Destroying Peroxides of Isopropyl Ether
Destroying Peroxides of Isopropyl Ether
Destroying Peroxides of Isopropyl Ether
H A
peroxide levels can be accomplished by treating it
with amines or with a variety of inorganic reagents
DESTROYING PEROXIDES OF
ISOPROPYL ETHER
T he explosive n a t u r e of peroxides of isopropyl ether
and the risk of fatal accidents have led to a consumer
diethylenetriamine complete removal of the peroxides occurred
within one hour.
d e m a n d for a product with more stringent peroxide —Aqueous 20% solutions of hydroxylamine hydrochloride
limitations. Some customers now specify a peroxide- destroyed peroxides when used in cold washing treatments.
free ether, or an isopropyl ether with a peroxide content
of no more t h a n a few parts per million. Because of the —Sodium metabisulfite solution was the most effective of the
extreme susceptibility of isopropyl ether to oxidation, it is inorganic reagents tried. This reducing reagent was found to
often difficult for the producer to meet these new require- react quickly with the peroxides in isopropyl ether in essentially
ments. T h e producer may find it necessary to reprocess a stoichiometric manner.
stored ether which has acquired some peroxides in order —Stannous chloride was the only inorganic reagent which was
to meet this more exacting grade. effective in the solid form.
Peroxides of isopropyl ether have a higher boiling
point t h a n the ether itself, and the ether therefore cannot Because of other variables that would be introduced, it
be safely dry-distilled unless peroxides are known to be is not possible from the laboratory experience to choose a
absent. T h e temperature and concentration at which single " b e s t " reagent, or to give specific instructions for
explosion becomes probable have never been authorita- removal of peroxides by the use of the inorganic reducing
tively stated. Even a very small amount of peroxide may agents in plant-scale operations. Plant equipment for
be dangerous, since it is concentrated in the still system. conducting the treatment would influence the results,
T h e need for reprocessing a large volume of isopropyl and the peroxide content of the two liquid phases during
ether stored at one of our plants led to a laboratory the mixing operation.
study to find effective and economic means of destroying For our purposes, the best system was chosen as cold-
peroxides. Peroxide content of the stored ether was washing with sodium metabisulfite solutions. While
well below the customary m a x i m u m limit, but did not details of the operations were not given, it was recently
meet the new requirements. It was hoped that per- reported by the Production D e p a r t m e n t of the Chemicals
oxides could be converted to harmless substances while Division that isopropyl ether having 0.017 meq. O2 per
using an a m o u n t of reagent small enough to have no gram of ether was successfully treated with 3.0%
deleterious effect on ether quality. This goal was not sodium metabisulfite solution. After treatment and
attained—any material which completely destroyed subsequent refining, the peroxide content of the ether
peroxides also appreciably degraded the ether. Sub- was a m a x i m u m of 0.0007 meq. O2 per g r a m of ether.
sequent redistillation or dehydration is required to Stainless steel equipment is preferred for this process.
restore the ether to the specifications of a refined product.
Effect of Amines on Decomposition of Peroxides
Materials tested for their effect on peroxides of iso-
propyl ether included several amines, used both in un- Amines and amine-like compounds are known to be
diluted form and as aqueous solutions. Inorganic effective antioxidants. O n e of these chemicals, mor-
pholine, has been successfully used by the Chemicals
reagents previously recommended in the literature were
Division of Union Carbide Corp. to prevent per-
also tested. These included solutions of sodium hy-
oxidation of freshly distilled isopropyl ether. O t h e r
droxide, sodium metabisulfite, ferrous sulfate, and
amines, for example, diethylenetriamine, triethylene-
stannous chloride. In addition, solid stannous chloride,
tetramine, and tetraethylenepentamine appear to be
ferrous sulfate, and sodium metabisulfite were tested.
equally effective as peroxide inhibitors for this ether {6).
Results of these tests showed t h a t :
T h e decomposition of peroxides by amines has been
—In a cold washing operation, high concentrations of peroxides reported by a n u m b e r of investigators. C a p p and
were completely removed by treatment with 25% by weight of Hawkins (2) studied the reaction of organic amines with
undiluted triethylenetetramine, and reduced to very low concen- alkaryl hydroperoxides and stated that these peroxides
trations by treatment with this amount of undiluted diethylene- could be converted smoothly into the corresponding
triamine and propylenediamine. alcohol with loss of one atom of oxygen by treatment
—When ether which contained peroxides was refluxed in the with m a n y amines. Others (7, IG) have reported on
presence of 25% by weight of undiluted triethylenetetramine and the decomposition of peroxides by amines and have
TABLE I. EFFECT OF U N D I L U T E D A M I N E S
Amine
Peroxiues, Meg. Oi/G. Ether
%
Trial Name Added" Treatment Start End
1 Butylamine 1 Shaken 15 min.; heated 2 hr., 55° C. 0.012 0.012
2 5 Shaken 15 min. 0.036 0.033
" Per cent by weight of pure amine added to the isopropyl ether.
" After shaking together with the isopropyl ether these amine solutions remained in contact with the ether overnight {about 17 hours) before analysisfor peroxide
content. h A heavy white precipitate of hydrazine sulfate formed in the sulfuric acid solution used in the peroxide analysis. ' Eighty-five per cent hydrazine in
water. Strong evolution of gas when shaken with the ether.
V O L 56 NO. 6 JUNE 1 9 6 4 39
stannous chloride, and recommend steam distillation of TABLE I I I . EFFECT OF R E F L U X I N G WITH
A M I N E SOLUTIONS
ether over dilute sodium hydroxide solutions as a safe pro-
cedure for purification. T h e United K i n g d o m Atomic Peroxides, Meq.02/G.
Ether
Energy Authority recently investigated the known
After Reflux"
methods for treating ethyl ether containing high con-
centrations of peroxides after use in purification of ore Trial Amine Solution Start Ether Water
solutions, and reported that the best method was to 1 Butylamine 0.036 0.009 0.001
wash the ether with a n acid solution of ferrous sulfate 2 Diethylenetriamine 0.036 0.003 Nil
(3). 3 Dipropylenetriamine 0.058 0.004 Nil
4 Diethylaminopropylamine 0.004 0.001 Nil
Although sodium hydroxide and sodium carbonate 5 Ethylenediamine 0.036 0.003 Nil
solutions have been suggested as ether peroxide de- Hydroxylamine hydrochloride 0.210 0.010 Nil
composition agents, they were not very effective in 6 Hydroxylamine hydrochloride 0.004 Nil Nil
removal of the peroxides developed in isopropyl ether 7 Hydroxyethyl piperazine 0.035 0.004 Trace
under the test conditions employed (Trials 1, 2, 3, and 8 Morpholine 0.024 0.003 0.001
9 iV-Methylmorpholine 0.058 0.008 0.004
4, T a b l e I V ) .
10 A"-Aminopropylmorpholine 0.058 0.005 Nil
Among the most effective inorganic reagents tried for 11 A r -Aminoethylpiperazine 0.058 0.007 Nil
removal of the peroxides in isopropyl ether were sodium 12 Propylenediamine 0.034 0.001 0.001
metabisulfite, ferrous sulfate, and stannous chloride. 13 Triethanolamine 0.052 0.002 0.001
A comparison of the effect of these three reducing agents 14 Tetraethylenepentamine 0.035 0.003 Nil
15 Triethylenetetr amine 0.036 0.002 Nil
may be observed from the data of T a b l e I V .
16 Triethylenetetramine 0.101 Nil Nil
T h e peroxides of isopropyl ether at either low or high 17 Triethylenetetramine 0.210 Nil Nil
concentrations were quickly and completely removed
by washing the ether with cold, very dilute aqueous " One volume of 20% aqueous amine solution was refluxed for two hours
with two and one-half volumes of peroxidic isopropyl ether.
solutions of sodium metabisulfite. T h e a m o u n t of bi-
sulfite present in the wash water, it was determined, need
only be equivalent in a m o u n t or slightly in excess of the
a m o u n t of peroxides calculated to be present in the quired to remove a given a m o u n t of peroxides in iso-
ether that is being washed (compare amounts of sodium propyl ether than are required to destroy the peroxides
bisulfite and peroxide in Trials 8 and 10, Table I V ) . using aqueous sodium metabisulfite. Estimates m a d e
T h e ease and rapidity by which peroxides were de- from d a t a given in T a b l e I V indicate that these reagents
composed in isopropyl ether by use of sodium meta- must be used in excess of a b o u t five times the a m o u n t
bisulfite solution are illustrated in the data obtained in the of peroxides present.
corrosion tests of T a b l e V, where the peroxides in a high- It will be observed that the peroxides of isopropyl ether
peroxide ether were completely destroyed by stirring 10 were destroyed by shaking the ether with stannous
volumes of the ether with 1 volume of the aqueous 5 % chloride in solid form. Sodium metabisulfite and ferrous
solution for 15 minutes. Some later experiments were sulfate did not decompose peroxides when used in the
conducted to determine the actual time required for solid state.
decomposition of the peroxides to take place by reaction T h e mechanism by which the peroxides are removed
with dilute sodium metabisulfite solution. I n these by treatment with inorganic reducing agents is stated
experiments [data not included in the tables] 200 ml. of in its simplest form to be that of reaction of the oxygen
a 5 % solution of sodium metabisulfite was added to 800 of the peroxide with the reducing agent, the reagent
ml. of isopropyl ether containing 0.575 meq. O2 per being simultaneously oxidized (3). W h e n freshly
g r a m of ether in a 2-liter Pyrex reaction vessel equipped prepared, aqueous ferrous sulfate solution is a pale green
with a n electric stirrer and provided with a n access for color. After shaking with the peroxidized ether, oxida-
sampling the ether and water phases. I n this case, the tion of the ferrous ion soon occurred as indicated by a
a m o u n t of bisulfite was only in slight excess of the brown coloration of the solution and precipitation of
a m o u n t of ether peroxide present. Results of these finely divided red ferric hydroxide on the sides of the
tests showed that decomposition of the peroxides was reaction flask. A very fine white precipitate of tin oxide
very rapid. An appreciable rise in temperature oc- (SnO 2) or stannic hydroxide formed when the stannous
curred u p o n addition of the bisulfite solution to the chloride was first added to the ether. This precipitate
ether (about 10° F.). No peroxides were found in coagulates to some extent with continued shaking
samples of the ether or water phase withdrawn at any of but is difficult to remove by filtration.
the test intervals down to as low a time interval as 30 For removal of peroxides, one should consider both
seconds. W h e n the peroxide content is high, consider- the concentration of the reducing solution, and the
able heat is evolved in the treatment with sodium meta- ratio of volume used to t h a t of the ether that is being
bisulfite and the reagent should be added slowly and with
cooling.
AUTHOR A. C. Hamstead is a Staff Associate in the Re-
Larger quantities of ferrous sulfate a n d stannous search and Development Department of the Chemicals Division
chloride in the form of their aqueous solutions are re- Union Carbide Corp., South Charleston, W. Va.
Aqueous 70%
Hydroxylamine
Treatment Conditions Aqueous 5% Sodium Bisulfite Hydrochloride
Vol. reagent, ml. 120 120 250
Isopropyl ether, vol. ml. 1200 1200 800
Peroxides start, meq. 0 2 / g . ether 0.179 0.17S 0.037
Peroxides end, meq. Oz/g. ether Nil* Nil" Nil 1
Time of treatment, hr. 2 66 4
Corrosion Rate of Metals, Inch/Year Liquid Vapor Liquid Vapor Liquid Vapor
Mild steel 0.654« 0.283" 0.406 0.013 3.850 0.625
Copper 0.026 0.114 0.029 0.054 0.064 0.150
Aluminum 3003 0.035 Nil d 0.019 0.011 6.150' Nil
Stainless steel type 304 Nil Nil Nil Nil Nil Nil
" The peroxide content nil after agitating the solution 75 ? The peroxide content at the end of 7 hr. was 0.007 meq. 0%/g. ether; nil after agitating
the solution 2 hr. ' Black deposit on metal; pitted. d Shallow pitting. ' Deeply etched; heavy corrosion deposit; solution contaminated with metal salts.