LWT - Food Science and Technology 62 (2015) 319e324
Contents lists avai
LWT - Food Science and Technology
journal homepage: www.elsevier .com/locate/ lwt
Effect of heat-moisture treatment with maltitol on physicochemicalproperties of wheat starch
Qingjie Sun*, Chong Nan, Lei Dai, Liu XiongCollege of Food Science and Engineering, Qingdao Agricultural University, Qingdao, Shandong Province, 266109, China
a r t i c l e i n f o
Article history:Received 29 August 2013Received in revised form27 November 2014Accepted 16 January 2015Available online 24 January 2015
Effect of heat-moisture treatment (HMT) with maltitol on physicochemical properties of wheat starchwas investigated. Compared with the mixture of maltitol and wheat starch (MAWS), peak viscosity,trough viscosity, final viscosity, breakdown and setback of MAWS modified by heat-moisture treatment(HMT-MAWS) was decreased by 119.29, 63.37, 84.50, 55.92, 21.12 RVU, respectively. The viscosities ofHMT-MAWS were affected more remarkably than that of the mixture of maltitol and heat-moisturetreatment modified wheat starch (MA-HMTWS). Gelatinization temperature (To, Tp, Tc) of HMT-MAWS increased significantly than that of MAWS. Scanning electron microscope showed that a layerof membrane-like substance adhered to the smooth surface of HMT-MAWS, but that of MAWS becamerough with a lot of small particles. After gelatinization and freeze drying, the gel structure of HMT-MAWSwas tighter than that of MAWS. According to X-ray diffraction pattern, the area of amorphous region ofHMT-MAWS was higher than that of MAWS.
Starch is the main polysaccharide in plant-origin food products.It is commonly used as thickener, colloidal stabilizer, gelling,bulking and water preserving agent. The granule swelling, gelati-nization, pasting and the subsequent gel properties of starch playimportant roles in the foodmanufacturing industry.With the use ofstarch in food industry becoming increasingly broader and moreextensive, native starch does not always possess the physico-chemical properties appropriate for certain types of processing.Starch is often modified by physical, chemical, and enzymaticprocesses to promote specific functional properties. Heat-moisturetreatment (HMT) is one of the more important physical methods byusing simple and environmentally safe processes, with lowcost andwithout by-products of chemical reagents (Adebowale, Olu-Owolabi, Olayinka, & Lawal, 2005; Sun, Wang, Xiong, & Zhao,2013; Zavareze & Dias, 2011). HMT involves treatment of starchgranules at lowmoisture levels (<35 g/100 g) within a certain timeperiod (15 mine16 h) and at a specific range of temperature(84e120 �C) above the glass transition temperature but below the
ce and Engineering, Qingdaod, Chengyang District, Qing-030449.
Additives such as sugar are commonly used in starch-basedfoods in order to optimize the processing operation and enhancefood quality by influencing the gelatinization and retrogradation ofstarch. However, food high in sugar and calories can lead to diseasessuch as obesity and diabetes. In recent times, people pay moreattention to healthy eating habits and low-calorie diets. The pos-sibility of using sugar alcohol instead of sucrose as a sweetener hasattracted more and more researchers' attention (Kommineni,Amamcharla, & Metzger, 2012; Lin, Lee, Mau, Lin, & Chiou, 2010).Maltitol, a sugar alcohol used to replace table sugar as a sugarsubstitute, has fewer calories, does not promote tooth decay andhas a lesser effect on blood glucose. The chemical property ofmaltitol is known as 4-O-a-glucopyranosyl-D-sorbitol.
Some researchers have already studied the effects of sugaralcohol on the physicochemical properties of starch. Lourdin, Bizot,Colonna, and Coignard (1997) indicated that sorbitol was one of thenatural plasticizers for starch vitreous behavior whereas other re-searches held that sorbitol, depending on its concentration inthe starch system, behaved as an antiplasticizer (Gaudin,Lourdin, Forssell, & Colonna, 2000; Mantzari, Raphaelides, &Exarhopoulos, 2010). The freezing behavior of corn starch gelswith xylitol, mannitol, and sorbitol exhibited higher freezing levelsthan that of corresponding sugars (Baek, Yoo, & Lim, 2004; Kim,
Fig. 1. Pasting properties of WS in the presence of maltitol with heat-moisturetreatment. WS: Wheat starch; HMT-WS: Wheat starch modified by heat-moisture;MAWS: Mixture of maltitol and wheat starch; HMT-MAWS: Mixture of maltitol andwheat starch modified by heat-moisture; MA-HMTWS: Mixture of maltitol and heat-moisture treatment modified wheat starch.
Q. Sun et al. / LWT - Food Science and Technology 62 (2015) 319e324320
Yoo, Cornillon, & Lim, 2004). However, few studies have reportedthe effect of maltitol on the physicochemical properties of wheatstarch (WS), especially with HMT.
The objective of this research is to study the effect of HMT withmaltitol on pasting properties, thermal properties, morphologicaland structural properties of WS. This study will give new ways toapplication of sugar-free starchy products.
2. Materials and methods
2.1. Materials
Wheat starch (WS) was supplied by Tianjin Tingfung StarchDevelopment Co., Ltd (Tianjin, China). Maltitol was purchased fromFutaste Co., Ltd (Shandong, China). All other reagents used were ofanalytical grades.
2.2. Methods
2.2.1. Samples preparationMaltitol was mixed with WS (MAWS) at a solid weight ratio of
0.5:1 based on starch (dry weight), and then modified by HMT(HMT-MAWS) according to Hoover & Vasanthan, 1994. The contentof moisture of samples was adjusted to 20 g/100 g by adding theappropriate amounts of distilled water into the containers. Thecontainers were hermetically sealed and equilibrated at ambienttemperature for 24 h. Resulting samples were then placed in anelectric oven at 100 �C for 12 h, dried at 40 �C, ground and sieved(0.15 mm). WS modified by HMT (HMT-WS) was mixed with mal-titol at a solid weight ratio of 0.5:1 based on starch (dryweight), thefinal sample maltitol-HMTWS (MA-HMTWS) was then obtained.
2.2.2. Pasting propertiesPasting properties of samples were determined by using RVA-4
(Newport Scientific Pvt. Ltd., Warriewood, Australia) according tothemethods (Singh, Isono, Srichuwong, Noda,&Nishinari, 2008). Asuspension of 3 g (12 g/100 g moisture basis) starch in 25 g ofaccurately weighed distilled water underwent a controlled heatingand cooling cycle under constant shear. The slurry was thenmanually homogenized by using a plastic paddle to avoid lumpformation before the RVA run. The rotating speed was maintainedat 160 rpm along the process. Parameters including peak viscosity(PV), viscosity at the end of hold time at 95 �C or trough viscosity(TV), final viscosity (FV) at the end of cooling, breakdown(BD ¼ PV � TV), setback (SB ¼ FV � TV) and pasting temperaturewere recorded.
2.2.3. Thermal propertiesThermal properties of samples were assessed by a Pyris-1 dif-
ferential scanning calorimeter (DSC) (PerkineElmer Co., Norwalk,CT, USA). The instrument was calibrated with indium and an emptypan was used as reference. Samples were hermetically sealed andequilibrated at ambient temperature for 24 h. Sample pan washeated from 25 �C to 120 �C at 10 �C/min. The onset (To), peak (Tp)and conclusion (Tc) gelatinization temperature as well as enthalpy(DH) were computed.
2.2.4. Scanning electron microscopy (SEM)The surface topography of samples was observed with scanning
electron microscopy (SEM) by using the method of Kim et al.(2008). WS, HTM-WS, MAWS and HTM-MAWS were cooked inboiling water for 20 min to prepare the gel samples, and then putinto an ultralow temperature freezer and freeze dried for 48 h.Freeze-drying was carried out in the vacuum chamber where thechamber pressure was set to 0.13 Pa at room temperature (21 �C).
The freeze-dried samples and uncooked granule samples werefinely milled and then placed under vacuum on a double-sidedScotch tape, mounted on an aluminum specimen holder andcoated with a thin film of gold. Samples were observed under a Jeolscanning electron microscope (JSM 840, Jeol, Tokyo, Japan).
2.2.5. X-ray diffractionX-ray diffractograms of samples were obtained with an X-ray
diffractometer (XRD-6000, Shimadzu, Tokyo, Japan). The scanningregion of the diffraction ranged from 5 to 60� with a target voltageof 45 KV, current of 40 mA and scan speed of 1�/min.
2.2.6. Statistical analysisAll experiments were conducted at least three times, for which
mean values and standard deviations were determined. Addition-ally, experimental datawere analyzed by using Analysis of Variance(ANOVA), and expressed as mean values ± standard deviations.Differences were considered at significant level of 95% (p < 0.05).Pearson's correlation coefficients among other parameters werecalculated by using SPSS 17.0 software.
3. Results and discussion
3.1. Pasting properties
RVA profiles of WS, HMT-WS, MAWS, HMT-MAWS and MA-HMTWS are presented in Fig. 1 and the corresponding pastingparameters are summarized in Table 1. The pasting temperatureand setback of HMT-WS increased comparedwith that of the nativeWS. However, its peak viscosity, trough viscosity and breakdowndecreased. In previous works, HMT-modified starches displayed anincreased pasting temperature and a decreased pasting viscosity,regardless of origin (Hoover & Vasanthan, 1994; Stute, 1992;Varatharajan, Hoover, Liu, & Seetharaman, 2010;Watcharatewinkul, Puttanlek, Rungsardthong, & Uttapap, 2009),these findings were consistent with our results. The pasting tem-perature and breakdown of MAWS slightly decreased, but the peakviscosity, through viscosity, final viscosity and setback increasedsignificantly. The addition of maltitol formedmore hydrogen bondswith starch and inhibited the movement of starch chains whichstrengthened system dynamics, so the viscosity of starch pasteincreased. Previous researches showed that when the concentra-tion of sorbitol was low (<60 g/100 g), the addition of sorbitolreinforced association among the starch chain, polyol and water
Table 1Pasting parameters of WS in the presence of maltitol with heat-moisture treatment.
WS: Wheat starch; HMT-WS: Wheat starch modified by heat-moisture; MAWS: Mixture of maltitol and wheat starch; HMT-MAWS: Mixture of maltitol and wheat starchmodified by heat-moisture; MA-HMTWS: Mixture of maltitol and heat-moisture treatment modified wheat starch.All data represent the mean of three determinations.Means with the same letter in each column are not significantly different (p < 0.05).
Q. Sun et al. / LWT - Food Science and Technology 62 (2015) 319e324 321
thus decreasedwater mobility in the dispersion and so the viscosityof starch paste increased (Gaudin et al., 2000; Mantzari et al., 2010).
Compared to MAWS, the pasting temperature of HMT-MAWSincreased from 72.8 �C to 82.1 �C, but the peak viscosity, throughviscosity, final viscosity, breakdown and setback decreased byabout 120 RVU, 63 RVU, 84 RVU, 56 RVU and 22 RVU respectively.The values of pasting viscosity, breakdown and setback were ar-ranged in the order of MAWS > MA-HMTWS > HMT-MAWS. Theresults indicated that the interplay between starch chain and polyolwas reinforced during heat-moisture treatment. During pasting,the swelling of WS was inhibited significantly by the maltitol withheat-moisture treatment, so the pasting viscosity decreased andthe pasting temperature increased significantly. HMT strengthenedthe stability of MAWS paste during continued shearing and heating.
3.2. Thermal properties
Fig. 2 shows DSC thermograms of WS, HTM-WS, MAWS andHTM-MAWS and Table 2 illustrates the onset, peak and conclusiontemperatures of gelatinization (To, Tp, Tc) and the gelatinizationenthalpies (DH). All samples exhibited only a single endothermicpeak. The endothermal transition of HTM-WS and HTM-MAWSshifted to the higher temperature.
HMT increased the To, Tp, Tc, Tc-To but decreased the DH of WS.Increases in gelatinization temperatures on heat-moisture treat-ment have also been observed by other researchers (Adebowale &Lawal, 2003; Hoover & Vasanthan, 1994). The decrease in DH onheat-moisture treatment ofWS attributed to the disruption of someof the original double helix.
The addition of maltitol increased the To, Tp, Tc but decreasedDH ofWS. Similar results were found in the mixture of starches andsugars (Maaurf, Che Man, Asbi, Junainah, & Kennedy, 2001;Thirathumthavorn & Trisuth, 2008). These were due to the
Fig. 2. DSC curves of WS (a), HMT-WS (b), MAWS (c) and HMT-MAWS (d). WS: Wheatstarch; HMT-WS: Wheat starch modified by heat-moisture; MAWS: Mixture of maltitoland wheat starch; HMT-MAWS: Mixture of maltitol and wheat starch modified byheat-moisture.
formation of hydrogen bonds between maltitol and starch chains.The amorphous and crystallites region were rearranged and starchswelling was inhibited by maltitol. The gelatinization temperaturewas increased and thermal enthalpy was decreased. Compared toMAWS, To, Tp, Tc, Tc-To and DH of HMT-MAWS increased signifi-cantly. Some maltitol formed hydrogen bonds with starch inamorphous and/or crystallites region, and restricted the mobility ofchains, but other free maltitol inhibited the mobility of starchchains and reduced the availability of water in the system duringHMT, so that higher temperature and thermal enthalpy wereneeded to gelatinize starch.
3.3. Analysis of scanning electron microscopy (SEM)
Scanning electron micrographs of WS, HTM-WS, MAWS andHTM-MAWS before gelatinization are presented in Fig. 3. It showedthat the WS granules had two components: bigger wheat A-typeand smaller wheat B-type granule. NativeWS granules had oval andspherical shapes with smooth surfaces. The surface of modifiedstarch granules by HMT appeared hollow and irregularly shaped.The concavities and indentation of the modified starch granuleswere evidence of the weak tissue affected by heat-moisture treat-ment. Similar results were found by other researchers(Jiranuntakul, Puttanlek, Rungsardthong, Punchaarnon, & Uttapap,2011; Lin et al., 2010; Vanier et al., 2012; Zavareze, Stork, Castro,Schirmer, & Dias, 2010). The surface of HTM-MAWS was smoothand irregular on which a layer of membrane-like substance wasadhered to, but the surface of MAWS was rough with a significantquantities of maltitol particles adhered to it. These findings indi-cated that maltitol interacted with the starch granules which pre-vented particles forming as a result of HMT.
After being cooked in boiling water for 20 min and then freeze-dried, the network of WS, HTM-WS, MAWS and HTM-MAWS wereobserved by SEM (presented in Fig. 4). The figure showed that aspongy structure, and at the same time, a fibrillar network of WSgels were formed after the freeze-dry process due to ice crystalformation. The microstructure of WS gel was characterized byfragmented structure due to numerous pores. The matrix sur-rounding the pores was very thin. The pore size of HTM-WSnetwork was much bigger than that in WS network, and the ma-trix surrounding the pores was thicker. The heat-moisture treat-ment decreased the ability of water combining, so there was morewater forming larger ice crystals, leading to bigger holes in networkof HTM-WS.
Compared to the MAWS gel, the pores of HTM-MAWS gel weresmaller and the matrix surrounding the pores was thicker. The gelnet structure of HTM-MAWS was tighter than MAWS. The phe-nomenon revealed that heat-moisture treatment induced the cor-relation between maltitol and starch. Similar results regardingeffects of sugars on tapioca starch were reported by Zhang, Tong,and Ren (2012). They found that the pore sizes of gels containing
Table 2Thermal properties of WS, HMT-WS, MAWS and HMT-MAWS.
WS: Wheat starch; HMT-WS: Wheat starch modified by heat-moisture; MAWS: Mixture of maltitol and wheat starch; HMT-MAWS: Mixture of maltitol and wheat starchmodified by heat-moisture.All data represent the mean of three determinations.Means with the same letter in each column are not significantly different (p < 0.05).
Fig. 3. Scanning electron micrographs of WS, HMT-WS, MAWS and HMT-MAWS. WS: Wheat starch; HMT-WS: Wheat starch modified by heat-moisture; MAWS: Mixture of maltitoland wheat starch; HMT-MAWS: Mixture of maltitol and wheat starch modified by heat-moisture.
Q. Sun et al. / LWT - Food Science and Technology 62 (2015) 319e324322
trehalose were the smallest and the matrices surrounding themwere the thickest. The starchestarch and/or starchemaltitol in-teractions were facilitated to form thick matrices due to the highsolid concentration in the amorphous regions and water moleculescoagulated into ice crystals and formed a separate phase. Uponfreeze drying, ice sublimated and left the starch gels in a sponge-like state. The sizes of pores and the thickness of the matriceswere correlated with the intermolecular hydrogen bonds betweenmaltitol and starch molecules.
3.4. Analysis of X-ray diffraction (XRD)
X-ray diffractograms of WS, HTM-WS, MAWS and HTM-MAWSwere presented in Fig. 5. WS showed the typical A-pattern withstrong peaks at 15� and 23� and a doublet at 17� and 18� 2q. Heat-moisture treatment did not change the crystalline pattern. Therelative crystallinity of HTM-WS (29.2%) was higher than the nativestarch (27.5%). The reasonwas that HMTcaused associations of newcrystallization with small existing crystalline regions of the starchgranule. The increased X-ray intensity is in agreement with
previous study of HMT starch from wheat, oat (Hoover &Vasanthan, 1994) and mung bean starch (Li, Ward, & Gao, 2011).Compared to the native WS and HTM-WS, addition of maltitolincreased the numbers of diffraction peaks. The intensity ofdiffraction peaks at 14.6�, 15.3�, 18.7�, 20.6� and 21.8� 2q of HTM-MAWS were stronger than that of MAWS. The figure also showedthat the area of the amorphous region of HMT-MAWS wasincreased than that of MAWS. The phenomenon could be explainedby maltitol particles melted during HMT and then interacted withstarch that could not recrystallize after cooling.
4. Conclusion
Physicochemical properties, such as pasting properties, thermalproperties, and structural properties of WS were significantlyaffected by HMT with maltitol. HMT strengthened the interactionbetween maltitol and starch. The properties of HMT-MAWS wereaffected more remarkably than that of MAWS and MA-HMTWS,indicating it was more useful to heat-moisture treat the mixtureof maltitol and WS. The results suggested that the interplay
Fig. 4. Scanning electron micrographs of WS, HMT-WS, MAWS and HMT-MAWS gels after freeze drying. WS: Wheat starch; HMT-WS: Wheat starch modified by heat-moisture;MAWS: Mixture of maltitol and wheat starch; HMT-MAWS: Mixture of maltitol and wheat starch modified by heat-moisture.
Fig. 5. X-ray diffractograms of WS, HMT-WS, MAWS and HMT-MAWS. WS: Wheatstarch; HMT-WS: Wheat starch modified by heat-moisture; MAWS: Mixture of maltitoland wheat starch; HMT-MAWS: Mixture of maltitol and wheat starch modified byheat-moisture.
Q. Sun et al. / LWT - Food Science and Technology 62 (2015) 319e324 323
between starch chain and polyol was reinforced during HMT. TheSEM photographs indicated that HMT with maltitol strengthenedWS gel structure. Maltitol instead of sugar used in food industry notonly can improve the taste of food, but also promote its structure,especially with heat-moisture treatment.
References
Adebowale, K. O., & Lawal, O. S. (2003). Microstructure, physicochemical propertiesand retrogradation behavior of Mucuna bean (Mucuna pruriens) starch on heatmoisture treatments. Food Hydrocolloids, 17, 265e272.
Adebowale, K. O., Olu-Owolabi, B. I., Olayinka, O. O., & Lawal, O. S. (2005). Effect ofheat moisture treatment and annealing on physicochemical properties of redsorghum starch. African Journal of Biotechnology, 4, 928e933.
Baek, M. H., Yoo, B., & Lim, S. T. (2004). Effects of sugars and sugar alcohols on thethermal transition and cold stability of corn starch gel. Food Hydrocolloids, 18,133e142.
Gaudin, S., Lourdin, D., Forssell, P. M., & Colonna, P. (2000). Antiplasticisation andoxygen permeability of starchesorbitol films. Carbohydrate Polymers, 43, 33e37.
Gunaratne, A., & Hoover, R. (2002). Effect of heat-moisture on the structure andphysicochemical properties of tuber and root starches. Carbohydrate Polymers,49, 425e437.
Hoover, R., & Vasanthan, T. (1994). Effect of heat-moisture treatment on thestructure and physicochemical properties of cereal, legume, and tuber starches.Carbohydrate Research, 252, 33e53.
Jiranuntakul, W., Puttanlek, G., Rungsardthong, V., Punchaarnon, S., & Uttapap, D.(2011). Microstructural and physicochemical properties of heat-moisturetreated waxy and normal starches. Journal of Food Engineering, 104, 246e258.
Kim, M. J., Choi, S. J., Shin, S. I., Sohn, M. R., Lee, C. J., & Kim, Y. (2008). Resistantglutarate starch from adlay: preparation and properties. Carbohydrate Polymers,74, 787e796.
Kim, Y. R., Yoo, B. S., Cornillon, C., & Lim, S. T. (2004). Effect of sugars and sugaralcohols on freezing behavior of corn starch gel as monitored by time domain1H NMR spectroscopy. Carbohydrate Polymers, 55, 27e36.
Kommineni, A., Amamcharla, J., & Metzger, L. E. (2012). Effect of xylitol on thefunctional properties of low-fat process cheese. American Dairy Science Associ-ation, 95, 6252e6259.
Lin, S. D., Lee, C. C., Mau, J. L., Lin, L. Y., & Chiou, X. Y. (2010). Effect of erythritol onquality characteristics of reduced calorie danish cookies. Journal of Food Quality,33, 14e26.
Li, S., Ward, R., & Gao, Q. (2011). Effect of heat-moisture treatment on the formationand physicochemical properties of resistant starch from mung bean (Phaseolusradiatus) starch. Food Hydrocolloids, 25, 1702e1709.
Lourdin, D., Bizot, H., Colonna, P., & Coignard, L. (1997). Influence of equilibriumrelative humidity and plasticizer concentration on the water content and glasstransition of starch materials. Polymer, 38, 5401e5406.
Maaurf, A. G., Che Man, Y. B., Asbi, B. A., Junainah, A. H., & Kennedy, J. F. (2001).Gelatinization of sago starch in the presence of sucrose and sodium chloride as
Q. Sun et al. / LWT - Food Science and Technology 62 (2015) 319e324324
assessed by differential scanning calorimetry. Carbohydrate Polymers, 45,335e345.
Mantzari, G., Raphaelides, S. N., & Exarhopoulos, S. (2010). Effect of sorbitol additionon the physicochemical characteristics of starchefatty acid systems. Carbohy-drate Polymers, 79, 154e163.
Singh, N., Isono, N., Srichuwong, S., Noda, T., & Nishinari, K. (2008). Structural,thermal and viscoelastic properties of potato starches. Food Hydrocolloids, 22,979e988.
Stute, R. (1992). Hydrothermal modification of starches: the difference betweenannealing and heat/moisture treatment. Starch, 44, 205e214.
Sun, Q. J., Wang, T., Xiong, L., & Zhao, Y. X. (2013). The effect of heat moisturetreatment on physicochemical properties of early indica rice. Food Chemistry,142, 853e857.
Takaya, T., Sano, C., & Nishinari, K. (2000). Thermal studies on the gelatinization andretrogradation of heatemoisture treated starch. Carbohydrate Polymers, 41,97e100.
Thirathumthavorn, D., & Trisuth, T. (2008). Gelatinization and retrogradationproperties of native and hydroxypropylated crosslinked tapioca starches withadded sucrose and sodium chloride. International Journal of Food Properties, 11,858e864.
Vanier, N. L., Zavareze, E. R., Pinto, V. Z., Klein, B., Botelho, F. T., Dias, A. R. G., et al.(2012). Physicochemical, crystallinity, pasting and morphological properties ofbean starch oxidised by different concentrations of sodium hypochlorite. FoodChemistry, 131, 1255e1262.
Varatharajan, V., Hoover, R., Liu, Q., & Seetharaman, K. (2010). The impact of heat-moisture treatment on the molecular structure and physicochemical propertiesof normal and waxy potato starches. Carbohydrate Polymers, 81, 466e475.
Watcharatewinkul, Y., Puttanlek, C., Rungsardthong, V., & Uttapap, D. (2009).Pasting properties of a heat-moisture treated canna starch in relation to itsstructural characteristics. Carbohydrate Polymers, 75, 505e511.
Zavareze, E. R., & Dias, A. R. G. (2011). Impact of heat-moisture treatment andannealing in starches: a review. Carbohydrate Polymers, 83, 317e328.
Zavareze, E. R., Stork, C. R., Castro, L. A. S., Schirmer, M. A., & Dias, A. R. G. (2010).Effect of heat-moisture treatment on rice starch of varying amylose content.Food Chemistry, 121, 358e365.
Zhang, X., Tong, Q. Y., & Ren, F. (2012). Manuscript title: influence of glucose, sucroseand trehalose on the freeze-thaw stability of tapioca starch gels. Journal of FoodScience and Technology, 4, 225e230.
