Faculty of Health, Science and Technology Engineering Physics
30 hp
Supervisors: Niklas Kvarnlöf, BillerudKorsnäs & Björn Sjöstrand, Karlstad University Examinator: Lars Johansson
June, 2020
Pulp compositions and their influence on the production of dialcohol cellulose
Olika sammansättningar av pappersmassa och deras påverkan på produktionen av dialkoholcellulosa
Viktoria Carlsson
Abstract
The characteristics of products made from pulp can be modified through different methods. If the pulp is refined either laboratory or industrially, the fibres in the pulp become more flexible and therefore creates stronger bonds to each other, which results in a final product with a higher strength. The refining process also causes the formation of small fibre pieces that are called fines, which also contribute to the increased strength. The major component in pulp is cellulose, which can be chemically modified to materials with changed properties. Periodate oxidation of cellulose results in dialdehyde cellulose that can be further reduced with sodium borohydride to obtain dialcohol cellulose, which is a material with a higher ductility compared to regular cellulose.
In this thesis, different pulp compositions and their influence on the production of dialcohol cellulose (DALC) were investigated. The aim of the study was to find out how the ductility of paper sheets made from DALC were affected by the presence of fines in the pulp. Nine different pulp compositions were prepared for the modification: unrefined pulp, unrefined pulp with added fines, industrially refined pulp, dewatered industrially refined pulp, and pulp refined 1000, 3000, 5000, 10 000 and 15 000 revolutions with a PFI Mill.
Paper sheets were made with a Rapid Köthen sheet former and the mechanical properties of the sheets were tested with a Zwick Roell tensile tester. The surface of the sheets were analyzed using a scanning electron microscope (SEM).
The results obtained from the tensile tests showed that DALC made from unrefined pulp and DALC made from pulp highly refined with a PFI Mill, resulted in sheets with a high strain-at-break. For each increased degree of refining with the PFI Mill, the resulting DALC sheets showed an improved elongation and tensile strength. When DALC was produced from industrially refined pulp and from unrefined pulp with added fines, the resulting sheets had a lower strain-at-break. These findings indicate that the presence of fines in the pulp do have a negative effect on the ductility of the resulting DALC sheets.
Sammanfattning
Egenskaper hos produkter gjorda på pappersmassa kan modifieras med olika metoder. Genom att mala massan i en labb-, eller industrikvarn blir fibrerna i massan mer flexibla och kan därmed skapa starkare bindningar till varandra, vilket resulterar i en slutprodukt med en högre styrka. Malningsprocessen leder också till att det bildas små fiberpartiklar som kallas fines, de bidrar också till att öka styrkan hos bindningar mellan fibrerna. Pappersmassa består till störst del av cellulosa, som kan modifieras kemiskt till material med förändrade egenskaper. Cellulosa kan oxideras med natriumperjodat till dialdehydcellulosa, som kan reduceras vidare med natriumborhydrid till dialkoholcellulosa, vilket är ett material med högre töjbarhet än cellulosa.
Under det här arbetet undersöktes det hur olika sammansättningar av pappersmassa påverkar produktionen av dialkoholcellulosa (DALC). Syftet med studien var att ta reda på hur töjbarheten för pappersark gjorda av DALC påverkas av innehållet av fines i massan. Nio olika sammansättningar av pappersmassa användes för modifieringen: omald massa, omald massa med tillsatta fines, industrimald massa, avvattnad industrimald massa, och massa mald 1000, 3000, 5000, 10 000 och 15 000 varv i en PFI kvarn. Pappersark gjordes med en Rapid Köthen arkformare; de mekaniska egenskaperna hos arken testades med dragprov i en Zwick Roell och ytan hos pappersarken analyserades med ett svepelektronmikroskop.
Resultaten från dragproven visade att DALC gjord på omald massa och DALC gjord på massa mald till en hög malgrad i en PFI kvarn, resulterade i ark med en hög töjbarhet. För varje ökad malgrad med PFI kvarnen visade de resulterande DALC-arken en ökad töjbarhet och dragstyrka. När DALC producerades från industrimald massa och från omald massa med tillsatta fines erhölls ark med en lägre töjbarhet. Dessa resultat antyder att närvaron av fines påverkar töjbarheten hos de resulterande DALC-arken negativt.
Acknowledgements
This thesis work was performed at BillerudKorsnäs Technology Center during the first semester of 2020. I am very grateful to BillerudKorsnäs for giving me the opportunity to perform this project. Many thanks to everyone working at the Technology Center for making the time I spent there pleasant and for helping me with the equipment during the laboratory work. A special thanks to Niklas Kvarnlöf, my supervisor at BillerudKorsnäs, for introducing me to this field and for all the help and guidance throughout the project.
I would also like to express my gratitude to my supervisor at Karlstad University, Björn Sjöstrand, for all the support and feedback during this thesis work. Lastly, I would like to thank Christer Burman who helped me with the sample coatings and Dimitrios Nikas who helped me with the SEM equipment.
Viktoria Carlsson
Table of contents
1 Introduction ... 1
1.1 BillerudKorsnäs ... 1
1.2 Environmental aspects ... 1
1.3 Purpose of the study ... 2
2 Theory ... 3
2.1 The structure of wood ... 3
2.2 Cellulose ... 3
2.2.1 Structure ... 3
2.2.2 Properties ... 4
2.3 Chemical modification of cellulose ... 4
2.3.1 Periodate oxidation of cellulose ... 5
2.3.2 Borohydride reduction of DAC ... 6
2.4 Fibre properties ... 6
2.5 Refining ... 7
2.5.1 Fibrillation ... 7
2.5.2 Fines ... 7
2.5.3 Laboratory and Industrial refiners ... 8
2.6 Sheet forming ... 8
2.6.1 Dynamic Drainage Analyzer ... 8
2.7 Paper properties ... 9
2.7.1 Mechanical properties ... 9
2.7.2 Tensile test ... 9
2.7.3 Tensile strength and Tensile index ... 10
2.7.4 Ductility and Strain-at-break ... 11
2.8 Scanning Electron Microscope ... 11
2.8.1 The instrument ... 11
2.8.2 Electron signal ... 12
2.8.3 Contrast ... 13
2.8.4 Resolution and Depth of field ... 13
2.8.5 Sample preparation ... 14
3 Experimental ... 15
3.1 Materials ... 15
3.2 Methods ... 15
3.2.1 Refining ... 15
3.2.2 Fibre properties ... 15
3.2.3 Addition of fines ... 15
3.2.4 Dewatered industrially refined pulp ... 16
3.2.5 Chemical modification of cellulose ... 16
3.2.6 Preparation of sheets ... 16
3.2.7 Tensile test ... 17
3.2.8 SEM imaging... 17
4 Results ... 18
4.1 Pulp compositions ... 18
4.2 Addition of fines to unrefined pulp ... 18
4.3 Mechanical testing ... 19
4.4 SEM images ... 22
5 Discussion ... 24
5.1 Mechanical properties ... 24
5.2 SEM analysis ... 26
6 Conclusion ... 28
7 Future work ... 29
8 References ... 30
Appendix I ... 33
Fibre properties ... 33
Appendix II ... 35
SEM images ... 35
1
1 Introduction
1.1 BillerudKorsnäs
In November 2012 the two Swedish companies Billerud and Korsnäs merged and became BillerudKorsnäs AB, which is a leading supplier of paper and packaging solutions [1]. With eight production facilities they provide the 2000 customers in the over 100 countries with fibre-based renewable materials [2]. The several products produced by BillerudKorsnäs can be divided into the four segments; Food & Beverages, Consumer & Luxury, Medical & Hygiene and Industrial [3].
BillerudKorsnäs are constantly working towards a sustainable future, with the ambition to have a production that is completely fossil-free by year 2030. They also provide packaging solutions that are both recyclable and renewable which can reduce the use of fossil-based materials and thereby decrease the climate impact [4]. Innovation is significant for sustainability, BillerudKorsnäs, together with partners, is working on a paper bottle that is 100 % recyclable, with the possibility to replace plastic bottles in the future. Another ongoing project, which is a collaboration between BillerudKorsnäs and Uppsala university, is the paper battery that can use electrodes made from paper to store energy [5].
1.2 Environmental aspects
The climate on earth has always been changing, with varying surface temperatures, and with a number of ice ages during the last 800 000 years. But the changes occurring today are worrying since they are rapid and caused mainly by humankind. The global warming is a fact, and it has been predicted that the average global temperature will be increased with between 1.5-4.5 °C by the year 2050 [6]. The increased temperature is mainly due to the release of greenhouse gases caused by human activities, and around half of the greenhouse effect occurs due to the release of carbon dioxide (CO2). Most of the released carbon dioxide is due to the burning of fossil fuels, such as coal and oil [7].
To slow down the global warming, the emission of greenhouse gases caused by human activities needs to be decreased. One way to contribute to this decrease is by using materials with a low environmental impact, indicating that oil-based materials such as plastics need to be changed to materials that are more environmentally friendly. This has proven to be a challenging task since plastic materials are widely used today due to their properties, that makes them suitable for many packaging applications. The issue regarding plastics is that these materials often are derived from oils [8], which is a process that contributes to the release of carbon dioxide. Most plastic materials are not biodegradable or renewable, which makes their environmental impact high. The pulp and paper industry can instead provide the world with recyclable and biodegradable materials, made from renewable resources, such as wood and other plants [9]. These wood-based materials mostly consist of cellulose, which is a strong and light weighted material, suitable for several applications [10]. Unfortunately, cellulose has some limitations compared to plastics [9], and therefore the development of cellulose-based materials with improved properties is constantly under progress.
2
1.3 Purpose of the study
The characteristics of products made from pulp can be modified to obtain improved properties. Refining of the fibres in the pulp results in an increased strength of the fibre network due to an improved flexibility of the fibres and due to the formation of small fiber pieces called fines [11]. Another possibility is the chemical modification of cellulose to obtain materials with changed properties. Periodate oxidation of cellulose results in dialdehyde cellulose that can be further reduced with sodium borohydride to obtain dialcohol cellulose (DALC), which is a relatively new material with an increased ductility compared to cellulose [12].
BillerudKorsnäs have been doing several trials regarding DALC, by optimizing the reaction conditions and by using both unrefined and industrially refined pulp. When unrefined pulp have been used for the production of DALC the resulting sheets showed to have an elongation of around 7-10 %, but when DALC was made from industrially refined pulp, the resulting sheets had very low elongation of around 1-3 %. The aim of this thesis was to investigate why DALC made from industrially refined pulp has such a low ductility.
The main purpose was to find out if the low ductility is connected to the presence of fines in the pulp. It was also examined if the results were similar for DALC made from different degrees of laboratory refined pulp.
3
2 Theory
2.1 The structure of wood
There are two categories of wood, softwood, and hardwood. Softwood includes fir, pine, spruce etc. and examples of hardwood are birch, beech and oak [13]. Softwood consists of two cell types, tracheids and ray cells, while hardwood consists of a number of cell types including vessels and ray parenchyma cells [14]. The physical and mechanical properties of products made from wood are mostly governed by the cell wall structure, which consist to a major part of cellulose, lignin, and hemicellulose. The structure can simplified be described as cellulose fibres forming a skeleton surrounded by lignin and hemicellulose [13,15].
Depending on the type of wood these three components exist in different fractions but with cellulose as the major part. Hardwood consists of 43-47 % cellulose and softwood contains 40-44 % cellulose [16]. The structure of cellulose is hierarchical which means that the cellulose fibre is built up from several bundles consisting of numerous microfibrils, each microfibril contains several cellulose chains that are organized together [14], as illustrated in Figure 1.
Figure 1.The structure of the cellulose fibre, derived from Gumrah Dumanli [17] .
2.2 Cellulose
2.2.1 Structure
Cellulose is a polymer chain that consists of repeating β D-glucopyranose units, linked through covalent bonds between the carbon atom C1 and the OH group attached to carbon atom C4, illustrated inFigure 2.
The length of the linear chain can be expressed as the degree of polymerization (DP), which is defined as the number of anhydroglucose units (AGU) that are bonded together [18]. The DP value differs depending on the raw material, with 6000-10 000 for cellulose from wood and between 4000-6000 for bacterial cellulose [16].
The polymer chain contains one non-reducing end, carrying the C4-OH group, and one reducing end containing the C1-OH group, as indicated inFigure 2 [18].
Figure 2.A schematic illustration of a cellulose chain containing the non-reducing end, the reducing end, and the repeating unit (AGU) where n is equal to the degree of polymerization, edited from Klemm et al [18].
4
The cellulose chain described above is known as the primary structure of cellulose while the secondary structure is formed when these chains bond together. Native cellulose has a secondary structure called cellulose I, which contains several chains organized parallel through hydrogen bonds forming sheets. These sheets interact through Wan der Waal forces, creating the two different crystal structures of cellulose I, cellulose Iα and cellulose Iβ, illustrated in Figure 3[10]. Both structures are always present, but in different fractions depending on the type of cellulose. For bacterial cellulose the dominating part is cellulose Iα and for plants (wood and cotton) the major part is cellulose Iβ [19]. Cellulose I can be chemically treated to the three other existing secondary structures: cellulose II, III and IV. The most stable of these modified structures is cellulose II, with chains that are organized antiparallel [10,14,18].
Figure 3.An illustration of the two crystal structures, cellulose Iα and cellulose Iβ. Cellulose chains are layered on top of each other from the side, derived from Henriksson et al [10].
2.2.2 Properties
Cellulose is a strong, light weighted, renewable and inexpensive material. It is insoluble in water and has a hygroscopic character, which means that cellulose is hydrophilic and can absorb large quantities of water, which result in swelling [10,20]. Cellulose has some limited properties compared to other oil-based materials, which prevent the use in some applications. The low ductility compared to plastics limits the possibility to have a final product in a structure with a complex shape. Cellulose also has limited gas-barrier properties, which means that it cannot provide a strong barrier against oxygen or water vapor. Therefore, cellulose is not suitable to be used without coatings in some packaging solutions [12]. However, cellulose can be modified chemically to obtain properties that makes it suitable for more advanced applications [21].
2.3 Chemical modification of cellulose
Cellulose is a reactive molecule since all three hydroxyl groups, in each repeating unit within the chain, can react with added chemicals. The hydroxyl group attached to the primary carbon (C6) is the most reactive of the three, due to fewer neighboring substituents, while the two other hydroxyl groups attached to secondary carbons (C2 and C3) are less reactive [14]. However, the added chemicals and the conditions during the reaction determines which hydroxyl groups that are dissociated. During an oxidation with 2,2,6,6-tetramethylpyperidine-1-oxyl (TEMPO), the primary hydroxyl groups of the D-glucose units are converted to carboxylic acids, while a periodate oxidation converts the two secondary hydroxyl groups to aldehydes [22,23].
Since cellulose fibres have a low ductility compared to the commonly used fossil based plastics, many researchers have been trying to find methods that results in stretchable sheets. This increased ductility is desired to make it possible to use cellulose in packaging solutions with complex shapes. Khakalo et al. [24]
showed that a strain-at-break of almost 22 % could be obtained for sheets made from refined fibres with an addition of 20 wt % gelatin. Vuoti et al. [25] obtained an elongation of almost 16 % for sheets made with hydroxypropylated cellulose ethers. Larsson et al. [26] prepared films containing a core of crystalline nanofibrils enclosed by a dialcohol cellulose shell, which resulted in sheets with a strain-at-break of around 15 %. In another study performed by Larsson et al. [12], sheets were made purely from dialcohol cellulose, which resulted in an elongation of 11 %. Dialcohol cellulose is a ductile material relatively easy to obtain through periodate oxidation followed by borohydride reduction.
5 2.3.1 Periodate oxidation of cellulose
The oxidation of cellulose with sodium periodate (NaIO4) as the oxidizing agent, occurs by breaking the ring structure in the anhydroglucose unit, and thereby the bond between C2 and C3. Two hydroxyl groups are replaced by aldehyde groups, as shown in Figure 4 [23]. The resulting material is called dialdehyde cellulose (DAC) and can be used as a starting material for other applications. DAC is more reactive than regular cellulose since it contains dialdehyde groups and is therefore suitable as an intermediate, with the possibility to react with other groups, creating alternative materials, such as dialcohol cellulose or dicarboxylic cellulose [27,28].
Figure 4. Schematic illustration for the oxidation of cellulose to DAC, edited from Larsson et al [12].
The degree of oxidation, which is the number of aldehydes created during the reaction [29], is important for the properties of DAC. Both Larsson et al. [29] and Hou et al. [27] showed that the wet and dry tensile strength of sheets made from DAC fibres increased when the degree of oxidation was increased. The number of aldehydes formed during the oxidation are dependent on the conditions during the reaction: the amount of sodium periodate, the pulp concentration, the reaction time and the temperature will affect the result [30].
By increasing the amount of sodium periodate and the time of the reaction, the degree of oxidation will be increased [27,30]. The reaction is slow in room temperature [12], and if the temperature is increased to 55-75 °C, the reactivity will be increased, and thereby the reaction time and the amount of sodium periodate can be decreased [31]. The temperature should be kept below 55°C since periodate will be unstable at higher temperatures. When periodate is unstable it can be decomposed into liberate iodine, which will make it hard to determine the amount periodate used during the reaction [30].
There are different methods to determine the degree of oxidation, the most widely used is through the addition of hydroxylamine hydrochloride. The hydroxylamine hydrochloride will react with all the carbonyls present in the solution while protons are released. After 2 h the reaction is completed, and the solution is titrated back to the original pH of hydroxylamine hydrochloride using sodium hydroxide (NaOH). The degree of oxidation is determined from the required amount of sodium hydroxide [29].
DAC can be further sulfonated with sodium bisulfate as the sulfonating agent. The resulting material has shown an improved water absorbance of the fibres, improved fibre swelling and increased dry and wet tensile strengths of the final sheets [27]. The aldehyde groups can also be converted to carboxylic groups through oxidation with sodium chlorite, where the result is dicarboxylic cellulose (DCC), used in several applications [28]. DCC have been introduced to spherical cellulose gel, suitable for applications including chromatography due to mechanical strength and stability [32].
6 2.3.2 Borohydride reduction of DAC
A further reduction of DAC with sodium borohydride (NaBH4) is possible, and the result of the reaction is dialcohol cellulose (DALC), illustrated in Figure 5. DALC has an open-ring structure and is therefore expected to be more flexible and have different physical properties than cellulose [33]. DALC influences the sheet forming process since it is complicated to dewater and needs a longer dewatering time compared to cellulose [34]. Sheets made from DALC have shown to have a high ductility and a high tensile strength. These properties also showed to be affected by the degree of oxidation obtained from the periodate reaction. The strain-at-break for the DALC sheets showed to increase with an increased degree of oxidation, while the tensile index was increased up to a certain level of oxidation to then decrease [12].
Figure 5. The reduction of DAC to DALC, edited from Larsson et al. [12].
2.4 Fibre properties
The fibre properties are important for the sheet forming process and thereby for the properties of the final product. Depending on the origin of the fibres, the properties can differ. Softwood and hardwood fibres differ by shape, fibre length, fibre diameter, density of cell wall and wall thickness. The fibre length is an important parameter in papermaking and can range from 1-6 mm. Softwood fibres have an average length of 3 mm while the average length of hardwood fibres are 1 mm. Sheets made from pulp containing long fibres generally have a higher strength than sheets made from short fibre pulp. The reason is that each long fibre can create a larger number of bonds to several other fibres, which improves the strength of the fibre network.
However, if the fibres are too long the resulting sheets can have uneven structures and thereby lower strength [35,36].
Fibre conformability is a property that can be described as the combination of flexibility, swelling and the possibility for the fibres to collapse. A high conformability is equivalent to fibres that can easily be formed after each other, and thereby create a strong network due to increased fibre bonding. Flexibility is the ability to bend, change shape or to deform due to external forces, and swelling measures water retention. When fibres are exposed to lateral pressure, which can be during drying or wet pressing, the fibres can collapse.
Their structure changes from the usual tube-like structure to more of a double layered strip, as illustrated in Figure6. The collapsed fibres increase the strength properties due to a higher binding capacity but affects the optical properties negatively [36]. Collapsed fibres have less unbonded surface areas compared to regular fibres, and since the light is reflected at the fibre surface and scattered at the unbonded surface, the light scattering will be decreased with a higher number of collapsed fibres [37].
Figure 6. The figure shows an illustration of a fibre that is exposed to lateral pressure and thereby collapse, edited from Popa [36].
7
The strength of the individual fibres and the bonds between the fibres influences the strength of the fibre network and thereby the strength of the formed sheet. Both the specific bonding strength, defined as the strength in the z-direction, and the bonded area, which is the number of contact points between fibres, influences the strength of the bonds [35].
2.5 Refining
To increase the strength of the fibres, the pulp can be exposed to a mechanical treatment, called beating or refining. During the process the wet fibres are subjected to mechanical forces, which will reduce the fibre length, cause fibre curling or straightening and create both internal and external fibrillations of the fibres [11]. Generally, the fibres will be more flexible, and the possibility for bonding between the fibres will be increased. Refining will also contribute to the creation of small fiber pieces that are called fines [11,38,39].
2.5.1 Fibrillation
During internal fibrillation the two first layers of the fibre wall are exposed to delamination, which will increase the swelling potential of the fibres [39]. The resulting fibres will be more conformable and flexible, which increases the bonding between the fibres. Internal fibrillation is considered as one of the most significant effect of refining [35]. External fibrillation is when fibrils are partly removed from the fibre wall [11], resulting in a hairy appearance of the fibre surface [35]. The external fibrillation is mainly caused by surface shear strains from the refiner, and the resulting fibres will have an increased surface area and flexibility [35,40]. The formation of fines occurs through the same shear strains, but the fibrils are completely removed from the fibre wall [40].
2.5.2 Fines
Fines are defined as particles that can pass through a round hole with the diameter 76 µm (or equivalently a 200 mesh wire) according to ISO 10376:2011. These particles can consist of cellulose, hemicellulose, lignin and extractive substances, and can be in the form of fibre pieces, fibrils, ray cells, lamellae fragments and so on [41].
There are two different categories of fines depending on their shape. The first includes chunky, flake-like particles that are lignin-rich and have a small length to width ratio. These fines will decrease the possibility for bonding between fibres and will therefore have an unfavorable effect on the mechanical properties. The second one considers the more flexible particles with a high cellulose content and a large length to with ratio, that will increase the bonding between fibres and increase the strength of the paper [39,42,43].
Fines can also be divided into primary and secondary fines. Primary fines are present in the pulp before the refining process and can consist of middle lamella lignin, parenchyma-, and ray cells. Secondary fines are created during refining and can consist of fibrils and lamellae fragments from the fibre wall [38,39,41].
Primary and secondary fines affect the paper properties different. Secondary fines, with the large surface area combined with the high potential for bonding, increases the strength of the paper to a higher extent compared to primary fines [38,44].
Alince et al. [42] showed that the presence of fibrillar fines improves the interfibre bonding, due to the formation of bridges between crossing fibres. These bridges arise due to the collapse of the fibrillar fines on the fibre surface upon drying. This can be observed in SEM images by Motamedian et al. [39] which also shows the formation of a web structure consisting of fines.
8 2.5.3 Laboratory and Industrial refiners
There are two different categories of refining equipment, industrial and laboratory refiners. Industrial refiners can consist of different geometries, such as cylindrical, conical and disk refiners, and are used in the industry to refine large amount of pulps. Some of the different laboratory refiners are PFI Mills, Jokro Mills, and Valley Beaters, and they are all appropriate when small samples are refined [45]. It is possible to replicate an industrial refiner with a smaller laboratory equipment, namely an industrial pilot refiner with similar equipment and operating parameters to an industrial refiner but in a smaller size [46].
The PFI Mill consists of a smooth housing cylinder and a disk with bars on the edges, that are pressed against the housing to achieve refining action. The pulp is uniformly applied around the inner wall of the housing cylinder and the disk is submerged into the cylinder. During the refining process the disk and the housing are rotating in the same direction but with different speeds. The degree of refinement is denoted by revolutions [35,45].
An earlier research made by Somboon [46] showed that depending on the refiner used, the resulting pulp had different fibre structures. When the pulp was treated with an industrial pilot refiner it consisted of shorter fibres, a higher value of fines and an increased fibre wall removal. The pulp treated with a laboratory refiner contained a larger amount of long fibres, a lower number of fines and a greater change in the internal fibre structure. It was found that these differences affected the paper properties. Sheets made from the laboratory refined pulp showed a higher tensile strength and higher sheet density than sheets made from the pulp refined with the industrial pilot refiner.
2.6 Sheet forming
Several paper properties are determined from sheets made with a laboratory sheet former. The equipment consist of a section where the pulp solution is applied, a wire used for the forming process and a section where the process water that pass through the wire ends up. Another name for the process water is white water due to the fibres, fines, and other particles present in the water [35].
The applied pulp solution usually has a low concentration, which results in a uniform fibre distribution, and thereby an isotropic sheet. A white water system can be open or closed, which results in different retention of fines. An open system gives a low retention while a closed system accumulates fines and gives a higher retention. Most laboratory sheet formers have an open system, but Rapid Köthen (ISO standard 5269-2) is a commonly used laboratory sheet former where the white water system can be both open and closed. When the white water system is closed it is possible to re-use the white water [35].
Sheets made with an industrial paper machine are anisotropic, since the fibres are oriented in certain directions, and these sheets therefore have properties dependent on the direction [35]. An industrial paper machine is more complicated than a laboratory sheet former, and there are properties desirable to optimize in the laboratory for further use in the industry. Properties such as the drainage and retention during the paper production can be optimized using a Dynamic Drainage Analyzer [47].
2.6.1 Dynamic Drainage Analyzer
A Dynamic Drainage Analyzer (DDA) is a small laboratory device that is used to measure the drainage of pulp by simulating the conditions of a paper machine used in the industry. The initial drainage, the drainage speed and the temperature are obtained for specific time intervals during the measurements. It is possible to study the drainage conditions for different pulp compositions, effects of different amount of added chemicals and fillers or different degrees of refining can be obtained. The equipment simplified consist of a forming jar where the pulp solution is applied, an exchangeable wire attached under the forming jar, a stirrer, a turbidity sensor and a vacuum system [47].
9
2.7 Paper properties
Some of the properties that characterizes paper can be divided into optical, physical, and mechanical properties, and depending on the use of the final product the importance of these properties differs. The quality of a paper is affected by the moisture content, and therefore several paper properties are measured at 23 °C and 50 % relative humidity according to the standard ISO 187:1990 [48]. The paper properties are influenced by the properties of the individual fibres and of the fibre network [36].
2.7.1 Mechanical properties
The mechanical properties of a product made from pulp are important. When considering the mechanical properties for an anisotropic sheet, there are two different directions used, the machine direction (MD) and the cross-machine direction (CD). The machine direction is parallel to the movement direction of the paper making machine and the cross-machine direction is perpendicular to the machine direction. Several paper properties are different in the machine, and cross-machine direction, and therefore these paper properties are dependent on the paper orientation [48]. Sheets made with the Rapid Köthen sheet former are isotropic, which means that the properties are independent on the paper orientation [35].
2.7.2 Tensile test
A common way to test the mechanical properties of a sheet is through a tensile test, which is performed by using a strip of a paper clamped, either vertical or horizontal, at each end to a tensile tester. The paper is elongated until failure by applied forces, as illustrated in Figure 7 [49].
Figure 7.An illustration of a tensile test performed on a paper strip, edited from Ek et al [49].
During a tensile test, the applied loading is measured as a function of how much the materials is elongated and can result in a force-elongation curve that can be converted into a stress-strain curve. The typical behavior of a stress-strain curve for paper is shown in Figure 8. The stress is determined with Equation (1), where F is the applied force, b is the width, and t is the thickness of the paper [49].
𝜎 = 𝐹
𝑏𝑡 (N/m2 or Pa) (1)
The strain is determined by the ratio between the length change, ∆𝑙, and the original length, 𝑙, of the paper, as shown in Equation (2), and is usually expressed in percent [49].
𝜀 = ∆𝑙
𝑙 (2)
10
Figure 8.The figure shows the typical behavior of a stress-strain curve, derived from Ek et al. [49].
The stress-strain curve follows a linear behavior when the material is deformed elastically. When the plastic deformation starts the curve deviates from the straight line until the material breaks, which indicates the point of failure, as shown in Figure 8. The curve can be used to determine several different properties, including tensile strength, tensile strength index and strain-at-break [49]. If the material is homogeneous, the linear behavior of the curve can be characterized by Hooke´s law, shown in Equation (3) [50].
𝜎 = 𝐸 ∙ 𝜀 (N/m2 or Pa) (3)
Where E is the modulus of elasticity, defined in Equation (4) as the slope of the straight line describing the elastic behavior of the stress-strain curve in Figure 8 [50].
𝐸 = ∆𝜎
∆𝜀 (N/m2 or Pa) (4)
2.7.3 Tensile strength and Tensile index
When paper is exposed to a tensile force, the strength can be expressed in several ways, tensile strength and tensile strength index are two commonly used parameters. The tensile strength is defined in Equation (5), where 𝐹𝑇 is the maximum applied force [49]. For paper, the tensile strength is measured as the force/unit width and not as the force/unit area, which is the case for other materials [48]. The reason is that paper is a porous material with a density and pore volume that can change and thereby the tensile strength can be misleading if it is measured as the force/unit area [49].
𝜎𝑇𝑏 =𝐹𝑇
𝑏 (N/m) (5)
The tensile strength is influenced by the fiber strength and the degree of fibre bonding and is therefore a suitable measurement of the capacity of fibre bonding in wood pulp. The value is also dependent on the conditions during the test, e.g. the loading rate or the content of moisture in the paper [50]. The tensile index relates the strength to the amount of the material that is exposed to the loading and is defined in Equation (6), where w is the grammage or basis weight of the paper, defined as the mass per unit area. The tensile index can range from 10-100 kNm/kg for paper materials [49,50].
𝜎𝑇𝑤 = 𝜎𝑇
𝑏
𝑤 (Nm/kg) (6)
11 2.7.4 Ductility and Strain-at-break
Ductility is a measurement of how much a material, exposed to a force, can be deformed plastically without breaking. The ductility can be measured with the strain-at-break, which is the ratio between the increased length of the paper strip at the point of failure to the original paper length, expressed in percent [49,51]. For most papers the value lies between 1-5 %, but values above 20 % have been found for specific paper [50]. A high strain-at-break indicates that the material is ductile.
2.8 Scanning Electron Microscope
To analyze the surface of a paper sheet, a scanning electron microscope (SEM) can be used. The first commercial SEM was developed in the beginning of 1960 and is one of the most commonly used electron microscopes. It provides the user with a high resolution image, showing the microscopic structure of the specimen surface. The large depth of field makes the appearance of the image three-dimensional [52,53].
2.8.1 The instrument
A conventional SEM operates in vacuum, and includes an electron gun, several lenses, an electron detecting system and deflectors, as shown in Figure 9 [52].
Figure 9.A schematic illustration of a conventional SEM, edited from Khursheed [52] and Leng [53].
The electron gun emits electrons that are accelerated to energies between 1 and 30 keV. The choice of electron gun is important since the quality of the resulting SEM image depends on the amount of current provided by the gun. There are two commonly used types of electron guns, the thermionic gun and the field emission gun. The thermionic gun creates a crossover point located between the anode and cathode, as illustrated in part a of Figure 10. A Wehnelt electrode, positioned in front of the cathode, is negatively biased relative to the cathode voltage, and will focus the emitted electrons and the crossover point is formed. The tungsten gun and the LaB6 gun are two thermionic guns [52].
12
A field emission gun has suppressor electrodes that are negatively biased, positioned beside the cathode, pushing the electrons against the anode, as shown in part b of Figure 10. The emitted electrons will not form a crossover point but will leave the tip through quantum tunneling [52]. When using a field emission gun as the electron source the requirement regarding the vacuum is very strict and therefore a separate vacuum pumping system for the electron gun is needed [54].
Figure 10. The two commonly used electron guns a) The thermionic gun and b) The field emission gun, edited from Khursheed [52].
The electron beam passes through different electromagnetic lenses, which is shown in Figure 9. These lenses are used for the formation of the electron probe and include two condenser lenses followed by one objective lens. Usually an aperture is placed in front of each lens to prevent wide angle electrons to reach the specimen.
Deflectors are positioned inside the objective lens and creates the movement of the probe over the sample surface. The electron detector collects the electron signals created from the interaction between the electron beam and the sample [52,53].
2.8.2 Electron signal
There are two different types of electron signals that are of importance in SEM, secondary-, and backscattered electrons. When an electron penetrates a solid sample, the electron can be scattered both elastically and inelastically. For inelastic scattering, kinetic energy is transferred from an incident electron to an electron in an atom in the sample. An electron from the same atom with enough kinetic energy leaves and becomes a secondary electron, as illustrated in part a of Figure 11. In the case of elastic scattering an incident electron is scattered by an atom in the sample, and these electrons are called backscattered electrons, which is shown in part b ofFigure 11 [53].
Figure 11. The figure shows a schematic illustration of an electron exposed to a) Inelastic scattering and b) Elastic scattering, edited from Thermo Fisher Scientific Phenom-World BV [55].
13
Depending on the type of electron signal, the electrons escape from different positions in the sample. As illustrated in Figure 12, the electrons can be scattered from a pear-shaped interaction zone. Secondary electrons can escape from the volume within 5-50 nm from the specimen surface, while backscattered electrons can, due to their high energy, escape from the volume within 50-300 nm from the surface. The characteristic X-rays, also indicated in Figure 12, can escape from the volume below 300 nm, and these are used in chemical analysis [53].
Figure 12.The different depths that electrons can escape from within a specimen, derived from Leng [53].
2.8.3 Contrast
The contrast in SEM images can be compositional or topographic, depending on the type of electron signal.
Backscattered electrons are the primary signal for a compositional image, which shows a variation corresponding to different chemical compositions on the sample surface. The topographic image is formed mainly by secondary electrons and shows variations depending on the geometries on the sample surface [53].
2.8.4 Resolution and Depth of field
The resolution in SEM is dependent of the cross-sectional probe diameter, illustrated in Figure 13. High resolution is obtained by minimizing the probe diameter, 𝑑𝑝, determined with equation (7), where β is the brightness of the beam, 𝑖𝑝 the probe current and 𝛼𝑓 the convergence angle. To minimize the probe diameter the brightness should be increased, and the convergence angle should be optimized [53].
𝑑𝑝 = ( 4𝑖𝑝
𝛽𝜋2𝛼𝑓2)
1/2
(7)
Figure 13.The situation when the probe current hits the specimen surface, indicating the convergence angle, probe diameter, working distance and aperture diameter, derived from Leng [53].
14
The resolution can be adjusted with the probe current and the acceleration voltage, which are two operational parameters. If the acceleration voltage is increased, the result is an increased beam brightness and thereby a decreased probe size. But when the voltage becomes higher, the interaction zone is increased, and a lower lateral spatial resolution is obtained. The beam brightness also depends on the electron gun, and a field emission gun is brighter than a thermionic gun. By using a field emission gun, the beam brightness can be high without using a high acceleration voltage [53].
To get the appearance of the SEM image to be three-dimensional, the depth of field needs to be high. The two operational variables that affect the depth of field is the aperture size and the working distance, showed in Figure 13. To achieve a high depth of field the working distance should be long and the aperture size small, but this combination decreases the resolution of the image due to a decreased convergence angle. Therefore, the working distance and the aperture size should be intermediate to get both the desired depth of field and resolution [53].
2.8.5 Sample preparation
When analyzing a sample with SEM, the specimen can be of different forms: thin films, bulk material and powder. The specimen needs to conduct electricity, and samples that do not can be coated with a conductive film onto the surface. When analyzing a sample that does not conduct electricity, there is a risk for surface charging, which means that there are charged regions on the sample surface that will generate distortion and the resulting image will be misleading [53].
15
3 Experimental
3.1 Materials
Bleached softwood kraft pulp was supplied from BillerudKorsnäs AB. Unrefined pulp with a concentration of 4-5 % was supplied from Gruvön Mill Sweden, and industrially refined pulp with a concentration of 4 % was supplied from Skärblacka Mill Sweden. The chemicals used during the thesis was all supplied from VWR:
Sodium metaperiodate (NaIO4), Isopropanol (C3H8O), Sodium borohydride (NaBH4; 98 %), Sodium phosphate monobasic (NaH2PO4; 99 %), Hydroxylamine hydrochloride (NH2OHꞏHCl) and Sodium hydroxide (NaOH).
3.2 Methods
Nine different pulp compositions were prepared for the chemical modification to DALC: unrefined pulp, unrefined pulp with added fines, industrially refined pulp, dewatered industrially refined pulp, and pulp refined 1000, 3000, 5000, 10 000 and 15 000 revolutions with a PFI Mill. Two batches of DALC were made for each pulp composition except the pulp refined 15 000 revolutions since the time was limited and such a high degree of refinement is considered an extreme value.
3.2.1 Refining
The pulp was refined with a PFI-Mill according to the standard ISO 5264-2:2011. 300 g pulp containing 10 % fibres was prepared from 750 g of unrefined pulp with 4 % (30 g) fibres. The pulp was applied to the walls of the housing cylinder and the disk containing bars was submerged into the cylinder. The pulp was refined between the cylinder and the disk, rotating in the same directions but with different speeds. Five different levels of refining were used: 1000, 3000, 5000, 10 000 and 15 000 revolutions.
The refined pulp was transformed to a steel beaker and diluted with tap water to 2 L. The pulp solution was defibrillated in a L&W Pulp Disintegrator, using 10 000 revolutions, in order to release each fibre without destroying them. To prepare the pulp for the modification process, the pulp solution was first dewatered using a vacuum dewatering pump, a Buchner funnel, and a filter paper with a particle retention of 5-13 µm.
The pulp was diluted with tap water to a 4 % concentration.
3.2.2 Fibre properties
The fibre properties of unrefined and industrially refined pulp were tested with a L&W Fibre Tester Plus.
Prior to the test a pulp solution with 1 g fibres/L solution was prepared. 100-150 ml of the prepared solution was transferred to a 250 ml plastic bottle. The test resulted in the following mean fibre properties: length, width, shape, fibril area and fibril parameter, and the fibres with a length smaller than 0.2 mm was considered as fines in accordance with the standard ISO 10376:2011. The results were used to get an approximate number of fines/g fibres and fines/ml pulp water in unrefined and industrially refined pulp.
3.2.3 Addition of fines
Fines were collected through dewatering with a Dynamic Drainage Analyzer (DDA) v.5. A 1 L pulp solution of 1 % concentration (10 g fibres) was prepared by adding tap water to 250 g industrially refined pulp. A 60 mesh (0.25 mm) wire was mounted to the wire section of the DDA and the pulp solution was added to the forming jar. When the dewatering process was finished the white water was collected. To check the fines content in the white water, the L&W Fibre Tester Plus was used. The test was performed four times to get an approximate number of fines/L white water.
16
The estimated number of fines/g fibres and fines/ml pulp water in unrefined and industrially refined pulp were used to determine the number of fines added to unrefined pulp. These values were considered to make sure that the number of fines in unrefined pulp with added fines were significantly higher compared to unrefined pulp. The dewatering process was performed ten times to get the desired number of fines, and the white water from each performed process was collected in a bucket.
The unrefined pulp was prepared by dewatering the pulp to a concentration of around 20 %, using a regular strainer. The pulp was added to a 10 L bucket, the collected white water was added to the bucket and the solution was stirred for 30 min. The pulp solution was dewatered using a vacuum dewatering pump, a Buchner funnel, and a filter paper with a particle retention of 5-13 µm. The pulp was diluted with tap water to a concentration of 4 %.
3.2.4 Dewatered industrially refined pulp
Dewatered industrially refined pulp was prepared by filtering out the fines in the liquid phase by dewater the industrially refined pulp to a concentration of 20 % using a regular strainer. Tap water was used to dilute the pulp to 4 %.
3.2.5 Chemical modification of cellulose
The oxidation of cellulose to DAC, and the reduction of DAC to DALC were performed in a 1000 ml reactor vessel. The temperatures, the amount of chemicals and the reaction times used, have earlier been optimized by BillerudKorsnäs and are left out here due to a non-disclosure agreement with BillerudKorsnäs.
Prior to the periodate oxidation, 700-800 g of 4 % pulp was added to the vessel. A stirrer and a lid was mounted to the vessel and the pulp was adjusted to the correct temperature during stirring. When the pulp had reached the desired temperature, NaIO4 (Sodium meta periodate) and C3H8O (isopropanol) was added to the vessel. When the reaction was completed the pulp was washed with tap water until the conductivity of the filtrate was less or equal to the conductivity of tap water, 150 µS/cm. The washing was performed to be certain that the pulp no longer contained any chemicals from the reaction. The dry content of the pulp was measured.
To estimate the degree of oxidation in DAC, two beakers with 0.5 g wet pulp were prepared and hydroxylamine hydrochloride was added to each beaker. The two solutions were stirred for 2 h, and then the solutions were titrated back with NaOH (sodium hydroxide) to the original pH of hydroxylamine hydrochloride, and the number of aldehydes/g fibres was obtained.
The pulp was diluted back to a 4 % concentration using a solution of NaH2PO4 (monobasic sodium phosphate). The pulp solution was transformed to the vessel, the stirrer and the lid was mounted and NaBH4
(sodium borohydride) was added. Since the addition of sodium borohydride results in a rapid increase of the pH, monobasic sodium phosphate can be added, which decreases the initial pH, and thereby limits the pH increase. When the reaction was finished the pulp was washed with tap water until the filtrate had a conductivity ≤ 150 µS/cm, and the dry content was measured.
3.2.6 Preparation of sheets
Sheets were made using a Rapid Köthen sheet former according to a modified version of the standard ISO 5269-2 since DALC is a material with a slow dewatering process. The target grammage for each sheet was 100 g/m2. The pulp solution was prepared and stirred for 30 min prior to the sheet forming. A 400 mesh wire was attached on top of the original 200 mesh wire and the forming column was mounted. The sheets were dried between two 400 mesh wires for 15 min, and then stored in a room with 23 °C and 50 % relative humidity (ISO 187:1990) until further testing.
17 3.2.7 Tensile test
The mechanical properties of the prepared sheets were measured in 23 °C and 50 % relative humidity (ISO 187:1990), using a Zwick Roell Z005 tensile tester. The tests were performed according to ISO 1924-3, but with a test speed of 10 mm/min instead of 100 mm/min since a high test speed can easily destroy the DALC material. Fewer test pieces than according to the standard were used since the tests were performed to get an idea of the results rather than very accurate numbers.
Each sheet was cut into 6-8 pieces, with the dimensions 15 mm x 140 mm, and the thickness was measured.
The paper pieces were mounted vertical to the tensile tester one at the time and the tests were performed, resulting in several properties that could be chosen prior to the testing. The ones obtained were the tensile strength, the tensile strength index, and the strain-at-break.
3.2.8 SEM imaging
The instrument used for the SEM analysis was a SEM LEO Gemini 1530. Prior to the SEM imaging, the specimens were prepared by cutting one small paper piece from each DALC sheet used for the analysis. The specimens were coated with Au (gold) using a Jeol Fine Coat Ion Sputter. Several topographic SEM images with varying magnification were taken for each specimen.
18
4 Results
4.1 Pulp compositions
The different pulp compositions used for the chemical modification to DALC are presented in Table 1 below.
Table 1. The different pulp compositions used for the chemical modification.
Unrefined pulp
Unrefined pulp with added fines Industrially refined pulp
Dewatered industrially refined pulp Pulp refined 1000 revolutions Pulp refined 3000 revolutions Pulp refined 5000 revolutions Pulp refined 10 000 revolutions Pulp refined 15 000 revolutions
For each pulp composition in Table 1, except for the pulp refined 15 000 revolutions prior to the modification, two batches were made in order to see that the results were consistent.
4.2 Addition of fines to unrefined pulp
The number of fines that was added to unrefined pulp was decided through the results obtained from the L&W Fibre Tester Plus. The total number of fibres and the fraction of fines in the pulp solutions, were obtained from the fibre tester, and are presented in Appendix I. By using these results together with Equation (i) in Appendix I, the number of fines/g fibres and fines/ml pulp water were calculated. The test results were used to determine an approximate level of fines rather than an exact number, and these approximated values are presented in Table 2 below. For the modification to DALC, pulp containing approximately 32 g fibres was used and therefore the number of fines/32 g fibres for the two pulp compositions are presented in Table 2. The pulp water for both the unrefined, and the industrially refined pulp contained 150-200 fines/ml and were therefore not considered relevant when deciding the number of fines added to unrefined pulp.
Table 2. The number of fines in unrefined and industrially refined pulp.
Pulp Unrefined Industrially refined
Fines/g fibres 30 000 50 000
Fines/32 g fibres 0.96 ∗ 106 1.6 ∗ 106
Fines/ml pulp water 150-200 150-200
19
Fines were filtered out from industrially refined pulp with a DDA. The results from the tests of the white water performed with the L&W Fibre Tester Plus are shown in Appendix I, and these tests resulted in the approximate value of 50 000 fines/L white water. Approximately 0.5 ∗ 106 fines were added to unrefined pulp through 10 L of white water.
4.3 Mechanical testing
The tensile tests resulted in several parameters, but the relevant ones for this study were the strain-at-break and the tensile strength index. Mean values with a 95 % confidence interval were calculated for each batch and are shown in Table 3 below. The number of performed tests varied between 16-31 for the different batches, and therefore the statistical model appropriate for the calculations of the confidence interval was a t-distribution.
Table 3. The mean values for the tensile strength index and the strain-at-break for the Rapid Köthen sheets made from the different batches of DALC. Each batch have been assigned a number to make it easier to separate them.
DALC made from Batch Tensile strength index (kNm/kg) Strain-at-break (%)
Unrefined pulp 1a 66.5 ± 2.61 7.99 ± 1.06
1b 63.4 ± 1.46 6.33 ± 0.49
Unrefined pulp with
added fines 2a 65.0 ± 1.29 4.89 ± 0.38
2b 65.3 ± 1.69 5.58 ± 0.69
Industrially refined pulp 3a 70.4 ± 1.43 4.12 ± 0.39
3b 76.0 ± 1.58 4.61 ± 0.58
Dewatered industrially
refined pulp 4a 72.1 ± 1.80 4.98 ± 0.43
4b 72.4 ± 0.97 3.81 ± 0.29
Pulp refined 1000
revolutions 5a 66.2 ± 1.99 4.61 ± 0.49
5b 67.2 ± 1.60 5.39 ± 0.50
Pulp refined 3000
revolutions 6a 72.7 ± 1.57 5.14 ± 0.56
6b 77.2 ± 1.06 5.95 ± 0.40
Pulp refined 5000
revolutions 7a 80.9 ± 1.82 6.28 ± 0.56
7b 71.6 ± 1.60 6.44 ± 0.31
Pulp refined 10 000
revolutions 8a 77.9 ± 2.17 7.49 ± 0.76
8b 86.5 ± 2.38 6.45 ± 0.56
Pulp refined 15 000
revolutions 9 84.2 ± 3.40 7.97 ± 0.67
20
In Table 3, the exact values of the strain-at-break for the different batches of DALC are shown. The strain-at-break for DALC sheets can be divided into three intervals, with poor being the values smaller than 4 %, average the values between 4-7 %, and good the values larger than 7 %. This division is arbitrary and based on earlier trials performed by BillerudKorsnäs.
To indicate how the strain-at-break in Table 3 differs between DALC made from the different pulp compositions, the values are presented in the two bar diagrams below. In Figure 14 the strain-at-break for DALC made from unrefined pulp, unrefined pulp with added fines, industrially refined pulp and dewatered industrially refined pulp are shown. Figure 15 shows the values for DALC made from the different degrees of laboratory refined pulp.
Figure 14.The strain-at-break for DALC made from unrefined pulp, unrefined pulp with added fines, industrially refined pulp and dewatered industrially refined pulp. The bar heights represent the mean values and the error bars represent the 95 % confidence interval.
Figure 15. The strain-at-break for DALC made from pulp refined 1000, 3000, 5000, 10 000 and 15 000 revolutions prior to the modification. The bar heights represent the mean values and the error bars represent the 95 % confidence interval.
21
The tensile strength index for the sheets made from all the performed batches of DALC, from Table 3, are presented in the following two bar diagrams. Figure 16 shows the tensile strength index for the DALC sheets made from unrefined pulp, unrefined pulp with added fines, industrially refined pulp and dewatered industrially refined pulp. Figure 17 shows the tensile strength index for the DALC sheets made from the different degrees of laboratory refined pulp.
Figure 16. The tensile strength index for DALC sheets made from unrefined pulp, unrefined pulp with added fines, industrially refined pulp and dewatered industrially refined pulp. The bar heights represent the mean values and the error bars represent the 95 % confidence interval.
Figure 17.The tensile strength index for the DALC sheets made from pulp refined 1000, 3000, 5000, 10 000 and 15 000 revolutions prior to the modification. The bar heights represent the mean values and the error bars represent the 95 % confidence interval.
22
4.4 SEM images
Rapid Köthen sheets (ISO 5269-2) of DALC made from the nine different pulp compositions in Table 1, were used for the SEM imaging. Since two batches were made for each pulp composition except one, the batch with the highest strain-at-break was chosen for the SEM analysis. The images are presented in Figure 18 and Figure 19 below.
Since DALC is a ductile material the wire markings are easily imprinted into the soft pulp during the sheet forming process. All nine SEM images in Figure 18 and Figure 19 contain wire marks from the sheet forming, and there is an arrow (marked 1) pointing at typical wire marks in each image. Consistent in all nine images, except the wire marks, are external fibrillation and collapsed or processed fibres. Each image contains an arrow (marked 2) pointing at fibrils/fibrillation and an arrow (marked 3) pointing at collapsed/processed fibres. The images differ in the amount of fibrillation and to what level the fibres are collapsed.
Most of the images also contain a circle, showing small fibre pieces/fibrils bonding between fibres, these particles are potential fines. Image b, d and e in Figure 19, also contain a rectangle showing particles with a chunky shape, that could be another category of fines.
Figure 18. SEM images of DALC fibres made from a) Unrefined pulp b) Unrefined pulp with added fines c) Industrially refined pulp and d) Dewatered industrially refined pulp. The arrows marked 1 show wire marks from the sheet forming, the arrows marked 2 show fibrils/fibrillation and the arrows marked 3 show processed/collapsed fibres. The circles indicate particles that are possible fines.
23
Figure 19. SEM images of DALC fibres made from pulp refined a) 1000 revolutions b) 3000 revolutions c) 5000 revolutions d) 10 000 revolutions and e) 15 000 revolutions with a laboratory refiner. The arrows marked 1 show wire marks from the sheet forming, the arrows marked 2 show fibrils/fibrillation and the arrows marked 3 show processed/collapsed fibres.
The circles and the rectangles indicate particles that are possible fines.
24
5 Discussion
Two batches of DALC were performed for each pulp composition (Table 1), except for the pulp refined 15 000 revolutions. The obtained results therefore show important trends for the mechanical properties of the different DALC sheets. If more time would have been available, several batches could have been performed for each of the nine pulp compositions, which could have verified the obtained trends.
5.1 Mechanical properties
When the pulp is refined industrially prior to the modification, the resulting material is not very ductile. But when the pulp is highly refined with a PFI Mill prior to the modification, the resulting material shows a high ductility. The reason for this difference is probably the different formation of fines. It is known that fines are created during the refining process [11,38,39], but a previous study made by Somboon [46] showed that an industrial pilot refiner cut the fibres to a higher extent, resulting in shorter fibres and a higher value of fines, compared to laboratory refined pulp. Therefore, it could be the existence of fines that influences the ductility in a negative way. A theory made from the results obtained during this study, is that fines work as glue between the DALC fibres resulting in very strong bonds. The strong bonding will make it harder for the fibre network to be stretched out and separated, and the fibre network will therefore break more easily, resulting in lower strain-at-break. For a laboratory refined pulp the number of fines are comparatively lower than for an industrially refined pulp, and therefore the bonds between the DALC fibres are less strong and the fibre network can be stretched out to a higher extent before breaking, resulting in a higher strain-at-break.
The two bar diagrams in Figure 14 and Figure 15 show the mean strain-at-break for all the performed batches of DALC. There are significant differences between batches of the same pulp compositions. These differences occur between the two batches of DALC made from unrefined pulp (batch 1a and 1b), and between the two batches of DALC made from dewatered industrially refined pulp (batch 4a and 4b). The difference between the two batches of the other compositions are not significant, which means that the results for these batches can be considered consistent according to this study. However, the significant difference that did appear between batch 1a and 1b, and between batch 4a and 4b, are probably due to the uncertainty regarding DALC. The chemical modification process is delicate since the degree of oxidation and the actual yields can be very different between batches made from the same pulp. Both the oxidation and the reduction was performed using temperatures, amount of chemicals and reaction times optimized by BillerudKorsnäs. A higher reproducibility could possibly have been obtained by controlling these parameters more accurate and by using a more exact amount of pulp for the oxidation reaction. Another motivation to the uncertainty concerning DALC is the dispersion in the obtained values for the strain-at-break. For example, for batch 1a (DALC made from unrefined pulp) the lowest obtained value was 5.4 % while the highest obtained value was 11.5 %. The large dispersion is probably due to uneven sheets that could be caused by the sheet forming process and by an uneven pulp solution.
When comparing the strain-at-break for the DALC sheets made from the different pulp compositions in Figure 14 and Figure 15, the differences depend on which two batches that are compared, since the values differ somewhat between the two batches of the same pulp composition. However, from Figure 14 it is clear that the two values obtained for DALC made from unrefined pulp (batch 1a and 1b) are significantly higher than the values obtained for DALC made from industrially refined pulp (batch 3a and 3b) and dewatered industrially refined pulp (batch 4a and 4b). In Figure 15, showing the strain-at-break for DALC made from the different degrees of laboratory refined pulp, a trend can be distinguished between the different degrees of refinement. The strain-at-break is somewhat increased when the number of revolutions is increased. It is also clear from the figure that the value obtained for DALC made from the pulp refined 15 000 revolutions (batch 9) is significantly higher than the mean values obtained for DALC made from the pulp refined 1000, 3000 and 5000 revolutions (batch 5a, 5b, 6a, 6b, 7a and 7b).
