STUDIES ON THE CLARIFICATION OF JUICE FROM WHOLE

STUDIES ON THECLARIFICATION OF JUICE FROMWHOLE SUGAR CANE CROPCaroline C.D. ThaiSubmitted in fulfilment of the requirements for the degree ofDoctor of PhilosophySchool of Chemistry, Physics and Mechanical EngineeringScience and Engineering FacultyQueensland University of TechnologyOctober 2013

KeywordsSugar cane juiceWhole sugar cane cropTops and leavesTrashGreen caneBurnt caneJuice compositionJuice gationTurbidityColloidsCoagulantFlocculantSurface chemistryZeta potentialSize distributionFloc structureFractal dimensionDLVO theoryi

AbstractIn the Australian sugar industry, fluctuations in raw sugar prices haveresulted in interest to increase the amount of biomass, so that productdiversification to platform chemicals, ethanol and cogeneration can be realised.Harvesting the whole sugar cane crop, WC (i.e., green cane stalk and tops andleaves, viz. trash), will bring in significant amounts of biomass that could be usedto produce high value biocommodities. However, juice expressed from WC isdifficult to process into raw sugar of acceptable quality and in high yield. Thework outlined in this thesis examined the composition and colloidal chemistry ofparticles present in different types of juices, and the impact trash impurities haveon coagulation and flocculation processes. This is to provide insights as to why itis difficult to process juice expressed from WC, and provide options to improvethe clarification of WC juice.There are significant differences in the composition of juice expressedfrom the cane stalk, trash and WC. Data derived from principal componentsanalysis revealed three distinct clusters, each cluster belonging to one type ofjuice. In general, WC juices either prepared using a two-roll laboratory mill orobtained directly in the factory contain higher proportions of polysaccharides( 48000 mg/kg on brix), proteins ( 13500 mg/kg on brix) and starch( 1500 mg/kg on brix) by factors of 2.5, 1.6 and 1.3, respectively, in comparisonto stalk juices. The method developed to separate and identify oligosaccharides inWC and stalk juices using GC/MS revealed differences in the oligosaccharidecontent in these two juices. Furthermore, prime organic acid trans–aconitic acidii

is higher in WC juices than stalk juices and the concentration of inorganic ionsalso follow the same trend, which complements the electrical conductivity of thejuices.It was established in the present study, for the first time, that colloidaljuice particles consist of silica, an aluminosilicate compound and ironoxide/hydroxide. The type of aluminosilicate compound will vary from one areato another, as this depends on the soil from which the sugar cane is grown. Thesurface of these juice particles consists of proteins, polysaccharides and organicacids, and the proportions of these compounds are related to the origin of theseparticles.The ζ-potential of the particles, average size distribution of theaggregates, and floc structure are influenced by the presence of trash impurities.A method was developed that allowed accurate measurements of theelectrophoretic mobility and hence, the ζ-potential of sugar cane juice particles.The ζ-potential of WC juice particles was found to be slightly lower(–2.65 0.06 mV) than that of stalk juice particles (–2.39 0.06 mV), indicatingthat the former particles are more stable and hence less likely to coagulate.The fractal dimension (Df) values of primary floc particles formed in juicewith varying amounts of trash impurities are between 2.18 for flocs derived fromstalk and 2.29 for flocs derived from WC. As the difference in Df are small, thedifferences observed in clarification performance between juices expressed fromstalk and those from WC, is in part, due to the configuration of the macro-flocsformed. The macro-flocs formed with juice particles derived from stalk are in anetwork arrangement allowing suspended impurities to become readily trapped.iii

The theory devised by Derjaguin, Landau, Verwey and Overbeek (DLVO)was used to account for the interactive forces. The interactive forces betweenjuice particles are predominantly van der Waals forces, and the presence of trashimpurities increase the magnitude of the repulsive forces.As such, the netinteraction energy is more negative for WC particles than for stalk particles,resulting in the later particles being more stable in solution. The extent of chargeneutralisation on formed calcium phosphate is reduced, preventing the coagulationprocess from being effective.On-site clarification trials were conducted on juices expressed from sugarcane originating from a number of sugar factories in Queensland and New SouthWales in Australia to investigate the best liming procedure for effectiveclarification of juice expressed from WC. Variants of the liming process, wherelime saccharate was added to cold ( 52 C), intermediate ( 76 C) and hot( 100 C) juice, were investigated to determine the best liming technique. Bothsingle and dual clarification techniques were also investigated.The dualclarification process involved pre-treatment of the juice by applying heat(60–70 C) or by applying heat and lowering the juice pH (3.5–4.5) to reach theisoelectric point of proteins and adding a highly charged cationic coagulant.Results on the clarification of juice derived from burnt (BE) and greencane (GE) have shown that sugar factories can successfully process juiceexpressed from stalk containing a reasonable proportion of trash (i.e., nominallyhalf) without affecting clarification performance. Effective particle removal wasachieved with intermediate liming (i.e., liming juice at 76 C prior to clarification).The addition of phosphoric acid ( 100 mg/kg as P2O5) to GE juice was found toiv

produce clarified juice turbidity of acceptable quality, though it resulted in higherproportions of P, Mg and Si in clarified juice.A number of coagulants (including polydiamines and aluminium species)were investigated for the early stages of the clarification process in order toidentify which coagulant can be used to reduce clarified juice turbidity in WCjuice. None of these additives showed promise. However, one of the additives, apolydiallydimethyl ammonium chloride (polyDADMAC) gave relatively largeflocs in juice expressed from WC was selected for use in the dual clarificationprocess.The dual clarification process significantly improved impurity removal injuices derived from WC juice. However, the formed floc aggregates were fragile,loosely bound, and floated rather than settled.The process where thepre-treatment step involves only heating is the best option as there were noincreases in the level of Ca and S relative to the values obtained to normal juice.Sucrose degradation was seen to be minimal. Hot liming (i.e., addition of limesaccharate to juice heated to 100 C) in the 2nd stage of the dual clarificationprocess reduced the amount of impurities and gave lower clarified juice turbiditiesthan cold, cold/intermediate and intermediate liming techniques.Proteins,polysaccharides, P and Si removal were enhanced using the hot liming technique.To adopt the dual clarification method, ways to settle the flocs bysedimentation need to be found.Alternatively, flotation methods such asdissolved air flotation (DAF) can be used to remove the floating flocs. However,it is important to note that at extreme temperatures ( 100 C) of juice, the DAFv

system could interfere with the supply of dissolved air required for the removal offlocs by flotation.A key contribution from this thesis is an enhanced understanding of thecoagulation and flocculation processes associated with colloidal particles presentin sugar cane juice. The work contained in this thesis has provided evidence toshow why it is difficult to clarify juice expressed from WC. Strategies to improvethe processing of juice expressed from WC have been proposed which will reducethe cost of processing juice expressed from WC.vi

PublicationsJournal articlesThai, C.C.D., Bakir, H. and Doherty, W.O.S. (2012). Insights to the clarificationof sugar cane juice expressed from sugar cane stalk and trash. Journal ofAgricultural and Food Chemistry, 60(11), 2916-2923.Thai, C.C.D. and Doherty, W.O.S. (2012). Characterisation of sugar cane juiceparticles that influence the clarification process. International Sugar Journal,114(1366), 719–724.Thai, C.C.D. and Doherty, W.O.S. (2013). Effect of trash on the coagulation andflocculation of sugarcane juice. International Sugar Journal, 115.Peer-reviewed conference papersThai, C.C.D. and Doherty, W.O.S. (2013). Effect of trash on the coagulation andflocculation of sugarcane juice. In R.C. Bruce (Ed.), Proceedings of theAustralian Society of Sugar Cane Technologists, 35. Townsville, Qld, Australia.Thai, C.C.D. and Doherty, W.O.S. (2012). Characterisation of sugarcane juiceparticles that influence the clarification process. In R.C. Bruce (Ed.), Proceedingsof the Australian Society of Sugar Cane Technologists, 34. Palm Cove, Qld,Australia.Thai, C.C.D. and Doherty, W.O.S. (2011). The composition of sugarcane juicesderived from burnt cane and whole green cane crop. In R.C. Bruce (Ed.),vii

Proceedings of the Australian Society of Sugar Cane Technologists, 33. Mackay,Qld, Australia.Conference Poster PresentationThai, C.C.D. and Doherty, W.O.S. (2012). Effect of impurities on the coagulationand flocculation of sugarcane juice particles. Advances in Particle SeparationConference, 18–20 June 2012, Berlin, Germany.AwardThe Denis Foster Chemistry/Chemical Engineering Award for the best paperpresentation (tertiary level) at the 33rd Australian Society of Sugar CaneTechnologists Conference, May 2013.viii

Table of ContentsKeywords . iAbstract. iiPublications . viiConference Poster Presentation . viiiAward . viiiTable of Contents . ixList of Figures. xivList of Tables . xviiList of Abbreviations and Nomenclature. xxStatement of Original Authorship . xxiiiAcknowledgements . xxiv1 Introduction . 11.1Overview . 11.2Research Problem. 31.3Objectives . 41.4Scope. 52 Literature Review . 72.1Sugar Cane Plant . 72.1.1Sugar cane components . 72.1.2Sugar cane quality . 92.2The Production of Raw Sugar . 122.2.1Harvesting and transport . 122.2.2Milling . 152.2.3Clarification . 172.2.4Evaporation . 192.2.5Crystallisation. 192.2.6Centrifugation and drying . 192.3The Chemistry of Clarification . 202.3.1Particles in sugar cane juice . 202.3.2Chemical reactions . 21ix

2.42.3.2.1Calcium phosphate chemistry . 212.3.2.2Degradation of reducing sugars and organic acids . 242.3.2.3Hydrolysis of starch . 262.3.2.4Formation of colour . 28Colloid Interactions and Stability . 302.4.1Electrical double layer. 302.4.2DLVO theory . 322.4.3Stability of colloids . 362.5Coagulation and Flocculation . 372.5.1Coagulants and flocculants . 382.5.1.1Inorganic and organic coagulants . 392.5.1.2Polyaluminium coagulants . 402.5.1.3Poly(acrylamide-co-sodium acrylate) . 412.5.2Coagulation and charge neutralisation . 422.5.3Flocculation and bridging mechanism . 432.6Impact of Whole Crop Processing . 452.6.1Clarification of whole green cane crop . 463 Materials and Methods . 513.1Introduction . 513.2Materials . 533.2.1Components of the sugar cane plant. 533.2.2Factory sugar cane juices for compositional analyses . 543.2.3Factory sugar cane juices for clarification studies . 553.2.4Prepared WC juice for clarification studies . 573.2.5Coagulants and flocculants . 583.3Clarification Procedure . 613.3.1Batch settling kit . 633.3.2Simple defecation . 633.3.3Dual clarification . 643.3.4Initial settling rate . 653.3.5Final mud level . 653.3.6Juice turbidity . 663.4Jar Test Experiments . 663.5Characterisation Methods. 67x