You are on page 1of 9

Chemosphere 44 (2001) 855±863

www.elsevier.com/locate/chemosphere

The solubility of gases and vapours in dry octan-1-ol at 298 K


a,*
Michael H. Abraham , Joelle Le a, William E. Acree Jr. b, Peter W. Carr c,
Andrew J. Dallas d
a
Department of Chemistry, University College London, 20 Gordon Street, London, WC1H OAJ, UK
b
Department of Chemistry, University of North Texas, Denton, TX 76203-0068, USA
c
Department of Chemistry, University of Minnesota, 207 Pleasant Street S.E., Minneapolis, MN 55455, USA
d
Donaldson Corporate Technology, 9250 W. Bloomington Fwy, Bloomington, MN 55431, USA
Received 6 December 1999; accepted 29 June 2000

Abstract

Ostwald solubility coecients of 74 compounds in dry octan-1-ol at 298 K have been determined, and have been
combined with literature values and additional values we have calculated from solubilities in dry octan-1-ol and vapour
pressures to yield a total of 161 log LOctOH values at 298 K. These LOctOH values are identical to gas-to-dry octan-1-ol
partition coecients, often denoted as KOA . Application of the solvation equation of Abraham to 124 values as a
training set yielded a correlation equation with n ˆ 124, S:D: ˆ 0:125, r2 ˆ 0:9970 and F ˆ 7731. This equation was
then used to predict 32 values of log LOctOH as a test set, giving a standard deviation, S.D. of 0.131, an average absolute
deviation of 0.085 and an average deviation of )0.009 log units. The solvation equation for the combined 156 log LOctOH
values was
log LOctOH ˆ 0:1200:203R2 ‡ 0:560pH
X X2 H
‡ 3:560 aH
2 ‡ 0:702 b2
‡ 0:939 log L16 ;
n ˆ156; r2 ˆ 0:9972; S:D: ˆ 0:125;
F ˆ10573;

where, n is the number of data points (solutes), r the correlation coecient, S.D. the standard deviation and F is the F-
statistic. The independent
P H is an excess molar refraction, pH
variables are solute descriptors as follows: R2P 2 the dipolarity/
polarisability, a2 the overall or summation hydrogen-bond acidity, bH
2 the overall or summation hydrogen-bond
basicity and L16 is the Ostwald solubility coecient on hexadecane at 298 K. The equation is consistent with similar
equations for the solubility of gases and vapours into methanol, ethanol and propan-1-ol. It is suggested that the
equation can be used to predict further values of log LOctOH , for which the solute descriptors are known, to within 0.13
log units. Ó 2001 Elsevier Science Ltd. All rights reserved.

Keywords: Partition; Octanol±air partition; Hydrogen-bond; LFER; Octan-1-ol; Solubility; KOA

1. Introduction
*
Corresponding author. Tel.: +44-020-7679-4684; fax: +44-
020-7679-7463. The water±octanol partition system is widely used
E-mail address: m.h.abraham@ucl.ac.uk (M.H. Abraham). as a model for a wide variety of processes in which

0045-6535/01/$ - see front matter Ó 2001 Elsevier Science Ltd. All rights reserved.
PII: S 0 0 4 5 - 6 5 3 5 ( 0 0 ) 0 0 2 8 8 - 5
856 M.H. Abraham et al. / Chemosphere 44 (2001) 855±863

compounds are distributed between condensed phases or 2. Methodology


are transferred from one condensed phase to another
condensed phase. More recently, attention has been fo- The indirect method of obtaining log LOctOH values
cussed on processes in which compounds are distributed requires a knowledge of the solubility of the compound
between the gas phase and a condensed phase. The in octanol, COctOH , and the compound saturated vapour
water±octanol system is now not such a suitable model pressure, Po . From the latter, the concentration in the
for this gas-to-condensed phase transfer, but the gas± gas phase, CG , can be obtained. Then LOctOH is given by
octanol system could well prove to be very useful. The
partition of a compound between the gas phase and LOctOH ˆ COctOH =CG …2†
octanol can be described by the dimensionless Ostwald
solubility coecient, and if COctOH and CG are in the same units, say mol dm 3 ,
then LOctOH is a dimensionless quantity. Eq. (2) holds for
both liquid and solid solutes, but there are two important
‰concentration in octanolŠ provisos, (i) the solute must not be too soluble, because
LOctOH ˆ …1†
‰concentration in the gas phaseŠ then the secondary medium activity coecient of the
solute in the saturated solution may be far away from
Note that LOctOH is entirely equivalent to the gas-to-dry unity, and (ii) the solute must not form solvates with
octanol partition coecient, often denoted as KOW , octanol ± that is the solid or liquid in equilibrium with the
provided that the latter is expressed as a dimensionless saturated solution must be the unsolvated compound.
quantity. Nielsen et al. (1989, 1990) have correlated Results of calculations through Eq. (2) are given in
sensory irritation thresholds in mice to values of log Table 1, where most of the octanol solubilities are from
LOctOH and numerous workers (Paterson et al., 1991; Pinsuwan et al. (1995) and the saturated vapour pressures
Simonich and Hites, 1996; Tolls and MacLachlan, 1996; are from Stephenson and Malinowski (1987). In a num-
Keymeulen et al., 1997; Brown et al., 1998) have used ber of cases, these indirectly determined values of log
log LOctOH in correlations of gas to plant partitioning. LOctOH can be compared with directly determined values
Harner and Mackay (1995) have suggested that log (Harner and Mackay, 1995; Harner and Bidleman,
LOctOH would generally be useful as a physicochemical 1998); the agreement is very good, see Table 1. The
parameter to describe the partition of chemicals from complete list of 161 values of LOctOH is given in Table 2,
the gas phase to soil and to vegetation. and includes both the indirectly determined values from
However, the use of log LOctOH values has hitherto Table 1, directly determined values of log LOctOH (Cabani
been limited because of the paucity of published data. et al., 1991; Tse and Sandler, 1994; Gruber et al., 1997),
Welke et al. (1998) correlated gas to plant partitioning and values obtained from published activity coecients
with log LOctOH but were unable to obtain any more (Tewari et al., 1982; Thomas et al., 1982; Bhatia and
than 38 such values for a set of 50 comparatively simple Sandler, 1995) and saturated vapour pressures (Ste-
compounds. Values for gaseous solutes have been de- phenson and Malanowsky 1987). However, the majority
termined by Wilcock et al. (1978) and Bo et al. (1993), of the values in Table 2 are those we have directly de-
and a number of values have been published in the termined by the well-known method (Abraham, 1971) of
solubility data project. Several workers have obtained gas-chromatographic head-space analysis. We have not
log LOctOH for various other solutes (Cabani et al., 1991; included in Table 2, values for PCBs (Harner and Mac-
Tse and Sandler, 1994; Harner and Mackay, 1995; kay, 1995; Harner and Bidleman, 1996) and polychlori-
Gruber et al., 1997), but the total number of reported nated naphthalenes (Harner and Bidleman, 1998),
values is still very small. It is possible to combine gas- because at the moment we are unable to include these
to-water partitions with water-to-wet octanol parti- compounds in our solvation equation method.
tions, but this will provide values of the gas-to-wet We treat the data, as log LOctOH values at 298 K,
octanol partition, and not the required gas-to-dry oct- through the solvation equation (Abraham, 1993)
anol partition. At the moment we are working on an
log SP ˆ c ‡ rR2 ‡ spH
analysis of gas-to-wet octanol partitions, and hope to X
2
X H
report on this in the near future. In the present paper, ‡a aH
2 ‡ b b2 ‡ l log L16 ; …3†
our aim is to determine, both directly and indirectly,
values for partition between the gas phase and dry where the dependent variable, log SP, will be log LOctOH ,
octanol and to combine these with literature values in and the independent variables are solute descriptors as
order to obtain a reasonably comprehensive data set. follows (Abraham, 1993). R2 is an excessP H molar refrac-
We shall then analyse the data set using our solvation tion, pH
2 the dipolarity/polarisability, a2 the
P overall or
parameter model, in the hope that we can devise an summation hydrogen-bond acidity and bH
2 is the
equation that can be used for the prediction of further overall or summation hydrogen-bond basicity. In the
values of log LOctOH . ®nal descriptor, L16 is the Ostwald solubility coecient
M.H. Abraham et al. / Chemosphere 44 (2001) 855±863 857

Table 1
Calculation of log L in dry octan-1-ol at 298 K
Solute log Soct a log CG log Loct
b
Naphthalene )0.15 )5.34 5.19
Anthracene )1.91 )9.46b 7.55
Phenanthrene )0.45 )7.97b 7.52 7.57c
Acenaphthene )0.59 )6.9b 6.31
Fluorene )0.62 )7.45d; e 6.83 6.79c
Fluoranthene )0.76 )9.37e 8.61
Pyrene )0.90 )9.65b 8.75 8.80c
Biphenyl )0.13 )6.28e 6.15
1,4-Dichlorobenzene 0.20 )4.26e; f 4.46
1,3,5-Trichlorobenzene )0.16g )5.01e 4.85
1,2,3,5-Tetrachlorobenzene 0.15 )5.40d; e 5.55
1,2,4,5-Tetrachlorobenzene )0.92 )6.54h 5.62 (5.63h )
1,2,3,4-Tetrachlorobenzene (5.64h )
Pentachlorobenzene )0.56 )7.05h 6.49 (6.27h )
Hexachlorobenzene )1.86 )9.03h 7.17 (6.90h )
Hexamethylbenzene )0.89 )7.20e 6.31
1,4-Dibromobenzene )0.30 )5.51e 5.21
Hexachloroethane )0.28 )4.75d; e 4.47
4-Ethoxyacetanilide )0.84 )10.43e 9.59
Diphenylamine 0.03 )7.61e 7.64
trans-Stilbene )1.10i )8.58i 7.48
4-Hydroxybenzoic acid )0.17j; k )8.25b 8.08
2-Hydroxybenzoic acid 0.17l )7.27b 7.44
Methyl 4-hydroxybenzoate )0.10j; k )8.67b 8.57
a
All data from Tse and Sandler (1994), unless shown otherwise.
b
Abraham et al. (1999).
c
Directly determined values from Harner and Bidleman (1998).
d
Mackay and Shiu (1981).
e
Stephenson and Malinowski (1987).
f
Rordorf (1985).
g
Yalkowsky et al. (1994).
h
Harner and Mackay (1995). The values in parentheses are directly determined ones.
i
Abraham et al. (1998).
j
Martin et al. (1984).
k
Alexander et al. (1977).
l
Acree (unpublished).

on hexadecane at 298 K (Abraham et al., 1987), de®ned head-space is sampled to a Hewlett-Packard 5890 gas
similarly to Eq. (1). Descriptors for several hundred chromatograph by a vacuum transfer sampling system
solutes have been published (Abraham, 1993; Abraham connected to the sample cell through a series of heated
et al., 1994), and the entire data base of descriptors is fused silica transfer lines. The sampling system consists
now commercially available (Abraham, 2000). of a series of heated valves (Valco Instruments, Hous-
Eq. (3) has already been applied to the solubility of ton, TX) and a mechanical vacuum pump (Welch Duo-
gases and vapours in water (Abraham et al., 1994), Seal). Each of the homologous series of solutes were run
methanol (Abraham et al., 1998a), ethanol (Abraham as a group, and were separated using a methyl silicone
et al., 1998b), and propan-1-ol (Abraham et al., 1999), capillary column, with ¯ame ionisation detection; the
as well as to numerous sets of gas-chromatographic signal was recorded and integrated by a Hewlett-Pack-
data, so it is a well-tried and tested equation. ard 3396 integrator.
The sample cell (60 ml) was thermostatted by a water
jacket at 25:00  0:02°C and the cell temperature was
3. Experimental monitored with a calibrated calorimetric thermometer
(Brooklyn Thermometer, Brooklyn, NY). Anhydrous
The head-space gas-chromatographic apparatus, octan-1-ol (40 ml; < 0:005% water; Aldrich Chemical)
measurement techniques, and calculations are as previ- was pipetted into the cell, and stirred using a glass-
ously described in detail (Dallas, 1995). In brief, the encapsulated magnetic stirrer. Typically, ®ve increments
858 M.H. Abraham et al. / Chemosphere 44 (2001) 855±863

Table 2
Values of log L in dry octan-1-ol at 298 K
Solute log L References
a
Helium )1.72 Wilcock et al. (1978), Harner and Mackay (1995)
Neon )1.57 Bo et al. (1993)b
Argon )0.71 Bo et al. (1993)b
Krypton )0.24 Bo et al. (1993)b
b
Xenon 0.38
Hydrogena )1.29 b

Oxygen )0.77 Bo et al. (1993)


Nitrogen )1.02 Bo et al. (1993)
b
Nitrous oxide 0.33
Carbon monoxide )0.88 Wilcock et al. (1978)
Carbon dioxidea 0.16 Wilcock et al. (1978)
Methane )0.38 Bo et al. (1993)
b
Ethane 0.42
b
Propane 0.97
b
Butane 1.53
Pentanea 1.95 Tse and Sandler (1994)
Hexane 2.44 Bhatia and Sandler (1995)
Heptane 2.95 Gruber et al. (1997)
Octane 3.30 Cabani et al. (1991)
Cyclohexane 2.71 Tse and Sandler (1994)
Methylcyclohexanea 3.05 Gruber et al. (1997)
c
Ethene 0.28
Pent-1-ene 1.93 Gruber et al. (1997)
Hex-1-ene 2.41 Tewari et al. (1982), Gruber et al. (1997)
Oct-1-ene 3.35 Tewari et al. (1982)
Non-1-enea 3.83 Tewari et al. (1982)
2-Methylbuta-1,3-diene 2.06 Thomas et al. (1982)
Cyclohexene 2.83 Gruber et al. (1997)
Tetra¯uomethane )0.95 Bo et al. (1993)
d
Chloromethane 1.39
Dichloromethanea 2.27 This work
Trichloromethane 2.80 This work
Tetrachloromethane 2.79 This work
1,1-Dichloroethane 2.41 Tse and Sandler (1994)
1,2-Dichloroethane 2.78 This work
1,1,1-Trichloroethanea 2.70 This work
1,1,2-Trichloroethane 3.40 Tse and Sandler (1994)
Hexachloroethane 4.47 Table 1
1-Chloropropane 2.24 This work
1,2-Dichloropropane 2.96 Bhatia and Sandler (1995)
1-Chlorobutanea 2.72 This work
cis-1,2-Dichloroethene 2.56 Tse and Sandler (1994)
Trichloroethene 2.99 Tse and Sandler (1994)
Tetrachloroethene 3.48 This work
Dibromomethane 3.07 This work
Bromoethanea 2.11 Thomas et al. (1982)
Iodomethane 2.16 Thomas et al. (1982)
Iodoethane 2.59 Thomas et al. (1982)
Dibromochloromethane 3.59 Tse and Sandler (1994)
d
Dimethylether 1.37
Diethylethera 2.19 This work
Dipropylether 2.97 This work
Diisopropylether 2.66 This work
Dibutylether 3.89 This work
Diisobutylether 3.40 This work
Dipentylethera 4.80 This work
t-Butyl methyl ether 2.58 This work
M.H. Abraham et al. / Chemosphere 44 (2001) 855±863 859

Table 2 (Continued)
Solute log L References
Tetrahydrofuran 2.86 This work
Tetrahydropyran 3.22 This work
Dioxane 3.17 This work
Propanonea 2.31 This work
Butanone 2.77 This work
Pentan-2-one 3.19 This work
Pentan-3-one 3.20 This work
3-Methylbutan-2-one 3.04 This work
Hexan-2-onea 3.68 This work
Heptan-2-one 4.15 This work
Cyclopentanone 3.67 This work
Methyl formate 1.75 This work
Ethyl formate 2.19 This work
Propyl formatea 2.66 This work
Methyl acetate 2.31 This work
Ethyl acetate 2.70 This work
Propyl acetate 3.17 This work
Butyl acetate 3.65 This work
Pentyl acetatea 4.12 This work
Hexyl acetate 4.58 This work
Isopropyl acetate 2.93 This work
Isobutyl acetate 3.45 This work
Isopentyl acetate 3.94 This work
Ethyl propanoatea 3.15 This work
Ethyl butanoate 3.56 This work
Acetonitrile 2.31 This work
Proprionitrile 2.69 This work
Butanonitrile 3.12 This work
2-Cyanopropanea 2.87 This work
Pentanonitrile 3.60 This work
Hexanonitrile 4.08 This work
e
Butylamine 3.61
b
Methylamine 1.90
Dimethylaminea 2.00 b
e
Dipropylamine 3.59
Nitromethane 2.52 This work
Nitroethane 2.88 This work
Nitropropane 3.25 This work
Dimethylformamidea 4.38 This work
Methanol 2.84 This work
Ethanol 3.20 This work
Propan-1-ol 3.68 This work
Propan-2-ol 3.38 This work
Butan-1-ola 4.19 This work
2-Methylpropan-1-ol 3.93 This work
Butan-2-ol 3.80 Cabani et al. (1991), Gruber et al. (1997)
2-Methylpropan-2-ol 3.50 This work
Pentan-1-ol 4.69 This work
3-Methylbutan-1-ola 4.52 This work
Hexan-1-ol 5.18 This work
Octan-1-ol 6.03 Cabani et al. (1991)
Hexadecan-1-ol 9.90 Table 1
Octadecan-1-ol 10.93 Table 1
Eicosan-1-ola 12.06 Table 1
2-Chloroethanol 4.30 Table 1
Cyclohexanol 5.18 This work
Acetic acid 4.31 This work
Dimethylsulfoxide 4.96 This work
860 M.H. Abraham et al. / Chemosphere 44 (2001) 855±863

Table 2 (Continued)
Solute log L References
Sulphur hexa¯uoridea )0.30 Wilcock et al. (1978), Bo et al. (1993)
Carbon disulphide 2.28 Thomas et al. (1982)
f
Tetramethyltin 2.62
Benzene 2.80 This work
Toluene 3.31 This work
Ethylbenzenea 3.72 Tewari et al. (1982), Gruber et al. (1997)
o-Xylene 3.90 Tewari et al. (1982), Gruber et al. (1997)
m-Xylene 3.79 Tewari et al. (1982), Gruber et al. (1997)
p-Xylene 3.79 Gruber et al. (1997)
Propylbenzene 4.09 Tewari et al. (1982)
Isopropylbenzenea 3.98 Tewari et al. (1982)
Hexamethylbenzene 6.31 Table 1
trans-Stilbene 7.48 Table 1
Biphenyl 6.15 Table 1
Naphthalene 5.19 Table 1
Acenaphthenea 6.31 Table 1
Fluorene 6.83 Table 1
Anthracene 7.55 Table 1
Phenanthrene 7.52 Table 1
Fluoranthene 8.61 Table 1
Pyrenea 8.75 Table 1
1,2-Dichlorobenzene 4.36 Table 1
1,4-Dichlorobenzene 4.46 Table 1
1,2,3-Trichlorobenzene 5.19 Table 1
1,3,5-Trichlorobenzene 4.85 Table 1
1,2,3,4-Tetrachlorobenzenea 5.64 Table 1
1,2,3,5-Tetrachlorobenzene 5.55 Table 1
1,2,4,5-Tetrachlorobenzene 5.62 Table 1
Pentachlorobenzene 6.49 Table 1
Hexachlorobenzene 7.17 Table 1
1,4-Dibromobenzenea 5.21 Table 1
Anisole 4.01 This work
a
Benzonitrile 4.46
Diphenylamine 7.64 Table 1
4-Ethoxyacetanilide 9.59 Table 1
4-Hydroxybenzoic acid 8.08 Table 1
2-Hydroxybenzoic acida 7.44 Table 1
Methyl 4-hydroxybenzoate 8.57 Table 1
Piperidine 4.04 Cabani et al. (1991)
N-Methylpyrrolidine 3.64 Cabani et al. (1991)
Pyrrolidinea 4.07 Cabani et al. (1991)
Dimethylacetamide 5.33 This work
Propanal 3.02 This work
Butanal 3.39 This work
Hexanal 4.41 This work
Octanal 5.36 This work
a
Test set.
b
Solubility data project.
c
Boyer and Bircher (1960).
d
Gerrard (1972).
e
Personal communication from Prof. Cabani.
f
Abraham (1972).

of a solute mixture (100 ll), prepared gravimetrically, space (400 ll) is vacuum transferred to a sample loop
were added to the octanol using a computer controlled (100 ll) and injected into the chromatograph.
autotitrator (Model ABU-11, Radiometer, Copenhagen, External standards of the vapour phase were pre-
Denmark). After equilibration, a sample of the head- pared gravimetrically by addition of a known amount of
M.H. Abraham et al. / Chemosphere 44 (2001) 855±863 861

Table 3
Observed log LOctOH and calculated values on Eq. (4) for the test set
Solute Obs. Calc. Obs. ) calc.
Helium )1.72 )1.730 0.010
Hydrogen )1.29 )1.226 )0.064
Carbon dioxide 0.16 0.354 )0.194
Pentane 1.95 1.907 0.043
Methylcyclohexane 3.05 2.976 0.074
Non-1-ene 3.83 3.766 0.064
Dichloromethane 2.27 2.420 )0.150
1,1,1-Trichloroethane 2.70 2.669 0.031
1-Chlorobutane 2.72 2.688 0.032
Bromoethane 2.11 2.114 )0.004
Diethylether 2.19 2.220 )0.030
Dipentylether 4.80 4.893 )0.093
Propanone 2.31 2.324 )0.014
Hexan-2-one 3.68 3.651 0.029
Propyl formate 2.66 2.760 )0.100
Pentyl acetate 4.12 4.118 0.002
Ethyl propanoate 3.15 3.137 0.013
2-Cyanopropane 2.87 2.898 )0.028
Dimethylamine 2.00 2.269 )0.269
Dimethylformamide 4.38 4.046 0.334
Butan-1-ol 4.19 4.171 0.019
3-Methylbutan-1-ol 4.52 4.541 )0.021
Eicosan-1-ol 12.06 11.723 0.337
Sulphur hexa¯uoride )0.30 )0.439 0.139
Ethylbenzene 3.72 3.698 0.022
Isopropylbenzene 3.98 3.981 )0.001
Acenaphthene 6.31 6.385 )0.075
Pyrene 8.75 8.790 )0.040
1,2,3,4-Tetrachlorobenzene 5.64 5.954 )0.314
1,4-Dibromobenzene 5.21 5.164 0.046
2-Hydroxybenzoic acid 7.44 7.408 0.032
Pyrrolidine 4.07 4.175 )0.105

the solute mixture to a 12 l glass bulb, and allowed to pounds and a randomly chosen test set of 32
vapourise completely. A sample of the standard vapour compounds, as shown in Table 2. Application of the
is transferred to the chromatograph exactly as for the solvation Eq. (3) to the training set yielded the correla-
test samples. Each measurement of LOctOH was carried tion equation
out at ®ve di€erent solute concentrations. No depen-
dence on solute concentration was noted, and the LOctOH log LOctOH ˆ 0:108 0:177 R2 ‡ 0:567 pH
X X H2
values were taken to be those at in®nite dilution. The ‡ 3:565 aH2 ‡ 0:703 b2
octan-1-ol was stored over molecular sieve 4A, and the
‡ 0:932 log L16 ;
purity of the octan-1-ol and of the solutes was checked
by gas chromatography. n ˆ 124; r2 ˆ 0:9970;
S:D: ˆ 0:125; F ˆ 7731; …4†

4. Results and discussion


Table 4
We have assembled in Table 2, values of log LOctOH
The available solvation descriptors
for 161 solutes, covering a reasonably wide range of
compound type. Although this compilation is useful as it Descriptor No. Min Max
is, the possible prediction of log LOctOH for further R2 3870 )1.37 4.62
H
compounds would greatly enhance the value of the p
P2 3330 )0.58 5.60
compilation. A preliminary analysis revealed that ®ve aH 4145 0.00 2.10
P 2

compounds were outliers, and so we divided the re- bH


2 3015 0.00 4.52
maining 156 solutes into a training set of 124 com- log L16 2250 )1.74 29.97
862 M.H. Abraham et al. / Chemosphere 44 (2001) 855±863

Table 5
Coecients in Eq. (3) for gas±solvent partitions
Solvent c r s a b l
Water )1.271 0.822 2.743 3.904 4.814 )0.213
Methanol )0.004 )0.215 1.173 3.701 1.432 0.769
Ethanol 0.012 )0.206 0.789 3.635 1.311 0.853
Propan-1-ol )0.028 )0.185 0.648 4.022 1.043 0.869
Octan-1-ol )0.120 )0.203 0.560 3.576 0.702 0.939

where, n is the number of data points (solutes), r the are in Table 5. There is a gradual progression in the
correlation coecient, S.D. the standard deviation and coecients from solvent water to solvent octan-1-ol,
F is the F-statistic. The goodness-of-®t of Eq. (4) is ex- with the minor exception of the a-coecient for propan-
cellent. Of the ®ve outliers to Eq. (4), four were the 1-ol. The coecients in Eq. (5) are quite in line with
aliphatic aldehydes; it is known that these compounds those for the other alcohols we have studied, so that Eq.
form aldols with aliphatic alcohols, and so their exclu- (5) is chemically reasonable, as well as being statistically
sion is warranted on chemical grounds. The other outlier very good. As for the other dry alcohols in Table 5, bulk
was dimethylacetamide, but its log LOctOH value is so octan-1-ol is shown to be weakly dipolar/polarisable, a
large that there is considerable experimental uncertainty strong hydrogen-bond base, and only a moderate-to-
here. weak hydrogen-bond acid. The l-coecient indicates
A standard method of assessing the predictive capa- that as the alcohol chain length increases, the bulk al-
bility of an equation is to apply it to a test set of com- cohol functions more and more as a lipophilic solvent;
pounds not used to set up the equation. For the 32 data note that for solvent hexadecane l ˆ 1 by de®nition.
test set that we set aside, values of log LOctOH were
predicted through Eq. (4), and the predicted and ob-
served values were compared, see Table 3. The standard Acknowledgements
deviation between predicted and observed log LOctOH
values is only 0.131, the average absolute deviation is We are grateful to Roche Products Ltd., for a re-
0.085, and the average deviation is )0.009 log units. It is search studentship (to J.L.), and we thank the late Prof.
therefore clear that log LOctOH values can be predicted Sergio Cabani for a number of log LOctOH values.
through Eq. (4) to around 0.13 log units, for any solute
for which descriptors are available. In Table 4, we give
details of the number of solutes for which we have the References
relevant descriptors. Even with the proviso that the
compound descriptors are within the descriptor space of Abraham, M.H., 1971. Substitution at saturated carbon. Part
the solutes used to set up the correlation equation, it is IX. Free energies of transfer from methanol to aqueous
evident that values of log LOctOH can be predicted for a methanol of tetra-alkyltins and the transition states in the
huge number of compounds. bimolecular substitution of tetra-alkyltins by mercuric
chloride. J. Chem. Soc. A, 1061±1068.
We can combine the training set and the test set to
Abraham, M.H., 1972. J. Chem. Soc., Perkin Trans. II, 1343±
give the ÔbestÕ equation for the correlation and predic- 1357.
tion of log LOctOH values Abraham, M.H., 1993. Scales of solute hydrogen-bonding ±
their construction and application to physicochemical and
log LOctOH ˆ 0:120 0:203R2 ‡ 0:560pH biochemical processes. Chem. Soc. Rev. 22, 73±83.
X X H2
‡ 3:576 aH
2 ‡ 0:702 b2 Abraham, M.H., 2000. The ABSOLV data base, Sirius analyt-
ical instruments, Riverside. Forest Row Business Park,
‡ 0:939 log L16 ; Forest Row, East Sussex, RH18 5DW, UK.
n ˆ 156; r2 ˆ 0:9972; Abraham, M.H., Andonian-Haftvan, J., Whiting, G.S., Leo,
A., Taft, R.W., 1994. Hydrogen bonding. Part 34. The
S:D: ˆ 0:125; F ˆ 10573: …5†
factors that in¯uence the solubility of gases and vapours in
The coecients and statistics for Eq. (5) are very close to water at 298 K, and a new method for its determination. J.
Chem. Soc. Perkin Trans. 2, 1777±1791.
those for Eq. (4), but we suggest Eq. (5) be used for
Abraham, M.H., Green, C.E., Acree, W.E., Jr., Hernandez,
further prediction of log LOctOH values. C.E., Roy, L.E., 1998. J. Chem. Soc., Perkin Trans. 2, 2677±
We can compare the coecients in Eq. (5) with those 2681.
we have obtained previously for the solution of gases Abraham, M.H., Grellier, P.L., McGill, R.A., 1987. Determi-
and vapours in water (Abraham et al., 1994) and in nation of olive oil-gas and hexadecane-gas partition coef-
other alcohols (Abraham et al., 1998a,b, 1999); details ®cients, and calculation of the corresponding olive oil-water
M.H. Abraham et al. / Chemosphere 44 (2001) 855±863 863

and hexadecane-water partition coecients. J. Chem. Soc., compounds as environmental compartment. J. Chromatogr.
Perkin Trans. 2, 797±803. A 774, 213±221.
Abraham, M.H., Le, J., Acree Jr., W.E., Carr, P.W., 1999. Mackay, D., Shiu, W.Y., 1981. J. Phys. Chem. Ref. Data 10,
Solubility of gases and vapours in propan-1-ol at 298 K. J. 1175±1199.
Phys. Org. Chem. 12, 675±680. Martin, A., Wu, P.L., Beerbower, A., 1984. J. Pharm. Sci. 73,
Abraham, M.H., Whiting, G.S., Carr, P.W., Ouyang, H., 188±194.
1998a. Hydrogen bonding. Part 45. The solubility of gases Nielsen, G.D., Thomsen, E.S., Alarie, Y., 1990. Sensory irritant
and vapours in methanol at 298 K: an LFER analysis. J. receptor compartment properties. Equipotent vapour con-
Chem. Soc., Perkin Trans. 2, 1385±1390. centrations related to saturated vapour concentrations,
Abraham, M.H., Whiting, G.S., Shuely, W.J., Doherty, R.M., octanol-water, and octanol-gas partition coecients. Acta
1998b. The solubility of gases and vapours in ethanol ± the Pharm. Nord. 1, 31±44.
connection between gaseous solubility and water-solvent Nielsen, G.D., Yamagiwa, M., 1989. Structure-activity rela-
partition. Can. J. Chem. 76, 703±709. tionships of airway irritating aliphatic amines. Receptor
Acree, Jr.,W.E., unpublished work. activation mechanisms and predicted industrial exposure
Alexander, K.S., Mauger, J.W., Petersen, Jr., H., Paruta, A.N., limits.. Chem.-Biol. Interactions 71, 223±244.
1977. J. Pharm. Sci. 66, 42±48. Paterson, S., Mackay, D., Bacci, E., Calamari, D., 1991.
Bhatia, S.R., Sandler, S.I., 1995. Temperature dependence of Correlation of the equilibrium and kinetics of leaf air
in®nite dilution activity coecients in octanol and octanol- exchange of hydrophobic organic chemicals. Environ. Sci.
water partition coecients of some volatile halogenated Technol. 25, 866±871.
organic compounds. J. Chem. Eng. Data 40, 1196±1198. Pinsuwan, S., Li, A., Yalkowsky, S.H., 1995. Correlation of
Bo, S., Battino, R., Wilhelm, E., 1993. Solubilities of gases in octanol/water solubility ratios and partition coecients. J.
liquids. 19. Solubility of He, Ne, Ar, Kr, Xe, N2 , O2 , CH4 , Chem. Eng. Data 40, 623±626.
CF4 and SF6 in normal 1-alkanols n-Cl H2l‡1 OH Rordorf, B.F., 1985. Thermochim. Acta. 85, 435±438.
(1 < l < 11) at 298.15 K. J. Chem. Eng. Data 38, 611±616. Simonich, S., Hites, R.A., 1996. Organic pollutant accumula-
Boyer, F.L., Bircher, L.J., 1960. J. Phys. Chem. 64, 1330±1331. tion in vegetation. Environ. Sci. Technol. 28, 2905±2914.
Brown, R.H.A., Cape, J.N., Farmer, J.G., 1998. Partitioning of Stephenson, S.M., Malanowsky, S., 1987. Handbook of the
chlorinated solvents between pine needles and air. Chemo- Thermodynamics of Organic Compounds. Elsevier, New
sphere 36, 1799±1810. York.
Cabani, S., Conti, G., Mollica, V., Bernazzani, L., 1991. Free- Tewari, Y.B., Miller, M.M., Wasik, S.P., 1982. Calculation of
energy and enthalpy changes for the process of transfer aqueous solubility of organic compounds. J. Res. Nat. Bur.
from gas and from dilute aqueous solutions of some alkanes Standards 87, 155±158.
and monofunctional saturated organic compounds. J. Thomas, E.R., Newman, B.A., Long, T.C., Wood, D.A.,
Chem. Soc. Faraday Trans. 87, 2433±2442. Eckert, C.A., 1982. Limiting activity coecients of non-
Dallas, A.J., 1995. Ph.D. Thesis, University of Minnesota, polar and polar solutes in both volatile and non-volatile
Minneapolis, MN 55455. solvents by gas chromatography. J. Chem. Eng. Data 27,
Gerrard, W., 1972. J. Appl. Chem. Biotechnol. 22, 623±650. 399±405.
Gruber, G., Langenheim, D., Gmehling, J., Moollan, W., 1997. Tolls, J., MacLachlan, M.S., 1996. Partitioning of semivolatile
Measurement of activity coecients at in®nite dilution organic compounds between air and Lollum multi¯orum
using gas±liquid chromatography. 6. Results for systems (Welsh Ray Grass). Environ. Sci. Technol. 28, 159±166.
exhibiting gas±liquid interface adsorption with 1-octanol. J. Tse, G., Sandler, S.I., 1994. Determination of in®nite dilution
Chem. Eng. Data 42, 882±885. activity coecients and 1-octanol-water partition coe-
Harner, T., Bidleman, T.F., 1996. Measurements of octanol-air cients of volatile organic pollutants. J. Chem. Eng. Data 39,
partition coecients for polychlorinated biphenyls. J. 354±357.
Chem. Eng. Data 41, 895±899. Welke, B., Ettlinger, K., Riederer, M., 1998. Sorption of
Harner, T., Bidleman, T.F., 1998. Measurement of octanol-air volatile organic chemicals in plant surfaces. Environ. Sci.
partition coecients for polycyclic aromatic hydrocarbons Technol. 32, 1099±1104.
and polychlorinated naphthalenes. J. Chem. Eng. Data 43, Wilcock, R.L., Battino, R., Danforth, W.F., Wilhelm, E., 1978.
40±46. Solubilities of gases in liquids II. The solubilities of He, Ne,
Harner, T., Mackay, D., 1995. Measurement of octanol-air Ar, Kr, O2 , N2 , CO, CO2 , CH4 , CF4 and SF6 in n-octane, 1-
partition coecients for chlorobenzenes, PCBs, and DDT. octanol, n-decane and 1-decanol. J. Chem. Thermodyn. 10,
Environ. Sci. Technol. 29, 1599±1606. 817±822.
Keymeulen, R., De Bruyn, G., Van Langenhove, H., 1997. Yalkowsky, S.H., Myrdal, P., Dannenfeiser, R.-M., Simamora,
Headspace gas chromatographic determination of the plant- P., 1994. Chemosphere 28, 1675±1688.
cuticle-air partition coecients for monocyclic aromatic