Analysis of additives in polymers by thin-layer chromatography coupled with Fourier transform-infrared microscopy

Автор: Пользователь скрыл имя, 05 Июля 2013 в 15:14, статья

Описание работы

Department of Applied Chemistry, RMIT University, Melbourne, Vic. 3001, Australia
Received 10 September 2001; received in revised form 20 January 2002; accepted 25 January 2002
Abstract
A new, fast and convenient method based on coupled thin-layer chromatography (TLC) and Fourier transform-infrared
(FT-IR) microscopy is developed to separate, detect and identify the additives in polymers. After the TLC development, the
analytes were transferred on to a barium fluoride (BaF
2
) salt plate via a special capillary technique and analysed by FT-IR
microscopy.

Скачать полностью (332.42 Кб) Сколько стоит заказать работу?
Работа содержит 1 файл

Vibrational Spectroscopy, Volume 30, Issue 2, 4 November 2002, Pages 147-156.pdf

— 396.71 Кб (Скачать)
Page 1
Analysis of additives in polymers by thin-layer chromatography
coupled with Fourier transform-infrared microscopy
 Wenxuan He
 a,*
 , Robert Shanks
 b
 , Gandara Amarasinghe
 b
 a
 Fujian Institute of Testing Technology, Fuzhou 350003, Fujian, PR China
 b
 Department of Applied Chemistry, RMIT University, Melbourne, Vic. 3001, Australia
Received 10 September 2001; received in revised form 20 January 2002; accepted 25 January 2002
 Abstract
A new, fast and convenient method based on coupled thin-layer chromatography (TLC) and Fourier transform-infrared
(FT-IR) microscopy is developed to separate, detect and identify the additives in polymers. After the TLC development, the
analytes were transferred on to a barium fluoride (BaF
 2
 ) salt plate via a special capillary technique and analysed by FT-IR
microscopy. The additives used for stabilization of polypropylene and the plasticisers used for poly(vinyl chloride) were
analysed as examples to illustrate this technique. The overall time taken for the experiment including transferring three marked
spots and then identifying them was about 20 min. An amount as small as 0.5 mg can be easily detected and identified. It was a
very convenient and reliable method to separate and evaluate complex additives for polymers without the interference from TLC
adsorbent, because of a special transferring and identifying method, which is suitable to FT-IR microscopy.
# 2002 Elsevier Science B.V. All rights reserved.
 Keywords: TLC; FT-IR microscopy; Polymer; Additive
 1. Introduction
Identification of additives in polymers is important
in forensic investigations, scientific research and qual-
ity control, etc. [1]. Currently, a range of additives is
being used for protection during processing, increasing
lifetime and improving the performance of products.
Although high performance liquid chromatography
(HPLC) is a very good method for separation of com-
plex additives, there are some difficulties in identifica-
tion of some components [2]. Gas chromatography
coupled with mass spectroscopy (GC–MS) is very
beneficial for isolating and identifying some additives
but when dealing with antioxidants and stabilisers, it is
not feasible in many cases owing to their decomposition
or lack of volatility.
Thin-layer chromatography (TLC) remains one of
the most widely used of all the chromatographic
techniques for simple and rapid qualitative separation
 [3–5]. The use of a detection technique such as,
 Fourier transform-infrared (FT-IR) spectroscopy or
nuclear magnetic resonance (NMR) spectroscopy
arises from the need to reliably identify components
or to obtain valuable molecular structure information
of the components separated by TLC [6,7]. Therefore,
combined TLC and FT-IR has been very attractive for
the analysis of complex additive in polymers. The
combination of TLC and FT-IR has recently been
applied in two ways: in situ or transfer method with
 Vibrational Spectroscopy 30 (2002) 147–156
 *
 Corresponding author. Fax: þ86-591-781-4856.
E-mail address: access@pub2.fz.fj.cn (W. He).
0924-2031/02/$ – see front matter # 2002 Elsevier Science B.V. All rights reserved.
PII: S 0924-203 1(02)00024-3
Page 2

diffuse reflectance infrared Fourier transform
(DRIFT) spectroscopy as a method of the evaluation
 [8]. However, the major difficulty encountered with in
 situ DRIFT–TLC method is the strong background
absorption of the adsorbent, which means that it can
only be used in particular spectral regions according to
the TLC adsorbent being used [9]. On the other hand,
the transfer of analytes in DRIFT–TLC is usually time
consuming [10].
Infrared microscopy is a powerful technique [11,12]
that combines the image analysis capabilities of
optical microscopy with the chemical analysis cap-
abilities of infrared spectroscopy. The combination of
these two techniques allows infrared spectra to be
obtained from microspectroscopic-sized samples.
With the assistance of microscopy, samples as small
as 0.01 mg or even less (depending on the infrared
absorption characteristics of the components of inter-
est) can be easily located and detected. Therefore,
with all of these characteristics, FT-IR microscopy
is a good technique of identification for components
separated by TLC.
The aim of this research is to apply FT-IR micro-
scopy in combination with TLC for the identification
of additives in polymers. It was very convenient and
reliable because of the effective transferring technique
and identifying method that were suitable to FT-IR
microscopy. The advantage of this method is that it can
be used to evaluate each separated component by
various TLC sheets without the interference of TLC
adsorbent. In addition, it is not a time consuming
method. To the best of our knowledge, this coupling
technique has not been previously published. In this
study, we analysed antioxidants and ultraviolet stabi-
lisers used for polypropylene (PP) and the plasticisers
for PVC as examples to show how this technique can
be applied. The technique can be extended to analyse
various other additives in variety of polymers.
2. Experimental
2.1. Materials
TLC analysis was performed on TLC plates
(5 cm  10 cm) precoated with a 250 mm layer of silica
gel 60 F254 on a glass support (Merck, Darmstadt,
Germany). The solvents (methanol, acetone, formic
acid, diethyl ether and toluene) and plasticisers (dii-
sooctyl phthalate and dibutyl phthalate) were of analy-
tical reagent grade and purchased from Science Supply,
Vic., Australia. Antioxidant (Irganox1010) and ultra-
violet stabilisers (Tinuvin770 and Chimassorb 119 FL)
were from Ciba Specialty Chemicals Inc., Australia.
Their chemical structures are shown in Fig. 1.
2.2. Analytical procedures
2.2.1. Preparation additive solutions, standards
and other solutions for detection limit test
A simulated solution of a mixture of additives (I)
was prepared by weighing 0.1 g of each Irganox1010,
Tinuvin770 and Chimassorb 119 FL into a 10 ml
volumetric flask and diluted it with toluene. The
standard of solutions (0.5%) Irganox1010, Tinuvin770
and Chimassorb 119 FL were made by dissolving
0.05 g of each pure additive in a 10 ml toluene.
A series of dibutyl phthalate solutions were used
for the detection limit test. A stock solution of dibutyl
phthalate (0.5%) was prepared by dissolving 0.05 g
dibutyl phthalate into a 10 ml of diethyl ether. The
solutions of 0.1, 0.05 and 0.01% were obtained by
diluting the stock solution with diethyl ether 5, 10, and
50 times, respectively.
2.2.2. Chromatographic procedures
Before runs, the TLC plates were cleaned by devel-
oping in methanol–formic acid (1:1) solvent mixture.
The plates were dried in air, heated for 3 h at 160 8C
and stored in a desiccator until use. A 2.0 ml of
solution (I) was applied onto the plates with 5 ml
syringe. TLC sheet was then developed with a mixture
of toluene–diethyl ether (10:1) as the mobile phase to a
distance of 60 mm. The developed spot was marked
under UV light and R
 F
 value was 0.8. After the
removal of R
 F
 ¼ 0:8 spot, the second development
was carried out with the same TLC plate with a solvent
mixture of acetone–formic acid (4:6) as the mobile
phase to a distance of 60 mm. The separated sub-
stances were visualised by iodine vapour and marked.
The R
 F
 values were calculated to be 0.5 and 0.1,
respectively.
2.2.3. Transfer of analytes
Melting-point glass capillaries of 8 cm long and
1 mm i.d., purchased from Selby-BioLab, Australia
 148
W. He et al. / Vibrational Spectroscopy 30 (2002) 147–156
Page 3

were used to transfer the analytes. The 5 cm long
piece was removed from the closed end of the capil-
lary by breaking and the capillary (3 cm  1 mm)
was then packed with a small piece of facial tissue
(Kimberly-Clark Australia Pty. Limited), as illustrated
in Fig. 2a. The marked spot of the developed TLC
plate was directly transferred to the capillary by lightly
pressing and rubbing the capillary end on the marked
 Fig. 1. Chemical structure of: (a) Irganox1010; (b) Tinuvin770; (c) Chimassorb 119 FL.
W. He et al. / Vibrational Spectroscopy 30 (2002) 147–156
149
Page 4

area. After the silica gel containing the adsorbed
compound was transferred to capillary (Fig. 2b), the
capillary was placed into a 5 ml beaker containing the
eluent. Then methanol was used as an eluent for
the R
 F
 ¼ 0:8 spot and acetone–formic acid (2:8) for
the R
 F
 ¼ 0:5 and R
 F
 ¼ 0:1 spots. Due to the capillary
and penetration actions, the eluent travels upwards,
allowing the component to rapidly elute from the
adsorbent to eluent. When the eluent reached a level
about 1–2 mm above the tissue position as indicated in
 Fig. 2c, the capillary was removed from the beaker.
 A 3 ml glass capillary micropipete (d ¼ 0:5 mm),
made by Drummond Scientific Co., USA, was used to
draw about 0.1 ml of solution from the capillary (tak-
ing care not to touch the facial tissue). Then, the 0.1 ml
solution was dropped onto the BaF
 2
 plate, and the
BaF
 2
 plate was placed inside a fume cupboard to
evaporate the solvent. A stream of nitrogen was used
to remove any trace of solvent. After the evaporation
was completed, the position was marked with a fine
pen. By following this method, all of the other sepa-
rated materials were transferred onto the same BaF
 2
 plate. In order to remove any remaining formic acid,
a small drop distilled water was dripped onto the
corresponding positions of R
 F
 ¼ 0:5 and R
 F
 ¼ 0:1
components on the BaF
 2
 plate and distilled water
was then taken away by removing it with a piece of
tissue. This process was repeated three times to remove
any remaining formic acid from the BaF
 2
 plate.
2.2.4. FT-IR microscopy measurement
A Perkin-Elmer Spectrum GX/2000 FT-IR spectro-
meter equipped with an Auto Image Microscope was
used to identify the separated components. Perkin-
Elmer Spectrum Search Plus software was used for
library searching. The degree of conformity of the
sample and the reference spectra is described by
means of a hit quality, wherein the value 100 resem-
bles maximum fit. About 0.1 ml of each standard
additive solution was dripped onto the BaF
 2
 plate
and the solvent toluene was removed by evaporating in
air. After their FT-IR spectra were obtained (aperture
50 mm  50 mm, no. of scan 32, resolution 4 cm
 À1
 ),
the FT-IR spectra were added to the IR standard
material library. Figs. 3–5 show R
 F
 ¼ 0:8, 0.5 and
0.1 component FT-IR spectra, their best matched
standard spectra and corresponding microscopic zones
which were chosen for measurement.
3. Results and discussion
The efficiency of the coupled TLC–FT-IR micro-
scopy technique was investigated by analysing the PP
additives. Outdoor PP products usually contain anti-
oxidants and ultraviolet stabilisers and it is becoming
more and more common to add complex ultraviolet
stabilisers in order to obtain good performance of PP
outdoor products [13]. In general, polymer additives
are analysed after the extraction of additives from the
polymer and there are extensive reports [1,14] dealing
with how to extract additives from polymers. In these
studies, a simulative extraction solution of complex
additives for PP was selected to illustrate this techni-
que. This complex additive mixture was made up of a
mixture of Irganox1010, Tinuvin770 and Chimassorb
119 FL that are not easy to analyse qualitatively by
other chromatographic method such as GC–MS or
HPLC. Irganox1010 is a hindered phenolic additive,
whereas Tinuvin770 and Chimassorb 119 FL are
hindered amine additives.
The first step for this method is to separate each
additive from the mixture by TLC development by
choosing a suitable TLC plate and a mobile phase.
TLC plates are available with a wide range of chro-
matographic supports such as normal phase silica gel,
alumina (neutral, basic and acid), polyamide or cellu-
lose, reversed-phase, ion exchange and chiral supports.
 Fig. 2. Capillary transfer technique, showing the steps of transfer
technique: (a) with the filter plug; (b) with filter plug and silica gel
containing adsorbed compounds; (c) after eluting with the eluent,
the upper eluent layer contains separated component.
150
W. He et al. / Vibrational Spectroscopy 30 (2002) 147–156
Page 5

Fig. 3. (a) R
 F
 ¼ 0:8 spot FT-IR spectra (aperture 50 mm  50 mm; scan times: 32; resolution: 4 cm
 À1
 ) and the hit quality of 96 matched
Irganox1010 standard spectra (50 mm  50 mm; scan times: 32; resolution: 4 cm
 À1
 ). (b) Microscopic picture of R
 F
 ¼ 0:8 spot on BaF
 2
 plate
and selected zone marked as C for FT-IR analysis.
W. He et al. / Vibrational Spectroscopy 30 (2002) 147–156
151
Page 6

Fig. 4. (a) R
 F
 ¼ 0:5 spot FT-IR spectra (aperture: 100 mm  100 mm; scan times: 32; resolution: 4 cm
 À1
 ), the hit quality of 95 matched
Tinuvin770 standard spectra (50 mm  50 mm; scan times: 32; resolution: 4 cm
 À1
 ). (b) Microscopic picture of R
 F
 ¼ 0:5 spot on BaF
 2
 plate and
selected zone marked as B for FT-IR analysis.
152
W. He et al. / Vibrational Spectroscopy 30 (2002) 147–156
Page 7

Fig. 5. (a) R
 F
 ¼ 0:1 spot FT-IR spectra (aperture: 100 mm  100 mm; scan times: 32; resolution: 4 cm
 À1
 ), the hit quality of 95 matched
Chimassorb 119 FL standard spectra (50 mm  50 mm; scan times: 32; resolution: 4 cm
 À1
 ). (b) Microscopic picture of R
 F
 ¼ 0:1 spot on BaF
 2
 plate and selected zone marked as A for FT-IR analysis.
W. He et al. / Vibrational Spectroscopy 30 (2002) 147–156
153
Page 8

Nevertheless, a variety of solvents and solvent mix-
tures having different polarity are available as the
mobile phase. The selection of TLC plates and the
mobile phase mainly depends on the functional groups
present in the analytes. In this study, silica gel plates
were used with toluene–diethyl ether mixture (10:1) as
the mobile phase for Irganox1010 (R
 F
 ¼ 0:8) and with
acetone–formic acid (4:6) for Tinuvin770 (R
 F
 ¼ 0:5)
and Chimassorb 119 FL (R
 F
 ¼ 0:1). Since Tinuvin770
and Chimassorb 119 FL are hindered amine com-
pounds, the basic alumina plates are the best suitable
TLC plates. However, the separation of hindered
amine additives during the TLC development
(R
 F
 ¼ 0:1 and 0.5) was successfully achieved with
silica gel plates using a highly polar solvent mixture
containing formic acid.
Since FT-IR microscopy with a nitrogen cooled
MCT detector has very high sensitivity (0.01 mg is
enough in most cases), dimensions of 1 mm  3 cm
capillary was chosen to transfer separated TLC com-
ponents to the BaF
 2
 plate. During the process of
transferring of analytes, it is important to prevent silica
gel from being on BaF
 2
 plate, because silica gel will
interfere with FT-IR determination. In order to prevent
adsorbent from being drawn into micropipet during
application onto the BaF
 2
 plate, use of a proper filter
mediumisessential.AccordingtoworkdonebySzekely
 [15],KBr waschosenas a filter medium. Although KBr
 powder itself is IR transparent, it is difficult to distin-
guish KBr powder from the real sample under the
microscope. As a result, a high quality FT-IR spectrum
cannot be gained. Then, a piece of cotton was chosen to
block the adsorbent, but the result was not satisfactory
owing to its fluffy volume, and the high absorbency
capacity of the eluent. Since the eluting process was
carried out by capillary and penetration actions, it can
be estimated that just a small amount of adsorbent
would be in the upper eluent. Therefore, a piece of
facial tissue was used as the filter and it worked very
well. The presence of adsorbent was not seen because
the spectra displayed in Figs. 3–6 showed no Si–Opeak
at $1100 cm
 À1
 . No fibre of the tissue was found under
the microscope. However, emphasis should be made
not to contact the tissue with the micropipet during the
transferring, since trace amount of adsorbent may
deposit on the tissues.
After a marked spot was transferred to a capillary
tube, the capillary was dipped into the eluent. At this
stage, it is important to make sure that the separated
component would not diffuse into the eluent in the
beaker, instead it would elute from the adsorbent
quickly. Thus, a pre-test experiment was carried out
with a dye (bromophenol blue). The dye was applied
onto a TLC plate and the spot was transferred into the
capillary via the same method described previously.
The capillary was placed in methanol. The upper
eluent shown in Fig. 2c became blue and it was
immediately apparent that instead of diffusion, the
dye was eluted within a second from silica gel adsor-
bent to methanol. No colour change was seen inside
the methanol container. Therefore, it is possible to
transfer separated components easily and quickly
with almost no diffusion, if the right eluent is chosen.
As for the eluent, it also is very important to choose a
solvent that is polar enough to elute the separated
component quickly from adsorbent to eluent. Metha-
nol was used to elute the less polar Irganox1010 and
acetone–formic acid (2:8) was used for the amine
additives. Since all eluent need to be completely
removed from the BaF
 2
 plate before FT-IR determina-
tion, it is preferable to use a highly volatile solvent if it
is possible.
The salt plate used for FT-IR microscopy should be
hard, stable and resistant to most solvents. In addition,
it should have as far as possible a wide spectrum range
and a reasonable price. Of all window materials, a
BaF
 2
 plate is the optimal one, because it has a rela-
tively wide spectral range (50,000–770 cm
 À1
 ) and is
resistant to almost all organic solvents, most acids,
bases and moisture. Another advantage is its recycl-
ability. After it was cleaned with a proper solvent, it
can be reused. In this study, a BaF
 2
 plate of dimension
2cm  4cm  0:2 cm that could accommodate at
least 18 samples was used. The separated components
can be identified under the FT-IR microscope one by
one, without being interrupted by transfer of samples
prior to the FT-IR identification, saving much time.-
After the transferring of each separated additive onto
the BaF
 2
 plate, each spot was visualised under the
microscope and the FT-IR spectra were taken at
selective area as illustrated in images of Figs. 3b–
 5b for Irganox1010, Tinuvin770 and Chimassorb 119
 FL, respectively. In order to obtain a high quality FT-
IR spectrum, it is very essential to choose the correct
area. Usually, if there is a single drop of liquid, or a
crystal or powder whose area is larger than 2500 mm
 2
 154
W. He et al. / Vibrational Spectroscopy 30 (2002) 147–156
Page 9

(under the microscope it is very easy to locate the
area), the areas (shown in Figs. 3b and 6b) are the best
testing zones. If not, more concentrated sample area
should be chosen as shown in Figs. 4b and 5b, with
100 mm  100 mm aperture. As for the background
position, the closer to the sample position, the better
the spectrum will be. The collected spectra were
matched with those of FT-IR library, and the hit
 Fig. 6. (a) Dibutyl phthalate FT-IR spectrum (1 ml 0.05% dibutyl phthalate toluene solution was applied to TLC sheet and then dibutyl
phthalate was transferred to BaF
 2
 (aperture: 50 mm  50 mm; scan times: 32; resolution: 4 cm
 À1
 ). (b) Selected zone marked as E for FT-IR
analysis.
W. He et al. / Vibrational Spectroscopy 30 (2002) 147–156
155
Page 10

quality was 96 for Irganox1010 and 95 for both
Tinuvin770 and Chimassorb 119 FL.
Dibutyl phthalate that is one of plasticisers for PVC,
which has almost no absorption at 1100 cm
 À1
 , how-
ever, shows a lot of weak characteristic absorbent
peaks such as at 1600 cm
 À1
 . We choose a series of
dibutyl phthalate solutions for detection limit testing.
The detection limit was assessed by applying 1.0 ml of
0.5, 0.1, 0.05 and 0.01% dibutyl phthalate in diethyl
ether on the TLC sheet, and then using the same
transferring and evaluating method. All the samples,
except 0.01% dibutyl phthalate gave high quality
spectra suggesting that even 0.5 mg dibutyl phthalate
can be detected by this method without loss weak
benzene ring stretch vibration peak at 1600 cm
 À1
 , as
can be seen in Fig. 6. For 0.1 mg dibutyl phthalate, the
spectrum was not satisfactory owing to losing many
weak characteristic peaks.
Furthermore, the applicability of this method was
demonstrated by analysing the mixture of plasticisers
used for PVC. The mixture of dibutyl phthalate,
diisooctyl phthalate were separated using silica gel
TLC sheets, diethyl ether–toluene (1:9) mixture as
the mobile phase and methanol as the eluent. The R
 F
 values were 0.6 and 0.5, respectively. Very high
quality FT-IR spectra of dibutyl phthalate and diisooc-
tyl phthalate FT-IR spectra were obtained after using
above-mentioned transferring and evaluating meth-
ods. Thus, it is clear that the method described above
can be applied to a variety of additives in various
polymers qualitatively.
4. Conclusions
The results demonstrate that TLC–FT-IR micro-
scopy coupling is very convenient and reliable tech-
nique to detect and identify additives in polymer
without the TLC adsorbent interference. The total
time for transferring three marked spots and then
identifying them was about 20 min. The method has
been applied to polymer additives including hindered
amine light stabilisers, a hindered phenol antioxidant
and complex plasticisers for poly(vinyl chloride).
With a wide range of precoated TLC plates (silica
gel and alumina supporters were tested in this paper)
and a variety of mobile phases, a diversity of complex
additives in polymers such as antioxidants, cure accel-
erators in rubber and additives for coating and so on
can be analysed by this method. Moreover, a high
quality infrared spectrum can be obtained with as little
as 0.5 mg sample and the spectra are suitable for
directly matching using spectral library search soft-
ware. However, the shortage of this method is that
it cannot be used to analyse sample quantitatively.
After qualitative analysis, HPLC and GC are the best
methods for quantitative analysis if it is required.
Acknowledgements
We wish to thank China Scholarship Council for the
financial support.
References
 [1] J.W. Gooch, Analysis and Deformulation of Polymeric
Materials, Plenum Press, New York, 1997, Chapter 1, p. 1.
[2] J. Mink, E. Horváth, J. Kristóf, T. Gál, T. Veress, Mikrochim.
Acta 119 (1995) 129–130.
[3] G.W. Somsen, W. Morden, I.D. Wilson, J. Chromatogr. A 703
(1995) 614.
[4] G.C. Kiss, E. Forgács, T. Cserháti, J.A. Vizcaino, J.
Chromatogr. A 896 (2000) 61.
[5] T. Cserháti, M. Candeias, I. Spranger, J. Chromatogr. Sci. 38
(2000) 145.
[6] M.S.F. Lie Ken Jie, J. Mustafa, M.K. Pasha, Chem. Phys.
Lipids 100 (1999) 166.
[7] T. Tajima, K. Wade, K. Ichimura, Vibrat. Spectrosc. 3 (1992)
211.
[8] G.W. Somsen, W. Morden, I.D. Wilson, J. Chromatogr. A 703
(1995) 644.
[9] S.A. Stahlmann, J. Planar Chromatogr. 12 (1999) 5.
[10] K. Wade, T. Tajima, K. Ichimura, Anal. Sci. 7 (1991) 407.
[11] P.A.M. Smith, Vibrat. Spectrosc. 24 (2000) 47.
[12] B.O. Budevska, Vibrat. Spectrosc. 24 (2000) 37.
[13] T. Konstantinova, A. Bogdanova, S. Stanimirov, Polym.
Degradat. Stability 43 (1994) 187.
[14] C.G. Smith, N.E. Skelly, R.A. Solomon, C.D. Chow,
Handbook of Chromatography: Polymers, CRC Press,
Florida, 1982, Section V, p. 166.
[15] J.C. Touchstone, Practice of Thin Layer Chromatography,
Wiley, New York, 1992, Chapter 20, p. 347.
156
W. He et al. / Vibrational Spectroscopy 30 (2002) 147–156

Информация о работе Analysis of additives in polymers by thin-layer chromatography coupled with Fourier transform-infrared microscopy