Korean J. Chem. Eng.

, 19(3), 425-430 (2002)

Novel Evaluation Method for the Water-in-Oil (W/O) Emulsion Stability by
Turbidity Ratio Measurements
Myung-Geun Song, Sung-Ho Cho, Jong-Yun Kim * and 3ong-Duk Kim *
De p a l - ~ e n t o f Che mi c a l Engi neer i ng, Ko r e a Ad v a n c e d Instit~_~te o f Sci ence and Technol ogy,
373-1 Gus e ong- dong, Yuseong- gu, Da e j e on 305-701, Ko r e a
* LG Ho u s e h o l d & He a l t h Car e Re s e a r c h Par k, #84 J ang- dong, Yusung- gu, Da e j e on 305- 343, Ko r e a
( Recei ved 14 June 2001 9 accept ed 15 February 2002)
Ab s l r a c t - T h e turbidity ratio met hod o f evaluating the stabilities o f water-in-oil emulsions has been established with
two wavel engt hs (450 and 850 zm~) by ta!dng the intensity ratio o f two be a ns . The slopes o f turbidity ratio o f several
water-in-oil emulsions wi t h t i me were calculated to evaluate the emul si on stabilities at different HLB (Hydrophilie-
Lipophile Balance), t he amount s o f emulsifiers, and wat er contents. The restilts o f t he turbidity ratio technique were
consistent wi t h t he amount o f phase separat i on o f emul si ons i ncubat ed f or 30 days at r oom temperatm-e. Fr om the
turbidity ratio measm-ements, we detemamed that the required HLB o f diesel oil was about 6.0, and that t he stability
o f emulsion increased with the amount of emulsifier. The increasing amount o f the wat er showed a negative effect on
emulsion stability. Finally, this met hod pro~4des a useful tool for the qui ck evaluation of t he required HLB and the
condi t i on o f emul si fi cat i on t hr oughout this study.
Key words: Stability o f Emulsion, W/ O Emulsion, Turbidity Ratio, Required HLB, Droplet Size DishJbution
INTRODUCTION
The stabilities o f emul si ons are ver y i mpor t ant f or var i ous in-
dusNal processes and emulsion prcducts, but the evaluation o f emul-
sion stability is not easy. Genemlly, the kinetics o f the emulsion-
breaking processes is governed by tt~-ee different particle loss mecha-
nisms: Brownian flocculation, sedimentation flccculation, and cream-
ing [ Red@ and Foglei; 1981]. Particles under Br owni an mot i on
collide and t hen coalesce to form larger izarticles. At the same time,
particles are creaming out due t o the difference bet ween the densi-
ties o f particles and congr uous medi um. Sedimentation floccula,
t i on is due t o dikTerential creami ng rates of par t i cl es and cr eams out
at a faster rate t h a i the smaller ones; therefore, larger particles collide
wi t h s l owl y- movi ng s mal l er part i cl es as t hey cr eam out.
Such emulsion-breaking processes have been studied extensively
by ma l y itwestigators, mad va i ous methods o f determining the emul-
sion stability have been proposed such as droplet size analyses [Trei-
net et al., 1989; Ta&os, 1994], measuring physical i:~werties o f emul-
si on [ Ta nbe and Shan~aa, 1994; Dr eher et al., 1999], and acceler-
at ed tests [Forster et al., 1992; Latreille and Paquin, 1990]. Turbid-
ity me a s u r e me i ~ have also been used t o detem~ine the emul si on
stability [ Ki m and Ki m, 1988; Song et al., 2000; Lee et al., 1999],
and a technique utili~ng the spectral absorbance at several wa ve -
Ieigttzs was proposed t o allow the detennination within a relatively
short period o f t i me [Frenkel et al., 1982; Katfi-nan and Garb, 1981;
Gunj i et al., 1992]. The turbidity "c is defi ned in ten, as o f the at-
tenuation o f electromagnetic racliation at a scattering aagl e o f 0 ~
= } i n ( I / I ) ( I )
~To whom conespondence should be addressed.
E-mail: kjd@kaist.ac.kr
425
wher e l ks the pat h length, and I0 and I are the intet~sities o f the in-
cident and ~aansmitted bea~as, respectively. Expe~-hnentaI 0m-bidi-
ties coul d be measur ed wi t h a conventional UV/ VI S spec~-opho-
tometer, whi ch reads absorbance. I f there is no mol ecul ar absorp-
tion, the t ur bi ~t y is related to the sampl e absorbance measur ed a t a
wavel engt h,
9 /=2.303 x (absorbance) (2)
For a monc~i sperse syst em of nouabsorbi ng iso~ropic spheres in
the absence o f multiple and dependent scattering, the ~bi c l i t y is
g i v e n b y [Mel i k and Fogl ei \ 1983; Ker kei \ 1969]
9 -~Nr~Q(r, 9~, m) (3)
wher e "c depends o n the particle concenh-ation, N, the cross sec-
tionaI area o f partlcIes, m 2, and t he scattering coefficient, Q, whi ch
is defined as the ratio o f the scattezmg t o g e o m~ i c a l cross sections.
The scattering cross sect i on is the total radiation area scattered b y a
single sphere flora a n incident b e a n of unit intensity. This scatter-
ing coefticient is, in turn, a fvilction o f t he particle size, t he ratio of
the refractive index of the parhcle g~ t o the refi-active index o f the
medi um ~,,, denoted by m, and the wavelength of the incident wave
i n the dispei~ing medi um, ) ~ . The wavel engt h 9~ is equal t o kJ/.t,,,
wher e ~0 is the wavel engt h in a5i~ The scattea-ing coefficient is cal-
culated b y using the Mi e theory of light scattemag Since the turbid-
ity o f a pol ydi spersi on is the sum of all the conttibutions over the
various particle sizes, the turbiclity o f the total distribution expressed
i n integral fon~a is [Kerker, 1969]
N~ ~
= - ~ - ~ c~ f(c~)Q(c~, m)dc~ (4)
where cz(=2nr~) is dimensionless particle size, and f(cZ) is the distri-
but i on functioi1 Therefore, the turbidity o f emul si on can be eval u-
426 M.-G. Song et al.
ated with particIe size distribution (PSI)).
The changes in PSI) L~der conditions of simultaneous floccula-
don and creaming for vaz-ious times showed that the PSI) shifts to
large particle sizes, owing to flocculation of the small particles ihi-
tally. Since the stability to flocculation increases exponentially with
increasing particle size, the smaller particles flocculate quickly to
form relatively stabIe larger parttcles [Becher, 1988]. Rathermore,
the lager particles cream out at a much faster rate than the smaller
ones, thereby resulting in an observable shift of tile PSD back to
the smaller after long brae. Finally, the PSD becomes increasingly
na~xower for longer times because the total particle concentration
is now exlremely low, thereby resulting only in tile slow creaming
out of the smaller particles. If the prepared emulsiom were stable
relatively, these cyclic changes or shit~ in PSI) would occur slowly.
From this point of view, it is possible to evaluate the emulsion sta-
bility by measurement of the change of PSD with time and the ~r-
bidity method can be a useful one, because the ~bi ~t y measure-
ments are rapid and simple.
It is well established that the intemity of light scatteiqmg increases
with decreastng droplet diameter, and therefore we can estimate
the relative size dis~-ibution by measuring tile ratio of turbidity at
two widely separated wavelengths. In fact, the change of IJarbidity
ratio as a function of time exhibits the relatve emulsion brealdng
process by simultaneous flocculation and coalescence more exactly
compared with turbidity at single wavelength.
In order to appraise the turbidity ratio technique, several water-
in-oil emulsiom were tested by using co~m-nerdaUy available e~-nuI-
sifters. The results of this technique were compared with other meth-
ods such as &-oplet size dis~-ibution and the m'nount of phase separa-
tion of emulsion.
EXPERJMENTAL
1. Mat eri al s and Met hods
The oil phase was res~icted to dieseI (composed of C12-C20 hydro-
carbon mLxture) and was obtained from Hanwha Enelgy Co. The re-
fiactive index of oil meas~-ed with Abb6 reffactometer is 1.469 at 20
~ No a&-iitives were contained in oil and deioi/zed water was used
For emulsifiers, sorbitan moncoleate (Span 80), polyoxyethylene
sorbitan monooleate (Tween 80), polyoxyethylene 2-cetyl ether
(Brij 52), and polyoxyethylene 2- and 12-nonylphenyl ethers (NP2
and NP12) flora Sigma Chemicals were used without ftather puri-
fication. Emulsions were prepared by dissolving the hy&ophobic
emulsifiers in oil phase and mixed thoro~@aly with slowly adding
the water phase contmned with hydrophilic emulsifier. Laboratoly
homogenizer (Heidolph DIAX 900) was used at 10,000 rpm for
8 minutes. All the emulsions were aJiowed to age for 24 hours at
room temperature and diluted 500 ~' nes in oil within 1 minute. Spec-
tral absorbance was measured after dilv~on by using the Jasco V-
570 UV/V-LS speclrophotometei: Tile IJarbidity ratio was desigilated
as R(='cJ'cs, L: long wavelength, S: short wavelength).
Droplet size distributions were analyzed by means of dynamic
light scattering technique with a Zetaplus instrument (Brookhaven
Instruments Co.) aider dilv~on with 500 ames. Multlmodal size dis-
tribution was calculated based on Non-Negatively Comtrained Least
Square (NNLS) method and the results were plotted as Iognon-naI
size distributions. The droplet size variation was monitored with time.
The emulsions were incubated in order to measure the amount
of phase separation of the water phase from the oil phase. The fleshly
prepared emulsions were replaced into 25 ml-cylindticaI test tubes
and incubated at room t empera~e. After 30 days, tile sepa-ation
heights were measured and compared with ~trbidity ratio measure-
m ent.
RESULTS AND DI SCUSSI ON
Fig. 1 shows the speclrum of typical optical densities with wave-
1.0
06
<
0.8
0.6
0.4 I
0.2
0. 0
400 500 600 700 800 900
Wavelength / nm
Fig. 1. The spectral absorbance of diluted (x500) emul si on i)re -
pared wi th Span 80 and Tween 80 at HLB=5. 5 (2.5 wt%
of enmlsifier and 10 wt % of water).
11 ,
j 9
i I SpanS0
.......... i . . . . . . . . I 9 Span80+Tween80(l:0.07)
1.0 / I 9 NP2+Span80+Brij52+NP12
1 I ~ I (1:1:0.47:0.09)
0.8 ~ ~ ~. . . . . . . . . . . . .
g } " :
0.7 !
._Z-
4~
~_ 0.6
0.5 "
21_ 2
0 20 40 60 80 100 120 140 160
Time / minute
Fig. 2. Change of tm'bidity i~lio, R, in diluted emulsions prepared
from different blends of enmlsifiers wilil thne (2. 5wt% of
emul si fi er and 10 wt % of water).
May, 2002
Evaluation of Emtdsion Stability 427
lengths for the diluted emulsions prepared ti'om the blend of Span
80 and Tweat 80 at HLB=5.5. The spectral absorbance increases
with decreasing of wavelength in the range of 425 nm to 900 trot.
The spectral absorbance of diluted emulsions Ix'epmed ~om ditth'-
ent emulsifier blends shoves sinfilar trends vdth Fig. 1. Therefore,
the R-value was defined as the ratio of speclral absotbance (T) at
850 ant and at 450 nm.
Fig. 2 shows lhe changes of R-values of diluted emulsions pre-
pared fi~m three different emulsifier blends as afunction of time All
1. 2 t : 9 HLB=4. 3
9 HLB = 5. 0
: 9 HLB = 5. 5
~' ~ 10 ~ . . . . . . . . . . . . . . . . . i . . . . . . . . . . . . . . . . . . . . . . . . . . . . ~ HLB=6. 0
I - ~ ] ~ i 9 HLB = 6. 5
n, 0.8 . . . . . . . . . . . . . . . . . . . . . . . . .
. - ~ [ ] :
~- ' - - - - - ~- ~ i i .
0.4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
0 10 20 30 40 50
"rime / minute
Fig, 3. C~ange of ttwbidity ratio, R, h~ dihnted emldsloIts prepared
wi th Span 80 and Tween 80 (2.5 wt~ of emulsifier and 10
wt % of water) wi t h t nne as af undi on of HLB value.
emulsions containing 15 WO/o of water and 2.5 "we/~ of emulsifiers.
R-values of emulsions of(==) and ( ~) are more rapidly changed
than those of emulsion (~). Furthemmte, there ae poirts of inflec-
tion in changes of R with time for file unstable emulsions at around
50 minutes. Therefore, we defined slope of turbidity ratio (=AR/
At) as the initial variation up to 50 minutes.
Fig. 3 shows the changes of turbidity ratio for the diluted emul-
2.
9 T,
, ~:
.'d"
4C
2',i
/
/
I
.c-N:I IINII~ $; h*: ~
n-:l ,,1' ,l +o:,
F F
I
, P
9 | ~, t; .' A
t-., R ~'.:l, ~
Fig, 5. Inner phase separation of water ~n dicsd on emulslolLs pre-
pared with Span 80 and Twem 80 (2. 5~t% of r
and 10 wt% of water) as a fandion of H~B value after 30
days of incubation -at room tonperat~we.
6 ! . . . . . . . . . . . . . .
5 . . i . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
.2~ 4 ..i . . . . . . . . . . . . . ' . . . . . . . . . . . . . . " . . . . . . . . . . . . . . . : . . . . . . . . . . . . . . . : . . . . . . . !
._o
2
<D
_9
(O
I
o 1/ . . . . . . . . . . . . . . . . i +
7
HLB number
Fig, 4. Change of slope of Owbidity ratio, R, in diluted emulsions
prepared with Span 80 and Twe~ 80 (2.5 wtO/o of emulsi-
tier and 10 wt % of water) as a function of HI B value.
1.2
9 1.0wt%
D 1. 8wt %
9 2. 5wt %
1. 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . V 3. 5wt %
~ 4. 5wt %
"--- 9 6. 0wt %
E
z
06 . . . . . . . . . . . ~_ - ~r ~: - : _: ' ~
x_ I . . . . k . . . . . . . . . . i
0 10 20 30 40 50
Ti me / mi nut e
Fig, 6. ( ~an~ of t t wbi d~ ratio, R, i n d~uted emlflslons prepared
with Span 80 and ~vecf l 80 ( H~B=6. 0 and 10 wto/~ of wa-
ter) wi t h t i me as a function of the amount of emulsifier.
Korean J. Chem. Eng.(Vol. 19, No. 3)
428 M.-G. Song a al.
sions prepmed fi'om Span 80 and Tween 80 blends (Z5 wt% ofenml-
sifter and 10 wt% of water) at different HLB. The turbidity ratios
decrease linem]y with time The slopes vary with HLB mid are plot-
ted in Fig. 4. R can be seen that the ctmlge rate of slope mininfized
at HLB=6.0, at which the emulsion is the most stable and there-
fore, the required HLB ofoi l is 6.0.
At ~" 30 days at morn tanperaure, emulsions were sepm'ated
into three regions - sediment layer, middle phase, and oil phase -
and the miuimmn separation was occurred at 6.0 of HLB as shown
in Fig. 5. These obs~cations wa~e consistent with turbidity ratio
measurements and we concluded that enmlsion pi~pared with Span
80 and Tween 80 shows the maximum stability at 6.0 of HLB.
The effect of the amount ofenmlsifier (Span 80 and Tween 80
blend) hi emulsion stability was studied at 6.0 of HLB value and
10 wl% of water content. The variations of turbidity ratio as afunc-
tion of the amount of emulsifier were measured (Fig. 6), and the
slopes were calculated fiom file linear decreases of turbidity ratios
with time. Fig. 7 shows that the slopes of turbidity ratio dea~eased
with the hlaxase of the mmunt of enmlsi_tqet, which is consistent
with the stabifity hm~ease as the enmlsifier concentl'ation itlCrease.
It is a common phenomenon, found f w all emulsions, that stability
increases up to a certain emulsifier concenamion, and this is well
in accord w~l the result of the above experiment. In fact, the effect
of the amount of enmlsifier for stabilization of emulsion would be
saturated near 6 wt%.
Fig. 8 shows the phase ~oarations of enmlsions prepared fi~om
different emulsifier content after 30 days at room temperattae. The
separation ~notmt of enmlsion decreases with the amount ofenml-
sifter, and this result is compatible with the change of slope of tur-
bidity ratio.
Finally, Fig. 9 shows the effect of the amount of water Emulsifi-
ers were fixed at 6.0 of HLB value and 2.5 wt% of emulsifier con-
centration (Span 80 and Tween 80 used). As the amount of water
2. 5
~ 1. 5 . . . . . . . .
1.0
Q.
_o
(/)
0.5 . . . . . . . . . . . .
: !
- - 1 . . . . . I - - ~ . . . . . L - -
,4 6 8
Amount of emulsifier I wt %
Fig. 7. C]tange ~ slope ~ ~trbidity ratio, R, in diluted emifl~ions
prepared with Span 80 and Tween 80 (HLB =6.0 and 10
wt% of water) as a fundion of the muotmt of emulsifier.
I ,-A3
,' l l [
I
" 7
] ~ : ~: , ] ~l , ~' l -~);r3,. ~"
J , ?.~ i|:1 ~ I:hA t-J*
,,,: I :I - / - ' ' ~ ......
:i 4" 1
, ! _
.:11
fl -
_
:, 2 ]. .1 fi ~] T v]
a. . , , ULnl :.,[ . ~r t , s i [ w" ,' wl %
Fig. 8. Inner phase separation of water h~ diesel on emul~ons pre-
pared ~i th Span 80 and Tween 80 (HLB=6.0 and 10 wt%
water) as a function of the -amount ef enmlsifier -after 30
days ef incubation m room tmlperatm'e.
1.00
0.95
O
0.9o
iv-
"5 0.85
0.80
~- 0. 75
o
0.70
0.65 i
10 20 30 40
Amount of water I wt%
Fig. 9. Chan~ of slope of turbidity ratio, R, in dflmed omfl~ions
prepared front Spon 80 mad lVceen 80 (HLB=6.0 mad 2.5
wt% of emulsifier) ~s a fimetion of the amount of water.
increased, the stability decreased beyond 30 wt% of water. Appar-
ently, file increase in the watt" phase has a negative efl'ect on fire
stability of emulsion. However, the effect of w~er content on stabil-
ity ofw~ter-in-oil emulsion was relatively small conll0ared with the
effect of HLB or the anount of emulsifier, since fire slope oftm' -
bidity ratio changed only about 0.2 -3) as the ~noant of water
increased fi ~n 5 to 40 wt%, and equivalent to merely 0.5 wt%
change of the ~noant of anulsifier (fi~n 2.5 wt% of emulsifier).
May, 2002
EvaIuation of Emulsion Stability 429
t --e-- 0 min
100 - 0 15mi n
30 min
50 min
8o- C
__>,
_E 6o
40
0 500 1000 1500 2000
Di amet er / nm
Fig. 10. Time dependence of the droplet size distribution of an
emulsion prepared with Span 80 and Tween 80 (HLB=6.0,
2.5 wt% of emulsifie~; and 10 wt% of water) after 500
times dilution measm'ed by using a dynamic light scatter-
ing technique.
--e-- 0 mini
9 - o- 10 min :
100 ~ nnnnn. --~" 30 min [
?. ~ ~ = . -o_ 50mn I
t . # ~ ~ 9 . o 7o_mm]
.o N "xrR "
It'/ ...9 \ \ \ . -.
.... i t _. , __. _ 9
0 500 1000 ]500 2000 250o 3000 3500
Di amet er /nm
Fig. 11. Ti me dependence of the droplet size distribution of an
emulsion prepared with Span 80 (2.5wt% of emulsifie~;
and 10 wt% of water) after 500 times dilution measured
by using a dynamic light scattering technique.
Since tile tw-bidity ratio method provides indirect size distn"ou -
tioil, we observed tile ddroplet size distribution in diluted emulsion
system with time. It is woithwhile to comt:are two methods. There-
fore, tile ~ne dependence of tile droplet size dish-ibution was moni-
tored by using dynamic light scattering tectmique.
Figs. 10 and 11 show tile time dependencies of the size di~ibu-
tion of enmlsion di-oplets diluted with 500 ~nes in oil phase. EmuI-
sion with Span 80 and Tween 80 at HLB=6.0 was used in Fig. 10,
and with Span 80 only (HLB=4.3) in Fig. 11. From Fig. 4, enml-
sion at 6.0 of HLB ks more stable than at 4.3. Both emulsiom con-
tam 2.5 wt% of emulsifier and 10 wt% of water. In Fig. 10, the water
droplets are small and the size distribLaon is initially narrow, but
the size distribution shifts toward lager partMe sizes and becomes
broader, owing to simultaneons flocculation or coalescence of the
smaller particles with time. In Fig. 11, the di-oplet size distribution
also shifts toward l ager size, but much faster and broader than in
Fig. 10. Generally, the turbidity decreases with time at any wave-
length, because the number of droplets decreases by the emulsion-
breaking process. Furthermore, tile turbidity at longer wavelength
decreases much faster thai at shorter wavelength, since tile larger
par~cles cream out at a much faster rate than the smaller ones. The
more unstable tile enmlsion is, the faster the turbiclity ratio decreases.
Therefore, it is a usefi, iI, rapid, and relatively simple method for the
evaluation of tile emulsion stability.
CONCLUSI ONS
Tile validity of the ~-bidity ratio measurement to evaluate water-
in-diesel oil emulsion stability has been examined tt~-oughout t t ~
study. Tile turbidity ratios were measured with fine after dilution
of emulsion, and the slot:es of the tarbi~ty ratio were compared.
Vaiious fa~oz~ affecting emuIsion stability, such as HLB of emuI-
sifiez; tile amount of emuIsifier and water concentration were tested
by using this method. The turbidity technique has been compared
with sepaation anount of emulsion after 30 days incubation at room
temperatare.
As tile results of tile turbidity ratio measurements, the required
HLB of diesel oil was 6.0 and the stability of emulsion was increased
with the amount of emulsifier. The increasing amount of the water
showed a negative effect on emuIsion stability. These results were
well in accord with those of inner phase setx~ation measureme:~.
Throughout this study, turbidity ratio technique proved to be a
simple, rapid, and relatively accurate method for the evaluation of
the emulsion stability.
ACKNOWLEDGEMENT
This investigation was canied out in collaboration with Inchon
Oil Refinery Co. and was financially supported by tim company
and BK21 project.
REFERENCES
B~&er, t~, ~%ncyclopedia of Emulsion Tectmolog,f,,' Marcel Deld<er,
Inc., New York, 3, 41 (1988)9
I~-ehei; T. M., Glass, J., O'Cozmoz; A. J. and Steve~s, G. W, "Effect of
Rheology on Coalescence Rates and Emulsion Stabilits/,'A_fChE J.,
45(6), 1182 (1999)9
Foz~el; Z, SdlanbiI, F and Voi~ybu~sld, W., ~176 of Fine Dis-
perse and Long-Term Stable Oii-in-Water Emulsions by the Phase
Inversion Te~nperatm-e Method~' Ji Dispea Sci. Yeeh., 13(2), 183
(1992).
Frenkel, M., Shwatz, R. and Gati, N., ~~ Measurements as a
Tectmique for Evaluation of Water-in-(3iI Emulsion Stabili~' J Dis-
persion Sci. Yech., 3(2), 195 (1982)9
Gunji, M., Ueda, H., Ogata, M. and Nakagald, M., "Studies on the Oi1-
in4Vater Emulsiozls Stabilized with Gum-Arabic by Using the Tur-
bidity Ratio Method~'Nakagata, Eakugaku ZasshL 112(12), 906
(1992).
Kate-nan, K. R. and Garti, N., ~Spectral Absorption Meastaements for
Detelminaion of Ease of Formation and Stability of Oil in Water
Emulsioi:Ls~' Z Dispersion SoL Yeoh., 2(4), 475 (1981).
Kelker, M., ~~ Scattering of Light and Other Electromagnetic Radia-
Korean J. Chem. Eng.(Vol. 19, No. 3)
430 M.-G. Song et al.
tion~' Academi c th-ess, Ne w York (1969).
Kim, J. G. and Kim, J. D., "Turbidity and Stability o f Sonicated Leci-
thin Li posome7 HWAHAK KONGHAK, 26, 171 (1988).
Latreille, B. and Paquin, R, "Evaluation of Emulsion Stability by Cen-
trifugation with Conductivity Measurements," X Food Sot, 55(6),
1666(1990).
Lee, M. H., Tai, C. Y. and Lu, C. H., "Synthesis o f Spherical Zirconia
by Reverse Emul si on th-edpitation-Effects o f Surfactmlt Type:'
KoreanJ. Chem. Eng., 16, 818 (1999).
Melik, D. H. and Foglel; H. S., ~176 Detemm~ation of Parti-
cle Size Dis~ibutions ofColIoidal Systems~'~ Colloidlnterfaee Sei.,
92, 161 (1983).
Red@, S. R. and Fo~ea; H. S., "Emulsion Stability: Delineation of Dif-
ferent Parttcle Loss Med~a~isms~' J Colloid Interface Sr 79(1),
l a 8 (1981).
Song, M. G., Jho, S. H., Kim, J. Y and Kim, J. D., "Rapid Evaluation
of Water-in-Oii(W/O) Emulsion Stability by Turbidity Ratio Meas-
urements;',I Colloid & Int erf Sci., 230, 213 (2000).
Ta&-os, 11 F., ~176 th'indples of Emulsion Tedmol ogy and
Their Applicattoi~s~' Colloid Surfaces A.: Physieochem. Bng. Aspects,
91, 39 (1994).
Tanbe, D. E. and Shama, M. M., ~ Effect of Colloidal Parades on
Fluid-Fluid Interracial lh-op~aes and Emulsion St abi l i ~' Adv Col-
loidInterfae., 52, 1 (1994).
Treine~; C., Laure1~t, A., Fessi, H. and Orecchiozli, A. M., ~
Emulsion Investigation m the System ~ Bfij-35, Phenoxy, 2-Etha-
nol Sea-Water-Compa'ison o f Partitiozmag and Droplet Size Data~'
J. Colloid& I nt erf Sei., 129(1), 201 (1989).
May, 2002