Me3SiC≡C-CMe=CH2 copper(I) β-diketonates: Synthesis, solid state structure, and low-temperature chemical vapour deposition

Article abstract of DOI:10.1016/j.jorganchem.2005.04.054

The ene-yne copper(I) β-diketonates [(η²-TMSMB) Cu(β-diketonate)] (β-diketonate = acetylacetonate/acac, 5; = 1,3-di-tert-butylacetonate/dbac, 6; = 1,1,1,5,5,5-hexafluoroacetylacetonate/ hfac, 7; TMSMB = 4-TriMethylSilyl-2-Methyl-But-1-ene-3-yne, Me ₃SiCC-CMeCH₂) are accessible by the reaction of [(η²-TMSMB)Cu(μ-Cl)]₂ (3) with [Na(β- diketonate)] (β-diketonate = acac, 4a; = dbac, 4b; = hfac, 4c) in a 1:2 molar ratio. Complexes 6 and 7 are also formed, when Cu₂O (8) is reacted with H-hfac (9a) or H-dbac (9b), respectively. The solid state structure of 7 is reported. The copper(I) ion possesses a planar environment caused by the η²-coordinated TMSMB ligand and the chelate-bound hfac group, while the CMeCH₂ entity stays free. The thermal properties of 5-7 were determined by applying ThermoGravimetry (TG) and Differential Scanning Calorimetry (DSC). All complexes decompose in a two-step process beginning at ca. 85 °C. Elimination of TMSMB produces [Cu(β-diketonate)] which disproportionates to give [Cu(β-diketonate)₂] and elemental copper. Preliminary hot-wall Chemical Vapour Deposition experiments (CVD) were carried out with 7. Copper films were deposited onto TiN-coated oxidised silicon wafers at a precursor vaporisation temperature of 50 °C and a deposition temperature of 145 °C. The films were characterised by SEM and EDX.

Full text of DOI:10.1016/j.jorganchem.2005.04.054

Journal of Organometallic Chemistry 690 (2005) 3878–3885
www.elsevier.com/locate/jorganchem
Me3SiC„C–CMe@CH2 copper(I) b-diketonates: Synthesis,
solid state structure, and low-temperature chemical vapour deposition
a
Y. Shen , T. Ruffer , S.E. Schulz , T. Gessner , L. Wittenbecher ,
a
b
b
c
¨
d
H.-J. Sterzel , H. Lang
a,
*
a
Technische Universita¨t Chemnitz, Fakulta¨t fu¨r Naturwissenschaften, Institut fu¨r Chemie, Lehrstuhl Anorganische Chemie,
Straße der Nationen 62, 09111 Chemnitz, Germany
b
Technische Universita¨t Chemnitz, Fakulta¨t fu¨r Elektrotechnik und Informationstechnik, Lehrstuhl fu¨r Mikrotechnologie,
Zentrum fu¨r Mikrotechnologien, Reichenhainer Straße 70, 09126 Chemnitz, Germany
c
BASF Aktiengesellschaft, GC/Z-M 311, 67056 Ludwigshafen, Germany
BASF Aktiengesellschaft, ZAR/A-M 311, 67056 Ludwigshafen, Germany
d
Received 6 December 2004; received in revised form 22 April 2005; accepted 22 April 2005
Available online 18 July 2005
Abstract
The ene-yne copper(I) b-diketonates [(g2-TMSMB)Cu(b-diketonate)] (b-diketonate = acetylacetonate/acac, 5; = 1,3-di-tert-
butylacetonate/dbac, 6; = 1,1,1,5,5,5-hexafluoroacetylacetonate/hfac, 7; TMSMB = 4-TriMethylSilyl-2-Methyl-But-1-ene-3-yne,
Me3SiC„C–CMe@CH2) are accessible by the reaction of [(g2-TMSMB)Cu(l-Cl)]2 (3) with [Na(b-diketonate)] (b-diketonate = acac,
4a; = dbac, 4b; = hfac, 4c) in a 1:2 molar ratio. Complexes 6 and 7 are also formed, when Cu2O (8) is reacted with H-hfac (9a) or
H-dbac (9b), respectively.
The solid state structure of 7 is reported. The copper(I) ion possesses a planar environment caused by the g2-coordinated
TMSMB ligand and the chelate-bound hfac group, while the CMe@CH2 entity stays free.
The thermal properties of 5–7 were determined by applying ThermoGravimetry (TG) and Differential Scanning Calorimetry
(DSC). All complexes decompose in a two-step process beginning at ca. 85 ꢀC. Elimination of TMSMB produces [Cu(b-diketonate)]
which disproportionates to give [Cu(b-diketonate)2] and elemental copper.
Preliminary hot-wall Chemical Vapour Deposition experiments (CVD) were carried out with 7. Copper films were deposited onto
TiN-coated oxidised silicon wafers at a precursor vaporisation temperature of 50 ꢀC and a deposition temperature of 145 ꢀC. The
films were characterised by SEM and EDX.
ꢁ 2005 Elsevier B.V. All rights reserved.
Keywords: Copper(I); b-Diketonate; Ene-yne; CVD; SEM; EDX; X-ray
1. Introduction
as alternative possibility for interconnect metallisation
in future generations of sub-micron integrated circuits
[1–3], which is attributed, for example, to the low resistiv-
ity and superior electro-migration resistance of copper
[4–8]. Copper precursors should be volatile, preferen-
tially monomeric with low molecular weight, and should
be easy to handle before and during the deposition exper-
iments. Among them, [LnCu(b-diketonate)] complexes
(L = 2-electron donor; n = 1, 2) have recently attracted
The Chemical Vapour Deposition (CVD) of pure
copper films from Lewis-base-stabilised copper(I) com-
plexes is substantial, since this technique is considered
*
Corresponding author. Tel.: +49 371 531 1200; fax: +49 371 531
1833.
E-mail address: heinrich.lang@chemie.tu-chemnitz.de (H. Lang).
0022-328X/$ - see front matter ꢁ 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2005.04.054

Y. Shen et al. / Journal of Organometallic Chemistry 690 (2005) 3878–3885
3879
Table 1
Synthesis of 5–7 by the reaction of 3 with 4
attention as CVD precursors, since the O,O-chelate-
bound b-diketonate and the corresponding Lewis-bases
L can be straightforward varied to tailor the volatility
and stability of such systems [1,2,9–12]. As Lewis-bases
L, phosphines, phosphites [11–13], alkenes [14–18],
dienes [10,19], or alkynes [2,20,21] can be used to obtain
an assortment of precursors with variable properties. The
thermal decomposition of [Cu(b-diketonate)] occurs at
low temperatures, generally in the range between 150
and 250 ꢀC to give elemental copper along with [Cu(b-
diketonate)2] [22].
Compound
R
b-Diketonate
Yielda (%)
5
6
7
a
Me
tBu
CF3
acacb
dbacc
hfacd
87
84
79
Based on 3.
b
c
acetylacetonate.
1,3-di-tert-butylacetonate.
1,1,1,5,5,5-hexafluoroacetylacetonate.
d
Recently, [(g2-VTMS)Cu(hfac)] (VTMS = VinylTri-
MethylSilane, hfac = 1,1,1,5,5,5-hexafluoroacetylaceto-
nate) was commercially introduced as efficient, but
thermal labile CVD copper precursor [23]. Other com-
pounds of the same family have demonstrated valuable
results [7,24,25]. Within this family of copper(I) CVD
precursors, the physical properties can easily be varied
by structural modifications, i.e. substitution reactions.
Changing, for example, the Lewis-base L in [LnCu(b-
diketonate)] (n = 1, 2) can directly alter the physical state,
the vapour pressure, the decomposition temperature, and
the long-term stability of the respective precursors [1,10].
For example, [(g2-alkyne)Cu(hfac)] can be used as CVD
precursor, however, a less volatile compound is formed in
the bubbler along with elemental copper [21]. This unsta-
ble complex was thought to be bimetallic in which the
alkyne unit binds to two copper(I) ions [26]. The forma-
tion of such species may be avoided by introducing both,
p-conjugated vinyl or acetylenic units, and electron-with-
drawing groups such as OCH3 or CF3, respectively [27].
The best results so far could be obtained, when the alkyne
unit is conjugated with a vinyl entity [28].
1, Table 1). Complexes 6 and 7 can additionally be pre-
pared by the reaction of H-hfac (9a) and H-dbac (9b)
with Cu2O (8) in presence of TMSMB (2) (Scheme 1).
After appropriate work-up, mononuclear 5–7 can be
isolated in high yield (Table 1) as pale green (5), yellow
(6) or off-white (7) materials of which 5 and 6 are liq-
uids, while 7 melts at 32 ꢀC. Complexes 5–7 are soluble
in most common organic solvents, such as n-pentane,
toluene, chloroform and diethyl ether. They are stable
under inert gas atmosphere for months at ꢀ30 ꢀC. On
exposure to air they gradually decompose during days
to give red solids which are not anymore soluble in com-
mon organic solvents.
2.2. Characterisation
Complexes 5–7 give satisfactory elemental analyses
confirming the composition (TMSMB)Cu(b-diketonate).
1
In addition, FT-IR, H-, 13C{1H}- and 29Si{1H}-NMR
spectroscopic studies were carried out. The solid state
structure of 7 was established by X-ray structure
determination.
We here describe the synthesis and characterisa-
tion of three [(g2-alkyne-ene)Cu(b-diketonate)] species
based on the 4-TriMethylSilyl-2-Methyl-But-1-ene-3-yne
(TMSMB) ligand. The use of one sample as CVD pre-
cursor for the deposition of copper onto oxidised silicon
wafers covered with a CVD TiN film as a barrier is re-
ported as well.
For 5–7 one distinctive mC„C vibration is found at
1930 (5), 1931 (6) or 1964 (7) cmꢀ1 in the IR spectra.
These absorptions are shifted to lower values, when com-
pared with free non-coordinated TMSMB (2172 cmꢀ1
)
[28–30]. The magnitude of this change is related with
the bond order and the nuclearity of the complex and is
influenced by R of the O2C3HR2 units. For 5–7 relatively
small DmC„C values (5: 242, 6: 241, 7: 208 cmꢀ1) are ob-
tained showing that in 7 somewhat weaker copper–car-
bonalkyne bonds are present as in 5 and 6. These data
correspond to other [(hfac)Cu(b-diketonate)] complexes
and are in agreement with DmC„C characteristic for
[(MHY)Cu(b-diketonate)] (MHY = 2-Methyl-1-Hexen-
3-Yne) [27]. Furthermore, two typical ma(CO) and ms(CO)
2. Results and discussion
2.1. Synthesis
The TMSMB-copper(I) b-diketonates 5–7 (TMSMB=
4-TriMethylSilyl-2-Methyl-But-1-ene-3-yne, Me3SiC„C–
CMe@CH2) (Table 1) can be synthesised by treatment
of [(g2-TMSMB)Cu(l-Cl)]2 (3), which itself is accessible
by the reaction of [CuCl] (1) with stoichiometric
amounts of TMSMB (2) in dichloromethane solutions,
with [Na(O2C3HR2)] (4a, R = CH3, acetylacetonate,
acac; 4b, R = tBu, 1,3-di-tert-butylacetonate, dbac; 4c,
R = CF3, 1,1,1,5,5,5-hexafluoroacetylacetonate, hfac)
in the ratio of 1:2 in diethyl ether at ꢀ30 ꢀC (Scheme
absorptions are found between 1520 and 1660 cmꢀ1
,
whereby ma corresponds to the asymmetric and ms to the
symmetric stretching vibration [27,31–34]. DmCO
(DmCO = ma ꢀ ms) smaller than 70 cmꢀ1 points to symmet-
rical, while larger values are typical for a asymmetrical
bonding mode of the b-diketonates [32]. For 5–7 DmCO
resembles to 67 (5, 7) and 34 cmꢀ1 (6) indicating that
the b-diketonates are symmetrical O,O-chelate-bound

3880
Y. Shen et al. / Journal of Organometallic Chemistry 690 (2005) 3878–3885
Me
CH2
Cu
Me3
Si
C
C
C
Cl
Cl
C
C
C
TMSMB (2)
[CuCl]
1/2
Cu
1
[Na(O2C3HR2)] (4)
- NaCl
Si
Me3
H2C
Me
3
Me
CH2
Cu
C
C
C
R
R
O
O
Si
Me3
TMSMB (2)
- 1/2 H2O
5 - 7
1/2 Cu2O
+
C3H2O2R2
8
9
Scheme 1. Synthesis of 5–7 (Table 1).
to copper(I). This structural motif is also favoured in the
solid-state as it could be confirmed by single X-ray struc-
ture analysis (Fig. 1). When one takes the highest fre-
quency in the above mentioned region as an estimate of
the strength of the donor–acceptor interaction between
the b-diketonates and copper, than the strongest car-
bon–oxygen bonds/weakest copper–oxygen interactions
are found for 7 [36].
(5: 5.37, 5.31; 6: 5.55, 5.30; 7: 5.81, 5.22 ppm), when
compared with free TMSMB (2: 5.35, 5.25 ppm) [32].
This is attributed to the g2-coordination of the car-
bon–carbon triple bond to copper(I). Although, the
Me3Si resonance signal (5: 0.31, 6: 0.26, 7: 0.32 ppm)
is somewhat shifted to lower field (i.e. free TMSMB:
0.20 ppm) [29].
The 13C{1H} NMR spectra of 5–7 show that the
TMSMB ligand only binds by the carbon–carbon triple
bond to copper. This is in agreement with the observa-
tion made in the IR spectra of 5–7 (vide supra) and is
typical for other ene-yne species, i.e. [(MHY)Cu(b-
diketonate)] [27].
1
The H NMR spectra of 5–7 are consistent with the
1:1 TMSMB-copper(b-diketonate) stoichiometry. The
methine group gives rise to a distinct resonance signal
at 5.75 (5), 5.72 (6) or 6.13 ppm (7) which is typical
for b-diketonates. The resonance signals for the
TMSMB ene unit are somewhat shifted to lower field
The Ca and Cb TMSMB carbons (Me3SiCa„Cb) are
low-field shifted (with exception of Ca of 7) (for compar-
ison, free TMSMB: Ca = 93.3, Cb = 106.8; 5: Ca = 99.9,
Cb = 113.6; 6: Ca = 94.2, Cb = 110.5; 7: Ca = 92.9,
Cb = 112.9 ppm) confirming the g2-coordination of
TMSMB to copper(I) [35]. In contrast, the H2C@CMe
carbon atoms appear at almost the same chemical shift
as in 2 indicating that there is no interaction of the ene
unit with the group-11 metal ion (TMSMB: 123.2,
127.3 ppm; 5: 123.4, 130.0 ppm; 6: 123.2, 127.7 ppm; 7:
124.7, 127.1 ppm). The CH fragment of the b-diketo-
nates gives rise to one very characteristic resonance sig-
nal at 94.9 (5), 89.9 (6) or 89.6 ppm (7). The CO
carbons are found between 178 and 202 ppm as it is ex-
pected for this type of building blocks [27]. It is obvious
that the CO resonance signal for 7 (178.3 ppm) is signif-
icantly high field-shifted, when compared with 5
(190.6 ppm) and 6 (201.7 ppm). This trend is also found
for [(MHY)Cu(acac)] (190.3 ppm) and [(MHY)Cu(hfac)]
F1
C6
C12
F2
Si1
F3
C7
O1
C9
C10
C8
F6
C1
Cu1
C2
C5
C11
O2
C13
C4
C3
F4
F5
Fig. 1. XP PLOT (drawn at 50% probability level) of 7 with the
˚
adopted atom numbering scheme. Selected bond lengths [A] and angles
[ꢀ] are given in Table 2.

Y. Shen et al. / Journal of Organometallic Chemistry 690 (2005) 3878–3885
3881
Table 3
Deviations of defining atoms from the best-molecule-plane of 7a
(178.4 ppm) and is attributed to the electronegative CF3
groups [27].
In the 29Si{1H} NMR spectra of 5–7 a resonance sig-
nal is observed at ꢀ21.9 (5, 6) or ꢀ10.2 ppm (7) for the
Me3Si group. The silicon atoms follow the order of
deshielding (hfac < dbac, acac).
˚
Deviation (A)
Atom
Cu(1)
Si(1)
O(1)
ꢀ0.017(8)
ꢀ0.09(2)
0.008(2)
ꢀ0.058(2)
ꢀ0.044(2)
0.003(2)
0.05(2)
O(2)
C(1)
2.3. X-ray structure analysis
C(2)
C(3)
C(9)
0.019(2)
0.030(2)
ꢀ0.010(2)
0.0323
Single crystals of 7 could be grown from a saturated
n-hexane/diethyl ether solution containing 7 at ꢀ30 ꢀC.
The molecular structure of 7 is shown in Fig. 1. Selected
C(10)
C(11)
r.m.s.d.
˚
bond lengths [A] and angles [ꢀ] are given in Table 2 and
important crystal data and collection and refinement
parameter are summarised in Table 5 (Section 3).
C(2)
C(3)
C(4)
C(5)
r.m.s.d.
a
ꢀ0.0009(7)
0.003(2)
ꢀ0.0010(8)
ꢀ0.0008(7)
0.0016
ꢀ
Complex 7 crystallises in the triclinic space group P1.
7 contains a tri-coordinated copper(I) ion in a planar
environment, caused by the two chelate-bound O(1)
and O(2) hfac oxygen atoms and the g2-coordinated
TMSMB ligand (C(1), C(2)) (Fig. 1). This structural
motif is typical for such species and is also found in
[(g2-MeC„CMe)Cu(hfac)] [2], [(g2-BTMSA)Cu(hfac)]
(BTMSA = Bis(TriMethylSilyl)Acetylene) [24] or [(g2-
MHY)Cu(hfac)] (MHY = 2-Methyl-1-Hexene-3-Yne)
[27].
The molecule scaffold with Cu(1), O(1), O(2), C(1)–
C(3) and C(9)–C(11) is planar (max. deviation 0.09(2)
˚
A, Si(1)) which best can be explained by the conjugated
hfac and Me3SiC„C–CMe@CH2 units (Table 3). The
distortion of the C(2)–C(5) ene-yne plane (max. devia-
˚
tion 0.003(2) A, C(3)) to the latter plane resembles to
The figures in parentheses represent the standard deviation in units
of the last significant digit.
˚
Cu(hfac)] (1.94(1) and 1.95(1) A) [26]. However, they are
slightly shorter than the corresponding separations
typical for the [(g2-alkene)Cu(hfac)] family, e.g.
2.013(5) and 2.277(7) A for [(g -1,5-COD)Cu(hfac)]
(COD = CycloOctaDiene) [36,37], or 2.011(3) and
2.029(3) A for [(g -7-tert-butoxy-2,5-NBD)Cu(hfac)]
(NBD = NorBornaDiene) [38]. This indicates a stronger
interaction between the copper and the sp-hybridised
carbon atoms, rather than that of g2-coordinated alk-
enes. The C(1)–C(2) bond length with 1.237(4) A is
somewhat longer than the related separation in free 2-
2
˚
2
˚
˚
2
˚
butyne (1.211 A) which is typical for g -coordinated
alkynes in transition metal chemistry [39]. The
11.7(2)ꢀ (Table 3).
The copper-alkyne bond distances in 7 are with
˚
H2C@CMe entity is not interacting with Cu(1) (Cu(1)–
1.967(3) (Cu(1)–C(1)) and 1.973(3) A (Cu(2)–C(2)) (Ta-
ble 2) almost as long as those ones found in [(g2-MHY)-
˚
C(3) 3.115, Cu(1)–C(4) 3.453 A) (Fig. 1).
The copper–oxygen distances are 1.973(2) (Cu(1)–
˚
O(1)) and 1.960(2) A (Cu(1)–O(2)) indicating a nearly
Table 2
a
symmetric binding of the b-diketonate to Cu(1). This
structural behaviour is also found in solution (vide
supra).
As expected, the alkyne unit deviates from linearity
(cis-bending) upon2-coordination of TMSMB to cop-
per(I) (C(1)–C(2)–C(3) 158.5(3), C(2)–C(1)–Si(1)
166.5(2)ꢀ) (Table 2 and Fig. 1).
˚
Selected bond lengths (A) and angles (ꢀ) for 7
Bond lengths
Cu(1)–O(1)
Cu(1)–O(2)
Cu(1)–C(1)
Si(1)–C(1)
1.973(2)
1.960(2)
1.967(3)
1.871(3)
1.973(3)
1.455(4)
3.221(4)
C(3)–C(4)
C(3)–C(5)
C(9)–C(10)
C(10)–C(11)
O(1)–C(9)
O(2)–C(11)
C(1)–C(2)
1.329(4)
1.512(4)
1.403(4)
1.390(4)
1.257(3)
1.244(4)
1.237(4)
Cu(1)–C(2)
C(2)–C(3)
Cu(1)–C(3)/C(4)b
2.4. TG and DSC studies, and chemical vapour deposition
Bond angles
O(1)–Cu(1)–O(2)
O(1)–Cu(1)–C(1)
O(1)–Cu(1)–C(2)
C(1)–Cu(1)–O(2)
C(2)–Cu(1)–O(2)
Cu(1)–C(1)–C(2)
Cu(1)–C(2)–C(1)
Cu(1)–O(1)–C(9)
a
94.56(10)
112.25(11)
148.81(9)
153.04(10)
116.62(11)
71.96(16)
71.45(18)
122.61(17)
Cu(1)–O(2)–C(11)
C(1)–Cu(1)–C(2)
O(1)–C(9)–C(10)
O(2)–C(11)–C(10)
C(2)–C(3)–C(4)
C(1)–C(2)–C(3)
C(2)–C(1)–Si(1)
C(9)–C(10)–C(11)
122.45(18)
36.59(11)
128.2(3)
129.6(2)
121.8(3)
158.5(3)
166.5(2)
122.5(3)
ThermoGravimetric studies (TG) were carried out to
optimise the temperature at which 5–7 should best be
maintained during the CVD experiment [40]. Complex
7 starts to decompose at 85 ꢀC. The percentage of the res-
idue at 400 ꢀC is half as much as the theoretical percent-
age calculated for the formation of copper. Typical for 7
is a two-step decomposition in the temperature range of
85–250 ꢀC. Between 85 and 170 ꢀC the g2-coordinated
ene-yne ligand TMSMB is eliminated, which is evidenced
The figures in parentheses represent the standard deviation in units
of the last significant digits.
b
Midpoint of the C(3)–C(4) carbon–carbon double bond.

3882
Y. Shen et al. / Journal of Organometallic Chemistry 690 (2005) 3878–3885
by the weight loss of 33.8%. In the second step (190–
250 ꢀC) the disproportionation of the in-situ generated
[Cu(hfac)] species takes place to produce [Cu(hfac)2]
along with elemental copper. For 5 and 6 a similar behav-
iour is found which differs from the process displayed by
their Cu(II) analogues (Section 3) [41].
larger grains are the product of a homogeneous reaction
in the gas phase. The deposition mechanism is changed
from a heterogeneous reaction on the surface to a homo-
geneous reaction in the gas phase. There are several
reasons for this change, which are not longer compre-
hensible, e.g. too low carrier gas flow or temperature
variations.
Differential Scanning Calorimetry (DSC) shows the
change of enthalpy during thermolysis. Complexes 5–7
possess two exothermic steps (DH = ꢀ46.49/ꢀ115.77
(5); ꢀ26.07/ꢀ32.48 (6); ꢀ150.75/ꢀ186.44 J/g (7)). Com-
plex 7 possesses an exothermic peak between 75 and
175 ꢀC with the maximum at 137 ꢀC, which corresponds
to the dissociation of TMSMB from [(g2-TMSMB)-
Cu(hfac)]. The measured enthalpy for this process is
ꢀ150.75 J/g. The disproportionation of in-situ formed
[Cu(hfac)] to copper and [Cu(hfac)2] takes place between
195 and 282 ꢀC (peak maximum at 239 ꢀC) with DH =
ꢀ186.44 J/g. This observation corresponds to other
alkyne-stabilised copper(I) b-diketonates [12,21,27].
Based on the above mentioned data, 7 was selected to
act as potential CVD precursor. In a typical deposition
experiment, copper was deposited onto a TiN-coated
oxidised silicon wafer at 145 ꢀC by using a quartz tube
horizontal hot-wall CVD reactor. The precursor temper-
ature was maintained at 50 ꢀC with a nitrogen flow of
14 sccm. The total pressure during the CVD experiment
was maintained at 4.6 · 10ꢀ2 mbar with nitrogen as car-
rier gas. No reducing agent such as H2 was used for the
deposition experiment. The surface morphology and
composition of the obtained copper films were examined
with SEM and EDX.
EDX studies of the deposited copper films were
carried out (Fig. 3, Table 4).
Apart from copper, also Si, Ti and O as substrate
components could be determined. The reason for finding
the latter elements is the small film thickness of about
200 nm. The larger grains were also characterised by
using EDX. At those it was found that the percentage
of copper is 98.78% (Si: 1.22%).
The deposition experiments with 7 show that this spe-
cies is a very promising precursor for further detailed
CVD studies.
3. Experimental
3.1. General procedure
All reactions were carried out under an atmosphere
of purified nitrogen (O2 traces: CuO catalyst, BASF
˚
AG, Ludwigshafen; H2O: molecular sieve 4 A) using
standard Schlenk techniques. Diethyl ether was purified
by distillation from sodium/benzophenone ketyl; n-pen-
tane was purified by distillation from calcium hydride.
Infrared spectra were recorded with a Perkin Elmer
FT-IR 1000 spectrometer. 1H, 13C{1H} and 29Si{1H}
NMR spectra were recorded with a Bruker Avance
The film deposited with 7 is copper coloured. In pre-
cursor flow direction, the colour appears more lightly re-
lated to smaller film thickness. In Fig. 2 a SEM image of
this film is depicted.
The film is not completely closed and not homoge-
neous. Two kinds of copper grains are visible: small
grains of about 180 nm size, the nucleation layer, which
are evenly formed and are in the state of coalescence,
and secondly, larger, irregular shaped grains of ca.
720 nm which are situated on the nucleation layer. The
1
250 spectrometer. H NMR spectra were recorded at
250.130 MHz, 13C{1H} NMR spectra at 62.895 MHz
and 29Si{1H} NMR spectra at 49.662 MHz. Chemical
shifts are reported in d units (parts per million) down-
field from tetramethylsilane with the solvent as the refer-
ence signal (1H: C6D6, d = 7.24; 13C{1H}: C6D6,
d = 128.5; 29Si{1H}: Me4Si, d = 0.0). Melting points
were determined using sealed nitrogen purged capillaries
Fig. 2. SEM images showing the morphology of the copper film obtained from the chemical vapour deposition at 145 ꢀC of 7, deposited on a TiN-
coated oxidised silicon wafer.

Y. Shen et al. / Journal of Organometallic Chemistry 690 (2005) 3878–3885
3883
Fig. 3. EDX spectrum of the copper film obtained from 7 [Ts = 145 ꢀC, Ptotal = 4.6 · 10ꢀ2 mbar; carry gas N2; flow rate = 14 sccm].
on a Gallenkamp MFB 595 010 M melting point appa-
ratus. Thermogravimetrical studies were carried out
with the Perkin Elmer System Pyris TGA 6 and DSC
studies with the Perkin Elmer System Pyris DSC 6. A
constant heating rate of 8 K minꢀ1 under nitrogen gas
(1.0 cm3/min) was used. SEM images were recorded with
a SEM DSM 982 instrument, Gemini from LEO com-
pany. EDX analysis: EDX system EDWIN (Ro¨ntec
company).
Microanalyses were performed by the Institute of Or-
ganic Chemistry, Chemnitz, Technical University and
the Institute of Organic Chemistry, University of Hei-
delberg, Heidelberg.
For CVD experiments a quartz tube horizontal hot-
wall reactor with 40 mm in diameter was used. Heating
is achieved by a resistively heated tube oven (HORST
company). The temperature is set by a temperature con-
troller HT 520 (HORST company) and was calibrated
with a thermocouple type GTH 1150 digital thermome-
ter. The sample was placed in a glass container which is
in direct contact with the quartz tube. Heating was per-
formed with a heating-bend. Carrier gas flow was regu-
lated using a MKS type 247 mass flow controller. The
pressure control systems consist of an Edwards Active
Gauge Controller, a cooling trap, and a turbo and a
backing pump. The trap prevented the reactor effluents
from entering the vacuum pump.
Compounds 2, [29] 3, [15] and 4a–4c [12] were pre-
pared by published procedures. All other chemicals
were purchased by commercial providers (Aldrich,
VWR Chemicals) of highest grade and were used as
received.
3.2. Synthesis of [(g2-TMSMB)Cu(acac)] (5)
To 3 (2.38 g, 5.0 mmol) dissolved in 40 mL of diethyl
ether a suspension of Na(acac) (4a) (1.22 g, 10.0 mmol)
in 40 mL of diethyl ether is added drop-wise during
10 min at ꢀ30 ꢀC. After 4 h of stirring at this tempera-
ture all volatiles were removed in oil-pump vacuum.
To the remaining liquid 100 mL of n-pentane are added.
Filtration through a pad of Celite and evaporation of
the solvent from the filtrate in oil-pump vacuum gives
a pale green liquid in 2.55 g yield (8.65 mmol, 87% based
on 3).
Elemental analysis: Calc. for C13H21CuO2Si (300.94):
C, 51.89; H, 7.03. Found: C, 51.36; H, 6.79%. IR
(NaCl): m[cmꢀ1] 3078 (w), 2959 (s), 2920 (s), 1930 (s)
(mC„C), 1663 (w) (mC@C), 1586 (vs) (CO, asym), 1519
(vs) (CO, sym), 1440 (vs), 1393 (vs), 1260 (s), 1194
(m), 1096 (m), 1016 (s), 978 (m), 922 (s), 863 (vs), 846
(vs), 764 (s), 701 (m), 637 (m), 578 (m), 536 (w), 472
1
(w). H NMR (C6D6): d 5.75 (s, 1H, CH/acac), 5.37 (s,
1H, @CH2/TMSMB), 5.31 (s, 1H, @CH2/TMSMB),
1.95 (s, 3H, CH3/TMSMB), 2.03 (s, 6H, CH3/acac),
0.31 (s, 9H, CH3/SiMe3, TMSMB). 13C{1H} NMR
(C6D6): 190.6 (CO/acac), 130.0 (H2C@C/TMSMB),
123.4 (H2C@C/TMSMB), 113.6 (C„C/TMSMB), 99.9
(C„C/TMSMB), 94.9 (CH/acac), 28.3 (CH3/acac),
24.0 (CH3/TMSMB), 0.1 (CH3/SiMe3). 29Si{1H} NMR
(C6D6): d ꢀ21.9. TGA: Tbegin = 80 ꢀC, Tend = 170 ꢀC,
Dm = 44.15%; Tbegin = 170 ꢀC, Tend = 350 ꢀC, Dm =
36.37%. DSC: Peak 1 = 171.3 ꢀC (DH1 = ꢀ46.49 J/g);
Peak 2 = 190.1 ꢀC (DH2 = ꢀ115.77 J/g).
Table 4
EDX studies of the copper films obtained from 7 (Ts = 145 ꢀC,
Ptotal = 4.6 · 10ꢀ2 mbar; carrier gas N2; flow rate = 14 sccm)
Element
Smaller particles (%)
Larger particles (%)
Cu
Si
75.51
13.08
8.68
98.78
1.22
–
O
Ti
2.72
–

3884
Y. Shen et al. / Journal of Organometallic Chemistry 690 (2005) 3878–3885
3.3. Synthesis of [(g2-TMSMB)Cu(dbac)] (6)
3.5. X-ray structure analysis of 7
Complex 6 can be synthesised as described for the
preparation of 5 (Section 3.2). In this respect, 2.38 g
(5.0 mmol) of 3 were reacted with 2.06 g (10.0 mmol)
of 4b. After appropriate work-up, complex 6 can be iso-
lated as a pale yellow liquid in 3.24 g yield (8.4 mmol,
84% based on 3).
Single crystals of 7 could be grown from a n-hexane/
diethyl ether solution at ꢀ30 ꢀC. One crystal was picked
directly in the Schlenk flask under nitrogen flow and
glued with resin (Araldit) on top of a glass needle. X-
ray structure measurement was performed with a BRU-
KER SMART 1K CCD diffractometer. Crystal data,
data collection and refinement details are given in
Table 5.
The unit cell was determined with the program SMART
[42]. For data integration and refinement of the unit cell
the program SAINT [42] was used. The space group was
determined using the program ABSEN [43]. All data were
corrected for absorption using SADABS [44]. The structure
was solved using direct methods (SHELX-97 [45]), refined
using least square methods (SHELX-97 [45]). All non-
hydrogen atoms were fully refined in their located posi-
tions, the hydrogen atoms were taken from the electron
density difference map and refined freely.
Elemental analysis: Calc. for C19H33CuO2Si (385.10):
C, 59.26; H, 8.64. Found: C, 58.69; H, 8.26%. IR (NaCl):
m[cmꢀ1] 3099 (w), 2958 (vs), 2903 (s), 2868 (s), 1931 (m)
(mC„C), 1650 (m) (mC@C), 1571 (vs) (CO, asym), 1537
(s) (CO, sym), 1457 (vs), 1432 (vs), 1409 (vs), 1359 (s),
1270 (m), 1248 (s), 1222 (s), 1184 (m), 1126 (m), 978
(m), 867 (s), 844 (s), 794 (m), 761 (m), 736 (m), 636
1
(m), 593 (m), 476 (s). H NMR (C6D6): d 5.72 (s, 1H,
CH/dbac), 5.55 (s, 1H, @CH/TMSMB), 5.30 (s, 1H,
@CH/TMSMB), 1.18 (s, 3H, CH3/TMSMB), 1.05 (s,
t
18H, Bu/dbac), 0.26 (s, 9H, SiMe3). 13C{1H} NMR
(C6D6): d 201.7 (CO/dbac), 127.7 (H2C@C/TMSMB),
123.2 (H2C@C/TMSMB) 110.5 (C„C/TMSMB), 94.2
(C„C/TMSMB), 89.9 (CH/dbac), 41.8 (C(CH3)3/dbac),
29.0 (CH3/tBu, dbac), 23.9 (CH3/TMSMB), 0.3 [SiMe3/
TMSMB], 29Si{1H} NMR (C6D6): d ꢀ21.9. TGA:
Tbegin =90 ꢀC, Tend =240 ꢀC, Dm=37.2%; Tbegin =240 ꢀC,
Tend = 500 ꢀC, Dm = 45.6%. DSC: Peak 1 = 128 ꢀC
(DH1 = ꢀ26.07 J/g); Peak 2 = 219 ꢀC (DH2 = ꢀ32.48 J/g).
Table 5
Crystal data and collection and refinement parameter for 7
Formula
C13H15F6CuO2Si
408.88
Formula weight
Crystal colour
Crystal dimensions (mm)
Crystal system
Space group
Greenish
0.90 · 0.24 · 0.10
Triclinic
3.4. Synthesis of [(g2-TMSMB)Cu(hfac)] (7)
P1
ꢀ
˚
a (A)
7.574(8)
10.825(11)
11.238(12)
76.11(2)
88.50(2)
86.17(2)
892.4(16)
2
˚
b (A)
For the synthesis and work-up of 7 see Section 3.2. In
this respect, 2.38 g (5.0 mmol) of 3 were reacted with
2.3 g (10.0 mmol) of 4c. Complex 7 can be isolated as
a off-white solid. Yield: 3.23 g (7.89 mmol, 79% based
on 3).
˚
c (A)
a (ꢀ)
b (ꢀ)
c (ꢀ)
3
˚
V (A )
Z value
Dcalc (g cmꢀ3
Index ranges
Mp: 32 ꢀC. Elemental analysis: Calc. for
C13H15F6CuO2Si (408.88): C, 38.19; H, 3.70. Found:
C, 37.85; H, 3.39%. IR (NaCl): m[cmꢀ1] 2961 (w), 2926
(w), 2904 (w), 1964 (m) (mC„C), 1642 (s) (mC@C), 1619
(m) (CO, asym), 1552 (m) (CO, sym), 1523 (m), 1477
(s), 1254 (s), 1200 (s), 1148 (vs), 1104 (m), 974 (w), 921
(m), 850 (s), 797 (m), 763 (m), 671 (m), 635 (w), 589
)
1.522
ꢀ10 6 h 6 10,
ꢀ14 6 k 6 15,
ꢀ7 6 l 6 16
412
F (0 0 0)
l (Mo Ka) (cmꢀ1
Radiation, Mo Ka
Temperature (K)
)
1.350
˚
k = 0.71073 A
173(2)
61.34
7236
1
(w), 530 (w). H NMR (C6D6): d 6.13 (s, 1H, CH/hfac),
2hmax
5.81 (s, 1H, @CH2/TMSMB), 5.22 (s, 1H, @CH2/
TMSMB), 1.87 (s, 3H, CH3/TMSMB), 0.32 (s, 9H,
SiMe3). 13C{1H} NMR (C6D6): d 178.3 (CO/hfac,
JC–F = 38.7 Hz), 127.1 (H2C@C/TMSMB), 124.7
(H2C@C/TMSMB), 112.9 (C„C/TMSMB), 92.9
(C„C/TMSMB), 89.6 (CH/hfac), 118.5 (CF3/hfac,
JC–F = 281.3 Hz), 23.6 (CH3/TMSMB), ꢀ0.5 (CH3/
SiMe3, TMSMB). 29Si{1H} NMR (C6D6): ꢀ10.2. TGA:
No. of reflections measured
Independent reflections observed (I) P 2.00(r(I))
No. of refined parameters
R1 (I P 2r(I))/all dataa
WR2 (I P 2r(I))/all datab
Goodness-of-fit (F2)c
4880
212
0.0397/0.0736
0.0991/0.1146
0.999
ꢀ3
˚
Dqmax (e Aꢀ3
)
0.450
ꢀ0.313
0.009
˚
Dqmin (e A
)
D/rmax
n = number of reflections, p = parameters used.
Tbegin = 85 ꢀC, Tend = 170 ꢀC, Dm = 33.8%; Tbegin
=
=
h
i
1=2
P
P
i
P
P
2
a
b
c
R1 ¼ ðjF o ꢀjF ckÞ= jF oj; wR2 ¼
ðwðF 2o ꢀF 2cÞ Þ= ðwF o4Þ
.
190 ꢀC,
Tend = 250 ꢀC,
Dm = 32.13%;
Tbegin
2
275 ꢀC, Tend
=
350 ꢀC, Dm = 14.54%. DSC: Peak
W ¼ 1=½r2ðF 2oÞ þ ð0.0612PÞ ꢁ; P ¼ ðF o2 þ 2F 2c Þ=3.
h
1 = 137 ꢀC (DH1 = ꢀ150.75 J/g); Peak 2 = 239.0 ꢀC
P
2
1=2
S ¼
wðF 2o ꢀ F 2cÞ =ðn ꢀ pÞ
.
(DH2 = ꢀ 186.44 J/g).

Y. Shen et al. / Journal of Organometallic Chemistry 690 (2005) 3878–3885
3885
[18] S.W. Rhee, S.W. Kang, S.H. Han, Electrochem. Solid-state Lett.
3 (2000) 135.
4. Supplementary material
[19] P. Doppelt, T.H. Baum, L. Ricard, Inorg. Chem. 35 (1996)
1286.
Crystallographic data for the structural analysis have
been deposited with the Cambridge Crystallographic
Data Centre, CCDC-No. 275398 for 7. Copies of this
information may be obtained free of charge from The
Director, CCDC, 12 Union Road, Cambridge, CB2
1EZ UK (fax: +44-1223-336-033; e-mail: deposit@ccdc.
cam.ac.uk or http://www.ccdc.cam.ac.uk).
[20] T.H. Baum, C.E. Larson, J. Electrochem. Soc. 140 (1993)
154.
[21] P. Doppelt, T.H. Baum, J. Organomet. Chem. 517 (1996) 53.
[22] G.A. Pehrsen, J.E. Parmehr, C.A. Apblett, M.F. Gonzales, P.M.
Smith, T.R. Omslead, T.R. Norman, J. Electrochem. Soc. 142
(1995) 939.
[23] J.A.T. Norman, B.A. Muratore, P.N. Dyer, D.A. Roberts, A.K.
Hochberg, J. Phys. IV (Paris) (1991) C2–C271.
[24] P. Doppelt, T.H. Baum, MRS Bull. XIX (8) (1994) 41.
[25] P. Doppelt, Coord. Chem. Rev. 180 (1998) 1783.
[26] T.Y. Chen, J. Vaissermann, P. Doppelt, Inorg. Chem. 40 (2001)
6167.
Acknowledgement
[27] T.Y. Chen, J. Vaissermann, R. Ruiz, J.P. Senateur, P. Doppelt,
Chem. Mater. 13 (2001) 3393.
Financial support from the Deutsche Forschungs-
gemeinschaft is gratefully acknowledged.
[28] R.U. Kirss, D. Gordon, P.S. Kirlin, Mater. Res. Soc. Symp. Proc.
282 (1993) 275.
[29] C.S. Rouvier, S. Ldfrere, M. Santelli, Tetrahedron Lett. 40 (1999)
5491.
References
[30] Y.J. Chen, H.D. Kaesz, H. Thridandam, R.F. Hicks, Appl. Phys.
Lett. 53 (1988) 1591.
[1] H.K. Shin, K.M. Chi, M.J. Hampden-Smith, T.T. Kodas, M.F.
Paffett, J.D. Farr, Chem. Mater. 4 (1992) 788.
[31] Y.Z. Shen, M. Leschke, S.E. Schulz, T. Gessner, H. Lang,
Chinese J. Inorg. Chem. 20 (2004) 1257.
[2] T.H. Baum, C.E. Larson, Chem. Mater. 4 (1992) 365.
[3] D.B. Beach, F.K. LeGoues, C.K. Hu, Chem. Mater. 2 (1990) 216.
[4] J. Li, R. Blewer, J.W. Mayer, MRS Bull. 18 (1993) 18.
[5] T.L. Alford, J. Li, J.W. Mayer, S.Q. Wang, Thin solid Films 262,
VII (1995).
[32] J.K. OÕLoane, M.K. Wilson, J. Chem. Phys. 23 (1995) 1313.
[33] W.A. Anderson, A.J. Carty, G.J. Palenik;, G. Schreiber, Can. J.
Chem. 49 (1971) 761.
[34] T.H. Baum, C.E. Larson, G.J. May, J. Organomet. Chem. 425
(1992) 189.
[6] P.L. Pai, C.H. Ting, IEEE Electron Device Lett. 10 (1989) 423.
[7] T.T. Kodas, M.J. Hampden-Smith, The Chemistry of Metal
CVD, VCH, Weinheim, 1994.
[35] M.E. Gross, V.M. Donnelly, in: V.S. Rana, R.V. Joshi (Eds.),
Adv. Matallization for ULSI Applications, Materials Research
Society, Pittsburgh, 1991, p. 355.
[8] A. Jain, K.M. Chi, H.K. Chi, J. Farkas, T.T. Kodas, M.J.
Hampden-Smith, Semicond. Int. 16 (1993) 128.
[36] K.M. Chi, H.K. Shin, M.J. Hampden-Smith, E.N. Duesler,
Polyhedron 10 (1991) 2293.
[9] J.A.T. Norman, B.A. Muratore, P.N. Dyer, D.A. Roberts, A.K.
Hochberg, June 11–12, 1991 VMIC Conference, IEEE, P 123.
[10] S.K. Reynolds, C.J. Smart, E.F. Baran, T.H. Baum, C.E. Larson,
P.J. Brock, J. Appl. Phys. Lett. 59 (1991) 2332.
[37] R. Kumar, F.R. Fronczek, A.W. Maverick, W.G. Lai, G.L.
Griffin, Chem. Mater. 4 (1992) 577.
[38] K.M. Chi, H.C. Hou, P.T. Huang, Organometallics 14 (1995)
2641.
[11] A. Jain, K.M. Chi, M.J. Hampden-Smith, T.T. Kodas, M.F.
Paffett, J.D. Farr, Chem. Mater. 3 (1991) 995.
[39] E. Pignataro, B. Post, Acta Crystallogr. 8 (1955) 672.
[40] K.J. Eisentraut, R.E. Sievers, J. Inorg. Nucl. Chem. 29 (1976)
1931.
[12] K.M. Chi, T.S. Corbitt, M.J. Hampden-Smith, T.T. Kodas, E.N.
Deusler, J. Organomet. Chem. 449 (1993) 181.
[41] N.S. Borgharkar, G.L. Griffin, J. Electrochem. Soc. 145 (1) (1998)
347.
[13] H.K. Shin, M.J. Hampden-Smith, T.T. Kodas, E.N. Duesler,
Polyhedron 10 (1991) 645.
[42] Bruker AXS Inc., Madison, WI, USA, 1998.
[43] P. McArdle, J. Appl. Cryst. 29 (1996) 306.
[44] G.M. Sheldrick, SADABS V 2.01, Program for Empirical Absorp-
tion Correction of Area Detector Data, University of Go¨ttingen,
Germany, 2000.
[14] C.H. June, Y.T. Kim, J.T. Baek, H.J. Yoon, D.R. Kim, J. Vac.
Sci. Technol. A 14 (6) (1996) 3214.
[15] M.Y. Park, J.H. Son, S.W. Rhee, Electrochem. Solid-state Lett. 1
(1998) 32.
[16] J.H. Son, M.Y. Park, S.W. Rhee, Thin Solid Films 335 (1998)
229.
[17] Y.J. Senzaki, Electrochem. Soc. 145 (1998) 362.
[45] G.M. Sheldrick, SHELX-97, Programs for Crystal Structure Anal-
ysis (Release 97-2), University of Go¨ttingen, Germany, 1997.