March, 1946
INDUSTRIAL AND ENGINEERING CHEMISTRY
possess three free hydroxyls, each of which can bind one molecule of water. I n this calculation such factors as condensation in capillaries are assumed to be little changed by removal of the amorphous component. A possible cause of the high figure for viscose rayon may be an error in the amorphous estimate already discussed. The fact that noncrystalline or accessible cellulose appears to be divisible into two roughly equal parts on the basis of hygroscopicity and hydrolytic susceptibility may be important, especially where consideration is being given to possible fiber structure and properties relations. Clearly, in hygroscopic and hydrolytic behavior the mesomorphous component bears a closer resemblance to the crystalline than to the amorphous component. For this reason, in the determination of properties-structure relations it may prove advantageous to consider mesomorphous
301
and crystalline components together as a single component. This would tend to emphasize the importance of the amorphous component, but such emphasis appears to be justified by the present evidence. I n any case the amorphous component should be distinguished from the remaining accessible cellulose to which i t apparently is not closely akin in chemical behavior. LITERATURE CITED
(1) Badgley, W.,Frilette, V. J., and Mark, H., IND.ENO.CBEM.,37, 226 (1945).
(2) Conrad, C . C . , and Scroggie, A; G., Ibid., 37, 592 (1945). (3) Nickeraon, R.F.,Zbid., 34, 1480 (1942). (4) Ibid., 37, 1115 (1945)~. CONTRIBUTION from the Cotton Research Foundation Fellowship a t Menon Institute.
Volumetric Behavior of 1-Butene R. H. ObDS, B. H. SAGE, AND W. N. LACEY Calgornia Institute of Technology, Pasadena, Calif.
T h e volumetric and phase behavior of 1-butene was experimentally determined at temperatures from 100' to 340" F. and at pressures up to 10,000 pounds per square inch. The results of these measurements were found to correlate well with those of other investigators.
T
HE industrial importance of 1-butene has recently increased because of its role in the production of synthetic fuels and rubberlike polymers. The development of new products and the improvement of existing processes involving this substance will depend t o some extent upon the availability of accurate data relating t o its physical properties. Published literature concerning the volumetric and phase, behavior of 1-butene is not abundant. Coffin and Maass (5) reported values of the vapor pressure of 1-butene throughout the interval from -108" to 25" F., and determined the volume of the bubble-point liquid for temperatures from -51" to 50" F. Lamb and Roper (6) studied the vapor pressure of 1-butene between -70" and 32" F., and Roper (6) reported several measurements of the volumetric behavior of the gas for temperatures from -22" to 140" F. a t pressures less than 16.5 pounds per square inch absolute. The present work extends the range of the experimentally observed volumetric behavior *of 1-butene to 10,000 pounds per square inch and 340" F. The 1-butene was prepared by dehydration of n-butanol with activated alumina as a catalyst. Distilled technical n-butanol was vaporized and superheated t o about 660" F. at a pressure of approximately 15 pounds per square inch absolute. The gas was then passed through a catalyst chamber at 680" F., consisting of 24 inches of Pyrex tubing 1.4 inches in inside diameter, packed with 4-8 mesh granular alumina. Flow was controlled by a throttle in the feed line carrying liquid n-butanol, and a flow rate of about 35 drops per minute gave satisfactory resuIts. Greater rates resulted in the presence of excessive amounts of material believed t o be unreacted n-butanol in the condensate from the gas leaving the catalyst chamber. Lower flow rates gave a yellow, oily, nonvolatile liquid in the reaction products, which was assumed to be polymerized butene. The capacity of the catalyst a t 680" F. was roughly 2 pounds of n-butanol per hour per cubic foot of catalyst. A water-jacketed condenser
removed most of the water and reaction products of low volatility from the gas leaving the catalyst chamber. The gas was further dried by passage through granular anhydrous calcium chloride. The crude 1-butene vapor was passed directly from the dryer into the kett,le of a 4 f O O t , vacuum-jacketed fractionation column packed with helical glass rings. The material was distilled at atmospheric pressure with overhead withdrawal a t the rate of approximately 2% of the reflux in the column as long as the temperature of condensation at the head of the column stayed within 0.5" F. of the normal boiling point of 1-butene. The overhead product from the fractionation column was condensed at liquid air temperatures with continuous removal of noncondensable gases by means of a mercury diffusion pump backed by a Hyvac mechanical pump. This material was then subjected to two more fractionations in the same column, the initial and final tenth portions of the charge in the kettle being discarded .each time. A variation of less than 0.3 pound per square inch was observed in the vapor pressure of the final product of this process when a sample of it was isothermally vaporized from bubble point to dew point at 100" F. At the temperature of liquid air boiling at atmospheric pressure, the behavior of 1butene previously noted by Coffin and Maass (3) was observed by the authors, This substance apparently exhibits no definite freezing point, becomes increasingly viscous, and approaches B gel-like consistency as its temperature is lowered. An earlier publication (7) gave a detailed description of the equipment and procedures used in determining the volumetric properties of materials of this kind. No changes were found to be necessary for the present work. Three separately prepared and purified samples of 1-bubene were used. The data from the three samples were graphically correlated, and the results reported here were read from CUX-V~S drawn through the experimental points and smoothed in respect both to temperature and pressure. Figure 1 shows a series Of isotherms relating compressibility factor and pressure, UP $0 1000 pounds per square inch absolute. The experimentally observed states of the system are represented by circles. Alwgescale plot of this type was used t o establish the behavior of the system as a gas from infinite attenuation to dew point. The volumetric behavior of liquid 1-butene from bubble point to 10,000 pounds per square inch was obtained from isothermal
INDUSTRIAL AND ENGINEERING CHEMISTRY
302 I .c
Since the ratio P / P ' deviates from unity more than can be accounted for by the estimated iincertainty in this vc-ork, the quantity P' calciiltited from Equation 1 does not adequately reprcscnt' the vapor pressure of 1-butene over the riingr of temperature studied. However, the equation, together with t,he curve dranm through the experimental points in Figure 3, may be uscfiil for interpolation within the temperature intcrvnl from -50" t o +300° F. The curve is dotted outside the region studied by the authors and rcpresents their estimate of probable behavior. The molal volume of 1-butene at the bubble point is shown as a function of temperature in Figure 4. I n so far as sets of data in different temperature ranges can be compared, the authors' results in Figures 2, 3, and 4 seem to be in good agreement, with those of others. Wherever comparisons were possible, similar agreement was found with the volumetric behavior in the low-pressur(8 gaseous region reported by Roper (6).
0.E
LL
e
2
Vol. 38, No. 3
0.f
>
t
m
3 0.4
8
0.2
ESTIMATED ACCURACY
' The design of the volumetric apparatus was such as to permit changes smaller than 0.1% in the absolute magnitude of any of the pertinent vari200 400 600 800 ables involved to be detceted with ease. However, PRESSURE L B h Q IN reliable estimates of the accuracy of the data can Figure 1. Exuerimental Results for States in and near the be obtained only from a study of the consistency Two-Phase Region for 1-Butene and reproducibility of the measurements. Several years' experience in the operation of the volumetric apparatus has led the authors to consider the following curves relating volume and pressure. The critical state of the system was determined graphically by the method of rectilinear values as the uncertainties involved in this work: diameters ( 1 , 2). An attempt was made to detect the critical state visually by enclosing a sample of 1-butene in a thick0.17' or 0 . 5 Io./sq. in. Pressure Temperature 0.058 F. walled glass capillary tube whose effective volume could be 0 . 2 % or 0.01 cu. ft./Ib. mole Volume varied by the injection or withdrawal of mercury. However, before sufficient data were obtained, the capillary tube ruptured Where alternative values of the estimated uncertainties are listed and destroyed enough of the apparatus to make it inadvisable to the larger one for the conditions considered should be taken. rebuild. The volumetric data at 340" F. should be regarded with less VOLUMETRIC AND PHASE DATA confidence because the 1-butene showed a marked tendency to The smoothed volumetric data interpolated to even values of the pressure up to 10,000 pounds per square inch are recorded both as molal volumes and compressibility factors in Table I a t 800 60" intervals from 100' to 340" F. Table I1 presents supple600 mentary values of the compressibility factor for parts of the 400 gaseous region. Data for the boundaries of the two-phase region are recorded in Table 111. The authors obtained a value of 200 297" F. for the critical temperature of 1-butene; Coffin and Maass ( 3 )reported it to be 291 ' F. 100 z 80 Figures 2, 3, and 4 compare the authors' results with those of 60 other investigators. In Figure 2 the vapor pressure scale is 40 logarithmic, and the temperature scale is based upon the reciprow cal absolute temperature. I n accord with the approximate 3 20 integrated form of the Clapeyron equation (Q), this choice of 8 scales yields a nearly linear relation. a 10 T o show more clearly the magnitude of the experimental 8 inconsistencies, Figure 3 presents the experimentally observed e vapor pressure, P , in comparison with a reference pressure, 4 P', computed from the following equation which was arbitrarily chosen as approximately descriptive of the relation between 2 vapor pressure and temperature:
-
$
loglo P'
6.18466
2285 22 -- 0.00054633 T T
where P' = pounds per square inc: absolute T = absolute temperature, Rankine
I
I
(V
300
200
I
100
50
I
0
-50
TEMPERATURE 'F
Figure 2.
Comparison of Vapor-Pressures with Those Found by Other Investigators
INDUSTRIAL AND ENGINEERING CHEMISTRY
March, 1946
TABLE I. COMPRESSIBILITY FACTOR AND MOLALVOLUMEOB BUTENE
~:f/BzTf& 100" F. (62.5)a
Abs. z 0,890 Dew oint Bubbre point 0.0165
220' F. (282.6)
l6Oo F. (142.9)
z
V
V
0.812 0.C373
85.5 1.583 398.9 1.577 1,568 1.560 1.552 1.545 1.537 1.529 1.522 1.515 1.509 1.503 1.497
37.8 1.735 444.8 1.731 1.716 1.702 1.689 1.676 1.662 1.649 1.637 1.625 1.614 1.604 1.594
z
0.700 0.0755
V
18.07 1.949 490.2 29.62 1.930 1.901 1.875 1.852 1.825 1.801 1.779 1.759 1.741 1.724 1.709
280"
z
0.475 0.1608
Figurea in parentheses are va or pressures in pounds per square inoh absolute. P V / R T . V = molal volume. ou%ic feet . per . pound mole. a
F. (506) V
340' F.
z .... . . ..
7.45 2.523 535.3 34.3 13.40 2.345 2.195 2.123 2.061 2.013 1.972 1.937 1.906 1.878 1.853
Z
-
...... .V. . . 580.0 38.7 16.69 8.84 4.22 2.806
;2.271 : 82
2.193 2.129 2.078 2.F37 2.002 0.700 1.944 0,793 1.896 0.884 1.856 0.973 1.822 1.061. 1.768 1.236 1.408 1.726 1.692 1.577 1.661 1.742 1.633 1.903 compressibility factor
.
-
TABLE 11. SUPPLEMENTARY DATAON COMPRESSIBILITY FACTOR FOR BUTENE Pressure, Temp., Lb./Sq. In. O F. Abs. 100 25 50 , 160 50 100 220 100 250 280
300 loo 450 500
'
Temp.,
z
a
0.959 0.914 0.941 0.876 0.913 0.748 0.936 0.781 0.603 0.496
F.
290 340
4
Pressure, Lb. Sq. In. Abs. 450 500 540 554.6 700 900 1100
Z 0.634 0.557 0.463 0.399. 0.511 0.335 0.337
Dew point.
TABLE111. PROPERTIES OF COEXISTING LIQUIDAND GAS
PHASESOF BUTENE
Pressure Lb./Sq. Ih. Abs.
Temp., a F.
Bubble Point
Z
Dew Point
V
Z
V 54.1 36.0 26.59 20.82 16.82 13 84 11.46 9.45 7.67 5.95 3.85
a
303
*
,
deteriorate a t this temperature. Upon lowering the temperature, after having maintained the system at 340" F. for several hours, the bubblepoint pressures a t the lower temperatures were no longer the same and were not so sharply defined. Moreover, samples withdrawn from the apparatus after having been subjected to temperatures as high as 340' F. contained traces of a yellow, oily, nonvolatile liquid which resembled the material previously mentioned in connection with the preparation of 1-butene. I n the calculations the following numerical values were used: molecular weight of 1-butene, 56.105; universal gas constant, 10.732 (lb./scr. in.). (cu. ft./ . lb. mole)/' R.; 0' F. 2 459.69" R. /
ACKNOWLEDGMENT
Acknowledgment is gratefully made for encouragement and financial assistance from the Polymerization Process Corporation, interest and helpful suggestions from Manson Benedict of The M. W. Kellogg Company, and advice and assistance from H. H. Reamer in regard to the experimental work. LITERATURE CITED
(1) Cailletet and Mathias, Compt. rend., 102,1202 (1886). (2) Ibid., 104, 1563 (1887). (3) Coffin and Maass, J. Am. Chem. SOC.,50,1427 (1928). (4) Fermi, "Thermodynamics", p. 67, New York, Prentice-Hall, 1937. ( 5 ) Lamb and Roper, J. Am. Chem. Soc., 62,806 (1940). (6) Roper, J . Phys. Chem., 44,835 (1940). (7) Sage and Lacey, Trans. Am. Inst. Mining M e t . Engrs., 136, 136 (1940).
Czitical. 25
0 COFFIN
AND MAASS
2.0
1.5
TEMPERATURE
OF.
Figure 3.' Reference Plot of Observed Vapor Pressure as Compared to Values Given by Pbitrarily Chosen' Equation
0
100 TEMPERATURE
200
T
Figure 4. Comparison of Bubble-Point Volumes with Values Determined by Other Investigators