Chemical Foundations for the Understanding of ... - ACS Publications

3 Chemical Foundations for the Understanding

Downloaded by UCSF LIB CKM RSCS MGMT on December 3, 2014 | http://pubs.acs.org Publication Date: June 1, 1971 | doi: 10.1021/ba-1971-0100.ch003

of Natural Macrocyclic Complexes D. H. BUSCH, K. FARMERY, V. GOEDKEN, V. KATOVIC, A. C. MELNYK, C. R. SPERATI, and N. TOKEL Ohio State University, Columbus, Ohio 43210

Studies on template effects, condensation reactions, oxida tive dehydrogenations, catalytic hydrogenations and ligand deprotonations have made it possible to synthesize almost any tetradentate macrocyclic ligand one might choose. The principle of multiple juxtapositional fixedness is ex plained and applied to metal binding in enzymes. The inflexibility of the metal ion site in the porphyrin ring is contrasted to those in synthetic macrocycles. TAAB, a related tetraazaannulene, can vary its inner site while other rings fold when the metal ion is too large. Many model compounds for high-spin, five-coordinate deoxyhemoglobin and deoxymyoglobin are presented. The Fe -O-Fe bridge is extremely hard to break when iron is bound to a tetraazaannulene (TAAB). Iron is a better catalyst for oxi dative dehydrogenation of its own ligand than is nickel. III

III

' T p h e sustained development of the new synthetic macrocycle chemistry -•• is traceable mainly to the efforts of those i n two laboratories. N e i l Curtis first discovered ( J ) that monocarbonyl compounds condense with diamines to produce cyclic compounds. Table I represents the products of this general category of reaction for such diamines as ethylenediamine, propylenediamine, and trimethylenediamine. The structure of the carbonyl compound must contain an α-methyl group. Simultaneously, macrocyclic complexes were intentionally produced at The Ohio State Univer sity i n a program then dedicated to the demonstration of the coordination template effects (2, 3). O u r first successful cyclization of this class is shown i n Figure 1. Though the rapid progress of this field is largely the result of extensive studies by the two groups already mentioned, the excellent contributions of several others must be mentioned even i n so 44 In Bioinorganic Chemistry; Dessy, R., et al.; Advances in Chemistry; American Chemical Society: Washington, DC, 1971.

B U S C H

Natural Macrocyclic Complexes

E T A L .

Table I.

Macrocycles Prepared from Diamines and Carbonyl Compounds R

CH

f

R

2

? M ^ (CH,) / \

R

4

H

X

? CHR

(

H


B

H

Amine

Acetone Acetone Acetone Acetone Acetone Methylethyl ketone Acetone Acetone

l

2

)

n 4

M

/

R

Carbonyl Compound

/

2

V

Metal

R '

R

SR*

1

R

R*

s

en pn en pn en

Ni Ni Cu Cu Co

Me Me Me Me Me

Me Me Me Me Me

Me Me Me Me Me

H Me H Me H

en tm tm

Cu Ni Cu

Et Me Me

Et Me Me

Me Me Me

H H H

CH2CH2

I

N i

C \ CH

3

CH

/

\

\

:

BrCH

2

BrCH

2

X

/

CH2 CH2 Ni(PE)

CH0CH2

/ C

I

Î

I CH

3

\ /

3

/ CH

S

C

2 Br

S C H \ / CH2CH2 X

2

H

Ni

Ni(PEX)Br

2

2 2

2

Figure 1. Preparation of s s'-o-xylyl-2 3-pentanedionebis(mercaptoethylimine) nickel(II) bromide 9

9

In Bioinorganic Chemistry; Dessy, R., et al.; Advances in Chemistry; American Chemical Society: Washington, DC, 1971.


46

BIOINORGANIC

C H E M I S T R Y

cursory a consideration of the subject as is intended here. In 1962, both Schrauzer (4) and Thierig and U m l a n d (5) reported cyclization reac tions involving a borate ester and similar linkages. Also, the synthesis of macrocyclic ligands differing only slightly from natural rings was accomplished early by Johnson and his associates (6). A n extensive summary on the synthesis of macrocyclic complexes w i l l soon be avail able (7). It is particularly true in the case of tetradentate macrocycles that the synthetic techniques are rapidly reaching a stage of maturity such that one can set out to synthesize almost any previously unknown macrocyclic ligand of specific structure with a reasonable expectation of suc cess. The reactions of mono- and dicarbonyl compounds with pairs of adjacent amine groups in the presence of metal ions has provided much of this generality. Figure 2 shows the previously mentioned Curtis reac tion. It also shows the reaction of an α-diketone to form the a-diimine chelate ring of a macrocycle. Though the reaction has been known, its successful application to macrocyclization is owing to Baldwin and Rose (8). Jager has made extensive contributions to synthesis of macrocyclic ligands i n their complexes. His procedures first incorporated the ^-diketone moiety in such structures and he first showed that the expected

Figure 2.

Schiff base condensation in macrocyclization reactions



3.

B U S C H

E T

A L .

Figure 3.


47

Schiff base condensation in macrocyclization reactions

ionization of a proton occurs in such reactions (9), so that the presence of each such grouping in a structure is accompanied by a unit of negative charge. Figure 3 shows a number of variations on the use of the Schiff base condensation in macrocyclization reactions. One of the main products of self condensation of o-aminobenzaldehyde in the presence of a metal ion is the tetraanhydrotetramer (10). The complexes of this ligand (abbreviated T A A B ) are discussed below. The first example of such reactions using reagents incorporating noncondensing amines as parts of their structures was developed by Curry. H e produced a variety of macrocycles starting with 2,6-diacetylpyridine (11). A similar tetrafunctional reagent was developed by Tasker and Green (12), as shown in the last reaction of this figure.


48

BIOINORGANIC

C H E M I S T R Y


In addition to an almost unlimited array of condensations that might produce rings of the sorts described above, there exist systematic transformations that permit one to generate a variety of products b y reactions of a single condensation product. Curtis ( J ) first produced a hydrogenation-dehydrogenation sequence, as shown i n Figure 4. It is generally true that procedures can be found for the reduction of azomethine linkages i n complexes containing macrocyclic ligands; however, the oxidative cis isomers

Figure 4.

trans series

Hydrogenation-dehydrogenation sequences in macrocycles


3.

B U S C H

E T

A L .


49


α and β meso and racemic

Figure 5.

Hydrogenation-dehydrogenation sequence for Ni(II) complexes

dehydrogenation exhibits metal ion-specific effects. In general, the group — C H R N H — can be oxidized to — C R = N — i n nickel complexes. In deed, Curtis' work involved nickel (II), and BarefiekTs extensions (13) have shown that the nickel complexes are much more susceptible to such reactions than are the corresponding cobalt derivatives (Figure 5 ) . In fact, nickel (III) seems to play a significant role i n the mechanisms of such reactions. Goedken has recently found that iron is still more effec tive at promoting oxidative dehydrogenation of its cyclic ligand (14). W e shall return to the latter point shortly. A third class of reaction derives from the acidity of the coordinated ligands. A variety of the structural members may evidence acidity i n complexes. Figure 6 provides a useful example from the studies of Goedken (13). The methylene group that ionizes i n this ligand does so in fair measure because of the substantial electron derealization i n the resulting anion. One may couple deprotonations and the concomitant


50

BIOINORGANIC

C H E M I S T R Y


tautomerizations to oxidations to provide still more possibilities i n ligand synthesis. From studies such as these, it is safe to conclude that one is no longer exploring the possibilities but that one can now define synthetic goals and expect to achieve them. I n this regard, there are a few struc tures that must soon become as familiar as those of ethylenediamine and of E D T A to all who are concerned with coordination chemistry (Figure 7). Principal among these are the uncluttered inner great rings of the

(C10 ) 4

ciOr

/

Figure 6.

Όeprotonation

X N

i

2
H 0

NH CH CH NHCH CH NHCH CH NH 2

2

2

2

2

2

2

2

N-—-\

2

cyclam

\

Ν

Ν

'


(4)

3.

BUSCH E T A L .

53


Table II.

Rates of Formation of Cu * Complexes 2

Cu

a q

+ L

2 +

• CuL + 2

fc (0.5

L

k (pH =

0H~)

2

4.7)

2


8.9 Χ 10

4

5.8 Χ ΙΟ"

hematoporphyrin I X

Table III.

2.0 Χ

10"

2

2

Rates of Dissociation of Cu * Complexes 2

C u L + + nH+

> Cu

2

aq

2

+ +

L

LH

w

n +

kiisec" ) 1

The famihar octahedral complexes N i ( N H ) * and N i ( e n ) * (en — N H 2 C H 2 C H 2 N H 2 ) are destroyed in fractions of seconds i n strongly acidic solutions, showing the existence of a facile mechanism for ligand dis sociation. If all four donor nitrogens are made part of one ligand, trien, N H C H C H N H C H C H N H C H C H N H , the rate is diminished, but only just noticeably. In any event, most nitrogen ligands are stripped from nickel (II) in strong acid media in a very short time. In contrast, 3

2

2

2

2

2

2

2

6

2

3

2

2


54

BIOINORGANIC CHEMISTRY

when the nickel (II) complex of a cyclic ligand, N i ( cyclam ) , is placed i n strong acid (6N HC1) nothing happens. That is, the rate of loss of the ligand is unbelievably slow. Cabbiness and Margerum (15) have measured rates of formation and rates of dissociation of copper(II) complexes of linear tetradentate and closely related macrocyclic complexes. Table II shows that the formation rates are retarded by a few orders of magnitude for macrocyclic complexes; however, the greatest effect occurs for the dissociation rate (Table III). The copper complex of the closely related linear tetradentate ligand dissociates some 10 times faster than does the macrocyclic complex. The huge effect on the dissociation rate is readily illustrated for the case of the nickel (II) complexes. The usual mechanism of substitution at nickel (II) involves bond breaking i n the rate-determining step. As Figure 8 shows, this occurs at an end group i n a polydentate derivative. In acidic media, the dissociated groups are protonated quickly and the vacated site in the coordination sphere of the nickel atom is filled rapidly by solvent. A second atom then dissociates and the entire ligand is replaced by solvent in a sequence of S 1 steps. F r o m the structure of a ring, a simple dissociative step cannot occur because the ring has no end. It is not possible to extend the metal-nitrogen distance sufficiently to constitute bond rupture without additional bond rupture involving the ligand or extensive rearrangement within the co2+


7

N

(i)

(2)

Figure 8.

First step of removal of a linear polyamine from Ni in acidic media 2+



3.

BUSCH E T A L .


55

X Figure 9.

Model for dissociation of macrocyclic complexes

ordination sphere. Figure 9 suggests the general features of displacement of a macrocycle from nickel (II). These processes appear to involve a nucleophile, and therefore their rate processes may resemble bimolecular reactions. The scheme given reflects what is probably the other principal distinguishing feature of such reactions for flexible macrocycles. The ring almost certainly folds before the first bond between itself and the nickel atom is broken. It is relatively simple to move a donor atom away from the metal ion when the ring is in a folded configuration. Although no detailed model was suggested, Cabbiness and Margerum (15) have referred to the sluggishness of these processes and the attend-



56


ant stabilization of the macrocyclic complexes as the "macrocyclic effect/' It is our conviction that the phenomenon is well illustrated by, but not unique to, macrocyclic ligands. It is a simple matter to illustrate the fact that the metal ion need not be completely encircled by a ring in order to display this phenomenon. The tridentate cyclic ligand trisanhydro-o-aminobenzaldehyde ( T R I , Figure 10) coordinates to a single face of an octahedron. The metal ion is not surrounded by the ligand, but wears it like a crown (16,17). Despite this fact, the complex is extremely inert toward substitution, as witnessed by resolution of salts of the cation Ni(TRI)(H 0) into optical isomers (18). Further, these optical isomers show no signs of racemizing, being optically stable in solution for weeks. It is suggested that the inertness occurs when some minimum number of donor atoms, possibly three, is so arrayed that the usual mechanism for stepwise dissociation of the ligand is thwarted. That is, extensive rearrangements must occur, or an entirely different path must be followed for ligand removal. Since the phenomenon arises mainly because of the fixed geometric placement of a set of ligating functions, it is called multiple juxtapositional fixedness. 2

3

2 +

Figure 10. Tridentate ligand, trisanhydro-o-aminobenzaldehyde (TRI) Since the constraints on mechanism act more severely with regard to dissociation than formation, M J F has strong thermodynamic consequences—the stability constants for macrocyclic and other MJF-affected complexes should be abnormally high. M J F is probably of considerable importance in metal binding to enzymes, for the locations of the donor atoms may be so fixed in space that they cannot be removed from a metal ion in a stepwise fashion without greatly altering the protein conformation; i.e., without denaturing the enzyme. There is a further consequence of this phenomenon that must be considered. Vallee and Williams (19) have suggested an entatic state



3.

BUSCH E T A L .

57


for metal ions in active sites on enzymes. The state of binding of the metal ion should be such that its energy level contributes at par (or better) to an energetically poised domain. For enzymatic catalysis of nucleophilic processes, this could be accomplished most obviously by binding the metal ion in an extremely electron-poor condition. This suggests weak bonds (poor donor groups or usual donor groups and long bond distances). This seems acceptable but there is a problem. The binding of metal ions to enzymes is not notably weaker than that to simple ligands having comparable ligating groups. In view of the frugality of nature, it is difficult to believe that she has not chosen to use the entatic principle, but it appears that she would have to be doing so without the requisite weakening in the binding of the metal ion to an enzyme. A simple answer can be suggested by invoking M J F and entasis as opposing factors as regards the binding of metal ions. The extra stabilization attributed to M J F may serve to offset the weak bonds required by entasis (Table I V ) . Table IV.

MJF

Possible Role of Multiple Juxtapositional Fixedness in Enzyme Binding of Metal Ions ia) Greatly enhanced binding) »J (b) Normal electron density J

(a) Greatly weakened binding) Bond strain—>< \ ^ (b) Reduced electron density ) Effects of Ring Size and

Strong binding and Great electrophilicity

Flexibility

Though the porphyrin ring does ruffle a bit around the edges, the size of the site that accepts the metal ion remains very nearly constant for a wide range of compounds (Table V ) (20, 21). This inflexibility is associated with the aromatic character of the ring. It is useful to compare the porphyrin ring with a related macrocycle that was introduced above, T A A B , Figure 11. This ligand has a metal ion site that is contained in a sixteen-membered ring that constitutes an alternating nonaromatic structure. It is, in fact, tetrabenzotetraaza-16-annulene, and though the number of members in the inner ring is the same as in porphyrin, a very different set of properties is associated with its structure. Although the metal ion and its donors occupy a single plane, the crystal structure determination reveals a strongly nonplanar arrangement for the rest of the structure (Figure 12) (22). The relative flexibility of this structure can be seen from the data of Table V I . The first two entries


58



Table V .

Aetioporphyrin Deuteroporphyrin Haematoporphyrin Protoporphyrin Mesoporphyrin Coporphyrin Uroporphyrin

CH CH CH CH CH CH C H · COOH 3

3

3

6

3

3

6

3

6

a

2

CH

C2H5

H CH(OH)CH C H = CH2

3

C2H5

(CH ) · COOH (CH ) · COOH 2

2

2

2

Porphyrin

CH CH CH

CH

3

8 3 3 3

CH C H · COOH 3

2

Taken from "Thorpe's Dictionary of Applied Chemistry," Vol. X, Longmans, Green Co., London, 1950. Naturally-occuring porphyrins.

6

=

Figure 11.

TAAB

Tetrameric ligand, TAAB

represent examples of high-spin and low-spin nickel (II), both i n the center of the T A A B macrocycle. The high-spin nickel ion is 0.16Â, or some 8 % , larger than the low-spin nickel ion. In contrast, i n the porphyrin complex, the F e - N distance of high-spin iron (III) is only some 0.08Â greater than that of low-spin iron ( I I I ) ; further, this increase comes about mainly by expulsion of the high-spin atom about 0.5Â out of the plane of the four porphyrin nitrogens. Indeed, the distance between the center of the porphyrin and the center of one of its nitrogen atoms changes no more than 0.02Â. These differences should be kept i n mind m


3.

BUSCH E T A L .


Compounds

0

4

5

C2H5

CH3

H CH(OH)CH CH=CH C H (CH ) .COOH (CH ) .COOH 3

2

2

2

2

2

2

59


6

Figure 12.

6

CH CH CH CH CH CH .COOH 3

3

3

3

3

2

C2H5

(CH ) (CH ) (CH ) (CH ) (CH ) (CH ) 2

2

2

2

2

2

s

2

2

2

2

2

.COOH «COOH -COOH .COOH .COOH · COOH

7

8

C2H5

(CH ) (CH ) (CH ) (CH ) (CH ) (CH ) 2

2

2

2

2

2

2

2

2

2

2

2

Perspective view of TAAB in NifTAAB)^

.COOH · COOH · COOH · COOH · COOH · COOH

• HO t


CH3

C C C C C C

H H H H H H

3 3 3 3 3 3

60


Table VI.

Metal—Nitrogen Distances in Tetradentate Macrocyclic Complexes L

Ni + Ni +

TAAB TAAB Porphyrin Porphyrin Porphyrin Porphyrin

2

2

N

i

2 +

Pd + Fe + Fe + 2

3


3

«Réf.

T> - Ν M

1.90 2.06 1.96 2.00 2.07 1.99

Comments low spin, X = B F high spin, X = I low spin •low spin high spin, a — 0.5A low spin (bisimidazole) a = 0 4

M

for a section to follow on the oxidizability of porphyrins and the probable natures of the oxidation products. As has been indicated above, the twoelectron oxidation product of porphyrin should have much i n common with T A A B . Those macrocycles that are quite flexible readily fold when presented with metal ions that are too large for them to encompass. As Figure 13 shows, the critical ring sizes occur at different values for different sets of donor atoms. In the case of four nitrogen donors in a fully saturated ring, 13 members w i l l surround a first row transition element ion while a 12-membered ring must fold (24). However, with the larger sulfur donors, a 14-membered ring is required to encircle the metal ion (25). Some years ago, Brubaker (26) suggested a possible mechanical constrictive effect on metal ions because of the tight fit of an almost too large metal ion. The data of Table V I I provide some support for this view. The D q value indicating coordination within one plane of all the nitrogen donors of two molecules of dimethylethylenediamine is 1185 cm" . This may be considered a reasonable value for four such donors. However, 1,7-CTH, a closely related tetradentate macrocycle (30), also binds its four nitrogens in a single plane about the metal ion, and this ligand has been assigned a D q value of 1426 c m ' . If the values are essentially correct, the ring is exerting a much stronger ligand field strength than is anticipated. This may be the result of mechanical restriction which would shorten the N i - N distance, for D q is an inverse function of a large power of that distance. 1

1


3.

BUSCH E T A L .

61


N

Ν

v

Ni + 2

Ν

/

\

Ν

Square planar

I

13 members or more


i n

r i n

^

g

Folded 12 members or less in ring

Square planar 14 members or more in ring

Folded 1 3

m

e

m

b

e

r

s

o r

less in ring

Figure 13. Relationship of ring size to folding for tetradentate macrocyclic complexes Table VII.

Spectral Parameters for Some iraws-Diacido Complexes, M A X 2 4

Complex 3

4

2

2

α

2

2

4

2

Dficmr )

vi(kK)

Co(NH ) Cl + · Co(en) Br + Ni(Dimen) (N0 )2 Ni(Etu) Cl Ni(m-l,7-CTH)Cl 3

1

6

t

15.90 15.21 7.84 7.80 8.49

n +

1

21.00 21.68 11.85 8.55 14.73

2490 2530 1185 855 1426

Dq'icmr ) 1

1461 1277 383 705 420

* Ref. 27. Ref. 28. Ref. 29. 6

c

The Oxidation-Reduction and TAAB Complexes

Properties of

Porphyrins

Table V I I I summarizes the idealized electron accounting associated with electrochemical reduction of the divalent metal ion complexes of T A A B (Figure 11). F o r each metal ion, a distinct set of one-electron reduction waves occurs (31). In methanol solution, these multiple oneelectron processes are followed b y a very complex poly-electron wave


62


that is associated with hydrogénation of the azomethine functions of the ligand. From Table VIII, it is clear that the number of one-electron waves is such that the system acquires 42 electrons. This amounts to the eight d-electrons needed to fill the four low-lying d-orbitals on the metal ion Table VIII.

Idealized Electron Accounting for Final Reduction Products of M ( T A A B ) n

2 +


Number of Electrons from Complex

M

TAAB

Electrode

Total

9 8 7

32 32 32

1 2 3

42 42 42

u

Cu(TAAB)+ Ni(TAAB) Co(TAAB)0

T A A B " requires d* sq. planar requires 2

34e~ 8e~

Cu + + T A A B Ni + + T A A B -

42e"

Co+ + T A A B -

3

2

2

2

2

Table IX. Products of the Electrode Reaction M ( T A A B ) + ne'-> M ( T A A B ) n

2 +

( 2 w ) +

Analyses C

Η

~ o Calc. Found

54.15 53.53

3.25 3.29

1.94 Calc. Found

58.9 59.2

3.53 3.24

0.68 Calc. Found

71.1 70.4

4.28 4.13

0.99 Calc. Found

58.9 59.6

3.53 3.79

C o ( T A A B ) C H C N 2.84 Calc. Found

70.3 69.7

4.52 4.43

η

Compound

1

Cu(TAAB)PF

1

Ni(TAAB)C10

2

Ni(TAAB)

1

Co(TAAB)C10

2

6

4

4

3

Table X .

Compound Co (TAAB)(C104) Co ( T A A B ) Co(TAAB)CH CN n

3

1

2

Χ

M

9.02 9.04

4.99 4.84

10.23 9.96

9.82 9.96

6.21 6.50

10.2 10.1

Ν

11.9 11.9 9.82 9.96 13.7 14.4

12.5 12.3

-

fi. 21 6.71

-

-

11.5 11.6

Co"(TAAB)

2+

and

E

Elec. Chg.

+0.58 +0.70 +0.72

-1 -1 -1

vs. Ag-AgNO, (0.1M) in C H , C N .


3.

BUSCH E T A L .


63

in D (or related) symmetry and a p i electron system containing 34 electrons. That is, the ligand has been reduced to T A A B " , an aromatic dianion. Most of the electrode products expected to be formed via the oneelectron processes have been synthesized by controlled potential elec trolysis, isolated, and characterized. Table I X shows evidence for their existence as pure compounds. The one-electron product in the copper system is diamagnetic and has been formulated as C u ( T A A B " ) . Both products have been isolated from the nickel system. N i ( T A A B ) C 1 0 contains one unpaired electron and may be either a N i (III) derivative of T A A B " or a N i (II) complex of the radical anion T A A B " . N i ( T A A B ) is formulated as a N i (II) complex of the dianion T A A B . Only the first two reduction products have been isolated from the cobalt system. These are C o ( T A A B ) C 1 0 (essentially diamagnetic) and C o ( T A A B ) C H C N with a moment that is slightly high for one unpaired electron. Table X presents some of the electrochemical data for the cobalt complexes. It is my purpose at the moment to point out only a single relationship. The oxidation waves for the two reduced products correspond well among the two compounds, but they differ greatly from the oxidation and re duction waves of the starting material. This and other information cause us to formulate this behavior as given in Equations 5-8. 4 h

2


I H

2

+

4

2

2 +

4

3

Presumed Process upon Controlled Potential Electrolysis Co (TAAB) n

Co^TAAB)*

2 +

+ e-

electrode -—• C o ^ T A A B ) * reaction

rearrangement

(5)

> Co (TAAB -)+ i n

(6)

2

Probable Oxidation Processes for 1-Electron Reduction Product —e~~ — e~ • [Co (TAAB-)] + » [Co (TAAB)] +

Co (TAAB -)+ m

2

m

2

ni

3

Its Reduction Products Reduction Waves'

1

II

IV

///

Ε

Elec. Chg.

Ε

Elec. Chg.

Ε

Elec. Chg.

-0.87 +0.25 +0.25

+1 -1 -1

-1.23 -1.06 -0.92

+1 +1 -1

1.85 -2.00 -1.78

~+l +1 +1


(7)

64


Probable Reduction Processes for 1-Electron Reduction Product Co (TAAB -)+ i n

+e-

2

> Co (TAAB ) 1 1

2 -

+e> Co^TAAB -)-

(8)

2

Following the addition of one electron to the cobalt atom in C o T A A B , the product rearranges via intramolecular oxidation-reduction to produce the cobalt (III) complex of the reduced ligand, the dianion T A A B " . The electrode processes of this substance are then rationalized as shown in the figure. The structure of the dianion is that of an approximately planar aromatic dianion (Figure 14). It is a direct analog of a porphyrin, differing in the locations and kinds of fused rings but having the same inner great 16-membered ring. The most obvious result of the structural n

2 +


2

Figure 14. Table XI.

( TAAB f~ dianion

Oxidation Potentials for Tetraphenylporphyrin Complexes: Benzonitrile vs. SCE, Cyclic Voltammetry" [M(II)TPP]

Ε —-

[M(II)TPP]

+

Μ

Ει

H Zn Cu [M(II)TPP]

Ε —

[M(III)TPP]+

Μ

μ°

Ni Co Fe

0.0

Ει 1.00 0.52 -0.32

[M(II)TPP] + 2

E

[Lett

2

1.00 0.79 0.99

2

Ε —1.20 1.10 1.33

Ε —

2.88(E0

[M(III)TPP]

2+

Ε —

[M(III)TPP]*+

μι

Ε

μ

5.10

1.00 1.19 1.18

2.71

2

2

• Ref. 32.


Ε

ζ

1.40 1.42 1.50


3.

BUSCH E T A L .

65


The inner great ring of porphyrin

Figure 15.

Tetraaza-16-annulene

Porphyrin and annulene ring systems

differences is the very much greater reducing strength of T A A B " as compared with the porphyrin dianion. 2

To continue this comparison, the most recent work (32) on the electrochemical oxidation of complexes of tetraphenylporphyrin is summarized in Table X L These materials suffer two or three one-electron oxidation steps terminating in the formation of the metal ion complex of the neutral 2-electron oxidation product of the ligand. As shown in Figure 15, the 2-electron oxidation product is closely related to the annulenes; i.e., it contains the same 16-membered nonaromatic alternating heterocyclic hydrocarbon as its great ring as does T A A B . In this state, the ligand should be much more flexible, have a variable-sized metal ion site and, in general, be quite different from the parent dianion. In fact, one should look to the chemistry of T A A B complexes in order to anticipate the chemical behavior of these oxidized products. A chemical property of M ( T A A B ) that deserves special consideration is the addition of nucleophiles to the carbon atom of the azomethine group (33, 34). Equations 9-13 illustrate these addition reactions while the structure of a typical product is given in Figure 16. W +


66


ROH

Ni(TAAB)(BF ) + OR" 4

> N i ( T A A B ) (OR)

2

H 0

(9)

2

2

N i ( T A A B ) ( B F ) + NH4OH 4

2

> Ni(TAAB)(NH ) 2

H 0

(10)

2

2

Ni(TAAB)(BF ) + O H 4

2

> Ni(TAAB)(OH)

(11)

2


dmf Ni(TAAB)(BF ) + O H 4

> N i ( T A A B ) ( N M e ) · χ dmf

2

2

H 0

2

(12)

2

Ni(TAAB)(BF ) + M e N H 4

2

2

> Ni(TAAB)(NMe ) 2

2

(13)

Figure 16. Structure of the product of addition reactions of Ni(TAAB)? complexes +

Very recently the mono-methoxide derivative of 2-electron oxidized zinc tetraphenylporphyrin has been reported (35). The assignment of structures to the compounds that have been identified in peroxidase and catalase systems presents a strong challenge. Admittedly, the ultimate goal of studying such systems must be the understanding of mechanisms, and it may or may not be necessary to understand the chemistry of the prosthetic group i n order to understand the mechanism. From the orientation of the inorganic field, the structure problem would be tackled first since it is the static problem. Mecha nisms are difficult enough even when the structural chemistry is clear. In the brief section to follow, we discuss probable structures for the com pounds formed by the prosthetic group i n peroxidase. N o comment is offered at this time on the mechanism of action of these enzymes. Figure 17 summarizes the compounds derived from peroxidase after Yamazaki and Yokota (36). The iron i n peroxidase is trivalent and its


3.

BUSCH E T A L .

67

Natural Macrocyclic Complexes Compound I (+5)

H 0 2


Ferroperoxidase (+2)\

2

Compound II y (+4)

Peroxidase Peroxidase + (+3)

H 0 2

2

Peroxidase derivatives

interaction with hydrogen peroxide very quickly generates green Compound I, which contains two additional oxidizing equivalents and is written for convenience as ( + 5 ) . From what has been said above of two-electron oxidation of porphyrins, one should immediately conclude that the ring has experienced 2-electron oxidation, and that in Compound I, the least possible transformation is to the annulene-like state. The spectrum of Compound I is distinctive when compared with Compounds II and III; therefore, the state of either the metal ion or the ligand must differ from those in the reference compounds. Another point that requires emphasis at this stage relates to the comparison of spectra with model compounds. The annulene-like material should be only a weak chromophore as compared with the porphyrin. For this reason and in view of the rather large electrode potentials in these systems, the spectra should depend rather strongly on the metal ion. Though the spectra of Wolberg and Manassen do not strongly support this view, one can only wonder how much sharper the spectra of the purified materials might be. The large extinction coefficients and general characteristics of the absorption bands of Compounds I, II, and III (37) are much more consistent with structures within which nucleophiles have added to the most electrophilic sites of the oxidized ligand (33, 34, 35). The general characteristics of the spectra are far more likely to be preserved in going from metal to metal and from substituent to substituent than are the specific features. It is clear from the spectral properties that addition of nucleophiles to the carbon atoms of these coordinated tetraaza-16-annulenes generates new aromatic systems. Consequently, it is useful to suggest that a nucleophile has added to the oxidized ligand in Compound I. The nucleophile could be a proximal function from the protein in the enzyme.


68


It might also be the O H " produced by reduction of H 0 . A scheme is given (Figure 18) which surprisingly thoroughly accounts for the struc tures of all the peroxidase compounds and the reactions that convert them. Again it must be emphasized that no attempt is being made to discuss the mechanism of enzyme action and that these structures are presented as deserving equal consideration to that accorded the earlier unsubstantiated suggestions. In the suggested scheme, Compound I has the solvolyzed structure given. It is converted to Compound II merely by one-electron reduction. Compound II is formulated therefore as an iron (II) derivative having the same isoporphyrin ligand as Compound I.


2

2

R Ο

Figure 18.

Reaction scheme for conversion of peroxidase compounds

This is at variance with the popular view that Compound II is an iron ( I V ) porphyrin (38,39). This view is not supported by electrochemical studies or model compounds, for example tetraphenylporphyrin (Wolberg and Manassen, 32), where it is shown that the ligand, not the iron ( I I I ) , is oxidized. T o be sure, the present proposal also deviates from simple use of the electrochemical results; however, i n this hypothesis, though the oxidizing equivalents reside on the ring, the addition of nucleophiles would modify the potentials considerably. Having worked a great deal with magnetic susceptibilities, the author is aware of the uncertainties


3.

69


BUSCH E T A L .

of these values when determined under other than ideal conditions. Consequently, these data are not considered here. Further, the values reported for Compounds I and II are easily adapted to almost any theory. Certain relationships between Compounds II and III actually provide the main impetus for asking that some thought be given to the structural scheme presented here. Compound III is formed from 0 and ferroperoxidase ( F e P ~ ) . In the scheme presented here, the ligand undergoes simultaneous 2-electron oxidation and nucleophilic addition of the H 0 " ion that is an artifact of 0 reduction. This produces a compound having a structure very close to that of Compound II, differing only in that H 0 " , and not (presumably) O H " , is added to a ligand carbon atom. The total oxidation equivalents of 0 are preserved in the structure without invoking absurd degrees of oxidation of both ligand and metal ion. It is particularly pleasing that Compound III can be prepared from Compound II by addition of H 0 . Since H 0 " is an excellent nucleophile, it would be expected to displace O H " from the isoporphyrin. Compounds II and III react very similarly when subjected to 1-electron reduction. Both go to peroxidase; however, Compound III gives up a mole of H 0 . In the scheme given here, this is an easily understandable over-all reaction. Both compounds are iron (II) isoporphyrin derivatives. The addition of an electron occurs at the ring and the increased electron density leads to immediate dissociation of the added nucleophile. A t this point, the system would contain the iron (II) derivative of the radical monoanion P". This is not stable but reverts (if it ever existed) to F e P ~ , peroxidase. It appears that Compound III simply would have to exist as a metal complex of H 0 " or H 0 if one sticks to schemes based on iron ( I V ) or on iron (III) combined with radical anions. Perhaps this would not be sufficient change in structure to account for the shifts in wave length of the spectral bands in going from Compound II to Compound III. Since the transitions are mainly owing to the ligand, the incorporation of the electron-releasing H 0 " into the ligand structure is consistent with observations. Finally, in the event iron ( I V ) is established as existing in these systems, the probability of nucleophilic ligand addition is high for the oxidized ligand derivative of so highly electrophilic a metal ion as F e ( I V ) . 2


n

2

2

2

2

2

2

2

IH

2

2

2

2

2

2

2

2

Iron Complexes of the Synthetic

Macrocycles

Though some difficulties persist, it has been possible to synthesize a large number of iron complexes with synthetic macrocycles. Brief attention w i l l be given first to the derivatives of T A A B {40). As summarized in Figure 19, o-aminobenzaldehyde condenses in the presence of ferrous chloride to form a T A A B complex. Because of the ease of oxida-


70


FeCl

+ C

2

or

\—CHO

F e C l + Zn(TAAB)ZnCl4 2

H+,EtOH

\NH:

(N ), 6hrs.

/

2

( N ) , 6 hrs. 2

EtOH,H+


Fe(TAAB)FeCl HNO,,AgN0

4

3

MeOH H 0 +

HCIO4

2

Fe (TAAB) 0(N0 ) · 4H 0 2

2

NaOMe, CH OH 3

3

4

2

> Fe (TAAB) 0(C10 ) · 4 H 0 2

Λ00°€

4

4

2

HF,H 0 6 hrs.

HF,H 0 6 hrs.

2

2

H+

2

H 0 2

Fe(TAAB)F(C10 ) · 2 H 0

Fe{TAAB(OMe) }0

4

2

2

2

H 0 Fe(TAAB)F(N0 ) · 2H 0 2

3

Figure 19.

2

2

Iron complexes of the tetramer of aminobenzaldehyde

Table X I I .

Analysis of Iron—TAAB Dimer

Fe [TAAB(OCH ) ] 0 2

Calcd. Found

3

2

2

C

Η

Ν

Ο

Fe

OCH

Mol Wt

66.91 67.05

4.88 4.96

10.41 10.21

7.43 7.30

10.37 10.28

11.51 10.78

1077 1038, 1080

3


3.

BUSCH E T A L .

tion and the very great stability of the iron (III) oxo-bridged dimer F e ( T A A B ) 0 , that species is best characterized. Its magnetic properties are typical of spin-spin coupled dimers containing F e ( I I I ) of S = 5/2. However, the bridge is quite unusual in another regard. It is exceptionally difficult to break. It has been cleaved only by H F after extended treatment—it is like dissolving glass. The stability of the oxo bridge may be taken as evidence for the great reactivity of the iron (III) complex. F r o m the standpoint of models for the oxidized derivatives of peroxidase, the ethoxide derivative of the dimer is most interesting. Data characterizing this species is presented in Table X I I . 2


71


2

4 +

CRH

Figure 20.

Structures of macrocyclic complexes with iron and other metal ions

A variety of other macrocycles have been used in the synthesis of iron and other transition element compounds. The structures of some of these are presented in Figure 20. The abbreviations presented herein are used in the following discussion. Although iron complexes with several of these ligands have been prepared and characterized, only those of the cyclic diene 1,7-CT w i l l be considered here (41). The several


72


Table XIII.

Iron Complexes of Curtis' Diene

Preparation CH CN 3

a) F e ( O A c ) + 1,7-CT · 2 H X 2

b) Fe(C10 ) · 6 H 0 + T E O F Downloaded by UCSF LIB CKM RSCS MGMT on December 3, 2014 | http://pubs.acs.org Publication Date: June 1, 1971 | doi: 10.1021/ba-1971-0100.ch003

4

2

2

• [Fe(l,7-CT)(CH CN) ]X2 24-36 hrs. 3

CH CN 3

reflux

1,7-CT 2 H C 1 0 NEt

2

4

3

[Fe(l,7-CT)(CH CN) ](C10 )2 3

Other Compounds

2

4

[Fe(l,7-CT)C1]C10 very pale green, μ ίί = 5.05

4

β

L i

Br,MeOH(Ni_ Nal, MeOH (Ni)

[Fe(l,7-CT)Br]C10 pale yellow green, [Lett =

4

5.11

[Fe(l,7-CT)I]C10 iL ( = 5.15

4

et

— - ^ ^ C J , [Fe(l,7-CT)C1 ]C10 _ AieOH~~* " g g » Ρ··» = · 2

NCS~. n+

iL

h t

4

r e e n

2

3 0

[Fe(l,7-CT)(NCS)i]B*4 very deep blue, μ.Η = 2.14 [Fe(l,7-CT)(CH CN) ] (C10 ) deep yellow, \j. = 2.19 3

4

2

3

eit

classes of complexes formed by this ligand are indicated in Table XIII. The five-coordinate, high-spin complexes Fe( 1,7-CT ) X constitute ex amples of a broad series of such structures that we find to be formed with a variety of macrocycles. They share their structural characteristics with deoxyhemoglobin, for that extremely important iron (II) compound is five-coordinate, high-spin, and surrounded by a tetradentate macrocycle having four nitrogen donors. W e suggest that the structure of the natural product should not be considered unusual, for this is as common as any structure we have observed among F e ( I I ) complexes of tetra dentate macrocycles. The thiocyanate derivative of Fe( 1,7-CT ) is lowspin, diamagnetic, and six-coordinate, while all of the F e ( I I I ) complexes are low-spin and six-coordinate. The reaction of Fe( 1,7-CT ) with oxy gen is complicated and interesting (Equations 14, 15). +

2+

2+


3.

73


BUSCH E T A L .

H+, 0 [Fe (l,7-CT)(CH CN) ] + n

3

2

2

2

> [Fe (l,7-CT)(CH CN ] + in

CH CN

3

2

(14)

3

3

pink

yellow,

[i it e

=

2.2

0 „ H+ [Fe (l,7-CT)(CH CN) ] i n

3

2

> [Fe (l,4,7,ll-CT)(CH CN) ]

3 +

n

CH CN 24 hrs

3

2

2 +

(15)


3

red, [Left ~

0.2

The Fe(II) complex is very quickly oxidized to the corresponding F e ( I I I ) derivative. However, extended exposure to the air is followed by oxi dative dehydrogenation of the diene to a tetraene. There is evidence for formation of an oxygen adduct during the early phases of this process. The reaction has been put to novel use (Equations 16-19). (0)

Co (l,7-CT)X + i n

Fe (l,7-CT)(CH CN) n

3

[O] 2

2 +

Fe (l,4,7,ll-CT)(CH CN) n

3

(16)

>Fe (l,4,7,ll-CT)(CH CN) n

2

2 +

+ 3 phen

CH CN 3

1,4,7,11-CT + C o B r

>N.R.

2

2

>

3

CH CN 3

CIO4-,

0

2

2 +

(17)

• Fe(phen) + + 1,4,7,11-CT

(18)

> [Co(l,4,7,ll-CT) Br ]C10

(19)

24 hrs

HBr

2

3

2

2

4

Attempts to prepare the tetraene complex of cobalt (III) by oxidation of the cobalt complex failed; however, Goedken has succeeded in remov ing the tetraene from iron and then placing it on cobalt ( I I I ) . New Vitamin B

12

Coenzyme Models

W e think of coenzyme B rather i n the schematic way shown on the left i n Figure 21—cobalt (III) in a tetradentate nitrogen-donor macrocycle with a C o - a l k y l bond and a base i n axial sites. M u c h interest resides i n the determination of the characteristics of the macrocycles that lead to the stability of the cobalt-carbon bond. The burgeoning supply of new macrocycles presents the opportunity of studying such relation ships, and Table X I V provides the first simple correlation of this kind (42, 43). Two ligands from the list gave stable C o - C bonds. One of these, T I M , contains four imines arranged in α-pairs while the other coni 2


74

BIOINORGANIC CHEMISTRY OH OH I I

c—c

/

NH COCH CH 2

NH COCH 2

I CH


2

H\

H

ÇH CR

XH

2

2

2

Jl Β > C H C H C O N H N H N—/ 2

2

2

HNCOCH CH 2

2

2

Me

NH COCH —< D 2

2

Me CH CH CONH 2

2

2

2

I

CH CH Me \ 2

2

Me Me HO

0

/ H HC

HOCH

Figure 21.

H\| .CH

V

2

0

/

Schematic representation of coenzyme B

12

Table XIV. Crystal Field Parameters" of Tetragonal Dichloro-Cobalt ( III ) Complexes Complex

Dq v(cm- ) x

1

[Co(TIM)Cl ]+

2800

[Co(CR)Cl ]+

2830

2

2

[Co(l,7-CT)Cl ]+

2640

[Co(DIM)Cl ]+

2580

[Co(£-CRH)Cl ]+

2580

[Co(cyclam)Cl ]+

2480

2

2

2

2

Form cobalt alkyls

Do not form cobalt alkyls

Values obtained using the formulae (27) Dt = 4/35 (10 Dq*y - v - C) Dq* = (Dq** - 7/4 Dt) assuming Dq (Cl~) = 1460 cm" C = 3800 cm" E

z

1

1


3.

BUSCH E T A L .

75


Table X V . Formation of fre«s-Dialkylcobalt(III) Complexes of TIM and CR [CH Co(TIM)I]B P h 3

1) 2 N a B H 4

2) R X

4

• [R(CH )Co(TIM)]B P h 3

R

[C H CH Co(TIM)Br]B Ph Downloaded by UCSF LIB CKM RSCS MGMT on December 3, 2014 | http://pubs.acs.org Publication Date: June 1, 1971 | doi: 10.1021/ba-1971-0100.ch003

6

6

2

1) 2 N a B H 4

2) R X

=

4

C H , C6JH5CH2 3

4

> [R(C H CH )Co(TIM)]B Ph e

R

6

=

2

4

C H , C6H5CH2 3

Entirely analogous reactions occur i n the C R system. NMR Changes τ (Co-CHz)

Complex [CH Co(TIM)I]B P h [(CH ) Co(TIM)]B P h [CH Co(CR)Br]B P h [(CH ) Co(CR)]B P h 3

8.85 9.64 9.47 9.82

4

3

2

3

2

3

4

4

4

tains two imines and a pyridine ring. The ligands which have failed to yield such derivatives i n our hands contain two ( D I M and 1,7-CT) or fewer unsaturated nitrogen atoms. The data of Table X I V also suggest that only the ligands of greatest coordinating strength ( D q * ) are effec tive i n this regard, and the ability to p i bond is probably of first impor tance. Preparation of the monoalkyl compounds proceeds smoothly i n accord with classic procedures (Equations 20, 21) (42, 43). y

[Co(CR)Br]Br · H 0 +\ NaBH j f MeOH RY or ) >Co(I) • 1 (acetone) [Co(TIM)X ]+ + 1 2NaBH / 2

RCo(CR)Br+

4

2

or

(20)

RCo(TIM)X+

4

Co(II) or

> + N a / H g (1%)

Co(III) I

"

CH CN 3

» Co(I)

RY

>

RCo(MAC)

(21)

(CH CN) + 3

2

More interestingly, series of irans-dialkyl cobalt (III) complexes are read ily synthesized (Table X V ) . This represents only the second report of


76


such compounds (44). The formation of these substances implies the existence of an intermediate nucleophile containing both cobalt(I) and a C o - C bond, for the second alkyl bond is formed by first reducing the cobalt ( III )-monoalkyl complex and then letting it function as a nucleophile in displacing halide from an alkyl halide. Farmery (42) has suc-


Table XVI.

Formation of Alkyl-Cobalt(I) Complexes of CR and TIM

C H C N + MeOH (a) [R C o ( L ) X ] Y + 2 N a B H • R Co(L) (i) Product difficult to isolate (ii) Hydrogénation of ligand often occurs (Hi) Very useful "in situ" synthesis 3

4

(b) [R C o ( L ) X ] Y + N a / H g (1 %) (i) Quantitative reaction (ii) Product easily isolated L = Macrocyclic Ligand

CH CN > R Co(L) 3

X = halide or neutral ligand

Y = B P h " , C10 " , P F 4

4

6

ceeded in isolating, purifying, and characterizing the first complexes of this kind R C o ^ M A C ) (Table X V I ) . The crystals of [ C H C o ( T I M ) ] are red by reflected fight but green by transmitted light. The complex is insensitive to light, stable indefinitely in an inert atmosphere, and it does not decompose below its melting point, 109 °C. It is soluble in hydrocarbons, benzene, ether, alcohols, and other organic solvents and sublimes without decomposition at reduced pressure. It is notable that these alkyl cobalt(I) compounds are much more stable than those somewhat related C o ( I ) compounds, such as C H C o ( C O ) , that have previously been reported. This is a direct consequence of the inertness toward substitution in the planar positions that results from the presence of the macrocyclic ligands and the strong p i electron interactions between the metal ion and the imine nitrogens. Attention has not been directed toward the excellent studies by the groups led by Schrauzer and by Williams, H i l l , and Pratt because of the firm conviction that these matters are to be discussed elsewhere in this volume. 3

3

4

Acknowledgment The studies upon which this report has been based have been generously supported by grants from the National Institute of General


3.

BUSCH E T A L .


77

Medical Sciences of the U . S. Public Health Services, by the National Science Foundation, and by the Ohio State University. W e are exceed ingly grateful.


Literature Cited (1) Curtis, N. F., Coord. Chem. Rev. (1968), 3, 3. (2) Thompson, M. C., Busch, D. H., Chem. Eng. News (1962), Sept. 17, 57. (3) Busch, D. H., "Alfred Werner Commemoration Volume," p. 174, Verlag Helv. Chimica Acta, Basel, 1967. (4) Schrauzer, G. N., Chem. Ber. (1962), 95, 1438. (5) Thierig, D., Umland, F., Angew. Chem. (1962), 74, 388. (6) Johnson, A. W., Kay, I. T., J. Chem. Soc. (1961), 2418. (7) Lindoy, L. F., Busch, D. H., "Preparative Inorganic Reactions," Jolly, Ed., Vol. 6, Interscience, New York, in press. (8) Baldwin, D. Α., Rose, N . J., Abstr. 157th Meeting, ACS, Minneapolis, Minn., 1969, Inor 020. (9) Jager, E. G., Z. Anorg. Allgem. Chem. (1969), 364, 178. (10) Melson, G. Α., Busch, D. H., J. Am. Chem. Soc. (1964), 86, 4834. (11) Curry, J. D., Busch, D. H., J. Am. Chem. Soc. (1964), 86, 592. (12) Green, M., Tasker, P. Α., Chem. Commun. (1968), 518. (13) Barefield, Ε. K., thesis, the Ohio State University, 1969. (14) Goedken, V., Busch, D. H., unpublished results. (15) Cabbiness, D. K., Margerum, D. W., J. Am. Chem. Soc. (1970), 92, 2151. (16) Melson, G. Α., Busch, D. H., J. Am. Chem. Soc. (1965), 87, 1706. (17) Fleischer, Ε. B., Klem, E., Inorg. Chem. (1965), 4, 637. (18) Taylor, L. T., Busch, D. H., J. Am. Chem. Soc. (1967), 89, 5372. (19) Vallee, B. L., Williams, R. J. P., Proc. Natl. Acad. Sci. (1968), 59, 498. (20) Fleischer, Ε. B., Accts. Chem. Res. (1970), 3, 105. (21) Hoard, J. L., "Hemes and Hemoproteins," Chance, Esterbrook, and Yonetani, Eds., p. 9 ff., Academic, New York, 1966. (22) Hawkinson, S. W., Fleischer, Ε. B., Inorg. Chem. (1969), 8, 2402. (23) Countryman, R., Collins, D. M., Hoard, J. L., J. Am. Chem. Soc. (1969), 91, 5166. (24) Collman, J. P., Schneider, P. W., Inorg. Chem. (1966), 5, 1380. (25) Rosen, W., Busch, D. H., Inorg. Chem. (1970), 9, 262. (26) Brubaker, G. R., Busch, D. H., Inorg. Chem. (1966), 5, 2114. (27) Wentworth, R. A. D., Piper, T. S., Inorg. Chem. (1965), 4, 709. (28) Travis, Kenton, thesis, the Ohio State University, 1970. (29) Lever, A. B. P., ADVAN. C H E M . SER. (1967), 62, 430.

(30) Rowley, D. Α., Drago, R. S., Inorg. Chem. (1968), 7, 795. (31) Tokel, Ν. E., Katovic, V., Farmery, K., Anderson, L . B., Busch, D. H., J. Am. Chem. Soc. (1970), 92, 400. (32) Wolberg, Α., Manassen, J., J. Am. Chem. Soc. (1970), 92, 2982. (33) Katovic, V., Taylor, L. T., Busch, D. H., J. Am. Chem. Soc. (1969), 91, 2122. (34) Taylor, L. T., Urbach, F. L., Busch, D. H., J. Am. Chem. Soc. (1969), 91, 1072. (35) Dolphin, D., Felton, R. H., Borg, D. C., Fajer, J., J. Am. Chem. Soc. (1970), 92, 743. (36) Yamazaki, I., Yokota, K., Biochim. Biophys. Acta (1965), 105, 301. (37) Hayaishi, O., "The Oxygenases," p. 280, Academic, New York, 1962. (38) George, P., Biochem. J. (1953), 55, 220. (39) Brill, A. S., Sandberg, H. E., Biochemistry (1968), 7, 4254.


78


(40) Katovic, V., Busch, D. H., unpublished results. (41) Goedken, V., Busch, D. H., unpublished results. (42) Farmery, K., Busch, D. H., 2nd Central Regional Meeting, ACS, Columbus, Ohio, June 3-5, 1970. (43) Ochiai, E., Long, K. M., Sperati, C. R., Busch, D. H., J. Am. Chem. Soc. (1969), 91, 3201. (44) Costa, G., Mestroni, G., Licari, T., Mestroni, E., Inorg.Nucl.Chem. Letters (1969), 5, 561. June 26, 1970.


RECEIVED


Chemical Foundations for the Understanding of ... - ACS Publications

Recommend Documents