Catalyst-free synthesis of polyorganosiloxanes by high temperature and pressure water. II. Understanding of the reaction process

Catalyst-Free Synthesis of Polyorganosiloxanes by HighTemperature and Pressure Water. II. Understandingof the Reaction Process

TAKUYA OGAWA,1 JUN WATANABE,2 YOSHITO OSHIMA2

1Business and Technology Incubator, Dow Corning Toray Co. Ltd., Ichihara 299-0108, Japan

2Department of Environment Systems, Graduate School of Frontier Sciences,The University of Tokyo, Kashiwa 277-8563, Japan

Received 11 October 2008; accepted 11 February 2009DOI: 10.1002/pola.23350Published online in Wiley InterScience (www.interscience.wiley.com).

ABSTRACT: A catalyst-free polysiloxane synthetic process that uses high temperatureand pressure water for the hydrolysis and subsequent polycondensation of phenyltrime-thoxysilane was studied in detail to gain insights into the reaction mechanism. It wassuggested that this process is essentially composed of two stages: (1) oligomerization ofphenyltrimethoxysilane yielding low-molecular weight species with high contents ofsilanol and methoxy groups and (2) polycondensation of the oligomers yielding high-molecular weight species. The use of a preformed oligosiloxane as a starting materialwas informative to understand the polycondensation stage. A modified synthetic processin which a stop valve was introduced to control the internal pressure was developedbased on the understanding of the present process. This modified process enabled atwo-stage reaction resulting in a discernible increase of the molecular weight of polysi-loxane. VVC 2009 Wiley Periodicals, Inc. J Polym Sci Part A: Polym Chem 47: 2656–2663, 2009

Keywords: alkoxysilanes; catalyst-free; hydrolysis; polycondensation; polyorgano-siloxanes

INTRODUCTION

Polyorganosiloxanes or silicones are the most pop-ular silicon-based polymeric materials in whichthe backbone is composed of repeating SiAO link-ages. On account of the good thermal stability,low-temperature stability, weatherability, trans-parency, and electric insulation, the materials areused in almost all industries including auto-mobile, construction, electronics, personal andhousehold care, and chemical industries. On ofimportant synthetic methods yielding silicones ishydrolysis and subsequent polycondensation of

alkoxysilanes in an organic solvent in the pres-ence of an acid catalyst.1 This process is well-established, but a large amount of organic solventwaste is a potential issue from en environmentalstandpoint. In addition, quite long reaction timeis also a disadvantage for the incumbent siliconemanufacturing processes.2–4

We have proposed a new silicone synthetic pro-cess in which a mixture of an alkoxysilane andwater was heated in a sealed reactor with neithera catalyst nor an organic solvent.5 This is the firstexample of a catalyst-free silicone synthesis by asupercritical water technology. As this polymer-ization process can be regarded as a reactionunder autogenous pressure at elevated tempera-tures,6 no pressurizing equipment is required.The largest advantage of this new process is nocontamination of volatile organic compounds in

Journal of Polymer Science: Part A: Polymer Chemistry, Vol. 47, 2656–2663 (2009)VVC 2009 Wiley Periodicals, Inc.

Correspondence to: T. Ogawa (E-mail: [email protected])

2656

the product because the product is obtained as asolvent-free form. In addition, the present syn-thetic method is very simplified due to skip of theneutralization process being necessary for a con-ventional solution-based process. Both acidity ofwater at the subcritical state7 and high reactiontemperature proved to be critical factors for theprocess. Several essential process variablesincluding the temperature, pressure, and the[water]/[alkoxysilane] ratio were studied,5 but in-depth reaction understanding is necessary to seeif this new process is feasible from a practicalstandpoint. In this report, an impact of each ele-mental reaction on the polymer properties wasexamined for the synthesis of poly(phenylsilses-quioxane), PPSQ hereafter, from phenyltrime-thoxysilane, PTMS hereafter. A preformed oligosi-loxane was essentially used as a substrate to focuson the role of the reactants to the molecularweight increase in the synthesis. Based on theunderstanding, a study on a modified processyielding high-molecular weight polymers isdescribed in the final part of this report.

EXPERIMENTAL

Equipment

A smaller reactor system composed of a reactorand narrow tubing with the internal volumes of10 mL and 2 mL, respectively, was made ofSUS 316 BA tubing. The tubings with diameters

of 1/200 and 1/800 were used for the reactor and thenarrow tubing, respectively. The narrow tubingfunctions as a volatile material trap as reportedpreviously.5 On the other hand, a larger reactorsystem composed of a 100-mL reactor made of a100 tubing and a 30-mL narrow tubing was alsomade of the SUS 316 BA tubings. The 30-mLtubing was made by combination of 1/1600, 1/800,and 1/200 tubings. A photo of the larger reactorsystem and an illustration of the reaction systemwith a heat source are depicted in Figure 1(a,b),respectively. The stop valve in the narrow tubingis optional. Connectors were SwagelokVR products.A pressure gauge was a KH15 pressure transmit-ter of Nagano Keiki Co Ltd. A salt bath with thetemperature deviation within þ/�1 �C was usedas the heat source.

Materials and Characterization Methods

PTMS as a starting material supplied by DowCorning Corp. was used without further purifica-tion. Oligosiloxanes having silanol and SiAOMegroups were synthesized as reported elsewhere,5

whereas that with abundant silanol groups wassynthesized by hydrolysis of phenyltrichrolosi-lane.8 Analytical gas chromatography (GC) wasperformed using a Shimadzu GC-14B gas chroma-tograph equipped with a DB-5 capillary column (Jand W Scientific). Gel permeation chromatogra-phy (GPC) was performed using chloroform as aneluent with a Shimadzu LC-VP gel permeationchromatography system equipped with four

Figure 1. (a) and (b). Experimental set-up for the present catalyst-free synthesis ofpolyorganosiloxanes. A photo of the larger reactor system (a) and an illustration ofthe stop valve-attached reactor system with a heat source (b).

CATALYST-FREE SYNTHESIS OF POLYORGANOSILOXANES 2657

Journal of Polymer Science: Part A: Polymer ChemistryDOI 10.1002/pola

Shodex columns: K-G, K-800D, K-804L, and K-803L, and a refractometer. The weight averagemolecular weight (Mw) and the polydispersity(PDI) were calculated using polystyrene stand-ards. 29Si-NMR spectra of CD3COCD3 solutionwere recorded with a Bruker AC300P spectrome-ter. Tetramethylsilane was used as an externalstandard. The [Ph(RO)SiO2/2]/[PhSiO3/2þPh(RO)-SiO

2/2] molar ratio as a measure of the residual

alkoxy and silanol contents was calculated by anintegral ratio in the 29Si-NMR spectra.

Synthesis of PPSQ by the Reaction Using EitherPTMS or Oligosiloxane Containing AbundantSiAOH Groups

The synthesis was conducted by the smaller reac-tor system depicted in Figure 1(a) without thestop valve as was described elsewhere.5

Synthesis of PPSQ by a Modified Processwith a 100-mL Reactor

PTMS (32 mL; 172 mmol) and deionized water(16.24 mL; 0.9 mol: 3.5 equivalents to the theoreti-cal amount) were loaded in a one-end capped 100-mL reactor attached with a temperature gauge.The other end was sealed by a cap connected totubing with the volume of 30 mL as a volatile ma-terial trap. A stop valve and a pressure gauge wereinstalled in the middle and at the terminal of thetubing, respectively. The stop valve was kept clos-ing. The reactor was placed in a preheated saltbath at 300 �C to start the reaction. During heat-ing, the temperature and pressure inside the reac-tor was monitored periodically. After heating for 3min, the stop valve was opened with maintainingthe heating. The reactor was pulled out of the saltbath after a designated period of time, and pouredinto a water bath to terminate the reaction. A solidproduct inside and at a wall of the reactor was col-lected by using a spatula. Drying the solid at 80 �Cin vacuo yielded colorless solid (20.95 g; 94% yield).

29Si-NMR (CD3COCD3; ppm): d ¼ �69.0(C6H5Si(OR)O2/2; R: H and CH3) and �78.0(C6H5SiO3/2). Mw: 9930; PDI: 3.15.

RESULTS AND DISCUSSION

Reaction of PTMS Under Various Conditions

Hydrolysis and Subsequent Polycondensationof PTMS

The overall reaction scheme for the synthesis ofPPSQ via hydrolysis and subsequent polyconden-

sation reactions of PTMS is as follows:

C6H5SiðOCH3Þ3 þ 3=2H2O ! C6H5SiO3=2

þ 3CH3OH ð1ÞAs reported previously, an analysis of a nonvo-

latile product by GPC indicates that the sequen-tial reaction goes stepwise with increasing molec-ular weight. The increase, however, leveled off af-ter the reaction period of about 15 min becausethe polycondensation attains the equilibrium.5

Volatile products, which are methanol as a by-product and unreacted PTMS, were analyzed byGC in this study. The amount of PTMS as a func-tion of the reaction time is shown in Figure 2along with that of methanol formed by both hy-drolysis of PTMS and subsequent polycondensa-tion of the partially-hydrolyzed species. In addi-tion, a couple of elemental reactions for the pres-ent synthesis are listed in Scheme 1.

It was demonstrated that PTMS was consumedwithin 3 min, whereas the generation of methanolceased by 15 min heating. These results indicatethat this sequential reaction is essentially dividedinto two stages. Hydrolysis of PTMS as a primaryreaction, shown as 1a of Scheme 1, takes place atfirst. Condensation between the hydrolyzed spe-cies and PTMS indicated as 1b of Scheme 1 occursat the same time as a secondary reaction. Repeat-ing the reaction 1b and condensation reactionsbetween the low-molecular weight species thusformed yielding oligosiloxanes can be regarded asthe first stage of the synthesis. Polycondensationbetween the resulting oligomers listed as 1c in

Figure 2. The amounts of PTMS and methanol as afunction of reaction time. [Color figure can be viewedin the online issue, which is available at www.interscience.wiley.com.].

2658 OGAWA, WATANABE, AND OSHIMA


Scheme 1 seems to occur at the second stage toyield higher molecular weight polymers. Oligo-mers and polymers with nonlinear structures arealso formed in the course of both reactions 1b and1c.

Effect of the Substrate Amount on the PolymerMolecular Weight

The pressure inside the reactor increases whilethe reaction proceeds because volatile methanolas a byproduct is generated along with the desiredPPSQ formation. The effect of the pressure on theproduct’s molecular weight has been studied,5 butis discussed again by a different manner. Runswith smaller amount of substrates are consideredto be runs at lower pressures. The total substrateamount, which is the sum of PTMS and water,was reduced to a half of the original value withmaintaining the [PTMS]/[water] molar ratio. Asshown in Figure 3, the molecular weight of theproduct became nearly constant after 15 minheating as was observed for runs in the above sec-tion. Each molecular weigh value was, however,discernibly higher than that for the run withdoubled substrate amount. To rationalize the ob-servation, we assumed that the molecular weightof PPSQ is dominated by the phase of MeOH andwater as well as their amounts. An effect of addedeither MeOH or water was studied next to vali-date the assumption.

Reaction of a Preformed Oligosiloxane UnderVarious Conditions

As a study with PTMS suggests that the syntheticprocess is basically composed of two stages, syn-thesis of PPSQ using a preformed oligosiloxanewas examined to better understand the polycon-densation as the second stage of the process.

Reaction of the Oligosiloxane and MeOH

The reaction of a preformed oligosiloxane andMeOH was examined at 300 �C as summarized inTable 1.

The molecular weight decreased when theMeOH amount is as high as the theoreticalamount. This is consistent with the fact thatMeOH remained in a reactor lowered the molecu-lar weight of the product.5 On the other hand, themolecular weight increase was observed for therun with quite small amount of MeOH. The causeof the increase will be discussed in the later sec-tion.

Reaction of the Oligosiloxane and Water

As the molecular weight of the polymer stronglydepends on the MeOH amount, a reaction of theoligosiloxane and water was examined underessentially two conditions: w/ large amount ofwater and w/ small amount of water. The runswere made at various temperatures rangingbetween 250 and 400 �C. The water amount wasreduced as the temperature increases to adjustthe internal pressure to the corresponding satu-rated pressure value of water. The results are alsosummarized in Table 1.

Figure 3. The molecular weight as a function ofreaction time. Square data points: the [PTMS]/[water]loaded volume of 1.5/0.75 (v/v); diamond data points:the [PTMS]/[water] loaded volume of 3.0/1.5 (v/v).[Color figure can be viewed in the online issue, whichis available at www.interscience.wiley.com.].

Scheme 1. Elemental reactions for the synthesis of PPSQ.



It is very intriguing to note that the molecularweight of the oligosiloxane increased for all runsregardless of the reaction conditions. In fact, theincrease was not affected by the temperature, butstrongly depended on the water amount. A runwith the water amount of 0.24 mL yielded a solidpolymer with the molecular weight as high as46,000 indicating that self-condensation of theoligomers took place as discussed in the latersection. On the other hand, runs with a largeramount of water yielded the polymer’s molecularweight ranging between 3,000 and 5,000.

Reaction of the Oligosiloxane and PTMS

Batches were examined at 300 �C with both asmall and a large amount of PTMS expecting thatthe molecular weight increase takes place by poly-condensation between silanols of the preformedoligosiloxane and methoxy groups of PTMS. Theresults are also summarized in Table 1.

Again, a reaction of the oligosiloxane with atrace amount of PTMS increased the molecularweight, whereas other batches with largeramount of PTMS decreased the value. This mayindicate that polycondensation between the oligo-siloxane and PTMS did not occur, but redistribu-tion between these species would have beenbrought about under these conditions.

The molecular weight decrease observed forthe runs with larger amount of PTMS is consist-

ent with the result using MeOH as describedabove. On the other hand, these results exhibit anevident contrast compared with those with largeramount of water. This may be explained by theinteraction of these reactants with the oligosilox-ane. MeOH and PTMS can highly interact withthe oligosiloxane whereas water does not seem tohave large interaction. In fact, the oligosiloxane issoluble in both MeOH and PTMS. This largeinteraction may lead to either methanolysis of theoligosiloxane or redistribution between the oligo-siloxane and PTMS. As water has essentially nointeraction with the oligosiloxane, just self-condensation of the oligosiloxane yielding highermolecular weight species may have taken place.

Bulk Self-Condensation of the Oligosiloxane

As described in the above sections, reactions of thepreformed oligosiloxane with either MeOH, water,or PTMS yielded a solid product with increasedmolecular weight when the reagents’ amount issmall. The oligosiloxane was heated by itself atvarious temperatures expecting that highermolecular weight species be formed by bulk self-condensation evolving MeOH and water as shownin eq 2. The results are also listed in Table 1.

The molecular weights of all the solid productswere higher than those before heat treatmentindicating that bulk self-condensation of the

Table 1. Reaction of a Preformed Oligosiloxane Under Various Conditionsa

Run Oligomer (g)MeOH(mL)

Water(mL)

PTMS(mL)

Temp.(�C) Appearance Mwb Mwc

1 0.5 0.16 300 Solid 3,770 4,8402 0.5 1.57d 300 Solid 3,770 3,0803 0.5 0.24 350 Solid 1,150 46,0004 0.5 2.35 350 Solid 1,150 3,1005 0.5 8.0 250 Solid 2,470 4,3506 0.5 7.1 300 Solid 2,470 4,2007 0.5 5.75 350 Solid 2,470 3,1308 0.5 1.7 400 Solid 2,470 4,0809 0.5 0.1 300 Solid 1,370 2,990

10 0.5 0.47 300 Liquid 2,470 1,78011 0.5 2.35 300 Liquid 2,470 2,00012 0.5 250 Solid 2,470 3,70013 0.5 300 Solid 2,470 6,36014 0.5 350 Solid 2,470 108,13015 0.5 400 Solid 2,470 58,640

aReaction time is 10 min.bWeight average molecular weight of the starting oligosiloxane.cWeight average molecular weight of the product.d The amount corresponds to the theoretical generation amount of MeOH for the synthesis of 0.5 g PPSQ.



oligosiloxane took place.9 It proved that the molec-ular weight significantly increases at tempera-tures over 350 �C. The molecular weight at 350 �Cwas higher than that at 400 �C. The reason forthe phenomenon would be explained as follows.As the self-condensation is an equilibrium reac-tion, too high pressure and too large amount ofMeOH in a reaction vessel shift the equilibrationto the starting material-side. These results indi-cate that both the temperature and pressureaffect the degree of bulk self-condensation.

OH=OMe-containing oligomer

! Self-condensed oligomer

þMeOHand=orwater ð2Þ

Lastly the bulk self-condensation of an oligosi-loxane bearing abundant silanol groups withoutSiAOMe groups was examined. Glassy solid withthe molecular weight of 23,000 was formed byheating at 300 �C while that of the starting oligo-siloxane was 1,420. This indicates that dehydra-tive polycondensation as well as the demethanolpolycondensation is possible by the present bulkprocess.

A Modified Synthetic Process with a StopValve at the Narrow Tubing

Pressure control proved to be a critical factor tosynthesize high-molecular weight polymer by thestudies described above. In addition, the presentsynthetic process is essentially a two stage pro-cess. Hence, design of a new reactor was made tomodify the synthetic process. A new reaction sys-tem in which a stop-valve was introduced in thenarrow tubing as a volatile material trap wasdeveloped as depicted in Figure 1(b) to enable the

two-stage process by controlling the internal pres-sure. The results are summarized in Table 2.

The temperature and pressure profiles for runs#16 and #17 are depicted in Figure 4.

The temperature for run #16 monotonouslyincreased whereas that for run #17 exhibited asteep decrease at the valve opening and thesubsequent monotonous increase afterward. Thesteep drop is likely to correspond to the transferof the volatile components described above. Oneshould pay larger attention to the pressure pro-file. The pressure increases to reach the top andthen gradually decreased in run #16. Thedecrease is due to cooling of the materials in thetrap. On the other hand, a presumably large dropof the pressure was observed when the valve wasopened for run #17 although the actual pressurevalues before the valve opening can not be

Table 2. Synthesis of PPSQ by a Modified Processa

Run #Valve-OpeningTime (min)b Pc

IsolatedYield (%)

Mw(PDI)d [OR]e

16 0 6.0 84 7,110 (2.76) 22.417 3 6.4 94 9,930 (3.15) 15.718 1–3 6.3 97 24,600 (6.43) 17.6

aThe reaction temperature was 300 �C, while the total reaction time was 17 min.bValve remained open while heating for batch #16 while the valve was gradually opened

over a period of 2 min for batch #18.c The maximum pressure in MPa observed after the valve opened.d The weight average molecular weight and the polydispersity index in parenthesis.e The residual alkoxy and silanol contents calculated by 29Si NMR spectroscopy.

Figure 4. The temperature and pressure profiles forruns #16 (open data points) and #17 (filled data points).[Color figure can be viewed in the online issue, which isavailable at www.interscience.wiley.com.].



recorded. The pressures are roughly estimated tobe greater than 30 MPa by an equation of statefor mixed gases. The pressure profile evidentlyindicates that the two-stage process can bedesigned by this valve-attached system.

An idea of this new system is as follows. One canexpect oligomerization of PTMS without removingmethanol for certain length of time at higher pres-sure as the first stage of the reaction. The secondstage, polycondensation of the resulting oligomerswith removing methanol and water, will occur byfurther heating after the valve opened. This stagecan be regarded as bulk polycondensationdescribed in the above section because very lowcontent of the volatile species was present. Anotherpotential advantage of this system is to suppressvaporization and subsequent devolatilization ofunreacted PTMS in the trap resulting in low prod-uct yield. As a control experiment, the valveremained open while heating in run #16.

When the valve was opened in run #17, signifi-cant quantity of volatile components, which areessentially water and MeOH, was spontaneouslytransferred into the trap due to the pressure differ-ence between the reactor and the trap. In fact, col-orless solid with very small quantity of water andmethanol was obtained as a product from the reac-tor. The isolated yield was 94% while the molecularweight was �10,000. The molecular weight wasdiscernibly higher than that for run #16 presum-ably due to an accomplishment of the well-organ-ized two-stage process. It is notable that the molec-ular weight of the product in run #18, in which thevalve was gradually opened over a period of 2 minafter maintaining the valve-closed state for one mi-nute, was significantly higher than that for run#16. Volatile components were continuously trans-ferred into the trap during the valve opening asobserved in run #17. This suggests that control re-moval of the volatiles and subsequent polyconden-sation at bulk state are very critical to increase themolecular weight of the product.

As for the product yield, the values in runs #17and #18 are discernibly higher than that in run#16. This is definitely due to suppression of vapor-ization of PTMS.

Polymer Structures

We have reported that polymers synthesized bythe present high temperature and pressure pro-cess had high content of residual methoxy andhydroxy, OR hereafter, groups despite their highmolecular weight.5 The OR group content of the

products synthesized by the stop value-attachedprocess was studied by 29Si-NMR spectroscopy.The products in runs #17 and #18 had certainamount of OR groups, but the values were below20 mol % as listed in Table 2. This is discerniblylower than that synthesized by the process with-out the stop valve definitely due to the effectiveremoval of methanol and water generated by thepolycondensation. These results indicate that thisvalve-attached system is suitable to synthesizepolymers with lower content of the structuraldefects.10 Except for the OR group content, no dis-cernible structural difference was observedbetween products synthesized by a conventionalsolution-based process and the present process.

CONCLUSIONS

An in-depth study on the catalyst-free synthesisof PPSQ was carried out by a 10-mL batch reactorfocusing on how several elemental reactions affectthe polymer properties, essentially the molecularweight. The elemental reactions include reactionsof preformed PPSQ oligomer with either methanolor water. The important findings are as follows.

1. An amount of volatiles in the reactor isvery critical. A trace amount of the vola-tiles does not discernibly affect the poly-condensation of the oligosiloxane. Molecu-lar weight decrease is observed when alarger amount of MeOH and/or PTMS ispresent, whereas water does not cause themolecular weight decrease.

2. Bulk polycondensation of the preformedoligosiloxane takes place yielding PPSQwith an increased molecular weight.

According to these facts, the catalyst-free syn-thesis can be regarded as a two-stage process: hy-drolysis and subsequent condensation of PTMSyielding an oligosiloxane as the first step, andpolycondensation of the oligomer yielding a high-molecular weight PPSQ as the second step. Alarger amount of MeOH and/or PTMS preventsthe polycondensation resulting in formation of thelow-molecular weight species.

A modified reaction system in which a stop-valve was attached in a volatile material trapproved to be very effective to synthesize highermolecular weight polysiloxanes. The valve cancontrol the internal pressure to attain the two-stage process: heating with the closed valve at



higher pressure for oligomerization as the first stepand subsequent heating under the valve-openedstate at lower pressure for polycondensation as thesecond step. The molecular weight attained bythis process was as high as 24,600.

The financial support byNEDO is greatly appreciated.

REFERENCES AND NOTES

1. Noll, W. Chemistry and Technology of Silicones;Academic Press: New York, 1968; Chapter 5.

2. Abe, Y.; Kagayama, K.; Takamaru, N.; Gunji, T.;Yoshihara, T.; Takahashi, N. J Non-Cryst Solids2000, 261, 39–51.

3. Atkins, G. R.; Krolikowska, R. M.; Samoc, A.J Non-Cryst Solids 2000, 265, 210–220.

4. Baney, R. H.; Itoh, M.; Sakakibara, A.; Suzuki, T.Chem Rev 1995, 95, 1409–1430.

5. Ogawa, T.; Watanabe, J.; Oshima, Y. J. SupercritFluids 2008, 45, 80–87.

6. Prof. Gedanken’s group has been actively study-ing the reactions of metal alkoxides under auto-genous pressure at elevated temperature. Forexample: Pol, S. V.; Pol, V. G.; Seisenbaeva, G. A.;Kessler, V. G.; Gedanken, A. Chem Mater 2004,16, 1793–1798.

7. Savage, P. E. Chem Rev 1999, 99, 603–622.8. Takiguchi, T.; Fujikawa, E.; Yamamoto, Y.; Ueda,

M.; Nippon Kagaku Kaishi 1974, 108–111.9. Bulk polycondensation is a well-known technology

to increase the molecular weight of thermoplastic

polyesters. This process is typically carried out

under reduced pressure above 250 �C.10. OR groups are regarded as the structural defect of

polysilsesquioxanes because they are reactive to

form siloxane bonds by dehydrative and/or de-

methanol condensation reactions. In fact, polymers

bearing these groups have, in general, poor thermal

stability because of the weight loss via the conden-

sation reaction. In addition, these polymers have

lower glass transition temperatures compared to

polymers having lower content of OR groups.



Documents

Catalyst-free synthesis of polyorganosiloxanes by high temperature and pressure water. II. Understanding of the reaction process