Accepted Manuscript Linking rheology and printability of a multicomponent gel system of carrageenanxanthan-starch in extrusion based additive manufacturing Zhenbin Liu, Bhesh Bhandari, Sangeeta Prakash, Sylvester Mantihal, Min Zhang PII: S0268-005X(18)31168-8 DOI: 10.1016/j.foodhyd.2018.08.026 Reference: FOOHYD 4607 To appear in: Food Hydrocolloids Received Date: 27 June 2018 Accepted Date: 15 August 2018 Please cite this article as: Zhenbin Liu, Bhesh Bhandari, Sangeeta Prakash, Sylvester Mantihal, Min Zhang, Linking rheology and printability of a multicomponent gel system of carrageenanxanthan-starch in extrusion based additive manufacturing, Food Hydrocolloids (2018), doi: 10.1016 /j.foodhyd.2018.08.026 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. ACCEPTED MANUSCRIPT Linking rheology and printability of a multicomponent gel system of carrageenan-xanthan-starch in extrusion based additive manufacturing Graphical abstract ACCEPTED MANUSCRIPT 1 Linking rheology and printability of a multicomponent gel system of 2 carrageenan-xanthan-starch in extrusion based additive manufacturing 3 Zhenbin Liua, Bhesh Bhandarib, Sangeeta Prakashb, Sylvester Mantihalb, Min Zhanga,c* 4 5 aState 6 Key Laboratory of Food Science and Technology, Jiangnan University, 214122 Wuxi, Jiangsu, 7 8 9 China b School of Agriculture and Food Sciences, The University of Queensland, Brisbane, QLD 4072, Australia cInternational Joint Laboratory on Food Safety, Jiangnan University, 214122 Wuxi, Jiangsu, China 10 11 *Corresponding author: Dr. Min Zhang, Professor of School of Food Science and Technology, Jiangnan 12 University, 214122 Wuxi, P. R. China. 13 Tel.: 0086-510-85877225; Fax: 0086-510-85877225; 14 E-mail: [email protected] 15 1 ACCEPTED MANUSCRIPT 17 Abstract: 3D food printing is an emerging technology with a potential to influence the food 18 manufacturing sector. Rheological properties of food inks are critical for their successful 3D printing. 19 However, the relationships between rheological properties and 3D printability have not been clearly 20 defined in food systems. In this work, a gel model system composed of carrageenan-xanthan-starch was 21 prepared for an extrusion-based 3D food printer. The 3D printing process was divided into three stages 22 and the corresponding rheological properties of inks for each stage were determined, namely extrusion 23 stage (yield stress, viscosity and shear-thinning behaviour), recovery stage (shear recovery and 24 temperature recovery properties) and self-supporting stage (complex modulus G* and yield stress at room 25 temperature). Finally, 3D printability of the model inks was systematically studied starting with printing 26 lines/pentagram (one dimensional, 1D structure) to printing lattice scaffold (two dimensional, 2D 27 structure) and finally printing cylinders (three dimensional, 3D structure). Results demonstrated that 28 addition of starch and xanthan gum in k-carrageenan based inks increased inks’ gelation temperature 29 (Tgelation), viscosity (within shear rate of 0.01 ~ 100 1/s), yield stress, G*, enhanced shear-thinning 30 (thixotropic) behaviour and reduced time-dependence of modulus (temperature recovery). Rheological 31 responses of yield stress (cross-over point where G (elastic modulus) equals to G (viscous modulus) in 32 the stress sweep tests) and shear-thinning behaviour (viscosity decreased when shear rate increased) were 33 closely related to ink’s extrudability. Inks’ gelation temperature (Tgelation) and time-dependent 34 behaviour (gelation time, tgel) significantly affected their printability and shape retention performance. 35 The mechanical strength of the ink is important to be self-supporting, especially for 3D structures. 36 Insights achieved from this study could provide guidance on improving 3D printability of foods that use 37 hydrocolloids as a printing aid. 38 Keywords: 3D printing, rheological properties, gelation temperature, printability 2 ACCEPTED MANUSCRIPT 40 41 1. Introduction 42 3D printing enable the fabrication of structures in a layer-by-layer pattern that are created in pre- 43 designed files (Rayna & Striukova, 2016). Several advantages of 3D food printing have been reported, 44 such as, customization of food structures, alteration of food texture, personalization of nutrition and use of 45 various food sources (Liu, Zhang, Bhandari, & Wang, 2017; Sun, Zhou, Huang, Fuh, & Hong, 2015). 46 Currently, there are four printing techniques that have been applied in food sector, namely extrusion 47 based printing, selective sintering printing, binder jetting and inkjet printing (Liu et al., 2017). Extrusion 48 based printing is so far the most frequently used for food printing as it is able to utilize a wide range of 49 available food inks. It has been able to fabricate structures from slurry or mashed food materials like 50 mashed potatoes (Liu, Bhandari, Prakash, & Zhang, 2018; Liu, Zhang, Bhandari, & Yang, 2018), dough 51 (Severini, Derossi, & Azzollini, 2016), chocolate (Mantihal, Prakash, Godoi, & Bhandari, 2017), gel 52 system (Cohen et al., 2009; Wang, Zhang, Bhandari, & Yang, 2017), and a combination of mashed 53 potatoes and juice gel system (Liu, Zhang, & Yang, 2018). During fabrication process, the food inks are 54 dispensed from the material reservoir by pneumatic, piston or screw driven system (Hamilton, Alici, & in 55 het Panhuis, 2018; Lille, Nurmela, Nordlund, Metsä-Kortelainen, & Sozer, 2018; Yang, Zhang, Bhandari, 56 & Liu, 2018). To improve printability in extrusion based printing, a good understanding of material’s 57 rheological properties is required. The food ink must not only be easily extruded out through a narrow 58 nozzle tip, but also possess enough mechanical strength to minimize deformation once being printed. Two 59 strategies have been proposed to achieve the desired rheological behaviour. The first strategy is to prepare 60 the inks with a low viscosity and yield stress that can be easily extruded, rapidly set through gelation 61 process once extruded from the nozzle tip and possess a high mechanical strength that resists deformation. 3 ACCEPTED MANUSCRIPT 62 The second strategy is to prepare the inks with desired rheological properties, like shear-thinning 63 behaviour which is desirable during extrusion and possess sufficient mechanical strength to withstand 64 deposited structure (M’Barki, Bocquet, & Stevenson, 2017). In bio-printing, it has been reported that 65 shear-thinning and thermo-reversible behaviour of inks are highly desirable. The shear-thinning behaviour 66 enables the easy extrusion of inks. While the thermo-reversible behaviour enables the inks to quickly 67 achieve enough yield strength to be capable of self-supporting through gelation (Wilson, Cross, Peak, & 68 Gaharwar, 2017; Zhang et al., 2015). Usually, these rheological behaviour are often closely related with 69 each other and cannot be varied separately. Thus a deep understanding of rheological properties of inks is 70 important to enable a successful printing. Some researchers have investigated the effect of rheological 71 properties on 3D printing behaviour. They highlighted the importance of the rheological behaviour of inks 72 on 3D printability, like shear-thinning character, viscosity and yield stress (Li, Liu, & Lin, 2016; Liu et 73 al., 2018; Sweeney, Campbell, Hanson, Pantoya, & Christopher, 2017). For some materials with thermo- 74 responsive behaviour, like thermo-reversible hydrogels of k-carrageenan and chocolate, the 75 gelation/solidify temperature is critical to determine appropriate printing temperature (Chung et al., 2013; 76 Holzl et al., 2016; Mantihal et al., 2017; Suntornnond, An, & Chua, 2017). Although many researchers 77 have emphasized the importance of rheological properties on 3D printing behaviour, few researchers have 78 correlated them during the whole printing process, starting from extrusion to the final self-supporting 79 stage. 80 As 3D food printing is a very recent development, investigations focused on the fabrication of edible 81 gels by 3D food printing technique are very limited (Cohen et al., 2009; Valérie Vancauwenberghe et al., 82 2017; Wang et al., 2017). Insights achieved from investigations involving 3D printing of edible gels could 83 deepen our understanding about many 3D food printing processes which use hydrocolloids as additives to 84 improve printing performance. Carrageenan and xanthan gum are widely used in the food industry, 4 ACCEPTED MANUSCRIPT 85 especially in starch-based food products (Chaudemanche & Budtova, 2008; Fakharian et al., 2015; 86 Gladkowska-Balewicz, Norton, & Hamilton, 2014; Lin, Liang, & Chang, 2016; Liu et al., 2018; Mandala 87 & Bayas, 2004). The k-carrageenan is a linear sulfated polysaccharide derived from red algae with 88 alternating 3,6-anhydro-D-galactose and β-D-galactose-4-sulphate repetitive units (Tecante & Santiago, 89 2012). This polymer illustrates rapid thermo-reversible behaviour and can form a firm and brittle gel 90 structure. It can be easily dissolved in water when heated, presenting a random coil structure. During 91 cooling process, it undergoes a conformational transition from coils to double helices followed by the 92 aggregation of helices. This results in the formation of a brittle gel structure stabilized by hydrogen 93 bonding between galactose (Liu, Chan, & Li, 2015; Wilson et al., 2017). Xanthan gum is an extracellular 94 polysaccharide secreted by the micro-organism Xanthomonas campestris. Xanthan gum solutions are 95 highly pseudoplastic with a strong shear-thinning and recovery behaviour. Viscosity rapidly decreases 96 when shear stress increases, but after removal of shear force, the initial viscosity is recovered almost 97 immediately. This property related to the formation of intermolecular aggregates by xanthan molecules is 98 through hydrogen bonding in solution. The highly ordered network of entangled and stiff molecules 99 presents a high viscosity at low shear stress, while it is progressively disrupted when shear stress is 100 applied (Graham, 2009; Mandala et al., 2004; Morrison, Clark, Talashek, & Yuan, 2004). The strong 101 shear-thinning and rapid recovery behaviour are highly desirable in extrusion based 3D printing as shear- 102 thinning behaviour enables the easy extrusion of inks through a narrow opening and the rapid recovery 103 behaviour allows the inks to quickly achieve enough mechanical strength after printing to resist 104 deformation. 105 This work was aimed at improving our understanding of correlation between rheological properties and 106 3D printability and provide some information on 3D food printing of carrageenan-xanthan-starch 107 multicomponent system. In this study, we firstly investigated the thermo-responsive behaviour of inks, 5 ACCEPTED MANUSCRIPT 108 following which several formulations were selected for determination of rheological properties and 109 evaluation of 3D printability. The 3D printing process was divided into three stages, namely extrusion 110 stage, recovery stage and self-supporting stage, and corresponding rheological properties of inks for each 111 stage were determined. Finally, 3D printability of model ink formulations was evaluated based on the 112 ability to print 1D structure (line, pentagram), 2D structure (lattice scaffold) and 3D structure (cylinder). 113 114 2. Materials and methods 115 2.1 Materials 116 Food grade xanthan gum (Batch: 345216) and kappa-carrageenan (k-carrageenan, Batch: 336115) were 117 purchased from Melbourne Food Ingredient Depot Co. Ltd, Brunswick, Australia. Xanthan gum and k- 118 carrageenan contain no fillers, preservatives, bulking agents and flavors. Potato starch with moisture 119 content of 16.52±0.23 g/100g (w.b.) was provided by Simplot Co. Ltd, Devonport, Australia. 120 2.2 Inks preparation 121 Ink formulations were prepared by following the procedure as shown in Fig. 1A. Required amount of 122 k-carrageenan, xanthan gum, and potato starch were firstly well mixed and dissolved in water whilst 123 constant stirring. The mixture (100 mL) were stirred for 30 min at 1200 r/s using a stirrer ([email protected] RW 20, 124 John Morris Scientific Pty Ltd, Australia). During this process, 0.1% food color was incorporated to the 125 mixture system. Afterwards, the mixtures were incubated for 30 min at 90°C using a water bath and were 126 homogenized for 4 min at 3000 r/s (5702 R, Eppendorf Co. Ltd, Germany) to remove the air bubbles 127 introduced during the stirring process. For the formulation containing only potato starch, continuous 128 stirring was conducted when heating to avoid syneresis/separation of water. The prepared inks were then 129 stored in a beaker in a refrigerator at 4 °C and evaluated within two days. 6 ACCEPTED MANUSCRIPT 130 A 23 factorial design was used to prepare the ink formulations. Tab. 1 describes the concentration levels (-1, 131 +1) for k-carrageenan, xanthan gum, and potato starch. The experimental matrix including eight 132 formulations in total can be seen in Tab. 2. In addition, as a comparative study, inks containing only 0.25% 133 xanthan gum, 1% k-carrageenan, and 2% potato starch were also prepared. 134 2.3 Rheological analysis 135 The rheological behaviour of ink formulations were characterized using an AR-1500 rheometer (TA 136 Instruments Ltd, UK) with a plate–plate geometry (40 mm diameter, 0.2 mm gap) and a very thin layer of 137 silicon oil in the trap to prevent the moisture evaporation. The formulations were firstly heated up to 60°C 138 before testing to form a flowable state. A plastic dropper was used to load samples with same number of 139 drops. Afterwards, the samples were equilibrated at initial measurement temperature for 7 min to reach a 140 steady state, unless otherwise stated. The shear-viscosity tests were conducted in flow ramp mode with 141 the shear rate increasing from 0.01 to 100 1/s at temperature of 35°C, 40°C and 45°C. The yield stress 142 measurements were performed in oscillation stress sweep mode at 1 Hz from 1 to 1000 Pa (at testing 143 temperature of 35°C, 40°C and 45°C) or to the upper limit of stress achievable for the rheometer (at 144 25°C). The yield stress of inks was determined as the cross-over point where elastic modulus (G) equal to 145 viscous modulus (G) (Wilson et al., 2017). Temperature ramps were performed at 10 s-1 from 65°C to 146 30°C with a cooling rate of 3°C/min. Gelation temperature (Tgelation) was determined by extrapolating 147 the high and low temperature asymptotes of the viscosity and specifying the temperature at which these 148 intersects (Barrera, Florián-Algarin, Acevedo, & Rinaldi, 2010). Rotational recovery tests were applied to 149 characterize the shear recoverability of inks at 40°C by using a low shear rate of 1 1/s for 180 s, followed 150 by a high shear rate at 100 1/s for 120 s and finally at a low shear rate of 1 1/s for 180 s. Shear 151 recoverability of inks was determined as the percentage of viscosity obtained during the first 30s in the 152 third step after high shear rate (100 1/s) based on the average viscosity obtained in the first step 7 ACCEPTED MANUSCRIPT 153 (Achayuthakan & Suphantharika, 2008). Temperature recovery tests (coined by the authors) were 154 conducted to characterize the time dependence of complex modulus (G*, G*=( G2+ G2)0.5) when 155 temperature changed from printing temperature (35°C, 40°C, 45°C and 50°C) to room temperature 156 (25°C). Frequency sweep analysis was performed at 25°C with angular frequency ranging from 1 to 100 157 rad/s at a constant deformation of 0.1% strain (within the linear viscoelastic range, LVR). Three replicates 158 were conducted and the average data were used to plot the curves. 159 2.4 3D printing process 160 A syringe type extrusion-based 3D printer (Choc Creator 2.0 Plus, Choc Edge Co. Ltd, UK) was used 161 to assess the printability of the inks. The desktop printer comes with a syringe-based deposition unit. The 162 syringes is inserted into one material barrel. A heating unit is used to control the printing temperature. The 163 printer could be controlled by a computer or the graphical user interface unit. A photograph of the printer 164 was given in Fig. 1B. 165 Room temperature of 25°C was applied in the printing process. The printing temperature was set at 166 35°C, 40°C, 45°C and 50°C through the temperature-controlled nozzle and the platform was held at room 167 temperature (25°C). The inks were firstly heated up to 60°C to form a flowable state and were then 168 poured into a metal syringe. During this process, it was made sure that no air bubble was included. The 169 syringes filled with inks were then incubated at printing temperature (35°C, 40°C, 45°C and 50°C) for at 170 least 1 hour prior to be used for deposition process. Printability tests, that is 1D printing (line, one-layer 171 pentagram), 2D printing (three-layer of lattice scaffold) and 3D printing (hollow cylinder) were printed at 172 the same condition (nozzle diameter 0.8 mm, layer height 0.8 mm, federate and printing rate 22 mm/s, 173 and solid infill). The stereolithography (stl) models (line, pentagram, lattice scaffold, and hollow cylinder) 174 were designed with Rhinoceros 5.0 (educational version, Robert McNeel & Associates, Seattle, US) and 175 detailed information are shown in Fig. 1C and Fig. 1D. 8 176 2.5 Printability evaluation ACCEPTED MANUSCRIPT 177 The printability of inks under various printing temperature was evaluated by consistency of lines, 178 average area of interconnected channels in lattice scaffold, and wall thickness of hollow cylinder. Images 179 of the printed constructs were captured once the fabrication process was completed using a 16 megapixel 180 (16 MP (f/2.0, 1/2.8", 1.12µm) camera (Nubia Z11, Nubia Technology Co., Ltd., Shenzhen, China) with a 181 ruler beside the construct. The average area of interconnected channels within lattice scaffolds and wall 182 thickness of hollow cylinder were analyzed using the ImageJ software. The images were firstly set a scale 183 and were then converted to 8 bit. After adjusting the threshold, the area of interconnected channels of 184 lattice scaffold could be automatically measured and analyzed. The “Straight Line” tool was used to 185 calculate the wall thickness of hollow cylinder. 186 2.6 Statistical analysis 187 188 The statistical software SPSS was used to analyze data and Origin 9.0 was used to plot the graphs. Differences of p < 0.05 were considered to be significant with the application of Duncan’s test. 189 190 3 Results and discussion 191 3.1 Choice of materials and their properties 192 In extrusion-based 3D printing, inks must easily flow through a nozzle tip, and should also possess 193 enough mechanical strength to resist deformation after deposition. For the inks to have desired rheological 194 properties, previous researchers have prepared inks with sufficient yield strength that enable the printed 195 constructs capable of self-supporting, such as mashed potatoes (Liu et al., 2018), baking dough (Yang, 196 Zhang, Prakash, & Liu, 2018) and orange juice gel (Roknul, Zhang, Mujumdar & Yang, 2018). However, 197 the inks with high mechanical strength always characterized by high viscosity and yield stress, which 198 could easily block the nozzle tip and result in the failure of printing process. To address this problem, inks 9 ACCEPTED MANUSCRIPT 199 with a strong shear-thinning and a rapid shear recovery properties are required (Paxton et al., 2017). 200 Another desirable property of ink is thermo-responsive behaviour, which means that ink’s viscosity is 201 reduced considerably at a high temperature but rapidly develops significant mechanical strength to resist 202 deformation once being extruded out from a nozzle due to the temperature difference between nozzle 203 cavity and fabrication platform. This method is frequently used in chocolate printing (Mantihal et al., 204 2017) and thermo-reversible gels printing (Ouyang, Yao, Zhao, & Sun, 2016; Wilson et al., 2017). In our 205 study, the two strategies of strong shear-thinning behaviour/rapid shear recovery and thermo-responsive 206 behaviour were adopted to fabricate the multicomponent gel system comprised of k-carrageenan, xanthan 207 gum and starch. 208 k-carrageenan and xanthan gum are both widely used in the food industry, especially in starch-based 209 food products (Chaudemanche et al., 2008; Fakharian et al., 2015; Gladkowska-Balewicz et al., 2014; Lin 210 et al., 2016; Mandala et al., 2004). As previously mentioned in the introduction k-carrageenan has a rapid 211 thermo-reversible behaviour while xanthan gum is highly pseudoplastic with a strong thixotropic and 212 recovery behaviour. Hence, it would be interesting to observe how these two ingredients with the desired 213 characteristics required for good printability, perform in a multicomponent gel system of k-carrageenan- 214 xanthan-starch. 215 Because the ink illustrates a thermo-reversible characteristic, it is important to determine its gelation 216 temperature (Tgelation) to set a proper printing temperature in the nozzle cavity. Therefore, in our work, 217 Tgelation of inks were firstly determined to select suitable ink formulations for further rheological tests 218 and 3D printability evaluation as the upper limit temperature of nozzle cavity was 50°C. Subsequently, 219 the rheological properties of the selected inks were correlated with their 3D printability. We divided the 220 extrusion-based printing process into three stages, and rheological properties of inks corresponding to 221 each stage were evaluated (Fig. 1A). 10 ACCEPTED MANUSCRIPT 222 The first stage of printing is an extrusion stage. The inks’ viscosity, yield stress and thixotropic (shear- 223 thinning) behaviour were characterized to evaluate their extrudability during the extrusion process. The 224 second stage is the recovery stage. During this stage, the inks were exposed to high shear rates and 225 experienced temperature variations when they were extruded out from the nozzle cavity (printing 226 temperature: 35°C, 40°C, 45°C and 50°C) to the deposited platform at room temperature (25°C). Here, we 227 define the shear recovery to mimic the recoverability of inks after being experienced the high shear rates 228 and the temperature recovery to characterize the time dependence of viscoelastic behaviour of inks during 229 temperature variations. Therefore, in a rheological test mimicked shear recovery tests applying an 230 alteration of high and low shear rates to the inks and temperature recovery tests applying an alteration of 231 printing temperature and room temperature were conducted in this stage. The third is self-supporting 232 stage. After finishing the deposition process, inks would be exposed at room temperature. During this 233 stage, the yield stress and G* are both important in determining the self-supporting behaviour of 234 deposited parts. Finally, we evaluated the printability of inks based on 1D structure (line, pentagram), 2D 235 structure (lattice scaffold) and 3D structure (cylinder), and correlated the printability behaviour of inks 236 with their rheological properties (Fig. 1A). 237 238 3.2 Thermo-reversible characteristics of inks 239 The thermo-responsive behaviour of inks with different compositions were evaluated by rheological 240 tests. As shown in Fig. 2A, viscosity of 1% k-carrageenan solution (kC(1)) increased rapidly at 241 temperature below 242 transition from coils to double helices followed by the aggregation of helices (Liu, et al., 2015; Parker, 243 Brigand, Miniou, Trespoey, & Vallée, 1993; Shchipunov, 2003). The cross-linking in k-carrageenan gel is 244 associated with the formation of double helices and their aggregates. This process is partially governed by 35°C due to the gelation of k-carrageenan during cooling involving a conformational 11 ACCEPTED MANUSCRIPT 245 micro crystallization, and the double helices constraining further aggregation and propagation during 246 gelation evolution (Liu et al., 2015). Comparatively, the viscosity of the 2% starch solution (S(2)) and 247 0.25% xanthan solution (X(0.25)) changed little when temperature decreased, indicating that they were 248 not responsible for the thermo-responsive behaviour of multicomponent inks. In addition, the inks 249 containing k-carrageenan (kC(1)X(0.25) and kC(1)X(0.25)S(2)) represented similar viscosity change 250 profiles as kC(1), indicating that k-carrageenan played a more important role in the thermo-responsive 251 behaviour of the multicomponent inks. 252 As shown in Fig. 2B, the gelation temperature (Tgelation) was recorded from the viscosity 253 measurements from temperature ramp-down curve by extrapolating the high and low temperature 254 asymptotes of the viscosity and specifying the temperature at which these intersect (Barrera et al., 2010). 255 As expected, Tgelation increased significantly when k-carrageenan concentration increased (Tab. 3). The 256 Tgelation of inks containing 2% k-carrageenan occurred above 44°C, while the inks containing 1% k- 257 carrageenan showed a gelation point below 40°C. This was because an increased k-carrageenan 258 concentration would increase the amount of available interaction sites for hydrogen bonding between k- 259 carrageenan polymer chains and thus making it easier for the formation of double helices and their 260 subsequent aggregation (Fakharian et al., 2015; Gladkowska-Balewicz et al., 2014; Wilson et al., 2017). 261 During extrusion tests it was found that inks containing 2% k-carrageenan easily gelled due to the high 262 Tgelation and high viscosity, causing the nozzle tip to block. Therefore, only the inks containing 1% k- 263 carrageenan were selected for further evaluation of rheological properties and printability tests. 264 As seen from Fig. 2A and Tab. 3, starch addition led to the increase of viscosity and Tgelation. This 265 might be related to the excluded volume effect of starch during which the presence of swollen starch 266 molecules leads to the reduction of accessible water molecules which are held by k-carrageenan. This 267 phenomenon results in the formation of some zones with elevated k-carrageenan concentration, 12 ACCEPTED MANUSCRIPT 268 facilitating the increase of viscosity, elastic modulus and the rates of ordering and aggregation 269 (Chaudemanche et al., 2008; Mandala et al., 2004). The incompatibility between polymers when one of 270 the ingredients acts as gelling agent could result in the dramatic change in gel properties, including an 271 increased Tgelation and modulus (Fakharian et al., 2015; Lorenzo, Zaritzky, & Califano, 2015). 272 Addition of xanthan gum also increased the Tgelation and viscosity of inks (Fig. 2A and Tab. 3). This 273 was because the conformational transition of xanthan molecules when temperature decreased from a more 274 flexible and disordered structure to a rigid ordered state which can form intermolecular associations and 275 network structure of weakly bound molecules (Graham, 2009; Morris, Rees, Young, Walkinshaw, & 276 Darke, 1977; Morrison et al., 2004). 277 278 3.3 Rheological properties of inks governing extrusion stage 279 3.3.1 Flow initiation analysis determined by yield stress at printing temperature 280 The extrusion process in 3D printing is easier for ink with a low yield stress (Sweeney et al., 2017). 281 Yield stress also reflects the mechanical strength of inks as it supports subsequent stacked layers during 282 post-printing condition (Liu et al., 2018). In our study, as the inks went through a rapid temperature 283 change from printing temperature in nozzle cavity (35°C, 40°C, 45°C, and 50°C) to room temperature at 284 deposited platform (25°C), the flow stresses were determined at printing temperature and room 285 temperature, respectively. 286 Yield stress (flow stress) is critical to ink’s extrudability as it closely relates to the minimum force 287 required to initiate a flow of ink (Li et al., 2016; Paxton et al., 2017). The minimum pressure required to 288 start a flow for the ink with a yield stress can be described by 289 Sweeney et al., 2017). Where Pmin means the minimum pressure required, L means the nozzle length, D 290 means the nozzle diameter, and τyield means the yield stress of inks. During extrusion processes, a low 13 (Pospischil et al., 2014; ACCEPTED MANUSCRIPT 291 yield stress is highly desirable as the extrusion is not continuous but starts and stops frequently during 292 printing. A high yield stress requires the printer to produce a larger force to exceed the yield stress. This 293 necessities a more robust extrusion system to be capable of constantly generating such high pressure in a 294 controlled way. (Lewis, 2006; Lewis, Smay, Stuecker, & Cesarano, 2006). 295 Yield stresses of inks was determined as the cross-over point where G equals to G (Fig. 2C) (Wilson 296 et al., 2017). The yield stress of inks at different printing temperature are illustrated in Tab. 4. As 297 expected, yield stresses for inks were decreased when temperature increased. In particular, for the inks of 298 kC(1)X(0.5)S(2) and kC(1)X(0.25)S(2), yield stresses decreased sharply from 551.3 Pa to 36.9 Pa and 299 from 134.2 Pa to12.0 Pa when temperature increased from 35°C to 45°C, indicating that a higher 300 extrusion force is required to initiate the flow of inks at a low temperature. The thermo-responsive 301 behaviour of k-carrageenan contributed the decrease of yield strength. Tgelation of inks of 302 kC(1)X(0.5)S(2) and kC(1)X(0.25)S(2) were 38.2°C and 36.9°C (Tab. 3), respectively. The k- 303 carrageenan molecules dissolve in water and present a random coil structure when temperature is above 304 Tgelation, while, at below Tgelation, they change to double helices followed by the aggregation of helices 305 finally forming a gel (Tecante et al., 2012; Wilson et al., 2017). 306 The increased amount of xanthan gum increased the yield stress of inks. For example, at 35°C, the 307 yield stress was 39.2 Pa for kC(1)X(0.25)S(0) and 80.2 Pa for kC(1)X(0.5)S(0), respectively. Addition of 308 2% starch also increased the yield stress from 80.2 Pa for kC(1)X(0.5)S(0) to 551.4 Pa for 309 kC(1)X(0.5)S(2). This increase might be attributed to the excluded volume effect of starch, where the 310 existence of starch molecules in the mixture results in a decrease of the accessible water molecules that 311 can be captured by k-carrageenan. Thus, a relative increased k-carrageenan concentration enhanced its 312 aggregation rates resulting in increased yield stress and viscosity (Chaudemanche et al., 2008; 313 Gladkowska-Balewicz et al., 2014; Lin et al., 2016). 14 ACCEPTED MANUSCRIPT 314 315 3.3.2 Shear-thinning behaviour 316 After overcoming the yield stress, the required pressure to maintain a continuous flow of inks would 317 depend on their viscosity and shear-thinning behaviour. One model that frequently used for shear-thinning 318 fluids is the power-law ( 319 Where ƞ means viscosity, K means the consistency index, γ̇ means shear rate, and n means flow index that 320 describes the material’s properties of shear thinning, thickening or remain Newtonian. For n < 1 fluids 321 shear thin and a smaller n indicates a more highly shear-thinning behaviour (Paxton et al., 2017; Wilson et 322 al., 2017). Ink’s shear-thinning behaviour could be quantified by fitting the viscosity profiles (Fig. 3) to 323 Power Law equation. The coefficients K and n (Tab. 4), could help us to better understand the ink’s flow 324 behaviour according to the equation 325 tube, P means the pressure drop, and l means the length of tube length. A low K and n are advantageous 326 to require a low pressure for a continuous flow (Lewis, 2006; Lewis et al., 2006; Liu et al., 2018). The 327 shear viscosity and stress profiles of inks at different printing temperature are illustrated in Fig. 3. All inks 328 present shear-thinning behaviour, indicated by a reduction in viscosity and an increase in shear stress 329 when shear rate was increased. For all inks, K decreased over an increasing temperature. Specifically, for 330 the ink of kC(1)X(0.5)S(0), the K were 6.19 Pa·sn, 3.14 Pa·sn, and 2.68 Pa·sn at temperature of 35°C, 331 40°C, and 45°C, respectively. Again, this was related to the thermo-reversible behaviour of k-carrageenan 332 (Wilson et al., 2017; Yang, Yang, & Yang, 2018). ) (Rezende, Bartolo, Mendes, & Filho, 2009; Sweeney et al., 2017). . Where Q means the flow rate, R means the radius 333 Addition of xanthan gum increased K at all tested temperature (Tab. 4). Added 2% potato starch also 334 increased ink’s viscosity (Fig. 3) and K (Tab. 4). Again, this might be due to the excluded volume effect 335 of starch in the system as discussed above (Chaudemanche et al., 2008; Gladkowska-Balewicz et al., 336 2014; Lin et al., 2016). 15 ACCEPTED MANUSCRIPT 337 As seen from Tab. 4, the flow index (n) was less than one, suggesting that inks exhibits shear thinning 338 behaviour. The n presents no regular change with temperature variations. However, addition of xanthan 339 gum and potato starch obviously decreased the values of n, meaning that shear thinning was more 340 pronounced. Specifically, at 35°C and 45°C, n decreased from 0.58 and 0.71 for kC(1)X(0.25)S(0) to 0.50 341 and 0.61 for kC(1)X(0.5)S(0) due to the increased amount of xanthan gum, and from 0.58 and 0.71 for 342 kC(1)X(0.25)S(0) to 0.47 and 0.48 for kC(1)X(0.25)S(2) due to the addition of starch. As reported, the 343 xanthan gum solutions present highly pseudoplastic behaviour. An increase of shear rate will 344 progressively reduce the viscosity, but upon the removal of shear stress the viscosity will recover rapidly. 345 This was because xanthan molecules can form intermolecular aggregates through hydrogen bonding and 346 polymer entanglements that presents high viscosity at low shear rates, and the structure are disrupted 347 rapidly with the application of high shear rates (Graham, 2009; Morrison et al., 2004). A more 348 pronounced shear thinning effect was also reported after addition of starch in k-carrageenan system 349 (Fakharian et al., 2015). 350 351 3.4 Rheological properties of inks governing recovery stage 352 3.4.1 Shear recovery behaviour 353 During printing, the inks experience the shear force and temperature changes when being extruded out 354 from the nozzle tip. Thus, we divided the ink’s recovery properties into shear recovery and temperature 355 recovery. 356 Here, as illustration, the shear recovery tests were conducted at a constant temperature of 40°C to 357 investigate the effect of individual shear forces on the viscosity changes. Thixotropic inks are highly 358 desirable for extrusion printing. In these inks, the viscosity is reduced rapidly with the application of shear 359 force but recovers quickly after the removal of shear force (Lille, Nurmela, Nordlund, Metsä-Kortelainen, 16 ACCEPTED MANUSCRIPT 360 & Sozer, 2017). To mimic the shear conditions during extrusion printing, the inks were firstly subjected to 361 a low shear rate of 1 1/s to simulate the low extrusion rate of initial extrusion state in the syringe barrel 362 part. Subsequently, the shear rate was increased to 100 1/s to simulate the shear force generated in the 363 nozzle tip part during extrusion. Finally, the shear rate was again decreased to 1 1/s to test the recovery 364 behaviour of inks to simulate post-printing condition because the rheometer cannot test viscosity at zero- 365 shear. Fig. 4A illustrates the shear recovery behaviour of viscosity of the inks under an alterations of low 366 and high shear rates. All inks’ viscosity decreased significantly at high shear rate and recovered rapidly at 367 low shear rate, and quickly reached a stabilized state within 30s. In the case of kC(1)X(0.5)S(0), viscosity 368 dramatically decreased from 4.33 Pa.s at 1 1/s to 0.36 Pa.s at 100 1/s, and then rapidly built up to 3.92 369 Pa.s in 10s (90.53% recovery of the initial viscosity). If it was allowed 30s, a 98.91% recovery was 370 obtained. For the ink kC(1)X(0.5)S(2), the recovery percentage was 75.56% in 10s and was 83.09% in 371 30s, but it could not increase further even it was given a longer time. This rapid and reversible viscosity 372 response make the inks suitable for 3D printing as they could be easily extruded out while rapidly 373 recovered enough mechanical strength necessary to support the next extruded layer. Fig. 4B shows the 374 recovery percentage of inks within 30s after removal the high shear rate (100 s-1). After addition of starch, 375 the recovery decreased from 91.30% for kC(1)X(0.25)S(0) to 78.50% for kC(1)X(0.25)S(2), from 98.91% 376 for kC(1)X(0.5)S(0) to 83.09% for kC(1)X(0.5)S(2), respectively. This might be related to the excluded 377 volume effect of starch in the multicomponent gel system (Chaudemanche et al., 2008; Mandala et al., 378 2004). The incompatibility between polymers when one of the ingredients acts as gelling agent could 379 result in the dramatic change in recoverability (Fakharian et al., 2015; Lorenzo et al., 2015). 380 Advantageously, with the increased amount of xanthan gum, the recovery percentage increased form 381 91.30% for kC(1)X(0.25)S(0) to 98.91% for kC(1)X(0.5)S(0), from 78.50% for kC(1)X(0.25)S(2) to 382 83.09% for kC(1)X(0.5)S(2), respectively. The increase of cross-linking density by physical bonds 17 ACCEPTED MANUSCRIPT 383 because of the increased addition of xanthan gum might contribute to the reversible-structure behaviour. 384 In addition, this was also probably because the formation of intermolecular aggregates of xanthan 385 molecules stabilized by hydrogen bonding presents a high viscosity at low shear rates, and the structures 386 are disrupted rapidly with the application of high shear rates, but upon the removal of shear stress the 387 viscosity recovered rapidly (Graham, 2009; Morrison et al., 2004). 388 3.4.2 Temperature recovery behaviour 389 During 3D printing process, the inks experienced a rapid temperature change from printing temperature 390 (nozzle cavity temperature of 35°C, 40°C, 45°C, and 50°C) to the room temperature (25°C). The gelation 391 kinetics of inks (temperature recovery) were investigated through a temperature cycle test (Fig. 5). To 392 simulate the temperature changes experienced by the inks during 3D printing process, the complex 393 modulus (G*, it provides an overall deformation resistance (Huang, 2018)) of inks were firstly recorded at 394 printing temperature (35°C, 40°C, 45°C, and 50°C) for 120s, and the gelation behaviour of inks were then 395 recorded at room temperature (25°C) for 300s,. Generally, the inks responded rapidly to the decrease of 396 temperature indicated by the sharp increase of G*, but it took a certain time for the inks to reach a plateau 397 in G*. The time necessary to reach the G* plateau was defined as tgel. It was determined as the crossover 398 point of two tangents respectively fitted to the plateau-region and G*-increase regions of the curves at 399 room temperature (25°C), as shown in Fig. 5A. We emphasize that the length of tgel is critical to the 400 shape retention of extruded filaments and final printed parts. A long tgel would definitely cause the 401 spreading deformation of filaments and printed part as inks cannot achieve a sufficient mechanical 402 strength within a short time to be capable of self-supporting. The tgel of inks at different temperature is 403 shown in Fig. 5F. As expected, for all inks the tgel increased with an increasing temperature. In 404 particular, the tgel increased from 77s at 35°C to 98s at 50°C for the ink of kC(1)X(0.25)S(2). This was 405 obvious because biopolymer molecules require a long time to form a gel like structure after being 18 ACCEPTED MANUSCRIPT 406 subjected to a high temperature. Note that the tgel decreased with the addition of starch, for example at 407 45°C, the tgel decreased from 106s for kC(1)X(0.25)S(0) to 91s for kC(1)X(0.25)S(2) and from 96s for 408 kC(1)X(0.5)S(0) to 85s for kC(1)X(0.5)S(2), respectively. This is probably because starch molecules are 409 not thermo-responsive and their excluded volume effect lead to a reduction of the accessible water for k- 410 carrageenan molecules. This resulted in the formation of a relative partial high concentration of k- 411 carrageenan making it easier to form a gel (Chaudemanche et al., 2008; Gladkowska-Balewicz et al., 412 2014; Lin et al., 2016). In addition, the interfacial phenomena between starch granules (solid phase) and 413 the continuous phase medium might be also accounted for this fact, as the interfacial surface-tension is 414 greater at lower temperature which can produce a rigid interface at the boundary solid-liquid phases (Shi 415 & BeMiller, 2002). With the increase of xanthan gum, the tgel also decreased. This may be related to the 416 conformational transition of xanthan molecules from a more flexible and disordered state to a rigid rod 417 state of network structure (Graham, 2009; Morrison et al., 2004). 418 419 3.5 Rheological properties of inks governing self-supporting stage 420 To be capable of self-supporting after deposition, the complex modulus (G*) and yield stress are 421 important (Huang, 2018; Liu et al., 2018; Wilson et al., 2017). G* means solid-like property reflecting the 422 ability to resist compression deformation and the mechanical strength (Huang, 2018). Apparently, the inks 423 with sufficient mechanical strength and yield stress illustrate excellent self-supporting behaviour (Lille, et 424 al., 2017; Yang et al., 2018). As seen from Fig. 6, G′ was obviously higher than viscous modulus (G″) and 425 they both exhibited frequency-dependent behaviour. The values of G′, G″ and G* at 10 rad/s for inks are 426 presented in Tab. 3. Increased amount of k-carrageenan significantly increased the G* from 64.6 Pa for 427 kC(1)X(0.25)S(0) to 694.3 Pa for kC(2)X(0.25)S(0), while decreased the tan (tan= G″/ G′) from 0.77 to 428 0.20, respectively. Similar trend for all inks were observed. This was expected since at a high 19 ACCEPTED MANUSCRIPT 429 concentration k-carrageenan is easier to transform into a denser gel state due to the formation of a network 430 structure induced by the aggregation of double helices (Liu et al., 2015). Addition of potato starch also led 431 to the significant increase of G* for all inks. Again, this might be because the excluded volume effect of 432 starch in the system decreased the amount of accessible water molecules to k-carrageenan thus increasing 433 its partial concentration (Chaudemanche et al., 2008; Gladkowska-Balewicz et al., 2014; Lin et al., 2016). 434 The G* for the ink of kC(1)X(0.25)S(2) rapidly increased from 590.1 Pa to 836.6 Pa when xanthan gum 435 increased by 0.25 g (kC (1)X(0.5)S(2)), while their solid-like character was similar with tan value of 436 0.21 for kC(1)X(0.25)S(2) and 0.23 for kC(1)X(0.5)S(2). This means that the increase in the amount of 437 xanthan in presence of starch produced more rigid networks maintaining similar degree of viscoelasticity. 438 However, the increased amount of xanthan in the inks without starch significantly increased G′, G″ and 439 G* but tan noticeable decreased forming a more compact network. Thus, the presence of starch does not 440 improve the energy stability of bonds in carbohydrate network, which could be related with interfacial 441 effects. 442 Another important parameter reflecting the resistance to deformation of inks is yield stress at self- 443 supporting stage (Lille, et al., 2018; Liu et al., 2018). During this stage, yield stress reflects the ink’s 444 performance to retain its shape under gravity and under the force given by upper material layers on top of 445 it. Many publications have reported that G′, G* and yield stress were critical to self-supporting ability and 446 maintain deposited information (Lille et al., 2018; Zhang et al., 2015; Huang, 2018). Here, please note 447 that we tested the yield stress of inks at printing temperature (35°C, 40°C, 45°C, and 50°C) and at room 448 temperature (25°C). We emphasize that the yield stress at printing temperature was used to describe the 449 extrusion behaviour during extrusion stage (section 3.3.1), while the yield stress at room temperature was 450 used to describe the self-supporting behaviour during post-printing. An oscillation stress sweep at 1 Hz 451 from 1 to the upper limit of stress achievable for the rheometer (at 25°C) was conducted. It was found that 20 ACCEPTED MANUSCRIPT 452 the yield stress for all inks was above 1000 Pa (data not shown), while the upper limit of stress of the 453 rheometer was around 2000 Pa. Therefore, it is possible that the yield stress obtained at such condition 454 might not be reliable thus it was not reported. 455 It is worth mentioning that the G′, G* and yield stress obtained at room temperature (25°C) might not 456 really reflect the deformation resistance performance as they were determined at a steady state at 25°C. 457 While, in our tests, inks went through a rapid temperature change from printing temperature (35°C, 40°C, 458 45°C, and 50°C) to room temperature (25°C). It’s not possible for the inks to reach a steady state at such a 459 short time (tgel, section 3.4.2). However, we emphasize that they can still give us useful information 460 about ink’s self-supporting behaviour for the deposited objects after reaching a steady state and in the 461 subsequent storage period. 462 3.6 Printability evaluation of inks 463 3.6.1 One-dimensional structure (1D structure) 464 To evaluate ink’s printability, a variety of constructs was printed. Fig. 7 shows some images of 1D 465 structures of lines and pentagram with different angles. As seen from Fig. 7A, some broken lines with a 466 very rough surface were captured when the ink of kC(1)X(0.5)S(2) was printed at 35°C. This was because 467 its high yield stress and K at 35°C resulting in its poor extrudability. At 35°C, the yield stress and K of 468 kC(1)X(0.5)S(2) were 551.3 Pa and 23.88, respectively (Tab. 4), significantly higher than those values at 469 other temperatures. As described in section 3.3.1, the yield stress is important to enable ink’s extrudability 470 from a narrow opening as it reflects the minimum pressure required to initiate the flow of inks, below 471 which the inks show more solid-like behaviour rather than liquid-like behaviour. Inks cannot be extruded 472 out unless the driving pressure larger than yield stress (Li et al., 2016; Paxton et al., 2017). A high K also 21 ACCEPTED MANUSCRIPT 473 showed an adverse effect to maintain the flow of inks as discussed in section 3.3.2 (Lewis, 2006; Lewis et 474 al., 2006). 475 Pentagrams with different angles were printed using ink of kC(1)X(0.25)S(2) (Fig. 7B). The overlap 476 problem was more serious in the fabrication of pentagram with a sharper angle (Fig. 7B(b)). As seen from 477 Fig. 7B(a), some areas were overlapped and amount of inks was doubled, which would cause the 478 accumulation of layer height errors during printing process, especially in the fabrication of 3D structures. 479 To address the angle printing deviations, two methods have been proposed: first one was to avoid the 480 sharp angle printing and the second one was to reduce the extrusion rate to half value at this overlapping 481 area or double the nozzle moving speed (He et al., 2016). 482 Pentagrams with a sharp angle printed at different temperature using ink of kC (1)X(0.25)S(2) are 483 shown in Fig. 7C. The overlap problem was more prominent for the samples fabricated at a high 484 temperature, especially at 50°C. This was because, during printing, the inks experienced a rapid 485 temperature change from a printing temperature in the nozzle cavity (35°C, 40°C, 45°C and 50°C) to the 486 room temperature (25°C), during which the inks were required to rapidly turn into a gel-like structure. As 487 illustrated in Fig. 5, the gelation of inks is a time-dependent process and a high temperature requires a 488 long tgel time. From rheological study, an increased tgel time was obtained from 77s at 35°C to 98s at 489 50°C (Fig. 5F). Generally, if the inks were still in an liquid like form after being extruded onto the 490 platform they would experience continuous spreading as they lack enough mechanical strength to be 491 capable of self-supporting. 492 493 3.6.2 Two-dimensional structure (2D structure) 494 Lattice scaffold is commonly used in bio-printing field and many researchers have used it to evaluate 495 the printing behaviour of bio-inks (He et al., 2016; Li et al., 2016; Ouyang et al., 2016; Paxton et al., 22 ACCEPTED MANUSCRIPT 496 2017). Here, three layers of lattice scaffold (43 mm43 mm, crosshatch infill with 7 mm spacing between 497 fibers, 2.4 mm height, Fig. 1B) was printed. As seen from Fig. 8A and Fig. 8C, for all tested inks, the 498 area of interconnected channels decreased when printing temperature increased from 35°C to 50°C. This 499 was expected as inks being extruded out from the nozzle cavity at a higher temperature require a longer 500 gelation time tgel to achieve sufficient mechanical strength to self-support after being deposited onto 501 platform at room temperature (Fig. 5F). In particular, for the scaffolds printed object using 502 kC(1)X(0.25)S(0), the average area of interconnected channels decreased from 32.29 mm2 to 16.58 mm2 503 when temperature increased from 35°C to 50°C, even some interconnected channels were disappeared at 504 50°C. This was expected as tgel time increased from 86s at 35°C to 110s at 50°C for the ink of 505 kC(1)X(0.25)S(0). During the gelation process, the extruded filaments deformed continuously due to 506 insufficient yield strength to resist gravity and the force given by the subsequent layers. From 507 experiments, it was observed that inks showed three printing behaviour under different printing 508 temperature. If the ink is printed at a high temperature, it will present droplet morphological form at the 509 nozzle tip and the fabricated interconnected channels would show obvious chambers with a rounded 510 corner due to the fusion of adjacent layers, as the case of cylinders fabricated at 45°C and 50°C using ink 511 kC(1)X(0.25)S(0) and kC(1)X(0.5)S(0) (Fig. 8A and Fig. 8B). If the ink is deposited at a proper 512 temperature, the extruded filaments would be smooth and uniform, and an almost equal to square shapes 513 of interconnected channels would be created, like cylinders fabricated at 40°C using ink kC(1)X(0.25)S(2) 514 and kC(1)X(0.5)S(2) (Fig. 8A and Fig. 8B). In contrast, if the ink is fabricated at a low temperature, it 515 would result in irregular filaments forming fractured interconnected channels with a rough surface 516 structure. Sometimes it would even cause the congestion of nozzle tip, which was indicated by the 517 cylinder printed at 35°C using ink kC(1)X(0. 5)S(2) (Fig. 8A and Fig. 8B). 23 ACCEPTED MANUSCRIPT 518 Generally, it is recommended that the nozzle cavity temperature should be slightly higher than 519 Tgelation to ensure both the easy extrusion of inks and the rapid occurrence of gelation to form enough 520 mechanical strength to self-support once being extruded out from nozzle tip resulting from temperature 521 variation (Li, et al., 2016; Ouyang, et al., 2016). For instance, the scaffolds printed at 40°C using 522 kC(1)X(0.5)S(2) presented an smooth and uniform structure with an almost equal to square shapes of 523 interconnected channels. This was because the nozzle cavity temperature of 40°C was slightly higher than 524 the ink’s Tgelation (38.2°C), making it easier to gel rapidly once being extruded out form nozzle (tgel- 525 77s). In contrast, the cylinder printed at 35°C using kC(1)X(0.5)S(2) (Tgelation 38.2°C), presented a 526 rough surface structure and a larger interconnected channels (52.66 mm2) than desired (49.00 mm2) (Fig. 527 8A and Fig. 8C). Theoretically, when the ink of kC(1)X(0.25)S(0) was printed at 35°C, scaffolds with 528 regular square shapes should be obtained as the printing temperature was close to its Tgelation (35.4°C). 529 However, a smaller area of interconnected channels (32.29 mm2) than target (49 mm2) with a round 530 corner was still obtained (Fig. 8C). Several factors below might be accounted for this fact. Firstly, the 531 yield stress of kC(1)X(0.25)S(0) at 35°C (39.2 Pa, Tab. 4) is not enough to resist the deformation and it 532 takes a certain time to achieve enough mechanical strength after extruded onto deposited platform at room 533 temperature (25°C, tgel 86s, Fig. 5F). Secondly, Tgelation of materials obtained from rheological tests 534 closely related to the cooling rate (Gladkowska-Balewicz et al., 2014). The cooling rate of 3°C/min used 535 in the study might be resulting in an overestimation of Tgelation. 536 3.6.3 Three-dimensional structure (3D structure) 537 A hollow cylinder was printed to investigate inks’ ability to fabricate complex 3D structures (Fig. 9A). 538 The cylinders printed using ink kC(1)X(0.25)S(2) and kC(1)X(0.5)S(2) presented good shape retention 539 even under the load of many successive layers (19 layers), which is difficult to achieve using soft 540 hydrogel inks. Comparatively, the cylinder fabricated with ink kC(1)X(0.25)S(0) showed very poor shape 24 ACCEPTED MANUSCRIPT 541 retention and it collapsed dramatically during printing process. Several factors below may be accounted 542 for this fact. Firstly, Tgelation of ink kC(1)X(0.25)S(0) was 35.4°C (Tab. 4), obviously lower than the 543 printing temperature (40°C). As such condition, the ink was too liquid after extrusion and could not 544 rapidly recover enough mechanical strength to be able to self-support. Secondly, the ink 545 kC(1)X(0.25)S(0) with a longer tgel (89s) indicated that it would take longer to achieve gelation after 546 being extruded out, during which the extruded filaments flows continually until gelation occurred. Lastly, 547 the G of ink kC(1)X(0.25)S(0) was significantly lower than other inks with a value of 51.2 Pa at 10 rad/s 548 at 25°C. This poor mechanical strength might also exacerbate the deformation of printed cylinder. To 549 better characterize the deformation of cylinders, the parameter of wall thickness was used to evaluate 550 inks’ ability to fabricate complex 3D structures, as collapsing of cylinders would definitely increase it. As 551 seen from Fig. 9D, the wall thickness of cylinders printed with kC(1)X(0.25)S(0) and kC(1)X(0.5)S(0) 552 were significantly higher than that of cylinders fabricated with ink kC(1)X(0.25)S(2) and 553 kC(1)X(0.5)S(2). Again, the three factors described above might be accounted for this phenomenon, 554 namely Tgelation, gelation time tgel, and mechanical strength (G). Note that the wall thickness of 555 cylinders was larger than target thickness (3.5 mm), regardless of the used inks (Fig. 9D). Also note that 556 generally the width of extruded-out filament is bigger than theoretical value, while the height of filament 557 is lower, as a result of gravity and fusion between ink layers. As illustrated in Fig. 9B and Fig. 9C, the 558 schema of width difference (W) caused by first layer deformation and fusion between adjacent ink 559 layers definitely contributed to the larger wall thickness than desired value. Moreover, the wall thickness 560 deviation might also be related to the slicing software Skeinforge, which is specially designed for plastic 561 materials like polylactide, without any consideration for food materials and hydrogel properties. 562 4 Conclusions 25 ACCEPTED MANUSCRIPT 563 In this work, gel model multi-component systems composed of carrageenan, xanthan gum and starch 564 were prepared for an extrusion-based 3D food printer. The 3D printing process was divided into three 565 stages and corresponding rheological properties of inks for each stage were determined, namely extrusion 566 stage (yield stress, viscosity, shear-thinning), recovery stage (shear recovery, temperature recovery) and 567 self-supporting stage (G). Finally, printability of inks was systematically studied by printing 1D structure, 568 2D structure 3D structure. Addition of starch and xanthan gum increased inks’ Tgelation, viscosity, yield 569 stress, G*, and enhanced inks’ shear-thinning behaviour and reduced inks’ tgel. During extrusion stage, 570 inks’ extrudability was significantly affected by yield stress and shear-thinning behaviour. Tgelation, tgel 571 and G* are important for inks’ printability and shape retention performance. The insights achieved from 572 this study will provide guidance for food printing processes in which hydrocolloids are used as additives 573 to improve printing behaviour of the food formulations. 574 Acknowledgments 575 The authors acknowledge the financial support from the China State Key Laboratory of Food Science 576 and Technology Innovation Project (Contract No. SKLF-ZZA-201706), National First-class Discipline 577 Program of Food Science and Technology (No. JUFSTR20180205), Jiangsu Province (China) 578 “Collaborative Innovation Center for Food Safety and Quality Control” Industry Development 579 Program, Jiangsu Province Key Laboratory Project of Advanced Food Manufacturing Equipment and 580 Technology (Contract No. FMZ201803), which have enabled us to carry out this study. 26 ACCEPTED MANUSCRIPT 582 583 References 584 585 Achayuthakan, P., & Suphantharika, M. (2008). Pasting and rheological properties of waxy corn starch as affected by guar gum and xanthan gum. Carbohydrate Polymers, 71(1), 9-17. 586 587 Barrera, C., Florián-Algarin, V., Acevedo, A., & Rinaldi, C. (2010). Monitoring gelation using magnetic nanoparticles. Soft Matter, 6(15), 3662-3668. 588 589 Chaudemanche, C., & Budtova, T. (2008). Mixtures of pregelatinised maize starch and κ-carrageenan: Compatibility, rheology and gelation. Carbohydrate Polymers, 72(4), 579-589. 590 591 592 Chung, J. H. Y., Naficy, S., Yue, Z., Kapsa, R., Quigley, A., Moulton, S. E., & Wallace, G. G. (2013). Bio-ink properties and printability for extrusion printing living cells. Biomaterials Science, 1(7), 763. 593 594 595 Cohen, D. L., Lipton, J. I., Cutler, M., Coulter, D., Vesco, A., & Lipson, H. (2009). Hydrocolloid printing: A novel platform for customized food production. In 20th Annual International Solid Freeform Fabrication Symposium, SFF 2009 (pp. 807-818). 596 597 598 Fakharian, M.H., Tamimi, N., Abbaspour, H., Mohammadi Nafchi, A., & Karim, A. A. (2015). Effects of κ-carrageenan on rheological properties of dually modified sago starch: Towards finding gelatin alternative for hard capsules. Carbohydrate Polymers, 132, 156-163. 599 600 601 Gladkowska-Balewicz, I., Norton, I. T., & Hamilton, I. E. (2014). Effect of process conditions, and component concentrations on the viscosity of κ-carrageenan and pregelatinised cross-linked waxy maize starch mixed fluid gels. Food Hydrocolloids, 42, 355-361. 602 Graham, S. (2009). Xanthan Gum. In Food Stabilisers, Thickeners and Gelling Agents (pp. 185-193). 603 604 Hamilton, C. A., Alici, G., & in het Panhuis, M. (2018). 3D printing Vegemite and Marmite: Redefining “breadboards”. Journal of Food Engineering, 220, 83-88. 605 606 He, Y., Yang, F., Zhao, H., Gao, Q., Xia, B., & Fu, J. (2016). Research on the printability of hydrogels in 3D bioprinting. Scientific reports, 6, 29977. 607 608 Holzl, K., Lin, S., Tytgat, L., Van Vlierberghe, S., Gu, L., & Ovsianikov, A. (2016). Bioink properties before, during and after 3D bioprinting. Biofabrication, 8(3), 032002. 609 610 Huang, C. Y. (2018). Extrusion-based 3D Printing and Characterization of Edible Materials. University of Waterloo. 611 612 Lewis, J. A. (2006). Direct ink writing of 3D functional materials. Advanced Functional Materials, 16(17), 2193-2204. 613 614 Lewis, J. A., Smay, J. E., Stuecker, J., & Cesarano, J. (2006). Direct ink writing of three‐dimensional ceramic structures. Journal of the American Ceramic Society, 89(12), 3599-3609. 27 615 616 MANUSCRIPT Li, H., Liu, S., & Lin, L. (2016). ACCEPTED Rheological study on 3D printability of alginate hydrogel and effect of graphene oxide. International Journal of Bioprinting, 2(2), 163-175. 617 618 619 Lille, M., Nurmela, A., Nordlund, E., Metsä-Kortelainen, S., & Sozer, N. (2017). Applicability of protein and fiber-rich food materials in extrusion-based 3D printing. Journal of Food Engineering, 220, 20-27. 620 621 Lin, J.H., Liang, C.W., & Chang, Y.H. (2016). Effect of starch source on gel properties of kappacarrageenan-starch dispersions. Food Hydrocolloids, 60, 509-515. 622 623 Liu, S., Chan, W. L., & Li, L. (2015). Rheological properties and scaling laws of κ-carrageenan in aqueous solution. Macromolecules, 48(20), 7649-7657. 624 625 Liu, Z., Bhandari, B., Prakash, S., & Zhang, M. (2018). Creation of internal structure of mashed potato construct by 3D printing and its textural properties. Food Research International, 111, 534-543. 626 627 Liu, Z., Zhang, M., & Bhandari, B. (2018). Effect of gums on the rheological, microstructural and extrusion printing characteristics of mashed potatoes. Int J Biol Macromol, 117, 1179-1187. 628 629 Liu, Z., Zhang, M., Bhandari, B., & Wang, Y. (2017). 3D printing: Printing precision and application in food sector. Trends in Food Science & Technology, 69, 83-94. 630 631 Liu, Z., Zhang, M., Bhandari, B., & Yang, C. (2018). Impact of rheological properties of mashed potatoes on 3D printing. Journal of Food Engineering, 220, 76-82. 632 633 Liu, Z., Zhang, M., & Yang, C.H. (2018). Dual extrusion 3D printing of mashed potatoes/strawberry juice gel. LWT, 96, 589-596. 634 635 Lorenzo, G., Zaritzky, N., & Califano, A. (2015). Mechanical and optical characterization of gelled matrices during storage. Carbohydrate Polymers, 117, 825-835. 636 637 M’Barki, A., Bocquet, L., & Stevenson, A. (2017). Linking Rheology and Printability for Dense and Strong Ceramics by Direct Ink Writing. Scientific reports, 7(1), 6017. 638 639 Mandala, I. G., & Bayas, E. (2004). Xanthan effect on swelling, solubility and viscosity of wheat starch dispersions. Food Hydrocolloids, 18(2), 191-201. 640 641 642 Mantihal, S., Prakash, S., Godoi, F. C., & Bhandari, B. (2017). Optimization of chocolate 3D printing by correlating thermal and flow properties with 3D structure modeling. Innovative Food Science & Emerging Technologies, 44, 21-29. 643 644 645 Morris, E. R., Rees, D. A., Young, G., Walkinshaw, M. D., & Darke, A. (1977). Order-disorder transition for a bacterial polysaccharide in solution. A role for polysaccharide conformation in recognition between Xanthomonas pathogen and its plant host. Journal of Molecular Biology, 110(1), 1-16. 646 647 648 Morrison, N. A., Clark, R., Talashek, T., & Yuan, C. R. (2004). New forms of xanthan gum with enhanced properties. In P. A. Williams & G. O. Phillips (Eds.), Gums and Stabilisers for the Food Industry 12 (pp. 124-130): The Royal Society of Chemistry. 28 649 650 ACCEPTED MANUSCRIPT Ouyang, L., Yao, R., Zhao, Y., & Sun, W. (2016). Effect of bioink properties on printability and cell viability for 3D bioplotting of embryonic stem cells. Biofabrication, 8(3), 035020. 651 652 Parker, A., Brigand, G., Miniou, C., Trespoey, A., & Vallée, P. (1993). Rheology and fracture of mixed ιand κ-carrageenan gels: Two-step gelation. Carbohydrate Polymers, 20(4), 253-262. 653 654 655 Paxton, N., Smolan, W., Böck, T., Melchels, F., Groll, J., & Jungst, T. (2017). Proposal to assess printability of bioinks for extrusion-based bioprinting and evaluation of rheological properties governing bioprintability. Biofabrication, 9(4), 044107. 656 657 658 Pospischil, M., Specht, J., Konig, M., Horteis, M., Mohr, C., Clement, F., & Biro, D. (2014). Paste rheology correlating with dispensed finger geometry. IEEE Journal of Photovoltaics, 4(1), 498503. 659 660 661 Rayna, T., & Striukova, L. (2016). From rapid prototyping to home fabrication: How 3D printing is changing business model innovation. Technological Forecasting and Social Change, 102, 214224. 662 663 Rezende, R. A., Bartolo, P. J., Mendes, A., & Filho, R. M. (2009). Rheological behaviour of alginate solutions for biomanufacturing. Journal of Applied Polymer Science, 113(6), 3866-3871. 664 665 Roknul, A.S. M., Zhang, M., Mujumdear, A.S., & Yang, C.H. Study on 3D printing of orange concentrate and material characteristics. Journal of Food Process Engineering, 0(0), e12689. 666 667 Severini, C., Derossi, A., & Azzollini, D. (2016). Variables affecting the printability of foods: Preliminary tests on cereal-based products. Innovative Food Science and Emerging Technologies, 38, 281-291. 668 669 Shchipunov, Y. A. (2003). Sol–gel-derived biomaterials of silica and carrageenans. Journal of Colloid and Interface Science, 268(1), 68-76. 670 671 Shi, X., & BeMiller, J. N. (2002). Effects of food gums on viscosities of starch suspensions during pasting. Carbohydrate Polymers, 50(1), 7-18. 672 673 Sun, J., Zhou, W., Huang, D., Fuh, J. Y. H., & Hong, G. S. (2015). An Overview of 3D Printing Technologies for Food Fabrication. Food and Bioprocess Technology, 8(8), 1605-1615. 674 675 676 Suntornnond, R., An, J., & Chua, C. K. (2017). Bioprinting of Thermoresponsive Hydrogels for Next Generation Tissue Engineering: A Review. Macromolecular Materials and Engineering, 302(1), 1600266. 677 678 679 Sweeney, M., Campbell, L. L., Hanson, J., Pantoya, M. L., & Christopher, G. F. (2017). Characterizing the feasibility of processing wet granular materials to improve rheology for 3D printing. Journal of Materials Science, 52(22), 13040-13053. 680 681 Tecante, A., & Santiago, M. d. C. N. (2012). Solution properties of κ-carrageenan and its interaction with other polysaccharides in aqueous media. In Rheology: InTech. 682 683 Vancauwenberghe, V., Katalagarianakis, L., Wang, Z., Meerts, M., Hertog, M., Verboven, P., Moldenaers, P., Hendrickx, M. E., Lammertyn, J., & Nicolaï, B. (2017). Pectin based food-ink 29 684 685 ACCEPTED MANUSCRIPT formulations for 3-D printing of customizable porous food simulants. Innovative Food Science & Emerging Technologies, 42, 138-150. 686 687 Wang, L., Zhang, M., Bhandari, B., & Yang, C. (2017). Investigation on fish surimi gel as promising food material for 3D printing. Journal of Food Engineering. 688 689 690 Wilson, S. A., Cross, L. M., Peak, C. W., & Gaharwar, A. K. (2017). Shear-Thinning and ThermoReversible Nanoengineered Inks for 3D Bioprinting. ACS applied materials & interfaces, 9(50), 43449-43458. 691 692 693 Yang, F., Zhang, M., Bhandari, B., & Liu, Y. (2018). Investigation on lemon juice gel as food material for 3D printing and optimization of printing parameters. LWT - Food Science and Technology, 87, 6776. 694 695 Yang, F., Zhang, M., Prakash, S., & Liu, Y. (2018). Physical properties of 3D printed baking dough as affected by different compositions. Innovative Food Science & Emerging Technologies, 1, 1-7. 696 697 Yang, Z., Yang, H., & Yang, H. (2018). Effects of sucrose addition on the rheology and microstructure of κ-carrageenan gel. Food Hydrocolloids, 75, 164-173. 698 699 700 Zhang, M., Vora, A., Han, W., Wojtecki, R. J., Maune, H., Le, A. B., Thompson, L. E., McClelland, G. M., Ribet, F., & Engler, A. C. (2015). Dual-responsive hydrogels for direct-write 3D printing. Macromolecules, 48(18), 6482-6488. 30 ACCEPTED MANUSCRIPT 702 703 704 Fig. 1 A: Research framework diagram. Tgelation, gelation temperature; PT, printing temperature; RT, 705 room temperature. B: Photograph of the printer used in the study. C: Designed 43 mm43 mm lattice 706 scaffold model with a 7 mm7 mm spacing between 3 mm fiber. D: Designed 15 mm height hollow 707 cylinder with a 3.5 mm wall thickness and 25 mm inside diameter. 31 ACCEPTED MANUSCRIPT 709 710 711 Fig. 2 A: Temperature ramp tests with a cooling rate of 3°C/min for 2% starch (S(2)), 0.25% xanthan gum 712 (X(0.25)), 1% k-carrageenan (kC(1)), and mixture system kC(1)X(0.25) and kC(1)X(0.25)S(2); B: 713 Illustrative example of gelation temperature (Tgelation) for the composition kC(1)X(0.25)S(0) with a 714 cooling rate of 3°C/min. Tgelation was determined by extrapolating the high and low temperature 715 asymptotes of the viscosity and specifying the temperature at which these intersect. C: Illustrative 716 example of yield stress for the composition kC(1)X(0.25)S(2) at 45°C. The yield stress of inks has been 717 determined as the cross-over point where elastic modulus (G) equals to viscous modulus (G). 32 ACCEPTED MANUSCRIPT 719 720 721 Fig. 3 Viscosity and shear stress profiles for different inks at temperature of 35°C, 40°C and 45°C over 722 shear rate of 0.01~100 1/s. 33 ACCEPTED MANUSCRIPT 724 725 726 Fig. 4 A: Recovery tests of inks conducted at 40°C under alteration of high (100 1/s) and low (1 1/s) shear 727 rate. B: Shear recoverability of inks determined as the percentage of viscosity obtained during the first 30s 728 in the third step after high shear rate (100 1/s) based on the average viscosity obtained in the first step. 34 ACCEPTED MANUSCRIPT 730 731 732 Fig. 5 A: Illustrative example of gelation time (tgel) of kC(1)X(0.5)S(0) through temperature change from 733 printing temperature (PT, 40°C) to room temperature (RT: 25°C) indicated the response of complex 734 modulus (G*). B, C, D, and E: Gelation kinetics of kC(1)X(0.25)S(0), kC(1)X(0.5)S(0), 735 kC(1)X(0.25)S(2), and kC(1)X(0.5)S(2) through temperature change from PT to RT indicated the 736 response of G*. F: Gelation time (tgel) of different inks at different temperature. 35 ACCEPTED MANUSCRIPT 738 739 740 Fig. 6 Viscoelasticity properties of different inks within frequency of 1~100 rad/s conducted at room 741 temperature 25°C. A: Elastic modulus. B: Viscous modulus. C: Complex modulus. 36 ACCEPTED MANUSCRIPT 743 744 745 Fig. 7 A: Line printing behaviour of ink kC(1)X(0.5)S(2) at printing temperature of 35°C, 40°C, 45°C and 746 50°C. B: Schema of overlap (a) and printed pentagrams with different angles (b, c and d) using ink 747 kC(1)X(0.25)S(2). C: Printed pentagrams using ink kC(1)X(0.25)S(2) at temperature of 35°C, 40°C, 45°C 748 and 50°C. 37 ACCEPTED MANUSCRIPT 750 751 752 Fig. 8 A: Images of printed scaffolds using different inks at different temperature. B: Schema of printed 753 interconnected channel holes at different status. C: Average area of interconnected channels of printed 754 scaffolds. 38 ACCEPTED MANUSCRIPT 756 757 758 Fig. 9 A: Images of hollow cylinders printed at 40°C viewed from different direction. B: Schema of width 759 deformation (W) of first ink layer. C: Schema of width deformation of second ink layer and its fusion 760 with first layer. D: Wall thickness of hollow cylinders printed at 40°C using different inks. 39 ACCEPTED MANUSCRIPT 762 763 Tab. 1: Concentration levels of k-carrageenan, xanthan gum and potato starch applied by 23 factorial design Independent variables Lower level (-1) Upper level (+1) k-Carrageenan (%w/w) 1 2 0.25 0.5 0 2 Xanthan (%w/w) Potato Starch (%w/w) 764 40 ACCEPTED MANUSCRIPT 766 767 768 Tab. 2: Experimental matrix designed for the independent variables concentration of k-carrageenan, 769 xanthan gum and potato starch (%w/w) in water. Coded values k-Carrageenan (%w/w) Xanthan (%w/w) Potato Starch (%w/w) Ink Formulations (-1, -1, +1) (-1, -1, -1) (-1, +1, +1) (-1, +1, -1) (+1, -1, +1) (+1, -1, -1) (+1, +1, +1) (+1, +1, -1) 1 1 1 1 2 2 2 2 0.25 0.25 0.5 0.5 0.25 0.25 0.5 0.5 2 0 2 0 2 0 2 0 kC(1)X(0.25)S(2) kC(1)X(0.25)S(0) kC(1)X(0.5)S(2) kC(1)X(0.5)S(0) kC(2)X(0.25)S(2) kC(2)X(0.25)S(0) kC(2)X(0.5)S(2) kC(2)X(0.5)S(0) 770 * In the codes listed for formulation column; kC, X and S stands for k-carrageenan, xanthan gum and 771 potato starch, respectively. 41 ACCEPTED MANUSCRIPT 773 774 775 776 Tab. 3 Gelation temperature (Tgelation), elastic modulus (G), viscous modulus (G) and complex modulus 777 (G*) of different inks Formulations Tgelation(○C) G (Pa) G(Pa) G* (Pa) kC(1)X(0.25)S(0)* 35.4±0.3a 51.2±4.6a 39.4±1.3a 64.6±1.6a kC (1)X(0.5)S(0) 35.7±0.1ab 377.5±19.6b 91.6±5.6b 388.4±10.2b kC (1)X(0.25)S(2) 36.9±0.0bc 577.1±26.3c 123.2±4.9c 590.1±11.0c kC (1)X(0.5)S(2) 38.2±0.1c 814.8±34.6e 189.8±16.3e 836.6±6.4e kC (2)X(0.25)S(0) 44.2±0.9d 682.4±36.5d 133.3±18.6cd 694.3±13.3d kC (2)X(0.5)S(0) 44.9±1.1d 831.0±16.9e 140.5±11.2d 842.8±14.6e kC (2)X(0.25)S(2) 46.7±0.6e 1808.0±68.9f 260.5±4.9f 1827±21.2f kC (2)X(0.5)S(2) 47.5±0.2e 9017.0±48.6g 1026.0±5.6g 9075±18.3g 778 * In the codes listed for formulation column; kC, X and S stands for k-carrageenan, xanthan gum and 779 potato starch, respectively. Values with different lowercase letters in the same column are significantly 780 different (p<0.05). G, G and G* were recorded at 10 rad/s. 781 782 783 42 ACCEPTED MANUSCRIPT 785 786 787 788 Tab. 4 Yield stress, consistency index (K) and flow index (n) of inks at different temperature Formulations Yield stress (Pa) kC(1)X(0.25)S(0)* 35°C 39.24.3 40°C # 45°C # kC(1)X(0.5)S(0) 35°C 80.27.1 40°C # 45°C # kC(1)X(0.25)S(2) 35°C 134.23.7 40°C 13.70.6 45°C 12.00.5 kC(1)X(0.5)S(2) 35°C 551.324.0 40°C 60.53.1 45°C 36.91.8 K (Pa·sn) n R2 2.160.04 1.700.03 1.270.03 0.580.01 0.630.01 0.710.03 0.990 0.989 0.957 6.190.09 3.140.06 2.680.05 0.500.01 0.620.02 0.610.01 0.996 0.986 0.987 10.030.21 0.470.01 9.780.16 0.450.03 7.620.10 0.480.00 0.997 0.994 0.997 23.880.56 0.360.01 21.720.44 0.360.00 16.320.35 0.360.01 0.996 0.996 0.996 789 * In the codes listed for formulation column; kC, X and S stands for k-carrageenan, xanthan gum and 790 potato starch, respectively. In the column for yield stress, “#” indicates that the inks showed liquid character with 791 elastic modulus (G) smaller than viscous modulus (G) in the stress of 1~1000 Pa. 792 793 794 43 ACCEPTED MANUSCRIPT Linking rheology and printability of a multicomponent gel system of carrageenan-xanthan-starch in extrusion based additive manufacturing Highlights Thixotropic and thermo-reversible gel system was used as a model in 3D printing Ink’s rheology and 3D printability was correlated in the whole printing process 1D, 2D, 3D structures were fabricated and correlated with ink’s rheology.