EVERLYB. FLEISCHER, CLETAK. MILLER,AND LAWRENCE E. WEBB
2342
Xe03. Further support for this is given by the observation4 that the electrical conductivity of the solution is very low. [CONTRIBUTION FROM
THE
Vol. 86
The very weak Raman bands a t 933, 524, and around 460 cm.-' must then be assigned to another molecular species of low concentration, that is, as yet, unknown.
GEORGEHERBERTJONES LABORATORY, UNIVERSITY OF CHICAGO, CHICAGO 37, ILL.]
Crystal and Molecular Structures of Some Metal Tetraphenylporphinesl BY EVERLYB. FLEISCHER, CLETAK. MILLER,A N D LAWRENCE E . WE BE^ RECEIVEDDECEMBER 23, 1963 The crystal and molecular structures of copper tetraphenylporphine, palladium tetraphenylporphine, zinc tetraphenylporphine dihydrate, and ferric hydroxide tetraphenylporphine monohydrate were determined from three-dimensional X-ray diffraction data. These structures are all-tetragonal. The space group of both copper tetraphenylporphine and palladium tetraphenylporphine is I42d with four molecules per unit cell. The other two porphyrins have two molecules per unit cell. The space group of the zinc porphyrin is I4/m and the molecular symmetry is 4/m. The iron porphyrin has space group I4 and a molecular symmetry of 4. The molecular configuration changes from a planar to a buckled, nonplanar one depending on the crystal packing and on the substituents attached to the porphine nucleus. Copper and palladium tetraphenylporphine are very nonplanar, while the porphine nucleus of the zinc compound is planar. The iron porphyrin is nearly planar except t h a t the iron atom is 0.2 A. above the plane of the four nitrogen atoms in the porphyrin ring.
Introduction Previous structural determinations have shown that although some phthalocyanines are planar,3 nickel etioporphyrin I 4 and tetraphenylp~rphine~, are nonplanar. We reported in a previous communication6 that copper tetraphenylporphine is also nonplanar ; this paper includes the details of this structure determination. Nickel etioporphyrin 117 appears to be planar, but disorder in the crystal makes it impossible to detect suspected slight deviations from planarity.
A summary of some porphyrin crystal forms found by us and otherss-l0 is given in Table I. The large number of different crystal forms can be partially explained by the wide variety of solutions from which the crystals were grown.
Experimental The metal tetraphenylporphines were prepared by the usual methods." The P d T P P sample was kindly supplied by A . Martell.
TABLE I FORMS OF PORPHYRIN CRYSTALS 7 -
Compound
Form
a
Dimensions, b
A,------
-Oblique C
a
angles, deg.-
0
7
--Density-Obsd.
Calcd.
Za Space group Reference
Monoclinic 12.35 12.35 10.30 102 8 1.336 1.339 4 P2Ja Porphine d Triclinic TPP 1 Pi 19.2 14.7 1.26 Orthorhombic 1 2 . 0 TPP 4 9 15.12 15.12 13.94 TPP Tetragonal 5 4 I&d b 15.04 13.993 1.40 1.43 4 Iq2d 15.04c Tetragonal CuTPP b 15.088 13.987 1.48 1.51 PdTPP 15.088 4 I42d Tetragonal b 15.04 13.92 SiTPP 15.04 4 I32d Tetragonal Triclinic 9.89 13.0 101 108 93 1.29 1 P1 gh ZnTPP 6.03 b 1.29 17.2 14.6 Orthorhombic 1 4 . 8 4 P212121 ZnTPP b 13.440 13.440 9.715 ZnTPP,(H20)2 Tetragonal 2 I4/m 1.30 1.35 b 13.534 1.31 1.31 2 I4 13.534 9.820 FeTPPOH.HZ0 Tetragonal 1.17 10.3 19.5 6.75 98 Monoclinic 2 P21/c 9, 10 Etio I 1.35 1.35 14.61 14.61 12.38 Tetragonal Ni Etio I 4 I4l/amd 1.5 1.37 14.68 14.68 12.51 Tetragonal Xi Etio I1 4 I4l/amd 7 This work. Estimated uncertainties in A. for a and c, respectively, are: CuTPP, 0.02 and 0.004; a Molecules per unit cell. PdTPP, 0.002 and 0.001; FeTPPOH,H20, 0.006 and 0.006; ZnTPP.(Hz0)2,0.006 and 0.01. A. Tulinsky, private communication.
The Dresent work was undertaken t o determine if all tetraphenylporphines are nonplanar. This was accomplished by varying the metal atom in order to study the influence of changing the coordination number of the metal. The effects of changes in crystal packing were also studied. Tetraphenylporphine will be denoted by T P P , palladium tetraphenylporphine by PdTPP, etc.
Deep purple C u T P P and P d T P P crystals were grown from benzene, dark red ZnTPP .(H20)2 crystals from an acetoneethanol solution, and dark red FeTPPOH . H 2 0 crystals from a chloroform-ethanol solution.12 All of the crystals were well formed tetragonal bipyramids. The edges of the pyramid base varied from 0.1 to 0.3 mm. in length. A triclinic form of ZnTPP, listed in Table I, was grown from benzene. This crystal form decomposes slowly. Precession photographs taken a few days after preparation of the crystals showed evidence of decomposition.
( I ) T h i s research was supported by a Public Health Service G r a n t and t h e Louis Block Fund. ( 2 ) N A S A Predoctoral Fellow. (3) J M . Robertson, J . Chem. Soc., 615 (1935); 1195 (1936); J. M . Robertson and I. Woodward, i b i d , 219 (1937); 36 (1940). (4) E. B F l e i x h e r , J . A m . Chem. S o c , 86, 146 (1963). ( 5 ) J I, Hoard, hI. J . H a m o r , and T. A . Hamor, i b i d . , 86, 2334 (1963). if3) E. B. Fleischer, i b i d , 86, 1353 (1963). (7) M . B. C r u t e , Acta C v y s l . , 12, 24 (1959).
(8) C . Rimington, S . F. Mason, and 0. Kennard, Speclrochim A d a , 11, 65 (1958). (9) J . M. Goldstein, P h . D . Dissertation, University of Pennsylvania, 1959. (10) C . L. Christ a n d D . Harker, Amer. Mineralogisl, 27, 219 (1942). (11) D . T h o m a s a n d A. Martell, J . A m . Chem. SOL., 81, 5111 (1059). and
references given therein. (12) W e gratefully acknowledge t h e help of Mrs. S Choi, who skillfully grew the ZnTPP.(HzO)z a n d F e T P P O H . H z 0 crystals.
June 20, 1964 I
STRUCTURES OF ,METAL I
I
1
I
I
TETRAPHENYLPORPHINES
2343
I
60 50
-
40
-
$ c
N 30*
z
20
N-M-
V .II
I
.2
I
.3
I
A
2.
I ,5
I .G
I
*?
I
.8
I
1
Fig. 1.-Statistical test for center of symmetry. Solid curve a is the theoretical curve for noncentric structures; solid curve b is the theoretical curve for centric structures. Open circles are data for copper tetraphenylporphine, solid circles for zinc tetraphenylporphine dihydrate. The lattice constants, density, and number of molecules per unit cell for each compound are given in Table I. The densities were determined by the flotation method. The C u T P P lattice constants were measured on the G. E . XRD-5 Eulerian cradle. The powder method, with calibration by spinel ( a = 8.0833 was employed to determine the P d T P P lattice constants. The ZnTPP‘(H20)z and FeTPPOH “ 2 0 lattice constants were obtained from Weissenberg h01 films and precession films calibrated with spinel powder. Determination of S t r u ~ t u r e s : 1 ~CUTPP and PdTPP.-Precession photographs showed that the Laue class is 4/mmm and that the C u T P P and P d T P P have the same space group. The systematic extinctions are hkl: h k 1 = 2n and hhl: 2h-+ 1 = 4n. This allows two possible space groups, I&md and I42d. The I 4 m d space group was tried first, but complete lack of agreement between the calculated and observed structure fackors showed this choice to be incorrect. The final choice of the I42d space group was confirmed by the successful determination of the structure. A scintillation counter was employed to collect all intensity data. An Eulerian cradle and Cu K a radiation were used to collect 465 C u T P P reflections, all nonzero. The intensities were obtained by the stationary crystal-stationary counter method, using balanced filters. A total of 475 nonzero P d T P P reflections were collected by the Weissenberg equi-inclination technique. Two sets of P d T P P data were collected, one with Mo K a radiation, the other with Cu K a radiation and a different crystal. The intensities were obtained by integrating the recorder scan of each peak with a planimeter. Each independent intensity for both C u T P P and P d T P P is the average of two equivalent reflections.l6 Lorentz and polarization factor corrections were applied. The proportionality factor between each set of calculated and observed structure factor amplitudes, was allowed to vary in the least-squares refinement calculations. These proportionality factors were used to put the two sets of P d T P P data on the same scale before averaging them. Absorption corrections were small enough to neglect. The scattering factors were obtained from Table 3.3.1A in the “International Tables for X-Ray Crystallography,” Vol. 111. The space group fixes the metal atom a t x = 0, y = 0, z = 0 on a fourfold inversion axis. The metal atoms were used to calculate a first approximation to the phases of the structure factors in order
+ +
J. V . Smith, private communication. (14) The following computer programs were used for the calculations: A. Zalkin, “Fordap-3”; W. R . Busing, K . 0 Martin, and H A. I.evy, “ORFLS, A Fortran Crystallographic Least-Squares Program” and “A Crystallographic Function and Error Program for the IBM 704” (Fortran revision) The intensity correction program was written by C. Knowles (15) Tables of calculated and observed structure factors of CuTPP, PdTPP, ZnTPP.(HzO)z. and FeTPPOH.Hz0 have been deposited as Document 7869 with the American Documentation Institute, Auxiliary Publication Project, Photoduplication Service, Library of Congress, Washington 25, D . C . A copy may be secured by citing the document number and remitting in advance $6.25 for photoprints or $2.50 for 35 mm. microfilm, payable to Chief, Photoduplication Service, Library of Congress. (13)
I
\
.9
H
Fig. 2.-Atornic numbering system used for tetraphenylphorphines. In the zinc and iron tetraphenylporphines, the oxygen below the metal atom is number 21; the oxygen above the metal atom is number 22. to locate the other atoms. The final cycles of the least-squares refinement include anisotropic temperature factors and the hydrogen atoms. The observed C u T P P structure factors were weighted equally. For P d T P P , unit weights were used first. I n the final stages of refinement, the observed structure factors were weighted according to the system of Hughes16 ( w = l/IFl* for 1 FI greater than 4 I Fl m i n and w = ~ / I B I F (* m i n for 1 FI 5 41 F I m i r , , where w is the weight). Comparison of data from the two P d T P P crystals indicated that the standard deviation of 1 F h k l l was roughly proportional to 1 F I , Lmaking ~ ~ , the Hughes weighting scheme reasonable. The differences between atomic coordinates found using diiTerent weighting schemes were small compared to the errors of the coordinates. The final, conventional R-factor for C u T P P was 0.061, including all reflections. The P d T P P R-factor, not including unobserved reflections, was 0.051. Including all reflections, it was 0.161. Z n T P P ’ (H202)2.-The Laue group is 4/m and the limiting condition on reflections is h k I = 2n. Space groupsconsistent with this are I4 and 14,which are noncentric, and I4/m, which is centric. The space group I4/m was chosen, since the structure could be successfully determined and refined using it. Furthermore, the usual statistical test for centric crystal structures’’ indicates that the space group is centric, as shown in Fig. 1. Data were collected and the structure was determined in the way described above for P d T P P . The space group fixes the zinc atom a t O,O,O. Mo K a radiation was used, and the absorption correction was small enough to neglect. The refinement included 569 nonzero structure factors and the Hughes weighting schemelo was used. The hydrogens and anisotropic temperature factors were included in the final stages of the refinement. The final R factor was 0.080 omitting unobserved reflections and 0.131 including unobserved reflections. FeTPPOH.H20.-This crystal has the same baue group and systematic extinctions as Z n T P P . ( HzO)?. As the results below indicate, the structure refined successfully with the space group 14. The molecularsymmetry of this porphyrin will not fit into the space groups I4 or I 4 / m , allowing us to reject these space groups. The structure factors were calculated from 399 independent nonzero reflections collected with Mo K a radiation as described above for P d T P P . The space group fixes the iron atom, the hydroxide oxygen, and the water oxygen a t 0,0,z on a fourfold axis. The z coordinate for the iron atom was set equal to zero and the positions of the other atoms were then found by using the iron
+ +
(16) E. W. Hughes, J . A m Chem Soc , 6S, 1737 (19-11) (17) E . R . Howells, D C. Phillips, and D Rogers, Acln C r y s l . , S, 210 (1950).
EVERLYB. FLEISCHER, CLETAK. MILLER,AND LAWRENCE E. WEBB
2344
A. Atom Cu(13) N(12) C(1) C(2) C(3) C(4) C(5) C(B) C (7) C(8) C!9) C(10) C(11) H(14) H(15) H(16) H(17) H(18) H(19) H(20)
TABLE I1 FRACTIONAL COORDINATES AND TEMPERATURE FACTORS COPPERTETRAPHENYLPORPHINE C. ZINCTETRAPHENYLPORPHINE DIHYDRATE yo
xa
0 0000 0 , 0 0 0 0 ,1140 ,1971 ,2602 2174 ,1232 ,0573 ,0801 1322 ,1025 127R ,0872 1106 333 234 ,144 ,127 ,159 .04B ,100
,0659 0316 ,1082 ,1782 1565 ,2171 ,3108 ,4828 ,3324 ,4186 ,3761 ,4637 ,096 ,237 ,525 ,290 ,446 350 ,531
sb 0.0000 - ,0032 ,0186 ,0166 ,0064 ,0169 - 0299 - ,0543 - ,1011 - ,1471 - 1722 ,0166 - ,0101 ,051 023 - ,110 ,179 ,244 087 045
-
Bat 26
BiiC 26 14 17 27 32 19 15 26 56 61 52 64 65
14 29 27 25 28 16 14 26 12 38 23 25
Baa 34 32 31 53 50 36 31 27 73 50 52 42 46
BIZ
51s 0 5 5 -11 - 1 0 1 1 - 7 0
0 1 4 0 7 -12 3 1 -20 1 -12 6 1
-
-
-
-
-
8 5 3
Bra 0
-8 -4 -3 6 -6 6
7 18 -6 12 -4
2
-
13 ranged from 0.00044 to 0.00101 and averaged 0.00071. The standard deviations were calculated by computer programs listed in ref. 14. Standard deviations ranged from 0.00068 to 0.00109 and averaged 0.00083. ' The anisotropic temperature factors are dek*B22 PB33 2hkB~ fined by the expression, M = h2Bll 2hlBI:, 2klB23, where e-.M is the factor by which thermal vibration reduces the atomic scattering factor for the Bragg reflection, hkl. The values given are the experimental values multiplied by IO4. The range of the standard deviations for anisotropic temperature factors of these compounds is 0.5 X t o 15 X lo-' with the majority between 0.0005 and 0.0010. The isotropic temperature factor for the hydrogen atoms of all compounds was taken as 3.0.
'
+
+
B.
C(3) C(4)
C(5) C(6) C(7) C(8) C(9) C(10) C(11) H(14) H(15) H(l6) H(17) H(18) H(19) H(20)
x"
0.0000 1151 1974 ,2613 ,2179 ,1266 ,0562 0819 ,1323 10% ,1261 ,0864 ,1101 ,327 ,235 ,166 ,116 I55 ,074 ,103
Atom
zb 0.0000 0.0000 ,0610 ,0022 ,0328 ,0159 ,1037 ,0161 1794 - , 0 0 3 7 1565 - ,0172 2176 - , 0 2 7 3 ,0535 3107 4830 1030 3323 - 1464 ,4189 - 1696 ,3780 ,0128 ,4613 - , 0 0 7 3 ,089 048 238 ,043 ,533 - ,129 ,282 - ,178 446 - 238 ,382 089 ,532 039
-
-
sb
Bii
Bzz
Bsa
BIZ
0 0000 0000 0000
32 35 37 43 42 33 39 34 35 63 52
32 191 44 88 40 130 43 131 36 165 45 99 35 157 36 102 55 251 96 156 98 244
0
-
0000 0000 0000
0000 0000 0000 1206 1188 000 000 000 220 220 2519
284
284
Bia
Bta
3 2 3 - 3 - 5 - 6 -30 -28
0 0 0 0 0 0 0 0 0 -25 -14
0 0 0 0 0 0 0 0 0 -15 -27
0
0
0
4 - 5
-
301
Standard deviations ranged from 0.00049 to 0.00079 with a n average of 0.00062. Standard deviations were 0.00132 and 0.00152 with an average of 0.00142.
+
+
+
D.
~
i
27 35 25 22
30 42 36 39 45 48 64
60 70
i
Bn
27 35 34
39 40 38 22 23 55
23 44 46 28
at 58 55 92 87 93 70 48 70 102 70 74 68 92
~
i
0 3 7 -14 -11 - 5
8 -11 -34 6 -14 -11 0
z
Bia
Bta
0 -14 - 3 7 -21
0 0 16 -3 20 1 -2
-
5 4 1
- 8
-
9 6 5 3
8 23 -3 6 3 -3
FERRIC HYDROXIDE TETRAPHENYLPORPHINE MONOHYDRATE
Atom
X,J
Ya
zb
0 000 1444 2244 3119
0 000 0376 - 0227 0306 1277 1322 2193 3154 4896 3593 4505 3588 4454 0000 0000
0 000 - 0211 - 0242 - 0323 - 0337 - 0294 - 0210 - 0254 0216 - 1531 - 1505 0938 0890 - 3005 2219
2889 1825 1291 1834 2904 2116 2644 2085 2618 0000 0000
PALLADIUM TETRAPHESYLPORPHINE Yo
YY
0 0000 - 0356 - 1292 - 1239 0290 0292 1,340 1911 2986 2170 2885 - 200 - 020 350 240 290 000
xa
Zn(13) 0,0000 N(12) ,1477 C(1) ,1881 C(2) 2936 C(3) ,3187 C(4) ,2273 C(5) ,2185 C(6) ,3143 C(7) ,4921 C(10) ,3586 C(11) ,4508 H(14) ,350 H(15) ,390 H(16) 570 H(19) ,400 H(20) ,500 O(22) ,000 (I
' Estimated standard deviations for atoms 1 through
Atom Pd(13) N(12) C(1) C(2)
Vol. 86
-
BC 3 3 3 4 4 4 3 3 4 5 5 5 5 4 4
9 3 4 2 4 0 6 9 5 2 6 5 2 4 0
a Standard deviations of the atoms in the porphine ring (atoms 1-5, 12, and 13) ranged from 0.0013 to 0.0017 with a n average of 0.0016. Standard deviations of the atoms in the phenyl rings (atoms 6-11) ranged from 0.0016 to 0.0044 with a n average of 0.0026. Standard deviations of atoms 1-5, 12, and 13 ranged from 0.0039 to 0.0044 with a n average of 0.0042. Standard deviations of atoms 6-11 ranged from 0.0035 to 0.0049 with a n average of 0.0042. B is the individual isotropic temperature factor.
'
a Standard deviations ranged from 0.00059 to 0.00195 with a n average of 0.00081. /I Standard deviations ranged from 0.00090 to 0.00160 with an average of 0.00117,
atom to determine the phasesof the structure factors. The threedimensional F,-F, synthesis indicates large anisotropic temperature factors, especially for the iron atom. Anisotropic refinement was attempted but was unsuccessful probably because the anisotropic vibration is so great and the atoms do not vibrate independently. Since it was not possible to locate the disordered hydrogens on the oxygens, the hydroxide was not differentiated from the water. The data were weighted according to the weighting system of Hughes,lBand hydrogens were not included in the refinement. Final R-factors were 0.17 for nonzero reflections and 0.29 including the unobserved reflections.
Discussion of Results The atomic numbering system is given in Fig. 2. The final fractional coordinates and temperature factors are shown in Table 11. As expected, the anisotropic temperature factors become larger as distance from the
metal atom increases. The high degree of anisotropy supports the hypothesis of high deformability normal to the mean plane of the porphyrin ring, as pointed out by Hoard.5 The attempted anisotropic refinements of the FeTPPOH . H20 structure indicated that the iron atom is held more loosely than the metal atoms in the other porphyrins. PI1 for the iron is about 0.()03, whereas /333is about 0.048, which is much larger than any temperature factors found for other porphyrin metal atoms. The Fourier electron density maps indicate that the iron atom is definitely held on one side of the ring even though 0 3 3 is large. The bond lengths and bond angles for CuTPP, PdTPP, ZnTPP. (H20)2, FeTPPOH.H*O, and nickel etioporphyrin I are shown in Tables I11 and IV.
STRUCTURES OF METALTETRAPHENYLPORPHINES
June 20, 1964
TABLE I11 INTRAMOLECDLAR BONDLENGTHS IN --CuTPP--Length
--XI Et10 I--au Length
1 957 1 396 1 396 1 427 1 335 1 427 1 398 1 398
0 013 013 013 016 023
016 ,013 013
a
0 007 012 012 013 014 014 013 013 014 015 016 015 019 020 015
1 981 1 383 1 386 1 461 1 337 1 436 1 380 1 358 1 486 1 392 1 393 1 415 1 346 1 387 1 395 1 00 1 19 0 68 1 22 1 28 0 86 1 19
b
--PdTPP--Length
2 009 1 369 1 377 1 440 1 346 1 432 1 413 1 386 1 502 1 376 1 377 1 391 1 345 1 419 1 336 1 11 1 13 0 99 0 87 1 13 1 15 1 26
U
015 020 019 018 021 026 019
a S$andard deviations as calculated by computer programs listed in ref. 14. 0.12 A.
TABLE IV BONDANGLESIN DEGREES -pdTPP-
-CUTPP--
Anglea
Angle
ob
Angle
ZnTPP.(HIOh FeTPPOH.Hz0 u
Angle
u
Angle
a
1,12813 125 7 0 6 126 3 126 128 4.12,13 126 0 6 127 2 0 7 127 2 6 126 7 1 5 1,12,4 107 8 7 106 4 1 0 105 9 7 104 9 1 8 7 111.7 1 . 9 2,1,12 107.6 .8 109.1 1 . 0 1 1 0 . 5 7 124.2 2 . 1 .8 125.8 1 . 0 125.5 5,1,12 127.3 124.1 2 . 2 1 2 4 . 8 .8 124 9 1 2 1 2 4 . 0 . 8 2,1,5 . 7 108.0 2 . 1 1,2,3 107.5 . 9 107.7 1 . 0 107.1 .7 105.7 2 . 0 . 9 106.9 1 . 0 106.4 2,3,4 108.6 .7 109.6 2 . 0 3,4,12 1 0 8 . 1 .8 1 0 9 . 8 1 0 1 1 0 . 2 126 1 2 . 1 5,4,12 1 2 7 . 3 .8 1 2 4 . 0 1 . 1 1 2 4 . 4 . 7 123.9 2 . 1 .7 124.8 .8 1 2 5 . 3 1 . 1 124.4 3,4,5 124.1 2 . 1 1 2 3 . 1 .8 1 2 5 . 0 1 . 1 1 2 5 . 9 . 8 1,5,4 . 8 1 1 8 . 9 1 . 1 1 1 8 . 1 . 7 116 7 2 . 0 1,5,6 119 8 . 9 116.1 1 . 1 116.0 . 7 118.8 2 . 0 117.0 4,5,6 . 5 121 0 3 6 122 0 . 9 120 7 1 . 2 1 2 0 . 5 5,6,8 . 5 118.4 3 . 5 5,6.10 119.1 1 0 122.6 1 . 3 120.5 120.5 2 2 .5 8,6,10 119 0 1 0 1 1 6 . 7 1 1 1 1 9 . 0 .5 1201 2.1 9,7,11 121.2 1 . 2 1 1 8 1 1 . 2 1 1 9 0 6,8,9 119 4 1 . 1 1 2 0 . 1 1 . 3 1 1 9 . 8 1 . 1 1 1 8 . 3 3 8 7,9,8 120.5 1 . 2 122.1 1 5 121.2 1 . 2 117.7 3 . 7 6,10,11 1 2 1 . 3 1 . 0 124 5 1 . 7 1 1 9 . 8 1 1 118 1 3 . 4 7,11,10 1 1 8 . 6 1 2 1 1 8 . 6 1 . 7 121 2 1 . 2 1 2 5 . 2 3 . 6 Estimated standard a Middle number is vertex of angle. deviations calculated by computer programs listed in ref. 14.
’
A given bond distance, including the M-N distance, is does not change significantly as the changed, even though the substituents are different. Since Cs-cS is nearly a single bond, conjugation of the benzene rings and the porphine nucleus must be small, These facts demonstrate the small effect of substituents on the bond distances of the porphine nucleus in these compounds. The rather large number of porphyrins found to have nearly the Same bond lengths allows one to obtain a reliable estimate of bond orders. Bond orders calcu-
b.
-2nTPP Length
0 009 013 014 014 016 016 016 02
2345
U
-FeOHTPr Length
0 007 010 011 012 012 011 012 012 011 012 012 010 014 014 010
2 030 1 356 1 383 1 409 1 352 1 442 1 386 1 442 1 494 1 439 € 352 1 426 1 415 1 299 1 376
(Hz0)-
2 042 1 369 1 379 1 421 1 374 1 429 1 414 1 403 1 497 1 359 1 360 1 419 1 344 1 344 1 419 1 27 0 96 1 25 1 15 1 21 1 15 1 21 2 447 2 447 Average
’
€1~0)U
0 019 03 029 03 032 03 1 033 d32 033 055 053 057 063
Av. of all bond lengths
1.997 1.379 1.384 1.437 1.348 1.431 1.401 1 386 1.495
058 061
b
0 030 2.950 0.050 0 030 2,179 0 048 standard deviation for C-H b o n d s is about
lated from bond distances averaged over the Ni, Cu, Pd, and Zn porphyrins are given in Table V The table shows that bond orders of porphyrins calculated by theoretical methods’* l9 vary widely in their agreement with the bond orders calculatedJ0from our experimental bond lengths TABLE V AVERAGEBONDDISTASCESWITH THEORETICAL AND EXPERIMENTAL BONDORDERS Av. distance, Bond
Exptl bond orderb
A0
Double LCAO-MO bond Huckel characterC modeld
Four orbital modele
N-Ci CrC2
1.38 0.3 0.25 0.58 0.22 .25 .47 .61 1.44 ,23 cz-ca 1.35 .90 .75 .78 .62 c4-cs 1.40 .49 .50 58 .62 a Average includes the experimental bond distances of nickel etioporphyrin I , copper tetraphenylporphine, palladium tetraCalphenylporphine, and zinc tetraphenylporphine dihydrate. culated by a Pauling-type formula used by Cruickshank in ref. 20 and based on ethylene, benzene, and graphite bond distances and bond orders. An approximate correction for the srn,dler size of the nitrogen was made. Pauling method; bond order given by proportion of the Kekule structures in which a particular bond is double. d .Application of simple Huckel model to porphine by Longuet-Higgins, Rector, and Platt in ref. 18. e Calculated by Gouterman and Wagniere in ref. 19.
’
The nonplanarity of CuTPP6 and the isostructural tetragonal form of TPPj are described elsewhere. The TPP is as nonplanar as CuTPPi even though TPP has no atom. PdTPP has the Same structure within error. The benzene and pyrrole rings are flat accuracy. The Of the between the plane of atoms 61 and benzene ring and the horizontal 001 plane is 80.0 + 0.6’ (18) H. C 1.onguet-Higgins, C . w Rector, and J R P l a t t , J Chem si
“g9;‘he;. ~w (20) D.
~
~
t
h
~
and G. Wagni6re, , S p r r i r y , 11, 108 (1963) J Cruickshank, Tetvoheiiroii,17, 1.5.5 (lnez)
~
~
EVERLYB. FLEISCHER, CLETAK. MILLER,vi^ L.IWRENCE E. LVEBB
2346 -b
I
3 -
3 4
4
0
I
0
7 Fig. 8.-Projection of (me unit of C u T P P o n 001. Fractional z-coordinates of the metal atoms are shown. The C u T P P tnolecule a t the center of the cell itlashed line) is bent toward negative t in the first and third quadrants, and t o r a r d positive z in the second and fourth quadrants. This correlates with the fact t h a t a benzene ring from a neighboring molecule is directll- above the first and third quadrants and directll- below the second and fourth quadrants.
tween the plane of atoms 6, 8 , and LO of the benzene ring and the horizontal 001 plane is 80 =t 2'. In F e T P P O H , H 2 0 ,the iron is in a field which is very asymmetric in the direction perpendicular to the plane of the four nitrogens, since the hydroxide oxygen carries much more negative charge than the oxygen in water. This asymmetry is shown by the fact that one of the oxygens, probably in the hydroxide, is 2.1s A. from the iron while the other oxygen is 2.95 A. away from the iron. I t is possible that the asymmetry of charge changes the orbital geometry of the d2sp3hybridization of the iron a t o m llrith this change in geometry, the d-orbitals of the iron may ox-erlap better with the nitrogen orbitals when the iron is about 0.25 Ai.above the nitrogen. The extent of nonplanarity of the porphyrins discussed above can be illustrated by the distances of the atoms from the 001 plane, which passes through the metal atom. These distances, which are listed for several porphyrins in Table VI. also demonstrate the differences in the planarity of the various molecules. TABLE \-I DEVIATIOSS O F Atom
AI
in both C u T P P and P d T P P . -4 model shows t h a t this bending is caused by the interaction of benzene hydrogen, H(lli), with pyrrole hydrogen, H ( l 5 ) , and the interaction of benzene hydrogen, H(19), with the pyrrole hydrogen equivalent to H(14). The plane of atoms 12, 1, and 4 of the pyrrole ring makes an angle of 13.4 f 0,4' with the 001 plane in CuTPP. The corresponding angle in P d T P P is 12.3 =t 0.6'. Tulinsky is now determining the structure of the triclinic form of TPP, which, interestingly, is much more planar than the tetragonal form of TPP.?' The two hydrogens a t the molecular center are bonded to two pyrrole rings, which are not equivalent to the other two pyrrole I t should be mentioned that in the tetragonal form of TPP, the pyrroles may also be nonequivalent. The molecules may, by random distribution, attain a statistical fourfold symmetry not inherent in the individual molecule. By way of contrast to the other metal porphyrins, the porphine nucleus of Z n T P P ' ( H 2 0 ) 2is in a mirror plane, within experimental error. The phenyl groups are perpendicular to this mirror plane. It is interesting to note that the zinc forms an octahedral rather than a tetrahedral complex, apparently because, a t least in this crystal form, tetrahedral geometry would distorto the ring too much. The Zn-H20 distance of 2.45 A. is rather long, indicating that the water molecules are not bound tightly to the zinc. F e T P P O H . H 2 0is nearly planar except that the iron atom is 0 . 2 K. abox-e the plane formed by the four nibroKen atoms, even though the F e - S distance of 2.03 A. is not significantly longer than metal to nitrogen distances in other porphyrins. This phenomenon has been observed by Kendrew" in myoglobin. The angle be1 2 1 ) .% Tulinsky. piivate communication i 2 1 a l SOTI; A U U E I i) s P R O O F - F ~ >reiults I. r r f this w i i k see il Tulinsky. ~ ' I I P I I I Soi 86, $127 '1964) J C K e n d r e w .Sctt,nct,, 139, 1250 ~
