See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/222567714

A Simple Method for Production of Pure Silica

from Rice Hull Ash

Article in Bioresource Technology · July 2000

DOI: 10.1016/S0960-8524(99)00127-3

CITATIONS READS

234 10,800

3 authors, including:

Uruthira Kalapathy John L Shultz

Claflin University University of Arkansas
26 PUBLICATIONS 1,384 CITATIONS 35 PUBLICATIONS 1,401 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Interaction of Silica nanoparticles with fluorescent molecules. View project

Value Added Products View project

All content following this page was uploaded by John L Shultz on 23 April 2017.

The user has requested enhancement of the downloaded file. All in-text references underlined in blue are added to the original document
and are linked to publications on ResearchGate, letting you access and read them immediately.
 Bioresource Technology 73 (2000) 257±262

A simple method for production of pure silica from rice hull ash
U. Kalapathy a,*, A. Proctor a, J. Shultz b
a
Department of Food Science, University of Arkansas, 272 Young Ave, Fayetteville, AR 72704 ,USA
b
Arkansas Analytical Laboratory, High Density Electronics Center, University of Arkansas, Fayetteville, AR 72701, USA
Received 8 October 1998; received in revised form 2 September 1999; accepted 2 September 1999

Abstract
Rice hull ash (RHA), a waste product of the rice industry is rich in silica. A simple method based on alkaline extraction followed
by acid precipitation was developed to produce pure silica xerogels from RHA, with minimal mineral contaminants. The silica gels
produced were heated to 80o C for 12 h to obtain xerogels. Silica and mineral contents of xerogels were determined by energy
dispersive X-ray (EDX) and inductively-coupled plasma (ICP) emission spectrometers, respectively. Xerogels produced from RHA
had 93% silica and 2.6% moisture. The major impurities of silica produced from RHA at an extraction yield of 91% were Na, K, and
Ca. Acid washing prior to extraction resulted in silica with a lower concentration of Ca (<200 ppm). However, ®nal water washing
of the xerogel was more e€ective in producing silica with lower overall mineral content (Na < 200 ppm and K < 400 ppm). X-ray
di€raction patterns revealed the amorphous nature of silica xerogel. Fourier transform infrared (FTIR) data indicated the presence
of siloxane and silanol groups. Ó 2000 Elsevier Science Ltd. All rights reserved.

Keywords: Rice hull ash; Silica xerogels; Minerals

1. Introduction and Scherer, 1990). Mizuki et al. (1993) and Krishnaro

Burning rice hull as fuel to generate energy results in
the waste product, rice hull ash (RHA). RHA is rich in
silica (about 60%) and can be an economically viable
raw material for production of silica gels and powders
(Kamath and Proctor, 1998; Chakraverty and Kale-
emullah, 1991). RHA has been evaluated as an adsor-
bent of minor vegetable oil components (Proctor et al.,
1995; Proctor and Palaniappan, 1990). Although vari-
ous uses for rice hull and RHA have been suggested in
the literature, their disposal or utilization remains a
major concern.
Soluble silicates produced from silica are widely used
in the glass, ceramics, and cement as a major component
and in pharmaceuticals, cosmetics, and detergents in-
dustries as a bonding and adhesive agents (Anon., 1997;
Laxamana, 1982). Silica also has been used as a major
precursor for a variety of inorganic and organometallic
materials which have applications in synthetic chemistry
as catalysts, and in thin ®lms or coatings for electronic
and optical materials (Lender and Ruiter, 1990; Brinker
*
Corresponding author. Tel.: +1-501-575-4605; fax: +1-501-575-
6936.
and Godkhindi (1992) have investigated the formation
of silicon carbide from rice hulls at high temperatures
and Maeda and Komatsu (1996) studied the semicon-
ductor properties of silicon carbide produced from rice
hull. Semiconductor industries require pure silica, cur-
rently produced by smelting quartz in a high tempera-
ture furnace, to produce ultrapure polycrystalline silicon
and silicon hydride. Development of a simple low energy
chemical method for producing pure silica should lead
to a variety of industrial applications for RHA.
Acid leaching (Chakraverty and Kaleemullah, 1991)
and gasi®cation (Luan and Chou, 1990) methods have
been investigated for recovering silica from rice hull.
Amorphous silica from RHA can be extracted using low
temperature alkali extraction; the solubility of amor-
phous silica is very low at pH < 10 and increases sharply
pH > 10. This unique solubility behavior enables silica to
be extracted in pure form from RHA by solubilizing
under alkaline conditions and subsequently precipitat-
ing at a lower pH (Iler, 1979; Kamath and Proctor,
1998). This low energy method based on alkaline solu-
bilization of amorphous silica could be more cost ef-
fective compared to the current smelting method.
Silica gel is a rigid three-dimensional network of
colloidal silica, and is classi®ed as aquagel (pores are

0960-8524/00/$ - see front matter Ó 2000 Elsevier Science Ltd. All rights reserved.
PII: S 0 9 6 0 - 8 5 2 4 ( 9 9 ) 0 0 1 2 7 - 3
258 U. Kalapathy et al. / Bioresource Technology 73 (2000) 257±262
®lled with water), xerogel (aqueous phase in the pores is
removed by evaporation) or aerogel (solvent is removed
by supercritical extraction) depending on how they were
made. Xerogels are mainly used to prepare dense ce-
ramics. Further, their high porosity and surface area
leads to applications such as catalytic substrates, ultra-
®lters, and chromatography column packing materials
(Brinker and Scherer, 1990). Since impurities present in
silica gels can a€ect their surface adsorption properties,
purity of the silica is very important in characterizing
rice hull silica adsorption.
In previous work, silica gel produced from RHA
using an alkali extraction method had over 65% mois-
ture, 2% of sodium, 0.04% of potassium, 0.01% of cal-
cium and traces (<0.01%) of other elements (Kamath
and Proctor, 1998). On the basis of dry material, Na, K,
and Ca concentrations in this gel would be 6%, 0.12%,
and 0.03%, respectively. Concentrations of Na and K in
this gel were relatively high compared to Na, and K
concentrations in commercial silica. Hence, an e€ective
mineral leaching step is necessary to produce xerogels
with reduced Na and K concentrations from RHA.
The objective of this study was to investigate the ef-
fectiveness of: (1) washing RHA with acid prior to alkali
extraction, and (2) washing dried silica gel (xerogel) with
water, in reducing the mineral content of silica produced
by the alkali extraction method developed in our
laboratory.
2. Methods
2.1. Acid washing
An acid washing step was used to remove the small
quantities of minerals prior to silica extraction from
RHA in the following manner. Ten grams of RHA
samples were dispersed in 60 ml of distilled water, and
the pH was adjusted to 1, 3, 5 or 7 using 6 N and 1 N
HCl. These dispersions were stirred for 2 h, ®ltered
through Whatman No. 41 ashless ®lter paper and then
the RHA residues were washed with 100 ml of water.
The residues were used for silica extraction. The ®ltrate
and washings at each pH were collected and dried in an
evaporating dish. These dried ®ltrates containing soluble
minerals were re-dissolved to determine Na, K, Ca, Mg,
Fe, and Mn contents by inductively-coupled plasma
(ICP) emission spectrometer (Spectro Analytical In-
struments, Fitchburg, MA).
2.2. Silica extraction
Silica was extracted from RHA adapting the method
of Kamath and Proctor (1998). Sixty ml portions of 1N
NaOH were added to the washed and unwashed RHA
samples and boiled in covered 250 ml Erlenmeyer ¯asks
for 1 h with constant stirring to dissolve the silica and
produce a sodium silicate solution. The solutions were
®ltered through Whatman No. 41 ashless ®lter paper,
and the carbon residues were washed with 100 ml of
boiling water. The ®ltrates and washings were allowed
to cool to room temperature and were titrated with 1N
HCl with constant stirring to pH 7. Silica gels started to
precipitate when the pH decreased to <10. The silica gels
formed were aged for 18 h. Deionized water (100 ml)
was added to gels and then the gels were broken to make
a slurry. Slurries were then centrifuged for 15 min at
2500 rpm, the clear supernatants were discarded and the
washing step was repeated. The gels were transferred
into a beaker and dried at 80o C for 12 h to produce
xerogels. Selected silica xerogel samples were ground
and subjected to additional washing with water. All the
samples were stored in airtight plastic bottles. Major
modi®cations in the present method relative to that of
Kamath and Proctor (1998) were incorporation of: (1)
an initial acid washing, (2) longer drying time to pro-
duce the xerogel, and (3) ®nal washing of the xerogel. A
¯ow diagram of the procedure is shown in Fig. 1.
2.3. Moisture content of silica gels
Moisture content of the silica gels was determined
using an air oven method (AOAC, 1990). About 1 g of
each sample was heated in aluminum moisture pans at
130o C for 1 h. The samples were cooled in a dessicator
and weighed. The weight loss (%) was recorded as the
moisture content of samples.
2.4. Chemical analyses of silica gels
The silicon content of the samples was estimated us-
ing energy dispersive X-ray (EDX) spectroscopy (Kevex
Instruments, Valencia, CA). Mineral content of the xe-
rogels was determined by nitric acid digestion followed
by atomic emission measurements using ICP emission
spectroscopy (Spectro Analytical Instruments, Fitsburg,
MA).
X-ray di€raction patterns were obtained by a dual
goniometer XÕpert XRD system (Philips Electronic In-
struments, The Netherlands) using an acceleration
voltage of 40 kV and current of 45 mA. The di€raction
angle was scanned from 10o to 90o 2h, at a rate of 5o /
min. Fourier transform infrared (FTIR) data were ob-
tained by adding 100 interferograms using a Impact 410
Nicolet instrument (Analytical Instruments, Madison,
WI).
2.5. Experimental design and statistical analyses
Silica was prepared from unwashed and acid washed
RHA, with and without ®nal washing of xerogel with
water, to investigate the e€ect of acid washing of RHA
 U. Kalapathy et al. / Bioresource Technology 73 (2000) 257±262 259

Fig. 2. EDX spectrometric data of silica produced from RHA.

was estimated from the EDX data based on the as-

sumption that all of the silicon was in the form of silica.
The silica and moisture contents of xerogels produced
from unwashed RHA were 93% and 2.7%, respectively.
Compositions of minerals leached from RHA by acid
washing at pH 1, 3, 5, and 7 are shown in Table 1. Since
acid leaching at a pH of 1 was found to be more e€ective
in removing Ca, Fe, and Mn from RHA, a washing step
at pH 1 was used in all subsequent acid leaching treat-
ments. Unwashed RHA, and RHA washed at pH 7 were
used as controls to evaluate the e€ectiveness of acid
washing on the purity of the xerogel. The e€ect of acid
leaching and ®nal washing of xerogel on the mineral
content of RHA silica is shown in Table 2. Trace metal
ions such as polyvalent iron and manganese with 3+,
and 4+ charges, respectively, might have been strongly
Fig. 1. Flow diagram of the procedure used to extract silica from adsorbed onto silica surface and subsequently trapped
RHA. inside the silica network during gel formation from sil-
icate solution (Iler, 1979). As a result of this, Fe and Mn
might not be completely solubilized during the nitric
and washing xerogel with water on the purity of silica. acid digestion for ICP measurements. The Fe and Mn
Silica was also prepared from RHA washed at pH 7, as concentrations in the silica would be a€ected by the
an additional treatment control to establish the e€ect of factors such as surface area, porosity, and particle size
acid washing. All the treatments were performed in
triplicate in a completely randomized design. Data were
analyzed by ANOVA and means were separated by the Table 1
Mineral content of acid solubles from RHA
least signi®cant di€erence when signi®cant F (P < 0.05)
values were observed (SAS, JMP IN Version 3.2.1. SAS Minerals Acid leaching treatment
Institute, Cary, NC, 1997). pH 1 pH 3 pH 5 pH 7
Sodium (%) 2.42a 1.97a 1.81a 1.90a
Potassium (%) 29.5a 31.7a 39.4a 37.5a
Calcium (%) 8.14a 7.03a 2.77b 3.93b
3. Results and discussion Magnesium (%) 2.57a 2.85a 2.05a 2.41a
Iron (ppm) 3204a 539b 20c 17c
Manganese (ppm) 19745a 1557c 2228c 5823b
3.1. Silica extract composition *
Values are means of triplicate analyses. In each row, values with
di€erent superscripts are signi®cantly (P < 0.05) di€erent from each
EDX elemental spectra of xerogel produced from other. Average standard deviations of Na, K, Ca, Mg, Fe, and Mn
unwashed RHA showed major element silicon and im- measurements were 0.2%, 3%, 0.6%, 0.3%, 20 ppm and 120 ppm,
purities Na and Cl (Fig. 2). Silica content of the xerogels respectively.
260 U. Kalapathy et al. / Bioresource Technology 73 (2000) 257±262

Table 2
Mineral content of silica prepared from rice hull ash using various treatments
Minerals Treatments
1 2 3 4 5
Sodium (%) 4.486b 7.120a 6.577a 0.019c 0.021c
Potassium (%) 0.266a 0.133b 0.164b 0.036c 0.039c
Calcium (%) 0.047a 0.037a 0.018b 0.047a 0.019b
Magnesium (%) 0.008a 0.007a 0.002a 0.007a 0.006a
Iron (ppm) 19.8 33.7 14.3 34.6 32.2
Manganese (ppm) <1 <1 <1 <1 <1
Treatment 1 ± Silica was prepared from unwashed RHA. Treatment 2 ± Silica was prepared from RHA washed at pH 7. Treatment 3 ± Silica
was prepared from acid washed RHA. Treatment 4 ± Silica was prepared from unwashed RHA and washed after silica was dried. Treatment 5 ±
Silica was prepared from acid washed RHA and washed after silica was dried.
*
Values are means of triplicate analyses. In each row, values with di€erent superscripts are signi®cantly (P < 0.05) di€erent from each other. Average
standard deviations of Na, K, Ca, and Mg, measurements were 0.5%, 0.1%, 0.01%, and 0.001%, respectively.
**
See the note in Section 3 regarding Fe and Mn measurements.

of the silica. Hence, the following discussion will be

focused on the e€ect of washing on the major impurities
Na, K, and Ca. The initial acid washing did not improve
the purity of silica when compared to the purity of silica
obtained from unwashed RHA and RHA washed at pH
7. Incorporation of an additional washing step after
drying the extracted silica, resulted in a silica product
with even lower mineral (Na, and K) content. Washing
the dried silica (xerogel) with deionized water is more
e€ective than washing the silica gel (aquagel) prior to
drying, in removing minerals from the silica. This could
be due to entrapment of metal ions in the silica gel
network, which may not be leached out by water
washing. Drying silica gel to a very low moisture level
might result in disruption of the silica gel network
structure and would free the trapped mineral ions (Na, Fig. 3. EDX spectrometric spectra of non-carbon materials in RHA.
and K). Sodium and potassium ions might be easily
leached out by washing the dried gel with water. The Ca
content was comparatively lower for silica prepared content of RHA was 39% (determined by ashing at
from acid washed RHA. This may be due to the fact 660o C for 8 h). Since silica is insoluble at pH < 10, highly
that, Ca being the major divalent cation in RHA, will alkaline conditions were necessary for ecient extrac-
require lower pH washing as it is strongly adsorbed on tion of silica from RHA. Yields of silica extracted from
the silica surface at pH > 5 (Iler, 1979). Therefore, to 10 g of RHA using 0.1, 0.25, 0.5 and 1.0 N NaOH are
keep the Ca content in silica at a lower concentration, shown in Table 3. The pH of dispersion of RHA in 0.1,
an initial acid washing might be required. Further, it 0.25, 0.5 and 1.0 N NaOH were 12.0, 12.4, 12.6, and
should be noticed that the silica extraction yields were 12.9, respectively. During the course of extraction, the
similar for unwashed RHA, RHA washed at pH 7 and pH of these dispersions decreased to 9.8, 10.8, 11.0, and
acid washed RHA. These data indicate that ®nal 11.4, respectively at the end of the 1 h extraction period.
washing of xerogel is necessary to keep the overall Silica yield from RHA for 1 N NaOH extraction was
mineral content at a lower concentration. However, if a 91% (see Table 4). Extraction with more concentrated
very low concentration of Ca is required, acid pre-wash NaOH did not improve the silica yield. Therefore, 1 N
of RHA can be used without any adverse e€ect on silica NaOH was used to extract silica from RHA in this
yield. study.

3.2. Silica extraction yield 3.3. Structural features of RHA silica

EDX analysis of RHA non-carbon materials showed As shown in Fig. 4, the broad X-ray di€raction pat-
that silica content of RHA (assuming that all of the Si is tern of RHA silica, which is typical for amorphous
in the form of silica) was about 54% (Fig. 3). Carbon solids (Kamath and Proctor, 1998), con®rmed the ab-
 U. Kalapathy et al. / Bioresource Technology 73 (2000) 257±262 261

Table 3
E€ect of concentration of sodium hydroxide used for silica extraction
on the yield of silica
Extraction Yield of silica (%)
0.1 NaOH Not extracted
0.25 N NaOH 17c
0.5 N NaOH 35b
1.0 N NaOH 91a
*
Values are means of triplicate measurements. Values with di€erent
superscripts are signi®cantly (P < 0.05) di€erent from each other.
Average standard deviation of silica yield was 2.3%.

Table 4
Yield and moisture content of silica from washed and unwashed rice Fig. 5. Fourier transform infrared spectra of silica produced from
hull ash (RHA) RHA.
RHA Yield of silica Moisture content
(%) (%) 4. Conclusions
Acid washed 91a 2.6a
Washed at pH 7 91a 2.4a This study revealed that silica xerogels with 93% silica
Unwashed 89a 2.7a content and minimal mineral contaminants can be
*
Values are means of triplicate measurements. In each column, values produced from RHA using a simple low energy chemical
with similar superscripts are not signi®cantly (P < 0.05) di€erent from method. The silica extraction yield from RHA was 91%.
each other. Average standard deviations of silica yield (%) and mois- The initial acid washing of RHA resulted in a lower Ca
ture contents (%) were 4 and 0.5, respectively.
content in the silica xerogel. The ®nal washing of the
xerogel led to a silica product with lower Na, and K
contents. Incorporation of initial acid washing of RHA
and ®nal washing of the xerogel resulted in a pure silica
with total Na, K, and Ca contents of <0.1%.

Acknowledgements

We wish to thank Producers Rice Mill (Stuttgart,

AR) for providing RHA; and Bruce Cottingham,
(Tyson Foods, Springdale, AR) for his assistance during
this study.

References

Fig. 4. X-ray di€raction pattern of silica produced from RHA.
sence of any ordered crystalline structure. In a previous
investigation, X-ray di€raction studies of raw and
heated RHA, indicated relatively high disordered
structure of silica. Although heating RHA to tempera-
tures over 500o C could convert a less ordered structure
to a more highly ordered structure, the absence of a
highly ordered structure in RHA was attributed to the
thermal insulation properties of the ash (Proctor, 1990).
The major chemical groups present in silica are
identi®ed by the FTIR spectra shown in Fig. 5. The
broad band between 2800 and 3750 cmÿ1 was due to
silanol OH groups and adsorbed water. The predomi-
nant absorbance peak at 1320 cmÿ1 was due to siloxane
bonds (Si±O±Si). The peaks between 1200 and 700 cmÿ1
are attributed to vibration modes of the gel network
(Kamath and Proctor, 1998).
Anon., 1997. Soluble Silicates and their Applications. Cross®eld
Publication, Cross®eld, Warrington, UK, Issue No. 2.
AOAC, 1990. Ocial Methods of Analysis, 15th ed. Association of
Analytical Chemists, Washington, DC.
Brinker, C.J., Scherer, G.W., 1990. Applications. In: Sol±Gel Science,
The Physics and Chemistry of Sol±Gel Processing. Academic Press,
San Diego, CA, pp. 839±880.
Chakraverty, A., Kaleemullah, S., 1991. Conversion of rice husk into
amorphous silica and combustible gas. Energy Consers. Mgmt. 32,
565±570.
Iler, R.K., 1979. Silica gels and powders. In: Iler, R.K. (Ed.), The
Chemistry of Silica. Wiley, New York, pp. 462±599.
Kamath, S.R., Proctor, A., 1998. Silica gel from rice hull ash:
preparation and characterization. Cereal Chemistry 75, 484±487.
Krishnaro, R.V., Godkhindi, M.M., 1992. Distribution of silica in rice
husk and its e€ect on formation of silicon carbide. Ceramics Intl.
18, 243±249.
Laxamana, N.B., 1982. Binders from rice hull ash low-cost housing
materials. Forbride-Dig. College: Forest Products Research and
Industries Development Commission 11, 27±30.
Lender, P.W., Ruiter, R., 1990. Novel inorganic materials and
heterogeneous catalysis: preparation and properties of high surface
262 U. Kalapathy et al. / Bioresource Technology 73 (2000) 257±262
area silicon carbide and silicon oxynitrides. In: Sheats, J.E.,
Carraher, C.E., Pittman, C.U., Eldin, M., Currell, B. (Eds.),
Inorganic and Metal-Containing Polymeric Materials, Plenum
Press, New York, pp. 187±195.
Luan, T.C., Chou, T.C., 1990. Recovery of silica from the
gasi®cation of rice husk/coal in the presence of a pilot ¯ame
in a modi®ed ¯uidized bed. Ind. Eng. Chem. Res. 29, 1922±
1927.
Maeda, E., Komatsu, M., 1996. The thermoelectric performance of
silicon carbide semiconductor made from rice hull. Mater. Res.
Soc. Symp. Proc. 410, 77±82.
View publication stats
Mizuki, E., Okumura, S., Saito, H., Muro, S., 1993. Formation of
silicon carbide from rice husks using enzymatic methods for carbon
control. Biores. Technol. 44, 47±51.
Proctor, A., 1990. X-ray di€raction and scanning electron microscope
studies of processed rice hull silica. J. Am. Oil Chem. Soc. 67,
576±584.
Proctor, A., Clark, P.K., Parker, C.A., 1995. Rice hull ash adsorbent
performance under commercial soy oil bleaching conditions.
J. Am. Oil Chem. Soc. 72, 459±462.
Proctor, A., Palaniappan, S., 1990. Adsorption of soy oil free fatty
acids by rice hull ash. J. Am. Oil Chem. Soc. 67, 15±17.