Cobalt(I) Olefin Complexes: Precursors for Metal-Organic Chemical Vapor Deposition of High Purity Cobalt Metal Thin Films

Article abstract of DOI:10.1021/acs.inorgchem.6b01146

We report the synthesis and characterization of a family of organometallic cobalt(I) metal precursors based around cyclopentadienyl and diene ligands. The molecular structures of the complexes cyclopentadienyl-cobalt(I) diolefin complexes are described, as determined by single-crystal X-ray diffraction analysis. Thermogravimetric analysis and thermal stability studies of the complexes highlighted the isoprene, dimethyl butadiene, and cyclohexadiene derivatives [(C₅H₅)Co(η⁴-CH₂CHC(Me)CH₂)] (1), [(C₅H₅)Co(η⁴-CH₂C(Me)C(Me)CH₂)] (2), and [(C₅H₅)Co(η⁴-C₆H₈)] (4) as possible cobalt metal organic chemical vapor deposition (MOCVD) precursors. Atmospheric pressure MOCVD was employed using precursor 1, to synthesize thin films of metallic cobalt on silicon substrates under an atmosphere (760 torr) of hydrogen (H₂). Analysis of the thin films deposited at substrate temperatures of 325, 350, 375, and 400 °C, respectively, by scanning electron microscopy and atomic force microscopy reveal temperature-dependent growth features. Films grown at these temperatures are continuous, pinhole-free, and can be seen to be composed of hexagonal particles clearly visible in the electron micrograph. Powder X-ray diffraction and X-ray photoelectron spectroscopy all show the films to be highly crystalline, high-purity metallic cobalt. Raman spectroscopy was unable to detect the presence of cobalt silicides at the substrate/thin film interface.

Full text of DOI:10.1021/acs.inorgchem.6b01146

Article
pubs.acs.org/IC
Cobalt(I) Olefin Complexes: Precursors for Metal−Organic Chemical
Vapor Deposition of High Purity Cobalt Metal Thin Films
Jeff A. Hamilton, Thomas Pugh, Andrew L. Johnson,* Andrew J. Kingsley, and Stephen P. Richards
†
†
,†
‡
‡
†
Department of Chemistry, University of Bath, Bath BA2 7AY, United Kingdom
‡
SAFC-Hitech, Power Road, Bromborough, Wirral, CH62 3QF, United Kingdom
*
S Supporting Information
ABSTRACT: We report the synthesis and characterization of
a family of organometallic cobalt(I) metal precursors based
around cyclopentadienyl and diene ligands. The molecular
structures of the complexes cyclopentadienyl−cobalt(I)
diolefin complexes are described, as determined by single-
crystal X-ray diffraction analysis. Thermogravimetric analysis
and thermal stability studies of the complexes highlighted the
isoprene, dimethyl butadiene, and cyclohexadiene derivatives
4
4
[
(C H )Co(η -CH CHC(Me)CH )] (1), [(C H )Co(η -
5 5 2 2 5 5
4
CH C(Me)C(Me)CH )] (2), and [(C H )Co(η -C H )] (4) as possible cobalt metal organic chemical vapor deposition
2
2
5
5
6
8
(
MOCVD) precursors. Atmospheric pressure MOCVD was employed using precursor 1, to synthesize thin films of metallic
cobalt on silicon substrates under an atmosphere (760 torr) of hydrogen (H ). Analysis of the thin films deposited at substrate
2
temperatures of 325, 350, 375, and 400 °C, respectively, by scanning electron microscopy and atomic force microscopy reveal
temperature-dependent growth features. Films grown at these temperatures are continuous, pinhole-free, and can be seen to be
composed of hexagonal particles clearly visible in the electron micrograph. Powder X-ray diffraction and X-ray photoelectron
spectroscopy all show the films to be highly crystalline, high-purity metallic cobalt. Raman spectroscopy was unable to detect the
presence of cobalt silicides at the substrate/thin film interface.
INTRODUCTION
these parameters, controlled formation of thin films of cobalt
■
can be achieved using either chemical vapor deposition
Cobalt is used in a wide variety of technological applications:
with a higher sticking coefficient toward copper, cobalt has
been used to replace tantalum as a lining material within
interconnect trenches in microelectronic devices. The cobalt
layer allows for better copper seed layer coverage thereby
reducing copper agglomeration and void formation. Thin films
of cobalt, which are ferromagnetic, may be layered with
nonmagnetic metals to form devices that take advantage of
2
,14
9,15
(
CVD)
As previously noted by the authors, there are limited reports
of cobalt precursors for CVD or ALD of cobalt and cobalt
or atomic layer deposition (ALD).
1
,2
1
6
oxide. Of those reported compounds very few are liquid at
room temperature, an indication that they may have an
appreciable vapor pressure. High volatility is an important
characteristic for both CVD and ALD precursors, as it ensures
that the precursor vapors are easily and consistently transported
in the carrier-gas stream from the precursor vessel to the
deposition chamber of the CVD or ALD system.
3
4
,5
giant magnetoresistance.
These devices can be used in
magnetic data storage media. Cobalt is also an important
catalyst in the Fischer−Tropsch (FT) process. It has the
advantage of being cheaper than ruthenium (the most active
group VIII metal used in FT catalysis) and is more practical
than nickel, which leads to volatile carbonyl formation during
the process and subsequent loss of the metal catalyst. Typically
a form of mesoporous silica or alumina is used as the
support for the cobalt. Several methods of catalyst formation
Metal−organic CVD (MOCVD), which uses organometallic
precursors in the CVD process, offers the prospect of
significantly lower deposition temperatures by virtue of the
reactive, and often fragile, metal−carbon bonds. Despite this,
there is a relative paucity of reports concerning organometallic
cobalt precursors and especially low oxidation state species. To
date thin films of cobalt metal prepared by CVD have been
6
7
,8
9
10
exist and have been reviewed.
Previously reported preparation techniques for cobalt thin
1
1,17
reported using the precursors [Co (CO) ] (I),
[Co-
2
8
17b,18
17c
5
11
12
(CO) NO] (II),
[Co(CO) (CF )] (III), [(η -C H )-
films include sputtering, e-beam evaporation, and thermal
3
3
3
5
5
1
7a,c,19
20
n
1
3
Co(CO) ] (IV),
CoH(CO)3 (V), CoH[ PrOP-
evaporation. Sputtering and evaporation techniques are
generally considered to be “line-of-sight” techniques and are
therefore best suited to two-dimensional substrates. However,
where conformal coating of three-dimensional structures is
desired, such as in microelectronics applications and where the
trend of reduction in device sizes demands high control over
2
2
1
17c
22
2
(
OMe)2]3 (VI), [(C H ) Co] (VII),
[Co (CO) (η -
5
5
2
2 6
t
RCCR′)] (VIII) (R = H, R′ = Bu) (R = H, R′ =
Received: May 10, 2016
©
XXXX American Chemical Society
A
DOI: 10.1021/acs.inorgchem.6b01146
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Chart 1. Organometallic Cobalt Chemical Vapor Deposition Precursors
n
n
n
n
23
39
SiMe , C H , C H , C H , C H ), Co(Allyl)(CO)3
pentadienide were all made according to literature procedures.
3
4
9
5
11
6
13
7
15
25
2
4
t
2 t
The diolefins used in this study were purchased through commercial
(
IX),t [Co( BuNC(Me)NEt) ] (X), [Co(κ - BuNCH-
2
40
26 16
5 2 i i
sources. Compounds 3, 4, 5, and 6 have been synthesized previously,
CHN Bu) ] (XI), and [(η -C H )Co(κ - PrNCHCHN Pr)]
2
5
5
1
and H NMR and elemental analysis were used to confirm batch
(
XII; Chart 1) at various temperatures and conditions.
While recent focus has been directed toward lower
temperature growth methods, such as ALD and plasma-assisted
ALD, in an attempt to avoid dewetting and nanoparticle
growth, precursor development has not advanced at a similar
pace. Well-established “off-the-shelf” precursors such as
purity.
General Procedures. Elemental analyses were performed by
Elemental Analysis Service, London Metropolitan University, London,
U.K. H and C NMR spectra were recorded on a Bruker Advance
00 MHz FT−NMR spectrometer; chemical shifts are quoted in units
of parts per million, relative to Me Si ( H, C); coupling constants are
1
13
3
1
13
4
2
7
5
28
5
28,29
[
Co (CO) ], [(η -C H )Co(CO) ], [(η -C H ) Co],
and [Co(η -allyl)(CO) ], all originally developed for CVD,
in hertz.
2
8
5
5
2
5
5 2
3
30
All reactions were performed under an inert atmosphere and in the
absence of light, using standard Schlenk line techniques. Low-
temperature reactions were performed using an ultralow temperature
recirculating chiller and modified Schlenk tube with external cooling
jacket. Solvents were dried over activated alumina columns using an
Innovative Technology solvent purification system (SPS) and
degassed under an argon atmosphere. All other reagents were
purchased from commercial sources.
Synthesis of Complexes 1−11. Synthesis of 1. Method A. To a
tetrahydrofuran (THF; 30 mL) solution of tris-
triphenylphosphinecobalt(I) chloride (4.41 g, 5 mmol) at −78 °C, a
2 M solution of sodium cyclopentadienyl in THF (2.5 mL, 5 mmol)
was added. The solution was allowed to warm to room temperature
over a period of 1 h and was stirred for an additional hour. Isoprene
(0.75 mL, 7.5 mmol) was then added, and the reaction mixture was
refluxed for a period of 0.5 h, during which time a color change from
deep red to red-orange was observed. The THF was removed in vacuo,
the remaining material was extracted into hexane (3 × 20 mL), and the
combined extracts were filtered through diatomaceous earth. The
solvent was removed in vacuo, and the oily residue was distilled (5 ×
10 mbar, 50 °C) into a liquid nitrogen cooled Schlenk yielding 0.57
g (59%) of a red-orange oil.
Method B. Cobaltocene (2.2 g, 10 mmol) and potassium (0.4g 10
O and stirred at −60 °C for 6 h. Isoprene
1.5 mL, 15 mmol) was added dropwise to the stirred solution, and
the brown reaction mixture was stirred for 48 h at −60 °C. After this
time the reaction mixture was warmed to room temperature, and the
solvent was removed in vacuo. The residue was dissolved in dry
hexane and filtered through diatomaceous earth. The hexane was
removed in vacuo, and the oily residue was distilled (5 × 10 mbar,
0 °C) into a liquid nitrogen cooled Schlenk yielding 1.90 g (80%) of
3
have been explored as potential precursors for ALD of cobalt
metal. Only recently has there been development of new cobalt
metal precursors, with the publication of the synthesis of stable
and highly volatile complexes of cobalt such as the bis-
2
5,28,31
26
amidinate complex,
bis-diazadienyl, and bis-hydrazo-
3
2
nate complexes alongside their application in the ALD of thin
33
films of cobalt metal and cobalt oxide.
One family of complexes that has received almost no
attention as potential precursors is cobalt(I) diolefin systems.
Despite their application in catalysis cobalt(I) diolefin
complexes are often thought of as highly reactive and unstable
species, and while this may be the case for the parent ethene
system [(η -C H )Co(η -C H ) ] the use of diolefin ligands
such as butadiene, norbornadiene, and 1,5-cycloctadiene
provides the complexes with significantly higher stability.
Only very recently have Lang and co-workers exploited this
5
2
34
5
5
2
4 2
35
5
stability in identifying the Co(I) 1,5-hexadiene system [(η -
4
−2
C H )Co(η -H CCHCH CH CHCH ) ] as a potential
5
5
2
2
2
2 2
36
cobalt MOCVD precursor; however, the related iridium
5
complex [(η -C H Me)Ir(COD)] (COD = 1,5-cycloctadiene)
has been used previously for the MOCVD production of IrO
5
4
mmol) were dissolved in Et
2
2
37
(
thin films with oxygen as a coreagent.
Herein we report the synthesis of a range of cobalt(I)
diolefin precursors, alongside studies of their thermal
decomposition, volatility, and subsequent selection as suitable
candidates for the MOCVD of cobalt metal films. The influence
of both modification of the cyclopentadienyl and the diolefin
ligand was examined, and the effect these modifications have on
the physical properties such as the vapor pressure, the
decomposition temperature, and process as well as the
deposition parameters during the MOCVD was investigated.
−2
5
a red-orange oil.
Analysis for compound made by both methods was identical. Found
C 62.36%, H 6.69%, Calc.: C 62.51%, H 6.82%. H NMR (300.22
1
3
2
MHz, C
D , 298 K) δ: −0.39 (dd, 1H, CH, J = 9.2 Hz, J = 1.5 Hz),
6
2 3 2
6
−
1
0.27 (d, 1H, J = 1.6 Hz, CH), 1.77 (dd, 1H, CH, J = 6.5 Hz, J =
5
.5 Hz), 1.89 (d, 1H, CH), 2.05 (s, 3H, Me), 4.63 (s, 5H, η -Cp), 4.93
(dd, 1H, CH, J = 9.2 Hz, J = 6.5 Hz). C{ H} NMR (75.50 MHz,
C D , 298 K) δ: 23.8 (CH ), 30.7(CH ), 34.4 (CH ), 79.0 (CH), 80.6
EXPERIMENTAL
3
3
13
1
■
General Experimental Details. Tristriphenylphosphinecobalt(I)
chloride, sodium cyclopentadienide, and sodium methyl-cyclo-
6
6
3
2
2
38
(C H ), 94.6 (C-Me).
5 5
B
DOI: 10.1021/acs.inorgchem.6b01146
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Complexes 2, 3, 4, 5, and 6 were prepared using both Method A
and B. Complexes 6, 7, 8, 9, 10, and 11 were synthesized using
methodology A.
(300.22 MHz, C D , 298 K) δ: −0.40 (d, 2H, CH), 1.63 (br s, 5H,
6
6
5
Cp-Me & 2 CH) 1.97 (s, 6H, Me), 4.41 (s, 2H, η -Cp-H), 4.52 (s, 2H,
5
13
1
η -Cp-H). C{ H} NMR (75.50 MHz, C D , 298 K) δ: 13.6 (Cp-
6
6
Synthesis of 2. Complex 2 was prepared using both methods A and
B in an analogous fashion to that of 1 with the exception of the
addition of 2,3-dimethyl butadiene (Method A: 0.85 mL, 7.5 mmol,
Method B 1.7 mL, 15 mmol) in place of isoprene. Sublimation (5 ×
Me), 20.2 (C-Me), 36.3 (CH ), 80.1 (C H -Me) 82.1, (C H -Me)
2 4 4 4 4
91.0 (C-Me), 93.8 (Cipso).
Synthesis of 9. Complex 9 was prepared using Method A with of
the addition of sodium methyl cyclopentadienyl (0.44g, 5mmol) and
−
2
−2
1
0
mbar, 60 °C) of the residue yielded red-orange crystals. Method
0.72 mL (7.5 mmol) of 1,3-cyclohexadiene. Distillation (5 × 10
A: 0.66 g (64%), Method B: 1.42 g (69%). Analysis. Found C 63.97%,
H 7.36%, Calcd: C 64.08%, H 7.33%. H NMR (300.22 MHz, C D ,
98 K) δ: −0.57 (d, 2H, CH, J = 0.9 Hz), 1.66 (d, 2H, CH, J = 0.9
Hz) 1.88 (s, 6H, Me), 4.43 (s, 5H, η -Cp). C{ H} NMR (75.50
mbar, 60 °C) yielded 0.71 g (69%) of a red-orange oil. Analysis. Found
C 66.17%, H 7.12%, Calcd: C 66.06%, H 6.93%. H NMR (300.22
1
1
6
6
2
2
2
MHz, C
6 6 2 2
D , 298 K) δ: 0.93 (m, 2H, CH ), 1.56 (m, 2H, CH ), 1.81
5
13
1
5
(s, 3H, Cp-Me), 2.87 (m, 2H, CH), 4.40 (t, 2H, η -Cp-H), 4.57 (t, 2H,
5
13
1
MHz, C D , 298 K) δ: 20.7 (CH ), 35.2 (CH ), 81.0 (C H ), 91.8
η -Cp-H), 4.79 (m, 2H, CH). C{ H} NMR (75.50 MHz, C
K) δ: 14.4 (Cp-Me), 26.3 (CH ), 54.9 (CH), 78.0 (CH), 80.1 (C
Me), 81.1 (C -Me), 94.9 (Cipso).
6
D , 298
6
6
6
3
2
5
5
(C-Me).
2
H -
4 4
Synthesis of 3. Complex 3 was prepared using methods A and B in
the presence of cyclopentadiene (0.63 mL, 7.5 mmol (Method A) or
.26 mL, 15 mmol (method B)). Sublimation of 6 (5 × 10−2 mbar, 50
C) afforded a dark red solid. Method A: 0.68 g (72%), Method B:
.61 g (85%). Analysis. Found C 63.12%, H 5.80%, Calcd: C 63.17%,
H
4 4
Synthesis of 10. Complex 10 was prepared using Method A with of
the addition of sodium methyl cyclopentadienyl (0.44 g, 5 mmol) and
0.76 mL (7.5 mmol) of norbornadiene. Distillation (5 × 10− mbar, 80
°C) yielded 0.58 g (50% yield) of brown-yellow oil. Analysis. Found C
1
°
2
1
1
1
H 5.83%. H NMR (300.22 MHz, C D , 298 K) δ: 2.02 (m, 1H, CH ),
67.97%, H 6.69%, Calcd: C 67.83%, H 6.57%. H NMR (250.13 MHz,
6
6
2
5
5
2
2
4
.40 (m, 2H, CH), 2.56 (m, 1H, CH ), 4.56 (s, 5H, η -Cp), 5.23 (m,
C
6
D
6
, 298 K) δ: 0.89 (m, 2H, CH
2
), 1.35 (s, 3H, η -Cp-Me), 2.63 (m,
2
H, CH). 13C{ H}c NMR (75.50 MHz, C D , 298 K) δ: 39.4 (CH ),
1
5
5
4H, CH), 3.17 (m, 2H, CH), 4.22 (t, 2H, η -Cp-H), 4.83 (t, 2H, η -
6
6
2
Cp-H). 13C{ H} NMR (75.50 MHz, C
26.3 (CH ), 55.6 (CH), 81.0 (CH), 82.2 (C
Me), 93.9 (Cipso).
1
D
, 298 K) δ: 12.8 (Cp-Me),
-Me), 84.1 (C H -
4 4
1.7 (CH), 77.3 (CH), 79.8 (C H ).
6 6
5
5
Synthesis of 4. Complex 4 was synthesized using both Method A,
2
4
H
4
in the presence of 0.72 mL (7.5 mmol) of 1,3-cyclohexadiene, and
Method B in the presence of 1.43 mL (15 mmol) of 1,3-
Synthesis of 11. Complex 11 was prepared using Method A with of
the addition of sodium methyl cyclopentadienyl (0.44 g, 5 mmol) and
0.92 mL (7.5 mmol) of 1,5-cyclooctadiene. Complex 11 was isolated
−2
cyclohexadiene. Sublimation (5 × 10 mbar, 60 °C) yielded a red-
orange solid. Method A: 0.66 g (65%), Method B: 1.45 g (71%).
1
−2
Analysis. Found C 64.76%, H 6.48%, Calcd: C 64.71%, H 6.42%. H
by distillation (5 × 10 mbar, 80 °C) as 0.71 g (58%) of a yellow-
NMR (300.22 MHz, C D , 298 K) δ: 0.77 (d, 2H, CH ), 1.43 (m, 2H,
brown oil. Analysis. Found C 68.41%, H 7.84%, Calcd: C 68.29%, H
6
6
2
5
1
5
CH ), 2.91 (m, 2H, CH), 4.50 (s, 5H, η -Cp), 4.71 (m, 2H, CH).
7.78%. H NMR (250.13 MHz, C
6
D
6
, 298 K) δ: 1.40 (s, 3H, η -Cp-
), 3.29 (m, 4H, CH ), 4.15
(t, 2H, η -Cp-H), 4.47 (t, 2H, η -Cp-H). C{ H} NMR (75.50 MHz,
, 298 K) δ: 12.9 (Cp-Me), 32.9 (CH ), 83.6 (CH), 84.0 (C
Me), 85.6 (C -Me), 96.0 (Cipso).
2
Identification was confirmed by a comparison of the spectroscopic data
with literature.
Me), 1.74 (dd, 4H, CH), 2.48 (m, 4H, CH
2
2
5
5
13
1
Synthesis of 5. Complex 5 was synthesized using both Method A,
in the presence of 0.76 mL (7.5mmol) of norbornadiene, and Method
B in the presence of 1.53 mL (15 mmol) of norbornadiene.
Recrystallization of 5 from a saturated hexane solution at −28 °C
afforded dark brown-orange crystals. Method A: 0.76 g (69%), Method
C
6
D
6
2
H -
4 4
H
4
4
CRYSTALLOGRAPHY
■
−2
Experimental details relating to the single-crystal X-ray
crystallographic studies are summarized in Table 1S (see the
Supporting Information). For both structures, data were
collected on a Nonius Kappa CCD diffractometer at 150(2)
K using Mo Kα radiation (λ = 0.710 73 Å). Structure solution
B: 1.56 g (72%). 5 could be further purified by sublimation (5 × 10
mbar, 80 °C). Analysis. Found C 66.74%, H 6.09%, Calcd: C 66.68%,
1
H 6.06%. H NMR (250.13 MHz, C D , 298 K) δ: 0.78 (m, 2H, CH ),
2
6
6
2
5
.70 (m, 4H, CH), 3.03 (m, 2H, CH), 4.42 (s, 5H, η -Cp).
Identification was confirmed by a comparison of the spectroscopic data
with literature.
4
1
and refinements were performed using SHELX86 and
42
Synthesis of 6. Complex 6 was synthesized using both Method A,
in the presence of 0.92 mL (7.5mmol) of 1,5-cyclooctadiene, and
Method B in the presence of 1.84 mL (15 mmol) of 1,5-
cyclooctadiene. Recrystallization of 6 from a saturated hexane solution
at −28 °C afforded yellow-orange crystals. Method A: 0.84 g (72%),
Method B: 1.63 g (70%). Analysis. Found C 67.09%, H 7.47%, Calcd:
SHELX97 software, respectively. Corrections for absorption
were made in all cases. Data were processed using the Nonius
43
Software. Structure solution, followed by full-matrix least-
squares refinement, was performed using the WINGX-1.80
44
suite of programs throughout. For all complexes, hydrogen
atoms were included at calculated positions. Data for complexes
2, 5, and 6 are deposited with the Cambridge Structural
Database with CCDC reference numbers 1477351−1477353.
1
C 67.24%, H 7.38%. H NMR (250.13 MHz, C D , 298 K) δ: 1.79 (m,
6
6
5
4
H, CH), 2.56 (m, 4H, CH ), 3.55 (m, 4H, CH ), 4.46 (s, 5H, η -Cp).
2
2
Identification was confirmed by a comparison of the spectroscopic data
with literature.
Synthesis of 7. Complex 7 was synthesized using Method A using
MATERIALS CHEMISTRY
■
0
.44 g of sodium methyl cyclopentadienyl (5mmol) and 0.75 mL of
−2
Thermogravimetric analysis (TGA) of the complexes was
performed at SAFC Hitech, Bromborough, U.K., using a
Shimadzu TGA-51 Thermogravimetric Analyzer. Data points
isoprene (7.5 mmol). Distillation (5 × 10 mbar, 50 °C) yielded 0.64
g of a red-orange liquid (62%). Analysis. Found C 63.84%, H 7.44%,
1
Calcd: C 64.08%, H 7.33%. H NMR (300.22 MHz, C D , 298 K) δ:
6
6
5
−1
−
0.51 (dd, 1H, CH), −0.34 (d, 1H, CH), 1.62 (s, 3H, η -Cp-Me),
.72 (dd, 1H, CH) 1.85 (d, 1H, CH), 2.01 (s, 3H, Me), 4.46 (t, 2H,
η -Cp-H), 4.52 (t, 2H, η -Cp-H), 4.87 (dd, 1H, CH). C{ H} NMR
were collected every second at a ramp rate of 5 °C min in a
−1
1
flowing (50 mL min ) N stream. Isothermal TGAs were run
2
5
5
13
1
−1
at 73 °C in a flowing (10 mL min ) N stream, with run times
2
(
(
9
75.50 MHz, C D , 298 K) δ: 13.9 (Cp-Me), 23.3 (C-CH ), 32.3
6
6
3
of 12, 28, and 107 min for complexes 1, 2, and 4, respectively.
Thin films were deposited using a modified cold-walled CVD
system (Electrogas Systems Ltd, U.K). The system consisted of
a tubular quartz reactor containing a silicon carbide coated
graphite susceptor. The temperature of the susceptor was
monitored using a k-type thermocouple coupled with a
proportional−integral−derivative controller and heated with a
CH ), 35.4 (CH ), 79.0 (CH), 79.9 (C H -Me), 81.7 (C H -Me),
2
2
4
4
4
4
3.9 (Cipso), 94.2 (C-Me).
Synthesis of 8. Complex 8 was prepared using Method A with of
the addition of sodium methyl cyclopentadienyl (0.44 g, 5 mmol) and
−
2
0
.85 mL (7.5 mmol) of 2,3-dimethylbutadiene. Distillation (5 × 10
mbar, 50 °C) yielded 0.79 g of a red-orange liquid (72%). Analysis.
Found C 65.59%, H 7.86%, Calcd: C 65.45%, H 7.78%. H NMR
1
C
DOI: 10.1021/acs.inorgchem.6b01146
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
water-cooled IR lamp mounted externally beneath the reaction
tube. The pressure of the system was maintained at 760 torr
throughout the deposition run, with the reactor lines and
precursor, held in a bubbler, at 90 and 85 °C, respectively. The
vapor generated was delivered to the reaction zone using a
high-purity hydrogen (99.998%) carrier gas (300 mL min )
metered from a mass flow controller. Films were deposited
onto CVD-coated ruthenium on silicon wafers at a susceptor
temperature between 325 and 400 °C. The total deposition
time per sample was 30 min.
Raman spectroscopy was recorded on a Renishaw InVia
Raman using a 532 nm (green) laser. X-ray photoelectron
spectrsocopy (XPS) measurements were performed on a Kratos
Axis Ultra-DLD photoelectron spectrometer, utilizing mono-
chromatic Al Kα radiation (photon energy 1486.6 eV), at the
University of Cardiff. The instrument was precalibrated using
pure gold and copper samples. Samples were sputtered for a
predetermined set time over a 4 mm wide area using 4 kV
argon ions using a minibeam I ion source. Spectra were
collected at pass energies of 80 and 160 eV for high resolution
and survey scans, respectively, with the 100 μm apperature in
place to focus on the center of the etch pit. Field emission
scanning electron microscopy (FE-SEM) analysis of the films
was undertaken on a JEOL JSM-6480LV scanning electron
microscope with EDX capability. Atomic force microscopy
with triphenylphosphine, which could be successfully removed
by either careful sublimation or crystallization of the product.
For this reason, an alternative large-scale preparative method of
production was sought: Using a modification of the procedure
34
used to produce the Jonas reagent [CpCo(H CCH ) ], we
2
2 2
−1
were able to synthesize complexes 1−6 in high yield without
the need to utilize the gaseous reagent ethene or
3
5
trimethylsilylethene (Method B: Scheme 1). However,
because difficulties encountered in the large-scale production
of bis-methyl-cyclopentadienyl cobalt (dimethyl-cobaltocene)
complexes 7−11 were produced, exclusively, using Method A
and sodium methyl-cyclopentadienide (NaCpMe).
1
Room-temperature H NMR of complexes 1−6 showed all
of the complexes to be diamagnetic, indicative of a cobalt(I)
environment, and to possess characteristic singlet resonances
5
1
for η -Cp-H between δ = 4.4 and 4.7 ppm; the H NMR
spectra of complexes 1 and 2 show the expected methyl [1:
2
1
.05 ppm; 2: 1.88 ppm] and olefinic protons [1: −0.27, −0.39,
1
.77, 1.89, 4.93 ppm; 2: −0.5, 1.66 ppm]. The H NMR spectra
for complexes 3−6 show the presence of the expected olefinic
protons [3: 2.40, 5.23 ppm; 4: 2.91, 4.50 ppm; 5: 2.70 ppm; 6:
1
.79 ppm ] as well as the methene (CH ) protons [3: 2.02, 2.56
2
ppm; 4: 0.77, 1.43 ppm; 5: 0.78 ppm; 6: 2.56, 3.55 ppm].
1
The H NMR spectra for the complexes 7−11 are similar to
the related complexes 1−2 and 4−6 with respect to the
(
AFM) analysis was performed using a Digital Instruments
Nanoscope IIIa, with TAP300 tips in contact mode (tip radius
10nm). Powder XRD of the films was performed on a Bruker
1
resonances diolefin ligand. Each H NMR spectrum also
contains a multiplet resonance between δ = 4.15 and 4.83 ppm
for the C H -Me hydrogens, which correlate to resonances
<
5
4
D8 Powder Diffractometer, using a Cu anode X-ray source, (Kα
wavelength = 1.5406 Å) at the University of Bath. Sheet
resistance measurements were recorded using a Jandel
Multiheight four-point probe in combination with a Guardian
Surface Resistance Meter model SRM-232−100, with a sheet
resistance range of 0−100 ohm/sq.
between δ = 79.9 and 85.6 (for the {C H CMe} carbon atoms)
4
4
and 93.8 and 96.0 ppm (Cipso), respectively, for the three
5
distinct carbon atoms in the {η -C H -Me} rings in the
5
4
associated 13C NMR spectra.
Molecular Structures of 2, 5, and 6. Of the eleven
complexes reported in this study, single crystals suitable for
single-crystal X-ray diffraction studies of only three complexes,
RESULTS AND DISCUSSION
■
2
, 5, and 6, could be isolated. In the case of 2 slow sublimation
Synthesis. The cobalt diolefin complexes 1−11 can be
easily prepared by a number of synthetic routes;
onto a water-cooled coldfinger (1 °C) yielded crystals suitable
for X-ray studies. In the case of 5 and 6 crystals were grown
from saturated hexane solutions at −28 °C. All three complexes
were characterized by single-crystal X-ray diffraction to
unambiguously determine the solid-state molecular structures
of the products and are relatively rare examples of CpCo(I)
complexes with diolefin ligands. Experimental crystallographic
data are summarized in the Supporting Information, the
molecular structures of the complexes are shown in Figure1,
and selected bond lengths and angles for 2, 5, and 6 are shown
in Table 1.
34,35,40,45
however, focus was given to two specific methods (Scheme 1),
which could easily scale to provide larger quantities of
precursor from equally scalable starting materials.
While complexes 1−11 were initially prepared by the
reaction of ClCo(PPh ) with NaCp, followed by a thermally
3
3
driven substitution of PPh with the appropriate diolefin ligand
3
(Method A: Scheme 1), products were invariably contaminated
Scheme 1. Synthetic Methods A and B
The solid-state complexes 2, 5, and 6 all adopt mononuclear
structures, in which the metal centers possess a pseudo-
5
trigonal-planar environment, which comprises of the η -C H
5
5
ligand and the two alkene moieties. Crystallizing in the
monoclinic space group P2 /c with one molecule in the
1
asymmetric unit, the overall structure of 2 is as expected with
4
the 2,3-dimethylbutadiene molecule bound in an η -fashion via
the double bonds to a {CpCo} group, in an analogous fashion
4
to the related rhodium complex [CpRh(η -CH C(Me)C(Me)-
2
46
CH )]. The molecule itself possesses a pseudo mirror plane
2
bisecting both the cyclopentadienyl and dimethyl-butadiene
ligands (through atom C(1) and the midpoints of the C(2)−
C(3) and C(12)−C(13) bond). The C(1) end of the {Cp} ring
is tilted away from the plane of the butadiene ligand [C(11)−
C(12)−C(13)−C(14)] such that the two planes intersect at a
D
DOI: 10.1021/acs.inorgchem.6b01146
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 1. Molecular structures of 2 (left), 5 (middle), and 6 (right) in the solid state. Thermal ellipsoids are shown at 50% probability.
Table 1. Selected Bond Lengths and Angles for 2, 5, and 6
crystallizes in the monoclinic space group C2/c with two
molecules in the asymmetric unit, with rotational disorder
within the {C H } ligand based on one of the two molecules of
bond distances (Å)
5
5
a
2
5
6
5
(Co(2)). In contrast complex 6 crystallizes in the space group
5
Co−(η -C H )
1.68
1.70
1.71
P2 /c with one molecule in the asymmetric unit. The Cp−Co
5
5
(cent)
1
Co(1)−C(11)
Co(1)−C(12)
Co(1)−C(13)
Co(1)−C(14)
Co(1)−C(15)
Co(1)−C(16)
C(11)−C(12)
C(12)−C(13)
C(13)−C(14)
C(14)−C(15)
C(15)−C(16)
2.024
1.983
1.985
2.022
2.005
1.999
2.021
2.014
interactions in 5 and 6 are comparable to that observed in
complex 2 [2: Co−C(ave) = 2.065 Å, Co−Cp(cent) = 1.68 Å;
5: Co−C(ave) = 2.081 Å, Co−Cp(cent) = 1.70 Å; 6: Co−
C(ave) = 2.093 Å, Co−Cp(cent) = 1.71 Å] and is in the
expected range. As anticipated, coordination of the election-rich
olefin groups to the Co(I) metal center results in some
lengthening of the olefin bond [5: C(11)−C(12) = 1.41Å,
C(14)−C(15) = 1.41; 6: C(11)−C(12) = 1.42, C(15)−C(16)
= 1.42 Å] and a loss of some double-bond character [dNBD(C
C) = 1.34 Å].
2.012
2.014
2.026
2.016
1.415
1.432
1.437
1.429
1.410
1.407
1.414
Thermal Profile of Complexes and use as Deposition
Precursor. TGA of complexes 1−11 was performed to gain
insight into relative volatilities and thermal stabilities. These
analyses were performed with an instrument that was housed in
a nitrogen-filled purge-box to minimize reaction with
atmospheric moisture/air. Figure 2 shows TGA plots for
complexes 1−11, and Table 2 summarizes the TGA data for 1−
11. All compounds were found to undergo mass loss and yield a
stable residue of 3.7−0.3% at a temperature between 165 and
310 °C, respectively. All final mass values are significantly below
the expected values for cobalt for all complexes, indicating that
volatilization occurred with minimal thermal decomposition.
bond angles (deg)
b
Cpcent−Co−{CC}
145.20
142.89
141.83
75.28
134.49
133.86
91.65
1
47.00
b
b
{
CC} −Co−{CC}
66.31
C(11)−C(12)−C(13)
C(12)−C(13)−C(14)
115.29
115.18
0.3
C(11)−C(12)−C(13)−C(14)
C(11)−C(12)−C(14)−C(15)
C(11)−C(12)−C(15)−C(16)
0.1
8.8
a
Data for only one of the two molecules in the asymmetric unit is
reported for 5. Midpoint of the CC bonds.
b
small angle of 9.3°. The cobalt center is 1.68 and 1.57 Å from
the {Cpcent} and {C } planes, respectively.
4
Examination of the 1,3-diene moiety shows a uniform range
of C−C bond lengths [1.43, 1.44, and 1.43 Å for the formal
C(11)C(12), C(12)−C(13), and C(13)C(14)] indicative
of an effective Co →1,3-diene π-electron density back-
donation, where the double-single-double bond alterations
observed in free 2,3-dimethylbutadiene are transformed into
delocalized bonds intermediate between CC and C−C
bonds. As a result charge separation between the cobalt central
atom and the ligands are somewhat minimized, resulting in a
relatively nonpolar bonding situation.
Complexes 5 and 6 are similarly mononuclear structures, in
which the metal centers possess a pseudo-trigonal-planar
environment, with the conjugated diene ligand replaced by
the nonconjugated diene ligands norbornadiene (NBD) and
Figure 2. TGA traces of 1−11 from 20 to 500 °C. Experiments were
−1
−1
1,5-cyclooctadiene (COD). The norbornadiene derivative 5
run under N (50 mL min ) at a ramp rate of 5 °C min .
2
E
DOI: 10.1021/acs.inorgchem.6b01146
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Table 2. Expected % Residue, % of Nonvolatile Residue, and
75 °C in an oil bath between recording their NMR spectra. The
degree of compound decomposition was estimated by
comparison of the integrated areas under the {Cp−H} (or
Onset of Volatilisation/Decomposition Temperature for 1−
1
1
{
MeCp}) NMR peak(s) and normalized to the residual protio
a
precursor expected % for Co % nonvolatile residue (temp) onset temp
solvent in the C D capillary. Interestingly compounds 1, 2, 4,
6
6
1
2
3
4
5
6
7
8
9
1
1
30.7
28.6
31.0
28.9
27.3
25.4
28.6
26.8
27.0
23.6
23.9
1.2 (225 °C)
1.7 (195 °C)
2.8 (165 °C)
3.7 (240 °C)
2.8 (216 °C)
3.8 (270 °C)
1.2 (200 °C)
0.3 (168 °C)
2.4 (212 °C)
0.5 (310 °C)
2.6 (230 °C)
73 °C
73 °C
70 °C
92 °C
93 °C
119 °C
62 °C
64 °C
86 °C
95 °C
77 °C
5
, and 6 all showed relatively high thermal stabilities: in the case
of complexes 1, 2, and 4 only minimal decomposition after 7 d
6%, 2%, and 4%, respectively) was observed, indicating a high
(
thermal stability. For complexes 5 and 6 the presence of
paramagnetic species in the NMR sample prohibited further
monitoring after 3 to 4 d. However, compound 3 showed rather
rapid thermal decomposition with formation of a metallic
mirror on the inside of the NMR tube after only 6 h. Despite
being liquids/oils and as such potentially useful as precursors,
compounds 7−11 all showed significant decomposition and
production of a metallic mirror on the NMR tube after 24−48
h, indicating a significantly reduced thermal stability for the
methyl-derivatized complexes.
0
1
a
The temperature at which 1% mass loss has occurred.
Thermal stability at volatilization temperatures is an important
physical property, as it ensures consistent precursor doses and
therefore consistent film thicknesses from run to run.
We have previously investigated the relationship between
alkyl group functionalization of the cyclopentadienyl ligand and
both volatility and stability in copper cyclopentadienide
isocyanate complexes. If we compare the temperatures at
which the onset of volatilization is observed for the 10
compounds for which both the cyclopentadienyl and methyl
cyclopentadienyl complex was synthesized, the general trend
appears to be that the addition of a methyl substituent onto the
The liquid and semi-solid complexes 1, 2, and 4, all of which
showed good thermal stability and were capable of being
sublimed, were further analyzed by isothermal gravimetric
analysis (IT-TGA). Figure 4 shows an IT-TGA plot at 73 °C of
47
{
Cp} fragment results in a lowering of the temperature at which
mass loss begins [1(73 °C):7(62 °C); 2(73 °C):8(64 °C);
4
clearly be seen in Figure 3, which shows a close-up of the TGA
(92 °C):9(86 °C); 6(119 °C):11(77 °C)]. This effect can
Figure 4. IT-TGA traces of 1, 2, and 4 at 73 °C (5 min); IT-TGA for
compounds 1, 2, and 4 run for a total of 12, 28, and 107 min,
−1
respectively. Experiments were run under N (10 mL min ).
2
compounds 1, 2, and 4. The plot shows all precursors to
undergo mass loss at a constant rate. It also shows the order of
volatility to be 1 > 2 > 4, the former evaporating to 50% of the
starting mass within just 5 min, over twice the rate of the latter
two.
As part of our study, vapor-pressure measurements were
performed on 1, 2, and 4 using a previously reported method
Figure 3. Comparison of the TGA traces of related complexes 1 & 7
and 4 & 9.
47,48
and apparatus.
Details of these analyses are supplied in the
Supporting Information. The results of the vapor-pressure
measurements of 1, 2, and 4 are depicted in Figure 5, and the
numerical results are shown in Table 3.
plots for compounds 1−7 and 4−9, respectively: incorporation
of asymmetry into the {Cp} fragment (i.e., a methyl group)
results in a noticeable decrease in the temperature at which a
mass loss occurs [Δ = 9.6 °C (1−7); Δ = 14.6 °C (4−9)].
Each set of data was obtained at temperatures below the
decomposition points of the precursors. The vapor pressures of
1, 2, and 4 obey the general equation log P = A − B/T, where A
and B are free parameters, with the corresponding enthalpy of
vaporization being deduced from the parameter B (Table 3).
Data show that, by far, the most volatile of the complexes
studied here is 1, with a vapor pressure of 7.83 torr at 85 °C.
The vapor pressures recorded in Table 3 are all higher than
those previously described cobalt-CVD precursors including
T
T
Note that the addition of a methyl substituent onto the {Cp}
fragment can also lead to an increase in temperature at which
the onset of mass loss commences, specifically in the cases of
compounds 5 (93 °C) and 10 (95 °C).
The thermal stability of complexes 1−11 was studied by
placing the neat compound inside an NMR tube containing a
1
49
sealed capillary tube of C D . The H NMR spectra were
cobaltocene, [(C H ) Co], the cobalt amidinate, [Co-
6
6
5
5 2
t
25
recorded periodically (every 4 h) with the NMR tubes held at
( BuNC(Et)NEt) ], and the trimethylsilyl-substituted systems
2
F
DOI: 10.1021/acs.inorgchem.6b01146
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Table 4. Experimental Conditions Used for Cobalt
Deposition from 1
variable
value
temperature of substrate
bubbler temperature
path temperature
325, 350, 375, and 400 °C
85 °C
90 °C
bulk flow rate (gas)
flow rate over bubbler (gas)
pressure
0.7 L/min (H2)
0.3L/min (H2)
760 torr
deposition time
30 min
electron micrograph (inset Figure 7c). Films deposited at 325,
50, and 400 °C demonstrated excellent thickness uniformity,
whereas films deposited at 375 °C appeared more uneven
RMS from AFM = 10.751 nm). All deposited films were
3
(
Figure 5. Vapor-pressure data for complexes 1, 2, and 4.
analyzed by XRD to ascertain the crystallographic orientation
of the films (Figure 8). XRD of all films demonstrates crystal
growth to be highly selective, resulting in the detection of only
one single reflection at 2θ = 44.5°, corresponding the (111)
Miller plane associated with face-centered cubic (fcc) cobalt
metal; all other reflections can be associated with the silicon
substrate. There is precedence in the literature for the
preferential deposition of (111) fcc cobalt onto borosilicate
2
23
[
Co (CO) (η -Me SiCCR)] (R = SiMe , H, or n-C H ),
2
6
3
3
3
7
although notably significantly lower than both Co (CO) and
alkyl systems [Co (CO) (η -RCCR)] (R = n-C H to n-
2
9
2
2
6
3
7
2
3b
C H ).
7
15
Resultantly, compound 1 was chosen as a good candidate for
AP-MOCVD experiments owing to its attractive TGA profile,
high vapor pressure, and thermal stability. CVD experiments
were performed in-house on a modified Electrogas hot-walled
reactor utilizing H as a reducing gas. Substrates were silicon
100) wafers, precleaned by washing with aqua regia, water, and
17c,51
glass.
Deposition temperature clearly has an effect on the
crystallinity of the films, and in general, the intensity of the
(111) reflection increases with increasing temperature; the
most intense (111) reflection was associated with thin films
deposited at a substrate temperature of 375 °C, which is in
agreement with SEM data. The least intense (111) reflection
was associated with films deposited at a substrate temperature
of 350 °C, which is likely to be a consequence of the relative
thinness of the film, as indicated by SEM analysis. The average
crystallite sizes were calculated to be 88.4 nm (375 °C) and
2
(
isopropanol and dried under a nitrogen flow. Table 4
summarizes the deposition conditions for complex 1. The
deposition of cobalt metal was investigated at four different
substrate temperatures, 325, 350, 375, and 400 °C, respectively.
In all of the deposition runs the precursor and the carrier gas
lines were externally heated to 130 °C, and the pressure during
deposition was maintained at 760 torr. H gas was used as the
2
−
1
52
carrier gas at a flow rate of 0.3 L min . Table 4 shows the
general physical parameters used throughout the deposition
experiments.
Films deposited at 325 and 350 °C appeared reflective,
metallic, and continuous to the naked eye, whereas films
deposited at 375 and 400 °C, while also appearing reflective,
metallic, and continuous also showed some brown discolor-
ation. At temperatures below 325 °C, deposited films were
powdery gray/blue and appeared discontinuous.
SEM and AFM micrographs of films grown from 1 are shown
in Figure 7. SEM and AFM analysis showed the film
morphology and film thickness to be dependent on temper-
ature, and in general, particle sizes were found to increase with
increasing deposition temperature, which is in agreement with
previous studies reported by Ivanova et al. and general thin-
film growth.
73.4 nm (400 °C) using the Scherrer equation; these values
are in agreement with SEM data. Accurate calculations for films
deposited at 325 and 350 °C were unobtainable due to weak
diffraction.
Film composition was studied by XPS analysis to determine
the elemental composition of the thin films. Analysis of as-
deposited films demonstrated not only the presence of metallic
cobalt but also the presence of oxygen and carbon, likely to
come from adsorbed organic species and surface oxidation.
After the surface was etched for a period of five minutes with an
argon sputter, oxygen and carbon peaks reduced to trace levels,
and the peaks associated with metallic cobalt (2s = 926 eV, 2p
= 779 eV, LMM = 716 eV, 3s = 101 eV, and 3p = 60 eV)
increased in intensity. Table 5 shows the elemental profile of
etched films as a function of deposition temperature;
quantification of the film was conducted using a Shirley
background by integrating the area under the peaks and taking
into account the relative sensitivity factors of each element.
50
Films deposited at 375 °C exhibit a high degree of
crystallinity, with hexagonal particles clearly visible in the
a
Table 3. Vapor Pressure Data for the Precursors 1, 2, and 4
A
B
ΔHsub [kJ mol−1]
temp range of measurement [°C]
calcd vapor pressure at 85 °C (torr)
1
2
4
11.37 ± 0.04
11.89 ± 0.14
12.42 ± 0.04
26.79 ± 0.14
30.32 ± 0.44
32.70 ± 0.13
22.27 ± 0.12
25.21 ± 037
27.19 ± 0.11
28−67
33−60
33−75
7.83
2.69
1.95
a
Data sets are fitted to log P (Mtorr) = A − B/T. Error margins are established by a linear regression method and given at the 95% confidence
interval.
G
DOI: 10.1021/acs.inorgchem.6b01146
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 7. Top-down SEM (scale: 2 μm), cross-sectional SEM (×25 000 magnification) and AFM (2 μm × 2 μm) micrographs for films deposited at
a substrate temperature of (a) 325, (b) 350, (c) 375, and (d) 400 °C. Insert (7c) shows a hexagonal crystallite within the thin film (250 nm cross-
section).
Table 5. Atomic % of Elements in Cobalt Thin Films
Deposited between 325 and 400 °C Post-Ar Etching (15 s)
a
a
substrate temperature
atom % Co
96.8
atom % O
3.2
325 °C
3
50 °C†
375 °C
400 °C
97.0
99.7
3.0
0.3
a
Quantification was performed assuming zero contribution from the
†
silicon substrate. Values are expressed as atomic percentages. Films
were too thin to obtain reliable elemental profile across the sample.
metallic cobalt films with oxygen and carbon impurities of
14b
1
.0% and 1.2%, respectively. Considering the high solubility
53
of carbon in cobalt the fact that no carbon can be detected in
the deposited films at temperatures between 325 and 400 °C is
quite remarkable and shows that the hydrogen carrier gas is
effective in the hydrogenation of all ligands cleanly from the
substrate surface during deposition. There are, however, oxygen
impurities present in all films, and considering that the
precursor contains no oxygen, it is reasonable to assume that
oxygen contamination is coming either from the gas source, a
leak in the reactor, or ex situ contamination.
Figure 8. XRD patterns of cobalt thin films deposited between 325
and 400 °C clearly showing the strong reflection from the Co (111)
miller plane. The PXRD pattern of the (100) silicon substrate is also
shown.
Figure 9 shows an XPS spectrum for a sample produced at
00 °C showing the presence of photoelectron peaks associated
4
with metallic cobalt and the 1s regions for carbon (284 eV) and
oxygen (532 eV), which are common contaminants in
deposited cobalt films. Analysis indicates that, in general, purity
increases with increasing substrate temperature. Films
deposited at a substrate temperature of 400 °C exhibit the
highest purity (99.7%) with trace amounts of oxygen (0.3%)
and no carbon; this compares favorably to previous deposition
experiments, which have shown CpCo(CO)2 to deposit
17c
In stark contrast to our studies Dormans et al. suggest that
the growth of cobalt thin films using the precursor CpCo(CO)2
are significantly retarded in the presence of H gas resulting in
2
thin films at any temperature that are heavily contaminated
with carbon (∼50%), a feature attributed to a change in
deposition mechanism from reduction to disproportionation.
Similar observations have been made in the deposition of nickel
H
DOI: 10.1021/acs.inorgchem.6b01146
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 9. Survey XPS spectrum of a Co thin film deposited at a substrate temperature of 400 °C (above), with the 1s regions for oxygen (bottom
left) and carbon (bottom right) highlighted.
5
4
−1
thin films from nickelocene by Rhee et al. and Brissonneau et
growth rate (4.18 nm min ) was reached at 375 °C and
55
al. It is also noteworthy that while Raman spectroscopy shows
diminishes thereafter, possibly due to desorption of the
precursor from the substrate or precursor depletion on the
reactor walls at elevated temperatures.
deposited at 350 °C show a diminished growth rate, possibly
−1
a single Raman band at 520.9 cm , and PXRD analysis shows a
number of weak reflections, both associated with the silicon
substrate, there is no evidence from our analysis (including
XPS) of cobalt silicide systems (i.e., Co Si, CoSi, and
contaminating the thin films.
The calculated growth rates for films grown from precursor 1
14b
In contrast films
2
1
7b,18b,56
due to competition between hydrogen and precursor for surface
CoSi2)
57
sites at this temperature.
at 325, 350, 375, and 400 °C, respectively, are shown in Table
Room-temperature sheet resistance measurements were also
conducted on films deposited using 1, the findings are
summarized in Table 6. Bulk resistivity values were found to
be in the range of 62.1−12.5 μΩ cm, the lower of which
6
; the values comparing favorably with previous studies
reported in the literature by Haycock et al. using CpCo-
1
4b,c
−1
(CO)2
(1.38−1.40 nm min ), although there is a drop in
growth rate at 350 °C, which is unexplained. The highest
58
approaches the value for bulk cobalt 5.6 μΩ cm (at 0 °C).
a
Films deposited at 375 °C displayed the lowest resistivity value,
which may be rationalized by the high crystallinity and
thickness of the film (vide infra). Resistivity values of thin
films can be increased by a number of effects, including
increasing surface roughness, a higher degree of scattering at
grain boundaries and with decreasing film thickness, electron
Table 6. Average Bulk Resistivity (ρ) and Sheet Resistance
(RS) Values for Cobalt Films Deposited from 1
film thickness (nm) [growth rate]
RS
ρ
1
temperature
(nm min−
)
(Ω/□) (μΩ cm)
3
3
3
4
25
50
75
00
79.5 [2.65]
51.8 [1.94]
4.05
13.80
1.00
32.4
62.1
12.5
17.6
125.3 [4.18]
80.5 [2.68]
58
scattering at a films surface increases. In addition, it is likely
2.20
that ex situ oxidation of the films will lead to an increase in
a
58
15
The bulk resistivity of cobalt = 5.6 μΩ cm (at 0 °C).
resistivity.
I
DOI: 10.1021/acs.inorgchem.6b01146
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
CONCLUSIONS
REFERENCES
■
■
This paper reports the synthesis and thermal analysis of 11
CpCo(diolefin) complexes and reports the first study of their
use as metal deposition precursors in atmospheric pressure
CVD. Because of its high thermal stability and physical
properties compound 1 was identified as a potential candidate
for deposition studies. Thin films were deposited onto Si
substrates in the temperature range of 325−400 °C in the
(1) Jung, H. K.; Lee, H. B.; Tsukasa, M.; Eunji, J.; Yun, J. H.; Lee, J.
M.; Choi, G. H.; Choi, S.; Chung, C. IEEE Reliability Physics
Symposium (IRPS) 2011, 307.
2) Yang, C. C.; Flaitz, P.; Wang, P. C.; Chen, F.; Edelstein, D. IEEE
Electron Device Lett. 2010, 31, 728.
3) He, M.; Zhang, X.; Nogami, T.; Lin, X.; Kelly, J.; Kim, H.;
(
(
Spooner, T.; Edelstein, D.; Zhao, L. J. Electrochem. Soc. 2013, 160,
D3040.
presence of H . Films deposited at all temperatures were found
2
(4) Barthel
F.; Petroff, F.; Steren, L. B. Nanostruct. Mater. 1995, 6, 217.
5) Fert, A.; Barthelemy, A.; Galtier, P.; Holody, P.; Loloee, R.;
Morel, R.; Petroff, F.; Schroeder, P.; Steren, L. B.; Valet, T. Mater. Sci.
́ ́
emy, A.; Cros, V.; Duvail, J. L.; Fert, A.; Morel, R.; Parent,
to have very low carbon contamination by XPS analysis, with
varying degrees of oxygen incorporation. The lowest measured
resistivity was 12.7−17.6 μΩ cm for films deposited at 375−
(
́
́
́
4
00 °C, which is in broad agreement with the purity and
Eng., B 1995, 31, 1.
continuity of the films.
(6) Steynberg, A.; Dry, M., Fischer−Tropsch Technology; Elsevier:
Amsterdam, 2004.
7) Jung, J.-S.; Kim, S. W.; Moon, D. J. Catal. Today 2012, 185, 168.
8) Ohtsuka, Y.; Arai, T.; Takasaki, S.; Tsubouchi, N. Energy Fuels
003, 17, 804.
9) Taheri Najafabadi, A.; Khodadadi, A. A.; Parnian, M. J.;
Mortazavi, Y. Appl. Catal., A 2016, 511, 31.
10) Khodakov, A. Y.; Chu, W.; Fongarland, P. Chem. Rev. 2007, 107,
692.
11) Ko, Y. K.; Park, D. S.; Seo, B. S.; Yang, H. J.; Shin, H. J.; Kim, J.
As indicated by XRD analysis the thin films deposited appear
to be highly oriented, adopting the fcc crystal structure;
crystallinity was found to increse proportionally with increasing
deposition temperature at 375 °C, after which crystallinity
appears to diminish. More significantly, complex 1 represents a
nonair-sensitive precursor capable of depositing cobalt metal
onto silicon substrates at a low enough temperature to preclude
the concomitant formation of cobalt silicide.
(
(
2
(
(
1
(
These precursors offer significant advantages over a number
of previously reported MO-CVD precursors for the production
of cobalt thin films, which include the ease with which these
materials can be scaled-up and modified, physical properties
such as stability (thermal and chemical), the precursors general
liquid nature at room temperature and relatively high vapor
pressures, and finally ability to produce high-purity thin films
under the conditions we have described.
Y.; Lee, J. H.; Lee, W. H.; Reucroft, P. J.; Lee, J. G. Mater. Chem. Phys.
2003, 80, 560.
(12) Alameda, J. M.; Carmona, F.; Salas, F. H.; Alvarez-Prado, L. M.;
Morales, R.; Per
13) Szmaja, W.; Kozłowski, W.; Balcerski, J.; Kowalczyk, P. J.;
Grobelny, J.; Cichomski, M. J. Alloys Compd. 2010, 506, 526.
14) (a) Yang, J.; Li, K.; Feng, J.; Gordon, R. G. J. Mater. Chem. C
015, 3, 12098. (b) Chioncel, M. F.; Haycock, P. W. Chem. Vap.
́
ez, G. T. J. Magn. Magn. Mater. 1996, 154, 249.
(
(
2
Deposition 2005, 11, 235. (c) Chioncel, M. F.; Nagaraja, H. S.;
Rossignol, F.; Haycock, P. W. J. Magn. Magn. Mater. 2007, 313, 135.
Further studies into the utility of several of these complexes
in the CVD of cobalt thin films at lower temperatures are
presently underway.
(
(
15) Lim, B. S.; Rahtu, A.; Gordon, R. G. Nat. Mater. 2003, 2, 749.
16) (a) Pugh, T.; Cosham, S. D.; Hamilton, J. A.; Kingsley, A. J.;
Johnson, A. L. Inorg. Chem. 2013, 52, 13719. (b) Klesko, J. P.;
ASSOCIATED CONTENT
Kerrigan, M. M.; Winter, C. H. Chem. Mater. 2016, 28, 700.
■
(17) (a) Rhee, H. S.; Ahn, B. T. J. Electrochem. Soc. 1999, 146, 2720.
*
S
Supporting Information
(b) Papadopoulos, N.; Karayianni, C.-S.; Tsakiridis, P.; Sarantopoulou,
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.inorg-
chem.6b01146. This publication is supported by multiple data
sets. Data relating to the TGA, isothermal TGA, volatility, and
XPS spectra of the films deposited at 325, 350, 375, and 400
E.; Hristoforou, E. Chem. Vap. Deposition 2011, 17, 211. (c) Dormans,
G. J. M.; Meekes, G. J. B. M.; Staring, E. G. J. J. Cryst. Growth 1991,
114, 364. (d) Ye, D. X.; Pimanpang, S.; Jezewski, C.; Tang, F.;
Senkevich, J. J.; Wang, G. C.; Lu, T. M. Thin Solid Films 2005, 485, 95.
(
18) (a) Ivanova, A. R.; Nuesca, G.; Chen, X. M.; Goldberg, C.;
Kaloyeros, A. E.; Arkles, B.; Sullivan, J. J. J. Electrochem. Soc. 1999, 146,
139. (b) Lane, P. A.; Oliver, P. E.; Wright, P. J.; Reeves, C. L.; Pitt, A.
°
C, created during this research, are openly available from the
Univ. of Bath data archive at http://dx.doi.org/10.13039/
2
D.; Cockayne, B. Chem. Vap. Deposition 1998, 4, 183. (c) Londergan,
A. R.; Nuesca, G.; Goldberg, C.; Peterson, G.; Kaloyeros, A. E.; Arkles,
B.; Sullivan, J. J. J. Electrochem. Soc. 2001, 148, C21.
501100000266 free of charge.
Tabulated crystallographic information for complexes 2,
(19) (a) Rhee, H. S.; Ahn, B. T. Appl. Phys. Lett. 1999, 74, 3176.
5, and 6, vapor pressure reports (PDF)
(
(
b) Rhee, H. S.; Ahn, B. T.; Sohn, D. K. J. Appl. Phys. 1999, 86, 3452.
c) Rhee, H. S.; Jang, T. W.; Ahn, B. T. Appl. Phys. Lett. 1999, 74,
X-ray crystallographic files (CIF)
1
1
003. (d) Rhee, H. S.; Sohn, D. K.; Ahn, B. T. MRS Online Proc. Libr.
999, 564, 145. (e) Chioncel, M. F.; Haycock, P. W. Chem. Vap.
AUTHOR INFORMATION
■
*
Deposition 2005, 11, 235.
(20) West, G. A.; Beeson, K. W. Appl. Phys. Lett. 1988, 53, 740.
Corresponding Author
E-mail: A.L.Johnson@bath.ac.uk. Phone: 44 (0)1225 384467.
Fax: 44 (0)1225 386231.
(
(
21) Choi, H.; Park, S. Chem. Mater. 2003, 15, 3121.
22) Lee, K.; Park, T.; Lee, J.; Kim, J.; Kim, J.; Kwak, N.; Yeom, S.;
Jeon, H. Jpn. J. Appl. Phys. 2008, 47, 5396.
23) (a) Georgi, C.; Hildebrandt, A.; Tuchscherer, A.; Oswald, S.;
Notes
(
The authors declare no competing financial interest.
Lang, H. Z. Anorg. Allg. Chem. 2013, 639, 2532. (b) Georgi, C.;
Hildebrandt, A.; Waechtler, T.; Schulz, S. E.; Gessner, T.; Lang, H. J.
Mater. Chem. C 2014, 2, 4676.
ACKNOWLEDGMENTS
■
(24) Odedra, R.; Boag, N.; Anthis, J.; Kanjolia, R. Application: WO
We acknowledge the financial support of the EPSRC (EP/
I019278/1), Univ. of Bath (studentships to T.P.), and SAFC-
Hitech for case-award support (T.P).
2
011017068; Sigma-Aldrich Co: United States, 2011.
(25) Li, Z.; Lee, D. K.; Coulter, M.; Rodriguez, L. N. J.; Gordon, R.
G. Dalton Trans. 2008, 2592.
J
DOI: 10.1021/acs.inorgchem.6b01146
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(
26) Knisley, T. J.; Saly, M. J.; Heeg, M. J.; Roberts, J. L.; Winter, C.
H. Organometallics 2011, 30, 5010.
27) Kim, K.; Lee, K.; Han, S.; Jeong, W.; Jeon, H. J. Electrochem. Soc.
007, 154, H177.
28) Lee, H.-B.-R.; Kim, H. Electrochem. Solid-State Lett. 2006, 9,
G323.
(54) Kang, J.-K.; Rhee, S.-W. J. Mater. Res., 2000, 15, 1828.
(55) Brissonneau, L.; Reynes, A.; Vahlas, C. Chem. Vap. Deposition
1999, 5, 281.
(56) Paranjape, M. A.; Mane, A. U.; Raychaudhuri, A. K.; Shalini, K.;
Shivashankar, S. A.; Chakravarty, B. R. Thin Solid Films 2002, 413, 8.
(57) Martin, J. D.; Hogan, P.; Abboud, K. A.; Dahmen, K.-H. Chem.
Mater. 1998, 10, 2525.
(
2
(
(
29) (a) Lee, H.-B.-R.; Park, Y. J.; Baik, S.; Kim, H. Chem. Vap.
Deposition 2012, 18, 41. (b) Diskus, M.; Nilsen, O.; Fjellvag, H. Chem.
Vap. Deposition 2011, 17, 135. (c) Lee, H.-B.-R.; Kim, H. ECS Trans.
(58) Lane, P. A.; Oliver, P.; Wright, P. J.; Reeves, C. L.; Pitt, A. D.;
Cockayne, B. Chem. Vap. Deposition 1998, 4, 183.
2
(
008, 16, 219.
30) Kwon, J.; Saly, M.; Halls, M. D.; Kanjolia, R. K.; Chabal, Y. J.
Chem. Mater. 2012, 24, 1025.
31) (a) Lim, B. S.; Rahtu, A.; Park, J.-S.; Gordon, R. G. Inorg. Chem.
003, 42, 7951. (b) Wu, J.; Li, J.; Zhou, C.; Lei, X.; Gaffney, T.;
Norman, J. A. T.; Li, Z.; Gordon, R.; Cheng, H. Organometallics 2007,
(
2
2
6, 2803. (c) Lim, B. S.; Rahtu, A.; Gordon, R. G. Nat. Mater. 2003, 2,
7
49. (d) Kim, H. Microelectron. Eng. 2013, 106, 69.
(
32) (a) Kalutarage, L. C.; Martin, P. D.; Heeg, M. J.; Winter, C. H.
Inorg. Chem. 2013, 52, 5385. (b) Klesko, J. P.; Thrush, C. M.; Winter,
C. H. Chem. Mater. 2015, 27, 4918.
(
33) (a) Donders, M. E.; Knoops, H. C. M.; van, M. C. M.; Kessels,
W. M. M.; Notten, P. H. L. J. Electrochem. Soc. 2011, 158, G92.
b) Han, B.; Choi, K. H.; Park, J. M.; Park, J. W.; Jung, J.; Lee, W.-J. J.
(
Vac. Sci. Technol., A 2013, 31, 01A145. (c) Klepper, K. B.; Nilsen, O.;
Fjellvag, H. J. Cryst. Growth 2007, 307, 457. (d) Klepper, K. B.; Nilsen,
O.; Fjellvag, H. Thin Solid Films 2007, 515, 7772.
(
34) Jonas, K.; Deffense, E.; Habermann, D. Angew. Chem., Int. Ed.
Engl. 1983, 22, 716.
35) Hapke, M.; Weding, N.; Spannenberg, A. Organometallics 2010,
9, 4298.
36) Georgi, C.; Hapke, M.; Thiel, I.; Hildebrandt, A.; Waechtler, T.;
Schulz, S. E.; Lang, H. Thin Solid Films 2015, 578, 180.
37) (a) Chen, R. S.; Chen, Y. S.; Huang, Y. S.; Chen, Y. L.; Chi, Y.;
Liu, C. S.; Tiong, K. K.; Carty, A. J. Chem. Vap. Deposition 2003, 9,
01. (b) Zhang, F. Y.; Barrowcliff, R.; Stecker, G.; Pan, W.; Wang, D.
L.; Hsu, S. T. Jpn. J. Appl. Phys. 2 2005, 44, L398.
(
2
(
(
3
(
38) Wakatsuki, Y.; Yamazaki, H.; et al. Inorg. Synth. 1989, 26, 189.
(
39) Panda, T. K.; Gamer, M. T.; Roesky, P. W. Organometallics
2
(
1
(
003, 22, 877.
40) King, R. B.; Stone, F. G. A.; Treichel, P. M. J. Am. Chem. Soc.
961, 83, 3593.
41) Sheldrick, G. M. SHELX-86, Computer Program for Crystal
Structure Determination; University of Gottingen: Germany, 1986.
42) (a) Sheldrick, G. M. SHELX-97, Computer Program for Crystal
Structure Refinement; Univerity of Gottingen: Germany, 1997.
b) Sheldrick, G. M. Acta Crystallogr., Sect. A: Found. Crystallogr.
008, 64, 112.
̈
(
̈
(
2
(
(
43) Otwinowski, Z.; Minor, W. Methods Enzymol. 1997, 276, 307.
44) Farrugia, L. J. J. Appl. Crystallogr. 2012, 45, 849.
(
45) (a) Gutnov, A.; Drexler, H. J.; Spannenberg, A.; Oehme, G.;
Heller, B. Organometallics 2004, 23, 1002. (b) Lee, W. S.; Brintzinger,
H. H. J. Organomet. Chem. 1981, 209, 401.
(
1
(
46) Drew, M. G. B.; Sloan, M.; Nelson, S. M. J. Organomet. Chem.
972, 39, C9.
47) Willcocks, A. M.; Pugh, T.; Cosham, S. D.; Hamilton, J.; Sung,
S. L.; Heil, T.; Chalker, P. R.; Williams, P. A.; Kociok-Kohn, G.;
Johnson, A. L. Inorg. Chem. 2015, 54, 4869.
(
48) (a) Rushworth, S.; Davies, H.; Kingsley, A.; Leese, T.; Odedra,
R. Microelectron. Reliab. 2007, 47, 718. (b) Rushworth, S. A.; Smith, L.
M.; Kingsley, A. J.; Odedra, R.; Nickson, R.; Hughes, P. Microelectron.
Reliab. 2005, 45, 1000.
(
49) Vieyra-Eusebio, M. T.; Rojas, A. J. Chem. Eng. Data 2011, 56,
008.
50) Ivanova, A. R.; Nuesca, G.; Chen, X.; Goldberg, C.; Kaloyeros,
A. E.; Arkles, B.; Sullivan, J. J. J. Electrochem. Soc. 1999, 146, 2139.
5
(
(
(
&
51) Maruyama, T.; Tago, T. J. Mater. Sci., 28, 5345.
52) West, A. R. Solid State Chemistry and its Applications; John Wiley
Sons Ltd: Hoboken, NJ, 1984.
(
53) May, P. W. Endeavour 1995, 19, 101.
K
DOI: 10.1021/acs.inorgchem.6b01146
Inorg. Chem. XXXX, XXX, XXX−XXX