JFS: Food Engineering and Physical Properties

Preparation of Passion Fruit Puree

by Flash Vacuum-Expansion
P. BRAT, D. OLLE, M. REYNES, P.-O. COGAT, AND J.-M. BRILLOUET

ABSTRACT: Purple passion fruit were processed by flash vacuum-expansion in comparison with a single-strength
juice. A puree was obtained with about 50% / fruit weight yield, which is 2-fold that obtained for the reference juice.
Color and cell-wall polysaccharides of the products were analyzed, and their rheological properties were investi-
gated. The red-purple puree was enriched in anthocyanins and alcohol-insoluble residue. The puree had higher
consistency and viscosity, which was related to its alcohol-insoluble residue and starch contents.
Key Words: purple passion fruit, vacuum-expansion, puree, cell walls, viscosity

Introduction Materials and Methods (1) fruit (1 kg) were placed in the steam-

F LASH VACUUM- EXPANSION IS A PRO -

cess in which plant materials are
1st steam-heated to 60 to 90 8C then
Fruit
heating chamber and heated for 10 min
(final temperature = 85 8C); steam-heat-
Sound, mature purple passion fruit ed fruit and steam-heating liquors were
(Passiflora edulis Sims) (31.8 6 10.2 g; recovered. (2) Fruit (1 kg) were heated
Food Engineering and Physical Properties

instantly introduced in a vacuum

chamber (2 to 5 kPa) where they ex-
pand or disaggregate due to formation
of micro-channels inside the tissues
and instantaneous evaporation of con-
stitution water (Cogat 1994, 1995).
Steam heating induces a thermal dena-
turation of endogenous oxidases, while
the whole process, performing with
absence of oxygen, prevents oxidation
and subsequent browning of the prod-
ucts. Today it is increasingly used as a
pretreatment of grapes before fermen-
tation (Ageron and others 1995) or for
production of tomato puree. Pro-
cessed materials (that is, wine) have a
higher content in total solids and are
more colored than products obtained
by traditional processes (that is,
blanching and crushing). However, to
our knowledge, no detailed study was
ever conducted to characterize pre-
cisely the effects of the process on bio-
chemical constituents of fruit.
With the aim of developing new
intermediate food products, our ob-
jective was to investigate the effects
of flash vacuum-expansion on purple
passion fruit with emphasis on cell-
wall polysaccharides, color, and
rheological behavior of processed
products. Since the process includes
2 consecutive steps (steam-heating
and vacuum-expansion), we investi-
gated the specific effect of expansion
in contrast to heating by characteriz-
ing the products from steam-heated
fruit and steam-heated/vacuum-ex-
panded fruit.
n = 40) were air freighted from Kenya, in the steam-heating chamber for 10
and upon arrival 3 batches (10 kg) were min then introduced in the vacuum
selected. vessel at 3 kPa; vacuum-expanded fruit,
steam-heating liquors, and aromatic li-
Flash Vacuum-Expansion quors were recovered. Purees were
Equipment then prepared by pulping steam-heated
It consisted of a cylindrical stainless- and vacuum-expanded fruit. Finally, an
steel steam-heating chamber (f = 12 aliquot of refined vacuum-expanded
cm; h = 24 cm; v = 2.7 L, Figure 1) fed at puree (100 g) was added with its corre-
normal pressure with a water-steam sponding volume fraction of aromatic
generator and coupled through a man- liquors (10 mL) and thoroughly mixed.
ual pneumatic valve (103 kPa; opening
time 0.5 s) to a cylindrical quartz vacu- Preparation of Alcohol-Insoluble
um vessel (f = 30 cm; h = 48 cm; v = 34 Residues
L) where vacuum (3 kPa) is generated Alcohol-insoluble residues were pre-
by a vacuum pump cooled by a closed pared as follows: Ethanol (400 mL) was
water circuit connected to a condenser. added to aliquots of reference juice and
Steam-heating liquors, generated by purees (100 g), and after thorough mix-
condensation of steam on the fruit and ing, the slurry was boiled for 30 min
by exudation of some inner juice, are then filtered on a sintered glass crucible
collected at the base of the steam-heat- (porosity n8 4). The residue was then
ing chamber. Aromatic liquors, gener- washed successively with (ethanol/wa-
ated by the instantaneous evaporation ter, 80:20) (200 mL), ethanol (100 mL),
of water and volatiles, are collected af- acetone (100 mL), and ether (50 mL).
ter passage on a plate exchanger. Finally, the residue was dried (24 h) in a
vacuum oven (50 8C) and weighed.
Preparation of Passion-Fruit Starch was removed from the alco-
Purees hol-insoluble residues according to
A reference passion-fruit juice was Brillouet and others (1988): Briefly, al-
prepared as follows: Fruit were cut in cohol-insoluble residue (10 g) was cryo-
half with a knife and introduced into a milled in liquid nitrogen with a
rotating pulper (Auriol, Marmande, Dangoumill 300 freezer-mill (Prolabo,
France) equipped with a 1-mm screen. Paris, France) for 5 min (top impact
In order to determine the mass balance frequency) then dispersed in 90% dime-
(fresh and dry matters) for each step of thylsulfoxide (120 mL) at 100 8C and
the flash vacuum-expansion process stirred for 30 min. Boiling 0.1M acetate
(steam-heating then vacuum-expan- buffer (pH 5.0; 240 mL) was then added
sion), 2 experiments were performed: followed by 200 µL of heat-stable type

542 JOURNAL OF FOOD SCIENCE—Vol. 66, No. 4, 2001 © 2001 Institute of Food Technologists
Preparation of Passion-Fruit Puree . . .

ric acid, 25 8C, 45 min, then 1M sulfuric

acid, 100 8C, 2 h. Sugars were then de-
rivatized into their alditol acetates
(Blakeney and others 1983) and ana-
lyzed by gas chromatography according
to Hoebler and others (1989), with inos-
itol as internal standard. Uronic acids
were measured without desterification
after preliminary dissolution in concen-
trated sulfuric acid by the m-phe-
nylphenol procedure (Blumenkrantz
and Asboe-Hansen 1973; Ahmed and
Labavitch 1977). Proteins (N 3 6.25)
were determined by a micro-Kjeldahl
procedure (Bietz 1974).
Color of the samples (L, a, b) was de-
termined using a Minolta CR-300 chro-
mameter according to Askar and Trep-
tow (1993).
Anthocyanins were estimated as de-
scribed by Wrolstad (1976): Briefly, sam-
ple (20 g) was added to 80 mL of either
HCl-KCl buffer (0.2M, pH 1.0) or ace-
tate-HCl buffer (1M, pH 4.5), and after
thorough mixing with a blender, the
Figure 1—Schematic of flash vacuum-expansion process mixture was centrifuged (7000 3 g; 10

Food Engineering and Physical Properties

min). After appropriate dilution with
the 2 buffers, absorbancies were mea-
sured at 510 nm. After substraction of
XII-A a-amylase from Bacillus licheni- umn (300 3 8 mm) (Showa Denko, Sho- absorbancies due to turbidity (l 700
formis (382 µKat/mL, pH 6.9, 20 8C; Sig- ko. Co. Ltd., Japan) eluted with 0.1M nm), the anthocyanin concentration
ma Chemical Co., St. Louis, Mo., LiNO3 at 0.45 mL/min from a Spectra- was calculated with the following equa-
U.S.A.). After incubation for 15 min at SYSTEM P1000XR pump (Thermo Sepa- tion:
100 °C, the mixture was cooled to 40 °C, ration Products, San Jose, Calif., U.S.A.)
and 1 mL of AMG 400L type LP amylo- with online refractive index detection Anth = (A 3 103 3 MW 3 d)/e
glucosidase from Aspergillus niger (13 (Shodex RI-71 detector thermostated at
µKat/mL, pH 4.3, 25 8C; Novo-Nordisk, 25 8C). Calibration was performed with where: Anth: anthocyanin concentra-
Denmark) was added to the medium. narrow pullulan molecular-weight stan- tion (mg/L); A: difference of absorban-
Then the mixture was incubated with dards (P-5, Mw 5 5 800 ; P-10, Mw 5 12 cies at 510 nm at pH 1.0 and pH 4.5 (af-
stirring for 2 h at 40 8C and centrifuged 200 ; P-20, Mw 5 23 700 ; P-50, Mw 5 48 ter correction for turbidity); MW: mo-
(18500 3 g; 10 min). After addition of 4 000 ; P-100, Mw 5 100 000 ; P-200, Mw 5 lecular weight of cyanidin-3-glucoside
vol of ethanol, destarched alcohol-in- 186 000 ; P-400, Mw 5 380 000 ; P-800, (major anthocyanin in purple passion
soluble residue was recovered as above. Mw 5 853 000; Showa Denko) fruit) 5 445.2; e: molar absorbance of
cyanidin-3-glucoside = 29.600; d: dilu-
Isolation of Water-Soluble Analytical Methods tion factor.
Polysaccharides (WSPs) Soluble solids (8Brix) were deter- Intrinsic viscosities ([hint]) of WSPs in
Destarched alcohol-insoluble resi- mined with a hand refractometer at 0.155M NaCl were determined at 25 8C
dues (10 g) were stirred for 45 min at room temperature. The pH was mea- with an automatic Schott Geräte AVS
room temperature in 500 mL distilled sured by a pH meter (pH-Vision 6071, 400 viscometer (Schott Geräte GmbH,
water adjusted to pH 4.0 with 0.1N HCl Jenco Elec. Ltd., Taiwan), and total acid- Hofheim, Germany) (flow time of the
after centrifugation (18500 3 g; 10 min); ity was determined by titration with solvent 90 s).
the pellet was washed with 250 mL wa- 0.1M NaOH. Total solids were deter-
ter and centrifuged in the same condi- mined by drying sample (2 g) for 3 h at Rheological Measurements
tions. Both supernatants were pooled 60 8C then for 24 h in a vacuum oven at Consistency was measured with a
and added with 4 vol of ethanol. Precip- 50 8C. Fisher’s least significant differ- Bostwick consistometer at 20 8C by
itated polysaccharides were recovered ence (LSD) at P , 0.05 was used to measuring the distance (cm) product
by filtration on a sintered glass crucible compare sample means. flowed in 30 s (Takada and Nelson
then freeze-dried. Neutral monosaccharides were re- 1983). Apparent viscosity (ha) of the pu-
leased from destarched alcohol-insolu- rees was measured at 25 8C in a glass
High-Performance Size-Exclusion ble residues (5 mg) by hydrolysis with beaker (50 mL) using a Haake VT500
Chromatography 2M trifluoroacetic acid for 75 min at viscometer (Haake Inc., Saddle Brook,
The molecular size distribution of 120 8C (Albersheim and others 1967). N.J., U.S.A.) with the star-shaped FL 100
WSPs was examined by high-perfor- They were also submitted to Saeman sensor system (Haake Inc.) (f 5 564,
mance size-exclusion chromatography hydrolysis as described by Hoebler and M 5 0.209) at a shear rate (dg/dt) of
using an OHpak SB-804 HQ Shodex col- others (1989), that is, 72% (w/w) sulfu- 2.09 s-1. Yield stress (t0) was determined

Vol. 66, No. 4, 2001—JOURNAL OF FOOD SCIENCE 543

 Preparation of Passion-Fruit Puree . . .

Table 1—Some characteristics of passion-fruit processed purees

Sample
Reference Steam-heated Vacuum-expanded Vacuum-expanded
juice puree puree puree + aromatic liquors
Total solids 17.02 6 0.27 16.23 6 0.24 16.47 6 0.25 14.94 6 0.22
(%/fresh weight)
Soluble solids 14.51 6 0.29 13.02 6 0.22 13.05 6 0.24 11.84 6 0.19
(%/fresh weight)
pH 3.22 6 0.01 3.60 6 0.02 3.81 6 0.02 3.80 6 0.02
Titratable acidity b
(%/fresh weight) 2.84 6 0.05 2.22 6 0.04 2.02 6 0.04 1.84 6 0.04
Color
L 54.30 35.86 32.90 33.59
a 5.18 12.92 16.46 14.03
b 31.38 10.44 8.94 7.60
Anthocyanins (mg/L) 3.91 6 0.05 23.81 6 0.32 45.72 6 0.57 42.73 6 0.46
a Significance of differences was defined at p < 0.05.
b Expressed as anhydrous citric acid

Table 2—Yield and composition of alcohol-insoluble residues

Sample
by increasing shear rate from 1 to 50 s-1
and plotting shear stress (t) against Reference Steam-heated Vacuum-expanded
shear rate. Two mathematical models juice puree puree
(Midoux 1993) were tested to describe Alcohol-insoluble residue 3.2 4.8 6.3
the behavior of the different fluids: (%/fresh weight)
Herschel-Bulkley (t 5 K (dg/dt)n 1 t0) Starch in alcohol-insoluble 80.2 53.1 37.2
Food Engineering and Physical Properties

and Casson (t0.5 5 K (dg/dt)0.5 1 t00.5). residue (%/dry weight)

None of these models correctly de- Destarched alcohol-insoluble 0.6 2.3 3.9
scribed the rheological behavior of the residue (%/fresh weight)
products. Thus, yield stresses (t0) were Composition of destarched
alcohol-insoluble residue
estimated by tangential extrapolation
(%/dry weight)
with shear stress axis.
Uronic acids a 4.2 22.6 25.0
Model dispersions of initial and de-
Neutral noncellulosic 9.3 14.8 16.8
starched alcohol-insoluble residues
polysaccharides b
were obtained by dispersing residues in
water containing 12% sucrose (50 mL) Cellulose c 5.0 23.4 35.5
at their respective concentrations in the Proteins (N x 6.25) 69.0 26.0 16.3
corresponding purees. Apparent vis- Uronic acids d 20.5 34.3 29.7
cosity was measured as above. Rhamnose d 1.2 1.3 1.1
Fucose d 1.2 1.1 1.0
Results and Discussion Arabinose d 10.4 4.7 4.0

P URPLE PASSION FRUIT (P ASSIFLORA

edulis Sims) is composed of a thin,
hard exocarp; a thin, purple mesocarp
Xylose d
Mannose d
Galactose d
7.2
3.7
23.3
8.5
1.3
8.2
9.8
1.2
6.1
layer; a white, spongy endocarp sur- Glucose (non- cellulosic) d 5.8 2.0 1.3
rounding an orange-yellow, pulpy, juicy, Glucose (cellulosic) d 26.5 38.6 45.8
edible aril containing seeds (Figure 2). a Expressed as ‹‹anhydrogalacturonic acid››
Passion-fruit juice, a highly aromatic b Neutral polysaccharides obtained by hydrolysis with diluted acid (TFA or sulfuric acid) and GC of the
alditol acetates and expressed as sum of ‹‹anhydrosugars››
product, is usually obtained from the c Glucose obtained by difference between Saeman and dilute acid hydrolyses
aril portion, after cutting the fruit, by d Mole % of constituent monosaccharides

scooping out the pulp and separating

out the seeds either by sieving or ex-
pression through a cloth (Chan 1993).

Yields respectively. After reincorporation of Chemical Composition of Juice and

Mass balances along the process
were satisfied within 3% to 5%. The low-
est yield (26%/fruit weight) was ob-
tained for reference juice, a value simi-
lar to that obtained by Casimir and oth-
ers (1981) using a converging cone ex-
tractor and a brush finisher. Both
steam-heating and steam-heating/vac-
uum-expansion gave (about 23) higher
yields of refined puree, 48% and 49%,
aromatic liquors in the refined vacu-
um-expanded puree, yield reached
54%. Thus, both treatments softened
and partly disintegrated the fruit rind,
which then passed the refining screen
in an amount equivalent to the juice. It
must be noted that aroma of both
steam-heated and vacuum-expanded
purees were pleasant and different
from that of the reference juice.
Purees
Soluble solids and titratable acidity of
both steam-heated and vacuum-ex-
panded purees were lower than those of
the reference juice, while pH was higher
(Table 1). This could be explained by a
(about 23) dilution of the inner juice
from the aril by the rind (endocarp and
mesocarp) of the fruit. Indeed, separate
measurements on a rind puree gave: sol-

544 JOURNAL OF FOOD SCIENCE—Vol. 66, No. 4, 2001

Preparation of Passion-Fruit Puree . . .

Table 3—Yield and composition of water-soluble polysaccharides sion step was evidenced by the higher
Sample content of the red-purple vacuum-ex-
panded puree than that obtained from
Reference Steam-heated Vacuum-expanded
juice puree puree steam-heated fruit. In agreement with
previous reports (Cogat 1994, 1995), the
Non-starchy water-soluble 0.1 0.6 0.9 vacuum-expansion must loosen the tis-
polysaccharides (%/fresh weight)
sue cohesion by creating microchan-
Composition of non-starchy
nels, thus separating cells at their cell-
water-soluble polysaccharides
(%/dry weight) wall level and facilitating extraction of
Uronic acids a 28.0 63.2 68.3 color.
Neutral non-cellulosic 25.8 8.5 6.6 The single-strength juice had high L
and b values (Table 1), indicating a clear
polysaccharides b
and yellow appearance as usually ob-
Proteins (N x 6.25) — 12.4 10.7
served (Casimir and others 1981). The
Methanol c 2.0 (40.1) 9.9 (86.5) 9.9 (79.7)
orange-yellow color of passion-fruit
Uronic acids d 48.7 86.4 89.9
juice is due to carotenoids, mostly b-
Rhamnose d 1.5 1.0 1.3
carotene (Nakasone and Paull 1998) lo-
Fucose d 0.5 0.1 0.2 cated in the aril portion of the fruit.
Arabinose d 9.8 2.8 2.0 Pulping by itself did not degrade the
Xylose d 3.2 1.7 0.8 purple mesocarp. The purees from
Mannose d 5.9 1.0 0.9 steam-heating or flash vacuum-expan-
Galactose d 25.8 5.6 3.7 sion exhibited higher a values and lower
Glucose d 4.5 1.2 1.2 L and b values than the reference juice,
h (mL/g) — 111 127 with respectively orange-red and red-
a Expressed as ‹‹anhydrogalacturonic acid›› purple appearance.
b Neutral polysaccharides obtained by hydrolysis with diluted acid (TFA or sulfuric acid) and GC of the
The pulp contents (360 3 g; 10 min)
alditol acetates and expressed as sum of ‹‹anhydrosugars››

Food Engineering and Physical Properties

c Values in parentheses are the degrees of methylation (dm) calculated as the molar ratio of methanol
against «anhydrogalacturonic acid»
d Mole % of constituent monosaccharides
uble solids 12%/fresh weight (fw), pH
4.82, and titratable acidity 0.60%/fw.
Anthocyanin determination was car-
ried out with cyanidin-3-glucoside as
reference because it is the major antho-
cyanin of the purple passion fruit (Ma-
cheix and others 1990). As expected, the
orange-yellow reference juice had a
Figure 2—Cross-section of passion fruit
very low content (Table 1) since the
mesocarp containing most of anthocya-
nins was not disintegrated by the pulp-
ing step. The steam-heating treatment
softened the outer portion of the fruit;
thus, part of the mesocarp was recov-
ered in the orange-red puree. The spe-
cific effect of the flash vacuum-expan-
of steam-heated and vacuum-expanded
purees were not measurable due to
their high consistency; it was 32% for
the reference juice. Thus, alcohol-insol-
uble residues were prepared from the
different products. Because mature
passion fruit is known to contain some
starch (Pruthi 1958), alcohol-insoluble
residues were also enzymatically de-
starched, and their respective levels are
given in Table 2. The reference juice
had, as expected, the lower content in
both residues, while their levels in-
creased from the steam-heated to the
vacuum-expanded purees. Again, par-
tial disintegration of the outer portion
of the fruit explained this increase in al-
cohol-insoluble residues that included
cell walls. The specific effect of the flash
vacuum-expansion against the steam-
heating is again evidenced by about
30% and about 70% respective increas-
es in alcohol-insoluble and destarched
alcohol-insoluble residue contents.
Destarched alcohol-insoluble resi-
dues were analyzed for cell-wall
polysaccharides and proteins (Table 2).
Residue from the reference juice was
contaminated by a high proportion of
alcohol-precipitated soluble proteins; in-
deed, the juicy portion of the fruit con-
tains about 2% proteins (Purseglove
1974). The polysaccharidic moiety of the
residues had a typical cell-wall composi-
tion with dominant proportions of uron-
ic-acid-containing polysaccharides and
cellulose followed by decreasing
amounts of xylose, galactose, arabinose,
noncellulosic glucose, and minute pro-

Vol. 66, No. 4, 2001—JOURNAL OF FOOD SCIENCE 545

 Preparation of Passion-Fruit Puree . . .

Table 4—Rheological characteristics of passion-fruit processed products were present in similar proportions.
Sample Thus the specific effect of the flash
vacuum-expansion step against the
Vacuum-
Vacuum- expanded steam-heating step is again evident. It is
Reference Steam-heated expanded puree + possible that the vacuum-expansion
juice puree puree aromatic liquors provokes disintegration of additional
Bostwick consistency — 6.0 1.7 3.7 tissues, richer in cellulose and xylose-
(cm in 30 s) containing polysaccharides, with re-
Apparent viscosity (Pa.s) 2.4 6 0.3 33.0 6 1.4 54.8 6 4.1 045.2 6 1.5 gards to the steam-heating only. A sepa-
Yield stress (Pa) 2.1 6 0.3 72.8 6 2.1 188.7 6 11.7 103.6 6 5.2 rate analysis of cell walls from en-
Reconstituted puree docarp and mesocarp (exocarp was not
damaged) would be needed to ascertain
Alcohol insoluble 3.2 4.8 6.3 —
residue (%/fresh weight) this hypothesis.
Apparent viscosity (Pa.s) 2.4 6 0.3 19.1 6 0.8 51.0 6 3.7 — Water-soluble polysaccharides (WSPs)
Destarched alcohol 0.6 2.3 3.9 — were extracted from the destarched alco-
insoluble residue (%/fresh weight) hol-insoluble residues with water at am-
Apparent viscosity (Pa.s) — 6.5 6 0.4 38.2 6 2.4 — bient temperature. Again, their levels in-
a Significance of differences was defined at p < 0.05.
creased from reference juice to vacuum-
expanded puree (Table 3). They repre-
sented, respectively, 26% and 23% of the
portions of rhamnose and fucose. The and galactose. The residue from the vac- destarched alcohol-insoluble residues
residue from the reference juice had a uum-expanded puree was richer in cellu- from steam-heated and vacuum-expand-
composition very different from the 2 lose and xylose-containing polysaccha- ed purees. According to their polysaccha-
others, having relatively lower propor- rides and poorer in uronic-acid-contain- ride composition, WSPs from steam-
tions of uronic acids and cellulose and ing polysaccharides than that from the heated and vacuum-expanded purees are
much higher proportions of arabinose steam-heated puree. Other neutral sugars mainly pectic substances with dominant
Food Engineering and Physical Properties

proportions of galacturonic acid highly

esterified by methanol. Associated neu-
tral polysaccharides have similar relative
compositions. Again, WSPs from the ref-
erence juice have a different composition
with, as observed in destarched alcohol-
insoluble residues, a lower relative level in
galacturonic acid and much higher pro-
portions of arabinose and galactose. The
molecular size distribution of different
WSPs was examined by high perfor-
mance size exclusion chromatography
(Figure 3); WSPs from steam-heated and
vacuum-expanded purees were similar,
while WSPs from the reference juice were
richer in lower molecular-weight popula-
tions. Intrinsic viscosities of WSPs from
steam-heated and vacuum-expanded pu-
rees were similar.

Rheological Characteristics
Consistencies of the purees were
measured with a Bostwick consistome-
ter (Table 4). The consistency of the pu-
ree from the flash-expanded fruit was
more than 3-fold higher than that of
the puree from steam-heated fruit. This
increase cannot be explained by the
concentration due to the vacuum-ex-
pansion (loss of about 10% of water)
because, even after reincorporation of
aromatic liquors, the consistency was
still far higher than after simple steam-
heating.
Apparent viscosity (Table 4) in-
creased similarly to the consistency:
Figure 3—HPSEC profile of water-soluble polysaccharides (a: reference juice,
b: steam-heated puree, c: vacuum-expanded puree) on a Shodex OHpak SB- Vacuum-expanded puree exhibited
804 HQ column in 0.1M LiNO3. Elution times of pullulan standards (P5 ➜ P800) higher viscosity than steam-heated pu-
are also shown. ree and also after reincorporation of

546 JOURNAL OF FOOD SCIENCE—Vol. 66, No. 4, 2001

Preparation of Passion-Fruit Puree . . .
aromatic liquors. Purees from vacuum-
expanded fruit, even with reincorpora-
tion of aromatic liquors, had far higher
yield stresses than the puree from
steam-heated fruit (Figure 4). Yield
stress of vacuum-expanded puree was
similar to peach puree (Lozano and
Ibarz 1994). The different fluids had a
shear-thinning non-Newtonian behav-
ior and were pseudoplastic and non-
thixotropic.
Because starch was gelatinized dur-
ing the steam-heating step, the influ-
ence of the level in alcohol-insoluble
residues and in their starch contents on
the apparent viscosity of the purees was
studied by reconstituting model disper-
sions of the residues in water-sucrose
mixture. Apart from the puree from the
steam-heated fruit, a good agreement
was found between apparent viscosities
of native purees and purees reconsti-
tuted with alcohol-insoluble residues.
After destarching, the viscosity of the
steam-heated and vacuum-expanded
reconstituted purees decreased by
about 65% and 25%, respectively. Thus,
Figure 4—Influence of shear rate (s-1) on shear stress (Pa): (x) reference juice,
(j) steam-heated puree, (m) vacuum-expanded puree, (r) vacuum-expanded
puree + aromatic liquors
the viscosity of vacuum-expanded pas-
sion-fruit puree is mainly due to (about
75%) its destarched alcohol-insoluble
residue content.
Conclusions
T HE FLASH VACUUM -EXPANSION PRO -
cess, when applied to purple pas-
sion fruit, extends its possible uses by
giving a new intermediate food prod-
uct, a vacuum-expanded puree, with a
2-fold yield as compared to the single-
strength juice. The high viscosity of the
puree is related to its content in de-
starched alcohol-insoluble residue.
More work is needed to characterize
aroma compounds of this product in
comparison with the juice itself.
References
Ageron D, Escudier JL, Abbal P, Moutounet M. 1995.
Prétraitement des raisins par flash détente sous
vide poussé. Revue Française d’Œnologie 153:50-
53.
Ahmed AER, Labavitch JM. 1977. A simplified method
for accurate determination of cell wall uronide
content. J Food Biochem 1:361-365.
Albersheim P, Nevins DJ, English PD, Karr A. 1967. A
Vol. 66, No. 4, 2001—JOURNAL OF FOOD SCIENCE
method for the analysis of sugars in plant cell-wall
polysaccharides by gas-liquid chromatography.
Carbohydr Res 16:127-150.
Askar A, Treptow H. 1993. Measurement of colour. In:
Askar A, Treptow H, editors. Quality assurance in
tropical fruit processing. Berlin, Germany:
Springer-Verlag. p 57-60.
Bietz JA. 1974. Micro-Kjeldahl analysis by an im-
proved automated ammonia determination fol-
lowing manual digestion. Anal Chem 46:1617-
1618.
Blakeney AB, Harris PJ, Henry RJ, Stone BA. 1983. A
simple and rapid preparation of alditol acetates
for monosaccharide analysis. Carbohydr Res
113:291-299.
Blumenkrantz N, Asboe-Hansen G. 1973. New meth-
od for quantitative determination of uronic acids.
Anal Biochem 54:484-489.
Brillouet JM, Rouau X, Hoebler C, Barry JL, Carré B,
Lorta E. 1988. A new method for determination of
insoluble cell walls and soluble nonstarchy
polysaccharides from plant materials. J Agric Food
Chem 36:969-979.
Casimir DJ, Kefford JF, Whitfield FB. 1981. Technolo-
gy and flavor chemistry of passion fruit juices and
concentrates. Adv Food Res 27:243-295.
Chan HTJR. 1993. Passion fruit, papaya and guava
juices. In: Nagy S, Chen Shaw PE, editors. Fruit juice
processing technology. Auburndal, Fla: Agscience.
p 334-377.
Cogat PO, inventor; Aurore Développement, assign-
ee. 1994 Jun 16. Installation pour le traitement de
matières biologiques hydratées. European patent,
EP 0,727,948 B1, n° PCT/FR94/01297, European
Patent Office, Munich, Germany.
Cogat PO, inventor; Aurore Développement, assign-
ee. 1995 Apr 8. Appareillage continu pour désaér-
Food Engineering and Physical Properties
er, chauffer, maintenir en température, refroidir
sous vide des matières organiques solides. French
patent, FR 2,656,547, n° 8917414, INPI, Paris.
Hoebler C, Barry JL, David A, Delort-Laval J. 1989.
Rapid acid hydrolysis of plant cell wall polysac-
charides and simplified quantitative determina-
tion of their neutral monosaccharides by gas-liq-
uid chromatography. J Agric Food Chem 37:360-367.
Lozano JE, Ibarz A. 1994. Thixotropic behaviour of
concentrated fruit pulps. Lebensm-Wiss u-Tech-
nol 27:16-18.
Macheix JJ, Fleuriet A, Billot J. 1990. Anthocyanic pig-
ments. In: Fleuriet A, Billot J, Macheix JJ, editors.
Fruit phenolics. Boca Raton, Fla: CRC Press Inc. p
41-57.
Midoux N. 1993. Modes d’études des écoulements
réels. In: Midoux N, editor. Mécanique et rhéologie
des fluides. Paris, France: Tec & Doc, Lavoisier. p
123-243.
Nakasone HY, Paull RE. 1998. Passion fruit. In: Naka-
sone HY, Paull RE, editors. Tropical fruits. Walling-
ford, U.K.: CAB International. p 270-291.
Pruthi JS, Lal G. 1958. Carotenoids in passion fruit
juice (Passiflora edulis). Food Res 23: 505.
Purseglove JW. 1974. Passifloraceae. In: Purseglove
JW, editor. Tropical crops dicotyledons. London:
Longman. p 420-429.
Takada N, Nelson PE. 1983. A new consistency meth-
od for tomato products: The precipitate weigth ra-
tio. J Food Sci 48:1460-1462.
Wrolstad RE. 1976. Color and pigment analysis in
fruit products. Cornwalis, Ore.: State Bulletin 624
of Agricultural Experiment Station, Oregon State
Univ.
MS 20000224
Thanks to P. Bohuon (CIRAD-AMIS, Montpellier, France)
for help in rheological measurements and to the Conseil
Régional Région Réunion (France) for Doctoral scholarship
(n° 21700).
Authors Brat, Ollé, Reynes, and Brillouet are with
the Centre de Coopération Internationale en Re-
cherche Agronomique pour le Développement
(CIRAD), Département FLHOR, Montpellier, F-
34398, France. Author Cogat is with Aurore
Développement, BP 195, Saint-Pierre (Réunion),
F-97455, France. Direct correspondence to J.-M.
Brillouet (E-mail: brillouet@cirad.fr).
547