This document summarizes a study using near-infrared Fourier transform Raman spectroscopy to monitor the bulk radical homo-polymerization of three commercial acrylate monomers: butylacrylate, hydroxylpropylmethacrylate, and laurylmethacrylate. The effects of reaction temperature, monomer additives, and oxygen presence on the polymerization rate were examined. For butylacrylate specifically, removing oxygen or the inhibitor hydroquinone monomethyl ether decreased the induction time but did not affect the propagation stage after gelation. Hydroxylpropylmethacrylate could not be reliably monitored due to insoluble particulate formation. Laurylmethacrylate results depended on temperature similarly to butylacrylate.

1. Spa Acta, Vol. CIA, No. 9110,pp. MS-1351,199l O%4-8539/91s3.tm+o.tm
Printed in Great Britain ~1991PergtlmmPnspk
Bulk radical homo-polymetisation studies of commercial acrylate
monomers using near infrared Fourier transfwm Raman spectroscopy
J. CLARKSON
Department of Chemistry, University of Strathclyde, Glasgow, U.K.
and
S.>,M.MASON and K. P. J. WILLIAMS*
BP Research, Sunbury Research Ckntre, Sunbury-on-Thames, Middlesex TW167LN, U.K.
(Received 7 March 1991; in nvisedfom und accepred 18 March 1991)
Ab&act-In our previous work we have used near infrared Fourier transform Raman spectroscopy to provide
a reliable, reproducible and quantitative method for in situ reaction monitoring of homo-polymetisation.
Model systems such as styrene and methyhnethacrylate were studied. In this present work three commercial
monomers have been studied, namely butylacrylate, hydroxylpropylmethacrylate and lauryhnethacrylate. The
effects of reaction temperature, monomer additives and ‘thepresence of oxygen have been elucidated. These
reaction variables have been shown to have a significant effect on polymerisaton rate in particularfor the
butykcrylate system.
INTRODUCTION
RAMAN spectroscopy has been used previously to monitor in situ the polymerisation of
vinyl monomers fl]. Our inital work centred on developing a method for the qualitative
assessment of the reactivity rates. More recently, we have developed a reliable and
quantitative method for monitoring the reaction process [2]. The polymerisation process
of both styrene and methyhnethacrylate monomers have been monitored at a range of
temperatures,. and percentage conversion plots against time have been obtained. These
data have been shown to be comparable with those obtained using classical dilatometric
methods [2].
The co-polymerisation of styrene and methylmethylacrylate has also been assessed [3].
Quantitative data have been obtained and it has been shown that the percentage
conversions for the individual monomers can be eludicated.
The method that we have developed for in situ reaction monitoring of vinyl monomers
uses the Raman band which occurs at ca 1640cm-‘. This is attributed to the symmetric
stretching vibration of the vinyl moiety [4]. The method simply involves measuring the
intensity of this particular band which decreases as the monomer is consumed. We have
shown, contrary to previous studies, that there is no need for normalisation of the
Raman data and that the peak intensity varies linearly with monomer concentration [5].
The advantages of Raman spectroscopy over infrared spectroscopy for this type of
study are well known [6]. Briefly, the reaction cell design is simplified with glass being a
suitable window and sample cell material. The Raman method is a scattering process and
as such is routinely used as a bulk technique for non-absorbing materials, which permits
relatively large samples volumes to be studied. Infrared (IR) spectroscopy, however, is
an absorption method which requires a thin film of sample. This requirement leads to a
high surface-to-volume ratio which can lead to the observation of “skin” effects, which in
turn are known to provide misleading kinetic data.
The purpose of this paper is to extend the Ramanmethod to study, quantitatively, the
polymerisation of commercially available monomers. The effects of the monomer
stabilisers and the presence of oxygen also have been investigated.
*Author to whom correspondence should be addressed.
Editor’s note: another example of an IT-Raman study of a kinetic system appears in Ref. [lo].
1345

2. 1346 J. CLARKSONetal.
Thermocouple
///‘///A/
Monomer
- Diletometry cell
J
hv laser
* Raman ecatter
7 whldow
- Silicone fluid
Fig.1. Insitu Ramandilatometrycell.
EXPERIMENTAL
Butylacrylate (99%) was supplied by Aldrich Chemical Co. and contained 15-50ppm of
hydroquinone monomethyl ether (MEHQ) as inhibitor. The initiator, 2,2’azo-bis(isobutyronitrile)
(AIBN) was supplied by BDH and used as received at a level of 0.5% w/v in all experiments. The
other acrylate monomers used, namely, hydroxylpropylmethacrylate (HPMA) and laurylmeth-
acrylate (LMA) contained MEHQ inhibitor at a level of cu 100 ppm.
Raman spectra were obtained using a Perkin-Elmer Fourier transform (IV)-Raman spectrom-
eter [7], fitted with a cooled (77K) germanium detector and employing a 180” backscattering
sampling geometry. The laser wavelength used throughout this study was the 1.064~m line from a
Spectron Nd:YAG laser. The incident laser power at sample was -lOOmW. Ail spectra were
recorded at 8 cm-’ resolution and took ca 30 s to acquire. The data collection and transfer from the
polymerisation was fully automated using a Perkin-Elmer Model 7700 Computer system.
Monomer samples together with initiator were placed into a partly silvered small glass bulb (ca
0.3 ml volume). The Raman dilatometry cell was transferred into an in situ reaction cell, the
temperature control for which is better than +OS”C. A schematic of the in situ reaction cell is
shown in Fig. 1.
RESULTSAND DISCUSSION
The different stages observed in a bulk homo-polymerisation process have been
discussed in full elsewhere [8]. The possible initiation, propagation and termination
Vinyl C=C
1 I I I I I ’
1300 1450 1650 1800 2000
Wavenumber I cm”
Fig.2. Ramanspectrum recorded from butylacrylate between 1275 and 2050 cm-‘.

3. Fig. 3.
NIR IT-Raman studies of acrylate homo-polymerisation
0 5 10 15 20 25 30 35 40 46
time I minutes
Percentage conversion curves obtained from butylacrylate at different
temperature.
1347
reactions have also been discussed. In our present study the depletion of the monomers
in the bulk homo-polymerisation was followed to approximately W-90%. The fractional
conversion was calculated using the equation:
X=Wlo- fWMo
Where [Ml0is the initial monomer concentration and [M] is the monomer concentration
at time (r). No correction was made for viscosity changes as the monomer polymerised.
The monomer concentration was calculated using the Raman peak observed at ca
1640cm-‘, which is attributed to the vinyl moiety of the molecule.
The FI-Raman spectrum obtained from butylacrylate between 1275and 2025cm-’ is
illustrated in Fig. 2. It is apparent that two intense Raman peaks are observed in this
region of the spectrum, the higher frequency band (174Ocm-‘) is assigned to the
carbonyl stretching vibration from the acrylate part of the molecule.
Figure 3 illustrates the results obtained from the polymerisation of butylacrylate
(containing MEI-IQ) ‘over a range of reaction temperatures. The curves represent the
effect of reaction temperature on the percentage conversion plots and, as a consequence,
the rate of formation of the polymer. The reaction was monitored at a range of
100 -
SO-
0 2k.%f,~AA
I I I I I I I I I I I I
0 4 0 12 16 20 24 20 32 36 40 44
Thna I minutes
Fig. 4. Percentage conversion plot from butyhwrylate at different reaction temperatures with
oxygen removed from the monomer.

4. 1348
100
90
50
70
h
r 50
g 50
840
)P
30
20
10
0
J. CLARKSONet al.
I I I I I I I I I I I I I
0 5 10 15 20 25 30 35 40 45 50 55 50
lime / minutes
Fig. 5. Percentage conversion plots from butylacrylate at different reaction temperatures with
oxygen removed from the monomers.
temperatures between 60 and 80°Cwith 2.5”Cincrements. It is apparent from these data
that both the reaction temperature and the other variables in our experiment can be
controlled with a high degree of precision. Small fluctuations in the temperature stability
can lead to very dramatic effects on the rate of the polymerisation process. As a
consequence the insitucell we have designed for these experiments, as stated previously,
has a temperature control of better than + 0.5”C.
The curves shown in Fig. 3 are typical of the results obtained from a bulk radical
homopolymerisation reaction. They are very similar to those obtained from other vinyl
monomers analysed using chemical methods and depict a classical shape for bulk
polymerisation. In particular there is an induction period, prior to the onset of the
gelation effect, which shows an almost linear relationship between conversion and time
followed by an abnormal acceleration in the polymerisation rate and then termination of
the reaction at a reduced rate. The initial phase of the polymerisation is known to occur
due to the formation of radical species which then propagate the reaction. The second
phase is shown as the Trommsdorf or gel effect which is characterised by an abnormal
acceleration in the polymerisation process [9]. This is brought about by the higher
I I I I I I I I I I I I
0 4 5 12 15 20 24 25 32 36 40 44
time I ‘C
Fig. 6. Percentage conversion curve of butylacryiate at 65°C showing (a) as received, (b)
inhibitor removed, (c) oxygen removed.

5. NIR FT-Raman studies of acrylate homo-polymerisation 1349
26
24
22
a20
t 18
f 16
- 14
E 12
F IO
o (a) With InhIbitor
0 (b) Without lnhlbitar
l (0) NW pw@
55 60 65 70 75 so
Temperature I ‘C
Fig. 7. Time taken to reach the gel point against temperature for (a) as received, (b) inhibitor
removed, (c) oxygen removed.
viscosity and restricted motions of the polymeric radicals at higher conversions, which
consequently leads to a slower rate for the termination of growing polymer chains. The
final phase in the polymerisation process occurs when the reaction mixture viscosity
greatly increases and the propagation step becomes diffusion-controlled, marked by a
slowing of the reaction.
MEHQ is added to many commercial vinyl monomers to prevent polymerisation
during storage and acts as a radical scavenger. The initial phase of the polymerisation is
due to the formation of radical species which then propagate the reaction. .Clearly, if the
reaction temperature is lowered the radical species are not readily formed and, as a
consequence, the MEHQ inhibitor can prevent the polymerisation from occurring.
Naturally when the inhibitor is consumed the reaction proceeds. The rate of formation of
radical species as a function of temperature is shown in Fig. 3. It is usual to observe a
small amount of monomer conversion prior to the gel point; however, our data suggest
that only a minimal amount of monomer has reacted before the gel point is reached. This
is attributed to the presence of MEHQ which hinders the polymerisation.
To assess further the role of the inhibitor in the polymerisation process two separate
experiments were devised. Initially, an attempt was made to remove the MEHQ from
o (a) With inhibitor
o (b) Wlthout inhibltor
l (c) Nitrogen purged
55 60 85 70
Tcmperatum I lC
75 80
Fig. 8. Time taken to reach 50% monomer conversion after the onset of the gel effect (a) as
received, (b) inhibitor removed, (c) oxygen removed.

6. 1350 J. CLARKSON et al.
Time I minutes
Fig. 9. Percentage conversion curves obtained from LMA at different reaction temperatures.
the butylacrylate monomer. An ion exchange resin (inhibitor removal column) was
purchased (Aldrich), which had a capacity of 3 1at 100 ppm, and was used to remove the
MEHQ from the butylacrylate monomer. In situ Raman data were obtained from the
polymerisation reaction over a range of temperatures. The percentage conversion curves
obtained from our analysis are shown in Fig. 4. The general shape of the curves is very
similar to that obtained from the monomer-containing inhibitor (see Fig. 3). However,
significant differences are observed in the initial stage of the reaction where it is clear that
the induction time is less when the inhibitor is removed.
Second, it has been shown that the presence of oxygen is crucial to the effectiveness of
the inhibitor. As a result we purged the monomer samples as received with nitrogen, to
remove any dissolved oxygen from the system prior to the addition of the initiator.
Figure 5 illustrates the percentage conversion curves obtained from these samples over a
range of reaction temperatures. The general shape of the curves is similar in shape to that
previously obtained; however, the reaction times observed have been greatly reduced.
Figure 6 depicts the data obtained from the polymerisation of butylacrylate at 65°C (a) as
received, (b) no MEHQ and (c) nitrogen purged. This clearly illustrates their respective
effects on the overall reaction time.
The data can be represented in a variety of ways to illustrate the way the inhibitor
restricts polymerisation. Figure 7 shows the time taken for the gel point to be reached for
a range of reaction temperatures using monomers treated as above. It is apparent from
the data that the absence of oxygen dramatically decreases the reaction time compared
with the as-received monomer. However, it is clear from Fig. 8 that the time taken for
50% conversion to be reached from the onset of the gel effect is largely independent of
monomer treatment. This observation is entirely in keeping with the role of the inhibitor
which, in the main, affects the initial stage of the reaction and not the propagation stage.
This latter stage is known to be dominant after the onset of the gel point.
In summary, we have shown that the Raman method can be used to study the
polymerisation of commercial butylacrylate monomer. The data obtained using our
dilatometry cell provide quantitative information from which a meaningful comparison
between data sets is possible The effect of small changes in reaction temperature, the
removal of the monomer stabiliser and the removal of oxygen from the system have been
shown to dramatically alter the polymerisation rates.
As part of our study to assess a more general applicability for this type of system we
have also investigated other commercial monomers, namely, HPMA and LMA. The
Raman data obtained from the HPMA monomer were poor and not reproducible. In the
initial stages of the polymerisation the data obtained were consistent with our previous
studies. However, as the reaction proceeded the intensity measured from the Raman
bands became erratic. The monomer chosen for this work is known to produce “hard”
polymers which are cross-linked and insoluble in their monomer. As a consequence after

7. NIR FT-Raman studies of acrylate homo-polymerisation 1351
cu 5% conversion the polymer was seen to come out of solution and form particulates in
the reaction cell. It is thought that the presence of these particulates intermittently blocks
the laser beam and gives rise to poor quality and non-reproducibility in our data. Whilst
we have not been able to generate any kinetic information from this system using our
current experimental set-up, we are currently assessing alternative methods to overcome
these problems.
LMA also was studied and the results obtained are shown in Fig. 9. This particular
monomer is a far weaker Raman scatterer than those previously studied and the data
recorded had a lower signal-to-noise ratio. This explains the greater degree of scatter in
the data points plotted in Fig. 9. Clearly, the quality of the data will effect the precision
of our percentage conversion plots. In the case of the butylacrylate system we have
estimated a measurement error of cu f 3%, whereas for the LMA the error is
substantially greater at ca + 10%. However, the percentage conversion plots from this
particular monomer system are similar in shape to those observed from butylacrylate. As
with the butylacrylate it has been possible to show the effect of reaction temperature on
reaction time. The effects of the additives on monomer stability have been omitted from
this study and will be addressed at a later date.
CONCLUSIONS
We have shown that the Raman method for in situ monitoring of homo-polymerisation
reactions from commercial acrylate monomers is reliable, reproducible and quantitative.
The effect of reaction temperatures and monomer additives can be elucidated from the
corresponding change in reaction time. Whilst the method is widely applicable to
commercial systems it lacks universal usage. The main problem we have encountered is
one of polymer insolubility which in turn produces erratic, non-reproducible data.
However, minor cell modifications should overcome this problem.
Acknowledgement-Permission to publish this manuscript has been given by British Petroleum Research.
REFERENCES
[l] I. S. Biggin, H. J. Bowley and D. L. Gerrard, in Time-resolued VibrationalSpectroscopy, p. 194.
Springer, New York (1985).
[2] K. P. J. Wiiams, S. M. Mason and D. L. Gerrard; manuscript in preparation.
[3] K. P. J. Williams, S. M. Mason and D. L. Gerrard; manuscript in preparation.
[4] W. G. Fateley, G. L. Carlson and F. E. Dickson, Appl. Spectrosc. 22,650 (1968).
[5] K. P. J. Williams and S. M. Mason, Trends Analyt. Chem. 9, 119 (1990).
[6] E. Gulati, K. McKengue and K. Y. S. Ng, Macromolecules 17, 1822 (1984).
[7] A. Crookell, P. J. Hendra, H. M. Mould and A. J. Turner, J. Raman Spectrosc. 21,85(1990).
[8] W. Y. Chui, G. M. Carratt and D. S. Soong, Macromolecules 16,348 (1983).
[9] V. E. Trommsdorf, H. Kohe and P. Legally, Macromolec. Gem. 1, 169 (1948).
[lo] A. M. Tudor, C. D. Meiia, M. C. Davies, P. J. Hendra, S. Church, A. J. Domb and R. Langer,
Spectrochim. Acta 47A, 1335 (1991).