Anda di halaman 1dari 12

Fluid Phase Equilibria 362 (2014) 288299

Contents lists available at ScienceDirect

Fluid Phase Equilibria


journal homepage: www.elsevier.com/locate/fluid

A comprehensive framework for surfactant selection and design


for emulsion based chemical product design
Michele Mattei a,b , Georgios M. Kontogeorgis b , Raqul Gani a,
a
Computer Aided Process-Product Engineering Center (CAPEC), Department of Chemical and Biochemical Engineering, Technical University of Denmark,
Sltofts Plads, Building 229, DK-2800 Kgs. Lyngby, Denmark
b
Center for Energy Resources Engineering (CERE), Department of Chemical and Biochemical Engineering, Technical University of Denmark, Sltofts Plads,
Building 229, DK-2800 Kgs. Lyngby, Denmark

a r t i c l e

i n f o

Article history:
Received 23 July 2013
Received in revised form 11 October 2013
Accepted 16 October 2013
Available online 25 October 2013
Keywords:
Product design
Emulsions
Property prediction
Cloud point

a b s t r a c t
The manufacture of emulsied products is of increasing interest in the consumer oriented chemical industry. Several cosmetic, house-hold and pharmaceutical products are in the emulsied form when sold
and/or they are expected to form an emulsion when used. Therefore, there is a need for the development
of a methodology and relevant tools in order to spare time and resources in the design of emulsionbased chemical products, so that the products can reach the market faster and at a reduced cost. The
understanding and modeling of the characteristic behavior of emulsions and their peculiar ingredients
is consequently necessary to tackle this problem with computer-aided methods and tools. A comprehensive framework for the selection and design of surfactants, the main responsible for the formation
and the stability of emulsions, is presented here together with the modeling of the cloud point, a keyproperty of nonionic surfactants, with a group-contribution model. The mathematical formulation of a
standard product design problem is presented, together with the list of both the pure component properties (related to nonionic surfactants) and the mixture properties (relevant to the overall products as
an emulsion) needed for the solution of the design algorithm. These models are then applied together
with established predictive models for pure component properties of ionic surfactants and for standard
mixture properties such as the density, the viscosity, the surface and the interfacial tension, but also the
type of emulsion expected (through the hydrophiliclipophilic balance), and its stability (through the
hydrophiliclipophilic deviation), forming a robust chemical product design tool. The application of this
framework is highlighted for the design of some emulsion based chemical products.
2013 Elsevier B.V. All rights reserved.

1. Chemical product design


Recently, a substantial shift is observed from materials valued
for their purity, to materials sold for their performance behavior [1,2]. To meet these challenges, on a global and local scale,
while remaining protable and maintaining sustainable growth,
there has been an increasing interest in the formulation and solution of product design problems [3]. The chemical product tree
shown in Fig. 1 gives an idea of the size of this shift: the roots
of the tree consist in a limited number of Raw Materials which
are processed to obtain the commodity products (Basic Products).
Specialty chemicals (Intermediate Products) are then manufactured
from the commodities and nally the leaves of the tree represent a large portfolio of higher value products (Rened Chemicals
& Consumer Products) obtained by processing and/or combining
the chemicals of the previous product classes. As one ascends the

Corresponding author. Tel.: +45 45 25 28 82; fax: +45 45 93 29 06.


E-mail address: rag@kt.dtu.dk (R. Gani).
0378-3812/$ see front matter 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.uid.2013.10.030

product tree, the number of products belonging to each category


grows exponentially from around 10 for the raw material class, up
to almost 30,000 in the last class of higher value added products.
This last class is composed of formulations, devices and technology
based consumer goods. Formulated products include pharmaceuticals, paints, food, cosmetic, detergents, pesticides, in which 5 to
more than 20 ingredients are usually present, representing a wide
range of chemical compounds such as polymers, surfactants, solids,
solvents, pigments, and aromas [4]. The common practice, in the
development of such products, is still the experiment-based and
trial-and-error approach. However, a systematic procedure, able to
design a higher added value product with enhanced product qualities, represents an efcient alternative, with respect to time and
resources, speeding up the product development.
1.1. Formulation design
Many chemical-based personal care products of everyday life
such as sun lotions, shower creams, and insect repellents are liquid
formulations, while examples of non-personal care products are

M. Mattei et al. / Fluid Phase Equilibria 362 (2014) 288299

289

design problem, with the aid of adequate property models and


computer-aided tools. Raman and Maranas [7] addressed the problem incorporating topological indices for correlating the necessary
physico-chemical properties, while Chemmangattuvalappil et al.
[8] applied combined property clustering and group-contribution
techniques. Teixeira et al. [9] directed their attention toward structured products (more specically microencapsulate perfumes for
textile application), while Charpentier [10] focused on the multiscale problem generated by the introduction in the methodology
of economic, social and environmental constraints. Fig. 2 shows
the work-ow diagram of the computer-aided design/verication
stage, based on dene target match target paradigm as presented by Conte et al. [5], highlighting input, output and tools used
for each step. The methodology employs the reverse design technique. The dened target properties of the product are then the
known variables and input for the property models. Appropriate
property models are needed to estimate the target properties of the
candidates so that they are evaluated and then accepted or rejected.
At the same time the mixture compositions that satisfy the product
constraints are determined, using suitable mixture property models as well as phase stability algorithms. As shown in Fig. 2, if in any
task a solution is not found, it is possible to return to a previous
task to rene the problem denition.

Fig. 1. The chemical product tree: classication of chemical-based products [3].

1.2. Emulsion design


paints, pesticides and drugs. These can be classied as consumeroriented products since their needs, on the basis of which they are
designed, are dened by the consumers. Therefore these products
need to satisfy multiple needs of the consumers [5]. A sunscreen
lotion, for example, must provide protection against sunburns and
skin cancer, but must also prevent skin aging, and be, for example,
long lasting, safe, easily applicable, with good sensorial properties
(color and odor) [6]. Because a single chemical is unlikely to satisfy
these multiple needs, a blend of several chemicals is usually sought.
A formulation may then contain materials from different classes of
chemicals, such as polymers, surfactants, solvents, pigments and
aromas. These classes of chemicals are usually classied as follows
[5]:
Active ingredients: these chemicals are the most important ones in
the formulation, because they satisfy the main needs of the product, thus dening the function of the product itself. For example,
the function of a sunscreen lotion is to protect the skin against
the UV radiation.
Solvent mixture: it is usually present in high concentration in the
formulation and has the function of dissolving the active ingredients and other chemicals in the formulation, ensuring the product
to be a single liquid phase and to be properly delivered. The solvent mixture must evaporate after application.
Additives: these chemicals are usually present in low concentration and they satisfy the secondary needs of the product,
enhancing the end-use product properties. Examples are pigments and aromas, to enhance the sensorial properties of the
formulation.
The presence of several classes of chemicals to be included in
the formulation design leads to the necessity of a step-by-step
hierarchical design methodology, in order to avoid any combinatorial explosion due to the high number of possible candidate
formulations to be generated and screened, systematically and efciently, while at the same time excluding blind trial-and-error
solutions. Several methodologies and frameworks have been developed, in order to address the need for the solution of a formulation

Formulations can also have other physical forms [11]: suspensions containing insoluble chemicals dispersed in the liquid
with the help of a dispersant; emulsions where solid constituents
have been emulsied through selected emulsiers together with
solvents and additives; solid products such as pharmaceutical
tablets or soap bars. In chemical product design, Cussler and Moggridge [2] distinguish between commodities, chemical devices,
molecular products and microstructured products, where the term
microstructure refers to a chemical organization on the scale of
micrometers, belonging to the colloidal domain and incorporating
polymer solutions, foams, gels and emulsions. The performances of
such products are related not only to the presence of active ingredients and additives in the formulation, but also to the products
structural and material properties [12]. Among the microstructured
products, emulsied products are the most relevant, particularly in the food and cosmetic industries. Emulsions are dened
as mixtures of two normally immiscible liquids, kinetically stabilized by emulsiers (most often surface active agents, better
known as surfactants) that lie on the interface between the two
phases. Active ingredients and additives are usually dissolved in
the continuous and/or dispersed phases, according to the needs
of the products. Bernardo and Saraiva [13] proposed a simultaneous approach to address product and process design, with
special attention to cosmetic emulsions, while Bagajewicz et al. [14]
extended a generic approach [15] to consider price-competitive
markets. Recently, a systematic procedure, which is applicable to
the design of emulsied formulated products, has been proposed
by Mattei et al. [16] and it is further extended in this work (see
Fig. 2).
For emulsion-based chemical product design, the solvent(s)
design task (Task 3a) provides as output two non-miscible liquid
phases and an additional task (Task 3b: Surfactant(s) design) is
needed. Necessarily, some of the models applied for the denition
of the target properties might differ when considering an emulsied product, rather than a homogeneous formulation. In particular,
surfactants are key chemicals in most emulsied formulations and
a wide range of peculiar properties need to be considered when
designing or selecting chemicals such as surfactants [17].

290

M. Mattei et al. / Fluid Phase Equilibria 362 (2014) 288299

Fig. 2. Work-ow diagram for the computer-aided stage of the formulation design methodology.

2. Surfactant selection and design

form of hi (x), gi (x) and FOBJ is different and the property models
needed are also different in each case.

A surfactant design problem does not differ from any other


molecular and/or mixture/blend design problem and it can therefore be described through the following generic mathematical
representation [3]:
FOBJ = max{C T y + f (x)}

(1)

hi (x) = 0

(2)

h2 (x) = 0

(3)

h3 (x) = 0

(4)

l1 g1 (x) u1

(5)

l2 g2 (x) u2

(6)

l3 By + Cx u3

(7)

In the above equations, x represents the vector of continuous


variables (such as mixture compositions), y the vector of binary
integer variables (such as compound identity), hi (x) is a set of equality constraints (related to molecular structure generation, chemical
feasibility rules, etc.), gi (x) is a set of inequality constraints (related
to environmental constraints and/or special property constraints)
and FOBJ is the objective function (to be maximized) on the basis of
which the design choices are taken. In the specic case of surfactant design, the numerical constraints presented in the equations
above depend on the problem denition, such as on the need for
the surfactant system into the nal product. In fact, the main application of this category of chemical is the emulsion stabilization,
while surfactants can also be found in commercial products to
form microemulsions (particularly common in the pharmaceutical industry) or as active ingredients (as previously dened) in
detergents. Other formulated products which are not in the emulsied form have key-properties strongly inuenced by the surfactant
content: surfactants are usually added to formulations with the
goal of preventing the re-deposition of dispersed chemicals (as in
toothpastes) or of reducing the surface tension of the product (as in
enhanced oil recovery uids). Each of the various needs generate a
different product design problem and while it can be mathematicallydescribed with the same set of equations presented above, the

2.1. Surfactant property models


Surfactants are characterized by an amphiphilic nature, which
means that a part of them is hydrophilic (water-like), while another
part is hydrophobic, or lipophilic (oil-like). In order to describe
their behavior in relation to two non-miscible phases and the range
of temperatures at which they are active, some non-conventional
properties, such as cloud point and critical micelle concentration,
are needed. It should be noted, however, that model-based product design methodologies for emulsied products have not been
yet developed to a level as those for homogeneous formulations,
and so predictive models for some of the needed properties are
missing. However, although many of these properties are strictly
mixture properties since they refer to the mutual behavior of surfactants in water mixtures, they can be modeled as pure component
properties, depending only on the molecular structure of the surfactant involved, since either the temperature or the composition
are usually kept constant. Hence these properties are modeled as
primary properties and can be, as a rst approximation, estimated
using group-contribution methods. The set of properties needed
for surfactant selection and design in an emulsion-based chemical
product design problem and the models used for their estimation
are summarized in Table S1 [18,2026] for primary properties and
in Table S2 [56,2729] for secondary properties, in Supplementary
material.
The models used to estimate the properties may be classied, for
each class of properties, into those that are predictive by nature and
those that are not. For example, estimating properties only from
molecular structural information involves predictive models, such
as the group-contribution methods, while estimating properties
from compound specic coefcients involves the use of correlation which are not predictive by nature. In chemical product design,
both types of models are needed. During the evaluation of candidate
products, the models need to be predictive and computationally
fast and cheap, while during the verication of a small number
of candidates, correlative models may be used, if the correlation
coefcients are available. During the evaluation stage, the models
need to be, at least, qualitatively correct, while during the verication stage, the models also need to be quantitatively correct.
Group-contribution methods are extensively considered here since
they only need the molecular structure of the pure component and
they exhibit a good accuracy together with a wide range of applicability. In the Marrero and Gani group-contribution model [18], the

M. Mattei et al. / Fluid Phase Equilibria 362 (2014) 288299

property estimation is performed at three levels, and the property


prediction model has the form reported below:
f (X) =

Ni Ci +

Mj Di +

Ok Ek

(8)

where f(X) is a function of the property X and it may contain


adjustable model parameters depending on the property involved.
In (8), Ci is the contribution of the rst-order group of type-i that
occurs Ni times. Dj is the contribution of the second-order group of
type-j that occurs Mj times. Ek is the contribution of the third-order
group of type-k that occurs Ok times. The denition and identication of the second-order groups has a theoretical basis, where
the principle of conjugation has been employed [19]. The criteria
used for the identication of third-order groups are analogous to
those used for second-order groups except for the types of compounds intended to be represented [18]. The role of the second
and third-order groups is to provide more structural information
about the portions of the molecular structure of a compound,
where the description through the lower-order groups is insufcient. Through this addition, obstacles as partial description of the
proximity effects and the lack of distinction between isomers can
be overcome. The ultimate objective of this multilevel scheme is
to enhance the accuracy, reliability and the range of application
of the model [18]. For determination of contributions Ci , Dj and
Ek , Marrero and Gani [18] suggested a multi-level approach. As
an alternative to the step-wise regression method, a simultaneous
regression method can be applied, in which the regression is performed by considering all the terms containing rst, second and
third-order groups in a single regression step.
When addressing the reverse problem with group-contribution
models, computer-aided tools are necessary in order to reliably
and quickly identify the chemical satisfying the target properties
dened. It is also necessary, then, to provide connectivity rules so
that the groups collected can be combined in all the possible legal
combinations. When applying to surfactants, this stage becomes
very relevant since it is not only necessary to make sure to connect groups so that no free attachments are left in the molecules
generated, but also that two distinct moieties (the hydrophilic and
lipophilic parts) are generated. This introduces a number of extraconstraints to be satised in order to generate acceptable surfactant
candidates.
2.2. Mixture properties
The design and selection of surfactants for chemical product
design need pure component properties as well as mixture properties. In fact, the interaction of the chosen surfactant with the other
constituents of the product needs to be evaluated; in the specic
case of emulsied products, particularly, the interaction of surfactants with both the water and the oil phases are of great importance.
Proper mixture property models which are able to relate the formulation properties to those of the single ingredients are needed.
A comprehensive list of the models used is provided in Table S3
[2932] in Supplementary material.
As listed in Table S3, for some mixture properties linear mixing
rules can be used to predict the property. For the generic mixture
property , the mixture property model based on a linear mixing
rule is described by the following equation:
=

xi i

(9)

where xi is the composition of compound i in the mixture, and  i


the pure component property.
Linear property models give good predictions of mixture properties for chemical systems characterized by negligible excess

291

properties of mixing. For those chemical systems having large


excess properties of mixing, more detailed models are needed.
Emulsions are systems far from the ideality, and therefore dedicated models for the estimation of some of the mixture properties
(listed in the right column of Table S3) are necessary. Since these
models are quite complex, they can hardly be employed as models
for screening of alternative designs, but they should be rather used
in the verication stage, or on a second step, when the search space
for the candidates has been reduced by the application of proper
constraints on other mixture properties.
3. Modeling of the cloud point of surfactants with a
group-contribution method
One of the surfactant-related pure component properties
considered necessary for the development of a model-based
methodology for surfactant design is the cloud point, sometimes
called the cloud temperature. This property is specic of mixtures between water and nonionic surfactants and it is dened
as the temperature at which the mixture starts to phase separate and two phases appear, thus becoming cloudy [33]. This
phenomenon is of particular relevance for surfactants containing polyoxyethylene chains (thus nonionic surfactants), exhibiting
reverse solubility versus temperature behavior. This is clearly visible in Fig. 3, where a standard phase behavior of an aqueous mixture
of a polyoxyethylene-based nonionic surfactant is reported.
In Fig. 3, different regions can be recognized: L1 identies an
aqueous surfactant solution where the surfactant is organized in
ordinary or reverse spherical micelles; W represents a very diluted
surfactant solution (around the critical micelle concentration); S
indicates the presence of solid surfactant; L , H1 and V1 , instead,
are regions where the surfactant is aggregated in uncommon
structures such as, respectively, lamellar, normal hexagonal and bicontinuous cubic structures. The last three regions are sometimes
grouped together as viscous meso-phases, since their rheological properties and behavior are substantially different from those
of the ordinary and reverse spherical micellar solution. The line
determining the W + L1 two-phase area is also known as the cloud
point line. It is evident that, as a mixture property, the cloud point
does not depend only on the system considered, but it is inuenced by the surfactant content. However, it is common practice
to dene as cloud point the numerical value assumed by the cloud
point curve at a surfactant weight percentage of 1%, measured by
visual observation method: the temperature at which the visible
solubility changes to cloudy over a range of 1 C or less is taken as
cloud point [35].
Several efforts have been made in order to develop a model
to predict the cloud point, only on the basis of the molecular
structure of the surfactant involved and the quantitative structureactivity relationship (QSPR) models have been extensively applied
[3537]. The authors, on the other hand, are not aware of any groupcontribution based method developed in relation to this property.
These methods, however, apply very well to chemical process and
product design because they can provide accurate predictions without being computationally demanding. Moreover, they can be used
in computer-aided molecular design because they employ the same
building blocks for molecular representation [24]. In this work we
have applied the Marrero and Gani GC-method [18] to this property.
3.1. Data-set
An original data-set consisting of 86 nonionic surfactants have been collected from different sources [3538].
The data set contains linear alkyl, branched alkyl, alkyl
phenyl ethoxylates, carbohydrate-derivative ethoxylates, alkyl

292

M. Mattei et al. / Fluid Phase Equilibria 362 (2014) 288299

100

W + L1

Temperature [C]

80

60

L
40

L1

20

V1

H1

S
0
0

10

20

30

40

50

60

70

80

90

100

Surfactant weight %
Fig. 3. Phase diagram of an aqueous mixture of a polyoxyethylene-based nonionic surfactant over the temperature range 0100 C Data are relative to the system between
water and dodecyl-esaethylene oxide.
Source: Data are taken from [34].

3.2. Model development


In order to determine the most suitable form of f(X) of the constitutive equation of the Marrero and Gani method, as in Equation
8, it is necessary to observe the trend of the experimental data
of the property to be estimated as a function of the main representative groups of the chemicals under investigation. Considering
the largest family of nonionic surfactants: the linear alkyl ethoxylates, the trend of the cloud point as a function of the number of
ethoxylate groups (CH2 CH2 O) in the hydrophilic chain is analyzed,
as shown in Figs. S1 and S2 in the Supplementary material.
As seen in Fig. S1, the dependence of the cloud point on the
number of ethoxylate groups of linear alkyl ethoxylates is not linear.
On the other hand, Fig. S2 shows that the dependence of the square
of the cloud point is linear. This justies then the choice of the form
of f(X), as in following Eq. (10):
CP2 =


i

Ni Ci +


j

Mj Dj +

Ok Ek

(10)

where the cloud point is expressed in K. Given the equation above,


in order to represent the remaining 72 compounds, 13 rst order

groups and 1 second order group are needed, according to the original set of parameters by Marrero and Gani [18]. The results of
the parameter estimation step performed through the step-wise
regression method are illustrated in the parity plot of Fig. 4.
The results in Fig. 4 indicate that the accuracy of the Marrero and
Gani GC-methods using only rst and second order groups is not
satisfactory, as quantied in Table S4 in Supplementary material.
In particular, the maximum absolute errors (column AADmax )
are too high for many categories of surfactants considered to consider the model reliable enough to be implemented in the product
design methodology. It is therefore considered necessary to include
new third order groups in the set of parameters, in order to improve
the performances of the method, as described in [24,39], in particular relation to those compounds for which the correlation indices
were poor: branched alkyl ethoxylates and carbohydrate-derivate
ethoxylates primarily. According to Marrero and Gani [18], in fact,
the second order groups are strictly dened and one cannot arbitrarily add new second order groups as one can do with third
order groups, instead. A step-by-step systematical data-error analysis as in [24] has been performed, generating 5 new third order
groups taking care of complex structures peculiar of the available

120
100

Calc. Cloud Point [C]

polyoxyethylenepolyoxypropylene copolymers and ethoxylated


amides. All experimental data are measured by visual observation
method in 1% aqueous surfactant solutions and are reported in
Table 1, divided in different classes.
Before applying the Marrero and Gani CG-method to the dataset chosen for the parameter estimation step, it is necessary to
analyze the matrix of group occurrences to make sure that each
groups describes at least two of the surfactants presents in the dataset. A single occurrence would actually distort the performance of
the model, leading to a perfect match for the compounds with those
groups, providing uncertain extrapolation capabilities. Moreover,
some of the data of Table 1 are excluded from the data-set since
their experimental value for the cloud point is inconsistent with
other values and are therefore identied as outliers. The outliers
are identied as they are inconsistent with the assumption that
the cloud point of linear alkyl ethoxylates increases with increasing length of the ethoxylated chain and with decreasing length of
the carbon chain. These surfactants whose CP values are excluded
are highlighted in gray in Table 1.

80
60
40
20
0
0

20

40

60

80

100

120

Expt. Cloud Point [C]


Fig. 4. Parity plot relative to the correlation of 72 data-points regarding cloud point
(in C) of nonionic surfactants using the Marrero and Gani GC-method with only
rst and second order groups.

M. Mattei et al. / Fluid Phase Equilibria 362 (2014) 288299

293

Table 1
List of the original data-set of experimental cloud point (in C) of nonionic surfactants (1% weight percentage). Distinction in classes is based only on the molecular structure,
each of which is described. Compounds highlighted in gray have been excluded from the parameter regression step.
Code

CPexp ( C)

Code

CPexp ( C)

Code

CPexp ( C)

Linear alkyl ethoxylates


CnEm
C4E1
C5E2
C6E2
C6E3
C6E4
C6E5
C6E6
C7E3
C8E3
C8E4
C8E5
C8E6
C8E8
C8E9
C8E12
C9E4
C9E5

44.5
36
0
40.5
63.8
75
83
27.6
7
38.5
58.6
72.5
96
100
106
32
55

Cn H2n+1 O(C2 H4 O)m H


C9E6
C10E4
C10E5
C10E6
C10E7
C10E8
C10E10
C11E4
C11E5
C11E6
C11E8
C12E4
C12E5
C12E6
C12E7
C12E8
C12E9

75
19.7
41.6
60.3
75
84.5
95
10.5
37
57.5
82
6
28.9
51
64.7
77.9
87.8

C12E10
C12E11
C13E5
C13E6
C13E8
C14E5
C14E6
C14E7
C14E8
C15E6
C15E8
C16E6
C16E7
C16E8
C16E9
C16E10
C16E12

95.5
100.3
27
42
72.5
20
42.3
57.6
70.5
37.5
66
35.5
54
65
75
66
92

TCnEm
IC6E6
IC10E6

78
27

(C(n2)/2 Hn1 )2 CHCH2 O(C2 H4 O)m H (n = 6, 10)


(C(n1)/2 Hn )2 CHO(C2 H4 O)m H (n = 13)
(C(n1)/3 H(2n+1)/3 )3 CO(C2 H4 O)m H
TC10E7
22
IC13E9
35

TC13E9
TC16E12

34
48

Phenyl alkyl ethoxylates


CnPEm
TC8PE9
C8PE7
C8PE9
C8PE10
C8PE13

64.3
22
54
75
89

Cn H2n+1 C6 H4 O(C2 H4 O)m H


C9PE8
C9PE9
C9PE10
C9PE12
C9PE13

C12PE9
C12PE11
C12PE15

33
50
90

C12E5P4

29.8

C11COOE8

53

C12SE3

44

AGM-13(3)

29

Branched alkyl ethoxylates


ICnEm

34
56
75
87
89

Alkyl polyoxyethylenepolyoxypropylene copolymers


CnEmPk
22.1
C12E4P5

Cn H2n+1 C6 H4 O(C2 H4 O)m (C3 H6 O)k H


C12E3P6
10.6

Carbohydrate-derivate ethoxylates
CnCOOEmC
CnCOOEm
C9COOE7C
44
65
C9COOE10C

Cn H2n+1 COO(C2 H4 O)m CH3


Cn H2n+1 COO(C2 H4 O)m H
C9COOE12
C11COOE6

Ethoxylated amides
CnGEm
CnAEm
CnSEm
C12GE2
C12GE3

78
46

Cn H2n+1 NHCH2 COO(C2 H4 O)m H


Cn H2n+1 NHCHCH3 COO(C2 H4 O)m H
Cn H2n+1 NCH3 CH2 COO(C2 H4 O)m H
C12GE4
75
C12AE3
22.5

Alkyl branched ethoxylates


AGM-n(3)
AGM-7(3)

34

Cn H2n+1 CH(O(C2 H4 O)3 H)2


AGM-11(3)

74
54

30

Source: Data from [3538].

data-set. The introduction of structural parameters, in order to take


into account the peculiar molecular assemblies of the nonionic surfactants considered, as third order groups is necessary since the
original sets of rst and second order groups of the Marrero and
Gani method cannot be arbitrarily modied.
3.3. Results and discussion of results
Once the new set of groups has been identied, a nal parameter regression is performed, where all the group contributions are
estimated simultaneously. The parameter values are reported in
Table 2, while the performances of the improved methods are given
in Fig. 5, as a parity plot and in Table S5 in Supplementary material as statistical indices. The simultaneous approach gives better
model performances, compared to the step-wise approach, and it is
therefore preferred. Obviously, when this approach is chosen, the

absolute values of the third order contribution might be comparable or even exceed those of the rst and second order groups.
It has to be noted that the only second-order group used in this
work, AROMRINGs1 s4 , has a zero contribution since the regression performed with the simultaneous method did not nd any
non-zero value for it. Fig. S3 in Supplementary material highlights
the effect of the addition of dedicated third-order groups in the
reduction of the absolute error of the model. In particular, it can
be noticed that only 2 calculated values clearly differ of more than
10 K after the introduction of the third order groups, while in the
regression with only rst and second order groups 11 correlations
show high absolute errors.
By comparing the results before and after the addition of the new
dedicated third order groups, it can be seen that evident improvements have been achieved. In particular, after the addition of the
third order groups, linear and branched alkyl ethoxylates show

294

M. Mattei et al. / Fluid Phase Equilibria 362 (2014) 288299

Table 2
Marrero and Gani group denition and contributions after the regression based on 72 experimental data of cloud point.
1st order group (i)

Ci (K2 )

2nd order group (j)


1 4

6.4351e+04
2.2149e+03
6.5736e+04
1.4320e+05
5.8171e+03
0
3.0249e+03
2.7706e+03
3.7198e+04
8.9104e+03
0
0
3.3508e+04

CH3
CH2
CH
C
aCH
aC-CH2
OH
CH2 COO
CH3 O
CH2 O
aC-O
CH2 NH
OCH2 CH2 OH

AROMRINGs s

Dj (K2 )

3rd order group (k)

Ek (K2 )

(CH2 )n (OCH2 CH2 )m (m = 3, n < 8) and (m = 4, 5, n > 8)


(CH2 )n (OCH2 CH2 )m (n = 5)
(CH2 )m CO (OCH2 CH2 ) (m = 8)
(CH2 )m C6 H4 (OCH2 CH2 ) (m = 8)
((CH2 )n )m COC2 H4 (n > 2, m > 1)

1.1108e+04
6.7595e+03
2.1595e+04
6.0521e+03
2.4357e+04

Table 3
Statistical indices of performances relative to the correlation of 72 data-points regarding the cloud point of nonionic surfactants using Marrero and Gani method before and
after the addition of 3rd order groups, compared with 3 different QSPR models [3537].
Model

Data-points for the


regression

SD

AAD

AADmax

Marrero and Gani method without 3rd order groups


Marrero and Gani method with 3rd order groups
QSPR model [35]
QSPR model [36]
QSPR model [37]

73
73
81
68
78

8.91
5.65
9.31
5.89
7.46

7.65
4.62
7.09
4.69
3.13

25.90
15.83
50.2
17.98
52.8

improved statistical indices, and in general the absolute errors have


been strongly reduced. These results represent an improvement
also if compared with those obtained with different QSPR methods
[3537], as shown in Table 3.
An example of the application of the model developed here is
given in Table A1 in the Appendix. The availability of more reliable
experimental data regarding cloud points or nonionic surfactants
belonging to other families (such as alkanediols, ethers, esters and
uorinated linear ethoxylates) will broaden the application range
of the model. However, it can already be safely applied in the surfactant design methodology considering the limited maximum error,
with basic limitation the molecular structures available.
4. Case studies
The framework presented in Fig. 2, together with the property models reported in Sections 2.1 and 2.2 have been applied
to two different surfactant design and selection problems, as a
part of emulsion-based chemical product design. Also the newly
120

Calc. Cloud Point [C]

100
80
60

developed group-contribution models for the prediction of the


cloud point and the critical micelle concentration [24] have been
applied as a comprehensive set of group-contribution models
regarding nonionic surfactants, forming a reliable and consistent
tool for the surfactant design. Table 4 lists the properties needed
by the methodology and the models used in this work.
All the property models reported in Table 4 have been tested
against experimental values in order to verify their reliability. The
values relative to the coefcients of determination of the different models are high, except for the Hansen solubility parameters
and the toxicity parameter. However, there are no other predictive
models with comparable correlation performances, therefore they
are applied for screening of alternative designs, and more accurate
and complex models are employed only a few alternatives are available, to verify the reliability of the predictions. Once these models
have been veried, they have been implemented in the methodology for chemical product design, as described in the previous
chapters.
The rst case study presented is about the design of an emulsied UV sunscreen, where the surfactant system is necessary as
emulsier that is to stabilize the formulation. The second case
study, on the other hand, is about the design of an emulsied handwash, where the surfactant system acts simultaneously as active
ingredient, as a main function, and as emulsier. This contribution,
however, is not intended to give more than an overall picture of an
emulsion-based product design problem, while it focuses on the
surfactant design and selection as a part of the above-mentioned
framework. A comprehensive description of the whole step-bystep methodology is given by Mattei et al. [16].

40

4.1. Surfactant design as emulsier emulsied UV sunscreen


20
0
0

20

40

60

80

100

120

Expt. Cloud Point [C]


Fig. 5. Parity plot relative to the correlation of 72 data-points regarding cloud point
of nonionic surfactants using the Marrero and Gani GC-method after the addition of
dedicated third order groups.

As seen from Fig. 2, in the step-by-step methodology for emulsied product design, the surfactant system is designed after the
active ingredients and both the continuous and the dispersed
phases have been selected. Table S6 in Supplementary material provides the output of steps 2 and 3a, that is the chemicals chosen as
active ingredients and continuous and dispersed solvents, together
with consumer assessments and the target properties responsible

M. Mattei et al. / Fluid Phase Equilibria 362 (2014) 288299

295

Table 4
List of the surfactant property applied in the case studies reported together with the model used and statistical indices.
Surfactant property

Model used

Coefcient of determination

Cloud point
Critical micelle concentration
Hansen solubility parameters
Hydrophiliclipophilic balance
Hydrophiliclipophilic deviation
Krafft temperature
Open cup ash point
Surface tension reduction
Toxicity parameter

Group-contribution method [this work]


Group-contribution method [previously developed [24]]
Group-contribution method [21]
Denition [24]
Denition [32]
QSPR model [26]
Group-contribution method [29]
QSPR model [26]
Group-contribution method [22]

R2 > 0.94
R2 > 0.99
R2 > 0.78
a
a
R2 > 0.94
R2 > 0.96
R2 > 0.99
R2 > 0.77

Statistical indices for hydrophiliclipophilic balance and hydrophiliclipophilic deviation cannot be provided since relative experimental values cannot be found.

for their design (solvents) and selection (active ingredients), as the


output of the step 1 of the methodology.
The surfactant to be designed, as previously mentioned, needs
to act as emulsier, ensuring the formation of an oil-in-water emulsion, stable under the range of temperature in which the product
is supposed to be used.
With relation to the set of Eqs. (1)(7), then, proper boundaries
are set so that the designed surfactant is safe, non-toxic, not affected
by the presence electrolytes and able to generate a stable emulsion.
These targets are translated into target properties, as in Table 5.
Through a computer-aided molecular design (CAMD) technique, all
the above-mentioned constraints have been applied, thus reducing
the search space to a limited number of candidate surfactants. The
choice of the most advantageous between the candidates is taken
minimizing the cost connected to its use, which means the minimum cost at which the critical micelle concentration is reached.
Octyl esaethylene oxide has been designed as optimum surfactant
for this purpose and Table 5 provides the results of the models used
for the design.
4.2. Surfactant design/selection as active ingredients emulsied
hand wash
There are some classes of products which contain surfactants
not only to keep the formulation in the emulsied form, but also to
act as active ingredients. These are detergents in general, where
surfactants are responsible for the wetting of the surface to be
cleaned, for the dissolution of the dirt and for not allowing the
re-deposition of the dirt itself. When these products need to be
designed, then, surfactants are included at the second step of the
methodology of Fig. 2, before any other ingredients of the formulation. The consumer assessments to be satised might then
differ from those listed in the previous paragraph, as different are
the target properties and the relative boundaries set on them. A
comprehensive list of the consumer assessments for surfactants as
active ingredients, together with the property constraints and models used is reported in Table 6. Compared to the previous case-study,
here the surfactant system needs to satisfy the main needs of the
product. Between them, high foam-ability and non-irritability of
the skin. The foam-ability has been qualitative modeled by Pandey
et al. [40] as a function of the surface tension of the system and
the critical micelle concentration of the surfactant used. More

precisely, the lower the surface tension, the higher the foam-ability,
and the lower the critical micelle concentration, the higher the
foam-ability. The non-irritability of the skin, on the other hand, is
estimated through Hansen solubility parameters [31]. If the Hansen
solubility parameters of the designed ingredient are compatible
with the parameters characterizing the proteins of the skin, then
that ingredient is likely to partly dissolve the protein layer, consequently irritating the skin. Therefore, proper boundaries on the
Hansen solubility parameters are set in order to qualitatively satisfy
the constraint of non-irritability. Given the above, the foam-ability
is usually related to the presence of ionic surfactants (since ionic
surfactants are characterized by low surface tension as well as low
critical micelle concentration), while the skin-care is usually connected to the use of nonionic surfactants, due to the values of their
solubility parameters, usually far from those of the proteins of the
skin. Therefore, a mixture of ionic and nonionic surfactants is usually chosen, the stability of which needs to be checked once the
other ingredients of the formulation have been chosen. Numerical constraints on the properties, as reported in Table 6, are applied
and through a CAMD techniques a restrict number of candidates are
generated. Between these, octyl esaethylene oxide is designed as
the most advantageous nonionic surfactant to be used as a nonionic
surfactant, while Sodium Laureth Sulfate is selected with rulebased selection criteria as the most advantageous ionic surfactant
to be chosen. Calculated values for the key-properties considered
in the case-study are reported in Table 6.
The stability of the two surfactants in the formulation, though,
needs to be ensured after all the other ingredients have been
selected and the most advantageous overall composition of the
product has been chosen. Table S7 in Supplementary material lists
the ingredients chosen, together with the relative consumer assessments and target properties to be satised.
The overall composition of the formulation is needed in order to
qualitatively estimate the stability of the emulsion by calculating
the HLD-value of each of the two active-ingredients selected at the
second step of the procedure, that are expected to form a stable
emulsion.
5. Future perspectives: surfactant-related phase behaviors
The description of binary (water-surfactant and oil-surfactant)
and ternary (wateroil-surfactant) phase behaviors with a

Table 5
Target properties, relative physico-chemical properties and models needed, constraints and modeled value for surfactant design for an emulsied UV sunscreen.
Target properties

Properties considered and models used

Constraints

Modeled value

Oil-in-water emulsion desired


Thermal stability
Safety
Non-toxicity
Not inuences by electrolytes
Stability as an emulsion

Hydrophiliclipophilic balance [25]


Cloud point
Flash point [29]
Toxicity parameter [22]
Nonionic surfactants preferred
Hydrophiliclipophilic deviation [32]
Critical micelle concentration [24]

HLB > 12
CP > 55 C
Tf > 55 C
log(LC50 ) > 3.16 mol/m3

HLD =
/ 0

13.4
73 C
341 C
3.97 mol/m3

0.7
0.009 mol/L

296

M. Mattei et al. / Fluid Phase Equilibria 362 (2014) 288299

Table 6
Target properties, relative physico-chemical properties and models needed, constraints and modeled value for surfactant design for an emulsied hand-wash.
Target properties

Properties considered and models used

Constraints

Calculated
value nonionic
surfactant

Calculated
value ionic
surfactant

Foam-ability

Surface tension [26]


Critical micelle concentration [24]

 < 33 mN/m
CMC < 0.01 mol/L

31 mN/m
0.009 mol/L

27.8 mN/m
0.008 mol/L

Non-irritability

Solubility parameters [21]

D =
/ 22.4 1.5 Mpa0.5
P =
/ 9.8 1.5 Mpa0.5
H =
/ 11.9 1.5 Mpa0.5

16.9
9.2
13.7

a
a
a

Cleaning performances

Surface tension [26]


Hydrophiliclipophilic balance [25]

 < 29.3 mN/m


HLB > 10

31 mN/m
13.4

27.8 mN/m
40

Safety

Flash point [29]

Tf > 55 C

341 C

94 C

Non-toxicity

Toxicity parameter [2]

log(LC50 ) > 3.16 mol/m3

3.97 mol/m3

LD50 = 5 g/kgb

Thermal stability

Cloud point [this work]


Krafft temperature [26]

CP > 55 C
TK < 20 C

73 C

16 C

Stability as an emulsion

Hydrophiliclipophilic deviation [32]

HLD =
/ 0

0.9

1.5

Hansen solubility parameters are not available, but the surfactant (Sodium Laureth Sulfate) is known to be soluble in water, so it is unlikely to dissolve the proteins of
the skin, given their Hansen solubility parameters.
b
LC50 is not available, so a value for LD50 , another toxicity parameter, has been provided instead.

thermodynamic model, instead of the adoption of simplied


heuristic, is considered as a major progress to be implemented in
the product design methodology. This way, it is possible to identify
boundaries in terms of temperature and, especially, composition
for a surfactant to generate a stable emulsied formulation. This
chapter will highlight some perspectives for future development of
the above-mentioned analysis and a few preliminary results; to this
point, however, the application of surfactant-related phase behaviors is limited to the availability of experimental data relative to
the systems of interest. Therefore this approach can be applied for
product analysis as well as for verication of the design obtained
from the methodology.

5.1. Binary systems: water-surfactant and oil-surfactant


Water-surfactant phase diagrams are fundamental when the
surfactant is expected to be mainly dissolved in the water-phase
(high values of the hydrophiliclipophilic balance), which leads to
the formulation of oil-in-water emulsions. On the other hand, the
phase diagrams between oil and surfactants are to be considered
primarily when a water-in-oil emulsion is desired.
As an example of water-surfactant phase behaviors, Fig. 3
presents the phase diagram of the system between water and
dodecyl-esaethylene oxide. For emulsion-based chemical product
design, therefore, it is relevant to dene the boundaries in terms of
temperature and concentration, so that the designed formulation
lies in the area dened as L1 , where the models presented in Tables
S1S3 in Supplementary material can be safely applied. Mitchell
et al. [34] and Sjblom et al. [41] provide a satisfactory amount of
experimental data as well as theoretical explanations for the formation of the different meso-phases, relatively to aqueous surfactant
solutions. The methods proposed, however, cannot be used for the
prediction of these phase boundaries and therefore they cannot be
applied in the framework for chemical product design. Approximate predictions of these phase behaviors, based on the analysis
of several phase diagrams in parallel with the molecular structure
of the species involved, are however expected to be possible by
the authors and they are considered to be a potential important
development on the way for a fully model-based methodology for
emulsion-based chemical product design. The adoption of such correlations can lead to the calculation of simplied water-surfactant

phase behaviors as described in Fig. S4 in Supplementary material,


related to the system reported in Fig. 3.
When a water-in-oil emulsion is wanted, on the other hand,
the phase behavior of surfactant mixtures with oil is relevant. Fig. S5 [42] in Supplementary material shows an example
the phase behaviors between oils and surfactants, in terms of
liquid-liquid miscibility boundaries. The reference surfactant is
hexyl-pentaethylene oxide, while four different n-alkanes are considered as the oil-phase.
In Fig. S5, only the miscibility curve as function of surfactant
concentration and temperature is reported, while the formation
of micellar solutions of standard appearance or of viscous mesophases with different self-assemblies of the surfactant in absence of
water is debatable [43]. For use in emulsion-based chemical product design, it is necessary that the designed formulation lies above
the line of the miscibility gap. It is easy to determine in Fig. S5 a
trend of the miscibility curves as a function of the number of carbon
atoms of the n-alkane considered. This leads to the consideration
that a correlation based on the molecular structure of the chemicals involved in the phase equilibrium can approximately describe
these curves and therefore a correlative model may be developed
and applied for use in emulsion-based chemical product design.
In relation to both water-surfactant and oil-surfactant phase
behaviors, however, experimental measurements are needed in
order to dene numerical boundaries on the composition of the
desired formulation, since model-based generation of data is not
considered yet reliable.

5.2. Ternary systems: wateroil-surfactant


The understanding of the behavior of ternary water-oilsurfactant systems is also considered to be crucial, in order to
determine temperature and composition boundaries for a stable
emulsion. These type of phase envelops can be represented in
several ways, since many variables are involved; in relation to
emulsion-based chemical product design, the most useful alternative is represented by the use of the so-called Kahlweits sh phase
diagram [42]. Here, ternary wateroil-surfactant data are drawn in
an XY diagram, where the surfactant content (usually in weight
percentage) is in the X-axis, while the temperature is on the Y-axis.
These diagrams represent a valid tool for emulsion-based product design since different types of products can be recognized and

M. Mattei et al. / Fluid Phase Equilibria 362 (2014) 288299

360

6. Conclusion

2 - W/O

350
Temperature [K]

297

340
330

320
310
300

2 - O/W

290
0%

20%

40%

60%

80%

Surfactant weight %
Fig. 6. Calculated sh-diagram of the system water-tetradecane-2-butoxyethanol;
wateroil ratio: 7.03.
Source: Data taken from [44].

the possibilities for the formation of each of them are easily identied, given the temperature and the composition of the formulation.
An example of such as sh-type phase diagram for the system
water-tetradecane-2-butoxyethanol is shown in Fig. 6.
In Fig. 6 different areas are can be identied: the region dened
with the symbol 1 represents the area where a micro-emulsion
can be formed, the region dened with the symbol 2 (both 2
W/O and 2 O/W) is the area where an emulsied product can be
formulated, while the region identied by the symbol 3 is a hybrid
domain, where an emulsion and a microemulsion may coexist. The
emulsion domain (2) consists of two areas: one above the hybrid
domain, described by the symbol 2 W/O, where a water-in-oil
emulsion can be formed and another below the hybrid domain,
described by the symbol 2 O/W, where an oil-in-water emulsion
is favored instead. Consequently, the region of the emulsion domain
located at intermediate temperatures between the two abovementioned areas represents an unstable region where it is not
recommended to design an emulsied product, since its life time
is expected to be limited. As a part of an emulsion-based chemical
product design procedure, then, it is necessary to make sure that the
designed formulation lies in the 2 W/O area when a water-in-oil
emulsion is desired and in the 2 O/W if an oil-in-water emulsion is wanted. This type of diagrams can be used both during the
design of the surfactant and during the verication of the designed
product. The authors are not aware of any reliable model for the
prediction of such phase equilibria, and therefore this analysis can
be performed up to now only when experimental data are available.
When predictions are necessary, the hydrophiliclipophilic deviation (HLD) approach [32] is applied instead. This method consists
of the application of an experimental-based correlation for each
of the surfactant in the formulation, considering several variables
such as the presence of electrolytes, the nature of the oil as well
as of the water phases, the temperature, and the molecular structures of the surfactant. If the calculated HLD-value is zero, then
the formulation is located in the 3 domain of Fig. 6 and therefore
an unstable system is expected. On the other hand, if a positive
value is obtained, then a water-in-oil emulsion is favored, while if
a surfactant is characterized by a negative value of its HLD, then
an oil-in-water emulsion may be formed. The higher the absolute
value of the HLD of the surfactant is, the more stable the emulsion formed is expected to be, since it is located further away from
the unstable region identied by the hybrid domain. This method
does not have the thermodynamic basis of the representation of the
ternary phase diagram, but it can be used as a qualitative predictive
model when the needed experimental data are not available.

A comprehensive framework for surfactant design and selection


for emulsion-based chemical product design has been presented.
The necessary properties for the whole procedure have been
listed and the need of predictive models for a reliable solution
of the reverse problem has been highlighted. Moreover, a groupcontribution model, based on the Marrero and Gani method has
been developed for the correlation and further prediction of the
cloud point of nonionic surfactants. The addition of new dedicated
third order groups, in order to take into account the peculiar structures of the surfactants considered, has been necessary in order to
achieve acceptable performances. Compared to the existing QSPR
models, the group-contribution model here developed performed
better given the standard deviation and the average absolute deviation as statistical indices and the number of different families of
surfactant considered, as an indicator of the range of application of
the model.
The application of the methodology has been then illustrated
using two conceptual case-studies. In the rst, the surfactant system to be designed was needed in order to keep the formulation
in a stable emulsied form. So, the surfactant needs to be chosen
once the active ingredients and solvents were known. Boundaries
were set on the basis of effectiveness (related to the ingredients
previously designed and/or selected), safety and toxicity, while the
nal decision has been performed by minimizing the cost of the
ingredient. In the second case-study, instead, the surfactant system was supposed to act as emulsier and active ingredient at the
same time. The surfactant is then designed before any other ingredient of the formulation, and in order to satisfy the main needs
of the product, a surfactant mixture of ionic and nonionic surfactants was needed. The stability of the surfactants in relation to the
other chemicals present in the designed product is ensured after the
overall composition has been selected. Further perspective for the
application of rigorous thermodynamic descriptions of the phase
behavior of surfactant-related system, both in binary mixture with
water and oil, and in ternary mixture with both an oil- and a waterphase, is considered a relevant improvement that can help leaving
heuristics and rules of thumb which are usually applied to estimate the phase stability of emulsied products. The application
range of the methodology is wide, in the sense that other similar products can be designed, once the needs-property relations
are established. Further works will focus on the development of
group-contribution methods for the pure component properties of
ionic surfactants (adequate new rst order groups are needed) and
the development of predictive, reliable models able to efciently
describe both the binary and the ternary phase diagrams involving
surfactants.

Acknowledgments
Financial support from the Technical University of Denmark is
greatly acknowledged.
Advice given by Professor Michael Hill, from the Department
of Chemical Engineering of Columbia University, New York, in the
development of one of the case studies is greatly appreciated.

Appendix.
Appendix B. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.uid.2013.10.030.

298

M. Mattei et al. / Fluid Phase Equilibria 362 (2014) 288299

Table A1
Calculation of the cloud point of Nonyl phenyl octaethylene oxide after the introduction of new third order groups. Calculated value is compared with experimental value
and with the calculations with the GC-model without 3rd order groups and three QSPR models.
Nonyl phenyl octaethylene glycol Molecular structure
O

HO

O
O

O
O

O
O

Molecular formula: C33 H60 O10


First order groups

Group contribution (K2 )

Occurrences

CH3
CH2
aCH
aC-CH2
CH2 O
aC-O
OCH2 CHOH

1
16
4
1
7
1
1

6.4351e+04
2.2149e+03
5.8171e+03
0
8.9104e+03
0
3.3508e+04

Second order groups

Occurrences

Group contribution

AROMRINGs1 s4

Third order groups

Occurrences

Group contribution

(CH2 )m

6.0521e+03

(CP )calc =

C6 H4


i

(OCH2 CH2 ) (m = 8)

Ni Ci +


j

Mj Dj +

Ok Ek = 107576.3 K , (CP)calc = 327.99 K GC-model without 3rd order groups: (CP)calc = 322.04 K, QSPR models from Table 8:

(CP)calc = 320.65 K [35], 326.33 K [36], 328.45 K [37], (CP)exp = 329.15 K.

References
[1] J. Villadsen, Putting structure into chemical engineering: proceeding of
an industry/university conference, Chemical Engineering Science 52 (1997)
28572864.
[2] E.L. Cussler, G.D. Moggridge, Chemical Product Design, 2nd ed., Cambridge
University Press, Cambridge, UK, 2012.
[3] R. Gani, Chemical product design: challenges and opportunities, Computers
and Chemical Engineering 28 (2004) 24412457.
[4] J. Abildskov, G.M. Kontogeorgis, Chemical product design. A new challenge of
applied thermodynamics, Chemical Engineering Research and Design 82 (2004)
15051510.
[5] E. Conte, R. Gani, K.M. Ng, Design of formulated products: a systematic methodology, AIChE Journal 57 (2011) 24312449.
[6] E. Conte, R. Gani, Y.S. Cheng, K.M. Ng, Design of formulated products: experimental component, AIChE Journal 58 (2012) 173189.
[7] V.S. Raman, C.D. Maranas, Optimization in product design with properties correlated with topological indices, Computers & Chemical Engineering 22 (1998)
747763.
[8] N.G. Chemmangattuvalappil, C.C. Solvason, S. Bommareddy, M.R. Eden, Combined property clustering and GC(+) techniques for process and product design,
Computers & Chemical Engineering 34 (2010) 582591.
[9] M.A. Teixeira, O. Rodriguez, S. Rodrigues, I. Martins, A.E. Rodrigues, A case study
of product engineering: performance of microencapsulated parumes on textile
applications, AIChE Journal 58 (2012) 19391950.
[10] J.C. Charpentier, Perspective on multiscale methodology for product design and
engineering, Computers & Chemical Engineering 33 (2009) 936946.
[11] C. Wibowo, K.M. Ng, Product-centered processing: manufacture of chemicalbased consumer products, AIChE Journal 48 (2002) 12121230.
[12] B.V. Smith, M. Ierapepritou, Integrative chemical product design strategies:
reecting industry trends and challenges, Computers & Chemical Engineering
34 (2010) 857865.
[13] F.P. Bernardo, P.M. Saraiva, Integrated process and product design optimization: a cosmetic emulsion application, Computer Aided Chemical Engineering
20 (2012) 15071512.
[14] M. Bagajewicz, S. Hill, A. Robben, H. Lopez, M. Sanders, E. Sposato, C. Baade,
S. Manora, J.H. Coradin, Product design in price-competitive markets: a case
study of a skin moisturizing lotion, AIChE Journal 57 (2011) 160177.
[15] M. Bagajewicz, On the role of microeconomics, planning, and nances in product design, AIChE Journal 53 (2007) 31553170.
[16] M. Mattei, G.M. Kontogeorgis, R. Gani, A systematic methodology for design of
emulsion based chemical products, Computer Aided Chemical Engineering 31
(2012) 220224.
[17] G.J. Tiddy, Formulation science and technology surfactants needed!, Current
Opinion in Colloid & Interface Science 4 (2000) 379380.
[18] J. Marrero, R. Gani, Group-contribution-based estimation of pure component
properties, Fluid Phase Equilibria 183 (2001) 183208.
[19] L. Constantinou, R. Gani, New group-contribution method for estimating properties of pure compounds, AIChE Journal 40 (1994) 16971710.

[20] E. Conte, A. Martinho, H.A. Matos, R. Gani, Combined group-contribution and


atom connectivity index-based methods for estimation of surface tension and
viscosity, Industrial & Engineering Chemistry Research 47 (2008) 79407954.
[21] H. Modarresi, E. Conte, J. Abildskov, R. Gani, P. Crafts, Model-based calculation
of solid solubility for solvent selection a review, Industrial & Engineering
Chemistry Research 47 (2008) 52345242.
[22] A.S. Hukkerikar, S. Kalakul, B. Sarup, D.M. Young, G. Sin, R. Gani, Estimation
of environment-related properties of chemicals for design of sustainable processes: development of group-contribution (+) (GC(+)) property models and
uncertainty analysis, Journal of Chemical Information and Modeling 52 (2012)
28232839.
[23] R. Ceriani, R. Gani, A.J.A. Meirelles, Prediction of heat capacities and heats of
vaporization of organic liquids by group contribution methods, Fluid Phase
Equilibria 283 (2009) 4955.
[24] M. Mattei, G.M. Kontogeorgis, R. Gani, Modeling of the critical micelle
concentration (CMC) of nonionic surfactants with an extended groupcontribution method, Industrial and Engineering Chemistry Research 52 (2013)
1223612246.
[25] W.C. Grifn, Classication of surface-active agents by HLB, Journal of the
Society of the Cosmetic Chemists 1 (1949) 311326.
[26] Y. Li, G. Xu, Y. Luan, S. Yuan, X. Xin, Property prediction of surfactant by quantitative structureproperty relationship: krafft point and cloud point, Journal
of Dispersion Science and Technology 26 (2007) 799808.
[27] C.F. Spencer, R.P. Danner, Improved equation for prediction of saturated liquid density, Journal of Chemical and Engineering Data 17 (1972)
236241.
[28] A.L. Horvath, Molecular Design, Elsevier, Amsterdam, 1992.
[29] H.J. Liaw, Y.H. Lee, C.L. Tang, H.H. Hsu, J.H. Liu, A mathematical model for predicting the ash point of binary solutions, Journal of Loss Prevention in the
Process Industries 15 (2002) 429438.
[30] R. Pal, Rheology of simple and multiple emulsions, Current Opinion in Colloid
& Interface Science 16 (2011) 4160.
[31] C.M. Hansen, Hansen Solubility Parameters: A Users Handbook, 2nd ed., CRC
Press, Boca Raton, FL, 2007.
[32] J.L. Salager, Quantifying the concept of physico-chemical formulation in
surfactant-oil-water systems state of the art, Progress in Colloid and Polymer
Science 100 (1996) 137142.
[33] J. Hu, X. Zhang, Z. Wang, A review on progress in QSPR studies for surfactants,
International Journal of Molecular Science 11 (2010) 10201047.
[34] D.J. Mitchell, G.J.T. Tiddy, L. Wairing, T. Bostock, M. McDonald, Phase behaviour
of polyethylene surfactants with water, Journal of Chemical Society-Faraday
Transactions I 79 (1983) 9751000.
[35] C. Guo, Y. Ren, P. Zhou, J. Shao, X. Yang, Z. Shang, Toward a quantitative model
and prediction of the cloud point of normal nonionic surfactants and noverl
Gemini surfactants with heuristic method and gaussian process, Journal of
Dispersion Science and Technology 33 (2012) 14011410.
[36] J. Ghasemi, S. Ahmadi, Combination of genetic algorithm and partial lease
squares for cloud point prediction on nonionic surfactants from molecular
structures, Annali di Chimica 97 (2007) 6983.

M. Mattei et al. / Fluid Phase Equilibria 362 (2014) 288299


[37] Y. Ren, B. Zhao, Q. Chang, X. Yao, QSPR modeling of nonionic surfactant cloud
points: an update, Journal of Colloid and Interface Science 358 (2011) 202207.
[38] M.J. Rosen, Structure/Performance Relationships in Surfactants, American
Chemical Society, Washington, DC, 1984.
[39] A.S. Hukkerikar, R.J. Meier, G. Sin, R. Gani, A method to estimate the enthalpy of
formation of organic compounds with chemical accuracy, Fluid Phase Equilibria
348 (2013) 2332.
[40] S. Pandey, R.P. Bagwe, D.O. Shah, Effect of counterions on surface and foaming
properties of dodecyl sulfate, Journal of Colloid and Interface Science 267 (2003)
160166.
[41] J. Sjblom, P. Stenius, I. Danielsson, Phase equilibria of nonionic surfactants and
the formation of microemulsions, in: M.J. Schick (Ed.), Nonionic Surfactants

299

Physical Chemistry, Surfactant Science Series, vol. 23, Marcel Dekker, New York,
1987.
[42] M. Kahlweit, R. Strey, Phase behavior of ternary systems of the type H2 Ooilnonionic amphiphile (microemulsions), Angewandte Chemie International
Edition English 24 (1985) 654668.
[43] A. Ray, Solvophobic interactions and micelle formation in structure forming
nonaqueous solvents, Nature 231 (1971) 313315.
[44] B.J. Lin, L.J. Chen, Liquidliquid equilibria for the ternary system
water + tetradecane + 2-butoxyethanol, Fluid Phase Equilibria 216 (2004)
1320.