Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

10.1351 Pac198658111535

Download as pdf or txt
Download as pdf or txt
You are on page 1of 6

Pure & App!. Chem., Vol. 58, No. 11, pp. 1535—1540, 1986.

Printed in Great Britain.


© 1986 IUPAC

INTERNATIONAL UNION OF PURE


AND APPLIED CHEMISTRY
ANALYTICAL CHEMISTRY DIVISION
COMMISSION ON ELECTROANALYTICAL CHEMISTRY*

RECOMMENDED METHODS FOR


THE PURIFICATION OF SOLVENTS
AND TESTS FOR IMPURITIES
ACETONE
Prepared for publication by
J. F. COETZEE and T.-H. CHANG
Department of Chemistry,
University of Pittsburgh, PA 15260, USA

*Membership of the Commission during 1983—85 when the report was prepared was as follows:
Chairman: J. Jordan (USA); Secretary: K. Izutsu (Japan); Titular Members: A. K. Covington
(UK); J. Juillard (France); R. C. Kapoor (India); E. Pungor (Hungary); Associate Members:
W. Davison (UK); R. A. Durst (USA); M. Gross (France); K. M. Kadish (USA); R. Kalvoda
(Czechoslovakia); H. Kao (China); Y. Marcus (Israel); T. Mussini (Italy); H. W. NUrnberg
(FRG); M. Senda (Japan); N. Tanaka (Japan); K. Tóth (Hungary); National Representatives:
D. D. Perrin (Australia); B. Gilbert (Belgium); W. C. Purdy (Canada); A. A. Vlek (Czecho-
slovakia); H. Monien (FRG); M. L'Her (France); Gy. Farsang (Hungary); H. .Gaur (India);
W. F. Smyth (Ireland); E. Grushka (Israel); S. R. Cavallari (Italy); W. Frankvoort (Netherlands);
Z. Galus (Poland); G. Johansson (Sweden); J. Buffle (Switzerland); H. Thompson (UK); J. G.
Osteryoung (USA); I. Piljac (Yugoslavia).

Republication of this report is permitted without the need for formal JUPAC permission on condition that an
acknowledgement, with full reference together with JUPAC copyright symbol (© 1986 JUPAC), is printed.
Publication of a translation into another language is subject to the additional condition of prior approval from the
relevant JUPAC National Adhering Organization.
Recommended methods for the purification of
solvents and tests for impurities: acetone

Values are given for the most relevant physicochemical properties and solva—
tion parameters of acetone. The most significant solvent properties of acetone,
particularly the limitations imposed by aldol condensation, are discussed.
Typical examples of electrochemical studies carried out in this solvent are
listed. Methods are recommended for the purification of commercial acetone and
for the determination of residual water and other impurities.

INTRODUCTION
Acetone can be classified as a dipolar aprotic solvent although its protons are more reac—
tive than those of typical members of this class of solvents. (This is also the case for
nitromethame and, to some extent, for the formyl proton of dimethylformanide.) This pro—
ton reactivity constitutes the main limitation of acetone as a solvent. Other limitations
are its exceptionally high hygroscopicity, high volatility and formation of explosive mix—
tures with air. On the credit side, acetone is available commercially in a sufficiently
pure state for many (but not all) applications, it has a wide liquid range, a usefully low
viscosity, and a relative permittivity (dielectric constant) that is sufficiently high to
allow simple electrochemical measurements. Finally, its toxicity is low.

Physicochemical properties and solvation parameters of acetone are summarized in Table 1.


The electron donor strength of acetone (as measured by its donor number of 17.0 with
respect to antimony pentachloride as reference acid) is somewhat higher than those of the
weakly basic dipolar aprotic solvents acetonitrile (14.1) and propylene carbonate (15.1)
but well below those of the more basic solvents of this class, e.g., dimethylformamide
(26.6) and dimethyl sulf oxide (29.8). As a result of its moderate electron donor
strength, acetone is a fair solvent for inorganic substances, although its relative per—
mittivity is sufficiently low that considerable ion association occurs. Acetone is an
excellent solvent for many classes of organic compounds. Its exceptional solvent power
for acetylene is utilized in the storage of acetylene. Acetone is also used commercially
as a solvent for nitrocellulose (in the manufacture of explosives), cellulose acetate and
many synthetic resins such as methacrylates. The principal commercial use of acetone,
however, is in the manufacture of numerous chemicals, particularly such derivative
solvents as methylisobutyl ketone (1).

Acetone has received less attention than many other dipolar aprotic solvents have as a
medium for electrochemical and other solution chemistry studies. The most extensive
investigations have been concerned with the conductance of electrolytes, although some
voltammetric and potentiometric studies have also been carried out. The main results of
these investigations are summarized below.

KEY SOLVENT PROPERTIES OF ACETONE


Proton acceptor and proton donor strength
Acetone is half
protonated in 81.5% (by wt.) sulfuric acid (aq.) corresponding to a pKa —
value (aq.) of its conjugate acid of —7.2 (ref. 4);
cf. acetonitrile, —9.5. The protons
of acetone have considerable acidity, pKa 19 (ref. 5), owing to the electron withdrawing
properties of the carbonyl group and also the ease of charge delocalization from carbon to
oxygen in the resulting carbanion.

Solvating ability
Free energies of transfer of (mainly) alkali metal and halide ions have been estimated by
means of a number of extrathermodynamic procedures. The probable validity of different
procedures has been discussed by Marcus (ref. 6). As compared to the cyclic carbonyl com-
pound, propylene carbonate, acetone appears to be a rather good solvator of the smaller
alkali metal ions, in spite of its much lower relative permittivity which should weaken
the electrostatic (Born) component of solvation. On the other hand, it appears to be a
significantly weaker solvator of halide ions than propylene carbonate is.

1536
Purification of solvents and tests for impurities: acetone 1537

TABLE I. Selected properties and solvation parameters of acetone

Freezing temperature, T/K —273.15 (tm/°C) — 94.7

Boiling temperature at 101.325 kPa (1 atm), 56.29


T/K —273.15 (tb/°C)

Flash point, tf 11°C —18

Vapor pressure, p/kPa at 20°C 24.227

Density, p/kg dm3 0.78440

Dynamic viscosity, /mPa 0.3040

Surface tension, y/mN m', at 20°C 23.32

Enthalpy of vaporization, Hv/kJ mol, at tb 29.09

Molar heat capacity, C/J 1 mol', at 30°C 129.2

Refractive index, nD 1.35596

Dipole moment, p/D, in hexane at 20°C 2.69

Relative permittivity (dielectric constant), D = c/c0 20.70

Donor number (Gutmann), DN/kcal mol' 17.0

Acceptor number (Mayer—Gutmann), AN/dimensionless 12.5

Solvatochromic parameters (Kamlet—Taft)


—— polarity parameter, ,r' 0.72

—— hydrogen bond acceptor parameter, B 0.48

—— hydrogen bond donor number, a (0.07).

Values of properties are from Ref. 1 and are for 25°C, except
where otherwise noted. Donor and acceptor numbers are from
Ref. 2 and solvatochromic parameters are from Ref. 3; values
are for 25°C.

.. 1 cal = 4.184 J
Somewhat uncertain; see Ref. 3.

Enolization and aldol condensation


Acetone exists as a keto—enol tautomeric equilibrium mixture containing ca. 10% of the
enol form, as determined by bromometric titration (ref. 7). It is to be noted that the
enol concentration is too low to be observed in spectrophotometric or relaxation experi-
ments, which are applicable to such other carbonyl compounds as 8—diketones in which the
enol content is much higher. The principal limitation of acetone as a solvent is the ease
with which it undergoes aldol condensation, which is both acid and base catalyzed. These
reactions have been extensively studied (ref. 4, 5, 7). Base catalysis occurs as follows.

H3CCCH3 + 0H ± H3C—C—CH2 carbanion formation


0 0

CH3 CH3
H3C—C + H2C—C—CH3 H3C—C---CH2—C—CH3 direct carbonyl addition
0 0— 0

CH3 CR3
H3C—C—CH2—C—CH3 + H20 ± H3C—C—CH2—C—CH3 + 0H aldol formation
0 ö OH
1538 COMMISSION ON ELECTROANALYTICAL CHEMISTRY

The product, 4—hydroxy—4—methyl—2—pentanone, is easily dehydrated (e.g., by Lewis acids),


forming the very stable 4—ntethyl—3—pentene—2—one (niesityl oxide):

CH3 cH3
H3C—C—CH2--C—CH3 —* H3C—C=CH--C—CH3 + H20
OH 0

This series of reactions is of extreme importance in, among other things, the purification
of acetone. Many commonly used desiccants are sufficiently basic (or acidic) to cause
aldol condensation. For example, Riddick (ref. 1) found an increase in the water content
of acetone after treatment with alumina, which acts as a mild desiccant in other solvents.
In fact, deliberate condensation of liquid ketones can be accomplished by adsorption on
basic alumina at room temperature (ref. 8).

Acid catalysis of aldol condensation probably occurs via the enol tautomer (1) to form an
unstable —hydroxy ketone (2) which then rapidly undergoes acid—catalyzed dehydration to
the very stable mesityl oxide (3).

H3C-C-CH3 + H H3C-C—CH3 H3C—C = CH2 +


0 OH OH
+
(1)

CH3
H3C—C
OH
+ H2C = C—CH3
(H
—-H - 113
H3C—C—CH2—C—CH3
OH
+
(2)

CH3
—H2O H3C— =
CH——CH3

(3)

Further condensation may also occur. For example, in presence of hydrogen chloride not
only mesityl oxide, but also phorone, (CH3)2C=CH—C—CH=C(CH3)2, is formed, while in
presence of sulfuric acid 1,3,5—trimethylbenzene,°mesitylene, is produced. This instabi-
lity of solutions of especially the stronger acids in acetone is obviously a serious limi-
tation of this solvent.

TYPICAL ELECTROCHEMICAL STUDIES IN ACETONE


Acetone was one of the many solvents included in the early, extensive conductometric stu-
dies of salts by Walden, which also included methylethyl ketone (ref. 9), and by Kraus
(ref. 10). It was found that many alkali metal and tetraalkylammonium salts have ion pair
formation constants in the range 102 — iO
L mol'. The polarography of the lanthanide
metal ions, sodium, potassium and rubidium ions, and other inorganic substances has been
studied by Coetzee and Siao (ref. 11), while Gutmann et al. (ref. 12) investigated all
alkali metal ions and thallium(I) ion. The half—wave potentials were compared with simi-
lar data for a variety of other solvents by Coetzee et al. (ref. 13) and by Gritzner (ref.
14). Overall, acetone behaves as a differentiating solvent as compared to water. The
electroreduction of such acidic solutes as the lanthanide metal ions is complicated by
hydrogen evolution, either from residual water or from acetone itself.

Many potentiometric acid—base titrations have been carried out in acetone, which behaves
as a typical differentiating solvent in such applications; it should be reiterated,
however, that solutions of many acids and bases in acetone are not ideally stable owing to
aldol condensation (v.s.). Titration and other results obtained in acetone have been sum-
marized by Charlot and Trmillon (ref. 15).

MANUFACTURE OF AND IMPURITIES IN ACETONE


Until World War I, acetone was manufactured by dry distillation of calcium acetate,
obtained by neutralizing pyroligneous acid (from wood distillation) with lime. The
greatly increased demand for acetone as a solvent for nitrocellulose—based explosives led
to the development of the Weizmann process involving fermentation of starch (later:
molasses) by means of the bacillus clostridium acetobutylicum. The Weizmanm process, in
Purification of solvents and tests for impurities: acetone 1539

turn, was displaced by catalytic dehydrogenation of 2—propanol (using copper or various


netal oxides as catalysts) and, to some extent, by several other synthetic processes.
Commercial acetone is now relatively pure, with water as the main impurity. The ACS spe—
cifications for reagent grade acetone (ref. 16) include the following: density < 0.7857
kg dnf3 at 25°C; water 0.5% (by wt.?); methanol and 2—propanol 0.05%; aldehydes (as
HCHO) 0.002%; titratable acid (1:1 by vol. acetone—water mixture titrated with NaOH to
phenolphthalein end point) 3 x i0 mol 0H kg1 acetone; titratable base (ca. 1:1 by
vol. acetone—water mixture titrated with HC1 to methyl red end point) 6 x i0 mol H+
kg1 acetone.

PURIFICATION OF ACETONE

The relatively high purity of commercial acetone is an obvious advantage. On the other
hand, the ease with which acetone undergoes either acid— or base—catalyzed aldol conden—
sation complicates any additional purification. Removal of the main impurity, water, pre—
sents a problem because even such mildly acidic or basic desiccants as silica gel or
alumina cause aldol condensation (v.s.). Riddick (ref. 1) found that drying with
anhydrous calcium sulfate (Drierite) followed by careful fractional distillation is most
effective, although it is difficult to lower the water content of this very hygroscopic
solvent below 1O M. Coetzee and Siao (ref. 11, 17) used a related method which is
described below.

Dry the acetone by shaking with Drierite (25 g L1) for several hours.
Decant and distill from fresh Drierite (10 g L) through an efficient
column, e.g., a Stedman column. Maintain contact with the atmosphere
through a Drierite drying tube. (Caution: Do not use anhydrous magne-
sium perchlorate which may explode in contact with acetone vapor.)
Store the distillate in a dark container fitted with a siphon for
dispensing the solvent.

TESTS FOR RESIDUAL IMPURITIES IN ACETONE


Conventional Karl Fischer titration of water in acetone (and other active carbonyl com-
pounds) produces a drifting end point and serious overtitration owing to (slow) generation
of additional water as a result of ketal formation:

,OCH3
(CH3)2C=O + 2CH3OH * (CH3)2C + H20
OCH3

The problem is aggravated by the relatively slow reaction of Karl Fischer reagent with
water. Better results are obtained with recently introduced modifications of Karl Fischer
reagent in which pyridine is replaced by sodium acetate (ref. 18) or ethanolamine (ref.
19) and which react faster with water. At least one preparation is available commercially
as J. T. Baker Chemical Company ReAquant reagent. An automated stopped—flow determination
of water in acetone utilizing this reagent has been described (ref. 20). Further improve-
ment has been claimed when methanol is replaced by a chloroform—methanol mixture as titra-
tion medium (ref. 21). The following procedure was recommended for acetone containing
0.2% water; appropriate adjustments in volumes may be necessary for lower water con-
centrations.

Pretitrate 4—5 mL of 2:1 (by vol.) chloroform—methanol with Karl


Fischer reagent containing ethanolamine, using preferably the
"dead—stop" method for equivalence point detection. Inject 100 L of
acetone and record volume of reagent used as a function of time for a
sufficient length of time (up to a few minutes) to allow satisfactory
extrapolation of the (relatively flat) line to zero time.

Various other methods for the determination of water in acetone have been proposed. These
methods are reviewed in ref. 17. The infrared absorption band of water at 2.70 1jn can be
used for water concentrations greater than 0.01%. Water can also be determined by adding
excess acetyl chloride and titrating the acetic acid liberated, or by adding excess
calcium hydride and measuring the volume of hydrogen liberated as a function of time,
extrapolating to zero time to allow for the reaction of calcium hydride with acetone
itself (ref. 11, 17). It appears, however, that in general the modified Karl Fischer
titration described above is to be preferred.
1540 COMMISSION ON ELECTROANALYTICAL CHEMISTRY

DETERMINATION OF OTHER IMPURITIES


The density and refractive index, as well as the differential boiling—condensing tem-
perature (ref. 1), are useful criteria for contamination in general.

Methanol and 2—propanol can be determined by gas chromatography or by their reducing


action on permanganate ion, and aldehydes can be determined by the Schiff—Villavecchia
reaction (ref. 1).

Electroactive impurities can be determined voltammetrically. The polarographic window of


acetone at the dropping nrcury electrode with tetraethylammonium perchlorate as sup-
porting electrolyte extends from +0.58 to —2.46 V vs. SCE(aq.); the cathodic limit
appears to be determined by reduction of acetone itself (ref. 17).

ULTRAVIOLET CUTOFF
For a typical commercial reagent grade of acetone (intended for spectrophotometry), maxi-
mum absorbances (in a 1—cm optical cell) are specified as 1.00 at 330 nm and 0.10 at 340
nm.

REFERENCES
1. J. A. Riddick and W. B. Bunger, "Organic Solvents," Wiley—Interscience, New
York, N.Y., 3rd edition, 1970, pp. 242—243, 721—727.
2. V. Gutmann, "The Donor—Acceptor Approach to Molecular Interactions," Plenum
Press, New York, 1978.
3. J. L. M. Abboud, M. J. Kanlet, and R. W. Taft, in H. W. Taft, Ed.,
"Progress in Physical Organic Chemistry," Interscience—Wiley, New York,
N.Y., Vol. 13, 1981, pp. 485—630.
4. S. Patai, Ed., "The Chemistry of the Carbonyl Group," Interscience, New
York, N.Y., 1966.
5. C. D. Gutsche, "The Chemistry of Carbonyl Compounds," Prentice—Hall,
Englewood Cliffs, N.J., 1967.
6 • Y• Marcus, Pure Chem. 55, 977 (1983).
7. J. Zabicky, Ed., "The Chemistry of the Carbonyl Group," Vol. 2,
Interscience, New York, N.Y., 1970.
8. J. Muzart, ynthesis (!), 60 (1982).
9. P. Walden and E. J. Birr, Z. physik. Chen. 153A, 1 (1931).
10. M. J. McDowell and C. A. Kraus, J. Am. Chem. Soc. 73, 3293 (1953).
11. J. F. Coetzee and Wei—San Siao, Inorg. Chem. 2, 14 (1963).
12. 0. Duschek, V. Gutmann and P. Rechberger, Mh. Chem. 105, 62 (1974).
13. J. F. Coetzee, D. K. McGuire and J. L. Hedrick, J. Chem. 67, 1814
(1963).
14. G. Gritzner, Inorg. Chin. Acta 24, 5 (1977).
15. G. Charlot and B. Trenillon, Les Reactions Chimigues Dans Les Solvants Et
LesSels Fondus," Gauthier—Villars, Paris, 1963, pp. 386—396.
16. American Chemical Society, "Reagent Chemicals," American Chemical Society,
Washington, D.C., 6th edition, 1981, pp. 68—70.
17. Wei—San Siao, M. S. Thesis, University of Pittsburgh, 1961.
18. J. C. Verhoef and E. Barendrecht, Anal. Chin. Acta94, 395 (1977).
19. E. Scholz, Fresenius Z. Anal. Chen. 309, 30 (1981).
20. M. A. Koupparis and H. V. Malmstadt, Anal. Chen. 54, 1914 (1982).
21. A. Dirscherl, Microchin. Acta, 477 (1982 II).

You might also like