International Communications in Heat and Mass Transfer 38 (2011) 730–733

Contents lists available at ScienceDirect

International Communications in Heat and Mass Transfer

j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / i c h m t

Investigation of adhesion of CaCO3 crystalline fouling on stainless steel surfaces with

different roughness☆
Yida Liu a, Yong Zou a,b,⁎, Liang Zhao a, Wen Liu a, Lin Cheng a
a
Institute of Thermal Science and Technology, Shandong University, Jingshi Road 17923, Jinan, P.R. China
b
Key Lab of Liquid Structure and Heredity of Materials, Ministry of Education, Shandong University, Jingshi Road 17923, Jinan, P.R. China

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

Available online 13 April 2011 Fouling adhering experiments on AISI 304 stainless steel surfaces with different roughness had been
performed in boiling supersaturated calcium bicarbonate solution. The effect of surface roughness on
Keywords: adhesion of fouling is limited, and the adhesion of fouling does not have a simple linear relationship with the
Surface roughness surface roughness of samples. The surface with roughness in middle is more easily induced to form
Crystalline fouling “transitional interface” which connects the fouling and matrix surface. It is also found that the crystalline
Calcium carbonate
types of fouling are changed in the fouling process due to the variation of metallic ions in reaction solution.
Corrosion resistance
© 2011 Elsevier Ltd. All rights reserved.

1. Introduction
Calcium carbonate is the typical fouling adhered on the surface of
heat exchangers in China. Its adhesion depends on two classes of
parameters: physicochemical properties of water solution (types and
concentration of metal ions) and heat transfer surface properties (the
surface free energy, surface roughness and so on). Among the surface
properties, the reports focused on the surface roughness are limited.
In the previous studies, the experimental conclusions are appar-
ently contradictory. Keysar et al. [1] showed that the adhesion
strength of the fouling layer on the mild steel surface was remarkably
influenced by the degree of surface roughness. However, in recent
literature, Al-Anezi et al. [2] showed the stainless steel sample with
Ra = 0.33 μm had the lowest adhesion while sample with
Ra = 0.038 μm was secondary. Surprisingly, sample Ra = 0.12 μm,
whose roughness is in the middle, had the highest adhesion.
On the other hand, this problem confused the study of the fouling
forming mechanism. The net force of interaction between fouling and
the surface has been described predominantly by the sol particle
interactions in terms of colloid chemistry [3]. Furthermore, studies on
the fouling mass transfer modeling focus mainly on macroscopic
particles, such as Kern–Seaton Model. In spite of few special cases [4],
most studies have proved that the surfaces with higher surface free
energy would prompt the adhesion of crystalline fouling. These
conclusions have been widely used to further explore anti-fouling
surface materials [5]. However, fouling modeling based on the above
theories cannot predict the effect of surface roughness on the
precipitate fouling precisely. Förster and Bohnet [6] argued that
mechanical forces at the interface crystal/heat transfer surface are
responsible for discrepancies between reality and theory which is
described by current models. In fact, the idea of “mechanical
adhesion” was put forward in 1925 by Mcbain and Hopkins [7] and
has been substantiated unreasonable by empirical study by Packham
and Johnston [8]. Besides the research on the Ca1-xMgxCO3 [9]
solubilities formed in calcite growth in the presence of Mg2+, few
researchers noticed the application of quantum mechanical theory
and lattice structure in the field of fouling. In the previous studies, one
important problem is ignored, that is, whether the crystal type of
fouling is changeless in the fouling process. In other words, it is not
clear whether the type of crystalline fouling will be affected by the
surface corrosion or oxidation. At least, no opening literatures
reported this issue.
In our previous studies, we have reported “transitional interface”
connecting fouling and matrix, and this “transitional interface”
behaves selective absorption, that is, fouling tends to grow on it
more easily if fouling has similar lattice structure and parameters as
the “transitional interface” [10]. This “transitional interface” may be
the corrosion or oxidation products of the metal matrix. In the present
work, the adhesion behavior of CaCO3 crystalline fouling on AISI 304
stainless steel samples with different roughness was studied by using
fouling weight measurement. Furthermore, the change of crystalline
type of fouling was tracked.
2. Experimental procedures

In order to minimize the effect of oxide layer on the fouling

☆ Communicated by P. Cheng and W.Q. Tao.
⁎ Corresponding author at: Institute of Thermal Science and Technology, Shandong
adhesion, the stainless steel AISI 304 samples with a size of
University, Jingshi Road 17923, Jinan, P.R. China. 25 mm × 25 mm × 0.8 mm were employed as measured specimen.
E-mail address: yzou@sdu.edu.cn (Y. Zou). The samples were firstly prepared by wire-electrode cutting, and then

0735-1933/$ – see front matter © 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.icheatmasstransfer.2011.04.003
 Y. Liu et al. / International Communications in Heat and Mass Transfer 38 (2011) 730–733 731

No. 6, polished by polishing machine, did not show the obvious anti-
Nomenclature
fouling effect.
Polynomial fitting is applied to do the regression analysis for the
Ra arithmetical mean deviation of the profile (μm)
data shown in Fig. 1. The results are summarized in Table 1. In this
Iarag the integrated diffraction peak intensities of aragonite
formula, “b” means the initial slope of the curve at the x-value of
(111) at about 26.2° (a.u.)
t = 0 h, which represents the rate of fouling adhesion. The samples
Icalc the integrated diffraction peak intensities of calcite
with middle roughness exhibit a slight higher slope, although this
(104) at about 29.5° (a.u.)
difference is very small. The lowest surface roughness did not show
especial role on fouling adhesion. This means, even paying a
Greek symbols
considerable cost to polish the sample, it still cannot improve its
ωarag the mass fraction of aragonite phase (%)
anti-fouling behavior ideally. At the same time, the sample No.1 with
ωcalc the mass fraction of calcite phase (%)
the highest surface roughness did not show more adhesion of fouling.
Subscript
3.2. Surface morphology and crystalline phases of deposits
arag aragonite phase
calc calcite phase
Fig. 2 shows the SEM morphology of different samples surface after
being boiled in reaction solution for 15 min. No obvious morpholog-
ical difference can be seen. Compared to other area, the surface groove
grinded by abrasive papers with mesh size of 280, 400, 600, 800 and did not show selective fouling adhesion. That is, the fouling is
1000, respectively. These samples are numbered as No.1, No.2, No.3, randomly distributed on the surface. This conclusion is consistent
No.4 and No.5, respectively. Another sample polished by metallo- with Fig. 1. Previous studies [11] reported the complex change of
graphic polishing machine was denoted as No.6. Their surface crystal morphology and structures during the CaCO3 growth. McBain
roughness measured by Mitutoyo SJ 201 surface roughometer is and Hopkins [7] proposed that adhesion should be divided into two
0.246, 0.184, 0.156, 0.106, 0.062, and 0.018 μm, respectively. All the kinds: namely, specific adhesion that is related to adsorption and
samples were cleaned with acetone and distilled water in supersonic mechanical adhesion. In the further research, the latter has been
cleaner before fouling adhesion testing. The testing solution is boiling controversial [8]. It is well accepted that interfacial roughness has a
supersaturated calcium bicarbonate solution containing approximate- marked effect on the adhesion involving polymer-metal or polymer-
ly 360 mg/dm3 Ca and 1140 mg/dm3 HCO− 3 . These samples were polymer [12]. In the present study, we are interested in the phase
immersed to the test solution in the same condition and taken out to constitution of crystalline fouling and try to analyse the fouling layer
measure the weight every 4 h. In order to reduce the effect of gravity from fouling type perspective. Therefore, the precipitation fouling is
deposition on the adhesion of fouling, all the samples were suspended collected respectively for the different time and then examined by
into the boiling solution vertically. During the process of fouling XRD.
adhesion, the precipitate from the boiling solution was collected to Fig. 3 shows the XRD results for precipitation fouling at different
investigate the phase constitution of the fouling by using D8 Advance time. It is indicated that the fouling is calcium carbonate but a mixture
X-ray diffractometry (XRD) with CuKα radiation at 30 kV and 40 mA. of two crystalline phases, which are aragonite with a main peak at
The morphological features of the CaCO3 deposits on the sample 26.2° and calcite with a main peak at 29.5°. It also notices the relative
surface were observed by a SU-70 scanning electron microscopy intensity of these samples is varied. For calculating the relative
(SEM). content of calcite and aragonite precisely, the quantitative phase
contents of calcite and aragonite in a sample are used, which are
derived from mixing powder method [13].
3. Results and discussion
Iarag
ωarag = ð1Þ
3.1. Fouling weight data Iarag + 0:41Icalc

Fig. 1 shows the relations between the crystallization fouling Icalc

ωcalc = ð2Þ
weight per unit area and the reaction time for the samples. As 2:44Iarag + Icalc
illustrated in Fig. 1, the sample No. 1 with the roughest surface
exhibits lowest fouling weight in most test points, while the sample Calculating results are shown in Fig. 4.
The above results indicate that the form of crystalline fouling is not
0.006 unalterable in the fouling process. That is, there is a lot of aragonite in
fouling weight per unit area /(g·cm-2)

No.1 the early stages, and the relative amount of calcite increases gradually
No.2 with the increase of testing time, while the total amount of calcite and
No.3
No.4 aragonite keeps constant. In our recent report, it is found that only
0.004 No.5
No.6 calcite single phase is formed in the simulated hard water (CaCl2 and

Table 1
0.002 Regression results by using polynomial fitting.

Sample no. y = ax2 + bx + c Regression

error
a b c
0
No.1 − 0.00000157 0.00019046 0.00025167 0.9892460
0 4 8 12 16 20 24 28 No.2 − 0.00000421 0.00028483 − 0.00014733 0.9886456
No.3 − 0.00000506 0.00031596 0.00015100 0.9884100
time / hours
No.4 − 0.00000468 0.00030887 0.00016367 0.9948849
No.5 − 0.00000515 0.00032412 0.00022800 0.9911033
Fig. 1. Fouling weight per unit area versus time immersed in boiling reaction solution
No.6 − 0.00000449 0.00029240 − 0.00000533 0.9852714
for samples with different roughness.
732 Y. Liu et al. / International Communications in Heat and Mass Transfer 38 (2011) 730–733

Fig. 2. SEM micrographs depicting the temporal growth of deposits 15 min after the start of the run.

NaHCO3) if there is no other metal ion in reaction solution [14]. But in
this study, considerable aragonite was found in the early stages. This
means there are other metallic ions in the solution. The most
reasonable reason is some reactions like electrochemical corrosion
occur on the sample surfaces in the boiling solution, and then some
metallic ions enter into the solution surrounding the specimens.
Although the amount is very little, these metallic ions could prompt
the formation of aragonite. Similar results about the effect of metallic
ions on aragonite forming have been reported, such as the role of
Mg2+ [9,15]. Coetzee et al. [16] reported both Fe2+ and iron oxides
are responsible for the growth of aragonite crystals. Reffass et al.
[17] emphasized that as a consequence of the corrosion, the release
of iron into the bicarbonate/carbonate-containing solutions can
generate soluble Fe2+, and then act as many kinds of complexing

100
calcite
Content of Aragonite / mass%

aragonite
80
Intensity / a.u.

60

20h
40
12h

8h 20
4h

15 min 0
15min 4h 8h 12h 20h
20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60
Reaction Time

Fig. 4. Calculated results of aragonite content in samples at different reaction time (the
Fig. 3. X-ray diffraction patterns of crystalline fouling collected at different fouling time. total amount of calcite and aragonite as sum keeps constant).
 Y. Liu et al. / International Communications in Heat and Mass Transfer 38 (2011) 730–733
compounds with OH− and HCO− 2−
3 /CO3 . In this study, the corrosion
may be caused both by chloride ions and by CO2− 3 ions.
In our previous studies, it is found that a “transitional interface”
works as a “bridge” to connect the matrix and fouling, and its property
decides the adhesion of the crystalline fouling, and this “transitional
interface” may be the product of oxidation or corrosion. Therefore, the
effect of surface roughness on adhesion of fouling can be attributed to
the effect of roughness on corrosion resistance of sample surface. The
easily corroded surface also easily formed transitional interface which
can further prompt the adhesion of fouling. In fact, it had been
reported that the machining parts with Ra N 0.08–0.16 μm can easily
be oxidized in atmosphere [18]. This is caused by the formation of
microbattery due to the local moisture film and the local concentra-
tion difference of O2 in this range of roughness [19]. In contrast,
stainless steel samples with lower or higher than this range of
roughness can exhibit higher oxidation resistant when these samples
were ground by the silicon carbide papers. Meng et al. [20] reported
that the stainless steel specimens ground by the silicon carbide papers
of mesh size 800 show the lowest corrosion resistance in weak acid
surrounding. It is thought that the exposed cathodic metal particle of
specimen ground by mesh size 800 (Ra = 0.161 μm) is more than that
of abrasive grinding specimen (Ra = 0.424 μm). Therefore, the active
size of pitting corrosion is less than that of abrasive grinding
specimen. At the same time, the metastable and stable potential of
pitting corrosion for specimen grinded by mesh size 800 is lower than
that of grinded by mesh size 1000. It drew a conclusion that the
corrosion resistance of the metallic matrix did not increase monot-
onously with the decrease of surface roughness. When the surface
roughness is enough large or small, the surface of samples show a
relative good corrosion resistance. Therefore, the surface with
roughness in middle is more easily induced to form “transitional
interface.”
Considering the present experimental results, the adhesion of
fouling does not have a simple linear relationship with the surface
roughness, but is intimately related to the corrosion resistance of the
matrix. The effect of surface roughness on adhesion of fouling is
limited and thus the practical improvement of anti-fouling property of
surface is not available by machining with lower surface roughness. To
improve the anti-fouling property, surface modification for iron-base
materials is much more valuable.
4. Conclusions
Through the fouling adhering experiments on stainless steel
surfaces with different roughness in supersaturated calcium bicar-
bonate solution, some results are summarized as follows:
(1) The effect of surface roughness on adhesion of fouling is
limited, and the adhesion of fouling does not have a simple
linear relationship with the surface roughness of sample. The
polished surface with lowest surface roughness cannot obvi-
ously improve anti-fouling property. The anti-fouling property
of surface is related to the corrosion resistance of the matrix.
The sample with the middle roughness has the slight poor anti-
fouling property, and also has the poor corrosion resistance.
(2) The form of crystalline fouling is changed in the fouling process.
That is, there is a lot of aragonite in the early stages, and the
relative amount of calcite increases gradually with the increase
733
of testing time. The reason for this change is the formation of
metal ions in solution due to the oxidation or corrosion of the
sample surface. The effect of surface roughness on the fouling
adhesion is essentially the effect on the formation of “transi-
tional interface” via corrosion or oxidation. The surface with
roughness in middle is more easily induced to form “transi-
tional interface” which connects the fouling and matrix surface.
Acknowledgements
This work has been supported by the National Basic Research
Program of China (2007CB206900) and the Natural Science Founda-
tion of Shandong Province (Z2007F05).
References
[1] S. Keysar, R. Semiat, D. Hasson, J. Yahalom, Effect of surface roughness on the
morphology of calcite crystallizing on mild steel, Journal of Colloid and Interface
Science 162 (2) (1994) 311–319.
[2] K. Al-Anezi, D.J. Johnson, N. Hilal, An atomic force microscope study of calcium
carbonate adhesion to desalination process equipment: Effect of anti-scale agent,
Desalination 220 (1–3) (2008) 359–370.
[3] R. Oliveira, Understanding adhesion: A means for preventing fouling, Experi-
mental Thermal and Fluid Science 14 (4) (1997) 316–322.
[4] B.-m.Z. Newby, M.K. Chaudhury, H.R. Brown, Macroscopic evidence of the effect of
interfacial slippage on adhesion, Science 269 (5229) (1995) 1407–1409.
[5] H. Müller-Steinhagen, Q. Zhao, Investigation of low fouling surface alloys made by
ion implantation technology, Chemical Engineering Science 52 (19) (1997)
3321–3332.
[6] M. Förster, M. Bohnet, Modification of molecular interactions at the interface
crystal/heat transfer surface to minimize heat exchanger fouling, International
Journal of Thermal Sciences 39 (7) (2000) 697–708.
[7] J.W. McBain, D.G. Hopkins, On adhesives and adhesive action, The Journal of
Physical Chemistry 29 (2) (1925) 188–204.
[8] D.E. Packham, C. Johnston, Mechanical adhesion: were McBain and Hopkins right?
An empirical study, International Journal of Adhesion and Adhesives 14 (2)
(1994) 131–135.
[9] K.J. Davis, P.M. Dove, J.J.D. Yoreo, The role of Mg2+ as an impurity in calcite
growth, Science 290 (5494) (2000) 1134–1137.
[10] L. Cheng, Y. Zou, J.W. Song, W. Liu, Phase constitution of crystallization fouling and
interface bonding with the surface of heat exchanger, in: C.S.E.T. (Ed.),
Proceedings of the CSHMT Conference, Chinese Society of Engineering Thermo-
physics, Guangzhou, China, 2007, pp. 1919–1923.
[11] D. Gebauer, A. Völkel, H. Cölfen, Stable prenucleation calcium carbonate clusters,
Science 322 (5909) (2008) 1819–1822.
[12] D.E. Packham, Surface energy, surface topography and adhesion, International
Journal of Adhesion and Adhesives 23 (6) (2003) 437–448.
[13] L. Zhao, Y. Zou, Y.D. Liu, W. Liu, L. Cheng, Effect of temperature on formation of
crystalline fouling on heat exchanger surface, Journal of Chemical Industry and
Engineering (China) 60 (8) (2009) 1938–1943.
[14] L. Zhao, Y. Zou, Y.D. Liu, W. Liu, L. Cheng, Effect of reaction mode on properties of
CaCO3 crystalline fouling, in: C.S.E.T. (Ed.), Proceedings of the CSHMT Conference,
Chinese Society of Engineering Thermophysics, Qingdao,China, 2009, p. 442.
[15] M. Kitamura, Crystallization and transformation mechanism of calcium carbonate
polymorphs and the effect of magnesium ion, Journal of Colloid and Interface
Science 236 (2) (2001) 318–327.
[16] P. Coetzee, M. Yacoby, S. Howell, S. Mubenga, Scale reduction and scale
modification effects induced by zinc and other metal species in physical water
treatment, Water SA 24 (1) (1998) 77–84.
[17] M. Reffass, R. Sabot, C. Savall, M. Jeannin, J. Creus, P. Refait, Localised corrosion of
carbon steel in NaHCO3/NaCl electrolytes: role of Fe(II)-containing compounds,
Corrosion Science 48 (3) (2006) 709–726.
[18] J. Ren, Handbook of common difference and cooperation, 1 ed, Mechanical
Industry Press, Beijing, 1990, pp. 196–197.
[19] Z. Zeng, Anti-rust application for metals, 1 ed, New Times Press, Beijing, 1989,
pp. 28–29.
[20] B. Meng, W. Guo, Y. Jiang, Effect of surface morphology on corrosion resistance of
1Cr11Ni2W2MoV stainless steel, Materials for Mechanical Engineering (China) 32
(8) (2008) 8–11.