Food and Bioproducts Processing 1 1 8 ( 2 0 1 9 ) 130–138

Contents lists available at ScienceDirect

Food and Bioproducts Processing

journa l homepage: www.e lsev ier .com/ locate / fbp

Effects of Fluorolink® S10 surface coating on WPCfouling of stainless steel surfaces and subsequentcleaning

Jian Huo, Jie Xiao, Timothy V. Kirk ∗, Xiao Dong Chen ∗

Suzhou Key Laboratory of Green Chemical Engineering, School of Chemical and Environmental Engineering, Collegeof Chemistry, Chemical Engineering and Materials Science, Soochow University, Suzhou, Jiangsu 215123, PR China

a r t i c l e i n f o

Article history:

Received 15 May 2019

Received in revised form 9

September 2019

Accepted 11 September 2019

Available online 18 September 2019

Keywords:

Heat exchanger fouling

Cleaning

a b s t r a c t

It is reported here that stainless steel surfaces used for heat exchange in the food industry,

once modified with Fluorolink® S10, show an improvement in the removal of whey protein

fouling. The fouling was created under unfavorable operating conditions (i.e. at a very low

fluid velocity). Whey Protein Concentrate (WPC) was used, which is a classic model foulant.

Contact angle, surface energy and work of adhesion were measured to examine the surface

changes due to the Fluorolink® S10 modification. Surface free energy was reduced from 46.38

to 19.00 mN m−1. In contrast to some reports in the literature, little effect on the extent of

fouling was observed here. However, extent of fouling removal was increased ∼75 to 98%.

It is suggested that Fluorolink® S10 can be used to improve the cleaning properties of heat

exchange surfaces, despite of its much lesser impact on minimizing fouling.

WPC

Surface modification

Surface energy

© 2019 Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.

grafting (e.g. Excalibur® and Xylan®, PEEK + fluoropolymer, PEO grafted

1. Introduction

Fouling, or deposition of material, on heat exchanger surfaces is unde-

sirable for many reasons (Patel et al., 2013; Lv et al., 2015), and is

particularly so in dairy processing, as every product needs to be heated

at least once (Fryer et al., 1996; De Jong, 1997; Jun and Puri, 2005).

Undesirable effects of milk fouling include decreased heat transfer,

increased pressure drop, potential product losses, and diminished

heat exchanger efficiency. These losses have considerable impact upon

processing plant economics (Bansal and Chen, 2006; Sadeghinezhad

et al., 2013; Zouaghi et al., 2017). Subsequent cleaning processes are

required in order to remove the deposits, to restore production effi-

ciency, and to meet hygiene standards and regulations (Gillham et al.,

1999; Boxler et al., 2013). It is also well understood that from a food

safety perspective an important issue with fouling is its potential in

harboring heat-resistant bio-organisms (Bansal and Chen, 2006), which

also imposes the need for regular cleaning even if thermo-hydraulic

effects of fouling are negligible. Indeed, it has been reported that clean-

ing efforts have taken up to 15% of dairy plant operating costs (Merian

∗ Corresponding authors.E-mail addresses: tim@suda.edu.cn (T.V. Kirk), xdchen@mail.suda.e

https://doi.org/10.1016/j.fbp.2019.09.0050960-3085/© 2019 Institution of Chemical Engineers. Published by Elsev

and Goddard, 2012; Mauermann et al., 2009). Accordingly, developing

and characterizing anti-fouling and cleaning aids has been a focus of

research worldwide (Saikhwan et al., 2006; Molena et al., 2014).

Many methods have been examined for reducing fouling: improving

heat exchanger design (e.g. geometry), optimizing operating conditions

(e.g. temperatures and flow rates/shear stresses), and chemical modi-

fication of heat exchange surfaces (Müller-Steinhagen et al., 2011; Da

Cruz et al., 2015).The latter strategy, often referred to as anti-fouling

surface development, has attracted a considerable amount of effort

in recent years (Santos et al., 2013). Da Cruz et al. (2015) provided a

summary of many surface modifications, in table form, which was

very informative for this current work (Da Cruz et al., 2015) Different

surface modifications aiming at reducing dairy fouling have included

diamond-like carbon (DLC) (Santos et al., 2006), Ni-P-PTFE (Beuf et al.,

2003; Balasubramanian and Puri, 2008, 2009; Barish and Goddard, 2013;

Rosmaninho and Melo, 2008), ion implantation (SiF+, MoS2, Silica,

SiOx) (Beuf et al., 2003; Balasubramanian and Puri, 2008), and polymer

du.cn (X.D. Chen).

silicone, and some nanocomposites) (Mauermann et al., 2009; Da Cruz

ier B.V. All rights reserved.

Food and Bioproducts Processing 1 1 8 ( 2 0 1 9 ) 130–138 131

Table 1 – Approximate composition of whey proteinconcentrate (WPC80), as given by the supplier.

Components Percentages (wt. %)

Protein 80.3Moisture 5.8Lactose 9.0Fat 4.8Ash 2.5

e

p

o

r

e

b

a

u

f

c

s

m

t

q

m

s

R

c

N

s

e

c

t

i

t

a

(

h

e

p

i

t

t

b

e

f

c

2

t

e

F

s

2

2

W(BiTbr(

Fig. 1 – Molecular structure of the commercialperfluoropolyether (PFPE) Fluorolink® S10 (Oldani et al.,

t al., 2015; Santos et al., 2006; Zouaghi et al., 2019). Liquid infused

orous surfaces (Zouaghi et al., 2017) are another interesting devel-

pment, and anti-fouling coatings based on the effect of hydrophilic

epulsion have been used successfully at the laboratory scale (Zouaghi

t al., 2019). The latter displayed enhanced resistance to resistance to

ovine serum albumin (BSA) protein adsorption, but was not examined

t the elevated temperatures of dairy processing, nor was it evaluated

nder clean in place (CIP) processes.

It is intuitive to think that the modification of stainless steel sur-

aces affects mostly the ‘first layer’ of fouling, with this initial effect

ontributing to the structure of subsequent fouling. As the modified

urface (coating) becomes obscured from subsequent fouling deposit

olecules, the modification may become less important though, with

he formation of succeeding deposits depending more upon subse-

uent protein-protein interactions. However, the surface modifications

ight have an influence on the strength of adhesion between the

urface and foulants (Patel et al., 2013; Sadeghinezhad et al., 2013;

osmaninho and Melo, 2008; Britten et al., 1988). Accordingly, for

leaning processes surface modifications may have significant effects.

otably, many researchers have reported surface coatings to have a

tronger effect on cleaning than on fouling (Da Cruz et al., 2015; Santos

t al., 2004). Thus, assessments of surface coating performance should

onsider their contribution to the whole fouling-cleaning cycle.

Perfluoropolyether polymers were studied by Molena et al. (2014) for

heir properties as protein adhesion resistant treatments for microflu-

dic devices, using BSA as a model protein. Their results indicated

hat perfluoropolyether photopolymers had high potential as protein

nti-fouling/fouling-removal materials. In addition, perfluoropolyether

PFPE) functional oligomers cannot accumulate in animals or in the

uman body, and are considered to be non-toxic materials (Molena

t al., 2014; Gao et al., 2010; Oldani et al., 2015, 2016a). Combined, these

roperties make them good choices for food applications. Evaluated

n this report, Fluorolink® S10 (Solvay S.A., Belgium) is a bifunc-

ional perfluoropolyether with triethoxysilane terminal groups, and

he following claimed characteristics: low surface energy, hydropho-

ic/oleophobic properties, and high chemical and physical stability.

This work explores the effect of a Fluorolink® S10 coating on heat-

xchanger surface fouling and cleaning processes. In order to simulate

ouling during processing of milk, the often considered whey protein

oncentrate (WPC80) model protein solution was used (Christian et al.,

002; Hooper et al., 2006; Fickak et al., 2011, 2012). Surface modifica-

ion by Fluorolink® S10 grafting is expected to decrease surface-vapor

xcess surface energy, raise contact angle, and reduce adhesion force.

urther, the thin coatings typically obtained with S10 (Oldani, 2015)

hould have negligible effect on heat transfer.

. Materials and methods

.1. Materials

hey protein concentrate (WPC 80) produced by FonterraAuckland, New Zealand) was purchased from Xin Shen Aoiological Technology Co. Ltd. (Jiangsu, China). The approx-

mate composition, as provided by the company, is listed inable 1. The 6 wt% WPC80 aqueous solutions were preparedy dissolving powder in deionized water at 50 ◦C, and stir-

ing for 15 min using a dispersion type stirrer at 3000 rpm/minT18 digital ULTRA-TURRAX®, IKA®-Werke GmbH & Co.

2016b; Tonelli et al., 2002).

KG, Germany). Sodium hydroxide (NaOH), acetic acid, iso-propyl alcohol, and butanone solution were purchased fromSinopharm Chemical Reagent Co., Ltd. (Shanghai, China).Fluorolink® S10 was purchased from Solvay Specialty Poly-mers (Shenzhen, China). Its molecular formula is shown inFig. 1.

2.2. Experimental apparatus

A circulating apparatus, previously described by Lv et al. (2015)and Li et al. (2013), was employed in this study, which couldbe divided into the following units: pre-heating, monitoring,cooling, and pumping (see Fig. 2). The sample solution wascirculated using a peristaltic pump.

In the pre-heating section (Number 3 in Fig. 2) the sam-ple solution was heated using a coil immersed in a waterbath. This maintained the inlet temperature of the test fluidat an approximately constant temperature (60 ± 0.5 ◦C), priorto entering the monitoring section.

The monitoring section (Number 4 in Fig. 2) consisted ofa stainless steel (304) heating rod (Haojiang Electric HeatingEquipment Ltd., Changzhou City, Jiangsu Province, and PRChina), thermocouples, power supply, and data acquisitionsystem. As shown on the left hand side of Fig. 2, the heatingrod (diameter 20 mm, total length 300 mm, effective heatinglength 130 mm) was used as the heat exchange surface, withheat flux varied via a multi-meter (UNIT-T, China). The heat-ing rod was fixed in a sealed transparent plexiglass cylinder(ID = 45 mm and length 560 mm), with 5 mm internal diame-ter bottom inlet and top outlet (ID = 5 mm). In addition, twolayers of stainless steel balls (5 mm diameter) were packed atthe bottom of the cylinder to ensure uniform flow distributioninto the monitoring system. K-type thermocouples were usedto measure the inlet, outlet, bulk fluid temperatures, and thesurface temperature of the heating rod (which were stuck onexternally to the surface of the rod). The surface temperaturemeasurements were indicative of the heat exchange situation,though these could not be considered to be totally quantita-tive. Nevertheless, all measured temperatures were logged toa computer using a thermocouple data logger (USB TC-08, PicoTechnology, Cambridge Shire, UK).

In the cooling section (Number 6 in Fig. 2) outlet fluidtemperature was reduced to about 40 ◦C via a cooling watercirculating system (CA-1111, EYELA, Japan), before being recir-culated to the pre-heating section. It is worth noting that itis usually accepted that the whey proteins would not causefouling when the fluid temperature is below 60 ◦C (beyondwhich greater extent of denaturation and aggregation occur,leading to greater fouling) (Bansal and Chen, 2006). The fluidexiting the heater section may be at a temperature well intothe high range for denaturation and aggregation of the wheyproteins, which must be cooled down if the sample fluid is tobe reused/recycle. Fluid from the heating section may reside

in the feed tank for significant periods, so it is cooled down toan ‘inactive state’ for fouling; 40 ◦C is a reasonable choice.

132 Food and Bioproducts Processing 1 1 8 ( 2 0 1 9 ) 130–138

Fig. 2 – Flow diagram of the experimental setup: (1) sample tank; (2) peristaltic pump; (3) electric heater; (4) monitoringsection; (5) voltage transformer; (6) cooling device.

A peristaltic pump (Number 2 in Fig. 2) was used to circulatefluid through the system, with an average bulk fluid velocityof 3.45 × 10−3 m s−1 through the monitoring system.

2.3. Coating

2.3.1. Heating rod pre-cleaningNaOH and Butanone were used to remove organic and inor-ganic contamination from the heating rod’s surface. The rodwas immersed in Butanone solution for 20 min, followed by0.5 wt% NaOH for 20 min, and then in water for 20 min. Allcleaning steps were conducted under ultrasonication.

2.3.2. Preparation of coating solutionsThe coating solutions were prepared according to Table 2. Thesolution was stirred for 2 days at room temperature (RT).

2.3.3. Coating procedureThe coating was deposited on the heating rod’s heat exchangesurface via immersion of the piece of concern for 30 min at RT.The heating rod was then heat treated for 30 min at 100 ◦Cand for 15 min at 150 ◦C in an oven. Similar flat samples of 304stainless steel were coated for surface tension measurements.

2.4. Protein fouling conditions

Fouling tests of coated and uncoated rods were carried outunder similar conditions, as given in Table 3. All fouling (andcleaning) experiments were performed at least twice underthe same operating conditions. After Lv et al. (2015), the tem-perature of pre-heating section was set at 61 ± 0.5 ◦C to ensurean inlet temperature of 60 ± 0.5 ◦C in the monitoring section.The coated and uncoated heating rods were tested in sep-arate experiments. Each fouling experiment was conductedfor about an hour and heating power of heating rod wereset to pre-fixed constant flux (see Table 3). Surface temper-ature exceeded 100 ◦C towards the end of the higher heatflux experiment (refer to Fig. 5(a)), but otherwise conditionsremained within industrially relevant temperatures, with noboiling effects.

Fluid velocity (u = Va/Ac) was calculated for the annularregion between the heating rod and the plexiglass tube. Va

is the protein solution flow rate (which was 4.7 × 10−3 kg s−1;leading to an estimate of fluid recirculation of 6–7 times forthe 3 L feed liquid), and Ac the annular cross-sectional area.Reynolds number was calculated at ∼150, indicating that flowwas laminar. The inner diameter of plexiglass tube was 45 mmand the diameter of heating rod was 20 mm, giving hydraulicdiameter, H, of 25 mm, according to the formula H = 4Ac/P.Where Ac is the cross-sectional area, and P is the wettedperimeter.

The surface of heating rod (the fouling process was over)was hung and dried in an air-circulation laboratory oven setat 60 ◦C and 30 min.

2.5. Fouling monitoring

As mentioned earlier, the surface temperatures were recordedvia a thermocouple attached to the rod’s surface. These tem-peratures only represent the surface conditions qualitativelyas they may be interfered by the presence of fouling. Trackingthe surface temperature does, however, give a basic under-standing of the thermal situation in the fouling section. Assuch fouling can be monitored qualitatively via the overallheat transfer coefficient (U), which was calculated using thefollowing equation:

U = Q

A(Ts − Tb)(1)

where U is the overall heat transfer coefficient (W/(m2.K)), Qis the power input to the heating rod (W), A is the heat trans-fer area of the heating rod (m2), and Q/A is thus the heat flux(W/m2), Ts is the heating rod’s surface temperature (oC), andTb is the bulk fluid temperature (oC). Both Ts and Tb were mea-sured according to Lv et al. (2015) and Fickak et al. (2011, 2012).The experimental repeatability was well in line with theseprevious works using essentially the same technique and rig.

2.6. Cleaning process

The cleaning process maintained apparatus configuration, as

an analogue of clean in place operations, and the cleaningsolution’s flow route was identical to the protein solution’s.

Food and Bioproducts Processing 1 1 8 ( 2 0 1 9 ) 130–138 133

Table 2 – Composition of the coating solution.

Compound Water Isopropyl alcohol Acetic acid Fluorolink® S10

Weight percentage 2% 97% 0.5% 0.5%

Table 3 – Operating parameters during fouling and the cleaning experiments.

Parameter Fouling Cleaning

Solution concentration 6 wt% WPC(80) 0.5 wt% NaOHFeed liquid 3 L 3 LInlet temperature 60 ± 0.5 ◦C 60 ± 0.5 ◦CVelocity 3.45 × 10−3 m s-1 3.45 × 10−3 m s-1

Heating power flux (heat flux) 3.417 and 4.985 kW/m2 respectively –Circulation time 3500 s 600 s

Table 4 – Contact angle and surface energy of 304 stainless steel with no coating and coating surfaces. Measurementswere repeated at least three times on each sample surface; errors are given by standard deviation. *In this singleexperiment, a small and thin sample of flat plate was suitably attached on the heater rod surface which was thensubjected to the fouling (at the high heat flux case), cleaning, and water rinsing process as would have been in thenormal tests in this study).

Contact angle Surface energy (mN m−1) Work of adhesion (J m−2)

No coating surfaces 62.40 ± 1.26o

46.38 ± 0.76 106.52 ± 1.41Coating surfaces 106.74 ± 1.46

o19.00 ± 0.86 51.84 ± 1.78

o

AwTtaw5cfMpaw

2

Asfsiscosshh

baf(Xo

Surface after cleaning and rinsing (n = 1)* 85.59 ± 2.59

fter each fouling experiment, the sample protein solutionas drained, and the entire loop was rinsed with water.he feed tank was again set at 61 ± 0.5 ◦C to ensure an inlet

emperature of 60 ± 0.5 ◦C. Operating conditions for cleaningre outlined in Table 3. Cleaning was performed for 600 s,ith photographs of rod surfaces taken after 0, 100, 180, 300,

00, and 600 s. 0.5 wt% NaOH solution was chosen as theleaning agent because the highest rate of removal of WPCouling has been observed at this concentration (Li et al., 2013;

ercadé-Prieto and Chen, 2006; Xin et al., 2002). After cleaningrocessing, the solution was recycled back to the sample cellnd expelled. Likewise, the entire loop was rinsed with clearater until the circulation system was cleaned.

.7. Surface analysis

series of analytical techniques were used to characterize theurface coating. In order to observe the effect of coating on sur-ace characteristics, the heating rods were replaced with flattainless steel 304 sheets. The flat plates were used to facil-tate the droplet test for surface tension measurement. Theame procedure as that for the heater rod was used for theoating of the flat plates, hence it is expected that the surfacebtained was similar. In a single experiment, a small and thinample of flat plate was suitably attached on the heater rodurface which was then subjected to the fouling (at the higheat flux case), cleaning, and water rinsing process as wouldave been in the normal tests in this study).

Contact angle, surface free energy, and work of adhesionetween surfaces and water were measured with a drop shapenalyzer (DSA 30, KRÜSS GmbH, Hamburg, Germany). Sur-ace topography and elemental content were analyzed by SEMscanning electron micrographs) and EDS (Energy Dispersive

-Ray Spectroscopy), respectively. Measurements were takenn three separate regions of each sample.

31.98 ± 1.62 78.38 ± 3.27

3. Results and discussion

3.1. Surface analysis

The coating’s effects on contact angle, surface energy, andwork of adhesion for a stainless steel surface are givenin Table 4. Fluorolink® S10 modification of stainless steelsurfaces reduced surface free energy from 46.38mN m−1 to19.00 mN m−1, decreased adhesion work between foulingdeposits and stainless steel surfaces, and increased contactangle. These changes indicate that the coating made the sur-faces more hydrophobic.

Fig. 3 illustrates the effects of the coating on the sur-face morphology and uniformity of stainless steel. The SEMmicrographs in 3(a) and (b) indicate that surface morphologyhas not changed significantly. The F and Si EDS images indi-cate the surface concentrations of these elements increasedsignificantly and were evenly distributed on the stainlesssteel surfaces. This demonstrates that the coating had beensuccessfully deposited on the stainless steel surface, and dis-tributed evenly.

Oldani (2015) dip-coated 304 stainless steel with similar1 wt% S10 solutions, which resulted in visible formation ofclusters of ∼1 to 10 �m diameter polymer spheres, and anincrease in average surface roughness from 0.192 to 0.438 �m.As our S10 modification did not result in visible changes insurface morphology, we expect its effect on roughness to beless than Oldani observed. Coating thickness from our 0.5 wt%is also likely to be lower than their 2.7 �m value. At that level,it should have negligible effect on heat transfer. Future workwill likely characterize S10 coatings in more detail, both beforefouling and after cleaning.

3.2. Effects of Fluorolink® S10 coating on WPC fouling

The morphology of deposits on coated and uncoated surfaces,under different heat flux conditions, are shown in Fig. 4. With

134 Food and Bioproducts Processing 1 1 8 ( 2 0 1 9 ) 130–138

Fig. 3 – SEM micrographs and EDS images (F and Si), of (a) unmodified (original) and (b) modified stainless steel surfaces.

Fig. 4 – Photographs of WPC fouling on heating rods after fouling experiments, and subsequent 60 ◦C air drying. Note thatdeposits have peeled off the coated rods during drying.

Food and Bioproducts Processing 1 1 8 ( 2 0 1 9 ) 130–138 135

Fig. 5 – Surface temperatures and heat transfer coefficientsfor WPC fouling on unmodified and Fluorolink® S10modified heating rod surfaces. Error bars are given bys

3sdp4toucfpmt

adritwmfara

Fig. 6 – Mass of WPC fouling deposits for unmodified andFluorolink® S10 modified heating rod surfaces coatingsurfaces under different heat fluxes. Experiments wereperformed in triplicate. Error bars are given by standarderror of the mean.

tandard error of the mean.

.417 kW m−2 heat flux, after fouling, the coated and uncoatedurfaces were mainly milky white and visibly uneven. Afterrying of the fouled rods, deposits on the coated surfaceseeled off, while those on the uncoated surfaces did not. At.985 kW m−2 heat flux, surface deposits were yellowish andhe surfaces were visibly uneven. Again after drying, depositsn the surface of the coated rod peeled off, but those on thencoated rod did not. This indicates that the Fluorolink® S10oating may have reduced the adhesion strength betweenouling deposits and stainless steel surfaces. This sheddinghenomenon further suggests that deposits formed on theodified surfaces were easier to remove. Cleaning results for

he fouling deposits are examined in Section 3.3.Fig. 5 shows changes in surface temperature (Fig. 5(a))

nd U (Fig. 5(b)) for uncoated and coated heating rods underifferent heat fluxes. Noting that these temperatures only rep-esent the surface temperatures qualitatively as they may benterfered by the presence of fouling. It was useful, however,racking these temperature does help understanding whatas going on and indeed they behave following the com-on wisdom about what would be expected in our extensive

ouling research experiences in the last 2 decades. Temper-tures initially rose rapidly (up to ∼100 s), as the bulk of the

ods heated up to a pseudo-equilibrium temperature afterpplication of electrical heating. Fouling is indicated by the

subsequent slower, but steady, increase in temperature. Atthe end of the experiment conducted with a heat flux of4.985 kW m−1 K−1, the temperature was still rising steadily,indicating that fouling is continuing. This seems less apparentat the lower heat flux.

As expected, calculated heat transfer coefficients reducedover the course of the experiment, consistent with increasingfouling heat transfer resistance, with the lower heat flux tracesleveling out, and effectively losing sensitivity, after ∼2000 s.Little difference was observed between U values for coatedand uncoated rods, except before ∼600 s under lower heatflux, where the coated rod displayed lower U. Higher mag-nitude gradients would be expected at higher heat fluxes,in both Fig. 5(a) and (b), as both temperature difference andU are linearly proportional to Q . Further, more fouling maybe expected at higher temperatures, as protein absorption isoften an entropy driven process, with gains due to release ofsurface adsorbed water and ions, and configuration changeswithin protein molecules (Rabe et al., 2011).

The extent of protein fouling on the heating rods is shownin Fig. 6. Slightly more deposition on the coated rods wasobserved. Andrade (1973) identified a thermodynamic ratio-nale for this — the hydrophobic, fluorinated surface of theFluorolink® S10 coating has reduced interactions with water(a polar solvent), raising the solution-material interface’s freeenergy excess, and providing more driving force for proteinadsorption. Conversely, if the lower free energy of the flu-orinated surface limited adhesion strength, which dependson the natures of the surface, solution, and proteins, thena greater fluid shear would have induced greater foulantremoval if the Re is high. However, in these fouling experi-ments, the fluid velocities were deliberately set so low (hencevery small Re) to examine the process under the ‘worst-case-scenario’ for fouling (i.e. causing the largest amount of foulingpossible).

Small influence of coatings on dairy product fouling of heatexchangers has been reported previously, for both milk (Patelet al., 2013), and a simulated dairy dessert cream mix (Beufet al., 2003). In the former, Patel et al. (2013) studied diamond-like carbon coatings (DLC) on stainless steel heat exchangers,

under considerably more shear than here (Re ∼1100 to 21,000).

136 Food and Bioproducts Processing 1 1 8 ( 2 0 1 9 ) 130–138

Fig. 7 – Photographs of WPC fouling deposits remaining onheating rods during the cleaning process — for uncoatedand coated rods, once-used coated rods, and re-coatedrods. Deposits were formed under heat flux conditions of3.417 kW m−2 and 4.985 kW m−2 heat fluxes.

These more hydrophilic coatings did not show a significantadvantage with milk. Under lab conditions (Re ∼21,000) therewas reduced fouling from whey protein isolate (WPI) solu-tions though. However, the total masses measured were low,with the authors urging caution in interpreting this result. Noadvantage for the coating was demonstrated in the lower Re(∼1200) pilot scale studies. Beuf et al. (2003) showed no sig-nificant benefit for a wide variety of coatings, but did findthat surfaces with low polar component of surface energyperformed better during cleaning — a PTFE based coating per-formed best. Their fouling experiments were performed at0.142 ms−1 velocity, and Re ∼120.

3.3. Effect of Fluorolink® S10 coating on cleaning ofwhey protein isolate

Cleaning of the fouled heating rods is shown in Fig. 7. WPIremoval continued over the entire 600 s period, with furtherremoval upon discharge of the cleaning fluid and replace-ment with water. None of the cleaned rods were completelyfree of fouling deposits, and per Fig. 6, had similar levels offouling prior to cleaning. Fig. 8 reveals the extent of foulingremoval. Approximately 75% of the fouling was removed fromunmodified rods, and ∼98% from the modified rods. The exper-iments were repeated with previously used modified rods, thathad been re-coated after two stages of ultra-sonicated clean-ing, and again ∼98% of fouling was removed. Previously usedmodified rods that were re-used without undergoing this re-coating procedure only had ∼75% of deposits removed. Thismay be due to removal of the coating via silane bond cleav-age under the strongly basic cleaning conditions. Increasedsurface energy and work of adhesion, and reduced contactangle, of a plate tested after fouling and cleaning (Table 4) mayindicate degradation of the coating, but the potential effectsof residual foulant on surface properties may also have influ-enced these measurements.

The improved cleaning is in line with thermal peelingof deposits observed earlier, and is consistent with obser-vations in the literature — specifically that lower energysurfaces (19.00 mJ m−1 here, cf. PTFE’s critical surface tensionof 18.5 mJ m−1 (Smart, 1994) can offer improved cleaning prop-erties. Saikhwan et al. (2006) suggested a value of 22–26 mN/mas optimum for this purpose. Both Britten et al. (1988) andBeuf et al. (2003) identified low polar interactions with water asthe key factor here, which may indicate that for milk proteinsthe stronger polar interactions they experience dominate theweaker dispersive ones. The perfluorinated surface modifi-cation may benefit from this property, as suggested by itsreduced work of adhesion with water, 51.84 vs. 106.52 J m-2

for an uncoated surface. However, as a lipocalin protein(Sawyer and Kontopidis, 2000), beta-lactoglobulin containshydrophobic domains, and increased irreversible adsorptionwith increased surface hydrophobicity has been observed(Krisdhasima et al., 1992). An alternative explanation maylie with the fouling deposit structure, or interface chem-istry, being more amenable to NaOH induced dissolution.Accordingly, future work may focus on detailed analysis offouling-heat exchanger interface chemistry, fouling depositstructure, and elucidation of cleaning mechanisms.

4. Conclusions

Fluorolink® S10 surface modification, showed no significantbenefit in fouling mitigation when compared with unmodified

stainless steel surfaces, in low shear conditions. Additionally,it was found that mass of the fouling deposits was unrelatedto surface energies. These findings are in contradiction withsome studies from the literature, where large benefits in foul-ing mitigation from WPI, WPC, and other dairy solutions onlow energy surfaces have been reported. Of course, a muchgreater shear condition may yield benefits, which needs fur-ther confirmation.

Cleaning of WPC fouling does improve with surface modi-fication, as 98% of the deposits were removed from the coated

surfaces, compared to 75% from the uncoated. It was also

Food and Bioproducts Processing 1 1 8 ( 2 0 1 9 ) 130–138 137

Fig. 8 – Removal of WPC fouling deposits for uncoated andcoated heating rods, once-used coated rods, and re-coatedr

nbmi

A

TN(2o

R

A

B

B

B

B

B

B

B

ods. Error bars are given by standard error of the mean.

oticed that the fouling deposition process might be affectedy the initial state, or early fouling, of the surfaces, as re-usedodified surfaces offered similar cleaning extents to unmod-

fied surfaces.

cknowledgments

his work was supported by project funding from theational Key Research and Development Program of China

International S&T Cooperation Program, ISTCP, Project No.016YFE0101200), and the Priority Academic Program Devel-pment (PAPD) of Jiangsu Higher Education Institutions.

eferences

ndrade, J.D., 1973. Interfacial phenomena and biomaterials.Med. Instrum. 7 (2), 110.

alasubramanian, S., Puri, V.M., 2008. Fouling mitigation duringproduct processing using a modified plate heat exchangersurface. Transactions 51 (2), 629–639.

alasubramanian, S., Puri, V., 2009. Reduction of milk fouling in aplate heat exchanger system using food-grade surfacecoating. Transactions 52 (5), 1603–1610.

ansal, B., Chen, X.D., 2006. A critical review of milk fouling inheat exchangers. Compr. Rev. Food Sci. Food Saf. 5 (2), 27–33.

arish, J.A., Goddard, J.M., 2013. Anti-fouling surface modifiedstainless steel for food processing. Food Bioprod. Process. 91(4), 352–361.

euf M., Rizzo G., Leuliet J., Müller-Steinhagen H., Yiantsios S.,Karabelas A., Benezech T., Fouling and cleaning of modifiedstainless steel plate heat exchangers processing milkproducts. In: Heat Exchanger Fouling and Cleaning: Fundamentalsand Applications, Paul Watkinson, University of BritishColumbia, Canada; Hans Müller-Steinhagen, GermanAerospace Centre (DLR) and University of Stuttgart; M. RezaMalayeri, German Aerospace Centre (DLR) Eds, ECISymposium Series, 2003.https://dc.engconfintl.org/heatexchanger/14.

oxler, C., Augustin, W., Scholl, S., 2013. Cleaning of whey proteinand milk salts soiled on DLC coated surfaces athigh-temperature. J. Food Eng. 114 (1), 29–38.

ritten, M., Green, M.L., Boulet, M., Paquin, P., 1988. Deposit

formation on heated surfaces: effect of interface energetics. J.Dairy Res. 55 (4), 551–562.

Christian, G., Changani, S., Fryer, P., 2002. The effect of addingminerals on fouling from whey protein concentrate:development of a model fouling fluid for a plate heatexchanger. Food Bioprod. Process. 80 (4), 231–239.

Da Cruz, L.G., Ishiyama, E., Boxler, C., Augustin, W., Scholl, S.,Wilson, D., 2015. Value pricing of surface coatings formitigating heat exchanger fouling. Food Bioprod. Process. 93,343–363.

De Jong, P., 1997. Impact and control of fouling in milkprocessing. Trends Food Sci. Technol. 8 (12), 401–405.

Fickak, A., Al-Raisi, A., Chen, X.D., 2011. Effect of whey proteinconcentration on the fouling and cleaning of a heat transfersurface. J. Food Eng. 104 (3), 323–331.

Fickak, A., Hatfield, E., Chen, X.D., 2012. Influence of run time andaging on fouling and cleaning of whey protein deposits onheat exchanger surface. J. Food Res. 1 (1), 212–224.

Fryer, P., Robbins, P., Green, C., Schreier, P., Pritchard, A., Hasting,A., Royston, D., Richardson, J., 1996. A statistical model forfouling of a plate heat exchanger by whey protein solution atUHT conditions. Food Bioprod. Process. 74 (4), 189–199.

Gao, Y., Huang, Y., Feng, S., Gu, G., Qing, F.L., 2010. Novelsuperhydrophobic and highly oleophobic PFPE-modified silicananocomposite. J. Mater. Sci. 45 (2), 460–466.

Gillham, C., Fryer, P., Hasting, A., Wilson, D., 1999.Cleaning-in-place of whey protein fouling deposits:mechanisms controlling cleaning. Food Bioprod. Process. 77(2), 127–136.

Hooper, R., Paterson, W., Wilson, D., 2006. Comparison of wheyprotein model foulants for studying cleaning of milk foulingdeposits. Food Bioprod. Process. 84 (4), 329–337.

Jun, S., Puri, V., 2005. Fouling models for heat exchangers in dairyprocessing: a review. J. Food Process Eng. 28 (1), 1–34.

Krisdhasima, V., McGuire, J., Sproull, R., 1992. Surfacehydrophobic influences on �-lactoglobulin adsorptionkinetics. J. Colloid and Interface Sci. 154 (2), 337–350.

Li, L., Lv, H.T., Deng, R.P., Liao, Z.K., Wu, X.E., Chen, X.D., 2013.Experimental investigation of egg ovalbumin scaling onheated stainless steel surface and scale-removal comparedwith that of whey protein. Colloid Surf. B 107, 198–204.

Lv, H., Huang, S., Mercadé-Prieto, R., Wu, X.E., Chen, X.D., 2015.The effect of pre-adsorption of OVA or WPC on subsequentOVA or WPC fouling on heated stainless steel surface. ColloidSurf. B 129, 154–160.

Mauermann, M., Eschenhagen, U., Bley, T., Majschak, J.P., 2009.Surface modifications—application potential for the reductionof cleaning costs in the food processing industry. Trend FoodSci. Technol. 20, S9–S15.

Mercadé-Prieto, R., Chen, X.D., 2006. Dissolution of whey proteinconcentrate gels in alkali. AIChE J. 52 (2), 792–803.

Merian, T., Goddard, J.M., 2012. Advances in nonfouling materials:perspectives for the food industry. J. Agr. Food Chem. 60 (12),2943–2957.

Molena, E., Credi, C., De Marco, C., Levi, M., Turri, S., Simeone, G.,2014. Protein antifouling and fouling-release inperfluoropolyether surfaces. Appl. Surf. Sci. 309, 160–167.

Müller-Steinhagen, H., Malayeri, M., Watkinson, A., 2011. Heatexchanger fouling: mitigation and cleaning strategies. HeatTransfer Eng. 32 (3), 189–196, Taylor Francis.

Oldani, V., 2015. Hydrophobic Coatings Based on CommercialPerfluoropolyethers for Fouling Mitigation. Approach on aPilot Heat Exchanger Plant.,http://dx.doi.org/10.13130/oldani-valeria phd2015-11-30.

Oldani, V., Del Negro, R., Bianchi, C.L., Suriano, R., Turri, S., Pirola,C., Sacchi, B., 2015. Surface properties and anti-foulingassessment of coatings obtained from perfluoropolyethersand ceramic oxides nanopowders deposited on stainless steel.J. Fluorine Chem. 180, 7–14.

Oldani, V., Bianchi, C.L., Biella, S., Pirola, C., Cattaneo, G., 2016a.Perfluoropolyethers coatings design for fouling reduction onheat transfer stainless-steel surfaces. Heat Transfer Eng. 37(2), 210–219.

Oldani, V., Sergi, G., Pirola, C., Sacchi, B., Bianchi, C.L., 2016b.Sol-gel hybrid coatings containing silica and a

138 Food and Bioproducts Processing 1 1 8 ( 2 0 1 9 ) 130–138

http://dx.doi.org/10.1021/acssuschemeng.8b05835, PublicationDate (Web): 25 April 2019.

perfluoropolyether derivative with high resistance andanti-fouling properties in liquid media. J. Fluorine Chem. 188,43–49.

Patel, J.S., Bansal, B., Jones, M.I., Hyland, M., 2013. Foulingbehaviour of milk and whey protein isolate solution on dopeddiamond-like carbon modified surfaces. J. Food Eng. 116 (2),413–421.

Rabe, M., Verdes, D., Seeger, S., 2011. Understanding proteinadsorption phenomena at solid surfaces. Adv. ColloidInterface Sci. 162 (1–2), 87–106.

Rosmaninho, R., Melo, L.F., 2008. Protein–calcium phosphateinteractions in fouling of modified stainless-steel surfaces bysimulated milk. Int. Dairy J. 18 (1), 72–80.

Sadeghinezhad, E., Kazi, S.N., Badarudin, A., Zubair, M.N.M.,Dehkordi, B.L., Oon, C.S., 2013. A review of milk fouling onheat exchanger surfaces. Int. Rev. Chem. Eng. 29 (3), 169–188.

Saikhwan, P., Geddert, T., Augustin, W., Scholl, S., Paterson, W.,Wilson, D., 2006. Effect of surface treatment on cleaning of amodel food soil. Surf. Coat. Technol. 201 (3), 943–951.

Santos, O., Nylander, T., Rosmaninho, R., Rizzo, G., Yiantsios, S.,Andritsos, N., Gabet, C., 2004. Modified stainless steel surfacestargeted to reduce fouling––surface characterization. J. FoodEng. 64 (1), 63–79.

Santos, O., Nylander, T., Schillén, K., Paulsson, M., Trägårdh, C.,2006. Effect of surface and bulk solution properties on theadsorption of whey protein onto steel surfaces at hightemperature. J. Food Eng. 73 (2), 174–189.

Santos, O., Anehamre, J., Wictor, C., Tornqvist, A. and Nilsson, M.,2013. Minimizing crude oil fouling by modifying the surface of

heat exchangers with a flexible ceramic coating. In: Proc. HeatExchanger Fouling & Cleaning X (pp. 79–84).

Sawyer, L., Kontopidis, G., 2000. The core lipocalin, bovine�-lactoglobulin. Biochim. Biophys. Acta 1482 (October (1–2)),136–148.

Smart, B.E., 1994. Characteristics of C–F systems. In: Banks, R.E.,Tatlow, J.C., Smart, B.E. (Eds.), Organofluorine Chemistry:Principles and Commercial Applications. Plenum Press, NewYork, NY [Chapter 3].

Tonelli, C., Di Meo, A., Fontana, S., Russo, A., 2002.Perfluoropolyether functional oligomers: unusual reactivity inorganic chemistry. J. Fluorine Chem. 118 (1), 107–121.

Xin, H., Chen, X., Özkan, N., 2002. Whey protein-based gel as amodel material for studying initial cleaning mechanisms ofmilk fouling. J. Food Sci. 67 (7), 2702–2711.

Zouaghi, S., Six, T., Bellayer, S., Moradi, S., Hatzikiriakos, S.G.,Dargent, T., Thomy, V., Coffinier, Y., André, C., Delaplace, G.,2017. Antifouling biomimetic liquid-infused stainless steel:application to dairy industrial processing. ACS Appl. Mater.Interfaces 9 (31), 26565–26573.

Zouaghi, S., Frémiot, J., André, C., Grunlan, M.A., Gruescu, C.,Delaplace, G., Duquesne, S., Jimenez, M., 2019. Investigatingthe effect of an antifouling surface modification on theenvironmental impact of a pasteurization process: an LCAstudy. ACS Sustain. Chem. Eng.,