Applied Energy 113 (2014) 1512–1518

Contents lists available at ScienceDirect

Applied Energy

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

Supercooling suppression of microencapsulated phase change materialsby optimizing shell composition and structure

0306-2619/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.apenergy.2013.08.048

⇑ Corresponding author. Tel.: +1 301 405 6007.E-mail address: baoyang@umd.edu (B. Yang).

Fangyu Cao, Bao Yang ⇑Department of Mechanical Engineering, University of Maryland, College Park, MD, United States

h i g h l i g h t s

� A new method for supercooling suppression of microPCMs by optimizing the structure of the microcapsule shell.� Large effective latent heat (up to 213 J/g) of the microPCMs, much higher than those using additive as nucleating agents.� Change of shell composition and structure significantly affects the phase transition processes of the encapsulated PCMs.� The latent heat of the shell-induced phase transition is maximized, reaching 83.7% of the latent heat of bulk octadecane.� Hollow spheres with porous rather than solid resin shell are also formed when the SDS concentration is very high.

a r t i c l e i n f o

Article history:Received 22 April 2013Received in revised form 17 July 2013Accepted 19 August 2013Available online 2 October 2013

Keywords:MicrocapsulesPhase change materialSupercooling suppressionMelamine formaldehyde resinHeat capacity

a b s t r a c t

A new method for supercooling suppression of microencapsulated phase change materials (PCMs) hasbeen developed by optimizing the composition and structure of the microcapsule resin shell. The micro-capsules comprising paraffin octadecane encapsulated in melamine–formaldehyde resin shell were syn-thesized with the use the oil-in-water emulsion technique. These PCM microcapsules are 5–15 lm indiameter. The supercooling of these octadecane microcapsules can be as large as 13.6 �C, when the homo-geneous nucleation is dominant during the melt crystallization into the thermodynamically stable tri-clinic phase. It is discovered that the homogeneous nucleation can be mediated by shell-inducednucleation of the triclinic phase and the metastable rotator phase when the shell composition and struc-ture are optimized, without need of any nucleating additives. The effects of synthesis parameters, such asratio of melamine to formaldehyde, pH of pre-polymer, and pH of emulsion, on the phase transition prop-erties of the octadecane microcapsules have been investigated systemically. The optimum synthesis con-ditions have been identified in terms of minimizing the supercooling while maintaining heat capacity.Potential applications of this type of phase changeable microcapsules include high heat capacity thermalfluids, thermal management in smart buildings, and smart textiles.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Phase change materials (PCMs) have been promising in wideapplications in thermal energy storage and transfer because the la-tent heat storage method provides much higher energy densitycomparing with the sensible heat method [1–12]. Microencapsu-lated PCMs, which refer to PCMs encapsulated in micron-sizedshell, have been studied extensively as a route to avoid the possibleinteraction between the PCM and matrix and the PCM leakage. Par-affins, the n-alkanes (CnH2nþ2) with different numbers of carbonatoms in their chain, are common PCMs used as the core materialof microcapsules because of their appropriate phase transitiontemperature, large latent heat of fusion, chemical stability, andcapability of being microencapsulated.[3–5,12–14].

However, the supercooling remains a major obstacle to theindustrial application of microcapsule PCMs. Supercooling DTs ofa PCM, i.e., the melting–freezing hysteresis, can be evaluated by

DTs ¼ Tm � Tf ; ð1Þ

where Tm and Tf are the melting and freezing temperature, respec-tively. It has been reported that the supercooling of the PCM micro-capsules increase up to 13 �C when the microcapsule size decreasesfrom 100 lm to 5 lm [15]. In purpose of utilizing the latent heat ofPCMs, the operation temperature should at least cover the rangefrom Tm to Tf so as to complete the phase transition, i.e., the hottemperature limit Th > Tm, and the cold temperature limit Tc < Tf.Giving an operation temperature range DT ¼ Th � Tc , the apparentspecific heat of PCMs CPCM can be estimated by

CPCM ¼ Dhf þ Cp;sðTf � TcÞ þ Cp;lðTh � TmÞ� �

=DT ð2Þ

F. Cao, B. Yang / Applied Energy 113 (2014) 1512–1518 1513

in which Dhf, Cp,s, and Cp,l are the latent heat of fusion, specific heatof solid phase and liquid phase of the corresponding PCM. Becauseof the large latent heat of PCMs, in a reasonably small operationtemperature range, Dhf is usually the major contribution of theapparent heat capacity. Based on Eq. (2), the larger DT is, the smal-ler the elevation of CPCM owing to latent heat is. As DT > DTs is re-quired, an increase of DT can dramatically affect the ability ofthermal management of PCMs.

Many researchers have been reported to reduce the supercool-ing of microencapsulated PCMs in the past. The most commonmethod is to add nucleation agents, such as high-melting pointparaffin or alcohol and solid nanoparticle, to promote heteroge-neous nucleation during melt crystallization. For example, Leefound the derivatives of n-paraffin, such as 1-octadecylamine, 1-octadecanol, were suitable for reducing supercooling of octadecanemicrocapsules [16]. Fan et al. reported that paraffin with meltingpoint 60–65 �C (20 wt.% of the core materials) was able to elimi-nate the supercooling of n-octadecane in microcapsules [17,18].Alvarado et al. selected to use 6 wt.% tetradecane and 6 wt.% tetra-decanol for supercooling suppression of microencapsulatedtetradecane [19]. Those nucleating additives are miscible withthe PCM alkanes in the core at elevated temperature, but they willprecipitate from the solution when temperature drops below thefreezing point of the additives and thus promote heterogeneousnucleation. One drawback of this method is that the effective latentheat of the PCM microcapsules is reduced due to the relativelylarge amount of additive.

In this paper, a new method for supercooling suppression ofoctadecane microcapsule has been developed by optimizing the

N N

N NH2

NH2

NH2

+O

H H

+ OH H

N N

N NH2

NH2

NH2

+HH

OH

OH

OH-

N N

NHN

HN

NH2

+

OH

OH H

OHð3Þ

Table 1List of octadecane microcapsules prepared with different conditions.

SampleNo.

F:Mratio

Initial pH of pre-polymer

Initial concentration of HAc inemulsion (ml/L)

MC01 2.00 8.30 1.50MC02 1.50 8.30 1.50MC03 1.30 8.30 1.50MC04 1.25 8.30 1.50MC05 1.25 8.00 1.50MC06 1.25 8.25 1.50MC07 1.25 8.50 1.50MC08 1.25 8.75 1.50MC09 1.25 9.00 1.50MC10 1.25 8.50 0.75MC11 1.25 8.50 1.00MC12 1.25 8.50 1.25MC13 1.25 8.50 1.38MC14 1.25 8.50 1.50MC15 1.25 8.50 1.75

composition and structure of the melamine–formaldehyde resinshell. The effects of synthesis parameters, such as ratio of mela-mine to formaldehyde, pH of pre-polymer, and pH of emulsion,on the phase transition properties of the octadecane microcapsuleshave been investigated systemically, and the optimum synthesisconditions have been indentified in terms of minimizing the super-cooling while maintaining the heat capacity of the composite.

2. Material and methods

2.1. Materials

n-octadecane, C18H38, (99%), melamine (99%), sodium dodecylsulfate (SDS, 99%), and acetic acid (HAc) (99%) were purchasedfrom Sigma–Aldrich. Sodium carbonate (anhydrous) was pur-chased from J.T. Baker. Formaldehyde (37 wt.% aqueous solution)was purchased from Fluka. Synfluid polyalphaolefin (PAO) was

supplied by Chevron Philips Chemical Company LLC. Distilledwater was supplied by the University of Maryland, College Park.

2.2. Synthesis of PCM microcapsules

The octadecane microcapsules were synthesized with the emul-sion polymerization method [1,17,18,20–28]. First, the pre-poly-mer solution was prepared, for example, by adding 1 g ofmelamine and 1.25 g of 37% formaldehyde into 5 ml of distilledwater. The pH of the mixture was adjusted to 8.5 with diluted so-dium carbonate solution. The mixture was kept at 70 �C for 30 mintill it turned to clear. To prepare the octadecane emulsion, 1 g octa-decane was added in 20 ml distilled water with 0.2 g SDS as surfac-tant. The pH of this aqueous phase was adjusted to 4 with HAc. Themixture was emulsified mechanically using a magnetic stirrerat60 �C. Then, 2 g pre-polymer solution was added to the emulsionsystem to initiate the in situ polymerization. The reaction contin-ued at 60 �C for 3 h, so that the polymerization could be completed.Required reaction time may vary for different precursor ratios. Theresultant microcapsules were filtered and washed with distilledwater and acetone, and then dried in an oven at 60 �C. Table 1 listsoctadecane microcapsules prepared with different conditions, suchas ration of formaldehyde to melamine (F:M ratio), pH of pre-poly-mer solution, and concentration of HAc in emulsion.

2.3. Microcapsule resin shell

Two steps of synthesis are involved in the formation of themicrocapsule resin shell. The first step is melamine–formaldehydepre-condensation in a basic environment,

1514 F. Cao, B. Yang / Applied Energy 113 (2014) 1512–1518

and the second is the microencapsulation by in situ polymerizationof the pre-polymer in an acidic environment, in which the mela-mine–formaldehyde resin (MFR) shell is formed on the surface ofthe octadecane droplets,

H+

N N

NHN

HN

NH2

+

OH

OH

N N

NHN

HN

NH2

OH

OH

N N

NHN

NH

NH2HN

N N

NHN

NH

OHN

N N

NHN

NHHNN

N N

NH2

HN

ð4Þ

The microstructure of the MFR shell can be characterized by thedegree of crosslink and bridge types between two triazine rings,which can be either methylene ether ones or methylene ones[29,30]. The parameters of the encapsulation process, includingsurfactant concentration, F:M ratio, the pH values of pre-polymersolution, and HAc concentration in emulsion, affect significantlythe microstructure of the microcapsules and thus the phase transi-tion behavior of the microencapsulated octadecane PCM.

2.4. Characterization of PCM microcapsules

The morphology and size of the octadecane microcapsules weremeasured using a Scanning Electronic Microscopy (SEM, HitachiSU-70) and a Transmission Electronic Microscopy (TEM, JEM-2010). The phase transition behavior of these microcapsules wasexamined using a differential scanning calorimetry (DSC, TA-Q100). The heating and cooling rate was set to be 5 �C/min undera nitrogen atmosphere. The fraction of phase change octadecanecan be estimated according to the following equation,

Octadecane% ¼ DHmicrocapsule

DHoctadecane� 100 ð5Þ

where DHoctadecane and DHmicrocapsule are the latent heat of fusion ofthe pure octadecane and the microcapsules, respectively.

3. Results and discussion

3.1. Effects of surfactant SDS concentration on microcapsule shell

SEM images of octadecane microcapsules are shown in Fig. 1a–c. Itcan be seen in Fig. 1a and b that the concentration of surfactant SDSis critical for the formation of microcapsules. Hollow spheres withporous shells, rather than solid resin shell, are formed when theSDS concentration is relatively high, e.g., 50 g/L in emulsion, asshown in Fig. 1a. The raspberry-like porous shells are made of nano-sized spheres grown in the aqueous phase. When SDS concentration

drops to 10 g/L, octadecane microcapsules with solid resin shell areformed, as shown in Fig. 1b. These microcapsules have a diameterranging from 5 lm to 15 lm, and their resin shell is about 110 nmin thickness. Fig. 1c shows the submicron resin spheres without

octadecane encapsulated, which indicates polymerization reactionoccurs not only at the water–oil interface, but also inside the aque-ous phase. Thus, excessive amount of MFR precursors are requiredin encapsulating octadecane oil in order to compensate the con-sumption of the polymerization in aqueous phase.

3.2. Effects of F:M ratio on phase transition properties of microcapsules

Fig. 2 shows the DSC curves of octadecane microcapsules syn-thesized with various F:M ratio in the precursor. The DSC heatingcurves are not shown since the melting temperatures of the micro-encapsulated octadecane are little different in shape from those ofthe bulk.

Three peaks are observed on the DSC cooling curves of thesemicrocapsules, labeled a, b, and c, from high to low temperatures.Peak c can be attributed to the transition from liquid octadecaneto the thermodynamically-stable triclinic crystal phase based onhomogeneous nucleation. The homogeneous nucleation occurs ata supercooling as large as 13.6 �C (see Supplemental materials)due to the lack of nucleation sites in the microcapsules. The appear-ance of peak a and b implies that the microencapsulated octadecanefollows a two-step phase transition mechanism, liquid-rotatorphase transition (peak a) and rotator-triclinic phase transition (peakb) [31]. The peak a also include the contribution from the direct li-quid-triclinic phase transition induced by the heterogeneous nucle-ation at the shell. The rotator phases of alkanes are often observed inconfined geometry such as microcapsules, which are weakly-or-dered crystalline phases that lack long-range order with respect torotation about the long axis of the molecule [17,32–35].

It can be seen in Fig. 2 that the homogeneous nucleation, andthus the supercooling, can be efficiently suppressed in the octade-cane microcapsule when the F:M ratio in the precursor is tunedaround 1.25. Previous studies have found that the F:M ratio has lit-tle effect on the microcapsule size, but the molecular structure andcomposition of the resultant resin shell can be significantlydifferent [36,37]. As different molecular structures of the wall

Fig. 1. SEM and TEM images of the microcapsules with octadecane encapsulated inthe melamine–formaldehyde resin shell. (a) SEM image of the sample producedwith 50 g/L SDS. (b) SEM and TEM (inserted) image of the sample prepared with10 g/L SDS. The insert is a TEM image showing the shell thickness. (c) micro/nanoparticles in filtrate.

40200-0.5

0.0

0.5

1.0

1.5

2.0

2.5γ β

Exo

up

B

C

D

Hea

t flo

w (W

/g)

Temperature (oC)

D: MC04, F:M=1.25C: MC03, F:M=1.30B: MC02, F:M=1.50A: MC01, F:M=2.00

A

F:M

dec

reas

es

α

Fig. 2. DSC freezing curves of samples MC01-04 with various F:M ratio inprecursors. Curves are shifted along Y axis.

a

0

1

2

3

4

5

6

F

E

D

B

C

A: MC09, pH = 9.00 B: MC08, pH = 8.75 C: MC07, pH = 8.50 D: MC04, pH = 8.30 E: MC06, pH = 8.25 F: MC05, pH = 8.00

Hea

t Flo

w (W

/g)

Temperature (oC)

γ β α

21.2

22.0

22.2

21.3

pH in

crea

ses

23.4

A

b

0 10 20 30 40

8.0 8.2 8.4 8.6 8.8 9.00.0

0.2

0.4

0.6

0.8

1.0

Rel

ativ

e La

tent

Hea

t

pH of pre-polymer solution

γβα

α

β

γ

Fig. 3. (a) DSC freezing curves of samples MC04-09. Curves are shifted along Y axis.(b) Relative latent heat of phase transition peaks for various pH values of pre-condensation solution.

F. Cao, B. Yang / Applied Energy 113 (2014) 1512–1518 1515

supply different forms of heterogeneous nucleation sites for theencapsulated PCMs, some of the molecular structures on the innerwall of the resin shell could be more suitable than the others topromote the nucleation of the metastable rotator phase or the tri-clinic crystalline phase.

Two types of functional groups form on the triazine rings of thepre-polymer molecules, the amino groups and the hydroxylmeth-ylamino groups, as described in Eq. (4). These two functionalgroups may undergo two different condensation or polymerizationreactions in the acidic environment to form the resin shell, i.e., theamino-hydroxylmethylamino condensation and the hydroxylm-ethylamino-hydroxylmethylamino condensation:

—NH2 þ—NH—CH2—OH!Hþ

—NH—CH2—NH—þH2O ð6Þ

—NH—CH2—OHþ—NH—CH2—OH!Hþ

—NH—CH2—O—CH2—NH—þH2O ð7Þ

Meanwhile, some of the functional groups stay suspended dur-ing the reaction process due to steric hindrance. The two parallelreactions produce the MFR shell of the microPCMs jointly, as

0

1

2

3

4

5

6

A

B

C

D

E

Hea

t Flo

w (W

/g)

Temperature (oC)

A: MC10, 0.75ml/L B: MC11, 1.00ml/L C: MC12, 1.25ml/L D: MC13, 1.38ml/L E: MC14, 1.50ml/L F: MC15, 1.75ml/L

F 21.7

22.6

22.2

21.9

22.1

22.2

γ β α

0 10 20 30 40

0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.70.0

0.2

0.4

0.6

0.8

1.0

Rel

ativ

e La

tent

Hea

t

volume of HAc in 1 L emulsion (ml)

γ β α α

β

γ

a

b

Fig. 4. (a) DSC freezing curves of samples MC10-15. Curves are shifted along Y axis.(b) Relative latent heat of phase transition peaks for various HAc concentrations inthe emulsion.

1516 F. Cao, B. Yang / Applied Energy 113 (2014) 1512–1518

shown in Eqs. (5) and (6). Apparently, the more hydroxylmethylgroups produced in prepolymerization, the more condensationreactions happen between the hydroxyl groups and betweenhydroxylmethyl and amino groups, and thus the larger complete-ness the crosslink of the MFR polymer. On the other hand, thehydroxylmethyl groups are produced by the reaction of melamineand formaldehyde, as shown in Eq. (3). In the case of F:M ratio lar-ger than the stoichiometric F:M ratio of the prepolymerization (i.e.,>1.5), a large number of hydroxylmethyl groups are produced andthus leading to a higher degree of crosslink in the MF resin shell ofthe microcapsules. A large supercooling, associated with the li-quid-triclinic crystalline transition induced by homogeneousnucleation, is observed in these microcapsules prepared with lar-ger F:M ratio. By contrast, the supercooling of crystallization canbe largely eliminated when the F:M ratio reduces to 1.25, wherethe resin shell has lower cross-linking degree and more free-stand-ing functional groups. It is worth noting that further lowering theF:M ratio (<1.25) yields instable pre-polymer and not formationof microcapsules.

According to the aforementioned analysis, we could argue thatthe resin shell with less degree of crosslink and more free-standingfunctional groups could promote nucleation of the metastablerotator phase and the triclinic phases and thus suppress the super-cooling associated with homogenous nucleation, which is also pro-ven in the following discussions.

3.3. Effects of pH values in the prepolymer on phase transitionproperties of microcapsules

The pH value in the prepolymer solution also affects the molec-ular structure and composition of the MFR shells of the microcap-sules and thus their phase transition behavior. Both the pH valueand F:M ratio control the degree of hydroxylation and the balancebetween hydroxylated melamine and the precursor melamine andformaldehyde in the pre-polymer solution, as described in Eq. (3).Fig. 3a shows the DSC curves of the octadecane microcapsules pre-pared with various pH value but constant F:M ratio, and Fig. 3blists the relative latent heat of exothermic peaks a, b, and c. The rel-ative latent heat of an exothermic peak is determined as the ratioof its peak area to the total area of three peaks under the DSCcurves. It can be seen in Fig. 3a that the optimum pH value inthe pre-polymer is 8.50 where peak c (and its associated superco-oling of homogeneous nucleation) are eliminated and the area ofpeak a is maximized. The maximum relative latent heat of phasetransition a is 80.8% while the relative latent heat of peak b is19.2%, shown in curve 3C for sample MC07. In addition, the onsettemperature of freezing improves from 21.3 �C to 23.4 �C whenthe pH value tunes from 8.00 to 8.50 at the constant F:M ratio1.25. This result implies that to minimize the two low-temperaturepeaks (the homogeneous nucleation freezing peak c and the rota-tor-triclinic phase transition peak b) and so as to maximize the li-quid–solid peak a, a properly crosslinked quasi-linear polymericstructure of MFR is preferred for the nucleation of the solid phaseof encapsulated octadecane.

3.4. Effects of HAc concentration in emulsion on phase transitionproperties of microcapsules

HAc concentration in the emulsion is another parameter thataffects the molecular structure and composition of the microcap-sule shell and consequently the phase change behavior of theoctadecane PCM. FM polymers with two different bridges, i.e.,etheric (–C–O–C–) and methylene (–C–) bridges, can be fabri-cated, as shown in Eqs. (6) and (7). The HAc concentration andthe resulting pH value in the emulsion have a significant influ-ence on the ratio of these two types of bridges and the polymeric

structure of the MF resin shell. The formation of etheric bridges ismore preferable than that of methylene bridges in a low pH value(acidic) environment [37]. Fig. 4a shows the DSC curves of theoctadecane microcapsules prepared with different HAc concentra-tion in the emulsion, and Fig. 4b lists the relative latent heat ofexothermic peaks a, b, and c. It can be seen in Fig. 4a and b thatthe relative latent heat of peak c decrease to zero when the HAcconcentration is in the range from 1.38 ml/L to 1.75 ml/L withthe fixed F:M ratio (1.25) and pH value of the pre-polymer(pHprepolymer = 8.50), The relative latent heat of peak a ismaximized, reaching 83.7% at the concentration of HAc in theemulsion CHAc = 1.38 ml/L, while the relative latent heat of peakb is 16.6%. The effective latent heat of these octadecane microcap-sules can be as large as 213 J/g, and the corresponding weightratio of the phase change octadecane reaches 88 wt.%, which aresignificantly higher than those in the microcapsules that useadditive as nucleating agents [17,18].

3.5. Effects of supercooling suppression on specific heat elevation ofmicrocapsules

Based on latent heat of the microPCMs, the apparent heatcapacity enhancements of with various operation temperatureranges of the microPCMs can be calculated by Eq. (2), which are

2 4 6 8 10 12 14 16 18 20

10

15

20

25

30

35

40A

ppar

ent s

peci

fic h

eat i

ncre

ase

(J/g

K)

Operation temperature range (oC)

Fig. 5. Calculation of apparent specific heat increase by the latent heat ofmicroPCMs corresponding to various operation temperature ranges. The horizontalline shows the increase at an operation temperature range of 13.6 �C.

F. Cao, B. Yang / Applied Energy 113 (2014) 1512–1518 1517

shown in Fig. 5. The apparent heat capacity enhancement is reci-procal to the operation temperature range DT. A significant in-crease of apparent specific heat has been demonstratedcomparing with the increase by microPCMs without shell optimi-zation. For instance, with a reasonable DT of 5 K and only latentheat of phase transition a is utilized, the apparent specific heatcould be increased to 35.7 J/g, 16.6 times larger as the specific heatof solid octadecane without phase transition. In another case, whenthe operation temperature range is extended to 10 K to cover bothphase transition peaks a and b, more latent heat can be utilized,though the apparent specific heat increase drops to 21.3 J/g.

4. Conclusion

The supercooling of PCM microcapsules has been suppressedsignificantly without compromising their effective latent heat offusion, through optimization of the composition and structure ofthe microcapsule resin shell. The octadecane microcapsules aresynthesized by using the oil-in-water emulsion technique. The ef-fects of synthesis parameters, such as ratio of melamine to formal-dehyde, pH of pre-polymer, and pH of emulsion, on the phasetransition properties of the octadecane microcapsules have beeninvestigated systemically. Important observations are listed asfollows:

(1) The homogenous nucleation of the octadecane triclinic crys-tal and its associated supercooling can be eliminated by theshell-induced liquid–crystal transition. To provide appropri-ate nucleation sites at the inner wall of the microcapsule, theshell composition and microstructure can be optimized bytuning the synthesis parameters, such as F:M ratio, pH inthe prepolymer, and HAc concentration.

(2) Three peaks, labeled a, b, and c are observed on the DSC cool-ing curves of the octadecane microcapsules, which are attrib-uted to the shell-induced liquid-rotator and liquid-triclinictransition, rotator-crystal transition, and homogeneouslynucleated liquid-triclinic transition, respectively. The rela-tive latent of heat of peak a can be maximized, reaching83.7% in the octadecane microcapsules by shell optimization.

(3) The effective latent heat of these octadecane microcapsulescan be as large as 213 J/g, and the corresponding weightratio of the phase change octadecane reaches 88 wt.%, whichare significantly higher than those in the microcapsules thatuse additive as nucleating agents.

(4) Hollow spheres with porous, rather than solid resin shell, areformed when the SDS concentration is high, for example,50 g/L in emulsion. When the SDS concentration drops to10 g/L, octadecane microcapsules with solid resin shell canbe formed.

Acknowledgements

This research is financially supported by NSF under Grant1336778.

Appendix A. Supplementary materialy material

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.apenergy.2013.08.048.

References

[1] Delgado M, Lazaro A, Mazo J, Zalba B. Review on phase change materialemulsions and microencapsulated phase change material slurries: materials,heat transfer studies and applications. Renew Sust Energ Rev 2012;16:253–73.

[2] Diaconu BM, Varga S, Oliveira AC. Experimental assessment of heat storageproperties and heat transfer characteristics of a phase change material slurryfor air conditioning applications. Appl Energy 2010;87:620–8.

[3] Mehling H, Cabeza LF. Heat and cold storage with PCM: an up to dateintroduction into basics and applications. Berlin, Germany: Springer; 2008.

[4] Farid MM, Khudhair AM, Razack SAK, Al-Hallaj S. A review on phase changeenergy storage: materials and applications. Energ Convers Manage 2004;45:1597–615.

[5] Hasnain SM. Review on sustainable thermal energy storage technologies, partI: heat storage materials and techniques. Energ Convers Manage 1998;39:1127–38.

[6] Xu B, Li Z. Paraffin/diatomite composite phase change material incorporatedcement-based composite for thermal energy storage. Appl Energ 2013;105:229–37.

[7] Zhao CY, Zhang GH. Review on microencapsulated phase change materials(MEPCMs): fabrication, characterization and applications. Renew Sust EnergRev 2011;15:3813–32.

[8] Han ZH, Cao FY, Yang B. Synthesis and thermal characterization of phase-changeable indium/polyalphaolefin nanofluids. Appl Phys Lett 2008;92:243104–5.

[9] Oró E, de Gracia A, Castell A, Farid MM, Cabeza LF. Review on phase changematerials (PCMs) for cold thermal energy storage applications. Appl Energ2012;99:513–33.

[10] Mondal S. Phase change materials for smart textiles – An overview. ApplTherm Eng 2008;28:1536–50.

[11] Han ZH, Yang B. Thermophysical characteristics of water-in-FC72nanoemulsion fluids. Appl Phys Lett 2008;92:013118–9.

[12] Kalinowski P, Lawler J, Yang B, Cao F. Heat Transfer Performance of a PhaseChange Microcapsule Fluid. ASME 2012 3rd Micro/Nanoscale Heat and MassTransfer International Conference. Atlanta, Georgia, USA: ASME; 2012. p.MNHMT2012-75190.

[13] Chen Z-H, Yu F, Zeng X-R, Zhang Z-G. Preparation, characterization andthermal properties of nanocapsules containing phase change material n-dodecanol by miniemulsion polymerization with polymerizable emulsifier.Appl Energ 2012;91:7–12.

[14] Rao Z, Wang S, Peng F. Self diffusion of the nano-encapsulated phase changematerials: a molecular dynamics study. Appl Energ 2012;100:303–8.

[15] Yamagishi Y, Sugeno T, Ishige T, Takeuchi H, Pyatenko AT. An evaluation ofmicroencapsulated PCM for use in cold energy transportation medium. EnergyConversion Engineering Conference, 1996 IECEC 96, Proceedings of the 31stIntersociety 1996. p. 2077–83, vol. 3.

[16] Lee W.-m. Microcapsule containing phase change material and article havingsame. US Patent Application 10/611417 2004.

[17] Fan YF, Zhang XX, Wang XC, Li J, Zhu QB. Super-cooling prevention ofmicroencapsulated phase change material. Thermochim Acta 2004;413:1–6.

[18] Zhang XX, Fan YF, Tao XM, Yick KL. Crystallization and prevention ofsupercooling of microencapsulated n-alkanes. J Coll Interface Sci 2005;281:299–306.

[19] Alvarado JL, Marsh C, Sohn C, Phetteplace G, Newell T. Thermal performance ofmicroencapsulated phase change material slurry in turbulent flow underconstant heat flux. Int J Heat Mass Transf 2007;50:1938–52.

[20] Palanikkumaran M, Gupta KK, Agrawal AK, Jassal M. Effect of emulsionpreparation method on microencapsulation of n- octadecane using melamine-formaldehyde pre-polymers. Indian J Fibre Text Res 2010;35:101–6.

[21] Salaun F, Devaux E, Bourbigot S, Rumeau P. Influence of process parameters onmicrocapsules loaded with n-hexadecane prepared by in situ polymerization.Chem Eng J 2009;155:457–65.

1518 F. Cao, B. Yang / Applied Energy 113 (2014) 1512–1518

[22] Sumiga B, Knez E, Vrtacnik M, Savec VF, Staresinic M, Boh B. Production ofmelamine-formaldehyde PCM microcapsules with ammonia scavenger usedfor residual formaldehyde reduction. Acta Chim Slovenica 2011;58:14–25.

[23] Su J-F, Wang X-Y, Wang S-B, Zhao Y-H, Huang Z. Fabrication and properties ofmicroencapsulated-paraffin/gypsum-matrix building materials for thermalenergy storage. Energy Conser Manag 2012;55:101–7.

[24] Sun G, Zhang Z. Mechanical strength of microcapsules made of different wallmaterials. Int J Pharm 2002;242:307–11.

[25] Palanikkumaran M, Gupta KK, Agrawal AK, Jassal M. Highly stablehexamethylolmelamine microcapsules containing n-octadecane prepared byin situ encapsulation. J Appl Polym Sci 2009;114:2997–3002.

[26] Long Y, York D, Zhang Z, Preece JA. Microcapsules with low content offormaldehyde: preparation and characterization. J Mater Chem 2009;19:6882–7.

[27] Lee HY, Lee SJ, Cheong IW, Kim JH. Microencapsulation of fragrant oil viain situ polymerization: effects of pH and melamine-formaldehyde molar ratio.J Microencapsul 2002;19:559–69.

[28] Erkan G, Sariisik M, Pazarlioglu NK. The microencapsulation of terbinafine viain situ polymerization of melamine-formaldehyde and their application tocotton fabric. J Appl Polym Sci 2010;118:3707–14.

[29] Kumar A, Katiyar V. Modeling and experimental investigation of melamineformaldehyde polymerization. Macromolecules 1990;23:3729–36.

[30] Coullerez G, Leonard D, Lundmark S, Mathieu HJ. XPS and ToF-SIMS study offreeze-dried and thermally cured melamine-formaldehyde resins of differentmolar ratios. Surf Interface Anal 2000;29:431–43.

[31] Xie B, Liu G, Jiang S, Zhao Y, Wang D. Crystallization behaviors of n-octadecanein confined space: crossover of rotator phase from transient to metastableinduced by surface freezing. J Phys Chem B 2008;112:13310–5.

[32] Sirota EB. Supercooling, nucleation, rotator phases, and surface crystallizationof n-alkane melts. Langmuir 1998;14:3133–6.

[33] Sirota EB, Herhold AB. Transient rotator phase induced nucleation in n-alkanemelts. Polymer 2000;41:8781–9.

[34] Kraack H, Sirota EB, Deutsch M. Measurements of homogeneous nucleation innormal-alkanes. J Chem Phys 2000;112:6873–85.

[35] Wu XZ, Ocko BM, Sirota EB, Sinha SK, Deutsch M, Cao BH, et al. Surface-tensionmeasurements of surface freezing in liquid normal-alkanes. Science1993;261:1018–21.

[36] Luo W-j, Yang W, Jiang S, Feng J-M, Yang M-b. Microencapsulation ofdecabromodiphenyl ether by in situ polymerization: preparation andcharacterization. Polym Degrad Stabil 2007;92:1359–64.

[37] Jahromi S. Storage stability of melamine-formaldehyde resin solutions, 1. Themechanism of instability. Macromol Chem Phys 1999;200:2230–9.