Food analysis is required for a wide variety of reasons - from testing suitability for purpose, through checking shelf-life or authenticity, to assuring legal compliance.
Instrumentation has improved considerably in the past 10 years with advances in microelectronics and engineering offering analysts in-, on-, or at-line capability for measuring food quality. There are instruments which could effectively replace laboratory testing and offer real-time control of product quality and manufacturing control.
However, the take up of new rapid methods of instrumentation within the food industry has been limited due to the resources required to effectively evaluate the best options. Faced with the challenge of prioritising tasks, it is often easier for manufacturers to continue with less efficient off-line measurements than to test and validate new on-line systems.
Yet many of the new systems, although expensive to set up, offer rapid recovery of initial capital. In many companies, there has therefore been a sporadic installation of on-line measuring systems, which are often undertaken as a one-off project championed by an enthusiastic technical manager, rather than as part of an ongoing company strategy.
Campden BRI has been monitoring and assessing rapid methods through project ROBOT - a Campden BRI member-funded research project looking at the future for analysis in food factories. The project's objectives include:
- Reviewing cutting-edge rapid analytical tools.
- Evaluating the food and drink industry's needs and current applications for rapid analytical instrumentation.
- Assessing selected rapid analytical tools at Campden BRI's laboratories and reporting back important technology-driven opportunities to Campden BRI member companies.
The following examples illustrate the systems explored.
In one study, an analyser, which uses transmittance measurement and scans across the whole near-infrared spectroscopy (NIR) spectrum, was used to determine fat, protein and moisture in various meat and dairy products. The system was supplied pre-calibrated and replicate analysis was performed on a range of homogenised meat and dairy product samples. Results were obtained in minutes and compared to results produced by different reference methods - fat by solvent extraction, moisture by oven drying, and total protein by Kjeldahl nitrogen determination.
Results from the analyser generally showed good correlation with the results obtained using reference methods - though a couple of samples (frankfurters and breaded ham) did show that further calibration would be necessary to produce reliable results for protein. Interestingly, the method was fairly robust and there was no need to calibrate the instrument prior to making measurements in different sample types.
A second study used a reflectance based instrument which used a range of specific filters for scanning at different wavelengths. In this, the analyser was used to determine fat, protein and moisture in beef and pork mince samples. Although the instrument was pre-calibrated for measurement of parameters, the manufacturer recommended further calibration on a range of high, intermediate and low fat samples to improve the accuracy of measurement for the sample type. Again, results were compared to those obtained using the reference methods.
Results obtained for fat analysis on both mince types showed good correlation (>96 per cent) with those obtained by reference methods. The moisture results were more variable, with >90 per cent correlation with beef, but only 63 per cent with pork. The protein results gave a correlation of 86 per cent with pork mince and 72 per cent with beef mince. It was concluded that further work and calibration on samples covering a much wider range of fat, moisture and protein levels would be required to improve the accuracy of measurement using this instrument.
Testing worth its salt?
The sodium content of food is often declared on product packaging and therefore quick and accurate methods for its determination will benefit the industry. An instrument was evaluated which used thermometric titration to determine sodium levels in various food products. Results were compared to results produced by atomic absorption spectrophotometry (AAS).
The majority of the initial results obtained by thermometric titration showed reasonable correlation with the results obtained using AAS. Further evaluation of the approach on different food types and a higher number of replicates would be required to validate the technique so that it is proven to be suitable for use in food production.
In a separate study, a digital salt meter was used to measure salt in bakery products. The device detects and converts the conductance of a sample into sodium chloride concentration.
Bakery products were blended with water and then filtered prior to measurement. Reference analysis was performed on the same samples using both silver nitrate titration to determine chloride content and atomic absorption spectrophotometry (AAS) to determine sodium content.
Results from analysis showed reasonable correlation with results obtained using AAS.The measurement of sodium via chloride (using silver nitrate titration) produced much lower values. It seems that the conductivity based digital salt meter is most suited to measurements of liquid foods in laboratory conditions. For bakery products, an additional step of blending with water was required.
As the results obtained correlated with results obtained with AAS but slightly overestimated the salt content, further calibration with salt standards prepared in bakery matrices may make the method more accurate.
Another approach explored was the direct pH measurement of food products by non-glass pH systems using ion selective field effect transistor (ISFET) silicon chip sensors. Avoidance of glass is crucial in food production areas, to mitigate against potential contamination of the food. The ISFET silicon chip sensor can be used as an alternative to glass electrodes for pH measurement and is robust enough to be used on the production line. The probes are designed for rapid analysis of liquids or semi-solids and some systems can be used for in-line applications.
Various food samples - including prawns, sausage, poultry, gravy, vegetables, soups and fillings were analysed in duplicate under standard laboratory conditions and compared with a UKAS reference method with a relevant pH meter with appropriate glass electrode. Results from the laboratory analysis on the range of food products tested showed good correlation with the results obtained by the traditional method.
The ISFET pH probe was also trialled in a poultry factory and it became apparent that the environmental conditions and variable temperature of the chickens affected the probe measurement, with repeated calibration being needed. Thus it is important to evaluate any new instrumentation in a factory environment to be sure it is robust enough.
Looking to the future
As instrumentation continues to evolve, with increased options for in-, on- and at-line analysis of raw materials, ingredients and products, many manufacturers will need independent assistance in identifying, assessing and implementing the new technology. Conversely, manufacturers of instrumentation are seeking independent assessment of new kit in a variety of environments and with a wide range of products.
One of the key roles of Campden BRI is to help bring these parties together - to ensure that the latest developments become a practical reality as quickly as possible. It is to everyone's advantage to deploy the most powerful technology quickly and effectively to support the production of safe, wholesome food.
Steve Garrett, Molecular Biology Group, Campden BRI, and Julian South, Head of Chemistry and Biochemistry, Campden BRI, Chipping Campden Gloucestershire, UK. www.campden.co.uk.