Solubility of HFC32, HFC125, HFC134a, HFC143a, and HFC152a in a pentaerythritol tetrapentanoate ester

Article abstract of DOI:10.1021/je980235e

Solubilities of five different hydrofluorocarbons (HFCs) in a pentaerythritol tetrapentanoate ester (95% purity) have been measured at temperatures between 303.15 and 363.15 K and pressures between 0.05 and 1.9 MPa. Henry's constant and the activity coefficient for HFCs at infinite dilution were derived for measurements below 0.26 MPa. The measurements were made with an isochoric method with an uncertainty of less than 2% for Henry's constant and less than 3% at high pressure. Within the investigated temperature range, solubilities for HFCs decrease in the following order: HFC152a > HFC134a > HFC32 > HFC125 > HFC143a. The experimental data have been correlated with a Flory-Huggins model with an extended temperature dependence, which is able to describe the data with a deviation from measured data of less than 2%.

Full text of DOI:10.1021/je980235e

J . Chem. Eng. Data 1999, 44, 823-828
823
Solu bility of HF C32, HF C125, HF C134a , HF C143a , a n d HF C152a in a
P en ta er yth r itol Tetr a p en ta n oa te Ester
A° sa Wa h lstr o1 m * a n d Len n a r t Va m lin g
Department of Heat and Power Technology, Chalmers University of Technology, S-412 96 Gothenburg, Sweden
Solubilities of five different hydrofluorocarbons (HFCs) in a pentaerythritol tetrapentanoate ester (95%
purity) have been measured at temperatures between 303.15 and 363.15 K and pressures between 0.05
and 1.9 MPa. Henry’s constant and the activity coefficient for HFCs at infinite dilution were derived for
measurements below 0.26 MPa. The measurements were made with an isochoric method with an
uncertainty of less than 2% for Henry’s constant and less than 3% at high pressure. Within the investigated
temperature range, solubilities for HFCs decrease in the following order: HFC152a > HFC134a > HFC32
>
HFC125 > HFC143a. The experimental data have been correlated with a Flory-Huggins model with
an extended temperature dependence, which is able to describe the data with a deviation from measured
data of less than 2%.
In tr od u ction
As far as we know, no one has measured the solubility
of HFCs in a compressor oil consisting of only one substance
with a known structure. Such measurements are very
valuable in the evaluation and development process of a
predictive model. Therefore we have made measurements
of the solubility of five different HFCs in a pentaerythritol
tetrapentanoate ester. These measurements are the begin-
ning of a series of measurements with different penta-
erythritol esters of known structure.
Restriction on the use of chlorofluorocarbons as working
fluids in heat pumps and refrigeration systems has led to
the use of hydrofluorocarbons (HFCs) as alternatives.
However, to use HFCs as working fluids, a compressor oil
that is compatible with HFCs needs to be found. Important
factors for selection of a successful compressor oil are
viscosity and lubricity, which are dependent on the working
fluid’s solubility in the compressor oil. Knowledge of
solubility is also required to understand the influence on
heat transfer in the working fluid system’s evaporator and
condenser or when the working fluid’s oil-separation
system is to be sized.
Exp er im en ta l Section
Isoch or ic Tech n iqu e. The experimental technique is
an isochoric one where the amount of gas absorbed in a
known quantity of liquid solvent is calculated from the
pressure change in a gas system of known volume, observed
during the absorption of the gas. The experimental equip-
ment together with calibration and calculations is thor-
oughly described in Wahlstr o¨ m and Vamling (1997a).
In the literature polyalkylene glycols (PAG) and pen-
taerythritol tetraalkyl esters (POE) are frequently dis-
cussed as likely compressor oils for the more environmen-
tally safe working fluids, but data about these systems are
scarce. Solubility measurements have previously been
carried out by Thomas and Pham (1992) for HFC134a in
different PAG oils; Takaishi and Oguchi (1993, 1995) for
HFC32 and HFC134a in a POE oil; Grebner and Crawford
The apparatus consists of three sections (Figure 1): (1)
a liquid-bath thermostat 1 containing the equilibrium cell
with a magnetic stirrer; (2) another liquid-bath thermostat
2 with a gas bottle of calibrated volume; (3) an air-bath
thermostat with a pressure transmitter and connections
between sections 1 and 2 and to gas storage and a vacuum
pump.
(1993) for HFC134a in an ester oil and a PAG oil; Hend-
erson (1994) for HFC32, HFC134a, HFC125, HFC143a, and
HFC152a in different POE oils; Martz et al. (1996) for
HFC134a and HFC125 in a POE oil; and Cavestri et al.
(
1994) for HFC134a in POE and PAG oils. All authors,
The pure solvent was first degassed and weighed before
the equilibrium cell was placed into the liquid-bath ther-
mostat 1. The gas was fed into the evacuated gas system,
the gas bottle located in the liquid-bath thermostat 2, and
the connection tubes located in the air-bath thermostat.
The temperature was maintained the same in the air bath
and in the liquid bath, and the pressure was recorded. After
the interconnecting valve was opened, the gas was admit-
ted into the cell and absorbed by the stirred solvent. The
progress of absorption was indicated by the decrease of the
pressure. When the pressure remained constant, equilib-
rium was reached. The next step was to change the
temperature for the equilibrium cell. When the pressure
again became constant, a new equilibrium point had been
reached. The measurements for one pressure level were
finished when all the desired temperatures were reached.
except the last one, have also correlated their data using
models that can accurately describe the solubility of the
specific system measured. However, these authors did not
describe whether the oils consist of mixtures of different
oil molecules or give the structure of the oil molecules. It
is therefore difficult to use those correlation models for a
mixture which has not yet been measured, for example
when a mixture with a new possible compressor oil is
investigated. Experimental investigations are time-con-
suming and thus expensive. New thermodynamic tools that
can predict the behavior of a mixture of HFCs and
compressor oils will be very useful in the search for
successful compressor oils.
*
Corresponding author. E-mail asa.wahlstrom@hpt.chalmers.se.
1
0.1021/je980235e CCC: $18.00 © 1999 American Chemical Society
Published on Web 06/10/1999

8
24 J ournal of Chemical and Engineering Data, Vol. 44, No. 4, 1999
where x
1
is the amount of absorbed gas in the solvent,
calculated from the drop of pressure in the gas system of
known volume and temperature. The exact amount of gas
in the system is determined with the equation of state used
in the Refprop6 program (McLinden et al., 1998). The
exponential term is called the Poynting correction and
takes into account that the liquid is at equilibrium pressure
P, different from the saturation pressure of the solvent,
sat.
∞
P2 . The factor vj 1 is the partial molar volume in the
liquid phase at infinite dilution, which is calculated ac-
cording to Brelvi and O’Connell (1972). Taking into account
the equilibrium condition, the liquid-phase fugacity can be
written as
l
1
v
1
v
1 1
f ) f ) æ y P
(4)
Since the solvents have negligible vapor pressures, the
mole fraction y of the gas in the vapor phase is taken to
F igu r e 1. Solubility apparatus.
1
v
equal unity. The gas-phase fugacity coefficient æ is cal-
1
Then the procedure described above was repeated for the
next pressure level.
culated with the equation of state used in Refprop6.
The activity coefficient for the solute at infinite dilution
is calculated with
The density of the solute was determined with an Anton
Paar DMA 602 vibrating tube densitometer, described by
Wimby and Berntsson (1994). The tube was filled at 20 EC
with degassed solute, and then the temperature was
increased to the desired temperatures. Ethanol and twice-
distilled water were used as calibration liquids.
Ca lcu la tion of Absor bed Ga s. The mole fraction x
absorbed gas (1) in the solvent (2) is calculated according
to
He1,2
∞
1
γ )
(5)
l
(
f )*
1
l
1
of
The fugacity (f 1)* of a pure condensed component 1 at T
and P is obtained with
l
1
sat.
1
v (P - P
)
l
1
sat. sat.
n1
(f
)* ) P
1
æ
1
exp
(
)
(6)
RT
x1 )
(1)
n1 + n2
sat.
It is equal to P1 , the saturation pressure of component 1
at T, with two corrections. First, the fugacity coefficient
æ1 corrects for deviations of the saturated vapor from
where n
2
, the number of moles of the solvent, is calculated
sat.
from the mass. The number of moles of gas absorbed in
the solvent is calculated from
ideal-gas behavior, calculated with Refprop6. Second, the
Poynting correction takes into account that the liquid is
at equilibrium pressure P, different from the saturation
Vgas syst
Vgas syst
n1 )
-
+
sat.
l
1
v
1
l
1
[
pressure, P1 . The molar volume v of solute in the liquid
v (Tbeg.,Pbeg.
)
v (Tequilib gas syst,P)
phase is also calculated with the equation of state used in
Refprop6.
V2,cell - Vcell
vabs gas
/
1 -
(2)
v
1
v
]
[
]
Un cer ta in ty. The uncertainty of each experimental
quantity is given in Wahlstr o¨ m and Vamling (1997a), and
the total uncertainty of measurement is estimated to be
less than 2% at low pressure (Henry’s constant) and less
than 3% at high pressure. For the density measurements
the uncertainty in the temperature reading was ( 0.05 °C
and the uncertainty of the measured densities is estimated
to be less than 0.1%.
v (T,P)
v (T,P)
1
where Vgas syst and Vcell are the volume of the gas system
(sections 2 and 3) and the volume of the cell (section 1),
respectively. The volume V2,cell of the solvent in the cell is
calculated from its mass and measured density. The molar
v
volume v1 of the gas in the vapor phase at different
conditions is calculated with Refprop6 (McOLinden et al.,
Ch em ica ls. The purity according to the manufacturers
is 99.9% for all HFCs. All HFCs are supplied by DuPont
de Nemours, Dordrecht Works, The Netherlands, except
HFC134a which is supplied by ICI Chemicals & Polymers
Ltd., Runcorn, Cheshire, U.K.
1
998), where T and P are the conditions of the cell at
equilibrium, Tbeg. and Pbeg. are the conditions at sections 2
and at the beginning of the measurement, and
equilib gas syst is the temperature in sections 2 and 3 at
3
T
equilibrium. The volume vabs gas of the absorbed gas in the
solvent is calculated as the liquid molar volume of the gas
at equilibrium conditions T and P, if liquid exists during
those conditions; otherwise, as the liquid molar volume at
the bubble point at equilibrium temperature T, unless T
is above the critical temperature. Then it is calculated as
the partial molar volume in the liquid phase at infinite
dilution according to Brelvi and O’Connell (1972).
Henry’s constant of a solute (1) in a solvent (2) is
calculated according to the equation
Thepentaerythritoltetrapentanoateester,C(CH
2 2
OCOCH -
CH CH CH ,was synthesized by a reaction between a
2
2
3 4
)
carboxylic acid and an alcohol catalyzed by an acid. The
method used is basically the same as the one used by Black
and Gunstone (1990) except that the raw product was
vacuum-distilled instead of separated by chromatography
and that p-xylene was used as a solvent instead of m-
xylene. Pentaerythritol (Aldrich 98%, 0.64 mol, 89 g) was
mixed with valeric acid (Aldrich 99%, 2.82 mol, 310 mL),
p-toluensulfonic acid (40 mmol, 7.6 g), and 160 mL of
p-xylene in a 1000 mL round flask. A Dean and Stark
apparatus was filled with p-xylene, and the mixture was
reflux-boiled until no more water was trapped in the Dean
l
1
f
He1,2 ) lim
(3)
(
∞
1
sat.
2
)
x1f0
x exp( vj (P - P )/RT)
1

J ournal of Chemical and Engineering Data, Vol. 44, No. 4, 1999 825
and Stark trap (this took about 3.5 h). The reaction product
rx2
Φ * )
(11)
was cooled and mixed with diethyl ether into a 2 L
separating funnel. The mixture was washed twice with 5%
NaOH, twice with water, and once with NaCl. Thereafter
2
x1 + rx2
where r is the number of segments in a solvent molecule.
The last term in this expression is added to the activity
coefficient to apply the Flory-Huggins theory to a real
polymer solution, that is, a solution which has differences
in molecular interactions. The dimensionless parameter is
given in terms of the interchange energy by
2 4
the solution was dried with Na SO and the solvent was
removed with a cyclic evaporator. The small amount of
p-xylene left in the raw product was removed in the
subsequent vacuum distillation at 0.08 mbar and 190-205
EC. The product was a 274 g slightly yellow oil (90%
1
exchange). The distilled product was analyzed by H NMR,
1
3
C NMR, and IR spectroscopies, and the purity was
0
ø ) w /kT
(12)
estimated to be better than 95%.
Cor r ela tion of Exper im en ta l Da ta . Within the inves-
tigated temperature range, it was found that Henry’s
constant on a logarithmic scale versus the reciprocal
temperature 1/T is practically a straight line. Therefore we
have correlated the experimental data with the equation
0
where k is Boltzmann’s constant and w is the interchange
energy between solute molecules and solvent segments.
Previous work (Wahlstr o¨ m and Vamling, 1997b) has shown
that the temperature representation can be improved if an
1
interchange energy parameter, w , is introduced besides
0
the interchange energy parameter w . Equation 12 will
ref
1,2
1
1
ln He1,2 ) ln He + B
-
(7)
then be extended to
(T
ref
)
T
0
1
w
w
ref
,2
ø )
1 +
(13)
where He1 is Henry’s constant at a reference tempera-
ture T (333.15 K) and the slope of the line is represented
(
)
kT
T
ref
by the parameter B.
0
1
The empirical parameters w , w , and r were fitted to the
experimental data by regression, performed by minimiza-
tion of the sum of squares
Our previous research (Wahlstr o¨ m and Vamling, 1997b)
has shown that the Flory-Huggins model can describe
similar systems well, and thus we have chosen to correlate
the experimental data with that model. The basic equation
to represent vapor-liquid equilibria is the equality of
fugacities, where the vapor-phase fugacity is calculated
with eq 4 and the liquid-phase fugacity is calculated with
Np
2
Pcalc,i
SS )
1 -
(14)
∑
(
)
Pexpt,i
i)1
where Pcalc is the calculated pressure, Pexpt is the experi-
l
1
l
1
p
mental pressure, and N is the number of experimental
f ) (f )*x γ
(8)
1 1
points. The result of the parameter estimation may be
expressed as a relative deviation of the calculated equilib-
rium pressures for each mixture. The relative deviation DP
l
1
The fugacity (f )* of a pure condensed component 1 at T
1
and P is calculated according to eq 6, and x is the liquid
fraction of component 1. For HFCs above their critical
(%) is calculated with
temperature no liquid phase exists, and therefore it is
Np
Pcalc,i - Pexpt,i
l
necessary to estimate the fugacity, (f )*. Chappelow and
1
DP )
/Np
(15)
∑
|
|
Prausnitz (1974) gave an equation for extrapolation of the
fugacity for ethane, propane, and butane, also used satis-
factorily by Stelmachowski and Ledakowicz (1993). The
extrapolation was represented with four coefficients, but
in this work we have only used the first two coefficients as
in
[
]
Pexpt,i
i)1
Resu lts
The experimental solubilities for the five HFCs in a
pentaerythritol tetrapentanoate ester are presented in
Tables 1-5. Measurements were made at four tempera-
tures. Each temperature Henry’s constant has been derived
for points with equilibrium pressures below 0.26 MPa. For
the mixtures of HFC125, HFC134a, and HFC32 three
points of different equilibrium pressure were used for the
derivation at each temperature, while only two points were
used for the mixtures of HFC143a and HFC152a due to
some experimental problems with leaks. These values were
then fitted to eq 7, and the resulting parameters are given
in Table 6 together with their 95% confidence interval. The
representation can be seen in Figure 2 where the derived
experimental Henry’s constants are plotted on a logarith-
mic scale versus the reciprocal temperature 1/T together
with the calculated Henry’s constants from eq 7. Compari-
son of Henry’s constant at the reference temperature (Table
C2
l
1
ln (f )* ) C +
(9)
1
T
The coefficients C
1
and C
2
are determined for each P by
using the fugacities calculated with eq 6 for reduced
temperatures 0.93 and 0.96.
The deviation from ideality in eq 8 is expressed by an
1
activity coefficient in the liquid phase, γ , which is obtained
from an excess Gibbs energy model. Here we use the
Flory-Huggins model (Prausnitz et al., 1986). For a
mixture of fluids that have no difference in molecular
interactions or in free volume but have flexible molecules
that differ significantly in size, Flory and Huggins inde-
pendently derived an expression for the activity coefficient
6
) shows that the solubilities for HFCs in pentaerythritol
tetrapentanoate ester decrease in the following order:
HFC152a > HFC134a > HFC32 > HFC125 > HFC143a.
Values for the activity coefficient at infinite dilution are
given in Table 7. The activity coefficient has not been
derived for temperatures where the solute is above its
critical temperature.
1
r
1
r
2
ln γ ) ln 1 - 1 - Φ * + 1 - Φ * + ø(Φ *) (10)
1
(
(
)
2
)
(
)
2
2
2
The segment fraction Φ * of sites occupied by the solvent
is

8
26 J ournal of Chemical and Engineering Data, Vol. 44, No. 4, 1999
Ta ble 1. Mole F r a ction Solu bility of HF C32 in
P en ta er yth r itol Tetr a p en ta n oa te
Ta ble 4. Mole F r a ction Solu bility of HF C143a in
P en ta er yth r itol Tetr a p en ta n oa te
T/K
P/MPa
x
T/K
P/MPa
x
T/K
P/MPa
x
T/K
P/MPa
x
3
3
3
3
3
3
3
3
3
3
3
3
3
3
03.17
03.13
03.15
03.16
03.16
03.15
03.15
23.28
23.24
23.27
23.29
23.26
23.27
23.27
0.0711
0.0937
0.1238
0.4078
0.5830
0.7947
1.0738
0.0903
0.1189
0.1560
0.5237
0.7510
1.0375
1.3827
0.0734
0.0955
0.1216
0.3307
0.4290
0.5243
0.6335
0.0620
0.0812
0.1042
0.2911
0.3814
0.4770
0.5687
343.18
343.12
343.10
343.17
343.14
343.11
343.15
363.22
363.20
363.34
363.23
363.24
363.36
363.25
0.1072
0.1415
0.1844
0.6253
0.8948
1.2485
1.6178
0.1226
0.1612
0.2095
0.7140
1.0153
1.4276
1.7949
0.0523
0.0690
0.0894
0.2547
0.3373
0.4304
0.5041
0.0440
0.0588
0.0769
0.2227
0.2993
0.3884
0.4461
303.15
303.18
303.16
303.17
303.14
303.13
303.15
323.27
323.28
323.27
323.27
323.25
323.24
323.27
0.0877
0.1525
0.4084
0.5525
0.6900
0.8557
1.1403
0.1060
0.1846
0.4988
0.6867
0.8691
1.0906
1.4762
0.0630
0.1059
0.2541
0.3375
0.4058
0.4855
0.6239
0.0526
0.0889
0.2137
0.2903
0.3533
0.4249
0.5377
343.16
343.15
343.17
343.17
343.12
343.13
343.17
363.21
363.25
363.24
363.25
363.19
363.18
363.23
0.1216
0.2119
0.5729
0.7953
1.0133
1.2738
1.6965
0.1352
0.2350
0.6339
0.8838
1.1292
1.4093
1.8427
0.0440
0.0748
0.1808
0.2514
0.3073
0.3716
0.4604
0.0370
0.0636
0.1531
0.2172
0.2664
0.3269
0.3907
Ta ble 2. Mole F r a ction Solu bility of HF C125 in
P en ta er yth r itol Tetr a p en ta n oa te
Ta ble 5. Mole F r a ction Solu bility of HF C152a in
P en ta er yth r itol Tetr a p en ta n oa te
T/K
P/MPa
x
T/K
P/MPa
x
T/K
P/MPa
x
T/K
P/MPa
x
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
03.15
03.16
03.14
03.16
03.14
03.14
03.15
03.15
23.27
23.26
23.26
23.29
23.26
23.26
23.27
0.0731
0.0991
0.1372
0.2590
0.3688
0.5309
0.6858
0.9926
0.0936
0.1281
0.1787
0.3385
0.4953
0.7169
0.9611
0.0675
0.0896
0.1265
0.2222
0.3163
0.4208
0.5206
0.6805
0.0551
0.0725
0.1033
0.1823
0.2687
0.3627
0.4627
323.27
343.16
343.17
343.15
343.15
343.15
343.16
343.16
363.17
363.23
363.26
363.28
363.27
363.16
363.25
1.3920
0.1119
0.1525
0.2139
0.6055
0.8740
1.1990
1.6970
0.1268
0.1732
0.2437
0.6931
0.9985
1.3860
1.8970
0.6141
0.0444
0.0587
0.0841
0.2238
0.3062
0.4004
0.5192
0.0363
0.0475
0.0684
0.1862
0.2563
0.3414
0.4317
303.16
303.15
303.16
303.15
303.15
303.15
323.26
323.26
323.28
323.26
323.25
323.26
0.0585
0.0770
0.1552
0.2158
0.2681
0.3800
0.0844
0.1103
0.2250
0.3157
0.4009
0.5647
0.1167
0.1532
0.2900
0.3764
0.4497
0.5824
0.1023
0.136
0.2636
0.3445
0.4183
0.5387
343.15
343.15
343.17
343.15
343.16
343.17
363.26
363.23
363.22
363.24
363.26
363.24
0.1101
0.1441
0.2958
0.4174
0.5405
0.7473
0.1346
0.1758
0.3635
0.5144
0.6753
0.9043
0.0883
0.1186
0.2357
0.3097
0.3820
0.4845
0.0753
0.1026
0.2083
0.2744
0.3434
0.4264
Ta ble 6. Nu m ber of P oin ts a n d Sp ecific P a r a m eter s w ith
Th eir 95% Con fid en ce In ter va l for Ca lcu la tin g Hen r y’s
Con sta n t w ith Equ a tion 7, a t Tem p er a tu r es betw een 303
a n d 363 K, for HF Cs in P en ta er yth r itol Tetr a p en ta n oa te
Ta ble 3. Mole F r a ction Solu bility of HF C134a in
P en ta er yth r itol Tetr a p en ta n oa te
ref
solute
Np
He1,2/MPa
B/K
T/K
P/MPa
x
T/K
P/MPa
x
HFC32
12
12
12
8
1.710 ( 0.016
2.050 ( 0.023
1.138 ( 0.011
2.332 ( 0.020
0.979 ( 0.017
-1873 ( 47
-2167 ( 56
-2300 ( 48
-1750 ( 42
-2277 ( 86
3
3
3
3
3
3
3
3
3
3
3
3
3
3
03.14
03.15
03.14
03.15
03.16
03.14
03.15
23.26
23.26
23.11
23.28
23.27
23.26
23.26
0.0515
0.0902
0.1321
0.1765
0.2727
0.3258
0.4583
0.0731
0.1296
0.1815
0.2556
0.3997
0.4854
0.6835
0.0907
0.1551
0.2219
0.2737
0.3962
0.4593
0.5942
0.0781
0.1348
0.1877
0.2418
0.3562
0.4190
0.5414
343.15
343.16
343.27
343.15
343.16
343.14
343.15
363.24
363.26
363.25
363.19
363.21
363.25
363.26
0.0939
0.1679
0.2254
0.3326
0.5239
0.6451
0.8953
0.1129
0.2031
0.2615
0.4046
0.6350
0.7914
1.0650
0.0662
0.1151
0.1554
0.2095
0.3130
0.3734
0.4760
0.0558
0.0973
0.1275
0.1779
0.2708
0.3264
0.4079
HFC125
HFC134a
HFC143a
HFC152a
8
The experimental points were correlated with the Flory-
Huggins model with an extended temperature dependence.
In Table 8 the number of points, the specific parameters
0
1
w , w , r, and the relative deviation DP of calculated and
experimental pressures are given for the different mixtures.
The representation can be seen in Figures 3-7, where
experimental and calculated pressures are plotted versus
composition of HFCs in pentaerythritol tetrapentanoate
ester, for various temperatures.
The experimental densities at various temperatures for
the pentaerythritol tetrapentanoate ester are presented in
Table 9 and depicted in Figure 8, together with data
measured on the same substance by Kishore and Shobha
F igu r e 2. Henry’s constant versus temperature for different
HFCs in pentaerythritol tetrapentanoate.
(
1992).
are around unity. In the temperature range measured in
this work HFC32 shows a negative deviation from Raoult’s
law while HFC143a has a positive deviation.
Discu ssion
The mixtures examined in this work are all only slightly
nonideal since the activity coefficients at infinite dilution
The experimental points were correlated both with the
Flory-Huggins model according to eq 12, w and r being
0

J ournal of Chemical and Engineering Data, Vol. 44, No. 4, 1999 827
Ta ble 7. Activity Coefficien t a t In fin ite Dilu tion for
HF Cs in P en ta er yth r itol Tetr a p en ta n oa te
T/K
solute
303.2
323.2
343.2
0.71
363.2
HFC32
0.67
0.93
0.88
1.29
0.84
0.68
1.04
0.91
1.30
0.86
HFC125
HFC134a
HFC143a
HFC152a
0.95
1.39
0.88
1.01
0.91
Ta ble 8. Nu m ber of P oin ts, Sp ecific P a r a m eter s for th e
F lor y-Hu ggin s Mod el w ith Exten d ed Tem p er a tu r e
Dep en d en ce, a n d th e Rela tive Devia tion DP for HF Cs
Dissolved in P en ta er yth r itol Tetr a p en ta n oa te
w0k-1/K
w /K
1
r
DP/%
solute
Np
HFC32
28
30
28
28
24
806
967
878
849
482
-197
-262
-221
-183
-204
9.78
3.47
6.17
5.28
4.62
1.35
1.46
1.91
1.95
1.94
HFC125
HFC134a
HFC143a
HFC152a
F igu r e 5. Experimental and Flory-Huggins-calculated pressures
versus mole fraction composition for HFC134a in pentaerythritol
tetrapentanoate at various temperatures.
F igu r e 3. Experimental and Flory-Huggins-calculated pressures
versus mole fraction composition for HFC32 in pentaerythritol
tetrapentanoate at various temperatures.
F igu r e 6. Experimental and Flory-Huggins-calculated pressures
versus mole fraction composition for HFC143a in pentaerythritol
tetrapentanoate at various temperatures.
F igu r e 4. Experimental and Flory-Huggins-calculated pressures
versus mole fraction composition for HFC125 in pentaerythritol
tetrapentanoate at various temperatures.
F igu r e 7. Experimental and Flory-Huggins-calculated pressures
versus mole fraction composition for HFC152a in pentaerythritol
tetrapentanoate at various temperatures.
1
regressed for each mixture while w is set equal to zero,
highly improved when the extended temperature depen-
dence was introduced.
Comparison of the densities measured in this work and
the results of Kishore and Shobha (1992) shows that the
0
and with the Flory-Huggins model according to eq 13, w ,
w , and r being all regressed simultaneously for each
mixture. The representation of experimental points was
1

8
28 J ournal of Chemical and Engineering Data, Vol. 44, No. 4, 1999
Ta ble 9. Den sity of P en ta er yth r itol Tetr a p en ta n oa te
Ester a t Va r iou s Tem p er a tu r es
Ack n ow led gm en t
For synthesizing the Pentaerythritol tetrapentanoate the
authors are grateful to Dr. Magnus Eriksson, Department
of Organic Chemistry, Chalmers University of Technology,
Gothenburg, Sweden.
T/K
density/(kg‚m-3)
T/K
density/(kg‚m-3
)
2
3
3
3
3
3
3
3
98.22
03.13
08.20
13.85
18.02
23.16
28.23
33.09
1015.0
1011.2
1007.3
1002.4
999.3
995.2
991.2
987.3
338.24
342.85
347.96
352.98
358.59
363.33
363.33
983.1
979.8
976.1
972.5
967.9
964.4
964.4
Liter a tu r e Cited
Black, K. D.; Gunstone, F. D. The Synthesis and Spectroscopic
Properties of Some Polyol Esters and Ethers. Chem. Phys. Lipids
1
990, 56, 169-173.
Brelvi, S. W.; O’Connell, J . P. Corresponding States Correlations for
Liquid Compressibility and Partial Molar Volumes of Gases at
Infinite Dilution in Liquids. AIChE J . 1972, 18, 1239-1243.
Cavestri, R. C.; Munk, J .; Menning, M. Solubility, Viscosity, and
Density Measurements of Refrigerant-Lubricant Mixtures. Proceed-
ings of the Annual Meeting of ASHRAE Transactions, J une 25-29,
1
2
994, Orlando, FL; ASHRAE: Atlanta, GA, 1994; Vol. 100, pp 220-
38.
Chappelow, C. C.; Prausnitz, J . M. Solubilities of Gases in High-Boiling
Hydrocarbon Solvents. AIChE J . 1974, 20, 1097-1104.
Grebner, J . J .; Crawford, R. R. Measurement of Pressure-Temperature-
Concentration Relations for Mixtures of R-12/Mineral Oil and
R-134a Synthetic Oil. Proceedings of the 1993 Winter Meeting of
ASHRAE Transactions, J an 23-27, 1993, Chicago, IL; ASHRAE:
Atlanta, GA, 1993; Vol. 99, pp 387-396.
Henderson, D. R. Solubility, Viscosity and Density of Refrigerant/
Lubricant Mixtures; Final Technical Report; Spauschus Associates,
Inc.: Eagle’s Landing, Stockbridge, GA, 1994.
Kishore, K.; Shobha, H. K. Structural Dependence of Density in High
Molecular Weight Esters. J . Chem. Eng. Data 1992, 37, 371-376.
Martz, W. L.; Burton, C. M.; J acobi, A. M. Local Composition Modeling
of the Thermodynamic Properties of Refrigerant and Oil Mixtures.
Int. J . Refrig. 1996, 19, 25-33.
McLinden, M. O.; Lemmon, E. W.; Klein, S. A.; Peskin, A. P. Refprop6
(
Computer Program for Thermodynamic and Transport Properties
of Refrigerants and Refrigerant Mixtures); NIST Standard Refer-
ence Database23, Version 6.0; National Institute of Standards and
Technology: Washington, DC, 1998.
F igu r e 8. Comparison of experimental densities of pentaeryth-
ritol tetrapentanoate versus temperature between this work and
Kishore and Shobha (1992).
Prausnitz, J . M.; Lichtenthaler, R. N.; Gomes de Azevedo, E. Molecular
Thermodynamics of Fluid-Phase Equilibria, 2nd ed.; PTR Prentice-
Hall Inc: Englewood Cliffs, NJ , 1986.
deviation between the data is around 0.4%. This is a little
above the combined margins of error since the uncertainty
of the equipment used in this work is 0.1% while the
uncertainty of the equipment used by Kishore and Shobha
is 0.2%. An explanation of the deviation between the two
works might be that the syntheses of the esters are slightly
different and therefore the byproduct may be different. The
purity of the esters may also be different; in this work the
purity is more than 95% while Kishore and Shobha have
not mentioned the exact purity. Another factor that may
influence the measured values is how well the esters are
degassed before the measurements.
Stelmachowski, M.; Ledakowicz, S. Prediction of Henry’s Constants
by the UNIFAC-FV Model for Hydrocarbon Gases and Vapors in
High-Boiling Hydrocarbon Solvents. Fluid Phase Equilib. 1993, 90,
2
05-217.
Takaishi, Y.; Oguchi, K. Solubility of the Solutions of HFC-134a and
Polyolester Based Oil. Proceedings of the IIR Conference, May 12-
1
4, 1993, Ghent, Belgium; 1993; International Institute of Refrig-
eration: Paris, France, Commission B1/2, pp 141-148.
Takaishi, Y.; Oguchi, K. Solubility of R-32 and Polyolester Lubricant
Mixtures. Proceedings of the 19th International Congress of Refrig-
eration, Aug 20-25, 1995, The Hague, The Netherlands; 1995;
International Institute of Refrigeration: Paris, France, Commission
B1, pp 568-574.
Thomas, R. H. P.; Pham, H. T. Solubility and Miscibility of Environ-
mentally Safer Refrigerants/Lubricants Mixtures. ASHRAE Trans.
Symp. 1992, 5, 783-788.
Wahlstr o¨ m, Å.; Vamling, L. The Solubility of HFC125, HFC134a,
HFC143a, and HFC152a in n-Eicosane, n-Hexadecane, n-Tridecane
and 2,6,10,14-Tetramethylpentadecane. Can. J . Chem. Eng. 1997a ,
75, 544-550.
Wahlstr o¨ m, Å.; Vamling, L. Prediction of Solubility for HFC Working
Fluids in Model Substances for Compressor Oils. Can. J . Chem. Eng.
Con clu sion s
The solubilities of five binary systems of five HFCs in a
pentaerythritol tetrapentanoate ester have been measured
with an isochoric method. Correlation with the Flory-
Huggins model with extended temperature dependence
shows that the theory is able to describe these kinds of
mixtures with a deviation from measured data of less than
1
997b, 75, 551-561.
Wimby, M.; Berntsson, T. Viscosity and Density of Aqueous Solutions
of LiBr, LiCl, ZnBr , CaCl , and LiNO . 1. Single Salt Solutions. J .
Chem. Eng. Data 1994, 39, 68-72.
2
2
3
2
%.
Henry’s constant has been derived for low-pressure
measurements and within the investigated temperature
range the logarithm of Henry’s constant versus the inverse
temperature forms a straight line for all HFCs. Solubilities
for HFCs decrease in the following order: HFC152a >
HFC134a > HFC32 > HFC125 > HFC143a.
Received for review September 22, 1998. Accepted April 9, 1999.
Financial support from the Swedish National Board for Industrial
and Technical Development (NUTEK) and the Swedish Council for
Building Research (BFR) is gratefully acknowledged.
J E980235E