ARTICLE IN PRESS

Bioresource Technology xxx (2007) xxx–xxx

Preparation and properties of a double-coated slow-release

NPK compound fertilizer with superabsorbent and water-retention
a,b a,* a
Lan Wu , Mingzhu Liu , Rui Liang
a
Department of Chemistry and State Key Laboratory of Applied Organic Chemistry, Lanzhou University, Lanzhou 730000, PR China
b
College of Chemical Engineering, Northwest Minorities University, Lanzhou 730030, PR China

Received 28 September 2006; received in revised form 29 December 2006; accepted 30 December 2006

Abstract

A double-coated slow-release NPK compound fertilizer with superabsorbent and water-retention was prepared by crosslinked poly
(acrylic acid)/diatomite – containing urea (the outer coating), chitosan (the inner coating), and water-soluble granular fertilizer NPK (the
core). The effects of the amount of crosslinker, initiator, degree of neutralization of acrylic acid, initial monomer and diatomite concen-
tration on water absorbency were investigated and optimized. The water absorbency of the product was 75 times its own weight if it was
allowed to swell in tap water at room temperature for 2 h. Atomic absorption spectrophotometer and element analysis results showed
that the product contained 8.47% potassium (shown by K2O), 8.51% phosphorus (shown by P2O5), and 15.77% nitrogen. We also inves-
tigated the water-retention property of the product and the slow release behavior of N, P and K in the product. This product with excel-
lent slow release and water-retention capacity, being nontoxic in soil and environment-friendly, could be especially useful in agricultural
and horticultural applications.
 2007 Elsevier Ltd. All rights reserved.

Keywords: Diatomite; Chitosan; Slow release; Water-retention; NPK compound fertilizer

1. Introduction Chitosan is a highly deacetylated derivative of chitin, one

The growth of plants and their quality are mainly a
function of the quantity of fertilizer and water. So it is very
important to improve the utilization of water resources and
fertilizer nutrients. One method of reducing fertilizer nutri-
ent losses involves the use of slow- or controlled-release fer-
tilizers. There are three types of these fertilizers: slightly
soluble materials, such as urea-formaldehyde; materials
for deep placement, such as urea supergranules; and coated
fertilizers (Jarosiewicz and Tomaszewska, 2003). Coated
fertilizers are physically prepared by coating granules of
conventional fertilizers with various materials that reduce
their dissolution rate. The release and dissolution rates of
water-soluble fertilizers depend on the coating materials.
*
Corresponding author. Tel.: +86 931 8912387; fax: +86 931 8912582.
E-mail address: m-zliu@163.com (M. Liu).
of the most abundant natural and biodegradable polymers.
It has been widely applied in the biomedical, pharmaceuti-
cal, and agricultural fields. In many of these applications
chitosan is extremely attractive due to its biodegradability,
biocompatibility, and nontoxicity (Borzacchiello et al.,
2001). Therefore, the soluble fertilizer coated by chitosan
would be an ideal slow release formulation.
Superabsorbent polymers can absorb a large amount of
water compared with general water absorbing materials in
which the absorbed water is hardly removable even under
some pressure. Because of their excellent characteristics,
superabsorbent polymers had been widely used in agricul-
ture and horticulture (Lokhande and Gotmare, 1999).
Recently, research on the use of superabsorbent polymers
as water managing materials for the renewal of arid and
desert environment has attracted great attention, and
encouraging results have been observed as they can reduce
irrigation water consumption, improve fertilizer retention

0960-8524/$ - see front matter  2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.biortech.2006.12.027

Please cite this article in press as: Wu, L. et al., Preparation and properties of a double-coated slow-release ..., Bioresour. Technol.
(2007), doi:10.1016/j.biortech.2006.12.027
 ARTICLE IN PRESS
2 L. Wu et al. / Bioresource Technology xxx (2007) xxx–xxx
in soil, lower the death rate of plants, and increase plant
growth (Nge et al., 2004; Lokhande and Varadara, 1992).
However, its application in this field has met some prob-
lems because most of these polymeric superabsorbents
are based on pure poly(sodium acrylate), and then they
are too expensive and not suitable for saline-containing
water and soils (Li et al., 2007; Lin et al., 2001). Recently,
the preparation of polymer/clay superabsorbent composite
has attracted great attention because of their relative low
production cost, excellent water-retention and their consid-
erable applications in agriculture and horticulture. Diato-
mite (a kind of clay mineral), as a good substrate for
superabsorbent composite materials, is a kind of light-
weight sedimentary rock composed principally of silica
microfossils of aquatic unicellular algae. Compared with
the layer structure of silicates mentioned above, diatomite
has three-dimensional network configuration, and the void
is up to 80–90%. Besides, there are isolated and hydrogen
boned hydroxyl groups detected on the surface of diato-
mite (Gao et al., 2005). So it has strong hydrophilicity
and can be as one of component of superabsorbents
(Khraisheha et al., 2005).
On the basis of the above background and our previous
studies on superabsorbent polymers (Chen et al., 2005; Ma
et al., 2004) and slow-release fertilizers (Liu et al., 2004;
Guo et al., 2005; Liang and Liu, 2006), we prepared in this
study a double-coated slow-release NPK compound fertil-
izer with superabsorbent and water-retention (DSFSW),
whose inner coating was chitosan (CTS), and the outer
coating was crosslinked poly(acrylic acid)/diatomite-con-
taining urea (PAADU). The product we prepared not only
has slow-release property but also could absorb a large
amount of water and preserve the soil moisture at the same
time. In addition, the outer coating (PAADU) could pro-
tect the inner coating (CTS) from mechanical damage.
These were significant advantages over the normal slow-
release or controlled-release fertilizers, which generally
have only a slow-release property. The results indicated
that the DSFSW could be found an application in agricul-
ture and horticulture, especially in drought-prone regions
where the availability of water is insufficient.
2. Experimental
2.1. Materials
Chitosan (CTS) was kindly provided by Jinxing Bio-
chemical Co., Zhejiang, China. Its viscosity average molec-
ular weight is 6.0 · 105 and its N-deacelylation degree is
90%. Acrylic acid (AA, chemically pure, Beijing Eastern
Chemical Works, Beijing, China) was distilled under
reduced pressure before use. Potassium persulfate (KPS,
analytical grade, Xi’an Chemical Reagent Plant, Xi’an,
China) and N,N 0 -methylene bisacrylamide (NNMBA,
chemically pure, Shanghai Chemical Reagent Factory,
Shanghai, China) were used as purchased. Diatomite, sup-
plied by Shanghai Chemical Reagent Factory, Shanghai,
Please cite this article in press as: Wu, L. et al., Preparation and properties of a double-coated slow-release ..., Bioresour. Technol.
(2007), doi:10.1016/j.biortech.2006.12.027
China, was calcined at 450 C, and the particle sizes less
than 10 lm were about 90 wt%. It contains 14–16 wt% of
calcium sulphate, 0.02 wt% of chloride and the content of
iron is less than 0.05 wt%. The pH of the solution
(10 wt% diatomite) is 6.5–7.5. NPK compound fertilizer
was industrial grade and the others were all analytical
grade. All of the materials used were available from com-
mercial sources.
2.2. Preparation and characterization of PAADU
A series of samples with different amounts of diatomite,
crosslinker, initiator, urea, and AA with different degrees
of neutralization were prepared by the following proce-
dure. Typically, AA (8.4 g) was dissolved in 19 mL distilled
water and then neutralized with a proper amount of ammo-
nia in a four-neck flask equipped with a stirrer, condenser,
thermometer, and nitrogen line. Urea (3.8 g) and diatomite
powder were added to the partially neutralized monomer
solution. Under nitrogen atmosphere, a proper amount
of NNMBA aqueous solution as a crosslinker was added
to the solution, and the mixed solution was stirred on a
water bath at room temperature for 30 min. The water bath
was then heated slowly to 65 C with vigorous stirring after
the radical initiator KPS was charged to the mixed solu-
tion. After 4 h, the reaction was stopped. After it was
cooled, the product was ground. A white powder PAADU
was obtained, which was used as the outer coating mate-
rial. The nitrogen content of PAADU was 19.8% deter-
mined by element analysis.
The ingredients of PAADU were analyzed using a Fou-
rier Transform Infrared (FTIR) spectrophotometer (Amer-
ican Nicolet Crop., model 170-SX). The sample of
PAADU was ground with dried KBr powder. The KBr
disk was dried again and subjected to FTIR
spectrophotometry.
2.3. Preparation of DSFSW
NPK compound fertilizer granules ranging from 2.36 to
2.85 mm in diameter were sieved and used for further
experiments. During the sieving, we tried to remove gran-
ules that had defects on the surface.
The NPK compound fertilizer granule was placed into a
rotary drum, and the CTS powder was stuck on the gran-
ules by means of epoxy dissolved in acetone. The adhesive
was applied by spraying at regular time intervals. The pro-
cess was finished until compact and homogeneous coating
formed on fertilizer granule. The coated granules were
dried to a constant mass at room temperature for 6 h. Then
the CTS-coated NPK compound fertilizer granules were
obtained.
CTS-coated NPK compound fertilizer granules were
dipped in water and then were immediately placed on
PAADU powder and shaken. In this manner, PAADU
could adhere to the surface of CTS-coated NPK compound
fertilizer and form the outer coating. After the granules
 ARTICLE IN PRESS
L. Wu et al. / Bioresource Technology xxx (2007) xxx–xxx
were coated, the surface of the product was crosslinked by
spraying 1% (v/v) methanol solution of epoxy chlorolpro-
pane and then dried in a 70 C oven to obtain the final
product, i.e., the double-coated slow-release NPK com-
pound fertilizer with superabsorbent and water-retention.
2.4. Component analysis of DSFSW
Contents of nitrogen, phosphor, and potassium in the
DSFSW were determined by an element analysis instru-
ment (Germany Elemental Vario EL Crop., model 1106)
and atomic absorption spectrophotometer (American
TJA Crop., model 1100-B).
2.5. Measurement of water absorbency of DSFSW
The accurately weighted DSFSW (about 0.5 g) was
immersed into a certain amount of tap water and allowed
to soak at room temperature for 2 h. The swollen DSFSW
was filtrated through a 80-mesh sieve to remove non-
absorbed water and weighted. The water absorbency was
calculated using the following equation:
WA ¼ ðM  M 0 Þ=M 0
Here M and M0 denote the weight of swollen DSFSW and
the weight of the dry DSFSW, respectively, and WA is the
water absorbency per gram of dried DSFSW.
2.6. Slow release behavior of DSFSW in soil
To study the slow-release behavior of DSFSW in soil, the
following experiment was carried out: 1 g of DSFSW was
well-mixed with 180 g of dry soil (below 2 mm in diameter)
and kept in a 200 mL plastic beaker properly covered and
incubated for different periods at room temperature.
Throughout the experiment, the soil was maintained at
30 wt% water-holding capacity by weighing and adding dis-
tilled water if necessary, periodically. Blank and controlled
experiments, viz., without any fertilizers and with untreated
NPK compound fertilizers (the total contents of N, P, and
K were the same as that of 1 g DSFSW), respectively, were
also carried out. The remaining NPK compound fertilizer
granular and DSFSW in the beakers were picked out and
washed well by distilled water after each incubated period
(day 1, 3, 5, 10, 15, 20, 25 and 30), and then dried at
70 C to a constant weight to be estimated for the contents
of N, P and K. The contents of N, P and K were estimated
by an element analysis instrument and atomic absorption
spectrophotometer, respectively.
2.7. Largest water-holding ratio of the soil
The sandy soil used in this study was representative of
the area of Lanzhou, which lies in the northwest of China
and is a dry and semidesert region. A 2 g sample of DSFSW
was well blended with 200 g of dry soil (below 26-mesh) and
kept in a PVC tube of 4.5 cm diameter; the bottom of the
Please cite this article in press as: Wu, L. et al., Preparation and properties of a double-coated slow-release ..., Bioresour. Technol.
(2007), doi:10.1016/j.biortech.2006.12.027
3
tube was sealed by nylon fabric (with an aperture of
0.076 mm) and weighed (W1). The soil samples were slowly
drenched by tap water from the top of the tube until the
water seeped out from the bottom. The tube was weighed
(W2) again when there was no water seeping out at the bot-
tom. A control experiment, viz., with no DSFSW, was also
carried out. The largest water-holding ratio (W%) of the soil
was calculated using the following equation:
W % ¼ 100  ðW 2  W 1 Þ=ðW 2  W 1 þ 200Þ ð2Þ
2.8. Measurement of the water-retention of DSFSW in soil
A 2 g sample of DSFSW was well-mixed with 200 g of
dry sandy soil (below 2 mm in diameter) and kept in a plas-
tic beaker and then 200 g of tap water was slowly added into
the beaker and weighed (W1). A controlled experiment, i.e.,
without DSFSW, was also carried out. The beakers were
maintained at room temperature and were weighed every
5 days (Wi) over a period of 30 days. The water evaporation
ratio (W%) of soil was calculated using the following
equation:
ð1Þ W % ¼ 100  ðW 1  W i Þ=200 ð3Þ
2.9. Morphology of DSFSW
The granules of DSFSW were also subjected to a scan-
ning electron microscopy (SEM) study. They were split
into two halves, and the fractions obtained were adhered
to sample holders with carbon LIT-C glue. The samples
were metal coated with a layer of gold and observed in a
JSM-5600LV SEM manufactured in Japan.
3. Results and discussion
3.1. The composition and physical properties of DSFSW
The structure of DSFSW fertilizer granule was the three-
layer, which the core is NPK compound fertilizer (57 wt%),
the inner coating is CTS (6.5 wt%) and the outer coating
is PAA/diatomite-containing urea composite (36.5 wt%),
respectively. Element analysis results showed that the
nitrogen content of NPK compound fertilizer, CTS and
PAADU were 14.78, 1.80 and 19.80 wt%, respectively.
Atomic absorption spectrophotometer results showed that
P and K contents of NPK compound fertilizer were 14.90
(shown by P2O5) and 14.86 (shown by K2O) wt%. The
diameters of dry DSFSW, and swollen DSFSW granules
in tap water were in the ranges of 2.52–3.54, and 12.4–
14.6 mm.
3.2. FTIR analysis of PAADU (the outer coating material)
The infrared spectrum of PAADU shows the character-
istic peaks of poly(acrylic acid)/diatomite at 2925 and
2854 cm1 and the characteristic absorption bands of urea
 ARTICLE IN PRESS
4 L. Wu et al. / Bioresource Technology xxx (2007) xxx–xxx
at 1624 and 1462 cm1. The peak observed at 521 cm1
corresponded to the group of N–CO–N, the peak observed
at 799 cm1 corresponded to the group of Si-OH, and the
peak observed at 469 cm1 corresponded to O–Si–O.
Besides, the peaks observed at 3300–3500 cm1 could be
attributed to –OH group of poly(acrylic acid)/diatomite
and –CONH2 group of urea. These results proved that
the PAADU consisted of poly(acrylic acid)/diatomite and
urea.
3.3. Effect of initial AA concentration on water absorbency
The initial monomer concentration is one of the influ-
encing factors that affect the water absorbency of DSFSW.
To study the effect of initial AA concentration (CA, wt%)
on water absorbency, the other reaction variables were
adopted as the optimal levels: diatomite, 8 wt%; neutraliza-
tion degree, 70%; initiator, 0.9 wt%; crosslinker, 0.2 wt%;
reaction temperature, 65 C. The water absorbency in tap
water decreased from 73 to 18 g/g with the increase of
CA from 11 to 38 wt%. It is hard to obtain a composite
gel in the preparation when CA is less then 11 wt% and
water absorbency is hard to measure exactly because of
the excessive water soluble material. It was reported by
Baker et al. (Baker et al., 1992), that the effective crosslink-
ing density is probably less than the nominal crosslinking
density at a low total monomer concentration in the prep-
aration of cationic acylamid-based hydrogel. The result for
PAA/diatomite superabsorbent composite conformed to
the resulting cationic acrylamide-based hydrogel and simi-
lar observations were reported by others (Li and Wang,
2005).
3.4. Effect of neutralization degree of AA on water
absorbency
Neutralization degree was defined as the molar percent-
age of carboxyls in AA neutralized by ammonia (reaction
conditions: initial AA concentration, 11 wt%; diatomite,
8 wt%; initiator, 0.9 wt%; crosslinker, 0.2 wt%; reaction
temperature, 65 C). The water absorbency increased at
first and then decreased gradually with the increase of neu-
tralization degree and reached maximum at neutralization
degree = 70%. It was expected that the activity of acrylic
acid is higher than acrylate, when the neutralization degree
was lower than 70%, the lower neutralization degree was
faster than the rate of the polymerization, the content of
the oligomers would increase, and the content of acrylate
would decrease, so the soluble part increased and the elec-
trostatic repulsion between the attached carboxylate anions
decreased, so that the stretching extend of hydrogel net-
work decreased, and this resulted in water absorbency
increasing with the increase of neutralization degree. It
was expected that the activity of acrylic acid is higher than
acrylate so the lower the neutralization degree was, the fas-
ter the polymerization rate was, and the high polymeriza-
tion rate would result the increase of the content of
Please cite this article in press as: Wu, L. et al., Preparation and properties of a double-coated slow-release ..., Bioresour. Technol.
(2007), doi:10.1016/j.biortech.2006.12.027
oligomers (the soluble part) in DSFSW. At the same time,
the charge density of the network would increase with the
increase of neutralization degree, and this would result
the increase of the stretching extend of the hydrogel net-
work, so it is easy to understand the increase of water
absorbency as the neutralization degree increased from
0% to 70%.
3.5. Effect of the amount of diatomite on water absorbency
We examined the effect of the amount of diatomite on
the water absorbency in tap water, while keeping the other
reaction variables on the above mentioned conditions: ini-
tial AA concentration, 11 wt%; neutralization degree, 70%;
initiator, 0.9 wt%; crosslinker, 0.2 wt%; reaction tempera-
ture, 65 C. It is obvious that the amount of diatomite is
an important factor affecting water absorbency. The water
absorbency increases with an increase of diatomite till
reaching a maximum at 8 wt%. However, when the amount
of diatomite is larger than 8 wt%, the water absorbency
decreased with the increase of the amount of diatomite.
The behavior may be attributed to the reaction between
PAA and diatomite. When the amount of diatomite is
low, the reaction could improve the polymeric network,
as a result, the water absorbency increases. On the other
hand, when the amount of diatomite is high, the additional
diatomite results in the generation of more crosslinking
points that increase the crosslinking density of the compos-
ite and then the elasticity of polymer chains decrease.
Moreover, the content of hydrophilic groups is lower at
higher the amount of diatomite, and then the osmotic pres-
sure difference decreases, which results in the shrinkage of
the PAA/diatomite. The results are in conformity with
Flory’s network theory (Flory, 1953) and similar observa-
tion has been reported by others (Lee and Yeh, 1997).
3.6. Effect of the amount of initiator on water absorbency
The effect of the amount of initiator on water absor-
bency of DSFSW was investigated while the other variables
were kept on the above mentioned conditions: initial AA
concentration, 11 wt%; neutralization degree, 70%; diato-
mite, 8 wt%; crosslinker, 0.2 wt%; reaction temperature,
65 C. Evidently, water absorbency increased with the
increase of the amount of initiator at first and then
decreased. It was expected that the decomposition velocity
of the initiator was slow when the amount of initiator was
small, and this resulted in the slow initiator and polymeri-
zation reaction; the residual monomers in the product
increased during the same reaction time, so water absor-
bency was low. But the collision between monomer free
radicals would increase with the increase of the amount
of initiator, and so the rate of polymerization would be
increased with that. A high rate of polymerization would
result the increase of the content of oligomers in DSFSW,
viz., the soluble parts in DSFSW increased. So water
 ARTICLE IN PRESS

L. Wu et al. / Bioresource Technology xxx (2007) xxx–xxx 5

absorbency decreased with a further increase of the amount 100 a

of initiator. b

3.7. Effect of the amount of crosslinker on water absorbency 80

c

Nitrogen release (%)

Crosslinking density is an extremely important swelling 60
control element. Relatively small changes in crosslinking
density can play a major role in modifying the properties
of superabsorbent polymers. The amount of crosslinker 40

dependence of water absorbency of DSFSW in tap water

was also investigated (reaction conditions: initial AA con- 20
centration, 11 wt%; neutralization degree, 70%; diatomite,
8 wt%; initiator, 0.9 wt%; reaction temperature, 65 C).
As the amount of crosslinker increased, water absorbency 0
0 5 10 15 20 25 30
increased up to a certain level, but with higher amount of
crosslinker water absorbency decreased drastically. As we Time (day)
known, the amount of crosslinker determined the cross- Fig. 1. Release behaviors of nitrogen in soil. (a) Untreated NPK fertilizer;
linking density of the hydrogel network. When the amount (b) mixture of untreated NPK fertilizer and PAA; (c) DSFSW.
of crosslinker was below 0.2 wt%, water absorbency
increased with the increase of the amount of crosslinker.
This could be attributed to the lower crosslinking density 100 a
with a smaller amount of crosslinker, which would cause b
the solubility of DSFSW to increase. When the amount 80
Phosphor release (%)

of crosslinker was higher than 0.2 wt%, a further increase c

of the crosslinking density of the hydrogel network would
decrease the space among the grids of the three-dimen- 60

sional network. This would make it more difficult for the

network to be swollen by water, which was responsible
40
for the decrease of water absorbency. These results were
also in accordance with the theory of Flory (1953).
20
3.8. Slow release behavior of DSFSW in soil
0
One of the most important characteristics of the 0 5 10 15 20 25 30
DSFSW we prepared was its slow- release property. Figs. Time (day)
1–3 show the slow release behaviors of N, P, and K in
Fig. 2. Release behaviors of phosphorus in soil. (a) Untreated NPK
untreated NPK fertilizer, a mixture of untreated NPK fer-
fertilizer; (b) mixture of untreated NPK fertilizer and PAA; (c) DSFSW.
tilizer and superabsorbent polymer, and DSFSW in soil,
respectively. As shown in Figs. 1–3, the N, P, and K in
untreated NPK fertilizers had released out more than 100 a
80% and 87% within 2 and 5 days (Figs. 1a, 2a and 3a), b
respectively, while the release rate of the untreated NPK
Potassium release (%)

fertilizer mixed with superabsorbent polymers (Figs. 1b, 80

2b and 3b) decreased compared with the untreated NPK c

fertilizers, in agreement with the results of Smith and Har- 60
rison (Smith and Harrison, 1991). However, the release
rate was higher than that of DSFSW. As shown in
Fig. 1c, the N in DSFSW released 11.9%, 17.8%, and 40

73.2% by the 3rd, 5th, and 30th days. As shown in

Fig. 2c, the P in DSFSW released 3.9%, 8.2%, and 64%
20
by the 3rd, 5th, and 30th days. As shown in Fig. 3c, the
K in DSFSW released 2.3%, 6.4%, and 68% by the 3rd,
5th, and 30th days. The releases of the sum of nutrients 0
in DSFSW did not exceed 15 wt% by the 3rd day. After 0 5 10 15 20 25 30

30 days, about 75% of the sum of nutrients was released. Time (day)
These results indicated that the slow-release properties Fig. 3. Release behaviors of potassium in soil. (a) Untreated NPK
of DSFSW conformed to the standard of slow-release fertilizer; (b) mixture of untreated NPK fertilizer and PAA; (c) DSFSW.

Please cite this article in press as: Wu, L. et al., Preparation and properties of a double-coated slow-release ..., Bioresour. Technol.
(2007), doi:10.1016/j.biortech.2006.12.027
 ARTICLE IN PRESS
6 L. Wu et al. / Bioresource Technology xxx (2007) xxx–xxx
fertilizers of the Committee of European Normalization
(CEN) (Trenkel, 1997). It is well-known that water-soluble
NPK compound fertilizer is easily dissolved in water, so it
will quickly dissolve in the soil solution after being added
to the soil, and the nutrient will be quickly exhausted.
While the PAA superabsorbent can absorb a lot of water
in soil, the NPK compound fertilizer dissolved in soil solu-
tion can be absorbed into it and on its surface. The
absorbed NPK compound fertilizer can be released or des-
orbed slowly through the exchange of free water or miner-
als between soil solution and PAA. Therefore, the mixture
of PAA and NPK compound fertilizer had a slower release
rate than untreated NPK compound fertilizer.
The fertilizer nutrient release mechanism of DSFSW can
be described by the following steps: (1) the water in soil
penetrates the outer PAADU layer in the first stage. (2)
The outer PAADU layer is slowly swollen by the water
in soil and transforms to hydrogel. A dynamic exchange
between the free water in the hydrogel and the water in soil
will develop (David and Mark, 1994; Smyth et al., 1998).
(3) The urea dissolved in the hydrogel network will slowly
diffuse out of the granule through this dynamic water
exchange. (4) When the free water in the PAADU layer
migrates to the middle layer, i.e. CTS coating, the water
will form pores in the initial stage and biodegradation in
the last stage (Ge et al., 2002). (5) Water enters into the
NPK compound fertilizer core through the tiny pores
and dissolves it. (6) The dissolved NPK compound fertil-
izer diffuses out the CTS layer and enters into the PAADU
layer, and then releases into the soil through the dynamic
exchange of free water. Therefore nutrients in the core
must pass through the swelling of the PAADU coating
and the biodegradation of the CTS coating before it can
enter into solution. So, the existence of outer coating regu-
lated the release behavior of urea and minimized the burst
effect. Meanwhile, the urea in the outer coating can release
slowly during the lag time of core and meet the nutrient
requirement of plants during this time.
3.9. The largest water-holding ratio of the soil
The experiment results of the largest water-holding ratio
of the soil indicated that the largest water-holding ratio of
the soil without DSFSW was 29.08%, and that of the soil
with DSFSW was 39.76% (the mass ratio of DSFSW to soil
was 1:100), 10.68% higher than the former. This showed
that the DSFSW we prepared still had excellent water
absorbency in soil, could obviously improve the water-
holding capacity of the soil, and could efficiently store
rainwater or irrigation water resources. This was one of
the significant advantages over the normal slow-release
fertilizers.
3.10. Water-retention behavior of DSFSW in soil
Besides its slow-release property, another one of the
most important characters of the DSFSW we prepared is
Please cite this article in press as: Wu, L. et al., Preparation and properties of a double-coated slow-release ..., Bioresour. Technol.
(2007), doi:10.1016/j.biortech.2006.12.027
100
a
Water evaporation ratio (%)
80 b
60
40
20
0
0 5 10 15 20 25 30
Time (day)
Fig. 4. Water-retention behaviors of DSFSW. (a) Water + dry soil; (b)
water + dry soil + DSFSW.
its water-retention capacity or, in other words, its effective
utilization of water in arid and desert regions to transform
them into ‘‘green and fertile lands’’. In this regard the pre-
liminary testing of DSFSW for water-retention was also
carried out, and the result is shown in Fig. 4. From
Fig. 4 we find that the water transpiration rate of soil with
DSFSW obviously decreased compared to the soil without
it. The water transpiration ratio of soil without DSFSW
reached 56.0 and 94.5 wt% on the 15th and 30th days,
respectively, while that of the soil with DSFSW was 43.5
and 78.0 wt%, respectively. The time needed for 50.0 wt%
water evaporation was 13.4 days for the soil without
DSFSW, while it was 16.9 days for the soil with DSFSW.
After 30 days, the water content of the soil without
DSFSW was 5.5 wt%, while that of the soil with DSFSW
was 22.0 wt%. Thus, DSFSW had good water-retention
capacity in soil, and that with DSFSW use water can be
saved and managed so that they can be effectively used
for the growth of plants. These results showed that the
DSFSW we prepared had excellent water absorbency,
water-retention, and moisture preservation capacity
besides its slow-release property. These were the significant
advantages over the normal slow-release or controlled-
release fertilizers, which always only had a slow- or con-
trolled-release property. The reason was that the superab-
sorbent polymer (PAA/diatomite) in the outer coating of
DSFSW could absorb and store a large quantity of the
water in soil, and allow the water absorbed in it to be
slowly released with the decrease of the soil moisture.
The swollen DSFSW was just like an additional nutrient
reservoir for the plant-soil system. Consequently, it pro-
longed irrigation cycles, reduced irrigation frequencies,
and strengthened the ability of plants to fight against
drought. It was also noted that lots of granular structures
could from in the soil with DSFSW use. It has been
reported in the literature (Krysiak and Madigan, 2004) that
these granular structures contributed to stabilization of the
 ARTICLE IN PRESS
L. Wu et al. / Bioresource Technology xxx (2007) xxx–xxx
soil structure and improved aeration, permeability, and till
ability of the soil, reduced soil packing and cracking, min-
imization of soil crusting, and thus prevention of the soil
from hardening, providing a favorable environment for
crops to grow. Furthermore, as we known, the Chitosan
in the first layer of the coating material was a biodegrad-
able material (Borzacchiello et al., 2001) and the introduc-
tion of diatomite into poly(acrylic acid) renders the outer
coating biodegradable and environmental friendly (Li
et al., 2007). Therefore, the coating materials were not
harmful to the soil.
3.11. Morphology of DSFSW
From the SEM of the surface of the DSFSW, it can be
seen that the surface of DSFSW is coarse and porous,
which structurally increased the surface area of the
DSFSW. Therefore, when DSFSW is dipped in water, it
can absorb water quickly to form a swollen hydrogel,
which is responsible for the water-retention property of
DSFSW. The SEM of the cross-section of the DSFSW
showed the three-layer structure of DFSW. The rough
outer layer was PAADU, which could absorb a large
amount of water. The middle layer was CTS, which could
reduce the rate of water diffusion into the core and the
nutrient diffusion outside the core, this provided the
DSFSW with a good slow-release property. The inner core
was a water-soluble NPK compound fertilizer granule.
Thus, after the fertilizer core is dissolved by water, it must
pass through two layers, i.e., CTS coating and PAADU
coating, to release into the soil. This provides the DSFSW
with a good slow-release property. In summary, the outer
PAADU layer enables the water-retention property, and
the middle CTS layer enables the DSFSW slow-release
property.
4. Conclusion
A double-coated slow-release NPK compound fertilizer
with superabsorbent and water-retention (DSFSW) was
prepared. The core of it was water-soluble NPK compound
fertilizer granule, the inner coating was chitosan, and the
outer coating was crosslinked poly(acrylic acid)/diato-
mite-containing urea composite. The product was charac-
terized by FTIR, element analysis and atomic absorption
spectrophotometer, and the results showed that is con-
tained 15.77% nitrogen, 8.51% phosphorus (shown by
P2O5), and 8.47% potassium (shown by K2O). The water-
holding capacity experiment showed that the largest
water-holding ratio was about 10.68% higher than that of
without it, and the water-retention experiment showed that
it still had excellent water-retention property in soil. The
results of the slow release experiment showed that the
release ratio of the effective nutrient in it was not above
75% on the 30th day. This new approach showed promis-
ing in utilizing natural resource such as chitosan and diat-
omite in the production of coating material, which could
Please cite this article in press as: Wu, L. et al., Preparation and properties of a double-coated slow-release ..., Bioresour. Technol.
(2007), doi:10.1016/j.biortech.2006.12.027
7
significantly reduce the production cost and make the tech-
nique quite environmental friendly. Also, the product with
good slow release property and excellent water-retention
could be especially useful in agricultural and horticultural
applications.
Acknowledgement
This work was supported by Special Doctorial Program
Funds of the Ministry of Education of China (Grant No.
20030730013).
References
Baker, J.P., Hong, L.H., Blanch, H.W., Pransnitz, J.M., 1992. Effect of
initial total monomer concentration on the swelling behavior of
cationic acrylamide-based hydrogels. Macromolecules 27, 1446–
1454.
Borzacchiello, A., Ambrosio, L., Netti, P.A., et al., 2001. Chitosan-based
hydrogels: synthesis and characterization. Journal of Materials
Science-Materials in Medicine 12, 861–864.
Chen, Z.B., Liu, M.Z., Ma, S.M., 2005. Synthesis and modification of salt-
resistant superabsorbent polymers. Reactive and Functional Polymers
62, 85–92.
David, S.G.H., Mark, T.S.L., 1994. Water–polymer interactions and
critical phenomena of swelling in inhomogeneous poly-(acrylonitrile-
acrylamide-acrylic acid) gels. Polymer 35, 4416–4422.
Flory, P.J., 1953. Principles of Polymer Chemistry. Cornell University
Press, New York.
Gao, B.G., Jiang, P.F., An, F.Q., Zhao, S.Y., Ge, Z., 2005. Studies on the
surface modification of diatomite with polyethyleneimine and trapping
effect of the modified diatomite for phenol. Applied Surface Science
250, 273–279.
Ge, J.J., Wu, R., Shi, X.H., et al., 2002. Biodegradable polyurethane
materials from bark and starch. II. Coating material for controlled-
release fertilizer. Journal of Applied Polymer Science 86, 2948–
2952.
Guo, M.Y., Liu, M.Z., Zhan, F.L., et al., 2005. Preparation and
properties of a slow-release membrane-encapsulated urea fertilizer
with superabsorbent and moisture preservation. Industrial and Engi-
neering Chemistry Research 44, 4206–4211.
Jarosiewicz, A., Tomaszewska, M., 2003. Controlled-release NPK fertil-
izer encapsulated by polymeric membranes. Journal of Agricultural
and Food Chemistry 51, 413–417.
Khraisheha, M.A.M., Al-Ghoutib, M.A., Allenb, S.J., Ahmadb, M.N.,
2005. Effect of OH and silanol groups in the removal of dyes from
aqueous solution using diatomite. Water Research 39, 922–932.
Krysiak, M.D., Madigan, D.P., 2004. Granular fertilizer. US Patent
0,069,032.
Lee, W.F., Yeh, P.L., 1997. Superabsorbent polymeric materials III. Effect
of initial total monomer concentration on the swelling behavior of
crosslinked poly(sodium acrylate) in aqueous salt solution. Journal of
Applied Polymer Science 64, 2371–2380.
Li, A., Wang, A.Q., 2005. Synthesis and properties of clay-based
superabsorbent composite. European Polymer Journal 41, 1630–
1637.
Li, A., Zhang, J.P., Wang, A.Q., 2007. Utilization of starch and clay for
the preparation of superabsorbent composite. Bioresource Technology
98, 327–332.
Liang, R., Liu, M.Z., 2006. Preparation and properties of a double-coated
slow-release and water-retention urea fertilizer. Journal of Agricultural
and Food Chemistry 54, 1392–1398.
Lin, J.M., Wu, J.H., Yang, Z.F., et al., 2001. Synthesis and properties of
poly(acrylic acid)/mcia superabsorbent nanocomposite. Macromolec-
ular Rapid Communication 22, 422–424.
 ARTICLE IN PRESS

8 L. Wu et al. / Bioresource Technology xxx (2007) xxx–xxx

Liu, M.Z., Zhan, F.L., Wu, L., et al., 2004. Preparation of slow release
urea fertilizer with preservation of soil moisture. Journal of Polymer
Materials 21, 213–220.
Lokhande, H.T., Gotmare, V.D., 1999. Utilization of textile loomwaste as
a highly absorbent polymer through graft copolymerization. Biore-
source Technology 68, 283–286.
Lokhande, H.T., Varadara, P.V., 1992. A new Guargum-based superab-
sorbent polymer synthesized using gamma radiation as a soil additive.
Bioresource Technology 42, 119–122.
Ma, S.M., Liu, M.Z., Chen, Z.B., 2004. Preparation and properties of a
salt-resistant superabsorbent polymer. Journal of Applied Polymer
Science 93, 2532–2541.
Nge, T., Hori, N., Takemura, A., et al., 2004. Swelling behavior of
chitosan/poly(acrylic acid) complex. Journal of Applied Polymer
Science 92, 2930–2940.
Smith, J.D., Harrison, H.C., 1991. Evaluation of polymers for controlled-
release fertilize properties when incorporated with nitrogen fertilizer.
Communications in Soil Science and Plant Analysis 22, 559–564.
Smyth, G., Francis, X.Q., Vincent, J.M., 1998. Water in hydrogels. 2. A
study of water in poly(hdrooxyethl methacrylate). Macromolecules 21,
3198–3204.
Trenkel, M.E., 1997. International Fertilizer Industry Association,
Stratospheric Ozone. HMSO, London.

Please cite this article in press as: Wu, L. et al., Preparation and properties of a double-coated slow-release ..., Bioresour. Technol.
(2007), doi:10.1016/j.biortech.2006.12.027