Thioether coordinated metalloenediynes: Syntheses, structures and thermal reactivity comparison

Article abstract of DOI:10.1016/j.poly.2005.07.032

Two novel, thioether containing benzannulated enediyne ligands, 1,2-bis(3-(quinolin-8-ylsulfanyl)prop-1-ynyl)benzene (1), and 1,2-bis(3-(pyridin-2-ylmethylsulfanyl)prop-1-ynyl)benzene (2) have been synthesized and characterized along with their corresponding Cu(I), Ag(I), Cu(II), and Pd(II) metalloenediyne complexes (3-10). Single crystal X-ray structures of the d¹⁰ and d⁸ N₂S₂ complexes show thioether coordination leading to distorted tetrahedral and trans square planar geometries about the metal centers, respectively. Based on their X-ray structures, all complexes are expected to exhibit high Bergman cyclization temperatures (>200°C), and indeed this is the case for the Cu(I) and Ag(I) complexes (211-218°C). In contrast, the corresponding divalent, tetragonal and trans square planar complexes of Cu(II) and Pd(II) exhibit more facile reactivity and lower solid state Bergman cyclization temperatures (186-190°C) which more closely resemble values expected for cis-complexes. The unusual order in the thermal reactivities of these constructs can be attributed to a previously evaluated mechanism by which weak thioether coordination permits an elevated temperature trans-to-cis geometric rearrangement prior to Bergman cyclization.

Full text of DOI:10.1016/j.poly.2005.07.032

Polyhedron 25 (2006) 550–558
www.elsevier.com/locate/poly
Thioether coordinated metalloenediynes: Syntheses, structures
q
and thermal reactivity comparison
*
Sibaprasad Bhattacharyya, Maren Pink, John C. Huffman, Jeffrey M. Zaleski
Department of Chemistry and Molecular Structure Center, Indiana University, Bloomington, IN 47405, USA
Received 17 June 2005; accepted 15 July 2005
Available online 9 September 2005
Abstract
Two novel, thioether containing benzannulated enediyne ligands, 1,2-bis(3-(quinolin-8-ylsulfanyl)prop-1-ynyl)benzene (1), and
1,2-bis(3-(pyridin-2-ylmethylsulfanyl)prop-1-ynyl)benzene (2) have been synthesized and characterized along with their correspond-
ing Cu(I), Ag(I), Cu(II), and Pd(II) metalloenediyne complexes (3–10). Single crystal X-ray structures of the d10 and d8 N2S2
complexes show thioether coordination leading to distorted tetrahedral and trans square planar geometries about the metal centers,
respectively. Based on their X-ray structures, all complexes are expected to exhibit high Bergman cyclization temperatures
(>200 ꢀC), and indeed this is the case for the Cu(I) and Ag(I) complexes (211–218 ꢀC). In contrast, the corresponding divalent,
tetragonal and trans square planar complexes of Cu(II) and Pd(II) exhibit more facile reactivity and lower solid state Bergman cycli-
zation temperatures (186–190 ꢀC) which more closely resemble values expected for cis-complexes. The unusual order in the thermal
reactivities of these constructs can be attributed to a previously evaluated mechanism by which weak thioether coordination permits
an elevated temperature trans-to-cis geometric rearrangement prior to Bergman cyclization.
ꢁ 2005 Elsevier Ltd. All rights reserved.
Keywords: Metalloenediynes; Diradical; Bergman cyclization; Copper; Palladium; Thioether
1. Introduction
groups at both –ene– and –yne– positions, through space
electron lone-pair electron repulsion) to Bergman cycli-
zation [3]. One fundamental and common goal of much
of this work was to understand and gain control over
the reactivity of the enediyne unit. To this end, photo-
chemical cyclization strategies [9–12] have also been eval-
uated as a means to initiate cyclization in compounds
with pronounced thermal stability such as benzannu-
lated and aromatic acyclic enediyne constructs.
The principles derived from these geometric and elec-
tronic studies paved the way for the use of metal ions as
reagents to activated enediynes toward Bergman cycliza-
tion [3,13]. The first key examples demonstrated that
Bergman cyclization could be strongly influenced by me-
tal ion coordination to a ligating enediyne framework.
For example, binding of Hg2+ to a bipyridyl-enediyne
conjugate forced the ligand into a planar conformation
which labilized the enediyne toward cyclization [14,15].
For over 30 years [1,2] the (Z)-hexa-3-ene-1,5-diyne
unit has intrigued chemists due to the ability of this func-
tionality to cyclize to form a highly reactive 1,4-benze-
noid diradical intermediate [3]. The discovery of this
reactive species within natural products isolated from
microbial fermentation broths [4–8] reignited interest in
systematically evaluating the geometric (e.g., alkyne ter-
mini separation, acyclic versus cyclic structures, H-bond-
ing at the alkyne termini) and electronic contributions
(e.g., benzannulation, electron donating/withdrawing
q
In celebration of the 60th birthday of Professor Malcolm H. Chis-
holm, and his more than 35 years of outstanding chemical achieve-
ment.
*
Corresponding author.
E-mail address: zaleski@indiana.edu (J.M. Zaleski).
0277-5387/$ - see front matter ꢁ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.poly.2005.07.032

S. Bhattacharyya et al. / Polyhedron 25 (2006) 550–558
551
In parallel, 1,2-bis(diphenylphosphinoethynyl)benzene
was shown to cyclization spontaneously upon binding
to divalent platinum or palladium [16]. In this case, later
work on the same ligand showed that this effect was
dependent upon the geometry of the metal center
[17,18]; for chelating enediyne motifs, tetrahedral struc-
tures result in high cyclization temperatures relative to
the much more facile reactivity of cis tetragonal or pla-
nar analogues [17,19,20]. The same guidelines also apply
to a set of tetradentate enediyne ligands, but the effects
of geometry on the cyclization temperature are not as
pronounced [20–22]. Electronic effects, such as ancillary
halogen lone pair donation [23,24], p-complexation at
the alkyne carbons [25,26], as well as coordination of
electron withdrawing, high valent metals near the alkyne
termini (e.g., MoIV–Sꢀ coordination), have also been
shown to be capable of driving enediyne reactivity at
ambient temperature [27]. The emerging theme from this
body of work is that ligand design, and correspondingly,
choice of metal ion activator, are critically important to
the resulting Bergman cyclization kinetics and
thermodynamics.
The conceptual developments highlighted above have
fueled the syntheses of more diverse tetradentate chelat-
ing ligands containing mixed N,S coordination [28].
Structures containing coordinatively fluxional thiother
ligands have not been previously examined and may yield
unusual reactivity due to the weak thioether ligand. This
work reports the syntheses of two benzannulated tetrad-
entate thioether–imine (quinoline, pyridine) enediyne li-
gands and their corresponding Cu(I), Ag(I), Cu(II),
and Pd(II) complexes. The thermal reactivities of these
complexes are unusual and suggest an atypical reaction
mechanism based on fluxional thioether coordination.
resonance of the solvent as an internal reference. Elemen-
tal analyses on all samples were obtained from Robertson
Microlit Laboratories Inc., and mass spectra were mea-
sured at the Indiana University Mass Spectrometer Facil-
ity. Differential scanning calorimetry (DSC) traces were
recorded on a V4.1 Dupont 910 calorimeter coupled to
a Dupont thermal analyst 2100, or on a TA Instruments
Q10 series calorimeter at a heating rate of 10 ꢀC minꢀ1
.
2.2. Ligand syntheses
2.2.1. 1,2-Bis(3-(quinolin-8-ylsulfanyl)prop-1-ynyl)-
benzene (L, 1)
A flask was charged with thioxine hydrochloride (1.6 g,
8.12 mmol) and dry dimethylformamide (30 mL). Anhy-
drous potassium carbonate (4.0 g, 28.9 mmol) was then
added and the mixture was allowed to stir for 5 h at
20 ꢀC. A solution of 1,2-bis-(3-bromopropynyl)benzene
(1.2 g, 3.87 mmol) in dry dimethylformamide (20 mL)
was then added dropwise over 1 h. The reaction mixture
was stirred for 3 h at ambient temperature. After comple-
tion of the reaction, the solid was filtered and washed with
dimethylformamide (2 · 10 mL). The combined organic
layers were then poured into water (300 mL) and then ex-
tracted with methylene chloride (50 mL). The organic
layer was washed with water (3 · 200 mL) and dried over
anhydrous sodium sulfate. Finally, the solvent was re-
moved under reduced pressure. The crude product was
purified on a silica gel column with 1% methanol in meth-
ylene chloride. Yield: 0.70 g (35%); yellow solid. 1H NMR
(400 MHz, CDCl3) d (ppm) 8.87 (s, 2H), 8.09 (d, 2H), 7.67
(d, 2H), 7.56 (d, 2H), 7.48 (d, 2H), 7.36 (m, 2H), 7.25 (m,
2H), 7.17 (m, 2H), 3.88 (s, 4H); 13C NMR (400 MHz,
CD3CN) d (ppm) 149.45, 145.74, 137.32, 136.58, 132.12,
128.39, 127.98, 127.02, 125.82, 125.69, 124.81, 121.87,
89.18, 82.03, 20.95. MS (EI, HRMS) 472.1066, Calc.
472.1067 (M+).
2. Experimental
2.1. General considerations
2.2.2. 1,2-Bis(3-(pyridin-2-ylmethylsulfanyl)prop-1-
ynyl)benzene (L0, 2)
Compounds 1,2-bis-(3-bromopropynyl)benzene and
pyridin-2-yl-methanethiol were prepared according to
literature methods [29,30] All the metal salts (e.g., Cu-
(MeCN)4PF6, Cu(BF4)2 Æ XH2O, AgPF6 and Pd(MeCN)4-
(BF4)2) were purchased from Strem. Reactions were
carried out under nitrogen using Schlenk line and
drybox techniques. Hydrocarbon solvents such as
diethyl ether and benzene were dried by distillation over
sodium/benzophenone prior to use. Methylene chloride
and n-butylamine were dried by distillation over calcium
hydride. Organic enediynes were purified by flash chro-
matography using silica gel (200–440 mesh). All other
chemicals and solvents were used as received from
Aldrich, Strem, and Fluka.
A flask was charged with pyridin-2-yl-methanethiol
(1 g, 8.12 mmol) and dry dimethylformamide (30 mL).
Sodium hydroxide (0.36 g, 9.00 mmol) was then added
and the mixture was allowed to stir for 1 h at 0 ꢀC. A
solution of 1,2-bis-(3-bromopropynyl)benzene (1.2 g,
3.87 mmol) in dry dimethylformamide (20 mL) was then
added dropwise over 1 h. The reaction mixture was stir-
red for 3 h at 0 ꢀC. The combined reaction mixture was
poured into water (300 mL) and then extracted with
methylene chloride (50 mL). The methylene chloride
layer was washed with water (3 · 200 mL) and dried over
anhydrous sodium sulfate. Finally, the solvent was re-
moved under reduced pressure. The crude product was
purified on a silica gel column with 1% methanol in
methylene chloride. Yield: 0.60 g (50%); yellow viscous
1H NMR and 13C NMR spectra were recorded on a
VXR 400 NMR spectrometer using the residual proton
1
oil. H NMR (400 MHz, CDCl3) d (ppm) 8.53 (s, 2H),

552
S. Bhattacharyya et al. / Polyhedron 25 (2006) 550–558
1
7.61 (t, 2H), 7.40 (m, 2H), 7.35 (d, 2H), 7.22 (m, 2H), 7.12
(m, 2H), 4.03 (s, 2H), 4.04 (s, 4H); 13C NMR (400 MHz,
CD3CN) d (ppm) 158.17, 149.97, 136.78, 132.38, 128.07,
125.84, 123.35, 122.20, 89.39, 82.43, 37.69, 20.51. MS
(EI, HRMS) 400.5667, Calc. 400.5682 (M+).
[Ag(L2)]PF6 (8): Yield: 75%, H NMR (400 MHz,
CDCl3) d (ppm) 8.50 (d, 2H), 7.73 (t, 2H), 7.48 (d,
2H), 7.25 (d, 2H), 7.15 (m, 4H), 4.27 (s, 4H), 3.74 (s,
4H); MS (ESI, HRMS): m/z 507.0105 (107Ag), Calc.
507.0113 (M ꢀ PF6).
[Cu(L2)](BF4)2 (9): Yield: 70%, MS (ESI, HRMS):
m/z 463.0360(63Cu), Calc. 463.0358 (M ꢀ 2BF4).
[Pd(L2)](BF4)2 (10): Yield: 90%, 1H NMR (400 MHz,
CDCl3) d (ppm) 8.58 (d, 2H), 8.05 (t, 2H), 7.76 (d, 2H),
7.36 (d, 2H), 7.32 (m, 4H), 4.98 (d, 2H), 4.61 (d, 2H),
4.43 (d, 2H), 4.10 (d, 2H). Anal. Calc. for C24H20N2S2-
Pd(BF4)2: C, 42.35; H, 2.96; N, 4.11. Found: C, 42.11;
H, 3.07; N, 4.31%.
2.3. Metalloenediyne syntheses
Metalloenediyne complexes (3–10) are all prepared
by reaction of ligands 1 and 2 with appropriate metal
salts following the same general method. The case of
[Cu(L1)]PF6 (3) is described in detail below.
Ligand 1 (0.25 mmol) was dissolved in 6 mL chloro-
form and an equimolar amount of the (MeCN)4CuPF6
was separately dissolved in 6 mL of MeCN. To a 100
mL Schlenk flask containing 25 mL MeCN, these ligand
and metal salt solutions were slowly and simultaneously
added dropwise (to avoid polymer formation) under
nitrogen at ambient temperature. After the complete
addition, the mixture was stirred for 30 min at ambient
temperature. The volume of the solution was reduced to
15 mL and filtered to remove the insoluble components.
To the filtrate, 50 mL of dry ether was added and the
resulting solution kept in freezer for 3 h. The yellow pre-
cipitate was subsequently collected by filtration. Yield:
2.4. X-ray crystallography
X-ray quality crystals of 1 were grown by slow evap-
oration from a methanol:dichloromethane mixture (1:1).
For complexes 3, 6, 7, 8a, single crystals were obtained
by slow vapor diffusion of diethyl ether into an acetoni-
trile solution of the complex at 0 ꢀC. Suitable crystals
were placed onto the tip of a 0.1 mm diameter glass
capillary and mounted on a SMART6000 (Bruker) at
120 K. The data collection was carried out using Mo
Ka radiation (graphite monochromator) and a detector
distance of 5.0 cm. A randomly oriented region of reci-
procal space was surveyed to at least the extent of a
hemisphere, collection several major sections of frames
with 0.30ꢀ steps in x at 3 different / settings and a detec-
tor position of ꢀ43ꢀ in 2h. Final cell constants were cal-
culated from the xyz centroids of strong reflections from
the actual data collection after data reduction. The
intensity data were corrected for absorption [31].
Space groups were determined based on intensity sta-
tistics and systematic absences. The structures were
solved using SIR-92 [32] and refined with SHELXL-97
[33]. Direct-methods solutions were calculated which
provided most non-hydrogen atoms from the E-map.
Full-matrix least squares/difference Fourier cycles were
performed, which located the remaining non-hydrogen
atoms. All non-hydrogen atoms were refined with aniso-
tropic displacement parameters. The hydrogen atoms
were placed in ideal positions and refined as riding
atoms with individual or relative isotropic displacement
parameters.
1
80% H NMR (400 MHz, CD3CN) d (ppm) 8.81 (d,
2H), 8.37 (m, 4H), 8.11 (d, 2H), 7.81 (t, 2H), 7.51 (m,
2H), 7.10 (m, 2H), 6.99 (m, 2H), 4.23 (s, 4H). MS
(ESI, HRMS): m/z 535.0360, Calc. 535.0364 (M ꢀ PF6).
[Ag(L1)]PF6 (4): Yield: 70% 1H NMR (400 MHz,
CD3CN) d (ppm) 8.83 (d, 2H), 8.33 (d, 2H), 8.09 (d,
2H), 7.95 (d, 2H), 7.68 (t, 2H), 7.44 (m, 2H), 7.12 (m,
4H), 4.01 (s, 4H); 13C NMR (400 MHz, CD2Cl2) d
(ppm) 154.47, 146.55, 140.47, 138.62, 132.91, 132.53,
130.97, 130.89, 129.89, 128.77, 125.38, 123.69, 89.25,
85.42, 30.32. MS (EI): m/z 579 (107Ag), 581(109Ag)
(M ꢀ PF6). Anal. Calc. for C30H20N2S2AgPF6: C, 49.67;
H, 2.78; N, 3.86. Found: C, 49.81; H, 3.03; N, 3.64%.
[Cu(L1)](BF4)2 (5):Yield:65%, MS(EI):m/z535(63Cu),
537(65Cu) (M ꢀ 2BF4). Anal. Calc. for C30H20N2S2-
Cu(BF4)2: C, 50.76; H, 2.84; N, 3.94. Found: C, 50.25;
H, 2.84; N, 3.94%.
1
[Pd(L1)](BF4)2 (6): Yield: 90% H NMR (400 MHz,
CDCl3) d (ppm) 9.21 (d, 2H), 8.87 (d, 2H), 8.66 (d,
2H), 8.44 (d, 2H), 8.08 (t, 2H), 7.87 (m, 2H), 7.29 (m,
2H), 7.13 (m, 2H), 4.82 (d, 2H), 4.56 (d, 2H); 13C
NMR (400 MHz, CDCl3) d (ppm) 159.68, 153.76,
144.40, 139.03, 134.73, 132.57, 132.13,130.97, 127.22,
125.70, 123.63, 122.25, 89.88, 87.78, 39.32. Anal. Calc.
for C30H20N2S2Pd(BF4)2: C, 47.87; H, 2.67; N, 3.78.
Found: C, 47.11; H, 2.67; N, 3.65%.
3. Results
3.1. Synthesis and characterization
1
[Cu(L2)]PF6 (7): Yield: 80%, H NMR (400 MHz,
Acyclic enediyne ligands 1 and 2 were prepared by
mixing 1,2-bis-(3-bromopropynyl)benzene [29] in DMF
with the pre-formed Na+ or K+ salt of the desired thiol
in a nucleophilic substitution reaction (Scheme 1).
Strong base (NaOH) was employed for the preparation
CDCl3) d (ppm) 8.43 (d, 2H), 7.67 (d, 2H), 7.45 (d,
2H), 7.21 (d, 2H), 7.17 (m, 4H), 4.34 (s, 4H), 3.81 (s,
4H). Anal. Calc. for C24H20N2S2CuPF6: C, 47.33; H,
3.31; N, 4.60. Found: C, 47.11; H, 3.67; N, 4.35%.

S. Bhattacharyya et al. / Polyhedron 25 (2006) 550–558
553
Br
S
S
N
N
K2CO3, DMF
+
2
Br
N
SH
HCl
1(L)
S
S
N
Br
Br
NaOH, DMF
+
2
N
N
SH
2(L')
Scheme 1. Syntheses of tetradentate enediyne ligands 1 (L) and 2 (L0).
of 2 due to the sluggish reactivity of pyridin-2-yl-metha-
nethiol with K2CO3. Both ligands can be isolated in
moderate yields (35–50%).
general method. In order to avoid possible polymer for-
mation ligand and appropriate metal salt were dissolved
separately and simultaneously added at room tempera-
ture maintaining sufficient dilution. The reaction mix-
ture was subsequently filtered to eliminate insoluble
materials. The desired metalloenediyne complexes were
isolated by recrystallization from diethylether in very
good yields (70–90%).
Weakly coordinating, thioether metalloenediynes of
soft metals (e.g., Pd(II), Cu(I), Ag(I)) can be readily pre-
ꢀ
pared by complexation with the appropriate PF6 (e.g.,
Cu+, Ag+) or BF4ꢀ (Cu2+, Pd2+) salts (Scheme 2). These
ions form the tetrahedral or tetragonal/square planar
geometries that modulate the enediyne conformation.
All the metal complexes were prepared following same
1
Both ligands display well-resolved H and 13C NMR
1
spectra in solution. In H NMR spectra, all aromatic
2+
n+
S
S
N
-
Mn+ nX-
N
M
Pd(BF4)2
Pd
1
N
N
S
S
3 Mn+ = Cu+; nX- = PF6
-
6
4 Mn+ = Ag+; nX- = PF6
-
-
5 Mn+ = Cu2+; nX- = 2BF4
n+
2+
S
S
N
-
Mn+ nX-
N
M
Pd(BF4)2
Pd
2
N
N
S
S
7 Mn+ = Cu+; nX- = PF6
-
10
-
8 Mn+ = Ag+; nX- = PF6
-
9 Mn+ = Cu2+; nX- = 2BF4
Scheme 2. Preparation of metalloenediynes complexes 3–10.

554
S. Bhattacharyya et al. / Polyhedron 25 (2006) 550–558
resonances appear between 7 and 9 ppm while the meth-
ylene protons adjacent to the enediyne linkage are
detected below 3.8 ppm. Upon diamagnetic metal ion
complexation, downfield shifts of all proton signals are
observed. Due to thioether coordination and metal ion
proximity, the methylene protons show dramatic shifts
upon complexation. For Cu(I) and Ag(I), the downfield
shift of the methylene protons signal is ꢁ0.4 ppm, while
for the Pd(II) complex, the shift is slightly larger (ꢁ0.7
ppm). In both cases, the shift serves as a useful diagnos-
tic of metal coordination by the thioether linkage.
Similarly, in the 13C NMR spectra of these metal-
loenediyne complexes, the methylene carbon resonance
is also markedly shifted from the free ligand value
(ꢁ20 ppm) to ꢁ30 ppm for the Ag(I) complex and
ꢁ39 ppm for the Pd(II) analogue, thereby revealing
the Lewis acidity/electron withdrawing capability of
the metal ion as a function of oxidation state. All other
resonances are observed as expected.
C25
C27
C22
C8
N2
C7
C12
C9
C6
C3
S2
C5
C4
C11
C10
3.2. Molecular structure
C15
Ligand 1 was crystallized by slow evaporation of
methanol-chloroform solution (Fig. 1, Table 1). X-ray
quality crystals of metalloenediyne complexes of ligands
1 (3, 6), and 2 (7, 8a) were obtained by vapor diffusion
of ether into an acetonitrile solution of the complexes
(Figs. 2 and 3, Table 1). The Ag(I) complex 8 did not
readily yield X-ray quality crystals; however counter
ion exchange of PF6 for SbF6 afforded the desired single
crystals (8a). Selected bond parameters are listed in
Tables 2 and 3.
C13
S1
C2
C18
C1
C21
N1
1
Fig. 1. X-ray structure and atom labeling scheme for 1. Thermal
In the X-ray structure of ligand 1 (Fig. 1) the two
thioxine (8-mercaptoquinoline) groups lie in a syn con-
ellipsoids are illustrated at 40% probability. The alkyne termini
˚
separation (C2ꢂ ꢂ ꢂC7) is 3.76 A.
Table 1
Crystallographic data for 1,3 Æ CH2Cl2,6 Æ CH3CN, 7, and 8a
1
3 Æ CH2Cl2
6 Æ CH3CN
7
8a
Empirical formula
Formula weight
Color of crystal
Crystal system
Space group
C
472.60
30H20N2S2
C31H22Cl2N2S2CuPF6 C32H23B2F8N3PdS2 C24H20CuF6N2PS2 C24H20N2S2AgSbF6
766.04
793.67
609.05
744.16
colorless
monoclinic
P21/n
21.9054(19)
4.9336(4)
23.425(2)
yellow
monoclinic
P21/n
yellow
monoclinic
P21/c
yellow
monoclinic
P21/c
colorless
monoclinic
P21/c
˚
a (A)
12.8102(11)
12.0201(11)
21.365(2)
14.912(2)
11.8170(16)
17.597(2)
13.1335(5)
12.6210(5)
14.7333(6)
13.1428(7)
12.8945(7)
15.2240(8)
˚
b (A)
˚
c (A)
a (ꢀ)
b (ꢀ)
c (ꢀ)
115.647(2)
105.565(3)
91.245(4)
94.699(1)
94.344(1)
3
˚
V (A )
2282.1(3)
4
1.375
0.256
3169.2(5)
4
1.605
1.102
3100.0(7)
4
1.701
0.812
2433.95(17)
4
1.662
1.198
2572.62(2)
4
1.921
2.034
Z
qcalc (g cmꢀ3
)
Linear absorption
coefficient (mmꢀ1
)
Goodness-of-fita on F2
1.022
1.034
1.067
1.053
1.068
Final R indicesb [I > 2r(I)]
R1 = 0.0337,
wR2 = 0.0804
R1 = 0.0474,
wR2 = 0.0875
R1 = 0.0397,
wR2 = 0.0915
R1 = 0.0580,
wR2 = 0.1035
R1 = 0.0286,
wR2 = 0.0699
R1 = 0.0349,
wR2 = 0.0753
0.885 and ꢀ0.539
R1 = 0.0331,
wR2 = 0.838
R1 = 0.0424,
wR2 = 0.0903
0.533 and ꢀ0.314
R1 = 0.0414,
wR2 = 0.1317
R1 = 0.0519,
wR2 = 0.1385
2.188 and ꢀ1.939
R indicesc (all data)
Largesꢀt 3difference peak and hole 0.300 and ꢀ0.276 0.954 and ꢀ0.610
˚
(e A
)
P
2
Goodness-of-fit ¼ ½ ½wðF 2o ꢀ F 2Þ ꢃ=Nobservns ꢀ NparamsÞꢃ1=2, all data.
a
b
c
c
P
P
R1 ¼ ðjF oj ꢀ jF cjÞ= jF oj.
P
P
2
2 1=2
wR2 ¼ ½ ½wðF 2o ꢀ F c2Þ ꢃ= ½wðF 2oÞ ꢃꢃ
.

S. Bhattacharyya et al. / Polyhedron 25 (2006) 550–558
555
C1
C8
C17
C13
N1
C2
C13
C1
S1
C3
C2
C21
N1
C9 C4
S1
Cu1
C3
C4
C19
C7
C9
Pd1
C20
C5
C6
C29
C12
C23
N2
C29
C28
S2
C5
C6
C7
S2
C12
C8
C22
C26
3
6
Fig. 2. X-ray structures and atom labeling scheme for 3 and 6. Thermal ellipsoids are illustrated at 40% probability. Counter ions are omitted for
˚
˚
clarity. The alkyne termini separation (C2ꢂ ꢂ ꢂC7) 3: 4.05 A; 6: 3.98 A.
C8
C19
C8
S2
C7
C19
C7
S2
C6
C6
C12
C10
C9
C5
C5
N2
Ag1
N2
C2
C2
C1
C1
N1
C4
C4
C3
C3
C9
C24
Cu1
S1
C24
S1
N1
C16
C16
C13
C13
8a
7
Fig. 3. X-ray structures and atom labeling scheme for 7 and 8a. Thermal ellipsoids are illustrated at 40% probability. Counter ions are omitted for
˚
˚
clarity. Alkyne termini separation (C2ꢂ ꢂ ꢂC7) 7: 3.95 A; 8a: 4.09 A.
figuration with respect to the enediyne plane with their
is also a plausible structure, it is energetically uphill by
ꢁ6–9 kcal/mol from the trans geometry [28]. Despite
the coordination geometry differences between the com-
plexes, the alkyne termini separation (C2ꢂ ꢂ ꢂC7) is almost
rings oriented at ꢁ45ꢀ to each other. The alkyne termini
˚
(C2ꢂ ꢂ ꢂC7) separation is relaxed (3.76 A) and shows no
evidence of steric crowding as occurs in more rigid
enediyne constructs with large metal binding functional-
ities near the alkyne termini, due in part to the flexibility
engendered by the additional methylene linkage [20].
For d10 metalloenediynes 3, 7 (Cu(I)), and 8a (Ag(I)),
the N2S2 ligand donor atoms coordinate in a distorted
tetrahedral fashion. The N1–M–N2 and S1–M–S2 bond
angles range from ꢁ106ꢀ to 113ꢀ and ꢁ113ꢀ to 123ꢀ,
respectively, reflective of the structural flexibility of the
ligand and the elongated M–S bond lengths (ꢁ0.15–
˚
same for all complexes (ꢁ4 A). This arises from the large
S1–M–S2 bite angles that range from 113ꢀ to 175ꢀ across
all complexes.
3.3. Thermal Bergman cyclization
The thermal Bergman cyclization temperatures of
these enediyne compounds have been measured by dif-
ferential scanning calorimetry (DSC) and are given in
Table 4. DSC is a convenient technique for comparing
and correlating the thermal reactivities of enediynes,
and is especially useful for complexes that react with
solvent and decompose under thermal conditions in
solution.
˚
0.3 A) over their M–N counterparts (Tables 2 and 3).
In contrast, the Pd(II) complex 6 is square planar with
the two weakly coordinating thioether linkages in a
trans relationship leading to S1–Pd–S2 and N1–Pd–N2
bond angles of ꢁ175ꢀ. Although the cis conformation

556
S. Bhattacharyya et al. / Polyhedron 25 (2006) 550–558
Table 2
Table 3
˚
˚
Selected bond lengths (A) and angles (ꢀ) of 1, 3, and 6
Selected bond lengths (A) and angles (ꢀ) of 7 and 8a
Distances
1
Distances
7
S1–C1
C1–C2
C2–C3
C3–C4
C4–C5
S1–C13
1.8186(18)
1.457(2)
1.192(2)
1.438(2)
1.419(2)
1.7608(17)
Cu1–N1
Cu1–N2
Cu1–S1
Cu1–S2
S1–C1
C1–C2
C2–C3
2.039(2)
2.014(2)
2.2832(6)
2.3348(7)
1.829(2)
1.457(3)
1.199(3)
3
Cu1–N2
Cu1–S1
Cu1–N1
Cu1–S2
S1–C1
C1–C2
C2–C3
2.024(2)
2.2575(8)
2.043(2)
2.2851(8)
1.844(3)
1.452(4)
1.196(4)
8a
Ag1–N1
Ag1–N2
Ag1–S2
Ag1–S1
S1–C1
C1–C2
C2–C3
2.288(4)
2.341(4)
2.4905(12)
2.5635(13)
1.821(5)
1.465(7)
1.193(7)
6
Pd1–N2
Pd1–N1
Pd1–S1
Pd1–S2
S1–C1
C1–C2
C2–C3
2.0468(18)
2.0440(18)
2.3171(6)
2.2871(6)
1.860(2)
Angles
7
N1–Cu1–N2
N1–Cu1–S1
N2–Cu1–S1
S1–Cu1–S2
S1–C1–C2
Cu1–S1–C1
112.88(8)
88.19(6)
135.06(6)
112.61(2)
114.79(16)
103.35(8)
1.450(3)
1.189(5)
Angles
1
8a
C1–S1–C13
S1–C1–C2
C1–C2–C3
C2–C3–C4
C3–C4–C5
S1–C13–C21
102.41(8)
N1–Ag1–N2
N1–Ag1–S1
N2–Ag1–S1
S1–Ag1–S2
S1–C1–C2
Ag1–S1–C1
108.51(14)
146.23(11)
80.32(10)
123.04(4)
108.8(3)
113.61(13)
176.85(18)
174.16(18)
118.80(15)
114.89(12)
101.6(2)
3
N1–Cu1–N2
N1–Cu1–S1
N1–Cu1–S2
S1–Cu1–S1
S1–C1–C2
Cu1–S1–C1
105.92(10)
88.08(7)
120.60(7)
118.83(3)
107.1(2)
Table 4
Bergman cyclization temperatures of 1–10 (DSC)
Compound
Cyclization temperature (ꢀC)
L (1)
230
218
214
186
190
107.47(10)
[CuL]PF6 (3)
[AgL]PF6 (4)
[CuL](BF4)2 (5)
[PdL](BF4)2 (6)
6
N1–Pd1–N2
N1–Pd1–S1
N1–Pd1–S2
S1–Pd1–S2
S1–C1–C2
Pd1–S1–C1
174.91(7)
84.88(5)
93.79(5)
175.91(2)
108.65(16)
102.12(7)
L0 (2)
228
215
211
188
189
[CuL0]PF6 (7)
[AgL0]PF6 (8)
[CuL0](BF4)2 (9)
[PdL0](BF4)2 (10)
Upon heating, ligand 1 exhibits a sharp, endothermic
peak at 105 ꢀC characteristic of a melting transition. At
higher temperature, a broad exothermic peak is also
observed at 230 ꢀC, which is characteristic of and at
the expected temperature for the Bergman cyclization
reaction of the neat liquid (Fig. 4). Ligand 2, which is
a liquid at ambient temperature, displays a similar exo-
therm at 226 ꢀC, indicative of diradical formation by a
similar construct.
tures below the cyclization reaction. For these com-
plexes, cyclization occurs in the solid state, whereas
the ligands react as neat liquids. Thus, direct compari-
son of the specific cyclization temperatures between
the ligands and complexes cannot be readily made.
However, it should be noted that the ligand cyclization
temperatures would likely be considerably higher if
encapsulated within an inert matrix, thereby illustrating
the ability of the metal to reduce the cyclization temper-
atures relative to the free ligands [17,19,20].
Unlike 1–2, the metalloenediyne complexes are all
solids and do not show melting endotherms at tempera-

S. Bhattacharyya et al. / Polyhedron 25 (2006) 550–558
557
thioether-coordinated enediyne framework. These con-
structs are very intriguing because they generate possi-
bilities for using the thioether to geometrically,
electronically, or substitutionally modulate the bound
enediyne conformation and hence the Bergman cycliza-
tion temperature.
Earlier studies have shown that tetradentate amine/
imine enediyne ligands [21,22,34] could give rise to, in
some cases, extremely facile Bergman cyclization at
ambient temperature [34]. In this sense, use of the thio-
ether linkage as a mechanism to draw the alkyne termini
into close proximity has attractive potential. Yet similar
studies also revealed the importance of metal geome-
try in controlling coordinated enediyne conformation
[16–22,35]. It is now well-established that cis-planar
structures often have markedly lower cyclization tem-
peratures than their tetrahedral or more rarely, their
trans-planar [35] counterparts. As shown in Table 4,
these thioether coordinated complexes do not adhere
to this clearly defined paradigm, but rather exhibit cycli-
zation temperatures for the trans-planar complexes that
are ꢁ25 ꢀC lower than their tetrahedral counterparts.
Since these complexes exhibit the same trend as their
non-benzannulated structures, albeit ꢁ50–60 ꢀC lower
in temperature [28], the variational thermal reactivities
of these constructs is clearly not disrupted by benzann-
ulation. With the wealth of data of metalloenediyne
complexes and the established relationships with geo-
metric structure, why then do the trans planar com-
plexes react more readily than the tetrahedral structures?
Bergman cyclization temperatures also exhibit elec-
tronic contributions to reduction or enhancement of the
thermal barrier to cyclization [36–40]. More specifically,
electron withdrawing functionalities at the alkyne termini
have been shown both theoretically [38] and experimen-
tally [40] to decrease the onset temperature for the cycliza-
tion reaction. Although these effects are relevant in
metalloenediyne constructs, their observation typically
requires careful synthetic isolation of geometric and elec-
tronic components by locking the ligand conformation in
order to systematically evaluate such contributions [24].
Within this theme, such enhancements in reactivity have
recently been shown to be accessible by high valent metal
ion coordination (e.g., MoIV–Sꢀ) near the alkyne termini
[27]. Ligands such as 1 and 2 displa- ying thioether coordi-
nation in the vicinity of the alkyne termini have potential
for just this mode of electronic enediyne activation. How-
ever, in these cases, the neutral charge of the thioether
leads to long metal–ligand bond lengths and poor orbital
overlap, which is the basis for electronic activation. Addi-
tionally, the tetrahedral or trans planar geometry (e.g.,
Pd(II)) negates significant electronic enhancement via less
than favorable geometric considerations. Thus, these fun-
damental guidelines for metalloenediyne reactivity do not
explain the thermal chemistry of thioether-coordinated
constructs such as 3–10.
Fig. 4. Differential scanning calorimetry traces for the thermal
cyclization of (a) 1, 2 and (b) 6, 7 in the solid state.
To this end, the cyclization temperatures of monova-
lent metalloenediyne complexes such as those of Cu(I)
and Ag(I) displaying tetrahedral geometries (211–218 ꢀC)
are ꢁ20–25 ꢀC higher than their corresponding
divalent, tetragonal and trans square planar complexes
of Cu(II) and Pd(II) (186–190 ꢀC). This trend is invari-
ant to ligand and virtually identical for both sets of com-
plexes derived from ligands 1 and 2, despite the change
of pyridine for quinoline coordination. It is also note-
worthy that the alkyne termini separation in the X-ray
structures increases upon metal coordination, while the
cyclization temperatures actually decrease. This behav-
ior is in stark contrast to the typical relationship
between cyclization temperature and alkyne termini sep-
aration that has been well-established for analogous
chelating metalloenediyne constructs [18–22].
4. Discussion
Ligands 1 and 2, and their corresponding metal-
loenediyne complexes 3–10 represent a new class of

558
S. Bhattacharyya et al. / Polyhedron 25 (2006) 550–558
[3] D.S. Rawat, J.M. Zaleski, Synlett (2004) 393.
In very recent work on the non-benzannulated deriv-
[4] M.D. Lee, T.S. Dunne, C.C. Chang, G.A. Ellestad, M.M. Siegal,
G.O. Morton, W.J. McGahren, D.B. Borders, J. Am. Chem. Soc.
109 (1987) 3466.
[5] M.D. Lee, J.K. Manning, D.R. Williams, N.A. Kuck, R.T. Testa,
D.B. Borders, J. Antibiot. 42 (1989) 1070.
[6] M. Konishi, H. Ohkuma, T. Tsuno, T. Oki, G.D. VanDuyne, J.
Clardy, J. Am. Chem. Soc. 112 (1990) 3715.
[7] J. Golik, J. Clardy, G. Dubay, G. Groenewold, H. Kawaguchi,
M. Konishi, B. Krishnan, H. Ohkuma, K. Saitoh, T.W. Doyle,
J. Am. Chem. Soc. 109 (1987) 3461.
[8] J. Golik, G. Dubay, G. Groenewold, H. Kawaguchi, M. Konishi,
B. Krishnan, H. Ohkuma, K. Saitoh, T.W. Doyle, J. Am. Chem.
Soc. 109 (1987) 3462.
[9] R.L. Funk, E.R.R. Young, R.M. Williams, M.F. Flanagan, T.L.
Cecil, J. Am. Chem. Soc. 118 (1996) 3291.
[10] A. Evenzahav, N.J. Turro, J. Am. Chem. Soc. 120 (1998) 1835.
[11] T. Kaneko, M. Takahashi, M. Hirama, Angew. Chem., Int. Ed.
Engl. 38 (1999) 1267.
atives of 3–6, detailed NMR and DFT studies implicated
elevated temperature formation of a cis isomer with re-
duced alkyne termini separation by dissociation of the
thioether ligand and rotation about the M–N bond
[28]. This species then serves as the active precursor to
Bergman cyclization. In fact, ambient temperature cycli-
zation can be initiated via competitive ligand exchange
by addition of 1 eq of pyridine. Under these conditions,
the methylene protons from the dissociated thioether li-
gand appear in the 1H NMR (d 3.9 ppm) and new reso-
nances from the cyclized product between d 6 and 7 ppm
emerge [28]. In light of the increased cyclization temper-
atures and the parallel thermal reactivity trends between
related metalloenediyne complexes, these benzannulated
derivatives likely follow the same reaction pathway.
[12] M. Nath, M. Pink, J.M. Zaleski, J. Am. Chem. Soc. 127 (2005) 478.
[13] A. Basak, S. Mandal, S.S. Bag, Chem. Rev. 103 (2003) 4077.
[14] B. Ko¨nig, H. Hollnagel, B. Ahrens, P.G. Jones, Angew. Chem.,
Int. Ed. Engl. 34 (1995) 2538.
5. Conclusion
[15] B. Ko¨nig, W. Pitsch, I. Thondorf, J. Org. Chem. 61 (1996) 4258.
[16] B.P. Warner, S.P. Millar, R.D. Broene, S.L. Buchwald, Science
269 (1995) 814.
[17] N.L. Coalter, T.E. Concolino, W.E. Streib, C.G. Hughes, A.L.
Rheingold, J.M. Zaleski, J. Am. Chem. Soc. 122 (2000) 3112.
[18] E.W. Schmitt, J.C. Huffman, J.M. Zaleski, Chem. Commun.
(2001) 167.
[19] P.J. Benites, D.S. Rawat, J.M. Zaleski, J. Am. Chem. Soc. 122
(2000) 7208.
[20] D.S. Rawat, P.J. Benites, C.D. Incarvito, A.L. Rheingold, J.M.
Zaleski, Inorg. Chem. 40 (2001) 1846.
[21] T. Chandra, M. Pink, J.M. Zaleski, Inorg. Chem. 40 (2001) 5878.
[22] T. Chandra, R.A. Allred, B.J. Kraft, L.M. Berreau, J.M. Zaleski,
Inorg. Chem. 43 (2004) 411.
[23] S. Bhattacharyya, A.E. Clark, M. Pink, J.M. Zaleski, Chem.
Commun. (2003) 1156.
[24] S. Bhattacharyya, J.M. Zaleski, Curr. Top. Med. Chem. 4 (2004)
1637.
[25] J.M. OꢀConnor, L.I. Lee, P. Gantzel, A.L. Rheingold, K.-C.
Lam, J. Am. Chem. Soc. 122 (2000) 12057.
We have prepared two novel thioether-coordinated
enediyne ligands and a set of metalloenediyne complexes
of Cu(I), Ag(I), Cu(II), and Pd(II). The d10 species exhibit
distorted but near tetrahedral geometries with weak thio-
ether coordination, while the d9 and d8 species are likely
trans tetragonal and square planar, respectively. Remark-
ably, the latter structures exhibit Bergman cyclization
temperatures ꢁ25 ꢀC lower than their tetrahedral counter-
parts. These values are much lower than those expected
for complexes with trans geometries. The rationale for this
chemical behavior can be traced to the fluxional nature of
the thioether linkage and a trans to pseudo cis conforma-
tional change about the metal center via M–N bond rota-
tion. This reactivity reveals the complex and untapped
potential for using metal coordination to activate
enediyne ligands toward Bergman cyclization.
[26] J.M. OꢀConnor, S.J. Friese, M. Tichenor, J. Am. Chem. Soc. 124
(2002) 3506.
[27] S. Bhattacharyya, M. Pink, M.-H. Baik, J.M. Zaleski, Angew.
Chem., Int. Ed. Engl. 44 (2005) 592.
6. Supplementary material
[28] S. Bhattacharyya, D.F. Dye, M. Pink, J.M. Zaleski, Chem.
Commun. (submitted).
[29] H. Hopf, P.G. Jones, P. Bubenitschek, C. Werner, Angew.
Chem., Int. Ed. Engl. 34 (1995) 2367.
[30] R. Deziel, E. Malenfant, Bioorg. Med. Chem. Lett. 8 (1998) 1437.
[31] R. Blessing, Acta Crystallogr. A51 (1995) 33.
[32] A. Altomare, G. Cascarno, C. Giacovazzo, A. Gualardi, J. Appl.
Crystallogr. 26 (1993) 343.
[33] SHELXTL-PLUS V6.14, BrukerAnalyticalX-raySystems,Madison, WI.
[34] D.S. Rawat, J.M. Zaleski, J. Am. Chem. Soc. 123 (2001) 9675.
[35] Y.-Z. Hu, C. Chamchoumis, J.S. Grebowicz, R. Thummel, Inorg.
Chem. 41 (2002) 2296.
Crystallographic data for the X-ray structure analy-
ses have been deposited with the Cambridge Crystallo-
graphic Data Center, CCDC Nos. 268 255 (6), 275 063
(7), 275 064 (3), 275 065 (1), and 275 066 (8a).
Acknowledgement
The support of the National Institutes of Health (R01
GM62541) is gratefully acknowledged.
[36] G.B. Jones, G.W. Plourde II, Org. Lett. 2 (2000) 1757.
[37] G.B. Jones, P.M. Warner, J. Am. Chem. Soc. 123 (2001) 2134.
[38] M. Prall, A. Wittkopp, A.A. Fokin, P.R. Schreiner, J. Comp.
Chem. 22 (2001) 1605.
References
[39] I.V. Alabugin, M. Manoharan, S.V. Kovalenko, Org. Lett. 4
(2002) 1119.
[40] M. Nath, J.C. Huffman, J.M. Zaleski, J. Am. Chem. Soc. 125
(2003) 11484.
[1] N. Darby, C.U. Kim, J.A. Salaun, K.W. Shelton, S. Takada, S.
Masamune, Chem. Commun. (1971).
[2] R.G. Bergman, Acc. Chem. Res. 6 (1973) 25.