Fluid Phase Equilibria, 51 (1989) 257-268

Elsevier Science Publishers B.V., Amsterdam -

257
Printed in The Netherlands

CORRELATION OF SOLUBILITIES OF GASES AND HYDROCARBONS IN WATER

J.M. MOLLERUP and W.M. CLARK*

of Chemical Engineering, DTH, DK-2800 Lyngby Denmark

*Chemical Engineering Department, Worcester Polytechnic Institute, Worcester,

USA

ABSTRACT
The density dependent

local composition (DDLC) model for the Helmholtz energy of a

fluid mixture is shown to correlate accurately the solubility of inert and acidic gases and of
hydrocarbons from Cl to Cs in water at pressures to 300 bar. Further, the local composition model correctly reproduces the experimentally observed maximum in the Henrys law
constant curve for gas solubilities in water. The quadratic mixing rules on the other hand,
predict no maxima. The new model has been applied to flash problems related to wet
sour gas mixtures where the model reproduces the experimental data.
The model contains up to three temperature

independent

binary interaction

para-

meters. When the interaction parameter in the DDLC term is zero, this correction vanishes
and the model reduces to one with the conventional classical quadratic mixing rules usually applied to hydrocarbon -hydrocarbon

mixtures.

The DDLC mixing rule can be applied to any equation of state, which can be divided
into a repulsive and an attractive contribution.
tion with

the

Redlich-Kwong

equation

In this work, we have applied it in conjunc-

of state with

a temperature

dependent

a-

local composition concept (DDLC) combines the equation

of

parameter.
Kevwords: Gas solubility, hydrocarbons, water.
INTRODUCTION
The density dependent

state approach and the local composition concept in a model that retains the complex
composition

dependence

of the Wilson activity coefficient

model (Wilson,

extends it to the compressible fluid state where a density dependence

1964) but

is essential for a

correct description of phase behavior at high pressure. The model is thermodynamically

fluid

state. The model

is derived

manner the density dependence

1981 & 1983).

0378-3812/89/$03.50

0 1989 Elsevier Science Publishers B.V.

local composition

of the
concept

258

The model
applied

only modifies

to any equation

contribution.
action

The

model

contains

which

are obtained

parameters

interaction
model

parameter

reduces

the attractive

of state which

in the

to one with

unlike

interactions,

existing

parameter

mixtures.

equation
matrix

are present.

quadratic

mixing
term

without

for hydrocarbon-hydrocarbon

binary
data.

term

rules usually
is only

and

the existing

binary

the

and

the

applied

to

important

when

methanol-

term can be added

changing

inter-

When

vanishes

water-hydrocarbon

The DDLC correction

models

independent

this correction

like the

and can thus be

fit to binary

The DDLC correction

interactions,

of state

temperature

is zero,

function

into a repulsive

by a least squares

DDLC term

polar-nonpolar

hydrocarbon
the

up to three

the conventional

hydrocarbon-hydrocarbon
strong

part of the Helmholtz

to most of
interaction

interactions.

THE MODEL
The
with

DDLC model

mixing

rule where

quadratic
the

for the

the local composition

Helmholtz
concept.

it is assumed

energy

that the mixture

sums of the pure component

van der Waals

The reduced

mixing

residual

F = ArU,V,nYKT

- ng - (f/Tn)

n,

en (1
J

n,

where

unlike

n,

fluid

in the equation

of state are

is a correction

is written

theory

rule is a quadratic

interaction

for the DDLC model

mixing

The DDLC term

nonquadratic

energy
1

the van der Waals

parameters

parameters.

rule where

Helmholtz

combines

The classical van der Waals

terms

term

to

are added.

as

~~~(1 -k,,)

ti,,ln)

fl,

aoj; is the geometric

mean of the pure component

parameters
(2)

E,, = 4-a;,
The

(1) represent

last

term

Helmholtz

g=

Pn(l

f =

(nm?)

where

is the

energy

(1) is the

dependent

the van der Waals

correction

repulsive

contribution

part of the attractive

rule model
The

pure

term.

to the

Helmholtz

The first two

of the Helmholtz

component

terms

energy

equation

for

is thus
(4)

is a Redlich-Kwong

- B/v)

~:n,L

fluid (Redlich-Kwong,

1949) g and fare

Cn(1 +BIV)

V is the total

and

in equation

-g-a fIT
If the fluid

nH=

(ng)

and f is the density

equation

F=

/In

first term

energy

the

k;,

(6)

volume,

J J'

(7)

in

and
the

259

bj; = (bjj + b;J (1 - ejJl2

(8)

fj; = 0, equation

The last term

zero.

n,

is only necessary

If k); is close to zero we may expand

(1) then

Fdt = - C/lTn) 1
If k);

(I),the DDLC term,

in equation

from

part of equation

whenever

equation

k); deviates

(3). The attractive

becomes

n,a;,(l-k,,

- k;, Cl-+a;k;,

(10)

/-/TN

is small,

correction

(7) reducesto

the

effective

interaction

term can be omitted.

A further

parameter

expansion

is kj;+ k);

of the model

wherefore

the

was shown

DDLC

by Mollerup

(1985).
In the case of hydrocarbon-hydrocarbon
action

is kj;,and

parameter

acidic gases or hydrocarbon-nitrogen

are correlated

same accuracy
We

have

interaction
interacts

they

meters.

pure

The binary
the

DDLC

interaction
models,

model

coefficient

is in many

cases equal

in the

quadratic

of

from

inter-

Polar-polar

the

two

nonpolar

binary

molecule

k,, becomes

important,

interaction

para-

in Table

of state (Soave,

the vapor

the

1983 & 1985).

systems are shown

equation

is calculated

are fitted

using

the use of binary

for nonpolar-water

for water

(Mollerup,

a small

introduction

the Redlich-Kwong-Soave

parameters

parameter

in cases where

the

binary

For hydrocarbon-

kj;and (ii, with

parameters,
models
mixtures

1 summarizes

signs. Table

a-parameter

interaction

interaction

molecule,

parameters

we have applied

component

binary

of zero.

kj; is necessary.

a non-zero

methanol-hydrocarbon

polar

have opposite

the only non-zero

activity

k); and t?,;. However,

a strong

Binary interaction

this work

two

correlated

parameters
with

because

with

be assigned

interactions

as that of the common

previously

interactions

it may often

1972).

pressure

2. In
The

of water.

to binary data only. The sum of kj;and k); in

to kj; in the vdW

mixture

model.

term

That

is about

is, the

equal

effective

in the

two

see eq. (IO).

RESULTS
The model
various
tion

monoxide,

model

hydrogen

of the

Figures

correlation

is not within
mixing

Mollerup

the solubilities
and

carbon

(1985)

of the

the solubility

has demonstrated

of hydrogen,

dioxide

of the solubilities

mixture

in water.

methane,
In this

hydrocarbons

the applica-

nitrogen,

paper

of

we

carbon
show

Cl to Cs in water

the
and

calculations.

1 to 4 show the comparison

The DDLC model

der Waals

sulfide

(1) to (9) has been used to correlate

in water.

to correlate

some multicomponent

tion

by equations

gases and hydrocarbons

of the

results

described

does represent
the experimental

of experimental

the maximum
accuracy.

and correlated

in Henrys constant

Henrys

although

The results of the correlation

rules, i.e. k); = 0, are shown by the dash line.

constants.

the correlawith

the van

260

Figure

5 shows

propane-water.
propane-water
Table

of experimental

3 and 4 show

and butane-water,

5 shows

concentrations
data

a comparison

Tables

the

and flash

and carbon

calculations

The predicted

cially for methane.

correlations

ratios (L/F). This was not expected

predict

liquid

densities

very accurately.

to the model

three

ratios

phase

system

at

nearly

of

lines of

obtained

for the binary

does not improve

The calculated

binary

dioxide-methane

the DDLC correction

dew points,

systems.
term,

data

espe-

are retained

the calculated

as the Redlich-Kwong-Soave

since they are water

equimolar

correlated

by adding

systems. The DDLC model

to feed

insensitive

equilibrium

Tables 6 to 9 show

sulfide-carbon

ratios are improved

The improved

for the multicomponent

monoxide

monoxide.

in water-hydrogen

equilibrium

correlated

and correlated

respectively.

water-hydrogen-carbon

of hydrogen

and

experimental

equation
shown

dew points and the water

liquid

does not

in Table

10, are

rich phase is main-

ly water.
The binary

interaction

parameters

are independent

of temperature

and obtained

by a

least squares fit to binary data only.

DISCUSSION

AND CONCLUSIONS

The DDLC model

Its theoretical

cient models
that

cannot

from which

it covers the whole

region

as the activity

When

meter

Its advantage

and is not limited

matrix

sumption
water

region

coefficient

except

parameter
model

and methanol,

theory

of the liquid

of the activity

over the activity

coefficient

to the high density

k); in the DDLC correction

the correction

changing

for systems containing

of the DDLC correction

if all k)i are non-zero.

as that

coeffi-

models

is

low pressure

models are.

of state models without

The addition

from a fundamental

and is as sound

it is derived.

the interaction

equation

derived

fluid

reduces to the van der Waals

be rigorously

basis is semi-empirical

water

term

If, however,

one can take advantage

term

vanishes,

can be added

the existing

binary

the model

to most of the

interaction

para-

and alcohols.

more than

doubles

the computer

time

only two of the species are strongly

the last term

polar,

consay

in equation

(1) as

pl,Pn(
J

and thus
being

n, /II + (naEul + rp&)/rl)

-4B

reduce

extended

computer

time

to multicomponent

consumption
mixtures

substantially.
containing

The correlation

both water

is presently

and methanol.

ACKNOWLEDGEMENT
WMC
Doctoral

gratefully
Fellowship.

acknowledges

the

National

Science

Foundation

for

a NATO

Post

261

TABLE 1: Application of binary interaction parameters in the

kij

kii

(+I
+

HC-CO2,HzS,N2

(+I

polar-polar
HC- alcohols
HC -water

K:

eij

Hydrocarbon

TABLE 2: Binary interaction parameters with water

Component
Methane

k12

klz

p12

1.040

-0.413

0.198

0.743
0.878

Ethane

0.364
-0.246

0.670
0.980

Propane

0.330
-0.171

0.811
1.195

Butane

1.245

-0.138

1.504

Benzene

0.849

-0.124

0.754

-0.091

1.013

-0.091

0.611

1.000

1.700

1.250

0.990
Hydrogen

2.345

0.845
-0.141

2.308
Nitrogen

1.130
0.839
0.392

Carbon

0.545

0.103
0.142

-0.178

0.095

-0.243

0.143

0.097

0.211
Dioxide

0.270

0.169

-0.057

-0.303
HydrogenSulphide

0.159
0.506

-0.636

-0.238
CarbonMonoxide

0.682
0.839

1.387
Hexane

0.477
0.561

1.196
Pentane

0.260

0.149
0.212

50 -

am
0

;5

*o-

.q
P
I
i

t;
5

IO-

P
\
2-

Propane

.
1-

150

250

200

300

350

Fig.

1.

100

450

500

ST.2.

300

350

400

250

300

Temperature

Calculated
COnStanfS

250

Temperature

and smoothed
experrmental
Henry,
of ethane I water at saftat,a
prer. Croveto
(1977);
DLXC;
- dW.

350

400

450

F,g. 2. Calculated i"d expermental

t
:
\
=
\
n
\

Pentane

300

350

400

450

500

550

300

350

PO0
TemperIture

Fig. 4.

450

500

550

Calculated and mmothed

263

VLE
370K
311K
LLE
311K
370K

428~

-1

37OK
_(I.1

428K

rl

-cj.01 ':
w"

311K <
","
311K

"

13 370K

-0 .OOl

00

311K

-0 LOO03
2

10

20

50

100

200

500

Pressure, bar
Fig. 5.

Calculated
system. l

experimental
DDLC;

equilibrium

VLE,

LLE,

ratios

in the propane-water

@ three

VLE,a LLE, 370 K; A 428 K; (Kobayashi

1951).

phase

line, 311 K;

264

TABLE 3: Propane - Water Three Phase Line

Vapour phase

Propane rich liquid phase

AP

bar

bar

IQ.103

KI

exp

talc

2730
3790

2640
4050

4570
4920

5410
6170

talc

K4.102

K3

ev

talc

talc

0.105
0.187
0.344
0.615

0.152
0.250
0.421
0.691

2730
3780
4550
4900

2640
4050
5390
6150

0.140
0.218

0.144
0.208

0.335
0.465

0.300
0.416

6320

1.09

1.11

4990

6290

0.625

0.555

5970
5340
4980

1.84
2.91
3.58
4.46
5.58
7.23

1.75
2.73
3.41

4910
4670
4440
4190
3950
3740

5940
5313
4950
4570
4200
3840

0.813
1.00
1.10
1.19

0.713
0.878
0.958
1.03
1.08
1.09

279
289
300
311
322

5.67
7.48
10.1
13.2
17.0

0.009
0.03
0.01
0.04
0.1

5020

333
344
350

21.4
26.8
29.8

0.2
0.3
0.4

4940
4700
4470

355
361
367

33.3
37.1
41.5

0.2
-0.0
-0.6

4220
3980
3760

4600
4220
3850

4.29
5.45
7.16

1.28
1.24

A = calculated -experimental
The equilibrium

ratios are calculated with the water rich liquid phase as the reference

phase. Kobayashi and Katz (1953)

TABLE 4: Butane - Water Three Phase Line

Butane rich liquid phase

AP

bar

bar

w
311

3.62a
3.60b

0.07

16100

0.09

19600

8.65

0.15

11500

8.64

0.16

16100

378

17.9

0.2

7080

17.9

0.2

9710

411

33.8

-0.2

4420

33.9

-0.3

5309

344

Vapour phase

K2-103

KI

K3

K4.100

talc

talc

exp

talc

exp

talc

26800

0.500

0.510

15900

26400

1.67

1.59

1.86

11100

20600

3.38

0.620
21200

2.10
2.86

11100

8.50

15600
5.81

6760

17.7

4220

9.60
5180

26.8
28.4

1.51

19300

10700

5.41

980

7.08

9260
5070

3.07

3.21
4.70

5.55
5.53

7.18

A = calculated -experimental
The equilibrium

ratios are calculated with the water rich liquid phase as the reference

phase.
a) Reamer, Sage, & Lacey (1952)
b) Wehe, & McKetta (1961)

265
TABLE

5: Experimentalandcalculatedequilibriumratios*

Monoxide- WaterSystem
Pbar

Hydrogen

oftheHydrogen-Carbon

CarbonMonoxide

Water

TK

311

366

422
478
589

exp.

talc.

exp.

3.45
13.8
31.0
65.5
138.
3.45
13.8
31.0
65.5
138.
31.0
31.0
65.5
103.
138.

2.95
13.6
29.7
60.8
135.
3.4
13.5
30.4
63.3
134.
31.8
31.4
66.8
102.
136.

26200
5410
2420
1110
537
21100
6190
2440
1160
536
1650
980
474
312
36

talc.

exp.

23900
5190
2400
1180
545
24500
6200
2760
1330
635
1630
825
401
267
33

24300
5180
2340
1220
617
31900
6110
3070
1560
705
2180
1350
666
439
61

talc.
22000
4830
2270
1150
559
28600
7270
3270
1600
789
1990
992
486
326
37

exp.
0.0223
0.0050
0.0024
0.0013
0.0007
0.236
0.061
0.028
0.014
0.0077
0.154
0.571
0.284
0.194
0.843

IndependentvariablesT,y.Thevapourphaseconsistsofnearlyequimolar
concentrationofhydrogenandcarbon
monoxide.
Gillespieand Wilson (1980)

TABLE

T
K

6: Equilibrium ratios in the Carbon Dioxide

P
bar

289

302

348

394

477

Gillespieand

51
101
202
55
101
202
25
51
101
202
7
25
51
101
202
25
51
101
202

Carbon
exp

Dioxide

DDLC

39
41
38
49
42
633:
181
98
61
47
771
252
130
73
2::
125
63
41

Wilson (1982)

41
39
36
52
t:
654
194
109
F
883
253
135
77
49
204
106
56
31

- Water System
Water

vdW
s;
65
76
63
505:
148
82
3";
517
148
79
45
28
119
!f
18

exp

DDLC

0.00094
0.0028
0.0030
0.0016
0.0040
0.0045
0.060
0.018
0.011
0.0074
0.0096
0.311
0.093
0.052
0.033
0.028
0.677
0.378
0.227
0.150

0.00061
0.0029
0.0038
0.0013
0.0034
0.0048
0.058
0.018
0.010
0.0078
0.0107
0.306
0.090
0.051
0.032
0.028
0.699
0.381
0.223
0.153

vdW
0.00064
0.0026
0.003 1
0.0013
0.0033
0.0045
0.058
0.018
0.011
0.0082
0.0116
0.306
0.091
0.052
0.034
0.031
0.699
0.383
0.227
0.160

talc.
0.0223
0.0050
0.0024
0.0013
0.0007
0.236
0.061
0.028
0.014
0.0077
0.154
0.571
0.285
0.194
0.844

266
TABLE 7: Equilibrium ratios in the Hydrogen Sulphide - Water System

bar

311

366

422

477

4:
103
207
8
41
74
103
206
31
137
206
31
103
137
206

HydrogenSulphide

exp

DDLC

Water

vdW

93

87

29

$2

124
41

28
27
185
39

:A
165

zt
146

E
::

32
21
18
19

55

42

f:
20
z!

16

;:
20

::

81.;

8Y

16

12

9.7
39
9l.s
6.6

exp
0.0091
0.019
0.019
0.019
0.104
0.025
0.022
0.049
0.057
0.177
0.084
0.127
0.611
0.248
0.220
0.237

DDLC

vdW

0.0085

0.0085
0.017
0.018
0.028
0.101
0.026
0.020
0.043
0.048
0.170
0.081
0.111
0.586
0.240
0.213
0.216

o.ozo
0.022
0.024
0.101
0.025
0.020
0.044
0.054
0.170
0.080
0.107
0.586
0.238
0.211
0.210

GillespieandWilson(1982)

TABLE 8: Equilibrium ratios in the Methane

bar

323
A:
138
348
::
138
423

478
533
589

z
138
62
103
138
103
138
138
169

- Water System

Methane
exp
4460
1060
564
5160
1180
608
3730
915
458
553
343
265
173
134
508

GillespieandWilson(1982)

DDLC

4930
1240
652
6210
1510
767
3870
916
446
456
286
222
122
96
31
27

Water
vdW
8870
2160
1100
5730
1380
681
1960
460
222
236
147
114
73
57

22
19

exp
0.0096
0.0024
0.0014
0.029
0.0071
0.0037
0.355
0.085
0.043
0.304
0.201
0.162
0.511
0.404
0.835
0.751

DDLC

vdw

0.0092
0.0023
0.0013
0.029
0.071
0.0038
0.359
0.086
0.043
0.305
0.194
0.152
0.521
0.413
0.849
0.746

0.0091
0.0022
0.011
0.029
0.067
0.0034
0.358
0.084
0.041
0.303
0.191
0.148
0.519
0.409
0.849
0.745

0.096
0.018
0.086

0.0096

118 450+
172
173

311*

380*

450*

130

165
76

123
169
110
182

0.027
0.13
0.035
0.12

0.028
0.013

0.012

0.11
0.017
0.090

0.0017
0.023
0.0021
0.018

0.0019
0.0015

DDLC

Water

0.027
0.034
0.13
0.12

0.028
0.010

0.097

0.017
0.11
0.09

0.0016
0.023
0.0019
0.018

0.0019
0.0015

vdW

z:
66
31

89

Ii

%
48

40
85
39
62

64
49

exp

:i
z1

iz

28

:;

:;
86
:i

f;

::
23
30
18

z:

32
t:

Zf

iii

82
62

DDLC vdW

Carbon Dioxide
exp

21
::
14

5:

28

s:
18

:T
16
5:

20

20
19
17
12

::

31

16
18
14

16
:i
18

21
17

16
d.33
15

ii

37

:;
IO

21
21
:;

tz

319
312
199
130

763
101

104

377
368
251

381
767
305
507

1100
650

512
292
307
152

990
230

249

340
530
234

678
1030
516
687

1560
1080

DDLC

vdW

exp

DDLC

Methane

Hydrogen Sulphide

Equilibrium Ratios

+ 50% Wet Sour Gas Mixture No. 1, Mole per cent H20: 50, HIS: 5, CO2: 30, CH4: 15
*
50% Wet Sour Gas Mixture No. 2, Mole per cent H20: 50, HzS: 40, COz: 5, CH4: 5
Robinson, D.B., Huang, S.S., Leu,A.-D. and Ng, H.-J. (1982)
C02-H2S kij = 0.099; CO2-CH4 kii = 0.080; H2S-CH4 kij = 0.093

0.027
0.031
0.097
0.12

0.026
0.0094

0.0019
0.0020
0.023
0.020

48
76

125
169
84 380+
129

exp

0.0019
0.0017

TK

311+

Ibar

TABLE 9: Flash Calculations, Wet Sour Gas Mixtures

306
182
156
83

421
594

464

310
170
118

1320
985
610
403

3410
2220

vdW

:i
6.4
12

6.4

IfI

:.zs
8.7

15
5:04
8.7

4.1
7.3

exp

16
;.:
15

7.6

zz

7.1
11

:;
7.0
::

5.2
9.2

DDLC

::
78
15

73.36

33

:.:
11

:r:
7.0
11

5.2
9.2

vdVu

Volume % (L/F)

268

TABLE 10: Dew Points

50% Wet Sour Gas Mixture No.1

P bar

394
408
422
436
443
450
464
472

4.2
6.4
9.5
13.9
16.7
19.4
27.0
32.5

50% Wet Sour Gas Mixture No.2

&pLC

AP
vdW

4.2

-0.05

-0.04

422

9.6

-0.17

-0.16

450
464
478

19.5
27.0
37.0

0.008
0.11
-0.008

0.06
0.21
0.18

De&

*
vdW

TK

P bar

-0.007
0.01
0.003
0.015
-0.22
0.04
0.01
-0.1

0.007
0.01
0.003
0.015
-0.22
0.04
0.008
0.1

394

A = calculated -experimental
REFERENCES
Le Breton, J.G. and J.J. McKetta, 1964. Low pressure solubility of n-butane in water.
Hydrocarbon Processing, 43: 136-138.
Croveto, R., R. Fernandez-Prini and M.L. Japas, 1984. The solubility of ethane in water
up to 473 K. Berichte Bunsenges., Phys. Chem., 88: 484-488.
Gillespie, P.C., G.M. Wilson, 1980. Vapor-liquid equilibrium data on water-substitute
gas components. Gas Processors Association Research Report 41.
Gillespie, P.C., G.M. Wilson, 1982. Vapor-liquid and liquid-liquid equilibria.Watermethane, water-carbon dioxide, water-hydrogen sulphide, water-pentane.
Gas Processors Association Research Report 48.
Kobayashi, R. and D.L. Katz, 1953. Vapor-liquid equilibria for binary hydrocarbonwater systems. Ind. Eng. Chem., 45: 440-451.
Mollerup, J., 1981. A note on excess Gibbs energy models and equations of state.
Fluid Phase Equilibria,7: 121-138.
Mollerup, J., 1983. Correlation of thermodynamic properties of mixtures using a
random-mixture reference state. Ibid, 15: 189-207.
Mollerup, J., 1985. Correlation of gas solubilities in water and methanol at high
pressure. Ibid, 22: 139-154.
Reamer, H.H., R.H. Olds, B.H. Sage and W.N. Lacey, 1944. Compositions of the coexisting phases of n-butane-water system in three phase region. Ind. Eng. Chem, 36:
381-383.
Reamer, H.H., B.H. Sage and W.N. Lacey, 1952. n-butane-watersystem
in the two
phase region. Ind. Eng. Chem., 44: 609-615.
Redlich, 0.. J.N.S. Kwong, 1949. On the thermodynamics of solutions. Chemical
Reviews, 44: 223-244.
Robinson, D.B., 5.5. Haung, A.D. Leu and H.-J. Ng, 1982. The phase behavior of two
mixtures of methane, carbon dioxide hydrogen sulphide, and water.
Gas Processors AssociationResearch Report 57.
Soave, G., 1972. Equilibrium constantsfrom a modified Redlich-Kwong equation of
state. Chem. En . Sci., 27: 1197-1203.
Tsonopoulos, C. an 2 G.M. Wilson, 1983. High temperature mutual solubilities of hydrocarbonsand water I. Benzene, cyclohexane, and n-hexane. AlChE Journal, 29: 990999.
Wehe, A.H. and J.J. McKetta, 1961. n-butane-1-butene-water
in the three-phase region
J. Chem. Eng. Data, 6: 167-172.
Wilhelm, E., R. Battino and R.J. Wilcock, 1977. Low pressure solubility of gases in liquid
water. Chemical Reviews, 77: 219-262.
Wilson, G.M., 1964. A new expression forthe excess free energy of mixing. J. Chem.
Sot., 86: 127-I 30.