rnary System NicotineC. 0. BADGETT Eastern Regional Research Laboratory, Philadelphia 18, P a .
A three-phase system was noted in previous work on the distribution of nicotine between water and kerosene. This study was undertaken to determine whether a highly concentrated nicotine phase could be obtained by proper balance of solvent volumes and temperature within the range normally used for recovery of nicotine. The behavior of the ternary system at various temperatures is described. Phase diagrams of the system at 64", 6 7 O , and 70" C. are given. A partial space model plotted from the data shows the conditions under which three phases exist. A highly concentrated nicotine phase can be produced. The ratio of solvents and temperature required to produce such a phase is out of the range normally used for recovery of nicotine. URING studies a t this laboratory on distribution of nicotine between water and kerosene (1), occasionally a three-phase system was observed. This system was studied more closely in the hope that it might afford a means of obtaining a nicotine-rich phase from dilute aqueous nicotine solutions which would be useful in a process for recovery of nicotine. Time was not available for collection of data on the complete system, but it was decided that the data on hand should be reported for reference purposes.
WI
PRBC EDURE
NICOTINE
A 52-nd. glass-jacketed buret was connected to a corivtant
,Figure 1. Partial Space Model Diagram of the ThreeComponent System: Nicotine-Water-Kerosene
temperature bath. The buret was fitted n i t h a thermometer and water was pumped through the jacket by means of a small centrifugal pump. Fifteen milliliters of an aqueous nicotine solution DISCUSSIOR were placed in the buret and 15 ml. of kerosene were added. The Figure 1 is a partial space model diagram of the three-compokerosene was Deobase-a commercial, specially purified, and nent system-nicotine-water-kerosene. The apexes represent deodorized grade. The nicotine was Durified by vacuum distillation. The buret was stoppered and rocked at a constant temperature for 15 minutes to establish equilibrium. It TYas AND VOLCME OF PHASES TABLE I. COMPOSITION then set in an upright position to allow Composition of Observed Volumes of separation of the layers TKOor three Phases a t Equilibrium'" Phases a t Equilibrium Original Composition phases formed and the volumes of the of System Upper, Middle, Lower, Upper, Middle, Lower, wt. % mt. % wt. % ml. mi ml. phases were recorded. At 70' C A new temperature was selected and 62.68 38.66 16 24 V. Nicotine 37.64 the buret was again set in motion. Ob83.01 33.74 23.7 4.7 2.5 Water 18.53 0 75 servations were made on the phases 9.75 3.58 Kerosene 43,82 60.59 existing a t this temperature. The procAt 67O C . ess was repeated until the system passed W. Nicotine 3 3 70 22.46 6.33 1.9 Water 22.47 from two-phase to three-phase and back Kerosene 43 82 to two-phase again. The transition 59.76 35.66 19.35 X. Siootine 37.61 temperatures were recorded. 75.43 38.89 22.5 4.0 4.3 0.63 Water 18.53 5.22 1.35 43.82 63 71 Kerosene *kt certain temperatures the three Y . Xcotine 41.57 phases were separated by drawing them Water 14.60 22.6 1.9 6.3 off through a stopcock on the buret and Kerosene 43.52 their compositions were determined. At 6 4 O C. Kicotine was determined by the silico24.55 53.73 32.74 2. Nicotine 37.64 70.11 45.63 21.5 2.4 6.7 18 .53 0.48 Water tungstic acid procedure, water by the 5.34 0.64 43.82 66.78 Kerosene Karl Fischer method, and kerosene by a Volume at room temperature was 30 ml. Apparent increase in total volume was due to expansion difference. Nicotine was found to exhibit a t higher temperatures. b T t . % of sum of all three components. no interference to the determination of water by the Karl Fischer reagent. I
2370
October 1951
INDUSTRIAL AND ENGINEERING CHEMISTRY KEROSENE P
2371
KEROSENE D'
i
80
Figure 2.
Phase Diagram of Three-Component System at 70" C.
Figure 3.
Phase Diagram of Three-Component System at 67" C. KEROJENE
100% of any one component. Plane FGHK cuts vertically through the triangular prism a t 43.82% kerosene. Temperature is represented on the vertical axis and the percentage of nicotine on the horizontal axis, Planes NPO, N'P'O', and N"P"0" are normal to plane FGHK a t 70", 67", and 64" C., and lines CC', BB', AA' are a t the intersections of these planes, respectively. The pointa M , U , and L lie on plane NPO; M', U t , and L' are on plane N'P'O'; and M " , Ut', and L" are on plane N"P"0". The enclosed curve ABCDC'B'A'E on the 43.82% kerosene plane, FGHK, represents a vertical cross section of a solid figure, a portion of which is represented by MCLM'U"L" and its periphery represents the transition points from two to three phases in this plane. Three phases exist in equilibrium within the solid figure MULM"U"L'. The composition of an upper phase a t any particular temperature is represented by a point on the line UU"a t that temperature. Similarly, the compositions of a middle phase and a lower phase are represented by a point on the lines M M " and LL", respectively. Only one or two phases can exist outside the solid figure partially shown here. Table I gives the original composition of the over-all system, the composition of each phase, and the observed volume of each phase. These weight per cent compositions are plotted on the three plane figures NPO, N'P'O', and N!'P''O''. For simplicity and accuracy these planes are shown separately in Figures 2, 3, and 4, respectively. Figure 2 is a phase diagram a t 70" C. A system of the composition represented by point V was heated to this temperature, whereupon it separated into three phases. Analyses of the three phases were made for each component as shown in Table I. The composition of the upper phase is represented by point U, the middle phase by point M , and the lower phase by point L Similarly, systems of compositions W , X, and Y in Figure 3 gave phases of compositions U', M ' , and L' a t 67" C., and the composition Z in Figure 4 gave phases of compositions W " , M " , and L". At 64' C. the three compositions represented by 8, X , and Y in Figure 3 and shown in Table I were chosen to illustrate the changes in volume of the phases with composition a t 67" C. Between temperatures of 64" and 70" C., the upper phase consists of a 32 to 39% nicotine solution in kerosene with very little water. The middle phase consists mainly of a 16 to 25% nicotine solution in water with a small amount of kerosene. The lower phase consists of a 34 to 45% solution of water in nicotine with very little kerosene. The bulk of the nicotine available in the ternary system is found in the upper or kerosene-rich phase. The amount of nico-
jy
20
Figure 4.
BO
Phase Diagram of Three-Component System at 64" C.
tine in this phase is not as great as the amount of nicotine found in the phase containing the bulk of the nicotine in the two-phase water-nicotine system ( 2 ) . Only a trace of water is present in the kerosene-rich phase of the ternary system. In cases where water is undesirable, the ternary system offers an advantage over the two-phase water-nicotine system where water is present in both phases in appreciable amounts. Any system of a composition represented by the three-phase
TABLE 11. OBSERVED TRANSITION TEMPERATURES FROM Two TO THREE PHASES Nicotine in Systema,
%
Transition Temps., O
75.5 79.0 79.3 79.5 79.4 78.9 77.6 75.3 73.4 72.5 70.0 67.4 66.8 65.5 5
Kerosene held oonstant at 43.82%.
c.
75.5 68.6 68.9 68.4 68.1 66.0 64.4 62.5 61.6 61.3 61.2 61.7 62.5 65.5
2372
INDUSTRIAL AND ENGINEERING CHEMISTRY
triangles of Figures 2, 3, and 4 will separate into three phases of compositions represented by the apexes of these triangles. According to Gibb’s phase rule, F = C - P 2, there can be 4 degrees of freedom for any single phase and as temperature and pressure are fixed, the concentration of two of the three components in the phase adequately describes the system. the solid figure in Figure 1 where three phases exist, there can be 2 degreees of freedom, and as pressure is fixed temperature adequately describes the system. Table I1 shows the observed tramition temperatures between
+
Vol. 43, No. 18
two and three phases. These Kere used to plot the closed curve ABCDC‘B’A’E’ shown in the space model. LITERATURE CITED
(1) Claffey, J. B.,Badgett, C. O., Skahmera, J. J., and Phillips, G. T.T’. M., I N D . E S G . C H E M . , 42, 166-71 (1950). ( 2 ) Hudson, C. S.,Z. p h y s i k . Chem., 47, 113 (1903). RECEIVED February 23, 1951. The appearance of trade names in this manuscript does not constitute a recommendation by the Government over any other brand not mentioned.
esoreinol-Fo
a1
OBSERVED BY ULTRAVIOLET ABSORPTION MEASUREMENTS PAUL J. STEDRY Minnesota Mining & Manufacturing Co., S t . Paul, >Ifinn.
Synthetic fibers used in reinforcement of molded rubber goods are usually treated with an adhesive which is in part a condensation product of resorcinol and formaldehyde. This product is prepared in dilute water solution, using sodium hydroxide as catalyst. Because established methods of kinetics cannot be readily applied to control the manufacture, a convenient and accurate index of reaction velocity was sought. Ultraviolet transmittance of the solution decreases gradually and reproducibly as the condensation progresses. The change in transmittance was used to compare reaction rates between 1 mole of resorcinol and 1, 2, or 3
moles of formaldehyde in the temperature range from 55” to 110” F. Temperature coefficient and activation energy were found to be independent of the molar ratio, and the reaction time to a given end point could be expressed as an exponential function of temperature. The effects of various catalysts and concentrations were similarly studied. Transmittance data in the near-ultraviolet thus provide a tool for the rate study of phenolic condensations which take place in a single phase. Before this method can be applied to theoretical w-orli, however, several error-introducing factors must be appraised more thoroughly.
I
ing viscosity changes during condensation. In ensuing ycars, Russian chemists (7-10, 19) made extensive use of relative viscositr, refractive indeu, and electrical conductivity data to follow the formation of phenol-formaldehyde resins catalyzed by acid or alkali, and thus n’ere able to calculate some fundamental phr;sicochemical constants. Simultaneously in England, Megson and collaborators ( d J 1 I , 1 2 ) introduced the concept of “resinification time,” measured from start of reaction to appearance of permanent turbidity and taken as a measure of reactivity of the phenolic mixture employed. As an extension of this method, Finn and Rogers ( 3 ) followed the initial phenol-formaldehyde condensation stage by observing the temperatures a t which the reaction niixtuic just becomes turbid. Still another method of rate study is the determination of phenols by broniination, used by Sprung ( 1 4 , l i ; ) and other authors. The only spectrophotometric approach to the kinetic study of phenolic condensation was briefly described by Engeldinger (W), who measured light absorption a t 490 mp during acidcatalyzed reaction b e k e e n resorcinol and formaldehyde. He ] m a thus able to iollow micelle growth up to the time of flocculation, and to draw conclusions regarding the effect of p H and of conceiitration of reactants.
N USING continuous synthetic filaments for reiniorcing
molded rubber goods, it iq necessary to pretreat the fiber with an adhesive to assure good bonding to rubber. An adhesive commonly used for this purpose, known as the RFL treatment, is prepared by partially condensing 1 mole of resorcinol with 3 moles of formaldehyde in dilute sodium hydroxide solution, and adding this solution to a mixture of natural rubber and GR-S latices, The adheslve is aged for 18 to 24 hours at room temperature before use. Production experience has shown that the initial condensation between resorcinol and formaldehyde niust be controlled rigidly. An improper end point of the reaction often causes coagulation of the latex on mixing and is nearly always reflected in the properties of the finished rubber article. A reaction time of 6 hours a t 5 5 O F. has been found empirically to give good results, but this standard is difficult to maintain because of varying atmospheric and water temperatures In order to eliminate costly temperature control equipment, it was desirable to find a rapid and convenient method for ascertaining the status of the condensation a t any given time, regardless of temperature. HISTORICAL
The condensation of phenols with formaldehyde is one of the oldest polymerization reactions known to modern chemistry, and its kinetics have been studied by many investigators. In the earliest published attempts, Jablonower (6) determined reaction velocity by making density measurements a t regular intervals, and Drummond (1) made the method more sensitive by measur-
RESORCINOL-FORMALDEHYDE REACTION IN DILUTE SODIUM HYDROXIDE SOLUTION
None of the described methods is conveniently suitable to determine the end point in the preparation of RFL adhesive. Refractive index, measured during a simulated production run
