J. Chem. Eng. Data 1986, 3 1 , 396-406
396
Gas Solubilities, Vapor-Liquid Equilibria, and Partial Molal Volumes in Some Hydrogen-Hydrocarbon Systems and George A. Kandalic
John F. Connolly
Research and Development Department, Amoco Oil Company, Naperville, Illinois 60566
Gas soluMUtles were measured for hydrogen In n-pentane, 2,3-dlmethylbutane, cyclohexane, n-decane, m -xylene, 1,4-dlethyIbenrene, and l-methylnaphthalene. Vapor-llquid equlllbrium ratios (K’s) were measured for the second, third, flfth, and sixth of these systems. Partial molal volumes of hydrogen dissolved in the llquld hydrocarbons were measured for all systems except methylnaphthalene. Temperatures and pressures fell In the ranses 35-320 O C and 200-2400 PSI.
Vapor-Liquid Equilibrium Ratlos Vapor-liquid equilibrium ratios, or K’s, are defined by
I ntroductlon Because hydrogen is used in petroleum processing, physical properties of hydrogen-hydrocarbon systems are often necessary for the design of refinery units. This paper adds to the previous data ( 7 , 2)which we have reported on such systems. Gas solubilis for seven systems, vapor-liquid equilibrium ratios for four systems, and partial molal volumes for six systems are reported here. Experimental Section The hydrocarbons used were American Petroleum Institute standard samples with stated purities of 99.9-kmol %. The experimental method has been described in detail ( 7 ) . Briefly, a small sample of known mass and composition was confined above mercury in a glass capillary. Then, as the liquid mixture was expanded at constant temperature, the pressures at which the first trace of gas appeared (bubblepoint pressure) and the last trace of liquid disappeared (dew-point pressure) were measured. The bubblepoint pressures yield gas solubilities, and in a two-component system the combination of bubble- and dew-point pressures yields the vapor-liquid equilibrium ratios ( K ’ s )for the system. Partial molal volumes were derived by measuring the expansion which occurs when hydrogen is mixed with liquid hydrocarbon in a glass capillary (3). The temperature can be set from 35 to 370 O C , but it was usually kept below 300 OC to avoid thermal decomposition. Pressure can be set from 100 to 3000 psi, but it was usually kept below 2500 psi to avoid frequent capillary replacement. Gas Solubilltles Because gas solubilities are fairly linear with “partial pressure”, the bubble points can be represented by equations of the form
S=a
of bubble pressure. As a definition of partial pressure, p x , p l o works somewhat better than p - pro,but not enough better to warrant the added complexity in solving eq 1 for x , (given p ) . p r ois an approximate vapor pressure which is calculated from Table 11. Of course, this particular p ,O must always be used in conjunction with the constants in Table I. The experimental bubble pressures, along with the deviations from eq 1, are listed in Table I I I.
+ b(p - pro)+ c(p -
(1)
where a , b , and c are constants at each temperature and are given in Table I; S is a “solubility” defined by S = ( x 2 / x 1 ) / @ - p ,O); x is a mole fraction in the liquid; p is a bubble pressure, p r o is a vapor pressure, and subscripts 1 and 2 denote hydrocarbon and hydrogen, respectively. These definitions of “solubility” and “partial pressure” are, of course, only arbitrary devices to simplify the representation Presently at Amoco Corporation Corporate Research Department. P. 0. Box 400, Naperville, I L 60566.
0021-9568/86/1731-0396$01.50/0
where y and x are mole fractions in the vapor and liquid, and subscripts 1 and 2 refer to hydrocarbon and hydrogen. For a binary system, K is a function of temperature and pressure only. At each experimental dew point (listed under psi in Table IV) the vapor composition was known, and the corresponding liquid composition was calculated from eq 1. These compositions were then used to calculate the vapor-liquid equilibrium ratios listed in Table I V . The function pK, where p is the pressure of the system, varies more slowly with pressure than does K. Therefore, interpolation to convenient values of pressure was accomplished by plotting pK as a function or pressure at constant temperature on large-scale graph paper. Reduced versions of these plots are shown in Figures 1-8. Smoothed K’s, listed in Table V, were read from the pK plots. Those values in parentheses are dependent on short extensions of eq 1 (linear form only) to pressures higher than observed bubble pressures. Values of K for two other hydrogen-hydrocarbon systems have been reported (2). Partial Molal Volumes Partial molal volumes of gases in liquid solutions are necessary in the calculation of thermodynamic properties of solutions from liquid-vapor-equilibrium and gas-compressibility data at high pressures ( 7 ) . Because the literature usually reports only room-temperature partial molal volumes for gas-hydrocarbon systems, and because our apparatus easily combines these measurements with bubble points, we have made numerous high-temperature partial-molal-volume measurements. The volume-expansion method used has been described (3). At gas concentration below 10 or 15 mol YO,the volume expansion of liquid per mol of gas dissolved varies linearly with composition at constant temperature and pressure; i.e. A V / n 2 = P2*
+ c’x,
(3)
where A V is the measured volume expansion of the liquid when n 2 moles of gas are dissolved in pure hydrocarbon to give a mole fraction x p , and P2* is the gas partial molal volume at infinite dilution. P,’ and c’are functions of temperature and pressure only. That the intercept is P2* follows from the definition
P,’ = lim ( A V / n 2 ) , , n,-0
0 1986 American Chemical Society
000 r
I
I
1
I
i 600
-
zn g
400-
YF P
300 -
200
-
100 0
0
0
500
1000
1500
I
I
I
600
1000
1600
2000
2000
PRESSURE IN PSI Flgure 1. pK, for 2,3dimethylbutane. I
0
600
1000
2000
1600
PRESSURE IN PSI
Figure 4. pK1 for 1,rldiethylbenzene.
i.e., assume that the molal volume of the solution, V, can be representedas a Taylor serles in the mole fraction of hydrogen, x,, for small x,. Then V = Vi
+ b’x, + c’x: + ...
(4)
where V, is the molal volume of the pure hydrocarbon, b’and c’are functions of temperature and pressure, and we neglect cubic and higher terms. By definition we have
v/= (dnV/dn,), 0
600
1000
1600
2000
2600
PRESSURE IN PSI
Flgure 2. pK, for cyclohexane.
Values of c’ obtalned from eq 3 are approximately related to the composition variation of partial molal volumes Pi and P,;
,+ ,
(5)
where n = n n is the total number of moles in solution and V, the solution molal volume, is given by V = x , ~ , x 2 0 , . Applying eq 5 to eq 4 we obtain
+
PI - V, = -c’x,2
(6)
P, - 8,’ = 2c’x,
(7)
398 Journal of Chemical and Engineering Data, Vol. 3 1, No. 4, 1986 N 0
N
N 4
s
*I
I?
CI
0
m
3 N
m
!9
7 p h
10 0
in
9 VI
9
OI U
0
m
10
n a m
4 4 4
2
3 h
I?
m 0
8 9 I?
10
'9
0
I?
n
Ln
U
in
2
N
N 3
el
0
U
m
p 01 N
'9
m
0"
::
m m
h
Ln
. I
03
m
I 3
U
U
in
Ln
h
s
n
N
10
4
3
U (0
N
3
3
0 N
0
9 9
9
g, 9
U N
m 10 0
10 L
in 3
Ln n
3
n
s in
N * Ln
Ln
0 3
m n
0 N
i N
0 0
3 N
x
Q 0 4
in Ln
9
x
m
4-
2 3
10
0,
N
10
9
N
U
**
9
u.
3 h
n
9
0,
g 9 9 5 Q (9
m
N
3
3
?
0
9
E!
3
0
m
9
x
i 3 in
'9
'9
3 h
N 3
3
3
3
in
10
10
'0
10
R
x
'S
0 3
r.
n
3 i
,2 01
x
a
3
3 9 4
-
!
0
N
10 h
10
v
-
3
h
3 N
4 co U 3
N n
U
s
3
n
0 N .
U
U
Ln
L"
Ln
Ln
in
Ln
2 *
3
rl
co
0
L"
Q
h
2
3
0,
10
?
I? N
s
0 0
s
3
3
'f:
0
VI
I?
N.
N 3
s
0
U
U
u?
4 ?
in
9
9
N.
a
9
m
10
? ?
U
01
U
0
U
m
m h
9 5
3
N N
s
R
OI 3
10
in
L n
yl
Ln
m
h
3 0
n
9
3
9
3 9
L n
9
9
h
z ON
Q
3
2
VI
U
9
3
0
N
in h
4
h N
x
Ln
*.
i
9
9
I?
m
2
0 N .
3
Q5
4
3 3
N 3
e
c
N
?
* U 9 0
3
'9 "-
3
Ln
-
i l
3
m
h Ln
9
9
9
h
U 0
0
Ln
Ln
$
U h
- -
?
3
9 '9
3 Ln
10
k N .
y-i
2
x
2 v,
x
? Ln
m
'9
! Ln yl
s 3
h
yl
s 9 Y h
h
N
3
%
m
3
?
3
3
m
9 in
3 N
10
0:
0:
h
2
3
a 4
z9 23
h
10 N 3
h in
n
10
?
3 N
3 N
3 n
01
z
R
n h r. in
3
3
3
h
n N
x
N IC(
3
3 h
3 Ln
3
n
3 N
3
3 3
3
s
P
1
Ln
10
10
10
10
10
10
10
10
10
10
10
4
m
3
3 N
3 h
3 u?
2
2
i Ln
4
Ln N
m N
0,
10
N
9
3 9
2
g
:2
m
N
in
?
9
U n
U N
m in
E!
s
s
e
rl
3
m
10
(9 N
N
d
9
m
m
2
h
N
s
9
Y
-
m
0,
h
'9 3
3
?
5
3 n
U
Ln
Ln
in
in
in
in
Ln
in
Ln
s
P
5
m 3
N N
3
U
U
r.
m
m
s
10
h U
N 10 h N
8 3
Ln
rl n N
3
10 3
4 2
3 n
m
m
3
h
U
10
N n
U N
3 9 9 N
0
E!
0, U
3 n
N 3 h
N
i N
3 i
3
-
h h 3
z
2
10
h N
E
3
3
3
3
3
3
10 3
3 VI
in 3
3 N
3 N
3
m
U
in
in
Ln
Ln
in
Ln
Ln
2
3 h
2
8
3
m
h c
9
n
?
N
Z 9
H9
n h in N
3
yl
9 m b
E
d
n VI
s
N
m h
9 4
In n
4 0
w
3
Ln 3
rl
4
0:
3
4: ?
m
N
0,
x
N
N
9
9 m
Ln 3
h 3
N
N
01 3 N
01
'f:
9
4 m
10 10
m 0. in
N
N
0
s
h
7
3
i
?
2
5
on
s
E 2 5
h N
9
N
3
w
w
2
w
h
I?
-?
U
2 3 0 yl
9 N
10
10 0
9 9
h
4
3
h
in
9
Ln h
in n
9
m
J
0:
ON
0
::
0
A x 0
0 0
.-m
Journal of Chemical and Engineering Data, Vol. 3 1, No. 4, 1986 399
Table 11. VaDor Pressure Constants' ( 4 ) A B 6.852 21 1064.630 n-pentane 6.80983 1127.187 2,3-dimethylbutane 6.844 98 1203.526 cyclohexane 6.953 67 1501.268 n-decane 7.00908 1462.266 m -x y 1ene 7.000 54 1589.273 l,4-diethylbenzene 1-methylnaphthalene 7.068 99 1852.674
C T,, K 232.000 470 228.900 500 222.863 554 194.480 619 215.105 619 202.019 662 192.716 787
P
'log (760 p:) = A - B / ( T + C) + 0.50(T/Tc - 0.75)*. Omit last term when T/T, < 0.75. p: is vapor pressure in (lb/in.2)/14.696. 14000
12000
10000
n
r
N
5
u t
.
8000
1
2k.
eo00
1
4000
I
-0 32 0 so that P, decreases slowly as H, is added and P, increases fairly rapidly. The liquid-phase activii coefficients show this same asymmetry in behavior ( 7). Experimental values for the solution dilation A V l n , , are available as supplementary material. Because there are sev-
era1 thousand of these values, and because we wish to derive partial molal volumes, an analytical representation is desirable. Previous work ( 7 ) has shown that the reciprocals of the dilute solution partiil molal volumes, P,", are linear with pressure up to about 1600 psi. A small quadratic term is sufficient to take care of any curvature between 1600 and 2500 psi. The
PRESSURE IN PSI
400 Journal of Chemical and Engineering Data, Vol. 3 1, No. 4, 1986
Journal of Chemical and Engineering Data, Vol. 31, No. 4, 1986 401
.**
N V
4 .?O 0
C
.
.
-
9 . %
N
o o o o - o - N - ~ - c ~ 1
.
1
1
.
.
.
1
,
0 c 0 0 0 0 0 0 0 0 0 0 c
7
000000000000
39??93??31?9
C O O O O O O O C O C O OO---NNN**m.? .? ? . .? .? Lh m m 994 4 fm P 0 0%0 h N N N N N N N N h N N N N N N h N N h N N N N N N N N 'DID N N N N N NN
0
V 0
c: MH
0
0
0 m?
o N c.
I
.
l
l
,
.
.
o o o o c o I
3
---
V N
I
402 Journal of Chemical and Engineering Data, Vol. 3 1, No. 4, 1986 400
I
I
I
I
Y
Y
iE3
\
300
u
z
! 3 9
I
200
I
3
!
0
5
aF m
100
d
If-
120'-*-*-*
*-e-
a 0
I
I
I
I
500
1000
1500
ZOO0
0
PRESSURE IN PSI Figure 10. Partial molal volumes of hydrogen, at infinite dilution, in liquid
2,3dimethylbutane. 400
0
2500
-*-*i
35*-*-*-*
0-p.-
I 500
I 1000
I 1500
I 2000
2500
PRESSURE IN PSI Flgure 12. Partialmolal volumes of hydrogen, at infinite dilution, in liquid
cyclohexane. I
I
I
I 400
e
Y
i; I
\
0 V
z
140'-e-
e-*-'
i,
-e-
35"-*--.--.
---------36'--------0
I
I
I
I
psi). Then a second ft was made by using the complete eq 8 with a, b , and d fixed at the values derived in the first fi, and with no restriction on pressure. a , b , and d , as derived in the first fit, and c , as derived in the second fit, are listed in Table VI. Deviations listed in Table V I and in the supplementary material refer to the second fit. Partial molal volumes of hydrogen at infinite dilution in hydrocarbons, were calculated from constants of Table V I and plotted in Figures 9-14. The nearly vertical dashed lines represent vapor pressures of the pure hydrocarbons. Some lower pressure points are shown as open circles to indicate that they are the result of a double linear extrapolation, Le., into regions of both composition and pressure which are below
v2*,
Journal of Chemical and Engineering Data, Vol. 31, No. 4, 1986 403
4
.?4 x
404 Journal of Chemical and Engineering Data, Vol. 3 1, No. 4, 1986
.d
*
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 ~ m u m W w O N e W w O N u 3 3 3 3 3 N N N
U
10 0
2
.ri VI
P
V U
8 N
x
N
e
0 0 0 0 0 0 0 0 0 0 m 0 0 u r0p0i -0w0z0z 02 2020
m u
0
m u
EiN
??
O NIO N W i n
* .
N
U %
(3
I
N
u x m 0 3
N
U Y
0
3
d
U
s N
u x
...........
..........
. . . . . .
v v v v v v
W h u N w m O m O o * 3
M
...........
m h O I u O W J N , 4 O m inm"N333330
n
. . . . . . .G- G G 1
h 0 W 0 Y I W 0 0
3
x
........
m N 3 3 3 3 3 O
v
N
Y
x
0umlp.mrlom
3
? ?? O N W "3
m o m
".?
0 0
m
3
?r:
Journal of Chemical and Engineering Data, Vol. 31, No. 4, 1986 405
J . N -2 d
W
OI
\o
d
0
d
h
N
w
In
N
N
N
N
N
N
o'c9 '9 '9
o\ N 3
m d d d
u
m
u
9 9 9
u
-
"
N rD
n
h
u
)
o : ? i
N
d
-. *.
"
N. i iD N
9
d
LnwyI\Dm
9 9 9 9 9 N
m
2 d
'9
m " m
-. '9
? ? ? d
3
N
m u
9 9
J. Chem. Eng. Data 1986, 3 1 , 406-408
406 200
-
1
I
I
molal volumes of hydrogen are often larger than those of hydrocarbons.
I
3 300'C-0
-
2 8 0 ' - . ~ .
-7
;g
AZ
d
0-0-
P
4
0
230'-0-.~ 180~-0-.-,
OUo-
d
i
I
-.z=g
-0
i
1
Reglstry No. Hydrogen, 1333-74-0; n-pentane, 109-66-0; 2,3dimethylbutane, 79-29-8; cyclohexane, 110-82-7; ndecane, 12418-5; m-xylene, 108-38-3; 1,4diethylbenzene, 105-05-5: l-methylnaphthylene, 90-12-0.
Literature Cited (1) COnnOlly, J. F. J. Chem. Phys. 1962, 36, 2897. (2) Connolly. J. F. Roc. Am. Pet. Inst. 1965, 45, 111, 62. (3) Connolly, J. F.; Kandalic, G. A. Chem. €ng. Prog. Symp. Ser. 1963, 59, 8. (4) Rossini, F. D.; Pltzer, K. S.;Arnett, R. L.; Braum, R. M.; Pimntel, G. C.
Selected Values of Physical and Thermodynamic Properties of Hydrocarbons and Related Compounds: Carnegle: Pittsburgh, 1953.
Received for review May 8, 1985. Accepted April 28, 1986.
Supplementary Materlel AvaWabk Experimental values for solution dilation (13 pages). Ordering information is given on any current masthead page.
Solidification Behavior of the Cinnamic Acid-p -Nitrophenol Eutectic System N. Bahadur Slngh" and P. Kumar Chemistry Department, Gorakhpur University, Gorakhpur 273 009,India
The soiid-iiqukl equilibrium of the cinnamic acid-p -nitrophenol (CA-pNP) eutectic system has been investlgated. Heat of fusion, microstructure, and strength measurements have been made. Infrared spectral studies indicate molecular association in the formatlon of the eutectic. Thermodynamic functions such as h E , g', and SEhave been calculated and were found to be negative except g E . Statistical mechanical treatment shows that the surface nucteation theory holds good In the solidification of the present eutectic. Introduction Solid materials are of considerable interest both from the fundamental and from the technological point of view. Eutectic materials also come in this category. I n order to control the properties of such materials, studies of phase diagram, linear velocity of crystallization, microstructure, compressive strength, and thermodynamic properties are essential (7). The present paper describes the chemistry of the cinnamic acid-p-nitrophenol (CA-pNP) eutectic system with reference to the above properties.
Experlmental Sectlon
.
Materials and Purnlcatlon p-Nitrophenol (BDH) and cinnamic acid (BDH) were purified by repeated distillation under reduced pressure. The purity was checked by determining the melting points with the help of a mercury thermometer correct to f0.1 'C. The melting points are 112.0and 133.0OC, respectively. Cinnamic acid is represented as CA and p-nitrophenol is represented as pNP. Phase Diagram and Undercooling Study. The phase diagram of the CA-pNP system has been studied by the thaw melt method (2). Mixtures of various compositions were prepared 0021-956818611731-0406$01.50/0
in glass test tubes by repeated heating, chilling, and grinding in a glass mortar. The melting points and thaw points were determined with a mercury thermometer correct to fO.l OC. The undercooling study was made in a manner described by Rastcgi and Bassi (3). Study ot Mkrostructures A microscopic method was used for the study of microstructures of components and eutectic. A glass slide was kept in an oven at a temperature higher than the melting points of the eutectic, and a very small amount of the sample was placed on A. As the sample melted completely, the coverslip was glided on it. The slide was allowed to cool and nucleations were started from one side and this was then photographed with a camera under a microscope of desired magnification. The effect of 0.1% 8-hydroxyquinoline and 4chloroaniline on the microstructureof the eutectic has also been investigated. Heat of Fusion Measurements. The heats of fusion of the pure components and the eutectic were determined by differential thermal analyzer (Paulii-Paulik-Erdey MOM derivatograph, Hungary) using the method of Vold (4). The heating rate was maintained at 2 OC/min, and the temperature was measured by a Pt-Rh thermocouple. From this method only the relative value of the heat of fusion could be determined. Compressive Strength Measurements. Compressive strengths of components and eutectic in the form of pellets were determined by an universal testing machine. The pellets were prepared by solidifying the molten materials in glass test tubes of uniform diameter and the surfaces were smoothened by rubbing on emery paper. True stress was obtained by dividing the maximum load applied on the pellet (which was capable of breaking the pellet) by the area of the pellet. For calculating true strain, the total number of divisions on the graph (which indicates the maximum compression) was divided by the magnification (magnification was obtained by dividing chart speed by cross head speed) and the value thus obtained was finally divided by the original length of the pellet.
.
0 1986 American
Chemical Society
