Water activity, which provides information about product safety and quality, is often thought to be a more complicated measurement than moisture content. But making accurate, repeatable moisture content measurements is not as simple as it seems.
In theory, moisture content (MC) measurement is easy. Simply determine the amount of water in a product, and compare that to the weight of everything else in the product. In fact, it is actually a difficult and complex process to obtain an accurate percentage of water in a product. Here’s why.
Different reporting methods cause confusion
Moisture content (MC) is reported on either a wet or a dry basis. For the wet basis, the amount of water is divided by the total weight of the sample (solids plus moisture). For the dry basis, the amount of water is divided by the dry weight (solids only). Unfortunately, moisture content is often reported only as a percentage, without any indication of which method was used. Though it is easy to convert between wet and dry basis, confusion and potential problems occur when comparisons are made between moisture contents reported on a different basis. In addition, moisture content reported on a dry basis can actually result in a percentage value greater than 100%, causing more confusion.
Diverse measurement methods make comparisons impossible
The AOAC lists 35 different methods for measuring moisture content. These are classified as either direct or indirect measurement methods. Direct methods involve removing the water from the product (by drying, distillation, extraction, etc.) then measuring the amount of water by weighing or titrating. Direct methods provide the most reliable results but are usually labor intensive and time consuming. Some examples include air oven-drying, vacuum oven-drying, freeze-drying, distillation, Karl Fischer, thermogravimetric analysis, chemical desiccation, and gas chromatography.
Indirect methods do not remove the water from the sample. Instead, they involve measuring some property of the food that changes as moisture content changes. These methods require calibration to a primary or direct method. Their accuracy is limited by the accuracy of the primary method. Indirect methods are usually fast and require little sample preparation but are less reliable than direct measurement methods. Examples of indirect measurement methods include refractometry, IR absorption, NIR absorption, microwave adsorption, dielectric capacitance, conductivity, and ultrasonic absorption.
Further complicating the process of measuring moisture content is that one measurement method does not necessarily provide the same results as another, and the measurement method is normally not reported with the moisture content value.
Even direct measurement methods do not provide consistent results. Any method that requires heating the sample (i.e., loss-on drying) can lead to the loss of organic volatiles or decomposition of the sample (especially for samples containing high levels of sugar). For example, if organic volatiles are present in a sample or if the sample decomposes while being dried, a Karl Fischer analysis, which is not susceptible to volatile loss or decomposition, will give different results than a loss-on drying analysis.
Variability is difficult to avoid
One answer to these problems is to simply use a consistent method and only compare values that have been obtained in the same way. Unfortunately, consistency in measurement method for moisture content analysis still will not eliminate all problems. Consider, for instance, loss-on drying. This method seems simple enough. A sample is weighed, and the weight is recorded. The sample is then transferred to an oven, allowed to dry, and the dry weight is measured. The amount of water is determined by subtracting the dry weight from the initial weight, and the moisture content is then calculated as the amount of water divided by the dry weight or total weight, depending on the reporting method.
Even this simple loss-on-drying method is mined with potential variability traps. The most fundamental is that the term ‘dry’ has no real scientific meaning and has never been well defined. Instead, an arbitrary dryness that is reproducible has to be established for each sample. “Dryness” is often defined as the point at which weight loss ends. However, thermogravimetric graphs show that weight loss levels off at different temperatures for different products. Also, depending on the product, the length of time needed to achieve “dryness” will differ, and a temperature which produces “dryness” in one product may cause decomposition in another. This means that each sample has a unique ideal oven temperature and drying time. This ideal time/temperature combination is available in the literature for some products, but there are many for which it is not available. It is difficult to know which combination to use for untested products. If the same time/temperature combination is not used, the resulting moisture contents should not be compared.
Another complication is that many ovens set at one temperature can vary over time from that temperature by as much as 15 °C, and two ovens set to the same temperature can vary by as much as 40 °C.
Additional sources of variation for just the loss-on drying method include: oven vapor pressure, sample preparation methods, sample particle size, sample weighing, and post-drying treatment. It is interesting that despite the potential pitfalls, when a loss-on drying moisture content is reported in literature, it is immediately accepted as correct. In addition, when comparisons are made between moisture content methods and one of those methods is loss-on drying, it is always assumed that the loss-on drying measurement is correct.