An investigation of l,4,7-tri(4-alkynyl)-l,4,7-triazacyclononanes: Ligand synthesis and metal co-ordination chemistry

Article abstract of DOI:10.1039/b007324p

Three triazacyclononanes bearing pendant alkynyl groups [l,4,7-tri(4-pentynyl)-l,4,7-triazacyclononane (ptacn), 1,4,7-tri(5-phenyl-4-pentynyl)-l,4,7-triazacyclononane (pptacn) and 1,4,7-tri(4-hexynyl)-1,4,7-triazacyclononane (4htacn)] have been synthesize

Full text of DOI:10.1039/b007324p

View Article Online / Journal Homepage / Table of Contents for this issue
An investigation of 1,4,7-tri(4-alkynyl)-1,4,7-triazacyclononanes:
ligand synthesis and metal co-ordination chemistry†
Murray V. Baker,* David H. Brown, Brian W. Skelton and Allan H. White
Department of Chemistry, The University of Western Australia, Nedlands, WA 6907, Australia
Received 11th September 2000, Accepted 20th October 2000
First published as an Advance Article on the web 29th November 2000
Three triazacyclononanes bearing pendant alkynyl groups [1,4,7-tri(4-pentynyl)-1,4,7-triazacyclononane (ptacn),
1,4,7-tri(5-phenyl-4-pentynyl)-1,4,7-triazacyclononane (pptacn) and 1,4,7-tri(4-hexynyl)-1,4,7-triazacyclononane
(4htacn)] have been synthesized. Seven complexes, [Mo(CO)3(pptacn)], [Rh(cod)(pptacn)][PF6], [RhCl3(pptacn)],
[CuCl2(pptacn)], [Ni2(µ-Cl)3(pptacn)2]Cl, [Rh(CO)2(4htacn)][PF6] and [Mo(CO)3(ptacn)], each of which bears three
pendant alkynyl groups, have been prepared. Single-crystal structure determinations for [H(4htacn)][BPh4],
[Mo(CO)3(pptacn)], [Rh(cod)(pptacn)][PF6] and [CuCl2(pptacn)] are reported.
Introduction
Results and discussion
We recently reported the synthesis of 1,3,5-tri(4-pentynyl)-
1,3,5-triazacyclohexane and its Cr(CO)3, Mo(CO)3 and CrCl3
complexes.1 The instability of the triazacyclohexane ligand and
the metal complexes has prevented any substantial study into
intramolecular alkyne–metal bonding. We have now turned our
attention to the well known triazacyclononanes (tacn), a system
which has a more robust azacycle core and well known metal
co-ordination chemistry.2,3 The increased stability of the ligands
and their metal complexes should provide a platform for
continued study into the intramolecular co-ordination of
alkynes to metals, an area which still remains, in the most part,
unexplored. Reports of alkynyl functionalised triazacyclo-
nonanes are scarce. Latva and co-workers4,5 have studied the
fluorescence of lanthanide ions in the presence of an alkynyl–
tacn system, but did not isolate any metal–tacn complexes.
Peacock and co-workers have recently published the first isol-
ation and crystallographic study of alkynyl functionalised tacn
metal complexes,6 complexes of CuII, NiII, CoII, and Mo0 being
reported, most of which contained a single alkynyl group linked
to a tacn nitrogen by a methylene or ethylene spacer unit.
Although various metal complexes of simple alkyl substituted
tacns have been recorded, including Rh–CO,7 Rh–cod
(cod = cycloocta-1,5-diene),8 Rh-halide9 and Group 6 metal
carbonyl and halide adducts,10,11 the study of the co-ordination
chemistry of R3tacns, where R is an extended alkyl substituent,
generally remains limited. By contrast, there are many reports
of the co-ordination chemistry of R3tacn where R is methyl or
isopropyl.
Synthesis of alkynyl functionalised 1,4,7-triazacyclononanes
The majority of N-functionalised tacns are prepared with
reactive reagents such as aldehydes, acid chlorides, halogeno-
methylcarbonyl compounds and benzyl halides2,3 with a few
examples of alkyl halides12–15 being used. The easiest synthetic
method for the preparation of alkynyl–tacns appeared to be via
the alkyl halide route. Consequently, the alkynyl tacn ligands
were synthesized by the reaction of 1,4,7-triazacyclononane
trihydroiodide with the corresponding alkynyl iodide and
ethyldiisopropylamine in acetonitrile (Scheme 1). This
To investigate the possibility of metal–alkyne interactions in
alkynyl–tacn complexes, we have explored the use of a three-
carbon spacer unit between the alkynyl group and the tacn
nitrogens, rather than the shorter units examined by Peacock
and co-workers. Here we report the synthesis and some
co-ordination chemistry of three trialkynyl tacns, 1,4,7-tri(4-
pentynyl)-1,4,7-triazacyclononane (ptacn), 1,4,7-tri(5-phenyl-
4-pentynyl)-1,4,7-triazacyclononane (pptacn) and 1,4,7-tri-
(4-hexynyl)-1,4,7-triazacyclononane (4htacn).
Scheme 1 Ts = Toluene-p-sulfonyl.
method appears to be general for the high yield preparation of
tacns substituted with extended alkyl chains. Initial attempts at
the synthesis of the ligands began using tacnؒ3HCl instead of
tacnؒ3HI but the reactions were slow. Analysis of the reaction
† Electronic supplementary information (ESI) available: unit cell
contents of [Mo(CO)3(pptach)]. See http://www.rsc.org/suppdata/dt/
b0/b007324p/
DOI: 10.1039/b007324p
J. Chem. Soc., Dalton Trans., 2000, 4607–4616
This journal is © The Royal Society of Chemistry 2000
4607

View Article Online
mixtures indicated that reactions involving tacnؒ3HCl were
complicated by conversion of the alkynyl iodides into the less-
reactive alkynyl chlorides.
The salt tacnؒ3HI was prepared by a variation of the method
of Fabbrizzi and co-workers.16 It was isolated, in 62% yield, as a
white powder by treatment of free tacn with purified hydriodic
acid. It is far more soluble than the chloride analogue. The
alkynyl iodides were all easily prepared (see Experimental
section).
The phenyl-terminated alkynyl ligand pptacn was prepared
in ca. 60% yield from tacnؒ3HI, 5-iodo-1-phenyl-1-pentyne and
(i-Pr)2EtN in acetonitrile. If Et3N was used instead of (i-Pr)2-
EtN then the yield of pptacn was reduced, as the iodoalkyne
reacted with Et3N to form a quaternary ammonium species,
triethyl(5-phenyl-4-pentynyl)ammonium iodide. Purification of
pptacn was achieved by flash chromatography (silica with 5%
Et3N/CHCl3). The early fractions in the purification showed
traces of 1-phenylpent-4-en-1-yne, an elimination product. The
ligand pptacn is readily soluble in most organic solvents. The
syntheses of 4htacn and ptacn were analogous to that of
pptacn.
tacn system, and subsequently investigate reactions of pendant
alkynyl groups with the metal centre. Metal–tacn complexes
which contained CO, C2H4, cod, or Cl as ancillary ligands
were targeted, as there are precedents for removal of these
ancillary ligands from other metal–tacn systems and related
complexes.9,10,18,19
The methyl-terminated (i.e. 4-hexynyl) ligand 4htacn was
purified by rapid silica filtration and isolated in ca. 75% yield.
The basic nature of the tri-substituted tacns was exemplified by
the immediate precipitation of [H(4htacn)][BPh4] on addition
of a methanolic solution of sodium tetraphenylborate to a
methanolic solution of 4htacn. In this experiment the proton
source was the methanol, with the insoluble nature of the
product providing the driving force for the reaction. This pro-
tonation phenomenon has been noted previously for the bulky
systems 1,4,7-triisopropyl-1,4,7-triazacyclononane and 1,4,7-
triisobutyl-1,4,7-triazacyclononane in the presence of a protic
solvent (alcohol or water) and a suitably large counter anion
Molybdenum complexes. Pptacn reacted rapidly with
[Mo(CO)3(CH3CH2CN)3] in acetone-d6 to produce [Mo(CO)3-
(pptacn)] in quantitative yield (by 1H NMR spectroscopy). The
reaction solution appeared to be stable and showed no change
on standing. On a preparative scale the reaction was performed
in acetone and was complete within 1 h. The complex was isol-
ated in high yield (≈73%) as a pale beige powder. The high yield
suggested that there were no significant complications arising
from alkyne–metal interactions during the reaction. Attempted
preparation of [Mo(CO)3(pptacn)] by way of the literature
methods for tacn–Mo(CO)3 adducts, using Mo(CO)6 in a high
boiling solvent,11 resulted in intractable black-brown decom-
position products.
Ϫ
(ClO4 or FeCl4Ϫ).11,12 Wieghardt and co-workers reported a
large formation constant for the mono-protonated form of
1,4,7-triisopropyl-tacn [pKa (i-Pr3tacnؒHϩ) > 14], which has
been attributed to intramolecular hydrogen bonding. Conse-
quently it has been recommended that attempted syntheses of
transition-metal complexes of this ligand should avoid protic
species in the reaction (i.e. water, alcohols).11
[Mo(CO)3(pptacn)] showed high solubility in most polar
solvents (e.g. acetone and CH2Cl2) and was insoluble in non-
polar solvents (diethyl ether, hexane and benzene). The complex
was air stable in the solid state but air-sensitive in solution. The
infrared spectrum showed signals consistent with a facially
co-ordinated Mo(CO)3 complex.10,11 The 1H and 13C NMR
spectra showed signals typical of a tacn metal complex, the 1H
NMR signals for the tacn ring protons (NCH2CH2N) changing
from a singlet for the “free” ligand to a complex but sym-
metrical pattern for the metal-bound ligand. The symmetry of
this pattern suggests that it corresponds to an AAЈBBЈ spin
system, i.e. in each NCH2CH2N group the two “inner” protons
are chemically equivalent and the two “outer” protons are
chemically equivalent. Crystallographic studies indicate that in
the solid state the tacn ring adopts a conformation in which the
C–H bonds on the tacn ring are staggered so that each proton is
different. If such a conformation was maintained in solution
the NCH2CH2N groups should appear as an ABCD spin
system; the observation of an AAЈBBЈ pattern suggests the tacn
ring is fluxional, the two possible staggered conformations
interconverting rapidly on the NMR timescale.20 This dynamic
nature is also apparent in the 13C NMR spectrum, which
exhibits only one signal for the ring carbons.20
The structure of the protonated species [H(4htacn)][BPh4]
was confirmed by an X-ray study, suitable crystals being grown
from an EtOAc–MeOH solution. This structure also shows
intramolecular hydrogen bonding (see Structural characteris-
1
ation). The H NMR spectrum for solutions of [H(4htacn)]-
[BPh4] exhibit a downfield singlet for the hydrogen-bonded NH
(δ 9.85, acetone-d6) and a symmetrical multiplet, which we
assign as an AAЈBBЈ pattern, for the ring methylene groups.
The downfield chemical shift of the NH signal is consistent
with the persistence of strong hydrogen bonding in solution.
The apparent symmetry of the ring methylene signals is con-
sistent with a structure in solution where either the NH proton
migrates from one ring nitrogen to another rapidly on the
NMR timescale, or the proton is simultaneously bound to each
nitrogen atom, and where the environments of the exo and endo
ring protons are not undergoing rapid exchange, i.e. inversion
of the tacn ring is slow on the NMR timescale.
The terminal alkynyl ligand ptacn was isolated in ≈60% yield
after purification by rapid silica filtration. In methanol solu-
tions at room temperature, it undergoes a facile intramolecular
hydroamination of one of its pendant alkynyl groups, resulting
in formation of an azonia-spiro[4.8]tridecane (I), synthesis and
characterisation of which, including X-ray studies, are reported
elsewhere.17
Reaction of ptacn with [Mo(CO)3(CH3CH2CN)3] in acetone
afforded [Mo(CO)3(ptacn)] as a beige powder in 65% yield. This
complex is much less soluble than the phenyl terminated
1
alkyne analogue [Mo(CO)3(pptacn)]. The infrared, H and 13C
NMR spectra were consistent with its formulation as fac-
[Mo(CO)3(ptacn)]. It exhibited slight solubility in acetone,
CH2Cl2 and CH3NO2. The complex could be recrystallised from
both acetone and CH2Cl2. This crystallisation is in contrast
to our attempts with the analogous complex of (1,3,5-tri-
(4-pentynyl)-1,3,5-triazacyclohexane), which, while prepared
in an identical manner to [Mo(CO)3(ptacn)], showed such
instability that no crystals could be grown before the complex
decomposed.1 This result demonstrates the greater stability
Synthesis of tacn–metal complexes
To study intramolecular alkyne–metal co-ordination in molyb-
denum and rhodium systems, we aimed first to synthesize
complexes in which the metal was bound by the N3 core of the
4608
J. Chem. Soc., Dalton Trans., 2000, 4607–4616

View Article Online
of the tacn–metal complexes compared to the less robust
triazacyclohexane–metal complexes.
from the gelatinous sample the mass of the residue corre-
sponded to the combined masses of the reagents. When either
1-hexyne or 1,8-nonadiyne was treated with [Rh2(µ-Cl)2(cod)2]
under similar conditions no reaction occurred. It is possible
that ptacn reacted, through one or more of its nitrogens, with
[Rh2(µ-Cl)2(cod)2] to generate a species that subsequently
reacted with the terminal alkynes. The reaction of terminal
alkynes with rhodium complexes is well documented and often
results in Rh–acetylide or alkyne oligomerisation products.21,22
A literature method for the preparation of [RhCl3(tacn)]
complexes involves mixing alcoholic solutions of RhCl3ؒ3H2O
and the tacn ligand, with the product being isolated as a yellow
precipitate.9,23 Flood and co-workers have reported the syn-
thesis of [RhCl3(R3tacn)] (R = methyl, benzyl or neohexyl) and,
whereas the methyl and benzyl complexes were found to be
soluble in dmso, the neohexyl complex was insoluble in all
common organic solvents, including dmso. Their attempts to
prepare [RhCl3(np3tacn)] (np3tacn = 1,4,7-tri(neopentyl)-1,4,7-
triazacyclononane) failed, no reason being suggested.9
Rhodium complexes. [Rh(cod)(pptacn)][PF6] was prepared by
a variation of the literature methods for [Rh(cod)(Me3tacn)]-
[PF6]
(Me3tacn = 1,4,7-trimethyl-1,4,7-triazacyclononane)8
and [Rh(cod)(ttcn)][PF6] (ttcn = 1,4,7-trithiacyclononane).19
The reaction of pptacn and [Rh2(µ-Cl)2(cod)2] in the presence
of KPF6 in acetone yielded [Rh(cod)(pptacn)][PF6], in quanti-
1
tative yield (by H NMR analysis). On a preparative scale the
product was isolated as a yellow powder in 67% yield and was
recrystallised from acetone–ether solutions. It is readily soluble
in acetone, dmso, CHCl3 and CH2Cl2, air-stable in the solid
state, but decomposes slowly in solutions exposed to air. In the
absence of air it shows good thermal stability; a solution of
[Rh(cod)(pptacn)][PF6] in acetone showed no appreciable
1
change when heated at 80 ЊC for 5 h. The H and 13C NMR
spectra of solutions of [Rh(cod)(pptacn)][PF6] showed signals
characteristic of [Rh(cod)(tacn)][PF6] complexes.8 The 1H
NMR spectrum exhibited an AAЈBBЈ pattern for the tacn ring
protons, as well as 3 distinct signals for the cod, co-ordinated
through both its alkene groups. The 13C NMR spectrum
showed a doublet due to rhodium–carbon coupling (J(Rh–C)
14.4 Hz), for the cod alkene carbons. Solutions in acetone-d6
slowly decomposed over the course of 2 months.
The reaction of pptacn with RhCl3ؒ3H2O in ethanol resulted
in immediate precipitation of a very insoluble brown solid. Its
infrared spectrum exhibited broad signals, rather than the sharp
signals characteristic of pptacn which are seen in the spectra of
other pptacn metal complexes, which suggests the presence of
polymeric species. Analysis of the filtrate from the reaction
1
An “NMR tube” reaction of 4htacn with [Rh2(µ-Cl)2(cod)2]
and KPF6 in acetone-d6 resulted in the quantitative formation
of [Rh(cod)(4htacn)][PF6]. A similar reaction with ptacn,
however, resulted in a solution that exhibited large, broad
by H NMR spectroscopy showed broadened and downfield-
shifted signals for the methylene groups adjacent to the ring
nitrogens of pptacn, suggesting the presence of protonated
pptacn species. These results may be a consequence of metal–
alkyne interactions, but in aprotic solvents (see below)
[RhCl3(CH3CN)3] forms a simple mononuclear adduct with
pptacn. Perhaps a combination of metal–alkyne interactions
and complications arising from the basicity of the tacn core and
the presence of the protic solvent (ethanol), similar to those
1
signals in the H NMR spectrum, as well as signals consistent
with free cod, and on standing the sample became gelatinous.
These results suggested that a polymerisation reaction had
occurred in the ptacn experiment. The polymerisation did not,
however, involve acetone-d6; when the solvent was removed
J. Chem. Soc., Dalton Trans., 2000, 4607–4616
4609

View Article Online
encountered previously11 and in the reaction of 4htacn with
methanol (see above), result in the polymeric products in the
RhCl3ؒ3H2O case.
bright yellow crystals from benzene–hexane. The complex is,
however, unstable in solution; it has proven difficult to purify on
a preparative scale, and crystals suitable for X-ray studies have
not been obtained.
The reaction of pptacn with [RhCl3(CH3CN)3] in CH3CN at
room temperature rapidly afforded [RhCl3(pptacn)] as a yellow,
micro-crystalline product in ca. 65% yield. While the solvent
here was “aprotic”, it was not rigorously dry. Small amounts of
adventitious water did not affect the reaction, but the crystal-
line product did contain one molecule of water per molecule of
complex (verified by 1H NMR and microanalysis). The crystals
were air stable and only sparingly soluble in cold dmso. The 1H
and 13C NMR spectra of fresh dmso-d6 solutions of the crystals
showed signals consistent with pptacn bound to a metal
centre. The mass spectrum (FAB) of a fresh solution of
[RhCl3(pptacn)] in dmso showed strong peaks attributed to
{[RhCl3(pptacn)]ؒdmso}ϩ and [RhCl2(pptacn)]ϩ and a weaker
peak attributed to [RhCl2(dmso)(pptacn)]ϩ. [RhCl3(pptacn)]
appeared to undergo slow solvolysis over the course of 2 weeks
in dmso solutions, as indicated by changes in their NMR and
mass spectra.
Nickel and copper complexes. The reaction of pptacn with
CuCl2ؒ2H2O or NiCl2ؒ6H2O in ethanol yielded respectively the
mononuclear species [CuCl2(pptacn)] and a product tentatively
assigned as the dinuclear species [Ni2(µ-Cl)3(pptacn)2]Cl. There
were no apparent complications arising from metal–alkyne
interactions or the formation of (pptacn)Hϩ in these
experiments.
X-Ray studies (see below) confirmed the structure of
[CuCl2(pptacn)] in the solid state, and the mass spectrum (FAB)
for a CH2Cl2 solution of the complex showed a peak due to
[CuCl(pptacn)]ϩ but no peaks that could be attributed to any
dinuclear species. Recrystallisation of [Ni2(µ-Cl)3(pptacn)2]Cl
proved difficult and yielded very fine light green needles, unsuit-
able for X-ray work. The structure of this complex was
proposed on the basis of the mass spectrum (electrospray) of
its solution in CH3CN, which showed only a strong signal
attributed to [Ni2(µ-Cl)3(pptacn)2]ϩ. In principle, an alternative
possible structure is [Ni2Cl2(µ-Cl)2(pptacn)2], which contains
a pair of edge-sharing octahedra. Previous studies2,25 have
shown that, for dinuclear metal halide adducts of 1,4,7-
trimethyl-1,4,7-triazacyclononane, structures containing edge-
sharing octahedra exhibit unfavourable steric interactions,
so that structures containing face-sharing octahedra (as in
[Ni2(µ-Cl)3(pptacn)2]Cl) tend to be formed.
The 1H and 13C NMR spectra of a two week old solution of
[RhCl3(pptacn)] in dmso-d6 showed two sets of signals, one
attributed to [RhCl3(pptacn)] and the other to [RhCl2(dmso-d6)-
(pptacn)]Cl. The mass spectrum (FAB) of this solution showed
a strong signal attributed to the solvolysis product [RhCl2-
(dmso-d6)(pptacn)]ϩ [m/z 812.2530 (requires 812.2561)]. The 1H
NMR spectra of solutions of the complexes showed the
expected behaviours when the relative amounts of dmso-d6 or
chloride were changed; NMR spectra of more dilute solutions
indicated a higher concentration of [RhCl2(dmso-d6)(pptacn)]Cl
compared to [RhCl3(pptacn)], while addition of KCl to a
sample, followed by gentle warming, resulted in spectra
showing an increase in the concentration of [RhCl3(pptacn)]
compared to [RhCl2(dmso-d6)(pptacn)]Cl. Other examples of
solvolysis in RhCl3–tacn systems are known. During this work,
Südfeld and Sheldrick reported the solvolysis of [RhCl3(Me3-
tacn)] in dmso in the presence of Ag(CF3SO3).24 They found
that only one chloride could be displaced by dmso even when
an excess of Agϩ was present.
Structure determinations
The results of the single crystal structure determinations for
[H(4htacn)][BPh4], [Mo(CO)3(pptacn)], [Rh(cod)(pptacn)][PF6]
and [CuCl2(pptacn)], executed at ‘low’-temperature, are con-
sistent with the stoichiometries, connectivities and stereo-
chemistries indicated. In all cases one formula unit devoid of
crystallographic symmetry comprises the asymmetric unit; the
protonation pattern is unambiguous, and there is no reason to
believe the bulk compound to be other than a racemate despite
the fact that in two cases chiral space groups are adopted. In
all of the species the torsion angles adopt a common pattern
around the symmetrically substituted nine-membered ring of
essentially threefold symmetry, regardless of whether the ring is
co-ordinated or not. The substituent dispositions at the three
nitrogens are likewise similar (‘equatorial’) about each ring and
throughout the array, varying but slightly in pitch depending on
co-ordination status, the immediate array being very similar
throughout, e.g. taking the N3 plane as datum, deviations
δC(2,3; 5,6; 8,9; 11,41,71) are: 1.077(3), 0.608(3); 0.983(3),
0.506(3); 1.039(3), 0.561(3); 0.944(3), 0.734(4), 0.690(3) Å in
[H(4tacn)]ϩ and 1.072(8), 0.587(8); 1.061(8), 0.588(7); 1.046(7),
0.561(8); 0.491(9), 0.524(9), 0.533(9) Å in [Mo(CO)3(pptacn)],
perhaps the most conspicuous difference being a ‘flattening’ of
the pendant dispositions in the co-ordinated ligand cf. the
other. Within the ring, the torsion angle pattern is broadly in
conformity with that of the ‘twist–boat–chair’ (‘TBC’) form,
the subject of previous molecular mechanics studies.26
The reaction of ptacn with [RhCl3(CH3CN)3] resulted in a
complex mixture of species which could not satisfactorily be
characterised. On prolonged standing of the mixture many of
the species precipitated. The major product remaining in solu-
tion was tentatively identified as [RhCl3(ptacn)] on the basis of
1
its H NMR spectrum. It may be that in ptacn the terminal
alkynes and ring nitrogens compete for the metal centre: when
the alkynes bind to the metal first, other reactions subsequently
occur and lead to many products, but when the tacn ring nitro-
gens bind first a more stable complex is formed.
The reaction of 4htacn with [Rh2(µ-Cl)2(C2H4)4] and KPF6 in
1
acetone-d6 resulted in a dark brown solution. The H NMR
spectrum of the sample showed many signals but none that
could be attributed to a product of formulation [Rh(C2H4)2-
(4htacn)][PF6]. The analogous reaction with pptacn gave
similar results. The poor results in these experiments may be
due to alkyne–alkene competition for the rhodium centre.
When 4htacn was treated with [Rh2(µ-Cl)2(CO)4] and KPF6
in methanol-d4 or acetone-d6, [Rh(CO)2(4htacn)][PF6] was
1
formed. The H and 13C NMR spectra of the product showed
resonances attributed to 4htacn bound to Rh only via its ring
nitrogens. The 13C NMR spectrum also showed a CO signal (δC
189.2, d, J(RhC) = 72 Hz) indicative of CO groups bound to a
single Rh, i.e. the complex is mononuclear. In previous studies
of the reaction of Me3tacn with [Rh2(µ-Cl)2(CO)4] in methanol
both mononuclear and dinuclear products were reported.7
Increased steric hindrance associated with the larger substitu-
ents in 4htacn relative to Me3tacn may be responsible for the
absence of dinuclear products in the present work. On a pre-
parative scale, crude [Rh(CO)2(4htacn)][PF6] was isolated as
a dark brown-green powder which recrystallised as very thin
[H(4tacn)][BPh4]. This is a less usual example of the tacn
ring in monoprotonated form. The structure (Fig. 1(a))
shows the protonic hydrogen H(1), with close contacts to N(4,
7) (distances: 2.12(2), 2.21(2) Å), the symmetry of the torsions
suggesting that it does not impart undue perturbation to the
conformation, and with a conformation similar to that in its
tri-methyl substituted analogue,27 where, although the ring has
quasi-3 symmetry, torsions in the C–C bonds are 11(2), 18(1),
7(2)Њ and those either side Ϫ103(1), Ϫ108(1), Ϫ113(1); 91(1),
95(1), 91(1)Њ. The pendant rods, that at site 4 disordered, play
a significant, if unremarkable, role in crystal packing. Anion
4610
J. Chem. Soc., Dalton Trans., 2000, 4607–4616

View Article Online
Fig. 1 Molecular projections of: (a) The [H(4htacn)]ϩ cation (the minor component of disorder has single line bonds); (b) [Mo(CO)3(pptacn)];
(c) (i) and (ii) the [Rh(cod)(pptacn)]ϩ cation; and (d) [CuCl2(pptacn)].
geometries are normal, seemingly without untoward inter-
actions with the cation pendants.
which has a similar ring conformation to the present (C–C
torsions 50.5(4), 49.8(4), 50.8(3), N–C Ϫ129.9(3), Ϫ128.7(8),
Ϫ130.8(3), C–N 64.3(3), 64.9(3), 63.2(3)Њ), while in the tris-
(benzyl), 17-electron, cationic counterpart (225 K, 3 symmetry)
Mo–N is 2.281(6) and Mo–C 1.95(1) Å.10 In the latter the ring
conformation is similar to that of the trimethyl substituted
protonated ligand, conforming to 3 symmetry, but with no
mention of ligand disorder, having a torsion in the C–C
ring bond of 2(1)Њ, with those in the bonds to either side
of the nitrogen 98(1), Ϫ100(1)Њ, with ring carbon atom
deviations from the N3 plane 0.81(1), 0.81(2) and the pendant
atom 0.66(2) Å.
1
–
[Mo(CO3)(pptacn)]ؒ2(CH3)2CO. Although devoid of crys-
tallographic symmetry, this molecule, inclusive of substituents,
has putative threefold symmetry (Fig. 1(b)). The molecular axis
is oriented closely parallel to b, with interesting implications
for packing consequent upon Mo lying at (0.74755(1),
Ϫ0.02057(2), 0 (constraint)) (Fig. 2, deposited as ESI).
The metal atom environment is presented in Table 1, 〈Mo–N〉
2.351(9), 〈M–C〉 1.931(6) Å, cf. the tris(i-propyl)-substituted
analogue 〈Mo–N〉 2.369(7), 〈Mo–C〉 1.918(3) Å (T = 298 K),11
J. Chem. Soc., Dalton Trans., 2000, 4607–4616
4611

View Article Online
Table 1 Metal atom environments. r/Å is the metal–ligand atom distance; other entries in the matrices are the angles (Њ) subtended at the metal by
the relevant atoms at the head of the row and column
[Mo(CO)3(pptacn)]
Atom
r
N(4)
N(7)
C(01)
C(02)
C(03)
N(1)
N(4)
N(7)
C(01)
C(02)
C(03)
2.344(5)
2.344(4)
2.364(4)
1.924(5)
1.938(5)
1.930(5)
76.3(2)
76.5(2)
76.1(2)
169.6(2)
103.0(2)
93.2(2)
94.7(2)
170.8(2)
104.1(2)
86.3(2)
103.8(2)
94.7(2)
170.6(2)
86.6(2)
83.5(2)
Mo–C–O angles (carbonyls 1–3) are: 178.4(4), 177.5(4), 177.9(5)Њ.
[Rh(cod)(pptacn)]ϩ
Atom
r
N(4)
N(7)
C(01)
C(02)
C(05)
C(06)
N(1)
N(4)
N(7)
C(01)
C(02)
C(05)
C(06)
2.232(4)
2.415(4)
2.313(4)
2.114(5)
2.075(5)
2.139(5)
2.173(5)
79.7(1)
79.9(1)
75.6(1)
95.6(2)
143.9(2)
139.3(2)
93.7(2)
104.5(2)
173.4(2)
39.5(2)
167.0(2)
90.0(2)
105.3(2)
88.2(2)
81.3(2)
155.7(2)
119.0(2)
89.6(2)
78.3(2)
96.0(2)
37.2(2)
[CuCl2(pptacn)]
Atom
r
N(4)
N(7)
Cl(1)
Cl(2)
N(1)
N(4)
N(7)
Cl(1)
Cl(2)
2.115(1)
2.090(1)
2.228(1)
2.2847(5)
2.2805(5)
83.07(5)
83.00(5)
85.08(5)
90.36(4)
171.00(4)
100.27(3)
173.77(4)
93.55(4)
101.97(4)
92.41(2)
Table 2 Ring torsion angles (Њ). Atoms are designated by number only,
[Rh(cod)(pptacn)][PF6]. The cation in this complex is
intrinsically unsymmetrical, the potential 3 symmetry of the
{Rh(tacn)} moiety incompatible with that of the {Rh(cod)},
resulting in a diverse scatter of Rh–N and Rh–C distances; a
similar array has been described for the tris(methyl)-substituted
analogue8 in which Rh–N (2.339(2), 2.335(2), 2.198(2) Å),
Rh–C (2.083(3), 2.076(3), 2.158(3), 2.169(3) Å) have a number
of more conspicuously pairwise groupings, provoking the
description8 of the rhodium environment as ‘trigonal bipyr-
amidal (TBPY)’, the short Rh–N distance and the midpoint of
the double bond supporting the two long Rh–C distances
defining the TBPY axis, to our mind somewhat uncomfortably,
since the other two nitrogen atoms support an equatorial angle
diminished to 76.77(9)Њ. In the present complex, if there is to be
a similar ascription, the axial array must be N(1), again the
shortest Rh–N, and the midpoint of C(05)–C(06), again the
longest C(cod) distances, with N(1)–Rh–(centroid) 174.2Њ.
However, N(1)–Rh–C(05,06) here are much less symmetrical
(167.0(2), 155.7(2)Њ) than their counterparts in the tris(methyl)
analogue (162.4(1), 160.3(1)Њ), with N(7)–Rh–C(02) also
approaching linearity (173.4(2)Њ) and it is evident that the
array of the tris(methyl) derivative is much more symmetrical,
pseudo-m (but with ligand torsions pseudo-3), degrading
further to that of the present cation, shown in projection down
its ‘axis’, by a twist of the two ligands relative to each other,
Fig. 1(c).
N italicised
[Mo(CO)3-
(pptacn)]
[Rh(cod)
[CuCl2
(pptacn)]
[H(4htacn)]ϩ
(pptacn)]ϩ
1–2–3–4
2–3–4–5
3–4–5–6
4–5–6–7
5–6–7–8
6–7–8–9
7–8–9–1
8–9–1–2
9–1–2–3
8–9–1–11 Ϫ158.4(1)
3–2–1–11 92.3(2)
2–3–4–41 Ϫ166.4(2)
6–5–4–41 102.2(2)
5–6–7–71 Ϫ168.3(2)
9–8–7–71 105.7(2)
44.0(2)
66.7(2)
Ϫ132.3(2)
50.5(2)
65.0(2)
Ϫ128.3(2)
47.6(2)
52.2(6)
64.5(5)
Ϫ132.8(4)
51.0(6)
64.1(6)
Ϫ131.3(5)
51.7(6)
Ϫ49.5(5)
Ϫ65.4(5)
130.8(4)
Ϫ51.6(5)
Ϫ61.8(5)
134.0(4)
Ϫ50.2(5)
Ϫ65.3(5)
127.5(4)
175.1(4)
Ϫ114.5(4)
169.3(4)
Ϫ104.3(4)
Ϫ178.3(4)
Ϫ109.1(4)
45.3(2)
69.4(2)
Ϫ133.6(1)
50.7(2)
63.9(2)
Ϫ129.3(1)
49.1(2)
74.3(2)
Ϫ139.1(2)
63.5(6)
68.1(2)
Ϫ132.7(5)
Ϫ177.1(4)
110.1(5)
Ϫ176.1(4)
110.4(5)
Ϫ174.9(4)
110.4(5)
Ϫ134.4(1)
Ϫ169.1(5)
103.9(2)
Ϫ171.6(1)
107.4(2)
Ϫ163.5(1)
98.2(2)
in the N-(5-pent-2-ynyl)-1,4,7-triazacyclononane adduct,6 in
which a pair of large N–Cu–Cl trans angles (167.6(2), 172.6(2)Њ)
also describe the base of a square pyramid with the substituent
at the apical nitrogen (Cu–N 2.256(4), cf. 2.228(1) Å (the
present)); the other cis-basal Cu–unsubstituted N distances in
that compound (2.026(5), 2.045(4) Å) are rather shorter than in
the present (2.090(1), 2.115(1) Å).
[CuCl2(pptacn)]. In principle, at least, the array of this com-
plex is potentially rather similar. It has a metal environment,
which, apart from perturbations imposed by the unconforming
torsions of the tridentate, is a reasonable approximation to m
symmetry, with N(7) the apex of a square pyramid, the closure
of N(7)–Cu–N(1,4) by the macrocycle allowing enlargement of
N(7)–Cu–Cl(1,2). The geometry is rather similar to that found
Ring torsion angles for all the structures are given in Table 2.
Reactions involving pendant alkynes
Our initial studies of reactions of complexes with pendant
alkynes have focussed on Mo and Rh. For [Mo(CO)3(pptacn)]
in various solutions (CH2Cl2, acetone, thf, CH3NO2) we
4612
J. Chem. Soc., Dalton Trans., 2000, 4607–4616

View Article Online
attempted to labilise the CO groups by oxidation of the metal
centre with ferrocenium hexafluorophosphate or AgBF4.10 In
each case, examination of the product by infrared spectroscopy
180 ЊC. The mixture was stirred, with continued heating, for 10
minutes and then allowed to cool. It was diluted with ethanol
(150 cm3) then poured into diethyl ether (350 cm3). The solid
precipitated was collected and dissolved in water (110 cm3), and
the resulting solution treated with charcoal, heated and filtered.
The filtrate was concentrated in vacuo to 25 cm3 and then made
basic with sodium hydroxide solution (15 M, 30 cm3). The
mixture was filtered and the filtrate continuously extracted with
CHCl3 for 4 days. The organic extract was dried (Na2SO4) and
concentrated in vacuo. The residue was dissolved in ethanol
(25 cm3), filtered, and the filtrate cooled (ice-bath). Aqueous
hydroiodic acid (55% w/w, 30 cm3, 0.22 mol) and then ether
(40 cm3) were added to the ethanolic solution. The resulting
precipitate was collected, washed with ethanol (40 cm3) and
ether (80 cm3) and then dried in vacuo to yield 1,4,7-tri-
azacyclononane trihydroiodide as a white powder (11.1 g, 62%),
mp 248–250 ЊC (decomp.) (Found: C, 14.63; H, 3.39; N, 8.39.
C2H6IN requires C, 14.05; H, 3.54; N, 8.39%); δH (D2O) 3.18
(12H, s, 3 × NCH2CH2N); δC (D2O) 42.7 (NCH2CH2N).
showed CO stretching frequencies (νCO 2015, 1875(br) cmϪ1
)
consistent with three facially co-ordinated CO groups attached
to MoI.10 On photolysis of solutions of the oxidised complexes
a gas (presumably CO) was evolved, and CO bands were of
reduced intensity or absent in the infrared spectra of the
photolysis products. Unfortunately, we have been unable to
identify the photolysis products or obtain evidence for any
metal–alkyne interactions. The products were paramagnetic
and therefore not amenable to characterisation by NMR; they
᎐
could not be crystallised, and C᎐C stretching bands in the
᎐
infrared spectra of [Mo(CO)3(pptacn)] and the oxidation
and photolysis products were too weak to provide useful
information.
When [Rh(cod)(pptacn)][PF6] was heated (at reflux, in pro-
pionitrile solution) or irradiated (visible light, in acetonitrile
1
solution) analysis by H NMR spectroscopy indicated that the
pptacn remained bound to the metal and retained apparent C3
symmetry but the cod ligand had been displaced, presumably
by solvent molecules. There was no evidence for metal–alkyne
interactions in either case.
5-Chloro-1-phenyl-1-pentyne. Phenylacetylene (20.5 cm3,
0.183 mol) was added to a cooled (Ϫ70 ЊC) solution of n-
butyllithium (2.5 M in hexanes, 72 cm3, 0.18 mol) in thf (120
cm3). The mixture was stirred at Ϫ70 ЊC for 15 min and then
1-bromo-3-chloropropane (20 cm3, 0.20 mol) added. The result-
ing mixture was stirred at Ϫ70 ЊC for 30 min and then allowed
to warm to room temperature with continued stirring over-
night. It was then refluxed for 3 h, cooled and concentrated
in vacuo. The residue was diluted with water (100 cm3) and
extracted with ether (2 × 150 cm3). The ether extract was
washed with water (80 cm3) and brine (40 cm3), dried (MgSO4)
and concentrated in vacuo. The residue was distilled
(Kugelrohr) to afford 5-chloro-1-phenyl-1-pentyne as a colour-
less oil (13.4 g, 42%), bp 100 ЊC (1 mmHg). δH (CDCl3) 2.05 (2
H, tt, J(H3H4) 6.8, J(H4H5) 6.4 Hz, 2 × H4), 2.60 (2 H, t,
2 × H3), 3.71 (2 H, t, 2 × H5), 7.26–7.30 (3 H, m, aryl H) and
7.37–7.41 (2 H, m, aryl H); δC (CDCl3) 16.8 (C3), 31.4 (C4),
43.7 (C5), 81.5 (C1), 88.0 (C2), 123.5 (C, aryl), 127.7 (CH, aryl),
128.2 (CH, aryl) and 131.5 (CH, aryl). The 1H NMR spectrum
is consistent with that reported.35
[Rh(CO)2(4htacn)][PF6] appears moderately stable in
acetone, methanol, benzene and toluene solutions for short
periods of time (<1 week). On prolonged standing (>1 week)
1
the H and 13C NMR spectra of the samples begin to change,
becoming very complex. The changes could be hastened by
heating (80 ЊC) or irradiation (UV light, through a Pyrex vessel)
of a sample with provision for removal of evolved CO. The 13
C
NMR spectra of such samples typically show no signals in the
CO region but several doublets in the region δ 135–80, each
with splittings of 5–10 Hz which we tentatively attribute to
Rh–C coupling. These signals may be due to Rh-bound alkyne
1
groups,28 but the complexity of the H and 13C NMR spectra
(e.g. some 13C NMR spectra show more than 50 signals) pre-
vents definitive identification of the complexes present.
Conclusion
We have prepared various metal complexes of tacns bearing
pendant alkynes. Compared to 1,3,5-triazacyclohexanes and
their metal complexes the alkynyl–tacns and their metal com-
plexes show far greater stability, even at elevated temperatures.
The more robust tacn system allows a broader survey of metal
co-ordination than the more sensitive triazacyclohexane
system. Nevertheless, attempts to achieve intramolecular
co-ordination of the pendant alkynes to the metal centres in a
well defined manner have so far been unsuccessful, and many of
our attempts to investigate these systems have been thwarted by
difficulties in characterising products by spectroscopic means.
5-Iodo-1-phenyl-1-pentyne. A mixture of 5-chloro-1-phenyl-
1-pentyne (5.3 g, 30 mmol), sodium iodide (18.9 g, 126 mmol)
and acetone (30 cm3) was heated at reflux for 20 h. The mixture
was filtered and the solid residue washed with acetone (20 cm3).
The filtrate was concentrated in vacuo, diluted with water (40
cm3) and then extracted with ether (2 × 100 cm3). The ether
extract was washed with aqueous sodium thiosulfate solution
(10%, 50 cm3), dried (MgSO4) and concentrated in vacuo to
give 5-iodo-1-phenyl-1-pentyne as a colourless oil (7.0 g, 87%).
δH (CDCl3) 2.07 (2 H, tt, J(H3H4) 6.7, J(H4H5) 6.8 Hz,
2 × H4), 2.54 (2 H, t, 2 × H3), 3.35 (2 H, t, 2 × H5), 7.26–7.29
(3 H, m, aryl H) and 7.36–7.40 (2 H, m, aryl H); δC (CDCl3) 5.4
(C5), 20.4 (C3), 32.1 (C4), 81.6 (C1), 87.5 (C2), 123.5 (C, aryl),
127.8 (CH, aryl), 128.2 (CH, aryl) and 131.5 (CH, aryl).
Experimental
General procedures were as described previously.29 Micro-
analyses were performed by the Microanalytical Unit, Research
School of Chemistry, Australian National University,
Canberra, Australia. 1,4,7-Tri(p-tolylsulfonyl)-1,4,7-triaza-
cyclononane,16 5-iodo-1-pentyne,30 [RhCl3(CH3CN)3],31 [Rh2(µ-
Cl)2(CO)4]32 and [Mo(CO)3(CH3CH2CN)3]33 were prepared
according to literature methods. 5-Chloro-1-pentyne was
purchased from Aldrich. Aqueous hydroiodic acid was purified
by refluxing over red phosphorus, under nitrogen, followed
by filtration through a sintered glass funnel containing solid
carbon dioxide.34
6-Iodo-2-hexyne. n-Butyllithium (2.5 M in hexanes, 19.5 cm3,
49 mmol) was added to a cooled (Ϫ78 ЊC) solution of 5-chloro-
1-pentyne (5.0 cm3, 47 mmol) in thf (20 cm3). The solution was
stirred for 5 min, with cooling, and for 5 min without cooling.
The solution was then cooled to Ϫ78 ЊC and methyl iodide (3.3
cm3, 53 mmol) added. The resulting solution was stirred for 5 h
during which time it was allowed to come to room temperature.
Water (1.5 cm3) was added and the thf distilled off at atmos-
pheric pressure. The residue was extracted with pentane (100
cm3). The pentane extract was washed with water (30 cm3),
dried (MgSO4) and concentrated in vacuo to give a mixture of 6-
iodo-2-hexyne and 6-chloro-2-hexyne. Sodium iodide (25 g, 167
mmol) and acetone (40 cm3) were added and the resulting mix-
ture was heated at reflux for 2 days. The acetone was removed
Synthesis of alkynyl functionalised 1,4,7-triazacyclononanes
1,4,7-Triazacyclononane trihydroiodide. 1,4,7-Tri(p-tolyl-
sulfonyl)-1,4,7-triazacyclononane (20.8 g, 35 mmol) was added
portionwise to concentrated sulfuric acid (65 cm3) heated at
J. Chem. Soc., Dalton Trans., 2000, 4607–4616
4613

View Article Online
by distillation and the residue diluted with water (150 cm3) and
extracted with pentane (150 cm3). The pentane extract was
washed with water (30 cm3) and aqueous sodium thiosulfate
solution (10%, 20 cm3), dried (MgSO4) and concentrated
in vacuo to leave a yellow oil, distillation of which afforded
6-iodo-2-hexyne as a colourless oil (5.9 g, 63%), bp 90–98 ЊC
(32 mmHg). δH (C6D6) 1.46 (3 H, t, J(H1H4) 2.6, 3 × H1), 1.53
(2 H, m, 2 × H5), 1.97 (2 H, m, 2 × H4) and 2.83 (2 H, t,
J(H5H6) 6.8 Hz, 2 × H6); δC (C6D6) 3.2 (C1), 5.4 (C6), 19.8
(C4), 32.7 (C5), 76.7 (C2) and 77.2 (C3). The 1H NMR
spectrum for a CDCl3 solution was consistent with the liter-
ature,36 but the NMR spectrum for a C6D6 solution is a better
indicator of sample purity.
H, 10.61; N, 12.63. C24H39N3ؒ H2O (H2O content estimated
by NMR analysis) requires C,4 75.97; H, 10.17; N, 12.66%].
δH [(CD3)2CO] 1.62 (6 H, tt, J(H1H2) 6.8, J(H2H3) 7.1,
6 × H2), 2.26 (6 H, dt, J(H3H5) 2.7 Hz, 6 × H3), 2.30 (3 H, t,
3 × H5), 2.56 (6 H, t, 6 × H1) and 2.70 (12 H, s, 3 × NCH2-
CH2N); δC [(CD3)2CO] 16.5 (C3), 28.1 (C2), 57.2 (NCH2-
CH2N), 58.2 (C1), 69.8 (C5) and 85.0 (C4).
1
–
[H(4htacn)][BPh4]. A solution of NaBPh4 (24 mg, 70 µmol)
in methanol (2 cm3) was added to a solution of 4htacn (18.5
mg, 50 µmol) in methanol (4 cm3). The resulting mixture
was allowed to stand overnight, and the precipitate collected,
washed with methanol (3 cm3), and dried in vacuo to afford
H(4htacn)ؒBPh4 as fluffy colourless crystals in quantitative
yield, mp 120–122 ЊC (decomp.) (Found: C, 83.69; H, 8.17; N,
6.04. C48H60BN3 requires C, 83.54; H, 8.77; N, 6.09%).
δH [(CD3)2CO] 1.75 (9 H, t, J(H3H6) 2.6 Hz, 9 × H6), 1.79 (6 H,
m, 6 × H2), 2.20 (6 H, m, 6 × H3), 2.94–3.03 (12 H, m,
3 × NCH2CH2N), 3.04 (6 H, m, 6 × H1), 6.78 (4 H, m, aryl),
6.92 (8 H, m, aryl), 7.33 (8 H, m, aryl) and 9.85 (1 H, s, NH);
δC [(CD3)2CO] 3.2 (C6), 16.8 (C2), 26.4 (C3), 50.1 (NCH2-
CH2N), 55.2 (C1), 77.1 (C5), 78.3 (C4), 122.2 (aryl), 126.0 (q,
J(CB) 2.8, aryl), 137.0 (aryl) and 164.9 (q, J(CB) 49.4 Hz, aryl).
1,4,7-Tri(5-phenyl-4-pentynyl)-1,4,7-triazacyclononane
(pptacn). A mixture of ethyldiisopropylamine (13.5 cm3, 78
mmol) in acetonitrile (25 cm3) was added to a mixture of
5-iodo-1-phenyl-1-pentyne (7.26 g, 26.9 mmol) and 1,4,7-
triazacyclononane trihydroiodide (4.22 g, 8.2 mmol) in
acetonitrile (40 cm3). The resulting solution was stirred for 2
days at ambient temperature and then refluxed for 3 days. It was
concentrated in vacuo and then diluted with sodium hydroxide
solution (5 M, 50 cm3) and extracted with CH2Cl2 (200 cm3).
The organic extract was washed with water (50 cm3), dried
(MgSO4) and concentrated in vacuo. The resulting oil was
purified by flash chromatography (silica, 5% Et3N/CHCl3) to
yield 1,4,7-tri(5-phenyl-4-pentynyl)-1,4,7-triazacyclononane as
a light orange oil (3.0 g, 67%) (Found: C, 84.09; H, 7.89; N,
7.44. C13H15N requires C, 84.28; H, 8.16; N, 7.56%). δH (C6D6)
1.64 (6 H, tt, J(H1H2) 6.8, J(H2H3) 7.1 Hz, 6 × H2), 2.42 (6 H,
t, 6 × H3), 2.51 (6 H, t, 6 × H1), 2.63 (12 H, s, 3 × NCH2-
CH2N), 6.94–7.01 (9 H, m, aryl H) and 7.51–7.53 (6 H, m, aryl
H); δC (C6D6) 17.6 (C3), 27.8 (C2), 56.7 (NCH2CH2N), 58.0
(C1), 81.5 (C5), 90.8 (C4), 124.9 (C, aryl), 127.7 (CH, aryl),
128.5 (CH, aryl) and 131.9 (CH, aryl); m/z (EI) 556.3673
(M ϩ 1) (requires 556.3692).
Synthesis of tacn–metal complexes
[Mo(CO)3(pptacn)]. A solution of [Mo(CO)3(CH3CH2CN)3]
(0.179 g, 0.52 mmol) in acetone (2 cm3) was added to a solution
of pptacn (0.289 g, 0.52 mmol) in acetone (2 cm3). The solution
was stirred for 1 h, during which time a precipitate formed. The
mixture was diluted with ether (6 cm3) and the precipitate
collected, washed with ether (10 cm3) and dried in vacuo, to
yield [Mo(CO)3(pptacn)] as a pale beige powder (0.281 g, 73%)
(Found: C, 68.69; H, 6.48; N, 5.75. C42H45MoN3O3 requires
C, 68.56; H, 6.16; N, 5.71%). νmax/cmϪ1 (KBr) 1895 and 1754br
(CO); δH [(CD3)2CO] 2.17 (6 H, m, 6 × H2), 2.50 (6 H, t,
J(H2H3) 6.9 Hz, 6 × H3), 2.94–3.12 (12 H, m, 3 × NCH2-
CH2N), 3.40 (6 H, m, 6 × H1), 7.29–7.32 (9 H, m, aryl) and
7.41–7.44 (6 H, m, aryl); δC [(CD3)2CO] 17.8 (C3), 25.3 (C2),
55.1 (NCH2CH2N), 64.9 (C1), 81.9 (C5), 90.0 (C4), 124.8 (C,
aryl), 128.5 (aryl), 129.2 (aryl), 132.3 (aryl) and 230.3 (CO).
1,4,7-Tri(4-hexynyl)-1,4,7-triazacyclononane (4htacn). Ethyl-
diisopropylamine (4.1 cm3, 24 mmol) was added to a stirred
mixture of 6-iodo-2-hexyne (2.71 g, 13.0 mmol) and 1,4,7-
triazacyclononane trihydroiodide (2.02 g, 3.9 mmol) in
acetonitrile (30 cm3). The resulting solution was stirred for 4
days at ambient temperature and then refluxed for 16 h. It was
concentrated in vacuo and the residue diluted with aqueous
sodium hydroxide solution (3 M, 40 cm3) and then extracted
with CH2Cl2 (2 × 100 cm3). The organic extract was washed
with water (40 cm3), dried (MgSO4) and concentrated in vacuo.
The residue was purified by rapid silica filtration (5% Et3N/
ether) to yield 1,4,7-tri(4-hexynyl)-1,4,7-triazacyclononane as
a pale yellow oil (1.1 g, 76%) (Found: C, 78.04; H, 10.33; N,
11.10. C8H13N requires C, 77.99; H, 10.64; N, 11.37%). δH
(CDCl3) 1.58 (6 H, m, 6 × H2), 1.75 (6 H, t, J(H3H6) 2.5,
6 × H6), 2.14 (6 H, m, 6 × H3), 2.50 (6 H, t, J(H1H2) 7.2 Hz,
6 × H1) and 2.69 (12 H, s, 3 × NCH2CH2N); δC (CDCl3) 3.48
(C6), 16.7 (C3), 27.6 (C2), 56.0 (NCH2CH2N), 57.9 (C1), 75.5
(C5) and 79.2 (C4).
[Mo(CO)3(ptacn)]. A solution of ptacn (0.156 g, 0.48 mmol)
in acetone (5 cm3) was added to a solution of [Mo(CO)3(CH3-
CH2CN)3] (0.165 g, 0.48 mmol) in acetone (5 cm3). A precipi-
tate soon formed and the mixture was stirred for 3 h. It was
diluted with ether (10 cm3) and the precipitate collected, washed
with ether (5 cm3) and dried in vacuo, to give [Mo(CO)3(ptacn)]
as a beige powder (0.158 g, 65%) (Found: C, 55.93; H, 6.41; N,
8.06. C24H33MoN3O3 requires C, 56.80; H, 6.55; N, 8.28%). νmax
/
cmϪ1 (KBr) 1890 and 1743br (CO); δH [(CD3)2CO] 2.08 (6 H, m,
6 × H2), 2.25 (6 H, dt, J(H2H3) 7.2, J(H3H5) 2.7 Hz, 6 × H3),
2.37 (3 H, t, 3 × H5), 2.88–2.97 and 3.01–3.08 (12 H, m,
3 × NCH2CH2N) and 3.28 (6 H, m, 6 × H1); δC [(CD3)2CO]
16.9 (C3), 25.2 (C2), 55.0 (NCH2CH2N), 64.8 (C1), 70.4 (C5),
84.1 (C4) and 230.0 (CO).
1,4,7-Tri(4-pentynyl)-1,4,7-triazacyclononane (ptacn). Ethyl-
diisopropylamine (4.1 cm3, 24 mmol) was added to a stirred
mixture of 5-iodo-1-pentyne (2.4 g, 12 mmol) and 1,4,7-
triazacyclononane trihydroiodide (2.0 g, 3.9 mmol) in
acetonitrile (30 cm3). The resulting solution was stirred for 2
days at ambient temperature and then refluxed for 20 h. It was
concentrated in vacuo and the residue diluted with aqueous
sodium hydroxide solution (3 M, 40 cm3) and extracted with
CH2Cl2 (2 × 100 cm3). The organic extract was washed with
water (40 cm3), dried (MgSO4) and concentrated in vacuo. The
residue was purified by rapid silica filtration (5% Et3N/47.5%
ether/47.5% CH2Cl2) to yield 1,4,7-tri(4-pentynyl)-1,4,7-triaza-
cyclononane as a pale yellow oil (0.74 g, 58%) [Found: C, 76.04;
[Rh(cod)(pptacn)][PF6]. A solution of pptacn (0.163 g, 0.29
mmol), [Rh2(µ-Cl)2(cod)2] (72 mg, 0.15 mmol) and KPF6 (56
mg, 0.30 mmol) in acetone (20 cm3) was stirred for 2 h. The
solution was filtered and the filtrate concentrated in vacuo. The
residue was triturated in benzene (5 cm3) until a fine suspension
formed. The mixture was diluted with ether (10 cm3) and the
precipitate collected, washed with ether (10 cm3) and dried
in vacuo to leave [Rh(cod)(pptacn)][PF6] as a yellow powder
(178 mg, 67%) which was pure according to spectroscopy.
Recrystallisation from acetone–ether solutions yielded analyt-
ically pure yellow crystals (Found: C, 61.35; H, 5.77; N, 4.50.
C47H57F6N3PRh requires C, 61.91; H, 6.30; N, 4.61%). δH
[(CD3)2CO] 1.58 (4 H, m, cod CHH), 2.31 (6 H, m, 6 × H2),
4614
J. Chem. Soc., Dalton Trans., 2000, 4607–4616

View Article Online
2.53 (4 H, m, cod CHH), 2.61 (6 H, t, J(H2H3) 6.9 Hz, 6 × H3),
2.93–3.11 (12 H, m, 3 × NCH2CH2N), 3.34 (6 H, m, 6 × H1),
3.68 (4 H, br s, cod CH) and 7.31–7.40 (15 H, m, aryl);
δC [(CD3)2CO] 17.7 (C3), 24.8 (C2), 31.0 (cod CH2), 55.7
(NCH2CH2N), 61.1 (C1), 74.8 (d, J(Rh–C) 14.4 Hz, cod CH),
82.1 (C5), 89.6 (C4), 124.5 (C, aryl), 128.8 (aryl), 129.3 (aryl)
and 132.2 (aryl).
[CuCl2(pptacn)] from CH3NO2–toluene solutions yielded large
dark green crystals suitable for an X-ray study.
Full-spheres of ‘low-temperature’ CCD area-detector dif-
fractometer data (2θmax = 58Њ, ω scans; monochromatic Mo-Kα
radiation, λ = 0.71073 Å; T ca. 153 K) were measured yielding
Nt(otal) reflections, these merging to N unique (Rint quoted) after
‘empirical’/multiscan absorption correction, No with F > 4σ(F)
being considered ‘observed’ and used in the full matrix least
squares refinement, anisotropic thermal parameter forms for
the non-hydrogen atoms (x, y, z, Uiso)H being treated as indi-
cated. Conventional residuals R, Rw (weights: (σ2(F) ϩ 0.0004
F2)Ϫ1 are quoted at convergence. Neutral atom complex scatter-
ing factors were employed within the XTAL 3.4 program
system.37 Individual features, difficulties, idiosyncrasies are
noted below. In Fig. 1 50% displacement ellipsoids are shown
for the non-hydrogen atoms, hydrogen atoms having arbitrary
radii of 0.1 Å.
[RhCl3(pptacn)]. A solution of pptacn (0.114 g, 0.205 mmol)
in acetonitrile (4 cm3) was added to a solution of [RhCl3-
(CH3CN)3] (67 mg, 0.20 mmol) in acetonitrile (10 cm3). The
resulting solution was stirred rapidly for 10 s and then allowed
to stand for 4 days. The yellow crystals that formed were
collected, washed with benzene (4 cm3) and dried in vacuo to
yield [RhCl3(pptacn)] as a yellow microcrystalline solid (99 mg,
64%) (Found: C, 59.74; H, 5.49; N, 5.41. C39H45Cl3N3RhؒH2O
requires C, 59.82; H, 6.05; N, 5.37%). δH [(CD3)2SO] 1.92 (6 H,
m, 6 × H2), 2.45 (6 H, t, J(H2H3) 6.7 Hz, 6 × H3), 2.93–3.00
and 3.28–3.35 (12 H, m, 3 × NCH2CH2N), 3.75 (6 H, m,
6 × H1), 7.30–7.35 (9 H, m, aryl) and 7.40–7.45 (6 H, m, aryl);
δC [(CD3)2SO] 16.8 (C3), 21.4 (C2), 56.4 (NCH2CH2N), 59.5
(C1), 81.2 (C5), 89.7 (C4), 123.1 (C, aryl), 128.0 (aryl), 128.5
(aryl) and 131.3 (aryl); m/z (FAB) 844.1894 (M ϩ dmso)
(requires 844.1922) and 728.2049 (M Ϫ Cl) (requires 728.2046).
Crystal/refinement data. [H(4htacn)][BPh4]. C48H60BN3,
M = 689.8, monoclinic, space group P21/n (C52h, no. 14, variant),
a = 11.763(1), b = 26.994(3), c = 12.907(1) Å, β = 90.296(2)Њ,
V = 4098 Å3, Dc (Z = 4) = 1.118 g cmϪ3, µMo = 0.64 cmϪ1, speci-
men 0.38 × 0.35 × 0.14 mm, ‘T’min,max = 0.76, 0.93, Nt = 40295,
N = 10355 (Rint = 0.032), No = 7171, R = 0.048, Rw = 0.058,
|∆ρmax| = 0.31(3) e ÅϪ3
.
C(43,44) were modelled as disordered over two sets of sites,
occupancies refining to 0.657(5) and complement; associated
disordered hydrogen atom parameters were constrained at
estimated values, other (x, y, z, Uiso)H refined.
[Rh(CO)2(4htacn)][PF6]. A solution of 4htacn (0.141 g, 0.38
mmol) in acetone (5 cm3) was added to a solution of [Rh2(µ-
Cl)2(CO)4] (68.3 mg, 0.17 mmol) in acetone (5 cm3). A solution
of KPF6 (78 mg, 0.42 mmol) in acetone (5 cm3) was added and
the mixture stirred for 3 h, then filtered and diluted with ether
(10 cm3). The resulting precipitate was collected, washed with
ether (5 cm3) and dried in vacuo, to leave a dark off-green
powder (160 mg, 56%) (Found: C, 41.95; H, 5.30; N, 5.45.
C26H39F6N3O2PRhؒKCl requires C, 41.75; H, 5.26; N, 5.62%).
νmax/cmϪ1 (KBr) 2061 and 1982 (CO); δH [(CD3)2CO] 1.71 (9 H,
t, J(H4H6) 2.4 Hz, 9 × H6), 2.21 (6 H, m, 6 × H2), 2.25 (6 H,
m, 6 × H3), 3.15–3.22 (12H, m, 3 × NCH2CH2N) and 3.38
(6 H, m, 6 × H1); δC [(CD3)2CO] 3.7 (C6), 16.8 (C3), 28.5 (C2),
57.8 (NCH2CH2N), 65.9 (C1), 77.0 and 78.2 (C4 and C5) and
189.2 (d, J(RhC) 72 Hz, CO); m/z (FAB) 528.2069 (M Ϫ PF6)
(requires 528.2097) and 500.2147 (M Ϫ PF6 Ϫ CO) (requires
500.2148). Recrystallisation from benzene–hexane yielded
bright yellow, very thin plate-like crystals.
[Mo(CO)3(pptacn)]ؒ(CH3)2CO. C45H51MoN3O4, M =
21
2v
793.9, orthorhombic, space group Iba2 (C , no. 45), a =
25.283(5), b = 14.986(3), c = 21.337(4) Å, V = 8084(3) Å3, Dc
(Z = 8) = 1.304 g cmϪ3, µMo = 3.7 cmϪ1, specimen: 0.45 ×
0.30 × 0.15 mm, ‘T’min,max = 0.82, 0.90, Nt = 46797, N = 5249
(Rint = 0.048), No = 4249, R = 0.036, Rw = 0.045, |∆ρmax| =
1.10(4) e ÅϪ3. (x, y, z, Uiso)H constrained at estimates
throughout.
‘Friedel pairs’ were retained distinct, here (and in the next
determination), xabs refining to Ϫ0.02(5). The acetone solvent
comprises two moieties, one disposed on a crystallographic
2 axis, the other disordered about it.
[Rh(cod)(pptacn)][PF6]. C47H57F6N3PRh, M = 911.9,
orthorhombic, space group P212121 (D42, no. 19), a = 11.485(1),
b = 12.312(2), c = 30.350(4) Å, V = 4291 Å3, Dc (Z = 4) = 1.411 g
cmϪ3, µMo = 5.0 cmϪ1, specimen 0.25 × 0.15 × 0.04 mm, ‘T’min,max
= 0.82, 0.93, Nt = 50587, N = 6089 (Rint = 0.051), No = 5265,
R = 0.037, Rw = 0.040, |∆ρmax| = 0.82(7) e ÅϪ3, xabs = 0.02(3).
(x, y, z, Uiso)H constrained throughout.
[CuCl2(pptacn)]. A solution of CuCl2ؒ2H2O (8.6 mg, 50
µmol) in ethanol (1 cm3) was added to a solution of pptacn
(30 mg, 55 µmol) in ethanol (1 cm3) and stirred for 1 week. The
green precipitate that formed was collected, washed with
ethanol (2 cm3) and ether (5 cm3) and dried in vacuo to afford a
green powder (15 mg, 43%) (Found: C, 68.16; H, 6.33; N, 5.92.
C39H45Cl2CuN3 requires C, 67.86; H, 6.57; N, 6.09%); m/z
(FAB) 653.2590 (M Ϫ Cl) (requires 653.2598).
The fluorine atoms of the anion were modelled as disordered
over two sets of sites, occupancies refining to 0.85(1) and
complement.
[CuCl2(pptacn)]. C39H45Cl2CuN3, M = 690.3, monoclinic,
space group P21/n, a = 16.974(2), b = 12.331(2), c = 17.456(2) Å,
β = 101.945(2)Њ, V = 3575 Å3, Dc (Z = 4) = 1.282
g ,
cmϪ3
[Ni2(ꢀ-Cl)3(pptacn)2]Cl. A solution of NiCl2ؒ6H2O (10 mg,
43 µmol) in ethanol (1 cm3) was added to a solution of pptacn
(27 mg, 49 µmol) in ethanol (1 cm3) and stirred for 1 week. The
resulting mixture was concentrated in vacuo and then triturated
with ether (5 cm3). The solid was collected and dried to yield
[Ni2(µ-Cl)3(pptacn)2]Cl as a light green powder (21 mg, 71%).
Recrystallisation from CH2Cl2 gave green needles (Found: C,
67.84; H, 6.44; N, 5.98. C39H45Cl2N3Ni requires C, 68.34; H,
6.62; N, 6.13%). m/z (FAB) 1331.49978 (M Ϫ Cl) (requires
1331.49995).
µMo = 7.9 cmϪ1, specimen 0.50 × 0.45 × 0.30 mm, ‘T’min,max
= 0.66, 0.80, Nt = 34673, N = 8951 (Rint = 0.022), No = 7508,
R = 0.028, Rw = 0.039, |∆ρmax| = 0.44(3) e ÅϪ3, (x, y, z, Uiso)H
refined throughout.
CCDC reference number 186/2254.
See http://www.rsc.org/suppdata/dt/b0/b007324p/ for crystal-
lographic files in .cif format.
Acknowledgements
We thank the Australian Research Council for a small grant
(to M. V. B.) and an Australian Postgraduate Research Award
(to D. H. B.), Dr Alex Bilyk and Dr Mark Ogden for helpful
discussions, Dr George Koutsantonis for generous donations of
RhCl3ؒ3H2O and [Rh2(µ-Cl)2(cod)2], and Dr Anthony Reeder
for the mass spectra.
Structure determinations
Crystals of [H(4htacn)][BPh4] were grown from ethyl acetate–
methanol, of [Mo(CO)3(pptacn)] by slow cooling of an acetone
solution and of [Rh(cod)(pptacn)][PF6] by slow evaporation
of an acetone solution of the complex. Recrystallisation of
J. Chem. Soc., Dalton Trans., 2000, 4607–4616
4615

View Article Online
19 A. J. Blake, R. O. Gould, M. A. Halcrow and M. Schröder, J. Chem.
References
Soc., Dalton Trans., 1994, 2197.
1 M. V. Baker, D. H. Brown, B. W. Skelton and A. H. White, J. Chem.
Soc., Dalton Trans., 2000, 763.
2 P. Chaudhuri and K. Wieghardt, Prog. Inorg. Chem., 1987, 35, 329.
3 K. P. Wainwright, Coord. Chem. Rev., 1997, 166, 35.
4 H. Takalo, I. Hemmilä, T. Sutela and M. Latva, Helv. Chim. Acta,
1996, 79, 789.
5 M. Latva, H. Takalo, V.-M. Mukkala and J. Kankare, Inorg. Chim.
Acta, 1998, 267, 63.
6 D. Ellis, L. J. Farrugia and R. D. Peacock, Polyhedron, 1999, 18,
1229.
7 B. de Bruin, J. J. J. M. Donners, R. de Gelder, J. M. M. Smits and
A. W. Gal, Eur. J. Inorg. Chem., 1998, 401.
8 B. de Bruin, J. A. Brands, J. J. J. M. Donners, M. P. J. Donners,
R. de Gelder, J. M. M. Smits, A. W. Gal and A. L. Spek, Chem.
Eur. J., 1999, 5, 2921.
20 C. F. G. C. Geraldes, M. C. Alpoim, M. P. M. Marques, A. D.
Sherry and M. Singh, Inorg. Chem., 1985, 24, 3876.
21 P. R. Sharp, in Comprehensive Organometallic Chemistry II, eds.
E. W. Abel, F. G. A. Stone and G. Wilkinson, Pergamon, Oxford,
1995, vol. 8, p. 210.
22 M. P. Lamata, E. San José, D. Carmona, F. J. Lahoz, R. Atencio and
L. A. Oro, Organometallics, 1996, 15, 4852.
23 K. Wieghardt, P. Chaudhuri, B. Nuber and J. Weiss, Inorg. Chem.,
1982, 21, 3086.
24 M. Südfeld and W. S. Sheldrick, Inorg. Chim. Acta, 2000, 304,
78.
25 D. Hanke, K. Wieghardt, B. Nuber, R.-S. Lu, R. K. McMullan,
T. F. Koetzle and R. Bau, Inorg. Chem., 1993, 32, 4300.
26 D. G. Evans and J. C. A. Boeyens, Acta Crystallogr., Sect. B,, 1990,
46, 524.
9 L. Wang, C. Wang, R. Bau and T. C. Flood, Organometallics, 1996,
15, 491.
10 T. Beissel, B. S. P. C. Della Vedova, K. Wieghardt and R. Boese,
Inorg. Chem., 1990, 29, 1736.
27 K. Wieghardt, S. Brodka, K. Peters, E. M. Peters and A. Simon,
Z. Naturforsch., Teil B, 1987, 42, 279.
28 H. Werner, P. Schwab, A. Heinemann and P. Steinert, J. Organomet.
Chem., 1995, 496, 207.
11 G. Haselhorst, S. Stoetzel, A. Strassburger, W. Walz, K. Wieghardt
and B. Nuber, J. Chem. Soc., Dalton Trans., 1993, 83.
12 J. L. Sessler, J. W. Sibert and V. Lynch, Inorg. Chim. Acta, 1994, 216,
89.
13 J.-M. Vincent, A. Rabion, V. K. Yachandra and R. H. Fish, Angew.
Chem., Int. Ed. Engl., 1997, 36, 2346.
14 S. Delagrange and F. Nepveu, Tetrahedron Lett., 1999, 40, 4989.
15 H. Higashimura, M. Kubota, A. Shiga, K. Fujisawa, Y. Moro-oka,
H. Uyama and S. Kobayashi, Macromolecules, 2000, 33, 1986.
16 A. Buttafava, L. Fabbrizzi, A. Perotti, A. Poggi, G. Poli and
B. Seghi, Inorg. Chem., 1986, 25, 1456.
17 M. V. Baker, D. H. Brown, B. W. Skelton and A. H. White, Aust. J.
Chem., in press.
29 N. L. Armanasco, M. V. Baker, M. R. North, B. W. Skelton and
A. H. White, J. Chem. Soc., Dalton Trans., 1997, 1363.
30 G. Büchi and H. Wüest, J. Org. Chem., 1979, 44, 546.
31 B. Catsikis and M. L. Good, Inorg. Nucl. Chem. Lett., 1968, 4,
529.
32 J. A. McCleverty and G. Wilkinson, Inorg. Synth., 1990, 28, 84.
33 G. J. Kubas and L. S. van der Sluys, Inorg. Synth., 1990, 28, 29.
34 L. Bretherick, Handbook of Reactive Chemical Hazards,
Butterworths, London, 2nd edn., 1979, p. 1135.
35 W. E. Truce and T. C. Klingler, J. Org. Chem., 1970, 35, 1834.
36 B. A. Anderson, J. Bao, T. A. Brandvold, C. A. Challener, W. D.
Wulff, Y.-C. Xu and A. L. Rheingold, J. Am. Chem. Soc., 1993, 115,
10671.
18 O. Krampe, C.-E. Song and W. Kläui, Organometallics, 1993, 12,
37 S. R. Hall, G. S. D. King and J. M. Stewart, The Xtal 3.4 User’s
Manual, Lamb, Perth, 1995.
4949.
4616
J. Chem. Soc., Dalton Trans., 2000, 4607–4616