close

Вход

Забыли?

вход по аккаунту

?

Evaluation of the propylene content in ethylene-rich copolymers by infrared spectroscopy.

код для вставкиСкачать
Die Angewandte Makromolekula.re Chemie 32 (1973) 153-161 ( N r . 469)
From the Montecatini Edison S.p.A.,
Centro Ricerche Milano, Via G. Colombo 81, 20133 Milano and Centro Ricerche
Ferrara, 44100 Ferrara, Italy
Evaluation of the Propylene Content in Ethylene-Rich
Copolymers by Infrared Spectroscopy
By C. TOSI*and T. SIMONAZZI
(Received February 19, 1973)
SUMMARY:
An infrared method for the analysis of composition in copolymers of ethylene
with small amounts of propylene is described. It is based on the ratio between the
absorbance of the 7.25 p band and the product of the absorbance by the half-width
of the 6.85 ,u band. Spectra are recorded on molten samples to minimize crystallinity effects. The results obtained by this method are in good agreement with
those found by means of the near-infrared method by BUCCIand SIMONAZZI.
ZUSAMMENFASSUNG:
Eine Ultrarotmethode zur Analyse der Zusammensetzung von Copolymeren aus
Athylen mit geringen Mengen an Propylen wird beschrieben. Sie benutzt das Verhaltnis zwischen der Extinktion der Bande bei 7.25 p und dem Produkt der Extinktion mit der Halbwertsbreite der Bande bei 6.85 p . Die Spektren werden von
Proben im geschmolzenen Zustand aufgenommen, um Kristallinitiitseffekte zumindest herabzusetzen. Die Ergebnisse, die unter Verwendung diever Methode erhalten
werden, stimmen mit denjenigen nach der Methode von BUCCIund SIMONAZZI
im
nahen Ultrarot gut iiberein.
Introduction
A number of papers dealing with the infrared analysis of composition in
ethylene-propylene (C2 - C3) copolymers have appeared in recent years in the
spectrochemical literature ; they have been critically reviewed by one of the
authors of the present workl. None of them, however, gives a quite satisfactory
answer t o t,he problem of t h e composition determination in C2 - Cg copolymers
with low C3 content (smaller, say, than about 20% by weight): or a t least,
even those works t h a t report calibration curves covering the range up t o
polyethylene do not contain all information necessary t o carry out the analysis
so as t o attain the best possible results.
I n this paper we describe two independent procedures that, in our opinion,
are particularly befitting the evaluation of the C3 content in C2-rich copolymers.
* Present address: Centro Ricerche Bollate, Via S. Pietro 50, 1-20021 Bollate
153
C. TOSIand T. SIMONAZZI
Method Rased on the
A7.25p/A6.85/rRatio
The former of the two aforementioned procedures is based on the ratio of
the bands a t 7.25 and 6.85 p. A t low C3 content, the symmetrical methyl
bending a t 7.25 ,u is the most intense band characteristic of C3 units (we do
not take into account the methyl stretching bands, which are strongly overlapped by the methylene ones). The use of this band for all determinations on
branched polyethylenes2-6 stands in witness of its suitability for analytical
purposes : it should be remarked, however, that the absorptivity o f side methyls
is appreciably higher728 than the absorptivity of the terminal methyls of long
chain paraffins on which calibrations were mostly based. The 7.25 ,u band is
also employed in the method proposed by CORISH and TUNNICLIFFE~,which
is likely the one of widest applicability for C2 - C3 copolymers.
Owing to the difficulty of dissolution of C2-rich copolymers in solvents
suited for IR examination, the spectra must be recorded on copolymer films
pressed by die-casting. These films can be made very thin, on account of the
high absorptivity of the 7.25 ,u band, but in that cme they are rather inhomogeneous and their thickness is badly measurable; on the other hand, the use
of thicker sheets entails a consistent loss of transparency. It is thus convenient
to resort to the ratio technique, viz. to compare the 7.25 ,u band with a proper
band of not too different intensity: the choice is, so to say, fixed, in that we
take as a reference the 6.85 ,u band resulting from superposition of CHz and
asymmetrical CH3 bending vibrations.
A double-beam Perkin-Elmer Mod. 221 spectrometer is employed, with the
following settings :
Wavelength range: 6 t o 8 p.
Wavelength scale expansion: 10 cm/p.
Scan speed: 0.2 ,u/min.
Slit program : 9.27.
Samples are examined a t 160 "C in order to avoid any morphological effects
on the 6.85 ,u band (remember that, a t room temperature, this band splits in
polyethylene in two components, polarized along the two crystal axes).
Absorbance values are read with respect to base lines traced across the peak
shoulders, as shown in Fig. 1.
The calibration curve, based on a series of standard copolymers prepared
with 1%-labelled either ethylene or propylene, was first constructed (Fig. 2)
by plotting the A 7.25p /A 6.851" ratio against the C3 weight fraction. Calibration
points were rather scattered around this curve, whether resulting from different films of a given copolymer or even from the same film rotated to different
positions; in both cases, it was observed that the absorbance variations of the
6.85 ,u band were much more pronounced than those of the 7.25 ,u band, and
154
1 R-Spectroscopy of EP-Copolymers
w
U
z
c
m
WAVELENGTH (MICRONS 1
Fig. 1.
Spectrum of a copolymer containing 20.7% C 3 by weight, recorded a t
room temperature (left) and a t 160 "C (right). Arrows indicate the points
where the half-width of the 6.85 p band is measured. Note that the A7.251
A6.85 ratio is about the same in the two spectra (0.418 and 0.415, respectively), while LIA is strongly increased a t 160 "C (0.134 p, as compared to
0.089 p a t room temperature).
t h a t small values of A 6 . 8 generally
~~
corresponded t o a broadening of the band.
t o~
We therefore thought t h a t precision could be improved by ratioing A 7 . 2 ~
the product of the absorbance of the 6.85 p band by its half-width. For the
0.7
0.6
0.5
5
0.4
4
0.3
3
a2
2
0.1
1
0
10
20
30
40
50
C,,*rt-%
Fig. 2 .
Calibration curves for the determination of Cs in high-Cz copolymers. Full
points refer to models (hydrogenated poly-3 methyl alkenamers).
155
C. TOSIand T. SIMONAZZI
sake of simplicity (the instrument being calibrated linearily in wavelength),
the latter quantity is expressed in wavelengths instead of wavenumbers :
A l + A 6.85 is approximately 0.10 p a t room temperature and 0.14 p a t 160 “C.
Actually the 95% confidence limits relevant t o 8 labelled copolymers covering
the range of 14.2 t o 40.5:/, C3 (by weight) a n d t o two blends of t h e Cz-richest
copolymer with polyethylene are appreciably narrower in this case than by
using the A7.25r*/A6.85,, ratio. Furthermore, a smooth curve is better fitted t o
the former set of points than t o the latter one.
Table 1 lists our calibration data.
Table 1. Calibration points for the composition analysis of
(spectra recorded a t 160 “C).
3 137-25
3629-48
3629-44
3137-38
3 137-31
3297-55
3297-59
3 274-53
Blend No. 1
B!end No. 2
Polyethylene
40.5
32.9
32.0
30.7
20.7
18.2
18.0
14.2
11.Od
6.0d
0.7e
0.652
0.561
0.548
0.606
0.416
0.365
0.357
0.318
0.315
0.204
0.036
f 0.011
1.7
0.053
9.5
f 0.032
5.8
5 0.024C 4.0
f 0.018
4.2
& 0.037 10.0
f 0.015
4.3
0.015
4.8
f 0.022
6.8
f 0.013
6.4
f 0.007 18.5
Cg
4.80
4.07
3.92
3.80
3.02
2.62
2.59
2.25
2.00
1.26
0.28
- C3 copolymers
f 0.20
f 0.06
f 0.06
f 0.07
f 0.05
f 0.09
f 0.07
f 0.04
f 0.13
f 0.06
f 0.05
4.1
1.5
1.5
1.9
1.7
3.4
2.7
1.8
6.5
5.0
20 3
a Radiochemical analysis. Samples 3 137-25, 3629-44, 3 137-38, 3 137-31, and
3274-53 contain 14C-labelledethylene, and samples 3 629-48, 3 297-55, and 3297-59
contain 14C-labelled propylene.
h Aithough we have headed this column by the usual notation “2 a”, the standard
t 95% coefficient relevant to the
deviation has been multiplied by the STUDENT’S
number of replications (on the average 5) for each calibration point.
It is not clear why the absorbance ratio for this copolymer is so high. It is worth
remarking that the 6.85 p band is appreciably broader than in the other copolymers, so that division by the half-width of the band causes the point relevant to
this sample to fit the line passing through the other calibration points.
d Weighed average of the Ca content in polyethylene and in the copolymer 3 274-53.
e Evaluated by means of t,he ROHMER’S
procedure*, on the basis of the A7.25/A7.30
ratio, which gives for this polymer 0.23 CH3/100 CH2.
C
It is expedient t o compare these data to those obtained for three hydrogenated poly-3-methyl alkenamers, which are reported in Table 2.
156
IR-Spectroscopy of EP-Copolymers
Starting
poly-3-methyl
alkenamer
a
See footnote
b
c3
-47.25
20
A7.25
(%’a
A6.85 XdA
20
~
W-%)
A6.85
t o Tabie 1.
Hydrogenated poly-3-methyl alkenamers are models of Cz - C3 copolymers,
and their constitution is the more alike the constitution of the latter, the
higher is the alternatmg tendency of the copolymers, i.e. the lower is the product of reactivity ratios. Note, however, that even if it were possible to synthesize from ethylene and propylene a copolymer with rlrz = 0 containing the
two monomers in different amounts, say Cz :C3 = 3 : 1, the constitution of this
hypothetical sample would not coincide with t,hat of ths hydrogenated poly-3methyl octenamer, in that the latter has methylene sequences invariably
containing 7 CHz’s*,while the former would have methylene sequences of all
possible lengths, containing on the average 7 CHz’s.
I n the light of these considerations, it is to be expected t h a t the ratio
A7.25/&.85 in hydrogenated poly-3-methyl alkenamers will fit the line concerning C z - C3 copolymers better than the ratio A7.25/&85 x A l does: in fact
it is reasonable to think t h a t absorbances of IR bands depend essentially on the
copolymer composition, while half-widths are more sensitive to the sequence
distribution. A look a t Fig. 2 shows the rightness of this belief.
It is also worth comparing the upper curve of Fig. 2 to other calibration
curves for Cz - c3 copolymers based on the A7.25/&85 r a t i ~ g ? l l - ~This
~.
comparison, for not tco low C3 content, should hold irrespective of the recording
temperature: in fact we have found only minor differences in the A7.25/&85
* On condition t,hat 3-methyl cycloolefins, in the polymerization to poly-3-methyl
alkenamers, are coordinated t o the transition metal of the catalyst in such a
way as to give rise to a head-to-tail arrangement of the monomeric units (e.g. in
the case of 3-methyl cyclooctene
,CH3
CH3’
In the poly-3-methyl aikenamers used by us, after hydrogenation, as models of
Cz - C3 copolymers, this type of coordination is strongly favoured, representing
more than 90% of the transalkylidenation stepslo.
157
C. TOSIand T. SIMONAZZI
ratio a t room temperature and a t 160 “C (see e.g. the caption to Fig. 1). Data
for such comparison are given in Table 3 ; they were drawn from the following
sources :
Method by CORISH and TUNNICLIFFE
: Fig. 8 of Ref. 9.
Method by DAVISON
and TAYLOR:
Fig. 3 of Ref. 11.
Method by DRUSHEL
and IDDINGS:
Fig. 8 of Ref. 12 (see also Fig. 10 of Ref. 13).
Method by KISSINand CHIRKOV: eq. A7.25/A6.85 =
= [29 C, + 0.14 (1 - C,)] / [13.4
11.3 C,] (see also Table 3 of Ref. 1).
+
c3
(wt-70)
Ref.9
10
20
30
40
50
0.21
0.35
0.47
0.60
0.70
I
Ref.11
0.12
0.24
0.38
0.52
0.68
I
A7.25/&.85
Ref.12
._
0.44
0.53
I
Ref.14
0.15
0.28
0.41
0.53
0.65
I
This work
0.264
0.395
0.521
0.647
-
Data reported by various akthors differ strongly, and it is rather difficult
to account for such differences, which are much too large to be justified only
by the different choice of base lines. The variety of catalyst systems employed
to synthesize copolymers should not play an essential role here (according to
CORISH and TUNNICLIFFE~,
polyethylene, hydrogenated natural rubber, polypropylene and Cz - C3 copolymers fall on a unique curve). The type of spectrophotometer and the settings adopted to record the spectra probably exert an
appreciable influence on the A7.25/A6.85 ratio. Anyway, regardless of the
causes of the discrepancies pointed out in Table 3, it can be safely stated that
;his analysis must be carefully calibrated and cannot be directly transferred
from other works.
Bearing these limitations in mind, we can support the view of KIMMER
and
SCHMOLK
that
E ~ “the
~ two adjacent bands a t 7.25 and 6.85 ,u are the most
suited for routine analysis of Cz - C3 copolymers”.
Method Based on the A1.6gp/A1.76s Ratio
Up to now, C z - C 3 copolymers and terpolymers were analyzed in our
laboratories mostly by the near-IR method proposed by BUCCIand one of us
ten years agol6. As we already pointed 0 ~ t 1 7 , this method is preferred over
others utilizing bending vibrations because of its smaller dependence upon the
catalyst system employed to prepare the copolymers (in other words, upon the
sequence distribution). However, when the problem of evaluation of the C3
158
IR-Spectroscopy of EP-Copolymers
content in C2-rich copolymers was t o be tackled, we feared t h a t this method
could fail t o give reproducible results below ca. 25% C3 by weight, since a t
these compositions a n accurate measurement of the 1.69 p band absorbance
becomes very difficult, there being only a n imperceptible shoulder on the side
of a steep band, and the position where t o make the absorbance reading being
better determined by cutting the spectrum with a vertical line a t a fixed distance
from the 1.76 p reference band than by eye. Fortunately, with a high resolution
instrument such as the Cary mod. 14 spectrophotometer, the near-IR method
turnsd out t o be applicable also t o copolymers with a molar C3 content smaller
than ca. 15%. I n fact, extrapolation of the calibration curve formerly obtained
from standards with higher C3 content produced fairly accurate results for ( 1 )
five of the aforementioned labelled copolymers, (2) three mechanical mixtures
of C2 - Ca copolymers with polyethylene, and (3) the three hydrogenated
poly-3-methyl alkenamers (see Table 4). This agreement also proves t h a t t h s
met hod is barely influenced by C2 crystallinity.
Table 4.
Results of the near-IR analysis of
3274-53
3 297-59
3297-55
3 137-31
3 629-48
Blend No. 3
Blend No. 4
Blend No. 5
Hydrogenated poly-3-methyl octenamer
Hydrogenated poly-3-methyl decenamer
Hydrogenated poly-3-methyl dodecenamer
a
(2-2 - C3
copolymers.
Present8
Found b
14.2
18.0
18.2
20.7
32.9
25.4
19.9
12.4
33.3
27.3
23.1
14.2
17.6
17.6
20.7
32.7
24.5
20.0
12.1
32.7
26.5
24.1
Results of the radioactive analysis for the former five copoiymers. Mechanical
mixt.ures of polyethylene with the copolymers 3629-48 (blends No. 3 and 4)
and 3 137-31 (blend No. 5).
From the extrapolated portion of the calibration curve formerly employed for
the analysis of copolymers with C3 content higher than ca. 25 wt-yo.
The precision seems t o be independent of the copolymer composition :
Cn rwt-o/,)
16.5
22.1
32.0
standard deviation
0.59
0.32
0.44
159
C. TOSIand T . SIMONAZZI
The amount of copolymer generally used for this analysis is ca. 0.2 g, since
sheets ca. 0.5-1 mm thick are needed (possible inhomogeneity phenomena are thus
minimized); with use of the absorbance scale expander, the minimum amount of
sample can be lowered to ca. 0.03 g. A typical spectrum of a copolymer containing
20.7 wt-yoC3 is shown in Fig. 3.
Fig. 3. Near-IR spectrum (recorded at room temperature) of the copolymer of
Fig. 1.
Conclusions
The new analytic method based on the ratio between the absorbance of the
7.25 ,u band and t,he product of the absorbance and the half-width of the 6.85 p
band is probably the most suited for evaluation of the propylene content in
Cz-rich copolymers. Its greatest advantage consists of the very small amount
of sample required for the analysis (less than 0.01 grams are needed to obtain
sufficiently intense spectra), which allows to follow fractionation processes ;
examination of the samples in the molten state should ward off the danger of
scattered results due to copolymer inhomogeneity (e.g. because of the presence
of Cz microcrystallinity). Furthermore, consideration of the half-width of the
reference band permits, as seen, to avoid some aberrant values of the absorbance ratio, and thus t o place more confidence in the results.
Comparison with this new procedure does also lead to an experimental proof
of applicabilit,y of the near-IR method based on the A1.69,JA1.76w ratio t o
copolymers with low propylene content, though showing the critical dependence
of this latter method upon instrumentation.
160
IR-Spectroscopy of EP-Copolymers
1
2
3
4
5
6
7
8
9
10
l1
12
13
14
15
16
17
C. TOSIand F. CIAMPELLI,Advan. Polym. Sci. 12 in press
W. M. BRYANT
and R . C. VOTER,J. Amer. Chem. SOC.75 (1953) 6 1 1 3
E. J. SLOWINSKI
Jr., H. WALTER,
and R. L. MILLER,
J. Polym. Sci. 19 (1956) 353
M. ROHMER,
Z. Anal. Chern. 170 (1959) 147
H. A. WILLBOURN,
J. Polym. Sci. 34 (1959) 569
K . SHIRAYAMA,
S. I. KITA,and K. WATABE,
Makromol. Chem. 151 (1972) 97
M. R. BASILAand G.F. CRABLE,
J. Chern. Phys. 35 (1961) 306
L. C. KEY, F. M. TRENT,and M. E. LEWIS,Appl. Spectrosc. 20 (1966) 330
P. J, CORISH and M. E. TUNNICLIFFE,
J. Polym. Sci. C 7 (1964) 187
G. DALL’ASTA,
Makromol. Chem. 154 (1972) 1
S. DAVISON
and G. L. TAYLOR,
Brit. Polym. J. 4 (1972) 65
H. V. DRUSHEL
and F. A. IDDINOS,
Anal. Chem. 35 (1963) 28
H. V. DRUSHEL,
Critical Revs. Anal. Chem. 1 (1970) 161
Yu. V. KISSINand N. M. CHIRKOV,Private communication (1970)
W. KIMMER
and R . SCHMOLKE,
Plaste Kaut. 15 (1968) 807
G. BUCCIand T. SIMONAZZI,
Chirn. Ind. (Milan) 44 (1962) 262
C. TOSI,M. P. LACHI,and A. PINTO,Makromol. Chem. 120 (1968) 225
161
Документ
Категория
Без категории
Просмотров
1
Размер файла
419 Кб
Теги
contents, spectroscopy, propylene, evaluation, ethylene, copolymers, rich, infrared
1/--страниц
Пожаловаться на содержимое документа