Properties of Enzyme Modified Corn, Rice and Tapioca Starches

Food Research International 42 (2009) 1426–1433

Contents lists available at ScienceDirect

Food Research International

journal homepage: www.elsevier.com/locate/foodres

Properties of enzyme modiﬁed corn, rice and tapioca starches

Sakina Khatoon a, Y.N. Sreerama b, D. Raghavendra a, Suvendu Bhattacharya c, K.K. Bhat d,* a Department of Lipid Science and Traditional Foods, Central Food Technological Research Institute, (Council of Scientific and Industrial Research), Mysore 570 020, India b Department of Grain Science and Technology, Central Food Technological Research Institute, (Council of Scientific and Industrial Research), Mysore 570 020, India c Department of Food Engineering, Central Food Technological Research Institute, (Council of Scientific and Industrial Research), Mysore 570 020, India d Department of Sensory Science, Central Food Technological Research Institute, (Council of Scientific and Industrial Research), Mysore 570 020, India article info abstract

Article history: Corn, rice and tapioca starches were partially hydrolyzed by treating the starch dispersions with heat sta- Received 24 April 2009 ble a-amylase. Dextrose equivalent (DE) of 8–12 was achieved by hydrolyzing the starch samples (10– Accepted 26 July 2009 20% w/v) for 30 min at 90 ± 2 °C. Scanning electron micrographs showed that starch granules had broken down to smaller particles. High performance liquid chromatography with refractive index detection indicated that oligosaccharides with broad molecular weight distributions are present in the reaction prod- Keywords: ucts. Hydrolyzed starch dispersions were analyzed for their rheological properties. The storage modulus Hydrolyzed starches values (G0) for 20% solid containing slurries were 7373 and 1470 Pa for untreated and enzyme treated a-amylase samples, respectively, indicating a marked decrease in solid properties due to enzyme action. The com- Enzyme modified starches Corn starch plex viscosities (g ) for native corn starch and hydrolyzed corn starch were 8243 and 1637 Pas, respec- Rice starch tively, which indicate that the enzyme treatment decreases the overall resistance of the sample to flow Tapioca starch such that the product can spread easily. Further 13C CP/MAS NMR and FTIR studies revealed the loss of Dextrose equivalent ordered structures in the enzyme modified starches. Free flowing fat substitute in the form of fine powder Spray drying was prepared by spray drying the hydrolyzed starch slurry. Ó 2009 Elsevier Ltd. All rights reserved.

1. Introduction composition (Marchal, Beeftink, & Tramper, 1999). Industrially produced maltodextrins normally consist of a broad distribution Starch degrading enzymes have been used to modify the phys- of both linear and branched molecules. The objective of the present ico-chemical properties of polysaccharides to achieve the desired work is to study the hydrolysis of starch with enzymes at different functional properties. The most important tool in providing a sac- conditions and characterize the products in terms of structural, charide with a specific composition is the use of starch hydrolyzing functional and rheological aspects. enzymes. Starches from various botanical origins differ slightly in amylose content, chain length distribution, molecular weight and the number of chain per cluster. It has been reported that wheat 2. Materials and methods and corn maltodextrins with dextrose equivalent 2–3 could be prepared by heterogeneous bacterial a-amylase digestion (Mc Pher- 2.1. Materials son & Seib, 1997). A fat mimetic maltodextrin was also produced by heterogeneous hydrolysis of potato starch by a-amylase (Rich- Corn, tapioca and rice starches were procured from the local ter, Schierbaum, Augustat, & Knoch, 1976). The three most widely market of Mysore, India. Thermostable a-amylase from Bacillus used a-amylases are all isolated from Bacillus i.e. B. amyloliquefac- licheniformis was procured from Hi Media Laboratories Pvt. Ltd., iens, B licheniformis, and B. stearothermophilus, and differ with re- Mumbai, India. All other chemicals were procured locally and they spect to the specificity by which they hydrolyze the a(1 ? 4) were of analytical grade. linkages in starch. Their temperature optimum is in the range of 60–90 °C and the pH optimum 6–7. The starch hydrolysis products are industrially produced by enzyme reactions from a dissolved 2.2. a-amylase assay solution of starch (upto 40% w/w). However, the concentration at which the hydrolysis reaction takes place influence the saccharide Alpha amylase activity was measured according to the method of Bernfeld (1955). One unit of activity is defined as the amount of enzyme that catalyzed the liberation of reducing sugar equivalent * Corresponding author. Tel.: +91 821 2515842; fax: +91 821 2517233. to one micro mole of maltose or glucose per minutes under assay E-mail address: [email protected] (K.K. Bhat). conditions.

2.3. Enzyme hydrolysis and fructose were eluted together due to their close retention times. Corn, tapioca and rice starch dispersions of 10%, 15% and 20% solids (w/v; dry solid basis) were gelatinized in a steam jacketed 2.8. Determination of rheological parameters kettle separately at 95 ± 2 °C for 15 min with moderate and continuous stirring. Gelatinized starch dispersions were partially hydro- A controlled stress rheometer (Model # RT 10, Haake, Kar- lyzed with heat stable a-amylase (1.33 IU/g starch) dissolved in esruhe, Germany) was used to determine the dynamic oscillation 200 ppm calcium chloride solution, at 90 ± 2 °C for 30 min. The properties and flow behavior of samples (starch slurries/disper- reaction was terminated by adjusting the pH to 3.0 with 1 N HCl. sions). The parallel plate attachment was employed to determine The hydrolyzed starch samples were neutralized to pH 6.5–7 with the rheological behavior of these starch slurries. Their concentra- 1 N NaOH at 60 °C according to the method of Mc Pherson & Seib, tions are 10%, 15% or 20% (dry solid basis). They were then mixed 1997. manually in a gentle manner by a glass rod for 5 min to have uniform sample. Latter, they were carefully loaded between 30 mm 2.4. Dextrose equivalent (DE) parallel plates with a gap of 1 mm, and excess sample was trimmed off. The starch samples used for rheological measurements are The degree of hydrolysis was measured as an increase in the unmodified and enzyme modified corn starch slurries. A thin layer content of reducing sugars. The results were compared to a calibra- of paraffin oil was gently applied to the edge of the exposed sample tion curve based on standard glucose (Miller, 1959). DE represents to prevent loss of moisture. For dynamic oscillatory tests, the stor- the percentage of hydrolysis of the glycosidic linkages present in age modulus (G0), and complex viscosity (g) were determined starch. It was calculated using the following equation during a frequency sweep varying from 0.01 to 45 Hz at a constant stress of 25 Pa after determining the linear viscoelastic range Reducing sugar; expressed as glucose employing stress sweep tests; all experiments were performed in DE ¼ 100 Total carbohydrate the linear viscoelastic zone. The G0, and g values at an angular velocity of 6.28 radian (equivalent to a frequency of 1 Hz) were chosen for comparison of results. All tests, in triplicate, were con- 2.5. Spray drying of partially hydrolyzed starches ducted at a temperature of 25.0 ± 0.1 °C by employing a circulatory water bath, supplied by the rheometer manufacturer. Modified starch dispersions were dried in a spray drier (Model To determine the flow properties, all measurements were con- No. BE 1216, Bowen Engineering, NJ, UAS). The dispersion was suit- ducted at 25 ± 0.1 °C on duplicate samples. The flow curves were ably diluted to ensure proper atomization. Depending on the starch generated using a stress sweep from 50 to 500 Pa to generate 40 concentration, two to three fold dilution of dispersion was neces- shear-rate/shear-stress data points. The yield stress was noted sary for easy flow and uniform spray through the atomizer. Inlet from the flow curve as the stress to initiate flow. Apparent viscosity and outlet temperatures were set at 165–170 °C and 100–105 °C, was reported corresponding to a shear rate of 100 s1. respectively. The resulting spray dried powder was white in colour and had a moisture content of about 1.5%. The powder was free 2.9. Structural studies by scanning electron microscope (SEM), solid- flowing and hygroscopic. state NMR and FTIR

2.6. Estimation of sugars Spray dried partially hydrolyzed starch samples and their corresponding unhydrolysed samples were scanned at 500–2000 Total carbohydrate content of the starches and reducing sugar magnification to observe the effect of cleavage of glycosidic bonds content of partially hydrolyzed starch samples, drawn at 5 min on the ultrastructure of starch particles due to enzyme treatment time intervals during enzymatic treatment, were assayed by phe- according to the method of French, 1984. Starch particles were nol-sulphuric acid method (Dubois, Gilles, Hamilton, Rebers, & coated with gold in a sputter coater and scanned in a scanning Smith, 1956) and di-nitro salicylic acid method (Miller, 1959), electron microscope (Model#435 VP, Leo Electron Microscopy, respectively. The glucose released from partially hydrolyzed Cambridge, UK). Representative photomicrographs are presented starches was quantified by comparing with calibration curve of for comparison among samples. Bruker made, 1997 DSX- standard glucose. Moisture and lipid contents of starch samples 300 MHz equipment was used for solid-state NMR (13 C Cross were determined by the AOAC methods (1975). polarization Magic Angle spinning experiment) according to the method of Gidley & Bociek, 1985. FTIR spectra of unmodified and enzyme modified starches were obtained on FTIR spectrophotom- 2.7. HPLC analysis eter (Model # Nicolet 5700, Thermo Electron Corporation, USA). The dry starch powders were used to record the spectra in the The free sugars from partially hydrolyzed starches were ex- transmission mode from 4000 to 400 cm1 using deutrated triglyc- tracted with 70% aqueous ethanol, concentrated by rotary evapora- erine sulphate detector according to the method of Dupuy, Wojcie- tion and quantified by HPLC (Model LC-10ATVP, Shimadzu, Japan) chowski, Ta, Huvenne, & Legrand (1997). For each sample three according to the method of Mc Ginnis & Fang, 1980. The column spectra were recorded and computed single-beam spectra at used was aminopropyl (Phenomenex, CA, USA). Isocratic elution 4cm1 resolution before Fourier transform. was performed using acetonitrile-water (70:30, v/v) solvent system at a flow rate of 1 ml/min. The solvent was delivered using LC-10ATVP pump (Shimadzu, Japan). Refractive index detector 3. Results and discussion (RID-10A, Shimadzu, Japan) was used to detect the sugars. Data signals were acquired and processed on a PC running the Class 3.1. Enzymatic hydrolysis of starches VP software (Shimadzu, Japan). Sugars were extrapolated from pure glucose, maltose and fructose standard curves. Twenty micro- Moisture content of starches used for enzymatic hydrolysis liter injections were made in each run and peak areas were used ranged from 9.9% to 11.6% (d.b.), total carbohydrate from 77.6% for all calculations. Under the experimental conditions, glucose to 83.2% and lipid content varied from 0.1% to 0.5%. A progres- 1428 S. Khatoon et al. / Food Research International 42 (2009) 1426–1433 sive increase in glucose content was observed with an increase Enzymatic hydrolysis was carried out using a-amylase from B. in temperature upto 95 °C beyond which there was a drop of licheniformis by taking 10%, 15% and 20% (w/v) starch disper- glucose content. Hence, a temperature of 95 °C has been selected sions. The hydrolysis of glycosidic bonds in gelatinized starch as the optimum temperature for the enzymatic hydrolysis. by amylases would result in the reduction of viscosity. This

10 A

Dextrose equivalent (DE) 2

0 5101520253040 Time (min.)

10 B

2 Dextrose equivalent (DE)

0 5 101520253040 Time (min.)

14 C

4 Dextrose equivalent (DE) 2

0 5 101520253040 Time (min.)

Fig. 1. Dextrose equivalent (DE) of enzyme modified corn (A), tapioca (B) and rice (C) starches. Starch concentrations used were 10% ( ), 15% (N) and 20% (O). Results are average of three independent determinations. S. Khatoon et al. / Food Research International 42 (2009) 1426–1433 1429 low viscous free flowing hydrolyzed starch may find application starch at 10% concentration produced high DE value (>14.0). Mod- as a fat substitute. ified starches with very low or very high DE values are not suitable for substituting fat (Byars, 2002; Dupuy et al., 1997). 3.2. Dextrose equivalent of modified starches 3.3. Spray drying of modified starches Starch hydrolysis products are normally characterized by their Dextrose Equivalent (DE) value, which is related to the degree of The slurry of modified starches was too viscous to be sprayed hydrolysis. Pure glucose has a DE of 100, pure maltose has a DE through the automated spray dryer. Further dilution was done till of about 50 and starch has a DE of effectively zero. DE indicates an uniform spray could be achieved. Another problem associated the extent to which the starch has been cleaved. DE values ob- with spray drying was the deposition of dried powder on the walls tained for 10%, 15% and 20% corn, tapioca and rice starches at an of the spray dryer. Trials were carried out to arrive at the appropri- intervals of 5 min reaction time during the course of enzymatic ate feed rate, inlet air temperature and outlet temperature. It was hydrolysis are presented in Fig. 1. Corn starch at 10%, 15% and necessary to maintain a low feed rate of 4–5 l/h and the inlet air 20% dispersion produced DE values of 8.2, 6.3 and 5.3, respectively, temperature and outlet temperature were to be adjusted in the during the initial 5 min reaction time. The DE values of 10% and range of 165–170 °C and 100–105 °C, respectively. Since the pow- 15% starch dispersions increased progressively with reaction time der is hygroscopic, but free flowing, it was collected in a dehumid- upto 30 min and reached 11.4 and 8.6 for 10% and 15% starch, ified chamber and packed immediately. Spray dried powder was respectively. Thereafter, DE decreased to 10.8 and 4.9 for 10% white in colour and had a moisture content of 1.5%. and 15% starch, respectively, at 40 min reaction time. However, in case of 20% dispersion a marginal increase of DE from 5.3 to 3.4. Rheology of modified starch gels/dispersions 6.4 was observed during 5–30 min reaction time. At 40 min 20% dispersion produced a DE value of 6.4. Spray dried modified starches, after making the slurry of re- A progressive increase in DE values of tapioca starch was also quired solid concentration, were subjected to rheological charac- noticed in all the three concentrations of starch dispersions. Max- terization to understand the changes occurring due to processing. imum DE values of 10.2, 8.1 and 5.2 were obtained for 10%, 15% The parameters determined were storage modulus (G0) that reflect and 20% tapioca starch dispersions, respectively. Similar to corn its solid characteristics whereas complex viscosity (g) is an indi- starch, a decrease in DE values were noticed in all the concentra- cation of the overall rheological status of the system towards tions of tapioca starch at 40 min reaction time. Hydrolysis of 10% deformation and flow. rice starch resulted in increase in DE from 11.6 to 14.5 during 5– The corn starch samples showed a marked decrease in G0 and g 30 min reaction time. However, increase in DE from 8.1 to 9.4 values indicating the significant thinning effect due to enzyme was observed at 15% concentration during 5–30 min. In case of treatment (Table 1). For example, G0 values of 20% slurry were 20% rice starch dispersion, DE remained nearly constant in the 7373 and 1470 Pa for untreated and enzyme treated samples, range of 8.1–8.4. A decrease in DE at 40 min reaction time was no- respectively. This indicates a marked decrease in solid properties ticed in all the concentrations of rice starch, which is similar to due to enzyme action, and the rate of decrease is higher for disper- corn and tapioca starches. sions containing higher solids. On the other hand, the complex vis- Hydrolysis of corn and tapioca starches at 10% and 15% concen- cosity values for native and hydrolyzed corn starches were 8243 trations resulted in DE values in the range of 8–11.4 at 30 min reac- and 1637 Pas, respectively, which indicate that the enzyme treat- tion time. Similarly, rice starch at 15% concentration also resulted ment decreases the overall resistance of the sample to flow and in DE of 9.4. These values are comparable to starch-based commer- spread. A remarkable decrease in storage modulus (171 ± 21 Pa) cial fat substitutes, which are in the range of 8–12 (Byars, 2002; and complex viscosity (191 ± 27 Pas) was noted for enzyme Dupuy et al., 1997). However, corn, tapioca and rice starches at modified corn starch gel, when it was spray dried. Further, this 20% concentrations produced low DE values (<8.0), whereas, rice spray dried and enzyme modified starch was heated at 65 °C for

Table 1 Rheological parameters for control and modiﬁed corn starches before and after enzyme treatment.a

Starch Storage modulus (G0) (Pa) Complex viscosity (g) (Pas) Concentration (% db) 15 20 15 20 Control 3818 ± 119 7373 ± 79 4249 ± 98 8243 ± 109 Enzyme treated 1919 ± 104 1470 ± 64 2144 ± 87 1637 ± 53

a Values indicate mean ± standard deviation (SD).

Table 2 Composition of sugars and oligosacharides obtained by the hydrolysis of starches by a-amylase.a,b

Sugar/oligosacharides Corn starch Rice starch Tapioca starch Starch concentration (%) 10 15 20 10 15 20 10 15 20 Glucose 94.4 46.9 40.6 44.3 69.6 42.1 51.3 43.4 56.9 Maltose 5.6 18.5 21.3 18.0 9.7 11.0 28.7 24.1 16.9 Maltotriose NDc NDc NDc 31.1 9.3 19.1 16.5 21.8 14.7 Maltotetrose NDc 15.2 20.9 6.6 8.6 13.4 NDc NDc NDc Maltopentose NDc 6.0 11.6 NDc 2.7 6.6 3.4 8.2 7.6 Higher oligosacharides NDc 13.2 5.3 NDc NDc 7.8 NDc 2.5 3.9

a Sugars and oligosacharides were quantiﬁed by HPLC using aminopropyl column. b ND, not detected. c Values are mean of three independent determinations. 1430 S. Khatoon et al. / Food Research International 42 (2009) 1426–1433

15 min, these rheological parameters were increased to 3.5. High performance liquid chromatography 743 ± 93 Pa and 820 ± 104 Pas, respectively. Similar results were also observed for tapioca and rice starches. Many of the applica- HPLC-chromatograms of all three low DE maltodextrins showed tions of maltodextrins depend on the textural characteristics of peaks resolved for oligomers upto DE 10. Mc Ginnis and Fang the hydrolysates, which are attributed to the presence of higher (1980) showed that maltooligosaccharides with maltose or glyco- saccharides. These lower DE hydrolysates contribute chewiness syl stubs were eluted more slowly than the linear saccharides upto and mouthfeel to food products (Aime, Arntfield, Malcolmson, & DP 23 (Mc Ginnis & Fang, 1980). The low molecular weight mate- Ryland, 2001; Tamime, Kalab, Barrantes, & Sword, 1995). Hence, rial in fat mimetic corn, rice and tapioca maltodextrins consisted of a-amylase treatment of corn starch may be useful in making poly- mostly linear maltooligosaccharides (Alexander, 1992). The pres- saccharide based fat spreads. It is also concluded that higher solid ence of maltooligasaccharides and the peaks can be explained by content in dispersion is desirable as the rate of degradation of the practical hydrolysis of gelatinized starch pastes with a-amy- starch is faster. lase as first suggested by Ying, Chunguang, Wang, and Da-Wen The apparent viscosity of dispersions, measured at a shear rate (2006). When gelatinized starch paste is heated with a-amylase, of 100 s1, are 1984 ± 29.3, 18.9 ± 0.8 and 64.5 ± 2.1 for native amylose is hydrolyzed preferentially, either in the continuous li- unmodified starch, enzyme modified freeze dried starch and en- quid phase or inside the suspended gel phase of the paste. It is pos- zyme modified spray dried starch, respectively. Native unmodified sible that the preferential digestion of amylose occurred within the starch also has a very high yield stress (80–105 Pa) compared to granule, because those linear molecules occur in the amorphous the yield stress of 0.5–1.0 Pa for enzyme modified freeze dried phase and should be more accessible to the enzyme in swollen starch and 3–6 Pa for enzyme modified spray dried starch. Since starch granules. Linear maltooligasaccharides also may originate apparent viscosity and yield stress are very low for enzyme treated from a-amylase digestion of the A-chains of amylopectin. HPLC- spray dried starch, this modified starch possesses an easy to flow chromatograms showed that the maltodextrins contain higher lev- behavior. els of maltose and glucose, but reduced levels of oligosaccharides.

Fig. 2. Scanning electron micrographs of control and enzyme modiﬁed starches. A, B and C are controls and D, E and F are enzyme modiﬁed starches of corn, tapioca and rice, respectively. S. Khatoon et al. / Food Research International 42 (2009) 1426–1433 1431

Table 2 shows the composition of 70% aqueous ethanol extract of starches hydrolyzed by a-amylase. The enzyme modiﬁed starches contained disaccharides and oligosaccharides in consider- able amount particularly in 15% and 20% starch dispersions. These oligosaccharides may impart required functional properties in a fat substitute.

3.6. Scanning electron microscopy (SEM) of gelatinized and partially hydrolyzed starches

Unmodified starches showed that the granules are round to polygonal in shape (Fig. 2) similar to shapes reported in the literature (French, 1984). The granule surface is relatively smooth and free from pores, cracks or fissures (Fig. 2). Due to a-amylase treatment, the starch granule showed many pits or pores while size re- mains unaltered. However, some granules with larger pits were also observed. Enzymatic treatment resulted in weak structured granules with opened cracks and exposed pronounced layer structure. Similar observations were reported for corn starch digestion by a-amylase (Sreenath & Lafayette, 1992). Hydrolyzed corn, rice and tapioca starches contained some residual granular structure (Fig. 2). Presumably, most of the large granules were torn into small particles by a combination of high shearing forces in the atomizer of the spray drier and the warm temperature of the starch paste hydrolyzate. Excessively fine granulation is undesirable, because the fine particles form lumps when water is added to product (Cheetham & Tao, 1998).

3.7. 13C CP/MAS solid-state NMR structures of gelatinized and partially hydrolyzed starches

3.7.1. Chemical shift (Fig. 3) summarizes the 13C CP/MAS NMR spectra of native (control) and enzymatically modified corn, rice and tapioca starches. Chemical shift values for all resolved signals are given in Table 3. Assignments of the resonance were consistant with literature data (Sipahioglu, Alvarez, & Solano-Lopez, 1999; Napaporn, Sayavit, & Pavinee, 2004; Cheetham & Tao, 1998; Keith, Richard, & Nigel, Fig. 3. 13C CP/MAS NMR spectra of control and enzyme modified starches. A, B and 1995; Morgan, Furneause, & Larsen, 1995). The spectra indicate C are controls and D, E and F are enzyme modified starches of corn, tapioca and rice, that various carbon signals of starch samples. Among the 13C sig- respectively. nals observed, those for C1 and C6 were resolved between 102– 105 ppm and 62–64 ppm respectively. The C2 C3 and C5 signals were observed as a single large peak between 74–84 ppm. The Table 3 C4 signal observed 83–92 ppm was found to be merged with the Line width of various resonance peaks in 13C CP/MAS NMR spectra of native and signal for the C2, C3 and C5 for blank samples. modified starches. The chemical shift values of carbon atoms on modification Starch Line width (Hz) shows that all the starches after modification were shifted 1.6– C1 C2, C3, C5 C6 2.5 ppm up field to that of blank gelatinized starch. It indicates a Corn 0 change in dihedral angles (Ø 2) between the glucose ring on enzy- Native (control) 492 473 348 matic modification. The strongest resonance signal of the spectrum Modified starch 404 400 512 observed was that of combined C2, C3 and C5 atoms and the peak Tapioca appeared from 74–84 ppm. For corn starch C2, C3 and C5 signal Native (control) 503 488 346 Modified starch 454 396 449 was shifted 9.8 ppm after modification. Not much change was ob- Rice served for other starches. Native (control) 506 456 348 Various cyclomalto-oligosaccharide complexes (Veregin, Fyfe, Modified starch 380 399 423 Marchessault, & Taylor, 1987; Singh, Zakiuddin Ali, & Divakar, 1993) have found that the dihedral (torsion) angle C-1-0-4-C-4- formation. In the spectrum the peak signal of C4 shows broadening C50 (Ø02) bears linear correlation with the chemical shift values of all blank samples specially in tapioca and combined with C2, C3 for anomeric carbon atom. The chemical shift of C6 has been and C5 signal. After modification C4 signal resolutions for all the shown to be correlated with the torsion angles X0-6-C-6-C-5-C-4 starches were better. Not much change was shown in chemical i.e. the angle describing the orientation of primary hydroxyl group shift between blank and modified. and hence the conformation. The chemical shifts of C6 atom 60 ppm, 62 ppm and 66 ppm have been found to correspond to gg, gf and tg conformation respectively. The chemical shift value 3.7.2. Line width of C6 signal for the starches viz. corn, tapioca and rice are in the Line width (width of the spectral signal of half its height), data range of 62–64 ppm, and these value suggests the presence of con- for the various carbon atom signals are presented in Table 3.In 1432 S. Khatoon et al. / Food Research International 42 (2009) 1426–1433

Fig. 4. FTIR spectra of control and enzyme modiﬁed starches. A, B and C are controls and D, E and F are enzyme modiﬁed starches of corn, tapioca and rice, respectively.

general C1 signal was broadest. The C2, C3 and C5 were slightly 1080 cm1 and the intensity of the band at 1020 cm1 was in- narrowed, whereas C6 was broadened compared to blank. A nar- creased. These results suggest that the ordered structure of native rowed signal indicated that the relative mobility of the particular starch was disrupted as a result of enzymatic degradation and the carbon atom as compared to the other carbons. Moisture content structure of the modified starch is more amorphous in nature. Sim- of the samples was 1.5–1.8%. In case of C1 peak narrowed down ilarly, in corn and tapioca starches also the intensity of 1020 cm1 for all the starches viz. corn, rice and tapioca 75–102 Hz, 130– amorphous band increased upon enzymatic degradation. It is also 176 Hz and 47–52 Hz, respectively. C2, C3 and C5 peak also nar- interesting to note that the native starches has prominent band rowed down for all the starches. at 929 cm1. This band is sensitive to water and characteristic in- After modification, C6 atom resonance has broadened for corn dex of hydrophilicity of starches (Alexander, 1992). Upon enzy- starch which shows more change. In rice flour C1 and in corn matic modification, the intensity of this band is decreased in all starch C6 atom shows greater change during modification. The the starches. The results of the FTIR spectra suggest the formation C6 carbon atom which is involved in the branching. During modi- of amorphous structure in starch samples of rice, corn and tapioca fication of corn starch debranching effects are more pronounced with concomitant decrease in the ordered structure of starch than other starch. The narrowing of the line width (i.e. decrease (Sevenou, Hill, Farhat, & Mitchell, 2002). in chemical shift range) indicates attainment of a higher crystalline order (French, 1984; Veregin et al., 1987). Decrease in the line 4. Conclusions width suggests an increased crystallinity during modification. Low DE starch hydrolysate can be prepared by a-amylase treat- 3.8. Comparison of FTIR spectra of modified starches ment of starches at 95 °C. Ultrastructure of enzyme modified starch granules are weaker with open cracks and exposed layer. Fig. 4 shows the FTIR spectra of native and enzymatically mod- Oligosaccharides with broad molecular weight distributions were ified starches of corn, tapioca and rice starches. The FTIR spectra of detected in the partially hydrolyzed starches. Presence of these oli- native rice starch have three main modes with maximum absor- gosaccharides may impart required functional properties for the bance at 1156, 1080 and 1020 cm1. The bands at 1156 cm1 and fat substitute. Spectral studies of modified starches suggested the 1080 cm1 are associated to the ordered structures of starch, formation of amorphous structure with concomitant decrease in whereas, the band at 1020 cm1 is associated to the amorphous the ordered structure of starches. A marked decrease in solid con- structures of starch (Dupuy, Wojciechowski, Ta, Huvenne, & Le- tent and overall resistance to flow of the sample were observed in grand, 1997). Enzymatic degradation of rice starch with a-amylase the hydrolyzed starches. Therefore, these modified starches may be resulted in the decrease in the intensity of bands at 1156 and suitable to formulate products that can easily flow and spread. S. Khatoon et al. / Food Research International 42 (2009) 1426–1433 1433

Acknowledgements Gidley, M. J., & Bociek, M. S. (1985). Molecular organization in starches: A 13C CP/ MAS NMR study. Journal of American Oil chemists Society, 107, 7040–7044. Marchal, L. M., Beeftink, H. H., & Tramper, J. (1999). Towards a rational design of This work was supported by the Department of Biotechnology, commercial maltodextrins. Trends in Food Science and Technology, 10, 345–355. Government of India. The authors thank Dr. V. Prakash, Director, Mc Ginnis, G. D., & Fang, P. (1980). High performance liquid chromatography. In R. L. Central Food Technological Research Institute (CFTRI), Mysore for Whistler & J. N. Bc Miller (Eds.), Methods in carbohydrate chemistry (pp. 33–43). New York: Academic Press. his keen interest and the encouragement. The authors also thank Mc Pherson, A. E., & Seib, P. A. (1997). Preparation and properties of wheat and corn Dr. Suryaprakash, Principal Research Scientist, Sophisticated starch maltodextrin with a low dextrose equivalent. Cereal Chemistry, 74, Instruments Facility, Indian Institute of Science, Bangalore for 424–430. Miller, G. L. (1959). Use of dinitrosalycilic acid reagent for the determination of NMR analysis. Ms. M. Asha of Central Instrumentation Facility, reducing sugars. Analytical Chemistry, 31, 426–428. CFTRI assisted in FTIR operation. Morgan, K. R., Furneause, R. H., & Larsen, N. G. (1995). Solid-state NMR studies on the structure of starch granules. Carbohydrate Research, 276, 387–399. Napaporn, A., Sayavit, V., & Pavinee, C. (2004). A study of ordered structure in acid References modified tapioca starch by 13C CP/MAS solid state NMR. Carbohydrate Polymers, 58, 383–389. Aime, D. B., Arntfield, S. D., Malcolmson, L. J., & Ryland, D. (2001). Textural analysis Richter, M., Schierbaum, F., Augustat, S., Knoch, K. D. (1976). Method of producing of fat reduced vanilla ice cream products. Food Research International, 34, starch hydrolysis products for use as a food additive. US patent 396246. 237–246. Sevenou, O., Hill, S. E., Farhat, I. A., & Mitchell, J. R. (2002). Organization of the Alexander, R. J. (1992). Maltodextrins: Production properties and applications. In F. external region of the starch granules as determined by infrared spectroscopy. W. Scenck, R. A. Hebeda (Eds.), Starch hydrolysis products, worldwide technology International Journal of Biological Macromolecules, 31, 79–85. production and applications (pp. 233–276). Sipahioglu, O., Alvarez, V. B., & Solano-Lopez, C. (1999). Structure and physico- AOAC, Association of Official Analytical Chemists (1975). Official methods of analysis chemical and sensory properties of feta cheese made with tapioca starch and (11th ed.). Wahington, DC: AOAC. pp. 927–928. lecithin as fat mimetics. International Dairy Journal, 9, 783–789. Bernfeld, P. Amylases a and b. (1955). In S. P. Colowick, N.O. Kalpam. Methods in Sreenath, H. K., & Lafayette, W. (1992). Studies on starch granules digestion by a- enzymology (Vol. 1, pp. 149–158). amylase. Starch/stärke, 44, 61–63. Byars, J. (2002). Effect of a starch-lipid fat replacers on the rheology of soft serve ice Tamime, A. Y., Kalab, M., Barrantes, E., & Sword, A. M. (1995). The effect of starch- cream. Journal of Food Science, 67, 2177–2182. based fat substitutes on rheological properties and microstructure of set style Cheetham, N. W. H., & Tao, L. (1998). Solid state NMR studies on the structural and yogurt made from reconstituted skim milk powder. Scanning, 17(Suppl. V), conformational properties of natural maize starches. Carbohydrate Polymers, 36, V100. 283–292. Singh, Vasudeva, Zakiuddin Ali, S., & Divakar, S. (1993). 13C CP/MAS NMR Dubois, M., Gilles, K. A., Hamilton, J. K., Rebers, P. A., & Smith, F. (1956). Calorimetric Spectroscopy of native and acid modified starches. Starch/Stärke, 45, 59–62. determination of sugar and related substances. Analytical Chemistry, 28, Veregin, R. P., Fyfe, C. A., Marchessault, R. H., & Taylor, M. G. (1987). Correlation of 350–356. 13C-chemical shift with torsional angles from high-resolution 13C-C.P.-M.A.S. Dupuy, N., Wojciechowski, C., Ta, C. D., Huvenne, J. P., & Legrand, P. (1997). Mid- NMR studies of crystalline cyclomato-oligosaccharide complex, and their infrared spectroscopy and chemometrics in corn starch classification. Journal of relation to the structure of the starch polymorphs. Carbohydrate Research, 160, Molecular Structure, 410–411, 551–554. 41–56. French, D. (1984). Organisation of structure granules. In R. L. Whistler, J. M. Bc Ying, M., Chunguang, C., Wang, J., & Da-Wen, S. (2006). Enzymatic hydrolysis for Miller, & E. F. Paschall (Eds.), Starch chemistry and technology (2nd ed., producing fat mimetics. Journal of Food Engineering, 73, 297–303. pp. 184–247). New York: Academic Press.