Ultrasonics Sonochemistry 7 (2000) 187–192
www.elsevier.nl/locate/ultsonch
Emulsification processes: on-line study by multiple light
scattering measurements
B. Abismaı̈l, J.P. Canselier *, A.M. Wilhelm, H. Delmas, C. Gourdon
Laboratoire de Génie Chimique, UMR CNRS 5503 INPT/UPS, Ecole Nationale Supérieure d’Ingénieurs de Génie Chimique,
18 chemin de la Loge, 31078 Toulouse, France
Abstract
The use of ultrasound in various processes of the chemical industry has been a subject of research and development for many
years. As regards in emulsification, apart from formulation variables, power is the most important parameter. Efficiency of
emulsification processes may be followed and evaluated by measuring particle size distribution, which mainly governs the kinetic
stability of such dispersions. Unfortunately, this kind of measurement must be performed at high dilution ( low volume fraction
of dispersed phase). The present work is devoted to the on-line study of ultrasound emulsification by means of a newly developed
apparatus based on multiple light scattering, which allows us to determine average droplet diameter and its variations directly on
concentrated media. The model system was an oil (kerosene)-in-water emulsion stabilized by a polyethoxylated sorbitan
monostearate. © 2000 Elsevier Science B.V. All rights reserved.
Keywords: Back-scattered intensity measurements; Drop size distribution; Emulsification; Oil-in-water emulsion; Power ultrasound
1. Introduction
Emulsions occur naturally in the vegetable kingdom
as well as in the animal one (e.g. rubber tree latex and
milk, respectively). However, mixing two immiscible
liquids to produce oil-in-water (o/w) or water-in-oil
(w/o) emulsions almost always requires energy. Now,
emulsions play a major role in materials processing,
from metal working to textile finishing, and have found
an essential place in formulated cosmetic, pharmaceutical and food products [1]. Most of them are of the o/w
type. The quality of an emulsion is related to its stability.
This latter depends on several parameters, among which
are formulation variables (nature and amount of stabilizing agent governing interfacial tension, viscosity of the
continuous phase, density difference between continuous
and dispersed phases), droplet charge and sedimentation
or creaming rate, and process variables (order of mixing,
power output, type of contacting apparatus, flow
regime), controlling average droplet size and droplet
size distribution [2–4].
Since the first reports on ultrasound emulsification
[5–7] and the first patent granted in the 1940s [8], the
advantages of power ultrasound ( low frequency, high
* Corresponding author. Fax: +33-562-252318.
E-mail address: [email protected] (J.P. Canselier)
energy) for emulsification have been considered [9–18],
and its use in various processes of the chemical industry
has been a subject of research and development.
However, the related specific phenomena have not yet
been thoroughly explained. In Neduzhii’s experiments,
for instance, irradiation time was not taken into account
[17]. Besides, Li and Fogler’s approach led to more
significant results [19,20]. Our previous paper was
devoted to the comparison of oil-in-water emulsions
produced by mechanical agitation or power ultrasound
[9]. We are now going to recall the theory developed
for emulsion formation by mechanical agitation, expose
a model for drop size determination from multiple lightscattering measurements, and compare the average drop
size values obtained by two techniques.
2. Theory
The Weber number (We) of an oscillating droplet is
the ratio of its kinetic energy due to turbulent fluctuations, E , to the energy due to the interfacial tension,
v
E:
s
E =r u2(d )d3
(1)
v
c
E =cd2
(2)
s
1350-4177/00/$ - see front matter © 2000 Elsevier Science B.V. All rights reserved.
PII: S1 3 5 0 -4 1 7 7 ( 0 0 ) 0 0 04 0 - 7
188
B. Abismaı̈l et al. / Ultrasonics Sonochemistry 7 (2000) 187–192
We=E /E =r u2(d )d/c
(3)
v s
c
where r is the density of the continuous phase, u2(d)
c
the mean-square spatial fluctuation in the liquid velocity
over a distance d (drop size) and c the interfacial tension.
For agitated liquid–liquid dispersions, the theory of
drop breakup prediction under turbulent flow conditions
has been developed by Hinze [21]. This author suggested
that the breakup of a drop occurred when the Weber
number reaches a critical value, We , which corresponds
c
to the maximum stable drop size, d . In the turbulent
max
isotropic regime, according to Kolmogorov’s theory
[22], u2(d ) is function of the average power dissipated
per mass unit, e:
u2(d )=(ed )2/3.
(4)
Resolving Eq. (3) for d=d
yields:
max
d =C (r /c)−3/5e−2/5
(5)
max
1 c
where C is a constant generally of the order of unity.
1
Classical techniques of drop size measurement give d
32
(the Sauter diameter), usually proportional to
d
(d /d =C ) [23], so that d is also correlated
max 32 max
2
32
by Eq. (5).
3. Experimental
A new technology based on the analysis of multiple
light scattering has been turned to account. The ‘Online TURBISCAN’ includes an optical sensor composed
of a pulsed near infrared light source (wavelength
l =850 nm) and two synchronous detectors. The transIR
Table 1
Characteristics of surfactantsa
Surfactant
c (mN/m)
s
c (mN/m)
i
HLB
MONTANOX 60
C EO
12 8
43.5
32.2
9.5
4.5
14.9
11.8*
a c =surface tension of aqueous solution at the cmc; c =interfacial
s
i
tension between kerosene and aqueous solution. * Calculated.
mission detector receives the light which goes through
the sample (0° from the incident beam), while the backscattering detector receives the light scattered by the
sample at 135° from the incident beam.
In dense dispersed media, photons undergo many
scattering events before escaping the medium and entering the TURBISCAN back-scattering detector. Multiple
light scattering thus contributes significantly to the measured back-scattered flux [24]. Let us introduce the
photon transport length, l1:
l1=l/(1−g)=2d/3w(1−g)Q
(6)
S
where l is the photon mean free path, g(d, l , n , n )
IR p f
and Q (d, l , n , n ) optical parameters given by the Mie
S
IR p f
theory, d the particle average diameter and w the droplet
volume fraction. Statistical models and numerical simulations were developed to describe the back-scattered
flux [depending on l(d, w, Q )] and the transmitted flux
S
[depending on l1(d, w, g, Q )] measured by the
S
TURBISCAN. It allows us to derive the average droplet
diameter, d from a knowledge of the droplet volume
fraction, w.
Emulsification was carried out using a low frequency
Fig. 1. Experimental equipment (‘On-line’ TURBISCAN ): 1, wattmeter; 2, generator; 3, horn; 4, reactor; 5, pump; 6, detector; 7, indicator;
8, computer.
B. Abismaı̈l et al. / Ultrasonics Sonochemistry 7 (2000) 187–192
Fig. 2. Volume drop size distribution for various ultrasonic power
values P ( W ) (t=30 s, c=10 g/l, w=0.25, c =9.5 mN/m, S=
i
MONTANOX 60).
189
were used: a polyethoxylated sorbitan monostearate
(MONTANOX 60) and a commercial octaethoxylated
lauryl alcohol (~C EO ) (gifts from SEPPIC ). In
12 8
addition to a previous presentation of emulsion components [9,10], their properties are listed in Table 1. All
emulsions were prepared at room temperature (~20°C ).
$ Firstly, the effect of ultrasound power on the droplet
size (Sauter diameter) of oil (kerosene)-in-water emulsions was studied by means of a laser diffractometer
(Malvern MasterSizer S). In this case dilution was
necessary.
$ Secondly, the same system was analyzed as such by
means of the TURBISCAN. The equipment is shown
schematically in Fig. 1.
4. Results and discussion
4.1. Analysis by dilution method (Malvern)
Fig. 3. Variation of d as a function of power dissipated per mass unit
32
( Exp. No. 1).
(20 kHz) ultrasound horn coupled with an appropriate
generator ( Vibracell ). Kerosene (required HLB=12)
[25] was the oil phase. Two types of surfactant (S)
This part consists in studying the effect of ultrasound
power on emulsification with two oil volume fractions
(25% and 37.5%) and two surfactants (MONTANOX
60 and C EO ). Before ultrasound emulsification, the
12 8
mixture was pre-emulsified during at least 10 min by
mechanical agitation, down to a Sauter diameter
between 70 and 100 mm ( Fig. 2).
The volume drop size distribution shows the important effect of ultrasound, even for low power values
(after 30 s with a power of 30 W, it goes down from
100 mm to 0.7 mm). Applying Kolmogorov and Hinze’s
theories to ultrasound phenomena, we observe that,
according to those correlations, ultrasound emulsifica-
Fig. 4. Back-scattered intensity vs. time for different power values (v=800 ml, c=10 g/l, w=0.25, c =9.5 mN/m, S=MONTANOX 60).
i
190
B. Abismaı̈l et al. / Ultrasonics Sonochemistry 7 (2000) 187–192
Table 2
C values under various emulsification conditions
1
4.2. Analysis without dilution (TURBISCAN)
Exp. no.
t
(s)
C
(g/l )
w
c
i
(mN/m)
103C
1
[Eq. (5)]
1 (MONTANOX )
2 (MONTANOX )
3 (C EO )
12 8
30
60
30
10
10
10
0.25
0.375
0.25
9.5
9.5
4.5
7.09
6.88
7.10
tion admits the same exponent value for e (−0.4) in Eq.
(5) as mechanical agitation (Fig. 3). These results have
to be compared with Walstra’s ones showing the same
dependence of d upon the net energy input E (MJ/m3)
i
provided by an Ultra-Turrax rotor–stator device (batch
2 min) or an ultrasonic transducer (continuous flow)
when d is plotted vs. E [26 ].
43
i
However, the constant C found in this work for
1
ultrasound emulsification (~7×10−3) ( Table 2) is much
lower
than
that
for
mechanical
agitation
(~36.7×10−3). This means that, for a constant power
value, insonation, mainly involving cavitation, is far
more efficient than mechanical agitation, yielding smaller
droplets. Moreover, a much lower dissipated energy has
been evidenced in the case of ultrasound generation [9].
In this second type of experiments, the same oil
(kerosene)-in-water system with MONTANOX 60 has
been used. A 800 ml volume of emulsion was prepared
at room temperature (~20°C ). Pre-emulsification was
afforded by the pump (5) (Fig. 1). The volume fraction
of kerosene was 0.25. The parameters studied were the
time of emulsification and the power input. The backscattered intensity percentage (BS%) vs. time, measured
with the TURBISCAN, leads directly to the photon
mean path length value [l1 (mm)] as a function of time
( Figs. 4 and 5). With this information and a numerical
simulation [24], the average drop size, d, may be calculated [Fig. 6(a) and (b)].
Therefore, the TURBISCAN provides an efficient
way to determine average droplet diameters in emulsions. The results of those measurements are of the same
order of magnitude as the values obtained with the drop
size analyzer ( Table 3).
Compared with classical drop size measurements, this
new technique is particularly useful when dilution is not
convenient (e.g. w/o emulsions, requiring large amounts
of organic solvent) or when this operation modifies
emulsion properties or stability (e.g. very concentrated
Fig. 5. l1 vs. time for different power values (same conditions as Fig. 4).
Table 3
Lowest droplet diameters (‘equilibrium’ values) from MasterSizer S and TURBISCAN on-line
P (W)
62
91
125
173
225
d (mm) Malvern
32
d (mm) TURBISCAN
0.50±0.01
0.46±0.20
0.43±0.01
0.47±0.20
–
0.375±0.20
0.31±0.01
0.39±0.10
0.30±0.01
0.39±0.10
B. Abismaı̈l et al. / Ultrasonics Sonochemistry 7 (2000) 187–192
191
Fig. 6. (a) d vs. time for different power values; (b) enlargement of the box showing ‘equilibrium’ values (same conditions as Fig. 4).
emulsions). Fig. 7 shows the variations of d (‘equilibrium’ values) vs. e. Application of Eq. (5) yields a value
of 8.38×10−3 for C (Fig. 7), consistent with that
1
obtained previously from ultrasound process
(7×10−3). But fitting the curve of Fig. 7 shows that d
rather decreases as e−0.16. This is probably due to the
low size of the drops.
5. Conclusions
Several findings are reported in this work concerning
the effect of power in ultrasound emulsification and
on-line analysis of emulsions.
Fig. 7. Variation of d as a function of power (same conditions as
Fig. 4).
192
$
$
$
$
B. Abismaı̈l et al. / Ultrasonics Sonochemistry 7 (2000) 187–192
Very small drop size can be achieved with ultrasound
even at low power provided that a ‘pre-emulsification’
is effected by simple agitation.
The effects of power in ultrasound emulsification and
in mechanical turbulent isotropic regime are similar.
The correlation constant for ultrasound is 7×10−3.
As regards emulsion drop size analysis, both techniques (with and without dilution) yield average
diameters of the same order of magnitude.
Knowing the efficiency of ultrasound in making very
fine liquid–liquid dispersions, a possible application
could concern diffusion-controlled reactions.
Acknowledgements
The authors thank Formulaction S.A., especially K.
Puech and O. Mengual, for their help while operating
their specific equipment (On-line TURBISCAN ).
References
[1] J.C. Johnson, Emulsifiers and Emulsifying Techniques, Noyes
Data, Park Ridge, NJ, 1979.
[2] P. Becher, Emulsions, Theory and Practice, second ed.Reinhold,
New York, 1965.
[3] P. Becher, Encyclopedia of Emulsion Technology Vol. 1 Marcel
Dekker, New York, 1983.
[4] S.E. Friberg, S. Jones, in: J.I. Kroschwitz ( Ed.), Kirk-Othmer
Encyclopedia of Chemical Technology, fourth ed., Vol. 9, Wiley,
New York, 1994, pp. 393–413.
[5] R.W. Wood, A.L. Loomis, Philos. Mag. 4 (1927) 417–436.
[6 ] B. William, T. Richards, J. Am. Chem. Soc. 51 (6) (1929)
1724–1729.
[7] C. Bondy, K. Söllner, Trans. Faraday Soc. 31 (1935) 835–843.
[8] Swiss Patent 394,390, 1944.
[9] B. Abismaı̈l, J.P. Canselier, A.M. Wilhelm, H. Delmas, C.
Gourdon, Ultrasonics 6 (1999) 75–83.
[10] B. Abismaı̈l, Thesis, INP, Toulouse, 1999.
[11] S. Mujumdar, P.S. Kumar, A.B. Pandit, Ind. J. Chem. Technol.
4 (1997) 277–284.
[12] S.A. Neduzhii, Colloid Zh. 23 (4) (1961) 374–378.
[13] S.A. Neduzhii, Akust. Zh. 7 (1) (1961) 99–100.
[14] S.A. Neduzhii, Akust. Zh. 7 (2) (1961) 265–266.
[15] S.A. Neduzhii, Akust. Zh. 8 (4) (1962) 481–482.
[16 ] S.A. Neduzhii, Sov. Phys. Acoust. 9 (2) (1963) 195–196.
[17] S.A. Neduzhii, Sov. Phys. Acoust. 10 (4) (1965) 390–397.
[18] D.M. Skauen, B. Misek, J. Am. Pharm. Assoc. 47 (1958) 32–39.
[19] M.K. Li, H.S. Fogler, J. Fluid Mech. 88 (3) (1978) 499–511.
[20] M.K. Li, H.S. Fogler, J. Fluid Mech. 88 (3) (1978) 513–528.
[21] J.O. Hinze, AIChE J. 1 (1955) 289–395.
[22] A.N. Kolmogorov, Dokl. Akad. Nauk SSSR 66 (1949) 825–828.
[23] F.B. Sprow, Chem. Eng. Sci. 22 (1967) 435–442.
[24] P. Snabre, O. Mengual, G. Meunier, in: 2e Congrès Mondial de
l’Emulsion, Bordeaux, 23–26 September (1997) .
[25] W.C. Griffin, in: M. Grayson ( Ed.), Kirk-Othmer Encyclopedia
of Chemical Technology, third ed., Vol. 8, Wiley, New York, 1979,
pp. 900–930.
[26 ] P. Walstra, in: P. Becher ( Ed.), Encyclopedia of Emulsion Technology Vol. 1, Marcel Dekker, New York, 1983, pp. 57–127.