Structure and Physicochemical Properties of Starches from Sieve Fractionsof Oat Flour Compared with Whole and Pin-Milled Flour

David G. Stevenson,'-' Jay-un Jane,' and George E. Inglett'

ABSTRACT

Cereal Chem. 84(6):533-539

One Oat cultivar grown in Idaho (three field sites) was pin-milled andseparated by sieving to investigate whether starch from oat bran differsfrom the remainder of kernel. Ground oat particles were classified intothree sieve fractions: 300-850 pm, 150-300 om and <150 Om). Il-Glucancontent in sieve fractions was analyzed and starch was extracted fromkernels without milling and from kernels of each sieve fraction. -Glucancontents of 300-850. 150-300. and <150 pin fractions were 4.2,2.3, and 0.8%, respectively. Therefore, starch in bran (300-850 pinfraction) and endosperm (< 150 om sieve fraction) were separated. Starchisolated from entire kernels had significantly higher apparent and abso-lute amylose content than starch from the 300-850 .tm sieve fraction.

Starch from different sieve fractions was not significantly different in theapparent amylose, absolute amylose, amylopectin molecular weight, gy-ration radii, starch gelatinization, and amylose-lipid complex thermaltransition temperatures. Starch from the 150-300 pin fraction hadsignificantly lower peak, final, and setback viscosity compared with thestarch isolated from the 300-850 om and <ISO om sieve fractions. Starchremoved from the oat bran fraction during 13-glucan enrichment may havedifferent applications compared with starch obtained from other kernelcompartments. Because pin-milling decreased apparent amylose contentand shortened amylopectin branch chains, its potential to alter starch struc-ture should be considered.

The oat kernel consists of distinct tissues that exhibit structuraland chemical compartmentalization: the hull, bran, starchy endo-sperm, and germ. Starchy endosperm of oat groats (kernels withhulls removed) makes up 80% of the dry weight but the bran hasgreatest influence on overall quality characteristics (Fulcher 1986).Bran is composed of the kernel fibrous components includingpericarp, seed coat, and nucellus, the single-celled aleurone layerthat contains -glucans, and the subaleurone layer that contains afew cell layers of the starchy endosperm. Starch granules of oatkernels are typically clustered in amyloplasts as compound starch(Jane et al 1994). Differing amylose contents of starch have beenreported for inner and peripheral regions of waxy barley kernels(Oscarsson et al 1997; Andersson et al 1999).

Oats contain high levels of 3-glucan hydrocolloids that providehealth benefits to humans (Pick et al 1996; Pomeroy et al 2001).Oat products such as Nutrim-OB with concentrated 3-glucans havebeen developed to assist food processors produce foods that giveconsumers healthy options in their diet (Carrière and lnglett 2000).One approach to producing a highly concentrated 3-glucan ingre-dient is to remove the starch by sieving. To find potential indus-trial applications for starch removed from oat bran, it is importantto find unique properties of starch isolated from these sieve frac-tions of the kernels.

Several studies have been conducted to investigate the thermaland pasting properties of different sieved fractions of cereal flours.Red hard spring wheat flour was reported to have no differencesin peak viscosity for particles of 42-68 jim in size, but breakdownincreased with decreasing particle size (Kurimoto and Shelton1988). Red hard spring wheat flour with particles >53 pinhigher enthalpy change of starch gelatinization, but no differencein gelatinization temperatures, than those with particles <53 jtm(Scanlon et al 1988). Marshall (1992) showed that milled brown

Cereal Products & Food Science Research Unit, National Center for AgriculturalUtilization Research, ARS, USDA, 1815 N. University Street, Peoria. IL 61604.Names are necessary to report factually on available data: however, the USDAneither guarantees nor warrants the standard of the product, and the use of thename by the USDA implies no approval of the product to the exclusion of othersthat may also be suitable.

2 Corresponding author. Phone: 309-688-6447. Fax: 309-681-6685. E-mail address:David.Stevenson@ars.usda.govDepartment of Food Science and Human Nutrition. 2312 Food Sciences Building,Iowa State University, Ames, IA 50011.

doi:l 0.1 094/CCHEM-84-6-0533This article is in the public domain and not copyrightable. It may be freely re-printed with customary crediting of the source. AACC International, Inc., 2007.

rice with particle size <250 .im had lower onset gelatinization tem-perature compared with milled rice with particle size 250-1,400jim. Milled rice particles <500 pin lower peak gelatinizationtemperature than particle size 500-1,400 pm, and particles <355pin lower enthalpy change of gelatinization than particle size355-1,400 p.m. Chen et al (1999) studying different methods ofmilling waxy rice found dry hammer-milled flour with particles of197-215 p.m had the highest pasting temperature and semi-dryground flour with particle sizes of 126-145 p.m had lowest past-ing temperature and setback viscosity.

Levels of 3-glucan in sieved fractions of barley have been studied(Yoon et a! 1995). They found finely ground barley had the high-est 3-glucan content in flour with particle sizes 103-149 p.m andconsiderably lower content for flour particles >250 pin <103Vim.

Isolating starch from milled cereal sieve fractions and charac-terizing the structure and functional starch properties has had littleinvestigation. Vasanthan and Bhatty (1995) purified starch fromwaxy, normal, and high-amylose barley after pin-milling and foundfine sieve screens were successful at separating the small and largegranules exhibited as the bimodal granule size distribution of barleystarch. The most extensive studies of starch properties from sievedfractions have been conducted for corn (Dowd et al 1999) or wheat(Tang et al 2005) kernels. Gelatinization temperature and enthalpychange decreased for starch isolated from corn fibers (pericarp)compared with starch obtained from corn slurry (endosperm) duringcorn wet-milling. No differences in pasting viscosities were ob-served for all starch sources, but corn fiber and washed corn fiberstarch pasted at higher temperature than starch from the corn wet-milling slurry. Apparent amylose content of starch from wheatfractions was similar, but gelatinization temperature varied.

In this study, we investigate the structure and functional prop-erties of starch separated in the oat bran fraction during milling,in which we expect to find different characteristics compared withstarch found in the other mill fractions. Pin-milling oat kernelswithout sieving was included as a control treatment which, basedon lack of reports, we expected would not alter starch structure orfunctional properties. However, alteration in starch structure wasobserved after pin-milling and these findings are also discussed.

MATERIALS AND METHODS

Plant Material, Milling, and Starch IsolationOats (Avena sativa L. cv. Ajay), grown in 2003 at three differ-

ent field sites (replicates) near Aberdeen, ID, were stored one to

Vol. 84, No. 6, 2007 533

two months in a dry environment at 12°C, then ground three timesseparately for each replicate (1 kg of sample/milling) using a pinmill (Alpine, Kolloplex, Augsburg, Germany) for 6 mm, with norepeated milling. Each milled sample was sifted using a Rototapshaker developed by the USDA using sieve squares (40-cm dimen-sion) shaken by a 3/4 horsepower motor from General Electric (FortWayne, IN) for 1 hr (3 x 20 mm, top fraction sieved 2nd and 3rdtime) with a 20-min interruption involving cleaning of sieves andallowing sample to cool in between in a series of sieves consistingof 850-jim mesh (No. 20) on top, followed by a 300-jim mesh(No. 50), 150-pm mesh (No. 100), and a 75-pm mesh (No. 200).Starch was then isolated from sieved fractions, from pin-milled oatswithout sieving, and from intact kernels using the method byKasemsuwan et al (1995) with further modification (Stevenson2003) in which sieve fractions (150 g) or oat kernels (500 g) wereinitially ground in 1.5 L of 0.3% (w/v) sodium metabisulfite using acommercial blender (Waring, New Hartford. CT, high mode used)and filtered through a screen of 106-pm mesh. Filtrate was cen-trifuged at 10,400 x g for 40 mm, then the pellet was washed five toeight times (until supernatant was clear) with 1.5 L of 10% toluene(v/v) in 0.IM sodium chloride to remove protein and lipids, withat least 4 hr between washes. Starch was then washed three timeswith deionized water, then washed twice with ethanol, and re-covered by filtration with Whatman No. 4 filter paper. Purifiedstarch cake was dried in a convection oven at 35°C for 48 hr; finalmoisture of starch was 7-8%. All reagents used were obtainedfrom Fisher Scientific (Pittsburgh, PA).

Properties of Oat Starches13-Glucan content of oat sieve fractions was measured using a

3-glucan diagnostic kit (Megazyme International Ireland Ltd.,Wicklow, Ireland) based on Approved Method 32-23 (AACC Inter-national 2000) in which sieved fractions were hydrolyzed by lich-enase and 3-glucosidase and resulting glucose measured usingglucose oxidase/peroxidase reagent.

Scanning electron microscopy (JOEL model 6400V, Tokyo.Japan) was used to observe oat starches at 1,500x magnification.Oat starch powders were spread on silver tape and mounted on abrass disk, then coated with gold/palladium (60:40) for all threereplicates.

Weight-average molecular weight (Mw) and z-average gyrationradius (Ri) of amylopectin from oat starches were determined usinghigh-performance size-exclusion chromatography equipped withmulti-angle laser-light scattering and refractive index detectors(HPSEC-MALLS-RI). Oat starch (duplicate measurements of eachof three ground samples for each replicate) was prepared as des-cribed by Yoo and Jane (2002a). The HPSEC system consisted ofHP 1050 series isocratic pump (Hewlett Packard, Valley Forge,PA), multiangle laser-light scattering detector (Dawn DSP-F, WyattTechnology, Santa Barbara, CA) and HP 1047A refractive indexdetector. To separate amylopectin from amylose, a Shodex OH pakSB-G guard column and SB-804 and SB-806 analytical columns(JM Science, Grand Island, NY) were used. Operating conditionsand data analysis are described by Yoo and Jane (2002b), exceptthat flow rate was 0.6 mL/min and sample injection concentrationwas 0.5 mg/mL.

Apparent and absolute amylose contents of oat starches weredetermined following the procedure of Lu et al (1996). Analysiswas based on iodine affinities of defatted whole starch andamylopectin fraction using a potentiometric autotitrator (702 SMTitrino, Brinkmann Instrument, Westbury, NY). Starch sampleswere defatted using a 90% dimethyl sulfoxide (DMSO) solution,followed by alcohol precipitation. Determination of amylose con-tent was duplicated for starches of each of three ground samplesof each oat replicate.

Amylopectin was fractionated by the selective precipitation ofamylose with n-butanol as described by Schoch (1942). Amylo-pectin (2 mg/mL) was defatted in 90% DMSO at 100°C for I hr,

followed by stirring for 24 hr, and then debranched using isoamy-lase (EC 3.2.1.68 from Pseudomonas amyloderamosa) (EN 102,Hayashibara Biochemical Laboratories, Okayama, Japan) as des-cribed by Jane and Chen (1992). Branch chain length distributionof amylopectin was determined using an HPAEC system (Dionex-300 and Dionex-GP50 gradient pump, Sunnyvale, CA) equippedwith an amyloglucosidase (EC 3.2.1.3, from Rhizopus mold, A-7255, Sigma Chemical, St. Louis, MO) postcolumn, online reactorand a pulsed amperometric detector (Dionex-ED50) (HPAEC-ENZ-PAD) (Wong and Jane 1997). PA-100 anion exchange analyticalcolumn (250 x 4 mm, Dionex) and a guard column were used forseparating debranched amylopectin samples. Gradient profile ofthe eluents and operating conditions were described previously(McPherson and Jane 1999), except Chromeleon v.6.50 softwarewas used. HPAEC-ENZ-PAD analysis was duplicated for starchesfrom each of three ground samples of each oat replicate.

Thermal properties of oat starches were determined using dif-ferential scanning calorimetry (DSC 2920 modulated, TA Instru-ments, New Castle, DE). Approximately 6 mg of oat product pow-der was weighed in a stainless steel pan, mixed with 18 mg ofdeionized water, and sealed. Sample was allowed to equilibrate for2 hr and scanned at a rate of I O'C/min over a temperature rangeof 0-120°C. An empty pan was used as reference. Rate of starchretrogradation was determined using the same gelatinized samples,stored at 4°C for seven days, and analyzed using DSC as des-cribed previously (White et al 1989). All thermal properties weredetermined in triplicate for starches from each of three groundsamples of each oat replicate.

The pasting properties of oat starches were analyzed using aRapid Visco-Analyser (RVA-4, Foss North America, Eden Prairie,MN) (Jane et al 1999). Oat starch suspension (8%, w/w), induplicate for each of three ground samples of each oat replicate,was prepared by weighing oat starch (2.24 g, dsb) into an RVAcanister and making up the total weight to 28 g with deionizedwater. Oat starch suspension was equilibrated at 30°C for I mm,heated at a rate of 6.0°C/mm to 95°C, maintained at 95°C for 5.5mm, cooled to 50°C at a rate of 6.0°C/mm, and then maintainedat 50°C for S mm. Constant paddle rotating speed (160 rpm) wasused throughout the entire analysis except for a speed of 960 rpmfor the first 10 sec to disperse sample. The standard error of themean (RVU) for 300-850 jim, 150-300 jim, and <150 jim sievefractions was 2.0, 1.1, and 1.5, respectively, for peak viscosity;5.1, 2.9, and 2.4, respectively, for final viscosity; and 5.0, 3.0, and2. 1, respectively, for setback viscosity.

Statistical AnalysisAll statistical significance tests were calculated using SAS meth-

ods (SAS Institute, Cary, NC) and applying the Tukey differencetest (Ramsey and Schafer 1996).

RESULTS AND DISCUSSION

Oat flour fractions separated by sifting were primarily collectedusing No. 50, 100, and 200 sieves. Therefore, only particles withdiameter ranges of 300-850 jim. 150-300 pm, and <150 pm werestudied. The percentage of collected flour fractions based on initialkernel weight for the 300-850 jim. 150-300 pm, and <150 pmsieve fractions collected were 34.1. 26.5. and 24.5%, respectively.Very few particles (<0.1%) passed through the No. 200 sieve. Per-centage dry weight 3-glucan content for the oat sieve fractions300-850 jim. 150-300 jim. and <150 pm were 4.2, 2.3, and 0.8%,respectively (3-glucan content is low because hull fragments areincluded). Therefore the sieves were successful in separating a con-siderably higher proportion of the 0-glucan-rich bran, allowing usto investigate whether starch from the oat bran (endosperm ex-terior) differs in structure and physicochemical properties whencompared with starch from the remaining endosperm.

534 CEREAL CHEMISTRY

Starch Granule MorphologyScanning electron micrography (SEM) of oat starch showed

starch from entire kernels, pin-milled oats, and three sieve frac-tions had granule morphology consisting of larger spherical gran-ules with diameter ranges of 7-9 gm, smaller spherical granuleswith a range of 3.5-5 j.tm, and many highly irregular granules (7:1proportion of irregular to speherical granules). Some granules weredome-shaped, which combined with the highly irregular shapedgranules, is the typical morphology of compound starches. Oatstarch has previously been reported to be a compound (Jane et al1994). Starch granules from all sieve fractions and entire kernelsshowed no difference.

Amylose ContentsIodine affinities for defatted starches and the corresponding ap-

parent amylose contents were significantly different between starchfrom entire oat kernels and pin-milled oats without sieving or the300-850 jim oat sieve fraction (Table 1). Amylose contents thatwe observed are higher than previous reports (Wang and White1994; Tester and Karkalas 1996; Hoover et al 2003). Starch fromthe larger sieve fraction (300-850 pm) had significantly loweriodine affinity of the whole starch and corresponding apparentamylose content compared with the other two sieve fractions. How-ever, no differences in absolute amylose content were observedamong the sieve fractions, although the 300-850 pm fraction hadsignificantly lower absolute amylose content than starch extractedfrom entire kernels. Pin-milled oat starch and its subsequentfractionation by sieving had lower apparent and absolute amylosecontent compared with starch from entire kernels, suggesting thatheat or shear from pin-milling, and not heat produced from sieving,resulted in some amylose degradation. To our knowledge, we can-not find any previous study that extracted starch from any plantmaterial before and after pin-milling, and compared the starch struc-ture and functional properties. An abundance of studies on cerealshave employed pin-milling to grind kernels to produce flour thatis utilized for research, with little regard given to the effects of pin-milling. Our results suggest that alteration of starch characteristicsby pin-milling should be considered.

Amylopectin Molecular Weight (Mn)and Gyration Radius (R7)

Weight-average molecular weight (Mw ), gyration radii (R,), poly-dispersity (PD), and density of amylopectin are shown in Table II.No significant differences between the amylopectin M, Rj., PD,and density were observed among the sieve fractions. Amylopec-tin from entire kernels had significantly higher polydispersity andlower density than starch from the 300-850 pm sieve fraction.Although not significantly different, the higher M. and R observedfor starch isolated from entire kernels without pin-milling suggeststhat pin-milled oat kernels experience some degradation of amylo-pectin molecules. The increased density in starch extracted fromoats that were pin-milled, regardless of sieving, could be due topin-milling disrupting exterior long branch chains of amylopectinmolecules, leaving intact the compact short branch chains.

Amylopectin Branch Chain Length DistributionHPAEC-ENZ-PAD chromatograms of debranched amylopectin

of starch from each sieve fraction are shown in Fig. I. The chro-matograms show starch from entire kernels, pin-milled oats with-out sieving, or the three sieve fractions studied have peak chainlengths at DP 12-13. The chain length distribution of amylopectinis summarized in Table III. Starch extracted from entire kernelsdisplayed a shoulder at DP 18-21 that was largely mirrored forpin-milled oats without sieving and from sieve fractions <300 jim,but was less pronounced for starch from the 300-850 pm sievefraction. A shoulder in amylopectin branch chain length distri-bution at DP 18-21 was reported previously (Hanashiro et al 1996;Jane et al 1999). Average amylopectin branch chain lengths were

similar for starch from all three sieved fractions and pin-milled oatswithout sieving, but significantly shorter than starch from entirekernels. Pin-milled oat starch, regardless of sieving, had amylo-pectin molecules with a higher proportion of short branch chains(DP 3-9) and lower proportion of long branch chains (DP ^! 37).The shorter average amylopectin branch chain lengths observed forpin-milled oat starch compared with starch from entire kernels,corroborates analysis of amylopectin density that pin-milling maydegrade exterior chains that have longer polymerization. Presenceof long side chains in amylopectin have been reported (Reddy et al1993; Takeda et al 1993, 1999; Singh et al 2005), although Math-eson (1996) found amylopectin density can increase by lengtheningci-glucan chains in the external, internal, or core molecular regions.Repeated, lengthy ball-milling of the wheat starch converted someamylopectin to low molecular weight fragments that were derivedfrom shearing glycosidic bonds in the long 132, 133, and 134 internalchains (Morrison and Tester 1994). However, the milling proce-dure intentionally damaged starch and was very severe relative tothe short pin-milling we used.

TABLE IIodine Affinities, Apparent Am y lose, and Absolute Amylose

Contents for Starches Collected from DifferentOat Sieve Fractions

Iodine Affinity Apparent AbsoluteOat WholeAmylopectinAmyloseAmyloseFractionStarchFractionContent (%)h Content (%)r

Entire kernel7.88a1.8 lab39.6a33.6aPin-milled7.33bc1.74b36.8bc30.8ab300-850 .im7.24c1.90ab36.4c29.6b150-300 pm7.50ah2.06a37.7ab30.5ab<150 pin 2.1 Ia37.9ab30.6abProbability'P = 0.003P <0.0001P = 0.003P = 0.02

Values followed by different letters in each column denote differences at the5% level of significance for comparisons between oat starch sieve fractions.Apparent amylose contents were averaged from six analyses for each of threereplicates. Values were calculated from dividing iodine affinity by a factor of0.199.Absolute amylose contents were averaged from six analyses for each of threereplicates. Values were calculated by subtracting iodine affinity for the amylo-pectin fraction from the iodine affinity for the whole starch, divided by a factorof 0. 199.P represents probability of F statistic exceeding expected value for each com-parison between oat starch sieve fractions in each column.

TABLE IIAverage Amylopectin Molecular Weight (My,), Polydispersity,

Gyration Radius (R7), and Density of Oat Starch Collectedfrom Different Sieve Fractions

Oat M. x 108 Polydispersity Density (p)Fraction(gImol)'(M...,/M)R (nm)C(g/molInm3)1

Entire kernel8.372.60a43110.2bPin-milled4.322.15ab30215.7ab300-850 pm4.171.42b28717.2a150-300 pm5.161.66ab31915.3ab<150pin5.861.71ab32815.7abProbability-P = 0.47P = 0.01P = 0.20P = 0.04

Data were obtained from six injections each of all three replicates.All samples were dissolved in 90% DMSO solution and precipitated with 5vol of ethanol. Freshly prepared starch aqueous solution (100 j.tL, 0.8 mg/mL)was injected to HPSEC system.

C Values followed by different letters in each column denote differences at the5% level of significance for comparisons between the oat starch sieve fractions.Average amylopectin molecular weight.Gyration radius.Density = MJR,3. Values for density may not correspond directly to data intable due to rounding of M and R values.P represents probability of F statistic exceeding expected value for each com-parison between oat starch sieve fractions in each column.

Vol. 84, No. 6, 2007 535

Thermal Properties of StarchThermal properties of native oat starches are shown in Table IV.

Two thermal transitions were observed for all starches, correspond-ing to starch gelatinization and melting of amylose-lipid complex.

76- entire kernel

II Ili4

ItI

2I1111111.111

II11111111

IIIIIIIIIJIIIIIIIII!liiIllhllllllplI!lHhInn..,..._......,.3132333435363738393103

DP

300-850 tm

3132333435363738393103

DP

1d11IIIIIIIIhiIIIlIluJlllllIIIl!i:IHii.n,.._,_............

Starch from entire oat kernels had significantly lower peak gela-tinization temperature and melting of amylose-lipid complex, thanstarch from any of the sieve fractions, but not for pin-milledstarch without sieving. This result is surprising, as pin-milling and

ii pin-milled

3132333435363738393103DP

7 150-300 p.tm

.13;343536373

. llllllilllfflIllllllhiunrn,,.._

- 8393103np

<150tm

3132333435363738393103

DP

'1 JIIIllUujilllfflHUfflfflilJfflhIH

Fig. 1. Relative peak area distributions of amylopectins obtained from starch of entire kernel or from the sieve fractions 300-850 him, 150-300 rim, and<ISO tim. HPAEC-ENZ-PAD analysis was used. Error bars represent standard error of the mean for each individual DP from two analyses of threereplicates. DP, degree of polymerization.

TABLE IllBranch Chain Length (CL) Distributions of Oat Starch Sieve Fraction Amylopectins'

OatFractionI

Entire kernel12.3Pin-milled12.0300-850 tim12.4150-300ttm12.4<ISO m12.6Probability'

Peak DPAverage

tICL

49.224.9a

47.523.Ob

47.923.3b

48.923.2b

48.723.3bP=0.02

DP 3-5DP 6-9

0.32b4.7b0.50a6.7a0.58a6.8a0.36ab5.9ab0.44ab5.2ab

P=0.04P=O.02

Percent Distribution

DP 6-12DP 13-24 DP 25-36DP^!37

21.1

44.3b14.5a19.8a

23.6

44.6ab14.Oab17. lb

23.9

45.3a13.2b17.1b

23.146.5a13.3b16.8b

22.2

46.9a13.5ab17.Ob

P = 0.07

P=0.02P=0.002P=0.05

HighestDetectable DP

102100104104103

a Values followed by different letters in each column denote differences at the 5% level of significance for comparisons between oat starch sieve fractions.b Grouping of degree of polymerization (DP) numbers according to Hanashiro et al (1996).

P represents probability of F statistic exceeding expected value for each comparison between oat starch sieve fractions in each column.

TABLE IVThermal Properties of Native Oat Starch Sieve Fractions

Starch Gelatinization

T. (°C) T (°C) All (Jig)

59.9b 64.3c 10.160.2ab 64.8bc 11.460.9a 66.5a 10.861.1a 66.3ab 9.561.3a 66.3ab 10.0P=0.0l P=0.005 P=0.08

Amylose-Lipid Transition

T0 (°C) T (°C) All (Jig)

93.1 100.3b 2.12ab93.7 101.5ab 2.69a95.4 102.5a 1.49b95.7 102.8a 1.42b96.9 102.6a 1.30b

P=O.06 P=0.05 P=0.007

Oat Fraction

Entire kernelPin-milled300-850 tim150-300 l.lm<150 timProbability"a Starch samples (2.0 mg, dsb) and deionized water (6.0 mg) were used for analysis. T0, T, and Al-I are onset and peak gelatinization temperature, and enthalpy

change of gelatinization, respectively.b Values were calculated from nine analyses for each of three replicates.

Values followed by different letters in each column denote differences at the 5% level of significance for comparisons between oat starch sieve fractions.d P represents probability of F statistic exceeding expected value for each comparison between oat starch sieve fractions in each column.

536 CEREAL CHEMISTRY

051015202530

250

200

150>

100

50

100

80

60

40

20

0

St

St

ESt

sieving would be expected to increase levels of damaged starch,which is known to lower onset gelatinization temperature (Swans-ton et al 1994; Mousia et al 2004). Additionally, Kerr et al (2000)showed that finer particle sizes of cowpea flour absorbed morewater, resulting in lower onset gelatinization temperature. Resultssimilar to ours for gelatinization temperatures and enthalpy changefor oat starch have been reported (Paton 1987), but other studieshave found lower gelatinization temperatures and higher enthalpychange (Tester and Karkalas 1996; Hoover et al 2003). Less in-herent crystalline order within granules located at the peripheralendosperm region (comparable to the 300-850 j.im sieve fraction)has been reported in starch with lower enthalpy change (Dowd etal 1999), but we did not observe this in our study.

Although we speculate that changes in starch properties of oatstarch after pin-milling are most likely due to shredding of amyloseand amylopectin structure, it is also possible that localized regionsof high temperature and moisture were present in which starchgelatinized and retrograded.

Retrograded oat starches are shown in Table V. No differenceswere observed among oat starches for the temperature to melt theretrograded starch, but starch from entire oat kernels had loweronset temperature for the retrogradation of amylose-lipid complexthan starch pin-milled without sieving or from the three sievefractions studied.

Pasting Properties of StarchPasting properties of the oat starches showed significant differ-

ences (Table VI, Fig. 2). Starch isolated from the 150-300 pmsieve fraction had significantly lower peak, final, and setback vis-cosity (P <0.0001, P < 0.0001. and P = 0.003, respectively) thanstarch isolated from the 300-850 Vtm and <150 pin frac-tions, and from pin-milling without sieving or entire oat kernels.Trough of starch isolated from the 150-300 pin fraction wasalso significantly lower than trough of starch isolated from the<150 .tm sieve fraction (P <0.0001). Pasting properties of starchfrom entire oat kernels were not significantly different than that of

starch from 300-850 pin <150 lim sieve fractions, except fortrough viscosity.

The highly significant differences in oat starch pasting prop-erties among the different sieve fractions of pin-milled Oat kernelsare surprising and difficult to explain based on current knowledge.We can be confident these differences are real due to the thoroughexperimental design in which oats were grown at three differentfield sites in Idaho and milled three separate times, with each sam-ple measured in duplicate, and all samples showed similar trends.Myrback and Gjørling (1945) calculated that only 0.10% of starchmolecule bonds need to rupture to depress slurry viscosity byhalf, so it is possible that pasting differences observed in starchfrom the sieve fractions could be due to minor modifications instarch structure. However, no differences were observed among thesieve fractions for DSC analysis, which is sensitive to detectingmolecular changes.

The influence of milling on pasting profiles obtained from aRapid Visco-Anlyser has been investigated for maize and wheatproducts (Becker et al 2001). They found milled maize and wheat

-s-- entire kernel ---pin-milled -k-- 300.850cm150-100 urn-*-- < 150 urn

Time (mm)

Fig. 2.2. RVA pasting profiles of starch from entire oat kernels or from 300-850 M°. 150-300 m, and <150Mm sieve fractions (8.0% dsb, w/w).

TABLE VThermal Properties of Retrograded Oat Starch Sieve Fractions

Starch Gelatinization Amylose-Lipid TransitionOat Fraction T. (°C)

T (°C)AH (Jig) T. (°C)

T (°C)AH(J/g)

Entire kernel 41.7 49.6 1.49 94.9b 103.6

193Pin-milled 41.0 50.5

1.72 97.4a 104.! 1.63300-850 pin 40.2

51.3

1.92

97.7a 104.3 1.74150-300 gm 41.8 51.5 1.65 98.8a 104.5 1.64<150 pin 41.9

51.9

1.67

98.5a 104.5 1.50Probabjlityd P=0.39

P= 0. 18 P = 0.64P=0.002

P = 0.11P=0.61

a Same starch samples after gelatinization (see Table IV) were left for 7 days at 4C and rescanned using DSC; T0. T. and AH are onset and peak thermaltransition temperatures, and enthalpy change of thermal transition, respectively.Values were calculated from nine analyses for each of three replicates.

C Values followed by different letters in each column denote differences at the 5% level of significance for comparisons between oat starch sieve fractions.d P represents probability of F statistic exceeding expected value for each comparison between oat starch sieve fractions in each column.

TABLE VIPasting Properties of Oat Starch Sieve Fractions Measured by Rapid Visco.Analyser

Oat FractionPeak ViscosityTroughBreakdownFinal ViscositySetbackPasting Temp (°C)Entire kernel 110.9b 76.5c 34.5a 176 lb 99.7a 93.5Pin-milled 136.1a 110.9a 25.2ab 213.1a 102.1a 92.7300-850 m 113.5b 84.3bc 29.2ab 192.9b 108.6a 93.8150-300 gm 101.6c 79.8c 21.8ab 153.8c 74.1b 94.4<150 pin 115.5b 94.1b 21.4b 191.5b 97.4a 94.1Probability" P<0.000lP<0.000lP0.05 P<0.000lPO.003 PrrO.1la 8% (w/w) oat starch sieve fraction suspension measured in duplicate for all three replicates.

Viscosity measured in Rapid Visco-Analyser units where I RVU = 12 centipoise.C Values followed by different letters in each column denote differences at the 5% level of significance for comparisons between oat starch sieve fractions.

P represents probability of F statistic exceeding expected value for each comparison between oat starch sieve fractions in each column.

Vol. 84, No. 6, 2007 537

-I-

products with large particle size (>250 pm) had similar or consid-erably higher peak viscosity depending on whether the productwas in grit or pellet form, but all had higher setback and finalviscosity than products with smaller particle sizes. There was apositive relationship between particle size and final viscosity or set-back that is in contrast with our findings where starch from smalland large particle size sieve fractions had higher peak, final, andsetback viscosities than starch from intermediate particle size sievefractions. Differences observed between our study and other resultscould be attributed to setback viscosities including all compoundsin corn and wheat kernels that may influence pasting characteris-tics. Additionally, there was no separation of particles >250 jim,making it difficult to directly compare materials with our sieveseparations.

In the only previous study to isolate starch and measure pastingproperties from milled cereal fractions, Dowd et al (1999), unlikeour findings, found no significant differences between wet-milledcorn fractions for starch peak, final, and setback viscosities, butfound significant differences in pasting temperature. They attributedifferences in pasting temperature to varying degrees of starchdamage during milling. Because the wet-milled corn fractionswith varying starch damage had no differences in paste viscosity,it is unlikely that differences we observed in paste viscosities ofoat starches from the sieve fractions are due to starch damage.Starch from pin-milling without sieving had significantly higherpeak, trough, and final viscosity than starch from any of the sievefractions, which suggests that heat produced during sieving mayinfluence pasting characteristics.

CONCLUSIONS

Starch isolated from oat kernels cultivated at three locationsthat were pin-milled and subsequently sieved showed no differ-ence in starch structure and thermal properties among sieve frac-tions and consistent starch properties from all three locations.However, lower paste viscosity was observed for starch isolatedfrom ground oat particles 150-300 jim compared with starch iso-lated from particles 300-850 jim or <150 jim. This finding illus-trates that oat starch extracted during concentration of 3-glucanproducts may have different industrial applications compared withoat starch obtained from entire kernels, depending on sieve meshdimensions used during refining. Pin-milling of oat kernels causedno visual damage of starch granules, but lowered starch apparentamylose content and shortened amylopectin branch chain lengthscompared with starch isolated from oat kernels without pin-milling. Therefore the effects of milling on starch characteristicsshould be considered when extracting starch.

ACKNOWLEDGMENTS

We wish to thank Mike Bonman and Dave Burrup for supplying oats.We also thank Arthur Thompson for scanning electron microscopyassistance; Jane Schupp for 3-glucan analysis; and Bill Deadmond forassistance with Rototap procedures.

LITERATURE CITED

Andersson, L.. Fredriksson, H., Oscarsson Bergh, M., Andersson, R., andAman, P. 1999. Characterisation of starch from inner and peripheralparts of normal and waxy barley kernels. J. Cereal Sci. 30:165-171.

Becker, A., Hill, S. E., and Mitchell, J. R. 2001. Milling-A furtherparameter affecting the Rapid Visco Analyser (RVA) profile. CerealChem. 78:166-172.

Carrière, C. J., and lnglett, G. E. 2000. Effect of processing conditions onthe viscoelastic behavior of Nu-TrimX: An oat-based 3-glucan-richhydrocolloidal extractive. J. Text. Stud. 31:123-140.

Chen, J. J., Lu, S., and Lii, C. Y. 1999. Effects of milling on the physi-cochemical characteristics of waxy rice in Taiwan. Cereal Chem.76:796-799.

Dowd, M. K., Radosavljevic, M., and Jane, J. 1999. Characterization of

538 CEREAL CHEMISTRY

starch recovered from wet-milled corn fiber. Cereal Chem. 76:3-5.Fulcher, R. G. 1986. Morphological and chemical organization of the oat

kernel. Pages 47-74 in: Oats: Chemistry and Technology. F. H. Webster,ed. AACC International: St. Paul, MN.

Hanashiro, I., Abe, J., and Hizukuri, S. 1996. A periodic distribution ofthe chain length of amylopectin revealed by high-performance anion-exchange chromatography. Carbohydr. Res. 283:151-159.

Hoover. R., Smith, C., Zhou, Y., and Ratnayake. R. M. W. S. 2003. Phy-sicochemical properties of Canadian oat starches. Carbohydr. Polym.52:253-261.

Jane, J., and Chen, J. F. 1992. Effect of amylose molecular size andamylopectin branch chain length on paste properties of starch. CerealChem. 69:60-65.

Jane, J., Kasemsuwan, T., Leas, S., Zobel, H., and Robyt. J. F. 1994.Anthology of starch granule morphology by scanning electron micro-scopy. Starch 46:121-129.

Jane, J., Chen, Y. Y., Lee, L. F., McPherson, A. E., Wong, K. S., Rado-savljevic, M., and Kasemsuwan. T. 1999. Effects of amylopectin branchchain length and amylose content on the gelatinization and pastingproperties of starch. Cereal Chem. 76:629-637.

Kasemsuwan, T., Jane, J., Schnable, P., Stinar. P., and Robertson. D.1995. Characterization of the dominant mutant amylose-extender (Ael-5180) maize starch. Cereal Chem. 71:457-464.

Kerr, W. L., Ward, C. D. W., McWatters, K. H., and ResurrecciOn, A. V.A. 2000. Effect of milling and particle size on functionality and physi-cochemical properties of cowpea flour. Cereal Chem. 77:213-219.

Kurimoto. Y., and Shelton. D. R. 1988. The effect of flour particle size onbaking quality and other flour attributes. Cereal Foods World 33:429-433.

Lu, T., Jane, J., Keeling, P. L., and Singletary, G. W. 1996. Maize starchfine structures affected by ear developmental temperature. Carbohydr.Res. 282:157-170.

Marshall, W. E. 1992. Effect of degree of milling of brown rice and particlesize of milled rice on starch gelatinization. Cereal Chem. 69:632-636.

Matheson, N. K. 1996. The chemical structure of amylose and amylo-pectin fractions of starch from tobacco leaves during development anddiurnally-nocturnally. Carbohydr. Res. 282:247-262.

McPherson, A. E., and Jane, J. 1999. Comparison of waxy potato withother root and tuber starches. Carbohydr. Polym. 40:57-70.

Morrison, W. R., and Tester. R. F. 1994. Properties of damaged starchgranules. IV. Composition of ball-milled wheat starches and of fractionsobtained on hydration. J. Cereal Sci. 20:69-77.

Mousia, Z., Edherly, S., Pandiella, S. S., and Webb, C. 2004. Effect ofwheat pearling on flour quality. Food Res. Int. 37:449-459.

Myrback, K., and Gjørling, L. G. 1945. Starch cleavage by Bacillus,nacerans. Arkiv Kemi Mineral. Geol. A20: No. 5.

Oscarsson, M., Parkkonen, K., Autio, K., and Aman, P. 1997. Compositionand microstructure of waxy, normal and high amylose barley samples.J. Cereal Sci. 26:259-264.

Paton, D. 1987. Differential scanning calorimetry of oat starch pastes.Cereal Chem. 64:394-399.

Pick, M. E., Hawrysh, Z. J., Gee, M. I., Toth, E., Garg, M. L.. and Hardin,R. T. 1996. Oat bran concentrate bread products improve long-termcontrol of diabetes: A pilot study. J. Am. Diet. Assoc. 96:1254-1261.

Pomeroy, S., Tupper, R., Cehun-Aders, M.. and Nestel, P. 200 1. Oat betaglucan lowers total and LDL-cholesterol. Aust. J. Nutr. Diet. 58:51-55.

Ramsey, F. L., and Schafer, D. W. 1996. Page 154 in: The Statistical Sleuth:A Course in Methods of Data Analysis. Duxbury Press: Belmont, CA.

Reddy, K. R., Ali, S. Z., and Bhattacharya, K. R. 1993. The fine structureof rice-starch amylopectin and its relation to the texture of cooked rice.Carbohydr. Polym. 22:267-275.

Scanlon, M. G., Dexter, J. E., and Biliaderis, C. G. 1988. Particle-sizerelated physical properties of flour produced by smooth roll reductionof hard red spring wheat farina. Cereal Chem. 65:486-492.

Schoch, T. J. 1942. Fractionation of starch by selective precipitation withbutanol. J. Chem. Soc. 64:2957-2961.

Singh, N., Inouchi, N., and Nishinari. K. 2005. Morphological, structural,thermal, and rheological characteristics of starches separated fromapples of different cultivars. J. Agric. Food Chem. 53:10193-10199.

Stevenson, D. G. 2003. Role of starch structure in texture of wintersquash (Cucurbita maxima D.) fruit and starch functional properties.Pages 173-174. PhD thesis. Iowa State University: Ames, IA.

Swanston, J. S., Rao, N. S., Subramanian, V., and Taylor, K. 1994. Theinfluence of some aspects of grain quality on malting potential in sor-ghum. J. Cereal Sci. 19:91-97.

Takeda, C., Takeda, Y., and Hizukuri, S. 1993. Structure of the amylopec-tin fraction of amylomaize. Carbohydr. Res. 246:273-28 I.

Takeda, Y., Takeda, C., Mizukami, H., and Hanashiro, 1. 1999. Structuresof large, medium, and small starch granules of barley grain. Carbohydr.Polym. 38:109-114.

Tang, H., Mitsunaga. T., and Kawamura, Y. 2005. Functionality of starchgranules in milling fractions of normal wheat grain. Carbohydr. Polym.59:11-17.

Tester, R. F., and Karkalas, J. 1996. Swelling and gelatinization of oatstarches. Cereal Chem. 73:271-277.

Vasanthan, T., and Bhatty, R. S. 1995. Starch purification after pin millingand air classification of waxy, normal, and high amylose barleys.Cereal Chem. 72:379-384.

Wang, L. Z., and White. P. J. 1994. Structure and physicochemical prop-erties of starches from oats with different lipid contents. Cereal Chem.

71:443-450.White, P. J., Abbas, I. R., and Johnson, L. A. 1989. Freeze-thaw stability

and refrigerated-storage retrogradation of starches. Starch 41:176-180.Wong, K. S.. and Jane, J. 1997. Quantitative analysis of debranched amy-

lopectin by HPAEC-PAD with a post-column enzyme reactor. J. LiquidChromatogr. 20:297-310.

Yoo, S., and Jane, J. 2002a. Structural and physical characteristics ofwaxy and other wheat starches. Carbohydr. Polym. 49:297-305.

Yoo. S.. and Jane, J. 2002b. Molecular weights and gyration radii ofamylopectins determined by high-performance size-exclusion chroma-tography equipped with multi-angle laser-light scattering and refractiveindex detectors. Carbohydr. Polym. 49:307-314.

Yoon, S. H., Berglund, P. T., and Fastnaught, C. E. 1995. Evaluation ofselected barley cultivars and their fractions for f3-glucan enrichmentand viscosity. Cereal Chem. 72:187-190.

[Received June 16, 2006. Accepted April 11, 2007.]

Vol. 84, No. 6, 2007 539