252
Langmuir 2008, 24, 252-258
Tailor-made Surface Properties of Particles with a Hydrophilic or Hydrophobic Polymer Shell Mediated by Supercritical CO2 Samuel Marre, Franc¸ ois Cansell, and Cyril Aymonier* Institut de Chimie de la Matie` re Condense´ e de Bordeaux, ICMCB-CNRS and ENSCPB, UniVersite´ Bordeaux I, 87 aVenue du docteur Albert Schweitzer, 33608 Pessac cedex, France ReceiVed July 18, 2007. In Final Form: September 25, 2007 Controlling the surface characteristics of inorganic materials with an organic shell is of great interest for control of the properties of the final material. A challenge is thus to be able to deposit a polymer shell with different solvation properties onto the surface of inorganic particles and to have a good control of the thickness of the organic layer without a prefunctionalization of surfaces. We demonstrate, in this paper, a method for coating silica particles (170-550 nm), used as model substrates, with either a hydrophilic (polyethylene glycol) or a hydrophobic polymer (polybutadiene hydroxy terminated) using a supercritical antisolvent process (precipitation from a compressed antisolvent). Several operating parameters were studied to control precisely the thickness of the deposited layer (from 2 to 30 nm), which was characterized using TEM, FESEM, XPS, and UV-visible techniques. This work demonstrates that supercritical antisolvent processes are powerful methods and good alternatives to conventional coating techniques toward the development of hybrid and/or core-shell nanomaterials.
1. Introduction In recent years, much research has focused on the control of nanoparticles surface properties. As particle dimension decreases, the surface composition and the surface area of particles become dominant factors in determining their end-use application. In particular, polymer-coated micro- and nanoparticles (core-shell particles) have been highly investigated due to their unique properties derived from the combination of several materials.1-3 Coating and/or encapsulation of nanoparticles are general approaches, which can be used to tune a large range of physical and chemical properties, such as solubility, chemical reactivity, biological efficiency, and optical, mechanical, electrical, thermal, and catalytic properties. The limiting factor in controlling the deposition of a polymer shell onto the surface of nanoparticles is still good control of the nature and the thickness of the coating material. The deposition of thin polymer layers on nanomaterials such as carbon nanotubes, inorganic nanorods or nanospheres, is usually performed using predominantly wet chemical methods,4 such as solvent evaporation,5,6 surface adsorption of polymers,7-9 selfassembled polymer layers,10,11 heterocoagulation,12-14 conventional polymerization,15-22 or dry CVD-assisted polymerization method.23 * To whom correspondence should be addressed. E-mail: aymonier@ icmcb-bordeaux.cnrs.fr. (1) Boal, A.; Ilham, F.; DeRouchey, J.; Thurn-Abrecht, T.; Russell, T.; Rotello, V. Nature 2000, 404, 746-748. (2) Bourgeat-Lami, E. J. Nanosci. Nanotech. 2002, 2, 1-24. (3) Schmidt, G.; Malwitz, M. Curr. Opin. Colloid Interface Sci. 2003, 8, 103108. (4) Caruso, F. AdV. Mater. 2001, 13 (1), 11-22. (5) Cohen, H.; Levy, R.; Gao, J.; Kousaev, V.; Sosnowski, S.; Slomkowski, S.; Golomb, G. Gene Ther. 2000, 7, 1896-1905. (6) Wang, D.; Robinson, D.; Kwon, G.; Samuel, J. J. Controlled Release 1999, 57, 9-18. (7) Kang, Y.; Taton, A. Macromolecules 2005, 38, 6115-6121. (8) Ditsch, A.; Laibinis, P.; Wang, D.; Hatton, T. Langmuir 2005, 21, 60066018. (9) Leventis, N. Acc. Chem. Res. 2007, 40, 874-884. (10) Caruso, F.; Lichtenfeld, H.; Donath, E.; Mo¨hwald, H. Macromolecules 1999, 32, 2317-2328. (11) Caruso, F.; Mo¨hwald, H. Langmuir 1999, 15, 8276-8281. (12) Caruso, R.; Susha, A.; Caruso, F. Chem. Mater. 2001, 13, 400-409. (13) Caruso, F.; Spasova, M.; Salgueirin˜o-Maceira, V.; Liz-Marza´n, L. AdV. Mater. 2001, 13, 1090-1094.
While these processes can produce encapsulated particles or aerogels,9 they have some inherent drawbacks to chemical processes. As an example, polymerization processes generally require prefunctionalization of inorganic surfaces by the adsorption or the covalent bonding of either coupling agents or monomers. However, it is not possible to prefunctionalize every kind of surface and this can lead to incomplete surface coating. In addition, these methods require large quantities of solvents, surfactants, and other additives, which may create end-use problems such as contamination of products. Thus, for certain applications, like the synthesis of nanovectors for drug delivery, many post-treatment steps are generally required to remove either remaining solvents or nondesirable compounds. Moreover, in the case of pharmaceutical applications, the temperatures and pH used in these coating processes are critical factors to preserve, intact, the nature of the core material.24 Other processes, such as the self-assembly of charged polymers or the heterocoagulation, allow only the deposition of polymers that have an electrostatic charge, which limits the number of species that can be deposited or requires a chemical modification of the coating material.25 Because of all the above-mentioned drawbacks, coating methods using supercritical fluids are useful and powerful alternatives over the conventional chemical coating routes. These (14) Fleming, M.; Mandal, T.; Walt, D. Chem. Mater. 2001, 13, 2210-2216. (15) Wolfe, D.; Oldenburg, S.; Westcott, S.; Jackson, J.; Paley, M.; Halas, N. Langmuir 1999, 15, 2745-2748. (16) Mandal, T.; Fleming, M.; Walt, D. Nano Lett. 2002, 2 (1), 3-7. (17) Cheng, X.; Chen, M.; Zhou, S.; Wu, L. J. Polym. Sci. A 2006, 44, 38073816. (18) Luna-Xavier, J.; Guyot, A.; Bourgeat-Lami, E. J. Colloid Interface Sci. 2002, 250, 82-92. (19) Castelvetro, V.; De Vita, C. AdV. Colloid Interface Sci. 2004, 108-109, 167-185. (20) Reculusa, S.; Poncet-Legrand, C.; Perro, A.; Duguet, E.; Bourgeat-Lami, E.; Mingotaud, C.; Ravaine, S. Chem. Mater. 2005, 17 (13), 3338-3344. (21) Bourgeat-Lami, E.; Lang, J. J. Colloid Interface Sci. 1998, 197, 293308. (22) Zhang, K.; Chen, H.; Chen, X.; Chen, Z.; Cui, Z.; Yang, B. Macromol. Mater. Eng. 2003, 288, 380-385. (23) Kenneth, K.; Lau, S.; Gleason, K. AdV. Mater. 2006, 18, 1972-1977. (24) Fu, K.; Griebenow, K.; Hseih, L.; Klibanov, V.; Langer, R. J. Controlled Release 1999, 58, 357-366. (25) Yap, H.; Quinn, J.; Ng, S.; Cho, J.; Caruso, F. Langmuir 2005, 21 (10), 4328-4333.
10.1021/la702154z CCC: $40.75 © 2008 American Chemical Society Published on Web 11/30/2007
Surface Properties of Particles with Polymer Shell
processes can indeed be used to deposit organic materials using physical processes26-28 and inorganic materials from chemical transformation of matter.29-33 Physical processes using supercritical fluids (SCF) are divided into three main classes depending on the role of the SCF:34,35 (i) as a solvent (RESS, rapid expansion of a supercritical solution),36 (ii) as an antisolvent (SAS, supercritical antisolvent),37,38 or (iii) as a solute (PGSS, particle from a gas-saturated solution).39 Of all the three classes mentioned above, SAS processes such as the precipitation from a compressed antisolvent (PCA), which is described in detail below, are the most widely used processes for the deposition of polymers. Indeed, these processes allow a wide range of solvent choice and polymers generally have low solubilities in supercritical fluids especially in supercritical CO2 (scCO2).40 The principle is to put a solution [solvent + solute] in contact with the supercritical antisolvent. To create the deposition, the supercritical antisolvent has to be dissolved into the solution to decrease its density and its ability to solubilize materials. The injection of the supercritical antisolvent supersaturates the solvent with the solute, causing it to precipitate. The solvent has to be chosen as (i) a good solvent for the coating material and (ii) a nonsolvent for the core material and being (iii) miscible with the antisolvent (ideally, they should be completely miscible) and (iv) easily removable from the coated material at the end of the process. Several organic materials have been successfully encapsulated with polymers using the SAS process, mainly for pharmaceutical41,42 or biomedical43,44 applications, to protect the core materialsgenerally an active substanceswith a biocompatible polymer shell. In addition, Zhang et al. have reported the coating of inorganic materialssTiO2 nanoparticlessby polystyrene using an ultrasonic-assisted SAS process.45 They have demonstrated that the ultrasonic probe reduces the agglomeration of particles during the coating process. Other studies have reported on the coating of surface-modified silica particles by poly(lactic-coglycolic acidsPLGA) using the PCA process.46,47 However, despite the coating’s importance in the characteristics of the (26) Reverchon, E.; Adami, R. J. Supercrit. Fluids 2006, 37, 1-22. (27) Hertz, A.; Sarrade, S.; Guizard, C.; Julbe, A. J. Eur. Ceram. Soc. 2006, 26 (7), 1195-1203. (28) Reverchon, E.; Della Porta, G. Pure Appl. Chem. 2001, 73 (8), 12931297. (29) Aymonier, C.; Loppinet-Serani, A.; Reveron, H.; Garrabos, Y.; Cansell, F. J. Supercrit. Fluids 2006, 38 (2), 242-251. (30) Marre, S.; Cansell, F.; Aymonier, C. Nanotech. 2006, 17, 4594-4599. (31) Pessey, V.; Garriga, R.; Weill, F.; Chevalier, B.; Etourneau, J.; Cansell, F. Ind. Eng. Chem. Res. 2000, 39 (12), 4714-4719. (32) Cansell, F.; Aymonier, C.; Loppinet-Serani, A. Curr. Opin. Solid State Mater. Sci. 2003, 7, 331-340. (33) Reveron, H.; Aymonier, C.; Elissalde, C.; Weill, F.; Maglione, M.; Cansell, F. J. Nanosci. Nanotech. 2005, 5 (6), 980-983. (34) Reverchon, E.; Della Porta, G. Pure Appl. Chem. 2001, 73 (8), 12931297. (35) Yeo, S.; Kiran, E. J. Supercrit. Fluids 2005, 34, 287-308. (36) Tu¨rk, M. J. Supercrit. Fluids 1999, 15, 79-89. (37) Reverchon, E.; Della Porta, G.; De Rosa, I.; Subra, P.; Letourneur, D. J. Supercrit. Fluids 2000, 18, 239-245. (38) Dixon, D.; Johnston, K.; Bodmeir, R. AIChE 1993, 39, 127-139. (39) El Vassore, N.; Flaibani, M.; Bertucco, A.; Caliceti, P. Ind. Eng. Chem. Res. 2003, 42 (23), 5924-5930. (40) O’Neill, M.; Cao, Q.; Fang, M.; Johnston, K.; Wilkinson, S.; Smith, C.; Kerschner, J.; Jureller, S. Ind. Eng. Chem. Res. 1998, 37, 3067-3079. (41) Elvassore, N.; Bertucco, A.; Caliceti, P. J. Pharm. Sci. 2001, 90 (10), 1628-1636. (42) Tu, L.; Dehghani, F.; Foster, N. Powder Technol. 2002, 126, 134-149. (43) Elvassore, N.; Bertucco, A.; Caliceti, P. Ind. Eng. Chem. Res. 2001, 40, 795-800. (44) Falk, R. R.; Randolph, T.; Meyer, J.; Kelly, R.; Manning, M J. Controlled Release 1997, 44, 77-85. (45) Zhang, J.; Liu, Z.; Buxing, H.; Li, J.; Li, Z.; Yang, G. J. Nanosci. Nanotech. 2005, 5 (6), 945-950. (46) Wang, Y.; Dave, R.; Pfeffer, R. J. Supercrit. Fluids 2004, 28, 85-89. (47) Wang, Y.; Pfeffer, R.; Dave, R.; Enick, R. AIChE 2005, 51 (2), 440-455.
Langmuir, Vol. 24, No. 1, 2008 253 Table 1. PBHT and PEG Characteristics
final core-shell material, some key characteristics of the coating remain uninvestigated, such as the thickness of the polymer layer or the nature of the polymer that can be deposited without surface modification of the substrate. We report in this work the coating of silica spheres by either a hydrophilic (polyethylene glycol) or a hydrophobic polymer (polybutadiene hydroxy terminated). By doing this, we aim to answer these open questions: is it possible to control precisely the thickness of the polymer shell? In which range of thickness? Do the shell characteristics depend on (i) the hydrophilic or hydrophobic nature of the polymer? (ii) The size of the particles to coat? The coating was performed at low temperature, without any pre-functionnalization of surface, using a PCA supercritical antisolvent process. First, the experimental setup and the fundamentals of the supercritical antisolvent processes are described. Then the influence of several key parameters like the initial polymer fraction, the speed of stirring in the reaction media, the injection flow rate of the solution, or the temperature is discussed in correlation with the characteristics of the coating of silica spheres. 2. Experimental Section 2.1. Chemicals. The support material was silica spheres either purchased from Alfa Aesar (550 nm) or homemade using the Sto¨ber process (170 nm).48 The polymers used as coating agents, hydrophilic PEG and hydrophobic PBHT, were both purchased from Aldrich and used as received. Their main characteristics are listed in Table 1. At ambient conditions, PEG is a solid and PBHT is a viscous liquid. Ethanol and dichloromethane (DCM), purchased from Aldrich, were used as the solvents for PEG and PBHT, respectively. The CO2 was used as received from Air Liquide. 2.2. Method and Experimental Setup. The principle of the PCA coating of particles with a polymer shell results from the combination of different phenomena (hydrodynamic, precipitation kinetics, thermodynamic, and mass-transfer effects); it can be described using a [solvent/antisolvent/polymer] ternary diagram (Figure 1). This ternary diagram is composed of two main areas: (1) the polymer fraction dissolved in the solvent in the presence of antisolvent; (2) the polymer phase in which small proportions of solvent and antisolvent are dissolved. Between these two areas, two phases coexist, one polymer-rich and one solvent/antisolvent-rich. The particles to be coated are dispersed in a solution containing a known quantity of polymer dissolved in the solvent (Figure 1, step a). This solution is injected inside the reactor filled with supercritical CO2 (scCO2). This injection induces the formation of solution droplets inside the continuous scCO2 phase. An interdiffusion starts between the two phases and the droplets are expanded by scCO2.49 The increase of the scCO2 molar fraction in the solution droplets causes a displacement in the ternary diagram following the dotted line. The (48) Sto¨ber, W.; Fink, A.; Bohn, E. J. Colloid Interface Sci. 1968, 26, 62-69. (49) Reverchon, E. J. Supercrit. Fluids 1999, 15, 1-21.
254 Langmuir, Vol. 24, No. 1, 2008
Marre et al.
Figure 1. Different steps of the particles coating with polymers using antisolvent processes. Experimental setup: (1) High-pressure hightemperature stirred reactor, (2) scCO2 inlet, (3) heater, (4) solution inlet, and (5) filter.
solvent loses its solvent power with respect to the polymer (Figure 1, step b). The reaction medium inside the droplets [solvent + antisolvent] becomes supersaturated with polymer as the solubility curve is crossed, inducing a phase transition with the formation of a polymer-rich phase encapsulating the silica spheres and a solvent/ antisolvent-rich phase (Figure 1, step c). Thereafter, the precipitation of the polymer occurs on the surface of the silica particles, which act as nuclei in the reaction media.46 The solvent is removed continuously by a cleaning step with the scCO2. The disappearance of the solvent causes the precipitation of the entire polymer (Figure 1, step d, arrow). The system is finally purged with CO2 to recover the polymer-coated silica spheres (Figure 1, step e). The setup used to coat the silica spheres with a polymer shell is described in Figure 1. It consists of a high-pressure, high-temperature stirred reactor (50 cm3) made of 316 stainless steel (1), which is equipped with a pressure sensor and a thermocouple, placed inside the reactor. In a typical experiment, a solution containing the dissolved polymer and the silica spheres in suspension in a solvent is prepared first. In the meantime, the antisolvent (scCO2) is pumped by a feeding system (2) composed of a bottle of CO2, a cooling apparatus, a high-pressure pump, and a preheater (50 °C). Thus, the CO2 is in its supercritical conditions before its injection at a constant flow (11 g min-1) inside the stirred reactor, which is kept at constant temperature using a heating element (3). Once the operating conditions are reached (50 °C, 15 MPa), the purge valve is opened so that scCO2 can flow through the reactor, while maintaining the system at constant pressure. Thereafter, the solution (4) is sprayed through a nozzle using an injection line which continuously feeds the reactor with the solution at different flow rates ranging from 1 to 4 mL min-1. The solvent/antisolvent mixture flows through a filter (5) which allows the recovery of the coated particles. Once the injection of the solution is done, the flow of scCO2 is maintained constant to remove the residual traces of solvent. Finally, the system is slowly depressurized and the coated silica particles are collected at the bottom of the reactor. The influence of four parameters on the characteristics of the polymer shell, especially the thickness of the shell, was investi-
gated: the injection flow rate of the solution (Fr), the stirring speed (Sstir), the temperature (T), and the initial polymer mass percentage (ξ), defined as: ξ)
mpol (mpol + mSiO2)
(1)
where mSiO2 and mpol are the mass of silica spheres and polymer initially introduced inside the solution, respectively. Each initial solution was prepared by mixing 100 mL of solvent (ethanol or DCM depending on the polymer), 500 mg of SiO2 spheres, and a variable amount of polymer defined as [ξ/(1 - ξ)] × 500 mg. 2.3. Characterization Techniques. The presence of polymer on the surface of silica spheres was first identified by X-ray photoelectron spectroscopy (XPS). The XPS measurements were performed with an ESCALAB 220i-XL apparatus on powder samples deposited onto indium foil. Since it is difficult to determine the thickness of thin organic layers deposited on a surface, we chose to estimate the thickness of the polymer shell by three methods: (1) By a direct measurement of the diameter of the particles before and after coating from micrographs of field emission scanning electron microscopy (FESEM, JEOL 6700F). The samples were prepared by deposition of a SiO2@polymer colloidal solution drop on a coppercarbon TEM grid and then evaporation at ambient conditions. (2) By an indirect measurement based on the weight of polymer deposited per surface of silica. The specific surface area of silica spheres was first determined by the BET method (Brunauer, Emmett, Teller - ASAP 2010 Micromeritics apparatus). The used silica sphere material (544 ( 22 nm in diameter) has a specific surface area of 7 m2 g-1. After the coating experiment, the particles are placed in ethanol or DCM to dissolve the deposited PEG or PBHT, respectively. The polymers are thereafter proportioned by UV-visible spectroscopy (Cary 1C UV-visible spectrophotometer apparatus). The deduced polymer volume is divided by the specific surface area to give a thickness value, assuming that the deposition is homogeneous.
Surface Properties of Particles with Polymer Shell
Langmuir, Vol. 24, No. 1, 2008 255
Table 2. Experimental Conditions and Thicknesses of Polymer Shells for the Different Experiments δ polymer thickness (nm) sample
polymer
1 2 3 4 5 6 7 8
PEG
ξ (%)
Fr (mL/min)
1 2 5 10 20
4 50 4 1 4
1 2 5 10 20 PBHT
300
800
4 2 1
800
4
1800 1300 800 300
(3) By theoretical calculation from eq 2 assuming that the shell thickness of polymer (δ) is uniform on the surface of all the silica particles, δ)
x(
1+
)
dSiO2 ξ × × r3 - r 1-ξ dpoly
(
)
calculated 1.6 3.3 8.0 15.5 29.6 8.0
80
4
5
3
T (°C)
800
5
9 10 11 12 13 14 15 16 17 18 19 20
Sstir (rpm)
(2)
where ξ is the initial mass fraction of polymer, dSiO2 and dpoly the silica and polymer density, respectively, and r is the mean radius of the uncoated silica particles (obtained from TEM measurements).
3. Results and Discussion Table 2 lists the experiments that were realized by varying the experimental parameters to coat the core silica particles (544 ( 22 nm). At constant flow rate Fr ) 4 mL min-1, stirring speed Sstir ) 800 rpm, and temperature T ) 50 °C, the polymer fraction ξ was tuned in the range of 1-20% by weight, with regards to the silica (tests 1-5 and 9-13). The other parameters were tuned at constant weight fraction of polymer ξ ) 5% as follows: Fr ) 1-4 mL min-1, Sstir ) 300-1800 rpm, and T ) 50-80 °C. The results are summarized in Table 2. Figure 2a shows a FESEM picture of typical silica particles before coating, whereas Figures 2b and 2c present PEG- and PBHT-coated silica spheres, respectively (tests 4 and 12, ξ ) 10%, T ) 50 °C, P ) 15 MPa, Sstirr ) 800 rpm, Fr ) 4 mL min-1). The FESEM micrograph of uncoated silica particles shows that they are spherical with a narrow size distribution centered at 544 nm, which allows us to measure the thickness of the polymer shell. Then, for both samples of polymer coatings, the average size of particles increases to 574 ( 24 and 573 ( 22 nm for PEG- and PBHT-coated particles, respectively (ξ ) 10%). In addition, one can see that the initial roughness of the silica particles tends to disappear after the coating. The presence of polymer on the surface of silica spheres was confirmed by XPS analyses (Figure 3). An Ar+ ions etch was performed to scan the variation of the atomic percentage of Si, O, and C as a function of the etching depth. Figure 3 shows this evolution for uncoated silica particles and for test 12. On one hand, the resulting atomic percentage shows that the uncoated silica particles are polluted by a small amount of carbon, probably coming from the synthesis method. Indeed, after the etching of an initial carbon layer on the surface, the carbon percentage is stable (8%, Figure 3a).
2.0 4.0 9.7 18.8 35.4 50 9.7
MEB results
UV/BET results
3.0 ( 1 4.5 ( 1 8.0 ( 0.5 16.0 ( 1 28.0 ( 1 7.5 ( 0.5 8.5 ( 1 8.0 ( 1
1.5 ( 0.5 3.0 ( 0.5 6.5 ( 1 14.0 ( 1 29.0 ( 1
2.5 ( 1 4.5 ( 1 9.0 ( 1 14.5 ( 1 22.0 ( 1 7.5 ( 0.5 8.5 ( 1 7.0 ( 1 9.5 ( 0.5 8.5 ( 1 7.5 ( 0.5 9.0 ( 0.5
2.0 ( 0.5 3.0 ( 0.5 7.0 ( 1 15 ( 1 27 ( 1
On the other hand, at the surface of the PBHT-coated silica particles the carbon percentage is 85% (PBHT is 98% of carbon), meaning that the surface is completely covered by the polymer. Further, the carbon percentage decreases with an increasing etching depth. At an etching depth of 60 nm, the observed Si and O percentages correspond almost exactly to those of the uncoated silica particles (Figure 3b). Nevertheless, as X-rays cannot be well-focused, the results represent the average over several silica spheres so that the PBHT thickness of one silica particle cannot be precisely determined by this technique. Experiments on smaller silica spheres made in our lab (average diameter ) 167 ( 12 nm) show similar results, for the same value of polymer fraction (ξ ) 10%), except for a thinner polymer shell (δ ) 4.5 ( 1 nm) than the one obtained for the 544 nm silica spheres (δ ) 14.5 ( 1 nm). This is in good agreement with the fact that the smaller silica particles developed a larger surface area (21 m2 g-1 for 167 nm silica particles instead of 7 m2 g-1 for the 544 nm ones). Figure 4 shows TEM pictures obtained for SiO2@PBHT particles with ξ ) 20%. The polymer film is visible between the silica particles, as indicated. This shows that the coating process can be equally applied to particles of different sizes.46 Figure 5 shows the influence of the initial polymer fraction on δ, all other parameters being constant (tests 1-5). The values of δ obtained by theoretical calculations for the different values of ξ are in good agreement with those obtained from FESEM micrographs measurements or from UV-BET data. This indicates that both methods are suitable for measuring polymer layer thickness and validates the assumptions that the polymer is uniformly deposited on the surface of the silica particles. In addition, these results show that the PEG is entirely present at the surface of the silica spheres. From Figure 5, one can see that the PEG-coated silica particles are well-separated with little agglomeration, even at high polymer fraction. To sum up the results that have been presented, we were able to obtain a precise control of the polymer thickness between 2 and 30 nm by varying ξ between 1 and 20%. The same set of experiments was performed with the PBHT coating (tests 9-13). The influence of ξ on the δ values is shown in Figure 6, with all other parameters held constant. As in the case of the PEG shell, the thickness of the PBHT shell (δ) on the silica spheres can be precisely tuned at the nanometer scale,
256 Langmuir, Vol. 24, No. 1, 2008
Marre et al.
Figure 3. Atomic percentage as a function of the depth of etching obtained from XPS analysis of initial silica particles (a) and test 12 (b).
Figure 2. FESEM pictures and size distribution measurements of the initial silica particles (a) and of typical SiO2@PEG and SiO2@PBHT particlesstests 4 and 12 (b, c).
in the range of 2-22 nm, by varying the value of ξ from 1 to 20%. For values of ξ < 10%, the measured values of δ are in good agreement with the theoretical calculations. However, at values of ξ > 10%, FESEM measurements give values of δ below those obtained by the UV-BET method, both of which are less than the theoretical values of δ. The observed variations between the theoretical values and those from the UV-BET method are probably due to an incomplete precipitation of the PBHT on the silica particles surface. It is possible that a fraction of the PBHT remains dissolved in the solvent/antisolvent mixture. When CO2 flow is increased, more polymer can be forced out of the solvent on the silica surface. The variations observed between the theoretical values of δ and FESEM measurements might be caused by an agglomeration of the coated silica particles induced by the viscous liquid behavior
of the PBHT. To reduce particle agglomeration, all the experiments were performed with polymer concentrations between 0.05 and 1 mg mL-1, as suggested in the literature.47 However, the nature of the polymer also plays a significant role. A closer look at the agglomeration of PBHT-coated silica particles as a function of ξ (Figures 7a-7d) shows that the higher the value of ξ, the larger the agglomerates. PBHT is a viscous liquid under ambient conditions, so for high ξ values, the PBHT may spread between the silica particles, leading to agglomeration (Figure 7e). Under these conditions, the PBHT that is located between the silica particles is proportioned by the UV-visible measurement but does not participate in the increase of the real deposited thickness obtained by the FESEM measurements. This behavior was not observed for the PEG-coated silica particles. The influence of three other operating parameters was investigated (tests 6-8 and 14-20). The temperature was varied to study the effect of the glass transition temperature of the polymer. For PEG, experiments were performed at 50 and 80 °C, which is above and below Tg (54 °C). All PBHT experiments were performed at 50 °C because the Tg of PBHT is -70 °C, well below the critical temperature of CO2. A comparison of the results of tests 6 and 8 suggests that the temperature does not influence δ. The remaining investigated parameters, i.e., the stirring speed (Sstir) and the flow rate (Fr), showed no influence on the thickness of the deposited polymer shell in our operating conditions (tests 6-7 and 14-20). To sum up the results obtained in our conditions of the investigated parameters, the initial mass fraction of polymer had the greatest influence on the thickness of the polymer coating.
Surface Properties of Particles with Polymer Shell
Langmuir, Vol. 24, No. 1, 2008 257
Figure 6. Thicknesses of the PBHT layers (δ) either calculated or measured (FESEM or UV/BET method) as a function of the initial mass fraction of polymer in the solution (ξ) (tests 9-13).
Figure 4. TEM micrographs of 167 nm silica particles coated with PBHT (ξ ) 20%).
Figure 5. Thicknesses of the PEG layer (δ) either calculated or measured (FESEM or UV/BET method) as a function of the initial mass fraction of polymer in the initial solution (ξ) (tests 1-5).
Figure 7. SEM micrographs of PBHT-coated (a-d) silica particles for different values of ξ. (a) ξ ) 2%, (b) ξ ) 5%, (c) ξ ) 10%, (d) ξ ) 20, and (e) explanation of the differences observed for high values of ξ between the δ obtained by SEM and UV/BET measurement methods.
The variation of ξ allows us to control precisely the thickness of the shell down to the nanometer scale, in the range of 2-30 nm. In general, the deposition of two different kinds of polymer (hydrophilic and hydrophobic) on the surface of silica particles can be achieved using the same process and results in similar morphologies. The apparition of agglomerated particles for coatings with PBHT at high value of ξ is inherent to the nature of the PBHT. Finally, the coatings were stable in time, with no variations observed in the 3 weeks after the deposition.
The PCA process, as all SAS processes, turns out to be a powerful tool for advanced material synthesis, allowing the deposition of different materials, with different characteristics, with a control of the polymer shell thickness in the range of 2-30 nm by varying ξ between 1 and 20%. However, thicker thicknesses could be obtained for higher values of ξ. The large range of thicknesses that can be deposited in a singlestep process is a great advantage compared to that of conventional coating processes. Indeed, methods like polymerization or selfassembled polymer layers either do not allow such precise control
258 Langmuir, Vol. 24, No. 1, 2008
at the nanometer scale or require several steps to obtain thicker shells.
4. Conclusion We have demonstrated in this work the coating of silica particles, used as model substrates, by PEG and PBHT, hydrophilic and hydrophobic polymers, respectively. The coating was performed without any prefunctionalization of the surface of silica spheres and at temperature as low as 50 °C, with precise control of the thickness of the polymer shell ranging from 2 to 30 nm in the studied conditions. The agglomeration of the particles appears for PBHT coatings at high polymer fraction (above 10%), but can be prevented by working at low polymer concentration.
Marre et al.
The physical supercritical antisolvent processes turn out to be good alternatives to conventional coating processes as they allow depositing polymers on substrates irrespective of their initial chemical properties. These processes open a route for the synthesis of hybrid and/or core-shell multifunctional materials via control of the surface properties by coating. Acknowledgment. The authors thank C. Marraud, J. Renouard, G. Chounet, M. S. Amiet, and O. Roushdy for the attention they have shown to this work. Financial support from the European Community in the frame of the SUPERMAT network is gratefully acknowledged. LA702154Z
