Design Principles for Liquid-Liquid Extraction


Many factors must be evaluated when developing a liquid-liquid extraction process. Here are some of the important parameters to consider as you go from laboratory testing to commercial-scale operation.

Liquid-liquid extraction is an important separation technology for a wide range of applications in the chemical process industries (CPI). Unlike distillation, which is based on boiling point differences, extraction separates components based on their relative solubilities in two immiscible liquids. Extraction is typically chosen over distillation for separation applications that would not be cost-effective, or even possible, with distillation.

This article discusses the basics of liquid-liquid extraction and provides guidance on how to select the appropriate solvent and extraction equipment. It introduces key concepts associated with designing a liquid-liquid extraction process. The article also touches on unit design and the importance of placing temperature, pressure, and level control devices in the correct location.

The basics

In a liquid-liquid extraction unit, a liquid stream (carrier) containing the component(s) to be recovered (solute) is fed into an extractor, where it contacts a solvent. The two liquids must be immiscible or only slightly miscible; this allows them to form a dispersion, with one liquid dispersed as droplets in the other.

Mass transfer occurs between the droplets (dispersed phase) and the surrounding liquid (continuous phase). In order for the two liquids to be subsequently separated, they must have different densities. The droplets then accumulate above or below the continuous phase, depending on the liquids’ relative densities. The boundary between the continuous phase and the droplet dispersion is referred to as the interface, and can be at the top or bottom of the extraction column.

Figure 1 illustrates the general concept of liquid-liquid extraction. More fundamental information about liquid-liquid extraction can be found in engineering textbooks as well as in Perry’s Handbook(1).


Figure 1. An extraction unit has two inlet streams (the liquid carrier containing solute molecules, and the solvent) and two outlet streams (the raffinate and the solute-rich extract).

Extraction is typically carried out in continuous, staged units, which can be operated in either of two modes: with co-current mixing or with counter-current mixing. The co-current mixing mode is generally limited to one theoretical stage per extraction unit, whereas counter-current mixing is amenable to multiple stages per unit. For this reason, counter-current mixing is usually preferred over co-current mixing.

Counter-current extractors can be arranged in one of two ways (Figure 2), the choice of which depends on the density of the solvent relative to that of the solute carrier. If the solvent is less-dense than the carrier liquid, the solvent is fed into the bottom of the column, the solute is carried upward to the top of the extractor, and the carrier liquid is removed from the bottom of the unit (Figure 2, left). If the solvent is more-dense than the carrier liquid, the solvent is fed into the top of the column, the solute is carried downward to the bottom of the extractor, and the carrier stream is removed from the top (Figure 2, right).


Figure 2. Counter-current extraction units can be set up in one of two ways. Left: When the solvent is lighter than the carrier liquid, the solvent is introduced at the bottom of the column and the solute is carried up toward the top of the extractor. Right: When the solvent is heavier than the carrier, the solvent is introduced at the top of the column and the solute is carried downward by the solvent toward the bottom of the column. The red dots represent the region in which the solute is transferred to the solvent.

Selecting a solvent

Solvents used in liquid-liquid extraction are chosen to achieve maximum transfer of the solute from the carrier into the solvent. They must not be completely miscible with the carrier liquid and should have a high affinity for the solute molecules.

An ideal solvent for liquid-liquid extraction will typically have the following properties:

  • high solubility for the solute and low solubility for the carrier liquid
  • density difference vs. the carrier liquid greater than 150 kg/m3
  • mid-level interfacial tension (5–30 dyne/cm)
  • high resistance to thermal degradation
  • nonreactive with other chemicals involved in the extraction process
  • high boiling point (for ease of material handling)
  • low viscosity (for ease of handling)
  • nontoxic, nonflammable, and not corrosive to process equipment (for ease of handling)
  • low cost.

Engineers usually choose several solvents for process development. Often, the solvent candidates violate one or more of the criteria for an ideal solvent, so a comparison study is required to identify the best solvent for a particular application.

Dispersed-phase droplet size

An extraction column can be designed with either liquid as the dispersed phase: The solvent can be dispersed as droplets in the carrier liquid, or the carrier liquid can be dispersed as droplets in the solvent.

In packed columns, it is generally desirable to disperse the inlet stream with the higher volumetric flowrate into the continuous phase in order to maximize the interfacial area for mass transfer. In agitated columns, the liquid phase with the lower volumetric flowrate is typically dispersed, so that both phases (dispersed and continuous) have similar residence times. It is generally desirable to disperse the liquid containing the solute in the extracting solvent, so that the direction of mass transfer is out of the dispersed droplets and into the continuous phase.

The dispersion in the extraction unit is unstable, so the droplets will eventually separate into a bulk phase. Droplet size is a critical parameter in determining the rate at which that occurs. The maximum stable size of droplets in the contacting section of static (nonagitated) extractors depends on the internal energy and physical properties of the fluids, and can be determined through statistical measurements taken during extraction tests. Droplets in mechanically agitated extractors will be smaller than those in static extractors. The maximum stable droplet diameter (DM) for static extractors can be calculated by:


where c is the proportionality constant, P is the interfacial tension, Δρ is the difference between the two fluids’ densities, and g is the gravitational constant.

The direction of mass transfer can influence the behavior of liquid-liquid dispersions, thereby affecting the proportionality constant in Eq. 1. The Marangoni effect (2) explains how mass transfer of solute from one phase to the other creates interfacial tension gradients on the surface of the dispersed phase. Transfer of solute from the droplet outward into the continuous phase tends to create larger droplets and higher interfacial area of contact between the two phases, and is represented by a larger proportionality constant. Transfer of solute from the continuous phase into the droplets tends to form smaller droplets and the proportionality constant is smaller. The mass-transfer rate across the droplet surface can also influence droplet behavior, with high flux rates causing the droplets to shatter. Figure 3 illustrates these effects.


Figure 3. The solute can move either outward from the droplets to the continuous solvent phase, or inward from the continuous phase into the dispersed solvent. Left: The transfer of solute outward causes the droplets to grow and can cause them to shatter into smaller droplets. Right: The transfer of solute from the solvent into the droplets can create stable droplet surfaces.

Mechanically agitated extractors add energy to the system to influence the droplet size and thereby generate sufficient interfacial area of contact between the two phases. The high shear rate produced by turbulent agitation breaks the droplets into smaller droplets and distorts their shape. Equations have been developed to calculate the maximum drop size for agitated liquid-liquid extraction (1).

Axial mixing

Axial (longitudinal) mixing causes the extractor to deviate from ideal plug-flow operation. As a result of axial mixing, the dispersed phase and the continuous phase will exhibit a range of residence times, which tends to reduce the concentration driving force for mass transfer in the column. Axial mixing can be a particular problem in large-diameter extraction columns, because redistributors have less profound effects in larger columns than they do in small-diameter extractors.

Several mechanisms and conditions can lead to axial mixing, including:

  • Vessel walls or other fixed components can exert frictional drag on the liquids and cause turbulent mixing.
  • Nonuniform droplet size distribution in the contacting section of the extractor can cause the extractor to deviate from plug-flow operation.
  • Chaotic eddy patterns can form when fluids...

Would you like to access the complete CEP Article?

No problem. You just have to complete the following steps.

You have completed 0 of 2 steps.

  1. Log in

    You must be logged in to view this content. Log in now.

  2. AIChE Membership

    You must be an AIChE member to view this article. Join now.

Copyright Permissions: 

Would you like to reuse content from CEP Magazine? It’s easy to request permission to reuse content. Simply click here to connect instantly to licensing services, where you can choose from a list of options regarding how you would like to reuse the desired content and complete the transaction.