Improving acetic acid production of acetobacter pasteurianus AC2005 in hawthorn vinegar fermentation by using beer for seed culture

Some studies Oliveira et al., 2001a,b; Ramı´rez-Ca´rdenas et al., 2008 found a greater reduction in the content of tannins, phytates and oligosaccharides in beans that were soaked and co

Review article

Influence of soaking on the nutritional quality of common beans

(Phaseolus vulgaris L.) cooked with or without the soaking water:

a review

Ana Carolina Fernandes, Waleska Nishida & Rossana P da Costa Proenc¸a*

Graduate Program in Nutrition, Nutrition in Foodservices Research Group (NUPPRE), Department of Nutrition, Federal University of Santa Catarina (UFSC), Campus Trindade, CEP 88.040-970, Floriano´polis, SC, Brazil

(Received 12 April 2010; Accepted in revised form 26 July 2010)

the need to discard the soaking water before cooking Thus, the present study proposes to review theinfluence of maceration on the nutritional quality of common beans (Phaseolus vulgaris L.) cooked with orwithout the soaking water, in an attempt to achieve agreement among scientists The article search was done

in a systematic way and eleven studies were found Of these, three compared the use or not of the soakingwater for cooking, seven of them discarded the soaking water and one used the soaking water This reviewdiscusses each nutrient and antinutrient regarding the effects of soaking and compares them with otherstudies done with legumes The results were not unanimous but there was a greater advantage to discardingthe soaking water before cooking

Keywords Antinutrients, bioavailability, cooking, dry beans, food quality, nutritional aspects, processing effects.

Introduction

The common bean (Phaseolus vulgaris L.) is consumed

worldwide, especially in Latin America and Africa

(FAO, 2009) Because of its cultural and nutritional

importance, the Brazilian food pyramid shows beans in

a group of their own (Philippi et al., 1999) and the Food

Guide for the Brazilian Population recommends the

consumption of at least one portion of beans per day

(Vasconcellos et al., 2006)

However, beans contain compounds that can

nega-tively affect their nutritional value, such as trypsin

inhibitors, lectins, phytates, polyphenols (especially

tannins in beans) and oligosaccharides (raffinose and

stachyose) Some of these are thermolabile, disappearing

after proper cooking, such as trypsin inhibitors and

lectins Others are thermostable, but their

concentra-tions are reduced by dissolution in water (Haro, 1983;

Silva & Silva, 1999, 2000)

Soaking the beans in water and discarding the water

may eliminate a percentage of these compounds Some

studies (Oliveira et al., 2001a,b; Ramı´rez-Ca´rdenas

et al., 2008) found a greater reduction in the content

of tannins, phytates and oligosaccharides in beans that

were soaked and cooked without the soaking water

However, Ramı´rez-Ca´rdenas et al (2008) pointed outsome studies that state that low concentrations ofphytates and phenolic compounds can be protectiveagainst cancer and cardiovascular diseases Meanwhile,oligosaccharide fermentation may have positive resultssuch as production of short-chain fatty acids anddecrease in intestinal pH (Muzquiz, 2008; Campos-Vega

et al., 2009)

The positive or negative effects of these compoundsseem to be more closely associated with their concen-tration in the beans, which varies according to type ofbean, as well as their interaction with other components

of the diet (Muzquiz, 2008; Ramı´rez-Ca´rdenas et al.,2008)

Bean soaking before cooking seems to be mously recommended by scientists; however, there is noconsensus regarding the discarding of the soaking water.Although many authors recommend the soaking water

unani-to be discarded unani-to eliminate antinutritional facunani-tors,others seek to prove the beneficial effects of thesefactors, which have been associated with the prevention

of diseases In this sense, it would be advantageous not

to discard the soaking water Yet, the published studiespresent contradicting and inconclusive results, which,according to Muzquiz (2008), can be attributed to theuse of different methodologies and parameters There-fore, a consensus regarding the fate of the soaking water

is yet to be achieved

*Correspondent: E-mail: rossana@mbox1.ufsc.br

doi:10.1111/j.1365-2621.2010.02395.x

The objective of the present study is to perform a

systematic review of the influence of soaking on the

nutritional quality of common beans (P vulgaris L.)

cooked with or without the soaking water, to assess and

compare the preparation methods and results and search

for concordant recommendations among the studies

Method

A systematic search of articles that discuss the influence

of soaking on the nutritional quality of common beans

(P vulgaris L.) cooked with or without the soaking

water, published between January 2004 and March 2009

was done The following databases were searched: Scielo

(Scientific Electronic Library Online), Lilacs (Latin

American and Caribbean Centre on Health Sciences

Information) and Scopus – which includes 100% of the

publications of the Medline (National Library of

Med-icine) database The keywords used for the search are

listed in Table S1 The search was done separately for

each language, using the keywords of the first line in

combination with the keywords of the lower lines

A total of twenty-two articles were found in Scielo,

nine in Lilacs and 392 in Scopus Nineteen repeated

texts were removed, totalling 404 studies

Based on the systematic search model, inclusion and

exclusion criteria were established to meet the objectives

of the research The inclusion criteria of the articles were

(i) original articles; (ii) articles in Portuguese, English or

Spanish; (iii) studies with the common bean (P vulgaris

L.); (iv) studies that analysed the effects of soaking the

beans on its composition, digestibility or bioavailability

in vitroor in vivo The exclusion criteria were (i) review

articles; (ii) articles in languages other than the ones

mentioned earlier; (iii) studies with coffee (called coffee

beans in English); (iv) studies with only other types of

legumes or with beans of different species; (v) studies

that analysed the effect of soaking on the properties of

the seeds, the bean plant or bean characteristics other

than the nutritional and sensorial characteristics; (vi)

studies that analysed the effects of soaking beans for

preparations other than the traditional preparations

(such as bean sweets, flours for supplements, animal

feeds); (vii) studies that only compared bean varieties or

cultivars, or compared different legumes, and did not

compare different processing methods; (viii) studies that

only covered the influence of different processing

meth-ods on the quality of the bean; (ix) articles that were not

complete, even when they were ordered from the

authors

After the abstracts of all the articles were read, those

that did not meet the inclusion criteria were excluded

Only eleven studies were specifically about the influence

of soaking on the nutritional quality of common beans

(P vulgaris L.) cooked with or without the soaking

water

The studies were analysed according to their year ofpublication, country of origin, objectives, variables,preparation methods and analyses, results and conclu-sions and⁄ or recommendations of the authors

Characteristics of the analysed articles

The characteristics verified in the selected articles aredescribed in Table S2 Most of these studies (27.3%)were done in Brazil (Oliveira et al., 2008; Ramı´rez-Ca´rdenas et al., 2008; Toledo & Canniatti-Brazaca,2008), followed by the United States of America(18.2%) (Luthria & Pastor-Corrales, 2006; Xu & Chang,2008); then came Mexico (Carmona-Garcı´a et al., 2007);Spain (Pujola` et al., 2007); Turkey (Nergiz & Go¨kgo¨z,2007); Ethiopia (Shimelis & Rakshit, 2007); Sudan(Elmaki et al., 2007); and Pakistan (Rehman & Shah,2004), each with 9.1%

Regarding the objectives and variables, three studies(Oliveira et al., 2008; Ramı´rez-Ca´rdenas et al., 2008;Toledo & Canniatti-Brazaca, 2008) assessed the effects

of cooking the beans with or without the soaking water.The other studies analysed the effects of different beanprocessing methods (raw, soaked, soaked and cooked,cooked without soaking), but did not discuss the use ofthe soaking water for cooking Of these eight studies,only one (Nergiz & Go¨kgo¨z, 2007) used the soakingwater to cook the beans, while the other seven studies(Rehman & Shah, 2004; Luthria & Pastor-Corrales,2006; Carmona-Garcı´a et al., 2007; Elmaki et al., 2007;Pujola` et al., 2007; Shimelis & Rakshit, 2007; Xu &Chang, 2008) discarded the soaking water

All articles analysed more than one variable Thus,the studies also assessed the effects of different cookingmethods (vapour, boiling, pressure cooking, microwave(Toledo & Canniatti-Brazaca, 2008; Xu & Chang, 2008;Shimelis & Rakshit, 2007; Nergiz & Go¨kgo¨z, 2007;Rehman & Shah, 2004); of different soaking solutionsother than pure water – sodium chloride (NaCl), sodiumbicarbonate (NaHCO3) and mixed (NaCl + NaHCO3)(Rehman & Shah, 2004; Carmona-Garcı´a et al., 2007;Shimelis & Rakshit, 2007); of different types of beans,different varieties, colours and cultivars (Luthria &Pastor-Corrales, 2006; Elmaki et al., 2007; Pujola` et al.,2007; Oliveira et al., 2008; Ramı´rez-Ca´rdenas et al.,2008); of different soaking times (Elmaki et al., 2007; Xu

& Chang, 2008) and of germinating the beans in theprepreparation phase (Shimelis & Rakshit, 2007).All articles also had more than one outcome variable.The outcomes investigated most often were changes inphytate content (Elmaki et al., 2007; Nergiz & Go¨kgo¨z,2007; Shimelis & Rakshit, 2007; Ramı´rez-Ca´rdenas et al.,2008; Toledo & Canniatti-Brazaca, 2008), followed bytannins content (Nergiz & Go¨kgo¨z, 2007; Shimelis &Rakshit, 2007; Ramı´rez-Ca´rdenas et al., 2008; Toledo &Canniatti-Brazaca, 2008), phenol content (total, polyphe-

nols – which also include tannins and phenolic acids)

(Luthria & Pastor-Corrales, 2006; Elmaki et al., 2007;

Nergiz & Go¨kgo¨z, 2007; Xu & Chang, 2008) and mineral

content (Elmaki et al., 2007; Pujola` et al., 2007; Oliveira

et al., 2008; Ramı´rez-Ca´rdenas et al., 2008); and in vitro

protein digestibility (Nergiz & Go¨kgo¨z, 2007; Shimelis &

Rakshit, 2007; Toledo & Canniatti-Brazaca, 2008)

Other changes were also verified such as centesimal

composition (Ramı´rez-Ca´rdenas et al., 2008; Toledo &

Canniatti-Brazaca, 2008); starch (total, available starch

and resistant starch, amylose) (Carmona-Garcı´a et al.,

2007; Pujola` et al., 2007); fibres (Rehman & Shah, 2004;

Ramı´rez-Ca´rdenas et al., 2008); trypsin-inhibiting

activ-ity (Nergiz & Go¨kgo¨z, 2007; Shimelis & Rakshit, 2007);

oligosaccharides (Shimelis & Rakshit, 2007); in addition

to the capacity to extract minerals with HCl (Elmaki

et al., 2007), among others The outcome variables

associated with the nutrients and antinutrients are shown

separately in Tables 1 and 2, which also show the effects

of different bean preparation methods on these variables

Phytates and phytic acid

The authors of all studies that assessed phytates stated

that a reduction of these compounds is desirable The

greatest reduction of phytates and phytic acid was

achieved by soaking and cooking without the soaking

water (Elmaki et al., 2007; Nergiz & Go¨kgo¨z, 2007;

Ramı´rez-Ca´rdenas et al., 2008; Toledo &

Canniatti-Brazaca, 2008) Toledo & Canniatti-Brazaca (2008)

stated that phytate reduction was equal in samples with

and without soaking, however, as shown in a table of

their study, the phytate content varied according to

cooking method On average, the greatest phytate

content was found in beans that were cooked with the

soaking water, followed by beans cooked without

soaking and finally beans cooked without the soaking

water Among soaked beans and for all cooking

methods, beans cooked without the soaking water

always had statistically lower phytate content than

those cooked with the soaking water Similar results

were found by Oliveira et al (2001b) in an older study

with common beans, and by Boateng et al (2007), who

studied the phytate content in another species of bean

However, phytic acid reduction may not be needed for

the utilisation of some nutrients A study done by

Oliveira et al (2003) showed that phytic acid in

concentrations as high as eight times of that found in

raw common bean did not compromise the utilisation of

casein by rats during a 10-day period

Studies found that soaking and cooking had different

effects on different legumes For example, Aranda et al

(2004) concluded that high consumption of phytate

from beans (Vicia faba L.) had no negative effects on the

digestion of calcium (Ca) and magnesium (Mg) by rats

However, through another mechanism, soaking and

cooking increased the metabolic utilisation of Ca and

Mg Meanwhile, Chopra & Sankhala (2004) found asignificant association between soaking and reducedphytate contents, concomitant with increased ironbioavailability in horse gram (Dolichos biflorus) andmoth bean (Phaseolus aconitifolius)

The reduction of phytates and phytic acid (phytatesalt) may not be necessary to improve the utilisation ofall nutrients However, their presence may impair theutilisation of some micronutrients, thus their reduction

is desirable In this sense, soaking, especially if thesoaking water is discarded, can be recommended, as itproved to be an effective way to reduce phytates andphytic acid

Total phenolic compounds

In all the studies that assessed total phenolic compounds(Luthria & Pastor-Corrales, 2006; Elmaki et al., 2007;Nergiz & Go¨kgo¨z, 2007; Toledo & Canniatti-Brazaca,2008; Xu & Chang, 2008), the loss of these compoundswas greater in soaked beans cooked without the soakingwater and proportional to the length of soaking Asimilar reduction was obtained for velvet beans (Mucunapruriens) by Vadivel & Pugalenthi (2008, 2009), bysoaking and discarding the water not absorbed by thebeans, followed by autoclaving

However, in the study by Luthria & Pastor-Corrales(2006), only 2% of the total phenolic compounds are lost

in the soaking water, while 83% remain in the beans and15% are probably lost during cooking The effect ofsoaking on the total amount of phenolic compounds wasalso discussed by Anton et al (2008) and Boateng et al.(2007) In the first study, there were no significantchanges in the total content of phenolic compounds insoaked but uncooked navy and pinto beans In the study

by Boateng et al (2007), there was a significant reductiononly in the total content of phenolic compounds of pintobeans The same was not observed for kidney beans aftersoaking and discarding the soaking water, without cook.There is no consensus regarding the reduction of totalphenolic compounds in beans when the inherent reduc-tion of their antioxidant activity is assessed Ranilla

et al.(2009) found a relationship between the reduction

of phenolic compounds and reduced antioxidant activity

in soaked beans; however, the greatest loss was found insamples where the cooking water was discarded, whichmay indicate that great loss is because of cooking andmay be avoided by consuming the beans with thecooking water

Xu & Chang (2009) also found a relationship betweenthe content of total phenolic compounds and antioxi-dant activity of beans However, there was no associ-ation between total phenolic acids and antioxidantactivity in black beans, only in pinto beans The authorsconcluded that the greatest loss of phenolic compounds

and consequently, of the antioxidant activity of the

studied beans, is because of heat They also conclude

that these changes depend upon the type of beans and

processing conditions and that different phenolic

con-tents might contribute to different degrees to the overall

antioxidant activity

The protective effect of beans against certain chronic

diseases has been associated with the presence of

phenolic compounds (Boateng et al., 2007; Xu et al.,

2007) However, high levels may become undesirable

when they impair digestion and protein absorption,

inhibiting the activity of digestive enzymes such as

a-amylase and trypsin (Vadivel & Pugalenthi, 2008)

In this context, associating a partial reduction of total

phenolic compounds with better absorption of bean

proteins, soaking and discarding the water not absorbed

by the beans before cooking, seems to be more

appro-priate

Tannins

Tannins are the most studied phenolic compounds of

beans Usually the studies assess total phenolic

com-pounds or tannins Among the selected articles, the

reduction of tannins was considered desirable by all

authors that analysed their content In one of the

studies (Ramı´rez-Ca´rdenas et al., 2008), there was a

greater tannin reduction in beans that were soaked and

cooked without the soaking water While comparingsoaked beans cooked with the soaking water vs.unsoaked beans, Nergiz & Go¨kgo¨z (2007) found lowertannin content in soaked beans On the other hand,Toledo & Canniatti-Brazaca (2008) found the lowertannin contents in all unsoaked samples and allcooking methods The authors justify that the greaterloss is because of a longer cooking period, requiredwhen the beans are not previously soaked However,when the soaked beans are compared, the same studyshows lower values for soaked beans cooked withoutthe soaking water

In the study by Oliveira et al (2001b), for whomtannin reduction was desirable, a greater reduction intannin content was also obtained by discarding thesoaking water In other studies with rojo bean (Mosha &Vicent, 2004), horse gram and moth bean (Chopra &Sankhala, 2004), soaking reduced the tannin levelssignificantly However, such compounds did not affectthe bioavailability of zinc and iron in the study done byHemalatha et al (2007)

Tannins are also considered bioactive compoundsbecause of their antioxidant capacity (Xu et al., 2007;

Xu & Chang, 2009); however, they may have beneficial

or adverse nutritional effects (Xu et al., 2007)

In this sense, even though tannins do not alwaysinterfere with the utilisation of nutrients, their reductionwas considered desirable by all authors as they are

Table 2 Selected studies, preparation methods and results regarding the antinutrients, indicating the methods that resulted in the greatest

Toledo &

Canniatti-Brazaca, 2008

NS CWS COS

For author:

Soaked = NS table: COS

NS COS > CWS

Xu & Chang, 2008 COS

Soaking length (SL)

COS Longer SL Ramı´rez-Ca´rdenas

et al., 2008

RAW NS CWS COS

Greatest content: NS

Nergiz & Go¨kgo¨z, 2007 NS

CWS

Shimelis & Rakshit, 2007 COS (H 2 O; NaHCO 3 )

Germination (G) Cooking (CC, PC)

All fl

G > NaHCO 3 > H 2 O

PC > CC Pujola` et al., 2007 RAW

S COS

S

Elmaki et al., 2007 COS

Soaking length (SL)

COS Longer SL

COS Longer SL Luthria &

Pastor-Corrales, 2006

RAW COS

2% in the soaking water

NS, not soaked; CWS, cooked with soaking water; COS, cooked without the soaking water; H 2 O, water; NaHCO 3 , sodium bicarbonate solution; CC,

cooking in common pot or Mattson cooker; PC, cooking in pressure cooker or autoclave; S, only soaked.

primarily an antinutritional factor Soaking and

discarding the soaking water was the most effective way

to reduce tannins Thus, this procedure can be

recom-mended in the preparation of beans, also because soaking

does not completely eliminate tannins from beans, thus

the antioxidant activity attributed to this compound is

partially preserved Thus, part of the antioxidant

potential attributed to this compound can be preserved

Oligosaccharides

Only one of the selected studies assessed the

oligosac-charide content of beans and how it changed with

different preparation methods Shimelis & Rakshit

(2007) studied the reduction of raffinose, stachyose and

a-galactosides in two bean varieties (kidney bean) after

soaking in water or a solution of sodium bicarbonate

(NaHCO3) and cooking without the soaking water in a

pot or autoclave The authors consider that reducing

these oligosaccharides is desirable as they cause

flatu-lence

According to Shimelis & Rakshit (2007), both soaking

solutions reduced the raffinose, stachyose and

a-galac-toside contents in both bean varieties The germination

process, also investigated by the study, was the most

effective method to reduce these compounds However,

soaking also reduced their levels significantly, especially

when a NaHCO3solution was used Both soaking and

cooking independently reduced the levels of all

oligo-saccharides Consequently, when the two processes were

associated, there was a greater reduction of these sugars,

which was even more effective when the beans were

cooked in an autoclave (Shimelis & Rakshit, 2007)

According to Granito et al (2007) in their study with

Phaseolus lunatusbeans, 1⁄ 3 of the raffinose content and

1⁄ 5 of the stachyose content are lost in the soaking

water and the rest remained in the cooking water

In another two studies that aimed to reduce stachyose

and raffinose in bean-based processed products, soaking

and cooking were effective for some bean varieties

(Matella et al., 2005) or for some types of

oligosaccha-rides (Siddiq et al., 2006)

Matella et al (2005) found that soaking followed by

discarding the soaking water reduced the

oligosaccha-ride content of Michigan black beans and but did not

affect the oligosaccharide contents of red and navy

beans The analysis was done only in raw beans

On the other hand, Siddiq et al (2006) found a

significant reduction in the raffinose and stachyose

contents of red kidney beans after soaking and

discard-ing the soakdiscard-ing water Cookdiscard-ing further reduced the

raffinose content but did not affect the stachyose

content

Although there are differences in the effectiveness of

bean processing to reduce oligosaccharides, which

depend on the specific oligosaccharide or bean variety,

soaking with subsequent discarding of the soaking waterbefore cooking seems to reduce these compounds,something considered desirable in the reviewed studies

Proteins and protein digestibility

Protein content depended on preparation method andvaried from study to study Toledo & Canniatti-Brazaca(2008) did not observe differences in the protein content

of soaked beans cooked with or without the soakingwater Meanwhile, Ramı´rez-Ca´rdenas et al (2008) foundgreater absolute protein contents in beans cooked withthe soaking water; this difference was not confirmedstatistically Pujola` et al (2007) found a greater proteincontent in soaked beans cooked without the soakingwater than in raw beans or uncooked soaked beans.Other studies with legumes investigated if differentpreparation methods, such as soaking, extrusion andespecially thermal treatments, led to protein loss (Os-man, 2007; Teguia & Fon Fru, 2007; Huma et al., 2008).Rehman & Shah (2005) found that the protein content

of lentils, chick peas, red kidney beans, white kidneybeans and black grams (Vigna mungo) may not beaffected by soaking, discarding the soaking water andcooking

Toledo & Canniatti-Brazaca (2008) found that proteindigestibility was lowest in beans cooked without thesoaking water, but there was no difference betweenunsoaked cooked beans and beans cooked with thesoaking water On the other hand, Nergiz & Go¨kgo¨z(2007) found that protein digestibility was greater inbeans cooked with the soaking water than in unsoaked,cooked beans, but they did not investigate beans cookedwithout the soaking water

A common observation is that bean processingreduces protein content but increases protein digestibil-ity While studying chick peas, lentils and different types

of beans, Martı´n-Cabrejas et al (2009) and Rehman &Shah (2005) found that protein digestibility increasedafter soaking and cooking without the soaking water.Soaking and cooking may reduce the contents of someantinutrients as, according to Shimelis & Rakshit (2005),tannins, trypsin inhibitors and some oligosaccharidescorrelate with lower protein digestibility in the haricotbean (P vulgaris L.)

Although antinutritional factors are associated withlower protein digestibility, studies do not agree onwhether the soaking water should be discarded Prep-aration method does not seem to change protein contentand digestibility

Ashes, loss of solids, minerals and bioavailability

Beans cooked with the soaking water had the highestash contents (Ramı´rez-Ca´rdenas et al., 2008; Toledo &Canniatti-Brazaca, 2008); however, soaking caused a

greater loss of total solids, regardless of cooking with or

without the soaking water (Pujola` et al., 2007)

The lower ash content of soaked beans may be because

of not only mineral lixiviation but also antinutritional

factors Shimelis & Rakshit (2005) found a positive

correlation between ash content and zinc and phytic acid

contents in haricot beans (P vulgaris L.) Thus, a

reduction of the ash content may be desirable Cooking

also seems to reduce ash content (Osman, 2007)

Mineral content varied from study to study Oliveira

et al (2008) found that the mineral content of beans

cooked with or without the soaking water were equal;

Ramı´rez-Ca´rdenas et al (2008) found higher contents of

zinc and iron in unsoaked beans and calcium and copper

in beans cooked with the soaking water Elmaki et al

(2007) found that increasing the soaking length of beans

or discarding the soaking water resulted in greater loss

of minerals However, these treatments were also

associated with greater HCl-extractability Thus,

although minerals are lost in the soaking water, soaking

and discarding the soaking water increases the

bioavail-ability of the minerals that remained in the beans This is

probably caused by a reduction of antinutrients that

chelate minerals, as they are also reduced when beans

are soaked and the soaking water discarded

Studies with other types of beans and legumes also

found differing mineral contents Huma et al (2008)

found that soaking and cooking can reduce the amount of

minerals significantly Granito et al (2007) observed that

there was a greater loss of calcium, magnesium,

potas-sium, zinc and iron in cooked beans than in soaked beans

cooked without the soaking water However, minerals

lost during cooking lixiviate to the cooking water; (Huma

et al., 2008) consequently, bean preparations consumed

with the cooking water retain those minerals

Meanwhile, Chopra & Sankhala (2004) found that

soaking decreases the tannin and phytate contents of

horse gram (D biflorus) and moth bean (P

aconitifo-lius), but calcium and magnesium contents are not

reduced significantly by dissolution; the digestibility and

metabolism of both minerals also increased with

soak-ing Aranda et al (2004) also observed that soaking and

discarding the soaking water decreases tannin and

phytate contents, which improves iron bioavailability

Hence, studies with legumes in general and this review

are concordant regarding mineral bioavailability: it

increases with soaking, especially when the soaking

water is discarded and is associated with a reduction of

antinutritional factors

Carbohydrates

As observed with proteins, studies are not concordant in

relation to carbohydrate content Ramı´rez-Ca´rdenas

et al (2008) found a greater carbohydrate content in

unsoaked, cooked beans and lower content in beans

cooked with the soaking water; however, statisticalanalyses were not done In relation to starch fractions,Carmona-Garcı´a et al (2007) found greater proportions

of total starch and available starch in beans cookedwithout the soaking water, considering the averagefound for samples soaked in different solutions Therewere divergences regarding resistant starch: a sodiumchloride (NaCl) solution was more effective in reducingresistant starch than a sodium bicarbonate (NaHCO3)solution In both cases, the beans were cooked withoutthe soaking water The starch, amylopectin and resistantstarch contents of raw beans and the amylose content ofsoaked beans were higher than those of beans cookedwithout the soaking water (Pujola` et al., 2007) How-ever, these results are not relevant because beans are noteaten raw, or soaked without subsequent cooking

Different results were also obtained by other authorswhile studying the carbohydrate content of beans.Oliveira et al (2001b) found that cooking soakedcommon beans without the soaking water reduced thestarch content by 26.8% Salgado et al (2005) found agreater resistant starch content in macassar beans(Vigna unguiculata L Walp) when they were soakedbut cooked without the soaking water Kutosˇ et al.(2003) found that unsoaked, cooked beans and soakedbeans cooked without the soaking water had equalresistant starch contents The authors of the two studies(Kutosˇ et al., 2003; Salgado et al., 2005) did notinvestigate beans cooked with the soaking water

In agreement with Pujola` et al (2007), Oliveira et al.(2001b) found that soaking associated or not withcooking, slightly reduced the starch content of beans.Additionally, Apata (2008) states that cooking reducesthe carbohydrate content even if the beans are notpreviously soaked Other authors second the influence ofcooking on starch content and also mention otherfactors that influence starch content, such as postcoo-king handling, cooking method, bean variety, matura-tion stage of the seeds and length of time stocked frozen(Osorio-Dı´az et al., 2002 and Salgado et al., 2005)

Most authors agree that cooking without the soakingwater reduces the carbohydrate content of beans, butresistant starch content remains unchanged, which isdesirable as resistant starch resembles soluble fibre(Salgado et al., 2005) If only starch is taken intoaccount, it would not be recommended to cook soakedbeans without the soaking water However, when allcarbohydrates are considered, it may be advantageous

to discard the soaking water, as this reduces the contents

of undesirable sugars, such as sucrose and the charides that cause flatulence

without the soaking water However, when soluble and

insoluble fibre fractions are analysed separately, their

contents vary in beans cooked with or without the

soaking water (Ramı´rez-Ca´rdenas et al., 2008; Toledo &

Canniatti-Brazaca, 2008) Rehman & Shah (2004)

stud-ied soaked beans cooked without the soaking water and

found that the cellulose, hemicellulose and lignin

con-tents were higher when the beans were soaked in a

sodium bicarbonate solution (NaHCO3) and cooked in

a microwave oven or regular pot

According to Kutosˇ et al (2003), soaking and

cook-ing pinto beans increase soluble fibre content, but a

higher increase was found in unsoaked, cooked beans

On the other hand, processing decreased the insoluble

fibre content, which was less affected by cooking

without the soaking water than by cooking without

soaking Total fibre content decreased discretely and

was less affected by cooking without soaking, as found

by Vidal-Valverde et al (1998) in their study with faba

beans (V faba L major) For this reason, Kutosˇ et al

(2003) believe that it is better not to soak beans to

maintain total fibre content It is important to

empha-sise that resistant starch content, which resembles

soluble fibre, was similar between unsoaked, cooked

beans and cooked beans without the soaking water

(Kutosˇ et al., 2003)

Chopra et al (2009) studied five different types of

uncooked legumes where the soaking water was

dis-carded and found that all fibre fractions increased with

soaking, Thus, legume soaking is beneficial to health

because it increases the dietary fibre content, especially

soluble fibre content

Considering the findings on nutrients and

antinutri-ents covered in the studies, the different effects of

prepreparation and preparation are summarised in

Table S3

Conclusion

The articles reviewed in this paper are based on studies

that analyse the soaking of common beans (P vulgaris

L.) in water or other solutions (e.g sodium bicarbonate,

sodium chloride, acetic acid) to reduce the

antinutri-tional and flatulence factors, as well as to increase

nutrient availability They also investigated if the losses

were significant during the preparation processes The

results of these articles were systematically analysed by

comparing the statistically analysed data

Discarding the soaking water before cooking was

found to be advantageous This procedure seems to

reduce some carbohydrate fractions of beans and can

reduce, maintain or increase fibre content Meanwhile,

resistant starch content remains unchanged, whose

function is similar to that of soluble fibres This method

also reduced phytates, phytic acid, total phenolic

com-pounds and tannins Even though mineral content was

also reduced, the bioavailability of most studied als increased Furthermore, the different preparationmethods do not seem to affect the protein content anddigestibility of the studied beans

miner-Soaking before cooking and discarding the soakingwater also seems to be an effective way to reduce theamounts of oligosaccharides that cause flatulence This

is an important issue because an excess of theseoligosaccharides can lead an individual to avoid eatingbeans altogether, because of the intestinal discomfort

So, despite the fact that these compounds do presentsome functional properties, if beans are not consumed

to avoid intestinal discomfort, these compounds willalso not be consumed and their benefits will not beenjoyed

It should be emphasised that although thermalprocessing of beans is by far the factor that mostreduces antinutrient and nutrient contents, beans are noteaten raw, especially because they contain toxic sub-stances, so cooking is mandatory Finally, the contents

of the analysed compounds in beans can be affected bybean variety, crop location and stocking and distribu-tion methods As these factors will always be present, wesuggest that beans should always be soaked and thesoaking water discarded before cooking when preparingbeans to improve their nutritional quality

ReferencesAnton, A.A., Ross, K.A., Beta, T., Gary Fulcher, R & Arntfield, S.D (2008) Effect of pre-dehulling treatments on some nutritional and physical properties of navy and pinto beans (Phaseolus vulgaris L.) LWT – Food Science and Technology, 41, 771–778.

Apata, D.F (2008) Effect of cooking methods on available and unavailable carbohydrates of some tropical grain legumes African Journal of Biotechnology, 7, 2940–2945.

Aranda, P., Lo´pez-Jurado, M., Fernande´z, M., Moreu, M.D.C., Porres, J.M & Urbano, G (2004) Bioavailability of calcium and magnesium from faba beans (Vicia faba L var major), soaked in different pH solutions and cooked, in growing rats Journal of the Science of Food and Agriculture, 84, 1514–1520.

Boateng, J., Verghese, M., Walker, L.T & Ogutu, S (2007) Effect of processing on antioxidant contents in selected dry beans (Phaseolus spp L.) LWT – Food Science and Technology, 41, 1541–1547 Campos-Vega, R., Reynoso-Camacho, R., Pedraza-Aboytes, G et al (2009) Chemical composition and in vitro polysaccharide fermen- tation of different beans (Phaseolus vulgaris L.) Journal of Food Science, 74, T59–T65.

Carmona-Garcı´a, R., Osorio-Dı´az, P., Agama-Acevedol, E., Tovar, J.

& Bello-Pe´rez, L.A (2007) Composition and effect of soaking on starch digestibility of Phaseolus vulgaris (L.) cv ‘Mayocoba’ International Journal of Food Science and Technology, 42, 296–302 Chopra, S & Sankhala, A (2004) Effect of soaking and sprouting on tannin, phytate and in vitro iron in underutilized legumes – horse gram (Dolichos biflorus) and moth bean (Phaseolus aconitifolius) Journal of Food Science and Technology, 41, 547–550.

Chopra, H., Sa, U & Ghugre, P (2009) Dietary fibre content of selected legumes: varietal differences and effect of processing Journal of Food Science and Technology, 46, 266–268.

Elmaki, H.B., Abdelrahaman, S.M., Idris, W.H., Hassan, A.B., Babiker, E.E & El Tinay, A.H (2007) Content of antinutritional

factors and HCl-extractability of minerals from white bean

(Phase-olus vulgaris) cultivars: influence of soaking and ⁄ or cooking Food

Chemistry, 100, 362–368.

FAO FAOSTAT (2009) Consumption: Crops Primary Equivalent,

Food and Agriculture Organisation Statistics Division Rome, Italy:

FAO (Food and Agriculture Organisation of the United Nations).

http://faostat.fao.org/site/609/default.aspx#ancor (last accessed 16

August 2010).

Granito, M., Brito, Y & Torres, A (2007) Chemical composition,

antioxidant capacity and functionality of raw and processed

Phaseolus lunatus Journal of the Science of Food and Agriculture,

87, 2801–2809.

Haro, A (1983) La calidad nutritiva de las leguminosas: grano y su

control gene´tico In: Leguminosas de grano (edited by J.I Cubero &

M.T Moreno) Pp 213–224 Madrid, Spain: Ediciones

Mundi-Prensa.

Hemalatha, S., Platel, K & Srinivasan, K (2007) Zinc and iron

contents and their bioaccessibility in cereals and pulses consumed in

India Food Chemistry, 102, 1328–1336.

Huma, N., Anjum, F.M., Sehar, S., Khan, M.I & Hussain, S (2008).

Effect of soaking and cooking on nutritional quality and safety of

legumes Nutrition and Food Science, 38, 570–577.

Kutosˇ, T., Golob, T., Kacˇ, M & Plestenjak, A (2003) Dietary

fibre content of dry and processed beans Food Chemistry, 80, 231–

235.

Luthria, D.L & Pastor-Corrales, M.A (2006) Phenolic acids content

of fifteen dry edible bean (Phaseolus vulgaris L.) varieties Journal of

Food Composition and Analysis, 19, 205–211.

Martı´n-Cabrejas, M.A., Aguilera, Y., Pedrosa, M.M et al (2009) The

impact of dehydration process on antinutrients and protein

digest-ibility of some legume flours Food Chemistry, 114, 1063–1068.

Matella, N.J., Dolan, K.D., Stoeckle, A.W., Bennink, M.R., Lee, Y.S.

& Uebersax, M.A (2005) Use of hydration, germination, and –

galactosidase treatments to reduce oligosaccharides in dry beans.

Journal of Food Science, 70, C203–C207.

Mosha, T.C.E & Vicent, M.M (2004) Nutritional value and

acceptability of homemade maize ⁄ sorghum-based weaning mixtures

supplemented with rojo bean flour, ground sardines and peanut

paste International Journal of Food Sciences and Nutrition, 55, 301–

315.

Muzquiz, M (2008) Conference: Componentes nutricionalmente

activos en leguminosas: implicaciones en nutricio´n y salud In:

Scientific memories – 1st International Scientific Congress and

National Bean Fair (edited by A.C Herna´ndez) ISBN:

978-970-43-0376-1 Pp 226–227 Celaya, Me´xico: Instituto Nacional de

Investigaciones Forestales, Agrı´colas y Pecua´rias.

Nergiz, C & Go¨kgo¨z, E (2007) Effects of traditional cooking

methods on some antinutrients and in vitro protein digestibility of

dry bean varieties (Phaseolus vulgaris L.) grown in Turkey.

International Journal of Food Science and Technology, 42, 868–873.

Oliveira, A.C., Carraro, F., Reis, S.M.P.M et al (2001a) The

elimination of the not absorved water during common bean soaking

resulted in weight gain in rats Brazilian Journal of Nutrition, 14,

153–155.

Oliveira, A.C., Queiroz, K.S., Helbig, E., Reis, S.M.P.M & Carraro,

F (2001b) The domestic processing of the common bean resulted in

a reduction in the phytates and tannins antinutritional factors, in the

starch content and in the raffinose, stachiose and verbascose

flatulence factors Archivos Latinoamericanos de Nutricio´n, 51,

276–283.

Oliveira, A.C., Reis, S.M.P.M., Carvalho, E.M et al (2003)

Increas-ing quantities of phytic acid in the diet did not affect casein

digestibility and weight gain in rats Brazilian Journal of Nutrition,

16, 211–217.

Oliveira, V.R., Ribeiro, N.D., Jost, E & Londero, P.M.G (2008).

Nutritional and microbiological quality of common beans

(Phase-olus vulgaris L.) cooked with or without the use of soaking water.

Revista Cieˆncia e Agrotecnologia, 32, 1912–1918.

Osman, M.A (2007) Effect of different processing methods, on nutrient composition, antinutrional factors, and in vitro protein digestibility of Dolichos lablab bean [Lablab purpuresus (L) sweet] Pakistan Journal of Nutrition, 6, 299–303.

Osorio-Dı´az, P., Bello-Pe´rez, L.A., Agama-Acevedo, E., Torres, A., Tovar, J & Paredes-Lo´pez, O (2002) In vitro digestibility and resistant starch content of some industrialized commercial beans (Phaseolus vulgaris L.) Food Chemistry, 78, 333– 337.

Vargas-Philippi, S.T., Latterza, A.R., Cruz, A.T.R & Ribeiro, L.C (1999) Adapted food pyramid: a guide for a right food choice Brazilian Journal of Nutrition, 12, 65–80.

Pujola`, M., Farreras, A & Casan˜as, F (2007) Protein and starch content of raw, soaked and cooked beans (Phaseolus vulgaris L.) Food Chemistry, 102, 1034–1041.

Ramı´rez-Ca´rdenas, L., Leonel, A.J & Costa, N.M.B (2008) Effect of domestic processing on nutrient and antinutritional factor content in different cultivars of common beans Cieˆncia e Tecnologia de Alimentos, 28, 200–213.

Ranilla, L.G., Genovese, M.I & Lajolo, F.M (2009) Effect of different cooking conditions on phenolic compounds and antiox- idant capacity of some selected Brazilian bean (Phaseolus vulgaris L.) cultivars Journal of Agricultural and Food Chemistry, 57, 5734–5742.

Rehman, Z.-U & Shah, W.H (2004) Domestic processing effects on some insoluble dietary fibre components of various food legumes Food Chemistry, 87, 613–617.

Rehman, Z.U & Shah, W.H (2005) Thermal heat processing effects

on antinutrients, protein and starch digestibility of food legumes Food Chemistry, 91, 327–331.

Salgado, S.M., Melo Filho, A.B., Andrade, S.A.C., Maciel, G.R., Livera, A.V.S & Guerra, N.B (2005) Modification of the concen- tration of resistant starch in macassar bean (Vigna unguiculata L Walp) hydrothermal process and freezing Cieˆncia e Tecnologia de Alimentos, 25, 259–264.

Shimelis, E.A & Rakshit, S.K (2005) Antinutritional factors and in vitro protein digestibility of improved haricot bean (Phaseolus vulgaris L.) varieties grown in Ethiopia International Journal of Food Sciences and Nutrition, 56, 377–387.

Shimelis, E.A & Rakshit, S.K (2007) Effect of processing on antinutrients and in vitro protein digestibility of kidney bean (Phaseolus vulgaris L.) varieties grown in East Africa Food Chemistry, 103, 161–172.

Siddiq, M., Nyombaire, G., Dolan, K.D., Matella, N.J & Harte, J.B (2006) Processing of sugar-coated red kidney beans (Phaseolus vulgaris): fate of oligosaccharides and phytohemagglutinin (PHA), and evaluation of sensory quality Journal of Food Science, 71, C521–C526.

Silva, M.R & Silva, M.A.A.P (1999) Nutritional aspects of phytates and tannins Brazilian Journal of Nutrition, 12, 21–32.

Silva, M.R & Silva, M.A.A.P (2000) Antinutritional factors: protease inhibitors and lectins Brazilian Journal of Nutrition, 13, 3–9.

Teguia, A & Fon Fru, S (2007) The growth performances of broiler chickens as affected by diets containing common bean (Phaseolus vulgaris) treated by different methods Tropical Animal Health and Production, 39, 405–410.

Toledo, T.C.F & Canniatti-Brazaca, S.G (2008) Chemical and nutritional evaluation of Carioca beans (Phaseolus vulgaris L.) cooked by different methods Cieˆncia e Tecnologia de Alimentos, 28, 355–360.

Vadivel, V & Pugalenthi, M (2008) Effect of various processing methods on the levels of antinutritional constituents and protein digestibility of Mucuna pruriens (L.) DC var utilis (Wall ex Wight) Baker ex Burck (velvet bean) seeds Journal of Food Biochemistry,

antinutritional properties of velvet bean seeds Journal of Food

Processing and Preservation, 33, 60–73.

Vasconcellos, A.B., Pinheiro, A.R.O., Recine, E & Carvalho,

M.F.C.C (2006) Dietary Guidelines for the Brazilian Population.

Pp 61–64 Brası´lia, Brazil: Ministry of Health.

Vidal-Valverde, C., Frias, J., Sotomayor, C., Diaz-Pollan, C.,

Fer-nandez, M & Urbano, G (1998) Nutrients and antinutritional

factors in faba beans as affected by processing Zeitschrift fu¨r

Lebensmittel-Untersuchung und -Forschung, 207, 140–145.

Xu, B.J & Chang, S.K.C (2008) Total phenolic content and

antioxidant properties of eclipse black beans (Phaseolus vulgaris

L.) as affected by processing methods Journal of Food Science, 73,

H19–H27.

Xu, B.J & Chang, S.K.C (2009) Total phenolic, phenolic acid,

anthocyanin, flavan-3-ol, and flavonol profiles and antioxidant

properties of pinto and black beans (Phaseolus vulgaris L.) as

affected by thermal processing Journal of Agricultural and Food

Chemistry, 57, 4754–4764.

Xu, B.J., Yuan, S.H & Chang, S.K.C (2007) Comparative analyses

of phenolic composition, antioxidant capacity, and color of cool

season legumes and other selected food legumes Journal of Food

Science, 72, S167–S177.

Supporting Information

Additional Supporting Information may be found in the

online version of this article:

Table S1 Descriptors in Portuguese, English andSpanish used to search articles for the systematicresearch on the influence of soaking on the nutritionalquality of common beans

Table S2 Author, year, country and predictor ables of the selected articles, according to the soakingand cooking methods used

vari-Table S3 Summary of the results found by thesystematic review on the influence of soaking on thenutritional quality of common beans (Phaseolus vulgarisL.) cooked with or without the soaking water

Please note: Wiley-Blackwell are not responsible forthe content or functionality of any supporting materialssupplied by the authors Any queries (other than missingmaterial) should be directed to the corresponding authorfor the article

Original article

Characterisation of pore properties of deep-fat-fried chicken

nuggets breading coating using mercury intrusion porosimetry

technique

Akinbode A Adedeji & Michael Ngadi*

Department of Bioresource Engineering, McGill University, Ste-Anne-de-Bellevue, H9X 3V9 QC, Canada

(Received 22 October 2009; Accepted in revised form 25 May 2010)

Summary The objective of this study was to characterise the pore properties of deep-fat-fried chicken nuggets coating

under different processing conditions namely frying temperatures (170, 180 and 190C) and time (0–240 s)using porosimetry technique Porosity range obtained was between 39.93 and 68.99% Porosity of the freeze-dried samples decreased with frying time The main effect of temperature on porosity was significant(P < 0.05) Porosity showed a high positive and negative correlation with moisture and fat contents, and thecorrelation coefficients ranged between 0.88 and 0.96 and 0.78 and 0.8, respectively Bulk density increasedwith frying time, while apparent density was relatively the same Pore distribution showed bimodality Therewas no significant effect of temperature on pore size distribution Over 70% of the pore volume is made up ofpores greater than 1 lm Pore volume ranged between 0.54 and 1.5 cm3g)1, and it decreased with fryingtime Mean pore diameter was between 0.006 and 389 lm, while with frying time, it ranged between 0.25 and8.32 lm Total pore area was between 2.53 and 16.53 m2g)1 Hysteresis phenomenon showed that some ofthe pores were not perfectly cylindrical in shape

Keywords Breading, chicken nuggets, coating, deep-fat-frying, porosimetry, porosity.

Introduction

Deep-fat-fried foods remain a major part of our diet

although there are concerns about consumption of

high-calorie-content foods Fried foods are easy to prepare,

and they possess unique flavour resulting in a huge

multi-million dollar market that keeps growing

(Suderman, 1996; Garcia et al., 2004; Agriculture and

Agri-Food Canada, 2009) The need to reduce fat

uptake has become the target of several researches

Techniques that have been investigated include

predry-ing, precooking (par-frypredry-ing, microwave and oven

cook-ing) and application of surface coating (Krokida et al.,

2001; Pedreschi & Moyano, 2005; Adedeji et al., 2009)

Some of the coatings used include flour base products

such as wheat, rice and corn flour; starch; seasoning

agents; and hydrocolloids (cellulose derivatives:

carb-oxymethyl cellulose, methyl cellulose, hydroxypropyl

methylcellulose and others like guar gum, xanthan gum)

(Rimac-Brncic et al., 2004) Food coatings are usually

made of batter and⁄ breading system Batters are

leav-ened flour paste, while breading is made of bread

crumbs or ground biscuit Batter adds value to friedproducts by the development of unique surface texture,flavour and colouration They also provide protectivebarriers to fat absorption and moisture loss duringfrying owing to their increased water-holding capacityand structural transformation such as gelation

Food structure is developed naturally or by ing Structure of food has been related to quality changeexperienced during food processing such as drying anddeep-fat-frying, especially at microscopic scale (Mell-ema, 2003; Donald, 2004; Aguilera, 2005) There hasbeen renewed interest in studying microstructural prop-erties of food such as porosity and pore size distributionprimarily because of their importance in defining foodquality and in better understanding of transport prop-erties of foods and optimisation of various processesthat lead to production of porous foods (Pinthus et al.,1995; Rahman et al., 2002; Aguilera, 2005; Witrowa-Rajchert & Lewicki, 2006; Adedeji & Ngadi, 2009)

process-There are several techniques that have been used tostudy pore characteristics of foods These includepycnometry (McDonald & Sun, 2001; Rahman &Sablani, 2003; Kassama & Ngadi, 2004, 2005a; Taiwo

& Baik, 2007), microscopy (Bouchon & Aguilera, 2001;Liang et al., 2006), x-ray computed tomography (Barc-

*Correspondent: Fax: +1 514 398 8387;

e-mail: michael.ngadi@mcgill.ca

doi:10.1111/j.1365-2621.2010.02324.x

elon et al., 1999; van Dalen et al., 2003, 2007; Lim &

Barigou, 2004; Miri et al., 2006; Kim et al., 2007),

magnetic resonance imaging (MRI) (Bows et al., 2001;

Wagner et al., 2008) and porosimetry (Karathanos &

Saravacos, 1993; Karathanos et al., 1996; McDonald &

Sun, 2001; Ngadi et al., 2001; Rahman et al., 2002;

Rahman & Sablani, 2003; Kassama & Ngadi, 2005b)

The objectives of this study were to elucidate pore

characteristics of deep-fat-fried chicken nuggets coating

using mercury intrusion porosimetry and to relate these

pore properties to pretreatment condition, moisture and

fat contents

Principle of mercury intrusion porosimetry (MIP)

The principle of porosimetry is based on capillary law,

non-reactive and non-wetting characteristics of certain

liquid such as mercury It is such that this liquid will not

penetrate a pore until certain pressure is applied The

relationship between the applied pressure and the

diameter of the pore is described by the Young–Laplace

equation also referred to as Washburn expression

where DP is the applied pressure, r is the radius of the

pore (r1and r2are the curvature of the liquid interface),

c is the surface tension of the liquid given as

0.485 N m)1 and h is the contact angle between the

intruding liquid and the wall of the material The

contact angle is usually between 130 and 140 for most

food materials and mercury (Giesche, 2006) A major

assumption in mercury intrusion porosimetry is that the

pore shape is cylindrical This assumption is generally

accepted to simplify what would seemingly have been a

complex problem (Ngadi et al., 2001) According to eqn

1, the pressure differential is usually proportional to the

surface tension of mercury, the contact angle between

the material of interest and the intrusion fluid, mercury,

and inversely proportional to the radius of the pore

Pore size distribution is determined based on the

relationship between the applied pressure, pore radius

and pore volume (eqn 2):

D vð Þ ¼P

r

d vð t vÞ

where D(v)is the pore size volume distribution function,

P is the applied pressure at the point when mercury is

forced into the sample, r is the pore radius that

corresponds to P, vt is the total pore volume and v is

the pore volume at pressure P

Porosimetry, unlike some other techniques, provides

information on bulk and apparent densities, pore size

distribution, pore area, pore shape (hysteresis

phenom-enon) and a wide range of porosity between 0.005 and

360 lm Porosimetry has been used extensively to studypore characteristics of dried foods (Karathanos et al.,1996; Rahman et al., 2002) The method has also beenused for pore characterisation of certain fresh⁄ moistureand partially dried foods (McDonald & Sun, 2001;Rahman et al., 2005) even though a clear explanation onhow moisture present in their sample, which couldhinder accurate acquisition of result, was evacuated wasnot given (Micromeritics, 1999; Giesche, 2006)

Materials and methods

MaterialsThe chicken nuggets used for this study was obtainedfrom a local manufacturer (Olymel, Boucherville, QC,CA) The composition of the breading and battercoating of the chicken nuggets included wheat flour,wheat crumbs, spices, guar gum and salt The chickennugget samples were kept at sub-freezing temperature()50 C) prior to use and brought to refrigerationcondition (4C) for about 4 h before use

Sample preparationChicken nugget samples were fried in fresh canola oil at

170, 180 and 190C for 10 time intervals between 0 and

240 s in a Henny Penny Computron 7000 Pressure Fryer(Model 500C; HP Corporation, Eaton, OH, USA) thathas a capacity for 30 L Three pieces of chicken nuggetswith an average volume of 75 cm3⁄ nugget were fried ineach run, giving an oil to product ratio of approximately130:1 Excess oil on the surface was mopped off usingpaper towel Fried samples were allowed to cool underambient conditions for 20 min Surface coating of thefried chicken nuggets was carefully peeled with handand quick-frozen in liquid Nitrogen The frozen chickennugget samples were then kept in the deep freezer at)40 C for 24 h before they were freeze-dried using alaboratory freeze-dryer (Thermo Savant Modulyod-115,Holbrook, NY, USA) for 36 h

Mercury intrusion porosimetryPore structure characteristics were determined followingthe methods described by Ngadi et al (2001) and(Micromeritics, 1999) Freeze-dried fried chicken nug-gets breading coating were weighed into the penetro-meter of the porosimeter (Auto Pore III series 9400;Micromeritics Inst Co., Norcross, GA, USA) Theporosimeter is capable of measuring more than0.005 lm pore sizes and operates at maximum pressure

of 228 MPa (33000 psi) The penetrometer is made ofglass material in a cylindrical shape with an enlargement

at the one end in bulb shape, which comes in varioussizes for different sample structure (powder or solid)

The particular one used for this study has a 15-cm3bulb

volume, a total stem volume of 1.131 cm3 and a

maximum measurable intrusion volume of 1.057 cm3

The size of penetrometer was chosen to ensure

maxi-mum measurement accuracy because it was estimated

that the total pore volumes of freeze-dried fried chicken

nugget coating samples should be within 25–90% of the

maximum measurable intrusion (stem) volume of the

penetrometer The amount of pore volume equivalent to

a pore size was obtained by eqn 2 A plot of cumulative

pore volume over the pore size range of the fried chicken

nuggets breading coating is presented as the pore

distribution graph Porosity (e) was obtained by

sub-tracting the ratio of apparent and bulk volume from one

where Vapp, qapp, Vbkand qbkare apparent volume and

density, and bulk volume and density, respectively To

measure pore properties such as porosity and pore size

distribution, freeze-dried fried chicken nuggets coating

samples were weighed, placed in the penetrometer bulb

and loaded in the low-pressure port of the porosimeter

After running the coating samples at the low-pressure

port, the penetrometer assembly with the sample inside

was then transferred to the high-pressure port according

to the standard protocol (Micromeritics, 1999) The

generated data were output to a computer system

interphased with the equipment

Moisture content

Moisture content was determined as the mass of

moisture in the fried chicken nuggets breading coating

to the overall weight on wet weight basis (wb) Fried

chicken nuggets coating samples taken at different time

intervals were heated in an oven (Isotemp 700, Fisher

Scientific, Pittsburgh, PA) at 105C for 24 h, and the

difference in weight before and after was used for

computation (Adedeji et al., 2009)

Fat content

Fried chicken nuggets breading samples were

freeze-dried and ground in a coffee grinder (C-Mill 5679,

Bodum USA Inc., New York, NY, USA) Fat content

was determined following the protocol of 960.39 (AOAC,

1990) The freeze-dried fried breading coating samples

(3–5 g) were each weighed into a thimble for fat

extraction in a solvent extractor (SER; Velp Scientifica,

Usmate, Italy) using petroleum ether Fat content was

determined as the ratio of the mass of extracted fat and

dry matter of the sample The whole process of extraction

took approximately two hours

Statistical analysisAll treatments were applied in triplicate; linear regres-sion and analysis of variance were carried out at 5%probability on sas Version 8.2 (SAS, 1999) software.Correlation analysis was performed on Microsoft Excelspreadsheet (Microsoft-office, 2007)

Results and discussion

PorosityThere was a significant effect (P < 0.05) of frying time

on porosity of the fried chicken nuggets breadingsamples Frying temperature also significantly affectedthe porosity values Mean separation showed that there

is significant difference (P < 0.01) between porosities at

170 C, compared to the values at 180 and 190 C.However, there was no interaction effect of temperatureand time on pore development in fried breading andbatter coating of chicken nuggets Porosity values fordeep-fat-fried and freeze-dried chicken nugget breadingcoating samples obtained using mercury intrusionporosimetry (MIP) are presented in Fig 1 The porosityvalues ranged between 39.93% and 68.99% There was ageneral decrease in porosity with frying time Thebreading and batter coating material comprised ofvarious ingredients including water and air pocketsprior to frying Some of the water was evaporatedduring frying, creating pores on its way out of theproduct Parts of these pores were subsequently filledwith fat as frying progressed Thus, pore spaces in thefried chicken nuggets breading coating samples were

30 35 40 45 50 55 60 65 70 75

occupied by moisture, fat and air There is also complex

expansion and collapse of the physical matrix of the

breading and batter system Normally, these phenomena

should translate to increase in porosity during frying

However, it has to be noted that the fried chicken

nuggets breading samples were freeze-dried prior to

porosity measurement as required by the MIP technique

used in this study (Micromeritics, 1999 and Giesche,

2006) Freeze-drying removed moisture but not fat

from the pore spaces Thus, the MIP measurements

accounted for the spaces occupied originally by air and

the spaces occupied by the water that was removed by

freeze-drying The high porosity recorded initially was

attributed largely to water that was removed during the

freeze-drying step The more the frying time, the less

the amount of water available for removal through the

freeze-drying step hence the decreasing trend in

poros-ity Barutcu et al (2009) used the methods of fluid

displacement and nitrogen stereopycnometer to measure

densities and estimated final porosities in the range of

34.8–39.2% for batters fried at 180C for 5 min The

data reported by the authors are in the range of the final

porosities obtained at the end of frying in this study

Adedeji & Ngadi (2010) also presented final porosity for

deep-fat-fried chicken nuggets batter coatings in the

range of 18.21–32.18% as determined by the X-ray

microcomputed tomography (X-ray lCT) technique

Different measurement techniques can lead to different

values of porosity The X-ray lCT technique may have

yielded lower values of porosity owing to its limitation

to measuring pores smaller than 5 lm

Kassama & Ngadi (2005b) also used MIP

measure-ment technique and reported a decreasing trend in

porosity of deep-fat-fried freeze-dried chicken meat

Apart from the measurement technique used, fat

intru-sion into the chicken nuggets during the frying could

also contribute to decrease in porosity (Pinthus et al.,

1995) The interrelationship between pore development,

moisture and fat transfer is illustrated in Fig 2a, b

where moisture loss and fat gained were plotted against

porosity, respectively There were linear correlations

between porosity versus moisture and fat contents

Decreasing moisture content resulted in decreasing

porosity (correlation coefficient was between 0.88 and

0.96 for the different frying temperatures) Fat content

presents a negative correlation with porosity, and the

correlation coefficient R ranged from 0.78 to 0.85 All

were statistically significant at P < 0.05

Pore size distribution

Pore size distribution curves for chicken nuggets

bread-ing coatbread-ing fried at 180C are shown in Figs 3 and 4

The mean pore size ranged between 0.006 and 389 lm

Similar range was obtained at the other temperatures

Final pore volume ranged between 0.54 and 1.5 cm3g)1

(Fig 3) There has not been any report in literature onpore volume of breading coating materials However,Farkas & Singh (1991) obtained a cumulative volumerange of 0.3 and 1.4 ml g)1for air-dried and freeze-driedchicken meat, respectively Kassama & Ngadi (2005b)had reported a range of 0.34–2.08 cm3g)1 for friedchicken meat Ngadi et al (2001) presented pore volumerange of 0.19–0.37 cm3g)1 for oven-cooked soy-ex-tended beef patties McDonald & Sun (2001) reportedvalues of 0.63 and 0.71 cm3g)1 for vacuum-tumbledand non-vacuum-tumbled cooked beef samples, respec-tively There was a general decrease in pore volume withfrying time This is in concert with the porosity trendobtained as discussed previously Similar trend was

0 10 20 30 40 50 60 70 80

reported by Kassama & Ngadi (2005b) for fried chicken

meat Over 70% of the pore volume was constituted by

pores greater than 1 lm (Fig 4) A plot of incremental

pore volume as a function of pore size distribution is

shown in Fig 5 Peak points in the curves indicate

prevalence of certain pore sizes within the distribution

(Rahman & Sablani, 2003) There were two peaks

(bimodal distribution) shown in most of the frying

curves, the first around 0.3–6 lm pore diameter and the

second between 6 and 389 lm, depending on the frying

time Adedeji & Ngadi (2010) also reported bimodal

pore distribution, with two peaks around 94–169 lmand 544 to 582 lm for deep-fat-fried chicken nuggetsbatter coatings formulated with different levels of wheatand rice flours The authors applied the X-ray lCTtechnique, and peak pore size ranges were different fromthose obtained in this study There was a diminishingeffect of frying time on micropore (<10 lm) constituent

of the chicken nuggets breading coating After 60 s offrying, the position of the first peak moved from around0.1 lm towards the 10-lm mark, and this pattern ofdistribution was shown up to 240 s frying time (Fig 5).Bulk density of the chicken nuggets breading coatingranged between 0.46 and 0.78 g cm)3, while the appar-ent density is from 1.21 to 1.48 g cm)3(Table 1) Therewas an increase in bulk density with frying time,indicating an increase in mass per unit volume ofcoating sample as a result of fat intrusion during frying.Kassama & Ngadi (2005b) reported similar trend forfreeze-dried fried chicken meat Apparent densityremained relatively unchanged during the frying oper-ation ranging between 1.21 and 1.48 g cm)3 Taiwo &Baik (2007) reported apparent density in the range of1.03–1.28 g cm)3 for fried sweet potato The rangeobtained in this study is also within the range reported

by Kawas & Moreira (2001) for fried tortilla chips and

by Kassama & Ngadi (2005b) for fried chicken meat.The increasing bulk density with frying time explainedthe decreasing porosity with frying time as reportedearlier The average pore diameter varied with fryingtime between 0.25 and 8.32 lm and did not show anyparticular pattern with frying time Kassama & Ngadi(2005b) had reported a range of 0.25 and 1.17 lm forchicken meat fried for up to 360 s Total pore area was

Figure 3 Frying time effect on pore size distribution as a function

of cumulative volume when frying temperature was 180 C.

Figure 4 Effect of frying time on per cent pore volume

distribution (180 C frying temperature) of chicken nuggets

breading coating.

0.000 0.050 0.100 0.150 0.200 0.250 0.300 0.350 0.400

between 2.53 and 16.53 m2g)1 There was a general

increase in pore area with frying time Kassama &

Ngadi (2005b) had reported a range of 5.86–8.24 m2g)1

for deep-fat-fried chicken meat The difference in sample

composition could be the reason for the difference seen

Analysis of variance showed that influence of frying

temperature on the cumulative pore volume distribution

for deep-fat-fried chicken nuggets coating was

insignif-icant (P < 0.05) (Fig 6) The pore size distribution for

the chicken nuggets breading coating fried at 170 and

190C followed similar trends described for 180 C

frying temperature as mentioned earlier

Hysteresis

The hysteresis phenomenon is shown for both control

(unfried) and fried (15 and 240 s at 180C) chicken

nuggets breading coating (Fig 7) The occurrence of

hysteresis is as a result of change in advancing and

receding contact angles of mercury and the material

during intrusion and extrusion process, respectively The

‘ink bottle’ and ‘energy barrier (snap-off)’ phenomenahave been described to influence the development ofhysteric trend in mercury intrusion porosimetry The

‘ink bottle’ phenomenon describes the occurrence ofpores that are wider at the base than at the throat, whilethe ‘energy barrier’ occurrence describes the differencebetween the intrusion and extrusion pressure, whichoccurs as a result of structural dimensions of pores(Giesche, 2006) These two phenomena confirmed thelikelihood of different geometry formation aside cylin-drical shape in biological materials However, evalua-tion of entrapped mercury in the samples indicates thatminimal disparity exists between intrusion and extrusionvolumes (Fig 7) Kassama & Ngadi (2005b) reportedsimilar result for fried chicken meat, hence justifying theassumption of cylindrical shape made in eqn 1 Tsaki-roglou & Payatakes (1998) and Mason & Morrow(1984, 1991, 1994) all used systems with model pores toconfirm and explain cause of hysteresis phenomenon inMIP analysis of non-axisymmetric pores The occur-rence of hysteresis phenomenon was evident in all theother coating samples fried at every temperature

Conclusion

Mercury intrusion porosimetry technique was used tocharacterise pore properties of deep-fat-fried chickennuggets breading coating Although frying temperatureaffected porosity of fried chicken nugget breadingcoating, it did not affect pore volume distribution.Porosity positively correlates with moisture content butnegatively correlates with frying time There was abimodal pore volume distribution indicating differentpoints of pore prevalence in the distribution in thesamples More than 70% of the pore volume isconstituted by pores greater than 1 lm Bulk density

Figure 6 Temperature effect on pore volume distribution in

deep-fat-fried chicken nuggets coating fried for 240 s.

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6

Table 1 Pore characteristics of chicken nuggets coating fried at

180 C obtained from mercury intrusion porosimetry

Total pore area,

m 2

g –1

Average pore diameter, lm

decreased with frying time, while apparent density

remained somewhat the same There was hysteresis

phenomenon in the distribution, which showed that

some of the pores are not completely cylindrical

Acknowledgment

The authors acknowledge the funding support of

Natural Science and Engineering Research of Canada

(NSERC) in the execution of the project

References

Adedeji, A.A & Ngadi, M.O (2009) 3-D Imaging of deep-fat fried

chicken nuggets breading coating using X-ray micro-CT

Inter-national Journal of Food Engineering, 5, Art 11.

Adedeji, A.A & Ngadi, M.O (2010) Microstructural properties of

deep-fat fried chicken nuggets coating with different batter

formu-lation International Journal of Food Properties, doi: 10.1080/

10942910903131423.

Adedeji, A.A., Ngadi, M.O & Raghavan, G.S.V (2009) Kinetics of

mass transfer in microwave precooked and deep-fat fried chicken

nuggets Journal of Food Engineering, 91, 146–153.

Agriculture and Agri-Food Canada (2009) Canada’s Snack Food

Industry (pp 1-2): Agriculture and Agri-Food Canada Available

from: http://ats-sea.agr.gc.ca/supply/3320_e.pdf Accessed 6th of

August, 2009.

Aguilera, J.M (2005) Why food microstructure? Journal of Food

Engineering, 67, 3–11.

AOAC (1990) Official Methods of Analysis Washington,

DC: Association of Official Analytical Chemists.

Barcelon, E.G., Tojo, S & Watanabe, K (1999) X-ray CT imaging

and quality detection of peach at different physiological maturity.

Transactions of the American Society of Agricultural Engineers, 42,

435–441.

Barutcu, I., Sahin, S & Sumnu, G (2009) Effects of microwave frying

and different flour types addition on the microstructure of batter

coatings Journal of Food Engineering, 95, 684–692.

Bouchon, P & Aguilera, J.M (2001) Microstructural analysis of

frying potatoes International Journal of Food Science and

Technol-ogy, 36, 669–676.

Bows, J.R., Patrick, M.L., Nott, K.P & Hall, L.D (2001)

Three-dimensional MRI mapping of minimum temperatures achieved in

microwave and conventional food processing International Journal

of Food Science and Technology, 36, 243–252.

van Dalen, G., Blonk, H., Aalst, H & Hendriks, C.L.L (2003) 3-D

imaging of foods using X-ray Microtomography GIT Imaging and

Microscopy, 3, 18–21.

van Dalen, G., Nootenboom, P., van Vliet, L.J., Voortman, L &

Esveld, E (2007) 3D Imaging and analysis of porous cereal

products using X-ray microtomography Image Analysis and

Ster-eology, 26, 169–177.

Donald, A (2004) Food for thought Nature Materials, 3, 579–581.

Dullien, F.A.L (1992) Porous Media: Fluid Transport and Pore

Structure, 2nd edn San Diego: Academic Press.

Farkas, B.E & Singh, R.P (1991) Physical Properties of Air-Dried

and Freeze-Dried Chicken While Meat Journal of Food Science, 56,

611–615.

Garcia, M.A., Ferrero, C., Campana, A., Bertola, N., Martino, M &

Zaritzky, N (2004) Methylcellulose coatings applied to reduce oil

uptake in fried products Food Science and Technology International,

10, 339–346.

Giesche, H (2006) Mercury porosimetry: a general (practical)

overview Particle and Particle Systems Characterization, 23, 9–19.

Karathanos, V.T & Saravacos, G.D (1993) Porosity and pore size distribution of starch materials Journal of Food Engineering, 18, 259–280.

Karathanos, V.T., Kanellopoulos, N.K & Belessiotis, V.G (1996) Development of porous structure during air drying of agricultural plant products Journal of Food Engineering, 29, 167–183.

Kassama, L.S & Ngadi, M.O (2004) Pore development in chicken meat during deep-fat frying Lebensmittel-Wissenschaft und-Tech- nologie, 37, 841–847.

Kassama, L.S & Ngadi, M.O (2005a) Pore development and moisture transfer in chicken meat during deep-fat frying Drying Technology, 23, 907–923.

Kassama, L.S & Ngadi, M.O (2005b) Pore structure characterization

of deep-fat-fried chicken meat Journal of Food Engineering, 66, 369– 375.

Kawas, M.L & Moreira, R.G (2001) Effect ofdegree of starch gelatinization on quality attributes of fried tortilla chips Journal of Food Science, 66, 300–306.

Kim, J., Moreira, R.G., Huang, Y & Castell-Perez, M.E (2007) 3-D dose distributions for optimum radiation treatment planning of complex foods Journal of Food Engineering, 79, 312–321.

Krokida, M.K., Oreopoulou, V., Maroulis, Z.B & Marinos-Kouris,

D (2001) Effect of pre-drying on quality of french fries Journal of Food Engineering, 49, 347–354.

Liang, B., Sebright, J.L., Shi, Y., Hartel, R.W & Perepezko, J.H (2006) Approaches to quantification of microstructure for model lipid systems Journal of the American Oil Chemists’ Society, 83, 389–399 Lim, K.S & Barigou, M (2004) X-ray micro-computed tomography of cellular food products Food Research International, 37, 1001–1012 Mason, G & Morrow, N.R (1984) Coexistence of menisci and the influence of neighboring pores on capillary displacement curvatures in sphere packings Journal of Colloid Interface Science, 100, 519–535 Mason, G & Morrow, N.R (1991) Capillary behavior of a perfectly wetting liquid in irregular triangular tubes Journal of Colloid Interface Science, 141, 262–274.

Mason, G & Morrow, N.R (1994) Effect of contact angle on capillary displacement curvatures in pore throats formed by spheres Journal of Colloid Interface Science, 168, 130–141.

McDonald, K & Sun, D.-W (2001) Pore size distribution and structure of a cooked beef product as affected by vacuum cooling Journal of Food Process Engineering, 24, 381–403.

Mellema, M (2003) Mechanism and reduction of fat uptake in deep-fat fried foods Trends in Food Science and Technology, 14(9), 364–373.

Micromeritics (1999) Autopore III: Operator’s Manual Norcross, GA: Micromeritics Instrument Corporation.

Microsoft-office (2007) Microsoft excel.

Miri, T., Bakalis, S., Bhima, S.D & Fryer, P (2006) Use of X-ray micro-CT to characterize structure phenomena during frying Paper number: 10.1051 presented at 13th World Congress of Food Science and Technology held at Nantes, France from 17–21 September 2006.

Ngadi, M.O., Kassama, L.S & Raghavan, G.S.V (2001) Porosity and pore size distribution in cooked meat patties containing soy protein Canadian Biosystems Engineering, 43, 17–24.

Pedreschi, F & Moyano, P (2005) Effect of pre-drying on texture and oil uptake of potato chips LWT – Food Science and Technology, 38, 599–604.

Pinthus, E.J., Weinberg, P & Saguy, I.S (1995) Oil Uptake in deep fat frying as affected by porosity Journal of Food Science, 60, 767–769 Rahman, M.S & Sablani, S.S (2003) Structural characteristics of freeze-dried abalone: porosimetry and puncture test Food and Bioproducts Processing: transactions of the Institution of Chemical Engineers, Part C, 81, 309–315.

Rahman, M.S., Al-Amri, O.S & Al-Bulushi, I.M (2002) Pores and physico-chemical characteristics of dried tuna produced by different methods of drying Journal of Food Engineering, 53, 301–313.

Rahman, M.S., Al-Zakwani, I & Guizani, N (2005) Pore formation

in apple during air-drying as a function of temperature: porosity and

pore-size distribution Journal of the Science of Food and Agriculture,

85, 979–989.

Rimac-Brncic, S., Lelas, V., Rade, D & Simundic, B (2004).

Decreasing of oil absorption in potato strips during deep fat frying.

Journal of Food Engineering, 64, 237–241.

SAS (1999) SAS procedures guide Version 8.2, 3rd edn Cary, N.C:

SAS Institute, Inc.

Suderman, D.R (1996) Effective use of flavourings and seasonings in

batter and breading systems In: Batters and Breadings in Food

Processing (edited by K Kulp & R Loewe) Pp 73–91 Minnesota,

USA: American Association of Cereal Chemists.

Taiwo, K.A & Baik, O.D (2007) Effects of pre-treatments on the shrinkage and textural properties of fried sweet potatoes LWT – Food Science and Technology, 40, 661–668.

Tsakiroglou, C & Payatakes, A (1998) Mercury intrusion and retraction in model porous media Advanced Colloid Interface Science, 75, 215–253.

Wagner, M.J., Loubat, M., Sommier, A et al (2008) MRI study of bread baking: experimental device and MRI signal analysis Inter- national Journal of Food Science and Technology, 43, 1129–1139 Witrowa-Rajchert, D & Lewicki, P.P (2006) Rehydration properties

of dried plant tissues International Journal of Food Science and Technology, 41, 1040–1046.

Original article

Acerola and cashew apple as sources of antioxidants and dietary

fibre

Maria do Socorro M Rufino,1,2a Jara Pe´rez-Jime´nez,2b Marı´a Tabernero,2c Ricardo E Alves,3* Edy S de Brito3

& Fulgencio Saura-Calixto2

1 Federal Rural University of the Semi-Arid, Mossoro´-RN, Brazil

2 Department of Metabolism and Nutrition, Institute for Food Science and Technology and Nutrition (ICTAN-CSIC), Madrid, Spain

3 Postharvest Physiology and Technology Laboratory, Embrapa Tropical Agroindustry, Fortaleza-CE, Brazil

(Received 6 April 2010; Accepted in revised form 22 July 2010)

Summary Several tropical fruits have been described as natural sources of dietary fibre (DF) and phenolic compounds,

associated with different health effects The aim of this work was to ascertain the DF, phenolic compoundscontent (including non-extractable polyphenols, mostly associated with DF) and antioxidant capacity inacerola fruits and cashew apples from selected clones ‘BRS 236’ acerola fruits presented a high antioxidantcapacity because of the combination of both extractable polyphenols and l-ascorbic acid (providing together

a Folin value of 170 kg)1g d.m.) ‘CCP 76’ cashew apples contained 28 g kg)1d.m of extractablepolyphenols and 13 g kg)1d.m of ascorbic acid as well as a high amount of non-extractable condensedtannins (52 g kg)1d.m.) DF content was of 260 g kg)1d.m in acerola fruit and of 209 g kg)1d.m incashew apple Acerola fruits and cashew apple should therefore be considered as new natural sources of DFand phenolic compounds

Keywords Anacardium occidentale, antioxidant capacity, dietary fibre, Malpighia emarginata, polyphenols.

Introduction

Significant amount of bioactive constituents, such as

dietary fibre (DF) or natural antioxidants, i.e

polyphe-nols, have been found during last years in many tropical

fruits and related by-products (Jime´nez-Escrig et al.,

2001; Vasco et al., 2008; Rufino et al., 2010a) In many

studies, all these constituents have been related to the

prevention of several chronic diseases, such as

cardio-vascular disease or certain cancers (Bingham et al.,

2003; Arts & Hollman, 2005) However, as tropical

fruits include a wide variety of products, there still

remain many fruits commonly consumed in tropical

countries, but whose content in bioactive compounds is

still totally or partially unknown

Acerola (Malpighia emarginata), originally from theAntilles, can be found from South Texas, throughMexico and Central America to northern South Amer-ica and throughout the Caribbean, being Brazil now theworld’s largest producer, consumer and exporter (Assis

et al., 2008) Acerola is characterised by a high vitamin

C content, which is many times higher than that of otherfruits that are considered good sources of this vitamin,such as guava, cashew apple, orange or lemon (Alves

et al., 1995) Besides its consumption as a fruit, acerolapresents a wide potential to be used in different juicesand beverages rich in bioactive compounds, as it hasbeen described in several works (Freitas et al., 2006;Mezadri et al., 2008; Lima et al., 2009; Sampaio et al.,2009)

The cashew (Anacardium occidentale) is native toTropical America Originating in Brazil, it has becomenaturalised in many tropical countries such as Vietnam,India, Nigeria, Tanzania, Ivory Coast, Mozambiqueand Benin Cashew is formed by the developed peduncle(apple) attached to the nut (actual fruit composed ofshell + kernel) The peduncle, which is also calledpseudo-fruit, false fruit, cashew apple or simply cashew,represents the edible portion, in natura and also as

*Correspondent: Fax: (55) 85 3391 7222;

e-mail: elesbao@cnpat.embrapa.br

a Present address: Food Technology Department, Federal University of

Ceara, Av Mister Hall, 2977, Pici, 60356-000 Fortaleza, CE, Brazil.

b

Present address: Institute of Advanced Chemistry of Catalonia-CSIC,

c ⁄ Jordi Girona, 18-26, 08034 Barcelona, Spain.

c

Present address: Experimental Surgery Department, University

Hos-pital La Paz, Paseo de la Castellana, 261, 28046 Madrid, Spain.

doi:10.1111/j.1365-2621.2010.02394.x

juices, pulp and preserves (Filgueiras et al., 1999) Apart

from the usual presence of sugars and organic acids, one

important characteristic of cashew apple is its high

vitamin C content, four times higher than sweet orange

(Akinwale, 2000)

For both acerola and cashew, several genetic

improve-ment programmes have been developed to obtain clones

allowing higher productivity In particular, the clone

‘BRS 236’ acerola (Cereja) was obtained from the

programme developed in Embrapa Tropical

Agroindus-try (Paiva et al., 1999) and the clone ‘CCP 76’ cashew

was obtained at the Ceara´ Agricultural Research

Cor-poration (Paiva et al., 2005) These particular clones

were compared to others of the same fruits in previous

works (Abreu et al., 2009; Sampaio et al., 2009) The

study of nutritional characteristic of clones allows more

reproducible results when compared to native fruits,

besides the production advantages of clones

Antioxidant capacity of acerola and cashew apple was

recently determined by various techniques (Alves et al.,

2008; Mezadri et al., 2008; Abreu et al., 2009; Sampaio

et al., 2009; Rufino et al., 2010a) However,

determina-tions were performed only in the supernatants of

aqueous-organic extractions, without considering the

antioxidant capacity associated with the residues of

these extractions, which may have specific health

prop-erties (Serrano et al., 2009)

The aim of this work was to ascertain the DF and

total phenolic contents of ‘BRS 236’ acerola fruits and

‘CCP 76’ cashew apples, and their antioxidant capacity

Based on these results further it would be possible to

evaluate the application of these tropical fruits as

antioxidants and sources of DF in functional foods or

as naturally antioxidant dietary supplements

Materials and methods

Chemicals and samples

Pepsin, glucose, tiourea and anthrone were obtained

from Merck (Darmstadt, Germany) Amyloglucosidase,

pancreatin, lipase, a-amylase,

2,2¢-Azino-bis(3-ethyl-benz-thiazoline-6-sulphonic acid) [ABTS],

6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid (Trolox),

2,2¢-Azobis(2-methylpropionamidine) dihydrochloride

[AAPH], 2,2 Diphenyl-1-picrylhydrazyl [DPPH•] gallic

acid and galacturonic acid were obtained from

Sigma-Aldrich Quı´mica, S.A (Madrid, Spain)

2,4,6-Tri(2-pyridyl)-5-triazine [TPTZ] was from Fluka Chemicals

(Madrid, Spain)

3,6¢-dihydroxy-spiro-[isobenzofuran-1-[3H],9¢[9H]-xanthen]-3-one (Fluorescein) and iron

III-clorure-6-hydrate were from Panreac, Castellar del

Valle´s (Barcelona, Spain) All reagents used were of

analytical grade

Samples analysed were ‘BRS 236’ acerola and ‘CCP

76’ cashew apple ‘BRS 236’ acerola was harvested in

Limoeiro do Norte, CE, Brazil and ‘CCP 76’ cashewapple in the Experimental Station of Embrapa TropicalAgroindustry, Pacajus, CE, Brazil Two kilograms ofeach fruit was harvested in the second semester of 2007 atthe commercial maturity stage (completely ripe), accord-ing to several parameters previously measured in them(Rufino et al., 2010b) Samples were taken from tendifferent trees and from different regions of them, toachieve a homogeneous sample nevertheless, as thesefruits were clones, there were no high differences betweenthem After harvesting, the fruits were transported to thelaboratory and processed in a domestic blender (Walita,Varginha, Brazil) to obtain a pulp and the seeds werediscarded Each fruit was divided into three subsamplesand the pulp was freeze-dried (LH 4500; Terroni Fauvel,Sa˜o Carlos, Sa˜o Paulo, Brazil) and milled to a particlesize of <0.5 mm in a centrifugal milling

MethodsDietary fibre determinationDietary fibre was measured as indigestible fraction,based on a procedure previously described (Saura-Calixto et al., 2000) This method combines enzymatictreatments and separation of digestible compounds bydialysis using physiological conditions (temperature andpHs), obtaining the fraction of food that is not digested

in the small intestine and reach the colon, composed oftwo fractions: a soluble fraction (supernatant) and aninsoluble fraction (residue of enzymatic digestion)(Fig 1) Total DF was calculated as the sum ofinsoluble DF components [resistant starch, non-starchpolysaccharides (NSP), Klason lignin, resistant protein,ash, extractable polyphenols, proanthocyanidins, andhydrolysable phenols] plus soluble DF components(NSP and extractable polyphenols)

Samples (0.3 g) were incubated with pepsin (0.2 mL

of a 300 mg mL)1 solution in 0.08 m HCl–KCl buffer,

pH 1.5, 40C, 1 h), pancreatin (1 mL of a 5 mg mL)1solution in 0.1 m phosphate buffer, pH 7.5, 37C, 6 h)and a-amylase (1 mL of a 120 mg mL)1 solution in0.1 m Tris–maleate buffer, pH 6.9, 37C, 16 h) Sam-ples were centrifuged (15 min, 3000 g) and superna-tants removed Residues were washed twice with 5 mL

of distilled water, and all supernatants were combined.Each supernatant was incubated with 100 lL ofamyloglucosidase for 45 min at 60C before beingtransferred to dialysis membranes (12 000–14 000molecular weight cutoff, Visking dialysis tubing; Med-icell International Ltd., London, UK) and dialysedagainst water for 48 h at 25C to eliminate digestiblecompounds

NSP were hydrolysed with 1 m sulphuric acid at

100C for 90 min and spectrophotometrically measured

by anthrone assay (Loewus, 1952) The residue wasweighed to determine insoluble DF, and resistant

protein and ash were determined in it (see ‘Other

Determinations’) In this residue, after treatment with

sulphuric acid (12 m, 20C for 3 h; dilution to 1 m and

incubation for 2 h, 100C), NSP were determined

spectrophotometrically as neutral sugars (by anthrone

assay) and uronic acids and klason lignin was

deter-mined gravimetrically (Loewus, 1952; Scott, 1979)

Extraction of antioxidants

Figure S1 shows a scheme of the treatment applied to the

samples to determine antioxidant capacity In a capped

centrifuge tube, 0.5 g of sample was placed; 20 mL of

acidic methanol⁄ water ⁄ HCl (50:50, v ⁄ v; pH 2) was added

and the tube was thoroughly shaken at room temperature

for 1 h The tube was centrifuged at 2500 g in a Thermo

Heraeus Megafuge 11 (Thermo Fisher, Waltham, MA,

USA) for 10 min and the supernatant was recovered

Twenty millilitres of acetone⁄ water (70:30, v ⁄ v) was

added to the residue, and shaking and centrifugation are

repeated Methanolic and acetonic extracts were

com-bined and used to determine the antioxidant capacity

associated with extractable antioxidants (Figs 1 and S1)

The residues of these extractions were subjected either to

hydrolysis with H2SO4in methanol to release

hydroly-sable tannins (Figs 1 and S1) or to treatment with

butanol⁄ HCl ⁄ FeCl3to determine proanthocyanidins or

condensed tannins (Figs 1 and S1)

Determination of antioxidant compounds

Total polyphenols in extracts were determined

accord-ing to the Folin-Ciocalteu method in supernatants

(Figs 1 and S1) (Singleton et al., 1999) Test sample(0.5 mL) was mixed with 1 mL of Folin-Ciocalteureagent and swirled After 3 min, 10 mL of sodiumcarbonate solution (75 g L)1) was added and mixed.Additional distilled water was mixed thoroughly byinverting the tubes several times After 1 h, theabsorbance at 750 nm was recorded The results wereexpressed as g gallic acid equivalents (GAE) g kg)1.Proanthocyanidins (condensed tannins) not extracted

by the previous aqueous-organic procedure weremeasured at 555 nm after hydrolysis with buta-nol⁄ HCl ⁄ FeCl3 (3 h, 100C) in supernatants (Figs 1and S1) (Reed et al., 1982) Results were comparedwith carob pod (Ceratonia siliqua) proanthocyanidinstandard (Nestle´, Ltd., Vers-Chez-les Blanes, Switzer-land), which has been reported to be a suitablestandard for proanthocyanidins determination (Pe´rez-Jime´nez et al., 2009) Hydrolysable tannins were mea-sured by hydrolysis with methanol and sulphuric acidfor 20 h at 85C (Figs 1 and S1) (Hartzfeld et al.,2002) Concentration was estimated by the Folin–Ciocalteu method and expressed as g GAE kg)1(Singleton et al., 1999)

Antioxidant capacity methodsDPPH• (Free-Radical Scavenging) Assay (Brand-Wil-liams et al., 1995; Sa´nchez-Moreno et al., 1998): Afteradjusting the blank with methanol, 0.1 mL of thesample was mixed with 3.9 mL of a DPPH•methanolicsolution (60 lm) The absorbance at 515 nm was mea-sured until the reaction reached the plateau depending

Acerola fruit and cashew apple pulps

Enzymatic treatments

Centrifugation

Supernatant

Dialysis Retentant (soluble dietary fibre)

(insoluble dietary fibre) Resistant protein, ashes Methanol/H2SO4

Neutral sugars, uronic acids, klason lignin, AC, HT Neutral sugars, uronic acids

Supernatant 4

Residue

AC, CT AC: antioxidant capacity; EPP: extractable polyphenols; HT: hydrolysable tannins;

CT: condensed tannins.

Butanol/HCl/FeCl3

Figure 1 Scheme of the determination of

dietary fibre and associated antioxidants in

acerola fruit and cashew apple.

on the sample, this time may be between 1 and 60 min.

A calibration curve at that wavelength was made to

calculate the remaining DDPH• The parameter EC50,

which reflects 50% depletion of DPPH•free-radical, was

expressed in terms of grams of fruit equivalent per gram

of DPPH• in the reaction medium The time taken to

reach the steady state at EC50(tEC50) and the antiradical

efficiency [AE] (AE = 1⁄ (EC50*tEC50) were also

deter-mined

ABTS•+ assay (Re et al., 1999): ABTS•+ radical

cation (ABTS•+) was produced by reacting 7 mm

ABTS•+ stock solution with 2.45 mm potassium

per-sulfate and allowing the mixture to stand in the dark at

room temperature for 12–16 h before use The ABTS•+

solution was diluted with methanol to an absorbance

of 0.70 ± 0.02 at 658 nm After the addition of 100 lL

of sample or Trolox standard to 3.9 mL of diluted

ABTS•+solution, absorbance readings were taken every

20 s, using a Beckman DU-640 (Beckman Instruments

Inc Fullerton, CA, USA) spectrophotometer The

reaction was monitored during 6 min The percentage

inhibition of absorbance versus time was plotted, and

the area below the curve (0–6 min) was calculated

Solutions of known Trolox concentrations were used for

calibration

Ferric reducing antioxidant power (FRAP) assay

(Benzie & Strain, 1996; Pulido et al., 2000): FRAP

reagent (900 lL), freshly prepared and warmed at

37C, was mixed with 90 lL of distilled water and

either 30 lL of test sample or standard or appropriate

reagent blank Reading at the absorption maximum

(595 nm) was taken every 15 s, using a

spectrophotom-eter The readings at 30 min were selected for

calcula-tion of FRAP values Solucalcula-tions of known Trolox

concentrations were used for calibration

Oxygen radical absorbance capacity (ORAC) assay

(Ou et al., 2001): Sample⁄ blank was mixed with PBS

buffer, AAPH and fluorescein Fluorescence was

re-corded until it reached zero (excitation wavelength

493 nm, emission wavelength 515 nm) in a fluorescence

spectrophotometer Perkin–Elmer LS 55 (Waltham, MA,

USA) at 37C Results were calculated using the

differences of areas under the fluorescein decay curve

between the blank and the sample and were expressed as

Trolox equivalents

Other determinations

Protein was determined using an automated nitrogen

analyser FP-2000; Dumas Leco Corp (St Joseph, MI,

USA) Fat content was determined using a Soxhlet

System HT (FOSS, Ho¨gana¨s, Sweden) extractor with

petroleum ether Ash content was determined with an

electric muffle furnance for 16 h at 550C quantified

gravimetrically and ascorbic acid

spectrophotometri-cally by using 2,6-diclorophenol (Strohecker & Henning,

1967)

Statistical analysisAll analyses were performed in triplicate and resultswere expressed as mean values ± standard deviation ondry matter (d.m.) basis

Results and discussion

Dietary fibreThe content and composition of DF in acerola andcashew apple, including neutral sugars, uronic acids,Klason lignin, resistant protein, ash and polyphenols,are presented in Table 1

Acerola fruits and cashew apple had a total DFcontent of 260 and 209 g kg)1d.m., respectively, most

of it insoluble DF (79% on acerola fruits and 88% incashew apple) This content is in the same range as thevalues reported for some common fruits such as apples,oranges or bananas, in which it ranges from 170 to

360 g kg)1d.m (Saura-Calixto et al., 2000)

Klason lignin content in acerola was 135.4 g kg)1d.m Klason lignin is the gravimetric residueobtained from the sulphuric acid treatment typicallyperformed to solubilize and hydrolyse polysaccha-rides in DF analysis; it is made up of lignin with

an associated mixture of charides

protein–polyphenols–polysac-Total neutral sugars and uronic acids represent themajor carbohydrate fraction of DF, reaching values of

76 and 57.5 g kg)1d.m for acerola fruits and cashewapple, respectively This portion of non-disgestiblecarbohydrates reaches the colon where it is potentiallyfermentable by colonic microbiota Some of the metab-olites generated during colonic fermentation of carbo-hydrates, as short chain fatty acids (especially butyrate),have been described as beneficial for intestinal health(Wong et al., 2006) Finally, cashew apple DF con-tained 41.2 g kg)1d.m of resistant or indigestibleprotein

Table 1 also shows that polyphenols are present on

DF in acerola and cashew apples, lending antioxidantcapacity to it, as discussed in the next paragraph

Polyphenols and antioxidant capacityExtractable polyphenol content determined in aqueous-organic extracts from acerola and cashew applecan be seen in Table 2 Cashew apple exhibited highpolyphenol content (28 g kg)1d.m.) In the case ofacerola, the extraordinarily high value obtained(170 g kg)1d.m.) may be an overestimation as ascorbicacid (present in acerola at a concentration of

150 g kg)1d.m.), like many other compounds, can reactwith Folin-Ciocalteau reagent (Prior et al., 2005).Cashew apple also contained ascorbic acid, but at amuch lower concentration (13 g kg)1d.m.)

Both acerola and cashew apple contained significant

amount of hydrolysable tannins in the residues of

aqueous-organic extractions (3.9 g kg)1d.m in acerola

and 12.1 g kg)1d.m in cashew apple) Acerola also

contained 52 g kg)1g non-extractable condensed

tan-nins With regard to cashew apple, these non-extractable

polyphenols were even more abundant than extractable

polyphenols Although non-bioavailable in the small

intestine, these non-extractable polyphenols linked to

DF would reach the colon, where the microbiota may

release them and modify their structure, giving place to

some absorbable metabolites and to antioxidant

metab-olites that can improve the colonic antioxidant status

(Gonthier et al., 2003; Cerda´ et al., 2005)

Antioxidant capacity associated with phenolic

com-pounds in the different supernatants obtained (Fig S1)

was determined by FRAP, ABTS•+, DPPH• and

ORAC (Table 2) It was high in both acerola fruits

and cashew apples in comparison with other South

America fruits known for their high antioxidant

con-tents or with the mix of fruits consumed in the Spanish

Mediterranean diet (Serrano et al., 2007; Vasco et al.,

2008; Rufino et al., 2010a)

Hydrolysable and condensed tannins in cashew appleslikewise presented high antioxidant capacity OnlyABTS•+ could be applied to condensed tannins, asbutanol⁄ HCl interferes in the other antioxidant capac-ity assays Results were lower than the antioxidantcapacity measured in the extractable polyphenolfraction

Polyphenols and antioxidant capacity associated withdietary fibre

Antioxidant capacity and phenolic compounds ated with DF (soluble and insoluble) are shown inTable 3 Extractable polyphenols associated with solu-ble DF and extractable polyphenols and hydrolysabletannins associated with insoluble DF were all observed

associ-in acerola fruits, whereas associ-in cashew apples all thepolyphenols associated with DF were associated with itsinsoluble fraction Although values in Table 3 cannot bedirectly compared with those in Table 2, as the extrac-tion methods are different (aqueous-organic vs enzy-matic extraction), a significant proportion of thepolyphenols is clearly associated with DF

Table 1 Content and composition of dietary fibre (g kg)1d.m.) of acerola fruit and cashew apple a

Mean value ± standard deviation, n = 3.

Table 2 Polyphenols and antioxidant capacity of acerola fruit and cashew apple in aqueous-organic extracts and their residues a

FRAP (lmol Trolox g)1d.m.) 1640.29 ± 94.51 245.44 ± 14.65 14.87 ± 0.82 56.37 ± 4.15 n.d n.d.

ORAC (lmol Trolox g)1d.m.) 816.55 ± 57.59 287.31 ± 57.63 173.92 ± 20.74 344.65 ± 65.25 n.d n.d.

Antiradical Efficiency, AE = 1 ⁄ (EC 50 *t EC50 ).

Also, all these polyphenol fractions associated with

DF exhibited high antioxidant capacity according to the

different methodologies employed For comparative

purposes, antioxidant capacity associated with DF of a

mixture of the fruits consumed in the Spanish diet was

17.7 lmol Trolox g)1d.m by FRAP assay (Serrano

et al., 2007), much lower than antioxidant capacity

associated with DF either of acerola or of cashew apple

Moreover, the antioxidant capacity associated with DF

in ‘BRS 236’ acerola fruits and ‘CCP 76’ cashew apples

may be of nutritional significance as these antioxidant

compounds would reach the colon intact and they could

produce various beneficial effects as discussed earlier

Conclusions

In summary, this study provides new nutritional data on

the composition of ‘BRS 236’ acerola fruits and ‘CCP

76’ cashew apples Acerola fruits presented a high

antioxidant capacity as determined by various assays

because of the combination of high concentrations of

extractable polyphenols (170 g kg)1d.m.) together with

an exceptional amount of ascorbic acid Cashew apples

contained 28 g kg)1d.m of extractable polyphenols

and a high amount of non-extractable condensed

tannins (52 g kg)1d.m.), both conferring it as well a

high antioxidant capacity A significant part of these

polyphenols, especially in the case of cashew apples,

were associated with DF, showing specific health

prop-erties Acerola fruits and cashew apples also presented a

high DF content (about 200 g kg)1d.m.) Owing to

their content in bioactive compounds, acerola fruits and

cashew apples have considerable nutritional and health

potential Moreover, these two clones present tion advantages and a higher reproductibility of resultswhen compared to native fruits

produc-Acknowledgments

The authors acknowledge the financial support from theCAPES, CNPq, ICTAN-CSIC, EMBRAPA, UFERSAand European Union (FP6 Contract no.: 0015279)

ReferencesAbreu, C.R.A., Maia, G.A., Figueiredo, R.W et al (2009) Quality, bioactive compounds, and antioxidant activity of cashew apple (Anacardium occidentale L.) from commercial early dwarf clones Acta Horticulturae, 841, 451–454.

Akinwale, T.O (2000) Cashew apple juice: its use in fortifying the nutritional quality of some tropical fruits European Food Research and Technology, 211, 205–207.

Alves, R.E., Chitarra, A.B & Chitarra, M.I.F (1995) Postharvest physiology of Acerola (Malpighia emarginata D.C.) fruits: matura- tion changes, respiratory activity and refrigerated storage at ambient and modified atmospheres Acta Horticulturae, 370, 223–230 Alves, R.E., Brito, E.A., Rufino, M.S.M & Sampaio, C.G (2008) Antioxidant activity measurement in Tropical fruits: a case study with acerola Acta Horticulturae, 773, 299–305.

Arts, I.C & Hollman, P.C (2005) Polyphenols and disease risk in epidemiological studies American Journal of Clinical Nutrition, 81, 317–325.

Assis, S.A., Fernandes, F.P., Martins, A.B.G & Faria Oliveira, O.M.M (2008) Acerola: importance, culture conditions, production and biochemical aspects Fruits, 63, 93–101.

Benzie, I.F.F & Strain, J.J (1996) The ferric reducing ability of plasma (FRAP) as a measure of antioxidant power: the FRAP assay Analytical Biochemistry, 239, 70–76.

Bingham, S., Day, N.E., Luben, R et al (2003) Dietary fibre in food and protection against colorectal cancer in the European Prospec-

Table 3 Polyphenols and antioxidant capacity associated with dietary fibre of acerola fruit and cashew apple a

Extractable polyphenols b

tive Investigation into Cancer and Nutrition (EPIC): an

observa-tional study Lancet, 361, 1496–1501.

Brand-Williams, W., Cuvelier, M.E & Berset, C (1995) Use of a free

radical method to evaluate antioxidant activity Lebensmittel

Wis-senchaft und Technologie, 28, 25–30.

Cerda´, B., Periago, P., Espı´n, J.C & Toma´s-Barbera´n, F.A (2005).

Metabolism of antioxidant and chemopreventive ellagitannins from

strawberries, raspberries, walnuts, and oak-aged wine in humans:

identification of biomarkers and individual variability Journal of

Agricultural and Food Chemistry, 53, 227–235.

Filgueiras, H.A.C., Alves, R.E., Mosca, J.L & Menezes, J.B (1999).

Cashew apple for fresh consumption: research on harvest and

postharvest technology in Brazil Acta Horticulturae, 485, 155–

160.

Freitas, C.A.S., Maia, G.A., Sousa, P.H.M., Brasil, I.M & Pinheiro,

A.M (2006) Storage stability of acerola tropical fruit juice obtained

by hot fill method International Journal of Food Science and

Technology, 41, 1216–1221.

Gonthier, M.P., Donovan, J.L., Texier, O., Felgines, C., Remesy, C &

Scalbert, A (2003) Metabolism of dietary procyanidins in rats Free

Radicals in Biology and Medicine, 35, 837–844.

Hartzfeld, P.W., Forkner, R., Hunter, M.D & Hagerman, A.E.

(2002) Determination of hydrolyzable tannins (gallotannins and

ellagitannins) after reaction with potassium iodate Journal of

Agricultural and Food Chemistry, 50, 1785–1790.

Jime´nez-Escrig, A., Rinco´n, M., Pulido, R & Saura-Calixto, F (2001).

Guava fruit (Psidium guajava L.) as a new source of antioxidant

dietary fiber Journal of Agricultural and Food Chemistry, 49, 5489–

5493.

Lima, A.S., Maia, G.A., Sousa, P.H.M., Prado, G.M & Rodrigues, S.

(2009) Storage stability of a stimulant coconut water–acerola fruit

juice beverage International Journal of Food Science and Technology,

44, 1445–1451.

Loewus, F.A (1952) Improvement in the anthrone method for

determination of carbohydrates Analytical Chemistry, 24, 219.

Mezadri, T., Villan˜o, D., Ferna´ndez-Pacho´n, M.S., Garcı´a-Parrilla,

M.C & Troncoso, A.M (2008) Antioxidant compounds and

antioxidant activity in acerola (Malpighia emarginata DC.) fruits

and derivatives Journal of Food Composition and Analysis, 21, 282–

290.

Ou, B., Hampsch-Woodill, M & Prior, R.L (2001) Development and

validation of an improved oxygen radical absorbance capacity assay

using fluorescein as the flourescent probe Journal of Agricultural and

Food Chemistry, 49, 4619–4626.

Paiva, J.R., Correia, M.P.F., Freire, F.C.O., Braga Sobrinho, R &

Juca´, W (1999) Selec¸a˜o massal de acerola em plantio comercial.

Pesquisa Agropecua´ria Brasileira, 34, 505–511.

Paiva, J.R., Barros, L.M., Cavalcanti, J.J.V et al (2005) Selection of

dwarf cashew clones for commercial plantation in Aracati County,

Ceara´ State, Brazil Revista Cieˆncia Agronoˆmica, 36, 338–343.

Pe´rez-Jime´nez, J., Arranz, S & Saura-Calixto, F (2009)

Proantho-cyanidin content in foods is largely underestimated in the literature

data: an approach to quantification of the missing

proanthocyani-dins Food Research International, 42, 1381–1388.

Prior, R.L., Wu, X & Schaich, K (2005) Standardized methods for

the determination of antioxidant capacity and phenolics in foods

and dietary supplements Journal of Agricultural and Food

Chemis-try, 53, 4290–4302.

Pulido, R., Bravo, L & Saura-Calixto, F (2000) Antioxidant activity

of dietary polyphenols as determined by a modified ferric reducing

antioxidant power assay Journal of Agricultural and Food

Chemis-try, 48, 3396–3402.

Re, R., Pellegrini, N., Proteggente, A., Pannala, A., Yang, M &

Rice-Evans, C (1999) Antioxidant activity applying an improved ABTS

radical cation decolorization assay Free Radicals in Biology and Medicine, 26, 1231–1237.

Reed, J.D., McDowell, R.T.E., Van Soest, P.J & Horvath, P.R.J (1982) Condensed tannins: a factor limiting the use of cassava forage Journal of the Science of Food and Agriculture, 33, 213– 220.

Rufino, M.S.M., Alves, R.E., Brito, E.S., Silveira, M.R.R.S & Moura, C.F.H (2010b) Quality for fresh consumption and processing of some non-traditional tropical fruits from Brazil Fruits, 64, 361–370 Rufino, M.S.M., Alves, R.E., Brito, E.S., Pe´rez-Jime´nez, J., Saura- Calixto, F & Mancini Filho, J (2010a) Bioactive compounds and antioxidant capacities of 18 non-traditional tropical fruits from Brazil Food Chemistry, 121, 996–1002.

Sampaio, C.G., Morais, S.M., Rufino, M.S.M., Alves, R.E & Brito, E.S (2009) Quality, bioactive compounds, and antioxidant activity of Brazilian acerola clones Acta Horticulturae, 841, 463–466.

Sa´nchez-Moreno, C., Larrauri, J.A & Saura-Calixto, F (1998).

A procedure to measure the antiradical efficiency of nols Journal of the Science of Food and Agriculture, 76, 270– 276.

polyphe-Saura-Calixto, F., Garcia-Alonso, A., Gon˜i, I & Bravo, L (2000) In vitro determination of the indigestible fraction in foods: an alterna- tive to dietary fiber analysis Journal of Agricultural and Food Chemistry, 48, 3342–3347.

Scott, R.W (1979) Colorimetric determination of hexauronic acids in plant materials Analytical Chemistry, 51, 936–941.

Serrano, J., Gon˜i, I & Saura-Calixto, F (2007) Food antioxidant capacity determined by chemical methods may underestimate the physiological antioxidant capacity Food Research International, 40, 15–21.

Serrano, J., Puuponen-Pimia¨, R., Dauer, A., Aura, A.M & Calixto, F (2009) Tannins: current knowledge of food sources, intake, bioavailability and biological effects Molecular Nutrition and Food Research, 53, s310–s329.

Saura-Singleton, V.L., Orthofer, R & Lamuela-Ravento´s, R.M (1999) Analysis of total phenols and other oxidation substrates and antioxidants by means of Folin-Ciocalteu reagent Methods in Enzymology, 299, 152–178.

Strohecker, R & Henning, H.M (1967) Analisis de vitaminas: metodos comprobados Madrid: Paz Montalvo.

Vasco, C., Ruales, J & Kamal-Eldin, A (2008) Total phenolic compounds and antioxidant capacities of major fruits from Ecua- dor Food Chemistry, 111, 816–823.

Wong, J.M., De Souza, R., Kendall, C.W., Emam, A & Jenkins, D.J (2006) Colonic health: fermentation and short-chain fatty acids Journal of Clinical Gastroenterology, 40, 235–243.

antioxi-Please note: Wiley-Blackwell are not responsible forthe content or functionality of any supporting materialssupplied by the authors Any queries (other than missingmaterial) should be directed to the corresponding authorfor the article

Original article

Effect of calcium carbonate, calcium citrate, tricalcium phosphate, calcium gluconate and calcium lactate on some physicochemical properties of soymilk

Pattavara Pathomrungsiyounggul, Alistair S Grandison & Michael J Lewis*

School of Chemistry, Food and Pharmacy, The University of Reading, Whiteknights, PO Box 226, Reading, RG6 6AP, United Kingdom

(Received 29 January 2010; Accepted in revised form 22 July 2010)

Summary Soymilk fortified with 25 mm Ca (Ca carbonate, Ca citrate, triCa phosphate, Ca gluconate or Ca lactate) was

compared with the properties of unfortified soymilk (control) Calcium carbonate, Ca citrate and triCaphosphate did not significantly affect [Ca2+], absolute viscosity and particle size of soymilk, but Cagluconate and Ca lactate significantly increased these properties The pH of soymilk was significantlyincreased by adding Ca carbonate but significantly reduced by adding Ca gluconate and Ca lactate Drysediment of soymilk increased significantly with the addition of all Ca salts excluding triCa phosphate.Freezing point depression increased significantly only for Ca gluconate and Ca lactate, mainly owing to theirhigher solubility

Keywords Calcium, freezing point depression, particle diameter, pH, sediment, soymilk, viscosity.

Introduction

Soymilk contains much less calcium compared to cows’

milk, so calcium fortification of soymilk is now widely

practiced One issue of importance in its production is

its stability, both to calcium addition and to subsequent

heat treatment The pH of soymilk is about 6.4–6.7

Addition of calcium (Ca) salts to soymilk during the

production of Ca-fortified soymilk or tofu manufacture

reduces its pH Normally, soymilk does not coagulate

when pH is >6.0 (Yazici et al., 1997) When soymilk

pH was reduced, its ionic calcium concentration

increased (Pathomrungsiyounggul et al., 2009)

Many forms of Ca are generally recognised as safe by

US Food and Drug Administration, including Ca

carbonate, Ca chloride, Ca lactate, Ca phosphate, Ca

citrate and Ca gluconate (Fairweather-Tait & Teucher,

2002) One major difference in these salts relates to their

solubility Inorganic salts such as Ca carbonate and Ca

phosphate are cheaper, less soluble and mainly mixed

into solid foods, where solubility is not an issue, whereas

Ca citrate, Ca gluconate, Ca lactate are normally more

soluble and better absorbed (Fabri, 2004)

Calcium carbonate-fortified soymilk and triCa

phos-phate-fortified soymilk have been developed by

Chiw-anon et al (2000) They added 25 mm of these Ca salts

to raise the total Ca of soymilk to a level similar to cow’smilk, but visible precipitation of Ca took place Thisundesirable effect was inhibited by adding potassiumcitrate and carrageenan TriCa phosphate (25 mm Ca)was added to develop a pasteurised soy beverage,comparable to cow’s milk in its total calcium, withoutaffecting flavour or chalkiness (Weingartner et al.,1983) During cold storage for 10 days at 1C, boththe control sample (no added Ca) and Ca-fortifiedsample showed no significant differences in proteinstability, viscosity, flow behaviour index, consistencycoefficient, flavour and chalkiness

Calcium citrate is claimed to be moderately soluble(Gerstner, 2003) The addition of Ca citrate (25 mm Ca),

to improve nutritional profile, was found not toadversely affect the flavour or chalkiness of a sterilisedsoy beverage; it also had little effect on the consistencycoefficient and flow behaviour index (Weingartner et al.,1983) Calcium citrate used at concentration of 0.1–0.50% did not coagulate soymilk (Kamel & deMan,1982)

Calcium chloride is also used to fortify soymilk.However, its addition was found to coagulate soymilk(Lu et al., 1980; Shun-Tang et al., 1999; Pathomrungsi-younggul et al., 2007, 2009) Pasteurised Ca chloride-

coagulated; this could be prevented by adding chelatingagents (Pathomrungsiyounggul et al., 2007, 2009)

*Correspondent: E-mail: m.j.lewis@reading.ac.uk

Calcium lactate is well suited in full-flavoured

applica-tions because it has a neutral taste, very high

bioavail-ability, high solubility and high stability in solution

(Oort, 2004) Adding Ca lactate (14.4 mm Ca) caused

soymilk coagulation after sterilisation, but coagulation

did not occur when chelating agent was added

(Prabharaksa et al., 1989) Calcium gluconate shows

good solubility, a neutral taste and dissolves quickly at

higher temperatures (Gerstner, 2003) Again, soymilk

coagulated when more than 0.20% of Ca gluconate was

added but was prevented by the addition of chelating

agent (Rasyid & Hansen, 1991)

There is limited information on the role of calcium ion

concentration [Ca2+] in soymilk When it was increased

progressively in soymilk, it coagulated on pasteurisation

when its [Ca2+] was about 0.40 mm

(Pathomrungsi-younggul et al., 2007) The effectiveness of chelating

agents for reducing [Ca2+] was studied in Ca

gluconate-fortified soymilk (Rasyid & Hansen, 1991) and Ca

chloride-fortified soymilk (Pathomrungsiyounggul

et al., 2007, 2009) For Ca chloride-fortified soymilk, a

relationship between [Ca2+] and pH was obtained, and

it was found that soymilk pH decreased when [Ca2+]

increased (Pathomrungsiyounggul et al., 2007, 2009)

Thus, soymilk pH is an important property that

influences consumer acceptability as well as heat

stabil-ity The pH of soymilk was investigated at various

concentrations for several Ca salts (Lu et al., 1980;

Kamel & deMan, 1982) Another study (Weingartner

et al., 1983) compared pH of soymilk that was fortified

with the same Ca concentration (25 mm) of triCa

phosphate and mixtures of triCa phosphate and Ca

citrate

The particle diameter of Ca chloride-fortified

soymilk increased as [Ca2+] increased

(Pathomrungsi-younggul et al., 2007) The particle size of cereal milk

including soymilk should be 400–500 nm for a

favour-able mouth-feel (Saeseaw et al., 2005) For Ca

chlo-ride-fortified soymilk, the kinematic viscosity of

soymilk was constant when [Ca2+] was <0.25 mm,

but it increased as [Ca2+] of soymilk increased from

0.25 to 0.59 mm (Pathomrungsiyounggul et al., 2007)

Addition of 25 mm Ca as triCa phosphate or mixtures

of triCa phosphate and Ca citrate to soy beverage prior

to pasteurisation was performed by Weingartner et al

(1983) They reported that samples containing Ca

citrate had lower viscosity than samples without Ca

citrate Viscosity was the main property used by

Prabharaksa et al (1989) for evaluating their product

during the development of a formulation and process

for Ca-fortified soymilk One of their findings was that

added Ca lactate (0.05–0.125% w⁄ v) raised the

viscos-ity of soymilk Yazici et al (1997) formulated a Ca

lactogluconate-fortified soymilk and found that

viscos-ity was important for determining a suitable amount of

chelating agent

The amount of sediment provides an indication of itsstability Analysing this sediment was a good methodfor establishing the effectiveness of chelating agents(Rasyid & Hansen, 1991; Pathomrungsiyounggul et al.,2009) It was found that coagulated Ca chloride-fortifiedsoymilk had a higher dry sediment content thannoncoagulated sample and that dry sediment in nonco-agulated sample increased as [Ca2+] increased(Pathomrungsiyounggul et al., 2007)

Freezing point depression (FPD) of bulk milk ismeasured to detect added water to milk Measurement

of FPD of Ca-fortified soymilk has not been reported Itmay provide useful information about the low molecularweight soluble components in soymilk

This study aims to compare the effects of addition ofsome soluble and insoluble calcium salts on factorsrelated to quality and stability of soymilk, such as pH,particle diameter, absolute viscosity, dry sediment andFPD of soymilk

Materials and methods

Preparation of soymilkDried soybeans, Glycine max (product of Canada),were purchased from a supermarket in Reading, UK.De-ionised water was used throughout the experiment.One hundred gram of soybeans was washed in water.The cleaned beans were soaked in 1 kg of water for

14 h at 4C The swollen soybeans were drained andground with 1 kg of boiling water in a blender (ModelA707A; Kenwood MFG, Surrey, UK) at low speed for

5 min The slurry was heated for 10 min at 80C Thehot mixture was filtered through four layers of cheesecloth, and the filtrate (850 g) was collected Thesoymilk was cooled to room temperature and then kept

at 4C

Preparation of samplesSoymilk (200 mL) was heated to 45C on a heatedmagnetic stirrer, and Ca carbonate (CaCO3; Sigma-Aldrich, Gillingham, Dorset), Ca lactate ([CH3CH(OH)COO]2Ca Æ 0.5 H2O; Sigma-Aldrich), triCa phosphate(Ca3O8P2; Fluka, Buchs, Switzerland), Ca citrate (Ca3

(C6H5O7)2.4H2O; BDH Laboratory Supplies, Poole,UK) or Ca gluconate (C12H22CaO14.H2O; Fisher Scien-tific Ltd., Loughborough, UK) was added at concentra-tion of 25 mm Ca The mixtures were stirred on a heatedmagnetic stirrer for 10 min at 45C Then, Ca-fortifiedsoymilk (in a 250-mL glass bottle) was pasteurised at

72C for 15 s in a water bath-controlled temperature

at 75C and subsequently cooled to room temperature.The samples were kept at 4 C for 2 h to equilibrate.Control sample was the pasteurised soymilk with noadded Ca salts Samples were prepared in triplicate

Analysis of samples

Soymilk, control and Ca-fortified soymilk were analysed

as mentioned in the following paragraphs The analysis

of each property was carried out in triplicate

Calcium ion concentration

Determination of [Ca2+] was carried out using a Ca2+⁄

pH analyser (Model 634; Siemens Healthcare

Diagnos-tics, Frimley, UK), described by Lin et al (2006) The

mV value for standards and samples was obtained

directly from the analyser Five ionic Ca standards

(0.25–3.0 mm) were used to produce a calibration curve

(log [Ca2+] (mm) and mV) on each day of analysis

pH

A pH meter (Model SA 720; Orion Research Inc.,

Massachusetts, MA, USA) was used to measure pH of

samples at 25C

Absolute viscosity

Kinematic viscosity of samples at 25C was measured

with an Ostwald U-tube using method as described by

(Lewis, 1996) The density of samples at 25C was

analysed using a specific gravity bottle (Lewis, 1996)

Absolute viscosity was calculated using the following

equation:

Absolute viscosity¼ kinematic viscosity  density

Particle diameter

Particle diameter of diluted samples (100 lL sample in

3 mL water) was measured with a Zetasizer Nano-S

(Malvern Instruments Ltd., Malvern, UK)

Dry sediment

Dry sediment was measured as an indicator of stability

to calcium addition and pasteurisation An accurately

weighed (A g) sample (23 g) was added to a centrifugetube The tube was then centrifuged at 4200 rpm(2760 g) for 30 min in a centrifuge (Model Centaur 2;Fisons, Loughborough, UK) The supernatant wasdiscarded The pellet was transferred to a moisture dishand dried in a hot air oven at 105C to constant weight(B g) The following equation was applied to find % drysediment

% Dry sediment¼ B

A

 

 100Freezing point depression (FPD)

Samples were measured for FPD by Cryoscope (Model4L2; Advanced Instruments Inc., New Jersey, NJ,USA)

Statistical analysisStatistical analysis of data was performed by 1-wayanova using spss 15.0 for Windows Mean differenceswere analysed using Duncan’s multiple range test at

P£ 0.05

Results and Discussion

Calcium ion concentrationSoymilk prepared for this study contained [Ca2+]0.13 ± 0.10 mm In a previous study (Pathomrungsi-younggul et al., 2007), [Ca2+] in soymilk was

<0.02 mm The difference was probably due to ent batches of soybean used to produce soymilk Total

differ-Ca content of soybeans varied with soybean variety(Sun & Breene, 1991; Cai et al., 1997) Minerals insoybeans are influenced by location of growth andseason (Liu, 1999)

Table 1 shows [Ca2+] of control (pasteurised, tified soymilk) and Ca-fortified soymilk The [Ca2+] ofcontrol was higher than that of unheated soymilk,indicating that pasteurisation increased [Ca2+] of soy-milk In cows’ milk and ewes’ milk, heating decreased

unfor-Table 1 Properties of control and soymilk added with various types of Ca salts

Means ± standard deviation in the same row followed by the same letter are not significantly different (P > 0.05).

*Coagulated sample was examined visually after pasteurisation.

†For Ca lactate, addition was 4% lower.

[Ca2+] (Mittal et al., 1990; De La Fuente et al., 1998).

The reason why ionic calcium increases following

pasteurisation of soymilk is not clear and is worthy of

further investigation The slight decrease found for

cow’s milk samples is attributed to calcium phosphate

precipitation Adding Ca carbonate, Ca citrate and

triCa phosphate did not significantly affect [Ca2+] of the

control These samples did not coagulate and had

[Ca2+] less than 0.40 mM, which was in agreement with

our previous conclusion (Pathomrungsiyounggul et al.,

2007) that controlling [Ca2+] below 0.40 mm prevented

coagulation in Ca-fortified soymilk In heated soymilk,

each protein molecule or its polymer remains soluble

because they are negatively charged, which causes

electrostatic repulsive force between the molecules

(Fukushima, 1991) In the presence of added Ca salt,

the negatively charged soy proteins bind with Ca ions

(Kohyama et al., 1995), thereby reducing their charge

and making them more susceptible to heat-induced

coagulation Shun-Tang et al (1999) reported that the

addition of Ca at lower levels scarcely reduces soymilk

protein solubility and so it did not coagulate

Addition of Ca gluconate and Ca lactate led to a

significant (P£ 0.05) increase in [Ca2+] when

com-pared to the control and other Ca-fortified soymilk

The maximum [Ca2+] was observed in Ca

gluconate-fortified soymilk, and its [Ca2+] was significantly

(P£ 0.05) higher than Ca lactate-fortified soymilk

The [Ca2+] in Ca gluconate-fortified soymilk and Ca

lactate-fortified soymilk were 18 and 14 times higher,

respectively, than that in control This is explained by

their higher solubilities Both Ca-fortified soymilk

coagulated on pasteurisation, and their [Ca2+] was

higher than 0.40 mm Gerstner (2003) has reported that

with higher soluble salts, more free Ca ions are in

solution and available for reaction than those with

lower solubility He showed that Ca solubilities at

25C for Ca gluconate and Ca lactate were 3.5 and

9.3 g L)1, respectively, whereas Ca carbonate was

insoluble Soymilk coagulation was reported to occur

because Ca ions neutralise the charge on the protein

molecules, resulting in the aggregation of the protein

molecules by hydrophobic interactions (Kohyama

et al., 1995)

Cow’s milk is considered to be a rich source of Ca

Ionic Ca levels in cow’s milk have been reported by

several researchers Cow’s milk at different stages of

lactation contained 1.43–2.50 mm [Ca2+] (Lin et al.,

2006) In middle lactation, it was reported to be

2.3 mm (De La Fuente et al., 1998) and reconstituted

skimmed milk was found to be 2.04 mm (Sievanen

et al., 2008) In other milk, ionised Ca in human milk

varied between 2.3 and 4.0 mm across individuals at

90 days of lactation (Neville et al., 1994) and ewes’

milk in the middle lactation had 2.8 mm (De La

Fuente et al., 1998) Thus, all noncoagulated soymilk

samples in this study contained considerably lower[Ca2+], (>4.6 times), compared to these mammalianmilk types

pHThe prepared soymilk had pH 6.59 ± 0.01 Soymilk pHhas been reported by several investigators and a widerange has been found: values of 6.0 (Nsofor &Maduako, 1992), 6.55–6.66 (Poysa & Woodrow, 2002)and 6.6 (Forster & Ferrier, 1979) They all preparedsoymilk by different methods In our previous study(Pathomrungsiyounggul et al., 2007), soymilk pH was6.73 Factors affecting soymilk pH are soybeans geno-type, growing location and growing year (Poysa &Woodrow, 2002) and soybeans storage condition (rela-tive humidity, temperature and time) (Kong et al., 2008;Liu & Chang, 2008)

Table 1 shows pH of control and Ca-fortified milk Pasteurisation did not change soymilk pH in thisstudy More severe heat treatment, such as sterilisation,was reported to reduce soymilk pH from 7.5 to 6.0(Nsofor & Maduako, 1992) There was an insignificantreduction in pH of control compared to Ca citrate andtriCa phosphate-fortified soymilk (Table 1) In the case

soy-of pasteurised soy beverage containing sugar andstored at 1 C for 10 days, the control (no addedCa), triCa phosphate-fortified soymilk (added 25 mmCa) and soymilk fortified with 20 mm Ca as Ca citrateplus 5 mm Ca as triCa phosphate had pH values of7.55, 7.63 and 7.35, respectively (Weingartner et al.,1983) This is higher than the range of valuesmentioned earlier

Addition of Ca carbonate significantly increased pH

of the control (P£ 0.05), and it had the highest pHvalue (Table 1) Lu et al (1980) found that soymilk with

a pH 6.40 showed no change in pH on addition of

30 mm Ca carbonate The lowest pH was found when

Ca gluconate and Ca lactate were added (Table 1).These data indicated that soymilk pH varied depending

on type of Ca salt This study agreed with the vation of Lu et al (1980) that soymilk began tocoagulate when the pH dropped to about 6.0 Thus,

obser-pH is an important parameter affecting heat stability,although it has not been investigated in detail in thispaper The salts selected fell into two categories, the saltswith poor solubility which have very little effect on pHand the soluble salts which significantly reduce pH

Absolute viscosityPrepared soymilk had absolute viscosity 1.87 ± 0.02 cP,which was lower than that of control which was thepasteurised soymilk (Table 1) Previous reports gave theviscosity of soymilk before and after sterilisation as 2.83and 2.63 cP, respectively (Prabharaksa et al., 1989)

Factors reported to influence soymilk viscosity include

processing method, storage time and storage

tempera-ture (Iwuoha & Umunnakwe, 1997), and total solid

content (Oguntunde & Akintoye, 1991) The absolute

viscosities of control and Ca-fortified soymilk are

presented in Table 1 Addition of Ca carbonate, Ca

citrate or triCa phosphate did not significantly change

the absolute viscosity of control Ca citrate-fortified

soymilk showed a significantly (P£ 0.05) higher

abso-lute viscosity than Ca carbonate-fortified soymilk and

triCa phosphate-fortified soymilk, but significantly

(P£ 0.05) lower than Ca gluconate-fortified soymilk

and Ca lactate-fortified soymilk Weingartner et al

(1983) measured viscosity of several pasteurised soy

beverages after 10 days storage at 1C and found that

the control (unfortified) and the beverage containing

triCa phosphate 25 mm Ca had a viscosity of 5.7 cP,

whereas the beverage containing Ca citrate (20 mm) plus

triCa phosphate (5 mm) had a viscosity of 4.9 cP Our

study showed that there was no significant difference in

absolute viscosity between the Ca carbonate-fortified

sample and triCa phosphate-fortified sample, whereas

their viscosity was significantly lower (P £ 0.05) than

sample made with Ca gluconate and Ca lactate

(Table 1) It was clear that the absolute viscosity of

control was significantly (P£ 0.05) lower than soymilk

with added Ca gluconate and Ca lactate Calcium

gluconate-fortified soymilk had a significantly lower

(P£ 0.05) absolute viscosity than Ca lactate-fortified

soymilk (Table 1) Ranjan et al (2005) reported that Ca

fortification (12.5 mm Ca) of buffalo milk with

gluco-nate and lactate salts led to a significant increase in

viscosity of that milk They showed that viscosities of

unfortified buffalo milk, Ca gluconate-fortified buffalo

milk and Ca lactate-fortified buffalo milk were 1.99, 2.16

and 2.10 cP, respectively It was reported that the

sterilised Ca lactate-fortified soymilk (14.4 mm Ca)

containing 0.2% w⁄ v sodium citrate as chelating agent

had viscosity 3.0 cP (Prabharaksa et al., 1989), which

was lower than absolute viscosity of Ca lactate-fortified

soymilk produced in this study

This study indicated that ‘insoluble’ Ca salts did not

change soymilk absolute viscosity, but soluble salt

addition led to soymilk becoming significantly more

viscous This could have arisen because of the increase in

particle size that was observed

Particle diameter

Particle size is one parameter affecting food texture and

taste The prepared soymilk had a particle diameter

328 ± 13 nm The mean particle size was measured for

three commercial soymilk samples and found to be 380,

420 and 440 nm (Saeseaw et al., 2005) Soymilk is an

aqueous extract, and one of its main solid components is

protein The size distribution of proteins in heated

soymilk was investigated by Ono et al (1991), and theyfound three protein particle fractions of >100 nm, 100–

40 nm and <40 nm The first two fractions constituted50% of the protein in heated soymilk Oil globules areanother component of soymilk, and they have a meandiameter as big as 350 nm (Odo, 2003)

Table 1 shows particle diameter of control andCa-fortified soymilk Fortification of soymilk with Cacarbonate, Ca citrate and triCa phosphate did notsignificantly change the particle diameter of control.However, particle diameter of Ca carbonate-fortifiedsoymilk was significantly (P£ 0.05) lower than that of

Ca citrate-fortified soymilk The two samples thatshowed coagulation had a very large particle size(>1000 nm) Chocolate milk and fresh full-fat milkhave been reported to have mean particle sizes between

500 and 530 nm (Saeseaw et al., 2005) Thus, all agulated Ca-fortified soymilk in this study had a smallerparticle diameter than these bovine milk samples

nonco-Dry sedimentSediment measurement is considered to be a goodindicator of heat stability of a product The preparedsoymilk contained a dry sediment of 0.36 ± 0.02%.Such a low dry sediment level indicates a high level ofstability for Ca-fortified soymilk, whilst values of about5% were indicative of coagulation in Ca-fortifiedsoymilk (Pathomrungsiyounggul et al., 2007) Table 1shows dry sediment content in control and Ca-fortifiedsoymilk The control contained the lowest level of drysediment and was not significantly different from triCaphosphate-fortified soymilk Addition of other Ca saltssignificantly increased (P£ 0.05) dry sediment levelwhen compared to the control, although increases weresmall for Ca carbonate and Ca citrate additions Thetwo samples that coagulated had the same content ofdry sediment (about 4%) and its dry sediment contentwas the highest, compared to other samples Theamounts of dry sediment of these two samples wereslightly less than those in coagulated soymilk, to which

25 mm Ca chloride was added and pasteurised, in thestudy of Pathomrungsiyounggul et al (2007), whichcontained 4.88% dry sediment

Freezing point depressionThe FPD of this particular batch of soymilk was

149 ± 2 moC, which is influenced by the concentration

of low molecular weight water-soluble solutes (Lerici

et al., 1983) Soymilk contains soluble constituents such

as protein, sugar, oligosaccharide and minerals (Odo,2003), which will contribute towards FPD, according totheir concentrations and molecular weights Table 1shows FPD of the control and Ca-fortified soymilk.Addition of Ca carbonate and triCa phosphate did not

affect significantly FPD, compared to that of the

control In water and soymilk, Ca carbonate and triCa

phosphate have a poor solubility (Gerstner, 2003); thus,

they did not change FPD The control and these two

samples had significantly (P£ 0.05) lower FPD than the

other samples The increase in FPD was significant

(P£ 0.05) for other Ca salts and was as follows: Ca

citrate < Ca lactate < Ca gluconate Calcium citrate,

Ca gluconate and Ca lactate are reported to have Ca

solubilities of 0.2, 3.5 and 9.3 g L)1, respectively, in

water at 25C (Gerstner, 2003) The highest FPD

(From Table 1) observed in soymilk was fortified by Ca

gluconate, which was still much lower than FPD of

mammalian milk FPD is worthy of further

investiga-tion as an indicator of soymilk composiinvestiga-tion, although

this has not been reported A wide range of FPD was

observed for different test formulations of soymilk,

ranging from 110 to over 400 moC (results not shown)

This results from the different ingredients used in the

formulation, such as soy flours, concentrates or isolates

and any other added compounds, e.g sugars or

stabil-isers FPD might provide a useful measure of product

consistency on a day-to day basis It could also be useful

to measure changes associated with binding of low

molecular weight components in soy bean to its protein

fraction The wide range in FPD that can be found for

soymilk is interesting, especially in comparison with

milk of animal origin, where FPD is controlled by

biological mechanisms to within a narrow range This

makes FPD as a useful test for detecting adulteration of

these milk samples with water, which is not the case for

soymilk An increase in milk FPD may arise because of

decrease in pH (Mittal et al., 1990)

Relationships between some of these properties

Relationships between some properties of soymilk

samples were obtained (Figs 1 and 2) Because Ca

carbonate increased soymilk pH, whereas the other

Ca salts decreased soymilk pH; the properties of Ca

carbonate-fortified soymilk were excluded from these

relationships For these other calcium salts, it was found

that an increase in pH led to a decrease in dry sediment

of samples (Fig 1) Freezing point depression appears

to increase as [Ca2+] increases (Fig 2), probably due tothe lower molecular weight of the calcium ions Aprevious study has also shown that lowering pH ofsoymilk will lead to an increase in ionic calcium and hashighlighted changes taking place

Conclusion

This study has shown that properties of soymilk weredependent on type of added Ca salts Calcium gluconateand Ca lactate produced a significant increase in [Ca2+],which bound protein molecules leading to coagulation,

as well as a significant reduction in pH, and significantincreases in absolute viscosity and FPD Addition of Cacarbonate, Ca citrate and triCa phosphate did notsignificantly affect [Ca2+], resulting in no coagulationand no change in absolute viscosity and particle size ofsoymilk Addition of Ca carbonate caused a significantincrease in pH, but addition of Ca citrate and triCaphosphate showed a small decrease in pH Dry sediment

of control was increased dramatically by the addition of

Ca gluconate and Ca lactate and slightly by the addition

of Ca carbonate, Ca citrate and triCa phosphate.Adding Ca gluconate and Ca lactate significantlyincreased FPD of soymilk It is proposed that theconsistency and stability of calcium-fortified soymilkcan be better controlled by monitoring pH, ioniccalcium and FPD

ReferencesCai, T.D., Chang, K.C., Shih, M.C., Hou, H.J & Ji, M (1997) Comparison of bench and production scale methods for making soymilk and tofu from 13 soybean varieties Food Research International, 30, 659–668.

Chiwanon, P., Puwastien, P., Nitithamyoung, A & Sirichakwat, P.P (2000) Calcium fortification in soybean milk and In Vitro bioavail- ability Journal of Food Composition and Analysis, 13, 319–327.

De La Fuente, M.A., Olano, A & Jua´rez, M (1998) Effects of heat treatments and subsequent storage on the mineral balance of ewes’ milk Journal of Dairy Research, 65, 457–464.

Figure 2 Relationship between [Ca2+] and FPD.

Fabri, D (2004) Fortification of beverages with calcium Soft Drinks

International, February, 24–25.

Fairweather-Tait, S.J & Teucher, B (2002) Iron and calcium

bioavailability of fortified foods and dietary supplements Nutrition

Reviews, 60, 360–367.

Forster, L.L & Ferrier, L.K (1979) Viscometric characteristics of

whole soybean milk Journal of Food Science, 44, 583–585, 590.

Fukushima, D (1991) Recent process of soybean protein foods:

chemistry, technology, and nutrition Food Reviews International, 7,

323–351.

Gerstner, G (2003) How to fortify beverages with calcium Drink

Technology & Marketing, November, 8–11.

Iwuoha, C.I & Umunnakwe, K.E (1997) Chemical, physical and

sensory characteristics of soymilk as affected by processing method,

temperature and duration of storage Food Chemistry, 59, 373–379.

Kamel, B.S & deMan, J.M (1982) Composition and properties of

bean curd made from Ontario soybeans Canadian Institute of Food

Science and Technology, 15, 273–276.

Kohyama, K., Sano, Y & Doi, E (1995) Rheological characteristics

and gelation mechanism of tofu (soybean curd) Journal of

Agricul-tural and Food Chemistry, 43, 1808–1812.

Kong, F., Chang, S.K.C., Liu, Z & Wilson, L.A (2008) Changes of

soybean quality during storage as relate to soymilk and tofu making.

Journal of Food Science, 73, S134–S144.

Lerici, C.R., Piva, M & Rosa, D (1983) Water activity and freezing

point depression of aqueous solutions and liquid foods Journal of

Food Science, 48, 1667–1669.

Lewis, M.J (1996) Physical properties of foods and food processing

systems England: Woodhead Publishing Ltd.

Lin, M.-J., Lewis, M.J & Grandison, A.S (2006) Measurement of

ionic calcium in milk International Journal of Dairy Technology,

59, 192–199.

Liu, K (1999) Soybeans: chemistry, technology, and utilization USA:

Aspen Publishers, Inc.

Liu, Z.S & Chang, S.K.C (2008) Optimal coagulant concentration,

soymilk and tofu quality as affected by a short-term model storage of

proto soybeans Journal of Food Processing Preservation, 32, 39–59.

Lu, J.Y., Carter, E & Chung, R.A (1980) Use of calcium salts for

soybean curd preparation Journal of Food Science, 45, 32–34.

Mittal, S.B., Hourigan, J.A & Zadow, J.G (1990) Effect of added

sodium hexametaphosphate on certain technological aspects of

UHT recombined milk The Australian Journal of Dairy Technology,

45, 1–4.

Neville, M.C., Keller, R.P., Casey, C & Allen, J.C (1994) Calcium

partitioning in human and bovine milk Journal of Dairy Science, 77,

1964–1975.

Nsofor, L.M & Maduako, O (1992) Stabilized soymilk for ambient

tropical storage: a preliminary report International Journal of Food

Science and Technology, 27, 573–576.

Odo, T (2003) Soy (soya) milk In: Encyclopedia of Food Science and Nutrition (edited by B Caballero, L.C Trugo & P.M Finglas).

pp 5403–5405 London, UK: Academic Press.

Oguntunde, A.O & Akintoye, O.A (1991) Measurement and comparison of density, specific heat and viscosity of cow’s milk and soymilk Journal of Food Engineering, 13, 221–230.

Ono, T., Choi, M.R., Ikeda, A & Odagiri, S (1991) Changes in the composition and size distribution of soymilk protein particles by heating Agricultural and Biological Chemistry, 55, 2291–2297 Oort, S (2004) Calcium fortification of beverages Fruit Processing,

14, 33–35.

Pathomrungsiyounggul, P., Grandison, A.S & Lewis, M.J (2007) Effects of calcium chloride and sodium hexametaphosphate on certain chemical and physical properties of soymilk Journal of Food Science, 72, E428–E434.

Pathomrungsiyounggul, P., Grandison, A.S & Lewis, M.J (2009) Effects of calcium-chelating agents and pasteurisation on certain properties of calcium-fortified soymilk Food Chemistry, 118, 808– 814.

Poysa, V & Woodrow, L (2002) Stability of soybean seed sition and its effect on soymilk and tofu yield and quality Food Research International, 35, 337–345.

compo-Prabharaksa, C., Olek, A.C & Steinkraus, K.H (1989) Enrichment of soybean milk with calcium Acta Biotechnology, 9, 9–16.

Ranjan, P., Arora, S., Sharma, G.S., Sindhu, J.S., Kansal, V.K & Sangwan, R.B (2005) Bioavailability of calcium and physicochemi- cal properties of calcium-fortified buffalo milk International Journal

Shun-Tang, G., Ono, T & Mikami, M (1999) Incorporation of soy milk lipid into protein coagulum by addition of calcium chloride Journal of Agricultural and Food Chemistry, 47, 901–905.

Sievanen, K., Huppertz, T., Kelly, A.L & Fox, P.F (2008) Influence

of added calcium chloride on the heat stability of unconcentrated and concentrated bovine milk International Journal of Dairy Technology, 61, 151–155.

Sun, N & Breene, W.M (1991) Calcium sulfate concentration influence on yield and quality of tofu from five soybean varieties Journal of Food Science, 56, 1604–1607, 1610.

Weingartner, K.E., Nelson, A.I & Erdman, J.W.J (1983) Effects of calcium addition on stability and sensory properties of soy beverage Journal of Food Science, 48, 256–257, 263.

Yazici, F., Alvarez, V.B., Mangino, M.E & Hansen, P.M.T (1997) Formulation and processing of a heat stable calcium-fortified soy milk Journal of Food Science, 62, 535–538.

Original article

A novel emulsion coating and its effects on internal quality and

shelf life of eggs during room temperature storage

Damir D Torrico1, Wannita Jirangrat1, Hong Kyoon No2, Witoon Prinyawiwatkul1*, Beilei Ge1& Dennis Ingram3

1 Department of Food Science, Louisiana State University Agricultural Center, Baton Rouge, LA 70803-4200, USA

2 Department of Food Science and Technology, Catholic University of Daegu, Hayang 712-702, Republic of Korea

3 School of Animal Sciences, Louisiana State University Agricultural Center, Baton Rouge, LA 70803-4200, USA

(Received 27 May 2010; Accepted in revised form 26 July 2010)

ratios) as coating materials in preserving internal quality of eggs were evaluated during a 5-weeks storage at

25C Consumers (n = 109) evaluated surface properties and purchase intent of freshly coated eggs Asstorage time increased, Haugh unit and yolk index values decreased whereas weight loss increased.Noncoated eggs rapidly changed from AA to B and C grades after 1 and 3 weeks, respectively However, allemulsion-coated eggs maintained their A-grade quality for 4 weeks Compared with noncoated eggs, allemulsion coatings reduced weight loss of eggs by at least seven times (0.88–1.03% vs 7.14%) OnlyMO:CH = 25:75 emulsion-coated eggs were not sensorially glossier than noncoated eggs All emulsion-coated eggs had >80% positive purchase intent and were negative for Salmonella This study demonstratedthat MO:CH emulsion coatings preserved internal quality and prolonged shelf life of eggs

Keywords Chitosan, egg quality, emulsion coating, Haugh unit, mineral oil, Salmonella, sensory discrimination, yolk index.

Introduction

Eggs are an excellent source of high quality protein

(Watkins, 1995) Furthermore, eggs hold important

functional properties such as coagulation, solidification,

aeration, emulsification, coloration and texturisation

(Stadelman, 1999) According to USDA, the production

of table eggs in the United States in 2008 was 6.40

billion dozen with a value of approximately 8.22 billion

dollars Of this production volume, broken eggs totalled

2.05 billion dozen that represent an increment of 2.54%

compared with that in 2007 (USDA, 2010) In addition

to the breakage of shell eggs, interior quality

deterior-ation and microbial contamindeterior-ation during storage cause

a serious economic loss to the poultry industry

(Stad-elman, 1995b; Wong et al., 1996) Certain

microorgan-isms such as Salmonella enterica serovar Enteritidis and

Salmonella entericaserovar Typhimurium that exist on

the shell surface can penetrate into the interior of eggs

and contaminate the internal content (Padron, 1990;

Berrang et al., 1999) During storage, the loss of

moisture and carbon dioxide via the shell pores causes

quality changes in albumen and yolk as well as weight

loss of eggs (Stadelman, 1995b) Thus, a protectivebarrier against the loss of moisture and carbon dioxidethrough the shell is necessary to preserve the egg quality.Low temperature refrigeration is considered the singlemost important treatment for preserving eggs In UnitedStates, eggs are required to be refrigerated at 45F(7 C) or below Nonetheless, in some developingregions of the world where refrigeration of eggs isseldom practiced, coating of eggs is thus an alternativeway to preserve the internal quality

Various coating materials have been applied to thesurface of egg shells for preserving the internal quality ofeggs These include synthetic polymers (Meyer &Spencer, 1973), polysaccharides (Xie et al., 2002; Bhale

et al., 2003; No et al., 2005; Kim et al., 2006; Caner &Cansiz, 2008), proteins (Herald et al., 1995; Cho et al.,2002; Xie et al., 2002; Rhim et al., 2004) and oils(Knight et al., 1972; Kamel et al., 1980; Obanu &Mpieri, 1984; Waimaleongora-Ek et al., 2009) Chitosan

is a natural biopolymer derived by deacetylation ofchitin, a major component of shells in crustacean such ascrab, shrimp and crawfish Chitosan generates a semi-permeable coating that modifies the internal atmosphereand decreases transpiration rates in food products(Nisperos-Carriedo, 1994) Recent studies (Bhale et al.,2003; No et al., 2005; Kim et al., 2007, 2008) revealed

*Correspondent: Fax: +1 225 578 5300;

e-mail: wprinya@lsu.edu

doi:10.1111/j.1365-2621.2010.02396.x

that chitosan coating preserved the internal quality of

eggs and extended the shelf life for at least 3 weeks

longer than noncoated eggs at 25C Butler et al (1996)

reported that chitosan films are efficient barriers against

permeation of oxygen but act as low water barriers

because of their strong hydrophilic properties

Oil is another coating material currently used to

preserve the internal quality of eggs Waimaleongora-Ek

et al.(2009) reported that coating with mineral oil (MO)

(26 mPa s) reduced the weight loss of eggs by more than

ten times (0.85% vs 8.78%) and extended the shelf life

of eggs by at least three more weeks compared with

noncoated eggs during 5 weeks of storage at 25C

Homler & Stadelman (1963) also proved that oil-coated

eggs had higher Haugh units and lower weight loss than

noncoated eggs after 3 weeks of storage at 22C Oils

used for egg coating must be odourless, colourless and

free of fluorescent materials (Stadelman, 1995b)

How-ever, shell colour and visual appearance of eggs may be

altered by oil used as a coating material Wong et al

(1996) reported that eggshells coated with MO possessed

a higher L* value (lightness) than noncoated eggs (87.05

vs 83.90), possibly because of glossier surface

A problem associated with MO coating is that oil

dries very slowly (1 day or longer without forced-air

blowing) compared with chitosan solution (CH)

(<15 min) when applied on the surface of the eggshell

without wiping it dry Thus, coating of eggs with

emulsion of MO and CH may considerably reduce the

drying time However, the emulsion may act differently

as a protective barrier against transfer of moisture and

carbon dioxide through the shell surface of eggs,

compared with MO and chitosan To date, there is no

information available on the effects of emulsion of MO

and CH on the internal quality and shelf life of eggs

during room temperature storage

The objectives of the present research were to evaluate

the effects of MO, CH and their three emulsions

(MO:CH = 75:25, 50:50 and 25:75 ratios) as coating

materials in preserving the internal quality (weight loss,

Haugh unit, yolk index, albumen pH) of coated eggs

during 5 weeks of storage at 25C and to evaluate

consumer perception and purchase intent of freshly

coated eggs by a sensory discrimination test Total plate

count (TPC) and Salmonella detection of the coated eggs

were also evaluated before and after 5 weeks of storage

Materials and methods

Materials

MO (viscosity = 26 mPa s; transparent, odourless and

food-grade) was obtained from Penreco (Karns city,

PA, USA) Chitosan (molecular weight = 223 kDa),

acid soluble and white-coloured powder prepared from

crab leg shell, was purchased from Biotech (Mokpo,

Korea) Emulsifier Tandem552K (a mixture of glycerides and diglycerides, polysorbate, water andpropyl gallate) was obtained from Caravan ingredients(Lenexa, KS, USA) Unwashed, faeces-free, white-shelleggs (from 51-week-old, Hyline W-36 hens; a weightrange of 50–70 g) were obtained from Cal-Maine Foods(Jackson, MS, USA) Immediately after collected fromthe farm and screened for defects and desirable weightrange, eggs were stored in the cold room (approximately

mono-7C) before the next day coating Before coating, eggswere kept at room temperature (approximately 25C)for 2 h to avoid water condensation on the egg surfacethat could interfere with coating

Preparation of mineral oil⁄ chitosan solution emulsionsChitosan coating solution was prepared by dissolvingchitosan in 1% (v⁄ v) acetic acid at 2% (w ⁄ v) concen-tration (Kim et al., 2009) Three emulsions of MO and

CH were prepared by adding 1% of emulsifier Tandem552K to three different ratios of MO and CH(MO:CH = 75:25, 50:50 and 25:75) and mixing using

an ultrasonic processor (CPX 500, Cole Parmer, VernonHills, IL, USA) for 8 min at 10C The CH solutionand all emulsions were prepared on the day of thecoating experiment

Coating treatment and storage of eggsEggs were individually weighed with a balance (TS400;Ohaus Corp., Florham Park, NJ, USA), coated with

MO, CH or one of the three emulsions using a spongebrush, and dried overnight at room temperature(25 ± 2C) Seven coating treatments were evaluatedthroughout the storage period: Control = noncoatedeggs; MO (U) = unwiped after coating with 100% MO;

MO (W) = wiped after coating with 100% MO; 75:25MO:CH = coating with MO:CH emulsion at a ratio of75:25; 50:50 MO:CH = coating with MO:CH emulsion

at a ratio of 50:50; 25:75 MO:CH = coating withMO:CH emulsion at a ratio of 25:75; CH = coatingwith 100% CH All eggs (fifty eggs⁄ treatment) wereplaced in a small-end down position (Kim et al., 2009)

in cardboard egg racks and stored at room temperature(25 ± 2C) and averaged 60–65% RH for 5 weeks.Ten eggs per each treatment were taken at 1-weekintervals for determination of weight loss, Haugh unit,yolk index and albumen pH

Determination of weight lossWeight loss (%) of the coated whole egg during storagewas calculated as {[initial whole egg weight (g) aftercoating at day 0) whole egg weight (g) after storage] ⁄initial whole egg weight (g) after coating at day 0}· 100.Weight loss (%) of the control noncoated whole egg was

calculated as {[initial whole egg weight (g) at day

0) whole egg weight (g) after storage] ⁄ initial whole egg

weight (g) at day 0}· 100 The weight of whole eggs was

measured with a balance (TS400S; Ohaus Corp.) Ten

measurements per treatment were taken

Determination of Haugh unit and yolk index

The height of albumen and yolk was measured with a

tripod micrometer (Model S-6428; B.C Ames Inc.,

Melrose, MA, USA) The yolk width was measured with

a digital caliper (General Tools & Instruments, New

York, NY, USA) The Haugh unit was calculated as 100

log (H) 1.7 W0.37

+ 7.57), where H is the albumenheight (mm) and W is the weight (g) of egg (Haugh,

1937) The yolk index was calculated as yolk height⁄ yolk

width (Stadelman, 1995a; Lee et al., 1996) Ten

mea-surements per treatment were taken

Measurement of albumen pH

After measurement of Haugh unit and yolk index, the

albumen was separated from the yolk The thin and

thick albumen were mixed thoroughly prior to

measur-ing pH with a pH meter (IQ150; IQ Scientific

Instru-ments, San Diego, CA, USA) Ten measurements per

treatment were taken

Sensory discrimination and consumer purchase intent

Consumers (n = 109) were recruited from Baton

Rouge, Louisiana, to participate in the sensory

discrim-ination of the coated eggs [with MO (unwiped or wiped),

CH and⁄ or one of the three emulsions] compared with

the control noncoated eggs at day 0 Consumers were

first presented with the labelled control egg, followed by

six unlabelled coated eggs and one unlabelled control (to

ascertain the ‘noise’ level) The unlabelled eggs were

individually compared to the labelled control for

spec-ified attributes For surface glossiness, consumers were

asked to indicate whether the unlabelled coated and

unlabelled control eggs were perceived as ‘more,’ ‘the

same’ or ‘less’ in the specified attribute compared with

that of the labelled control, and whether they were ‘sure’

or ‘unsure’ about their decision; in this case, as the

direction of a given attribute was of interest, the bipolar

R-index was used For surface odour and colour, and

overall surface appearance, consumers were asked

whether the unlabelled coated and unlabelled control

eggs were ‘different from’ or ‘the same as’ the labelled

control, and whether their decision was ‘sure’ or

‘unsure’; in this case, as the direction of a given attribute

was not measured, the unipolar R-index was used The

test protocol followed that of Bhale et al (2003)

Consumers self-paced their evaluation (but not longer

than 30 min.) Afterwards, these consumers evaluated

purchase intent for all eggs on a yes⁄ no scale andreported as % positive purchase intent

Microbiological analysisThe control noncoated eggs and eggs coated with MO(unwiped or wiped), CH and one of the three emulsionswere analysed for TPC and Salmonella at day 0 andafter 5 weeks of storage The internal content (yolk andalbumen) of egg samples was homogenised using astomacher (STO-400; Tekmar Company, Cincinnati,

OH, USA) in a dilution of 1:10 of 0.1% bufferedpeptone water (BD Difco, Sparks, MD, USA) ForTPC, viable cells (CFU⁄ g of egg) were enumerated onplate count agar (PCA) (BD Difco) by the pour plateand spread plate methods followed by incubation at

35C for 24 h For Salmonella spp detection, genates of egg samples were enriched by using Tetra-thionate broth (BD Difco) and incubated at 35 C for

homo-24 h Following enrichment, subcultures were platedonto XLT4 agar (BD Difco) at 35 C for 24 h prior todetection All microbiological assays were done induplicate for each treatment

Statistical analysisFor internal quality (weight loss, Haugh unit, yolk indexand albumen pH) of eggs, mean ± standard deviationvalues were reported based on ten measurements (eggs)per treatment Data were analysed using Analysis ofVariance, followed by the Tukey’s studentized range test(a = 0.05) using the SAS software (SAS, 2003)

The data obtained from the sensory discriminationtest were converted into frequency counts, and then theR-index was calculated for each attribute and expressed

as a percentage of sensory discrimination The bipolarR-index for surface glossiness and the unipolar R-indexfor odour, colour and overall surface appearance werecomputed from the equations as in Bhale et al (2003).The significance of the R-index was determined usingthe table provided by Bi & O’Mahony (2007) At thesignificance of 5%, the observed R-index value wassignificant if it exceeded the critical R-index of 56.65%for the unipolar R-index test For the bipolar R-index,the result was significant if it exceeded the criticalR-index of 57.89% for R-index more or fell behind thecritical R-index of 42.11% for R-index less

Results and discussion

Effects of mineral oil, chitosan solution and their emulsions

as a coating material on weight lossDifferences in the weight loss among the controlnoncoated eggs and those coated with MO, CH and⁄ ortheir three emulsions (MO:CH = 75:25, 50:50, and

25:75) were found (interaction between coating

treat-ments· storage periods, P < 0.0001) during 5 weeks of

storage at 25C (Table 1) Overall, the weight loss

progressively increased with increased storage periods

Without exception, all eggs coated with MO (unwiped

or wiped) and⁄ or three MO:CH emulsions had

signif-icantly (P < 0.05) lesser weight loss than noncoated

and CH-coated eggs throughout the 5 weeks of storage

period However, there were no significant differences

(P > 0.05) in weight loss observed among five

treat-ment groups of eggs coated with MO (unwiped or

wiped) and three emulsions throughout 5 weeks of

storage After 5 weeks, eggs coated with MO (unwiped

or wiped) and⁄ or three emulsions had at least seven

times lesser weight loss (%) compared with that of the

control eggs (0.69–1.03% vs 7.14%)

Evaporation of water and, to a much lesser extent,

loss of CO2from the albumen through the shell lead to

overall weight loss of the whole egg (Obanu & Mpieri,

1984) Table 1 shows that the weight loss of eggs coated

with MO (unwiped or wiped) (0.69–0.70%) and three

emulsions (0.88–1.03%) after 5 weeks of storage was

lower than that (1.43%) of noncoated eggs after 1 week

of storage Similarly, Waimaleongora-Ek et al (2009)

reported that, at 25C storage, the weight loss (0.85%)

of eggs coated with MO (wiped; 26 mPa s) after

5 weeks was lower than that (1.97%) of noncoated

eggs after 1 week Moreover, Obanu & Mpieri (1984)

reported that vegetable oil coatings significantly

re-duced (eleven times less) the weight loss (0.013–0.016 g)

of coated eggs, compared to that (0.186 g) of noncoated

eggs after 35 days of storage at 25–32C Slight

differences in weight loss among these studies may be

because of different coating materials used, storage

period and temperature, egg size and shell porosity

(Muller, 1958)

In our present study (Table 1), no significant

differ-ence (P > 0.05) in weight loss was observed between

noncoated (7.14%) and CH-coated (6.82%) eggs after

5 weeks These values are similar to those reported fornoncoated (7.84%) and CH-coated (6.69–7.66%) eggsafter 5 weeks of storage at 25C by Bhale et al (2003)

It was obvious that CH coating was less effective inminimising weight loss than MO and MO:CH emulsioncoating (Table 1) Because chitosan films are cationics,water molecules can interact with the matrix andincrease the water vapour permeability rate because oftheir highly hydrophilic nature (Wong et al., 1992;Butler et al., 1996), thus reducing the film’s waterbarrier capability

According to FAO (2003), a weight loss of 2–3% iscommon in marketing eggs and is hardly noticeable toconsumers This study demonstrated that MO (unwiped

or wiped) and MO:CH emulsion (irrespective of theMO:CH ratio) coatings can equally (P > 0.05) offer aprotective barrier against the loss of moisture throughthe eggshell, thus minimising weight loss (<1.03%,Table 1)

Effects of mineral oil, chitosan solution and their emulsions

as a coating material on Haugh unitThe Haugh unit, an expression relating egg weight andheight of the thick albumen, is a measurement of thealbumen quality The higher the Haugh unit value, thebetter the albumen quality of eggs (Stadelman, 1995a).Changes in the Haugh unit of noncoated and coatedeggs during 5 weeks of storage at 25C were observed(interaction between coating treatments· storage peri-ods, P < 0.0001) (Table 2) Overall, the Haugh unitsignificantly decreased with increased storage periods;however, this decrease progressed at a much slower ratefor eggs coated with MO (wiped or unwiped) and⁄ orMO:CH emulsions than for noncoated and CH-coatedeggs Compared with noncoated eggs, eggs coated with

MO (wiped or unwiped) and three emulsions hadsignificantly higher Haugh unit throughout 5 weeks ofstorage (P < 0.05) No significant differences in Haugh

Table 1 Weight loss (%)* of eggs coated with mineral oil (MO), chitosan (CH) and⁄ or three emulsions during 5 weeks of storage at 25 C

0.40 ± 0.07 CD,b

0.46 ± 0.14 BC,b

0.67 ± 0.14 AB,b

0.88 ± 0.35 A,b 50:50 MO:CH 0.21 ± 0.04 C,b

0.37 ± 0.09 BC,b

0.47 ± 0.12 B,b

0.83 ± 0.32 A,b

0.90 ± 0.30 A,b 25:75 MO:CH 0.34 ± 0.21 B,b

*Means ± standard deviations of ten measurements A–E

Means with different superscripts within a row indicate significant differences (P < 0.05) a–b

Means with different superscripts within a column indicate significant differences (P < 0.05).

†

Control, noncoated eggs; MO (U), unwiped after coating with 100% MO; MO (W), wiped after coating with 100% MO; 75:25 MO:CH, coating with MO:CH emulsion at a ratio of 75:25; 50:50 MO:CH, coating with MO:CH emulsion at a ratio of 50:50; 25:75 MO:CH, coating with MO:CH emulsion at a ratio of 25:75; CH, coating with 100% chitosan solution (CH) CH at 2% (w ⁄ v) was prepared in 1% (v ⁄ v) acetic acid.

unit were observed among five treatment groups of eggs

coated with MO and⁄ or MO:CH emulsions throughout

the 5 weeks of storage The Haugh unit of CH-coated

eggs was significantly higher than that of the control

eggs during 2–4 weeks of storage, but was comparable

to that of the control eggs after 5 weeks (Table 2)

The Haugh unit of noncoated eggs decreased from an

initial value of 83.79 to 58.79 after 1 week and to 37.00

after 2 weeks of storage The Haugh unit (53.23–59.12)

of eggs coated with MO (unwiped or wiped) and⁄ or

three emulsions after 5 weeks of storage was higher than

that (37.00) of noncoated eggs after 2 weeks of storage;

this implies that coating with MO (unwiped or wiped) or

MO:CH (irrespective of the ratio) could preserve the

albumen quality for at least three more weeks compared

with noncoated eggs at 25C (Table 2) On the other

hand, the Haugh unit (38.16) of CH-coated eggs after

4 weeks of storage was comparable to that (37.00) of

noncoated eggs after 2 weeks of storage; this implies

that CH coating was also effective in preserving the

albumen quality of eggs for at least two more weeks

compared with noncoated eggs at 25C These results

were substantiated by previous observations for

MO-coated eggs (Homler & Stadelman, 1963; Kamel et al.,

1980; Waimaleongora-Ek et al., 2009) and CH-coated

eggs (Lee et al., 1996; Bhale et al., 2003; Kim et al.,

2007, 2008)

Based on the Haugh unit, eggs can be classified into

four grades: AA (above 72), A (72–60), B (59–31) and C

(below 30) (Lee et al., 1996) Changes in classified egg

grade during 5 weeks of storage at 25C are shown in

Table 2 The grade of noncoated eggs decreased rapidly

from AA to B and C after 1 and 3 weeks, respectively.However, eggs coated with MO (unwiped or wiped) andthree emulsions changed from AA to B grade after

5 weeks, thus preserving the A-grade quality up to

4 weeks The CH-coated eggs changed from AA to Bgrade after 3 weeks and to C grade after 5 weeks

Effects of mineral oil, chitosan solution and their emulsions

as a coating material on yolk indexThe spherical nature of egg yolk can be expressed as ayolk index value, an indication of freshness, by measur-ing the yolk height and width (Stadelman, 1995a).Generally, the yolk index values decreased with in-creased storage periods (Table 3) This decrease wasaffected by the coating treatments and storage period at

25C (interaction between coating treatments · storageperiods, P < 0.0001) (Table 3) and indicated a progres-sive weakening of the vitelline membranes and liquefac-tion of the yolk caused mainly by the diffusion of waterfrom the albumen (Obanu & Mpieri, 1984)

All eggs coated with MO (unwiped or wiped) andthree emulsions had significantly higher yolk indexvalues (0.33–0.37) than noncoated (0.24) and CH-coatedeggs (0.27) after 5 weeks of storage (P < 0.05) Undersimilar storage time and temperature, Waimaleongora-

Ek et al (2009) reported a higher yolk index value (0.37)

of MO (26 mPa s)-coated eggs compared with that(0.21) of noncoated eggs

In our present study (Table 3), the yolk index values(0.33–0.37) of eggs coated with MO (unwiped or wiped)and⁄ or three emulsions after 5 weeks of storage were all

Table 2 Haugh unit* and grade † of eggs coated with mineral oil (MO), chitosan (CH) and ⁄ or three emulsions during 5 weeks of storage at 25 C

*Means ± standard deviations of ten measurements A–E

Means with different superscripts within a row indicate significant differences (P < 0.05).

higher than that (0.30) of noncoated eggs after 2 weeks

of storage Data from Tables 2 (Haugh unit) and 3 (yolk

index) imply that coating with MO (unwiped or wiped)

or MO:CH (irrespective of the ratio) could preserve the

albumen and yolk quality of eggs for at least three more

weeks compared with noncoated eggs at 25C

Effects of mineral oil, chitosan solution and their emulsions

as a coating material on albumen pH

Besides the Haugh unit, albumen pH can also be used as

an indicator for the albumen quality of eggs (Scott &

Silversides, 2000) Freshly laid eggs contain

1.44–2.05 mg CO2g)1of albumen (Keener et al., 2001;

Biladeau & Keener, 2009) and have an albumen pH

value of 7.6–8.7 (Goodwin et al., 1962; Rhim et al.,

2004; Waimaleongora-Ek et al., 2009) In this study, thealbumen pH values of all noncoated and coated eggsranged from 7.91 to 8.76 (Table 4)

During storage, carbon dioxide escapes via eggshellpores, resulting in thinning of the albumen and anincreased albumen pH value up to 9.6 (Knight et al.,1972; Heath, 1977; Kemps et al., 2007) The albumen

pH values of all eggs coated with MO (unwiped orwiped) and⁄ or three emulsions were significantly lowerthan that of noncoated and CH-coated eggs throughoutthe 5 weeks of storage (Table 4) This implies that MOand MO:CH emulsions as coating materials could retard

a loss of carbon dioxide through eggshell pores by acting

as a gas barrier No significant differences in albumen

pH values were observed among five treatment groups

of eggs coated with MO or MO:CH emulsions after

5 weeks of storage

The pattern for changes in albumen pH during

5 weeks of storage somewhat differed with coatingtreatments The albumen pH of noncoated and CH-coated eggs gradually increased from an initial value of8.28 to 8.66 and 8.63, respectively, after 5 weeks ofstorage However, the opposite was observed for thealbumen pH of eggs coated with MO (unwiped orwiped) and⁄ or three emulsions, with the pH decreasingfrom 8.28 to 7.91–8.04 (Table 4) Kamel et al (1980)reported that the albumen pH of noncoated eggsincreased from the initial value of 8.64 to 9.51 after ca

5 weeks of storage at 25C Biladeau & Keener (2009)observed that the albumen pH of MO-coated eggsdecreased from the initial value of 8.35 to 7.96 after

12 weeks of storage at 7C Jirangrat et al (2010)observed that the albumen pH of noncoated eggsmarkedly (P < 0.05) increased from 8.71 to 9.42 whilethat of MO-coated eggs slightly decreased (but notsignificant, P > 0.05) from 8.71 to 8.64 after 5 weeks ofstorage at 25C The decrease in albumen pH duringstorage may be because of the continuing breakdown ofthe constituents in egg white and⁄ or a change in the

Table 3 Yolk index* of eggs coated with mineral oil (MO), chitosan

(CH) and ⁄ or three emulsions during 5 weeks of storage at 25 C

Coating †

0 week 1 week

2 weeks

3 weeks

4 weeks 5 weeks

0.40 B,ab

0.40 B,ab

0.37 BC,a

0.37 BC,a 0.35 C,ab 50:50 MO:CH 0.45 A

0.41 B,a

0.38 BC,a

0.37 CD,a 0.33 D,ab 25:75 MO:CH 0.45A 0.38B,ab 0.38B,bc 0.37B,a 0.37B,a 0.33C,b

CH 0.45A 0.37B,b 0.35BC,c 0.33CD,b 0.30DE,b 0.27E,c

*Means of ten measurements with the standard deviation range of 0.01–

0.04 A–E

Means with different superscripts within a row indicate

signif-icant differences (P < 0.05) a–d

Means with different superscripts within a column indicate significant differences (P < 0.05).

†

Control, noncoated eggs; MO (U), unwiped after coating with 100% MO;

MO (W), wiped after coating with 100% MO; 75:25 MO:CH, coating with

MO:CH emulsion at a ratio of 75:25; 50:50 MO:CH, coating with MO:CH

emulsion at a ratio of 50:50; 25:75 MO:CH, coating with MO:CH emulsion

at a ratio of 25:75; CH, coating with 100% chitosan solution (CH) CH at

2% (w ⁄ v) was prepared in 1% (v ⁄ v) acetic acid.

Table 4 Albumen pH* of eggs coated with mineral oil (MO), chitosan (CH) and⁄ or three emulsions during 5 weeks of storage at 25 C

*Means ± standard deviations of ten measurements A–D

Means with different superscripts within a row indicate significant differences (P < 0.05) a–c

Means with different superscripts within a column indicate significant differences (P < 0.05).

†

Control, noncoated eggs; MO (U), unwiped after coating with 100% MO; MO (W), wiped after coating with 100% MO; 75:25 MO:CH, coating with MO:CH emulsion at a ratio of 75:25; 50:50 MO:CH, coating with MO:CH emulsion at a ratio of 50:50; 25:75 MO:CH, coating with MO:CH emulsion at a ratio of 25:75; CH, coating with 100% chitosan solution (CH) CH at 2% (w ⁄ v) was prepared in 1% (v ⁄ v) acetic acid.

bicarbonate buffer system (Sharp & Powell, 1931;

Obanu & Mpieri, 1984; Biladeau & Keener, 2009)

However, differences in initial egg quality, egg size and

storage conditions (temperature and period) may affect

albumen pH before and after storage (Muller, 1958;

Goodwin et al., 1962; Sabrani & Payne, 1978; Scott &

Silversides, 2000; Silversides & Scott, 2001)

Results from Tables 1–4 collectively indicate that

coating with MO (unwiped or wiped) and⁄ or MO:CH

emulsions (irrespective of the MO⁄ CH ratios) effectively

reduced weight loss and preserved the albumen and yolk

quality of eggs for at least 3 weeks longer than observed

for the noncoated eggs at 25C

Sensory discrimination and purchase intent of noncoated

and coated eggs

The R-index (%) was used to measure the degree of

difference between the control noncoated eggs and

freshly coated eggs (Table 5) A value of 100% indicates

perfect discrimination, whereas a chance value of 50%

indicates that the two samples cannot be differentiated

(Bhale et al., 2003) As shown in Table 5, more

con-sumers indicated that the coated eggs were perceived to

be significantly (P < 0.05) glossier than the noncoated

control, except for eggs coated with emulsion of

MO:CH = 25:75 (not significantly different from thecontrol, P > 0.05, with the R-more of 56.93%) For thesurface odour and colour, and overall surface appear-ance, the unipolar R-index values for all coated eggs fellbetween 37.38 and 55.16, indicating that consumerscould not significantly (P > 0.05) differentiate thecoated eggs from the control noncoated eggs Table 5also shows that the purchase intent of the MO:CH-coated eggs was above 80% compared with 67% for theCH-coated eggs The purchase intent of MO-coated wasnot determined because of its less practicality because ofthe longer drying time Based on Tables 1–5, theMO:CH = 25:75 emulsion would have more potential

as a coating material for eggs because it was more costeffective, yet performed similarly in preserving theinternal quality of eggs, compared to other MO andMO:CH coatings

Microbiological analysisTPC is a quality indicator of the raw material beforeprocessing (ICMSF, 1986) Bacteria including Salmo-nellacan readily penetrate the shell and membranes of

an intact hatching egg (Berrang et al., 1999; Messens

et al., 2005) Results of TPC and Salmonella detectionfor internal noncoated eggs and eggs coated with MO(unwiped or wiped), CH and⁄ or MO:CH emulsionsbefore and after 5 weeks of storage at 25C are shown

in Table 6 At day 0, TPC of noncoated eggs (control)Table 5 R-index (% sensory discrimination)* comparing noncoated

eggs with freshly coated eggs and their purchase intent

Coating †

Surface

glossiness

Surface odour Surface colour

Overall surface appearance

Purchase intent (%) R-index

more

R-index less R-index R-index R-index

*Based on 109 consumers At a = 0.05, the critical R-index value for a

bipolar test is 57.9% for R-index-more and 42.1% for R-index-less, and

the critical R-index value for a unipolar test is 56.65% Italicised R-index

values indicate significant difference (P < 0.05).

†

Control, noncoated eggs; MO (U), unwiped after coating with 100% MO;

MO (W), wiped after coating with 100% MO; 75:25 MO:CH, coating with

MO:CH emulsion at a ratio of 75:25; 50:50 MO:CH, coating with MO:CH

emulsion at a ratio of 50:50; 25:75 MO:CH, coating with MO:CH emulsion

at a ratio of 25:75; CH, coating with 100% chitosan solution (CH) CH at

2% (w ⁄ v) was prepared in 1% (v ⁄ v) acetic acid.

à

More responses were selected by consumers Therefore, the R-index

less was not considered.

Treatments*

TPC † (Pour plate)

TPC (Spread plate)

Salmonella spp.

detection CFU g 21

at a ratio of 25:75; CH, coating with 100% chitosan solution (CH) CH at 2% (w ⁄ v) was prepared in 1% (v ⁄ v) acetic acid.

† TPC, total plate count Values represent the average of two replicates.

ND, not detectable.