Silver(I) aryloxide-triphenylphosphine complexes. An assessment of their potential as CVD precursors and the structure of [Ag(OC6H4Me-2)(PPh3)3]·2- MeC6H4OH·C6H5Me

Article abstract of DOI:10.1016/S0020-1693(03)00077-X

The silver aryloxide-triphenylphoshine complexes [Ag(OR)(PPh₃)₂] (R=C₆H₂Cl₃-2,4,6 (1), C₆H₄Me-2 (2) or C₆H₂(CH₂NMe₂)₃-2,4,6 (3)) and [Ag(OR)(PPh₃)₃]·ROH (R=Ph (4) or C₆H₄Me-2 (5)) have been prepared and characterised. Complexes 1, 3 and 5 were tested as precursors for the growth of silver films on glass substrates using an aerosol-assisted chemical vapour deposition method. Although films were produced, they were of poor quality compared to those produced from other silver complexes as precursors. A crystal structure determination of 5 as a toluene solvate shows the o-cresolato ligand to be monodentate and together with the triphenylphosphine ligands generates a distorted tetrahedral array around the silver. The additional o-cresol molecule is hydrogen-bonded to the o-cresolato oxygen atom and not directly bonded to silver.

Full text of DOI:10.1016/S0020-1693(03)00077-X

Inorganica Chimica Acta 349 (2003) 37Á44
/
www.elsevier.com/locate/ica
Silver(I) aryloxideÁ
their potential as CVD precursors and the structure of
[Ag(OC6H4Me-2)(PPh3)3]×2-MeC6H4OH×C6H5Me
/
triphenylphosphine complexes. An assessment of
/
/
Dennis A. Edwards, Mary F. Mahon, Kieran C. Molloy *, Virginie Ogrodnik
Department of Chemistry, University of Bath, Claverton Down, Bath BA2 7AY, UK
Received 17 August 2002; accepted 30 December 2002
Abstract
The silver aryloxideÁ
/
triphenylphoshine complexes [Ag(OR)(PPh3)2] (Rꢀ/C6H2Cl3-2,4,6 (1), C6H4Me-2 (2) or C6H2(CH2NMe2)3-
2,4,6 (3)) and [Ag(OR)(PPh3)3]×
/
ROH (RꢀPh (4) or C6H4Me-2 (5)) have been prepared and characterised. Complexes 1, 3 and 5
/
were tested as precursors for the growth of silver films on glass substrates using an aerosol-assisted chemical vapour deposition
method. Although films were produced, they were of poor quality compared to those produced from other silver complexes as
precursors. A crystal structure determination of 5 as a toluene solvate shows the o-cresolato ligand to be monodentate and together
with the triphenylphosphine ligands generates a distorted tetrahedral array around the silver. The additional o-cresol molecule is
hydrogen-bonded to the o-cresolato oxygen atom and not directly bonded to silver.
# 2003 Elsevier Science B.V. All rights reserved.
Keywords: Crystal structures; Silver complexes; Aryloxide complexes
1. Introduction
The alternative approach utilises the reaction of silver
oxide and an aryl alcohol in aqueous acetone.
Although metal alkoxides and aryloxides are an
extremely well established group of compounds [1],
relatively little is known of such compounds of silver(I)
reflecting its low affinity for oxygen donor ligands.
Some 25 years ago, a study [2] of reactions between
silver ions and alkoxide ions illustrated the instability of
silver alkoxides. For example, reactions of AgClO4 with
Silver aryloxides appear to be polymeric via bridging
aryloxide ligands as shown [4] by a structure determina-
tion of [Ag(OC6H2Cl3-2,4,5)]n in which the silver is
bonded to three oxygens from three different aryloxide
ligands and to one carbon from another aryl ring which
also uses its oxygen to bond to a further silver in the
polymer. The geometry around each silver is approxi-
mately trigonal pyramidal with respect to the oxygens
LiOR (Rꢀ
of LiOR with Rꢀ
/
Me or Et) afforded pure silver whereas use
t-Bu or i-Pr produced a mixture of
/
but a further weak AgÃ
/
Cl interaction gives the silver an
silver, silver alkoxide and other uncharacterised silver
species. However, since then a few isolated reports of
silver aryloxides have appeared, two general methods of
preparation having been employed [3]. The first involves
the reaction of a silver salt and a sodium or lithium
aryloxide in the parent alcohol or in tetrahydrofuran.
overall trigonal bipyramidal coordination.
Attempts to generate monomeric silver aryloxides
have involved the use of additional ‘soft’ ligands which
have a greater affinity for silver(I) than aryloxides. Thus
monomeric [Ag(OC6H2Cl3-2,4,6)(PPh3)2] has monoden-
tate aryloxide and triphenylphosphine ligands arranged
in a trigonal planar fashion around the metal with again
an additional weak intramolecular AgÁ Á ÁCl interaction
[5]. Apart from this phosphine example, the macrocyclic
polythiaether ligand 1,4,7,10,13,16-hexathiacyclotetra-
* Corresponding author. Tel.: ꢁ
386 231.
E-mail address: chskcm@bath.ac.uk (K. Molloy).
/
44-1225-386 382; fax: ꢁ44-1225-
/
0020-1693/03/$ - see front matter # 2003 Elsevier Science B.V. All rights reserved.
doi:10.1016/S0020-1693(03)00077-X

38
D. Edwards et al. / Inorganica Chimica Acta 349 (2003) 37Á44
/
decane (18S6), has been used [6] to stabilise the picrate
[Ag2(pic)2(18S6)] and a dimeric quinolinolato complex
films on scrupulously clean glass substrates under an
atmospheric pressure of dinitrogen using a horizontal
cold wall CVD reactor. The precursor was delivered to
the gas phase as an aerosol generated by nebulising an
anhydrous tetrahydrofuran solution, the aerosol being
swept into the reactor using dinitrogen as carrier gas.
Film quality was estimated by EDXS techniques using a
JEOL Superprobe instrument operating at an accelerat-
[Ag2(C9H6NOH)3(C9H6NO)(PPh3)2]ꢁ[CH3CO2]ꢂ
×
/
H2O containing
a
N,O-chelating quinolinolato
C9H6NO ligand [7] is also known. In addition, we
have reported the only crystallographically-authenti-
cated example of a silver siloxide, viz {[Ag(PPh3)2(m4-
OMe2SiOSiMe2O)}2 [8].
We have already reported [9] that good quality silver
films can be produced by the aerosol-assisted chemical
vapour deposition technique (AACVD) using silver
carboxylate, fluorocarboxylate, b-ketoenolate or b-ke-
toiminate-trimethyl- or triphenyl-phosphine complexes
as precursors. It was also felt worthwhile to attempt the
ing voltage of 5 or 10 kV with a beam current of 5ꢃ
/
10ꢂ8 or 5ꢃ10ꢂ9 A.
/
2.2. Syntheses
As far as possible, syntheses were carried out in the
absence of light and under an atmosphere of dintrogen
using previously dried solvents.
synthesis of some silver aryloxideÁ/phosphine complexes
containing analogous weak AgÃO and strong AgÃ
/
/P
interactions and subject them to AACVD to assess the
possibility of producing coherent silver films with
desirable features such as high reflectivity and hardness.
Alkoxides and aryloxides of several metals e.g., Zr, Sn,
2.2.1. (2,4,6-Trichlorophenolato)bis(triphenyl-
phosphine)silver(I) (1)
This compound was prepared by an adaptation of the
literature method [5]. Silver oxide (0.23 g, 1.0 mmol),
2,4,6-trichlorophenol (0.43 g, 2.2 mmol) and triphenyl-
phosphine (1.15 g, 4.4 mmol) were heated together for 2
h in refluxing toluene (75 cm3). After filtration of the hot
Nb have been used as precursors for either solÁgel or
/
CVD production of metal oxide films, but, given the
thermal instability of silver oxide above approximately
160 8C, a metal film would be expected from the
AACVD of silver aryloxide complexes. Indeed, analo-
gous copper metal films have been deposited by CVD of
[Cu(OBut)]4 [10] and the adduct [Cu(OBut)(PMe3)] [11].
mixture, 60Á808 boiling range petroleum spirit was
/
added to the filtrate until a precipitate appeared. This
colourless solid was filtered, washed with toluene and
dried in vacuo. Yield 1.34 g, 81%. Anal. Found (Calc.
for C42H32AgCl3OP2): C, 60.9 (60.9); H, 3.87 (3.89)%.
2. Experimental
1
IR: 1583w, 1558w, 1516w, 1435m, 1244m [n(CO)]. H
NMR (CDCl3): 7.38Á
Cl3C6H2O). 13C NMR (CDCl3): 160.0 (CO), 133.9 (d,
2Jꢀ16.5 Hz, PC6H5), 132.6 (Cl3C6H2O), 132.1 (d, 1Jꢀ
27.0 Hz, PC6H5), 130.1 (PC6H5), 128.8 (d, 3Jꢀ
10.0 Hz,
PC6H5), 127.1 (Cl3C6H2O), 123.7 (Cl3C6H2O), 112.7
/
7.24 (m, 30H, C6H5), 6.97 (s, 2H,
2.1. General
/
/
IR spectra were recorded as Nujol or hexachlorobu-
tadiene mulls using a Nicolet 510P FT spectrometer
within the range 4000Á
400 cmꢂ1: only selected data are
/
/
(Cl3C6H2O). 31P NMR (ꢂ
/
25 8C, CDCl3/CH2Cl2); 9.38
quoted later in this section. Microanalyses for carbon,
hydrogen and nitrogen were carried out using a Carlo-
Erba Model 1106 microanalyser. JEOL GX 270 or EX
400 FT spectrometers were used to obtain 1H, 13C
(SiMe4 reference) and 31P (85% H3PO4 reference) NMR
spectra. FAB (LSIMS) mass spectra were recorded by
the EPSRC mass spectrometry service at the University
of Wales, Swansea using caesium ion bombardment at
25 kV, the samples being in a 3-nitrobenzyl alcohol
matrix, dissolution being assisted by use of dichloro-
methane co-solvent.
Thermal analysis experiments were carried out at
atmospheric pressure in a helium-purged atmosphere
from room temperature (r.t.) to 450 8C at a ramp rate of
25 8C minꢂ1 using a Dupont 951 TGA module, a
Dupont 910 DSC cellbase and a TA Instruments 2100
controller. Details of the apparatus and experimental
methodology for the AACVD of films generated from
silver compounds as precursors have been given in full
elsewhere [9]. Essentially, attempts were made to grow
1
(unresolved d, J(AgP)ꢀ
/
345 Hz).
2.2.2. (o-Cresolato)bis(triphenylphosphine)silver(I)
(2)
A mixture of silver oxide (1.00 g, 4.3 mmol), o-cresol
(2-methylphenol) (0.93 g, 8.6 mmol) and triphenylpho-
sphine (4.46 g, 17.0 mmol) was stirred overnight at r.t. in
toluene (40 cm3). The resulting light grey product was
filtered and washed with toluene before drying in vacuo.
Yield 4.77 g, 75%. Anal. Found (Calc. for C43H37-
AgOP2): C, 69.3 (69.8); H, 5.06 (5.04)%. IR: 1622w,
1583m, 1435m, 1294w, 1278m [n(CO)]. 1H NMR
(CDCl3): 7.40Á
/
7.24 (m, 34H, C6H5ꢁMeC6H4O), 2.21
/
(s, 3H, CH3). 13C NMR (CDCl3): 154.1 (CO), 134.0 (d,
2Jꢀ
/
16.5 Hz, PC6H5), 132.5 (d, Jꢀ
/
25.4 Hz, PC6H5),
3
1
130.8 (MeC6H4O), 129.9 (PC6H5), 128.8 (d, Jꢀ10.0
/
Hz, PC6H5), 126.9 (MeC6H4O), 123.8 (MeC6H4O),
120.2 (MeC6H4O), 114.9 (MeC6H4O), 15.7 (CH3). 31P
NMR (ꢂ25 8C, CDCl3/CD2Cl2): 6.77.
/

D. Edwards et al. / Inorganica Chimica Acta 349 (2003) 37Á
/44
39
2.2.3. [2,4,6-Tris(dimethylaminomethyl)phenolato]bis-
(triphenylphosphine)silver(I) (3)
MeC6H4OHꢁ
(s, 6H, CH3C6H4Oꢁ
(CDCl3): 156.3 (CO), 136.4 (C6H5CH3), 133.9 (d, Jꢀ
16.5 Hz, PC6H5), 133.3 (d, Jꢀ17.6 Hz, PC6H5), 132.6
/
C6H5CH3), 2.35 (s, 3H, C6H5CH3), 2.24
CH3C6H4OH). 13C NMR
/
2
A mixture of silver oxide (1.00 g, 4.3 mmol), 2,4,6-
tris(dimethylaminomethyl)phenol (2.28 g, 8.6 mmol)
and triphenylphosphine (4.46 g, 17.0 mmol) was stirred
in toluene (40 cm3) overnight at r.t. Filtration of the
resulting darkly coloured mixture gave a clear filtrate
which deposited a very light grey solid on standing at r.t.
This product was filtered, washed with toluene and
dried in vacuo. Yield 5.09 g, 66%. Anal. Found (Calc.
for C51H56AgN3OP2): C, 67.6 (68.3); H, 6.19 (6.29); N,
4.45 (4.69)%. IR: 1604m, 1435w, 1311s, 1229m [n(CO)],
1170w [n(CN)], 1142w [n(CN)]. 1H NMR (CDCl3):
/
1
/
(MeC6H4O), 132.1 (MeC6H4O), 132.0 (MeC6H5O),
130.4 (MeC6H4OH), 129.7 (PC6H5), 129.0 (C6H5CH3),
3
128.7 (d, Jꢀ
/
8.8 Hz, PC6H5), 128.4 (C6H5CH3), 128.2
(MeC6H4O), 126.7 (MeC6H4OH), 125.2 (MeC6H4O),
124.3 (MeC6H4OH), 118.6 (MeC6H4OH), 115.5 (Me-
C6H4OH), 21.4 (CH3C6H5), 16.2 (CH3C6H4Oꢁ
/
CH3C6H4OH). 31P NMR: (ꢂ
4.51.
/
25 8C, CDCl3/CD2Cl2),
7.43Á
/
7.23
(m,
30H,
C6H5),
6.94
(Me2NCH2)3C6H2O], 3.52 (s, 4H, CH2N), 3.29 (s, 2H,
[s,
2H,
2.3. X-ray crystallography
CH2N), 2.28 [s, 12H, (CH3)2N], 2.21 [s, 6H, (CH3)2N].
13C NMR (CDCl3): 155.5 (CO), 134.0 (d, Jꢀ
2
/16.5 Hz,
Crystallographic quality crystals of 5 were obtained
from a toluene solution cooled in a refrigerator. Crystal
1
PC6H5), 132.5 (d, Jꢀ
/
26.4 Hz, PC6H5), 130.0 (PC6H5),
128.3
128.7 (d,
3Jꢀ
/8.9
Hz,
PC6H5),
data: C75H68AgO2P3, Mꢀ
/
1202.07, triclinic, space
13.347(2), bꢀ14.991(2), cꢀ
81.65(2), bꢀ74.42(2), gꢀ65.27(2)8,
3150.9 A , Zꢀ2, Dcalc 0.442
1.267 g cmꢂ3, mꢀ
mmꢂ1, F(000)ꢀ
1252. Data were collected at 293(2) K
on a CAD-4 automatic four-circle diffractometer in the
¯
[(Me2NCH2)3C6H2O], 123.0 [(Me2NCH2)3C6H2O], 63.8
(CH2N), 60.3 (CH2N), 45.1 [(CH3)2N], 44.8 [(CH3)2N].
group P1 (no. 2), aꢀ
/
/
/
˚
18.012(3) A, aꢀ
/
/
/
31P (ꢂ
/
25 8C, CDCl3/CD2Cl2): 7.32.
Vꢀ
/
/
ꢀ
/
/
3
˚
/
2.2.4. (Phenolato)tris(triphenylphosphine)silver(I)×
/
˚
phenol (4)
range 2.30B
radiation. A crystal of dimensions 0.35ꢃ
mm was used for data collection. The index ranges were
05h 515, ꢂ155k 517 and ꢂ195l 520. A total of
10 336 reflections were measured of which 9853 were
unique, [Rint 0.0412].
/
u B
/
23.928 using Mo Ka (0.71069 A)
This compound was prepared in a similar manner to
the copper(I) analogue [12]. A mixture of silver oxide
(2.32 g, 10.0 mmol), phenol (1.88 g, 20.0 mmol) and
triphenylphosphine (10.49 g, 40.0 mmol) was heated in
toluene (80 cm3) at 75 8C for 10 h. The pale brown
solution was filtered while still hot and the filtrate
cooled to r.t. After standing overnight in the dark, the
product precipitated as a light beige solid which was
filtered, washed with toluene and dried in vacuo. Yield
7.79 g, 72% (based on phenol as limiting reagent). Anal.
Found (Calc. for C66H56AgO2P3): C, 73.0 (73.2); H, 5.45
(5.18)%. IR: 3368w,br, 3184w [n(OH)], 1622m, 1606w,
1585w, 1435s, 1248m [n(CO)].
/
0.35ꢃ0.30
/
/
/
/
/
/
/
/
/
ꢀ
/
The diffraction intensities were corrected for Lp but
not for absorption. The solution of the structure and
refinement (SHELX) converged to a conventional [i.e.
based on 5532 reflections with Foꢀ
and wR2 of 0.1457 for a weighting scheme of wꢀ
[s2(Fo)2ꢁ(0.0907P)2ꢁ (Fo2ꢁ2Fc2)/
1.9303P] where Pꢀ
/
4s(Fo)] R1 of 0.0693
/
1/
/
/
/
/
3. The goodness-of-fit was 1.066, with maximum and
ꢂ3
˚
0.564 e A
minimum residual densities of 0.833 and ꢂ
/
,
1H NMR (CDCl3): 7.31Á
/
6.71 (m, 55H, PC6H5ꢁ
/
respectively.
C6H5Oꢁ
/
C6H5OH), 6.40 (s, 1H, OH), 13C NMR
The asymmetric unit contains one molecule of the
silver aryloxide complex, one molecule of o-cresol and
one disordered toluene molecule. The disorder was
extreme, the most satisfactory refinement being achieved
by placing the toluene electron density as two rigid
phenyl groups in the occupancy ratio of 52:48 for partial
2
(CDCl3): 160.5 (CO), 133.7 (d, Jꢀ
/15.4 Hz, PC6H5),
132.4 (d, 1Jꢀ
3Jꢀ
8.8 Hz, PC6H5), 128.1 (C6H5O), 116.8 (C6H5O),
116.6 (C6H5O). 31P NMR (ꢂ
25 8C, CDCl3/CD2Cl2):
5.08.
/17.6 Hz, PC6H5), 130.1 (PC6H5), 128.9 (d,
/
/
carbons C(69)Ã
/
C(74) and C(75)Ã
/
C(80), respectively.
2.2.5. (o-Cresolato)tris(triphenylphosphine)silver(I)×
/
o-
The partial methyl carbons of the toluene could not be
convincingly located. In the final least-squares cycle all
atoms were allowed to vibrate anisotropically except for
cresol×toluene solvate (5)
/
This compound was prepared by the same method as
2 using the same ratio of silver oxide (1), triphenylpho-
sphine (4) and o-cresol (2) in toluene. After removing
the insoluble 2 which precipitates from solution, crystals
of 5 are formed on standing. Anal. Found (Calc. for
C75H68AgO2P3): C, 74.9 (74.9); H, 5.65 (5.70)%. IR:
1581w, 1435m, 1244m [n(CO)]. 1H NMR (CDCl3):
the partial carbons C(69)Ã
/C(80) in the solvent region of
the electron density map. Phenyl rings were treated as
rigid hexagons throughout and hydrogen atoms were
included at calculated positions where relevant in the o-
cresolato and o-cresol moieties except for H2(A) of the
o-cresol which is involved in hydrogen bonding to the o-
cresolato oxygen atom O(1). This hydrogen was located
7.35Á/7.25
(m,
58H,
PC6H5ꢁ
/
MeC6H4Oꢁ
/

40
D. Edwards et al. / Inorganica Chimica Acta 349 (2003) 37Á44
/
Table 1
stretching vibrations (see Section 2). These are in a
similar region to the analogous modes of copper(I) and
gold(I) aryloxides [12,14] but provide no information
which can be related to co-ordination type.
˚
Selected bond lengths (A) and angles (8) for 5
˚
Bond lengths (A)
Ag(1)Ã
Ag(1)Ã
/
O(1)
2.386(5)
2.539(2)
2.565(9)
0.98(4)
Ag(1)Ã
Ag(1)Ã
/
P(1)
P(3)
2.577(2)
2.540(2)
1.59(4)
NMR spectra (1H, 13C and 31P) were recorded for all
five compounds (see Section 2), the proton and carbon
spectra confirming the stoichiometries but are otherwise
unexceptional. In the aryl proton region of the 1H NMR
spectra of complexes 2, 4 and 5, the triphenylphosphine,
aryloxide and aryl alcohol (and for 5 additionally the
toluene) resonances all extensively overlap. Conversely,
for complexes 1 and 3 the resonances of the triphenyl-
phosphine protons are clearly differentiated from those
of the aryl protons of the trichlorophenoxo or tris(di-
methylaminomethyl)phenoxo ligands. The ortho- and
para-dimethylaminomethyl protons in 3 are magneti-
cally inequivalent, both the methyl and methylene
protons displaying the expected 2:1 ratios. This is in
contrast with the situation found for dimeric [Ga-
Cl2{OC6H2(2,4,6-CH2NMe2)3}]2 [15] in which the
ortho- and para-dimethylaminomethyl protons are ap-
parently magnetically equivalent. In the 13C NMR
/P(2)
/
O(1)Ã
O(2)Ã
O(2)Ã
/
O(2)
O(1)Ã
/
H(2A)
C(55)
/H(2A)
O(1)Ã
/
1.297(8)
/C(62)
1.354(8)
Bond angles (8)
O(1)Ã
O(1)Ã
P(1)ÃAg(1)Ã
O(1)ÃH(2A)Ã
/
Ag(1)Ã
/
P(1)
98.36(14) O(1)Ã
/
Ag(1)Ã
Ag(1)Ã
Ag(1)Ã
C(55)ÃO(1)Ã
/
P(2)
P(2)
P(3)
105.2(2)
111.08(7)
116.41(7)
116.4(4)
/
Ag(1)Ã
/
P(3)
107.8(2)
115.73(6)
170(10)
P(1)Ã
/
/
/
/
P(3)
P(2)Ã
/
/
/
/O(2)
/
/
Ag(1)
in a difference Fourier map based on low Bragg data
˚
and refined at a distance of 0.98 A from O(2).
The asymmetric unit, omitting the disordered solvent
molecule, produced using ORTEX [13], is shown together
with the labelling scheme in Fig. 1. Selected bond
lengths and angles are given in Table 1.
spectra, all aryloxide CÃ
/
O resonances are found in the
3. Results and discussion
154Á160 ppm range, values similar to those of the free
/
aryl alcohols. Generally, the triphenylphosphine carbon
resonances are easily distinguishable from those of the
aryloxide or aryl alcohol because of the presence of
Five silver aryloxideÁtriphenylphosphine complexes
have been prepared from the reactions in toluene of
silver oxide and the appropriate aryl alcohol in the
/
carbonÁphosphorus couplings with characteristic JCP
/
presence of the phosphine. Yields were in the range 66Á
82%. The products were of two types, the bis-phosphine
complexes [Ag(OR)(PPh3)2] with Rꢀ2,4,6-Cl3C6H2 (1),
2-MeC6H4 (2) or 2,4,6-(Me2NCH2)3C6H2 (3), and the
tris-phosphine complexes which also contained an
/
coupling constants. The magnetic inequivalence of the
methyl and methylene carbons of 3 is again evident. The
13C NMR spectrum of 5 was difficult to assign convin-
cingly, resonances associated with the o-cresol, o-
cresolato and toluene moieties overlapping consider-
ably.
/
additional
[Ag(OR)(PPh3)3]×
molecule
of
the
aryl
Ph (4) or 2-MeC6H4
alcohol
/
ROH with Rꢀ
/
The 31P NMR spectra of the five complexes were
(5). Complex 5 was actually isolated as a toluene solvate
and its structure confirmed by a single crystal structure
determination (loc. cit.). Syntheses were carried out in
the absence of light as the products were found to be
somewhat light-sensitive turning grey or pale brown on
standing. Under normal laboratory conditions, at-
tempted recrystallisations of compounds 2, 3 and 4
from anhydrous toluene invariably produced silver films
on the surfaces of the glass equipment. All five
compounds also tended to decompose when heated
above 80 8C and reacted slowly with chlorinated sol-
recorded at room temperature and at ꢂ
CDCl3/CD2Cl2 mixed solvent. At room temperature, all
resonances were broad singlets in the range d 5.5Á9.5
with no indication of AgÁP coupling, suggesting triphe-
/
25 8C in a
/
/
nylphosphine lability in solution as noted in previous
similar NMR studies [9,16]. Consequently, in Section 2
we merely report d values at ꢂ
range d 4.51Á9.38 with Dd co-ordination shift values
(from free triphenylphosphine) of 9.99Á14.71. The
/
25 8C which are in the
/
/
closely similar Dd values reflect the nature of the
phosphine and are clearly little influenced by change
in the type of aryloxide present. The spectra recorded at
vents e.g. CHCl3 to form [AgCl(PPh3)n] (nꢀ2 or 3).
/
However, compound 5 could be recrystallised from
toluene.
ꢂ
/
25 8C again showed no AgÁP coupling, except for 3
/
which exhibited a rather broad doublet at d 9.38 with a
1JAgP of 345 Hz. Resolution of the individual 107Ag and
109Ag components was not achieved.
The FAB (LSIMS) mass spectra of compounds 1, 3, 4
and 5 have been recorded and metal-containing frag-
ments identified using the characteristic patterns im-
posed by the natural abundances of the 107 and 109
isotopes of silver. Only 5 showed a parent ion, albeit of
We have previously discussed [9] the value of IR
spectroscopy in providing evidence to support specific
co-ordination modes for carboxylate ligands in silver
carboxylates, perfluorocarboxylates and their bis- and
tris-phosphine complexes. Unfortunately, IR spectro-
scopy is of less significance for these silver aryloxide
complexes, although bands can be assigned to CÃ
/O

D. Edwards et al. / Inorganica Chimica Acta 349 (2003) 37Á
/44
41
Table 2
Thermal stabilities of silver aryloxideÁtriphenylphosphine complexes
/
Decomposition temperature (8C)
% Mass loss
Found a Total Calc. b
Start
Finish
[Ag(OC6H2Cl3-2,4,6)(PPh3)2] (1)
[Ag(OC6H4Me-2)(PPh3)2] (2)
195
330
130
55
320
380
310
270
315
180
290
140
300
62.4
17.5
75.1
63.1
11.1
17.2
68.2
7.1
77.9
75.1
87.0 c
85.4
[Ag{OC6H2(CH2NMe2)3-2,4,6}(PPh3)2] (3)
[Ag(OC6H5)(PPh3)3×C6H5OH] (4)
[Ag(OC6H4Me-2)(PPh3)3×2-MeC6H4OH]×C6H5CH3 (5)
290
110
180
110
150
74.2
85.4
81.5
88.0
90.0
91.0
/
/
/
74.4
a
b
c
From TGA plot.
Assuming Ag as final product.
82.7 for AgCl as residue.
low abundance. Fragments of the type [Ag(PPh3)n]ꢁ
(nꢀ1 or 2), which have also been identified in the mass
spectra of silver carboxylateÁtriphenylphosphine com-
plexes [9], are again detected in high abundance for all
four aryloxides tested, the fragment with nꢀ2 being the
most abundant metal-containing ion for 1, 4 and 5 but
nꢀ1 being the most abundant for 3. The high abun-
(OC6H4Me-4)(HOC6H4Me-4)] [19] which has O(1)Ã
/
˚
˚
/
O(2) of 2.62 A and HÁ
O(2)ÃH(2A)Ã
170(10)8]. The AgÃ
quite similar to those found in [Ag(OC6H2Cl3-2,4,5],
/
O(1) of 1.4(1) A. In 5, the
/
/
/
O(1) system is essentially linear [angle
˚
/
O bond length of 2.386(5) A in 5 is
/
˚
2.543(3) A] [4] and in [Ag(OC6H2Cl3-
˚
2,4,6)(PPh3)2] [2.255 A] [5] but shorter than
[2.317(3)Á
/
/
dance of these ions reflects the tendency of silver(I) to
form linear species in the gas phase [17]. Each com-
pound also shows [AgO(PPh3)2]ꢁ fragments, highly
abundant for 1 (63% relative to [Ag(PPh3)2]ꢁ) but
only of between 2 and 4% relative abundances for the
other aryloxide complexes. This fragment has previously
those
in
the
quinolinolato
cation
[7]
[Ag2(C9H6NO)(C9H6NOH)3(PPh3)2]ꢁ [2.529(4) and
˚
2.546(4) A]. The AgÃ
/
P bond distances of 2.539(2),
˚
2.540(2) and 2.577(2) A in 5 are longer than those of
˚
[Ag(OC6H2Cl3-2,4,6)(PPh3)2] [2.444(1) and 2.451(1) A]
[5], reflecting the presence of the additional third bulky
triphenylphosphine ligand. A comparison with, for
been observed in the mass spectra of [Ag(CÅ
/
CPh)(PPh3)] [18] and triphenylphosphine adducts of
silver b-ketoenolates and fluorocarboxylates [9]. No
fragment ions containing discrete aryloxide groups
were detected for any of the compounds. The spectrum
of 5 also contains low-abundant di-silver fragment ions
such as [Ag2(CO)(PPh3)2]ꢁ and [Ag2(OH)(PPh3)3]ꢁ
presumably formed by gas phase reactions in the
spectrometer chamber, given that 5 is mononuclear in
the solid state.
example, [Ag(FBF3)(PPh3)3] [20] which has AgÃ
/P
˚
bond distances of 2.506(3), 2.543(3) and 2.577(3) A
provides confirmation of this steric effect. The distortion
from regular tetrahedral geometry is evident on con-
sideration of the PÃ
1). The regular tetrahedral angle is exceeded for all three
PÃAgà angles, feature also found for
[Ag(FBF3)(PPh3)3] [20], whereas the PÃAgÃO angles
are all reduced from the regular tetrahedral angle. As
expected, the PÃAgÃP angles are smaller than that
/
AgÃ
/
P and PÃ
/
AgÃ
/
O angles (Table
/
/
P
a
/
/
The mononuclear structure of 5 (Fig. 1) has been
established by a single crystal X-ray structure determi-
nation. The silver atom is bonded to the oxygen of a
monodentate o-cresolato ligand and to the phosphorus
atoms of three triphenylphosphine ligands to generate a
distorted tetrahedral array. The neutral o-cresol mole-
cule in the structure is not bonded directly to silver but is
hydrogen-bonded via the hydroxyl proton H(2A) to
O(1) of the o-cresolato group. The proton H(2A) was
/
/
found [5] for [Ag(OC6H2Cl3-2,4,6)(PPh3)2] [131.1(1)8]
where the AgOP2 array is essentially distorted trigonal
planar.
Thermal analyses (Table 2) of all five aryloxide
complexes were carried out under inert helium atmos-
pheres to obtain information on decomposition tem-
peratures and sequences and on the relative stabilities of
the compounds. Such information gives an indication of
the viability of these compounds as useful silver film
precursors. Under our experimental conditions, two-
stage decompositions were evident for all complexes
except 2 although the temperature ranges varied widely
found during the solution of the structure, the O(1)Ã
/
O(2) separation and H(2A)ÃO(1) bond distances of
/
˚
2.565(9) and 1.59(4) A, respectively, being similar to
those of the p-cresolato complex [Rh(PMe3)3-

42
D. Edwards et al. / Inorganica Chimica Acta 349 (2003) 37Á44
/
Table 3
Conditions for AACVD testing and the nature of the films produced
[Ag(OC6H2Cl3-2,4,6)(PPh3)2] (1)
[Ag{OC6H2(CH2NMe2)3}(PPh3)2]
(3)
[Ag(OC6H4Me-2)(PPh3)3×
MeC6H4OH]×C6H5CH3 (5)
/
2-
/
Concentration range (g lꢂ1
)
50Á
300Á
0.80Á
17Á88
/
120
596
1.05
63Á
300
0.95Á
41Á106
thick, yellowÁ
/
65
13Á36
/
Reactor temperature (8C)
/
250Á398
0.50Á1.00
20Á104
brown, not reflective, very thin, yellowÁ
/
N2 flow rate (l minꢂ1
Run times (min)
)
/
/
1.00
/
/
/
/
Visual appearance of film a thin, pale grey, not reflective,
/
/
brown, not reflective,
soft, easily damaged
carbon
soft, easily damaged
carbon
soft, easily damaged
carbon
Detected impurities b
a
For 1 appearance relates to a film grown at a concentration of 120 g lꢂ1, a reactor temperature of 300 8C and a run time of 88 min; for 3
concentration 65 g lꢂ1; 300 8C; 106 min run time; for 5 concentration 65 g lꢂ1; 300 8C; 75 min run time.
EDXS analysis.
b
from one compound to another. Complex 2 merely
displayed an overall decomposition starting at 130 8C
and terminating at 310 8C. The mass loss of 75.1% is not
in accord with the final formation of simply silver metal
for which a mass loss of 85.4% would be expected. The
decomposition of complex 1 started at a higher tem-
perature than all the others with a large first stage mass
loss of 62.4% which corresponds to the loss of both
triphenylphosphine ligands (calc. mass loss 63.3%). The
overall mass loss of 79.9% again suggests that silver
metal is not the sole final product (calc. mass loss 87.0%)
but is closer to a calculated mass loss of 82.7% if the
final product is silver chloride. Indeed, qualitative tests
showed the presence of silver chloride in the final
residue, the chloride arising from decomposition of the
trichlorophenolato ligand and the presence of weak
intramolecular ClÁ Á ÁAg interactions inherent within the
structure. Decomposition of complex 3 commenced at
Fig. 1. The asymmetric unit of 5, showing the labelling scheme used in the text and Table 1. The disordered molecule of toluene solvent has been
omitted for clarity. Thermal ellipsoids are at the 30% probability level.

D. Edwards et al. / Inorganica Chimica Acta 349 (2003) 37Á
/44
43
the very low temperature of 55 8C with a large first stage
loss of the triphenylphosphine ligands and a smaller
second stage loss leading to a final total% mass loss of
74.2%. This is much less than expected if the final
residue is silver metal (calc. 88.0%). The decomposition
profiles of the two tris-triphenylphosphine complexes 4
and 5 differ from those of the bis-triphenylphosphine
complexes in that the first stage involves a small mass
loss followed by a much larger second stage mass loss.
The overall% mass loss for 4 of 85.4% does not
correspond with the final formation of solely silver
metal (calc. 90.0%). The first stage in the decomposition
of 5 appears to correlate with the% mass loss for the
toluene solvate (found 7.1%; calc. 7.7%) but again the
overall loss of 81.5% is too low to suggest the final
residue is merely silver (calc. 91.0%). These results,
therefore, do not support the view that AACVD of the
aryloxide complexes should lead to the formation of
good quality, pure silver films. Actual AACVD testing
confirmed that the aryloxide complexes were unsuitable.
Attempts were made to grow silver films using
AACVD under varied conditions but using a constant
dinitrogen atmosphere of one bar pressure and the same
horizontal cold wall reactor. Details of our experimental
protocol and equipment are given elsewhere [9] and
results from the use of silver carboxylate, fluorocarbox-
ylate, b-ketoenolate and b-ketoiminate complexes as
precursors have already been discussed. Solutions of
complexes 1, 3 and 5 in anhydrous tetrahydrofuran were
nebulised and the resulting aerosols passed over heated
glass substrates using dinitrogen as carrier gas. Each
precursor was tested using differing concentrations,
temperatures, dinitrogen carrier gas flow rates and run
times (see Table 3) but none of the precursors generated
good quality silver films as assessed by their visual
appearance, thickness, softness and ease of damage by
touch. In the light of the poor quality of the films, their
reflectivities and resistivities were not measured and
scanning electron micrographs were not obtained, un-
like films produced from other silver complexes [9].
However, the films were examined by EDXS which
confirmed the presence of silver with trace levels of
carbon, oxygen and silicon but the absence of phospho-
rus. The oxygen and silicon clearly arise from the
underlying glass substrates.
use as precursors for the growth of silver films by
AACVD. The results were disappointing, the silver films
obtained being vastly inferior to those grown from silver
carboxylates, fluorocarboxylates, b-ketoenolates and b-
ketoiminates reported earlier [9].
5. Supplementary material
Crystallographic data for structure have been depos-
ited with the Cambridge Crystallographic Data Centre,
CCDC No. 190513. 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 www:
http://www.ccdc.cam.ac.uk).
Acknowledgements
We thank the University of Bath for a research
studentship (to V.O.), the EPSRC mass spectrometry
service (University of Wales, Swansea) for provision of
mass spectra and Pilkington p.l.c. for their interest and
support.
References
[1] (a) D.C. Bradley, R.C. Mehrotra, D.P. Gaur, Metal Alkoxides,
Academic Press, New York, 1978;
(b) K.C. Malhotra, R.L. Martin, J. Organomet. Chem. 239 (1982)
159;
(c) R.C. Mehrotra, Adv. Inorg. Chem. Radiochem. 26 (1983) 269;
(d) D.C. Bradley, Chem. Rev. 89 (1989) 1317;
(e) R.C. Mehrotra, A. Singh, Progr. Inorg. Chem. 46 (1997) 239;
(f) H.E. Bryndza, W. Tam, Chem. Rev. 88 (1988) 1163;
(g) W.A. Hermann, N.W. Huber, O. Runte, Angew. Chem., Int.
Ed. Engl. 34 (1995) 2187.
[2] R.S. Macomber, J.C. Ford, J.H. Wenzel, Syn. React. Inorg.
Metal-Org. Chem. 7 (1977) 111.
[3] D. Sevdevic, L. Fekete, L. Meider, J. Inorg. Nucl. Chem. 42
(1980) 885.
[4] G. Smith, E.J. O’Reilly, B.J. Reynolds, C.H.L. Kennard, T.C.W.
Mak, J. Organomet. Chem. 331 (1987) 275.
[5] G. Wulfsburg, D. Jackson, W. Ilsley, S. Dou, A. Weiss, J.
Gagliardi, Z. Naturforsch., Teil A 47 (1992) 75.
[6] N. Galesic, M. Herceg, D. Sevdevic, Acta Crystallogr., C 44
(1988) 1405.
[7] A.H. Othman, S.C. Goh, H.K. Fun, K. Sivakumar, Acta
Crystallogr., C 52 (1996) 2760.
4. Conclusions
[8] D.A. Edwards, R.M. Harker, M.F. Mahon, K.C. Molloy, J.
Chem. Soc., Dalton Trans. (1997) 3509.
Some new silver(I) aryloxide complexes containing
two or three triphenylphosphine ligands have been
prepared and characterised. Very few related complexes
have been reported previously. The single-crystal X-ray
structure analysis of an o-cresolato complex shows it to
be mononuclear with a monodentate aryloxide ligand.
Three of the complexes have been tested for potential
[9] (a) D.A. Edwards, R.M. Harker, M.F. Mahon, K.C. Molloy, J.
Mater. Chem. 9 (1999) 1771;
(b) D.A. Edwards, R.M. Harker, M.F. Mahon, K.C. Molloy,
Inorg. Chim. Acta 328 (2002) 134.
[10] (a) P.M. Jefffries, G.S. Girolami, Chem. Mater. 1 (1989) 8;
(b) P.M. Jeffries, L.H. Dubois, G.S. Girolami, Chem. Mater. 4
(1992) 1169.

44
D. Edwards et al. / Inorganica Chimica Acta 349 (2003) 37Á44
/
[11] M.J. Hampden-Smith, T.T. Kodas, M.F. Paffett, J.D. Farr, H.K.
Shin, Chem. Mater. 2 (1990) 636.
[16] E.L. Muetterties, C.W. Alegranti, J. Am. Chem. Soc. 92 (1970)
4114.
[12] M. Kubota, A. Yamamoto, Bull. Chem. Soc. Jpn. 51 (1978) 2909.
[13] P. McArdle, J. Appl. Crystallogr. 27 (1994) 435.
[14] T. Sone, M. Iwata, N. Kasuga, S. Komiya, Chem. Lett. (1991)
1949.
[17] L.D. Detter, R.G. Cooks, R.A. Walton, Inorg. Chim. Acta 115
(1986) 55.
[18] M.I. Bruce, M.J. Liddell, J. Organomet. Chem. 427 (1992) 263.
[19] S.E. Kegley, C.J. Schaverien, J.H. Freudenberger, R.G. Bergman,
J. Am. Chem. Soc. 109 (1987) 6563.
[15] D.A. Atwood, A.H. Cowley, R.D. Schluter, M.R. Bond, C.J.
Carrano, Inorg. Chem. 34 (1995) 2186.
[20] M. Camilli, F. Caruso, Inorg. Chim. Acta 127 (1987) 209.