Drying and Denaturation Kinetics of Whey Protein Isolate (WPI) During Convective Air Drying Process

This article was downloaded by: [RMIT University]On: 17 July 2014, At: 15:38Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK

Drying Technology: An International JournalPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/ldrt20

Drying and Denaturation Kinetics of Whey ProteinIsolate (WPI) During Convective Air Drying ProcessM. Amdadul Haque a b , Aditya Putranto c , Peter Aldred a , Jie Chen d & Benu Adhikari aa School of Health Sciences, University of Ballarat , Mount Helen , Victoria , Australiab Department of Agro-processing , BSMRAU , Gazipur , Bangladeshc Department of Chemical Engineering , Monash University , Clayton , Victoria , Australiad State Key Laboratory of Food Science and Technology, Jiangnan University , Wuxi , ChinaPublished online: 26 Sep 2013.

To cite this article: M. Amdadul Haque , Aditya Putranto , Peter Aldred , Jie Chen & Benu Adhikari (2013) Drying andDenaturation Kinetics of Whey Protein Isolate (WPI) During Convective Air Drying Process, Drying Technology: An InternationalJournal, 31:13-14, 1532-1544, DOI: 10.1080/07373937.2013.794832

To link to this article: http://dx.doi.org/10.1080/07373937.2013.794832

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) containedin the publications on our platform. However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of theContent. Any opinions and views expressed in this publication are the opinions and views of the authors, andare not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon andshould be independently verified with primary sources of information. Taylor and Francis shall not be liable forany losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoeveror howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use ofthe Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Drying and Denaturation Kinetics of Whey Protein Isolate(WPI) During Convective Air Drying Process

M. Amdadul Haque,1,2 Aditya Putranto,3 Peter Aldred,1 Jie Chen,4 andBenu Adhikari11School of Health Sciences, University of Ballarat, Mount Helen, Victoria, Australia2Department of Agro-processing, BSMRAU, Gazipur, Bangladesh3Department of Chemical Engineering, Monash University, Clayton, Victoria, Australia4State Key Laboratory of Food Science and Technology, Jiangnan University, Wuxi, China

The denaturation and drying kinetics of whey protein isolate(WPI) in a convective drying (CD) environment was measured usingsingle droplet drying experiments. The moisture content and tem-perature histories during drying of WPI droplets were predictedusing reaction kinetics–based models. The denaturation kinetics ofWPI in the CD process was predicted using first-order reactionkinetics considering the denaturation rate constant to be moisturecontent and temperature dependent. Single droplets of WPI (10%[w/v], 2.0� 0.1mm initial diameter) were used throughout theseexperiments. The drying experiments were carried out at two tem-peratures (65 and 80�C) at a constant air velocity (0.5m/s) for600 s. The extent and nature of the denaturation of WPI duringthe CD was compared with those in isothermal heat treatments(IHT) at the same medium temperatures. The denaturation ofWPI was 68.31% in convective air drying at 65�C and 600 s andit was 10.79% in the IHT at the same temperature and time. Thestress due to dehydration and the exposure time were found to beresponsible for the denaturation of WPI in the CD process and longexposure time was found to be responsible for its denaturation in theIHT process. At the media temperature of 80�C, the denaturationloss of WPI was 90.00 and 68.73% in IHT and CD processes,respectively. Both the thermal (moist heat) and dehydration stresseswere found to be responsible for denaturation of WPI during CDprocess and very high thermal stress was found to be responsiblefor denaturation of WPI during the IHT. There was good agree-ment between the experimental and reaction engineering approach(REA)-predicted moisture content and temperature histories. Theexperimental moisture content and temperature histories were fol-lowed by the respective REA predictions within 6.5% (R2¼ 0.995)and 3% (R2¼ 0.981) errors, respectively. The denaturation kineticsof WPI during CD was predicted well (R2¼ 0.95 – 0.98; averageerror¼ 6.5� 0.5%) by a first-order reaction kinetics model.

Keywords Denaturation kinetics; Drying kinetics; Isothermalheat treatment; Reaction engineering (REA)approach; Single droplet drying; Whey protein isolate

INTRODUCTION

Proteins derived from various natural sources such asplant, animal, and milk are converted into dry powderform to enhance their stability and long-term storage. Eventherapeutic proteins such as antibodies, which are usuallydelivered after reconstitution, are also converted intopowder.[1] Dry therapeutic proteins are being increasinglyused in native form for inhalation, pulmonary, and trans-dermal delivery.[2] Nontherapeutic proteins such as milkproteins are important ingredients in manufactured func-tional and health foods. They are also converted into drypowder form for convenience of use and longer shelf life.

Proteins are first extracted into water from their sourcesand subsequently the water is removed by drying. Whendried, proteins are less sensitive to heat and other stressesand become chemically more stable. Spray drying is awell-established method for drying food as well as phar-maceutical powders and it has the advantages of beinghygienic, having high throughput and simple operation,and producing powders with good particle size distri-bution.[3] However, the process of conversion of proteinsolutions into dry particulate form is complex because ofthe sensitivity of proteins to heat, especially when they arestill in solution.[4] Spray drying of proteins has been studiedby pharmaceutical and food science researchers.[5–7] Thesestudies show that a considerable amount of protein is inac-tivated or denatured during the powder formation process.

The implication of denaturation of protein is the irre-versible damage to both the secondary (a-helix, b-sheet,and random coil) and tertiary structures.[8,9] Thesedenatured protein powders do not redissolve well and losetheir therapeutic effectiveness and other functional proper-ties such as gelling and emulsification. The denatured=aggregated proteins may give rise to malfunctioning ofliving systems and hence may cause functional problemsin vital organs such as brain, liver, spleen, and skeletaltissues.[10–12] In addition, denaturation of protein

Correspondence: M. Amdadul Haque, School of HealthSciences, University of Ballarat, S 113, Mt. Helen Campus,Ballarat, VIC 3353, Australia; E-mail: [email protected]

Drying Technology, 31: 1532–1544, 2013

Copyright # 2013 Taylor & Francis Group, LLC

ISSN: 0737-3937 print=1532-2300 online

DOI: 10.1080/07373937.2013.794832

1532

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

15:

38 1

7 Ju

ly 2

014

negatively impacts the properties of consumer productswhen they are used as ingredients.

Whey protein is a mixture of globular proteins. It is oneof the most commonly used proteins in the dairy, bakery,and health care industries. This protein is also quite com-monly used as a nutritional supplement or additive in for-mulated foods because of its ability to supply all of theessential amino acids. In addition, whey proteins are con-sidered to be highly nutritious proteins due to the presenceof essential minerals such as calcium (present in the whey)and sulfur and they have a very high digestibility score.[13,14]

However, it has been reported that a considerable amountof whey proteins are denatured during the dehydrationprocess. Singh and Creamer carried out a thorough inves-tigation on the denaturation pattern of whey proteinsduring skim milk powder production.[15] They found thatabout 90% of b-lactoglobulin and 40% of a-lactalbuminpresent in the whey proteins were denatured. Theseauthors found that most of the denaturation occurred atthe preheating section of skim milk manufacturing.Anandharamakrishnan et al. found that up to 70% dena-turation of whey protein isolates (WPI) occurred in powdermanufacturing processes, including drying.[13] Hence, it isof practical importance to minimize denaturation of wheyproteins in drying processes in order to maintain theirnatural characteristics and to broaden the range of theirapplication. Therefore, it is essential to quantify the extentand nature of drying and denaturation kinetics of proteinsduring the formation of dried protein particles. A betterunderstanding of the physics of drying and denaturationof whey proteins allows development of process protocolsthat result in minimal protein denaturation.

Reversed-phase high-performance liquid chromato-graphy (RP-HPLC) is commonly used as a suitable anal-ytical method for measuring protein denaturation.[16,17]

The strong affinity of the hydrophobic groups of denaturedproteins to the long nonpolar alkyl chains (i.e., C8, C18)used in the RP-HPLC column makes it possible for thismethod to separate the denatured and undenaturedproteins.

The reaction engineering approach (REA) applies theprinciples of chemical reaction engineering in modelingand simulation of drying processes. It was proposed byChen in 1996[18,19] and has been successfully used to modela number of complex drying processes.[20–22] In REA, theevaporation is modeled as a zero-order reaction with acti-vation energy and the condensation is described as afirst-order reaction with respect to the concentration ofwater vapor in air without activation energy. The REAdoes not assume the uniform moisture content but itevaluates the average moisture content during drying. Ithas been proven to be a simple, accurate, and effectiveapproach in modeling the simultaneous heat and masstransfer processes that occur during drying.[23,24]

It is not yet possible to measure the denaturationkinetics of protein droplets subjected to spray drying dueto its closed configuration, circulation of particles withinthe drying chamber, and the presence of millions ofdroplets=particles at any cross section of the drying cham-ber. Therefore, the denaturation kinetics data of proteins inthe spray-drying process are not available; hence, the dena-turation behavior of proteins during the particle formationprocess of spray drying is not well understood. Because it isnot yet possible to measure the drying kinetics of proteindroplets in situ in the spray-drying process, researchers fre-quently use single droplet drying (SDD) to measure thesedrying kinetics and to monitor the surface morphology ofthe drying droplets.[25–27] However, the drying and dena-turation kinetics of whey proteins are not measured andmodeled in an integral way even though the drying kineticsof whey proteins are available in very limited instances.[28]

In this context, this study had three objectives: firstly, tomeasure the drying and denaturation kinetics of WPI in theconvective drying (CD) process using an SDD instrument.We also compared the extent of denaturation of WPI inCD with that in isothermal heat treatment (IHT) at thesame medium temperature; secondly, to predict the dryingkinetics (moisture and temperature histories) of WPIsolution droplets in a CD environment and presentthis in an integral way with the experimentally measuredand mathematically modeled denaturation kinetics; and,finally, to observe the morphological changes in the WPIdroplets=particles during CD and the dissolution processes.

EXPERIMENTAL

Materials

a-Lactalbumin (a-lac), b-lactoglobulin (b-Lg A and B),and bovine serum albumin (BSA) with �95% purity werepurchased from Sigma-Aldrich (New South Wales,Australia). The HPLC solvents acetonitrile and trifluoroa-cetic acid (TFA) were purchased from Sigma-Aldrich andFisher Scientific (Victoria, Australia), respectively. Thewhey protein isolate (WPI-895) was donated by FonterraCooperative (New Zealand). All proteins and chemicalswere used as received without further purification.

Methods

The protein droplets=particles subjected to SDD andIHT processes were collected and the amount of proteindenaturation in these treated samples was measured byusing an RP-HPLC. The drying kinetics and denaturationkinetics of WPI were predicted and the data obtainedfrom SDD experiments were used to determine the modelparameters and to validate the model predictions. Theevolution=development of morphology in WPI droplets=particles during drying and dissolution processes wereobserved during the same SDD instrument.

WHEY PROTEIN ISOLATE 1533

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

15:

38 1

7 Ju

ly 2

014

Single Droplet Drying

The design, working principle, and other details of thisinstrument are provided elsewhere.[29] The aqueous dropletsof (5� 0.1) mL of WPI (10% w=v) solution were suspendedon the tip of a glass filament of 0.2� 0.01mm diameter.The WPI droplets, which were suspended from the glassfilaments, were held inside the drying chamber using a boredTeflon cylinder. The drying air was allowed to pass througha series of columns containing dried molecular sieves anddried silica gels. This dried air was finally passed throughthe drying chamber. Two drying temperatures (65 and80�C) were used and these temperatures were maintainedusing a proportional–integral–derivative controller. Therelative humidity values of the drying air were measuredby using an EL-USB relative humidity data logger (LascarElectronics, Salisbury, UK) and were found to be 2� 0.5%at these two drying temperatures. The wet bulb temperatureswere 26 and 30�C at air temperatures of 65 and 80�C,respectively. The droplets were dried for 60 to 480 s at 60-sintervals and the final reading was made at 600 s. The velo-city of the air past the stationary droplet was maintained at0.5m=s in all experiments. Mass loss and temperaturehistories were recorded in a computer. The morphologicalchanges of the droplets=particles were recorded to the com-puter using a digital camera (Arecont Vision, Los Angeles,CA, USA) that was attached to the SDD instrument.

Isothermal Heat Treatment

The IHT experiments were carried out in a preciselytemperature-controlled water bath (Thermoline, Australia).A heater=circulator was used to pump the water so that theheat can be circulated evenly throughout the water bath.The temperature was regulated using a temperature controlknob and a thermometer fitted with the water bath. Asample volume of 0.5mL WPI solution (used in SDDexperiments) was poured in 4-mL glass vials. These vialswere then heated at 65 and 80�C within the same timeregime used during SDD. Thermocouples were inserted intwo additional vials containing the same amount of WPIsolution during the IHT experiments to record the tempera-ture profiles during the heating of the samples.

Determination of Denaturation of WPI Using RP-HPLC

The dried droplets were collected carefully in Eppendorfmicrotubes and diluted to 100 mg=mL with deionized water.Then, the pH of the diluted protein solutions was loweredto 4.6 by adding 0.1M HCL to separate the native and irre-versibly denatured whey proteins.[17,30] After keeping theprotein solution samples at this pH for 30min, they werecentrifuged at 13,000 rpm (11,500 g) for 10min. The super-natant was collected and injected onto the reversed-phasecolumn to separate the denatured and undenatured pro-teins as described by Ferreira and Cacote.[16] The sameamount of WPI solution not subjected to the drying or

thermal treatment was used as a control. The same pHtreatment and centrifugation procedures were followedfor the control samples as well. The supernatant of the con-trol samples was also injected in the same way to determinethe denatured and undenatured fractions. This procedurewas also followed in the case of isothermal heat-treatedWPI samples to determine the extent of denaturation.

A reversed-phase column (Aeris WIDEPORE 3.6XB-C18, 150� 2.1mm, Phenomenex Pty. Ltd., Sydney,Australia) was used to separate the denatured and undena-tured protein fractions. In these tests a 0.1% (v=v) aqueoussolution of TFA was used as solvent A and 0.1% TFA in80% aqueous acetonitrile (v=v) was used as solvent B. Alinear increment of solvent B, from 36 to 56% over20min and 56 to 60% over 10min, was maintained andthe process was completed by lowering solvent B from 60to 36% in 5min. Finally, the column was reequilibratedfor 3min before the next run started. The temperature ofthe column and the flow rate of the solvent were main-tained at 40�C and 0.5mL=min, respectively. The injectionvolume of the sample was 40 mL. An ultraviolet detector atwavelength of 215 nm was used in these tests.

The percentage of denatured protein was calculatedusing Eq. (1) using the sum of areas under the peaks ofindividual protein fractions (BSA, a-lac, b-LgA, andb-LgB) in treated WPI and the sum of the peaks of theseprotein fractions in control samples, as suggested by Parrisand Baginski.[17] The denatured protein was quantified bymeasuring the loss of protein by first precipitating outand then centrifuging and finally passing the supernatantto the RP-HPLC column. The elution profiles of the wheyprotein fractions were standardized by using the standardindividual proteins (BSA, a-lac, b-Lg A, and b-Lg B).

Denatured protein %ð Þ ¼ 1� RAtreat=RAcontrolð Þ � 100;

ð1Þ

where RAtreat is the sum of areas under peaks of the proteinfractions of treated sample, and RAcontrol is the sum ofareas under peaks of the protein fractions of control sam-ple, These experiments were carried out in triplicate and theaverage values are reported in the ensuing sections.

Drying Kinetics Modeling Using the Reaction EngineeringApproach

The general REA is an application of chemical reactionengineering principles to model drying kinetics. A sum-mary of the developments of the lumped approach ofREA was given by Chen.[19]

Generally, the drying rate of a material can be expressedas given by Eq. (2):

msdX

dt¼ �hmAðqv;s � qv;bÞ; ð2Þ

1534 HAQUE ET AL.

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

15:

38 1

7 Ju

ly 2

014

where ms is the dry solid mass of the sample (kg), X is theaverage moisture content on a dry basis (kg � kg�1), t is time(s), qv,s is the water vapor concentration at the interface(kg �m�3), qv,b is the water vapor concentration in thedrying medium (kg �m�3), hm is the mass transfer coeffi-cient (m � s�1), and A is the surface area of the droplet=particle (m2).

Equation (2) is a basic mass transfer equation. The sur-face vapor concentration (qv,s) can then be scaled againstsaturated vapor concentration (qv,sat) using Eq. (3):

qv;s ¼ exp�DEv

RT

� �qv;satðTÞ; ð3Þ

where DEv represents the additional difficulty to removemoisture from the material when it is not free water orwater activity is not 1. The parameter DEv depends onthe average moisture content (X). R is the universal gasconstant (J �mol�1 �K�1). T is the temperature of thedroplet=particle being dried (K). The qv,sat (kg �m�3) forwater vapor can be estimated using Eq. (4)[28,31]:

qv;sat ¼ 4:844� 10�9ðT � 273Þ4 � 1:4807

� 10�7ðT � 273Þ3 þ 2:6572� 10�5ðT � 273Þ2

� 4:8613� 10�5ðT � 273Þ þ 8:342� 10�3: ð4Þ

The mass balance (Eq. (2)) can be rearranged into Eq. (5)by replacing qv,s by Eq. (3).

msdX

dt¼ �hmA exp

�DEv

RT

� �qv;satðTÞ � qv;b

� �: ð5Þ

From Eq. (5), it can be observed that the REA is expressedin the first-order ordinary differential equation with respectto time. It must be noted that the REA does not assume theuniform moisture content but it evaluates or predicts theaverage moisture content.

The activation energy (DEv,b) can be calculated byrearranging Eq. (5) into Eq. (6):

DEv ¼ �RT ln�ms

dXdt

1hmA

þ qv;bqv;sat Tð Þ

" #: ð6Þ

The rate of moisture loss dX=dt is experimentally deter-mined. The dependence of activation energy on averagemoisture content (dry basis, X) can be normalized as

DEv

DEv;b¼ f X � Xbð Þ; ð7Þ

where f is a function of the difference in water content(X �Xb), and DEv,b is the equilibrium activation energyrepresenting the maximum DEv determined by the relative

humidity and temperature of the drying air as given byEq. (8):

DEv;b ¼ �RTb lnðRHbÞ: ð8Þ

RHb is the relative humidity of drying air, Xb is the equilib-rium moisture content under the condition of the drying air(kg � kg�1), and Tb is the drying air temperature (K).

In order to yield the temperature profiles during drying,the mass balance equation represented by Eq. (5) needs tobe coupled with the heat balance. For the SDD systemmentioned above, the heat balance can be written as

mCpddT

dt¼ hAðTb � TÞ þms

dX

dtDHv

þ 0:5pdfffiffiffiffiffiffiffiffiffiffiffiffihdf kf

pðTb � TÞ; ð9Þ

where m is the mass of the droplet=particle (kg), Cpd is thespecific heat capacity of droplet=particle (J � kg�1 �K�1), his the heat transfer coefficient (W �m�2 �K�1), and DHv isthe latent heat of vaporization of water (J � kg�1). The thirdterm on the right-hand side of Eq. (9) represents the heatconducted through the filament to the droplet assumingthe filament as an infinite fin, as given by the equation.[29]

The mass transfer coefficient, hm, and heat transfer coef-ficient, h, can be determined by using the establishedRanz-Marshall correlations for Sherwood number (Sh)and Nusselt number (Nu) of spherical droplets subjectedto convective air drying.[32] In terms of Sh and Nu, boththe hm and h can be expressed as given by Eqs. (10) and(11), respectively:

Sh ¼hmd

Dv¼ 2þ 0:6R0:5

e S0:33c ð10Þ

Nu ¼ hd

kb¼ 2þ 0:6R0:5

e P0:33r ; ð11Þ

where d is the diameter of the evaporating droplet=particle,Dv is the diffusivity of water vapor (m2=s), and kb is thethermal conductivity of air (W �m�1 �K�1). Re and Sc inEq. (10) and Pr in Eq. (11) are dimensionless Reynolds,Schmidt, and Prandtl numbers, respectively, whichwere calculated from the thermophysical properties ofdrying air.

Because the droplet diameter continuously changes dur-ing drying, a shrinkage model is incorporated in the mod-eling. The relationship between the variation in dropletswith time can be calculated from the variation of moisturecontent of the droplet with time using Eq. (12)[33]:

d

do¼ aþ 1� að Þ X � Xb

Xo � Xb; ð12Þ

WHEY PROTEIN ISOLATE 1535

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

15:

38 1

7 Ju

ly 2

014

where d is the droplet diameter (m) at time t, do is the initialdroplet diameter (m), and a is the model constant. X(t)were taken from experimental moisture content values asa function of time, and Xo is the initial moisture content(t¼ 0). By using least-squares method implemented inMS Excel, the values of the model constant a were foundto be 0.532 and 0.514 at drying temperatures of 65 and80�C, respectively, which are not significantly different(p> 0.05). For the same protein type with similar initialsolid concentration (dried by similar relative humiditybut different air temperatures), an average model constanta can be considered.[31,33] In this model the average value0.523 was used in calculating the droplet diameter.

The drying conditions set for the SDD of the WPI dro-plets are listed in Table 1. The air velocity and the dryingair temperature were controlled using a precision rotameter(Influx AI29, Influx Measurements Ltd., Hampshire, UK)and a proportional–integral–derivative controller (C91BrainChild Electronics, OneTemp Pty Ltd., Melbourne,Australia), respectively. The relative humidity of the dryingair was determined experimentally as detailed previously inthe Single Droplet Drying section of this article. The initialmoisture content of the droplet was determined by know-ing the mass of the droplet and the solid concentration ofthe solution. The diameter of the droplet was calculatedby knowing the mass of the droplet and density of the sol-ution and assuming the droplet to be spherical. The physio-chemical properties of WPI solution used in the dryingkinetics modeling are provided in the Appendix.

Modeling of the Denaturation Kinetics

The inactivation kinetics of microorganisms or dena-turation kinetics of enzymes during CD can generally beexpressed by the first-order reaction kinetics model(Eq. (13)).[34–36] This model has also been used in predictingthe denaturation kinetics of whey proteins duringhigh-temperature and high-pressure–high-temperaturetreatments.[37–40]

dC

dt¼ �kdC; ð13Þ

where kd is the denaturation rate constant (s�1), and C isthe protein concentration (%) at each time.

The denaturation rate constant, kd, can be expressed asa function of temperature and water content of the dryingdroplet as given by Eq. (14)[36]:

kd ¼ ko exp aX � Ea þ bX

RT

� �; ð14Þ

where T is the droplet temperature (K), X is the moisturecontent of the droplet (kg � kg�1), R is the universal gasconstant, Ea is the apparent activation energy (J �mol�1),and ko (s

�1), a, and b are the model constants.The parameters a, b, k, and Ea were estimated consider-

ing the simultaneous effects of moisture content (X) andtemperature (T) of the drying droplet. A least-squaresalgorithm (Microsoft Solver) was used in determining thevalues of these constants using the experimental WPIdenaturation kinetics data. The predicted values of X andT through the REA model were used in Eq. (14) in orderto minimize the noise.

RESULTS AND DISCUSSION

Denaturation of Protein During Convective Drying

The standard proteins were used to locate and define thepeaks of individual protein components (BSA, a-lac, b-LgB, and b-Lg A) on the chromatograms of WPI. It can beseen from Fig. 1A that the elution times of BSA, a-lac,b-Lg B, and b-Lg A were 18.80, 18.97, 20.94, and21.70min, respectively. This order of elution of these pro-teins in RP-HPLC is in good agreement with the elutiontime of these proteins reported by previous research-ers.[16,17] As can be seen from Fig. 1B, when the denatura-tion of protein occurs, the peak area and the retention timeof the denatured protein differed from those of the controlsample. The denaturation of proteins resulted in dimin-ished peak area under the curve and longer elution time.

It can be seen from Figs. 2A and 2B that there was aremarkable difference in the trends of denaturation ofWPI caused by the CD and IHT. When the IHT was car-ried out at a temperature at which thermal denaturationcan reportedly start (i.e., 65�C),[17,41] no irreversible dena-turation of WPI was observed up to 350 s of heating. How-ever, up to 10.79% of WPI was found to be irreversiblydenatured when the heating was prolonged for 600 s. Onthe other hand, when the WPI solution droplets were sub-jected to CD at air temperature of 65�C, a much higherpercentage (68.31%) of irreversible denaturation of WPIoccurred. The irreversibly denatured protein is the proteincontent that has been removed=separated through centrifu-gation and subsequent reversed-phase column extraction.The massive difference of denaturation of WPI betweenCD and IHT (at the same medium temperature) can be

TABLE 1Drying conditions during SDD of WPI

Experimental conditions65�C

Drying air80�C

Drying air

Initial temperature (�C) 23.804 24.896Initial moisture content(kg=kg dry solid)

9.75 9.75

Drying air velocity (m=s) 0.5 0.5Relative humidity (%) 2� 0.5 2� 0.5Initial droplet diameter (m) 0.002 0.002

1536 HAQUE ET AL.

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

15:

38 1

7 Ju

ly 2

014

attributed to the different stresses that occur in drying andisothermal heating (absence of drying) processes. In theCD process, the proteins experience both thermal (heat)

and dehydration (water evaporation) stresses, whereas theyexperience only thermal stress during IHT. Furthermore, ithas been previously reported that the interfacial stress also

FIG. 1. A. Reversed-phase HPLC chromatograms of whey protein standards: (I) bovine serum albumin, (II) a-lactalbumin, (III) b-lactoglobulin B,

and (IV) b-lactoglobulin A; B. Diminished peak heights with increased retention time of reversed-phase HPLC column for the 65�C convective-dried

sample (color figure available online).

WHEY PROTEIN ISOLATE 1537

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

15:

38 1

7 Ju

ly 2

014

irreversibly denatures proteins at the air–water interfaceduring the drying process.[42,43] Hence, it is comprehensiblethat the dehydration and interfacial stresses acting togetherare responsible for causing a higher degree of protein dena-turation during CD even at relatively lower air tempera-tures such as 65�C.

At 80�C both the SDD and the IHT processes resultedin substantial denaturation of WPI. Interestingly, at thistemperature the IHT process caused a much greater extentof WPI denaturation (90.00%) than CD (68.73%) at theend of 600 s. The IHT process caused 21% more denatura-tion of WPI than the CD process did. This relatively lowerdenaturation of WPI during CD (at 80�C) can be attribu-ted to two important factors. Firstly, the droplet tempera-ture does not rise as quickly in convective SDD as thesolution temperature does in IHT (temperature profiles in

Figs. 2A and 2B). In the CD process, the droplet tempera-ture remains well below the air temperature for a consider-ably longer time due to the evaporative cooling. Thedroplet temperature remains close to the wet bulb tempera-ture as long as sufficient water is available for evaporation.When less water is available for evaporation, the droplettemperature starts to rise toward the air temperature. Itcan be seen from Fig. 3 (moisture content profiles) thatthe moisture evaporation process stagnated after 400 and360 s of drying at 65 and 80�C, respectively. When the lossof water in the drying process became negligible, especiallyat low residual moisture contents, further denaturation ofWPI also became negligible. It is generally accepted thatthe presence of water is essential for protein denaturationreactions to occur. It appears that the higher solid con-centration in the particle is capable of providing better

FIG. 2. WPI denaturation profile during IHT and SDD at (A) 65�C and (B) 80�C medium temperature.

1538 HAQUE ET AL.

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

15:

38 1

7 Ju

ly 2

014

protection against the denaturation due to lower molecularmobility. It also appears that the moist heat, above thedenaturation temperature of WPI (�65�C),[17,41] denaturesa much greater amount of WPI at a much faster rate thanthe dry heat supplied in the droplet drying process.

It can also be seen from Fig. 3 (denaturation profiles)that after carrying out CD for 400 s or longer the amountsof irreversibly denatured protein at both 65 and 80�C weresimilar (p> 0.05). At times longer than 400 s the increase inthe drying air temperature did not cause further denatura-tion of WPI, perhaps due to the higher solid concentrationand unavailability of water to cause structural changes.After 400 s of drying, the drying rates were very low. Forexample, the drying rates were 0.0035 and 0.0015 kg=kg=sat 65 and 80�C, respectively (Fig. 4) and these low dryingrates meant that further denaturation did not occur.

Prediction of Drying Kinetics by REA

For the application of REA, the relative activationenergy, which is the fingerprint of REA, needs to be deter-mined as a function of moisture content. Using experi-mental data on moisture content, temperature, andsurface area of the droplet=particle during drying, the acti-vation energy was evaluated using Eq. (6). The activationenergy was generated from the CD run of WPI carriedout at 80�C. The activation energy was divided by the equi-librium activation energy as shown in Eq. (8) to calculatethe relative activation energy as shown in Eq. (7). Therelative activation energy was then correlated with thedifference between average moisture (X) content and equi-librium moisture content (Xb). A least-squares algorithmimplemented in MS Excel was used to calculate therelative activation energy as a function of moisture content

FIG. 3. Experimental and predicted moisture content (MC) and denaturation profiles of WPI droplets during drying at 65 and 80�C.

FIG. 4. Drying rate profiles as a function of water content for drying of WPI droplets at 65 and 80�C air temperature.

WHEY PROTEIN ISOLATE 1539

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

15:

38 1

7 Ju

ly 2

014

difference (X �Xb). The relative activation energy of WPIwas found to be represented by Eq. (15):

DEv

DEv;b¼ 1þ 306:803ðX � XbÞ1:9 exp½�5:87ðX � XbÞ0:404�

ð15Þ

It can be seen from Fig. 5 that Eq. (15) provides a goodfit to the experimental relative activation energy data with�7% error (R2¼ 0.996). The profile of relative activationenergy shown in Fig. 5 also follows the trends presentedin earlier publications.[28,31] The relative activation energywas zero at high moisture contents (in the early stage ofdrying), which indicates that the water was available read-ily for evaporation. The value of the relative activationenergy continued to increase as drying progressed, whichindicates that the removal of water became progressively

difficult as drying progressed. When equilibrium moisturecontent was attained, the relative activation energy was 1.

In order to predict the moisture content and tempera-ture profiles during drying, the mass and heat balanceequations represented by Eqs. (5) and (9), respectively,were simultaneously solved. The equilibrium and relativeactivation energy shown in Eqs. (8) and (15), respectively,were used for this purpose. The results of model predic-tions were then compared with the experimental moistureand droplet temperature history data.

The experimental moisture history and the REA-predicted moisture history are presented in Fig. 3. Simi-larly, the experimental temperature history and the REA-predicted temperature histories are presented in Fig. 6. Itcan be seen from these two figures that the model predic-tions followed the experimental moisture content–timeand temperature–time data reasonably well. The average

FIG. 5. Relative activation energy curves for drying of WPI droplets at 65 and 80�C drying temperature.

FIG. 6. Experimental and predicted temperature and denaturation profiles of WPI droplets during drying at 65 and 80�C.

1540 HAQUE ET AL.

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

15:

38 1

7 Ju

ly 2

014

errors between the experimental and predicted data forthe moisture content–time profile and the temperature–time profile were �6.38% (R2¼ 0.995) and �3.48%(R2¼ 0.981), respectively, at a drying temperature of65�C. Similarly, the average errors between the experi-mental and predicted moisture content–time and tempera-ture–time profiles were �6.65% (R2¼ 0.996) and �2.51%(R2¼ 0.986), respectively, at 80�C. These results indicatethat the REA-based temperature and moisture kineticsmodels are capable of predicting the moisture and tempera-ture histories of WPI droplets at both 65 and 80�C reason-ably well. In both drying temperatures, a higher degree oferror occurred in temperature–time profiles when theexperimental droplet temperatures rose very rapidlytoward the air temperature. The errors in prediction ofmoisture content–time data can be partly attributed tothe effect of drag on mass measurements, although weendeavored to minimize the effect of drag on the mass

measurements (see Methods section on SDD). The abovecomparisons suggest that the REA model can be confi-dently used to predict the drying kinetics of WPI in a CDenvironment.

Prediction of Denaturation Kinetics

The experimental denaturation kinetics of WPI at bothdrying air temperatures (65 and 80�C) are presented inFigs. 3 and 6. The same experimental denaturation–timeprofile data are presented in these two figures to illustratethe effect of moisture evaporation and droplet temperatureon the denaturation kinetics. The predicted lines passingthough the experimental denaturation data were generatedusing the first-order reaction kinetics equation (Eq. (13))and Eq. (14), which provide the temperature and moisturecontent dependence of the denaturation constant. Theparameters used for these predictions are presented inTable 2. It can be seen from Figs. 3 and 6 that the predicted

TABLE 2Kinetic parameters in prediction of denaturation of WPI during droplet drying

Drying temperature (�C) Ko (s�1) a b Ea (kJ �mol�1) R2 % Error

65 59,897 119 328,715 50.12 0.95 6.9580 59,704 114 328,673 54.2 0.985 5.98

FIG. 7. Morphology of 5� 0.1mL WPI during CD at 65�C and successive dissolution at room temperature (20�C) (color figure available online).

WHEY PROTEIN ISOLATE 1541

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

15:

38 1

7 Ju

ly 2

014

denaturation kinetics data followed the experimental dena-turation kinetics data reasonably well. The coefficients ofdetermination (R2) for these predictions were 0.95 and0.985 at 65 and 80�C, respectively. The prediction wasweaker in the early stage of drying as reversible denatura-tion occurred; for example, before 240 s (Fig. 2A). Thisreversible denaturation gave rise to the fluctuation in thehydrophobicity in determination of denaturation usingRP-HPLC. This reversible denaturation occurred primarilydue to evaporation stress. The activation energies requiredfor the denaturation of WPI were found to be 52�2 kJ �mol�1 and these values were not statistically differentat 65 and 85�C (p> 0.05). The magnitude of this denatura-tion energy was comparable to the inactivation energy ofprobiotic bacteria.[34] Using the same kinetics models,they[34] found that the Ea values for bacterial inactivationwere 39� 1 kJ �mol�1 during the drying of probiotic bac-teria in skim milk droplets. The closeness of the denatura-tion energy of WPI and inactivation energy of probioticbacteria suggests that the death of probiotics perhapsoccurred due to denaturation of their protein componentwithin the cell.

Morphology of WPI Droplets During Convective Dryingand Dissolution

The changes in the morphology of WPI solutiondroplets obtained while drying at 65�C and the subsequentdissolution process (in water at 20�C) are presented inFig. 7. In dissolution experiments 5� 0.1 mL of Milli-Qwater was syringed into the particle at the end of the CDprocess. The dissolution process was carried out for 720 sin still air at ambient temperature. The morphologicalchanges were captured and recorded by the same imageacquisition system used in droplet drying experiments. Itcan be seen from Fig. 7 that at the end of the 720-s dissol-ution process some small part of the protein remainedsettled (undissolved) at the bottom of the droplet. Thisinsoluble mass might have been due to the irreversibledenaturation of WPI.

CONCLUSIONS

The drying and denaturation kinetics of WPI dropletswere measured in the CD process and compared with thoseobtained from the IHT process. The CD and the IHT pro-cesses were carried out at 65 and 80�C for 600 s. TheREA-based models were applied to predict both the experi-mental moisture content and temperature histories. Thedenaturation kinetics was also predicted considering thedenaturation as a first-order reaction. The results showedthat there was no denaturation of WPI during IHT at65�C medium temperature up to 350 s, whereas 44% WPIwas denatured irreversibly by this time in the case of CD.At the end of treatment (600 s) at 65�C, the denatura-tion was 68.31 and 10.79% in CD and IHT processes,

respectively. The stress due to dehydration and theexposure time were found to be responsible the denatura-tion of WPI at 65�C during CD, whereas only exposuretime was found to be responsible for its denaturation inthe IHT process. At 80�C, the denaturation of WPI wasmuch higher in the IHT process (90%) than in the CD pro-cess (68.73%). At this temperature, both the thermal (moistheat) and dehydration stresses were found to be responsiblefor denaturation of WPI in the CD process, whereas thevery high thermal stress was found to be responsible fordenaturation of WPI in the IHT process. REA-basedmathematical models were able to predict the moistureand temperature histories of the WPI droplets reasonablywell in the CD process. The average error between theexperimental and predicted moisture histories was 6.5%,whereas the average error between the experimental andpredicted temperature histories was 3.0%. The predicteddenaturation kinetics followed the experimental denatura-tion kinetics data within 6.5� 0.5% error.

ACKNOWLEDGMENTS

The first author thanks the Australian government forproviding him with an Endeavour Postgraduate Scholar-ship. The authors thank Bruce Armstrong for technical helpduring the experiments. The Fonterra Cooperative (NewZealand) is acknowledged for the donation of whey proteinisolate (WPI). This work was also partially supported by theAustralian government’s Collaborative Research Network(CRN) initiative.

REFERENCES

1. Sane, S.U.; Wong, R.; Hsu, C.C. Raman spectroscopic characteriza-

tion of drying-induced structural changes in a therapeutic antibody:

Correlating structural changes with long-term stability. Journal of

Pharmaceutical Sciences 2004, 93, 1005–1018.

2. Johnson, K.A. Preparation of peptide and protein powders for

inhalation. Advanced Drug Delivery Reviews 1997, 26, 3–15.

3. Jayasundera, M.; Adhikari, B.; Adhikari, R.; Aldred, P. The effect of

food-grade low-molecular-weight surfactants and sodium caseinate

on spray drying of sugar-rich foods. Food Biophysics 2010, 5(2),

128–137.

4. Abdul-Fattah, A.M.; Kalonia, D.S.; Pikal, M.J. The challenge of

drying method selection for protein pharmaceuticals: Product

quality implications. Journal of Pharmaceutical Sciences 2007, 96(8),

1886–1915.

5. Schmitz, I.; Gianfrancesco, A.; Kulozik, U.; Foerst, P. Kinetics of

lysine loss in an infant formula model system at conditions applicable

to spray drying. Drying Technology 2011, 29, 1876–1883.

6. Adler, M.; Unger, M.; Lee, G. Surface composition of spray-dried par-

ticles of bovine serum albumin=trehalose=surfactant. Pharmaceutical

Research 2000, 17, 863–870.

7. Oldfield, D.J.; Taylor, M.W.; Singh, H. Effect of preheating and

other process parameters on whey proteins reactions during skim

milk powder manufacture. International Dairy Journal 2005, 15,

501–511.

8. Ameri, M.; Maa, Y.-F. Spray drying of biopharmaceuticals:

Stability and process considerations. Drying Technology 2006, 24,

763–768.

1542 HAQUE ET AL.

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

15:

38 1

7 Ju

ly 2

014

9. Tzannis, S.T.; Prestrelski, S.J. Activity–stability considerations of

trypsinogen during spray drying: Effects of sucrose. Journal of

Pharmaceutical Sciences 1999, 88(3), 351–358.

10. Dobson, C.M. The structural basis of protein folding and its links

with human disease. Philosophical Transactions of the Royal Society

of London 2001, 356, 133–145.

11. Horwich, A. Protein aggregation in disease: A role for folding inter-

mediates forming specific multimeric interactions. The Journal of

Clinical Investigation 2002, 110, 1221–1232.

12. Thomas, P.J.; Qu, B.H.; Pedersen, P.L. Defective protein folding as

a basis of human disease. Trends in Biochemical Sciences 1995, 20,

456–459.

13. Anandharamakrishnan, C.; Riellya, C.D.; Stapleya, A.G.F. Effects of

process variables on the denaturation of whey proteins during spray

drying. Drying Technology 2007, 25(5), 799–807.

14. Anandharamakrishnan, C.; Raghavendra, S.N.; Barhate, R.S.;

Hanumesh, U.; Raghavarao, K.S.M.S. Aqueous two-phase extraction

for recovery of proteins from cheese whey. Food and Bioproducts

Processing 2005, 83(C3), 191–197.

15. Singh, H.; Creamer, L.K. Denaturation, aggregation and heat stab-

ility of milk protein during the manufacture of skim milk powder.

Journal of Dairy Research 1991, 58(3), 269–283.

16. Ferreira, I.M.P.L.V.O.; Cacote, H. Detection and quantification of

bovine, ovine and caprine milk percentages in protected denomination

of origin cheeses by reversed-phase high-performance liquid chroma-

tography of beta-lactoglobulins. Journal of Chromatography A 2003,

1015, 111–118.

17. Parris, N.; Baginski, M.A. A rapid method for the determination of

whey protein denaturation. Journal of Dairy Science 1991, 74, 58–64.

18. Chen, X.D.; Xie, G.Z. Fingerprints of the drying behaviour of

particulate or thin layer food materials established using a reaction

engineering model. Transactions of the Institute of Chemical

Engineers – Part C: Food and Bio-Product Processing 1997, 75,

213–222.

19. Chen, X.D. The basics of a reaction engineering approach to

modeling air-drying of small droplets or thin-layer materials. Drying

Technology 2008, 26, 627–639.

20. Patel, K.C.; Chen, X.D. Drying of aqueous lactose solutions in a

single stream dryer. Food and Bioproducts Processing 2008, 86,

185–197.

21. Rogers, S.; Wu, W.D.; Lin, S.X.Q.; Chen, X.D. Particle shrinkage and

morphology of milk powder made with a monodisperse spray dryer.

Biochemical Engineering Journal 2012, 62, 92–100.

22. Mezhericher, M.; Levy, A.; Borde, I. Theoretical drying model of

single droplets containing insoluble or dissolved solids. Drying

Technology 2007, 25, 1025–1032.

23. Fu, N.; Woo, M.W.; Lin, S.X.Q.; Zhou, Z.; Chen, X.D. Reaction

engineering approach (REA) to model the drying kinetics of droplets

with different initial sizes—Experiments and analyses. Chemical

Engineering Science 2011, 66, 1738–1747.

24. Putranto, A.; Chen, X.D.; Xiao, Z.; Webley, P.A. Simple,

accurate and robust modelling of various systems of drying

foods and biomaterials: A demonstration of the feasibility of

the reaction engineering approach (REA). Drying Technology

2011, 29, 1519–1528.

25. Sano, Y.; Keey, R.B. The drying of a spherical particle containing

colloidal material into a hollow sphere. Chemical Engineering Science

1982, 37(6), 881–889.

26. Che, L.; Li, D.; Chen.X.D. Convective drying kinetics of single dro-

plets of aqueous glucose. Drying Technology 2012, 30, 1029–1036.

27. Ghandi, A.; Powell, I.; Chen, X.D.; Adhikari, B. Drying kinetics and

survival studies of dairy fermentation bacteria in convective air drying

environment using single droplet drying. Journal of Food Engineering

2012, 110, 405–417.

28. Lin, S.X.Q.; Chen, X.D. The reaction engineering approach to

modelling the cream and whey protein concentrate droplet drying.

Chemical Engineering and Processing 2007, 46, 437–443.

29. Adhikari, B.; Howes, T.; Bhandari, B.R. Use of solute fixed coordi-

nate system and method of lines for prediction of drying kinetics

and surface stickiness of single droplet during convective drying.

Chemical Engineering and Processing 2007, 46, 405–419.

30. Anandharamakrishnan, C.; Rielly, C.D.; Stapley, A.G.F. Loss of

solubility of a-lactalbumin and b-lactoglobulin during the spray dry-

ing of whey proteins. Swiss Society of Food Science and Technology

2008, 41, 270–277.

31. Lin, S.X.Q.; Chen, X.D. A model for drying of an aqueous lactose

droplet using the reaction engineering approach. Drying Technology

2006, 24, 1329–1334.

32. Ranz, W.E.; Marshall, W.R. Evaporation from drops. Chemical

Engineering and Processing 1952, 48(3), 141–146.

33. Lin, S.X.Q.; Chen, X.D. Changes in milk droplet diameter during dry-

ing under constant drying conditions investigated using the glass-

filament method. Transactions of the Institute of Chemical Engineers

Part C 2003, 82(C3), 213–218.

34. Li, X.; Lin, S.X.Q.; Chen, X.D.; Chen, L.; Pearce, D. Inactivation

kinetics of probiotic bacteria during the drying of single milk droplets.

Drying Technology 2006, 24, 695–701.

35. Chen, X.D.; Patel, K.C. Micro-organism inactivation during drying of

small droplets or thin-layer slabs—A critical review of existing kinetics

models and an appraisal of the drying rate dependent model. Journal

of Food Engineering 2007, 82, 1–10.

36. Meerdink, G.; Van’tRiet, K. Prediction of product quality during

spray drying. Transactions of the Institution of Chemical Engineers –

Part C: Food and Bioproducts Processing 1995, 73, 165–170.

37. Oldfield, D.J.; Singh, H.; Taylor, M.W.; Pearce, K.N. Kinetics of

denaturation and aggregation of whey protein in skim milk heated

in an ultra-high temperature (UHT) pilot plant. International Dairy

Journal 1998, 8, 311–318.

38. Levieux, D.; Levieux, A.; El-Hatmi, H.; Rigaudiere, J.-P. Immuno-

chemical quantification of heat denaturation of camel (Camelus

dromedarius) whey proteins. Journal of Dairy Research 2006, 73,

1–9.

39. Hinrichs, J.; Rademacher, B. High pressure thermal denaturation

kinetics of whey proteins. Journal of Dairy Research 2004, 71,

480–488.

40. Hinrichs, J.; Rademacher, B. Kinetics of combined thermal and

pressure-induced whey protein denaturation in bovine skim milk.

International Dairy Journal 2005, 15, 315–323.

41. Ferreira, I.M.P.L.V.O.; Mendes, E.; Ferreira, M.A. HPLC=UV

analysis of proteins in dairy products using a hydrophobic

interaction chromatographic column. Analytical Sciences 2001, 17,

499–501.

42. Maa, Y.-F.; Hsu, C.C. Protein denaturation by combined effect of

shear and air–liquid interface. Biotechnology and Bioengineering

1997, 54, 503–512.

43. Maa, Y.-F.; Nguyen, P.T.; Hsu, S.W. Spray-drying of air–liquid

interface sensitive recombinant human growth hormone. Journal of

Pharmaceutical Sciences 1998, 87(2), 152–159.

44. Foster, K.D.; Bronlund, J.E.; Paterson, A.H.J. The prediction of

moisture sorption isotherms for dairy powders. International Dairy

Journal 2005, 15, 411–418.

45. Lin, S.X.Q.; Chen, X.D.; Pearce, D. Desorption isotherm of milk

powders at elevated temperatures and over a wide range of relative

humidity. Journal of Food Engineering 2005, 68(2), 257–264.

46. Choi, Y.; Okos, M.R. Effects of temperature and composition on

the thermal properties of foods. In Food Engineering and Process

Applications: Vol. 1. Transport Phenomenon; Maguer, L.; Jelen, P.,

Eds.; Elsevier: New York, 1986; 93–101.

WHEY PROTEIN ISOLATE 1543

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

15:

38 1

7 Ju

ly 2

014

APPENDIX: PHYSIOCHEMICAL PROPERTIES OF WPISOLUTION USED IN THE DRYING KINETICSMODELING

The equilibrium moisture contents (Xb) of the WPIpowder were calculated from the sorption isotherm usingthe Guggenheim-Anderson-deBoer (GAB) isotherm asgiven by Eq. (A1):

Xb ¼mockaw

1� kawð Þ 1� kaw þ ckawð Þ ; ðA1Þ

where mo is the monolayer moisture content (kg=kg solid),aw is the water activity (dimensionless), and c and k areGAB constants. The GAB model parameters for WPI wereobtained from Foster et al.[44] The temperature dependenceof c and k was estimated using Eqs. (A2) and (A3),respectively[45]:

c ¼ co expDH1

RT

� �ðA2Þ

k ¼ ko expDH2

RT

� �; ðA3Þ

where T is the drying air temperature (K), R is the gas con-stant (J �mol�1 �K�1), and DH1 and DH2 are the heat ofsorption of water in a droplet=particle. The GAB constantsfor WPI at the two drying temperatures were extrapolatedusing Eqs. (A2) and (A3). The values for c and k were esti-mated to be 8.43 and 0.784, respectively, for 65�C and 8.39and 0.782, respectively, for 80�C, which yielded Xb valuesof 0.013 (kg=kg solid) and 0.01 (kg=kg solid) for 65 and80�C, respectively.

The temperature dependence of density (qp) and specificheat capacity (Cp) of WPI were estimated using the Choiand Okos models as given by Eqs. (A4) and (A5)[46]:

qp ¼ 1329:9� 0:5184 T � 273:15ð Þ ðA4Þ

Cp ¼ 2008:2þ1:2089 T�273:15ð Þ�0:0013129 T�273:15ð Þ2

ðA5Þ

1544 HAQUE ET AL.

Dow

nloa

ded

by [

RM

IT U

nive

rsity

] at

15:

38 1

7 Ju

ly 2

014