[4] and the method of Rice and Boleracki [5]. In the Spencer method the drying is effected in a specially designed electric oven. The method is applicable to solid sugars as well as liquid sugar products. The samples are placed in aluminum capsules fitted with metallic gauze bottoms which permit free passage of air. When liquid or semiliquid products are to be dried, the sample is absorbed on asbestos [6]. Air, heated by passing over an electric heating element, is passed through the sample and the moist air is continuously withdrawn by

suction.

The Rice and Boleracki method [5] consists in spreading a very thin film of molasses or sirup on thin sheets of silver and drying in a vacuum oven at 70° C. In the hands of an experienced operator the method appears to yield concordant results and has the added advantage of rapidity. It is necessary to exercise some care in preparing the film for drying and also in the manipulation during drying and weighing. The authors recommend the method for the determination of moisture in such products as honey, invert sirup, and blackstrap molasses.

5. REFERENCES

[1] U. S. Customs Regulations (see p. 797 of this Circular).

[2] Official and Tentative Methods of Analysis of the Association of Official Agricultural Chemists, 5th ed. p. 484, 485 (1940).

[3] J. Assn. Official Agri. Chem. 21, 89 (1938).

[4] G. L. Spencer, J. Ind. Eng. Chem. 13, 70 (1921).

[5] E. W. Rice and P. Boleracki, Ind. Eng. Chem., Anal. Ed. 5, 11 (1933). [6] G. P. Meade, J. Ind. Eng. Chem. 13, 924 (1921).

XVI. DETERMINATION OF ASH

Although the determination of ash in sugars and sugar products is subject to considerable uncertainty, it is still widely used as an indication of the mineral constituents present. In routine analysis the socalled "sulfated-ash" method is employed, largely on account of its simplicity and reproducibility. In this method any volatile constituents, such as Cl, NO3, CO2, etc., are driven off and are replaced by SO. This replacement is considered advantageous in that it compensates for the losses. A further advantage is the conversion of volatile salts, such as KCl, into the nonvolatile sulfate. The method is as follows:

1. SULFATE METHOD

Weigh 2 to 5 g of the sample in a 50- to 100-ml platinum dish, add 0.5 ml of concentrated H2SO, or 1.0 ml of 1:1 H2SO4, heat gently on a hot plate until the sample is well carbonized, and then heat in a muffle furnace at a low red heat until all carbon is burned. Cool and add a few drops more of H2SO4, heat until this is fully volatilized, then cool in desicator and weigh. Reignite in the muffle furnace to constant weight. Express the result as the percentage of sulfated ash.

The general practice in many laboratories is to deduct one-tenth of the amount of sulfate ash to reduce to the normal ash. This deduction of 10 percent has been studied by many investigators and found to be in error for cane products. Jamison and Withrow [1] found that the value for sulfate ash in Cuban raw sugar, even with the customary 10-percent correction, was about 34 percent higher than the ash by

direct incineration. Ogilvie and Lindfield [2] found the correction factor to be from 12 to 15 percent for beet sugars and from 6 to 26 percent for cane sugars.

Jamison and Withrow proposed a modification of the sulfate method, in which they added 2 ml of sulfuric acid (2:1) to the sample of sugar, heated the sample on a hot plate until completely carbonized, and finally ignited it in the muffle furnace to a white ash. After cooling the ash, they added 3 or 4 drops of sulfuric acid (2:1) and heated it until the excess acid was driven off. They again ignited the ash in the muffle for 15 minutes.

2. METHODS OF THE ASSOCIATION OF OFFICIAL AGRICULTURAL CHEMISTS [3]

In addition to the sulfate-ash method, the AOAC has adopted as official two methods for carbonated ash, as follows:

Method I.-Heat 5 to 10 g of the sample in a 50- to 100-ml platinum dish at 100° C until the H2O is expelled, add a few drops of pure olive oil, and heat slowly over a flame until swelling ceases. Then place the dish in a muffle and heat at low redness until a white ash is obtained. Treat the residue with a little (NH)2CO, solution, reevaporate, and heat again in the muffle at a very dull red heat to constant weight.*

Method II.-Carbonize 5 to 10 g of the sample in a 50- to 100-ml platinum dish at a low heat and treat the charred mass with hot water to dissolve the soluble salts. (In low-purity products the addition of a few drops of pure olive oil may be desirable.) Filter through an ashless filter, ignite filter and residue to a white ash, add the filtrate of soluble salts, evaporate to dryness and ignite to about 525° C to constant weight.*

3. ADDITIONAL METHODS

A number of other methods of determining ash have been proposed by various investigators, the details of some of which have been collected by Jamison and Withrow [1]. The methods are as follows: Oxalic acid method of Grobert [4].

Quartz sand modification of Alberti and Hempel [5].

Benzoic acid modification of Boyer [6].

Zinc oxide modification [7].

Lixiviation modification [8].

Von Lippman advocates taking the dried-out sample on which the water determination has been made, saturating it with vaseline oil (having a boiling point of about 400° C), and igniting the mixture. The carbonized mass is then to be burned to ash in a mixed current of air and oxygen.

Since certain insoluble materials, such as sand and clay, which may be present in the sugar, and would therefore be included in the ash as determined by incineration, have no appreciable effect on the sugar in the process of refining, it is frequently necessary to determine the percentage of soluble ash. This may be accomplished by dissolving the sugar in hot water, filtering, washing the filter thoroughly with hot water, and evaporating the combined filtrate and washings in a platinum dish to dryness. The ash in the dry residue is then determined by one of the standard methods.

The use of (NH1);CO, was dropped in 1940.

The determination of ash by means of conductivity measurements is treated in chapter XVII, p. 275.

4. REFERENCES

[1] U. S. Jamison and J. R. Withrow, J. Ind. Eng. Chem. 15, 386 (1923).

[2] J. P. Ogilive and J. H. Lindfield, Int. Sugar J. 20, 114 (1919).

[3] Official and Tentative Methods of Analysis of the Association of Official Agricultural Chemists, 5th Ed., p. 487 (1940).

[4] J. de Grobert, J. Chem. Soc. 58, 670 (1890).

[5] Alberti and Hempel, Deut. Zuckerind. 16, 1069 (1891).

[6] E. Boyer, J. Chem. Soc. 58, 1472 (1890).

[7] A. H. Allen, Commercial Organic Analysis, 4th ed., p. 346 (P. Blakiston's Sons & Co., New York, N. Y., 1909.)

[8] Methods of Sugar Analysis, p. 10 (The Great Western Sugar Co., Denver, Colo., 1920).

XVII. ELECTRICAL CONDUCTANCE OF SUGAR

SOLUTIONS

1. INTRODUCTION

Since the electrical conductance of a liquid may be made a measure of the concentration and mobility of conducting particles in solution, its measurement is useful in determining these qualities in sugar products. It has been used extensively, principally as a rapid method of estimating the ash content of solutions [1, 2, 3] at various stages of manufacture in order to predict their performance in subsequent stages or as an index of the quality of the finished product. Many have found, however, that it is not always necessary to convert the conductance measurement to the more familiar value of sulfate or carbonate ash. In such cases, conductance values are reported in units of specific conductance.

More recently, conductance measurements have come into use in the sugar-manufacturing process to control such operations as boiling [4], crystallization [5], centrifuging [6], and others [7], and in the laboratory to determine purity [8, 9] and concentration of solute [10].

Briefly, electrical conductance is the reciprocal of electrical resistance. In sugar solutions it is expressed in units of specific conductance or reciprocals of units of specific resistance, which in turn may be defined as the resistance in ohms of a column of liquid 1 cm long and having a uniform cross-sectional area of 1 cm 2.

To determine the specific conductance of any volume of liquid containing an electrolyte, it is therefore necessary to measure (a) the resistance in ohms of the liquid, and (b) the dimensions of the volume of liquid causing this resistance. This measurement may be performed with great precision, provided errors depending on the temperature of the solution, the construction of the bridge, oscillator, balancing capacitor, resistance standards, and change of apparent resistance with frequency, have been eliminated or corrected [11, 12]. The resistance of the solution is determined by connecting the cell, Zz, figure 46, in one arm of an alternating-current bridge and adjusting R, and C, until no current flows between the points C and D, as detected by means of head phones, T. Then, if Z, and Z2 are electrically the same, the value of R, may be used to determine the resistance of the column of solution in the cell.

The volume of liquid causing this resistance is most conveniently measured indirectly as explained later under the section on Conductivity Cells, p. 268.

The possibility of electrolysis and oxidation-reduction in sugar solution has dictated the use of alternating current for the measurement of electrical conductance. Although alternating current does tend to reverse the effects of electrolysis, this reversal is not complete except under special conditions. Furthermore, the use of alternating current has introduced complicated electrical phenomena in the bridge network. The conductivity cell, being a part of this network, behaves in a manner that can be approximately simulated by a circuit made up of resistors and capacitors [12, 44]. The resistance of the column of liquid in the cell can be balanced by the variable resistor,

FIGURE 46.-Direct-reading bridge circuit for electrical conductance.

R,, figure 46, and the capacitance is most conveniently balanced by the variable capacitor, C,, placed in parallel with the variable resistor. There still remains, however, a residual effect, which will be referred to as electrode-polarization reactance. The magnitude of this effect, which tends to increase the apparent resistance of the solution, is influenced by the degree of platinization, the material of the electrodes, the composition of the electrolytes, and the frequency of the alternating current [12, 17].

2. APPARATUS

(a) BRIDGE

Most of the bridge circuits of recent design for studies of sugars are modifications of the type used by Kohlrausch [13], figure 47. The most important change, aside from improvements in apparatus, is that the majority read directly either in ohms or ash percentage.

The Sandera bridge [14], figure 48, although basically a Wheatstone

bridge, depends not on the variation of resistance to accomplish balance but on the variation of the distance between electrodes. The cell, C, and the calibration resistor, R, which remains unchanged during any one measurement, form two arms of the bridge. The other two arms are incandescent

lamps, L1 and L2. These lamps are arranged in such a manner that light falls on opposite sides of an opaque wedge, P. The bridge is balanced by modifying the distance between the electrodes of the cell until the light reflected from the wedge is of equal intensity on both sides. After proper adjustment, the distance between the electrodes indicates ash percentage.

(b) NULL-POINT INDICATOR

In most precision bridges, a telephone receiver, sensitive to the frequency of the alternating current flowing through the bridge, is connected as shown in figure 46 to indicate the absence of current across CD when capacitance branches are balanced.

R

Ry

CELL

FIGURE 47.--Kohlrausch bridge circuit for electrical conductance. R, fixed resistance; R., balancing resistance; T, headphones.

both the resistance and

An electron-ray tube (6E5) has been successfully employed in a bridge circuit as a visual null-point indicator. This circuit, figure 49, may be used at any frequency between 1,000 and 15,000 cycles, and if the tuned circuit, L, and C13, is properly selected, it is sensitive to a change in cell resistance of 1 part in 100,000, with a potential difference of less than 2 volts between the bridge terminals, A and B, figure 46.

Other types of indicators are those which are sensitive to frequencies of 60 cycles or lower, as, for example, an alternating-current galvanometer, milliameter, etc. Such types are particularly useful when observations must be made in noisy surroundings. However, the polarization reactance resulting from these low frequencies must be ignored or compensated for in the manner described under Checking Čells, p. 274