No abstract

No abstract

No abstract

No abstract

A method was developed and a collaborative study was performed under the auspices of the International Commission of Uniform Methods for Sugar Analysis. The collaborators used high-performance anion-exchange chromatography (HPAEC) to determine trace amounts of glucose and fructose in 3 raw and 3 refined sugar samples provided as blind duplicates. Fourteen laboratories participated in the study. Although difficulties were experienced by a few analysts, 10 laboratories reported positive results. Average repeatabilities and reproducibilities for glucose and fructose in raw sugar were slightly above 5 and 10%, respectively, and the average Horwitz ratios were well under 2. Average repeatabilities and reproducibilities for glucose and fructose in refined sugar were 10 and 22%, respectively; although the Horwitz ratios were >2, they were marginal (2.8). The HPAEC results agreed with results obtained by gas chromatography in an independent laboratory. Because the method was collaboratively studied according to the protocol of the International Union of Pure and Applied Chemistry for collaborative studies, and the results meet its criteria, it is recommended that the method be adopted Official First Action by AOAC INTERNATIONAL.

Near infrared spectroscopy (NIRS) for the simultaneous analysis of brix, pol, sucrose, glucose, fructose and ash in sugar factory processing streams has many advantages. The most important are the almost immediate reporting of results which can be used for process control, and the ability to analyse many samples over a short period for troubleshooting purposes. The Sugar Milling Research Institute NPC has developed and refined near infrared transmission spectroscopy models for the prediction of these analytes in sugarcane processing streams. The uptake of the technology in the sugar factory laboratories has proved challenging, requiring improved analyte prediction ranges, understanding the transfer of instrument results into factory information management systems and the introduction of a simplified sample dilution procedure. The NIRS technology has allowed for the development of toolkits for routine factory scale sucrose inversion studies and measurement of losses across centrifuges, previously difficult to implement.

The International Commission for Uniform Methods of Sugar Analysis Ltd (ICUMSA) is the only international organisation involved with the development and testing of analytical methods for use within the sugar industry. Analysis methods undergo a rigorous approval procedure, progressing from Tentative to Official or, Official (Reference) method status before being published in the Methods Book. As an example of these processes, the comparison of two alternative computations (Berding and Pollock; Hamna) used in a proposed hydraulic-press method for cane evaluation is described. The press method was compared to the Official Method for cane analysis (GS5/7-1 (2011)) that is based on wet disintegration of cane, by comparing the results obtained from 79 samples split into three subsamples analysed by the three methods. Correlations of greater than 0.9 were recorded for the press method’s sugar and fibre results compared against the existing method. A recommendation was made to ICUMSA, based on the study, that a draft method should be written for cane evaluation using the press method and Berding and Pollock computations.

The Sugar Milling Research Institute NPC (SMRI) has developed a simple to use near-infrared spectroscopy (NIRS) transmission-based analysis method as an alternative to conventional methods for analysis of sugarcane factory stream samples. The technology provides rapid, simultaneous analysis of refractometric dry substance (rds), polarimetric sugar, sucrose, glucose, fructose, conductivity ash contents as well as colour and pH for all streams and additionally, dry solids for final molasses and eliminates the need for sample clarification chemicals. The analyte prediction equations were developed using conventional results of samples from 14 South African factories, analysed at SMRI using SANAS/ISO17025 accredited test methods, and NIRS scans of the same samples using up to 16 different NIRS instruments. The NIRS analyte prediction equations were validated against more than 1,500 independent factory samples that had been analysed by conventional methods of analysis, including samples from factories outside South Africa. The reproducibility of the NIRS results were equivalent to existing conventional analysis reproducibility values (juice and final molasses) and previously undocumented values determined for this study for conventional raw house analysis methods. Correlation coefficients of greater than 0.97 were recorded for all major analytes and greater than 0.9 for minor analytes when predicted results were compared against conventional results. A maintenance protocol was also developed to ensure that the prediction equations remain robust and can account for sample matrix variations that can occur from season to season. The SMRI-NIRS technology was installed at all 14 South African factories and found to be robust and give equivalent results to conventional methods of analysis.