No thanks! I would like to know more about CHROMacademy

 Over 3000 E-Learning topics / 5000 Articles & Applications
Troubleshooting Liquid-Liquid Extraction

Liquid-liquid extraction (LLE) is the most widely used extraction technique for liquid samples.  The information provided in this article will mainly focus on liquid-liquid extraction (LLE) using a separatory funnel, however, many of the principles are applicable to the more modern miniaturized approaches to LLE i.e. single drop microextraction (SDME), flow-injection extraction (FIE), membrane extraction techniques, and supported liquid extraction (SLE).

Some of the potential issues with liquid-liquid extraction (LLE) are

  • Emulsion formation
  • Analytes strongly adsorbing to particulates
  • Analytes bound to high MW compounds (protein-drug interactions)
  • Mutual solubility of the two phases
  • Labor intensive and time consuming
  • Produces large volumes of waste (traditional LLE using a separatory funnel)
    • Expensive to purchase organic solvent and dispose of waste
    • Personal safety
  • Potential for solvent mediated decomposition
  • LLE may have method robustness and transfer issues due to manual processing
    • Being superceded by solid phase extraction (SPE), solid phase microextraction (SPME), and miniaturized modes of LLE
  • Difficult to automate but not impossible


Emulsions occur when a sample contains a high amount of surfactant-like compounds (i.e. phospholipids, free fatty acids, triglycerides, proteins etc.).  These surfactant like molecules are large and will have mutual solubility in the aqueous and organic solvents which results in the formation of an emulsion where there is a mid-zone between the two phases which has intermediate solubility in each of the two phases making it difficult to quantitatively collect one phase or another (Figure 1).  The emulsion can also trap some of the analyte of interest leading to quantitative problems. 

Emulsions often occur with samples where the animal (or human) diet is high in fats.  Thus, emulsions sometimes appear when passing from pre-clinical trials, with animals on low-fat controlled diets, to clinical trials with humans who may be on high-fat diets.  This characteristic problem makes LLE a less dependable procedure if it is expected that the same extraction protocol will be used for both pre-clinical and clinical samples.  If this problem is anticipated it is worth trying high-fat samples during method development in addition to the standard test matrices.


Content on this page requires a newer version of Adobe Flash Player.

Get Adobe Flash player

Figure 1 Emulsion formation.

There are a few tricks of the trade to either stop emulsions forming in the first place or to disrupt them if they do form.  However, as a general rule it is easier to prevent the formation of an emulsion than to break it after it has been formed.  The simplest way to prevent the formation of an emulsion is to gently swirl instead of shake the separatory funnel.  By swirling the separatory funnel the agitation which can cause the emulsion to form is reduced, however, the surface area of contact between the two phases is maintained to allow for extraction to occur. 

If an emulsion is formed it can be disrupted by addition of brine or salt water which increases the ionic strength of the aqueous layer and will force separation of the two phases by forcing the surfactant-like molecule to separate into one phase or the other - this technique is known as salting out (Figure 1).
The individual layers or emulsion can often be separated from each other via filtration through a glass wool plug (to remove the emulsion) or a phase separation filter paper (to isolate a specific layer).  Phase separation filter papers are highly silanized and depending on the treatment (or type) of paper it will allow either the aqueous or organic phase to pass through and be isolated.

Addition of a small amount of different organic solvent will adjust the solvent properties of the separation and can result in the surfactant-like molecule being solubilized in either the organic or aqueous layer to a greater extent which breaks the emulsion.

Centrifugation of the separation can be used to isolate the emulsion material in the residue of the centrifugation.
Heating or cooling the extraction vessel can be used to break the emulsion but should be the method of last resort as heating glassware can be problematic.  For example, if a separatory funnel is heated it needs to be open to the atmosphere so that pressure does not build up inside the funnel causing a safety hazard.  Glassware also needs to be suitable for heating i.e. Pyrex or Chemex, as some of the softer glasses will not hold up when heated. 

Analyte Adsorption

In LLE, if the analyte has active sites (i.e. if there are acidic or basic polar sites) adsorption to any particulate matter within the sample matrix can occur.  This can be quite common when extracting aqueous samples into an organic solvent.  Adsorption of the analyte(s) during the extraction process can cause low recovery. 

Prior to carrying out a LLE the sample can be filtered to remove any particulates, these should also be washed with a strong organic solvent in order to remove any analytes which are already adsorbed to the particulates; this extract can then be combined with the organic analyte phase from the LLE.  To fully recover analytes which are strongly adsorbed to particulates may require a change in pH, an increase in ionic strength, or the use of a more polar organic solvent depending on the chemical nature of the analyte.


Solute Binding

As well as binding to particulates in the sample, analytes can also bind to other components within the sample matrix.  This can be particularly problematic when extracting biological samples which contain proteins as these are notorious for binding small analytes.
There are several strategies for disrupting protein/analyte binding.

  • Addition of detergent which would preferentially bind to or extract the analyte of interest
  • Addition of organic solvent, chaotropic agent, or a strong acid will denature the protein which in turn minimizes the protein/analyte interaction
  • Dilution with water which will drive more analyte into the aqueous phase
  • Displacement with a more strongly binding component i.e. a chelating reagent such as EDTA (ethylenediaminetetraacetic acid)

All detergents contain a hydrophilic “head” and a hydrophobic “tail” region.  It is these structural characteristics which allow detergents to aggregate in aqueous media and be used to break up protein/analyte interactions.  At high concentration, the polar hydrophilic region of each molecule is oriented toward the polar solute (water) while the hydrophobic regions are grouped together to form thermodynamically stable micelles with hydrophobic cores (Figure 2).  The hydrophobic core region of the detergent micelle associates with the hydrophobic surfaces of proteins and results in soluble protein-detergent complexes. Actual micelle structures are more complex and dynamic, and can change due to detergent concentration and solution composition.

Detergent micelle

Figure 2: Detergent micelle structure.


A chaotropic agent is a molecule in water solution which can disrupt the hydrogen bonding network between water molecules.  This has an effect on the stability of the native state of other molecules in the solution, mainly macromolecules (proteins, nucleic acids) by weakening the hydrophobic effect.  

For example, a chaotropic agent reduces the amount of order in the structure of a protein formed by water molecules, both in the bulk and the hydration shells around hydrophobic amino acids, and may cause its denaturation biochemistry, which is turn minimizes the protein/analyte interaction.


Mutual Phase Solubility

LLE is normally carried out using an aqueous phase and a water immiscible organic solvent.  However, even solvents which are considered to be immiscible in water will have some solubility in water (Table 1).  This mutual phase solubility can change the relative volumes of the two phases.  In order to avoid this it is good practice to saturate each phase with the other so that the volume containing the analyte can be accurately known in order to correctly determine analyte recovery.  The process of saturating the two phases is carried out in a separatory funnel without the sample; ultimately the two phases are mixed as you would during the extraction process and allowed to separate.  Either phase can then be used for the LLE of the analyte.

Solvent Solubility in Water
(g/100 g)
Density (g/cm3)
Benzene 0.18 0.8765
n-Butanol 7.7 0.8095
Chloroform 0.8 1.4778
Cyclohexane < 0.1 0.8110
Dichloromethane 1.32 1.3266
Ethyl acetate 8.7 0.9003
Diethyl ether 7.5 0.7138
Heptane 0.01 0.6795
Hexane 0.14 0.6606
Methyl tert-butyl ether (MBTE) 4.8 0.7353
Pentane 0.04 0.6262
Toluene 0.05 0.8668

Table 1: Water “immiscible” solvents.

Download a pdf copy of a Solvent Miscibility Chart here


Salting Out

Salting out is a useful strategy when trying to extract, for example, drugs which are polar from biological fluids, heavy metal ions from aqueous solutions, or pesticides in fruits and vegetables (QuEChERS - salting out is the basis of this technique).  Addition of an inorganic salt (i.e. NaCl) to a mixture of water and water miscible organic solvent will change the ionic strength which can affect the solubility of the analyte and causes separation of the two phases. 

The use of different salts and different salt concentrations will result in a different degree of phase separation.1-2  This technique is useful as it allows the use of highly polar water miscible solvents to extract analytes that cannot be extracted using traditional LLE (aqueous/water immiscible solvent).


Solvent Selection

Selecting the correct solvent to perform LLE extraction is key to the success of this technique.  As has been mentioned previously, it is common to use a water/water immiscible organic solvent as there is a polarity difference between the phases which allows for analyte extraction and phase separation. 

However, the use of the salting out technique described above does allow LLE to be extended to the extraction of analytes using water/water miscible solvent systems.  It is important that the analyte of interest has preferential solubility in one solvent over the other to facilitate optimum recovery.  When selecting a solvent it is important to note the density of the solvent and water with respect to each other so that the phases can be identified i.e. higher density solvents will be the lower layer (Table 2).  It is also worth noting that it is good practice to keep both layers throughout the LLE process so that the wrong layer is not incorrectly discarded or if further extraction is required if the recovery is low.

Often if a known analyte is being analyzed the literature will already contain methods for extraction and these can be useful starting points which may only need a small amount of optimization for a particular sample.

It should be noted that in all areas of chemistry (especially those which use a large volume of solvent and, therefore, produce large volumes of solvent waste) there is a push to use “greener” solvents.  For example, there are many methods which utilize hexane as the extracting solvent, which has inherent toxicity issues; therefore, switching to heptane which has similar solvent properties but a less severe toxicity profile would be advantageous.  An alternative solvent for extractions which currently use dichloromethane is ethyl acetate.

Solvent Density (g/cm3)
Hexane 0.6606
Diethyl ether 0.7138
Acetone 0.7845
Acetonitrile 0.7857
Toluene 0.8668
Water 0.9970
Dichloromethane 1.3266
Chloroform 1.4788

Table 2: Solvents listed by density.


Distribution Ratio

An important question in any LLE procedure is where is the analyte going and how can we measure analyte recovery?  LLE is an equilibrium technique and the analyte may not be able to be quantitatively extracted in one extraction, therefore, multiple extractions are carried out (often at undergraduate level we are taught that three extractions are required).  However, this needs to be quantified better in order to guide method development. 

This can be done using the distribution coefficient, K (Equation 1) which provides guidance in determining extraction efficiency by allowing a determination of the analyte concentration in solution following a certain number of extractions.  Ultimately the distribution ratio allows us to determine the concentration of analyte in the organic phase ([A]org) versus the concentration of analyte in the aqueous phase ([A]aq).  The value of K does not necessarily need to be measured as literature will often provide Ko/w (distribution coefficient for water/octanol) values for known analytes which will suffice.  Equation 2 can be used to quantify the extraction efficiency of a LLE.  


[A]org = concentration of analyte in the organic layer
[A]aq = concentration of analyte in the aqueous layer
[A]i = analyte concentration which remains in the aqueous phase after i number of extractions
Vaq = volume of the aqueous layer
Vorg = volume of the organic layer
[A]o = original concentration of analyte in the aqueous phase
i = number of extractions
K = distribution coefficient

Using Equation 2 the effect that organic solvent volume and distribution coefficient (K) can have on method development can be investigated (Figure 3).  Developing a LLE method will be primarily concerned with extraction efficiency, volume of extracting solvent used, and labor intensiveness of the process - all of these parameters need to be balanced. 

We can look at each of the scenarios illustrated by Figure 3.

The experiment depicted by the black line uses a solvent in which the analyte has a distribution coefficient (K) of 2.  A fairly standard volume of aqueous solvent (100 mL) and an equal volume of organic solvent is used for the LLE.  As the analyte has a distribution coefficient K = 2, this means that on extraction two parts of the analyte will go into the organic phase and 1 part will remain in the aqueous phase.  Therefore, after one extraction a third of the analyte will remain in the aqueous phase.  A second extraction will result in 11% of the analyte remaining in the aqueous phase (this corresponds to 89% analyte recovery), and after a third extraction an extraction yield of 96-97% can be achieved which can be considered quantitative.  This protocol could achieve quantitative extraction using three extraction steps and 300 mL organic solvent.
If the amount of organic solvent is doubled (blue line) then theoretically the quantitative extraction of the analyte (K = 2) could be achieved in fewer extractions.  In this case two extractions give quantitative extraction but using 400 mL of organic solvent.  While this process is slightly less labor intensive a greater amount of solvent is used, therefore, it has to be decided which is more important, labor intensiveness or solvent consumption.

If using less organic solvent is deemed the most important parameter then this can be achieved by halving the volume of organic solvent used in each extraction step (green line), however, this is a  much more labor intensive (5 extractions) process.  The total volume of solvent used would be 250 mL.
Ideally a solvent would be selected which provides the highest possible value or K.  For example if an analyte had K= 10 quantitative extraction could be achieved in one or two extractions (red line).  However, K=10 is an unusual circumstance as those types of analytes will not be soluble in the aqueous phase.

Another thing to keep in mind with this approach is that LLE techniques are all based on equilibrium constants, the fore, two things need to be considered.

  1. Equilibrium does not tell us anything about kinetics - the flask must be swirled to help drive the analyte from one phase to the other phase.  Although a higher K is more likely to speed up the kinetics.
  2. In most cases we do not go completely to equilibrium.  In a real world situation the curve shown in Figure x would exhibit lower extraction efficiencies.
Distribution ratio

Figure 3: Optimization of LLE - effect of extracting solvent volume and analyte distribution coefficient K.


Acid, Base, Neutral Extractions - Where’s the Analyte?

As previously mentioned it is worth saving both phases during the extraction process so that the analyte is not accidently discarded if the wrong phase is identified.  Or if poor extraction efficiency is found subsequent extractions may need to be carried out.

There are several tools which can be used to enhance extraction efficiency such as selecting the correct solvent (consider analyte solubility and K), with ionizable analytes using the effects of pH to alter the solubility profile (force the analyte preferentially into one phase or the other), altering the volume of organic solvent used etc.
Most real world samples may contain different molecule types (acid, base, neutral) and how these are extracted, and the location of the analytes (organic or aqueous phase) needs to be carefully considered to avoid quantitation and separation errors. 

For example, when working with a sample which contains neutral, basic, and acidic analytes (Figure 4) the location of each analyte can be manipulated - this can be useful if all analytes need separated, if the analyte was required to be in either an organic or aqueous matrix for compatibility with the subsequent analytical technique (HPLC, GC etc.), or if the analyte needs to be separated from other matrix components which have preferential solubility in one or other of the phases (i.e. moving the analyte to the organic phase may prove the most successful way of achieving a clean separation).  

When molecules are in their non-ionized form they are less polar and will be more soluble in less polar solvents (i.e. the organic phase), whereas, ionized compounds are more polar and soluble in polar solvents (i.e. water).
It can be seen from Figure 4 how the use of pH can be used to manipulate the solubility of the ionizable analytes and separation of all four molecules can be achieved.

Acid base neutral LLE flow chart

Figure 4: Acid, base, neutral LLE flow chart.




  1. //
  2. Frankforter, G. B.; Cohen, L. J. Am. Chem. Soc. 1914, 36, 1103–1134.

This article was based in the CHROMacademy webcast - Troubleshooting Sample Preparation

You may also like:

e-Learning Modules
Primary Sample Preparation Techniques
Liquid-Liquid Extraction Techniques
Microextraction Techniques

Choosing the Correct Sample Preparation Technique

Quick Guides
Troubleshooting Filtration
Investigating SPE Recovery Problems

loading data
loading data
loading data
loading data
loading data

group  subsCHROMacademy can deliver to corporate clients on a multi-user subscription basis.
Served up from secure servers to the corporate intranet or individual desktops.

  • Microsite - your own learning site powered by CHROMacademy
  • Your Landing Pages -with your logo and branding
  • Customized Assessments - Based on content agreed upon Certificate of Completion
  • Certification Programs - Offer your learners a goal to strive towards
  • LMS : Connect - Our Learning Management System is S.C.O.R.M. compliant and will connect to your system
  • Engagement Package - E-newsletter stimulation program derived from your content and ours
  • Full archive of Essential Guide webcasts & tutorials
  • 1000’s of eLearning topics - HPLC / GC / Sample Prep / Mass Spec
  • Ask the Expert - our experts will answer your chromatography questions within 24 hrs.
  • Assessments - test your knowledge
  • Application notes & LCGC articles
  • Troubleshooting and virtual lab tools

Request a quote


 Home | About UsContact Us | SubscribeTerms and Conditions | Advertise | Privacy Policy 

loading data

loading data

loading data


loading data

loading data