Application Note: Sample Preparation 


A common question for using of water activity instrumentation is how do I prepare my samples?

For example should they be ground, sliced, mashed, emulsified etc. The answer is not simple and depends on the wider application and reasons for testing.


Why is understanding sample preparation important?

To grind or not to grind - tough call?

To grind or not to grind - tough call?

Firstly, we should consider best practice in performing aw measurements. For most users the goal is comparable and repeatable processes, so that measurements of a product today can be compared to measurements tomorrow. As such consistent and repeatable test procedures are vitally important, and this includes sample preparation. Repeatable tests ensure that any differences in readings can be attributed to something outside the test procedure; for example has the product formulation changed, is there an issue within the production process or perhaps the storage conditions are not within requirements. Any changes may affect important properties of a product; for foods these include shelf life, taste and appearance; for pharmaceuticals or chemicals these include viability or anti caking.

Underpinning repeatable measurements is repeatable sample preparation. It has been well documented that different preparation methods can significantly affect aw results. For example Harper et al. (2010) demonstrated that meat samples cut into 3.2 cm hexagons versus the same product diced into 0.4 cm2 cubes gave highly significant variations (+/-0.016 aw; P-value 0.001). For a relatively homogeneous product this is a large variation. For more heterogeneous products one can expect even larger variations.

Which preparation method should I use?

Which sample preparation method to use depends on what you are comparing the water activity measurements too. As we explained in our introductory presentation (Available Here) water activity is a qualitative value so should be used in combination with other testing.

Slicing may expose parts of a product with higher Aw. But if the product is not stored sliced is this relevant? 

Slicing may expose parts of a product with higher Aw. But if the product is not stored sliced is this relevant? 

For example we test a food product for shelf life and at the same time aw. You should then ensure that the food is prepared in the same way for aw testing as it is for the shelf life tests. Spoilage typically starts on the outside surface of food products so homogenisation and grinding may not be representative of how the product will be stored. Having identified the required aw tolerances for the product all future aw testing should be performed in the same manner. Clearly aw testing instruments have only limited sample capacity as such preparation should be kept to a minimum and performed in the same way for each future batch test.

However there are occurrences when limited sample preparation is not ideal. Testing product consistency and quality in-relation to aw may require products to be ground to provide aw results representative of the entire product. An excellent example is coffee. It may be that whole beans are being sold but testing the coffee in a ground state represents firstly how the coffee will ultimately be used but also provides an understanding of the overall aw of the product. Tablet testing is also typically performed with ground product.

Controlling temperature is an important aspect of ensuring repeatability

Controlling temperature is an important aspect of ensuring repeatability

It can be seen that there is no right answer, our advice is;

  1. Be consistent in your procedures
  2. Record what method you use and why
  3. Be detailed with your preparation procedures
  4. Check how the product is processed for other tests
  5. Limit handling and exposure of the product (which might result in moisture loss or gain)


N.M. Harper, K.J.K. Getty, E.A.E. Boyle, Evaluation of sample preparation methods for water activity determination in jerky and kippered beef: A research note, Meat Science, Volume 86, Issue 2, October 2010, Pages 527-528, ISSN 0309-1740