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
dr. dr. Janssen Prof. Moscicki ISBN : 978‐90‐367‐3592‐6 ISBN : 978‐90‐367‐3593‐3 (electronic version)
. Promotores : Prof. dr. dr. Picchioni Beoordelingscommissie : Prof. Broekhuis Prof. ir.B. dr.A. Minnaard Prof. dr. H.M. ir. A.J. Heeres Prof. F. A.P. L. L.J.
to: Tresna. and my sister
. my parents.
188.8.131.52. Plastic applications and waste issues 1. Typical example for the synthesis of polydioxanone end‐capped with 1.2. Starch 1.1. Starch‐based blends by melt mixing 1.1. Starch‐based blends by in situ polymerization 1.2. Materials 184.108.40.206.4. Starch modifications to improve product properties 1.Table of Contents
Chapter 1: Introduction
1.1.3. Materials and Methods 2. Starch ‐ biopolymer blends and graft co‐polymers 1.4‐di‐O‐isopropylidene‐α‐D‐galactopyranose (2) 27 28 26 27 27 27
2.2. Methods 2. Starch production processes 1.3. Thermoplasticized starch 1. Biomaterials from starch 220.127.116.11.1.4. Cross‐linked starch 1.3.3. 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
18.104.22.168.22.214.171.124.126.96.36.199.1. The potential of biodegradable plastics 1.3. Introduction 2. Product analyses
.2. Starch Esters 1. Biodegradable plastics from starch 1. Thesis Outline 1.5.2. Structure and properties of starch 188.8.131.52.
Chapter 3: Synthesis of Poly‐(ε)‐caprolactone Grafted Starch Co‐polymers by Ring Opening Polymerisation using Silylated Starch Precursors
184.108.40.206. 2.4.3. Methods 3. Typical example of in situ polymerization of ε‐CL with silylated starch 50 3.1.1. Conclusions 2. Systematic studies 220.127.116.11.3.1. De‐silylation of poly‐(ε)‐caprolactone grafted silylated starch co‐polymers 51 51 52 52 52 53 54 48 49 49 50 50
3. Nomenclature 18.104.22.168.2. Screening experiments 2.2. Mechanistic aspects 2. Product analyses 2. Synthesis of silylated starch
.3.1. Materials and Methods 22.214.171.124. Results and Discussions 2. Nuclear Magnetic Resonance (NMR) 3.3.2. Typical example of the starch silylation procedure 126.96.36.199. Peracetylation of silylated starch 3. Calculations 3. Results and Discussions 188.8.131.52.184.108.40.206. Calculation of average degree of polymerization 220.127.116.11.2.2. Introduction 3. Analytical methods 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
2.1. Materials 18.104.22.168.3. Product yield 2.3.
Catalysts screening 22.214.171.124. Introduction 126.96.36.199.2.3. 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
188.8.131.52. Materials and Methods 4.1. Nomenclature 4.3. References
.5.2.3. Peracetylation procedure 4. Effect of the addition of toluene as a co‐solvent 4.3.
3. Exploratory experiments 184.108.40.206.2.3. Methods 4.6.2. Nomenclature 220.127.116.11. Conclusions 4. Determination of the Degree of Substitution (DS) 4.3.3. Materials 4. Product characterisation 4. Analytical equipment 4.2.3. FT‐IR measurements 4.2. Systematic studies 18.104.22.168.3.2. References
Chapter 4: Synthesis of Higher Fatty Acid Starch Esters using Vinyl Laurate and Stearate as Reactants
4. 1H‐ and 13C‐NMR analyses 22.214.171.124.2. Deprotection of silylated‐starch‐g‐PCL 3.5. Effect of vinyl ester to AHG ratio on the product DS 4.4. Results and Discussion 126.96.36.199. Conclusions 188.8.131.52. In situ ring opening polymerization of ε‐caprolactone with silylated starch 57 63 64 64 65
Results and Discussion 5.2. Materials 6.1.4. Work‐up of PCL‐g‐GMA products 6. Analytical Methods 110
.3.2.1. Introduction 5.2.3. Nomenclature 5. Compatibilizer synthesis 6.1. Materials and Methods 184.108.40.206. Product properties 5.2. Introduction 6.2. Materials 5.2. References
Chapter 6: Synthesis and Properties of Reactive Interfacial Agents for Polycaprolactone‐Starch Blends
6. Experimental Design 5.2.2. Materials and Methods 220.127.116.11.1.1.2. Methods 6.2.3. Analytical Equipment 18.104.22.168.2.3. Conclusions 22.214.171.124. Mathematical modelling 5.1.2. Peracetylation procedure and Degree of Substitution (DS) determination 88 88 89 92 96 102 102 103
126.96.36.199. Preparation of PCL‐starch blends with the reactive compatibilisers 110 6.2.2. Work‐up of PCL‐g‐DEM 108 109 109 109 109 110 110
188.8.131.52. Methods 5.5. Typical example of the preparation of laurate and stearate starch Esters 88 86 87 87 87 88
5.Chapter 5: Experimental and Modeling Studies on the Synthesis and Properties of Higher Fatty Esters of Corn Starch
184.108.40.206.2. Calculation of the Degree of Functionalization (FD) of the reactive compatibilizers 6.1. 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
220.127.116.11. Ternary blends compatibilized with PCL‐g‐GMA 6.4.4. References Summary Samenvatting (Dutch Summary) Acknowledgements List of Publications
. Nomenclature 6.2. Conclusions 6. Statistical mdeling 6.2. Effect of the BPO intake on the product FD 18.104.22.168.22.214.171.124. Preparation of the ractive compatibilizers 6.2.3. Ternary blends compatibilized with PCL‐g‐DEM 126.96.36.199.1.2.3. Synthesis and Properties of Starch‐ PCL Blends 6.1.
188.8.131.52. Effect of substrate (GMA/DEM) to PCL Ratio on the FD 6. Binary blends of starch and PCL 6.4.1. Thermal properties of the compatibilizers 6. Results and Discussions 6.
. 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. Finally. Different routes to modify starch to improve the product properties and to extend the application range will be provided.
1. The second and third starch producer regions are Europe and Asia [4‐5]. Starch
75% of all organic material on earth is present in the form of polysaccharides. rice. Most of the starch is produced in the USA. maize. but other types of starch such as cassava. 4‐5]. Most of the starch produced worldwide is derived from corn. .9 ton starch per hectare .1. wheat. present and a forecast of starch production  o : Europe □ : USA : Total
∆ : the Rest of the World
. . Plants synthesise and store starch in their structure as an energy reserve. Starch is found in seeds (i.e. potato. with an average yield of 4.45 (estimated from the total volume and value of the US corn starch export ). It is generally deposited in the form of small granules or cells with diameters between 1‐100 µm .Chapter 1
80 70 60 50 40 30 20 10 0
Starch Production (Million Tons)
Figure 1. potato or cassava) of the plants [2‐3]. current and future estimations of worldwide starch production are given in Figure 1. while corn seeds consist of 65‐80% starch by weight. sorghum. sweet potato. An important polysaccharide is starch. barley. Most starch crops are very productive. The worldwide production of starch in 2008 is estimated to be around 66 million tons . Past.1. corn. and wheat starch are also produced in large amounts [2. or peas) and in tubers or roots (i. The current price of corn starch is around $0. Past. Potato accumulates starch to approximately 75 % of the dry weight in the tubers with a yield up to 21 ton starch per hectare.e.
the corn kernel is softened by steeping in warm water containing SO2 until the volume of the kernel increases with 55‐65%. and is embedded in a proteinaceous cellular matrix as is shown in Figure 1. The fiber is removed using washing screens. Corn starch production  a. Cross‐sectional view of a corn kernel
.2. After a coarse milling.1. Typical corn‐milling operation b. which is again fed to a second milling process. The resulting suspension from the mills contains fiber.
a. As an example.2. the mixture is fed to a hydrocyclone to separate the germ and the rest of the kernel. The resulting starch is further washed in a cyclone and finally dried. The low density gluten is separated from the starch suspension by centrifugation. sand. and other foreign materials. a typical corn starch production process is given in Figure 1. After initial cleaning to remove cob. Starch production processes Starch is generally extracted from the plant by wet milling processes [9‐10]. gluten.1.b. the debris is filtered from the slurry. and starch granules are obtained after centrifugation from the suspension.
b. The plant material is grounded in water.a.2.
Figure 1. Introduction
1. and starch. The starch is present in the endosperms (floury and horny).
O H OH
H OH H
O H OH
O H OH
H O .6‐7 x 105 Da .1..Chapter 1
1.2. The linear
. Structure and properties of starch Starch is a polymer consisting of anhydroglucose (AHG) units (see Figure 1.3. H OH H
O H OH
H OH H
c.... Typically. containing periodic branches linked with the backbones through α‐D‐(1. the amylose content is between 18‐28% .. Amylose c.
Figure 1. Two types of AHG polymers are usually present in starch: amylose and amylopectin [2‐3.
H2COH H OH H O H
b. but usually in the range of 1. Chemical Structure of Starch a.. The content of amylose and amylopectine in starch varies and largely depends on the starch source. The molecular weight of amylose is a function of the plant source and processing method. The molecular weight of amylopectin is higher than that of amylose and is typically 4‐5 x 108 Da .6)‐ glucosidic bonds ..
H . . 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 .. 6].3.) .a. The starch granules will swell rapidly to many times of its original volume. The amylose content of several common starches is given in Table 1. Amylose is essentially a linear polymer in which AHG units are predominantly connected through α‐D‐(1....4)‐glucosidic bonds. Amylopectin is a branched polymer....
H2COH H .1.
. H OH H
H2COH H2COH H O H H OH H H2COH H O H H OH H
O H OH
H O .
H2COH H . Each branch contains about 20‐30 anhydroglucose units. but it is very hygroscopic and binds water reversibly. Anhydroglucose (AHG) unit b.
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 (%) 20.2.8 16. Starch gelatinization temperature range 
Starch Potato Tapioca Corn Waxy corn Wheat Gelatinization Temperature Range [o C] 59‐68 58. Parker in 1838. and. The gelatinization temperature range of various starch sources is given in Table 1.2. Table 1. Biomaterials from starch
1.8 17. depending on the amount of water present.2. The first generation of commercial plastics was derived from cellulose nitrate and is known as celluloid [12‐14]. Introduction
amylose molecules leach out of the granules into the solution. and celluloid was
.1.7 20 18. 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 .7 26
Table 1. and granule fragments.5 28 52 70‐75 27 15.1.5 25.2. Plastic Applications and Waste Issues Plastic is the general term for a wide range of synthetic or semisynthetic polymerisation products. Plastics are used in a wide range of applications and the demand is still increasing every year . swollen granules.5‐70 62‐72 63‐72 58‐64
1. Cellulose nitrate was first prepared by A. The resulting suspension contains a mixture of linear amylose molecules. will form a thick paste or gel.
and 37% of the total plastics demand in EU  is used as packaging materials.5
0 1950 1960 1970 1980 Year 1990 2000 2010
Figure 1. The latter is important when using the material for packaging purposes to save weight. Plastics are very attractive materials.4. and energy during transportation of goods.)
Included: Thermoplastics. and PP-Fibers
2006: 245 2002: 200
200 Plastics Production (Million Tons)
Not Included: PET-. Polyurethanes. and polycarbonates (PC). Elastomers. About 29% of the total plastics produced in the USA . for instance to be used in packaging and paper coatings . Sealants. Important polymers used for packaging are polyethylene (HDPE and LDPE). the worldwide plastics production has reached 245 million ton per annum  (see Figure 1. polyethylene terephtalate (PET). electrical.5. the oldest purely synthetic plastic is Bakelite. polystyrene (PS). PE is a very versatile plastic because it can be shaped easily into various forms. PA-. and consumer products (see Figure 1. and therefore need less energy to shape it into useful materials . space.). and PolyAcryl. Hyatt in 1870 . Coatings. While celluloid is derived from a natural polymer (cellulose). Plastics are also used for building materials and automotive. 15].4. polyvinyl chloride (PVC). Worldwide Plastics Production (1950‐2006) 
. Adhesives. when polyethylene (PE) was invented. polypropylene (PP).Fibers
1950: 1. A dramatic increase in demand for plastics began after World War II. In 2006. Plastics have lower melting temperatures compared to glass and metals.Chapter 1
patented by J. They have a low density and can be shaped in thin films while maintaining good properties. The production and consumption of plastics has increased significantly with a rate of almost 10% every year since 1950. Thermosets. The largest application of plastics is for packaging purposes. discovered by Baekeland in 1907 [12.
and other public works . Plastic litter is hazardous to wildlife [1. and lakes. In the USA. The disposal of plastics materials in municipal solid waste (MSW) is a serious issue in many parts in the world. Most of the waste is still disposed (50%) by landfill or incinerated to recover the energy (30%) . consisting of 25%‐v of the total waste . In Europe (2006). however. The volume fraction of plastic in MSW is much larger due to the low density of plastics. Plastics have been polluting sea [1.5.7% in 2006 . Plastic applications and waste treatment in Europe  1.2. hydropower plants operation. irrigation. which are usually found in waste streams. threatening fishery. ship navigation. the amount of plastics in MSW increased from less than 1% in the 1960 to 11.2. rivers. Recycling (part of) the plastics to reduce the amount of MSW also has limitations. It is not applicable for thermoset resins . plastics waste was ranked as the second major source of MSW after paper and paperboard. Introduction
Plastic waste. Biodegradable plastics are polymeric materials capable of decomposing when given an
. only 20% of the plastic is recycled. In 1996. is causing serious environmental problems. 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). 6]. soil.
Figure 1. and accumulation of its residues in soil cause significant reductions in agricultural yields. Co‐mingled plastics. and is only effective for single plastic sources or simple plastic formulations. It has a high volume to weight ratio and is resistant to degradation. 16]. are not easy to recycle . and may be more than twice the weight fraction .
96 proposal ‐ Degradable plastics are plastic materials that undergo bond scission in the backbone of a polymer through chemical.2 working group on biodegradable polymers ‐ Biodegradation of a plastic material is a process leading to naturally occurring metabolic end products. Biodegradable plastics have gained considerable interest since the 1980s. Several authorities have provided definitions for biodegradable plastics . and polyesters from micro‐organisms . For instance. It may also be treated in sewage sludge water treatment plants or buried in the soil [20‐21].5/
. The change in chemical structure results from the action of naturally occurring micro‐ organisms. ASTM sub‐committee D20. animals. Biodegradable plastic waste may be treated in composting facilities. because they take part in nature’s cycle of renewal. and since 2000 they have become competitive with traditional materials in some applications [20‐21]. The market price of biodegradable plastics (mainly from starch and PLA) has decreased the last years. Biopolymers are inherently biodegradable. 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. biopolymers from nature (from plants. 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. Biodegradable plastics may be classified into two general groups. Nowadays new types of biodegradable plastics with improved properties and lower costs have been developed .Chapter 1
appropriate environment and sufficient amount of time . and microorganisms) and biodegradable synthetic polymers . The price of these biodegradable plastics is expected to be reduced considerably in the next decade . 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 . the price of PLA based plastics in 2010 is estimated to be around € 1. DIN 103. biological and/or physical forces in the environment at a rate which leads to fragmentation or disintegration of the plastics. together with food and yard waste as well as paper. proteins (present abundantly in plants and animals). Common natural biopolymers are carbohydrates.
blending with other materials.45 per kg . consumer products and food products. € 1. The price of starch in 2007 was about $ 0. when recycling is not practical or uneconomical [20‐21]. Bastioli [20‐21] showed that nearly all biodegradable plastics available in the market are derived from starch.50‐4.2. either by plasticization. or when environmental impacts have to be minimized. chemical modification or combinations of them . breathable fabrics. biofillers for tires. 20]. Starch may become an attractive raw material for plastics in the future. . hygienic materials (diapers. half of its price in 2003 (around 3 Euro/kg).
1. pots). agricultural products (mulch films.50/ kg in 2005) is also expected to be lower in the future due to considerable reduction in modification costs . Introduction
kg. Examples are the use as packaging material for carrier bags. and chewable items for pets [1. alone or complexed with natural or synthetic biodegradable polymers) or from polylactic acid which originates from the fermentation of a starch feedstock.3. Biodegradable plastics are also used for food servicewares . napkins) . starch must be modified. It cannot be shaped in films with adequate mechanical properties (high percentage elongation. Biodegradable Plastics from Starch Starch is a very attractive source for the development of biodegradable plastics. 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 . These are summarized briefly in the next sections.2‐ 1. before they can be applied as biodegradable plastics.4 per kg in 2007‐2008 ). Consequently. The price of starch‐based biodegradable plastics (€ 1. either from starch‐based materials (slightly modified starch. Biodegradable plastics are especially very useful for single‐use applications. because the price of oil based polymers may still increase due to the rise in the crude oil prices . Virgin starch is not suitable as a packaging material. utensils .3. such as polyethylene (PE.
. towels. This price is much lower than conventional plastics derived from oil. 1.8‐ 2. tensile and flexural strength) and is too sensitive to water [26‐27]. Starch modifications to improve product properties
Several techniques may be applied to develop starch based biomaterials with improved properties.4 or $ 1.
30. water content. phase separation may still occur. For instance. glycol. because the resulting product will be brittle when equilibrated with ambient humidity . 39]. Although thermoplasticization seems to be a promising method. This problem is mainly caused by the presence of strong inter‐ and intra‐molecular hydrogen bonds between the starch macromolecules [28. the semicrystalline structure of starch and its granular form are lost and the starch polymers are partially depolymerized. 35. 37]. and blow moulding [28. Another issue is the poor water resistance and low strength. The thermoplasticization process will decrease the interactions of the molecular chain and destruct the structure of the starch . which results in embrittlement. Thermoplasticized starch Virgin starch is brittle and difficult to be processed into articles due to its relatively high glass transition temperature (Tg. At intermediate water levels. 30]. 40].3. and therefore are processable by conventional polymer processing techniques such as injection. The brittleness is known to increase in time due to free volume relaxation and retrogradation. The use of other plasticizers (for example glycerol) results in a rubbery material. 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. Starch can be modified to obtain materials which melt below the decomposition temperature . 33]. The modification involves break down of the starch granular structure by the use of plasticizers at high temperatures (90‐180 oC) and shear. The use of water as a plasticizer is not preferable. extrusion. sugars) [30. and the properties and amount of plasticizers. . The plasticized starch properties may be tuned by changing the temperature of processing. TPS synthesized from polyol and sugar plasticizers have the tendency to re‐crystallize (retrogradation) after being stored for a period of time. As the result. destructed. with better properties than virgin starch in various applications. which will result in a continuous phase in the form of a viscous melt [29‐30. which still limits their use. resulted in the formation of an amorphous mass [35. The plasticizers are also usually hydrophilic and can be
. 32‐ 36]. There are several substances used as plasticizer for the preparation of thermoplastic starch (TPS). Yu. such as water and polyols (glycerol. The modified products are known as thermoplastic. glycerol. approximately 230 oC) . plasticized.1. the thermal properties of glycerol‐plasticized starch are a function of water content [31. sorbitol. or melted starch [29. 40].Chapter 1
1. and hexylene glycol. et al  discovered that the elongation of break of the thermoplastic starch is significantly improved by plasticization with glycol. 32. which is even above the thermal degradation temperature.
3. paper. The esterification reaction is catalyzed by H+ and proceeds with the formation of water . and particularly with acetic acid . As a result. 1. Introduction
washed out by water . Starch Esters The development of starch esters started in the mid 19th century [44. Already in 1865. epoxy. et al  studied the mechanical properties of films from maize starch cross‐linked by sodium trimetaphosphate (SMTP). These hydroxyls can react easily with a wide range of compounds such as acid anhydrides. 1. aldehydes. Schuetzenberger acetylated starch with acetic anhydride . Medium DS starch acetates (0. the use of formyl esters of starch is much lower than that of acetate esters. differing in Degree of Substitution (DS). Cross‐linking results in a reduction of the solubility in water and to thickening. Solutions to improve the properties of TPS are blending. while highly substituted starch acetate (DS of 2‐3) are soluble in organic solvents. Breakdown of the starch and the formation of low molecular weight products occur to a significant extent .3.2) are commercially available and used as food additives and in the textile industry. organic chloro‐compounds. The cross‐linked products are therefore more rigid (and less elastic) materials than virgin starch . Low DS products (0. textile.3‐1) are still soluble in water. or coating with hydrophobic polymers [33.2.01‐0. The cross‐linked starches have found many applications. and adhesive industry. Substitution of the hydroxyl groups of starch with acetate groups makes the esters more hydrophobic than native starch. leading to higher viscosities [42‐43]. High DS starch acetate can be easily casted
. especially in the food. 39]. Three different types of starch acetates may be distinguished. 46]. The DS is defined as the moles of substituents per mole of AHG units . Silva. Chemicals of these classes having two or more of the reactive groups may react with two or more hydroxyls of the starch molecules. Most studies dealt with the synthesis of starch esters of C1‐C4 carboxylic acids. Higher processing temperatures generally led to higher cross‐linking levels. synthesized by direct addition of formic acid to starch at room temperature. resulting in an increase of the Young’s modulus and tensile strength of the products and a decrease in elongation at break. The products are called cross‐linked starches. The simplest starch ester is starch formate (C1). The most popular starch ester is starch acetate . Cross‐linked starch An anhydroglucose molecule of starch contains two secondary and one primary hydroxyl group. and ethylenic compounds.3.
DMF. Starch triacetate has been successfully synthesized using acetic anhydride in combination with pyridine‐gelatinized starch . DMSO. but also C‐6 and C‐ 3 substitution . Several synthetic routes have been developed for starch esters. while alkaline catalysts (such as carbonate. High DS starch acetates are thermoplastic materials suitable to be used as biodegradable plastics .1. xylene. hydrogencarbonate. Mechanism of starch acetylation using vinyl acetate
. allowing the polymer to be produced with a range of hydrophobicities . et al [55‐56]. The degree of acetylation of starch acetates can be easily controlled. The kinetics of the reaction between gelatinized aqueous potato starch and vinyl acetate was studied by De Graaf.
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. Starch esters have also been synthesized using alkanoyl chlorides [49‐53]. acetate. Esterification may be performed using acid anhydrides in aqueous media or organic solvents (pyridine. Neutral and weak acid/ alkaline catalysts result in regioselective substitution at the C2 hydroxyl groups of starch. and phosphate) will result in C‐2. Acetylation of starch in water and DMSO using vinyl acetate has been studied lately by Mormann and Al‐Higari  as well as Dicke . Another attractive route involves the use of vinyl esters as reagents .Chapter 1
into films using organic solvents. or isopropanol) with acidic (hydrochloric or sulfuric acid) or alkaline (NaOH or triethylamine) catalysts . and is shown in Scheme 1.1.
interfacial agents (e.61) products could be obtained using this approach. polymers are rarely miscible with each other  so that. Starch ‐ Biopolymer Blends and Graft Co‐Polymers Blending of different polymers is an established method to obtain products with improved properties. the latter are mixed together. By working at temperatures above the melting point or glass transition temperature of the two components. compatibilizers) might be used. The first involves simple melt mixing of the two components for example by extrusion. 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 . 1. in the simplest case of a binary blend. Despite this advantage. it is frequently used in order to chemically graft the polymerized component on the other one (polystyrene on polybutadiene in our example above ). which in turn provide a very strong adhesion at the molecular level between the dispersed phase and the matrix. Two main methodologies are applied for the production of polymer blends.3. dodecanoyl. As a result. Although the in situ polymerization process is not as technologically straightforward and economically convenient as melt‐mixing. in order to achieve also a good molecular adhesion between the phases by melt mixing. but only relatively low‐DS (0. however. the two components are chemically linked to each other. Introduction
The synthesis of long chain fatty acid ester of starch has attracted much interest lately. However. 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 . 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. i. availability of mixing equipment and no necessity to use any organic solvent (often employed for the in situ polymerization).34‐0. one component will be dispersed into the other. octadecanoyl) chlorides [49‐53]. a chemical reaction might take place upon processing (reactive extrusion).4.g. However. melt‐mixing remains the preferred route to polymer blends mainly because of very practical reasons: low costs. relatively expensive and rather corrosive. Their role is comparable to the one of a surfactant in emulsion formation . If the right combination of chemical groups on the two components is present along the polymers backbone. The introduction of longer acid chain is expected to reduce the brittleness of virgin starch and to increase its hydrophobicity . they locate themselves at the interface between the
.e. The fatty acid chloride reactants are. Fatty‐acid starch esters have been synthesized using fatty acid (octanoyl. 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 .
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.Chapter 1
two components stabilizing the dispersion (most probably by a steric repulsion mechanism ) and providing an improved adhesion at the interface.). 66‐81].
Figure 1. The most often used synthetic polymer for blending with starch is polyethylene [27. In the past. 64‐98]. The uncompatibilized blends of starch and
. there have been efforts to blend as well as to graft synthetic polymers onto starch.3.4.6. The products have been synthesized in the lab as well as on industrial scale [32. 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. The synthesis of compatibilizer and its use for starch/ synthetic polymer blending has also been studied.6. Illustration of the role of interfacial agent in compatibilizing blends 1. Starch‐based blends by melt mixing Starch‐synthetic polymer blending has been studied as early as in 1973 [64‐65]. 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.1. 80‐ 81] or food packaging . 39.
dextran‐g‐PCL . poly‐dioxanone. To obtain completely biodegradable products. PCL‐g‐maleic anhydride (PCL‐g‐MA) [91. synthetic biodegradable polymers have been applied. The mechanical properties of these blends (tensile strength and elongation at break) decrease at higher starch content. Introduction
polyethylene show a coarse phase separation due to differences in polarity of starch (hydrophilic) and polyethylene (hydrophobic). 96]. poly‐valerolactone. PCL‐g‐pyromellitic anhydride [94‐95]. As a result a graft copolymer (PE‐g‐Starch). while the tensile strength of the compatibilized ones decreases only slightly with the starch content. 39. and poly‐caprolactone). which both can react in situ with the hydroxyl groups of starch [70. poly‐lactides. As is the case of blends with PE. and poly‐lactones (such as poly‐butyrolactone. among which synthetic polyesters are considered very promising materials . Despite these good results. while PE‐g‐GMA posseses epoxide groups. which combines excellent biodegradability with acceptable mechanical properties. Studies on the blending of starch with PCL have been already described [32. 84‐98]. 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. since the conventional synthetic polymers are usually poorly or non‐biodegradable. The tensile strength of the uncompatibilized blends is drastically reduced when starch content is increased. the compatibilizer. thus improving the dispersion of starch in the PE matrix.e. Examples of these biodegradable polyesters are poly‐glycolide. The use of conventional synthetic polymer such as PE for blending with starch will only result in a partially biodegradable material. The mechanical properties of the blend are also improved. Some of these polyesters also have very good mechanical. and EVOH . The use of reactive compatibilizer precursors PCL‐g‐glycidyl methacrylate (PCL‐g‐GMA) [89‐90]. is formed upon mixing. the uncompatibilized blends of starch with PCL give coarse phase separation and a reduction in the mechanical properties when the starch content is increased. PE‐g‐MA contains reactive anhydride sites.
. These biodegradable polyesters will finally decompose into non‐toxic products . 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. The ester bonds of these polymers are susceptible to attack by water and this leads to enhanced biodegradability. Polycaprolactone (PCL) is a well‐known synthetic biodegradable polyester. thermal and water/gas permeability properties that are even comparable to bulk non‐biodegradable polymers such as PE and PP. 72‐73]. 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. i.
1. 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 . Three routes have been studied in detail: starch esterification. and the presence of liquid‐solid reaction system leads to reduced reaction rates between starch and ε−caprolactone. and difficult to handle since it releases ethane. Starch is a very hydrophilic material that always contains moisture. a very flammable by‐product. Thesis Outline
The objective of this thesis is to study synthetic pathways to obtain starch derivatives with the potential to be used as bioplastic. The water in starch granules competes with the hydroxyl groups of starch in the initiation step of the polymerization reaction. The results of this study have been used as input for the synthesis of the starch‐g‐PCL.2. In Chapter 3.2. The method applied basically consists of three steps (temporary partial protection of starch‐hydroxyl
. In this reaction. 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. the hydroxyl groups of starch are supposed to function as initiating sites. defined as the percentage of grafted polyester to starch compared to the total amount of homopolymer and grafted polyester) . The highest GE value (up to 90%) has been achieved when using triethylaluminium as catalyst [97‐98]. starch‐g‐PCL has been synthesized using toxic materials such as isocyanates .
1.4‐di‐O‐isopropylidene‐α‐D‐galactopyranose) using i Al(O Pr)3 as the catalyst. Starch is insoluble in the typical organic solvents used for ROP (such as toluene or THF). the synthesis of starch‐g‐PCL is reported. starch/ biopolymer blending. leading to formation of PCL homopolymers rather that starch‐g‐PCL.4. thus resulting eventually in low GE values . Another approach consists of the Ring Opening Polymerization (ROP) of ε‐caprolactone monomer in the presence of starch. A new strategy for the in situ ROP of ε−caprolactone on starch with the use of common ROP catalyst is therefore highly desirable. Another possible cause for the low GE values is the heterogenous nature of the reaction. and starch‐g‐ biopolymer formation. during the reaction. Previous studies showed that common ROP catalysts such as tin octoate or aluminium isopropoxide gave low (0‐14%) grafting efficiencies (GE. Chapter 2 describes the ROP of p‐dioxanone initiated by a protected monosaccharide (1.3. In the past. which however is extremely air‐ and water‐sensitive.3.
University of Wales. are also reported. 2000‐2005. Corn. a systematic experimental study on the effect of the process variables on the starch ester DS has been performed.. . and removal of the silyl groups). USA. 1995.B. C. C. Washington D. An Introduction to the New Science of Biodegradable Plastics. Wurzburg). starch. UK. Denmark: http://www. Fowler: Industrial markets for starch. UK. Mol. together with the mechanical and thermal properties of the synthesized starch esters are reported in Chapter 5. 2005. Aarhus.B.H. Breed. addition of co‐solvent. References
. Website of International Starch Institute. R. London. A preliminary study on the effect of several process variables (reactant ratio. Whistler. R.html. Nawrath. C. . Somerville: Mini review: Plant polymers for biodegradable plastics: cellulose. . 2002. 1986.F. The synthesis of two interesting compatibilizers (PCL‐g‐GMA and PCL‐g‐DEM) for starch/PCL blends is discussed in Chapter 6. Sand Hutton. Wurzburg: Introduction.A.5. The Biocomposites Centre. CRC Press. Bemiller. Bangor. Passchall).. de Bragança. Based on these preliminary results. . Part of A Global Economy. 105‐122. USA. 2008. 2007
. York. and polyhydroxyalkanoates. Poirier. O. P. Academic Press. Y. 2004. 1984. . including the mechanical and thermal properties. Holmes: Interactive European Network for Industrial Crops and their Applications.starch. in Modified Starches: Properties and Uses (Ed.C. E. . Princeton University Press. ROP of ε−CL on the remaining starch‐hydroxyl groups. UK.dk/isi/stat/rawmaterial. Inc. accessed on August 01. J. and application of different catalysts) on starch esterification using vinyl fatty esters (vinyl laurate and vinyl stearate) is provided in Chapter 4. Stevens: Green Plastics. .L. New Jersey. Boca Raton. Corn Refiners Association. The model. USA. 1. A. Summary Report for the European Union.S. Central Science Laboratory. The use of these compatibilizers in starch/PCL blends.N. O. Introduction
by trimethylsilyl groups.
1. Gywnedd. Corn Refiners Annual Report 2007. Young: Fractionation of starch. Agricultural and Rural Strategy Group.M. Inc. E. in Starch: Chemistry and Technology (Ed. The results were quantified using a statistical model..
Polym. in Kirk Othmer Encyclopedia of Chemical Technology Volume 22. Patel.M. J. . USA. Germany. USA: http://www. Properties. R. . Inc. Institute for Prospective Technological Studies and European Science and Technology Observatory. Hanser Publishers.351‐355. 2008. Bemiller. Applications. M. Bastioli: Global status of the production of biobased packaging materials. Watson: Corn and sorghum starches: production. accessed on August 19. G. Department of Dairy and Food Science. X. Marscheider‐Weidemann. 1990. London. 2000.N.Chapter 1
. Strong: Plastics: Materials and Processing. R. 1273‐1335 .epa. Husing. C. M. . J. Technomic Publishing Company. Denmark.. John Wiley & Sons. Inc.gov/garbage/plastic. S. B. Applications.E. . 1996. . Techno‐economic Feasibility of Large‐Scale Production of Bio‐based Polymers in Europe. Ehrenstein: Polymeric Materials: Structure.htm. Munich. . Seidel). the Food Biopack Conference Proceeding (Ed. 557‐562. A. 2000. J. 105‐109 . A. 2000. R. 27‐40. Processing. 1998.M. Passchall).D. 2003. Academic Press. Rustgi: Biodegradable polymers. Chandra. Selke: Packaging and the Environment. . J. 2008 .
. Inc. The compelling facts about plastics: An analysis of plastics production. Weith: Plastics. F. Englewood Cliffs. Association of Plastics Manufacturers.. and Regulations. Whistler. . R. 23. Munich. The Royal Veterinary and Agricultural University. Germany.F. A. Wolf. Eng. Ind. USA. Chem. Lancaster. Website of Environmental Protection Agency. R.L. 1984. Frederiksberg. Selke. Link. Starch/ Starke 2001. 11..J. S. Fritz. Hanser Publishers.E. Ren: Biodegradable plastics: a solution or challenge?. J. Fifth Edition (Ed. Whistler. Culter: Plastics Packaging. . New Jersey.B. U. Starch/ Starke 2001. Weber). Bastioli: Global status of the production of biobased packaging materials.J. S. E. USA. 2005.W. G.A. Sci. Inc. Schleich. Cleaner Prod.L. C. Braun: Environmental impacts of biobased/ biodegradable packaging. 31. Brussels. Belgium. Daniel: Starch. Angerer: Technical Report Series. UK. Hoboken.J. Prentice‐Hall. C. Washington DC. Crank. Properties. Prog. in Starch: Chemistry and Technology (Ed. 53 . . Hernandez. .R. 53. Plastics Europe. O. R.. 1939. J. demand and recovery for 2006 in Europe.
A. P. Dev. 1996. and Olsen. Macromol. measured by differential scanning calorimetry and solid state NMR spectroscopy. Westhoff. J. Part C: Polym. N. Forssella. Y. B. X. J. Yang. Polym. Rev. Starch/ Starke 2001. a model for thermoplastic starch. Ind. of concentrated barley starch‐glycerol‐water mixtures.M. Yu. P. M.F. Carbohydr. Denmark. epoxy functions and stearoyl chloride. Gao. .V. 53. properties. Chem. Ma: The effects of citric acid on the properties of thermoplastic starch plasticized by glycerol.H. R. Prod. Kruiskamp. Interfaces 2006. Crops Prod.M. 34.Lin: Note. Vliegenthart: Interaction between dry starch and plasticisers glycerol or ethylene glycol. polylactate and polyhydroxybutyrate. 21. J. Parker: Phase and glass transition behaviour. 331–336 . 62. Chambre Syndicale des Emballages en Matiere Plastique (French Plastic and Flexible Packaging Association). 356‐361 . Weber). 4.F. 385–409 . Starch/Stärke.J. T. Eng. X. G.K. Compos. Seville. Schwach. P. Bagdi.M. Appl. Introduction
European Commission Directorate General of Joint Research Centre. Avérous. 53. Mikkilti. Spain. Frederiksberg C. A. 2005. Physical and mechanical properties of biobased materials – starch. J. 1980. 2000.. R. Polym. A. 372–380 . 13. 2005. J. . Biodegradable thermoplastic starch.G. Doane: Starch‐based blown films. 2003.G. K. 2005. J. 1491‐1494 . Ind. Smits. Biobased packaging materials for the food industry. 494–504 . Yu. A. the Royal Veterinary and Agricultural University. 409–416 . 1997. C.. Carvalho. Otey. J. report of Food Biopack Project (Ed. 2003. status and perspectives. . 57. Starch/ Stärke 2001.B. . interaction. K. Wang. O. van Soest. and Couturier: Properties of biodegradable multilayer films based on plasticized wheat starch. and Pukanszky: Thermoplastic starch/layered silicate composites: structure. Moates.J. Petersen.. Polymer Sci. Curvelo. 1–17 . 53.M. Wang. W. 592‐595. 2008. Department of Dairy and Food Science. L. Muller.P. Press Release: Major Price Hikes in Plastics. July 15. E. Sci. Gandini: Surface chemical modification of thermoplastic starch: reactions with isocyanates.H. J. K.S. Y. Martin.L. Carbohydr.A. F. Res. Wang: Properties of starch blends with biodegradable polymers. Nielsen. P. 43. A Real Case of “Force Majeure” that should be passed through to Packaging Prices Immediately. 275‐282
J. USA.J. Bertrand. B. Gattin. 1. Mullen. 57. . B. 2004.P. E. 139–147 . W. J. C. destructuration and hydrolysis during O‐formylation. 2594–2604 . Ramsay.J. Polym. 44. Polym. 1999. 56. Wurzburg. S.B. J. J. Carbohydr. Wurzburg: Cross‐linked starches. Chem. 55‐78. C. M. X.C. B. 58 (9). B. Sagar. Coutuier: Comparative biodegradation study of starch‐ and polylactic acid‐based materials. P. E. Appl. Degrad. . Environ. Aburto. Silva.D. Thiebaud. Rodriguez‐Gonzalez. Boca Raton. chapter 4 in book Modified Starches: Properties and Uses.A. W.D.W. R. 705‐721. Polym. Vikman. 11–17. O. Jarowenko: Acetylated starch and miscellanous organic esters. Panayiotou: Processing and characterization of starch/polycaprolactone products. 34. F. Journal of Applied Polymer Science 1995.
. T. Polym. Wurzburg). Chem. A. 2001.A. Inc. Jiang. Prinos.. Sci. . Sci. E. O. K. Ramsay.W.H. V. S. Favis: High performance LDPE/ thermoplastic starch blends: a sustainable alternative to pure polyethylene. M. CRC Press. Matzinos. Feil: Morphology and enzymatic degradation of thermoplastic starch– polycaprolactone blends. Boca Raton. Pilin. Kontoyiannis. Favis: Rheological and thermal properties of thermoplastic with high glycerol content. A. Myllarinen. Borredon. 2002.A. F.W. (editor). 1647‐1656. Appl. J. 58. T.B. 1997. 2004. Carvalho. I. USA. Ind. J.W. Andrade: Reactive processing and mechanical properties of cross‐linked maize starch. C. Rodriguez‐Gonzalez. Pacsu: Starch studies: Possible industrial utilization of starch esters. Y. van der Zee. 46–51 . Bikiaris. 17–24 . 65. Hulleman. in Modified Starches: Properties and Uses (Ed. Grohens: Starch modifications. C. 1986. 1943. 74. Alric. J. 1986. 381‐384 . I. Mullen.Chapter 1
. Tserki. Crops Prod. O. G. 1517–1526 . .F. Ind. 2006. Polymer 2003. Ribeiro.T. 1209‐1217. Eng. Levesque. E. M. C. Ibezim. Copinet.. . CRC Press.A. Feller. Y. Pacsu: Starch studies: Preparation and properties of starch triesters. H. J. 581–586 . Ind. 24. 77. Divers. Qiao. Merrill: Properties of fatty‐acid esters of starch. E. P. Starch/Stärke 2005.D. Panayiotou: Properties of fatty‐acid esters of starch and their blends with LDPE. A.B. 389‐398 .C. 1942. Sun: Reinforced thermoplastic acetylated starch with layered silicates. Stab. Inc.D. Eng. Polym. 35. D.
Symp. C. Alric. A. Aburto. A. A. Scott. Macromolecules 1996. Panayiotou: Synthesis. Alric. 1997. Wiley‐VCH. D. A. E. 137‐144. 175. J. 34. P. Aburto. . Hamaili. J. 1991. E. E. E. Borredon: Organic solvent‐free transesterification of various starches with lauric acid methyl ester and triacyl glycerides. Borredon: Free‐solvent synthesis and properties of higher fatty esters of starch ‐ Part 2. P. Inoue: Compatibilizers for melt blending: Premade block copolymers. I. J. Smith. A. Thiebaud. 1440‐1451. Alric. Borredon. Thiebaud. J. Dicke.Graf. Aburto. Macosko. C. Cellulose 2004. P. Mouysset‐Baziard. W.M. Hanser Publishers. Germany. W. Starch‐Starke 2004. C. . Introduction
. 1999. J.C. S.Painter: Specific Interactions and the Miscibility of Polymer Blends. . Khandpur.D. L. Mormann. Prinos. de Graaf. Bakker and G. Guegan. C. 1996 .H. 118‐121. Alric. Lancaster.K. in Reactive Polymer Blending (Ed. G. Duquesne. Technomic Publishing Company. 74. E. de Graaf: Ph. Alric. Polym. R.. P. . Starch‐Starke 1999. . 28. M. . M. Degee. R.E.V. Polym. 255‐263. Bikiaris. 132‐135.828.F. D. I. 5590‐5598 . Borredon: Preparation of long‐chain esters of starch using fatty acid chlorides in the absence of an organic solvent. Bakker. D. 2001. Hu. M. Nakayama. Appl. E.E. USA. Senocq. P. 145‐152. 1997. . J. 56. P. Dubois: Synthesis and characterization of compatibilized poly(ε‐caprolactone)/ granular starch composites.E. Introduction. characterization.J. Rijksuniversiteit Groningen. New York. 2001. USA. J.Stokes. A straight way to regioselectively functionalized polysaccharide esters. Thesis. . J.Fennel Evans: Fundamental of Interfacial Engineering. . 29. T. 51. Macromol. G. Bikiaris.E. H. I. Broekroelofs. I. Hu). Starch‐Starke 1999. . I. 51. 57. D. . 11.928. Inc. Rutot. Panayiotou: Properties of octanoated starch and its blends with polyethylene. G. Aburto. R.W. J. S. A. US Patent 2. Starch‐ Starke 2005. Carbohydr.Coleman. A. Prinos. Carbohydr. Janssen. Borredon. 101‐112. J. F. 33‐43
. 302‐307. 1960 . C. Sci. W. Aburto. M. and biodegradability of fatty‐acid esters of amylose and starch.H. R. Marechal. Tuschoff: Acylation of hydroxy compounds with vinyl esters. C. B. 1995. Polym. Beenackers: The kinetics of the acetylation of gelatinised potato starch. Al‐Higari: Acylation of starch with vinyl acetate in water. A. Munich.
N. Jane. . Sci. 1991. Danjaji. Pedroso. 647‐655
. Almeida. J. N. N. Bhattacharya: Properties of injection moulded starch/ synthetic polymer blends. H. W. Adv. Polym.A. 16. Degr. 1‐9 . K. Testing. Ismail. Sun: Effects of polyethylene‐grafted maleic anhydride (PE‐g‐MA) on thermal properties. Wang. Polym. Koutsopoulos. 88. Pavlidou.S. 2904–2911. Polvm. Polymer. Gelina. Polym. L. 2005. 162–163.A. 2003. Nawang. Frangis. J. E. Res. Mei: Study and development of LDPE/starch partially biodegradable compounds.A. C. D. Nikolov. Polym. Ind. 310–317 . M. D. 30. 2005. Ishiaku. Nakamura. 88. A. Degradation studies and moisture uptake of sago‐starch‐ filled linear low‐density polyethylene composites. Favis. Appl.D. 1998. A. 1242‐1251 . morphology. Appl. Chem. L. Sung. Z. 287‐291 . I. J. 59. L. J. Z. Panayiotou: LDPE/Starch blends compatibilized with PE‐g‐ MA copolymers. Carbohydr. Ribes‐Greus: Thermal and viscoelastic properties of some commercial starch products. Ramsay. Panayiotou: LDPE/plasticized starch blends containing PE‐g‐MA copolymer as compatibilizer.D. 1503–1521 .S. Mater. R. D.S. Effect of compounding and starch modification on properties of starch‐filled Low‐ Density Polyethylene. 2002. 38.G. Mani. Contat‐Rodrigo. Ramsay. Prinos. I.L. R. 236–241 .S. thermal and morphological properties of virgin and reprocessed LDPE/corn starch blends. Ishiaku. H.G. 1998. U. 70. Polym.G. Polym. U. Sci.‐J.M. Duran. Z. Pedroso. G. 1997. 1467‐1475. . 2003.I. Stab. Sci. Bikiaris. 21. 167–172 . Ismail. 2001. Rosa: Effects of the compatibilizer PE‐g‐GMA on the mechanical. J. J. Nawang. Technol.M.. J. 2005. and tensile properties of Low‐Density Polyethylene (LDPE) and corn starch blends. Rosa: Mechanical. N. J. Eng. R. Vouroutzis.S. R. Verhoogt: Processing and characterization of thermoplastic starch/polyethylene blends. III. thermal and morphological characterization of recycled LDPE/corn starch blends. 1841‐1846 . A.A. D.H. Eur. R. Mohd Ishak: Material behaviour. Appl. Mohd Ishak: Material properties.D.J. 59. 75–81 . Cordi. B. Testing.Chapter 1
. Effect of amylopectin to amylose ratio in starch. E. Danjaji. H. L. St‐Pierre. Polym. 1998. 20. C. Polym. Evangelista. Z. Tech. Liu. Y. Mechanical properties of sago starch‐filled linear low density polyethylene (LLDPE) composites. Proc. 34. W. B. Bikiaris.
K. M.M. Kim. Jung. Sci. PHB‐V. K. K. J.S..M. B. Chem. J. 1999. Liu. Res. S. Dev. 9177–9190 . Adv. Polym. 1997. Eng.H. . U. Park: Modification of aliphatic polyesters and their reactive blends with starch. D. A. Sci. J. Bhattacharya Residual stresses and density gradient in injection molded starch/synthetic polymer blends Polymer 2000. M. Polym. Siddaramaiah. Ind. T. 3976–3987 . Sci. 1659‐1663 . E. 2002. D. Chem. 23. Polym. Preliminary diffusion evaluation. Westhoff: Starch‐Based Films. Garg.P. Introduction
. Polym. Sci. 35. L. and their blends with starch in soil compost. 4157–4167 . Jana: Studies on the properties and characteristics of starch– LDPE blend films using cross‐linked. Raj. 1298‐1305. C. Karlsson: Polymers from Renewable Resources. 43. Stridsberg. 2004. Rodrigues. 157. Sen.. S.H. 23. O. S. 2001. Appl. 1507– 1516. Otey. K. Bhattacharya: Synthesis and characterization of anhydride‐functional polycaprolactone. Tang. 89. Jana. Tech. Calil: Effect of thermal aging on the biodegradation of PCL. M. A. J. Res. Appl. dos Santos Rosa. Eur. 12.S. L. R. 42‐65. Ind. Chem. Polym. Kim.K. Polym. Sci. Odusanya. Chem. A‐C. 81. Polymer 2000.G.R. A. . Fringant: Properties of thermoplastic blends: starch–polycaprolactone. Appl.W. Prod. U. Ryner. B. 2004. 2002. 1984. Maiti: A biodegradable film.S.Y. 2000. 267. S.‐C. Doane: Starch‐based blown films. U. J. Lindblad. 179‐187. Sci.. in Polym. R. Ranucci. J. A.D. Otey. Azemi: Effect of starch predrying on the mechanical properties of starch/poly(e‐caprolactone) composites. 2003. B. B. M. 139‐161. C. Kim. Eng. 87. H. Sci. Albertsson: Controlled Ring‐Opening Polymerization: Polymers with designed Macromolecular Architecture. Ang. Guedes.M.K. Moro. J. 1987. Pilot plant investigation for production of the biodegradable film. 1139–1148. F. 41.C. Ishiaku. 26. Z. 16–19 . Cho. Kammer: On mechanical properties of sago starch/poly(ε‐caprolactone) composites. Dole.M. 1. Makromol. . Westhoff. Y. Park: Reactive blends of gelatinized starch and polycaprolactone‐g‐glycidyl methacrylate. P. Adv. N.P. C. 3539–3546 . John. Yang. O. W.H. Adv. 877–884. Polym. T. Low density polyethylene/ starch blend films for food packaging applications. 40. Sankar. . Polym. 41. M. 2. 284‐287 .H. S. 32–45 . M. Ishiaku.M. M. 2003. Part A: Polym. cross‐linked and glycerol modified starch. J. Environ. J.F. Azemi. Manan.S. Polym. Eng. 157. F. glycerol modified. J. . Odusanya. 2007. .N. Manan. J.
. Averous. Albertsson. C.
M. 1998. Rapid Commun. Rutot. 1432–1442. M. E. Narayan: Aliphatic polyester‐grafted starch‐like polysaccharides by ring‐opening polymerization. P. P. 2001. Polym. D. Macromol. Degrad. Krishnan. Macromol.E. E. R. 283–286. Rimedio: Polymer communication: Preparation and characterisation of compatibilised polycaprolactone/ starch composites. M.
. P. Dubois: Aliphatic polyester‐grafted starch composites by in situ ring opening polymerization. R. Avella. C. P. R. Errico. 2002. Raimo.E. Avella. Laurienzo. Dubois: Synthesis and characterization of compatibilized poly (e‐caprolactone)/ granular starch composites. Stab. Degee. R. and P. Composite Interfaces. 7. Narayan. D. 2003. Appl. Rimedio. Dubois. M.S. 3875–3881 . 41. 127–134.Chapter 1
. Symp. J. Polymer 2000. R. . 80. . Errico. M. 33–43 . P. M. Polymer 1999. M. Wu: Physical properties and biodegradability of maleated‐ polycaprolactone/ starch composite. 215–225. Duquesne. 40. . 19. Bhattacharya: Synthesis and characterization of starch‐ graft‐polycaprolactone as compatibilizer for starch/polycaprolactone blends. 175. . Polym. Martuscelli. Mani. Sadocco: Preparation of biodegradable polyesters/high‐amylose‐starch composites by reactive blending and their characterization. Sci. 83. P. 2000. Tang. 3091‐3100. J. Rutot. Degée.
3. also polydioxanone with an OiPr end‐group was formed (20‐30 %).Chapter 2
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. Systematic studies showed that the polymer yield is a function of the reaction temperature and the reaction time. polyesters. Besides the desired polydioxanone functionalised with a monosaccharide end‐group. ring‐opening polymerization
. and MALDI‐TOF mass spectrometry. with higher temperatures (100°C) leading to lower yields. 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. The products were synthesized at 60‐100°C and characterized by 1H‐ and 13C‐NMR. Keywords: biodegradable.2. 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).4‐di‐O‐isopropylidene‐α‐D‐galactopyranose)/Al(OiPr)3 initiator system to yield polydioxanone with a protected monosaccharide end‐ group is described.
2. and is tougher than polylactides and even HDPE . The relatively low melting point of polycaprolactone limits its applicability.4‐dioxan‐2‐ one)). It is fully degraded in the body within a period of 180 days . This melting point is considerably higher than typically found for polycaprolactone (60 oC) and lower than that of polylactides (at least 175°C). Nishida et al  reported that polydioxanone decomposes to non‐toxic gases (CO2 and H2O) by microorganisms. which is a unique compromise between application and processing temperature. polyglycolide. or zinc lactate has been reported. Biodegradable aliphatic polyesters end‐capped with sugar molecules have been studied extensively for use in biomedical applications.1. The catalysts are metal alkoxides with Lewis acidic character [8. 9]. polydioxanone also shows good performance. 2). Exchange of the alkoxide group by e. Its melting temperature is close to 110 oC. The synthesis of protected monosaccharide end‐capped biodegradable polymers is usually performed via Ring Opening Polymerization (ROP). Besides
. 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. stannous octoate [Sn(Oct)2].g.Chapter 2
2. are interesting polymers because of their good product performance and biodegradability . Polydioxanone (poly(p‐dioxanone) or poly(1. Introduction
Aliphatic polyesters. stigmasterol and ergocalciferol)  and also with protected monosaccharides (galactopyranose/ glucofuranose) [6‐7].3 MPa and an elongation at break of 500‐600%. quinine. such as polycaprolactone. probably because the p‐dioxanone monomer has become commercially available only recently . testosterone. particularly for nano‐ encapsulation systems for drug delivery . which to the best of our knowledge is the first study to functionalise polydioxanone with a monosaccharide. The ring‐opening polymerization (ROP) of p‐dioxanone using metal catalysts such as aluminum isopropoxide [Al(OiPr)3]. tocopherol. From a biodegradability point of view. and polylactides. This chapter describes experimental studies on the catalytic ROP of p‐ dioxanone in the presence of a protected glucose molecule (1. reaction of the metal‐alkoxide with an appropriate alcohol allows the synthesis of end‐capped poly‐lactones. Despite its good properties.4‐di‐O‐ isopropylidene‐α‐D‐galactopyranose. only limited information about the synthesis and properties of polydioxanone is available in the open literature. whereas the high temperature for polylactides results in thermal degradation and undesirable transfer reactions during synthesis and processing steps . Polydioxanone has a tensile strength close to 48. pregnenolone.3. has interesting product properties compared to other aliphatic polyesters. accessible by the polymerization of p‐dioxanone (1). Several polymers with bioactive alcoholic and phenolic end‐groups of interest for drug‐related applications were synthesized (for example geraniol.
98+% (Aldrich) were used as received. However. (20 mmol) of aluminum isopropoxide to 20 ml of toluene.184.108.40.206.8 g. poorly soluble heterogeneous systems. 3]. NMR) due to the poor solubility of the products in standard organic solvents.08 mmol aluminum isopropoxide) was added to pure 1 (0.2.4‐di‐O‐ isopropylidene‐α‐D‐galactopyranose (2). The reaction mixture was clear and colorless during the reaction. the results of this study are also of interest for the preparation of starch/polydioxanone polymers using ROP. which was pre‐heated till about 60°C to obtain it in a liquid state. 0. containing 0.8 ml of a solution of an aluminum isopropoxide stock solution was added. 1. To this solution. Materials and Methods
2.2. part of this solution (350 µl. The hot solution was
. CDCl3 was obtained from Sigma and was used as received.1. 2. 97% (Sigma) and aluminum isopropoxide. and diethylether (Labscan) were used as received.07 g.5 mmol) was dissolved in toluene (1 ml) at 50 oC. Particularly interesting in this field are starch polymers grafted with polydioxanone. the mixture was brought to room temperature and several drops of HCl (1 N) were added to stop the reaction. heptane (Acros).4‐di‐O‐ isopropylidene‐α‐D‐galactopyranose (2) 2 (1. This stock solution was prepared by adding 4. 2. Boehringer Ingelheim.2. the synthetic pathways and the soluble. After the pre‐determined reaction time.3. The resulting clear solution was stirred for 2 h at 50 oC.44 g. 8 mmol).1.2.2. Typical example for the synthesis of polydioxanone end‐capped with 1. As such. Methods All polymerization experiments were carried out under a protective nitrogen atmosphere using standard Schlenk‐ and glovebox techniques. Materials p‐Dioxanone monomer (1.55 mmol of 2 and 0. relatively low molecular weight compounds reported in this study may be viewed as model systems for more complex. Next.
2. Subsequently. Germany) was purified according to the procedure described by Raquez et al [2. Toluene (Labscan) was dried and stored on molecular sieves 3 Å (Labscan) under nitrogen. grafting efficiencies are difficult to determine in this system by standard analytical techniques (e. 5. hot dichloroethane (20‐25 ml) was added to completely dissolve the partly solid polymer at room temperature. Analytical grade dichloroethane (Labscan). Ring Opening Polymerization of p‐dioxanone using an Al(OiPr)3‐Monosaccharide Initiator System
potential applications in biomedical products.g. The polymerization was allowed to proceed for 16 h at 100 oC.
H‐NMR was applied to determine the experimental average degree of polymerization. 2. X n exp of the product.3.
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). Calculation of Average Degree of Polymerization The Theoretical Average Degree of Polymerization. one initiated on a galactopyranose molecule and the other on an isopropoxide group. The isolated yield at this condition was 68%.)
Here. 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
H-isopropoxide end group
. 2. it is assumed that all available initiator is used effectively. As will be shown later.4.2. Processing of the raw data was performed using VNMR software. 1H‐ and 13C‐NMR spectra were recorded on a Varian AMX 400 NMR. The white solid was finally separated from the liquid by decantation and dried in a vacuum oven (5 mbar. 2D‐NMR spectra were recorded on a Varian Unity 500 NMR. MALDI‐TOF spectra were recorded on an Applied Biosystems Voyager DE‐PRO machine using dithranol/ NaI as the matrix (linear mode).2. 40 oC) until constant weight.2. 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
(2. two polymers with different end‐groups are present. the above equation simplifies to
= monomer conversion ×
[monomer ]0 [ sugar ]0
precipitated in a heptane/ ether mixture (300‐400 ml. Product Analyses NMR analyses were performed in CDCl3. 4:1 by volume) at 4‐8 oC.1.
6 ppm + A 5.3.1. R gp .1.7. The exchange reaction typically takes place at 50°C for 2 h.6: 96 was applied.6 ppm ] / 6 A 5.)
2.6 ppm A 5.1 ppm
(2.).3.). Typically.4‐di‐O‐isopropylidene‐α‐D‐ galactopyranose (2) and Al(OiPr)3 as the catalyst precursor (Scheme 2.7 ppm − 6 × A 5. an Al ‐ monosaccharide 2 ‐ monomer 1 mol ratio of 1: 6. 2. is calculated using:
R gp = A
H − galactopyranose end group H − isopropoxide end group
The ratio of the two different types of polymers (either end‐capped with galactopyranose or an isopropoxide group).1. 2.)
or. Schematic representation of the polymerization reaction including atom numbering scheme
. is rewritten as:
[ A 3.2.3. Screening Experiments Initial experiments to synthesize protected galactopyranose end‐capped polydioxanone (3) were performed using 1. in term of the NMR resonances :
R gp = A 5. To avoid the formation of isopropoxide end groups. In the first step.3.5. 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. The polymerization reaction was performed as a two step process. an excess of monosaccharide 2 on Al was used (see eq. the protected mono‐saccharide 2 reacts with Al(OiPr)3 to form the actual catalyst for the polymerization reaction.6.4.
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. Results and Discussions
2. To compensate for this effect. eq.
respectively. H‐NMR spectra (Figure 2. together with the CH3 groups from the protecting groups of the sugar appear in the range 1.1 ppm compared to the monomer. Product analyses The products were analyzed using 1H‐ and 13C‐ NMR and MALDI‐TOF..7.78 to 4.15‐1. the in‐situ formed catalyst was reacted with p‐dioxanone monomer 1 at 100 oC for 16 h.) are not particularly informative. The typical proton resonances of the p‐dioxanone (3.2. It is difficult to determine the end‐groups of the polymer on the basis of 1H‐NMR. although it is evident that p‐dioxanone polymerisation occurred.) In the next step.51 ppm. It confirms that polymer initiation not only occurs with the monosaccharide but also with the remaining isopropoxide group of the catalyst precursor (vide infra).1.1 and 2.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
(2.3. Typical 1 H‐ and 13C‐NMR spectra of 3 are shown in Figure 2. This multiplet is characteristic for the CH proton of an isopropoxide end‐group.
.1 ppm. The OiPr : 2 end‐group ratio for the standard experiment was 1: 2. 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.1. of interest is the presence of a small multiplet at about 5. Typical isolated product yields are 68% at these conditions.40 ppm) unit are broadened and shifted up to 0.1. Proton resonances of the CH3 group of the isopropoxide end‐group. The off‐white solid reaction product was collected after a dissolution‐reprecipitation process using dichloroethane and a heptane/ diethyl ether mixture.67. However. Although the spectra clearly indicate the presence of the monosaccharide 2. 2.
1 (c).) is more informative and clearly shows the presence of a polydioxanone polymer backbone and a monosaccharide end group.4‐di‐O‐isopropylidene‐α‐D‐galactopyranose‐end‐capped polydioxanone.4‐di‐O‐isopropylidene‐α‐D‐galactopyranose. Particularly the C2‐C6 carbons in the range δ = 66. 68. H‐NMR Spectra of: (a).2.3. 2 (b). C‐6 is shifted from δ = 62. In addition to the mono‐saccharide end‐ group.2.52 ppm (β’).1.2 ppm. For instance.8. Ring Opening Polymerization of p‐dioxanone using an Al(OiPr)3‐Monosaccharide Initiator System
Figure 2.3.2 ppm in product 3. 3
C‐NMR (Figure 2. The carbon resonances of the polydioxanone backbone are present at δ = 63.2‐70.2. 1.61 (γ’) and 73. characteristic resonances of an ‐O‐C(=O)‐CH2‐O‐CH2‐CH2‐OH end group are present at δ = 61.2 and 96. p‐dioxanone monomer.2 ppm.17 and 69.
.28 ppm in 2 to δ 66.3 ppm are shifted considerably. 1. Carbon resonances arising from the monosaccharide end‐group are present between δ = 66.
Figure 2.4‐di‐O‐isopropylidene‐α‐D‐galactopyranose‐end‐capped polydioxanone. 1 (c).2.. C‐ NMR Spectra of: (a). D‐NMR (HSQC) was applied for complete peak assignment of the product. p‐dioxanone monomer.73 and 109. the numbering scheme of carbons and protons is given in Scheme 2. 1. The resonances of the protecting group of the sugar appears at δ = 108. 2 (b). 1. and 2.2.76‐25.70 (>C‐(CH3)2) and between δ = 21.3.48 ppm (‐CH‐ (CH3)2) and between δ = 21.1.4‐di‐O‐isopropylidene‐α‐D‐galactopyranose.97 (>C‐(CH3)2).97 ppm (‐CH‐(CH3)2).3.76‐25.2.
. An overview of the data is given in Tables 2.2.1. 3
Carbon resonances of the isopropoxide groups from the by‐product (isopropoxide end‐capped polydioxanone) are present at δ = 68. A typical example of a part of the 2D‐NMR spectra is given in Figure 2.
59 4.32 * 4.
. except values with *.732 broad.1. which were determined from HSQC spectrum due to overlapping resonances.78 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 5.54 4. HSQC 13C‐1H spectra of 3 Table 2.00 4.17 3.8
Reactant peaks (ppm) Galactopyranose (2) H‐1 H‐2 H‐3 H‐4 H‐5 H‐6 5. around 1.40
all values were determined using 1H NMR.301 3.60 4.74 3.25 3.50 4.68 4.16* 4.09 4.26 3.77 3.3.31 4.83 p‐dioxanone Monomer (1)
H‐α H‐β H‐γ
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. The latter likely stems from the matrix used to ionize the sample.33 70. The molecular weight distribution of the major peaks may be represented by the following relation:
m / z = 23 + 260 + 102n
(2.11/ 170.4.28 70.92 68.Chapter 2
Reactant Peaks (ppm) Galactopyranose (2) C‐1 C‐2 C‐3 C‐4 C‐5 C‐6 96.28 p‐dioxanone monomer (1)
C‐α C‐β C‐γ C‐carbonyl
61.09 65. A typical molecular weight distribution is observed.57 62.52 63.
.58 68. corresponding with an average number molecular weight (Mn) of about 1600.54 166.21
NMR analyses also allow calculation of the molecular weight of the products. MALDI‐TOF was also applied to characterize the products. the ratio of the intensity of the end groups and the polymer backbone peaks is determined.00 171.31 70.)
This series represents a dioxanone polymer end capped with 2 and an additional Na ion.8. The product obtained at screening conditions (16 h reaction time at 100°C) contains on average 13 monomer units. A detailed procedure is given in the experimental section.17 69.2 68.2 73.21 70.63 68.75 70. For this purpose.61 170. An example of a MALDI‐TOF spectrum of 3 recorded in a dithranol/NaI matrix is given in Figure 2.2. The difference in molecular weight between the main peaks is 102 g/mol.10 71.80 61.92 66. which is the molecular weight of a repeating dioxanone unit.
see Figure 2.5.)
This relation is indicative for the presence of a polydioxanone polymer containing an isopropoxide end‐group. although both with a very low intensity (Figure 2.).5 +102
Figure 2.) and imply the presence of dioxanone polymers with carboxylic end groups.) and (2.10. Furthermore.
m / z = 23 + 18 + 102n m / z = 1 + 18 + 102n
(2.11. This series may be described by the following relation:
m / z = 23 + 60 + 102n
(2.4. ionized with either Na+ or H+.11.10. two other distributions are present.9. These distributions may be represented by equation (2.)
. Typical MALDI‐TOF spectra of 3 A second molecular weight distribution is also clearly visible when enlarging the spectra. 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. for details.5. in line with the NMR data.) (2.
is not particularly suitable for polydioxanone. The observed differences in the molecular weight distributions of the various samples were only marginal. Most likely the matrix (dithranol/NaI). In addition. Enlarged MALDI‐TOF spectra for 3 Although MALDI‐TOF clearly demonstrated the presence of various types of end‐groups in the product. Various samples with. In the first step.g.7
858. 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. 2‐(4′‐ hydroxybenzeneazo)benzoic acid (HABA)). small but significant amounts (20‐30%‐mol) of polydioxanone chains with an OiPr end group are present. the catalyst precursor is treated with monosaccharide 2 resulting in an alcohol exchange
. low resolution spectra were obtained.1.6.Chapter 2
792.7 796. 2.3
836. different molecular weights were analysed.13]. according to NMR.2.1
0 600 700 800 900 Mass (m/z) 1000 1100 1200
Figure 2. Mechanistic aspects Both NMR and MALDI‐TOF measurements imply that the main product is indeed the desired monosaccharide end‐capped polydioxanone. it proved not suitable to determine the average molecular weight of the products. but in all cases poor quality. [6. although successfully applied for galactopyranose‐end‐capped polycaprolactone .3. Various other matrices were tested (e.5.
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. During the reaction. termination of the chain growth may occur by reaction with an alcohol. Simplified reaction scheme for the ROP of p‐dioxanone catalyzed by Al(OiPr)3. 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. Subsequent coordination and insertion of dioxanone molecules leads to the formation of a polymer chain with a monosaccharide end‐group.
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 2.6. The termination reaction is known to be reversible and the formed polymer may again react with an aluminum center and continue to grow . Irreversible termination of the polymerisation is performed at the end of reaction period by adding dilute acid to the polymerisation mixture. a p‐dioxanone molecule will coordinate to the Lewis acidic aluminium center followed by an insertion step. (R = monosaccharide 2)
). Chain Length (Xn
Yieldb (%) 30.74 18.11 22.28 33.3 6.96
For each experiment.5 81.3‐6.9 2.6 86. will be formed when the polymerization starts with an aluminum alkoxide with a remaining Al‐OiPr group (eq.81 16.4 2. 2.Chapter 2
The minor product.6 t (h) 16 16 16 16 16 16 16 16 16 16 1. time and the mol ratio of monomer to monosaccharide) on the yield. An overview of the experiments and the results are given in Table 2.1 3. OiPr‐end‐capped poly(p‐dioxanone).1 86.6 6.50 25. two different reaction times (1.2. Table 2. 2.5 1.). a (nearly) fixed Al(OiPr)3 : monosaccharide ratio of 1 : 6.7 2.3 6. These may be present in the reaction mixture because the exchange reaction between Al(OiPr)3 and 2 was incomplete.6 6. A total of 15 experiments were performed at two polymerization temperatures (60 and 100 oC).6 6.6 2.22 57.50 35.8 3.4
2/OiPr ratio Rgp (mol/mol) 2.3 and 62.0 87.3.36 17.45 19.55 Product Properties Avg.6 6.3 6.08 30.1 2.74 52.3 6.22 16. degree of polymerisation of the product and the end group distribution was determined. In all cases. despite the excess of 2.6 6.8 2.70 58.6 91.5 1.10 15.21 54.5 67. 2.5 1.3.16 46.12 42.53 59.7.3 6.6 89.29 14.4 2.5 92.6 81.22 13.93 8.5 1.5. 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
.6 6.5 hrs and 16 hrs) and a p‐dioxanone to monosaccharide 2 molar ratio ranging between 3.3.9
7. the end group may also be formed by a termination reaction with free isopropanol.3 3.3 6.6 6.1 2.3 81.6 was applied.6 80.7.
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) 6.30 62.78 10.4 3.74 23. formed in the first step of the polymerisation reaction (eq. However.85 14.6 96.2 91.93 37.8 2.04 19.6 6.5 84. Systematic Studies The effect of important process variables (temperature.5 T (°C) 100 100 100 100 100 60 60 60 60 60 60 60 60 60 60 dioxanone/ 2 ratio (mol/mol) 3.49 13.5 2.
3.). the yields for reactions conducted at 60 oC are always higher than those performed at 100 oC.3. as suggested by Raquez et al . Clearly.5 and 16 h reaction time). For instance. the yields are higher when performing the reaction at 16 h reaction times. The effect of the reaction time on the product yield may be derived from the data provided in Table 2. equilibrium conversion is not yet achieved within 1.3. which reduced to 75% when increasing the temperature to 120oC. 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.3. see Table 2.5 h is higher than the yield at 16 h. A more detailed analysis on the actual nature and composition of the polymerization products. 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. Product yield The isolated yields of the reactions are all but one between 67 and 96% (see Table 2. the yield at 1. leading to a low average chain length of the dioxanone polymer (7. Higher polymerisation yields were obtained at lower temperatures. the monomer conversion reaches a maximum value before going down to the equilibrium monomer conversion.1. For the two experiments with a dioxanone: 2 ratio higher than 16 (≅ 23 and 60). Ring Opening Polymerization of p‐dioxanone using an Al(OiPr)3‐Monosaccharide Initiator System
2. with an enthalpy of polymerization of approximately ‐15 kJ/ mol . One of the experiments (S111) resulted in a very low yield (31%) compared to the other reaction.). It was shown that at lower monomer to catalyst ratios. a low dioxanone: monosaccharide 2 ratio was applied.3.5 h. when the reaction is performed at a low dioxanone: 2 ratio (≅ 16). Evidently. et al . However. and particularly when comparing the data in set 2 and 3 (60oC. The latter is due to the slight exothermicity of the reaction .2. the equilibrium conversion of dioxanone at 80oC was about 80%. For this particular experiment. 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. 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. no detailed explanation has been put forward to explain this anomalous behaviour. which is expected to be more pronounced at lower temperatures. On the basis of our experimental data and in line with literature data. will be required. Another possibility is the occurrence of polymer crystallization. 1. It is likely that this relatively low molecular weight compound dissolves partly during isolation/ purification procedure leading to lower isolated yields.2. 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. To the best of our knowledge.
3.2.4 and 3.). and 2.2.2. the X n theo (eq.
Temperature: 100oC. reaction time and temperature on the X n exp of the products is shown in Figure 2. to reduce the number of OiPr endgroups in the product may be the removal of isopropyl alcohol formed in step
. as expected for a typical ring opening polymerisation .5 hr Xn
40 Xn exp
30 40 50 dioxanone : 2 ratio (mol/mol)
Figure 2.) for all experiments are scattered randomly between 2. time. The R gp values (Table 2.6. in line with the theoretical predictions. A possible strategy to increase the Rgp values i. Effects of process conditions on the average chain length ( X n exp ) and end group distribution The effects of the p‐dioxanone: 2 ratio.e.3.5.3. and dioxanone/ 2 mol ratio) is absent..Chapter 2
2. 2.) as a function of the dioxanone : 2 ratio at 90% and 100% monomer conversion is also provided. 2.7. Most experimental points are scattered along these lines. time: 16 hr Temperature: 60oC. A clear trend between R gp and the process conditions (temperature.7. The X n exp increases linearly with respect to the 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. time: 16 hr
Temperature: 60oC. In Figure 2. time: 1.7.
) from the reaction mixture by e.459*10‐7
. A linear model proved adequate to describe the effects of the independent variables on the X n exp :
= 8.097 × (T )
(2. The 2/ Al ratio was not included in the model as the experimental range (6. Statistical Data Analysis Quantification of the influence of the experimental factors (temperature.3.8. vacuum distillation before adding the dioxanone monomer .177 × (t ) − 0. indicating that the model describes the experimental data reasonably well. Ring Opening Polymerization of p‐dioxanone using an Al(OiPr)3‐Monosaccharide Initiator System
1 of the polymerisation process (eq.854 × (dioxanone : 2 ratio) + 0.4. Analysis of variance (ANOVA) for linear model of X n function of experimental parameters
SS 4103 93 4010 DF 3 12 15 MS 1337 7.3. The low P‐value clearly indicates that the model is statistically significant.7. Table 2.g. A parity plot of the modeled versus experimental values of the average chain length Xn confirms this statement (Figure 2.6) was too limited to draw sound conclusions.3‐220.127.116.113 F 173.974). The analysis of variance for the model is given in Table 2.2.977 (with an adjusted R2 value of 0.)
where t and T are respectively the polymerization time and temperature. 2. The R2 value for the model is 0.12.543
Model Error Total
P‐value 3.49 + 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. 2.
11] and due to the fact that the reaction is an equilibrium polymerization with a slight exothermic effect. 16. with high ratios leading to a higher average chain length.
The ROP of p‐dioxanone in the presence of a monosaccharide (1. The isolated yields of the off‐white solid products were between 30 and 96%. 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). Within the experimental ranges. with higher ratios leading to higher molecular weights.8. This is in line with literature data [5. 17]. 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. which is in agreement with the available data on ring opening polymerization [15.4‐di‐O‐ isopropylidene‐α‐D‐galactopyranose.3.Chapter 2
40 Xn model
Figure 2. Furthermore. the model predicts that the X n exp is positively influenced by the polymerization time. 2) with Al(OiPr)3 as the catalyst is reported. Both NMR and MALDI‐TOF measurements indicate that the products are mixtures of polymers and significant
. In addition. 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.4.2.
Biomed. 8419‐8425. 1993. Phys. 34. Dubois: Some thermodynamic. Res. Rapid Commun.J. Chu.M.M. Lin. starch). Chem. Macromol. Narayan.4‐dioxan‐2‐one and controlled synthesis of diblock copolymers with epsilon‐caprolactone. A statistical model has been developed to quantify the effects of process variables (time. P. References
. P. temperature and monomer: monosaccharide ratio) on the X n exp . R. kinetic.R. Damrau: Polylactones. Macromolecules 2001. Raquez. 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. D. Macromol.g. and mechanistic aspects of the ring‐opening polymerization of 1.L. J. P.6. H.4‐dioxan‐2‐ one initiated by Al((OPr)‐Pr‐i)(3) in bulk. 153‐166.
. J. Kricheldorf. Narayan. P. 2000. 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
2.4‐dioxan‐2‐one.O. H.5. .C. . Degee. 42. 199. Degee. C. 1063‐1071. 1089‐ 1097. 1998.
2. Moreover. . Raquez. Grubb: Hydrolytic degradation and morphologic study of poly‐p‐dioxanone. These studies will be reported in the next chapter. 21. 27. D. Zn‐lactate‐catalyzed polymerizations of 1. Dubois: ʺCoordination‐insertionʺ ring‐ opening polymerization of 1. R. 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. Mater.
. 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.
45‐0. ε‐ Caprolactone is grafted to silylated starch by a ring‐opening polymerisation catalysed by Al(OiPr)3 in THF at 50oC. The DS of the polycaprolactone graft is between 0. biodegradable polymers. grafting. The silylation reaction was performed using hexamethyl disilazane (HMDS) as the reagent in DMSO at 70°C. polycaprolactone.21 and 0. The grafting efficiency varies between 28 and 58%. silylation
. 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. Keywords: starch.72. Silylated starch with a degree of substitution (DS) between 0.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.7 was obtained. The poly‐(ε)‐caprolactone side chains consist of 40‐55 monomer units and is a function of the reagent intakes. 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 remainder being homopolymers of ε‐caprolactone.
4‐glucosidic bonds) and amylopectin (a branched polymer with α‐D‐1. Starch is abundantly available. Well known examples are polyesters derived from cyclic lactones (polycaprolactone. 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. Starch films are known to have good oxygen barrier properties. There are two types of biopolymer in starch. resulting
. However. it is water sensitive. for instance grafting of monomers (like styrene and methyl methacrylate) to the starch backbone [6. Starch has also been thermoplasticized with the help of plasticizers such as glycerol and other polyalcohols . The polymeric structure of starch consists of repeating anhydroglucose units. 7]. and this value increases with about 10% per year . However. The development of green biodegradable polymers for e. amylose (a linear polymer of anhydroglucoses with α‐D‐1. and the mechanical properties of starch‐based films are generally inferior to those derived from synthetic polymers [4. Introduction
Worldwide. Aerobic soil‐burial experiments showed that the mechanical properties of PCL films decreased rapidly in time and were fully degraded after 4 weeks . the product properties are in most cases still not up to standards and blending with other polymers is required . The amylose‐amylopectin ratio in native as well as modified starches has a strong impact on the product properties.g.6‐glucosidic bonds besides α‐D‐1. a natural biopolymer. The content of amylose in starches depends on the plant and typically varies between 18‐28%. roots) as granules or cells with typical particle sizes between 1‐100 µm. Global production of starch is 60 million ton per year in 2004 . Starch is present in the body of many plants (tubers.1. However. Polycaprolactone (PCL) is easily degraded by micro‐organisms . the future generation of packaging materials is highly desirable. A wide variety of synthetic biodegradable polymers have been prepared. 5]. in almost all cases. Various modification strategies have been explored. as starch is highly hydrophilic. These plastics are mainly synthetic polymers from fossil resources. Starch.4‐ glucosidic bonds). polyvalerolactone. 245 million tons of plastics are produced per year. Starch modification is therefore needed to meet the product properties in a number of application areas. Blends of thermoplastic starch and PCL are not fully miscible. and polybutyrolactone). is one of the potential candidates for future biodegradable polymer products. which are known to degrade with difficulty and cause serious environmental problems . the used monomers and the corresponding grafted chains are not easily biodegradable. Chapter 3
This approach has also been applied successfully to graft PCL and polylactide on dextran [19‐20]. ε‐Caprolactone monomer (Fluka) was dried over Calcium Hydride (CaH2) for 48 h. A stock solution was prepared by dissolving 1. The highest GE (90%) was achieved using triethylaluminium as catalyst . distilled under reduced pressure at 100 oC and stored under a protective nitrogen atmosphere. In this chapter. Common Ring Opening Polymerization (ROP) catalysts such as tin octoate or aluminium isopropoxide gave low grafting efficiencies (GE. during the reaction. This catalyst is extremely air and water sensitive. Aluminium triisopropoxide [Al(OiPr)3] (Acros) was used without further purification. In this way. Starch is insoluble in the typical organic solvents used for ROP (such as toluene or THF). 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]. All available data indicate that the presence of water reduces the GE. 0‐14%). DMSO (Acros) and toluene (Labscan) were dried overnight over molecular sieves 3 Å (Merck) and stored under a protective nitrogen atmosphere. thus to a reduction in the GE.2. Hydrochloric acid (HCl) 1 N was prepared from Titrisol concentrated hydrochloric acid solution (Merck) and distilled (Milipore) water. leading to a liquid‐solid system. a very flammable by‐product. Another possible cause for the low grafting efficiencies is the heterogenous nature of the reaction. Materials Corn starch (Sigma) was dried at high vacuum (~1 mbar) at 100 oC for one day before use. Dry tetrahydrofuran (THF) and toluene for polymerization experiments were obtained in closed vessels from Aldrich and were used as received.
3. This is expected to lead to reduced reaction rates between starch and CL compared to CL homopolymerisation.1. Materials and Methods
3. Hexamethyldisilazane (HMDS. To increase the miscibility of starch and polycaprolactone. it has been proposed to chemically graft caprolactone onto the hydroxyl groups of starch using ring‐opening polymerisation .
. an alternative method to synthesize poly‐(ε)‐caprolactone grafted starch co‐polymers (starch‐g‐PCL) is reported. the ring opening polymerisation occurs solely in the liquid phase and this is expected to lead to higher GE values.2. Acros) and methanol (Labscan) were used as received.67 gram (8 mmol) Al(OiPr)3 in 50 ml of dry toluene in a glove box. therefore difficult to handle and releases ethane. Synthesis of Poly‐(ε)‐caprolactone Grafted Starch Co‐polymers using Silylated Starch Precursors
in undesirable phase separation . This is rationalised by assuming that water competes with the hydroxyl groups of starch in the initiation step of the polymerization reaction. thus leading to the formation of PCL homopolymers rather that starch‐g‐PCL .
For each experiment.
3. silyl‐CH3). The samples were characterized by 1H‐NMR. 3‐6 ppm (m.2. 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. Typical example of in situ polymerization of ε‐CL with silylated starch The silylated product from the first step was dissolved in THF (0.1 mbar. Silylated Starch (1. 3‐6 ppm (m. 50 oC): δ 0. 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. The pre‐determined amount of HMDS (typically 24 ml. This procedure was repeated three times.12 (m.6 mL) at 50 oC (1‐2 h). toluene (40 ml) was added to solubilize the precipitated (partially silylated) starch. before peracetylation.7‐2. A mol ratio of ‐OH groups to catalyst of 10:1 was used.2. THF (4. After 6 h. Sample SN‐3. DS = 0. starch). To this homogenous solution. Subsequently. Typical example of the starch silylation procedure The procedure for corn starch silylation was adapted from that published for dextran [19‐23].1.2.60): H‐NMR (CDCl3.2. 80 oC). The reaction was carried out at 70 oC. 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. acetate‐CH3). Chapter 3
3. Traces of DMSO trapped in the product were removed by dissolving the product in a small amount of toluene and re‐precipitation in methanol. DS = 0. 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. Sample SN‐3. broad peaks.60):
H‐NMR (CDCl3. 3. The white‐to‐transparent solid products were stored under vacuum in a desiccator at 6‐8 oC. broad peaks. another portion of toluene (20 ml) was added. 40 oC) until constant weight.2.111 mol) was added to the gelatinized mixture to initiate the silylation reaction. a predetermined amount of ε‐caprolactone monomer (molar ratio of monomer to
.5 (m.2. the solvents were removed from the silylated starch product under reduced pressure (~ 20 mbar) at 70 oC. after peracetylation. 50 oC): δ 0.12 (m.
Silylated Starch (1.5 ml) and a predetermined amount of the stock solution of Al(OiPr)3 in toluene were added. 0. After removal of the solvents under reduced pressure (0. the silylated starch (1) product was dried in a vacuum oven (~5 mbar. After 24 h. After 2 and 4 h reaction time. starch). silyl‐CH3). 1.
31 (m.3‐4.88 ppm (m.75. broad peaks. α‐PCL). starch). α‐PCL).98 (t.5‐3.) was >99 %.4 (m.2. DSPCL=0.4 (m. 2.34): H‐NMR (DMSO. The peracetylation procedure applied in this study was adapted from the literature [24. ε‐PCL).16 (d.88 ppm (m. The product was collected as a white solid and characterised by 1H‐NMR.25 (t. The silylated starch‐g‐PCL (2) product was precipitated from the solution by the addition of heptane (about 250 ml) at ‐18 o C. DSsilylation = 0.21): H‐NMR (DMSO d‐6.75.54 (m. 60 oC): δ 0.68. and 5‐5.2.12 (s. β and δ‐ PCL). 1. 1. After stirring for 2 h.2. Subsequently. filtrated. 4. and 5‐5. Typically. ‐CH. γ‐PCL). 25].2.3. Sample SN1CL1. 4. broad peaks. the desilylated starch‐g‐polycaprolactone product (3) was precipitated using heptane.37 (t. iPr).4.25 (t. 3. 60 oC): δ 1.5‐ 3. β and δ‐PCL). The samples were characterized by 1H‐NMR. ‐CH3 .
3.16 (d. 4.5. 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). Starch‐g‐PCL (3. ‐CH3. γ‐ PCL). 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. DSsilylation = 0. 2. 3. 1.1 g) was suspended in THF (4%‐w/v) and stirred at 55 oC until it was fully dissolved (typically 3 h). 1 (0. 3. silyl‐CH3).68. Deprotection product of Sample SN1CL2. 3. The total isolated yield at this condition (see Table 3. The solid precipitate was filtered and dried under vacuum (~ 5 mbar) at 40 oC for 48 h. 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. ε’‐PCL). 3. The yield was measured gravimetrically and is based on the weight of the product and the total weight of reactants charged to the reactor. iPr). 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. Peracetylation of the remaining OH groups of modified starch is a well established procedure to improve characterisation of the products by NMR . ‐CH. iPr).2. ε‐PCL). DSPCL=0.54 (m. ε’‐PCL). Silylated Starch‐g‐PCL (2. 1. starch).3‐4. 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. 3.5.31 (m. 4. the peracetylating reagents
. and vacuum dried at 40 oC for 24 h. iPr).37 (t. 1.
ACH 3 silyl / 9 1 7 A−0. acetic anhydride and pyridine in a DMAP : acetic anhydride : pyridine : AHG molar ratio of 1: 10: 22 : 1) were added. 3. When assuming that all ε‐CL monomer is converted.1. the DSPCL may be calculated using eq. This leads to the following equation:
ACH 2 −ε + ACH 2 −ε ' ACH 2 −ε '
A3. The peracetylation reaction was conducted for 7 h at 50 oC.3.3−3.)
(3.4. 3. Chapter 3
(DMAP. it is assumed that the average length of the grafted chain is equal to the chain length of the homopolymer.
3.2. 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.). 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 3.2. 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.2 ppm A3.2. The product was precipitated by the addition of methanol and washed several times with methanol.)
where Ax‐y stands for the 1H‐NMR peak area in the range δ x‐y ppm.8− 4.6−0.4 ppm
(3. The assumption of high conversions (>95%) was correct for all experiments (see Table 3.1.6 ppm DSsilylation = × = × 3 Astarch proton / 7 27 A3−5.3. In this calculation.3. 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. Analytical Methods 18.104.22.168.)
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 was finally dried overnight in a vacuum oven at 70 oC and 5 mbar until constant weight. DSsilylation may be calculated using 1H‐NMR spectra of the products after peracetylation using eq.1.
( A3. and Vm stands for molar volume.3.8− 4. A3. 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.3.3−3.9 ⎟ × 100 % ⎜ ⎜ ⎟ ACH 2 −ε ' / 2 ⎠ A3.9
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).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. 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. 2007)
3.4 / 2 ⎟ ⎠ ⎝ ⎝
where Ax‐y stands for the peak area in the range δ x‐y ppm.2 ) ⎤ 1 × ⎢ (3 − DSsilylation )⎥ ⎢ ε . Synthesis of Poly‐(ε)‐caprolactone Grafted Starch Co‐polymers using Silylated Starch Precursors
DS PCL =
0. T stands for absolute temperature.3−3.)
⎛ ⎞ ⎞ ⎛ A A = ⎜1 − CH −iso ⎟ × 100 % = ⎜1 − 4. Although all steps have been investigated.4 + A3. The values of ∆Hv and Vm were obtained from the SciFinder Scholar database (American Chemical Society.8− 4 .4 − A4 .4.5. the focus of this chapter will be on the first two steps of the procedure. 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).CL/OH ratio ⎥ ⎣ ⎦
(3.8− 4 . This leads to the following equation:
ACH 2 −ε ' / 2 − ACH −iso ACH 2 −ε ' / 2
× 100 %
b. 24 h). Synthesis of Silylated Starch The silylation of corn starch was performed with HMDS as the silylating agent (eq.
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.3. the starch peaks are very broad and this feature hampers accurate determination of the DSsilylation.5. Chapter 3
3. a homogeneous reaction mixture was maintained throughout the reaction. The 1H‐ NMR spectrum of silylated starch (Figure 3. The silylation procedure was adapted from that previously reported for dextran [19‐23].). The degree of substitution of the silyl groups (DSsilylation) was determined by 1H‐ NMR. In this way.
Instead of using DMSO as solvent. a product with a DS of 0. Using standard conditions (HMDS: AHG molar ratio of 3.60 was obtained. 3. An additional peracetylation procedure to substitute the free hydroxyl groups with acetate groups was performed to improve the quality of the NMR spectra.1. The 1H‐ NMR peaks from the AHG unit after peracetylation were indeed considerably sharper and allowed more accurate DS calculations (Figure 3. However.a.1. mixtures of toluene and DMSO were applied to avoid precipitation of the silylated starch during the reaction. 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.).1.). It is defined as the average number of silyl substituted OH groups on the anhydroglucose (AHG) unit of starch.1.) shows the presence of silyl groups at about δ 0 ppm. C
O H 2OR
[R = H or Si(CH3)3] Starch Hexamethyldisilazane Silylated Starch (1)
(3. The silylated products were characterised by NMR.a.5. as suggested by Einfeldt et al . 70 oC.
. Synthesis of Poly‐(ε)‐caprolactone Grafted Starch Co‐polymers using Silylated Starch Precursors
Figure 3.2. not peracetylated (b).60. in CDCl3. and Figure 3. 50 oC): (a).5‐4.1.5) on the product DS was investigated by performing experiments with a constant starch and a variable HMDS intake.1. The results are given in Table 3. Typical 1H‐NMR Spectra of Silylated Starch (Sample SN3. peracetylated The effect of the HMDS to starch molar ratio (1. DSsilylation=0.
Experiment SN1 SN2 SN3 SN4
Experiments were performed in DMSO at 70°C.5
Figure 3. For all experiments. mmol) 12 (56) 18 (83) 24 (111) 36 (167) HMDS: AHG mol ratio 1.65
0.5 Product DS (DSsilylation) 0. Its presence will reduce the polarity of the reaction medium (DMSO.
Surprisingly. The experimentally observed lowering at higher HMDS intakes is likely due to a decrease in the polarity of the reaction mixture. HMDS is a rather apolar compound (Hildebrand solubility parameter of 6. The silylation reaction
.5 HMDS : AHG Ratio [mol/ mol]
0. At a mol ratio of HMDS to AHG of 4.2. the volumetric HMDS intake is about half of the DMSO intake.5
2. Chapter 3
0. solubility parameter = 11.5:1.68 0.5 3 3.4
1.25 3 4.36 cal1/2cm‐3/2) considerably. a positive effect of higher reagent intake on the reaction rate and thus the DS is expected. 70 oC) Table 3.25 cal1/2cm‐3/2) due to the presence of the bulky apolar SiMe3 groups.45
0.1. a fixed starch intake of 6 g (37 mmol AHG) starch was applied.67 0.60 0. Effect of HMDS: AHG mol ratio on DS of the silylation product a
HMDS intake (ml. DS of the silylated products at different HMDS: AHG ratios (DMSO. the DSsilylation decreases for higher intakes of HMDS. For reactions with an order higher than zero.55
68. 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.
O O H2C O R H H OR H O O H OR H O Catalyst: Al(OiPr)3 Toluene. To the best of our knowledge. the silylation of starch in pure DMSO using HMDS as the reagent has not been reported to date.6‐glucosidic for dextran).6. 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 use of HMDS to silylate dextran (Mw=6000‐40000) in DMSO (HMDS to OH molar ratio of 0. The products are soluble in DMSO as well as in less‐polar solvent such as chloroform and THF. Petzold et al  reported that silylation of starch with trimethylsilylchloride (TMSCl) in pyridine yielded trimethyl‐silyl substituted starch with DS values between 0. After precipitation with heptane and vacuum drying.0 .25‐5. 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=0.2. liquid ammonia.7‐3. A schematic representation of the reaction is provided in eq. Silylation of starch with HMDS in formamide. 3. white solid products with isolated yield > 96% were obtained. DMF. 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. 3.)
.3‐2. and DMSO/pyridine mixtures yielded silylated starch with DS values ranging between 0.45‐0.7) of the products is in the range of those published for other starch silylation systems.2.0 [18.0) gave silylated dextran with DS values between 1. 29].6. 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)
(3. DMA/LiCl.1 and 3. sample SN‐1) in THF at 50 oC for 24 h using Al(OiPr)3 as catalyst.3. The DSsilylation (0. pyridine.
The homopolymers are formed by direct polymerization of ε‐CL initiated on isoproproxide moieties attached to the Al catalyst.2 ppm.4.8 and 5. broad peaks in the region δ 3. Apparently. DSsilylation =0. not all of the caprolactone is grafted to starch and large amounts of PCL homopolymers (72%) were formed. Resonances from the starch peaks are observable as small.4. The formation of CL homopolymers for this type of reactions has been observed before [15.
Figure 3. Typical 1H‐NMR spectrum of a silylated starch‐g‐PCL sample. A typical spectrum is shown in Figure 3.68. The peaks from the polycaprolactone units are clearly present in the range of δ 1. Further process optimization studies allowed the synthesis of products with less than 42% of homopolymers (vide infra).2‐4 ppm and imply that the ring‐opening polymerisation reaction of CL indeed occurred.4 ppm. 26]. the exchange reaction between Al(OiPr)3 and the OH groups of silylated starch is not quantitative under the conditions applied in this study.4‐22.214.171.124) in DMSO‐d6 at 60 oC. However. Coding of the peaks is given in Figure 3.
. as is shown in Figure 3.0‐5.9 and 1. DSPCL=0. Chapter 3
The products were characterized using 1H‐NMR analysis in DMSO‐d‐6 as the solvent. (Sample SN1CL1. This is clearly indicated by resonances of the iPr end‐group of the PCL homopolymer at δ 4.
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. Exchange Reaction: iPr O Al O iPr iPr O + H2O iPr O iPr O Al OH + iPr OH
2. 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. 3. residual water may also initiate the polymerisation reaction. Synthesis of Poly‐(ε)‐caprolactone Grafted Starch Co‐polymers using Silylated Starch Precursors
1. However. peaks arising from a ‐COOH unit could not be detected in 13C‐ NMR spectra (δ 175‐180 ppm).4. 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. This leads to the formation of carboxylic end groups (see eq.7.)
1. Product formation for the reaction between silylated starch and ε‐ caprolactone Besides the ‐OH group of starch.7.
In the case of only homopolymerisation.0 15.68) in THF at 50 oC with Al(OiPr)3 as the catalyst (1 mol Al(OiPr)3 per 10 mol‐starch‐OH groups). Five experiments were performed to study the effect of different ε‐CL to silylated starch ratio. in all samples. Determined gravimetrically and defined as the total weight of the isolated product divided by the total intake of reactants (silylated starch and CL). and Figure 3.2. and reaches 58% for a ε‐CL to starch–OH ratio of 29. The ACL of the polymer and the DSPCL increase almost linearly with the ε‐ caprolactone intake (Figure 3.5 29.21 0. (SN1CL1) is 28%. Chain Length [mon. The results are given in Table 3.2. This implies that the ε‐CL conversion is essentially quantitative in all cases.
Overview of results for the grafting of ε‐CL on silylated starch a
ε‐CL/ OH [mol/ mol] 13.5 and 100%. This implies that grafting of caprolactone to starch also occurs to a significant extent.
Experiment SN1CL1 SN1CL2 SN1CL3 SN1CL4 SN1CL5
. the intensity of the resonance ε’ was considerably higher that that of the ‐CH‐ iPr peak. The GE increases with higher ε‐CL intakes.5. However. units] 40 43 44 49 54 DSPCL 0. The yield of the products was measured gravimetrically and varies between 96.0 18. the intensity of the peak from the ‐CH2‐ end group of the homopolymer (ε’ at δ 3.).34 0. Chapter 3
The ratio of homopolymerisation versus grafting on starch may be obtained by comparing the integrals of selected peaks in 1H‐NMR spectra.5 Avg.9 ppm). The grafting efficiency (GE) for the sample given in Figure 3.3.
b.2 Total Yieldb [%] >99 >99 >99 99 96.2.5. Table 3.72 Grafting Efficiency [%] 28 43 48 55 58
All reactions were performed using the same intake of SN1 silylated starch (DS=0. 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.3 ppm) should be twice the intensity of the ‐CH‐ proton of the isopropoxide end group (δ 4.58 0.9 22.
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)
the percentage of PCL grafted to starch compared to the total amount (grafted and homopolymer) of polymerized CL [%] : gas constant. 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. The products may have interesting applications as compatibilizers for starch‐polymer blends.72. Considerable amounts of ε‐CL hompolymers with isopropyl end‐groups were also formed.68) was obtained using a DMSO/toluene mixture as the solvent and HMDS as the silylating agent. 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. Silylated starch with a low‐medium DS (0.5. This was achieved by making the starch more hydrophobic by partial substitution of the OH groups by trimethylsilyl groups.
3. The DS of the PCL chains was between 0. The ROP with ε‐CL was performed using Al(OiPr)3 as catalyst in THF as the solvent. 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. 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.21‐0.46‐0. depending on the ε‐CL‐starch ratio. Chapter 3
3. The key feature is the use of a homogeneous reaction mixture for the ROP of starch with ε‐CL. The grafting efficiency varied between 28‐58%. indicating that homogeneous reaction conditions are favorable for the grafting reaction.4. Conclusions
The successful synthesis of poly‐(ε)‐caprolactone grafted corn starch co‐ polymers using a three step approach is reported.986 cal mol‐1 K‐1 : temperature [K]
molar volume [cm3/ mol]
J. Karlsson. . Aarhus. 2003. Association of Plastics Manufacturers. Polym. 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. . Yavuz. 51.dk/isi/stat/rawmaterial. 311‐316. 1998. Aly. F. M. Y.. 1996. 53‐57. accessed on August 01. 61‐67. 11. C. . 56.R. Solaro: Biodegradable polymeric materials. Denmark: http://www. Polym. Eng. Polym.. Chiellini. the Royal Veterinary and Agricultural University. A. Chandra. 305‐313. Biobased packaging materials for the food industry. 17. R. 407‐412. Sci. Starch‐Starke 2004. Krochta. Environmen.A.F. D. Yang.html. E.
. Brussels. 3. R. J. 2000.Sci. demand and recovery for 2006 in Europe. C. H. R. G.M. The compelling facts about plastics: An analysis of plastics production. L. . Environ. 2008. Goldberg: Review of the biodegradability and utility of poly(caprolactone). Burr. .starch. K.C. J. Wang. 61‐74. E. Polym. Rustgi: Biodegradable polymers. Degrad. 1998. Russell: Graft copolymers of polysaccharides with thermoplastic polymers ‐ new type of filled plastic. W.A. Beliakova. S. Frederiksberg C. . 107‐113. Fanta. Sci.6. 1997. 1977. References
. Prog. Polym. 23. C. . Belgium. J. R. Starch‐Starke 2000. 43. Website of International Starch Institute. Denmark. status and perspectives. Wang: Properties of starch blends with biodegradable polymers. Rev. Bagley. C. 52. Sci. Food Technol. report of Food Biopack Project (Ed.C. 1273‐1335. Weber). De Mulder‐Johnston: Edible and biodegradable polymer films: Challenges and opportunities.M.K. Babac: Preparation and biodegradation starch/polycaprolactone films. Linke: Water vapor sorption determination of starch based porous packaging materials. Mater. 2008. . 2003.A. 385–409 . A. Adv. J. Tsiapouris. 1251‐1253. Abdel‐Mohdy: Grafting of poly(methacrylic acid) on starch and poly(vinyl alcohol). 1995. Albertsson: Biodegradable polymers and environmental interaction. X. .B. Plastics Europe. A. . Department of Dairy and Food Science. Eng. 8. 38. of
. Part C: Polym. Doane. Macromol.
K. I. . K. Chem. D. Biosci. J. P. Dellacherie. Petzold. Klemm: Preparative and 1H NMR investigation on regioselective silylation of starch dissolved in Dimethyl Sulfoxide. Degee. E. . Gunther. Macromolecules 2000. . . Dubois. Dellacherie. A. Degee. D. Stab. Averous. Six: Silylation reaction of dextran: Effect of experimental conditions on silylation yield. M. Duquesne. Stein. P. 415‐426. Monomers Polym. Klemm. . P. Six: Controlled synthesis of amphiphilic biodegradable polylactide‐grafted dextran copolymers. 35‐47. Symp. Fringant: Properties of thermoplastics blends: starch‐polycaprolactone. 4. I. 73. 41. Petzold. Stein. Klemm: Regioselective functionalization of starch: Synthesis and 1H NMR characterization of 6‐O‐ silyl ethers. 965‐969. P. Rutot. 6713‐ 6721. Part A: Polym. Ydens. Ydens. Degrad. Einfeldt: Structure design of polysaccharides: novel concepts. Dubois: Aliphatic polyester‐ based biodegradable materials: new amphiphilic graft copolymers. 561‐566. J. Pantiru. C. selective syntheses. Klemm. H.L. 2001. M. Dellacherie. . Ydens. Polym. 2001. 4157‐4167. P.L. L. Polymer 2002. . Dubois. Einfeldt. Fessi. Nouvel. Boullanger: Ring‐opening polymerisation of epsilon‐caprolactone with monosaccharides as transfer
. K. L. J. 2001. R. P. A. D. 163. W. Gunther: Synthesis and NMR characterization of regiocontrolled starch alkyl ethers. . high value applications. K. Chapter 3
. W. Dellacherie. . structure analysis. L. Krishnan. P. D. 42. Einfeldt. 2002. L. Macromol.L. 1443‐1450. Dubois. Dubois. J. Sci. 251‐269. 2. Hamaide. B. Nouvel. P. E. M. Koschella. regioselectivity. Petzold. Biomacromolecules 2001. 5. C. Klemm. Cellulose 2003. Polymer 2000. Narayan: Aliphatic polyester‐grafted starch‐like polysaccharides by ring‐opening polymerization. 2004. Moro. J. Degee. T. 33. 1. Six. and chemical stability of silylated dextrans. L. Polymer 1999. 40. P. D. P. Stein. Dubois: Controlled synthesis of poly(epsilon‐caprolactone)‐grafted dextran copolymers as potential environmentally friendly surfactants. 3091‐3100. Macromol. Polym.. Kussler.L. W. D. P. A. Gunther. Petzold. Biomacromolecules 2003. . 10. . and subsequent reactions. 43. C. 2577‐2588. 1735‐ 1743. Six: Partial or total silylation of dextran with hexamethyldisilazane. I. Heublein: Silylation of cellulose and starch ‐ selectivity. E. D. Rutot. 341‐347. E. . A. Nouvel. E. Dole. D. Des.
Junistia. 553‐556. Hertz Jr. Kricheldorf. Synthesis of Poly‐(ε)‐caprolactone Grafted Starch Co‐polymers using Silylated Starch Precursors
agents. T. . . presented at ACS Rubber Division meeting. Heeres: Synthesis of higher fatty acid starch esters using vinyl laurate and stearate as reactants.R.K.4‐ dioxan‐2‐one. Rapid Commun.J. A.P. A.L. . 199. H. 941‐950. Macromolecul. Macromolecules 1996. Horton. L. Rapid Commun. R. D. . Carbohydr. J. 659‐663. A novel route to functionalised nanoparticles. Manurung.
. Penczek. Mexico City. Res. D. 61. 1399‐1406. Lehmann: Selective 6‐O‐acetylation of amylose. S. Picchioni. F. Duda: Polymerization of ε‐caprolactone initiated by aluminum isopropoxide carried out in the presence of alcohols and diols. Sugih.O.B. 1989 . 1089‐1097. . 1978. Phys. A. Damrau: Zn‐lactate catalyzed polymerizations of 1. Macromol. H. 1998. L. Biela. Duda: Living polymerization with reversible chain transfer and reversible deactivation: the case of cyclic esters.: Solubility parameter concepts – a new look. 22. Kinetics and mechanism. 2001. Macromol Chem. 21. 29.M. Janssen. accepted for publication in Starch‐Starke. 2000. D. 2008.
vinyl ester to starch ratio and the type of catalyst.96). presumably due to a reduction of the polarity of the reaction medium. the DS decreases. 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. K2CO3. The yellowish products were characterized by 1 H‐. When performing the reactions using Na2HPO4 as the catalyst. and Na‐acetate are superior catalysts with respect to activity compared to Na2HPO4 and products with DS values close to 3 were obtained. The fatty esters were prepared by reacting the starch with vinyl laurate or vinyl stearate in the presence of basic catalysts (Na2HPO4. the DS for the starch‐laurate compounds is higher than for the corresponding starch‐ stearates. an increase in the vinyl‐ester concentration leads to higher product DS values. 13C‐NMR and FT‐IR. vinyl stearate
. esterification.24 ‐ 2. For low vinyl‐ester to starch ratios. K2CO3. vinyl laurate. At higher ratios. The DS of the products is a function of the carbon number of the fatty acid chain. Keywords: corn starch.
most of the studies performed to date use short chain carboxylic acids (C1‐C4). Introduction
Green biodegradable polymers derived from natural resources are potentially very interesting substitutes for non‐biodegradable petroleum‐based polymers.1. cellulose or proteins are potentially very interesting starting materials for biodegradable packaging materials. Further modification of starch is therefore required to introduce hydrophobicity and to improve mechanical and moisture barrier properties. Their
. and consequently. The hydrophobicity increases with the degree of acetate substitution (DS.7 [1. An attractive field of application for these polymers is the use as packaging materials. The mechanical properties and hydrophobicity of the products were significantly improved when using these longer chain fatty acid precursors [1. defined as the moles of substituents per mole of anhydroglucose (AHG) units) . However. poor processibility (high viscosity). and even fatty acid derivatives (C12‐C18) have been used in the modification reaction [5. The first paper on the acetylation of starch was already published in 1865 . 5]. which are relatively expensive and rather corrosive . However. The introduction of acetate groups on starch makes the product more hydrophobic. However. the use of native starch for packaging materials is limited due to its low moisture resistance.34‐0.61) products were obtained using this approach. For the current petrochemical based products recycling is often neither practical nor economically feasible . Recently Mormann et al  explored the possibility of using vinyl esters and particularly vinylacetate as reagents for the preparation of starch esters. and incompatibility with hydrophobic polymers. more water‐resistant products may be obtained [3‐4]. The major obstacle is the pronounced brittleness of the materials. 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. even after the addition of plasticizers . However. high brittleness. An alternative method using methyl and glyceryl laurate esters in the absence of solvent has been recently developed . the fatty ester substituents [1. 7]. 5]. In particular starch is attractive as it is relatively cheap and abundantly available. and particularly acetic acid derivatives (C2) [2‐4].Chapter 4
4. Esterification of starch with low molecular weight fatty acid derivatives is one of the oldest modification technologies to improve starch properties. Relatively low‐DS (0. 5‐6] were introduced using fatty acid chloride reagents. To improve the mechanical properties. resulting in products with DS values up to 2. Natural polymers such as starch. higher molecular weight carboxylic starch esters (C4‐C6) .
1. 4.N‐dimethylaminopyridine (DMAP).2.01 when using vinyl‐laurate. and tetrahydrofuran (THF) were supplied by Acros (Belgium) and were also used as received. In DMSO. 4‐N. 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 starch was dried before use for 48 h at 105 oC under vacuum (approx. IR spectra were recorded on a Spectrum 2000 FT‐IR Spectrometer (Perkin Elmer). Materials and Methods
4. The reactions were either performed in water or in DMSO using a basic catalyst (Na2HPO4). Germany) and acetic anhydride (Merck. 73% amylopectine and 27% amylose) was purchased from Sigma (Germany).2. starch esters with a substantially higher DS value (up to 1. The effects of the starch to vinylester ratio on the reaction rates and DS have been explored. In addition. not been reported to date. as recommended by Laignel et al . the use of basic catalysts other than Na2HPO4 has been investigated. ~1 mbar). to the best of our knowledge. The spectra were recorded at 50 oC.6 for starch acetate ester) were obtained. The synthesis of starch stearate esters using vinyl ester reagents has. Analytical grade vinyl stearate (Aldrich.2. Germany) were used as received. This solvent effect is likely caused by the higher solubility of the vinyl esters in DMSO compared to water. Japan). The maximum attainable DS of starch acetate in water was below 1 and limited to 0. 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. Technical grade dimethyl sulfoxide (DMSO). the Netherlands).
4. vinyl laurate (Fluka. Materials Corn starch (approx. Germany) were used without further purification. leading to a moisture content of 2 %‐wt (measured gravimetrically). Analytical Equipment H‐ and 13C‐NMR spectra were recorded in CDCl3 on a 400 MHz Varian AMX NMR machine. The products were placed directly on the diamond plate and 50 scans with a resolution of 4 cm‐1 were recorded. sodium acetate (Merck.
. Potassium carbonate (Boom. 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.2. leading to higher reaction rates. Germany) and disodium hydrogenphosphate (Merck.
C4‐11). After cooling. 3S. 2H. The product was washed twice with methanol (50 and 25 ml.2. 3H. 68‐74 (broad. 172‐174 ppm (C=O.
FT‐IR (cm‐1): 2920 (C‐H stretching).3 (m. Starch‐laurate (Sample 17. C18). 7H. C1S‐6S)
. 3‐6 ppm (m. approximately 5 mbar) for 24 h until constant weight. broad peaks. typical spectra are given in Figure 4.Chapter 4
4. 7H. C1S). C4‐11).3 (m.5 (m. broad peaks. 1740 (C=O). 2. 1. overlap with CDCl3 (C4S). 7H.3. (13C‐NMR) and Figure 4. 1. 12). C12). 24. C3). 760.4.5 (m. C4‐C17). C2). Table 4. broad peaks. Typical example of the synthesis of laurate and stearate esters of corn starch Corn starch (0. (FT‐IR). 2H. 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. C3).3.0 (C12).9 (t. 1. 3H. 1150 (C‐O stretching).5 (m..3.5 g) was first gelatinized in DMSO (5 ml) at 70 oC for 3 h. 1110 (C‐O stretching). the product was dried in a vacuum oven (70 oC.1.1. (1H‐NMR). C2). 1.8‐2. 76‐78 ppm. broad peaks. 3‐6 ppm (m. broad peaks.7 (C11).6 (m. 2H. 3‐6 ppm (m. 28‐32 (C4‐9). respectively).52): H‐NMR (before peracetylation.0 (m. and O‐C6S). C2’). C6S).96) H‐NMR (before peracetylation. 95. Table 4. 34.
H‐NMR (after peracetylation.9 (t.9 (t. 16H.2. 2. O‐C3S. C3). 2850 (C‐H stretching). CDCl3): δ 0. 1. CDCl3): δ 0. Subsequently. 1. 1020 (C‐O stretching). resulting in the formation of a homogenous transparent solution. 1230 (C‐O stretching). 31.9 (broad. attached to O‐C2S. C1S‐6S). broad peaks. 2H. 935 (C‐O stretching). 1. Finally.3 (m. 61. 16H. 2H. Methods 4.. 28H. broad peaks. C2). CDCl3): δ 0. The samples were characterized by 1H‐ and 13C‐NMR and FT‐IR. Starch‐stearate (Sample 19. DS = 2. the product was precipitated using methanol (100 ml) and separated from the liquid phase by decantation.1 (m. 1295. CDCl3): δ 14. 2H. Figure 4.1. C1S‐6S). 720.4 (m.1. C2S.9 (C3). DS = 2. 1350 (C‐H bending).
C‐NMR (before peracetylation.2. 1410 (C‐H bending).4 (broad.1 (C11). 3H. broad peaks. 22.9 (C10).. 1370 (C‐H bending). 3H. 5S). The atom numbering scheme is given in Figure 4. 2.
2 (C2). 5S). C18). 7H. 32.2. 1150 (C‐O stretching). attached to O‐C2S.4 (m.6 (m. 1370 (C‐H bending).4 (broad. C2). 28H. 75.8‐2. Synthesis of Higher Fatty Acid Starch Esters using Vinyl Laurate and Stearate as Reactants
H‐NMR (after peracetylation. 1295. 3S. 2H. 61. C1S). 2. 3H. O‐C3S. 1020 (C‐O stretching). C3). 865. broad peaks. 95. broad peaks. 1.3 (m. CDCl3): δ 14. 34. 68‐74 (broad. CDCl3): δ 0.0 (C18).0 (C16). 172‐174 ppm (C=O.9 (t.7 (C17). 1740 (C=O). 1.7 (C4S). Numbering scheme for carbon atoms of products 4. 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. 3H. 950 (C‐O stretching). C2’). 720. C2S. 1100 (C‐O stretching).3. 22.5 (m. 1. and O‐ C6S)
FT‐IR (cm‐1): 2920 (C‐H stretching).0 (C3).2. 1410 (C‐H bending). 760.5 (broad. 2850 (C‐H stretching). 1455 (CH2). 2H.1.
O R O
1S 2S 1
CH3 2' O
O R O
O R O
Figure 4. C1S‐6S)
C‐NMR (before peracetylation. A peracetylation reaction to substitute all of the remaining
. 26‐32 (C4‐15). C6S). C4‐C17). 3‐6 ppm (m. 1350 (C‐H bending). 25.
8− 2.5.5. It was finally dried overnight in a vacuum oven at 70oC and 5 mbar until constant weight. Determination of the Degree of Substitution (DS) The DS of the product was calculated using 1H‐NMR spectra of the products after peracetylation. The procedure was described earlier by Elomaa et al  and the relevant equations are shown in equation 4.8 ppm (protons attached to the fatty acid carbons C3‐C12 for laurate and C3‐C18 for stearate) using eq. The peracetylation procedure by Einfeldt et al  was applied. The DS of the fatty acid esters was calculated from the DS of the products after peracetylation (eq 4.:
.3.5‐1. Typically.126.96.36.199‐2. 4.6 ppm).)
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 0.6 − AC 2 3
(4. The product was precipitated by the addition of methanol and washed several times with methanol.
AH − acetate = A1.7.6−5.) (4.:
AH − acetate AH − starch
A3.1 g) was suspended in THF (4%‐w/v) and stirred at 55 oC until the starch was fully dissolved (typically 3 h). 4. 4.Chapter 4
hydroxyl groups with acetate groups was applied to obtain reliable DS data.3. The peracetylation reaction was conducted for 7 h at 50 oC.6‐5. the peractylating reagents (DMAP.).).188.8.131.52 7
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. acetic anhydride and pyridine in a DMAP : acetic anhydride : pyridine : AHG molar ratio of 1: 10: 22 : 1) were added.)
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 3.2.6 ppm range) and a correction has to be made (eq 4.4. and 4.)
AH − starch =
DS fatty ester = 3 − DS acetate
(4.. the starch ester (0. Subsequently. and 4.
The reaction was performed in two discrete steps.)
OH O HO OH
R O CH2
O HO O R
(1) R = laurate (C12) (2) R = stearate (C18)
Scheme 4. the vinylester and the catalyst were added and the reaction mixture was heated to 110°C. the starch was gelatinised in DMSO at 70°C for 3 h to make the hydroxyl groups of starch more accessible for reaction. 4.1.
. After 24 h.6.8 = (stearate) 33 33
(4.. the esterified starch started to separate from the medium in the form of a gel.1). 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 4.3.7. After 2‐3 h.7.1. The products of these exploratory reactions are insoluble in water and DMSO.5−1. but swell in organic solvents such as toluene and THF. Results and Discussion
4. the brownish gel was precipitated with methanol and the product was collected after vacuum drying in the form of a transparent. light yellow solid. Synthesis of Higher Fatty Acid Starch Esters using Vinyl Laurate and Stearate as Reactants
AH − fattyester = AH − fattyester =
AC 3−C12 A0.‐4. A schematic representation of the esterification reaction of starch with the vinyl esters is provided in Scheme 4. 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.3. Initially.8 = (laurate) 21 21 AC 3−C18 A0.3. Subsequently.5−1.1. Ax‐y stands for the peak area in the range δ x‐y ppm.
2. A reaction with vinylstearate at similar conditions resulted in a stearate starch ester with a DS of 2. (c) peracetylated starch laurate. When using a vinyl‐laurate : AHG molar ratio of 3 and K2CO3 as the catalyst. (b) starch laurate. 4.52 (Sample 17.3. and 4. Medium DS starch laurate and starch stearate (1 < DS < 2) dissolve poorly in DMSO‐d6 and CDCl3.) in CDCl3 at 50oC.
. Typical 1H‐NMR spectrum of: (a) native starch in DMSO d‐6 at 60oC.1. Table 4. Product Characterisation 4. Higher DS products have a higher solubility in CDCl3 and good quality 1H‐ and 13C‐NMR spectra could be obtained (Figure 4.2. a product DS of 2.1.3. 4.52 was obtained.96.
Figure. DS = 2.2.1. Table 4.Chapter 4
The DS of the products was determined by using NMR (vide infra). even at higher temperatures (50 oC).) in CDCl3 at 50oC. For atom numbering scheme: see Figure 4.1. 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.3.).52 (Sample 17.2. DS = 2.
required for DS determinations. Typical 13C‐NMR spectra of: (a) native starch. This feature hampers the DS determination by NMR.2. Typical 13C‐NMR spectra of the products are given in Figure 4. (c) starch stearate. (b) starch laurate. 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 4.52 (Sample 17) in CDCl3 at 50oC. in DMSO d‐6 at 60oC. The starch peaks (δ 3‐5. The same
Figure 4.1. Two of the carbon resonances (1S and 4S) are considerably shifted compared to native starch. The resonances arising from the anhydroglucose unit of starch are broadened. NMR spectra of the peracetylated products are considerably improved in terms of peak resolution and allow a more reliable calculation of the DS.3 ppm.5 ppm) are broad and overlapping . For atom numbering scheme: see Figure 4.2.c.5 ppm). and therefore a peracetylation procedure to substitute all of the remaining OH groups with acetate groups was applied [8. 10‐11]. 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).3.96 (Sample 19) in CDCl3 at 50oC. Clearly visible are the peaks arising from starch and the aliphatic hydrogen atoms of the fatty acid chain (δ 0. The 1H‐NMR spectrum of a typical peracetylated starch laurate is shown in Figure 4. The proton signals of the acetate methyl group.3. DS = 2.8‐2. are together with the CH2 groups of the acid chain adjacent to the ester moiety in the range δ 1. DS = 2.8‐2.
1.4.4. Sample 19. FT‐IR measurements The FT‐IR spectra of starch laurate and starch stearate are shown in Figure 4.4. starch laurate C-H C=O
a. The near absence of remaining hydroxyl
. For comparison. In addition.4. Table 4. a spectrum of native starch (Figure 4.2. starch stearate
b. Table 4.4. and 4. 4.b. Characteristic peaks of the polysaccharide backbone are visible in the 1250‐900 cm‐1 region (C‐O stretching) .2.Chapter 4
phenomenon was observed by Dicke for starch acetate .3.). native starch -OH C-H
Figure 4.96.) is also included.1. starch stearate (DS= 2.
c. FT‐IR Spectra of starch laurate (DS= 2. Sample 17.a.c.b.) show characteristic bands of the carbonyl group of the fatty esters in the 1750‐1700 cm‐1 region. and 4.c.4. the C‐H stretching vibrations of the alkyl groups of the fatty ester chain are clearly present at 2920 and 2850 cm‐1.) and native starch FT‐IR spectra of both starch laurate and starch stearate (Figure 4.4.52. 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.
07 0.60 1.52 2.68 0.24 2.90 1.3.01 0. Most of the experiments (14) were performed using Na2HPO4 as the catalyst. type of catalyst and the effect of the addition of co‐solvents on the product DS was studied in more detail.99 1.44
Exp.1. in line with the NMR data.2.96 2.13 1. four experiments were performed with two alternative basic catalysts (K2CO3 and Na‐acetate).57 0.54 2.1.91 0. 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 1.05 0. The results are shown in Table 4.08 1. Table 4. 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. 4.90 0. Systematic Studies The effect of important process variables like the vinyl‐ester to AHG ratio. In addition.
. 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.23 0.
2 3 4 Vinyl Ester: AHG Starch Ratio [mol/mol]
Figure 4.5. this positive effect dominates the reaction rate and the DS of the products will therefore increase at higher vinyl ester intakes.1.Chapter 4
4. A further increase leads to a reduction in the DS. A further increase in the vinyl ester intake leads to a reduction in the DS. 2 wt% catalyst intake on starch).5.23 for vinyl‐laurate at an intermediate vinyl‐ester‐AHG ratio of 3. The DS values are increasing with higher vinyl‐ ester/AHG molar ratio until a certain maximum. The highest DS value was 1.). This is likely due to a reduction of the polarity of the reaction medium. This reduced polarity is expected to lead to ○ : starch‐laurate □ : starch stearate (lines for illustrative purposes only)
. At low to medium vinyl ester/AHG ratios (0‐3).5
Degree of Substitution (DS)
0. The DS of the products is a clear function of the vinyl laurate and stearate ester intakes (Figure 4.3.5. At a ratio of 1:6.
1.3. the ester intake is equal on a weight basis to the DMSO intake. Higher concentrations of vinyl esters are expected to lead to higher esterification reaction rates. This behaviour is likely the result of two opposing effects. 7‐10). DS of the product as a function of the type of vinyl ester and the vinyl ester‐starch ratio (24 h reaction time. The results are presented in Figure 4. 110 oC. 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.
This effect is particularly evident at higher vinyl ester/ AHG ratios (>3) (Figure 4. These negative effects dominate the reaction performance at higher vinyl ester/AHG ratios and lead to a reduction in the DS values. When using Na2HPO4 as the catalyst.e. The latter factors appears to have a strong effect on reaction rates (vide supra).8 for stearoyl chloride.1. It is clear that Na‐acetate and K2CO3 are considerably more active than Na2HPO4 and products with a significantly higher DS were obtained. with high carbon numbers leading to a reduction in the DS. the DS of the product is also a function of the chain length of the fatty acid. 4.3. leading to enhanced reactivity.2. For starch
. Effect of the addition of toluene as a co‐solvent A number of reactions were performed using a co‐solvent. 11‐14 in Table 4. the starch laurate esters display higher DS values than the starch stearates. The addition of toluene for the esterification of starch with vinyl‐esters surprisingly did not lead to improved DS values. Thus.5. Catalysts screening A number of alternative basic catalysts for Na2HPO4.3.3. with reductions in polarity leading to lower reaction rates. 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.3. et al  for the silylation of starch.1. The results are given in Table 4. These findings were rationalised by assuming that the co‐solvents increase the solubility of the silylated products. Here. i.). 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. Aburto. 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 . In this case.). Although toluene may positively effect the reaction by (partly) re‐dissolving starch ester precipitates. and illustrated in Figure 4. the addition of co‐solvents (toluene/ THF) led to an increase in the DS.7 for lauroyl chloride to 0.1. 4. A similar trend in reactivity pattern was observed in this study and the DS decreased from 1.3.6. 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. 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. it also results in a dilution of the reaction mixture and a reduction in the polarity.). K2CO3 and Na‐ acetate were tested. A similar procedure was proposed by Nouvel.
1.5 were obtained.6. K2CO3 gave products with a significantly higher DS (2. The products were synthesised in DMSO using vinyl‐esters in the presence of basic catalysts (Na2HPO4. function of the type of catalyst.
. K2CO3 and Na‐acetate are superior with respect to activity when compared with Na2HPO4. Comparison of DS values with different catalysts (Vinylester: AHG molar ratio = 3:1. and Na‐acetate). K2CO3. For starch stearate. The DS of the products is a clear function of the chain length of the fatty ester and the type of catalyst.96) compared to Na‐acetate (DS=2.5
Degree of Substitution
1. and FTIR and confirm the presence of chemically bound fatty acid chains.4. DMSO).
2. the statement that the DS for the laurate esters is always higher than for the stearate esters is not generally valid and a.Chapter 4
0. when using K2CO3 as the catalyst. the DS for the laurate ester is lower than the stearate ester.44). in line with the findings for NaH2PO4. 110oC.6. the two catalysts are equally effective and products with a DS of about 2.o. Conclusions
A study on the synthesis of corn starch fatty acid esters with high DS values is reported. catalyst amount = 2%‐w. Thus. the DS of the product is also tunable by proper catalyst selection. and Figure 4.). The yellow products were characterized by 1H‐ and 13C‐NMR. gray: Na2HPO4 black: K2CO3 white: Na‐acetate
4. The DS of the laurate ester when using Na‐acetate is higher than for the stearate ester (Table 4. Apparently. However.5
S. The DS of the products may also be tuned with the vinyl ester/AHG molar ratio. Koch. characterization. thermal characteristics and biodegradability of extruded starch acetate foams. 21. Merrill: Properties of Fatty‐Acid Esters of Starch. Aburto. D. Synthesis of Higher Fatty Acid Starch Esters using Vinyl Laurate and Stearate as Reactants
With these catalysts. Polym. Sci. Crops Prod. Eng. Borredon. 2003. Chem. . Sci. J. C. E. E. . 1440‐1451. Polym. 58. Prinos. Ind. Aburto. Important product properties will be described in the next chapter. Xu. J. products with a DS > 2. S.
. 65. Panayiotou: Properties of fatty‐acid esters of starch and their blends with LDPE. Appl. 1997. . A. I. Borredon. At low vinyl ester/ AHG ratio. 1999. 705‐721. Thiebaud. 54. 361‐368. Appl. Sagar. Dzenis. J. J.5. . Polym. C. 74. the DS of the product increases at higher vinyl ester intakes. Mullen and E. 34. J. W. . and biodegradability of fatty‐acid esters of amylose and starch. J. 1942. 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. References
. Panayiotou: Synthesis. J. Y. D. Thiebaud. Bikiaris. M. A. Alric. Higher ratios led to a reduction in the DS. Appl. 1208‐1217. Sci. Pacsu: Starch Studies: Preparation and Properties of Starch Triesters. Carbohydr. Alric. E. Prinos. Hanna: Water solubility.4 could be obtained for both laurate and stearate esters. presumably due to a reduction of the polarity of the reaction medium.W. and P. Polym. 1‐11. J. Bengtsson. 1995. Ind.
4. D. 1647‐1656. 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. K. Gatenholm: Surface octanoylation of high‐ amylose potato starch films. M. A maximum was observed at a vinyl ester: AHG ratio between 2 and 4. I.6. Y. Bikiaris. 2005.
Mater. T. Kussler. Gunther. Aburto. Soininen. Dubois. Six: Partial or total silylation of dextran with hexamethyldisilazane. Mouysset‐Baziard. 302‐307. G. E. 1735‐1743. P. R. . W. . Klemm: Preparative and H‐1 NMR investigation on regioselective silylation of starch dissolved in dimethyl sulfoxide. E. . Polym. Nouvel. Med. Laatikainen. R. J. M. M. D. 118‐121. Starch‐Starke 2004. L. Sci. .. I. R. Starch‐Starke 2005.Chapter 4
. I. . P. 145‐152. . Macromol. H. Al‐Higari: Acylation of starch with vinyl acetate in water. Peltonen. L. Nuzillard: Proton NMR spectroscopy assignment of D‐glucose residues in highly acetylated starch. . Mano. W. 57. 255‐263. 43. J. 2003. Koniarova. H‐1 NMR and TGA/IR. 1997. 261‐267. S. F. Reis: Thermal properties of thermoplastic starch/synthetic polymer blends with potential biomedical applicability. K.L. Asplund. J. Ydens.F. Dicke: A straight way to regioselectively functionalized polysaccharide esters. Carbohydr. Massiot. Elomaa. Laignel. Aburto. 2004. Hamaili. and E. Alric. Mormann. M. 56. 1. Borredon: Organic solvent‐free transesterification of various starches with lauric acid methyl ester and triacyl glycerides. 298. Senocq. Mater. 251‐260. I. Dellacherie. D. . J. M. 57. A. Carbohydr. Einfeldt. Starch‐Starke 1999. 11. 341‐347. Polymer 2002. Stein. C. G. J. Urtti: Determination of the degree of substitution of acetylated starch by hydrolysis. B. P. Bliard. J. Borredon: Free‐solvent Synthesis and Properties of Higher Fatty Esters of Starch – Part 2. 51. Biosci. Petzold. 2001. Degee. S. Hyvarinen. C. 14(2) 127‐135.
. Res. A. Cellulose 2004. Alric.
K2CO3. and Na‐acetate). The thermal stability of the esterified products is higher than that of native starch.45‐1. trans‐esterification.39) are totally amorphous whereas the low‐DS ones (DS= 1. and modulus of elasticity of 46‐113 MPa.26‐2. The results were adequately modeled using a non‐linear multivariable regression model (R2≥0.75) are still partially crystalline. High‐DS products (DS= 2.7‐3. 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 effect of the process variables (vinyl ester to starch ratio. vinyl laurate. elongation at break of 3‐26%. catalyst intake.96). vinyl stearate
.5 MPa. Mechanical tests show that the products have tensile strength (stress at break) between 2.Chapter 5
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. Keywords: starch esters. The basicity of the catalyst and the reaction temperature have the highest impact on the product DS. The reactions were carried out in DMSO using basic catalysts (Na2HPO4. The thermal and mechanical properties of some representative product samples were determined.
Esterification is one of the oldest methods used to improve starch properties . It is cheap and abundantly available . the mechanical properties were still not satisfactorily. 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. and particularly on the synthesis of starch acetate (C2) [11‐12]. however. However. The use of vinylesters has also been explored [20‐22]. The largest application area of plastics is the use as packaging material (37%) . Examples are agricultural mulch . The limited use is mainly because of a number of unfavourable properties of starch such as low moisture resistance. Unfortunately. The mechanical properties of products with longer ester chains (C4 and C6) products and plasticized acetate esters were considerably improved compared to native starch.65) . the physicochemical properties of the products were not mentioned. Alternative routes to avoid the use of acid‐chloride/pyridine combinations have been developed. Sagar and Merill  studied the synthesis and properties of C4‐C6 esters of high amylose starch. The development of novel biodegradable plastic materials capable of decomposing when given an appropriate environment and time is of utmost importance . et al  and Aburto. Most early studies mainly focused on the use of short‐chain carboxylic acid (C1‐C4). The use of native starch as a building block for novel biodegradable polymers.1. is rather limited to date. Thiebaud. the research activities were mainly limited to the use of vinylacetate. et al [15‐19] synthesized longer chain fatty esters (C8‐C18) of potato starch and corn amylose using fatty acid chlorides and pyridine.Chapter 5
5. Typical examples are blending with polyvinyl alcohol . packaging material. The products were biodegradable. 3‐4]. The high DS esters showed interesting properties. Two examples using higher vinyl esters (vinyl laurate) were reported by Mormann. 12] studied the synthesis and properties of of C1‐C6 esters of starch. most of the plastic waste is not biodegradable. et al . The global starch production was estimated at 60 million tons in 2004 . and food service‐ware . The products were hydrophobic and the mechanical properties were considerably improved compared to native starch. polyethylene  and polycaprolactone  and chemical modification by esterification with organic acids [11‐19]. high brittleness. Introduction
Plastics are very useful materials and are used in large amounts (245 million ton per annum) [1‐2].
. However. and incompatibility with hydrophobic polymers . Starch is an attractive feedstock for the synthesis of novel biodegradable plastics. Chemical or physical modification is required for successful applications as biodegradable polymers. 0. Mullen and Pacsu [11. and this causes serious environmental problems [1. however.
This chapter describes systematic studies using Design of Experiments (DOE) to gain insights in the effect of process variables (temperature. and Tg) of each sample were determined from the spectra related to the cooling run and the second heating one. In addition. as recommended by Laignel et al . After a first heating run from room temperature up to 200 oC to delete the thermal history of the material. the thermal and mechanical properties of representative examples of the highly hydrophobic materials are described and discussed. Analytical grade vinyl stearate (Aldrich. 73% amylopectine and 27% amylose) was purchased from Sigma (Germany). Materials Corn starch (approx. 4‐N. and acetic anhydride (Merck. DSC analyses were performed on a TA Instruments DSC 2920. 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. each sample was cooled to ‐50°C and then heated again to 200 oC (heating rate 10 oC/min).2. Germany). the Netherlands). Germany) were used without further purification. and tetrahydrofuran (THF) were supplied by Acros (Belgium).1. sodium acetate.N‐ dimethylaminopyridine (DMAP). Japan). All these chemicals were used as received. Germany). The thermal properties (Tm. The samples were heated to 700 oC in a nitrogen atmosphere with a heating rate of 10°C/min. The spectra were recorded at 50 oC. 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 . Technical grade dimethyl sulfoxide (DMSO). Long fatty esters of corn starch with a broad range in degree of substitution (DS= 0. 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. 5. The catalysts for the esterification reaction were analytical grade potassium carbonate (Boom.2. Analytical grade methanol. TGA measurements were performed using a Perkin Elmer TGA 7 Thermogravimetric Analyzer.2. pyridine. The samples (about 10 mg) were placed in sealed aluminum cells.2. Tc. Analytical Equipment H‐ and 13C‐NMR spectra were recorded in CDCl3 on a 400 MHz Varian AMX NMR machine.
5. It was dried for 48 hour at 105 oC under vacuum (~1 mbar) to reduce the moisture content below 2 % before use.24 ‐ 2. and disodium hydrogenphosphate dodecahydrate (both Merck. and toluene were obtained from Labscan (Ireland). vinyl ester to anhydroglucose ratio. Materials and Methods
5. vinyl laurate (Fluka.
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 184.108.40.206. 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. 220.127.116.11. 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.
00 5.00 5.56 1.00 5.35 0.96 0.00 5.00 5.00 3.11 2.00 3.27 1.00 5.00 5.00 4.93 0.00 3.17 2.00 4.00 3.00 2.00 5.00 2.00 5.00 2.87 1. Experimental and Modeling Studies on the Synthesis and Properties of Higher Fatty Esters of Corn Starch
Table 5.00 5.00 5.40 1.2.00 5.12 0.00 5.00 5.62 0.00 3.00 5.00 3.00 2.00 2.50 2.00 3.00 3.27 0.00 5.00 3.00 3.00 5.82 2.79
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
.50 2.54 2.00 Catalyst Temperature [%‐w] [˚C] 2.00 3.00 2.00 3.00 2.00 5.00 5.94 2.00 2.00 2.00 5.00 5.52 2.46 1.00 4.00 5.00 3.00 2.00 5.00 3.00 5.00 5.88 2.00 5.47 1.09 0.07 0.00 5.00 3.00 5.12 2.00 5.50 0.00 5.00 2.00 5.53 1.00 5.00 3.44 2.00 2.00 3.84 2.50 2.54 2.28 2.00 5.23 2.00 5.00 5.42 2.00 5.00 2.00 5.48 0.00 3.00 3.00 3.00 5.72 2.27 1.
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 2.24 2.00 3.00 3.00 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 0.44 2.70 2.32 1.67 2.00 3.00 5.41 1.52 2.50 2.78 2.00 4.00 2.00 5.59 2.75 1.
00 2.2183 0.0001 0.98 2.
stearate stearate stearate stearate stearate stearate stearate stearate stearate stearate
CH3COONa K2CO3 K2CO3 K2CO3 K2CO3 K2CO3 K2CO3 K2CO3 K2CO3 K2CO3
4.).0003 0.1082 ‐0.00 3.55 2.0066 0.00 3.1568 ‐2.)
Table 5.Chapter 5 45 46 47 48 49 50 51 52 53 54
95 80 80 110 110 80 80 110 110 95
1.00 2.0008 0.3.1.00 5.00
3.00 5.50 2.0249 ‐0.64 2.00 5.0011 0.3.The modelled values for βij for the two esters are given in Table 5.0857 ‐0.0017 0.21 2.00 3.00 3.0009 0.6939 0.1748 0.59 2.00 5.1104 ‐0.0027 ‐0. Values for the coefficients of the DS model for starch laurate and stearate
Coefficient Starch Laurate 9. Mathematical Modeling Modelling of the DS data for both vinylesters was performed using non‐linear multivariable regression based on the 4 independent variables (Table 5.0327 ‐2.0023 0.00 4.0008 ‐0.90 2.0010 ‐0.00 5.0732 ‐0. The experimental data for both starch laurate and starch stearate esterification are best described with a quadratic model including interaction terms (eq.00 5.).
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 3.1979 Starch Stearate 17.3.1947
β0 β1 β2 β3 β4 β5 β6 β7 β8 β9 β10 β11 β12 β13
.1.0006 ‐0. The type of catalyst was quantified using the basicity constant (pKb) in water [26‐27].0470 0.5650 0.00 2.00 5.0016 0.0120 0.41
mol ratio of the vinyl ester to AHG units of starch
5.24 2.3.96 2.00 5.3.64 2.
4. and confirm this statement.6118 0.967 0. the adjusted‐R2 values are very close to the R2 ones.5
a.945 0.8386 DF 13 14 27 MS 1.5
0. which represents an “internal” validation method for the model.5
0. The very low P‐ values indicate that the models are statistically significant.5
0.1. Parity plots of the DS models for starch laurate and stearate Moreover.5 DS actual
2.2569 0. starch laurate
1.9531 0.2903 P‐value 1. starch stearate
Figure 5.0356 F 35.970 0.5
0.5 DS actual
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 5.805
2.939 0.875) indicate that the model correctly predict the
.970 (starch laurate) and 0.4.629 P‐value 1. The parity plots of both models are given in Figure 5. which indicate  that all significant variables are included in the model.8167 21.4. Analysis of variance for DS models for starch esterification
Starch Laurate Model Error Total SS 16. we also performed a PRESS analysis  (see corresponding R2 values in Table 5.5
Starch Stearate Model Error Total SS 20.).026 x 10‐7 R2 values R2 R2adjusted R2press 0. respectively and it may be concluded that the models describe the experimental data well.3399 0. Also in this case the reasonable R2 values (0.967 (starch stearate). The R2 values for the models are 0.96 x 10‐8 R2 values R2 R2adjusted R2press 0.5
2 DS predicted
1. Table 5.0583 F 27. Finally.7699 DF 13 14 27 MS 1.
The product DS for Na2HPO4 was considerably lower (1. On the basis of the sequence. The pKb for the latter (6. It is assumed that the first step in the reaction sequence is activation of the starch OH groups by deprotonation by a base [21. the type of catalyst has the largest effect on the product DS for both the laurate and stearate esters (Figure 5. Of all the variables studied.Chapter 5
products DS as function of the process variables within the range of experimental variables. 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. A similar trend was observed by Dicke  for the acetylation of Hylon VII starch using vinylacetate.8) is intermediate between that of the other two catalysts. the base strengths of the catalysts in DMSO are not known.82). The resulting anion will react with the vinyl ester to from the product.
. it can be rationalised that the rate of the reaction (and thus the product DS) will be higher when using a stronger base.18) and CH3COONa (1. K2CO3 is a stronger base (pKb 3.00) than for K2CO3 (2. The highest product DS values were obtained using K2CO3.2. the performance of Na2HPO4 does not follow this trend. two plots showing the dependence of the DS on reaction temperature and catalyst amount (in %‐w).25) and this leads to higher product DS values for K2CO3.66) than CH3COONa (pKb 9. for both starch laurate and starch stearate are given in Figure 5. 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. However. 22]. Unfortunately. This was not the case for alkaline catalysts. Catalyst performance of CH3COONa was slightly less than for K2CO3 whereas the lowest DS products were produced when using Na2HPO4 as the catalyst. such as carbonate or acetate salts leading to higher DS esters. These results may be rationalised by considering the role of the catalyst in the modification reaction. This is indeed the case when comparing the performance of K2CO3 with CH3COONa.).2. whereas catalyst performance is considerable lower. Another explanation may be related to the regio‐chemistry of the reaction. To evaluate the effect of each variable on the product DS. 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.
2. although to a lesser extent.). the catalyst intake and reaction temperature also affect the product DS (Figure 5. 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. starch stearate
Figure 5. starch laurate
b. As expected and in
. Experimental and Modeling Studies on the Synthesis and Properties of Higher Fatty Esters of Corn Starch
5. and the results for all samples are summarised in Table 5. DSC analysis of starch stearate esters ( ) DS 1. .3.3. higher reaction temperatures and catalyst intakes lead to higher product DS values.75. DSstearate = 1. Product Properties The thermal and mechanical product properties for two starch‐laurate and stearate samples with medium (DSlaurate = 1. the vinyl ester to AHG ratio has the smallest effect on the product DS.) DS ( 18.104.22.168) and high DS values (DSlaurate = 2. Of all variables studied.
Tm = 21 oC
Tc = 12 oC
10 Temperature ( C)
Figure 5.39) were determined.2. Representative DSC spectra for the starch stearate are given in Figure 5.39)
.45 (. . DSstearate = 2.Chapter 5
line with studies on starch acetates .26.
45 Tm [˚C] n.26 1.5. Thermal transitions at relatively high temperature (melting. Changes in the transition temperatures of native starch by esterification have been reported in the literature [13‐14.
n.d. close to the value obtained in our study for the medium DS starch stearate (21 °C).d.d. For instance a Tm of 32 oC was reported for a potato starch stearate ester with a DS of 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. in the present work.= not detectable
The thermal behavior of the products is a strong function of the DS.a 21 Tc [˚C] n.
. 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. In any case. or glass transition above 0 oC) were even absent for the high DS products.a 27 n. especially in the amylopectin part [15. which is known to contain crystalline areas.a 22 n. 22].4.). 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. However.d. octadecane.a 12
a. Unfortunately the thermal transitions for the virgin material lie above the degradation temperature and can therefore not be accurately determined . Experimental and Modeling Studies on the Synthesis and Properties of Higher Fatty Esters of Corn Starch
Table 5. esterification of starch with vinyl laurate and stearate reduces the crystallinity of the starch considerably and leads to a lowering of the transition temperatures. To the best of our knowledge. In particular the behaviour at relatively high temperatures (>0 oC) is further discussed.39 1. crystallization. implying that these products are fully amorphous. these are the first examples of completely amorphous starch esters. TGA analysis was performed to study the thermal degradation behavior of the starch laurate and stearate samples (Figure 5.75 2. The melting point of the potato starch stearate was close to that of model compounds for the stearate side chains (methyl stearate. 40‐42 oC. At moderate DS values the products still display a melting temperature (at 20‐30 oC) and a crystallization temperature (at 10‐25 oC). 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). These values are reduced considerably compared to virgin corn starch. 17].8 .d.
Similar trends were reported for amylose octanoate‐ stearate‐laurate esters  and potato starch octanoate‐ laurate esters . The difference in thermal stability between the laurate and stearate sample is limited. .Chapter 5
Temperature ( C)
Figure 5.26) (. elongation at break and elasticity modulus) of the products were determined (Table 5.39)
The TGA (Figure 5. .) curves clearly show starch esterification with either laurate or stearate results in products with enhanced thermal stability.
.). ( ): native starch ( ): starch laurate (DS 2.
TGA analysis of various starch samples.4.6.4. The mechanical properties (stress at break.): starch stearate (DS 2. Native corn starch degrades at lower onset temperatures (250‐300 oC) than the corresponding esters (300‐350 oC).
1 Modulus [MPa] 59. independently of the DS values.05 for both laurate samples. On the other hand.9 ± 14.3 ± 2. however.2 8.7 ± 4.0 3.2 ± 0.7 ± 1.5 MPa. 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.1 Elongation at break [%] 26.7 ± 3. Both starch laurate samples display. observed for the high DS starch stearate sample. Starch stearate with a relatively low DS (1.3 21. and modulus of elasticity between 46‐113 MPa. elongation at break between 3‐26%.2 2. This product is more rigid compared to the product with medium DS value (higher tensile strength and elasticity modulus. a clear plastic behavior with a maximum in the stress‐strain curves. This indicates that their mechanical behavior is comparable to that of typical engineering polymers .5 ± 1.5. starch stearate at relatively high DS values (2. but higher elongation at break.8
Starch Ester Starch laurate Starch laurate Starch stearate Starch stearate
All materials generally show tensile strengths between 2. 17]. is roughly 0. plastic deformation and yield strain of about 0.e.1 82.6.45) displays a very similar behavior with respect to the laurate samples (i.0 3. thus indicating a relatively more rigid material with respect to all other samples.26 1.7 ± 4. calculated as the ratio between the stress at yield (σy.3 3.0 ± 2. Experimental and Modeling Studies on the Synthesis and Properties of Higher Fatty Esters of Corn Starch
Table 5.39) does not show any plastic behavior but only an elastic one. For starch laurate. A deeper understanding of the mechanical behavior can be gained by investigating the shape of the stress‐strain curves (typical examples in Figure 5.7 ± 14.
Mechanical properties of the starch esters
DS 2. An inverse behaviour is.45 Stress at break [MPa] 3.05).1 ± 15. and lower elongation at break).1 ± 1. The corresponding yield strain (roughly 0.).0 112.9 46.39 1. The values of the yield strain. determined from the stress‐strain curves as the stress at which the material ceases to be linearly elastic) and the modulus.025) lies still in the typical range of engineering polymers but also close to that typical of composite materials .75 2. an increase in product DS results in lower tensile strengths and modulus of elasticity.
. This is in agreement with previous studies regarding the mechanical properties of starch esters with different chain length of the fatty acid moiety [13.7‐3.
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 stearate (DS 2. Stress‐strain curves for starch esters.7.45) (‐ .26) (.39). Although there are many differences in the synthetics methods as well as in testing conditions and procedures. .
.5.): starch stearate (DS 1.75) ( ): starch laurate (DS 2. ( ): starch laurate (DS 1.).Chapter 5
4 stress (MPa)
0 0 5 10 15 20 strain (%) 25 30 35
for potato starch exactly the opposite is observed: relatively low stress (0. High DS (2.26) 3. In agreement with our own data.1 ±1.2 8.3 Potato Starch Stearate Ester Low DS (1.7.7) 0.39) 2. confirms that the corn starch esters synthesized in our research are relatively rigid materials as compared to the potato starch esters.5. In this respect the difference in stress and elongation at break for the high‐DS starch laurate sample is striking. 3. an unexpected behavior was also observed for the mechanical properties of the starch esters from potato starch.45) 3. n.7 ±3.
High DS (2.7 MPa) coupled with a significant (1500%) elongation. 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) .0 26.7 ±4.). DS) which mainly determines the thermal and mechanical behaviour.6 9 ±2 High DS (2. While for corn starch (DS 2.= not available
Inspection of the mechanical properties of the starch esters as shown in Table 5.3 10 ±2
Low DS (1.6 Corn Starch Stearate Ester Low DS (1.a. as confirmed by DSC.7 ±0. 9% for potato starch).7 ±0.a.5 ±1. Here.1 21.0 3.0 ±2.9 ±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. They related this inverse property of starch stearate to crystallization of C18 side chains. Experimental and Modeling Studies on the Synthesis and Properties of Higher Fatty Esters of Corn Starch
n. 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. we show that although no crystallization of C18 occurred (see DSC data of the high DS material in Table 5. the high DS material is still more rigid.2
Stress at Break [MPa] Elongation at Break [%]
a.7 MPa for potato starch ones) are coupled with significant differences in the elongation (22% for corn starch esters.8) 3.2 MPa for corn starch esters.1 ±15.4 1500 ±8. merely because of the structural property of the C18 side chain with respect to the starch backbone. 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.26) relatively high stress at break (3. On the other hand at low DS values comparable stress at break (3.a. Moreover.75) 3.5 MPa) is coupled with a low elongation (26 %). 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).7 ±1.a.1 High DS (2.2 ±0. but the interplay between those factors.7) 1.
1‐2. including statistical modelling. reaction temperature. average value of mole of substituted –OH per mole of anhydroglucose (AHG) units [‐]
.5 for starch laurate.07‐1. and Na‐acetate).
5. The starch esters were synthesized by reacting gelatinized starch with vinyl laurate or vinyl acetate in DMSO in the presence of basic catalysts (Na2HPO4. The use of K2CO3 and CH3COONa catalysts resulted in medium‐high DS products (2.5 for starch stearate).3‐1. Statistically adequate (R2 ≥0. catalyst intake. Comparison with literature data indicates that the properties may be (fine) tuned also by the starch source. and catalyst basicity) to the DS of the starch ester products were developed. on the synthesis of corn starch esters with long chain fatty acids is described. Nomenclature
: regression coefficients obtained by a multiple regression procedure [various units.4‐3 for starch stearate). 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.9 for starch laurate. Conclusions
A systematic study. The models may be used to determine the appropriate process conditions to obtain a product with a pre‐defined DS.4. 0. K2CO3. 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 DS of the products is a strong function of the basicity of the catalyst. 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. 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.5. 1. Thus. while the use of Na2HPO4 resulted in low‐medium DS products (0.Chapter 5
C. 2002. and Regulations. Denmark. Blackie Academic & Professional. E.dk/isi/stat/rawmaterial. Plastics Europe. Belgium. . Hanser Publishers. in Packaging in the Environment (Ed. R.J. and Biopolymers. Stevens: Green Plastics. Otey. Association of Plastics Manufacturers. R. Denmark. Glasgow. . D. in The Food Biopack Conference: Foodstuffs. Culter: Plastics Packaging.html. C. 19. 2008 . The Royal Veterinary and Agricultural University. 2000. 2000. 1980. Department of Dairy and Food Science. A. Marsden: The challenge of domestic waste disposal. W. demand and recovery for 2006 in Europe. Aarhus. Hernandez. Res. S. UK. E. Westhoff. An Introduction to the New Science of Biodegradable Plastics.6. G. Eng. . Selke. Dev. . New Jersey. G. F.starch. . Munich. S. M. Weber). Brussels. Ehrenstein: Polymeric Materials. 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]
5. Levy). 27‐29 August 2000 (Ed. Munich. P. USA. Properties. accessed on August 01. Germany. Copenhagen.
. Processing. The compelling facts about plastics: An analysis of plastics production.W. 592‐595. Denmark: http://www. Prod. Structure – Properties ‐ Applications. References
. M. M. Applications. Doane: Starch‐based blown films. Ind. Packaging. Germany. H. Bastioli: Global Status of the Production of Biobased Packaging Materials. . Princeton University Press. Hanser Publishers. 2000. Conference Proceedings. 2008. Website of International Starch Institute. J. Frederiksberg C. J. Chem. 1986.
Pacsu: Starch studies: preparation and properties of starch triesters. Polym. Prinos. E. I. M. Sci. Al‐Higari: Acylation of starch with vinyl acetate in water. Pacsu: Starch studies: possible industrial utilization of starch esters. 74. de Graaf. Hamaili. Rev. A. Thiebaud. A. Weber). Carbohydr. W. 381‐384. Alric. J. . J. Thiebaud. . D. J. report of Food Biopack Project (Ed. A. E.. C. Ind. J. 65. Aburto. W. Panayiotou: Synthesis. K. W. . Borredon. Merrill: Properties of fatty‐acid esters of starch. Bikiaris. J. . Starch‐ Starke 2005. 1997. 1942. Polym. J. Eng. Part C: Polym. Aburto. J. Mullen. X. 57. 118‐121. C. J. 28. 2000. Bikiaris. Alric. S. C.D. Bikiaris. . Beenackers: The kinetics of the acetylation of gelatinised potato starch. Wang: Properties of starch blends with biodegradable polymers. characterization. de Graaf: Ph. 137‐144. . I. Mouysset‐Baziard. Aburto. Wang. 1943. E. Polym. . Carbohydr. M. S. R. Panayiotou: Properties of fatty‐acid esters of starch and their blends with LDPE. L. Chem. . A. 1997. 51. Thiebaud. F. Alric. Macromol. E. Sci. A. S. Sagar. Starch‐Starke 1999. G. Starch‐Starke 2004. C. Frederiksberg C. and biodegradability of fatty‐acid esters of amylose and starch. Borredon: Organic solvent‐free transesterification of various starches with lauric acid methyl ester and triacyl glycerides. 132‐135. 2003. L. 35. Janssen. M. Sci. Borredon. J. Thesis. 1440‐1451. E. I. Senocq. G.Chapter 5
. K. 1995. I. Z. J. Eng. A. 1999. H. Yang. Denmark. Appl. J. Alric. 51. D. D. R. I. J. Borredon: Preparation of long‐chain esters of starch using fatty acid chlorides in the absence of an organic solvent. Panayiotou: Properties of octanoated starch and its blends with polyethylene. 145‐152. E. Chem. C. Aburto. . 302‐307. P. . 34. Rijksuniversiteit Groningen. Biobased packaging materials for the food industry. Y. 705‐721. 58. 101‐112. J. 34. Borredon: Free‐solvent synthesis and properties of higher fatty esters of starch ‐ Part 2. D. Polym. . Alric. 385– 409. Broekroelofs. 1647‐1656. the Royal Veterinary and Agricultural University. Appl. C43. J. J. Appl. E. E. Mullen. B. Ind. Department of Dairy and Food Science. . E. status and perspectives. Sci. Aburto. 1209‐1217. Prinos. Starch‐Starke 1999. 56. Polym. . Prinos. Alric. I. Mormann. Aburto. 1996
.. J. Borredon. W. 1995.
Res. Ashby. . A. Junistia. K. G. Shercliff. 2007. D. Cellulose 2004. J. John Wiley & Sons Inc. Prentice‐Hall. Janssen. Montgomery: Design and Analysis of Experiments 5th edition. Cebon. M. R.. Heeres: Synthesis of higher fatty acid starch esters using vinyl laurate and stearate as reactants. 2008. Laignel. Materials: Engineering. Picchioni. Bliard. . M. 11. A. . A straight way to regioselectively functionalized polysaccharide esters. New Jersey. L. Nuzillard: Proton NMR spectroscopy assignment of D‐glucose residues in highly acetylated starch. H. C. M. Dicke. New York. Manurung. Amsterdam. R. B. . B. 298. 2001. 255‐263. A. the Netherlands. 1986. L. Science. A. Carbohydr. Sugih. 1997. D. J. Underwood: Quantitative Analysis 5th edition. L. R. Butterworth‐Heinemann. Day. accepted for publication in Starch‐Starke. C. 251‐260. . USA. Massiot. F. P.
. H. USA. Processing and Design. Experimental and Modeling Studies on the Synthesis and Properties of Higher Fatty Esters of Corn Starch
Highest values of the FD (up to 45 %) were observed for 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. at relatively high GMA and BPO intakes. the mechanical behavior of the compatibilized blends is in general different from that of pure PCL and of the corresponding uncompatibilized blends.Chapter 6
Synthesis and Properties of Reactive Interfacial Agents for Polycaprolactone‐Starch Blends
The synthesis of two reactive interfacial agents for starch‐PCL blends. The reactive interfacial agents were tested for their performance in PCL‐starch blends. 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. compatibilizer. 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. At relatively low starch intakes. grafting
. The FD values for PCL‐g‐DEM were considerably lower (up to 7 %). As a result. Keywords: starch‐PCL blend. diethyl maleate. In particular the elastic modulus for the compatibilized blends is significantly higher than that of uncompatibilized ones. PCL‐g‐ glycidyl methacrylate (PCL‐GMA) and PCL‐g‐diethyl maleate (PCL‐g‐DEM) is decribed. Both act as compatibilizers for PCL/starch blends by improving the interfacial adhesion between the starch particles and the PCL matrix. PCL‐g‐DEM has at least a comparable performance with respect to PCL‐g‐GMA. despite the expected differences (favorable to PCL‐g‐GMA) in the in situ formation of the compatibilizers.
to an ideal compatibilizer having both PCL and starch blocks. 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. it would be actually more accurate to define the block copolymer as an “interfacial agent”.1. along the backbone. Aerobic soil‐burial experiments showed that the mechanical properties of PCL films decrease rapidly in time . To reduce the tendency for phase separation. synthetic biodegradable polymers have been applied. blends of hydrophilic starch and hydrophobic biodegradable polyesters exhibit phase separation  due to differences in polarity of the building blocks. of polar groups (usually epoxides or anhydrides [13‐18]) able to react with the – OH groups along the starch backbone.Chapter 6
6. upon mixing) by using a functionalized PCL.e. One of the well known modification strategies is blending the starch with polymers displaying a stronger hydrophobic character and better mechanical properties.
. polycaprolactone (PCL). The latter displays the presence. However. When using ungelatinized starch as a component in the blend. Unfortunately. PCL has gained considerable interest for possible applications in the fields of packaging materials and medical applications [10‐11]. This phenomenon is highly undesirable and limits the application range considerably . an interfacial agent) may be used to improve the interfacial association between the two polymer phases.e. such as polyethylene (PE) or polystyrene (PS) [3‐6]. Blending of starch and synthetic biodegradable polyesters has been widely applied for the synthesis of fully‐biodegradable products. as in this study. for starch/PCL blends. 2]. Unfortunately. these synthetic polymers are poorly or non‐biodegradable. Such structure is rather difficult to achieve by simple copolymerization methods and it is usually prepared in situ (i. The ester bonds are susceptible to attack by water and this leads to enhanced biodegradability. As a consequence. This leads. is known to be degraded with ease by microorganisms widely distributed in nature . a compatibilizer (i. To overcome this issue. Starch modifications to improve the product properties like enhanced hydrophobicity and mechanical properties were already reported in the early 19th century [1. Among these. polyesters are considered very promising alternatives . 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 ). Introduction
Starch is a cheap and abundantly available natural polymer with very good application perspectives in the area of biodegradable plastics. A well‐known biodegradable polyester. 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.
2. The synthesis and application of PCL‐g‐DEM is. 16]. Mw=50000) from Solvay Caprolactone. systematic studies and subsequent quantification of the functionalization reaction has not been reported to date. Intakes for each experiment are given in Table 1 and 2. The various compatibilizers have been tested for their performance in PCL‐starch blends.
. Mw=3000 from Solvay Caprolactones. an absolute novelty of the present research. PCL‐g‐glycidyl methacrylate (PCL‐g‐ GMA) and PCL‐g‐diethyl maleate (PCL‐g‐DEM). This chapter describes a systematic study on the synthesis of two reactive interfacial agents for starch‐PCL systems. Belgium. Materials Polycaprolactone (PCL) CAPA 2304. Germany and methanol (99.2. This low molecular weight PCL grade was used without further purification. after which the equipment was stopped and the chamber was opened to collect the samples. Compatibilizer synthesis The compatibilizers were prepared in a Brabender Plasticorder PL2000 batch‐ kneader (chamber volume 35 cm3). The starch was dried for at least 24 h at 110oC under vacuum (approx. diethyl maleate ≥97% purity (Fluka). a solution of BPO in GMA or DEM was added drop by drop over a period of 5 minutes. Corn starch (with approx. The intake of reagents was maximally 75‐80% of the volumetric volume to ensure proper mixing.1. 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. Glycidyl methacrylate 97% purity (Aldrich). Methods 6.8%) from Merck. Materials and Methods
6. After the PCL was melted (1‐2 minutes).2. UK was used for the preparation of the interfacial agents. Exploratory studies on the synthesis of PCL‐g‐GMA and its applications for starch/PCL blends have been published [13. The materials were mixed for another 5 minutes.1.8%) from Labscan. xylene (99. >99%) was obtained from Acros. ~1 mbar) prior to use. Tetrahydrofuran (THF.2.2.2.
6. and benzoyl peroxide 75% (Merck. The kneader was heated to 130 oC and PCL (CAPA 2304) was added while maintaining a rotational speed of 80 rpm. Germany) were used as received. UK. to the best of our knowledge. however. 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. Ireland. 73% amylopectin and 27% amylase) was obtained from Sigma and high molecular weight PCL (CAPA 6503. 6.
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). 22.214.171.124. The error on the transition temperature is assumed to be ± 1 oC and 5 % of the calculated values for the corresponding enthalpies. The resulting suspension was filtered and precipitated using methanol (450 mL) at 6‐8°C. 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 .3. 5 mbar). 6.2. Tensile tests were performed using an Instron 4301
. Methanol (450 mL) was added to the filtrate. and the product was precipitated at 6‐8 oC. Before analysis. Digital Scanning Calorimetry (DSC) measurements were performed using a Q1000 TA Instruments equipped with a TA Instruments DSC cooling system. created using a Cressington 208 sputter coater. Subsequently. 6. The content was blended for 15 minutes.4. The solvent was decanted and the solid product was dried in a vacuum oven (40 oC.5 h. An operation temperature of 170°C and a rotation speed of 80 rpm were applied .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.2. PCL‐g‐DEM (5 g) was dissolved in xylene and refluxed at 150 oC for 2 h. and then filtered. Infrared spectra were collected with a FT‐IR apparatus in the ATR mode using a Spectrum 2000 instrument from Perkin Elmer. 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).2. the chamber was opened and the resulting material was collected. PCL‐g‐GMA (5 g) was dissolved in 50 ml THF. The solvent was decanted and the solid product was dried in a vacuum oven (40 oC.Chapter 6
6. Analytical Methods H‐NMR measurements were performed using a 400 MHz Varian AMX Oxford NMR apparatus with CDCl3 (99. Scanning Electron Microscopy (SEM) was performed using a Jeol 6320 F Scanning Electron Microscope. stirred for 1. the samples were covered with a paladium/platinum conductive layer of 3 µm thickness. Aldrich) as the solvent.2.2.8%. PCL was added to the chamber followed by the addition of the starch and the reactive compatibilizer. 5 mbar). Work‐up of PCL‐g‐GMA products  To remove unreacted GMA monomer and GMA homopolymer.
For every T‐bone. This procedure is described in detail in the literature  and consists of calculating the sum of squares (SS) for the model and the error. strain at break (ε).2 ppm) side chains with that of a characteristic proton resonance of the PCL backbone (‐CH2‐ protons at δ 4. A 5% relative error in the peak area of the NMR spectra was assumed. leading to a 10% relative error in the FD values. The T‐bone samples were prepared using a Fontijne Holland TH 400 hot‐ press. Table 3).1. When the relative degrees of freedom (DF) are known. The latter is a measure of the statistical significance of the model.3.1. As a result we were able to obtain a model for the FD of the reaction. 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. 6.2 ppm) or DEM (‐CH2‐ protons at δ 4. 8 different T‐bones were used. For a given sample/blend. 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). 19].
. 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.0 ppm [13.2. stress at break (σ) and modulus (E) were measured. 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 FD was calculated using 1H‐NMR by comparing the area of protons belonging to the GMA (‐CH< proton at δ 3.2. On the basis of the latter values.4. The validity of the model was determined by performing an analysis of variance (ANOVA. Synthesis and Properties of Reactive Interfacial Agents for Polycaprolactone‐Starch Blends
machine. 6. the F‐ value for the model is determined followed by the P‐value. 15‐16.
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.
COOEt COOEt *
GMA BPO *
Scheme 126.96.36.199.1. Results and Discussions
6.1. 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.
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
This trend is independent of the BPO amount and matches with data reported by other groups [13.2. subsequent reactions of MA to an already grafted MA molecule hardly occur. It is most probably related to the fact that GMA is able to propagate to longer grafted chains when reacted with PCL (Figure 6.6% mol/mol-CL units DEM.6% mol/mol-CL units 15 GMA.1. Thus an increase in the GMA intake will provide more monomer available for the growing of the grafted chains leading to higher FD values.1% mol/mol-CL units 10
Substrate Intake (mol% on CL units)
Figure 6.Chapter 6
chemical structure of DEM.3. The results are graphically provided in Figure 6. Our experimental findings. showed that MA reacts easily with a radical on the PCL backbone. higher FD values for GMA than for DEM are in line with this explanation and support the proposed molecular reaction mechanisms. 130°C) Higher GMA intakes lead to higher FD values.
GMA. i. BPO=0.1. BPO=0. Effect of GMA and DEM to CL‐units ratio (mol/mol) on the FD of the products (constant PCL intake. 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.). 6.
. BPO=1. the length of a MA graft is always unity whereas longer grafts are possible for GMA.3. 22].3% mol/mol-CL units DEM. Hence.3. However. BPO=0.e.
for which a levelling off and eventually a decrease of the FD values for relatively higher MA amounts has been observed. at relatively lower GMA intakes.1
0. constant substrate to PCL ratio) For high GMA to CL ratio (24%‐mol/ mol CL units).4. DEM=30% mol/mol-CL units
0. These results are in line with earlier work [15. Effect of the BPO intake on the product FD (130°C. Such discrepancy is probably related to the relatively low intake of DEM as well as to the different reactivity of DEM compared to MA .2.2
BPO Intake (mol% on CL units)
Figure 6. although a slight increase in the FD values might be appreciated.7
0.4. doubling the amount of initiator results in considerable higher product FD. Apparently.).9
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. This will lead to a higher proportion of PCL radicals by hydrogen abstraction from the polymer backbone and thus to higher FD values. 17].6
0. GMA=12% mol/mol-CL units DEM. The FD values are within a rather narrow range. no detectable influence of the BPO amount on the FD is observed (Figure 6.2 GMA.4.1.
50 45 40 35 30
25 20 15 10 5 0 0. Such behavior is slightly in contrast to what observed for maleic anhydride [13. The use of higher initiator concentrations will result in an increase in the number of formed radicals. 17].3. 15. However. Synthesis and Properties of Reactive Interfacial Agents for Polycaprolactone‐Starch Blends
The experimental trend for DEM is different. GMA=24% mol/mol-CL units GMA. 6.
If many macroradicals are created in the system (relatively high BPO intakes) at low GMA intake. This phenomenon is likely related to the different nature of the radical grafting mechanism of GMA and DEM on PCL as mentioned before. the BPO and monomer intakes are considered as independent parameters. 6.Chapter 6
is an optimum ratio between the BPO concentration and available monomer (GMA in this case) on the FD.3. In addition. The model provided in eq. 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).2. as well as several transfer reactions might be responsible for the observed trend . defined as the difference in solubility parameters between PCl and the substrates.)
. 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.0325nmδ + 0. 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). δ )
(6. ni .1. This leads to the following equation:
FD = f (nm .3. 17].3. namely an increase in the BPO intake results in higher product FD.6022nm ni − 0. 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. gives the best description of the experimental data:
FD = −1. Here. To quantify synergic effects of monomer and initiator molar intakes on the FD of GMA and DEM on the PCL backbone.e. 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. recombination of (macro) radicals. a statistical model has been developed by performing a multivariable linear regression on the data reported in Table 6. ni the molar amount of initiator in the feed and δ the difference in solubility parameters calculated using group contributions .3.)
where nm is the molar amount of monomer in the feed. the possibility of side reactions will become significant.2. the mutual solubility of DEM in PCL and DMA in PCL was included in the model by using a parameter δ.2729nm niδ
(6. However.1. i.8875 + 0. The effect is however much less pronounced than for GMA. and 6. In particular the occurrence of “cage effects”.5431niδ + 1. 6.
BPO Intake (mol% on CL units) 0.957) and its closeness to the adjusted R2 (0.4 0. Graphical representation of the FD model for PCL‐g‐GMA..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 ). a.3.
SS 1828 81.6
60 FD (%) 40
20 0 1 0.4 BPO Intake (mol% on CL units) 0.8 0.3.3
0.6.17 F 44.358 DF 4 8 12 MS 456.
Model Error Total
Analysis of variance for the FD model provided in eq 6.5. Graphical representation of the FD models for PCL‐g‐GMA and PCL‐g‐DEM are given in Figure 6. Table 6.2 10
25 30 35 40 GMA Intakes (mol% on CL units)
a. 3D plot. b.
1 0.3. the interaction between (PCL‐monomer) intake and solubility parameter difference ( nmδ ) and the interaction between the latter factor and the initiator intake ( ni δ ).2 10 20 GMA Intake (mol% on CL units) 40 30 50
0. Contour plot.5 0.358 1909. respectively. This was also confirmed by inspection of the residue distribution by a normal probability plot (not reported here for brevity) .8 0.
b.7 0. and Figure 6.
. The R2 value for the model (0.941) also suggests that all important variables have been included in the model.6 0.908 10.5.
Figure 6. Synthesis and Properties of Reactive Interfacial Agents for Polycaprolactone‐Starch Blends
The analysis of variance data are given in Table 6. The very low P‐value implies that the model is statistically significant.
6 0. Furthermore. Similar observations have also been made by Kim. a higher BPO intake leads to a reduction of the FD values while an opposite trend is observed at lower DEM intakes. 6. 3D plot.3
1 0. 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.2 10
25 30 35 40 DEM Intake (mol% on CL units)
a. Remarkable is once more the different trends for the two substrates.6
2 0 1 0. The random behavior of the Tc as function of the FD values is probably the result of two concurring effects: the presence of irregularities. for DEM a clear transition is observed. a.) of the compatibilizers were determined by DSC.8 0.4 BPO Intake (mol% on CL units) 0. although both properties are significantly lower than those of pure PCL. Graphical representation of the FD model for PCL‐g‐DEM. et al  working with PCL‐g‐GMA. Such changes in the thermal properties compared to pure PCL may be caused by the introduction of grafts on the PCL chains. For all samples.4 0.4.
Figure 6.5 0. From a more practical point of view. which is expected to lead to a Tc reduction. 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.6. For relatively high DEM intakes (> 30 % mol/mol).7
6 FD (%) 4
0.1.8 0. the melting temperature and the relative enthalpy decrease with respect to pure PCL. Thermal properties of the compatibilizers The thermal properties (Table 6.
b. the crystallization temperature and enthalpy are not a clear function of the FD values.9
BPO Intake (mol% on CL units) 0. radical initiator and PCL intakes). and 6.1. While for GMA an increase in BPO or monomer intakes invariably leds to higher FD values. This induces irregularities and is expected to result in a lowering of the Tc and the Tm.2 10 20 DEM Intake (mol% on CL units) 40 30 50
0.3. Contour plot.2. and a favored nucleation
7 10.5 357.2.3 10.8 372.1 11.6 11.6 168.5 425.7 %.1 13.6 5.0 332.3.3
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. 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).0 341. 6.9 368.0 430.5 7.6 386. PCL‐g‐GMA has an FD of 9.2 321.
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. 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.8 384.6 305. A total of 12 blends were prepared: a series of binary ones (not containing any compatibilizers) constituting our reference points. 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. An overview of all prepared blends together with their thermal and mechanical (tensile tests) properties is given in Table 6.4 431.9 424.6%.2 342.
.5 489.9 379.4. 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. Table 6.8 371.9
E (MPa) 270.8 329.9 11.5
ε (%) 640.4.0 9.0 401.4 230.1 10.2 343.3 15.3 380.2 9. 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) 16.2 337.
8 MPa) clearly indicates that the rigidity of the blends increases at higher starch contents.2.
75 70 65 60 55 50 45 40 35
∆ H (J/ g-PCL)
10 15 20 Starch content (%)
Figure 6. The simultaneous increase in the modulus (from 270.8 % at 30 % starch content.7.2 MPa up to 341.3.43 MPa and 640. respectively from 16.). 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.7. The thermal and mechanical properties indicate no or very little interaction of the starch particles with the PCL ○ : ∆Hc □ : ∆Hm
. Melting and crystallization enthalpy as function of the starch content for binary blends with PCL.1 MPa and 341. 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. 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.Chapter 6
6. This is most probably due to a nucleation effect of the starch on PCL as observed also for other kinds of polymer blends .1.5 % for pure PCL to 7. 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 .
Morphology of starch/PCL binary blends.e. Synthesis and Properties of Reactive Interfacial Agents for Polycaprolactone‐Starch Blends
matrix.4. Evaporation of the water during blend preparation.
b. The presence of these voids was also observed for sago starch/PCL blends [25.8.2. 26] and explained by assuming that the voids are formed by water in the starch. However. the starch particles are simply inserted into voids of the PCL matrix. for which a 6 oC drop is observed when using PCL‐g‐DEM.9.
. The latter trend is also valid for the corresponding enthalpies: i. the corresponding binary blend (S/PCL 20/80) contitutes a better reference point compared to pure PCL. The thermal behavior is characterized (Table 6.
a. either by heating or in combination with mechanical stress .3. This is confirmed by examination of the blend morphology by SEM (Figure 6. In this respect the thermal properties remain substantially unchanged with the exception of the Tc. 6. This is visually confirmed by the morphology of the prepared blends (Figure 6. a decrease with respect to pure PCL is observed and then a substantial invariance as function of the PCL‐g‐DEM content. Such behavior is consistent with the hypothesis that the starch particles hinder the melting and crystallization processes of the PCL chains.
Figure 6. Thus.). and the lack of interface adhesion cause void formation.2. in order to fully understand the role of PCL‐g‐DEM.) 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. whose synthesis is described for the first time in this work. As consequence one might expect a slight different structure of these ternary blends as compared to the corresponding binary one. (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. was used as compatibilizer for the preparation of ternary blends with starch and PCL. Ternary blends compatibilized with PCL‐g‐DEM PCL‐g‐DEM.8.).
a. This can be explained by the improved interfacial adhesion  between PCL and starch.9. Indeed starch particles in the ternary blends display a smoother interface with the PCL matrix compared to the corresponding binary blends (Figure 6.
c.10.). resulting in more rigid material with higher modulus.).
Figure 6. 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.).
b. 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.9. As a result of the different morphology also the mechanical behavior display differences with the virgin PCL and the binary blend. (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
.10. 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. which will hinder the flowability and fibre forming ability of PCL matrix under cold drawing. 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.
the crystallization temperature experiences a drop of about 5‐6 oC. in all cases (i. 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.3.e. 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). In particular also in the case of PCL‐g‐GMA. at all PCL‐g‐GMA contents) there is a clear increase of the modulus with respect to the
. Modulus of ternary blends S/PCL/PCL‐g‐DEM as function of PCL‐g‐ DEM amount. However. the modulus is higher and hardly a function of the compatibilizer amount. 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
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. 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). by taking as reference the corresponding binary blend (S/PCL 20/80).3. the stress and strain at break are lowered and hardly a function of the PCL‐g‐GMA amount.2. 6.10.
binary blend. 29]. and morphologies in Figures 6.12. These differences in compatibilizing effect are not yet fully understood.11. (left) the structure of PCL‐g‐GMA is inhomogeneous at the molecular level with long poly(GMA) branches pending from the PCL backbone. presence of voids. 28.
Figure 6.7 % for PCL‐g‐DEM.) than PCL‐g‐GMA for blends containing 20 %‐wt of starch. As given schematically in Figure 6.4. and 6.8 % for PCL‐g‐GMA against 1.11.). partial interfacial adhesion) which render these ternary blends a kind of “intermediate” case between the binary ones and those compatibilized with PCL‐g‐DEM.
c. Thus. Morphology of ternary blends S/PCL/PCL‐g‐GMA. PCL‐g‐DEM is at least as active as compatibilizer (compare modulus values in Table 6. As a result.11.
. This confines all reactive GMA groups in relatively “isolated” spots along the PCL backbone. (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. 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. 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. 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. This is surprising if one takes into account the different FD values: 9. 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.
when comparing ternary blends (compatibilized with PCL‐g‐ GMA) with the binary ones.8 MPa for S/PCL 30/70 up to 430. thus probably ensuring a better coverage of the surface.12. Moreover.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 %) . 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.12. As consequence.) PCL‐g‐DEM has much less reactive groups (lower FD) than PCL‐g‐GMA.4. Synthesis and Properties of Reactive Interfacial Agents for Polycaprolactone‐Starch Blends
PCL-g-DEM PCL-g-GMA OH O OH O
Figure 6. the distribution of the reactive group along the PCL backbone is more “homogenous”. On the other hand (right of Figure 6. This is in agreement with the hypothesis made above (Figure 6. 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. but since DEM is preferentially grafted as monomer. In order to check this we compared blends with the same amount of PCL‐g‐GMA (2 %‐wt) but with different starch intakes (10. at higher starch contents more –OH groups would be available for reaction with the GMA groups. The latter result (from a modulus of 341.) and relating the “coverage” of the starch particle surface upon reaction with PCL‐g‐GMA with the compatibilization efficiency. We can therefore conclude that the efficiency of PCL‐g‐GMA in the
. other groups will presumably remain available (arrows in Figure 6.12. 20 and 30 %‐wt respectively). for all starch amounts (Table 6. Schematic representation of the in situ reaction between functionalized PCL and starch. 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.4. once PCL‐g‐DEM has reacted with one –OH groups on the starch particle.).12. Indeed. thus attenuating the effect discussed above. Concerning the thermal behavior (Table 6.) for further reaction.).
The latter hypothesis is indirectly confirmed by the observation that PCL‐g‐GMA becomes more efficient at relatively higher starch contents in the blends.
6.957. The PCL‐g‐GMA and PCL‐g‐DEM compatibilizers display lower melting temperatures and melting enthalpies compared to virgin PCL. 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. has been performed on the synthesis of two reactive compatibilisers. The reactive compatibilizers were used in blends of starch with PCL.
6. firstly reported in this work. The proposed model to quantify the effects of process variables (monomer and initiator intake. The most important variable for the final product FD is the interaction between the amount of monomer and initiator used. mutual solubility of the monomer into the molten polymer) on the FD adequately described the experimental results (R2=0. 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. P‐value ≤10‐9). 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]
.4. 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.5. Conclusions
A systematic study. including statistical modeling. PCL‐g‐glycidyl methacrylate (PCL‐g‐ GMA) and PCL‐g‐diethyl maleate (PCL‐g‐DEM).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. Nomenclature
E FD : initial modulus [MPa] : functionalization degree.
Eur. Ranucci. USA.W. 33. Liu. 1998. 2002. 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. Albertsson.G. . M.Chem. O. S. CRC Press. 1209‐1217.. 157. Bhattacharya: Properties of injection moulded starch/ synthetic polymer blends – IV. W. Pacsu: Starch Studies: Preparation and properties of starch triesters. Part 2.6. S.D. Adv. E. Ramsay. Effect of amylopectin to amylase ratio in starch. . 34. 1942. J. Karlsson. E. A. Polym. Kawasaki: Biodegradable films made from low density polyethylene (LDPE). Lindblad. A‐C. . 1477‐1487. Eur.S. M. Ogawa. Mani. Polym. Albertsson: Biodegradable Polymers and Environmental Interaction. Eng. Polym. Mani. Polym. Psomiadou. 34. 139‐161. J. . Carbohydr. 1986. R. 1467‐1475.A. Ioannis. N. Y. R. F. B.B. 227‐242. Jarowenko: Acetylated starch and miscellanous organic esters. Sci. M.J. Davis: High performance LDPE/ thermoplastic starch blend: a sustainable alternative to pure polyethylene. in Modified Starches: Properties and Uses (Ed. Thermal and morphological properties. Inc. H. Bhattacharya: Properties of injection moulded starch/ synthetic polymer blends – III. Rodriguez‐Gonzales. Polymer 2003. E. Mullen. 1997.
. Biliaderis. 1517‐1526. . A‐C. . J. 44. wheat starch and soluble starch for food packaging applications. Ind. Karlsson: Polymers from Renewable Resources. Boca Raton. Sci. 34. References
. Polym. C. 1251‐1253. 1998. . Wurzburg). 38. B.Eng. 1998.
Mater. C. M. John Wiley & Sons Inc.Chapter 6
. J. . Sci.S. Macromolecules 1996. Montgomery: Design and Analysis of Experiments 5th edition. Park: Modification of Aliphatic Polyesters and Their Reactive Blends with Starch.M. E.H. Part A: Polym. Tang: Functionalization of polyesters with maleic anhydride by reactive extrusion. 41. 1999. R. M. J. Wu: Physical properties and biodegradability of maleated‐ polycaprolactone/ starch composite. 1995. John. 5135‐ 5142.H. 1998. 5590‐5598. Environ. C.. R. Polym. . Raimo. Park: Reactive blends of gelatinized starch and polycaprolactone‐g‐glycidyl methacrylate. Jung.E. K. P. T. 81. Z. Martuscelli. J. Marechal. 4157‐4167.K. 2003. Technomic Publication. Kim. C. Cho. D. 29. Ruggeri. 1997. Fringant: Properties of thermoplastic blends: starch‐polycaprolactone. 127–134. L. 42. J. Khandpur.K. M. Polymer 2001. . Chem. Kim. J. Nakayama. Cho. 3. Mani. 8. P.S.C. Sci. Coleman. 1991. Avella. 41.W. Tang.K.. Sci. 1693‐1702. G. M. A.F. 1996. Chiellini. 1997.Y. . . . M. 35. E.M. K. Kim. R. Bhattacharya: Synthesis and characterization of anhydride‐functional polycaprolactone. Polym. 2001. Degrad. E. 127. New York. 80.K. Prog. J. R. K.Y. 2004. Inoue: Compatibilizers for melt blending: Premade block copolymers. Degrad. 3875‐3881. It. Guegan. L. Chem. Polymer 2000. Kim. Bhattacharya. Rustgi: Biodegradable polymers. R. Aglietto: Chemical reactions affecting the free‐radical grafting of diethyl maleate onto ethylene polymers. Solaro: Biodegradable polymeric materials. J. C. . Errico. USA. P. Polym. Marrucci. 1273‐1335. Environ. Stab.
. 23. Goldberg: A review of the biodegradability and utility of poly(caprolactone). . . Laurienzo. . D. C. J. Polym. Polymer 2000. Yang. M. C. .. P. 179‐187. Passaglia. USA. Adv. Macosko. . Moro.H. J. P. Rimedio: Preparation and characterization of compatibilised polycaprolactone/ starch composites. Part A: Polym. M. Polym. J. Park: Grafting of glycidyl methacrylate onto polycaprolactone: preparation and characterization. Chandra. Gazz. Graf. Appl. . Chim. 12. Averous. 91‐95. 305‐313. Painter: Specific Interactions and the Miscibility of Polymer Blends. Pennsylvania. 1139‐1148. 2001. C. Polym. 61‐67. Dole. Sci. M. J. A. 37. J. Polym. 1507‐ 1516. . J.
.S. Ruggeri. 9. E. 2003. Polym. Ishak: Mechanical properties and enzymic degradation of thermoplastic and granular sago starch filled poly(ε‐ caprolactone). B. B. Manan. Adv. D. Ishiaku. 11‐15.A. U.A. Eng. R. Sci. Kammer: On mechanical properties of sago starch/ poly(ε‐caprolactone) composites. . Passaglia. Synthesis and Properties of Reactive Interfacial Agents for Polycaprolactone‐Starch Blends
. Sci.M. Technol. 38. 273‐281. J. . 21. 293‐304. U.S. S. Ishiaku. Picchioni: Formation and compatibilizing effect of the grafted copolymer in the reactive blending of 2‐ diethylsuccinate containing polyolefins with poly‐ε‐caprolactam (nylon‐6). Eur. 40. Manan.N. . D. O. Lee. 2002. 877‐884. J. Odusanya. Ishiaku. Pang.D.W. Test. S. Tedesco.H. . M. Mauler: Comparative study of PP‐MA and PP‐GMA as compatibilizing agent on polypropylene/nylon 6 blends. Azemi: Effect of starch predrying on the mechanical properties of starch/ poly(ε‐caprolactone) composites.S.W. 393‐401. F. M.M. U.B. Polym. Odusanya. 57.M. Nachtigall. Polym.V. O. W. Bull.M. Aglietto.S.Safa. 2006. R.M. Barbosa.S. . H. 2002. 87. Polym. A. 1998.N. G.Nasirtabrizi: Ring opening reactions of glycidyl methacrylate copolymers to introduce bulky organosilicon side chain substituents. 1298‐1305. K. M. Polym. K.S. 2000. Azemi. Polym. Appl.
Starch is relatively cheap and available from a broad range of plants. In combination with the current high prices for petrochemical products. There are two types of polymers present in starch: amylose and amylopectin. blending with other materials. Its production has increased significantly since 1950. 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. 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. Starch is a polymer consisting of anhydroglucose (AHG) units. tensile and flexural strength). Starch is also too sensitive to water. For instance.o oil and gas) are very attractive materials for a broad range of applications. 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. with a rate of almost 10% every year. plastics may end up in the environment and cause serious environmental problems. The content of amylose and amylopectine in starch varies and largely depends on the starch source. There are several starch modification methods available. 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. To gain insight in the potential of this approach.6)‐glucosidic bonds. chemical modification or combinations thereof. The application of a number of prospective methods for developing starch‐based biomaterials will be described in this thesis.Summary
Plastics made from fossil resources (a. Starch‐based biodegradable materials are considered interesting candidates to replace certain types of conventional plastics. containing periodic branches linked with the backbones through α‐D‐(1. After use. Examples are the use as packaging and construction materials. 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. we have
. it has a high volume to weight ratio and is generally resistant to degradation. Starch must therefore be modified before it can be applied as a biodegradable plastic. Amylose is essentially a linear polymer in which AHG units are predominantly connected through α‐D‐(1. such as thermoplasticization. there is a strong need for renewable alternatives for plastics from fossil resources.4)‐glucosidic bonds.
4‐di‐O‐isopropylidene‐α‐D‐galactopyranose ratio (at constant Al(OiPr)3 intake). depending on the ε‐CL to starch ratio. The use of higher carboxylic (fatty) acid to esterify starch resulted in products with significantly improved
. with higher ratios leading to higher molecular weights. even in the presence of plasticizers. 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. are described in Chapter 2. Most of the previous starch esterification studies involved the use of short chain carboxylic acids (C1‐C4).4‐di‐O‐isopropylidene‐α‐D‐ galactopyranose) using Al(OiPr)3 as the catalyst. and off‐white solid products with isolated yields of 30‐96% were obtained. Esterification of starch with carboxylic acid derivatives is one of the oldest strategies to improve starch properties. The silylation reaction was performed using hexamethyl disilazane (HMDS) in DMSO/ toluene mixtures at 70°C.3. 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. The grafting efficiency for the desired reaction was 28‐58%. Poly‐(ε)‐ caprolactone grafted silylated starch co‐polymers with average chain length of 40‐ 55 monomer units (molecular weight of 4500‐6300) were obtained.Summary
initially performed research on the ROP with a simpler model system involving a simple protected mono‐saccharide instead of starch. with higher temperatures (100°C) leading to lower yields. 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. The polymerizations were performed at 60‐100°C.21‐0. and particularly acetic acid derivatives (C2) have received considerable attention. The results for the model system involving the ROP of p‐dioxanone initiated by hydroxyl groups of a protected monosaccharide (1. and subsequent removal of the silyl groups.2. 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. The yield of the polymers was a function of the reaction temperature and the reaction time.3. A statistical model has been developed to quantify the effects of process variables (time. temperature and monomer: monosaccharide ratio) on the average degree of polymerization. The hydrophobicity of starch acetates is higher than virgin starch. The synthetic procedure may be divided into three steps: hydrophobisation of starch by the introduction of SiMe3 groups followed by in situ ROP.2. Considerable amounts of ε‐CL homopolymers with isopropyl end‐groups were also formed. The DS of the PCL chains was between 0.46‐0. The results are described in detail in Chapter 3.68) was obtained. but the products are still very brittle.72. Silylated starch with a low to medium DS (0.
and the type of catalyst. K2CO3. Mechanical tests show that the products have tensile strength (stress at break) between 2. In Chapter 5. 1. vinyl ester to starch ratio. The DS of the products is a strong function of the basicity of the catalyst.5 for starch stearate).39) are totally amorphous whereas the low‐DS ones (DS= 1.26‐2. the DS decreased. however. The starch esters were prepared by reacting starch with vinyl laurate or vinyl stearate in the presence of basic catalysts (Na2HPO4. 0.3‐1. usually performed using fatty acid chlorides which are relatively expensive and rather corrosive.1‐2.75) are still partially crystalline.07‐1.4 could be obtained for both laurate and stearate esters.4‐3 for starch stearate). With these catalysts. reaction temperature. The yellowish products were characterized by 1H‐. Statistically adequate (R2≥0. High‐DS products (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. products with a DS > 2. and Na‐ acetate).5 MPa. The use of methyl and glyceryl esters results in products with only relatively low DS. The products have a broad range of degree of substitution (DS = 0.7‐3. The last part of this thesis deals with reactive blending of starch and biodegradable polymers using reactive interfacial agents.5 for starch laurate. and modulus of elasticity of 46‐113 MPa. The synthesis of these fatty starch esters is. K2CO3 and Na‐acetate are superior with respect to activity when compared with Na2HPO4.96 and P‐value of ≤10‐7) second‐order mathematical models correlating the effect of process variables (vinyl to AHG‐ starch mol ratio. At higher ratios. elongation at break of 3‐26%. Reaction temperature and catalyst intake also affect the product DS but only to a lesser extent. the synthesis of two polycaprolactone (PCL)‐based reactive interfacial agents. presumably due to a reduction of the polarity of the reaction medium. a systematic study. The use of Na2HPO4 resulted in low‐medium DS products (0. 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.96). The starch esters were synthesized by reacting gelatinized starch with vinyl laurate or vinyl acetate in DMSO in the presence of basic catalysts (Na2HPO4. The thermal and mechanical properties of some representative product samples are also described. while the use of K2CO3 and CH3COONa catalysts resulted in medium‐high DS products (2. For low vinyl‐ester to starch ratios. In Chapter 6. and Na‐acetate) in DMSO at 110°C. an alternative method for higher fatty starch esters by using fatty vinyl esters is reported. catalyst intake. 13C‐NMR and FT‐IR.45‐1. PCL‐g‐
.9 for starch laurate. Summary
mechanical properties and hydrophobicities. The DS of the products was a clear function of the chain length of the fatty ester. K2CO3. and catalyst basicity) to the DS of the starch ester products were developed.24 ‐ 2. The thermal stability of the esterified products is higher than that of native starch. an increase in the vinyl‐ester concentration led to higher product DS values. In Chapter 4‐5.
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). mutual solubility of the monomer into the molten polymer) on the functionalisation degree (FD) of the GMA and DEM units to the PCL backbone. at relatively high GMA and BPO intakes. As the result. The latter hypothesis is indirectly confirmed by the observation that PCL‐g‐GMA becomes more efficient at relatively higher starch contents in the blends. PCL‐g‐DEM seems to have slightly more efficient compatibilising effect than PCL‐g‐GMA as shown by blends morphology and elastic modulus values. At a fixed starch content (20 %‐wt). 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).
. and their use in compatibilising starch‐PCL blends is described. A statistically adequate model (R2=0. P‐value ≤10‐9) has been developed to quantify the effects of process variables (monomer and initiator intake.957. 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. 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. the mechanical behavior of the compatibilised blends is in general different from that of pure PCL and of the corresponding uncompatibilised blends. Highest values of the FD (up to 45 %) were observed for PCL‐g‐GMA. Both of the PCL‐based compatibilisers improve the interfacial adhesion between the starch particles and the PCL matrix in starch/ PCL blends.
Zetmeel is relatief goedkoop en te winnen uit een breed scala aan planten. De verhouding amylose en amylopectine in zetmeel varieert en hangt grotendeels af van de plant waaruit het zetmeel gewonnen wordt. Er zijn verschillende methodes beschikbaar om zetmeel te modificeren. Twee type polymeren zijn dominant aanwezig in zetmeel: amylose en amylopectine. Voorbeelden zijn thermo‐ plastificatie. 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. 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. De AHG eenheden kunnen op een aantal manieren aan elkaar gekoppeld worden. Biologisch afbreekbare polymeren op basis van zetmeel kunnen mogelijk bepaalde conventionele kunststoffen vervangen.v.4)‐glucoside bindingen verbonden zijn.6)‐glucoside bindingen met de hoofdketen zijn verbonden. De synthese van dergelijke producten met een in situ katalytische ringopening polymerisatie (ROP) van geschikte monomeren als b. ε‐caprolactone op de hydroxy‐groepen van zetmeel is helaas niet eenvoudig. Amylose is een lineair polymeer waarin de AHG eenheden hoofdzakelijk via α‐D‐(1. Het gebruik van zetmeel voor verpakking materialen is momenteel beperkt. chemische modificaties en combinaties van deze methoden. heeft geleid tot een grote interesse voor biologisch afbreekbare alternatieven gemaakt uit hernieuwbare grondstoffen. De productie van kunststoffen is sinds 1950 met bijna 10% per jaar gestegen. Modificatie van zetmeel kan het aantal toepassingen sterk vergroten. Zetmeel is een polymeer bestaande uit anhydroglucose (AHG) eenheden. in combinatie met de huidige hoge prijzen voor de grondstoffen. De toepassing van een alternatieve methode waarbij zetmeel eerst minder hydrofiel gemaakt wordt door de introductie van grote hydrofobe SiMe3 groepen. gevolgd door de ROP met een polyester precursor en verwijdering van de SiMe3 groepen
. Na gebruik kunnen de materialen in het milieu eindigen en grote problemen veroorzaken. Dit proefschrift beschrijft een aantal methodes om nieuwe polymere materialen op basis van zetmeel te maken. Het zetmeel is daarnaast ook erg gevoelig voor water. Het amylopectine is een vertakt polymeer waarbij de vertakkingen via α‐D‐(1. mede omdat het niet mogelijk is om films met goede mechanische eigenschappen te maken. menging met andere materialen.Samenvatting
Kunststoffen gemaakt van fossiele grondstoffen als olie en gas zijn zeer aantrekkelijke materialen voor een breed scala aan toepassingen. Dit.
De gevolgde procedure bestaat uit drie stappen: hydrofobisering van het zetmeel door de introductie van SiMe3 groepen. De effecten van procesvariabelen (tijd.4‐di‐O‐isopropylidene‐α‐D‐galactopyranose (bij constante i Al(O Pr)3 concentratie). De resultaten voor het modelsysteem waar de ROP van p‐dioxanone wordt geïnitieerd door de hydroxy‐groepen van een beschermde mono‐saccharide (1. gevolgd door in situ ROP met ε‐ caprolacton en daarna de verwijdering van de silyl groepen.3.2. waarin hogere temperaturen (100°C) tot lagere opbrengsten leiden. De grafting van ε‐ caprolactone aan het gesilyleerde zetmeel werd in THF bij 50°C uitgevoerd met Al(OiPr)3 als katalysator. De opbrengst van de polymeren was een functie van de reactietemperatuur en de reactietijd.
.72 en was afhankelijk van de ε‐caprolacton/zetmeel verhouding. 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. 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%).46‐0.Samenvatting
zou in theorie moeten leiden tot een hogere grafting efficiency.2.68). Er werden ook aanzienlijke hoeveelheden homopolymeer van ε‐ caprolacton met isopropyl eindgroepen gevormd. temperatuur en monomeer: monosaccharide verhouding) op de gemiddelde polymerisatie graad zijn gekwantificeerd met een statistisch model. De resultaten worden in Hoofdstuk 3 in detail beschreven. De kennis verkregen met het modelsysteem is toegepast om ε‐caprolactone te graften op zetmeel.3. De polymerisaties werden tussen 60‐100°C uitgevoerd en leverde vaste producten in geïsoleerde opbrengsten tussen de 30 en 96%. 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. Een hogere verhouding leidde tot hogere molecuul massa’s. DS = 0. De silylerings reactie werd bij 70°C uitgevoerd met hexamethyl disilazane (HMDS) in DMSO/tolueen mengsels.21‐0. De DS van de PCL ketens lag tussen de 0.4‐di‐O‐isopropylidene‐α‐D‐galactopyranose) met Al(OiPr)3 als katalysator worden in Hoofdstuk 2 beschreven. De grafting efficiency van de gewenste reactie was 28‐58%. 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 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. De substitutiegraad van de SiMe3 groepen varieerde van laag tot middelhoog (degree of substitution.
De effecten van belangrijke procesvariabelen (vinyl ester/zetmeel molverhouding. In de meeste zetmeel esterificatie studies wordt gebruikt gemaakt van kleine carbonzuren (C1‐C4).24 tot 2. De temperatuur en de katalysator concentratie beïnvloeden de DS ook maar in mindere mate. en in het bijzonder azijnzuur derivaten (C2). naar de synthese van lange vetzuur esters uit maïszetmeel (zetmeel laureaat en zetmeel stearaat. De thermische en mechanische eigenschappen van sommige representatieve producten zijn uitgebreid bepaald. K2CO3 en Na‐acetaat) in DMSO bij 110°C te laten reageren. Met deze katalysatoren worden voor zowel de laureaat als stearaat esters producten met een DS van boven de 2. Echter. 13C‐NMR en FT‐IR gekarakteriseerd. 0. wordt een systematische studie. De substitutiegraad varieert van 0. inclusief statistische modellering. K2CO3. de vinyl‐ ester/zetmeel verhouding en het type katalysator. zelfs in aanwezigheid van plastificeermiddelen. De acetaat esters zijn in het algemeen minder hydrofiel dan zetmeel. Het gebruik van hogere (vet‐) zuren resulteerde in producten met beduidend betere mechanische eigenschappen en hydrofobiciteit.07‐1. Echter de producten zijn nog steeds zeer bros. vermoedelijk als gevolg van een verlaging van de polariteit van het reactiemengsel. De producten werden met 1H‐. De DS van de producten is een sterke functie van de basiciteit van de katalysator.5 voor zetmeel laureaat.96.5 voor zetmeel stearaat). 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. De synthese van deze esters wordt in het algemeen uitgevoerd met vetzuurchloriden die vrij duur en sterk corrosief zijn. K2CO3 en Na‐acetaat vertonen superieure activiteiten vergeleken met Na2HPO4. bij verdere verhoging van de verhouding verminderde de DS. Het gebruik van methyl en glyceryl esters resulteert in producten met een lage substitutiegraad. en Na‐acetaat) te laten reageren. De DS van de producten was een duidelijke functie van de keten lengte van de vetzure vinylester.3‐1.96 en P‐ waarde van ≤ 10‐7). Het gebruik van Na2HPO4 resulteerde in producten met een laag tot middelhoge DS (0. In Hoofdstuk 5. wordt een alternatieve methode voor de synthese van zetmeel esters met lange vetzuren beschreven. katalysator concentratie en katalysator basiciteit) op de DS zijn gekwantificeerd met een statistisch model (R2 ≥ 0. De zetmeel esters werden bereid door zetmeel met vinyl laureaat of vinyl stearate in DMSO in de aanwezigheid van basische katalysatoren (Na2HPO4. De zetmeel esters werden bereid door het zetmeel met vinyl laureaat of vinyl stearaat in aanwezigheid van basische katalysatoren (Na2HPO4.4 verkregen. terwijl het gebruik van de K2CO3 en Na‐acetaat katalysatoren
. In Hoofdstuk 4‐5. Samenvatting
De esterificatie van zetmeel met carbonzuur derivaten is één van de oudste strategieën voor het verbeteren van zetmeeleigenschappen. Bij lage vinyl‐ester/zetmeel verhoudingen leidde een verhoging van de vinyl‐ester concentratie in eerste instantie tot hogere DS waarden. reactietemperatuur.
De producten met een hoge DS (2.Samenvatting
resulteerde in producten met een middelhoog tot hoge DS waarden (2. De hoogste FD waarden (tot 45%) werden gevonden voor PCL‐g‐GMA bij relatief hoge GMA en BPO concentraties.39) zijn volledig amorf terwijl de producten met een DS van 1. Beide reactieve compatibilisers verbeteren de interactie tussen de zetmeeldeeltjes en de PCL matrix. Er is een statistisch model (R2 =0.5 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. De thermische stabiliteit van de geësterificeerde producten is hoger dan die van zetmeel. Het laatste hoofdstuk van dit proefschrift (Hoofdstuk 6) behandelt de menging van zetmeel en biologisch afbreekbare polymeren met reactieve compatibilisers. Zo is bijvoorbeeld de elastische modulus van de gecompatibiliseerde mengsels beduidend hoger (tot 430 MPa) dan die van niet‐gecompatibiliseerde mengsels (320‐340 MPa). 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. De FD waarden voor PCL‐g‐DEM waren aanzienlijk lager (tot 7%). een uitrekking tot breuk van 3‐26%. De synthese van deze verbindingen wordt uitgebreid beschreven. 1.45‐1. Dit leidt tot verschillen in mechanische eigenschappen tussen gecompatibiliseerde mengsels en van de niet‐ gecompatibiliseerde mengsels van zetmeel en PCL.957 en P‐ waarde van ≤ 10‐9) ontwikkeld waarmee de effecten van procesvariabelen (monomeer en initiator concentratie. en een elasticiteit modulus van 46‐113 MPa.0 voor zetmeel stearaat).26‐2. Als reactieve compatibilisers zijn twee polycaprolactone derivaten. 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).4‐3. gebruikt.1‐2.7‐3.75 nog gedeeltelijk kristallijn zijn.9 voor zetmeel laureaat.
Francesco Picchioni. Anne Appeldoorn. for the inspiring ideas during the discussions. Gert Alberda van Ekenstein and Harry Nijland helped me a lot with mechanical and thermal property measurements. I also want to give my deep appreciation to my reading committee. for his efforts to make every foreign students feel at home in Groningen. and for helping me arranging many things when I was not able to do it from Indonesia. for her hospitality and solving administrative matters. who was my first daily supervisor for about two months. This thesis is a summary of my research on developing novel biodegradable polymers from starch.J. Although it is indeed a very interesting topic. A. People from other groups have also been very kind and supportive to my research. for upgrading my writing quality. L. and Laurens Bosgra for being always helpful with technical issues and Jan Henk Marsman for the help with analytical equipment. Thank you very much for spending your valuable time reading my thesis and for the constructive comments and corrections you have made. and for always encouraging me to finish this thesis. I want to thank my three promoters.A. for valuable discussions during my early PhD period and for handing me over to the right hands. Erik Heeres. not in the last place because of the problems related with conventional plastics. since the mechanical properties as well as hydrophobicity of this starting material is relatively poor. the raw material for fossil derived plastics. These include the current high oil prices. 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. Broekhuis. Prof. I want to thank Marya van der Duin‐de Jonge. Prof. First of all. I want to thank Erik Heeres for being more than a good supervisor. Anno Wagenaar and Wim Kruizinga taught me how to operate the NMR machines and to interpret the spectra. I found out that developing a better fully‐biodegradable‐material from starch is not an easy job. A. together with difficulties in managing plastic waste. and Prof. and Prof. as well as taking the SEM pictures. Leon Janssen. Mosiscki. especially on his opinions about the statistical modelling and mechanical properties. Anno was also always there when
The synthesis of biodegradable materials is currently receiving a lot of attention. Prof. Marcel de Vries. I want to thank Francesco Picchioni for being always there when I needed him to discuss anything. Thank you Erik. Minnaard. I want to thank Leon Janssen for his hospitality. He has also been a good friend and always supported me during the hard times. for your consistent support. Erwin Wilbers. which improved the quality of this thesis. I also want to thank Ineke Ganzeveld. Prof.
Niels van Vegten. Ais Jenie. Judy Retti Witono. Taufik. Anna Nizniowska. Oscar Rojas.
. Thank you Tresna. I want to thank Edwin A. friendly. Wita Sondari. R. Yongki. Tante Caroline. and Tante Alma for being like my parents and family in Groningen. and many others for the good old times! Thank you Tante Ietje. Ardiyanti. Henky Muljana (my paranymph) and Anindhita Widyadhana. and for my wife and family. Francesca Gambardella. Rasrendra. Tante Smith. in the office. Yao Jie. Taufik. Puspita Kencana. Patrick and Oppie. Gerard Kraai. mbak Mia. Sari Fabianti. support. Jasper Huijsman. and Iwan Kustiawan for becoming my family. for your never‐ending blessings. mas Pandu. for your love. and on many occassions. Vincent Nieborg and Fesia Lestari. Boelo Schuur. and Henk van de Bovenkamp for cheering my days and being my friends during summer and winter. Thyara. Albert Kieviet helped me with the MALDI‐TOF measurements. pak Harry. You all have always been my inspiration and motivation. Iwan Harsono. Ignacio Melian Cabrera. Francesca Fallani. Ita. Annette Korringa. Thank you my parents and sister. Marcel Wiegman. Jelle Wildschut. Egi. Chalid. August Kurniawan. and Mahesa for being my brothers and sisters in Groningen. and the ISD staff helped me a lot with visum and residence permit arrangements as well as tax papers. Louis Daniel. Nidal Hammoud Hassan. Many master and PhD students. Tiara. I have met a lot of Indonesian students in Groningen. Hans Heeres. Togi. Wuri Raspati.Acknowledgements
I had questions or needed specific equipments. Nadia Gozali. The chemical and glassware magazijn as well as the FWN library personnel were always patient. patience. bu Ida Susanti. My special thanks are for Laura Junistia and Jan‐Pieter Drijfhout. Mahfud. Teguh. Abdul Osman. kak Roga. and passion. and although we seldom met afterwards. Mita. Thank you AW. our memories will never fade away. Farchad H.B. Ronny Sutarto. Zhang Youchun. Arjan Kloekhorst. Diana Jirjis. Tante Oppie. for everything. Tenny. Bima and Wisnu (also my paranymph). Endy. Kuslan. and assistant professors have shared the days with me in Groningen. Thank you God. Wahyudin. spring and autumn in Groningen. A lot of thanks for Henk Stegeman and Poppy Sutanto. Our collaboration has also become part of this thesis. Teddy. in the lab. Anneke Toxopeus. Asal Hamarneh. kak Atha. Agnes R. for never leaving me alone. Erna Subroto. Samuel. Thanks also for Marcel and Connie. Hofman. mbak Ika. C. visiting researchers. master students who worked with me. and helpful. Groningen is also a place where I found true friendship. Buana Girisuta and Rina Karina. Dado. Jaap Bosma. Wangsa. Diana Santangelo. M.
26‐28 November 2006 (oral presentation).K. Junistia.P.). Junistia. Heeres.J.J. presented in 11th Asia Pacific Confederation of Chemical Engineering (APCChE) Congress.M Janssen and H. J. Picchioni. Junistia. Manurung. Sugih. Heeres. Polym. A. 2008 (submitted to Carbohydr. Heeres.Janssen. Sugih.10. L. Broekhuis. Picchioni. presented in “Polymers for Advanced Technologies”.P.K. Picchioni. Sugih.K. J. A. L. Sugih. L. L. J.K. F. L. 9th National Conference organized by the Society for Polymer Science. Picchioni. and F. F. L. Appl. Sugih. Picchioni. A. Guangzhou (China).J. A. Sugih. 25‐28 September 2007 (oral presentation). Experimental Studies on the ring opening polymerization of p‐dioxanone using an Al(OiPr)3‐monosaccharide initiator system. Picchioni: Functionalization and blending of polycaprolactone. Heeres.P. 17‐20 December 2006 (oral presentation).B.A. H. A.J.K. Janssen.M.K. J. Synthesis and properties of reactive interfacial agents for polycaprolactone‐starch blends. F. F. H. Bratislava (Slovak Republic).J.J. F. Eur.J.M. Experimental and modeling studies on the synthesis and properties of higher fatty esters of corn starch. A. presented in the 1st International Symposium on Sustainable Chemical Product and Process Engineering.
. 2.K.) 6.P.A. presented in International Workshop: “From Polymer Modification to Multicomponent Polymer Systems”.P. Sci. A. A.010). Junistia. H. 2008 (accepted for publication in Starch‐Starke). 9.2008.J.List of Publications
1. DOI: 10. H. Pune Chapter. Broekhuis.P. (in press. A.J. Heeres. 2008 (submitted to J.Picchioni. Janssen. 4. and H.K. F.Heeres. F. starch‐based innovative product: The synthesis of grafted polycaprolactone‐starch by ring opening polymerisation.B. H. H. R. Sugih. 2008. and H. Polym. and F. 8. Manurung. R. Heeres: Synthesis of higher fatty acid starch esters using vinyl laurate and stearate as reactants. 5.B. A. Mumbai (India).K.1016/j. India. L. Heeres: Synthesis and properties of novel green biopolymers derived from starch and fatty acids. Polym. Drijfhout. Kuala Lumpur (Malaysia). 27‐30 August 2006 (oral presentation) 7.B. South China University of Technology.Drijfhout. A. Sugih. Heeres: Green.eurpolymj. Synthesis of poly‐ (ε)‐caprolactone grafted starch co‐polymers by ring‐opening polymerisation using silylated starch precursors. Picchioni: Biodegradable polymer materials based on polyesters and starch/proteins. 3. Sugih. Picchioni. Drijfhout.P. 2008 (accepted for publication in Starch‐ Starke). L.M.