Understanding the Effectiveness of Fluorocarbon Ligands in

Aug 31, 2010 - Introduction. A proper understanding and control of the synthesis and assembly ...... selected/characterized on the basis of cloud-poin...
0 downloads 0 Views 3MB Size
Download PDF
J. Phys. Chem. C 2010, 114, 15553–15561

15553

Understanding the Effectiveness of Fluorocarbon Ligands in Dispersing Nanoparticles in Supercritical Carbon Dioxide Vishwanath Haily Dalvi,† Vibha Srinivasan,‡ and Peter J. Rossky*,†,‡,§ Department of Chemical Engineering, UniVersity of Texas, Austin, Texas 78712, Department of Chemistry and Biochemistry, UniVersity of Texas, Austin, Texas 78712, and Institute for Computational Engineering and Sciences, UniVersity of Texas, Austin, Texas 78712 ReceiVed: March 6, 2010; ReVised Manuscript ReceiVed: July 17, 2010

To augment the experimental search for a nonfluorous capping ligand that is effective in dispersing nanoparticles in supercritical carbon dioxide, we have developed a simulation protocol, involving atomistic molecular dynamic simulations, which semiquantitatively reproduces empirical observations and hence can be used to gain insight into the relevant physical phenomena. We have used this protocol to examine the reasons behind the exceptional effectiveness of perfluoroalkanethiols, compared to alkanethiols, as capping ligands for nanoparticles in supercritical carbon dioxide. From these simulations, we infer that the principal reason for this enhanced effectiveness is that the C-F bond is much more polar than the C-H bond and hence the fluorocarbons can interact with the quadrupolar carbon dioxide in addition to interacting with its van der Waals centers. For the models studied, the effect of the electrostatic interaction offsets the fact that the dispersion forces exerted by a dense fluorocarbon layer are weaker than in the case of the alkanethiols due to the sparser packing of the former, larger, ligands. The sparse packing and a molecular geometry that ensures that the C-F dipoles of two fluoroalkane molecules on opposing surfaces are always adVerse makes two fluorocarbon-passivated surfaces less attractive to each other than the corresponding hydrocarbon-passivated surfaces. Hence, fluoroalkanethiol-passivated surfaces can be both more CO2-philic and less autophilic than alkanethiol-passivated surfaces, which we conclude leads to the effectiveness of fluorocarbon ligands in supercritical carbon dioxide. 1. Introduction A proper understanding and control of the synthesis and assembly of nanoparticles is essential in order to cheaply and efficiently implement the great variety of promising new technologies collectively termed “nanotechnology”. To collect, separate, and deposit nanoparticles after synthesis, it is particularly important to be able to create and control their dispersions in fluid media.1 Bare nanoparticles of many interesting materials (metals, semiconductors, etc.) have high surface energies and need their surfaces passivated by capping ligands to disperse, at reasonable conditions of pressure and temperature, in common solvents.2 The stability of dispersions of such ligand-passivated nanoparticles depends critically on the properties of the ligand tail and the solvent as well as conditions of temperature and pressure. Understanding the physical phenomena governing the stability of ligand-passivated nanoparticle dispersions is of academic as well as industrial importance, since certain solvents and ligands are more suited to large-scale applications than others.3-6 Among the solvents, supercritical fluids, whose solvation power is a strong function of pressure, would allow reversible and tunable stabilization of nanoparticle dispersions and are very desirable for large scale use. Of these, supercritical carbon dioxide (ScCO2) with its easily accessed critical region combined with nonflammability and nontoxicity is especially attractive. However, only nanoparticles capped with ligands whose tails are composed of perfluoropolyethers7 and perfluo* Corresponding author. Phone: (512) 471-3555. Fax: (512) 471-1624. E-mail: [email protected]. † Department of Chemical Engineering. ‡ Department of Chemistry and Biochemistry. § Institute for Computational Engineering and Sciences.

roalkanes8 have so far been identified as forming stable dispersions in ScCO2 at convenient conditions of temperature (near ambient) and pressure ( 150 atm are linear.

15556

J. Phys. Chem. C, Vol. 114, No. 37, 2010

Figure 3. Plot of Fpore vs 1/L for perfluorooctanethiol-passivated slit pore with CO2 as solvent. The symbols are the result of solutions of eq 4, while the solid lines are the straight line fits through them. Two bulk pressures are represented with Pbulk ) 150 atm (circles, lower line) and Pbulk ) 300 atm (diamonds, upper line). The error bars for the fit are less than the size of the symbols.

Figure 4. Plot of Ppore vs L calculated as described for perfluorodecanethiol-passivated pores solvated by CO2 for various values of bulk pressures. The symbols are the result of said calculation, while the dashed lines are meant to guide the eye. The four bulk pressures represented are Pbulk ) 150 atm (circles), Pbulk ) 200 atm (squares), Pbulk ) 250 atm (triangles), and Pbulk ) 300 atm (diamonds). The solid line is the corresponding curve for interactions in a vacuum, and the dot-dashed line corresponds to interaction in a vacuum if the F atoms had the same charge as the H atoms of n-dodecanethiol. For clarity, only error bars for Pbulk ) 150 atm and Pbulk ) 300 atm are shown. The error bars for the other pressures are similar. Interactions for L < 40 Å are strongly repulsive (high Ppore), and the corresponding values are not shown.

Using the methods described in section 2 and ref 26, Ppore vs L curves for different values of Pbulk are calculated for the CO2 solvated pore passivated by perfluorodecanethiol (Figure 4) and perfluorooctanethiol (Figure 5) so that, using eq 2, we get WPMF vs L plots for various values of Pbulk (see Figures 6 and 7). Density profiles in a large gap (L ) 100 Å) slit pore for both solvent and ligand appear in Figures 8 and 9 for perfluorodecanethiol- and perfluorooctanethiol-passivated pores, respectively. Identifying Leq ) 43 Å (in a vacuum, Leq-vac ) 40 Å) for perfluorodecanethiol-passivated pores and Leq ) 35 Å (Leq-vac ) 33 Å) for perfluorooctanethiol-passivated pores, we get, from eq 1, the W vs Pbulk plots seen in Figure 10. In the absence of solvent, equilibrium occurs at Leq-vac, and for L < Leq-vac, the overlap of ligand regions causes large repulsive interactions. Upon fitting the equation of state, we get characteristic parameters for CO2 (εf ) 0.807 kcal/mol, σf ) 2.9 Å).26 After first discussing the consistency of our results with experimental observations, we will comment on the applicability

Dalvi et al.

Figure 5. Plot of Ppore vs L calculated as described in the text for perfluorooctanethiol-passivated pores solvated by CO2 for various values of bulk pressures. The symbols are the result of calculation, while the dashed lines are meant to guide the eye. The four bulk pressures represented are Pbulk ) 150 atm (circles), Pbulk ) 200 atm (squares), Pbulk ) 250 atm (triangles), and Pbulk ) 300 atm (diamonds). The solid line is the corresponding curve for interactions in a vacuum, and the dot-dashed line corresponds to interaction in a vacuum if the F atoms had the same charge as the H atoms of n-dodecanethiol. For clarity, only error bars for Pbulk ) 150 atm and Pbulk ) 300 atm are shown. The error bars for the other pressures are similar. Interactions for L < 33 Å are strongly repulsive (high Ppore), and the corresponding values are not shown.

Figure 6. Potential of mean force of the slit pore for perfluorodecanethiol-passivated slit pores solvated by CO2 at various pressures eff is evaluated using Ltrunc ) 60 Å. The free calculated using eq 2. Pbulk energy corresponding to a single periodic image is attributed to a pair of particles to calculate the free energy per mole. The symbols are the result of said calculation for various values of L, while the dashed lines are meant to guide the eye. The four bulk pressures represented are Pbulk ) 150 atm (circles), Pbulk ) 200 atm (squares), Pbulk ) 250 atm (triangles), and Pbulk ) 300 atm (diamonds). The solid line is the corresponding curve for interactions in a vacuum, and the dot-dashed line corresponds to interaction in a vacuum if the F atoms had the same charge as the H atoms of n-dodecanethiol. For clarity, only lower error bars for Pbulk ) 150 atm and upper error bars for Pbulk ) 300 atm are shown.

of previous models developed for polymer brushes in describing systems of densely packed short chain ligands such as those considered here. We will then examine the results of a simple interpretive model26 to describe the behavior of the two systems. Using this model, we will infer the molecular level causes of the effectiveness of the fluorocarbon ligands in ScCO2. 3.1. Consistency with Experimental Observations. We have included in Figure 10 the W vs Pbulk plot for the dodecanethiol-passivated pore (Leq ) Lvac ) 43 Å) using data from our previous paper.26 It is quite apparent that the systems are stabilized much better (more repulsive at smaller L) by the

Dispersing Nanoparticles in ScCO2

Figure 7. Potential of mean force of the slit pore for perfluorodecanethiol-passivated slit pores solvated by CO2 at various pressures eff calculated using eq 2. Pbulk is calculated using Ltrunc ) 50 Å. The free energy corresponding to a single periodic image is attributed to a pair of particles to calculate the free energy per mole. The symbols are the result of said calculation for various values of L, while the dashed lines are meant to guide the eye. The four bulk pressures represented are Pbulk ) 150 atm (circles), Pbulk ) 200 atm (squares), Pbulk ) 250 atm (triangles), and Pbulk ) 300 atm (diamonds). The solid line is the corresponding curve for interactions in a vacuum, and the dot-dashed line corresponds to interaction in a vacuum if the F atoms had the same charge as the H atoms of n-dodecanethiol. For clarity, only lower error bars for Pbulk ) 150 atm and only upper error bars for Pbulk ) 300 atm are shown.

Figure 8. Density profile for a slit pore of gap width L ) 100 Å passivated by perfluorodecanethiol ligands and solvated by CO2 at a bulk pressure of Pbulk ) 169 atm. The density profile of the C atoms of the ligand (Fligand) is shown by the dashed line, while the solvent density profile Fp is the solid line. The density profile is determined using the entire cross section of the periodic image of the slit pore.

perfluorocarbon ligands than by the n-dodecanethiol ligands, consistent with experimental observations. The experiments on redispersion of perfluorodecanethiol-passivated particles showed that, at 353 K (80 °C), the pressure for redispersion (flocculation pressure) of 20-40 Å diameter nanoparticles is 272 atm. Our simulations are done at 308 K and seek to approximate a 60 Å diameter nanoparticle to better compare with our earlier results. To roughly estimate the anticipated flocculation pressure at 308 K, one route is to use the equality of the Flory-Huggins interaction parameter χ ) (Vsolvent/RT)(δsolvent - δsolute)2 (which should be less than 0.5 for good solvent conditions). Here, νsolvent is the molar volume, R is the universal gas constant, T is the temperature, and δsolvent and δsolute are the Hildebrand solubility parameters of the solvent (ScCO2) and solute (ligand-passivated nanoparticle), respectively. Using a standard database,52 we can calculate53 at T ) 353 K and Pbulk ) 272 atm νsolvent(353 K) ) 60.7 cm3/mol and δsolvent(353 K) ) 11.2 (J/cm3)0.5. Using

J. Phys. Chem. C, Vol. 114, No. 37, 2010 15557

Figure 9. Density profile for a slit pore of gap width L ) 100 Å passivated by perfluorooctanethiol ligands and solvated by CO2 at a bulk pressure of Pbulk ) 155 atm. The density profile of the C atoms of the ligand (Fligand) is shown by the dashed line, while the solvent density profile Fp is the solid line. The density profile is determined using the entire cross section of the periodic image of the slit pore.

Figure 10. Plot of W vs Pbulk for CO2 solvated slit pores passivated by n-dodecanethiol (circles and Leq ) 43 Å, Leq-vac ) 43 Å),26 perfluorodecanethiol (diamonds and Leq ) 43 Å, Leq-vac ) 40 Å), perfluorooctanethiol (triangles and Leq ) 35 Å, Leq-vac ) 33 Å). The interactions for a single periodic image are attributed to a single particle pair. The thin solid horizontal line corresponds to -1.5RT, so that W values greater than this imply a thermodynamically stable dispersion. The dotted lines are guides to the eye.

dodecane at 500 atm as the solute, we get δsolute(353 K) ) 15.8(J/ cm3)0.5, so that χ(353 K) ) 0.432. Now, since the packing density of the ligands is independent of temperature here at the given temperature range, we calculate for δsolute the value for dodecane at 80 atm at 308 K (same density as dodecane at 500 atm and 353 K), i.e., δsolute(308 K) ) 16.0(J/cm3)0.5. Using this value to get χ(308 K) ) 0.432, we find the solvent pressure of 105 atm at 308 K. Alternatively, the value obtained by simply equating solvent densities at the two conditions is nearly the same at 100 atm. In our simulations, we take the flocculation pressure to be that at which W < -1.5kBT, which is a reasonable criterion.14 From the W vs Pbulk plots (Figure 10) for a 60 Å diameter nanoparticle passivated by perfluorodecanethiol, we find the flocculation pressure is between 100 and 200 atm. Since the experimental nanoparticles are smaller in diameter (20-40 Å) than the one our simulation (60 Å) represents, it is expected that the experimental flocculation pressure would be lower.54 Hence, the results are consistent with the experimental observations to the extent that they can be directly compared. Since,

15558

J. Phys. Chem. C, Vol. 114, No. 37, 2010

Dalvi et al.

from our simulations, the two fluorinated ligands appear to perform equally well, we may infer that the inability of perfluorooctanethiol-passivated nanoparticles to redisperse after precipitation27 likely has to do with phenomena in the flocculated state, which our method is not designed to address. 3.2. Solvent Does Not “Solvate” Ligands. As we have noted in the previous paper,26 the ligands are too short and pack too densely to use models developed for polymer brushes14,54-56 or sparsely packed ligands6 to apply. It is clear from the density profiles (Figures 8 and 9) that the ligand region is not penetrated by the solvent. The passivated surface in this case is much harder and smoother than it would be in the case of a polymer brush, and the solvent essentially interacts only with the ligand tips. Further, and more importantly, the onset of interfacial repulsion occurs before ligands can overlap. Repulsion is, to use standard terminology,54 osmotic and not elastic. 3.3. Contribution to W. The linearity of Fpore vs 1/L plots in Figures 2 and 3 implies that the density profile in the slit pore can be approximated by a step function whose value is Fbulk at z > lo and z < L - lo and zero elsewhere. Here, lo is a measure of the thickness of the passivating layer. Since the solvent density profile is fairly uniform across the pore and can be adequately approximated by a step function, it makes sense to try to use a simple model by which to better understand the contributions to W for the two systems. Such a model is developed in the previous paper26 based on refs 56 and 57. We consider W to be composed of three types of contributions: Wvac (due to direct interaction between ligands on opposite faces), Wself (due to solvent cohesive interaction with its own bulk), and Wwall (contribution due to solvent interaction with the walls). Of these, we already know Wvac from our simulations in a vacuum. Wwall is the difference in wall-fluid interaction free energy when the gap width is very large and when the gap width is some finite value of L and we need a route to estimate it. If ufw(z′) is the interaction free energy of a single solvent molecule at a distance z′ from a ligand-passivated wall, then26

Figure 11. Wall-solvent interaction potential calculated using density profiles for wide slit pores (L ) 100 Å). The solid line is for perfluorodecanethiol, the dashed line is for perfluorooctanethiol, the dot-dashed line is for perfluorodecanethiol with charges on F equal to those on H of dodecanethiol, while the dotted line is the same for perfluorooctanethiol.

(6)

Figure 12. Plots illustrating the effect of electrostatics. The solid line is the difference between Wvac + Wwall of the H-charged and F-charged perfluorooctanethiol ligand (Leq ) 35 Å, Leq-vac ) 33 Å); F-charged value minus H-charged value. The dashed line is the same quantity for perfluorodecanethiol ligands (Leq ) 43 Å, Leq-vac ) 40 Å).

The procedure to obtain a characteristic ufw from a density functional theory and the solvent density profile at wide-gap pores is described in detail in the previous paper.26 The plots of ufw for the two fluorocarbon ligands are displayed in Figure 11. 3.4. Effect of Electrostatics. To illustrate the specific effect of electrostatic interactions, we have run wide-gap simulations (L ) 100 Å) for perfluorodecanethiol and perfluorooctanethiol slit pores but with the charges on C-F atoms being replaced with the charges on C-H atoms of dodecanethiol, while keeping all other things unchanged. We will call these “H-charged” ligands (with a charge of 0.06e on each “F”) as opposed to the fully charged or “F-charged” ligands (with a charge of -0.17 on each F of CF3 and -0.12 on each F of CF2). From the density profile, we have obtained the ufw value for each case (Figure 11). The well depth of the wall-solvent interaction for H-charged and F-charged perfluorodecanethiol ligands is approximately the same, but that for the F-charged ligand has the greater persistence in its range. For the perfluorooctanethiol ligands, the well depth is also deeper for the F-charged ligands than for the H-charged ligands. The effect of charge on ufw is different for the two types of ligands most likely because of

differences in their structure at the ligand-solvent interface. It appears that the CO2 molecule attractive dispersion interactions in the hollows formed between neighboring ligand tips at the ligand-solvent interface are enhanced for the longer chains. This fact emphasizes that the electrostatic and dispersion interactions are not simply separable if there is any change in structure associated with the change in interaction. Finally, one could compare the interaction of CO2 with the H-charged fluorocarbon with that of dodecanethiol ligands studied in ref 26. The hydrocarbon result is given in Figure 12 of that reference. The minimum occurs as expected at somewhat longer distance for the longer hydrocarbon, but most notably, the well depth is about 10% deeper for the hydrocarbon, consistent with the higher packing density of van der Waals attractive centers for the hydrocarbon, already pointed out. The effect of electrostatics is demonstrated in Figure 12 where the difference in the Wvac + Wwall values between the H-charged and F-charged ligands is plotted (H-charged value is subtracted from the F-charged value). It is clear that, without the electrostatic interactions with quadrupolar CO2, the fluorocarbon ligands would not have been as effective; they would have attracted each other more strongly and attracted CO2 more

Wwall(L) ) AporeFbulk

∫l

L-ll

o

[ufw(z′) + ufw(L - z′)] dz′ 2AporeFbulk

∫l ∞ ufw(z′) dz′ o

Dispersing Nanoparticles in ScCO2

Figure 13. Space filling models and cartoons for fluorocarbon ligand (perfluorooctanethiol). It is readily apparent that the fluorocarbons cannot orient in such a way that the dipoles align favorably. Also shown are schematics of surfaces on which dipoles are adverse (bottom lefthand side) and where the dipoles have reoriented to relieve repulsion (bottom right-hand side).

weakly. The subtle primary advantage of the electrostatic interactions is (as seen from Figure 11) that they extend the range of attractive interaction of the wall with the solvent, making it that much more attractive for the solvent in the pore. 3.5. Reasons for Effectiveness of Fluorocarbons in CO2. From our analysis, we can point out three reasons for the greater effectiveness of the fluorocarbon ligands over the hydrocarbon ligands: (1) Higher C-F bond dipole moment compared to C-H bond dipole moment allows fluorocarbon ligands to interact more strongly with the quadrupole of CO2, making them more CO2-philic. Further, the range of interaction with solvent also increases, allowing osmotic repulsion to become significant at wider gap widths (or particle separations). (2) The unique geometry of fluoroalkanes (see Figure 13) where the positive charges are buried in a “shell” of negative charges which ensures that the fluorocarbons always present an adVerse dipole to themselves. This electrostatic repulsion results in a weaker net ligand-ligand attractive interaction. Figure 13 schematically illustrates that there is a potential disadvantage of having a ligand with a strongly dipolar group which may not always be adversely oriented to groups from the other wall. The ability to reorient adverse dipoles is a possible reason why polyethyleneglycol based ligands58 are not effective capping ligands in ScCO2. (3) An additional subtle advantage of the fluorine group is that the packing density of the perfluorocarbons (∼20 Å2/ligand) is significantly lower than that of the hydrocarbons (∼16 Å2/ligand27) owing to the fact that fluorocarbons, with a molecular cross section of 28.3 Å2,48 are fatter than the corresponding hydrocarbon whose molecular cross section is 18.9 Å2.43 Hence, the dispersion interactions of a fluorocarbon-ligand-passivated surface with another such surface are weaker than those with n-dodecanethiol ligands. It is well-known that the best stabilizations of colloidal systems are achieved when the colloids have a high affinity for the solvent and a low affinity for themselves.59-61 Hence, the fluorocarbons behave like “ideal” ligands in these respects. The mix of geometry, electrostatics, and dispersion interactions for fluoroalkanes all favor these over hydrocarbons for

J. Phys. Chem. C, Vol. 114, No. 37, 2010 15559 CO2 solvent. The CO2-philicity can be explained entirely on the basis of electrostatic and dispersion interactions, precluding the need to postulate a “special” interaction between CO2 and fluorocarbons. The significance of electrostatic interactions to the enhanced CO2-philicity of the fluorocarbons is the subject of some controversy with some groups finding no significant electrostatic interactions between CO2 and fluorocarbons62,63 and attributing the enhanced solubility of CO2 in fluorocarbons to the relative ease with which cavities form in fluorocarbon fluids due to lowered cohesive energy density (which, as we have seen, is itself significantly due to electrostatics). Other groups claim a significantly stronger interaction between CO2 and fluorocarbons which is attributed to electrostatics64 though “site-specific” interactions have been invoked.65 The controversy can be somewhat resolved by noting that electrostatics (a) play a rather unimportant role in the properties of bulk fluorocarbon fluids66 and (b) have their effect masked, at short range, by the dispersion interactions. Hence, calculations/measurements in bulk fluids or between pairs of molecules need not reveal the effect which electrostatics will contribute in a heterogeneous environment. This is especially true if the heterogeneity is extended; the effect due to the addition of the relatively slowly decaying electrostatic interactions can play a much more significant role in that case. 3.6. Explanation of Experimental Observations on Alternative Ligands. The various attributes of perfluorocarbons mentioned above also appear to account for the success of fluoroacrylate-based polymeric surfactants in stabilizing dispersions of organic liquids in ScCO2.67 The nonfluorous CO2-philic functional groups that have been proposed in the literature incorporate either the oxygenated functional groups such as ethers,58 carbonates,3 and sugars4 or stubby hydrocarbons incorporating tertiary alkanes.5,58 Of these, only the last have seen some limited success as ligands for nanoparticle emulsions in ScCO2, with some recent notable advances for stubby hydrocarbons which also incorporate oxygenated tails.68 The failure of the ethers to be effective nanoparticle ligands58 is probably due to their flexibility and hence ability to reorient dipoles, resulting in stronger direct attractive interactions between particles (see Figure 13). The various ligands have been selected/characterized on the basis of cloud-point measurements, i.e., their behavior as free ligands in bulk CO2. However, the behavior of ligands restricted to a surface is quite different from their behavior in bulk liquid, and this point must be noted when screening ligands. 4. Conclusion In this paper, we have used a molecular simulation protocol to examine the reasons behind the remarkable effectiveness of perfluoroalkanethiols, compared to alkanethiols, as capping ligands for nanoparticles in supercritical carbon dioxide. We have determined that the principal reason for this enhanced effectiveness is that the C-F bond is much more polar than the C-H bond, and, hence, the fluorocarbons can interact with the quadrupole of carbon dioxide in addition to interacting with it by dispersion forces. The effect of the electrostatic interaction offsets the fact that the dispersion interaction is weakened by the relatively sparse packing of fluoroalkanethiols. The molecular geometry of fluorocarbons ensures that the C-F dipoles of two fluoroalkane molecules are always adVerse, making two fluorocarbon-passivated surfaces much less attractive to each other than the corresponding hydrocarbon-passivated surfaces. Hence, fluoroalkanethiol-passivated surfaces are both more CO2philic and less autophilic than alkanethiol-passivated surfaces, which is the reason for the effectiveness of fluorocarbons in supercritical carbon dioxide.

15560

J. Phys. Chem. C, Vol. 114, No. 37, 2010

Acknowledgment. This work was supported by funding to the NSF STC on Environmentally Responsible Solvents and Processes (CHE-9876674). The authors are grateful to Brian Korgel, Keith Johnston, and Aaron Saunders for stimulating discussions and to Sangik Cho for help with calculations. We are grateful for support from the Texas Advanced Computing Center and a major research instrumentation grant from the NSF (MRI-0619838). Additional support to P.J.R. was provided by the R. A. Welch Foundation (F-0019). Supporting Information Available: A comprehensive summary of the potential parameters used in this work. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) Murray, C. B.; Norris, D. J.; Bawendi, M. G. Synthesis and Characterization of Nearly Monodisperse CdE (E)S, Se, Te) Semiconductor Nanocrystallites. J. Am. Chem. Soc. 1993, 115, 8706–8715. (2) Shah, P. S.; Hanrath, T.; Johnston, K. P.; Korgel, B. A. Nanocrystal and Nanowire Synthesis and Dispersability in Supercritical Fluids. J. Phys. Chem. B 2004, 108, 9574–9587. (3) Sarbu, T.; Styranec, T.; Beckman, E. J. Non-fluorous polymers with high solubility in supercritical CO2 down to low pressures. Nature 2000, 165–168. (4) Potluri, V. K.; Xu, J.; Enick, R.; Beckman, E.; Hamiltion, A. D. Peracylated Sugar Derivatives Show High Solubility in Liquid and Supercritical Carbon Dioxide. Org. Lett. 2002, 2333–2335. (5) Anand, M.; Bell, P. W.; Fan, X.; Enick, R. M.; Roberts, C. B. Synthesis and Steric Stabilization of Silver Nanoparticles in Neat Carbon Dioxide Solvent Using Fluorine-Free Compounds. J. Phys. Chem. B 2006, 110, 14693–14701. (6) Patel, N.; Egorov, S. A. Interactions between sterically stabilized nanoparticles in supercritical fluids: A simulation study. J. Chem. Phys. 2007, 126, 054706. (7) Saunders, A. E.; Shah, P. S.; Park, E. J.; Lim, K. T.; Johnston, K. P.; Korgel, B. A. Solvent Density Dependent Steric Stabilizaton of Perfluoropolyether-Coated Nanocrystals in Supercritical Carbon Dioxide. J. Phys. Chem. B 2004, 108, 15969–15975. (8) Shah, P. S.; Husain, S.; Johnston, K. P.; Korgel, B. A. Role of Steric Stabilization on the Arrested Growth of Silver Nanocrystals in Supercritical Carbon Dioxide. J. Phys. Chem. B 2002, 106, 12178–12185. (9) Resnati, G. Synthesis of Chiral and Bioactive Fluoroorganic Compounds. Tetrahedron 1993, 49 (42), 9385–9445. (10) Pelley, J. Canada Moves to Eliminate PFOS Stain Repellants. EnViron. Sci. Technol. 2004, 452A. (11) Brust, M.; Walker, M.; Bethell, D.; Schiffrin, D. J.; Whyman, R. Synthesis of Thiol-Derivatized Gold Nanoparticles in a Two phase LiquidLiquid System. J. Chem. Soc., Chem. Commun. 1994, 801–802. (12) Shah, P. S.; Holmes, J. D.; Doty, R. C.; Johnston, K. P.; Korgel, B. A. Steric Stabilization of Nanocrystals in Supercritical Carbon Dioxide Using Fluorinated Ligands. J. Am. Chem. Soc. 2000, 122, 4245–4246. (13) Clarke, N. Z.; Waters, C.; Johnson, K. A.; Satherley, J.; Schiffrin, D. J. Size-Dependent Solubility of Thiol-Derivatized Gold Nanoparticles in Supercritical Ethane. Langmuir 2001, 17, 6048–6050. (14) Shah, P. S.; Holmes, J. D.; Johnston, K. P.; Korgel, B. A. SizeSelective Dispersion of Dodecanethiol-Coated Nanocrystals in Liquid and Supercritical Ethane by Density Tuning. J. Phys. Chem. B 2002, 106, 2545– 2551. (15) Gupta, G.; Shah, P. S.; Zhang, X.; Saunders, A. E.; Korgel, B. A.; Johnston, K. P. Enhanced Infusion of Gold Nanocrystals into Mesoporous Silica with Supercritical Carbon Dioxide. Chem. Mater. 2005, 17, 6728– 6738. (16) Chang, S.-C.; Lee, M.-J.; Lin, H.-M. Nanoparticles formation for metallocene catalyzed cyclic olefin copolymer via a continuous supercritical anti-solvent process. J. Supercrit. Fluids 2007, 420–432. (17) Dixon, D. J.; Johnston, K. P.; Bodmeier, R. A. Polymeric Materials Formed by Precipitation with a Compressed Fluid Antisolvent. AIChE J. 2004, 127–139. (18) da Rocha, S. R. P.; Johnston, K. P.; Rossky, P. J. Surfactant Modified CO2-Water Interface: A Molecular View. J. Phys. Chem. B 2002, 106, 13250–13261. (19) Stone, M. T.; da Rocha, S. R. P.; Rossky, P. J.; Johnston, K. P. Molecular Differences between Hydrocarbon and Fluorocarbon Surfactants at the CO2/Water Interface. J. Phys. Chem. B 2003, 107, 10185–10192. (20) Rossen, W. R. Foams: Theory, Measurements and Applications; Prudhomme, R. K., Khan, S. A., Eds.; Dekker: New York, 1996. (21) Johnston, K. P.; Harrison, K. L.; Clarke, M. J.; Howdle, S. M.; Heitz, M. P.; Bright, F. V.; Carlier, C.; Randolph, T. W. Water in Carbon

Dalvi et al. Dioxide Microemulsions: A New Environment for Hydrophiles including Proteins. Science 1996, 271, 624. (22) Harrison, K.; Goveas, J.; Johnston, K. P.; O’Rear, E. A. Water in Carbon Dioxide Microemulsion with a Fluorocarbon-Hydrocarbon Hybrid Surfactant. Langmuir 1994, 10, 3536. (23) DeSimone, J. M. Practical Approaches to Green Solvents. Science 2002, 297, 799–803. (24) Sagisaka, M.; Koike, D.; Mashimo, Y.; Yoda, S.; Takebayashi, Y.; Furuya, T.; Yoshizawa, A.; Sakai, H.; Abe, M.; Otake, K. Water/ Supercritical Carbon Dioxide Microemulsions with Mixed Surfactant Systems. Langmuir 2008, 24, 10116–10122. (25) Johnston, K. P.; da Rocha, S. R. P. Colloids in Supercritical Fluids over the last 20 years and future directions. J. Supercrit. Fluids 2009, 47, 523–530. (26) Dalvi, V. H.; Srinivasan, V.; Rossky, P. J. Molecular Simulation Study of Effectiveness of Alkanethiol Ligands in Stabilizing Nanoparticle Emulsions in Supercritical Carbon Dioxide and Ethane. J. Phys. Chem. C, in press, 2010. (27) Shah, P. S.; Husain, S.; Johnston, K. P.; Korgel, B. A. Nanocrystal Arrested Precipitation in Supercritical Carbon Dioxide. J. Phys. Chem. B 2001, 105 (39), 9433–9440. (28) Smith, W. F.; Forester, T. R. DLPOLY 2.15; Daresbury Laboratories: Daresbury, Warrington, U.K., 2003. (29) Humphrey, W.; Dalke, A.; Schulten, K. VMD-Visual Molecular Dynamics. J. Mol. Graphics 1996, 33–38. (30) Jorgensen, W. L.; Maxwell, D. S.; Tirado-Rives, J. Development and Testing of the OPLS All-Atom Force Field on Conformational Energetics and Properties of Organic Liquids. J. Am. Chem. Soc. 1996, 118, 11225–11236. Watkins, E. K.; Jorgensen, W. L. Perfluoroalkanes: Conformational analysis and liquid-state properties from ab initio and Monte Carlo calculations. J. Phys. Chem. A 2001, 105, 4118–4125. (31) Harris, J. G.; Yung, K. H. Carbon Dioxide’s Liquid-Vapour Coexistence Curve and Critical Properties as predicted by a Simple Molecular Model. J. Phys. Chem. 1995, 99, 12021–12024. (32) Berry, R. S.; Rice, S. A.; Ross, J. Physical Chemistry; Oxford University Press: Oxford, U.K., 2000; Table 21.13. (33) Vidali, G.; Ihm, G.; Kim, H. Y.; Cole, M. W. Potentials of Physical Adsorption. Surf. Sci. Rep 1991, 12 (4), 133–181. (34) Gatica, S. M.; Li, H. I.; Trasca, R. A.; Cole, M. W.; Diehl, R. D. Xe adsorption on C60 monolayer on Ag(111). Phys. ReV. B 2008, 77, 045414. (35) Li, X.; Li, J.; Eleftheriou, M.; Zhou, R. Hydration and Dewetting near Fluorinated Superhydrophobic Plates. J. Am. Chem. Soc. 2006, 128, 12439–12447. (36) Peguin, R. P. S.; da Rocha, S. R. P. Solvent-Solute Interactions in Hydrofluoroalkane Propellants. J. Phys. Chem. B 2008, 112, 8084–8094. (37) Raveendran, P.; Wallen, S. L. Exploring CO2-Philicity: Effects of Stepwise Fluorination. J. Phys. Chem. B 2003, 107, 1473–1477. (38) Jang, S. S.; Blanco, M.; Goddard, W. A., III; Cladwell, G.; Ross, R. B. The Source of Helicity in Perfluorinated N-Alkanes. Macromolecules 2003, 36, 5331–5341. (39) Potter, S. C.; Tildesley, D. J.; Burgess, A. N.; Rogers, S. C. A transferable potential model for the liquid-vapour equilibria of fluoromethanes. Mol. Phys. 1997, 92 (5), 825–833. (40) Gough, C. A.; Pearlman, D. A.; Kollman, P. Calculations of the relative free energy of aqueous solvation of several fluorocarbons: A test of the bond potential of mean force correction. J. Chem. Phys. 1993, 99 (11), 9103–9110. (41) Schmidt, M. W.; Baldridge, K. K.; Boatz, J. A.; Elbert, S. T.; Gordon, M. S.; Jensen, J. H.; Koseki, S.; Matsunaga, N.; Nguyen, K. A.; Su, S.; Windus, T. L.; Dupuis, M.; Montgomery, J. A. General Atomic and Molecular Electronic Structure System. J. Comput. Chem. 1993, 14, 1347– 1363. (42) Dalvi, V. H.; Rossky, P. J. Molecular Origins of Fluorocarbon Hydrophobicity. Proc. Natl. Acad. Sci. U.S.A. 2010, 107, 13603-13607. (43) Fenter, P.; Eisenberger, P.; Li, J.; Camillone III, N.; Bernasek, S.; Scoles, G. Surface of n-octadecanethiol self-assembled on Ag(111) surface: An Incommensurate Monolayer. Langmuir 1991, 7, 2013–2016. (44) Rovida, G.; Pratesi, F. Sulfur overlayers on the low-index faces of silver. Surf. Sci. 1981, 104 (2-3), 609–624. (45) Sellers, H.; Ulman, A.; Shnidman, Y.; Eilers, J. E. Structure and Binding of Alkanethiolates on Gold and Silver Surfaces: Implications for Self-Assembled Monolayers. J. Am. Chem. Soc. 1993, 115, 9389–9401. (46) Korgel, B. A.; Fullam, S.; Connolly, S.; Fitzmaurice, D. Assembly and Self-Organization of Silver Nanocrystal Superlattices: Ordered “Soft Spheres”. J. Phys. Chem. B 1998, 102, 8379–8388. (47) Broniatowski, M.; Dynarowicz-Latka, P. Iso-branched semifluorinated alkanes in Langmuir monolayers. J. Colloid Interface Sci. 2006, 299, 916–923. (48) Jadzinsky, P. D.; Calero, G.; Ackerson, C. J.; Bushnell, D. A.; Kornberg, R. D. Structure of a Thiol Monolayer-Protected Gold Nanoparticle at 1.1A Resolution. Science 2007, 318, 430–433.

Dispersing Nanoparticles in ScCO2 (49) Frenkel, D.; Smit, B. Understanding Molecular Simulation; Academic Press: London, 2002. (50) Schoen, M.; Diestler, D. J. Analytical Treatment of Simple Fluid Adsorbed in Slit Pore. J. Chem. Phys. 1998, 5596–5606. (51) Widom, B. Some topics in the theory of fluids. J. Chem. Phys. 1963, 39, 2808–2812. (52) http://webbook.nist.gov/chemistry/fluid/ (accessed 2009). (53) Allada, S. R. Solubility parameters of supercritical fluids. Ind. Eng. Chem. Process. Des. DeV. 1984, 23, 344–348. (54) Fernandez-Nieves, A.; Fernandez-Barbero, A.; Vincent, B.; de las Nieves, F. J. Reversible Aggregation of Soft Particles. Langmuir 2001, 17 (6), 1841–1846. (55) Meredith, J. C.; Sanchez, I. C.; Johnston, K. P.; de Pablo, J. J. Simulation of Structure and Interaction Forces for Surfaces Coated with Grafted Chains in a Compressible Solvent. J. Chem. Phys. 1998, 6424– 6434. (56) Evans, R.; Marconi, U. M. B.; Tarazona, P. Fluids in narrow pores: Adsorption, capillary condensation, critical points. J. Chem. Phys. 1986, 84 (4), 2376–2399. (57) Tarazona, P.; Marconi, U. M. B.; Evans, R. Phase Equilibria of Fluid Interfaces and Confined Fluids. Mol. Phys. 1987, 60 (3), 573–595. (58) Fan, X.; McLeod, M. C.; Enick, R. M.; Roberts, C. B. Preparation of Silver Nanoparticles via Reduction of a Highly CO2 Soluble Hydrocarbon Based Metal Precursor. Ind. Eng. Chem. Res. 2006, 45, 3343–3347. (59) Israelachvilli, J. Intermolecular and Surface Forces, 2nd ed.; Academic Press: London, 1992. (60) Louis, A. A.; Allahyarov, E.; Lowen, H.; Roth, R. Effective force in colloidal mixtures: From depletion attraction to accumulation repulsion. Phys. ReV. E 2002, 61407–61411.

J. Phys. Chem. C, Vol. 114, No. 37, 2010 15561 (61) Rabani, E.; Egorov, S. A. Integral Equation Theory for the Interactions between Passivated Nanocrystals in Supercritical Fluids: Solvophobic and Solvophilic Cases. J. Phys. Chem. B 2002, 106, 6771– 6778. (62) Costa Gomes, M. F.; Padua, A. A. H. Interactions of CO2 with Liquid Fluorocarbons. J. Phys. Chem. B 2003, 107, 14020–14024. (63) Yee, G. G.; Fulton, J. L.; Smith, R. D. Fourier Transform Infrared Spectroscopy of Molecular Interactions of Heptafluoro-1-butanol or 1-Butanol in Supercritical Carbondioxide and Supercritical Ethane. J. Phys. Chem. 1992, 96 (15), 6172–6181. (64) Cece, A.; Jureller, S. H.; Kerschner, J. L.; Moschner, K. F. Molecular Modelling Approach for Contrasting the Interaction of Ethane and Hexafluoroethane with Carbon Dioxide. J. Phys. Chem. 1996, 100, 7435–7439. (65) Dardin, A.; DeSimone, J. M.; Samulski, E. T. Fluorocarbons dissolved in Supercritical Carbon Dioxide. NMR Evidence for Specific Solute-Solvent Interactions. J. Phys. Chem. B 1998, 102, 1775–1780. (66) Song, W.; Rossky, P. J.; Maroncelli, M. Modelling alkane+pefluoroalkane interactions using all-atom potentials: Failure of usual combining rules. J. Chem. Phys. 2003, 119 (17), 9145–9162. (67) DeSimone, J.; Wells, S. L. CO2 Technology Platform: An important tool for environmental problem solving. Angew. Chem., Int. Ed. 2001, 40, 518–527. (68) Eastoe, J.; Gold, S.; Steytler, D. C. Hydrocarbon Surfactants for CO2: An Impossible Dream? Aust. J. Chem. 2007, 60, 9, 630–632.

JP102047E