CC BY
38
YflK 544.3+544.7
BecTHHK Cnôry. Cep. 4. 2013. Bun. 1
M. Hebabcha, M. Miltgen, A. Modaressi, P. Magri, A. Ait-Kaci, M. Rogalski
AGGREGATION OF NANOPARTICLES IN POLYPHASE MIXTURES — IMPACT ON PHASE EQUILIBRIA
Introduction. The application of nanosized solid particles, NPs, is of growing importance in many domains of science, human life and technology [1-5]. Particles with a diameter in the range 1-100 nm occupy an intermediate state between solid and molecular states. When the number of atoms in the particle lies in the range of 1000 or above, their properties gradually evolve from molecular to solid type. Such particles are characterized by high ratio of surface-to-volume atoms. Thus the surface effects become increasingly important when the size of NPs decreases. Attractive/repulsive interactions between particles depend on the geometrical factors and change strongly with decreasing particle size. Thus, Nps display some exceptional properties and may influence not only such dynamic properties as nucleation and particle growth [6, 7], but also the phase behaviour of liquid—liquid and liquid—solid equilibria. The formulation of filler — polymer blends drawn attention on the phase equilibria of liquid systems containing Nps. This issue was investigated by the group of Lipatov [8-12]. A change in the cloud point curves and phase compositions was reported and explained by preferential adsorption existing between the filler and one of the components of the blend. Nesterov et al. observed that the LCST curve of poly(vinyl acetate)/poly(methyl methacrylate) (PVA/PMMA) mixtures were shifted to higher temperature, the miscibility window being enlarged in the presence of filler [9] and Lipatov et al. [10] found that introducing silica filler in chlorinated polyethylene/poly(ethylene-co-vinyl acetate) (EVA) blends led either to the increase or to the decrease in the temperature of phase separation depending on the filler concentration.
Ginzburg [13] formulated a theory to predict how nanoparticles can influence the behavior of the blend. According to this theory, the addition of Nps stabilize the solution reducing the number of unfavorable polymer/polymer interaction if the radius of the Np is smaller than the radius of gyration of the polymer. In the case of Nps larger than the polymer radius of gyration, the particle-rich phase segregates from the blend. In a immiscible polymer blend the nanofiller may distribute unevenly among the phases [14].
However, the physical reasons explaining significant shift of equilibrium diagram of macromolecules are not relevant in the case of mixtures of molecules. The influence of Nps on phase equilibria is not well understood in this case. A shift of the phase equilibrium may occur due to a selective adsorption of the mixture components on the Np surface. However, the usual, monolayer adsorption would significantly influence the equilibrium state only if the total surface of Nps is very high i.e. if the mass of Nps is high. Recent studies concerning the separation of acids from [15, 16] aqueous solutions using ionic liquids and Nps
Mustapha Hebabcha — assistant, LTMM, Université Houari Boumédiène des Sciences et Technologie, Algeria.
Morgane Miltgen — student, LCP-A2MC, Universite de Lorraine, France.
Ali Modaressi — Dr., associate professor, LCP-A2MC, Universite de Lorraine, France.
Pierre Magri — Dr., researcher, LCP-A2MC, Universite de Lorraine, France.
Ahmed Ait-Kaci — Dr., professor, LTMM, Universite Houari Boumediene des Sciences et Technologie, Algeria.
Marek Rogalski — Dr., professor, LCP-A2MC, Universite de Lorraine, France; e-mail: [email protected]
©M. Hebabcha, M. Miltgen, A. Modaressi, P. Magri, A. Ait-Kaci, M. Rogalski, 2013
showed that the extraction process is influenced by Nps. It was observed that in aqueous solutions containing an ionic liquid and Nps of alumina stable nanoaggregates were formed. Therefore, the Np dispersion in a given solvent corresponds to a dispersion of nanoaggregates filled with the solvent. Moreover, the ionic liquid distribute unevenly between the bulk solution and the inner space of the aggregate. These phenomena were explained using the theory of solvation forces between colloidal nanoparticles as was proposed by Israelachvili [17] and recently by Fichthorn et al. [18-22]. These authors found that interaction between solid surfaces are solvent mediated and induce the solvent ordering in the confined space [19]. The interaction forces between colloidal particles in solution induce the formation of aggregates and are mediated by liquid filling the inner space of aggregates.
In the present work the influence of Nps on the phase equilibria was dealt with. At first the mechanism of Np assembly was investigated and the size and the stability of aggregates formed was determined. Next, the phase behaviour of selected mixtures containing Np aggregates was concerned.
Materials. The NPs Al2O3, d =80 nm, TiO22, d < 100 nm, and Fe3O4, d = 10^20 nm) were supplied by Sigma-Aldrich.
Ethanol (99.5 %), cyclohehane (99.7 %), acetonitrile (99.9 %), dimethyslsulfoxide (DMSO) (99.7 %), decan-1-ol (99%) were purchased from Sigma-Aldrich and were used without further purification.
Dynamic Light Scattering (DLS). The mean particle size was determined using a Malvern 48000 photon correlation spectrometer (Malvern Instruments, Malvenrn, UK) with a light source from a 35-mW, 633 nm He-Ne laser. The autocorrelation function of the scattered light was determined with 7032 Malvern UK, 8-bit, high-speed, 72 channels correlator. The data were analysed by the standard program using CONTIN method [23] obtaining the average value of the diameter of dispersed solid particles. Measurements were performed using samples conditioned at first at constant temperature during 24 h, and just before experiment in an ultrasonic bath during 20 min. Ultrasonic treatment disintegrates aggregates and allows to follow the aggregation occurring during the experiment. Therefore, DLS experiments give information about the size and the stability of aggregates formed in the solutions studied.
Aggregate formation. The stability of Ag and SiO22 Nps was studied in DMSO and in acetonitrile. All experiments reported here were performed without stabilizing additives such as surfactants that might modify the mechanism of aggregate formation and the phase equilibria.
Dispersion of Nps in the solvent was performed using ultra sound cavitation and the mechanical dispersion.
Ultrasonic cavitation is often used to disperse nano-size particles into liquids and to break particle aggregation. Ultrasonic waves propagating into the liquid alternate high-pressure and low-pressure cycles causing high speed liquid jets colliding with particles and particle aggregates. This applies mechanical stress on the attracting forces between the individual particles. This causes not only the disaggregation of particles but also elimination of gas microbubbles attached to the surface.
The rotor stator mixer used in this work (ultra turrax) is a mechanical disperser with a high angular speed going up to 30000 rpm.
The aggregate size was analysed using a Malvern 48000 photon correlation spectrometer just after sample homogenization and in certain case 48 h later. Measurements were carried out during 15 min. The stability of aggregates was rated on the base of the size changes of aggregates. and corresponds to a difference between the maximal and the minimal
size of aggregates, Ad = dmax — dmin, observed during 15 min. The stability scale defined on this base is listed in Table 1.
Stability of aggregate prepared using different combination of ultrasonic and mechanical dispersing methods was checked and results are reported in Tables 2-4.
The most relevant information that may be observed from the inspection of these tables is the possibility to obtaining stable aggregates in various solvents without surfactant additives. The average aggregate size was mostly within the interval 100-200 nm independently of the solvent used. Therefore, aggregates obtained by combining the mechanical and the ultrasonic stirring were suited to check the Np influence on solid—liquid or liquid—liquid diagrams.
Table 2
Stability of SiO2 aggregates dispersed in acetonitrile (the stability analysis was carried out: 0 — immediately after the sample preparation; 48 h after the sample preparation)
Analyse Time/h Solution treatment Stability Aggregate size, nm
Ultra sound Ultra turax
0 - - VU 2000-4000
0 - + G ~ 400
0 + - F ~ 200
0 + + G ~ 150
48 - - F ~ 130
48 - + G ~ 130
Table 3
Stability of SiO2 aggregates dispersed in DMSO (the stability analysis was carried out: 0 — immediately after the sample preparation; 48 h after the sample preparation)
Analyse Time/h Solution treatment Stability Aggregate size, nm
Ultra sound Ultra turax
0 - - U ~ 160
0 - + G ~ 200
0 + - U 150-200
0 + + G ~ 200
48 - - F ~ 400
48 - + G ~ 700
Cryoscopic depression in presence of Nps. The solute addition produces a freezing-point depression of the solvent. For low solute concentrations, the freezing point depression becomes a colligative property and depends solely of the concentration of the solute. For ideal solutions the freezing depression is proportional to the solute concentration. In the case of Np suspension the solute molality is very low and the expected freezing depression is negligible.
Table 1
The stability scale established on the base of the size changes of aggregates (the size of Nps was determined by dynamic light scattering)
Ad, nm Stability
< 10 Excellent (E)
10-50 Good (G)
50-200 Fair (F)
200-1000 Unstable (U)
1000-5000 Very unstable (VU)
> 5000 Flocculating (FL)
Table 4
Stability of Ag aggregates dispersed in DMSO (the stability analysis was carried out: 0 — immediately after the sample preparation; 24 and 48 h
after the sample preparation)
Time, h Solution treatment Stability Aggregate size, nm
Ultra sound Ultra turax
0 - - G ~ 130
0 - + F 100-140
0 - - F 80-120
0 - + G 85-90
24 + - F ~ 160
24 + + F 200-300
48 - - U 170-400
48 - + F ~ 160
Cryoscopic experiments were performed using Nps of three materials i. e. silver (d < 100 nm), alumina (d < 50 nm) and silica (10 nm < d < 20 nm) in cyclohexane and dimethysulfoxide (DMSO). Results are presented in Fig. 1, 2 and compared with cryoscopic depression calculated earlier. One observes a significant depression of the temperature of fusion that is of the same order as the cryoscopic depression. This finding may be explained by aggregation of Nps. Indeed, as was discussed in precedent chapter, Nps form small, monodispersed aggregates the inner space of which is filled with the solvent that mediates interactions between colloidal particles. Therefore, the chemical potential of the solvent in the confined space is modified in respect to the chemical potential in the bulk. It may be thought that this difference of the chemical potential may modify the fusion temperature of the solvent in the confined space.
The fusion temperature depression depends probably of the mechanism of aggregate formation, i.e. of interaction between Nps and between Nps and the solvent. In all experiments
17.8
17.7
17.6
17.5
17.4
o
17.3
17.2
17.1
17
16.9
0 0.2 0.4
0.6 m, ,
0.8
D> g
1.0 1.2 1.4
17.8
17.7
17.6
17.5
p 17.4
h 17.3
17.2
17.1
17
16.9
\
. \
\\
V \
\
1
0.2
0.4
«Siû»
1
0.6 0.8
2
Fig. 1. The cryoscopic depression occuring in DMSO with Al2O3 Np dispersion (a)
SiO2 Np dispersion (b): 1 — depression occurring with Nps; 2 — cryoscopic depression of DMSO
0
6.5 6.41 6.3 p 6.2 ^6.1 6 5.9 5.8
N ■— N
N. \ 1
N N
N \
k 1 ■ 2
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
6.5 6.4 6.3
O 6.2
E>f 6.1
5.8
\
\ N X
\ N \ 1
\
\ 1
— k 1 ■ 2 \
0. 02 0. mA 06 g , g 0. 10 0.
6.5 6.4 6.3 p 6.2 ^6.1 6 5.9 5.
*
N
N N
\ \
N
k 1 .2
0 0.02 0.04 0.06 0.08 0.10 0.12 0.14
mAl,O» g
Fig. 2. The cryoscopic depression occuring in cyclohexane with SiO2 Np dispersion (a), Al2O3 Np dispersion (b), Ag Np dispersion (c):
1 — depression occurring with Nps; 2 — cryoscopic depression of cyclohexane
b
a
c
6
performed a break was observed on the fusion temperature curve that correspond probably to a change in aggregation mechanism. No cryoscopic depression was observed for low Np concentration in cyclohexane solutions. In this case there exists probably a critical concentration of aggregation of Nps.
Ternary system Acetonitrile+cyclohexane+ethanol+Ag Nps at 298,2 K. The liquid—liquid solubility curve of the ternary system acetonitrile+cyclohexane+ethanol was measured at 298.2. A series of acetonitrile+cyclohexane mixtures was titrated with ethanol up to the homogenisation of the two-phase binary mixture. Results obtained were in good agreement with evaluated data taken from the IUPAC-NIST Solubility Data Series publication. The resulting solubility curve is plotted in Fig. 4. Next, the same experience was performed with the amount of Nps varying up to 0.1-0.2 % of the mass of binary mixture. This experiment was performed using a sample with the acetonitrile/cyclohexane mole ratio of 0.88. Acetonitrile-cyclohexane solutions were titrated with ethanol up to the vanishing of one of two liquid phases. Fig. 3 presents results obtained with acetonitrile+cyclohexane mixtures containing varying amounts of Nps of Al2O3, SiO2 and Ag. It may be observed that the amount of ethanol needed for homogenisation of binary mixture was lower in presence
Fig. 3. Solubility curve of the system acetonitrile+cyclohexane+etha-nol+Nps at 298,2 K:
mole fraction of ethanol on the solubility curve in function of the mass of dispersed Nps; acetoni-trile/cyclohexane mole ratio was of 0.88
0.376 0.374
0.372 0.37 0.368 I 0.366 * 0.364 0.362 0.36 0.358
\
\
, I
\ J SiO, -
J > — Al2OO3 —Ag -1-
0.05
0.1 0.
mNp , g
15
0.2
0.25
Ethanol 1 * 0
0
1 0.9 0. Acetonitrile
0.7 0.6 0.5 0.4 0.3 0.2 0.1 0
Cyclohexane
Fig. 4. Solubility curve of the system:
1 — acetonitrile + cyclohexane + ethanol; 2 — acetonitrile + cyclohexane + ethanol + Ag Nps (0.05 w/w % ) at 298,2 K
of Nps. Results obtained with AI2O3, SiO2 are very similar in spite of different chemical properties of these oxides.
0
One notice that the smallest amount of ethanol was needed when 0.05 g of the oxide was added to the mixture. The amount of 0.2 w/w % of SiO22 lowers of 1.4 w/w % the amount of ethanol. Higher amounts of Nps increase the amount of ethanol. Results obtained with Ag Nps display a different pattern. In this case 0.03 w/w % is sufficient to lower of 1.4 w/w % the amount of ethanol. Increasing the mass of Ag Nps does not change this result.
These results may be explained assuming the partition of mixture components between the bulk and the liquid confined in the inner space of nanoaggregates. The more ordered fluid in the interparticle gap is in equilibrium with the bulk fluid, as in the case of Liquid—Liquid Crystal equilibrium. All components of the mixture present either in the bulk or in the interparticle gap have the same chemical potential.
^¿(bulk) = Hi (gap). This implies partition of components between the gap space and the bulk:
Xi (gap) _ Yi(bulk)
K =
xi (bulk) Yi(gap)
Thus, the liquid in the interparticle gap might have a different concentration than in the bulk. This mechanism allows to explain the shift of liquid—liquid equilibrium presented in Fig. 3.
At low NP concentration the formation of aggregates results in the partition of mixture components between the gap space and the bulk. A critical concentration of nanoaggregates corresponds to the maximal number of aggregates that may form a stable dispersion. This concentration is attained at 0.2 w/w % for SiO22 and 0.05 w/w % for Ag. Higher amounts of Nps do not modify the phase equilibrium and probably remain at the liquid—liquid interface.
Further discussion of thermodynamic aspects of observed phenomena is not possible due to the complexity of physical phenomena occurring in this system. Indeed, the phase equilibrium is conditioned not only by chemical properties of the mixture components but also by chemical properties, size and shape of Nps used in experiment.
In Fig. 4 the solubility curves corresponding to the ternary liquid—liquid equilibria in presence of 0.05 w/w % of Nps of Ag are plotted.
A small systematic shift caused be the presence of Ag nanoaggregates was observed for all concentrations studied. However it was higher for the cyclohexane rich than for the acetonitrile rich phase.
Binary system acetonitrile+n-decanol+Nps Ag. The binary system acetonitri-le+n-decanol was selected because this mixture is partly miscible at low temperatures and displays the critical miscibility temperature of about 25 °C. This experiment showed the influence of Nps on the phase equilibrium in the vicinity of the critical miscibility point. Experimental results are presented in Fig. 5 and are in good agreement with evaluated data published in the IUPAC-NIST Solubility Data Series [23, 24].
In the same figure the liquid—liquid diagram corresponding to acetonitrile+n-deca-ne+0.05 w/w % of Nps of Ag was plotted. Neither the acetonitrile rich branch of diagram nor the critical miscibility point are not affected by the presence of Nps. A significant shift of the equilibrium temperature was found in the n-decanol rich branch. This suggests that Nps were dispersed in the alcohol phase only. Moreover, the equilibrium temperature becomes significant at the vicinity of the critical mixing temperature.
Conclusion. Experimental results presented in this study confirms a significant influence of Nps on phase equilibria. Indeed, a shift of phase diagrams were observed as well
25
20
O
o
15
10
Fig. 5. Solubility curve 0.000 0.200 0.400 0.600 0.800 1.000
of the system x(1)
with liquid—solid as liquid—liquid equilibria. This finding cannot be explained solely by adsorption on the surface of Nps. It was demonstrated that Nps dispersed in the liquid phase form small aggregates, the size of which (the aggregation ratio) depends on the physical and chemical properties of the dispersing and dispersed phases. The liquid that fills the inner space of aggregates displays thermodynamic properties different than in the bulk. This difference leads to the partition of the solution components between the bulk and the confined space. It was suggested, that the component partition is the main factor which determines the shift of phase equilibrium diagram observed in Np containing liquids.
References
1. GleiterH. // Acta Mater. 2000. Vol. 48. P. 1.
2. El-Sayed M. A. // Acc. Chem. Res. 2001. Vol. 34. P. 257.
3. Sun Y. G., Xia Y.N. // Science. 2002. Vol. 298. P. 2176.
4. MargueritatJ., GonzaloJ., Afonso C. N. et al. // Nano Lett. 2006. Vol. 6. P. 2037.
5. Kumar A., VemulaP. K., Ajayan P. M., JohnG. // Nat. Mater. 2008. Vol. 7. P. 236.
6. YouH., ChenC., YangS. et al. // Cryst. Growth Des. 2011. Vol. 11. P. 5449.
7. BakshiM. S., SacharS., KaurG. et al. // Cryst. Growth Des. 2008. Vol. 8. P. 1713.
8. NesterovA. E., Lipatov Y. S., Horichko V. V., Gritsenko O. T. // Polymer. 1992. Vol. 33. P. 619.
9. NesterovA. E., Lipatov Y. S. // Polymer. 1999. Vol. 40. P. 1347.
10. Nesterov A. E., Lipatov Y. S., Ignatova T. D., Nesterov D. A. // Polymer. 2002. Vol. 43. P. 875.
11. Lipatov Y. S. // Prog. Polym. Sci. 2002. Vol. 27. P. 1721.
12. Lipatov Y. S. // J. Macromol. Sci. (B). 2006. Vol. 45. P. 871.
13. GinzburgV. V. // Macromolecules. 2005. Vol. 38. P. 2362.
14. HoreM. J., LaradjiM. //J. Chem. Phys. 2007. Vol. 126. 244903.
15. ModaressiA., BoulmokhA., MuteletF. et al. // Coll. Surf. (A). 2009. Vol. 338. P. 47.
16. Hossein SajjadiH., ModaressiA., MagriP. et al. // J. Mol. Liq. 2012. (In press.)
17. Israelachvili J. N. Intermolecular and Surface Forces: 2nd edition. New York: Academic Press, 1992.
k Nps free ■ Ag Nps
■ aa
18. FichthornK. A., QinY. // Gran. Matt. 2008. Vol. 10. P. 10105.
19. QinY., FichthornK. A. // Phys. Rev. (E). 2006. Vol. 73. 020401.
20. FichthornK. A., Qin Y. // Ind. Eng. Chem. Res. 2006. Vol. 45. P. 5477.
21. FichthornK. A., QinY. // J. Chem. Phys. 2007. Vol. 127. 144911.
22. ProvencherS. W. // Comput. Phys. Commun. 1982. Vol. 27. P. 229.
23. Sazonov V. P., ShawD., SkrzeczA., SazonovN. // IUPAC-NIST Solubility Data Series. 83. Acetonitrile: Ternary and Quaternary Systems. 2007. P. 908.
24. Sazonov V. P., ShawD., SkrzeczA., SazonovN. // IUPAC-NIST Solubility Data Series. 78. Acetonitrile: Binary Systems. 2002. P. 407.
Статья поступила в редакцию 3 октября 2012 г.
