Supercritical Fluids – Fundamentals and Applications

Introduction

A fluid heated to above the critical temperature and compressed to above the critical pressure is known as a supercritical fluid. Frequently the term, compressed liquid, is used to indicate a supercritical fluid, a near-critical fluid, an expanded liquid or a highly compressed gas. The phenomena and behavior of supercritical fluids has been the subject of research right from 1800’s.

Two supercritical fluids are of particular interest, carbon dioxide and water. Carbon dioxide has a low critical temperature of 304 K and a moderate critical pressure of 73 bar. It is non-flammable, non-toxic and environmentally friendly. It is often used to replace toxic freons and certain organic solvents. Further, it is miscible with a variety of organic solvents and is readily recovered after processing. It is also a small and linear molecule and thus diffuses faster than conventional liquid solvents.

Water has a critical temperature of 647 K and a critical pressure of 220 bar due to its high polarity. The character of water at supercritical conditions changes from one that supports only ionic species at ambient conditions to one that dissolves paraffins, aromatics, gases and salts. Due to this unique property, research has been carried out on supercritical water for reaction and separation processes to treat toxic wastewater. Further, the dielectric constant of water changes from about 78 at room temperature and atmospheric pressure to roughly 6 at critical conditions, enabling control of reactions that depend on the dielectric constant of the medium.

Supercritical fluids such as water and carbon dioxide are substances that are compatible with the earth's environment. However, several other supercritical fluids can be used, but the final choice would depend on the specific application and additional factors such as safety, flammability, phase behavior and solubility at the operating conditions and the cost of the fluid. In the following sections, a brief outline of the properties, fundamentals and applications of supercritical fluids is provided. 

I. Properties and fundamentals of supercritical fluids

I.1. Solvent strength.

The density of a supercritical fluid is extremely sensitive to minor changes in temperature and pressure near the critical point. The density of the fluids are closer to that of organic liquids but the solubility of solids can be 3-10 orders of magnitude higher. The enhancement of solubilities was discovered in 1870’s for the potassium iodide-ethanol system. The solvent strength of a fluid can be expressed by the solubility parameter, d, which is the square root of the cohesive energy density and is defined rigorously from first principles. A plot of the solubility parameter for carbon dioxide versus pressure would resemble that a plot of density versus pressure. This confirms that the solvation strength of a supercritical fluid is directly related to the fluid density. Thus the solubility of a solid can be manipulated by making slight changes in temperatures and pressures.

Another attractive feature of supercritical fluids is that the properties lie between that of gases and liquids. A supercritical fluid has densities similar to that of liquids, while the viscosities and diffusivities are closer to that of gases. Thus, a supercritical fluid can diffuse faster in a solid matrix than a liquid, yet possess a solvent strength to extract the solute from the solid matrix.

I.2. Phase behavior.

The phase behavior of ternary systems of carbon dioxide and the solubilities of over 260 compounds in liquid carbon dioxide were studied in a monumental work published in 1954. Though this data is for liquid carbon dioxide, it provides a first approximation to solubilities in supercritical fluids. An understanding of the phase behavior is important since the phase behavior observed in supercritical fluids considerably differ from the behavior observed in liquids. One such behavior is the retrograde region. For an isobaric system, an increase in the temperature of a solution increases the solubility of the solute over certain ranges of pressure (consistent with the typical liquid systems) but decreases the solute solubility in other pressure ranges. This anomalous behavior wherein the solubility of the solute decreases with a temperature increase is called the retrograde behavior. Thus, the following generalizations may be made regarding the solute solubilities in supercritical fluids. Solute solubilities approach and may exceed that of liquid solvents. Solubilities generally increase with increase in pressure. An increase in the temperature of the supercritical fluid may increase, decrease or have no effect on the solubility of the solute depending upon the pressure.

Carbon dioxide is not a very good solvent for high molecular weight and polar compounds. To increase the solubility of such compounds in supercritical carbon dioxide, small amounts (ranging from 0 to 20 mol %) of polar or non-polar cosolvents called modifiers may be added. The cosolvent interacts strongly with the solute and significantly increases the solubility. For example, addition of a small amount (3.5 mol%) of methanol to carbon dioxide increases the solubility of cholesterol by an order of magnitude.

Compressed gases and fluids have the ability to dissolve in and expand organic liquid solvents at high pressures (50 to 100 bar). This expansion usually decreases the solvent strength of the liquid. Eventually the mixture solvent strength is comparable to that of the pure compressed fluid. Knowledge of when a solute would precipitate can be important and helps one to determine when heavy hydrocarbons would precipitate in an oil reservoir when carbon dioxide is injected.

The phase behavior of binary systems follows the typical six classes of binary diagrams. The Class I binary diagram is the simplest case. The pressure-temperature diagram consists of a vapor-pressure curve for each pure component, ending at the pure component critical point. The loci of critical points for the binary mixtures are continuous from the critical point of component one to the critical point of component two. More complicated behavior exists for other classes, including the presence of upper critical solution temperature (UCST) lines, two-phase immiscibility lines, and even three-phase immiscibility lines.

Modeling of phase behavior cannot be done using relatively simple thermodynamics because extreme non-idealities occur in the supercritical region. One of the simplest cases of phase behavior modeling is that of modeling the solubility of crystalline solids in supercritical fluids. Thermodynamic models are based on the principle that the fugacities of a component are equal for all phases at equilibrium under constant temperature and pressure. Associations resulting from hydrogen bonding or donor-acceptor interactions, can have a pronounced effects on supercritical fluid phase behavior. Understanding of hydrogen bonding among mixtures in supercritical fluids is important because of the increased interest in supercritical water solutions, and in polar cosolvents for supercritical fluid carbon dioxide. Various equations of state such as the statistical association fluid theory and the lattice fluid hydrogen bonding model are often used to describe these associations.

Experimental confirmation of phase behavior is often necessary to account for unaccounted and complex behavior that can not be modeled apriori. The most useful tool for examining phase behavior is the variable-volume view cell, whose contents can be viewed safely through a sapphire window by means of a mirror or a video camera. The apparatus contains a piston to separate the pressurizing fluid from the sample and allows for manipulation of temperature, pressure, or composition.

I.3. Extraction with supercritical fluids

Supercritical extraction has been applied to a large number of solid matrices. The desired product can be either the extract or the extracted solid itself. The advantage of using supercritical fluids in extraction is the ease of separation of the extracted solute from the supercritical fluid solvent by simple expansion. In addition, supercritical fluids have liquid like densities but superior mass transfer characteristics compared to liquid solvents due to their high diffusion and very low surface tension that enables easy penetration into the porous structure of the solid matrix to release the solute.

Extraction of soluble species (solutes) from solid matrices takes place through four different mechanisms. If there are no interactions between the solute and the solid phase, the process is simple dissolution of the solute in a suitable solvent that does not dissolve the solid matrix. If there are interactions between the solid and the solute, then the extraction process is termed as desorption and the adsorption isotherm of the solute on the solid in presence of the solvent determines the equilibrium. Most solids extraction processes, such as activated carbon regeneration, fall in this category. A third mechanism is swelling of the solid phase by the solvent accompanied by extraction of the entrapped solute through the first two mechanisms, such as extraction of pigments or residual solvents from polymeric matrices. The fourth mechanism is reactive extraction where the insoluble solute reacts with the solvent and the reaction products are soluble hence extractable, such as extraction of lignin from cellulose. Extraction is always followed by another separation process where the extracted solute is separated from the solvent.

Another important aspect in supercritical extraction relates to solvent/solute interactions. Normally the interactions between the solid and the solute determine the ease of extraction, i.e., the strength of the adsorption isotherm is determined by interactions between the adsorbent and the adsorbate. However, when supercritical fluids are used, interactions between the solvent and the solute affect the adsorption characteristics due to large negative partial molar volumes and partial molar enthalpies in supercritical fluids.

The thermodynamic parameters that govern the extraction are found to be temperature, pressure, the adsorption equilibrium constant and the solubility of the organic in supercritical fluid. Similar to the retrograde behavior of solubility in supercritical fluids, the adsorption equilibrium constants can either decrease or increase for an increase in temperature at isobaric conditions. This is primarily due to the large negative partial molar properties of the supercritical fluids. In addition to the above factors, the rate parameters like the external mass transfer resistances, the axial dispersion in the fluid phase, and the effective diffusion of the organics in the pores also play a crucial role in the desorption process. A thorough understanding of these governing parameters is important in the modeling of supercritical fluid extraction process and in the design, development and future scale-up of the process. 

I.4. Polymers and supercritical fluids

A number of studies have examined solubilities of polymers in supercritical fluids. With the exception of a few polymers, such as fluoropolymers, most high molecular weight polymers do not dissolve in carbon dioxide. However, polymers can uptake a significant amount of the supercritical fluid. As the concentration of the compressed fluid is increased in the polymer phase, the sorption and subsequent swelling of an amorphous polymer can cause a glass- to liquid-phase transition. The glass transition temperature of the polymer may be drastically reduced and this behavior may be exploited in polymer processing to produce extremely small voids only a few micrometers in diameter.

I.5. Biocatalysis and supercritical fluids

The advantages of using enzymes in supercritical fluids have been investigated. The densities of supercritical fluids are comparable to that of liquids, while the viscosities and diffusion coefficients are comparable to that of gases. This enhances the rates for diffusion controlled reactions. The low water activity environment will shift the thermodynamic equilibrium of hydrolytic reactions to favor synthesis. Also, reactions in which water is a product can be driven to completion. Gaseous reactants are completely miscible in supercritical fluids and thus reactions involving gaseous substrates will not be limited by solubility.

Since enzymes are insoluble in supercritical fluids, recovery is straightforward and immobilization is unnecessary. Further, product fractionation, purification etc. from the reaction mixture are feasible by reducing the pressure in a sequence of separators. Enzyme catalysis has been well characterized for physical properties of aqueous systems and detailed reaction mechanisms have been derived. However, knowledge of aqueous systems can not be extrapolated to supercritical fluid catalysis because the physical properties of the solvent such as dielectric constant, viscosity, diffusivity etc. are widely different.

Enzymatic reactions in non-aqueous media, especially supercritical fluids, are gaining acceptance. The initial paper on this subject appeared in 1985 in which a study of the hydrolysis of p-nitrophenyl phosphate to p-nitrophenol with alkaline phosphatase was reported. Subsequently, rates of oxidation of cholesterol with cholesterol oxidase were determined. It was also found that cosolvents facilitate cholesterol aggregation in supercritical carbon dioxide and the enzyme is known to be more active toward aggregates than cholesterol monomers. The most extensively studied enzymatic reactions in supercritical carbon dioxide systems have involved lipases. However, the use of supercritical fluids as the reaction media for enzymatic catalysis is still in its infancy.

I.6. Dispersions in supercritical fluids

The ability to design surfactants for the interface between water (or organics) and supercritical fluids offers new avenues in protein and polymer chemistry, separation science, reaction engineering, waste minimization and treatment. Surfactant design, which is reasonably well understood for conventional reverse micelles and water-in-oil microemulsions for alkane solvents, is more difficult for carbon dioxide because the properties of carbon dioxide are much different from those of water or nonpolar organic solvents. Carbon dioxide has no dipole moment and weaker van der Waals forces than hydrocarbon solvents. It is possible, however, to form dispersions of either hydrophilic or lipophilic phases in a carbon dioxide continuous phase. Organic-in-carbon dioxide dispersions may be stabilized using surfactants like fluorinated compounds, which are carbon dioxide-philic. 

II. Applications and commercial processes of supercritical fluids

Any commercialization of a process that uses supercritical fluids must involve a cost analysis that should indicate that the advantages in the new process offsets the penalty of high pressure operations. A variety of supercritical fluid processes have been commercialized. Details of a few such processes are given below. Many other processes have been investigated on a lab or pilot plant scale and have the potential to be scaled up in the near future.

II.1. Supercritical Fluid Chromatography.

Supercritical fluid chromatography is now often used as an analytical tool. The density is used as the controlling feature. Separations are based on a user programmed density profile with the supercritical fluid as the mobile phase. This analytical technique has been successfully used to separate oligomers and high molecular weight polymers. Supercritical fluids are used as the extracting solvents for the removal of polyaromatic hydrocarbons from soil. It is now a standard method for gas chromatography sample preparation because the extraction is considerably faster than Soxhlet extraction.

II. 2. Fractionation

Supercritical fluids can be used to fractionate low vapor pressure oils and polymers. This fractionation is difficult to achieve in distillation because the impurities have about the same volatility as the primary components reducing the overall selectivity. Kerr-McGee Inc. has developed a commercial process for the separation of heavy components of crude oil. Fractionation with respect to chemical composition is possible and has been investigated to produce polymer fractions of low polydispersity starting from a parent material of high polydispersity.

II.3. Reactions

 Supercritical fluids are attractive media for several chemical reactions. The properties of supercritical fluids mentioned earlier can be used to advantage. By small adjustments in pressure, the reaction rate constants can be altered by two orders of magnitude. Equilibrium constants for reversible reactions can also be changed 2-6 fold by small changes in pressure. This dramatic control over the reaction rates has led to the design of several reactions in different areas of biochemistry, polymer chemistry and environmental science. In bioreactions, increased solubilities of hydrophobic material and the potential to integrate the separation and reaction steps has led to research in this area. The use of lipase and synthesis of mondisperse biopolymers holds commercial promise. Carbon dioxide has also been extensively studied for homogeneous polymerization of a few polymers such as fluoroacrylates. The feasibility of free radical polymerization of polystyrene and the polymerization of polyethylene has also been investigated. Carbon dioxide is also often used as a swelling agent for a polymer substrate. Though highly corrosive and a high critical temperature and pressure, supercritical water has been one of the most studied medium for chemical reactions. Supercritical water has the ability to dissolve many nonpolar organic compounds such as alkanes and chlorinated biphenyls and can dissolve in several gases. It is thus an attractive media for oxidative reactions and has been used to treat a wide variety of waste water streams from chemical, petroleum, textile industries. Huntsman Corporation has commercialized a hydrothermal oxidation unit to treat alcohol and amine contaminated water.

II.4. Applications in the material and polymer industry

Supercritical fluids are used extensively in the material and polymer industry. Rapid expansion from supercritical solutions across an orifice or nozzle is used commercially to precipitate solids. In this technique, a solute dissolved in supercritical fluid is depressurized rapidly. By controlling the operating variables carefully, the desired precipitated morphology can be attained. In an another process, called gas anti-solvent, a supercritical fluid is rapidly added to a solution of a crystalline solid dissolved in an organic solvent. Since the solute has limited solubility in the fluid, the supercritical fluid acts as an anti-solvent to precipitate solid crystals. By varying the density of the fluid, the particle size distribution of final crystals can be finely controlled. Another process is the precipitation using a compressed fluid anti-solvent. In this process, the solution is sprayed through a nozzle into a compressed fluid and the solvent diffuses rapidly into the supercritical fluid while the fluid swells the solution to precipitate the solute. This process has been used commercially to form nanometric monodisperse microspheres of polymers. Another process that has been commercialized is the usage of supercritical fluid carbon dioxide to produce foamed parts. Since supercritical fluids depress the glass transition temperature of the polymer, polymer foams can be formed at room temperature by directly adding the supercritical fluid into the extruder. 

II.4. Food applications

Carbon dioxide is the most common supercritical fluid in the food industry. Due to the non-toxicity and low critical temperature, it can be used to extract thermally labile food components and the product is not contaminated with residual solvent. Further, the extract’s color, composition, odor, texture are controllable and extraction by supercritical fluid carbon dioxide retains the aroma of the product. Supercritical carbon dioxide extraction is used as a replacement for hexane in extracting soybean-oil and has been tested for extraction from corn, sunflower and peanuts. Supercritical fluid extraction provides a distinct advantage not only in the replacement but also extracts oils that are lower in iron and free fatty acid. To satisfy the consumer's need for 'lighter' foods, developmental work on supercritical extraction of oils from potato chips and other snack foods are been carried out. In addition, supercritical carbon dioxide has also been used to extract lilac, essential oils, black pepper, nutmeg, vanilla, basil, ginger, chamomile, and cholesterol.

A large amount of research has been concentrated on the decaffeination of coffee by supercritical carbon dioxide. Thus, it is not surprising to note that this was the first process to be commercialized (in 1978), whose primary step is supercritical extraction. Dry carbon dioxide cannot extract caffeine from dry coffee effectively and that the beans should be pre-wetted by water. Soaking of about 2 hours is necessary for efficient extraction of caffeine from coffee beans by supercritical carbon dioxide. Commercial processes for decaffeination of coffee include the Kraft General Foods in USA and several processes in Germany.

II.5. Pharmaceutical applications

Since the residual solvent present in the extracted material is of critical importance in the pharmaceutical industry, supercritical fluid carbon dioxide has found several applications. The extraction of vitamin E from soybean oil and a purification method for vitamin E has been well studied. The latter process avoids the step of vacuum distillation, which usually results in the thermal degradation of the product. Solubilities and recrystallization of various drugs has been demonstrated in supercritical fluids.

II.6. Environmental applications

Due to strict environmental regulations, supercritical fluids are used as replacements for conventional hazardous chemicals such as hexane. Supercritical fluid extraction has been proposed as an alternative technique for soil remediation and activated carbon regeneration. Over 99% of a majority of organics can be removed from contaminated soil. Organics that have been successfully extracted include PAHs, PCBs, DDT and toxophene. Carbon dioxide has been used with entrainers for the extraction of highly polar compounds. A commercial process to separate oils from refinery sludge and contaminated soil has been developed by CF Systems Corporation, USA. Chelating moieties that dissolve into carbon dioxide have been developed for the extraction of heavy metals from soil.  

References:

  1. Dixon, D. J. and Johnston, K.P., "Supercritical Fluids," In Encyclopedia of Separation Technology; Ruthven, D.M., Editor, John Wiley, 1997, 1544-1569.
  2. Akgerman, A. and Giridhar, M., "Fundamentals of Solids Extraction by Supercritical Fluids," In Supercritical Fluids - Fundamentals for Applications; Sengers, J.M.H.; Kiran, E., Eds., Klüwer Academic Publishers, 1994, 669-696.