Methyl ester extraction of sunflower oil in a w i n screw extruder
METHYL ESTER EXTRACTION O F RIINFI,OWER OIL IN A W I N SCREW EX'I'RUDER
clitian melalui
Unggulan I erpndu XI1
15-2006.
Ika Antaliw Kartika
bcpikrtcmcut 0fAgroiodustrial Technology FATErA-IPB
Darn~ugaCar~~pus.
P.O. Sox !ZV Rogor f60(IZ lndonaia
r~l:
+62 251 621 971;Fax: t62 251 621 974
-
E-mail a d d r ~ikntW&ythoo.com
:
Lampe.
H.R. dan
y
&mistry.
ContcmpuPrcntici-1 Iall, In 12%) while the rnuisture content
was iow (< 1.5%). For further uutilizatiun of the cake
meal, those qualities were very favorable, particularly for cxtraction o f residual oil wntained in the cake
meal, as solvent extraction is an adaplable methud to
treat this t)pc of cake meal. In addition, thc charac:cristic parlicle size distribution of cake meal, nhich
was dominated by particles less than 0.5 mrn in
dismeter 80%) facilitates the applica-tion oI' this
rr,ciho~l.
This study set out to evaluate the applicatinn
o i twin-screw cxtrudcr to conduct linoleic methyl
ester cxtraction of sunflower oils in a continuous
mode. The characterization of extraclion perforrnance was obscrved by the determinations of oil extrartion yield and mechanical energy input as function
of screw configuration and solvent-ro-solid ratio.
J.
Twin-Screw Extruder
Experiments
were
conductcd
with
o
CLEXTRAL BC 45 co-rotating txvin-screw extruder.
It was built with seven modular barrels, csch 200
mrn in length, and diffcrent twin-screu,s which had
segmental screw elements cach 50 and 100 mm in
length. The modules were hestcd by thermal induction and cooled by water circulation. Seeds werc Icd
into the extruder inlet port by 3 volumic screw
feeder, and linoteic methyl esters werc injected on
module S by a piston pump. The filter section consisting of six hemisphericat dishes with perforations
1 mm in diameter was outfitted on module 4 to
separate extracted oil. Figure 1 shows three scre~v
profiles tested.
I'rrrlcbo arjd ICS.
j,
Experimental
For all experiments, the ternpcrature along the
barrel and the feed rate were fixed at 80°C and 25
kg/h, respectively. The screw rotation speed was
fixcd at 166 - 210 rpm, while the solvenl-to-solid
ratio was varied from 0 to 0.63. To ensure a stable
flow rate and barrel temperaturc, Ihe extruder was
operated fur 20 25 minutes before processing the
actual samples. Upon achieving steady operation,
filtrate (oil.1inolcic methyl ester mixture containing
the foot) and cake meal samplcs were imrnediatcly
collected over a period ~f 20 minutes. The filtraic
dnd cake meal were urighed. The filtrat:: was
liifiher centrifuged to separate the foot from Ihc
uil/linoleic methyl ester rnixturc. The rnolsturc and
rcsidual nil contents of the cake meal were mzasu~cd
accocding to sbndards NF V03-903 and NF V03908. The linoleic methyl ester content of filtratc and
residual oil contained in cake mcal was determined
by gas chromatography using F A M E methcd. O i l
extraction yield was calculated from the following
relationship :
-
hIATERIALS AND METHODS
All trials were carried out using whole and uncleaned sunflower seeds, which were supplied by La
Toulousaine de Cereales. The sunflower seeds were
of :he classic and oleic types. The oil content of
srcds used, expressed in relation to the dry matkr
contcnt uf ur~leanedseed, was 48.5% (classic type)
and 42.2% (oleic type). The seed moisture conlent at
storage was 6.6% (classic type) and 6.2% (oleic
type). The linoleic methyl esters were of the sunflvwcr oils type, which was supplied by COGNIS.
AH sol~entsand chemicals were analytical grades
tlial were obtained from Sigma-Aldrich. Fluka:
10x10
where R is the oil ehbaction yield based on rrsidual
oil content of cake mcal (% mass). & is the inlet
flow rate of the dry s e d (kg%) and Qcd is the outlet
flow rate of the dry cake meal (kg%). Tw and Trdare
the oil contents of the seed (%) and the cake meal
(%), respectively, in relation to the dry matter.
Sccds
J.
2
1
U
g
T2F
.'
.-
,;
U
2
profiles ttsted
nlcchanical
The
rncchanical cnergj
calculatrd
SME
[P/Qd and
[(460 I r 0.C
where I
clectiic currcnt
f'
(W),Qs
flow ratt
S,$
spel
(Volt)
efficiencx of
lincleic
extracti
'esidual
The quality parameters of a crude oil includcd
b:
ti) the acid value, expressed in mg of KOH/g o f oil
J Tek. Ind Pprr 1'01 I # ( l ) . 1 7 - j 2
10310 tj
&
Oil Quality Analysis
(NF T 60-204), which is an indication of the free
fatl! acid mnterit of the oil, and (ii) the iodinc L ~ I U C ,
cxprtssed in terms of [lie number of cr.ntigr,.i~~~.r
01'
iodine absorbed pcr gram of nil (.AOCS-Td 1 d-'12 ).
ttllizh i s a nlensul-c ot'tllc unsaturaliot~oj I b i l . ..
VI
C2F
!,icld was
289b. More oil
!!IC
nl
r ~ ~ l ~ r ~ ~ t . r t r -( sS i! hS )i dmtio x i s it
Seeds
L
riments
,LBC
wae
with
c o - d n gtwin-sacw d e r
sevm d u l a r
gth, and diffaenl Wn-srews
scrcwelermntsd50and100min
e mdutes m
mled
:much id&
Lhermal induc-
circulation. S
~ S
valwic w e \ v
+
Seeds
1 lindeic d y l cJters w r e injacted
by
Tht
wn-
h m i s w disk
Filtrate
$&ms
diameter was &mad
XW
oil.
1
sttd.
t
Filtrate
experiments,
Seeds
tern-
3
feed rate wac fixcd
nxtively. Tht sacw M i m sped
66
rpm, while
from 0 to
slablc
barrel tempatwe, the t m d e r
br
minutes befwe pmssing
npIes. Upon
steady
Ulinoltic m h y l etu mixture
md cake m d sample were
over
20
filtratc
w m weigh& The f~ltrat-,was
lo separate
fml
Ihs
:
ester
molsture
1conten& h c =kc mcal vicrc ~nciuurccl
&
NF
V03conminod
J.
11
Filtrate
Figure I . Scrcxv profiles ic.ird
Specific n~echanicalenergy
,
.
I he specific mechanical energy (SME) waq
calculatrd from the following equation :
cake meal w x
romatogruphy
methcd.
rollo~ving
:
oil exhction
: cake meal I% niass). &is t t ~ ci r l i c ~
bf the dfj ssed (kgh)
QCd is the
~f
cake
(kgh). Tm
Toare
ltcnts
seed
md
meal
:tively,
to
u ~ l i t yparamettrs
1
exprwsed
2041,
mnter,! thc
KOHlg
the
the i d i n c ~ ~ i u c ,
d ~ cnumbcr
c e n t i g ~ d t ~~~I h
I'
pcr
( A ( ~ ( ' s - l ' d 1 I!-"? 1.
mcasurc of thc u n l l u ~
,illnln o t ' r ~ ~ l \
tcms
7rkJ
J Tek, [rid Per, Loi ISi I t . J:-j-'
f n r oi! exlraclion of sunflower scrds in a twin-scrcw e ~ l r u d e BC
r 45
S,tfE
{C17/'U,dand P = r(460 I r 0.95S33.'600j
..\'hcrc I i s t:11; el~clriccurrent (A), P is ~ h cnlotvr
power (W), Qs is the inlet flow rdte of thr seed
(kglli) and SJ i s thc scrcw 1.0taiioi1speed (rpm). 160
(Voll) and 600 (rpm) are used to deszribe the
efficiency of the ektruiler.
Results and discussion
The i~jcction of Iin~leic methyl ester on
module 3 improved rhe oil extraction yield and
decreased the residual oil content of czke nlcal
(Tablc 1). I he injection o f linoleic rnethyl ester
ticilita~cdthe extraction processing by solubilizing
thc oil cuiltilined in ma:criais, the oil cxteaction
y e l d was thus increa*d by approximately 1 0 2xn4. hiclrc oil c x t m c ~ i u nyield was obser~edwhcu
:!L, ::$d\-rf~t-t+
-\(did (5:s) r ~ ~ i~i o 2 i %x r r;~evittori, C., Ciumy, [I.,
tf.,
Lnt~ibclct, 2000.
extrac~ion
frun~
Uhem. SOC.
Table 2. Influence of screw*configuration and sulvent-to-solid ( S I S ) ratio on oil qualiry
I
Profile
1
2
I
Secd
ss
type
(rprn)
Olcic
166
23.79
Oleic
166
Oleic
Oleic
Iodinc value
Solvent content
(mgiodinelIO0 g oil)
("4
0
Acid value
(mg KOHlg oil)
3.15
23.07
0.54
1.28
1 12.02
59.39
166
26.36
0.63
2.27
1 14.90
65.73
210
25.28
0.49
1.47
111.18
50.64
Qs
S/s
(kgh) ratio
1 1 1 1 1
Oleic
Oleic
Oleic
110
26.67
27.30
22.30
0.62
0
0.56
1.64
4
85.00
113.73
85.16
114.80
0
54.87
.I,7'ek Itrd Pert. L'ol I !?(I), 47-r2
Tec
G.S., 1999a.
Ciaukm, A.,
gcwtion efTect
starch
flo
Ciautam, A., Ctroudhury,
199%.
raidence tim,
mi?cing
extrus;on
flour.
22,263-285.
Solvants industriels: s
I
Masson,
Gercel, FI.F.,
pj~olysis(
oilcakr. Riornaw
5(1
573-579.
Ijzie~ak.1 11, 1989.
tbod
163-174.
61.12
I
config
cxiruder:
(K)
OIelc
23 33
in
of
'Table 1. lnfiuence of screw configuration and solvcnt-to-solid (SIS) ratio on methyl ester extraction of sunflower
oils in a twin-screw extruder BC 45
pro..
M.S..
W
Bioenergy :
clitian melalui
Unggulan I erpndu XI1
15-2006.
Ika Antaliw Kartika
bcpikrtcmcut 0fAgroiodustrial Technology FATErA-IPB
Darn~ugaCar~~pus.
P.O. Sox !ZV Rogor f60(IZ lndonaia
r~l:
+62 251 621 971;Fax: t62 251 621 974
-
E-mail a d d r ~ikntW&ythoo.com
:
Lampe.
H.R. dan
y
&mistry.
ContcmpuPrcntici-1 Iall, In 12%) while the rnuisture content
was iow (< 1.5%). For further uutilizatiun of the cake
meal, those qualities were very favorable, particularly for cxtraction o f residual oil wntained in the cake
meal, as solvent extraction is an adaplable methud to
treat this t)pc of cake meal. In addition, thc charac:cristic parlicle size distribution of cake meal, nhich
was dominated by particles less than 0.5 mrn in
dismeter 80%) facilitates the applica-tion oI' this
rr,ciho~l.
This study set out to evaluate the applicatinn
o i twin-screw cxtrudcr to conduct linoleic methyl
ester cxtraction of sunflower oils in a continuous
mode. The characterization of extraclion perforrnance was obscrved by the determinations of oil extrartion yield and mechanical energy input as function
of screw configuration and solvent-ro-solid ratio.
J.
Twin-Screw Extruder
Experiments
were
conductcd
with
o
CLEXTRAL BC 45 co-rotating txvin-screw extruder.
It was built with seven modular barrels, csch 200
mrn in length, and diffcrent twin-screu,s which had
segmental screw elements cach 50 and 100 mm in
length. The modules were hestcd by thermal induction and cooled by water circulation. Seeds werc Icd
into the extruder inlet port by 3 volumic screw
feeder, and linoteic methyl esters werc injected on
module S by a piston pump. The filter section consisting of six hemisphericat dishes with perforations
1 mm in diameter was outfitted on module 4 to
separate extracted oil. Figure 1 shows three scre~v
profiles tested.
I'rrrlcbo arjd ICS.
j,
Experimental
For all experiments, the ternpcrature along the
barrel and the feed rate were fixed at 80°C and 25
kg/h, respectively. The screw rotation speed was
fixcd at 166 - 210 rpm, while the solvenl-to-solid
ratio was varied from 0 to 0.63. To ensure a stable
flow rate and barrel temperaturc, Ihe extruder was
operated fur 20 25 minutes before processing the
actual samples. Upon achieving steady operation,
filtrate (oil.1inolcic methyl ester mixture containing
the foot) and cake meal samplcs were imrnediatcly
collected over a period ~f 20 minutes. The filtraic
dnd cake meal were urighed. The filtrat:: was
liifiher centrifuged to separate the foot from Ihc
uil/linoleic methyl ester rnixturc. The rnolsturc and
rcsidual nil contents of the cake meal were mzasu~cd
accocding to sbndards NF V03-903 and NF V03908. The linoleic methyl ester content of filtratc and
residual oil contained in cake mcal was determined
by gas chromatography using F A M E methcd. O i l
extraction yield was calculated from the following
relationship :
-
hIATERIALS AND METHODS
All trials were carried out using whole and uncleaned sunflower seeds, which were supplied by La
Toulousaine de Cereales. The sunflower seeds were
of :he classic and oleic types. The oil content of
srcds used, expressed in relation to the dry matkr
contcnt uf ur~leanedseed, was 48.5% (classic type)
and 42.2% (oleic type). The seed moisture conlent at
storage was 6.6% (classic type) and 6.2% (oleic
type). The linoleic methyl esters were of the sunflvwcr oils type, which was supplied by COGNIS.
AH sol~entsand chemicals were analytical grades
tlial were obtained from Sigma-Aldrich. Fluka:
10x10
where R is the oil ehbaction yield based on rrsidual
oil content of cake mcal (% mass). & is the inlet
flow rate of the dry s e d (kg%) and Qcd is the outlet
flow rate of the dry cake meal (kg%). Tw and Trdare
the oil contents of the seed (%) and the cake meal
(%), respectively, in relation to the dry matter.
Sccds
J.
2
1
U
g
T2F
.'
.-
,;
U
2
profiles ttsted
nlcchanical
The
rncchanical cnergj
calculatrd
SME
[P/Qd and
[(460 I r 0.C
where I
clectiic currcnt
f'
(W),Qs
flow ratt
S,$
spel
(Volt)
efficiencx of
lincleic
extracti
'esidual
The quality parameters of a crude oil includcd
b:
ti) the acid value, expressed in mg of KOH/g o f oil
J Tek. Ind Pprr 1'01 I # ( l ) . 1 7 - j 2
10310 tj
&
Oil Quality Analysis
(NF T 60-204), which is an indication of the free
fatl! acid mnterit of the oil, and (ii) the iodinc L ~ I U C ,
cxprtssed in terms of [lie number of cr.ntigr,.i~~~.r
01'
iodine absorbed pcr gram of nil (.AOCS-Td 1 d-'12 ).
ttllizh i s a nlensul-c ot'tllc unsaturaliot~oj I b i l . ..
VI
C2F
!,icld was
289b. More oil
!!IC
nl
r ~ ~ l ~ r ~ ~ t . r t r -( sS i! hS )i dmtio x i s it
Seeds
L
riments
,LBC
wae
with
c o - d n gtwin-sacw d e r
sevm d u l a r
gth, and diffaenl Wn-srews
scrcwelermntsd50and100min
e mdutes m
mled
:much id&
Lhermal induc-
circulation. S
~ S
valwic w e \ v
+
Seeds
1 lindeic d y l cJters w r e injacted
by
Tht
wn-
h m i s w disk
Filtrate
$&ms
diameter was &mad
XW
oil.
1
sttd.
t
Filtrate
experiments,
Seeds
tern-
3
feed rate wac fixcd
nxtively. Tht sacw M i m sped
66
rpm, while
from 0 to
slablc
barrel tempatwe, the t m d e r
br
minutes befwe pmssing
npIes. Upon
steady
Ulinoltic m h y l etu mixture
md cake m d sample were
over
20
filtratc
w m weigh& The f~ltrat-,was
lo separate
fml
Ihs
:
ester
molsture
1conten& h c =kc mcal vicrc ~nciuurccl
&
NF
V03conminod
J.
11
Filtrate
Figure I . Scrcxv profiles ic.ird
Specific n~echanicalenergy
,
.
I he specific mechanical energy (SME) waq
calculatrd from the following equation :
cake meal w x
romatogruphy
methcd.
rollo~ving
:
oil exhction
: cake meal I% niass). &is t t ~ ci r l i c ~
bf the dfj ssed (kgh)
QCd is the
~f
cake
(kgh). Tm
Toare
ltcnts
seed
md
meal
:tively,
to
u ~ l i t yparamettrs
1
exprwsed
2041,
mnter,! thc
KOHlg
the
the i d i n c ~ ~ i u c ,
d ~ cnumbcr
c e n t i g ~ d t ~~~I h
I'
pcr
( A ( ~ ( ' s - l ' d 1 I!-"? 1.
mcasurc of thc u n l l u ~
,illnln o t ' r ~ ~ l \
tcms
7rkJ
J Tek, [rid Per, Loi ISi I t . J:-j-'
f n r oi! exlraclion of sunflower scrds in a twin-scrcw e ~ l r u d e BC
r 45
S,tfE
{C17/'U,dand P = r(460 I r 0.95S33.'600j
..\'hcrc I i s t:11; el~clriccurrent (A), P is ~ h cnlotvr
power (W), Qs is the inlet flow rdte of thr seed
(kglli) and SJ i s thc scrcw 1.0taiioi1speed (rpm). 160
(Voll) and 600 (rpm) are used to deszribe the
efficiency of the ektruiler.
Results and discussion
The i~jcction of Iin~leic methyl ester on
module 3 improved rhe oil extraction yield and
decreased the residual oil content of czke nlcal
(Tablc 1). I he injection o f linoleic rnethyl ester
ticilita~cdthe extraction processing by solubilizing
thc oil cuiltilined in ma:criais, the oil cxteaction
y e l d was thus increa*d by approximately 1 0 2xn4. hiclrc oil c x t m c ~ i u nyield was obser~edwhcu
:!L, ::$d\-rf~t-t+
-\(did (5:s) r ~ ~ i~i o 2 i %x r r;~evittori, C., Ciumy, [I.,
tf.,
Lnt~ibclct, 2000.
extrac~ion
frun~
Uhem. SOC.
Table 2. Influence of screw*configuration and sulvent-to-solid ( S I S ) ratio on oil qualiry
I
Profile
1
2
I
Secd
ss
type
(rprn)
Olcic
166
23.79
Oleic
166
Oleic
Oleic
Iodinc value
Solvent content
(mgiodinelIO0 g oil)
("4
0
Acid value
(mg KOHlg oil)
3.15
23.07
0.54
1.28
1 12.02
59.39
166
26.36
0.63
2.27
1 14.90
65.73
210
25.28
0.49
1.47
111.18
50.64
Qs
S/s
(kgh) ratio
1 1 1 1 1
Oleic
Oleic
Oleic
110
26.67
27.30
22.30
0.62
0
0.56
1.64
4
85.00
113.73
85.16
114.80
0
54.87
.I,7'ek Itrd Pert. L'ol I !?(I), 47-r2
Tec
G.S., 1999a.
Ciaukm, A.,
gcwtion efTect
starch
flo
Ciautam, A., Ctroudhury,
199%.
raidence tim,
mi?cing
extrus;on
flour.
22,263-285.
Solvants industriels: s
I
Masson,
Gercel, FI.F.,
pj~olysis(
oilcakr. Riornaw
5(1
573-579.
Ijzie~ak.1 11, 1989.
tbod
163-174.
61.12
I
config
cxiruder:
(K)
OIelc
23 33
in
of
'Table 1. lnfiuence of screw configuration and solvcnt-to-solid (SIS) ratio on methyl ester extraction of sunflower
oils in a twin-screw extruder BC 45
pro..
M.S..
W
Bioenergy :