V O L U M E 19, NO. 6
406 Table V.
Effect of Time on Degree of Reduction of Human Hair
[Commercial cold waving solution (DCR-3) a t room temperature] Reduction Time Cystine Found Average Man. % % 17.0,16.0 0, control 16.5 4 8 14
3.0, 2.6, 2.4, 2.5,
20
Table VI.
2.9 2.6 2.6 2.4
3.0 2.6 2.5 2.5
Effect of Time on Degree of Reoxidation of Reduced Human Hair
Oxidation Time Min. 0, control 3 6
Cystine Found
Average
%
%
9 15
on reduced hair was studied. In Table VI the results of a typical experiment are tabulated. The hair was reduced a t room temperature for 15 minutes by the immersion technique using a commercial cold waving solution. Samples of the hair were then transferred t o a 3y0 aqueous solution of potassium bromate, corresponding to a commercial cold waving oxidizing solution. They were left in the solution for varying lengths of time as indicated in Table VI, and then “alkylated” by the iodoacetate method, and the amount of residual cystine was estimated by the Sullivan method.
If one takes into consideration that a 15-minute reduction period brings the residual cystine value down to 37, or less, this experiment shows that most reoxidation takes place in the first 3 minutes. Other oxidizing agents like hydrogen peroxide act much the same as bromate. ACKNOWLEDGMENT
rlus cysteine value of the original hair, and subtracting the value or the residual cystine of the modified hair yields the amount of cysteine formed in the reduction step. Application of this method of analysis to the process of cold waving has yielded some interesting results. I n Table V the effects of reducing a sample of hair by immersion for varying lengths of time in a commercial cold waving solution (0.72 N in thioglycolate and pH 9.2) are listed. The results obtained show that by the immersion technique the reduction is practically complete within 4 minutes and that further treatment has little effect upon the residual cystine. However, this holds true only with solutions such as were used in this experiment. Stronger and more alkaline solutions will disintegrate the hair completely. I n another series of experiments the effect of oxidizing agents
The authors wish to take this opportunity to thank Milton Harris for his interest in this work and for his many helpful suggestions. LITERATURE CITED
(1) Brdzcka, Czechoslov. Chem. Commun., 5,238 (1933). (2) Folin and Looney, J . Biol. Chem., 51,421 (1922). (3) Hessand Sullivan,Ibid., 151,635 (1943). (4) Miller and Du Vigneaud, Ibid., 118,101 (1937). (5) Mirsky and Anson, J. Gen. Physiol., 18,307 (1935). (6) Okuda, J . Biochem. Japan, 5,201 (1925). (7) Patterson, Geiger, Mirell, and Harris, J . Research National Bur. Standards, 27, 89 (1941); Research Paper RP 1405. (8) Sullivan, U.S. Pub. Health Service, Pub. Health Repts., 41,1030 (1926); 44,1421(1929). (9) Sullivan, Hess, and Howard, J . Biol. Chem., 145,621 (1942). PREEENTEDbefore the Division of Biological Chemistry a t the 110th CHEMICAL SOCIETY, Chicago, 111. Meeting of the AMERICAN
Measuring the Volumetric Expansion of Solid Materials W. W. PENDLETON AND H. M. PHILOFSKY, Westinghouse Research Laboratories, East Pittsburgh, Pa.
A method for measuring the coefficient of cubical expansion of solid materials is described. Similar in principle to the A.S.T.M. test for the volumetric expansion of bituminous materials, it presents modifications in both apparatus and procedure which allow use of laboratory equipment and shorten time of testing. Both expansion and contraction may be measured with this method.
T
HIS paper describes a simple method for measuring the volumetric expansion of materials which are in the solid state a t some temperature above the freezing point of mercury ( -38.9 ’ (2.). The method is applicable to measuring the coefficient of expansion of a substance over the temperature range -35” to 300” C. (-31” to 572“ F.). Thermoplastic materials, as well as crystalline solids, may be measured. In principle, the method proposed is the same as that described by Abraham ( I ) , which is the standardized A.S.T.M. test for measuring the coefficient of expansion of bituminous compounds (2). In the standard test, a special steel cylinder, closed a t one end and threaded a t the other, is used for the sample cell. A steel cap is screwed onto the cell and gasketed against a shoulder on the cylinder. Into the cover is welded a steel capillary with such a construction that, when the sample and mercury fill the cylinder and the cap is in place, enough excess mercury is extruded through the capillary to ensure void-free conditions within the cell. During the heating, which is carried out in a liquid bath on the inverted cell, the escaping mercury is collected and carefully weighed. Account is taken of the mercury expansion and steel expansion in the calculation of the expansion of the specimen.
The material may be previously degasified by vacuum and heat in a separate container or in the steel cell with a second cover fitted with a vacuum outlet. Since weighing is the most accurate method of determining volume change, this method should result in very good accuracy. The modifications to the standard test, which are described in this paper, permit the use of ordinary laboratory equipment, shorten the time of sample preparation, and make possible a study of contraction of a material as well as expansion. The method may be used t o measure both “true” and “apparent” volume expansion. As explained by Abraham ( I ) , engineers may wish to know the expansion of a material either as received or after special pretreatment. If the sample is not pretreated, and contains foreign inclusions such as water and dissolved gases, these inclusions will produce an apparent expansion greater than the true expansion. Thermoplastic materials can be degasified t o give true expansions but, with solid materials having no melting points below disintegration temperatures, only apparent expansions may be measured. Although this method has been used only for measuring the volumetric expansion of asphalts (which are thermoplastic), it is applicable to any solid material
407
JUNE ‘1947
r-
--
VI n! I
e
continued until bubbles cease to appear on the asphalt surface. After degasification, the material is allowed to solidify under vacuum, whereupon the flask is inverted and mercury in A is allowyed to fill slowly the void space in the flask. Any sample which does not adhere to the glass will float on the mercury to the top of the sample holder as shown in Figure 1. Two readings on A before and after filling the void, determine the volume of mercury in the flask. A correction for the external tubing from D to G should be made. With the stopcocks turned as shown in Figure 1, the system is ready for expansion measurements. MEASUREMENT PROCEDURE AKD CALCULATIONS
The temperature is raised in suitably small intervals and B is read after temperature has reached equilibrium. The time iilterval for each 10” C. rise should be about 4 hours. At any point on the expansion curve the temperature I II may be lowered to check a previous point or inI I vestigate the curve shape. In the work on asL------J phalts, the writers followed each expansion Figure 1. Volume ExDansion RIeasurement of Solids and curve by a contraction curve. A hysteresis loop TheEmoplastic Rlaterials was noted for some materials not previously degasified. It was found necessary to read B to three significant figures to keep the final error in coefficient of expansion not affected by contact with mercury. Measurements on solids to less than 1%. not melted by heat and univet by mercury may be subject to Calculations of coefficient of cubical expansion are made on minor errors due to void formations between particles vihich are the basis of the following derivation. adjacent to the glass container.
-
DESCRIPTION OF APPARATUS
Figure 1 is a schematic diagram of the modified method for measurement of the volumetric expansion of solid materials.
A large buret, A , for filling the system with mercury and a small buret, B , for measuring expansion are shown connected to a Florence flask, E , and the vacuum system attached to trap M . The capillary tubing, H , is connected outside oven I to the buret and vacuum systems by means of a small rubber tubing, G . The buret system and vacuum manometer, L, may be mounted on a
panel board as shown by the dotted lines. The capillary tubing should extend well beyond the oven wall, so that temperature gradients will affect only a small volume of mercury and the mercury in B will not be above room temperature. C and D are double-bored stopcocks for controlling the evacuation and mercury filling Within the circulating air oven I, a support, K ,is provided which is strong enough to hold the flask which contains a large mass of mercury in addition to the sample; fan J provides for even temperature distribution. The thermometer, F , should be so placed that its bulb is close to sample container E . Temperature errors should be checked carefully and minimized as much as possible, since expansion is directly proportional to the temperature change. Because of the large mercury volume which is intimately in contact with the sample, small oven-temperature fluctuations do not seriously affect the sample temperature. For accuracy of reading, B should be closely graduated. It has been found that, for the measurement of expansion of asphalts from 25“ to 200” C., a sample weight of 50 grams (about 50 ml.) requires a flask of 125 ml., a buret, A , of 250 ml., and buret B , of 25 ml. -4is graduated in 1-ml. divisions, while B is graduated in 0.1-ml. divisions. SAMPLE PREPARATION
The sample, in the form of small pieces, is weighed and placed in the flask whose neck has been previously shaped to allow rapid sealing to the capillary tubing, H . Prior to the connection of this tubing to the buret system, the stopcocks and tubulations down to D in Figure 1 must be filled with mercury as shown. The rest of the system remains free from mercury until after evacuation of E. All horizontal tubing above D should be avoided in order to facilitate filling of that part of the system with mercury. The sample flask and tube are now connected to the vacuum system by means of G . The flask is first placed in an upright position (the reverse of that shown in Figure 1) to aid the evacuation and degasification of the sample. For complete degasification of asphalt, a temperature of about 10’ C. above its ball and ring softening point for a 2-hour period is necewary. Degasification is
The rise of mercury in B due to expansion of the mercury and the sample is related to the actual volume change in the following manner: Aeo =
.AVO’f A v o
- (AVO’+ AVdp’Ato
(1)
where Aeo = difference of B readings with sample a t t and a t t o AVO’= change in mercury volume between t and to AVO = change in sample volume between t and t o 8’ = coefficient of cubical expansion of mercury Ato = temperature difference betneen t and to t = temperature within oven t o = temperature of mercury outside oven The negative term in Equation 1,which accounts for the change in volume of the mercury displaced from the oven as it cools from t to t,, has been found to be negligible and may be dropped from the equation. The change in mercury volume AVO’ over the temperature interval At0 is given by: Avo‘ =
VO’DO’AVO‘
( 2)
where VO’ = original mercury volume Do’ = mercury density a t t o Avo’ = change in specific volume of mercury between to and t Since the temperature dependence of the mercury is negligible the symbols Avo‘, Aeo, and AVOcan be redefined in tei-ms of any desired temperature interval, At. Substituting Equation 2 in Equation 1 and solving for AI’, the change in volume of the sample, we have:
AV
=
Ae
- T70’L)o’~v’
(3)
shere Ae = difference of B readin. s a t tl and t2 Au’ = change in specific volume of mercury between t l and t2 Since the coefficient of cubical expansion is defined as:
p = - AV AtV V = volume of sample a t 0’ C. and, since the room temperature volume, VO,of the sample may be expressed as:
Bo
=
1; (1
+ No)
(51
408
V O L U M E 19, NO. 6
the expression for the coefficient of cubical expansion using ,Equations 3,4,.and 5 can be shown to be: =
Ae - VO’Do’Av’ AtVo - to ( Ae - VO’DO’AV’)
Equation 6 may be simplified to: 1
p=tA
- to
where L.020 30-176 ,00053(‘C)-’
I40 0
40
80
I20
160
200
TEMPERATURE (‘C)
I20
Figure 3.
Volume Expansion per Milliliter of Asphalt A
i ao
.I20
2
E
8 1)
v)
z0 a
W
a 6.0 W
-
.080
I-
+
W
a
3
z
4.0
.060
W V
z : 2.c W
,040
E 0 TEMPERATURE ( O C)
Figure 2. Plotted Data from Measurement of Coefficient of Expansion of Asphalt A
a X w
0
The value of V Ois found from the room temperature density and mass of the sample. The original mercury volume, V O, is determined from the difference in A readings as the vacuum void is filled with mercury. The volume Ae is measured from the difference in the mercury levels in B b,etween tl and tP. The change in specific volume for mercury, Av , between these temperatures as well as the value of DO’can be obtained from handbook tables. Equation 6 neglects the correction due to the expansion of the glass container. This correction is, however, far smaller than the accuracy of the method and need not be considered here. For the case of a linear relationship between AV and temperature, Equation 6 may be used a t any convenient value of At, including Ato. Linearity may be tested by a plot of Ae versus temperature. Figure 2 shows a linear relationship for asphalt A. Figure 3 shows the change in volume per unit volume for the same asphalt. Here the slope of the line is approximately equal to the coefficient of cubical expansion of the material based on room temperature volume instead of volume a t zero C. The case of a nonlinear expansion curve may be treated as a succession of straight-line segments, in Yhich A t in Equation 6 spans the segment, or the tangent to the curve a t any point may be drawn and the slope placed in the following equation:
40
80 TEMPERATURE
I20 (“C)
160
200
Figure 1. Volume Expansion per Milliliter of Asphalt B
where
de = slope of tangent to Le versus t plot a t any t dt
p’ = coefficient of cubical expansion of mercury at temperature t, AV I n Equation 7b the -is replaced by /3’, since Af is assumed to Af be infinitely small. Figure 4 is an example of a nonlinear expansion curve from which two values of coefficient of expansion have been calculated. This curve represents the apparent expansion of an undegassed asphalt. Apparent expansions of some asphalts containing tightly held water vapor have shown large increases above certain high temperatures ranging from 150” to 200” C. The viscosity of the sample appears to have some effect on this critical temperature. This device is, therefore, a convenient method for studying the gas evolution properties of various materials. This method is also applicable to a study of contraction of materials below room temperature. In place of the oven in Figure 1, a suitable cold chamber may be used. Readings on B would be made starting at the top and going down. In this case,
409
JUNE 1947 the signs of Equations 6 and 7b must be reversed. I n some applications for asphalts and like materials, engineers are concerned with a so-called "pull-away" temperature, a t which the forces of contraction overcome the adhesion forces within the material. An interesting study of this action might be made on various materials with the method described for contraction. The device may also be used to study the change in expansion characteristics near the freezing points of materials showing sharp melting points and near the boiling points of liquids which may be solidified a t a temperature above -35" C. SUMMARY
The method described is similar in principle to the A.S.T.M. test for volumetric expansion of bituminous materials. Certain modifications in both apparatus and procedure result in the following advantages over the standard method : Results can be obtained in a shorter time because the system is easier to manipulate, and the sample preparation is simplified.
The method requires only ordinary laboratory glassware and equipment. Introduction of mercury under vacuum conditions ensures complete elimination of gaseous voids external to the sample. Contraction as well as expansion measurements can be made. The method avoids the health hazard of vaporized mercury. ACKNOWLEDGMENT
The writers wish to express their appreciation to C. F. Hill, manager, and to L. J. Berberich, section head, of the Insulation Department of Westinghouse Research Laboratories for their helpful suggestions in this work. LITERATURE CITED (1) Abraham, H., "Asphalts and Allied Substances," Vol. 11, 5th ed., pp. 1033-50, New York, D. Van Nostrand Co., 1945. (2) Am. SOC.Testing Materials, A.S.T.M. Standards, 1'01. 111, p. 1126, Specification D 176-421,1942.
Laboratory Low-Temperature Fractional Distillation Optimum Charging Rates C. E. STARR, JR., J . S. kNDERSON, AND V. JZ. D.IVIDSOU Esso Laboratories Standard Oil C o m p a n y of .Yew Jersey, Louisiana Division, Baton Rouge, La. The analysis of mixtures of light hydrocarbon gases and gases such as hydrogen, nitrogen, oxygen, and carbon monoxide is generally conducted first by segregation of fractions by low-temperature distillation and second by analysis of the individual fractions. In such procedure the lightest components are taken overhead from the distillation column during charging of the samples. Published procedures for operation of low-temperature distillation equipment do not ordinarily include sufficiently detailed instructions for charging samples of widely varying compositions to ensure obtaining overhead fractions of maximum purity with minimum charging time.
T
H E analysis of mixtures of light hydrocarbon gases and gases such as hydrogen, nitrogen, oxygen, and carbon monoxide is generally conducted first by segregation of fractions by low-temperature distillation and second by analysis of the individual fractions. In such type of analytical procedure thr lightest components are taken overhead from the distillation column during the charging of the Pample. Published procedures (3, 4,5 ) for operation of lowtemperature distillation equipment do not ordinarily include sufficiently detailed instructions for charging samples of widely varying compositions to ensuw ohtaining overhead fractions of maximum purity x i t h minimum charging time. With the aim of improving such analyses a study a-as conducted to establish a more uniform procedure for handling samples of complex gas mixtures when conducting low-temperature fractional distillation, a necessary step in the segregation of compounds for specific chemical and physical tests. When such a gas mixture is charged to low-temperature distillation equipment, usually cooled by liquid nitrogen (boiling point - 195.8" C.), an overhead fraction is first segregated to contain, if present, all the hydrogen (boiling point, -252.7' C.), nitrogen, carbon monoxide (boiling point, -192" C.), and oxygen (boiling point,
A study has been made of charging rates for samples ordinarily encountered in petroleum refining plants. Overhead samples taken at different charging rates have been analyzed by mass spectrometer to determine the amount of contamination w-ith higher boiling constituents. The results obtained indicate that charging rates much higher than ordinarily employed may be used with equal accuracy; use of the higher charging rates effects considerable economy of time. Supplementary data obtained also show- that for special applications distillation rates may be increased ' severalfold over rates now generally employed.
- 183CC.),andaportio~ioftheniethane (boilingpoint, - 161.4"C.), ethylene (boiling point, - 103.7' C.), and ethane (boiling point, -89.0' (2.). During the charging period and while the gases listed above are being removed, the balance of the hydrocarbons are condensed into the still pot, and upon completion of the charge, are fractionally distilled. For accurate analysis of the overhead fraction it is desirable that i t be segregated in such a manner that hydrocarbons of three or more carbon atoms are excluded. Likewise for accurate determination of hydrocarbons of three or more carbon atoms in lean gases, it is important that the overhead fraction should not contain even small amounts of the heavier hydrocarbons, since such small amounts may represent appreciable portions of the total present. The overhead fraction is generally analyzed by chemical and combustion analyses on such equipment as the U. S. Steel (6) or Burrell ( 2 ) apparatus. I n this type of analysis ethylene is removed by chemical absorption and methane and ethane are determined by combustion. When methane is burned with oxygen the sample undergoes a contraction in volume equal to twice the volume of methane; the carbon dioxide produced is equal in volume to the methane. The water formed condenses and is of negligible volume. CH, 202 +COn 2H20 (1)
+
+
