8166
Ind. Eng. Chem. Res. 2007, 46, 8166-8175
Experimental Measurement of the Solubility and Diffusivity of CO2 in Room-Temperature Ionic Liquids Using a Transient Thin-Liquid-Film Method Ying Hou and Ruth E. Baltus* Department of Chemical and Biomolecular Engineering, Clarkson UniVersity, Potsdam, New York 13699-5705
In this paper, results from an experimental investigation of carbon dioxide (CO2) solubility and diffusivity in the ionic liquids 1-n-butyl-3-methylimidazolium bis(trifluoromethylsulfonyl)imide ([bmim][Tf2N]), 1,2dimethyl-3-propylimidazolium bis(trifluoromethylsulfonyl)imide ([pmmim][Tf2N]), 1-butyl-3-methylpyridinium bis(trifluoromethyl sulfonyl)imide ([bmpy][Tf2N]), 1-(3,4,5,6-perfluorohexyl)-3-methylimdazolium bis(trifluoromethyl sulfonyl)imide ([perfluoro-hmim][Tf2N]), and 1-n-butyl-3-methylimidazolium tetrafluoroborate ([bmim][BF4]) are reported. A transient thin liquid film method was developed, which enables one to determine the Henry’s law constant and the diffusivity at low pressure simultaneously. Measurements were performed at temperatures in the range of 283-323 K. Henry’s law constants were in the range of 25.5-84 bar and were in general agreement with measurements reported by other researchers for these and similar ionic liquids. The entropies and enthalpies of absorption were determined to increase as gas solubility decreased. The measured CO2 diffusion coefficients in the five ionic liquids were ∼10-6 cm2/s, which is an order of magnitude smaller than the coefficients for CO2 diffusion in traditional organic solvents. In contrast to the gas solubility results, measured diffusion coefficients were determined to be dependent on the ionic liquid cation as well as the anion. In addition, CO2 diffusion coefficients were considerably more sensitive to temperature than were CO2 solubilities in these ionic liquids. Results were used to develop a correlation relating CO2 diffusion to ionic liquid properties and system temperature. 1. Introduction Room-temperature ionic liquids (RTILs) are organic salts that melt near room temperature. RTILs generally have bulky lowsymmetry organic cations with a delocalized charge, which leads to their low melting point.1-4 The ionic liquid anions are more symmetrical and are generally smaller than the cations. Common anions include hexafluorophosphate (PF6-), tetrafluoroborate (BF4-), and (bis(trifluoromethylsulfonyl)imide) (Tf2N-). RTILs have negligible vapor pressure;5-7 therefore, they are considered to be “green” solvents that may potentially replace conventional volatile organic compounds (VOCs) in many reaction and separation processes.8 The properties of RTILs, such as melting point, viscosity, density, and gas and liquid solubility, can be adjusted via the appropriate choice of cation and anion.9,10 RTILs are also nonflammable, have high thermal stability, and have a wide liquid range.11,12 To design ionic-liquid-based processes, knowledge of the solubilities and diffusivities of gases in room-temperature ionic liquids is needed. However, only limited information about these properties is available in the literature. Carbon dioxide (CO2) has been used in many previous studies of solubility in ionic liquids, because it is of interest for a variety of ionic liquid applications. Therefore, it was a convenient solute choice for this study. The objective of this work was to develop an experimental approach to measure both solubility and diffusivity of CO2 in RTILs, as well as examine relationships between RTIL structure and these properties. In this work, the solubility and diffusivity of CO2 in five different ionic liquids were determined. An experimental method based on the one-dimensional diffusion of solute gas into a thin ionic liquid film was developed. A major advantage of this * To whom correspondence should be addressed. Tel.: 315-2682368. Fax: 315-268-6654. E-mail address:
[email protected].
method over other techniques is that both the solubility and diffusivity of the target gas in a RTIL can be determined in a single experiment that requires only a small sample of ionic liquid (∼250 µL). The effect of temperature on solubility and diffusivity was examined by conducting measurements at different temperatures, ranging from 283 K to 323 K. Because experiments were conducted at low pressure (∼1-2 bar), measured diffusion coefficients can be considered to be infinite dilution values. An expression relating CO2 diffusivity to ionic liquid properties was developed. Ionic liquid density and viscosity were also measured for the ionic liquids where literature values were not available. 2. CO2 Uptake Model The experimental method involves tracking the decrease in pressure that results following the introduction of CO2 into a small closed chamber that contains a thin film of ionic liquid. The solubility and diffusivity were determined by fitting the pressure decay to a one-dimensional diffusion model for solute uptake into the liquid. Before each experiment, vacuum was applied to the chamber. At time t ) 0, CO2 was introduced at pressure P0 and the system was sealed. The decay in pressure above the liquid film resulting from CO2 absorption into the IL was monitored as a function of time. The system geometry is defined such that z ) 0 corresponds to the base of the liquid film and z ) L corresponds to the gas/ liquid interface. The assumptions of the model are as follows: (i) one-dimensional diffusion of CO2 in the z (vertical)-direction only; (ii) no convective transport in the system; (iii) the ionic liquid has negligible vapor pressure; (iv) equilibrium is established at the gas/liquid interface, with the CO2 concentration in the liquid phase described using Henry’s Law; (v) the thickness of the ionic liquid film and the liquid viscosity are constant and spatially uniform; and (vi) the CO2 diffusion coefficient is
10.1021/ie070501u CCC: $37.00 © 2007 American Chemical Society Published on Web 10/18/2007
Ind. Eng. Chem. Res., Vol. 46, No. 24, 2007 8167
independent of CO2 concentration. Combining Fick’s first law of diffusion with a mole balance written on the liquid-phase yields
∂CCO2 ∂t
) -DCO2
∂2CCO2
(1)
∂z2
where CCO2 is the molar concentration of CO2 in the ionic liquid. The initial and boundary conditions are given as follows. Initial conditions:
t ) 0, P ) P0, CCO2 ) 0
(2a)
Boundary conditions:
∂CCO2
t > 0, z ) 0, z ) L, CCO2 )
∂z
)0
(2b)
FIL P ) K‚P HCO2MWIL
(2c)
where HCO2 is the Henry’s Law constant for CO2 in the ionic liquid, FIL the density, and MWIL the molecular weight of the ionic liquid. The solution of eq 1, subject to the conditions in eq 2, is
CCO2 CCO2|z)L
)1-
() 4
π
∞
(-1)n
cos ∑ n)0 (2n + 1)
[
(
)
(2n + 1)πz 2L
×
]
(2n + 1)2π2DCO2t
exp -
4L2
(3) 3. Experimental Methods
A mole balance written on the gas phase yields
( )(
| )
∂CCO2 dP RT dn RT VIL -DCO2 ) ) dt V dt V L ∂z
(4)
z)L
where V is the volume of gas and VIL is the volume of the ionic liquid in the chamber. Substituting the derivative (∂CCO2/∂z)z)L from eq 3 into eq 4 yields
dP
)
dt
( ) () [
RT VIL
-
V
L
DCO2‚KP‚
2
L
]
∞
(2n + 1)2π2DCO2t
n)0
4L2
∑ exp -
(5)
Rearranging eq 5 and integrating from t ) 0 to t yields
ln
P
)
( )
P0
k
HCO2
∞
∑
n)0
1 (2n + 1)2
{[
exp -
where
k)
] }
(2n + 1)2π2DCO2t
8RTVILFIL π2VMWIL
4L2
-1
Using a nonlinear least-square method, the two unknown variables HCO2 and DCO2 were simultaneously fit to the experimental P(t) data. From an examination of results obtained using a different number of terms in eq 6, it was found that 55 terms in the summation in eq 6 were sufficient to determine HCO2 and DCO2 accurately.13 A comparison of the HCO2 and DCO2 values determined from data collected over different time intervals shows that results obtained using the first ∼8 h of data are essentially the same as results obtained using longer time intervals.13 A similar experimental approach was used by Camper et al.14 to measure the solubility and diffusivity of different gases, including CO2, in ethylmethylimidazolium bis(trifluoromethylsulfonyl)imide ([emim][Tf2N]). The experiments performed by Camper involved a larger volume of ionic liquid (∼6 mL). Pressure decay data were collected for the first 20 min of the experiment, during which time the liquid was not stirred. After sufficient data were collected to determine the solute diffusion coefficient, the ionic liquid was stirred to accelerate the approach to equilibrium. Gas solubility was determined from the difference between initial and equilibrium pressures in the system. A model for diffusion into a semi-infinite volume was used to determine the solute diffusion coefficient in the ionic liquid from the initial pressure data. In the model used by Camper to interpret the initial transients, it was assumed that the penetration depth of CO2 into the ionic liquid is short and that CO2 concentration in liquid far from the interface remains at zero. This differs from the model used in this study where CO2 penetrates into the entire liquid film. Later in this paper, we will discuss the advantages and disadvantages of these, as well as other experimental approaches, to determine the solubility and diffusivity in ionic liquids.
(6)
3.1. Materials. 3.1.1. Ionic Liquids. Five different ionic liquids were studied in this work: 1-n-butyl-3-methylimidazolium bis(trifluoromethylsulfonyl)imide ([bmim][Tf2N]), 1,2dimethyl-3-propylimidazolium bis(trifluoromethylsulfonyl)imide ([pmmim][Tf2N]), 1-butyl-3-methylpyridinium bis(trifluoromethylsulfonyl)imide ([bmpy][Tf2N]), 1-(3,4,5,6-perfluorohexyl)3-methylimdazolium bis(trifluoromethylsulfonyl)imide ([perfluorohmim][Tf2N]), and 1-n-butyl-3-methylimidazolium tetrafluoroborate ([bmim][BF4]). The ionic liquids [bmim][Tf2N], [pmmim][Tf2N], and [bmpy][Tf2N] were obtained from Covalent Associates (Woburn, MA) (electrochemical grade: >99.5% purity,
