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
Promotores : Prof. dr. Moscicki ISBN : 978‐90‐367‐3592‐6 ISBN : 978‐90‐367‐3593‐3 (electronic version)
. L. A. F. Picchioni Beoordelingscommissie : Prof. dr. dr.B. H.J.J.A. Heeres Prof. Janssen Prof. ir. L. Minnaard Prof. ir. dr.P. Broekhuis Prof. dr. A. dr.M.
and my sister
to: Tresna. my parents.
2.2.4. Starch‐based blends by melt mixing 126.96.36.199‐di‐O‐isopropylidene‐α‐D‐galactopyranose (2) 27 28 26 27 27 27
2.1.3. Cross‐linked starch 1. Starch production processes 1. Starch modifications to improve product properties 1. Plastic applications and waste issues 1. Biomaterials from starch 1.1.1.Table of Contents
Chapter 1: Introduction
188.8.131.52.184.108.40.206. Materials 2.3. Starch‐based blends by in situ polymerization 1. Thermoplasticized starch 220.127.116.11.2.2. Starch 1. Methods 18.104.22.168.22.214.171.124. Starch ‐ biopolymer blends and graft co‐polymers 1. Structure and properties of starch 1.4. Introduction 2.1.3. Typical example for the synthesis of polydioxanone end‐capped with 1. References
2 3 4 5 5 7 9 9 10 11 11 13 14 16 16 17
Chapter 2: Experimental Studies on the Ring Opening Polymerization of p‐dioxanone using an Al(OiPr)3‐ Monosaccharide Initiator System
2.3. Thesis Outline 1. Product analyses
. Starch Esters 1.1. The potential of biodegradable plastics 126.96.36.199. Biodegradable plastics from starch 1. Materials and Methods 188.8.131.52.2.3.
2.4.2. Typical example of the starch silylation procedure 3. Mechanistic aspects 2.2. References
Chapter 3: Synthesis of Poly‐(ε)‐caprolactone Grafted Starch Co‐polymers by Ring Opening Polymerisation using Silylated Starch Precursors
3.2.3. Methods 184.108.40.206. Peracetylation of silylated starch 3.1. 2.3.2. Introduction 3.2. Calculations 3. Screening experiments 220.127.116.11. Calculation of average degree of polymerization 2.2.2. Systematic studies 18.104.22.168.2.1.3. Typical example of in situ polymerization of ε‐CL with silylated starch 50 3. Results and Discussions 22.214.171.124.3.4.2. Nomenclature 2.3. Product analyses 126.96.36.199.1. Materials 188.8.131.52.1. De‐silylation of poly‐(ε)‐caprolactone grafted silylated starch co‐polymers 51 51 52 52 52 53 54 48 49 49 50 50
3. Analytical methods 3.2. 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
2.2. Synthesis of silylated starch
.6.3. Results and Discussions 184.108.40.206. Product yield 220.127.116.11. Conclusions 2.3. Materials and Methods 3. Nuclear Magnetic Resonance (NMR) 3.
1.3. Exploratory experiments 4.3. Systematic studies 4.3. Peracetylation procedure 4.3. Nomenclature 4.4. Conclusions 3. 1H‐ and 13C‐NMR analyses 18.104.22.168. Catalysts screening 22.214.171.124. Nomenclature 126.96.36.199. Materials 4. Deprotection of silylated‐starch‐g‐PCL 188.8.131.52.3. Product characterisation 4.3. Materials and Methods 4.5. Introduction 184.108.40.206. Effect of the addition of toluene as a co‐solvent 220.127.116.11.3.3. FT‐IR measurements 18.104.22.168.2.2.2. Analytical equipment 4. Typical example of the synthesis of laurate and stearate esters of corn starch 72 73 74 75 75 76 76 78 79 80 81 81 82 83 84 70 71 71 71 72
4.1.2. Determination of the Degree of Substitution (DS) 4.3.1.
3.2. In situ ring opening polymerization of ε‐caprolactone with silylated starch 57 63 64 64 65
3. Conclusions 4. Methods 4.3. References
Chapter 4: Synthesis of Higher Fatty Acid Starch Esters using Vinyl Laurate and Stearate as Reactants
4.3. Effect of vinyl ester to AHG ratio on the product DS 22.214.171.124.3.1. Results and Discussion 4.2. References
126.96.36.199. Work‐up of PCL‐g‐DEM 108 109 109 109 109 110 110
6.2. Analytical Equipment 5.2. Nomenclature 188.8.131.52.1.5. Methods 184.108.40.206.1.1. Introduction 6.2.2. Methods 5. Preparation of PCL‐starch blends with the reactive compatibilisers 110 6.1. Mathematical modelling 5.2. Compatibilizer synthesis 6.2. Experimental Design 5. Peracetylation procedure and Degree of Substitution (DS) determination 88 88 89 92 96 102 102 103
5.2. Materials and Methods 5.2.2. Analytical Methods 110
Chapter 6: Synthesis and Properties of Reactive Interfacial Agents for Polycaprolactone‐Starch Blends
6.3.3. Results and Discussion 220.127.116.11.2. Typical example of the preparation of laurate and stearate starch Esters 88 86 87 87 87 88
5. Materials 18.104.22.168.1. Product properties 5.6. Materials and Methods 22.214.171.124. Conclusions 5. Materials 126.96.36.199.1. Work‐up of PCL‐g‐GMA products 6.2. Introduction 5.Chapter 5: Experimental and Modeling Studies on the Synthesis and Properties of Higher Fatty Esters of Corn Starch
Effect of substrate (GMA/DEM) to PCL Ratio on the FD 6. Binary blends of starch and PCL 188.8.131.52.184.108.40.206.2.2. Calculation of the Degree of Functionalization (FD) of the reactive compatibilizers 6.5. 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
6. References Summary Samenvatting (Dutch Summary) Acknowledgements List of Publications
. Conclusions 6.
6.3. Thermal properties of the compatibilizers 6.4.3. Ternary blends compatibilized with PCL‐g‐GMA 6. Synthesis and Properties of Starch‐ PCL Blends 220.127.116.11.2.3.3. Effect of the BPO intake on the product FD 6.3.3. Preparation of the ractive compatibilizers 6.3.2. Ternary blends compatibilized with PCL‐g‐DEM 6.1.3. Nomenclature 18.104.22.168.1.1. Statistical mdeling 6.2. Results and Discussions 22.214.171.124.
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.
. an outline of this thesis is given. Different routes to modify starch to improve the product properties and to extend the application range will be provided. Finally.
The worldwide production of starch in 2008 is estimated to be around 66 million tons . Most of the starch produced worldwide is derived from corn.e.
80 70 60 50 40 30 20 10 0
Starch Production (Million Tons)
Figure 1. . corn. sorghum. rice. while corn seeds consist of 65‐80% starch by weight.1. Starch
75% of all organic material on earth is present in the form of polysaccharides.1. maize. . but other types of starch such as cassava.Chapter 1
1. present and a forecast of starch production  o : Europe □ : USA : Total
∆ : the Rest of the World
. or peas) and in tubers or roots (i. wheat. potato or cassava) of the plants [2‐3]. Most of the starch is produced in the USA. Plants synthesise and store starch in their structure as an energy reserve. with an average yield of 4. An important polysaccharide is starch. barley. Starch is found in seeds (i. The second and third starch producer regions are Europe and Asia [4‐5].1. The current price of corn starch is around $0. Potato accumulates starch to approximately 75 % of the dry weight in the tubers with a yield up to 21 ton starch per hectare. current and future estimations of worldwide starch production are given in Figure 1.9 ton starch per hectare . Past. and wheat starch are also produced in large amounts [2. sweet potato.e. It is generally deposited in the form of small granules or cells with diameters between 1‐100 µm . Most starch crops are very productive. Past.45 (estimated from the total volume and value of the US corn starch export ). 4‐5]. potato.
the debris is filtered from the slurry.
Figure 1. The plant material is grounded in water.a. which is again fed to a second milling process. gluten.2. After a coarse milling. the corn kernel is softened by steeping in warm water containing SO2 until the volume of the kernel increases with 55‐65%. The starch is present in the endosperms (floury and horny).b. Cross‐sectional view of a corn kernel
. Corn starch production  a. The low density gluten is separated from the starch suspension by centrifugation.2. and other foreign materials. sand. and starch granules are obtained after centrifugation from the suspension. Starch production processes Starch is generally extracted from the plant by wet milling processes [9‐10].
b. a typical corn starch production process is given in Figure 1. Typical corn‐milling operation b. The resulting starch is further washed in a cyclone and finally dried.1. Introduction
a.1. The fiber is removed using washing screens. and is embedded in a proteinaceous cellular matrix as is shown in Figure 1. After initial cleaning to remove cob. the mixture is fed to a hydrocyclone to separate the germ and the rest of the kernel. The resulting suspension from the mills contains fiber. As an example.2. and starch.
... Two types of AHG polymers are usually present in starch: amylose and amylopectin [2‐3. Anhydroglucose (AHG) unit b.
a.3. The starch granules will swell rapidly to many times of its original volume..Chapter 1
1.. Each branch contains about 20‐30 anhydroglucose units. H OH H
O H OH
H OH H
H .. The content of amylose and amylopectine in starch varies and largely depends on the starch source.) .6‐7 x 105 Da .
H2COH H OH H O H
b. Amylopectin is a branched polymer. Amylose is essentially a linear polymer in which AHG units are predominantly connected through α‐D‐(1. Chemical Structure of Starch a.a. Structure and properties of starch Starch is a polymer consisting of anhydroglucose (AHG) units (see Figure 1...
H2COH H .6)‐ glucosidic bonds . the amylose content is between 18‐28% . containing periodic branches linked with the backbones through α‐D‐(1.
O H OH
H OH H
O H OH
O H OH
H O . Amylopectin
Starch is insoluble in cold water. Typically.
H2COH H H OH H O H OH O H OH O CH2 H O H H OH H O H OH H O . but it is very hygroscopic and binds water reversibly..2.
H2COH H .. The amylose content of several common starches is given in Table 1. ..4)‐glucosidic bonds.3. but usually in the range of 1.
.1. H OH H
H2COH H2COH H O H H OH H H2COH H O H H OH H
O H OH
H O . 6].1.
Figure 1. Heating a starch solution leads to loss of hydrogen bonding in the interior of the starch granule and the starch will start to gelatinize. The linear
.. Amylose c. The molecular weight of amylopectin is higher than that of amylose and is typically 4‐5 x 108 Da .. The molecular weight of amylose is a function of the plant source and processing method.
will form a thick paste or gel.2. Cellulose nitrate was first prepared by A. and. and celluloid was
. Biomaterials from starch
1.2. Amylose content of common starches 
Starch Arrowroot Corn Hybrid amylomaize Class V Hybrid amylomaize Class VII Oat Manioc Potato Rice Sago Sweet potato Tapioca Wheat Amylose (%) 20.5‐70 62‐72 63‐72 58‐64
1. Table 1.1.7 26
Table 1.5 28 52 70‐75 27 15. depending on the amount of water present.1. The resulting suspension contains a mixture of linear amylose molecules. Starch gelatinization temperature range 
Starch Potato Tapioca Corn Waxy corn Wheat Gelatinization Temperature Range [o C] 59‐68 58.7 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 . Plastics are used in a wide range of applications and the demand is still increasing every year . The gelatinization temperature range of various starch sources is given in Table 1. and granule fragments. swollen granules. The first generation of commercial plastics was derived from cellulose nitrate and is known as celluloid [12‐14].8 16.2. Introduction
amylose molecules leach out of the granules into the solution. Parker in 1838.2. Plastic Applications and Waste Issues Plastic is the general term for a wide range of synthetic or semisynthetic polymerisation products.8 17.5 25.
Important polymers used for packaging are polyethylene (HDPE and LDPE). when polyethylene (PE) was invented.Fibers
1950: 1. discovered by Baekeland in 1907 [12. electrical. and consumer products (see Figure 1. the worldwide plastics production has reached 245 million ton per annum  (see Figure 1. and energy during transportation of goods. Sealants. A dramatic increase in demand for plastics began after World War II. About 29% of the total plastics produced in the USA .4. polypropylene (PP). polyvinyl chloride (PVC). Thermosets. Adhesives. Hyatt in 1870 .5. space. Worldwide Plastics Production (1950‐2006) 
. 15]. The largest application of plastics is for packaging purposes.).)
Included: Thermoplastics. Coatings. Plastics have lower melting temperatures compared to glass and metals. and polycarbonates (PC). and 37% of the total plastics demand in EU  is used as packaging materials. PA-. The latter is important when using the material for packaging purposes to save weight. The production and consumption of plastics has increased significantly with a rate of almost 10% every year since 1950. Elastomers.5
0 1950 1960 1970 1980 Year 1990 2000 2010
Figure 1. polyethylene terephtalate (PET).4. They have a low density and can be shaped in thin films while maintaining good properties. and therefore need less energy to shape it into useful materials . and PP-Fibers
2006: 245 2002: 200
200 Plastics Production (Million Tons)
Not Included: PET-.Chapter 1
patented by J. for instance to be used in packaging and paper coatings . While celluloid is derived from a natural polymer (cellulose). PE is a very versatile plastic because it can be shaped easily into various forms. the oldest purely synthetic plastic is Bakelite. Plastics are also used for building materials and automotive. Polyurethanes. Plastics are very attractive materials. polystyrene (PS). and PolyAcryl. In 2006.
which are usually found in waste streams. are not easy to recycle . and accumulation of its residues in soil cause significant reductions in agricultural yields. only 20% of the plastic is recycled. and other public works . hydropower plants operation.7% in 2006 . however. Recycling (part of) the plastics to reduce the amount of MSW also has limitations. 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). Introduction
Plastic waste. It is not applicable for thermoset resins . The disposal of plastics materials in municipal solid waste (MSW) is a serious issue in many parts in the world. soil. and may be more than twice the weight fraction . Plastic applications and waste treatment in Europe  1. Co‐mingled plastics. Biodegradable plastics are polymeric materials capable of decomposing when given an
. irrigation. rivers. The volume fraction of plastic in MSW is much larger due to the low density of plastics. consisting of 25%‐v of the total waste . and is only effective for single plastic sources or simple plastic formulations. and lakes. plastics waste was ranked as the second major source of MSW after paper and paperboard.2. is causing serious environmental problems. In Europe (2006).5. 6]. threatening fishery. Plastic litter is hazardous to wildlife [1. In 1996. In the USA. ship navigation. Most of the waste is still disposed (50%) by landfill or incinerated to recover the energy (30%) . It has a high volume to weight ratio and is resistant to degradation. the amount of plastics in MSW increased from less than 1% in the 1960 to 11.2.
Figure 1. 16]. Plastics have been polluting sea [1.
animals.96 proposal ‐ Degradable plastics are plastic materials that undergo bond scission in the backbone of a polymer through chemical. and since 2000 they have become competitive with traditional materials in some applications [20‐21]. The change in chemical structure results from the action of naturally occurring micro‐ organisms.2 working group on biodegradable polymers ‐ Biodegradation of a plastic material is a process leading to naturally occurring metabolic end products. biopolymers from nature (from plants. together with food and yard waste as well as paper. For instance. because they take part in nature’s cycle of renewal. proteins (present abundantly in plants and animals). and microorganisms) and biodegradable synthetic polymers . The price of these biodegradable plastics is expected to be reduced considerably in the next decade . Nowadays new types of biodegradable plastics with improved properties and lower costs have been developed .5/
. ASTM sub‐committee D20.Chapter 1
appropriate environment and sufficient amount of time . The market price of biodegradable plastics (mainly from starch and PLA) has decreased the last years. biological and/or physical forces in the environment at a rate which leads to fragmentation or disintegration of the plastics. Common natural biopolymers are carbohydrates. the price of PLA based plastics in 2010 is estimated to be around € 1. Biopolymers are inherently biodegradable. The considerable growth of interest in composting as a means to replace landfill due to the decreasing disposal spaces (especially in Europe) may also help the progress of biodegradable plastics development . Biodegradable plastic waste may be treated in composting facilities. and polyesters from micro‐organisms . Several authorities have provided definitions for biodegradable plastics . 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 have gained considerable interest since the 1980s. It may also be treated in sewage sludge water treatment plants or buried in the soil [20‐21]. DIN 103. Biodegradable plastics may be classified into two general groups. 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.
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 . tensile and flexural strength) and is too sensitive to water [26‐27]. Starch may become an attractive raw material for plastics in the future. pots).4 per kg in 2007‐2008 ). 20].2.3. Introduction
kg. Virgin starch is not suitable as a packaging material. before they can be applied as biodegradable plastics. blending with other materials. These are summarized briefly in the next sections. half of its price in 2003 (around 3 Euro/kg). alone or complexed with natural or synthetic biodegradable polymers) or from polylactic acid which originates from the fermentation of a starch feedstock. Biodegradable Plastics from Starch Starch is a very attractive source for the development of biodegradable plastics. Examples are the use as packaging material for carrier bags. utensils . when recycling is not practical or uneconomical [20‐21]. either from starch‐based materials (slightly modified starch. or when environmental impacts have to be minimized.50/ kg in 2005) is also expected to be lower in the future due to considerable reduction in modification costs . chemical modification or combinations of them . hygienic materials (diapers. napkins) . biofillers for tires. The price of starch‐based biodegradable plastics (€ 1. Biodegradable plastics are also used for food servicewares . . towels.3. breathable fabrics.8‐ 2. Bastioli [20‐21] showed that nearly all biodegradable plastics available in the market are derived from starch. starch must be modified.2‐ 1. because the price of oil based polymers may still increase due to the rise in the crude oil prices .
. This price is much lower than conventional plastics derived from oil. It cannot be shaped in films with adequate mechanical properties (high percentage elongation. such as polyethylene (PE.45 per kg . € 1. Consequently. consumer products and food products.4 or $ 1.50‐4.
1. Biodegradable plastics are especially very useful for single‐use applications. and chewable items for pets [1. either by plasticization. Starch modifications to improve product properties
Several techniques may be applied to develop starch based biomaterials with improved properties. The price of starch in 2007 was about $ 0. 1. agricultural products (mulch films.
The thermoplasticization process will decrease the interactions of the molecular chain and destruct the structure of the starch . water content. 33]. 32.1. which still limits their use. There are several substances used as plasticizer for the preparation of thermoplastic starch (TPS). 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. with better properties than virgin starch in various applications. 40]. The plasticized starch properties may be tuned by changing the temperature of processing. resulted in the formation of an amorphous mass [35. plasticized. the semicrystalline structure of starch and its granular form are lost and the starch polymers are partially depolymerized. As the result. such as water and polyols (glycerol. sugars) [30. TPS synthesized from polyol and sugar plasticizers have the tendency to re‐crystallize (retrogradation) after being stored for a period of time. 30. and hexylene glycol.3. The use of other plasticizers (for example glycerol) results in a rubbery material. and therefore are processable by conventional polymer processing techniques such as injection. The modification involves break down of the starch granular structure by the use of plasticizers at high temperatures (90‐180 oC) and shear. sorbitol. because the resulting product will be brittle when equilibrated with ambient humidity .Chapter 1
1. . extrusion. Thermoplasticized starch Virgin starch is brittle and difficult to be processed into articles due to its relatively high glass transition temperature (Tg. 32‐ 36]. This problem is mainly caused by the presence of strong inter‐ and intra‐molecular hydrogen bonds between the starch macromolecules [28. which results in embrittlement. The brittleness is known to increase in time due to free volume relaxation and retrogradation. which will result in a continuous phase in the form of a viscous melt [29‐30. 39]. 35. and blow moulding [28. which is even above the thermal degradation temperature. destructed. Starch can be modified to obtain materials which melt below the decomposition temperature . Although thermoplasticization seems to be a promising method. approximately 230 oC) . Another issue is the poor water resistance and low strength. glycol. Yu. the thermal properties of glycerol‐plasticized starch are a function of water content [31. phase separation may still occur. 40]. 30]. The plasticizers are also usually hydrophilic and can be
. For instance. et al  discovered that the elongation of break of the thermoplastic starch is significantly improved by plasticization with glycol. glycerol. and the properties and amount of plasticizers. 37]. The modified products are known as thermoplastic. At intermediate water levels. or melted starch [29. The use of water as a plasticizer is not preferable.
These hydroxyls can react easily with a wide range of compounds such as acid anhydrides. The cross‐linked products are therefore more rigid (and less elastic) materials than virgin starch . et al  studied the mechanical properties of films from maize starch cross‐linked by sodium trimetaphosphate (SMTP). and adhesive industry. 46]. Substitution of the hydroxyl groups of starch with acetate groups makes the esters more hydrophobic than native starch. Breakdown of the starch and the formation of low molecular weight products occur to a significant extent . or coating with hydrophobic polymers [33.3. Schuetzenberger acetylated starch with acetic anhydride . especially in the food. textile. Silva. As a result. The most popular starch ester is starch acetate . The DS is defined as the moles of substituents per mole of AHG units . High DS starch acetate can be easily casted
. Medium DS starch acetates (0.2) are commercially available and used as food additives and in the textile industry. Cross‐linked starch An anhydroglucose molecule of starch contains two secondary and one primary hydroxyl group.3‐1) are still soluble in water. paper. 39].01‐0. differing in Degree of Substitution (DS). epoxy. leading to higher viscosities [42‐43]. Most studies dealt with the synthesis of starch esters of C1‐C4 carboxylic acids.3. Three different types of starch acetates may be distinguished. the use of formyl esters of starch is much lower than that of acetate esters. Low DS products (0. Solutions to improve the properties of TPS are blending. Cross‐linking results in a reduction of the solubility in water and to thickening. The simplest starch ester is starch formate (C1).3. 1. and particularly with acetic acid . Introduction
washed out by water . 1. and ethylenic compounds. aldehydes.2. Chemicals of these classes having two or more of the reactive groups may react with two or more hydroxyls of the starch molecules. Higher processing temperatures generally led to higher cross‐linking levels. while highly substituted starch acetate (DS of 2‐3) are soluble in organic solvents. Already in 1865. The products are called cross‐linked starches. organic chloro‐compounds. The esterification reaction is catalyzed by H+ and proceeds with the formation of water . Starch Esters The development of starch esters started in the mid 19th century [44. synthesized by direct addition of formic acid to starch at room temperature. The cross‐linked starches have found many applications. resulting in an increase of the Young’s modulus and tensile strength of the products and a decrease in elongation at break.
DMF. DMSO. and phosphate) will result in C‐2. but also C‐6 and C‐ 3 substitution . Another attractive route involves the use of vinyl esters as reagents . Esterification may be performed using acid anhydrides in aqueous media or organic solvents (pyridine.1. xylene. hydrogencarbonate. Acetylation of starch in water and DMSO using vinyl acetate has been studied lately by Mormann and Al‐Higari  as well as Dicke . et al [55‐56]. Starch esters have also been synthesized using alkanoyl chlorides [49‐53]. Neutral and weak acid/ alkaline catalysts result in regioselective substitution at the C2 hydroxyl groups of starch. Several synthetic routes have been developed for starch esters. The kinetics of the reaction between gelatinized aqueous potato starch and vinyl acetate was studied by De Graaf. while alkaline catalysts (such as carbonate. High DS starch acetates are thermoplastic materials suitable to be used as biodegradable plastics . Mechanism of starch acetylation using vinyl acetate
. The degree of acetylation of starch acetates can be easily controlled. and is shown in Scheme 1.
Starch OH + OH
O Starch O
+ H3C C O CH
O H3C C O
C O CH O Starch
O Starch + H3C C H
O H3C C O Starch
C O OH
O H3C C OH
Scheme 1. allowing the polymer to be produced with a range of hydrophobicities . Starch triacetate has been successfully synthesized using acetic anhydride in combination with pyridine‐gelatinized starch .1. or isopropanol) with acidic (hydrochloric or sulfuric acid) or alkaline (NaOH or triethylamine) catalysts . acetate.Chapter 1
into films using organic solvents.
polymers are rarely miscible with each other  so that. However. which in turn provide a very strong adhesion at the molecular level between the dispersed phase and the matrix. The introduction of longer acid chain is expected to reduce the brittleness of virgin starch and to increase its hydrophobicity . 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 . The first involves simple melt mixing of the two components for example by extrusion.g. in order to achieve also a good molecular adhesion between the phases by melt mixing. 1.3. The fatty acid chloride reactants are. one component will be dispersed into the other. Introduction
The synthesis of long chain fatty acid ester of starch has attracted much interest lately.e. the two components are chemically linked to each other. Two main methodologies are applied for the production of polymer blends. dodecanoyl. Starch ‐ Biopolymer Blends and Graft Co‐Polymers Blending of different polymers is an established method to obtain products with improved properties. If the right combination of chemical groups on the two components is present along the polymers backbone.34‐0.4. it is frequently used in order to chemically graft the polymerized component on the other one (polystyrene on polybutadiene in our example above ). 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. Fatty‐acid starch esters have been synthesized using fatty acid (octanoyl. As a result. octadecanoyl) chlorides [49‐53]. the latter are mixed together. however. i. in the simplest case of a binary blend. a chemical reaction might take place upon processing (reactive extrusion). Although the in situ polymerization process is not as technologically straightforward and economically convenient as melt‐mixing. relatively expensive and rather corrosive. By working at temperatures above the melting point or glass transition temperature of the two components. 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 . Their role is comparable to the one of a surfactant in emulsion formation . 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 . interfacial agents (e. but only relatively low‐DS (0. availability of mixing equipment and no necessity to use any organic solvent (often employed for the in situ polymerization).61) products could be obtained using this approach. melt‐mixing remains the preferred route to polymer blends mainly because of very practical reasons: low costs. they locate themselves at the interface between the
. Despite this advantage. compatibilizers) might be used. However.
). 39. there have been efforts to blend as well as to graft synthetic polymers onto starch.
Figure 1. The synthesis of compatibilizer and its use for starch/ synthetic polymer blending has also been studied. 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.4. 64‐98].3. The starch/polyethylene blends are used for agricultural mulch [27. The products have been synthesized in the lab as well as on industrial scale [32. 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 most often used synthetic polymer for blending with starch is polyethylene [27.Chapter 1
two components stabilizing the dispersion (most probably by a steric repulsion mechanism ) and providing an improved adhesion at the interface.6. 80‐ 81] or food packaging . The uncompatibilized blends of starch and
.1. 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 1.6. 66‐81]. 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]. In the past.
Some of these polyesters also have very good mechanical. 72‐73]. while the tensile strength of the compatibilized ones decreases only slightly with the starch content. synthetic biodegradable polymers have been applied. and poly‐lactones (such as poly‐butyrolactone. poly‐lactides. The tensile strength of the uncompatibilized blends is drastically reduced when starch content is increased. Introduction
polyethylene show a coarse phase separation due to differences in polarity of starch (hydrophilic) and polyethylene (hydrophobic). Polycaprolactone (PCL) is a well‐known synthetic biodegradable polyester. As is the case of blends with PE. The use of reactive compatibilizer precursors PCL‐g‐glycidyl methacrylate (PCL‐g‐GMA) [89‐90]. thus improving the dispersion of starch in the PE matrix. and poly‐caprolactone). and EVOH . dextran‐g‐PCL . while PE‐g‐GMA posseses epoxide groups. 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. The use of conventional synthetic polymer such as PE for blending with starch will only result in a partially biodegradable material. PCL‐g‐pyromellitic anhydride [94‐95]. poly‐dioxanone. is formed upon mixing. Despite these good results. 84‐98]. PCL‐g‐maleic anhydride (PCL‐g‐MA) [91. the compatibilizer. which combines excellent biodegradability with acceptable mechanical properties. and of a premade starch‐g‐PCL  for starch/PCL blends resulted in a better dispersion between the phases and in turn in improved mechanical properties compared to the uncompatibilized blends. The mechanical properties of these blends (tensile strength and elongation at break) decrease at higher starch content. As a result a graft copolymer (PE‐g‐Starch). 96].
. These biodegradable polyesters will finally decompose into non‐toxic products . PE‐g‐MA contains reactive anhydride sites. thermal and water/gas permeability properties that are even comparable to bulk non‐biodegradable polymers such as PE and PP. The mechanical properties of the blend are also improved. the uncompatibilized blends of starch with PCL give coarse phase separation and a reduction in the mechanical properties when the starch content is increased. Examples of these biodegradable polyesters are poly‐glycolide. 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. since the conventional synthetic polymers are usually poorly or non‐biodegradable. i. The ester bonds of these polymers are susceptible to attack by water and this leads to enhanced biodegradability. Studies on the blending of starch with PCL have been already described [32. 39. poly‐valerolactone.e. among which synthetic polyesters are considered very promising materials . To obtain completely biodegradable products. which both can react in situ with the hydroxyl groups of starch [70.
and the presence of liquid‐solid reaction system leads to reduced reaction rates between starch and ε−caprolactone. starch‐g‐PCL has been synthesized using toxic materials such as isocyanates . In the past. 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.
1. defined as the percentage of grafted polyester to starch compared to the total amount of homopolymer and grafted polyester) . A new strategy for the in situ ROP of ε−caprolactone on starch with the use of common ROP catalyst is therefore highly desirable. The water in starch granules competes with the hydroxyl groups of starch in the initiation step of the polymerization reaction. In Chapter 3. Starch is insoluble in the typical organic solvents used for ROP (such as toluene or THF). Chapter 2 describes the ROP of p‐dioxanone initiated by a protected monosaccharide (1. the hydroxyl groups of starch are supposed to function as initiating sites. Three routes have been studied in detail: starch esterification.2.2. starch/ biopolymer blending. Previous studies showed that common ROP catalysts such as tin octoate or aluminium isopropoxide gave low (0‐14%) grafting efficiencies (GE. the synthesis of starch‐g‐PCL is reported.3. 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 . which however is extremely air‐ and water‐sensitive.4‐di‐O‐isopropylidene‐α‐D‐galactopyranose) using i Al(O Pr)3 as the catalyst. thus resulting eventually in low GE values . a very flammable by‐product. Another possible cause for the low GE values is the heterogenous nature of the reaction.3. The highest GE value (up to 90%) has been achieved when using triethylaluminium as catalyst [97‐98]. Starch is a very hydrophilic material that always contains moisture.4. In this reaction. and starch‐g‐ biopolymer formation.Chapter 1
1. and difficult to handle since it releases ethane. The results of this study have been used as input for the synthesis of the starch‐g‐PCL.4. The method applied basically consists of three steps (temporary partial protection of starch‐hydroxyl
. leading to formation of PCL homopolymers rather that starch‐g‐PCL. Thesis Outline
The objective of this thesis is to study synthetic pathways to obtain starch derivatives with the potential to be used as bioplastic. Another approach consists of the Ring Opening Polymerization (ROP) of ε‐caprolactone monomer in the presence of starch. during the reaction.
are also reported. de Bragança. Central Science Laboratory. UK.. . York. 2002. The model. .B. Passchall). . Corn Refiners Annual Report 2007..A. in Starch: Chemistry and Technology (Ed. 2004.5. CRC Press. The synthesis of two interesting compatibilizers (PCL‐g‐GMA and PCL‐g‐DEM) for starch/PCL blends is discussed in Chapter 6. E.dk/isi/stat/rawmaterial. USA. Y.S. J. starch.L. 2007
. Boca Raton. Wurzburg: Introduction. Bemiller. Summary Report for the European Union. Aarhus. and polyhydroxyalkanoates. 2005. and removal of the silyl groups). A. 2000‐2005. O. together with the mechanical and thermal properties of the synthesized starch esters are reported in Chapter 5. Academic Press. 1984. University of Wales. Bangor. Breed.M.. 1. New Jersey. R.
1. . accessed on August 01. Holmes: Interactive European Network for Industrial Crops and their Applications. Stevens: Green Plastics. in Modified Starches: Properties and Uses (Ed. C. addition of co‐solvent. ROP of ε−CL on the remaining starch‐hydroxyl groups.html. E. 105‐122. . USA. 1986. References
. Somerville: Mini review: Plant polymers for biodegradable plastics: cellulose. Inc. Denmark: http://www.starch. The results were quantified using a statistical model. London. P. R. 2008. Website of International Starch Institute.N. Sand Hutton. Wurzburg). Part of A Global Economy. Corn Refiners Association. Fowler: Industrial markets for starch. Gywnedd. including the mechanical and thermal properties. C. Agricultural and Rural Strategy Group. Introduction
by trimethylsilyl groups. The Biocomposites Centre. a systematic experimental study on the effect of the process variables on the starch ester DS has been performed. Washington D. .F. Nawrath. USA. The use of these compatibilizers in starch/PCL blends. Whistler. and application of different catalysts) on starch esterification using vinyl fatty esters (vinyl laurate and vinyl stearate) is provided in Chapter 4. Young: Fractionation of starch. Mol.H. UK. An Introduction to the New Science of Biodegradable Plastics. UK. 1995. Inc. . Corn. Based on these preliminary results. A preliminary study on the effect of several process variables (reactant ratio. C. Princeton University Press. O. Poirier.B.C.
Cleaner Prod. 31. Germany. J. Prog. 2003. 1939. . USA. Germany. 23. Husing. in Kirk Othmer Encyclopedia of Chemical Technology Volume 22. 2000. accessed on August 19. S. Frederiksberg. O.Chapter 1
. demand and recovery for 2006 in Europe.D. Eng. . R. 1984. Wolf. . Inc. Whistler. Strong: Plastics: Materials and Processing. Properties.J. R. Bemiller. . Whistler.N. Fritz. Association of Plastics Manufacturers. 2008. J. Angerer: Technical Report Series. Washington DC.M. A. Rustgi: Biodegradable polymers. R.J. The compelling facts about plastics: An analysis of plastics production. . Applications. 1996. Fifth Edition (Ed. . 2008 . Marscheider‐Weidemann. in Starch: Chemistry and Technology (Ed. 557‐562. . London. 1990.W. 27‐40. USA. R. UK. . Bastioli: Global status of the production of biobased packaging materials. Chem. Watson: Corn and sorghum starches: production. Polym.B. Starch/ Starke 2001. Inc.epa.E.htm. the Food Biopack Conference Proceeding (Ed.. R. Seidel).. Lancaster. Starch/ Starke 2001. 53.A. 1273‐1335 . Department of Dairy and Food Science. C. J. Braun: Environmental impacts of biobased/ biodegradable packaging. U. S. USA. Hanser Publishers. Ehrenstein: Polymeric Materials: Structure. Chandra. G. Culter: Plastics Packaging. C. Link. 2000. Hanser Publishers. Applications. Weber). Hernandez.L. M. Daniel: Starch. Selke. Academic Press. 2005. Technomic Publishing Company. J. Passchall).351‐355. C. J. John Wiley & Sons. Ind. and Regulations. F.M. . Selke: Packaging and the Environment. The Royal Veterinary and Agricultural University. Inc.gov/garbage/plastic. Prentice‐Hall.. Ren: Biodegradable plastics: a solution or challenge?.E. S. Englewood Cliffs. Munich. Techno‐economic Feasibility of Large‐Scale Production of Bio‐based Polymers in Europe. .
. A. Sci. A. R. Brussels. M. B. Crank. Weith: Plastics. Munich. Schleich. Belgium. Plastics Europe. 1998. 2000. 105‐109 . . G.F.R. 53 . Patel.J. Hoboken. Denmark.. E. USA: http://www. Processing. 11. J. New Jersey. X. Website of Environmental Protection Agency. Bastioli: Global status of the production of biobased packaging materials. Inc. Institute for Prospective Technological Studies and European Science and Technology Observatory.L. Properties.
57. 21. 1491‐1494 . K. Muller. Prod. R. Polym. B. Eng. of concentrated barley starch‐glycerol‐water mixtures. P. X. C. Appl.P. a model for thermoplastic starch. the Royal Veterinary and Agricultural University. Carbohydr. 2003. measured by differential scanning calorimetry and solid state NMR spectroscopy. 1980. Y. P.G. 62.S. 53. 4. Westhoff. Starch/ Starke 2001. 43. J. Otey. Weber).. Ind. J. Carvalho. Macromol. properties.J. R. 2000. G.A. Sci. P. Vliegenthart: Interaction between dry starch and plasticisers glycerol or ethylene glycol.M. 2008.M. Yu. 13. Spain. Gandini: Surface chemical modification of thermoplastic starch: reactions with isocyanates. Martin. Schwach. . 356‐361 . Compos. 372–380 . M. Physical and mechanical properties of biobased materials – starch. 2003.H. K.H. 409–416 . epoxy functions and stearoyl chloride. Gao. J. Biobased packaging materials for the food industry. Mikkilti. Forssella. . Dev. J. .M. 1996. J. Polymer Sci.F..G. polylactate and polyhydroxybutyrate. J. 2005. W. . van Soest.F. 53. Wang. 2005. Curvelo. Ind. J. 331–336 . and Pukanszky: Thermoplastic starch/layered silicate composites: structure. E. Chambre Syndicale des Emballages en Matiere Plastique (French Plastic and Flexible Packaging Association). Introduction
European Commission Directorate General of Joint Research Centre. Polym. P.. Biodegradable thermoplastic starch. Bagdi. Petersen. Department of Dairy and Food Science. J. 53. A. X. O.K. 2005. Ma: The effects of citric acid on the properties of thermoplastic starch plasticized by glycerol. J. 1–17 . status and perspectives. 494–504 . Nielsen. 592‐595. Seville. 385–409 . Carbohydr. 1997.Lin: Note. and Couturier: Properties of biodegradable multilayer films based on plasticized wheat starch. L.B. Crops Prod. and Olsen. Parker: Phase and glass transition behaviour. Moates. A Real Case of “Force Majeure” that should be passed through to Packaging Prices Immediately. Chem. Press Release: Major Price Hikes in Plastics. A. Part C: Polym. A. Wang. Doane: Starch‐based blown films. Yang. T. 34.V. Smits. Kruiskamp. F. N.M. July 15. Frederiksberg C. interaction. K. Wang: Properties of starch blends with biodegradable polymers. Avérous. Interfaces 2006. Starch/Stärke.J.L. A. Res. Starch/ Stärke 2001. Rev. report of Food Biopack Project (Ed. 275‐282
. Denmark. Yu. Y.
139–147 . Wurzburg: Cross‐linked starches. Alric. J. M.D. 55‐78.D. Ind. E. G. P. Appl. B. P. 35.B. Jiang. 77. Prinos. B. Polym. Polym. CRC Press. C. Divers. R.B.C. K. H.H. X. Appl. . van der Zee. Chem. Mullen. J. S. M. . 1647‐1656. Inc. Eng. Coutuier: Comparative biodegradation study of starch‐ and polylactic acid‐based materials. Ibezim. B. Boca Raton. Polymer 2003. Eng. I.A. Pacsu: Starch studies: Preparation and properties of starch triesters. Mullen. Ramsay. S. Matzinos. . F.Chapter 1
. Favis: High performance LDPE/ thermoplastic starch blends: a sustainable alternative to pure polyethylene. A. M. 1. 1999. .
. 74. Aburto. C.W. (editor). 17–24 . Merrill: Properties of fatty‐acid esters of starch. 381‐384 . Polym. destructuration and hydrolysis during O‐formylation. Ind. A. 2004. 1942. Sun: Reinforced thermoplastic acetylated starch with layered silicates. 58.A. E. Polym. Grohens: Starch modifications. B. I. Feil: Morphology and enzymatic degradation of thermoplastic starch– polycaprolactone blends. E. W.A. Degrad..C.J. 24. Rodriguez‐Gonzalez. A. 11–17. Bikiaris. 46–51 . 2006. Wurzburg. Favis: Rheological and thermal properties of thermoplastic with high glycerol content. 56. Carbohydr.B. Kontoyiannis. 1986. Y.A. 65. 2594–2604 . Inc. Sagar. Pacsu: Starch studies: Possible industrial utilization of starch esters. 57. Y. Boca Raton. Feller. 1209‐1217. USA. J. Starch/Stärke 2005. Sci. Hulleman. Ribeiro. 58 (9). 2002. C. F. Carvalho. chapter 4 in book Modified Starches: Properties and Uses. Ind. . T.D. Wurzburg). 1943.W. in Modified Starches: Properties and Uses (Ed. 44. E. Vikman. 389‐398 . Journal of Applied Polymer Science 1995. Thiebaud. Levesque. O. Tserki. 705‐721. J. J. Myllarinen. W. Rodriguez‐Gonzalez. C. Gattin. Silva.P. T. Panayiotou: Properties of fatty‐acid esters of starch and their blends with LDPE. Jarowenko: Acetylated starch and miscellanous organic esters. Sci.W. 1517–1526 . J. Crops Prod.W. 2001. Qiao. C. Environ. CRC Press. 2004. J. E. O. Pilin.F. 34.T. D. Borredon. V. USA.D. Stab. Ramsay.J. Copinet. Andrade: Reactive processing and mechanical properties of cross‐linked maize starch. 1997. J.. 581–586 . O. Bertrand. Panayiotou: Processing and characterization of starch/polycaprolactone products. Polym. 1986. Chem.
P.F. 1996 . Aburto. 137‐144. Borredon. W. A. Al‐Higari: Acylation of starch with vinyl acetate in water. 145‐152. Introduction. 28. 132‐135. 101‐112. Mormann. I. 1440‐1451. USA.828. Thiebaud. W. P. New York.H. Technomic Publishing Company. Broekroelofs. 118‐121. Polym. Polym. J. . Panayiotou: Properties of octanoated starch and its blends with polyethylene. 255‐263. Borredon: Free‐solvent synthesis and properties of higher fatty esters of starch ‐ Part 2. E. G. I. 1999.E. . R. M. 2001. Macromolecules 1996.Graf. Nakayama. R. Rutot. Munich. Starch‐Starke 1999. Thesis. S.Painter: Specific Interactions and the Miscibility of Polymer Blends. Panayiotou: Synthesis. Inc.M. I. A. Bakker. Bikiaris. Macosko. E. de Graaf: Ph. Alric. E. Polym. Dubois: Synthesis and characterization of compatibilized poly(ε‐caprolactone)/ granular starch composites.. Bikiaris. Aburto. H. . Hu). A. P. P. Marechal. 302‐307. 1991. Mouysset‐Baziard.H. Prinos. Appl. Senocq. Aburto. US Patent 2. C. Degee. Carbohydr. Scott. M. 57. Alric. J. T. G.E. A. 1997. J. Thiebaud. I. Bakker and G. . . and biodegradability of fatty‐acid esters of amylose and starch. Sci. 51.W.Stokes. A. R.928.C. Alric. C. Germany. M. E. A. . Lancaster.Fennel Evans: Fundamental of Interfacial Engineering. Cellulose 2004. . W. Prinos. Starch‐ Starke 2005. Borredon. Hanser Publishers. 175. E. A. D. Hamaili. J. Borredon: Preparation of long‐chain esters of starch using fatty acid chlorides in the absence of an organic solvent. Janssen. J. in Reactive Polymer Blending (Ed. Borredon: Organic solvent‐free transesterification of various starches with lauric acid methyl ester and triacyl glycerides. C.V. Dicke. A. Aburto. E. A straight way to regioselectively functionalized polysaccharide esters. Symp. 29. . P. D. de Graaf. Carbohydr.E. 34. 33‐43
. B. Aburto. . 56. Tuschoff: Acylation of hydroxy compounds with vinyl esters. Introduction
. Alric. 2001.J.D. J. J. J. USA. 51. C. C. Hu. P. J. 74. 1960 . Duquesne. Beenackers: The kinetics of the acetylation of gelatinised potato starch. Alric. Macromol. . 1995. D. Starch‐Starke 1999. F. Smith. Inoue: Compatibilizers for melt blending: Premade block copolymers. J. Guegan. characterization. G. Wiley‐VCH. Rijksuniversiteit Groningen. 1997. . 5590‐5598 . R. L. Starch‐Starke 2004. 11. Khandpur. I.E. S.Coleman. D. M.K. C.
21. Z. Effect of amylopectin to amylose ratio in starch. Z.G.M. R.G. L. 38. 20. R. 88. J. Verhoogt: Processing and characterization of thermoplastic starch/polyethylene blends. 88. Nawang. N. A. Liu. 75–81 . .S. Ramsay. morphology.D. Res. 310–317 . 59. Mechanical properties of sago starch‐filled linear low density polyethylene (LLDPE) composites. H. J. G. III. H. Koutsopoulos.D. 1998. Chem. R. 1998. 59.Chapter 1
. Rosa: Mechanical. R. Polym. Polym.J. Ismail. Ind. 2904–2911. Almeida. Gelina. and tensile properties of Low‐Density Polyethylene (LDPE) and corn starch blends. 2003. A.‐J. Ramsay. Contat‐Rodrigo. Bhattacharya: Properties of injection moulded starch/ synthetic polymer blends. Stab. Polym.G. Degr. Wang. 1998. thermal and morphological characterization of recycled LDPE/corn starch blends. Bikiaris. D. Appl. 30. 1841‐1846 . Polym. Mater. C. Danjaji. Panayiotou: LDPE/Starch blends compatibilized with PE‐g‐ MA copolymers. Ribes‐Greus: Thermal and viscoelastic properties of some commercial starch products. St‐Pierre. C. L. Ismail. Ishiaku.I. Polym. 1503–1521 .S. Polymer. 647‐655
. B. W. N. J. 1467‐1475. E. Adv. Sci. Frangis. 167–172 . Nakamura. M. 2005.S. 2005. A. I. 1‐9 . 2001. L. Mohd Ishak: Material behaviour. Testing. Sung.A. Eng. Pedroso. W. Y. Z. 2005.S.L. D. U.A. 1242‐1251 . Pavlidou. Jane. Evangelista. J.A. Duran. Testing. Favis. Ishiaku. Panayiotou: LDPE/plasticized starch blends containing PE‐g‐MA copolymer as compatibilizer. J. 16. L.H. 34. . Z. Mei: Study and development of LDPE/starch partially biodegradable compounds. Mani. Sci.A. Polym. Nikolov. thermal and morphological properties of virgin and reprocessed LDPE/corn starch blends. Bikiaris. U. R. N. Eur. J. E. B. 1997. Polvm. Cordi. I. 2003. Pedroso. 236–241 .. Sun: Effects of polyethylene‐grafted maleic anhydride (PE‐g‐MA) on thermal properties.M. Appl. Proc. Nawang. 287‐291 . Degradation studies and moisture uptake of sago‐starch‐ filled linear low‐density polyethylene composites.S. Effect of compounding and starch modification on properties of starch‐filled Low‐ Density Polyethylene. Polym. Rosa: Effects of the compatibilizer PE‐g‐GMA on the mechanical. 70. Vouroutzis. 162–163. K. 1991. Polym.D. Technol. 2002. Tech. D. Appl. N. Prinos. Danjaji. D. H. J. Sci. J. Carbohydr. Mohd Ishak: Material properties.
K. Moro. Ind. Sci. N. Sci. K. U. Ishiaku. 179‐187. Otey.S.P. H. 26. 43.K. S. Park: Modification of aliphatic polyesters and their reactive blends with starch. Polym. J. K. L. 35.S. Lindblad. Cho. Part A: Polym. S. 1659‐1663 . F. Calil: Effect of thermal aging on the biodegradation of PCL. 41. 41. 2004. Bhattacharya Residual stresses and density gradient in injection molded starch/synthetic polymer blends Polymer 2000. B. 23. 284‐287 . 2002. J. 157. W. Ang. 1507– 1516. M. 4157–4167 . Sankar.H. Albertsson: Controlled Ring‐Opening Polymerization: Polymers with designed Macromolecular Architecture. Ryner. Dev. 40. Polym. C. A‐C. Guedes. D. Pilot plant investigation for production of the biodegradable film. F. L. 2000. Doane: Starch‐based blown films. Otey. Eng. Azemi. Westhoff. . Liu. A. 157. 2003.M. 1999. Westhoff: Starch‐Based Films. Sen. 23. 1. in Polym. Z. Makromol. C. M. Appl. Y. Averous. Adv. Adv.M. PHB‐V. 3976–3987 .. Polymer 2000. A. S. cross‐linked and glycerol modified starch. Preliminary diffusion evaluation.F. M. Maiti: A biodegradable film. 3539–3546 . T. J. B. . 2001. Raj.‐C. Ind. Tech. J. 89. Res. dos Santos Rosa. 87. Chem. Kim. J. C. S. Sci. Dole. Eng. Appl. Ranucci.K. . Polym. Kim. Polym.H. S. .W. 2. B. Polym. and their blends with starch in soil compost.K.P.G. U. Polym. Sci. 2003. 1987. Polym.M. J. 81. 2004.H. A. T. Odusanya. M. M. E. glycerol modified. P. Kim. D. Polym. Environ.
. O.C. Sci. Albertsson. Low density polyethylene/ starch blend films for food packaging applications. Prod. 2002. 267.. A.N. M. 42‐65. J. Jung. J. Ishiaku. 877–884. Polym. 16–19 . John. Jana.H. 9177–9190 . 1298‐1305. Karlsson: Polymers from Renewable Resources. 32–45 .M.S. Eng.. J. O. R. J. 1984. . Chem. C. Tang. R.D. . Manan. Appl. Res. Kammer: On mechanical properties of sago starch/poly(ε‐caprolactone) composites. Manan. Fringant: Properties of thermoplastic blends: starch–polycaprolactone. Introduction
. Azemi: Effect of starch predrying on the mechanical properties of starch/poly(e‐caprolactone) composites. Chem.R. J.Y. Rodrigues. Bhattacharya: Synthesis and characterization of anhydride‐functional polycaprolactone. 1997. Jana: Studies on the properties and characteristics of starch– LDPE blend films using cross‐linked. Sci. U.S. Eur. B. Sci.M. 12. 2007. 139‐161. Adv. 1139–1148. Park: Reactive blends of gelatinized starch and polycaprolactone‐g‐glycidyl methacrylate. Stridsberg. Garg. Odusanya. Chem. M. Siddaramaiah. Yang.
S. Narayan: Aliphatic polyester‐grafted starch‐like polysaccharides by ring‐opening polymerization. . Errico.
. 2003. 19. Macromol. P. J. Krishnan. M. P. D. 127–134. 41. 2000. E. Martuscelli. Dubois. Mani. Degrad. Raimo. Rimedio. Sadocco: Preparation of biodegradable polyesters/high‐amylose‐starch composites by reactive blending and their characterization. .E.E. 2001. Dubois: Synthesis and characterization of compatibilized poly (e‐caprolactone)/ granular starch composites. Rapid Commun. 3875–3881 . Degée. R. M. 2002. P. 7. . M. 3091‐3100.Chapter 1
. R. 40. Laurienzo. Stab. R. Appl. 1998. Avella. P. 33–43 . Tang. Duquesne. and P. Errico. Macromol. Rimedio: Polymer communication: Preparation and characterisation of compatibilised polycaprolactone/ starch composites. Composite Interfaces. M. M. J. Degee. 1432–1442. Wu: Physical properties and biodegradability of maleated‐ polycaprolactone/ starch composite. Polym. C. P. P. 83. Polym. D. Polymer 2000. M. E. 175. Polymer 1999. M. Dubois: Aliphatic polyester‐grafted starch composites by in situ ring opening polymerization. R. Symp. Sci. Bhattacharya: Synthesis and characterization of starch‐ graft‐polycaprolactone as compatibilizer for starch/polycaprolactone blends. Narayan. 215–225. . Avella. Rutot. 283–286. Rutot. 80. R.
Systematic studies showed that the polymer yield is a function of the reaction temperature and the reaction time. ring‐opening polymerization
. with higher temperatures (100°C) leading to lower yields. also polydioxanone with an OiPr end‐group was formed (20‐30 %). and MALDI‐TOF mass spectrometry.3.4‐di‐O‐isopropylidene‐α‐D‐galactopyranose)/Al(OiPr)3 initiator system to yield polydioxanone with a protected monosaccharide end‐ group is described. polyesters.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. The products were synthesized at 60‐100°C and characterized by 1H‐ and 13C‐NMR. 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). Besides the desired polydioxanone functionalised with a monosaccharide end‐group. 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.2. Keywords: biodegradable.
The ring‐opening polymerization (ROP) of p‐dioxanone using metal catalysts such as aluminum isopropoxide [Al(OiPr)3]. polyglycolide. polydioxanone also shows good performance. The synthesis of protected monosaccharide end‐capped biodegradable polymers is usually performed via Ring Opening Polymerization (ROP). which to the best of our knowledge is the first study to functionalise polydioxanone with a monosaccharide. Introduction
Aliphatic polyesters.3 MPa and an elongation at break of 500‐600%. Exchange of the alkoxide group by e.Chapter 2
2. pregnenolone. Its melting temperature is close to 110 oC. The catalysts are metal alkoxides with Lewis acidic character [8. 9]. only limited information about the synthesis and properties of polydioxanone is available in the open literature. and is tougher than polylactides and even HDPE . This chapter describes experimental studies on the catalytic ROP of p‐ dioxanone in the presence of a protected glucose molecule (1. stigmasterol and ergocalciferol)  and also with protected monosaccharides (galactopyranose/ glucofuranose) [6‐7]. reaction of the metal‐alkoxide with an appropriate alcohol allows the synthesis of end‐capped poly‐lactones. quinine.4‐dioxan‐2‐ one)). stannous octoate [Sn(Oct)2]. This melting point is considerably higher than typically found for polycaprolactone (60 oC) and lower than that of polylactides (at least 175°C).3. Polydioxanone (poly(p‐dioxanone) or poly(1. whereas the high temperature for polylactides results in thermal degradation and undesirable transfer reactions during synthesis and processing steps . tocopherol.4‐di‐O‐ isopropylidene‐α‐D‐galactopyranose. probably because the p‐dioxanone monomer has become commercially available only recently . 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. which is a unique compromise between application and processing temperature. Polydioxanone has a tensile strength close to 48. 2). Biodegradable aliphatic polyesters end‐capped with sugar molecules have been studied extensively for use in biomedical applications. and polylactides. particularly for nano‐ encapsulation systems for drug delivery . From a biodegradability point of view. or zinc lactate has been reported. accessible by the polymerization of p‐dioxanone (1). such as polycaprolactone. testosterone. Several polymers with bioactive alcoholic and phenolic end‐groups of interest for drug‐related applications were synthesized (for example geraniol.1. It is fully degraded in the body within a period of 180 days . Besides
. are interesting polymers because of their good product performance and biodegradability . The relatively low melting point of polycaprolactone limits its applicability. has interesting product properties compared to other aliphatic polyesters. Nishida et al  reported that polydioxanone decomposes to non‐toxic gases (CO2 and H2O) by microorganisms.g.2. Despite its good properties.
part of this solution (350 µl. which was pre‐heated till about 60°C to obtain it in a liquid state.08 mmol aluminum isopropoxide) was added to pure 1 (0.5 mmol) was dissolved in toluene (1 ml) at 50 oC. the mixture was brought to room temperature and several drops of HCl (1 N) were added to stop the reaction.1. Materials p‐Dioxanone monomer (1.2.55 mmol of 2 and 0.2.2. hot dichloroethane (20‐25 ml) was added to completely dissolve the partly solid polymer at room temperature. Germany) was purified according to the procedure described by Raquez et al [2. The reaction mixture was clear and colorless during the reaction. and diethylether (Labscan) were used as received. 97% (Sigma) and aluminum isopropoxide. Particularly interesting in this field are starch polymers grafted with polydioxanone. relatively low molecular weight compounds reported in this study may be viewed as model systems for more complex. 3]. Subsequently.2. 0. As such. the results of this study are also of interest for the preparation of starch/polydioxanone polymers using ROP.g.8 ml of a solution of an aluminum isopropoxide stock solution was added. Toluene (Labscan) was dried and stored on molecular sieves 3 Å (Labscan) under nitrogen. poorly soluble heterogeneous systems. grafting efficiencies are difficult to determine in this system by standard analytical techniques (e.4‐di‐O‐ isopropylidene‐α‐D‐galactopyranose (2) 2 (1.07 g.2. CDCl3 was obtained from Sigma and was used as received.2. The hot solution was
. the synthetic pathways and the soluble. The resulting clear solution was stirred for 2 h at 50 oC. 1. This stock solution was prepared by adding 4. Materials and Methods
2.2. heptane (Acros). 2.3. To this solution. NMR) due to the poor solubility of the products in standard organic solvents. Ring Opening Polymerization of p‐dioxanone using an Al(OiPr)3‐Monosaccharide Initiator System
potential applications in biomedical products. 8 mmol). After the pre‐determined reaction time.2.3. containing 0.
2. Analytical grade dichloroethane (Labscan).4‐di‐O‐ isopropylidene‐α‐D‐galactopyranose (2). Typical example for the synthesis of polydioxanone end‐capped with 1.1. Boehringer Ingelheim. (20 mmol) of aluminum isopropoxide to 20 ml of toluene.8 g. 98+% (Aldrich) were used as received. However. Next. 5. 2. Methods All polymerization experiments were carried out under a protective nitrogen atmosphere using standard Schlenk‐ and glovebox techniques. The polymerization was allowed to proceed for 16 h at 100 oC.44 g.
precipitated in a heptane/ ether mixture (300‐400 ml. This leads to the following equation:
X n exp is
H-galactopyranose end group
H-isopropoxide end group
. X n exp of the product.4. As will be shown later. two polymers with different end‐groups are present.2. one initiated on a galactopyranose molecule and the other on an isopropoxide group. If the amount of sugar is in excess with respect to the aluminum catalyst. Product Analyses NMR analyses were performed in CDCl3.2. 4:1 by volume) at 4‐8 oC.)
Here. 2D‐NMR spectra were recorded on a Varian Unity 500 NMR. 40 oC) until constant weight.
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. Processing of the raw data was performed using VNMR software. MALDI‐TOF spectra were recorded on an Applied Biosystems Voyager DE‐PRO machine using dithranol/ NaI as the matrix (linear mode). The isolated yield at this condition was 68%. 1H‐ and 13C‐NMR spectra were recorded on a Varian AMX 400 NMR.3.2. it is assumed that all available initiator is used effectively. 2. 2. The white solid was finally separated from the liquid by decantation and dried in a vacuum oven (5 mbar. the above equation simplifies to
= monomer conversion ×
[monomer ]0 [ sugar ]0
(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
H‐NMR was applied to determine the experimental average degree of polymerization.1.
2.4. Typically. eq.)
(2.3. To avoid the formation of isopropoxide end groups. the protected mono‐saccharide 2 reacts with Al(OiPr)3 to form the actual catalyst for the polymerization reaction. an excess of monosaccharide 2 on Al was used (see eq. 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. R gp .6: 96 was applied.2.6 ppm ] / 6 A 5.3.6. is calculated using:
R gp = A
H − galactopyranose end group H − isopropoxide end group
(2.5. is rewritten as:
[ A 3. Screening Experiments Initial experiments to synthesize protected galactopyranose end‐capped polydioxanone (3) were performed using 1. 2.)
The ratio of the two different types of polymers (either end‐capped with galactopyranose or an isopropoxide group).4‐di‐O‐isopropylidene‐α‐D‐ galactopyranose (2) and Al(OiPr)3 as the catalyst precursor (Scheme 2.4− 4. The polymerization reaction was performed as a two step process. in term of the NMR resonances :
R gp = A 5.6 ppm A 5. 2. To compensate for this effect.126.96.36.199 ppm
(2.). In the first step.1. The exchange reaction typically takes place at 50°C for 2 h.
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 2.6 ppm + A 5.). an Al ‐ monosaccharide 2 ‐ monomer 1 mol ratio of 1: 6.3.7 ppm − 6 × A 5. Results and Discussions
2.1. Schematic representation of the polymerization reaction including atom numbering scheme
15‐1. Product analyses The products were analyzed using 1H‐ and 13C‐ NMR and MALDI‐TOF. Typical isolated product yields are 68% at these conditions.) are not particularly informative.3.
. H‐NMR spectra (Figure 2. The OiPr : 2 end‐group ratio for the standard experiment was 1: 2.1 and 2.1.1 ppm.) In the next step.7. together with the CH3 groups from the protecting groups of the sugar appear in the range 1.51 ppm.78 to 4. It confirms that polymer initiation not only occurs with the monosaccharide but also with the remaining isopropoxide group of the catalyst precursor (vide infra). Although the spectra clearly indicate the presence of the monosaccharide 2.2. However.. The off‐white solid reaction product was collected after a dissolution‐reprecipitation process using dichloroethane and a heptane/ diethyl ether mixture.67.1 ppm compared to the monomer. the in‐situ formed catalyst was reacted with p‐dioxanone monomer 1 at 100 oC for 16 h.40 ppm) unit are broadened and shifted up to 0.1. although it is evident that p‐dioxanone polymerisation occurred. Typical 1 H‐ and 13C‐NMR spectra of 3 are shown in Figure 2. It is difficult to determine the end‐groups of the polymer on the basis of 1H‐NMR.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
(2.1. respectively. Proton resonances of the CH3 group of the isopropoxide end‐group. The typical proton resonances of the p‐dioxanone (3. 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. This multiplet is characteristic for the CH proton of an isopropoxide end‐group. of interest is the presence of a small multiplet at about 5.
2 and 96. H‐NMR Spectra of: (a).) is more informative and clearly shows the presence of a polydioxanone polymer backbone and a monosaccharide end group.2 ppm.28 ppm in 2 to δ 66. 68. For instance.3.2. characteristic resonances of an ‐O‐C(=O)‐CH2‐O‐CH2‐CH2‐OH end group are present at δ = 61. 1.2 ppm in product 3.52 ppm (β’). Particularly the C2‐C6 carbons in the range δ = 66. In addition to the mono‐saccharide end‐ group.61 (γ’) and 73.2 ppm. The carbon resonances of the polydioxanone backbone are present at δ = 63.3 ppm are shifted considerably.4‐di‐O‐isopropylidene‐α‐D‐galactopyranose. Carbon resonances arising from the monosaccharide end‐group are present between δ = 66. p‐dioxanone monomer. 2 (b).8.17 and 69. 3
C‐NMR (Figure 2.2.2‐70. Ring Opening Polymerization of p‐dioxanone using an Al(OiPr)3‐Monosaccharide Initiator System
Figure 2. 1.
. 1 (c).4‐di‐O‐isopropylidene‐α‐D‐galactopyranose‐end‐capped polydioxanone.2. C‐6 is shifted from δ = 62.3.1.
p‐dioxanone monomer. and 2.76‐25.3.1. A typical example of a part of the 2D‐NMR spectra is given in Figure 2. The resonances of the protecting group of the sugar appears at δ = 108. C‐ NMR Spectra of: (a). 2 (b).3.
Figure 2.4‐di‐O‐isopropylidene‐α‐D‐galactopyranose‐end‐capped polydioxanone. 3
Carbon resonances of the isopropoxide groups from the by‐product (isopropoxide end‐capped polydioxanone) are present at δ = 68.4‐di‐O‐isopropylidene‐α‐D‐galactopyranose.97 (>C‐(CH3)2). An overview of the data is given in Tables 2.48 ppm (‐CH‐ (CH3)2) and between δ = 21.1. the numbering scheme of carbons and protons is given in Scheme 2.70 (>C‐(CH3)2) and between δ = 188.8.131.52.3.76‐25. 1.2. 1.97 ppm (‐CH‐(CH3)2). D‐NMR (HSQC) was applied for complete peak assignment of the product. 1 (c)..73 and 109.
60 4. HSQC 13C‐1H spectra of 3 Table 2.00 4.59 4.
.68 4.31 4.40
all values were determined using 1H NMR.732 broad.78 4.301 3.1.74 3.54 4.09 4.8
Reactant peaks (ppm) Galactopyranose (2) H‐1 H‐2 H‐3 H‐4 H‐5 H‐6 5.32 * 4. Ring Opening Polymerization of p‐dioxanone using an Al(OiPr)3‐Monosaccharide Initiator System
Figure 2.26 3. around 1.25 3.50 4. which were determined from HSQC spectrum due to overlapping resonances.17 3.16* 4.
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 5.3.83 p‐dioxanone Monomer (1)
H‐α H‐β H‐γ
4. except values with *.77 3.
10 71.75 70. which is the molecular weight of a repeating dioxanone unit.Chapter 2
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. corresponding with an average number molecular weight (Mn) of about 1600.31 70.92 66.00 171. A typical molecular weight distribution is observed.52 63.61 170.28 70.57 62.54 166.21
NMR analyses also allow calculation of the molecular weight of the products. The latter likely stems from the matrix used to ionize the sample.17 69.4.
.21 70.2. For this purpose.63 68.11/ 170.2 68.28 p‐dioxanone monomer (1)
C‐α C‐β C‐γ C‐carbonyl
61. An example of a MALDI‐TOF spectrum of 3 recorded in a dithranol/NaI matrix is given in Figure 2.33 70.92 68.2 73.80 61.07
Reactant Peaks (ppm) Galactopyranose (2) C‐1 C‐2 C‐3 C‐4 C‐5 C‐6 96.8. the ratio of the intensity of the end groups and the polymer backbone peaks is determined. MALDI‐TOF was also applied to characterize the products. 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.09 65.58 68. The difference in molecular weight between the main peaks is 102 g/mol. The molecular weight distribution of the major peaks may be represented by the following relation:
m / z = 23 + 260 + 102n
This series represents a dioxanone polymer end capped with 2 and an additional Na ion.
4. see Figure 2.10.)
This relation is indicative for the presence of a polydioxanone polymer containing an isopropoxide end‐group.11. two other distributions are present.). Typical MALDI‐TOF spectra of 3 A second molecular weight distribution is also clearly visible when enlarging the spectra. These distributions may be represented by equation (2.) (2.) and imply the presence of dioxanone polymers with carboxylic end groups.5.
m / z = 23 + 18 + 102n m / z = 1 + 18 + 102n
(2. ionized with either Na+ or H+. This series may be described by the following relation:
m / z = 23 + 60 + 102n
(2. 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
1098.9.11. Furthermore.5.5 +102
Figure 2. in line with the NMR data. although both with a very low intensity (Figure 2.)
.) and (2. for details.10.
g. although successfully applied for galactopyranose‐end‐capped polycaprolactone .Chapter 2
858. In addition. small but significant amounts (20‐30%‐mol) of polydioxanone chains with an OiPr end group are present. Most likely the matrix (dithranol/NaI). the catalyst precursor is treated with monosaccharide 2 resulting in an alcohol exchange
.2. Mechanistic aspects Both NMR and MALDI‐TOF measurements imply that the main product is indeed the desired monosaccharide end‐capped polydioxanone. [6. low resolution spectra were obtained.3
836.13]. 2. different molecular weights were analysed. 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 2.1. Enlarged MALDI‐TOF spectra for 3 Although MALDI‐TOF clearly demonstrated the presence of various types of end‐groups in the product. Various other matrices were tested (e.7 796. according to NMR.5.3. but in all cases poor quality. Various samples with. is not particularly suitable for polydioxanone.1
0 600 700 800 900 Mass (m/z) 1000 1100 1200
Figure 2. it proved not suitable to determine the average molecular weight of the products. In the first step. 2‐(4′‐ hydroxybenzeneazo)benzoic acid (HABA)).6. The observed differences in the molecular weight distributions of the various samples were only marginal.
(R = monosaccharide 2)
. a p‐dioxanone molecule will coordinate to the Lewis acidic aluminium center followed by an insertion step. Irreversible termination of the polymerisation is performed at the end of reaction period by adding dilute acid to the polymerisation mixture. During the reaction. The termination reactions lead to the formation of ‐O‐C(=O)‐CH2‐O‐CH2‐CH2‐OH end groups.
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. Subsequent coordination and insertion of dioxanone molecules leads to the formation of a polymer chain with a monosaccharide end‐group. 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. Simplified reaction scheme for the ROP of p‐dioxanone catalyzed by Al(OiPr)3. The termination reaction is known to be reversible and the formed polymer may again react with an aluminum center and continue to grow . termination of the chain growth may occur by reaction with an alcohol.6. Next. The resulting Al‐alkoxide may again initiate a polymerisation reaction.
29 14.53 59. These may be present in the reaction mixture because the exchange reaction between Al(OiPr)3 and 2 was incomplete. Systematic Studies The effect of important process variables (temperature.5 67.8 2.3.28 33.22 16.5 hrs and 16 hrs) and a p‐dioxanone to monosaccharide 2 molar ratio ranging between 3. An overview of the experiments and the results are given in Table 2.45 19.3.96
For each experiment.).50 25.5 T (°C) 100 100 100 100 100 60 60 60 60 60 60 60 60 60 60 dioxanone/ 2 ratio (mol/mol) 3.3 81.8 2.30 62. 2.5 1.74 18.9 2.1 2.3 6.6 2. However.12 42.6 6.5 1. a (nearly) fixed Al(OiPr)3 : monosaccharide ratio of 1 : 6.08 30.6 89.55 Product Properties Avg. will be formed when the polymerization starts with an aluminum alkoxide with a remaining Al‐OiPr group (eq.6 6.50 35. degree of polymerisation of the product and the end group distribution was determined.5 2. A total of 15 experiments were performed at two polymerization temperatures (60 and 100 oC).74 52.11 22. 2.4 2. Chain Length (Xn
Yieldb (%) 30.3 3.1 86.10 15.Chapter 2
The minor product.4
2/OiPr ratio Rgp (mol/mol) 2.6 81.4 2.5 1.6 6.2.9
7.7.6 6.5 1.5 84.6 6.3‐6. despite the excess of 2.6 was applied.6 6. OiPr‐end‐capped poly(p‐dioxanone). Table 2.6 96.3 and 62.4 3. 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
.21 54.3 6.70 58.1 2.6 80.3 6. 2.6 91.8 3.0 87.3 6. time and the mol ratio of monomer to monosaccharide) on the yield.93 37.6 86.2 91.85 14.).93 8.5 81. formed in the first step of the polymerisation reaction (eq. In all cases.6 t (h) 16 16 16 16 16 16 16 16 16 16 1.22 57.04 19.49 13.81 16.1 3.3 6.5 92.16 46.
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) 6.78 10.3 6. two different reaction times (1.3.22 13.7 2.74 23. the end group may also be formed by a termination reaction with free isopropanol.7.5.6 6.6 6.36 17.
when the reaction is performed at a low dioxanone: 2 ratio (≅ 16).5 h. For the two experiments with a dioxanone: 2 ratio higher than 16 (≅ 23 and 60). 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. Clearly. Evidently.2. no detailed explanation has been put forward to explain this anomalous behaviour. 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. the yield at 1.3. will be required.5 and 16 h reaction time). which reduced to 75% when increasing the temperature to 120oC. However.3. 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. One of the experiments (S111) resulted in a very low yield (31%) compared to the other reaction. The effect of the reaction temperature (60 and 100°C) on the product yield at three different dioxanone: 2 ratios is given in Table 2.3.
. Product yield The isolated yields of the reactions are all but one between 67 and 96% (see Table 2. Ring Opening Polymerization of p‐dioxanone using an Al(OiPr)3‐Monosaccharide Initiator System
2. which is expected to be more pronounced at lower temperatures. For this particular experiment. the yields are higher when performing the reaction at 16 h reaction times. leading to a low average chain length of the dioxanone polymer (7. To the best of our knowledge. the monomer conversion reaches a maximum value before going down to the equilibrium monomer conversion. as suggested by Raquez et al . The latter is due to the slight exothermicity of the reaction . the equilibrium conversion of dioxanone at 80oC was about 80%. and particularly when comparing the data in set 2 and 3 (60oC. with an enthalpy of polymerization of approximately ‐15 kJ/ mol . The effect of the reaction time on the product yield may be derived from the data provided in Table 2.). et al . Higher polymerisation yields were obtained at lower temperatures.1. 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. the yields for reactions conducted at 60 oC are always higher than those performed at 100 oC. It was shown that at lower monomer to catalyst ratios. see Table 2. a low dioxanone: monosaccharide 2 ratio was applied. equilibrium conversion is not yet achieved within 1. It is likely that this relatively low molecular weight compound dissolves partly during isolation/ purification procedure leading to lower isolated yields. Another possibility is the occurrence of polymer crystallization. On the basis of our experimental data and in line with literature data.3.5 h is higher than the yield at 16 h.).3.2. For instance. 1. A more detailed analysis on the actual nature and composition of the polymerization products.
time: 16 hr Temperature: 60oC. 2.3.).2. reaction time and temperature on the X n exp of the products is shown in Figure 2.2.) as a function of the dioxanone : 2 ratio at 90% and 100% monomer conversion is also provided. in line with the theoretical predictions. In Figure 2. A possible strategy to increase the Rgp values i.7.
Temperature: 100oC.3.5. time: 1. the X n theo (eq..7. and 2. time: 16 hr
Temperature: 60oC. as expected for a typical ring opening polymerisation .4 and 3.2. Most experimental points are scattered along these lines. Effects of process conditions on the average chain length ( X n exp ) and end group distribution The effects of the p‐dioxanone: 2 ratio.6. A clear trend between R gp and the process conditions (temperature. 2.) for all experiments are scattered randomly between 2. The X n exp increases linearly with respect to the dioxanone : 2 ratio. to reduce the number of OiPr endgroups in the product may be the removal of isopropyl alcohol formed in step
.3. and dioxanone/ 2 mol ratio) is absent.e. The R gp values (Table 2.7. 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.Chapter 2
2.5 hr Xn
40 Xn exp
30 40 50 dioxanone : 2 ratio (mol/mol)
Figure 2. time.
3.703 F 173. The low P‐value clearly indicates that the model is statistically significant. 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. time and dioxanone/ 2 ratio) on the X n exp has been performed by multivariable linear regression  on the data given in Table 2. 2.854 × (dioxanone : 2 ratio) + 0.459*10‐7
.3‐6.177 × (t ) − 0.2.543
Model Error Total
P‐value 3.8.) from the reaction mixture by e. Table 2. The analysis of variance for the model is given in Table 2. The R2 value for the model is 0.g.7.4.3.)
where t and T are respectively the polymerization time and temperature.4.3. indicating that the model describes the experimental data reasonably well.49 + 0. A linear model proved adequate to describe the effects of the independent variables on the X n exp :
= 8. vacuum distillation before adding the dioxanone monomer .977 (with an adjusted R2 value of 0. 2.974).097 × (T )
(2. The 2/ Al ratio was not included in the model as the experimental range (6.). A parity plot of the modeled versus experimental values of the average chain length Xn confirms this statement (Figure 2. Ring Opening Polymerization of p‐dioxanone using an Al(OiPr)3‐Monosaccharide Initiator System
1 of the polymerisation process (eq.6) was too limited to draw sound conclusions.12. Statistical Data Analysis Quantification of the influence of the experimental factors (temperature.
The ROP of p‐dioxanone in the presence of a monosaccharide (1. the model predicts that the X n exp is positively influenced by the polymerization time. The isolated yields of the off‐white solid products were between 30 and 96%.3.2. 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. In addition. Within the experimental ranges. the X n exp is negatively influenced by temperature. 16. Both NMR and MALDI‐TOF measurements indicate that the products are mixtures of polymers and significant
.8. the model allows determination of the process variables to obtain a product with the desired degree of polymerization. with higher ratios leading to higher molecular weights. Furthermore. with high ratios leading to a higher average chain length.4‐di‐O‐ isopropylidene‐α‐D‐galactopyranose. This is in line with literature data [5.4. 2) with Al(OiPr)3 as the catalyst is reported. 17]. 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). 11] and due to the fact that the reaction is an equilibrium polymerization with a slight exothermic effect.Chapter 2
40 Xn model
Chu.4‐dioxan‐2‐one and controlled synthesis of diblock copolymers with epsilon‐caprolactone. and mechanistic aspects of the ring‐opening polymerization of 1. Degee. Degee. . Damrau: Polylactones. Phys. Macromolecules 2001.R. 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. 27. Biomed.5.M. 199. 153‐166.
2.J. Raquez. 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. Kricheldorf. Rapid Commun. H.M.O. 21.6. P. P. 1089‐ 1097. 8419‐8425. P. Dubois: Some thermodynamic. A statistical model has been developed to quantify the effects of process variables (time. References
. starch). 1063‐1071. 1998. . 2000. Macromol. D. D. 1993. These studies will be reported in the next chapter. C. R. 42.4‐dioxan‐2‐one. Macromol. Moreover. Lin. Res. Narayan. R. temperature and monomer: monosaccharide ratio) on the X n exp . P.C. J.4‐dioxan‐2‐ one initiated by Al((OPr)‐Pr‐i)(3) in bulk. Narayan. 34. J. .g. Dubois: ʺCoordination‐insertionʺ ring‐ opening polymerization of 1. kinetic. Grubb: Hydrolytic degradation and morphologic study of poly‐p‐dioxanone.
. Mater. H. 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. Raquez. Zn‐lactate‐catalyzed polymerizations of 1. Chem.L.
. 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.
21 and 0. The silylation reaction was performed using hexamethyl disilazane (HMDS) as the reagent in DMSO at 70°C. The grafting efficiency varies between 28 and 58%. polycaprolactone. ε‐ Caprolactone is grafted to silylated starch by a ring‐opening polymerisation catalysed by Al(OiPr)3 in THF at 50oC. biodegradable polymers.7 was obtained. 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. Silylated starch with a degree of substitution (DS) between 0. silylation
. The poly‐(ε)‐caprolactone side chains consist of 40‐55 monomer units and is a function of the reagent intakes. Keywords: starch. the remainder being homopolymers of ε‐caprolactone. grafting.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. The DS of the polycaprolactone graft is between 0.72.45‐0. Removal of the silyl groups of the poly‐(ε)‐ caprolactone grafted starch co‐polymers is possible using a mild acid treatment with diluted hydrochloric acid in THF at room temperature.
The amylose‐amylopectin ratio in native as well as modified starches has a strong impact on the product properties.4‐glucosidic bonds) and amylopectin (a branched polymer with α‐D‐1. However. However. the product properties are in most cases still not up to standards and blending with other polymers is required .4‐ glucosidic bonds). and polybutyrolactone). Global production of starch is 60 million ton per year in 2004 . is one of the potential candidates for future biodegradable polymer products. Polycaprolactone (PCL) is easily degraded by micro‐organisms .g. Aerobic soil‐burial experiments showed that the mechanical properties of PCL films decreased rapidly in time and were fully degraded after 4 weeks . 7]. the used monomers and the corresponding grafted chains are not easily biodegradable. roots) as granules or cells with typical particle sizes between 1‐100 µm. Starch. A wide variety of synthetic biodegradable polymers have been prepared. a natural biopolymer. The polymeric structure of starch consists of repeating anhydroglucose units. it is water sensitive. The development of green biodegradable polymers for e. There are two types of biopolymer in starch. However. Starch has also been thermoplasticized with the help of plasticizers such as glycerol and other polyalcohols . and this value increases with about 10% per year . Various modification strategies have been explored. The content of amylose in starches depends on the plant and typically varies between 18‐28%. 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. amylose (a linear polymer of anhydroglucoses with α‐D‐1. Chapter 3
3. Starch films are known to have good oxygen barrier properties.1. and the mechanical properties of starch‐based films are generally inferior to those derived from synthetic polymers [4. the future generation of packaging materials is highly desirable. Starch modification is therefore needed to meet the product properties in a number of application areas. which are known to degrade with difficulty and cause serious environmental problems . 245 million tons of plastics are produced per year. Starch is abundantly available.6‐glucosidic bonds besides α‐D‐1. 5]. for instance grafting of monomers (like styrene and methyl methacrylate) to the starch backbone [6. resulting
. Blends of thermoplastic starch and PCL are not fully miscible. polyvalerolactone. Starch is present in the body of many plants (tubers. as starch is highly hydrophilic. Well known examples are polyesters derived from cyclic lactones (polycaprolactone. Introduction
Worldwide. These plastics are mainly synthetic polymers from fossil resources. in almost all cases.
In this way. All available data indicate that the presence of water reduces the GE. Aluminium triisopropoxide [Al(OiPr)3] (Acros) was used without further purification. ε‐Caprolactone monomer (Fluka) was dried over Calcium Hydride (CaH2) for 48 h. Common Ring Opening Polymerization (ROP) catalysts such as tin octoate or aluminium isopropoxide gave low grafting efficiencies (GE. This is rationalised by assuming that water competes with the hydroxyl groups of starch in the initiation step of the polymerization reaction. Hexamethyldisilazane (HMDS.2. This is expected to lead to reduced reaction rates between starch and CL compared to CL homopolymerisation. Starch is insoluble in the typical organic solvents used for ROP (such as toluene or THF). Dry tetrahydrofuran (THF) and toluene for polymerization experiments were obtained in closed vessels from Aldrich and were used as received. the ring opening polymerisation occurs solely in the liquid phase and this is expected to lead to higher GE values. distilled under reduced pressure at 100 oC and stored under a protective nitrogen atmosphere.67 gram (8 mmol) Al(OiPr)3 in 50 ml of dry toluene in a glove box.2. a very flammable by‐product. leading to a liquid‐solid system. Materials Corn starch (Sigma) was dried at high vacuum (~1 mbar) at 100 oC for one day before use. Acros) and methanol (Labscan) were used as received. To increase the miscibility of starch and polycaprolactone. The highest GE (90%) was achieved using triethylaluminium as catalyst . DMSO (Acros) and toluene (Labscan) were dried overnight over molecular sieves 3 Å (Merck) and stored under a protective nitrogen atmosphere. A stock solution was prepared by dissolving 1.
. Synthesis of Poly‐(ε)‐caprolactone Grafted Starch Co‐polymers using Silylated Starch Precursors
in undesirable phase separation . an alternative method to synthesize poly‐(ε)‐caprolactone grafted starch co‐polymers (starch‐g‐PCL) is reported. This approach has also been applied successfully to graft PCL and polylactide on dextran [19‐20]. therefore difficult to handle and releases ethane. Hydrochloric acid (HCl) 1 N was prepared from Titrisol concentrated hydrochloric acid solution (Merck) and distilled (Milipore) water. In this chapter. during the reaction. 0‐14%). thus leading to the formation of PCL homopolymers rather that starch‐g‐PCL . Another possible cause for the low grafting efficiencies is the heterogenous nature of the reaction.1. 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]. Materials and Methods
3. thus to a reduction in the GE. it has been proposed to chemically graft caprolactone onto the hydroxyl groups of starch using ring‐opening polymerisation .
3. This catalyst is extremely air and water sensitive.
1. Silylated Starch (1. 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. THF (4. 50 oC): δ 0. Sample SN‐3. Typical example of the starch silylation procedure The procedure for corn starch silylation was adapted from that published for dextran [19‐23]. another portion of toluene (20 ml) was added. the silylated starch (1) product was dried in a vacuum oven (~5 mbar. silyl‐CH3). 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.
3. 80 oC). 50 oC): δ 0. A mol ratio of ‐OH groups to catalyst of 10:1 was used. This procedure was repeated three times. The reaction was carried out at 70 oC. 40 oC) until constant weight.2.
Silylated Starch (1. After removal of the solvents under reduced pressure (0. 3‐6 ppm (m.12 (m.6 mL) at 50 oC (1‐2 h).2. The white‐to‐transparent solid products were stored under vacuum in a desiccator at 6‐8 oC. Subsequently.2.60):
H‐NMR (CDCl3. broad peaks. a predetermined amount of ε‐caprolactone monomer (molar ratio of monomer to
. 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. After 24 h. broad peaks.111 mol) was added to the gelatinized mixture to initiate the silylation reaction.5 (m. Traces of DMSO trapped in the product were removed by dissolving the product in a small amount of toluene and re‐precipitation in methanol. 3‐6 ppm (m. 1. after peracetylation. silyl‐CH3). starch). before peracetylation. DS = 0. After 6 h. To this homogenous solution.2. toluene (40 ml) was added to solubilize the precipitated (partially silylated) starch. starch).2.60): H‐NMR (CDCl3. the solvents were removed from the silylated starch product under reduced pressure (~ 20 mbar) at 70 oC. acetate‐CH3). After 2 and 4 h reaction time. Chapter 3
3.1 mbar. Sample SN‐3. For each experiment.2. The pre‐determined amount of HMDS (typically 24 ml.5 ml) and a predetermined amount of the stock solution of Al(OiPr)3 in toluene were added. DS = 0. Typical example of in situ polymerization of ε‐CL with silylated starch The silylated product from the first step was dissolved in THF (0. 3.12 (m. 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. 0. The samples were characterized by 1H‐NMR.7‐2.
98 (t. the peracetylating reagents
3.68. Starch‐g‐PCL (3. 60 oC): δ 0. 3.37 (t.88 ppm (m.5. and 5‐5. and vacuum dried at 40 oC for 24 h. iPr). γ‐PCL). Deprotection product of Sample SN1CL2.3‐4. 4. The samples were characterized by 1H‐NMR.88 ppm (m. the desilylated starch‐g‐polycaprolactone product (3) was precipitated using heptane. filtrated. broad peaks.1 g) was suspended in THF (4%‐w/v) and stirred at 55 oC until it was fully dissolved (typically 3 h).4 (m.54 (m. The silylated starch‐g‐PCL (2) product was precipitated from the solution by the addition of heptane (about 250 ml) at ‐18 o C. iPr). DSsilylation = 0. 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. 2.37 (t. α‐PCL).2.31 (m.4 (m. 3. After stirring for 2 h. The solid precipitate was filtered and dried under vacuum (~ 5 mbar) at 40 oC for 48 h. 4.2.2. 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. 3. The peracetylation procedure applied in this study was adapted from the literature [24. Subsequently. ‐CH.) was >99 %.5‐3. iPr). silyl‐CH3). 1 (0.34): H‐NMR (DMSO. DSsilylation = 0.75. Silylated Starch‐g‐PCL (2.75. γ‐ PCL). 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. ε‐PCL). 1. ‐CH3. The yield was measured gravimetrically and is based on the weight of the product and the total weight of reactants charged to the reactor. 3. 3. 4.21): H‐NMR (DMSO d‐6. β and δ‐PCL).12 (s. DSPCL=0. DSPCL=0. β and δ‐ PCL).68. ‐CH3 . 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). 25]. starch).3‐4. 1. The product was collected as a white solid and characterised by 1H‐NMR.3. ε’‐PCL). The total isolated yield at this condition (see Table 3.5. ε’‐PCL). iPr). 2.54 (m. α‐PCL). 3.16 (d. Sample SN1CL1. starch). 60 oC): δ 1. 4. ‐CH.25 (t. Typically.25 (t. 1.31 (m. ε‐PCL). and 5‐5.2. 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.16 (d. Peracetylation of the remaining OH groups of modified starch is a well established procedure to improve characterisation of the products by NMR .5‐ 3. broad peaks.2. 1. 1.
).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. 3. The assumption of high conversions (>95%) was correct for all experiments (see Table 3. 3. Analytical Methods 3. It was finally dried overnight in a vacuum oven at 70 oC and 5 mbar until constant weight.1.8− 4.)
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.3.1. Chapter 3
where Ax‐y stands for the 1H‐NMR peak area in the range δ x‐y ppm. The product was precipitated by the addition of methanol and washed several times with methanol.
3. the DSPCL may be calculated using eq. 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.3−3.3. DSsilylation may be calculated using 1H‐NMR spectra of the products after peracetylation using eq.2 ppm A3.4 ppm
(3.2.2. 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 3.2.1. This leads to the following equation:
ACH 2 −ε + ACH 2 −ε ' ACH 2 −ε '
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. 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.3.2.6 ppm DSsilylation = × = × 3 Astarch proton / 7 27 A3−5.4.8 ppm
(3. In this calculation. it is assumed that the average length of the grafted chain is equal to the chain length of the homopolymer. The peracetylation reaction was conducted for 7 h at 50 oC.2.6−0. When assuming that all ε‐CL monomer is converted.
8− 4 .2 ) ⎤ 1 × ⎢ (3 − DSsilylation )⎥ ⎢ ε . This leads to the following equation:
ACH 2 −ε ' / 2 − ACH −iso ACH 2 −ε ' / 2
× 100 %
(3. The values of ∆Hv and Vm were obtained from the SciFinder Scholar database (American Chemical Society.3−3.3−3.
.3. T stands for absolute temperature.3−3.4. Synthesis of Poly‐(ε)‐caprolactone Grafted Starch Co‐polymers using Silylated Starch Precursors
DS PCL =
3.3.8− 4.9 ⎟ × 100 % ⎜ ⎜ ⎟ ACH 2 −ε ' / 2 ⎠ A3. 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). 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. 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.)
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).9
⎡ 0.4 − A4 .4 + A3.5. A3. the focus of this chapter will be on the first two steps of the procedure. and Vm stands for molar volume.8− 4 .CL/OH ratio ⎥ ⎣ ⎦
(3.4 / 2 ⎟ ⎠ ⎝ ⎝
where Ax‐y stands for the peak area in the range δ x‐y ppm. Although all steps have been investigated.5.( A3.)
⎛ ⎞ ⎞ ⎛ A A = ⎜1 − CH −iso ⎟ × 100 % = ⎜1 − 4.
5. The silylation procedure was adapted from that previously reported for dextran [19‐23].1. It is defined as the average number of silyl substituted OH groups on the anhydroglucose (AHG) unit of starch.).a.5. 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. An additional peracetylation procedure to substitute the free hydroxyl groups with acetate groups was performed to improve the quality of the NMR spectra. 70 oC.) shows the presence of silyl groups at about δ 0 ppm.a.
. 3. mixtures of toluene and DMSO were applied to avoid precipitation of the silylated starch during the reaction. 24 h).60 was obtained.
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. The silylated products were characterised by NMR. C
O H 2OR
[R = H or Si(CH3)3] Starch Hexamethyldisilazane Silylated Starch (1)
(3. The degree of substitution of the silyl groups (DSsilylation) was determined by 1H‐ NMR.b.1. The 1H‐ NMR peaks from the AHG unit after peracetylation were indeed considerably sharper and allowed more accurate DS calculations (Figure 3. The 1H‐ NMR spectrum of silylated starch (Figure 3. as suggested by Einfeldt et al .3. a product with a DS of 0.). Using standard conditions (HMDS: AHG molar ratio of 3. the starch peaks are very broad and this feature hampers accurate determination of the DSsilylation. Chapter 3
3. a homogeneous reaction mixture was maintained throughout the reaction. However.).1. In this way.1. Synthesis of Silylated Starch The silylation of corn starch was performed with HMDS as the silylating agent (eq.)
Instead of using DMSO as solvent.
not peracetylated (b).
. in CDCl3. peracetylated The effect of the HMDS to starch molar ratio (1. and Figure 3.1. Typical 1H‐NMR Spectra of Silylated Starch (Sample SN3.1.5‐4. Synthesis of Poly‐(ε)‐caprolactone Grafted Starch Co‐polymers using Silylated Starch Precursors
Figure 3. 50 oC): (a).5) on the product DS was investigated by performing experiments with a constant starch and a variable HMDS intake.2.60. DSsilylation=0. The results are given in Table 3.
At a mol ratio of HMDS to AHG of 4.5 HMDS : AHG Ratio [mol/ mol]
0.60 0. HMDS is a rather apolar compound (Hildebrand solubility parameter of 6. the volumetric HMDS intake is about half of the DMSO intake.4
1.25 cal1/2cm‐3/2) due to the presence of the bulky apolar SiMe3 groups. For reactions with an order higher than zero. Chapter 3
0.5 Product DS (DSsilylation) 0. 70 oC) Table 3. The silylation reaction
. Effect of HMDS: AHG mol ratio on DS of the silylation product a
HMDS intake (ml. a positive effect of higher reagent intake on the reaction rate and thus the DS is expected. Its presence will reduce the polarity of the reaction medium (DMSO.1.55
0. a fixed starch intake of 6 g (37 mmol AHG) starch was applied.25 3 4.46
Experiment SN1 SN2 SN3 SN4
Experiments were performed in DMSO at 70°C. For all experiments.
Surprisingly.2.5:1. solubility parameter = 11.65
0.5 3 3.6 DSsilylation
0. the DSsilylation decreases for higher intakes of HMDS.5
Figure 3. The experimentally observed lowering at higher HMDS intakes is likely due to a decrease in the polarity of the reaction mixture. mmol) 12 (56) 18 (83) 24 (111) 36 (167) HMDS: AHG mol ratio 1.67 0.5 2.68 0.36 cal1/2cm‐3/2) considerably. DS of the silylated products at different HMDS: AHG ratios (DMSO.
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. white solid products with isolated yield > 96% were obtained.6. DMF. sample SN‐1) in THF at 50 oC for 24 h using Al(OiPr)3 as catalyst. A schematic representation of the reaction is provided in eq.3‐2. 29]. The use of HMDS to silylate dextran (Mw=6000‐40000) in DMSO (HMDS to OH molar ratio of 0. Silylation of starch with HMDS in formamide.7‐3.3. The products are soluble in DMSO as well as in less‐polar solvent such as chloroform and THF. the silylation of starch in pure DMSO using HMDS as the reagent has not been reported to date.68. 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‐1.0 [18. Synthesis of Poly‐(ε)‐caprolactone Grafted Starch Co‐polymers using Silylated Starch Precursors
involves charged‐ionic species  and a reduction in the polarity of the reaction medium is expected to lead to a lowering of the silylation reaction rates.25‐5. pyridine. 3. and DMSO/pyridine mixtures yielded silylated starch with DS values ranging between 0.0) gave silylated dextran with DS values between 1.0 . 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. To the best of our knowledge.7) of the products is in the range of those published for other starch silylation systems.2.45‐0.)
. Petzold et al  reported that silylation of starch with trimethylsilylchloride (TMSCl) in pyridine yielded trimethyl‐silyl substituted starch with DS values between 0. After precipitation with heptane and vacuum drying. DMA/LiCl.
O O H2C O R H H OR H O O H OR H O Catalyst: Al(OiPr)3 Toluene.6‐glucosidic for dextran). 3.2. 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 DSsilylation (0. liquid ammonia.6.1 and 3.
Resonances from the starch peaks are observable as small. The homopolymers are formed by direct polymerization of ε‐CL initiated on isoproproxide moieties attached to the Al catalyst. Chapter 3
The products were characterized using 1H‐NMR analysis in DMSO‐d‐6 as the solvent. DSPCL=0.3. This is clearly indicated by resonances of the iPr end‐group of the PCL homopolymer at δ 4. broad peaks in the region δ 3.
Figure 3. Coding of the peaks is given in Figure 3.9 and 1.4. (Sample SN1CL1.3.0‐5.
.68. DSsilylation =0. The formation of CL homopolymers for this type of reactions has been observed before [15. However. Typical 1H‐NMR spectrum of a silylated starch‐g‐PCL sample. 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.4.21) in DMSO‐d6 at 60 oC.8 and 5.4‐3.2 ppm. the exchange reaction between Al(OiPr)3 and the OH groups of silylated starch is not quantitative under the conditions applied in this study. The peaks from the polycaprolactone units are clearly present in the range of δ 1. as is shown in Figure 3. A typical spectrum is shown in Figure 3. Apparently. not all of the caprolactone is grafted to starch and large amounts of PCL homopolymers (72%) were formed. 26].
peaks arising from a ‐COOH unit could not be detected in 13C‐ NMR spectra (δ 175‐180 ppm). 3. Exchange Reaction: iPr O Al O iPr iPr O + H2O iPr O iPr O Al OH + iPr OH
2. residual water may also initiate the polymerisation reaction. 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 3.4. This leads to the formation of carboxylic end groups (see eq.
1.7.). Product formation for the reaction between silylated starch and ε‐ caprolactone Besides the ‐OH group of starch.7.)
. Polymerization Reaction: O iPr O Al OH iPr O PCL homopolymer with carboxylic end-group + O O α H β χ δ ε O α' β' χ' δ' ε' [O C CH2 CH2 CH2 CH2 CH2 ]n O C CH2 CH2 CH2 CH2 CH2 OH
(3. Synthesis of Poly‐(ε)‐caprolactone Grafted Starch Co‐polymers using Silylated Starch Precursors
1. However. Exchange Reaction:
iPr O Al O iPr H2C OH
iPr O Al O iPr iPr O
H2C O O O H O H OR H H OR [R = H or Si(CH3)3 ] H O
H OH H
+ iPr OH
Silylated Starch (1) 2.
2. and Figure 3. units] 40 43 44 49 54 DSPCL 0.3 ppm) should be twice the intensity of the ‐CH‐ proton of the isopropoxide end group (δ 4. Table 3.5 and 100%.5. in all samples.
Experiment SN1CL1 SN1CL2 SN1CL3 SN1CL4 SN1CL5
a.0 18.47 0. Chain Length [mon. This implies that grafting of caprolactone to starch also occurs to a significant extent.2 Total Yieldb [%] >99 >99 >99 99 96.5 29.0 15. However.58 0.5 Avg. The yield of the products was measured gravimetrically and varies between 96. The ACL of the polymer and the DSPCL increase almost linearly with the ε‐ caprolactone intake (Figure 3.9 22. The grafting efficiency (GE) for the sample given in Figure 3.72 Grafting Efficiency [%] 28 43 48 55 58
All reactions were performed using the same intake of SN1 silylated starch (DS=0.3.34 0.68) in THF at 50 oC with Al(OiPr)3 as the catalyst (1 mol Al(OiPr)3 per 10 mol‐starch‐OH groups).2.9 ppm). In the case of only homopolymerisation. (SN1CL1) is 28%. Determined gravimetrically and defined as the total weight of the isolated product divided by the total intake of reactants (silylated starch and CL).
Overview of results for the grafting of ε‐CL on silylated starch a
ε‐CL/ OH [mol/ mol] 13. Five experiments were performed to study the effect of different ε‐CL to silylated starch ratio.
.21 0. the intensity of the resonance ε’ was considerably higher that that of the ‐CH‐ iPr peak. the intensity of the peak from the ‐CH2‐ end group of the homopolymer (ε’ at δ 3. Chapter 3
The ratio of homopolymerisation versus grafting on starch may be obtained by comparing the integrals of selected peaks in 1H‐NMR spectra.). and reaches 58% for a ε‐CL to starch–OH ratio of 29. 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.2.5. The results are given in Table 3. The GE increases with higher ε‐CL intakes. This implies that the ε‐CL conversion is essentially quantitative in all cases.
Synthesis of Poly‐(ε)‐caprolactone Grafted Starch Co‐polymers using Silylated Starch Precursors
Average Polymer Length [mon. units]
50 0.7 45 0.6 0.5 40 0.4 35 0.3 0.2 30 12 14 16 18 20 22 24 26 28 30 DSPCL
ε-CL : Starch-OH Ratio [mol/ mol]
Figure 3.5. Average Chain Length (ACL) and the DS for as a function of the CL‐ starch intake. ∆ : Average Chain Length (ACL); □ : DSPCL The mechanism of ROP of cyclic esters such as caprolactone in the presence of an alcohol (ROH, silylated starch in our case) is provided in Scheme 3.1. [30‐31]. Higher monomer intakes are expected to lead to higher polymerization rates as shown in Scheme 3.1.b. This will result in longer PCL chains in the final product, in line with the experimental observations (see Table 3.2. and Figure 3.5.). At higher ε‐caprolactone intakes not only the ACL of the grafted chain increases but also higher values for the DSPCL are observed. This finding may be rationalised by assuming that the rate of chain transfer (Scheme 3.1.c.) with starch is increased at higher ε‐caprolactone intakes. The rate of this reaction is expected to be a function of both the starch and the concentrations of Al‐species with a growing PCL chain. The starch intake for all experiments was equal, meaning that the concentration of Al‐species with a growing PCL chain should be higher at higher ε‐CL intakes. This is indeed predicted by the mechanism given in Scheme 3.1.b.; higher caprolactone intakes will increase the rate of this reaction and lead to higher concentration of Al‐species with a growing PCL chain.
Al O iPr
R OH O O kp kd
Al O R
O Al O [(CH2)5 C O]n R
Al O R
O Al O [(CH2)5 C O]n R + R OH
O Al O R
+ HO [(CH2)5 C O]n R
Scheme 3.1. The observation that a higher monomer to alcohol ratio leads to higher amounts of PCL chains with an alcohol end group and thus a higher DS was also reported for the polymerization of p‐dioxanone with Zn‐lactate as catalyst and α‐tocopherol as the alcohol . To show the potential of our approach to use hydrophobised starch instead of native starch for the ring opening polymerisation of cyclic esters, several ring openings polymerisations of native starch with ε‐CL monomer were performed either in pure ε‐CL or in a mixture of ε‐CL and toluene (80‐100oC, 24 hr). At the start of the reaction, the starch was always insoluble in the reaction medium. After reaction the product was isolated, washed thoroughly with toluene and dried. The weight of the product, however, was very close to the initial starch intake. Examination of the products by FT‐IR does not show the presence of caprolactone vibrations. Thus, it may be concluded that solubilisation of starch is of key importance to obtain poly‐(ε)‐caprolactone grafted starch co‐polymers. These findings are in line with earlier studies on the ROP of ε‐caprolactone on native granular starch using Al(OiPr)3 as catalyst . Here, caprolactone polymerization did not occur after 18‐24 h reaction time and only liquid ε‐CL was recovered. Only when performing the reaction with high concentrations of the aluminium catalyst, a product with a GE of about 13% was obtained. This low GE was explained by assuming that the reaction between starch and Al(OiPr)3 is slow and due to the heterogeneous nature of the reaction mixture. Our study, together with the result of Dubois et al  indicate that homogenous reaction conditions are required for the successful ROP of ε‐CL to obtain poly‐(ε)‐caprolactone grafted starch co‐polymers when using Al(OiPr)3 as the catalyst. When performing the reaction under heterogenous conditions, a high grafting efficiency is only achievable when using triethylaluminium as the catalyst .
Synthesis of Poly‐(ε)‐caprolactone Grafted Starch Co‐polymers using Silylated Starch Precursors
3.3.3. Deprotection of Silylated‐Starch‐g‐PCL A preliminary experiment was performed to remove the silylated groups of the silylated starch‐g‐PCL (eq. 3.8.) using a mild acid treatment with diluted hydrochloric acid in THF at room temperature. All silyl groups were removed successfully, as is clearly seen from an NMR spectrum given in Figure 3.6.
Figure 3.6. H‐ NMR Spectrum of Starch‐g‐PCL (reaction product after desilylation of sample SN1CL2, DSsilylation =0.68, DSPCL=0.34), in DMSO‐d6
O O H2C H H OR H C (CH2)5 O
O H O CH2 H H OH H C (CH2)5 O
O H OR
THF, 2 hr H
O H OH
[R = H or Si(CH3)3 ]
Silylated Starch-g-Polycaprolactone (2)
1.21‐0.986 cal mol‐1 K‐1 : temperature [K]
molar volume [cm3/ mol]
.72. Chapter 3
3. 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. indicating that homogeneous reaction conditions are favorable for the grafting reaction. the highest value was obtained with a ε‐CL‐AHG ratio of 29.4.68) was obtained using a DMSO/toluene mixture as the solvent and HMDS as the silylating agent. the percentage of PCL grafted to starch compared to the total amount (grafted and homopolymer) of polymerized CL [%] : gas constant.46‐0. Nomenclature
: peak area of certain proton in 1H‐NMR spectra [‐] : peak area of certain peak at x until y ppm in 1H‐NMR spectra [‐] : average number of CL repeating units in a grafted polymer chain or homopolymer [monomer units] : Degree of Substitution. The products may have interesting applications as compatibilizers for starch‐polymer blends. depending on the ε‐CL‐starch ratio.2. Conclusions
The successful synthesis of poly‐(ε)‐caprolactone grafted corn starch co‐ polymers using a three step approach is reported.5. 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. Considerable amounts of ε‐CL hompolymers with isopropyl end‐groups were also formed. This was achieved by making the starch more hydrophobic by partial substitution of the OH groups by trimethylsilyl groups. 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%. 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. Silylated starch with a low‐medium DS (0.
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.
1998. 3. Degrad. 1977. R. .C. 1998. . 51. R. C.Sci. . 2003. 56. 17. 305‐313. . R. J. A. Denmark. Linke: Water vapor sorption determination of starch based porous packaging materials. 1273‐1335. Department of Dairy and Food Science. J. Biobased packaging materials for the food industry. . Goldberg: Review of the biodegradability and utility of poly(caprolactone).
. Fanta. 2000. K. Wang. E. 107‐113. W. Denmark: http://www. De Mulder‐Johnston: Edible and biodegradable polymer films: Challenges and opportunities. 2008.K. The compelling facts about plastics: An analysis of plastics production. Bagley. Synthesis of Poly‐(ε)‐caprolactone Grafted Starch Co‐polymers using Silylated Starch Precursors
: heat of vaporization [kJ/mol] :
Hildebrand solubility parameter [cal1/2cm‐3/2]
3. . Food Technol. Polym. Albertsson: Biodegradable polymers and environmental interaction. 61‐74. Polym. A. Polym. Yavuz. Sci. References
. 385–409 . C. F. Sci. 61‐67. Eng. Abdel‐Mohdy: Grafting of poly(methacrylic acid) on starch and poly(vinyl alcohol). Frederiksberg C.A. status and perspectives.M.A. Chandra. H. D.html. Y. Doane. Solaro: Biodegradable polymeric materials. 1995. the Royal Veterinary and Agricultural University. . S. accessed on August 01. Environmen. 2003. 407‐412. R..C.starch. Website of International Starch Institute. 11.R. Prog. G. Beliakova. J. 1251‐1253. J. L. M. . 1997. Part C: Polym. Rev. Adv. 1996. . C. demand and recovery for 2006 in Europe. Krochta..6. report of Food Biopack Project (Ed. Rustgi: Biodegradable polymers. 311‐316. Aarhus.dk/isi/stat/rawmaterial. Polym. Yang. A. C. Chiellini. Aly. Babac: Preparation and biodegradation starch/polycaprolactone films. Polym.F. Association of Plastics Manufacturers. X. Eng. Starch‐Starke 2004. 38. E. . Brussels.B. 43. Weber). 52. Sci. 2008. Starch‐Starke 2000. Macromol. Belgium.M. Environ. 23. Plastics Europe.A. Russell: Graft copolymers of polysaccharides with thermoplastic polymers ‐ new type of filled plastic. Burr. J. of
. Tsiapouris. Karlsson. Mater. Wang: Properties of starch blends with biodegradable polymers. 8. 53‐57.
415‐426. J. . P. 2004. 73. Dellacherie. W. Einfeldt. Pantiru. D. Krishnan. Biomacromolecules 2001. E. . Monomers Polym. Rutot. Dubois. . D.. selective syntheses. Macromolecules 2000. Nouvel. Biosci. . Dubois: Controlled synthesis of poly(epsilon‐caprolactone)‐grafted dextran copolymers as potential environmentally friendly surfactants. T. E. and chemical stability of silylated dextrans. 1735‐ 1743. Degee. Dole. C. 3091‐3100. P. L. . 341‐347. structure analysis. J. . H. . Biomacromolecules 2003. K. A. Moro. A. Koschella. I. Petzold. Averous. 41. I. 1. Stein. Klemm. Polymer 2000. Degrad. W. Six. Fringant: Properties of thermoplastics blends: starch‐polycaprolactone. Gunther. L. high value applications. . Dellacherie. Einfeldt. K. Macromol. Klemm. L. Stab.L. Fessi. Part A: Polym. 40. Ydens. D. Degee. D. P. Nouvel. Ydens. and subsequent reactions. Polymer 2002.L. Boullanger: Ring‐opening polymerisation of epsilon‐caprolactone with monosaccharides as transfer
. C. . 2002. 561‐566. D. P. Stein. Petzold. 163. M. I. Klemm. Klemm: Preparative and 1H NMR investigation on regioselective silylation of starch dissolved in Dimethyl Sulfoxide. Ydens. Polymer 1999. 251‐269. 2577‐2588. Six: Controlled synthesis of amphiphilic biodegradable polylactide‐grafted dextran copolymers. J. Cellulose 2003. Gunther: Synthesis and NMR characterization of regiocontrolled starch alkyl ethers. Dubois. 5.L. 2001. Gunther. 10. Symp. Sci. D. Degee. regioselectivity.L. L. Des. W. K. P. L. Chapter 3
. . E. 43. P. 6713‐ 6721. M. Nouvel. Stein. Six: Partial or total silylation of dextran with hexamethyldisilazane. J. Dubois. Petzold. J. D. 35‐47. Petzold. E. Klemm: Regioselective functionalization of starch: Synthesis and 1H NMR characterization of 6‐O‐ silyl ethers. 2001. 4. . Dellacherie. Six: Silylation reaction of dextran: Effect of experimental conditions on silylation yield. Dubois: Aliphatic polyester‐ based biodegradable materials: new amphiphilic graft copolymers. P. B. Polym. E. C. 4157‐4167. 33. 965‐969. M. Heublein: Silylation of cellulose and starch ‐ selectivity. Dellacherie. Macromol. . Chem. Duquesne. R. Hamaide. Einfeldt: Structure design of polysaccharides: novel concepts. P. 2001. P. P. Dubois. Kussler. 42. Rutot. Narayan: Aliphatic polyester‐grafted starch‐like polysaccharides by ring‐opening polymerization. A. A. D. 2. 1443‐1450. K. Polym. P.
22. 199. Kricheldorf. .J. A. . 941‐950. 61. F. H. Duda: Living polymerization with reversible chain transfer and reversible deactivation: the case of cyclic esters. Junistia. Manurung. Heeres: Synthesis of higher fatty acid starch esters using vinyl laurate and stearate as reactants. 1998.B. L. Macromolecul. L. Picchioni. Duda: Polymerization of ε‐caprolactone initiated by aluminum isopropoxide carried out in the presence of alcohols and diols. R. presented at ACS Rubber Division meeting. Mexico City.O. .
.K. 2000. accepted for publication in Starch‐Starke. A novel route to functionalised nanoparticles. S. Macromol. 553‐556. 1399‐1406.P. Biela. 1089‐1097. A. T. J. A. Kinetics and mechanism.4‐ dioxan‐2‐one. H. Rapid Commun. Carbohydr. Macromolecules 1996. . Synthesis of Poly‐(ε)‐caprolactone Grafted Starch Co‐polymers using Silylated Starch Precursors
agents. Penczek. 659‐663. 2008.R. Sugih.L. 1978. Damrau: Zn‐lactate catalyzed polymerizations of 1. 29. Horton. D. 2001. Rapid Commun.: Solubility parameter concepts – a new look. D.M. Lehmann: Selective 6‐O‐acetylation of amylose. Hertz Jr. Phys. Res. D. 1989 . 21. . Macromol Chem. Janssen.
an increase in the vinyl‐ester concentration leads to higher product DS values. 13C‐NMR and FT‐IR.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. For low vinyl‐ester to starch ratios. At higher ratios. and Na‐acetate) in DMSO at 110°C.24 ‐ 2. esterification. K2CO3. The fatty esters were prepared by reacting the starch with vinyl laurate or vinyl stearate in the presence of basic catalysts (Na2HPO4. The DS of the products is a function of the carbon number of the fatty acid chain. Keywords: corn starch. vinyl ester to starch ratio and the type of catalyst. vinyl stearate
. the DS decreases. and Na‐acetate are superior catalysts with respect to activity compared to Na2HPO4 and products with DS values close to 3 were obtained. When performing the reactions using Na2HPO4 as the catalyst. vinyl laurate. The yellowish products were characterized by 1 H‐. the DS for the starch‐laurate compounds is higher than for the corresponding starch‐ stearates. K2CO3.96). presumably due to a reduction of the polarity of the reaction medium.
An alternative method using methyl and glyceryl laurate esters in the absence of solvent has been recently developed . However. high brittleness. An attractive field of application for these polymers is the use as packaging materials. 5‐6] were introduced using fatty acid chloride reagents. 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. the use of native starch for packaging materials is limited due to its low moisture resistance.61) products were obtained using this approach. Further modification of starch is therefore required to introduce hydrophobicity and to improve mechanical and moisture barrier properties. To improve the mechanical properties. Relatively low‐DS (0. resulting in products with DS values up to 2. The major obstacle is the pronounced brittleness of the materials. defined as the moles of substituents per mole of anhydroglucose (AHG) units) . The introduction of acetate groups on starch makes the product more hydrophobic. 5]. Their
. more water‐resistant products may be obtained [3‐4]. In particular starch is attractive as it is relatively cheap and abundantly available. However. The mechanical properties and hydrophobicity of the products were significantly improved when using these longer chain fatty acid precursors [1.34‐0. poor processibility (high viscosity). 7]. However. The hydrophobicity increases with the degree of acetate substitution (DS. and even fatty acid derivatives (C12‐C18) have been used in the modification reaction [5. However. higher molecular weight carboxylic starch esters (C4‐C6) .Chapter 4
4. For the current petrochemical based products recycling is often neither practical nor economically feasible . Esterification of starch with low molecular weight fatty acid derivatives is one of the oldest modification technologies to improve starch properties. cellulose or proteins are potentially very interesting starting materials for biodegradable packaging materials. and particularly acetic acid derivatives (C2) [2‐4]. and consequently. Introduction
Green biodegradable polymers derived from natural resources are potentially very interesting substitutes for non‐biodegradable petroleum‐based polymers. most of the studies performed to date use short chain carboxylic acids (C1‐C4). which are relatively expensive and rather corrosive . The first paper on the acetylation of starch was already published in 1865 . even after the addition of plasticizers . the fatty ester substituents [1.7 [1. Recently Mormann et al  explored the possibility of using vinyl esters and particularly vinylacetate as reagents for the preparation of starch esters.1. and incompatibility with hydrophobic polymers. 5]. Natural polymers such as starch.
This solvent effect is likely caused by the higher solubility of the vinyl esters in DMSO compared to water. not been reported to date. Technical grade dimethyl sulfoxide (DMSO). Japan). the Netherlands). Materials Corn starch (approx. 4‐N. to the best of our knowledge.01 when using vinyl‐laurate. In DMSO.2. Germany) were used as received. The effects of the starch to vinylester ratio on the reaction rates and DS have been explored. The maximum attainable DS of starch acetate in water was below 1 and limited to 0. 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. leading to higher reaction rates. The products were placed directly on the diamond plate and 50 scans with a resolution of 4 cm‐1 were recorded. Analytical Equipment H‐ and 13C‐NMR spectra were recorded in CDCl3 on a 400 MHz Varian AMX NMR machine. 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. and tetrahydrofuran (THF) were supplied by Acros (Belgium) and were also used as received. ~1 mbar). 4.
4. Materials and Methods
4. The starch was dried before use for 48 h at 105 oC under vacuum (approx. Germany) were used without further purification. IR spectra were recorded on a Spectrum 2000 FT‐IR Spectrometer (Perkin Elmer). 73% amylopectine and 27% amylose) was purchased from Sigma (Germany). The synthesis of starch stearate esters using vinyl ester reagents has. sodium acetate (Merck.2.2. Analytical grade vinyl stearate (Aldrich. Germany) and acetic anhydride (Merck. as recommended by Laignel et al .6 for starch acetate ester) were obtained.2. Potassium carbonate (Boom. the use of basic catalysts other than Na2HPO4 has been investigated. The spectra were recorded at 50 oC. In addition.
. leading to a moisture content of 2 %‐wt (measured gravimetrically). starch esters with a substantially higher DS value (up to 1. vinyl laurate (Fluka.N‐dimethylaminopyridine (DMAP).1. The reactions were either performed in water or in DMSO using a basic catalyst (Na2HPO4). Germany) and disodium hydrogenphosphate (Merck. 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.
5S). 24. C2). Table 4. 2. 28H. the product was precipitated using methanol (100 ml) and separated from the liquid phase by decantation. Starch‐stearate (Sample 19. 1. 3H. After cooling. and O‐C6S). 935 (C‐O stretching). 3‐6 ppm (m. C1S‐6S). C6S).5 (m..5 (m. 1350 (C‐H bending). 3S. CDCl3): δ 14. C1S‐6S)
.1.1 (m. C4‐11). 2H.4 (broad. CDCl3): δ 0. 1. broad peaks. 720. broad peaks. typical spectra are given in Figure 4. 2H.1. 2. O‐C3S. 1295. 1. CDCl3): δ 0.3. 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. (13C‐NMR) and Figure 4. 2H. broad peaks. C1S‐6S). C12).0 (m.5 g) was first gelatinized in DMSO (5 ml) at 70 oC for 3 h.9 (t.3 (m. 2H.2. 34. C4‐C17).2. DS = 2. overlap with CDCl3 (C4S). 7H. the product was dried in a vacuum oven (70 oC. 760.9 (t.
C‐NMR (before peracetylation.1..1 (C11). 61. 1455 (CH2). 1. C1S).8‐2. C2).96) H‐NMR (before peracetylation. 172‐174 ppm (C=O.7 (C11). 2H. 2.1. 1110 (C‐O stretching).3. 28‐32 (C4‐9).9 (broad. broad peaks. broad peaks. 1.9 (C3). 1020 (C‐O stretching).52): H‐NMR (before peracetylation. C18). 1.9 (C10). resulting in the formation of a homogenous transparent solution. attached to O‐C2S. 76‐78 ppm. Table 4. C4‐11). 3H. 1. C2). 31.9 (t. (1H‐NMR). broad peaks. C2’). 3H. broad peaks. The atom numbering scheme is given in Figure 4. 2H. 7H. 16H. C3).2. 3‐6 ppm (m. 68‐74 (broad. C2S.3 (m. Figure 4. (FT‐IR).0 (C12).3 (m. Methods 4. 3H. Subsequently. Starch‐laurate (Sample 17. respectively). 1740 (C=O).
H‐NMR (after peracetylation.6 (m. 3‐6 ppm (m.
FT‐IR (cm‐1): 2920 (C‐H stretching). 12).4 (m. 1410 (C‐H bending). 1230 (C‐O stretching). DS = 2. C3).5 (m. 2850 (C‐H stretching). Finally. CDCl3): δ 0. The samples were characterized by 1H‐ and 13C‐NMR and FT‐IR. C3). 1150 (C‐O stretching). 7H.. 22. 1370 (C‐H bending). The product was washed twice with methanol (50 and 25 ml. Typical example of the synthesis of laurate and stearate esters of corn starch Corn starch (0.4. 95.3. broad peaks. 16H. approximately 5 mbar) for 24 h until constant weight.Chapter 4
C1S). 25. 2850 (C‐H stretching). 2H. 28H. 95. 3H. 7H. 1740 (C=O). attached to O‐C2S. C18). 172‐174 ppm (C=O.2 (C2). 68‐74 (broad.5 (m. 950 (C‐O stretching). 865. Numbering scheme for carbon atoms of products 4. 1100 (C‐O stretching). 1410 (C‐H bending).3.5 (broad. A peracetylation reaction to substitute all of the remaining
.0 (C16). C2’). 3‐6 ppm (m. 760. O‐C3S. broad peaks. 75.
O R O
1S 2S 1
CH3 2' O
O R O
O R O
Figure 4. 1.0 (C18).2. Synthesis of Higher Fatty Acid Starch Esters using Vinyl Laurate and Stearate as Reactants
H‐NMR (after peracetylation. C4‐C17).0 (C3).4 (broad. 1370 (C‐H bending). 22.2. CDCl3): δ 14. 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.4 (m.7 (C17). 2. 61. 720. and O‐ C6S)
FT‐IR (cm‐1): 2920 (C‐H stretching). 34. 3S. C6S). C1S‐6S)
C‐NMR (before peracetylation.8‐2. C3). 1455 (CH2). 3H. 1295. C2). CDCl3): δ 0.1. 1020 (C‐O stretching).3 (m. 32. 1350 (C‐H bending). C2S.9 (t. 26‐32 (C4‐15).6 (m. 1. 5S). 1150 (C‐O stretching). 2H. broad peaks.7 (C4S). 1.
. 4.3. acetic anhydride and pyridine in a DMAP : acetic anhydride : pyridine : AHG molar ratio of 1: 10: 22 : 1) were added.2.6 − AC 2 3
(4. The peracetylation procedure by Einfeldt et al  was applied.5‐1.4. 4. The peracetylation reaction was conducted for 7 h at 50 oC.4..)
AH − starch =
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. Typically.Chapter 4
hydroxyl groups with acetate groups was applied to obtain reliable DS data.8− 2.). Subsequently. the peractylating reagents (DMAP.8 ppm (protons attached to the fatty acid carbons C3‐C12 for laurate and C3‐C18 for stearate) using eq.) (4.1.3. The DS of the fatty acid esters was calculated from the DS of the products after peracetylation (eq 4.)
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.).6 7
(4.6 ppm range) and a correction has to be made (eq 4. the starch ester (0.2. 4.
DS fatty ester = 3 − DS acetate
AH − acetate AH − starch
A3. The procedure was described earlier by Elomaa et al  and the relevant equations are shown in equation 4. The product was precipitated by the addition of methanol and washed several times with methanol.
AH − acetate = A1.5.1 g) was suspended in THF (4%‐w/v) and stirred at 55 oC until the starch was fully dissolved (typically 3 h).6−5.6‐5.6 ppm).2. It was finally dried overnight in a vacuum oven at 70oC and 5 mbar until constant weight.184.108.40.206.8‐2. and 4.)
Determination of the unit area of the acetate H‐atoms (AH‐acetate) is hampered by peak overlap with the H‐atoms attached to C2 (AC2) of the fatty acid chains (δ 1. Determination of the Degree of Substitution (DS) The DS of the product was calculated using 1H‐NMR spectra of the products after peracetylation. and 4.
After 24 h.‐4. Results and Discussion
OH O HO OH
R O CH2
O HO O R
(1) R = laurate (C12) (2) R = stearate (C18)
Scheme 4..1).7. Initially. light yellow solid. the brownish gel was precipitated with methanol and the product was collected after vacuum drying in the form of a transparent.1. The products of these exploratory reactions are insoluble in water and DMSO.6. the starch was gelatinised in DMSO at 70°C for 3 h to make the hydroxyl groups of starch more accessible for reaction.3.)
. 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.8 = (stearate) 33 33
(4.3. 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.)
In eq.8 = (laurate) 21 21 AC 3−C18 A0.7. Synthesis of Higher Fatty Acid Starch Esters using Vinyl Laurate and Stearate as Reactants
AH − fattyester = AH − fattyester =
AC 3−C12 A0.5−1.1.
4. the esterified starch started to separate from the medium in the form of a gel. The reaction was performed in two discrete steps. After 2‐3 h. the vinylester and the catalyst were added and the reaction mixture was heated to 110°C. Ax‐y stands for the peak area in the range δ x‐y ppm.5−1.3. Subsequently. but swell in organic solvents such as toluene and THF. 4. A schematic representation of the esterification reaction of starch with the vinyl esters is provided in Scheme 4.
2. 4.52 was obtained.).3. Table 4. Medium DS starch laurate and starch stearate (1 < DS < 2) dissolve poorly in DMSO‐d6 and CDCl3. When using a vinyl‐laurate : AHG molar ratio of 3 and K2CO3 as the catalyst.1. 4. 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.96.) in CDCl3 at 50oC.52 (Sample 17.) in CDCl3 at 50oC.1. and 4.1. DS = 2. Table 4.52 (Sample 17.2. Typical 1H‐NMR spectrum of: (a) native starch in DMSO d‐6 at 60oC. Higher DS products have a higher solubility in CDCl3 and good quality 1H‐ and 13C‐NMR spectra could be obtained (Figure 4. even at higher temperatures (50 oC). A reaction with vinylstearate at similar conditions resulted in a stearate starch ester with a DS of 2.
. Product Characterisation 4.1. a product DS of 2. (b) starch laurate.
Figure. (c) peracetylated starch laurate. DS = 2. For atom numbering scheme: see Figure 4.2.2.Chapter 4
The DS of the products was determined by using NMR (vide infra).3.3.
8‐2.2.52 (Sample 17) in CDCl3 at 50oC. and therefore a peracetylation procedure to substitute all of the remaining OH groups with acetate groups was applied [8. NMR spectra of the peracetylated products are considerably improved in terms of peak resolution and allow a more reliable calculation of the DS. required for DS determinations. 10‐11]. DS = 2. The proton signals of the acetate methyl group. DS = 2.3.c. Clearly visible are the peaks arising from starch and the aliphatic hydrogen atoms of the fatty acid chain (δ 0.2.5 ppm) are broad and overlapping .3. This feature hampers the DS determination by NMR. The resonances arising from the anhydroglucose unit of starch are broadened. (c) starch stearate.96 (Sample 19) in CDCl3 at 50oC. The same
.3 ppm.1. are together with the CH2 groups of the acid chain adjacent to the ester moiety in the range δ 1.8‐2. For atom numbering scheme: see Figure 4. Typical 13C‐NMR spectra of the products are given in Figure 4. in DMSO d‐6 at 60oC.
Figure 4. (b) starch laurate. The starch peaks (δ 3‐5. Typical 13C‐NMR spectra of: (a) native starch. The 1H‐NMR spectrum of a typical peracetylated starch laurate is shown in Figure 4. Synthesis of Higher Fatty Acid Starch Esters using Vinyl Laurate and Stearate as Reactants
A typical 1H‐NMR spectrum of starch laurate is shown in Figure 4.5 ppm). Clearly visible are the carbon resonances of the fatty ester chains (δ 10‐35 ppm) and C atom of the ester group (δ 170‐175 ppm). Two of the carbon resonances (1S and 4S) are considerably shifted compared to native starch.
FT‐IR measurements The FT‐IR spectra of starch laurate and starch stearate are shown in Figure 4. starch laurate C-H C=O
c.4.4.4. Table 4. starch stearate (DS= 2.) is also included. Table 4. and 4. For comparison. and 4.1.b. The near absence of remaining hydroxyl
.4. FT‐IR Spectra of starch laurate (DS= 2.1.Chapter 4
phenomenon was observed by Dicke for starch acetate .a.4. Sample 19.52. native starch -OH C-H
Figure 4. 4. In addition.4.).) show characteristic bands of the carbonyl group of the fatty esters in the 1750‐1700 cm‐1 region.4.2. starch stearate
b.2. Characteristic peaks of the polysaccharide backbone are visible in the 1250‐900 cm‐1 region (C‐O stretching) .c.3.96.) and native starch FT‐IR spectra of both starch laurate and starch stearate (Figure 4. the C‐H stretching vibrations of the alkyl groups of the fatty ester chain are clearly present at 2920 and 2850 cm‐1. a spectrum of native starch (Figure 4.b. 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. Sample 17.
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 1.3.91 0.54 2.90 1. type of catalyst and the effect of the addition of co‐solvents on the product DS was studied in more detail. The results are shown in 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.90 0.57 0.07 0. 4.52 2.2. Systematic Studies The effect of important process variables like the vinyl‐ester to AHG ratio.96 2.23 0.08 1.60 1.99 1.13 1.05 0. In addition.
.24 2. in line with the NMR data. Synthesis of Higher Fatty Acid Starch Esters using Vinyl Laurate and Stearate as Reactants
vibrations in the range 3000‐3600 cm‐1 and at 1640 cm‐1 indicates that the DS of the product is high. Most of the experiments (14) were performed using Na2HPO4 as the catalyst.1.44
Exp. Table 4. four experiments were performed with two alternative basic catalysts (K2CO3 and Na‐acetate).1.68 0.01 0.
23 for vinyl‐laurate at an intermediate vinyl‐ester‐AHG ratio of 3. A further increase in the vinyl ester intake leads to a reduction in the DS. DS of the product as a function of the type of vinyl ester and the vinyl ester‐starch ratio (24 h reaction time.
4. this positive effect dominates the reaction rate and the DS of the products will therefore increase at higher vinyl ester intakes. At a ratio of 1:6. 7‐10). This reduced polarity is expected to lead to ○ : starch‐laurate □ : starch stearate (lines for illustrative purposes only)
.). The results are presented in Figure 4. This behaviour is likely the result of two opposing effects.1.5. At low to medium vinyl ester/AHG ratios (0‐3). This is likely due to a reduction of the polarity of the reaction medium. 110 oC.5
2 3 4 Vinyl Ester: AHG Starch Ratio [mol/mol]
Figure 4. 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. The highest DS value was 1. The DS of the products is a clear function of the vinyl laurate and stearate ester intakes (Figure 4.3. A further increase leads to a reduction in the DS.5.5
Degree of Substitution (DS)
0. the ester intake is equal on a weight basis to the DMSO intake. The DS values are increasing with higher vinyl‐ ester/AHG molar ratio until a certain maximum. 2 wt% catalyst intake on starch).5. Higher concentrations of vinyl esters are expected to lead to higher esterification reaction rates.3.
5. The latter factors appears to have a strong effect on reaction rates (vide supra).1.2. The results are given in Table 4. Aburto.3. A similar trend in reactivity pattern was observed in this study and the DS decreased from 1. Here. 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.e. When using Na2HPO4 as the catalyst. i.).1. 4. 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. Effect of the addition of toluene as a co‐solvent A number of reactions were performed using a co‐solvent. The addition of toluene for the esterification of starch with vinyl‐esters surprisingly did not lead to improved DS values. et al  for the silylation of starch. the starch laurate esters display higher DS values than the starch stearates.). with high carbon numbers leading to a reduction in the DS.1. it also results in a dilution of the reaction mixture and a reduction in the polarity. K2CO3 and Na‐ acetate were tested. with reductions in polarity leading to lower reaction rates. and illustrated in Figure 4. the addition of co‐solvents (toluene/ THF) led to an increase in the DS.3. For starch
.3. It is clear that Na‐acetate and K2CO3 are considerably more active than Na2HPO4 and products with a significantly higher DS were obtained. These negative effects dominate the reaction performance at higher vinyl ester/AHG ratios and lead to a reduction in the DS values. In this case.8 for stearoyl chloride. Catalysts screening A number of alternative basic catalysts for Na2HPO4.3.6. This effect is particularly evident at higher vinyl ester/ AHG ratios (>3) (Figure 4. 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 . 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.7 for lauroyl chloride to 0. These findings were rationalised by assuming that the co‐solvents increase the solubility of the silylated products. 11‐14 in Table 4. 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 4‐6.3. Thus.). leading to enhanced reactivity. the DS of the product is also a function of the chain length of the fatty acid. Although toluene may positively effect the reaction by (partly) re‐dissolving starch ester precipitates. 4. A similar procedure was proposed by Nouvel.
and FTIR and confirm the presence of chemically bound fatty acid chains. 110oC. catalyst amount = 2%‐w.6. K2CO3 and Na‐acetate are superior with respect to activity when compared with Na2HPO4. gray: Na2HPO4 black: K2CO3 white: Na‐acetate
A study on the synthesis of corn starch fatty acid esters with high DS values is reported. K2CO3. The yellow products were characterized by 1H‐ and 13C‐NMR.1. and Na‐acetate).o. and Figure 4.5
Degree of Substitution
1.5 were obtained. the statement that the DS for the laurate esters is always higher than for the stearate esters is not generally valid and a.5
0. The DS of the laurate ester when using Na‐acetate is higher than for the stearate ester (Table 4.5
Figure 4. the DS of the product is also tunable by proper catalyst selection. in line with the findings for NaH2PO4. the two catalysts are equally effective and products with a DS of about 2. K2CO3 gave products with a significantly higher DS (2.44).Chapter 4
laurate esterification. Comparison of DS values with different catalysts (Vinylester: AHG molar ratio = 3:1.).
. 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. DMSO). when using K2CO3 as the catalyst. Thus.
2. The products were synthesised in DMSO using vinyl‐esters in the presence of basic catalysts (Na2HPO4.4. the DS for the laurate ester is lower than the stearate ester. function of the type of catalyst. Apparently.6.96) compared to Na‐acetate (DS=2. However.
E. Polym. 54. . Sagar. J. W. Eng. M. I. J. presumably due to a reduction of the polarity of the reaction medium. D. Chem. Dzenis. Synthesis of Higher Fatty Acid Starch Esters using Vinyl Laurate and Stearate as Reactants
With these catalysts. Ind. . M. . J. Sci. . Aburto. Gatenholm: Surface octanoylation of high‐ amylose potato starch films. Sci. S. Xu. Panayiotou: Synthesis. Bikiaris. Alric. A maximum was observed at a vinyl ester: AHG ratio between 2 and 4. 1999. 705‐721. D. K. Aburto. Merrill: Properties of Fatty‐Acid Esters of Starch.
. Important product properties will be described in the next chapter. 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. Y. 34. Prinos. thermal characteristics and biodegradability of extruded starch acetate foams. Bengtsson. J. Polym. Crops Prod. 1440‐1451. C.4 could be obtained for both laurate and stearate esters. J. Thiebaud.6. Appl. 21. 361‐368. The DS of the products may also be tuned with the vinyl ester/AHG molar ratio. D. J. and biodegradability of fatty‐acid esters of amylose and starch. 2003. 1208‐1217. 1995. Carbohydr. Y. Mullen and E. 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. characterization. 1‐11. Appl. Polym. I. Alric. X. 58. 74. 1942. Sci. Panayiotou: Properties of fatty‐acid esters of starch and their blends with LDPE. J. Hanna: Water solubility. and P. Appl. Borredon. 1997. Prinos. A. S. Borredon. C. . Koch. Ind. References
. Thiebaud. Polym. 2005. the DS of the product increases at higher vinyl ester intakes. J.5. A. Bikiaris. At low vinyl ester/ AHG ratio.
4. Pacsu: Starch Studies: Preparation and Properties of Starch Triesters. E. E. 1647‐1656. products with a DS > 2.W. 65. Higher ratios led to a reduction in the DS.
298. . Mater. K. Alric. 302‐307. Soininen. Peltonen. E. L. . Degee. Starch‐Starke 1999. Ydens. C. A. 251‐260. Dicke: A straight way to regioselectively functionalized polysaccharide esters. F. L. S. A. 2004. J. . P. T.F. 57. 145‐152. D.Chapter 4
. Nuzillard: Proton NMR spectroscopy assignment of D‐glucose residues in highly acetylated starch. Med. 1735‐1743. Mano. Macromol. E. Biosci. I. Mormann. W. B. 2003. 56. . P. Petzold. 261‐267. 1997. I. 57. 51. Mater. G. 255‐263. Borredon: Organic solvent‐free transesterification of various starches with lauric acid methyl ester and triacyl glycerides. Senocq. R. Cellulose 2004. Koniarova. Reis: Thermal properties of thermoplastic starch/synthetic polymer blends with potential biomedical applicability. Polymer 2002. Dubois. Laatikainen. Starch‐Starke 2005. Hyvarinen. M. Alric. 118‐121. I. Laignel. Asplund. Nouvel. Elomaa. J. G. M. . R. Stein. Gunther. Bliard. Mouysset‐Baziard. M. Aburto.
. P. 341‐347. R. . J. . H. Polym. Six: Partial or total silylation of dextran with hexamethyldisilazane. . W. H‐1 NMR and TGA/IR. S. Hamaili. 11. 14(2) 127‐135. D. Borredon: Free‐solvent Synthesis and Properties of Higher Fatty Esters of Starch – Part 2. Res.. C. Aburto. and E. Dellacherie. Massiot. J. Carbohydr. J. J. 43. Carbohydr. Starch‐Starke 2004. M. 1. Kussler. Al‐Higari: Acylation of starch with vinyl acetate in water. 2001. Klemm: Preparative and H‐1 NMR investigation on regioselective silylation of starch dissolved in dimethyl sulfoxide.L. Urtti: Determination of the degree of substitution of acetylated starch by hydrolysis. Sci. Einfeldt.
Mechanical tests show that the products have tensile strength (stress at break) between 2.96).5 MPa.7‐3. High‐DS products (DS= 2. The thermal and mechanical properties of some representative product samples were determined. Keywords: starch esters. The reactions were carried out in DMSO using basic catalysts (Na2HPO4. vinyl laurate. The effect of the process variables (vinyl ester to starch ratio. K2CO3. 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.45‐1. The results were adequately modeled using a non‐linear multivariable regression model (R2≥0. catalyst intake.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.39) are totally amorphous whereas the low‐DS ones (DS= 1. vinyl stearate
. and Na‐acetate). The basicity of the catalyst and the reaction temperature have the highest impact on the product DS. elongation at break of 3‐26%. and modulus of elasticity of 46‐113 MPa. The thermal stability of the esterified products is higher than that of native starch.75) are still partially crystalline. trans‐esterification.26‐2.
and incompatibility with hydrophobic polymers . The limited use is mainly because of a number of unfavourable properties of starch such as low moisture resistance. packaging material. the physicochemical properties of the products were not mentioned. however. The development of novel biodegradable plastic materials capable of decomposing when given an appropriate environment and time is of utmost importance . 12] studied the synthesis and properties of of C1‐C6 esters of starch. Esterification is one of the oldest methods used to improve starch properties . Unfortunately. et al [15‐19] synthesized longer chain fatty esters (C8‐C18) of potato starch and corn amylose using fatty acid chlorides and pyridine. The global starch production was estimated at 60 million tons in 2004 . et al  and Aburto. and particularly on the synthesis of starch acetate (C2) [11‐12]. and food service‐ware . Thiebaud. the mechanical properties were still not satisfactorily. polyethylene  and polycaprolactone  and chemical modification by esterification with organic acids [11‐19]. Introduction
Plastics are very useful materials and are used in large amounts (245 million ton per annum) [1‐2]. However.
. The products were biodegradable. high brittleness. and this causes serious environmental problems [1. Starch is an attractive feedstock for the synthesis of novel biodegradable plastics.1. most of the plastic waste is not biodegradable.Chapter 5
5. Alternative routes to avoid the use of acid‐chloride/pyridine combinations have been developed. Most early studies mainly focused on the use of short‐chain carboxylic acid (C1‐C4). The use of vinylesters has also been explored [20‐22]. Typical examples are blending with polyvinyl alcohol . however.65) . Two examples using higher vinyl esters (vinyl laurate) were reported by Mormann. The products were hydrophobic and the mechanical properties were considerably improved compared to native starch. Sagar and Merill  studied the synthesis and properties of C4‐C6 esters of high amylose starch. is rather limited to date. The largest application area of plastics is the use as packaging material (37%) . The mechanical properties of products with longer ester chains (C4 and C6) products and plasticized acetate esters were considerably improved compared to native starch. 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. 0. Mullen and Pacsu [11. 3‐4]. Chemical or physical modification is required for successful applications as biodegradable polymers. the research activities were mainly limited to the use of vinylacetate. It is cheap and abundantly available . However. The high DS esters showed interesting properties. et al . Examples are agricultural mulch . The use of native starch as a building block for novel biodegradable polymers.
the Netherlands). DSC analyses were performed on a TA Instruments DSC 2920. Japan). The experiments were modeled using non‐linear multivariable regression.2. The samples were heated to 700 oC in a nitrogen atmosphere with a heating rate of 10°C/min. and acetic anhydride (Merck. sodium acetate. and tetrahydrofuran (THF) were supplied by Acros (Belgium). 4‐N. vinyl ester to anhydroglucose ratio.
5. All these chemicals were used as received.2. It was dried for 48 hour at 105 oC under vacuum (~1 mbar) to reduce the moisture content below 2 % before use. The catalysts for the esterification reaction were analytical grade potassium carbonate (Boom. Analytical Equipment H‐ and 13C‐NMR spectra were recorded in CDCl3 on a 400 MHz Varian AMX NMR machine. each sample was cooled to ‐50°C and then heated again to 200 oC (heating rate 10 oC/min). In addition.96) were prepared by reacting the starch with vinyl laurate or vinyl stearate in the presence of basic catalysts in DMSO.N‐ dimethylaminopyridine (DMAP). This chapter describes systematic studies using Design of Experiments (DOE) to gain insights in the effect of process variables (temperature. and disodium hydrogenphosphate dodecahydrate (both Merck. Materials Corn starch (approx. 5.2. After a first heating run from room temperature up to 200 oC to delete the thermal history of the material. TGA measurements were performed using a Perkin Elmer TGA 7 Thermogravimetric Analyzer. pyridine. Germany).2. and Tg) of each sample were determined from the spectra related to the cooling run and the second heating one.24 ‐ 2. Tc. T‐Bones samples (with thickness of 2 mm) for determination of the tensile properties were prepared using a melt press
. Technical grade dimethyl sulfoxide (DMSO). the thermal and mechanical properties of representative examples of the highly hydrophobic materials are described and discussed. Long fatty esters of corn starch with a broad range in degree of substitution (DS= 0. The spectra were recorded at 50 oC. The samples (about 10 mg) were placed in sealed aluminum cells. The thermal properties (Tm. vinyl laurate (Fluka. 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 . Analytical grade vinyl stearate (Aldrich. Analytical grade methanol. catalyst type and intake) on the DS of the products. 73% amylopectine and 27% amylose) was purchased from Sigma (Germany). Germany) were used without further purification. Germany).1. Materials and Methods
5. as recommended by Laignel et al . and toluene were obtained from Labscan (Ireland).
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 220.127.116.11. 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. 18.104.22.168. 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 3.00 3.56 1.00 3.00 3.00 5.52 2.40 1.41 1.00 5.00 2.09 0.88 2.00 3.00 3.00 2.00 5.00 2.00 3.00 4.00 5.48 0.00 5.50 2.00 5.00 5.00 2.42 2.12 0.59 2.00 5.00 2.00 4.94 2.54 2.00 5.00 3.00 5.00 5.00 Catalyst Temperature [%‐w] [˚C] 2.00 5.2.00 3.00 3.00 4.
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 5.00 5.44 2.00 3.00 3.00 5.00 2.00 3.00 5. Experimental and Modeling Studies on the Synthesis and Properties of Higher Fatty Esters of Corn Starch
Table 5.00 5.00 5.00 2.27 1.00 3.00 5.00 2.50 2.00 5.00 5.00 5.00 5.53 1.52 2.78 2.00 3.87 1.00 4.67 2.00 5.00 3.50 2.35 0.07 0.00 5.47 1.00 5.27 1.24 2.44 2.00 3.00 3.84 2.00 5.00 5.00 5.75 1.00 3.00 3.00 3.00 2.00 2.72 2.00 5.70 2.17 2.00 2.00 5.12 2.50 0.82 2.00 5.54 2.00 2.00 5.00 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 0.23 2.00 3.00 5.00 3.00 5.79
Experiment Vinyl Ester 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 laurate laurate laurate laurate laurate laurate laurate laurate laurate laurate laurate laurate laurate laurate laurate laurate laurate laurate laurate laurate laurate laurate laurate laurate laurate laurate laurate stearate stearate stearate stearate stearate stearate stearate stearate stearate stearate stearate stearate stearate stearate stearate stearate stearate
.00 2.00 5.00 5.00 5.62 0.00 2.32 1.27 0.93 0.28 2.00 2.96 0.50 2.46 1.11 2.00 5.00 5.
21 2.3.00 5.1947
β0 β1 β2 β3 β4 β5 β6 β7 β8 β9 β10 β11 β12 β13
.64 2. 5.00
3.0003 0.00 5.The modelled values for βij for the two esters are given in Table 5.0016 0.00 5.)
Table 5.00 3.0017 0.1.00 3.0249 ‐0.00 5.96 2.90 2.1748 0.0732 ‐0.00 3.50 2.6939 0.1104 ‐0.00 2.3.1082 ‐0.).0857 ‐0. The type of catalyst was quantified using the basicity constant (pKb) in water [26‐27].0008 0.55 2.0009 0.3. Values for the coefficients of the DS model for starch laurate and stearate
Coefficient Starch Laurate 9.3.0011 0.2183 0.0023 0.0470 0.00 2.1.1568 ‐2.00 5.00 5.0006 ‐0.00 3.00 2.41
mol ratio of the vinyl ester to AHG units of starch
5.0327 ‐2.00 5.).
stearate stearate stearate stearate stearate stearate stearate stearate stearate stearate
CH3COONa K2CO3 K2CO3 K2CO3 K2CO3 K2CO3 K2CO3 K2CO3 K2CO3 K2CO3
4.59 2.1979 Starch Stearate 17.0066 0.00 4.0120 0.3.5650 0.00 3.64 2.0008 ‐0. 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. The experimental data for both starch laurate and starch stearate esterification are best described with a quadratic model including interaction terms (eq.24 2.50
95 80 80 110 110 80 80 110 110 95
1.0010 ‐0.98 2.
DS = β 0 + β1 x1 + β 2 x 2 + β 3 x3 + β 4 x 4 + β 5 x1 x 2 + β 6 x1 x3 + β 7 x1 x 4 + β 8 x 2 x3 + β 9 x 2 x 4
2 2 2 + β10 x3 x 4 + β11 x 2 + β12 x3 + β13 x 4
(5.00 5.Chapter 5 45 46 47 48 49 50 51 52 53 54
a.0027 ‐0.0001 0.
Starch Stearate Model Error Total SS 20.). which represents an “internal” validation method for the model.5
2 DS predicted
1.629 P‐value 1.8167 21.5
1.967 0.5 DS actual
2. The parity plots of both models are given in Figure 5.970 (starch laurate) and 0. starch laurate
b.4. respectively and it may be concluded that the models describe the experimental data well. The R2 values for the models are 0.1. 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 5.970 0.5 DS actual
0.2903 P‐value 1. and confirm this statement. the adjusted‐R2 values are very close to the R2 ones.945 0.6118 0.026 x 10‐7 R2 values R2 R2adjusted R2press 0. Table 5.5
1.805‐0. we also performed a PRESS analysis  (see corresponding R2 values in Table 5.5
1.0583 F 27.5
0.0356 F 35.967 (starch stearate).4.939 0. which indicate  that all significant variables are included in the model. starch stearate
2.96 x 10‐8 R2 values R2 R2adjusted R2press 0.3399 0.875) indicate that the model correctly predict the
0. Also in this case the reasonable R2 values (0.5
2.7699 DF 13 14 27 MS 1.2569 0.4986 16.4.9531 0. The very low P‐ values indicate that the models are statistically significant.1. Finally.5
a. Parity plots of the DS models for starch laurate and stearate Moreover.8386 DF 13 14 27 MS 1. Analysis of variance for DS models for starch esterification
Starch Laurate Model Error Total SS 16.
22]. Unfortunately. The product DS for Na2HPO4 was considerably lower (1. The resulting anion will react with the vinyl ester to from the product. for both starch laurate and starch stearate are given in Figure 5. it can be rationalised that the rate of the reaction (and thus the product DS) will be higher when using a stronger base.
.25) and this leads to higher product DS values for K2CO3.00) than for K2CO3 (2.). 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. Catalyst performance of CH3COONa was slightly less than for K2CO3 whereas the lowest DS products were produced when using Na2HPO4 as the catalyst. whereas catalyst performance is considerable lower.66) than CH3COONa (pKb 9. This was not the case for alkaline catalysts. This is indeed the case when comparing the performance of K2CO3 with CH3COONa.82). 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. two plots showing the dependence of the DS on reaction temperature and catalyst amount (in %‐w). However.8) is intermediate between that of the other two catalysts. the performance of Na2HPO4 does not follow this trend.2.2.Chapter 5
products DS as function of the process variables within the range of experimental variables. On the basis of the sequence. 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. Of all the variables studied. The pKb for the latter (6. K2CO3 is a stronger base (pKb 3.18) and CH3COONa (1. It is assumed that the first step in the reaction sequence is activation of the starch OH groups by deprotonation by a base [21. A similar trend was observed by Dicke  for the acetylation of Hylon VII starch using vinylacetate. Another explanation may be related to the regio‐chemistry of the reaction. the type of catalyst has the largest effect on the product DS for both the laurate and stearate esters (Figure 5. 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. To evaluate the effect of each variable on the product DS. the base strengths of the catalysts in DMSO are not known. The highest product DS values were obtained using K2CO3.
although to a lesser extent. starch stearate
Figure 5. the catalyst intake and reaction temperature also affect the product DS (Figure 5.). Experimental and Modeling Studies on the Synthesis and Properties of Higher Fatty Esters of Corn Starch
a.2. 3D contour plot of the DS as a function of reaction temperature and catalyst intake (at constant vinylester to AHG mol ratio of 4) Besides the type of catalyst. As expected and in
.2. starch laurate
the vinyl ester to AHG ratio has the smallest effect on the product DS.Chapter 5
line with studies on starch acetates .39)
. DSstearate = 2. Of all variables studied.) DS ( 2. 5.
Tm = 21 oC
Tc = 12 oC
10 Temperature ( C)
Figure 5.26. Product Properties The thermal and mechanical product properties for two starch‐laurate and stearate samples with medium (DSlaurate = 1. .3. and the results for all samples are summarised in Table 5.3. . Representative DSC spectra for the starch stearate are given in Figure 5.45 (. DSC analysis of starch stearate esters ( ) DS 22.214.171.124) and high DS values (DSlaurate = 2.3. higher reaction temperatures and catalyst intakes lead to higher product DS values.39) were determined. DSstearate = 1.5.
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). TGA analysis was performed to study the thermal degradation behavior of the starch laurate and stearate samples (Figure 5. Thermal transitions at relatively high temperature (melting. 22]. These values are reduced considerably compared to virgin corn starch.a 27 n. 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.a 21 Tc [˚C] n.a 22 n. Unfortunately the thermal transitions for the virgin material lie above the degradation temperature and can therefore not be accurately determined . these are the first examples of completely amorphous starch esters.
.d. In any case.75 2.45 Tm [˚C] n. close to the value obtained in our study for the medium DS starch stearate (21 °C).8 . esterification of starch with vinyl laurate and stearate reduces the crystallinity of the starch considerably and leads to a lowering of the transition temperatures. For instance a Tm of 32 oC was reported for a potato starch stearate ester with a DS of 1.
n. Changes in the transition temperatures of native starch by esterification have been reported in the literature [13‐14. 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. 17].26 1. The melting point of the potato starch stearate was close to that of model compounds for the stearate side chains (methyl stearate. In particular the behaviour at relatively high temperatures (>0 oC) is further discussed.5.). or glass transition above 0 oC) were even absent for the high DS products. Experimental and Modeling Studies on the Synthesis and Properties of Higher Fatty Esters of Corn Starch
Table 5.d. At moderate DS values the products still display a melting temperature (at 20‐30 oC) and a crystallization temperature (at 10‐25 oC). crystallization. in the present work.
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 2.4.d. However.39 1. 40‐42 oC. especially in the amylopectin part [15.a 12
a. which is known to contain crystalline areas.= not detectable
The thermal behavior of the products is a strong function of the DS. octadecane. implying that these products are fully amorphous.d. To the best of our knowledge.d.
4. Similar trends were reported for amylose octanoate‐ stearate‐laurate esters  and potato starch octanoate‐ laurate esters .6. .).4.): starch stearate (DS 2.Chapter 5
Temperature ( C)
Figure 5. The difference in thermal stability between the laurate and stearate sample is limited. . ( ): native starch ( ): starch laurate (DS 2. elongation at break and elasticity modulus) of the products were determined (Table 5. The mechanical properties (stress at break. Native corn starch degrades at lower onset temperatures (250‐300 oC) than the corresponding esters (300‐350 oC).) curves clearly show starch esterification with either laurate or stearate results in products with enhanced thermal stability.26) (.
The TGA (Figure 5.
TGA analysis of various starch samples.
45) displays a very similar behavior with respect to the laurate samples (i. independently of the DS values.7 ± 4.
. calculated as the ratio between the stress at yield (σy. This product is more rigid compared to the product with medium DS value (higher tensile strength and elasticity modulus. Starch stearate with a relatively low DS (1.1 Elongation at break [%] 26.3 21. and lower elongation at break). This indicates that their mechanical behavior is comparable to that of typical engineering polymers .1 ± 15.7‐3.26 1.7 ± 4. This is in agreement with previous studies regarding the mechanical properties of starch esters with different chain length of the fatty acid moiety [13. observed for the high DS starch stearate sample.5 ± 1.05). A deeper understanding of the mechanical behavior can be gained by investigating the shape of the stress‐strain curves (typical examples in Figure 5.2 8.0 3. The corresponding yield strain (roughly 0.7 ± 3.1 ± 1. 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 82. determined from the stress‐strain curves as the stress at which the material ceases to be linearly elastic) and the modulus.7 ± 14. starch stearate at relatively high DS values (2.5 MPa. Experimental and Modeling Studies on the Synthesis and Properties of Higher Fatty Esters of Corn Starch
Table 5. elongation at break between 3‐26%.025) lies still in the typical range of engineering polymers but also close to that typical of composite materials .). however. thus indicating a relatively more rigid material with respect to all other samples.2 2. but higher elongation at break.39) does not show any plastic behavior but only an elastic one. The values of the yield strain.e. An inverse behaviour is.9 ± 14.5.3 ± 2. 17].
Mechanical properties of the starch esters
DS 2. For starch laurate.6.45 Stress at break [MPa] 3. plastic deformation and yield strain of about 0.9 46. a clear plastic behavior with a maximum in the stress‐strain curves.75 2.0 112.39 1.2 ± 0. On the other hand.0 ± 2.1 Modulus [MPa] 59.7 ± 1.8
Starch Ester Starch laurate Starch laurate Starch stearate Starch stearate
All materials generally show tensile strengths between 2.05 for both laurate samples. Both starch laurate samples display. is roughly 0. and modulus of elasticity between 46‐113 MPa.0 3.3 3. an increase in product DS results in lower tensile strengths and modulus of elasticity.
7.75) ( ): starch laurate (DS 2.5. Stress‐strain curves for starch esters.
.45) (‐ . Although there are many differences in the synthetics methods as well as in testing conditions and procedures.39). 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. .Chapter 5
4 stress (MPa)
0 0 5 10 15 20 strain (%) 25 30 35
Figure 5. ‐): starch stearate (DS 2. ( ): starch laurate (DS 1.26) (.).): starch stearate (DS 1. .
3 Potato Starch Stearate Ester Low DS (1. as confirmed by DSC.3
High DS (2.1 High DS (2. we show that although no crystallization of C18 occurred (see DSC data of the high DS material in Table 5. On the other hand at low DS values comparable stress at break (3.1 21.7 MPa for potato starch ones) are coupled with significant differences in the elongation (22% for corn starch esters. n.75) 3. 3.
.1 ±1. 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.9 ±0.5.4 1500 ±8. merely because of the structural property of the C18 side chain with respect to the starch backbone.26) relatively high stress at break (3.2 8.0 3. Experimental and Modeling Studies on the Synthesis and Properties of Higher Fatty Esters of Corn Starch
Table 5. They related this inverse property of starch stearate to crystallization of C18 side chains.).
n.7 ±3. 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.7 ±0.7) 0. The same considerations are valid for starch stearate at high DS values (higher stress and lower elongation when using corn starch instead of potato one).a.3 10 ±2
Low DS (1.7 MPa) coupled with a significant (1500%) elongation. In this respect the difference in stress and elongation at break for the high‐DS starch laurate sample is striking.0 26. confirms that the corn starch esters synthesized in our research are relatively rigid materials as compared to the potato starch esters. While for corn starch (DS 2.a.7 ±4.5 ±1.6 Corn Starch Stearate Ester Low DS (1.a. DS) which mainly determines the thermal and mechanical behaviour.45) 3.0 ±2.7 ±1. 9% for potato starch). High DS (2.1 ±15.39) 2. for potato starch exactly the opposite is observed: relatively low stress (0. Here.= not available
Inspection of the mechanical properties of the starch esters as shown in Table 5.2
Stress at Break [MPa] Elongation at Break [%]
a.7. but the interplay between those factors.5 MPa) is coupled with a low elongation (26 %).26) 3.2 ±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. an unexpected behavior was also observed for the mechanical properties of the starch esters from potato starch.6 9 ±2 High DS (2.a.8) 3.7 ±0. the high DS material is still more rigid. 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) .7) 1. Moreover.a.2 MPa for corn starch esters.7. In agreement with our own data.
5 for starch stearate). and Na‐acetate). The use of K2CO3 and CH3COONa catalysts resulted in medium‐high DS products (2. catalyst intake. Thus.5. The models may be used to determine the appropriate process conditions to obtain a product with a pre‐defined DS.96 and P‐value of ≤10‐7) second‐order mathematical models correlating the effect of process variables (vinyl to AHG‐starch mol ratio. average value of mole of substituted –OH per mole of anhydroglucose (AHG) units [‐]
. 1.Chapter 5
5. 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. K2CO3. on the synthesis of corn starch esters with long chain fatty acids is described. Comparison with literature data indicates that the properties may be (fine) tuned also by the starch source. while the use of Na2HPO4 resulted in low‐medium DS products (0.4‐3 for starch stearate). Reaction temperature and catalyst intake also affect the product DS although to a lesser extent than the type of catalyst.
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 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. The DS of the products is a strong function of the basicity of the catalyst. including statistical modelling. reaction temperature.5 for starch laurate. 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. Statistically adequate (R2 ≥0.4. Nomenclature
: regression coefficients obtained by a multiple regression procedure [various units.3‐1. 0.07‐1. Conclusions
A systematic study.1‐2.9 for starch laurate. and catalyst basicity) to the DS of the starch ester products were developed.
W. Properties.W. Prod. Munich. UK. The compelling facts about plastics: An analysis of plastics production. Westhoff. and Biopolymers. The Royal Veterinary and Agricultural University. Denmark. . . Marsden: The challenge of domestic waste disposal. 1980. Structure – Properties ‐ Applications. D.starch. R. 2000. Chem. C. Hanser Publishers. Copenhagen. E. S. Frederiksberg C. Levy). 1986. Website of International Starch Institute. M. Dev. Otey. Culter: Plastics Packaging.6.html. A. S. Ehrenstein: Polymeric Materials. Germany. Denmark. . Belgium. Munich. Blackie Academic & Professional. Stevens: Green Plastics.dk/isi/stat/rawmaterial. 2008 . J. Glasgow. M. 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. Denmark: http://www. P. Princeton University Press. Germany. Res. J. C. . Eng. accessed on August 01. New Jersey. E. in Packaging in the Environment (Ed. G. Processing. Brussels. Hernandez. References
. Conference Proceedings. Applications. USA. 2000. in The Food Biopack Conference: Foodstuffs. Plastics Europe. M. 2000. Bastioli: Global Status of the Production of Biobased Packaging Materials. . Ind. Weber). 592‐595. H. Packaging.J. demand and recovery for 2006 in Europe. G. F. 27‐29 August 2000 (Ed. Doane: Starch‐based blown films. 2008. 19. and Regulations. Hanser Publishers. An Introduction to the New Science of Biodegradable Plastics. Association of Plastics Manufacturers. . Selke. Aarhus. 2002. Department of Dairy and Food Science.
Ind. W. Bikiaris. M. E.. Starch‐Starke 2004.. Ind. J. Sagar. J. 74. . Alric. Chem. 101‐112. Yang. J. Borredon: Preparation of long‐chain esters of starch using fatty acid chlorides in the absence of an organic solvent. A. Thiebaud. 145‐152. Appl. S. Eng. Panayiotou: Properties of fatty‐acid esters of starch and their blends with LDPE. L. Mullen. C. Part C: Polym. Weber). X. J. D. K. Starch‐Starke 1999. Chem. J. 1647‐1656. Z.D. J. Macromol. 34. E. Sci. Department of Dairy and Food Science. Carbohydr. . Polym. M. Panayiotou: Properties of octanoated starch and its blends with polyethylene. Y. L. I. 302‐307. Appl. Thesis. Alric. A. 58. J. . . 1995. E. de Graaf: Ph.Chapter 5
. J. Mouysset‐Baziard. Prinos. E. Sci. H. Bikiaris. Senocq. C. Borredon. 1999. J. A. E. . I. 385– 409. C. Rev. D. 1440‐1451. 2000. Polym. . Borredon: Free‐solvent synthesis and properties of higher fatty esters of starch ‐ Part 2. Janssen. A. 51. 28. 1942. C. P. . Pacsu: Starch studies: preparation and properties of starch triesters. Mormann. R. Aburto. Frederiksberg C. Aburto. A. E. Aburto. C. 1209‐1217. Borredon: Organic solvent‐free transesterification of various starches with lauric acid methyl ester and triacyl glycerides. R. J. 118‐121. . F. and biodegradability of fatty‐acid esters of amylose and starch. Hamaili. Aburto. Alric. I. status and perspectives. characterization. Pacsu: Starch studies: possible industrial utilization of starch esters. K. Aburto. D. Wang. 381‐384. Thiebaud. . J. Alric. Wang: Properties of starch blends with biodegradable polymers. Starch‐Starke 1999. I. . Polym. G. A. E. G. Polym. Prinos. W. 57. Aburto. Borredon. B. 132‐135. Polym. J. 65. J. Borredon. 137‐144. Broekroelofs. Biobased packaging materials for the food industry. Sci. Bikiaris. report of Food Biopack Project (Ed. 705‐721. 56. 35. . Eng. Prinos. D. Sci. Starch‐ Starke 2005. 1943. Denmark. Panayiotou: Synthesis. Appl. 1997. J. J. 1997. the Royal Veterinary and Agricultural University. S. Alric. Alric. de Graaf. . 34. E. Thiebaud. Mullen. M. S. Rijksuniversiteit Groningen. E. 51. Merrill: Properties of fatty‐acid esters of starch. I. J. 1996
. Beenackers: The kinetics of the acetylation of gelatinised potato starch. I. Al‐Higari: Acylation of starch with vinyl acetate in water. . Carbohydr. 1995. C43. W. W. 2003.
. New York. C. Underwood: Quantitative Analysis 5th edition. L. Bliard. L. accepted for publication in Starch‐Starke. A. Day. M. 2001. USA. H. R. Res. K. Nuzillard: Proton NMR spectroscopy assignment of D‐glucose residues in highly acetylated starch. Science. R. Cellulose 2004. A. John Wiley & Sons Inc. USA. Junistia. H. 2008. Ashby. 251‐260. Heeres: Synthesis of higher fatty acid starch esters using vinyl laurate and stearate as reactants. . Amsterdam. 255‐263. 1986. 298. Picchioni. Montgomery: Design and Analysis of Experiments 5th edition. Experimental and Modeling Studies on the Synthesis and Properties of Higher Fatty Esters of Corn Starch
. Dicke. Massiot. the Netherlands. D. B. 11.. B. Prentice‐Hall. J. Sugih. F. Janssen. R. . Butterworth‐Heinemann. C. New Jersey. A. Cebon. Manurung. 1997. M. 2007. M. Laignel. L. . Carbohydr. . Shercliff. A.
. J. A straight way to regioselectively functionalized polysaccharide esters. Materials: Engineering. D. P. G. Processing and Design.
at relatively high GMA and BPO intakes. Both act as compatibilizers for PCL/starch blends by improving the interfacial adhesion between the starch particles and the PCL matrix. The FD values for PCL‐g‐DEM were considerably lower (up to 7 %). The reactive interfacial agents were tested for their performance in PCL‐starch blends. PCL‐g‐ glycidyl methacrylate (PCL‐GMA) and PCL‐g‐diethyl maleate (PCL‐g‐DEM) is decribed. the mechanical behavior of the compatibilized blends is in general different from that of pure PCL and of the corresponding uncompatibilized blends. In particular the elastic modulus for the compatibilized blends is significantly higher than that of uncompatibilized ones. At relatively low starch intakes. grafting
. compatibilizer. As a result. 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.Chapter 6
Synthesis and Properties of Reactive Interfacial Agents for Polycaprolactone‐Starch Blends
The synthesis of two reactive interfacial agents for starch‐PCL blends. Highest values of the FD (up to 45 %) were observed for PCL‐g‐GMA. Keywords: starch‐PCL blend. PCL‐g‐DEM has at least a comparable performance with respect to PCL‐g‐GMA. diethyl maleate. despite the expected differences (favorable to PCL‐g‐GMA) in the in situ formation of the compatibilizers. 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. glycidyl methacrylate. This discrepancy could be explained on the basis of the functional groups (GMA or DEM) distribution along the PCL backbone.
PCL has gained considerable interest for possible applications in the fields of packaging materials and medical applications [10‐11]. upon mixing) by using a functionalized PCL. The latter displays the presence. these synthetic polymers are poorly or non‐biodegradable. polycaprolactone (PCL). polyesters are considered very promising alternatives . Unfortunately. Starch modifications to improve the product properties like enhanced hydrophobicity and mechanical properties were already reported in the early 19th century [1.1. This leads. When using ungelatinized starch as a component in the blend.Chapter 6
6.e. As a consequence. is known to be degraded with ease by microorganisms widely distributed in nature . Unfortunately. This phenomenon is highly undesirable and limits the application range considerably . Such structure is rather difficult to achieve by simple copolymerization methods and it is usually prepared in situ (i. However. To reduce the tendency for phase separation.e. 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. an interfacial agent) may be used to improve the interfacial association between the two polymer phases. such as polyethylene (PE) or polystyrene (PS) [3‐6]. 2]. The ester bonds are susceptible to attack by water and this leads to enhanced biodegradability. 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. as in this study. of polar groups (usually epoxides or anhydrides [13‐18]) able to react with the – OH groups along the starch backbone. a compatibilizer (i. to an ideal compatibilizer having both PCL and starch blocks. along the backbone. Among these. it would be actually more accurate to define the block copolymer as an “interfacial agent”. Introduction
Starch is a cheap and abundantly available natural polymer with very good application perspectives in the area of biodegradable plastics. for starch/PCL blends. Aerobic soil‐burial experiments showed that the mechanical properties of PCL films decrease rapidly in time . Blending of starch and synthetic biodegradable polyesters has been widely applied for the synthesis of fully‐biodegradable products. blends of hydrophilic starch and hydrophobic biodegradable polyesters exhibit phase separation  due to differences in polarity of the building blocks. One of the well known modification strategies is blending the starch with polymers displaying a stronger hydrophobic character and better mechanical properties. synthetic biodegradable polymers have been applied. To overcome this issue. 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 ).
. A well‐known biodegradable polyester.
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. 73% amylopectin and 27% amylase) was obtained from Sigma and high molecular weight PCL (CAPA 6503. however. 16]. Ireland. Methods 6. Exploratory studies on the synthesis of PCL‐g‐GMA and its applications for starch/PCL blends have been published [13. Materials and Methods
6. 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. 6.2. Mw=50000) from Solvay Caprolactone.2. Belgium.
. xylene (99. The materials were mixed for another 5 minutes.2. Germany and methanol (99. to the best of our knowledge. The kneader was heated to 130 oC and PCL (CAPA 2304) was added while maintaining a rotational speed of 80 rpm. >99%) was obtained from Acros.2. systematic studies and subsequent quantification of the functionalization reaction has not been reported to date. Mw=3000 from Solvay Caprolactones. After the PCL was melted (1‐2 minutes).8%) from Merck.2. an absolute novelty of the present research. Tetrahydrofuran (THF. and benzoyl peroxide 75% (Merck. diethyl maleate ≥97% purity (Fluka). This chapter describes a systematic study on the synthesis of two reactive interfacial agents for starch‐PCL systems.1.8%) from Labscan. PCL‐g‐glycidyl methacrylate (PCL‐g‐ GMA) and PCL‐g‐diethyl maleate (PCL‐g‐DEM). The various compatibilizers have been tested for their performance in PCL‐starch blends. The synthesis and application of PCL‐g‐DEM is. This low molecular weight PCL grade was used without further purification. Glycidyl methacrylate 97% purity (Aldrich). Compatibilizer synthesis The compatibilizers were prepared in a Brabender Plasticorder PL2000 batch‐ kneader (chamber volume 35 cm3).
6.2. after which the equipment was stopped and the chamber was opened to collect the samples. The starch was dried for at least 24 h at 110oC under vacuum (approx. Materials Polycaprolactone (PCL) CAPA 2304. ~1 mbar) prior to use. UK. Corn starch (with approx. Germany) were used as received. UK was used for the preparation of the interfacial agents. Intakes for each experiment are given in Table 1 and 2. The intake of reagents was maximally 75‐80% of the volumetric volume to ensure proper mixing. a solution of BPO in GMA or DEM was added drop by drop over a period of 5 minutes.1.
4. the chamber was opened and the resulting material was collected. 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). The solvent was decanted and the solid product was dried in a vacuum oven (40 oC. Subsequently. The error on the transition temperature is assumed to be ± 1 oC and 5 % of the calculated values for the corresponding enthalpies. 6.2. 6. Scanning Electron Microscopy (SEM) was performed using a Jeol 6320 F Scanning Electron Microscope. PCL‐g‐DEM (5 g) was dissolved in xylene and refluxed at 150 oC for 2 h.3. Methanol (450 mL) was added to the filtrate.Chapter 6
6. created using a Cressington 208 sputter coater.2.2. and then filtered. the samples were covered with a paladium/platinum conductive layer of 3 µm thickness. Tensile tests were performed using an Instron 4301
.2. Before analysis. Work‐up of PCL‐g‐GMA products  To remove unreacted GMA monomer and GMA homopolymer.8%. Infrared spectra were collected with a FT‐IR apparatus in the ATR mode using a Spectrum 2000 instrument from Perkin Elmer. 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 . 6.2. 5 mbar). Each sample was first heated from 0 oC to 100 oC (heating rate 10 oC/min) to remove the thermal history of the material. The resulting suspension was filtered and precipitated using methanol (450 mL) at 6‐8°C. stirred for 1. Digital Scanning Calorimetry (DSC) measurements were performed using a Q1000 TA Instruments equipped with a TA Instruments DSC cooling system.5 h. PCL‐g‐GMA (5 g) was dissolved in 50 ml THF. and the product was precipitated at 6‐8 oC. The content was blended for 15 minutes. Analytical Methods H‐NMR measurements were performed using a 400 MHz Varian AMX Oxford NMR apparatus with CDCl3 (99. 5 mbar).2. PCL was added to the chamber followed by the addition of the starch and the reactive compatibilizer. The solvent was decanted and the solid product was dried in a vacuum oven (40 oC. An operation temperature of 170°C and a rotation speed of 80 rpm were applied . Preparation of PCL‐starch blends with the reactive compatibilisers The PCL‐starch blends were prepared in a Brabender Plasticorder PL2000 batchkneader (chamber volume 35 cm3).2. Aldrich) as the solvent.2.3.
2. As a result we were able to obtain a model for the FD of the reaction. The T‐bone samples were prepared using a Fontijne Holland TH 400 hot‐ press. 6. 15‐16. The validity of the model was determined by performing an analysis of variance (ANOVA. stress at break (σ) and modulus (E) were measured. For a given sample/blend.2 ppm) or DEM (‐CH2‐ protons at δ 4.
. leading to a 10% relative error in the FD values. On the basis of the latter values. For every T‐bone.4.2. Table 3).0 ppm [13. strain at break (ε). 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. 8 different T‐bones were used. the F‐ value for the model is determined followed by the P‐value. 19]. 6. 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. Synthesis and Properties of Reactive Interfacial Agents for Polycaprolactone‐Starch Blends
machine. The latter is a measure of the statistical significance of the model. This procedure is described in detail in the literature  and consists of calculating the sum of squares (SS) for the model and the error.1. 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).)
The FD was calculated using 1H‐NMR by comparing the area of protons belonging to the GMA (‐CH< proton at δ 3. it is possible to calculate the mean square (MS) for the model and the error.3. When the relative degrees of freedom (DF) are known.1. A 5% relative error in the peak area of the NMR spectra was assumed. 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)
(6.2 ppm) side chains with that of a characteristic proton resonance of the PCL backbone (‐CH2‐ protons at δ 4.
1. 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 6.1.Chapter 6
6.).1.3. 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 6.3.
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
It is most probably related to the fact that GMA is able to propagate to longer grafted chains when reacted with PCL (Figure 6.2.6% mol/mol-CL units 15 GMA. Thus an increase in the GMA intake will provide more monomer available for the growing of the grafted chains leading to higher FD values.3. BPO=1. The results are graphically provided in Figure 6. 130°C) Higher GMA intakes lead to higher FD values.3% mol/mol-CL units DEM. 22]. 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% mol/mol-CL units DEM. BPO=0.). BPO=0. subsequent reactions of MA to an already grafted MA molecule hardly occur. 6.
GMA. higher FD values for GMA than for DEM are in line with this explanation and support the proposed molecular reaction mechanisms. Our experimental findings. Effect of GMA and DEM to CL‐units ratio (mol/mol) on the FD of the products (constant PCL intake. showed that MA reacts easily with a radical on the PCL backbone.Chapter 6
chemical structure of DEM.
. However. Hence.1.1% mol/mol-CL units 10
Substrate Intake (mol% on CL units)
Figure 6. the length of a MA graft is always unity whereas longer grafts are possible for GMA.1.3.3. i.e. BPO=0. This trend is independent of the BPO amount and matches with data reported by other groups [13.
4.).2 GMA. for which a levelling off and eventually a decrease of the FD values for relatively higher MA amounts has been observed.1. although a slight increase in the FD values might be appreciated.4. The use of higher initiator concentrations will result in an increase in the number of formed radicals. at relatively lower GMA intakes. These results are in line with earlier work [15.). 6.
50 45 40 35 30
25 20 15 10 5 0 0. 15. The FD values are within a rather narrow range.4. 17]. Such behavior is slightly in contrast to what observed for maleic anhydride [13. GMA=24% mol/mol-CL units GMA. Synthesis and Properties of Reactive Interfacial Agents for Polycaprolactone‐Starch Blends
The experimental trend for DEM is different. doubling the amount of initiator results in considerable higher product FD. Effect of the BPO intake on the product FD (130°C.2
BPO Intake (mol% on CL units)
0.2. Such discrepancy is probably related to the relatively low intake of DEM as well as to the different reactivity of DEM compared to MA .8
0. constant substrate to PCL ratio) For high GMA to CL ratio (24%‐mol/ mol CL units).4
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 6.1
1. no detectable influence of the BPO amount on the FD is observed (Figure 6. GMA=12% mol/mol-CL units DEM.3. This will lead to a higher proportion of PCL radicals by hydrogen abstraction from the polymer backbone and thus to higher FD values. However.3
0. DEM=30% mol/mol-CL units
1. the BPO and monomer intakes are considered as independent parameters. as well as several transfer reactions might be responsible for the observed trend .2.2. 6. In particular the occurrence of “cage effects”.6022nm ni − 0. ni the molar amount of initiator in the feed and δ the difference in solubility parameters calculated using group contributions . 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. namely an increase in the BPO intake results in higher product FD. i. δ )
where nm is the molar amount of monomer in the feed. The model provided in eq.8875 + 0.3. 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).3.)
. the possibility of side reactions will become significant.e. the mutual solubility of DEM in PCL and DMA in PCL was included in the model by using a parameter δ.Chapter 6
is an optimum ratio between the BPO concentration and available monomer (GMA in this case) on the FD. In addition. To quantify synergic effects of monomer and initiator molar intakes on the FD of GMA and DEM on the PCL backbone. ni . 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). defined as the difference in solubility parameters between PCl and the substrates. This phenomenon is likely related to the different nature of the radical grafting mechanism of GMA and DEM on PCL as mentioned before. The effect is however much less pronounced than for GMA. 6.5431niδ + 1.3. This leads to the following equation:
FD = f (nm .0325nmδ + 0.2729nm niδ
(6. 17]. gives the best description of the experimental data:
FD = −1. a statistical model has been developed by performing a multivariable linear regression on the data reported in Table 6. 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. Here. 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.1.3. and 6. recombination of (macro) radicals. If many macroradicals are created in the system (relatively high BPO intakes) at low GMA intake.
b.8 0.358 1909. b. Table 6.908 10.5.957) and its closeness to the adjusted R2 (0. Contour plot.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 ). respectively. 3D plot.4 BPO Intake (mol% on CL units) 0.2 10
25 30 35 40 GMA Intakes (mol% on CL units)
a..358 DF 4 8 12 MS 456. The R2 value for the model (0.8 0.
Figure 6. Graphical representation of the FD model for PCL‐g‐GMA.
1 0. The very low P‐value implies that the model is statistically significant.7 0.
Model Error Total
Analysis of variance for the FD model provided in eq 6.3
0.6 0.3.941) also suggests that all important variables have been included in the model.6
60 FD (%) 40
20 0 1 0.3. and Figure 6.5.2 10 20 GMA Intake (mol% on CL units) 40 30 50
SS 1828 81. the interaction between (PCL‐monomer) intake and solubility parameter difference ( nmδ ) and the interaction between the latter factor and the initiator intake ( ni δ ).17 F 44.3.4 0.5 0.9
BPO Intake (mol% on CL units) 0. Graphical representation of the FD models for PCL‐g‐GMA and PCL‐g‐DEM are given in Figure 6. Synthesis and Properties of Reactive Interfacial Agents for Polycaprolactone‐Starch Blends
The analysis of variance data are given in Table 6. This was also confirmed by inspection of the residue distribution by a normal probability plot (not reported here for brevity) .
2 0 1 0.7
6 FD (%) 4
0. Graphical representation of the FD model for PCL‐g‐DEM. et al  working with PCL‐g‐GMA. Thermal properties of the compatibilizers The thermal properties (Table 6. and a favored nucleation
1 0. the crystallization temperature and enthalpy are not a clear function of the FD values.4 BPO Intake (mol% on CL units) 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.3. This induces irregularities and is expected to result in a lowering of the Tc and the Tm.1.6 0.
b. the melting temperature and the relative enthalpy decrease with respect to pure PCL. 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. From a more practical point of view. Remarkable is once more the different trends for the two substrates. although both properties are significantly lower than those of pure PCL. Furthermore. For all samples.8 0.
Figure 6. Such changes in the thermal properties compared to pure PCL may be caused by the introduction of grafts on the PCL chains. radical initiator and PCL intakes). Similar observations have also been made by Kim. 3D plot.9
BPO Intake (mol% on CL units) 0.4 0.4. which is expected to lead to a Tc reduction.) of the compatibilizers were determined by DSC.2 10 20 DEM Intake (mol% on CL units) 40 30 50
0. and 6.5 0. 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. Contour plot. for DEM a clear transition is observed. a. While for GMA an increase in BPO or monomer intakes invariably leds to higher FD values.8 0. 6.1. a higher BPO intake leads to a reduction of the FD values while an opposite trend is observed at lower DEM intakes.6.2 10
25 30 35 40 DEM Intake (mol% on CL units)
a. For relatively high DEM intakes (> 30 % mol/mol).
3 380.6 168.8 372.2 337.6 11.8 384.8 329.4.7 10.0 9.1 11.9
E (MPa) 270.3
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 total of 12 blends were prepared: a series of binary ones (not containing any compatibilizers) constituting our reference points. 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.9 368.0 430. 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.1 13.5 425.2 342. Table 6.4.6%.0 401.5
ε (%) 640.0 332. 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) 16.5 357. 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).4 431. PCL‐g‐GMA has an FD of 9.2 9.3 15. 6.6 305. 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.5 7.8 371. An overview of all prepared blends together with their thermal and mechanical (tensile tests) properties is given in Table 6.6 5.9 424.2 343.2 321.0 341.4 230.2.7 %.
.3 10.9 11.9 379.6 386.5 489.1 10.3.
Our experimental design allows comparisons of thermal and mechanical properties as function of the starch content for binary blends (no compatibilizer) and intake of PCL‐g‐DEM and of PCL‐g‐GMA.
).7. 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. 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.43 MPa and 640. The thermal and mechanical properties indicate no or very little interaction of the starch particles with the PCL ○ : ∆Hc □ : ∆Hm
.7.1 MPa and 341. 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.
75 70 65 60 55 50 45 40 35
∆ H (J/ g-PCL)
10 15 20 Starch content (%)
Figure 6.Chapter 6
6.8 MPa) clearly indicates that the rigidity of the blends increases at higher starch contents.2.8 % at 30 % starch content. Melting and crystallization enthalpy as function of the starch content for binary blends with PCL. The simultaneous increase in the modulus (from 270.3.5 % for pure PCL to 7. This is most probably due to a nucleation effect of the starch on PCL as observed also for other kinds of polymer blends .2 MPa up to 341. 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 .1. respectively from 16.
3.) 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.
c.). The presence of these voids was also observed for sago starch/PCL blends [25.). This is visually confirmed by the morphology of the prepared blends (Figure 6. for which a 6 oC drop is observed when using PCL‐g‐DEM.e.
Figure 6. The latter trend is also valid for the corresponding enthalpies: i. Synthesis and Properties of Reactive Interfacial Agents for Polycaprolactone‐Starch Blends
matrix. However.2. In this respect the thermal properties remain substantially unchanged with the exception of the Tc.8.4. the starch particles are simply inserted into voids of the PCL matrix. Morphology of starch/PCL binary blends. Ternary blends compatibilized with PCL‐g‐DEM PCL‐g‐DEM. a decrease with respect to pure PCL is observed and then a substantial invariance as function of the PCL‐g‐DEM content. Evaporation of the water during blend preparation. either by heating or in combination with mechanical stress . and the lack of interface adhesion cause void formation. Such behavior is consistent with the hypothesis that the starch particles hinder the melting and crystallization processes of the PCL chains. This is confirmed by examination of the blend morphology by SEM (Figure 6.2. 6. 26] and explained by assuming that the voids are formed by water in the starch. the corresponding binary blend (S/PCL 20/80) contitutes a better reference point compared to pure PCL. As consequence one might expect a slight different structure of these ternary blends as compared to the corresponding binary one.8.
b. was used as compatibilizer for the preparation of ternary blends with starch and PCL. in order to fully understand the role of PCL‐g‐DEM. (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.
. whose synthesis is described for the first time in this work. Thus. The thermal behavior is characterized (Table 6.9.
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. (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
.9.). 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. which will hinder the flowability and fibre forming ability of PCL matrix under cold drawing.10. As a result of the different morphology also the mechanical behavior display differences with the virgin PCL and the binary blend. 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.
a. resulting in more rigid material with higher modulus.9.
Figure 6. Morphology of ternary blends compatibilized with PCL‐g‐DEM.). This can be explained by the improved interfacial adhesion  between PCL and starch. Indeed starch particles in the ternary blends display a smoother interface with the PCL matrix compared to the corresponding binary blends (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 6.10.
Modulus of ternary blends S/PCL/PCL‐g‐DEM as function of PCL‐g‐ DEM amount.3. at all PCL‐g‐GMA contents) there is a clear increase of the modulus with respect to the
.10. In particular also in the case of PCL‐g‐GMA.
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).e. by taking as reference the corresponding binary blend (S/PCL 20/80). 6. The first comparison as function of the PCL‐g‐GMA intake (ternary blends with starch/PCL ratio of 20/80) results in quite similar considerations as the ones made for PCL‐g‐DEM. Synthesis and Properties of Reactive Interfacial Agents for Polycaprolactone‐Starch Blends
390 380 370 360 E (MPa) 350 340 330 320 0 1 2 3 PCL-g-DEM amount (wt%) 4 5
Figure 6. the stress and strain at break are lowered and hardly a function of the PCL‐g‐GMA amount. in all cases (i.3. 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). the crystallization temperature experiences a drop of about 5‐6 oC. the modulus is higher and hardly a function of the compatibilizer amount. However.2.
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.
Figure 6. 28. one might speculate that the longer length of the GMA grafts compared to DEM (vide supra) plays a negative role in the in situ formation of the compatibilizer. As a result.7 % for PCL‐g‐DEM. 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.Chapter 6
binary blend. Morphology of ternary blends S/PCL/PCL‐g‐GMA.11. PCL‐g‐DEM is at least as active as compatibilizer (compare modulus values in Table 6. 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. and morphologies in Figures 6.4.11. However. As given schematically in Figure 6. partial interfacial adhesion) which render these ternary blends a kind of “intermediate” case between the binary ones and those compatibilized with PCL‐g‐DEM.11. 29].
a. This is surprising if one takes into account the different FD values: 9. 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.) than PCL‐g‐GMA for blends containing 20 %‐wt of starch. Thus. presence of voids.9.
. These differences in compatibilizing effect are not yet fully understood.8 % for PCL‐g‐GMA against 1. (left) the structure of PCL‐g‐GMA is inhomogeneous at the molecular level with long poly(GMA) branches pending from the PCL backbone. and 6.12.
12.3 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 %) . Concerning the thermal behavior (Table 6.) PCL‐g‐DEM has much less reactive groups (lower FD) than PCL‐g‐GMA.12.4.12. 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. In order to check this we compared blends with the same amount of PCL‐g‐GMA (2 %‐wt) but with different starch intakes (10. This is in agreement with the hypothesis made above (Figure 6. other groups will presumably remain available (arrows in Figure 6. Indeed. The latter result (from a modulus of 341.4. Schematic representation of the in situ reaction between functionalized PCL and starch. 20 and 30 %‐wt respectively). but since DEM is preferentially grafted as monomer.). Moreover. thus attenuating the effect discussed above. As consequence. once PCL‐g‐DEM has reacted with one –OH groups on the starch particle.). We can therefore conclude that the efficiency of PCL‐g‐GMA in the
. when comparing ternary blends (compatibilized with PCL‐g‐ GMA) with the binary ones. at higher starch contents more –OH groups would be available for reaction with the GMA groups. for all starch amounts (Table 6. On the other hand (right of Figure 6.12.8 MPa for S/PCL 30/70 up to 430.) and relating the “coverage” of the starch particle surface upon reaction with PCL‐g‐GMA with the compatibilization efficiency. Synthesis and Properties of Reactive Interfacial Agents for Polycaprolactone‐Starch Blends
PCL-g-DEM PCL-g-GMA OH O OH O
Figure 6. thus probably ensuring a better coverage of the surface. 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. the distribution of the reactive group along the PCL backbone is more “homogenous”.) for further reaction. 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.
957.4. 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).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.
6. The latter hypothesis is indirectly confirmed by the observation that PCL‐g‐GMA becomes more efficient at relatively higher starch contents in the blends. Nomenclature
E FD : initial modulus [MPa] : functionalization degree. The PCL‐g‐GMA and PCL‐g‐DEM compatibilizers display lower melting temperatures and melting enthalpies compared to virgin PCL. including statistical modeling. 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. PCL‐g‐glycidyl methacrylate (PCL‐g‐ GMA) and PCL‐g‐diethyl maleate (PCL‐g‐DEM). This represents an unreported mathematical confirmation of the fact that these kinds of functionalization reactions are mainly governed by the synergy between the different process variables and only slightly by their individual values. This is in contrast with chemical reactivity and amount of chemical groups along the PCL backbone (both factors favorable to GMA as compared to DEM) but it is explainable on the basis of the group distribution along the PCL backbone. Conclusions
A systematic study. firstly reported in this work. 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. 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]
. has been performed on the synthesis of two reactive compatibilisers. The reactive compatibilizers were used in blends of starch with PCL. The proposed model to quantify the effects of process variables (monomer and initiator intake.
1986. Polym. 227‐242. Eng. Mani. H. Albertsson. 157. R. B. Polym. Liu.B. Rodriguez‐Gonzales. . Ogawa. 1998..G. 1209‐1217. Ramsay. Bhattacharya: Properties of injection moulded starch/ synthetic polymer blends – IV. . Wurzburg). 34. Kawasaki: Biodegradable films made from low density polyethylene (LDPE). Polymer 2003. Polym. Polym. Eur. R. Adv. wheat starch and soluble starch for food packaging applications. USA. . 34. 1998. . 1477‐1487. Psomiadou. M. A‐C. 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. J. Mullen. Lindblad.S. Y. Sci. E. Ind. . References
. Karlsson. W. 139‐161. 38. 1467‐1475. 1251‐1253. E. Part 2. 1998. 1997. O. S. Polym. Sci. A‐C. J. B. Ranucci. 2002. Effect of amylopectin to amylase ratio in starch.W. N. S. CRC Press. 44. Karlsson: Polymers from Renewable Resources. 34. Jarowenko: Acetylated starch and miscellanous organic esters.J. M. C.6. 33. in Modified Starches: Properties and Uses (Ed. Davis: High performance LDPE/ thermoplastic starch blend: a sustainable alternative to pure polyethylene. . F. J. Biliaderis. Boca Raton. Eur. A.D. Bhattacharya: Properties of injection moulded starch/ synthetic polymer blends – III. 1942.Chem. Albertsson: Biodegradable Polymers and Environmental Interaction. E.
. Thermal and morphological properties. 1517‐1526. Mani. M.A. . Inc. Carbohydr. Ioannis.Eng. Pacsu: Starch Studies: Preparation and properties of starch triesters.
29. J. Bhattacharya. It.C. Tang. J.K. P. E. 61‐67. Chem.H. R. Environ. 37. 1139‐1148. M. Polym. 41. G.Y. Polym. Painter: Specific Interactions and the Miscibility of Polymer Blends. Polym. M. John. 2003. 1999. Polymer 2001. J. E. C. L. Polym. J. Guegan.H. Polym. 3875‐3881. . Errico.K. J. 81. Sci.Y. Chem. Yang. Jung. Bhattacharya: Synthesis and characterization of anhydride‐functional polycaprolactone. 127.M. Fringant: Properties of thermoplastic blends: starch‐polycaprolactone. . Mater.M. New York. D. 2001. 80. Chiellini. 5590‐5598. . 2001. 3. A. Sci. Gazz. P.W. 5135‐ 5142. Park: Modification of Aliphatic Polyesters and Their Reactive Blends with Starch. Macromolecules 1996.K. M. Coleman. C. 1995. C. Park: Grafting of glycidyl methacrylate onto polycaprolactone: preparation and characterization. 305‐313. Kim. Sci. Sci. L. 42. Cho. R. Part A: Polym. 179‐187. USA. M. Stab.. Wu: Physical properties and biodegradability of maleated‐ polycaprolactone/ starch composite. Rustgi: Biodegradable polymers. Goldberg: A review of the biodegradability and utility of poly(caprolactone). Ruggeri. Mani. Cho.E. . R. Degrad. Averous. 1273‐1335. USA. Laurienzo. Solaro: Biodegradable polymeric materials. . 35. . Chandra. Dole. Adv. P. Polymer 2000. Environ. J. Park: Reactive blends of gelatinized starch and polycaprolactone‐g‐glycidyl methacrylate. J.
. Moro. 12. J. 1996. P. J. Avella. 8.K. 1693‐1702. . Kim. . M. M.F. C. Prog. Degrad. J. Rimedio: Preparation and characterization of compatibilised polycaprolactone/ starch composites. K. 1507‐ 1516. J. Kim. 127–134.Chapter 6
. Martuscelli. Marrucci.. Part A: Polym. Pennsylvania.S. P. . Khandpur. Passaglia. Montgomery: Design and Analysis of Experiments 5th edition.H. 23. Marechal. C. E. . Technomic Publication. 1998. A. 4157‐4167. . Graf. Aglietto: Chemical reactions affecting the free‐radical grafting of diethyl maleate onto ethylene polymers. K. 1991.S. Polym. . Nakayama. 1997. Raimo. Appl. . Polymer 2000. M. 91‐95. D. T. J. C. 2004. M.. R. Polym. C. 1997. Kim. John Wiley & Sons Inc. Z. K. J. R. Tang: Functionalization of polyesters with maleic anhydride by reactive extrusion. Chim. Macosko. Inoue: Compatibilizers for melt blending: Premade block copolymers. 41. .
Ishak: Mechanical properties and enzymic degradation of thermoplastic and granular sago starch filled poly(ε‐ caprolactone). A. Polym. U.M. Manan. 877‐884. Ishiaku.A. E. . Nachtigall. Test. 2003. Polym.W. Polym. Appl.
. W. Sci. R. 393‐401. O. B. 273‐281. Mauler: Comparative study of PP‐MA and PP‐GMA as compatibilizing agent on polypropylene/nylon 6 blends. Kammer: On mechanical properties of sago starch/ poly(ε‐caprolactone) composites.H.Nasirtabrizi: Ring opening reactions of glycidyl methacrylate copolymers to introduce bulky organosilicon side chain substituents. Pang. 40. F. J. M. 293‐304.S. K. Synthesis and Properties of Reactive Interfacial Agents for Polycaprolactone‐Starch Blends
. 38. Aglietto.M. . . 87.S. Odusanya.N. 2006. 2002. H. U.B.M.S. Azemi: Effect of starch predrying on the mechanical properties of starch/ poly(ε‐caprolactone) composites. Sci. Manan. .N.A. 9. Polym. Technol.S. Eur.M. Tedesco. 57. G. M.D. B. S.S.S. Polym.V. K. Eng. 2000. J. D. O. Barbosa. 2002. Picchioni: Formation and compatibilizing effect of the grafted copolymer in the reactive blending of 2‐ diethylsuccinate containing polyolefins with poly‐ε‐caprolactam (nylon‐6). Ishiaku. 11‐15. .W. Lee. S. U. R. Ishiaku. Azemi. Adv.M. D.Safa. Odusanya. Passaglia. Ruggeri. Polym. 1298‐1305. 1998. M. Bull. 21.
Plastics made from fossil resources (a. Starch is a polymer consisting of anhydroglucose (AHG) units. containing periodic branches linked with the backbones through α‐D‐(1. Its production has increased significantly since 1950. There are several starch modification methods available.o oil and gas) are very attractive materials for a broad range of applications. Starch is also too sensitive to water. Starch‐based biodegradable materials are considered interesting candidates to replace certain types of conventional plastics. with a rate of almost 10% every year. it has a high volume to weight ratio and is generally resistant to degradation. After use. Examples are the use as packaging and construction materials. 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. such as thermoplasticization. The use of virgin starch for packaging materials is restricted because it cannot be shaped to films with adequate mechanical properties (high percentage elongation. chemical modification or combinations thereof. blending with other materials. 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 content of amylose and amylopectine in starch varies and largely depends on the starch source. For instance.4)‐glucosidic bonds. plastics may end up in the environment and cause serious environmental problems. Starch is relatively cheap and available from a broad range of plants. The application of a number of prospective methods for developing starch‐based biomaterials will be described in this thesis. 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. In combination with the current high prices for petrochemical products. There are two types of polymers present in starch: amylose and amylopectin. we have
. tensile and flexural strength). while amylopectin is branched polymer. there is a strong need for renewable alternatives for plastics from fossil resources. Amylose is essentially a linear polymer in which AHG units are predominantly connected through α‐D‐(1. 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 must therefore be modified before it can be applied as a biodegradable plastic. To gain insight in the potential of this approach.6)‐glucosidic bonds.
21‐0. Considerable amounts of ε‐CL homopolymers with isopropyl end‐groups were also formed.72. and particularly acetic acid derivatives (C2) have received considerable attention.3. 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 DS of the PCL chains was between 0. temperature and monomer: monosaccharide ratio) on the average degree of polymerization. The polymerizations were performed at 60‐100°C. depending on the ε‐CL to starch ratio.2. Silylated starch with a low to medium DS (0. Poly‐(ε)‐ caprolactone grafted silylated starch co‐polymers with average chain length of 40‐ 55 monomer units (molecular weight of 4500‐6300) were obtained. Esterification of starch with carboxylic acid derivatives is one of the oldest strategies to improve starch properties. The use of higher carboxylic (fatty) acid to esterify starch resulted in products with significantly improved
. but the products are still very brittle. even in the presence of plasticizers. The results are described in detail in Chapter 3. The yield of the polymers was a function of the reaction temperature and the reaction time. The grafting of ε‐caprolactone to the silylated starch by a ring‐opening polymerisation catalysed by Al(OiPr)3 was performed in THF at 50oC. The grafting efficiency for the desired reaction was 28‐58%. with higher ratios leading to higher molecular weights. 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 silylation reaction was performed using hexamethyl disilazane (HMDS) in DMSO/ toluene mixtures at 70°C.4‐di‐O‐isopropylidene‐α‐D‐ galactopyranose) using Al(OiPr)3 as the catalyst.4‐di‐O‐isopropylidene‐α‐D‐galactopyranose ratio (at constant Al(OiPr)3 intake). 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. with higher temperatures (100°C) leading to lower yields. and off‐white solid products with isolated yields of 30‐96% were obtained.Summary
initially performed research on the ROP with a simpler model system involving a simple protected mono‐saccharide instead of starch.2. are described in Chapter 2. A statistical model has been developed to quantify the effects of process variables (time. The synthetic procedure may be divided into three steps: hydrophobisation of starch by the introduction of SiMe3 groups followed by in situ ROP. 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. Most of the previous starch esterification studies involved the use of short chain carboxylic acids (C1‐C4).68) was obtained.46‐0.3. The hydrophobicity of starch acetates is higher than virgin starch.
The yellowish products were characterized by 1H‐. High‐DS products (DS= 2. an increase in the vinyl‐ester concentration led to higher product DS values.96 and P‐value of ≤10‐7) second‐order mathematical models correlating the effect of process variables (vinyl to AHG‐ starch mol ratio. and the type of catalyst. K2CO3. the synthesis of two polycaprolactone (PCL)‐based reactive interfacial agents. In Chapter 5.26‐2.5 for starch stearate). For low vinyl‐ester to starch ratios. a systematic study. The thermal stability of the esterified products is higher than that of native starch. K2CO3. 13C‐NMR and FT‐IR. K2CO3 and Na‐acetate are superior with respect to activity when compared with Na2HPO4.7‐3. Summary
mechanical properties and hydrophobicities. Statistically adequate (R2≥0. In Chapter 6. Mechanical tests show that the products have tensile strength (stress at break) between 2. PCL‐g‐
.39) are totally amorphous whereas the low‐DS ones (DS= 1.4 could be obtained for both laurate and stearate esters.24 ‐ 2. products with a DS > 2. In Chapter 4‐5. the DS decreased.5 MPa. The synthesis of these fatty starch esters is. and catalyst basicity) to the DS of the starch ester products were developed. presumably due to a reduction of the polarity of the reaction medium. catalyst intake. The thermal and mechanical properties of some representative product samples are also described. an alternative method for higher fatty starch esters by using fatty vinyl esters is reported. 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.45‐1.96). usually performed using fatty acid chlorides which are relatively expensive and rather corrosive. The starch esters were synthesized by reacting gelatinized starch with vinyl laurate or vinyl acetate in DMSO in the presence of basic catalysts (Na2HPO4. and Na‐acetate) in DMSO at 110°C. and Na‐ acetate). At higher ratios.07‐1.9 for starch laurate. 1. The starch esters were prepared by reacting starch with vinyl laurate or vinyl stearate in the presence of basic catalysts (Na2HPO4. and modulus of elasticity of 46‐113 MPa. With these catalysts. The products have a broad range of degree of substitution (DS = 0. The use of Na2HPO4 resulted in low‐medium DS products (0. reaction temperature. elongation at break of 3‐26%. The DS of the products was a clear function of the chain length of the fatty ester.4‐3 for starch stearate). while the use of K2CO3 and CH3COONa catalysts resulted in medium‐high DS products (2. Reaction temperature and catalyst intake also affect the product DS but only to a lesser extent.3‐1. however.75) are still partially crystalline. The DS of the products is a strong function of the basicity of the catalyst.1‐2. vinyl ester to starch ratio.5 for starch laurate. The use of methyl and glyceryl esters results in products with only relatively low DS. The last part of this thesis deals with reactive blending of starch and biodegradable polymers using reactive interfacial agents. 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.
At a fixed starch content (20 %‐wt). The latter hypothesis is indirectly confirmed by the observation that PCL‐g‐GMA becomes more efficient at relatively higher starch contents in the blends. mutual solubility of the monomer into the molten polymer) on the functionalisation degree (FD) of the GMA and DEM units to the PCL backbone. In particular the elastic modulus for the compatibilised blends is significantly higher (reaching up to 430 MPa) than that of uncompatibilised ones (320‐340 MPa).Summary
glycidyl methacrylate (PCL‐GMA) and PCL‐g‐diethyl maleate (PCL‐g‐DEM). 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 FD values for PCL‐g‐DEM were considerably lower (up to 7 %). the mechanical behavior of the compatibilised blends is in general different from that of pure PCL and of the corresponding uncompatibilised blends. at relatively high GMA and BPO intakes. A statistically adequate model (R2=0. and their use in compatibilising starch‐PCL blends is described. As the result. PCL‐g‐DEM seems to have slightly more efficient compatibilising effect than PCL‐g‐GMA as shown by blends morphology and elastic modulus values. Both of the PCL‐based compatibilisers improve the interfacial adhesion between the starch particles and the PCL matrix in starch/ PCL blends. P‐value ≤10‐9) has been developed to quantify the effects of process variables (monomer and initiator intake. 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.
. Highest values of the FD (up to 45 %) were observed for PCL‐g‐GMA.957.
De productie van kunststoffen is sinds 1950 met bijna 10% per jaar gestegen. Het gebruik van zetmeel voor verpakking materialen is momenteel beperkt. in combinatie met de huidige hoge prijzen voor de grondstoffen. 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. Het zetmeel is daarnaast ook erg gevoelig voor water. menging met andere materialen. Zetmeel is een polymeer bestaande uit anhydroglucose (AHG) eenheden.4)‐glucoside bindingen verbonden zijn. Twee type polymeren zijn dominant aanwezig in zetmeel: amylose en amylopectine. De verhouding amylose en amylopectine in zetmeel varieert en hangt grotendeels af van de plant waaruit het zetmeel gewonnen wordt. Modificatie van zetmeel kan het aantal toepassingen sterk vergroten. Voorbeelden zijn thermo‐ plastificatie. De synthese van dergelijke producten met een in situ katalytische ringopening polymerisatie (ROP) van geschikte monomeren als b. Dit proefschrift beschrijft een aantal methodes om nieuwe polymere materialen op basis van zetmeel te maken. De toepassing van een alternatieve methode waarbij zetmeel eerst minder hydrofiel gemaakt wordt door de introductie van grote hydrofobe SiMe3 groepen. chemische modificaties en combinaties van deze methoden. ε‐caprolactone op de hydroxy‐groepen van zetmeel is helaas niet eenvoudig. Na gebruik kunnen de materialen in het milieu eindigen en grote problemen veroorzaken. Amylose is een lineair polymeer waarin de AHG eenheden hoofdzakelijk via α‐D‐(1. Biologisch afbreekbare polymeren op basis van zetmeel kunnen mogelijk bepaalde conventionele kunststoffen vervangen.Samenvatting
Kunststoffen gemaakt van fossiele grondstoffen als olie en gas zijn zeer aantrekkelijke materialen voor een breed scala aan toepassingen. Het amylopectine is een vertakt polymeer waarbij de vertakkingen via α‐D‐(1. heeft geleid tot een grote interesse voor biologisch afbreekbare alternatieven gemaakt uit hernieuwbare grondstoffen. mede omdat het niet mogelijk is om films met goede mechanische eigenschappen te maken.6)‐glucoside bindingen met de hoofdketen zijn verbonden. Zetmeel is relatief goedkoop en te winnen uit een breed scala aan planten. Er zijn verschillende methodes beschikbaar om zetmeel te modificeren. gevolgd door de ROP met een polyester precursor en verwijdering van de SiMe3 groepen
. De AHG eenheden kunnen op een aantal manieren aan elkaar gekoppeld worden. De belangrijkste redenen zijn de watergevoeligheid van de katalysatoren en het feit dat zetmeel slecht of zelfs onoplosbaar is in de typische organische oplosmiddelen die voor de ROP gebruikt worden.v. Dit.
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/1.3.4‐di‐O‐isopropylidene‐α‐D‐galactopyranose (bij constante i Al(O Pr)3 concentratie). DS = 0.21‐0. 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 DS van de PCL ketens lag tussen de 0. De grafting van ε‐ caprolactone aan het gesilyleerde zetmeel werd in THF bij 50°C uitgevoerd met Al(OiPr)3 als katalysator. De silylerings reactie werd bij 70°C uitgevoerd met hexamethyl disilazane (HMDS) in DMSO/tolueen mengsels. De resultaten voor het modelsysteem waar de ROP van p‐dioxanone wordt geïnitieerd door de hydroxy‐groepen van een beschermde mono‐saccharide (1.2. 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.2. gevolgd door in situ ROP met ε‐ caprolacton en daarna de verwijdering van de silyl groepen.
. Een hogere verhouding leidde tot hogere molecuul massa’s. De polymerisaties werden tussen 60‐100°C uitgevoerd en leverde vaste producten in geïsoleerde opbrengsten tussen de 30 en 96%. De substitutiegraad van de SiMe3 groepen varieerde van laag tot middelhoog (degree of substitution.3. De resultaten worden in Hoofdstuk 3 in detail beschreven. De grafting efficiency van de gewenste reactie was 28‐58%. De kennis verkregen met het modelsysteem is toegepast om ε‐caprolactone te graften op zetmeel. waarin hogere temperaturen (100°C) tot lagere opbrengsten leiden. 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%).Samenvatting
zou in theorie moeten leiden tot een hogere grafting efficiency. Er werden ook aanzienlijke hoeveelheden homopolymeer van ε‐ caprolacton met isopropyl eindgroepen gevormd.68). De gevolgde procedure bestaat uit drie stappen: hydrofobisering van het zetmeel door de introductie van SiMe3 groepen.72 en was afhankelijk van de ε‐caprolacton/zetmeel verhouding.4‐di‐O‐isopropylidene‐α‐D‐galactopyranose) met Al(OiPr)3 als katalysator worden in Hoofdstuk 2 beschreven. De opbrengst van de polymeren was een functie van de reactietemperatuur en de reactietijd. temperatuur en monomeer: monosaccharide verhouding) op de gemiddelde polymerisatie graad zijn gekwantificeerd met een statistisch model. Om inzicht in het potentieel van deze benadering te krijgen is onderzoek verricht naar de ROP met een eenvoudig modelsysteem bestaande uit een beschermde mono‐saccharide in plaats van zetmeel. De effecten van procesvariabelen (tijd.46‐0.
De esterificatie van zetmeel met carbonzuur derivaten is één van de oudste strategieën voor het verbeteren van zetmeeleigenschappen. wordt een alternatieve methode voor de synthese van zetmeel esters met lange vetzuren beschreven. terwijl het gebruik van de K2CO3 en Na‐acetaat katalysatoren
. De effecten van belangrijke procesvariabelen (vinyl ester/zetmeel molverhouding. vermoedelijk als gevolg van een verlaging van de polariteit van het reactiemengsel.3‐1. De substitutiegraad varieert van 0.5 voor zetmeel stearaat). Het gebruik van methyl en glyceryl esters resulteert in producten met een lage substitutiegraad. Echter de producten zijn nog steeds zeer bros.96 en P‐ waarde van ≤ 10‐7). In de meeste zetmeel esterificatie studies wordt gebruikt gemaakt van kleine carbonzuren (C1‐C4). Het gebruik van Na2HPO4 resulteerde in producten met een laag tot middelhoge DS (0. wordt een systematische studie. In Hoofdstuk 5. De thermische en mechanische eigenschappen van sommige representatieve producten zijn uitgebreid bepaald. inclusief statistische modellering. In Hoofdstuk 4‐5.5 voor zetmeel laureaat. de vinyl‐ ester/zetmeel verhouding en het type katalysator. De synthese van deze esters wordt in het algemeen uitgevoerd met vetzuurchloriden die vrij duur en sterk corrosief zijn. Het gebruik van hogere (vet‐) zuren resulteerde in producten met beduidend betere mechanische eigenschappen en hydrofobiciteit. De acetaat esters zijn in het algemeen minder hydrofiel dan zetmeel. reactietemperatuur.07‐1. De zetmeel esters werden bereid door het zetmeel met vinyl laureaat of vinyl stearaat in aanwezigheid van basische katalysatoren (Na2HPO4. 13C‐NMR en FT‐IR gekarakteriseerd. K2CO3 en Na‐acetaat vertonen superieure activiteiten vergeleken met Na2HPO4. De producten werden met 1H‐. De zetmeel esters werden bereid door zetmeel met vinyl laureaat of vinyl stearate in DMSO in de aanwezigheid van basische katalysatoren (Na2HPO4. K2CO3. Met deze katalysatoren worden voor zowel de laureaat als stearaat esters producten met een DS van boven de 2. Echter. De DS van de producten is een sterke functie van de basiciteit van de katalysator.4 verkregen. 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. bij verdere verhoging van de verhouding verminderde de DS. en in het bijzonder azijnzuur derivaten (C2). zelfs in aanwezigheid van plastificeermiddelen. 0. naar de synthese van lange vetzuur esters uit maïszetmeel (zetmeel laureaat en zetmeel stearaat. katalysator concentratie en katalysator basiciteit) op de DS zijn gekwantificeerd met een statistisch model (R2 ≥ 0. De DS van de producten was een duidelijke functie van de keten lengte van de vetzure vinylester. De temperatuur en de katalysator concentratie beïnvloeden de DS ook maar in mindere mate. K2CO3 en Na‐acetaat) in DMSO bij 110°C te laten reageren. en Na‐acetaat) te laten reageren.24 tot 2.96. Bij lage vinyl‐ester/zetmeel verhoudingen leidde een verhoging van de vinyl‐ester concentratie in eerste instantie tot hogere DS waarden.
en een elasticiteit modulus van 46‐113 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 hoogste FD waarden (tot 45%) werden gevonden voor PCL‐g‐GMA bij relatief hoge GMA en BPO concentraties. een uitrekking tot breuk van 3‐26%. PCL‐g‐glycidyl methacrylaat (PCL‐GMA) en PCL‐g‐diethyl maleaat (PCL‐g‐DEM).4‐3.75 nog gedeeltelijk kristallijn zijn. gebruikt.26‐2. 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. Als reactieve compatibilisers zijn twee polycaprolactone derivaten.
. 1. Beide reactieve compatibilisers verbeteren de interactie tussen de zetmeeldeeltjes en de PCL matrix.1‐2.7‐3. Zo is bijvoorbeeld de elastische modulus van de gecompatibiliseerde mengsels beduidend hoger (tot 430 MPa) dan die van niet‐gecompatibiliseerde mengsels (320‐340 MPa). Mechanische tests tonen aan dat de producten een treksterkte hebben variërend van 2. De synthese van deze verbindingen wordt uitgebreid beschreven.39) zijn volledig amorf terwijl de producten met een DS van 1.5 MPa.9 voor zetmeel laureaat. Dit leidt tot verschillen in mechanische eigenschappen tussen gecompatibiliseerde mengsels en van de niet‐ gecompatibiliseerde mengsels van zetmeel en PCL.45‐1.Samenvatting
resulteerde in producten met een middelhoog tot hoge DS waarden (2. Er is een statistisch model (R2 =0. De thermische stabiliteit van de geësterificeerde producten is hoger dan die van zetmeel.957 en P‐ waarde van ≤ 10‐9) ontwikkeld waarmee de effecten van procesvariabelen (monomeer en initiator concentratie. De FD waarden voor PCL‐g‐DEM waren aanzienlijk lager (tot 7%). Het laatste hoofdstuk van dit proefschrift (Hoofdstuk 6) behandelt de menging van zetmeel en biologisch afbreekbare polymeren met reactieve compatibilisers.0 voor zetmeel stearaat). De producten met een hoge DS (2.
and for helping me arranging many things when I was not able to do it from Indonesia. Prof. 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. Broekhuis. These include the current high oil prices. as well as taking the SEM pictures. I want to thank my three promoters. This thesis is a summary of my research on developing novel biodegradable polymers from starch.A. I want to thank Marya van der Duin‐de Jonge. Mosiscki. and Prof. I want to thank Erik Heeres for being more than a good supervisor. I want to thank Leon Janssen for his hospitality. for her hospitality and solving administrative matters. I want to thank Francesco Picchioni for being always there when I needed him to discuss anything. for upgrading my writing quality. for the inspiring ideas during the discussions. together with difficulties in managing plastic waste. Minnaard. and Prof. since the mechanical properties as well as hydrophobicity of this starting material is relatively poor. L. and Laurens Bosgra for being always helpful with technical issues and Jan Henk Marsman for the help with analytical equipment. Erwin Wilbers. Gert Alberda van Ekenstein and Harry Nijland helped me a lot with mechanical and thermal property measurements. which improved the quality of this thesis.J. for your consistent support. Erik Heeres. 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. and for always encouraging me to finish this thesis. Thank you Erik. the raw material for fossil derived plastics. not in the last place because of the problems related with conventional plastics. 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. He has also been a good friend and always supported me during the hard times. A. Leon Janssen. Thank you very much for spending your valuable time reading my thesis and for the constructive comments and corrections you have made. Prof. for valuable discussions during my early PhD period and for handing me over to the right hands. who was my first daily supervisor for about two months. Marcel de Vries. especially on his opinions about the statistical modelling and mechanical properties.Acknowledgements
The synthesis of biodegradable materials is currently receiving a lot of attention. Anno was also always there when
. Although it is indeed a very interesting topic. Anne Appeldoorn. People from other groups have also been very kind and supportive to my research. I also want to thank Ineke Ganzeveld. A. First of all. Prof. Prof. Francesco Picchioni.
Groningen is also a place where I found true friendship. master students who worked with me. and many others for the good old times! Thank you Tante Ietje. Diana Santangelo. Nadia Gozali. Vincent Nieborg and Fesia Lestari. our memories will never fade away. Kuslan. Anna Nizniowska. and on many occassions. Wuri Raspati. for your love. Ardiyanti. Hofman. Sari Fabianti. A lot of thanks for Henk Stegeman and Poppy Sutanto. mbak Ika. Annette Korringa. Jelle Wildschut. visiting researchers. in the lab. Yao Jie. The chemical and glassware magazijn as well as the FWN library personnel were always patient. for your never‐ending blessings. Ronny Sutarto. Tante Oppie. Abdul Osman. and assistant professors have shared the days with me in Groningen. Ita. Albert Kieviet helped me with the MALDI‐TOF measurements. Endy. and although we seldom met afterwards. Arjan Kloekhorst. Boelo Schuur. kak Roga. Bima and Wisnu (also my paranymph). mbak Mia. Wangsa. Teddy. Many master and PhD students. I have met a lot of Indonesian students in Groningen. Samuel. Henky Muljana (my paranymph) and Anindhita Widyadhana. Thanks also for Marcel and Connie. Ais Jenie. Patrick and Oppie.Acknowledgements
I had questions or needed specific equipments. and helpful. Dado. C. August Kurniawan. My special thanks are for Laura Junistia and Jan‐Pieter Drijfhout. Mahfud. Zhang Youchun. and passion. Tante Caroline. Tante Smith. friendly.B. Chalid. Buana Girisuta and Rina Karina. M. for everything. Mita. Teguh. and Tante Alma for being like my parents and family in Groningen. Iwan Harsono. Puspita Kencana. Jasper Huijsman. Erna Subroto. Marcel Wiegman. and the ISD staff helped me a lot with visum and residence permit arrangements as well as tax papers. Thank you Tresna. patience. Yongki. Niels van Vegten. Our collaboration has also become part of this thesis. and Henk van de Bovenkamp for cheering my days and being my friends during summer and winter. Francesca Gambardella. R. for never leaving me alone. Taufik. support. mas Pandu. Ignacio Melian Cabrera. Tenny.
. spring and autumn in Groningen. Hans Heeres. Nidal Hammoud Hassan. Asal Hamarneh. Tiara. Francesca Fallani. Anneke Toxopeus. Judy Retti Witono. Thank you my parents and sister. Farchad H. in the office. Diana Jirjis. and for my wife and family. Oscar Rojas. Togi. Louis Daniel. Wita Sondari. Jaap Bosma. Rasrendra. bu Ida Susanti. Egi. and Iwan Kustiawan for becoming my family. pak Harry. Agnes R. I want to thank Edwin A. Gerard Kraai. kak Atha. Thank you AW. Thyara. Thank you God. Wahyudin. and Mahesa for being my brothers and sisters in Groningen. You all have always been my inspiration and motivation. Taufik.
F. Picchioni. India. Sugih. Sugih. Janssen. L. 8. A. 9. Bratislava (Slovak Republic).K. A.P.J. Sugih.Picchioni.K. H. Pune Chapter.B. Sugih. H. Drijfhout. A. presented in 11th Asia Pacific Confederation of Chemical Engineering (APCChE) Congress. H. H. Junistia. A. Appl. 2008 (submitted to J. 3. Mumbai (India). Polym. F.M.Heeres. Heeres.) 6. and H.K. Heeres. Kuala Lumpur (Malaysia). A. 2. A. presented in “Polymers for Advanced Technologies”.J.P.).A. A. Sugih. H. Sugih. Junistia. Heeres. L. Heeres: Synthesis of higher fatty acid starch esters using vinyl laurate and stearate as reactants. Junistia. Picchioni: Biodegradable polymer materials based on polyesters and starch/proteins.K. South China University of Technology. Broekhuis.Drijfhout. R.010). J. 4. Picchioni. 5.P.M. and F. Experimental Studies on the ring opening polymerization of p‐dioxanone using an Al(OiPr)3‐monosaccharide initiator system.K.P.J. L. Guangzhou (China). (in press. J. Janssen. Sugih. Sci.10.B. A. Sugih. 2008 (submitted to Carbohydr. 27‐30 August 2006 (oral presentation) 7.J. Picchioni. and F. Broekhuis. L. Polym. F. L.K. 2008 (accepted for publication in Starch‐Starke).J. L. Experimental and modeling studies on the synthesis and properties of higher fatty esters of corn starch. starch‐based innovative product: The synthesis of grafted polycaprolactone‐starch by ring opening polymerisation. DOI: 10.K.J. presented in the 1st International Symposium on Sustainable Chemical Product and Process Engineering.List of Publications
1. 17‐20 December 2006 (oral presentation). J. L. F.P.A. 2008.B.B. Synthesis of poly‐ (ε)‐caprolactone grafted starch co‐polymers by ring‐opening polymerisation using silylated starch precursors. Eur. F. L. A. J. 2008 (accepted for publication in Starch‐ Starke).J. Heeres. Polym. F. 25‐28 September 2007 (oral presentation). Manurung.1016/j. Drijfhout.eurpolymj. Heeres. A. R.K. Picchioni. H.K.P.M Janssen and H. Sugih.P.Janssen. Picchioni.2008. F.M. Picchioni. Heeres: Green. presented in International Workshop: “From Polymer Modification to Multicomponent Polymer Systems”. Manurung.J. and H. Junistia.
. Picchioni: Functionalization and blending of polycaprolactone.J. 26‐28 November 2006 (oral presentation). A. 9th National Conference organized by the Society for Polymer Science. Synthesis and properties of reactive interfacial agents for polycaprolactone‐starch blends. Heeres: Synthesis and properties of novel green biopolymers derived from starch and fatty acids.