Journal of Sol-GelScienceand Technology,5, 77-82 (1995) © 1995 KluwerAcademicPublishers,Boston. Manufacturedin The Netherlands.

Transparent Inorganic/Organic Copolymers by the Sol-Gel Process: Thermal Behavior of Copolymers of Tetraethyl Orthosilicate (TEOS), Vinyl Triethoxysilane (VTES) and (Meth)acrylate Monomers ANNA B. WOJCIK AND LISA C. KLEIN

Rutgers--the State University of New Jersey, Ceramics Department, P.O. Box 909, Piscataway, NJ 08855-0909 Received March 24, 1995; Accepted June 9, 1995 Abstract. Three types of inorganic/organic copolymers have been prepared in a one-step sol-gel process, and their thermal stabilities have been studied. The one-step sol gel process was carried out in mixtures of three monomeric components: HEMA (hydroxyethyl methacrylate)-VTES (vinyl triethoxysilane)-TEOS (tetraethyl orthsilicate), HDDA (hexanediol diacrylate)-VTES-TEOS and GPTA (glycerol propoxy triacrylate)-VTES-'IEOS. Copolymers with HEMA, which is able to form a linear organic polymer, were the least thermally stable materials. They lost the highest proportion of weight during heat treatment to 600°C and exhibited the lowest decomposition temperatures. Copolymers with HDDA and GPTA, which are able to form crosslinked organic polymers, had higher decomposition temperatures, and their weight loss during heat treatment to 600°C was small. The skeletal densities of all copolymers increased slightly during heat treatment. Keywords: inorganic/organic copolymers, one-step sol-gel process, linear polymers, crosslinked polymers, silica/polyacrylate copolymers 1

Introduction

Among the inorganic/organic hybrid materials prepared by sol-gel processes, those that have the inorganic and organic constituents covalently bonded to each other have received the most attention [1-3]. The reason is that covalent inorganic/organic linkages reduce the tendency for phase separation [4]. In the absence of phase separation, it is possible to combine the hardness of the inorganic component with the flexibility of the organic component while maintaining good optical properties [1]. In addition to the optical and mechanical properties of the hybrid materials, a very important parameter is their thermal stability. It is essential to study the thermal stability before selecting the temperature at which materials can be used without degradation. Thermal stability is critical when the hybrid materials are intended for use at elevated temperatures over long periods of time or for many temperature cycles. The thermal stability and hardness of some polyacrylate/silica copolymers by sol-gel processes have been reported [5-7]. In these cases, the polyacrylates formed linear chains covalently attached to silica. Recently, we reported the synthesis of three

component hybrid materials that were prepared by an acid catalyzed sol-gel process in one step, using cocondensation and copolymerization of TEOS, VTES and (meth)acrylate monomer [8]. These hybrid materials consist of inorganic/organic copolymers, with the (meth)acrylate monomers being HEMA (hydroxyethyl methacrylate), HDDA (hexanediol diacrylate) and GPTA (glycerol propoxy acrylate). HEMA copolymers have poly(HEMA) linear chains that are linked to silica, whereas HDDA and GPTA copolymers are expected to form crosslinked organic networks. Originally, we prepared two component hybrids by infiltrating silica gels with GPTA, but later found that simultaneous interpenetrating networks could be prepared that had equally high transmission [9]. For further comparison with infiltrated gels and interpenetrating network gels, we have concentrated on gels with direct Si--C linkages, leading to the use of the third component VTES [8]. The main interest of our studies continues to be improved optical transmission. The highest optical transmission was found with HDDA. Transmission decreased as the content of HDDA in the copolymer increased, though the differences were not large. For HEMA copolymers, we saw the reverse. For the highest

78

Wojcik and Klein CH3 I OH-CH2-CH2-O-CO- CH=CH2

PIHEMA) -

HEMA

PIHDDA)

CH2=CH-CO-O-(CH2)6-O-CO-CH=CH2 HDDA

CH2-O-(C3H60)- CO-CH--CH2 I

CH-O-IC3H60)- CO-CH=CH2

P(GPTA)

I CH2-O-(C3H60)-CO-CH=CH2 GPTA

Fig. 1.

Chemistry of the organic monomers and schematic structures of their polymers: linear for poly(HEMA) and crosslinked networks for poly(HDDA) and poly(GPTA).

content of HEMA in the copolymer structure, we found the highest transmission. The reason for this is the structure and relative size of the HEMA molecule. The HEMA molecule is small in relation to the HDDA and GPTA molecules, and can be uniformly distributed throughout the silica network. Its hydroxyl groups, together with carbonyl groups, can contribute to hydrogen bonding with S i - - O H species, whereas HDDA and GPTA molecules, like methylmethacrylate (MMA) [10], can interact with silanol groups only via their carbonyl groups. Optical transmission of the GPTA copolymers was the lowest, following the same dependence with the copolymer content as with HDDA copolymers. Based on a comparison of optical transmission, it was concluded that the copolymerization rate of HEMA with VTES must be higher than that for HDDA and GPTA with VTES. This meant that, under the conditions used in the one-step process, HEMA formed more homogeneous hybrid copolymers than either HDDA or GPTA. Our continuing interest is to use these copolymers as optical materials. Therefore, we need to evaluate their thermal stability. In this paper, we report some thermal properties of the these three types of copolymers on the basis of pyrolysis, thermogravimetric weight loss and effect of heat treatment on skeletal density.

2

Experimental

Procedures

Inorganic/organic copolymers of HEMA-VTESTEOS, HDDA-VTES-TEOS and GPTA-VTES-TEOS were prepared by a sol-gel method, as described previously [8]. The chemistry of the organic monomers and schematic structures of their polymers are shown in Fig. 1. A schematic diagram of the polymerization of the three monomer system, represented by HEMAVTES-TEOS, is shown in Fig. 2. The compositions of the monomer mixtures are listed in Table 1. VTES 1, TEOS 2, ethanol (equal volume to TEOS), water representing 1.2 moles per mole of ethoxysilyl groups, and hydrochloric acid (0.01 M) were mixed for 24 hours. Then the organic monomer, HEMA, HDDA or GPTA 3, was added along with benzoyl peroxide as an initiator (0.5% of HEMA weight; 1% of HDDA weight and 1.5% of GPTA weight). After half an hour mixing, 10 ml of the solution was poured into polypropylene tubes and the tubes were sealed. The tubes were heated to 60°C. Usually after 24 hours, the contents of the tubes gelled, and heating was continued for another 5-6 days. Then the caps on the tubes were slightly loosened to allow the solvents to evaporate slowly. The samples were dried at 50°C for one week. All of the dried samples were monolithic. Following drying, some samples

Inorganic/Organic Copolymers by the Sol-Gel Process i i

i ./OC2Hs CH 2=CH *SI -OC2H 5

VTES

\

OC2Hs

+

HEMA +

initiator

+ TEOS

CH2 i.

CzHsO,N.

C-CH 3 I

C=O I o I CH2 I CH=OH

/

OCzH5

Si C,HsO/

free radicalpolymerization of~s 'SE = vinyl = groups ~ and ~ HEMA ~ ~lV

linearpolymerchain ~ ~ ' ~ A k

~OC 2Hs ~ HCI/water

hydrolysisof ethoxysilyl groups

condensationof Si-OH groups q

/

-R ~. ~

silicaparticles

79

Pyrolyses of the hybrid materials were performed in an electric furnace in air for 4 hrs at 600°C. After pyrolysis, transparent porous, monolithic rods were obtained. Their weight constitutes the silica content in the hybrid materials, and these measurements are shown in Table 1 under the heading Silica Content--Pyrolysis. Thermogravimetric analysis (TGA) of the heat treated gels was performed on a Perkin Elmer TGA 7 system in air over thetemperature range 25-650°C at a heating rate of 15°C/min. Decomposition onset temperatures were determined from TGA curves, and these temperatures are listed in Table 1. The residual weight at 650°C constitutes the silica content in these samples. The TGA results are listed in Table 1 under the heading Silica Content--TGA. Differential scanning calorimetry (DSC) measurements were performed in air with the heating rate 5°C/min to determine the glass transition temperature of the organic polymer. The skeletal densities of the heat treated samples were determined with helium pycnometry (Micromeritic Accupyc 1330). FTIR spectra were collected for selected samples. More complete descriptions of the FTIR results were reported previously [8].

Fig. 2.

Schematic diagram of the polymerization of the three monomer system: HEMA-VTES-TEOS.

3 were subjected to an additional drying treatment in an oven in air for 24 hrs at 115°C. The weight loss during this treatment is shown in Table 1.

Table 1.

Results and Discussion

DSC analysis showed no detectable glass transition in the range 0-200°C. This means that movement of the

Compositions of the inorganic/organic sol-gel copolymers and their thermal behavior.

Sample code

Composition tool : tool : mol

HEMA HEMA HEMA HEMA

1(#1) 2(#2) 3(#3) 4(#4)

HDDA HDDA HDDA HDDA

1(#5) 2(#6) 4(#7) 3(#8)

HEMA : VTES : TEOS 1.67 : 1 : 1.89 0.9 : 1 : 1.57 2 : 1 : 1.89 3 : 1 : 1.72

Silica content Pyrolysis TGA (wt.%) (wt.%)

Decomposition onset T (°C)

Weight loss (drying, %)

68.9 66.7 44.4 31.4

78.1 76.2 34.5 27.7

278 270 230 220

19.0 20.7 11.3 9.8

83.4 52.3 38.7 32.4

83.2 56.7 39.3 34.1

272 285 336 340

7.5 7.9 5.2 5.1

68.2 56.7 28.2 90.3

68.2 61.0 28.5 90.0

345 350 350

4.0 6.1 7.0 9.2

HDDA : VTES : TEOS 0.56 : 1 : 1.89 0.8 : 1 : 1.48 1.11 : 1 : 1.31 1.11 : 1 : 1.89 GPTA : VTES : TEOS GPTA I(#9) GPTA 2(#10) GPTA 3(#11) Silica(#12)

0.1 : 1 : 2.1 0.15 : 1 : 0.87 0.40 : 1 : 2 100% TEOS

80

Wojcik and Klein 1.7

1.6[

HEMA After

"G" ~>,

==

1.5

aJ

Before

o

1,4

3 1"30

I

I

I

I

.40

50

60

70

80

90

Silica Content (wt%)

Fig. 3. Heliumpycnometerdensities of HEMAcopolymersvs silica content before and after pyrolysis. 1.7

After ~ /

HDDA

1.6

.~ 1.5 c~

1.4

°

1,3 30

i

40

[]

50

i

i

i

60

i

70

i

,

80

i

90

Silica Content (wt%)

Fig. 4. Heliumpycnometerdensities of HDDAcopolymersvs silica content before and after pyrolysis.

polymer chains has been hindered by copolymerizing them with the silica network. This behavior has been observed before in infiltrated gels and is not surprising in a hybrid material that is a nanocomposite [11]. It is useful to measure the skeletal density as a reflection of the physicochemical structure of the sol-gel materials. The relationships between the silica content and the skeletal densities before and after heat treatment are shown in Fig. 3 for HEMA copolymers and in Fig. 4 for HDDA copolymers. In both cases, the skeletal densities increase with the increase of the silica content. The skeletal density is not a simple mixing rule of the densities. The skeletal density for silica xerogel is about 2 g/cm 3 and for the HEMA, HDDA or GPTA less than 1.2 g/cm 3. The hybrid matrix density is closer to that of the organic polymer than to that of silica xerogel.

It has been reported before that both the extent and speed of drying of sol-gel hybrid materials affect their densification [12]. We found that after drying at 50°C for one week, followed by 24 hours at 115°C, all of the copolymers remained crack-free, transparent monoliths with little visible change in their light transmission. We examined the influence of the heat treatment on densities of the copolymers. Upon heat treatment a slight increase not exceeding 8% was observed in the skeletal densities. This increase in the density of the hybrid materials is due to the removal of the volatiles and not due to densification of the covalently linked network. When the effect of heat treatment on hardness was examined by indentation, the changes were within the range of the experimental error. The copolymer weight loss during drying treatment (Table 1) is attributed to the evaporation of volatiles such as water, alcohol or unreacted organic monomers. The weight loss varies from sample to sample with the highest weight losses recorded in HEMA copolymers. The drying weight losses for HEMA copolymers are more than twice the weight loss for HDDA and GPTA copolymers with similar molar ratios of components. This may be due to the hydrophilic character of the HEMA molecule that binds to water molecules strongly, leading to increased water retention in the gels. Part of the large drying weight loss is likely the evaporation of unreacted monomer. HEMA is the smallest of the three monomer molecules, having only one double bond. In comparison, HDDA with two double bonds and GPTA with three double bonds are monomers that are able to crosslink. For HDDA and GPTA copolymers, the weight loss is less than the weight loss observed for silica xerogels prepared from TEOS alone. This suggests that the copolymers with HDDA and GPTA are less hydrophilic than pure silica. The silica content found by pyrolysis of copolymer samples that were dried for one week at 50°C and the silica content from TGA of samples following the additional 24 hr drying at 115°C were similar. The trends were the same and the silica contents were comparable. In the case of the HEMA copolymers, there was some discrepancy for the samples with high silica contents. When the copolymer composition is calculated from the starting monomer mixture and compared to the composition determined by pyrolysis, there is good agreement in the cases of the HDDA and GPTA copolymers. In contrast, when the compositions for HEMA copolymers are calculated from the monomer mixtures and compared to the polymer content estimated from

Inorganic/Organic Copolyrners by the Sol-Gel Process

81

Ioo.o

~.o

9o.o.

9rJ, o

so, n

\

7O, 0 •

tto.o

oo. o

5o.n

50.0.

4n. o

40. O

~o. o

so.o.

-

22 ;7 \

\

\ \

20.o

20. 0

0.o

IkO "r~=n!~mtum (, C)

r~mm

Fig. 5.

Thermogravimetric weight loss curves for HEMA copolymers (#1-4). Pure silica gels are included for comparison (#12).

#.

7o. o

v e0,u-

50.o -

S

~o. o -

~o, o o.o

,,a°

.ao

,~0

,~°

d..

.I

Tempera,me ( ' C )

Thermogravimetricweight loss curves for HDDA copolymers (#5-8). Fig. 6.

the difference determined by: 100% - (TGA silica%) - (drying weight loss%) = polymer content% the polymer content is less than that calculated from the monomer mixture. This means that some amount of monomer or low molecular weight oligomeric HEMA, loosely connected to the matrix, has been lost during the copolymer drying. The thermal stabilities of the hybrid materials were determined by the TGA analysis. The TGA curves are shown in Fig. 5 for the HEMA copolymers, in Fig. 6 for the HDDA copolymers and Fig. 7 for the GPTA copolymers. The temperature for the onset of decomposition is listed in Table 1. Thermal stability of the copolymers was strongly dependent on the kind and the composition of the copolymers. For HEMA copolymers, for the lowest silica content, the onset temperature was as low as 220°C. This temperature is close to the decomposition temperature for

(DC)

Fig. 7. Thermogravimetricweight loss curves for GPTA copoly-

mers (#9-11).

polymethacrylates and indicates the decomposition of the thermally weakest ester groups [11]. Depolymerization processes that can occur at low temperatures are likely to contribute to this decomposition. In the case of the HEMA copolymers, the onset temperature increased with increasing silica content in the copolymer. An increase of onset temperature with increasing silica content was observed for polyacrylonitrile (PAN)/silica sol-gel glasses where the onset temperatures were somewhat higher. This is because of the higher thermal stability of PAN which is able to convert to thermally stable species during heating [6]. For HDDA and GPTA copolymers, unlike HEMA copolymers, the onset decomposition temperature increases with the increase of the organic polymer content. This means that the crosslinking of the organic polymer contributes to the ultimate thermal stability of the hybrid network. It is well known that crosslinked polyacrylate networks exhibit higher thermal stability than linear polyacrylate chains. Furthermore, depolymerization processes are less likely to occur in crosslinked networks. The highest temperatures recorded for the onset of decomposition are those for Sample #8 for the HDDA copolymers and Sample #11 for the GPTA copolymers. The temperatures were 340°C or greater. This temperature is higher than the use temperature expected for these hybrids, suggesting that the materials would have sufficient thermal stability. Finally, to determine if there was evidence for structural changes in any of the copolymers that had been heat treated to 115°C for 24 hrs, FTIR spectra were collected of the samples before and after heat treatment. The unreacted vinyl group absorption of VTES at 1608 cm -1 was not changed during heat treatment.

82

Wojcik and Klein

structure of the organic component. In H D D A and GPTA copolymers, crosslinking of the organic polymers increased the thermal stability o f hybrid materials. Thermal treatment of the copolymers for 24 hrs at 115 ° resulted in a loss of weight that was not accompanied by structural changes in the copolymer chemistry. This weight loss increased the copolymer skeletal density only slightly without affecting its hardness or transparency.

~o

s~o

sooo

~

laeo

tTim

in,o

tm

lm

Acknowledgment

Wuwnumber (cm-l)

Fig. 8. Fourier-transformedinfrared spectra for HEMA copolymer

(#3) before and after heat treatment showing the only difference is the water content. Similarly the trace bands at 1640 cm -1 attributed to unreacted double bonds of acrylic groups were not changed by heat treatment. The only differences observed in the F T I R spectra before and after heat treatment were due to some dehydration of the heat treated samples indicated by a reduction of OH stretching of H-bonded H 2 0 and H-bonded S i - - O H around 3 2 0 0 3500 cm -1 and the appearance of more pronounced bands for SiO2 overtones at 1840 cm -1 . Also, the narrowing of the band in the vicinity of 950 cm -1 and the band at 800 cm -~ may indicate that the silicate network is losing adsorbed water, and, in effect, the absorption of the free or non H-bonded silanol groups is becoming more pronounced [10]. The spectra are shown for a H E M A copolymer in Fig. 8, and the differences are typical of most pairs of heat treated and unheat treated copolymers. The absence of major differences in the spectra supports the statement that the weight losses associated with the 115°C treatment were due to removal of volatile species that were not structurally part of the copolymer.

4

Conclusions

The thermal stabilities of the three types of threecomponent inorganic/organic copolymers depend strongly both on the silica content and the type of organic monomer. The least thermal stability was exhibited by H E M A copolymers because of the linear

The authors acknowledge the assistance of the staff of the Fiber Optic Materials Research Program (FOMRP) and the Center for Ceramic Research (CCR), Advanced Technology Centers funded in part by the NJ Commission on Science and Technology.

Notes 1. VTES from Aldrich Chemical Company, Inc., Milwaukee, WI. 2. Dynasil A, Dynamit-Nobel, Huls Co., Piscataway,NJ. 3. HEMA, HDDA, GPTA from Radcure Specialties, Louisville, KY.

References 1. H. Schmidt in Sol-Gel Optics, edited by L.C. Klein (Kluwer Academic Publishers, Boston, 1994), pp. 451-481. 2. K.J. Shea and D.A. Loy, Chemistry of Materials 1,572 (1989). 3. H. Huang, B. Orler, and G.L. Wilkes, Macromolecules20, 1322 (1987). 4. Y. Chujo and T. Saegusa in Advances in Polymer Science Vol. 100 (Springer-Verlag,Berlin, 1992), p. 12. 5. Y. Wei, D. Yang, and R. Bakthavatchalam,Materials Letters 13, 261 (1992). 6. Y. Wei, R. Bakthavatchalam,and C.K.Whitecar, Chem. Mater. 2, 337 (1990). 7. Y. Wei, D. Yang, and L. Tang, Makromol. Chem., Rapid Commun. 14, 273 (1993). 8. A.B. Wojcik and L.C. Klein, J. Sol-Gel Sci. Technol. 4, 57 (1995). 9. A.B. Wojcik and L.C. Klein, J. Sol-Gel Sci. Technol. 2, 115 (1994). 10. X. Li and T.A. King, J. Sol-Gel Sci. Technol. 4, 75 (1995). 11. B. Abramoff and L.C. Klein, SHE Vol. 1328 241 (1990). 12. M. Nandi, J.A. Conklin, L. Salvati Jr., and A. Sen, Chemistry of Materials 3, 201 (1991).