Triangular tricopper(i) clusters supported by donor-substituted triazacyclohexanes

View Article Online / Journal Homepage / Table of Contents for this issue

PAPER

www.rsc.org/dalton | Dalton Transactions

Triangular tricopper(I) clusters supported by donor-substituted

triazacyclohexanes†

Randolf D. Ko¨hn,* Lorena Toma´s Laudo, Zhida Pan,‡ Fredy Speiser§ and Gabriele Kociok-Ko¨hn

Received 30th October 2008, Accepted 24th February 2009

First published as an Advance Article on the web 2nd April 2009

DOI: 10.1039/b819268e

Triazacyclohexanes (R₃TAC, 1a–i) with pyridyl or thioether functionalities (R) in the N-substituents

react with three equivalents of CuX (X = Cl (2), Br (3) or I (4)) in MeCN to give the triangular

tri-copper clusters [R₃TAC(CuX)₃] (R = 2-pyridylmethyl (2a, 3a), 5-^tbutyl-2-pyridyl (3b),

2-(3-phenylpropylthio)ethyl (3c), 2-(2-ethyl-butylthio)ethyl (3d), 2-(4-heptylthio)ethyl (2e, 3e),

2-(1-heptylthio)ethyl (3f), 2-(2,4,6-trimethyl-benzylthio)ethyl (3g), 2-(o-methyl-benzylthio)ethyl (3h) and

2-(o-ﬂuoro-benzylthio)ethyl (2i, 3i, 4i)). The thioether complexes are stable towards air and water. The

bromide bridge in the clusters can be replaced by chloride (2c, e, f, i) or iodide (4c, e, f, i) by the reaction

of a dichloromethane solution of the cluster with aqueous NaI or AgCl, respectively. Crystal structures

of 2a, 3a, 3b, 2e, 3h and 4i show triangular halide-bridged Cu₃clusters capped by the triazacyclohexane

and stabilised by the coordination of one pyridyl or thioether arm to each copper atom. DFT

calculations conﬁrm the NMR assignments and reveal the electronic structure of the copper triangle.

Introduction

Copper-containing oxidases have gained importance during the

last decade with the elucidation of some of their crystal structures.

Trinuclear copper centres are present in many of the blue copper

oxidases such as ascorbate oxidase, ceruplasmin and laccase. The

Fig. 1 Previously reported triangular tri-copper clusters.³

trinuclear copper cluster is the active site of these enzymes, where

O₂is reduced to water in a four electron process. The active site is

located next to a fourth copper atom, a T1 or blue copper, which

mediates the electron transfer from the oxidised substrate to the

trinuclear copper cluster site.¹

Results and discussion

Synthesis

In our research, we have focused on reproducing well-deﬁned

triangular tri-copper complexes. We have previously shown that

a triazacyclohexane can act as a bridging ligand between copper

atoms.²The highly sensitive and insoluble complexes could not

be studied further. Kickelbick et al.³have described the two

complexes shown in Fig. 1 where the triangular array of three

Cu(I) atoms mutually bridged by bromides is stabilised by the

coordination of an ether or amine functionality at the N-

substituent of the triazacyclohexane. Both of these complexes

were still not soluble without decomposition so that no solution

chemistry or properties were reported. In this paper, we report

the syntheses and characterisation of more stable and soluble

complexes with pyridyl or thioether donor groups.

The pyridylmethyl and alkylthioethyl substituted triazacyclo-

hexanes were prepared from the corresponding amines and

formaldehyde as shown in Scheme 1. 1a, previously described

as oil, was found to crystallise after long storage and was

characterised by X-ray crystallography as shown in Fig. 2. The

other triazacyclohexanes were isolated as viscous oils.

Department of Chemistry, University of Bath, Bath, UK BA2 7AY.

E-mail: r.d.kohn@bath.ac.uk

† Electronic supplementary information (ESI) available: Plots for struc-

tures not shown in the article, DFT results including a typical input ﬁle,

additional orbital contour plots and xyz ﬁles for optimised structures and

calculated ¹H and ¹³C NMR shifts. CCDC reference numbers 707388–

707395. For ESI and crystallographic data in CIF or other electronic

format see DOI: 10.1039/b819268e

‡ Current address: Henkel (China) Company Ltd., Henkel Asia-Paciﬁc

and China Headquarters, No. 928 Zhangheng Road, Pudong, Shanghai,

P.R. China, 201203.

Scheme 1 Synthesis of the thioether ligands 1c-i.

Copper(I) halides react readily in acetonitrile in a 3 : 1 ratio

with 1 to give the triangular tri-copper complexes 2–4 (Scheme 2).

Best results were obtained for the bromides 3 where the product

precipitates or even crystallises shortly after the addition of 1 to

the acetonitrile solution of CuBr. The chlorides 2 and iodides 4

§ Current address: Badstr. 8, D-88131 Lindau, Germany.

4556 | Dalton Trans., 2009, 4556–4568

This journal is

The Royal Society of Chemistry 2009

©

View Article Online

NMR and ESI-MS characterisation

All complexes are soluble enough in polar solvents like

MeCN or MeNO₂to be characterised by NMR spec-

t

troscopy. The Bu-pyridylmethyl-subtituted complex 2b and the

thioether-substituted complexes are also soluble in halogenated

hydrocarbons (chloroform, ﬂuorobenzene, dichloromethane,

dichloroethane or o-dichlorobenzene). Apart from small NMR

shifts upon coordination relative to the free ligand, the most

characteristic change is observed for the ring protons: the broad

singlet for the equatorial and axial proton in free triazacyclo-

hexane becomes two separate doublets as is generally observed

for triazacyclohexane complexes.⁴Assignment of the two ring

positions was possible by 1D NOESY and, in some cases, ROESY

e.g. irradiation at the 6-py position in 3b (8.6 ppm) in PhNO₂

led to NOE enhancement of signals at 4.0 and 3.1 ppm in

a 2 : 1 ratio as expected for the closer equatorial and axial

Fig. 2 Thermal ellipsoid (50%) plot of ligand 1a. Hydrogen atoms

omitted for clarity.

can also be obtained from the bromides 3 by halide exchange

in dichloromethane–water with in situ-generated AgCl or NaI,

respectively. The pyridylmethyl complexes are intense yellow

while all thioether complexes are colourless. All complexes gave

satisfactory elemental analyses with the exception of 2c, f which

may contain traces of ﬁne particles of Ag or AgCl that could not

be removed. However, ESI-MS and NMR showed the purity of

the solutions even in these cases.

The donor substituents on the triazacyclohexane ligands were

varied to improve the solubility of the clusters. A tert-butyl group

in the pyridylmethyl-substituted complex 3b improved the solubil-

ity relative to 3a at the cost of substantial synthetic effort. Variation

of the thioether substituent was much easier, especially for longer

alkyl chain substituents in 1e, f, which gave complexes of good

solubility (>0.1 M) in chlorinated hydrocarbon solvents. Among

the aromatic thioether substituents tested, ortho-ﬂuorobenzyl in

1i gave the best solubility for the complexes with the additional

advantage of a ¹⁹F NMR handle for future investigations. The use

of wet solvents containing several equivalents of water per cluster

was found to further improve the solubility.

1

position, respectively, based on the structure shown later. J_CH

coupling constants were determined for a few complexes. While,

in most cases, the resolution did not allow the extraction of

different coupling constants for the equatorial and axial ring

C–H bonds, in the case of the highly soluble 4e, a doublet of

doublets was observed with 141 and 153 Hz coupling. Weak ¹³C

1

satellites at about 153 Hz around the well-isolated H signal at

4 ppm allowed an assignment of the larger coupling constant

to the equatorial hydrogen. DFT optimised C–H bonds also

˚

show this difference (1.098 A (eq) and 1.113 A (ax)) and the

weaker axial C–H bond indicates involvement in the Cu–N_TAC

interaction. Surprisingly, and especially for the pyridylmethyl-

substituted complexes, the two ring proton signals were difﬁcult to

observe, being broadened and shifted depending upon the solvent,

temperature and concentration, and even coalescing to a single,

broad peak at elevated temperatures. 1D ROESY spectroscopy

conﬁrmed that chemical exchange was occurring between the

two ring positions. Due to the large size of the complexes, 1D

NOESY also gave strong correlation signals but of the same sign as

correlation due to dipolar NOE. The mechanism of this surprising

fast exchange of the equatorial and axial ring positions is currently

under detailed experimental and computational investigation and

All complexes tolerate air in the solid state, at least for a short

time, and are unaffected by water. Solutions of the pyridylmethyl-

substituted complexes 2a, 3a and 3b are readily oxidised to

green solutions on contact with air while all thioether substituted

complexes tolerate air, even in solution, for extended periods

before turning green or yellow (4).

Scheme 2 Synthesis of the complexes.

The Royal Society of Chemistry 2009

This journal is

Dalton Trans., 2009, 4556–4568 | 4557

©

View Article Online

will be published later. An analogous process in the thioether

substituted complexes occurs much slower or not at all and clean

1

pairs of doublets for the ring protons are observed in H NMR

spectra in halogenated solvents. A variable-temperature study on

3f showed some evidence of a dynamic process within the thioether

arm but no exchange of the ring hydrogen positions. This slow

conformational exchange also leads to complex ¹H NMR patterns

for the S–CH₂CH₂–N bridge in the thioether complexes.

The Cu₃cluster complexes could be characterised by electro-

spray mass spectrometry in acetonitrile or ﬂuorobenzene solution.

All complexes [LCu₃X₃] show a common pattern of the highest

mass signal for [LCu₃X₂]⁺(thus [M - X]) and two major signals

for the loss of one and two CuX to give [LCu₂X]⁺and [LCu]⁺.

The isotope pattern and high-resolution mass conﬁrm these

assignments. A typical spectrum for 3f is shown in Fig. 3. No ion

for the complete cluster (or proton or sodium adduct) was found

for samples in acetonitrile solution. A solution of 3b in the less

polar ﬂuorobenzene gave a signal for the complete cluster at about

0.1% of the intensity of the major ion [LCu₂X]⁺([LCu₃X₂]⁺also ob-

served). Interestingly, in several cases of [LCu₃Br₃] complexes, ESI-

MS also showed ions for fragments containing chloride in place of

a bromide. This chloride must come from a facile substitution from

ubiquitous chloride in the MS instrument. No bromide containing

ions were detected in the mass spectra of the products of halide

exchange reactions. This proves there was complete exchange.

Fig. 5 Thermal ellipsoid (50%) plot of 3b. Hydrogen atoms omitted for

clarity.

Fig. 6 Thermal ellipsoid (20%) plot of 2e. Hydrogen atoms omitted for

clarity.

Fig. 3 ESI-MS of 3h.

Description of crystal structure

Crystals suitable for X-ray crystallography were grown for 2a, 3a,

3b, 2e, 3h and 4i. The crystals of 3h were very poor but still gave a

reasonable structure to allow comparison to the other structures, at

least for bond distances to copper. An even poorer structure for 3i

conﬁrms the connectivity and a coordination environment around

copper similar to 3h or 4i. Sample molecular structures are shown

in Fig. 4–7. Two different crystal types were obtained for 3a—one

Fig. 7 Thermal ellipsoid (50%) plot of 4i. Hydrogen atoms omitted for

clarity.

solvent-free structure 3a-A and a second, 3a-B, with MeCN in the

lattice and two independent molecules of different conformations.

The crystal structure for 2a was isostructural to the latter with

two conformers. The two conformers are shown in Fig. 8. One is

nearly C₃symmetric (sym) with the three pyridyl arms pointed in

a trikelion fashion and all three halide bridges at a similar distance

from the Cu₃plane. The other (asym) does not have this symmetry

and has two pyridyl groups pointed at each other and one of the

Fig. 4 Thermal ellipsoid (50%) plot of the asym conformer found for

3a-A. Hydrogen atoms omitted for clarity.

4558 | Dalton Trans., 2009, 4556–4568

This journal is

The Royal Society of Chemistry 2009

©

View Article Online

˚

The copper–copper distances are in the range of 2.8–2.9 A for

˚

the pyridyl complexes and 2.9–3.0 A for the thioether complexes

and indicate a trend of increasing distance with stronger donors

compared to the weaker donor complexes of Kickelbick³(Cu–Cu

˚

distance 2.79 A with amine donor and 2.67 A with ether). These

distances are in the range observed for weak “cuprophilic” closed-

shell interactions.⁵

The average distances from copper to the donor atoms shown in

Table 1 are not unusual compared to other copper complexes. No-

ticeable is the trend for the bond distance to the triazacyclohexane

˚

nitrogen atoms. The Cu–N distances are longer than 2.21 A for

pyridyl complexes (as for Kickelbick’s amine and ether complexes)

˚

and shorter than 2.20 A for thioether complexes.

Fig. 8 Two conformers found in the triangular clusters shown in 2a:

nearly C₃symmetric (left) with all halide bridges at a similar distance from

the Cu₃plane (sym) and asymmetric (right) with the halide labelled “Br”

about twice as far below the Cu₃plane than the other two halide bridges

(asym).

DFT calculations

The pyridylmethyl-substituted complexes 2a and 3a and

methylthioethyl-substituted complexes 2*, 3* and 4* were in-

vestigated by DFT calculations on the RI-BP86/TZVP level

with ZORA and COSMO solvent corrections using the ORCA⁶

programme. Reasonable agreement of the optimised structure

with the crystal structures (bond distances to copper within

halide bridges about twice as far below the Cu₃plane than the other

two halide bridges. A similar asymmetric structure is found in the

other crystal form of 3a and in 3b. This asymmetry in the halide

bridges is much less pronounced in the thioether complexes as

well as in the amine and ether complexes of Kickelbick. Our DFT

calculations described below found minima with an even larger

asymmetry in the halide bridge (for both pyridyl and thioether

substituents) as well as another minimum with symmetrical halide

bridges (for the thioether substituent at almost the same energy).

Thus, bending of the halide bridge along the Cu–Cu axis has

a rather soft energy potential and can be inﬂuenced by crystal

packing effects. This structural lability of the halide bridge may

also aid the facile substitution of the bridge by another halide.

˚

0.1 A of X-ray data) and of the calculated NMR shifts with

the observed shifts (except for 4* where the basis set used for

iodine was not suitable for property calculations) were found. As

expected for weak dispersion interactions, our DFT calculations

give the largest deviation from the experiment in the Cu–Cu

˚

contacts which are optimised at about 2.6 A. The difference

in the triazacyclohexane to copper bond distances between the

complexes with the soft thioether substituent and the harder

substituents is well reproduced by the DFT calculations (about

˚

2.15 A for thioether and 2.20 A for pyridyl complexes). Calculation

˚

Table 1 Average bond lengths (standard deviation of average) and elevation D of the halide bridge X below the Cu₃plane in A. Donor = N(pyridyl) or

S(thioether). For comparison the corresponding distances from DFT optimisation are given in italics

Cu–Cu

Cu–X

Cu–N_TAC

Cu–Donor

D

2a

3a

sym

sym, DFT

asym

2.87(8)

2.62

2.84(6)

2.57

2.63

2.31(7)

2.36

2.31(4)

2.41

2.38

2.25(1)

2.21

2.27(2)

2.18

2.22

2.07(2)

2.03

2.08(4)

2.04

2.05

1.21, 0.96, 0.89

1.87, 0.95, 0.72

1.55, 0.76, 0.70

1.68, 1.48, 0.90

1.78, 1.06, 0.86

1.36, 1.08, 0.98

1.73, 0.82, 0.80

1.59, 0.87, 0.81

2.03, 0.76, 0.71

1.95, 1.01, 0.87

1.53, 1.02, 0.67

1.04, 0.95, 0.80

1.68, 1.01, 0.59

1.76, 0.79, 0.78

1.19, 1.17, 1.10

1.05, 0.84, 0.78

1.98, 0.68, 0.63

1.83, 0.90, 0.88

1.23, 1.22, 1.17

1.22, 1.05, 1.05

2.07, 0.66, 0.62

0.77, 0.77, 0.73

1.19, 1.05, 0.73

1.03, 0.86, 0.75

asym, DFT

asym, DFT, no VDW

sym

2.86(7)

2.80(7)

2.87(10)

2.56

2.44(7)

2.44(3)

2.49

2.23(1)

2.24(1)

2.22(1)

2.20

2.10(1)

2.10(3)

2.11(4)

2.05

asym

asym, other cell

asym, DFT

asym, DFT, no VDW

asym

2.59

2.49

2.21

2.05

3b

2e

2*

2.89(10)

2.98(10)

2.63

2.44(4)

2.29(3)

2.36

2.23(1)

2.19(1)

2.16

2.11(2)

2.38(1)

2.38

DFT

DFT, no VDW

2.65

2.67

2.88(8)

2.61

2.64

2.66

2.91(3)

2.57

2.37

2.36

2.41(3)

2.48

2.49

2.47

2.58(3)

2.54

2.52

2.17

2.16

2.15(8)

2.16

2.17

2.16

2.19(1)

2.15

2.20

2.39

2.40

2.42(1)

2.36

2.41

2.42

2.41(1)

2.38

2.45

2*-Cl-sym^a

3h

3*

DFT

DFT, no VDW

3*-Br-sym^a

4i

4*

DFT

DFT, no VDW

R = NMe₂, X = Br

R = OMe, X = Br

Ref. 3

2.79(3)

2.66(1)

2.43(2)

2.40(2)

2.23(2)

2.20(1)

2.27(1)

2.36(1)

^aOptimised minimum for symmetrical halide bridges (all halides at similar distances below Cu₃plane).

This journal is

The Royal Society of Chemistry 2009

Dalton Trans., 2009, 4556–4568 | 4559

©

View Article Online

of the NMR parameters for a structure with all non-hydrogen

atoms kept in the positions found in the best crystal structure

(3a) gave an even better agreement with experiment. The ¹H

NMR shifts for the ring positions conﬁrmed the assigned axial

and equatorial positions. Single point calculations for 2a and

2* with the B3LYP functional gave NMR shifts even closer to

the experiment at the cost of much longer computation times.

Calculated structural and NMR data are listed alongside the

experimental values for comparison in Table 1 and in the ESI.† In

all cases, HOMO and HOMO-1 are close in energy and mainly

copper-centered orbitals (Fig. 9). The HOMO energy decreases

from the chloride to the iodide complexes (2a (Cl): -3.48, 2b

(Br): -3.54; 2* (Cl): -3.61, 3* (Br): -3.75 and 4* (I): -3.90 eV)

and is much lower for thioether complexes relative to pyridyl

complexes. Thus, pyridyl-substituted chloride complexes should

be easiest, and thioether-substituted iodide complexes hardest,

to oxidise as observed for their air-sensitivity. For the thioether

complexes 2* and 3*, the LUMO was located at the Cu₃X₃ring

with a signiﬁcant contribution in the centre of the Cu₃triangle

(Fig. 10). The position of the LUMO indicates that a nucleophile

may attack at the centre of the copper triangle as proposed for

the initial attack of O₂in multi-copper oxidases.¹The pyridyl

complexes 2a and 3a have unoccupied orbitals of similar shape and

energy but the LUMOs are among two sets of three orbitals of p*

symmetry centered in the pyridine rings (Fig. 11). This leads to a

much smaller HOMO–LUMO gap and explains the yellow colour

of pyridyl complexes versus the colourless thioether complexes.

Fig. 11 Contour plot¹⁹of the LUMO of 2a (-2.26 eV).

experimental trends. The Mulliken charge⁸on the copper atoms

also drops signiﬁcantly from pyridyl (Cl: +0.22, Br: +0.29) to

thioether (Cl: +0.17, Br: +0.20) complexes. Calculation at the X-

ray atom positions for 3a and at the B3LYP level gives smaller

bond orders (0.14 and 0.07) and slightly larger copper charges

(0.28 and 0.34, respectively) without changes in the trends.

Experimental section

General methods and instrumentation

All manipulations were carried out using standard Schlenk line or

dry-box techniques under an atmosphere of argon or of dinitrogen.

Solvents were reﬂuxed over the appropriate drying agent, distilled

and stored in Teﬂon valve ﬂasks in the dry-box. NMR samples

were prepared under dinitrogen in the dry-box. ¹H and ¹³C NMR

spectra were recorded on Bruker Avance 300 MHz, 400 MHz or

500 MHz spectrometers at 298 K and assignments were conﬁrmed

by COSY, HSQC, HMBC or NOESY (ax. or eq. ring CH₂)

spectra. Residual protio solvent was used as reference for ¹H

(CDCl₃, 7.26 ppm; D₂O, 4.85 ppm) and ¹³C spectra (CDCl₃,

77.16 ppm) or the solvent signal for non-deuterated NMR spectra

(neat solvent peak referenced vs. TMS stated). Values are quoted

in ppm. Coupling constants are quoted in Hz. C–H coupling

constants are stated where they were obtained by non-decoupled

¹³C or HMBC spectra. The concentration of (mostly saturated)

solutions in the NMR tube was obtained by integration relative to

the solvent signal in non-deuterated NMR spectra and estimated

by integration against the residual CHCl₃in CDCl₃(0.1% H)

and listed with the NMR solvent. High-resolution electrospray

mass spectra were obtained on a Bruker TOF instrument. Isotope

patterns match the assignments and calculated exact m/z values

are given for the most intense ion. Elemental analyses were

carried out by Mr Alan Carver (University of Bath) on an Exeter

Analytical Instruments CE-440 Elemental Analyser.

Fig. 9 Contour plots¹⁹of the HOMO of 2* (-3.63 eV) (left) and

HOMO-1 of 2* (-3.67 eV) (right).

Starting materials

The triazacyclohexanes 1 were prepared analogous to pyridyl-

methyl-triazacyclohexane 1a⁹from the corresponding amine and

paraformaldehyde or formalin solution. The previously described

oily 1a crystallised after a year and could be washed with hexane

and isolated as a solid. Its structure was conﬁrmed by X-ray

analysis.

Fig. 10 Contour plots¹⁹of the LUMO of 2* (-1.27 eV) (views from the

side and from below).

2-(4-tert-Butyl)pyridylmethylamine. 2-Cyano-4-tert-butylpyr-

idine (prepared according to Shuman et al.¹⁰) (14.74 g, 92 mmol)

was dissolved in 500 mL of THF and added to 92 mL of 2 M

BH₃·THF in THF (184 mmol). The mixture was heated to reﬂux

overnight and worked up to yield 30% of the amine.

The copper–copper interaction as indicated by the optimised

distances and Mayer bond orders⁷(2a/3a 0.21, 2*/3* 0.17) show

little difference between chloride and bromide complexes but a

decrease from pyridyl to thioether complexes and conﬁrms the

4560 | Dalton Trans., 2009, 4556–4568

This journal is

The Royal Society of Chemistry 2009

©

View Article Online

¹H NMR (CDCl₃): 8.37 d (1H, J = 5.3 Hz, 6-py), 7.17 d (1H,

J = 1.9 Hz, 3-py), 7.06 dd (1H, J = 5.3 and 1.9 Hz, 5-py), 3.87 s

1

¹³C-{ H} NMR (CDCl₃): 41.57 (CH₂–NH₂), 41.30 (CH), 37.03

(S–CH₂), 36.34 (CH₂–S), 25.33 (CH₃–CH₂), 11.20 (CH₃).

1

(2H, py-CH₂), 1.97 br (2H, NH₂), 1.22 s (9H, ^tBu). ¹³C-{ H} NMR

Tris(2-(2-ethylbutylthio)ethyl)triazacyclohexane

(1d). 2-(2-

(CDCl₃): 161.4 and 160.4 (2- and 4-py), 148.8 (6-py), 118.7 and

117.9 (3- and 5-py), 47.8 (py-CH₂), 34.4 (C), 30.3 (CH₃).

Ethylbutylthio)ethylamine (700 mg, 4.34 mmol) was dissolved in

toluene and reacted with p-formaldehyde (130 mg, 4.34 mmol)

for one hour. The solvent was distilled off in vacuo affording a

colourless oil in 68% yield.

Tris(2-(4-tert-butyl)pyridylmethyl)triazacyclohexane (1b). 2-

(4-tert-Butyl)pyridylmethylamine (1.30 g, 7.91 mmol) was dis-

solved in toluene (20 mL). Parafomaldehyde (237 mg, 7.91 mmol)

was added and the mixture was stirred overnight. Then the solvent

was distilled off and the residual yellow oil dried in vacuo to give

1.20 g yield (86%).

¹H NMR (CDCl₃): 3.41 br (6H, ring CH₂), 2.66 m (6H, CH₂–

NH₂, J = 5.9), 2.58 m (6H, S–CH₂), 2.48 d (6H, J = 5, CH₂–S),

1.35–1.42 m (15H, CH₃–CH₂and CH), 0.8 t (18H, J = 6.9, CH₃).

1

¹³C-{ H} NMR (CDCl₃): 74.26 (ring CH₂), 52.93 (CH₂–N), 41.01

¹H NMR (CDCl₃): 8.33 dd (3H, J = 5.3 and 0.7 Hz, 6-py),

7.37 d (3H, J = 1.9 Hz, 3-py), 7.03 dd (3H, J = 5.3 and 1.9 Hz,

5-py), 3.83 s (6H, py-CH₂), 3.54 br (6H, ring CH₂), 1.19 s (27H,

(CH), 36.69 (S–CH₂), 31.16 (CH₂–S), 25.18 (CH₃–CH₂), 10.97

(CH₃).

1

2-(4-Heptylthio)ethylamine. Prepared analogously to the pre-

cursor of 1d from 2-mercaptoethylamine hydrochloride (4 g,

35.2 mmol), KOH (4 g, 70 mmol) and 4-bromoheptane (6.5 mL,

35 mmol) as a colourless oil in 75% yield.

^tBu). ¹³C-{ H} NMR (CDCl₃): 160.4 (2-py), 158.5 (4-py), 149.0

(6-py), 119.3 (3-py), 119.0 (5-py), 74.1 (CH₂, ring), 58.9 (py-CH₂),

34.6 (C), 30.5 (CH₃).

¹H NMR (CDCl₃): 2.85 t (2H, J = 6.4, CH₂–NH₂), 2.59 t

2-(3-Phenylpropylthio)ethylamine. Sodium hydroxide (1.2 g,

30 mmol) was dissolved in MeOH (50 mL) followed by the addition

of mercaptoethylamine hydrochloride (1.71 g, 15 mmol). After

30 min, 1-bromo-3-phenylpropane (3 g, 15 mmol) was run into

the solution. The reaction mixture was vigorously stirred under

nitrogen overnight. The solution was ﬁltered and the solvent

removed under vacuum. The remaining product was dissolved

in ether, ﬁltered and the solvent evaporated affording a yellow oil

(79% yield).

(2H, J = 6.4, S–CH₂), 2.58 m (1H, CH), 1.43–1.55 m (10H, CH–

CH₂CH₂–CH₃and NH₂), 0.91 t (6H, J = 7.2, CH₃). ¹³C-{ H}

1

NMR (CDCl₃): 45.3 (CH), 41.6 (CH₂–NH₂), 37.2 (CH₂CH₂CH),

34.6 (CH₂–S), 19.9 (CH₃–CH₂), 13.9 (CH₃).

¹H NMR (CDCl₃): 3.4 br (6H, ring CH₂), 2.65 m (6H, CH₂–

N), 2.58 m (9H, CH–S–CH₂), 1.4–1.55 m (24H, CH₃–CH₂CH₂),

0.90t (18H, J = 7.1, CH₃) ppm. ¹³C-{ H} NMR (CDCl₃): 74.27

1

(ring CH₂), 53.19 (CH), 45.75 (CH₂–N), 37.30 (CH₂–S), 28.66

(CH₂CH₂CH), 20.10 (CH₃–CH₂), 14.18 (CH₃) ppm.

¹H NMR (CDCl₃with some CD₃OD): 7.3 m and 7.15 (5H,

Ph), 2.9 t (2H, J = 6.01, CH₂–NH₂), 2.85 t (2H, J = 7.72, S–

CH₂), 2.6 t (2H, J = 6.05, CH₂–S), 2.5 t (2H, J = 7.72, Ph–CH₂),

¹H NMR (7.0 w% in DCM): 3.33 br (6H, ring CH₂), 2.58 m

(6H, CH₂–N), 2.56 m (9H, S–CH₂and CH), 1.4–1.5 m (24H,

1

1.9 m (2H, J = 7.54, Ph–CH₂–CH₂). ¹³C-{ H} NMR (CDCl₃

CH₃–CH₂C_H2), 0.88 t (18H, J = 7.1, CH₃) ppm. ¹³C-{ H}

1

with some CD₃OD): 141.33, 128.41, 128.03, 125.93 (i, m, o, p-Ph),

40.37 (CH₂–N), 35.20 (S–CH₂), 34.69 (PhCH₂CH₂CH₂–S), 31.15

(PhCH₂CH₂CH₂–S), 30.96 (PhCH₂CH₂CH₂–S).

NMR (CDCl₃): 74.0 (ring CH₂), 52.8 (CH₂–N), 45.6(CH), 37.1

(CH₂CH₂CH), 28.4 (CH₂S), 20.0 (CH₃–CH₂), 13.9 (CH₃) ppm.

Tris(2-(4-heptylthio)ethyl)triazacyclohexane

(1e). Prepared

Tris(2-(3-phenylpropylthio)ethyl)triazacyclohexane (1c). 2-(3-

Phenylpropylthio)ethylamine (700 mg, 3.57 mmol) was dissolved

in toluene and one equivalent of paraformaldehyde (111 mg,

3.57 mmol) was added. The reaction ran overnight and the solvent

was removed by distillation. The remaining oil was dried under

reduced pressure, redissolved in ether, ﬁltered and the solvent

removed under vacuum yielding 85% of yellow oil.

analogously to 1d from the amine (1.58 g, 9 mmol) and

paraformaldehyde (271 mg, 9 mmol) in toluene as oil in 87%

yield.

2-(1-Heptylthio)ethylamine. Prepared analogously to the pre-

cursor of 1d from 2-mercaptoethylamine hydrochloride (4 g,

35.2 mmol), NaOH (2.82 g, 70.4 mmol) and 1-bromoheptane

(6.5 mL, 35 mmol) as a yellow oil in 81% yield.

¹H NMR (CDCl₃): 7.2 and 7.1 (15H, Ph), 3.3 br (6H, ring CH₂),

2.65 t (6H, S–CH₂), 2.55 m (12H, CH₂–S and CH₂–N), 2.45 t (6H,

¹H NMR (CDCl₃): 2.8 t (2H, J = 6.40, CH₂–N), 2.5 t (2H, J =

6.40, S–CH₂), 2.4 t (2H, J = 7.35, CH₂–S), 1.85 s (2H, NH₂),

1.5 q (2H, J = 7.35, CH₂–CH₂–S), 1.2–1.3 (8H, set of signals for

1

J = 7.54, Ph–CH₂), 1.85 m (6H, Ph–CH₂–CH₂). ¹³C-{ H} NMR

(CDCl₃): 141.58, 128.58, 128.30, 126.04 (i, m, o, p-Ph), 74.14 (ring

CH₂), 52.75 (CH₂–N), 35.19 (S–CH₂), 31.75 (PhCH₂CH₂CH₂–S),

31.32 (PhCH₂CH₂CH₂–S), 30.36 (PhCH₂CH₂CH₂–S) ppm.

1

CH₂–CH₂–CH₂–CH₂), 0.9 t (3H, J = 7.16 CH₃). ¹³C-{ H} NMR

(CDCl₃): 41.02 (CH₂–NH₂), 36.15 (CH₂–S), 32.14 (CH₂), 31.83

(CH₂), 29.78 (CH₂), 29.19 (CH₂), 28.83 (CH₂), 22.58 (CH₂), 14.02

(CH₃).

2-(2-Ethylbutylthio)ethylamine. To

a solution of NaOH

(614 mg, 15.3 mmol) in 30 mL of methanol, 2-mercaptoethylamine

hydrochloride (0.81 g, 7.14 mmol) was added. The solution was

stirred under nitrogen, 1-bromo-2-ethylbutane (1 mL, 1.179 g,

7.14 mmol) added and left stirring overnight. Filtration and

solvent removal in vacuo afforded a colourless oil that was

dissolved in diethyl ether, ﬁltered, and dried under vacuum. A

colourless oil was obtained in a 61% yield.

¹H NMR (CDCl₃): 2.79 t (2H, J = 6.3, CH₂–NH₂), 2.51 t (2H,

J = 6.3, S–CH₂), 2.40 d (2H, J = 5, CH₂–S), 1.30–1.35 (5H, CH₃–

CH₂and CH), 1.27 s (2H, NH₂), 0.86 t (6H, J = 7.0 Hz, CH₃).

Tris(2-(1-heptylthio)ethyl)triazacyclohexane

(1f). Prepared

analogously to 1d from the amine (6.3 g, 36 mmol) and

paraformaldehyde (1.08 g, 36 mmol) in toluene (50 mL). The

residue was taken up with Et₂O, decanted from insoluble material

and isolated by solvent removal in vacuo as an oil in 65% yield.

¹H NMR (CDCl₃): 3.38 br (6H, ring CH₂), 3.03 m (6H, CH₂–N),

2.58 m (6H, S–CH₂), 2.48 t (6H, J = 7.54, CH₂–S), 1.53 q (6H, J =

7.54, CH₂–CH₂S), 1.25–1.45 (18H, set of signals for CH₃–CH₂–

1

CH₂–CH₂), 0.83 t (9H, J = 7.35, CH₃). ¹³C-{ H} NMR (CDCl₃):

This journal is

The Royal Society of Chemistry 2009

Dalton Trans., 2009, 4556–4568 | 4561

©

View Article Online

74.26 (ring CH₂), 52.89 (CH₂–N), 32.55 (CH₂–S), 31.88 (CH₂),

30.50 (CH₂), 29.90 (CH₂), 29.06 (CH₂), 29.02 (CH₂), 22.75 (CH₂),

14.21 (CH₃).

40.7 (CH₂–N), 35.7 (CH₂–S), 28.6 d (J = 3.0, Ph–CH₂) ppm.¹⁹F

NMR (CDCl₃): -118.3 ppm.

¹H NMR (CDCl₃, 7.26 ppm): 7.30 t (1H, J = 7.3, 3-Ph), 7.17 m

(1H, 5-Ph), 7.05 t (1H, J = 6.7, 4-Ph), 6.98 t (1H, J = 10, 6-

Ph), 3.69 s (2H, Ph–CH₂–S), 2.80 m (2H, CH₂–N), 2.51 m (2H,

2-(2,4,6-Trimethylbenzylthio)ethylamine. Prepared analogo-

usly to the precursor of 1d from 2-mercaptoethylamine hydrochlo-

ride (3.37 g, 29.6 mmol), NaOH (2.37 g, 59.5 mmol) and 2,4,6-

1

S–CH₂), 1.26 s (2H, NH₂). ¹³C-{ H} NMR (CDCl₃, 77.16 ppm):

160.7 d (J_CF= 246, 1-Ph), 130.9 d (J_CF= 3.5, J_CH(d) = 159, 3-Ph),

128.7 d (J_CF= 9, J_CH(dd) = 162 and 9, 5-Ph), 125.7 d (J_CFand

2

trimethylbenzylchloride (a -chloroisodurene) (5 g, 29.6 mmol) as

a yellow oil in 70% yield.

J

_CH(dddd) all about 5, 2-Ph), 124.2 d (J_CF= 3.5, J_CH(dd) = 162

¹H NMR (CDCl₃): 6.85 s (2H, Ph), 3.8 s (2H, Ph–CH₂–S), 2.95t

(2H, J = 6, CH₂–N), 2.7 t (2H, J = 6, CH₂–S), 2.4 s (6H, 2,6-CH₃),

and 8, 4-Ph), 115.4 d (J_CF= 22, J_CH(dd) = 162 and 9, 6-Ph), 40.8

(J_CH(tt) = 136 and 3, CH₂–N), 35.8 (J_CH= 138, CH₂–S), 28.6 d

(J_CF= 2.4, J_CH(td) = 141 and 3, PhCH₂).

1

2.25 s (3H, 4-CH₃), 1.3 br (2H, NH₂). ¹³C-{ H} NMR (CDCl₃):

136.91 (o-Ph), 136.68 and 131.16 (p- and i-Ph), 129.17 (m-Ph),

47.72 (CH₂–NH₂), 41.36 (S–CH₂), 37.33 (CH₂–S), 21.03 (4-CH₃),

19.74 (2,6-CH₃).

Tris(2-(2-ﬂuorobenzylthio)ethyl)triazacyclohexane (1i). Pre-

pared analogously to 1f from the amine (5.85 g, 31.9 mmol) and

paraformaldehyde (0.948 g, 31.6 mmol) in toluene (50 mL) as a

yellowish oil in 87% yield.

Tris(2-(2,4,6-trimethylbenzylthio)ethyl)triazacyclohexane (1g).

Prepared analogously to 1d from the amine (2.9 g, 13.9 mmol)

and paraformaldehyde (0.416 g, 13.9 mmol) in toluene (30 mL) as

a yellow oil in 77% yield.

¹H NMR (CDCl₃): 7.1–7.6 (12H, multiple signals for Ph), 3.85 s

(6H, Ph–CH₂), 3.4 br (6H, ring CH₂), 2.7 m (6H, CH₂–N), 2.6 m

1

(6H, S–CH₂). ¹³C-{ H} NMR (CDCl₃): 162 d (J = 249, C–F),

¹H NMR (CDCl₃): 6.8 s (6H, Ph), 3.75 s (6H, Ph–CH₂–S), 3.35

br (6H, ring CH₂), 2.65 m (12H, CH₂–N and CH₂–S), 2.35 s (18H,

131.63 d (J = 4.1, 6-C), 130.1 d (J = 8.4, 4-C), 126.6 d (J =

15.3, 1-C), 124.94 d (J = 3.6, 5-CH), 115.5 d (J = 21.35, 3-CH),

74.05 (ring CH₂), 42.01 (CH₂–N), 36.05 (CH₂–S), 26.82 (Ph–CH₂)

ppm.¹⁹F NMR (CDCl₃): -118.0.

1

2,6-CH₃), 2.2 s (9H, 4-CH₃). ¹³C-{ H} NMR (CDCl₃): 136.97

(o-Ph), 136.63 and 131.26 (p- and i-Ph), 129.15 (m-Ph), 74.22

(ring CH₂), 52.95 (CH₂–N), 31.32 and 31.16 (CH₂–S–CH₂), 21.05

(CH₃), 19.82 (2,6–CH₃).

¹H NMR (CDCl₃, 7.26 ppm): 7.33 td (3H, J = 7.5 and 1.5,

3-Ph), 7.22 m (3H, 5-Ph), 7.09 td (3H, J = 7.5 and 1.0, 4-Ph), 7.02

ddd (3H, J = 9.3, 8.5 and 1.5, 6-Ph), 3.75 s (6H, FPh–CH₂), 3.33

br (6H, ring CH₂), 2.63 m (6H, N–CH₂), 2.54 m (6H, S–CH₂).

2-(2-Methylbenzylthio)ethylamine. Prepared analogously to

the precursor of 1d from 2-mercaptoethylamine hydrochloride

(3.07 g, 27.0 mmol), NaOH (1.62 g, 54.0 mmol) and 2-

methylbenzylchloride (5 g, 27.0 mmol) as a yellow oil in 69%

yield.

1

¹³C-{ H} NMR (CDCl₃, 77.16 ppm): 161.0 d (J_CF= 247, 1-Ph),

131.1 d (J_CF= 3.5, J_CH(d) = 159, 3-Ph), 128.9 d (J_CF= 9, J_CH(dd) =

162 and 9, 5-Ph), 125.9 d (J_CFand J_CH(dddd) all about 5, 2-Ph),

124.3 d (J_CF= 3.5, J_CH(dd) = 162 and 8, 4-Ph), 115.6 d (J_CF

=

¹H NMR (CDCl₃): 7.2 m (4H, Ph), 3.85 s (2H, Ph–CH₂–S), 2.9 t

(2H, J = 6.78, CH₂–N), 2.6 t (2H, J = 6.78, S–CH₂), 2.4 s (3H,

22, J_CH(ddd) = 162, 8 and 2, 6-Ph), 74.0 (J_CH(tt) = 143 and 4,

ring CH₂), 52.4 (J_CH(t) = 133, CH₂–N), 29.9 (J_CH(tt) = 139 and 4,

CH₂–S), 28.9 d (J_CF= 3, J_CH(td) = 141 and 3, PhCH₂).

1

CH₃), 1.4 br (2H, NH₂). ¹³C-{ H} NMR (CDCl₃): 136.71 and

136.03 (C, Ph), 130.77, 129.67, 127.41 and 125.91 (CH, Ph), 41.06

(CH₂–NH₂), 35.98 (S–CH₂), 34.09 (CH₂–S), 19.20 (CH₃) ppm.

(Triazacyclohexane)tris(copper halide) complexes

Tris(2-(2-methylbenzylthio)ethyl)triazacyclohexane (1h). Pre-

pared analogously to 1d from the amine (700 mg, 3.86 mmol)

and paraformaldehyde (0.116 g, 3.86 mmol) in toluene (20 mL) as

a yellow oil in 87% yield.

CuCl, CuBr and CuI were purchased from Aldrich and used

as received. In some cases indicated, CuBr was recrystallised

as colourless CuBr(MeCN). No signiﬁcant improvement of the

cluster synthesis was found using this purer starting material.

¹H NMR (CDCl₃): 7.1–7.2 m (12H, Ph), 3.75 s (6H, Ph–CH₂–

S), 3.3 br (6H, ring CH₂), 2.60 m (6H, CH₂–N), 2.55 m (6H,

1

S–CH₂), 2.4 s (9H, CH₃). ¹³C-{ H} NMR (CDCl₃): 136.83 and

2a. 1a (333 mg, 0.92 mmol) and CuCl (274.4 mg, 2.77 mmol)

were suspended in 100 mL of MeCN. The CuCl slowly dissolved.

While stirring the mixture overnight, a yellow solid precipitated.

The solid was ﬁltered off and dried in vacuo to yield 230 mg.

Removal of the solvent from the solution yielded another 273 mg

of yellow product (total yield 83%). The product was slightly

soluble in MeCN and DCM but not in THF. Crystals suitable

for X-ray crystallography were grown from an MeCN solution.

Anal. found (calcd for C₂₁H₂₄N₆Cl₃Cu₃): C, 37.8 (38.36); H,

3.69 (3.68); N, 13.4 (12.78)%; recrystallised material contained

one equivalent of MeCN: found (calcd for C₂₃H₂₇N₇Cl₃Cu₃): C,

39.45 (39.55); H, 3.89 (3.90); N, 14.05 (14.04)%.

135.91 (C, Ph), 130.84, 129.76, 127.47 and 125.94 (CH, Ph), 74.13

(ring CH₂), 52.58 (CH₂–N), 34.76 (S–CH₂), 29.95 (CH₂–S), 19.34

(CH₃).

2-(2-Fluorobenzylthio)ethylamine. Prepared analogously to

the precursor of 1d from 2-mercaptoethylamine hydrochloride

(3.92 g, 34.6 mmol), NaOH (2.76 g, 69.2 mmol) and 2-

ﬂuorobenzylchloride (4.12 mL, 34.6 mmol) as a yellow oil in 91%

yield.

¹H NMR (CDCl₃): 7.35 t (1H, J = 7.5, 3-Ph), 7.25 dd (1H, J =

5.7, 5-Ph), 7.15 t (1H, J = 7.4, 4-Ph), 7.05 (1H, J = 8.5, 6-Ph),

3.75 s (2H, Ph–CH₂–S), 2.85 br (2H, CH₂–N), 2.6 t (2H, J = 6.2,

¹H NMR (CD₃CN, 1.94 ppm): 8.54 d (3H, J = 4.4 Hz, 6-py),

7.80 td (3H, J = 7.8, 1.6 Hz, 4-py), 7.36 m (6H, 3- and 5-py), 3.97 s

(6H, py-CH₂), 3.65 br (3H, eq. ring CH₂), 3.32 br (3H, ax. ring

1

S–CH₂), 1.8 br (2H, NH₂). ¹³C-{ H} NMR (CDCl₃): 160.8 d (J =

246, 1-Ph), 130.9 d (J = 3.9, 3-Ph), 129.1 d (J = 8.1, 5-Ph), 125.7 d

(J = 14.8, 2-Ph), 124.2 d (J = 3.6, 4-Ph), 115.4 d (J = 21.9, 6-Ph),

1

CH₂). ¹³C-{ H} NMR (CD₃CN, 1.32 ppm): 156.4 (2-py), 150.4

4562 | Dalton Trans., 2009, 4556–4568

This journal is

The Royal Society of Chemistry 2009

©

View Article Online

(6-py), 138.8 (4-py), 125.8, 125.0 (5- and 3-py), 74.5 (CH₂, ring),

58.3 (py-CH₂).

¹H NMR (CD₃NO₂, 4.33 ppm): 8.585 d (3H, J = 5.44, 6-py),

7.463 d (3H, J = 5.09, 5-py), 7.391 s (3H, 3-py), 3.970 d (3H, J =

8.3, eq. ring CH₂), 3.798 s (6H, pyCH₂), 3.088 d (3H, J = 8.3, ax.

¹H NMR (CH₃CN, 1.93 ppm, 1.9 mM): 8.53 d (3H, J = 5.0,

6-py), 7.78 t (3H, J = 7.8, 4-py), 7.36 d (3H, J = 7.8, 3-py), 7.33 t

(3H, J = 6.5, 5-py), 3.94 s (6H, py-CH₂), 3.58 br (3H, eq. ring

1

ring CH₂), 1.323 d (27H, J = 1.57, ^tBu). ¹³C-{ H} NMR (CD₃NO₂,

62.8 ppm): 163.7 (2-py), 154.9 (4-py), 150.5 (6-py), 122.6, 121.7 (3-

and 5-py), 76.8 (ring CH₂), 58.8 (pyCH₂), 36.0 (C), 30.6 (CH₃).

¹H NMR (PhNO₂, p: 7.68 ppm, 5.0 mM): 8.86 d (3H, J = 5,

6-py) other py signals obscured by solvent, 4.31 d (3H, J = 8, eq.

ring CH₂), 3.92 s (6H, pyCH₂), 3.29 d (3H, J = 8, ax. ring CH₂),

1.25 s (27H, ^tBu).

¹H NMR (PhF, 0.6 mM): 8.77 br (3H, 6-py), (3,5-py covered by

solvent), 4.05 br (3H, eq. ring CH₂), 3.40 s (6H, pyCH₂), 2.73 br

(3H, ax. ring CH₂), 1.09 s (27H, ^tBu).

1

CH₂), 3.29 br (3H, ax. ring CH₂). ¹³C-{ H} NMR (CH₃CN, 1.28

ppm): 156.3 (2-py), 150.0 (6-py), 138.5 (4-py), 125.4 (5-py), 124.6

(3-py), 73.9 (CH₂, ring), 58.0 (py-CH₂).

3a. CuBr (108 mg, 0.753 mmol) was dissolved in 5 mL of

MeCN. This produced a slightly greenish solution with some

insoluble solids. The solution was allowed to settle and the clear

solution was decanted onto 1a (66.8 mg, 0.185 mmol). A yellow

solid was formed immediately. The mixture was shaken for 5 min,

allowed to settle, the solution decanted and the residue washed

with further 4 mL of MeCN, two portions of 4 mL Et₂O and dried

in vacuo giving 122 mg of 3a (83%). A saturated solution in MeCN

was about 1–2 mM by NMR.

EI-HRMS: m/z found (calcd for C₂₁H₂₄N₆Br₂Cu₃[M -

Br]⁺), 710.8311 (710.8275). Anal. found (calcd for 3a·MeCN,

C₂₃H₂₇N₇Br₃Cu₃): C, 33.1 (33.21); H, 3.22 (3.27); N, 11.7 (11.79)%.

¹H NMR (MeCN, 1.93 ppm, 2.2 mM): 8.57 d (3H, J = 4.2,

6-py), 7.76 t (3H, J = 7.8, 4-py), 7.32 m (6H, 3- and 5-py), 3.83 s

(6H, pyCH₂), 3.71 br (3H, eq. ring CH₂), 3.16 br (3H, ax. ring

CH₂).

2c. A saturated solution of NaCl (300 mg, 5.2 mmol) in

water was prepared and degassed by bubbling nitrogen through

it. In another ﬂask, 3c (300 mg, 0.285 mmol) was dissolved in

dichloromethane and added into the saturated sodium chloride

solution. Silver nitrate (145 mg, 0.855 mmol) was added to the

mixture. The two phase mixture was stirred for 2 h. The organic

phase (grey) was collected and the solvent evaporated leaving a

white-grey solid in 62% yield.

ESI-HRMS: m/z found 880.0573 (880.0505 calcd for

C₃₆H₅₁Cl₂Cu₃N₃S₃[M - Cl]⁺).

¹H NMR (CDCl₃, 0.9 mM): 7.10–7.25 (15H, sets of signals

for Ph), 4.07 br (3H, eq. ring CH₂), 2.95 br (3H, ax. ring CH₂),

2.45–2.80 (24H, set of signals for CH₂–S, S–CH₂and CH₂–N),

¹H NMR (MeCN, 1.93 ppm, 1.0 mM): 8.65 d (3H, J = 4.2,

6-py), 7.81 t (3H, J = 7.8, 4-py), 7.40 t (3H, J = 6, 3-py), 7.32 d

(3H, J = 7, 5-py), 3.86 d (3H, J = 8, eq. ring CH₂), 3.77 s (6H,

pyCH₂), 3.05 br (3H, ax. ring CH₂) ppm. ¹³C-{HMBCGPND}

NMR (MeCN, 1.28 ppm): 155.5 (2-py), 150.0 (6-py), 138.3 (4-py),

124.9 (3-py and 5-py), 74.6 (ring CH₂), 58.0 (pyCH₂).

1

1.95 br (6H, Ph–CH₂–CH₂). ¹³C-{ H} NMR (CDCl₃): 140.02,

129.69, 127.54 and 125.5 (Ph), 72.98 (ring CH₂), 53.10 (CH₂–N),

35.22 (S–CH₂), 32.09 (Ph–CH₂), 31.67 (CH₂CH₂CH₂–S), 30.72

(CH₂–S).

¹H NMR (MeCN–DCM (60 : 40 w%), 1.93 ppm, 2.0 mM):

8.59 d (3H, J = 4.7, 6-py), 7.75 dt (3H, J = 1.3/7.8, 4-py),

7.34 t (3H, J = 6.4, 5-py), 7.29 d (3H, J = 7.2, 3-py), 3.82 s

(6H, pyCH₂), 3.77 br (3H, eq. ring CH₂), 3.12 br (3H, ax. ring

CH₂). ¹³C-{HMBCGPND} NMR (MeCN, 1.28 ppm): 154.4 (2-

py), 149.7 (6-py), 137.8 (4-py), 124.2 and 124.4 (3-py and 5-py),

74.6 (ring CH₂), 57.5 (pyCH₂).

3c. 1c (140 mg, 0.228 mmol) was dissolved in dry acetonitrile

(25 mL) and added to CuBr (98.2 mg, 685 mmol). The reaction ran

overnight affording a green solution over a colourless precipitate.

The solution was decanted and the colourless solid was washed

with diethyl ether and dried in vacuo yielding 76% of the cluster.

ESI-HRMS: m/z 966.9420 found (966.9421 calcd for

C₃₆H₅₀Br₂Cu₃N₃S₃[M - Br]⁺). Anal. found (calcd for

C₃₆H₅₀Br₃Cu₃N₃S₃): C, 41.2 (41.09); H, 4.84 (4.88); N, 3.87

(3.99)%. Mp 180 ^◦C.

¹H NMR (CH₂Cl₂, 5.30 ppm, 0.9 mM): 8.75 d (3H, J = 4.3,

6-py), 7.69 t (3H, J = 7.7, 4-py), 7.34 t (3H, J = 6.1, 3-py), 7.12 d

(3H, J = 7.4, 5-py), 3.96 d (3H, J = 8.0, eq. ring CH₂), 3.66 s (6H,

pyCH₂), 2.84 d (3H, J = 7, ax. ring CH₂) ppm.

¹H NMR (CDCl₃, 0.5 mM): 7.15–7.2 (15H, sets of signals

for Ph), 4.23 d (3H, J = 7.6, eq. ring CH₂), 2.6–2.8 m (27H,

Ph–CH₂, CH₂–S–CH₂CH₂–N, ax. ring CH₂), 1.98 m (6H, Ph–

¹H NMR (PhNO₂, para-H 7.68 ppm, 4.5 mM): 8.93 d (3H, J =

4, 6-py), (other pyridyl signal hidden by solvent), 4.25 br (3H,

eq. ring CH₂), 3.98 s (6H, pyCH₂), 3.27 br (3H, ax. ring CH₂).

1

CH₂–CH₂), 2.5 t (6H, Ph–CH₂). ¹³C-{ H} NMR (CDCl₃): 138.31,

128.92, 128.67 and 125.5 (Ph), 74.49 (ring CH₂), 53.10 (CH₂–N),

35.22 (S–CH₂), 32.09 (Ph–CH₂), 31.66 (CH₂CH₂CH₂–S), 30.71

(CH₂–S).

1

¹³C-{ H} NMR (PhNO₂, p-CH 135.32 ppm): 154.3 (2-py), 150.7

(6-py), 138.0 (4-py), (3-py and 5-py obscured by solvent), 75.9

(ring CH₂), 58.2 (pyCH₂).

¹H NMR (CDCl₃–CD₃NO₂, 4.33 ppm, 5 mM): 7.1–7.25 (15H,

sets of signals for Ph), 4.2 d (3H, J = 8.1, eq. ring CH₂), 2.6–2.8 m

(27H, Ph–CH₂, CH₂–S–CH₂CH₂–N, ax. ring CH₂), 2.0 m (6H,

Ph–CH₂–CH₂).

3b. 1b (1.20 g, 2.26 mmol) was dissolved in MeCN (60 mL)

and CuBr (972 mg, 6.78 mmol) added. The mixture was stirred

overnight and then allowed to settle. The clear yellow solution was

decanted, layered with THF and then hexane and left standing in

a glovebox. After several months, large yellow crystals formed. A

saturated solution in ﬂuorobenzene at 298 K is 0.6 mM.

EI-HRMS (ﬂuorobenzene): m/z found (calcd for

C₃₃H₄₈N₆Br₃Cu₃[M]⁺), 957.9277 (957.9337). Anal. found

(calcd for C₃₃H₄₈N₆Br₃Cu₃): C, 41.0 (41.32); H, 5.01 (5.04); N,

8.44 (8.76)%.

4c. 3c (200 mg, 0.19 mmol) was dissolved in dichloromethane.

A saturated aqueous solution of NaI (200 mg, 1.34 mmol) was

added over this solution. Both solutions were stirred for 1 h. The

colourless organic phase was collected and the solvent evaporated

under vacuum. A white solid was obtained in 71% yield.

ESI-HRMS: m/z 1063.9151 found (1063.9217 calcd

for C₃₆H₅₁Cu₃I₂N₃S₃[M - I]⁺). Anal. found (calcd for

This journal is

The Royal Society of Chemistry 2009

Dalton Trans., 2009, 4556–4568 | 4563

©

View Article Online

C₃₆H₅₁I₃Cu₃N₃S₃): C, 36.3 (36.23); H, 4.34 (4.31); N, 3.50

(3.52)%.

C₃₀H₆₃Br₃Cu₃N₃S₃): C, 36.2 (36.31); H, 6.31 (6.40); N, 4.23

(4.23)%. Mp 163 ^◦C.

¹H NMR (CDCl₃, 4 mM): 7.10–7.25 (15H, sets of signals for

Ph), 3.99 d (3H, J = 8.0, eq. ring CH₂), 3.23 d (3H, J = 8.0, ax. ring

CH₂), 2.64 (18H, CH₂–S and S–CH₂CH₂N), 2.48 t (6H, J = 7.2,

¹H NMR (CDCl₃, 2 mM): 4.28 d (3H, J = 7.7, eq. ring CH₂),

3.0 q (3H, J = 6, CH), 2.80 m (6H, CH₂–N), 2.65 m (6H, CH₂–S),

2.61 d (3H, J = 7, ax. ring CH₂), 1.65 m, 1.60 m, 1.45 m and 1.40 m

(6H each, CH₃–CH₂CH₂–CH), 0.91 t (18H, J = 7.3, CH₃). ¹³C-

1

Ph–CH₂), 1.82 m (6H, Ph–CH₂–CH₂). ¹³C-{ H} NMR (CDCl₃):

1

141.48, 128.65, 128.63 and 126.20 (Ph), 75.49 (ring CH₂), 52.79

(CH₂–N), 35.04 (S–CH₂), 33.66 (Ph–CH₂), 31.9 (CH₂CH₂CH₂–

S), 30.25 (CH₂–S).

{ H} NMR (CDCl₃): 73.2 (ring CH₂), 53.76 (CH₂–N, J_CH(t) =

138), 46.76 (CH, J_CH(d) = 142), 34.78 (CH₂CH₂CH, J_CH(t) =

127), 28.90 (CH₂–S, J_CH(t) = 137), 19.29 (CH₃–CH₂, J_CH(t) =

125), 14.45 (CH₃, J_CH(q) = 117) ppm.

3d. 1d (200 mg, 0.394 mmol) was dissolved in acetonitrile

(20 mL) and CuBr (MeCN) (0.169 mg, 1.18 mmol) was added.

After stirring overnight the solvent was removed in vacuo and the

colourless solid washed with hexane and dried in vacuo yielding

70% of 3d.

4e. 3e (0.20 g, 0.2 mmol) was converted with aq. NaI to 4e

analogous to 4f in 69% yield.

ESI-HRMS: m/z 1004.0271 found (1004.0271 calcd

for C₃₀H₆₃I₂Cu₃N₃S₃[M - I]⁺). Anal. found (calcd for

C₃₀H₆₃I₃Cu₃N₃S₃): C, 32.0 (31.79); H, 5.58 (5.60); N, 3.68

(3.71)%.

Anal. found (calcd for C₂₇H₅₇Br₃Cu₃N₃S₃): C, 35.3 (34.12); H,

6.23 (6.05); N, 4.58 (4.42)%.

¹H NMR (CDCl₃): 4.22 d (3H, J = 6.4, eq. ring CH₂), 3.05 d (3H,

J = 6, ax. ring CH₂), 2.95 q (3H, J = 5, CH), 2.80 br (12H, CH₂–N

and S–CH₂), 1.65 m (12H, CH₃–CH₂CH₂), 1.45 m (12H, CH₃–

¹H NMR (CDCl₃, 7 mM): two broad peaks corresponding to

ring CH₂4.0 and 3.3, 2.75 m (12H, S–CH₂–CH₂–N), 2.56 d (6H,

J = 5.9, CH₂–S), 1.47 m (3H, CH), 1.41 m (12H, CH₃–CH₂), 0.87 t

1

CH₂CH₂), 0.92 t (18H, J = 7.2, CH₃). ¹³C-{ H} NMR (CDCl₃):

1

(18H, J = 7.3, CH₃). ¹³C-{ H} NMR (CDCl₃): 70.69 (ring CH₂),

76.6 (ring CH₂), 55.60 (CH₂–N), 46.64 (CH), 34.59 (CH₂CH₂CH),

28.31 (CH₂–S), 19.33 (CH₃–CH₂), 14.53 (CH₃).

52.68 (CH₂–N), 40.39 (CH), 37.70 (S–CH₂), 31.32 (CH₂–S), 25.07

(CH₃–CH₂), 11.02 (CH₃).

¹H NMR (CH₂Cl₂, 5.31 ppm, 97 mM): 4.18 d (3H, J = 7.5, eq.

ring CH₂), 2.95 br (3H, ax. ring CH₂), 2.91 q (3H, J = 5.3, CH),

2.78 m (6H, S–CH₂), 2.71 m (6H, CH₂–N), 1.62 m and 1.58 m

(6H each, CH₃–CH₂CH₂), 1.41 m and 1.39 m (6H each, CH₃–

2e. Direct method: 1e (100 mg, 0.178 mmol) was dissolved in

dry acetonitrile (20 mL) and degassed. The solution was added to

CuCl (53 mg, 0.534 mmol) and the solution was stirred overnight.

The solution was reduced under vacuum and the precipitate

isolated. Crystallisation by slow evaporation of a solution in

CHCl₃–MeNO₂resulted in a small amount of crystals.

1

CH₂CH₂), 0.90 t (18H, J = 7.3, CH₃). ¹³C-{ H} NMR (CH₂Cl₂,

53.73 ppm): 76.77 (ring CH₂, J_CH(dd) = 141 and 153), 53.69 (CH₂–

N, J_CH(dd) = 138), 46.23 (CH, J_CH(d) = 140), 34.27 (CH₂CH₂CH,

J

_CH(t) = 125), 27.90 (CH₂–S, J_CH(t) = 140), 19.08 (CH₃–CH₂,

_CH(t) = 125), 14.07 (CH₃, J_CH(t) = 125).

Anal.

found

(calcd

for

2e

with

1.5

CHCl₃,

C_31.5H_64.5Cl_7.5Cu₃N₃S₃): C, 36.5 (36.45); H, 6.33 (6.26); N,

4.05 (4.05)%.

Method via 3e: 3e (300 mg, 0.302 mmol) was converted to 2e

analogously to 2c with saturated NaCl and AgNO₃(153.69 mg,

0.904 mmol) in 58% yield.

2f. 3f (300 mg, 0.302 mmol) was converted to 2f analogous to

2c with saturated NaCl and AgNO₃(153.69 mg, 0.906 mmol) in

60% yield.

ESI-HRMS: m/z 722.2471 found (722.2459 calcd for

ESI-HRMS: m/z 820.1474 found (820.1444 calcd for

C₃₀H₆₃Cl₂Cu₃N₃S₃[M - Cl]⁺).

¹H NMR (1,2-dichloroethane, 3.73 ppm, 50 mM): 4.07 d (3H,

J = 8, ax. ring CH₂), 2.71 br (6H, CH₂–N), 2.63 d (3H, J = 8,

eq. ring CH₂), 2.57 br (6H, S–CH₂), 2.53 t (6H, J = 7.3, CH₂–

S), 1.58 m (6H, CH₂–CH₂–S), 1.36 m and 1.2–1.3 (24H, sets of

signals for CH₂–CH₂–CH₂–CH₂), 0.86 t (9H, J = 7.5, CH₃). ¹³C-

¹H NMR (DCM, 5.31 ppm, 33 mM): 4.10 br (3H, eq. ring

CH₂), 2.93 q (3H, J = 6.0, CH), 2.74 m (6H, S–CH₂), 2.61 m (6H,

N–CH₂), 2.59 br (3H, ax. ring CH₂), 1.60 m (12H, CH₂–CH),

1

1.40 m (12H, CH₃–CH₂), 0.90 t (18H, J = 7.3, CH₃). ¹³C-{ H}

NMR (DCM, 53.73 ppm): 76.42 (J_CH(t) = 142, ring CH₂), 53.09

(J_CH(t) = 136, CH₂–N), 46.09 (J_CH(d) = 137, CH), 34.58 (J_CH(t) =

124, CH₂CH₂CH), 29.00 (J_CH(t) = 140, CH₂–S), 19.02 (J_CH(t) =

123, CH₃–CH₂), 13.92 (J_CH(qt) = 125 and 7.3, CH₃).

1

{ H} NMR (CDCl₃): 76.21 (ring CH₂), 51.51 (CH₂–N), 32.37

(CH₂), 31.10 (CH₂), 29.31 (CH₂–S), 28.30 (CH₂), 28.18 (CH₂),

27.74 (CH₂), 22.02 (CH₂), 13.33 (CH₃).

¹H NMR (PhF, 0.5 mM): 4.11 br (3H, eq. ring CH₂), 2.96 m

(3H, CH), 2.61 m (6H, S–CH₂), 2.40 m (6H, N–CH₂), 2.59 br

(3H, ax. ring CH₂), 1.63 m (12H, CH₂–CH), 1.40 m (12H, CH₃–

3f. 1f (2 g, 3.5 mmol) was dissolved in dry acetonitrile (35 mL)

and CuBr (1.53 g, 10.7 mmol) was added. The cluster was soluble

in acetonitrile. The solvent was removed in vacuo and the white

solid was washed with hexane. After drying under vacuum, 2.5 g

of a colourless solid (71% yield) was obtained.

ESI-HRMS: m/z 908.0416 found (908.0433 calcd for

C₃₀H₆₃Br₂Cu₃N₃S₃[M - Br]⁺). Anal. found (calcd for

C₃₇H₅₄Br₃Cu₃N₃S₃): C, 36.15 (36.31); H, 6.13 (6.40); N, 4.1

(4.23)%. Mp 155 ^◦C.

¹H NMR (CDCl₃, 3 mM): 4.19 d (3H, J = 7, ax. ring CH₂),

3.12 d (3H, J = 7, eq. ring CH₂), 2.83 m (12H, S–CH₂and CH₂–

N), 2.63 t (6H, J = 6.5, CH₂–S), 1.65 m (6H, CH₂–CH₂–S), 1.2–1.5

(24H, set of signals for CH₂–CH₂–CH₂–CH₂), 0.9 t (9H, J = 7.3,

1

CH₂), 0.85 t (18H, CH₃). ¹³C-{ H} NMR (PhF): 77.0 (ring CH₂),

54.0 (CH₂–N), 47.4 (CH), 35.4 (CH₂CH₂CH), 29.3 (CH₂–S), 19.7

(CH₃–CH₂), 14.4 (CH₃).

3e. 1e (0.60 g, 1.0 mmol) was dissolved in acetonitrile (15 mL)

and CuBr (0.46 g, 3.2 mmol) was added. After stirring overnight

the solution was decanted, the solvent removed in vacuo yielding

68% of colourless solid 3e.

ESI-HRMS: m/z 908.0416 found (908.0433 calcd for

C₃₀H₆₃Br₂Cu₃N₃S₃[M - Br]⁺). Anal. found (calcd for

4564 | Dalton Trans., 2009, 4556–4568

This journal is

The Royal Society of Chemistry 2009

©

View Article Online

1

CH₃). ¹³C-{ H} NMR (CDCl₃): 74.45 (ring CH₂), 53.08 (CH₂–

127.84 (Ph), 75.72 (ring CH₂), 53.45 (CH₂–N), 35.67 (S–CH₂),

32.02 (CH₂–S), 21.0 (CH₃).

N), 33.09 (CH₂–S), 32.44 (CH₂), 32.11 (CH₂), 30.71 (CH₂), 30.23

(CH₂), 29.92 (CH₂), 29.23 (CH₂), 14.45 (CH₃).

2i. Direct method: 1i (2 g, 3.375 mmol) was dissolved in MeCN

and 3 equivalents of CuCl (1 g, 10.125 mmol) was added. A white

precipitate was formed instantaneously. The white product was

ﬁltered, washed with hexanes and dried under vacuum affording

a 85% yield of the cluster.

Method via halide exchange: a saturated solution of NaCl

(300 mg, 1.76 mmol) in water was prepared and degassed by

bubbling nitrogen through it. In another ﬂask, 3i (300 mg,

0.258 mmol) was dissolved in dichloromethane and added to

the saturated sodium chloride solution. Silver nitrate (131 mg,

0.774 mmol) was added to the mixture. The two phase mixture

was stirred for 2 h. The organic phase (grey) was collected and the

solvent evaporated, leaving a white-grey solid in 54% yield.

ESI-HRMS: m/z 849.9251 found (849.9283 calcd for

C₃₀H₃₆Cl₂Cu₃F₃N₃S₃[M - Cl]⁺). Anal. found (calcd for

C₃₀H₃₆Cl₃Cu₃F₃N₃S₃): C, 40.5 (40.54); H, 4.03 (4.08); N, 5.00

(4.73)%.

¹H NMR (o-dichlorobenzene, 7.19 ppm, 21 mM): 4.17 d (3H,

J = 7.5, eq. ring CH₂), 2.71 br (6H, CH₂–N), 2.58 d (3H, J = 7.5,

ax. ring CH₂), 2.54 br (6H, CH₂–S), 2.53 t (6H, J = 7.5, SCH₂),

1.53 q (6H, J = 7.3, CH₂–CH₂–S), 1.1–1.2 (24H, CH₃–CH₂CH₂–

CH), 0.83 t (9H, J = 7.3, CH₃).

4f. 3f (200 mg, 0.2 mmol) was dissolved in dichloromethane. A

saturated aqueous solution of NaI (200 mg, 1.3 mmol) was added

and the mixture stirred for 1 h. The colourless organic phase was

collected and the solvent evaporated under vacuum. A white solid

was obtained in 65% yield.

ESI-HRMS: m/z 1004.0271 found (1004.0156 calcd

for C₃₀H₆₃I₂Cu₃N₃S₃[M - I]⁺). Anal. found (calcd for

C₃₀H₆₃I₃Cu₃N₃S₃): C, 33.3 (31.79); H, 5.82 (5.60); N, 3.87

(3.71)%.

¹H NMR (CDCl₃, 40 mM): 4.1 d (3H, J = 7, eq. ring CH₂),

3.5 d (3H, J = 7, ax. ring CH₂), 2.90 m and 2.85 m (12H, S–CH₂

and CH₂–N), 2.5 t (6H, J = 7.5, CH₂–S), 1.6 q (6H, J = 7.1,

CH₂CH₂S), 1.2–1.4 m (24H, set of signals for CH₂–CH₂–CH₂–

¹H NMR (CDCl₃, 6 mM): 7.4, 7.2, 7.1, 7.0 (12H, set of signals

for Ph), 4.3 br (3H, eq. ring CH₂), 3.95 br (6H, Ph–CH₂), 3.3

br (3H, ax. ring CH₂), 2.9 br (12H, S–CH₂–CH₂–N). ¹⁹F NMR

(CDCl₃): -117.2.

CH₂), 0.90 t (9H, J = 6.9, CH₃). ¹³C-{ H} NMR (CDCl₃): 74.65

1

(ring CH₂), 52.75 (CH₂–N), 34.05 (CH₂–S), 31.87 (CH₂), 29.61

(CH₂), 29.10 (CH₂), 28.94 (CH₂), 28.57 (CH₂), 22.74 (CH₂), 14.25

(CH₃).

¹H NMR (DCM, 5.31 ppm, 10 mM): 7.45 t (3H, J = 7.40, 3-

Ph), 7.25 q (3H, J = 6.6, 5-Ph), 7.13 t (3H, J = 7.4, 4-Ph), 7.03 t

(3H, J = 9.2, 6-Ph), 4.18 d (3H, J = 7.9, eq. ring CH₂), 3.87 s (6H,

FPh–CH₂), 2.70 m (6H, S–CH₂), 2.66 m (6H, N–CH₂), 2.63 (3H,

3g. 1g (175 mg, 0.263 mmol) was dissolved in acetonitrile

(25 mL) and stirred while CuBr (113 mg, 0.79 mmol) was added

into the solution. The reaction mixture was stirred overnight. The

solution was decanted and the solvent pumped off, affording a

white solid in 69% yield.

1

ax. ring CH₂). ¹³C-{ H} NMR (DCM, 53.73 ppm): 160.8 d (J_CF

246, 1-Ph), 131.6 d (J_CF= 3.5, 3-Ph), 129.2 d (J_CF= 8.2, J_CH

163, 5-Ph), 124.5 d (J_CF= 3.5, J_CH= 163, 4-Ph), 123.8 d (J_CF

=

15, 2-Ph), 115.2 d (J_CF= 22, J_CH= 169, 6-Ph), 76.7 (J_CH= 150,

ring CH₂), 51.7 (J_CH= 124, CH₂–N), 29.6 (J_CH= 130, CH₂–S),

29.5 d (J_CF= 2.4, J_CH= 143, PhCH₂).

ESI-HRMS: m/z 1009.9936 found (1009.9969 calcd

for C₃₉H₅₇Br₂Cu₃N₃S₃[M - Br]⁺). Anal. found (calcd for

C₃₉H₅₇Br₃Cu₃N₃S₃): C, 40.0 (42.80); H, 4.77 (5.25); N, 3.61

(3.84)%. Mp 190 ^◦C.

3i. 1i (4 g, 6.75 mmol) was dissolved in dry acetonitrile (35 mL)

and a solution of CuBr (2.9 g, 20.27 mmol) was added. A white

solid precipitated instantaneously. This solid was ﬁltered and

washed several times with hexane. The solution was decanted and

placed in the fridge. A crystalline solid precipitated. Both solids

were dried under vacuum to give a combined yield of 94%.

ESI-HRMS: m/z 796.0228 found (795.9793 calcd for [M -

Br]⁺). Anal. Found (calcd for C₃₀H₃₆Br₃Cu₃F₃N₃S₃): C, 36.0

(35.25); H, 3.66 (3.55); N, 4.70 (4.11)% (calcd for 3j·MeCN,

C₃₂H₃₉Br₃Cu₃F₃N₄S₃: C, 36.15; H, 3.70; N, 5.27%).

¹H NMR (o-dichlorobenzene, 7.0 mM): 6.63 s (6H, Ph), 4.24 d

(3H, J = 7.5, eq. ring CH₂), 3.83 s (6H, Ph–CH₂–S), 2.70 m (6H,

CH₂–N), 2.53 m (9H, CH₂–S and ax. ring CH₂), 2.31 s (18H, 2,6-

CH₃), 2.13 s (9H, 4-CH₃). ¹³C-{ H} NMR (o-dichlorobenzene):

1

136.8 (o-Ph), 136.6 and 130.9 (i and p-Ph), 129.36 (m-Ph), 77.02

(ring CH₂), 53.70 (CH₂–N), 32.59 (S–CH₂), 31.19 (CH₂–S), 20.96

(CH₃), 20.55 (CH₃).

3h. 1h (212 mg, 0.36 mmol) was dissolved in acetonitrile

(15 mL) and stirred while three equivalents of CuBr (127 mg,

1.09 mmol) were added into the solution. The reaction mixture

was stirred overnight. The solution was decanted and the solvent

pumped off, affording a white solid that was washed with hexane

and dried in vacuo. Yield 54%.

ESI-HRMS: m/z 925.9021 found (995.9030 calcd for

C₃₃H₄₅Br₂Cu₃N₃S₃[M - Br]⁺). Anal. found (calcd for

C₃₃H₄₅Br₃Cu₃N₃S₃): C, 39.12 (39.23); H, 4.77 (4.49); N,

4.15 (4.16)%.

¹H NMR (CDCl₃, 13 mM): 6.95–7.40 (12H, 4 sets of signals for

Ph), 4.25 br (3H, eq. ring CH₂), 3.95 s (6H, Ph–CH₂), 3.75 br (3H,

ax. ring CH₂), 3.0 m (12H, S–CH₂–CH₂–N). ¹⁹F NMR (CDCl₃):

-117.3.

¹H NMR (CHCl₃, 7.26 ppm, 8 mM): 7.38 t (3H, J = 7, 3-Ph),

7.18 q (3H, J = 6, 5-Ph), 7.07 t (3H, J = 7.5, 4-Ph), 6.99 t (3H, J =

9.3, 6-Ph), 4.21 d (3H, J = 7.8, eq. ring CH₂), 3.81 s (6H, FPh–

CH₂), 3.36 d (3H, J = 7.8, ax. ring CH₂), 2.91 m (6H, N–CH₂),

1

2.72 m (6H, S–CH₂). ¹³C-{ H} NMR (CHCl₃, 77.36 ppm): 161.0 d

(J_CF= 247, 1-Ph), 131.8 d (J_CF= 3.9, 3-Ph), 129.4 d (J_CF= 7.3,

5-Ph), 124.6 d (J_CF= 3.8, 4-Ph), 124.1 d (J_CF= 15.0, 2-Ph), 115.5 d

(J_CF= 21.5, 6-Ph), 76.9 (ring CH₂), 52.1 (CH₂–N), 30.1 (PhCH₂),

29.6 (CH₂–S). (+ signals for 1 equiv. MeCN and 4 equiv. H₂O.)

¹H NMR (DCM, 5.31 ppm, 17 mM): 7.41 td (3H, J = 7.6 and

1.6, 3-Ph), 7.24 q (3H, J = 7.9, 5-Ph), 7.11 td (3H, J = 7.5 and

¹H NMR (o-dichlorobenzene, 0.6 mM): 4.26 d (3H, J = 7.8,

eq. ring CH₂), 3.75 s (6H, Ph–CH₂–S), 2.52 m (6H, CH₂–N),

2.45 m (6H, S–CH₂), 2.37 d (3H, J = 7.8, ax. ring CH₂), 2.18 s

1

(9H, CH₃) (aromatic signals obscured by solvent). ¹³C-{ H} NMR

(o-dichlorobenzene): 136.80, 132.24, 130.93, 130.74, 127.21 and

This journal is

The Royal Society of Chemistry 2009

Dalton Trans., 2009, 4556–4568 | 4565

©

View Article Online

1.0, 4-Ph), 7.03 t (3H, J = 9.2, 6-Ph), 4.22 d (3H, J = 8.2, eq.

ring CH₂), 3.83 s (6H, FPh–CH₂), 2.75 d (3H, J = 8.0, ax. ring

were optimised for both the sym and asym conformer. The energy

difference at the RI-BP86 level was small (1.76 kJ mol^-1) without a

signiﬁcant difference in the NMR parameters or the bond lengths

to copper . The values given are those for the slightly more stable

1

CH₂), 2.68 m (6H, N–CH₂), 2.68 m (6H, S–CH₂). ¹³C-{ H} NMR

(DCM, 53.73 ppm): 160.8 d (J_CF= 246, 1-Ph), 131.5 d (J_CF= 3.7,

3-Ph), 129.2 d (J_CF= 8.1, 5-Ph), 124.5 d (J_CF= 3.5, 4-Ph), 123.8 d

(J_CF= 15, 2-Ph), 115.3 d (J_CF= 22, 6-Ph), 77.1 (ring CH₂), 52.0

(CH₂–N), 29.7 d (J_CF= 2.5, PhCH₂), 29.1 (CH₂–S). (+ signals for

1 equiv. MeCN.)

˚

asym form. Bond lengths were found to differ up to 0.1 A (Cu–Cu

˚

up to 0.3 A) compared to the crystal structures. Only hydrogens

were optimised with all other atoms at the ﬁxed coordinates of

the best crystal structure (3b) to estimate the effect of different

bond lengths on the electronic structure and calculated NMR

parameters.

4i. Direct method: 1i (4 g, 6.75 mmol) was dissolved in MeCN

and 3 equivalents of CuI (2.9 g, 20.27 mmol) were added. A white

precipitate was formed instantaneously. The white product was

ﬁltered, washed with hexanes and dried under vacuum affording

a 94% yield of the cluster.

Crystal data and reﬁnement details for 1a, 2a, 3a-A, 3a-B, 3b, 2e,

3h and 4i. Intensity data for all structures were collected at 150 K

on a Nonius Kappa CCD diffractometer equipped with an Oxford

Method via halide exchange: 3i (200 mg, 0.19 mmol) was

dissolved in dichloromethane. A saturated aqueous solution of

NaI (300 mg, 2 mmol) was added over this solution. Both solutions

were stirred for 1 h. The colourless organic phase was collected and

the solvent evaporated under vacuum. A white solid was obtained

in 76% yield.

ESI-HRMS: m/z 1043.8915 found (1043.8783 calcd

for C₃₀H₃₆Cu₃F₃I₂N₃S₃[M - I]⁺). Anal. found (calcd for

C₃₀H₃₆Cu₃F₃I₃N₃S₃): C, 31.20 (30.98); H, 3.56 (3.12); N, 3.54

(3.61)%.

cryostream, using graphite monochromated Mo Ka radiation (l =

16

˚

0.71073 A). Data were processed using the Nonius Software.

For 3a-B, 3b, 2e, 3h and 4i, a symmetry-related (multi-scan)

absorption correction was applied. Crystal parameters and details

on data collection, solution and reﬁnement for the complexes are

provided in Table 2. Structure solution, followed by full-matrix

least-squares reﬁnement was performed using the WINGX-1.70

suite of programs throughout.¹⁷

2a and 3a-B crystallise with two molecules of acetonitrile in

the asymmetrical unit. One solvent molecule shows disorder

with 50% occupation for both parts. 3a-A contains solvent-

accessible voids which could be ﬁlled with two solvent molecules

of acetonitrile. However, these were so severely disordered that

the PLATON programme SQUEEZE¹⁸was employed to take

this model into the reﬁnement process. Data collection of 2a

and 3h resulted in weak data at high theta angles due to poor

crystal quality despite numerous attempts on different crystals

from several crystal batches. Reﬁnement for 2a still gave reasonable

structural parameters but 3h gave high R_intand R values with some

unreasonable bond lengths. Hence, for 3h, the phenyl rings of

the ligands were idealised and bond lengths of C21–C22, C1–N3

and C3–N3 restrained. The asymmetrical unit of 4i contains one

solvent molecule of CH₂Cl₂.

¹H NMR (CDCl₃, 8 mM): 7.37 t (3H, J = 7, 3-Ph), 7.23 m (3H,

5-Ph), 7.10 t (3H, J = 7, 4-Ph), 7.01 t (3H, J = 9, 6-Ph), 4.3 d

(3H, J = 8, eq. ring CH₂), 3.8 s (6H, Ph–CH₂), 3.3 d (3H, J = 8,

ax. ring CH₂), 2.9 m (6H, CH₂–N), 2.7 m (6H, CH₂–S). ¹⁹F NMR

(CDCl₃): -117.7.

¹H NMR (DCM, 5.31 ppm, 4.4 mM): 7.40 t (3H, J = 7.6, 3-Ph),

7.26 q (3H, J = 7.9, 5-Ph), 7.13 t (3H, J = 7.5, 4-Ph), 7.04 t (3H,

J = 9.2, 6-Ph), 4.29 d (3H, J = 7.9, eq. ring CH₂), 3.81 s (6H,

FPh–CH₂), 2.66 m (6H, N–CH₂), 2.70 d (3H, J = 8.6, ax. ring

1

CH₂), 2.66 m (6H, S–CH₂). ¹³C-{ H} NMR (DCM, 53.73 ppm):

160.8 d (J_CF= 246, 1-Ph), 131.4 d (J_CF= 3.5, 3-Ph), 129.2 d (J_CF

=

8.1, 5-Ph), 124.6 d (J_CF= 3.1, 4-Ph), 123.9 d (J_CF= 15, 2-Ph),

115.2 d (J_CF= 22, 6-Ph), 77.6 (ring CH₂), 51.7 (CH₂–N), 30.7

(CH₂–S), 28.1 d (J_CF= 5, PhCH₂).

Conclusions

Computational details

All calculations reported in this paper have been obtained with the

ORCA electronic structure program version 2.6.35.⁶A Wachters

basis set¹¹was used for copper and contracted triple-z quality basis

sets with a polarisation for all other atoms (TZVP¹²). Ahlrich’s

auxiliary basis sets TZV/J¹³for all atoms were used for the RI

method. The ORCA implementation of zero-order relativistic

correction (ZORA) and of a COSMO solvent model with inﬁnite

e was used for all calculations.

DFT calculations were performed with the BP86 and B3LYP

functionals.¹⁴Geometry optimisation was done with the former

using the RI method as implemented in ORCA. Maximum

integration grid 7 was used for the heavier atoms Cu, Br and

I. The use of Grimme’s van der Waals corrections¹⁵gave shorter

bond distances without substantial improvements relative to the

experimental structures. The much slower B3LYP functional was

used in some cases in single point calculations of the NMR shifts

for comparison. NMR shifts were calculated using the IGLO

method as implemented in ORCA relative to TMS calculated with

the same method as reference. In the case of 2a, the structures

Well deﬁned triangular tricopper complexes supported by pyridyl-

and thioether-functionalised triazacyclohexanes have been syn-

thesised and characterised. Their much improved solubility and

stability allows investigations in solution and gives access to the

exploration of their chemistry. The cluster complexes are robust

enough to allow exchange of the halide bridges in aqueous media.

Future studies will explore the introduction of non-halide bridges

by this method to make these triangular clusters more similar to

the (hydr)oxo bridges typical of natural enzymes.

The tricopper complexes were well characterised by NMR

spectroscopy. In particular the characteristic equatorial and axial

ring hydrogen signals of the triazacyclohexanes were very sensitive

to changes in the N-substituents, halide bridges and even the

solvent, and showed, at least in some cases, a surprising exchange

on the NMR time scale. Relatively low-level DFT calculations

were able to reproduce the structures and NMR spectra reasonably

well to aid the assignments and indicate an empty orbital in the

centre of the copper triangle that could be involved in reactions

with nucleophiles.

4566 | Dalton Trans., 2009, 4556–4568

This journal is

The Royal Society of Chemistry 2009

©

View Article Online

This journal is

The Royal Society of Chemistry 2009

Dalton Trans., 2009, 4556–4568 | 4567

©

View Article Online

9 R. D. Ko¨hn, Z. Pan, M. F. Mahon and G. Kociok-Ko¨hn, J. Chem.

Soc., Dalton Trans., 2003, 2269–2275.

Acknowledgements

10 R. T. Shuman, P. L. Ornstein, J. W. Paschal and P. D. Gesellchen, J. Org.

Chem., 1990, 55, 738–741.

We thank the EPSRC for support of the project (studentship

for L.T.L. and postdoctoral positions for P.Z. and F.S.) and the

National Computational Service (NSCCS, especially Sarah Wilsey

and Helen Tsui) for supporting the DFT studies.

11 A. J. H. Wachters, J. Chem. Phys., 1970, 52, 1033; A. J. H. Wachters,

IBM Tech. Rept. RJ584, 1969; F exponents; C. W. Bauschlicher, Jr,

S. R. Langhoff and L. A. Barnes, J. Chem. Phys., 1989, 91, 2399.

12 A. Schaefer, H. Horn and R. Ahlrichs, J. Chem. Phys., 1992, 97, 2571;

R. Ahlrichs and K. May, Phys. Chem. Chem. Phys., 2000, 2, 943;

A. Scha¨fer, C. Huber and R. Ahlrichs, J. Chem. Phys., 1994, 100,

5829.

References

13 K. Eichkorn, O. Treutler, H. Ohm, M. Haser and R. Ahlrichs, Chem.

Phys. Lett., 1995, 240, 283; K. Eichkorn, F. Weigend, O. Treutler and

R. Ahlrichs, Theor. Chem. Acc., 1997, 97, 119.

14 A. D. Becke, Phys. Rev. A, 1988, 38, 3098; J. P. Perdew, Phys, Rev. B,

1986, 33, 8822; A. D. Becke, J. Chem. Phys., 1993, 98, 5648; C. Lee, W.

Yang and R. G. Parr, Phys. Rev. B, 1988, 37, 785.

15 S. Grimme, J. Comput. Chem., 2006, 27, 1787–1799 and references

therein.

16 Z. Otwinowski and W. Minor, in Methods in Enzymology, Volume

276: Macromolecular Crystallography, part A, ed. C. W. Carter, Jr. and

R. M. Sweet, Academic Press, 1997, pp. 307–326.

17 L. J. Farrugia, J. Appl. Crystallogr., 1999, 32, 837–838.

18 Sluis and A. L. Spek, Acta Crystallogr., Sect. A, 1990, 46, 194–201.

19 Orbital contour plots were generated using gOpenMol version 3.00

(Leif Laaksonen et al.). The software is available free of charge at:

http://www.csc.ﬁ/english/pages/g0penMol (accessed 2008). Positive

levels are shown in yellow (small) and red (large), negative levels in blue

(light small and dark large.).

1 E. I. Solomon, A. J. Augustine and J. Yoon, Dalton Trans., 2008, 3921–

3932 and references therein.

2 R. D. Ko¨hn, Z. Pan, M. F. Mahon and G. Kociok-Ko¨hn, Chem.

Commun., 2003, 1272–1273; R. D. Ko¨hn, Z. Pan, M. Haufe and G.

Kociok-Ko¨hn, Dalton Trans., 2005, 2793–2797.

3 G. Kickelbick, D. Rutzinger and T. Gallauner, Monatsh. Chem., 2002,

133, 1157–1164.

4 e.g. R. D. Ko¨hn, P. Kampe and G. Kociok-Ko¨hn, Eur. J. Inorg. Chem.,

2005, 3217–3223.

5 R. D. Ko¨hn, G. Seifert, Z. Pan, M. F. Mahon and G. Kociok-Ko¨hn,

Angew. Chem., Int. Ed., 2003, 42, 793 and references therein.

6 F. Neese, ORCA, an Ab Initio, Density Functional and Semiempirical

Electronic Structure Program Package, Version 2.6, revision 35, Max-

Planck Institute for Bioinorganic Chemistry, Mu¨lheim an der Ruhr,

Germany, 2008.

7 A. J. Bridgeman, G. Cavigliasso, L. R. Ireland and J. Rothery, J. Chem.

Soc., Dalton Trans., 2001, 2095–2108 and references therein.

8 R. S. Mulliken, J. Chem. Phys., 1955, 23, 1833–1840.

4568 | Dalton Trans., 2009, 4556–4568

This journal is

The Royal Society of Chemistry 2009

©

Article Doi

Triangular tricopper(i) clusters supported by donor-substituted triazacyclohexanes

DOI: 10.1039/b819268e

Source and publish data:

Authors:

Article abstract of DOI:10.1039/b819268e

Full text of DOI:10.1039/b819268e

Products guided by the article

R&D Labs maybe for 1174513-02-0

Relevant to this article

Hot Product