Silicate cages: precursors to new materials

ELSEVIER

Journal of Organometallic Chemistry 542 (1997) 141 - 183

Review

Silicate cages: precursors to new materials P h i l i p G. H a r r i s o n Del~u'tment of Chemisn3; Unicersit3."of Nottingham. Unicersit3."Park. Nottingham. NG7 2RD. UK Received 26 September 1996

I. Background and the condensation process Second only to carbon, silicon forms the largest number of bonds with other elements. Unlike carbon, however, where C-C, C-O, and C-H bond energies are approximately equal, the Si-O bond is considerably stronger than the Si-H bond and much stronger than the St-St bond. Therefore, chains of S i - O ~ S i - O - S i make up the skeletons of silicate chemistry. In silicates all silicon atoms exist in tetrahedrai coordination by tour oxygen atoms, and these ate the molecular units from which naturally occurring silicates ate constructed. Silicates can occur as rings, connected in chains and layers, or in the form of cages [i]. The structure of cage silicates is based on S i l O linkages tbrming a cage with a silicon atom at each vertex. Substituents coordinate around the silicon vertices tetrahedndly. The nature of tile exo ca~e substituent in a compound can determine many physical properties. The number of {XSiO~} ullits sets the shape of the f!'ame, which is uniquely Uno strained lbr eight and ten units. Tile silicate cage como pounds (RSiOt ~)0,, (R ~ organic or inorganic group), n ~ 6 (I), n ~ 8 (11), n ~ 10 (!11) or n ~ 12 (IV) (Fig. I), represent a rather versatile class of potential threedimensf, onal building block units for the synthesis of new materials, and therelbre they are of considerable theoretical and practical interest. Silicate cages of types !! and IV are present as subunits in zeolite A ill) and zeolites type X and Y (IV) respectively (Fig. 2). A cage structure involves several rings connected together in a finite three-dimensional molecular skeleton. These compounds are known as polyhedral silsesquioxanes [2,3]. The nomenclature most commonly used is given by a trivial systemalit; system. Sil-ses-quioxane, denotes that each silicon atom is connected to three oxygen atoms, {SiO3}. The prefix 'oligo' is often used to indicate a small number of siisesquioxane links. Otherwise, prefixes such as 'octa-, penta-, deca-, etc.' are used to indicate a specific ,lumber of these links, e.g. (HSiO~.~)H [octasilsesquioxane or 0022-328X/97/$17.00 © 1997 Elsevier Science S.A. All rights reserved. Pil S 0 0 2 2 - 3 2 8 X ( 9 6 ) 0 6 8 2 1 - 0

octa(hydrosilsesquioxane)] and (CH3)7C6H~(SiOIs) a [heptamethylphenyloctasilsesquioxane]. The general formula of p:,lyhedral silsesquioxanes is (XSiO~ 5), where n >_4, and X = H, organyl, halogen, etc., which are called hydrosilsesquioxanes (also known ~,~ hydridospherosiloxanes), organylsilsesquioxanes, and halosiisesquioxanes, etc. respectively. Oligosilsesquioxanes are generally colourless crystalline substances. With growing chain length of the alkyl group in octa(alkylsilsesquioxanes) the melting points and densities decrease, whereas volatility and solubility in organic solvents increase. Volatility of these compounds is extremely high. For the same substituent the octamers have the highest melting point and are the least volatile. Polyhedral silsesquioxanes are potentially a very use° ful class of compounds. Their chemistry has, however, remained underdeveloped for a long time I~eause of a hick ot' facile methods for their synthesis in useful quantities. It has taken a long time to develop suitable syntheses. The most common process now used to obtain polyhedral silicon~oxygen skeletons of oligosilsesquioxanes with R ~ organic group is by the hydrolytic condensation of trifunctional monomers XSiY3. where X is a chemically stable organic sub° stituent and Y is a highly reactive substituent, such as Ci or alkoxy, and is strongly dependent on several factors including: I. concentration of initial monomer in the solution; 2. nature of the solvent; 3. character of substituent X in the initial monomer; 4. nature of fimctional group Y in the initial monomer; 5. type of catalyst employed; 6. temperature; 7. rate of addition of water and quantity of water added; 8. solubility of the polyhedral oligomers tbrmed. Tile polycondensation process is very difficult to understand because of strong mutual effects of the above-mentioned factors. However, by an examination

142

P.G. Hurt@m1/ Journal ~' Org,lnomet~llih"Chcmistr3.'542 (1997) 141- i~3

{ /

7,\ •-

III

j,

IV

FiB. I. Repte~elltalion~ of [R~iO~/~], c~ge .~tril¢lU~ where n ~ (~ (!). n ~ ~ (ii). n ~~, IO (!11). ;rod n ~ 12 (IV). (The.~e are idealized repre~fll~lion~; the $i0 ~l~lBle~ in the re~l ~lrU¢lUre,~ t~re very ¢lo~e io letr~,hedr~fl ~mgle~, The Si-O-Si ~m~le is ol~ned to c~,, 15(}°),

of each individual factor ilself, qualitative trends can be deduced. Increase in tcm~r'ature results in the fornlao tion of high!y condensed polymers, Hence the tempcrao lure should be kept low, at r ~ m teml~rature or prefer° ably subambient. Slow, cnrcful addition of water will keep the concentration of silanol groups tbrmed low. On the other hand, the choice of a suitable ~olvent such as an alcohol or ether can stabilize silanol groups by hydrosen-bondin8 of the St=OH to the alcohol or ether function as well as to the remaining silalkoxy groups. The nature of the functional group Y in the initial monomer can also have an effect. For Y = Ci the hydrolysis is very rapid compared to when Y ~ OEt, OMe or CHACO0, and the reaction becomes autocat. alytic with the formation of HCI. The character of the substituent X in the initial monomer plays a role due to

.LI$chenle I, Pml~),~cd m¢chamsm for Ih¢ hydr~,lylk" condens~klion of FiB, 2, The strt~tu~ el" ~ ) l i t e A (~) ~nd zeolile.~ X and Y (b),

RSi(OR)~.

P.G. Harrison/.hmrmd ¢!1"Or.eumm~('l,dlic Chemistry 542 ¢1997J 141-

I&¢

!43

Scheme 2.

Scheme4.

both steric as well as + l-inductive effects. For example, in the aikyichlorosilane series, the hydrolysis rate is known to slow down with increasing hydrocarbon substituent length [4]. When X = CH=CH., drr-prr effects have consequences for the stability of the five-coordinated silicon intennediate. Therefore, in most cases, the mechanism of polycondensation of trifunctional monomers XSiYa bearing a different substituent X is also different. Also. it is not surprising that the size of the silicon atom is of great importance, and for the heavier Group 14 dements germanium and tin such cubic sesquioxane structures typical lbr silicon are not known. A close look at the condensation and hydrolysis equilibria involved in the dimerization of silanols (Scheme I) shows how complex the mechanisms are. and will give an initial idea for a qualitative interpretation of the above-mentioned factors. As seen in Scheme !. a five-coordinated silicon c(')mplcx' is proposed to be involved as an intermediate. The formation of oligosilsesquioxanes in the course of hydrolytic polycondensation of XSiY~ monomers in dilute solvents can be represented by the overall equalion

sation (Schemes 4 and 5). Interestingly. the polycondensation of t BuSi(OH)3 has been reported to give a unique example of the fully condensed, tetranuclear silsesquioxane, [tBuSiO]4, which could be isolated in 94% yield [8]. In this case, the compound which results from the first step, [' BuSi(OH)-, ]20, comprises hydrogen-bonded sheets in the solid state [9]. The mechanism by which oligo(ethylsilsesquioxanes) are lbrmed in the hydrolysis of ethyltrichlorosilane [10, I ! ] and ethyldichlorosilane [ 12] has been established by GC-mass spectrometry. The tbrmation of oligoethylsilsesquioxanes through hydJ'olysis of C.,H~SiCla in aqtJeous butanol is shown in Scheme 6, Investigation of the hydrolysis of C.,H.~SiCI~ in aqueous butanol by GC-mass spectrornetry [13] shows the influence of the substituent X, and is shown in Scherne 7.

2. Trigonal prismatic {Si~Og} cages

th)wever, it is in reality it i|mltistep and ra|her colnplicat,'d process, Sprung and Guenther [5] were the first to postulate that the hydrolysis of orgmionyltrifunctional monouters XSiY~ involves tile forlllalion consecutively of linear, cyclic, polycyclic, and fitmlly polylledral siloxanes, attd assumed tile chain growtll to be of a nmdom character, This Ilypothesis was further devel° oped by Brown and coworkers, who studied the hydrolysis of cyclohexyl- [6] and phenyltrichlorosilane [7] and suggested the tbnnation of polyhedral silsesquioxanes and their honio derivatives occurred as a result of consecutive stepwise polycondensation of cyclic macromolecules, as illustrated in Schemes 2 and 3 respectively. On the basis of the polycondensation intermediates observed, cyclization at the initial step of the reaction and simultaneous lbrmation of linear and cyclic oligosiloxanes were proposed to occur. Co-condensation of the latter was believed to lead to polycyclosilox.'mes [7] which are inert to further intermolecular polyconden-

The first double three ring (D3R) silicate (type I. see Section i) c o m p o u n d s . (CH~)~,Si,O,~ and (C 2H,~),Si~,O,,, were first synthesized in 1955 by Sprung and Guenther [5.10,14]. The homologous compounds (C.~It~).Si,O,, [15,16], (CxHtT),Si,O,, [15-171, (iC,)IoI~,,).Si,O,, [15,16], (C,,Ht~),Si,O,~ [6.18] at,d (C,, H ~), S i.O,, [ Iq,20] were aim) successfully synlhe~. sized in the Ibllowing years by hydrolysis of their corresponding RSiX~ IIlOilOIl|ei',~ where X ~C! or alkoxy, In 19700 Smolin [21] was able to synthesize and determine the structure of the D3R silicate (type I with R = O :~) by reacting an ethylenediamine solution of nickel hydroxide with a 2,5% solution of ,~ilica in elhylenedianline. The structure of the crystals obtained exhibits two {Si,O~'; } cage anions surrounded by three [Ni(en)a]: + cations (en = ethylendimnine), hi between the anions and cations are situated hydrogenobonded water molecules (Fig. 3). Comp.ared to the idealized structure I (see Section I) the crystal structure of {Sic,O~; } is very distorted so that most of the O=Si-O angles are close to the tetrahedral angle whereby the D3R system is stabilized. hi the 1970s many investigations of acidified sodium monosilicate solutions by 2')Si NMR gave evidence that D3R silicates are also present in these systems [22~28],

Sd~eme 3.

Schtmle 5.

,XSiY~ + 1.5nH:O ~ (XSiOI ~),, + 3nHY

(I)

144

P.G. Harrison/ Journal of Organomemllic Chemisto' 542 (1997) 141-183 CH~CH(lluO)SiO 2 0

CHxzCH(lluO)§tCl "--*" CHe-CH(BuOhSi -

-XCI

.-lhOI(CH~-CMuO)SiO]2Bu-:

_._ llaO{(CH2nCMeO~iOhll u :

BuO[(CH~-CH)(BuO)SIOg011u

BuOI(CH~--CH)(BuO)SiO]sBu

-,,.

--:,-_ .

/

_

/

t ",~

/ )

II i

l I

I

11--

,''"

*

t

t

I

t t

...

Scheme

Subsequent investigations of silicate solutions with tetraethylammonium (TEA) as the cation showed that this cation had a stabilizing el'feet on the D3R silicate system [29=33]. It was tbund that using TEA as the cation in a silicate solution favoured the formation of D3R silicates to more than 95% if the optimum concentration range of Si and the optimum St:NEt2 ratio was chosen [29]. :~Si NMR of solutions of various Si concentrations and different St:NEt2 ratios are shown in Fig. 4. For a concentration of 0.9 M SiO, and a ratio of TEA:St- 3.7 there is a resonance at ca. -88ppm for the D3R silicate {Si~,O~'f}. In addition, small signals at ca. ~ 82 ppm and ca. - 7 0 p p m arc present which are due to traces of disilicate and monosilicate respectively. At an SiO: concentration of 3.07 M and a ratio of TEA:St-0.81, ~veral changes are observable in the spectrum, Besides the strong signal at ca. - 8 8 ppm for the D3R silicate, many small signals arc seen around that peak as well as many other small peaks in the disilicate and silicate t h e - and four-ring region at a~und ca. -80ppm and in the double four ring ( ~ R ) silicate region around ca, = 98 ppm. At a SiO, concentration of 3.95 M and a ratio of TEA:Si ~ 0,62"there are

I

I I I l I I I l I I

h,

major changes in the ~p¢ctrum, In addition to several sharp ~aks, broad hands arise in the disilicate, D3R, and D4R regions. Finally, at a SiO~ concentration of 5.1M and a ratio of TEA:St ~ 0.25 only broad hands centred at ca, - 79, - 87, - 97, and - 108 ppm are

present and the only sharp peak in the spectrum is due to monosilicate at ca. -71 ppm. So!id state NMR and X-ray crystallography show that the D3R structure is preserved on crystallization, The solid state ~¢Si NMR spectrum of crystals obtained from a TEA silicate solu. tion [29] with a ratio NEt ~ :St ~ 1,26 is shown in Fig. 5. The strong resonance at - 9 0 p p m indicates that the D3R silicate is dominant beside a small resonance at ca. - 6 7 ppm due to monosilicate. Similarly, Groenen et al. [34], investigating waterDMSO (l:l) solutions of different tetraalkylammonium/silicate ratios, lound that Ibr TEA as the cation the ratio of TEA and silicate had a significant effect (see Fig. 6(at and Fig. 6(b)). As seen in Fig. 6(at, for an NEt~:Si ratio of 0.5:1, besides small amounts of monosilicate and D3R-silicate the main peak observed in the ~Si NMR spectra is that of the silicon atoms of the D4R silicate (type II) (see Section I). Changing the

P.G, H(u'rison/ Jounud of Organomemflic Chemistry S4~ 11997) 141-183 EtSiC! z

.~ EI(BuO) SiCl~

: El(BuO)~ $iCI

Et~BuO) CI$iOSi(OBu). El ,,,

v

.//

I 1[

/ ~"

!

/ !

~.-

/

Tx

t

t

.,"

I 1

J

i

/

I

,L EtSi(OBu)~

I 1

"

\

,,

! \

¢ i

S c h e m e 7,

rtdio Io I:1 ,~sulls in tile stabilization of tile D3R-silicare, and only waces of D4R.silicate are now observed (Fig, 6(b)). Subsequently, Hoebbel el al. [35] synlheo sized and characterized by X-ray crystallography the trimethylsilylester of the D3R silicate, [(CH~)~Si],Si~,Ol,~, which had only been synthesized previously in dilute solutions by the trimethylsilylation of TEA/silicate or silicate solutions [29,30]. However. difficulties were encountered in preparing the crystalline ester due to the high instability of the quite strained 5i-O-Si bonds of the D3R structure towards proton attack [36]. The structure of [(CH.~)~Si]t,Sit,O|~ is shown in Fig. 7. As seen from the crystal structure of [(CH~)~Si]6Si~,OI~. SiMe~ groups are attached to the e x o oxygen of the {Sit,Ors} cage. and it is the steric effect of the SiMe~ groups which results in the stabilization of the D3R silicate compound. Si,nilar distortions are found in the structure of (Ct, Ht~)~,Si~,O,~ [37].

3. Cubane {SiaO,2} cages The most studied type of double-ring-silicate system is the cubane {SisO~:} silicate. Tile first representative

145

was (CH 3)sSisOl_, [38-41,14] synthesized by Scott [41 ], but many more IMR silicate compounds with alkyl substituents including C,H s [5,10,12,38,40,42A3], C3H 7 [38,40], CH(CH3) ", [40], C4H,) [38A0], csa.~ [441, C~,H,3 [15-171, /-C,~H,,~ [171, C6H., [38,45], lCioH 7 [20], 2-C4H3S [46] and aryl substiments including C6H 5 [6,40,44A7,48], and 4-CH3C6H 4 [48], have been synthesized. Some polyhedral oligoorganylsilsesquioxanes have been prepared by thermal depolymerization of organyltrichloro and o~anyltrialkoxysilane hydrolysis products. However, this method is not popular and so has not been used widely. The thermolysis of polymeric hydrolysis products of XSiY3 compounds is performed by heating at 200-400°C in the presence of a base catalyst [38,39,47,48,46] (sometimes in vacuum). Alkali metal hydroxides and in some cases tdethylamine are the most frequently used catalysts [48,46]. The silicate cages with saturated alkyl groups are chemically quite inert species. Compounds like (CH_, =CH)sSisOi, [10,49,50] (1) or HxSi80~,[51-53] (2), bearing a reactive group on the silicon atom. offer the possibility of obtaining new compounds by reaction of the exo cage function. However, these seemingly reactive groups, hydride or vinyl, on the D4R silicate cage unit are extremely unreactive compared to analogous Si-substituted linear, cyclic or polycyclic oligosiloxanes, presumably due to the rigid siliconoxygen framework and different electron density on the D4R silicate cage unit. Two reactions are known for the octavinylsilsesquioxane compound (I). Electron-beamor X-ray-induced polymerization yields thin films of the corresponding polymer (Scheme 8) used in a dry vac° uum submicrometre lithography process [54]. Bromina° tion of [I) in carbon tetrachloride is complex [55]. The main reaction products in the mixtures obtained were (CH ~ = CH)7(CHBr=CH ~)SiHO i~ and (CH~ =CH)r,(CHBr=CH~)BrSiHO~2. Altempts to achieve a more extensive bro.nination of (CH: =CH)xSixO~.~ under the same conditions failed. Octahydridosilsesquioxane HxSiNO~_~ (2) was first prepared fortuitously in about 0.1% yield in 1959 by Mtlller et al. [51] while studying the preparation of poly (h y d rid o silsesq u io x a n e s ) . Me~SiO[HSi(OSiMe~)O],SiMe~. in 1970 Frye and Collins [52] increased the yield to 13% by an improved method involving the careful hydrolysis of HSiCI~ in a benzene=concentrated H~SO4 mixture. A further improvement was made in 1991 by Agaskar [53]. who developed a nt:w synthetic procedure which gave a mixture of HHSiKOI: and Hi,SiloOl.~ in high yield (about 27%) by using partially hydrated FeCI~ as the source of water for the hydrolysis of HSiCI:~. isolation of itxSixO~ was very easy to implement by crystallization, giving a yield of ca. 17.5% of pure HMSi,Ot~. However, the recognition in 1988 that the {SiHO~:}

I~

P.G. Httrrison / Jourmil of OrganomeUl//ic Chemisrn" .~42 f 1997~ 141- I,~3

,
Fig, J, Cry~lal slriiclUi'e ot°[Nilcn)~]t Sir,()l~ ' 2~14,O,

{.~7--64] would atl~Ird a l~tter uaderstandiu~ of zeolites and I~ad to new chenlistry of rchited .,~tructures. For .oaocyclic silallcs ittally i~actions of th~ Si=H

framework I~re a close resemblance in the D4R fi~und in Linde Type A and ia Co APO=~0 lyp~ zeoliles [~(~] (FI~, 8) renewed a particularinterest in octahydrio dosilse~uioxane. In consequence, it was coilsider~d that li d~tuiled spectroscopic investigation of H~Si~O~:

t'oilCtion arc k n o w . , iucluding i~aciioii with orgtillO° lithium r.ji,lgt~lils [~i.s,6(q. alkeucs [01l, i~.z~.~ 16 -761,

IiA'I, Sl~O I

li. t

'-

~ .

.

.

.

.

.

.

.

.

.

.

I

iiiii

tll I.,

--

ii

ii

.

_

i

_

,i

.

i

i

i

-

_

~

ii

!

i

i

Fig. 4, ~Si NMR sff~tra of solulion.~ oF varh~s $i c~a~.'cntr,ztions and diff~r,~nt Si;NEI~ nltios.

,

P. G. Harrison/,hmrmd o,l"Or,qamm~etailic Chemitto" 542 ~i 997; 14 I - I &¢

147

TEAOH

--60

-80

TEAOH

-100 -120

-60

(a)

-80

-100 -120

(b)

Fig. 6. -")Si NMR spectra of TEA silicate water-DMSO (!:1) solutions with NEt~ :Si ratio of 0.5:! (a)and NEt~ :Si ratio of l:l(b).

iiiii

.io

.-;o

.d0

I'1

"100 "1;0

•

"1;0

"1;0

-140

I

(2) in CCl 4 (Scheme 9) results in > 95% yield of ClsSisOi2 (3), which is less reactive and therefore easier to handle than many other chlorosilanes [99, I00]. The corresponding octamethoxy compound (MeO)~Si~Ot., (4), is obtained in 45% yield by further reaction of ClsSi8Ot., (3) with MeONO (Scheme 9). The crystal structure of (MeO)~Si~Ot., is shown in Fig. 9 and illustrates the {SisO=.,} cubane cage with MeO groups attached to each silicon vertex. The Si-O angles are close to tetrahedral angles due to distortion of the cubane cage. The reactions of octahydridosilsesquioxane (2) with Me aSiOSnMe.a, Ph4SbOSbMe~, Me aSiOSbMe~, and Me~SnOSnMe.a were successful only in the two latter cases (Scheme 9) giving Me~SiOand Me aSnO-substituted {Si~Ot,} cages respectively.

•

-150 -160 6 [ppml

Fig, 5. Solid state "~"Si NMR spectrum of crystals obtained from a TEA silicate mlution with a ratio NEt 4 :Si = 1.26.

diazomethane [77], alkali metals [78-80], alcohols [8186], alkali-metal alkoxides [87], aldehydes [88], ketones [89-91], carboxylic acids [92,93] and halogens [94-98]. In contrast, however, the hydrido function on octahydridosilsesquioxane (2) is relatively unreactive. It is quite remarkable that it took nearly 40years alter the initial synthesis of HsSisOt., before the first Si=H substitution reaction of this cage compound was reported by Day et al. in 1985 [99]. Photochemical chlorination of H, Si ~O~2

Cr

Co

C,,

Sq

Os Ctlr ~

-qF

°' $L~j

C,,

Cts e,,'-"

0

,=

Fig. 7. Crystal structure of [(CH .~)~Si]r, Si,,Ot ~.

14g

P.G. Harris¢m~Journal of Organometallic Chemisto' 542 (1997~ 141-183

(~tm

I

•"

,__p...~ ZoO°

i

\

"-,

|

(I)

polymer

mixture (main products:Hepmvinyl(bmmoethyl. and hexavinylbisbromoelhyl ocuuflscsquioxane) Scheme 8,

With Ph4SbOSbPh~ the Ph.tSbO-substimted spherosilicare could not be obtained, and only mixtut'es of (C~Hs)~Sb. C~H~,. St=H-containing products and an intractable solid were found [101]. H~Si~Ot~ reacts with the silylation agents (CH ~ ~ CH)Si(CH ~):CI/M e~NO or to give the silylated sphemsilicate~ St~Oj~[OSi(CH ~),(CH, -~CH]~ ( ~ ) and Si,Ot~[OSi(CH~)~CI]N (gb) resl~ctively [!i)2], However, Me, ShOo and Me~SbOosubstiluled {Si~O~:} cage~ could not be obtained by the corresponding ~aetion of H~Si,Oi, with M e ~ S n C I / M e ~ N O or Me~SbCI/Me~NO [101], However, octahydrio dosilsmsquioxane reacts with Me~SnOMe to afford a complex mixture of incompletely trimethylstannylated (SisOt:] c ~ s (i,e, Hs=,~(Me~SnO),,SisO~:) [103]~ Feher and Weller [104] showed that reaction of octahydridosil~uioxane (2) with Me~SnOSnMe~ resulted in the formation of compound (6), which could be transformed into compound (7) using either Me~SIOSbMe~ or M¢~COSbMe~ as r e ~ n t (Scheme 9). Bth'gy and Calzaferri [105] have reported the palladium~cataiysed H/D exchange by bubbling D: into a pentane solution containing octahydridosilsesquioxane (2) and Pd/C catalyst at ambient temperature (Scheme I0), Several platinumq:ataly~l hydrosilylation reactions have ~ n carried out using H~PtCI~ as catalyst [106]~ ~ta(lo~xylsil~squioxane) and octogeyclohexylme;hylsil~uioxane) have ~ n obtained by reacting ~tahydridosil~squioxane (~) with hex-I~ene and methylenecyclohexene respectively [107]. This offe~ a new syn0~tic route to polyhedral organylsil~squioxaaes, and the general reaction shown in Scheme I0 can lead to a large number of new oligosilsesquioxanes [108-110], Bassindaie and Gentle [111] have employed

(CICH~)Si(CH~!~CI/M¢~NO

the hydrosilation method using octahydridosilsesquiox' ane to ob 'tatn siloxane and hydrocarbon octopus molecules with sUse~uioxane cores. An akernative pathway to (¢-C~Hlj)sSisOj2 is by the catalytic hydrogenation of PhsSisO~, [112] at 34 arm and 200°C. The compound obtained is reported to be identical in all respects to that prepared by the hydrolyric condensation of (c°C¢,H~ )SiCl~ [6]. Tack¢ ¢t al, [I 13] have described two synthetic routes to ~tminoalkylosubstituted cages. The reaction of (I la) in refluxing acetonitrile tbr 18h yields octa{[24dimethyloamino)phenyl]silsesquioxan¢} ( 11) in 00% yield. A yield of 83% was t~poned for the hydrolysis of (1 lb): however, for this the reaction time needed to be increa~d to ~veral weeks (Scheme II). Both C~H~ and 2-C4H.~S substituents on the silicon atom of octasilse~uioxane can be modified. Treatment of phenyl groups with HNO~ [48] results in nitration to give NO~C~H~ s~.' ~tiwuents,and treatment of 2-C~H ~S with Br: in HBr as solvent [46] gives 2-C.~Br~S as the substituent. A broad range of new highly functionalized organyl-

Ca~

Co) o P

Fig, 8, Structure of HsSisOl: (a) and part of the zeolite A structure (b),

P.G. It, n'ison / Jounml o.[Orga,om~.mllic Chemisto" 54211997) 141- I&'~

149

R'(:H:I)2SIC'~ MeaMSieOSbMe4Mos (_llo,.' ~) "OSiL~ "\

~~-. ('

.

H

hv ~ C C 1 4 / Ck

CI

c~

8 CHsONO

C12

(2)

f

(a)

NMesSnOSnMes "~esSnH UesSnO~.. '~_ OSnMes 0

'~Ol

L~,,s.o

MesSiOSbMe4 or MesCOSbMe4

- 8 NOCI

)SbMe4

Oh

MeO

eT~OMe

Me

~JO~~MekOMe MeO

"u'lmS'~08nl~s

(e)

~0 O~bMOd~ Me,SbO

(4)

8bMo4

Me,8g (7)

Scheme9.

silsesquioxanes was reported by Feher and Budzichowski [I 14] based on (p-CICH,C¢,H4)sSisOI: (12). formed by the hydrolytic condensation of pCICH2Ct, H4SiCI.~ in aqueous acetone. (12) is a synthet-

ically useful precursor tbr the synthesis of several octafunctional polyhedral oligosiisesquioxanes (PXCH:Cc, H4)sSisOj:, including X = ! (13). X =OH (14), X - - O N O , (15). X = O A c (16). X = p-

I~

P.G. Harrison/Jourmd rq'Orgwu)metaili,"Chemisto' .542 (1997~ 141-183

Ct

vj

Fig., 9, Crystalstructureof (McO)KSiS.0t2.

nitrobenzoyl (17) and X = methylterephthaloyi z18). A summary of the reaction is given in Scheme 12. Most of the heterosubstituted oligosil~uioxanes known, however, have been obtained by co-hydrolysis using different XSiR~ (R = Ci. CH~COO) monomers. A summary of heterosubstituted ~tasil~squioxanes X,,X~,,,SisO,~ obtained by coohydrolysis using differ° ent XSiR~ monomers is given in Table I. The co-hydrolysis of a mixture of trifunctional XSiR and X'SiR~ monomer, usually gives a mixture of

heterosubstituted oligosilsesquioxanes with all possible combinations of substituents, X and X' [ I 15. 116]. When used for the preparation of heterosubstituted octasilsesquioxanes, for example, this reaction leads to a mixture containing nine compounds of the general formula X,,,X~_,,,Sis012 with re=O-8. Compounds of this type with m = 2-6 may, in turn, represent mixtures differing not only in the composition, but also by the substituent position. The relative yield of particular heterosubstituted oligosilsesquioxanes with various m values depends mainly on the molar ratio and reactivity of the initial monomer. In the case of an equimolar ratio of both monomers and similar rates of hydrolysis, the compound with an equal number of X and X', X4X4SisOi,, is formed in highest yield. However, variations of the molar ratio of the initial monomers results in different products and yield. Thus co-hydrolysis of XSiR~/X'SiR.~ mixtures in the molar ratio !:7 leads to the formation of XX'7SisOI: as the major product. The yields of heterosubstituted oligosiisesquioxanes are significantly affected by dift'etences in reactivity of the initial monomers. Martynova and Chupakina [120] have shown by GLC-mass spe~'trometric analysis that the preparation of mixed (CH ~),,(CH: =CH)~_ .SisO~: compounds .,ia the reaction CH ~SiCi~ + CH: =CHSiCI~ + Ho,O ~, (CH~),,(CH, =CH)~,,Si~01:

in alcoholic medium does not lead to reproducible

H #

H~a~C~ I, OH2~CH(CH~sCH~ N

14

(2)

o~U=l<

Of

~. C H ~

"

i

R 1, R a (CHa)sCH~ Of

2, R.CH n

D O

(0) Scheme 10,

(lOa)

P. G Harrison / Journal of Orgmmmetallic Chemisto" 542 ¢1997) i 4 I - 183

compositions of the oligomers. Reproducible composition however can be achieved in changing using the corresponding acetates, CH3Si(OCOCHg)3 and CH 2 =CHSi(OCOCH.0.~. The results are shown in Table 2. The efficient synthesis of compounds with only a single heterosubstituent is rather difficult. The best method appears to be by the reaction of one Si-H function of HsSisOl,, (2) with an alkene in a 1:1 molar ratio, whereby compounds (19a, 19b, 19e) have been obtained [109-111,121] (Scheme 13). Octacarbonyldicobalt Co2(CO) s reacts readily with silanes at room temperature (Eq. (12)) [122,123] to give silylcobaltcarbonyi compounds. This type of hydrogen elimination accompanied by cleavage of a metal bond is common in organometallic chemistry and many of the known Si-Co(CO) 4 compounds can be synthesized by applying this reaction [124-126]. The reaction between HsSisOi2 and Co2(CO) s in a 1:1 molar ratio leads to the formation of the first monosubstituted hydridosilsesquioxane HTSisOI2Co(CO)4 [127] with a silicon-metal bond as shown in Scheme 14. The crystal structure of ((Co(CO)~)-(H~SisO,.,) [127] (Fig. 10) illustrates the {SisOt.~} cage unit with hydrogen atoms

! 51

attached to seven of the eight silicon cage atoms. The eighth silicon cage atom is attached to the Co(CO)4 substituent via a silicon-cobalt bond. The coordination of the cobalt atom in the structure is trigonal bipyramidal. When a 1:2 molar ratio of reactants is employed, however, the bis-substituted cage (20b) is formed which has been characterized by NMR data [128]. Despite this success, no other reaction between HsSisO~2 and a metal carbonyl compound, like the very common reactions of a non-cyclic silane R3Si-H with CpMn(CO) 3 [129] or Fe3(CO)t2 [130], has been reported. (c-CtHll)7HSisOl2 (22) [131,132] and (cCtHtl)TCISisOl2 (23) [131] have been prepared by the reaction of the incomplete condensed cage (cCtHtl)TSi~Og(OH) 3 [132] (21) with SiCI 4 and HSiCI 3 respectively (Scheme 14). (c-CtHit)7(Me3SnO)SisOl2 (2Aa) and (c-C~Hll)7(Me4SbO)SisOl2 (24b) have been synthesized by reacting (c-CtHit)THSi80!2 (22) with Me3SnCI/Me3NO or Me4SbCI/Me3NO respectively [ 101] (Scheme 14). Like octahydridosilsesquioxane (2), (22) undergoes palladium-catalysed deuterium exchange which results in the formation of (25) [131] (Scheme 14). All three compounds (22, 23, and 25) undergo Wittig-reactions with R'3P-- CH 2 reagents (Scheme 14),

&

a

.24 obox

"R

R

8 Q41~)'---" e-"'--O(~ll

(11) (lib) Schetne II.

P.G. Harrison/ Journal of Organometallic Chemisto, 542 (1997) 141-183

152

CNm

H~O, &celone

o~_

+m.,.,/~

111mO,--:

o?

~'~.~o-,~.-0-0+

~,,,~?-O-o'-.~.,~ jo_.+_~ ~

Nal

~-~'~--~-o-s~

~.c.~

02)

(13) I(IOCI'I I

.--,~~,-;-+.~,~;~ ,&o_# ~+.+.+,,

J

(1,4)

0 G -,,0CI41

i01

o + ~ ~l,-,,Om~: - ° . ,

ti

Ore| It°

~

ll~ ,0+'-|

I,'

~l"~..o-a..l

(16) R : Me (IT) p.O=NCeH4 (18) p-MeO2CCsH~

(15) ~hen~ 12,

i 53

P,G Harrison/ Joumol of Organometallic ChemistD"542 ¢1997~ 141-183

Table I Heterosubstituted octasilsesquioxanes X mX~_ mSi sO t2 X X' m Reference C 6H I I CH~ C2H s CH~ t-C.iH9 CH=CH, CH~

H C~Hs CH=CH, C6Hs C6H s CH.,CH, Br CH=CH 2

7 I-7 I-7 4, 6, 7 6 7 !-7

Table 2 Co-hydrolysis of CH .~Si(OCOCH+~)~and CH 2 =CHSi(OCOCH~); Ratio of the inital reactant Major component in product CH ~- and CH 2 =CHSi(OCOCH.~)~

[45] [115] [116] [I 17,118] [I 19] [55] [120]

i/7 2/6 3/5 4/4

(CH.~XCH2 =CH)7SisOI: (CH.~)2(CH2=CH)6SisOl2 (CH 3)~(CH2 = CH)5Si sO 12 (CH ~)4(CH2 =CH)4SisOl2 (CH .~)s(CH2= CH)3SisO 12 (CH .~)6(CH2= CH)2Si8OI2 (CH 3)7(CH2 =CH)SisOI2

5/3 6/2 7/I

and the proposed mechanism is illustrated in Scheme 15. The silsesquioxane-substituted phosphorane (26a) can also be used as Wittig reagent and yields many monofunctionalized alkenyl-silsesquioxanes with a wide range of aldehydes [131] (Scheme 16). The product derived from (26a) and p-(PhCH20)C<+H4CHO affords compound (29) in high yield after hydrogenation with Pd/C [131] (Scheme 17). The cubane silicate (30) was first synthesized by Hoebbel and Wieker [133] in 1971, after being characterized by X-ray structure in the mineral ekanite (K ~CaThSisO.~0) by Mokejewa and Goiowastikov [ 134]. The reaction of a tetramethylammonium hydroxide solution with silica gel in the ratio N:Si - 1:1 results in the formation o f crystals of c o m p o s i t i o n [!CH~)4N]~[SisO+o]. 69H+O. Since the development of "+St NMR was at its initial stage at this time, the

condensation rate of the silicate anions in the tetramethylammoniumsilicate was investigated by the molybdate blue method [ 135] and paper chromatography [ 136,137]. However, the use of 2qSi NMR in the following years resulted in a better understanding of this and related systems. Silicate silicon atoms can have different chemical environments and are subdivided in five groups which are described by QO Q I Q 2 Q 3 Q4 as shown in Fig, 11 w i t h 2~Si chemical shift ranges in Table 3. A

typical example of how complex the constitution of a solution may be is illustrated in Fig. 12, which shows the spectrum of a 4.1 M sodium silicate solution with Na:Si ratio of 1:1 in which many different silicon sites may be distinguished, The copper-complex [Cu(en) 2]4Sii,lO2ti , 38H.~O has been obtained by reacting an alkaline aqueous

H

"\

HlPtCle

• ,,

H

HI c

i --

H

i

/a

-o,,

H # H

(2)

I/2Co2(C0)e

F-

Toluene 60oC 4h

R., CH2--CHr"CH2-" CHs

R- ~ )

H Re

14

(2oa) Scheme 13.

@ @ Fe

(1St))

(19C)

H I O =g ! H

Ooa)

1:$4

P.G. Harrison / Journal of Organometallic Chemist~. 542 ( ! 997) 141-183 R

R

SiC~ n

R

n

(al)

n

(2s)

HSiCIs

I1

n\~

R H

.cHin',

~, R'sPmCH;e

n n

~

Cells(2eb,

II

"

(22)

m~mm.(:mmmNO

m.~H,,

R

~Pu/o~ n'~P,~H2

O

ml~rtb) # R

R

(24)

(as) Scheme 14,

(Cu(en)[ + } solution with silica gel [138,139], Fig, 13 shows a projection of the crystal structure onto the (100) plm~ and illustrates the isolated nature of the (Cu(en)~+ } cations and {Si~O~° } silicate cage anion within the structure, Hydrogen.bonded water molecules are situated between the (Cu(en)~+} cations and the {$i,0~= } silicate cage anion. To exami~ the influence of ~ cation on the silicate cage structure, Hoebbel et aL [140] reacted an aqueous solution of cobalKlllkliam~ n e hydroxide with silica gel or tetramethoxysilane which yielded a compound of the composition [Com(en)2]~H=SI,O~, The X-my structure [141] (Fig.

14) unexpectedly showed the existence of two cage silanols located in opposite corners of the cubane silicate, illustrating how easily the cage oxides can be protonated in order to compensate for the charge of cation. Similar results were obtained when [Co(pn1,2)~]~* and [Co(pn-l,3)~] ~* [pn ~ propylendiamine] were employed as cations. Wiebke and Hocbbel [142] obtained c r y s t a l s of the c o m p o s i t i o n [(CH~).~N]|<,[Si~O:o](OH),~ • 116H20 by fractional crystallization at room temperature from an aqueous tetramethylammonium silicate solution with a molar ratio N:Si = 3:1 (concentration of Si = 0.56 tool I- z; concen-

155

P.G. Harrison~Journal of Organontemflic Chemisto, 542 (1997) 141-183

H

o(~a)~

.,~~'ColCO14

H H

(20b)

o~1

R41

o02)

Fig. 10. Structureof H.tSiHOt:(Co(CO)4). silicate {SisO~o } anions, OH--ions and H:O molecules which are linked via hydrogen bonds. Fig. 15 shows layers within the host structure. Nitrogen atoms of the guest species NMe2 are shown as filled ellipsoids.

tration of NMe4OH 3.47 mol I- t ). X-ray studies of suitable single crystals showed the compound as a hostguest compound whereby the NMe~ cation is suited in the cavity of the host structure built up by oligomeric

"R

g

" 4,

.

Me~P=CH2

e,,,CHl~

g R

(2s) R - Cell.

MeaPICHaD i

~..---

~

in

, ,,m

," R m

(2ee,

Scheme1.5.

R

i

"*

t56

P.G. Harrison/ Jou~l of OrganometallicChemistry542 (!997) 141-183

R

R

.

T

'-:

R C= R' = Cells

R'CN0

(27)

R

(2~) R = CeHI I

m m

m

R' " ~ s C(CHdOH2 (CHdICHCH= p~.~H40~HaG6Hs p,~OC(O)C(CHs)CHa

I~-~H48r Scheme 16,

Using the cations NPhMe~. NBnMe~ (Bn - benzyl) or DMPI (DMPI - I,l~lin~thylpipeddiniun0 respectively, instead of the tetramethylammonium cation, yielded three new compounds of the composition

[NPhMe~]~[SisO,s(OH)~.], [NBnMeo~]~[Si~O~.] . 53.6H~O and [DMPI]~[SiKO,~(OH)~].48.5H~O whose crystal structures aim showed host-guest features

[143,144],

R m

OH

(n) errant,3 Scheme 17,

P. G. Harrison / Jounml of Organomeudlh" Chemistr3., 542 f 1997) 14I - id3

HO

"OH

HO--

O--"--O-SJ--OH H

Qo

H

HO - - Si - - O --

H

H

°X

--O-H

Qt

i 57

OH H

Q3

-o

o-

"

__,

l

o-

/

s,o

Uessid

"O~

Q3

Q4

(30) Fig. I I. Different types of Q silicate silio.m atom.

Table 3 ~'~Si chemical shift ranges of different envimnmems Nun)bet of S i - O - S i bridges Monosili¢~ltes Si(OH)~ End [ffoups Si(OSi){011) Middle Stoups Si(OSi)~ R)H)~ Bran~:h~d tlroups Si(OSi)~(OH ) Net I]mups SI(OSt)~

-70

..

ldentil'icmion

0

Q"

I

Q'

2 3 4

Q~' Q~ Q~

<6

-8i~

-gs 1o .....

Signal group (see Fig, 12) A (ca. - 71 ppm) B (ca, -711 to = 81 ppm) C,D (ca, ~ 81 to = 82 and ~ 80 to = 91 ppm) E (ca, = 93 to - 97 ppm) F (ca, - I(~)to - 120ppm)

-o0 7S

-os tb~,,,,) - W _ z'o

a~

¢'bpm]-

Fig. 12. 2~Si NMR spectrum of an aqueous sodium silicate solution 4. I M and ratio Na:Si ~ i:1,

158

P.G. Harri~tm/ Journal of Organometallic Chemistr3., 542 (1997) 141-183

(,,)

Fig. 13. Crystal structure of [Cu(en): ]4SisO2o' 38H :O.

"~Si NMR studies have shown that the (SisO~o} silicate cage (30) is not stable in alkaline aqueous solution with tetramethylammonium as cation. A significant advance, however, was made in 1986 by Groenen et al. [145] who showed that the [Si~O~;| silicate cage (30) could he stabilized in solution by using a I:1 mixture of DMSO=H;O. The 'clean' nature of these solutions is illustrated by the ~Si NMR s ~ t r a (Fig. 16). Spectra for both 0.$:1 and I:1 ratios of tetrametho ylammonium/Si exhibit only one sharp peak characteristic of Si in the (SimOny) cage structure, and no resonMces due to other species ~ apparent. The reason for ~is stabiliution was not commented upon, and remains open m speculation, The c ~ may also he stabilized against cleavage by silylafion of the exo cage oxy~ns. Hoebbel and Wieker [133] have m~fied the {SisO~ ) silicate cage by Lentz tH~ylstlylatton [146] of [(CH ~), N]s[SisO~o], 69H,,O

~¢Q

Fig, 14, The (H~Si~O~} ~" unit in the crystal structure of [C°#t(en)~]~H~Si,O~ • "

to yield the stable trimethylsilylester [SisO~o](SiMe.~)s Old). Silylation of [(CH~)4N]s[SisO:o].69H:O with vinyldimethylchlorosilane (or divinyltetramethylsilane), allyldimethylchlorosilane and dimethylchlomsilane leads to the three functionalized spherosilicates (31a-e) [147,148] (Scheme 18). These new compounds (31a~c) have good ~lubility in organic solvents and are very stable towards hydrolysis and condensation, so that cleavage of the silicate cage does not occur. I~rther, the capped silicates (31a~e) hear reactive functional groups which should allow the compounds to be precursors for the synthesis of new inorganic~-organic polymeric mate° rials. Agaskar [ 149] has reported a facile one-pot, high

-

" X

Fig, 15, Crystal s,*ngture of [(CHO~N]I~[SisO~o](OH),.II6H~O. Nitrogen atoms of(CH~),N" are shown as filled ellipsoids. "

P.G. Harrison/ Journal of Organmnemllic Chemisl~, 542 f 1997J 141- I&~ TMAOH

TMAOH

8 .-6O

-8O

-I00 -120

(ppml

tppml

Fig. 16. ~Si NMR spectra of a tetramethylammonium silicate DMSO-H20 (l:l) solution with tetrmnethylammonium/Si ratio of 0.5: I (a) and l: I (b).

yield, route for the synthesis of the compounds [SisOao](SiMe3)s, [SisO20]SiMe2CH=CH2)s and [SisOao](SiMeaCHaCI)s respectively. In the procedure, three solutions are mixed in sequence and the product extracted from the reaction mixture with an immiscible hydrocarbon solvent. The constitutions of the three solutions are as follows: Sl: H20 (20ml)+ Me4NOH (25% aqueous, 20ml) + (MeO).~Si (8.1 mi) + Me.,SO (42 ml) + H20 (H20:Me2SO - 1:1; [Si] ~ 0.65 M) $2: (H.~CO)2C(CH3) 2 (70ml) + HCI ( ~ 10M, 3.0ml) + YSi(CH3)2OSi(CH3)2Y (10ml) (Y -- -CH~, -CHCH:, CH:Cl) $3: (CH~)~NCHO (20ml) + YSi(CH 02OSi(CH.~),Y (10ml) + Y(CH3)2SiCI 3 (5 ml) (Y ffi -CH ~, -CHCH 2. CH:CI) The procedure employed is as follows: solution St which contains the [SigO~o]t~ anion is added dropwise over a period of 20 rain to the solution $2. This mixture is then stirred for 15 rain. During this time the reaction of H,O in solution SI and (H~CO)2C(CH3) 2 in solution S2 leads to the formation of methanol and acetone, whilst the [SiaO~0]H~ anion is partially silylated. Solution $3 is then added to complete the silylation reliclion, and the product is separated by usual methods. More recently, novel highly functionalized D4R cages have been synthesized. Silylation of the {SittO~o } cage with HMeaSiel followed by catalytic hydrosilylation of the Si-H function with appropriate alkenes yields the

159

four new compounds (32a-d) [ 150-154] (Fig. i 7). The functionalities on the exo-cage groups are also available for the reaction; for example, compound (32¢) can function as chelating ligand towards Zr(O"Pr)4 giving the complex (33) [151,153] (Scheme 19). A similar complex (34) has been obtained by the addition reaction of the I ~ R cage [SisO2o](SiMe2H) s (31a) with (allylacetylacetonato)bis(butoxy)aluminimn [153] (Scheme 20). Additional derivatives based on (31a) have also been reported, Hydrosilylation with vinylferrocene affords both mono- and octasubstimted products [154], whilst hydrosilylation reactions with l,l'-divinylferrocene and l,l'-divinyl(octamethyi)ferrocene result in the formation of poly(ferrocenyloctasilsesquioxane) polymers. Unsymmetrical cages which have more than one different group attached to the silicon atoms have been obtained by a number of different methods. In 1992 Hasegawa and Motojima [155] investigated the synthesis of [Si80:o](SiMe~CH=CH:) ~ by silylating {SigO~o } in a methanolic solution using dimethylvinylchlorosilane. The gas chromatogram (Fig. 18) showed that, besides [Si~O~o](SiMe2CH=CH2) 8 (peak A), incompletely dimethylvinylsilylated derivatives of {Si80~o } with one silanol group [SigO2o]SiMe:CH=CH~)TH (peak B) and two silanol groups [Si~O:o](SiMe:CH=CH2)6H~ (peak C) were formed. Furthermore, silylation of the {Si80~o } cage with a mixture of vinyldimethylchlorosilane and trimethyichlorosilane yields, depending on the ratio of the silylating agents, mixed (vinyldimethylsilylXtrimethylo sily l)-su b stitu ted c a g e s [Si,O:,](SiMe.~),,,(SiMe2CH =CH 2)H~,, [ 147]. Capillary gas chromatography of the silylated mixture using the two silylation reagents in a I:l ratio showed nine peaks (Fig. 19(A)). The two extremity peaks (peaks a and b) are due to the totally trimethylsilylated {SiaO~7} cage [SiaO~o](SiMe~), and totally vinyldimethylsilylated ~'"-{SiaO~., ~ } cage [SiHO20](SiMe2CH~CH~)a. The inter-

"0

*4

(~o)

($1)

SiMe2H SiMe2CH=CH2 c : R' = SiMe2CH2-CH=CH2 d : R' = CH3 a : R' =

b : R' =

Scheme 18.

I~

PoG. Harrism~ / Journal ~" Organomet~dlic Chemistr3" 542 ~i 9~7) 14 ! - 183 0 I

.

0

i_If-I

-

,

(~+,a}

OH|

~,

o"

"°"s,.,_..~,~ f " t ' I

o

~klr CHr ~I--OCH~ o

CI,,I~

¢141___

p

Ho o-

~

o

t CHI

o

#

i@Q'

.

Fig, I 7, D4R°~:~es obt~ir.ed by ctttalytl~ ,~ilylationof 131#t with ~lkenes,

mediate peaks can be assigned to the ~ven mixed species [Si~O~)](SiMe~)+(SiMe,CH~CH,) (peak I) to [SisO~](SiMe~XSiMe~CH~CI~)7 (peak~ 7). The ma° jot products ate [SisO~.t~]~SiMe~).,(SiMe~CH~CH~)~ (peak 4.)and [Si~O~.oI(SiMe~)s(SiMe,CH~CH~)~ (peak 3), Ch~ging of the ratio of silylation agents (CH ~)~SiCI and CHz ~CH(CH~)~SiCI to 10:1 results in only three main peaks visible in the capillary gas chromatogram due to [Si~O+J(SiMe~)~ (49%), and the mixed silylated compounds [SisO,~]~$iMe~)~(SiMe+CH=CH: ) (25%)

and [Si~O~](SiMe~)~,(SiMe:CI°I~CH,), (7%) (Fig. 19(B)). D4R cages with different functionalities can be oblained by the silylation of {SimOny} cage with a mixture of dimethylchlorosilane and dimethylvinylchlorosilane [153,154,156], Five peaks are seen in the gas chromatogram, the lurgest peak being due to [Si~O,~](SiMe:H)~(SiMe~CH=CH~)~ (peak 3)(Fig. 20). The others p e a k s are due to [SiaO:o](SiMe:H)~(SiMezCH=CH2) z (peak I), [Si~O2o](SiMe2H)~(SiMe,CH=CH2)3 (peak 2),

P.G. tt++rrison/ J+mn+,l +~'Organomemllh" Chemislr)" 542 fl997) 141-183

I

,,o \

o" ~

~

P" c ~ - c ~ o---T-

/c c t ~ - o-- c II

0

I61

14= \C--CH= II 0

+ e Zr(OPr"),= (a2c)

.o. .

H

T-

R'

~:H+

(:HI"-(:HI--CHI--O-- ( C , ~ CH:I

--

i

\,_/

CH+

0

0

'H

¢sHT CmHT R'

I

,.m,

|

R' R'

(") Scheme 19.

.o\

-

~

~. +8

H

0-- / ~

(re+a)

,

H

p.,

HoC

R'

(a4) Scheme 20.

0¢dtt

-'CH+

I~

P.G. Harrison / Journal c~ Organometallic Ci~emistr3."542 f 1997) 141-183

B 1

H(~

i

1

I

!

SilC~'-CH2

T

20 25 3O RmrCk~Tm~ I n'm Rg. lg. Gas chromalogram of the products of the situation of ($i~O~;" } using di~ylvinylchlorosihme after reaction time of I h and dimelhylvinyichlowe~ileme/($i,O~o ) ratio of 32:! (a) and 48:1

Q HMaV

(b).

[Si~O20](SiMe.,.H)3(SiMe:CH=CH,) s (peak 4) and

[Si~O:o](SiMe.,H):(SiMe:CH =CH: )~, (peak 5). 4. Double five Hng (DSR) [Sl,oO,s} cages

.

.

.

14,0

The D5R cage silicate of general formula R t0Si toOls has five-fold symmetry and comprises two pentasiloxane rings (see Fig. l). DSR-cages where R ~ organic ~ p are more common than those DSR-silicates bearing inorganic groups on the cage silicon atom. Relative to D4R=cage silicates, few DSR-silicates have been pNpared and characterized, i.e. HloSijoOl~ [52.~3].

.

15,0

,

I 0

17.0 IIItlm~ m

Fig. ~ . Gas ¢hromatogram of silylation products of the {SisO~~} cage with a mixture of dimethylchlorosUane and dimethlyvinylchlorosilane ratio I: I.

(i-C~H~))loSi~.O,~ [17]. (CH=CH.)i.Si,.O~ [50]. (C~H~)II~SiwOI~ [47.48], Only two compounds arc known which possess a reactive functional group on the

#e.

~

m

i

A

B

j ~

St,

L. "

t

.~.¢smmsm, t,l~...mare

,!0

- *~

0

Fq~, 19, G¢~ dcoma~sram of silyhtion products of the {$i~O~} cage with a mixture of vinyldimethylchlo~ilan¢ and trimethylchlorosilane r,llio I:1 ( A ) ~ 1 l ~ l (B).

P.G. Harrison / Jounm! ofOrgm~mnemlliv Chemisw3'542 fl997) 141-183

163

Scheme 2 I,

DSR-cage silicon atom, H and CHa-CHa, and only ~oactions involving H .~Si t,O~ have been described. No easy one-pot route is known lbr the preparation of H i,Sit,Ol~ (35). Its preparation is always accompao nied by mtJjor quantities of HHSixOt,~ and higher hydro° silsesquioxanes [52,53], and hence time.consuming purification procedures need to be adopted. It is perhaps remarkable that 29 years elapsed from the initial discovery of H t,Sit,ot~ to the description of the first reaction of this compound. Using the reagent (CH~)~NO. CISi(CHa)~(CH=CH:) Agaskar [102] in 1989 converted the Si-H function of H toSi toOt~ thereby obtaining the [SiloO,5](Si(CH3)2CH=CH2)io (36). In the

same way [Si.IO~](Si(CH~)~CH~CI).o (3?) has been synthesized by conversion of the Si-H function of H.~Si,IOl~ (35) with (CH~)~NO. CISi(CH~),CH~CI (Scheme 21 ). Very little work has been carried out so far on compounds containing the anionic {Si~0,~,:~:~"~=}silicate unit. In 1975 Hoebbel et al. [159] reported a novel silicate consisting of DSR {S11oO25 ' too } anions in [n(C~H,~)4N]OH-SiO2wH20 and [i-(C~Htt)4N]OHSiO:-H20 solutions. The corresponding trimethylsilyl ester [SitoO.~](Si(CH3).~)to [159] (39) has also been 5 } (38) using the synthesized by silylation of {St,~O,to~ method of Gtitz and Masson [160] (Scheme 22). Like '

re, (Cll~)~i09i(Cll~)¢ (CHj)~ICI

Oil

Scheme 22.

P.G. Itarrison / Journal of Organometallic Chemisto, 542 (1997) 141-183

!64

3

Fig. 21. Geometrical i.u3me~ of dodecahydridosil~squioxane (40).

Table 4 Summary of known D6R-silicates and higher silsesquioxanes (RSiOt.s), R n Reference H CH.~ C6H~.~ CsH~7 i-C,~H to C~H s C~H~

12,14,16 12 12,20 12,24 18,28 12,22,24 12

[52,161l [5,157l [16] [17] [! 7] [6A7,162] [112]

the D4R (SisO~'} anion, the D5R {SttoO:.~ ' to- ) anion is labile in aqueous solution. However. like the D4R compound, the presence of DMSO has a stabilizing effect on the D5R {Si IOV25 nm~} anion.

5. Double-six ring (D6R) {SinO~a} cages and higher silsesquioxanes For dodecasilse~uioxane two unstrained geometrical isomers are possible (Fig. 21) and the possibilities for isomeric complexity increase sharply for the higher sil~uioxanes (RSiO as)t4 o.,. Reactions of these compounds are sparse. They are prepared by hydrolysis of their corresponding trifunctional monomers RSiCI3 (Y CI. CH.~COO). Hydrogenation of Si~,OlsPh,: in an autoclave at 34arm ~nd 200°(7 yields 100% conversion of Si,~O,~Ph!, to Si,~O,~(¢'°C~H**),~ (39)[I 15]. Table 4 summarizes known D6Rosilicates and higher silse~uioxanes. Silylation of Hl:Sil~Ol~ (40) and the two (D.~h- and C~v-) isomers of Hi4Sil,,O~l (41) using the (CH :~)~NO/CISi(CH ~)~ reagent yields the new trimeth. ylsilylated cages [Sim,Ol~](OSiMe~)l~ (42), D , : [Sit4Oal](OSiMe~)t,~ (43) and C~-[Si,.40:,[(OSiMe~)l,t (44). The crystal structures of the two isomers ~of Hl.,Sil~O~l (411) and (41b) are shown it) Fig. 22 [161]. Fi~. ~ , e~st~| stature of D~-[Si~O~]{OSiMe~)~ 14111 111 and C~ 4S*,,O~, ~OSiMe~ ),~ 141b) {b),

6. Silseuluioxanes with an extra vertex The frameworks of silsesquioxanes will) an extra vertex as illustrated in Fig. 23 have beet) obtained as side products in the synthesis of other silsesquioxanes.

1

i0>

0--5; o

N

Fig, 23, Sil~,~uiox~s with an exwa ~'¢nex,

~

oa

-$i

RSIX:

Fig. 24. Model of the silylation of a silica surface with RSiX ~.

P.G. liarriso, / Journal ¢!f Orgmmmetallic Chenffsto" 542 { 1997) 141-183

i 65

,

j,

(21) (,IS)

(46)

R - c-CO'It t Scheme 23.

However, they appear to be as yet totally uninvestigated

chemically attaching organic groups to the °~,rface silanol groups, most commonly by silylation using methoxy- or chlorosilanes (Fig. 24). Despite its apparent simplicity, the silylation of silica is an enormously complex process. Silylation of incompletely condensed silsesquioxanes such as (21), however, aftbrds a suitably simplified model system. Examples of compounds (45) and (46) derived from the silylation of (21) [I 32,170] are illustrated in Scheme 23, where the analogy with the reactivity at silica surfaces may be seen. Although the preparation of (21) is straightforward, it requires an inconveniently long gestation period (36weeks) before synthetically useful quantities of the trisilanol compound (21) can be obtained. A facile synthesis of new incompletely condensed polyhedral oligosilsesquioxanes by hydrolysing (c-C ~H,~)SiCI~ and (c°C 7H ~:~)SiCI~, respectively, in an H ,O~acetone envio ronment affords compounds (21a and~21b) in a much shorter time (7 days) [171] (Scheme 24). In the case of R ~ ('-C 7H t~, the incompletely condensed silsesquioxo axe with two vacant vertices (47) was also obtained in ca. 40% yield and can be extracted from (21b) with

[551. 7. Incompletely condensed silsesquioxanes The incompletely condensed silsesquioxane (21) can be prepared in surprisingly good yields by the hydrolytic condensation of (c-C~,Hjt)SiCI 3 [132,163], and has been used as a molecular model for bydroxylated silica surfaces, as ligands in homogeneous analogues of silica-supported catalysts, and as 'building blocks' for the systematic construction of structurally well-defined S i / O / M clusters. Surface-modified silicas [132] are an important class of m~tterials wllicll have been used extensively as stationary phases in chn)matography [164.165] and have fimnd many other useful applications such as ion toilet° tion [165], heterogeneous hydrogenation catalysts [167], inorganic polymer fillers [168], and drug delivery agents [169]. For all these applications, the surface properties of the silica can be tailored by modifying the surthce by

R

l Acetone/HsO RSlCls

+ PH R

R (21)

R/

(47) 21a: R = c-C5H9 21 : R = c-C6Hll 21b: R = c.C7H13 Scheme 24.

~t

R ;'; q'CTHIs

166

P.G. lta rri,~n / Journal of OrRanometallic Chemistn, 542 ( ! 997) 14I- 183

J

MeMOs

R

R

(48) ~

R ac.CSl~

" (53)

M=Sn

R

- " =CHs

44b: R u 0,C41'111 M " h,GO

CH2C6Hs)s

4k: R ~ (:~CTHlS M-Sn

"\ fle4.C~l,

fl

R

\

~c.,

R

R. 1

CpT~ t0,

\

.

R

R

(So) MmSn (30

R R

(s2) ~hen~ 25,

P, G, Harrison / Joun~al of Organometallic ChemLv~t3,"542 (1997) 14 ! - 183

Ill s(I It

R,_

t, p"|' Os~(COhoLz +

,o.,

o

..

o

I ...'s,--o.~--s:' t

.. ~'

,,

a

i 67

t

o.

-H,O

",

. ..o

R

R"~I~(')'~"~ R

- 2 L

,,.s,-.

o

\

,,

I ..s,--n'l--si. 2.,, ~

~t =o,"

.

~

.o

!..6

it, a , ~ O ~ a ~

a

R

Scheme 26.

CH,CI 2. This silsesquioxane (47) should be capable of accommodating two hetero metal ions forming a completely condensed silsesquioxane framework which would be useful as a model to study bimetallic silicasupported catalysts, e.g. surface dichromates or dimolybdates [ 171]. Both (21) and (45) under cyclodehydration reactions readily to afford (21a) and (45a) [172]. Such base-assisted reactions are most commonly observed for high valent metal halides and rationalize the failure of (21) and its analogues to react cleanly with reagents such as MoO.~C!2 and CrO.~Cl~. Lanthanide complexes of (21) have also been synthesized and structurally characterized [ 173].

Io R . O'CiHII

1~1~1

results in the synthesis of compound (49), and a large number of reactions of (49) are shown in Scheme 25 leading to the formation of (50), (51), and (52) [174]. In 1986 Feher [175] synthesized compound (53) incorporating Zrcp [cp--cyclopentadiene] into the cube-like silsesquioxane framework as a model of silica-supported catalysts by reaction of (21) with (CsMes)Zr(CH~C~Hs) 3 (Scheme 25). From the ease with which zirconium, one of the larger transition metals [176], can be incorporated into the cubane siloxane framework, Feher suggested that it would be possible to synthesize more such materials with a wide variety of different metals. Reaction of Os3(CO)I~,(CsH~4)2 with the trisilanol compound (21) yields compound (54) [177] (Scheme 26), and the reaction between trisilanol (21) and Mo;[(O'Bu)6 or W2(OtBu)6 results in the formation of [(c-Ct~Hil)7Si7Oi2]:Mo: (55) (Scheme 27) or [(c-C6Hll):Si7Ol:]:W,(/z-HXOtBu) (56) respectively [178]. Compound (55) reacts with NO with cleavage of the Mo~Mo triple bond to give the NO-

(4Sa)

f.

The vacant vertex in compound (21) may be filled with a variety of hetero elements. For example. (21). (21a). (21b) react with MeMCI3 (M = Ge. Sn) to give the Main Group 14 capped hetero cage compound~ (48a. 48b. 48e) [132.171] (Scheme 25). Reaction of (21) with excess of pentamethylantimony (CH~)~Sb

(21)

2 NO

,o--"~=,, ",...

°'°! (58a)

oN

\/=

\ =\/jHo

t: (58b)

Fig. 25. The two isomers o1' the molybdenum complexes (58a) and

(58b).

°0

(55)

(57) Scheme 27.

P.G, H, rri~ot,t/ Joto'nal t~"Organonwtallw Chtmt.~tty 542 11997~ 14 I.-183

l~ R

R

{45)

R.,~.J~,,

(sg)

Scheme 28. complex (57) (Scheme 27). Like anany other Mo and W complexes [179.180], the two isomers (58a and 581)) (Fig. 25) catalyse rapidly the metathesis of olefins. Compound (45). which can be obtained by monosilylalion of (21) (see Scheme 25). yields the chromate compound (59) with CrO~ [181] (Scheme 28). The reaction of (21) with VOCI~/NH~. ("PrO).NO or (Me~SiCH:)~VO [182,183] affords compound (60) and a dimer (61), As shown in Scheme 29, a monomer/dimer equilibrium exists, The dimerization of (M)) is enthalpy favoured, but at concentrations less than lOmM and/or temperatures greater than 25 °C, the major (> 95%) V.containing species is the monomer (60), The monomeric vanadium compound (61)) and the chromium compound (59) serve as models of silica-sup~ ported vanadium or chromium catalysts tbr alkene polyo meri~ation, Compound (60) with AIMe~ as coocataly~t shows considerable act.i~i!y ~ow,~rds ethylene polymerization [183], In order to examine tile table of aluo minium durin$ ethylene polymerizatio, the ltllXed vana~ dium/alumii|ium compound~ (62a~e) sho~vn in Fig, 20 have ~ e n prepared [184], Inde.,d, conll~,~und (62c) sho~'~ similar activity toward~ ethyle.¢ pol~ ,nertzatio. a~ observed tb¢ the (~)/AIMe~ catalyst system, ill addition, several sil~esquiosane Ti, V, and Cr COmo plexe~ (63) and ( ~ f ) have been synthesized h~ which two cag¢~ are linked together by two transitio, metal atoms as shown in Scheme 30 [185,186], Man:~ investigatio.s have been carried out on the

synthesis of anionic A l / S i / O frameworks [187-189]. Polyhedral aluminosilsesquioxanes are potentially useful models for such alunni.osilicates. Scheme 31 gives a summary of tile synthetic routes used to obtain unto,tic polyhedral aluminosilsesquioxanes (65), (66), (67), (68), and (69). In addition, a boron silsesquioxane dinner (70) ix also known [190]. and its synthesis ix illustrated in Scheme 32.

8. Metal-rich siisesq.ioxanes Several metal-rich cage silicaves ilavc been reported. which may also serve as models lk>r similar silica-supported transition metal oxide catalyst.,,~ Tile reaction of 'BuSi(OH)a with Re:()7 in a 2:1 ratio leads |o the Re°oxide Ikmr ring silicate (71)j 191,192]: 4'BuSiit)H)~ + 2Re:qt) .....,

,-4

TIIc cD~lal ,,,truclmc of (71} (l°ig~ ?/! ¢xhd~lls a ~llicon°oxy~el| ei~l|lqaca~bcl*ed ti. B, Th~ IOul" siticon atolils of the rh| B are each I~ulld by all (ReO a ) al|d a 'Bu group. All the (ReO~) aud 'Bu ~roups ale iu mutually ~'is co.ligutatio.s, A simple sy.thetic route {Scheme 33). using silicon colllpound~ t72a) and (72b) vdth bulky organic glx~t|ps

R :O--m~f._ A

.~,..T/O 111

(21)

,tt t )

?1

iil-.l.j..O.i

(6L)

(60) Sc|~cmc2t).

P.G. Harrison/.hm,~al ofOrganometallit" Chemisu3" 542 ¢1997~ 141-1¢¢3

(62a)

(62b)

the compounds are obtained in excellent yields (up to 70% after recrystallization) and have an identical threelayered structure shown in Fig. 28(a). The middle layer, formed by six metal atoms, is connected via O-atoms to the Si-atoms of the two outer laying six-membered siloxane rings. One CI- is accommodated within the cage. To compensate for the negative charge, an Na" ion is situated in the outer sphere of the cage near the silsesquioxane 'crown'. The best characterized representative is (PhSiO~..~)L~(CuO).~(NaOo~).~ (74) whose crystal structure exhibits two four-membered {Cu:O.~}, tour six-membered {Si2CuO3}, and two 12-membered {Si~CuO~,} rings (Fig. 28(b)).

(62c)

Fig, 26. Examples of mixed vanadium-almninium compounds (62ac),

affords several cubane-metallasilsesquioxanes (73) with a high titanium content [193]. Aluminium analogues of (73) are also known [194], and can be obtained by reaction of a silanetriol with an organoaluminium compound (Scheme 34(a)). in addition to these neutral cages, the similar anionic cage has also been prepared by employing Na[Et_,AIH, ] (Scheme 34(b)). Both types of cage have been characterized by X-ray crystallography. The sodium salt, PhSiOONa, reacts readily with transition metal chlorides to afford new spherosilicate materials the compositions of which depend upon the ratio of reactants [ 195]:

9, Reactions and applications of silicate cages

Polyoligoorganyisilsesquioxanes are hiehly stable both towards thc:molysis and the action of nucleophilic and electrophilic agents compared to analogous linear, cyclic and polycyclic oligosiloxanes. This is due to the rigid Si-O framework and shorter Si-O bond distances in oligo(organylsiisesquioxanes), in polyhedral oligomers with a strained structure, e.g. in tetra+ and hexasilsesquioxanes, the Si-O-Si units are tie;wed much more readily. The relative stability of ,,ilsc,,tlUiOX° anes with n = 6. 8. I0. 12. and 14 is presumed to be mainly determined by the degree of distortion of the Si=O-Si angle and the nature of the intramolecular interaction of the atoms forming the ~mgles concerned. X°ray diffraction has shown that in oligo.~ilsequioxanes the O=Si-O angles are all close to the tetrahedral mlglc (ca. 1()().5'~). Fairly high aFtgular .~[|°aili~ may t,ccur in

12/r(PhSiOONa) ,, + 6MCI:

( PhSiO~ s)~z(MO). + 12NaCI

(4)

(n ~ 3 - 6 . M ~ Cu, Ni, Mn. Co. Cd)

12/.(PhSiOONa) ,, + 4MCI,

(5)

~, (PhSiO, ~),:(MO),(NaO. ~)~ + NaCI

169

(n ~ 3=6, M ~ Cu. Ni. Mn, Cd)For the forlller reaction

I•II..@. IIISI

I

I

II

...#0s,..~.o- ~-;-a T",-_ 0 e

t?

0.2

!;

/

X

IR

A

1 R = ¢.C,Ht

R

MX: EIIN CoHt

o'I :" ! ii u ,,~o- '/

I1~, i O

,,-,.o,-o, ~o' o-

,**...

o o

III/

~11

c,.s.

T/m/

X

II

( 6 4 ) II

(63)

(21)

Scheme 30.

t l t ,, TI

M t t 11'1

b

v

v

¢

TI

V

d o f

v v Cr

?l Cr Cr

170

P.G. Harrison/ Journal of OrganometaUic Chem~stn'542 (1997) 141-183

ll~~r~l~~--'-"

=

.

. .

.

.

• ......

.........

=

o~

~

o2 It

'

=

o',0

gt-I-II,.-o-_l,-li,~_ O_* M~ m

1211 16S)

R s il.4~Hl!

t661

167)

~ .so©oo i

gl•-.0

i

i~

_

jil.,,,V

•

"X"

r" :

m

168) (22)

a Y = SiMe3 b =SnMe3 e = SbMe4

(69)

Scheme 3I.

hex~il~uioxmles which accounts for the lower stabil+ iqt of Ibis skeleton [55]. Infrared spectroscopic data reveals that the thermolyo sis of ocla(methylsilsesquioxane) leads to lhe formation of a non+volatile solid phase similar to SiO~. In a helium atmosphere (CH~)~$i,O~ sublimes almost complelely in the temperature range 150=300+C with+ oul marked decompo~ilion, However, in air it decomIX~es slowly at 270+C, and at ;500+C 66% is oxidized to SaO~ [196], Octa(ethylsilsesquioxane) decompo~s slowly at its melting point (285 +C), and octa(propylsil~luioxane) decomposes slowly in air at 150°C, In contrast, octaisopropylsilsescluioxane does not change under the same conditions (150°C) [40], Octa(vinylsil°

sesquioxane) begins to react with oxygen at 170°C exothermically, and at 550°C it is completely oxidized to SiO+. In vacuo octa(vinylsilsesquioxane) de,:omposes at 290~300°C fi)rming a nonovolatile solid [197]. Octa(phenylsil~squioxane) is the most heat resistant oligosilsesquioxane known, and does not change when heated in air to its melting temperature (SIX)+C) [47]. The Si-O=Si group in oligoorganylsilsequioxanes is stable toward sulphuric acid, but oligo(hydro+ silsesquioxanes are susceptible to attack by sulphuric acid [40,52]. However, octa(hydro- and organyl.. silsesquioxanes) are stable towards both concentrated hydrochloric and acetic acids [6A0,52], Octa(phenyl+ silse~uioxane) polymerizes under the action of sul-

,i.e., I l,,~.

{2t) Scheme 32,

P.G. Hare'son / Journal of OrgammwtallicChemislo"542 (199D 141-183

171

branched polyphenylsilsesquioxanes. In the presence of catalytic amounts of alcoholic alkalis, octavinylsilsesquioxane in chloroform solution undergoes rearrangement to higher silsesquioxanes [204] such as (CH 2 =CH)noSinoOl5 and (CH 2 =CH)n2Sil2Ois. Due to their interesting properties, SiO 2 films have been used for a variety of applications. In semiconductor technology, SiO 2 films are used for a wide range of applications, for example as masks against impurity diffusion, electrical isolation between contacts and substrates, protection from environmental degradation, as

phuric acid and reacts with fuming nia'ic acid [40]. The siloxane bonds in oligo(organylsilsesquioxanes) are cleaved when heated with alkali hydroxides. Heating of oligosilsesquioxanes with a catalytic amount of alkali at 200-250°(2 leads to formation of ladder polymers (Scheme 35) [47,162,198-203]. These polymers are very heat resistant, e.g. polyphenylsilsesquioxane starts to decompose above 600°(2. At 900°(2 only loss of the phenyl groups occurs, with the Si-O framework of the molecule ~maining intact [200]. This behaviour distinguishes sharply the above polymers from linear and

!

si(og)=

/ (72a)

(72b) R'

4 (72a)or (72b) + Ti(O/Pr)4

.... - 12 HO~r

03)

03) (a)

R = 2,4,6-t-BusCeH20 R' = O./.Pr

4 (72b) + TiCI(O/Pr)s

+ 4 NEt3

(73)

. 4 NEt3HCI

(b)

.8 HO~r

R = 2,6-Pr=CeHsn(SiMes) R' = O-/.Pr

(73)

4 tBuSi(OSnM=3)3 ÷ 4 MeCsH4TiCI3

.12 Me3SnCl

(=)

R = t-Bu R' = MeCsH4

Schem~ 33.

172

P.G. H, rris,m / J, umal ~" Organomemllic Chemislo' 542 ¢ i ~¢)7) 141-183

0

0

!

12 r,

ii

fY T T ,1,/,!1!

C141~

-

7 :$

6

<

7

Fig. 28. Crystal structures of (PhSiOt.5)t,.(MnO) ~ (a) and (PhSiO|..~)I 2(CuO)4(NaOo .~)-* (b).

Fig, 27. Crystal structureof (71). Most of this work involves the use of tetraethoxysilane or ethyltriethoxysilane to prepare the low temperature dielectric films [207]. The use of cage-like organosiloxanes formula (RSiOt s) , (R = organic substituent, i.e. -CH.e. -C:H.~, - C 2 H s o r - C 3 H 5 and n -- 6, 8, 10, or 12) have also been investigated [54,209]. These compounds have a higher C:Si ratio than tetraethoxysilane, and all the compounds can be synthesized by hydrolysis of organyltrichlorosilane in an alcoholic medium. Octavinylsiimsquioxane is the most sensitive to radiation, The process used for dielectric film

anti-reflective coatings, high temperature optical filters, and much more [205,206]. In the deposition of dielectric, ~mi-conducting and metallic layers, volatile element-organic compounds (E.O.C.s) of various types have been utilized [207.208]. The E.O.C.s used to prepare ~ l i d films must satisfy two main requirements: a sufficiently high vapour pressure, and they need to contain a minimal relative content of non-depositing elements (particularly organic con,~tituents),

(a)

R'.$:.~O~.

~eT'

,. R 2

;~1

,siM,,

J,!, - 4 H~, - 8 C4Hto

\0/

Rt'Si~o~AI.R2

(2,6-iPr~C,H~XSiMe~)N, R ~ . 1,4-diox-

Ra =

(b)

e

m

R'.~i.~O-...^l. Et ~,.A(O.~o~ ,.0/\ . S H~, - 4

C:H~

-

!

~a(~h

?,/o

R~.Si~,~ 0.~. AI~ Et ms.

R I = (2,6-,Yr2CGH~XSiMeDN, scheme 34,

m

P.G. H~rri,~on/ .t,~,~ml o.¢Orgammu*tutlir Ct~emis~rv 542 ~1907~ 14 I - h~¢~

h t

173

-~-oc~

(o)

I, ....i iriln Scheme 35.

fi~brication using this compound consists of two consecutive steps: vacuum condensation of octavinyisilsesquioxane film; plasma treatment of the film. Using a nickel cell, films of polyvinylsilsesquioxanes have been obtained after a polymerization time of 4 min with optimum condensation conditions achieved fi)r 500nm of fihn at I Pa. current density ( 5 - 7 ) × I0 -~ A cm- ~ and substrate potential of - 50 V. Investigations show that. by such treatment, the polymerization of octavinylsilsesquioxanes leads to films exhibiting different physicochemicai and electrophysical characteristics than for films derived from tetraethoxysilane. The film deposition temperature of octavinylsilsesquioxane (25 °C) is significantly lower than that of tetraethoxysio lane at 300'~C. Generally the dielectric properties of the amorphous films prepared by glow discharge treatment of polycrystalline octavinylsilsesquioxane at room temperature have better dielectric properties than SiO, or polymer films obtained from tetr~ethoxysilane. Studies of [CH~ ==CH-CH:]~[SisO~.] have been shown to have also excellent film forming properties [210]. Iolydridospherosiloxanes are quite intriguing as precursors because the Si~O-Si structure of silica exists it) the polyhedral ¢~lges. Therefore. i| is !~asible that ther~ real decomposition oI' the hydridosphcrosilox,ne,~ it) oxygen to form siliea involves simple a~:tivation of the Si-H bond and subsequent linkage of the polyhedra by oxygen bridges. Additionally. since the size of the polyhedral silsesquioxanes varies, it is possible that the films deposited will vary in porosity, density, and other ,'elated properties as long as the polyhedra are indeed preserved during the CVD process. The mechanism of thermal decomposition of H sSisO,, in an oxygen atmosphere may be described by the overall equation

H,SisOi, + 40~ ~ 8SiO: + 4H,O

.(6)

Characterization of the SiO: films grown by CVD Rein H xSisOt, precursors showed that the SiO: structure of these films were quite simihu" to those deposited fi'om an HxSisOi,/H.)Sit.Ol.~ mixture [211]. X-ray diffi'action shows that the fihns are amorphous with good specular reflectance, and appear to have smooth surfaces as seen by SEM. Auger depth profiling shows there is tin)fern! distribution el" oxygen and silicon throughout the bulk of the film. in 1985 Day et al. attempted to obtain ceramic materials by the hydrolysis of SisOI~(OCH~)8. :"St

--~tl -OH I

L___ (b)

'

I

•

~

,

-IO0 0

,

l

. . . .

,

-tO0 g

. . . .

,,,tOl.O

i

i~,

•

-IOI.O

I~ (etSI) Fig. 29. >'St NMR spectra recorded during the hydrolysi~ of Si sOt ~(()CH ~)~.

NMR studies (Fig. 29) during hydrolysis, however. showed that significant side reactions also occur, indi° caring that tile cubane cage structure did not rem~,in intact [99]. L~,er Agaskar tried to obtain microporous macromolecular materials by reacting the decavinylo D5R cage [Si I,~O,~](SiMe,CH ~CH ~)t,, with ~, ~toichioo metric amottnt of the bifunction|,l sil~lne [HSi(CH~):CrHa)O.~]~ in the presence of PtCI,(C,H~CN): as catalyst. The resultnnt ~olid is described as a hard. clear resilient, thermally stable (up to 350°C) material which absorbs > 50~h, by weight of tetrahydrofltran. :'JSi CP/MAS confirmed that the struco tural integrity of the spherosilicate cores was retained° XRD traces, however, showed that the polymer (75) is

Step i

~,m.__~

.ph.o.,.~tu

J

Crou.llnklng Step ll

[OrRanolJthlcmacromoleculsr materl~ 7 Pyrolylds

Step !I! ~i ii NImocompositecer~a J I Dlffm'enUal Step IV Fig. 30. Illustration of' the controlled synthesi~ rottte for the prep;,r~riot1 of" inicropot'ou,~ ct~r:ittli¢~ fret11 fullctiotlalized ~phtm~ilicat¢~,

174

P.G. H~irri+on/ Journ+d +y+Or~anomet+#lic ~hemi,s+tO•~42 ( I ~ 7 ) 74 I - 183 o 0

(e)

o,°.o 0,%o

// I

I

I

,

6

~

Fig, 31. Solid slate ~Si NMR spectra recorded before (a) and after (b) leaching of the pyroly.'.+dpolymer (75) with HF.

~~s. Unfortunately. low temperature argon ad+on measurements gave no evidence of microporosit), [212]. Microporosity with high surface area could. however, be achieved by the controlled synthesis route

illustrated in Fig. 30. using [SisO+,o](SiMe++CH=CH++))s as the starting precursor. After leaching out the SiO.+-rich phase of the pyrolysv~! polymer ('I$) with hydrofluoric acid. the intensity of the resonance in the solid state 2~)Si

Y

,, ~'-

~,+++

+++,"

"~+

(?6) /

40

20

0

-20

..40 .60 ppm

~

-100 -120--

FiB, 35, S~td +~te +'~Si NMR ,,+pectin of polymer (?6) obtained by Ihe Pt.+al~13~+d hydm,silylalion of [Si~O:ol(SiMe:H) . will)

175

P,G, Harrison / Joun~al o,f Or,p,anonu~zc81ic Chemis~o" 542 q 1997) 141-183

(77) . o ~ _ . ~ . ~ "~

~-o-~"

,

-..~-~L-'~, '°

.o~-o---~,~

"QSt_--O--S~" , ~,

~~:~-~

~

I ~-~-,:.,

~ - ?I - ~

.,_:_o_~o_:_~_~_~__ ~-,~-~

gfg

-~"~ .~-o-~

C~-, .o.~-~

..

.o1"-° %

.

(79)

( 7 8)

Fig. 33, Polymers with chain-like (??), cyclic (78) or branched ('/9) bridges between the D4R cages.

, L

L

: ~...~ N- ,,,

INIOL,

,,o

L

,%

L,

O-

,,,o,

\

~.

,

......~ o _ ~ O - - - o , , .

,.

(81) (so)

.

20 0

40

-80

-120

.

.

.

.

J

.

.

.

.

.

v-

20 0

,

•

. .

i~

.

-40

.

•

.

.

-80

_

•

_

,

_

-120

[ppml Fig. 34. Solid state "~Si NMR spectra of polymers (80) a,d (!'~).

176

t'.G, Hm+ris+em/,h,l.'n+ll +J.fOr,+ptln+mler+llli(, Chetnistrv 542 ¢i 0 9 7 / I 4 1 - I<<+,"

FI

It0

0

p,

pH,,

4~,13

0

0

c H r ctt-c--o--- CH=-- C11=--CHIn--Ti--.O¢I.I= 0

(:H=

+ H20 (ama)

TI

,I"2 01,I= Z !

.OH=, 0

X=~'~il=-qDi

0,~ ~¢~ Of'It I¢I~I¢I--0 -+0.I,- OH+=..OH, +

(83)

phi o,~ ~i,-- CHi.=CHi (84) IP.MI, + H~O

",~-~LL

+

f T - = ' - = ; ~ ~°='

P.G, Harrison I lounia! o,f Oit~anm,emllic ChemistO" 542 f 1997l 141-183

information about the influence of the bridge length on the porosity was obtained using two additional polymers, (80) and (81) [156], whose bridging length is similar to the microporous polymer (76): these were prepared by the reaction of [SisO2o](SiMe2H) s with [SigO_~,](SiMe2CH2CH=CH2)) s and SigOl2H 8 with [SisO2o](SiMe2CH=CH2))8 respectively, and contain seven-membered (80) and four-membered (81) bridges. However, unlike polymer (76) polymers (80) and (81) do not exhibit significant specific surface areas. Fig. 34 shows the solid state 29Si NMR of polymers (80) and (81). Platinum-catalysed hydrosilylation of a mixed hydro-vinyl-silylated D4R-silicate results in the formation of the polymer (82) in which the cubane silicate cage units are preserved and which has a specific surface area of 200 m: g- I [ 156]. Silanol-rich polymers (83) and (84) have been obtained via hydrolysis/condensation reactions of (328) and (32b) (Scheme 36). The average number of unreacted silanol groups in both polymers is 8.8. Both polymers, however, showed a low surface area [153,154]. The solid state -~'~Si NMR of the obtained silanol-rich polymers (83) and (84) is shown in Fig. 35. New porous polymers have been synthesized from {Sit,Og} and {Si~tOl+} cage precursors and characterized by solid-state-"'+Si and -~'~C NMR. Preparative routes to polymers (84)-(88) are shown in Eqs. (7)-(! I). Two methods have been employed: (i) H+PlC~++calalysed hydrosilylation of vinyl-substituted cages, and (it) hydrolysis of bromosilane functions: the general structures of tile polyme/formed are illustrated schematically in the scheme. The hydrosilylation reactions were carried out in refluxing tohiene, gehition occurring after ca. 15 thin. Hydrolysis of the bronlosilanes was effecled by ice+water. After separating off and drying in vacuo, the polymers were isolated as insoluble glassy powders.

Ta

get, 120 rain

+°o'o-_

L_

Fig. 35. Solid state -"~SiNMR specm.n of the silanol-rich polymer (83). NMR assigned to [SiO4] is diminished relative to the other resonances (Fig. 31), The result after leaching is a microporous (pore size in nanometre range) ceramic material with high surface area (594 m -~g- ~) [213]. A similar microporous material has been obtained by Hoebbel et al. by the Pt-catalysed hydrosilylation of [SigO~o](SiMe.,H) ~ with [SisO_,o](SiMe_,CH=CH .,))~. The polymer (76) obtained is microporous with a specific surface area of about 300m-' g-~ without posttreatment. Solid state -'~Si NMR of polymer (76) (Fig. 32) shows that on average two functional groups remain unreacted [214]. Heat treatment of polymer (76) shows that it is stable in air up to 250°C. However, degradation occurs at 342 °C. Thermal oxidation of the methyl groups occurs at about 450°C, which finally leads to new Q units at the silicon (SiOo,5)4 with different degrees of polymerization [215]. In order to examine the influence of the bridge structure and bridge length on the porosity and structure, hybrid polymers have been synthesized with d i f i~l'¢at types of bridging molecule unit between cubane silicate cages [216]. Fig. 33 shows hybrid polymers with chain+like (77). cyclic (715) or brandied (79) bridges between the D4R cages. However, only low specific surthce areas of about 4-7 m ~ g= ~ were found. More

0.560 .99.02

.17.44

/

-1.24

!

--"'~

O--

,

~H+,

177

CH~

+

+o/

CH)

CH)

Hs

Hs

B~ B

(84)

i

I~

PoG. Harrkfon / Jim,vital t~ Or#~anomelallic Chemisl~" 542 (1997) 14#- #83 -0.113 /

:

i"1,,.,,I--.-

~ C

..

o.l:ll

f.ll(

.

..

3.688

-80.3tt

~._ \-" ~'c.=c.2 •O,IhlS

J'

i~.qoI

I

($s) 0.669 -18.04

~"

...........

"-~,~

/

o---"~!1

s~k

~1 ~

-109.29

/

i

~,

,~.~-o--~'-o,_/~/-'~":o

........

0

(86) 0,16,1

-I,Ill

(87)

0,3t9

P. G. Han'ison / Jou rmd or"Orgmumwmllic Chemistr3, 542 ¢i 997) 14 I - l&¢ -66,19

.

~

-..---3,10

_ (

n

,./

.' 3 ~ = * l - k _ I

.

/r,~

(

.

:,

179

..

/

{

.

(SS) H,O

2{Sit~O,)}(OSiMe:ar)~, 2~ (84)

(7)

4{Si~,O,)}](OSiMe2 H)~, + 3{Si~O,,}(CH=CH,)~

(s)

- , (as) H ,O

2{SisO,_~}(OSiMe.~Br)8 2, (86)

(9)

was presumed that the tetramethylammonium salts encapsulated in the polymer during the symhesis evaporated at higher temperatures leaving cavities which were responsible for the porosity [218]. The reaction of (Me~SnO)sSisOt, with PCI 3 in a benzene solution produces a gelatinous suspension via Eq. (12) within 15 min. 3( Me ~SnO)~Si ~Ol:

{Sit~Ot:l(OSiMe~H)s + (SisO,~}(CH=CH~)a

(s7) {Si~O,.,}H~ + {Si~O,,}(CH=CH~)~ --*(88)

+ 8PCI

(io) (ll)

Characterization of the structures of the polymer is most easily accomplished using solid-state--"~9C; o, and - 13 C NMR, and structures (84)-(88) are annotated with aopropriate chemical shift data (parts per million). Spectra for polymers (8[;) and (86) exhibit only very small resonances due to un,'eacted cage functions, indicating that ca. 89% and 95% respectively of the available cxocyclic functions (determined fi:om integrated peaks areas) participate in polymer linkage lbrmation, h! contrast, Sl~ctra for the other polymers show that the degree of polymerization is much less. and only ca. 60~68% of linkage formation occurs under the condi° tions employed. The two polymers derived from {Si00,)} cages, (83) aitd (84), as well as polymer (86) exhibit type li nitrogen adsorption isotherms with hysteresis on the desorption arm, behaviour characteristic of mesoporous materials. Specific surface areas of (84) and (85) arc relatively low at 12.2 m 2 g- t and 63.4 m "~g~ J respectively, but that of (86) is significantly higher (218.3m 2 g~ t). The adsorption isotherms of (87) and (88) arc both type I but also exhibit hysteresis indicative of mixed microporous-mesoporous character. Specific surface area values for these polymers arc 147.0 m" g~ t and 573.7m 2 g °t respectively, the latter being comparable to values found in zeolites and much higher than the highest previously reported for polymers derived from molecular (Si aO:~ ) cage precursors [217]. Silylation of [(CH.~)4N],[Si~O2o] with dichlorodimethylsilane gives a polymer with low specific surface area. However, heating at 300°C yields a porous material with a surface area of 300m 2 g-t. it

[Si~Ot,(OSnMe~)M_,,(OPcI:),,tO.~Pci)q]. (12)

- nMe~SnCI

n ~, 2 q + p

Additional heating afforded a thick slurry. Approximately 60-75% of the available tin was liberated as M%SnCI as indicated by the t H NMR spectrum. The BET surface area of this material measured by nitrogen adsorption is 500m 2 g-t with a pore volume and pore diameter calculated to be 0.3 cm ~g~ t and 24,~ respeco tively [219]. Lichtenhan and coworkers [220,221] emo ployed (89) as a bifunctional building block for the construction of linear silsesquioxane-siloxane copoly~ mers via silanolocentred reactions with bifunctional silane and siloxane precursors, rransition-metalo.cono taining silsesquioxane polymers have also been reported using a similar approach. ov I 0 .._8,i

HO ~, $I"

t$1,

O,,

I ,.,

0 0,~ I L

Cy

(89)

Further application of current interest Ibr polyhedral siisesquioxanes is in the production of liquid crystalline (LC) copolymers. These materials have been produced by hydrosilylative coupling of oligo(dimethyisiloxanes) with allyioxy-mesogens. The oligo(dimethyisiloxane)

I~

P.G. H~trrison/ Journal of Orgmmmetallk, Chemisto' 5421 i 997) 141-183

units serve as flexible spacers. The usefulness of LCinorganic copolymers to combine the superior properties of both components warrants further attention. In addition, polyhedral silsesquioxanes show excellent adhesion to a wide variety of surfaces, including dentin, and are also extremely abrasion resistant. Therefore, hybrid c o p o l y ~ of LC units and silsesquioxane units offer t ~ opportunity to create single phase copolymers that will ~ v e good-to-excellent adhesion, abrasion, resistance, and high processability.

References

{~] D,A, Armitage. Inorganic Rings and Cages. Edward Arnold. L~mdon, 1972. Baney. Chem. Rev.. 95 (1995) IJ,(~. Loy and K J, Shea. Chem. Rev.. 9~ (1995) 1431. Anddanov. The Methods of Elementorganic Chemistry, Nauka. Mo~ow. 1968. p, 589 [~l M.M, Sprung and F.O, G,en~ner, J. Am. Chem. S¢x:,, 77 (195~) 3996. [6] J.E, Brown and L.H. Vogt. J. Am, Chem, Soc.. 87 (1%~) 4313. [7l J.F. Brown, J. Am. Chem. See.. 87 (1965) 4317. [~] E, Wiberg and W. Simmler. Z. Anorg, AIIg. Chem.. 282 (19~S) 330. {9] P,D, Lickiss, S.A. Lister, A,D, Redhou.~ and C.J. Wi~ner, J. Chcm. S~. Chem. Commun.. ( 1991 ) 173, I~0] V.I, LavNnt'yev. Te~isydokl ill Repuhl. Konfcrents, Mokglykh uchenykh khimikov. Tallin. 1979, p. 87. [i~] V.I.L.r¢fll'yev, V,M, Keratin and G,G, Tr¢cr, Zh, Oh~hch, Khim. o';I (1981) 123, l~-~] V,i, Lavrcnt'yev lind V,t~, Kostrot~kii, I~', Sih, Old, Akad, Nauk SSSR S~r. ghim,, 1=~(Iq79l 14, li,q V.M. Keratin and V.I. Lavrent'ycv. Zh. Oh~h¢ll. ghim...~q (1989) 3?% !i41 M.M. Spftln~I {lnd F,O. Goenlher, J, Am. Chem. S~'. 77 ( 19~ t 39~), ll,q K.A. Andflaaov and B.A. lemailor. Zh. Ohsh,'It.ghim., 3~ ( 1 ~ ) 341, I~61 I(.A, Andtiallov alld B.A, Itmailor, J Or~anolnet, Chent,, 3 (1907) 43S, i17] K.A, At~lrianov and B,A. Izmailor. ZI~, Obshch. Khim,, 46 (1976) 329. F.J. Feher. D.A Ned.man and J.F. Wal~er. J. Am. Cl~eni.Sot.. 111(19t49) 1741. [~vl KA. Andrianov al~d V.A. (ginet~. It~. Akad, N0~ik SSSR Old, Khim Nauk, (lt)Sg) 4~), I~o1 K.A. Andriano~, and A.A. Zhdanov, Plastma,~sy, 7 (1962) 24, ~l'u.I. Smolin. Soy. Phys. Cryslallogr.. 15" (197(l) 23. [2.~ID. H ~ e ~ l and ~. Wicker. Z. Anorg. AIIg, Chem.. 4Oo (19731 148, I.~,q(3, En~lhardt, D, g¢igan, H, Jancke, D, th~:h~t and W, Wiekerr, Z A~or~, A11g, Cl~m, 41,~ (IqTS) I?, I?.41G, [%~lh~,It, N, Jancke, D, II~bI~I and W, Wieb:r, Z,

[21 R.H. [3] D.A. [4] K.A.

12H H.C. Ma|Mi~anm Z. Natud~m, cl'i.~Feil B:. 29 (1974)495. G~ild, B,M L~we and N.A McGilh, J, Chore, S~', CI~nt, Comntutl. (1974) ?~I, 12sl D, t t ~ l , G. Gaffe. G, Engdhurd! und A Till. Z. A~,~. Allg. Chem.. 4.';0 (I~7~)) ~.

[:?l R,O,

[29] D. Hoebbei, G. Garzo, G. Engelhardt, R. Ebert, E. Lippmaa and M. Alia, Z. Anon. Allg. Chem., 465 (1980) 15.

[3oi D. Hoebbel, A. Vargha, B. Falke and G. EngelhardL Z. Anorg, Alig. Chem., 521 (1985) 61.

[31] R.K. Harris and C.T.G, Knight, J. Mol, StrucL 78 (1982) 273, [321 KJ. Carell, A,F. Masters and K.G. Wilshier, Zeolites, 2 (1982) 244. [33] Yu,I. Smolin, Kristaliografia, 29 (1984) 712, [34] EJ J. Groenen, A.G,T.G, Kortbeck, M, Mackay and O. Sudo meijer, Zeolites, 6 (1986) 403, [351 D. Hoebbel, (]. Engelbert, A. Samoson, K. Ujszaszy and Yu.I. Smolin, Z. Anorg. Alig. Chem,, 552 (1987) 236. [36] G. Garzo, D. Hoebbel, A. Vargha and K. Ujszaszy, J. Chem, Soc. Dalton Trans., (! 984) (1857). [371 H. Behbehani, BJ. Brisdon, M,F. Mahon and K,C, Molloy, J. Organomet. Chem,, 469 (1994) 19. [3s] AJ. Ban), et al,, J. Am, Chem, Sot,, 77 (1955) 4248, [39] M,M. Sprung and F.O. Guemher, J. Am, Chem, S~:,, 77 (1955) 6045. [40] K. Oisson, Arkiv. Kemi,, 13 (1958) 30?. [41] D,W, Sc(~t, J. Am, Chem. Soc., 68 (1946) 356, [42] E, Wiberg and W. Simmler, Z, Anorg. Allg, Chem,, 283 (1956) 401. [43] V.i, Lavrent'yev, M.G. Voronkov and V,M. Kovrigin, Zh, Obshch. Khim.. 50 (1980) 382. [44] M,M. Sprung and F.O. Guenther, J. Polym, Sci,, 28 (1958) 17. [45] J.F. Brown, J. Am, Chem, S~x:,, ~7 (1965) 4317, [46l K, Olsson and C. Axcn, Arkiv. Kemi,, 22 (1964) 237, [47] J.F. Brown, L.H. Vogl and P.I. Prescott, J, Am. Cllem, S(~.'., 86 (l~R~4) 1120, [4X] K. Ois.~m and C. Gronwall, Arkiv, Kemi,, 17 (1961) 529, [491 K.A. Andfianov el al,, Zh. Ohshch, Khim., 49 (1~)78) 2692, [sol M,G. Vomnkov el al., Zh. Ohsheh, Khim., 4V (1979) 1522, [s~] R, Mtiller, R. Kohn¢ and S, Sliwinski, J. Prakt, Chem,, 9 (19~9) 71, Is.~]C,L, Frye and W.T, Collins, J. Am, Chem, S~'., V2 (1970)

is,~lP.A. A~a~k~m Inor~, Chem.. 30 ( 1~)91) ~TIIX. I,~I V.P. Ko~.'hko~. T.N. Ma(lvno~a a0d V.S I)~milo~k'h. Thin Solid Film~, t01 (19X~I) :169,

ls~] M G Voh~oko~ and V.I. Lm~nlyc~., Top. Curt. ('hem..

liJ~ (1982) I~, I~r,l W.M. Meier and D.H. Ohon, Alias oI'Z~olit¢ Slructm~ Tyff,,.~, Butt¢~oMh.~, London, 2,d cd,., 1988. [svl P, Bomhauser and G. Cal~al'em, Sl'~eclt~:him, Acta Part A:, 46 (1990) 1045. [s~l J, Baumanfl, R, Beer, G, ('al~afem and B, Waldeck, J, Pllys, Cheln.. 93 (1989) 2292. IS~l K. Tai No. D. Halt Bar ~lnd M. Shik Jhon. J. Ph)~s.Chenl.. ~¢} { 198(0 1772, I~] G, Caltal'~m and R, HotTnlann, J, Chei111,Sty, Dalton Trails,, (1~)1) ~)17, 161] G, Cal~,al'em and L, Fo~s, Heir, Chem, Acta, 7¢) (1{)87) 4fi5. I~,21G, Calealcm and L, For,~s, Heir, Cheln, At:la, 69 (1986) 873, I(,:qM. B~lilm:h. P. Bomhauser. H. BUr~.,yand G. Cal~a|em. SI~CII~hiln, Ac|a Pan A:, 47 (1~1) Ih27, l~i M, l}~Jllheh, P, BOl-llll{lll~@l" and G, Cal/a|~:llOi, SPIE J., 1575 i1992) ~88, l~si S.P. Massi¢. Jr.. J, Am. Chem. S~c.. 6/¢ (l~)46l 112S. [(~I R.N. Meals. J. Ain. Chen~. S~:.. ¢~'¢(Iq46) 1~80. [~nl J.~. Webster. J. Org. Chem.. 21 (1956) 1044. [6~1 A.J. Barry (l_~w Coming). US Patent 2626266. 16.10.19.1S. [6~1 A J. Btlri3,. L. D~pree aud D.E. H ~ k (Dew Cor,i,g). US Patc~t 2 5 I0~53, 6.6.194fl. I~ol A J. Barry. L. Depree am| I).E. H~a~k (I)ow Corningt. (IS Patent 2546330, 11,12,1948,

P.G. Hm-rison / . h m r m d of Orgammwlallic ('he.Us(O" 542 ~i997~ 141 - 183

17il J.W. Giikey and D.E. Hick, ACS Adv. Chem. Sen. 23 (19591 246. [721 J.W. Gilkey. L. Depree alld D.E. tlook ( l ~ w Coming). US Patent 2591668. 4.11.19.17. [731 E.A Chemysher and A.D. Petrify. Bull. Ac;.id. Sci. USSR Div. ChelU, Sci, (Engi,), (1956) 637. [741 G.H. Wagael aud P.W. Shal~e[ (Uniofl Carbide). US Paletlt - 775 oft6. -++.3.1954. ~" 175] A.J. Ba.rfy (Dow Coming). US Patent 2 572 302. I 6. I O. 1948. I761 E,P. Mecheev, A.L, Klebansky, G.N. Mellnova and K.K. Popkov, Plast. Massy, I ( 196 ! ) 19. [771 K, Kramer and A. Wright. Angew. Chem.. 74 ( 19621 468. 17S] R.A. Benkeser and D.I. FoL'ster. J. Am. Chem. She.. 74 119521 42011. [79] R.A. Benkeser and D.I. Forster. J. Am. Chem. She.. 74 11952) 5314. I80] R.A. Benkeser. H. l+andesmann and D.I. For..,ter..I. Am Chem. So¢,, 74 11952) 648. 18~! Ytl.I, Khudobin tad N.P. Kharitom~v, Bull. Acad.Sct. (ISSR Div. Chem. Sci. (Engl.), (19581 112. 1821 C. Friedel and A. Ladenburg. Justus Liebigs Ann. Chem.. 14.¢ 118671 118. [83] S. Nitsche, Angex~,. Chem.. h3 ( 1951 ) 4911. Is41 M.E. Havill. H. Joff¢ alld lt.W. Post. J. Organonlel. Che,u.. 13 11948) 280. !s51 S. Nttsche and E. Pit'stm (Wacker Chemic), Gern)an Patent 896414. 23.5.19511. I861 J.S. Peake. W,It. Nebergall and ~(,!', Chen. J. Am, Chem, Stw.. 74 11952) 1526. 187] B.N. Dolgov. N.P. KllarilollOV mid M.G, Voft)tikov. J. (ieti, Chem. USSR (En$l.). 24 ( 19541 1169. Issl P, Chevalier (Rh6ae Poulenc). DAS, 11131521) (F. Prior 13,7,1956). 1891 R, ('alas mid N. l)uffaul. (,,R, Acad. Sci,. 2.15 ( 19571 911h. l'ml R, ('alas. N. I)ltl'fatlt. I:~, Fraillet and J. I}onastre. C,R, Acad, St'i,, 251 (19hO) 2987, r,Nr Ilaeeh~oud IMC{'I+Ixx I:1ct'11'i¢). [!,% Pah:ltt 2727~4~O, 22,1,195], It21 M,(i, Vt.tmkot, ,I, {iVti, ('betli |!SSI~ (Engl.), 24 (1~1541 859 I',q N,I ~, Ol°h!x, RA, l+l~/~alktll,~+1+ S¢l+Wexa ;tlltl M ( I . Vtlt't.lko+,. ,I, l l c n Chclu, I'SSR (Iqigl,l, o¢2 l10h21 2S21+, I)R, I)¢{tti~ ~ttld 1', I +d+tlt+tt, ,I, Chel|t, ,%tw Cll¢llt ('tltltilttlll,+ ( 19.~41 31 fdJ~ I,W~ St~tlkills illtd it, W', ih~,~l, ,I, ()l'~{lllOlllel, Chetll, I; % (19bID

552, P.A, Mc('tt,,ker ~ttttl li,l++, r¢llly, ,I, All|, ('bolt, So~+',, 75 (ItJ5 +,) 1583. I',?1 A.C. Taylor mid B.V,(i, Walden. J. Atn, Chem. Soc. oo 19441 842. [,,sl YuJ, Khmlobin. Bull, Acad, Sci, USSR Div, Chem, Sci, (l~n~..,I,), 11956) 821 1991 V.W, Day. W.G. Klempcrer, V.V.M.'tinz and D,M. Millar. J, Am, ('bent, She.0 Io7 ( IO851 8.,6... +~ '~ i ltXq R, Beer, I+1, BUrg). G, (,altal'errl at|d I, Kmnber, J, Eh;tll'tUl Spcclros¢. Rcl, Phcatun,, 4.1 (1987) 12 i. !1otl F.J, Feher aml K,J. ~'ellcr. ()rga+|oni¢lallics. ') (199()) 2038, IIo21 P.A. A~2;,l~kdr. S)11111, 1{¢;.i~.'1 h|ol'~, Mcl, ()l'~, ('hcltl,, 21~ (IOt)l)) 41~3, D. MilhtL Phi). 'l+he.,,i,,. |!lli~'ct~ily o/ Illillois. (3~ampai~m. Urlmna, IL. 1987. !104] F,.!. Feh,:r and K,I, W¢llci. htt.'~,, ('hen(,, 30 (19911 SS(), riobj tl. Bilrgy arid ft. (';,Izaf~rri. Ilelv. Chitn. Acla, 73 (1991))fieF: i~(ml .l+l.+ Speier, ,I,A, \Vcbsler ;llld (;.ll, l}arttc~,0J, Anl, ('hem+ St)c, 79 119571 974 flit71 I). Ilerrcn, II, llLlrg)' and (i, (',llzalerri, I ieh. Chin( A¢la. 7.1 (19911 24. lit)s] I. Oiilna. in S. P.ttai illld Z. l,tappopt.l (¢ds.). "1"i1¢('hemish) of

! 81

O~anic Silicon Compounds, Wiley, Neg' York, 1989, Part H, pp, 1479-1526 and references cited therein, 11091 H, Biirov and G, Calzafem, J, Chmmatogr., 507 11990)481 lilol G, Calzaferri and R, imhof, J. Chem, S~'. Dalton Trans.. (1992) 3391, [ill] A,R, Bassindale and T.E Gentle, ], Mater. Chem., 3 (1993) 1319, [li21 FJ, Feher and T.A. Budzichogski. J. Organomet. Chem., 373 (1989) 153. [1131 R. Tacke, A. Lopez-Mms, W.S. Sheldrick and A. Seebald, Z. Anorg. Allg. Chem., 6/9 (1993) 347. [1141 FJ. Feller and T.A, Budzichowski. J, Organomet. Chem., 379 (1989) 33. Ill5] V.I. Larent'yev, V.M. Kovrigin and G.G. Treer, Zh. Obshch. Khim., 51 (1981) 123. [I 161 M.G, Voronkov et al.. Dokl, Akad. Nauk SSSR, 58 (198 +~ 642, liiTl K,A, Andrianov and N.N. Makarova, lzv. Akad. Nauk SSSR Set'. Khim,, (1967) 1381. 11181 K.A. Andrianov, V.S. Tikhonov and G,P, Makhneva, lzv. Akad. Nauk SSSR Ser. Khim., (1973) 956. I119] N.N. Makarova, Ref, Kand. Dis.s, Moskva INEOS, 1972. [I.'o] T.N. Martynova and T.l, Chupakina, J. Organomet. Chem. ,745 (1988) I I. ll2il G. Calzaferri, I). Herren and R, lmhof, Heh. Chim. Acla A, 74 (1991) 1278. [1221 J.F. Harrod and A.J. Chalk, J. Am. Chem. Sot,, 87 (1965) 1133. li231 B.J, Agletl and J M. Campbell, J. Chem. Soc. Chem. Comtlllttl,, (1965) 217. I1241 T.D. Tilley, in S. Patai and Z. Ral)poport (eds.), Silicon=Helel't~alOtll Bond, Wile). Ne~ York. 1991, Chapter 9, pp. 245+ 307. LI25] C.S, Cundy. B.M, Kmg,,,ton and M,F. Lappent Adv, ()t'b,anomet. Cht:rn,, I I (1973) 253. 112hi I+,1t, Sommer m+d j,E, l+>+ons. J, A m ChenL She,, 9o ( I + h S ) 4197,

112'7! G. C;tl/:dem. R, llilhtff a,d K.W. T(hm+t,t,,. ,l+ ('hem. S+~' i)allon ]'ran,~,. (1993) 374 I. 112Sl P,G I,larrisoll alld R K+tn,wngies,,¢r, ,1, Ch¢tlt, S,~P Chel., COiltiiilllt,, I lO0h) 20t+5, l12ql W A Ih:rrm+mn a,.I I! Vu~s. I (hgmtmn¢l Chore. ~' (10851 47. [1311] (), Kulm aml M, liit~mgne. ,I. ()r/~+t.otlt¢! Cllem,, I0 (19h7) 137, l l31J F J, ig.,Jl+t, K,,I, Waller and ,I,J, Schxb,al+), ()l'~al|Otl|¢lalJic~, 14 (199~) 2(X)9, 11321 F J+ Feher+ I).A. Ntdwtltlttl arid ,I,F, Walz¢l', ,l, AIII, Ch¢til, Silt,, I l l 119891 1741, [133] I), lh+ebb¢l and W, Wicker. Z, Atlotg, AIIg, Chem,, 3?¢4 11971) 43, 1134] W,I, Mokejcwil arid N,I, (]oltiv+aslikov, Dilkl+ Akild. Nat(k, SSSR [Bet', Aead, Wi,,,s, tidSSR], 167 (l%f+) 113, 11351 t', Tliilo, W, Wicker ittltl H, Strait Z, Atlot+[L,',,lit+,Ch¢tlt .... +40 (19t~5) 2 h i [13h] W, WK'ker mid !) Ih.:blwl. Z ,&tt,~r~, AIIg, ('llem...¢hO 119h9) 139. 11371 !), I-Ioel~hel. rhesis. lluml~oldt I lniver,,,ilil|, Berhn. 1%9, 11,:~8] Yu,I, Stuolin Yu,F, Shcpclev and I,K+ Ilutikoxa, Soy Phy~ Crvslallt~gr.. 17 119721 I0, I I.~01 11, Iloebbcl and W, Wicket', Z, Chem,, 11972) 2tin, 1140] D,l,lt~etfl}el. W, Wicket+. J,I, SmolirL J,F, Sepelev ;md R, i'mnc~+. .+.+ Z, Anorg, AIIg, ('h¢tlL, 423 I It)Th) .++5, 11411 .6.1, SIiwlitl. JF, Sepetev, R, P+mles, I) IIoebbel m.! W, Wicker, Krist.llo+grafia. 20 11975) 5h7, 11421 M. Wiebke and D. i+luebl~el. J, Chem, St~c. l)alltm 'l'ratl,,+,. ( 19921 245 I.

182

P.G. Harrison / Journal ~' Organometallic Chemistr3"542 (1997) 14I- 183

[~4~1 M. Wiebke, J. Emmer and J. Fel~he. J. Chem. Sac. Chem. Commun.. (1993) 16(¼. [I,14] M. Wiebke and J. Emmet, J. Chem, Soc. Chem. Commun., (1994) 2079. [1451 EJ.J. Gmenen. A.G.T.G. Konbeck. M. Mackey and O. Sudme~r, Zeolite,s, 6 (1986) 403. [1461 C.W. Lentz. Inorg. Chem., 3 (1964) 574. [147] D. HoebbeL 1. Pitsch. T. Reffier. W. Hiller. H. Jancke and D. Muller. Z. Anorg. Allg. Chem.. 576 (1989) 160. [1~8] D. Hoebbel, I. Pitsch, A.R. Grimmer. W. Hiller. H. Jancke and R.K. ~ s , Z. Chem.. 29 (1989) 260. [1491 P,A. Agaskan inorg. Chem.. 29 (1990) 1603. [1~] I. PiLqchoD. Hoebhel. H. Jancke and W. Hiller. Z. Anorg. AIIg. Clam., 596 (1991) 63. [151] D, 14(~bhel.T. l~in~n. K. Endr~s, H. Schmidt. A, Kayan and E, Arp~rc. Proc. First E u r o l ~ Workshop on Hybrid Organic-lnceganic Malerials. Bier~ille. France. 1993. pp. 319323. [1521 I, Pi¢~h. Thesis, Univ¢~rsit~tRostock. 1992. [i~3] D. Hoebhel. T. Reinen. K. Endres and H. Schmidt, Mater. Rcs. Sac. Syrup. Prec.. 346 (1994) 863. {is4] M. Moran. C.M, C a ~ . I. Cuadrado and J. Losada. Organometallics. 12 (1993) 4327. [i85] I. Hasegawa and S. Molojima. J. Organomet. Chem.. 441 (1992) 373. [I861 D. Hoebbel, K. Endres. T. Reinert and l, Pit.~h. J. NonoCryst, Solids, 176 (1994) 179. [157] LH. Vogt and J.F. Brown. Inorg. Chem.. 2 (1963) 189. 11581 M,G. Voronkov. V.I. Larem'yev and VM. Kovrigin, J. Org~mometal, Chem.. 220 (1981) 285. [i891 D, Hoebhel. W, Wicker. P. Francke and A, Otto. Z. Anorg. AII~. Chum.. 418 (1975) 3~. [1~1 J, ~z and C,R) Mas~n~ J, Chem. Soc, Set, A'.. (1970) 2683, 11611 P.A. Agask~r. V,W, Day and W,G, Klemperet. J, Am, Chem. So~,. 10~ (1987) ~54, [Io2] J,F, Brown, J, Polym, Sci,. h ~(1%3) 83. {102~] J,F, Brown .nd L,H, Vo~l. J, Am, Chem, S(~,, 87 (1%S) 4313, g,g, Ufl~er,R, Jan~efl.(3,Jil~e,K,D, Lork a~d IL AnH)ach. ill C~, Horvalh ledJ, High Pcftbfmaace Liquid Cllrom~tto~raghy. Vol. ~. Academic ~ss, New York, 1988, pp, 1=93, [I o-_q g,g, Un~r, ~ u s Silica, I~lsevict, N~w YorL 1979, O,~, Leyden, G.H. LulW¢ll, A.E. Sloan and N, ~ Angeiis. j. Anal. ~htm, Acla, 84 (1970) 97. [~71 Y. Iwasawa (~.), Tailored Metal Calalyst.,~, R~idel, Boston, MA, 198~, [~1 E,P. Pl~demann, Silane Coupling Agents, Plenum, New York, 1982. [~91 K.K. Unger. H, Rupp~ht. B. Valentin and W, Kirchner.Drug Dev. lad, Ph~rm,. ¢ (1983) 69. [17ol F,J, ~her and D,A, Newman, J, Am. Chem, See,. 113 (1990) 1931. [iT11 F.J. Feher, T.A. Budzichowski, R.L. Blanski, g.J. Welter and J.W, giller, Organometailics, i0 (1991) 1516, [1721 F J, Feher and R.A, Budzichowski, Polyhedron, 14 (1995l 3239, [~31 W.A. Hem~nn, R. Anwander, V, Oufaed ~nd W, Scherer, Anew, Chum,, I ~ (1994) 1338, 11741 F J, F~her. T,A, Bud~ichowskl. K, Rahimian and J.W, Ziller, J. Am, Chem, See,, 114 (1992) 3~59, [~Tsl F J, ~h~r, J, Am, Chum, S~:,,/~ (1986) 38~S0, [1~1 R,C. Weast, ihndbook of Chemis~ ai~lPhysics,CRC Press, Cleveland, OH, 571hedn,, 1976, pp, F213=F.214, [~71 FJ, I ~ r , J,.C. Liu. S,R, Wi~l~,m and J,R, S~pley, [norg,

C~m,. 29 (199~)) 513~,

[178] FJ. Feher, T.A. Budzichowsky, S.T. Chacon, M.H. Chisholm and W. Streib, J. Am. Chem. Sac., ]15 (1991) 689.

[1791 K.J. Ivin, Olefin Metathesis. Academic Press, New York. 1983.

[18o1 V. Dragutan and A.T. Balaban, Dimonie Olefin Metathesis and Ring Opening Polymerisation of Cyclo-Olefins, Wiley-interscience, New York, i 98 i. [181] FJ. Feher and R.L Blansld, J, Chem. Sac, Chem. Commun., 0990) 1614. [182] FJ. Feller and J,F. Walzer, Inorg. Chem., 30 (! 991 ) !689. [1831 FJ, Feher. J.F. Wa|zer and R.L Blanski, J. Am. Chem, So(:., 113 (1991) 3618, [184] F.J. Feher and R.L. Blanski, J. Am. Chem. Sac., 114 {1992) 5886. 11851 FJ. Feher, S.L Gonzales and J.W. Ziller, Inorg, Chem,, 27 (1988) 3440, [186] F.J. Feher and J.F. Walzer, lnorg. Chem., 29 (1990) 1604, [187] FJ. Feher and K.J. Weller, Organometallics. 9 (19~))2638. [1881 FJ. Feher, T,A, Bud~ichowski and KJ, Weller, J. Am. O)em. Sac.. i l l (1989) 7288. [189] FJ. Feher, KJ. Weller and J.W. Ziller, J. Am. Chem. Sac., 114 (ir992) 9686. [~9o] FJ. Feher, T.A. Budzichowski and J.W. Ziller. ino~. Chem., 31 (1992) 5100. [191] N. Winkhofer, H.W. Roesky. M. Noltemeyer and W.T. Robinson. Angew, Chem. Int. Ed. Engl., 31 (1992) 599. [192] F.T. Edelmann, Angew. Chem. lnt. Ed, Engl., 31 (1992) 580. [193] N. Winkhofer, A. Voigt, It, Dam, H,W, Roesky, A. Steiner, D. Stalke and A. Relier, Angew. Chem. Int. Ed. Engl,, M (1994) 1352, [1941 M.L Montem, A. Voigt, M. Teichert, !, Uson and H.W. Rt~sky, Angew. Chem, Int. Ed, Engl,, 34 (1995) 2504. [ZOS] H. Biirgy, G. Calzaferri, D. Herren and A. Zhdanov, Chimia, 45 (1991) 3. 11961 A.N. Kanev. S.A, Pro~hor~va and G.A. Kokovin, Top, Curt, Cheo, 102 (1982) 243. [I~71 R.A, M~hcalskaya, ~lckiron, Tekh0 Set, M~Ie|°. Vyp,, 7 (1979) 61). (198] Ao galchmann. ~h. Khinl,. Ic (19h7)738, i1991 K,A, Aodrlanov, Vys,komol, Stud0.7 (1')6,~) 147":, K,A. Aildftaflov. I~kl, Akad, N,tfk SSSR, 4 (1%~) 8 ~ [:(HI V.S, Papkov, Vysokomol, Seed. Ser, A:, I? (197~S) 2?00, [2o.~IV,S, Papkov, Vysokomol. SL~d, Set, A;, 29 ( 1 9 7 ? ) ~ 1 , [2o31 V,S, Pltpkov, Vysokomol, Sl~d, Ser, A:, 23 (1980) II?, [2~I T.N, M~mynova, V,P. Korchkov and P.P. Semyannikov. Zh. Ohshch, Khim., 53 (1983) 1587. [2051 D.C. Bradley, R.C. Mehrota, W. Wardlaw, J. Chem. See,, 1952, 12, 5020, [2~1 BJ, Baliga, S,K, Cl~andhi, J, El¢ctrt~hem, Sac,, 1976, 123, 941. [20~] V,Ph, Koran, V.A. Kurwchkin, V.P. ~min, Plenki it ElemenI~anicheskih Soedineniy v Radi~leMronike, Energija, Moscow, 1973. [2o81 S,A, Neuestmev, 2nd, Syrup, po Plasmochimii, Riga, 1975. Vol I pp, 201-204. [2091 V,P, Korchkov, T,N. Marlyno~a, V.I, B¢lyi, Thin Solid Films. 1983, I01, 369, [-'telT,N, Marty~va, V.P. Korchkov, P,P Semiannikov. J. OtBa~m~et, Chem., 1983, 25~, 277, [2111 S,B, ~ s u , C,H, Peng, T, Shi, P,A. Agaskar, J. Electr(~:hem. See., 1992. 139, 2682, [2121 P.A, Ag~skttr. J, Am. Chem. S~c., I l l (1989) 6859. [2131 P.A, ABask~r, J. Chem, S~...Chem. Commun.. (It}92) I024. [2141 D, Ho~bb¢l, 1,Pitsch,D. H¢idem:mn. H, Jancke a.d W. Hiller. Z, Anorg. Allg. Chem.. 58.~(1990) 133.

P.G Itarrison / Jounlal ~" Orgammwtallic Clwmislr3"542 ( 1997~ 14I-18.¢ [215] P. Kolsch, !. Pitsch, D. Schultze, D. Heidemann and D. Hoebbel, J. Therm. Anal,, 4, {1994) 789. [216] D. Hoebhel, L Pitsch and D, Heidemann, Z. Anorg, AIIg. Chem., 592 (1991) 207. [217] P.G. Harrison and R. Kannengeisser, J. Chem. Soc. Chem. Commun., (1996) 415. [218] i. Ha~gawa, M. Ishida and S. Motojima, Proc. First European Workshop on Hybrid O~anic-lnorganic Materials, Bierviile, France, 1993, pp. 329-332.

183

[219] F.H. Feher and K.J. Weller, Chem, Mater., 6 (1994) 7. [220] R,A. Manaz, P.F, Jones, K.P. Chaffee, J.D. Lichtenhan, LW. Gilman, I.M.K. lsmail and M.J. Burmeister, Chem. Mater., 8 (1996) 1250. [221] J,D, Lichtenh~m, N.Q, Vu, LA. Carter, LW. Gilman and F,J. Feher, Macromolecules, 26 (1993) 2141.