PHÂN TÍCH NHIỆT PMMASILICA GEL TIẾNG ANH

Seediscussions,stats,andauthorprofilesforthispublicationat:
/>
ThermalanalysisofPMMA/gelsilicaglass
composites
ArticleinJournalofSol-GelScienceandTechnology·January1996
DOI:10.1007/BF00401038

CITATIONS

READS

30

41

Someoftheauthorsofthispublicationarealsoworkingontheserelatedprojects:

InvestigatingcertaintypesofdirectBrain-MachineInteractionViewproject

AllcontentfollowingthispagewasuploadedbyFotiniPallikarion14September2017.
Theuserhasrequestedenhancementofthedownloadedfile.


Journal of Sol-Gel Science and Technology 7,203-209 (1996)
@ 1996 Kluwer Academic Publishers. Manufactured in The Netherlands.

Thermal Analysis of PMMA/Gel Silica Glass Composites
FOTINI PALLIKARI-VIRAS
Physics Department, University ofAthens, Zografos, Panepistimiopolis, Athens 157 84, Greece
XIAOCHUN LI AND ‘IERENCE A. KING
Department of Physics and Astronomy, Schuster Laboratory, University of Manchestel; Manchester Ml3 9PL, UK

Received June 13, 1995; Accepted February 12, 1996

Abstract.
Thermal analysis of poly-methylmethacrylate (PMMA) impregnated porous gel silica glasses confirms
that the PMMA chains form hydrogen bonds with the pore surface silanol groups. The adopted conditions for the insitu polymerisation result in about 4% of residual monomers trapped in the polymer, most of them in the amorphous
structure. The polymer and monomer mixture takes up the whole of the free pore volume. Most of the residual
monomer polymerises during the DSC scans above the glass transition temperature providing an excellent probe
for the weak glass transition. Polymerisation in the gel silica glass medium affects the glass transition temperature,
the length of polymer chains, and the degree of polymerisation.
Keywords:
crylate

1.

gel silica glass, thermal analysis, glass transition temperature, polymerisation, poly-methylmetha-

Introduction

Composites based on porous gel silica glasses impregnated with PMMA have been prepared by the formation of a porous sol-gel glass followed by in-situ
polymerisation from the monomer [l-4]. These composites are used as host media in optical applications
such as lasers, optics, sensors and nonlinear optical
studies where matching of refractive index and reduced
scatter is achieved and also where optically active components can be incorporated [5, 61. The composites
have improved optical quality and mechanical strength,
which is valuable for these applications. The in-situ
polymerisation conditions leads to the presence of a
small percentage of residual monomer, There is evidence that hydrogen bonding, between the pore surface silanol groups and the polymer side chains, affects
the stretching vibrational frequency of the C=O ester
carbonyl group [3]. It is noted that the degree of insitu polymerisation of MMA monomer inside the silica
glass pores, is further affected by the silica gel pore

environment. The presence of residual monomer will
affect the optical, thermal and mechanical properties
of the composite. It is also of concern, that the composite’s structure should provide an inert host for the
optically active dopants, and that heating of the samples
will not alter their physical characteristics, for example by encouraging monomer to polymer conversion to
take place, which changes the properties while in use.
The effect of residual monomer on the thermal properties of the composite glass is, therefore, of great interest and will be explored in this work by thermal
analysis. The hydrogen bonding hypothesis between
the polymer and pore surface functional silanol groups
will be further investigated by simple considerations
regarding the molecular dynamics in the composite
system on the basis of thermal analysis data. Previous thermal analysis reports [4, 71 have shown that
the molecular dynamics of the polymer and the silica
glass, typically of lo-15 nm pore size, affect the glass
transition temperature of the polymer. As a result the
glass transition temperature of the PMMA decreases


204

Pallikari-Viras, Li and King

by about 20°C when it polymerises inside the silica
glass pores. Such a behaviour will affect the temperature range of applications of these composite materials
and, similarly, their performance. It is, therefore, of
interest to examine the previous thermal analysis observations, in a range of composites of smaller pore
size, in this case of the order of 5 nm.
It will be shown that, unlike in previous observations,
the glass transition temperature of the polymer in the

glass pores was found to be higher than that of the
bulk. An explanation of this difference will be given.
For analysis purposes the benefits of the presence of a
small percentage of residual monomer, as a probe of
weak structural transformations, will be discussed.

source was a Coherent Innova 90 argon laser operated
at 514.5 nm and 700 mW. Typical operating conditions were bandwidth = 4 cm-‘, scanning increment
= 1 cm-‘, integration time = 2 s. The frequency scale
was calibrated by reference to the spectra of L-cystine
and the 812 cm-’ line of PMMA. Thermal analysis
measurements of comparable mass samples (ranging
from 2.28 to 4.74 mg) were performed with a DSC
220°C Seico Instruments Inc calorimeter. The heating
and cooling rates were maintained at 10 degrees/min.

2.

Thermal analysis on a number of PMMA/gel silica
glass composites and PMMA bulk samples show repeatable characteristics in the region from 25 to 250°C
Figs. 1 to 3. There are three temperature regions where
exothermic peaks are observed: the first and stronger
peak having a maximum at about 116°C in the PMMAonly samples and at about 120°C in the PMMA/gel silica glass composites, the second at about 160°C and
the third at about 200°C Table 1. These are attributed,
as it will be shown later, to the polymer in the composite.
The most prominent feature of the DSC curves, the
strong exothermic peak, is found at the lower temperature region near the glass transition temperature to
amorphous PMMA (105C). The samples’ glass transition temperatures have been recorded in the second
heating (and previous cooling) run, occurring at about
105°C for the PMMA-only, Figs. 2(d), 3(c) and about

114°C in the composite samples, Figs. 2(c), 3(d). It was
observed that the onset of the first exothermic peak, Ti ,
Figs. 2 and 3 coincides with the corresponding samples glass transition temperature. In fact, the composite

3.
3.1.

Experimental

Porous gel silica glass samples partially densified at
temperatures of 600°C and 700°C and 800°C were supplied by Geltech Inc. (USA). The procedure used for
the in-situ polymerisation of MMA in silica glass is
described elsewhere [3,8]. The polymer samples were
polymerised (a) inside the gel silica glass pores, for
example, labelled PMMA/silica 600 for sol-gel glass
samples densified to 600°C and similarly for the 700°C
and 800°C densified samples, (b) polymerised external
to the gel silica glass but inside the same polymerisation
vessel, e.g., labelled PMMA only-600 for conditions
corresponding to the 600°C glass sample.
The instrumental and measurement details for the
gel silica glass pore characteristics obtained by nitrogen adsorption and for the polymer molecular weights
obtained by GPC (see Table 2), are described in reference [3]. Raman spectra were recorded by means of
a computer operated Spex Ramalog spectrometer fitted with a 1403 double monochromator and a 1442U
third monochromator in the scanning mode. The light
Table

Results
Thermal Analysis Measurements

1, Data evaluated from DSC and nitrogen adsorption measurements.
Tl

T2

T3

(“C)

(“C)

-ho
J/g

-h,

(“C)

Sample

fl

fl

fl

f1.5

f1.5

PMMA only-600
PMMA only-700
PMMA only-800
PMMAkilica 600
PMMAkilica 700
PMMA/silica 800

104
105
103
118
106
113

142
176
160
159
164
160

191
200
200
200
200
200

22.0
28.0

20.0

J/g

11.7
14 7
10.2

b/H

n=

(%)

k/h/>

f0 03

10.20

Vd/(l + Vd)

39
49
3.6

053
0.52
0.51

0.514
0.484
0.481


Thermal Analysis

I

I

I

50

100

150

I

200

205

i

Heatingtemperature(“C)
Figure I.
DSC first heating curves of PMMA/gel

silica glass composite samples from substrates
densified at: (a) 600°C. mass of
composite
m = 3.14 mg, (b) 7OO”C, m = 4.15 mg, (c) 800°C
m = 2.85 mg

50

100

150

200

0

Heatingtemperature(“C)
Frgure 3. DSC heating curves of: (a) PMMALsilica
600. m =
3.14 mg, first run, (b) PMMA only-600, mass of bulk, m = 2.28 mg,
first run, (c) PMMA only-600,
second run, (d) PMMA/silica
600.
second run.

samples glass transition and Tt onset temperatures shift
to higher values by the same amount.
The two other peaks found in the DSC curves, having maxima at T2 and T3, are generally much weaker
in intensity and broader and appear at about the same
temperature for both PMMA-only and PMMA/gel silica glass composite samples. These temperatures are

near the melting points of the crystalline isotactic
(160°C) and syndiotactic (200°C) PMMA [9]. The
consecutive heating (and previous cooling) curves do
not display the above exothermic peaks. No exothermic peaks were observed in the DSC scans of silica
only samples in the temperature region of 25250°C.
TY
a

3.2.
,
I
50

100

Molecular Weight Measurements

,\\
150

200

Heatingtemperature(“C)
Figure 2. DSC heating curves of: (a) PMMA/silica
800, m =
2.85 mg, first run, (b) PMMA only-800, m = 4.74 mg, first run, (c)
PMMA/silica
800, second run, (d) PMMA only-800,
second run.

The samples’ molecular weights determined by GPC
are shown in Table 2 with nitrogen adsorption measurements of surface area and pore size. The GPC
measurements indicate that the polymer chains inside
the glass pores grow smaller than the chains in the corresponding PMMA-only sample. Although the pore


206

Pallikari-Viras, Li and King

Table

2.

Data evaluated

Surface
area (m*/g)
f20

Sample
PMMA

from nitrogen

adsorption

Pore volume
(cm343
f0.02

and GPC measurements.
Mean pore
diameter
(nm)

only-600

M,
flO%

Mw
flO%

290,000

840,000

PMMAkilica

600

605

0.89

5.9

4,000

56,000

PMMAkilica

700

587

0.79

5.4

7,000

73,000

PMMA/silica

800

557

0.78

5.6

6,000

67,000

environment would create polymerisation conditions
under excess pressure, a situation which will encourage the formation of higher molecular weights [lo], at
the same time it will reduce the propagation mobility
of the free radicals. Since there is less available space
inside the pores the free radical propagation is restricted and smaller polymer chains may be expected
to be formed. The molecular size will also depend
on the degree of polymerisation, which is lower in the
PMMA/gel silica glass composites, according to the
Raman data.
The molecular size of the polymer inside the glass
pores, compared with those in the PMMA-only samples, will depend on one or more of the above factors,
i.e., the early termination of the polymerisation due to
the effect of glass on mobility, the excess pressure effect and incomplete polymerisation due to the restricted
space. The resulting molecular size will depend upon
which factors most dominate the polymerisation process. It should be noted, however, that the reverse phenomenon (i.e., the polymer in PMMA/gel silica glass
samples having higher molecular weight than in the respective PMMA-only samples) has been observed in
previous work, for the case where the glass pore size
was smaller (d = 4.4 nm) [3] and where there were
probably excess pressure effects influencing predominantly the chain length.

relative intensity of the two peaks may be analyzed as
an approximate measure of the percentage of monomer
in the monomer/polymer mixture. The Raman spectra
show that not only is polymerisation incomplete in all
samples, but also that the degree of conversion is lower
in the PMMA/gel silica glass composites.

3.3.

4.

Raman Spectroscopy Measurements

A typical Raman spectrum of a PMMA-only 800 sample is shown in Fig. 4 over the region 1200-1800 cm-‘.
The most interesting features in this frequency region
for this study, are the two peaks at -1642 cm-’ and
w 1725 cm-‘. The peak at 1642 cm-’ indicates the presence of monomer (MMA), H2C=C(CHs)COOCHs,
which has not been converted to PMMA during polymerisation. The peak at 1725 cm-’ is assigned to
the C=O stretching mode of the ester carbonyl. The

B
2
8
S

,*I

10

Raman shift (cm-‘)
Figure 4. Raman spectrum of sample
the 1642 cm-’ and 1725 cm-’ lines.

PMMA

only-800,

showing

Discussion

The DSC exotherms are attributed to a polymerisation reaction. There are several reasons that support
this explanation. Ran-ran spectroscopy and NMR data
show that the PMMA is not fully polymerised in either the PMMA/gel silica glass composites or the bulk
PMMA, prepared under the same temperature, pressure and initiator concentration conditions [3]. The
other possibility, that the peak may be due to crystallisation reaction, can be excluded on the basis that no


Thermal Analysis

melting peaks are observed around the expected melting points, releasing the same amount of energy as
the exothermic peaks. A third reason in favour of the
polymerisation against crystallisation interpretation, is
that the exothermic peak disappears upon consecutive
heating (and cooling) of the samples, Figs. 2 and 3. The
glass transition is, however, present in the re-heating
(and cooling) curves. This behavior is to be expected
if a near 100% conversion of monomer to polymer
has occurred upon first heating, rather than crystallisation.
We assume, therefore, that further polymerisation
occurs during the first DSC heating run, during which
the monomer is more fully converted into polymer.
There is, nonetheless, the presence of the two weak
exothermic peaks, which notably occur around the
melting point temperatures of the iso- and syndioPMMA. The above observations lead to the following
interpretation. The in-situ polymerisation temperature
of 60°C of our samples is well below the glass transition
temperature, which is 105°C for amorphous PMMA
[9]. As the polymer concentration increases during polymerisation the reaction medium becomes more
viscous at T < T,, the polymerisation rate decreases

and the conversion is forced to stop before it is fully
completed (cf Trommsdorff or glass effect) [lo]. As
a result, there will be monomer units trapped within
the polymer chains, the greatest proportion of them in
amorphous polymer regions. While the temperature
rises near the glass transition, during the first heating
run, the amorphous polymer chains become flexible,
the monomer is freed and the prior incomplete polymerisation now continues until almost all monomer
trapped in the amorphous region is fully converted.
The polymerisation would be expected to commence at
the glass transition of the monomer/polymer mixture,
which will be slightly lower (due to presence of unconverted monomer units acting as plasticiser [Ill) than
the glass transition of the pure polymer. This is confirmed, as the glass transition observed in there-heating
(and cooling) curves of the samples coincides with the
onset temperature of their first exothermic peak, Figs. 2
and 3. The weak exotherms at about 160°C and 200°C
are likely to have the same origin as that of the strong
exothermic peak, i.e., they are polymerisation peaks,
but come from fewer monomers trapped in areas of
different chain structure. It is obvious that, although
the incomplete polymerisation is not desirable, as far as
the reinforced silica glass optical quality is concerned,
it serves, however, as a sensitive probe of the weak

207

glass transition. More information can be drawn from
the polymerisation peaks, about (a) the percentage of
monomer to polymer in the in-situ polymerisation, (b)
the extent by which the pores are filled with the polymer, and (c) providing some insight into the polymer

chain dynamics within the gel silica glass pores.
4.1.

Estimation of the Extent of Polymerisation

The degree of conversion of monomer to polymer in
the samples can be estimated from the enthalpies of
polymerisation. The enthalpy of conversion of liquid
monomer to amorphous or (slightly) crystalline polymer is H = -560 J/g of polymerised monomer (91.
We will assume that a fully converted monomer under
the conditions met in this work, has the same enthalpy
of polymerisation H.
If m, is the mass of monomer which has not been
converted and m,, is the mass of the polymer in the
(pre-heated) PMMA-only samples then the ratio of
monomer will be: mm/(mp + m,). If hb(Jg-‘) is
the enthalpy of the bulk PMMA-first
polymerization
peak (most monomers are trapped in the amorphous
region), then the ratio hh/H will yield the percentage of monomer in the amorphous part of the polymer,
or, approximately, in all the polymer. The estimated
percentage of unconverted monomer in PMMA-only
samples from the thermal analysis data, hb/H, is of
the order of 4%, Table 1. This value is of the same
order as previous estimation from NMR data of similar
samples 131.
4.2.

Estimation of the Polymer Content in Glass

The enthalpies of polymerisation of the PMMA/gel
silica glass composites combined with pore geometry
characteristics can yield information about the extent
of filling of the glass pores with polymer.
If V = volume of pores/mass of glass, M and m represent the mass of the glass and polymer is the composite, respectively, and II = m/(M + m), is the mass ratio of polymer content in the composite samples, then,
V 2 volume of polymer/mass of glass, since, volume
of pores 2 volume of polymer in glass. If the density of the polymer is d = mass of polymer/volume of
polymer, then:
Vd
Vdtl

=-l-l
-

(1)


208

Pallikari-Viras, Li and King

The value of II can be estimated from the enthalpies of
polymerisation, h, and hb, of the composite and bulk
PMMA samples, respectively as
(2)
Also correcting the value of h, for the mass of the
polymer, yields the enthalpy of the PMMA-only polymerisation peak,
h c M+m

m

=h

h

(3)

Table 1 shows that the data of the last two columns
satisfy relation (1) where the polymer density is taken
as d = 1.19 gcmp3 [9]. According to (1) and the data
of Table 1, the pore volume appears to be fully taken
up by the polymer, (which is about 50% of the composite), since the values of the two last columns are in
agreement. The mass of the small percentage of liquid
monomer present may give a noticeable polymerisation peak, but it is negligible as far as its volume in the
pores is concerned.
4.3.

Chain Dynamics and Sutjace Phenomena

It has been suggested in earlier work that H-bonding
occurs between the polymer side chain and the silanol
(SiOH) groups inside the glass pores [3]. The Hbonding hypothesis was based on Raman spectroscopy
data. The DSC results of this work support the above
hypothesis. The shift of Tg of the polymer in composites to higher temperatures, taking into account the
expected shift to lower temperatures due to the presence of the monomer, is the result of the hindering of
the segmental motions of the macromolecular chain
due to the H-bonding. These increased excess surface forces, through the pressure that they exert, reduce
the flexibility of the polymer chains and lead to higher
glass transition values. H-bonding effects are predominant in glasses of very small pore sizes. In larger pore
sizes other effects (e.g., the lower degree of cross linking) may cause an overall decrease of the transition

temperature.
In gel silica glass samples having very similar surface areas, the average pore size may play an important
part in the molecular interaction, such as the H-bonding
between the silanol groups and polymer chains [12]
and, therefore, affect the glass transition temperature

of the polymer in the pores. The densification temperatures of 600, 700 and 800°C of the measured silica
glasses are relatively close and can result in similar
pore sizes, as shown in Table 2.
Heating the composite above 200°C will result in a
small degree of depolymerisation of PMMA. The process is slow compared to the thermal treatment rates
[ 131. The few monomers created will be in equilibrium
with the rest of the polymer inside the sealed vessel.
The subsequent cooling of the sample to ambient temperature, will result in a near fully converted polymer.
The glass transition step is, therefore, observed in the
second heating run.
It is important for a number of applications that the
glass transition temperature of PMMA in composites
is not affected by thermal treatments. Higher glass
transition temperatures will be favourable. In that
sense, the in-situ polymerised samples have slightly
improved stability as well as optical properties, having increased their glass transition by approximately
10-15 degrees.
5.

Conclusions

Thermal analysis measurements support the hypothesis of hydrogen bonding between polymer side groups
and surface silanol groups in PMMA/gel silica glass
composites as suggested in earlier work. The Hbonding causes the glass transition of PMMA to increase by 10-15 degrees and, therefore, improves its

thermal stability against deformation while heating
leading to useful improvement in its optical quality. Allowance should be made for the small depression of
the glass transition due to the presence of residual
monomer. The observed glass transition temperature
of the bulk PMMA samples are characteristic of amorphous PMMA.
The polymer chains inside the silica glass were found
to be smaller than the chains of the bulk PMMA prepared under the same conditions of polymerisation.
As the mobility of the polymer chains and of the
initiator inside the glass pores decreases, due to reduced available space, smaller molecular chains are
formed.
The whole volume of the pores is practically taken
up by the polymer and contains about 4% of unconverted monomer, which further polymerises during the
first heating DSC run. The full polymerisation of the
residual monomer takes place above three characteristic temperatures: the glass transition temperature of


Thermal Analysis

the amorphous polymer and the melting points of isoand syndiotactic PMMA. The last two polymerisation
peaks are due to monomer trapped within crystallites
of iso- and syndiotactic chains. Therefore, the unconverted monomer provides a useful, sensitive and simple
probe, in a DSC scan, to pick up the weak glass transition and the presence of crystallinity in the amorphous
polymer.
Acknowledgments
This work has been financially supported by the European Science Exchange Program of the Royal Society
and by the University of Manchester. We express special thanks to Dr Colin Booth and Dr Peter M. Budd
of the Chemistry Department, University of Manchester, and Professor Alan H. Windle of the Department
of Material Science and Metallurgy, Cambridge University, for elucidating discussions on polymer properties.

View publication stats

209

References
1. E J.A. Pope and J.D. Mackenzie,
Mat. Res. Sot. Bull. 12, 29,
(1987).
2. E.J.A. Pope, J. Sol-Gel Sci. &Tech. 2, 717 (1994).
3. X. Li, T.A. King, and E Pallikari-Viras,
J. Non-Cryst.
Solids
170,243 (1994).
4. E.J.A. Pope, M. Asami, and J.D. Mackenzie,
J. Mat. Res. 4,
1018 (1989).
5. M. Rahn and T.A King, Appl. Opt. (1996), 34, 8260.
6. G J. Gall, X. Li, and T.A. King, J. Sol-Gel SCI & Tech 2, 775
(1994).
7 B. Abramoff
and L.C. Klem, m Chemical
Pmcessq
of Advanced Materials,
edited by L.L Hench and J.K. West (Wiley,
New York, 1992). p. 73
8. X. Li and T.A. King, J. Sol-Gel Sci. & Tech 4,75(1994)
9. J. Brandrup
and E.H. Immergut
(eds.), Polymer
Handbook
(Wiley, New York, 1989).

10. G. Odian. Prmciples
of Polymerisation
(Wiley,
New York,
1981).
11. H.-G. Elias, Mucmmolecules
(Wtley, London, 1977), p. 412.
12. L.L. Hench and J.K. West, Chem. Rev. 90, 33 (1990).
13 S. L. Madorsky,
Thermal Degradation
of Orpmc
Polymers
(Wiley, New York, 1964).