on starch suspensions. The experim ental evidence indicates th a t such tables can be constructed m a th e m a tic a lly i f th e absolute density o f th e specific starch is k n o w n .
F
O R many years the wet-milling industry in this country has employed the hydrometer for measuring the density of starch suspensions or slurries in their manufacturing operations. Various companies have adopted dry substance tables for these Baumd values, but no common table exists for use within or without the industry where starch is employed.
The use of a hydrometer for the measurement of solids in a suspension or slurry may seem strange to chemists who regard it solely from the standpoint of determination of density in true solutions. Its success in starch suspensions is based on the relatively slow settling of the starch and on the fact that the concentration of the starch suspensions at the hydrometer bulb is substantially constant. A n analyst familiar with these properties acquires the technique readily and results on a common sample agree within 0.1 136. or 0.2 per cent dry substance.
Tables have been published for the B aum 6-dry substance relationship for potato starch suspensions (4, 5, 6). Other papers have been published on the densities of the various starches (S).
The purpose of this paper is to present such tables for corn
starch, known in the trade as pearl starch, for the hydrometer in use in this country— 145 Modulus, standardized at 60° F.
E x pe rim e n tal
This work is based on methods previously published on the determination of the Baum 6-dextrose equivalent-dry sub
stance for corn sirup and corn sugar (1, 2) which should be
* Present address, Goodyear Tire & Rubber Co., Akron, Ohio.
consulted for the basic experimental methods, as only the departures from this work are discussed here.
C o r n s t a r c h . The starch used for these tables was taken from the final washing filter in the factory. The cake from this filter is either dried to form the pearl starch of commerce or suspended in softened water for subsequent hydrolysis to corn sirup or corn sugar. A quantity of this suspension was washed by decantation or dewatered on a Büchner funnel, washed, and finally suspended in distilled water to approximate the various densities required.
The starch had a crude protein content of 0.3 per cent on a dry basis.
B a u m £ b y H y d r o m e t e r . The Baumd readings were made at three temperatures— 60°, 100°, and 140° F. (15.56°, 37.78°, and 60° C.), the temperature of the bath being maintained within
±0.02° F. (0.011° C.). Approximately a gallon (3.785 liters) of the sample was placed in a glass bottle in the water bath and the suspension was maintained by a motor-driven stirrer so adjusted that the agitation was not vigorous enough to cause air to be occluded. When the temperature of the suspension equalled that of the bath, the sample was quickly transferred to
hydrom-T a b l e I. Bau.mê-Pek C e n t Dr y S u b s t a n c e S t a b c h
Ta b l e II. Co r r e c t i o n s t o Be Su b t r a c t e d f r o m As s i g n e d
eter cylinders and five 4-ounce (0.12-liter) screw-top bottles.
The former were used for the determination of Baum6 by hy
drometer and the latter for determination of density by pycnom- eter.
The starch in the cylinders was kept in suspension by means of a long rod which had a perforated disk at one end, and by care
fully introducing and manipulating this agitator, no air was carried into the suspension. The hydrometer to be used was kept in a cylinder of distilled water adjacent to the cylinders containing the starch suspension, and when the temperature of the starch suspension was the same as that of the bath, the hy
drometer was removed, quickly wiped free from the film of water, and placed in the starch suspension. When the spindle was at rest a small drop of methylene blue solution (in water) was added at the stem to accentuate the meniscus and the reading made at the upper edge of the blue line around the stem. Previous work had indicated that the differences between the reading of this point and the plane surface of the liquid (correct reading) was 0.08° Bd. Thus 0.08° Be. was added to the observed reading and this value appears in all tabulated data for Baumd
D e n s i t y b y P y c n o m e t e r . Settling difficulties presented a problem in the determinations of density by pycnometer and as a result three methods were used to overcome this trouble.
None is completely free from criticism, although the latter two are regarded as more desirable. The pycnometers used have been described previously (2).
a. The 4-ounce samples mentioned above were kept in the bath while the Baum6 was determined by hydrometer. Then the temperature was raised or lowered a few degrees, depending on relationship of bath to room temperature, to compensate for heat changes during the transfer of the sample to the pycnometer.
The bottle was agitated every 5 minutes, then returned to the bath, reagitated, etc. After five such agitations, the sample was transferred rapidly to the pycnometer, stoppered, and placed in the bath which was now set at the desired temperature. The pycnometers were kept in the bath for 30 minutes, then removed and weighed.
Using this method, difficulty was encountered at the higher concentrations in that starch collected at the ground surfaces.
The precision was about 0.05° B6. and as a result this method was discarded for the heavier Baum6s and method b used instead.
The moisture on the starch was determined on one of the re
maining 4-ounce samples, as described below.
b. The pycnometers were filled about three-fourths full at room temperature, heated to about 130° F., and boiled very gently under high vacuum to remove occluded gas. The starch was then allowed to settle and the remaining space was filled with cool, recently boiled distilled water. The pycnometer was placed in the water bath and weighed at the end of an hour.
From this weight it was possible to calculate the specific gravity on the basis of complete suspension.
This method eliminated the difficulty of the collection of starch between seals and also enabled any loss through the capillary of
the stoppers to be taken as pure water— i. e., no solids. The moisture test was made on a separate bottle as before.
c. The pycnometer was used with an added feature consisting of a test tuoe with standard-taper joint as shown in Figure 1.
The starch suspension was run into the pycnometer until it was nearly full and then allowed to settle. When the top part of the liquid was essentially clear, the stopper was inserted ana the appa
ratus placed in the bath. After an hour the pycnometer was re
moved from the bath and weighed. Then the stopper and outer cap were removed and the test tube was applied. The suspension was shaken until uniform and a moisture determination made on this material.
This method eliminated the difficulties of starch at the seals;
the liquid that was lost through the capillary was essentially pure water; and settling was not a dominant factor, since the moisture determination was made on the suspension actually present in the pycnometer. Tests were made to estimate the error introduced by loss of the film of water left on the stopper and cap which were removed when the test tube was attached for resuspending the starch. This was found to be 20 to 30 mg. on a 95- to 100-gram sample weight. This involves an error of 0.02 to 0.03 per cent and can be neglected for all ordinary purposes or be compensated for in work of highest precision.
D e t e r m i n a t i o n o f M o i s t u r e . The method used is basically the same as that employed for the determination of moisture in corn sirup and corn sugar (./) which should be consulted for details.
Diatomaceous Silica, prepared Jolms-Manville Iiy-Flo (1, 2).
Apparatus. Wide-moutn Erlenmeyer flasks, 250-ml. capacity, with 40/12 standard taper, with stoppers. Pyrex test tubes
100 X 15 mm. (i).
Procedure. Twenty-five to 35 grams of diatomaceous silica were run into duplicate flasks, the test tube was introduced, and weight constancy was obtained under oven conditions identical with those of the test. Weighings were made with an empty flask as a tare. The starch samples, either in the 4-ounce bottles or the pycnometer equipped with test tube, were shaken until homogeneous suspensions were obtained. Sample portions, sufficiently large to yield 5 to 8 grams of dry substance, were transferred quickly to the flasks by means of a pipet from which the tip had been cut. The sample was then worked into the diatomaceous silica by means of the test tube, which yielded a damp powdery mass.
The flasks were placed in a vacuum oven at 100° C., and thé pressure was reduced first by an efficient water pump until most of the water was removed and then by means of a Megavae pump.
After approximately 4 hours of drying the flasks were removed
336 I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y Vol. 15, No. 5
from the oven and cooled and the mass in the flasks was reworked into a fine powder. The flasks were returned to the oven, and a pressure of less than 1 mm. maintained until weight constancy was obtained. An overnight period has been found adequate.
Because of the very hygroscopic property of dry starch, an efficient train (1) must be used to introduce air into the oven and the flask stoppers must be inserted immediately upon opening the oven.
This method gave reproducible results within very narrow tolerance. Microscopic examination of the dried starch revealed no visible change in the granules through rupture or gelatinization. Tests indicated that oven temperatures could be carried as high as 110° C. with no appreciable effect on the results.
The method of reporting data was the same as that previ
ously used (2). The net pycnometer weight was corrected first to vacuum. In the case of the 60°/60° F. data, the density (vacuum) was obtained and the corresponding Baumé value assigned. This procedure was changed for 140°/60° F.
and 100°/60° F. data. After the net pycnometer weights had been corrected to vacuum, an additional weight correction (glass expansion and hydrometer) was added before calculat
ing the density in order to make these values agree with the corresponding hydrometer readings. The reason for this has been previously discussed (2). B y this correction observed Baumé readings and Baumé values obtained by pycnometer always agreed within 0.05° Bé. and with an average deviation of 0.03° Bé.
The Baumé-dry substance data were plotted in two ways:
Baumé vs. dry substance and Baumé vs. factors. In both cases the best curve was a straight line, as it should be if the slurry were a true suspension. The previously used method of factors was employed for purposes of tabulation.
The factors were:
Factor for 60760°F.(F60760°F.) = 1.7770 JLJé.
Factorfor 100760°F. (n 00°/6 0°F .) = — M i — = 1.7700
Factor for 140760° F. (F I40°/G0° F.) = + f ' Q1° = L753
In Table I, the Baum 6s in all cases are the observed values and the specific gravities (air) are those corresponding to the Baum 6s.
In the above data the specific gravities at 60 760 ° F. in air can be converted to vacuum by the usual calculations. The I00°/G0° F. and the 140760° F. data require not only the calculation for air to vacuum but also those for glass expansion and for hydrometer. This has been discussed previously (2) and an abridged table (Table I I ) is given here for those who may desire to express the data on some other basis.
Since most Baum 6 tests in factory processes are not made at a fixed temperature, a table of temperature corrections is necessary (Table I I I ) .
The determination of each point in the table furnished enough data for calculation of the density of the actual starch solids. lienee a series of independent starch density values were obtained over the range of suspension gravities. These were observed to check very closely in the 60°/60° F. series and to average 1.636. Employing this value for the Baum 6- dry substance relationship, a calculated table was found to check the experimental table within the limits of the experi
mental methods used. I t would appear, therefore, that tables for water suspensions of the whole gravity range could be pre
pared for any starch from an accurate density value of starch solids determined at a single point. Each temperature, of course, would require a separate determination.
A table has been prepared for general factory use based on 100 ° F (Table IV ). I n this table the Baumds are the observed readings, the specific gravities are those in air assigned to the Baumd values and the weights per gallon are based on these specific gravities. (To bo added to or subtracted from observed Baumd to adjust to 100° F.)
Temperature,
Experimental results indicate that tables for Baumé-dry substance starch can be made mathematically, provided the density of any starch is known.