ICUMSA News n°32 – 1997

Message from the President

We are nearing the end of the 22nd Session and Referees’ reports for the Berlin meeting should now be completed. As these reports become available this office will mail copies to Chairmen of National Committees and it is expected that Chairmen will distribute them to interested members. Members attending the Berlin meeting should bring their copies to Referees’ reports to the conference.

The arrangements for the Berlin meeting have been communicated to Chairmen of National Committees and these should now be available to all members. Anyone seeking further information should contact the conference organisers:

Verein der Zuckerindustrie, Attention Mrs. Eva Sawadski

Postfach 2545, D-53015 Bonn, Germany

Phone: +49 228 228 50, Fax +49 228 228 5100 (or 5101),


At this time, we have two nominations for President. The invitation to those candidates to prepare manifestos for members to consider before voting at the Executive Committee has met with a response by one candidate, Professor Giuseppe Vaccari. His contribution follows.

Professor Giuseppe Vaccari’s Manifesto

At the time that nominations were called to fill the Office of President after the 22nd Session in 1998, Dr. Player may have imagined there being a long list of candidates for the voter’s attention at the Berlin meeting. This did not eventuate, and the number of candidates is in fact very limited: The Indian National Committee has nominated a candidate and I have been nominated by the Czech National Committee. I wish to thank them sincerely for the confidence they have placed me in.

The reasons for the low number of candidates surely does not reflect a diminishing interest in the Commission’s work nor the lack of distinguished scientists able to hold this office. One of the reasons could be the responsibility that such an office calls for and the time demands now placed on scientists and technologists employed at the various Research Institutes and Sugar Companies. In contemplating tackling such a prestigious task one has to see it as a sacrifice to be made in order to serve the objectives of the Commission rather than an acknowledgement of personal talent. It was on this basis that I have carefully considered the Czech National Committee’s nomination.

As a University Professor I am largely free to organise my own work and since the work of ICUMSA is compatible with my own research interests, I believe that I could undertake the role of President if elected.

Our President, Dr. Player, has during his 12 years achieved all the reforms of the Commission that were called for. These included the introduction of new Refereeships as well as the achievement of goals established at the time of his installation. These include the publication of a new Methods Book, the periodical newsletter “ICUMSA News” and the history of ICUMSA. I believe that the new President should continue along the course that Dr. Player has charted while implementing any changes made necessary by changed circumstances or requested by National Committees. The large amount of work carried out over many years has been achieved thanks to the invaluable cooperation of many contributors and this is well documented in the Proceedings of 1990 and 1994 and in “ICUMSA News” since 1987.

It is clear that to successfully achieve the President’s goals the past cooperation needs to be maintained and increased. It is of course es sential that the President’s leadership inspire such cooperation. Recently, it has been proposed that the President’s workload might be relieved by installing a permanent secretariat. These proposals have not received unanimous agreement and this matter will be discussed again in Berlin. I believe that the new President must be able to accept whatever solution might be adopted and I am hoping that a compromise can be reached with unanimity. In the event that a permanent secretariat is not adopted I propose that my colleague and mentor Giorgio Mantovani be my General Secretary. Concerning the management of ICUMSA’s publications following John Duttons retirement as Publications Manager, I am working with Dr. Player in exploring a number of options and would expect to make a proposal shortly.

Giuseppe Vaccari

In Memoriam Albert Emmerich

NIR Polarimetry project finished succesfully before his death

It was the intention of the Editor to publish a progress report for Subject 4: Polarimetry and Quartz plates written by Dr. Albert Emmerich, however Dr. Emmerich died on August 17, 1997 at the age of 76 years. He was head of the Department for Analytical and Physical Chemistry in the Institute for Agricultural Technology and Sugar Industry at the Technical University in Braunschweig from 1949 to 1984. In the final years of his professional career he also was Assistant Manager of the Institute.

The main points of his scientific work were based – according to his specialisation and interest – on the physical properties of sugar solutions and the development of analytical methods for the sugar industry. For many years he collaborated closely with the national physical laboratories (especially the PTB) and this lasted practically until his death. The internationally accepted results of his work were precise basic values (tables) for the optical rotation, refractive index and density of sugar solutions, which now are closely connected with his name.

In his last years Dr. Emmerich concentrated his work on finishing the density tables for invert sugar and its components fructose and glucose in cooperation with the German national physical laboratories in Braunschweig (PTB). The ICUMSA project on NIR Polarimetry provided for extension of the International Sugar Scale to longer wavelengths of light was his main occupation in the last years of his life. He supervised the preparation of the ultrapure standard solutions of sucrose destined for the precision measurements in the PTB reference laboratory for Polarimetry. He was always supported generously by the Braunschweig Sugar Institute. Fortunately, Dr. Emmerich had the privilege of finishing this important work completely after clearing all the hurdles in the new territory of NIR measurements. Though suffering strongly from a severe illness in the last months of his life Dr. Emmerich crowned his life’s work by reviewing important chapters of his former research work in the book on 100 years of ICUMSA history (in print). The survey of the ICUMSA Sessions 1982 and 1986, when he was General Secretary of ICUMSA and in this function the spiritus rector of the organisation, was written by himself into the history book.

To his death Dr. Emmerich was Honorary Life Secretary of ICUMSA and Referee for Polarimetry. With him ICUMSA and the scientific world lost an outstanding personality and an untiring doer.

Klaus Thielecke and R.F. Madsen

Progress Report for Subject 4:

Polarimetry in the NIR

by Joachim Keitel, Acting Referee

Polarimetry is still the most important analytical method in the sugar industry. One of the advantages over alternative methods is, that the instruments can be checked and calibrated easily using certified quartz plates. To get the readings at different wavelengths, Bünnagel’s formulae for the so called Optical Rotary Dispersion (ORD) of quartz and sucrose are used.

The main drawback of standard polarimeters is, that they are not able to handle very dark solutions without a prior clarification. Normally, this is done by adding the toxic basic lead acetate. To overcome the need for lead clarification, some manufacturers introduced polarimeters using near infrared (around 880 nm) light source instead of the typical yellow sodium light (589.44 nm). Most sugar solutions can be read in these instruments when filtered to remove turbidity. Until now, no certified value for the quartz plates in the NIR could be given. The recommendation by ICUMSA stated, that Bünnagel’s formulae for the ORD of quartz and sucrose, which are valid up to 650 nm only, should provisionally be extrapolated to the near infrared (Proc. 20th Session ICUMSA, 1990, p. 206).

At this point, the PTB, the Sugar Institute in Braunschweig, and the company Schmidt + Haensch undertook a project to obtain data for the calibration of polarimeters working in the near infrared. This project was funded by ICUMSA, Senate of Berlin, PTB and the Braunschweig Sugar Institute.

The rotation values for three sucrose solutions and for PTB 49 mm quartz plate were measured at four temperatures ranging from 17 to 30 °C and four wavelengths. To do so, the high precision polarimeter of the PTB had to be equipped with a new detector. Moreover, appropriate light sources for the NIR wavelengths had to be found, which was not easy, as a wavelength accuracy of better than 0.01 nm not can be achieved with commercially available light sources. These were custom build by special work groups at the PTB.

The results confirmed preliminary measurements presented by W. Altenburg, C.C. Chou (1991) and N.H. Paton, M.R. Player, R.M. Urquhart, M. Dong (1993). In a NIR polarimeter, calibrated using the extrapolated Bünnagel’s formulae, the reading of a pure sucrose solution is too low by 0.06%. New formulae for both quartz and sucrose have been calculated.

The effect of temperature on ORD measurements in the NIR is similar to that experienced at visible wavelengths, so for practical use existing temperature corrections apply equally for polarisations at NIR wavelengths.

The detailed results will be presented at the ICUMSA meeting in Berlin 1998 in a paper by A. Emmerich, J. Keitel, M. Mösche and W. Seiler as a part of the Referees report.

The work on this NIR project was closely supervised by Dr. Emmerich despite the grave illness which led to his death in August 1997. He worked to see its completion only weeks before he died thus fulfilling a life’s ambition.


Altenburg, W.; Chou, C.C.: Zuckerind 116 (1991) 1041–1046

Paton, N.H.; Player, M.R.; Urquhart, R.M.; Duong, M.: Zuckerind. 118 (1993) 705–709

Progress Report for Subject 3:

Method format, collaborative testing and statistical treatment of data

by Mary An Godshall (USA), Referee

During this session, quite a few collaborative studies were undertaken, These tests have provided valuable information about performance of the methods, even when the outcome was not as good as hoped for. Although the results of these tests will not be finalized until the reports are presented in Berlin, the tests are listed here with some preliminary commentary.

Subject GS2. Referee Geoff Parkin has conducted a test on the revised Knight & Allen method for reducing sugars in white sugar

Subject GS3. Referee Malcolm Burge has proposed several tests for speciality sugars, to fulfil the 1994 recommendations and to coincide with finalization of the Codex Standard. These include enzymatic analysis of sulfite, colour in brown sugars using different filters, and an assay of anticaking agents by a gravimetric method and a polarimetric method.

Subject GS8. Several methods pertaining to beet sugar processing were tested, directed by Referee Jean-Pierre Lescure and the French National Committee. A preliminary screen with five laboratories and six samples tested enzymatic methods for glucose, fructose, D- and L-lactic acids, acetic acid and raffinose in several matrices. The impetus for this test is the need for fast results during processing. The results were so encouraging that the Referee hopes to organise a full-scale collaborative study by the end of 1997. Other tests included determination of calcium in beet processing samples by EDTA titration (modification of existing method) and determination of dry substance in beet pulp (a full-scale international collaborative test).

Subject 2. Referee Klaus Thielecke conducted two collaborative tests for raffinose in molasses, one enzymatic and the other HPAEC.

Subject 5. Determination of water content in beet molasses, cane molasses and raw sugar by Karl Fischer method was conducted by Guiseppe Vaccari. The method performed well for cane and beet molasses with high water content (>10%), but raw sugar, with a lower water content (<1.5%), demonstrated poor precision.

Subject 7. A series of screening tests on colour and turbidity in white sugars was conducted by Richard Riffer. One test examined the feasibility of using NTU to determine turbidity in white sugar. Another test looked at the effect of three different membrane types on ICUMSA colour results. The latter test falls into the category of ruggedness testing of the Official colour method.

Subject 8. A test on determination of glucose and fructose in raw and white sugars by HPAEC was conducted by Kevin Schäffler.

Subject 9. Referee Paul Bourlet has conducted a preliminary test for the determination of betaine in molasses by HPLC. The Referee points out that the test was conducted for technical purposes, as betaine is not at present a required commercial test for molasses.

Subject 15. Collaborative tests for determination of reducing sugars in raw sugar by the Lane & Eynon, Luff-Schoorl and Berlin Institute methods and the Modified Ofner methods for white sugars were conducted by Mel Carter. The Lane & Eynon methods had very high variability and the values tended to be considerably higher than those obtained with the other methods. The reasons for these wide discrepancies should engender discussion. Laboratories showed a 3 to 5-fold range in values for the same samples. At the same time, most laboratories were able to duplicate themselves (with blind duplicates) quite well, indicating that the Lane & Eynon method is highly operator dependent. The two other raw sugar methods showed adequate performance, while the Ofner method for white sugar showed a concentration dependence.

Subject 16. A collaborative test to compare the official ICUMSA method for ash using two incinerations with two procedures using single incineration, the Australian and the WAP methods, were conducted by Jean-Pierre Ducatillon. The three methods performed very well.

In at least two instances, the Referees conducted tests on several methods at once, allowing possibility of determining the equivalence of methods. Determining the equivalence of methods becomes relevant in view of the recommendation by Associate Referee Roger Wood to Codex that instead of prescribing specific methods of analysis that method characteristics instead be prescribed. Presumable this means that more than one method would be acceptable and knowing whether or not methods are equivalent should be helpful in this regard. This is a new item of interest within Subject 3, and I hope to have more information about it in my final report at the Berlin Session.

A continuing difficulty experienced by Referees is obtaining sufficient laboratories as collaborators. Perhaps it would be more accurate to say that receiving results back is as much a part of the difficulty as anything. The success with which collaborators have conducted their tests does show that it is possible to remain within the guidelines accepted by ICUMSA/IUPAC/ISO/AOAC for conducting collaborative studies and is a credit to their persistence and commitment. The guidelines are precisely that, and a rigid, invariant adherence to them serve no one well, as each test must be evaluated on its own merits. Nevertheless, this Referee feels that every attempt should be made to remain within the guidelines.

A perennial question is how to evaluate the results. Two helpful ratios are available to the Referee. These include the Horwitz ratio, a simple calculation which can be used to evaluate the reproducibility relative standard deviation (RSDR) obtained in the test. It allows to calculate the expected RSDR for a given concentration of analyte in the sample. Another helpful ratio, also provided by Horwitz, is that of repeatability to the reproducibility (r/R). The expected ratio is around 0.5; a much lower ratio indicates that individual analysts can repeat themselves well, but not from lab to lab. This could mean the test has not been written well, and analysts have introduced variations into the method. It can also mean that the method is highly operator dependent. I have noted in ICUMSA collaborative tests, the r/R ratio for adequately performing tests tends to be around 0.4.

Safety issues

It has become apparent that safety in all aspects of laboratory life has become a huge issue. All the accrediting organizations (IUPAC, ISO, AOAC) now stress that all safety issues in a method should be addressed in the method write-up. Some have gone so far as to recommend that disposal methods for waste generated in the test should also be included, although that has not yet been accepted. It is recommended that all future ICUMSA method write-ups include a section on safety for applicable items, and that these be added whenever a method is re-issued, or the Method Book is updated.

Outlier treatment

In 1994 IUPAC proposed a revision to the “Harmonized Guidelines” for collaborative studies in treatment of outliers. It now recommends that outlier’s probabilities be changed from 1 to 2.5%. The new statistics discs from AOAC for evaluating collaborative study data now reflect this change, and critical tables are available. IUPAC and AOAC have slightly different protocols for evaluating outliers, but in the main, they give similar results, especially if the outliers are large.

Workshop for Referees proposed

When Referees undertake to conduct a collaborative test, they take on a large responsibility, and do not always have very much in the way of guidelines or assistance. Sometimes, after the fact, a problem is noted that might have been prevented or would have made the test more powerful, if a small change had been made in the planning stage. There is a need to organize the various guidelines in one place, and a proposal is made to produce a booklet that can serve as a guide for referees planning a collaborative test. Also, a workshop on conducting collaborative tests is proposed for the ICUMSA meeting in Berlin. The booklet and the seminar would emphasize the importance of adhering to the protocol, especially with regard to blind duplicates or split-level samples, rather than known duplicates, and other important issues. The proposed booklet could contain information on the types of methods recognized by ICUMSA, the collaborative testing procedure, how to do the statistics, how to interpret the statistics, evaluating the various precision parameters, the importance of safety information, the method write up, the responsibilities of a Referee and the responsibilities of Associate Referees. This workshop could also be a forum for the many issues that are arising, for example, laboratory proficiency testing and the new Codex recommendation to prescribe method characteristics instead of specific methods.

Editor: Rud Frik Madsen, Strandpromenaden 38, DK-4900 Nakskov, Denmark – Fax +4554951675; Tel. +4553921675; E-mail