Extracting Phosphorus from Soil Using the Olsen Bicarbonate Method

So how does one measure the amount of plant-available P in soil?
soil testing
analytical chemistry

December 13, 2017


During the agricultural revolution of the 19th century, soil scientists began learning about the role of certain elements as nutrients in plants. Phosphorus (P) is one of these. P is considered a macronutrient, meaning that a deficiency of plant-available P in the soil will lead to stunted growth for the plant. Stunted growth means reduced yields when it comes to cultivated crops. A reduced yield was bad news in the 1800’s just as it is today. So how does one measure the amount of plant-available P in soil? If one measures total P, how much of that can a plant really access?

Thus began the search for measuring phosphorus in soil such that the measurement would correlate with the uptake of P by a given crop. This is the basis of soil testing by the middle of the 20th century: to extract and measure P in soil and relate it to the P requirements of the crop, so that agronomic practices (e.g. fertilizing) can be implemented and no losses in yield occur (Beegle, 2005).

As one might predict, the relationship between extractable-phosphorus in soil and phosphorus uptake in the plant are highly variable across the vast range of soils and crops. In this reality, soil scientists have long worked to develop different extraction methods for measuring soil P, each attempting to represent (usually through some field-crop calibration studies) the true amount of plant-available P in a given soil (e.g. Holford, Mallarino, Saunders, etc.). The nature of this soil testing universe is that some extractions will work well in a given soil type, but fail in another.

The most notable example of this is related to the procedure developed by Bray and Kurtz (Bray, 1945). The Bray Method, developed in the eastern US, uses a dilute acid to extract phosphorus from soils. The extracted-P calibrates well with plant uptake on many of the acid-to-neutral soils of the eastern US. It does not, however, calibrate well in the calcareous soils of the western US, basically due to the neutralization of the acidic extractant (Smith, 1957; Blanchar, 1964). Another extraction would need to be developed.


An increase in H+ activity will increase phosphorus (P) solubility. A decrease in Ca2+ activity also increases P solubility. With these two considerations in mind, one can see that an effective extraction would seek, ideally, to both remove Ca2+ and add H+ to the soil solution. Prior to the Olsen Method, some P extractions utilized weak acids. The complicating issue was that, in calcareous soils, Ca2+ was released, as acid dissolved CaCO3. Some extracted P then formed calcium phosphate in a secondary precipitation reaction, effectively removing the P from the extract before analysis.

By using sodium bicarbonate (NaHCO3) at pH 8.3-8.5, the concentration of HCO3- ion in solution is 63 times higher than CO3-. In a 0.5 M solution of NaHCO3, the equilibrium concentration of Ca2+ is 857 times smaller than in a CaCO3-water system at equilibrium. Essentially, at a pH 8.5, CaCO3 is formed and maintained. Given that a decrease of calcium activity yields an increase in phosphorus solubility, the Olsen bicarbonate extraction sufficiently overcomes the challenge of preventing calcium phosphate precipitates by keeping the calcium preoccupied, as it were, with a carbonate ion.

In neutral or acidic soils, calcium-bound P still tends to be most readily available, compared to the hard-to-get-at-P bound up in iron or aluminum compounds. The NaHCO3 extraction, carried out at a pH of 8.3-8.5, again helps to decrease Ca++ activity, which in turn increases P solubility. In fact, the bicarbonate ions introduced in the extraction performed better at replacing sorbed phosphorus on soil particles than did acetate and sulfate ions from acid extractions methods.

When compared to the P extracted by the Mehlich or Bray methods, the Olsen method generally extracts less P, except in highly buffered soils. Still, the Olsen Method has shown to correlate very well the applied P in field and greenhouse experiments to soil P measurements and plant response, and across a wide range of soil pH.


  1. Prepare 0.5 M solution of NaHCO3 adjusted to pH 8.30 - 8.50.
  2. Shake 2.5 g of soil with 50 ml of NaHCO3 solution for 30 minutes in 125 ml Erlenmeyer flask.
  3. Constant speed and room temperature should be maintained.
  4. Filter extracts through filter paper such that filtrate is clear.
  5. For most soils, dilute the filtrate 1:3 before colorimetric analysis to prevent issues measuring high concentrations.
  6. Measure P concentration in extract by standard colorimetry (e.g., ascorbic acid method, Murphy 1962) or inductively coupled plasma spectrometer, depending on research question.
  7. Calculate Olsen-extractable P from concentration measured on spectrometer.

Method Trap: Shaking

Shaking a soil in an extractant (e.g. sodium bicarbonate solution) is essential to the chemistry of this method. Shaking a soil properly is essential to achieving reliable results every time a soil is tested. Three components of shaking will determine the amount of P extracted each time from the same soil: duration of shaking, temperature of the extractant, and speed of the shaker.

An increase in the duration of shaking will increase the amount of phosphorus extracted from the soil. This is logical because the increased shaking duration will allow more time for soil chemical equilibria to shift P from one form to a soluble, extractable form. Olsen studied this in his development of the sodium bicarbonate extraction method (see chart below).

For some soils, an increase of shaking time steadily increases the amount of P removed from the soil, but for many others, an extraction of more than 30 minutes does not extract significantly more P. Additionally, in commercial soil testing labs that analyze tens of thousands of soil samples yearly, shaking each sample for more than 30 minutes could mean financial losses because the added marginal accuracy in the estimation of soil P isn’t worth the extra labor and waiting time.

The temperature of the extractant will influence the amount of phosphorus extracted. In chemical reactions, an increase in temperature usually increases the rate of the reaction since there is more energy in the system. In the case of soil P extractions, this means that, given two extractions lasting 30 minutes each, the extraction conducted at a higher temperature would extract more P from the soil. The simple solution is to conduct all extractions at room temperature, since room temperature is (ideally) constant from day to day. The reagents should all be at room temperature, too.

Like an increase of temperature, an increase in shaking speed will also increase the rate of chemical reactions or equilibrium shifts, since there is more energy in the system. Another way to think of it is that an increase of shaking brings the soil particles into contact with “fresh” solution more frequently, which expedites the extraction process.

Method Trap: Filtration

For two reasons can improper or inadequate filtration interfere with the analysis. One, if after filtration the samples still contain fine, suspended particles or colloids, then those suspended materials will scatter the light beam in the spectrophotometer regardless of the presence of phosphorus in the sample. Essentially, the instrument will “detect” phosphorus when it is really just detecting absorbance/scattering of light. Two, if after filtration the samples still contain suspended particles and the samples are destined for an instrument like an ICP-OES, there could potential be issues with the suspended particles blocking the sample feed line to the instrument.

Another issue that might occur when extracting phosphorus is the development of a colored extract, common on soils high in organic matter. Basic filtration will not remove this orange/dark yellow color. If the color is so dark that it is interfering with the transmittance of light in the spectrophotometer, then try adding ground, black carbon powder to the soil before the extraction; the carbon can absorb much of the color. If the coloring of the extract is minimal, then the readings from the spectrophotometer should be largely unaffected. This is because the wavelength of light is set around 880 nm for the Murphy molybdenum blue reactants and not for the yellowish color of the P extract.

Method Trap: Reagents

Possible problems with reagents include improperly preparation, improper storage, contamination, and use after expiration.

Obviously, improper preparations of reagents will affect the entire chemistry of the extraction and analysis. When the sodium bicarbonate extractant is to be between pH 8.3 and 8.5, it should be made a certain pH in that range every time. Sodium bicarbonate extractant set to a pH of 7.5, for example, will not perform at the optimal level established by the method. Worse perhaps would be improper preparation of the colorimetric reagents, such that the reaction of these with the extracted P doesn’t not occur properly, resulting in either sub-optimal or false color development. Likewise, storage recommendations for the colorimetric reagents should be followed. Some in the Murphy-ascorbic acid method require storage in a dark environment, such to preserve the sensitivity of the reagents.

Contamination is, in my experience, rare in soil P testing because the concentration of extractable P is orders of magnitude higher than the concentration of P residue that might or might not be present on the glassware and vessels used. In other words, a low soil test P results of 3 or 5 mg/L is still much higher than potential P in a run blank (0-0.2 mg/L)

And, of course, reagents should always be made with “fresh” chemicals that have not reached expiration.