Zinc thiolate complexes [ZnLn(SR)]+ with azamacrocyclic ligands: Synthesis and structural properties

Article abstract of DOI:10.1002/ejic.200500947

A series of 24 novel zinc thiolate complexes containing saturated macrocyclic polyamine ligands with the general formula [Zn(L_n)(SR)] ClO₄ (L_n: n-dentate azamacrocyclic ligand) has been synthesized. Two different thiolate ligands (R = phenyl, benzyl) and 12 macrocyclic ligands with ring sizes varying from 11 to 16 atoms and containing three, four, and five nitrogen donors (including a new macrocyclic triamine ligand, 8-methyl-[11]aneN₃) have been used. The solid-state structures of 20 of the 24 complexes have been characterized by X-ray diffraction. The crystal lattice of all compounds is comprised exclusively of monomers. The tetracoordinate N₃S compounds exhibit Zn-S bond lengths ranging from 2.2259(5) to 2.2492(9) A and the pentacoordinate N ₄S systems from 2.266(2) to 2.3200(7) A. In addition, two complexes with the very rare N₅S coordination environment for zinc have been found. They exhibit Zn-S bond lengths of 2.3782(7) and 2.42 A (mean value of two independent molecules). The thiolate complexes are suitable model systems for the precursor proteins of matrix metalloproteases (pro-MMPs) as well as for thiolate alkylating enzymes such as cobalamine-dependent/ independent methionine synthases, farnesyl transferase, and the Ada repair protein. Wiley-VCH Verlag GmbH & Co. KGaA, 2006.

Full text of DOI:10.1002/ejic.200500947

FULL PAPER
DOI: 10.1002/ejic.200500947
Zinc Thiolate Complexes [ZnLn(SR)]+ with Azamacrocyclic Ligands: Synthesis
and Structural Properties
Johannes Notni,[a] Helmar Görls,[b] and Ernst Anders*[a]
Keywords: Enzyme models / Macrocyclic ligands / S ligands / Solid-state structures / Zinc
A series of 24 novel zinc thiolate complexes containing satu-
rated macrocyclic polyamine ligands with the general for-
mula [Zn(Ln)(SR)]ClO4 (Ln: n-dentate azamacrocyclic ligand)
has been synthesized. Two different thiolate ligands (R =
phenyl, benzyl) and 12 macrocyclic ligands with ring sizes
varying from 11 to 16 atoms and containing three, four, and
five nitrogen donors (including a new macrocyclic triamine
ligand, 8-methyl-[11]aneN3) have been used. The solid-state
structures of 20 of the 24 complexes have been characterized
by X-ray diffraction. The crystal lattice of all compounds is
comprised exclusively of monomers. The tetracoordinate N3S
compounds exhibit Zn–S bond lengths ranging from
2.2259(5) to 2.2492(9) Å and the pentacoordinate N4S sys-
tems from 2.266(2) to 2.3200(7) Å. In addition, two complexes
with the very rare N5S coordination environment for zinc
have been found. They exhibit Zn–S bond lengths of
2.3782(7) and 2.42 Å (mean value of two independent mole-
cules). The thiolate complexes are suitable model systems for
the precursor proteins of matrix metalloproteases (pro-
MMPs) as well as for thiolate alkylating enzymes such as co-
balamine-dependent/independent methionine synthases,
farnesyl transferase, and the Ada repair protein.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2006)
ates.[11–13] Darensbourg et al. have investigated the alky-
lation of a zinc complex with a bis(thiolate) [N2S2]2– chelate
ligand.[14] They also obtained the thioether derivatives, but
did not discuss the mechanism.
Introduction
Zinc thiolate complexes have received considerable atten-
tion in the past decade. This interest derives from the fact
that many enzymes feature the alkylation (methylation) of
a zinc–thiolate bond.[1,2] The most prominent among them
are the Ada repair protein,[3] cobalamine-dependent/inde-
pendent methionine synthases,[4,5] and farnesyl transfer-
ase.[6] The studies focusing on zinc thiolate complexes have
thus paid special attention to the mechanism of the alky-
lation of thiols. Two different possible reaction modes are
discussed. On the one hand, the alkylation may occur intra-
molecularly and the thiol is alkylated in the zinc-bound
state, whereas the reaction could also start with an initial
heterolytic dissociation, followed by a reaction of free thiol-
ate. Wilker and Lippard have employed the anionic tetrathi-
olate [Zn(SPh)4]2– as a functional model for Ada.[7] They
found a methylation of the complex by (MeO)3PO to yield
PhSMe and postulated a dissociative reaction mechanism.
Vahrenkamp et al. have studied the methylation of [tris(pyr-
azolyl)borato]zinc thiolates (TpR,RZnSR) and have pro-
vided strong evidence (gained from kinetic measurements)
that these reactions occur intramolecularly and feature a
four-center transition structure.[8–10] A similar interpret-
ation has been provided by Carrano et al., who have investi-
gated the methylation of a series of (scorpionato)zinc thiol-
The above (incomplete and arbitrary) list of examples
provides insight into some established applications of zinc
thiolate complexes as enzyme models. They also resemble,
however, the catalytic domain of another important class of
enzymes, the matrix metalloproteases (MMPs or matrixins).
These play an important role in the degradation of extra-
cellular matrix components, such as collagen.[15,16] The en-
zymes are secreted into the extracellular space in a latent
(inactive) form, the so-called proMMPs, therefore an ad-
ditional activation step is required to induce protease ac-
tivity. This is a sensitive process, since the active form can
digest its own framework and thus the timing of activation
is critical. Furthermore, an imbalance of the active enzyme
and its inhibited form is implicated in the pathology of dis-
eases such as arthritis, cancer, multiple sclerosis, and cardio-
vascular diseases.[17] Thus, it is vitally important to gain a
detailed understanding of the activation mechanism.[18]
The catalytic zinc() ion in the latent enzymes − the
MMP precursor proteins (proMMPs) − is bound to the de-
protonated sulfhydryl group of a cysteine and to three histi-
dine residues.[19,20] In vitro experiments have shown that an
activation can be achieved with a variety of reagents, but
the initial step is always a hydrolytic cleavage of the zinc–
sulfur linkage.[21] The coordination sphere of the zinc ion is
then completed by a water molecule to form the catalyti-
cally active His3ZnOH motif.[22] A model reaction for this
[a] Institut für Organische Chemie und Makromolekulare Chemie,
Humboldtstr. 10, 07743 Jena, Germany
E-mail: ernst.anders@uni-jena.de
[b] Institut für Anorganische und Analytische Chemie,
Lessingstr. 8, 07743 Jena, Germany.
1444
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2006, 1444–1455

Zinc Thiolate Complexes with Azamacrocyclic Ligands
FULL PAPER
process has been reported by Parkin et al., who performed
the hydrolytic cleavage of a tripodal S3Zn–SR complex and
obtained free thiolate.[23] To the best of our knowledge, no
model complex that resembles the His3Zn–Cys coordina-
tion motif of the inactive enzyme has been used for these
investigations so far. Our complexes might thus prove to be
useful in this field of research.
Furthermore, recent investigations have suggested that
carbonic anhydrase[24,25] plays a pivotal role in the fixation
of carbonyl sulfide according to the equation LnZnOH +
COS Ǟ LnZnSH + CO2.[26] Quantum mechanical calcula-
tions have revealed that a zinc hydrosulfide moiety His3Zn–
SH is formed in the course of this interconversion.[27]
A
suitable structural model − the neutral hydrosulfide com-
plex TpPh,MeZnSH − has been found by Vahrenkamp et
al.[28,29] Our efforts to obtain an equivalent zinc hydrosul-
fide complex with an azamacrocyclic ligand were not suc-
cessful because these compounds are not stable. Neverthe-
less, due to their positive overall charge, our thiolate com-
plexes also could be regarded as suitable model systems for
this species.
In this article we focus on the solid-state structures of
cationic zinc monothiolates bearing a polyazamacrocyclic
ligand. Azamacrocyclic ligands are available with many ring
sizes and different numbers of donor atoms, therefore they
allow a wide range of complexes to be synthesized with dif-
ferent electronic properties and steric demands. Thus, one
can study the influence of the “protein-imitating” ligand on
the binding mode and reactivity of a zinc-bound thiolate in
great detail. With regard to future investigations or possible
synthetic applications, it is convenient to choose the most
suitable thiolate complex out of a large pool of substances
with different reactivities and solubilities. We cover com-
plexes formed with a wide range of azamacrocycles and re-
veal insights into both zinc thiolate coordination chemistry
and coordination geometries of polyazamacrocyclic ligands
on zinc centers. In subsequent articles, we will report on the
reactions that these thiolates are capable of undergoing.
Figure 1. Macrocyclic polyamine ligands used for complexation ex-
periments.
with captopril, a hypertensive drug containing a thiolate
function, but they did not characterize this substance. Their
work illustrates, however, the biological significance of this
type of compounds. In addition, Addison and Sinn have
reported the complex trans-[Zn(cyclam)(SC6F5)2], which
bears two pentafluorothiophenolate ligands.[32] Initially, we
prepared our complexes in a similar way by treating equi-
molar amounts of the macrocyclic ligand with zinc perchlo-
rate in methanol. This was followed by addition of an equi-
molar amount of sodium thiolate in methanol. The desired
compounds [Zn(Ln)(SR)]ClO4 precipitated with good
yields. The reaction mode is outlined in Equation (1).
Ln + Zn(ClO4)2 + NaSR Ǟ [Zn(Ln)(SR)]ClO4 + NaClO4
(1)
Results and Discussion
A higher purity and better yields were achieved with po-
tassium thiolates since the precipitation of potassium per-
chlorate forces the equilibrium towards the desired prod-
ucts.
However, this synthesis has its limits. The nature of the
thiolate has a marked influence on this reaction: only arene-
and phenylmethanethiolates yield the desired thiolate com-
plexes. This is in contrast to the [tris(pyrazolyl)borato]zinc
thiolates, for which aryl- and alkyl derivatives are avail-
able.[9] Alkyl thiolates (e.g. MeSH, EtSH, tBuSH) only af-
ford the zinc bis(thiolates) Zn(SR)2 and the bis(perchlorate)
complexes [Zn(Ln)](ClO4)2.
Starting Materials
Figure 1 illustrates the macrocyclic ligands that we used
for complexation experiments. They feature ring sizes vary-
ing from 9 to 16 atoms and contain three, four, or five nitro-
gen donors. In addition, we report the synthesis of a new
azamacrocycle, 8-methyl-[11]aneN3 (4). All azamacrocycles
were prepared using a slightly modified Richman–Atkins
procedure.[30]
Complexation Experiments
Restrictions are also encountered when the ligand does
Zinc monothiolates with macrocyclic amine ligands are not possess the proper ring size and the right type and
a novel class of compounds, and only two such examples number of nitrogen donor sites. If the macrocycle is smaller
are documented in the literature. Koike, Kimura et al.[31] than 11 atoms (ligands 1 and 2), the synthesis again yields
postulated the formation of a complex of [Zn(cyclen)]2+ only the bis(thiolate) salts Zn(SR)2 and complexes of the
Eur. J. Inorg. Chem. 2006, 1444–1455
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
1445

J. Notni, H. Görls, E. Anders
FULL PAPER
composition [Zn(L3)2](ClO4)2. The reason is that small ring ure 2), and all four Zn–S bond lengths are in the very nar-
systems form extremely stable metal complexes[33] that are row range of 2.2295(5)–2.2394(6) Å. The Zn–N bonds have
much more stable than complexes with larger rings.[34] In distances between 2.037(2) and 2.076(2) Å. There is no sig-
addition, 9- and 10-membered cyclic triamines do not react nificant correlation with the nature of the N donor, i.e. sec-
according to a 1:1 stoichiometry with zinc ions, which pre- ondary or tertiary. The IR spectra show a single N–H band
vents the coordination of an additional monodentate li-
gand. Instead, a 1:2 complex precipitates from solutions,
thus leaving behind noncomplexed zinc(), which then
forms a bis(thiolate) salt. A similar observation was made
in the case of L4 (tetramethylcyclam, 13). This ligand al-
ways exhibits an unfavorable steric arrangement in its com-
plexes with zinc as well as with other metal ions [it normally
adopts the (+ + + +) configuration;[35] see below for de-
tails]. These complexes are generally less stable than com-
plexes of its unsubstituted analogue, cyclam (8).[36] In the
present case, the zinc() ion is removed from the macrocycle
by formation of the more stable bis(thiolate) of zinc, releas-
ing the free ligand.
Structures
As already mentioned, of all the mercaptans tested only
phenylmethane- and arenethiols form the desired com-
plexes. Therefore we synthesized two series of isomeric com-
plexes [Zn(Ln)(SR)]ClO4 employing phenylmethanethiol (a;
R
= CH2C6H5) and 4-methylbenzenethiol (b; R =
C6H4CH3). Twenty four novel thiolate complexes were ob-
tained, of which 20 were characterized by single-crystal X-
ray diffraction. (In the structure representations, the hydro-
gen atoms on carbon atoms and the perchlorate ions have
been omitted for clarity.)
[Zn(3)(SR)]ClO4 (17) and [Zn(4)(SR)]ClO4 (18)
Figure 2. Molecular structures of 17a, 17b, 18a, and 18b. Selected
bond lengths [Å]: 17a: Zn–S 2.2353(9), Zn–N1 2.044(3), Zn–N2
2.047(3), Zn–N3 2.073(7); 17b: Zn–S 2.2394(6), Zn–N1 2.056(2),
Zn–N2 2.066(4), Zn–N3 2.048(2); 18a: Zn–S 2.2295(5), Zn–N1
2.076(2), Zn–N2 2.060(2), Zn–N3 2.066(2); 18b: Zn–S 2.2358(5),
The complexes derived from the 11-membered ring li-
gands 3 and 4 are remarkably similar regarding their indi-
vidual structural parameters. The N3ZnS unit exhibits a
slightly distorted tetrahedral coordination geometry (Fig- Zn–N1 2.068(2), Zn–N2 2.065(2), Zn–N3 2.037(2).
Figure 3. Molecular structures of 19a, 19b, and 20a. Selected bond lengths [Å]: 19a: Zn–S 2.245(2), Zn–N1 2.046(3), Zn–N2 2.056(3),
Zn–N3 2.052(3); 19b: Zn–S 2.2492(9), Zn–N1 2.036(3), Zn–N2 2.032(3), Zn–N3 2.052(3); 20a: Zn–S 2.2367(7), Zn–N1 2.075(2), Zn–N2
2.050(2).
1446
www.eurjic.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2006, 1444–1455

Zinc Thiolate Complexes with Azamacrocyclic Ligands
FULL PAPER
in the range of 3261–3271 cm–1, which is broadened to a
The nitrogen atoms in these complexes are chiral in na-
varying extent. This indicates the high degree of chemical ture and we describe their actual configuration using estab-
analogy of these hydrogen atoms. Interestingly, the ligand’s lished conventions.[37,38] The positions of hydrogen atoms
methyl substitution in 18a has a marked effect on its solu- or methyl groups are denoted relative to a plane (or near-
bility; it is readily soluble in dichloromethane, while the un- plane) defined by the nitrogen atoms. If the N-bound resi-
substituted analogue, 17a, is not.
dues are found on the side of the complex that bears the
thiolate, their position is indicated with a (+). N-bound resi-
dues opposite the thiolate side are denoted with a (–). It is
obvious that the configuration of the N3 complexes is al-
ways (+ + +), as is expected for ligands whose cavities are
too small for the respective metal ion.[37]
[Zn(5)(SR)]ClO4 (19) and [Zn(6)(SR)]ClO4 (20)
The three structurally characterized compounds 19a,
19b, and 20a (Figure 3), synthesized with ligands 5 and 6,
show the same geometrical equivalence as observed for the
complexes 17 and 18. Slightly longer Zn–S bonds than in
the case of the 11-membered cycles are encountered. How-
ever, the values for the Zn–N bonds are found in the same
range. The IR spectra again show single bands in the region
between 3252 and 3259 cm–1. If the complex is asymmetric,
they are broadened as mentioned above. Compound 20a is
symmetric along the C5–Zn–S plane (apart from a disorder
of the atoms C1, C1A, N1, and C1M); therefore, the two
hydrogen atoms attached to N2 and N2A are identical and
give rise to a single, very narrow IR absorption. This shows
the usefulness of IR absorption bands of N–H groups in
determining the symmetry properties of macrocyclic amine
complexes, a concept which will be referred to again later.
Complex 20b also exhibits a single band at almost the same
value as 20a, but it is noticeably broadened. This indicates
the absence of a strict symmetry here. However, the struc-
ture is likely to follow the same motif. Similar to 18a, the
additional methyl group induces a solubility of 20a in
dichloromethane.
[Zn(7)(SR)]ClO4 (21) and [Zn(8)(SR)]ClO4 (22)
The ring size of ligands 7 and 8 is rather small and all
four structures accordingly exhibit the (+ + + +) configura-
tion (Figure 4). Although compounds 21 both have a pair
of opposing “long” and “short” bonds, the difference of
both “axes” is too small to consider the geometries as being
Coordination Geometries in N4 Complexes
Whether a five-coordinate complex belongs to the trigo-
nal-bipyramidal or the tetragonal-pyramidal type is a ques-
tion which is tricky to answer in the present case. Justifying
an answer solely by the values of the bond angles, and thus
seeking a more or less distorted trigonal bipyramidal axis,
does not suffice. Upon first inspection, all the structures
would appear to be a form of distorted square pyramid (see
Figures 4–7 below). One thus needs to pay closer attention
to the Zn–N bond lengths. In some cases (23a, 23b, and to
a lesser extent 27a), the Zn–N bonds have almost the same
lengths and the coordination polyhedra should therefore be
regarded as being tetragonal-pyramidal with the sulfur
atom in the apical position. The other complexes exhibit
strongly differing values for pairs of opposing Zn–N bonds.
They can therefore be classified as trigonal-bipyramidal. In
these cases the thiolate ligand assumes an equatorial posi-
tion. This effect is most strongly pronounced in complex
26a (Figure 7): the mean deviation between adjacent Zn–N
bond lengths is nearly twice as large (0.14 Å) as the differ-
ence between the average Zn–N and Zn–S bond length
(0.08 Å). The species 21a and 21b must be classified as be-
ing “in between”, since they exhibit neither a peculiar
“axis” nor sufficiently small deviations in the Zn–N bond
lengths.
Figure 4. Molecular structures of 21a, 21b, 22a, and 22b. Selected
bond lengths [Å]: 21a: Zn–S 2.266(2), Zn–N1 2.159(4), Zn–N2
2.205(5), Zn–N3 2.151(4), Zn–N4 2.176(4); 21b: Zn–S 2.2796(7),
Zn–N1 2.175(2), Zn–N2 2.143(3), Zn–N3 2.201(3), Zn–N4
2.162(2); 22a: Zn–S 2.2853(9), Zn–N1 2.218(3), Zn–N2 2.127(3),
Zn–N3 2.187(3), Zn–N4 2.114(4); 22b: Zn–S 2.2957(5), Zn–N1
2.200(2), Zn–N2 2.109(2), Zn–N3 2.208(2), Zn–N4 2.138(2).
Eur. J. Inorg. Chem. 2006, 1444–1455
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
1447

J. Notni, H. Görls, E. Anders
FULL PAPER
trigonal-bipyramidal. The reason for this distortion is the sesses a rigorous molecular symmetry (Cs) in the solid state,
small ring size of the ligand 7, which enforces a marked with the plane of symmetry being defined by C1, Zn, and
rigidity in the coordination polyhedron. Although the sys- S. The IR spectrum of 23b thus exhibits two extraordinarily
tem obviously strives for bipyramidality, steric constraints sharp N–H stretching bands at 3270 and 3259 cm–1 with
imposed by the ring force a tetragonal-pyramidal structure. exactly the same intensity. In contrast, the absorptions of
In the case of complexes 22, the ring is considerably larger 23a at 3272 and 3259 cm–1 are not as sharp and their inten-
and allows the formation of a (severely distorted) trigonal sities show a marked difference.
bipyramid. In agreement with this assignment, the IR spec-
[Zn(10)(SR)]ClO4 (24)
tra show a single N–H stretching band for complexes 21a
(3292 cm–1) and 21b (3301 cm–1), emphasizing the degree of
equivalence of the nitrogen atoms. In contrast to this, the
N–H protons of compounds 22 are chemically inequivalent
and three stretching bands are observed for both 22a (3276,
3288, and 3306 cm–1) and 22b (3272, 3289, and 3322 cm–1).
Complex 24a possesses a (+ – – –) configuration which
complies with the structural symmetry of the ligand (Fig-
ure 5). The symmetry of 24a results in essentially two amine
bands in its IR spectrum. One is located at 3268 cm–1 (br.,
N1–H, N2–H, N3–H) and apparently consists of two
closely overlapping bands, and the second band at
3309 cm–1 is quite sharp (N4–H). The IR spectrum of 24b
[Zn(9)(SR)]ClO4 (23)
With regard to the other [N4] ligands, cyclam (9) yields is very similar. It also exhibits two bands at 3248 and
an unusual result as complexes of 9 are the only ones which 3305 cm–1. We therefore assume that 24b is very likely to
exhibit an almost perfect square-pyramidal geometry (Fig- have the same configuration.
ure 5). This is particularly surprising since the ligand is
[Zn(11)(SR)]ClO4 (25)
clearly large enough to allow the formation of a trigonal
The most significant difference between the two com-
pounds derived from ligand 11 is the configuration of the
tertiary nitrogen atom, which is found to be (+) in complex
25a and (–) in 25b. This results in differing symmetries for
the complexes (Figure 6). A related observation has been
bipyramid. The influence of the thiolate ligand on the con-
formation is minimal. All the Zn–N bond lengths are al-
most identical, with their maximum deviation being less
than 0.02 Å for 23a and 0.01 Å for 23b. The two Zn–S bond
lengths are identical within experimental error. Both com-
plexes exhibit a (+ + – –) configuration which has been
reported as being generally the most stable geometry.[39] Of
the investigated complexes, 23b is the only one which pos-
1
reported by Alcock et al.,[40] who have determined from H
NMR spectroscopic data that
a D2O solution of
[Zn(11)](NO3)2 contains two isomers, one of which is sym-
metric and the other asymmetric. In the present case, this
effect can be observed in the 1H NMR spectrum of 25a but
not for 25b due to the lack of an isolated signal. The ben-
zylic methylene protons of 25a yield two peaks at δ = 3.80
and 3.76 ppm with a relative intensity of 5:1. Its 13C NMR
spectrum shows a predominant set of signals with five ring
methylene peaks. For 25b, two different 13C NMR spectro-
scopic data sets could be detected for the methylene carbon
atoms of the macrocycle, a major set with five peaks and a
Figure 5. Molecular structures of 23a, 23b, and 24a. Selected bond
Figure 6. Molecular structures of 25a and 25b. Selected bond
lengths [Å]: 23a: Zn–S 2.3161(7), Zn–N1 2.155(2), Zn–N2 2.151(2), lengths [Å] and angles [°]: 25a: Zn–S 2.3200(7), Zn–N1 2.118(2),
Zn–N3 2.149(2), Zn–N4 2.139(2); 23b: Zn–S 2.3174(8), Zn–N1 Zn–N2 2.235(2), Zn–N3 2.133(2), Zn–N4 2.202(2); N2–Zn–N4
2.139(2), Zn–N2 2.147(2); 24a: Zn–S 2.2902(6), Zn–N1 2.112(2),
Zn–N2 2.187(2), Zn–N3 2.123(2), Zn–N4 2.218(2).
165.81(9); 25b: Zn–S 2.3182(9), Zn–N1 2.106(3), Zn–N2 2.171(3),
Zn–N3 2.105(3), Zn–N4 2.191(3); N2–Zn–N4 154.3(2).
1448
www.eurjic.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2006, 1444–1455

Zinc Thiolate Complexes with Azamacrocyclic Ligands
FULL PAPER
minor set consisting of ten peaks. We thus assume that in ment, which causes the multiplicity of their NMR signal.
both cases the predominant species is of the same symmet- (Why this is only observed for 25a and not for the other
ric configuration as found in the solid-state structure of 25b. complexes of asymmetric configuration remains a mystery.
The minor component is of an asymmetric configuration, A possible explanation could be found in different velocities
analogous to the solid-state structure of 25a.
of configurational change, which might cause enhanced
The fact that the crystal lattices of 25a and 25b contain line-widths for the other complexes and render the detec-
different configurational isomers is also reflected in their tion of signal multiplicities impossible. However, we are not
IR spectra. Three N–H bands at 3239, 3308 and 3346 cm–1 sure about this.) The IR spectrum shows only three bands
are present in the case of the unsymmetrically configured at 3251, 3266, and 3286 cm–1, which is not surprising since
25a and only two at 3246 (br.) and 3293 cm–1 for the sym- the structural environments of N2 and N4 are very similar.
metric ligand configuration found in 25b.
For compound 26b, only one structural motif could be ob-
tained by X-ray crystallography. It reveals, however, that the
configuration of the ligand is the same as found in 26a.
[Zn(12)(SR)]ClO4 (26)
As mentioned above, 26a (synthesized from ligand 12)
has the largest deviations in its Zn–N bond lengths. It
clearly adopts a trigonal-bipyramidal geometry and a
(+ – – –) configuration of the nitrogen atoms (denoted in
the order N1–N2–N3–N4; see Figure 7). Similar to 25a, the
NMR spectra reveal that a chloroform solution of complex
26a contains two isomers. The 13C NMR spectrum exhibits
11 methylene carbon peaks for the main component and
seven for the minor component. The major component pos-
sesses an asymmetric configuration of the chiral nitrogen
atoms, which is very likely identical to the solid-state struc-
ture of this compound. The symmetric configuration of the
minor component might be (+ – – +) or all-(–), but this
question could not be answered from the routine NMR
[Zn(14)(SR)]ClO4 (27)
Ligand 14 is known to form complexes that are less
stable than complexes formed from smaller macrocycles. A
detailed study has shown that the steric strain in the six-
membered chelate rings is responsible for this.[37] Zinc()
tends to form the complex [Zn(14)](ClO4)2 with a (+ – + –)
configuration. Its stability stems from the fact that the me-
tal ion’s size and the Zn–N bond length ensure a chair con-
formation in each of the four chelate rings. A mixture of
this compound with 27a was obtained when sodium thiol-
ate was used for synthesis; only the potassium salt afforded
the pure thiolate complex. In the case of 27a, the central
Zn2+ ion is pentacoordinate (Figure 7), which obviously
does not allow the nitrogen atoms to arrange in a tetrahe-
dral manner. However, the ligand folds in such a way that
only chelate rings with a chair conformation are formed.
The N–Zn–N angles of the adjacent nitrogen atoms vary
between 78 and 89°. Obviously, ring strain is low enough
to render the complex more stable than [Zn(14)](ClO4)2. In
accordance with the observed molecular symmetry, the IR
spectrum of 27a shows only one N–H absorption at
3249 cm–1. Two IR bands were observed for compound 27b,
which comply with two pairs of chemically nonequivalent
hydrogen atoms. This does not, however, necessarily indi-
cate the presence of another configuration.
1
spectroscopic data. The H NMR spectrum shows a mul-
tiplet for the benzylic methylene protons. Upon closer in-
spection, this multiplet consists of a singlet at δ = 3.85 ppm
(minor component, 15% of total integral value of benzyl
protons) overlapping with an AB spin system at δ =
3.81 ppm (main component, J = 14.3 Hz, ∆δ = 9.7 Hz).
This observation fits perfectly, since the asymmetric config-
uration of the nitrogen atoms ensures a chiral zinc atom.
The benzyl protons are thus in a diastereotopic environ-
[Zn(15)(SR)]ClO4 (28)
When sodium thiolate was used to synthesize complex
28a, the sulfur-free complex [Zn(15)](ClO4)2 (16) was ob-
tained instead (Figure 8). Only the use of the potassium salt
afforded complex 28a, while 28b was also obtained with
sodium p-methylbenzenethiolate. These compounds and
complex 16 are the first reported zinc complexes of the li-
gand [15]aneN5 (15). They all show a high degree of struc-
tural equivalence regarding the configuration of the macro-
cycle. The unit cell of 28b contains two independent mole-
cules with almost superimposable coordination polyhedra,
although noticeable deviations in the bond lengths are en-
countered. Since the two molecules have nearly the same
appearance, only one of them is depicted in Figure 8.
These hexacoordinate zinc complexes with the rare N5S
coordination environment (only one structurally charac-
Figure 7. Molecular structures of 26a and 27a. Selected bond
lengths [Å]: 26a: Zn–S 2.28(2), Zn–N1 2.271(3), Zn–N2 2.115(3),
Zn–N3 2.256(3), Zn–N4 2.123(3); 27a: Zn–S 2.299(2), Zn–N1
2.208(3), Zn–N2 2.182(3), Zn–N3 2.194(3), Zn–N4 2.189(3).
Eur. J. Inorg. Chem. 2006, 1444–1455
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
1449

J. Notni, H. Görls, E. Anders
FULL PAPER
Figure 8. Molecular structures of 16, 28a, and 28b. Selected bond lengths [Å]: 16: Zn–N1 2.100(2), Zn–N2 2.067(2), Zn–N3 2.170(2),
Zn–N4 2.179(2), Zn–N5 2.168(2); 28a: Zn–S 2.3782(7), Zn–N1 2.199(3), Zn–N2 2.366(3), Zn–N3 2.192(3), Zn–N4 2.368(3), Zn–N5
2.221(3); 28b (molecule A): Zn–S 2.405(2), Zn–N1 2.206(7), Zn–N2 2.331(6), Zn–N3 2.152(5), Zn–N4 2.385(6), Zn–N5 2.194(5); (molecule
B, not depicted): Zn–S 2.435(2), Zn–N1 2.265(5), Zn–N2 2.309(6), Zn–N3 2.199(5), Zn–N4 2.307(6), Zn–N5 2.205(5).
terized example has been reported until now[41]) do not have
much in common with the other complexes discussed in this
work. Their coordination polyhedra cannot be assigned to
a specific structure type. Of the complexes discussed in this
study, 28a reveals the largest mean Zn–N bond length
(2.267 Å), and 28b shows the largest deviation within a sin-
gle molecule (0.23 Å). However, the most significant char-
acteristics are probably the zinc–sulfur bond lengths, which
are 2.3782(7) Å for 28a and 2.405(2) and 2.435(2) Å for 28b,
respectively. These are extraordinarily long Zn–S bonds
compared to the average bond length of 2.29 Å for terminal
zinc thiolates listed in the Cambridge Structural Datab-
ase.[1] This makes complexes 28 somewhat equivalent to the
zinc tetrathiolate anion [Zn(SPh)4]2–, which has a Zn–SPh
bond length of 2.36 Å,[42,43] and accentuates their excep-
tional structural properties.
Comparative Considerations
Figure 9. Zn–S bond lengths [Å] (mean value for 28b).
Figure 9 shows that the Zn–S bond lengths depend on
the number or the nature of the nitrogen donor sites in only
a loose way. It is much more obvious that there is a certain
correlation with the ring size: as it increases from 11 to 13
atoms, the Zn–S bond lengths increase from approx. 2.23
to 2.29 Å. The Zn–S bonds in complexes with 14-membered
or larger rings (compounds 23 to 27) seem to scatter around
a “standard” length of approx. 2.3 Å, which is close to the
average zinc thiolate linkage of 2.29 Å as mentioned above.
The fact that compounds 28 are much less stable than the
other complexes is reflected in their extremely long Zn–S
bonds.
In contrast to the previous case, it is not the ring size of
the macrocyclic ligand but rather the number of nitrogen
donors that seems to have the pivotal influence on the
average Zn–N bond lengths. The plot depicted in Figure 10
reveals three distinct groups of compounds, corresponding
to the number of nitrogen atoms involved. Their secondary
or tertiary nature does not contribute to a Zn–N bond
length alteration in a noticeable way. As far as can be Figure 10. Average Zn–N bond lengths [Å].
1450 Eur. J. Inorg. Chem. 2006, 1444–1455
www.eurjic.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Zinc Thiolate Complexes with Azamacrocyclic Ligands
FULL PAPER
judged from this incomplete data (four structures are miss- tensity) of the mentioned degradation products. For the
ing), the type of thiolate has no significant influence either. compounds with 13- to 16-membered rings (28 not consid-
ered), the peak intensities of the solvent complexes increase
with the ring size, reaching a maximum for compounds 27.
NMR Spectra
In addition, the aromatic thiolate yields complexes of con-
siderably higher stability than phenylmethanethiolate. This
agrees with the observations made during the preparation
of 27a and 28a. The “overcrowded” complexes 28 are the
least stable ones. In the case of the phenylmethanethiolate
complex 28a, the molecular peak could not even be de-
tected. In no case was any evidence for a dissociation of the
macrocycle from the zinc center found.
1
The H NMR spectra generally show only nonspecific
multiplets and broadened peaks for the hydrogen atoms at-
tached to the macrocycle. This is very likely caused by dy-
namic effects. Measurements of gradually diluted solutions
with concentrations ranging from 0.1 to 4 m yield unal-
tered spectra. This allows the conclusion that there is no
distinct contribution of intermolecular interactions. Only
the benzyl methylene, the tolyl methyl groups, and the aro-
matic hydrogen atoms exhibit the usual line shape. How-
ever, their shifts show only very little dependence on macro-
cycle substitution. Nevertheless, these signals are suitable
for reaction monitoring.
Conclusions
Azamacrocycles have been found to be suitable ligands
to obtain stable, soluble thiolate complexes of ZnII. In re-
trospect, this cannot be taken as a matter of course. The
fact that only certain ring sizes and thiolates yield the de-
sired complexes illustrates how sensitive these systems are
to steric and electronic influences. (It is worth mentioning
that substituted arene- and phenylmethanethiolates are gen-
erally suitable for these syntheses, as some additional ex-
periments have shown. However, a more detailed investiga-
tion of the issue has not been performed because we did
not expect it to provide any fundamentally enhanced insight
into the matter.) It is surprising that the crystal lattices of
all complexes comprise only monomeric structures. We as-
sume that the steric demand of the benzyl and aryl residues
of the thiolate ligands is a substantial factor behind this
behavior. Less sterically hindered thiolates such as MeS–
would presumably promote the formation of dimers or
oligomers. However, this will probably remain speculation
because these compounds are elusive and our efforts to de-
velop alternative synthetic routes were not successful.
The main advantage of this class of complexes is its
structural diversity. A large number of thiolate–zinc link-
ages that differ in length, stability, and bond order can be
studied. With the exception of 28, their solubility in non-
and semipolar organic solvents allows the investigation of
reactions under conditions that resemble the hydrophobic
cavities of enzyme proteins. In addition, there are two char-
acteristics that make them ideally suited for studies combin-
ing experimental work and computational methods. First,
their molecular structure data are, except for a few exam-
ples, available with high accuracy. Second, they contain a
comparably small number of non-hydrogen atoms. Thus,
quantum mechanical calculations of these compounds can
actually be performed at sufficiently high levels of theory
to gain insight into reaction mechanisms. Finally, their easy
accessibility, long-term stability, and the variety of charac-
terized compounds renders this class of complexes useful
for biomimetic studies wherever zinc thiolates are involved.
The 13C NMR spectra are resolved much better, even
though they sometimes exhibit considerably increased line-
widths for the methylene carbon atoms of the macrocyclic
ligands. For all complexes but 26, the number of detected
methylene peaks is smaller than the number of methylene
carbon atoms of the macrocycle and correlates directly with
the valence symmetry of the macrocyclic ligand. This allows
the conclusion that either the complexes adopt a symmetri-
cal structure in solution or the dynamic conformational
changes proceed quickly enough to override the asymmetry
of the macrocycle that originates in the conformation en-
forced by the metal coordination. Nevertheless, compounds
19, 20b, 22, 23a, 25b, and 26 exhibit additional signals of
small intensities for the carbon atoms of the macrocycle.
This indicates the presence of at least one additional isomer.
In case of the more rigid systems 19, 20b, and 22 it is pos-
sible that conformational changes in the bridging ethylene
and propylene units require comparably high energy barri-
ers and hence are slow on the NMR timescale, thus al-
lowing the detection of the conformers as separate signal
sets. Another possible interconversion between species in
solution is the configuration inversion of a chiral nitrogen
atom. This seems to hold for 23a, 26, and particularly for
25b (see discussion of structures above).
Mass Spectra
Mass spectra, recorded in ESI mode with a mixture of
MeOH and MeCN as solvents, provide valuable infor-
mation on the stability of the Zn–S bond. The weaker the
thiolate is bound, the more species are detected with the
monodentate thiolate ligand being replaced by HO–, MeO–,
or MeCN, or simply without the ligand. Thus, the intensity
of these signals as compared to the molecular peak
[Zn(Ln)(SR)]+ may be considered as an indicator of com-
plex stability.
According to this approach, the mass spectra reveal the
complexes containing small macrocyclic ligands to be the
most stable. All complexes with rings of up to 12 atoms
Experimental Section
Materials and Methods: All reagents used were of analytical purity.
(17–21) show none or only small amounts (up to 10% in- Solvents were not dried prior to use, except where indicated. NMR
Eur. J. Inorg. Chem. 2006, 1444–1455
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
1451

J. Notni, H. Görls, E. Anders
FULL PAPER
spectra were recorded using a Bruker AC 250 spectrometer at a
temperature of 30 °C. IR spectra were recorded from the neat sub-
stances using a Nicolet Avatar 320 FT-IR spectrometer. Elemental
analyses were performed using a Heraeus Vario EL III system.
Melting points were determined with a Büchi Melting Point B 545
apparatus and are uncorrected. Mass spectra were recorded using
a Finnigan MAT SSQ 710 or a Finnigan MAT 900XL TRAP.
tolylsulfonyl)aminopropyl]methylamine[44] (88 mmol) and 1,2-
bis(p-tolylsulfonyloxy)ethane (88 mmol) in dry DMF (1.5 L) was
stirred at 140 °C for 3 h. Three-quarters of the solvent was then
distilled off in vacuo and the mixture was poured into 1 L of an
ice/water mixture. The product was filtered off, washed with plenty
of water, and recrystallized from ethanol to yield 56% of 8-methyl-
1,4-bis(p-toluenesulfonyl)-1,4,8-triazacycloundecane as a colorless
1
solid. M.p. 161 °C. H NMR (250 MHz, CDCl3): δ = 1.78 (q, J =
Preparation of the Azamacrocycles: The preparation essentially fol-
lowed a published procedure.[30] Small alterations were made to the
reaction conditions of the cyclization (2–3 h at 125 °C) and the
workup procedure of the deprotected amine. The preparation of 8-
methyl-[11]aneN3 (4) serves as an example.
5.65 Hz, 4 H, CH2), 2.19 (s, 3 H, N–CH3), 2.45 (s, 6 H, Ph–CH3),
2.58 (t, J = 6.25 Hz, 4 H, CH2), 3.16 (t, J = 5.73 Hz, 4 H, CH2),
3.37 (s, 4 H, CH2), 7.30 and 7.69 (d, J = 6.56 Hz, 4 H each, Ph–
H) ppm. 13C NMR (62.9 MHz, CDCl3): δ = 21.5, 24.4, 42.4, 48.5,
50.8, 52.5, 127.5, 129.7, 134.9, 143.4 ppm. C23H33N3S2O4 (479.67):
Preparation of 8-Methyl-1,4,8-triazacycloundecane (8-Methyl- calcd. C 57.59, H 6.39, N 8.76, S 13.37; found C 57.32, H 7.17, N
[11]aneN3) (4): A solution of the disodium salt[30] of N,N-bis[3-(p-
8.60, S 13.24. For detosylation, the solid was heated in concd.
Table 1. Analytical data for compounds 17–28.
Complex
Composition calcd. [%]
Complex
Composition found [%]
Yield
[%]
M.p.[a]
[°C]
C
H
Cl
N
S
C
H
Cl
N
S
17
40.46
5.98
7.96
9.44
7.20
17a
17b
18a
18b
19a
19b
20a
20b
21a
21b
22a
22b
23a
23b
24a
24b
25a
25b
26a
26b
27a
27b
28a
28b
40.47
40.31
41.78
41.77
41.67
41.75
43.03
43.00
39.01
39.05
40.36
40.41
41.66
41.66
41.80
41.64
42.88
42.90
42.95
42.67
43.58
44.07
40.21
40.48
6.24
5.86
6.19
6.16
6.17
6.05
6.45
6.33
5.88
6.98
6.25
6.23
6.32
6.36
6.16
6.47
6.58
6.77
7.19
6.73
7.22
6.98
6.31
6.53
8.00
7.82
7.94
7.99
7.83
7.54
7.30
7.69
7.54
7.41
7.50
7.51
7.01
6.99
7.05
7.01
6.88
6.96
7.40
6.77
6.58
6.96
7.36
7.25
9.48
9.36
9.09
9.09
9.02
9.07
8.82
8.78
12.10
12.16
11.82
11.90
11.49
11.39
11.49
11.33
11.14
11.24
11.15
11.15
10.21
10.85
13.97
13.92
6.95
6.72
6.55
6.53
6.50
6.61
6.51
6.32
6.44
6.66
6.28
6.45
6.17
6.22
6.40
6.09
6.04
6.09
6.12
5.75
5.60
6.10
6.00
6.06
37
53
53
21
46
65
53
49
35
73
31
50
61
79
60
67
41
56
38
47
60
54
20
66
228–230
310 d
162–164
228–229
187–190
309 d
158–160
218–220
253–256
310d
216–218
292 d
204–206
276 d
221–223
275 d
211–213
230–232
193–195
272 d
173–175
270 d
18, 19
41.84
6.14
7.72
9.15
6.98
20
21
43.39
39.41
40.52
41.81
6.39
5.91
6.16
6.4
7.49
7.70
7.47
7.26
8.87
12.17
11.81
11.47
6.77
6.97
6.76
6.57
22
23, 24
25, 26
43.03
6.62
7.06
11.15
6.38
27a
27b
28
43.80
44.19
40.56
7.17
6.83
6.41
6.46
6.86
7.04
10.21
10.84
13.91
5.85
6.21
6.37
182–183
248 d
[a] d: decomposition.
Table 2. Spectroscopic data for compounds 17–28.
Complex
MS [m/z (%)][a]
IR [cm–1][b]
Complex
MS [m/z (%)][a]
IR [cm–1][b]
17a
17b
18a
18b
19a
19b
20a
20b
21a
21b
22a
22b
344 (100)
344 (100)
358 (100)
358 (100)
358 (100)
358 (100)
372 (100)
372 (100)
359 (100)
3271
3246
3268
3261
3254
3259
3256
3252
3292
23a
23b
24a
24b
25a
25b
26a
26b
27a
27b
28a
28b
387 (100), 281 (94)
387 (100), 281 (35)
387 (100), 281 (64)
387 (100), 281 (34)
401 (100), 295 (46)
401 (100), 295 (54)
401[c] (75), 295 (100)
401 (100), 295 (63)
415[c] (33), 309 (100)
415 (100), 309 (87)
278 (100),[d] 139 (46)
402[c] (29), 278 (100)
3272, 3259
3270, 3259
3309, 3268
3305, 3248
3346, 3308, 3239
3293, 3246
3286, 3266, 3251
3275, 3244
3249, (3199, O–H)
3274, 3226
3353, 3338, 3278
359 (100)
3301
373 (100), 267 (20)
373 (100), 267 (16)
3306, 3288, 3276
3322, 3289, 3272
3352, 3329, 3306, 3268
[a] Only the molecular peak and the peak of the main degradation product [Zn(L)(OH)]+ are given. [b] Only the data of the N–H
absorptions are denoted. [c] Molecular peak (where different from base peak). [d] No molecular peak detected.
1452
www.eurjic.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2006, 1444–1455

Zinc Thiolate Complexes with Azamacrocyclic Ligands
FULL PAPER
H2SO4 (100 mL) for 72 h. A saturated NaOH solution was used to
neutralize the H2SO4 until a pH of 12 was reached. The amine was
extracted with chloroform (5×100 mL), the combined extracts
were dried with Na2SO4, the solvent was evaporated, and the crude
H 12.19, N 22.83 (hygroscopic material; no exact determination of
sample weight possible).
Preparation of [Zn([15]aneN5)](ClO4)2 (16): [15]aneN5 (15; 1 mmol)
was added to a 0.1  solution of Zn(ClO4)2·6H2O (10 mL) in H2O.
After heating for 10 min, the mixture was cooled. A colorless pre-
product distilled using
a kugelrohr apparatus (b.p. 100 °C,
0.1 mbar) to give 2.75 g (32%) of 4 as a colorless oil. IR: ν =
˜
3289 m (NH), 2918 s, 2833 (CH2), 2869 s, 2789 s (CH3), 1677 m, cipitate separated within several hours. The crystals were filtered
1456 s, 1144 s, 1119 s, 1058 s, 737 s cm–1. 1H NMR (250 MHz,
CDCl3): δ = 1.48 (q, J = 5.26 Hz, 4 H, CH2–CH2–CH2), 2.09 (s, 3
H, N–CH3), 2.42 (t, J = 5.43 Hz, 4 H, MeN–CH2), 2.63 (s, 4 H,
CH2–CH2), 2.68 (t, J = 5.29 Hz, 4 H, HN–CH2) ppm. 13C NMR
(62.9 MHz, CDCl3): δ = 25.65, 41.76, 47.43, 48.13, 57.36 ppm.
C9H21N3 (171.26): calcd. C 62.74, H 12.87, N 24.39; found C 59.28,
off, washed twice with ethanol, and dried in vacuo to give 179 mg
of 16 (37%) as colorless crystals, m.p. 182–184 °C.
C10H25Cl2N5O8Zn (479.36): calcd. C 25.04, H 5.25, Cl 14.78, N
14.60; found C 24.95, H 5.18, Cl 14.60, N 14.65. MS: m/z (%) = 378
(100), 278 (28), 160 (18), 140 (31). 1H NMR (250 MHz, DMSO): δ
= 2.1–2.65 (m, 10 H), 2.65–3.15 (m, 10 H), 3.4–3.9 (br. m, 5 H,
Table 3. 1H NMR spectroscopic data for compounds 17–28 (δ [ppm], 250 MHz, 30 °C). Due to dynamic effects only unspecific multiplets
are obtained for the hydrogen atoms of the macrocyclic ligand. Thus, only the data of the protons of the thiolate ligands are given.
Complex
C6H5CH2
C6H5CH2
Complex
C6H5CH2
C6H5CH2
17a[a]
18a[a]
19a[a]
20a[a]
21a[a]
22a[a]
23a[b]
24a[b]
25a[b]
26a[b]
27a[b]
28a[c]
3.71 (s)
3.70 (s)
3.73 (s)
3.73 (s)
3.65 (s)
3.66 (s)
3.79 (s)
3.73 (s)
7.21 (t), 7.33 (t), 7.47 (d)
7.21 (t), 7.33 (t), 7.43 (d)
7.22 (t), 7.33 (t), 7.45 (d)
7.21 (t), 7.34 (t), 7.45 (d)
7.22 (t), 7.30 (t), 7.43 (d)
7.19 (t), 7.31 (t), 7.44 (d)
7.19 (t), 7.31 (t), 7.48 (d)
7.15 (t), 7.23 (t), 7.41 (d)
7.17–7.33 (m), 7.40–7.48 (m)
7.15–7.32 (m), 7.41–7.50 (m)
7.15 (t), 7.22 (t), 7.29 (d)
7.09 (t), 7.22 (t), 7.29 (d)
17b[a]
18b[a]
19b[a]
20b[a]
21b[a]
22b[a]
23b[b]
24b[b]
25b[b]
26b[b]
27b[b]
28b[c]
2.25 (s)
2.25 (s)
2.25 (s)
2.25 (s)
2.24 (s)
2.24 (s)
2.26 (s)
2.26 (s)
2.25 (s)
2.26 (s)
2.25 (s)
2.15 (s)
6.98 (d), 7.29 (d)
6.98 (d), 7.27 (d)
6.98 (d), 7.31 (d)
6.99 (d), 7.33 (d)
6.94 (d), 7.24 (d)
6.95 (d), 7.26 (d)
6.99 (d), 7.28 (d)
6.95 (d), 7.33 (d)
6.95 (d), 7.29 (d)
6.89–7.43 (m)
3.80 (s), 3.76 (s) (20%)[d]
3.81,[e] 3.85 (s) (15%)[d]
3.84 (s)
6.90 (d), 7.42 (d)
6.82 (d), 7.21 (d)
3.57 (s)
[a] In CD3CN. [b] In CDCl3. [c] In [D6]DMSO. [d] Multiple peaks detected, percentage of minor peak given in brackets. [e] AB spin
system, J = 14.3 Hz, ∆δ = 9.7 Hz.
Table 4. 13C NMR spectroscopic data (62.3 MHz, 30 °C) for compounds 17–28.
Complex
Solvent
δ [ppm]
17a
CD3CN
CD3CN
CD3CN
CD3CN
CD3CN
CD3CN
CDCl3
CDCl3
CDCl3
CDCl3
CDCl3
[D6]DMSO
CD3CN
CD3CN
CD3CN
CD3CN
CD3CN
CD3CN
CDCl3
23.5, 28.6, 45.1, 48.3, 126.0, 127.9, 128.3, 146.1
22.6, 28.6, 45.0, 48.0, 48.2, 58.4, 126.0, 127.9, 128.3, 146.0
126.0, 127.9, 128.2, 146.0
23.4, 23.9, 27.1, 28.6, 45.9, 48.0, 48.8, 126.0, 127.9, 128.3, 146.0
29.0, 43.7, 125.6, 127.9, 128.0, 146.3
18a
19a[a]
20a
21a
22a[b]
23a[b]
24a
27.2, 29.0, 43.5, 45.5, 47.7, 49.5, 125.6, 127.9, 128.0, 146.3
27.9, 28.2, 30.3, 47.4, 48.7, 49.1, 51.1, 126.6, 128.5, 128.6, 145.9
28.1, 30.2, 48.2, 49.4, 50.2, 51.6, 126.4, 128.3, 128.4, 145.8
24.6, 30.7, 39.9, 47.3, 49.7, 50.1, 60.1, 126.6, 128.4, 128.5, 145.4
24.8, 28.1, 28.2, 30.2, 47.0, 49.2, 50.2, 51.6, 52.0, 54.0, 54.6, 126.4, 128.1, 128.4, 145.6
27.4,[c] 30.3, 53.8,[c] 126.2, 128.1, 128.6, 145.2
25a
26a[b]
27a
28a
29.7, 45.9,[c] 125.1, 127.7, 128.2, 146.3
19.6, 23.6, 45.2, 48.2, 48.4, 129.1, 133.3, 133.4, 135.0
19.6, 22.7, 45.1, 47.8, 48.2, 58.6, 129.2, 133.3, 133.4, 134.9
19.6, 24.1, 49.2, 129.2, 133.4, 133.6, 135.3
19.6, 21.7, 24.1, 46.4, 48.0, 48.2, 57.6, 129.2, 133.5, 133.7
19.6, 43.7, 128.9, 132.5, 133.4, 137.9
17b
18b
19b[b]
20b[b,d]
21b
22b[b]
23b
19.6, 27.2, 43.5, 45.4, 47.7, 49.6, 129.0, 132.6, 133.6, 138.2
20.8, 24.9, 27.8, 47.8–51.1,[c] 130.0, 133.8, 134.6, 138.2
20.8, 28.1, 48.0, 49.3, 50.2, 51.7, 129.5, 133.8, 135.0, 137.5
20.8, 24.6, 39.7, 47.3, 49.6, 50.0, 60.2, 129.7, 133.9, 135.0, 137.4
24.0, 24.5, 45.1, 46.0 (2 C), 47.3, 49.0, 50.3, 51.2, 56.2, 62.6
20.8, 24.8, 28.2, 47.0, 49.5, 50.5, 51.6, 52.1, 52.9, 54.0, 54.7, 129.5, 133.7, 134.8, 137.5
20.8, 27.3,[c], 53.7,[c] 129.0, 133.5, 135.1, 136.8
24b
CDCl3
CDCl3
25b[e]
26b[b]
27b
28b
CDCl3
CDCl3
[D6]DMSO
20.4, 45.5,[c] 128.6, 130.5, 133.8, 142.2
[a] Due to dynamic effects, too many signals were detected for the ring methylene carbon atoms; therefore, only aromatic signals are
given. [b] Multiple conformers present; only signals of main component given. [c] br. [d] One aromatic signal not detected due to dynamic
processes. [e] Two components due to configuration inversion; ring ligand shifts of the second species (approx. 20%) shown in the second
line.
Eur. J. Inorg. Chem. 2006, 1444–1455
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
1453

J. Notni, H. Görls, E. Anders
FULL PAPER
N–H) ppm. 13C NMR (62.9 MHz, DMSO): δ = 44.6, 44.8, 45.3,
[2]
[3]
E. Kimura, T. Koike, Adv. Inorg. Chem. 1997, 44, 229–261.
L. C. Myers, M. P. Terranova, A. E. Ferentz, G. Wagner, G.-L.
Verdine, Science 1993, 261, 1164.
R. G. Matthews, C. W. Goulding, Curr. Opin. Chem. Biol. 1997,
1, 332–339.
K. Peariso, Z. S. Zhou, A. E. Smith, R. G. Matthews, J. E.
Penner-Hahn, Biochemistry 2001, 40, 987–993.
H.-W. Park, S. R. Boduluri, J. F. Moomaw, P. J. Casey, L. S.
Beese, Science 1997, 275, 1800–1805.
J. J. Wilker, S. J. Lippard, Inorg. Chem. 1997, 36, 969–978.
R. Burth, H. Vahrenkamp, Z. Anorg. Allg. Chem. 1998, 624,
381–385.
46.0 ppm. IR: ν = 3298 m, 2944 m, 2887 m, 1467 m, 1342 w,
˜
1071 vs, 973 s, 945 s, 798 s, 621 vs cm–1.
[4]
[5]
[6]
Preparation of the Thiolate Complexes: The ligand (4 mmol) was
added with stirring to 40 mL of a hot 0.1  solution of Zn(ClO4)2·
6H2O in methanol. Stirring and heating were continued for an
additional 15 min. Depending on the ligand, some precipitation oc-
curred. A solution of 4 mmol of the thiol in 8 mL of 0.5  KOH
was then added dropwise, whereupon potassium perchlorate pre-
cipitated. The solution was filtered and, after a varying amount of
time, colorless crystals separated spontaneously. The precipitates
were filtered off and, if necessary, recrystallized from methanol.
All complexes revealed a 1:1:1:1 composition of zinc/macrocyclic
amine/thiolate/perchlorate, consistent with the general formula
[Zn(L)(SR)]ClO4.
[7]
[8]
[9]
U. Brand, M. Rombach, J. Seebacher, H. Vahrenkamp, Inorg.
Chem. 2001, 40, 6151–6157.
M. Ji, B. Benkmil, H. Vahrenkamp, Inorg. Chem. 2005, 44,
3518–3523.
[10]
[11]
B. S. Hammes, C. J. Carrano, Inorg. Chem. 1999, 38, 4593–
4600.
B. S. Hammes, C. J. Carrano, Inorg. Chem. 2001, 40, 919–927.
J. N. Smith, Z. Shirin, C. J. Carrano, J. Am. Chem. Soc. 2003,
125, 868–869.
C. A. Grapperhaus, T. Tuntulani, J. H. Reibenspies, M. Y. Dar-
ensbourg, Inorg. Chem. 1998, 37, 4052–4058.
J. F. Woessner Jr, FASEB J. 1991, 5, 2145–2154.
H. Nagase, J. F. Woessner Jr, J. Biol. Chem. 1999, 274, 21491–
21494.
Note: It is important that the precipitation starts as quickly as pos-
sible. Prolonged standing of concentrated solutions lowers the
yields and the purity of the products considerably, most likely due
to oxidation processes. Thus, it is best to induce the formation of
crystals by adding a seed crystal or by ultrasound treatment. If the
precipitation of the complex occurs instantly upon addition of the
potassium thiolate, which is not normally the case, separation from
the solid potassium perchlorate is difficult. In such cases, the syn-
thesis should be carried out using a sodium thiolate solution. The
desired complex will then be obtained in a good yield, although
without the removal of the alkali metal perchlorate.
[12]
[13]
[14]
[15]
[16]
[17]
H. Matter, M. Schudok, Curr. Opin. Drug Discovery Dev. 2004,
7, 513–535.
[18]
[19]
N. Borkakoti, Biochem. Soc. Trans. 2004, 32, 17–20.
S. P. Salowe, A. I. Marcy, G. C. Cuca, C. K. Smith, I. E.
Kopka, W. K. Hagmann, J. D. Hermes, Biochemistry 1992, 31,
4535–4540.
R. C. Holz, S. P. Salowe, C. K. Smith, G. C. Cuca, L. Que Jr,
J. Am. Chem. Soc. 1992, 114, 9611–9614.
E. B. Springman, E. L. Angleton, H. Birkedahl-Hansen, H. E.
Van Wart, Proc. Natl. Acad. Sci. USA 1990, 87, 364–368.
H. E. Van Wart, H. Birkedahl-Hansen, Proc. Natl. Acad. Sci.
USA 1990, 87, 5578–5582.
General Properties and Solubility: All thiolate complexes 17–28
form colorless crystalline solids. Substances featuring macrocycles
with ring sizes from 14 to 16 atoms and four nitrogen atoms (i.e.
23–27) are readily soluble in chloroform. All complexes, with the
exception of 28, are soluble in acetonitrile and nitromethane, but
only partially in methanol. Compounds 28 show only low solubility
in all solvents mentioned, but are soluble in DMSO. Other data
[elemental analytical data, melting points, and yields (Table 1), IR
and mass spectroscopic data (Table 2), and 1H (Table 3) and 13C
NMR (Table 4) data] are given in the respective tables.
[20]
[21]
[22]
[23] B. M. Bridgewater, T. Fillebeen, R. A. Friesner, G. Parkin, J.
Chem. Soc., Dalton Trans. 2000, 4494–4496.
[24]
[25]
[26]
[27]
[28]
D.-W. Christianson, C. A. Fierke, Acc. Chem. Res. 1996, 29,
331–339.
T. Wingo, C. Tu, P. J. Laipis, D. N. Silverman, Biochem. Bio-
phys. Res. Commun. 2001, 288, 666–669.
G. Protoschill-Krebs, C. Wilhelm, J. Kesselmeier, Atmos. Envi-
ron. 1996, 30, 3151–3156.
S. Schenk, J. Kesselmeier, E. Anders, Chem. Eur. J. 2004, 10,
3091–3105.
M. Bräuer, E. Anders, S. Sinnecker, W. Koch, M. Rombach,
H. Brombacher, H. Vahrenkamp, Chem. Commun. 2000, 647–
648.
M. Rombach, H. Vahrenkamp, Inorg. Chem. 2001, 40, 6144–
6150.
J. E. Richman, T. J. Atkins, J. Am. Chem. Soc. 1974, 96, 2268–
2270.
T. Koike, M. Takamura, E. Kimura, J. Am. Chem. Soc. 1994,
116, 8443–8449.
A.-W. Addison, E. Sinn, Inorg. Chem. 1983, 22, 1225–1228.
R. Yang, L.-L. Zompa, Inorg. Chem. 1976, 15, 1499–1502.
L.-L. Zompa, Inorg. Chem. 1978, 17, 2531–2536.
M. T. S. Amorim, S. Chave, R. Delgado, J. J. R. F. da Silva, J.
Chem. Soc., Dalton Trans. 1991, 3065–3072.
E. K. Barefield, F. Wagner, Inorg. Chem. 1973, 12, 2435–2439.
R. Luckay, T. E. Chantson, J. H. Reibenspies, R. D. Hancock,
J. Chem. Soc., Dalton Trans. 1995, 1363–1367.
F. Wagner, M. T. Mocella, M. J. D’Aniello Jr, A. H.-J. Wang,
E. K. Barefield, J. Am. Chem. Soc. 1974, 96, 2625–2627.
B. Bosnich, C. K. Poon, M. L. Tobe, Inorg. Chem. 1965, 4,
1102–1108.
Crystal Structure Determination:[45] The intensity data for com-
pound 23b were collected with a Nonius CAD4 diffractometer and
for the other compounds with
a Nonius KappaCCD dif-
fractometer, using graphite-monochromated Mo-Kα radiation.
Data were corrected for Lorentz and polarization effects, but not
for absorption.[46–48] The structures were solved by direct methods
(SHELXS[49]) and refined by full-matrix least-squares techniques
2
against Fo (SHELXL-97[50]). For all compounds except for 17b,
19a, 20a, and 21b, the hydrogen atoms of the amine groups were
located by difference Fourier synthesis and refined isotropically. All
other hydrogen atoms were included at calculated positions with
fixed thermal parameters. All non-hydrogen atoms were refined an-
isotropically.[50] The absolute structures of compounds 16, 19a, 21a,
and 28b could not be determined because they show racemic twin-
ning. XP (SIEMENS Analytical X-ray Instruments, Inc.) was used
for structure representations. The drawings in this paper were gen-
erated using PLATON.[51]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
Acknowledgments
[36]
[37]
Financial support of this work by the Deutsche Forschungsgemein-
schaft (SFB 436: “Metal-Mediated Reactions Modeled After Na-
ture”) is gratefully acknowledged.
[38]
[39]
[1] G. Parkin, Chem. Rev. 2004, 104, 699–767.
1454
www.eurjic.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2006, 1444–1455

Zinc Thiolate Complexes with Azamacrocyclic Ligands
FULL PAPER
[40] N. W. Alcock, P. Moore, C. Pierpoint, J. Chem. Soc., Dalton
Trans. 1984, 2371–2376.
[41] J. Marek, P. Kopel, Z. Travnicek, Acta Crystallogr., Sect. C
2003, 59, 558–560.
[46] MOLEN, An Interactive Structure Solution Procedure, Enraf–
Nonius, Delft, The Netherlands, 1999.
[47] COLLECT, Data Collection Software, Nonius B.V., The Ne-
therlands, 1998.
[42] D. Swenson, N. C. Baenziger, D. Coucouvanis, J. Am. Chem.
Soc. 1978, 100, 1932–1934.
[43] N. Ueyama, T. Sugawara, K. Sasaki, A. Nakamura, S. Yama-
shita, Y. Wakatsuki, H. Yamazaki, N. Yasuoka, Inorg. Chem.
1988, 27, 741–747.
[44] A. Bencini, M. I. Burguete, E. Garcia-Espana, S. V. Luis, J. F.
Miravet, C. Soriano, J. Org. Chem. 1993, 58, 4749–4753.
[45] CCDC-257864 to -257884 contain the supplementary crystallo-
graphic data for this paper. These data can be obtained free of
charge from The Cambridge Crystallographic Data Centre via
http://www.ccdc.cam.ac.uk/data_request/cif.
[48] Z. Otwinowski, W. Minor, in Methods in Enzymology (Eds.:
C. Carter, R. Sweet), Academic Press, New York, 1997, pp.
307–326.
[49] G. Sheldrick, Acta Crystallogr., Sect. A 1990, 46, 467–473.
[50] G. Sheldrick, SHELXL-97, Release 97-2, University of
Göttingen, Germany, 1997.
[51] A. L. Spek, PLATON, A Multipurpose Crystallographic Tool,
Utrecht University, Utrecht, The Netherlands, 2004 (http://
www.cryst.chem.uu.nl/platon).
Received: October 26, 2005
Published Online: February 21, 2006
Eur. J. Inorg. Chem. 2006, 1444–1455
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
1455