FBP-470; No.

of Pages 10
ARTICLE IN PRESS
food and bioproducts processing x x x ( 2 0 1 4 ) xxx–xxx

Contents lists available at ScienceDirect

Food and Bioproducts Processing

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

Cleaning studies of coconut milk foulants formed

during heat treatment process

P. Saikhwan ∗ , S. Thongchan, N. Jumwan, P. Thungsiabyuan,

J. Sakdanuphap, S. Boonsom, P. Kraitong, P. Danwanichakul
Department of Chemical Engineering, Faculty of Engineering, Thammasat University, 99 Paholyothin Road,
Klongluang, Pathumthani 12121, Thailand

a b s t r a c t

To elucidate cleaning of coconut milk foulants, swelling and dissolution of model coconut milk foulants found in
batch and continuous heat treatment processes were investigated. The model coconut milk foulants were immersed
in aqueous sodium hydroxide (NaOH) solutions at pH range of 7–12. Both model deposits showed little swelling at
pH less than 10; increasing pH beyond 10 resulted in a rise of extent of swelling. Proteins and fats in the deposits
were removed by NaOH solution. Although some components in the deposits could not be removed by soaking the
deposit in NaOH, swollen deposits were less cohesive. Strength of swollen deposits was measured using fluid dynamic
gauging (FDG) technique and it was found that the strength decreased with increasing pH of NaOH solution (shear
stresses reduced from ∼5 to ∼3.5 Pa). Hence, using NaOH solution at appropriate pHs could improve the cleaning
efficiency of coconut milk foulants. However, the strength of the deposit formed from a continuous heat treatment
process, which was stronger than that obtained from a batch process, could not be measured using FDG (normal stress
∼800 Pa, shear stress ∼30 Pa). This suggests that other cleaning agents or cleaning conditions should be adopted in
cleaning of the deposit formed during continuous heat treatment.
© 2014 The Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.

Keywords: Cleaning; Coconut milk; Fluid dynamic gauging (FDG); Fouling, Swelling

1. Introduction cohesive breakdown where rupture is at deposit–deposit

Coconut milk is an oil-in-water emulsion obtained from aque-
ous extract of coconut meat. Moisture and fat are major
components in coconut milk and the reported composition
(in wt%) are: moisture, 54.1; fat, 32.2; protein, 4.4 and car-
bohydrate, 8.3 (Popper et al., 1966). Coconut milk is used in
traditional Asian dishes and is available worldwide in vari-
ous forms. In order to preserve coconut milk, heat treatment
is required. For instance, pasteurization involves heating the
milk to temperature of 72 ◦ C for 20 min (Seow and Gwee,
1997) whereas ultra-high temperature (UHT) treatment of
coconut milk requires heating the milk at 121 ◦ C for 20 min
(Arumughan et al., 1993).
Deposits form on heating surfaces during heat treat-
ments of food products and daily cleaning is generally usual
(Changani et al., 1997). Cleaning or removal of deposit lay-
ers from equipment surfaces is observed in two modes: (i)
bonds and (ii) adhesive breakdown where rupture is at
deposit–substrate bonds resulting in a clean substrate. Sim-
ilar to other processes, equipment fouled with coconut milk
deposits is usually cleaned by a combination of chemical and
mechanical methods. First, the deposit is softened by the
action of cleaning solution, and then is removed by a mechan-
ical action such as scouring and pigging. Alkali solutions are
common in cleaning of food soils as they are known to break
down proteins (Lelieveld et al., 2003) and saponify triglycerides
found in fat into water soluble fatty acid salts (Plett, 1985;
Fryer and Asteriadou, 2009). In addition to alkali solution, large
amount of surfactant content and elevated temperature are
used in cleaning of triglycerides (Jurado-Alameda et al., 2012;
Cunault et al., 2013). Other cleaning agents used in cleaning of
fatty food soils are enzymes and ozone (Jurado-Alameda et al.,
2012). Cleaning of coconut milk deposits is also done by alkali
solution. Nevertheless, the cleaning is poorly understood and

∗
Corresponding author. Tel.: +66 0 898914807.
E-mail address: psaikhwan@engr.tu.ac.th (P. Saikhwan).
Received 20 April 2013; Received in revised form 16 December 2013; Accepted 24 December 2013
0960-3085/$ – see front matter © 2014 The Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.fbp.2013.12.011

Please cite this article in press as: Saikhwan, P., et al., Cleaning studies of coconut milk foulants formed during heat treatment process. Food
Bioprod Process (2014), http://dx.doi.org/10.1016/j.fbp.2013.12.011
FBP-470; No. of Pages 10
ARTICLE IN PRESS
2 food and bioproducts processing x x x ( 2 0 1 4 ) xxx–xxx
Nomenclature
h clearance (m)
m0 initial mass of a deposit (kg)
mf final mass of a deposit (plateau) (kg)
M mass flow rate (kg s−1 )
Ps suction stress (Pa)
r FDG nozzle radius (m)
R initial rate of swelling (min−1 )
t time (min)
ı thickness of a deposit (m)
ıf final thickness of a deposit (plateau) (m)
ı0 initial thickness of a deposit (m)
m mass change (kg)
εm maximum extent of swelling estimated from
gravimetric measurements
ε␦ maximum extent of swelling estimated from
FDG measurements
 viscosity (Pa s)
 density (kg m−3 )
 shear stress (Pa)
many of the protocols used are anecdotal. For instance, soft-
ening of the deposit during chemical cleaning due to swelling
is controlled by pH and optimum pHs for rate of swelling and
swelling suppression of protein deposits have been reported
(Bird and Fryer, 1991; Mercadé-Prieto et al., 2007; Saikhwan
et al., 2010). However, there is no report of optimum pH used
in cleaning of coconut milk foulant.
Unlike dairy deposits, little work has been done on coconut
milk foulants. Nonetheless, literature shows some similari-
ties between the two deposits. When milk is heated, milk
proteins start to denature and aggregate, whereas ions will
tend to precipitate, both result in fouling (Rosmaninho et al.,
2007). Denaturation of proteins in coconut milk upon heating
was also reported. As protein that is not water soluble acts
as an emulsifying agent for the oil–water emulsion of coconut
milk, the instability of the emulsion upon heating is a result of
denaturation of coconut milk proteins (Seow and Gwee, 1997).
Denaturation temperatures of proteins in coconut milk were
observed at about 92 and 110 ◦ C (Seow and Goh, 1994). The
denaturation at 92 ◦ C was suggested to be due to the denat-
uration of globulin whereas the denaturation at 110 ◦ C was
attributed to the denaturation of albumins and some globulins
(Kwon et al., 1996). Fats were reported to have little or no effect
on milk fouling (Visser and Jeurnink, 1997). Although compo-
sitions of milk (87% water, 5% lactose, 3.06% protein, 4.0% fat,
0.4% mineral: Walstra and Jenness, 1984) and coconut milk are
different and the liquids show different rheological behaviours
(Narataruksa et al., 2010), the aforementioned experimental
findings suggest the denaturation of coconut milk proteins as
the cause of fouling.
The swelling behaviour of a given deposit in cleaning solu-
tion is of importance in understanding chemical cleaning of
the deposit. For instance, swelling studies of ␤-lactoglobulin
(␤-Lg) gels and whey protein gels were used to elucidate clean-
ing of milk deposits (Mercadé-Prieto et al., 2007; Saikhwan
et al., 2010). This paper reports the study of swelling of
coconut milk deposits in sodium hydroxide (NaOH) in order
to understand the cleaning of the deposit. Furthermore the
information obtained from this work could be used to optimize
cleaning protocols of coconut milk foulants. The technique
Please cite this article in press as: Saikhwan, P., et al., Cleaning studies of coconut milk foulants formed during heat treatment process. Food
Bioprod Process (2014), http://dx.doi.org/10.1016/j.fbp.2013.12.011
of fluid dynamic gauging (FDG) was employed to measure
swelling kinetics and strengths of coconut milk foulants
soaked in NaOH solutions of varying pH. A detailed descrip-
tion of FDG and procedures of using FDG in measuring swelling
kinetics have been reported elsewhere (Chew et al., 2004;
Saikhwan et al., 2007; Gordon et al., 2010).
2. Experimental
2.1. Preparation of coconut milk
Coconut milk was obtained by squeezing fresh shredded
coconut meat in distilled water at 45 ◦ C. The ratio of coconut
meat to distilled water used was 500 g coconut meat to 200 ml
distilled water; this is a typical ratio used in extracting coconut
milk. Amounts of fat and protein in the extracted coconut
milk were estimated using liquid extraction as described by
Narataruksa et al. (2010) and Kjeldahl method (AOAC, 1990)
respectively. It was found that on average in 100 ml, the
extracted coconut milk contained 2.63 ± 0.04 g of protein and
31.6 ± 1.3 g of fat respectively.
2.2. Formation of deposits
Coconut milk deposits were generated on sample plates (SS
304, thickness of 0.6 mm). Two methods of deposit formation
were employed to investigate coconut milk foulants formed
during heat treatments in batch and continuous processes.
2.2.1. Batch process (BP)
Heat treatment of coconut milk in a batch system is com-
mon in small scale production. To simulate a batch heat
treatment system, coconut milk deposits were formed by
heating coconut milk in an apparatus (Fig. 1(a)) from room
temperature (heating rate of ∼1 ◦ C min−1 ). A thermometer was
immersed in coconut milk to measure temperature of the
coconut milk. Once the coconut milk was at 70 ◦ C, heating
was continued for 20 min. After that, the sample plates (SS304,
10 mm × 70 mm, 0.6 mm thickness) were removed from the
apparatus, rinsed in distilled water to remove excess coconut
milk, drained and weighed. During heating, an overhead stir-
rer was used (100 rpm) to ensure uniform temperature of
coconut milk and avoid film forming at the top surface of
coconut milk. Larger rotational speed was avoided as this
is avoided in the real process; high rotational speed results
in instability of coconut milk after heating. The sample
plate surface temperature measured by a thermocouple was
recorded as 80–90 ◦ C which is slightly less than the reported
denaturation temperature of coconut milk protein (Seow and
Goh, 1994).
2.2.2. Continuous process (CP)
Coconut milk deposits were generated using an apparatus
similar to that reported by Hooper et al. (2006) simulating a
continuous pasteurization process. Coconut milk was recir-
culated through the deposition cell, a laboratory co-current
heat exchanger, shown in Fig. 1(b). A sample plate (SS304,
25 mm × 120 mm, 0.6 mm thickness) was placed inside the
deposition cell as substrates for coconut milk deposits. Ther-
mocouples were used to record temperature at different
positions (Fig. 1(b)). The coconut milk was preheated in a
beaker held in a water bath (∼50 ◦ C) and in silicone tubing
immersed in hot water (95 ◦ C) before entering the deposition
FBP-470; No. of Pages 10
ARTICLE IN PRESS
food and bioproducts processing x x x ( 2 0 1 4 ) xxx–xxx 3

Fig. 1 – Schematic of apparatus used in formation of coconut milk deposits: (a) batch system (BP) and (b) continuous system
(CP) using a deposition cell, thermocouples (T1–T4) can be inserted at the inlet and outlet of utility-side and coconut
milk-side for temperature measurement.

cell at 65 ◦ C. Hot water was circulated at 95 ◦ C on the utility
side (sample plate temperature ∼92 ◦ C). As heat transfer area
of the deposition cell is small, the preheated beaker and tub-
ing immersed in hot water were used to ensure a coconut
milk outlet temperature of 70 ◦ C. The coconut milk at the out-
let was then circulated back to the preheated beaker and into
the deposition cell. The circulation was continued for 20 min.
Sample plates then were removed from the apparatus, rinsed
in distilled water to remove excess coconut milk, drained
and weighed. Only the deposit formed close to the outlet of
the deposition cell was used in swelling and cleaning stud-
ies because this deposit would simulate the deposit formed
in real industrial equipment. The average velocity of coconut
milk was at 0.048 m s−1 (Re ∼ 3088); this velocity was chosen
to obtain similar Reynolds number used in Butapetch and
Narataruksa (2004) and Narataruksa et al. (2010) (Re ∼ 3278).
The average wall stress was 16 Pa.
Measurements of total proteins and fats in the coconut
milk deposit were made using Kjeldahl method and soxhlet
method respectively (AOAC, 1990). All chemicals used were
analytical grade.
FTIR spectra of deposits were obtained (Spectrum 2000,
Perkin-Elmer) before and after immersion in NaOH solu-
tion (pH 12) to detect any chemical changes of the deposits.
Before each scan, the deposit was dried at 60 ◦ C until no
further change in weight and prepared using thin film
technique.
2.3. Swelling studies
All studies of swelling kinetics were performed on freshly
made deposits by monitoring thickness of the deposits soaked
in NaOH solutions at room temperature. Aqueous NaOH
solutions were prepared by dissolving NaOH pellets in dis-
tilled water. Solution pH was measured using a pH electrode
(Hannah) and experiments were performed at least twice to
confirm reproducibility.
The thickness of BP deposits could not be measured using
FDG because the deposits were disrupted by the gauging flow.
Hence, assuming thickness of deposits being proportional to
their weights, gravimetric measurements (see Blackadder and
Le Poidevin, 1978; Mercadé-Prieto et al., 2007) were employed.
The deposit was immersed in a beaker (1 L) containing NaOH
solution and was removed from the beaker at every 5 min. Fil-
ter paper was used to absorb excess liquid from the deposit
and the sample weight was measured using a gravimet-
ric balance (±0.0005 g). Then the deposit was returned to
the beaker. These steps were repeated until no change in
deposit weight was observed (the indication of the end of
swelling). A pH electrode was placed to monitor the pH of
the NaOH solution in the beaker and when any change in
pH (larger than 0.5) was observed, 1 M NaOH solution was
added to adjust the pH. The pH range used in the investi-
gation was 7–12. Densities of fresh CP foulant and swollen
foulants were measured and they were comparable. Therefore,

Please cite this article in press as: Saikhwan, P., et al., Cleaning studies of coconut milk foulants formed during heat treatment process. Food
Bioprod Process (2014), http://dx.doi.org/10.1016/j.fbp.2013.12.011
FBP-470; No. of Pages 10
ARTICLE IN PRESS
4 food and bioproducts processing x x x ( 2 0 1 4 ) xxx–xxx

the use of gravimetric measurement to monitor swelling was

valid.
Swelling studies of the deposits obtained from CP were
performed using the 5 mm diameter nozzle FDG device (See
Saikhwan et al., 2006). A small hydrostatic head (20 mm)
was used to minimize disruption of the deposit. The gaug-
ing flow was then established without the deposit present.
After that the gauge nozzle was withdrawn, the deposit
sample was placed underneath the gauge nozzle and the clear-
ance (h) between the gauge nozzle and the deposit surface
was adjusted to give flow rate of 5.5–5.9 g s−1 . This flow rate
range is in the range corresponding to values of clearance
of 1.2–1.3 mm (the limit of reliable thickness measurement
of this FDG configuration, Saikhwan, 2008). The gauging flow
was checked regularly in order to maintain constant clearance
to avoid any disruption caused by gauging flow during the
Fig. 2 – Images of coconut milk deposits obtained from (a)
swelling of the deposit. Thickness measurements using FDG
batch and (b) continuous heat treatment processes.
were conducted twice for each sample. Before any swelling
studies, an initial thickness of sample was measured using
water in the FDG device. Then water was drained and FDG where M is mass flowrate of gauging flow,  is viscosity of
was set-up with NaOH solution as gauging flow to monitor gauging fluid,  is density of gauging fluid and r is a nozzle
thickness of the sample during swelling. The timing started radius of an FDG device.
when the deposit was in contact with NaOH solution and the Normal stress on the gauged deposit was estimated from
FDG measurement was stopped at least 15 min after no further measured normal stresses (see Saikhwan et al., 2006).
swelling was observed. All swelling studies were conducted
at room temperature. In addition to obtaining continuous 3. Results and discussion
thickness-time profiles of deposits, another thickness-time
profile was monitored only at the beginning and towards the 3.1. Coconut milk deposits
end of swelling. This additional thickness measurement gave
similar profile (not shown) to those obtained from the contin- Fig. 2 illustrates BP and CP deposits; both deposits have the
uous measurements and hence it was ensure that the swelling same colour as coconut milk. As seen in the figure, the
process was not disrupted by gauging flow during swelling formation of CP deposit is more even. In addition, the BP
studies. deposit was observed as a soft deposit whereas the CP deposit
was more gel-like. This difference could be due to higher
fat content found in BP deposit and higher protein content
2.4. Deformation testing
found in CP deposit (Table 1). Moreover, different structure and
strength could be due to different flow dynamics at heated sur-
The deposit that was soaked in NaOH solution to maximum
face during fouling as configurations of the two systems were
thickness or equilibrium thickness was then tested for defor-
different. Nevertheless, the two deposit types show the same
mation using the FDG device with water as the gauging fluid.
major components as shown in Fig. 3 where both deposits
For deformation testing, a hydrostatic head was set to the test
show peaks associated with protein (e.g. N H) and fat (e.g.
value and the gauge was moved progressively closer to the
C O). According to Fig. 3(a)(i), there are peaks of COO− (peak
deposit and its thickness, ı, measured. A delay of 120 s was
7) and OH (peaks 2, 8, 9). These peaks suggest the presence
made between consecutive measurements to ensure that the
of carboxylic acid in BP deposit. Moreover, FTIR spectra of BP
system had reached equilibrium. The procedure was repeated
deposit exhibited peaks in the frequencies of 3800–3200 cm−1
with a larger hydrostatic head until no deposit was visible on
which appeared in FTIR spectra of soybean oil subjected to
the surface. Tests were performed at least twice.
thermal oxidation (Goburdhun et al., 2001). However, these
Shear stress exposed on the deposit was estimated using
peaks were not observed with CP deposit (Fig. 3(b)(i) and
the following equation (Chew et al., 2005):
Table 2) which could be due to higher fat content in BP deposit
(Table 1).
 3M 1 Peaks around 700 cm−1 were evident in both deposit types.
= (1)
 h2 r These peaks could be identified as the amine group and alkyl

Table 1 – Average mass and weight composition (rounded to ±0.1) of each component found from coconut milk deposits
(from batch process: BP and continuous process: CP) before and after swelling in NaOH solution at pH 12.
Component Before g (wt%) After g (wt%) Extent of removal (%)

BP CP BP CPs BP CP

0.59 0.03 0.20 0.015

Protein 66 50
(9.5) (23.0) (8.1) (19.0)
4.43 0.08 1.65 0.045
Fat 63 44
(71.0) (62.0) (67.0) (57.0)
1.22 0.02 0.61 0.02
Others 50 0
(20.0) (15.0) (25.0) (25.0)

Please cite this article in press as: Saikhwan, P., et al., Cleaning studies of coconut milk foulants formed during heat treatment process. Food
Bioprod Process (2014), http://dx.doi.org/10.1016/j.fbp.2013.12.011
FBP-470; No. of Pages 10
ARTICLE IN PRESS
food and bioproducts processing x x x ( 2 0 1 4 ) xxx–xxx 5

50
(a) (i) before

13

Transmittance (%)
12
1011
89
7
10 1 2 3 5 6
(ii) after

3 4

Transmittance scale of 10-50%

2 Transmittance scale of 41.5-46.5%
1

4000 3400 2800 2200 1600 1000 400

Wave number (cm-1)
(b) 40 (i) before
Transmittance (%)

4 5
1 2
20 3
40 (ii) after

4 5
2 3
1

20
4000 3400 2800 2200 1600 1000 400
Wave number (cm-1)

Fig. 3 – FTIR of coconut milk deposits before and after immersion in NaOH solution (pH 12, room temperature): (a) batch
processed foulant and (b) continuous processed foulant (FTIR transmittance bands and their characteristic functional groups
are described in Table 2).

group found in carbohydrate (Kačuráková and Wilson, 2001). suggests that the unidentified components in Table 1 could be
The peak around 700 cm−1 was also observed from virgin carbohydrate.
coconut oil sample (723 cm−1 , Lu and Tan, 2009).
Effect of cleaning with NaOH was also investigated using
3.2. Swelling profiles
FTIR. Signal obtained from BP sample that had been immersed
in NaOH was so weak that the graph plotted using the same
Fig. 4 shows various swelling behaviours of coconut milk
transmittance scale as Fig. 3(a)(i) would show no remain-
deposits observed. Fig. 4(a) shows the constructions used
ing organic components (see Fig. 3(a)(ii), black line). In the
to extract the characteristic parameters, namely the initial
enlarged graph (Fig. 3(a)(ii), grey line), the remaining peaks
swelling rate, R, and the maximum extents of swelling, ε␦ and
suggests that fat and protein were removed (peaks 6–12 in
εm , defined as
Fig. 3(a)(i) disappeared). Comparing Fig. 3(b)(i) and 3(b)(ii), only
peak 1 in Fig. 3(b)(i) disappeared whereas peaks 3, 4 and 5 in
ıf m
Fig. 3(b)(i) were at higher transmittance than peaks 2, 4 and 5 εı =  εm = f (2)
ı0 m0
in Fig. 3(b)(ii). Hence, cleaning the deposit with NaOH removed
the major components to some extent.
where ε␦ and εm are maximum extents of swelling obtained
As shown in Table 1, components other than protein and
from FDG and gravimetric measurements respectively, ı0 and
fat were observed in deposits both before and after swelling.
ıf are the initial thickness and the final thickness of a deposit
Similarly, peaks around 700 appear in FTIR results obtained
measured using FDG; m0 and mf are initial mass and final
from deposits both before and after swelling. This observation
mass of a deposit when swelling is monitored by gravimetric
measurements.

Please cite this article in press as: Saikhwan, P., et al., Cleaning studies of coconut milk foulants formed during heat treatment process. Food
Bioprod Process (2014), http://dx.doi.org/10.1016/j.fbp.2013.12.011
FBP-470; No. of Pages 10
ARTICLE IN PRESS
6 food and bioproducts processing x x x ( 2 0 1 4 ) xxx–xxx

Table 2 – FTIR transmittance bands and their characteristic functional groups observed from Fig. 3.
Figure Peak (s) Wavelength (cm−1 ) Characteristic groups and modes of vibration

3(a)(i) before 1 3279 N H stretching

2 2969, 2945 C H stretching
3 2855 O H stretching
4 2346, 2363 Carbon dioxide
5 1744 C O stretching
6 1656 N H bending
7 1560 COO−
8, 9 1459, 1400 O H bending
10 1244 C O stretching/C N stretching
11, 12 1162, 1114 C N stretching
13 721 N H wagging, C H bending

3(a)(ii) after 1 3487 N H stretching

2 2359 Carbon dioxide
3 1760 C O stretching
4 710 N H wagging, C H bending

3(b)(i) before 1 2919 C H stretching

2 2359, 2342 Carbon dioxide
3 1747 C O stretching
4,5 883, 721 N H wagging, C H bending

3(b)(ii) after 1 2361 Carbon dioxide

2 1748 C O stretching
3 1560 COO−
4,5 883, 720 N H wag, C H bending

3.2.1. BP deposits According to Table 1, the removal of fats from BP deposit is

The deposits obtained from BP showed insignificant swelling
in water (pH 7) but they showed swelling with falling rate
behaviour in NaOH solutions as shown in Fig. 4(a). The data
were fitted to Fick’s law (not shown) to confirm Case I swelling
(see Crank, 1975) and this swelling behaviour suggests that dif-
fusion of NaOH into the deposit was much slower than that of
relaxation of the deposit (Alfrey et al., 1966). The results could
also be interpreted as the swelling was controlled by diffusion
and reaction, as discussed by Mercadé-Prieto et al. (2008). Dur-
ing swelling in NaOH solution, an oil layer was observed in
the solution. Light absorption of the solution was measured
using UV-spectrophotometer (UV-1201, Shimadzu, Japan)
and maximum absorption was found at 395 nm indicating
existence of phenolic compounds in the solution (Mebirouk
et al., 2007). Change in optical density (OD) of NaOH solution
with time at this wavelength is in accord with change in mass
of the deposit, m, compared to the initial mass, m0 , as shown
in Fig. 5. Hence, it could be concluded that swelling of the
deposit results in a release of the compound observed in the
NaOH solution.
3.2.2. CP deposits
Swelling of CP deposit in water follows Case II as depicted
in Fig. 4(b). Case II behaviour was also observed from the
sample in NaOH solution at low pHs (8–10). However, the
deposits immersed in NaOH at high pHs (11–12) exhibited Case
I swelling (Fig. 4(c)). Since, swelling rate of Case I swelling
decreases with time due to the growth in thickness of the
deposit, the shift from Case II swelling to Case I swelling at
pH around 10 could be explained by the much larger extent
of swelling at pH values beyond 10 (Fig. 6(b)). Since swelling
experiments of the deposit obtained from the CP were con-
ducted using the FDG device where the amount of NaOH
solution used was significantly larger compared to the mass
of the deposit, no change in OD of the solution was observed.
Nevertheless, fat layer in the solution was also visible.
higher than from CP deposit. This could be due to the higher fat
content of BP deposit. Table 1 also shows that NaOH removed
both fat and protein from deposits to a similar extent. How-
ever, the other unidentified compounds present in deposits
could not be easily removed by NaOH.
3.3. Effect of pH
Fig. 6 illustrates effect of pH on the maximum extent of
swelling of the deposits. The figure shows little swelling at pH
values less than 10. Numerous literature discussed the onset
of swelling of polymeric deposits as a result of deprotonation
and hydrolysis which occurred around the constituent amino
acid pKa (Mercadé-Prieto et al., 2007; Saikhwan et al., 2007).
As the onset of swelling at pH 10 is close to pKa of glutamic
acid, the major amino acid in coconut milk, at 9.8 (Kwon et al.,
1996), swelling of coconut milk deposits could be due to depro-
tonation and subsequent charge repulsion. The effect of pH
on initial swelling rate is demonstrated in Fig. 7 with a trend
similar to Fig. 6.
During swelling, fat layers were observed at the surface
of NaOH solution and this could be due to the effect of
hydrolysis of fat and saponification. Since the extent of
hydrolysis increases with increasing concentration of NaOH
solution, thicker oil layer observed in NaOH solutions during
swelling at high pH confirmed the hydrolysis of fat during
the swelling. The hydrolysis was also tracked by FTIR spectra
of deposits immersed in NaOH as peaks associated with fat
disappeared (see Section 3.1). It should be noted that the
peak at 1560 cm−1 (peak 3 in Fig. 3(b)(ii)) appeared in the FTIR
spectrum of CP deposit after swelling in NaOH. Since this
peak was not shown in the fresh CP deposit, this peak was not
identified as fat presenting in the deposit. It is more appro-
priate to identify this peak as COO− vibration associated
with deprotonation of protein or sodium soap formation
(frequencies specific for RCOO− Na+ vibration:
−1
1570–1550 cm , Rohman and Man, 2011). The peak is

Please cite this article in press as: Saikhwan, P., et al., Cleaning studies of coconut milk foulants formed during heat treatment process. Food
Bioprod Process (2014), http://dx.doi.org/10.1016/j.fbp.2013.12.011
FBP-470; No. of Pages 10
ARTICLE IN PRESS
food and bioproducts processing x x x ( 2 0 1 4 ) xxx–xxx 7

(a) 1.15 (a) 1.09

1.08

1.10 1.07
δf/δ0
1.06
εm 1.05
δ/δ0 1.05 R
1.04
1.03
1.00
1.02
1.01
0.95 1
0 20 40 60 80 100 120 7 8 9 10 11 12 13
t (min) pH
(b) 1.15 (b) 2
1.9

1.10 1.8
1.7
1.6
δ/δ0 1.05 1.5
εδ
1.4

1.00 1.3
1.2
1.1
0.95 1
0 20 40 60 80 100 120 7 8 9 10 11 12 13
t (min)
pH
(c) 1.95
1.85 Fig. 6 – Effect of pH on equilibrium extents of swelling of
1.75 coconut milk deposits obtained from (a) a batch heat
1.65 treatment process using gravimetric measurements, εm
1.55 and (b) a continuous process using FDG measurements, ε␦ .
δ/δ0 1.45
(a) 0.0030
1.35
1.25 0.0025

1.15
1.05 0.0020

0.95
R (min-1)

0 100 200 300 0.0015

t (min)

Fig. 4 – Swelling at room temperature (∼30 ◦ C) of (a) batch 0.0010

processed deposit immersed in NaOH solution at pH 12; (b)

0.0005
continuous processed deposit immersed in water and (c)
continuous processed deposit immersed in NaOH solution
0.0000
at pH 12. The construction lines show how the various 7 8 9 10 11 12 13
pH
parameters were extracted.
(b) 0.012

0.07 0.30
0.010
0.06 0.25
0.008
0.05
0.20
R (min-1)

Δm/m0
OD
0.04 0.006
0.15
0.03
0.004
0.10
0.02
0.002
pH 11 (1) 0.05
0.01
OD
0.000
0 0.00 7 8 9 10 11 12 13
0 10 20 30 40 50 60 70 pH
t (min)
Fig. 7 – Effect of pH on initial swelling rate, R, of coconut
Fig. 5 – Changes of mass (filled triangle) and optical density milk deposits obtained from (a) a batch heat treatment
(OD, open circle) of coconut milk deposit from a batch heat process using gravimetric measurements and (b) a
treatment process immersed in NaOH solution at pH 11. continuous process using FDG measurements.

Please cite this article in press as: Saikhwan, P., et al., Cleaning studies of coconut milk foulants formed during heat treatment process. Food
Bioprod Process (2014), http://dx.doi.org/10.1016/j.fbp.2013.12.011
FBP-470; No. of Pages 10
ARTICLE IN PRESS
8 food and bioproducts processing x x x ( 2 0 1 4 ) xxx–xxx

weak and this could be because sodium soap or protein 5

dissolved in NaOH solution during swelling. Since BP deposit
was more porous compared to CP deposit, dissolution of
4
sodium soap and protein would be faster. This could be the
reason why peak in the frequency range of 1570–1550 cm−1 is
not evident in Fig. 3(a)(ii) although deprotonation of protein 3

τ (Pa)
and fat hydrolysis occurred.
Another evidence of fat hydrolysis was the significant drop 2
in pH values of the NaOH solution at low pHs during the
swelling of coconut milk deposits. Apart from disappearance
1
or decreased in intensity of peaks associated with carboxylic Cohesive
acids and protein, there is no other significant change of FTIR Adhesive
spectra of the deposits after immersion in NaOH solutions. 0
This suggests that molecular transformation during swelling No soaking pH 7 pH 12
could be only due to hydrolysis of proteins and fats. Fig. 8 – Shear stresses () observed with cohesive and
adhesive breakage of the deposit from a batch process;
3.4. Deposit strengths deposits tested were under three conditions: (i) no soaking;
(ii) soaking in water (pH 7) and (iii) soaking in NaOH
3.4.1. BP deposits solution at pH 12.
Surfaces of these deposits in the area underneath the noz-
3.4.2. CP deposits
zle rim were sheared off by gauging flow and the measured
These deposits exhibited partial bending in the region directly
shear stresses were reported as cohesive strengths of the
underneath the gauge nozzle as shown in Fig. 9(a). Upon
deposits could be removed easily, strength measurements
removal of stresses, the bending remained. Therefore, the
were performed on the deposits under three conditions: (i)
bending could be interpreted as cohesive breakage of the
freshly made deposit that was not soaked in water/NaOH
deposits. The pH effect on cohesive strength of these deposits
solution; (ii) deposit immersed in water (pH 7) until equilib-
plotted in Fig. 9(b) suggested similar trend as the deposits
rium thickness was reached (∼40 min) and (iii) deposit soaked
obtained from the batch process.
in NaOH solution at pH 12 until equilibrium thickness was
Nevertheless, the deposits from the continuous process
reached (∼50 min). Fig. 8 shows that soaking reduced cohe-
could not be removed adhesively even at the maximum
sive and adhesive strengths and the strengths decreases with
stress that could be imposed by the FDG device used (nor-
increasing pH of a solution used in soaking. Nevertheless,
mal stress ∼800 Pa, shear stress ∼30 Pa). This indicates strong
the adhesive strengths of deposit samples tested under the
deposit–substrate interactions of these deposits and clean-
three conditions were comparable. This suggests that the
ing using NaOH solution at room temperature without hard
use of NaOH at low pHs would be sufficient in the clean-
mechanical cleaning was not possible.
ing.
Since both CP and BP deposits consist of same major com-
ponents, different swelling and cleaning behaviours are more

Fig. 9 – (a) Deformation observed with the deposit from a continuous process; (b) suction (Ps ) and shear stresses (t) observed
with cohesive breakage of the deposit soaked in NaOH solution at various pH values.

Please cite this article in press as: Saikhwan, P., et al., Cleaning studies of coconut milk foulants formed during heat treatment process. Food
Bioprod Process (2014), http://dx.doi.org/10.1016/j.fbp.2013.12.011
FBP-470; No. of Pages 10
ARTICLE IN PRESS
food and bioproducts processing x x x ( 2 0 1 4 ) xxx–xxx
likely to be due to different conditions used in heat treat-
ment processes such as heating rate and flow dynamic. Effects
of coconut milk temperature at different positions in a heat
exchanger (varies from 50 to 70 ◦ C), fouling period and ageing
of deposits should also be studied in future work.
4. Conclusions
Swelling studies of coconut milk foulants soaked in aque-
ous caustic soda solutions were conducted. Simulated BP and
CP deposits appeared similar and both deposits showed dis-
solutions of proteins and fats during swelling. The protein
dissolution could be from hydrolysis whereas the fat disso-
lution could be due to saponification of fat. In addition, FTIR
spectra suggest no other reaction involved during swelling of
both deposits.
The equilibrium thickness and deposit strength increased
with increasing pH values for both deposit types. Neverthe-
less, adhesive strength of the CP deposit was considerably
larger than that of the BP. The results suggested that cleaning
of the CP deposit could not be done using sodium hydroxide
solution alone at room temperature and other cleaning agents
or cleaning conditions should be adopted to improve cleaning
efficiency.
Acknowledgements
Funding from Thailand Research Fund and Faculty of Engi-
neering at Thammasat University is gratefully acknowledged.
References
Alfrey, T., Gurnee, E.F., Lloyd, W.G., 1966. Diffusion in glassy
polymers. J. Polym. Sci. C 12, 249–261.
Arumughan, C., Balachandran, C., Sundaresan, A., 1993.
Development of a process for coconut cream on commercial
scale. J. Food Sci. Technol. 30, 408–412.
Association Official Analytical Chemistry (AOAC), 1990. Official
Method of Analysis. The Association of Official Agricultural
Chemists.
Bird, M.R., Fryer, P.J., 1991. An experimental study of the cleaning
of surfaces fouled by whey proteins. Trans. IChemE 69C,
13–21.
Blackadder, D.A., Le Poidevin, G.J., 1978. Dissolution of
poly-propylene in organic solvents. 4. Nature of the solvent.
Polymer 19, 483–488.
Butapetch, A., Narataruksa, P., 2004. Monitoring of coconut milk
fouling at pasteurization temperatures. In: Proceedings of the
Regional Symposium of Chemical Engineering (RSCE). The
Grand Hotel Bangkok, Bangkok, Thailand.
Changani, S.D., Belmar-Beiny, M.T., Fryer, P.J., 1997. Engineering
and chemical factors associated with fouling and cleaning in
milk processing. Exp. Therm. Fluid Sci. 14,
392–406.
Chew, J.Y.M., Paterson, W.R., Wilson, D.I., 2004. Fluid dynamic
gauging for measuring the strength of soft deposits. J. Food
Eng. 65 (2), 175–187.
Chew, J.Y.M., Hölfling, V., Augustin, W., Paterson, W.R., Wilson,
D.I., 2005. A method for measuring the strength of scale
deposition on heat transfer surfaces. Dev. Chem. Eng. Miner.
Process. 13 (1/2), 21–30.
Crank, J., 1975. The Mathematics of Diffusion, 2nd ed. Clarendon
Press, UK, pp. 256–257.
Cunault, C., Coquinot, Y., Burton, C.H., Picard, S., Pourcher, A.M.,
2013. Characteristics and composition of fouling caused by pig
slurry in a tubular heat exchanger – recommended cleaning
systems. J. Environ. Manage. 117, 17–31.
Please cite this article in press as: Saikhwan, P., et al., Cleaning studies of coconut milk foulants formed during heat treatment process. Food
Bioprod Process (2014), http://dx.doi.org/10.1016/j.fbp.2013.12.011
9
Fryer, P.J., Asteriadou, K., 2009. A prototype cleaning map: a
classification of industrial cleaning processes. Trends. Food
Sci. Technol. 20, 255–262.
Gordon, P., Brooker, A.D.M., Chew, Y.M.J., Wilson, D.I., York, D.W.,
2010. Studies into the swelling of gelatine films using a
scanning fluid dynamic gauge. Food Bioprod. Process. 88,
357–364.
Goburdhun, D., Jhaumeer-Laullo, S.B., Musruck, R., 2001.
Evaluation of soybean oil quality during conventional frying
by FTIR and some chemical indexes. Int. J. Food Sci. Nutr. 52,
31–42.
Hooper, R.J., Paterson, W.R., Wilson, D.I., 2006. Comparison of
whey protein model foulants for studying cleaning of milk
fouling deposits. Food Bioprod. Process. 84 (C4),
329–337.
Jurado-Alameda, E., García-Román, M., Altmajer-Vaz, D.,
Jiménez-Pérez, J.L., 2012. Assessment of the use of ozone for
cleaning fatty soils in the food industry. J. Food Eng. 110,
44–52.
Kačuráková, M., Wilson, R.H., 2001. Developments in
mid-infrared FT-IR spectroscopy of selected carbohydrates.
Carbohydr. Polym. 44, 291–303.
Kwon, K., Park, K.H., Rhee, K.C., 1996. Fractionation and
characterization of proteins from coconut. J. Agric. Food
Chem. 44, 1741–1745.
Lelieveld, H.L.M., Mostert, M.A., Holah, J., White, B., 2003. Hygiene
in Food Processing, 1st ed. Woodhead Publishing, UK,
pp. 239.
Lu, H., Tan, P.P., 2009. A comparative study of storage stability in
virgin coconut oil and extra virgin olive oil upon thermal
treatment. Int. Food Res. J. 16, 343–354.
Mebirouk, M., Sbai, L., Lopez, M., Gonzalez, J., 2007. The
absorption of polyphenols from olive oil mill wastewaters by
sawdust and biodegradation by the fungus Phanerochaetae
chrysosporium. Grasas Y Aceites 58 (4), 366–371.
Mercadé-Prieto, R., Falconer, R.J., Paterson, W.R., Wilson, D.I.,
2007. The pH threshold in the dissolution of ␤-lactoglobulin
and aggregates in alkali. Biomacromolecules 8, 1162–1170.
Mercadé-Prieto, R., Paterson, W.R., Wilson, D.I., 2008.
Caustic-induced gelation of beta-lactoglobulin. Int. J. Food Sci.
Technol. 43, 1379–1386.
Narataruksa, P., Pichitvittayakarn, W., Heggs, P.J., Tia, S., 2010.
Fouling behaviour of coconut milk at pasteurization
temperatures. Appl. Therm. Eng. 30, 1387–1395.
Plett, E.A., 1985. Cleaning of fouled surfaces. In: Lund, D.B., Plett,
E., Sandu, C. (Eds.), Proceedings of fouling and cleaning in
food processing. Madison, Wisconsin, USA, pp. 286–311.
Popper, K., Notter, G.G., Nurry, F.S., 1966. New products from
coconut: non-dairy chip dips and ‘cream cheese’. Food
Process. Marketing 27 (12/92), 96 (cited by Seow and Gwee,
1997).
Rohman, A., Man, Y.B.C., 2011. Determination of sodium fatty
acid in soap formulation using Fourier Transform Infrared
(FTIR) spectroscopy and multivariate calibrations. J. Surfact.
Deterg. 14, 9–14.
Rosmaninho, R., Santos, O., Nylander, T., Paulsson, M., Beuf, M.,
Bénézech, T., Yiantsios, S., Andritsos, N., Karabelas, A., Rizzo,
G., Müller-Steinhagen, H., Melo, L.F., 2007. Modified stainless
steel surfaces targeted to reduce fouling: evaluation of fouling
by milk components J. Food Eng. 80, 1176–1187.
Saikhwan, P., (PhD Dissertation) 2008. Cleaning behaviour of
some polymeric and proteinaceous fouling layers. University
of Cambridge.
Saikhwan, P., Geddert, T., Augustin, W., Scholl, S., Paterson, W.R.,
Wilson, D.I., 2006. Effect of surface treatment on cleaning of a
model food soil. Surf. Coat. Technol. 201, 943–951.
Saikhwan, P., Chew, J.Y.M., Paterson, W.R., Wilson, D.I., 2007.
Swelling and its suppression in the cleaning of polymer
fouling layers. Ind. Eng. Chem Res. 46 (14), 4846–4855.
Saikhwan, P., Mercadé-Prieto, R., Chew, Y.M.J., Gunasekaran, S.,
Paterson, W.R., Wilson, D.I., 2010. Swelling and dissolution in
cleaning of whey protein gels. Food Bioprod. Process. 88,
375–383.
FBP-470; No. of Pages 10
ARTICLE IN PRESS
10 food and bioproducts processing x x x ( 2 0 1 4 ) xxx–xxx

Seow, C.C., Goh, S.L.,1994. A differential scanning calorimetric
study of the thermal stability of coconut milk proteins. In:
Proceedings of the 5th ASEAN Food Conference. Malaysian
Institute of Food Technology, Kuala Lumpur, Malaysia.
Seow, C.C., Gwee, C.N., 1997. Coconut milk: chemistry and
technology. J. Food Sci. Technol. 32, 189–201.
Visser, J., Jeurnink, T.J.M., 1997. Fouling of heat exchangers. Exp.
Therm. Fluid Sci. 14, 407–424.
Walstra, P., Jenness, R., 1984. Dairy Chemistry and Physics. Wiley,
New York.

Please cite this article in press as: Saikhwan, P., et al., Cleaning studies of coconut milk foulants formed during heat treatment process. Food
Bioprod Process (2014), http://dx.doi.org/10.1016/j.fbp.2013.12.011