Journal of Food Engineering 136 (2014) 19–27

Contents lists available at ScienceDirect

Journal of Food Engineering

journal homepage: www.elsevier .com/ locate / j foodeng

Heat treatment and turbo extractor rotational speed effects onrheological and physico-chemical properties of varietal applesauce

http://dx.doi.org/10.1016/j.jfoodeng.2014.03.0160260-8774/� 2014 Elsevier Ltd. All rights reserved.

Abbreviations: d43, volume based mean particle diameter; AIR, alcohol insolubleresidue; BIC, Bayesian information criterion; HAC, hierarchical ascendant classifi-cation; PCA, Principal Component Analysis; PDM, pectin degree of methoxylation;Span, particle size distribution span; TSP, total soluble pectin.⇑ Corresponding author. Tel.: +1 315 787 2259; fax: +1 315 787 2284.

E-mail address: oip1@cornell.edu (O.I. Padilla-Zakour).

Nongnuch Athiphunamphai a, Haim Y. Bar b, Herbert J. Cooley a, Olga I. Padilla-Zakour a,⇑a New York State Agricultural Experiment Station, Department of Food Science, Cornell University, 630 West North Street, Geneva, NY 14456, United Statesb Department of Statistical Science, Cornell University Ithaca, 173 Comstock Hall, Ithaca, NY 14583, United States

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

Article history:Received 4 September 2013Received in revised form 26 January 2014Accepted 19 March 2014Available online 27 March 2014

Keywords:ApplesauceExtractor rotational speedHeat treatmentRheologyParticle sizePectin

We studied the effect of variety, ripening and processing parameters on applesauce rheology. Four vari-eties at 3 ripening stages were processed into applesauce. Apples were diced, heated to 85 �C for hotbreak process (no heating for cold break), fed to a turbo extractor (400–1800 rpm) and hot-packed. Sam-ples were analyzed for rheological and physico-chemical properties. Results were analyzed by ANOVAand Tukey’s test (p 6 0.05). Variety, ripening, heating and extractor speed, significantly affected sauceproperties. Increasing speed produced thicker sauce. Ripening improved consistency for Crispin and Cort-land cold break sauces. Hot break produced consistent quality sauce over time with 60–100% less syner-esis, 4–10% higher pectin, 20–45% smaller mean particle diameter, and 30–70% higher distribution spanthan cold break; thus, it could overcome variations in consistency from variety and ripening. Consistencyand free-liquid flow could be predicted as functions of particle size, pectin content and pectin degree ofmethoxylation (R2 = 0.80, 0.93).

� 2014 Elsevier Ltd. All rights reserved.

1. Introduction

An important quality attribute of applesauce is its consistency,which is dependent on chemical components and physical proper-ties such as size and distribution of apple tissue particles. Thesecomponents can be affected by many factors, including varietyand ripening of the fruit and any operations involved in post-harvest handling, storage and processing (Mohr, 1989). Godfreyet al. (1995) found that blanching Idared and Rome apples in hotwater at 59–71 �C before pulping to sauce led to thicker saucesand believed that the effect was caused by pectin methylesteraseactivity. Additionally, Schijvens et al. (1998) reported that longercooking times in a screw cooker with steam injection for GoldenDelicious apple slices resulted in increasing applesauce apparentviscosity, applesauce serum capillary viscosity and water insolublesolids, but decreasing applesauce mean particle size. At the pulpingstep, increasing finisher screen opening size resulted in largermean particle size, leading to higher apparent viscosity and

consistency index (Nogueira et al., 1985; Rao et al., 1986; Schijvenset al., 1998). However, increasing finisher speed (500–900 rpm)caused an increase in mean particle size (Nogueira et al., 1985)but no effect on consistency index (Rao et al., 1986) with the sameapple varieties and similar firmness ranges. Therefore, the process-ing-induced changes in applesauce rheology were likely resultingfrom changes in chemical and physical properties and seem to bevery specific to the processing parameters involved. Colin-Henrionet al. (2009) have studied processing effects on dietary fibers andcell wall polysaccharides in applesauce, yet without informationon applesauce consistency and particle size distribution.

In addition to processing, the ripening stage and variety of thefruit may affect applesauce consistency. Variations in raw materialcomposition affect industrial processing of fruits, as nowadays ap-ples are processed all year long due to availability from cold stor-age (0–2 �C, 90–95% relative humidity (RH)) and controlledatmosphere (CA) storage (�1 to 4 �C, 1–3% O2, 4% CO2)(Hardenburg et al., 1986; USDA, 2012). The main changes in applecomposition from storage include loss of neutral sugars, solubiliza-tion and depolymerization of cell wall polysaccharides due to thecombined action of several cell-wall-degrading enzymes on pecticand cellulosic fractions (Billy et al., 2008; Goulao and Oliveira,2008). The effect of these changes on applesauce processing isyet to be determined. Additionally, commercial applesauceprocessing methods have changed from hot break to cold break

20 N. Athiphunamphai et al. / Journal of Food Engineering 136 (2014) 19–27

processes with the utilization of turbo extractors that can pulp thewhole fruit to sauce or puree at room temperature without anythermal pretreatment. Understanding the influence of both rawmaterials and processing conditions on sauce properties will helpto improve product quality. Our goals were to study the impactof apple variety, ripeness, and key processing parameters (heattreatment and extractor rotational speed) on rheological propertiesof applesauce, following current industrial procedures, and tomodel the relationship between physical, chemical and rheologicalproperties of applesauce.

2. Material and methods

2.1. Applesauce processing

Four apple varieties (Malus domestica) from the 2010 harvest,Crispin, Cortland, Jonagold, and Idared were used for the study. Ap-ples were sourced from upstate New York farms and had been keptin controlled atmosphere (CA) storage for 6–8 months. Immedi-ately after receiving the fruit at the New York State AgriculturalExperimental Station (Geneva, NY), apples were placed in coldstorage at 10 �C and 95% RH) to accelerate fruit ripening, and pro-cessed into applesauce at three different ripening times; 0 (begin-ning), 18 (middle) and 36 (matured) days of storage.

Prior to processing, apples were weighed and tested for firm-ness using a hand-held penetrometer model FT 327 (WagnerInstruments, Greenwich, CT). Applesauce processing is shown inFig. 1. Apples (�15 kg) were diced (1.27 cm) by a Dicer ModelNo. C (Urschel Laboratories Inc., Valparaiso, IN). Apple dices weresteam heated in a Vulcan Steam Oven (Model 3TE; Diagon Devices,Inc., Pullham, WA) to 85 �C for 16 min (representing hot breaksauce), and pulped by a turbo extractor (Bertocchi CX5, BertocchiSLR., Parma, Italy) fitted with a 3.2 mm opening screen at threeextractor rotational speeds: 400 (low), 800 (medium) and 1300(high) rpm. The cold break sauce was unheated and pulped bythe Turbo extractor, at higher rotational speeds due to the firmnessof uncooked dices, at 800 (low), 1300 (medium) and 1800 (high)rpm. Water was added (10% and 15% for hot and cold break sauces,respectively) to standardize the soluble solids, and a small amountof 10% ascorbic acid solution was added to prevent browning prior

Fig. 1. Flow chart of (a) cold and (b)

to pasteurization. Sauce was pasteurized (t = 6 min, T = 90 �C) in anagitated steam kettle (Groen MFG. Co., Chicago, IL), and hot-filledinto 16 oz glass jars. Jars were immediately capped, turned upsidedown for cap sterilization, held hot for 3 min and quickly cooled ina water bath to room temperature.

2.2. Rheology measurements

The measurements were carried out on a Brookfield DV-III Ultraprogrammable rheometer (Brookfield Engineering Laboratories,INC. Middleboro, MA) with V73 plus spindle to obtain values ofshear stress for each applesauce sample. The measurements weredetermined by applying shear rate from 0.5 to 3.0 s�1 and thenfrom 3.0 to 0.5 s�1 all at 25 �C. USDA applesauce consistency wasmeasured according to the Grading Manual for Canned applesauce(USDA, 2005). USDA consistency sauce and liquid values weremeasured at room temperature by averaging the readings takenat the four quadrants of the USDA Consistometer flow sheet No.1,a plastic chart containing concentric circular markings 0.5 cmapart. The readings are taken at the edge of the spread of sauceand liquid separated from sauce after 1 min. USDA consistencyfree-liquid is obtained by subtracting sauce from liquid flow. Goodquality sauce should have USDA consistency sauce value <6.5 cmand USDA consistency free liquid value <0.7 cm.

2.3. Particle size measurements

Particle size distribution measurements were performed usinglaser diffraction (Mastersizer model MS-2000, Malvern InstrumentInc., Westborough, MA). Applesauce samples were added into awater-continuous dilution accessory (2000 Hydro-S) filled withdeionized water. The particle size distribution was calculated fromthe intensity profile of the scattered light. The volume based (d43)mean particle diameter and the width of distribution (span) wereobtained for every sample:

d43 ¼X

nid4i

.Xnid

3i where ni is the number of particles

of diameter di ð1Þ

Distribution span ¼ ðd90th percentile � d10th percentileÞ=d50th percentile ð2Þ

hot break applesauce processing.

N. Athiphunamphai et al. / Journal of Food Engineering 136 (2014) 19–27 21

2.4. pH, total soluble solids, and moisture

Applesauces were analyzed for pH using a pH meter (AccumetBasic AB15 pH meter; Fisher Scientific Waltham, MA), total solublesolids using a refractometer (Leica Auto ABBE, Leica Inc., Buffalo,NY) and reported as �Brix, and moisture content following AOAC(Horwitz, 2000).

2.5. Total soluble pectin content and pectin degree of methoxylation

For alcohol insoluble residue (AIR) assessment, applesauce washomogenized with 3 volumes of 95% ethanol. The precipitate wasfiltered through Whatman 55 mm filter paper (Whatman,Piscataway, NJ), dried in an oven at 70 �C until stable, weighedand recorded as% AIR. Soluble pectin content (as galacturonic acidequivalent) was determined according to Fraeye et al. (2009) andpectin degree of methoxylation was determined by using alcoholoxidase and Purpald as described by Anthon and Barrett (2004)for water-soluble and chelator-soluble pectin fractions of apple-sauce, which were proportionally combined as fractions of totalsoluble pectin, and overall degree of methylation of sauces.

2.6. Statistical analysis

Two apple batches were processed for each processing treat-ment, and two jars of applesauce were collected for each batch.Statistical analysis was performed using JMP 10.0 (SAS InstituteInc., Cary, NC). Before running the ANOVA and regression models,we tested for potential multicollinearity and performed PrincipalComponent Analysis (PCA) for rheological properties and chosethe number of components so that 90% of variability is explainedby the top components. Since the variables showed strong correla-tions, this step was essential to increasing the power of our analy-sis. To test differences between mean effects of apple variety,ripening days, extractor speed and heating, we used Tukey’s Hon-estly Significant Difference (HSD) at 95% confident interval. ThenPCA was used to calculate multivariate product spaces to examineall variable effects and hierarchical ascendant classification (HAC)was performed. We selected the final model to predict rheologicalproperties based on the maximal Bayesian Information Criterion(BIC) in a stepwise regression procedure.

Table 1Eigenvectors of principle component analysis of applesauce rheological properties byfirst two components.

Parameters Consistencycomponent

Free liquidflow

Consistency index 0.59 0.03Yield stress 0.59 0.051/(USDA consistency sauce2) 0.55 0.07USDA consistency free-liquid flow �0.08 1.00

3. Results and discussion

3.1. Apple and applesauce analysis

Firmness could be used as an indicator for apple ripening stage(Johnston et al., 2002). In this study, during accelerated ripening at10 �C for up to 36 days, apple firmness significantly decreased from50–60 to 35–40 N for Cortland and from 58–62 to 45–50 N for Cris-pin, showing the change from fresh to ripened stage. However,there was no significant change for Idared (40–50 N) and Jonagold(38–45 N). Idared apples maintain good firm fruit condition for avery long time (8 months or longer at 0 �C air storage) as reportedby Mohr (1989) and notably resist cell wall separation. AlthoughIdared apples firmness did not change within the range of thisstudy, a decrease in firmness might be observed if ripening timeswere longer. Jonagold firmness after controlled atmosphere (CA)storage was lower than values from fresh apples from the sameharvest year reported in another study by our group (60–65 N).Therefore, Jonagold apples may have ripened during CA storageand at day 36, apples were believed to fall in the overripe stage.

Applesauce had 9.5–11.5 �Brix, pH 3.4–3.8, and moisturecontent 86–88% depending on apple variety and ripening stages.Firmness and other fruit ripening indicators and trends were in

agreement with previous reports (La Belle, 1981; Massey, 1989).During ripening, most applesauce, except for Jonagold, had highersoluble solids and pH due to higher sugar levels and loss of acid.The lower soluble solids for Jonagold sauce supported the hypoth-esis that Jonagold apples went to senescent stage. Additionally, noeffect of heat treatment and extractor rotational speed were foundon soluble solids, pH, and moisture content.

Extractor rotational speed, heat treatment and their interac-tions significantly affected applesauce yield (p < 0.0001). No effectsof apple variety and ripening days were found. Applesauce yieldranged 27–53% for cold break sauce and 30–70% for hot breaksauce, and it significantly increased with extractor rotationalspeed.

3.2. Rheological properties

Rheological properties of applesauce were well explained withmore than 95% confidence of fit by the power law model withthe parameter consistency index, and Casson’s model with theparameter yield stress. Those two models have been previouslyused for the description of applesauce and apple puree (Cantu-Loz-ano et al., 2000; Godfrey et al., 1995; Qiu and Rao, 1988; Rao et al.,1986). All applesauce in this study had consistency index and yieldstress ranging 30–270 Pa�s and 30–160 Pa, depending on variety,ripening days and heat treatment, and those values were withinthe ranges reported by previous studies (Godfrey et al., 1995; Qiuand Rao, 1988; Rao et al., 1986).

Applesauce rheological properties were described by the firsttwo components of the Principal Component Analysis, which ac-counted for 68% and 25% of the total sum of squares and kept foranalysis (Table 1). The first two components were named consis-tency component, which is positively related to high viscosity,and free-liquid flow, which is linearly related to USDA consistencyfree-liquid. The desired applesauce quality should have a high con-sistency component and low free-liquid flow. Apple variety, ripen-ing days, extractor rotational speed, heat treatment and theirinteractions significantly affected consistency component andfree-liquid flow (p < 0.0001).

3.2.1. Extractor speed effectThe effect of extractor rotational speed on consistency compo-

nent and free-liquid flow was similar for all varieties and ripeningdays. Therefore, Fig. 2 shows the averaged values of consistencycomponent and free-liquid flow from all varieties and ripeningdays by extractor speed for cold and hot break sauce. For coldbreak sauce, consistency component significantly increased andfree liquid flow decreased when extractor speed increased fromlow to medium with 1–1.5 times higher consistency index, 1–4times higher yield stress, 10–50% lower USDA consistency sauceand 10–20% lower USDA consistency free-liquid. No significantchange was observed when changing from medium to high speed.For hot break sauce, consistency component increased significantlyfrom low to high speed with 1–2 times higher consistency index

(a) (b)

0

2

4

6

8

0

1

2

3

Con

sist

ency

com

pone

nt

Fre

e-liq

uid

flow

(cm

)

Fig. 2. Effect of extractor rotational speed on (a) consistency component (dimensionless) and (b) free-liquid flow of cold and hot break applesauce averaged from fourvarieties and three ripening days.

22 N. Athiphunamphai et al. / Journal of Food Engineering 136 (2014) 19–27

and yield stress, and 10–20% lower USDA consistency sauce. How-ever, there was no change for free-liquid flow.

The extractor rotational speed is correlated to centrifugal force,the driving force pushing apples against the screen (F / speed2)(McLellan and Padilla-Zakour, 2005). The high turning speed (highforce) breaks up the tissue and quickly pushes the liquid and pulpyparts through the screen. At low speed for cold break sauce, forceapplied was insufficient to disintegrate apple tissue; thus, less pul-py parts were added to the extractor output, causing low consis-tency component and high free-liquid flow. Speed had lessimpact once the force reached the maximum threshold, explainingwhy less changes were observed between medium and high speed.Hot break sauce was generally less affected by speed, as it requiredless force to push products through the screen since apples weresoftened during heat treatment. Inadequate speed for hot breaksauce could result in undesirable products as found in this study.At low speed for hot break sauce, product turned out to be pulpyapple juice when processed from high firmness apples (Crispin,Cortland) since apple dices were not softened enough from heat-ing, and less force was applied at the extracting step. Therefore,producers should select optimum speed based on high firmnessapples and ensure that apple dices have been softened enough dur-ing heating. Due to the highest consistency component and lowestfree-liquid flow, applesauce processed at high speed for both cold(1800 rpm) and hot break sauces (1300 rpm) were chosen to con-duct further chemical analysis.

3.2.2. Ripening and heat treatment effectsThe effect of heat treatment on consistency component and

free-liquid flow depended on ripening days and apple variety. ForCortland and Crispin cold break sauce, increasing ripening daysfrom 0 to 36 significantly increased consistency component, anddecreased free-liquid flow, showing better sauce quality (Figs. 3and 4). Cold break sauce processed at day 36 had 1–4 times higherconsistency index and yield stress, 8–16% lower USDA consistencysauce and 50–100% lower USDA consistency free-liquid than fromday 0 for Cortland and Crispin. However, there was no significantdifference on consistency component and free-liquid flow for bothIdared and Jonagold cold break sauces, indicating there was no ef-fect from ripening on sauce processed from these two varieties.Free-liquid flow for Jonagold cold break sauce increased with rip-ening days, showing that sauce processed from overripe appleshad worse quality than ripened apples. The change in consistencycomponent and free-liquid flow produced the same results as thechange in apple firmness. Thus, apple firmness could be used asa quick indicator to track the change in applesauce rheology. Applevarieties responded to accelerated ripening differently and there-fore applesauce samples produced over time showed consistencydifferences.

Compared to cold break sauce processed at the same ripeningday, hot break sauce had lower consistency component when

processed from ripened apples (30–70% reduction of consistencyindex and yield stress and 15–30% increase in USDA consistencysauce), while no significant difference was observed when pro-cessed from fresh apples (Fig. 3). Hot break sauce had 60–100% sig-nificantly lower free-liquid flow for all applesauce, regardless ofvariety and ripeness (Fig. 4). Waldron et al. (1997) reported thatif the forces holding a cell together were stronger than the cellwalls, then failure occurred in the cell walls; if the forces holdingthe cells together were weaker than the cell walls, then the cellswill separate. Therefore, in firm, unheated apples, cell adhesionwas strong and tissue fracture involved rupture across the cellreleasing the cell content as free liquid. While processes such asheating or fruit ripening resulted in a considerable softening of ap-ple tissue and decreased cell adhesion, the rupture of tissue oc-curred by cell separation with no destruction of cell wallmaterials and no cell content was released, causing minimumfree-liquid separation. Additionally, hot break sauces were less af-fected by apple ripening stage as both rheological properties hadsimilar values for all ripening days within the same variety, thusheat treatment could be used to overcome the variation associatedwith apple ripeness. The change in applesauce rheology due toboth ripening and heating effects was generally accompanied bychanges in particle size distributions and chemical composition.

3.3. Particle size distributions

The extractor rotational speed, apple variety, ripening days,heat treatment and their interactions had significant effects on vol-ume weighted mean diameter (d43) and distribution span. For allvarieties and ripening days, increasing speed from low to mediumfor cold break sauce caused an increase of 6–12% d43 and a decreaseof 6–13% distribution span due to the addition of more apple pulpto the final product. There was no effect of extractor speed on d43

and distribution span for hot break sauce.Cold break sauces had higher d43 (680–950 lm) and lower dis-

tribution span (1.3–2.0), exhibiting a sharp and narrow distribu-tion than hot break sauces (400–650 lm for d43 and 2.0–2.8 fordistribution span) (Fig. 5). The decrease in d43 for hot break sam-ples has been reported by a previous study by Vetter et al.(2001). When heating was applied to apple dices, it caused signif-icant tissue softening resulting in easier cell separation along themiddle lamella due to pectin solubilization. Hot break samplesshowed a high amount of smaller particles while cold break sam-ples had larger particles and fewer smaller particles due to thestronger cell wall adhesion of the unheated tissue. For cold breaksauce, particle size characteristics depended on apple variety andripening stages. Increasing ripening days from 0 to 36 resulted inlower d43 and higher distribution span for Crispin and Cortland.The opposite result was found for Jonagold and there was no differ-ence for Idared. However, hot break sauces were not significantlydifferent over ripening within the same variety, except for Jonagold

Fig. 3. Consistency component (dimensionless) of varietal applesauce made from cold break process (rotational speed 1800 rpm) and hot break process (rotational speed1300 rpm).

Fig. 4. Free-liquid flow of varietal applesauce made from cold break process (rotational speed 1800 rpm) and hot break process (rotational speed 1300 rpm).

N. Athiphunamphai et al. / Journal of Food Engineering 136 (2014) 19–27 23

at day 36, which could be explained by the use of over ripe apples.A higher number of smaller size particles could enhance surface–surface particle contacts and decrease the mean distance betweenthe particles, increasing the potential of particle–particle interac-tion (Afoakwa et al., 2008; Agarwala et al., 1992; Yoo and Rao,1994). This could result in less free-liquid flow, as found in saucesprocess from cold break sauce made from ripened apples, and hotbreak sauces. However, the smaller particles present in hot breaksauces fit between the larger particles, thus they act as a lubricantfor the flow of larger particles, causing a decrease in overall viscos-ity (Servais et al., 2002). Even though a shift to a smaller mean par-ticle diameter was observed with ripening, the cold break saucestill had a higher consistency component since the smaller parti-cles found in cold break sauce were still too large to fit between

the larger particles. Additionally, the change in total soluble pectinmay affect the rheological properties, which will be described inSection 3.4.

3.4. Total soluble pectin (TSP) and pectin degree of methoxylation(PDM)

Pectin changes that occur during ripening and heat treatmentare relevant to applesauce consistency and free-liquid flow. TheTSP and PDM values measured in varietal applesauce are given inTable 2. Range of results (0.14–0.33%) was similar to those re-ported in the literature (0.17–0.55%) for apples, over apple storagetime, and for varietal applesauce (Le Bourvellec et al., 2011; Scalzoet al., 2005; Vanoli et al., 2009). The general trend was that during

Cortland – cold break Cortland – hot break

Crispin – cold break Crispin – hot break

Idared – cold break Idared – hot break

Jonagold – cold break Jonagold – hot break

Day 0 Day 18 Day 36

Fig. 5. Particle size distribution of varietal applesauce made from cold break process (rotational speed 1800 rpm) and hot break process (rotational speed 1300 rpm).

24 N. Athiphunamphai et al. / Journal of Food Engineering 136 (2014) 19–27

ripening, TSP increased significantly for Cortland and Crispin coldbreak sauce, while no significant change was observed for Idaredand Jonagold. Nevertheless, a decrease in TSP was found for Jona-gold cold break sauce, indicating that apples were over ripe as re-ported by a previous study (Lozano, 2006).

During the ripening process, pectin undergoes extensive struc-tural changes associated with the general decrease in firmnessand loss of cell cohesion (Bartley and Knee, 1982; Redgwell et al.,1997). The main structural changes include the rupture of pectinbonds to other cell wall polysaccharides and depolymerization ofthe galacturonate backbone, resulting in the conversion of proto-pectin to soluble pectin, caused by enzymes, including endo- andexo-polygalacturonase and glycosidase (Billy et al., 2008; Gross

and Sams, 1984; Gross and Wallner, 1979). Even though therewas no significant difference on TSP from heat treatment, hot breaksauce generally had higher TSP than cold break sauce when com-pared for the same variety and ripening days. Heat could solubilizepectin polymers, similarly to ripening, through a non-enzymaticb-elimination reaction (Sakai et al., 1993; Thakur et al., 1997;Voragen et al., 1995).

The values for the PDM in applesauce are shown in Table 2. Ourresults (33–70%) were comparable to those reported in the litera-ture (47–88%) for apple varieties, with progress of ripening andfor varietal applesauce (Anthon and Barrett, 2008; Johnston et al.,2002; Le Bourvellec et al., 2011). Ripening and heating treatmentdid not have a significant effect on PDM for most varieties except

Table 2Total soluble pectin (g galacturonic acid equivalent per 100 g applesauce) and pectin degree of methoxylation (%) of varietal applesauce made from cold break process (1800 rpmrotational speed) and hot break process (1300 rpm rotational speed).

Variety Ripening daysA Total soluble pectin (%) Pectin degree of methoxylation (%)

Cold break sauce Hot break sauce Cold break sauce Hot break sauce

Cortland 0 0.18cde 0.20bcde 55abcd 60ab

18 0.22bcde 0.22bcde 56abcd 55abcd

36 0.33a 0.22bcde 55abcde 58abc

Crispin 0 0.14e 0.15e 65a 59ab

18 0.16de 0.17cde 56abc 53bcdef

36 0.21bcde 0.20cde 48bcdefg 46cdefgh

Idared 0 0.26abc 0.32cde 40ghi 44defghi

18 0.26abcd 0.33cde 42fghi 39ghi

36 0.30ab 0.33cde 35hi 40ghi

Jonagold 0 0.21abc 0.20a 33hi 38ghi

18 0.19abcd 0.16a 40ghi 37ghi

36 0.14ab 0.16a 39ghi 43efghi

Values followed by different letters differ significantly at the 0.05 level (Tukey’s HSD test compared by variety, ripening days and heat treatment).A Apples were ripened at 10 �C, 95% relative humidity and processed into applesauce after 0, 18, and 36 days.

N. Athiphunamphai et al. / Journal of Food Engineering 136 (2014) 19–27 25

for Crispin applesauce (PDM > 60%), which showed a significantdecrease during ripening (15–30%) in cold break sauce. PDMstrongly depended on apple variety, which could lead to differentsauce properties despite the similar pectin content.

3.5. Relationships between rheological properties, particle sizecharacteristics, TSP and PDM

Principal Component Analysis (PCA) was performed to correlatethe physical, chemical and rheological properties of applesauce.The first three components explained 92% of the variability in thedata. The first two components are presented in Fig. 6a. TSP andPDM had polar influences on PC3 (12.7% variation). Along PC1,free-liquid flow is positively correlated to mean particle diameter(d43) (r = 0.83) and negatively correlated to distribution span(r = �0.81). Further examination of the loading plot between PC2and PC3 shows that free-liquid flow is also correlated to TSP andPDM (r = �0.38, 0.20). The second component represents the con-sistency component, which positively correlated with TSP(r = 0.61) and negatively correlated with PDM (r = �0.37). Addi-tionally, the consistency component, when compared betweenPC1 and PC3, is correlated to d43 and distribution span (r = 0.33,�0.42). A high correlation was found between distribution span

(a) (

Fig. 6. Principal Component Analysis of four varieties of applesauce for rheological and pplot and (b) product map; ellipses represent samples grouped by hierarchical ascendan

and d43 (r = �0.92), showing that when applesauces have smallermean particle diameter, they normally have wider particledistribution.

The hierarchical ascendant classification (HAC) performed onthe PCA data outlined three groups of applesauce products (shownby ellipses in Fig. 6b). The first group consisted of cold break sauceprocessed from less ripened (Cortland and Crispin at day 0, 18) andoverripe apples (Jonagold at day 18, 36). These applesauces werecharacterized by having low consistency component, and highfree-liquid flow. The second group was cold break sauce processedfrom ripened apples (Cortland, Crispin at day 36, Jonagold at day 0,Idared at day 0, 18, 36) and hot break Idared sauce. These apple-sauces had the highest consistency component, but showed a widerange of free-liquid flow. The third group was composed of Cort-land, Crispin and Jonagold hot break sauce for all three ripeningdays. They had the lowest consistency component and free-liquidflow. The results clearly indicate that hot break sauces were less af-fected by variety and ripening days.

Since PCA showed that consistency component and free-liquidflow had high correlation with distribution span, d43, TSP andPDM, the change in rheological properties might be attributed tothe change in these explanatory variables and their interactions.All data from both hot and cold break sauce were subjected to

b)

hysico-chemical properties in plane defined by first two components. (a) Correlationt classification).

Consistency com

ponent Con

sist

ency

com

pone

nt

Total soluble pectin

Total soluble pectin Degree of methoxylation

Degree of methoxylation

Fig. 7. Response surface for the effect of total soluble pectin (g galacturonic acidequivalent per 100 g applesauce) and pectin degree of methoxylation (%) onconsistency component of the averaged values from four varieties of applesauce atdistribution span = 1.7.

26 N. Athiphunamphai et al. / Journal of Food Engineering 136 (2014) 19–27

stepwise regression to predict consistency component and free-liquid flow. The final models, which include only significant terms,are shown below:

Consistency component¼13:95�2:3�span�23:79�TSP�0:22 �PDM

þ0:90�TS�PDM ðR2¼0:80Þð3Þ

Free-liquid flow¼ 9:96� 0:015� d43 � 42:21� TSP� 0:30� PDMþ 0:13� TSP� d43

þ 0:000096� PDM� d43 þ1:11� TSP� PDM

� 0:002� TSP� PDM� d43 ðR2 ¼ 0:89Þ ð4Þ

The model (Eq. (3)) indicated that consistency component was lin-early correlated to distribution span, TSP, PDM and the interactionbetween TSP and PDM. Consistency component is inversely related

Total soluble pectin

Total soluble pectin

Free - liquid flow

Me

Free

- liq

uid

flow

Mean particle diameter (D43)

Mean particle diameter (D43)

Fig. 8. Response surface for the effect of total soluble pectin (g galacturonic acid equivafrom four varieties of applesauce on free liquid flow (cm) at (a) pectin degree of metho

to distribution span since no interaction term between span and theother two values was found. The extent of increase or decrease ofconsistency component at different TSP or PDM was different be-cause of the interaction between these independent variables. Theeffect of TSP and PDM on consistency component is shown inFig. 7 when applesauce has a distribution span = 1.7. IncreasingTSP causes an increase in consistency component when PDM ishigher than 40%, but no significant effect is seen when PDM is lessthan 40% as there are enough available free methoxyl groups tobind with calcium to form a strong network. Overall, consistencycomponent increased when sauce had lower distribution span andhigher TSP, which could be achieved by using ripened apples withthe cold break process. The effect of PDM on consistency compo-nent represented the variation found due to apple variety, and itwas less affected by ripening and heat treatment.

Eq. (4) shows that d43, TSP, PDM and their interactions influ-enced free-liquid flow. Since there is a three-way interaction term,PDM was held constant to explain the change from other two vari-ables, as it is an intrinsic factor depending mostly on variety. Fig. 8aand b shows a surface plot for free-liquid flow at PDM values of40% and 55%, respectively. When the sauce has pectin DM = 40%,free-liquid flow decrease with d43 for all TSP levels. However, athigher PDM, the change in free-liquid flow depends on d43 andTSP. When sauce has PDM = 55%, free-liquid flow decreases withincreasing TSP when d43 is more than 600 lm, or decreases if d43

is smaller when TSP is less than 0.30%. Applesauce withd43 < 600 lm or high TSP > 0.30% has low free-liquid flow repre-senting ideal values in the 0–0.5 cm range. Applesauce with lowTSP and large d43, will have high free-liquid flow, especially whenPDM is higher than 40%. Therefore to achieve low free-liquid flowvalues, applesauce should be made from ripened apples and/or bythe hot break process to achieve the targeted values.

4. Conclusion

Applesauce consistency depends on apple variety, post-harvestripening stage, and processing conditions (heat treatment andextractor speed), which are related to the differences in chemicalcomposition and physical properties. Changes in consistency andliquid separation in applesauce can be explained by soluble pectincontent, pectin degree of methoxylation and particle size distribu-tion. Controlled post-harvest ripening and heat treatment could beused to improve sauce quality (thicker sauce with less syneresis)

Free - liquid flow

Total soluble pectin

Total soluble pectin Mean particle diameter (D43)

an particle diameter (D43)

Free

- liq

uid

flow

lent per 100 g applesauce) and mean particle diameter (lm) of the averaged valuesxylation = 40% and (b) pectin degree of methoxylation = 55%.

N. Athiphunamphai et al. / Journal of Food Engineering 136 (2014) 19–27 27

by increasing total soluble pectin and by attaining targeted particlesize distributions (smaller mean particle diameter and wider parti-cle size distribution). Hot break sauce had consistent qualityregardless of apple variety and ripening; therefore, the hot breakprocess could be used to overcome the variation associated withthese two factors.

Acknowledgements

This research project was supported by the Fulbright Scholar-ship and the College of Agriculture and Life Sciences at CornellUniversity.

References

Afoakwa, E.O., Paterson, A., Fowler, M., Vieira, J., 2008. Particle size distribution andcompositional effects on textural properties and appearance of dark chocolates.J. Food Eng. 87 (2), 181–190.

Agarwala, M.K., Patterson, B.R., Clark, P.E., 1992. Rheological behavior of powderinjection molding model slurries. J. Rheol. 36 (2), 319–334.

Anthon, G.E., Barrett, D.M., 2004. Comparison of three colorimetric reagents in thedetermination of methanol with alcohol oxidase. Application to the assay ofpectin methylesterase. J. Agric. Food Chem. 52, 3749–3753.

Anthon, G.E., Barrett, D.M., 2008. Combined enzymatic and colorimetric method fordetermining the uronic acid and methylester content of pectin: application totomato products. Food Chem. 110 (1), 239–247. http://dx.doi.org/10.1016/j.foodchem.2008.01.042 ER.

Bartley, I.M., Knee, M., 1982. The chemistry of textural changes in fruit duringstorage. Food Chem. 9 (1–2), 47–58.

Billy, L., Mehinagic, E., Royer, G., Renard, C.M.G.C., Arvisenet, G., Prost, C., Jourjon, F.,2008. Relationship between texture and pectin composition of two applecultivars during storage. Postharvest Biol. Technol. 47 (3), 315–324. http://dx.doi.org/10.1016/j.postharvbio.2007.07.011.

Cantu-Lozano, D., Rao, M.A., Gasparetto, C.A., 2000. Rheological properties ofnoncohesive apple dispersion with helical and vane impellers: effect ofconcentration and particle size. J. Food Process Eng. 23 (5), 373–385.

Colin-Henrion, M., Mehinagic, E., Renard, C.M.G.C., Richomme, P., Jourjon, F., 2009.From apple to applesauce: processing effects on dietary fibres and cell wallpolysaccharides. Food Chem. 117 (2), 254–260.

Fraeye, I., Knockaert, G., Van Buggenhout, S., Duvetter, T., Hendrickx, M., Van Loey,A., 2009. Enzyme infusion and thermal processing of strawberries: pectinconversions related to firmness evolution. Food Chem. 114 (4), 1371–1379.http://dx.doi.org/10.1016/j.foodchem.2008.11.041.

Godfrey, U.A.M., Bourne, M.C., Rao, M.A., 1995. Blanch temperature/time effects ofrheological properties of applesauce. J. Food Sci. 60 (6), 1289–1291. http://dx.doi.org/10.1111/j.1365-2621.1995.tb04576.x ER.

Goulao, L.F., Oliveira, C.M., 2008. Cell wall modifications during fruit ripening: whena fruit is not the fruit. Trends Food Sci. Technol. 19 (1), 4–25.

Gross, K.C., Sams, C.E., 1984. Changes in cell wall neutral sugar composition duringfruit ripening: a species survey. Phytochemistry 23 (11), 2457–2461.

Gross, K.C., Wallner, S.J., 1979. Degradation of cell wall polysaccharides duringtomato fruit ripening. Plant Physiol. 63 (1), 117–120.

Hardenburg, R.E., Watada, A.E., Wang, C.Y., 1986. The Commercial Storage of Fruits,Vegetables, and Florist and Nursery Stocks. Agricultural Research Service,Washington, DC.

Horwitz, W., 2000. Official Methods of Analysis of AOAC International. AOACInternational, Gaithersburg, MD.

Johnston, J.W., Hewett, E.W., Hertog, M.L.A.T.M., 2002. Postharvest softening of apple(Malus domestica) fruit: a review. New Zeal. J. Crop Horticult. Sci. 30 (3), 145–160.

La Belle, R.L., 1981. Apple quality characteristics as related to various processedproducts. ACS Symp. Ser. 170, 61–76.

Le Bourvellec, C., Bouzerzour, K., Ginies, C., Regis, S., Plé, Y., Renard, C.M.G.C., 2011.Phenolic and polysaccharidic composition of applesauce is close to that of appleflesh. J. Food Compos. Anal. 24 (4–5), 537–547.

Lozano, J.E., 2006. Fruit Manufacturing Scientific Basis, Engineering Properties, andDeteriorative Reactions of Technological Importance. Springer, New York, NY.

Massey Jr., L.M., 1989. Harvesting, storing and handling processing apples. In:Downing, D.L. (Ed.), Processed Apple Products. Van Nostrand Reinhold, NewYork, NY, pp. 31–51.

McLellan, M.R., Padilla-Zakour, O.I., 2005. Juice processing. In: Barrett, D.M.,Somogyi, L.P., Ramaswamy, H.S. (Eds.), Processing Fruits: Science andTechnology. CRC, Boca Raton, FL, p. 73.

Mohr, W.P., 1989. Influence of cultivar, fruit maturity and fruit anatomy on applesauce particle size and texture. Int. J. Food Sci. Technol. 24 (4), 403–413.

Nogueira, J.N., McLellan, M.R., Anantheswaran, R.C., 1985. Effect of fruit firmnessand processing parameters on the particle size distribution in applesauce. J.Food Sci. 50 (3), 744–746. http://dx.doi.org/10.1111/j.1365-2621.1985.tb13786.x ER.

Qiu, C.G., Rao, M.A., 1988. Role of pulp content and particle size in yield stress ofapple sauce. J. Food Sci. 53 (4), 1165–1170.

Rao, M.A., Cooley, H.J., Nogueira, J.N., McLellan, M.R., 1986. Rheology of apple sauce:effect of apple cultivar, firmness, and processing parameters. J. Food Sci. 51 (1),176–179.

Redgwell, R.J., MacRae, E., Hallett, I., Fischer, M., Perry, J., Harker, R., 1997. In vivoand in vitro swelling of cell walls during fruit ripening. Planta 203 (2), 162–173.

Sakai, T., Sakamoto, T., Hallaert, J., Vandamme, E.J., 1993. Pectin, pectinase andprotopectinase: production, properties, and applications. Adv. Appl. Microbiol.39, 213–294.

Scalzo, R.L., Forni, E., Lupi, D., Giudetti, G., Testoni, A., 2005. Changes of pecticcomposition of ‘Annurca’ apple fruit after storage. Food Chem. 93 (3),521–530.

Schijvens, E.P.H.M., Van Vliet, T., Van Dijk, C., 1998. Effect of processing conditionson the composition and rheological properties of applesauce. J. Texture Stud. 29(2), 123–144.

Servais, C., Jones, R., Roberts, I., 2002. The influence of particle size distribution onthe processing of food. J. Food Eng. 51 (3), 201–208, http://dx.doi.org.proxy.library.cornell.edu/10.1016/S0260-8774(01)00056-5.

Thakur, B.R., Singh, R.K., Handa, A.K., 1997. Chemistry and uses of pectin – a review.Crit. Rev. Food Sci. Nutr. 37 (1), 47–73.

U.S. Department of Agriculture, 2005. Grading Manual for Canned Applesauce.<http://www.ams.usda.gov/AMSv1.0/getfile?dDocName=STELPRD3105662>(retrieved 01.04.13).

U.S. Department of Agriculture, Economic Research Service, 2012. U.S. AppleStatistics. <http://usda.mannlib.cornell.edu/MannUsda/viewDocumentInfo.do?documentID=1825> (retrieved 04.04.13).

Vanoli, M., Zerbini, P.E., Spinelli, L., Torricelli, A., Rizzolo, A., 2009. Polyuronidecontent and correlation to optical properties measured by time-resolvedreflectance spectroscopy in ‘jonagored’ apples stored in normal and controlledatmosphere. Food Chem. 115 (4), 1450–1457.

Vetter, S., Kunzek, H., Senge, B., 2001. The influence of the pre-treatment of applecell wall samples on their functional properties. Z. Lebensm. Unters. Fors. 212,630–635.

Voragen, A.G.J., Pilnik, W., Thibault, J.F., Axelos, M.A.V., Renard, C.M.G.C., 1995.Pectins. In: Stephen, A.M. (Ed.), Food Polysaccharides and their Applications.Marcel Dekker Inc., New York.

Waldron, K.W., Smith, A.C., Parr, A.J., Ng, A., Parker, M.L., 1997. New approaches tounderstanding and controlling cell separation in relation to fruit and vegetabletexture. Trends Food Sci. Technol. 8 (7), 213–221. http://dx.doi.org/10.1016/S0924-2244(97)01052-2 ER.

Yoo, B., Rao, M.A., 1994. Effect of unimodal particle size and pulp content onrheological properties of tomato puree. J. Texture Stud. 25 (4), 421–436.