The Stereochemistry of Cobalt(III) Complexes

Welcome to my discussion of Cobalt(III) Complexes! I hope you find this as fascinating a study is I do! I began working on the NMR spectroscopy of these systems as part of graduate research, and have retained an interest in this topic. I believe that there is still much to be learned from these systems. I wanted to present here a general overview of the different types of ligands which have been synthesized in past decades and the isomers they form, as well as to offer some ideas as to where future research could proceed.


Quadridentate Ligands

Pentadentate Ligands

Quadridentate Ligands

A ligand is a molecule which has unshared electron pairs. Usually ligands contain the elements nitrogen, oxygen, or sulfur. A vast number of ligands are tetradentate, meaning that they have four donor atoms. When coordinated to Co(III), which has an octahedral geometry, six coordination sites are available. This means that tetradentate ligands occupy four of the six sites, leaving two sites open. These two open sites may be occupied by two monodentate groups such as chloride, or by a bidentate ligand such as carbonate or oxalate. The vast majority of tetradentate ligands fall under two categories: linear ligands and tetradentate ligands. Linear tetradentate ligands are exactly what their name implies; they are linear rather than branched. In contrast, tripodal tetradentate ligands have a central donor atom to which three separate arms are attached. The differences between these two types of tetradentate ligands and their geometries are discussed in the sections which follow.

Linear Tetradentate Ligands

Triethylenetetramine: The "Prototype" Linear Tetradentate Ligand

One of the first linear tetradentate to ligands be studied was 2,2,2-tet or NH2CH2CH2NCH2CH2NCH2CH2NH2. Many other tetradentate ligands can be considered to be derivatives of triethylenetetramine, and for the purposes of discussion this is considered to be the "prototype" tetradentate ligand.

Very early on it was recognized that octahedral complexes containing this ligand can adopt any one of three geometrical isomers, the geometry being entirely dependent upon the way in which the ligand coordinates to the cobalt center. These three isomers are the trans isomer, the symmetrical cis isomer, and the unsymmetrical cis isomer [1,2]. In the trans isomer, all four of the donor atoms lie in one plane with the cobalt ion. The two monodentate groups are situated 180 degrees apart, and the complex possess a mirror plane of symmetry. The symmetrical cis isomer contains a C2 axis and the two monodentate ligands are 90 degress apart. The unsymmetrical cis isomer contains no symmetry elements at all.

trans isomer

symmetrical-cis isomer

unsymmetrical-cis isomer

The symmetry of these complexes are important from a spectroscopic standpoint. Since the symmetrical-cis isomer contains a C2 axis and the trans isomer contains a mirror plane, one would expect only three carbon signals arising from the triethylenetaramine. These two isomers could prove difficult to distinguish from their NMR standpoint. On the other hand, the trans isomer does not possess any symmetry elements, and so six carbon signals would be observed for any complex adopting this geometry.

In later years, researchers began to investigate ligands which would coordinate in a more stereospecific way. Could some variation of triethylenetetramine be made which adopts to form single isomer, rather than all three? A variety of approaches have been used. One of the first modifications was the introduction of symmetric methyl goups at the (2,9), (3,8), and (5,6) position. Another modification was an increase in the size of the carbon chains between nitrogen donors This led to the ligands 2,3,2-tet, 3,2,3-tet and 3,3,3-tet. Still another area of research has involed the replacement of nitrogen donors with sulfur, yielding ligands with NSSN and SNNS donor sequences. In the following sections these approaches will be discussed.

Ligands with Methyl Groups on Carbon Chain

One early attempt at producing a more sterospecific ligand was the symmetrical addition of methyl groups onto the ligand. Three different dimethyl derivatives of 2,2,2-tet have been synthesized, shown below, and each adopts a different geometric isomer.

NH2(CH3)CHCH2NCH2CH2NCH2CH(CH3)NH2 2,9-dimethyltriethylenetetramine
NH2CH2CH(CH3)NCH2CH2N(CH3)CHCH2NH2 3,8-dimethyltriethylenetetramine
NH2CH2CH2NCH(CH3)(CH3)CHNCH2CH2NH2 5,6 dimethyltriethylenetetramine

The first, studied by Robert Asperger and Chui Fan Liu, [3], was reported to form the all three isomers. NMR spectroscopy was used to determine the isomerism of the isolated products. The symmetrical cis isomer was reported to be purple in color, whereas the trans isomer was reporetd to be green. A red color was reported for the unsymmetrical cis isomer.

The second ligand, studied by Yoshikawa, Sekihara, and Goto [4] was found to adopt only the trans configuration. The complexes were found to preferentially form the unsymmetrical cis isomer when isolated as the dinitrite and oxalate species. Proton NMR helped to establish the identity of these isomers. As a symmetrical-cis or trans isomer, the methyl groups would have been equivalent and contributed to a single signal. However, two distinct signals were observed for the dinitrite and oxalate species, indicating the formation of the unsymmetrical isomer. A third chloro species was found to be a mixture of isomers.

Complexes of the third ligand are red in color and when isolated as the nitro salt, the results of NMR measurments indicate that this ligand coordinates to form only the unsymmetrical cis isomer [5].

The effect of placement of the methyl groups upon steroeochemistry is clearly illustrated by these studies. Placement of the methyl groups into the 2,9 position yields has little effect upon the preferred isomer, whereas placement of methyl groups into the 3,8, and 5,6 positions yield and unsymmetrical cis isomers.

Molecular Model of 3,8 dimethytriethyleme Cobalt(III) Chloride. Note the trans configuraton.

Molecular Model of symmetrical cis 2,9 dimethyltriethylenetetramine Cobalt(III) Chloride.

Complexes with Lengthened Carbon Chains

A second avenue of approach centered around changing the length of the carbon chains joining the donor atoms in triethylenetetramine. Would longer carbon chains between the nitrogen donors provide a ligand which would prefer one geometric isomer over the other? Considering only the symmetrical molecules, there are three possible variations called 2,3,2-tet, 3,2,3-tet, and 3,3,3-tet. The chemical formulas for these ligands are listed in the table below.

NH2CH2CH2NCH2CH2CH2NCH2CH2NH2 (2,3,2-tet)
NH2CH2CH2CH2NCH2CH2NCH2CH2CH2NH2 (3,2,3-tet)

Dale Alexander and Hobart Hamilton [6,7] studied the ligands 2,3,2-tet and 3,2,3-tet. Based upon the results of visible absorption spectroscopy, both complexes were reported to prefer the trans configuration. This is in contrast to the prototype triethylenetetramine, which is known to adopt all three geometric isomers as previously discussed. It was discovered that (3,2,3-tet) could be forced to to adopt a symmetrical cis geometry only if a bidentate ligand such as carbonate or oxalate was used to occupy the remaining two coordination sites.

Nearly two decades later, Yamamoto, Kudo, and Toyota began a series of extensive NMR studies of complexes [Co(2,3,2-tet)Cl2]Cl, [Co(3,2,3-tet)Cl2]Cl, and [Co(3,3,3-tet)Cl2]Cl. All were observed to form the trans isomer. Hoever, the complexes [Co(2,3,2-tet)(sal)]Cl, [Co(3,2,3-tet)(sal)]Cl, and [Co(3,3,3-tet)(sal)]Cl were also reported, where sal represents the bidentate ligand salicylate. These complexes, in contrast to the chloro ismers, were shown to adopt the unsymmetrical cis isomer.

Molecular Model of trans [Co(2,3,2-tet)Cl2]X

Complexes with NSSN Donor Sequences

A third attempt to create a more stereospecific ligand centered around the replacement of donor atoms in triethylenetetramine. One possible repplacement is Sulfur, which should serve as an acceptable replacment for nitrogen as an electron pair donor. Work in this area was pioneered by Worrell and Busch in the mid 1960s. One of the first sulfur-containing ligands to be investigeated was 1,8-diamno, 3-6 dithiaoctane or NH2CH2CH2SCH2CH2SCH2CH2NH2 [9,10]. This ligand is often abbreviated eee (for ethyl-ethyl-ethyl). Several complexes with cobalt(III) have been isolated, including the nitro and chloro salts. The nitro salt [Co(eee)(NO2)2]Cl is rust-red in color and whereas the chloro salt [Co(eee)(Cl2)]Cl is dark blue in color. The ligand is known to adopt only the symmetrical cis isomer, a fact which has attributed to the increased size of the sulfur atoms relative to nitrogen. The x-ray structures of this complex has been determined, and it has recently been investigated by modern high-resolution NMR techniques. The sister ligand epe, NH2CH2CH2SCH(CH3)CH2SCH2CH2NH2 [11], contans a single methyl group. Complexes of this ligand also adopt the symmetrical cis geometry. The geometry had been determined from X-ray crystallography, and complexes containing this ligand have also been studied by modern NMR techniques.

Molecular model of symmetrical cis [Co(epe)Cl2]X

This work was also pursued by Bosnich, Keene, and Phillip, [12] who report complexes with the following nitrogen-sulfur ligands:


The latter two can be viewed as derivatives of (2,3,2-tet) and (3,2,3-tet). The complex [Co(ETE)Cl2]+ was observed to form a red-violet unsymmetrical cis and a green trans isomer. However, the symmetrical cis isomer was not observed. Recall that the trans isomer is preferred by the all-nitrogen analog of ETE. Therefore the length of the carbon chains joining the donor atoms is clearly seen to have an effect upon coordination geometry. In contrast, the complex [Co(TET)Cl2]+ was also reported to form two isomers; a red-violet unsymmeytrical cis isomer and a blue symmetrical cis isomer. In contrast to the all-nitrogen analog (3,2,3-tet), this ligand was not observed to adopt the trans configuration.

Complexes With SNNS Donor Sequences

More than a decade later, other researchers [13,14] studied the efects of an SNNS donor sequence upon coordination geometry. Four such ligands have been investigated:


Since these ligands place the sulfur atoms at the ends of the ligand, the stereospecifity characteristic of many similar ligands is not observed here. The first two ligands form a dark green symmetrical cis isomer and a dark violet unsymmetrical cis isomer, [Co(SNNS)(en)]+ where en represets the ligand ethylene diamine. The third ligand is not reported to form any stable complexes. The fourth was isolated as [Co(SNNS)(en)]+ and [Co(SNNS)(R,R-chxn)]+ where (R,R-chxn) stands for the bidentate ligand 1,2-cyclohexanediamine. This ligand was reported to form only the unsymmetrical cis isomer.

Tripodal Tetradentate Ligands

Tris(2-aminoethyl)amine: the prototype tripodal ligand

Just as triethylenetetramine can be considered the prototype linear tetradentate ligand, tris(2-aminoethyl)amine could be considered to be the prototype tripodal ligand. This ligand has the chemical formula N(CH2CH2NH2)3 and is usually abbreviated tren Complexes with tris(2-aminoethyl)amine have totally different symmetry considerations than for complexes with triethylenetetramine. Before proceeding with a discussion of the literature, we will begin with a discussion of some symmetry considerations.

First, setting aside any asymmetry which might arise from the monodentate or bidentate ligands occupying the remaining two coordination sites, that tris(2-aminoethyl)amine can form only one isomer. So, for example, the complex [Co(tren)]Cl3 forms only a single isomer. Second, consider that the complex [Co(tren)]Cl3 possesses a mirror plane of symmetry but no C2 axis like the symmetrical-cis [Co(trien)Cl2]Cl system. Therefore the symmetry of these systems is quite different. Consulting the illustration below for the [Co(tren)Cl2]Cl system, the two C1 carbons would contribute to the same C-13 signal, the two C2 carbons would contribute to the same signal, and the C3 and C4 carbons would each have individual signals. Therefore four carbon signals would be observed in the C-13 NMR spectrum, as opposed to the symmetrical-cis or trans isomers of [Co(trien)Cl2]Cl, where three carbon signals are observed.

Model of [Co(tren)Cl2]+
Showing Equivalent Carbons

The differences in the symmetry between the [Co(trien)Cl2]Cl and [Co(tren)Cl2]Cl arise again when considering complexes with bidentate ligands such as [Co(tren)(C2O4)]+, illustrated below. While these systems form a single isomer, the two carbons of the oxalate ligand are not equivalent; one is trans to the central nitrogen donor of the tren ligand and one is trans to a terminal nitrogen donor of the tren ligand. These are labeled A and B in the model shown below. The noneqivalence of these carbons could become significant when considering the NMR spectra of complexes with large bidentate ligands such as 2,2-bipyridine or 1,10-phenanthroline.

Model of [Co(trn)(ox)+

Lastly, consider the geometrical isomerism which arises with the use of asymmetric bidentate ligand or systems with different monodentate ligands, such as [Co(tren)ICl]+. In this system there are actually two different geometric isomers, depending upon the way in which the two monodentate groups are coordinated. In one isomer, the I- is coordinated trans to the terminal nitrogen of the tren ligand and the Cl- is coordinated trans to a terminal nitrogen; in another isomer the Cl is coorinated trans to the central nitrogen of the tren ligand and the I is coordinated trans to a terminal nitrogen. These two isomers are illustrated below.

Complexes of this ligand have been studied independently by several different researchers. Kimura, Young, and Collman studied a series of amino acid complexe [15]. W.G. Jackson and A.M. Sargeson also studied and charaterized several cobalt(III)-tren complexes with cysteine, cyteine, and thiother derivatives [16]. Eiko Toyota, Yutaka Yamamoto, and Yoshihsa Yamamoto [17] studied a series of complexes of the formula [Co(tren)(sal)]+ were sal represents salicylate and several methyl salicylate derivatives. Most of these complexes are asymmetric, depending upon the way in which the salicylate and salicylate derivatices are coordinated. Detailed carbon and proton assignments are reported for these systems. Massoud and Milburn [18, 19] also studied complexes with this tripodal ligand; systems studied by thir group include [Co(tren)(CO3)]+, [Co(tren)(C2O4)]+. [Co(tren)(en)]+, as well as several others. Carbon NMR assignments are reported for these systems, but proton NMR assignments are not reported.


Whereas a large number of variations on the linear tetradentate ligand triethylenetetramine have been reported, there have been few derivatives of tris(2-aminoethyl)amine. One variation which has been studied by Massoud and Milburn [18,19] is the ligand N(CH2CH2CH2NH2)3, which is abbreviated trpn. The only difference between the two ligands lies in the fact that in trpn the nitrogen donors are joined by three-carbon linkages, whereas in tren they are joined by two-carbon linkages. As with tren, compelxes with trpn contain a mirror place of symmetry and six carbon signals are observed rather than the nine which might be expected. Complexes with this ligand have not been as widely studied as for the tren ligand.

Molecular Model of [Co(trpn)Cl2]+

N,N-bis (2-aminoethyl)glycinate

Recall that one approach to the synthesis of trien derivatives was development of ligands with alternate donor sequences such as SNNS and NSSN. Similarly one appraoch to the synthesis of tren derivatives includes ligands in which one or more of the terminal amines are replaced with other donors. In one ligand, N,N-bis (2-aminoethyl) glycinate (which has been given the designation i-dtma) one of the terminal amine is replaced with an carboxyl group; this ligand therefore coordinates with three nitrogen and one oxygen donors. Unlike the prototyle ligand tren, this ligand can coordinate in two different ways. In the complex S-[Co(i-dtma)Cl2]+ (shown below on the left) the ligand coordinates in such a way that the mirror plane is retained. Therefore the usual four donor signals would be observed in the C-13 NMR spectrum of this compelx. In the complex U-[Co(i-drma)Cl2]+ (shown below on the right) the mirror place is not retained. In this complex all of the carbon atoms are nonequivalent and so six signals would be predicted in the C-13 NMR spectrum. When the two monodentate groups are nonequivalent or when an asymmetric bidentate ligand is used, then the number of possible isomers is increased to four. Complexes with this ligand have been studied by Akamatsu, Komorita, and Shimura [20]. Complexes with alanine, glycine, ethylenediamine, and several other bidentate ligands are reported. The various geometrical isomers were identified by visible absoprtion spectroscopy, circular dichroism spectra, and 1-H NMR. The NMR was used only as a "fingerprint" to help in the identification of the various isomers.

    Molecular Model of S-[Co(i-dtma)Cl2]+

Molecular Model of U-[Co(i-dtma)Cl2]+

Linear Pentadentate Ligands

In contrast to the extensive studies performed regarding tetradentate ligands, only a handful of pentadentate compelxes have been studied. Only two pentadentate complexes with nitrogen and sulfur donors have been studied. Whereas tetradentate ligands can possess any one of three isomers, four isomers are theoretically possible for pentadentate ligands.


Ligand Q was studied by Worrell and Jackman [18,19]. Several different compounds have been isolated, including the cloro, bromo, nitrite, and perchlorate salt. This ligand preferentially forms the (alpha, alpha) isomer, which is illustrated below. This stereospecifity is again attributed to the size of the sulfur donor atoms relative to nitrogen. Considering three consecutive donor atoms, a facial configuration is preferred to a meridional configuration.

Molecular model of [Co(Q)Cl]2+

The sister ligand, QS, differes only by the center donor and was studied by Worrell and Goddard [20]. Like Q, complexes with ligand QS prefferentiallfy form the (alpha,alpha) isomer. Both complexes have been studied by x-ray crystallography [21] as well as two-dimensional NMR techniques [22].


Triethylenetetramine Complexes

1. Sargeson and Searle. Inorganic Chemistry. 4, 45 (1965)

2. Sargeson and Searle. Inorganic Chemistry. 4, 787 (1967)

3. Buckingham, Marzilli, and Sargeson. Inorganic Chemistry, 6, 1032 (1966)

4. D.A. Buckingham, P.J. Creswell, R.J. Dellaca, M. dwyer, G.J. Gainsford, L.G. Marzilli, I.E. Maxwell, Ward T. RObinson, A.M. Sargeson, and K.R. Turnbull. Journal of the American Chemical Society. 96:6, 1713 (1973)

5.Yoshihisa Yamamoto and Eiko Toyota. Bulletin of the Chemical Society of Japan. 52, 2540 (1979)

6.Yoshihisa Yamamoto, Hiroko Kudo, and Eiko Toyota. Bulletin of the Chemical Society of Japan. 56, 1051(1983)

7.Yoshihisa Yamamoto, Eiko Toyota, and Yutaka Yamamoto. Bulletin of the Chemical Society of Japan. 56, 2721 (1983)

8.Yoshihisa Yamamoto, and Hiroko Kudo. Bulletin of the Chemical Society of Japan. 57, 287 (1984)

9.Yoshihisa Yamamoto, Eiko Toyota, and Shoko Tsukuda. Bulletin of the Chemical Society of Japan. 58, 1595 (1985)

10.Yoshihisa Yamamoto, Kayoko Yoshii, Eiko Toyota, and Kazuko Konno. Bulletin of the Chemical Society of Japan. 62, 724 (1989)

Derivatives of Triethylenetetramine

11. Asperger and Liu. Inorganic Chemistry. 4, 1395 (1965)

12. Yoshikawa, Sekihara, and Goto. Inorganic Chemistry. 6, 169 (1967)

13. Goto, Saburi, and Yoshikawa. Inorganic Chemistry. 8, 358 (1968)

14. H.G. Hamilton and D. Alexander. Inorganic Chemistry. 5, 2060 (1966)

15. D. Alexander and H.G. Hamilton. Inorganic Chemistry 5, 2131 (1969)

NSSN Ligands

16. J.H. Worrell and D.H. Busch. Inorganic Chemistry. 8, 1563 (1969)

17. J.H. Worrell and D.H. Busch. Inorganic Chemistry. 8, 1572 (1969)

18. J.H. Worrell, T.E. MacDermott, and D.H. Busch. J. Am. Chem. Soc. 92, 3318 (1970).

19. B. Bosnich, W.R. Keene, and A.T. Phillip. Inorganic Chemistry. 8, 2567, (1969)

20. M.R. McClure. M.S. Thesis. University of South Florida, 1995.

21. M.R. McClure. Ph.D. Dissertation. University of South Florida, 1999.

22. M.R McClure and Jay H. Worrell. Journal of Coordination Chemistry.

23. M.R. McClure and Jay H. Worrell. Coordination Chemistry Reviews.

SNNS Ligands

24. Yamanari, Takeshita, and Shimura. Bull. Chem. Soc. Jpn. 57, 1227 (1984)

25.Yamanari, Takeshita, and Shimura. Bull. Chem. Soc. Jpn. 57, 2852 (1984)

Tripodal Ligands

15 Eiichi Kimura, Stefan Young, and James P. Collman. Inorganic Chemistry. 9, 1183 (1970).

16. W.G. Jackson and A. M. Sargeson. Inorganic Chemistry. 17, 2165 (1978)

17. Yamamoto and Yamamoto. Bull. Chem. Soc. Jpn. 61, 3175 (1988)

18. Massoud and Milburn. Inorganica Chimica Acta. 154, 115 (1988)

19. Massoud and Milburn. Polyhedron. 8, 2389 (1989)

20. Keji Akamatsum, Takashi Komorita, and Yoichi Shimura. Inorganic chemistry. 21, 2223 - 2226

Pentadentate Ligands

21. J.H. Worrell and T.A. Jackman. J. Inorg. Nucl. Chem. 39, 981 (1977)

22. J.H. Worrell and T.A. Jackman. Inorganic Chemistry. 17, 3358 (1979)

23. Roger A. Goddard, Ph.D. Dissertation, University of South Florida

24.Thoms Li. Ph.D. Dissertation. University of South Florida