Measurement of Reflectance of White Sugar

Energy Series, Division VIII, Vol. 2, New York, McGraw-. Hill Book Co.,1951. (3) Holler, A. C., and Yeager, J. P., Ind. Eng. Chem., Anal.Ed.,. 14, 719...
0 downloads 0 Views 3MB Size
V O L U M E 2 4 , N O . 5, M A Y 1 9 5 2 recovering manganese from the cathode, the mercury being distilled off in a stream of nitrogen. LITERATURE CITED

(1) Furman, N. H., Bricker, C. E., and McDuffie, B., J . Wash. Acad. Sci., 38, 159 (1948). (2) Grimes, VV. R., and Casto, C. C., U. S. Atomic Energy Commission, Collected Paper 79 (March 1947) : Kational Nuclear Energy Series, Division VIII, Vol. 2, iYew York, McGrawHill Book Co.. 1951. (3) Holler, A. C., and Yeager, J. P., ISD. ENG.CHEM.,ASAL. ED., 14, 719 (1942). (4) Khlopin, N. Y., 2. anal. Chem., 107, 104 (1936). (5) Latimer, W. hf., “Oxidation Potentials,” S e w York, PrenticeHall, 1938. (6) Lingane, J. J.. and Neites, L., ANAL. C m x . , 19, 159 11947).

( 7 ) Lundell, G. E. F., and Hoffman, J. I., “Outlines of Methods

829 of Chemical Analysis,” pp. 94-5, Sew York, John FYiley tk Sons, 1938. (8) Maxwell, J. A., and Graham, R. P., Chem. Revs., 46, 471 (1950). (9) Rodden, C. J., “A4nalytical Chemistry of the Manhattan

Project,” National Xuclear Energy Series, Division VIII, Vol. 1, p. 515, Xew York, McGraw-Hill Book Co., 1950. (10) Russell, A. S., Evans, D. C., and Rowell, S. Ti’., J . Chem. SOC., 1926, 1872. (11) Sinclair, E. E., and Casto, C. C., U. S. Atomic Energy Commission, R e p t . C-4.360.6(March 1946). (12) Willard, H. H., and Greathouse, L. H., J . Am. Chem. Soc., 39, 2366 (1917). RECEIVED for review August 18, 1951. -4ccepted February 18. 1952. From B thesis presented by Leven S. Haalegrore t o the Graduate School of Emory University, December 1949, in partial fulfillment of t h e requirements for the degree of master of science.

Measurement of Reflectance of White Sugar T. R . GILLETT AND P. F. MEADS California and Hawaiian Sugar Rejining Corp., Ltd., Crockett, Calif. Color is an extremely important factor in determining the qualit). of w-hite granulated sugars. Up to the present time no suitable devices have been developed for use in plant control of the color of w-hite sugars. An investigation designed to develop a suitable control procedure led to the development of a photoelectric reflectance method for determining the color of granulated sugars controlled w-ithin reasonably narrow grain size distributions. This method has been used for a number of years for plant control purposes w-ith considerable satisfaction. Production personnel know at all times the color of the refined sugar and are able to correct production methods promptly in order to maintain uniform quality in the product. This is very desirable in meeting product specifications and customer requirements.

B

ECAUSE color is an extremely important factor in the evaluation of the quality of white sugars, a great deal of study and investigation has been carried on in the industry t o develop suitable means for measuring t h e color of such products. This has involved measurement of t h e color of dry white sugar as well as of solutions of white sugar. Both types of measurement are significant for quality evaluation and for control purposes. Solution colors are of particular importance where the white sugar is dissolved prior t o use in industrial food processes. On the other hand, the white appearance of granulated sugar is of significance where i t is used or compared in t h e dry form. The primary concern is the visual appearance of the dry sugar or the sugar solution. Accordingly, all color measurements in this laboratory have been placed on a visual basis-Le., referred t o suitable visual procedures. I n this limited sense, the color methods may be considered someJThat empirical. However, the photoelectric transmittance and reflectance measurements utilized are basic and their fundamental nature is restricted only by equipment limitations, which have had no effect on the usefulness of the methods for control and comparison. Both dry and solution methods of color measurement have been employed in this laboratory. Solution colors are used as a basis for controlling as well as evaluating color. This type of measurement is generally more precise because the color is accentuated in solution, and transmittance measurements are more readily made than reflectance measurements. The method developed in this laboratory for photoelectrically measuring the solution color of white sugars has been described (5’). This instrument and procedure have proved very satisfactory for routine color work in the laboratory, but they have not been particularly adaptable t o color control in the refinery. Considerable study was given t o the possibility of developing

an instrument that would give a close approximation of the color of a dry granulated sugar, and would be sufficiently simple and rapid for use by operating personnel. This investigation ultimately led t o the design and construction of a photoelectric reflectometer which has proved very satisfactory in service for the past 10 years. The instrument essentially replaced t h e previous method of visual comparison against a series of dry standard sugars. With this change, inaccuracies resulting from visual observation and personal judgment have largely been eliminated. The problems involved in developing an instrument for measuring the color of dry white sugar are obvious. I n the first place, the amount of color involved is extremely small-for example, high quality white sugars, of normal granulated grain size will reflect 94% or more of the incident light. Lower quality sugars, which occasionally may he enrountered in the trade, might reflect as little as 90% of the light. Obviously, measurement of the small differences in light reflected by various sugars is very difficult, and differences in grain size of white sugars seriously affect the nature of the reflection. 1Iore finely divided particles appear whiter than coarpe particles both to the eye and t o photoelectric devices Furthermore, differences in luster influence results, and differences in apparent grayness sometimes confuse the detection of yellowness in the sugar. These and other factors make the measurement of color of dry sugar an exceedingly complicated problem. Grain size variations are particularly important. Keane and Brice ( 5 ) found that, in general! the coarse-grained sugars had lower reflectances than the fine-grained sugars and that variations in coloring matter appeared t o have a more pronounced effect on the reflectance of the coarse sugars than of fine sugars. They concluded that a plot of reflectance against grain size should yield

ANALYTICAL CHEMISTRY

830 a family of curves, one curve for eaeh value of apparent color. The present investigation has substantiated this conclusion. Browne and Zerhan ( 1 ) have a180 commented on the effect of grain size on reflectance.values, referring pmtieularly t o the work a t the Java Sugar Experiment Station, In the initial stages of this investigation a number of different methods were tried. Among these were experimental arrangement8 involving measurement of multiplereflectance, fluorescence, transmittance through a thin layer of dry sugar, and transmittance through a thin layer of moist sugar. None of these methods appeared particularly promising and the investigation was therefore directed t o development of a method for measuring diffuse reflectance from the surface of the sugar. This method wa8 adopted as a hasis for the instrument that was ultimately developed.

The slide-wire units have been expanded by the u8e of a series resistance, in a manner similar t o that used in the instrument developed for solution color measurements (S). This multiplying effect results in t h e t o t d 10% absorption range of white sugmrs being spread over the greater part of the full slidewire scale, thus giving much greater sensitivity in the instrument. As a matter of fact, the scale spread is increased about seven times. T o Secure optimum sensitivity, a multiple reflecting lamp and scale galvanometer are employed.

DESCRIPTION OF INSTRUMENT

The general arrangement of the photoelectric reflectometer is shown in Figure 1.

Figure 3.

Photoe1eotrie:Retleetometer

~

~~~~

~~

--.,

~ ~ - -

absolute refleetances of 'ahout 87%"as related t o magnesium carbonate are used. For initial balancing of the circuit, the whit.e ~

Figure 1. Optical Arrangement of Photoelectric Reflectometer

glai8 plate, replacing one of the tiles. It was determided that the clear photo glass plate had no significant effect on the measurements.

A photograph of the completed instrument is shown in Figure 3. In principle, light is directed upward through a color filter a t OPERATION OF INSTRUMENT

the photoele&ic celk. This'is the same ([eneral arrangement hitlance point is in the direction of great.er reflectance, ~

~~~~

~~

~~

~~

watt, llO-<, projection la& The color filter is a Corning glass Daylight type No. 590. This li$ht blue filter modifies the spectral character of the light to simulate conditions under which white sugars are generally examined visually under daylight from the north, ~

mol0 J CELL

ai granu-

~ . . lated sugars generally have higher reflectancos (whiter) than the

-PYoTo CELL

reference tile. The lefehand tile of the instrument is then raised, the steel ring placed on the glass plate, and a sample of sugar poured into the metal ring and then covered with a suitable opaque cover. The slide-wire is rotated until the galvanometer again indicates a. balance. The color reading is taken from the slidewire scale. The sugar is then swept off the glass plate by a movement of the steel ring t o the left, and drops through the hopper to a receiving pan. Last trace8 of the sugar are removed with a soft cloth. CALIBRATION

Figure 2.

Photoelectric Reflectometer Circuit

A number of variations utilizing two photoelectric cells in balanced circuits were tried. The hasic photoelectric circuit which WSB finally adopted (Figure 2) is a modification of similar circuits in this lahoboratory (2.4).

As it wa6 intended that solution colors would remain the standard for color determination, the reflectance instrument is calibrated in terms of solution colors. For this purpose the color8 of a number of special samples of sugars of three different grain sizes were carefully determined on both the laboratory solution color instrument and the new reflectometer, The reflectance measurements have been converted from slidewire scale readings to -log R values and are compared in Figure 4 with the solution color values. Each of the three types of sugar gave a linear relationship in terms of -log R and -log T. Also, as would heexpected, thehesugarsshowa higherreflectance(1ower -1ogR value) for a particular solution color than do the coarser grain eugars. Table I shows the average screen analysis and particle size index of these sugars.

V O L U M E 2 4 , NO. 5, M A Y 1 9 5 2

831

The presence of colloidal matter has a marked effect on the relationship between dry and solution colors. Certain types and quantities of colloidal matter impart a gray cast to the dry sugar and influence the amount of light absorbed and hence affect the color reading indicated by the reflectometer. On the other hand, the solution instrument corrects for the presence of colloidal matter and thus accentuates the differences between the two color values. The luster of the cryatals may also have some bearing on the reflected color, although this hap not been conclusively demonstrated. I t seem? reasonable, however, that a lusterless sugar with chalky crystal surfaces nil1 show different reflecstance values than clear, dust-free sugar crystals.

Table 1. Average Screen inalyses of Sugar Samples ( P e r cent) On 20

On 28

On 35

On

On

48

80

On

100

On

150

Balance 0.1

Calculated Particle Size Index

Coarse granulated T 3 . 7 7 6 . 1 18.5 1.3 . , . , , 33 Regular granulated . T 9 . 4 40.3 43.2 7.1 48 Fine granulated , . . 0 1 2 . 1 14.3 34:8 34'8 13 6 98 T = Trace. Particle size index is average particle size, expressed a6 Tyler mesh value, calculated froiri screen analysis, assuming sugar retained on a particular screen has average size midway between t h a t screen and t h e next coarser screen.

In view of this linear relationship it is possible to calibrate the instrument scale in terms of -log R, -log T , or some other proportional unit. In this particular case the instrument is also calibrated in terms of C and H color units ( 6 ) ,since the solution color instrument is calibrated in these same units. Rhich are used for routine control in the refinery. DISCUSSION OF RESU-LTS

Initial calibration involving the samples mentioned above has been checked by the subsequent determination of colors on the solution colorimeter compared with those on the reflectometer on consecutive-shift composite production samples of granulated sugar. In this particular list, the standard deviation is about 15 C and H color units, the maximum amounting to 30 C and H units. For the most part, this instrument will give results within 25 units of the colors determined on the solution colorimeter.

Even though the instrument has limitations associated with grain size variations and distribution of color within the crystal lattice, it has been of very considerable assistance to operating personnel in the maintenance of production color within proper quality specifications. Its use has been restricted to sugars with the normal granulated sugar particle size, although, as is evident from the curves shown in Figure 4, appropriate scales could be developed for products of other grain sizes. However, it is very essential that the particle size distribution be maintained unifornily within narrow limits for each puticular grade. The use of this instrument has reduced considerably the amount of sugar returned for reprocessing because of off-color characteristics. Obvioudy, this has provided considerable monetary benefit s. SUMMARY

An investigation of methods of measuring t,he apparent color of dry white sugars led to the development of a photoelectric reflectometer which utilizes a balanced photoelectric circuit and the principle of diffuse reflectance. I n its development numeroue

Table 11. Typical Coniparative Results on Dry Sugar Reflectometer Sample 1 2 3 4 5

Y k

. L O G T.

832

ANALYTICAL CHEMISTRY

factors were considered relating t o the effect of grain size, luster, distribution of color in the crystal, etc., on the reflectance value of different types of white sugars. Good correlation has been shown between reflected color and solution color of a white sugar, providing the grain size and other characteristics are uniform. The instrument is calibrated in terms of solution color and has been used for over 10 years by operating personnel for determining the reflected color of production sugars. Results are obtained quickly and with sufficient reliability t o avoid packing of offcolored sugars. LITERATURE CITED

(1) Browne, C. A., and Zerban, F. W., “Physical and Chemical

(2) (3) (4)

(5) (6)

Methods of Sugar Analysis,” 3rd ed., p. 620, Kew York, John U’iley & Sons, 1941. Gillett, T. R., and Holven, -4.L., Ind. Eng. Chem., 35,210 (1943). Gillett, T. R., Meads, P. F., and Holven, A. L., ANAL.CHEM., 21,1228 (1949). Holven, A. L., and Gillett, T. R., Facts About Sugar, 30, 169 (1935). Keane, J. C., and Brice, B. A , , IXD.EXG.CHEM.,ANAL.ED.,9, 258 (1937).. Spencer, G. L., and hleade, G. P., “Cane Sugar Handbook,” 8th ed., p. 552, New York, John Wiley & Sons, 1945.

RECEIVEDfor review June 15, 1951. Accepted Januaay 9, 1952. Presented before the Division of Sugar Chemistry, Symposium on Measurement of Color of Sugar Products, at the 119th Meeting of the AMERICAN CHEMIc . 4 ~SOCIETY, Boston, Mass.

Color Standards for Sugar Cane Sirup and Edible Sugar Cane Molasses C. B. BROEG AND-C.F. WALTON, JR. Sugar Branch, Production and Marketing Administration, U.S. Department of Agriculture, Washington, D. C. Although color has been very important in the commercial marketing of sugar cane sirup and edible sugar cane molasses, no w-idely accepted method for measuring color has been in use. Some distributors have their own color standards, but these are not in general use. Others depend entirely on visual inspection. Comparison of color measurements by means of Lovibond glasses, liquid color standards, other glasses, and colorimeters indicated that a liquid color standard which is flexible, easily used, and inexpensive would be the most useful until further work developed standards of a permanent nature. Liquid color standards have been recommended for use in grading these sugar cane products in connection with the C . S. Department of Agriculture’s permissive grades for sirup and molasses. This fills a gap in the marketing of these products, and the results indicate that the color standards are adaptable to a number of liquid sugar products.

I

N CONNECTIOS with the development of specifications for grading sugar cane sirup and edible sugar cane molasses, the need for color standards became a major factor. Preliminary specifications contained only general and somewhat vague descriptions of color, and producers and distributors of these products pointed out that such color specifications were confusing, inadequate, and a source of dispute in transactions based on such general terms. Consequently, it was decided t h a t definite color standards should be developed for these products. Because sugar cane sirup and edible sugar cane molasses are produced generally by individual farmers or small factories and relatively small units of production are involved, the color standards should be as simple as possible. Color standards have long been used for grading maple sirup. Liquid standards of caramel solutions were devised by Balch ( 8 ) and for a number of years were distributed by the Department of Agriculture. Recently, Brice and his associates ( 5 ) have developed permanent glass standards which have largely replaced the liquid standards. Some of the large sirup distributors have developed color Standards for their own use (Penick & Ford, Ltd., Inc., and .Imerican Molasses Co.). Sugar cane sirup buyers in the southeastern Gulf States use color as one of the criteria for purchasing sirup, but the practice has been t o rely on visual inspection alone and not t o use standards. This condition has led

to uncertainty and confusion in the grading of sirups. A number of methods for measuring or comparing colors were investigated, in the hope of finding color standards that would be widely applicable and simple t o use. Research on the color of sugar products and on methods for measuring color has generally been along two lines: measurement of color by comparison with a standard color, and measurement of absolute color. Among the many investigators of color are Ventzke (II), Stammer ( I O ) , Bates and associates ( S ) , Peters and Phelps ( 9 ) , Brewster ( 4 ) , Keane and Brice (a), Zerban (12), Zerban et al. (13), and Gillett, Meads, and Holven (7). Because in grading it was not considered important t o measure the absolute color of sirups and molasses as such, but rather t o compare the amount of color in a sample with that of a known color standard, this work was largely directed toward the investigation of color comparison and development of a suitable color standard for making the comparison. Attention was given t o availability and reproducibility of the color standards, simplicity of use, cost, flexibility, and applicability t o a variety of sugar products. It seemed desirable t o have standards which would be immediately available and easily reprodGcible, could be used a t any place without special laboratory facilities, would be reasonable enough in cost t o permit even the smallest producer to use them, and would be flexible enough to permit changes to be made in a short time and a t little expense. Brief studies were made of several glasses and li uid color standards and several colorimeters. The standard 8asses investigated were: Lovibond glasses series 52 (yellow) and brewers’ red glasses and combination of these two series, Hellige amber C and D series glasses, and glasses developed by Penick & Ford, which consisted of two yellow and t v o red glasses of different optical densities for use singly or in combinations. The glass standards for maple sirup developed by Brice and his associates were not investigated. A standardized caramel solution of various concentrations and standards prepared from a Rolution of mineral salts were the only liquid colors studied. An Am5nco neutral wedge colorimeter and a Duboscq color comparator were used for comparative purposes, and a Coleman photoelectric colorimeter was used for measuring the transmittance of samples. Liquid colors, instruments, and numerous samples of sugar cane sirup and edible sugar cane molasses were collected and their “color” was determined by transmittance and by comparison with various standard colors. 4 clear cell providing a 0.125-inch depth of sample was prepared for general use except in the rase of the Hellige standards, where a 1-mm. depth was used. The