Aliphatic polyester-grafted starch composites by in situ ring opening

Aliphatic polyester-grafted starch composites by in situ
ring opening polymerization *
Abstract- Catalyzed Ring-Opening Polymerization (ROP) of lactones, such as c-caprolactone
and ii-vnlerolactone. ha\ been investigated for preparing aliphatic polycsrer-grafted corn starch
compositions. The polynieri7ntion has been carried out either i n bulk (without solvent) or in
I0 wtQ grunular starch i n toluene \uspension. aftcr adequate activation of thc starch wrface
arnylosc/amylopectin hydroxyl groups into aluminum alkoxides. This activation has been done by
it1 si/rr reaction with triethylaltiininuin and has proved to be very effective i n promoting fast ROP and
nlw covalent grafting o f the polyestcr chains onto thc starch xurfacc. Drying of the starch granules
constitutes a key-step of the p r o p o d process. The actual lixation of the initiator onto the \tarch
wrface has been chccked by XPS analysis whereas good adhesion between the two cornponenls has
been evidenced by SEM observations. Growth of the polyester chains on the starch granules has
bccn followed by thc increase of thc inean diameter o f the 'encapsulated' granulcs a\ attested by laser
scattering granulometry.
Koywotdv: Starch: aliphatic polycstei-s: poly(c-caprolactone): ringopening polymrriz;rtinn. composites.
1. INTRODUCTION
Natural polysaccharides such as starch (granular or plasticized) are often used in
polymer blends and composites to produce the widely sought biodegradable property [ I]. However, conventional melt-processing usually provides the starch-based
materials with very poor mechanical properties, mainly due to thermal decomposition of starch before melting, strong water absorption and poor interfacial adhesion
This paper was presented at thc / i i ~ ~ , r . ~ / c i t i o i i eEut-o-/T//c,r.\
r/
'YY
Villeurbannc. France. Scptcrnber 6-9, 1999.
' F,-niaiI : pi1iI ippc.du boi 5 @ ti in h .iic .be
'
.Yyrripo,\i/itt/,
held i n Lyori-
12, 31. In order to overcome some of these drawbacks. physical modifications like
hydrophobic coating, cross-linking and addition of an external plasticizer [4, 51
can be performed. Another alternative consists of the chemical modification of
starch, such as non-degradative substitution of the hydroxyl groups with functional
groups like esters, ethers, isocyanates, anhydrides and the like 15-71, Starchgraft copolymers have also been produced from various monomers, e.g. styrene,
methyl (meth)acrylate, butyl acrylate, acrylatnide and acrylonitrile, by using a solution process in which the starch grafting was initiated either by radiation or by
cerium ions (CeJ+) [8- 101. Anionic polymerization of ethylene oxide on starch
has also been reported [ l 11. Other starch graft copolymers have been obtained by
reactive extrusion, the extrusion shear generates starch macroradicals spread along
starch molecules, which react with unsaturated monomers and/or preformed polymers [ 121. Some of us have recently reported the it7 situ grafting of aliphatic polyester chains by ring opening polymerization (ROP) of r-caprolactone (F-CL)aftei
starch surface activation by Sn(Oct)?, Ti(OnBu)d, A1(OiPr)3or AIEt, [ 131. From
these preliminary observations, it emerged that the best results in terms of grafting
level and polymerization kinetics were obtained by activating the starch hydroxyl
groups with AIEt?.
This paper reports on the synthesis and characterization of aliphatic polyestergrafted Etarch (Scheme 1 ) by in s i f i i ROP of E-CL (or 8-valerolactone) in the presence of granular starch. The in sirii ROP is initiated by the amylose/amylopectin hy.
droxyl groups available at the starch surface, after adequate activation in aluminum
alkoxides. The polymerization is conducted either in bulk (without solvent) or in i
10 wt%l toluene suspension, leading to chemical grafting of the hydrophobic poly.
esters chains onto the hydrophilic starch. Key advantages of the proposed process
are: the use of bulk monomer for the ROP in the presence of starch, control ovei
the level of grafting, ease of control of the final material properties, and enhancec
interfacial adhesion leading to homogeneous encapsulation of the initially tiloistun
sensitive substrate by a hydrophobic polyester shell. Furthermore, polycaprolac.
tone (PCL) presents the rarely met property of being miscible with inany othei
polymers (such as poly( vinyl chloride), styrene/acrylonitrile, ABS, bisphenol-P
polycarbonate, nitrocellulose, cellulose butyrate) and is also mechanically coinpat
ible with other - polypropylene, polyethylene, natural rubber, poly(viny1 acetale)
Scheme 1.
ethylene/propylene rubber, for example [ 141. Therefore, the PCL-grafted starch
could be used as cotnpatibilizer in composites of such polysaccharide substrates
with various matrix resins, wherein the matrix of the composite is a component that
is miscible or mechanically compatible with PCL.
2. EXPERIMENTAL
2. I. Mctlzocl.y (?1'(.Izrrmt.tt.r-i,ritiori
Differential Scanning Calorimetry thertnograrns were recorded on a MDSC 2920
from TA Instruments. Samples (w. 10 mg) in a hermetic aluminum pan underwent
two consecutive scans from - 120°C to 160°C at 10"C/min (under nitrogen flow).
Laser Scattering Granulornetry was performed onto water dispersion of the coated
starch granules with a Coulter LS 230 equipped with a small volume cell allowing
the size, and size distribution of the particles to be determined from 40 nm to 2 mm.
Secondary Ion Mass Spectrometry analysis was performed on a TOF-SIMS, IONTOF IV.
The SEM images were obtained with a JEOL JSM 6100 microscope, after surface
coating with gold by using a sputter coater.
2.2. Mlite ricr 1.\
e-Caprolactone ( r -C L )and 6-valerolactone (6-VL) monomers obtained from Acros
were dried over calcium hydride for two days and distilled under reduced pressure.
Triethylaluminutn was obtained from Fluka in a toluene solution (I.8 M ) and was
used as received. Toluene was dried by refluxing over calcium hydride and distilled
under nitrogen just prior to use. Commercial corn starch having a particle size of
about 14 p m was provided by Biopastics Inc. and was used after intensive drying.
2.3. Po ly m tjr i w ion p roc~tliiI P
2.3.1. B d k ROP In a typical experiment, 5 g e-CL (6-VL) was added to
5 g granular starch (SO/S0 wt/wl) in a SO ml round-bottomed flask previously
purged with nitrogen, and equipped with a stirrer, and oil valve for removal of
volatiles. A predetermined ainount of triethylalurninuin in toluene solution was
then introduced under nitrogen via a flame-dried syringe at room temperature. The
reaction medium was then heated up to the desired temperature. Fast cooling down
to room temperature stopped polymerization.
2.3.2. Sirspmyion ROP In a 100 ml round-bottomed flask equipped with a
stirrer, an oil valve, and previously purged with nitrogen, the polymerization was
carried out by adding the adequate quantity of triethylaluminurn to a suspension of
S g starch in S O ml dried toluene ( I0 ~ 1 % )and
. then 5 g of s-CL via a flame-dried
syringe under nitrogen slight overpressure. The medium was then heated up to the
218
n. Kutot et al
desired temperature. Polymerization was stopped by fast cooling down to room
temperature.
Efficiency of the iri situ generation of the aluminum alkoxides on the granular
starch was indicated by the presence of the characteristic aluminum peaks (Alzpand
Al?,) in the X-ray photoelectron spectrum (XPS, VG ESCALAB 2201 XL) of the
starch surface treated by AIEt, in toluene solution (in absence of s-CL).
2.4. Deterrnimition ofpolvc2ster gruftirig cffk.iencj1 and
nio11o117er
co~version
Selective extraction of the non-grafted polyester chains was performed in toluene at
room temperature. Crude product (2 g) was added to 25 ml of dried toluene. After
16 h agitation, the insoluble fraction was filtered off and dried until constant weight
was reached. The insoluble part contains both starch and the polyester chains that
were physically and/or chemically grafted on it. The soluble part was precipitated
from ti-heptane and contains non-grafted polyester chains. The remaining soluble
fraction corresponds to the unreacted lactone monomer. The grafting efficiency is
defined as follows:
9% grafted polyester
Grafting efficiency =
9% grafted polyester 9% non-grafted polyester
+
3. RESULTS AND DISCUSSION
3.1. Polq'merixitiorz procwlurc.
The polymerization procedure is depicted i n Scheme 2. The polymerization strategy
is based on fixation of aluminum alkoxide functions onto the starch surface by
it7 situ reaction of triethylaluminum and hydroxyl groups of starch. The reaction
equilibrium is shifted towards the corresponding alkoxides by the evolution of
ethane. This method proved to be very efficient in promoting fast polymerization as
well as covalent grafting of the growing polyesters chains onto the starch phase [ 131.
It is worth pointing out that the drying stage of the granular starch is a key-step
of the proposed process. Indeed, residual water should have a strong influence on
the ability to generate (or to maintain the activity) of the active aluminum alkoxides
species at the starch surface. In fact, it is well known that alkylaluminum can readily
react with water molecules with the concomitant formation of active alurninoxane
whcrc 111 = 4 or 5
Scheme 2.
derivatives. This promotes the lactone polymerization directly in solution with
formation of non-grafted polyester chains. It comes out from a detailed study that
showed that the drying of corn starch granules successively at 100°C in a ventilated
oven for 24 h, at 90°C for 7 h under 5 x lo-' mbar and then at 120°C for 16 h
under l o p 3 mbar, leads to a content of residual water of 0. I3 wt% as determined by
Karl-Fisher analysis. Such residual water content has proved to be low enough to
avoid the formation of too many polyester chains free in solution. For instance, a
grafting efficiency higher than 50% is reached in bulk ROP of E-CL at 65°C when
promoted by these dried starch granules treated with 0.3 wt% AIEt?.
Table I reports some of the results obtained under different experimental conditions. Depending on the polymerization conditions (particularly bulk or suspension
ROP, reaction temperature, relative quantity of triethylaluminum used to treat the
poly saccharide surface, nature of the lactone monomer), the monomer conversion
and the grafting efticiency can vary dramatically.
The polymerization in bulk is more effective than the polymerization in toluene
dispersion (lower grafting efficiency even after longer reaction time, entries 1 and 2).
I t has also been noted that an increase i n temperature, from 65°C to 1 10"C, results
in an enhancement of the grafting efficiency, from 32% to 41 %, while the monomer
conversion remains constant at a. 78% (entries 3-5). . Under the same bulk
polymerization conditions, the grafting efficiency is higher with 8-valerolactone
than with E-caprolactone, 58% and 325%.respectively (entries 3 and 6). That can
be attributed to the fact that 6-valerolactone ROP initiated by aluminum alkoxides
species is known to be slower, leaving more time for the triethylaluminum to react
with the hydroxyl groups available at the starch surface. Therefore, more active
sites are formed at the surface allowing the fixation of more polyester chains and
thus a higher grafting efficiency.
Tahle 1.
Bulk ROP of F-CL promoted by the AtEtj surface treated starch granules (residual water content =
0.13 wt%). Effect of the experimental conditions
~~
~
Al content
Tcmpct dturc
(wt%)
("C)
I
2
3
4
5
6
0.33
0.36"
65
65
65
1
1 .OO'
Entry
1 .oo
I .oo
I .oo
I .oo"
90
1 I0
65
65
Reaction
time (min)
Monomer
converwn ( % )
Grafting
efficiency ( % )
95
300
I20
I20
I20
I20
I20
54
6
76
52
33
32
40
41
S8
62
XO
14
16
52
I0 wt% sujpension in toluene (jee experimeiitul).
" PoIymcriiration of S-VL.
Two-stcp procedure: ( i ) stiri'ace treatment i n toluene suspension and (ii)bulk polymerization after
solvent reinoval.
In the last entry of Table I , the reaction of starch with triethylaluminum has
been carried out according to a two-step procedure. Indeed, starch granules have
been first reacted with 1.0 wt% AIEt, in toluene suspension (10 wt%) followed
by removal of the supernatant via a flame-dried capillary under nitrogen pressure,
washing once again with fresh dried toluene, and drying under vacuum. In a second
stcp, E-CL is added and the reaction medium is heated up to the desired temperature.
Under such conditions, the bulk ROP conducted at 65°C for 2 h yields a higher level
of PCL grafting (62% against 32%, entries 3 and 7). This observation might be
explained by the removal of the active species free in solution.
3.2. Chnrac.terization ofthe PCL-grafted starch compositions
DSC thermograms of the polyester-grafted starch composites show that the molecular weight is high enough to allow the polyester chains to crystallize. Indeed, a
PCL-grafted starch composite with 38 wt% PCL (Fig. 1) is characterized by a glass
transition temperature and a melting temperature of -60°C and 53"C, respectively,
in good agreement with the commonly met values for homoPCL.
Interestingly enough, laser scattering granulometry has been investigated for
checking the growth of the aliphatic polyesters grafted onto the starch surface.
It results that the mean diameter of the coated particles increases with grafted
polyester content as obtained after selective extraction on the non-grafted polymeric
- 60 73 "C
5 1 "C
16.16 Jig
1
0
-50
-100
Figure 1. DSC thcrmograin o f
0
50
'Iemperaturc (T)
I xo up
il
PCL-grafted \larch with 38 wt% PCL
100
150
22 I
Differential volume
Starch-g-PCL ;WtYGCL
Mean dameter ( p )
0
13 45
6 57
7
23 12
19 85
14
23 42
18 31
29
27 70
20 03
(F)
has been characterized by Secondary Ion Mass Spectrometry. Typical fragments,
including the monomeric ones at C6HIlO?(rn/z = 115) and CSHuO:(m/c =
101) respectively for PCL and PVL, were detected not only for the grafted
homopol yesters, but also for sequentially copolymerized PVL-b-PCL grafts onto
starch granules [ 1.51.
Scanning Electron Microscopy (SEM) images of the so-obtained composites attest to the very good interfacial adhesion between starch and the aliphatic polyester
chains (Fig. 3b and 3c). For the sake of comparison, a total lack of adhesion between the components has been observed in simple starch/polycaprolactone melt
blend (Fig. 3a).
The SEM images of the composites obtained after extraction of non-grafted polyesters show a tight coating of the starch granules by the polyester chains (Fig. 4).
Actually, every starch particle is still embedded in a shell of anchored polyester
chains, confirming the mean diameter increase measured by laser scattering granuI om et ry.
4. CONCLUSIONS
Polyester-grafted starch composites have been synthesized by iiz situ ROP of the
corresponding lactone niononier ( E - C L and S-VL) in the presence 01 granular
starch. Polymerization was carried out either in bulk (without solvent) or in toluene
dispel-sion after adequate modi tication of the hydroxyl groups, available at the starch
surface, into aluminum alkoxides. This process proved to be very effective in
promoting fast polymeriza~ion and covalent grafting of the polyester chains onto
the starch surface. The grafting of the polyineric chains has been evidenced by
Differential Scanning Caloriinetry, Laser Scattering Granulometry, Secondary Ion
Mass Spectrometry. The very good interfacial adhesion between the polyester
coating and starch sureace has been attested by SEM observations achieved on
samples before and after selective extraction experiments.
A~,hriocc.letl,~tiii~ii~~
Thi5 worh wa5 \upported by both Rkgion Wallonne and Fond5 Social Europ6en in
the frame of Objectii 1-Hainaut: Materia Nova program.
REFERENCES