SEPTEMBER, 1974 Vol. 99, NO. 1182. THE ANALYST The Rapid Determination of the Total Dry Extract of Wines BY P. J. WAGSTAFFE (De9avlment of Trade and Industry, Laboratory of the Government Chemist, Cornwall House, Stamford Street, London, SE1 9NQ) A method is described whereby the total dry extract of wines, including sweetened and fortified wines, can be determined from knowledge of their relative densities and alcoholic strengths. The method is designed to be used in conjunction with the rapid refractometric method used in this laboratory for the determination of the alcoholic strength of wine and requires no further measurements, other than that of temperature, to be made. THE total dry extract of wine is usually determined by a process involving the removal of volatile material and, after suitable readjustment of volume, measurement of the relative density of the resulting solution.The relative density thus found is expressed in terms of the equivalent concentration of sucrose by reference to tables. Generally, this operation is effected on the residual solution resulting from distillation of the wine when determining its alcoholic strength. The total dry extract found by this procedure differs only slightly from that found by the more accurate method of evaporating a known volume of wine and weighing the residue. In fact, the densimetric procedure is the official method of the European Economic; Community (EEC).l To distil all wines in order to determine their total dry extract could be inconvenient,' particularly as the alcoholic strength of most of them can be determined by a rapid refracto- metric method2 with an accuracy sufficient for many purposes.Several workers394 have attempted to devise methods in which distillation is avoided, most of them involving measure- ment of refractive index or relative density, or both. Unfortunately, none of these methods satisfy our requirements as regards accuracy, compactness of tables and, most important, the range of composition of wines with which they can cope. The present work describes the development and use of tables that overcome these difficulties. EXPERIMENTAL A large number of aqueous ethanoIic sucrose solutions of accurately known composition were prepared covering, in uniform increments, the range 0 to 350 g 1-1 of sucrose and 0 to 25 per cent.VjV of ethanol. Their relative densities, D, were measured by pycnometry and their refractive indices, R, by means of immersion refractometers calibrated in the arbitrary Zeiss scale, all measurements being made at 20 "C. The results of these determinations were plotted in various ways as shown in Figs. 1, 2 and 3. Each of these relationships was expressed mathematically by means of a computerised regression analysis, which fitted third-order equations to the experimental data, each equation expressing sucrose concentration as a function of the ethanol concentration and ( R + D), R or D. The situation was complicated to some extent €or the relationships involving refractive index. In order to cover the whole range of compositions it is necessary to use two refracto- meter prisms, each covering a different range of refractive indices.Unfortunately, a graph of refractometric scale readings versus true refractive indices exhibits a discontinuity where the two scales overlap, which is revealed as the breaks in the lines representing 150 and 200 g 1-1 in Figs. 1 and 2. It was therefore necessary to derive two equations for each of the relation- ships involving refractive index, one relating to the range covered by the No. 1 prism ai;d the other to that covered by the No. 2 prism. The suitability of these equations for determining the total dry extract was assessed by substituting into each the appropriate results obtained by analysis of a series of table wines. By comparing the values of dry extract thus found with those obtained by the densimetric method it was demonstrated that the equation relating dry extract to alcoholic strength and 537 0 SAC; Crown Copyright Reserved.538 40 WAGSTAFFE : THE RAPID DETERMINATION 0 - - No.1 I I I I [Analyst, VOl.99 Sucrose concentratiodg 1-1 240 2 200 - m > 160 120 80 No. 2 Ethanol, per cent. V/V Fig. 1. Variation of the sum R + D with concentration of ethanol at various sucrose concentrations relative density gave by far the best results, the errors obtained having the lowest mean-value and a distribution whose standard deviation was less than one third of those resulting from the use of the relationships involving refractive index. In view of this finding and because of the previously mentioned complications associated with immersion refractometers, it was decided to derive tables that relate dry extract to alcoholic strength and relative density.20 0 Sucrose 1 concentration/g 1-1 Q, > 1 120- VJ e, rr .- N cy 8 0 - No. 1 Y 40 - NO. 1 I I I I I 1 5 10 15 20 25 0 Ethanol, per cent. V/V Fig. 2. Variation of the refractive index (Zeiss scale) with concentration of ethanol a t various sucrose concentrations The equation relating sucrose concentration to the concentration of ethanol and relative density of aqueous ethanolic sucrose solutions is given as it provides in itself a means of determining dry extract (see below).September, 19741 OF THE TOTAL DRY EXTRACT OF WINES 539 Sucrose (g 1-l) = (399.443A x + (259.3490 x lo-?) - (642*503Ae x + (1211.760A3 x + (1.297n3 i< - (18.005AD x + (148.071A2D2 x lo-') -!- (488*320A30 x lo-') - (378*544A3O3 x lo-'') .. (1) where and A = ethanol concentration, per cent. V/V I ) == 1000 (relative density, 20 OC/ZO "C) - 1000. DERIVATION OF TABLES FOR THE DETERMINATION OF TOTAL DRY EXTRACT FROM KNOWLEDGE OF THE RELATIVE DENSITY AND ALCOHOLIC STRENGTH OF WINE- The principal difficulty in producing tables of the type required lies in reaching a com- promise between accuracy and compactness. This compromise was finally achieved as follows: Fig. 3 demonstrates the decrease in relative density that occurs as the concentration of ethanol in an aqueous ethanolic sucrose solution is increased. It can be shown that the lines representing solutions of equal sucrose concentration become increasingly parallel to each other as the sucrose concentration increases.Now, in any given aqueous ethanolic sucrose solution the relative density of the corresponding ethanol-free solution can be considered to be the relative density of the mixture increased by an amount dependent on the concentration of ethanol. If all the constant sucrose lines were entirely evenly spaced, as measured in a direction parallel to the D axis, the influence of the ethanol on the relative density would be independent of the actual sucrose concentration ; therefore, provided the relative density and the concentration of ethanol in the mixture were known it would be possible to deduce the relative density of the corresponding ethanol-free solution. This would be, in effect, a more precise form of the Tabarie e q ~ a t i o n , ~ but allowing for contraction in volume when sucrose solutions and ethanol solutions are mixed together. 0 5 10 15 20 25 Ethanol, per cent.V/V Fig. 3. Variation of relative density with con- centration of ethanol a t various sucrose concentrations As, in fact, the lines representing equal sucrose concentrations are not completely parallel, a mathematical analysis was undertaken to determine the line whose slope best represented the slopes of all the constant sucrose lines represented in Fig. 3. The equation representing the line thus derived is DA = -1*149012A + 0.02059A2 - 0*00042A3 . . . . . . . , (2) Equation (2) was used to calculate the correction terms DA corresponding to unit increments of ethanol concentration over the range 0 to 25 per cent, V/V.Thus, in any given aqueous540 WAGSTAFFE : THE RAPID DETERMINATION [Analyst, VOl. 99 ethanolic sucrose solution the relative density of the corresponding ethanol-free solution can be determined by adding to the relative density of the mixture an amount DA, determined from equation (2), dependent on the concentration of ethanol. In order to avoid the necessity of converting relative densities thus determined into their corresponding sucrose concentrations it was decided to express all relative densities directly in terms of grams of sucrose per litre of solution. This was achieved by use of Savage's table,5 which relates the relative density, 20 "C/20 O C , to grams of sucrose per litre of solution at 20 "C. Fortunately, this relationship between relative density and sucrose concentration is virtually linear over the range of interest, hence the sucrose concentrations obtained by conversion of the relative densities can be assumed to be additive.TABLE I CORRECTION TERMS (g 1-1 OF SUCROSE) CORRESPONDING TO CONCENTRATION OF ETHANOL, Ethanol, per cent. v/v 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0 12.0 13.0 14.0 15.0 16.0 17.0 18.0 19.0 20.0 21.0 22.0 23.0 24.0 25.0 7 0.0 3.8 7.4 11.0 14.6 18.0 21.3 24.6 27.9 31.1 34.2 37.3 40.3 43.4 46.4 494 52.4 55.4 58.4 61.4 64.5 67-5 70-6 73.7 76.9 80.1 The ethanol correction 0.1 0.4 4.2 7.8 11.4 14.9 18-3 21.6 24.9 28.2 31.4 34-5 37.6 40.6 43.7 46.7 49.7 52.7 55.7 58.7 61.7 64.8 67.8 70.9 74.0 77.2 0.2 0.8 4.5 8.1 11.7 15.3 18.7 22.0 25.3 28.5 31.7 34.8 37.9 40.9 44-0 47.0 50-0 53.0 56-0 59.0 62.0 65.1 68.1 71.2 74.3 77.5 PER CENT.v,/v, AT 20 "c Intermediate ethanol concentrations, per cent. V / V - 0.3 1-1 4.9 8.5 12.1 15.6 19.0 22.3 25.6 28.9 32.0 35.1 38.2 41-2 44.3 47.3 50.3 53.3 56.3 59.3 62-3 65.4 68.4 71.5 74.7 77.9 0.4 1.5 5.2 8.8 12.4 16.0 19.3 22.6 25.9 29.2 32-3 35.4 38.5 11.5 44.6 47.6 50.6 53-6 56.6 59.6 62.6 65.7 68.7 71.8 75.0 78.2 0-5 1.9 5.6 9.2 12-8 16.3 19.7 22.9 26.3 29.5 32.7 35.7 38.8 41.9 44.9 37.9 50.9 53.9 56-9 59.9 62.9 66.0 69.1 72.2 75-3 78.5 0.6 2.3 6.0 9.6 13.2 16.6 20.0 23.3 26.6 29.8 33.0 36.1 39-1 42.2 45.2 48.2 51.2 54.2 57.2 60.2 63.3 66.3 69.4 72.5 75.6 78.8 0.7 2-7 6.3 9.9 13.5 17.0 20.3 23.6 26.9 30.1 33.3 36.4 39-4 42.5 45.5 48.5 51-5 54.5 57.5 60.5 63.6 66.6 69.7 72.8 75-9 79.1 terms derived from equation 0.8 3.0 6.7 10.3 13-9 17.3 20.6 23.9 27.2 30-5 33.6 36.7 39.7 42.8 45.8 48.8 51-8 54.8 57-8 69.8 63.9 66.9 70.0 73.1 76.3 79.5 0.9 3-4 7.0 10.6 14.2 17.7 21.0 24-3 27-6 30.8 33.9 37.0 40.0 43.1 46.1 49.1 52.1 55.1 58.1 61.1 64.2 67.2 70.2 73.4 76.6 79.8 (2) for unit increments of ethanol concentration were converted into the corresponhing sucrose concentration, intermediate values corresponding to tenths of a per cent.VjV being obtained by linear interpolation and the results expressed as in Table I. Table I1 is a rearranged version of Savage's table and gives the apparent sucrose concentrations corresponding to the relative density of the mixture (in the tables used in practice, entries of D are given to 0.1 unit). Hence the sucrose con- centration of an aqueous ethanolic sucrose solution can be determined by adding, algebraically, the value found in Table I corresponding to the ethanol concentration to the value found in Table TI corresponding to the relative density of the mixture.TABLE I1 VALUES FOR SUCROSE CORRESPONDING TO Dzo oc/30 oc D Sucroselg 1-1 D Sucroselg 1-1 D Sucrose/g 1-1 D Sucrose/g 1-1 D Sucrose/g 1-1 970 -77.7 1005 12.9 1040 103.7 1075 195-1 1110 287.2 975 -64.7 1010 25.8 1045 116.7 1080 208.3 1115 300.5 980 -51.7 1015 38.7 1050 129.8 1085 221.3 1120 313.7 985 -38.7 1020 51.7 1055 142.8 1070 234.5 1125 326.9 990 -25.8 1025 64.7 1060 115.9 1095 247.5 1130 340.2 995 -12.9 1030 77.8 1065 169.0 1100 260.9 1000 0.0 1035 90.7 1070 182.0 1105 274.0 In the table used in practice, entries of D are given to 0.1 unit.September, 19741 OF THE TOTAL DRY EXTRACT OF WINES 541 The usefulness of these tables was examined critically.Table I11 illustrates the errors result- ing from applying the tables to the data obtained from measurement of aqueous ethanolic sucrose solutions of known composition. TABLE I11 ERRORS (gl-1 OF SUCROSE) PRODUCED BY APPLICATION OF TABLES I AND I1 TO AQUEOUS ETHANOLIC SUCROSE SOLUTIONS OF KNOWN COMPOSITION Ethanol concentration, per cent. V / V Sucrose/ I A \ g 1-1 0 2.0 5.0 10.0 15.0 20.0 25.0 20 50 100 150 200 250 300 350 +0*1 +0*1 + 0.1 +0.1 + 0.2 0.0 + 0.2 0.0 - 0.5 - 0.5 - 0.3 - 0.1 -0.1 +0*1 + 0.2 + 0.2 - 0.2 + 0.1 - 0.1 - 0.1 0.0 + 0.2 + 0.3 + 0-5 - 0.5 - 0.3 - 0.2 - 0-2 - 0.3 0.0 - 0.2 0.0 0.0 + 0.2 +0.1 + 0.2 - 0.2 - 0.2 - 0.6 - 0.9 + 0.8 + 0.5 + 0.4 0-0 - 0.5 - 1.4 - 1.8 - 2.5 +2.1 + 1.9 + 1.2 + 0.5 - 0.6 - 2.0 - 3.3 - 4.6 Error = determined value - true value.As can be seen from Table 111, the error increases with increase in both the ethanol and sucrose concentrations. The standard deviation of the error over the whole range of compo- sitions is 1.05. If the region exceeding 250 g 1-1 of sucrose and 15 per cent. VlV of ethanol is excluded the standard deviation is 0.55, giving approximate 95 per cent. confidence limits Tables I and I1 were further examined by applying them to a wide variety of wine samples. These samples included some table wines but emphasis was placed on sweetened and fortified wines. The alcoholic strength of each wine was determined by distillation and by the refractometric method, ie., by measurement of refractive index and relative density at 20 "C.The total dry extract was determined from Tables I and I1 employing the alcoholic strengths determined by both methods. In addition, the suitability of equation (1) for determining total dry extract was examined by substituting into it the appropriate values, again employing alcoholic strengths determined by distillation and by the refractometric method. The results were compared with the dry extract values obtained by the official EEC method and are summarised in Table IV together with results obtained in a similar way on a series of table wines. of * l a 1 g 1-1. TABLE I V SUMMARY OF ERRORS (g I-') I N DETERMINATIONS OF TOTAL DRY EXTRACT BY VARIOUS PROCEDURES Samples Procedure A 7 7 From table From equation r - - - .~ ___h---------q A Ethanol by Ethanol by Ethanol by Ethanol by refractometer distillation refractometer distillation I \ Sweet, sweetened and fortijied wines (68 samples)-- Mean error . . .. . . . . - 1.1 - 0.5 - 0.8 - 0.3 95 per cent. confidence limits . . & 2-6 1.4 * 2.0 * 1.2 Mean error . . . . . . . . - 0-7 - 0.6 - 0.6 - 0.2 95 per cent. confidence limits . . * 1.9 1.4 1.7 0.8 Error = determined value -- true value. Table wines (36 samples)- The results shown in Table IV are much as would be expected. An increase in error of the value for total dry extract results from the use of the tables instead of the more precise equation and, to a greater extent, from the use of the refractometric method to determine the ethanol content.Nevertheless, a 95 per cent. confidence limit of *2.6 g l-l, as found for the sweetened and fortified wines when using the tables in conjunction with the refractometric method, should prove to be adequate for most purposes.542 WAGSTAFFE : THE RAPID IIETERMIN~ITIC'F [A?znlyst, 1701. 99 INFLUENCE OF TEMPERATURE ON THE DETERMINATION OF TOTAL DRY EXTRACT- The rapid refractometric method for determining the ethanol content is, effectively, temperature independent; the only condition that needs to be satisfied is that measurement of both the refractive index and relative density are made at the same temperature. Unfortu- nately, the method described here for determining dry extract does not exhibit this temperature independence.The need to attemperate all samples to 20 "C in order to determine the extract could be inconvenient and would greatly restrict the usefulness of the refractometric method. For this reason, temperature correction tables were devised, which enable allowance to be made for the influence of temperature on the determination of dry extract. The relative densities D,oc/zuoc of a series of aqueous ethanolic sucrose solutions were measured at 15, 20 and 25 "C, no allowance being made for the thermal expansion of the borosilicate pycnometer used. The differences between the relative densities at 15 and 20 "C and between thosc at 20 and 25 "C were plotted against ethanol concentration at several different sucrose concentrations. The differences in relative density at suitable increments in ethanol concentration were read off from the graphs, expressed in terms of sucrose coii- centration (grams per litre) employing Savage's table, and the results set out in tabular form.The intermediate points between 15 and 20 "C and between 20 and 25 "C were found by interpolation, assuming linearity over 5 "C intervals. Table V shows part of the temperature correction table, the full version of which extends from 0 to 350 g 1-1 in 50 g 1-1 increments. The appropriate correction term is found from the known ethanol concentration and approximate extract value as determined from Tables I and 11. It is evident from Table V that the actual magnitude of the correction term is not strongly dependent on the dry extract level, thereby justifying the use of the approximate value of dry extract in order to determine the correction.TABLE V TEMPERATURE CORRECTION TERMS (g 1-l) TO RE ADDED TO OR SUBTRACTED FROM THE APPARENT EXTRACT FOUND FROM MEASUREMENT O F D AT TEMPERATURES OTHER THAN 20 "c Ethanol, Temperature/"C Apparent per extract/ cent. g 1-1 v/v 50 0 5 10 12.5 15 17.5 20 22.5 25 100 0 5 10 12.5 15 17.5 20 22.5 25 Subtract -- 7 15 2.4 2.5 2.9 3.2 3.5 4.0 4.4 4.9 5.5 2.7 2.8 3.2 3-5 3.9 4.3 4.8 5.3 5.8 16 1.9 2.0 2-3 2.6 2.8 3.2 3.5 3.9 4.4 2.2 2-2 2.6 2.8 3.1 3.4 3.8 4.2 4.6 17 1.4 1-5 1.7 1.9 2.1 2-4 2.6 2.9 3.3 1.6 1.7 1.9 2.1 2.3 2.6 2.9 3.2 3.5 18 1.0 1.0 1.2 1.3 1-4 1.6 1.8 2.0 2.2 1.1 1.1 1-3 1.4 1.6 1.7 1.9 2.1 2.3 19 0-5 0.5 0.6 0.6 0.7 0.8 0.9 1.0 1.1 0.5 0.6 0.6 0.7 0.8 0.9 1.0 1.1 1.2 Add A , - 21 0-6 0.6 0.7 0.8 0.8 0.9 1.0 1.1 1.2 0.7 0.7 0-8 0.8 0.9 1.0 1.0 1.2 1.3 22 1.2 1.3 1.4 1.6 1.7 1-8 2.0 2.2 2-4 1.4 1.4 1-5 1.7 1.8 2.0 2.1 2.4 2.6 23 1.9 1.9 2.1 2.3 2.5 2.8 3.1 3.3 3.7 2.0 2.1 2.3 2.5 2.7 2.9 3.2 3.5 3.8 24 2-5 2.6 2.8 3-1 3.4 3.7 4.1 4.4 4.9 2.7 2.8 3.0 3.4 3.6 3.9 4.2 4.7 5.1 25 3-1 3.2 3.5 3.9 4-2 4.6 5-1 5.5 6.1 3.4 3.5 3.8 4.2 4.5 4.9 5.3 5.9 6.4 The tables used in practice extend from 0 to 350 g 1-1 of sucrose in 50 g 1-1 increments.Little additional error results from measuring the relative density at ambient temperature and applying the temperature corrections. Although Table V was derived for use with a borosilicate pycnometer it does not matter if soda-glass instruments are used instead. These remarks apply equally to results obtained with glass hydrometers, which can be used if preferred.September, 19741 OF THE TOTAL DRY EXTRACT OF WINES 543 This paper has been published with the permission of the Government Chemist. The assistance of Mr. 0. F. Newman and Mr. I. Telford with the statistical analysis and computer programming is gratefully acknowledged. REFERENCES 1. 2. 3. 4. 5. “Recueil des M6thodes Internationales d’Analyse des Vins,” Office International de la Vigne et du Vin, Paris, 1958, Part A3, p. 5. Cooke, J. R., Analyst, 1974, 99, 306. Petro-Turza, M., and Kovacs-Klement, M., Mitt. Rebe Wein, 1971, 4, 289. Ribdreau-Gayon, J., and Peynaud, E., “Analyse et Controle des Vins,” Second Edition, Beranger, Savage, R. I., Int. Sug. .J., 1972, 74, 167. Paris, 1968, p. 62. Received A9ril l s t , 1974 Accepted April 17fh, 1974