COMPARATIVE CHARACTERIZATION OF DIETARY FIBRE ENRICHED FROZEN/THAWED MASHED POTATOES

Transcription

1 COMPARATIVE CHARACTERIZATION OF DIETARY FIBRE ENRICHED FROZEN/THAWED MASHED POTATOES Mrí Dolores Alvrez, Cristin Fernández, Mrí Dolores Olivres nd Weneslo Cnet Deprtment of Plnt Foods Siene nd Tehnology, Instituto del Frío/ICTAN-CSIC, José Antonio Novis 10, E Mdrid, Spin Running title Fire enrihed F/TM pottoes Address orrespondene to Mrí Dolores Álvrez, Deprtment of Plnt Foods Siene nd Tehnology, Instituto del Frío/ICTAN-CSIC, José Antonio Novis nº 10, E Mdrid, Spin. E-mil:myoyes@if.si.es 1

2 ABSTRACT The potentil use of ommeril fires (pe fire (PF), inulin (I) nd their lends (PFI)), s fireenrihing gents in frozen/thwed mshed pottoes (F/TM pottoes), ws reported. PF nd I supplementtions onferred hrdness nd softness to the produt respetively. Differenes were ttriuted to the reltionship of the fire with the potto strh mtrix. The ssoition of PF t low onentrtion (< 15 g/kg mshed pottoes) nd I t high onentrtion (> 45 g/kg) is strongly enourged to fortify the diet without promoting negtive effets on texturl nd rheologil properties of F/TM pottoes or olour nd overll eptility of the resulting produts. Keywords: Dietry fire, Mshed pottoes qulity, Osilltory test, Rheology, Mirostruture 52 2

3 53 INTRODUCTION The helth enefits of fire onsumption re well reognized. A diet with low dietry fire is ssoited with spetrum of degenertive diseses, inluding dietes, oesity, oronry hert disese, owel ner nd gllstones; it is well-estlished ft tht the onsumption of dequte mounts of dietry fire signifintly redues the risk of these diseses. [1] Consumer wreness of these hrteristis is inresing; this influenes onsumer purhsing deisions, while the funtionl foods mrket is onstntly growing. A reent joint FAO/WHO report reommended reduing the intke of sugrs, ft nd lohol, nd inresing onsumption of fruits, vegetles nd erel produts, with the im of inresing the totl dietry fire intke to t lest 25 g dy -1. [2,3] The hllenge is to develop trditionl mshed pottoes using dietry fires to inrese the dily intke of fires Dietry fire is olletive term for group of sustnes of vrying hemil omposition, struture, physil properties nd physiologil effets. [4] A finl greement ws only rehed in Novemer 2008 on glol dietry fire definition for the Codex Alimentrius. [5] Now the Codex defines dietry fire s rohydrte polymers with 10 or more monomeri units, whih re not hydrolysed y the endogenous enzymes in the smll intestine of humns. Suh n greed definition filittes onsistent pplition of lelling nd helth lims, thus ensuring lrity nd enhning onsumer onfidene. [4] Nowdys, whole rnge of fires re ville in the mrket, ut sometimes it is diffiult to hoose properly euse of the vrition in their physiohemil properties. Numerous fires from ompletely different soures hve een isolted nd hrterized nd inorported into different food produts. [6] In ft there re now mny fire-enrihed produts on the mrket. [7] Fires hve een inorported in wide vriety of foods, like diry produts, [8-12] met or fish [13,14], ut kery produts re the preferred soure of dietry fire. [1,3,4,6,15-18] However, there re few referenes to mshed pottoes with dded dietry fires. Nonetheless, some fires otined from lge, like rrgeenns, or produed y Xnthomons mpestris, like xnthn gum, hve een used for tehnologil purposes in mshed pottoes. [19,20] The gretest improvement in overll eptility (OA) of F/TM pottoes hs een hieved y ddition of kpp-rrgeenn (κ-c) nd xnthn gum (XG) (eh hydroolloid t 1.5 g/kg); the improvement of F/TM 3

4 82 83 potto texture ws sried to the retrding of strh retrogrdtion, inresed wter inding pity nd enhnement of the prinipl hrteristis determining onsumer eptne. [21] Bering in mind the neessity of inresing dietry fire ingestion (espeilly in Western soieties) nd tht ustomers demnd not only helthier foods ut lso high sensory qulity, [15] this reserh work ws minly foused on the possiility of offering F/TM pottoes with improved nutritionl vlue nd with high onsumer eptility. This work inludes systemti study on the effet of fires of different origin insolule (pe fire) nd solule (inulin) dietry fires, dded singly (PF nd I respetively) nd in ssoited lends (PFI) on the texturl nd rheologil properties nd the mirostruture of proessed mshed pottoes. Thus, olour, totl solule solids ontent nd OA of the ove smples were investigted to evlute the produt s pperne nd mke onsumer-eptle fire-enrihed F/TM pottoes MATERIAL AND METHODS Mterils The pottoes used were tuers (Solnum tuerosum, L., v Kennee) from Aguilr de Cmpoo (Pleni, Spin). PF (Vitel Pe Fire EF 150, J. Rettenmier & Sohne GmH & Co, Rosenerg, Germny) ws n insolule fire with totl dietry fire ontent ~ 70 g/100 g (dry mtter), of whih: insolule dietry fire ontent 65 g/100 g, solule dietry fire ontent 0.5 g/100 g nd resistnt strh ontent 1.5 g/100 g, ording to produer s dt. Inulin (I) (Orfti HP, BENEO-Orfti, Tienen, Belgium) ws long-hin I with degree of polymeriztion, DP (totl numer of frutose or gluose units) > 23 nd purity of 99.5% (produer s dt). κ-c (GENULACTA rrgeenn type LP-60) nd XG (Keltrol F [E]) were donted y Premium Ingredients, S.L. (Giron, Spin). Following rnge finding experiments, the lower nd upper fire levels to e used were set t 0 nd 45 g/kg for PF lone nd t 0 nd 60 g/kg for either I lone or PFI lends. Conentrtions nd odes used for identifition of smples re shown in Tle Preprtion of F/TM Pottoes 4

5 Tuers were mnully wshed, peeled nd died. All mshed pottoes were prepred in ~ 1350-g thes from g/kg of pottoes, g/kg of semi-skimmed in-ottle sterilised milk (ft ontent, 15.5 g kg 1 ), g/kg of wter, 7.7 g/kg of slt (NCl) nd 1.5 g/kg of either κ-c or XG 21 using TM 31 food proessor (Vorwerk Espñ, M.S.L., S.C., Mdrid, Spin). I (onentrtion 0-60 g/kg) ws previously dissolved in g/kg of wter nd milk t 70 C for 15 min, stirring onstntly with mgneti stirrer. The ingredients were first ooked for 30 min t 90 C (lde speed: 0.10 g). The mount of liquid evported ws determined y weighing the ingredients efore nd fter first ooking. This liquid ws then repled y dding milk. In terms of proessility, there were serious diffiulties in ooking PF fire together the rest of the ingredients, espeilly in ses where PF levels exeeded 15 g/kg. PF (onentrtion 0-45 g/kg) ws previously hydrted t rtio of PF to wter of 1:6, nd then lso dded t this point. Next, ll the ingredients were ooked for n dditionl 5 min t 90 C. The msh ws immeditely ground for 40 s (lde speed: 80 g), 20 s (lde speed: 450 g) nd 20 s (lde speed: 1000 g) then homogenized through stinless steel sieve (dimeter 1.5 mm). Following preprtion, mshed pottoes were pled on flt freezing nd mirowve thwing trys nd then frozen y fored onvetion with liquid nitrogen vpour in n Instron progrmmle hmer (model ; - 70/+250 C) t -60 C until their therml entres rehed -24 C. After freezing, smples were pked in polyethylene plsti gs, seled under light vuum ( 0.05 MP) on Multiv pking mhine (Sepp Hggenmüller KG, Wolfertshwenden, Germny), pled in domesti freezer for storge t 24 C nd left there for t lest 1 month efore thwing to ssure the pproprite experimentl repetility of the freezing nd frozen storge proesses. Pked frozen smples were thwed in Smsung M1712N mirowve oven (Smsung Eletronis S.A., Mdrid, Spin). Smples were heted for totl of 20 min t n output power rting of 600 W, s detiled elsewhere. [22] After thwing, the temperture rehed t the produt therml entre ws mesured in ll ses (+50 ± 5 C). All smples were rought up to 55 C y pling them in Hetofrig CB60VS wter-th (Heto L Equipment A/S, Birker testing temperture ws 55 C. [22] Instrumentl Texturl Properties d, Denmrk). Smple Bk extrusion (BE) nd one penetrtion (CP) tests were performed using TA.HDPlus Texture Anlyser (Stle Miro Systems Ltd, Godlming, UK) equipped with 300 N lod ell. For performne 5

6 of BE tests, rig (model A/BE, Stle Miro Systems) ws used onsisting of flt 45 mm dimeter Perspex dis plunger tht moved within 50 mm inner dimeter Perspex ylinder smple holder 144 ontining 50 1 g of mshed pottoes. Produt ws extruded to distne of 20 mm t 2 mm/s ompression rte. Mximum positive fore of extrusion (firmness (N), BEF) ws reorded. For performing the CP tests, TTC spredility rig (model HDP/SR, Stle Miro Systems) ws used onsisting of 45 onil perspex proe (P/45 C) tht penetrted onil smple holder ontining g of produt. Produt ws penetrted to distne of 17.5 mm t 3 mm/s ompression rte. Mximum fore of penetrtion per grm of produt (firmness (N/g), CPF) ws reorded. Texture mesurements were performed in qudruplite nd results verged Osilltory nd Stedy Rheologil Mesurements A Bohlin CVR 50 ontrolled stress rheometer (Bohlin Instruments Ltd., Cirenester, Glouestershire, UK) ws used to ondut smll mplitude osilltory sher experiments nd stedy sher using plte-plte sensor system with 2 mm gp (PP40, 40 mm) nd solvent trp to minimize moisture loss during tests. Before mesurements were tken, smples were left etween the pltes for 5 min equilirtion time. [21] Temperture ontrol t 55 C ws hieved with Peltier Plte system (-40 to +180 C; Bohlin Instruments). Dynmi rheologil tests were performed under the following onditions: (1) liner visoelsti domin determined for eh smple from stress sweeps t 1 rd/s, nd (2) three frequeny sweeps over the rnge rd/s. The pplied stress ws seleted to gurntee the existene of liner visoelsti response ording to previous stress sweeps rried out. A new smple ws used eh time for rheologil mesurements, whih re therefore verge vlues of four determintions. Vlues of phse ngle (δ, deg), storge modulus (G, P) nd loss modulus (G, P) were reorded t 1 rd/s. Flow urves were otined t sher rtes of /s pproximtely, whih is the rnge of interest in food texture studies. [23] Visosity vlues in the upwrd visosity/sher rte urves t sher rte of 50 1/s (η pp,50, P s) were tken s the pprent visosity of the smples. Dt were lso fitted to the Ostwld de Wele model, [24-26] where n is the flow ehviour index nd K (P s n ) is the onsisteny index Other Qulity Prmeters 171 6

7 Instrumentl mesurement of olour of the F/TM pottoes in the pots ws rried out with HunterL model D25 (Reston, VA, USA) olour differene meter fitted with 5 m dimeter perture. Results were expressed in ordne with the CIELAB system with referene to illuminnt D65 nd visul ngle of 10. A olour index, the yellowness index (YI), ws lulted s */L*. [27] The totl olour differene ( E*) etween respetive ontrols mde without dded fires (C1 onentrtions, Tle 1) nd the rest of F/TM smples mde with the orresponding dded fires ws lulted s follows: E * L n K L S L L 1 2 C n K C S C C 1 2 H n K H S H H 1 2 1/ where L 1 *, C 1 * nd H 1 * re the referene (fresh smple) olour prmeter vlues nd S L = 1.0, S C = C 1 * nd S H = C 1 *. K L, K C, nd K H re prmetri ftors nd vry with the experimentl onditions. All were given fixed unit vlue for the omintion of our speifi referene onditions. [19,18] Totl solule solids (TSS) ontent (g/100g (w/w)) s mesured y refrtive index ws determined with n Atgo (Itshi-ku, Tokyo, Jpn) dx-30 refrtometer. Mesurements of olour nd TSS ontent were performed in qudruplite nd the results verged Sensory Anlysis Sensory ssessment ws onduted y 14-memer untrined pnel. F/TM pottoes were sujeted to n overll eptility (OA) test sed on ll sensory ttriutes (texture, olour, tste), on nine-point hedoni sle (with 8 m) lelled t eh nhor: (left nhor: 1 = dislike extremely; right nhor: 9 = like extremely) Snning Eletron Mirosopy (SEM) F/TM potto mirostruture ws exmined y SEM using Hithi model S mirosope (CENIM- CSIC). MP smples were ir-dried then mounted on metl holders, followed y gold sputter-oting (200 Å pprox.) in SPI diode sputtering devie. Photomirogrphs were tken t vrious mgnifitions with digitl system Snvision 1.2 of RONTEC (800x1.200 pixel). 7

8 Sttistil Anlyses A one-wy ANOVA ws pplied to evlute how fire types (PF nd I) nd lends (PFI) ffeted the texturl nd rheologil properties, olour prmeters, TSS ontent nd the OA of the F/TM pottoes. Liner disriminnt nlysis (LDA) ws performed to fous on the possile ojetive disrimintion etween the different types of fires inorported into the smples on the sis of their qulity ttriutes. One-wy ANOVA were pplied to eh fire type (PF, I nd PFI lends) to evlute how the onentrtion of eh fire y itself or of lends of these (Tle 1) ffeted the ove mentioned qulity prmeters of the produts. Minimum signifint differenes were lulted using Fisher s lest signifint differene (LSD) tests with 99% onfidene intervl for the omprison of instrumentl prmeters, nd 95% onfidene intervl for omprison of OA. Sttistil nlyses were performed with Sttgrphis softwre version 5.0 (STSC In., Rokville, MD, USA). 215 RESULTS AND DISCUSSION Effet of Fire Types nd Blends on Instrumentl Mesurements nd OA of F/TM Pottoes For omprtive nlysis of the influene of the two insolule nd solule dietry fires (PF nd I respetively) nd their inry lends (PFI) on the BEF, CPF, δ, G, G, η pp,50, n, K, YI, E* nd TSS vlues nd the OA sores of the F/TM pottoes, n ANOVA of one ftor ws performed (Tle 1). One n see tht the smples with only solule fire (I) presented signifintly lower CPF, G, G, K nd YI vlues nd higher δ vlues nd OA sores thn their ounterprts with either dded PF lone or PFI lends. None of the smples differed sttistilly in the vlues of pprent visosity nd n index. With regrd to onsumer eptne, long-hin I n t s ft mimeti thnks to its pity to form mirorystls, whih intert with eh other to form smll ggregtes tht olude lrge mount of wter; this retes fine nd remy texture tht gives mouthfeel similr to tht of ft. [8,9] In ontrst, the lowest wter sorption ws found with the ddition of I to whet flour. [16] 229 8

9 Aording to their mehnil spetr (Fig. 1), the response of smples with oth 30 g/kg dded I or PF lone, or those ontining 30 g/kg of eh fire (C7 lend, Tle 1), ws typil of ross-linked polymer network; [1] i.e. the storge modulus G ws higher thn the loss modulus G t ny given frequeny, whih is hrteristi of visoelsti solids. The shpe of the rheologil spetrum ws similr for ll the other smples. It is worth noting the effet of I ddition tking into ount the smples with PF inorportion. The ddition of ny I onentrtion produed softening of the struture, nd in the smples with dded I lone there ws slightly higher frequeny dependene (prtiulrly t low frequeny) with more signifint drop in G vlues thn in G vlues. This result shows tht the ddition of I onferred different lending properties on the F/TM pottoes, surely due to the frutose polyshrides in its omposition. [16] One of the possile mehnisms y whih I inresed the phse ngle whih whih gives reltive mesure of the energy lost versus energy stored in the yli deformtion ( δ of 90 deg indites the mteril is fully visous) is the presene of smll I prtiles. [29] Suh smll prtiles n e produed y ontrolling heting nd ooling of the dissolved I to indue nuletion nd thus smll insolule rystl formtion. [30] The presene of I hs een rported to filitte the redution of snk r hrdness, [4] nd negtive effet of I hs lso een oserved, mking mkerel surimi gels softer when inorported t 40 g/kg, wheres PF inresed hrdness (n inrese of lmost 60% with 40 g/kg). [14] Furthermore, other studies hve shown tht inorportion of low levels ( 40 g/kg) of inner PF nd hiory root I my lter the texture of restrutured fish produts; i.e. inner PF fire fvoured greter gel strength nd hrdness, [13] wheres I fire redued hrdness With regrd to YI index, ommerilly ville I is spry-dried white powder whih gives n opque gel t higher onentrtions when mixed properly with wter. [31] This whitish powder nd the opque gel mde from it redued the yellowness nd inresed the E* of the smples (Tle 1). In ddition, the TSS ontent inresed when I ws dded either lone or in ssoited lends. Supplementtion with I introdued some more reduing sugrs nd led to inresed totl rohydrte ontents. [10] The sme ehviour ws oserved in diry drinks with dded oligofrutose [11] nd I [8] respetively In ontrst, no signifint differenes in CPF, δ, G, G nd K vlues were deteted etween smples with dded PF lone nd with dded PFI lends (Tle 1, Fig. 1). Therefore, the presene of insolule fire inresed the firmness per grm, elstiity, visosity nd onsisteny of the smples, evidening tht I is 9

10 unle to ountert the effet produed y the high proportion of ellulose present in PF fire. This is likely due to intertions etween the fire struture nd the wter. In whet dough, the highest wter sorption ws found with the ddition of PF, followed y ro fire. [16] Note, however, tht firmness (BEF vlue) inresed when PF ws dded lone, wheres PF ddition in ssoited lends left BEF silly unhnged s ompred to the ddition of I lone. The pprent lk of orreltion etween the two types of texturl mesurements (BEF nd CPF per grm) indites tht oth prmeters re 266 mesuring different effets in the mtrix of the F/TM pottoes. Aording to Lee nd Chung, [32] while the ompression test mesures the overll inding property of the gel mteril, the penetrtion test evlutes the degree of omptness. If we ept this interprettion, PF onsiderly inresed either overll inding or the degree of omptness in the mshed potto gel network, ut the presene of I onsiderly redued overll inding nd the struture of the gel eme weker. In ddition, other uthors hve reported tht t very high I onentrtions oth potto strh nd I proly gel independently, with potto strh remining in the morphous stte [33] nd I rerystllizing. [34] In spite of the ft tht ANOVA onfirmed tht there ws little vriility etween some of the instrumentl mesurements of F/TM pottoes formulted with either dded PF lone or PFI lends, LDA using fire type s the disriminnt ftor showed tht the three smples type investigted were lerly disriminted. LDA ws performed to develop two funtions to disriminte etween the two types of fire nd their lends sed on liner omintions of the twelve qulity ttriutes shown in Tle 1. The reltive perentges of vrine ounting for disriminting funtions 1 nd 2 were quite similr (59 nd 41% respetively) inditing tht none of the funtions on its own n disriminte well etween the 3 levels of the ftor. The stndrdized oeffiients of the funtions used to disriminte mongst the types of fire re not shown for the ske of revity. However, for the first disriminnt funtion, YI nd TSS ontent re lerly the ttriutes whih est seprte the PF, I nd PFI smples, wheres for the seond funtion, BEF, G nd OA re the ttriutes tht est seprte the smples y fire ontent. Fig. 2 shows the projetions of the oservtions onto the two xes extrted y the LDA. There ppers to e ler disrimintion etween smples with oth dded PF nd I lone or their lends. However, verge vlues (group entroids) of the two disriminnt funtions for eh of the three fires type re inluded in Fig. 2; they re mrked y smll plus sign. Centroid vlues show tht funtion 1 disrimintes minly etween F/TM smples with dded PFI lends nd those with either PF or I lone (it seprtes them y 10

11 more thn 3 units). Note tht the first funtion hrdly disrimintes etween smples with either dded PF or I lone. On the other hnd, funtion 2 disrimintes minly etween smples with dded PF lone nd those with dded I lone (it seprtes them y more thn 4 units), leit the seond funtion lso disrimintes quite well etween smples with dded PFI lends nd those with either dded PF or I lone (it seprtes them y more thn 2 nd 1 unit respetively). These results indite tht the ddition of oth fires nd their lends ppers to interfere with the struture of F/TM pottoes differently Effet of Fires Conentrtion on Qulity Attriutes of F/TM Pottoes To study the influene of the onentrtion on the qulity ttriutes nd the OA sores of the F/TM pottoes, three different one-wy nlyses were performed on eh fire type nd lends of these. Figures 3 nd 4 show the effet of fire onentrtion on the texturl (BEF nd CPF) nd rheologil properties (δ, G, G, η pp,50, n nd K) of F/TM pottoes with either PF or I dded lone or PFI lends. In terms of texture nd osilltory dynmi mesurements, fire onentrtion signifintly ffeted (P 0.01) the mjority of these properties. Only in the se of the δ vlue (Fig. 3) s funtion of onentrtion ws there no signifint hnge in F/TM pottoes with dded I lone In F/TM smples, dding g/kg PF lone (C2-C4 onentrtions respetively; Tle 1) inresed oth BEF (Fig. 3) nd CPF (Fig. 3) properties s ompred to mshed pottoes without dded PF fire (C1 onentrtion). These results indite tht muh stronger strutures re produed when F/TM pottoes ontin insolule fire. The mximum BEF nd CPF vlues were reorded in smples ontining 15 nd 30 g/kg respetively, leit differenes in BEF nd CPF vlues etween smples with 15 nd 45 g/kg dded PF were not signifint. With regrd to I ddition, the highest BEF vlue ws reorded in the smples ontining 15 g/kg I. There were no signifint differenes either in BEF vlues for I t g/kg or in CPF for I t 15 g/kg (Fig. 3) ompred with their ontrol (C1), wheres dding higher levels of I (30-60 g/kg) signifintly redued the CPF vlues of the smples With regrd to the ddition of PFI lends, there were no signifint differenes in the BEF vlues of the smples with dded C2 nd C5 lends (oth with 15 g/kg dded PF, Tle 1) nd the C1 ontrol without fire. The vrition in the CPF vlue (Fig. 3) shows tht firmness per grm ws lowest in the smples with dded C5 lend (whih hd the lowest (15 g/kg) nd the highest (45 g/kg) PF nd I ontents 11

12 respetively), even lower thn in the C1 ontrol. Furthermore, smples with higher PF onentrtions (C4, C5, C6 nd C7 lends) hd higher BEF nd CPF vlues thn C1 ontrol, leit differenes etween smples were not signifint. The texture mesurements show tht firmness tended to inrese with PF ontent nd derese with I ontent, exept in the se of the BEF vlue with the lower level of I (Fig. 3). Results show tht the texture tests did not dequtely disriminte the hnges in smple struture produed y the fire ddition. In the se of the PF fire, this effet my hve een due to defiient wetting of the fires t the higher onentrtions euse of their high wter holding pity, nd lso to the formtion of lusters owing to the reltive length of the pe fires. [1] Smples with dded PF hd lower δ vlues (Fig. 4), inditing greter elstiity, lthough differenes etween phse ngles of F/TM pottoes with dded PF lone were not signifint. Conversely, the effet of the onentrtion ws liner nd elstiity (G vlues) inresed t the three onentrtions used (Fig. 4). In the smples with dded I, the effet of the onentrtion ws lso liner, with δ inresing nd G deresing t the four test onentrtions tested. Nevertheless, there were non-signifint differenes etween δ vlues due to greter dispersion of the results; lso, the smples ontining 15 nd 30 g/kg I were more elsti thn the ontrol. This ft is proly relted to I low onentrtion, s the system does not hve enough prtiles nd/or moleulr density of I hins to reh ritil rowding effet. [35] For their prt, smples with dded C3, C6 nd C7 lends, ll with 30 g/kg dded PF, hd signifintly lower δ vlues thn the ontrol (Fig. 4), inditing tht high levels of PF fvoured lower visosity. [36] G inresed when ny of the PFI lends studied ws dded, nd only the smples formulted with the lend ontining the lowest PF onentrtion (C2) hd similr elstiity to the ontrol (Fig. 4). On the other hnd, F/TM pottoes with dded C6 lend ontining the highest PF onentrtion exhiited the gretest elstiity with higher mgnitude of G, followed y smples ontining 30 g/kg dded PF (C3 nd C7 lends) With respet to the visosity of the smples (Fig. 4), the mximum loss modulus G ws likewise reorded with the highest level of PF dded lone. Addition of the highest I onentrtion (C5, 60 g/kg) redued the visosity, leit differenes with ontrol were non-signifint. Agin, the mximum G vlue ws reorded in smples ontining 15 g/kg I. The hnges in the vlues of G on dding PFI lends were similr to those oserved in elstiity (Fig. 4). In ddition, no signifint differenes were found in the 12

13 visous properties of the smples without dded fires nd those with dded C2 nd C5 lends (with lower PF ontents nd with 30 nd 45 g/kg I respetively) Stedy sher mesurements: the effet of onentrtion on the onsisteny index (K) ws not signifint for smples with dded PF lone (Fig. 5). There ws no signifint hnge in the n index (Fig. 5) s funtion of onentrtion in F/TM pottoes with dded I lone. Conentrtion likewise hd no signifint effet (P 0.01) η pp,50 (Fig. 5) nd K (Fig. 5) in F/TM pottoes with dded PFI lends. In F/TM smples with dded PF lone, the effet of the onentrtion ws liner, with η pp,50 deresing (Fig. 5) t the three onentrtions used. The n index, whih indites the extent to whih sher-thinning devites from 1, ws lso redued y dding 45 g/kg (Fig. 5), lthough in this se the pseudoplstiity of smples ontining lower PF onentrtions (15 nd 30 g/kg) ws not signifintly lower thn in the smples ontining 45 g/kg I With regrd to I ddition, the mximum η pp,50 nd K vlues were reorded in smples ontining 15 g/kg I. Addition of the highest I onentrtion (C5, 60 g/kg) lso signifintly redued onsisteny (Fig. 5). Only the ddition of I t 15 g/kg inresed pprent visosity (Fig. 5), wheres the η pp,50 of the smples with 30, 45 nd 60 g/kg ws similr ut signifintly lower thn in the ontrol. In the former se, there re vrious hypotheses tht ould explin this ehviour. Aording to Zimeri nd Kokini, [37] it ould e due to the ft tht the struture of I hs no entnglements t ll, with gglomertes (rystls) sliding one on top of the other. The ltter uthors lso indited tht the rheologil properties of mixed I-wxy mize strh (WMS) systems t rtios of I to WMS of 50:50 presented derese in the moduli s mgnitude due to WMS network disruption y I, when ompred to gels with lower I onentrtions. Severl studies hve ddressed mixed rohydrte intertions, nd it hs een found tht their misiility in solution dereses with inresing onentrtion Addition of PFI lends: the nlysis of vrine showed tht the effet of the different onentrtions of fires on oth η pp,50 nd K ws only signifint t the 0.05 level. In ft, unexpetedly ll the smples with dded PFI lends hd lower pprent visosity thn the ontrol (Fig. 5); however, when 45 g/kg of PF nd 15 g/kg of I (C6) ws dded, the η pp,50 vlue deresed signifintly with respet to the C1 ontrol, inditing tht ddition of high PF onentrtion produed slight wekening of the initil 13

14 struture of the system. With respet to the flow index (n) (Fig. 5), on inresing the PF onentrtion nd reduing the of I onentrtion (C3, C4, C6 nd C7 lends), the n vlue deresed, lthough the inrese in pseudoplstiity ws greter in the flow of the smples with the lower I onentrtions. Similrly, the lowest smple onsisteny ws reorded for the F/TM pottoes with dded C2 lend, inditing tht in the presene of low PF ontent the ddition of 30 g/kg of I used slight softening of the originl struture of the system. At PF onentrtions ove 30 g/kg, on the other hnd, suh ddition inresed smple onsisteny lthough the differenes with the ontrol were not signifint. Note tht in the presene of PF fire there ws no orreltion etween the η pp,50 nd K vlues, whih my e ttriuted to erly fixing of the struture nd high level of wter in the produt. These results re in disgreement with those reported y Fernández et l. [20] fter dding five different hydroolloids nd two diry proteins to mshed pottoes. The high orreltion etween η pp,50 nd K ws ssoited with the presene of shift from visoelsti regime to purely visous one s found for seleted gum solutions, [38] sine K is the sher stress t sher rte of 1.0 s -1. In ontrst, when pplying extensionl rheometry t lrge strin deformtions, Piteir et l. [1] oserved tht PF inorportion in ookie dough did not seem to ffet the dough s unixil trnsient flow t strins ove pproximtely 0.1. On the ontrry, etween pproximtely 0.01 nd 0.1 strin units, the trnsient extensionl visosity seems to derese the more is inorported. Also, it hs een reported tht oth mixing tolerne index nd elstiity were redued y the ddition of PF to whet dough. [16] Moreover, the extensiility of the dough nd the deformtion energy were gretly redued y PF, yielding redution of out 42% in the extensiility of the whet flour, nd hene smller ule efore filure, proly due to the high ellulose ontent in this fire. Thurston nd Pope [39] suggested tht the derese in the visous omponent t high sher rte reflets loss of the ility of ompounds to store elsti energy in the sher deformtion proess. A similr sene of orreltion etween smll nd lrge deformtion tests hs een reported for other gels. [14] Figure 6 shows the effet of fire onentrtion on olour prmeters, TSS ontent nd OA of F/TM pottoes with PF or I dded lone or PFI lends. In generl terms, when dded singly or in ssoited lends fire ontent signifintly ffeted (P 0.01) oth the olour prmeters nd the TSS ontent of the smples. Only in the se of F/TM pottoes with I dded lone ws there no signifint hnge in OA sore s funtion of onentrtion. In smples with dded PF lone, dding the lowest level of PF 14

15 slightly redued the YI vlue s ompred with ontrol (Fig. 6). Conversely, yellowness inresed signifintly when 30 nd 45 g/kg PF ws dded. Cod susges hve lso een reported to eme yellower with pe protein. [40] In smples with I dded lone, dding g/kg I redued the YI vlue s ompred with the ontrol, lthough there ws no signifint differene in the yellowness of the different mshed pottoes formulted with higher I onentrtions (30-60 g/kg). Conversely, it hs een found tht I did not ffet the olour vlues of low-ft ie rem. [41] And gin, ddition of oligofrutose to diry drinks did not ffet the L*, * nd * vlues. [11] Oviously, the olour of the ingredients used influenes the olour of the produt. The PF ingredient ws rown powder, nd the F/TM pottoes with dded PF were drker rown in olour, wheres the smples without PF were yellower in olour. In ft dding inresing levels of PF inresed oth L* nd * vlues, nd s result, the F/TM pottoes with dded PF fire were drker (lower L*/* rtio) thn the C1 ontrol. On the other hnd, I is white powder nd produed lighter-oloured mshed potto (higher L*/* rtio) thn the ontrol, reduing the YI vlue. In the se of lend ddition, ll smples exept the ones with the lowest PF ontent (C5) hd signifintly higher YI vlues thn the ontrol (Fig. 6), inditing tht the drkening produed y PF ddition ws stronger thn the lightening used y I inorportion. Curiously, s Fig. 6 shows, the vlue of E* (the referene tken in the three ses ws the olour of the respetive C1 ontrols) inresed linerly when the onentrtion of either PF or I inresed. Comprison of the totl olour differenes suggests tht the ddition of the sme onentrtion of either fire lone to F/TM pottoes ffeted the finl olour in the sme wy, leit PF nd I used smples to drken nd righten respetively. In smples with dded PFI lends, the vlue of E* with respet to the C1 ontrol ws muh higher t the lowest PF onentrtion (C2 lend, Tle 1), suggesting tht in this se the lightening used y I ddition msked the effet of PF ddition TSS ontent deresed linerly with PF ontent (Fig 6). PF ontins mostly insolule fire with high wter sorption nd some solule solids my e hrd to detet. However, s ws expeted, totl solule solids were higher in smples ontining I thn in the ontrol, whih grees with the findings reported y other uthors [11] in proioti yogurts supplemented with fruto-oligoshride. Also, in the se of the ddition of lends, those smples with lower PF onentrtions nd higher I ontents (C2 nd C5) hd higher TSS ontents

16 440 Sensory hrteristis re influened y mirostruture. Addition of PF fire redued the OA sore of 441 the F/TM pottoes (Fig. 6d), wheres Wng et l. [16] found tht reds ontining PF were judged eptle y the pnellists. In this study, pnellists preferred F/TM pottoes without dded PF. There ws no differene (P 0.05) in the verge sores of smples ontining PF fire, inditing tht the different pe fire onentrtions used did not interfere in the eptility of the smples. Sores were notiely lower in these ses, proly due to the onsiderle differenes in texture, olour, flvour nd odour of these smples s result of the originl properties of the PF fire. The undesirle effet of fire on red onsists minly of n ojetionle gritty texture nd unsuitle tste nd mouthfeel. [7] In ft PF signifintly influened smple texture, whih ws pereived s sndy euse of the high proportion of ellulose. Also, PF inresed pereived hrdness when dded to F/TM pottoes either lone or t the mximum tested levels. The sensory nlysis indites tht PF lone nnot e used s n ingredient in mshed pottoes to fortify the diet. There were no signifint differenes in OA sores for F/TM pottoes without nd with dded I t ny onentrtion. Unlike other reports, [8,12] there ws no ler reltionship etween the sensory sore nd the quntity of solule fire dded individully, whih is likely relted to the presene of XG in ll the systems. Addition of XG t ny of the onentrtions used, either individully or lended with κ-c, inresed the reminess of F/TM pottoes. [21] However, severl reports lso orroorte the influene of I on suh mouthfeel prmeters s reminess nd thikness. The ddition of I to low-ft nd whole milk set yoghurt influenes the pereption of reminess. With rising I onentrtion, the pereption of reminess inreses s well. [12] Also, milk-everge model systems with dded I were pereived s signifintly remier thn smples without I. [8] With respet to the intervl of onentrtions onsidered in this work, inorportion of g/kg I did not signifintly ffet the pereived reminess nd mouthfeel of the produts, nd the resulting texture were pereived s smoother The vrition in the OA sore for PFI lends indites tht with the lowest PF onentrtion nd I t 30 g/kg (C2 lend), onsumer eptne ws omprle to ontrol without dded ingredients. Also, smples with 15 g/kg dded PF nd 45 g/kg dded I (C5 lend) were judged y the onsumer pnellists s very eptle, soring > 7 for OA (Fig. 5d). Therefore, the ssoition of oth fires t minimum levels of PF nd mximum levels of I lrgely ounterted the hrdening effet of ddition of PF lone. On the 16

17 other hnd, smples supplemented with inry lends ontining 30 g/kg PF (C3, C6 nd C7 lends) sored onsiderly less thn those with 10 nd 15 g/kg ddition SEM Mirostruture ws nlysed y SEM (Fig. 7 nd 8). The photomirogrphs in Fig. 7, d nd f nd in Fig. 8, d nd f show lose-up views of eh smple. Those F/TM pottoes ontining PF lone exhiited ler differenes. When food nd other iomterils re frozen, the expnsion ssoited with phse hnges nd susequent ontrtion due to further ooling will use mehnil stresses to develop, whih my use deformtion nd lrge-sle rking or miro-struturl dmge. This topi hs een reltively negleted in the food literture, ut it ws lerly visile in ll the smples formulted with PF lone either t low (Fig. 7 nd ) or t high onentrtions (Fig. 7 nd d). Kim nd Hung [42] suggested tht the propensity to rk in liquid nitrogen immersion freezing depends on severl interrelted physil properties suh s the porosity nd frozen density of the food. Density influenes rking euse it is proportionl to moisture ontent, nd produts with higher moisture ontent will hve higher volume expnsion rtio during freezing nd hene higher rking suseptiility. In this study moisture ontents were not mesured, ut it is definitely prole tht the F/TM smples with dded PF lone hd higher moisture ontent thn the other smples, [14,34] espeilly given tht these smples ontined more wter in order to ssure hydrtion of PF fire In this study drip loss fter entrifugtion ws lwys zero in ll smples; this ft is ttriuted to the presene of XG in the systems, inditing tht the wter inding pity of the F/TM pottoes ontining this gum ws sslute. [21] However, in their respetive ounterprts without dded ryoprotetnts (dt re not shown), the presene of PF inresed wter inding pity s ompred to I. Similrly, wter sorption in whet dough ws highest when PF ws dded. [16] This is likely used y the lrge numer of hydroxyl groups in the PF fire struture, whih llow more wter intertions through hydrogen onding. Therefore, F/TM pottoes with PF dded lone re more rigid (s evidened y instrumentl mesurements), hve more moisture ontent nd hve no or very little expnsion due to phse hnge; this would rete extremely high tngentil tensile stress t the surfe [42] s the inner lyers freeze nd expnd, strething the rigid outer rust. Thus, rking is unlikely to our during thwing. [43] It is 17

18 possile, however, tht in the produts with high PF ontent, the rks tht hve formed nd propgted during freezing will ontinue towrds the entre during thwing nd omplete the splitting proess, espeilly givn the limited wter ville nd restrition of wter movement during thwing of the produts, whih is quikly entrpped y the PF fire. Unlike PF, I inorportion did not use rking (Fig. 7e nd f), nd in the se of smples with dded PFI lends, some rking ws only detetle (Fig. 8) with the lowest I onentrtion (C3 lend, Tle 1). The higher the I ontent, the greter is the hne tht internl stress will dissipte insted of umulting. This is proly relted to tht ft tht I inorportion inreses porosity. F/TM smples onsisted minly of whole single potto ells nd of some ruptured ells nd ell frgments emedded in n extrellulr strh phse, whih ws lended with gelled κ-c nd XG. This gum ws effetive in stilizing F/TM pottoes ginst strh retrogrdtion, minimizing the spongy struture formtion of frozen gel. [44] Smples with 15 g/kg dded PF (Fig. 7 nd ) present more dehydrted pperne, sine prt of the intrellulr wter ws drwn out osmotilly when the produt ws frozen euse of freezing-indued onentrtion of the ell mss. Moreover, these smples seemed to present some whitish rnds, whih were most ovious in the smples ontining higher PF onentrtions either lone (Fig. 7 nd d) or lended with I (Fig. 8-d). The ddition of PF to the produt ppered to disrupt the ontinuity of the potto strh mtrix, produing mirostruturl rrngement, undulting surfes (Fig. 7, 8 nd ) nd, unlike other repots, [14,18], more ompt morphology. A lose exmintion of the smples with dded PF t quite high onentrtion lso reveled the formtion of wrinkles, ridges nd folds (Fig. 7d, 8 nd d). Mromoleules show preferene for eing surrounded y their own type in mixed solutions, nd onsequently self-ssoition is intensified in the presene of other mromoleules. [3] F/TM potto mirostruture is explined y dietry fire omposition, nd PF ontins insolule polyshrides. [14] The texture of the mshed pottoes ws possily reinfored y intertions etween the PF s polyshrides-i.e. insolule dietry fire forming ellulose-rih kone struture-nd strh. [16] Photomirogrphs of the F/TM pottoes with 30 g/kg dded PF nd 10 g/kg dded I show the PF-potto strh mtrix to e well formed, with strong, ontinuous fire strnds entrpping some strh grnules (Fig. 8 nd ) However, even though it is not esy to differentite I from potto strh grnules, smples with 60 g/kg dded I (Fig. 6e nd f) nd with 15 g/kg dded PF nd 45 g/kg dded I (Fig. 7e nd f) ll hd n I-rih 18

19 phse with smll I rystllites forming ontinuous network. These rystllites re hrd to mke out when I ontent is lower (Fig. 7d). At 45 nd 60 g/kg, I presentsed mirostruture formed y lusters of smll I rystls. In the se of dded I fire the I-potto strh mtrix ws more ontinuous thn in the 532 smples ontining PF, onfirming previous findings. [18] The mgnitude of the overll texturl nd rheologil properties of these smples deresed, due minly to the integrtion of I rystls in developed potto strh network; this in turn resulted in smoother texture, whih n proly e explined y trhe orreltion etween the presene of I rystls nd the development of gelling properties of I. This reinfores the hypothesis of n I geltion mehnism proposed in other sientifi ppers. [29,31,35] Fig. 6f nd 7f lso onfirm the hypothesis of Zimeri nd Kokini. [37] These uthors stted tht I s struture is formed y gglomertes (rystls), with no interonnetions with one nother. Redution of hrdness of pst hs een reported with inresing I onentrtion, [18], ssoited with the wy in whih the I ws inorported into the struture of the pst. And gin, the strh grnules in psts ontining solule dietry fire ppered to e oted in muilginous-like lyer. [3] CONCLUSIONS F/TM potto qulity ws determined y mens of texture, olour nd TSS ontent. The rheologil ehviour of the F/TM pottoes ws studied y mens of osilltory nd stedy mesurements, estlishing the influene of fire type on the OA of F/TM pottoes. The originl olour of the ingredients lerly influened the olour of F/TM pottoes. Osilltory dynmi tests proved more sensitive thn stedy, k extrusion nd penetrtion tests. Thus, G nd G ws found to e suitle to desrie the hnge in F/TM potto struture in terms of differenes in the ehviour of the produts s funtion of fire type onentrtion. LDA showed tht YI nd TSS ontent re the ttriutes whih est seprte smples enrihed with either PF or I nd PFI-enrihed smples, wheres BEF, G nd OA re the ttriutes tht est distinguish the smples with PF nd I when dded singly. Results show tht F/TM potto texture, struture nd potentil sensoril qulity re intrinsilly linked to the integrtion of fire in F/TM pottoes systems. Mjor differenes in struture my e relted to the soluility of the fire dded. In prtiulr, PF (minly insolule) nd I (solule) fires show mrked differenes in their potentil to ffet the OA of the produts. These differenes my e ttriuted to the ehviour of the fire within the system nd the reltionship of the fire with the potto strh mtrix. This study ppers to onfirm tht 19

20 the ddition of insolule PF to mshed pottoes uses disruption nd rking of the potto strh mtrix, nd onsequently hrdening. Conversely, the ddition of solule I fire results in the entrpment of smll I rystls within soft I-potto strh network ACKNOWLEDGEMENTS The uthors wish to thnk the Spnish Ministry of Siene nd Innovtion for finnil support (AGL ), nd lso P. Adev, I. Amurrio nd A. Grí of the Eletron Misrosopy Lortory (CENIM-CSIC)

21 REFERENCES 1. Piteir, M.F.; Mi, J.M.; Rymundo, A.; Sous, I. Extensionl flow ehviour of nturl fire filled dough nd its reltionship with struture nd properties. Journl of Non-Newtonin Fluid Mehnis 2006, 137 (1-3), WHO. Diet, nutrition nd the prevention of hroni diseses. Report of Joint Fo/Who Expert Consulttion, World Helth Orgnistion Brennn, C.S.; Tudori, C.M. Evlution of potentil mehnisms y whih dietry fire dditions redue the predited glyemi index of fresh psts. Interntionl Journl of Food Siene nd Tehnology 2008, 43 (12), Sun-Wterhouse, D.; Teoh, A.; Mssrotto, C.; Wiisono, R.; Wdhm, S. Comprtive nlysis of fruit-sed funtionl snk rs. Food Chemistry 2010, 119 (4), Alinorm 09/32/26. Report of the 30 th session of the Codex ommittee on nutrition nd foods for speil dietry uses 3-7 Novemer 2008 (p. 46, Appendix II) Rosell, C.M.; Sntos, E.; Collr, C. Physio-hemil properties of ommeril fires from different soures: omprtive pproh. Food Reserh Interntionl 2009, 42 (1), Collr, C. Novel high-fire nd whole grin reds. In Tehnology of Funtionl Cerel Produts; Hmker, B.R.; Ed.; Woodhed Pulishing Limited: Cmridge, 2008; Villegs, B.; Cronell, I.; Costell, E. Inulin Milk Beverges: Sensory differenes in thikness nd reminess using r-index nlysis of the rnking dt. Journl of Sensory Studies 2007, 22 (4) Villegs, B.; Tárreg, A.; Cronell, I.; Costell, E. Optimising eptility of new preioti low-ft milk everges. Food Qulity nd Preferene 2010, 21 (2) Crdrelli, H.R.; Argon-Alegro, L.C.; Alegro, J.H.A.; Cstro, I.A.; Sd, M.I. Effet of inulin nd ltoillus prsei on sensory nd instrumentl texture properties of funtionl hoolte mouse. Journl of the Siene of Food nd Agriulture 2008, 88 (8) Cstro, F.P.; Cunh, T.M.; Brreto, P.L.M.; Amoni, R.D.M.C.; Prudênio, E.S. Effet of oligofrutose inorportion on the properties of fermented proioti lti everges. Interntionl Journl of Diry Tehnology 2009, 62 (1),

22 Guggiserg, D.; Cuthert-Steven, J.; Piinli, P.; Bütikofer, U.; Eerhrd, P. Rheologil, Mirostruturl nd sensory hrteriztion of low-ft nd whole milk set yogurt s influened y Inulin Addition. Interntionl Diry Journl 2009, 19 (2), Crdoso, C.M.L.; Mendes, R.; Pedro, S.; Nunes, M.L. Qulity hnges during storge of fish susges ontining dietry fire. Journl of Aquti Food Produt Tehnology 2008, 17 (1), Crdoso, C.M.L.; Mendes, R.; Vz-Pires, P.; Nunes, M.L. Effet of dietry fire nd mtgse on the qulity of mkerel surimi gels. Journl of the Siene of Food nd Agriulture 2009, 89 (10), Brennn, C.S.; Smyue, E. Evlution of strh degrdtion nd texturl hrteristis of dietry fier enrihed isuits. Interntionl Journl of Food Properties 2004, 7 (3), Wng, J.; Rosell, C.M.; Benedito de Brer, C. Effet of the ddition of different fires on whet dough performne nd red qulity. Food Chemistry 2002, 79 (2), Gómez, M.; Rond, F.; Blno, C.A.; Cllero, P.A.; Apesteguí, A. Effet of dietry fire on dough rheology nd red qulity. Europen Food Reserh nd Tehnology 2003, 216 (1), Tudori, C.M.; Kuri, V.; Brennn, C.S. Nutritionl nd physiohemil hrteristis of dietry fier enrihed pst. Journl of Agriulture nd Food Chemistry 2002, 50 (2), Alvrez, M.D.; Cnet, W.; Fernández, C. Effet of ddition of iopolymers on the mehnil properties, olour nd sensory ttriutes of fresh nd frozen/thwed mshed pottoes. Europen Food Reserh nd Tehnology 2008, 226 (6), Fernández, C.; Alvrez, M.D.; Cnet, W. Stedy sher nd yield stress of fresh nd frozen/thwed mshed pottoes: effet of iopolymers ddition. Food Hydroolloids 2008, 22 (7), Alvrez, M.D.; Fernández, C.; Cnet, W. Enhnement of freezing stility in mshed pottoes y the inorportion of kpp-rrgeenn nd xnthn gum lends. Journl of the Siene of Food nd Agriulture 2009, 89 (12), Alvrez, M.D.; Fernández, C.; Cnet, W. Effet of freezing/thwing onditions nd long-term frozen storge on the qulity of mshed pottoes. Journl of the Siene of Food nd Agriulture 2005, 85 (14),

23 Bistny, K.L.; Kokini, J.L. Dynmi visoelsti properties of Foods in texture ontrol. Journl of Rheology 1983, 27 (6), Ro, M.A. Flow nd Funtionl Models for Rheologil Properties of Fluid Foods. In Rheology of Fluid nd Semisolid Foods. Priniples nd Applitions; Bros-Cánovs, G.V.; Ed.; Aspen Pulishers In: Githersurg, MD, 1999; Au-Jdyil, B.; Jumh, R.Y.; Shker, R.R. Rheologil properties of onentrted fermented produt, lneh, produed from ovine milk: effet of prodution method. Interntionl Journl of Food Properties 2002, 5(3) Alvrez, E.; Cnel, M.A.; Delgdo-Bstids, N.; Meirs, R. Rheologil hrteriztion of ommeril y fruit purees. Interntionl Journl of Food Properties 2008, 11(2) Frnis, F.J.; Clydesdle, F.M. Food Colorimetry: Theory nd Applitions; AVI Pulishing Compny, Westport, Bixuli, R.; Slvdor, A.; Fiszmn, S.M.; Clvo, C. Effet of the ddition of orn flour nd olornts on the olour of fried, ttered squid rings. Europen Food Reserh nd Tehnology 2002, 215 (6), Ronkrt, S.N.; Pquot, M.; Deronne, C.; Fougnies, C.; Beses, S.; Bleker, C.S. Development of gelling properties of inulin y mirofluidiztion. Food Hydroolloids 2010, 24 (4), Héette, C.L.M.; Delour, J.A.; Koh, M.J.H.; Booten, K.; Kleppinger, R.; Mishenko, N.; Reyners, H. Complex melting of semirystlline hiory (Cihorium intyus L.) root inulin. Crohydrte Reserh 1998, 310 (1-2), Bot, A.; Erle, U.; Vreeker, R.; Agterof, W. Influene of rystllistion onditions on the lrge deformtion rheology of inulin gels. Food Hydroolloids 2004, 18 (4), Lee, C.M.; Chung, K.H. Anlysis of surimi gel properties y ompression nd penetrtion tests. Journl of Texture Studies 1989, 20 (3), Zimeri, J.E.; Kokini, J.L. Morphologil hrteriztion of the phse ehviour of inulin-wxy mize strh systems in high moisture environments. Crohydrte Polymer 2003, 52 (3), Zimeri, J.E.; Kokini, J.L. The effet of moisture ontent on the rystllinity nd glss trnsition temperture of inulin. Crohydrte Polymer 2002, 48 (3),