Synthesis and Properties of Starch Based Biomaterials
Asaf Kleopas Sugih
The author thanks the University of Groningen for the financial support through an Ubbo Emmius Scholarship.
Synthesis and Properties of Starch Based Biomaterials
Proefschrift ter verkrijging van het doctoraat in de Wiskunde en Natuurwetenschappen aan de Rijksuniversiteit Groningen op gezag van de Rector Magnificus, dr. F. Zwarts in het openbaar te verdedigen op vrijdag 12 december 2008 om 13.15 uur door Asaf Kleopas Sugih geboren op 4 juli 1975 te Bandung, Indonesië
H. Moscicki ISBN : 978‐90‐367‐3592‐6 ISBN : 978‐90‐367‐3593‐3 (electronic version)
.B. A. dr. L. ir. Broekhuis Prof. dr. dr.J. Promotores : Prof.A. L.M. ir. F.P. dr. Minnaard Prof. dr. A. dr. Janssen Prof.J. Heeres Prof. Picchioni Beoordelingscommissie : Prof.
to: Tresna. and my sister
. my parents.
1. Thermoplasticized starch 1. Product analyses
.3.3. Structure and properties of starch 1.1. Methods 2. Introduction 2.4. Starch ‐ biopolymer blends and graft co‐polymers 1.1.2. References
2 3 4 5 5 7 9 9 10 11 11 13 14 16 16 17
Chapter 2: Experimental Studies on the Ring Opening Polymerization of p‐dioxanone using an Al(OiPr)3‐ Monosaccharide Initiator System
2.2. Starch 1.1. Materials and Methods 2.1. Biodegradable plastics from starch 18.104.22.168.Table of Contents
Chapter 1: Introduction
1.2. Cross‐linked starch 1. Starch‐based blends by melt mixing 1.2. Typical example for the synthesis of polydioxanone end‐capped with 1.1. Biomaterials from starch 22.214.171.124. Starch Esters 126.96.36.199.4‐di‐O‐isopropylidene‐α‐D‐galactopyranose (2) 27 28 26 27 27 27
2. The potential of biodegradable plastics 1.2. Starch production processes 188.8.131.52.2.3.2. Thesis Outline 184.108.40.206.4.5. Plastic applications and waste issues 220.127.116.11.2.2.3. Materials 2.1. Starch‐based blends by in situ polymerization 1. Starch modifications to improve product properties 1.2.
3.2. Systematic studies 18.104.22.168.2.3. Typical example of the starch silylation procedure 3.5. Conclusions 2. 22.214.171.124.2.1. Calculation of average degree of polymerization 2.2.3. Analytical methods 3. Materials 3.2.3. Typical example of in situ polymerization of ε‐CL with silylated starch 50 3.3. References
Chapter 3: Synthesis of Poly‐(ε)‐caprolactone Grafted Starch Co‐polymers by Ring Opening Polymerisation using Silylated Starch Precursors
126.96.36.199.2.1.2. Mechanistic aspects 188.8.131.52. Product yield 2. Peracetylation of silylated starch 184.108.40.206.4. Introduction 3. Product analyses 2. Results and Discussions 3.2. Materials and Methods 3.3. Calculations 220.127.116.11.2. Nuclear Magnetic Resonance (NMR) 3. Nomenclature 2. Methods 3.3.2. Results and Discussions 18.104.22.168.4. Screening experiments 2.2.2. Synthesis of silylated starch
.1.2.2. De‐silylation of poly‐(ε)‐caprolactone grafted silylated starch co‐polymers 51 51 52 52 52 53 54 48 49 49 50 50
3. Effects of process conditions on the average chain length ( X n exp ) and end group distribution
28 29 29 30 36 38 39
40 42 43 44
4. Conclusions 4. 1H‐ and 13C‐NMR analyses 4.4. Effect of the addition of toluene as a co‐solvent 4.6. Methods 22.214.171.124. Product characterisation 126.96.36.199.3.3. Nomenclature 3.1. In situ ring opening polymerization of ε‐caprolactone with silylated starch 57 63 64 64 65
3. Analytical equipment 4.3.3. Exploratory experiments 4. Systematic studies 4. Nomenclature 4.2. Materials and Methods 4.2. Typical example of the synthesis of laurate and stearate esters of corn starch 72 73 74 75 75 76 76 78 79 80 81 81 82 83 84 70 71 71 71 72
4.3.2. Determination of the Degree of Substitution (DS) 4. Peracetylation procedure 188.8.131.52.3.3.3. Results and Discussion 184.108.40.206.2.2. FT‐IR measurements 220.127.116.11.2.5. Conclusions 18.104.22.168. References
Chapter 4: Synthesis of Higher Fatty Acid Starch Esters using Vinyl Laurate and Stearate as Reactants
.22.214.171.124.1. Materials 4. Deprotection of silylated‐starch‐g‐PCL 3.3.
3.3. Introduction 126.96.36.199.2. Catalysts screening 4.3. Effect of vinyl ester to AHG ratio on the product DS 4.
2. Work‐up of PCL‐g‐DEM 108 109 109 109 109 110 110
6.2.4. Nomenclature 5.6. Analytical Equipment 5. Experimental Design 5.1.2. Analytical Methods 110
Chapter 6: Synthesis and Properties of Reactive Interfacial Agents for Polycaprolactone‐Starch Blends
188.8.131.52.2.2. Compatibilizer synthesis 6. Product properties 184.108.40.206.3.3. Preparation of PCL‐starch blends with the reactive compatibilisers 110 6.2. Typical example of the preparation of laurate and stearate starch Esters 88 86 87 87 87 88
5. Work‐up of PCL‐g‐GMA products 6. Materials and Methods 5. Results and Discussion 220.127.116.11.Chapter 5: Experimental and Modeling Studies on the Synthesis and Properties of Higher Fatty Esters of Corn Starch
18.104.22.168.3. Conclusions 5.2.2. Mathematical modelling 5. Materials 6. Materials and Methods 6. Introduction 22.214.171.124. Peracetylation procedure and Degree of Substitution (DS) determination 88 88 89 92 96 102 102 103
5.3. Introduction 6. Methods 126.96.36.199.4. Methods 188.8.131.52.2.3.2. Materials 184.108.40.206.
220.127.116.11.18.104.22.168. Effect of the BPO intake on the product FD 6. Effect of substrate (GMA/DEM) to PCL Ratio on the FD 6.1.
6.3.1. Statistical mdeling 6. Ternary blends compatibilized with PCL‐g‐GMA 6.1.2. Synthesis and Properties of Starch‐ PCL Blends 6. Results and Discussions 6. Binary blends of starch and PCL 6.1.3. Modeling of the combined effects of the GMA/ DEM and BPO intakes on the FD
111 116 117
118 120 121 122 123 125 128 128 129
22.214.171.124. Thermal properties of the compatibilizers 6.6. Preparation of the ractive compatibilizers 126.96.36.199.188.8.131.52. References Summary Samenvatting (Dutch Summary) Acknowledgements List of Publications
.3. Conclusions 184.108.40.206. Ternary blends compatibilized with PCL‐g‐DEM 6.2.1. Nomenclature 6.3.3. Calculation of the Degree of Functionalization (FD) of the reactive compatibilizers 6.
Finally. an outline of this thesis is given.Chapter 1
A general overview of starch properties will be given and the potential use of starch as a starting material for a wide range of green biomaterials will be reviewed and discussed. Different routes to modify starch to improve the product properties and to extend the application range will be provided.
An important polysaccharide is starch. present and a forecast of starch production  o : Europe □ : USA : Total
∆ : the Rest of the World
.1.45 (estimated from the total volume and value of the US corn starch export ). potato or cassava) of the plants [2‐3]. potato. It is generally deposited in the form of small granules or cells with diameters between 1‐100 µm .Chapter 1
1. Starch is found in seeds (i. current and future estimations of worldwide starch production are given in Figure 1. and wheat starch are also produced in large amounts [2. Most starch crops are very productive.e. .1. maize.e. rice.
80 70 60 50 40 30 20 10 0
Starch Production (Million Tons)
Figure 1. with an average yield of 4. Past. Most of the starch is produced in the USA. Potato accumulates starch to approximately 75 % of the dry weight in the tubers with a yield up to 21 ton starch per hectare. The current price of corn starch is around $0. Most of the starch produced worldwide is derived from corn. or peas) and in tubers or roots (i. sweet potato. barley. while corn seeds consist of 65‐80% starch by weight. Plants synthesise and store starch in their structure as an energy reserve. corn. Starch
75% of all organic material on earth is present in the form of polysaccharides. The worldwide production of starch in 2008 is estimated to be around 66 million tons . sorghum. . but other types of starch such as cassava. wheat.1. Past.9 ton starch per hectare . The second and third starch producer regions are Europe and Asia [4‐5]. 4‐5].
b. The fiber is removed using washing screens. and starch granules are obtained after centrifugation from the suspension.
a. Starch production processes Starch is generally extracted from the plant by wet milling processes [9‐10]. Typical corn‐milling operation b. Cross‐sectional view of a corn kernel
. gluten. and other foreign materials.1.2. After a coarse milling. a typical corn starch production process is given in Figure 1. sand.2. the mixture is fed to a hydrocyclone to separate the germ and the rest of the kernel. The plant material is grounded in water.2.1.a. Introduction
1. The resulting suspension from the mills contains fiber. the corn kernel is softened by steeping in warm water containing SO2 until the volume of the kernel increases with 55‐65%. which is again fed to a second milling process. the debris is filtered from the slurry. The resulting starch is further washed in a cyclone and finally dried. Corn starch production  a.
b. and is embedded in a proteinaceous cellular matrix as is shown in Figure 1. and starch. The starch is present in the endosperms (floury and horny).
Figure 1. The low density gluten is separated from the starch suspension by centrifugation. As an example. After initial cleaning to remove cob.
Structure and properties of starch Starch is a polymer consisting of anhydroglucose (AHG) units (see Figure 1. Each branch contains about 20‐30 anhydroglucose units. 6]. but usually in the range of 1.4)‐glucosidic bonds... containing periodic branches linked with the backbones through α‐D‐(1. Amylose c. Amylose is essentially a linear polymer in which AHG units are predominantly connected through α‐D‐(1. The content of amylose and amylopectine in starch varies and largely depends on the starch source.6‐7 x 105 Da .
a. The starch granules will swell rapidly to many times of its original volume.. The linear
. The molecular weight of amylopectin is higher than that of amylose and is typically 4‐5 x 108 Da . Chemical Structure of Starch a..2. the amylose content is between 18‐28% .) . . Heating a starch solution leads to loss of hydrogen bonding in the interior of the starch granule and the starch will start to gelatinize. Amylopectin
Starch is insoluble in cold water.
H2COH H H OH H O H OH O H OH O CH2 H O H H OH H O H OH H O . The molecular weight of amylose is a function of the plant source and processing method. Two types of AHG polymers are usually present in starch: amylose and amylopectin [2‐3.
O H OH
H OH H
O H OH
O H OH
H O .3.
H2COH H OH H O H
b.1... Anhydroglucose (AHG) unit b... but it is very hygroscopic and binds water reversibly.1. The amylose content of several common starches is given in Table 1.6)‐ glucosidic bonds .Chapter 1
H2COH H .
Figure 1.. H OH H
O H OH
H OH H
H2COH H . Typically.. H OH H
H2COH H2COH H O H H OH H H2COH H O H H OH H
O H OH
H O .3. Amylopectin is a branched polymer...
swollen granules. The first generation of commercial plastics was derived from cellulose nitrate and is known as celluloid [12‐14]. Plastic Applications and Waste Issues Plastic is the general term for a wide range of synthetic or semisynthetic polymerisation products. and granule fragments.5‐70 62‐72 63‐72 58‐64
1.1.1. Plastics are used in a wide range of applications and the demand is still increasing every year . will form a thick paste or gel. The resulting suspension contains a mixture of linear amylose molecules. and.7 26
Table 1. Introduction
amylose molecules leach out of the granules into the solution. Biomaterials from starch
1.8 16. Cellulose nitrate was first prepared by A. and celluloid was
. Parker in 1838. Amylose content of common starches 
Starch Arrowroot Corn Hybrid amylomaize Class V Hybrid amylomaize Class VII Oat Manioc Potato Rice Sago Sweet potato Tapioca Wheat Amylose (%) 20.5 25. Table 1.8 17. Starch gelatinization temperature range 
Starch Potato Tapioca Corn Waxy corn Wheat Gelatinization Temperature Range [o C] 59‐68 58.7 20 18.2. The gelatinization temperature range of various starch sources is given in Table 1.2.5 28 52 70‐75 27 15.2.2. depending on the amount of water present. The gelatinization temperature range can be defined as the temperature at which granular swelling begins until the temperature when nearly 100% of the granules are gelatinized .
and therefore need less energy to shape it into useful materials .5
0 1950 1960 1970 1980 Year 1990 2000 2010
Figure 1. space. discovered by Baekeland in 1907 [12.)
Included: Thermoplastics. polyethylene terephtalate (PET). and polycarbonates (PC). Plastics are also used for building materials and automotive. for instance to be used in packaging and paper coatings . Plastics are very attractive materials.).Chapter 1
patented by J. While celluloid is derived from a natural polymer (cellulose). electrical. polypropylene (PP). polystyrene (PS). Thermosets. Important polymers used for packaging are polyethylene (HDPE and LDPE).4. They have a low density and can be shaped in thin films while maintaining good properties. Polyurethanes. In 2006. the oldest purely synthetic plastic is Bakelite. Sealants. The production and consumption of plastics has increased significantly with a rate of almost 10% every year since 1950. and PolyAcryl. Plastics have lower melting temperatures compared to glass and metals. Elastomers. and energy during transportation of goods. The largest application of plastics is for packaging purposes. 15]. PA-. About 29% of the total plastics produced in the USA .5. The latter is important when using the material for packaging purposes to save weight. the worldwide plastics production has reached 245 million ton per annum  (see Figure 1. when polyethylene (PE) was invented. and PP-Fibers
2006: 245 2002: 200
200 Plastics Production (Million Tons)
Not Included: PET-.4. Coatings. Hyatt in 1870 . Adhesives. and 37% of the total plastics demand in EU  is used as packaging materials. A dramatic increase in demand for plastics began after World War II. and consumer products (see Figure 1. Worldwide Plastics Production (1950‐2006) 
. PE is a very versatile plastic because it can be shaped easily into various forms.Fibers
1950: 1. polyvinyl chloride (PVC).
It is not applicable for thermoset resins . are not easy to recycle . In the USA. however. plastics waste was ranked as the second major source of MSW after paper and paperboard. Plastic applications and waste treatment in Europe  1. The volume fraction of plastic in MSW is much larger due to the low density of plastics. and accumulation of its residues in soil cause significant reductions in agricultural yields. In 1996. Most of the waste is still disposed (50%) by landfill or incinerated to recover the energy (30%) . soil. The potential of biodegradable plastics The application of biodegradable plastics could be an attractive solution for the problems related to the use of conventional plastics (vide infra).2. threatening fishery.7% in 2006 . rivers. Co‐mingled plastics. only 20% of the plastic is recycled.5. 16]. Recycling (part of) the plastics to reduce the amount of MSW also has limitations. irrigation. Introduction
Plastic waste. the amount of plastics in MSW increased from less than 1% in the 1960 to 11.2. The disposal of plastics materials in municipal solid waste (MSW) is a serious issue in many parts in the world. is causing serious environmental problems. In Europe (2006). Plastic litter is hazardous to wildlife [1. and is only effective for single plastic sources or simple plastic formulations. and lakes. consisting of 25%‐v of the total waste . 6]. Plastics have been polluting sea [1. Biodegradable plastics are polymeric materials capable of decomposing when given an
. which are usually found in waste streams. hydropower plants operation. and other public works .
Figure 1. and may be more than twice the weight fraction . It has a high volume to weight ratio and is resistant to degradation. ship navigation.
and these are shown below: ISO 472: 1998 ‐ A plastic designed to undergo a significant change in its chemical structure under specific environmental conditions resulting in a loss of some properties that may vary as measured by standard test methods appropriate to the plastics and application in a period of time that determines its classification. The considerable growth of interest in composting as a means to replace landfill due to the decreasing disposal spaces (especially in Europe) may also help the progress of biodegradable plastics development . Biodegradable plastics may be classified into two general groups. The change in chemical structure results from the action of naturally occurring micro‐ organisms. animals.2 working group on biodegradable polymers ‐ Biodegradation of a plastic material is a process leading to naturally occurring metabolic end products. because they take part in nature’s cycle of renewal. Biopolymers are inherently biodegradable. biopolymers from nature (from plants. biological and/or physical forces in the environment at a rate which leads to fragmentation or disintegration of the plastics. Biodegradable plastic waste may be treated in composting facilities. and since 2000 they have become competitive with traditional materials in some applications [20‐21].96 proposal ‐ Degradable plastics are plastic materials that undergo bond scission in the backbone of a polymer through chemical. Japanese Biodegradable Plastic Society draft proposal ‐ Biodegradable plastics are polymeric materials which are changed into lower molecular weight compounds where at least one step in the degradation process is through metabolism in the presence of naturally occurring organisms.5/
. the price of PLA based plastics in 2010 is estimated to be around € 1. ASTM sub‐committee D20.Chapter 1
appropriate environment and sufficient amount of time . and microorganisms) and biodegradable synthetic polymers . together with food and yard waste as well as paper. and polyesters from micro‐organisms . The price of these biodegradable plastics is expected to be reduced considerably in the next decade . Several authorities have provided definitions for biodegradable plastics . For instance. DIN 103. The market price of biodegradable plastics (mainly from starch and PLA) has decreased the last years. Biodegradable plastics have gained considerable interest since the 1980s. It may also be treated in sewage sludge water treatment plants or buried in the soil [20‐21]. Nowadays new types of biodegradable plastics with improved properties and lower costs have been developed . Common natural biopolymers are carbohydrates. proteins (present abundantly in plants and animals).
1. The price of starch in 2007 was about $ 0.2‐ 1. before they can be applied as biodegradable plastics. either from starch‐based materials (slightly modified starch.
1. Consequently. Starch modifications to improve product properties
Several techniques may be applied to develop starch based biomaterials with improved properties. when recycling is not practical or uneconomical [20‐21]. 20]. Biodegradable Plastics from Starch Starch is a very attractive source for the development of biodegradable plastics.
. or when environmental impacts have to be minimized. These are summarized briefly in the next sections.4 or $ 1. Introduction
kg. hygienic materials (diapers. chemical modification or combinations of them . . consumer products and food products.4 per kg in 2007‐2008 ). agricultural products (mulch films. Virgin starch is not suitable as a packaging material. Examples are the use as packaging material for carrier bags. breathable fabrics.45 per kg . The price of starch‐based biodegradable plastics (€ 1. pots).8‐ 2. napkins) . Starch may become an attractive raw material for plastics in the future. blending with other materials. utensils . Biodegradable plastics are especially very useful for single‐use applications. either by plasticization. Bastioli [20‐21] showed that nearly all biodegradable plastics available in the market are derived from starch. € 1. because the price of oil based polymers may still increase due to the rise in the crude oil prices . The global production capacity of starch‐based bioplastics in 2010 is estimated to increase to 200‐300 kiloton per year from 77‐200 kiloton in 2003 . This price is much lower than conventional plastics derived from oil. tensile and flexural strength) and is too sensitive to water [26‐27]. towels.2. Biodegradable plastics are also used for food servicewares .50/ kg in 2005) is also expected to be lower in the future due to considerable reduction in modification costs . half of its price in 2003 (around 3 Euro/kg). alone or complexed with natural or synthetic biodegradable polymers) or from polylactic acid which originates from the fermentation of a starch feedstock. biofillers for tires. such as polyethylene (PE. starch must be modified.3.50‐4. It cannot be shaped in films with adequate mechanical properties (high percentage elongation. and chewable items for pets [1.3.
3. 30]. approximately 230 oC) . and the properties and amount of plasticizers. sorbitol. A biodegradation study according to ISO/CEN 14852 and ASTM D5209‐92 standards  on films made from starch–glycerol–water mixtures confirmed that the films are easily biodegradable. The plasticizers are also usually hydrophilic and can be
. This problem is mainly caused by the presence of strong inter‐ and intra‐molecular hydrogen bonds between the starch macromolecules [28. water content. because the resulting product will be brittle when equilibrated with ambient humidity . which will result in a continuous phase in the form of a viscous melt [29‐30. 37]. and hexylene glycol. At intermediate water levels. The modification involves break down of the starch granular structure by the use of plasticizers at high temperatures (90‐180 oC) and shear. which still limits their use. phase separation may still occur. There are several substances used as plasticizer for the preparation of thermoplastic starch (TPS).1. extrusion. such as water and polyols (glycerol. 40]. Another issue is the poor water resistance and low strength. TPS synthesized from polyol and sugar plasticizers have the tendency to re‐crystallize (retrogradation) after being stored for a period of time. with better properties than virgin starch in various applications. sugars) [30. 35. et al  discovered that the elongation of break of the thermoplastic starch is significantly improved by plasticization with glycol.Chapter 1
1. For instance. 33]. The use of water as a plasticizer is not preferable. the thermal properties of glycerol‐plasticized starch are a function of water content [31. 40]. 30. Starch can be modified to obtain materials which melt below the decomposition temperature . and blow moulding [28. Thermoplasticized starch Virgin starch is brittle and difficult to be processed into articles due to its relatively high glass transition temperature (Tg. 32. Although thermoplasticization seems to be a promising method. and therefore are processable by conventional polymer processing techniques such as injection. The use of other plasticizers (for example glycerol) results in a rubbery material. . 39]. which is even above the thermal degradation temperature. The brittleness is known to increase in time due to free volume relaxation and retrogradation. The thermoplasticization process will decrease the interactions of the molecular chain and destruct the structure of the starch . the semicrystalline structure of starch and its granular form are lost and the starch polymers are partially depolymerized. As the result. The plasticized starch properties may be tuned by changing the temperature of processing. which results in embrittlement. glycol. Yu. plasticized. glycerol. The modified products are known as thermoplastic. or melted starch [29. resulted in the formation of an amorphous mass [35. 32‐ 36]. destructed.
As a result. Starch Esters The development of starch esters started in the mid 19th century [44. and ethylenic compounds. The esterification reaction is catalyzed by H+ and proceeds with the formation of water . resulting in an increase of the Young’s modulus and tensile strength of the products and a decrease in elongation at break. et al  studied the mechanical properties of films from maize starch cross‐linked by sodium trimetaphosphate (SMTP). High DS starch acetate can be easily casted
. 39]. leading to higher viscosities [42‐43]. The cross‐linked products are therefore more rigid (and less elastic) materials than virgin starch .3‐1) are still soluble in water. paper. The simplest starch ester is starch formate (C1). Solutions to improve the properties of TPS are blending. The cross‐linked starches have found many applications. Schuetzenberger acetylated starch with acetic anhydride . textile.3. These hydroxyls can react easily with a wide range of compounds such as acid anhydrides. Low DS products (0. epoxy. or coating with hydrophobic polymers [33. Cross‐linking results in a reduction of the solubility in water and to thickening. the use of formyl esters of starch is much lower than that of acetate esters. 1. aldehydes. Medium DS starch acetates (0.2. Substitution of the hydroxyl groups of starch with acetate groups makes the esters more hydrophobic than native starch. Already in 1865. Most studies dealt with the synthesis of starch esters of C1‐C4 carboxylic acids. synthesized by direct addition of formic acid to starch at room temperature. Higher processing temperatures generally led to higher cross‐linking levels. The products are called cross‐linked starches. differing in Degree of Substitution (DS). Three different types of starch acetates may be distinguished. Chemicals of these classes having two or more of the reactive groups may react with two or more hydroxyls of the starch molecules. Breakdown of the starch and the formation of low molecular weight products occur to a significant extent .2) are commercially available and used as food additives and in the textile industry. especially in the food.3. Silva.01‐0. Cross‐linked starch An anhydroglucose molecule of starch contains two secondary and one primary hydroxyl group. The most popular starch ester is starch acetate . and particularly with acetic acid . and adhesive industry. organic chloro‐compounds.3. 46]. Introduction
washed out by water . The DS is defined as the moles of substituents per mole of AHG units . 1. while highly substituted starch acetate (DS of 2‐3) are soluble in organic solvents.
Another attractive route involves the use of vinyl esters as reagents . Starch esters have also been synthesized using alkanoyl chlorides [49‐53].Chapter 1
into films using organic solvents. allowing the polymer to be produced with a range of hydrophobicities . while alkaline catalysts (such as carbonate. High DS starch acetates are thermoplastic materials suitable to be used as biodegradable plastics . Starch triacetate has been successfully synthesized using acetic anhydride in combination with pyridine‐gelatinized starch . and is shown in Scheme 1. DMSO. Esterification may be performed using acid anhydrides in aqueous media or organic solvents (pyridine. Mechanism of starch acetylation using vinyl acetate
. Acetylation of starch in water and DMSO using vinyl acetate has been studied lately by Mormann and Al‐Higari  as well as Dicke . hydrogencarbonate. Several synthetic routes have been developed for starch esters. DMF.1.
Starch OH + OH
O Starch O
+ H3C C O CH
O H3C C O
C O CH O Starch
O Starch + H3C C H
O H3C C O Starch
C O OH
O H3C C OH
Scheme 1.1. et al [55‐56]. The degree of acetylation of starch acetates can be easily controlled. Neutral and weak acid/ alkaline catalysts result in regioselective substitution at the C2 hydroxyl groups of starch. acetate. and phosphate) will result in C‐2. The kinetics of the reaction between gelatinized aqueous potato starch and vinyl acetate was studied by De Graaf. or isopropanol) with acidic (hydrochloric or sulfuric acid) or alkaline (NaOH or triethylamine) catalysts . xylene. but also C‐6 and C‐ 3 substitution .
the two components are chemically linked to each other. interfacial agents (e. The use of methyl and glyceryl esters of fatty acid (in the absence of solvent) to synthesize starch fatty acid esters has also been studied . but only relatively low‐DS (0. As a result.34‐0. Introduction
The synthesis of long chain fatty acid ester of starch has attracted much interest lately. The fatty acid chloride reactants are. the latter are mixed together.e. The second methodology for producing polymer blends involves the in situ polymerization of one component (thus originally present in the blend in monomeric form) in the presence of the second one. Although the in situ polymerization process is not as technologically straightforward and economically convenient as melt‐mixing. however. 1. However. The introduction of longer acid chain is expected to reduce the brittleness of virgin starch and to increase its hydrophobicity . polymers are rarely miscible with each other  so that. octadecanoyl) chlorides [49‐53]. Despite this advantage. Two main methodologies are applied for the production of polymer blends. it is frequently used in order to chemically graft the polymerized component on the other one (polystyrene on polybutadiene in our example above ). relatively expensive and rather corrosive. in order to achieve also a good molecular adhesion between the phases by melt mixing. Starch ‐ Biopolymer Blends and Graft Co‐Polymers Blending of different polymers is an established method to obtain products with improved properties. The degree of adhesion (binding) between the dispersed phase and the matrix is dependent on the molcular interactions between the two components and represents a crucial factor in determining the morphology of the blends and eventually the product performance . dodecanoyl. in the simplest case of a binary blend. However. The first involves simple melt mixing of the two components for example by extrusion. The classical example of such process is represented by the production of High Impact Poystyrene (HIPS) obtained by styrene polymerization in the presence of polybutadiene . i. a chemical reaction might take place upon processing (reactive extrusion). compatibilizers) might be used. If the right combination of chemical groups on the two components is present along the polymers backbone. By working at temperatures above the melting point or glass transition temperature of the two components.4. melt‐mixing remains the preferred route to polymer blends mainly because of very practical reasons: low costs. one component will be dispersed into the other. they locate themselves at the interface between the
.61) products could be obtained using this approach.g. availability of mixing equipment and no necessity to use any organic solvent (often employed for the in situ polymerization). Their role is comparable to the one of a surfactant in emulsion formation .3. Fatty‐acid starch esters have been synthesized using fatty acid (octanoyl. which in turn provide a very strong adhesion at the molecular level between the dispersed phase and the matrix.
The blending and grafting of starch with synthetic polymers is usually performed to achieve higher hydrophobicity and to improve the mechanical and thermal properties as well as to obtain cheaper and more biodegradable products. The starch/polyethylene blends are used for agricultural mulch [27. The most often used synthetic polymer for blending with starch is polyethylene [27.1.6.Chapter 1
two components stabilizing the dispersion (most probably by a steric repulsion mechanism ) and providing an improved adhesion at the interface.4. 80‐ 81] or food packaging . The uncompatibilized blends of starch and
.). there have been efforts to blend as well as to graft synthetic polymers onto starch. 64‐98]. Interfacial agents already available on the market can be used as such or can be produced upon mixing (in situ) by chemical reaction of the two components. The products have been synthesized in the lab as well as on industrial scale [32. 39. Starch‐based blends by melt mixing Starch‐synthetic polymer blending has been studied as early as in 1973 [64‐65].6. A suitable interfacial agent for the blends of two polymeric materials is a block copolymer for which the chemical structure of every block is the same (or very similar) to the one of the individual components to be blended  (illustrated in Figure 1.3. 66‐81]. The synthesis of compatibilizer and its use for starch/ synthetic polymer blending has also been studied. In the past.
Figure 1. Illustration of the role of interfacial agent in compatibilizing blends 1.
These biodegradable polyesters will finally decompose into non‐toxic products . since the conventional synthetic polymers are usually poorly or non‐biodegradable. Some of these polyesters also have very good mechanical. and poly‐caprolactone). Polycaprolactone (PCL) is a well‐known synthetic biodegradable polyester.e.
. To obtain completely biodegradable products. The ester bonds of these polymers are susceptible to attack by water and this leads to enhanced biodegradability. is formed upon mixing. the compatibilizer. among which synthetic polyesters are considered very promising materials . PCL‐g‐pyromellitic anhydride [94‐95]. 96]. Despite these good results. and of a premade starch‐g‐PCL  for starch/PCL blends resulted in a better dispersion between the phases and in turn in improved mechanical properties compared to the uncompatibilized blends. The use of reactive compatibilizer precursors PCL‐g‐glycidyl methacrylate (PCL‐g‐GMA) [89‐90]. 84‐98]. 39. synthetic biodegradable polymers have been applied. The mechanical properties of these blends (tensile strength and elongation at break) decrease at higher starch content. and poly‐lactones (such as poly‐butyrolactone. poly‐dioxanone. systematic studies on the compatibilizer precursor synthesis as well as on the influence of the chain topology and chemical reactivity (for the in situ compatibilizer formation) have not yet been reported. poly‐valerolactone. while the tensile strength of the compatibilized ones decreases only slightly with the starch content. Studies on the blending of starch with PCL have been already described [32. Examples of these biodegradable polyesters are poly‐glycolide. The tensile strength of the uncompatibilized blends is drastically reduced when starch content is increased. thus improving the dispersion of starch in the PE matrix. while PE‐g‐GMA posseses epoxide groups. 72‐73]. thermal and water/gas permeability properties that are even comparable to bulk non‐biodegradable polymers such as PE and PP. Polyethylene‐g‐maleic anhydride (PE‐g‐MA) and polyethylene‐g‐glycidyl methacrylate (PE‐g‐GMA) have been used as reactive compatibilizer for starch/PE blends. poly‐lactides. As a result a graft copolymer (PE‐g‐Starch). i. Introduction
polyethylene show a coarse phase separation due to differences in polarity of starch (hydrophilic) and polyethylene (hydrophobic). The mechanical properties of the blend are also improved. the uncompatibilized blends of starch with PCL give coarse phase separation and a reduction in the mechanical properties when the starch content is increased. and EVOH . As is the case of blends with PE. PCL‐g‐maleic anhydride (PCL‐g‐MA) [91. which both can react in situ with the hydroxyl groups of starch [70. dextran‐g‐PCL . PE‐g‐MA contains reactive anhydride sites. The use of conventional synthetic polymer such as PE for blending with starch will only result in a partially biodegradable material. which combines excellent biodegradability with acceptable mechanical properties.
Three routes have been studied in detail: starch esterification.3. leading to formation of PCL homopolymers rather that starch‐g‐PCL. Thesis Outline
The objective of this thesis is to study synthetic pathways to obtain starch derivatives with the potential to be used as bioplastic. Starch is insoluble in the typical organic solvents used for ROP (such as toluene or THF). defined as the percentage of grafted polyester to starch compared to the total amount of homopolymer and grafted polyester) . starch/ biopolymer blending. Chapter 2 describes the ROP of p‐dioxanone initiated by a protected monosaccharide (1. Previous studies showed that common ROP catalysts such as tin octoate or aluminium isopropoxide gave low (0‐14%) grafting efficiencies (GE. thus resulting eventually in low GE values . Starch‐based blends by in situ polymerization The synthesis of starch based graft copolymers by in situ polymerization represents not only an alternative route to melt blending (vide supra) for the production of novel biomaterials.4.4‐di‐O‐isopropylidene‐α‐D‐galactopyranose) using i Al(O Pr)3 as the catalyst. Another approach consists of the Ring Opening Polymerization (ROP) of ε‐caprolactone monomer in the presence of starch. Starch is a very hydrophilic material that always contains moisture. and difficult to handle since it releases ethane. Another possible cause for the low GE values is the heterogenous nature of the reaction.2. but it can also provide an efficient synthetic methodology for the production of a compatibilizer to be used for melt blending of starch with a biopolymer .3.2. and starch‐g‐ biopolymer formation. In this reaction. which however is extremely air‐ and water‐sensitive. the synthesis of starch‐g‐PCL is reported. the hydroxyl groups of starch are supposed to function as initiating sites. The results of this study have been used as input for the synthesis of the starch‐g‐PCL. The highest GE value (up to 90%) has been achieved when using triethylaluminium as catalyst [97‐98]. In the past.Chapter 1
1. and the presence of liquid‐solid reaction system leads to reduced reaction rates between starch and ε−caprolactone. The water in starch granules competes with the hydroxyl groups of starch in the initiation step of the polymerization reaction. starch‐g‐PCL has been synthesized using toxic materials such as isocyanates . a very flammable by‐product.
1. In Chapter 3. during the reaction. A new strategy for the in situ ROP of ε−caprolactone on starch with the use of common ROP catalyst is therefore highly desirable.4. The method applied basically consists of three steps (temporary partial protection of starch‐hydroxyl
Passchall). Based on these preliminary results. Poirier. Princeton University Press. Wurzburg).L. Washington D. O. A preliminary study on the effect of several process variables (reactant ratio. York. Wurzburg: Introduction. P. 2000‐2005. are also reported.S. . . C. Holmes: Interactive European Network for Industrial Crops and their Applications. A. Summary Report for the European Union. .C. The results were quantified using a statistical model. Inc. Corn. Bangor. 2007
. Bemiller. Sand Hutton. Academic Press.starch. The synthesis of two interesting compatibilizers (PCL‐g‐GMA and PCL‐g‐DEM) for starch/PCL blends is discussed in Chapter 6. Stevens: Green Plastics. Breed. Boca Raton. 105‐122. The use of these compatibilizers in starch/PCL blends. together with the mechanical and thermal properties of the synthesized starch esters are reported in Chapter 5. C.
1. 1984. Mol. UK. Gywnedd.B. O. addition of co‐solvent. C. accessed on August 01. and polyhydroxyalkanoates. Fowler: Industrial markets for starch. References
. . Aarhus.. The Biocomposites Centre. London. . 2008. . 1. Introduction
by trimethylsilyl groups. UK. CRC Press. Website of International Starch Institute. Somerville: Mini review: Plant polymers for biodegradable plastics: cellulose. . Corn Refiners Annual Report 2007.html. a systematic experimental study on the effect of the process variables on the starch ester DS has been performed. ROP of ε−CL on the remaining starch‐hydroxyl groups. USA. R. USA. E. 2002. 2005. USA.5. 1986.dk/isi/stat/rawmaterial.B. in Starch: Chemistry and Technology (Ed.H. 1995. Nawrath. Y. UK.N. starch.. E.M. and removal of the silyl groups). Denmark: http://www. J. An Introduction to the New Science of Biodegradable Plastics. New Jersey. Central Science Laboratory. Young: Fractionation of starch. R. University of Wales. and application of different catalysts) on starch esterification using vinyl fatty esters (vinyl laurate and vinyl stearate) is provided in Chapter 4.F. Part of A Global Economy. Agricultural and Rural Strategy Group.. Inc. Corn Refiners Association. including the mechanical and thermal properties. Whistler. in Modified Starches: Properties and Uses (Ed. The model. de Bragança.A. 2004.
Selke: Packaging and the Environment. Hernandez. demand and recovery for 2006 in Europe. Prentice‐Hall. Chandra. USA. Bastioli: Global status of the production of biobased packaging materials. Germany. Cleaner Prod. in Kirk Othmer Encyclopedia of Chemical Technology Volume 22. Website of Environmental Protection Agency. Eng. R. Weith: Plastics.M. Munich. C. Crank.E. Washington DC. Lancaster.N. Culter: Plastics Packaging. R. Marscheider‐Weidemann. M. Weber). Sci.R. USA. J. C. Hanser Publishers.B. U. . J. S. 1998. Patel. .A. Watson: Corn and sorghum starches: production. Bemiller. Whistler. A. R.Chapter 1
.gov/garbage/plastic. Inc. Technomic Publishing Company. Inc. J.. Whistler. Academic Press. Passchall). S.epa.L. Link. . Munich.J. . Properties. 2000. A. 1990. in Starch: Chemistry and Technology (Ed. X. . Properties. Daniel: Starch. . 2000. Inc. accessed on August 19.E. A. USA: http://www. . 1273‐1335 . Denmark. and Regulations. the Food Biopack Conference Proceeding (Ed.M. Ren: Biodegradable plastics: a solution or challenge?. 23. UK. . Wolf. Fifth Edition (Ed. Hoboken. Applications. Frederiksberg. Strong: Plastics: Materials and Processing. Bastioli: Global status of the production of biobased packaging materials. Braun: Environmental impacts of biobased/ biodegradable packaging. Inc. 27‐40. Applications. 53 . 11. London. S. Schleich.F. Selke... G. The Royal Veterinary and Agricultural University. The compelling facts about plastics: An analysis of plastics production.D. John Wiley & Sons. .W.J. Institute for Prospective Technological Studies and European Science and Technology Observatory.. Englewood Cliffs. Brussels. Angerer: Technical Report Series. Association of Plastics Manufacturers. C. Starch/ Starke 2001. 2008 .L. J. Germany. Ind. 105‐109 . Husing. Techno‐economic Feasibility of Large‐Scale Production of Bio‐based Polymers in Europe. Department of Dairy and Food Science. F. Fritz. Hanser Publishers. Plastics Europe. M.htm.J. 2003. J. Processing. Belgium. 557‐562. 2008. J. Prog. . O. Ehrenstein: Polymeric Materials: Structure. . E. G. 2000. USA.
. 1984. R. Seidel). B. Chem. Polym. 2005. Rustgi: Biodegradable polymers.351‐355. 31. 1939. Starch/ Starke 2001. New Jersey. 1996. R. 53. R.
F. J. Press Release: Major Price Hikes in Plastics. 494–504 . status and perspectives. 62. 2005. 1980. 409–416 . . 1996. A. 2000. Y. 592‐595. Starch/ Starke 2001. Parker: Phase and glass transition behaviour. Schwach. Martin. Chambre Syndicale des Emballages en Matiere Plastique (French Plastic and Flexible Packaging Association). R. Ind. Sci. Mikkilti. 21. 13. Wang: Properties of starch blends with biodegradable polymers. Yang. Ma: The effects of citric acid on the properties of thermoplastic starch plasticized by glycerol. J. Petersen. C. Smits. 2005. Introduction
European Commission Directorate General of Joint Research Centre. 2003. Frederiksberg C. K. K. 356‐361 . 275‐282
.B. Gandini: Surface chemical modification of thermoplastic starch: reactions with isocyanates. Yu. P. measured by differential scanning calorimetry and solid state NMR spectroscopy. Compos. July 15. X. a model for thermoplastic starch.H. Carvalho. Muller. Vliegenthart: Interaction between dry starch and plasticisers glycerol or ethylene glycol..G. O. Avérous. Y.M. the Royal Veterinary and Agricultural University.A. Carbohydr. P. Wang. interaction. Res.. Dev.M. M. Yu.J. Ind. Prod. B. 2008.G. 53.M. J. Interfaces 2006.V. F. 1–17 . report of Food Biopack Project (Ed. Starch/ Stärke 2001. and Pukanszky: Thermoplastic starch/layered silicate composites: structure. K. 34. Department of Dairy and Food Science. Physical and mechanical properties of biobased materials – starch. Wang. Weber). Otey. A. Moates. Chem. of concentrated barley starch‐glycerol‐water mixtures. 53. Spain.M. epoxy functions and stearoyl chloride. Eng. R.H. Macromol. van Soest. 1491‐1494 . Biobased packaging materials for the food industry. 57. J. G. 43. P.K. Kruiskamp. polylactate and polyhydroxybutyrate. Forssella. properties. 1997. Part C: Polym. Nielsen.. Biodegradable thermoplastic starch. 331–336 . . T. 4.J. Starch/Stärke. A Real Case of “Force Majeure” that should be passed through to Packaging Prices Immediately. A. Polym.P. Appl. Polym. 385–409 . Westhoff. 2005. Doane: Starch‐based blown films. L. . X.S. P. J. A. E. J. Crops Prod. Polymer Sci. Gao.F.L. W. J. N. and Olsen. J.Lin: Note. 53. Curvelo. Carbohydr. Bagdi. Seville. J. Rev. 2003. and Couturier: Properties of biodegradable multilayer films based on plasticized wheat starch. 372–380 . . Denmark.
.A. O. Polym. 55‐78. Ramsay. 11–17. 1. W. 1999. 1986. Crops Prod. E. Wurzburg. .C.F. CRC Press.B. Rodriguez‐Gonzalez. B. D. Appl.B. 1986. 389‐398 . Copinet. Wurzburg). Matzinos. Divers. W. Polym. Gattin. 2004. R. Environ. in Modified Starches: Properties and Uses (Ed.H. Boca Raton. V. I. 58. Coutuier: Comparative biodegradation study of starch‐ and polylactic acid‐based materials. CRC Press.
. F. E. Mullen. Aburto.W. Ibezim. Pacsu: Starch studies: Possible industrial utilization of starch esters. USA. Polym. E. A. C. Polymer 2003. 46–51 .. M. C. A. Sun: Reinforced thermoplastic acetylated starch with layered silicates. 1517–1526 . Borredon. Carvalho.D. 44. M. Wurzburg: Cross‐linked starches. J. 34. USA. Inc. K. 17–24 . E.D. Carbohydr. 2006. Eng.J. Ind. Boca Raton. Favis: Rheological and thermal properties of thermoplastic with high glycerol content. Levesque.D. 1942. 705‐721. C. Y. Sci. Polym. Starch/Stärke 2005. C. 2002. destructuration and hydrolysis during O‐formylation. Polym.W. S. J.W. Bikiaris. G. M. O.. 1647‐1656.J. J. H. A. Kontoyiannis. B.C.D. Jarowenko: Acetylated starch and miscellanous organic esters. B. S. Silva. T. Tserki. Myllarinen. Mullen. Sci. J. Ind. Chem. van der Zee.A.A. Vikman. Feller. P. Pacsu: Starch studies: Preparation and properties of starch triesters. P. T. B. 74. . 77. Merrill: Properties of fatty‐acid esters of starch. Degrad. 35. Ribeiro.W.B. Grohens: Starch modifications. 2594–2604 . . Pilin. 1997. Ind. C. Stab. Panayiotou: Properties of fatty‐acid esters of starch and their blends with LDPE. 2001. Thiebaud. Andrade: Reactive processing and mechanical properties of cross‐linked maize starch. 139–147 .A. F. Y. (editor). 24. Journal of Applied Polymer Science 1995. O. 581–586 . . Qiao. 58 (9). Ramsay. Hulleman.P. Chem. Bertrand. Prinos. Eng. 65. J. Appl. I. . Rodriguez‐Gonzalez. Panayiotou: Processing and characterization of starch/polycaprolactone products. Alric. 381‐384 . Feil: Morphology and enzymatic degradation of thermoplastic starch– polycaprolactone blends. Favis: High performance LDPE/ thermoplastic starch blends: a sustainable alternative to pure polyethylene. Inc. Jiang. J. E. Sagar. 1209‐1217. X. 57. 56. chapter 4 in book Modified Starches: Properties and Uses. 1943. J.T. 2004.
New York. 11. M. Tuschoff: Acylation of hydroxy compounds with vinyl esters. E. M. P. Mouysset‐Baziard. S. Hu. Appl. A. .F. C. R. .Stokes. G. Introduction. E. Starch‐Starke 1999. Duquesne. P. 74. J. P. Senocq. C. R. M. G. Alric. Aburto. Mormann. Starch‐Starke 1999. W. P. Guegan. A.M. S. 5590‐5598 . A. J. P. Aburto. Inoue: Compatibilizers for melt blending: Premade block copolymers. 1997. B. Bikiaris. Alric. D. H.E. Aburto. Panayiotou: Synthesis. Carbohydr.H. E. 51.V. Degee. 1995. 1991. I. Polym.Painter: Specific Interactions and the Miscibility of Polymer Blends. Symp. Lancaster. Rijksuniversiteit Groningen.J. 34. Dubois: Synthesis and characterization of compatibilized poly(ε‐caprolactone)/ granular starch composites. Nakayama. 101‐112. Macosko. 1999. A. R. Prinos. 1960 . 145‐152. Dicke. USA. 57.828. Alric. C. . D. 56. E. 28. Polym. Introduction
. and biodegradability of fatty‐acid esters of amylose and starch. . Broekroelofs. W. D. Aburto. Smith. Technomic Publishing Company. . Panayiotou: Properties of octanoated starch and its blends with polyethylene. A. Macromolecules 1996. 132‐135. 302‐307. Scott. 51. Borredon. Aburto. Borredon: Preparation of long‐chain esters of starch using fatty acid chlorides in the absence of an organic solvent. 33‐43
. G. Germany. Starch‐ Starke 2005. Borredon: Organic solvent‐free transesterification of various starches with lauric acid methyl ester and triacyl glycerides. 2001. 137‐144. Alric. A. . Hanser Publishers. characterization. F. 1997. Sci. C. J. 255‐263. J. R. .D.H. US Patent 2. W. A. Al‐Higari: Acylation of starch with vinyl acetate in water. I. Rutot. Beenackers: The kinetics of the acetylation of gelatinised potato starch. Starch‐Starke 2004.E. in Reactive Polymer Blending (Ed. Thesis. J. I. Borredon: Free‐solvent synthesis and properties of higher fatty esters of starch ‐ Part 2. E. Cellulose 2004.928. 2001. de Graaf. Polym. T. Wiley‐VCH. P. Hu). .. Thiebaud. E.Coleman. D. Marechal. USA. Carbohydr. I. J. C. Prinos. A.W.Fennel Evans: Fundamental of Interfacial Engineering.Graf. J.C. Borredon. Khandpur. J. Bikiaris. Thiebaud. J.E. A straight way to regioselectively functionalized polysaccharide esters. M. Bakker and G. Hamaili. I. Bakker. Macromol. . 29. 1440‐1451. J. de Graaf: Ph.K. C. Janssen. Inc. L. 118‐121. 1996 . . . Alric.E. Munich. 175.
S. 88. U. 2005. 1998. R.A. Ismail. Proc. D. Jane.M. Mohd Ishak: Material properties.G. Rosa: Mechanical. Ismail. Res. N. Polym. Tech. Pavlidou. Cordi. 236–241 . Stab. J. 1242‐1251 . Danjaji. Adv. Ribes‐Greus: Thermal and viscoelastic properties of some commercial starch products. G. Ramsay. L. 1467‐1475. 75–81 . 21.Chapter 1
. Appl. Polym. Ramsay.S. K. A. R. Degr.I. Pedroso. Mei: Study and development of LDPE/starch partially biodegradable compounds.S. Testing. Danjaji. 2003. J.S. N. 1841‐1846 . 1998. 162–163. B. Polym. L. Nawang. 1997. Mater. Verhoogt: Processing and characterization of thermoplastic starch/polyethylene blends. Polym. W. Polvm. thermal and morphological properties of virgin and reprocessed LDPE/corn starch blends. morphology. Bikiaris. Appl. Ind. 2001. Evangelista. Z. L. 2005. Polym. St‐Pierre. I. Polym. Z. D.A. Polymer. . Testing. H. 1998. J. Polym. Rosa: Effects of the compatibilizer PE‐g‐GMA on the mechanical. 2002. Liu. C. Technol. Nikolov. Gelina. Carbohydr. 2904–2911. 59. 16. R.S. E. Panayiotou: LDPE/plasticized starch blends containing PE‐g‐MA copolymer as compatibilizer. Z. 70. Koutsopoulos. H. Mohd Ishak: Material behaviour. Panayiotou: LDPE/Starch blends compatibilized with PE‐g‐ MA copolymers. Wang.G. 38.. Polym. 1991. J. Nakamura. 59. 2005. H. 647‐655
. Frangis. J. 88. Effect of amylopectin to amylose ratio in starch.M. Degradation studies and moisture uptake of sago‐starch‐ filled linear low‐density polyethylene composites. and tensile properties of Low‐Density Polyethylene (LDPE) and corn starch blends.L. A. 2003. Bhattacharya: Properties of injection moulded starch/ synthetic polymer blends.H. Sung. W. Almeida. Sun: Effects of polyethylene‐grafted maleic anhydride (PE‐g‐MA) on thermal properties. E. Appl. 30. D. Prinos. Effect of compounding and starch modification on properties of starch‐filled Low‐ Density Polyethylene.A. Sci. 167–172 . R. Mechanical properties of sago starch‐filled linear low density polyethylene (LLDPE) composites. Ishiaku. A. 20. Z. Eng.G. R. 1‐9 . thermal and morphological characterization of recycled LDPE/corn starch blends. Favis. Vouroutzis. U. Chem. Nawang. 287‐291 . J. Y. D. Eur. I. Sci. Pedroso.D. Duran. M. J. .J. N.‐J. III. 310–317 . Mani. 34. 1503–1521 . Contat‐Rodrigo. J. Sci. L. Ishiaku. B. N. Bikiaris.D.D.A. C.
cross‐linked and glycerol modified starch. . Kim. Ryner. . L. 43. W. 87.S. 26. Otey. J. J.P. 284‐287 . 179‐187. Polym. Albertsson: Controlled Ring‐Opening Polymerization: Polymers with designed Macromolecular Architecture. 2002. Kammer: On mechanical properties of sago starch/poly(ε‐caprolactone) composites. A. Preliminary diffusion evaluation. Jung. A‐C. Sci. 1999.S. Polym. U.G. 35. 89.M. 81.. Westhoff: Starch‐Based Films. Ind. Adv. . J. K. Tech. 877–884. J. B. Z. Polym. Sci.‐C. Environ. N. Ishiaku. Sci. B. F. Manan. 157. Liu. M. 4157–4167 .H. Ishiaku. 2004. Azemi. Jana. Azemi: Effect of starch predrying on the mechanical properties of starch/poly(e‐caprolactone) composites. Bhattacharya: Synthesis and characterization of anhydride‐functional polycaprolactone. M. Polym. M. U. 1139–1148.P. 1984. Manan.M. Odusanya. 3539–3546 . D. 23. and their blends with starch in soil compost.M. Averous. Introduction
. Garg.M.H. J. F. Jana: Studies on the properties and characteristics of starch– LDPE blend films using cross‐linked. 267. Park: Reactive blends of gelatinized starch and polycaprolactone‐g‐glycidyl methacrylate. 3976–3987 .W.M. 41. Polymer 2000. Westhoff. Res. Siddaramaiah.H. A.K. Chem. dos Santos Rosa.H. Kim. Res. Rodrigues. . Sankar. J. in Polym. J. John. PHB‐V. 1997.N. Karlsson: Polymers from Renewable Resources. Stridsberg. Polym. 2007. R. 42‐65. Ind. Ang. J. Lindblad. M. J. B. Polym.S. Sci.S. A. K. O. Eng. . Odusanya. 23. D. B. 1. Appl. Dole. Moro. 12.K. Appl. E.
. Kim. 139‐161. C. Guedes. 1659‐1663 . Tang. Part A: Polym. C. Yang. Maiti: A biodegradable film. Sci. Chem. H. 157. T. Prod. L. Otey. M. 32–45 . Polym. Appl.K. 2004.Y. Adv. Pilot plant investigation for production of the biodegradable film. P. 2. 41. Eng. Bhattacharya Residual stresses and density gradient in injection molded starch/synthetic polymer blends Polymer 2000. 2001. S. Dev. K. 1987. Low density polyethylene/ starch blend films for food packaging applications. T. Polym. Cho.R. 2002. Albertsson. Makromol. Polym. 40. Ranucci. S. M. Doane: Starch‐based blown films. J. Calil: Effect of thermal aging on the biodegradation of PCL. 2003. glycerol modified. 1298‐1305. 16–19 . 2003. R..C.F. Eng.. C. Adv. Sen. Park: Modification of aliphatic polyesters and their reactive blends with starch. . S. Fringant: Properties of thermoplastic blends: starch–polycaprolactone. U. 2000. 1507– 1516. O. Y. Eur. S. S. M.D. J. Raj. Chem. C. Chem. Sci. A. 9177–9190 . Sci.
Macromol. Narayan. Stab. and P. Degee. 41. M. Errico. R. R. Dubois. Sci. 3091‐3100. Tang.
. Polym. Raimo. 2003. 80. Polymer 1999. Rutot. P. Rimedio. 215–225. Mani. 3875–3881 . Sadocco: Preparation of biodegradable polyesters/high‐amylose‐starch composites by reactive blending and their characterization. Degée. J. M. 33–43 . 1432–1442. P. M. P. M. M. R. 19. Dubois: Synthesis and characterization of compatibilized poly (e‐caprolactone)/ granular starch composites. 127–134. M. Symp. Duquesne. 175. D. Wu: Physical properties and biodegradability of maleated‐ polycaprolactone/ starch composite.S. D. . .E. J. P. Dubois: Aliphatic polyester‐grafted starch composites by in situ ring opening polymerization.Chapter 1
. R. 40. Avella. . Appl. Polymer 2000. Martuscelli. Avella. 1998. Rutot. E. Rapid Commun. P. Krishnan. Polym. 2000. 2001. R. 83. M. P. Narayan: Aliphatic polyester‐grafted starch‐like polysaccharides by ring‐opening polymerization.E. 2002. Laurienzo. 283–286. Bhattacharya: Synthesis and characterization of starch‐ graft‐polycaprolactone as compatibilizer for starch/polycaprolactone blends. Degrad. E. Rimedio: Polymer communication: Preparation and characterisation of compatibilised polycaprolactone/ starch composites. C. Macromol. Errico. Composite Interfaces. . 7.
The products were synthesized at 60‐100°C and characterized by 1H‐ and 13C‐NMR. A statistical model has been developed that successfully describes the experimentally observed relation between the average chain length of the functionalized polymer and reaction parameters. Besides the desired polydioxanone functionalised with a monosaccharide end‐group. and MALDI‐TOF mass spectrometry. ring‐opening polymerization
Experimental Studies on the Ring Opening Polymerization of p‐dioxanone using an Al(OiPr)3‐Monosaccharide Initiator System
The ring opening polymerization (ROP) of p‐dioxanone using a protected monosaccharide (1. also polydioxanone with an OiPr end‐group was formed (20‐30 %). with higher temperatures (100°C) leading to lower yields. Keywords: biodegradable.3.4‐di‐O‐isopropylidene‐α‐D‐galactopyranose)/Al(OiPr)3 initiator system to yield polydioxanone with a protected monosaccharide end‐ group is described. polyesters. Systematic studies showed that the polymer yield is a function of the reaction temperature and the reaction time. The average chain length of the polymers is between 7 and 58 repeating units and may be tuned by the monomer to monosaccharide ratio (at constant Al(OiPr)3 intake).
The catalysts are metal alkoxides with Lewis acidic character [8. probably because the p‐dioxanone monomer has become commercially available only recently . particularly for nano‐ encapsulation systems for drug delivery .3 MPa and an elongation at break of 500‐600%. Several polymers with bioactive alcoholic and phenolic end‐groups of interest for drug‐related applications were synthesized (for example geraniol. This melting point is considerably higher than typically found for polycaprolactone (60 oC) and lower than that of polylactides (at least 175°C). Polydioxanone (poly(p‐dioxanone) or poly(1. The synthesis of protected monosaccharide end‐capped biodegradable polymers is usually performed via Ring Opening Polymerization (ROP). pregnenolone. which is a unique compromise between application and processing temperature. 9]. The relatively low melting point of polycaprolactone limits its applicability. reaction of the metal‐alkoxide with an appropriate alcohol allows the synthesis of end‐capped poly‐lactones. Biodegradable aliphatic polyesters end‐capped with sugar molecules have been studied extensively for use in biomedical applications.2.3. From a biodegradability point of view. has interesting product properties compared to other aliphatic polyesters. polyglycolide. and is tougher than polylactides and even HDPE . stigmasterol and ergocalciferol)  and also with protected monosaccharides (galactopyranose/ glucofuranose) [6‐7]. Exchange of the alkoxide group by e. testosterone.g. Despite its good properties. polydioxanone also shows good performance. or zinc lactate has been reported. This chapter describes experimental studies on the catalytic ROP of p‐ dioxanone in the presence of a protected glucose molecule (1.4‐di‐O‐ isopropylidene‐α‐D‐galactopyranose. quinine. stannous octoate [Sn(Oct)2]. Nishida et al  reported that polydioxanone decomposes to non‐toxic gases (CO2 and H2O) by microorganisms. only limited information about the synthesis and properties of polydioxanone is available in the open literature. and polylactides. tocopherol. The ring‐opening polymerization (ROP) of p‐dioxanone using metal catalysts such as aluminum isopropoxide [Al(OiPr)3]. Polydioxanone has a tensile strength close to 48. whereas the high temperature for polylactides results in thermal degradation and undesirable transfer reactions during synthesis and processing steps . Besides
. which to the best of our knowledge is the first study to functionalise polydioxanone with a monosaccharide. such as polycaprolactone.Chapter 2
2.1. are interesting polymers because of their good product performance and biodegradability . 2). accessible by the polymerization of p‐dioxanone (1). It is fully degraded in the body within a period of 180 days .4‐dioxan‐2‐ one)). Its melting temperature is close to 110 oC. The alkoxide group will end up as an ester end‐group in the polymer and in this way at least one of the end‐groups may be easily controlled and varied. Introduction
Toluene (Labscan) was dried and stored on molecular sieves 3 Å (Labscan) under nitrogen.2. Subsequently. 5.55 mmol of 2 and 0. 0. The hot solution was
.2.2. the results of this study are also of interest for the preparation of starch/polydioxanone polymers using ROP.8 ml of a solution of an aluminum isopropoxide stock solution was added. 2. However. As such. heptane (Acros). The polymerization was allowed to proceed for 16 h at 100 oC.4‐di‐O‐ isopropylidene‐α‐D‐galactopyranose (2). Materials p‐Dioxanone monomer (1. the synthetic pathways and the soluble. Boehringer Ingelheim. Particularly interesting in this field are starch polymers grafted with polydioxanone. This stock solution was prepared by adding 4. relatively low molecular weight compounds reported in this study may be viewed as model systems for more complex. which was pre‐heated till about 60°C to obtain it in a liquid state. Next.2. part of this solution (350 µl. Methods All polymerization experiments were carried out under a protective nitrogen atmosphere using standard Schlenk‐ and glovebox techniques. containing 0.1.08 mmol aluminum isopropoxide) was added to pure 1 (0.4‐di‐O‐ isopropylidene‐α‐D‐galactopyranose (2) 2 (1.2. 2.3. The reaction mixture was clear and colorless during the reaction.8 g. Analytical grade dichloroethane (Labscan). 97% (Sigma) and aluminum isopropoxide.g.
2.5 mmol) was dissolved in toluene (1 ml) at 50 oC. 3]. 8 mmol). poorly soluble heterogeneous systems. and diethylether (Labscan) were used as received.44 g. Materials and Methods
2.1.2.07 g.3. the mixture was brought to room temperature and several drops of HCl (1 N) were added to stop the reaction. hot dichloroethane (20‐25 ml) was added to completely dissolve the partly solid polymer at room temperature. 1. Germany) was purified according to the procedure described by Raquez et al [2.2. After the pre‐determined reaction time. (20 mmol) of aluminum isopropoxide to 20 ml of toluene. NMR) due to the poor solubility of the products in standard organic solvents.2. grafting efficiencies are difficult to determine in this system by standard analytical techniques (e. 98+% (Aldrich) were used as received. The resulting clear solution was stirred for 2 h at 50 oC. To this solution. Ring Opening Polymerization of p‐dioxanone using an Al(OiPr)3‐Monosaccharide Initiator System
potential applications in biomedical products. Typical example for the synthesis of polydioxanone end‐capped with 1. CDCl3 was obtained from Sigma and was used as received.
The isolated yield at this condition was 68%.
calculated by comparing the peak area of characteristic end‐group protons with that of a proton of the repeating unit in the polymer (A H‐polymer). 40 oC) until constant weight. MALDI‐TOF spectra were recorded on an Applied Biosystems Voyager DE‐PRO machine using dithranol/ NaI as the matrix (linear mode). 2. The white solid was finally separated from the liquid by decantation and dried in a vacuum oven (5 mbar.2. Product Analyses NMR analyses were performed in CDCl3.3.)
Here. X n exp of the product. Calculation of Average Degree of Polymerization The Theoretical Average Degree of Polymerization. the above equation simplifies to
= monomer conversion ×
[monomer ]0 [ sugar ]0
(2. Processing of the raw data was performed using VNMR software. 1H‐ and 13C‐NMR spectra were recorded on a Varian AMX 400 NMR. If the amount of sugar is in excess with respect to the aluminum catalyst. This leads to the following equation:
X n exp is
H-galactopyranose end group
(2. 2.4.1. two polymers with different end‐groups are present.)
H-isopropoxide end group
precipitated in a heptane/ ether mixture (300‐400 ml. 4:1 by volume) at 4‐8 oC. it is assumed that all available initiator is used effectively.3. As will be shown later. X n theo in terms of number of monomer units is calculated as follows :
= monomer conversion ×
[monomer ]0 [total initiator ] [monomer ]0 = monomer conversion × 3[ Al (O i Pr) 3 ] + [ sugar ]0
H‐NMR was applied to determine the experimental average degree of polymerization.2.2. one initiated on a galactopyranose molecule and the other on an isopropoxide group. 2D‐NMR spectra were recorded on a Varian Unity 500 NMR.
7 ppm − 6 × A 5.1 ppm
(2.6 ppm ] / 6 A 5.4. an excess of monosaccharide 2 on Al was used (see eq.220.127.116.11. an Al ‐ monosaccharide 2 ‐ monomer 1 mol ratio of 1: 6. R gp .4− 4. The exchange reaction typically takes place at 50°C for 2 h. is calculated using:
R gp = A
H − galactopyranose end group H − isopropoxide end group
(2. the protected mono‐saccharide 2 reacts with Al(OiPr)3 to form the actual catalyst for the polymerization reaction.6 ppm + A 5. is rewritten as:
[ A 3.2.)
2.6: 96 was applied.7.3.)
or.1. In the first step. To compensate for this effect.)
The ratio of the two different types of polymers (either end‐capped with galactopyranose or an isopropoxide group).
CH2OH 6 O 4 H H5 O 3 H O O H 2 O H O 1 α O β O γ Al(OiPr)3 catalyst O O α γ γ' β' β α' H2C O [ C CH2 O CH2 CH2 O ]n C CH2 O CH2 CH2 OH 6 O H O H5 1 4 O H O H 2 3 H O
Scheme 2.1 ppm
(2. To avoid the formation of isopropoxide end groups. eq.). The polymerization reaction was performed as a two step process. in term of the NMR resonances :
R gp = A 5.3.5. Schematic representation of the polymerization reaction including atom numbering scheme
. Screening Experiments Initial experiments to synthesize protected galactopyranose end‐capped polydioxanone (3) were performed using 1. 2.1.4‐di‐O‐isopropylidene‐α‐D‐ galactopyranose (2) and Al(OiPr)3 as the catalyst precursor (Scheme 2.). Typically.6 ppm A 5. Results and Discussions
2. Ring Opening Polymerization of p‐dioxanone using an Al(OiPr)3‐Monosaccharide Initiator System
The NMR peaks of the repeating unit overlap partially with those of the galactopyranose end group. 2.
Proton resonances of the CH3 group of the isopropoxide end‐group. However. of interest is the presence of a small multiplet at about 5.) are not particularly informative. it is not possible to determine whether this is truly an end‐group due to overlapping peaks with protons from the poly(p‐ dioxanone) backbone. Although the spectra clearly indicate the presence of the monosaccharide 2.1.1 ppm. This multiplet is characteristic for the CH proton of an isopropoxide end‐group.7.2. It is difficult to determine the end‐groups of the polymer on the basis of 1H‐NMR.40 ppm) unit are broadened and shifted up to 0. although it is evident that p‐dioxanone polymerisation occurred. H‐NMR spectra (Figure 2.. The off‐white solid reaction product was collected after a dissolution‐reprecipitation process using dichloroethane and a heptane/ diethyl ether mixture. Typical isolated product yields are 68% at these conditions.67.15‐1. 2. Product analyses The products were analyzed using 1H‐ and 13C‐ NMR and MALDI‐TOF. It confirms that polymer initiation not only occurs with the monosaccharide but also with the remaining isopropoxide group of the catalyst precursor (vide infra).1 ppm compared to the monomer.Chapter 2
H H CH2OH iPr O Al O iPr iPr O O O H O H H O H O O H
O H O O H O Al O CH2 O O H O H H O O H O H H O
O 3 H
H2C H2C O H H O H
3 iPr OH
(2. The typical proton resonances of the p‐dioxanone (3.1.78 to 4.1 and 2.3. The OiPr : 2 end‐group ratio for the standard experiment was 1: 2. together with the CH3 groups from the protecting groups of the sugar appear in the range 1. Typical 1 H‐ and 13C‐NMR spectra of 3 are shown in Figure 2. respectively.) In the next step. the in‐situ formed catalyst was reacted with p‐dioxanone monomer 1 at 100 oC for 16 h.51 ppm.1.
Carbon resonances arising from the monosaccharide end‐group are present between δ = 66. For instance.2 and 96. 1. 68.17 and 69.) is more informative and clearly shows the presence of a polydioxanone polymer backbone and a monosaccharide end group.61 (γ’) and 73. H‐NMR Spectra of: (a).2.28 ppm in 2 to δ 66.
. p‐dioxanone monomer.2 ppm in product 3.2 ppm. In addition to the mono‐saccharide end‐ group.1. Particularly the C2‐C6 carbons in the range δ = 66. characteristic resonances of an ‐O‐C(=O)‐CH2‐O‐CH2‐CH2‐OH end group are present at δ = 61. Ring Opening Polymerization of p‐dioxanone using an Al(OiPr)3‐Monosaccharide Initiator System
Figure 2.3. The carbon resonances of the polydioxanone backbone are present at δ = 63. 1 (c).2.4‐di‐O‐isopropylidene‐α‐D‐galactopyranose‐end‐capped polydioxanone. 3
C‐NMR (Figure 2. 1.3.4‐di‐O‐isopropylidene‐α‐D‐galactopyranose.2‐70.2 ppm.3 ppm are shifted considerably. C‐6 is shifted from δ = 62.8. 2 (b).2.52 ppm (β’).
Figure 2. 1.4‐di‐O‐isopropylidene‐α‐D‐galactopyranose‐end‐capped polydioxanone.70 (>C‐(CH3)2) and between δ = 21.
Carbon resonances of the isopropoxide groups from the by‐product (isopropoxide end‐capped polydioxanone) are present at δ = 68.3. 1 (c). the numbering scheme of carbons and protons is given in Scheme 2. The resonances of the protecting group of the sugar appears at δ = 108.97 (>C‐(CH3)2).48 ppm (‐CH‐ (CH3)2) and between δ = 21.73 and 109. D‐NMR (HSQC) was applied for complete peak assignment of the product.1.3.2. p‐dioxanone monomer. 2 (b)..2.97 ppm (‐CH‐(CH3)2). C‐ NMR Spectra of: (a).76‐25.76‐25. An overview of the data is given in Tables 2. A typical example of a part of the 2D‐NMR spectra is given in Figure 2. and 2.4‐di‐O‐isopropylidene‐α‐D‐galactopyranose. 1.
09 4.68 4. which were determined from HSQC spectrum due to overlapping resonances.3.83 p‐dioxanone Monomer (1)
H‐α H‐β H‐γ
4.301 3.732 broad.8
Reactant peaks (ppm) Galactopyranose (2) H‐1 H‐2 H‐3 H‐4 H‐5 H‐6 5. HSQC 13C‐1H spectra of 3 Table 2.78 4.25 3.
.74 3.1.32 * 4.59 4.60 4.26 3.00 4.16* 4.17 3.40
all values were determined using 1H NMR. except values with *.77 3. around 1.54 4. Ring Opening Polymerization of p‐dioxanone using an Al(OiPr)3‐Monosaccharide Initiator System
H‐NMR peak assignments a
Product peaks (ppm) Poly(p‐dioxanone) end‐capped with protected galactopyranose (3) H‐1 H‐2 H‐3 H‐4 H‐5 H‐6 H‐α H‐β H‐β’ H‐γ H‐γ’ H‐OH 5.31 4.50 4.
A typical molecular weight distribution is observed.17 69.80 61.31 70.2 73.Chapter 2
Table 2.92 68. MALDI‐TOF was also applied to characterize the products. corresponding with an average number molecular weight (Mn) of about 1600. An example of a MALDI‐TOF spectrum of 3 recorded in a dithranol/NaI matrix is given in Figure 2.54 166. which is the molecular weight of a repeating dioxanone unit. the ratio of the intensity of the end groups and the polymer backbone peaks is determined. The molecular weight distribution of the major peaks may be represented by the following relation:
m / z = 23 + 260 + 102n
.57 62.75 70.92 66.11/ 170. The product obtained at screening conditions (16 h reaction time at 100°C) contains on average 13 monomer units.07
Reactant Peaks (ppm) Galactopyranose (2) C‐1 C‐2 C‐3 C‐4 C‐5 C‐6 96.)
This series represents a dioxanone polymer end capped with 2 and an additional Na ion.2.58 68.21
NMR analyses also allow calculation of the molecular weight of the products.52 63.2 68. For this purpose.33 70. The difference in molecular weight between the main peaks is 102 g/mol.10 71.09 65.4.
C‐NMR peak assignments
Product Peaks (ppm) Poly(p‐dioxanone) end‐capped with protected galactopyranose (3) C‐1 C‐2 C‐3 C‐4 C‐5 C‐6 C‐α C‐β C‐β’ C‐γ C‐γ’ C‐carbonyl C‐carbonyl‐iPr/gal 96.00 171.63 68.21 70. The latter likely stems from the matrix used to ionize the sample.8.28 p‐dioxanone monomer (1)
C‐α C‐β C‐γ C‐carbonyl
61.28 70. A detailed procedure is given in the experimental section.61 170.
Furthermore. ionized with either Na+ or H+.5 +102
This relation is indicative for the presence of a polydioxanone polymer containing an isopropoxide end‐group. These distributions may be represented by equation (2.).
m / z = 23 + 18 + 102n m / z = 1 + 18 + 102n
. although both with a very low intensity (Figure 2.10. in line with the NMR data.4. Ring Opening Polymerization of p‐dioxanone using an Al(OiPr)3‐Monosaccharide Initiator System
100 90 80 70 60 %-Intensity 50 40 30 20 10 0
600 800 1000 1200
1098.5.9.) and imply the presence of dioxanone polymers with carboxylic end groups. This series may be described by the following relation:
m / z = 23 + 60 + 102n
(2. Typical MALDI‐TOF spectra of 3 A second molecular weight distribution is also clearly visible when enlarging the spectra.11.) (2. two other distributions are present.5. see Figure 2.) and (2.11.10. for details.
2‐(4′‐ hydroxybenzeneazo)benzoic acid (HABA)).2.1
0 600 700 800 900 Mass (m/z) 1000 1100 1200
Figure 2. according to NMR.7
858.7 796. In the first step. the catalyst precursor is treated with monosaccharide 2 resulting in an alcohol exchange
.1. A mechanistic proposal for the ROP of p‐dioxanone with Al(OiPr)3 as the catalyst precursor leading to the desired monosaccharide end‐ capped polydioxanone is given in Figure 2.13].3.3
836. although successfully applied for galactopyranose‐end‐capped polycaprolactone . is not particularly suitable for polydioxanone.6. In addition.g. Various samples with. it proved not suitable to determine the average molecular weight of the products. Most likely the matrix (dithranol/NaI). [6. but in all cases poor quality.Chapter 2
792. Mechanistic aspects Both NMR and MALDI‐TOF measurements imply that the main product is indeed the desired monosaccharide end‐capped polydioxanone. Various other matrices were tested (e. small but significant amounts (20‐30%‐mol) of polydioxanone chains with an OiPr end group are present. different molecular weights were analysed.5. 2. Enlarged MALDI‐TOF spectra for 3 Although MALDI‐TOF clearly demonstrated the presence of various types of end‐groups in the product. The observed differences in the molecular weight distributions of the various samples were only marginal. low resolution spectra were obtained.
termination of the chain growth may occur by reaction with an alcohol.
iPr iPr O Al O iPr + R OH (in excess) O
O Al O R + O iPr OH
O Al O O iPr
O O O O
O Al [O CH2 CH2 O CH2 C ] n OR O O O O Al [O CH2 CH2 O CH2 C ] O n+m OR
+ HX (termination) O Al O R O O Al X O O
O H [O CH2 CH2 O CH2 C ] l by-product O iPr
H [O CH2 CH2 O CH2 C ]n OR
H [O CH2 CH2 O CH2 C ] n+m OR
Figure 2. a p‐dioxanone molecule will coordinate to the Lewis acidic aluminium center followed by an insertion step. Irreversible termination of the polymerisation is performed at the end of reaction period by adding dilute acid to the polymerisation mixture. The termination reaction is known to be reversible and the formed polymer may again react with an aluminum center and continue to grow . During the reaction. Simplified reaction scheme for the ROP of p‐dioxanone catalyzed by Al(OiPr)3.6. Subsequent coordination and insertion of dioxanone molecules leads to the formation of a polymer chain with a monosaccharide end‐group. The termination reactions lead to the formation of ‐O‐C(=O)‐CH2‐O‐CH2‐CH2‐OH end groups. Next. The resulting Al‐alkoxide may again initiate a polymerisation reaction. Ring Opening Polymerization of p‐dioxanone using an Al(OiPr)3‐Monosaccharide Initiator System
reaction and the formation of the desired active catalyst with preferentially all three OiPr exchanged with 2. (R = monosaccharide 2)
6 2.7.).3.28 33.6 t (h) 16 16 16 16 16 16 16 16 16 16 1.6 was applied.8 3.45 19. These may be present in the reaction mixture because the exchange reaction between Al(OiPr)3 and 2 was incomplete.8 2. two different reaction times (1.5 T (°C) 100 100 100 100 100 60 60 60 60 60 60 60 60 60 60 dioxanone/ 2 ratio (mol/mol) 3. Table 2.6 6.5 1.12 42.
Set Sample S111 S112 S113 S114 S115 S211 S212 S213 S214 S215 S221 S222 S223 S224 S225
Overview of experiments a
Processing Condition 2/Al ratio (mol/mol) 6.5 84. formed in the first step of the polymerisation reaction (eq.22 57.3.3 6.10 15.1 3.8 2.3 81.5 92.96
For each experiment.29 14.21 54.3 6.53 59. the exchange reaction between the catalyst and the protected sugar was performed for 2 hours at 50°C The yield is the isolated yield of product 3
.16 46.3 3.22 16. A total of 15 experiments were performed at two polymerization temperatures (60 and 100 oC).6 6.7 2. 2. In all cases.7.1 86.1 2.93 37. a (nearly) fixed Al(OiPr)3 : monosaccharide ratio of 1 : 6.2.85 14. OiPr‐end‐capped poly(p‐dioxanone).11 22. 2.6 6.4 2. Systematic Studies The effect of important process variables (temperature.50 25.2 91.74 18.3‐6.9
7.5 1.08 30.3 and 62.70 58.6 86.5 2.4
2/OiPr ratio Rgp (mol/mol) 2.04 19.).6 6.5 67.5 hrs and 16 hrs) and a p‐dioxanone to monosaccharide 2 molar ratio ranging between 3.6 6.50 35.74 52.49 13.55 Product Properties Avg.81 16.93 8.6 6.6 96.74 23. despite the excess of 2.78 10. the end group may also be formed by a termination reaction with free isopropanol.5 81.3 6.3 6. An overview of the experiments and the results are given in Table 2.0 87. will be formed when the polymerization starts with an aluminum alkoxide with a remaining Al‐OiPr group (eq. Chain Length (Xn
Yieldb (%) 30. 2.6 81.6 89.6 91. However.6 6.36 17.30 62.22 13.3 6.4 2.6 6.6 80.5. time and the mol ratio of monomer to monosaccharide) on the yield.4 3.5 1.9 2. degree of polymerisation of the product and the end group distribution was determined.3 6.Chapter 2
The minor product.1 2.3.5 1.
which is expected to be more pronounced at lower temperatures.). and particularly when comparing the data in set 2 and 3 (60oC. These results were explained by assuming that the reaction is an equilibrium polymerisation and that the equilibrium is shifted to the monomer side at higher temperatures. as suggested by Raquez et al . The effect of the reaction temperature (60 and 100°C) on the product yield at three different dioxanone: 2 ratios is given in Table 2.2. the monomer conversion reaches a maximum value before going down to the equilibrium monomer conversion. Clearly. we conclude that equilibrium conversion is achieved after 16 h and that the lower polymer yields at higher temperatures are due to the slight exothermicity of the reaction. see Table 2. Higher polymerisation yields were obtained at lower temperatures. A more detailed analysis on the actual nature and composition of the polymerization products.3. Ring Opening Polymerization of p‐dioxanone using an Al(OiPr)3‐Monosaccharide Initiator System
2. which reduced to 75% when increasing the temperature to 120oC. One of the experiments (S111) resulted in a very low yield (31%) compared to the other reaction. To the best of our knowledge.
. Similar observations were made by Raquez et al  and Kricheldorf et al  for dioxanone polymerisations using Al(OiPr)3 in the absence of a external alcohol or using benzyl alcohol.5 h is higher than the yield at 16 h.1.). 1. It was shown that at lower monomer to catalyst ratios.3. Another possibility is the occurrence of polymer crystallization.3. For this particular experiment.3. the yields are higher when performing the reaction at 16 h reaction times. with an enthalpy of polymerization of approximately ‐15 kJ/ mol . the equilibrium conversion of dioxanone at 80oC was about 80%.3. For the two experiments with a dioxanone: 2 ratio higher than 16 (≅ 23 and 60). et al . The effect of temperature on the yield of the bulk polymerization of dioxanone using Sn(Oct)2 and Al(OiPr)3 as catalysts (without the use of a second alcohol) has been studied by Nishida et al  and Esteves. However. The effect of the reaction time on the product yield may be derived from the data provided in Table 2. For instance. no detailed explanation has been put forward to explain this anomalous behaviour. the yields for reactions conducted at 60 oC are always higher than those performed at 100 oC.5 and 16 h reaction time). equilibrium conversion is not yet achieved within 1. It is likely that this relatively low molecular weight compound dissolves partly during isolation/ purification procedure leading to lower isolated yields. a low dioxanone: monosaccharide 2 ratio was applied. will be required.2. leading to a low average chain length of the dioxanone polymer (7. On the basis of our experimental data and in line with literature data. the yield at 1. when the reaction is performed at a low dioxanone: 2 ratio (≅ 16). Evidently.5 h. The latter is due to the slight exothermicity of the reaction . Product yield The isolated yields of the reactions are all but one between 67 and 96% (see Table 2.
In Figure 2..6. A clear trend between R gp and the process conditions (temperature. as expected for a typical ring opening polymerisation .) as a function of the dioxanone : 2 ratio at 90% and 100% monomer conversion is also provided. time: 1.7. time: 16 hr Temperature: 60oC. 2. Effects of process conditions on the average chain length ( X n exp ) and end group distribution The effects of the p‐dioxanone: 2 ratio. Average Chain Length ( X n exp ) of the product as a function of the dioxanone : 2 mol ratio The end group distribution was determined using NMR and is expressed in terms of R gp (eq. reaction time and temperature on the X n exp of the products is shown in Figure 2.7.Chapter 2
2.3.) for all experiments are scattered randomly between 2. Most experimental points are scattered along these lines. and 2. The X n exp increases linearly with respect to the dioxanone : 2 ratio.5 hr Xn
40 Xn exp
30 40 50 dioxanone : 2 ratio (mol/mol)
Figure 2.3. and dioxanone/ 2 mol ratio) is absent.2.2. to reduce the number of OiPr endgroups in the product may be the removal of isopropyl alcohol formed in step
. time.4 and 3.
Temperature: 100oC.2. in line with the theoretical predictions.).e. time: 16 hr
Temperature: 60oC. 2. The R gp values (Table 2.5. A possible strategy to increase the Rgp values i.3. the X n theo (eq.7.
177 × (t ) − 0.g.). The R2 value for the model is 0. time and dioxanone/ 2 ratio) on the X n exp has been performed by multivariable linear regression  on the data given in Table 2.6) was too limited to draw sound conclusions.4.3. Statistical Data Analysis Quantification of the influence of the experimental factors (temperature. The analysis of variance for the model is given in Table 2.703 F 173.) from the reaction mixture by e. vacuum distillation before adding the dioxanone monomer .974).4.49 + 0.2. Table 2. The 2/ Al ratio was not included in the model as the experimental range (6.543
Model Error Total
P‐value 3.854 × (dioxanone : 2 ratio) + 0.3. 2. 2. indicating that the model describes the experimental data reasonably well.097 × (T )
. A parity plot of the modeled versus experimental values of the average chain length Xn confirms this statement (Figure 2.3‐6.7. The low P‐value clearly indicates that the model is statistically significant. A linear model proved adequate to describe the effects of the independent variables on the X n exp :
where t and T are respectively the polymerization time and temperature. Ring Opening Polymerization of p‐dioxanone using an Al(OiPr)3‐Monosaccharide Initiator System
1 of the polymerisation process (eq.8.3. Analysis of variance (ANOVA) for linear model of X n function of experimental parameters
SS 4103 93 4010 DF 3 12 15 MS 1337 7.977 (with an adjusted R2 value of 0.
the model predicts that the X n exp is positively influenced by the polymerization time. Both NMR and MALDI‐TOF measurements indicate that the products are mixtures of polymers and significant
. Molecular weights (NMR) of the product were between 970 and 6200 and are a clear function of the p‐dioxanone/2 ratio (at constant Al(OiPr)3 intake).4‐di‐O‐ isopropylidene‐α‐D‐galactopyranose. which is in agreement with the available data on ring opening polymerization [15. 11] and due to the fact that the reaction is an equilibrium polymerization with a slight exothermic effect. Conclusions
The ROP of p‐dioxanone in the presence of a monosaccharide (1.3. 2) with Al(OiPr)3 as the catalyst is reported.8. Within the experimental ranges.
2. the model allows determination of the process variables to obtain a product with the desired degree of polymerization. the X n exp is negatively influenced by temperature. 16. 17]. with higher ratios leading to higher molecular weights. This is in line with literature data [5. Modeled versus experimental values for the average chain length Xn The model predicts that the X n is a clear function of the p‐dioxanone : 2 mol ratio. In addition. Furthermore. with high ratios leading to a higher average chain length.2. The isolated yields of the off‐white solid products were between 30 and 96%.4.Chapter 2
40 Xn model
Dubois: ʺCoordination‐insertionʺ ring‐ opening polymerization of 1. R. Raquez. Degee. These studies will be reported in the next chapter. 153‐166. 1063‐1071. and mechanistic aspects of the ring‐opening polymerization of 1. P. Rapid Commun. 199.R. Macromol. . Raquez. Dubois: Some thermodynamic.4‐dioxan‐2‐one and controlled synthesis of diblock copolymers with epsilon‐caprolactone.C. Biomed. Narayan. 2000. Chu. 21. 1993.
. Narayan.M.4‐dioxan‐2‐ one initiated by Al((OPr)‐Pr‐i)(3) in bulk. Damrau: Polylactones. Phys. 27. Zn‐lactate‐catalyzed polymerizations of 1. Nomenclature
x − y ppm
: peak area of certain proton in 1H‐NMR spectra [‐] : peak area of certain peak at x until y ppm in 1H‐NMR spectra [‐] : mol ratio of galactopyranose end‐capped polydioxanone and isopropoxide‐end‐capped polydioxanone [‐] : number of monomer unit in the polymer products [‐] : time [hour] : temperature [oC] : experimental average degree of polymerization the polymer products [monomer units] : theoretical average degree of polymerization of the polymer products [monomer units]
n t T
X n exp X n theo
2.O.L. . Res. P. R. C. 1998. 42. H. Macromolecules 2001. Lin. Degee. 34. . J.4‐dioxan‐2‐one. temperature and monomer: monosaccharide ratio) on the X n exp .J. P. J. Kricheldorf.M.
2.6. P.g. kinetic. Grubb: Hydrolytic degradation and morphologic study of poly‐p‐dioxanone. starch). Chem. H. Mater. References
. A statistical model has been developed to quantify the effects of process variables (time. D. 8419‐8425. Macromol. 1089‐ 1097. D. Moreover. Ring Opening Polymerization of p‐dioxanone using an Al(OiPr)3‐Monosaccharide Initiator System
amounts of p‐dioxanone polymers with an isopropoxide end group (20‐30%) were present.5. the findings of this study have proven to be valuable input for synthetic studies on the preparation of novel biodegradable polymers consisting of polydioxanones and polycaprolactones grafted on oligo‐ and polysaccharides (e.
. H. Nishida, M. Yamashita, T. Endo, Y. Tokiwa: Equilibrium polymerization behavior of 1,4‐dioxan‐2‐one in bulk. Macromolecules 2000, 33, 6982‐6986. . T. Hamaide, M. Pantiru, H. Fessi, P. Boullanger: Ring‐opening polymerisation of epsilon‐caprolactone with monosaccharides as transfer agents. A novel route to functionalised nanoparticles. Macromol. Rapid Commun. 2001, 22, 659‐663. . K. Bernard, P. Degee, P. Dubois: Regioselective end‐functionalization of polylactide oligomers with D‐glucose and D‐galactose. Polym. Int. 2003, 52, 406‐411. . H.R. Kricheldorf: Syntheses and application of polylactides. Chemosphere 2001, 43, 49‐54. . H.R. Kricheldorf, M. Berl, N. Scharnagl: Poly(lactones). 9. Polymerization mechanism of metal alkoxide initiated polymerizations of lactide and various lactones. Macromolecules 1988, 21, 286‐293. . H.R. Kricheldorf, I. Kreisersaunders: Polylactones. 30. Vitamins, hormones and drugs as co‐Initiators of AlEt3‐initiated polymerizations of lactide. Polymer 1994, 35, 4175‐4180. . L.M. Esteves, L. Marquez, A.J. Muller: Optimization of the coordination‐ insertion ring‐opening polymerization of poly(p‐dioxanone) by programmed decrease in reaction temperatures. J. Appl. Polym. Sci. 2005, 97, 659‐665. . D.C. Montgomery: Design and Analysis of Experiments, 5th Edition, John Wiley & Sons Inc., New York, USA, 2001. . P. Dubois, R. Jerome, P. Teyssie: Aluminum alkoxides ‐ A family of versatile initiators for the ring‐opening polymerization of lactones and lactides. Makromol. Chem. Macromol. Symp. 1991, 42/43, 103‐116. . A. Duda, A. Kowalski, J. Libiszowski, S. Penczek: Thermodynamic and kinetic polymerizability of cyclic esters. Macromol. Symp. 2005, 224, 71‐84. . A. Kowalski, A. Duda, S. Penczek: Polymerization of L,L‐lactide initiated by aluminum isopropoxide trimer or tetramer. Macromolecules 1998, 31, 2114‐ 2122. . A. Duda, A. Kowalski, S. Penczek, H. Uyama, S. Kobayashi: Kinetics of the ring‐opening polymerization of 6‐, 7‐, 9‐, 12‐, 13‐, 16‐, and 17‐membered lactones. Comparison of chemical and enzymatic polymerizations. Macromolecules 2002, 35, 4266‐4270. . A. Duda: Polymerization of epsilon‐caprolactone initiated by aluminum isopropoxide carried out in the presence of alcohols and diols. Kinetics and mechanism. Macromolecules 1996, 29, 1399‐1406.
Ring Opening Polymerization of p‐dioxanone using an Al(OiPr)3‐Monosaccharide Initiator System
. S. Penczek, T. Biela, A. Duda: Living polymerization with reversible chain transfer and reversible deactivation: The case of cyclic esters. Macromol. Rapid Commun. 2000, 21, 941‐950.
ε‐ Caprolactone is grafted to silylated starch by a ring‐opening polymerisation catalysed by Al(OiPr)3 in THF at 50oC. polycaprolactone. The silylation reaction was performed using hexamethyl disilazane (HMDS) as the reagent in DMSO at 70°C. Experiments with native starch under similar conditions do not result in the desired poly‐(ε)‐caprolactone grafted corn starch co‐polymers and unreacted starch was recovered after work‐up. The DS of the polycaprolactone graft is between 0. Removal of the silyl groups of the poly‐(ε)‐ caprolactone grafted starch co‐polymers is possible using a mild acid treatment with diluted hydrochloric acid in THF at room temperature. The grafting efficiency varies between 28 and 58%.45‐0. The poly‐(ε)‐caprolactone side chains consist of 40‐55 monomer units and is a function of the reagent intakes. silylation
.7 was obtained. Keywords: starch. grafting. Silylated starch with a degree of substitution (DS) between 0. the remainder being homopolymers of ε‐caprolactone. biodegradable polymers.Chapter 3
Synthesis of Poly‐(ε)‐caprolactone Grafted Starch Co‐polymers by Ring Opening Polymerisation using Silylated Starch Precursors
Poly‐(ε)‐caprolactone grafted corn starch co‐polymers were synthesized using a hydrophobised silylated starch precursor.21 and 0.72.
1. Starch modification is therefore needed to meet the product properties in a number of application areas.g. polyvalerolactone.4‐ glucosidic bonds). the used monomers and the corresponding grafted chains are not easily biodegradable. as starch is highly hydrophilic. Starch has also been thermoplasticized with the help of plasticizers such as glycerol and other polyalcohols . Starch films are known to have good oxygen barrier properties. A wide variety of synthetic biodegradable polymers have been prepared. and polybutyrolactone). The content of amylose in starches depends on the plant and typically varies between 18‐28%. Well known examples are polyesters derived from cyclic lactones (polycaprolactone. roots) as granules or cells with typical particle sizes between 1‐100 µm. it is water sensitive. Various modification strategies have been explored. 5]. However. The polymeric structure of starch consists of repeating anhydroglucose units. Introduction
Worldwide. Global production of starch is 60 million ton per year in 2004 . 7]. for instance grafting of monomers (like styrene and methyl methacrylate) to the starch backbone [6. the future generation of packaging materials is highly desirable. amylose (a linear polymer of anhydroglucoses with α‐D‐1. The development of green biodegradable polymers for e. There are two types of biopolymer in starch. a natural biopolymer.6‐glucosidic bonds besides α‐D‐1. Polycaprolactone (PCL) is easily degraded by micro‐organisms . Starch is abundantly available. resulting
. which are known to degrade with difficulty and cause serious environmental problems . Aerobic soil‐burial experiments showed that the mechanical properties of PCL films decreased rapidly in time and were fully degraded after 4 weeks . is one of the potential candidates for future biodegradable polymer products. However.4‐glucosidic bonds) and amylopectin (a branched polymer with α‐D‐1. The amylose‐amylopectin ratio in native as well as modified starches has a strong impact on the product properties. the product properties are in most cases still not up to standards and blending with other polymers is required . and this value increases with about 10% per year . Starch. Starch is present in the body of many plants (tubers. PCL has gained much interest for possible applications in the medical field as well as in the area of packaging materials [12‐ 13] Several studies to combine the properties of starch and PCL have been performed to obtain fully biodegradable materials with improved product properties. However. 245 million tons of plastics are produced per year. in almost all cases. Blends of thermoplastic starch and PCL are not fully miscible. Chapter 3
3. These plastics are mainly synthetic polymers from fossil resources. and the mechanical properties of starch‐based films are generally inferior to those derived from synthetic polymers [4.
leading to a liquid‐solid system.2. Acros) and methanol (Labscan) were used as received.
. the ring opening polymerisation occurs solely in the liquid phase and this is expected to lead to higher GE values. A stock solution was prepared by dissolving 1. Hydrochloric acid (HCl) 1 N was prepared from Titrisol concentrated hydrochloric acid solution (Merck) and distilled (Milipore) water. DMSO (Acros) and toluene (Labscan) were dried overnight over molecular sieves 3 Å (Merck) and stored under a protective nitrogen atmosphere. a very flammable by‐product. Aluminium triisopropoxide [Al(OiPr)3] (Acros) was used without further purification. therefore difficult to handle and releases ethane. Materials and Methods
3. thus leading to the formation of PCL homopolymers rather that starch‐g‐PCL . 0‐14%). during the reaction. The starch source is made less hydrophilic and thus more soluble in organic solvents by substituting part of the OH groups of starch by a bulky silyl group [16‐18]. distilled under reduced pressure at 100 oC and stored under a protective nitrogen atmosphere. an alternative method to synthesize poly‐(ε)‐caprolactone grafted starch co‐polymers (starch‐g‐PCL) is reported. The highest GE (90%) was achieved using triethylaluminium as catalyst . This is expected to lead to reduced reaction rates between starch and CL compared to CL homopolymerisation. Common Ring Opening Polymerization (ROP) catalysts such as tin octoate or aluminium isopropoxide gave low grafting efficiencies (GE. All available data indicate that the presence of water reduces the GE. In this way. it has been proposed to chemically graft caprolactone onto the hydroxyl groups of starch using ring‐opening polymerisation . Dry tetrahydrofuran (THF) and toluene for polymerization experiments were obtained in closed vessels from Aldrich and were used as received. This is rationalised by assuming that water competes with the hydroxyl groups of starch in the initiation step of the polymerization reaction. To increase the miscibility of starch and polycaprolactone. Another possible cause for the low grafting efficiencies is the heterogenous nature of the reaction. Starch is insoluble in the typical organic solvents used for ROP (such as toluene or THF). In this chapter.1. This approach has also been applied successfully to graft PCL and polylactide on dextran [19‐20]. thus to a reduction in the GE. This catalyst is extremely air and water sensitive. Synthesis of Poly‐(ε)‐caprolactone Grafted Starch Co‐polymers using Silylated Starch Precursors
in undesirable phase separation . Hexamethyldisilazane (HMDS. Materials Corn starch (Sigma) was dried at high vacuum (~1 mbar) at 100 oC for one day before use.2.67 gram (8 mmol) Al(OiPr)3 in 50 ml of dry toluene in a glove box. ε‐Caprolactone monomer (Fluka) was dried over Calcium Hydride (CaH2) for 48 h.
THF (4.6 mL) at 50 oC (1‐2 h).7‐2. starch). after peracetylation.
Silylated Starch (1. For each experiment.5 ml) and a predetermined amount of the stock solution of Al(OiPr)3 in toluene were added. before peracetylation.12 (m. Chapter 3
3. DS = 0.
3. a predetermined amount of ε‐caprolactone monomer (molar ratio of monomer to
. The reaction was carried out at 70 oC.2. 40 oC) until constant weight. After 2 and 4 h reaction time.2.2. The pre‐determined amount of HMDS (typically 24 ml. This procedure was repeated three times. The white‐to‐transparent solid products were stored under vacuum in a desiccator at 6‐8 oC. silyl‐CH3). A mol ratio of ‐OH groups to catalyst of 10:1 was used. broad peaks. Silylated Starch (1. acetate‐CH3). 0. Traces of DMSO trapped in the product were removed by dissolving the product in a small amount of toluene and re‐precipitation in methanol. The mixture was stirred at 50 oC for 4 h to promote the exchange reaction between the isopropoxide groups of Al(OiPr)3 and the free ‐OH groups of starch.2. silyl‐CH3). To this homogenous solution. Typical example of in situ polymerization of ε‐CL with silylated starch The silylated product from the first step was dissolved in THF (0. the solvents were removed from the silylated starch product under reduced pressure (~ 20 mbar) at 70 oC. 3‐6 ppm (m. The samples were characterized by 1H‐NMR. another portion of toluene (20 ml) was added. After 24 h.12 (m. broad peaks. Sample SN‐3. After 6 h. Typical example of the starch silylation procedure The procedure for corn starch silylation was adapted from that published for dextran [19‐23].111 mol) was added to the gelatinized mixture to initiate the silylation reaction.1 mbar. Methods All reactions and manipulations with air‐sensitive materials were carried out under a protective nitrogen atmosphere either using standard Schlenk techniques or in a glove box. 3‐6 ppm (m. toluene (40 ml) was added to solubilize the precipitated (partially silylated) starch.2. the silylated starch (1) product was dried in a vacuum oven (~5 mbar. Subsequently. 50 oC): δ 0.5 (m. 80 oC).2. After removal of the solvents under reduced pressure (0. DS = 0. 1.1.60):
H‐NMR (CDCl3.2. starch). pre‐dried corn starch (6 g) and dry DMSO (75 ml) were stirred at 70oC for about 3 h until a clear solution was formed.60): H‐NMR (CDCl3. The intake of 1 depended on its DS and was adjusted to obtain a solution with 5 x 10‐5 mol free‐OH groups/ml of THF. 50 oC): δ 0. Sample SN‐3. 3.
Silylated Starch‐g‐PCL (2. The silyl group was removed by adding a slight excess (with respect to the number of the silyl functionalities) of 1N HCl to a solution of ‐starch‐silylated‐g‐PCL in THF (10 % w/v). 3. 1. α‐PCL). 3. and vacuum dried at 40 oC for 24 h. 3. 4. The peracetylation procedure applied in this study was adapted from the literature [24. The product was collected as a white solid and characterised by 1H‐NMR. iPr).37 (t. Starch‐g‐PCL (3. γ‐PCL). starch). 4.98 (t.5‐3. DSPCL=0.98 (t. 3.4 (m.2. 2.3‐4.2. broad peaks. The reaction was stopped by cooling down the mixture to room temperature and the addition of 2‐3 drops of 1 N HCl to deactivate the catalyst.54 (m.25 (t. 1 (0.31 (m. The samples were characterized by 1H‐NMR.) was >99 %. ε‐PCL). β and δ‐PCL). After stirring for 2 h. filtrated. ‐CH3 .88 ppm (m.34): H‐NMR (DMSO. and 5‐5.5.4. starch). ‐CH3. ‐CH.37 (t. 1. iPr). Sample SN1CL1. ε‐PCL). 1. Subsequently.68. Typically.21): H‐NMR (DMSO d‐6. γ‐ PCL). 4. 4.2. DSsilylation = 0. α‐PCL). ε’‐PCL).75.
3. broad peaks.25 (t.54 (m. De‐silylation of poly‐(ε)‐caprolactone grafted silylated starch co‐polymers Desilylation of the 2 was performed using a procedure described by Ydens et al  for desilylation of silylated dextran‐g‐PCL. The silylated starch‐g‐PCL (2) product was precipitated from the solution by the addition of heptane (about 250 ml) at ‐18 o C. 60 oC): δ 0. 1. The yield was measured gravimetrically and is based on the weight of the product and the total weight of reactants charged to the reactor. The solid precipitate was filtered and dried under vacuum (~ 5 mbar) at 40 oC for 48 h. DSPCL=0. DSsilylation = 0. β and δ‐ PCL). ε’‐PCL).31 (m.16 (d. 3. Synthesis of Poly‐(ε)‐caprolactone Grafted Starch Co‐polymers using Silylated Starch Precursors
OH‐groups was 13 : 1 for a standard experiment) was added and the ring opening polymerization reaction was allowed to proceed for 24 h at 50 oC. iPr).4 (m. ‐CH.88 ppm (m. 1.16 (d. the desilylated starch‐g‐polycaprolactone product (3) was precipitated using heptane.12 (s.2. Deprotection product of Sample SN1CL18.104.22.168 g) was suspended in THF (4%‐w/v) and stirred at 55 oC until it was fully dissolved (typically 3 h). 3. and 5‐5. the peracetylating reagents
. silyl‐CH3). The total isolated yield at this condition (see Table 3. Peracetylation of silylated starch Characterisation of the silylated starch by NMR proved very difficult due to the presence of very broad and overlapping resonances arising from starch. 60 oC): δ 1.2.
3.5‐ 3.68. 2.75.3‐4. Peracetylation of the remaining OH groups of modified starch is a well established procedure to improve characterisation of the products by NMR . iPr). 25].
The degree of substitution of the PCL graft on 2 (DSPCL) is defined as the average number of PCL polymer chains present on an AHG unit of starch. it is assumed that the average length of the grafted chain is equal to the chain length of the homopolymer.2. DSsilylation may be calculated using 1H‐NMR spectra of the products after peracetylation using eq. 3.3.
3. The ACL is calculated from 1H‐NMR spectra by comparing the peak area of protons attached to ε‐carbon atoms in a repeating CL unit with that of the characteristic ε’ protons of the last CL unit in a PCL chain  (see Figure 3.).3.8 ppm
where Ax‐y stands for the 1H‐NMR peak area in the range δ x‐y ppm. 3.8− 4.
ACH 3 silyl / 9 1 7 A−0.3. When assuming that all ε‐CL monomer is converted.2. The peracetylation reaction was conducted for 7 h at 50 oC.3. acetic anhydride and pyridine in a DMAP : acetic anhydride : pyridine : AHG molar ratio of 1: 10: 22 : 1) were added. The assumption of high conversions (>95%) was correct for all experiments (see Table 3.2.1. Calculations The DS of the silylated starch (DSsilylation) is defined as the average number of silyl groups present on an anhydroglucose unit (AHG) of starch. The Average Chain Length (ACL) of the Poly‐(ε)‐caprolactone chain is defined as the average number of CL repeating units in a grafted polymer chain.6 ppm DSsilylation = × = × 3 Astarch proton / 7 27 A3−5. In this calculation. This leads to the following equation:
ACH 2 −ε + ACH 2 −ε ' ACH 2 −ε '
A3. It was finally dried overnight in a vacuum oven at 70 oC and 5 mbar until constant weight.4 ppm
(3.2. The product was precipitated by the addition of methanol and washed several times with methanol.6−0.2. Nuclear Magnetic Resonance (NMR) H‐ NMR spectra were recorded in CDCl3 at 50oC or in DMSO d‐6 at 60 oC on a Varian AMX 400 NMR machine. Analytical Methods 3.1.2 ppm A3.1.)
.3−3. the DSPCL may be calculated using eq. Chapter 3
8− 4.3−3. Synthesis of Poly‐(ε)‐caprolactone Grafted Starch Co‐polymers using Silylated Starch Precursors
DS PCL =
0.8− 4 .9 ⎟ × 100 % ⎜ ⎜ ⎟ ACH 2 −ε ' / 2 ⎠ A3.5. followed by an in‐situ Ring Opening Polymerization (ROP) of ε‐ caprolactone monomer on the hydrophobized starch and subsequent silyl group removal by a mild acid treatment. Although all steps have been investigated.8− 4 .4 + A3.CL/OH ratio ⎥ ⎣ ⎦
⎛ ⎞ ⎞ ⎛ A A = ⎜1 − CH −iso ⎟ × 100 % = ⎜1 − 4. It is calculated by comparing the area of protons related to the PCL grafted to starch with the area of the protons of all PCL chains present in the product.9
⎡ 0.4 / 2 ⎟ ⎠ ⎝ ⎝
where Ax‐y stands for the peak area in the range δ x‐y ppm.3−3. The Hildebrand solubility parameter of HMDS and DMSO were calculated using the following equation :
⎛ ∆H v − RT ⎞ ⎟ δ =⎜ ⎟ ⎜ Vm ⎠ ⎝
where ∆Hv stands for heat of vaporization. Results and Discussions
The overall procedure to synthesize poly‐(ε)‐caprolactone grafted starch co‐ polymers (3) consists of three steps and includes hydrophobization of starch by silylation of part of the hydroxyl groups of starch using hexamethyl disilazane (HMDS).4 − A4 .
. This leads to the following equation:
ACH 2 −ε ' / 2 − ACH −iso ACH 2 −ε ' / 2
× 100 %
The grafting efficiency (GE) is defined as the percentage of PCL grafted to starch compared to the total amount of polymerized CL (grafted and PCL homopolymer).( A3. T stands for absolute temperature.2 ) ⎤ 1 × ⎢ (3 − DSsilylation )⎥ ⎢ ε .4. the focus of this chapter will be on the first two steps of the procedure. The values of ∆Hv and Vm were obtained from the SciFinder Scholar database (American Chemical Society.3.3.5. A3. and Vm stands for molar volume.
1.1.1. 24 h). 70 oC. Synthesis of Silylated Starch The silylation of corn starch was performed with HMDS as the silylating agent (eq. In this way. Using standard conditions (HMDS: AHG molar ratio of 3.
6 H2C OH 4H 3 H5 OH H 6 H2C O R 4H H5 OR 3 H
CH3 CH3 CH3
Si NH Si
DMSO 50 deg. However. An additional peracetylation procedure to substitute the free hydroxyl groups with acetate groups was performed to improve the quality of the NMR spectra.5. a homogeneous reaction mixture was maintained throughout the reaction.
. as suggested by Einfeldt et al .)
Instead of using DMSO as solvent. The 1H‐ NMR spectrum of silylated starch (Figure 3. The 1H‐ NMR peaks from the AHG unit after peracetylation were indeed considerably sharper and allowed more accurate DS calculations (Figure 3. It is defined as the average number of silyl substituted OH groups on the anhydroglucose (AHG) unit of starch. Very broad peaks of starch protons at δ 3‐5 ppm and a sharp peak of the methyl substituents of the silyl group at about δ 0 ppm were observed (see Figure 3.) shows the presence of silyl groups at about δ 0 ppm.1. The silylated products were characterised by NMR. the starch peaks are very broad and this feature hampers accurate determination of the DSsilylation. mixtures of toluene and DMSO were applied to avoid precipitation of the silylated starch during the reaction.5.a.60 was obtained. C
O H 2OR
[R = H or Si(CH3)3] Starch Hexamethyldisilazane Silylated Starch (1)
(3.b.). The degree of substitution of the silyl groups (DSsilylation) was determined by 1H‐ NMR. Chapter 3
3.). a product with a DS of 0.a.). 3.3. The silylation procedure was adapted from that previously reported for dextran [19‐23].
.60. The results are given in Table 3.5) on the product DS was investigated by performing experiments with a constant starch and a variable HMDS intake. Synthesis of Poly‐(ε)‐caprolactone Grafted Starch Co‐polymers using Silylated Starch Precursors
Figure 3. not peracetylated (b).5‐4.2. 50 oC): (a). peracetylated The effect of the HMDS to starch molar ratio (1. DSsilylation=0. and Figure 3. Typical 1H‐NMR Spectra of Silylated Starch (Sample SN3. in CDCl3.1.
0.25 3 4.4
1. the DSsilylation decreases for higher intakes of HMDS. mmol) 12 (56) 18 (83) 24 (111) 36 (167) HMDS: AHG mol ratio 1. For all experiments.5 Product DS (DSsilylation) 0.5 2. a positive effect of higher reagent intake on the reaction rate and thus the DS is expected.1. The silylation reaction
. a fixed starch intake of 6 g (37 mmol AHG) starch was applied. For reactions with an order higher than zero.65
0.68 0.2. Effect of HMDS: AHG mol ratio on DS of the silylation product a
HMDS intake (ml.5
Figure 3. the volumetric HMDS intake is about half of the DMSO intake. At a mol ratio of HMDS to AHG of 4.5:1.67 0.5 3 3.60 0.36 cal1/2cm‐3/2) considerably.55
0. solubility parameter = 11.
Experiment SN1 SN2 SN3 SN4
Experiments were performed in DMSO at 70°C. Its presence will reduce the polarity of the reaction medium (DMSO. The experimentally observed lowering at higher HMDS intakes is likely due to a decrease in the polarity of the reaction mixture.6 DSsilylation
0. DS of the silylated products at different HMDS: AHG ratios (DMSO. 70 oC) Table 3.5 HMDS : AHG Ratio [mol/ mol]
4. HMDS is a rather apolar compound (Hildebrand solubility parameter of 6.5
0.25 cal1/2cm‐3/2) due to the presence of the bulky apolar SiMe3 groups. Chapter 3
2.3. The DSsilylation (0.45‐0. Petzold et al  reported that silylation of starch with trimethylsilylchloride (TMSCl) in pyridine yielded trimethyl‐silyl substituted starch with DS values between 0.0 . and DMSO/pyridine mixtures yielded silylated starch with DS values ranging between 0. DMA/LiCl.
O O H2C O R H H OR H O O H OR H O Catalyst: Al(OiPr)3 Toluene. 50 oC H CH2 H OR H C (CH2)5 O
O H OR
(R = H or Si(CH3)3 Silylated Starch (1) ε-Caprolactone Monomer
[R = H or Si(CH3)3 ] Silylated Starch-g-Polycaprolactone (2)
.6‐glucosidic for dextran).7) of the products is in the range of those published for other starch silylation systems. DMF. 29]. In situ Ring Opening Polymerization of ε‐Caprolactone with Silylated Starch A number of in situ ROP experiments with ε‐caprolactone (CL) were carried out using a typical silylated starch sample (DS=0.3‐2. After precipitation with heptane and vacuum drying.1 and 3.68.6. 3.6. liquid ammonia. The use of HMDS to silylate dextran (Mw=6000‐40000) in DMSO (HMDS to OH molar ratio of 0.0) gave silylated dextran with DS values between 1. To the best of our knowledge.25‐5. A schematic representation of the reaction is provided in eq. sample SN‐1) in THF at 50 oC for 24 h using Al(OiPr)3 as catalyst. pyridine. the silylation of starch in pure DMSO using HMDS as the reagent has not been reported to date.7‐3. white solid products with isolated yield > 96% were obtained. The products are soluble in DMSO as well as in less‐polar solvent such as chloroform and THF.0 [18. Synthesis of Poly‐(ε)‐caprolactone Grafted Starch Co‐polymers using Silylated Starch Precursors
involves charged‐ionic species  and a reduction in the polarity of the reaction medium is expected to lead to a lowering of the silylation reaction rates.2. Silylation of starch with HMDS in formamide. Similar reductions in the reaction rates when working at higher reactant intakes were observed for the esterification of starch with vinyl laurate and stearate . The much higher DS values obtained for dextran compared to starch may be related to differences in molecular weights between starch and dextran and the type of AHG linkages (mainly α‐1.
as is shown in Figure 3. DSPCL=0.9 and 1. Chapter 3
The products were characterized using 1H‐NMR analysis in DMSO‐d‐6 as the solvent. However. broad peaks in the region δ 3.
Figure 3. Coding of the peaks is given in Figure 3. The homopolymers are formed by direct polymerization of ε‐CL initiated on isoproproxide moieties attached to the Al catalyst.2‐4 ppm and imply that the ring‐opening polymerisation reaction of CL indeed occurred. Further process optimization studies allowed the synthesis of products with less than 42% of homopolymers (vide infra). 26].2 ppm.0‐5. Resonances from the starch peaks are observable as small.
. The formation of CL homopolymers for this type of reactions has been observed before [15. Typical 1H‐NMR spectrum of a silylated starch‐g‐PCL sample. (Sample SN1CL1.3. This is clearly indicated by resonances of the iPr end‐group of the PCL homopolymer at δ 4.4.21) in DMSO‐d6 at 60 oC.4 ppm.3. A typical spectrum is shown in Figure 3. the exchange reaction between Al(OiPr)3 and the OH groups of silylated starch is not quantitative under the conditions applied in this study. Apparently. not all of the caprolactone is grafted to starch and large amounts of PCL homopolymers (72%) were formed.4‐3. The peaks from the polycaprolactone units are clearly present in the range of δ 1.8 and 5.4.68. DSsilylation =0.
This leads to the formation of carboxylic end groups (see eq.
1. Synthesis of Poly‐(ε)‐caprolactone Grafted Starch Co‐polymers using Silylated Starch Precursors
. However.7.). Polymerization Reaction: O iPr O Al OH iPr O PCL homopolymer with carboxylic end-group + O O α H β χ δ ε O α' β' χ' δ' ε' [O C CH2 CH2 CH2 CH2 CH2 ]n O C CH2 CH2 CH2 CH2 CH2 OH
(3. residual water may also initiate the polymerisation reaction. 3. peaks arising from a ‐COOH unit could not be detected in 13C‐ NMR spectra (δ 175‐180 ppm). Exchange Reaction:
iPr O Al O iPr H2C OH
iPr O Al O iPr iPr O
H2C O O O H O H OR H H OR [R = H or Si(CH3)3 ] H O
H OH H
+ iPr OH
Silylated Starch (1) 2. Product formation for the reaction between silylated starch and ε‐ caprolactone Besides the ‐OH group of starch. Polymerization Reaction: iPr O Al O iPr H2C O O H O H OR H H OR H O O
O O α β χ δ ε α' β' χ' δ' ε' H2C [O C CH2 CH2 CH2 CH2 CH2 ]n O C CH2 CH2 CH2 CH2 CH2 OH O O H O H OR H H OR H O [R = H or Si(CH3)3 ]
Silylated Starch-g-PCL (2) O
iPr O Al O iPr iPr O
O n+1 iPr
[O C CH2 CH2 CH2 CH2 CH2 ]n O C CH2 CH2 CH2 CH2 CH2 OH
Figure 3. Exchange Reaction: iPr O Al O iPr iPr O + H2O iPr O iPr O Al OH + iPr OH
The ratio of homopolymerisation versus grafting on starch may be obtained by comparing the integrals of selected peaks in 1H‐NMR spectra.68) in THF at 50 oC with Al(OiPr)3 as the catalyst (1 mol Al(OiPr)3 per 10 mol‐starch‐OH groups). This implies that the ε‐CL conversion is essentially quantitative in all cases.5 29. The results are given in Table 3. The yield of the products was measured gravimetrically and varies between 96. units] 40 43 44 49 54 DSPCL 0.5 Avg.2.2 Total Yieldb [%] >99 >99 >99 99 96.34 0.0 18. The grafting efficiency (GE) for the sample given in Figure 3. and Figure 3. and reaches 58% for a ε‐CL to starch–OH ratio of 29.3.5.
b. (SN1CL1) is 28%. the intensity of the peak from the ‐CH2‐ end group of the homopolymer (ε’ at δ 3. Five experiments were performed to study the effect of different ε‐CL to silylated starch ratio.9 22.
. Determined gravimetrically and defined as the total weight of the isolated product divided by the total intake of reactants (silylated starch and CL).2. The ACL of the polymer and the DSPCL increase almost linearly with the ε‐ caprolactone intake (Figure 3. In the case of only homopolymerisation. This implies that grafting of caprolactone to starch also occurs to a significant extent.47 0.5. Table 3.9 ppm). However.3 ppm) should be twice the intensity of the ‐CH‐ proton of the isopropoxide end group (δ 4. This indicates that higher ε‐CL concentrations during the reaction lead to longer PCL grafts as well as to higher levels of initiation of the grafting reaction on free hydroxyl group of silylated starch.72 Grafting Efficiency [%] 28 43 48 55 58
All reactions were performed using the same intake of SN1 silylated starch (DS=0.5 and 100%. the intensity of the resonance ε’ was considerably higher that that of the ‐CH‐ iPr peak.).
Overview of results for the grafting of ε‐CL on silylated starch a
ε‐CL/ OH [mol/ mol] 13.2. in all samples.21 0. Chain Length [mon. The GE increases with higher ε‐CL intakes.0 15.58 0.
Experiment SN1CL1 SN1CL2 SN1CL3 SN1CL4 SN1CL5
Synthesis of Poly‐(ε)‐caprolactone Grafted Starch Co‐polymers using Silylated Starch Precursors
Average Polymer Length [mon. units]
50 0.7 45 0.6 0.5 40 0.4 35 0.3 0.2 30 12 14 16 18 20 22 24 26 28 30 DSPCL
ε-CL : Starch-OH Ratio [mol/ mol]
Figure 3.5. Average Chain Length (ACL) and the DS for as a function of the CL‐ starch intake. ∆ : Average Chain Length (ACL); □ : DSPCL The mechanism of ROP of cyclic esters such as caprolactone in the presence of an alcohol (ROH, silylated starch in our case) is provided in Scheme 3.1. [30‐31]. Higher monomer intakes are expected to lead to higher polymerization rates as shown in Scheme 3.1.b. This will result in longer PCL chains in the final product, in line with the experimental observations (see Table 3.2. and Figure 3.5.). At higher ε‐caprolactone intakes not only the ACL of the grafted chain increases but also higher values for the DSPCL are observed. This finding may be rationalised by assuming that the rate of chain transfer (Scheme 3.1.c.) with starch is increased at higher ε‐caprolactone intakes. The rate of this reaction is expected to be a function of both the starch and the concentrations of Al‐species with a growing PCL chain. The starch intake for all experiments was equal, meaning that the concentration of Al‐species with a growing PCL chain should be higher at higher ε‐CL intakes. This is indeed predicted by the mechanism given in Scheme 3.1.b.; higher caprolactone intakes will increase the rate of this reaction and lead to higher concentration of Al‐species with a growing PCL chain.
Al O iPr
R OH O O kp kd
Al O R
O Al O [(CH2)5 C O]n R
Al O R
O Al O [(CH2)5 C O]n R + R OH
O Al O R
+ HO [(CH2)5 C O]n R
Scheme 3.1. The observation that a higher monomer to alcohol ratio leads to higher amounts of PCL chains with an alcohol end group and thus a higher DS was also reported for the polymerization of p‐dioxanone with Zn‐lactate as catalyst and α‐tocopherol as the alcohol . To show the potential of our approach to use hydrophobised starch instead of native starch for the ring opening polymerisation of cyclic esters, several ring openings polymerisations of native starch with ε‐CL monomer were performed either in pure ε‐CL or in a mixture of ε‐CL and toluene (80‐100oC, 24 hr). At the start of the reaction, the starch was always insoluble in the reaction medium. After reaction the product was isolated, washed thoroughly with toluene and dried. The weight of the product, however, was very close to the initial starch intake. Examination of the products by FT‐IR does not show the presence of caprolactone vibrations. Thus, it may be concluded that solubilisation of starch is of key importance to obtain poly‐(ε)‐caprolactone grafted starch co‐polymers. These findings are in line with earlier studies on the ROP of ε‐caprolactone on native granular starch using Al(OiPr)3 as catalyst . Here, caprolactone polymerization did not occur after 18‐24 h reaction time and only liquid ε‐CL was recovered. Only when performing the reaction with high concentrations of the aluminium catalyst, a product with a GE of about 13% was obtained. This low GE was explained by assuming that the reaction between starch and Al(OiPr)3 is slow and due to the heterogeneous nature of the reaction mixture. Our study, together with the result of Dubois et al  indicate that homogenous reaction conditions are required for the successful ROP of ε‐CL to obtain poly‐(ε)‐caprolactone grafted starch co‐polymers when using Al(OiPr)3 as the catalyst. When performing the reaction under heterogenous conditions, a high grafting efficiency is only achievable when using triethylaluminium as the catalyst .
Synthesis of Poly‐(ε)‐caprolactone Grafted Starch Co‐polymers using Silylated Starch Precursors
3.3.3. Deprotection of Silylated‐Starch‐g‐PCL A preliminary experiment was performed to remove the silylated groups of the silylated starch‐g‐PCL (eq. 3.8.) using a mild acid treatment with diluted hydrochloric acid in THF at room temperature. All silyl groups were removed successfully, as is clearly seen from an NMR spectrum given in Figure 3.6.
Figure 3.6. H‐ NMR Spectrum of Starch‐g‐PCL (reaction product after desilylation of sample SN1CL2, DSsilylation =0.68, DSPCL=0.34), in DMSO‐d6
O O H2C H H OR H C (CH2)5 O
O H O CH2 H H OH H C (CH2)5 O
O H OR
THF, 2 hr H
O H OH
[R = H or Si(CH3)3 ]
Silylated Starch-g-Polycaprolactone (2)
1. The products may have interesting applications as compatibilizers for starch‐polymer blends. depending on the ε‐CL‐starch ratio.5.46‐0.21‐0. Considerable amounts of ε‐CL hompolymers with isopropyl end‐groups were also formed. Silylated starch with a low‐medium DS (0. The ROP with ε‐CL was performed using Al(OiPr)3 as catalyst in THF as the solvent.
3.68) was obtained using a DMSO/toluene mixture as the solvent and HMDS as the silylating agent. The grafting efficiency varied between 28‐58%.4. average value of mole of substituted –OH per mole of anhydroglucose (AHG) units [‐]
x − y ppm
DSsilylation : DS of silyl group substituents [‐]
DS of PCL chain substituents [‐]
: Grafting Efficiency.986 cal mol‐1 K‐1 : temperature [K]
molar volume [cm3/ mol]
The successful synthesis of poly‐(ε)‐caprolactone grafted corn starch co‐ polymers using a three step approach is reported.72. indicating that homogeneous reaction conditions are favorable for the grafting reaction.2. This was achieved by making the starch more hydrophobic by partial substitution of the OH groups by trimethylsilyl groups. Poly‐(ε)‐caprolactone grafted silylated starch co‐polymers with average chain length of 40‐55 monomer units (polymer molecular weight of 4500‐6300) were obtained. Nomenclature
: peak area of certain proton in 1H‐NMR spectra [‐] : peak area of certain peak at x until y ppm in 1H‐NMR spectra [‐] : average number of CL repeating units in a grafted polymer chain or homopolymer [monomer units] : Degree of Substitution. the highest value was obtained with a ε‐CL‐AHG ratio of 29. Control ROP experiments of ε‐CL with native starch under similar conditions did not produce the desired poly‐(ε)‐caprolactone grafted corn starch co‐polymers. Chapter 3
3. The DS of the PCL chains was between 0. The key feature is the use of a homogeneous reaction mixture for the ROP of starch with ε‐CL. the percentage of PCL grafted to starch compared to the total amount (grafted and homopolymer) of polymerized CL [%] : gas constant.
K. . 407‐412.. Burr. C. R. 2000. Mater. Synthesis of Poly‐(ε)‐caprolactone Grafted Starch Co‐polymers using Silylated Starch Precursors
: heat of vaporization [kJ/mol] :
Hildebrand solubility parameter [cal1/2cm‐3/2]
3. Food Technol. . Aarhus. .starch. 311‐316. 61‐67. Association of Plastics Manufacturers.
. Biobased packaging materials for the food industry. Beliakova. References
. Russell: Graft copolymers of polysaccharides with thermoplastic polymers ‐ new type of filled plastic. the Royal Veterinary and Agricultural University. . Starch‐Starke 2004. M.A. Brussels.M. 56. Adv. Rustgi: Biodegradable polymers. 51.6. 43. Wang. Karlsson. Goldberg: Review of the biodegradability and utility of poly(caprolactone). 1273‐1335. Albertsson: Biodegradable polymers and environmental interaction.M. Yang. 61‐74. L. 23. A. J. Abdel‐Mohdy: Grafting of poly(methacrylic acid) on starch and poly(vinyl alcohol). . Aly. of
. 107‐113. Bagley. Chandra. Part C: Polym. Website of International Starch Institute. . 1997. J. D. R. Weber). Tsiapouris.C. F. Sci. Environmen. Polym. C. Doane. 53‐57. Sci. 305‐313. Belgium. 3. report of Food Biopack Project (Ed. 11.C. Y. Polym.R.A. G. Environ. Prog. Krochta. 38. Solaro: Biodegradable polymeric materials. 2003. Denmark: http://www. Frederiksberg C. 17. E. Starch‐Starke 2000. 385–409 . 1996. S. E. 1977. 2008. Eng. Degrad. accessed on August 01. Eng.F.B. R. 1998. K. Babac: Preparation and biodegradation starch/polycaprolactone films. Rev. Chiellini. . 1995. J. X.A. Polym. Sci. .Sci. Denmark. 1998. Polym. W. A. 52. C. J.dk/isi/stat/rawmaterial. Plastics Europe. A. Yavuz. 2008. H. Fanta. status and perspectives. 1251‐1253. R. . De Mulder‐Johnston: Edible and biodegradable polymer films: Challenges and opportunities. C. 2003. Linke: Water vapor sorption determination of starch based porous packaging materials. demand and recovery for 2006 in Europe. . J. The compelling facts about plastics: An analysis of plastics production.. Macromol. Polym. Wang: Properties of starch blends with biodegradable polymers.html. Department of Dairy and Food Science. 8.
Hamaide. 1. . Degee. D. Petzold. K.L. 5. P. Fessi. Des. D. Klemm. 40. P. Stab. Fringant: Properties of thermoplastics blends: starch‐polycaprolactone. Petzold. R. C. structure analysis. Degee. Six: Partial or total silylation of dextran with hexamethyldisilazane. W. Chem. . Ydens. E. Klemm: Preparative and 1H NMR investigation on regioselective silylation of starch dissolved in Dimethyl Sulfoxide. W. Dellacherie. and chemical stability of silylated dextrans. Sci. Degee. P. D. W. 1735‐ 1743. . Dubois. Kussler. Macromolecules 2000. . P. Degrad. 33. Dellacherie. Nouvel. . Macromol. A. D. 43. Boullanger: Ring‐opening polymerisation of epsilon‐caprolactone with monosaccharides as transfer
. Stein. Rutot. Einfeldt. J. A. 4157‐4167. M. 6713‐ 6721. I. Heublein: Silylation of cellulose and starch ‐ selectivity. Dubois. J. Biosci.L. Gunther. Petzold. 561‐566. L. Rutot. Biomacromolecules 2001. Symp. E. A. 2001. 2. A. 41. Nouvel. Stein. 415‐426. Ydens. Narayan: Aliphatic polyester‐grafted starch‐like polysaccharides by ring‐opening polymerization. Part A: Polym. 73. Polymer 2002. Six. D. L. Einfeldt. Gunther: Synthesis and NMR characterization of regiocontrolled starch alkyl ethers. Macromol. M. L. Polym. J. J. D. B. P. . E. Klemm: Regioselective functionalization of starch: Synthesis and 1H NMR characterization of 6‐O‐ silyl ethers. P. 341‐347. 2001. and subsequent reactions. Klemm. K. . I. P. T. C. Biomacromolecules 2003. L. M. Six: Silylation reaction of dextran: Effect of experimental conditions on silylation yield. Chapter 3
. 2004. C. D. 4. Ydens. Dellacherie. Averous. Polymer 1999. Monomers Polym. 2001. P. E. 2577‐2588. 35‐47. Dellacherie. Dubois. . 3091‐3100..L. 2002. Six: Controlled synthesis of amphiphilic biodegradable polylactide‐grafted dextran copolymers. Gunther. P. . 163. Moro. K. I. Cellulose 2003. 1443‐1450. K. high value applications. Einfeldt: Structure design of polysaccharides: novel concepts. 965‐969. 10. 251‐269. Duquesne. Stein. regioselectivity. . Polymer 2000. 42. Pantiru. E.L. . P. Klemm. Petzold. D. Nouvel. Dubois: Aliphatic polyester‐ based biodegradable materials: new amphiphilic graft copolymers. L. Polym. Dubois: Controlled synthesis of poly(epsilon‐caprolactone)‐grafted dextran copolymers as potential environmentally friendly surfactants. Krishnan. selective syntheses. Dole. . P. H. Koschella. J. Dubois.
D. Phys. Janssen. 1089‐1097. 659‐663. accepted for publication in Starch‐Starke. A.M. Lehmann: Selective 6‐O‐acetylation of amylose. A.4‐ dioxan‐2‐one. Duda: Living polymerization with reversible chain transfer and reversible deactivation: the case of cyclic esters. S. Junistia. 2008. T. Res.L. Biela. 1998. Horton.: Solubility parameter concepts – a new look. D. 941‐950. H. 1399‐1406.J. Macromolecul. Macromolecules 1996. J. Mexico City. D. Penczek. 199. Picchioni. L. . 553‐556.R. 2000. . Macromol Chem. Carbohydr. R. Duda: Polymerization of ε‐caprolactone initiated by aluminum isopropoxide carried out in the presence of alcohols and diols. Kricheldorf.B. L. . Hertz Jr. Rapid Commun. A novel route to functionalised nanoparticles. . Sugih. Damrau: Zn‐lactate catalyzed polymerizations of 1. Kinetics and mechanism.K.
. F.O. 21.P. 22. 1989 . Heeres: Synthesis of higher fatty acid starch esters using vinyl laurate and stearate as reactants. Rapid Commun. 1978. Manurung. 2001. presented at ACS Rubber Division meeting. A. . Macromol. Synthesis of Poly‐(ε)‐caprolactone Grafted Starch Co‐polymers using Silylated Starch Precursors
agents. 61. H. 29.
the DS for the starch‐laurate compounds is higher than for the corresponding starch‐ stearates. the DS decreases.96). an increase in the vinyl‐ester concentration leads to higher product DS values. presumably due to a reduction of the polarity of the reaction medium. Keywords: corn starch. The fatty esters were prepared by reacting the starch with vinyl laurate or vinyl stearate in the presence of basic catalysts (Na2HPO4. and Na‐acetate are superior catalysts with respect to activity compared to Na2HPO4 and products with DS values close to 3 were obtained. For low vinyl‐ester to starch ratios. 13C‐NMR and FT‐IR. vinyl stearate
. K2CO3. The DS of the products is a function of the carbon number of the fatty acid chain. vinyl ester to starch ratio and the type of catalyst. When performing the reactions using Na2HPO4 as the catalyst. and Na‐acetate) in DMSO at 110°C.Chapter 4
Synthesis of Higher Fatty Acid Starch Esters using Vinyl Laurate and Stearate as Reactants
This chapter describes the synthesis of long fatty esters of corn starch (starch‐ laurate and starch‐stearate) with a broad range in degree of substitution (DS= 0.24 ‐ 2. vinyl laurate. esterification. K2CO3. At higher ratios. The yellowish products were characterized by 1 H‐.
However. For the current petrochemical based products recycling is often neither practical nor economically feasible . the fatty ester substituents [1. 5‐6] were introduced using fatty acid chloride reagents. resulting in products with DS values up to 2. which are relatively expensive and rather corrosive . more water‐resistant products may be obtained [3‐4]. Further modification of starch is therefore required to introduce hydrophobicity and to improve mechanical and moisture barrier properties.1. 7]. However.Chapter 4
4. 5]. defined as the moles of substituents per mole of anhydroglucose (AHG) units) . poor processibility (high viscosity).61) products were obtained using this approach. Natural polymers such as starch. However. even after the addition of plasticizers . An attractive field of application for these polymers is the use as packaging materials. the use of native starch for packaging materials is limited due to its low moisture resistance. The first paper on the acetylation of starch was already published in 1865 . and even fatty acid derivatives (C12‐C18) have been used in the modification reaction [5. Their
. and incompatibility with hydrophobic polymers.7 [1.34‐0. higher molecular weight carboxylic starch esters (C4‐C6) . high brittleness. The major obstacle is the pronounced brittleness of the materials. most of the studies performed to date use short chain carboxylic acids (C1‐C4). and particularly acetic acid derivatives (C2) [2‐4]. In particular starch is attractive as it is relatively cheap and abundantly available. To improve the mechanical properties. the mechanical properties of high‐ DS starch derivatives of low chain carboxylic acids still need considerable improvements before large scale application as packaging materials becomes within reach. An alternative method using methyl and glyceryl laurate esters in the absence of solvent has been recently developed . However. Esterification of starch with low molecular weight fatty acid derivatives is one of the oldest modification technologies to improve starch properties. Recently Mormann et al  explored the possibility of using vinyl esters and particularly vinylacetate as reagents for the preparation of starch esters. Introduction
Green biodegradable polymers derived from natural resources are potentially very interesting substitutes for non‐biodegradable petroleum‐based polymers. The mechanical properties and hydrophobicity of the products were significantly improved when using these longer chain fatty acid precursors [1. The hydrophobicity increases with the degree of acetate substitution (DS. and consequently. 5]. cellulose or proteins are potentially very interesting starting materials for biodegradable packaging materials. Relatively low‐DS (0. The introduction of acetate groups on starch makes the product more hydrophobic.
not been reported to date. the Netherlands). Analytical Equipment H‐ and 13C‐NMR spectra were recorded in CDCl3 on a 400 MHz Varian AMX NMR machine. 4‐N. Materials and Methods
4. The starch was dried before use for 48 h at 105 oC under vacuum (approx. We here report an investigation on the synthesis of higher fatty acid esters of starch with an emphasis on the introduction of laurate and stearate ester side chains. The effects of the starch to vinylester ratio on the reaction rates and DS have been explored.
. starch esters with a substantially higher DS value (up to 1. IR spectra were recorded on a Spectrum 2000 FT‐IR Spectrometer (Perkin Elmer). The synthesis of starch stearate esters using vinyl ester reagents has. and tetrahydrofuran (THF) were supplied by Acros (Belgium) and were also used as received. The effect of the addition of a non‐polar solvent (toluene) to the reaction medium to solubilise the products and thus to enhance the reaction rates has also been studied. the use of basic catalysts other than Na2HPO4 has been investigated. ~1 mbar). Synthesis of Higher Fatty Acid Starch Esters using Vinyl Laurate and Stearate as Reactants
research focused on the synthesis of starch acetates and only two examples with a higher molecular weight fatty acid vinyl ester were reported.2. 4.2. Technical grade dimethyl sulfoxide (DMSO). The spectra were recorded at 50 oC. Germany) and disodium hydrogenphosphate (Merck. Japan). 73% amylopectine and 27% amylose) was purchased from Sigma (Germany). leading to a moisture content of 2 %‐wt (measured gravimetrically).2. Germany) and acetic anhydride (Merck. sodium acetate (Merck.2. Materials Corn starch (approx. Analytical grade vinyl stearate (Aldrich. In DMSO.
4.6 for starch acetate ester) were obtained. vinyl laurate (Fluka.N‐dimethylaminopyridine (DMAP). The maximum attainable DS of starch acetate in water was below 1 and limited to 0. This solvent effect is likely caused by the higher solubility of the vinyl esters in DMSO compared to water. In addition. Germany) were used without further purification.1. Potassium carbonate (Boom. leading to higher reaction rates. to the best of our knowledge. The reactions were either performed in water or in DMSO using a basic catalyst (Na2HPO4). Germany) were used as received.01 when using vinyl‐laurate. as recommended by Laignel et al . The products were placed directly on the diamond plate and 50 scans with a resolution of 4 cm‐1 were recorded.
7H. Typical example of the synthesis of laurate and stearate esters of corn starch Corn starch (0. 24. 16H. 1.9 (C10).0 (m.2. 95. 7H. 760. 61. CDCl3): δ 0. C12).
FT‐IR (cm‐1): 2920 (C‐H stretching).2.96) H‐NMR (before peracetylation. C6S).3 (m.4. broad peaks. (1H‐NMR). 34. C4‐11). broad peaks.2.0 (C12). C2’). 3S.. 1455 (CH2). vinyl laurate or vinyl stearate (3 mol/mol AHG units in starch) and potassium carbonate catalyst (2 %‐wt with respect to starch) were added and the mixture was stirred at 110 oC for 24 h. 12). C18). 76‐78 ppm. 2850 (C‐H stretching).7 (C11). broad peaks.52): H‐NMR (before peracetylation. Methods 4. 31. 28H. 1. Starch‐laurate (Sample 17. After cooling.5 (m. broad peaks. The product was washed twice with methanol (50 and 25 ml. 1. Table 4.1. the product was precipitated using methanol (100 ml) and separated from the liquid phase by decantation. 3H. Figure 4. C1S‐6S).1.5 (m. 3H.. 1370 (C‐H bending). C4‐11). broad peaks. attached to O‐C2S. 1230 (C‐O stretching). 2. C4‐C17).1.1.3. 3‐6 ppm (m. 28‐32 (C4‐9). 2H. 1740 (C=O).6 (m.5 (m. C2). The atom numbering scheme is given in Figure 4. and O‐C6S).9 (t. C2). C2).
H‐NMR (after peracetylation.9 (t. broad peaks. CDCl3): δ 14. 1150 (C‐O stretching).9 (C3). resulting in the formation of a homogenous transparent solution. 1. 3H. typical spectra are given in Figure 4.9 (broad. (FT‐IR). the product was dried in a vacuum oven (70 oC. 1. C3). 720. 1295. 935 (C‐O stretching). C3). CDCl3): δ 0. DS = 2. (13C‐NMR) and Figure 4.3 (m. The samples were characterized by 1H‐ and 13C‐NMR and FT‐IR.. DS = 2. 2H.Chapter 4
4. respectively). 2H. C2S. CDCl3): δ 0. 1.4 (m. C1S‐6S)
.3. approximately 5 mbar) for 24 h until constant weight.
C‐NMR (before peracetylation. 1020 (C‐O stretching). 22. C3).5 g) was first gelatinized in DMSO (5 ml) at 70 oC for 3 h.1 (m. 7H. 3‐6 ppm (m. broad peaks. 3H. C1S‐6S). C1S). 172‐174 ppm (C=O. O‐C3S. 2. 68‐74 (broad. 16H. Finally.4 (broad. overlap with CDCl3 (C4S). 1350 (C‐H bending).9 (t. 2H.3. 2H. Subsequently. 3‐6 ppm (m. Table 4.8‐2. 1110 (C‐O stretching).3 (m. 2H. 2. 1. 5S).1 (C11). 1410 (C‐H bending). broad peaks. Starch‐stearate (Sample 19.
C18).2.9 (t. and O‐ C6S)
FT‐IR (cm‐1): 2920 (C‐H stretching). 1350 (C‐H bending). C2). 1020 (C‐O stretching).0 (C16). 61. 1100 (C‐O stretching). CDCl3): δ 0. 3‐6 ppm (m. broad peaks. 22. 3H. C1S). 68‐74 (broad. C3). A peracetylation reaction to substitute all of the remaining
.7 (C4S). 3H.6 (m. 2H.5 (m.5 (broad. 1295. C2’). Synthesis of Higher Fatty Acid Starch Esters using Vinyl Laurate and Stearate as Reactants
H‐NMR (after peracetylation. 865.7 (C17). 95. 32.2 (C2). 2H. 1150 (C‐O stretching).8‐2. 26‐32 (C4‐15). C2S. 720. 760.0 (C18).4 (broad. 7H. 1. 1. 1410 (C‐H bending).4 (m. Numbering scheme for carbon atoms of products 4. C4‐C17). C6S).2. 75.3 (m. CDCl3): δ 14. 1740 (C=O). Peracetylation procedure The presence of remaining hydroxyl groups in the products resulted in broad and overlapping starch resonances in 1H‐NMR spectra  and hampered calculation of the DS. 3S. 25. 5S). 2850 (C‐H stretching).
O R O
1S 2S 1
CH3 2' O
O R O
O R O
Figure 4. 2. 172‐174 ppm (C=O. 1. 1370 (C‐H bending).3.0 (C3).1. 34. O‐C3S. broad peaks. attached to O‐C2S. 28H. 950 (C‐O stretching). 1455 (CH2). C1S‐6S)
C‐NMR (before peracetylation.
The procedure was described earlier by Elomaa et al  and the relevant equations are shown in equation 22.214.171.124 ppm (protons attached to the fatty acid carbons C3‐C12 for laurate and C3‐C18 for stearate) using eq.3.2.6 ppm range) and a correction has to be made (eq 4.6−5.).1.6‐5.6 7
DS fatty ester = 3 − DS acetate
AC 2 = 2 × AH − fattyester
The AC2 values of both the laurate and stearate side chains were calculated from the peak intensity in the range δ 0.)
The DS of the acetate groups of the products may be calculated by comparing the unit area of the acetate protons (AH‐acetate) to the unit area of the starch protons (calculated from the intensity of the starch peaks at δ 3.)
AH − starch =
(4. Typically. and 4.
AH − acetate = A1.6 − AC 2 3
(4.) (4.4.8‐2.1 g) was suspended in THF (4%‐w/v) and stirred at 55 oC until the starch was fully dissolved (typically 3 h). The peracetylation reaction was conducted for 7 h at 50 oC.). acetic anhydride and pyridine in a DMAP : acetic anhydride : pyridine : AHG molar ratio of 1: 10: 22 : 1) were added.2. and 4..3. The peracetylation procedure by Einfeldt et al  was applied. Subsequently.5.7.Chapter 4
hydroxyl groups with acetate groups was applied to obtain reliable DS data. It was finally dried overnight in a vacuum oven at 70oC and 5 mbar until constant weight. 4.)
Determination of the unit area of the acetate H‐atoms (AH‐acetate) is hampered by peak overlap with the H‐atoms attached to C2 (AC2) of the fatty acid chains (δ 1. The product was precipitated by the addition of methanol and washed several times with methanol.2.5‐126.96.36.199.8− 2.6 ppm).3. the starch ester (0. the peractylating reagents (DMAP. The DS of the fatty acid esters was calculated from the DS of the products after peracetylation (eq 4. 4.:
AH − acetate AH − starch
A3. Determination of the Degree of Substitution (DS) The DS of the product was calculated using 1H‐NMR spectra of the products after peracetylation.:
A schematic representation of the esterification reaction of starch with the vinyl esters is provided in Scheme 4.)
In eq. Initially. 188.8.131.52 = (stearate) 33 33
(4. After 2‐3 h.7.8 = (laurate) 21 21 AC 3−C18 A0.1).
4. while ACx‐Cy is the area of the H‐atoms attached to carbons in the range Cx‐Cy (carbon numbering scheme is given in Figure 4. Synthesis of Higher Fatty Acid Starch Esters using Vinyl Laurate and Stearate as Reactants
AH − fattyester = AH − fattyester =
AC 3−C12 A0. The products of these exploratory reactions are insoluble in water and DMSO. the starch was gelatinised in DMSO at 70°C for 3 h to make the hydroxyl groups of starch more accessible for reaction.‐4.3. the brownish gel was precipitated with methanol and the product was collected after vacuum drying in the form of a transparent. After 24 h. Exploratory Experiments A number of exploratory experiments were carried out with vinyl‐laurate and vinyl stearate (vinyl ester: AHG molar ratio of 1 : 3) at 110°C for 24 h in DMSO using K2CO3 as the catalyst. light yellow solid.5−1.5−1. Results and Discussion
4. Ax‐y stands for the peak area in the range δ x‐y ppm. The reaction was performed in two discrete steps.1.6.3. the esterified starch started to separate from the medium in the form of a gel.1. Subsequently. the vinylester and the catalyst were added and the reaction mixture was heated to 110°C.. but swell in organic solvents such as toluene and THF.7.
OH O HO OH
R O CH2
O HO O R
(1) R = laurate (C12) (2) R = stearate (C18)
52 (Sample 17.).1. 1H‐ and 13C‐NMR analyses The solubility of the products in common NMR solvents (DMSO‐d6 or CDCl3) is a function of the product DS.1.3. (c) peracetylated starch laurate.) in CDCl3 at 50oC. Medium DS starch laurate and starch stearate (1 < DS < 2) dissolve poorly in DMSO‐d6 and CDCl3.3. and 4.
Figure.2. 4. Product Characterisation 4. DS = 2. For atom numbering scheme: see Figure 4. (b) starch laurate. even at higher temperatures (50 oC).1. Typical 1H‐NMR spectrum of: (a) native starch in DMSO d‐6 at 60oC.) in CDCl3 at 50oC. Table 4.96. A reaction with vinylstearate at similar conditions resulted in a stearate starch ester with a DS of 2.1. Higher DS products have a higher solubility in CDCl3 and good quality 1H‐ and 13C‐NMR spectra could be obtained (Figure 4. DS = 2. 4.
. a product DS of 2.Chapter 4
The DS of the products was determined by using NMR (vide infra).184.108.40.206 was obtained.2.52 (Sample 17. When using a vinyl‐laurate : AHG molar ratio of 3 and K2CO3 as the catalyst. Table 4.
Figure 4. Two of the carbon resonances (1S and 4S) are considerably shifted compared to native starch. NMR spectra of the peracetylated products are considerably improved in terms of peak resolution and allow a more reliable calculation of the DS.3. For atom numbering scheme: see Figure 4. Synthesis of Higher Fatty Acid Starch Esters using Vinyl Laurate and Stearate as Reactants
A typical 1H‐NMR spectrum of starch laurate is shown in Figure 4. required for DS determinations.1. (b) starch laurate.2. in DMSO d‐6 at 60oC.8‐2. The 1H‐NMR spectrum of a typical peracetylated starch laurate is shown in Figure 4. Typical 13C‐NMR spectra of: (a) native starch. are together with the CH2 groups of the acid chain adjacent to the ester moiety in the range δ 1. The starch peaks (δ 3‐5.5 ppm) are broad and overlapping .c. and therefore a peracetylation procedure to substitute all of the remaining OH groups with acetate groups was applied [8.8‐2. The proton signals of the acetate methyl group.5 ppm). Clearly visible are the carbon resonances of the fatty ester chains (δ 10‐35 ppm) and C atom of the ester group (δ 170‐175 ppm). The resonances arising from the anhydroglucose unit of starch are broadened. DS = 2. DS = 2.96 (Sample 19) in CDCl3 at 50oC. This feature hampers the DS determination by NMR. The same
.3. 10‐11]. Clearly visible are the peaks arising from starch and the aliphatic hydrogen atoms of the fatty acid chain (δ 0.2. (c) starch stearate. Typical 13C‐NMR spectra of the products are given in Figure 4.3 ppm.52 (Sample 17) in CDCl3 at 50oC.
Table 4.Chapter 4
phenomenon was observed by Dicke for starch acetate .2. the C‐H stretching vibrations of the alkyl groups of the fatty ester chain are clearly present at 2920 and 2850 cm‐220.127.116.11.96.4. starch stearate
b.). native starch -OH C-H
Figure 4.) and native starch FT‐IR spectra of both starch laurate and starch stearate (Figure 4. FT‐IR measurements The FT‐IR spectra of starch laurate and starch stearate are shown in Figure 4. For comparison.2.b.4.4. starch laurate C-H C=O
a. The near absence of remaining hydroxyl
. FT‐IR Spectra of starch laurate (DS= 2. starch stearate (DS= 2. a spectrum of native starch (Figure 4.1.c.) is also included.1. In addition.4. Table 4. 4. Sample 17.) show characteristic bands of the carbonyl group of the fatty esters in the 1750‐1700 cm‐1 region. and 4.
c.c. The shift of the starch peaks and the presence of peaks arising from the fatty ester chains clearly indicate that the esterification reaction with vinyl laurate and vinyl stearate was successful.3.4.a.b. Sample 19. Characteristic peaks of the polysaccharide backbone are visible in the 1250‐900 cm‐1 region (C‐O stretching) . and 4.
Synthesis of Higher Fatty Acid Starch Esters using Vinyl Laurate and Stearate as Reactants
vibrations in the range 3000‐3600 cm‐1 and at 1640 cm‐1 indicates that the DS of the product is high. Most of the experiments (14) were performed using Na2HPO4 as the catalyst.99 1.08 1. The results are shown in Table 4. in line with the NMR data. Overview of the esterification of starch using vinyl‐esters and basic catalysts a
Catalyst Na2HPO4 Na2HPO4 Na2HPO4 Na2HPO4 Na2HPO4 Na2HPO4 Na2HPO4 Na2HPO4 Na2HPO4 Na2HPO4 Na2HPO4 Na2HPO4 Na2HPO4 Na2HPO4 K2CO3 CH3COONa K2CO3 CH3COONa Vinyl Ester: AHG ratio [mol/ mol] 2 3 6 2 3 6 2 3 4 6 2 3 4 6 3 3 3 3 Amount of Toluene added [ml] ‐ ‐ ‐ 5 5 5 ‐ ‐ ‐ ‐ 5 5 5 5 ‐ ‐ ‐ ‐ DS 1. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Vinyl Ester laurate laurate laurate laurate laurate laurate stearate stearate stearate stearate stearate stearate stearate stearate laurate laurate stearate stearate
All experiments were performed at 24 h at 110°C in DMSO with a catalyst intake of 2 wt% based on starch.01 0. 4.52 2.3.60 1. Table 4.23 0.54 2.1.96 2.07 0.13 1.68 0.90 0. In addition.2.1. type of catalyst and the effect of the addition of co‐solvents on the product DS was studied in more detail.05 0.
.90 1. four experiments were performed with two alternative basic catalysts (K2CO3 and Na‐acetate).91 0.24 2. Systematic Studies The effect of important process variables like the vinyl‐ester to AHG ratio.57 0.44
The highest DS value was 1.23 for vinyl‐laurate at an intermediate vinyl‐ester‐AHG ratio of 3. 2 wt% catalyst intake on starch). This is likely due to a reduction of the polarity of the reaction medium.1. The DS of the products is a clear function of the vinyl laurate and stearate ester intakes (Figure 4.5.5
Degree of Substitution (DS)
4. This reduced polarity is expected to lead to ○ : starch‐laurate □ : starch stearate (lines for illustrative purposes only)
. this positive effect dominates the reaction rate and the DS of the products will therefore increase at higher vinyl ester intakes. Effect of vinyl ester to AHG ratio on the product DS The effect of the vinyl‐ester to AHG molar ratio on the product DS was determined for both types of vinyl‐esters with Na2HPO4 as catalyst (samples 1‐ 3.5
2 3 4 Vinyl Ester: AHG Starch Ratio [mol/mol]
Figure 4. A further increase leads to a reduction in the DS. The DS values are increasing with higher vinyl‐ ester/AHG molar ratio until a certain maximum.5. A further increase in the vinyl ester intake leads to a reduction in the DS.3. At low to medium vinyl ester/AHG ratios (0‐3). the ester intake is equal on a weight basis to the DMSO intake. 7‐10). At a ratio of 1:6. DS of the product as a function of the type of vinyl ester and the vinyl ester‐starch ratio (24 h reaction time. Higher concentrations of vinyl esters are expected to lead to higher esterification reaction rates.).
1.5. The results are presented in Figure 4. This behaviour is likely the result of two opposing effects. 110 oC.
1. leading to enhanced reactivity.). 4. Although toluene may positively effect the reaction by (partly) re‐dissolving starch ester precipitates. These findings were rationalised by assuming that the co‐solvents increase the solubility of the silylated products. A similar trend in reactivity pattern was observed in this study and the DS decreased from 1.3.). The products have about the same DS value for vinyl‐laurate when using only DMSO and even reduced DS values for vinyl‐stearate (see Table 4. The latter factors appears to have a strong effect on reaction rates (vide supra). 4. the starch laurate esters display higher DS values than the starch stearates. The difference was explained by assuming that the reaction rate is reduced for larger reagents due to steric effects and this explanation likely also holds for the reactions with the vinylesters . with high carbon numbers leading to a reduction in the DS. it also results in a dilution of the reaction mixture and a reduction in the polarity. Thus. et al  for the silylation of starch. et al  reported the synthesis of fatty esters of starch using alkanoyl chloride reactants (C8‐C18) with reactant ratio of 6 mol alkanoyl chloride/mol AHG.7 for lauroyl chloride to 0. the addition of co‐solvents (toluene/ THF) led to an increase in the DS. These negative effects dominate the reaction performance at higher vinyl ester/AHG ratios and lead to a reduction in the DS values. the DS of the product is also a function of the chain length of the fatty acid.3. A similar procedure was proposed by Nouvel.1. with reductions in polarity leading to lower reaction rates.2.6. The results are given in Table 4.3. Here. Effect of the addition of toluene as a co‐solvent A number of reactions were performed using a co‐solvent. i. In this case. 11‐14 in Table 4. Catalysts screening A number of alternative basic catalysts for Na2HPO4. When using Na2HPO4 as the catalyst. The addition of toluene for the esterification of starch with vinyl‐esters surprisingly did not lead to improved DS values. For starch
.3. It is clear that Na‐acetate and K2CO3 are considerably more active than Na2HPO4 and products with a significantly higher DS were obtained. and illustrated in Figure 4.e.8 for stearoyl chloride.). Synthesis of Higher Fatty Acid Starch Esters using Vinyl Laurate and Stearate as Reactants
a lowering of the reaction rates due to a reduction of the solubility and degree of ionization of the starch reactant as well as the base catalysts. K2CO3 and Na‐ acetate were tested.3. Aburto.5. the reactions were initiated in DMSO and toluene was added after 12 h reaction time to re‐dissolve the poorly soluble partially‐esterified starch products (entry 4‐6.1. This effect is particularly evident at higher vinyl ester/ AHG ratios (>3) (Figure 4.
and FTIR and confirm the presence of chemically bound fatty acid chains.4. For starch stearate. 110oC. K2CO3 gave products with a significantly higher DS (2. Comparison of DS values with different catalysts (Vinylester: AHG molar ratio = 3:1.). in line with the findings for NaH2PO4.5 were obtained. The yellow products were characterized by 1H‐ and 13C‐NMR. the two catalysts are equally effective and products with a DS of about 2. and Figure 4. The DS of the products is a clear function of the chain length of the fatty ester and the type of catalyst.5
Degree of Substitution
1. DMSO). and Na‐acetate).96) compared to Na‐acetate (DS=2. when using K2CO3 as the catalyst.5
Figure 4. gray: Na2HPO4 black: K2CO3 white: Na‐acetate
laurate esterification. Thus. K2CO3.44).6.6.5
2.1. K2CO3 and Na‐acetate are superior with respect to activity when compared with Na2HPO4. The products were synthesised in DMSO using vinyl‐esters in the presence of basic catalysts (Na2HPO4. catalyst amount = 2%‐w. Conclusions
A study on the synthesis of corn starch fatty acid esters with high DS values is reported. The DS of the laurate ester when using Na‐acetate is higher than for the stearate ester (Table 4. function of the type of catalyst. the DS for the laurate ester is lower than the stearate ester. However. the statement that the DS for the laurate esters is always higher than for the stearate esters is not generally valid and a. Apparently.o. the DS of the product is also tunable by proper catalyst selection.
. Appl. the DS of the product increases at higher vinyl ester intakes. Nomenclature
x − y ppm
: peak area of certain proton in 1H‐NMR spectra [‐] : peak area of certain peak at x until y ppm in 1H‐NMR spectra [‐] : Degree of Substitution. X. Polym. D. Thiebaud. J. Gatenholm: Surface octanoylation of high‐ amylose potato starch films. products with a DS > 2. Higher ratios led to a reduction in the DS. J. A maximum was observed at a vinyl ester: AHG ratio between 2 and 4. Synthesis of Higher Fatty Acid Starch Esters using Vinyl Laurate and Stearate as Reactants
With these catalysts. thermal characteristics and biodegradability of extruded starch acetate foams. Sci. Bengtsson. I. Aburto. Borredon. presumably due to a reduction of the polarity of the reaction medium. Alric. and P. D. Sci. Prinos. Chem. 21. Important product properties will be described in the next chapter. E. Mullen and E. J.
4. 74. Alric. Pacsu: Starch Studies: Preparation and Properties of Starch Triesters. 1997. Polym. Thiebaud. Ind. Panayiotou: Synthesis.W. I. M.
. Appl. J. 2005. Polym. 705‐721. M. Crops Prod. characterization. Appl. A. S. The DS of the products may also be tuned with the vinyl ester/AHG molar ratio. W. Aburto. Hanna: Water solubility. 1440‐1451. 1999. and biodegradability of fatty‐acid esters of amylose and starch. Ind. Carbohydr. E. average value of mole of substituted –OH per mole of anhydroglucose (AHG) units [‐]
DS fatty ester : DS of fatty ester (laurate or stearate) group substituents [‐]
DS of acetate group substituents after peracrtylation [‐]
4. 361‐368. Panayiotou: Properties of fatty‐acid esters of starch and their blends with LDPE. At low vinyl ester/ AHG ratio. . Xu. K. J. Merrill: Properties of Fatty‐Acid Esters of Starch. Eng. Borredon. Prinos. J. . 58.5. J. C.6. Bikiaris. C. 1208‐1217. 1995. 1‐11. D. . 1647‐1656. J. A. Dzenis. Sagar. Sci. 65. 34. 1942. Bikiaris.4 could be obtained for both laurate and stearate esters. Koch. Y. 54. S. References
. 2003. . Polym. E. Y.
Alric. . 57. Bliard. 118‐121. Hyvarinen. 302‐307. G. J. Polymer 2002. Koniarova. Laatikainen. P. Starch‐Starke 2005. T.
. Mano. Dubois. Petzold.. I.L. J. R. Starch‐Starke 2004. . 57. 298. Borredon: Free‐solvent Synthesis and Properties of Higher Fatty Esters of Starch – Part 2. Soininen. Biosci. P. Aburto. Hamaili. Mouysset‐Baziard. E. Al‐Higari: Acylation of starch with vinyl acetate in water. P. G. Borredon: Organic solvent‐free transesterification of various starches with lauric acid methyl ester and triacyl glycerides. Laignel. Mormann. 1. 251‐260. 2001. I. Nouvel. B.Chapter 4
. C. M. Elomaa. Carbohydr. S. M. A. 255‐263. S. Ydens. . L. Einfeldt. 11. Med. Senocq. M. J. R. F. . J. Starch‐Starke 1999. K. Carbohydr. Macromol. M. 2004. 1735‐1743. Mater. J. J. Sci. . Six: Partial or total silylation of dextran with hexamethyldisilazane. Peltonen. W. . 43. . 51. D. Gunther. 261‐267. Kussler. Degee. E. Urtti: Determination of the degree of substitution of acetylated starch by hydrolysis. Mater. Cellulose 2004. C. I.F. Dicke: A straight way to regioselectively functionalized polysaccharide esters. Massiot. Dellacherie. W. Reis: Thermal properties of thermoplastic starch/synthetic polymer blends with potential biomedical applicability. Polym. Nuzillard: Proton NMR spectroscopy assignment of D‐glucose residues in highly acetylated starch. L. 341‐347. Res. 1997. H‐1 NMR and TGA/IR. D. and E. Stein. R. H. Klemm: Preparative and H‐1 NMR investigation on regioselective silylation of starch dissolved in dimethyl sulfoxide. Asplund. Alric. 2003. 145‐152. Aburto. 14(2) 127‐135. 56. . A.
Experimental and Modeling Studies on the Synthesis and Properties of Higher Fatty Esters of Corn Starch
This chapter describes a systematic study on the synthesis of higher fatty esters of corn starch (starch‐laurate and starch‐stearate) by using the corresponding vinyl esters. Mechanical tests show that the products have tensile strength (stress at break) between 2. vinyl stearate
. reaction temperature and type of the catalyst) on the Degree of Substitution (DS) of the starch laurate and starch stearate esters was determined by performing a total of 54 experiments. The reactions were carried out in DMSO using basic catalysts (Na2HPO4. High‐DS products (DS= 2. K2CO3.45‐1.96).7‐3.26‐2. The thermal and mechanical properties of some representative product samples were determined. The results were adequately modeled using a non‐linear multivariable regression model (R2≥0. The thermal stability of the esterified products is higher than that of native starch.39) are totally amorphous whereas the low‐DS ones (DS= 1. The effect of the process variables (vinyl ester to starch ratio. catalyst intake. and modulus of elasticity of 46‐113 MPa.5 MPa. The basicity of the catalyst and the reaction temperature have the highest impact on the product DS. elongation at break of 3‐26%.75) are still partially crystalline. Keywords: starch esters. and Na‐acetate). vinyl laurate.
Two examples using higher vinyl esters (vinyl laurate) were reported by Mormann. 0. and particularly on the synthesis of starch acetate (C2) [11‐12]. The limited use is mainly because of a number of unfavourable properties of starch such as low moisture resistance. most of the plastic waste is not biodegradable. It is cheap and abundantly available .
. is rather limited to date. Sagar and Merill  studied the synthesis and properties of C4‐C6 esters of high amylose starch. The use of vinylesters has also been explored [20‐22]. however. 3‐4]. the research activities were mainly limited to the use of vinylacetate. Most early studies mainly focused on the use of short‐chain carboxylic acid (C1‐C4). The use of methyl and glyceryl esters to prepare starch laurate (C12) ester was investigated but only yielded products with relatively low DS values (max. However. and incompatibility with hydrophobic polymers . The products were biodegradable. Starch is an attractive feedstock for the synthesis of novel biodegradable plastics.Chapter 5
Plastics are very useful materials and are used in large amounts (245 million ton per annum) [1‐2]. Alternative routes to avoid the use of acid‐chloride/pyridine combinations have been developed. The mechanical properties of products with longer ester chains (C4 and C6) products and plasticized acetate esters were considerably improved compared to native starch. Unfortunately. Esterification is one of the oldest methods used to improve starch properties . Mullen and Pacsu [11. packaging material.65) . The use of native starch as a building block for novel biodegradable polymers. The global starch production was estimated at 60 million tons in 2004 . et al . Examples are agricultural mulch . and food service‐ware . et al [15‐19] synthesized longer chain fatty esters (C8‐C18) of potato starch and corn amylose using fatty acid chlorides and pyridine. The high DS esters showed interesting properties. However. high brittleness. the physicochemical properties of the products were not mentioned. and this causes serious environmental problems [1. Chemical or physical modification is required for successful applications as biodegradable polymers. Thiebaud.1. 12] studied the synthesis and properties of of C1‐C6 esters of starch. The products were hydrophobic and the mechanical properties were considerably improved compared to native starch. et al  and Aburto. The development of novel biodegradable plastic materials capable of decomposing when given an appropriate environment and time is of utmost importance . however. Typical examples are blending with polyvinyl alcohol . The largest application area of plastics is the use as packaging material (37%) . polyethylene  and polycaprolactone  and chemical modification by esterification with organic acids [11‐19]. the mechanical properties were still not satisfactorily.
Germany) were used without further purification. and Tg) of each sample were determined from the spectra related to the cooling run and the second heating one. vinyl ester to anhydroglucose ratio. and disodium hydrogenphosphate dodecahydrate (both Merck.24 ‐ 2. Japan). 5. 73% amylopectine and 27% amylose) was purchased from Sigma (Germany).2. Analytical grade vinyl stearate (Aldrich. Germany). Analytical grade methanol.2. each sample was cooled to ‐50°C and then heated again to 200 oC (heating rate 10 oC/min).N‐ dimethylaminopyridine (DMAP).2. the thermal and mechanical properties of representative examples of the highly hydrophobic materials are described and discussed. TGA measurements were performed using a Perkin Elmer TGA 7 Thermogravimetric Analyzer. Materials Corn starch (approx. as recommended by Laignel et al .
5. The samples (about 10 mg) were placed in sealed aluminum cells. Analytical Equipment H‐ and 13C‐NMR spectra were recorded in CDCl3 on a 400 MHz Varian AMX NMR machine. The experiments were modeled using non‐linear multivariable regression.96) were prepared by reacting the starch with vinyl laurate or vinyl stearate in the presence of basic catalysts in DMSO. sodium acetate.1. Technical grade dimethyl sulfoxide (DMSO). The samples were heated to 700 oC in a nitrogen atmosphere with a heating rate of 10°C/min. and acetic anhydride (Merck. pyridine. Tc. Long fatty esters of corn starch with a broad range in degree of substitution (DS= 0. All these chemicals were used as received. It was dried for 48 hour at 105 oC under vacuum (~1 mbar) to reduce the moisture content below 2 % before use. 4‐N. and toluene were obtained from Labscan (Ireland). Germany).2. After a first heating run from room temperature up to 200 oC to delete the thermal history of the material. The spectra were recorded at 50 oC. In addition. Experimental and Modeling Studies on the Synthesis and Properties of Higher Fatty Esters of Corn Starch
We have recently performed exploratory studies on the synthesis of starch esters of higher fatty acids using vinyl laurate and vinyl stearate as the reagents . the Netherlands). DSC analyses were performed on a TA Instruments DSC 2920. This chapter describes systematic studies using Design of Experiments (DOE) to gain insights in the effect of process variables (temperature. vinyl laurate (Fluka. and tetrahydrofuran (THF) were supplied by Acros (Belgium). The catalysts for the esterification reaction were analytical grade potassium carbonate (Boom. Materials and Methods
5. The thermal properties (Tm. T‐Bones samples (with thickness of 2 mm) for determination of the tensile properties were prepared using a melt press
. catalyst type and intake) on the DS of the products.
apparatus (Fontijne, Holland), operated at 150 oC and 150 bar for 3 minutes. The tensile tests were performed using an Instron Series IX Automated Materials Testing System 1.09 at 20oC and a crosshead speed of 30 mm/min. 5.2.3. Methods 18.104.22.168. Typical example of the preparation of laurate and stearate starch esters The corn starch (0.5 g) was gelatinized in DMSO (5 ml) at 70oC for 3 h. After the mixture became a homogenous, colourless solution, vinyl laurate or vinyl stearate (3‐5 mol/mol with respect to the AHG units) and the catalyst (potassium carbonate, or sodium acetate, or disodium hydrogenphosphate, 2‐5%‐w with respect to starch) were added. The reactor content was kept at 80‐110oC for 24 h. The product was precipitated using methanol (100 ml), and the liquid phase was removed by decantation. This precipitation‐decantation procedure was repeated twice using 50 and 25 ml of methanol, respectively, to purify the reaction product. Finally, the product was dried in a vacuum oven (70oC, 5 mbar) for 24 h or until constant weight. 22.214.171.124. Peracetylation procedure and Degree of Substitution (DS) determination The presence of un‐substituted hydroxyl groups in the products resulted in broad and overlapping starch resonances in 1H‐NMR spectra. A peracetylation reaction to substitute all of the remaining hydroxyl groups with acetate groups was therefore applied to obtain reliable DS values. Typically, the starch ester (0.1 g) was added to THF (4%‐w/v) and stirred at 55 oC until dissolution (typically 3 h). Subsequently, the peracetylating reagent (1 mol DMAP, 10 mol acetic anhydride, and 22 mol pyridine per mol AHG units) were added. The peracetylation reaction was conducted for 7 h at 50 oC. The product was precipitated by the addition of methanol and washed several times with methanol before finally dried overnight in a vacuum oven at 70oC and 5 mbar. The DS of the products was calculated using a procedure given in previous work . 5.2.4. Experimental Design The experiments were carried out in 6 blocks. The three variables used in each block were the vinyl ester to AHG starch mol ratio ( x1 ), catalyst intake ( x 2 ), and reaction temperature ( x3 ). All experiments in a block were conducted using a 3‐ variable, 2‐level Full‐Factorial Experimental Design with one center point, giving a total of 23+1 = 9 experiments per block. For each block, a given type of catalyst (Na2HPO4, K2CO3 or Na‐acetate) and vinyl ester (either vinyllaurate or vinylstearate) was applied. This gave a total of 54 experiments (2 types of vinyl
Experimental and Modeling Studies on the Synthesis and Properties of Higher Fatty Esters of Corn Starch
esters x 3 types of catalysts x 9 experiments per block). In a later stage, the type of catalyst was quantified using the pKb of the catalyst ( x 4 ). The ranges for the individual variables ( x1 ‐ x 4 ) are shown in Table 5.1. The mathematical analysis of the experimental data was performed with the software package “Matchad 13” (Mathsoft). The response (y, DS of the products) was modelled using the following expression:
y = β 0 + ∑ β i xi + ∑ β ii xi2 + ∑∑ β jk x j x k
i i j k≠ j
Here, βi, βii and βjk are the regression coefficients obtained by a multiple 2 regression procedure. One of the terms, namely x 4 , caused singularity of the matrix used for the modelling, and was therefore excluded. A t‐statistic was used to rank the individual regression coefficients (βi, βii, or βjk) according to their relative importance . An analysis of variance (ANOVA) was performed to check the adequacy of the model. The applied procedure is well described in the literature  and consists of calculating the sum of squares (SS) for the model and the error, together with the total sum of squares. In combination with the relative degrees of freedom (DF) it is possible to calculate the mean square (MS) for the model and the corresponding error. On the basis of the latter values, the F‐value for the model is calculated. With this information the P‐value for the model is determined. The latter value is related to the statistical significance of the model.
5.3. Results and Discussion
A schematic representation of the esterification reaction of starch with a vinyl ester is provided in eq. 5.2. The starch was gelatinized before the addition of the vinylester to make the starch OH groups more accessible for reaction. The reactions were carried out in DMSO for 24 h using three different basic salts (Na2HPO4, K2CO3, or Na‐acetate) as the catalysts. The products of the reaction were brownish gels which became lighter in colour after product precipitation and washing with methanol. After drying, the products were isolated as transparent, yellowish solids. The products were soluble in organic solvents such as toluene and THF.
OH O HO OH
R O CH2
O HO O R O
(1) R = laurate (C12) (2) R = stearate (C18)
The products were characterized by 1H‐NMR in CDCl3. The peaks of starch (δ 3‐ 5.5 ppm) and aliphatic hydrogen atoms of the fatty acid chain (δ 0.8‐2.5 ppm) were clearly present. The starch peaks were broad and overlapping, and only after the starch esters were peracetylated, the resolution of these peaks was considerably improved. The DS of the products was determined by using 1H‐NMR spectra of the peracetylated starch esters. The effect of three reaction parameters (molar ratio of vinyl ester to AHG units of starch x1 , catalyst intake x2 , and reaction temperature x3 ) on the DS of starch was studied using a full factorial experimental design with one center point. The ranges of the values of the independent variables ( x1 ‐ x3 ) are shown in Table 5.1. An additional variable x 4 (related to the basicity of the catalyst) was also included in order to obtain a general model for starch esterification. A total of 54 experiments were conducted. Table 5.1. Experimental Design Variables
Independent Variables Full Factorial Variables Vinyl ester to AHG‐ starch ratio [mol/mol] Catalyst intake [%‐w]a Reaction temperature [°C] Additional Variable Catalyst basicity (pKb)b
Low (‐1) Middle (0) High (+1)
2.0 3.5 5.0
80 95 110
K2CO3 (3.66) Na2HPO4 (6.8) CH3COONa (9.25)
3 4 5
in %‐wt based on starch intake, in the model (eq. 5.3.) it is transformed to 105 x catalyst mol amount. taken from ref .
The results of the experiment are given in Table 5.2.
41 1.00 2.00 5.00 5.00 3.00 5.00 3.00 3.00 3.32 1.00 5.79
Experiment Vinyl Ester 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 laurate laurate laurate laurate laurate laurate laurate laurate laurate laurate laurate laurate laurate laurate laurate laurate laurate laurate laurate laurate laurate laurate laurate laurate laurate laurate laurate stearate stearate stearate stearate stearate stearate stearate stearate stearate stearate stearate stearate stearate stearate stearate stearate stearate
.00 4.00 5.12 0.72 2.00 3.00 5.46 1.00 5.00 3.54 2.00 4.00 80 80 110 110 80 80 110 110 95 80 80 110 110 80 80 110 110 95 80 80 110 110 80 80 110 110 95 80 80 110 110 80 80 110 110 95 80 80 110 110 80 80 110 110 DS 0.27 1.40 1.00 2.07 0.82 2.00 5.52 2.00 5.00 5.00 5.00 5.48 0.56 1.00 5.00 5.87 1.84 2.00 5.42 2.62 0.00 5.54 2.00 3.44 2.00 2.00 5.00 5.00 5.11 2.23 2.00 3.00 5.00 4.00 3.27 0.00 5.00 3.00 3.00 5. Experimental and Modeling Studies on the Synthesis and Properties of Higher Fatty Esters of Corn Starch
Table 5.00 3.67 2.00 3.00 3.93 0.2.00 3.75 1.50 2.00 5.00 2.94 2.00 5.00 5.53 1.00 4.50 2.00 2.24 2.00 3.09 0.78 2.00 5.00 5.27 1.35 0.00 3.00 5.96 0.00 5.00 2.70 2.00 2.00 2.00 2.17 2.00 5.00 2.00 3.
Overview of experimental data
Catalyst Na2HPO4 Na2HPO4 Na2HPO4 Na2HPO4 Na2HPO4 Na2HPO4 Na2HPO4 Na2HPO4 Na2HPO4 CH3COONa CH3COONa CH3COONa CH3COONa CH3COONa CH3COONa CH3COONa CH3COONa CH3COONa K2CO3 K2CO3 K2CO3 K2CO3 K2CO3 K2CO3 K2CO3 K2CO3 K2CO3 Na2HPO4 Na2HPO4 Na2HPO4 Na2HPO4 Na2HPO4 Na2HPO4 Na2HPO4 Na2HPO4 Na2HPO4 CH3COONa CH3COONa CH3COONa CH3COONa CH3COONa CH3COONa CH3COONa CH3COONa Vinyl : AHGa [mol/mol] 3.00 2.00 3.50 2.52 2.00 5.00 2.00 2.00 3.88 2.00 3.00 3.00 5.50 0.00 5.50 2.00 5.47 1.00 5.00 5.00 Catalyst Temperature [%‐w] [˚C] 2.00 5.12 2.00 3.28 2.59 2.00 2.00 5.00 5.00 5.44 2.00 2.
00 3.3.1104 ‐0.00 3.00 4.96 2.00 2.0027 ‐0.1568 ‐2.0006 ‐0.
stearate stearate stearate stearate stearate stearate stearate stearate stearate stearate
CH3COONa K2CO3 K2CO3 K2CO3 K2CO3 K2CO3 K2CO3 K2CO3 K2CO3 K2CO3
β0 β1 β2 β3 β4 β5 β6 β7 β8 β9 β10 β11 β12 β13
.0011 0.0001 0.00 2.00
3.1979 Starch Stearate 17.21 2.
DS = β 0 + β1 x1 + β 2 x 2 + β 3 x3 + β 4 x 4 + β 5 x1 x 2 + β 6 x1 x3 + β 7 x1 x 4 + β 8 x 2 x3 + β 9 x 2 x 4
2 2 2 + β10 x3 x 4 + β11 x 2 + β12 x3 + β13 x 4
(5.00 5.0016 0. The type of catalyst was quantified using the basicity constant (pKb) in water [26‐27].00 3.0120 0.55 2.00 3.00 5. Mathematical Modeling Modelling of the DS data for both vinylesters was performed using non‐linear multivariable regression based on the 4 independent variables (Table 5.00 5. 5.50
95 80 80 110 110 80 80 110 110 95
1.).0249 ‐0.1.3.).64 2.Chapter 5 45 46 47 48 49 50 51 52 53 54
a.3.00 3.00 5. The experimental data for both starch laurate and starch stearate esterification are best described with a quadratic model including interaction terms (eq.5650 0.0470 0.00 5.0327 ‐2.50 2.)
Table 5.2183 0. Values for the coefficients of the DS model for starch laurate and stearate
Coefficient Starch Laurate 9.59 2.0023 0.0008 0.0009 0.00 2.0008 ‐0.98 2.0017 0.00 5.00 5.1082 ‐0.6939 0.00 5.1748 0.41
mol ratio of the vinyl ester to AHG units of starch
5.3.90 2.1.0066 0.64 2.0003 0.3.0732 ‐0.The modelled values for βij for the two esters are given in Table 5.0010 ‐0.0857 ‐0.
Analysis of variance for DS models for starch esterification
Starch Laurate Model Error Total SS 16.6118 0.5
Starch Stearate Model Error Total SS 20. Also in this case the reasonable R2 values (0.967 (starch stearate).5
0.945 0. and confirm this statement.5 DS actual
2.2903 P‐value 1.4.970 0.9531 0. respectively and it may be concluded that the models describe the experimental data well.875) indicate that the model correctly predict the
.3399 0.96 x 10‐8 R2 values R2 R2adjusted R2press 0.5
a.2569 0.970 (starch laurate) and 0. The parity plots of both models are given in Figure 5.1.5 DS actual
2.0356 F 35.5
2. The very low P‐ values indicate that the models are statistically significant.8167 21.8386 DF 13 14 27 MS 1. we also performed a PRESS analysis  (see corresponding R2 values in Table 5.939 0.4986 16. which represents an “internal” validation method for the model.7699 DF 13 14 27 MS 1.1. Parity plots of the DS models for starch laurate and stearate Moreover.805
2.026 x 10‐7 R2 values R2 R2adjusted R2press 0. Table 5. starch stearate
Figure 5. The R2 values for the models are 0.629 P‐value 1.5
2 DS predicted
1. Experimental and Modeling Studies on the Synthesis and Properties of Higher Fatty Esters of Corn Starch
The result of the analysis of variance is given in Table 5.4.5
1.967 0.4. which indicate  that all significant variables are included in the model. the adjusted‐R2 values are very close to the R2 ones. starch laurate
b.0583 F 27.5
K2CO3 is a stronger base (pKb 3. However. This was not the case for alkaline catalysts.Chapter 5
products DS as function of the process variables within the range of experimental variables.82). On the basis of the sequence. A possible explanation of this peculiar behaviour of Na2HPO4 is the fact that the pKb values in water are used for quantification whereas the actual solvent for the reaction is DMSO.18) and CH3COONa (1. such as carbonate or acetate salts leading to higher DS esters.25) and this leads to higher product DS values for K2CO3. The pKb for the latter (6. The highest product DS values were obtained using K2CO3. Dicke  showed that Na2HPO4 has a strong tendency to selectively deprotonate the OH group at the C2 position of starch leading to C2 substituted acetate esters. the type of catalyst has the largest effect on the product DS for both the laurate and stearate esters (Figure 5. The explanation for the high preference of C2 substitution for the Na2HPO4 catalyst is not yet known and needs to be established by mechanistic studies. These results may be rationalised by considering the role of the catalyst in the modification reaction. Another explanation may be related to the regio‐chemistry of the reaction.2. for both starch laurate and starch stearate are given in Figure 5. The product DS for Na2HPO4 was considerably lower (1.00) than for K2CO3 (2. To evaluate the effect of each variable on the product DS. whereas catalyst performance is considerable lower. the base strengths of the catalysts in DMSO are not known. A similar trend was observed by Dicke  for the acetylation of Hylon VII starch using vinylacetate. This is indeed the case when comparing the performance of K2CO3 with CH3COONa. the performance of Na2HPO4 does not follow this trend.8) is intermediate between that of the other two catalysts.2. It is assumed that the first step in the reaction sequence is activation of the starch OH groups by deprotonation by a base [21. Unfortunately. Catalyst performance of CH3COONa was slightly less than for K2CO3 whereas the lowest DS products were produced when using Na2HPO4 as the catalyst. Of all the variables studied.
.66) than CH3COONa (pKb 9.). it can be rationalised that the rate of the reaction (and thus the product DS) will be higher when using a stronger base. two plots showing the dependence of the DS on reaction temperature and catalyst amount (in %‐w). The resulting anion will react with the vinyl ester to from the product. 22].
the catalyst intake and reaction temperature also affect the product DS (Figure 5. starch laurate
b. As expected and in
.2. starch stearate
Figure 5. Experimental and Modeling Studies on the Synthesis and Properties of Higher Fatty Esters of Corn Starch
a.2. 3D contour plot of the DS as a function of reaction temperature and catalyst intake (at constant vinylester to AHG mol ratio of 4) Besides the type of catalyst.). although to a lesser extent.
Representative DSC spectra for the starch stearate are given in Figure 5. Of all variables studied. 5.3. and the results for all samples are summarised in Table 5.45 (. . the vinyl ester to AHG ratio has the smallest effect on the product DS.39)
.26. higher reaction temperatures and catalyst intakes lead to higher product DS values. DSstearate = 1.39) were determined.2.75.3.Chapter 5
line with studies on starch acetates . Product Properties The thermal and mechanical product properties for two starch‐laurate and stearate samples with medium (DSlaurate = 1.
Tm = 21 oC
Tc = 12 oC
10 Temperature ( C)
Figure 5.5. DSstearate = 2. DSC analysis of starch stearate esters ( ) DS 1. .3.) DS ( 2.45) and high DS values (DSlaurate = 2.
75 2. 22]. 17].d.= not detectable
The thermal behavior of the products is a strong function of the DS. Experimental and Modeling Studies on the Synthesis and Properties of Higher Fatty Esters of Corn Starch
Table 5. crystallization.
. These values are reduced considerably compared to virgin corn starch. TGA analysis was performed to study the thermal degradation behavior of the starch laurate and stearate samples (Figure 5.26 1.
Melting point (Tm) and crystallization temperature (Tc) of representative starch laurate and stearate samples
Starch ester Starch laurate Starch laurate Starch stearate Starch stearate DS 2.
n.a 21 Tc [˚C] n. At moderate DS values the products still display a melting temperature (at 20‐30 oC) and a crystallization temperature (at 10‐25 oC). 28‐30 oC) and the authors concluded that the transition temperatures are solely determined by the side chain without any significant contribution of the starch backbone. octadecane. these are the first examples of completely amorphous starch esters. However. In particular the behaviour at relatively high temperatures (>0 oC) is further discussed. especially in the amylopectin part [15.39 1.a 22 n.a 27 n. Unfortunately the thermal transitions for the virgin material lie above the degradation temperature and can therefore not be accurately determined .5. close to the value obtained in our study for the medium DS starch stearate (21 °C). The melting point of the potato starch stearate was close to that of model compounds for the stearate side chains (methyl stearate. To the best of our knowledge. or glass transition above 0 oC) were even absent for the high DS products. This clearly indicates that the observed thermal transitions are not only determined by the side chains but are an interplay between that of the starch backbone and ester side chains.8 .d. in the present work.d. 40‐42 oC. implying that these products are fully amorphous.d. Changes in the transition temperatures of native starch by esterification have been reported in the literature [13‐14. For instance a Tm of 32 oC was reported for a potato starch stearate ester with a DS of 1.). the Tm for the medium DS starch laurate (27 oC) is much higher than the corresponding side chain model compounds (‐10 oC for dodecane and 5 oC for methyl laurate). which is known to contain crystalline areas. esterification of starch with vinyl laurate and stearate reduces the crystallinity of the starch considerably and leads to a lowering of the transition temperatures. Thermal transitions at relatively high temperature (melting.45 Tm [˚C] n. In any case.a 12
The mechanical properties (stress at break.) curves clearly show starch esterification with either laurate or stearate results in products with enhanced thermal stability. .6. Similar trends were reported for amylose octanoate‐ stearate‐laurate esters  and potato starch octanoate‐ laurate esters .39)
The TGA (Figure 5.Chapter 5
Temperature ( C)
Figure 5.4. The difference in thermal stability between the laurate and stearate sample is limited. elongation at break and elasticity modulus) of the products were determined (Table 5.
.4.): starch stearate (DS 2. Native corn starch degrades at lower onset temperatures (250‐300 oC) than the corresponding esters (300‐350 oC).26) (.). ( ): native starch ( ): starch laurate (DS 2.
TGA analysis of various starch samples. .
0 112.1 ± 15.2 8.5 MPa.9 ± 14.0 3.5 ± 1.9 46.6. plastic deformation and yield strain of about 0.
. independently of the DS values.025) lies still in the typical range of engineering polymers but also close to that typical of composite materials . however.7 ± 14.1 ± 1.45 Stress at break [MPa] 3. observed for the high DS starch stearate sample.7 ± 4.7 ± 4.7 ± 1.1 Elongation at break [%] 26.39) does not show any plastic behavior but only an elastic one. Both starch laurate samples display.8
Starch Ester Starch laurate Starch laurate Starch stearate Starch stearate
All materials generally show tensile strengths between 2.). On the other hand.39 1.1 82. This product is more rigid compared to the product with medium DS value (higher tensile strength and elasticity modulus.0 3. and modulus of elasticity between 46‐113 MPa. This indicates that their mechanical behavior is comparable to that of typical engineering polymers . elongation at break between 3‐26%. thus indicating a relatively more rigid material with respect to all other samples. This is in agreement with previous studies regarding the mechanical properties of starch esters with different chain length of the fatty acid moiety [13.2 2.
Mechanical properties of the starch esters
DS 2.05 for both laurate samples. The values of the yield strain. starch stearate at relatively high DS values (2. a clear plastic behavior with a maximum in the stress‐strain curves.05). determined from the stress‐strain curves as the stress at which the material ceases to be linearly elastic) and the modulus. The corresponding yield strain (roughly 0.0 ± 2.1 Modulus [MPa] 59.e. An inverse behaviour is. These considerations strongly point out the fact that the mechanical behavior can be coarsely (plastic vs elastic behavior) but also finely (stress and strain at break but also modulus values) tuned by the chemical structure of the fatty acid chains as well as the DS values. but higher elongation at break.3 3.3 21.26 1.75 2. A deeper understanding of the mechanical behavior can be gained by investigating the shape of the stress‐strain curves (typical examples in Figure 5. is roughly 0.7 ± 3.3 ± 2.5. Starch stearate with a relatively low DS (1.7‐3. 17]. Experimental and Modeling Studies on the Synthesis and Properties of Higher Fatty Esters of Corn Starch
Table 5.45) displays a very similar behavior with respect to the laurate samples (i. an increase in product DS results in lower tensile strengths and modulus of elasticity. and lower elongation at break). calculated as the ratio between the stress at yield (σy.2 ± 0. For starch laurate.
Although there are many differences in the synthetics methods as well as in testing conditions and procedures.): starch stearate (DS 1. Stress‐strain curves for starch esters. .39).).
4 stress (MPa)
0 0 5 10 15 20 strain (%) 25 30 35
Figure 5. .7.26) (.75) ( ): starch laurate (DS 2. ‐): starch stearate (DS 2. a rough comparison can be made between the mechanical properties of the esters described in this chapter and those for reported for related starch esters synthesized with alkanoyl chloride as reagent and pyridine as catalyst  (Table 5. ( ): starch laurate (DS 1.5.45) (‐ .
They related this inverse property of starch stearate to crystallization of C18 side chains.5 MPa) is coupled with a low elongation (26 %). DS) which mainly determines the thermal and mechanical behaviour.).3
High DS (2.26) 3.7 ±0.7 ±0.
Comparison of the mechanical properties of corn starch ester (this study) with potato starch ester 
Corn Starch Laurate Ester Potato Starch Laurate Ester Low DS n.1 21.1 ±1.7.
.6 9 ±2 High DS (2.75) 3.2 ±0.1 ±15.3 10 ±2
Low DS (1.7 ±4. this comparison also confirms our conclusion derived from the analysis of thermal properties (vide supra): it is not only the individual factors (kind of starch and fatty acid.7.6 Corn Starch Stearate Ester Low DS (1. 9% for potato starch). as confirmed by DSC.26) relatively high stress at break (3.45) 3. Moreover.a.39) 2. Here.7 MPa) coupled with a significant (1500%) elongation. but the interplay between those factors.8) 3.3 Potato Starch Stearate Ester Low DS (1.7) 1.2
Stress at Break [MPa] Elongation at Break [%]
a.2 8.a. 3. High DS (2.7 ±3.= not available
Inspection of the mechanical properties of the starch esters as shown in Table 5. an unexpected behavior was also observed for the mechanical properties of the starch esters from potato starch. In this respect the difference in stress and elongation at break for the high‐DS starch laurate sample is striking. the high DS material is still more rigid.a.a.0 ±2.1 High DS (2. for potato starch exactly the opposite is observed: relatively low stress (0.4 1500 ±8. While for corn starch (DS 2.7 ±1. The same considerations are valid for starch stearate at high DS values (higher stress and lower elongation when using corn starch instead of potato one).a. we show that although no crystallization of C18 occurred (see DSC data of the high DS material in Table 5.9 ±0. On the other hand at low DS values comparable stress at break (3.
n. Their medium or high DS stearate esters of potato starch are more rigid compared to the starch esters from lower chain fatty acids (octanoate and laurate) . confirms that the corn starch esters synthesized in our research are relatively rigid materials as compared to the potato starch esters.2 MPa for corn starch esters. In agreement with our own data.7 MPa for potato starch ones) are coupled with significant differences in the elongation (22% for corn starch esters.5.7) 0. n.5 ±1.0 26. Comparison with the literature data implies that the mechanical behaviour of starch esters of higher fatty acids is also a clear function of the type of starch used. merely because of the structural property of the C18 side chain with respect to the starch backbone.0 3. Experimental and Modeling Studies on the Synthesis and Properties of Higher Fatty Esters of Corn Starch
Thermal and mechanical behaviour of the samples of different DS values clearly shows that the DS represents one of the crucial factors affecting the final product properties.4.4‐3 for starch stearate). Reaction temperature and catalyst intake also affect the product DS although to a lesser extent than the type of catalyst.96 and P‐value of ≤10‐7) second‐order mathematical models correlating the effect of process variables (vinyl to AHG‐starch mol ratio. The starch esters were synthesized by reacting gelatinized starch with vinyl laurate or vinyl acetate in DMSO in the presence of basic catalysts (Na2HPO4. Statistically adequate (R2 ≥0. Comparison with literature data indicates that the properties may be (fine) tuned also by the starch source. The use of K2CO3 and CH3COONa catalysts resulted in medium‐high DS products (2.5 for starch stearate). reaction temperature. while the use of Na2HPO4 resulted in low‐medium DS products (0.5. catalyst intake.Chapter 5
5.9 for starch laurate.5 for starch laurate.07‐1. The DS of the products is a strong function of the basicity of the catalyst. and Na‐acetate). Conclusions
A systematic study. 0. depend on the unit of related variable x] : constant regression coefficients obtained by a multiple regression procedure [‐] : regression coefficients of variable xi obtained by a multiple regression procedure [various units] : regression coefficients of quadratic variable xi2 obtained by a multiple regression procedure [various units] : regression coefficients of interaction variable xjxk obtained by a multiple regression procedure [various units]
β0 βi β ii
: Degree of Substitution.
5. 1.1‐2. the chemical composition of the starting materials (either starch or the ester precursor) as well as the processing parameters affect the final product DS values and as such provide an effective toolbox to modulate the desired product properties for a given application. on the synthesis of corn starch esters with long chain fatty acids is described. including statistical modelling. K2CO3. and catalyst basicity) to the DS of the starch ester products were developed.3‐1. The models may be used to determine the appropriate process conditions to obtain a product with a pre‐defined DS. Thus. average value of mole of substituted –OH per mole of anhydroglucose (AHG) units [‐]
: regression coefficients obtained by a multiple regression procedure [various units.
accessed on August 01. Experimental and Modeling Studies on the Synthesis and Properties of Higher Fatty Esters of Corn Starch
DS fatty ester : DS of fatty ester (laurate or stearate) group substituents [‐]
: base dissociation constant [‐] : temperature [oC] : crystallization temperature [oC] : melting temperature [oC] : experimental variables used for mathematical modelling using multiple regression procedure [various units]
: stress at yield [MPa]
. 592‐595. P.dk/isi/stat/rawmaterial. . Princeton University Press. R. in The Food Biopack Conference: Foodstuffs. W. . in Packaging in the Environment (Ed. . Weber). Frederiksberg C. Culter: Plastics Packaging. Germany. R. Westhoff. Munich. 19. and Regulations. M. 2008 .6. 2008. Munich. Otey. Belgium. The compelling facts about plastics: An analysis of plastics production.
. Denmark: http://www. Ind. 2000. Blackie Academic & Professional. Doane: Starch‐based blown films. C. Website of International Starch Institute. S. .J. F. Chem. Stevens: Green Plastics. . UK. Ehrenstein: Polymeric Materials. M. 2000. 1980.W. Bastioli: Global Status of the Production of Biobased Packaging Materials. An Introduction to the New Science of Biodegradable Plastics. H. Glasgow. Applications. Eng. Department of Dairy and Food Science.html. Denmark. 2002. A. Hernandez. Germany. E. M. Prod. J. Hanser Publishers. Structure – Properties ‐ Applications. . Packaging. C. demand and recovery for 2006 in Europe. E. Dev. J. USA. Res. Levy). Processing. Properties. and Biopolymers. 27‐29 August 2000 (Ed. The Royal Veterinary and Agricultural University. Marsden: The challenge of domestic waste disposal. Selke. 2000. G. D. Hanser Publishers.starch. Denmark. 1986. Conference Proceedings. Plastics Europe. S. Copenhagen. Brussels. G. Association of Plastics Manufacturers. New Jersey. Aarhus.
Alric. D.D. Macromol. Merrill: Properties of fatty‐acid esters of starch. J. Y. Sagar. Janssen. 1943. S. Alric. X. J. J. J. Aburto. Mullen. Ind. K. Borredon: Organic solvent‐free transesterification of various starches with lauric acid methyl ester and triacyl glycerides. Borredon: Preparation of long‐chain esters of starch using fatty acid chlorides in the absence of an organic solvent. 1647‐1656. 57. J. 137‐144. Sci. C. J. Bikiaris. Ind. . Borredon. A. 705‐721. Bikiaris. Mouysset‐Baziard. Sci. S. . D. Aburto. Alric. J. 145‐152. 51. A. D. Appl. F. . L. Borredon: Free‐solvent synthesis and properties of higher fatty esters of starch ‐ Part 2. E. Rijksuniversiteit Groningen. Beenackers: The kinetics of the acetylation of gelatinised potato starch. Wang: Properties of starch blends with biodegradable polymers. W. J. . de Graaf. . M. Wang. Thiebaud.. Al‐Higari: Acylation of starch with vinyl acetate in water. C. 1996
. R. E. Department of Dairy and Food Science. Senocq. Panayiotou: Properties of fatty‐acid esters of starch and their blends with LDPE. E. . Polym. J. 65. Denmark. Biobased packaging materials for the food industry. Mullen. 2000. Mormann. 28. 1999. A. Sci. Thesis. 1942. . . 385– 409. I. Polym. . 34. 51. Starch‐Starke 1999. W. A. Starch‐Starke 2004. Starch‐Starke 1999. Alric. Panayiotou: Properties of octanoated starch and its blends with polyethylene. 74. Weber). Pacsu: Starch studies: possible industrial utilization of starch esters. B. 2003. I. E. 1997. Chem. 1995. Carbohydr. Hamaili. 381‐384. Panayiotou: Synthesis. E. E. Pacsu: Starch studies: preparation and properties of starch triesters. W. J. Z. Eng. Broekroelofs. report of Food Biopack Project (Ed. I. J. Chem. 1997. Prinos. S. Thiebaud. Polym. Polym. M. Aburto. Thiebaud. . C. the Royal Veterinary and Agricultural University. J.. K. J.Chapter 5
. de Graaf: Ph. I. Bikiaris. . E. J. Prinos. E. 132‐135. and biodegradability of fatty‐acid esters of amylose and starch. Aburto. Alric. J. 34. Appl. A. 1209‐1217. Borredon. W. 58. Polym. Borredon. L. . Aburto. P. G. . A. R. Rev. Frederiksberg C. 118‐121. status and perspectives. D. Yang. Alric. Carbohydr. Eng. H. 101‐112. C. 1440‐1451. Aburto. Starch‐ Starke 2005. I. 1995. Part C: Polym. 56. Appl. M. G. J. characterization. 35. C. E. 302‐307. I. Sci. C43. Prinos.
Butterworth‐Heinemann. Janssen. . USA. B. Prentice‐Hall. New Jersey. C. Ashby. J. Experimental and Modeling Studies on the Synthesis and Properties of Higher Fatty Esters of Corn Starch
. Laignel. Cellulose 2004. Underwood: Quantitative Analysis 5th edition. Day. M. A. M. . Picchioni. Processing and Design. 11. M. K. Science. accepted for publication in Starch‐Starke. New York. Manurung. . Carbohydr. 2007. D. C. 1997. Massiot. L. H. Res. 298. John Wiley & Sons Inc. L. J. P. Bliard.. G. Cebon. USA. L. Montgomery: Design and Analysis of Experiments 5th edition. 1986. 255‐263. Nuzillard: Proton NMR spectroscopy assignment of D‐glucose residues in highly acetylated starch. Heeres: Synthesis of higher fatty acid starch esters using vinyl laurate and stearate as reactants. Junistia. Sugih. A. Amsterdam. H. Materials: Engineering. R. Shercliff. . 2008. D.
. A. R. Dicke. the Netherlands. A. B. F. 251‐260. . 2001. R. A straight way to regioselectively functionalized polysaccharide esters.
at relatively high GMA and BPO intakes. PCL‐g‐DEM has at least a comparable performance with respect to PCL‐g‐GMA. glycidyl methacrylate. This discrepancy could be explained on the basis of the functional groups (GMA or DEM) distribution along the PCL backbone. The effect of important process variables (intiator and monomer intakes as well as estimated solubility of the monomer in molten PCL) on the degree of grafting (FD) of the GMA and DEM units to the PCL backbone was explored in detail and quantified using multivariable linear regression. the mechanical behavior of the compatibilized blends is in general different from that of pure PCL and of the corresponding uncompatibilized blends. The reactive interfacial agents were tested for their performance in PCL‐starch blends. Keywords: starch‐PCL blend. The FD values for PCL‐g‐DEM were considerably lower (up to 7 %). PCL‐g‐ glycidyl methacrylate (PCL‐GMA) and PCL‐g‐diethyl maleate (PCL‐g‐DEM) is decribed. Highest values of the FD (up to 45 %) were observed for PCL‐g‐GMA. diethyl maleate. despite the expected differences (favorable to PCL‐g‐GMA) in the in situ formation of the compatibilizers. Both act as compatibilizers for PCL/starch blends by improving the interfacial adhesion between the starch particles and the PCL matrix. The compounds were prepared by reacting a low molecular weight PCL (Mw 3000) with glycidyl methacrylate or diethyl maleate in the presence of benzoylperoxide at 130°C. At relatively low starch intakes. In particular the elastic modulus for the compatibilized blends is significantly higher than that of uncompatibilized ones. grafting
. compatibilizer.Chapter 6
Synthesis and Properties of Reactive Interfacial Agents for Polycaprolactone‐Starch Blends
The synthesis of two reactive interfacial agents for starch‐PCL blends. As a result.
6. One of the well known modification strategies is blending the starch with polymers displaying a stronger hydrophobic character and better mechanical properties. 2]. the native material is very hydrophilic and important mechanical properties are inferior compared to most synthetic polymers and this hampers its direct use as packaging materials. such as polyethylene (PE) or polystyrene (PS) [3‐6]. Introduction
Starch is a cheap and abundantly available natural polymer with very good application perspectives in the area of biodegradable plastics. it would be actually more accurate to define the block copolymer as an “interfacial agent”. along the backbone.e. polycaprolactone (PCL). In general a compatibilizer is a block‐copolymer where each block displays a chemical structure equal or very similar to that of the polymers to be mixed. Unfortunately. Among these. PCL has gained considerable interest for possible applications in the fields of packaging materials and medical applications [10‐11]. Unfortunately. blends of hydrophilic starch and hydrophobic biodegradable polyesters exhibit phase separation  due to differences in polarity of the building blocks. Aerobic soil‐burial experiments showed that the mechanical properties of PCL films decrease rapidly in time . these synthetic polymers are poorly or non‐biodegradable. an interfacial agent) may be used to improve the interfacial association between the two polymer phases. Blending of starch and synthetic biodegradable polyesters has been widely applied for the synthesis of fully‐biodegradable products. This leads. Such structure is rather difficult to achieve by simple copolymerization methods and it is usually prepared in situ (i. polyesters are considered very promising alternatives .
. to an ideal compatibilizer having both PCL and starch blocks. upon mixing) by using a functionalized PCL.1. To reduce the tendency for phase separation. is known to be degraded with ease by microorganisms widely distributed in nature . However. The ester bonds are susceptible to attack by water and this leads to enhanced biodegradability.e. Starch modifications to improve the product properties like enhanced hydrophobicity and mechanical properties were already reported in the early 19th century [1. To overcome this issue. as in this study. for starch/PCL blends. This phenomenon is highly undesirable and limits the application range considerably . When using ungelatinized starch as a component in the blend. As a consequence. A well‐known biodegradable polyester. The latter displays the presence. of polar groups (usually epoxides or anhydrides [13‐18]) able to react with the – OH groups along the starch backbone. a compatibilizer (i. synthetic biodegradable polymers have been applied. It must be stressed here that the word “compatibilizer” is correctly used only when the block copolymer is actually able to significantly influence the dispersion of the polymers to be mixed (most probably through a steric stabilization mechanism ).
~1 mbar) prior to use. Methods 6.
. Materials Polycaprolactone (PCL) CAPA 2304.2. an absolute novelty of the present research. Intakes for each experiment are given in Table 1 and 2. Belgium.8%) from Labscan.
6. Mw=3000 from Solvay Caprolactones. The starch was dried for at least 24 h at 110oC under vacuum (approx. 73% amylopectin and 27% amylase) was obtained from Sigma and high molecular weight PCL (CAPA 6503. 6. UK. This chapter describes a systematic study on the synthesis of two reactive interfacial agents for starch‐PCL systems. after which the equipment was stopped and the chamber was opened to collect the samples. This low molecular weight PCL grade was used without further purification. 16].2.2.1. Corn starch (with approx. Glycidyl methacrylate 97% purity (Aldrich). Synthesis and Properties of Reactive Interfacial Agents for Polycaprolactone‐Starch Blends
which is able to mainly improve the interfacial adhesion between the polymer and starch itself. diethyl maleate ≥97% purity (Fluka). Germany) were used as received. Ireland. The synthesis and application of PCL‐g‐DEM is. to the best of our knowledge.2.8%) from Merck. Materials and Methods
6. Germany and methanol (99. systematic studies and subsequent quantification of the functionalization reaction has not been reported to date. Exploratory studies on the synthesis of PCL‐g‐GMA and its applications for starch/PCL blends have been published [13. Mw=50000) from Solvay Caprolactone. PCL‐g‐glycidyl methacrylate (PCL‐g‐ GMA) and PCL‐g‐diethyl maleate (PCL‐g‐DEM). Compatibilizer synthesis The compatibilizers were prepared in a Brabender Plasticorder PL2000 batch‐ kneader (chamber volume 35 cm3). >99%) was obtained from Acros. xylene (99.2.2. After the PCL was melted (1‐2 minutes). The various compatibilizers have been tested for their performance in PCL‐starch blends. a solution of BPO in GMA or DEM was added drop by drop over a period of 5 minutes. The materials were mixed for another 5 minutes. UK was used for the preparation of the interfacial agents.1. The kneader was heated to 130 oC and PCL (CAPA 2304) was added while maintaining a rotational speed of 80 rpm. Tetrahydrofuran (THF. The intake of reagents was maximally 75‐80% of the volumetric volume to ensure proper mixing. The effect of important process variables on the degree of grafting of the GMA and DEM units to the PCL backbone has been explored in detail and quantified using multivariable linear regression. and benzoyl peroxide 75% (Merck. however.
PCL‐g‐DEM (5 g) was dissolved in xylene and refluxed at 150 oC for 2 h.Chapter 6
6.3. the samples were covered with a paladium/platinum conductive layer of 3 µm thickness. created using a Cressington 208 sputter coater. Infrared spectra were collected with a FT‐IR apparatus in the ATR mode using a Spectrum 2000 instrument from Perkin Elmer.8%. The solvent was decanted and the solid product was dried in a vacuum oven (40 oC.2. The solvent was decanted and the solid product was dried in a vacuum oven (40 oC. Aldrich) as the solvent. The content was blended for 15 minutes.2. 5 mbar). Work‐up of PCL‐g‐GMA products  To remove unreacted GMA monomer and GMA homopolymer. stirred for 1.2. 6. and the product was precipitated at 6‐8 oC. Tensile tests were performed using an Instron 4301
. The error on the transition temperature is assumed to be ± 1 oC and 5 % of the calculated values for the corresponding enthalpies.2. Methanol (450 mL) was added to the filtrate.2.2. 6. 6. Subsequently. Before analysis.5 h.3. Work‐up of PCL‐g‐DEM products Purification of the PCL‐g‐DEM product was performed according to a modified procedure for PCL‐g‐maleic anhydride . Analytical Methods H‐NMR measurements were performed using a 400 MHz Varian AMX Oxford NMR apparatus with CDCl3 (99. Preparation of PCL‐starch blends with the reactive compatibilisers The PCL‐starch blends were prepared in a Brabender Plasticorder PL2000 batchkneader (chamber volume 35 cm3).2. Each sample was first heated from 0 oC to 100 oC (heating rate 10 oC/min) to remove the thermal history of the material. Scanning Electron Microscopy (SEM) was performed using a Jeol 6320 F Scanning Electron Microscope. PCL was added to the chamber followed by the addition of the starch and the reactive compatibilizer. Digital Scanning Calorimetry (DSC) measurements were performed using a Q1000 TA Instruments equipped with a TA Instruments DSC cooling system.4. PCL‐g‐GMA (5 g) was dissolved in 50 ml THF.2. The resulting suspension was filtered and precipitated using methanol (450 mL) at 6‐8°C. An operation temperature of 170°C and a rotation speed of 80 rpm were applied . and then filtered. The transition temperatures of each sample were further determined by first cooling down the samples from 100 oC to 0 oC and subsequently heating up back to 100 oC (cooling and heating rate were 10 oC/min). 5 mbar). the chamber was opened and the resulting material was collected.
When the relative degrees of freedom (DF) are known.)
The FD was calculated using 1H‐NMR by comparing the area of protons belonging to the GMA (‐CH< proton at δ 3. A 5% relative error in the peak area of the NMR spectra was assumed. Table 3). On the basis of the latter values.4.2. This procedure is described in detail in the literature  and consists of calculating the sum of squares (SS) for the model and the error. Calculation of the Degree of Functionalization (FD) of the reactive compatibilizers The number of moles of GMA or DEM present on the PCL backbone was quantified using the degree of functionalisation (FD).1. 19]. 15‐16.2 ppm) side chains with that of a characteristic proton resonance of the PCL backbone (‐CH2‐ protons at δ 4. As a result we were able to obtain a model for the FD of the reaction. For every T‐bone. Statistical Modeling The influence of different processing parameters on the FD values has been determined by performing a multivariable regression procedure on the available data. it is possible to calculate the mean square (MS) for the model and the error. The corresponding value for every blend was calculated as an average of the 8 measurements while the standard deviation was taken as absolute error on the average values. For a given sample/blend. 6. The validity of the model was determined by performing an analysis of variance (ANOVA. The FD is defined as:
number of moles of GMA/DEM attached to PCL (mol) × 100% number of repeating units of the PCL backbone (mol)
(6. 8 different T‐bones were used. the F‐ value for the model is determined followed by the P‐value. Synthesis and Properties of Reactive Interfacial Agents for Polycaprolactone‐Starch Blends
.2.3. strain at break (ε). leading to a 10% relative error in the FD values. 6.1. stress at break (σ) and modulus (E) were measured.0 ppm [13. The latter is a measure of the statistical significance of the model.2 ppm) or DEM (‐CH2‐ protons at δ 4. The T‐bone samples were prepared using a Fontijne Holland TH 400 hot‐ press.
1.3. Functionalization reactions (only showing reactivity for the >CH2 in α position on PCL backbone) Typical 1H‐NMR spectra for the products are shown in Figure 6.1.1.
COOEt COOEt *
GMA BPO *
Scheme 6.3. Results and Discussions
6. Preparation of the Reactive Compatibilizers Twelve compatibilizers were prepared by either reacting glycidyl methacrylate (GMA) or diethyl maleate (DEM) with low molecular PCL as the starting polymer and benzoyl peroxide (BPO) as the radical initiator (Scheme 6.).1.
Synthesis and Properties of Reactive Interfacial Agents for Polycaprolactone‐Starch Blends
Figure 6.1. Typical 1H‐NMR spectra of PCL‐g‐GMA (top) and PCL‐g‐DEM (bottom) Peak assignments were based on available data reported for related products [13, 15‐16, 20]. The FD values and the thermal properties of the products are shown in Table 6.1. (PCL‐g‐GMA) and Table 6.2. (PCL‐g‐DEM).
Overview of experiments for the PCL‐g‐GMA compatibilisers a
Intake FD (%) ‐ 5.7 15.2 27.9 3.2 45.2 c 6.3 T cryst. ( C)
(%‐mol) b GMA BPO ‐ 0.6 0.6 0.6 0.3 1.1 0.3
ΔH cryst. (J/ g‐PCL) 75 72 68 67 73 66 71
T melt. ( C)
ΔH melt. (J/ g‐PCL) 79 72 69 68 73 68 67
PCL PCL‐g‐GMA 1 PCL‐g‐GMA 2 PCL‐g‐GMA 3 PCL‐g‐GMA 4 PCL‐g‐GMA 5 PCL‐g‐GMA 6
‐ 12 24 36 6 24 12
27 22 19 29 25 20 28
51 47 47 48 47 46 47
Experiments were carried out at 130 °C. %‐mol with respect to the CL repeating units in the PCL c based on the soluble fraction of the compatibilizer
Table 6.2. Overview of experiments for the PCL‐g‐DEM compatibilisers a
Intake Sample (%‐mol) b DEM PCL PCL‐g‐DEM 1 PCL‐g‐DEM 2 PCL‐g‐DEM 3 PCL‐g‐DEM 4 PCL‐g‐DEM 5 PCL‐g‐DEM 6
FD (%) ‐ 3.5 2.1 7.2 3.6 6.5 0.9
T cryst. ( C)
ΔH cryst. (J/ g‐PCL) 75 70 68 64 61 63 70
T melt. ( C)
ΔH melt. (J/ g‐PCL) 79 72 62 66 60 66 69
BPO ‐ 0.6 0.6 0.6 1.1 1.1 0.3 27 27 30 21 26 25 30 51 44 47 45 42 41 47
‐ 15 30 45 30 60 7.5
Experiments were carried out at 130 °C %‐mol with respect to the CL repeating units in the PCL
One of the PCL‐g‐GMA compatibilizers (PCL‐g‐GMA 5, see Table 6.1.), was only partly soluble in CDCl3, and therefore the FD is based on the soluble fraction of the compatibilizer only. The presence of an insoluble fraction, combined with a relatively broad molecular weight distribution (as shown by GPC, but not shown here for brevity), suggests that cross‐linking occurred during this experiment. In general, the FD of the PCL‐g‐GMA (3.2‐45.2%) products is much higher than those of PCL‐g‐DEM (0.9‐7.2%). This difference may be either due to the difference in mutual solubility of the GMA and DEM in PCL or differences in the molecular
Synthesis and Properties of Reactive Interfacial Agents for Polycaprolactone‐Starch Blends
mechanism of the grafting reaction. The mutual solubility may be expressed in terms of the differences of solubility parameters of PCL and the reagents . The difference between the solubility parameters of GMA and PCL is 0.29 cal1/2cm‐3/2, while it is much higher (6.3 cal1/2cm‐3/2) for DEM and PCL . Thus, GMA is likely better soluble in PCL, leading to higher values for FD of the products, as confirmed by our experiments. However, the higher experimental grafting efficiencies for PCL‐GMA may also be rationalised by considering the reaction mechanism for the preparation of the compatibilisers. GMA molecules may either react directly with a radical at the PCL backbone or with a radical present on an already coupled GMA molecule. The latter leads to longer GMA grafts on a PCL backbone . A simplified representation of the reactivity of GMA is shown in Figure 6.2.
O (CH2)4 CH2 C O + RO
O (CH2)4 CH
+ GMA O (CH2)4 CH H2C H3C C
C O CH2 HC O O
Termination + H or + R or + RO (CH2)4
+ k GMA O CH C O
H2C H3C C
C O CH2 HC O O
Figure 6.2. Simplified scheme of the GMA Grafting Reaction Mechanism  The reactivity of DEM in radical reactions is expected to be different from that of GMA. Previous studies on maleic anhydride (MA), a compound resembling the
Thus an increase in the GMA intake will provide more monomer available for the growing of the grafted chains leading to higher FD values. Our experimental findings. BPO=0. This trend is independent of the BPO amount and matches with data reported by other groups [13.6% mol/mol-CL units 15 GMA. 6.3. i.6% mol/mol-CL units DEM.1% mol/mol-CL units 10
Substrate Intake (mol% on CL units)
Figure 6. The results are graphically provided in Figure 6. 22]. higher FD values for GMA than for DEM are in line with this explanation and support the proposed molecular reaction mechanisms.e.3. Hence. 130°C) Higher GMA intakes lead to higher FD values. Effect of substrate (GMA/DEM) to PCL ratio on the FD The effect of the substrate (GMA or DEM) to PCL ratio on the FD was studied by changing the GMA/DEM intake at different BPO amounts. BPO=0. the length of a MA graft is always unity whereas longer grafts are possible for GMA.3% mol/mol-CL units DEM.1.3. BPO=1.2. Effect of GMA and DEM to CL‐units ratio (mol/mol) on the FD of the products (constant PCL intake.
GMA. BPO=0. However.). It is most probably related to the fact that GMA is able to propagate to longer grafted chains when reacted with PCL (Figure 6.
chemical structure of DEM. subsequent reactions of MA to an already grafted MA molecule hardly occur.1. showed that MA reacts easily with a radical on the PCL backbone.
The FD values are within a rather narrow range.
50 45 40 35 30
25 20 15 10 5 0 0. no detectable influence of the BPO amount on the FD is observed (Figure 6.5
0. GMA=24% mol/mol-CL units GMA. However. The use of higher initiator concentrations will result in an increase in the number of formed radicals.2 GMA. for which a levelling off and eventually a decrease of the FD values for relatively higher MA amounts has been observed. This will lead to a higher proportion of PCL radicals by hydrogen abstraction from the polymer backbone and thus to higher FD values.6
0.1. although a slight increase in the FD values might be appreciated. constant substrate to PCL ratio) For high GMA to CL ratio (24%‐mol/ mol CL units).4.8
1. GMA=12% mol/mol-CL units DEM.3.2
BPO Intake (mol% on CL units)
Figure 6. 17].2.3
0. Such behavior is slightly in contrast to what observed for maleic anhydride [13.7
0. doubling the amount of initiator results in considerable higher product FD. Apparently.9
1. 15. there
. Effect of the BPO intake on the product FD (130°C.4.4. 17].). DEM=30% mol/mol-CL units
0. Effect of the BPO intake on the product FD The effect of the BPO intake on the FD was studied by using different intakes of BPO (Figure 6.4
0. Such discrepancy is probably related to the relatively low intake of DEM as well as to the different reactivity of DEM compared to MA . 6. These results are in line with earlier work [15. at relatively lower GMA intakes. Synthesis and Properties of Reactive Interfacial Agents for Polycaprolactone‐Starch Blends
The experimental trend for DEM is different.
the BPO and monomer intakes are considered as independent parameters.1. The data for PCL‐g‐DEM shows a similar trend as for the experiments with a high GMA to CL ratio (24%‐mol/ mol CL units). However. defined as the difference in solubility parameters between PCl and the substrates. The results described in the previous paragraph imply that the mutual interaction between these variables and not the absolute value itself determines the final FD values to a great extent. the mutual solubility of DEM in PCL and DMA in PCL was included in the model by using a parameter δ.)
. ni .3.3.8875 + 0. If many macroradicals are created in the system (relatively high BPO intakes) at low GMA intake.)
where nm is the molar amount of monomer in the feed. namely an increase in the BPO intake results in higher product FD.2.Chapter 6
is an optimum ratio between the BPO concentration and available monomer (GMA in this case) on the FD. 6. and 6. To quantify synergic effects of monomer and initiator molar intakes on the FD of GMA and DEM on the PCL backbone.1. The effect is however much less pronounced than for GMA.6022nm ni − 0. The model provided in eq.3.2. recombination of (macro) radicals. 6. In particular the occurrence of “cage effects”. the possibility of side reactions will become significant. δ )
(6.0325nmδ + 0. a statistical model has been developed by performing a multivariable linear regression on the data reported in Table 6.e. Here.2729nm niδ
(6. these investigations focused on a better understanding of the individual variables by studying the effect of higher peroxide and monomer intakes while keeping for example their ratio constant. A doubling of the initiator intake for PCL‐g‐DEM only results in a 70% increase in the FD (compared to 300% for PCL‐g‐GMA). Modeling of the combined effects of the GMA/ DEM and BPO intakes on the FD Previous studies showed the importance of the initiator to monomer ratio on the FD values [15. as well as several transfer reactions might be responsible for the observed trend . In addition. i.5431niδ + 1. 17]. This leads to the following equation:
FD = f (nm . ni the molar amount of initiator in the feed and δ the difference in solubility parameters calculated using group contributions .3. This phenomenon is likely related to the different nature of the radical grafting mechanism of GMA and DEM on PCL as mentioned before. gives the best description of the experimental data:
FD = −1.
5 0.4 0.5. a.941) also suggests that all important variables have been included in the model. and Figure 6.928 P‐value <10‐9
The value of the coefficients in the model imply that the FD is positively influenced by the interaction between monomer and initiator intake ( nm ni ). Table 6.
Model Error Total
Analysis of variance for the FD model provided in eq 6.3.
Figure 6.2 10 20 GMA Intake (mol% on CL units) 40 30 50
60 FD (%) 40
20 0 1 0.8 0.358 1909.7 0.6.
1 0. Synthesis and Properties of Reactive Interfacial Agents for Polycaprolactone‐Starch Blends
The analysis of variance data are given in Table 6.9
BPO Intake (mol% on CL units) 0.17 F 44.957) and its closeness to the adjusted R2 (0. b.6 0. 3D plot. Contour plot.3. respectively. This was also confirmed by inspection of the residue distribution by a normal probability plot (not reported here for brevity) .
SS 1828 81. Graphical representation of the FD models for PCL‐g‐GMA and PCL‐g‐DEM are given in Figure 6. the interaction between (PCL‐monomer) intake and solubility parameter difference ( nmδ ) and the interaction between the latter factor and the initiator intake ( ni δ ). Graphical representation of the FD model for PCL‐g‐GMA.358 DF 4 8 12 MS 456. The R2 value for the model (0.2 10
25 30 35 40 GMA Intakes (mol% on CL units)
b. The very low P‐value implies that the model is statistically significant.4 BPO Intake (mol% on CL units) 0.
BPO Intake (mol% on CL units) 0. a. the crystallization temperature and enthalpy are not a clear function of the FD values. This induces irregularities and is expected to result in a lowering of the Tc and the Tm.2 10
25 30 35 40 DEM Intake (mol% on CL units)
a.6 0. radical initiator and PCL intakes). and a favored nucleation
. Similar observations have also been made by Kim. These differences in dependency of the FD values on the DEM and GMA intakes may be explained on the basis of the different grafting mechanisms as previously discussed. although both properties are significantly lower than those of pure PCL. the statistical model provides besides a reliable description of the experimental data also a good mathematical description to modulate the FD values of the two substrates by changing the chemical composition (monomer. et al  working with PCL‐g‐GMA.2 10 20 DEM Intake (mol% on CL units) 40 30 50
0.4 0. Contour plot.
Figure 6. and 6. For relatively high DEM intakes (> 30 % mol/mol).7
6 FD (%) 4
0.8 0. From a more practical point of view.6. 3D plot.3
0.1.3. The random behavior of the Tc as function of the FD values is probably the result of two concurring effects: the presence of irregularities. Such changes in the thermal properties compared to pure PCL may be caused by the introduction of grafts on the PCL chains.) of the compatibilizers were determined by DSC.
2 0 1 0.8 0. For all samples. a higher BPO intake leads to a reduction of the FD values while an opposite trend is observed at lower DEM intakes. for DEM a clear transition is observed. 6.5 0. which is expected to lead to a Tc reduction.4 BPO Intake (mol% on CL units) 0. the melting temperature and the relative enthalpy decrease with respect to pure PCL.1. Furthermore. While for GMA an increase in BPO or monomer intakes invariably leds to higher FD values. Remarkable is once more the different trends for the two substrates. Graphical representation of the FD model for PCL‐g‐DEM.Chapter 6
1 0.2.4. Thermal properties of the compatibilizers The thermal properties (Table 6.
PCL‐g‐GMA has an FD of 9.2 342.2 9.9 424.0 401.6 168.9 368.
. A total of 12 blends were prepared: a series of binary ones (not containing any compatibilizers) constituting our reference points.1 10. An overview of all prepared blends together with their thermal and mechanical (tensile tests) properties is given in Table 6.4.3
Tc (oC) 35 36 36 36 30 30 30 29 31 30 29 28 29
∆Hc (J/gPCL) 55 69 51 40 52 54 55 53 58 57 63 54 55
T m (oC) 57 57 57 57 57 57 57 57 58 57 57 58 56
∆Hm (J/gPCL) 62 68 53 43 46 50 52 51 54 50 60 51 48
PCL‐g‐DEM has an FD of 1.0 341.4 230.6 305.2 321.8 384. a series with PCL‐g‐DEM (different intakes of the latter at fixed starch/PCL ratio) and two series with PCL‐g‐GMA (one with different intakes of PCL‐g‐GMA at a fixed starch/PCL ratio and one with a fixed compatibilizer intake at three different starch/PCL ratios).0 332.9
E (MPa) 270.3. Thermal and mechanical properties of starch/PCL blends a
Sample PCL S/PCL 10/90 S/PCL 20/80 S/PCL 30/70 S/PCL/PCL‐g‐DEM 20/80/1 S/PCL/PCL‐g‐DEM 20/80/2 S/PCL/PCL‐g‐DEM 20/80/5 S/PCL/PCL‐g‐GMA 20/80/1 S/PCL/PCL‐g‐GMA 20/80/2 S/PCL/PCL‐g‐GMA 20/80/5 S/PCL/PCL‐g‐GMA 20/80/10 S/PCL/PCL‐g‐GMA 10/90/2 S/PCL/PCL‐g‐GMA 30/70/2
σ (MPa) 16.8 371.0 430.5 7.2 337.6%.7 %.2 343.5 357. The lower crystallization and melting enthalpy of the compatibilizers compared to virgin PCL is possibly caused by the presence of side chains on the PCL backbone which are expected to have a negative effect on the crystallinity of the products. 6.5 489. Synthesis and Properties of Starch‐ PCL Blends The two compatibilizers (PCL‐g‐GMA and PCL‐g‐DEM) were further used as interfacial agents in blends of PCL with starch. Synthesis and Properties of Reactive Interfacial Agents for Polycaprolactone‐Starch Blends
of the PCL chains in the compatibilizers compared to virgin PCL (higher Tc) induced by the presence of polar groups.6 386.
Our experimental design allows comparisons of thermal and mechanical properties as function of the starch content for binary blends (no compatibilizer) and intake of PCL‐g‐DEM and of PCL‐g‐GMA.7 10.5 4126.96.36.199 372.5
ε (%) 640.0 9.3 15.1 13.9 11.6 5.1 11.3 10. Table 6.3 380.4 431.6 11.9 379.8 329.
75 70 65 60 55 50 45 40 35
∆ H (J/ g-PCL)
10 15 20 Starch content (%)
Figure 6.1 MPa and 341.5 % for pure PCL to 7.8 MPa) clearly indicates that the rigidity of the blends increases at higher starch contents. The two trends are remarkably mirroring each other and are both characterized by an increase of ∆Hm and ∆Hc with respect to pure PCL at 10 % starch content followed by a monotonous decrease of both quantities as function of the starch intake. respectively from 16.Chapter 6
6. Melting and crystallization enthalpy as function of the starch content for binary blends with PCL.3.7.8 % at 30 % starch content. Binary blends of starch and PCL For binary blends (no compatibilizer is added) a monotonous decrease of the stress and strain at break is observed as function of the starch intake. This is in agreement with the data previously reported on PCL/starch blends and related to the lack of interfacial adhesion between the starch particles and the PCL matrix . This is most probably due to a nucleation effect of the starch on PCL as observed also for other kinds of polymer blends . The simultaneous increase in the modulus (from 270.2.43 MPa and 640.2 MPa up to 341. The thermal and mechanical properties indicate no or very little interaction of the starch particles with the PCL ○ : ∆Hc □ : ∆Hm
.1.7. The thermal behavior is characterized by no significant changes in the Tm and Tc values but unreported trends of the corresponding enthalpies as function of the starch content (Figure 6.
Such behavior is consistent with the hypothesis that the starch particles hinder the melting and crystallization processes of the PCL chains. either by heating or in combination with mechanical stress . (a) S/PCL 10/90 (b) S/PCL 20/80 (c) S/PCL 30/70 Indeed starch particles are clearly dispersed in the PCL matrix but no interaction (adhesion) between the two phases is actually detected. This is confirmed by examination of the blend morphology by SEM (Figure 6. The thermal behavior is characterized (Table 6. in order to fully understand the role of PCL‐g‐DEM. Evaporation of the water during blend preparation. The presence of these voids was also observed for sago starch/PCL blends [25.9.e. As consequence one might expect a slight different structure of these ternary blends as compared to the corresponding binary one.2. for which a 6 oC drop is observed when using PCL‐g‐DEM.3.
a.8. Ternary blends compatibilized with PCL‐g‐DEM PCL‐g‐DEM.4. the starch particles are simply inserted into voids of the PCL matrix.) by constant values of the Tm (with respect to pure PCL) while the Tc is slightly lower (30 vs 35 oC) than that of pure PCL and it is independent of the compatibilizer content.
. However. and the lack of interface adhesion cause void formation.2. a decrease with respect to pure PCL is observed and then a substantial invariance as function of the PCL‐g‐DEM content. This is visually confirmed by the morphology of the prepared blends (Figure 6. was used as compatibilizer for the preparation of ternary blends with starch and PCL. 6. Thus.
b.).). In this respect the thermal properties remain substantially unchanged with the exception of the Tc. 26] and explained by assuming that the voids are formed by water in the starch.8.
Figure 6. Synthesis and Properties of Reactive Interfacial Agents for Polycaprolactone‐Starch Blends
matrix. whose synthesis is described for the first time in this work. The latter trend is also valid for the corresponding enthalpies: i. Morphology of starch/PCL binary blends.
c. the corresponding binary blend (S/PCL 20/80) contitutes a better reference point compared to pure PCL.
This can be explained by the improved interfacial adhesion  between PCL and starch. A closer inspection of the SEM pictures reveals that the starch particles are clearly embedded in the PCL matrix with almost no void spaces at the interface.). Indeed starch particles in the ternary blends display a smoother interface with the PCL matrix compared to the corresponding binary blends (Figure 6.10.). (a) S/PCL/PCL‐g‐DEM 20/80/1 (b) S/PCL/PCL‐g‐DEM 20/80/2 (c) S/PCL/PCL‐g‐DEM 20/80/5
. At 1 %‐wt of PCL‐g‐DEM in the blends the modulus increases with respect to the binary blend (0 %‐wt compatibilizer in Figure 6. resulting in more rigid material with higher modulus.
Figure 6. At higher PCL‐g‐DEM intakes the lower average molecular weight of the compatibilizer as compared to the one of the unmodified PCL used in the blends (3000 vs 50000) is probably responsible for the observed decrease in the E values.9.
c. The stress and strain at break remain constant as function of the compatibilizer intake while the modulus display an optimum as function of the compatibilizer intake (Figure 6. Morphology of ternary blends compatibilized with PCL‐g‐DEM. As a result of the different morphology also the mechanical behavior display differences with the virgin PCL and the binary blend.10.Chapter 6
a. which will hinder the flowability and fibre forming ability of PCL matrix under cold drawing.).9.
the stress and strain at break are lowered and hardly a function of the PCL‐g‐GMA amount.3. The first comparison as function of the PCL‐g‐GMA intake (ternary blends with starch/PCL ratio of 20/80) results in quite similar considerations as the ones made for PCL‐g‐DEM. Synthesis and Properties of Reactive Interfacial Agents for Polycaprolactone‐Starch Blends
390 380 370 360 E (MPa) 350 340 330 320 0 1 2 3 PCL-g-DEM amount (wt%) 4 5
Figure 6. Ternary blends compatibilized with PCL‐g‐GMA When using PCL‐g‐GMA as compatibilizer two possible comparisons can actually be made: one at fixed starch/PCL ratio and changing the amount of PCL‐ g‐GMA and one at different starch/PCL ratios but with fixed amount of PCL‐g‐ GMA (2 %‐wt).e. 6.
2) 3) 4)
Remarkable is that in the case of PCL‐g‐GMA no optimum is found in the modulus as function of the compatibilizer intake.2. at all PCL‐g‐GMA contents) there is a clear increase of the modulus with respect to the
. the crystallization temperature experiences a drop of about 5‐6 oC. it can be observed that: 1) the melting temperature as well as the crystallization and melting enthalpies do not change significantly (discrepancies in the values are within the boundary defined by the experimental error). However. Modulus of ternary blends S/PCL/PCL‐g‐DEM as function of PCL‐g‐ DEM amount. by taking as reference the corresponding binary blend (S/PCL 20/80).10.3. the modulus is higher and hardly a function of the compatibilizer amount. In particular also in the case of PCL‐g‐GMA. in all cases (i.
(left) the structure of PCL‐g‐GMA is inhomogeneous at the molecular level with long poly(GMA) branches pending from the PCL backbone.). Thus.
Figure 6. As a result.4.7 % for PCL‐g‐DEM. partial interfacial adhesion) which render these ternary blends a kind of “intermediate” case between the binary ones and those compatibilized with PCL‐g‐DEM. This confines all reactive GMA groups in relatively “isolated” spots along the PCL backbone. presence of voids. upon reaction of PCL‐g‐GMA with one of the –OH groups on the surface of the starch particles the remaining GMA groups are actually scarcely available for further reaction with other –OH groups spatially distant from the reacted one. As given schematically in Figure 6. 29].
c.) than PCL‐g‐GMA for blends containing 20 %‐wt of starch. clearly indicating that also PCL‐g‐GMA (like PCL‐g‐DEM) acts as compatibilizer in the blends Such hypothesis is partially and qualitatively confirmed by the blends morphology (Figure 6.
b.12. This is surprising if one takes into account the different FD values: 9.11. despite a much more favourable FD value for PCL‐g‐GMA with respect to PCL‐g‐ DEM and even despite a relative low reactivity of ester groups towards the –OH groups on starch and nucleophilic groups in general [27. (a) S/PCL/PCL‐g‐GMA 20/80/1 (b) S/PCL/PCL‐g‐GMA 20/80/2 (c) S/PCL/PCL‐g‐GMA 20/80/3 In this case we observe structural features (partially smooth interface between the components. PCL‐g‐DEM is at least as active as compatibilizer (compare modulus values in Table 6.11.8 % for PCL‐g‐GMA against 1.9. 28. one might speculate that the longer length of the GMA grafts compared to DEM (vide supra) plays a negative role in the in situ formation of the compatibilizer. and morphologies in Figures 6. and 6. Morphology of ternary blends S/PCL/PCL‐g‐GMA. These differences in compatibilizing effect are not yet fully understood.
.11. However.Chapter 6
at higher starch contents more –OH groups would be available for reaction with the GMA groups. Concerning the thermal behavior (Table 6.).12. the Tc decreases (with respect to the corresponding binary blends) while all other factors (Tm and enthalpies) remain virtually unchanged with respect to the binary blends and also as function of the starch intake.4. The proposed explanation implies however that at relatively higher starch contents (> 20 %‐wt) the segregation of the poly(GMA) chains in “isolated” spot along the PCL backbone would become less important. As consequence.12.12. Moreover. The latter result (from a modulus of 341. when comparing ternary blends (compatibilized with PCL‐g‐ GMA) with the binary ones. Synthesis and Properties of Reactive Interfacial Agents for Polycaprolactone‐Starch Blends
PCL-g-DEM PCL-g-GMA OH O OH O
Figure 6. Schematic representation of the in situ reaction between functionalized PCL and starch.) for further reaction. for all starch amounts (Table 6. thus attenuating the effect discussed above. On the other hand (right of Figure 6. Indeed. but since DEM is preferentially grafted as monomer.3 MPa for S/PCL/PCL‐g‐GMA 30/70/2) clearly indicates that the efficiency of PCL‐g‐GMA as compatibilizer becomes more relevant at relatively higher starch contents (>30 %) .) PCL‐g‐DEM has much less reactive groups (lower FD) than PCL‐g‐GMA. We can therefore conclude that the efficiency of PCL‐g‐GMA in the
. other groups will presumably remain available (arrows in Figure 6.) and relating the “coverage” of the starch particle surface upon reaction with PCL‐g‐GMA with the compatibilization efficiency.). This is in agreement with the hypothesis made above (Figure 6. the stress and strain at break decrease while the modulus is unchanged at low starch contents and increases significantly for S/PCL blends with 30 %‐wt starch. the distribution of the reactive group along the PCL backbone is more “homogenous”. once PCL‐g‐DEM has reacted with one –OH groups on the starch particle. 20 and 30 %‐wt respectively). thus probably ensuring a better coverage of the surface. In order to check this we compared blends with the same amount of PCL‐g‐GMA (2 %‐wt) but with different starch intakes (10.12.8 MPa for S/PCL 30/70 up to 430.4.
has been performed on the synthesis of two reactive compatibilisers.4.5. Conclusions
A systematic study. including statistical modeling. PCL‐g‐glycidyl methacrylate (PCL‐g‐ GMA) and PCL‐g‐diethyl maleate (PCL‐g‐DEM). mutual solubility of the monomer into the molten polymer) on the FD adequately described the experimental results (R2=0. moles of GMA or DEM present per mole of CL repeating units
amount of initiator intake [mol% on CL units] amount of monomer GMA or DEM intake [mol% on CL units]
. P‐value ≤10‐9). Nomenclature
E FD : initial modulus [MPa] : functionalization degree.
6. At a fixed starch content (20 %‐wt) PCL‐g‐DEM seems to have sligthly more efficient compatibilizing effect than PCL‐g‐GMA as shown by blends morphology and elastic modulus values. The proposed model to quantify the effects of process variables (monomer and initiator intake. The latter hypothesis is indirectly confirmed by the observation that PCL‐g‐GMA becomes more efficient at relatively higher starch contents in the blends. can replace PCL‐g‐GMA as compatibilizer at relatively low starch contents offering at the same time the advantage of a less pronounced modification of the polymer backbone and a reduced consumtpion of polar groups to be grafted on PCL.Chapter 6
compatibilization of starch/PCL blends can be significantly improved by changing the blend composition. From a more practical point of view it can be concluded that the newly synthesized PCL‐g‐DEM. This represents an unreported mathematical confirmation of the fact that these kinds of functionalization reactions are mainly governed by the synergy between the different process variables and only slightly by their individual values. This is in contrast with chemical reactivity and amount of chemical groups along the PCL backbone (both factors favorable to GMA as compared to DEM) but it is explainable on the basis of the group distribution along the PCL backbone.957. firstly reported in this work. The reactive compatibilizers were used in blends of starch with PCL. The most important variable for the final product FD is the interaction between the amount of monomer and initiator used. The PCL‐g‐GMA and PCL‐g‐DEM compatibilizers display lower melting temperatures and melting enthalpies compared to virgin PCL.
R. Polym.D. . J. 1998. W. Polym. 1997. Carbohydr. M. Jarowenko: Acetylated starch and miscellanous organic esters. E. Bhattacharya: Properties of injection moulded starch/ synthetic polymer blends – IV. Boca Raton. Sci. Psomiadou. 44. Adv. Pacsu: Starch Studies: Preparation and properties of starch triesters. Ogawa. 227‐242. 1986. R. Biliaderis. N. J. Ioannis. M. Karlsson. Inc. Bhattacharya: Properties of injection moulded starch/ synthetic polymer blends – III. Ranucci. S. Polym. . USA. O. Ind. Polymer 2003. S. . 1998. F. CRC Press. Ramsay. Thermal and morphological properties. 1251‐1253. wheat starch and soluble starch for food packaging applications. 157.W. 1998. Eng. Mani. 1477‐1487. References
. 34. Synthesis and Properties of Reactive Interfacial Agents for Polycaprolactone‐Starch Blends
: temperature [oC] : crystallization temperature [oC] : melting temperature [oC]
: enthalpy of crytallization [J/g‐PCL] : enthalpy of melting [J/g‐PCL] : solubility parameter [cal1/2cm‐3/2] : stress at break [MPa] : strain at break [%]
6. 34. Mani.A. A. C. 1209‐1217. 1517‐1526. E. M.6. Part 2. Sci. . Albertsson: Biodegradable Polymers and Environmental Interaction. Mullen. Eur. 38. B.Chem. . A‐C. Y.Eng. B. 1467‐1475. E. Davis: High performance LDPE/ thermoplastic starch blend: a sustainable alternative to pure polyethylene. Effect of amylopectin to amylase ratio in starch. 2002. 34. Albertsson. Eur. Lindblad. .
. Kawasaki: Biodegradable films made from low density polyethylene (LDPE). 1942. J.G.S.J. A‐C. Wurzburg). Polym. Rodriguez‐Gonzales. H. Polym. . Karlsson: Polymers from Renewable Resources.B.. 33. Liu. in Modified Starches: Properties and Uses (Ed. 139‐161.
W. Pennsylvania.Y. Polym. J. K. 1507‐ 1516. 91‐95. Degrad. Chandra. Passaglia. Yang. 2003. D.Chapter 6
. Bhattacharya. Macosko. R. Environ. 41. Dole. Kim. Solaro: Biodegradable polymeric materials. J. Cho. J. . Polym. . Averous. Mater. R. M. M. Inoue: Compatibilizers for melt blending: Premade block copolymers. K. 2001.E. Moro. J. Sci. 1996. M. . Montgomery: Design and Analysis of Experiments 5th edition. 81. 5590‐5598. . C. Polym. USA. Bhattacharya: Synthesis and characterization of anhydride‐functional polycaprolactone. A.F. Marrucci.C. R. Tang: Functionalization of polyesters with maleic anhydride by reactive extrusion. D. C. Z. Prog. Mani. 179‐187. C. Cho. . P. 127. Laurienzo. Gazz. Degrad. Marechal. Macromolecules 1996.S. Park: Reactive blends of gelatinized starch and polycaprolactone‐g‐glycidyl methacrylate. Jung. G. Chim. P. C. .K. J. J. Chem. Martuscelli. 80. C. Kim. Sci. . Goldberg: A review of the biodegradability and utility of poly(caprolactone). 2004. Rustgi: Biodegradable polymers. 3. Chem. Technomic Publication. 1997. 127–134. . Polym. T.. Park: Grafting of glycidyl methacrylate onto polycaprolactone: preparation and characterization. Polym. 1693‐1702. E.M. It. Sci. Environ. John Wiley & Sons Inc. Polymer 2000. Wu: Physical properties and biodegradability of maleated‐ polycaprolactone/ starch composite. . John. 1991. R. 4157‐4167. Polym. J. Painter: Specific Interactions and the Miscibility of Polymer Blends. 305‐313. . Coleman. Kim. Avella. Errico. P. Part A: Polym. USA. J. . Stab. M. J.K. 3875‐3881. . 35. J. Polym.S. Guegan.K. 5135‐ 5142. 1139‐1148. Chiellini.K.. M. 61‐67.
.H. 1998.H. 12. New York. . Polymer 2001.Y. 8. . R. 1273‐1335.. Sci. Aglietto: Chemical reactions affecting the free‐radical grafting of diethyl maleate onto ethylene polymers. Park: Modification of Aliphatic Polyesters and Their Reactive Blends with Starch. 42. Raimo. Part A: Polym. Kim. Tang. L. Polymer 2000. E. J. 1997. M. C. 29. A.M. P. K. P. Appl. Nakayama. 1995. Ruggeri. C. 2001. J. Rimedio: Preparation and characterization of compatibilised polycaprolactone/ starch composites. M. 1999. 23. Adv. 41. L. Graf. Khandpur. E. J. Fringant: Properties of thermoplastic blends: starch‐polycaprolactone.H. 37. M.
K. O.S. E. Polym.S. Polym. R. D. D. Polym. Manan. A.N. 2000.S.W. 1298‐1305. . 57. . 40. Odusanya. M. K. Ruggeri. Appl. Pang.W.S. Azemi.Nasirtabrizi: Ring opening reactions of glycidyl methacrylate copolymers to introduce bulky organosilicon side chain substituents.D. S. U.M.S. G. Passaglia. Eur. 38. S.S.H. Technol. Azemi: Effect of starch predrying on the mechanical properties of starch/ poly(ε‐caprolactone) composites. Polym. Polym.N. Lee.A. J. 877‐884. U. 293‐304. 21.A. .V. Ishiaku. 1998. 2002. Barbosa. W. Nachtigall. J. Adv. 393‐401.M. 11‐15. 2006. Test. M. Mauler: Comparative study of PP‐MA and PP‐GMA as compatibilizing agent on polypropylene/nylon 6 blends. 9. 87. 2003. Polym. B. Aglietto. Kammer: On mechanical properties of sago starch/ poly(ε‐caprolactone) composites. Tedesco. Picchioni: Formation and compatibilizing effect of the grafted copolymer in the reactive blending of 2‐ diethylsuccinate containing polyolefins with poly‐ε‐caprolactam (nylon‐6). 2002. Bull. . B. . H. F. Sci. Sci. R. U. Manan.M.M. Ishiaku. O. Ishak: Mechanical properties and enzymic degradation of thermoplastic and granular sago starch filled poly(ε‐ caprolactone).M. Eng.Safa. M.B. 273‐281. Ishiaku.
. Odusanya. Synthesis and Properties of Reactive Interfacial Agents for Polycaprolactone‐Starch Blends
Starch‐based biodegradable materials are considered interesting candidates to replace certain types of conventional plastics. chemical modification or combinations thereof. Amylose is essentially a linear polymer in which AHG units are predominantly connected through α‐D‐(1.o oil and gas) are very attractive materials for a broad range of applications. There are several starch modification methods available. There are two types of polymers present in starch: amylose and amylopectin. Starch must therefore be modified before it can be applied as a biodegradable plastic. The content of amylose and amylopectine in starch varies and largely depends on the starch source. The application of a number of prospective methods for developing starch‐based biomaterials will be described in this thesis. The use of virgin starch for packaging materials is restricted because it cannot be shaped to films with adequate mechanical properties (high percentage elongation. while amylopectin is branched polymer.4)‐glucosidic bonds. Its production has increased significantly since 1950. plastics may end up in the environment and cause serious environmental problems. The main reasons are the water sensitivity of common catalysts and the fact that starch is poorly or even insoluble in the common organic solvents used for ROP. Examples are the use as packaging and construction materials. it has a high volume to weight ratio and is generally resistant to degradation. such as thermoplasticization. blending with other materials. we have
. Starch is also too sensitive to water. there is a strong need for renewable alternatives for plastics from fossil resources. After use.6)‐glucosidic bonds. with a rate of almost 10% every year. containing periodic branches linked with the backbones through α‐D‐(1. The application of an alternative method involving hydrophobisation of starch by the introduction of large hydrophobic SiMe3 groups followed by a ROP with a polyester precursor and subsequent removal of the SiMe3 groups is expected to result in higher grafting efficiencies. For instance. Starch is a polymer consisting of anhydroglucose (AHG) units. tensile and flexural strength). Starch is relatively cheap and available from a broad range of plants.Summary
Plastics made from fossil resources (a. Chemical grafting of biodegradable polyesters on the starch backbone is expected to result in less hydrophilic and thus less water sensitive materials with improved mechanical properties. To gain insight in the potential of this approach. The synthesis of such grafted products by the in situ ring opening polymerization (ROP) of the monomers on the hydroxyl groups of starch is unfortunately not very straightforward. In combination with the current high prices for petrochemical products.
The grafting efficiency for the desired reaction was 28‐58%.2.2.4‐di‐O‐isopropylidene‐α‐D‐galactopyranose ratio (at constant Al(OiPr)3 intake). but the products are still very brittle. The grafting of ε‐caprolactone to the silylated starch by a ring‐opening polymerisation catalysed by Al(OiPr)3 was performed in THF at 50oC. The synthetic procedure may be divided into three steps: hydrophobisation of starch by the introduction of SiMe3 groups followed by in situ ROP. The polymerizations were performed at 60‐100°C.68) was obtained. even in the presence of plasticizers. and off‐white solid products with isolated yields of 30‐96% were obtained. Characterization of the products using 1H‐ and 13C‐NMR as well as MALDI‐TOF mass spectrometry showed the presence of significant amounts of p‐dioxanone polymers with an isopropoxide end group (20‐30%) The knowledge obtained from the model system was applied for the synthesis of starch‐g‐poly‐ε‐caprolactone using hydrophobised silylated starch.4‐di‐O‐isopropylidene‐α‐D‐ galactopyranose) using Al(OiPr)3 as the catalyst. The DS of the PCL chains was between 0. Silylated starch with a low to medium DS (0.21‐0.46‐0.72.Summary
initially performed research on the ROP with a simpler model system involving a simple protected mono‐saccharide instead of starch. Average molecular weights of the products were between 970 and 6200 (7 – 58 monomer units) and were a clear function of the p‐ dioxanone/ 1. A statistical model has been developed to quantify the effects of process variables (time. The results are described in detail in Chapter 3. The yield of the polymers was a function of the reaction temperature and the reaction time. are described in Chapter 2. The silylation reaction was performed using hexamethyl disilazane (HMDS) in DMSO/ toluene mixtures at 70°C.3. with higher temperatures (100°C) leading to lower yields. The silyl groups of the poly‐(ε)‐caprolactone grafted starch co‐polymers were finally successfully removed using a mild dilute hydrochloric acid treatment in THF at room temperature. Esterification of starch with carboxylic acid derivatives is one of the oldest strategies to improve starch properties. and subsequent removal of the silyl groups. The hydrophobicity of starch acetates is higher than virgin starch. Most of the previous starch esterification studies involved the use of short chain carboxylic acids (C1‐C4).3. Poly‐(ε)‐ caprolactone grafted silylated starch co‐polymers with average chain length of 40‐ 55 monomer units (molecular weight of 4500‐6300) were obtained. Considerable amounts of ε‐CL homopolymers with isopropyl end‐groups were also formed. The results for the model system involving the ROP of p‐dioxanone initiated by hydroxyl groups of a protected monosaccharide (1. depending on the ε‐CL to starch ratio. temperature and monomer: monosaccharide ratio) on the average degree of polymerization. The use of higher carboxylic (fatty) acid to esterify starch resulted in products with significantly improved
. with higher ratios leading to higher molecular weights. and particularly acetic acid derivatives (C2) have received considerable attention.
The thermal stability of the esterified products is higher than that of native starch. the synthesis of two polycaprolactone (PCL)‐based reactive interfacial agents.96). The synthesis of these fatty starch esters is. For low vinyl‐ester to starch ratios. High‐DS products (DS= 2. while the use of K2CO3 and CH3COONa catalysts resulted in medium‐high DS products (2. presumably due to a reduction of the polarity of the reaction medium. The starch esters were synthesized by reacting gelatinized starch with vinyl laurate or vinyl acetate in DMSO in the presence of basic catalysts (Na2HPO4.75) are still partially crystalline. With these catalysts.07‐1. The use of Na2HPO4 resulted in low‐medium DS products (0.4‐3 for starch stearate). The products have a broad range of degree of substitution (DS = 0.4 could be obtained for both laurate and stearate esters. usually performed using fatty acid chlorides which are relatively expensive and rather corrosive. an alternative method for higher fatty starch esters by using fatty vinyl esters is reported. In Chapter 5. vinyl ester to starch ratio.3‐1. elongation at break of 3‐26%. PCL‐g‐
.7‐3.26‐2. The last part of this thesis deals with reactive blending of starch and biodegradable polymers using reactive interfacial agents. and Na‐ acetate).5 for starch stearate). The thermal and mechanical properties of some representative product samples are also described.24 ‐ 2. Summary
mechanical properties and hydrophobicities. 0.1‐2.9 for starch laurate. products with a DS > 2. Chapter 4 describes the results of a preliminary study on the synthesis of long fatty esters of corn starch (starch‐laurate and starch‐stearate) using vinyl fatty esters. Reaction temperature and catalyst intake also affect the product DS but only to a lesser extent. however. and the type of catalyst. and Na‐acetate) in DMSO at 110°C. At higher ratios. the DS decreased. 1. 13C‐NMR and FT‐IR. Statistically adequate (R2≥0. Mechanical tests show that the products have tensile strength (stress at break) between 2. an increase in the vinyl‐ester concentration led to higher product DS values. The use of methyl and glyceryl esters results in products with only relatively low DS. In Chapter 4‐5. The DS of the products is a strong function of the basicity of the catalyst. K2CO3 and Na‐acetate are superior with respect to activity when compared with Na2HPO4.45‐1. The starch esters were prepared by reacting starch with vinyl laurate or vinyl stearate in the presence of basic catalysts (Na2HPO4. including statistical modelling on the synthesis of long fatty esters of corn starch (starch‐laurate and starch‐stearate) using corresponding vinyl fatty esters is reported. a systematic study.5 for starch laurate.39) are totally amorphous whereas the low‐DS ones (DS= 1. The DS of the products was a clear function of the chain length of the fatty ester.5 MPa. reaction temperature. In Chapter 6. and catalyst basicity) to the DS of the starch ester products were developed. and modulus of elasticity of 46‐113 MPa. K2CO3.96 and P‐value of ≤10‐7) second‐order mathematical models correlating the effect of process variables (vinyl to AHG‐ starch mol ratio. K2CO3. The yellowish products were characterized by 1H‐. catalyst intake.
Highest values of the FD (up to 45 %) were observed for PCL‐g‐GMA. The PCL‐based compatibilisers were prepared by reacting a low molecular weight PCL (Mw 3000) with glycidyl methacrylate or diethyl maleate in the presence of benzyolperoxide initiator at 130°C. PCL‐g‐DEM seems to have slightly more efficient compatibilising effect than PCL‐g‐GMA as shown by blends morphology and elastic modulus values.
.957. Both of the PCL‐based compatibilisers improve the interfacial adhesion between the starch particles and the PCL matrix in starch/ PCL blends. A statistically adequate model (R2=0. This is in contrast with chemical reactivity and amount of chemical groups along the PCL backbone (both factors favorable to GMA as compared to DEM) but it is explainable on the basis of the group distribution along the PCL backbone. In particular the elastic modulus for the compatibilised blends is significantly higher (reaching up to 430 MPa) than that of uncompatibilised ones (320‐340 MPa). the mechanical behavior of the compatibilised blends is in general different from that of pure PCL and of the corresponding uncompatibilised blends. The FD values for PCL‐g‐DEM were considerably lower (up to 7 %).Summary
glycidyl methacrylate (PCL‐GMA) and PCL‐g‐diethyl maleate (PCL‐g‐DEM). The latter hypothesis is indirectly confirmed by the observation that PCL‐g‐GMA becomes more efficient at relatively higher starch contents in the blends. P‐value ≤10‐9) has been developed to quantify the effects of process variables (monomer and initiator intake. mutual solubility of the monomer into the molten polymer) on the functionalisation degree (FD) of the GMA and DEM units to the PCL backbone. As the result. and their use in compatibilising starch‐PCL blends is described. at relatively high GMA and BPO intakes. At a fixed starch content (20 %‐wt).
Dit. heeft geleid tot een grote interesse voor biologisch afbreekbare alternatieven gemaakt uit hernieuwbare grondstoffen.6)‐glucoside bindingen met de hoofdketen zijn verbonden. Er zijn verschillende methodes beschikbaar om zetmeel te modificeren. menging met andere materialen. Het zetmeel is daarnaast ook erg gevoelig voor water.Samenvatting
Kunststoffen gemaakt van fossiele grondstoffen als olie en gas zijn zeer aantrekkelijke materialen voor een breed scala aan toepassingen. chemische modificaties en combinaties van deze methoden. Het chemisch grafting van biologisch afbreekbare polyesters op de zetmeel ketens zou kunnen leiden tot minder hydrofiele en dus minder water gevoelige materialen met betere mechanische eigenschappen. mede omdat het niet mogelijk is om films met goede mechanische eigenschappen te maken. Na gebruik kunnen de materialen in het milieu eindigen en grote problemen veroorzaken. De toepassing van een alternatieve methode waarbij zetmeel eerst minder hydrofiel gemaakt wordt door de introductie van grote hydrofobe SiMe3 groepen. Dit proefschrift beschrijft een aantal methodes om nieuwe polymere materialen op basis van zetmeel te maken. ε‐caprolactone op de hydroxy‐groepen van zetmeel is helaas niet eenvoudig. Het amylopectine is een vertakt polymeer waarbij de vertakkingen via α‐D‐(1. Twee type polymeren zijn dominant aanwezig in zetmeel: amylose en amylopectine. Biologisch afbreekbare polymeren op basis van zetmeel kunnen mogelijk bepaalde conventionele kunststoffen vervangen.4)‐glucoside bindingen verbonden zijn. De AHG eenheden kunnen op een aantal manieren aan elkaar gekoppeld worden. Voorbeelden zijn thermo‐ plastificatie. gevolgd door de ROP met een polyester precursor en verwijdering van de SiMe3 groepen
. in combinatie met de huidige hoge prijzen voor de grondstoffen. De verhouding amylose en amylopectine in zetmeel varieert en hangt grotendeels af van de plant waaruit het zetmeel gewonnen wordt. De synthese van dergelijke producten met een in situ katalytische ringopening polymerisatie (ROP) van geschikte monomeren als b. Amylose is een lineair polymeer waarin de AHG eenheden hoofdzakelijk via α‐D‐(1. De belangrijkste redenen zijn de watergevoeligheid van de katalysatoren en het feit dat zetmeel slecht of zelfs onoplosbaar is in de typische organische oplosmiddelen die voor de ROP gebruikt worden.v. Modificatie van zetmeel kan het aantal toepassingen sterk vergroten. Zetmeel is relatief goedkoop en te winnen uit een breed scala aan planten. Zetmeel is een polymeer bestaande uit anhydroglucose (AHG) eenheden. De productie van kunststoffen is sinds 1950 met bijna 10% per jaar gestegen. Het gebruik van zetmeel voor verpakking materialen is momenteel beperkt.
De opbrengst van de polymeren was een functie van de reactietemperatuur en de reactietijd. De resultaten voor het modelsysteem waar de ROP van p‐dioxanone wordt geïnitieerd door de hydroxy‐groepen van een beschermde mono‐saccharide (1. De resultaten worden in Hoofdstuk 3 in detail beschreven. De silylerings reactie werd bij 70°C uitgevoerd met hexamethyl disilazane (HMDS) in DMSO/tolueen mengsels. De grafting efficiency van de gewenste reactie was 28‐58%.21‐0. De gevolgde procedure bestaat uit drie stappen: hydrofobisering van het zetmeel door de introductie van SiMe3 groepen.68). De substitutiegraad van de SiMe3 groepen varieerde van laag tot middelhoog (degree of substitution.4‐di‐O‐isopropylidene‐α‐D‐galactopyranose) met Al(OiPr)3 als katalysator worden in Hoofdstuk 2 beschreven. gevolgd door in situ ROP met ε‐ caprolacton en daarna de verwijdering van de silyl groepen. De polymerisaties werden tussen 60‐100°C uitgevoerd en leverde vaste producten in geïsoleerde opbrengsten tussen de 30 en 96%. Karakterisering van de producten met 1 H‐ en 13C‐NMR en MALDI‐TOF massaspectrometrie toonden de aanwezigheid aan van de gewenste producten en daarnaast significante hoeveelheden van p‐ dioxanone polymeren met een isopropoxide eindgroep (20‐30%). De effecten van procesvariabelen (tijd. temperatuur en monomeer: monosaccharide verhouding) op de gemiddelde polymerisatie graad zijn gekwantificeerd met een statistisch model.46‐0.3. Er werden ook aanzienlijke hoeveelheden homopolymeer van ε‐ caprolacton met isopropyl eindgroepen gevormd.2.2.
. Om inzicht in het potentieel van deze benadering te krijgen is onderzoek verricht naar de ROP met een eenvoudig modelsysteem bestaande uit een beschermde mono‐saccharide in plaats van zetmeel. De DS van de PCL ketens lag tussen de 0. De silyl groepen van het poly‐ε‐caprolactone gesilyleerde zetmeel werden uiteindelijk succesvol verwijderd door middel van een milde behandeling met verdund zoutzuur in THF bij kamertemperatuur. DS = 0.72 en was afhankelijk van de ε‐caprolacton/zetmeel verhouding.3. Een hogere verhouding leidde tot hogere molecuul massa’s. De gemiddelde molecuul massa van de producten lag tussen de 970 en 6200 (7 ‐ 58 monomeereenheden) en was een functie van de verhouding p‐ dioxanone/1.4‐di‐O‐isopropylidene‐α‐D‐galactopyranose (bij constante i Al(O Pr)3 concentratie).Samenvatting
zou in theorie moeten leiden tot een hogere grafting efficiency. De polycaprolacton eenheden (PCL) van het verkregen poly‐ε‐caprolactone gesilyleerde zetmeel hadden een gemiddelde ketenlengte van 40‐55 monomeereenheden (molecuul massa 4500‐6300). De kennis verkregen met het modelsysteem is toegepast om ε‐caprolactone te graften op zetmeel. De grafting van ε‐ caprolactone aan het gesilyleerde zetmeel werd in THF bij 50°C uitgevoerd met Al(OiPr)3 als katalysator. waarin hogere temperaturen (100°C) tot lagere opbrengsten leiden.
en in het bijzonder azijnzuur derivaten (C2). De DS van de producten is een sterke functie van de basiciteit van de katalysator. De producten werden met 1H‐.07‐1.5 voor zetmeel stearaat). katalysator concentratie en katalysator basiciteit) op de DS zijn gekwantificeerd met een statistisch model (R2 ≥ 0. Hoofdstuk 4 beschrijft de resultaten van een voorstudie naar de synthese van lange vetzuur esters uit maïszetmeel (zetmeel laureaat en zetmeel stearaat) met de vinyl esters van de vetzuren al reagens. Het gebruik van hogere (vet‐) zuren resulteerde in producten met beduidend betere mechanische eigenschappen en hydrofobiciteit. 13C‐NMR en FT‐IR gekarakteriseerd. In Hoofdstuk 4‐5. De synthese van deze esters wordt in het algemeen uitgevoerd met vetzuurchloriden die vrij duur en sterk corrosief zijn.24 tot 2. K2CO3. Het gebruik van methyl en glyceryl esters resulteert in producten met een lage substitutiegraad. Het gebruik van Na2HPO4 resulteerde in producten met een laag tot middelhoge DS (0. Met deze katalysatoren worden voor zowel de laureaat als stearaat esters producten met een DS van boven de 2. K2CO3 en Na‐acetaat) in DMSO bij 110°C te laten reageren. Samenvatting
De esterificatie van zetmeel met carbonzuur derivaten is één van de oudste strategieën voor het verbeteren van zetmeeleigenschappen. De acetaat esters zijn in het algemeen minder hydrofiel dan zetmeel. De zetmeel esters werden bereid door het zetmeel met vinyl laureaat of vinyl stearaat in aanwezigheid van basische katalysatoren (Na2HPO4.5 voor zetmeel laureaat. Echter de producten zijn nog steeds zeer bros. De DS van de producten was een duidelijke functie van de keten lengte van de vetzure vinylester. zelfs in aanwezigheid van plastificeermiddelen.96 en P‐ waarde van ≤ 10‐7). terwijl het gebruik van de K2CO3 en Na‐acetaat katalysatoren
. In de meeste zetmeel esterificatie studies wordt gebruikt gemaakt van kleine carbonzuren (C1‐C4). vermoedelijk als gevolg van een verlaging van de polariteit van het reactiemengsel. K2CO3 en Na‐acetaat vertonen superieure activiteiten vergeleken met Na2HPO4. De thermische en mechanische eigenschappen van sommige representatieve producten zijn uitgebreid bepaald. De substitutiegraad varieert van 0. De effecten van belangrijke procesvariabelen (vinyl ester/zetmeel molverhouding. inclusief statistische modellering. wordt een systematische studie. bij verdere verhoging van de verhouding verminderde de DS.96. 0. De zetmeel esters werden bereid door zetmeel met vinyl laureaat of vinyl stearate in DMSO in de aanwezigheid van basische katalysatoren (Na2HPO4. naar de synthese van lange vetzuur esters uit maïszetmeel (zetmeel laureaat en zetmeel stearaat. De temperatuur en de katalysator concentratie beïnvloeden de DS ook maar in mindere mate. reactietemperatuur. Echter.3‐1. In Hoofdstuk 5. de vinyl‐ ester/zetmeel verhouding en het type katalysator. wordt een alternatieve methode voor de synthese van zetmeel esters met lange vetzuren beschreven. en Na‐acetaat) te laten reageren.4 verkregen. Bij lage vinyl‐ester/zetmeel verhoudingen leidde een verhoging van de vinyl‐ester concentratie in eerste instantie tot hogere DS waarden.
4‐3. Er is een statistisch model (R2 =0. Dit leidt tot verschillen in mechanische eigenschappen tussen gecompatibiliseerde mengsels en van de niet‐ gecompatibiliseerde mengsels van zetmeel en PCL.
.9 voor zetmeel laureaat.7‐3. De synthese van deze verbindingen wordt uitgebreid beschreven. De hoogste FD waarden (tot 45%) werden gevonden voor PCL‐g‐GMA bij relatief hoge GMA en BPO concentraties. Mechanische tests tonen aan dat de producten een treksterkte hebben variërend van 2. PCL‐g‐glycidyl methacrylaat (PCL‐GMA) en PCL‐g‐diethyl maleaat (PCL‐g‐DEM).957 en P‐ waarde van ≤ 10‐9) ontwikkeld waarmee de effecten van procesvariabelen (monomeer en initiator concentratie.0 voor zetmeel stearaat).45‐1. Het laatste hoofdstuk van dit proefschrift (Hoofdstuk 6) behandelt de menging van zetmeel en biologisch afbreekbare polymeren met reactieve compatibilisers.1‐2.75 nog gedeeltelijk kristallijn zijn. 1. Beide reactieve compatibilisers verbeteren de interactie tussen de zetmeeldeeltjes en de PCL matrix. De producten met een hoge DS (2.39) zijn volledig amorf terwijl de producten met een DS van 1. wederzijdse oplosbaarheid van het monomeer in het gesmolten polymeer) op de functionalisatie graad (FD) van de GMA en DEM eenheden aan de PCL keten adequaat gekwantificeerd kunnen worden. gebruikt.Samenvatting
resulteerde in producten met een middelhoog tot hoge DS waarden (2. Als reactieve compatibilisers zijn twee polycaprolactone derivaten. Zo is bijvoorbeeld de elastische modulus van de gecompatibiliseerde mengsels beduidend hoger (tot 430 MPa) dan die van niet‐gecompatibiliseerde mengsels (320‐340 MPa). De PCL gebaseerde compatibilisers werden bereid door reactie van laag moleculair PCL (Mw 3000) met glycidyl methacrylaat of diethyl maleaat in de aanwezigheid van een benzyolperoxide initiator bij 130°C.5 MPa.26‐2. een uitrekking tot breuk van 3‐26%. De thermische stabiliteit van de geësterificeerde producten is hoger dan die van zetmeel. De FD waarden voor PCL‐g‐DEM waren aanzienlijk lager (tot 7%). en een elasticiteit modulus van 46‐113 MPa.
which improved the quality of this thesis. A. Prof. Francesco Picchioni. together with difficulties in managing plastic waste. for valuable discussions during my early PhD period and for handing me over to the right hands. I want to thank Francesco Picchioni for being always there when I needed him to discuss anything.Acknowledgements
The synthesis of biodegradable materials is currently receiving a lot of attention. for her hospitality and solving administrative matters. Thank you Erik. Broekhuis. Thank you very much for spending your valuable time reading my thesis and for the constructive comments and corrections you have made.A. and Prof. This thesis is a summary of my research on developing novel biodegradable polymers from starch. Erwin Wilbers. and for always encouraging me to finish this thesis. especially on his opinions about the statistical modelling and mechanical properties. for your consistent support. Minnaard. People from other groups have also been very kind and supportive to my research. L. I want to thank Leon Janssen for his hospitality. Prof. and Prof. not in the last place because of the problems related with conventional plastics. who was my first daily supervisor for about two months. I also want to thank Ineke Ganzeveld. Anno Wagenaar and Wim Kruizinga taught me how to operate the NMR machines and to interpret the spectra. since the mechanical properties as well as hydrophobicity of this starting material is relatively poor. A.J. for upgrading my writing quality. as well as taking the SEM pictures. and for helping me arranging many things when I was not able to do it from Indonesia. Prof. Mosiscki. I found out that developing a better fully‐biodegradable‐material from starch is not an easy job. Marcel de Vries. I also want to give my deep appreciation to my reading committee. Prof. Anne Appeldoorn. Leon Janssen. I want to thank Erik Heeres for being more than a good supervisor. Gert Alberda van Ekenstein and Harry Nijland helped me a lot with mechanical and thermal property measurements. I want to thank the persons of Building 18 in Nijenborgh 4 for making me feel at home in Groningen and for their help during my research. Anno was also always there when
. I want to thank Marya van der Duin‐de Jonge. First of all. Erik Heeres. the raw material for fossil derived plastics. Although it is indeed a very interesting topic. He has also been a good friend and always supported me during the hard times. I want to thank my three promoters. These include the current high oil prices. for his efforts to make every foreign students feel at home in Groningen. and Laurens Bosgra for being always helpful with technical issues and Jan Henk Marsman for the help with analytical equipment. for the inspiring ideas during the discussions.
for everything. Thanks also for Marcel and Connie. I want to thank Edwin A. C. Ronny Sutarto. master students who worked with me. R. and many others for the good old times! Thank you Tante Ietje. Niels van Vegten. Hofman. Nidal Hammoud Hassan. our memories will never fade away. and assistant professors have shared the days with me in Groningen. Mita. Anna Nizniowska. Asal Hamarneh. Yongki. A lot of thanks for Henk Stegeman and Poppy Sutanto. mbak Mia. I have met a lot of Indonesian students in Groningen. for your love. Wuri Raspati. Puspita Kencana. Albert Kieviet helped me with the MALDI‐TOF measurements. Dado. August Kurniawan. for your never‐ending blessings. patience. Agnes R. in the lab. My special thanks are for Laura Junistia and Jan‐Pieter Drijfhout. The chemical and glassware magazijn as well as the FWN library personnel were always patient. and Mahesa for being my brothers and sisters in Groningen. Bima and Wisnu (also my paranymph). Diana Jirjis. Francesca Fallani. friendly. Henky Muljana (my paranymph) and Anindhita Widyadhana. Abdul Osman. and Tante Alma for being like my parents and family in Groningen. Ita. and Henk van de Bovenkamp for cheering my days and being my friends during summer and winter. Thank you God. Farchad H. M. Louis Daniel. Tante Smith. Chalid. Hans Heeres. Tante Oppie. Teddy. Marcel Wiegman. Buana Girisuta and Rina Karina. Thank you my parents and sister. Patrick and Oppie. pak Harry. Tiara. Wahyudin. Vincent Nieborg and Fesia Lestari. mas Pandu. Kuslan. Ais Jenie. kak Roga. and for my wife and family. and although we seldom met afterwards. Erna Subroto. and passion. for never leaving me alone. Arjan Kloekhorst. Togi. Iwan Harsono. Diana Santangelo. Wita Sondari. Wangsa. Thank you Tresna. Jaap Bosma.B. Ignacio Melian Cabrera. and helpful. Groningen is also a place where I found true friendship. bu Ida Susanti. Boelo Schuur. Taufik. Many master and PhD students. Gerard Kraai. Egi. Francesca Gambardella. spring and autumn in Groningen. Zhang Youchun. Oscar Rojas. Tenny. and Iwan Kustiawan for becoming my family. Teguh. Annette Korringa. Jasper Huijsman. Samuel. in the office. kak Atha. visiting researchers. Anneke Toxopeus. Taufik. and on many occassions.
. Ardiyanti. Thank you AW. Endy. mbak Ika. You all have always been my inspiration and motivation. Yao Jie. Jelle Wildschut. Judy Retti Witono.Acknowledgements
I had questions or needed specific equipments. support. Rasrendra. Thyara. Sari Fabianti. and the ISD staff helped me a lot with visum and residence permit arrangements as well as tax papers. Mahfud. Nadia Gozali. Tante Caroline. Our collaboration has also become part of this thesis.
9th National Conference organized by the Society for Polymer Science. and H.2008.010). Sugih.) 6. Picchioni.B.B. 3.K. Picchioni.K. A. J. L. Picchioni: Functionalization and blending of polycaprolactone.1016/j.Picchioni. J. L. Sugih. Picchioni. presented in 11th Asia Pacific Confederation of Chemical Engineering (APCChE) Congress. 2008 (submitted to J. India. 26‐28 November 2006 (oral presentation). Experimental Studies on the ring opening polymerization of p‐dioxanone using an Al(OiPr)3‐monosaccharide initiator system. 2008 (accepted for publication in Starch‐ Starke). Broekhuis.Janssen. and F. Sugih.J. L. Manurung. Synthesis of poly‐ (ε)‐caprolactone grafted starch co‐polymers by ring‐opening polymerisation using silylated starch precursors. presented in the 1st International Symposium on Sustainable Chemical Product and Process Engineering. Junistia. 17‐20 December 2006 (oral presentation). A. 8. Picchioni: Biodegradable polymer materials based on polyesters and starch/proteins. F. Mumbai (India). 4. Broekhuis.M. and F.J.K. F. Sugih. Picchioni. (in press. Heeres. Polym. Heeres: Synthesis of higher fatty acid starch esters using vinyl laurate and stearate as reactants. and H. Heeres: Synthesis and properties of novel green biopolymers derived from starch and fatty acids. presented in International Workshop: “From Polymer Modification to Multicomponent Polymer Systems”.J. Manurung. F.P. R. A.P. 2008 (submitted to Carbohydr. Appl. H. Kuala Lumpur (Malaysia). Heeres. 27‐30 August 2006 (oral presentation) 7. L. Synthesis and properties of reactive interfacial agents for polycaprolactone‐starch blends. Drijfhout. Picchioni. Pune Chapter. Junistia. Eur. Polym.M.B. H. starch‐based innovative product: The synthesis of grafted polycaprolactone‐starch by ring opening polymerisation.J.P. L. 2008.B. Sci. F. H. F. presented in “Polymers for Advanced Technologies”.J. Bratislava (Slovak Republic). 25‐28 September 2007 (oral presentation). Heeres: Green. Polym. A. H. A. 9.). 5.Heeres. A.eurpolymj.J.J.K. Sugih. J. A. R. Janssen.P. DOI: 10. Sugih.P.10. Junistia. Experimental and modeling studies on the synthesis and properties of higher fatty esters of corn starch.K. L.J.List of Publications
1. Janssen. Heeres. H. 2008 (accepted for publication in Starch‐Starke). South China University of Technology.P. Picchioni.J. Sugih.K.K. L. Drijfhout.A. A. F.K. L. J. H.M Janssen and H. A. Guangzhou (China).Drijfhout.A. F. 2. A. Sugih.M. A. Sugih.P. Junistia. Heeres.K.