Receptacle model of salting-in by tetramethylammonium ions

Abstract

Water is a poor solvent for nonpolar solutes. Water containing ions is an even poorer solvent. According to standard terminology, the tendency of salts to precipitate oils from water is called salting-out. However, interestingly, some salt ions, such as tetramethylammonium (TMA), cause instead the salting-in of hydrophobic solutes. Even more puzzling, there is a systematic dependence on solute size. TMA causes the salting-out of small hydrophobes and the salting-in of larger nonpolar solutes. We study these effects using NPT Monte Carlo simulations of the Mercedes-Benz (MB) + dipole model of water, which was previously shown to account for hydrophobic effects and ion solubilities in water. The present model gives a structural interpretation for the thermodynamics of salting-in. The TMA structure allows deep penetration by a first shell of waters, the dipoles of which interact electrostatically with the ion. This first water shell sets up a second water shell that is shaped to act as a receptacle that binds the nonpolar solute. In this way, a nonpolar solute can actually bind more tightly to the TMA ion than to another hydrophobe, leading to the increased solubility and salting-in. Such structuring may also explain why molecular ions do not follow the same charge density series as atomic ions do.

Figure 1

The MB-dipole model: the water-water and water-tetramethylammonium ion interaction.

Figure 2

The experimental Setchenov coefficients, k_S, in mole/kg for different salts as a function of a hydrophobe size (upper panel), and ΔΔG/k_BT as obtained with the MB+dipole TMA model (lower panel). The experimental data for hydrocarbon gases (From ref. 9) are shown by dashed line, and the data for 2-ketones (from Ref. 8, 39) are shown by solid lines. n denotes the number of carbon atoms in a molecule.

Figure 3

The experimental Setchenov coefficients, k_S, in mole/kg, for gaseous methane and ethane in water TMABr solution as a function of temperature (From ref. 9) (upper panel), and ΔΔ G/k_BT as obtained with the MB+dipole TMA model for two different hydrophobe size (lower panel).

Figure 4

The average (left panel) and orientation dependent (right panel) water-tetramethylammonium ion pair distribution function as obtained from MC simulation for MB+dipole.

Figure 5

The first two shells water-tetramethylammonium ion angle distribution function (upper panel), and a snapshot of first two shells (indicated by thin broken lines) waters (lower panel) around the tetramethylammonium ionas obtained from MC simulation for MB+dipole model.

Figure 6

The Monte Carlo simulation results for average (right panel) and orientation dependent (left panel) hydrophobe-tetramethylammonium ion pair distribution function for hydrophobe of size σ_LJ=1.5. Monte Carlo simulation results.

Figure 7

The orientation dependent pair distribution function for water-tetramethylammonium ion (left panel) and hydrophobe-tetramethylammonium ion (right panel) for hydrophobe size σ_LJ=1.5. Green circles denote the most probable location of water (left) and hydrophobe (right) molecules. Monte Carlo results.

Figure 8

The same as in Figure 6 but for the hydrophobe size σ_LJ=1.8

Figure 9

The orientation dependent water-methanol ion (left panel) and hydrophobe-methanol (right panel) pair distribution function for hydrophobe size σ_LJ=1.5. Monte Carlo results. The less intense color that appears on the left hand-sides of the the panels, compared to the right hand-sides, is due to the asymmetry of the alcohol molecule.