
tensor are reported in Table 3. The δ33 component is found to
lie perpendicular to the plane of the porphyrin system, as is
always the case in aromatic systems. The other two components
which lie in the plane are designated the tangential component
(δt), lying perpendicular to the metal-nitrogen bond, and the
radial component (δr), lying along this bond. It is apparent that
when the D95/LanL2DZ basis set is used in the calculations,
the agreement between the calculated and experimental values
is improved over the calculations using the 3-21g basis. With
the larger basis set the agreement is such that there are no
ambiguities in the spatial assignments of the principal compo-
nents. Unfortunately, the calculation in Ni-porphyrin with
the D95/LanL2DZ basis would not converge, so only the
chemical shift values calculated at the 3-21g basis set level are
available. In this case the larger error in the 3-21g basis set
calculations coupled with the small difference between the two
in-plane shift components renders the assignment of these
orientations tentative for Ni-porphyrin. In spite of this problem,
it is apparent from the table that as the strength of the bonding
between the metal and the porphyrin ligand increases, as
indicated by a shortened metal-nitrogen distance, the 15N
chemical shift tensors move from the characteristic 5n type
nitrogen (as in the pyrrole anion) to the 5p type nitrogen (in
pyrrole), not only in the value of the tensor components but
also in the orientation of these components.46 The tangential
component is the component undergoing large changes. The
crossover of the δt and δr components is expected at a metal-
nitrogen length of about 1.95 Å, which is close to the value for
the Ni-porphyrin.
13C Chemical Shifts. The principal values of the 13C
chemical shift tensors obtained using the MAT experiment are
reported in Table 4. The assignments of the chemical shifts in
the solution were made by recording an INADEQUATE
spectrum on the ZnTPP. Unfortunately, neither NiTPP nor
MgTPP were soluble enough to obtain an INADEQUATE
spectrum; therefore, the assignments were made by a comparison
of the 13C-1H J-coupling patterns of the two compounds with
that measured in the ZnTPP. The solid state chemical shift
assignments are based on comparison of solution and solid state
isotropic chemical shifts. Due to the high degree of signal
overlap, as may be observed in the MAS spectra in Figure 3, it
was not possible to obtain all of the chemical shift tensor data
for all of the carbons.
Unfortunately, at this time a comparison cannot be made
between the metallotetraphenylporphyrins and H2TPP, due to
the increased number of resonances in H2TPP because of its
lower molecular symmetry and the additional complication of
the dynamic rearrangement of the two inner hydrogens. In
addition, the dynamics of the hydrogen rearrangement in the
free base TPP is complicated by two crystalline forms.40,41 The
dynamic process of the hydrogen rearrangement has previously
been studied by variable temperature 15N CP/MAS NMR in both

NMR Study of Metal-5,10,15,20-Tetraphenylporphyrins
J. Am. Chem. Soc., Vol. 119, No. 30, 1997 7119
Table 5. Calculation of 13C Chemical Shift Tensors Principal
Values of Porphyrin and Several Metal-Porphyrin Complexesa
when replacing Ni by either Mg or Zn observed in the calculated
chemical shift component is not observed in the experimental
values. The lack of sensitivity of the 13C chemical shift to the
metal substitution is observed in the calculations. The only large
changes (on the order of 15 ppm) observed between the free
base porphyrin and the metal-porphyrin complexes are in the
CR
Câ
Cmeso
compound
δ11 δ22 δ33 δ11 δ22 δ33 δ11 δ22 δ33
porphyrin
N-H ring
porphyrin
N ring
197 158 43 207 124 38
165 130 33 174 109 32
209 176 70 214 119 45
177 147 56 180 106 39
δ22 components of CR and Cmeso
.
The effect of the quadrupolar coupling constant of the 14N
on the dipolar coupled spectrum of the CR carbon is observed
in the CR slice of the MAT spectra. This provides an
opportunity to obtain the axes of the CR principal values from
the experiment by exploiting the spatial term in the quadrupolar
coupling Hamiltonian. As the signal to noise ratio required to
accurately fit dipolar patterns is high, only the CR pattern of
the ZnTPP was possible to fit. However, the patterns for the
NiTPP and the MgTPP possess the same features as that of the
ZnTPP, indicating that the results discussed below can be
applied to all three cases. The results of the spectral fitting
gave the principal values reported in Table 4. It should be noted
that the calculated values for CR of Zn-porphyrin agree with
the experimental values in ZnTPP, indicating that the phenyl
ring is too far away to effect the chemical shift components of
this carbon. The spectral fitting determined that the dipolar
coupling is 818 Hz, corresponding to a CR-Ν distance of 1.387
Å. No thermal corrections or anisotropy in the J coupling was
included in the estimation of this distance.47 This bond distance
is in good agreement with both the experimental distance of
1.376 Å42 and the optimized value (in Zn-porphyrin) of 1.369
Å.45 The δ33 component of the 13C chemical shift tensor for
CR is found to be perpendicular to the plane of the porphyrin
ring as expected, thereby restricting the remaining two com-
ponents to lie in the plane of the porphyrin system. In addition,
the δ22 component was determined to lie in the pyrrole plane at
an angle of 32° relative to the CR-Ν bond. This corresponds
to an orientation of the δ11 component for CR of either 10° from
the CR-Câ bond, oriented outside of the pyrrole ring, or 53°
from the CR-Câ bond, oriented inside of the pyrrole ring. These
angles have an uncertainty of about 10°. The first of these two
orientations is in reasonable agreement with the calculated
orientation for the Zn-porphyrin, which is 19°, oriented outside
of the pyrrole ring, with the D95/LanL2DZ basis set and 18°
with the 3-21g basis set. The fit of the spectrum is fairly
insensitive to the quadrupolar coupling constant of the 14N which
is estimated as less than 5 MHz.
porphyrin
203 167 57 211 122 42 174 107 19
avgb
171 139 45 177 108 36 145
92 15
Ca-porphyrin 199 183 55 211 127 48 178 123 29
168 154 44 178 115 41 151 107 23
Mg-porphyrin 202 183 55 212 125 47 182 117 29
171 153 44 178 110 40 152
99 23
Zn-porphyrin 203 182 54 212 124 47 181 116 28
172 153 45 179 111 39 149 98 22
Ni-porphyrin
-
-
-
-
-
-
-
-
-
181 147 37 180 112 42 152
87 26
a The first row of values correspond to results with the D95/LanL2DZ
basis set and the second row to those with the 3-21g basis set. b All
four meso carbons are the same.
the pure, triclinic form of free base TPP, as well as in the
tetragonal form with a mixed crystal of the free base and a metal-
complexed TPP.2,11 It is not understood if the tetragonal form
is only obtained in cases of impurities or mixed crystals,
however. A direct comparison between the experimental and
calculated 13C chemical shift tensor components is also not
possible as the calculations were performed on the metal-
porphyrin complexes instead of on the metal-TPP compounds,
for the reasons presented in the section describing the quantum
chemical calculations.
From the 13C MAS spectra in Figure 3 and isotropic chemical
shifts given in Table 4 it may be seen that, except for CR, there
is little change between the isotropic chemical shifts of three
metal-TPP complexes. The largest change is a 7 ppm
difference between the isotropic chemical shifts of the CR carbon
in NiTPP as compared to either the Zn or Mg compounds. The
remainder of the carbons have isotropic chemical shifts that
agree to within 3 ppm between the three different metals.
Again, the largest changes are observed between the Ni and
either the Zn or Mg complexes, with the latter two having
chemical shifts that differ by less than 1 ppm. Essentially, the
effect of changing the metal center is isolated in the nitrogen
chemical shifts.
From the data in Table 4 it is obvious that there are only
minor changes in the principal values of the 13C chemical shift
tensors as the metal is changed. Even though principal values
could not be extracted from the slices for the meta and para
carbons of the phenyl ring, the line shapes are essentially
identical, indicating that no significant differences exist between
the three metal complexes. In addition, the principal values
obtained for the phenyl ring carbons are typical values for
aromatic carbons, indicating that there is no interaction between
the phenyl π-system and that of the porphyrin ring. This is
expected as the angle between the plane of the porphyrin ring
and that of the phenyl group, as measured by X-ray, does not
allow for interaction between the two π-systems.42-44 Among
the porphyrin ring carbons, the variation of any shift component
of a given carbon is less than 10 ppm among the three metal
complexes.
Conclusions
In this paper the principal values of the 15N chemical shift
tensors along with the majority of those of the 13C chemical
shift tensors are reported for the Mg, Zn, and Ni complexes of
5,10,15,20-tetraphenylporphyrin. Calculations of the chemical
shift tensor are also provided for the Ca, Mg, Zn, and Ni
complexes of porphyrin. DFT-GIAO chemical shielding cal-
culations using moderate or even small basis sets provide
adequate results to interpret the experimental results for these
large, complex molecular systems containing metals. It is shown
that the calculations for the 15N chemical shift tensor are
independent of the substituent at the meso position. In addition,
the 13C chemical shift tensors were very similar for all of the
compounds, both in the experimental and calculated data, with
differences between the four metal complexes generally being
less than 5 ppm. The only significant changes observed due to
the presence of the different metals are in the nitrogen chemical
shift tensor. These changes are very large, with differences of
over 80 ppm measured in both the δ11 and the δ33 components.
The same conclusion is reached from the calculated compo-
nents in the metalloporphyrins given in Table 5. The variations
for a given carbon (CR, Câ, or Cmeso) are all about 5 ppm or
less. The only exception to this observation is the δ22
component of Cmeso, which decreases as the metal-nitrogen
distance is decreased. However, the decrease of about 10 ppm
(47) Zilm, K. W.; Grant, D. M. J. Am. Chem. Soc. 1981, 103, 2913.

7120 J. Am. Chem. Soc., Vol. 119, No. 30, 1997
Strohmeier et al.
The chemical shift tensor of the nitrogen changed from being
similar to that observed in the pyrrole anion to being similar to
that in pyrrole as the metal-nitrogen separation decreased. A
change is also observed in the calculated orientation of the two
in-plane components. Calculations on porphyrin complexes
with longer metal-nitrogen distances indicate that the δ11
component is tangential to the metal-nitrogen bond, and as the
metal-nitrogen separation is decreased the two in-plane com-
ponents become nearly degenerate and then crossover such that
the δ22 component becomes the tangential component. This
crossover appears to occur at a metal-nitrogen separation of
about 1.95 Å. This conclusion is also supported by the
15N chemical shift tensor of these two compounds are within
experimental error of each other.
Acknowledgment. This work was funded in part by the
National Institutes of Health from the Institute of General
Medical Sciences through Grant GM08521-34. Funding for
purchase of the 15N-labeled pyrrole was provided by a research
and training grant from Exxon Corporation. Computational
resources were provided by the Center for High Performance
Computing at the University of Utah. Mark Strohmeier is
grateful for a fellowship from the Technical University of
Braunschweig, funded by the German Academic Exchange
Service (DAAD), to come to the University of Utah. Special
thanks are extended to Jamie Manson for obtaining the powder
diffraction on the ZnTPP.
experimental data, as the two downfield components of the 15
N
chemical shift tensor measured for NiTPP, where the Ni-N
distance is calculated to be 1.953 Å, are only separated by 22
ppm. In addition, ZnTPP and MgTPP have similar metal-
nitrogen separations, and the measured principal values of the
JA970447G