You are on page 1of 14

Renewable Energy 108 (2017) 555e568

Contents lists available at ScienceDirect

Renewable Energy
journal homepage:

Investigation of the corrosive properties of phase change materials in

contact with metals and plastic
Maria C. Browne*, Ellen Boyd, Sarah J. McCormack
Department of Civil, Structural and Environmental Engineering, Trinity College Dublin, University of Dublin, Dublin 2, Ireland

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

Article history: Phase change material (PCM) is used as a means of storing thermal energy. The method of PCM
Received 11 July 2016 containment in thermal energy storage (TES) systems is important as the material is in direct contact
Received in revised form with metal piping, plates and housing units which can be damaged if the material is corrosive. The
22 February 2017
compatibility of PCM with various materials for use in TES applications has been investigated using the
Accepted 27 February 2017
immersion corrosion method and suggestions to minimise their corrosive behaviour or to mitigate it by
using an alternative PCM have been presented and are described in the following paper. The metals and
plastics tested were copper, brass, aluminium, stainless steel, mild steel and Perspex. These were
Phase change material
immersed in ve different PCM, three fatty acids, a salt hydrate and Micronal for a period of 722 days.
Thermal energy storage Stainless steel was determined to be the most suitable of all the PCM investigated. Aluminium can be
Containment used with all fatty acids, however caution is advised as the maximum corrosion rate was found to be
Corrosion 12.4 m g/cm2year. Copper and brass can be encapsulated using fatty acid and again caution is advised as
Metal and plastic corrosion rates were found to be 22.15 g/cm2year and 1.67 g/cm2year, respectively.
2017 Elsevier Ltd. All rights reserved.

1. Introduction mild, boiler and austentic steels and the inuence of water and
other impurities was carried out and found LiCl-LiNo3-NaNO3 salt
Phase change materials (PCM) absorb thermal energy as latent mixtures lead to surface corrosion in all steels, NaCl-NaNO3 is
heat at a constant phase change temperature. PCM with a suitable compatible with all steels and KCl-LiCl are much more aggressive
phase transition temperature can be used to as a means of storing [17].
thermal energy for later use [1e3] or as a method of temperature Cabeza et al. carried out a series of short- [18], medium- [19] and
regulation of PV [4e7]. There has been considerable research on long- [20] term corrosion tests. The process involved immersing the
types of PCM, their applications and thermophysical properties metal sample in melted PCM in a test tube which remained in a
[8e12]. water bath for the duration of the experiment to ensure the PCM
PCM encapsulation can take the form of micro or macro remained in a liquid state. The short-term investigation involved
encapsulation [13e15]. With regard to macro-encapsulation, it is leaving the metal samples immersed in the following salt hydrates;
important the encapsulation material in the TES system is resistant zinc nitrate hexahydrate, sodium hydrogen phosphate dodecahy-
to corrosion by the PCM to ensure the system remains fully func- drate and calcium chloride hexahydrate with melting temperatures
tional for its lifetime. of 36  C, 35  C and 32  C respectively for a period of two weeks [18].
The availability of data on the corrosive nature of PCM was very The zinc nitrate hexahydrate emerged as a very corrosive material
limited until an investigation into the corrosion resistance of as aluminium and steel were highly corroded. Calcium chloride
metallic alloys to hydrated salts was carried out where stainless hexahydrate produced a pitting effect in aluminium and stainless
steel appeared to be the most corrosion resistant and aluminium steel. The samples continued to be immersed for 75 days [19].
the least [16]. Further investigation into the compatibility of high The metals previously mentioned were immersed in two PCMs
phase change temperature (160  C - 420  C) salt eutectics with TH29 and a chemical composition of MgCl2$6H20 for a period of
500 days in the long term corrosion tests [20]. As graphite can be
used as a method to increase the thermal conductivity of PCM, the
* Corresponding author. study immersed both the graphite and metal pieces in PCM.
E-mail addresses: (M.C. Browne),
Aluminium and steel with either PCM are not recommended, brass
(E. Boyd), (S.J. McCormack).
0960-1481/ 2017 Elsevier Ltd. All rights reserved.
556 M.C. Browne et al. / Renewable Energy 108 (2017) 555e568

and copper are recommended with both PCM and stainless steel is the corrosion rate. Finally, there has been no investigation into the
recommended for containment of PCM with caution. effect of PCM on Perspex a commonly used material which can be
Investigations into the compatibility of commercially available easily manipulated. The corrosive nature of PCM in TES systems is
PlusICE E17 and Climsel C18, two salt hydrates with melting tem- an important factor to be considered during system design as it will
peratures of 17  C and 18  C respectively, with an aluminium alloy impact on the lifetime of the system and thus, the economic ad-
and copper separately and aluminium in the presence of copper vantages of the system.
while immersed in an electrolyte [21]. The samples were immersed The methodology presented involves immersing ve materials,
in liquid PCM for 17 h. It was found copper samples experienced including metals and a plastic, in ve different PCMs for an
greatest mass loss varying between 0.8 and 0.145 g/(m2h) for extended period of time. The investigation records the rate of mass
ClimSel C18 and PlusICE E17, respectively. Aluminium showed loss of the metals and plastic at specied intervals over the course
pitting corrosion at a mass loss of 7  105 g/(m2h) while in contact of the experiment. In addition to this, a series of Scanning Electron
with ClimSel C18 with negligible mass loss when immersed in Microscopy (SEM) images were taken to assess the corrosion on a
PlusICE E17. For combined samples of aluminium and copper, no microscopic level.
mass loss occurred on the copper due to the cathodic protection but As the study presented is carried out under real time conditions,
the corrosion rate increased by a factor of 8 in Climsel 18 compared it presents a more detailed account of the natural corrosion process
to aluminium alone [21]. Garca-Romero, Delgado, Urresti, Martn of material samples immersed in PCM compared to accelerated
and Sala [22] presented the corrosion behaviour of four aluminium methods. The paper shows the calculation of the rate of corrosion of
alloys (EN AW 2024, 3003, 6063 and 1050) in the commercial salt each sample by the individual PCM. This provides an insight into
hydrate Glaubers salt (Na2SO4$10 H20) which has a melting point of the extent of the corrosive nature of PCM and can be used to ensure
32  C. Samples were fully and partially immersed in the PCM and the optimum encapsulation material is chosen during system
experiments were conducted for a period of up to 90 days. It was design. Furthermore, SEM imaging can be used for a greater insight
found the aluminium alloy 2024 was not compatible with this into the initial stages of corrosion which are not initially visible to
specic salt hydrate, however, 3003, 6063 and 1050 were shown to the naked eye.
be fully compatible.
An investigation into the corrosion resistance of metal samples, 2. Materials and methodology
stainless steel, carbon steel, aluminium and copper to fatty acids
stearic, palmitic, myristic and lauric acid with melting tempera- This section describes the material samples assessed for corro-
tures between 40  C - 63  C [23]. Following 910 thermal cycles of sion resistance, in addition to the PCM investigated. The method-
being fully immersed in each PCM, the metal samples were ology of the experiment is presented to show the approach
removed and a gravimetric analysis carried out. The highest undertaken. Finally, a brief description of scanning electron mi-
corrosion rates calculated were 3.4  102 and 4.9  102 mg/cm2 croscopy is given as images of the samples were taken on specic
for carbon steel in contact with myristic acid and copper in contact days throughout the experiment to assess the level of corrosion at
with stearic acid. Stainless steel with a chromium oxide layer and an increased magnication.
aluminium with an aluminium oxide layer were found to be
compatible with fatty acids. 2.1. Material samples
An investigation into the compatibility of four organic and one
salt hydrate PCM (namely a molecular alloy C16-C18, RT20, RT26, Five materials that are commonly used in TES system designs
RT25 and TH24) and plastics (high density polyethylene, low den- and construction [5,29e31] were tested in the experiment. The
sity polyethylene, polyethylene terephthalate and polypropylene) metals were aluminium, brass, copper, stainless steel, steel and the
was carried out [24]. A furnace was used to cycle the plastic bottles plastic used was Perspex. The chemical compositions of the mate-
to increase the rate of phase change. LSPE was discarded as an rials tested are presented in Table 1 using a Tescan Mira Scanning
encapsulation material due to the mass variations and large de- Electron Microscope. The samples were cut from larger pieces of
formations caused by the PCM. Polyethylene terephthalate and materials and each had dimensions of 50 mm  14 mm and a depth
polypropylene were shown to be the best encapsulation material of of 2 mm.
the organic PCM and high density polyethylene for inorganic PCM.
Polypropylene, polyethylene terephthalate and low density poly- 2.2. Phase change material (PCM) samples
ethylene are not recommended for using with TH24.
Further immersion corrosion tests have been carried out for a Fatty acids are an organic PCM and have many superior prop-
period of 84 days on aluminium, stainless steel, carbon steel and erties including high heat capacity, congruent melting, little or no
with salt hydrates, salt hydrate eutectics and fatty acid eutectics supercooling, non-toxic, good thermal and chemical stability, low
[25e27]. The results suggest that stainless is recommended for use cost, non-ammable and small volume change [32e34]. They are
with any of the PCM studied and all other metals would either be known to have high melting temperatures which are outside the
recommended with caution or not be recommended at all. Fatty range of thermal energy storage applications [35]. To overcome
acid eutectic of capric:palmitic (75.2:24.8%) show that copper is the this, a eutectic mixture of fatty acids is two or more fatty acids
only material tested which is not suitable for containment of the combined to form a binary fatty acid with a lower phase change
PCM. However, these tests were carried out for a short period of temperature and feature the same stability of a single component.
time and further analysis should be carried out to draw accurate The advantage of fatty acid eutectics is the ability to gain a lower
conclusions. Oro , Barreneche, Farid and Cabeza [28] has also melting temperature with all the material advantages of a fatty acid
investigated the compatibility of metal and polymers with PCM by varying the mass ratio of the mixed fatty acids [35,36]. Capric
with lower phase change temperatures (22  C to 16  C). acid and two fatty acid eutectics, capric:palmitic and capric:lauric
There has been no research carried out on eutectic fatty acids, acid have been investigated. Capric:palmitic acid is a mixture of
particularly capric, palmitic and lauric acid which have often been 75% capric and 25% palmitic by weight and capric:lauric acid is a
used in thermal energy storage systems. The phase change process mixture of 45% capric and 55% lauric acid.
has been accelerated using furnaces and thermal cycling however, Inorganic salt hydrates consist of salt and water which combine
due to the reduced time immersed in PCM may potentially affect in a crystalline matrix when the material solidies. Different salts
M.C. Browne et al. / Renewable Energy 108 (2017) 555e568 557

Table 1
Composition of metals and plastic samples.


Aluminium Brass Copper Stainless steel Steel Perspex

Chemical composition (%) Aluminium (Al) >99.08 <0.5

Titanium (Ti) <0.5
Zinc (Zn) <0.5 >35.2
Silicon (Si) <0.5
Magnesium (Mg) <0.5
Copper (Cu) <0.5 <64 >99.95
Iron (Fe) <0.5 <0.5 >69 >99.26
Manganese (Mn) <0.5 <2.0 <0.5
Nickel (Ni) <0.5 <14
Lead (Pb) <0.5 >0.5
Phosphorous (P) <0.045 <0.5
Bismuth (Bi) >0.5
Carbon (C) <0.08 <0.5
Chromium (Cr) <18
Sulfur (S) <0.03 <0.5
Molybdenum (Mo) <2
Nitrogen (N) <0.1
Other Synthetic polymers

are used in combination with water to alter the phase change 2.3.2. Corrosion test of metal and plastic materials
temperature. As the crystal is heated it melts, releasing the water of Following sample preparation, each material sample was placed
crystallisation and allowing the salt to dissolve in the water [37]. in a test tube and immersed in PCM. The samples were completely
They are attractive materials for TES due to their high thermal covered in PCM and the test tube was sealed to avoid ingress of
conductivity [38] and as a result the commercially available salt moisture and water from the water bath. There was no attempt
hydrate SP22 was used in the investigation. made to avoid contact with oxygen in the air so as to match as
Micronal is a material designed for temperature regulation closely as possible the expected on-site conditions within a thermal
applications and its performance has been investigated within storage unit. The test tubes were then placed in a water bath at a
buildings [39]. The PCM wax mixture, classied as organic parafn, constant temperature of 45  C to ensure all PCMs remained in a
is contained in small vessels (particle diameter 5 mm) with the core liquid phase and were not stirred.
PCM and a hard covering material [40].
A full list of PCM investigated is presented in Table 2 along with 2.3.3. Sample weighing
thermophysical properties. The phase change range and heat of The samples were then cleaned with acetone to remove any
fusion have been found using the differential scanning characteri- impurities from the material surface. Each material sample was
sation technique and are in agreement with previously reported removed from the PCM every 7 days, cleaned with tap water to
literature [6,8,41]. remove any excess PCM and then weighed using a Kerns PCB to the
nearest 0.01 g. This process was repeated every 7 days for 722 days
in total. In addition to this the samples were visually evaluated for
2.3. Methodology bubbles, precipitates, surface changes and pitting.

The methodology used in this study was based on the Standard 2.3.4. Corrosion assessment
Practice, Cleaning, and Evaluating Corrosion Test Specimens (ASTM The evaluation of the corrosion results involved completing a
G1-01) and has been applied to corrosion investigations of PCM and gravimetric analysis which consisted of the mass loss of the sample
metals [18e20,25]. The process was divided into four steps. with respect to the initial mass as in Eq. (1).

Dm mt0  mt (1)
2.3.1. Preparation of the sample The corrosion rate (CR) (mg/cm year) of the samples was then
The preparations of the samples required cutting them from determined by taking into account the mass loss (Dm), the area of
larger pieces of materials to approximately 50 mm  14 mm and a the material (A) and the experimental time (t0 - t) and applying
depth of 2 mm. The samples were then cleaned with acetone to them in Eq. (2).
remove any impurities from the material surface. Prior to the
experiment the samples were weighed using weighing scales (Kern CR (2)
scales accuracy of 0.01 g) and the values were recorded. A$t0  t

Table 2
List of materials and PCM used in corrosion experiments.

PCM Material classication Phase change range ( C) Heat of fusion (kJ/kg)

Capric acid Fatty acid 27.5e32.7 167

Capric: palmitic acid Eutectic fatty acid 17.7e22.8 191
Capric: lauric acid Eutectic fatty acid 18.9e22.3 116
SP22 Eutectic of fatty acid/parafn wax 16.9e25.3 146
Micronal Parafn 20.9e25.8 119
558 M.C. Browne et al. / Renewable Energy 108 (2017) 555e568

Another method to calculate corrosion loss on an hourly basis

determined by the American Society for Testing Materials [42] has
been applied [21] in previous studies. However, as the presented
investigation is conducted over a period of time beyond a year, Eq.
(2) is used to calculate the corrosion rate.
Recommendations were made based on the results of the
corrosion tests and suggested which materials are most compatible
as a containment method for specic PCM. The corrosion weight
loss guide used in industry practices was used as a reference, in
Table 3.

2.4. Scanning electron microscope (SEM)

SEM is a scanning electron microscope that produces images by Fig. 1. Material mass loss due to immersion in capric acid.
scanning it with a focused beam of electrons. The electrons interact
with atoms in the sample material, producing various signals that
can be detected and that contain information about the samples
surface topography. To ensure the samples were fully visually
examined a Tescan MIRA scanning electron microscope (SEM) was
used to take images on Day 1, Day 35, Day 200 and Day 500. SEM
allowed for images to be taken at magnications of 200, 3000
and 6000 which allowed for the inspection of initial signs of
corrosion not visible to the naked eye.

3. Results and discussion

The results and discussion of the experiment undertaken are

presented in this section in two parts, (i) results of the corrosion
experiment and (ii) the metallurgical examination using SEM. The Fig. 2. Corrosion rate during immersion tests of material samples with capric acid.
mass loss and corrosion rate of some encasement material-PCM
pairs is negligible in many cases and as a result these materials are
not shown in the graphs or SEM images but the results are pre- the mass loss increases to 0.19 g for a short period before reaching
sented in the tables. steady state for the remainder of the experiment. Over the initial
300 days, aluminium forms a thin oxide layer which minimises the
mass loss. However, following this period there is a sudden loss in
3.1. Laboratory corrosion results
mass, this is caused by pitting corrosion. The pits are caused by local
ruptures in the metal. Following the pitting corrosion, the propa-
The results of the corrosion tests are presented based on the
gation stage is reached. This occurs when the formation of
effect of each individual PCM on the different material samples. As a
aluminium ions at the bottom of the pit which, under certain
result each graph shows the time dependent effects of one PCM
conditions, will form a lm of aluminium chloride or aluminium
type on the material samples.
oxychloride in the pit and stabilise it. After a while the aluminium
The results of the experiment with capric acid are shown in
chloride will hydrolyse into aluminium hydroxide which pre-
Figs. 1 and 2 for mass loss and corrosion rate, respectively.
cipitates at the rim of the pit and covers the opening, which
The mass loss of mild steel and copper can be seen to increase
eventually hinders exchange of ions and slows down the corrosion
relatively constantly throughout the duration of the experiment,
process [43].
with mild steel having the highest mass loss of all the material
The corrosion rate of each material due to being in contact with
samples losing 0.93 g by Day 722 and copper the second highest
capric acid over the course of the experiment is presented in Fig. 2.
with a mass loss of 0.65 g. Brass has one of the higher mass losses in
Stainless steel and Perspex remain unaffected by the PCM with
the rst 200 days reaching 0.08 g, however the rate of mass loss
corrosion rates of of less than 0.01 g/cm2year. The corrosion rate of
decreases due to the presence of zinc as an alloy element and re-
all other material samples vary for the rst 400 days before sta-
sults in mass loss of 0.37 g following 722 days. There is negligible
bilising and maintaining a constant corrosion rate for the
mass loss (less than 0.01 g) recorded from stainless steel or Perspex
remainder of the experiment which were found to be 0.022 g/
due to capric acid.
cm2year, 0.013 g/cm2year, 0.01 g/cm2year and 0.03 g/cm2year for
Aluminium shows minimal mass loss of approximately 0.01 g in
copper, brass, aluminium and mild steel, respectively.
the rst 300 days of the experiment; however following this period

Table 3
Guide for corrosion weight loss used in the industry.

mg/cm2 yr mm/yr Recommendation

>1000 2 Completely destroyed within days

100 to 999 0.1e1.99 Not recommended for service greater than one month
50 to 99 0.1e0.19 Not recommended for service greater than one year
10 to 49 0.02e0.09 Caution recommended, depending on the specic application
0.3 to 9.9 e Recommended for long term service
<0.2 e Recommended for long term service; no corrosion, other than as a result of surface cleaning was evidenced
M.C. Browne et al. / Renewable Energy 108 (2017) 555e568 559

The mass loss of material samples due to the immersion in the

eutectic fatty acid capric:palmitic is presented in Fig. 3. Stainless
steel and Perspex have been unaffected by the fatty acid eutectic
with mass loss values of less than 0.05 g. Mass loss of mild steel is
high, 1.09 g, compared to copper, brass and aluminium materials
which have losses of 0.48 g, 0.27 g and 0.34 g at the end of the
The corrosion rate of the samples immersed in capric:palmitic
acid (Fig. 4) showed the corrosion rate of stainless steel and Perspex
was negligible (less than 0.005 g/cm2year). However, the remaining
sample materials show that the corrosion rate is consistently
increasing for the duration of the experiment up until Day 400
where upon the graphs reach steady state after reaching at 0.02 g/
cm2year, 0.01 g/cm2year, 0.01 g/cm2year and 0.038 g/cm2year for Fig. 4. Corrosion rate during immersion tests of material samples with capric:palmitic
copper, brass, aluminium and mild steel, respectively. Capric acid is
more corrosive than capric:palmitic acid when considering copper
and brass as the corrosion rate is higher for capric acid by
approximately 0.006 g/cm2year and 0.004 g/cm2year, respectively.
However, capric:palmitic acid presented corrosion rates higher by
approximately 0.005 g/cm2year and 0.006 g/cm2year for
aluminium and mild steel than that of capric acid.
The aggressive nature of capric:lauric acid can be seen in Fig. 5
which shows the mass loss of each material sample due to im-
mersion in the fatty acid eutectic. Perspex and stainless steel have
negligible mass loss of less than 0.001 g. Similar to the material
reaction to immersion in previous fatty acids, aluminium is initially
unaffected by the PCM, however following Day 300 the material
showed an initial mass loss of 0.01 g which increases to 0.1 g on Day
400 which remains at approximately that value for the remainder
of the experiment. As can be seen from the graph in Fig. 5 the mass
Fig. 5. Material mass loss due to immersion in capric:lauric acid.
loss of mild steel, copper and brass samples was increasing
constantly over the test period which showed mass losses of 0.83 g/
cm2year, 0.48 g/cm2year and 0.32 g/cm2year by Day 722.
The corrosion rate of capric:lauric acid (Fig. 6) on all of the
material samples is similar to that of the other fatty acid materials
presented above. Stainless steel and Perspex remain unaffected by
the PCM as the mass loss is approximately 0 g.
The corrosion rate for aluminium was initially relatively low for
the rst 300 days at 0.001 g/cm2year before increasing to approx-
imately 0.004 g/cm2year for the remainder of the experiment. Mild
steel has the highest corrosion rate, stabilising at approximately
0.03 g/cm2year followed by copper and brass with corrosion rates
of 0.017 g/cm2year and 0.01 g/cm2year, respectively.
It was noted during the early stages of the experiment, test
tubes containing copper and brass immersed in a fatty acid pre-
Fig. 6. Corrosion rate during immersion tests of material samples with capric: lauric
sented a blue colour of the copper and brass salts (sulphates or
chlorides) in the test tubes compared to the stainless steel sample.
This was seen in all three fatty acid samples (capric, capric: palmitic
acid and capric: lauric acid) as shown in Fig. 7.
SP22 was found to be less aggressive than the fatty acid PCM
(Fig. 8). Aluminium was found to be most affected by the salt hy-
drate having one of the largest mass losses at the end of the
experiment at 1.18 g. The mass loss of mild steel increased after 400
days to approximately 0.08 g. The remaining material samples have
been unaffected by the salt hydrate immersion tests.
The corrosion rate of aluminium (0.041 g/cm2year) was much
higher than any of the other material samples (Fig. 9). Nearly all had
negligible corrosion rates except mild steel which recorded a
corrosion rate of 0.002 g/cm2year.
It was observed that the salt hydrate in the test tube with
aluminium expelled a gas to cause the PCM to create air voids. This
was caused by a chemical reaction between the aluminium and salt
which emits hydrogen [44].
Mild steel with Micronal presented much higher mass loss
Fig. 3. Material mass loss due to immersion in capric:palmitic acid.
560 M.C. Browne et al. / Renewable Energy 108 (2017) 555e568

Fig. 7. Copper, brass and stainless steel samples in (a) capric acid (b) capric:palmitic acid (c) capric:lauric acid.

Fig. 8. Material mass loss due to immersion in SP22. Fig. 11. Corrosion rate during immersion tests of material samples with Micronal.

rate is high however following 150 days it stabilises at approxi-

mately 0.015 g/cm2year. Mild steel is most susceptible to corrosion
by Micronal.
On comparing the corrosion results of the fatty acids, capric acid
appears to be the most aggressive overall. SP22 has been shown to
undergo aggressive corrosion when paired with aluminium. How-
ever, when combined with other sample materials the corrosive
effect was minimal which is in contrast with the ndings of Mor-
eno, Miro , Sole
, Barreneche, Sole
, Martorell and Cabeza [25].
However, it must be noted that SP22 is a mixture of salt hydrate and
organic compounds which would likely reduce the corrosive
Fig. 9. Corrosion rate during immersion tests of material samples with SP22.
3.2. Scanning electron microscope metallographic images

than any other material tested, as shown in Fig. 10. The mass loss of The samples were investigated under a microscope for signs of
mild steel reached a value of approximately 0.4 g compared to the corrosion on Days 1, 25, 200 and 540 using magnications of 200,
corrosion weight loss of all other material samples being less than 3000 and 6000. However, due to the transparent nature of
0.04 g. Perspex this material could not be used in the SEM imaging.
The evolution of corrosion rate of materials immersed in Stainless steel has not been affected by the PCM in any immersion
Micronal is presented in Fig. 11. Initially, the mild steel corrosion tests and as a result there is no presentation of images of stainless
steel. From photographic images, the rate of corrosion can be
visually observed as the size of the pitting in some samples in-
creases with the length of time it was immersed in the PCM.

3.2.1. Capric acid

As shown in Fig. 12e28, capric acid is one of the more aggressive
PCMs tested. Pitting can be seen in the copper (Fig. 12) following 25
days of immersion in capric acid which reected the mass loss of
0.02 g at this stage of the experiments considering the initial mass
was 10.51 g. The entire surface layer on the copper has been
removed by Day 540 along with a mass loss of 0.57 g (Fig. 12c) as
there are no vertical indents as seen in Fig. 12a which was taken on
Day 1.
Similarly, brass had an initial mass of 10.12 g and was corroded
in the same manner as copper. However, this occurred at a slower
Fig. 10. Material mass loss due to immersion in Micronal. rate as on Day 540 the material sample has evidence of severe
M.C. Browne et al. / Renewable Energy 108 (2017) 555e568 561

(a) (b) (c)

Fig. 12. Copper immersed in capric acid on (a) Day 1 6000 magnication (b) Day 25 6000 magnication (c) Day 540 6000 magnication.

pitting (as seen in Fig. 13) although there still some evidence of the surface can be seen to be full of large areas of pitting (Fig. 17c) by
surface layer being intact due to the vertical indents which is ex- Day 540. The corrosion shown in the images reect the mass loss
pected as the mass loss of 0.29 g is less than that of copper and found over the course of the immersion corrosion tests. The initial
capric acid. mass of 10.19 g was reduced by 0.05 g by Day 200 and by 0.19 g on
Aluminium had an initial mass of 3.13 g and can be seen to avoid Day 540.
excessive corrosion up to Day 200 as the horizontal textured lines The aluminium sample immersed in capric:palmitic acid can be
are evident this is also shown in the mass loss of 0.01 g (Fig. 14b). seen in Fig. 18. The sample withstands the corrosive effect of the
The pitting is clearly seen in the images taken on Day 540 Fig. 14c material until approximately 200 days when pitting can be seen in
when the corrosion got progressively worse as the mass loss was image (b) along with some of the intact surface material of the
0.2 g. sample which would reect a mass loss of 0.01 g. By Day 540 the
The mass loss and corrosion rate of mild steel in capric acid surface has been badly corroded as no vertical lines can be seen and
which was 0.83 g and 0.031 g/cm2year on Day 722 showed that the this is shown in mass loss of 0.27 g, an 8.6% loss on the initial mass.
material was adversely affected by the PCM. The images shown in Mild steel can be seen to be corroded from early on in the
Fig. 15 reect this and although there is no deep pitting, the layer on experiment, as on Day 25 small areas of pitting are obvious which
the surface of the material can be seen to change over time as it is was also seen in the mass loss of 0.1 g (Fig. 19) and the surface of the
being worn away by the corrosive PCM. sample can be seen to have vast areas of pitting (Fig. 19c) by Day
540 which showed a mass loss of 0.85 g, 9.1% of the total initial
3.2.2. Capric:palmitic acid
Capric:palmitic has been shown to corrode metals when
considering the mass loss and corrosion rate results of the gravi- 3.2.3. Capric:lauric acid
metric study. Capric:lauric acid showed the same corrosion patterns to that of
Copper had an initial mass of 10.65 g and can be seen to be the previous two fatty acid PCM. Copper and brass samples, as
affected by the corrosive nature of capric:palmitic acid by Day 200 presented in Figs. 20 and 21; appear to have signs of pitting by Day
the mass had reduced by 0.07 g. The vertical lines are barely visible 25 which got increasingly worse as the experiment progressed.
and vast areas of pitting can be seen to be forming as shown in By Day 540 it can be seen the surface of the samples are
Fig. 16b. Day 540 shows the surface of the sample has been completely corroded. The images showed the corrosion process
corroded by the PCM Fig. 16c which was a result from the corrosion whereby there was a mass loss of 0.01 g on Day 25 by both mate-
of the surface and a mass loss of 0.37 g. rials. Copper has shown to be more susceptible to corrosion by
Similar to the copper results, brass immersed in capric:palmitic capric:lauric acid than that of brass as the mass loss was shown to
acid shows areas of pitting after Day 200 (Fig. 17b) and the entire be 0.43 g by Day 540 for copper and 0.26 g for brass which is

(a) (b) (c)

Fig. 13. Brass immersed in capric acid on (a) Day 1 6000 magnication (b) Day 200 6000 magnication (c) Day 540 6000 magnication.
562 M.C. Browne et al. / Renewable Energy 108 (2017) 555e568

(a) (b) (c)

Fig. 14. Aluminium immersed in capric acid on (a) Day 25 6000 magnication (b) Day 200 6000 magnication (c) Day 540 3000 magnication.

(a) (b) (c)

Fig. 15. Mild steel immersed in capric acid on (a) Day 1 6000 magnication (b) Day 25 6000 magnication (c) Day 540 6000 magnication.

(a) (b) (c)

Fig. 16. Copper immersed in capric:palmitic acid on (a) Day 1 6000 magnication (b) Day 200 3000 magnication (c) Day 540 3000 magnication.

(a) (b) (c)

Fig. 17. Brass immersed in capric:palmitic acid on (a) Day 1 6000 magnication (b) Day 200 3000 magnication (c) Day 540 200 magnication.
M.C. Browne et al. / Renewable Energy 108 (2017) 555e568 563

(a) (b) (c)

Fig. 18. Aluminium immersed in capric:palmitic acid on (a) Day 1 3000 magnication (b) Day 200 3000 magnication (c) Day 540 3000 magnication.

(a) (b) (c)

Fig. 19. Mild steel immersed in capric:palmitic acid on (a) Day 1 3000 magnication (b) Day 25 3000 magnication (c) Day 540 3000 magnication.

(a) (b) (c)

Fig. 20. Copper immersed in capric:lauric acid on (a) Day 1 3000 magnication (b) Day 25 3000 magnication (c) Day 540 3000 magnication.

reected in the images taken using SEM. Fig. 22 shows aluminium of the material almost entirely corroded where the mass loss was
immersed in capric:lauric acid, the images show the process of found to be 0.71 g.
corrosion had started by Day 200 as one large area of pitting can be
seen in Fig. 22b. By Day 540 the entire surface of the metal has been
completely broken down which is seen in Fig. 22c and reected in 3.2.4. Salt hydrate SP22
the mass loss results as 3.2% of the initial weight had been corroded SP22 has been shown to result in severe corrosion on aluminium
by this stage of the experiment. in the gravimetric analysis and minor corrosion on the remainder of
The microscopic evaluation of mild steel immersed in capri- the samples. It is interesting to note that corrosion may not be
c:lauric acid is presented in Fig. 23. The grain boundaries of mild evident in the gravimetric analysis but the initial stages of corrosion
steel can be seen in (b) which indicates the weak points of the can be seen using SEM imaging. An example of this is copper
material that are most susceptible to corrosion by the fatty acid immersed in SP22 as pitting can be seen in Fig. 24c and had a mass
eutectic and have reported a mass loss of 0.18 g by Day 200. The loss of 0.02 g by Day 540. This is due to the large magnication of
material is seen to be further corroded by Day 540 with the surface the SEM images which showed the initial stages of corrosion which
were not large enough to be determined by the weighing scale.
564 M.C. Browne et al. / Renewable Energy 108 (2017) 555e568

(a) (b) (c)

Fig. 21. Brass immersed in capric:lauric acid on (a) Day 1 6000 magnication (b) Day 25 6000 magnication (c) Day 540 3000 magnication.

(a) (b) (c)

Fig. 22. Aluminium immersed in capric:lauric acid on (a) Day 1 6000 magnication (b) Day 200 6000 magnication (c) Day 540 3000 magnication.

(a) (b) (c)

Fig. 23. Mild steel immersed in capric:lauric acid on (a) Day 1 6000 magnication (b) Day 200 6000 magnication (c) Day 540 3000 magnication.

Images of aluminium which has been immersed in SP22 are corrosion can be seen in Figs. 26 and 27 for these materials. Small
presented in Fig. 25. The aggressive corrosion is evident as pitting pockets of pitting in the images taken on Day 540 suggest given
can be seen in (b) after only 25 days. time the material will further corrode which is also shown by the
The grain boundaries can be seen to be depleted in (c) which has progressive mass loss of 0.01 g on Day 200 and 0.04 g on Day 540
been taken on Day 540 when mass loss of 0.01 g was reported. The experienced by both materials.
delay in mass can be explained by the initial low mass of the ma- Aluminium was extensively corroded from exposure to Micro-
terial sample, 3.14 g. Such extensive corrosion can compromise the nal. Numerous small areas of pitting can be seen in Fig. 28b which
mechanical properties of materials. increase and grow in size by Day 540 this reects the reported mass
loss of 0.01 g (Fig. 28c). It is worth noting this image is taken at a
3.2.5. Micronal magnication of 200 to emphasis the aggressive nature of
Although, copper and brass immersed in Micronal did not Micronal on aluminium.
show signs of corrosion in the gravimetric analysis, early stage
M.C. Browne et al. / Renewable Energy 108 (2017) 555e568 565

(a) (b) (c)

Fig. 24. Copper immersed in SP22 on (a) Day 1 6000 magnication (b) Day 200 6000 magnication (c) Day 540 6000 magnication.

(a) (b) (c)

Fig. 25. Aluminium immersed in SP22 on (a) Day 1 3000 magnication (b) Day 25 3000 magnication (c) Day 540 200 magnication.

(a) (b) (c)

Fig. 26. Copper immersed in Micronal on (a) Day 1 3000 magnication (b) Day 200 3000 magnication (c) Day 540 6000 magnication.

3.3. Evaluation and recommendation up to 22.15 mg/cm2year compared to the corrosion rate of 1.67 mg/
cm2year of copper in Micronal. Aluminium is affected by fatty
The nal results of the mass loss and corrosion rate of each acids which have shown corrosion rates of up to 12.04 mg/cm2year
material immersed in PCM after 722 days are summarised in although this is less than the corrosion rate of copper, brass and
Table 4. Overall, it can be seen that mild steel is an unlikely mild steel which have shown corrosion rates of up to 37.58 mg/
candidate for encapsulation of PCM particularly fatty acids and cm2year in fatty acids. The material loss of stainless steel and
Micronal which was shown to have the highest mass loss of Perspex is negligible and it can be assumed these materials are
0.88 ge1.09 g and 0.4 g for fatty acids and Micronal, respectively. compatible for use with PCM.
Aluminium showed the highest mass loss of all materials immersed The corrosion rate was measured in g/cm2year which has been
in SP22 of 1.17 g; the salt hydrate was shown to aggressively attack changed to mg/cm2year to be used in combination with the
the passivation layer it produces for self-protection causing further corrosion weight loss standard guide. The document is used in in-
corrosion of the material. Brass and copper are seen to be corroded dustry to dene the degree of corrosion considered excessive and
by the fatty acids more than the other PCM with corrosion rates of gives an indication of the period of recommended use. The results
566 M.C. Browne et al. / Renewable Energy 108 (2017) 555e568

(a) (b) (c)

Fig. 27. Brass immersed in Micronal on (a) Day 25 3000 magnication (b) Day 200 3000 magnication (c) Day 540 3000 magnication.

(a) (b) (c)

Fig. 28. Aluminium immersed in Micronal acid on (a) Day 1 3000 magnication (b) Day 200 3000 magnication (c) Day 540 3000 magnication.

Table 4
Corrosion results obtained by gravimetric analysis after 722 days.

Metal Capric acid Capric: palmitic acid Capric: lauric acid SP22 Micronal
Mass loss Corrosion rate (mg/ Mass loss Corrosion rate (mg/ Mass loss Corrosion rate (mg/ Mass loss Corrosion rate (mg/ Mass loss Corrosion rate (mg/
(g) cm2yr) (g) cm2yr) (g) cm2yr) (g) cm2yr) (g) cm2yr)

Copper 0.65 22.15 0.48 16.39 0.5 17.4 0.02 0.68 0.05 1.67
Brass 0.39 13.57 0.27 9.31 0.32 10.9 0.02 0.7 0.03 1.02
Aluminium 0.2 7.08 0.34 12.04 0.12 4.25 1.17 41.45 0.01 0.33
Stainless 0.01 0.37 0 0 0 0 0 0.0 0 0
Mild steel 0.93 31.07 1.09 37.58 0.88 30.34 0.08 2.75 0.4 14.72
Perspex 0 0 0 0 0 0 0.02 0.63 0 0

of the suitability of tested encasement material-PCM pairs are overall economic feasibility of the system when considering the
presented in Table 5. The metallographic investigation is also taken suitability of a PCM - encasement material pair as it as suggests the
into account when making the decision. container will not have to be replaced during the lifetime of the
Samples have been categorised into recommended, not rec- system. Ideally, a material with a recommended relationship with a
ommended and caution. Recommended suggests the material-PCM PCM would be the primary choice for integration into a PV/T/PCM
pair is compatible for long term service as the corrosion rate was system. However, other factors such as thermal conductivity and
less than 0.3 mg/cm2yr. A Recommended result improves the implications of manufacture should be also considered.

Table 5
Suitability of material e PCM pairs studied for use in a TES system.

Capric acid Capric: palmitic acid Capric: lauric acid SP22 Micronal

Copper Caution Caution Caution Recommended Recommended

Brass Caution Caution Caution Recommended Recommended
Aluminium Caution Caution Caution Not recommended Recommended
Stainless steel Recommended Recommended Recommended Recommended Recommended
Mild steel Caution Caution Caution Recommended Caution
Perspex Recommended Recommended Recommended Recommended Recommended
M.C. Browne et al. / Renewable Energy 108 (2017) 555e568 567

Caution is advised (as shown in Table 5) when using material- 4. Conclusion

PCM pairs with a recorded corrosion weight loss of 10 mg/cm2yr
e 49 mg/cm2yr. Extra care must be taken to assess the specic Initial investigation into the use of sealants as a method of
application and the potential for corrosion under these conditions. fabricating a container in a PV/T/PCM system found all sealants
It would be recommended to carry out an initial investigation such used have been shown to react with PCM, making them unsuitable
as building a prototype to determine the full extent of potential adhesives for PCM containment. As a result of the effect of the PCM
corrosion. It would not be recommended to use an encasement on adhesives, there has been no investigation of the corrosive effect
materialePCM pair which advises caution as the container may not of the particular fatty acids studied. As a result investigation into
last the lifetime of the TES system. the corrosive properties of the PCM was carried out by immersing
Container materialePCM pairs with corrosion rates great than them in six commonly used materials to assess the materials which
50 are not recommended for service. However if applications would be compatible for PCM containment in TES systems. The
necessitate the use of certain materials then depending on the level experiment involved immersing the material samples in liquid PCM
of corrosion, encasement materialePCM pairs can be used for pe- for up to 722 days.
riods of less than one year (corrosion rate 50 mg/cm2yr e 99 mg/ The ndings of this study presented compatible PCM-
cm2yr) or less than one month (corrosion rate 100 mg/cm2yr e 999 encasement material pairs which were recommended based off
mg/cm2yr). This is application specic as it may not be economi- the experimental results in conjunction with the industry guide for
cally feasible to replace parts of a system following a month or year corrosion weight loss. Stainless steel has been shown to be the most
in service. resistant to corrosion of all the metals investigated as such it is
Stainless steel and Perspex are the most resistant to corrosion or recommended for use with all the PCM. Similarly, Perspex showed
deformation by the fatty acid PCM as they have shown to be the minimal corrosion when immersed in the PCM deeming it a suit-
only two materials of the ve materials investigated to be recom- able option for an encasement material. However, the function of
mended for use with all PCM. the material would have to be considered in the system design, as
This was the rst investigation into the behaviour of fatty acids Perspex has a low thermal conductivity. Aluminium is not recom-
and metals and it has been found to show caution when using mended for use with SP22 due to high corrosion rates. Although
copper, brass, aluminium and mild steel with capric acid and fatty aluminium has showed signs of corrosion when immersed in fatty
acid eutectics, capric:palmitic and capric:lauric. The corrosion rate acids, it was not sufcient enough to impair functionality and as
for these metal-PCM pairs remained within 4.5 mg/cm2/year and such can be used if caution is exerted. Both copper and brass were
37 mg/cm2/year. recommended for integration with SP22 and Micronal, however
Aluminium is not recommended for use with SP22 as has been caution is advised when encapsulating fatty acids.
shown with other salt hydrate PCM which has been reported in Throughout the corrosion tests images were taken using a SEM
previous studies [20]. Despite this, it has been recommended for to assess the level of corrosion on the surface of the material. The
use with all other material samples. images showed the initial stages of surface corrosion which were
The corrosive behaviour of Micornal has not been investigated too small to effect the change in weight of the material samples.
previous to this study and it has shown that it is compatible with all They also allowed the initial stages of corrosion to be monitored
materials tested except mild steel where caution is recommended. and assessed. In some cases, such as the fatty acids and copper pairs
One of the main corrosion prevention measures is to remove and the fatty acids and aluminium pairs, areas of pitting were
contact between the PCM and the metal using macro- captured in the SEM images. As the area of pitting increased over
encapsulation methods. The most cost effective methods in use the course of the experiment it followed that the rate of corrosion
are polyethylene and polypropylene plastic bottles [45]. However also increased, emphasising the vulnerability of the material once
the plastic packaging hinders heat transfer to and from the PCM. pitting was evident. The images assisted in pinpointing the initial
Thin plated metal and mild steel cans have also been used for this signs of corrosion and helped in drawing the presented conclusions
application although ingress of moisture can drastically impact on the most appropriate PCM-encasement material pairs for use in
their performance [46]. TES systems.
One example of plating surfaces to minimise corrosion is
copper heat pipe exchangers encased in aluminium, onto which Acknowledgements
aluminium ns are attached thereby plating the copper surface
with aluminium. However, if minor cracks form in such casings The authors would like to acknowledge the Centre for Micro-
the copper will become cathodic to the aluminium and the PCM copy and Analysis (CMA) Trinity College Dublin, EU Structural
will act as an electrolyte resulting in galvanic corrosion [21]. Funds, ERDF, the Department of Jobs, Enterprise and Innovation and
Pendyala [14] investigated the potential to encapsulate PCM with Higher Education Authority through GREP Funding, Science Foun-
a metal oxide to produce a coated PCM pellet using electro- dation Ireland through their Research Frontiers Program Code 07/
chemical coating. It is recommended metals be fused together RFP/ENEF719, the Research Support Unit at Dublin Institute of
using metal joining techniques such as welding and materials Technology, European Unions Horizon 2020 Research and Inno-
which cannot be welded, such as that of Perspex be cast to avoid vation Programme under grant agreement No. 657466 (INPATH-
possible leaks. TES), COST Action TU0802 Next Generation Cost Effective Phase
Aluminium, copper and brass have shown higher corrosion rates Change Materials for Increased Energy Efciency in Renewable
than that of Perspex and stainless steel in fatty acid PCM. The Energy Systems in Buildings (NeCoEPCM) and COST Action TU1205
corrosion rate varies from 425 mg/cm2year to 22 mg/cm2year Building Integration of Solar Thermal Systems (BISTS).
depending on the encasement materialePCM pair. Aluminium can
experience galvanic corrosion whereby tin plating can reduce the References
effect and an oxide inhibiting compound can also be used on the
surface to avoid the ingress of moisture on the contact surface. Tin [1] M.C. Browne, B. Norton, S.J. McCormack, Heat retention of a photovoltaic/
and nickel plating is also an option to consider reducing the effects thermal collector with PCM, Solar Energy 133 (2016) 533e548.
[2] L. Navarro, A. de Gracia, D. Niall, A. Castell, M. Browne, S.J. McCormack,
of corrosion on copper. P. Grifths, L.F. Cabeza, Thermal energy storage in building integrated thermal
systems: a review. Part 2. Integration as passive system, Renew. Energy 85
568 M.C. Browne et al. / Renewable Energy 108 (2017) 555e568

(2016) 1334e1356. stability and corrosion of metals with respect to thermal cycling, Renew.
[3] L. Navarro, A. de Gracia, S. Colclough, M. Browne, S.J. McCormack, P. Grifths, Energy 28 (2003) 939e948.
L.F. Cabeza, Thermal energy storage in building integrated thermal systems: a [24] A. Lazaro, B. Zalba, M. Bobi, C. Castello n, L.F. Cabeza, Experimental study on
review. Part 1. active storage systems, Renew. Energy 88 (2016) 526e547. phase change materials and plastics compatibility, AIChE J. 52 (2006)
[4] M.J. Huang, P.C. Eames, N.J. Hewitt, The application of a validated numerical 804e808.
model to predict the energy conservation potential of using phase change [25] P. Moreno, L. Miro , A. Sole
, C. Barreneche, C. Sole , I. Martorell, L.F. Cabeza,
materials in the fabric of a building, Sol. Energy Mater. Sol. Cells 90 (2006) Corrosion of metal and metal alloy containers in contact with phase change
1951e1960. materials (PCM) for potential heating and cooling applications, Appl. Energy
[5] A. Hasan, S. McCormack, M. Huang, B. Norton, Energy and cost saving of a 125 (2014) 238e245.
photovoltaic-phase change materials (PV-PCM) system through temperature [26] G. Ferrer, A. Sole , C. Barreneche, I. Martorell, L.F. Cabeza, Corrosion of metal
regulation and performance enhancement of photovoltaics, Energies 7 (2014) containers for use in PCM energy storage, Renew. Energy 76 (2015) 465e469.
1318e1331. [27] A. Sole, L. Miro
, C. Barreneche, I. Martorell, L.F. Cabeza, Corrosion of metals and
[6] A. Hasan, S.J. McCormack, M.J. Huang, B. Norton, Evaluation of phase change salt hydrates used for thermochemical energy storage, Renew. Energy 75
materials for thermal regulation enhancement of building integrated photo- (2015) 519e523.
voltaics, Sol. Energy 84 (2010) 1601e1612. [28] E. Oro , C. Barreneche, M.M. Farid, L.F. Cabeza, Experimental study on the se-
[7] M.C. Browne, B. Norton, S.J. Mc Cormack, Phase change materials for photo- lection of phase change materials for low temperature applications, Renew.
voltaic thermal management, Renew. Sustain. Energy Rev. 47 (2015) Energy 57 (2013) 130e136.
762e782. [29] J.R. Turnpenny, D.W. Etheridge, D.A. Reay, Novel ventilation cooling system
[8] B. Zalba, J.M. Marn, L.F. Cabeza, H. Mehling, Review on thermal energy storage for reducing air conditioning in buildings.: Part I: testing and theoretical
with phase change: materials, heat transfer analysis and applications, Appl. modelling, Appl. Therm. Eng. 20 (2000) 1019e1037.
Therm. Eng. 23 (2003) 251e283. [30] U. Stritih, V. Butala, Experimental investigation of energy saving in buildings

[9] A. Sari, H. Sari, A. Onal, Thermal properties and thermal reliability of eutectic with PCM cold storage, Int. J. Refrig. 33 (2010) 1676e1683.
mixtures of some fatty acids as latent heat storage materials, Energy Convers. [31] M.C. Browne, B. Norton, S.J. McCormack, Heat retention of a photovoltaic/
Manag. 45 (2004) 365e376. thermal collector with PCM, Sol. Energy 133 (2016) 533e548.
[10] A. Sharma, V.V. Tyagi, C.R. Chen, D. Buddhi, Review on thermal energy storage [32] D. Rozanna, T. Chuah, A. Salmiah, T.S. Choong, M. Saari, Fatty acids as phase
with phase change materials and applications, Renew. Sustain. Energy Rev. 13 change materials (PCMs) for thermal energy storage: a review, Int. J. Green
(2009) 318e345. Energy 1 (2005) 495e513.
[11] Y. Yuan, N. Zhang, W. Tao, X. Cao, Y. He, Fatty acids as phase change materials: [33] D. Feldman, M.M. Shapiro, D. Banu, C.J. Fuks, Fatty acids and their mixtures as
a review, Renew. Sustain. Energy Rev. 29 (2014) 482e498. phase-change materials for thermal energy storage, Sol. Energy Mater. 18
[12] T. Kousksou, P. Bruel, A. Jamil, T. El Rhaki, Y. Zeraouli, Energy storage: ap- (1989) 201e216.
plications and challenges, Sol. Energy Mater. Sol. Cells 120 (Part A) (2014) [34] A. Hasan, A.A. Sayigh, Some fatty acids as phase-change thermal energy
59e80. storage materials, Renew. Energy 4 (1994) 69e76.
[13] T.E. Alam, J.S. Dhau, D.Y. Goswami, E. Stefanakos, Macroencapsulation and [35] H. He, Q. Yue, B. Gao, X. Zhang, Q. Li, Y. Wang, The effects of compounding
characterization of phase change materials for latent heat thermal energy conditions on the properties of fatty acids eutectic mixtures as phase change
storage systems, Appl. Energy 154 (2015) 92e101. materials, Energy Convers. Manag. 69 (2013) 116e121.
[14] S. Pendyala, Macroencapsulation of Phase Change Materials for Thermal En- [36] Y. Yuan, N. Zhang, W. Tao, X. Cao, Y. He, Fatty acids as phase change materials:
ergy Storage, 2012. a review, Renew. Sustain. Energy Rev. 29 (2014) 482e498.
[15] A. Jamekhorshid, S.M. Sadrameli, M. Farid, A review of microencapsulation [37] Rubiterm, 2015.
methods of phase change materials (PCMs) as a thermal energy storage (TES) [38] Y. Wu, T. Wang, Hydrated salts/expanded graphite composite with high
medium, Renew. Sustain. Energy Rev. 31 (2014) 531e542. thermal conductivity as a shape-stabilized phase change material for thermal
[16] F.C. Porisini, Salt hydrates used for latent heat storage: corrosion of metals and energy storage, Energy Convers. Manag. 101 (2015) 164e171.
reliability of thermal performance, Sol. Energy 41 (1988) 193e197. [39] L.F. Cabeza, C. Castello n, M. Nogue s, M. Medrano, R. Leppers, O. Zubillaga, Use
[17] M. Groll, O. Brost, D. Heine, Corrosion of steels in contact with salt eutectics as of microencapsulated PCM in concrete walls for energy savings, Energy Build.
latent heat storage materials: inuence of water and other impurities, Heat 39 (2007) 113e119.
Recovery Syst. CHP 10 (1990) 567e572. [40] BASF, Micronal PCM - Intelligent Temeperature Management for Buildings,
[18] L.F. Cabeza, J. Illa, J. Roca, F. Badia, H. Mehling, S. Hiebler, F. Ziegler, Immersion 2012.
corrosion tests on metal-salt hydrate pairs used for latent heat storage in the [41] A. Sharma, L.D. Won, D. Buddhi, J.U. Park, Numerical heat transfer studies of
32 to 36 C temperature range, Mater. Corros. 52 (2001) 140e146. the fatty acids for different heat exchanger materials on the performance of a
[19] L.F. Cabeza, J. Illa, J. Roca, F. Badia, H. Mehling, S. Hiebler, F. Ziegler, Middle latent heat storage system, Renew. Energy 30 (2005) 2179e2187.
term immersion corrosion tests on metal-salt hydrate pairs used for latent [42] ASTM, Standard Practice for Preparing, Cleaning and Evaluating Corrosion Test
heat storage in the 32 to 36 C temperature range, Mater. Corros. 52 (2001) Specimen, in: Annual Book of ASTM Standards, Section 3, Metal Test Methods
748e754. and Analytical Procedures, American Society for Testing and Materials
[20] L.F. Cabeza, J. Roca, M. Noguee s, H. Mehling, S. Hiebler, Long term immersion (ASTM), West Conshohocken, PA, US, 1994.
corrosion tests on metal-PCM pairs used for latent heat storage in the 24 to [43] S. Gustafsson, Corrosion Properties of Aluminium Alloys and Surface Treated

29 C temperature range, Mater. Corros. 56 (2005) 33e39. Alloys in Tap Water, 2011.
[21] A.J. Farrell, B. Norton, D.M. Kennedy, Corrosive effects of salt hydrate phase [44] C. Porcincula, N. Marcilio, I. Tessaro, M. Gerchmann, Production of hydrogen
change materials used with aluminium and copper, J. Mater. Process. Technol. in the reaction between aluminum and water in the presence of NaOH and
175 (2006) 198e205. KOH, Braz. J. Chem. Eng. 29 (2012) 337e348.
[22] A. Garca-Romero, A. Delgado, A. Urresti, K. Martn, J.M. Sala, Corrosion [45] PCM Products, 2015.
behaviour of several aluminium alloys in contact with a thermal storage phase [46] A.F. Regin, S.C. Solanki, J.S. Saini, Heat transfer characteristics of thermal en-
change material based on Glaubers salt, Corros. Sci. 51 (2009) 1263e1272. ergy storage system using PCM capsules: a review, Renew. Sustain. Energy
[23] A. Sar, K. Kaygusuz, Some fatty acids used for latent heat storage: thermal Rev. 12 (2008) 2438e2458.