Feb., 1920
T H E J O U R N A L OF I N D U S T R I A L A N D ENGINEERING CHEMISTRY
t h e solution was titrated with normal HzS04 and required only 31.00 cc., showing t h a t a n acidity of s . 78 cc. N / I HzS04 had developed. T h e more dilute t h e alkali; t h e less marked is this reaction. CONCLUSIONS
T h e rapid formaldehyde method for ammonium nitrate invariably gives low results by direct titration. By using a n excess of alkali and titrating back, less reliable results are obtained, presumably owing t o t h e formation of formic acid. Direct titration with a N / 3 NaOH solution gives results which are consistently about 0 . 2 5 per cent low. It would seem, therefore, t h a t this method m a y be used for accurate work if a suitable correction is made b y adding a constant factor t o t h e results. This factor is obtained by running a Kjeldahl on a representative sample and comparing t h e result with t h a t obtained by t h e formaldehyde method, using N / 3 NaOH.
THE PREPARATION OF HEMATOXYLJN By Percy A. Houseman and Clement K. Swift LABORATORY OP MACANDREWS & FORBES Co., CAMDEN, N. J
Received August 7, 1919
Hematoxylin has hitherto been made almost exclusively i n Germany. During t h e war, a relatively large amount of hematoxylin was required for use in surgery as a bacterial stain, and in this emergency practically none was obtainable. A method was therefore sought t o manufacture pure hematoxylin in commercial quantity for supply t o t h e American government. T h e apparatus developed for this purpose is shown in t h e cut.
I73
T h e raw material used is a powdered extract of logwood which is prepared from a mild aqueous extract of unoxidized logwood chips. T h e powdered extract is made into small briquettes which are broken into pieces and t h e dust screened out. T h e pieces are then filled into t h e extractors A. B are flasks containing ether, surrounded by oil which is directly heated by electrical heaters made of ?\/lone1metal wire wound on slate. T h e ether vapor passing up t h e pipes C enters t h e top of t h e extractors A, is condensed b y t h e copper coils D, and drops back on t h e logwood extract, dissolving hematoxylin from it, a n d enters t h e flasks B again through t h e constant level syphons E . G are mercury manometers from which a n electrical connection is made t o relay H, which cuts off t h e current from t h e heaters if pressure develops in t h e apparatus through failure t o condense all t h e ether. T h e apparatus runs day and night. Every fourth d a y t h e valves I are closed and all of t h e ether distilled from t h e flasks into t h e extractors A. About 1500 t o 2 0 0 0 cc. of cold water are then introduced a t K t o each of t h e flasks. T h e water is heated by t h e oil baths and dissolves t h e crude hematoxylin in t h e flasks. The solution is blown out through t h e tubes L, a n d filtered hot. After standing for several hours pure hematoxylin crystallizes from this solution in colorless needles which are filtered a t t h e pump, washed with ice water and dried on a porcelain tray. The extractors are changed every t e n days, t h e exhausted extractor being transferred t o t h e position M and t h e ether recovered from a hot water bath. T h e apparatus described produces I O O g. of pure hematoxylin per day, t h e comparatively low capacity being caused by t h e slight solubility of t h e hematoxylin in ether which, however, seems t o be t h e only practicable solvent. T h e loss of ether is almost
I
/ I
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
I74
The apparatus functions practically as a large-scale Soxhlet extractor with constant instead of intermittent syphon. For extractors i t has been found very convenient t o use 3-gal. copper fire extinguishers. T h e apparatus described in this article has been found very useful for other continuous extractions with organic solvents when i t is desired t o treat several kilograms of material.
METHOD OF PREPARING PHLOROGLUCINOL REAGENT FOR THE KREIS TEST By Henry W. Dixon MEATINSPECTION
LABORATORY,
BUREAUOF ANIMALINDUSTRY,
SANFRANCISCO, CAL. Received August 26, 1919
T h e Kreisl test has been found t o be a useful chemical test for t h e detection and estimation of rancidity in fats when considered together with t h e taste, odor, and flavor. If a supply of phloroglucinol is not available, t h e Kreis reagent may be conveniently prepared by t h e following method, based on Barth and Schreder’s synthesis of phloroglucinol from resorcinol.2 Dissolve IOO g . of sodium hydroxide i n 40 cc. of water by heating slowly in an iron or nickel container. Cool t o a temperature a t which a few crystals of resorcinol dropped into t h e liquid do not become brown, and stir in 1 5 g. of resorcinol. Cover with a watch glass t o exclude air and a t t h e same time to permit observation, and heat for 2 or 3 hrs. in a sand b a t h until the melt is a dark chocolate-brown. (If t h e melt is heated too fast or too long, charring will t a k e place with low yield of phloroglucinol. If t h e reaction is not carried to completion, unchanged resorcinol will interfere with t h e color of t h e Kreis test.) Cool, still excluding air. Dissolve t h e melt in about 500 cc. of water, heating gently if necessary. Transfer t o a beaker, cool in ice and cautiously add concentrated hydrochloric acid until t h e solution is acid. Extract repeatedly by shaking with IOO cc. portions of ether in a separatory funnel. Combine t h e ether extracts and bleach by shaking with animal charcoal. T h e filtered ether extract should be kept as a stock solution. T o prepare t h e Kreis reagent dilute a portion of t h e stock solution with ether until it gives a good reaction with a strongly rancid f a t b u t a less intense reaction on further dilution. Fifteen grams of resorcinol ordinarily yield I O 1. or more of the Kreis reagent. This reagent appears t o be a satisfactory substitute for t h a t prepared from pure phloroglucinol. A COLLOIDAL NETHOD FOR INCREASING THE VOLUME OF ADHESIVE WATER-GLASS By J. D. Malcolmson MELLON
OF 1 NDUSTRIAI. RESEARCH, PITTSBURGH. PITTSBURGH, PA. Received May 28, 1919
INSTITUTE
UNIVERSITY
OF
Commercial silicate of soda, or water-glass, with 4 2 O B&, is used i n very large quantities as a n adhesive in t h e fiber container industry. a density of about
THrs JOURNAL, 10 (1918), 471. Ber., 12, 503.
1 Kerr, 2
Vol.
12,
No.
2
In f a c t , this material is t h e only adhesive employed for gluing together t h e components of both doublefaced corrugated board and laminated solid fiber board. During a recent emergency created by freight embargoes and other war-time causes of delay, t h e stocks of water-glass in t h e box-makers’ hands began t o run low. This laboratory was a t t h a t time confronted with t h e problem of increasing t h e volume of t h e 42 O B6. adhesive without injuring its essential properties. These properties, which make water-glass of especial value t o t h e box-board industry, may be classed as follows: ( I ) Adhesive value. (A powerful adhesive is not required, but it must be equal to or greater than the “felted” strength of the paper fibers ) ( 2 ) Proper viscosity. (This will be discussed below.) (3) Low penetration. (This prevents loss of adhesive through absorption by the relatively porous corrugated strawboard. This property obviously depends upon the viscosity.) (4) Fast setting and drying rate. (The boards are run through the pasting machines at a speed of upwards of 60 It. per min.) ( 5 ) Relative insolubility when dry. (6) Relative flexibility when dry. ( 7 ) Negligible effect of age. (8) Relatively low cost. (This is necessary because container board is a product of relatively low cost.) Water-glass a t about 4 2 ’ BC. is a fairly viscous liquid. High viscosity is one of its desirable features for use in this industry, a n d is only limited b y t h e fact t h a t it must be fluid enough t o be pumped t o various parts of t h e factory and t o flow over t h e troughs a n d rollers as fast as needed. High viscosity prevents loss b y absorption a n d insures a satisfactory drying rate and adhesive power. Initial attempts*to increase t h e volume of a given amount of t h e adhesive were made b y diluting with water. This naturally lowered t h e viscosity t o a very marked ext e n t and t h e drying rate and adhesive power were correspondingly lowered t o a prohibitive degree. Much of this adhesive, moreover, was absorbed b y t h e strawboard. These experiments clearly showed t h a t t h e desirable properties of water-glass as a n adhesive are, in general, a function of t h e viscosity, and as this is lowered these properties suffer proportionately. T h e problem was finally solved b y diluting t h e water-glass with brine.l It was found possible b y this means t o increase t h e volume of t h e solution b y 25 per cent without lowering t h e original viscosity or detracting from any of t h e essential qualities of t h e adhesive as listed above. This process was perfected on a laboratory scale a n d subsequently tried out on a large scale in several factories. I t was found t o work satisfactorily, t h e only disadvantages being: ( I ) The possibility of a drop in the market price of water-glass. (2) The increased floor and storage space required. (3) The installation of equipment. Water-glass is a complex mixture including sodium silicate (probably hydrated), water and silica (or silicic acid) in t h e colloidal state. T h e actual proportions of NazO and Si02 t o form t h e compound NazO.xSi02, and t h e proportion of Si02 i n t h e colloidal form, are questions about which much uncertainty 1 The writer desires to acknowledge his indebtedness to Dr. E. W . Tillotson, Jr , assistant director of Mellon Institute, for his suggestions and assistance in the development of this method.
