Polymer Bulletin 22,599-602 (1989) P o l y m e r B u l l e t i n

9 Springer-Verlag 1989
Optimization of pectin extraction from dried peel of citrus grandis
Dang Mai Huong* and Dang Van Luyen
Institute for Polymer Chemistry, Academy of Sciences, Nghia do, Tu liem, Hanoi, Vietnam
Summary
The influence of acid concentrati on, temperature, and extracti on time on pectin ex-
tracted from dri ed peel of Citrus grandi s have been investigated, The dependence of
pectin yi el d and of the relative vi scosi t y of pectin solution on experimental condi t i ons
has been descri bed by mathematical models.
To obtai n suitable pectin for immobilizing biocatalysts, the extracti on shoul d be car-
ried out at l ow aci d concentrati on, l ow temperature, and for long peri ods of time.
Introduction
The pectin cont ent of citrus peel is usually high, 25 - 30 % of the dri ed peel mass 1).
In Vietnam a large quant i t y of 2 - 3 cm t hi ck Citrus grandi s peel is wasted. It is, there-
fore, worthwi l e t o get pectin from thi s seemi ngl y waste material t o different uses in
f ood i ndust ry 1), ent rappi ng microbial cells 2) etc.
Until now pectin is mostl y used in f ood industry. For thi s purpose a relatively wi de
range of molecular wei ght s of pectin is allowed. However, in order t o form stable po-
lymer beads in fi xi ng biocatalysts, pectin shoul d be in the first place long enough. To
thi s end, this attempt has been made to get the pectin extracti on under control by
means of mathematical models.
Changes of pectin yi el d and sol uti on vi scosi t y as functi ons of extracti on condi t i ons
have been established. The knowl edge of how these factors effect molecular wei ght
of pectin, can make it possi bl e to produce pol ysacchari de carrier for immobilized bio-
catalysts.
Materials and Methods
Extracti on
100 g of Citrus grandi dri ed peel was pretreated with 1 I of hot water. 500 ml of dilute
hydrochl ori c aci d were added t o the 600 g wet material. After mixing, the mixture was
put into a water bath for a fi xed peri od of time.
The liquid phase was then separated. When an equal volume of ethanol was poured
into it under stirring, pectin was precipitated. This gel ati nous preci pi tate was filtered
and dried. The yi el d of raw pectin was determi ned gravimetrically.
Determination of Sol uti on Vi scosi t y
1 g of dri ed pectin was di ssol ved in 100 ml of 0.9 % NaCI sol uti on under slightl X alkali-
ne condi ti on. The relative vi scosi t y of pectin sol uti on was determi ned at 25 ~' C by
means of an Ubbel ohde vi scosi meter No. 3.
*To whom offprint requests should be sent
600
Choi ce of Mai n I nf l uences and Experi ment al Matri x
The ext ract i on can be affected by many f act ors such as t he rati o of l i qui d and sol i d
phase, part i cl e si zes atc. There is evi dence of t he f act t hat t he t hr ee vari abl es - aci d
concent r at i on, t emper at ur e and react i on t i me - pl ay a deci si ve role.
To avoi d di st ur bi ng react i ons of pect i n in aci d sol ut i on like decar boxyl at i on and ri ng
f or mi ng react i ons, t he ext ract i on was done at a t emper at ur e much l ower t han appl i ed
in usual pect i n ext ract i on f rom ci t rus peel 1).
A 3-factor st at i st i cal desi gn was devel oped occor di ng t o 3), The l evel s cor r espondi ng
t o t he f act or s in t hi s pl an were;
aci d concent r at i on 0.72 < c < 2.16 x 1 = ( c - 1. 44) / 0, 72
t emper at ur e ( ~ C) 50 < t < 70 x 2 = ( T - 6 0 ) / 2
react i on t i me( h) 5 < t < 9 x 3 = ( t - 7 ) / 2
The measur ed r esponses were: Yl " pect i n yi el d (%) Y2 " relative vi scosi t y of
pect i n sol ut i on wi t h t he concent r at i on of l g / ml in compar i son wi t h 0,9 % NaCI sol ut i on
R e s u l t s a n d D i s c u s s i o n
15 exper i ment s were carri ed out under condi t i ons accor di ng t o t he mat ri x in Tabl e 1.
The model whi ch cor r esponds t o t he r esponse val ue of t he stati sti cal desi gn ment i o-
ned above has t he f ol l owi ng quadr at i c expr essi on:
Y = b o + b l X 1 + b2x 2 + b3x 3 + b12Xl X 2 + b13Xl X 3 + b23x2x33
+ b11x12 + b22x22 + b33x33x32
(1)
A pr ogr am f or quadr at i c r egr essi on has been r un on a Personal Comput er Appl e II t o
- cal cul at e coef f i ci ent s bi
- veri fy t he si gni f i cance of each coef f i ci ent by t he t -t est at 95 % conf i dence level
f ormul at e mat hemat i cal model wher eby i nsi gni f i cant coef f i ci ent s are el i mi nat ed
- cal cul at e r esponse val ue accor di ng t o t he model and pr ove whet her t he model
can be accept ed
pr edi ct t he opt i mum val ue vy usi ng si mpl ex met hod.
For t he si gni f i cance t est var i ances are needed. They were det er mi ned by a f our-t i me
repeti ti ve measur ement under t he condi t i ons:
x 1 = x 2 = x 3 = 0 c = 1.44 % T = 70 ~ C t = 7 h
f or pect i n yi el d Sl 2 = 0.216 f = 3
relative vi scosi t y s22 = 0.0369 f = 3
After cal cul at i ng all t he coef f i ci ent s of (1) we have obt ai ned t he model s f or pect i n yi el d
and rel ati ve vi scosi t y in t he f ol l owi ng form:
Y1 = 9.37 - 0. 07x 1 + 0. 63x 2 + 0. 16x 3 - 0.17Xl X 2 - 0. 18Xl X 3 - 0. 14x2x 3
+ 0. 05x12 + 0. 03x22 + 0. 31x32 (2)
Y2 = 0.75 - 0. 53x 1 + 0.04x 2 + 0. 07x 3 - 0.06Xl X 2 - 0.07Xl X 3 - 0. 26x2x 3
+ 0. 34x12 + 0. 005x22 + 0. 15x32 (3)
601
Table 1" Results of pectin extraction under different experimental conditions
Run acid T t coded variables pectin yield rel. viscosity
c (%) (~ (h) x 1 x2 x3 Yl Y1 Y2 Y2
1 0.72 50 5 -1 -1 -1 8.63 8.74 1.37 1.35
2 2.16 50 5 1 -1 -1 8.33 8.74 0.50 0.30
3 0.72 70 5 -1 1 -1 10.09 9.99 1.85 1.88
4 2.16 70 5 1 1 -1 10.22 9.99 0.89 0.82
5 0.72 50 9 -1 -1 1 8.74 8.74 1.93 1.88
6 2.16 50 9 1 -1 1 8.81 8.74 0.43 0.82
7 0.72 70 9 -1 1 1 10.75 9.99 1.01 1.35
8 2.16 70 9 1 1 1 9.07 9.99 0.13 0.30
9 2.31 60 7 1.215 0 0 9.97 9.37 0.79 0.60
10 0.56 60 7 -1.215 0 0 9.14 9.37 2.09 1.89
11 1.44 72.2 7 0 1.215 0 10.03 10.13 0.90 0.75
12 1.44 47.8 7 0 -1.215 0 9.02 8.61 1.00 0.75
13 1.44 60 9.34 0 0 1.215 9.68 9.37 0.85 0.75
14 1.44 60 4.57 0 0 -1.215 8.35 9.37 0.58 0.57
15 1.44 60 7 0 0 0 9.71 9.37 0.60 0.75
y 9 experimental response value Y 9 response value calculated by equ. (4) and (5)
By eliminating insignificant coefficients the models are simplified:
Y1 = 9.37 + 0.63x 2 (4) Y2 = 0.75 - 0.53x 1 + 0.34x12 - 0.26x2x 3 (5)
The response values are calculated according to the simplified models (Y1 , Y2 in Ta-
ble 1). The Fisher test indicates that the models are accepted.
Equ.(4) shows that under the employed experimental conditions pectin yield depends
strongly on temperature only. Greater amount of pectin is available at higher tempera-
ture, perhaps due to the acceleration of transfer processes in this solid-liquid reaction.
The great values of b 1 and b 11in equation (5) indicate that the relative viscosity of
pectin in solution depends greatly on acid concentration, x 1. The effect of acid con-
centration on the viscosity is illustrated in Fig. 1. In the investigated variable range of
the acid concentration, 0.72 - 2.16 %, lower acid concentration garantees the produc-
tion of higher viscosity pectin.
A study of the negative coefficient b23 = - 0.26 in equation (5) shows that tempera-
ture and extraction time have contradictory effects on the viscosity of pectin solution.
In an extraction for short period of time - Fig. 1. a: x 3 = -1 ; t = 5 h - the greatest
viscosity is reached at higest temperature - x 2 = I ; t -- 70 ~ C. This means that
the long pectin macromolecules can not quickly be extracted at low temperature. If
the temperature increases, the transfer process of great macromolecules into the li-
quid phase becomes better, so that the pectin solution viscosity increases with in-
creasing temperature.
In the 9 hour extraction - Fig. 1. b: x 3 = 1 ; t = 9 h -, however, the viscosity de-
creases with increasing temperature. The glucosidic linkage of pectin molecules can
be disrupted more easyly in the acid liquid phase than in the solid phase. Thus, in this
case the viscosity is determined primarily by the depolymerization process. So, the lo-
west temperature - x 2 = - 1 ; T = 50 ~ C - yields pectin of the highest viscosity.
602
0!
rel . . vi sc. A / ' ~ +I a 21
0
y2
J
x 2 = - I ~' "~
\ 1
x 2 { t emp)
e w 0 -
- 1 +1 x 1 {acid}
x 2 = -1 b
+1 ~
- 1 - - - - ~ - - ~ - ~ - - { t emp)
x2
' ' v
-'1 ~+1 x 1 {acid)
F ,qure 1 : Changes of t he rel ati ve vi scosi t y of pect i n sol ut i on at t he di f f erent ext ract i on
condi t i ons accor di ng t o e q u . ( 5 ) , a~ x 3 = - 1 ; t = 5 h b._~. x 3 = 1 ; t = 9 h
Many ki nds of r eact i ons of pol ysacchar i de in aci d medi um can resul t in pect i n mol e-
cul es of ver y di fferent char act er or wi t h di fferi ng f unct i onal gr oups. Therfore, t o pr odu-
ce pol ymer f or f i xi ng bi ocat al yst s we have chosen f rom t he t wo opt i mum condi t i ons
t he one, by whi ch t he t emper at ur e is kept low: x 1 = - 1 x 2 = - 1 x 3 = 1.
After decodi f i cat i on, we obt ai n t he f ol l owi ng opt i mum val ues:
c -- 0.72 % T = 50 ~ C t = 9 h
Pecti n ext ract i on as per f or med at t hese opt i mum condi t i ons pr ovi des us wi t h pect i n of
hi gh mol ecul ar wei ght . From its sol ut i on it is abl e t o f orm l arge, t hi n and st abl e fi l ms
f or IR anal ysi s 4). Membr anes of pect i n ext ract ed under ot her condi t i ons are not st ab-
le and can be obt ai ned onl y wi t h t he hel p of anot her pol ymer l i ke pol yvi nyl al cohol .
We al so have used hi gh vi scosi t y pect i n f or ent r apment of yeast cells. Mechani cal l y
st abl e i mmobi l i zed cel l s have been built. It is pr oved by t hese resul t s t hat pect i n f rom
dri ed peel of Ci t rus gr andi s may become a good carri er f or mi crobi al cells.
Ac k n o wl e d g e me n t
Our speci al t hanks go t o Mr. Di nh Xuan Ba and t o Prof. Dr. Nguyen Xuan Dung f or
maki ng val uabl e suggest i ons t o t hi s work.
Re f e r e n c e s
1 UIImanns Enzykl op&di e der Techni schen Chemi e 13, 167 - 192 (1962)
Urban und Schwar zenber g, MLi nchen - Berlin
2 A. Schel l enber ger and R. Ul bri ch, Z. Chem. 25, 10, 349 (1985)
3 K. Hart mann, E. Lezki and W. Sch&fer, St at i st i sche Ver suchspl anung und Auswer -
t ung in t i er Stoffwi rtschaft, VEB Verl ag f or Grundst of f i ndust ri e, Lei pzi g (1974)
4 Dang Mai Huong, Nguyen Xuan Dung and Dang Van Luyen
Vi et nam J. Chem., in pr ess
Accepted November 7, 1989 C