Chemical Engineering Journal 287 (2016) 390–397

Contents lists available at ScienceDirect

Chemical Engineering Journal

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

Slow release fertilizers based on urea/urea–formaldehyde polymer

nanocomposites
Cintia F. Yamamoto a, Elaine I. Pereira b, Luiz H.C. Mattoso b, Teruo Matsunaka c, Caue Ribeiro b,⇑
a
Federal University of São Carlos, Department of Materials Engineering, Washington Luiz Highway, km 235, São Carlos, SP 13565-905, Brazil
b
EMBRAPA Instrumentation, 1452 XV de Novembro Street, São Carlos, SP 13560-970, Brazil
c
Rakuno Gakuen University, Department of Dairy Science, Ebetsu, Hokkaido 0698501, Japan

h i g h l i g h t s g r a p h i c a l a b s t r a c t

Urea–formaldehyde in the presence

of MMT is effective to obtain
fertilizer.

Fertilizer nanocomposites were
produced by cold plastic extrusion.

Fertilizer nanocomposites presented
good mechanical resistance.

Efficient results were found in water-
solubility and soil incubation tests.

Intermediate urea–paraformaldehyde
molar ratio showed most efficient
results.

a r t i c l e i n f o a b s t r a c t

Article history: A novel nanocomposite material based on the exfoliation of montmorillonite into a matrix of urea/
Received 12 August 2015 urea–formaldehyde polymer was developed to be used as nitrogen-loaded slow release fertilizers.
Received in revised form 23 October 2015 The nanocomposites were produced by cold plastic extrusion, a simple and high-productive method,
Accepted 8 November 2015
using a formaldehyde precursor to accomplish polymerization in situ during extrusion processing.
Characterizations showed that the extruded fertilizer nanocomposites presented good mechanical
resistance, and that the urea release was noticeably controlled by the extent of polymerization. Water
Keywords:
immersion experiments demonstrated that the nanocomposite structure was responsible for the effective
Slow release fertilizer
Nanocomposite
slow urea release behavior. A dependence of the availability of other nitrogenous compounds (NH+4 and
Urea NO 3 ) on the polymerization degree was also demonstrated by soil incubation tests. The urea/

Urea–formaldehyde polymer urea–formaldehyde polymer nanocomposites were found to be versatile and smart materials capable
of supplying the enormous demand for novel efficient nitrogen fertilizers.
Ó 2015 Elsevier B.V. All rights reserved.

1. Introduction important commercially available nitrogenous fertilizers [1]. Its

The growing use of fertilizers has enabled high gains of produc-
tivity in agriculture. However, the growing price of these inputs
has led to an increasing cost in the crops production. From an
economic point of view, urea (CO(NH2)2) is one of the most
⇑ Corresponding author. Tel.: +55 16 2107 2915; fax: +55 16 2107 2903.
E-mail address: caue.ribeiro@embrapa.br (C. Ribeiro).
application at a pristine form to soil involves significant losses of
nitrogen through volatilization of NH3 to atmosphere and emission
of greenhouse gases (GHG) [2].
In recent decades, the enormous demand for fertilizers and the
urgent need for a more efficient use of these materials, have led to
development of various types of so-called controlled or slow
release fertilizers (SRFs). SRFs is a class of fertilizers whose funda-
mental role is to release nutrients to the soil for a period of time

http://dx.doi.org/10.1016/j.cej.2015.11.023
1385-8947/Ó 2015 Elsevier B.V. All rights reserved.
 C.F. Yamamoto et al. / Chemical Engineering Journal 287 (2016) 390–397
longer than a conventional fertilizer so that nutrient availability for
absorption by plants is substantially prolonged. In an earlier report,
we have developed a nanocomposite based on dispersion of
montmorillonite (MMT) within a urea matrix, and this hybrid
material exhibited slower dissolution compared with pristine urea,
which allowed a better control of the fertilizer release due to a
hindered diffusion of urea through the nanocomposite structure
[3]. The MMT/Urea nanocomposites presented good potential for
practical uses, since it was comprised of small granules, thereby
being suitable for application as solid material on soil. However,
these nitrogenous nanocomposites presented low mechanical
strength and poorly understood relationship between clay con-
tent and final nitrogen release behavior, which have motivated fur-
ther research in order to obtain SRF materials with optimized
performances.
Urea–formaldehyde (UF) polymer has attracted special interest
among other products tested for controlled release of nitrogen to
soil. UF-based products belong to the first group of materials espe-
cially developed for slow release of nitrogen [1]. It has been shown
that agricultural use of this organic polymer occurs at stages
similar to those adopted for urea [4–7]. Based on patent search,
it was identified some related documents, such as WO2013090287-
A1 [8], WO2011137393-A1 [9], WO2010093462-A1 [10],
US2008141747-A1 [11], US2006196241-A1 [12], US2005144997-A1
[13]. Some of these patents comprise liquid compositions made up
of urea formaldehyde resins, as for example the US2006196241-A1
document, which relates to a fertilizer product corresponding to
a liquid composition where urea formaldehyde is one of the possi-
ble components; other patents describe formulations of particulate
urea formaldehyde polymers (such as the WO2010093462-A1
document); and a third group of patents relates to granulate fertil-
izers produced by the mixture of different components in urea
formaldehyde resin, which are solidified to obtain the final
material (US2008141747-A1 document, for instance). Despite
these products have some features similar to our proposed
material, the main difference lies on the application of a liquid,
pre-formulated urea formaldehyde resin. However, the prepara-
tion of UF granules at large scale is generally complex because it
involves reactions in liquid medium where the control of polymer-
ization is essentially done by the molar ratio between components.
Formaldehyde solutions, commonly used as a precursor, contain
70–73% of water (by weight), which hinders their processing
by methods such as extrusion. Furthermore, the low solubility
of UF polymer has led to extremely low rate of nitrogen release
in several practical applications. Therefore, it is highly preferable
the development of fertilizer products with complex architectures
comprised of partially polymerized urea, urea-rich domains
limited by structural barriers and easily available urea. In this
sense, the relative contents of UF/unreacted urea could be used
to adjust the final profile of urea release for a specific agronomical
application, thereby resulting in an intelligent fertilizer. To the best
of our knowledge, composite materials based on UF/unreacted
urea have not been supplied to the fertilizer market yet.
Hence, we report in this paper a strategy to prepare nanocom-
posite fertilizers based on urea–formaldehyde polymer (UF)/unre-
acted urea and montmorillonite (MMT) exfoliated at nanoscale
using a formaldehyde precursor (paraformaldehyde) to polymerize
UF in situ. This hybrid nanocomposite fertilizer was prepared
through cold plastic extrusion, a simple and high-productive
method, where MMT not only acted as a processing agent (plasti-
cizer), but also as a structural component and diffusional barrier
against urea dissolution. Our results show that this intelligent
material, due to its easy processing and versatile polymerization
control, may be a promising strategy to produce complex granules
with adjusted dissolution, and possibly to be used as a carrier of
other nutrients such as K+ and PO3 4 .
391
2. Materials and methods
2.1. Materials
The raw materials used in the nanocomposite formulations
were: urea AP (Synth), paraformaldehyde (Sigma–Aldrich) and
montmorillonite clay (Bentonita, Drescon S/A – Produtos de
Perfuração). Urea was previously milled to a mesh grain size of
300 mm in a TE-330 hammer mill (Tecnal, Brazil). All the other
materials were used as received.
2.2. Preparation of nanocomposites
Nanocomposites were prepared with a fixed montmorillonite:
urea (M/U) weight ratio of 1:1, and at different urea:paraformalde-
hyde (U/Pf) molar ratios of 1:0.25, 1:0.5, and 1:1. Urea–formalde-
hyde polymer is formed by reaction of formaldehyde with an
excess of urea with conditions of pH, temperature, molar ratio,
reaction time, etc., well controlled [1]. This reaction basically
consists in a stepwise mixing of urea and formaldehyde with
subsequent release of water. Urea suffers hydroxymethylation
under both acidic and alkaline conditions, and the formation of
urea–formaldehyde (UF) occurs by condensation of these hydrox-
ymethylated compounds. Paraformaldehyde is the polymer of
formaldehyde with a degree of polymerization between 8 and
100. This polymer depolymerizes into formaldehyde in presence
of water and heating [14]. Thus, the use of paraformaldehyde
allows minor addition of water to the MMT/Urea formulations,
leading to a better control of plasticity in the extrusion process.
The nanocomposites were prepared through the following
steps: pre-mixing, extrusion, shaping and curing. The precursor
materials (montmorillonite, urea and paraformaldehyde) and
water (15 wt.%) were pre-homogenized and then extruded at
35 °C in a counter-rotating double screw extruder (Coperion
Werner and Pfleiderer) running at 60 rpm. Part of the extruded
mass was pelletized and another part was molded into standard-
ized specimens to conduct mechanical test. The optimum amount
of water used in the preparation of nanocomposites, to achieve
adequate plasticity, was beforehand determined by torque
rheometry. After shaping, the samples were cured at 80 °C in an oven
for 12 h, and subsequently stored at 90 °C until complete drying. The
nomenclature given to each composition is listed in Table 1.
2.3. Characterizations
2.3.1. X-ray diffraction (XRD)
X-ray diffractograms were registered using a XRD 6000 diffrac-
tometer (Shimadzu, Japan), in order to verify occurrence of mont-
morillonite exfoliation due to the shear processing. Relative
intensity of diffraction was registered in the angular range (2h) of
3–40°, using a Cu Ka incident beam (k = 0.1546 nm) and scanning
speed of 1° min1. The voltage and current of the X-ray tube were
30 kV and 30 mA, respectively. The basal spacing (d) of montmoril-
lonite was calculated using the Bragg equation, n  k = 2  d  sin h,
where n (= 1) is the reflection order and h angle of diffraction.
Table 1
Nanocomposites and their nomenclatures. The raw materials used were identified by
their initial letters: M – montmorillonite, U – urea and Pf – paraformaldehyde.
Nomenclature Mass ration M:U Molar ration U:Pf
M/U 1:1 1:1 –
M/U/Pf 1:1:0.25 1:1 1:0.25
M/U/Pf 1:1:0.5 1:1 1:0.5
M/U/Pf 1:1:1 1:1 1:1
392 C.F. Yamamoto et al. / Chemical Engineering Journal 287 (2016) 390–397

2.3.2. Thermal analysis
For thermogravimetry (TG), samples were heated from room
temperature to 800 °C under nitrogen atmosphere flowing at
60 ml min1 using a Q500 thermal analyzer (TA Instruments,
USA). For differential scanning calorimetry (DSC), samples were
heated from 25 to 300 °C under nitrogen flow of 60.0 ml min1
on a DSC Q100 (TA Instruments, USA). The linear heating rate used
in both analyses was 10 °C min1.
2.3.3. Mechanical test: Three point flexural test
Three points flexural test was used to measure the mechanical
resistance of the samples at room temperature. The plastic paste
samples obtained by extrusion were re-processed on a piston
extruder in order to produce cylindrical specimens with diameter
of 9 mm and length of 100 mm, as protocoled by the ASTM C158.
Mechanical tests were performed with a speed of 0.03 mm min1
on a DL-3000 universal testing machine (EMIC, Brazil) equipped
with a 3000 kgf load cell. The flexural strength or rupture modulus
was calculated by Eq. (1):
16Pa
rrf ¼
pD3
where D is the specimen diameter, P is the total force applied, a is
the moment of the arm (in the case of three point flexural tests, a
is the distance between the point of force application to the axis
of rotation. The results were obtained from an average of twenty
specimens.
The modulus of elasticity or Young’s modulus (E), which is the
constant of proportionality between the stress (r) and the strain
(e), was estimated as the ratio of Hooke’s law [15] in Eq. (2):
r ¼ Ee
2.3.6. Determination of total inorganic N release in incubated soil
The release of total inorganic N was carried out using a Nopporo
soil (Haplic Gley Upland Soil) taken from the surface layer
(0–20 cm depth) of an experimental farm at Rakuno Gakuen
University (Japan). After removal of vegetation and roots, the soil
was allowed to dry at room temperature, sieved through a 2 mm
sieve, and submitted to chemical analyses (Table 2). Nanocompos-
ite samples were then milled into powders (including urea), and
mixed to this reference soil, which was also tested with a witness
sample in order to evaluate the influence of microorganisms
naturally present in the raw soil (not fertilized).
Two parallel experiments of soil incubation were performed
[19]. In the first experiment, the soil samples were incubated at
30 °C and the humidity was adjusted to 60%, 70%, and 80% depend-
ing on the soil water holding capacity. In the second experiment,
the soil samples were incubated under humidity of 60%, and the
temperature of incubation was varied in 25 °C, 30 °C, and 35 °C.
Three replicates were prepared for each treatment condition and
the moisture was adjusted daily. The nitrogen amount used in
the incubation tests was 10 mg of N per 20 g of soil placed in
ð1Þ 100 ml glass flasks. The mass of each composition used in the
experiment was determined with basis on the total N content of
the samples (Kjeldahl method – ASTM International) [17]. Specific
ammonium (NH+4) or nitrate (NO 3 ) content was determined
through agitated extraction with 1 M KCl solution for 30 min.
The soil suspension was filtered, and NO +
3 and NH4 contents were
analyzed colorimetrically by a flow injection method (FIAstar
5000, FOSS). The apparent release rate was calculated as follows
(Eqs. 3–6):

mg  nh
a  i  o
ppm
NNHþ or NNO3 ¼  d  100 mlðKClÞ þ CRAsoil
4 blottle 1000
ð2Þ ð3Þ

2.3.4. X-ray micro-tomography (XMT) TIN ¼ NNHþ þ NNO3 ð4Þ

4

Imaging by X-ray micro-tomography was performed by placing

samples into a rotating steel bracket of a microtomography instru- DðTINÞ
RR ¼ ð5Þ
ment 1172 (SkyScan). In the image acquisition process, the follow- input N
ing parameters were adopted: no filter, spatial resolution (voxel
size) of 7.9 lm, rotation step of 0.4°, 180° rotation, and 12 frames DðTINÞ ¼ TINtreatment  TINcontrol ð6Þ
to averaging process. The image reconstruction process of the
where appm is the result (to NH+4
or NO
3)
obtained in the flow injec-
tomographic sections was done by using the NRecon software
tion, d is the dilution, CRAsoil is the water content of soil, RR is the
(SkyScan) in which the following parameters were used: smooth-
apparent release rate, and TIN is the total inorganic nitrogen
ing – 5; ring artifact correction – 5, and beam hardening correction
content.
of 60%.
Comparison of the treatments was performed by analysis of
variance (ANOVA) and Duncan’s multiple comparison tests at a
2.3.5. Evaluation of urea release in water significance level of 5%. The software used in the analysis was R,
The kinetic of urea release in water at 25 °C and natural pH for version 3.1.3.
each nanocomposite was examined following a method adapted
from Tomaszewska and Jarosiewicz [16]. Before testing, the total
3. Results and discussion
N content of each nanocomposite formulation was determined
by the Kjeldahl method (ASTM International) [17]. These results
Fig. 1 shows XRD patterns of the nanocomposite materials with
were used to calculate the amount of material used in the release
different compositions. The pattern of neat MMT in Fig. 1a
tests. Briefly, samples were soaked into water at 25 °C, and were
evidences a typical (0 0 1) basal reflection which corresponds to a
constantly homogenized by gentle stirring using an orbital shaker.
d-value of 12.8 Å for pristine MMT. It can also be evidenced that
The recipients were sealed with plastic film to reduce water loss by
this basal reflection does not overlap with any other reflections
evaporation. Aliquot parts of 1 ml were collected and centrifuged
(15 min at 14,000 rpm, MiniSpin Plus) at different periods of time Table 2
up to 4 days. Pure urea was also tested as a control experiment. Chemical properties of Haplic Gley Upland Soil used in the incubation experiment.
The urea concentration in solution was determined by UV–Vis
pH N-total C-total Pa available CEC Exchangeable
spectrophotometry in a 1601PC spectrophotometer (Shimadzu, (H2O) (mg kg1) cations (mg kg1)
Japan) according to the method reported by With et al. [18].
(g kg1) (cmol(+) K+ Ca2+ Mg2+
Profiles of percentage of urea solubilized as a function of time
kg1)
were then obtained. The measurements were replicated three
5.98 3.2 41.8 69 21.1 178 760 58
times under identical laboratorial conditions for each type of
a
nanocomposite formulation. Truog method.
 C.F. Yamamoto et al. / Chemical Engineering Journal 287 (2016) 390–397 393

(a) (b)
1500
Montmorillonite
M/U/Pf 1:1:1

1200

M/U/Pf 1:1:0.5
Intensity

Urea M/U/Pf 1:1:0.25

Intensity / a.u.
900

Montmorillonite
600

300

Paraformaldehyde

0
4 6 8 10 3 4 5 6 7 8 9 10
2θ / degree 2θ / degrees

Fig. 1. XRD patterns as a detailed view of the 2h = 3–10° angular range for neat (a) montmorillonite, urea, and paraformaldehyde; (b) extruded formulations of
montmorillonite/urea/paraformaldehyde at different molar ratios.

of paraformaldehyde or urea. This is an important aspect for the
nanocomposites because any superposition involving the (0 0 1)
peak could be interpreted as changes in the interlamellar space
of MMT. Interlamellar expansion of MMT is clearly seen in the
patterns of the MMT/Urea/paraformaldehyde formulations,
Fig. 1b. One can notice that the basal reflection (0 0 1) was shifted
to lower angles (higher d-values). According to Bragg law, the inter-
lamelar spacing was increased to about 17.7 Å in all the samples.
These results suggest that the extrusion process was efficient in
intercalating urea molecules into the MMT lamellar structure, and
therefore, producing nanocomposites with intercalated morphology.
Thermal analyses were used to evaluate the polymerization
step. Fig. 2 presents TG and DTG curves for all studied materials.
Mass losses of urea and paraformaldehyde occurred mainly at
low temperatures, such behaviors are characteristic of these
compounds. By DTG analyses in Fig. 2b, it was observed that
paraformaldehyde shows a single mass loss step at 130 °C,
which is related to a direct volatilization process. Urea presents
four degradation stages with the principal mass loss at 210 °C
(between 105 and 225 °C). For MMT mass loss steps ascribed
to non-constitutional water (20–100 °C), structural hydroxyls
(200–400 °C) and strongly bounded water and organic matter

100 (a) (b)

M/U/Pf 1:1:1
Montmorillonite
90 M/U/Pf 1:1:1 M/U/Pf 1:1:0.5
M/U/Pf 1:1:0.5
80
M/U/Pf 1:1:0.25
M/U/Pf 1:1:0.25
70 M/U 1:1
M/U 1:1
60
dm/dT
Weight / %

Montmorillonite
50

40 Urea

30 Paraformaldehyde

20
Urea
10

0 Paraformaldehyde

100 200 300 400 500 600 700 800 100 200 300 400 500 600
o o
Temperature / C Temperature / C

Fig. 2. (a) TG curves and (b) DTG curves of MMT/urea/paraformaldehyde formulations and their neat precursors.
394 C.F. Yamamoto et al. / Chemical Engineering Journal 287 (2016) 390–397

(400–600 °C) were observed [20]. It was also observed that the
degradation of MMT forms a large amount of residual mass
(ffi90%) up to 600 °C. In the non-polymerized sample (M/U 1:1),
the principal mass loss stage of urea is still present. This was
expected since urea is unreacted in that condition. However, in
the partially and fully polymerized MMT/Urea/paraformaldehyde
samples the urea mass loss stage was displaced towards higher
temperatures, approximately 250 °C. For the partially polymerized
MMT/Urea/paraformaldehyde samples (1:1:0.5 and 1:1:0.25) both
stages were observed, suggesting the presence of a urea–formalde-
hyde polymer and unreacted urea. The thermal profile of the
sample M/U/Pf 1:1:1 shows only one degradation peak in the
DTG curve, related to the presence of urea–formaldehyde polymer
[21]. It is important to notice that there was no considerable
amount of formaldehyde in all nanocomposites samples, as indi-
cated by the total suppression of the peaks at 135 °C in the TG/
DTG curves. This was also consistent with the expected pathway
of polymerization, since formaldehyde has a very high reactivity
with urea. The residual mass presented by these materials is
closely related to the MMT content.
DSC experiments confirmed the thermal behavior of the MMT/
Urea/paraformaldehyde nanocomposites. Fig. 3 shows melting
(135 °C) and degradation (192 °C) of urea, as well as the depolymer
ization/volatilization (144 °C) of paraformaldehyde. For the M/U/Pf
1:1:1 nanocomposite (which is supposed to be a fully polymerized
material), only a thermal event at 246 °C is observed, confirming
complete polymerization of urea in this material. For the interme-
diate samples, peaks allusive to urea degradation were still pre-
sent, but it is important to notice that any peak related to
paraformaldehyde was absent, which is indicative of total conver-
sion to formaldehyde during the curing process. The temperatures
values assigned to the thermal events detected by TG and DSC
have been summarized in Table 3.
Considering that the polymerization would increase the
mechanical resistance of the M/U/Pf nanocomposites in compar-
ison with pristine urea, the flexural mechanical test was performed
to indicate the extent of polymerization. Also, it is worth mention-
ing the key aspect of mechanical resistance on practical application
of fertilizers in field. Brittle materials tend to be powdered during
application, which is not desirable due to segregation, machine
damage and also occupational aspects related to workers responsi-
ble for handling the fertilizer. On the other hand, a much plastic
material may impose damages to machines typically used in the
application of fertilizers by obstruction or agglomeration. Results
of 3-point flexural mechanical tests for some selected samples
are reported in Table 4. One can notice that the polymerization
step greatly influenced the mechanical properties, since an
increase in elastic modulus was observed to occur with increasing
polymer content (supposed to be proportional to the
paraformaldehyde content) of the samples. However, a decrease
in rupture modulus was observed for M/U/Pf 1:1:1, probably due
to the presence of internal defects which strongly influence on
the maximum load supported by hard/brittle materials. In fact, it
is greatly possible that the material polymerized at the highest
extent contains a greater number of critical defects because the
curing process may cause some structural contraction, or because
of anisometric shrinkage during the drying step. Despite the differ-
ence observed among samples, it is noticed the high deviation
characteristic of fragile materials (especially for M/U/Pf 1:1:1,
which is visually brittle and, consequently, tends to suffer catas-
trophic fractures – see Fig. 4).
In order to examine the influence of the curing process on the
nanocomposite microstructures, a non-destructive analysis was
proposed by X-ray tomography. In this case, a full 3D image of
the internal structure of the samples was generated, which was
useful to analyze the porosity and phase segregation resulting from
the extrusion process. Typical cross-sectional views shown in
Fig. 4, revealed that the polymerization process influenced on the
formation of large porosity and cracks. The non-polymerized sam-
ple (M/U 1:1) (Fig 4a) showed a homogeneous microstructure with
some regular porosity maybe related to moisture through the dry-
ing process. The M/U/Pf 1:1:0.25 sample presented similar
microstructure, however, the samples with ratio 1:1:0.5 and
1:1:1 showed large cracks and porosity. As previously discussed,
these cracks are probably originated from structural contraction
caused by drying, which cause premature mechanical failure in a

Table 3
Summary of most significant thermal events observed on the TG and DSC
experiments of urea, paraformaldehyde and their nanocomposites with MMT.

Sample Thermal Description

events
(°C)
TG DSC
Urea – 135 Melting
207 192 Volatilization
238 – Loss of amine groups
300 – Degradation of polymerized fractions
Paraformaldehyde 133 144 Depolymerization/volatilization
UM 1:1 199 185 Urea volatilization
351 – Degradation of polymerized fractions
M/U/Pf 1:1:0.25 183 195 Volatilization of unreacted urea
260 – Depolymerization/volatilization/
decomposition of U:Pf polymer
M/U/Pf 1:1:0.5 191 220 Volatilization of unreacted urea
264 – Depolymerization/volatilization/
decomposition of U:Pf polymer
Fig. 3. DSC curves of (a) urea, (b) paraformaldehyde, (c) montmorillonite, and
M/U/Pf 1:1:1 252 246 Depolymerization/volatilization/
compositions (d) M/U 1:1, (e) M/U/Pf 1:1:0.25, (f) M/U/Pf 1:1:0.5, and (g) M/U/Pf
decomposition of U:Pf polymer
1:1:1.
 C.F. Yamamoto et al. / Chemical Engineering Journal 287 (2016) 390–397 395

Table 4 the non-polymerized sample, caused a retard in the urea release,

Flexural strength and elasticity modulus data obtained by three-point flexural test for if compared with the use of pristine urea. Additionally, polymeriza-
different MMT/urea nanocomposite formulations shaped into cylindrical rods in an
extruder.
tion is shown to play a key role in this process, since the release
time and the equilibrium value over 100 h were directly dependent
Formulation rrf (MPa)a E (MPa)b on the polymer content. After this period, it was not possible to
M/U 1:1 10.0 ± 2.5 16.7 ± 3.0 affirm that the urea content remaining in the nanocomposites
M/U/Pf 1:1:0.25 17.5 ± 4.2 37.8 ± 10.6 was unavailable, but the almost linear trends observed in this
M/U/Pf 1:1:0.5 25.7 ± 4.0 58.9 ± 7.2
M/U/Pf 1:1:1 14.2 ± 3.5 67.1 ± 14.1
study suggest that the time for total urea release for the M/U/Pf
1:1:1 nanocomposite may be superior to 300 h, i.e., 13 days of
a
rrf: Flexural strength. complete immersion into water. Altogether, this full immersion
b
E: Elasticity modulus.
test is very aggressive and may not correspond to real conditions
in field, where the moisture level is limited by the soil uptake
capacity.

1,5 (a)

1,2

0,9

NH4 / mg
+
0,6

Urea
0,3 M/U 1:1
M/U/Pf 1:1:0.25
M/U/Pf 1:1:0.5
0,0 M/U/Pf 1:1:1

0 5 10 15 20 25 30
Time/ Days

1,5
Urea (b)
M/U 1:1
M/U/Pf 1:1:0.25
1,2
M/U/Pf 1:1:0.5
M/U/Pf 1:1:1

Fig. 4. Cross-sectional X-ray microtomography images of the nanocomposites 0,9

NO3 / mg

(pellets) obtained by cold extrusion: (a) M/U 1:1, (b) M/U/Pf 1:1:0.25, (c) M/U/Pf
1:1:0.5 and (d) M/U/Pf 1:1:1.
-

0,6

more brittle matrix. This is consistent with the results obtained by

mechanical tests. In the non-polymerized M/U 1:1 and M/U/Pf 0,3

1:1:0.25 samples, on the other hand, the matrix is probably more

flexible and could have suffered structural re-arrangement during 0,0
drying.
0 5 10 15 20 25 30
Fig. 5 shows the urea release profile in water (full immersion)
for all extruded samples. It is noteworthy that all materials, even Time / Days

100
Urea 1,5 (c)
Total inorganic nitrogen / mg

1,2
80
% Cumulative release

M/U 1:1
M/U/Pf M/U/Pf
1:1:0.25 1:1:0.5 0,9
60

0,6
40

Urea
0,3 M/U 1:1
20 M/U/Pf
1:1:1 M/U/Pf 1:1:0.25
M/U/Pf 1:1:0.5
0,0 M/U/Pf 1:1:1
0

0 20 40 60 80 100 0 5 10 15 20 25 30
Time / h Time / Days

Fig. 5. Kinetic of urea release at 25 °C in aqueous medium (neutral pH) for Fig. 6. Release kinetics for (a) NH+4, (b) NO 3 , and (c) total inorganic nitrogen
commercial urea and M/U/Pf nanocomposites. obtained through soil incubation test at 25 °C and 60% soil moisture.
396
Therefore, in order to analyze the release of nitrogen for the as-
prepared materials in near-real conditions, incubation experiments
were performed, and the influence of moisture (determined by
field capacity) and temperature on the evolution of NH+4 and NO 3 released nitrogen content at equilibrium as the value observed
were analyzed – as seen in Fig. 6, and Figs. 1–3 of the supplemen-
tary material. The samples exhibited behaviors clearly influenced
by the polymerization extent. The non-polymerized sample
presented a nitrogen release behavior very similar to pristine urea,
indicating that the urea release in this material is mainly con-
trolled by a diffusion processes through its granular structure as
previously reported [2]. Conversely, the M/U/Pf 1:1:0.25 sample
showed a reduced NH+4 evolution over the first 5 days of incuba-
tion. This was also observed for M/U/Pf 1:1:0.5, indicating that
there is a chemical effect associated with polymerization, i.e.,
despite the release of nitrogen to the incubating medium may be
fast, the conversion of urea into NH+4 is significantly delayed.
The conversion into NO 3 is important for the nitrogen absorp-
tion process of plants, and it is also related to losses of nitrogen
by volatilization. The NO 3 conversion was found to be governed
by the total NH+4 content available, as displayed in Fig. 6b. All the
evolution trends for NO +
3 follow the NH4 release profile of the
samples, i.e., the samples with high degree of polymerization
generated less NO +
3 due to the low NH4 content available for
conversion. The total N content in the samples, defined by the
sum of the NH+4 and NO 3 contents, follows the same trend, but it
is shown that depending on the polymerization extent, not
all nitrogen will be available after 30 days of incubation. It is
Cumulative total inorganic nitrogen / mg
1,2
0,9
0,6
0,3
0,0
25 30
Temperature / C
Cumulative total inorganic nitrogen / mg
1,2
0,9
0,6
0,3
0,0
60 70
Soil moisture / %
Fig. 7. Statistical analysis performed on data of soil incubation experiments at
different (a) incubation temperatures and (b) moisture soil with respect to the
cumulative total inorganic nitrogen content. The superposition of error bars
indicates that the data are not statistically different at that condition (temperature
or moisture).
C.F. Yamamoto et al. / Chemical Engineering Journal 287 (2016) 390–397
noteworthy that less than half of the total nitrogen content was
released by the M/U/Pf 1:1:1 nanocomposite. Furthermore, the
samples M/U/Pf 1:1:0.25 and 1:1:0.5 attained the same total
for pristine urea and the non-polymerized M/U 1:1 sample, mean-
ing that, even for these polymerized nanocomposites, the total
release of nitrogen is still reachable. This indicates mostly probably
that urease in soil is also active for that short-chain urea–formalde-
hyde polymers. However, this does not mean that the M/U/Pf 1:1:1
nanocomposite is water insoluble, because longer incubation times
could have resulted in greater nitrogen release rates for this
formulation.
Statistical analyses (Fig. 7 and Supplementary Fig. 4) showed
that the results of samples were significantly different, except for
nitrate under different moisture conditions. It is noticed that at
higher temperatures the NO 3 conversion follows to the same pro-
file, i.e., the conversion process is temperature-dependent. On the
other hand, the NH+4 release is shown to be dependent on the type
of nanocomposite structure for all temperature conditions. This
reflects on the total inorganic nitrogen, which is shown to be sta-
tistically different for all samples, except for urea and M/U 1:1
nanocomposite, which are shown to be statistically similar.
Regarding moisture, the samples were shown to behave differently
under lower humidity levels, since the statistical significance is
reduced at higher moistures values. However, even under these
conditions the samples were, in general, statistically different,
which emphasizes the nanocomposite structure effect on the urea
release. The largest deviation among samples is noted for the M/U/
Pf 1:1:1 nanocomposite probably due to its highest degree of
(a) polymerization.
4. Conclusions
This study has shown that formation of urea–formaldehyde
polymer from paraformaldehyde in the presence of clay mineral
montmorillonite is an effective method to obtain fertilizer granules
by extrusion. The presence of montmorillonite was essential to
Urea assure suitable processing, since the nanocomposite matrix itself
M/U 1:1 is the fertilizer. Fully polymerized material showed partial solubil-
M/U/Pf 1:1:0.25
M/U/Pf 1:1:0.5 ity, indicating that in situ polymerization forms a material that
M/U/Pf 1:1:1 immobilizes urea chemically, thereby preventing its easy conver-
35
sion into other nitrogenous compounds. The intermediate urea–
0
paraformaldehyde molar ratio was found to result in the most
efficient rates of nitrogen release in soil incubation tests, indicating
the optimum compositional range to obtain the nanocomposite
(b) fertilizers. These results confirm the initial assumptions of this
research and create a real possibility of manufacturing intelligent
materials for soil fertilization.
Acknowledgements
The authors thank CNPq, CAPES (Program Science with no
Borders), FAPESP, FAPED, Finep, JICA, and Rede Agronano/Embrapa
Urea for their financial support.
M/U 1:1
M/U/Pf 1:1:0.25
M/U/Pf 1:1:0.5 Appendix A. Supplementary material
M/U/Pf 1:1:1
Supplementary data associated with this article can be found, in
80
the online version, at http://dx.doi.org/10.1016/j.cej.2015.11.023.
References
[1] M.E. Trenkel, Slow- and Controlled-Release and Stabilized Fertilizers: An
Option for Enhancing Nutrient Use Efficiency in Agriculture, IFA, International
Fertilizer Industry Association, Paris, 2010.
 C.F. Yamamoto et al. / Chemical Engineering Journal 287 (2016) 390–397
[2] E.I. Pereira, C.C.T. da Cruz, A. Solomonc, A. Led, M.A. Cavigellid, C. Ribeiro, Novel
slow release nanocomposite nitrogen fertilizers: the impact of polymers on
nanocomposite properties and function, Ind. Eng. Chem. Res. 54 (14) (2015)
3717–3725.
[3] E.I. Pereira, F.B. Minussi, C.C.T. da Cruz, A.C.C. Bernardi, C. Ribeiro, Urea–
montmorillonite-extruded nanocomposites: a novel slow-release material, J.
Agric. Food Chem. 60 (21) (2012) 5267–5272.
[4] A.M. Dave, M.H. Mehta, T.M. Aminabhavi, A.R. Kulkarni, K.S. Soppimath, A
review on controlled release of nitrogen fertilizers through polymeric
membrane devices, Polym.-Plast. Technol. Eng. 38 (4) (1999) 675–711.
[5] G.B. Neyman, E.A. Derr, Homogeneous granules of slow-release fertilizer and
method of making the same, 2002/0011087 A131-Jan-2002.
[6] T. Jahns, H. Ewen, H. Kaltwasser, Biodegradability of urea-aldehyde
condensation products, J. Polym. Environ. 11 (2003) 155–159.
[7] Shanghai Chemical Industry Research Institute, Manufacture of slow release
fertilizer involves heating and mixing solid urea or urea solution with
formaldehyde solution to obtain mixed slurry, and mixing phosphorous or
nitrogen fertilizer, with urea formaldehyde solution; W. Cao, M. Chen, L. Duan,
J. Liang, H. Yang, W. Zhang, D. Zhu, CN101475414-A, 2009 Jul 08.
[8] Koch Agronomic Services LLC, Improving flow of urea-based granular fertilizer
composition, where the urea-based granular fertilizer has been exposed to
aqueous solution, comprises contacting urea-based granular fertilizer
composition with conditioning agent; K.D. Gabrielson, WO2013090287-A1,
2013 Jun 20.
[9] Koch Agronomic Services LLC, K.D. Gabrielson, M.L. Epling, Fertilizer, useful for
reducing nitrification in soil, comprises a nitrogen source comprising urea,
ammonia and/or ammonium nitrate, and a reaction product of formaldehyde,
an ammonia source and a nitrification inhibitor; M.L. Epling, Gabrielson,
WO2011137393-A1, 2011 Nov 03.
[10] A.R. Sutton, M. Case, T.J. Healey, Method of odor control involves applying a
urea formaldehyde polymer additive to manure to reduce or eliminate odor
where the polymer is polymethyl urea resin; A.R. Sutton, M. Case, T.J. Healey,
WO2010093462-A1, 2010 Aug 19.
397
[11] Georgia-Pacific Chem LLC, Koch Agronomic Services LLC, Granular-type urea
fertilizer for plants including grasses comprises mixture of urea and non-
thermosetting urea–formaldehyde resin solids; J.C. Phillips, United States,
US2008141747-A1, 2008 Jun 19.
[12] Georgia Pacific Resins INC, Koch Agronomic Services LLC, Aqueous nitrogen
fertilizer solution for plants, comprises ammonium nitrate and urea–
formaldehyde resin; J.C. Phillips, S.L. Wertz, A.K. Cothran, K.D. Gabrielson, Y.
A.C. Phillips, United States, US2006196241-A1, 2006 Sep 07.
[13] Georgia Pacific Resins INC, Making granular fertilizer comprises contacting
phosphoric acid and ammonia with alkaline liquid containing resin made by
reacting formaldehyde and urea, where heat removes water from liquid; J.C.
Phillips, C.B. Pace, T.G. Holt, K.D. Cochran, United States, US2005144997-A1,
2005 Jul 7.
[14] J.A. Kiernan, Formaldehyde, formalin, paraformaldehyde and glutaraldehyde:
what they are and what they do, Microsc. Today (1) (2000) 8–12.
[15] W.D. Callister Jr., D.G. Rethwisch, Materials Science and Engineering: An
Introduction, eighth ed., John Wiley & Sons, Hoboken, NJ, 2010.
[16] M. Tomaszewska, A. Jarosiewicz, Use of polysulfone in controlled-release NPK
fertilizer formulations, J. Agric. Food Chem. 50 (16) (2002) 4634–4639.
[17] ASTM International, Standard Test Method for Total Nitrogen in Organic
Materials by Modified Kjeldahl Method, ASTM.
[18] T.K. With, T.D. Petersen, B. Petersen, A simple spectrophotometric method for
the determination of urea in blood and urine, J. Clin. Pathol. 14 (1961) 202–
204.
[19] J.T. Spargo, M.A. Cavigelli, S.B. Mirsky, J.E. Maul, J.J. Meisinger, Mineralizable
soil nitrogen and labile soil organic matter in diverse long-term cropping
systems, Nutr. Cycl. Agroecosys. 90 (2) (2011) 253–266.
[20] A.E.I. Elkhalifah, S. Maitra, M.A. Bustam, T. Murugesan, The effect of the molar
mass of ammonium cations on the thermal behavior of
organomontmorillonites, Int. J. Chem. Environ. Eng. 3 (1) (2012) 39–44.
[21] A. Fereidoon, M. Ghorbanzadeh Ahangari, M. Jahanshahi, Effect of
nanoparticles on the morphology and thermal properties of self-healing poly
(urea–formaldehyde) microcapsules, J. Polym. Res. 20 (2013) 1–8.