Foaming of Ethyl Hydroxyethyl Cellulose

Foaming of Ethyl Hydroxyethyl Cellulose Effects of chain length shortening by Marc Carrillo Aguilera Master Thesis No. 175/2016 at Department of Materials and Manufacturing Technology Chalmers University of Technology Gothenburg, Sweden Master Thesis as an Erasmus student Performed at: Department of Materials and Manufacturing Technology Chalmers University of Technology, SE - 412 96 Gothenburg Supervisor(s): Kristina Karlsson Department of Materials and Manufacturing Technology Chalmers University of Technology, SE - 412 96 Gothenburg Leif Karlson Akzo Nobel Functional Chemicals AB Hamnvägen 2, SE - 444 85 Stenungsund Mats Stading SP - Food and Biotechnology Frans Perssons väg 6, SE - 402 29 Gothenburg Examiner: Mikael Rigdahl Department of Materials and Manufacturing Technology Chalmers University of Technology, SE - 412 96 Gothenburg Foaming of Ethyl Hydroxyethyl Cellulose Effects of chain length shortening Marc Carrillo Aguilera ©MARC CARRILLO AGUILERA 2016 Diploma work no 175/2016 Department of Materials and Manufacturing Technology Chalmers University of Technology SE-412 96 Gothenburg Sweden Telephone + 46 (0)31-772 1000 Cover: a short text explaining the picture (if used) Typeset in L A T EX Chalmers Reproservice Gothenburg, Sweden 2016 Foaming of Ethyl Hydroxyethyl Celullose (EHEC) Effects of chain length and degradation MARC CARRILLO AGUILERA Department of Materials and Manufacturing Technology Chalmers University of Technology Abstract The current depletion of petroleum resources together with environmental issues have led to new approaches in plastic manufacturing. This trend involves using ecofriendly materials coming from renewable resources. Good candidates for this, due to their properties and availability, are the cellulose derivatives. Some of them, such as hydroxypropyl methylcellulose (HPMC), showed in previous studies a promising behavior when making polymeric foams. Unfortunately, the corresponding results with methyl ethyl hydroxyethylcellulose (MEHEC) were not as promising and the rather high molecular weight of the used MEHEC grade was here believed to be one of the important factors, affecting the foaming ability. Therefore, in this report, the effect of chain scission and thus the chain length on the foaming behavior of MEHEC was studied. The reduction in molecular weight was achieved through an addition of a degrading agent and salt to aqueous solutions of the original cellulose derivative (MEHEC). After drying, the resulting polymers were characterized, both in the solid state e.g. in the form of films, and as aqueous solutions. This characterization was focused on the rheological behavior. Techniques such as rotational rheometry and dynamic mechanical thermal analysis (DMTA) were used for this purpose. The results pointed to a decrease in molecular weight when adding encreasing amounts of the degrading agent. However, the addition of salt seemed to have an opposite effect, and an increase in viscosity was observed for increasing amounts of salt. A hot-mold process and density measurements were used for assessing the foaming ability. The reduction in molecular weight allowed aqueous solutions with up to 20 and 25 %polymer content to be used in the foaming process (the optimal concentration for the original MEHEC was 5 %). A higher polymer content would be beneficial with regard to large scale processing, like extrusion, one of the most common polymer processing techniques. Additionally, shrinkage was observed in some of the foams and was believed to be associated with the salt content, affecting the foaming ability negatively. Finally, the molecular weight was determined by size-exclusion chromatography (SEC). These SEC-results confirmed the reduction in molecular weight of the MEHEC-grade used. Keywords: methyl ethyl hydroxyethylcellulose, cellulose derivatives, Foaming ability, DMTA, Rheological properties, Tensile tests, Shortening of chain length. iii Acknowledgements There are a number of people without whom this thesis might not have been possible, and to whom I am really grateful. First, I would like to express the deepest gratitude to my supervisor Kristina Karlsson, for always being willing to help and available to answer all my doubts. It is unnecessary to say that you have been indispensable to overcome all the obstacles. My sincere thanks to my examiner, Mikael Rigdahl, for all the valuable discussions we have shared together. Many thanks to Leif Karlson, Mats Stading, Linda Härdelin, Marco Berta and Waqas Qazi for helping me with the laboratory equipment and for more than valuable advice. Even being far from here, I would like to thank my fabulous family and girlfriend for being an inexhaustible source of encouragement and support throughout writing this thesis and my life in general. Finally, I would also like to thank everybody that, in one way or another, have helped me during my Master Thesis. You all have enriched my stay in the wonderful country of Sweden. Marc Carrillo Aguilera, Gothenburg, June 2016 v Contents Abbreviations ix List of Figures xi List of Tables xiii 1 Introduction 1 1.1 Objective/Aim .............................. 2 2 Theoretical Background 3 2.1 Cellulose derivatives (CDs) . . . . . . . . . . . . . . . . . . . . . . . . 3 2.1.1 Ethyl hydroxyethyl cellulose . . . . . . . . . . . . . . . . . . . 4 2.2 Foam characterization . . . . . . . . . . . . . . . . . . . . . . . . . . 4 2.2.1 Glass transition temperature (Tg)................ 4 2.2.2 Storage and loss modulus . . . . . . . . . . . . . . . . . . . . 5 2.2.3 Viscosity.............................. 6 2.2.4 Viscoelasticity........................... 7 2.2.5 Cloudpoint............................ 7 2.2.6 Size-exclusion chromatography(SEC) . . . . . . . . . . . . . . 8 2.3 Hot-moldfoaming............................. 9 2.4 Rheology.................................. 10 2.4.1 Rotational rheometer . . . . . . . . . . . . . . . . . . . . . . . 10 2.4.2 Dynamic Mechanical Thermal Anlysis (DMTA) . . . . . . . . 11 3 Materials 13 3.1 Water ................................... 13 3.2 Sodium sulfate (Na2SO4)......................... 13 3.3 EHEC ................................... 13 3.4 Sodium hypochlorite (NaClO) . . . . . . . . . . . . . . . . . . . . . . 14 4 Experimental procedure 15 4.1 Chain shortening of cellulose derivatives . . . . . . . . . . . . . . . . 15 4.2 Viscosity measurements using TA instruments AR-G2 . . . . . . . . . 17 4.3 DMTA measurements using RSA II . . . . . . . . . . . . . . . . . . . 18 4.3.1 Linear Viscoelastic Region (LVR) . . . . . . . . . . . . . . . . 19 4.3.2 Glass transition temperature (Tg)................ 19 4.3.3 Baking simulation . . . . . . . . . . . . . . . . . . . . . . . . . 19 4.4 Foaming with hot-mold machine . . . . . . . . . . . . . . . . . . . . . 20 4.5 Size-Exclusion Chromatography (SEC) . . . . . . . . . . . . . . . . . 20 4.6 Density .................................. 21 5 Results and discussion 23 5.1 Chain shortening of cellulose derivatives . . . . . . . . . . . . . . . . 23 vii Contents 5.2 Viscosity measurements . . . . . . . . . . . . . . . . . . . . . . . . . 24 5.3 DMTAmeasurements........................... 26 5.3.1 Linear Viscoelastic Region (LVR) and glass transition temperature (Tg) ........................... 26 5.3.2 Baking simulation . . . . . . . . . . . . . . . . . . . . . . . . . 27 5.4 Foaming with the hot-mold equipment . . . . . . . . . . . . . . . . . 28 5.5 Size-Exclusion Chromatography (SEC) . . . . . . . . . . . . . . . . . 30 5.6 Density .................................. 30 6 Conclusions 33 6.1 Chain shortening of ethyl hydroxyethyl cellulose . . . . . . . . . . . . 33 6.2 Foamingability.............................. 33 6.3 Size-Exclusion Chromatography . . . . . . . . . . . . . . . . . . . . . 33 7 Future work 35 References 37 Appendix I viii Abbreviations CAGR Compound Annual Growth Rate CD Cellulose Derivative DMTA Dynamic Mechanical Thermal Analysis DSC Differential Scanning Calorimetry EHEC Ethyl hydroxyethyl cellulose EPS Expanded Polystyrene HPMC Hydroxypropyl methylcellulose LDPE Low Density Polyethylene LVR Linear viscoelastic region MEHEC methyl ethyl hydroxyethylcellulose PLA Poly (lactic acid) PU Polyurethane ix 1. Introduction ypropyl methylcellulose (HPMC) and methyl ethyl hydroxyethylcellulose (MEHEC). 1.1 Objective/Aim In previous studies, five samples of HPMC and four of EHEC were analyzed in order to evaluate their foaming ability [6]. The first derivative showed promising foaming properties whereas only one EHEC sample did (MEHEC), and even so not obtaining as good results as the HPMC. One of the factors that could be causing the different behavior is the molecular chain length. Therefore, the main goal of this thesis is to validate or non-validate such a hypothesis. Initially, some shortening of the chain length of MEHEC materials is necessary in order to reduce the molecular weight. Thereafter, observations of the influence on the chain length on the foaming behavior will be made. In this context, measurements of the polymer solutions, determination of the foam density as well as evaluations of the molecular weight of the MEHEC-specimens are of importance and interest. 2 2 Theoretical Background The aim of this section is to provide basic knowledge to the reader for a better representation and understanding of the results which will be subsequently discussed. 2.1 Cellulose derivatives (CDs) The biodegradable polymers made from renewable resources have received increased attention of polymer scientists, plastics manufacturers and government agencies among others. Most of the interest in biodegradable plastics is focused at developing low cost materials that are economical in high-volume applications. Some important applications, such as packaging and consumer products, are good examples to illustrate that. Starch-based materials have been of particular interest because of its low cost [7]. As starch, poly (lactide acid) or polylactide (PLA) is a biodegradable and biocompatible polymer produced from renewable resources (cornstarch and sugarcane). It is a thermoplastic aliphatic polyester that is synthesized through a ring-opening polymerization of lactide and lactic acid monomers. During the last decade this material has been extensively studied due to its potential for replacing fossil-based polymers. Such features as being green and biodegradable, as well as the non-release of toxic components during its manufacture, endow PLA with all a material needs to become a very interesting candidate[8]. Following the same pattern, some studies are devoted to manufacturing foam from cellulose, the most abundant and widely used organic polymer in the world. After a chemical treatment cellulose can be made water-soluble. The water-soluble cellulose ethers have many applications in different areas of industry. Some of them are listed below[9]. •Water-borne paints. CDs are used as thickening and rheology modifiers. Additionally, they take an active part in the particle stabilization in the paint. •Construction. The main objective of the cellulose derivatives in this field is to increase the water retention. •Pharmaceutical formulations. They can be used as one component of the excipient or as a coating layer on the outside of the tablet. •Food. Cellulose ethers are able to retain moisture evaporation, thus they are used for extending the shelf life of bread and other baked goods. They are also used as substitutes for gluten. 3 2. Theoretical Background Although there are already many applications for CDs in general, HPMC and MEHEC have a special interest due to its potentially good properties to form different kinds of foam. 2.1.1 Methyl ethyl hydroxyethylcellulose MEHEC is a water-soluble cellulose derivative in which both ethyl and hydroxyethyl groups are attached to the anhydroglucose units by ether linkages. Alkali, ethylene oxide and ethyl chloride are needed to manufacture MEHEC from cellulose. Figure 1: Schematic representation of the chemical structure of MEHEC. Figure adapted from [6] Although MEHEC already has other applications in fields such as painting, this study will be focused on the manufacturing of foams from this material. 2.2 Foam characterization Attributes related to the molecular structure of polymers can make them more or less successful for processing. Unlike metals or ceramics, polymeric materials often consist of very long chain-like macromolecules. For this reason, it is important to know the different material properties defining the polymeric material before studying it. 2.2.1 Glass transition temperature (Tg) One of the fundamental parameters describing polymers is the glass transition temperature, Tg. The material behaves in an increasingly brittle manner if the temperature drops below Tgwhereas its behavior is more rubber-like above this temperature. Materials which are formed by long chains, networks of linked atoms, or those having a complex molecular structure are the ones exhibiting a glass structure. When such materials (which usually have a high viscosity) undergo a rapid cooling from above Tgto a temperature below that, the structure cannot reorganize itself into a crystalline structure. During this cooling the molecular movement is too sluggish, or the molecular structure too awkward, to take up a crystalline conformation. 4 2. Theoretical Background Consequently, the random arrangement distinctive of the liquid persists down to temperatures where the viscosity is so high that the material is considered to be solid. The term glassy is recognized as a permanent non-equilibrium state and in some cases even a way to achieve a state of lowest energy might not be available. Despite this scientific significance of glass forming-materials, the glass transition itself is poorly understood. In particular, it is not clear whether the glass transition is a thermodynamic transition [11] or a purely kinetic phenomenon [12]. Figure 2: The temperature dependence of the tensile modulus E. The glass transition temperature Tgas well as the melting temperature Tmare indicated. As can be seen in figure 2, there are three regions clearly differentiated. Bellow Tg, there is a temperature range where the material exhibits a more or less constant value of the tensile modulus E, followed by a decrease of Ewhen Tgis approached. Once this temperature is exceeded a second plateau is reached and the last region corresponds to the melting temperature (if the polymer is semi-crystalline). Using dynamic mechanical thermal anylisis (DMTA) technique (which will be explained in more detail in the next sections), Tgcan be found. Tgis determined by the large drop in the chart log(E’)/Temp (at the same time, in the tan(δ) curve, a peak can be seen). 2.2.2 Storage and loss modulus When a sinusoidal strain is applied to the sample, two components are defining the corresponding stress. These components represent the elastic part of the response (also called storage modulus and represented by E’ when a tensile strain is applied or G’ for shear strain) and, on the other hand, the viscous part (or loss modulus, represented by E” or G” depending on the kind of strain applied). The value of these parameters depend on the phase δ(phase shift between the stress and the strain on the sample). For an ideal elastic material δ=0◦(energy stored) and δ=90◦ for a liquid viscous material (energy dissipated). The usual way to represent them 5 2. Theoretical Background is from a right-angle triangle, having tanδas a quantifier of the balance between energy loss and stored energy. Therefore, if the material is more viscous the value of tanδis greater than 1 and lower when the material has a more solid behavior. The parameter tanδcan also be understood as the phase difference between the stress and the strain applied. The ratio E”/E’ or G”/G’ (which is tanδ), is also a quantifier of the internal friction of the material. 2.2.3 Viscosity In short, the viscosity quantifies fluids resistance to flow. It is represented by the symbol “eta” and it is calculated as the ratio between the shear stress and the velocity gradient. η=F A÷ ∂v ∂y !(2.1) Although the equation shown above is the general expression, the viscosity is normally represented by Newton’s equation. This equation states that the resulting shear ratee in a fluid is directly proportional to the force applied and inversely proportional to its viscosity. f A!=η ∂v ∂y !(2.2) The unit used by the International System of Units to define viscosity is the pascal second [Pa s]. Nevertheless, another common unit is the dyne second per square centimeter also known as Poise referring to the French physiologist Jean Poiseuille (1799-1869). Since ten Poise equal one Pa·s, the conversion of units is straightforward. What has been discussed up to now is also known as the dynamic viscosity. Another expression for viscosity is sometimes used. It is called kinematic viscosity and it is the ratio between viscosity of a fluid and its density: ν=η ρ(2.3) Viscosity describes the resistance to flow at a particular temperature. A high or low internal resistance to flow will define high and low viscosities respectively. Normally, longer chain molecules lead to higher viscosity (such as plastics). As a general rule for polymers, the viscosity decreases when increasing the temperature (the opposite happens with gases). When measuring the viscosity, it is usual to represent it in a chart log(η)/shear rate. It is easy to distinguish two basic parts in the characteristic curve for polymers: 6 2. Theoretical Background first a plateau at low shear rates and, after a threshold value of the shear rate, the viscosity decreases. 2.2.4 Viscoelasticity The classical theory of elasticity considers mechanical properties of elastic solids following Hooke’s law. The deformation is proportional to the applied strain. On the other hand, the hydrodynamic theory understands the properties of viscous liquids according to Newton’s law, that is, the resulting strain rate is proportional to the applied stress. Both are idealized models, many solids follow Hooke’s law when the deformation is small and many liquids follow Newton’s law when the deformation rate is low. Viscoelastic materials combine both behaviors, thus the strain can partly recover when the stress is released. For Newtonian fluids, viscosity ηis a material constant and not dependent of the rate of deformation. For elastic materials the strain is directly proportional to the stress with a proportional factor called shear modulus G(for shear) or Young’s modulus E(for tension), and the elastic energy is stored in the substance when deforming it. The strain is totally recoverable after the load is removed, provided that the limit for onset of plastic deformation was not exceeded. For purely viscous materials all the energy is dissipated in the form of heat, thus the recoverable strain is zero [13]. Polymeric materials are in general non-Newtonian since they consist of long molecules that entangle with each other, creating flexible and reversible “joints”. The polymeric behavior is Hookean and Newtonian, making polymers viscoelastic materials. Moreover, to have a more straightforward characterization of the materials it is convenient to test them in the linear viscoelastic region (LVR) where the viscoelastic properties are independent of applied stress or strain levels. Due to the complex behavior of viscoelastic materials and in order to study the flow properties of foams, rheology will be used. This science, from the Greek reos (flow) and logos (study) as well as defined by Heraclitus as panta rei (all things flow), is useful to describe such properties. 2.2.5 Cloud point Cloud point measurements is a common method to determine phase boundaries and phase separation [14]. At temperatures above the cloud point the polymer is no longer completely soluble in the fluid, giving a cloudy appearance. In order to obtain good mixtures, i.e. a polymer solution, the temperature during the mixing process should be above the cloud point. One way to ensure that the mixing takes place above a certain temperature could be by adding some extra chemical (e.g. NaCl, Na2SO4, CaCl2, etc.) as has previously been done with some non-ionic surfactants [15][16]. At the same time, one has to be careful and ensure 7 2. Theoretical Background that the chemicals added are not negatively affecting the reaction since it could affect the expected outcome. 2.2.6 Size-exclusion chromatography(SEC) The concept of separation by size using chromatography was first discussed by Synge and Tiselius [18]. They observed that small molecules could be excluded from the small pores of zeolites depending on their molecular size [19]. The first examples of SEC were introduced by Wheaton and Bauman [20] in their work on ion-exclusion chromatography. Chromatographic techniques separate molecules by taking advantage of the distribution of the different components between two parts: the steady phase and the mobile phase. SEC is a liquid type of chromatography where the steady part is solid and the mobile is liquid. This technique is also called Gel Permeation Chromatography (GPC). Its principal application is molecule separation according to the size in order to study the molecular weight and the polymer distribution. The material which fills the column is acting as a sieve and different molecules can be distinguished by size [21]: •Permeable molecules: small molecules which go slowly through the porous phase since they are withheld in the pores. •Divisible molecules: molecules which enter partially in the pores. •Excluded molecules: the molecules bigger than the pores which can go through the column faster than the smaller ones. The sieve, which represents the steady phase, is made by a porous material and varies depending on many variables such as pore’s size, temperature, accuracy, range of molecular weight, etc. The pore size, together with the geometry, restricts access of molecules based on their Stokes’ radius. Consequently, the largest polymers, which are excluded from the pores, leave the column first, that is, polymer elute in order of decreasing size. But there are also other factors which affect the measurements such as shape or molecule linearity. Nevertheless, this technique allows the use of calibrated curves, derived from a set of known analytes, to be used to estimate the molecular weight of an unknown analyte. Typical calibration curves are based on polymers of known molecular weights. The mobile phase (fluid) has to drag the dissolved polymers without reacting neither with it nor with the steady phase. At the end of the column is a device which calculates some features of the particles leaving the system. Light scattering sensors and refraction index measurements are examples of such devices. Below, a schematic figure of a whole system can be seen. 8 2. Theoretical Background Figure 3: Schematic representation of a system used for SEC. Due to the range of chain lengths that is at hand in a polymer, its molecular weight can not be defined with a single value. Therefore, the molecular weight of a polymer must be described using an average taking into account all the different chain lengths of the sample. The most common measures of the average molecular weights are Mn(number average) and Mw(weight average) and their definitions are: Mn=PNiMi PNi Mw=PNiM2 i PNiMi (2.4) The ratio between Mwand Mnis called polydispersity index. It is used as a measure of the broadness of molecular weight distribution of a polymer. 2.3 Hot-mold foaming Although there are many foaming techniques such as baking, freeze-casting or by extrusion among others, just the first one, which is also called compression molding, will be used during this master thesis. This technique has been intensively investigated for processing starch-based polymers, particularly foamed containers [22]. Therefore it was considered a good option when testing the foaming ability of the cellulose derivatives. Not only is it an easy way to foam the different samples but also a relatively fast technique. The process principally consists of applying pressure on the sample, which is poured 9 2. Theoretical Background in a preform or mold, by means of two plates that can be heated individually. Normally, the test material needs to be mixed with an external blowing agent which creates the foam when it is activated either due to temperature, pressure or both. For example, when water is the blowing agent, due to its evaporation over 100 ◦C the bubbles created give the foam structure. Figure 4: Schematic picture of hot-mold foaming. 2.4 Rheology Rheology is the science of flow and deformation of materials. A strain or strain rate creates a deformation and indicates the distance that a certain material element moves under the influence of an external force, or stress. As has been mentioned before, the behavior of materials under stress or strain varies from purely viscous to pure elastic (or a combination of both, viscoelastic materials). This behavior is reflected in measurable properties such as modulus, viscosity and elasticity. Many materials present a complex rheological behavior. Material processability as well as end-use performance are determined by the rheological properties. This shows how important rheological measurements are for the whole range of different materials including ceramics, petroleum products, elastomers, pharmaceuticals, etc [23]. Rotational rheometers as well as DMTA instruments are commonly used equipment for evaluating the rheological properties and will be described further in the next sections. 2.4.1 Rotational rheometer Most commercial rheometers measure the viscoelastic properties using a rotational geometry. There are three very common geometry set-ups, although the most used are the parallel plates and the cone-plate. These two set-ups, together with the one with concentric cylinders, are shown in Figure 5. The experimental procedure to 10 2. Theoretical Background obtain the results is similar in all cases. Depending on the type of material being evaluated, one or the other set-up should be used: concentric cylinders for very low to medium viscosity, cone-plate for very low to high viscosity and plate-plate for low viscosity to soft solids. Moreover, most commercial rheometers use a mode called simple shear deformation. In this mode, one surface is fixed whereas the other is rotated (either with a strain or a given strain rate). The result obtained from the rheometer can be a shear stress if the strain is being controlled or the strain if the shear stress is being controlled. Figure 5: From left to right: A schematic display of the rheometer gap fixtures; cone-plate, plate-plate and concentric cylinders. 2.4.2 Dynamic Mechanical Thermal Anlysis (DMTA) The first attempts to do oscillatory experiments to characterize the elasticity of a material was made by Poynting in 1909 [25]. The utilization of oscillatory experiments is today a basic procedure to measure properties of polymers, such as viscosity or storage and loss moduli and at the same time taking into account the influence of temperature. Rheology, has a reputation of being difficult to understand, involving a fair degree of mathematical sophistication [26]. Nevertheless, dynamic mechanical thermal analysis is more straightforward. For this reason, it is widely used to characterize material properties as a function of different variables such as temperature, frequency, stress or a combination of them. During the test, a small deformation is applied to the sample, with a known geometry, being studied. Either controlled stress or strain can be applied on the subjected sample. The stress applied in a DMTA is sinusoidal and depending on the stiffness the sample will deform a certain amount. One of the concerns has always been the stabilizing bearing of the shaft due to the fact that it is activated by a force motor and it is not always easy to hold it in position. This technique allows to measure parameters such as modulus, tan(δ) and glass transition temperature and it is also helpful to find the LVR (linear viscoelastic region). These parameters change with temperature and transitions in the sample can be seen due to the temperature control. The stress applied is sinusoidal, thus, the modulus can be expressed by the storage and loss components. 11 4. Experimental procedure 0,1 1 10 100 0,01 0,1 1 Viscosity (Pa·s) shear rate (s -1 ) 0,01 0,1 1 10 100 0,01 0,1 1 Viscosity (Pa·s) shear rate (s -1 ) Figure 8: Example of two viscosity charts obtained using a rotational rheometer ARG2 from TA Instruments Ltd. Moreover, to ensure a better accuracy of the results, each measurement was repeated twice and four times if the results were ambiguous. Apart from the better accuracy, they were repeated to see if any drastic change in viscosity took place over time. Fortunately, the variable time did not show a negative effect on the measurements. 4.3 DMTA measurements using RSA II Three different experiments were carried out with Rheometrics Scientific RSA II. Each one of them had a different objective. The first one was used to establish the Linear Viscoelastic Region (LVR). The second one was to determine the glass transition temperature. The last set of experiments was a baking simulation to observe the evolution of the foam phases with temperature. For the first two tests, films are required and, for the third one, solutions. The recommendations for the film dimensions were defined as follow (according to the manual of the instrument): 35 mm length, width up to 12.7 mm (6 mm optimal) and thickness up to 1.5 mm and not thinner than 0.01 mm. Due to the use of a Petri dish to obtain the film, it was easy to fulfill the first two dimensions and for the thickness a simple calculation as shown in equation 4.1 was necessary. Initially, a 0.1 mm thickness was chosen. dry film thickness =wet film thickness x polymer content(%) 100 (4.1) For the solutions, deionized water was heated to 90◦C before the powder was added. A higher temperature favors the mixing. Once the solution was homogeneous it was let to cool down before being used. 18 4. Experimental procedure 4.3.1 Linear Viscoelastic Region (LVR) In order to find LVR, a strain sweep was applied as part of a tensile test. The different samples (films with the according dimensions) were placed in the clamps and the tensile test could start. Once the test was finished, a chart representing storage and loss moduli and tanδas a function of the strain applied, was obtained. Apart from being an important parameter to define the material behavior it was also measured due to the fact that it was needed for the temperature sweep. Another fact that needs to be taken into consideration when measuring LVR is the static force since a tensile test is performed. It is necessary to maintain a value greater than the lowest force level reached in the dynamic oscillation, if not, buckling will occur and part of the stress signal is truncated and data quality is negatively affected. 4.3.2 Glass transition temperature (Tg) In order to determine an approximate value of the Tgand study the temperature dependence of the viscoelastic behavior of the different samples in a dry state, a ramp temperature test was carried out (oscillatory measurements at a frequency of 1Hz, heating rate from 5 to 7 ◦C/min and using a stable value of the strain(%) from the LVR obtained in previous tests). The films were placed in the clamps in the same way as before and the output was a chart representing the temperature dependence of the storage and loss moduli and tanδ. A first approach to find the glass transition temperature was through the definition - Tgis the temperature at which the log(E’) - temperature curve starts to decrease substantially. This was done by drawing two lines in the log(E’) curve, one from the beginning of the flat part and the other one following the slope where the substantial decrement takes place. The approximation of Tgcan be found drawing a vertical line from their intersection. 4.3.3 Baking simulation The configuration with two parallel heated plates was used for the baking simulation. It allowed the study of different phases as the solution was heated and also the evolution of the changes in stiffness while the temperature increased. Apart from the normal simulations where the oven was closed during all the time, some tests were repeated opening the oven every 20 ◦C in order to take a photo and follow the course of events visually. 19 4. Experimental procedure 4.4 Foaming with hot-mold machine As has been noted earlier, this is a uncomplicated technique which is, taking into account previous studies, suitable for foaming. During the experiments, the concentration of polymer was changed from 5 to 25%depending on the amount of NaClO used and the viscosity. First, 10%solutions were used and depending on the results, the amount of polymer was decided. Here, water was use, both as plasticizer and blowing agent. The mixture was poured on one of the plates, filling the preform made of aluminum foil (circular profile with a diameter between 25 and 30 mm). The hot-press machine used for baking was a Franz Haas Waffelmaschinen together with a temperature control system (Figure 9). The temperature of the plates was varied from 140 to 220 ◦C and the molding times from 8 to 15 minutes. No additional pressure, except for the plate’s weight, was applied, thus, the external agent was activated due to the elevated temperature (above water evaporation temperature) and time. Figure 9: Press-hot machine used for the "baking" experiments. 4.5 Size-Exclusion Chromatography (SEC) These experiments were carried out at the Department of Chemistry and Chemical Engineering, Chalmers. The different solutions made by mixing miliQ water and the cellulose derivatives with shortened molecules were poured into plastic vials. The set of samples was placed in a Waters 717 Plus Autosampler, where the experiments were performed. Sample is taken from each vial and, after being filtered twice, arrives at the column. Next, the size of the particles are analyzed by the unit Wyatt DAWN Heleos II and Optilab T-REX. All the data gathered by the last unit is sent to the computer system where data processing takes place, and the results are 20 4. Experimental procedure obtained, i.e. the molecular weight distribution. 4.6 Density In order to determine the density of the produced foams two different methods can be considered. Depending on the shape of the resulting foam, one method or the other will be appropriate. The solutions are placed in cylindrical preforms made of aluminum foil inside a “baking” machine in order to obtain cylindrical shape. It permits a further comparison between different samples since the same procedure is followed in all cases. If the foam has filled the preform and completely adapted the cylindrical shape it is easy to determine the volume and hence the density (it is just necessary to weigh it). This method is faster but the approximations made when measuring the dimensions can drastically affect the accuracy of the result. If the formed foam is far from the mold shape the density can be measured using a sand displacement method. It consists of measuring the volume when the sample is immersed in a measuring cylinder full of sand. Dividing the weight by the volume, the density is obtained. Although it is a longer procedure, it might be more accurate than the previous one since geometrical shapes can be difficult to measure with sufficient accuracy. 21 4. Experimental procedure 22 5 Results and discussion In this section, all the results obtained from the experiments will be shown and discussed. 5.1 Chain shortening of cellulose derivatives The amounts of chemicals used for lowering of the molecular weight of MEHEC are given in the following table. Table 2: Summary of chemicals used in each experiment. Experiment Na2SO4(g) NaClO(g) %wt(NaClO) 1 100 15 2 2 50 30 4 3 0 30 4 4 20 60 8 5 20 7.5 1 6 20 15 2 7 20 0 0 8 0 7.5 1 9 20 30 4 10 20 0 0 11 0 7.5 1 12 20 3.75 0.5 13 20 15 2 Some of the amounts were fixed for the whole set of experiments, for example 1000 g of water and 100 g of M10. The main reason why these amounts were fixed is because the study was focused on how the molecular weight of M10 was reduced and sodium hypochlorite had the greatest effect in this respect and therefore the one that should be changed. Following a similar approach, the temperature was set to 93◦C and 30 minutes of duration for each experiment. Fixing these four variables the number of experiments were limited, making it easier to evaluate the effect of the variable parameters. Addition of sodium sulfate reduce the cloud point of the MEHEC and is used to simplify the stirring and mixing in the reactor and the filtration after the reaction. To find the optimal amount of sodium sulfate required to give both good mixing and filtering, the first four experiments were focused on optimizing this amount. Both 100 g and 50 g resulted in easy mixing and filtering. In contrast, some problems were 23 5. Results and discussion detected especially with filtering when no sodium sulfate was added. Addition of 20 g sodium sulfate resulted in good mixing and acceptable filtering and was selected as the optimal addition. After the amount of sodium sulfate had been determined, the only remaining degree of freedom was the amount of sodium hypochlorite. To determine how much of this chemical was needed, the experiments were combined with viscosity measurements (the corresponding results and discussion will be given in the next subsection). It was decided that five different amounts of sodium hypochlorite would be added: 0.5, 1, 2, 4 and 8 weight-%(%wt) (plus one reference experiment without addition of sodium hypochlorite). As can also be seen in the table 2, three experiments were repeated. Number seven and eight had either low content of salt or degrading agent and the first time they were carried out, the filtering time was not adequate. As a result, after they were dried in the oven, the product obtained did not meet the expectations of being separable into a powder form again. Nevertheless, the second time (experiment number 10) a powder suitable for further tests was obtained. Experiment number six was done twice to evaluate the repeatability. The quality of the powder (how easily it could be obtained again after the procedure and how fine it was) also deserves a comment since many properties are evaluated from solutions and a finer powder means better mixing and hence quality of the solutions. The main findings from the experiment was that the more salt that was added, the easier it was to obtain the final product in the form of a fine powder. Consequently, number one and two showed the best performance in this respect and the rest of the samples exhibited similar performance. 5.2 Viscosity measurements Viscosity tests were carried out using 2 %solutions and the values, which were obtained at a shear stress of 1 Pa, are given in table 3. For the experiments with higher contents of sodium hypochlorite, the tests were repeated with 5%solutions but the results did not differ from the first trials. In general, when the level of sodium hypochlorite was increased, the viscosity decreased. As can be seen in figure 10(a), for low contents of sodium hypochlorite, the reduction is not very pronounced whereas for 2 %and 4 %the change in viscosity was more drastic. Between 4 and 8 %only a small change can be observed. 24 5. Results and discussion Table 3: Viscosities of the polymer solutions for the different experiments including the original M10. The polymer concentration was 2 ·%wt. Experiment Viscosity (mPa·s) Experiment Viscosity (mPa·s) 1 98 8 - 2 22 9 10 3 7 10 763 4 7 11 107 5 536 12 671 6 52 13 62 7 - M10 19720 A similar trend was also observed in case of the amount of salt added. The measurements revealed that, although it favored the mixing process and a better powder was obtained, the amount of salt added affected the chain-length shortening negatively giving a higher molecular weight. Therefore, higher viscosities were obtained for the same percentages of NaClO (shown in Figure 10(b) where, for the same amount of sodium hypochlorite, 4%, the amounts of sodium sulfate were changed to 0, 20 and 50 g). 02468 1 10 100 Viscosity (mPa·s) %wt (NaClO) 0 10 20 30 40 50 5 10 15 20 25 Viscosity (mPa·s) Sodium Sulfate (g) Figure 10: Left(a): Viscosity at different levels of sodium hypochlorite (20g of sodium sulfate). Right(b): Viscosity at different levels of sodium sulfate (at 4 %wt sodium hypochlorite). Considering the effect of the salt on the viscosity, it was considered appropriate to support the result with the original M10. Consequently, 20 g of salt was added to the M10. The viscosity was expected to be slightly lower than the reference. A small lowering of the viscosity could be expected due to the sample being exposed to an elevated temperature for 30 minutes. Possibly, some interaction with the salt could also result in a lowered viscosity. However, the result was that the viscosity 25 5. Results and discussion was clearly below that of the original M10, see figure 10(a). Further studies are needed in order to find an explanation to this phenomena. Further discussion is needed for the similar results obtained for experiments three and four; figure 10(b). One explanation for this similarity could be the accuracy of the machine but this option was dismissed. The second hypothesis, the more likely one, was to assume that there exists a certain limit of sodium hypochlorite after which no lower molecular weight is achieved. To verify such a hypothesis, further studies in the matter are needed. 0,01 0,1 1 10 100 1000 0,01 0,1 1 10 100 M10 Exp10 Viscosity (mPa·s) shear rate (s -1 ) 1 10 100 1000 1 10 100 Exp3 Exp4 Viscosity (mPa·s) shear rate (s -1 ) Figure 11: Left(a): Original M10 with and without salt (Exp10). Right(b): The same for experiments 3 and 4. For experiments three and four, it was observed that the values of viscosity tended to exhibit a slight shear-thickening behavior at higher shear rates. However, that behavior is believed to be an artifact due to the low viscosity and high shear rates, the results under such circumstances could therefore not be trusted. 5.3 DMTA measurements 5.3.1 Linear Viscoelastic Region (LVR) and glass transition temperature (Tg) As Table 4 shows, the Tgdidn’t vary significantly between the specimens from the experiments listed in Table 1. Thus, changes in molecular weight did not drastically affect the softening temperature. Moreover, the values measured (between 170 and 180 ◦C) for the glass transition temperature was comparable with values reported for similar cellulose derivatives [27]. 26 5. Results and discussion Table 4: Glass transition temperatures (Tg) of the different specimens from experiments and the original M10. Experiment Tg (◦C) 1 175 2 176 3 176 4 172 5 174 6 179 9 168 10 178 12 177 13 179 M10 175 5.3.2 Baking simulation As has been discussed earlier, baking with the hot-press machine is a fast way to evaluate the foaming ability. However, in order to evaluate the different transitions occurring during the heating, a simulation using the DMTA machine is a good option. The oven was opened every 20 ◦C and a photo was taken. The result is shown in Figure 12. During the first 40 ◦C an emergence of blurry bubbles inside the solution was observed. Around 100 ◦C (evaporation temperature of water), there was a transition from a transparent to an opaque appearance (white color shade) and the foam formation took place. Then, from 120 to 160 ◦C the foam stabilizes. Finally, different stages of degradation took place (the higher temperature, the more degradation). Figure 12: Baking simulation in the DMTA equipment using experiment number 4. 27 6. Conclusions 34 7 Future work As the different tests were carried out and the properties were evaluated, new questions emerged. However, more time would be needed to look into all of them. Some of all the interesting questions that were not answered are given below and can serve as a possible guide for further studies. •During the viscosity measurements it was mentioned that the values obtained for experiment 10 (M10 plus 20 g of salt) were surprising. In order to find an explanation to this observation, the experiment should be repeated again. Doing so, it could be verified if the experiment was not done correctly or if there were other factors contributing to the result. •The baking simulations indicated that the polymer content strongly affects the result. Therefore, trying to find the optimal processing conditions for the different experiments could be interesting and valuable. •A comparison with HPMC should have been carried out. Unfortunately, it was not possible to do that, thus, it could be a part of a future study. Since the foaming ability of HPMC has been shown to be good in previous studies, it would have been interesting to analyze how a reduction in molecular weight would affect the foaming ability of this derivative and if the previous results obtained could be further improved. 35 7. Future work 36 References [1] Polymer Foam Market worth $131.1 Billion By 2018. (n.d.). Retrieved from http://www.marketsandmarkets.com/PressReleases/polymer-foam.asp [2] D. Klemm, B. Heublein, H.P. Fink, A. Bohn.Cellulose: Fascinating biopolymer and sustainable raw material. Angewandte Chemie International Edition, 44 (22), pp. 3358–3393. (2005). [3] Ray Shill. The plastics trade. In Workshop of the World: Birmingham’s Industrial Heritage. United Kingdom. The History Press. (2006). [4] M. Kjellin and I. Johansson. Surface-Active Polymers from Cellulose. In Surfactants from Renewable Resources. Wiley. p. 254. (2010). [5] K. Makuuchi and S. Cheng. Chain scission and oxidation. In Radiation Processing of polymer materials and its industrial applications. Canada: Wiley. p. 228. (2012). [6] K. Karlsson, E. Schuster, M.Stading and M. Rigdahl. Foaming behavior of water-soluble derivatives – hydroxypropyl methylcellulose and ethyl hydroxyethyl cellulose. Cellulose Volume 22, Number 4, pp. 2651-2664 (2015). [7] E.S.Stevens, A. Klamczski, G.M. Glenn. Starch-lignin foams. eXPRESS Polym. Lett, 4. pp. 311-320. (2010). [8] M. Nofar and C.B. Park. Poly(lactic acid) foaming. Elsevier Ltd. pp. 1722-1733 (2014). [9] M. Kjellin and I. Johansson. Surface-Active Polymers from Cellulose. In Surfactants from Renewable Resources. Wiley p. 253. (2010). [10] Smith B.M., Bean SR, Herald T.J. and Aramouni F.M. Effect of HPMC on the quality of wheat-free bread made from carob germ flour-starch mixtures. Journal of Food Science. 77. pp. 684-688.(2012). [11] W. Götze and L. Sjögren, Relaxation Processes in supercooled liquids. Rep. Prog. Phys. 55, p. 241 (1992). [12] J.H. Poynting, Proc. Roy. Soc. Ser. A 82:546 (1909). [13] Delay JM and Wissbrun, KF. Melt rheology and its role in plastic processing – Theory and applications. Dordrecht: Kluwer Academic Publishers. p. 16. (1999). [14] Y.C. Bae, S.M. Lambert, D.S. Soane and J.M. Prausnitz. Cloud-point curves of polymer solution from thermooptical measurements. Macromolecules. pp. 4403-4407. (1991). [15] Scott, H., Royce, E. A. and Han, K. S. Effect of inorganic additives on solutions of non-ionic surfactants. J. Colloid Interface Sci. 98. pp. 196-201. (1984). 37 References [16] Singh, D. Effect of different additives on cloud point of non-ionic surfactant. Thesis. National Institute of Technology, Rourkela. (2011). [17] P. van de Witte, P. J. Dijkstra, J. W. A. van den Berg, and J. Feijen, Phase behavior of polylactides in solvent-nonsolvent mixtures. J. Polym. Sci., Part B: Polym. Phys. 34, 2553–2568 (1996). [18] Synge, R. L. M. and Tiselius, A. Fractionation of hydrolysis products of amylose by electrokinetic ultrafiltration in an Agar-Agar Jelly. Biochemical Journal. 46. pp. xli-xlii. (1950). [19] Barrer, R. M. General and physical chemistry. III. Zeolites as adsorbents and molecular sieves. Annu. Rep. Prog. Chem. 41. pp. 31–46. (1944). [20] Wheaton, R. M. and Bauman, W. C. Non-Ionic Separations with Ion Exchange Resins. Annals of the New York Academy of Sciences. 57 (3). pp. 159–176. (1953). [21] M.C. Guitiérrez-Bouzán, A. Burdó and J. Cegarra. La cromatografía de exclusión: análisis de la distribución de pesos moleculares en siliconas por GPC. Boletín Intexter (UPC). Num 135. (2009). [22] S. Thomas, P.H. Visakh, A.P. Mathew. Advances in Natural Polymers. Composites and nanocomposites. Spinger p. 131. (2013). [23] H.A.Barnes, J.F.Hutton and K.Walters. An Introduction to Rheology. Elsevier BV. United States of America. p. 6. (1989). [24] K.P. Menard. Dynamic Mechanical Analysis. A practical Introduction. CRC. United States of America. P 13 (1999). [25] Gotro, J. (2014). Rheology of Thermosets Part 2: Rheometers (Figure). Retrieved from: http://polymerinnovationblog.com/rheology-thermosets-part-2- rheometers/ . [26] H. Lobo and J.V. Bonilla. Handbook of Plastics Analysis. CRC Press. p. 154. (2005). [27] P. Sakellariou, R.C. Rowe and E.F.T. White. The thermomechanical properties and glass transition temperatures of some cellulose derivatives used in film coating. Int. J. Pharm. 27. pp. 267-277 (1985). [28] J.A. Diosdado, H. Plasencia, M. Torres, M. de L. Mondragon. Microstructure characterization of low density EPS. Trans Tech Publiccations. Vol 420, pp. 167-176. (2013). [29] Saiz-Arroyo, C., de Saja, J.A. and Rodríguez-Pérez, M.A. Production and Characterization of crosslinked low-density polyethylene foams using waste of foams with the same composition. Polym Eng Sci, 52: 751–759. (2012). [30] Calvert K, Trumble K, Webster T, Kirkpatrick L. Characterization of commercial rigid polyurethane foams used as bone analogs for implant testing. J Mater Sci Mater Med 21(5):1453–1461. (2010). 38 Appendix I Appendix. Figure 16: All the baked samples during the first trial. II Appendix. Figure 17: All the baked samples during the second trial. III