Reaction of metal alkoxides with 3-alkyl-substituted acetylacetone derivatives - Coordination vs. hydrodeacylation

Article abstract of DOI:10.1039/b400364k

Reaction of Ti(OⁱPr)₄ or Zr(OPr)₄ with 1 or 2 molar equiv of the 3-alkyl-substituted acetylacetone derivatives 3-acetyl-6-trimethoxysilylhexane-2-one or 3-acetylpentane-2-one not only gives the corresponding β-diketonate complexes but also results in about 15% hydrodeacylation of the β-diketone. With the stronger Lewis acid Al(O ^sBu)₃ hydrodeacylation prevails. Hydrodeacylation is suppressed when a 1:5 ratio of metal alkoxide and β-diketone is reacted.

Full text of DOI:10.1039/b400364k

View Online / Journal Homepage / Table of Contents for this issue
P A P E R
Reaction of metal alkoxides with 3-alkyl-substituted acetylacetone
derivatives—coordination vs. hydrodeacylation
Michael Puchberger, Wolfgang Rupp, Ulrike Bauer and Ulrich Schubert*
Institute of Materials Chemistry, Vienna University of Technology, Getreidemarkt 9/165,
A-1060 Wien, Austria. E-mail: uschuber@mail.zserv.tuwien.ac.at
Received (in Montpellier, France) 9th January 2004, Accepted 7th April 2004
First published as an Advance Article on the web 7th October 2004
i
Reaction of Ti(O Pr) or Zr(OPr) with 1 or 2 molar equiv of the 3-alkyl-substituted acetylacetone
4
4
derivatives 3-acetyl-6-trimethoxysilylhexane-2-one or 3-acetylpentane-2-one not only gives the corresponding
b-diketonate complexes but also results in about 15% hydrodeacylation of the b-diketone. With the
s
stronger Lewis acid Al(O Bu)
3
hydrodeacylation prevails. Hydrodeacylation is suppressed when a 1 : 5 ratio
of metal alkoxide and b-diketone is reacted.
Introduction
at the carbonyl carbon of b-diketones or b-keto esters, as in
allyl acetoacetate or 2-(methacryloyloxy)ethyl acetoacetate, or
at the carbon between the keto groups (3-R-acac derivatives) as
in 3-allyl-2,4-pentanedione, a number of bromo-substituted
7
3-alkylpentanedione derivatives and 3-acetyl-6-trimethoxy-
Acetylacetone is an often used additive in sol-gel processes to
lower the reactivity of metal alkoxides, M(OR) . Upon reac-
x
tion with acetylacetone, part of the alkoxide groups is sub-
stituted by acetyl acetonate ligands and a new molecular
5
,6
precursor M(OR)
tained with a different structure and a lower reactivity towards
water. For example, reaction of Ti(OR) with acetylacetone,
4
(acac)y (acac ¼ acetyl acetonate) is ob-
silylhexane-2-one (1; Scheme 1). Compound 1 was reacted with
titanium alkoxides to prepare the new mixed-metal precursors
x ꢀ y
8
2a,b (a monomeric formula is drawn for 2a in Scheme 1 for
simplicity, although the compound should be dimeric—see
one of the best investigated chemical modification of metal
1
alkoxides, provides dimeric [Ti(OR) (acac)] when a 1 : 1
3
above), analogous to the Ti(OR)
tained with unsubstituted acetylacetone. The purpose of pre-
(acac)x derivatives ob-
2
4 ꢀ x
molar ratio is reacted, and monomeric cis-Ti(OR)
2
(acac)
2
for
2
i
paring the derivatives 2 was to link a Si(OMe) and a Ti(O Pr)
a 1 : 2 ratio. When the acac derivatives are employed as
precursors for sol-gel processing, the OR groups are preferen-
tially hydrolyzed. The same applies to other protic compounds
3
x
group by means of the organic spacer and thus to prevent
phase separation upon sol-gel processing. Homogeneous
gels were indeed obtained, from which anatase and rutile
nanocrystals formed upon calcination at unusually high
3
with bi- or multidentate anions.
Besides controlling the reactivity of the metal alkoxides, the
bidentate ligands can also be used to introduce functional
8
temperatures.
The derivatization of Ti(OR) and other metal alkoxides by
4
organic groups into gels. Carboxylates are mainly used for
4
this purpose, while b-diketones with functional organic groups
are less often employed. The functional group can be attached
acetyl acetonate ligands is well-investigated in sol-gel chemis-
try. Neither complications nor by-products were reported in
Scheme 1
T h i s j o u r n a l i s & T h e R o y a l S o c i e t y o f C h e m i s t r y a n d t h e
C e n t r e N a t i o n a l d e l a R e c h e r c h e S c i e n t i f i q u e 2 0 0 4
N e w J . C h e m . , 2 0 0 4 , 2 8 , 1 2 8 9 – 1 2 9 4
1289

View Online
Table 1 H and C NMR data of 3-acetyl-6-trimethoxysilylhexane-2-one (1) and the O-alkylation by-product 4, and their assignment. The ratio
1
13
1
a : 2a : 4 was approximately 13 : 5 : 1 [by integration of the dCH
2
(1a, 1b) and fCH
2
(4) signals, respectively].
1
13
C d
a
H d
Assignment
0.67
1.35
1.85
2.16
3.64
3.55
–
10.0
20.9
31.4
29.0
68.2
50.5
204.3
m, f
m, e
3
q, d, J ¼ 7.38 Hz
s, a
3
t, c, J ¼ 6.87 Hz
g
b
0
1
2
2
–
3
–
.65
.48
.22
.11
9.1
23.7
m, f
m, e
3
30.6
29.0
t, d, J ¼ 4.1 Hz
s, a
c
110.3
50.5
191.1
.55
g
b
0.70
1.81
2.11
2.26
3.78
3.55
5.48
–
–
11.0
22.1
31.6
19.4
70.0
50.5
99.4
171.7
196.7
m, i
m, h
s, a
s, e
3
t, f, J ¼ 3.9 Hz
g
m, c
d
b
a
The bold letters refer to the labels in the structural formulas in the left column.
1
Table 2 H, C and Si NMR data of the reaction mixture of 1 and Ti(O Pr)
13
29
i
4
, and their assignment
29
1
13
C d
a
Si d
H d
Assignment
ꢀ45.5
0.69
9.9
m, f
m, e
s, a
1.53
2.03
2.32
3.55
–
–
–
24.1
20.3
3
32.3
50.5
t, d, J ¼ 4.1 Hz
g
c
113.1
185.5
190.8
b
h
ꢀ47.4
0.60
10.0
22.3
27.3
m, f
m, e
m, d
s, a
1.40
1.55
2.11
2.42
3.55
–
n. a.
3
43.2
50.5
t, c, J ¼ 3.6 Hz
g
b
208.5
—
—
1.35
21.5
21.1
m, d
s, a
1
4
–
.99
.95
3
67.3
170.1
q, c, J ¼ 3.2 Hz
b
Isopropoxy groups
1.15
25.2
77.5
25.3
63.9
21.5
62.7
21.5
65.1
CH
broad, CH
CH
quint, J ¼ 3.0 Hz
CH
broad, CH
CH
broad, CH
3
4.72
1.20
3.98
1.23
4.15
1.23
4.22
3
3
3
3
a
The bold letters refer to the labels in the structural formulas in the left column.
1
290
N e w J . C h e m . , 2 0 0 4 , 2 8 , 1 2 8 9 – 1 2 9 4

View Online
with 2 equiv of unsubstituted
was performed for comparison.
i
The reaction of Ti(O Pr)
acetylacetone (acac-H) in CDCl
4
3
1
3
Whereas the C NMR spectrum of acetylacetone is character-
ized by two CO signals at 201.8 ppm for the keto form and
1
91.1 ppm for the enol form, the C NMR spectrum of cis-
3
1
Ti(O Pr) (acac) showed two signals at 187.1 and 191.2 ppm,
i
2
2
1
0
due to the stereochemically inequivalent carbon atoms. In the
1
i
H NMR spectrum of Ti(O Pr) (acac) , a singlet at 5.49 ppm
2
2
was observed for the CH group and two singlets at 1.92 and
.02 ppm for the two methyl groups. There was no indication
for unreacted acetylacetone.
The analogous reaction of Ti(O Pr) with 2 equiv of 1 in
2
i
4
3
CDCl led to a rather complicated mixture, whose NMR data
1
are given in Table 2. The H spectrum is shown in Fig. 1. Seven
different carbonyl signals were observed in the carbonyl region
1
3
of the C NMR spectrum (Fig. 2). Two of them (at 171.7 and
196.7 ppm) were assigned to the O-silylated by-product 4 (see
Table 1), two (at 185.5 and 190.8 ppm) to a diketonate complex
1
i
Fig. 1 H spectra of 1 (top) and the reaction mixture of Ti(O Pr)
equiv of 1 (bottom) in CDCl
4
with
2
3
.
(2) and one (204.3 ppm) to an uncoordinated b-diketone,
possibly unreacted 1a (resonances of the enol form 1b were
not identified). There were two additional carbonyl signals at
1
the literature. A careful analysis of the NMR and IR spectra of
i
the reaction mixtures of 1 with Ti(O Pr)
70.1 and 208.5 ppm, indicating an ester and a ketone. From
the NMR spectra it was not possible to decide whether 2a or 2b
or both were formed.
4
led us to investigate
the reaction of 1, and 3-acetylpentane-2-one (3; Scheme 1) as a
s
i
4 4 3
simpler model, with Ti(O Pr) , Zr(OPr) and Al(O Bu) in
The appearance of an ester and a ketone can be explained by
a C–C cleavage reaction (hydrodeacylation) of 2 leading to
6-trimethoxysilyl-2-hexanone (5) and isopropyl acetate (6). The
ester 6 was clearly identified by long-range correlation signals
in the HMBC spectrum (Fig. 2) of both the isopropyl CH at
more detail. We report in this article that the reaction of 1
and 3 is less straightforward than the reactions of unsubsti-
tuted acetylacetone, because hydrodeacylation of the 3-sub-
stituted b-diketones is competing with the formation of
b-diketonate complexes.
4
sponding C carbonyl signal at 170.1 ppm. The HMBC
3
.95 ppm and the CH resonance (1.99 ppm) to the corre-
1
3
Results and discussion
technique is a sensitive method for the determination of
1
long-range (two- and three-bond) connectivity between H
and heteronuclei. The ketone 5 was also clearly identified by
the corresponding correlation signals in the TOCSY and
3-Acetyl-6-trimethoxysilylhexane-2-one (1) was prepared from
iodopropyltrimethoxysilane and acetylacetone by a modifica-
9
tion of the previously published procedure. The product
1
HMBC spectra. The H resonances of the butyl chain were
contained the O-silylated compound (4) as an impurity, which
it was not possible to separate from 1. All proton and carbon
signals of the NMR spectra were assigned (Table 1) and served
as a reference for the following discussion. The ratio of the keto
form (1a), the enol form (1b) and the O-silylated compound (4)
was 13 : 5 : 1. This mixture was further used for the following
NMR investigations.
at 0.60, 1.40, 1.55 and 2.42 ppm and both the CH
3
resonance at
2
range correlation signals to the carbonyl carbon at 208.6 ppm.
.11 ppm and the CH resonance at 2.42 ppm showed long-
2
i
Moreover, an OMe/O Pr exchange at the silicon atom was
i
observed, the isopropoxy groups originating from Ti(O Pr) .
4
The chemical shifts of some isopropoxy groups were indicative
of a bridging coordination. The approximate ratio between all
reaction products [eqn. (1)] was 5 : 6 : 1a : 2 ¼ 5 : 5 : 5 : 20 (with 4
still being present), which means that only two thirds of the
employed b-diketone 1 was coordinated to titanium.
i
The reaction of Ti(O Pr) with 1 molar equiv of 1 yielded
4
the same products as in the reaction with 2 equiv. The spectra
differed only in the relative concentrations of the products.
Although the initial experiments were carried out with the
alkoxysilyl-substituted derivative 1, we assumed that the ob-
served cleavage reaction is a consequence of the 3-alkyl sub-
i
stituent. We therefore also reacted Ti(O Pr) with 2 equiv of
4
3-acetylpentane-2-one (3; which contained approximately 5%
of the O-alkylated compound 7 as a by-product). The reaction
mixture was again investigated by NMR spectroscopy
(
Table 3). The same kinds of products were formed as in the
reaction of 1, that is, the b-diketonate complex(es) 8, isopropyl
acetate (6) and 2-pentanone (9).
The hydrodeacylation products 5 and 6 were also observed
when CD Cl , C D or CD OD were used as the solvent.
2
2
6
6
3
Reaction of Zr(OPr)
resulted in the same NMR spectra as that of the reaction with
4
with 2 molar equiv of 1 in CDCl
3
i
s
4 3
Ti(O Pr) . When Al(O Bu) was reacted with 1 or 2 equiv of 1,
the hydrodeacylation reaction was nearly quantitative.
Contrary to these findings, no hydrodeacylation, that is, only
formation of the corresponding b-diketonate complexes, was
i
observed in the NMR spectra when Ti(O Pr)
Al(O Bu) was reacted with 5 molar equiv of 1.
3
4
, Zr(OPr)
4
or
s
i
Fig. 2 Section of the HMBC spectrum of Ti(O Pr)
4
with 2 equiv of
1
3
in CDCl (carbonyl region).
N e w J . C h e m . , 2 0 0 4 , 2 8 , 1 2 8 9 – 1 2 9 4
1291

View Online
1
Table 3 H and C NMR data of the reaction mixture of 3 and Ti(O Pr)
13
i
4
, and their assignment
1
13
C d
a
H d
Assignment
3
0.85
1.85
2.12
3.50
–
11.9
21.5
t, e, J ¼ 7.46 Hz
3
t, d, J ¼ 7.37 Hz
29.0
70.3
s, a
3
t, c, J ¼ 7.20 Hz
204.5
b 12
3
0.99
2.09
2.22
–
–
14.7
22.5
t, e, J ¼ 7.56 Hz
s, a
m, d
c
20.3
111.8
190.8
b
1.30
2.14
2.27
3.80
–
–
15.0
22.2
t, g
s, a
s, e
q, f
d
22.2
63.7
170.6
197.1
b
3
1.03
1.99
2.31
–
–
–
15.3
25.3
t, e, J ¼ 7.42 Hz
s, a
m, d
c
22.3
114.2
185.2
190.5
b
b
0.89
1.53
2.13
2.38
–
13.6
24.8
t, e
3
sept, d, J ¼ 7.38 Hz
24.3
45.6
s, a
3
t, c, J ¼ 7.23 Hz
209.3
b
1.22
2.00
4.98
–
21.9
21.3
mult, d
s, a
3
67.6
170.5
quint, c, J ¼ 6.18 Hz
b
Isopropoxy groups
1.18
25.4
26.6
64.2
76.1
77.1
67.6
mult
mult
1.23
4.00
4.46
4.73
4.96
s (br), m
3
quint, m, J ¼ 5.85 Hz
3
sept, m, J ¼ 6.15 Hz
3
quint, m, J ¼ 6.18 Hz
a
The bold letters refer to the labels in the structural formulas in the left column.
1
292
N e w J . C h e m . , 2 0 0 4 , 2 8 , 1 2 8 9 – 1 2 9 4

View Online
Scheme 2
Conclusions
Experimental
While the reaction of 1 or 2 molar equiv of unsubstituted
acetylacetone with titanium, zirconium and aluminium alkox-
ides clearly gives acetyl acetonate substituted metal alkoxides,
reaction of the 3-R-acac derivatives 1 and 3 was distinctly
different in a two respects.
General
All reactions were carried out using standard Schlenk techni-
ques in a moisture- and oxygen-free argon atmosphere. Sol-
vents were made water- and oxygen-free by standard methods.
NMR spectra were recorded on a Bruker Avance 300, 300.13
First, 6-(trimethoxysilyl)-2-hexanone (5) and isopropyl acet-
ate (6) were identified in the reaction of 1 with 1–2 molar equiv
1
13
29
MHz { H}, 75.47 MHz { C} and 59.63 MHz { Si}, equipped
with a 5 mm broadband probe head and a z-gradient unit. The
solvents were CDCl (99.8%, euriso-top) and CD Cl (99.6%,
i
of Ti(O Pr)
4
, in addition to the expected substitution product 2
3
2
2
[
result from a cleavage of 3-R-acac-H. The reaction with 2
analogously for Zr(OPr) ]. The compounds 5 and 6 clearly
4
euriso-top). 2D spectra were measured with Bruker standard
pulse sequences: COSY (correlated spectroscopy), TOCSY (to-
tal correlation spectroscopy, mixing time was usually 160 ms),
s
molar equiv of Al(O Bu) only results in the cleavage of 1 and
3
only a minor proportion of a b-diketonate complex was
observed. The cleavage reaction probably proceeds as indi-
cated in Scheme 2. The dependence of the cleaved proportion
of 1 on the kind of metal alkoxide (Al c Ti E Zr) may partially
be due to the different Lewis acidity, but also to the higher
covalency of the Al–O interaction and the shorter distance for
the transfer of the OR group due to the smaller bonding radius
of aluminium. The 3-alkyl substitution probably facilitates the
cleavage reaction in a two-fold manner: by making the oxygen
more nucleophilic and the alkoxide/b-diketonate substitution
less facile. Therefore hydrodeacylation is not observed with
1
13
HSQC (heteronuclear single quantum correlation), H/ C- and
1
29
H/ Si-HMBC (heteronuclear multiple bond correlation).
Syntheses and reactions
Synthesis of 3-acetyl-6-trimethoxysilylhexane-2-one (1). An
amount of 15.0 g (0.1 mol) of NaI was dissolved in 60 ml of
acetone. Chloropropyltrimethoxysilane (19.9 g, 0.1 mol) was
added and the mixture was refluxed overnight. An amount of
13.8 g (0.1 mol) of K CO and 10.0 g (0.1 mol) of acetylacetone
was added and the mixture was again refluxed overnight. The
reaction mixture was then filtered and the solvent removed
in vacuo. The remaining liquid was distilled in vacuum (b.p.
90 1C/1 mbar). Yield: 12.1 g (46.1%). The by-product 4 could
2
3
unsubstituted acetylacetone under the same conditions.
i
In the case of Ti(O Pr)
4
and Zr(OPr)
proportion is larger than the cleaved proportion. This explains
4
, the coordinated
ꢀ1
why the use of 1 nevertheless has a pronounced influence in
8
getting homogeneous silica/titania mixed oxide gels.
not be removed by distillation. IR (cm ): 2943 (m, C–H), 2943
(m, C–H), 2840 (m, Si–OCH ), 1700 (m, C O), 1585 (w, C–C
3
Q
and C–O), 1456 (w, C–H), 1358 (w, C–H), 1291 (w, CH ), 1190
This is another example where a metal alkoxide reacts with a
potentially coordinating organic compound both by formation
of a metal complex and as a Lewis acidic catalyst in the
transformation of the organic compound. The other example
for such a dichotomy that we recently found was formation of
2
(m, Si–OCH ), 1077 (s, Si–OCH ), 995 (w), 955 (w), 896 (w),
3
3
2
807 (m), 758 (m, Si–C), 679 (w). Si NMR ( Si H HMBC
9
29
1
1
13
spectrum): ꢀ45.8 ppm. H and C NMR data: see Table 1.
a lysinate complex with Zr(OR)
1
formation.
4
vs. catalysis of lactame
Reaction of 3-acetyl-6-trimethoxysilylhexane-2-one (1) with
i
1
i
Ti(O Pr) . Ti(O Pr) was mixed with 2 molar equiv of 1 under
4 4
The second difference is that with 2 molar equiv of 3-R-acac-
H the reactions do not go to completion and small but
significant amounts of unreacted 3-R-acac-H were spectro-
scopically observed in the reaction mixtures. The substitution
of an alkoxide ligand of the alkoxide for a b-diketonate ligand
is therefore less straightforward than in the case of unsubsti-
tuted acetylacetone. This is caused by the lower acidity of
the proton in the 3-position due to the electron-donating
alkyl group in this position. Incomplete substitution was
previously observed in the reaction of 3-allyl-2,4-pentanedione
argon and stirred for 15 min. A portion of 20 mg of the
resulting mixture was diluted with 0.6 ml of CDCl and used
3
i
for the NMR experiments. The reaction of Ti(O Pr) with 1
4
molar equiv of 1 yielded the same products as in the reaction
with 2 equiv. Because of the higher quality of the spectra of the
reaction with 2 equiv of 1, they were used for the structure
ꢀ
1
assignments. IR (cm ): 2969 (m, C–H), 2840 (w, Si–OCH3),
1701 (w, C O), 1592 (m, C–C and C–O), 1460 (m, C–H), 1358
Q
(m CH), 1290 (w, CH ), 1122 (m, Ti–O Pr), 1086 (s, Si–OCH ),
i
2
3
i
i
i
1040 (s, O Pr), 992 (m, CO of O Pr), 884 (w, Ti–O Pr), 848 (w),
808 (m). The NMR data are given in Table 2.
s
6
3 4 4
with Al(O Bu) , Ti(OBu) and Zr(OBu) . When a larger
proportion of 3-R-acac-H is reacted, the equilibrium is shifted
to the side of the b-diketonate complex and the Lewis acid-
Lewis base interaction as in Scheme 2 is prevented. This
explains why no cleavage is observed with a higher proportion
of R-acac-H.
i
Reaction of 3-acetylpentane-2-one (3) with Ti(O Pr) . The
reactions with 1 or 2 molar equiv of 3 were performed as
described for the reaction with 1. The NMR data are given
in Table 3.
4
The results reported in this paper show that care must
be taken when 3-substituted acetylacetone derivatives are
used for the modification of metal alkoxides. The problems
arising from incomplete reaction and dehydroacylation clearly
originate from the presence of an alkyl substituent in the
Acknowledgements
¨
We thank the Fonds zur Forderung der wissenschaftlichen
Forschung (FWF), Vienna, for its support of this work and
Prof. Johannes Frohlich for valuable discussions.
¨
3
-position.
N e w J . C h e m . , 2 0 0 4 , 2 8 , 1 2 8 9 – 1 2 9 4
1293

View Online
J. Zhang, S. Luo and L. Gui, J. Mater. Sci., 1997, 32, 1469.
D. Hoebbel, T. Reinert, H. Schmidt and E. Arpac, J. Sol-Gel Sci.
Technol., 1997, 10, 115.
D. Holzinger and G. Kickelbick, Chem. Mater., 2003, 15, 4944.
References
5
6
1
R. J. Errington, J. Ridland, W. Clegg, R. A. Coxall and J. M.
Sherwood, Polyhedron, 1998, 17, 659.
X-Ray structure analyses of Ti(OR)
7
8
2
2
(b-diketonate)
2
derivatives:
W. Rupp, N. Hu
2594.
W. Urbaniak and U. Schubert, Liebigs Ann. Chem., 1991, 1221.
¨
sing and U. Schubert, J. Mater. Chem., 2002, 12,
J. L. Wang, F. M. Miao, X. J. Fan, X. Feng and T. J. Wang, Acta
Crystallogr., Sect. C, 1990, 46, 1633; U. Schubert, H. Buhler and
B. Hirle, Chem. Ber., 1992, 125, 999.
9
10 Two signals were also observed in the NMR spectrum of [Ti(O
Pr) (acac)] : A. Leaustic, F. Babonneau and J. Livage, Chem.
i
3
4
J. Livage, M. Henry and C. Sanchez, Progr. Solid State Chem.,
1988, 18, 259; C. Sanchez and J. Livage, New J. Chem., 1990
14, 513.
U. Schubert, J. Sol-Gel Sci. Technol., 2003, 26, 47.
3
2
Mater., 1989, 1, 240.
11 O. Metelkina and U. Schubert, Monatsh. Chem., 2003, 134,
1065.
1
294
N e w J . C h e m . , 2 0 0 4 , 2 8 , 1 2 8 9 – 1 2 9 4