R013 Determination of dissolved solids (Brix) - A comparison of methods based on refractometers and density meters

Transcription

1 Bulletin No: Title: R013 Determination of dissolved solids (Brix) - A comparison of methods based on refractometers and density meters This technical bulletin is based on a presentation given by Jeff Pedley of B+S at a soft drinks technical conference in St Louis, USA in Some of the comments and discussion relate to particular instruments in use at that time. Introduction High precision measurement of dissolved solids (Brix) in soft drinks and juices is an important element of quality control (QC) and quality assurance (QA). There are in principle many ways to measure dissolved solids, but refractometry and density-based methods are most commonly used for this purpose. Whichever method is selected, there are advantages, disadvantages and limitations. This paper addresses important differences between the methods. The basic principles are outlined and practical aspects of applying the methods are discussed with reference to modern digital instruments. Reference is made to the RFM 340 digital automatic refractometer from Bellingham & Stanley and the models DMA 48 and DMA 58 digital density meters from Paar. Before looking at instrument performance it is worth reviewing briefly some basic concepts: the Brix scale and how it is used and important elements of quality control/quality assurance. Some of these notes are not intended to be conclusive so much as a stimulus for further thoughts on instrument performance as it pertains to particular applications. Brix Scale The Brix scale is based on solutions of pure sucrose in water, with concentration expressed as weight %. Accurate measurements at 20 C of density and refractive index (at nm) of sucrose solutions and the relationships to concentration have been adopted by ICUMSA and are used throughout the world for measuring the dissolved solids (mainly sugars) of liquid food products. Food products can contain many types of sugar and other soluble components such as acids, salts, colour and flavour agents. However, the Brix scale is used regardless of the composition of the aqueous phase. Strictly speaking, for non-sucrose based products, measurements should perhaps be expressed as apparent Brix or sucrose equivalents. Refractometers and density meters measure refractive index and density respectively. Each of these fundamental properties can be converted to a Brix value using well established empirical relationships such as those approved by ICUMSA. For pure sucrose solutions, the relationship between density and refractive index is fundamentally meaningful. However, for non-sucrose based products, there is no straightforward relationship between the two properties and a density based method will not give the same Brix as a refractometry method. This is because different soluble substances affect density and RI to different degrees. This can be illustrated by the following simple example. Bellingham + Stanley Ltd/Technical Bulletin No: R013 1

2 Let us assume that we have a hypothetical drink that contains % wt of sugar supplied in the form of high fructose corn syrup (HFCS * ) and water with no other components. At 20 C this liquid will have a refractive index of Using the ICUMSA RI/Brix conversion table, a Brix value of 9.87 % is obtained. The same drink will have a density at 20 C of kg m -3. Using the ICUMSA Brix/density relationship, a Brix value of % is obtained - a difference of 0.19%. Which value is correct? Neither, or at least they may both be regarded as equally incorrect. Of course, either value is perfectly acceptable for the purpose of QC/QA, provided the method and target value is clearly specified and used consistently. A change of method will clearly necessitate a change in specification. The only exception is for pure sucrose solutions. Quality Control (QC) and Quality Assurance (QA) Reliable Brix measurements are needed during preparative stages such as component dosing or concentrate dilution operations (QC) and final product compliance testing (QA). Incorrect or erratic measurements are often blamed on the instrument. However, there are potentially a number of sources of error: - sample type - colour, homogeneity, viscosity can affect measurements - pulps and syrups are generally more problematic than finished drinks - sample preparation - degassing and/or filtration may be important - experimental protocol - instrument cleaning routine, sampling techniques, time delays for sample equilibration must be defined - instrument performance - adequate stability and accuracy/precision are limiting and last, but certainly not least: - human interaction - care and strict adherence to protocol is necessary. The instrument is important but is only one component in the chain. The quality (reliability) of the measurement will only be as good as the weakest component in the chain. Repeatability and Reproducibility Accuracy is a term often used in a generic fashion to mean different things. It is wrong to use accuracy for Brix measurements without some reference to the method or instrument used (NB density vs RI) and perhaps the product composition; for example the Brix will change if the sugar type/source is changed. The true accuracy of an instrument can be verified using appropriate calibration standards such as pure, freshly prepared, sucrose solutions or oils (RI scale). Having established that an instrument can measure accurately, there is a further requirement for repeatability (or precision) and reproducibility; both essential elements of reliable QA and QC methods. * HFCS - fructose 90.8%, dextrose 7.7%, maltose 1.2%, trace components 0.3% Bellingham + Stanley Ltd/Technical Bulletin No: R013 2

3 Repeatability may be defined as the difference between successive test results for identical test material obtained by the same operator with the same apparatus under constant operating conditions. Reproducibility is the difference between independent results for identical test material obtained by different operators working in different laboratories but with the same apparatus. Both definitions are of course subject to a further statistical constraint. A QA method may specify a requirement for an instrument with a repeatability to ±0.03 Brix. However, systematic errors between different laboratories could yield the following results for the same product: ±0.03 (lab 1) ±0.03 (lab 2) In this case repeatability is within specification but reproducibility is not acceptable. There could be a number of reasons for this discrepancy: Factors that affect repeatability (random errors), e.g. - sample stability - instrument stability (working at its limit) - inadequate temperature control - operator care Factors that affect reproducibility (systematic errors), e.g. - experimental protocol - instrument type or make - instrument set-up, e.g. incorrect calibration The resolution of an instrument is another aspect to consider. High resolution (e.g. two or three decimal places in Brix) does not necessarily mean equivalent accuracy and says nothing about repeatability. It simply specifies the form of the display. An erratic display could mean the instrument is unstable and perhaps unworthy of its resolution. It could, of course, mean the sample is unstable and the fluctuations are genuine. Conversely, a stable display could mean the instrument is insensitive! It is important to distinguish these effects. When deciding on a particular instrument type, it is clearly important to consider all aspects of the measurement operation. It may be that ease of use is more important than high resolution and certain samples may not justify a high resolution/high repeatability instrument. Sampling techniques and sample preparation may be the controlling factors in deciding the overall error in a method. A high resolution instrument may simply prove to be too sensitive. An instrument that is ideal for one type of sample may be totally unsuitable for another. In the light of the concepts and principles described above, we can now consider instruments and their performance. Refractometers and density meters - basic principles Dissolved solids affect both density and refractive index (RI) of a liquid. Thus when the concentration of sugar in water is increased, the density and RI increase. In this way density or RI can be used to determine solute concentration in a binary system. Refractometers and density meters are designed according to well-characterised fundamental physical phenomena. Bellingham + Stanley Ltd/Technical Bulletin No: R013 3

4 Refractometers measure refractive index of a material; most practical applications are with liquid samples. Most refractometer designs are based on the critical angle concept. A critical angle condition is established where an incident beam of known wavelength is totally internally reflected from a prism/sample interface. From the geometry of the optical path and the known RI of the prism material, the RI of the sample is determined. Modern automatic digital refractometers such as the RFM 340 from Bellingham & Stanley, utilise a photodiode array detection system to determine with high precision the position of the borderline between refracted and internally reflected light. A great advantage of this design is that there are no moving parts that can be subject to mechanical wear. Early digital refractometers were based on motorised designs. Modern digital density meters such as the DMA series from Paar, utilise the effect of density on the natural resonant frequency of an oscillating U-tube filled with the liquid under test. A simple equation relates the resonant frequency to the mass, volume and density of the test liquid. Modern digital refractometers and density meters are equipped with software that performs the necessary conversion of the measured quantity to a value on a related scale such as Brix. Many other scales can be programmed, for example glucose, fructose, invert sugar, sodium chloride and many more. There is no limit to the types of data processing possible other than by the memory limitations of the instrument. An alternative that is becoming of increasing interest is to transmit the data via an RS 232 interface to a PC or LIMS system. RI and density change with temperature and so good temperature control is an important requirement for high precision measurements with both density meters and refractometers. However, temperature control is more critical in the case of density measurement. The figure below shows an example for 10% sucrose. Achieving adequate thermal equilibrium in the measuring chamber can make density measurements rather slow. Automatic temperature compensation (TC) is available as a feature on most instruments, although this is generally based on Brix (sucrose data). For non-sucrose, sugar-based products the TC will still be a good approximation, although errors are then likely to be greater with density measurements because of the higher temperature coefficients. Many laboratories prefer to use TC rather than employ temperature control. However, high resolution refractometers such as the RFM 340 (resolution to 0.01 Brix) perform better with applied temperature control - thermostat/circulator. This is because the stable heat sink provided by the water serves not only to control the temperature of the prism plate assembly, but also to stabilise the internal environment of the instrument against external temperature fluctuations (heat transfer from surroundings or hot/cold samples). This enhanced stability enables the instrument to hold its calibration more reliably and reduces the need for frequent zeroing of the instrument. Density meters such as the DMA 48 and 58 models have the advantage of built-in high precision temperature control. The higher resolution DMA 58 is proportionally more dependent on good temperature control and this does reflect somewhat in the significantly higher cost of this instrument. Bellingham + Stanley Ltd/Technical Bulletin No: R013 4

5 Effect of Temperature on Refractive Index and Density of 10% Sucrose Solution (Relative to values at 20 deg C) % change relative to 20 degc Temperature (deg C) density rel to 20 C RI rel to 20 C Comparison of a digital refractometer with a digital density meter What should we look for in a good instrument? - repeatability (precision) - reproducibility - simple operation - speed - easy cleaning - simple calibration - friendly software - set-up security - durability - minimal maintenance - flexible data handling/recording - flexibility (multi-mode) - possibly, size - small footprint? - good supplier support And, of course, all at a low price! Another factor perhaps is familiarity. Companies have a tendency to invest in a particular technique and are then naturally reluctant to change. Uniformity of instrumentation throughout an organisation may be important, particularly if inter-laboratory cross referencing of product quality is required. Digital density meters appeared on the market in the seventies before digital refractometers. A number of companies therefore opted for density measurement because of this advance in technology. Bellingham + Stanley Ltd/Technical Bulletin No: R013 5

6 A choice between refractometry and densimetry is not simply a case of selecting the highest possible resolution/repeatability; other factors must be considered. The following table lists the important differences (and similarities) between an RFM 340 refractometer and DMA 48/58 density meters. Feature Refractometer Density Meter Sample type Any fluid Limited by viscosity and homogeneity Sample application Direct to prism or flow through Flow through Measurement time 6-30 seconds minutes Sample degassing Can be important Essential Temperature equilibration 6-30 seconds minutes Calibration time 1-2 minutes 10 minutes Resolution RI 0.01 Brix g cm -3 (DMA 48) g cm -3 (DMA 58) Repeatability (Brix) Samples 0-20 %Brix Repeatability (Brix) Syrups ± 0.02 ± 0.01 (DMA 58) ± (DMA 48) ±0.03 to 0.1 ± 0.1 (DMA 48) Cleaning Easy wash/wipe Difficult, ml sample flush Weight 6.2 kg 25 kg Servicing/maintenance No moving parts Instrument validation only Mechanical wear possible Annual check recommended Power 12 v remote Mains 220/110 v Fifteen to twenty years of practical experience with digital instruments has shown that, when used correctly, both refractometers and density meters will provide reliable measurements for QC/QA in the drinks industry. Instruments of both type are capable of the same sort of resolution/precision (repeatability) - for practical purposes, approximately ±0.01 to 0.03 Brix, depending upon sample type and measuring range. A density meter working on well behaved samples will repeat to perhaps ±0.005 Brix, but this cannot be regarded as a general specification. Similarly, a differential refractometer (different working principle to an RFM 340) working in a narrow Brix range on some products could perhaps achieve a repeatability better than ±0.01 Brix. Refractometers are more tolerant to sample type. Degassing may be important for carbonated drinks if bubbles have a tendency to accumulate on the prism surface but for Bellingham + Stanley Ltd/Technical Bulletin No: R013 6

7 density meters the absence of bubbles is essential. Inhomogeneous samples such as juices containing residual particles cannot be measured reliably with density meters. The same samples can be measured on a refractometer but repeatability may be dependent on a reliable sampling and measurement protocol. The RFM 340 has a special feature which gives the user a numerical measure of sample quality (a measure of the difficulty in discerning the critical angle borderline). This parameter can be a useful tool for testing difficult samples. Effective cleaning between measurements is essential for both methods to ensure maximum performance. Refractometers such as the RFM 340 with a readily accessible prism surface are easier to clean than flow through cells/density tubes. This is particularly so with viscous fluids that might require excessive flushing with clean water or solvent. Also, drying out of a sugar solution can leave behind films/deposits that cause large errors on re-solution. Cleaning can be a time consuming operation. The DMA density meters have the attractive feature of built-in heating and temperature control but thermal equilibration can be slow because of high dependence of density on temperature. Water circulation, although regarded as a clumsy method for temperature control, does provide a refractometer with a large heat sink that quickly absorbs heat transferred from the relatively small sample volume and maintains the instrument in a stable condition, thereby minimising calibration drift. Syrups are not so easily measured with either instrument type. Near saturation concentrations of sugar affect water structure and hence RI/density. High concentration syrups are therefore often inherently unstable and Brix repeatability is relatively poor. Refractometers are nevertheless easier to use with syrups; the high viscosity makes sampling and cleaning very slow with density meters. Partially inverted sugar is a problem and cannot be quantified by either refractometer or density meter alone. Sucrose is a chemically unstable disaccharide which hydrolyses to form a molecule of glucose and a molecule of fructose. In doing so, a molecule of water is absorbed thereby increasing both the density and refractive index of the solution. C 12 H 22 O 11 + H 2 0 C 6 H 12 O 6 + C 6 H 12 O 6 Sucrose Glucose Fructose The reaction is catalysed by acid and so sucrose-based drinks containing, for example, phosphoric or citric acids will undergo spontaneous inversion. If a drink is partially inverted during QA testing the apparent Brix value obtained with either a density meter or refractometer will be unreliable, being dependant upon sample history (ambient temperature, time of standing, ph etc). To obtain an unambiguous result, partial inverts can be dealt with in one of two ways: either by forcing complete inversion (heat, added acid) and measuring 100% invert sugar or by using a separate technique such as polarimetry to complement the refractometer/density measurement. Putting aside preference based on familiarity, and given the types of instruments currently available, choice will probably become a compromise between desired precision and the consequent constraints and cost this imposes on the overall measurement procedure. Time consumption and ease of use can be paramount and it is important to match the instrument capability with a realistic expectation of what can actually be achieved with a particular product and process line. And, of course, instrument price should equate with the expected cost savings through improved operating efficiency and/or tighter process control. For the majority of applications, the important differences between density meters and refractometers centre primarily around ease of use and speed of measurement. Higher Bellingham + Stanley Ltd/Technical Bulletin No: R013 7

8 precision density meters may give marginal improvements in resolution and repeatability for certain well behaved samples, but at a considerably higher price. Some thoughts for future development There was a time when manufacturers dictated instrument designs based on their perceived understanding of industrial requirements. In turn, presumably, industries developed quality tests based around the capability of available instrumentation. This is no longer the case. Nowadays, the customer demands the specification and hopefully the instrument manufacturer complies. In future, instrument manufacturers will have to be prepared for a greater degree of flexibility and customer orientation. Improvements in software and instrument communication interfaces in the last two decades have enabled many possibilities for data manipulation and recording and the scope is endless. This aspect of instrumentation will inevitably develop rapidly as companies perfect their quality management systems and inflict greater demands on the data management capabilities of instruments and associated computer systems. Vast amounts of product information storage, data correlation and statistical analysis will be possible thus enabling a greater degree of automated quality control. Hardware changes are a different matter. The current generation of automatic digital refractometers such as the RFM340 have been designed for benchtop applications with multiple use ranging from drinks to fruit pulps, sauces, jams, syrups, not to mention the many applications outside the food industry, in particular chemical, oil and pharmaceutical industries. If higher and higher measurement precision is a realistic scenario, given the varied nature of the candidate products and processes, then a change from this standard instrument type is likely. It may be possible to improve measurement of Brix by, say, an order of magnitude in precision for a finished drink, but a different instrument, operating on a different measuring range would then have to be prescribed for a concentrate or syrup. It may be necessary to design instruments specifically for one customer and maybe just one product, assuming this is a cost effective solution for both manufacturer and customer. The other likely trend will be towards greater process control, thereby relying more and more upon in-line monitoring rather than the slower process of turn around of data from the quality laboratory. There is considerable scope for co-operation between the drink producer and the instrument manufacturer in this respect to design the ideal measuring system with automatic data capture to an integrated process control information system. Such systems are already in use but more widespread progression towards this approach seems inevitable. One thing is for sure, instrument manufacturers will have to work more closely with customers than ever before. Bellingham + Stanley Ltd/Technical Bulletin No: R013 8