Methods in the Starch and Dextrose Industry W. R. FETZER, Clinton Foods Znc., Clinton, I o w a The methods of analysis used in the starch and dextrose industry have been greatly extended and improved during the past 2) years. .4 review paper covering all these advances is beyond the scope of this paper, which is confined to methods dealing largely with the sale of these products. BaumC-dry substance-dextrose equivalent tables have been published with corresponding data for refractive index. New methods have been suggested for the measurement of color in corn sirup. Viscosity tables for corn sirup have been found useful in the design and installation of equipment. The determination of reducing sugars has been improved through the Lane-
T
HE corn wet milling industry has been in existence for nearly 100 years. The earliest product was the familiar starch sirup, which is still referred to by the trade as “glucose.” Crude sugars of the “TO” and “80” types gradually made their appearance. Refined corn sugar, or dextrose, was produced commercially in 1926 and has become increasingly important as a sweetener. Today the industry produces annually 1,000,000 tons of corn s’irup, 45,000 tons of crude sugars, and 420,000 tons of dextrose. The first cornstarch plants were small factories that ground 2000 t o 3000 bushels of corn daily. They were widely scattered, and a company with a daily capacity of 5000 bushels was considered a large plant. Today there are nine companies producing starch, corn sirup, and dextrose, with an average capacity of 40,000 bushels daily; the largest plant grinds 100,000 bushels of corn per day. The early emphasis in manufacture was largely mechanical, involving the separation of the various components of the corn kernel. The earliest laboratories were what we would today call testing stations. Baume, acidity, and sulfur dioxide were the principal tests, and most of the foremen and superintendents were graduates of the ‘laboratory.” When the first graduate chemists entered the industry a t the turn of the century, they were not too favorably received by the mechanical group in charge of operations. However, the introduction of the use of pH by Sjostrom in 1922 (55),which solved many difficult questions, marked the turning point for more and more emphasis of a chemical nature in the wet milling operations. Today, most companies employ large groups of research chemists, the largest producer having several hundred such employees working in an ultramodern laboratory representing an investment of several million dollars. As would be expected, the early starch laboratories adopted the analytical procedures of the sucrose industry, even though these methods were not always strictly applicable to the products of starch hydrolysis. The individual efforts of the various laboratories led t o different interpretations of the same data, and there were discrepancies in the data given the uber of the various products. This situation produced so much confusion that in the middle thirties the management of the industry established the Corn Industries Research Foundation. Its original purpose was to obtain from the individual members, or by outside research, critical analytical data and methods which would be available to users of their products and useful to other laboratories engaged in research work involving starch products. To effect this end, a Technical Advisory Committee, composed of the key personnel of the laboratories from each member company, was formed, and this group has fostered research on analytical methods and apparatus. I t s final recommendation in many
Eynon procedures employing methylene blue. A special copper method for the determination of dextrose in the presence of other reducing sugars has had wide acceptance. Dry substance-refractive index tables for dextrose and BaumB-dry substance tables have been published. Starch has been shown to be stable over a wide range of drying temperatures. Orifice viscometers for the determination of the fluidity of starch pastes have been greatly improved. New recording viscometers have been suggested for the determination of rheological properties of starch pastes through a cooking and cooling cycle.
cases specifies an “official” and an optional method, the former being required in cases of a dispute on product specifications. Purchase specifications have been adopted for several pieces of equipment, and in one instance a complete assembly of equipment has been adopted and an agreement made with a vendor to stock or to manufacture it according to detailed specifications. The Technical Advisory Committee has worked with the Association of Official Agricultural Chemists in writing a chapter on starch hydrolytic products ( I ) , whereby authentic methods could be widely disseminated. The industry has been allotted one membership in the U. S. Committee for Uniform Methods of Sugar Analysis and has actively participated in its work. The Corn Industries Research Foundation has sponsored fellowships in universities and a t the Kational Bureau of Standards and has a list of projects which it desires to complete. Because a paper on all methods in the starc,h and dextrose industry viould be extremely long, this paper is confined to methods as applied t o finished products that reach the consumer-namely, starch, corn sirup, and dextrose. The methods that have been developed, widely accepted, and well covered in the literature, are covered briefly. However, on some methods, where a difference of opinion exists, more detailed information is given, in order that the points at issue may be better understood. SOLIDS BY HYDROMETER
Starch. The density and per cent solids can be determined upon starch slurries with considerable precision, if care is taken to assure a thorough suspension. Starch slurries or suspensions are mixtures of water and fine insoluble solids, and a BaumB-dry substance table may be constructed without experimental work, provided the density of starch is known. The density of starch is not a fixed constant, for it is influenced to some degree by the season, locality in which it is produced, and maturity of the parent corn. In addition, commercial st,arch carries residual corn glut’en,which is lighter than starch, to the extent of 0.20 to 0.35%, approximately 0.6% crude fat, and a trace of fine fiber. Cleland, Fauser, and Fetzer published complete tables for B a u d - d r y substance cornstarch (19). Their work was based on density by pycnometer, followed by dry substance on the suspension. The density of the cornstarch employed was 1.636. Senti ( 5 9 ) investigated the density of starch in a number of solvents and found the density of cornstarch t o be 1.628 in water. Corn Sirup. One of the items of analytical procedure taken from the sucrose industry was the old Dutch scale hydrometer with its temperature correction for heavy sirups, of 4’ F. equals 0.1 ’Baume. Because the viscosity of the commercial Baumks42 to 45-is very great, early producers added a concept of their 1129
1130
ANALYTICAL CHEMISTRY
own, that theBaum6readingshould beobtained at 140"F.,and to that reading should be added 1.00" Baume. This correction was purported to place the sale on a 100" F. basis, with an accepted trade definition as follows: Commercial B a u d = Baume reading a t 140' F.
+ 100"Baume
One of the first problems undertaken by the Technical Advisory Committee of the Corn Industries Research Foundation was t o change the hydrometer t o the American standard of 145 modulus. Considerable discussion centered on whether the hydrometer to be adopted should be the 145M-60' F. hydrometer of the Manufacturing Chemists' Association or the 145M-20" C. hydrometer advocated by the National Bureau of Standards for sugar sirups. The decision was made in favor of the former, largely to avoid confusion from the use of two types of hydrometers in the plant, as many supplies are purchased on the basis of the 145M-60" F. hydrometer. The preparation of Baumbdry substance tables proved involved, because the dry substance for any given Baume was found to increase as the dextrose equivalent increased. Deutrose equivalent is d e h e d within the industry as the percentage of reducing sugars, calculated as dextrose and expressed on a dry substance basis. As corn sirup is produced in several dextrose equivalents, more extended Baumhlextrose equivalent-dry substance tables were required, which necessitated a large amount of experimental work. These tables were placed in use in 1939 and published (244).Abbreviated tables are given by the Association of Official Agricultural Chemists (1). Dextrose. There are no published Baume-dry substance tables for dextrose, although Jackson (27, 29, 31) has published sufficient data for one to be calculated on the basis of 145M-20" C. One company has B table based on 145M-60' F., but has not released it for general use. SOLIDS BY OVEN BlOlSTURE
Starch. hloisture in starch can be determined by any of the distillation or standard oven methods. The distillation procedure requires the use of Filter Cel or asbestos t o avoid bumping (50); otherwise, the test can be carried out without difficulty. The determination by oven procedure presents no difficulty, if it is borne in mind that dry starch is extremely hygroscopic and that extreme care must be taken to avoid the readsorption of moisture (50). Hellman (26) has shown that stirch is unaltered physically over a wide range of temperatures. Corn Sirup. The determination of moisture in corn sirup was a controversial subject within the industry for over a decade. The basis for the controversy involved a belief by some chemists that the higher moisture values, tor a given density, by the newer methods, might result from a decomposition of the sirup. To prove that this did not exist required extended research. The determination of moisture in corn sirup proved to be simple, once the difficulty was recognized. Corn sirup contains dextrin which, on drying, forms a film which retards the release of any moisture beneath the film. If the film is thin, the error in the test from trapped water may be small; if the film is thick, the error may be as large as 1%. The solution to such a situation is simple dispersion, whereby the film thicknebs is reduced to a minimum. Fetzer, Evans, Cleland, and Fauser (20, 21, $3 investigated the determination of moisture in corn sirup and sugar products, by distillation, vacuum oven (employing Filter Cel ), filter paper, and sand procedures, a t different temperature levels, and found that corn sirup was an extremely stable material, provided the p H of the product was kept within t,he range of 4.0 to 5.5, Their work xlminated in a series of Baumbdextrove equivalent-dry substance tables (24)which are official for these products. The commercial ranges appear in the A.O.A.C. ILlethods ( I , p. 591, XXXIV, 142). The author prefers the Filter Cel and filter paper methods, the former for the average analyst and the latter for those with more analytical experience
Dextrose. Crystalline dextrose hydrate presents no particular drying problem, as its normal pH is within the stable pH range for moisture determinations. The test may be run in an air oven or vacuum oven, without dispersion of the sample. Hydrol. Hydrol, or corn sugar molasses, requires some precaution in a moisture determination. Commercial hydro1 has the following approximate percentage composition : Moisture Sodium chloride Total ash Crude protein Dextrose and other sugars by differenec
25-30
5-8 5
9-12 0 254.50 50-61
Hydrol is heat labile. The best procedure for determining moisture is by benzene distillation ( 2 2 ) ; next best is the vacuum oven (60"), Filter Cel dispersion method. A rapid control method can be employed by establishing an oven procedure standardized against the former procedures. SOLIDS BY REFRhCTIVE ISDEX
Corn Sirup. Refractive index offers the most rapid procedure for determination of solids. However, the refractive index for a given solids varies with the dextrose equivalent, which creates an additional variable, so that although the refractive index of an unknown sirup can be obtained, the corresponding solids cannot be obtained precisely unless the dextrose equivalent of the sirup is known. Fortunately, corn sirup is generally produced in the following dextrose equivalent ranges: L o a conversion Confectioner's sirup3 Extra sweet sirups
30-32 41-43 54-56
Complete refractive indexdextrose equivalent-dry substance tables have been published ( 1 8 ) for 20" and 45" C.; 45" C. is necessary for the commercial BaumBs, 42-46' Be (78 to 86% solids), for the sirup is so viscous at these concentrations that it would not be possible to close or separate the prisms. The refractive index-dextrose equivalent-Baum6 table appears in the A.O.A.C. Methods ( 2 ) and fortunately the data allow the Baume to be determined within 0.1 '* which is satisfactory for most commercial purposes. Dextrose. The refractive indexdry substance tables for dextrose published by Zerban and Martin (59) have been accepted by the Association of Official Agricultural Chemists and the U. S. Committee for Uniform Methods of Sugar Analysis. The industry appreciates the work done by these authors on a major product. REDUCIWG SUGARS
Starch. The methods for reducing poaer of starch fall into four classifications: ( 1 ) copper (9), (2) hypoiodite (48),(3) ferricyanide (S5), and (4) dinitrosalicylic acid (44). The reducing power of starches is important in the matter of end group assay and the estimation of the degree of polymerization. The paste characteristics of starch systems often render the determination difficult. For euample, in unmodified starches where the reducing power is small, larger sample weights are required, whereas, in modified starches in which the reducing power increases, smaller sample weights can be med. Therefore, a high degree of precision obtainable with a minimum of sample is an important consideration in the choice of a method. This requirement is well satisfied by Shaffer and Williams (5S), who propose the potentiometric measurement of the oxidationreduction potential of a ferri-ferrocyanide system. As to the mcuracv of these methods, Blom and Rosted (9) have raised objections to the ferricyanide, hypoiodite, and the usual copper tartrate complexes. because they fail to control overouidation. Instead, these authors propose a copper complex of copper sulfate, potassium carbonate, and potassium bicarbonate. The dinitrosalicylic method which has been developed by lfeyer and his Swiss collaborators ( 4 4 ) allows colorimetric measurement
1131
V O L U M E 24, NO. 7, J U L Y 1 9 5 2 of the reducing power and has receutly been used by Seboch (4s) and Kerr(35). Corn Sirup. The control of the mid hydrolysis of starch is entirely dependent upon the determination of reducing sugars. By adjusting starch concentration, amount of acid, and time, it is possible to produce a sirup within the range of 2 dextrose equivalents for the various types of sirup produced. Not only must t h e test be completed in a minimum time, but. the apparatus must be simple. A large number of samples must be sent t o the laboratory t o establish converter conditions, for the starch slurry going to the refinery varies considerably in buffer properties or acid requirements, and these changes must be known immediately if the final sirup is to fall within the specificationsset for the product. The laboratories employed their own titration procedure before Lane and Eynon (56-41) published their classic research on the reducing sugar determination which bears their name. Their introduction of methylene blue (56) as au internal indicator wm instantly received, and if a poll were taken as t o which laboratory procedure contributed mast to progress in this field of analysis, this simple technique would undoubtedly win first place. Figure 1. The laboratories in the industry use a special Demose Sehellbach buret (Figure 1) for the determinrtEquiva tion of reducing sugars by the Lane Eynon proBucedure. Because it is custom-made, some comret ment is necessary.
search Foundation and is used by all member companies. Recent chromatographic analyses substantiate these values on the corn sirups thus far analyzed. Dextrose. In the hydrolysis of starch to dextrose, the usual yardstick of dextrose equivalent may not represent the best converter conditions, for dextrose equivalent is the summation of all reducing substances citleulated as dextrose and may not represent the maximum dextrose content. Two methods for the direct determination of dextrose are used by the industry: Stoinboff (57). Sichert-Bleyer modification (54); and Steinhoff (57), Zerban-Sattler modification (60). Cantor and Smith (17)have proposed a modification of the Siehert-Bleyer methods, employing ceric sulfate. Both methods have proved useful in studying the hydrolysis of starch, although there is a belief that some minor degradation products, with reducing power, formed in the starch hydrolysis interfere with the test unless these are removed by carbon or other defecating agent
The buret has a capacity of 20 ml., of which the first 10 ml. are in the form of a bulb, resulting in a zero reading ahead of the bulb, and the second graduation a t 10.0 ml. below the bulb. The stem of the buret is graduated in 0.1 ml. from 10 to 20 ml. This special buret forces the analyst to choose the concentrsr tion of his unknown sugar to fall within the titration limits of 10 to 20 ml. Since he titrates t o 0.05ml., his precision falls within the limita of 1 in 200 at the lower titer, to 1 in 400 at the highest titer. The standardization of the Soxhlet-Fehling solution is 80 chosen that it falls within the range of 10 to 20 ml., thus establishing conditions wherein the amount of capper reduced is essentially a straight-line funotion.
Figure 3.
I
I
2s
I
30
I
35
I
.o
I
45
I
50
I
55
OEXIROSL E Q Y I W L E N I
Figure 2.
Carbohydrate Composition of Corn Sirup
The carbohydrate composition of corn sirup had been a controversial subject for nearly 20 years when the Corn Industries Reseamh Foundation commissioned Cantor and Moyer (16, $%) to establish the composition according t o the methylation and distillation technique of Cantor. These investigations estahlished a commercial carbohydrste tahle for acid converted sirups. which is shown in Figure 2. This has been officially adopted by the Corn Industries Re-
Iteating Bath
The research work on these methods has shown tho necessity for exact temperature control of the water bath. After erperiments with several modifications of water baths boiling at atmospheric pressure (99" to 100" C.), the conclusion was reached that the bath must be improved in order t o obtain precision. The bath adopted (Figure 3) bas a large water reservoir, heavyduty stirrer. and two heating elements, one of which functions when a cold Bask is introduced into the bath. The bath contents are water and ethylene glycol, enabling the bath to he kept a t 100" =t0.02" C. without. boiling. Before this bath w a ~ adopted, variations in analysis were obtained which were attributed to bath temperature, as the variation in boiling water baths between laboratories was as high as 0.7"C . COLOR
Starch. The starch industry dies not have an instrument to m w u r e the color of starch. Several instrument makers have s u p w t e d apparatus for this purpose, but all of them have eer-
1132
ANALYTICAL CHEMISTRY
t a b drawbacks, from the standpoint of either price or design, which makes an experimental trial period unattractive. The instrument designed by Gillett and Meads (925) for the measurement of reflectance of white sugars appears to have considerable merit. The time-honored method of measuring the color of starch is t o place two small portions of starch on a glass microscope slide, compress the two portions together by means of a second glass, and visually compare the whiteness of the two samples. Another useful test is to depress a quantity of starch by means of a pencil and examine the space cone 80 produced. This simple test shows up the yellowness of starch and gives quick visual evidence as t o whether the color pasaes a mental standard.
bottoms. The agreement between observers as to results obtained under these circumstances was far fromsrttisfactory. One of the variahles in the measurement of color in corn sirup is the amount of white light absorption. This characteristic is easily observed in samples of corn sirup. Some appear to have brilliance or sparkle; others appear dull. The author solved this problem by a modified Duboscq colorimeter, shown in Figure 4. There are two metal circular disks containing small circular sections of Lovibond tintometer glasses: Upper disk: 0.05, 1.00,2.00,3.00, Series 510 yellow Lowerdisk: 0.1,O.Z. 0.3. 0.4.0.5,C.B. 0.7,0.8,0.9, Series 510yellow
In addition, a rectangular slide contains glass standnrdfi: 0.1.0.2,0.3-Series
200 Red
I n operation, one 5-inch color tube is filled with dibtilled water; the other, with the unknown sirup. The lowest color slide is introduced above the water tube, and the mirror is moved back and forth until the light intensity is the same for both fields. If no matoh is obtained, the next color slide is introduced, and the operation is repeated until the color is matched. It i8 generally necessary to introduce red into the color a t about 0.8 yellow. Twomethoda areinusein theindustry, basedon spectrophotometers employiog the Lovibond system as the reference. Seallet (51)published a method for the photoelectric determination of the color of corn sirup. The method call8 for the use of the standard, shadowless bottom color tube (6 X 1 inch) in a Lumetron photoelectric colorimeter. The test is performed a t 420 mil and the results are reported either in per cent transmittance or the corresponding Lovibond values-sum of yellow and red values. Thia I . ... i n s t r u m e n t has been 7 ,~ - - * I *, . - /! used bp one refiner forr numher of years. -7Smith (56) has pro. posed a method of coloi I measurement which hw I been in use in the Corr 1-111 Products Refining Co I laboratories for severaI T L n .cp"L'I.L.6 -n..-4:*I J"L"'". "f the color is based on the RUBBER 7"BlNL SIB'' Lovibond t i n t o m e t e r color system, although ' the readings are mado an ,.,I+. I ,,+"0. a Coleman spectrophotometer employing 2-i--G A S " and 4-em. cells. The Figure 5. Buel Fluidity o p t i o a l dendies, obFunnel tained a t 650 and 450 (corrected) mfi, are suhtraoted, reduced t o a 1-em. basis, and multiplied by 160. This value is then referred t o a table which translates it to the usual Lovibond reading. ~~
,
Fipure 4. Modified Duboseq Colorimeter
Corn Sirup. The author believes that measurement of color should be divided between two types of apparatus: (1)apparatus for manufacturing control and (2) apparatus for research work. Apparatus for control work must be simple and rugged, and carry a low maintenance Go&. The primary purpose of auch color measurements is t o keep production under control. For this reason, if the color apparatus is sufficiently simple to be understood and even operated by the factory supervisor, the problem of converting laboratory data. to factory service is much simplified. The long-standing niethod of reparting the color of corn sirup is the measurement of the color of a &inch column of the sirup against Lovihond standard glasses-yellow 510 and red 200. If the corn sirup is a quality product, the color will not exceed 0.6 yellow. Higher colors, particularly if a small amount of red is introduced, usually indicate t h a t the product is below standard or has been overheated a t some stage in processing or handling. Much criticism has been directed against the Lovibmd system within the industry a6 well as in the American Oil Chemists' Society, which also employs these stmdards. Although many laboratories olaim t o employ or to have employed the Laviband system, this is not strictly correct. These laboratories purchased the old-style glasbes (50 X 17 mm.) and used them in a homemade comparator box, employing a home-made light source, with color tube8 (6 X 1 inch) which in many instances had shadowed
-"---)L -f---
..""""
~
&
VISCOSrrY
Starch. No subject in this industry has received more attention than the determination of the fluid properties of starch pastes. Starch pastes form pseudoplastic systems; nevertheless their consistencies are referred to in terms of either viscosity or its reciprocal, fluidity. Despite repeated proposals for the development of continuous recording viscometem, the simple one-point orifice type persists. When the old refuses to abdicate to the new, there must. be some fundamental reason for its existence. For'this reason, considerable attention is given here to the basic consideration of a method for viscosity or fluidity. These basic considerations m e involved: 1. The apparatus must be simple, rugged, compsrativrly low in cost, and operatad with a minimum of expense. 2. The amount of sample'rcquired should be smsll, not m i > -
1133
V O L U M E 24, N O . 1, J U L Y 1 9 5 2 from the standpoint of collection in the factory hut also from the standpoint of samples received from the outside. 3. The viscosity 01.fluidity of a starch paste is dependent upon the paste preparation and no result, regardless of apparatus, is any better than prepamt.ion of the paste. 4. It is much easier to secure proper paste preparation in the labamtory on a small sample than on a large sample. 5. The trade often uses and desires information on heavy pastes, ranging from 1 pound per gitllon (8.7Y0 by weight) to 2 ponndsper gallon (21.5% by weight).
qnired to operate the mechanism of a recording instrument. This can he largely offset if the energy required for movement of the paddle, cylinder, etc., in the paste is large in comparison to that required for the recorder. In other words, the sample of starch and the resulting paste must he large. This complicates thepastepreparation, as the heat transfer ispoar in heavierpastes.
The fluidity funnels determine the viscosity of the s t m h paste by measurement of rate of flow through an orifice. The orifice is standardized against water, a t a predetermined temperature, which is the same as that used for the starch paste. At least three types of orifice funnels BR used for determining caustic and hot fluidities: (1) B u d fluidity funnel (Corn Products Refining Co. type), (2) Clinton fluidity funnel, and (3) Scott cup. With the Brat two, the milliliters of paste passing through the orifice within the specified time is termed the "fluidity number." With the Scott, the time in seconds required for 50 ml. of paste t o pas8 the orifice is called the "Scott number." The data ohtained from $1 three apparatus carry less precision as the value of the paste approaches the value of water. To control this situation, the concentration of the paste i.l increased t o keep the resulting value in the more precise range of the funnel. This complicates the determination t o one who is not familiar with starch. To illustrate this point in greater detail, the following paste concentrations are used with the Clinton funnel: %Dry S u b s t a n c e S t a r c h by Weight Common
Acid modified Oxidired Oridised (gums)
Caustic Fluidity 1.47 4.26
.. ..
Hot Fluidity 2.77
...
5.10 12.9
A former criticism of the orificetype funnel was the difficulty in reproducing the orifice. This situation has been remedied with the Clinton funnel and the Scott cup. No one would seriously dispute the desirability of a viscometer that would record t.he viscosity of a paste through a range of temperatures over a single temperature value as obtained by the fluidity funnel or the Scott cup. Howes $way8 arise. y is dwaytys re-
- -1
1-+-.
Figure 8.
S c o t t Viscometer
The obvious way to solve this nroblem would he to meed the agitation of the paste, 4gitation results in rupture of the starch granules and the meohanical hreakThe outcome of dawn of the pa&. this situation is always a oompromise between bath temperature rate and type of agitation, and paste concen.UiDITI FUNNEL ND tration. JlDlTY FUNNEL
-
TYPES OF VISCOMETER S
Buel Fluidity (or Corn Products Refining Co.) Funnel. The exact history of this funnel is ohscure, hut it appears to have originated with the Corn Produots Refining Co., and some of its features were described h,y B u d in 1912 (IS) and by Morgan i n 1943 (47). It is shown in Figure 5 . C l i n t o n F l u i d i t y Funnel The Clinton funnel has undergone Iseveral changes since its inception in 1928, although the hasic design has n,It been altered. The originators wislhed to improve upon the Buel funnel by incorporating a longer stern, whtxehy a more constant head was oh tained. The funnel is shown in Figuire 6.
.
FLYIDIT" NUMB ER
The howl and part of the a tem of the master funnels itre of light weight nickel silver. Precis/on-hore elass i..L.:..^. ^F +L" ,lilX^..".^"" " 1. y",,.y6 "L Y r.l "..own 1s cemented inside the stern, and in of ~
I
Figure 7 .
Figure 6 . Clinton Fluidity Funnel
Perfortuauu; UI Clinton F u n n e l s
a , . "
il.c.l"lyL.I
.
1134
ANALYTICAL CHEMISTRY
sufficient length to maintain the master specifications. No stopcock is used and the stainless steel orifice tip is attached to the funnel by rubber tubing. The orifice is 1 mm. long and 1.49 mm. in diameter. In standardizing, the funnel is fillea with approximately 300 ml. of distilled water a t 75" F. The level is lowered to amraximatelv 1 om. below the funnel rim. The funnel is s t a d a d i z e d w6en 200 ml. DBSS the orifice in 55 to 60 seconds, the exact time becoming the-factor for the funnel. This flow is then defined as 100 fluidity. Thus if the amount of paste delivered with the funnel factor time is 120 ml., the fluidity of the paste is 60. IO
The behavior of the Buel and the Clinton funnels employing solutions of known visoosity is shown in Figure 7. The useful range of the Buel funnel is generally given as 20 t o 45. The useful range of the Clinton funnel is generally regarded as from 15 to 70.
i
2
~
187S.L o ~
o
n
2
0
!~ ~~
8,
~~
.b
YI*YIE*
Figure 11. Corn Industries Viscometer Curves for Unmodified Cornstarch 92' C. water bath
I
Fipure 9.
1-
FROM CORN INWST.
VISCOMETER
Improved Viscornetor TlMP JN HINYIES
Figure 12.
Effect of Con'cenwation
Hot Scott Viscometer (43). M t h this orifice viscometer (Figure 8) the paste viscosity is determined hot, which i., of considerable value when starch pastes are used hot. The basic design bas not been altered through the years and it is listed in supply catalogs, for the purchase of laundry starch by the Federal Government is based on specifications established for this equipment. It,s use requires a pasting procedure independent of the viscometer; the visoometer is not lagged and bath agitation and temperature control are manual. The oorrect volume of bot paste must be measured and transferred t o the cup. The faults of the old Scott cup have been removed in a new and improved model developed by the Corn Products Refining Co., shown in Figure 9. The paste preparation and the viscosity determination are combined in one pieoe of e uipment. The paste preparation bath is of large capacity, heavly insulated, and filled with water which is maintained a t vigorous boiling with live steam The starch and water in a nickel beaker are placed in the hath, and the standard paddle is inserted, rtnd mechanically stirred by a constant speed motor until the paste is formed. The paddle is then swung back rtnd the paste held ttt hath temperature for a definite time. It is next poured into the cup, which is equipped with an overflow, so that a uniform head is assured. The cup is contained in a chamber of the bath, heated by the bailing water. The viscosity is determined a t approximately 210" F.
Figure IO. Corn Industries Viscometer
Corn Industries Viscometer. The development of this viscometer was one of the projects of the Corn Industries Researoh Foundation, and its operation and resulting viscosity charts have been described (4-7,34). It is similar to other torque re-
1135
V O L U M E 24, N O , 7, J U L Y 1 9 5 2 cording viscometers in many respects, hut incorporates some new features. The physical appearance of the viscometer is shown in Figure 10. The bath temperature of the viscometer may he varied, but most data have been reported employing a bath temperature of 92" C., which has engendered some criticism, in so far as it does not represent true paste preparation. A st.rip chart on common starch (corn) is shown in Figure 11. The effects an the viscosity maximum produced by fixed bath temperature with variation in paste concentration and by a fined paste concentration with variation in t.emperature are shown in Figures 12 and 13. The type, shape, and magnitude of the maximum produced by a given starch slurry are largely a function of the swelling power
of the starch granules as influenced by the concentration and the rate of heat transfer t o the slurry, and these maxime. do not characterize a starch paste by the usual and accepted practices. Brabender Viscograph (10-12). This viscograph iS similar to the amylograph and is of the torque recording type, employing constant speed agitation (Figure 14). There is no bath in the customary sense. In its place, radiant heat from electric heating elements is transmitted to the beaker containing the paste a t a rate determined by a thermometer in the paste, such that the maximum temperature is always obtained according t o a predetermined heating curve. This system h m merit, but it does not function too well if the paste concentration is too high. Barham Viscometer (3). This instrument is a modified rotating cylinder viscometer, consisting of "doughnut"-shaped cup and a free cylinder, concentric with and suspended in the paste cup. The paste cup is immersed in an oil bath, so controlled that the temperature may be maintained from room temperature to 100' C. The doughnut-shaped cup enables the bath liquid t o contnot more surface of the paste cup, thereby reducing the heat transfer problems encountered in other viscometers. Viscosity is measured by evaluating the restoring torque, which must be applied to prevent the suspended cylinder from rotating. The viscometer and a typical paste curve are shown in Figures 15and 16.
T I M 1N YllUTLL
Figure 13
Effect of Temperature
Pmm Corn Industries riseometer cmses. 5Vo p a t e cd unmodi6od cornstarch. Maximum pasto temperatures
Figlire 15. Barham Viscometer
Figure 14. Brabender Yiscograph
Caesar Consistometer (14,15). The Caesar consistometer is a paddle, torque recording instrument in which the work is recorded m the wattage required for % constant speed motor. The bath is well lagged with excellent temperature control. A schematic drawing of the apparatus is shown in Figure 17 and a typical paste oume in Figure 18. Morgan Paste Photometer (46). This apparatus is essentially the combination of a small pasting bath, where the transparency of the paste is measured by a photoelectric cell and microammeter. The bath contains glycerol. The pasting cell is a test tube equipped with an agitator. The bath temperature is altered from room temperature to the desired paste temperature and the trans-
1136
ANALYTICAL CHEMISTRY
parency is plotted against time. Morgan states that character and degree of modification can be determined more satisfactorily by this equipment than by viscometer measurements. A diagram of the apparatus is shown in Figure 19 and a paste curve in Figure 20. This account of viscometers is not complete, but is sufficient to show the complexity of the problem. However, there still exists a wide gap between any laboratory method and industrial use of starch wherein pastes are produced on a large scale. 36
DATA FROM BARHAM'S VISCOMETER CORN STIRCH
Figure 18.
I
I
50
I
I1
70 DEGREES C.
Figure 16.
1
I
90
o
I
I
IO 20 MINUTES
I
~
3090
1I ro1 ~ I ~ 50I ~ I
J
DEGREES C.
Data from Barham Viscometer
Data on Cornstarch from Caesar Consistometer
portance in the industry for the design of pumps, pipe lines, and vacuum pans, and the adaptation of the product for a particular use. Unfortunately, these data have been difficult to obtain for corn sirup, in that the commercial densities above are pseudoplastic systems which are not conducive to the usual viscosity technique. Early data on the subject leave much in doubt, as the dextrose equivalent and Baume of the sirup are not defined. In 1932, Bishop and Young ( 8 )presented data for a limited range of Baumks and dextrose equivalents, employing the principle of the falling ball. The need for additional data on corn sirup and corn sugar liquors was soon felt and the Corn Industries Research Foundation sponsored an extensive project covering the viscosity of corn sirup and crude sugar sirups in all ranges of commercial dextrose equivalents and BaumCs. In this work, Miller and Mench (46) employed Oswald-Cannon-Fenske viscometers for
Riegel (68) investigated the rate of heating used in the preparation of a starch paste, with reference to the viscosity of the resulting pastes. This work is shown in Figure 21, where curve I shows the effect in attaining the bath temperature of 95" C. in 2 minutes, curve I1 in 1 hour, and curve I11 in 3 hours. Another factor often overlooked by the laboratory worker as well as the consumer of starch is the effect of small quantities of salts in the water used for paste preparation. Wiegel (68) shows this very clearly in Figure 22.
VARIABL
l-3 SPEED STIRRER
;,/rkMM
THERMOMETER
,
SAMPLER TUBE
2 LITER BEAKER
TEST TUBE
CORN SIRUP AND DEXTROSE
COLLIMATING LENS
Viscosity of Corn Sirup. The viscosity of corn sirup, particularly in the commercial BaumCs of 42 to 47, has been of im-
_I-,
L
~ RESISTANCE N
THERMOMETER STREAMLINED MIXER BLADES
,
F
INDICATOR MICROAMMETER RECORDING
Figure 19.
-: ' Figure 17.
Schematic Diagram of Caesar Consistometer
io
HOT PLATE
Schematic Diagram of &IorganApparatus
I 40
DEGREES
Figure 20.
I
I
50
60
r
m
I
80
I
90
CENTIGRADE
Pasting of Cornstarch in Morgan Apparatus
1137
V O L U M E 2 4 , NO. 7, J U L Y 1 9 5 2
TIME1 AT 95.1 IN MINUTES
OEGREES C
Figure 21. Effect nf Heating nn Viscosity
c
DKTILLED W4TER
N1100.000 HAC1
N 100000B4CL2
L
n Y
>
TAP WATER
20
I I
I I
40
60
I I
80
II 95
DEGREES C
I
30
I 60
TIHEIAT95.)
Figure 22.
I
90
I
120
I 150
IN MINUTES
Effect of Salt on Viscosity
viscosities under 1000 centipoises and a falling ball type for viscosities above this value. Unfortunately, certain irregularities in the data have been found which necessitate new data or a replotting of the original data to find and eliminate the sources of error that have developed. Dextrose. The determination of viscosity of dextrose solutions has presented no difficulty. Powell ( 4 9 ) has published data on solutions up to 50% by weight with a temperature range from 25’ to 50” C. These data are also given in the International Critical Tables (28). ACKNOWLEDGMENT
The author wishes to acknowledge the assistance and criticisms of the members of the research laboratory in the preparation of this paper. LITERATURE CITED
Bechtel, W.G., Cereal Chem., 24, 200 (1947). Bechtel, IT. G., TextileInds., 113, 93 (1949). Bechtel, W.G., and Fischer, E. K., J . Colloid Sci., 4, 265 (1949). Bechtel, IT. G., and Kesler, C. C., Paper Trade J., 125, 35 (1947). Bishop, IT. B., and Young, K., Ind. Eng. Chem., 24,1171 (1932). Blom. J., and Rosted, C. O., Acta Chem. Scand., 1, 32 (1947). Brabender, C. W., Muhlenlab., 7, 121 (1937). Brabender, C. W.,hlueller, G., and Heide, F., Milling, 90, 696 (1938). Brabender, C. W,, hlueller, G., and Koster, A., 2. ges. Getreide Muhlen- u. Bdckereiw., 24, 168 (1937). Buel, H., Orig. Com. 8th Intern. Congr. Appl. Chem., 13, 63 (1912). Caesar, G. V.,Ind. Eng. Chem., 24,1432 (1932). Caesar, G. V., and Moore, E. E., Ibid., 27, 1449 (1935). Cantor, S. >)I., and Moyer, W.IV.,Abstracts of Papers, 104th Meeting, AlfERIcAs CHEMICAL SOCIETY, 1). 1R. Buffalo, K.Y.. 1942. Cantor, S. SI., and Smith, R. J., private communication. Cleland, J. E., Evans. J. W.,Fauser, E. E., and Fetzer. W. R., IND. ENG.CHEM., ANAL. ED.,16, 161 (1944). Cleland, J. E., Fauser, E. E., and Fetser, IT. R., Ibid., 15, 334 (1943). Cleland, J. E., and Fetser, W ,R., Ibid., 14, 27 (1942). Ibid., p. 124. Ibid., p. 127. Evans, J. W., and Fetser, R. R., Ibid., 13, 855 (1941). Fauser, E. E., Cleland, J. E., Evans, J. W.,and Fetser, W.R., Ibid., 15, 193 (1943). Gillett, T. R., and Meads, P. F., ANAL.CHEM.,24,829 (1952). Hellman, N. N., Cereal Chem., 28,79 (1951). International Critical Tables. Vol. 2, p. 347, New York, hlcGrawHill Book Co., 1927. Ibid., Vol. 5, p. 23, 1929. Jackson, R. F., Bur. Standards, Bull. 13, 633 (1916). Jackeon, R. F., Bur. Standards, Sci. Papers 293 (1916). Jackson. R. F.. J. Wash. Acad. Sci.. 6. 530 (1916). Kerr, R. IT.,“Chemistry and Industry of Starch,” p. 379, iYew York, Academic Press, 1950. Kerr, R. TV., Cleveland, F. C.. and Katsberk, W. J., J . Am. Chem. Soc.. 73,111 (1951). Kesler, C. C., and Bechtel, W.G., IND.ENG.CHEaf., AXAL. ED., 19,lG (1947). Lampitt, L. H., Fuller, C. H. F., and Goldenburg, N.,J . SOC. Chem. I d . , 60,25 (1947). Lane, J. H., and Eynon, L., Ihid., 42, 32T (1923). Ibid., p. 143T. Ibid., p. 463T. Ibid., 44, 150T (1925). Ibid., 46,434T (1927). Ibid.. 50.85T (1931). Lansky,’L., Kooi, M e , and Schoch, T. J.. J . Am. Chem. SOC., 71,4066 (1949). MacNider, G. S l . , J . Ind. Eng. Chem.,4,417 (1912). Meyer, K. H., Bernfeld, P., Boissanas, R. A , , Gurlter, P., and Noelting. G.. J. Phvs. & Colloid Chem., 53, 319 (1949). Miller, B.D., hlench,-J. W., Degering, E. F., and Newton, R. F., “Viscosity of Corn Syrups,” Lafayette, Ind., Purdue University, 1946. Morgan, W. L., IKD.ESG. CHEM.,ASAL.ED.,12, 313 (1940). Morgan, IT, L., and Vaughn, N. L., Ind. Eng. Chem., 35, 233 (1943). Pigman, W. IT.,and Wolfrom, hl. L., Adiances i n Carbohydrate Chem., 3, 129 (1948). Powell, C. W.R., J . Chem. Soc., 105, 1 (1914). Sair, L., and Fetzer, W. R., ISD. ENG.CHEW..AXAL.ED., 14, 843 (1942). Scallet. B. L., AKAL.CHEM.,20,591 (1948). Senti, F. R., private communication. Shaffer, P. A., and Williams, R. D., J . Biol. Chem., 111, 707 (1933). Sichert, K., and Bleyer, B., Z. anal. Chem., 107, 328 (1936). Sjostrom, 0. A., J . Ind. Eng. Chem., 14, 941 (1922). Smith, R. J., private communication. Steinhoff. G., 2. Spiritwind., 56,64 (1933). Wiegel, E., Ibid., 50, 62 (1933). Zerban, F. W., and Martin, J., J. Assoc. O ~ CAor. . Chemists, 27,295 (1944). Zerban, F. IF’.,and Sattler. L., IYD.ENG.CHEM.,ANAL.ED.,10, 669 (1938).
(1) haec. Offic. Agr. Chemists, “Official and Tentative hlethods
of Analysia,” 1945. (2) Ibid., p. 591, XXXIV, 143.
(3) Barham, H. N., Wagoner, J. A., and Reed, G. N., Ind. Eng. C h a . . 34,1490 (1942).
RECEIvEDfor review October 26,1952. Accepted April 24, 1952. Presented before t h e Division of Sugar Chemistry, Symposium on Sugar Analytical Methods, a t the 120th Meeting of the AMERICAN CHEMICAL SOCIETY, New York, X. I’.
