INDUSTRIAL A N D ENGINEERING CHEMISTRY
February, 1927
215
Salts i n Successive Portions of Bone Char-Treated Solutions (50-cc. portions analyzed)
-
SAMPLE C a H I ( P 0 h
4
I~~~~~
NatC03
KazSOc Ca(C3H302)z NasCaHsOi
CaCh
KNO3
NaC1
Per cent
Per cent
Per cent
Per cent
Per cent
Per cent
Per cent
0.06
1.40 1.60
1.80 1.80
2 04 2 40
n
07
Per cent
SUCROSE Per cent
POLARIZATION FOR INVERT
SUGAR
Per cent
T7.
2 2.
The results in the accompanying table are given in the order of maximum adsorption in the first portions, 'except those for sucrose and invert sugar. The difference in adsorption is so great that any seeming inconsistencies do not materially influence the value of the results. Various unknown factors made it seem that the method of analysis was sufficiently accurate. The amount of water adsorbed by different grades of char will vary largely the apparent adsorption of the salts. The fact that char will adsorb substances a t one concentration and give them up at another, which has been proved before, was shown in several filtrations of a raw sugar solution through a 64-foot column of char. The invert sugar in the first portion of one solution was 0.63 per cent; in later portions it rose to 1.56
per cent and finally dropped back to 1.29 per cent, the original percentage in the solution. There is undoubtedly exchange adsorption between the solutes and the char, and also among several substances adsorbed a t one time. Hence the analysis of the portions already off may vary largely from that of the portion coming off, especially when a refiners' char containing many chemical substances is used. The refiner's chief question is not about the composition of the filtered material but about the effect upon crystallization of the sucrose by the quantity of other substance present in solution. It is apparent that complete purification of raw sugar is practically impossible by use of bone char alone.
Water-Resistant Animal Glue' By F. L. Browne and C. E. Hrubesky FOREST PRODUCTS LABORATORY, MADISON, WIS.
A
S A result of a recent investigation, the Forest Products
Laboratory has been successful in making animal glue water-resistant by utilizing the well-known tanning action of formaldehyde.2 The study had for its immediate objective the development of a water-resistant glue for woodworking purposes, particularly for use in naval aircraft, but the results obtained may well prove capable of wider industrial application. The problem in making use of the tanning action of fornialdehyde on glue arises in the control of the speed of the reaction in such a way as to permit the introduction of enough formaldehyde to tan the glue sufficiently and yet keep the mixture for a reasonably long period of time fluid enough to be applied by means of the usual machinery of the glue room. If a batch of animal glue is prepared in the customary manner by soaking the dry glue in the requisite proportion of cold water and melting by warming it to 60" C., and enough formalin is then stirred in to produce an irreversible jelly, the reaction takes place so rapidly that local coagulation result's before the formalin can be distributed unifornily through the glue sol. Lindauer3 has found that by using, instead of formalin, the equivalent amount of the formaldehyde polymer, paraformaldehyde, a mixture can be obtained which remains in a workable condition for B quarter of an hour to an hour (the period depending upon the proportions used and the Presented before the Division of Leather and Gelatin Chemistry at the 72nd Meeting of the American Chemical Society, September 5 to 11. 1926. 3 The vexed question of the nature of the reaction between formaldehyde and the glue proteins will not be discussed here. For a review of the theories see the ( B r i t . ) Adhesives, Research Committee First Regort, 1982, 118; see also Reiner and Marton, K o l l o i d - Z . , 88, 273 (1923); Thomas, et al., J . A m . Leather Chem. Assoc., 81, 71 (1926). 3 U. S. Patent 1,506,013 (1924).
temperature) and then changes quickly to an irreversible jelly. Plywood joints made with such a glue while it is in the fluid condition are water-resistant. Although glues of working life as short as this have been successfully used in woodworking industries, they are never likely to meet with general favor. Means of further extending the working life of the treated glue were therefore sought. The investigation was based upon the following theory : At constant temperature the rate of gelation of a formaldehyde-containing animal-glue sol is a function of the concentration of free formaldehyde. I n order to introduce enough formaldehyde to tan the glue thoroughly and yet retain a long working life, it must be added in some indirect fashion so that, while an ample amount is potentially available for reaction with the glue proteins, only a very small proportion of it is actually present as free formaldehyde a t any given instant. Three such indirect devices have been studied which utilize: 1-Formaldehyde polymers. 2-Hydrolyzable compounds of formaldehyde. 3-Reversible formaldehyde adsorption complexes.
Furthermore, in view of the great influence of the pH of its solutions upon the properties of glue and the fact that the equilibrium between free formaldehyde and its polymer or compounds is usually altered by changing pII, it was to be expected that the water resistance and working life of glues containing tanning agents of these types would be affected materially by the further addition of acids or bases. Experimental Methods
Ground glue, of a grade suitable for wood joint work, was soaked in cold water in a proportion of 100 parts by
I S D U S T R I d L 3 S D ESGILVEERILVGCHEMISTRY
216
weight of glue to 225 to 250 parts of water. When the water had been completely imbibed, the glue was melted a t about 60" C. by placing the container in a water bath. The temperature of the sol was then adjusted to the desired point and the tanning agent and any auxiliary substance added, usually as dry powders. As soon as they were dissolved ()r uniformly distributed through the sol bv stirring. -
5'01. 19, s o . 2
casein glue call for an ayerage joint strength of not less than 250 pounds per square inch when dry and 125 pounds per square inch when wet, procedure being very similar to the one described. Results
Tvpical results obtained with a number of tanning agents __ including polymers, compounds, and a&or$ion comdexes of formaldehvde. are shown ir, Table I. The substances listed tan the glue fairly thoroughly and react very much less rapidly than formalin. As would be expected, the working life observed with different tanning agents varies considerably, presumably because of differences in the concentration of free formaldehyde in the solutions with which they are in equilibrium. Methylene-ditannic acid and formaldehydesodium bisulfite did not tan glue in neutral solution; in thece cases it is probable that the compounds in question are in equilibrium with a smaller concentration of free formaldehyde than is characteristic of the tanned glue. The addition of sodium hydroxide to a glue mixture containing formaldehyde-sodium bisulfite breaks up the compound and causes the immediate coagulation of the glue. I n general, the addition of acids to glue sols containing any of the tanning agents fisted in Table I prolongs the working life but sometimes =.>.. r1gur'e 1-t Y y w o o d T e s t i n g M a c h i n e , Specia11 Grips. a n d T e s t results in diminishing water resistance of the S p e c i me n plywood joints d d d k o n of alkalies, as a rule, shortens the working life. the glue was considered ready for use. The period between T a b l e I-Results w i t h Different T a n n i n g A g e n t s that moment and the time when the viscosity began to inwater 225 to 250 grams, tanning agent as crease very rapidly as gelation set in was recorded as the (G!ue formula: animal glue 100,indicated) working life. The change from sol to gel took place so VORK. 'LYwoon JOINT TANNING -&GENT quickly that, even with the crude method of testing the conAUXILIARY ING STRENGTH MATERIAL J F E A' sistency by stirring with a rod, duplicate determinations of Kind Amount Wet 300 c. D r y the working life agreed within about 10 per cent. I Mzn. :bs./ Lbs./ An animal-glue mixture which gels a t a qonstant temperaGvarns Grams utes q. zn rp. zn. ture above the normal melting point of the untreated glue Polymers of Formaldehyde: 1 is said to be tanned, and is presumably much more resistant Commercial paraformalde- 20 4i5 113 hyde 5 to water than ordinary animal glue. It by no means follows, Same, heated 1 day a t 41 . . . ... 1000 c. a however, that it is tanned sufficiently to be "water-resistant" Same, heated 7 days a t 100 , . . ... 1000 c. 5 in the sense in which that term is applied to glue by the 25 ... ... a-Polyoxymethylene 5 modern woodworker-i. e., retaining sufficient strength @-Polyoxymethylene 41 ... ... J Mixture @a n d y-polyto hold glued joints together under severe exposure to mois90 ... ... oxymethylenes 5 ture. The water resistance was therefore measured by the Compounds of Formaldehyde: following method, which has long been standard procedure Hexamethylenetetramine E 300 333 46 125 426 82 Oxalic' aiid, 1 a t the Forest Products Laboratory: 570 62 I
I
I
'
0
'
One-sixteenth-inch birch veneer, selected for straight grain and freedom from defects, was cut to a size of 12 by 12 inches and stored in a conditioning room a t 80" F. and 60 per cent relative humidity until required for use. A test panel was made by gluing three of the 12-inch sheets together with the grain of the core piece a t right angles t o that of the face plies, as in the usual plywood construction. Glue was spread on both sides of the core piece a t a rate of about 27 sq. cm. per gram of glue solution. Five t o 10 minutes were allowed to elapse before applying a pressure of about 200 pounds per square inch. Usually a bundle of several panels was placed in the press. Panels were removed from the press the next day and stacked with stickers between them in the conditioning room for 7 days for seasoning. Ten test specimens of the dimensions shown in Figure 1 were then cut from the central portion of each panel. Five of them were tested to destruction while in their dry seasoned condition, and five were submerged in water for 2 days, after which they were tested while still wet. Several plywood panels were made with each glue tested, usually at different intervals during the working life of the mixture.
U. S. Navy Department specifications4 for water-resistant 4
U. S. Navy Dept. Specification 6 2 0 8 , July 1, 1924.
Formaldehyde-aniline Methylene-paratoluidine Formaldehyde-urea
16 4.3 8.5
.-
Borax, 1
...
NaOH.' 0.56
Oxalic acid, 5 . 5
120
482
73
15
...
62
'b
10 7 14
Formaldehyde-starchb 16.5 Formaldehyde-sugarc 20 Formaldehyde Adsorption
376 505 318 457 519
330 180 50 250 100 110 20 15 5
664
532 507 409 421
Sa
185 40a 21a 4lQ 35 60 20
15
Complexes:
CH20 adsorbed by charcoal CHlO adsorbed by silica re1
16.5
30
a Panels were not water-resistant unless pressed in the hot press 1 5 minutes a t 100' C. b Made b y method of German Patent 201 436 (1907). c Made b v method of Heidushka and Zirkel, Arch. Phorm., 2 6 4 , 467 (1916).
Formaldehyde C o m p o u n d s as T a n n i n g Agents
With most of the substances studied, the tanning action on the glue continues, although slowly, even if the glue sol
February, 1927
ISDCSTRI.1L A S D ESGINEERISG CHEMISTRY
is chilled to room temperature. Thus if a mixture containing hexamethylenetetramine or formaldehyde-urea, while in the fluid condition, is chilled to room temperature, it gels just a5 the untreated glue sol would do. If it is warmed to 60" C. again before the lapse of the period of its working life, it melts, remaining fluid until the time come3 for it to gel as a result of the tanning action. But if the chilled jelly is kept a t room temperature for a period longer than the working life, it cannot be melted again even on heating in a bath of boiling water. Formaldehyde-aniline and methylene-ptoluidine behave differently. Evidently the tanning action fails to continue when the glue mixture containing these materials is cooled to room temperature, because such jellies may be kept for a week or more and will melt again on being warmed to 60" C., remaining fluid for the remainder of their interrupted working life. Such tanning agents are not well adapted t o the production of a commercially practicable water-resistant glue because the rapid chilling of the glue upon being applied to the wood would stop the tanning reaction. I n the case of plywood panels this difficulty could be overcome by using a hot press, but hot-press gluing has not hitherto proved economical in American practice. The values for water resistance of plywood joints obtained with the formaldehyde compounds studied thus far are not as high as can be obtained with casein glue or with animal glue tanned with formaldehyde polymers. It should be remembered, however, that the water-resistance test by soaking plywood specimens is exceedingly severe because of the high stresses in the glue 47 joint resulting from-the Glue Formula opposition of the directhe core veneers. A glue that succeeds in holding the plywood together at, all under these circumstances r e p r e sents avery marked imp r o v e m e n t in water resistance as compared with ordinary animal glue. F o r m a l d e h y d e Adsorption Complexes as Tanning Agents
Activated c o c o n u t shell charcoal and silica gel adsorb formalde4 8 IZ I6 24 hyde in considerable Oxalic a o d - grams Figure 2-Influence of Proportion of a m o u n t s . Such adOxalic Acid Added on Working Life of a sorption complexes can G l u e Containing Paraformaldehyde be i n c o r p o r a t e d i n Upper curve-Glue held at 40° to 4 5 O C. Lower curve-Glue held at 50° to 5 5 O C. animal glue mixtures and will tan the glue slowly, giving water-resistant plywood joints. Formaldehyde Polymers as Tanning Agents
Paraformaldehyde proved capable of imparting t o animal glue a high degree of water resistance. It has certain other advantages over formaldehyde compounds and adsorption complexes. It is completely converted to formaldehyde on depolymerization, whereas only a fraction of a formaldehyde compound is potential tanning material. Moreover, the polymer is cheaper a t the present time than the compounds. The disadvantage of using paraformaldehyde is the fact that the commercial article of that name is not a single chemical
217
species or even a definite mixtu're, but supposedly a variable mixture of polyoxymethylenes of differing reactivity. The best results thus far, from the point of view of conibining practicable working life with a degree of mater resistance meeting requirements equal to those of the Navy Specification, have been obtained by using oxalic acid in conjunction with paraformaldehyde. The following factors hare important bearings on 9 the properties of the treated glue: I-Nature of the paraformaldehyde. 2 - P r o p o r t i o n of paraformaldehyde. 3 - P r o p o r t i o n of oxalic acid. 4-Temperature during working life. 5-Concentration of glue. 6-Grade of glue. x.4TURE
8\
7-
6-
5-
a \
OF THE
PARAFORMALDE HYDE -Table I shows the variation in working life of an animal glue containing 5 grams of various polymers of formaldehyde per 100 g r a m s of d r y glue. The so-called a-, p-, and r-polyoxymethyl30 70 ao enes, when u s e d in Temperature - d e g r e e s C. equal proportions, tan Figure 3-Relation between Working a n d Temperature of t h e Glue for Two glue less rapidly as we Life Glues Containing Paraformaldehyde, One proceed from the CY to Containing Oxalic Acid a n d t h e Other Not the y. The commercia1 paraformaldehyde gave a working life of similar length t o that with the a-compound, but heating it in a closed container a t 100" C. for several days reduced its rate of reaction to that of a mixture of p- and ypolyoxymethylenes.
\
in Working Life w i t h Different Commercial Paraf ormaldehydes (Formula: animal glue 100, water 225, paraformaldehyde 10, oxalic acid 5.5 grams)
Table 11-Variation
SOURCE
WORKING
LIFE
AT 46'
c.
Hours Laboratory stock Factory No. 1 Factory No. 2 Factory KO. 3 Factory KO.4
6.5 5 3 2.75 2.5
Table I1 gives the working life of glues containing oxalic acid and samples of paraformaldehyde obtained from different commercial sources. The variation is between two- and threefold. On heating the more reactive samples a t 100" C. for a week they became similar in behavior to the laboratory stock supply. Tests of the water resistance of plywood panels made with glues containing the several paraformaldehydes exhibited no significant variation. Since that is the case, it would be desirable to use the least reactive product in order to secure longer working life. Paraformaldehyde for use in this type of glues should be ground finely enough to pass completely through a 50-mesh screen. The amount required is greatly in excess of its solubility, so that most of it exists in mechanical suspension in the glue mixture during the period of working life. Hence, in order to keep the glue batch uniform, settling of the suspended paraformaldehyde must be prevented. The agitation of the roll spreader commonly used for applying glue is suffi-
INDUSTRIAL A N D ENGINEERING CHEMISTRY
218
cient to accomplish the purpose with 50-mesh material, or finer, provided the glue is of ordinary woodworking consistency. With glue sols of low viscosity more efficient stirring might be necessary. PROPORTION OF PARAFORMALDEHYDE-Since most O f the paraformaldehyde is present in the solid phase in mechanical suspension, one might expect the working life of the glue to be nearly independent of the proportion of paraformaldehyde added. Such is not the case. -4s Table I11 shows, increasing the amount of the tanning agent consistently results in decreasing the working life throughout the whole range of the experiments. The probable explanation is that the free formaldehyde in solution reacts with the glue proteins more rapidly than the paraformaldehyde passes into solution and depolymerizes, and that the velocity of solution of the paraformaldehyde thus controls the speed of the tanning action. I n that case increasing the amount of the solid phase present would shorten the working life by increasing the area of interface a t which solution takes place. Similarly differences in the size of the primary crystal particles, with resulting differences in specific surface, may account for the variation in the reactivity of commercial paraformaldehydes, as suggested by Descude.6 Table 111-Influence of t h e Proportion of Paraformaldehyde (Formula: animal glue 100. water 250 grams, paraformaldehyde as indicated) ~~
PARAFORMALDEHYDE
Grams 0.25 0.50 1.0 2.0
5.0 10.0 15.0 19 8
--
~
~~
W
52 26 15
;
PLYWOOD JOINT STRENGTH
~
~
Dry
Wet
Lbs./sp. in. 430 409 353 399 475 435
Lbs./sq. in.
...
0 0 0 0 113 136
Bull. SOL. chdm., 29, 87 (1908).
f
Table IV-Influence of Grade of Glue ~ animal glue 100, water 225,paraformaldehyde 10, oxalic acid 5.5 (Formula: grams) GRADBOF GLUE"
Viscosity
Jelly strength
Millipoises
Grams 372 299 240 190
...
118
. .
The water resistance first becomes sufficient to hold the plywood specimens together in the soaking test when about 2 parts of paraformaldehyde are added to 100 parts of dry glue, but to insure more satisfactory water resistance it is advisable to use about 10 parts of paraformaldehyde to 100 of glue. PROPORTION OF OXALICACID-The relation between the working life of paraformaldehyde-containing glue and the proportion of oxalic acid added is shown in Figure 2. The working life increases to a maximum as more acid is added and then rapidly falls off. Analogous curves can be obtained with any acid. With most acids it is found that the water resistance is diminished by the presence of the acid. That is especially true when strong mineral acids are employed. Certain organic acids, including citric and benzoic acids when used in sufficient proportions to lengthen the working life materially, impair the strength of the glue joint even when dry. With increasing proportions of oxalic and mucic acids, however, the strength and water resistance are not adversely affected until the point of maximum working life on the curve has been passed. TEMPERATURE DURING WORKINGLIFE-The great significance of the temperature a t which the glue is kept during its working life is shown by Figure 3. A temperature of about 60" C. is recommended for ordinary animal glue as a condition a t which harmful bacterial growth is inhibited and hydrolysis of the glue not serious. With the paraformaldehyde-containing glues it is desirable to use a temperature of 40" to 45" C. in order to obtain a longer working life and retard glue hydrolysis still further. Bacterial decomposition is prevented by reason of the antiseptic properties of the 8
tanning agent. Since the temperature coefficient of the working life is so great, a careful control of the temperature is necessary. Premature gelation in the glue spreader would not only waste glue but would involve much labor in cleaning the tough jelly out of the machine. CONCEKTRATION OF GLUE-The proportion of paraformaldehyde to glue remaining constant, the working life decreases as the concentration of glue increases. Since the concentration of glue is determined, for the woodworker a t least, chiefly by the necessity of maintaining the proper viscosity to suit the conditions of use, this factor cannot be varied greatly as a means of controlling the working life. GRADEOF GLm-The grade of animal glue employed, as measured in terms of viscosity and jelly strength, has much less influence on the working life and water resistance than might he expected. Within the limits of grade likely to be used for woodworking, the lower grade glues in general seem to gel more quickly as a result of the tanning action of paraformaldehyde than glues of higher grade. It is clear, however, that the behavior of a given glue depends chiefly upon factors other than the grade. Five glues covering the range in grades suitable for woodworking purposes were supplied especially for this study by a large glue manufacturer. The results obtained with them are given in Table IV.
~
~ Minutes 180 60 53
VOl. 19, KO. 2
95 76 65
-._
55
I 1 I
1
'IpE
Minutes 425
1
1
1 I 1
PLYWOOD TESTS Wet JOINT T E S T S b Dry
Lbs /
Lbs:/
sa. in.
sa. zn.
449 463 452
ji: 155
Lbs./ Per Cent sa. i n . woodfazlure 3560 31 3600 36 3385 38
...
' The grades of the glues were measured in terms of viscosity and jelly strength following the methods of the National Association of Glue Manufac. turers, as described in THISJOURNAL, 16, 310 (1924). b Joint tests are made by parallel-grain gluing of selected, seasoned hard maple blocks, conditioning thoroughly after gluing, and testing to de: struction in shear. The proportion of the broken area in which failure took place in the wood is estimated and recorded as "per cent wood failure.'' The test specimen and testing tool are described in detail in the Navy Specification cited. Recommended Formula
It is believed that the formula given below is practicable for woodworking uses where it is desirable to employ an animal glue and yet provide water resistance. For the last year the formula has been subjected to a thorough laboratory investigation carried on with semicommercial-scale glueroom equipment. It has not yet, however, been tried out under actual factory conditions. FORMULA
P a r i s b y weight 100 Animal gluea 225 Water 10 Paraformaldehyde Oxalic acid 5.5 Equivalent to National Association of Glue Manufacturers' Grade 12 or 13, or to Peter Cooper's Grade I or IX. If a higher or lower grade of glue is used the proportion of water in the formula should be increased or decreased accordingly.
Soak the glue in the water until the latter has been imbibed, then melt a t 60" C. Lower the temperature to between 40" and 45" C. and add the remaining ingredients. As soon as the oxalic acid dissolves the glue is ready for use. Enough agitation must be provided to keep the paraformaldehyde in suspension-a condition which will probably be met by most glue spreaders. The working life of the glue a t 40" to 45" C. is from 7 to 9 hours if a suitable type of paraformaldehyde is employed. Joints made with this glue
I.VDiY:STRIAL AND ENGINEERING CHEMISTRY
February, 1927
should be sufficiently water-resistant to meet requirements embodied in the Kavy specifications for casein glue. In parallel-grain gluing with hard maple lumber the joints should be strong enough to tear the wood when tested to destruction. The proportion of water used in this formula can be altered sufficiently to give any viscosity required for ordinary woodworking purposes. In testing plywood panels made with animal glue containing paraformaldehyde for mater resistance it is imperative
2 U
1
Glue formula animal due
loo
219
perature the tanning of the glue is a slow reaction. Since the values for water resistance given in Tables I, 111, and IV were obtained with panels conditioned for 7 days before testing, they underestimated somewhat the degree of water resistance which the glues are capable of developing. No attempt is made in this paper to draw a close comparison between the water resistance of the glues discussed and that of commercial types on the market. Both are undergoing constant improvement and their water resistance therefore is not a fixed property. For most woodworking purposes both types are sufficiently water-resistant, and choice between them will be governed by differences in properties other than water resistance, and by the relative costs. Conclusions
I
1
I
I
4
Seasoning Parlod
-
I 8
I 10
I
I
I2
14
I 16
Ik
days
Figure 4-Shou ing t h a t t h e Paraformaldehyde-Containing Glues Develop Their Water Resistance Slowly
that a seasoning period of at least 7 days, or preferably longer, be allowed after removing the panels from the press and before submitting specimens to the soaking test. -4s shown in Figure 4, the full water resistance of the glued joints is not developed for many days, presumably because at room tem-
1-Through the use of (a) formaldehyde polymers, ( b ) hydrolyzable formaldehyde compounds, (c) formaldehyde adsorption complexes, it is possible to introduce enough formaldehyde into animal glue sols to render glued joints water-resistant and a t the same time to retard the reaction sufficiently to give a working life practicable for woodworking. 2-The working life of such glues can be further extended by the addition of acids in small amounts, but often at the sacrifice of some of the water resistance. Oxalic acid extends the working life without loss in water resistance. 3-A glue of practicable working life and sufficient water resistance to meet requirements embodied in Navy specifications for casein glue is made by adding to a batch of animal glue, as ordinarily prepared, 10 parts by weight of paraformaldehyde and 5.5 parts of oxalic acid per 100 parts of dry glue.
Some Interesting Sugarhouse Incrustations’ By John W. Schlegel and J. P. Manley XEW YORK REFINERY. NATIONA SUGAR L REFININGCO., LONGISLAXD CITY. N. Y.
T
HE char filtration plant of a modern refinery is a fertile field for the investigation of sugarhouse scales, deposits, and incrustations. Frequently further investigation of incrustations which upon casual examination seem of least importance leads to results of prime importance t o the refining process. The incrustations discussed in this paper are representative of these commonly found. So far as the writers are aware, no analyses of scales of this character have yet been published. Char Filter Deposit
Several years ago it was noticed that in the filtration of high remelt sirups over bone black a heavy, sticky sludge ultimately formed upon the upper surfaces of the filters, retarding the flow of liquor through the filters and later affecting the efficiency of the sweetening-off process. Investigation showed that, although the high remelt sirups from the centrifugal machines were brilliant. a flocculent precipitate appeared upon later dilution and heating preparatory to char filtration. This precipitate was later found deposited upon the upper surfaces of the char filters. A portion of this smear from the char filter tops was washed free from sugar, dried, and analyzed, with the following results : Presented before the Division of Sugar Chemistry a t t h e 72nd Meeting of the American Chemical Society, Philadelphi?. Pa., September 5 to 1 1 , 1926
.
Organic and volatile..
Si02 ....................
.....................
SOa FerOa
...
+ AlzO3. . . . . . . . . . .
Per cent 37.60 44.05 5.48
3.99
Per cent PZOJ. . . . . . . . . . . . . . . . . . . 1.41 CaO . . . . . . . . . . . . . . . . . . . 4.66 MgO . . . . . . . . . . . . . . . . . . 2.82
--
TOTAL.. . . . . . . . . . . . . . . . 100.00
This scale evidently consists mainly of silica, combined with calcium sulfate, iron and alumina, and probably basic magnesium silicate. The very interesting feature of this deposit is its high percentage of silica, due probably to the use of diatomaceous earth in the clarification of the washed sugar liquors. During such clarification, especially if an excess of lime is used, small quantities of silica are dissolved from the earth, the degree of solubility depending largely upon the excess of lime used. Although normally, only mere traces of silica are dissolved, sometimes so much is dissolved that, upon the subsequent concentration of the impurities in the lower grade sirups, their limit of solubility is reached and the silica separates out. This is what seems to have happened in this case, for with more careful control over the defecation the deposition ceased. Liquor Pipe Incrustations
Scales found in the outlet pipes from low-grade char filters, used for the filtration of raw sugar washings through bone black, were analyzed, one lot about six years later than the other. These scales ultimately practically close up the pipes. In fact. it is the resultant slow flow of t>heliquor