THE JOURNAL OF INDVSTRIAL A N D ENGINEERING CHEMISTRY
Oct., 1918
809
BOLIVIANIMPORTS, ETC.(Concluded) ARTICLES 1913 I915 MINERALOILS AND THEIRDERIVATIVES.. . . . . . . . . . $ 86,403 $234,630 20,533 7,288 Germany 21,116 118,683 United States VEGETABLE OILS, EDIBLE.. 14,124 17,870 Italy 4,536 6,590 United States 521 4,048 OILS, OTHER.. 47,232 8,456 Germany ...................................... 14,823 48 11,146 6.254 United States.. SHEETAND F'LATE GLASS.. 46,664 18,337 Germany. 21,078 121 United States 5.129 9.437 GLASSWARE.. 103,845 2 1 , 518 Germany 53,500 3,232 United States 6,576 6,261 PAPER AND CARDBOARD (EXCEPT WALL PAPER). . . . . 167,391 91,012 Germany 70,353 6,598 United States.. 15.742 28.790 MANUFACTURE OF PAPER(EXCEPT BOOKS)... . . . . . . . 31,914 14,138 Germany 15,562 1,629 United States.. 414 1,516
AMERICAN PRODUCTS S O L D I N BOLIVIA(Concluded) 1917 ARTICLES 1914 Glass and glassware. .............................. $ 7,895 $19,469 Grease, lubricating.. .............................. 5,464 1,933 India-rubber manufactures. ........................ 31,618 5,024 Leather, patent.. ................................. 1,475 Oils: Refined mineral: Gasoline, etc.. ................................ 1,998 4,494 Illuminating.. ................................ 22,199 16,647 36,026 20,583 Lubricating, etc.. ............................. Naphthas, etc.. ............................... 1,998 59,797 Veeetable : 6,123 Cottonseed, . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 2,353 Linseed ........................ ............. 233 6,862 187 Allother ..................................... Paints pigments etc.: Ready-mixed i a i n t s . . ........................... 2,647 1,236 White lead.. ...................... 23,688 7,095 ........... 66,405 ........................... 83,227 85,045 ........................... 720 3,685 1,830 Perfumeries, cosmetics, etc . . . . . . . . . . . .
How the war has affected what little trade American manufacturers have with Bolivia can be ascertained from t h e following table, which is compiled from American statistics for the fiscal years 1914 and 1917:
Sugar, refined
..................................... ................................. ...................... ......................................... ................................. .................................. ............................... ....................... .................................... .................................. ................................... ..................................... ................................. ..................................... ............................... ..................................... ...............................
AMERICANPRODUCTS SOLDIN BOLIVIA ARTICLES 1914 1917 Blacking, shoe paste, etc. $ 1,136 $ 4,642 Candles .......................................... 3,841 Cement, hydraulic.. Chemicals, drugs, dyes, etc.: Acids 119 4,423 Calcium carbide.. 573 23,792 13,;;; Medicines, patent and proprietary . . . . . . . . . . . . . . . . . Soda salts and preparations.. 4.;t)i) Allother Explosives : 6,622 23,768 Cartridges, loaded Dynamite. ................................... 72,044 Gunpowder.. 12,556 .............................. 26 42,217 All other..
.......................... ............................... .......................................... ............................... ..................... ....................................... ............................... ....... ...................................
.......
... ...
2:;i$i
...
...
(.
................. ................. .....................................
7,050 330 6
15,914 5,781 817
The statistics supplied by our own Government do not show many imports from Bolivia, as shipments are made from the Pacific and Atlantic ports of neighboring countries and consequently credited t o them. Of t h e nearly $3,000,000 worth of t i n ore imported into t h e united States during h + e. fiscal year 191,, practically all was credited t o Chile in our statistics. Such imports jumped t o $IO,OOO,OOO in 1918. Imports of bismuth were valued a t $196,000 in 19x7, of which $32,000 worth is credited t o Argentina, Chile, and Uruguay, and is obviously of Bolivian origin. The preliminary statistics for the fiscal year 1918 do not show imports of bismuth.
ORIGINAL PAPERS VALUATION OF RAW SUGARS BY W. D. HORNE
I n the sugar trade it has long been customary t o buy and sell raw sugars on their polarization, adopting usually a baeis of 96' for centrifugal sugars and 89' for muscovado and molasses sugars. For every degree above the basis a certain additional increase is paid, while for each degree below the basis a double deduction is made. This system, while based on practical considerations a n d having much t o recommend it, is still far from satisfactory, because it does not take sufficient account of the many influences on refining of sugars introduced b y their endless variations. Efforts have therefore been made in recent years t o attempt a rough standardization of raw sugars on t h e part of some of the advanced manufacturers. Such methods consist of grading the sugars according t o the size of grain, hardness of grain, cleanliness of solution, odor, and reaction, as well as polarization and moisture, with perhaps some other determinations. Admirable as these efforts are, they still fall far short of what is desired, for they are based on assumptions which frequently are not borne out in practice and entirely overlook many important varia1 Read before the Division of Industrial Chemists and Chemical Engineers, 56th Meeting of the American Chemical Society, September 10 to 13, 1918.
tions in raw sugars which radically affect their value for refining purposes. It is the object of this paper t o direct attention t o the practical considerations involved in valuation of raw sugars for refining purposes a n d t o suggest some amplifications of the tests applied, with the .hope t h a t i t may lead to a full discussion of the subject and in the belief t h a t closer attention t o the points involved must inevitably lead t o greater efficiency of manufacture with a consequent decrease' of wasteful operation in both manufacture and refining. The refining value of a raw sugar depends upon its content of sucrose and the availability of t h a t sucrose, as modified by the nature and quantity of the impurities. The nature of the impurities will determine the ease of their separation from the sucrose during refining. Refining consists principally of ( I ) affination or washing the residual mother liquor of the raw sugar niassecuite from the solid grain of the sugar; ( 2 ) defecation and filtration, t o remove insoluble substances a n d some soluble impurities from the solution of the washed sugar or from the dilute washings; (3) decolorization boneblack or other means; (4) crystallization and separation of pure sugar from t h e other constituents. The response of any raw sugars t o tests for the first three of these operations can be determined readily
810
T H E J O U R N A L O F I N D U S T R I A L A N Q E N G I N E E R I N G C H E M I S T R Y Vol.
by operations herein described, giving a far closer grading of t h e sugar in its refining value t h a n b y t h e da5a a t present supplied. The proposed tests are novel only in some details and in their immediate application t o a precise valuation of raw sugars on a strictly mathematical and therefore scientific basis. The object of washing the raw sugar is t o remove t h e film of low-testing molasses from t h e surface of t h e grains. A 96" polarizing sugar is usually crystallized from a solution having a purity of something less t h a n 80, and t h e separation of t h e sucrose in crystals leaves a mother liqudr of between 60 and 7 0 purity. Purging t h e massecuite in centrifugals removes, i t is true, a large part of this syrup, b u t unless t h e sugar crystals are' sprayed with water or a sugar solution in t h e centrifugal machine, an appreciable amount of this mother liquor remains adherent t o t h e crystals. As the object of refining is t o separate t h e impurities from the sugar as promptly and thorouglfly as possible, the first operation of refining consists of washing this residual mother liquor from t h e faces of t h e sugar crystals, and t h e condition of the crystals is a factor of great importance in this purification. An even grain of large size is t h e easiest t o cleanse. Small crystals present relatively more surface than larger ones, carrying more syrup, requiring larger amounts of wash water, and present g:eater resistance t o t h e purging of the sugar. They also dissolve more freely in t h e wash water or sugar solutions used for mixing and washing, thus decreasing t h e yield of washed sugar a n d increasing t h e amount of washings. This carries relatively pure sugar of€ into impure washings and increases t h e labor of its recovery. Small grained sugars are slow t o purge and are a detriment t o rapid work. Clustered ,grain is another objectionable feature as these conglomerates hold a certain amount of low mother liquor which refuses t o wash out. The purity of the grain itself is of fundamental importance for if t h a t is not of high test, no amount of washing will yield a washed sugar of t h e desired quality. All raw centrifugal sugars should be boiled from clean solutions of sufficient purity t o insure a nucleus of t h e purest type, ranging over 99 in test. This can best be done b y starting grain on concentrated juice, building u p later if need be with syrups or mother liquors from other sugars. If a secondor third sugar is used as seed grain t h e resulting crystals will contain more impurity and can never be washed up t o t h e highest purity. One of t h e best attempts a t regulating t h e production of raw sugar so as t o meet reasonable requirements of refiners, has been made by the Cuban-American Sugar Company. Their tests include size of grain, taking a diameter of a little'less t h a n a millimeter as standard; hardness of grain, as observed when rolled between t h e t h u m b and finger; odor, divided into normal, musty, fruity, and sour; cleanli' ness, as indicated b y t h e milligrams of ir;lsoluble mat-
.
IO,
No.
IO
ter per I O O g., or b y t h e turbidity of 5 g. dissolved in 2 5 cc. of water, as well as polarization and moisture. The hardness of grain is a t best a questionable factor. I t s solubility is what counts, and t h e smaller i t is t h e more it dissolves. The refiner wants t o know what yield will. be had of washed sugar in washing any particular sample and how pure the washed sugar will be. The more directly and accurately these facts can be determined, t h e more useful t h e information will be. I n working out this question t h e matter of first importance was t o adopt t h e proper liquid for,washing the raw sugar. Water is not suitable as it dissolves too much of t h e grain. A pure sugar solution has the objection of tending t o obscure the difference between high and low sugars through adding t o each approximately t h e same absolute amount instead of the same relative amount of pure sugar. W h a t I have found t o be more satisfactory is a saturated solution of the raw sugar itself. This has the advantage of dissolving none of the grain and removing practically all of t h e mother liquor from t h e crystals, and finally of altering t o a minimum degree t h e purity of t h e washed sugar. The only other method would be t o use an alcohol washing method or its equivalent, or t o use a syrup or molasses obtained b y working back and washing several times, both of which are too slow. The low molasses also becomes too viscous t o use t o advantage. The method used is t o mix I O O g. of t h e raw sugar with 45 cc. of water, shake I O min. t o saturate, let settle, and decant 92 cc. of the resulting solution upon zoo g. of t h e raw sugar. This is well mixed into a magma and purged in a small laboratory centrifugal. I n t h e experiments here referred to, a 5-in. cyclone centrifugal has been used, with which about min. is needed t o get up full speed. Two minutes' steady running after this is all t h a t is needed, making about 9 revolutions of the handle shaft each I O seconds. The sugar is thus made comparatively dry, and t h e yield of washed sugar is found b y weighing the basket with t h e purged sugar in it. The sugar is then removed, dried z hrs. a t 98' C. in a water-jacketed air bath, and polarized. A good centrifugal sugar, not mixed with seconds, will polarize about 9 9 . 4 , which, of course, is the purity of t h e washed crystals, and t h e ordinary centrifugal sugars, such as constitute the greater part of t h e present-day supply, will have a purity of a t least 99". This is in exact accord with refinery practice, showing t h a t t h e method above described gives as pure a washed sugar as t h a t obtained under working conditions, where clear water is used in t h e final washing in t h e centrifugals. Washing an ordinary 96 " centrifugal sugar yields washings of 7 5 " t o 80" purity, or thereabouts, and i t is highly desirable t h a t all of t h e impurities possible should be separated from t h e grain and forced into t h e syrup, for these two are handled separately and each additional pound of impurities t h a t hangs back in t h e washed sugar necessitates just so much extra work for its final elimination. A pound of impuri-
Oct., 1918
T H E J O U R N A L OF 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
ties will lose about 0 . 4 of its mass in t h e boneblack, and carry with its remainder about 0.4 lb. of sugar into residual syrup, having a purity of about 40'. I n washings of 80' purity only about 2'/2 boilings or crystallizations are necessary t o effect this separation, whereas in washed sugar of 98' purity 6 or 7 boilings and crystallizations with all the correlated operations will benecessary t o eliminate t h e impuritie2:. I n a washed sugar of 98' purity, therefore, as compared with one of 99' purity taken as standard, there will be I per cent of t h e mass which will have t o be boiled, etc., four extra times. It follows t h a t each I per cent of impurities above t h e normal I per cent left in the washed sugar costs the refiner I per cent x 4 X t h e cost of boiling, etc., X price per lb. of raw sugar above t h e basic price of raw sugar, and should be valued accordingly. Conversely, a sugar washing up t o 99.5O purity under standard conditions is 0 . 5 per cent X 4 X boiling cost X price per lb. of raw sugar less expensive t o refine t h a n the normal and should be valued in accordance. The second point t o be taken into account is t h e amount of defecation required and the speed of filtration. These two items are rather closely related, as no raw sugar solutions will filter clear unless defecating material be added t o floccblate the fine suspended impurities into masses sufficiently large t o be caught in t h e meshes of t h e filter cloth or other medium employed. Raw sugars require differing amounts of defecant and consume widely varying lengths of time for filtration. It has always been customary t o neutralize any acidity with lime added in a thin cream. Formerly the flocculation required was obtained b y adding blood from slaughter houses t o the solution before heating to t h e coagulating point of serum, then raising t h e temperature and causing a strong flocculation. Later, phosphoric acid and acid phosphate of lime came into vogue and these, neutralized with lime, proved very satisfactory when using filtering bags. But the rapid work done in t h e beet sugar industry in filtering carbonate of lime in filter presses could not be followed in sugar refining on phosphate of lime because this precipitate is soft and chokes up t h e cloths when subjected t o t h e high pressure necessarily used in t h e filter press. Now, in the past few years, a suitable defecant for use in filtering sugar solutions through filter presses has been found in the great deposits of infusorial earth at Lompoc, Cal. One vein of this deposit, after special processes of treatment, is now being largely used, under t h e name of Filtercel, b y nearly all the refiners of the country. But despite the most careful treatment in defecation, some raw sugars will filter very slowly, delaying the operations of a refinery sometirnes in a n exaggerated and very costly manner. Such sugars are commorily called gummy, and frequently indeed they give an excessive amount of precipitate when treated, in solution, with an excess of alcohol and a little acetic acid. These sugars are in fact very common and their objectionable quality is quite evidently due t o the improper defecatiop of raw
81I
cane juice, a matter t h a t is capable of nearly complete correction, and t h a t at almost no added expense. Some raw sugars, on t h e other hand, filter very slowly, although containing no excessive gummy matter, probably from high sulfates, from clay, and other causes. I n order t o determine t h e relative filterability of a raw sugar, it has been found best t o use a 4 5 per cent solution defecated with the minimum amount of acid calcium phosphate and made slightly alkaline with a standard sucrate of lime. Into 18.I cc. of this solution, representing I O g. of raw sugar, introduce 0 . 2 cc. of a solution of acid calcium phosphate made u p t o contain I per cent Pz0b. This will represent 0 . 0 2 per cent PzOc on t h e dry sugar. Then add saccharate of lime standardized against the P205 solution so as t o balance it volume for volume when using phenolphthalein, until t h e solution is just faintly alkaline t o litmus. This is heated t o boiling in a test tube and allowed t o settle while inclined a t 45'. Other tubes are rapidly prepared in succession, using 0 . 0 3 per 0.04 per cent, and so on, noting the minicent PzOS, mum amount which gives a clear supernatant liquor. After thus determining the minimum amount of defecant t h a t will give a clear solution b y this preliminary trial, one adds t h e indicated amount of acid phosphate and lime t o IOO cc. of t h e 45' Brix sugar solution, heats gradually t o 190O F., lets stand half a minute off the hot surface, and pours slowly in a small stream upon t h e top of the triple thickness of a 6-in. bag filter cloth folded like a filter paper in a 3-in. perforated brass cone, with about 6 2 5 holes per sq. in., setting loosely by means of attached lugs, in a vulcanite funnel with no stem. I n about "4 minute, when the transfer is complete, t h e cloth is closely covered with a watch glass. The time is observed which is required for the filtration of 7 0 cc: If the cloth is of t h e right structure this filtrate will appear clear. Excessive amounts of defecant retard rather then aid filtration. I n t h e refinery the column of liquor constantly rises in t h e bags, while in this test the column, after the short period of introduction, constantly sinks, so t h a t conditions are quite different, and while they cannot be directly compared, there is found t o be this relation-that the slower the solution is t o filter in t h e refinery the longer t h e time required for 7 0 cc. t o pass through the small filter cloth. I n sugar solutions t h a t filter freely this test portion will run through in 5 min. or less. I n medium sugars from 5 t o I O min. will be taken, while poor sugars require I O t o 15 min., and bad samples take from I j t o 2 0 min., or even longer. As a very slow filtering sugar may easily increase t h e time of filtration in the refinery t o 1 2 0 per cent of the normal, it follows t h a t time lost may be approximated by adding 7 per cent t o t h e normal time for each additional 5 min. in the experimental filtration test. The percentage of excess time multiplied by the price of a pound of raw sugar multiplied b y t h e cost of normal filtration will give approximately the correction t o be deducted from the basic price
812
T H E J O U R N A L O F I N D U S T R I A L A N D E N G I N E E R I W G C H E M I S T R Y Vol.
t o arrive at t h e refining value of a slow filtering sugar. The third item t o be taken into account as affecting the value of a raw sugar is its readiness t o yield u p its color on filtration through boneblack or on subjection t o any other decolorizing process. After all, refining sugar is principally decolorizing it, and any system of standardization or valuation of raws which ignores this important feature is lacking in one of t h e essential details. Sugars may be of a wide range of shades and from many causes, as t h e variety of cane, burnt cane, caramelization during manufacture, over-liming, working back molasses and second or third product sugars, contamination by iron salts, and so on. Some of these coloring matters are more easily absorbed t h a n others and their absorption b y different agents varies widely. For instance, t h e natural color of cane juice is only slightly absorbed by boneblack, while Norit absorbs it quite freely; and this latter agent takes u p 9 5 parts of color due t o t h e action of lime on invert sugar as easily as it absorbs 2 3 parts of color due t o caramel. These and other considerations render a n empirical test desirable, based on general common practice. Such a determination of t h e decolorability of a raw sugar b y boneblack, for instance, may be arrived a t by dissolving I O g. of raw sugar in 30 cc. of water, adding 0 . 2 j g. of Filtercel and 2 g. of t h e best boneblack ground finer t h a n 60 mesh, bringing all gently t o a boil, and filtering through paper. A similar test, made as the first is, but without boneblack, affords t h e basis of comparison. After reading the colors of t h e filtrates either against a tintometer standard or by comparing the depths of columns t o give equal colors, one may readily calculate t h e amount of color absorbed b y the boneblack. I t will be foyrrd t h a t a fair average sugar will yield about 7 5 per cent of its color in this test, and any greater amount yielded means a proportional economy in char work required, while a smaller absorption designates a larger amount of char work t h a t will be required. One can easily calculate the amount t o be added t o or deducted from the basic price of a sugar t o arrive a t its value in respect t o filterability. Thus a sugar giving up only 6 0 per cent of color instead of 7 j per cent should have l s / a ~ of the normal char filtering expense deducted from the basic price t o recompense for the extra expense t h a t will be entailed in its char filtration. Other factors might be taken into account, as t h e amount of ash in t h e raw sugar, but as under present conditions it is of less importance how much melassigenic ash there is t h a n how much time and labor will have t o be expended in refining t h e sugar, these factors may, for t h e present, be disregarded. The extra refining expenses enumerated in t h e above examples are very small, i t is true, and would occur in relatively few cases, b u t with upward of $600,000 worth of raw sugar entering t h e port of New York alone, daily, even small decimals add up t o large aggregates and are certainly worth taking into account. Just now there is so ready a market for all raw sugar
IO,
NO.
IO
t h a t competition in its sale is slight, b u t when t h e present stress is over and Europe resumes her large production there will be a great surplus, with corresponding competition t o sell. Then the purchaser will pick and choose what suits him best, and i t is t h e part of caution for the raw-sugar maker t o consider what class of sugars will be most desired and t o manage his manufacture accordingly. It is in t h e hope of assisting in this very particular discrimination t h a t these suggestions are presented. YONKERS, NEWYORK
ON THE PREPARATION OF AN ACTIVE DECOLORIZING CARBON FROM KELP1 By F. W. ZERBAN AND
E. C. PREELAND
Dr. J . W. Turrentine, in charge of the United States Experimental Kelp Potash Plant a t Summerland, California, has for several years been engaged in working out methods for the commercial utilization of t h e giant kelps of t h e Pacific Coast. During the course of his investigations it occurred t o him t h a t t h e char obtained in the manufacturing process used might perhaps be converted into a decolorizing carbon. It appears, however. t h a t this question was not taken up actively, until one of t h e authors of this article conceived t h e same idea, while engaged in a study on carbons t h a t might be used in t h e cane sugar industry. At his request, Dr. Turrentine sent him some dried kelp t o experiment with. I n .the first test t h e kelp, after thorough drying and grinding, was carbonized in an iron retort provided with an outlet for gases, until no more fumes were given off. The char was then transferred t o a closed iron receptacle and heated for 2 hrs. t o a bright red heat. It was then cooled, boiled out with hydrochloric acid, washed with water, and dried. Upon examination it was found t h a t the resulting carbon reduced t h e color of a molasses test solution t o about one-third of t h a t obtained b y using a n equal quantity of our standard carbon, Norit. A sample of kelp char, also received from Dr. Turrentine, when treated in a similar manner as t h e dried kelp, produced only a very poor carbon. We therefore decided t o investigate this matter more thoroughly, a n d at our request Dr. Turrentine very kindly furnished us an ample supply of raw material for our further experiments, and we wish t o express t o him our thanks for this courtesy, as well as for t h e great interest he has taken in t h e progress of our work. The material received consisted of three different samples. The first, A, was kelp (Macrocystis p y r i f e r a ) dried in a rotary kiln; t h e second, B, was “incinerated” kelp, prepared as described below; and t h e third, C, was a sample made b y subjecting kelp t o destructive distillation. The last sample was kindly sent t o us through Dr. Spencer, who was investigating t h e destructive distillation of kelp at t h e Forest Products Laboratory, Madison, Wisconsin. None of t h e three samples had been leached with water. 1 Presented before the Division of Industrial Chemists and Chemical Engineers at the 56th Meeting of the American Chemical Society, Cleveland, September 10 t o 13, 1918.
