Effects of Milling on the Physicochemical Characteristics
of Waxy Rice in Taiwan1
J. J. Chen,2 S. Lu,2,3 and C. Y. Lii4
Cereal Chem. 76(5):796–799
Three types of mills and six milling methods were employed to mill
coarse particles (100–300 µm); cyclone and turbo milling led to a more
two waxy rice varieties (TCSW1, long grain; TCW70, short grain), and
even particle-size distribution, and the wet-milling gave the finest particles
the physicochemical and functional properties of rice flour were examined.
(10–30 µm). Dry hammer-milled rice had higher gelatinization and pasting
The results showed that dry-milling maintained a higher level of the
temperatures, and semi-dry grinding milling resulted in the lowest pasting
chemical components than other milling methods. Wet-milling slightly
temperature, setback viscosity, and enthalpy value among the mills. The
increased solubility as test temperatures increased, and significantly in-
final quality of the two waxy rice varieties was profoundly affected by the
creased swelling power at 75 and 85°C for TCSW1 and TCW70, respec-
mill type and milling method.
tively. Hammer and semi-dry hammer milling gave higher percentages of
Waxy rice (Oryza sativa L.), also called glutinous, sweet, or
For dry-milled flour, polished rice kernels were ground with turbo,
mochi rice, is characterized chiefly by very low amounts of amylose
cyclone, or hammer mills. For semi-dry milled flour, rice samples
in the starch. It can be classified as indica and japonica types which
(30% mc) were steeped for 1 hr and then centrifuged for ?1 min
are commonly used for food in Taiwan (Chang 1990). Besides
(2,000 rpm) to remove water. The rice was then ground with the
inherent preprocessing starch properties and storage history, dif-
hammer or plate mill, and dried in a hot air oven at 40°C for 12 hr
ferent types of mills or grinders profoundly affect the physico-
to reduce the moisture to 13%. For wet-milled flour, the rice kernels
chemical and functional properties of rice flour (Nishita and Bean
were steeped for 1 hr and then ground with four times its weight
1982, Bean and Nishita 1985, Lu and Lii 1989, Arisaka et al 1992,
in water with a double-disk stone mill. The slurry was poured into
Yang 1994, Chen 1995). However, wetting encourages uniform
a thick cloth bag and centrifuged (2,000 rpm) to remove the free
cooking instead of starchy, gummy cooking, and the steeped prep-
water. The wet-milled flour was then dried in a hot-air oven at
aration results in a smoother differential scanning calorimetry (DSC)
40°C for 12 hr to reduce the moisture to 13%.
scan than does dry milling (Chen 1995, Kohlwey et al 1995).
In some Asian countries, rice is milled to flour or to a coarse
meal as part of the process for making traditional rice-based baked
The moisture, crude protein, crude lipid, and ash for all the rice
or steamed products (Juliano and Sakurai 1985, Chen 1988). Wet-
flours were determined using Approved Methods 44-15A, 46-11A,
milled rice flour usually produces better texture than dry milling.
30-10, and 08-01 (AACC 1995). The conversion factor (N × 5.95)
However, part of the water-soluble vitamins, albumin, sugars, and
was applied to convert nitrogen to crude protein content. The color
some lipids are lost during wet-milling (Chen 1988, Yang 1994, Chen
of the waxy rice flours was determined by a ?80 Color System
1995, Juliano and Hicks 1996). Because of high costs and environ-
(Nippon Deshoku Industry Co., Japan). The L, a, b scale was used
mental concerns, dry or semi-dry methods have been used to produce
as the light to dark, red to blue, and yellow to green indices.
quality similar to that obtained from wet-milling (Lu and Lii 1989,
The particle-size distribution was measured with a Galai CIS-1
system (Galai Production Ltd., Israel). Sample (?100 mg) was mixed
In this study, two waxy rice varieties were milled to flour using
with 400 mL of ethanol, and the particle-size distribution was ex-
six milling methods. Changes in the resulting chemical compounds
pressed as a percentage (%) of the waxy rice flour. The median
and physicochemical characteristics of the waxy rice flour were
particle size (PS50) was determined as the estimated sieve size
through which 50% of the sample would pass.
Swelling power and solubility were determined at a test temper-
MATERIALS AND METHODS
ature range of 65–95°C according to Schoch (1964) with minor
modifications (Lii et al 1986).
Starch-pasting properties were measured with a Rapid Visco-
Two waxy type rice varieties, japonica type (Taichung Waxy 70,
Analyser (RVA model 3D, Newport Scientific Pty., Ltd., Narrabeen,
TCW70) and indica type (Taichung Sen Waxy 1, TCSW1) were
Australia). Rice flour (3 g, 14% mb) was weighed directly into an
obtained from the Taichung District Experimental Station, Chang-
aluminum RVA canister. Distilled water was added to a total of 28 g.
Haw county, Taiwan, in 1994, and packed in laminated polyethylene
Sample was held at 50°C for 1.5 min, heated to 95°C in 3.7 min,
film bags and then stored in a cold room at 4°C before milling
held at 95°C for 2 min, cooled to 50°C in 3.7 min, and then held
process. The rice was milled using turbo (made in Taiwan), cyclone
at 50°C for 6.1 min. Apparent viscosity was recorded in Rapid
(Udy Corp., Boulder, CO), hammer (Culatti, type MDCI, Swiss),
Visco-Analyser units (1 RVU = ?10 cps) (Welsh et al 1991).
plate (Straub Co., Philadelphia, PA), or stone (made in Taiwan)
Differential scanning calorimetry (DSC) was performed with a
Setaram DSC 121 (Setaram Co., France). The heating rate was
5°C/min from 20 150°C. Samples (110–120 mg) were premixed
with water (2:3, w/w) and kept at 4°C overnight to allow a uniform
distribution of water in the flours. These samples were placed in
1 Presented in part at the AACC 82nd Annual Meeting, San Diego, CA, October 1997.
stainless crucibles and reweighed before the DSC analysis. Onset
Department of Food Science, National Chung-Hsing University, Taichung, Taiwan
(To) and peak (Tp) transition temperatures were determined for each
3 Corresponding author. E-mail: email@example.com
endothermic by a computerized system developed by the Setaram
4 Institute of Chemistry, Academia Sinica, Nankang, Taipei, Taiwan, R.O.C.
Co. The transition enthalpies (?H) were determined from the peak
area of the endothermic and expressed as joules per gram of dry
Publication no. C-1999-0809-04R.
© 1999 American Association of Cereal Chemists, Inc.
matter (Huang et al 1994).
796 CEREAL CHEMISTRY
and ash than did wet-milling (P < 0.05). When soaked rice kernels
Data were analyzed using analysis of variance (ANOVA) to detect
are processed by wet-milling, some soluble protein, sugars, and
any differences in mean values from replicate runs of each treat-
nonstarch bound lipids are washed out (Medcalf and Lund 1985,
ment (SAS Institute, Cary NC). Duncan’s new multiple range test
Chen 1988, Yang 1994, Chen 1995, Juliano and Hicks 1996).
(? = 0.05) was used for the comparison of sample means. Pearson’s
Color measurement showed that the finer the flours, the brighter
correlation procedure was used to examine the degree of associ-
and whiter their color (Table I). Stone-milled flour gave the highest
ation between variables.
L value for whiteness and the lowest a and b values. The b value
indicated a tendency toward decreasing yellows as the particles
RESULTS AND DISCUSSION
became finer. The correlation coefficient showed a significance
between PS50 and L and b as well as white index (WI) of –0.77,
Effects of Milling on Chemical Composition and Color
0.86, and –0.75, respectively (P < 0.01). The results indicate that
The two waxy rice cultivars, TCW70 and TCSW1, contained
the sample particle size affected the color, and that the smaller
4.91–8.03% protein, 0.19–0.93% ash, and 0.30–2.50% lipids
flour particles resulted in a smoother surface. Since the samples
(Table I). Dry-milling resulted in higher contents of protein, lipid,
were different in particle size, the surface texture reflected from
Fig. 2. Effect of milling methods on the swelling power of TCW70 (A) and
Fig. 1. Effect of milling methods on the particle-size distribution of
TCSW1 (B) rice flours.
TCW70 (A) and TCSW1 (B) rice flours.
Effect of Milling Type on Chemical Composition (%) and Color of Rice Floursa,b
a Means of triplicates ± standard deviation.
b Means within row with different letters are different significantly at P < 0.05.
c Milling types A–F: turbo; cyclone; hammer; grinding; semi-dry hammer; wet stone, respectively.
d White index = 100 – ?(100 – L)2 + a2 + b2.
Vol. 76, No. 5, 1999 797
Fig. 3. Effect of milling methods on solubility of TCW70 (A) and TCSW1
(B) rice flours.
Fig. 4. Effect of milling methods on the pasting characteristics of TCW70
(A) and TCSW1 (B) rice flours.
the sample would vary among samples (Bean 1986, Kurimoto and
including peak viscosity (PV), maximum setback viscosity (SB),
Distribution of Particle Size
and pasting temperature (PT). There were no systematic differences
Dry and semi-dry hammer milled flours were coarser for both
in PV, SB, and PT for rice varieties. However, the angle and the
rice varieties, and the semi-dry hammer milled flour was finer than
sharpness of slope from initiation of viscosity to the peak viscosity
dry hammer-milled rice flour at 100–300 µm. Cyclone- and turbo-
were clearly different between the two samples. Dry hammer-milled
milled flours had more even distributions, and stone-milled had
flour showed the highest pasting temperature, and the semi-dry
more finer flour (10–30 µm), ?62.92 and 71.84% for TCW70 and
grinding milled flour had the lowest pasting temperature and
TCSW1, respectively (Fig. 1). The PS50 of all samples ranged
setback viscosity. The pasting curve for dry hammer-milled flours
from 12.3 to 215.1 µm (Table I), indicating that all samples were
showed negligible viscosity breakdown during a heating-hold cycle
significantly different from each other (P < 0.05). The turbo-mill
and higher viscosity during the cooling cycle, indicating that retro-
heated the sample to 50°C, the hammer mill also reached 42°C
gradation is rapid. In the relationship between particle size and
during milling, but the stone-milled samples stayed at room tem-
pasting temperature, the finest flours had the lowest initial onset
perature. Scanning electron microscopy showed that dry-milled
temperature, while the coarse flours had the highest. The breakdown
rice flour has clump starch granules, but the starch granules from
value (BD) decreased as PS50 increased significantly (r = –0.67, P
wet-milled samples were separated (Arisaka et al 1992).
< 0.05). However, the wet-milled rice flour from TCW70 had
higher peak (306 ± 20 RVU) and breakdown (178 ± 14 RVU)
Swelling Power and Solubility
viscosity than those from TCSW1. This is probably because the
The resulting data on swelling power and solubility for the dif-
flour with larger particles could not hydrate or expand as rapidly
ferent milling methods are shown in Figs. 2 and 3. For wet-milling,
and was inhibited by the starch on the particle surface (Jomduang
solubility increased slightly as the test temperature increased (Fig. 3),
and Mohamed 1994). Nishita and Bean (1982) reported that the
but the swelling power increased significantly at 75 and 85°C for
absence of peak viscosity is due to delayed swelling of the starch
TCSW1 and TCW70, respectively (Fig. 2). The grinding-milled
granules that are embedded in the relatively large endosperm
rice flour had higher solubility than the others. Possibly, the dif-
chunks in coarse flours.
ferences were caused by damaged starch produced during the grinding
process (Arisaka et al 1992, Chen 1995). The waxy type starch
Pasting Behavior Determined by DSC
granules unrestricted the swelling and resulted in the absence of a
Semi-dry plate milled flour showed the lowest ?H value.
network structure from amylose molecules that can hold the starch
TCSW1 rice flour had higher To and Tp values than TCW70, but a
molecules together (Tester and Morrison 1990).
nonsignificant difference in ?H (Table II). The To value decreased
as PS50 decreased significantly (r = 0.59, P < 0.05). The lowest
Pasting Behavior Determined by RVA
?H value came from the semi-dry plate milled flour, which compared
Pasting properties for the rice flours as measured by the RVA are
with highest ?H value obtained from the wet-milled flour. It
shown in Fig. 4. Milling method and cultivar affected all parameters,
reflects the considerable disruption of the native crystalline structure
798 CEREAL CHEMISTRY
Arisaka, M., Nakamura, K., and Yoshii, Y. 1992. Properties of rice
Effect of Milling Type on Thermal Behavior of Rice Flours
flour prepared by different milling methods. Denpun Kagaku
Determined by Differential Scanning Calorimetrya,b
Bean, M. M., and Nishita, K. D. 1985. Rice flours for baking. Page 539
in: Rice: Chemistry and Technology, 2nd ed. B. O. Juliano, ed. Am.
Assoc. Cereal Chem.: St. Paul, MN.
Bean, M. M. 1986. Rice flour—Its functional variations. Cereal Foods
Carpio, E. V., and Aco, E. V. 1990. Factors affecting the dry-milling
characteristic of rice flour using a pin mill. Philippine Agriculturist
Chang, P. Y., Lin, S. Y., Li, C. F., Chao, R. S., and Chen. S. C. 1990.
Studies on the utilization of rice flour. Res. Rep.131. Food Industry
Research and Development Institute: Taiwan, R.O.C.
Chen, J. J. 1995. Effect of milling methods on the physicochemical
properties of waxy rice flour. MS thesis. National Chung-Hsing Uni-
Chen, W. C. 1988. Studies on the effects of rice milling methods on rice
flours properties and development a new rice product. MS thesis.
National Chung-Hsing University: R.O.C.
a Means of triplicates ± standard deviation.
Huang, R. M., Chang, W. H., Chang, Y. H., and Lii, C. Y. 1994. Phase
b Means within row with different letters are different significantly at P < 0.05.
transition of rice starch and flour gels. Cereal Chem. 71:202-207.
c Milling types A–F: turbo; cyclone; hammer; grinding; semi-dry hammer;
Juliano, B. O., and Hicks, P. A. 1996. Rice functional properties and rice
wet stone, respectively.
food products. Food Reviews Int. 12:71-103.
Juliano, B. O., and Sakurai, J. 1985. Miscellaneous rice products. Page
occurred during the grinding process. The mechanical force caused
569 in: Rice Chemistry and Technology, 2nd ed. B. O. Juliano, ed.
starch damage during milling, with a lower ?H value in rice flour
Am. Assoc. Cereal Chem.: St. Paul, MN.
Jomduang, S., and Mohamed, S. 1994. Effect of amylose/amylopectin
(Bean and Nishita 1985, Park et al 1988, Lu and Lii 1989, Carpio
content, milling methods, particle size, sugar, salt and oil on the puffed
and Aco 1990, Arisaka et al 1992, Jomduang and Mohamed 1994,
product characteristics of a traditional Thai rice-based snack food
Yang 1994, Chen 1995). Marshall (1992) reported that a reduction
(Khao Kriap Waue). J. Sci. Food Agric. 65:85-93.
in particle size caused a large decrease in Tp, Tc, and ?H, but only
Kohlwey, D. E., Kendall, J. H., and Mohindra, R. B. 1995. Using the
a modest decrease in To. The cooking quality is affected by moistening
physical properties of rice as a guide to formulation. Cereal Foods
or steeping the rice flour, which also produces a smoother DSC
scan (Kohlwey et al 1995).
Kurimoto, Y., and Shelton, D. R. 1988. The effect of flour particle size on
baking quality and other flour attributes. Cereal Foods World 33:429-433.
Lu, S., and Lii, C. Y. 1989. The influences of various milling processes
on the physicochemical properties of rice flours and the rice flake prep-
aration. Food Sci. (R.O.C) 16:22-35.
In this study, we found that waxy rice variety, mill type, and
Lii, C. Y., Chang, S. M., and Yang, H. L. 1986. Correlation between the
milling process profoundly affected the physicochemical charac-
physico-chemical properties and the eating quality of milled rice in
teristics of rice flours. The differences in milling parameters resulted
Taiwan. Bull. Inst. Chem. Academia Sinica 33:55-62.
in changes in functional properties. Thus, rice variety and milling
Marshall, W. E. 1992. Effect of degree of milling of brown rice and
method are two major factors affecting the quality of rice flours.
particle size of milled rice on starch gelatinization. Cereal Chem.
Medcalf, S. L., and Lund, D. B. 1985. Factors affecting water uptake in
milled rice. J. Food Sci. 50:1676-1679.
We wish to thank the Taichung District Agricultural Improvement
Nishita, K. D., and Bean, M. M. 1982. Grinding methods: Their impact
Station, Taiwan, R. O. C for providing the two waxy rice varieties.
on rice flour properties. Cereal Chem. 59:46-49.
Park, N. K., Seog, H. M., Nam, Y. J., and Shin, D. H. 1988. Physico-
chemical properties of various milled rice flours. Korean J. Food Sci.
Schoch, T. J. 1964. Swelling power and solubility of granular starches.
American Association of Cereal Chemists. 1995. Approved Methods of
Pages 106-108 in: Methods in Carbohydrate Chemistry. R. L. Whistler
the AACC, 9th ed. Method 08-01 approved April 1961, revised October
ed. Academic Press: New York.
1981 and October 1986; Method 30-10 approved April 1961, revised
Tester, R. F., and Morrison, W. R. 1990. Swelling and gelatinization of
October 1975 and October 1981, reviewed October 1994; Method 44-
cereal starches. II. Waxy rice starches. Cereal Chem. 67:558-563.
15A approved October 1975, revised October 1981 and October 1994;
Welsh, L. A., Blakeney, A. B., and Bannon, D. R. 1991. Rapid viscometric
Method 46-11A approved October 1976, revised October 1982 and
analysis of rice flour. Int. Rice Newsl. 16(5):11-12.
September 1985, reviewed October 1994; Method 76-30A approved
Yang, C. H. 1994. Studies on the preparation, processing properties, and
May 1969, revised November 1972, October 1982, and October 1984,
the factors influencing the properties of semi-dry milled rice flour.
reviewed October 1994. The Association: St. Paul, MN.
PhD thesis. National Taiwan University: R.O.C.
[Received October 6, 1998. Accepted June 17, 1999.]
Vol. 76, No. 5, 1999 799