ARTICLE IN PRESS

Water Research 38 (2004) 10621068

The use of nutshell carbons in drinking water lters for

removal of trace metals
Mohamed Ahmednaa, Wayne E. Marshallb, Abdo A. Husseinyc,*, Ramu M. Raod,
Ipek Goktepea
a
Department of Human Environment and Family Sciences, North Carolina A & T State University, Greensboro, NC, USA
b
Commodity Utilization Research Unit, USDA-ARS,Southern Regional Research Center, New Orleans, LA, USA
c
Technology International Inc., LaPlace, LA 70068, USA
d
Department of Food Science, Louisiana State University, Baton Rouge, LA, USA
Received 19 May 2003; received in revised form 23 September 2003; accepted 29 October 2003

Abstract

Filtration of drinking water by point-of-use (POU) or point-of-entry (POE) systems is becoming increasingly popular
in the United States. Drinking water is ltered to remove both organic and inorganic contaminants. The objective of
this study was to evaluate the use of granular activated carbon from nutshells (almond, English walnut, pecan) in a
POU water ltration system to determine its effectiveness in removing select, potentially toxic metal ions, namely,
copper (Cu2+), lead (Pb2+) or zinc (Zn2+) found in drinking water. The nutshell-based carbon system was designated
Envirolter and was compared to four commercial POU systems with brand names of BRITA, Omni Filter, PUR %
and Teledyne Water Pik. Eight prototype Envirolters, consisting of individual or binary mixtures of carbons made
from acid-activated almond or pecan shells and steam-activated pecan or walnut shells were constructed and evaluated
for adsorption of the three metal ions. The results indicated that a binary mixture of carbons from acid-activated
almond and either steam-activated pecan or walnut shells were the most effective in removing these metals from
drinking water of all the POU systems evaluated. Binary mixtures of acid-activated almond shell-based carbon with
either steam-activated pecan shell- or walnut shell-based carbon removed nearly 100% of lead ion, 9095% of copper
ion and 8090% of zinc ion. Overall the performance data on the Envirolters suggest that these prototypes require
less carbon than commercial lters to achieve the same metal adsorption efciency and may also be a less expensive
product.
r 2003 Elsevier Ltd. All rights reserved.

Keywords: Granular activated carbon; Drinking water ltration; Point-of-use water lter; Pecan shell; Almond shell; English walnut
shell; Copper; Lead; Zinc

1. Introduction annually on bottled water and home water treatment

Despite the assurance offered by numerous regulatory
agencies that the domestic drinking water in the United
States is safe, many consumers are wary of tap water
and question its quality and health effects [1]. As a
result, American households spend over $5 billion
*Corresponding author. Tel.: +1-985-652-1127; fax: +1-
985-652-1196.
E-mail address: dr abdo@rtconline.com (A.A. Husseiny).
systems. The concern of American consumers about the
safety of their drinking water is warranted in many
instances, especially where the water is supplied in old
iron pipes systems, where lead is used for soldering
faucets and other connections. The leaching of lead from
these lead-based solders is thought to be one of the
major causes for the occasional ndings of lead in
drinking water. Lead heads the list of environmental
threats because, even at extremely low concentrations,
lead has been shown to cause brain damage in children

0043-1354/$ - see front matter r 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.watres.2003.10.047
 ARTICLE IN PRESS
M. Ahmedna et al. / Water Research 38 (2004) 10621068
[2]. Unfortunately, according to the Centers for Disease
Control and Prevention, nearly 30 million Americans
consume drinking water containing lead and copper in
excess of the standard recommended by the US EPA [2].
Additionally, many tap water sources, especially in
remote rural areas, have undesirable and sometimes
harmful color, odor, and hardness (from high zinc
concentrations) that prompt consumers to seek ways to
improve their tap water quality in their homes by
various ltration devices.
Currently, a variety of water treatment devises are
available in the market to the homeowner. These devises
differ in the types of lter media used, chemicals to be
removed, location within the home, and ease of
operation and cost per unit of lter medium and
maintenance and operating costs. Some of the materials
used in the lter media of these devices include bone
carbon or coal-based activated carbons, ion exchange
resins and reverse osmosis lters. However, lter systems
with ion exchange resin or activated carbon are the most
widely used. Each of these two systems has been shown
to be ineffective in completely removing all known
hazardous contaminants. Consequently, hybrid systems
have been introduced into the commercial markets. The
hybrid systems contain predominantly ion exchange
resin with the balance consisting of granular activated
carbon. Hybrid systems have been shown to be effective
in removing most inorganic contaminants in water.
However, their unit cost is high. For instance, ion
exchange resin will cost $5$20 per lb, depending on the
quality. Commercial granular activated carbon, on the
other hand, cost about $1.50 per lb but it is less effective
than ion exchange resin in removing inorganic consti-
tuents such as heavy metals [3].
Granular activated carbon can be prepared from a
large variety of carbon-containing feedstock by pyro-
lysis into char and subsequent physical or chemical
activation of the char. Almost any lignocellulosic
material, in some cases with the assistance of a binder,
can be converted to granular activated carbon. The nal
porous structure and properties of activated carbon are,
however, a function of the precursor used in its
preparation, the activation procedure, and the extent
of activation [4]. Activated carbon surfaces have a pore
size that determine its adsorption capacity, a chemical
structure that inuences its interaction with polar and
nonpolar adsorbates, and active sites which determine
the type of chemical reactions with other molecules [5];
all can be custom designed by selecting the right
precursor and choosing the pyrolysis and activation
conditions. Thus, high volume agricultural by-products
are lignocellulosic wastes that may offer an inexpensive
and renewable source of specialty granular activated
carbons [3,610]. This option is particularly attractive
since the yearly harvest and processing of various
agricultural crops grown in the United States yields
1063
considerable quantities of agricultural by-products.
Conversion of such plentiful by-products into activated
carbons that can be used in applications such as
drinking water purication or waste treatments would
add value to agricultural commodities, help reduce the
cost of waste disposal and provide a potentially
inexpensive replacement of existing commercial adsor-
bents from nonrenewable sources.
The objectives of this study were to create Envir-
olters from phosphoric acid-activated almond and
pecan shells and steam-activated pecan and English
walnut shells and compare the efcacy and estimated
cost of Envirolters to the efcacy and estimated cost
of commercial drinking water ltration systems.
2. Materials and methods
2.1. Materials
Four commercial drinking water ltration systems
were used in this study and were purchased at local retail
outlets. Omni Filter consists of 100% coal-based, steam-
activated carbon. It is produced by OMNI Industries,
Mulvane, KS, for the removal of drinking water
contaminants. Teledyne Water Pik is also composed of
100% coal-based, steam-activated carbon. This lter is
manufactured by Waterpik Technologies Inc., Fort
Collins, CO and is used specically for tap water
% water lter cartridges for their pitcher
ltration. PUR
lter system contain predominantly cation exchange
resin and some coal-based, granular activated carbon.
% cartridges are made by PUR,
PUR % Minneapolis, MN,
specically for tap water ltration. According to the
manufacturer, they are particularly efcient at removing
lead from water. BRITA water ltration cartridges are
similar to the PUR% cartridges, in that they consist of
predominantly cation exchange resin with some coal-
based, granular activated carbon and are used in pitcher
ltration systems. They are manufactured by Brita
Products Company, Oakland, CA and are effective at
lead removal, according to the companys literature.
Nutshell-based granular activated carbons that com-
prise the Envirolter were obtained from the South-
ern Regional Research Center, New Orleans, LA.
Phosphoric acid-activated almond and pecan shells were
produced by the method of Toles et al. [11]. Steam-
activated pecan and English walnut shells were devel-
oped using the method of Johns et al. [12].
2.2. Methods
2.2.1. Determination of physical and chemical properties
of nutshell-based granular activated carbons
Physical properties of the granular activated carbons
include percent yield and bulk density and chemical
 ARTICLE IN PRESS
1064 M. Ahmedna et al. / Water Research 38 (2004) 10621068

properties include pH and percent ash. Surface proper-
ties consist of surface area, micro-, meso- and macro-
pore volume and negative surface charge.
Percent yield, surface area and meso- and micropore
volume were determined following the methods of Toles
et al. [13] and Ahmedna et al. [7]. Surface charge was
assessed by the titration method described by Toles et al.
[13]. Bulk density, pH, and percent ash were determined
according to procedures described by Ahmedna et al. [6].
2.2.2. Preparation of envirofilter and commercial
filtration systems
Envirolter ltration systems were prepared ac-
cording to the descriptions provided in Table 1. In order
to keep the size and shape of the Envirolters similar
to the commercial point-of-use (POU) lters, product
containers manufactured by BRITA were used to
construct Envirolters and the four commercial
lters. The empty BRITA containers were lled with
nutshell-based granular activated carbons, either singly
or in binary mixtures while the reference product
containers were lled with the commercial products.
Care was taken to ensure that all lters had the same
column depth and packing level. However, since the
densities of the cartridge contents were different for each
adsorbent, different weights of adsorbent were present
in some cartridges as shown in Table 1. The amount
and type of adsorbent in each lter cartridge is given in
Table 1.
2.2.3. Determination of adsorption properties of the
envirofilter and commercial filtration systems
The adsorption properties of Envirolters and
commercial lters were compared for their ability to
adsorb Cu2+, Pb2+ and Zn2+ in simulated drinking
water according to the following procedure.
Ultra clean water was obtained by passing tap water
through a Millipore ELIX III water purication system
(Millipore, Bedford, MA) to remove all organics and
inorganic contaminants. Known concentrations of
metals as copper chloride, lead nitrate and zinc chloride
were added to a specic volume of contaminant-free
water obtained as described above. The nal metal
concentrations in the simulated drinking water were
3.0 ppm Cu2+, 0.150 ppm Pb2+ and 10.0 ppm Zn2+.
These metal ion concentrations were selected to
match those listed in the performance data sheet for
BRITA lters and the concentrations were above the
EPA primary standard drinking water limits of 1.3,
0.015, and 5.0 ppm for copper, lead and zinc ion,
respectively.
Test lters were thoroughly cleaned using contami-
nant-free water, then rinsed with 20% nitric acid prior to
use. Test water containing known concentrations of
metals was passed through the lters in two steps. First,
500 mL of test water were ltered through the cartridges
in a pour-through pitcher (gravity ow) and the ltrate
was discarded to minimize contamination. After the rst
run, 500 mL of the test water were passed through the
same lter and the ltrate was collected and evaluated
for metal ion concentrations. Metal ion concentrations
were evaluated using US EPA standard method 200.7
[14].
The lters metal removal efciency was expressed as
the amount of water contaminant removed. To normal-
ize the efciency data, the amount of water contaminant
removed per unit weight of adsorbent was calculated.
The experimental design was a completely rando-
mized design in which 12 lter types received equal
volumes of the same test water solutions. The levels of
lead, copper and zinc removed from water solution were
measured as dependent variables. Metal uptake data
was statistically analyzed by multivariate analysis of
variance (MANOVA) using SAS [15]. Tukeys multiple
comparison test was used to compare mean metal
uptakes by lters within each metal ion category.

Table 1
Filter identication and lter contents

Filter identication Filter contents Adsorbent in lter (g)

Envirolter I 100% acid-activated pecan shells (PA) 93

Envirolter II 100% steam-activated pecan shells (PS) 62
Envirolter III 100% acid-activated almond shells (AA) 73
Envirolter IV 100% steam-activated English walnut shells (WS) 63
Envirolter V 50% PA and 50% PS 70
Envirolter VI 50% PA and 50% WS 70
Envirolter VII 50% AA and 50% PS 60
Envirolter VIII 50% AA and 50% WS 60
Omni lter 100% coal-based activated carbon 76
Teledyne water Pik 100% coal-based activated carbon 74
%
PUR Mixture of cation exchange resin and coal-based activated carbon 100
BRITA Mixture of cation exchange resin and coal-based activated carbon 96
 ARTICLE IN PRESS
M. Ahmedna et al. / Water Research 38 (2004) 10621068 1065

Difference between metal uptakes of the test lters was
judged signicant at a 0:05 signicance level.
3. Results and discussion
3.1. Physical and chemical properties of nutshell-based
granular activated carbons
Physical, chemical and surface properties of the
nutshell-based carbons are presented in Tables 2
and 3. The method of activation had an effect on most
of the carbon properties regardless of the nutshell used
as starting material. Acid activation resulted in higher
yield, lower surface area, higher percentage of micropore
volume, lower percentage of meso-pore plus macro-pore
volume, higher bulk density, lower pH and greater
negative surface charge compared to steam activation.
The only property that appeared to be independent of
activation method was percent ash, which is usually
precursor dependent and not linked to activation
method. Therefore, choice of activation method pro-
duced considerably different carbons in terms of their
physical, chemical properties and surface properties.
Since the pH of the acid activated carbons were
considerably lower than for the other adsorbents, both
experimental and commercial, the pH was adjusted to 7
prior to their use in the water ltration experiments.
Some of the carbons physical and chemical properties
may not directly relate to a carbons effectiveness in
water purication but they are important to their
commercial utilization. For instance, percent yield is a
factor used in cost estimation and commercial potential
of activated carbons while bulk density and pH are
parameters that affect potential commercial use of
activated carbon in terms of surface area and negative
surface charge.
A carbons surface properties, however, may play
signicant roles in the utilization of nutshell-based
carbons in the ltration of drinking water. High surface
area may allow increased adsorption of metal ions if the
internal carbon surfaces can accommodate some ad-
sorption of the positively charged metals. High negative
surface charge should be more accommodating to
positively charged metal ions due to ionic attractive
forces. Surface charge of opposite adsorbate polarity
may play a major role in determining the suitability of
activated carbon in specic water ltration needs
because the electrical charge of the surface groups will
enhance adsorption of the target molecules. For
example, a negatively charged carbon surface would
attract positively charged pesticide residues and metal
ions such as copper, lead and zinc. While surface area
determines the extent of internal space available for
adsorption of target compounds, both the size and
distribution of micro-pores, meso-pores and macro-
pores determine the adsorptive properties of acti-
vated carbons. For instance, small pore size will not
trap large adsorbate molecules and large pores may not
be able to retain small adsorbates, whether they are
charged, polar molecules or uncharged, non-polar
compounds.

Table 2
Physical and chemical properties of nutshell-based activated carbonsa

Agricultural by-product Activation agent Yield (%) Bulk density (g/mL) pHb Ash (%)

Pecan shells Phosphoric acid 32 0.51 3.2 (6.8) 2.1

Pecan shells Steam 18 0.40 7.8 10.4
Almond shells Phosphoric acid 24 0.42 3.3 (7.1) 7.9
English walnut shells Steam 11 0.34 6.7 4.3
a
Values are the means of duplicate determinations where the standard deviations are less than 10% of the mean values.
b
Values in parentheses are adjusted pH values. The pH of the carbon was adjusted to pH 7 using 0.1 N NaOH before its use in the
water ltration experiments.

Table 3
Surface properties of nutshell-based activated carbonsa

Agricultural by- Activation agent Surface area Micropores Mesopores plus Surface charge
product (m2/g) (%) macropores (%) (mmol H+/g)

Pecan shells Phosphoric acid 682 92.7 7.3 2.43

Pecan shells Steam 724 71.4 28.6 0.23
Almond shells Phosphoric acid 708 86.6 11.4 2.46
English walnut shells Steam 1060 81.5 18.5 0.25
a
Values are the means of duplicate determinations where the standard deviations are less than 10% of the mean values.
 ARTICLE IN PRESS
1066 M. Ahmedna et al. / Water Research 38 (2004) 10621068

Table 4
Amount of metal ion removed per unit mass of adsorbent

Filter identication Adsorbent in Lead removed Copper removed Zinc removed

cartridge (g) (ppm/g) (ppm/g) (ppm/g)

Envirolter I 93 0.0018b 0.0316ab 0.0743ab

Envirolter II 62 0.0027ab 0.0420a 0.0270b
Envirolter III 73 0.0024ab 0.0450a 0.0996a
Envirolter IV 63 0.0008c 0.0141b 0.0070c
Envirolter V 70 0.0024ab 0.0421a 0.0877a
Envirolter VI 70 0.0023ab 0.0408a 0.0872a
Envirolter VII 60 0.0030a 0.0494a 0.0981a
Envirolter VIII 60 0.0030a 0.0525a 0.1082a
Omni Filter 76 0.0022ab 0.0410a 0.0596ab
Teledyne water Pik 74 0.0024ab 0.0435a 0.0750ab
%
PUR 100 0.0018b 0.0327ab 0.0666bc
BRITA 96 0.0019b 0.0320ab 0.0635bc
Values are the means of duplicate determinations where the standard deviations are less than 10% of the mean values. Column means
with different letters are signicantly different from each other based on Tukeys multiple comparison test with a 0:05:
See Table 1 for lter contents.

3.2. Removal of metal ions by envirofilters and
commercial filtration systems
Table 4 compares the metal removal efciencies of
eight Envirolters and four commercial lters con-
taining coal-based activated carbons or cation exchange
resin/activated carbon mixtures. The removal efcien-
cies are corrected for the mass of adsorbent in the lter
and are given as the amount of metal ion removed per
unit weight of adsorbent.
The data in Table 4 indicate that Envirolters VII
and VIII were the most effective adsorbents in removing
copper, lead or zinc ions from simulated drinking water.
Both of these Envirolters are binary systems
comprised of 50% phosphoric acid-activated almond
shell carbons and 50% steam-activated pecan or walnut
shell carbons. As an individual Envirolter 100%
acid-activated almond shells (III) performed well,
especially in the removal of zinc ion. This system
removed more zinc ion per g of adsorbent than any of
the commercial lters. In addition, the binary systems
present in Envirolters V and VI removed more zinc
ion than the commercial lters.
The data in Table 4 also show that copper and lead
ion were effectively removed by Envirolters II, III, V
and V, but only Envirolters VII and VIII removed
signicantly more lead and copper than the four
commercial lters. In terms of total amount of metal
ion removed, only two ltration systems, Envirolters
II and IV, did not reduce zinc ion levels below the US
EPA mandated level of 5.0 ppm. Envirolter IV also
failed to remove copper ion concentrations to below
1.3 ppm, the US EPA mandated limit. Lead ion was
particularly difcult to reduce below 0.015 ppm. Only
three Envirolters accomplished the task, namely, III,
VII and VIII. Three of the four commercial ltration
systems, the exception was Omni Filter, were able to
reduce lead concentrations below 0.015 ppm.
The relative success of Envirolters VII and VIII
may be attributed to a combination of high surface
charge exhibited by the phosphoric acid-activated
almond shell carbons and well-developed meso- and
macroporosity and high surface area exhibited by both
steam-activated pecan and walnut shell carbons. The
presence of high surface area and well-developed meso-
and macroporosity would allow faster diffusion of liquid
into the pore structure of the carbon and a greater
opportunity for ionic and physical adsorption of metals
to take place. The effect appears synergistic since
Envirolter IV (100% walnut shell-based carbon)
poorly adsorbed any of the three metal ions and
Envirolter II (100% pecan shell-based carbon)
poorly adsorbed zinc ion.
Binary mixtures of carbons consisting of acid
activated almond shells with either steam-activated
pecan or walnut shells removed close to 100% of lead
ion, 9095% of copper ion and 8090% of zinc ion. The
excellent metal uptake exhibited by these binary carbon
mixtures is highly signicant since their steam-activated
portion has low surface charge and high surface area
and could, therefore, be favorable for adsorption of
organics or chlorination by-products from water. This
may enable an Envirolter to remove organic and
inorganic contaminants with 100% agricultural by-
product-based activated carbons. A properly con-
structed Envirolter mimics the PUR % and BRITA
ltration systems in that acid-activated nutshell carbons
will assume the role of a cation exchange resin while
steam-activated nutshell carbons will assume the role of
steam-activated, coal-based carbons used in these
 ARTICLE IN PRESS
M. Ahmedna et al. / Water Research 38 (2004) 10621068 1067

commercially available water lters made from mixtures
of cation exchange resin and activated carbon.
3.3. Estimated product cost for envirofilter and
commercial POU water filtration systems used in this
study
Table 5 estimates the product cost for an Envir-
olter comprised of 50% acid-activated almond shell
carbon and either 50% steam-activated pecan shell or
walnut shell carbon. The estimated cost consists of raw
material costs (nutshells), carbon production costs,
which are based on estimated costs from Toles et al.
[16,17] for both acid-activated and steam-activated
nutshell carbons, plastic lter manufacturing costs based
on an estimate given by a plastics manufacturer and
miscellaneous costs, that include advertising, and
Table 5
Estimated production cost to manufacture Envirolter VII
or VIII
transportation costs. We estimate the total cost to be
$8.71/kg of carbon.
A cost comparison between the Envirolter and
two commercial POU water ltration systems are given
in Table 6. BRITA and PUR % were selected because they
are the two most popular brands of drinking water
ltration systems in the United States. Two costs are
generated. First, is the estimated cost per kg of
adsorbent, where a production cost is listed for
Envirolter and retail costs are given for the
commercial products. Second, is the estimated cost per
pitcher type lter unit. In both cases, the cost of the
Envirolter is considerably less than the retail prices
for the commercial systems. This cost differential would
allow an Envirolter manufacture to adjust the
wholesale and perhaps the retail price of the product
over a wide range. In fact, the manufacturer could sell
Envirolters at a retail cost similar to the commercial
units, since they appear to be more efcient at metal ion
removal on a per unit weight basis than the potential
competition.

Item Cost ($ per kg 4. Conclusions

of carbon)

Nutshells 0.11a The performance data obtained from the various

Granular activated carbon production 2.00b Envirolters show that Envirolters used less
Plastic lter cartridge production 6.30c adsorbent than commercial lters to achieve the same
Miscellaneous costs 0.30d metal ion adsorption efciency. Additionally, this super-
Total production cost 8.71 ior adsorption efciency per unit weight comes at a
a
Nutshell cost is based on a mixture of 50% acid-activated potentially cheaper price on the retail market. Effective-
almond shells and 50% steam-activated pecan shells or walnut ness and useful life of any water lters require accessible
shells. Almond shells cost $0.07/kg carbon based on a 30% surface area and high surface charge for ionic adsorp-
yield and pecan or walnut shells cost $0.15/kg carbon based on tion. These characteristics determine the nal amount of
a 15% yield. adsorbent per lter and ultimately the lter production
b
Carbon production cost is based on the cost to acid activate cost. The data suggest that the carbon mixtures used in
almond shells ($2.45/kg carbon) or steam activate nutshells the Envirolters examined in this study possessed
($1.54/kg carbon) which is derived from information provided
adequate surface area and suitable surface charge. The
by Toles et al. [16,17].
c
Cost based on information supplied by a manufacturer of
results of this investigation indicate that as prototype
plastic cartridges and assuming there are nine pitcher lters per lters, Envirolters have the potential to serve as
kg of carbon, since each pitcher lter contains 0.11 kg of alternates to commercial home lters. Moreover, the use
adsorbent. One lter cartridge was estimated to cost $0.70. of nut shell-based carbon lters have other advantages
d
These costs include advertising and shipping costs to retail such as they can be manufactured from renewable
outlets. material found in great abundance, the raw materials do

Table 6
Production and retail cost comparisons between Envirolter VII or VIII and the two leading commercial water ltration systems

Filtration system Cost ($)/kg of adsorbent Cost ($)/POU lter

a
Envirolter 8.71 0.97b
#
PUR 77.94b 8.66c
BRITA 71.91b 7.99c
a
Based on cost data provided in Table 5.
b
Based on 0.11 kg of adsorbent per POU lter. Therefore, 1 kg of adsorbent would ll nine lter cartridges.
c
Retail cost of each pitcher lter.
 ARTICLE IN PRESS
1068 M. Ahmedna et al. / Water Research 38 (2004) 10621068
not have to be imported, and the use of nutshells offer
solution to environmental problems and help the agri-
cultural economy with an additional market potential.
Acknowledgements
This research was funded by an USDA, CSREES,
Small Business Innovation Research Phase I
Award#2001-33610-20383 awarded to Technology In-
ternational Incorporated of Virginia, LaPlace, LA.
References
[1] Gelt J. Consumers increasingly use bottled water, home
water treatment systems to avoid direct tap water. Arroyo
1996;9(1):19.
[2] Lewis SA. Trouble on tap: if were so technologically
advanced, why cant we safely drink water from our
faucets, Sierra Club, 1995. www.exemplar. net/water/
technology.html.
[3] Marshall WE, Ahmedna M, Rao RM, Johns MM.
Granular activated carbons from sugarcane bagasse:
production and uses. Int Sugar J 2000;102:14751.
[4] Gergova K, Petrov N, Eser S. Adsorption properties and
microstructure of activated carbons produced from by-
products by stream pyrolysis. Carbon 1994;32:693702.
[5] Bansal RC, Donnet JB, Stoeckli F. Active carbon. New
York, NY: Marcel Decker; 1988.
[6] Ahmedna M, Johns MM, Clarke SJ, Marshall WE, Rao
RM. Potential of agricultural by-product-based activated
carbons for use in raw sugar decolorization. J Sci Food
Agric 1997;75:11724.
[7] Ahmedna M, Marshall WE, Rao RM. Surface properties
of by-product-based activate carbons and their effect on
raw sugar decolorization. Bioresource Technol 2000;71:
10312.
[8] Ahmedna M, Marshall WE, Rao RM. Production of
granular activated carbons from select agricultural by-
products and evaluation of their physical, chemical and
adsorption properties. Bioresource Technol 2000;71:
11323.
[9] Pendyal B, Johns MM, Marshall WE, Ahmedna M, Rao
RM. The effect of binders and agricultural by-products on
physical and chemical properties of granular activated
carbons. Bioresource Technol 1999;68:24754.
[10] Pendyal B, Johns MM, Marshall WE, Ahmedna M, Rao
RM. Removal of sugar colorants by granular activated
carbons made from binders and agricultural by-products.
Bioresource Technol 1999;68:4551.
[11] Toles CA, Marshall WE, Johns MM. Phosphoric acid
activation of nutshells for metals and organic remediation:
process optimization. J Chem Technol Biotechnol 1998;72:
25563.
[12] Johns MM, Marshall WE, Toles CA. The effect of
activation method on the properties of pecan shell-
activated carbon. J Chem Technol Biotechnol 1999;74:
103744.
[13] Toles CA, Marshall WE, Johns MM. Surface functional
groups on acid-activated nutshell carbons. Carbon
1999;37:120714.
[14] US EPA. Methods for the determination of metals in
environmental samples. Supplement I, Report#EPA/600/
R-94/111. Method 200.7: determination of metals and
trace elements in water and wastes by inductively coupled
plasma-atomic emission spectrometry, 1994.
[15] SAS, 2000. Release 8.10 ed. Cary, NC: SAS Institute Inc.
[16] Toles CA, Marshall WE, Johns MM, Wartelle LH,
McAloon A. Acid-activated carbons from almond shells:
physical, chemical and adsorptive properties and estimated
cost of production. Bioresource Technol 2000;71:8792.
[17] Toles CA, Marshall WE, Wartelle LH, McAloon A.
Steam- or carbon dioxide-activated carbons from almond
shells: physical, chemical, and adsorptive properties and
estimated cost of production. Bioresource Technol
2000;75:197203.