::.
.
.
en
TOF
ORNL P
160.
.
EEEEEEEE
|125 44
MICROCOPY RESOLUTION TEST CHART
NATIONAL BUREAU OF STANDARDS - 1963
.
en so 43-6 MASTERS
ORNU P-1601
Cont 650918-5
t here
ritenere righ
tranno
STABILITY OF AROMATIC DILUENIS AND SOLVENT EXTRACTION
REAGENTS IN RADIOCHEMICAL PROCESSING*
OCT 6 1965
Charles A. Blake, Jr. and J. M. Schmitt
.
a van
Oak Ridge National Laboratory
Oak Ridge, Tennessee

RELEASED FOR 200XCENT
1.
An
a contr
want t
w
nad O.
a st,
Nu, na salam" het
ud med utter, What
o homelement
n a wimbo what the hours
*. Innhwatervast. , .,..-'mon *** ria..,1 tai vasta
- me.
A Mag
The report womparetu
, **
w
n epretthon, grund af het
roccot el Govern
Wat monoty a batt
LEGAL NOTICE
teen count on
or
moet
wertHether to Balta
worth

IN NUCLE! SCIECI ABSTRACTS
QUAD-CORE*
To be presented at the
*. INTERNATIONAL CONFERENCE ON THE CHEMISTRY OF THE
SOLVENT EXTRACTION OF METALS
-- "*;***.' ... mit
on September 27-29, 1965; Harwell, England
-.
-
-
-
To be published after editorial review .
in the Proceedings of the Conference
-
-
-
-
-
- -
-
"Research sponsored by the V. S. Atomic Energy Commission
under contract with the Union Carbide Corporation
#
T
A
-
*
--
4-
-
-
-
•••••••••••
••
-
+
+-
--+-
--- -
--
--
-*
-
--
-
• • • • •
Tr
-
7
+ +
+
,. ?Vir 4
.
41
STABILITY OF AROMATIC DILUENTS AND SOLVENT EXTRACTION
REAGENTS IN RADIOCHEMICAL PROCESSING
Charles A. Blake, Jr. and J. M. Schmitt
Oak Ridge National Laboratory
Oak Ridge, Tennessee
ABSTRACT
The stabilities of some commercial aroma tic diluents, especially
the mixed isomers of diethylbenzene, DEB, are determined and examined
with respect to their effect on the use of the diluents in solvent
extraction processes. Degradation was by mixing the organic phase with
an acidic aqueous phase, usually '2M HNO3, in the presence or absence of
6°Сo gamma radiation. The degraded solvent was examined by severa i ...
techniques; measurement of extraction ability, gas liquid and column
chromatography, distillation, infra red and nuclear magnetic resonance
spectrometry, differential thermal analysis, and elemental analysis.
Exposure of IM TBP/DEB solution while stirring with 2M HNO3 to 2 watt
hr/1 radiation degraded about 0.3% of the DEB. At 70 watt hr/1 about
10% of the DEB degraded. The meta isomer was the most stable. The
principle DEB products were ethylacetophenones and l-nitroethyl-ethyl-
benzenes and are poor extractants for uranium and fission products.
Impurities frequently associated with DEB degraded to compounds which
extracted fission products strongly.. DEB degraded 2 to 5 times faster
when TBP was present in the organic phase, but metals dissolved in
both phases retarded degradation slightly. Products in test3 where
of the DEB was degraded depressed uranium extraction slightly, decreased
the rate of phase disengagement from acidic and alkaline solutions by
a factor of 2, did not change the solvent flash point. No satisfactory :
method of product removal from the organic phase was developed but addi.
tional testing is needed to determine rate of buildup during actual pro-
cessing. Measurement of the heat of degradation of solvents in a new
adaptation of the method of differential thermal analysis 18 described
and it is shown how this method should be useful in estimating the re- ;
quirements of processing apparatus to dissipate heat evolved during :
degradation. The calculations show little effect of 2 watt hr/1 DEB
exposure on extraction column temperature, or of degradation of en-
trained DEB in aqueous phases being concentrated by evaporation. Com-
mercial diisopropylbenzenes and detergent "alkylates" are le88 stable
than DEB. Data are included which summarize the ORNL observations
which pertain to the stability of some extraction reagents, @.8., at
amines, phenols, and a dialkyl phenyi phosphonate.
VW
#Research sponsored by the U. S. Atomic Energy Commission
under contract with the Union Carbide Corporation.
STABILITY OF AROMATIC DILUENTS AND SOLVENT EXTRACTION
REAGENTS IN RADIOCHEMICAL PROCESSING*
Charles A. Blake, Jr. and J. M. Schmitt
Oak Ridge National Laboratory
Oak Ridge, Tennessee
1.
INTRODUCTION
Interest in aromatic diluents stems from results of ORNL studies
which showed that significant improvement in the performance of a veri-
ety of solvent extraction-metal recovery processes would be realized if
the commonly used TBP-aliphatic diluent combinations could be replaced
with TBP-aromatic combinations. The benefits from such a change could
include increased TBP extraction power, improved solubility properties
of some of the extracted TBP-metal complexes, more highly selective ex-
traction of uranium and other metals, and increased TBP stability against
radiation-enhanced chemical degradation. Little was known, however,
about the behavior of the aromatic diluents themselves during radiochemi-
cal processing. For example, it was not known which structural types
were most useful, or which, if any, were suitably stable, or hat degra.
dation products are formed, or how such products and their formation
might affect solvent properties, including safety.
..
*Research sponsored by the U. S. Atomic Energy Comission under contract
with the Union Carbide Corporation
TASSAR
In 1962 a study at ORNL had just been finished of some of the
factors involved in determining the suitability of aliphatic diluents
for use in radiochemical processing. This program was reoriented to
incorporate testing of aromatic diluents, and preliminary evaluations
of the stabilities of some samples of alkylbenzenes, that were then
Ware
available, presented at Gatlinburg in 1962.2 It was evident (Table 1)
that several of the simple alkylbenzenes** compared favorably with n-
dodecane (the aliphatic standard) with respect to resistance to forma.
tion of fission product extracting degradation products when heated
together with nitric acid. But fission product extraction power is
only one measure of diluent degradation; indeed, because of its sensi-
tivity to trace impurities, it is probably not the best test if it is
desired to eva luate completely the performance of diluents of differ-
ent structure. The poor showing of Solvesso 100 (Table 1), 18, for ex-
ample, due to a large extent to compounds other than its principle con-
stitutents, which here are predominately Cao and C12 alkyl benzenes.
A significant change in the diluent price structure has occurred
since 1962. "Straight chain" alkylbenzenes in the dodecylbenzene range
are now approaching costs of $0.75 per gallon because of their use in
large-scale production of biodegradable 'detergents. Since the n-alkyl-
benzenes performed well in the tests of Table 1, it was hoped that these
**Candidates for future use as aromatic diluents are the liquid tetra-
methylbenzenes, prehnitene and 18 odurene (Table 1), which are not yet
commercially available, although there is considerable interest in the
polymethylbenzenes, a., Durene, a solid symmetrical tetramethylbenzene
is now available. .
TABLE 1.
PERFORMANCE OF DEGRADED DILUENTS
9520-ND Y
Diluent
Diluent
Flash point
(closed cup)
(F)
| (c/sec/ml)
(calcium test)
Alkyl Benzenes
110
100
100
100
100
>6000
200
1400
500
Methyl (toluene)
1,2-Dimethyl (o-xylene)
1,3-Dimethyl (m-xylene)
1,4-Dimethyl (p-xylene)
1,2,4-Trimethyl (pseudo-
cumene)
1,2,3-Trimethyl (hemimel:
litene) 73%
1,2,3,4-Tetramethyl (prehn-
itene)
1,2,3,5-Tetramethyl (180-
durene).
Ethyl
Diethyl (mixtures)
Triethyl (mixture)
Propy1
Isopropyl (cumene) ;
D118opropyi (mixture)
Triisopropy1
n-Butyl
sec-Butyl
tert-Butyl
180-Butyl
Tetralin
1-Methyl-4-18opropyl (para-
cymene)
sec-Amy1
tert-Amy! (1,1-dimethyl-
propyl)
n-Hexyl
Cyclohexy1
a-Nonyi
Solve880-100
Solve880-150
10Z
140
80, 450
30, 4000
*6000
>6000
170
1608
125
126
130
120
1408
45
.
171
138
decomposes
4000
170
130
120
>6000
180
>6000
*6000
118
150
Aliphatic Hydrocarbons
128
Amsco 125-82
Ameco 125-82 (H2804
scrubbed)
g-Dodecane
4000
100
128
165
125
Open cup.
"See "Experimental Methods".
"alkylates" would be useful diluents. Brief investigation of some com-
mercial products has, however, shown them to be mixtures of 20-30 COM-
pounds some of which degrade readily on contact with nitric acid.
nil
Most of the detailed studies described in this paper have been
made on diethylbenzene (DEB) hecause: (1) It is probably the simplest
of the commercially available high-flash point-aromatic diluente, can
be obtained reasonably free from other compounds, and costs $1.40 per
ga ilon; (2) Some of the commercial batches give low f188ion product ex-
traction after degradation with nitric acid; (3) The separated and puri-
fied DEB isomers are available for individual study. The first portion
of the paper describes and compares degradation of commercial and puri-
fied DEB's by nitric acid in the presence or abseace of ionizing radia.'
tion, with and without TBP present in the organic phase and with and
without metals in both phases. Then follow sections which present the
methods used to identify the principle degradation products and the
effects which the products have on extraction properties and solvent
flash point. Comments about solvent clean-up experiments are included.
Measurement of the heat of degradation of DEB in a new adaptation.cf
the method of differential thermal analysis is described and it is shown
how this inethod should be useful in estimating the requirements of pro-
cessing apparatus to dissipate heat evolved during the degradation. Pl.
nally, data are included which describe the poor stability of commercial
detergent alkylates and diisopropyl benzenes and summarize the ORNL ob-
servations which pertain to the stability of some extraction rougents,
6.8., amides, amines, phenols and a dialkyl phenýlphoqphonato.
1
r
.
-
.
.
.
.
.
.
.
.
.
.
2. EXPERIMENTAL METHODS
Two methods were employed to degrade the solvent phases: (1) A
6000 curie cobalt source was used to irradiate an agitated 2-phase
system, the aqueous phase being initially 2M HNO3, and the radiation
power being approximately 1 watt/11ter; (2) The solvent was boiled
under total reflux with aqueous acid, the agitation being accomplished
by the boiling. The acid was 2M HNO3 except for one test series with
chloride solutions.
· TBP was present during many of the degradations. Since the inter-
est was primarily with the diluent and its properties, the degraded
solvent was usually scrubbed with 0.2M Na2C0g to remove the low mole-
cular weight acid products derived from TBP. Such scrubbing restores
TBP, degraded by either of the above methods, nearly to its original
extraction behavior as measured in the present test procedure.
The degraded solvent was examined by several techniques; measure-
.
ment of metal extraction ability, gas liquid chroma tography (GLC), col-
umn chroma tography, distillation, infra red (IR) and nuclear magnetic
resonance (NMR) spectrometry, differential thermal analysis (DTA), and
elemental analyses.
In metal extraction tests the degraded solvent is customarily used
to extract 95zr-Nb or 181Hf, from aqueous HNO3, and the ability of the
extracted metals to withstand subsequent scrubbing with HNO3 18 one of
the common measures of degradation. However, it has been observed at
ORNL that the curves showing metal ion extraction and those showing
retention ability are nearly always parallel when using the test proce-
dure described below. To simplify the presentation of data, only the
metal extraction ability of the solvent is described in this paper.
Extraction test procedure (calcium test): (1) Make the degraded
solvent IM in TBP and scrub twice with an equal volume of 0.2aqueous
Na2CO3 to remove low-molecular-weight acids, primarily DBP and MBP (10
min each contact); (2) Contact 30 min with solid calcium hydroxide,
~50 g solid/11ter of organic phase, and separate from solid; (3) Con-
tact with metal tracer solution (e.8., 1894f) in nitric acid solution
and measure the extraction. An equal volume of tracer solution con-
taining 10*, C/sec ml. in 2M HNO3 18 customarily used. Extraction :
power 18 given in terms of the distribution coefficient, D, or of the
y activity reporting to the organic phase.
In gas chroma tography an Aerograph Autoprep Model A-700 (Wilkins
Instrument and Research, Inc.) gas chrona tograph (heated metal flla-
ment detector) equipped with an I and N Speedomax # recorder was used
in this work. The column was copper and 1/4 inch x 10-20 feet long
with a packing of 20% Dow Corning SE-30/Chromosorb W-HMDS.
Elemental analyses were performed by the ORNL Analytical Division.
IR and NMR measurements were made on Beckman IR7 and Varian A-60 in-
.
struments, respectively.
Differential thermal analyses (DTA) were made by comparing tempo.
era tures in two tubes immersed in a stirred oil bath equipped with a
programmed heater. The test media are described in a later section.
. &N
3. STABILITY OF BATCHES OF COMERCIAL DEB AND PURIFIED DEB ISOMERS
The studies described in the following sections are all of the
stability of DEB and associated impurities against attack by nitric
acid. Tests in which DEB was boiled with hydrochloric acid and chlo-
ride salts showed no degradation, 2.8., DEB boiled with 6M HC1 for 48
hr gave no evidence of by-product formation in GLC tests, and chloride
uptake by the treated DEB was not detected by a method in which 1 atom
of chlorine in 1000 molecules of DEB would have been found.
URN
Tracer Extraction Tests. Wide variations were observed in com-
parisons of HNO3-degraded coramercial DEB mixtures by the metal lon
extraction test procedure (e.8., Table 1). These and other DEB mix-
tures varied in composition (50-60% meta-DEB, 25-40% para-DEB, 5-10%
ortho-DEB, 5-10% other compounds related in structure). Empirically,
their stabilities, ae measured by extraction power after degradation,
decreased as the percentage of meta -DEB decreased. 3 Deviations from
this rather simple picture were shown, however, when a previously un- ...
tested commercial DEB and some, but not all, batches of each of the
sepa rated 18omers gave low tracer extraction when degraded (Table 2).
In brief gas chromatographic examination it was not possible in dif-
ferent lots of the same material to distinguish differences which re-
lated to the tracer extraction data. It is apparent that high tracer
extraction is related primarily to impurities and that these impurities
may associate more with para- and ortho-DEB 18omers than with meta-DEB.
In further testing of this hypothesis the poorly stable D-62 DEB
(Table 2) was degraded with 2M HNO2 for 4 hours, then distilled to
.
.
.
-
-
....
....
.
.
...
.
...
...
.....
my mom
m
y
YAN
aparati in
m
lni
mob
TABLE 2. COMPARISON OF SEVERAL DEB'S AFTER DEGRADATION
i
itet
Calcium Test
mit weine" - singolo in temelin intim
181Hafnium
Distribution
Coefficient
DEB
SAMPLE
was in the methoden -
COMMERCHAL
952r-Nba
(c/8/ml)'
>6000
4000
65
t
i
3.3
0.1
0.004
-webm non main, ito naman
ORTHO
.
.0.1
wm. ir .----on gram .
.
МЕТА
D-26
D-62
D-103
D-76-A
D-106
D-76-6
D-102-A
D-105
D-76-B
D-102-B
D-107
0.82
0.001
0.0005
0.0013
...........
100
120
200
*6000
680
4000
PARA
>1
0.0012
0.1
b
AIM TBP In DEB boiled under reflux for 4 hr with an equal vol-
ume of 2M HNO3.
PExtraction from 181Hf or 957r-Nb tracer solutions, 10%, c/sec/
ml, equal volumes organic and aqueous phases, after scrubbing
degraded organic twice with an equal volume of 0.2% aqueous
Na2CO3 and contacting scrubbed organic for 30 min with solid
calcium hydroxide.
2.
LILA
-"
,
h
;
:
:
!
recover undegraded DEB. This cycle was repeated 3 times. Samples of
the initial material and each of the distilled products were made to
IM with TBP and then subjected to the standard degradation test. The
hafnium extraction coefficient decreased from an initial D=3.3, to 1.1,.
0.25 and finally to D=0.13. Initially, the ratio of 18omers was (GLC)
meta :para :ortho=1:0.83:0.24. The final ratio, 1:0.69:0.24, showed a
preferential decomposition of the para isomer or more likely, of mate-
rial which reported with the para-DEB in the gas chromatogram. It was
not possible, however, to resolve the apparent para-DEB GLC peak into
more than 1 component of use by a GLC column specially developed to
resolve DEB isomers”.
The nitrogen analysis of each of the degraded solutions before dis-
tillation was essentially constant at 0.35M and represents nitration,
on the average, of 1 mole of DEB for each 18 originally present. Since
the hafnium distribution coefficients were independent of the nitrogen
uptake, it became apparent that additional test methods were required
to elucidate the degradation scheme.
Comparison of Degradation by_Boiling and Irradiation. Gas chro-
matograms after meta-DEB degradation by 2M HNO, with and without Irra-
diation are nearly identical (Fig 1). Six or more degradation products
are apparent and 3 of these (peaks 3, 4 and 5) predominate. Two of ..!
these 3 products (identified by methods described in a later section) are:
3-ethylacetophenone, (4); and l-nitroethyl-3-ethylbenzene, (5).
&CHE
NO
-CH-CH3
.
.
CzHS
COMPOUND 4
₂ HS
COMPOUND 5
10
Together with a small amount of ethylbenzoic acid isolated from the
aqueous phase, the 3 products account for greater than 95% of both the
total degradation and the nitrogen uptake. Chromatograms of degraded
para- and ortho-DEB are similar to those of Figure I in numbers of pro
ducts and in their distribution. Ortho-DEB 18oner products emerge from
the column about 5% faster than their meta-DEB counterparts, para-DEB
18omers emerge about 7% slower.
Degradation vs. Time.
The true extent of degradation and increase
in concentration of individual degradation products have been followed
-
-
-
principally by quantitative gas chromatography and, to a smaller extent,
by analysis of the nitrogen uptake by the degrading solvent phase. A
composite representation of data obtained by both methods is showa on
Figure 2 for each of the DEB 18omers when degraded separately by boile
ing with 2M HNO3 under reflux. About 5 and 20% meta-DEB decomposed
after 5 and 10 hours boiling, respectively, and about twice as much of
each of the ortho- and . para-DEB's degraded in the same time intervals.
(In these tests the aqueous HNO2 concentration was maintained at 2M by
periodic replacement of the depleted acid).
The 1somers of DEB, degraded with HNO in the presence of Boco ir-
radiation, were analyzed similarly and the data are shown in Figure 3.
The curves are those of Figure 2 drawn on the expanded scale shown at
the top of Figure 3. The Irradiation dose scale at the bottom of
Figure 3 was arbitrarily adjusted to make the meta-DEB chemical damage
curve fit the meta-DEB radiation damage data. The para -DEB radiation
data fit the upper curve nicely at 35 watt hr/1 dosage but then fa11
.
am
T
.
.
11
off. The ortho-DEB radiation damage data are everywhere lower than
the corresponding chemical damage curve. Gas chromatograms showed
that radiation and chemical damage produce the same products, but
apparently at rates which differ with the 1somer. Damage observed
at 70 watt hr/1 18, for example, equivalent to boiling the meta -DEB
for 5 hr, the para-DEB for 4 hr and the ortho-DEB for 2 hr.
The distribution of products changes with increased degradation.
At the highest exposures compounds 4 and 5 predominate. Compound 3
reached a maximum concentration (at about 5 hr boiling for meta-DEB)
and then decreased. .
1
Degradation in the Presence of TBP. Diethylbenzenes degrade 2 to
5 times faster with HNO3 when TBP 18 present in the organic phase dur-
ing the treatment (Table 3). As in the case with aliphatic diluents,
this '18 probably because of nitrate and nitrite extraction by the TBP.
When a IM TBP solution in n-dodecane was irradiated to 70 whr/1
In a test procedure identical to that used to obtain the data of Table
3., the organic phase nitrogen concentration was 0.25M, which 18 equiva-
:
lent to 0.078 moles (N/mole dodecane.
Degradation in the Presence of Dissolved Metals. A commercial DEB
(D-116), IM in TBP, was mixed with an aqueous 2M HNO3 solution contain-
ing dissolved metals and the stirred 2-phase system was irradiated.
The object of the experiment was to simulate in laboratory tests a con-
dition more nearly approximating the conditions prevailing in large
plant processing equipment. Periodic analyses of the organic phase
TABLE 3. COMPARISON OF DEB DEGRADATION BY HNO3
DURING IRRADIATION WITH AND WITHOUT TBP
35
DEB
200
220
watt.hr/1 BºCo Irradiation
70
w/1M TBP w/o TBP w/1M TBP W/0 TBP
Total Degradation Products In Arbitrary GLC Units/Mole DEB
meta (D-102A)
, 75
para (D-102B)
185
225 . 120
Commercial (0-62) 280
Moles Nitrogen/Mole DEB
meta (D-102A)
0.046 0.014 0.079 0.027
para (0-102B)
0.065 0.025 . . 0.093 . 0.040
0.087 0.035 0.133 0.078
80
225
"while stirring with 2M HNO3, dose rate = 1.5 watt/1
100% DEB = 6.4M, DEB = 4.6M In IM TBP/DEB
and that from a control test without metals present showed that the
rate of DEB degradation was slightly lower in the presence of the
metals (Table 4).
4.
ISOLATION AND IDENTIFICATION OF MAJOR DEGRADATION PRODUCTS
.-.-
Very little previous study has been made of the reactions of DEB
.-.
-.-
.
w
a
i son
with dilute nitric acid. Voswinkel, e.g., 4 describes formation of ethyl
benzoic acids and phthallic acids, but no reference was found which de...
scribed substitutions on the side chains. Methylisopropyl benzene 18
reported,5 however, to give methyl acetophenone when treated with di-
--,
.....
.
..
.
..
lute nitric acid, and l-nitroethylbenzene has been prepared from ethyl-
benzene. Since IR spectra of degraded DEB showed strong absorption at
wave lengths corresponding to side chain ke tone and nitro groups these
seemed strong clues to the likely degradation product structures. Little
nitration of the benzene ring itself was expected, 5,6 and IR spectra of
degraded DEB confirmed this expectation.
Because the small amount of major DEB products were resolved by
GLC (F18 1), it was hoped that preparative gas chromatography could be
used for their separation and identification. To provide suitable feeds
for the operation a concentrate of para-DEB degradation products was
prepared as follows:
.
100 g para-DEB were mixed with boiling 2M HNO3 for 16 hr.
GLC analysis showed that about 45% of the para -DEB degraded
in this time and that the principle products of this severe
treatment were most probably the para-DEB ana logs of com-
pounds 4 and 5 of (F1: 1). The a 8-degraded material was
scrubbed with an equal volume of 1.0M NaCO3 (Subsequently
about 1-2 8 of material tentatively identified as ethyl-
:
1:
TABLE 4. DEGRADATION OF IM TBP/DEB IN THE PRESENCE
OF DISSOLVED METALS
1M TBP/DEB Degraded by 2M HNOJ
B0C0 Dose a
(whr/1)
Without Meta 18
With Metals
Total, Products by GLC-Arbitrary Units
210
420
210
400
· 4706
590
Total Degraded DEB (%)
11 .
:
:
70
b.
"Dose rate = ' 1.0 w/hr
'Equilibrium metal concentrations
Organic pha se: Total metals = 0.
0.0155M zr)
Aqueous phase: Total meta 18 = 0.153M, (0.0059M U, 0.11M FØ,.
0.037Zr)
.
.
15
The
benzoic acid was recovered from the bicarbonate solution).
wa shed material was then distilled under vacuum
the following fractionations being made:
IBP to 70°C (predominantly DEB); 70-100°C (predominantly
ketone); 100-132°C (predominantly nitro compound). About
5% of the initial 100 g samples rema ined as pot residue.
Ethyl-Acetophenone (Compound 4). The 70-100°C fraction was then
further fractionated on a 20 ft column in the GLC ($ 2.0). Several
grams of product were prepared which gave a strong IR ketone band much
like that for acetophenone at 1680 wave numbers. The equivalent weight
of the ketone was determined to be approximately 154 and this with the
elemental analysis indicated C:8:0 ratios of 9.5:12.1:1.0. The actua 1
analyses and the theoretical analyses for 4-ethyl-acetophenone (C:H:0–
10:12:1) are given below:
.
Percent
'
.
Peak 3
4-Ethyl-Acetophenone
.
.
.
.
-
-
-
Carbon
Hydrogen
Oxygen
81.2
7.9
10.5
81.1
8.2
10.8
A sample of 4-ethyl-acetophenone was obtained from Eastman Kodak
Company and the material appeared to be identical to the isolated mate-
rial in GLC tests.
1-Nitroethyl-4-ethylbenzene (Compound 5). Attempts by GLC to
prepare pure material from the 100-132°C distillation fraction corres-
ponding to peak 5 (F18 1) were not successful. The product always con-
tained appreciable amounts of the 4-ethylacetophenone from the therma l
IND
decomposition of the nitro compound. Separation was, however, made by
the technique of column chroma tography. The sample was diluted in pe-
troleum ether and added to the top of a 30 cm x 1 cm column of neutral
a lumina wet with petroleum ether. Washing with petroleum ether removed
the small amount of DEB Initially in the sample and washing with ben-
zene eluted the nitro compound. The ketone and other products remained
on the column. The NMR spectrum of the sample after evaporation of the
petroleum ether gave excellent confirmation of the l-nitroethyl-4-ethyl-
benzene , structure. Elemental analyses verified the expected mole ratios
of C:H:0:N=10:13:2:1. Some of the product was sealed into a small vial
which was then heated at 110°C for 1 1/2 hr. Gas chroma tograms of the
heated material showed that 25% of the nitro compound had become 4-ethyl-
acetophenone. An additional 1 1/2 hr heating transformed another 25% of
the nitro compound. This explains the difficulty of purifying the altro
compound by GLC and is further confirmation of the structure of the two
.
.
compounds.
Other Products. Brief examination of silica gel and neutral a lumina
with their elutropic series (petroleum ether, benzene, diethyl ether, ace-
tone, methyl alcohol) showed that resolutions of the other and less signifi.
cant degradation products of DEB were possible. It appeared that mate-
rials which could not be eluted from the packed column by the solvents
amounted to less than 2% of the initial DEB sample before degradation. .
-o
-
'.•*
-.
5. EFFECT OF DEGRADATION PRODUCTS ON THE EXTRACTION PROCESS
-.
Extraction Power
Degradation products from pure DEB's have al-
..
-.
ready been shown to be poor extractants for 95zr-Nb and 101Rf (Table 2).
.
.-mern
.
-
' -
Insufficient supply of the compounds has limited further examination
of their f188ion product extraction power. In a few uranium extrac-
tion tests (Table 5) the nitro and ketone products showed very low
uranium extraction ability when used alone. In combination with TBP
the pure compounds depressed uranium extraction slightly as did de-
grading a IM TBP/DEB solution. The concentration of products used and
the 4 hr's boiling given the IM TBP/DEB solution are equivalent to ,
severe degradation of the organic phase.
Phase Separation. The presence of degradation products of DEB
and its impurities in an extractanc phase has an adverse effect on
disengagement from aqueous phases. For example, a 1M TBP/DEB solution
degraded by boiling with 2M HNOg for 4 hr takes about twice as long as
the undegraded solution to separate at room temperature from 2M HNO3
(90 sec vs 45 sec) and from 0.2M Na2CO3 (60 sec vs 30 sec).
Flash Point. The closed cup flash points of degraded and fresh
IM TBP/DEB were essentially the same:
Fresh IM TBP/DEB
Flash Point (c.c.) = 136-1387
Degraded IM TBP/DEB
Flash Point (c.c.) - 135-136°F
(70 watt hr/1 86Co
exposure while stirring
with 2M HNO3)
...
Solvent Cleanup. DEB of proper purity should present no more
difficulty from the standpoint of unwanted fission product extraction
than that encountered in use of g-dodecane ($ 5.1 and Table 2). As 18
the case with aliphatic diluents, it is difficult to remove completely
TABLE 5, EFFECT OF DEGRADATION PRODUCTS
ON URANIUM EXTRACTION
Uranium Distribution Coefficient
from
2M HNO3, 0.004M U(VI)
Extractant Phase
IM TBP in DEB
IM TBP in DEB
Degraded 4 hr, 2M HNO3 (boiled)
IM TBP in DEB
0.4M 1-Nitroethyl-4-ethylbenzene
IM TBP in DEB
0.5M 4-Ethylacetophenone
0.4M 1-Nitroethyl-4-ethylbenzene in DEB
0.5M 4-Ethylacetophenone in DEB
10-5
Nitrogen analysis of the organic phase after degradation = 0.48M
19
DIT
the DEB degradation products from the organic phase. Scrubbing with
water and aqueous alkaline solutions (NaOH, LLOH, Na2CO3) removes some
of Compound 5, but 18 Ineffective in Compound 4 removal. Scrubbing
with ethanolamine efficiently removes most of Compound 5, but this
method of cleanup 18 subject to the same limitations previously noted.2
(Expense, loss of TBP, distribution of ethanolamine to the organic
phase). Further testing 18 required to determine the extent of pro-
duct buildup in a continuous process and whether such buildup actually
constitutes a problem.
Mea surement of the Heat of Diluent Degradation by Differential
Thermal Analysis. When a diluent is used in a solvent extraction pro-
cess it is imperative to know whether its degradation may involve haz-
ardous chemical reactions or products which are explosively unstable
under process conditions. Much of the solvent study here and elsewhere
has been concerned primarily with the degree of degradation of various
diluents, and little attention has been paid to developing hazard eva l-
uation test procedures. Differential thermal analysis (DTA) has been
used recently? to determine quantitatively the kinetic parameters for
certain types of reactions and it seemed likely that DTA might be used
to measure heat evolved during diluent degradation. With the help of
W. H. Baldwin of the ORNL Chemistry Division, a system was developed
which would permit heating mixtures of either aliphatic or aroma tic dlo
luents, tributyl phosphate, nitric acid, and water in a single phase.
Nitrobenzene, which did not react with any of these components under
... the conditions of the test was used as the solvent. Gas chromatograms
:
:
20
.
of the DTA degraded para-DEB and DEB degraded by boiling with 2M HNO3
were an indication that the decomposition reactions were similar even
though the conditions for degradation were very different.
The quan-
tity of DEB degrading during the DTA test was estimated from the chro-
:
ma tograph and the heat evolved during its decomposition was estimated
from the thermograph. It was concluded that the heat of degradation
of para-DEB with dilute nitric acid 18 50-70 kcal/mole. (In compari-
son, the heat of nitration of benzene to nitrobenzene 18 27 kcal/mole,
the heat of oxidation of hydrogen to water 18 58 kca 1/mole, and the
heat of solution of hydrochloric acid is 17 kca 1/mole.) In parallel
testing di-180-propyl benzene gave a heat of degradation of about 40
kca 1/mole. Exothermic degradation in tests with n-dodecane was ob-
served on the thermograms, but the total material degraded was too
small to enable determination of the heat of reaction.
In another application of DTA, some concentrated DEB degradation
products (the sti 11 pot residue from a vacuum discillation of nitric
acid degraded DEB) were sorbed on a lumina and heated. Exothermic re-
· action began at 160°C.
It seems likely that with further refinements of the DTA techniques
this method could prove a useful adjunct to evaluation of diluent qua l-
ity and stabllity. Assume, for example, that 1M TBP-DEB solvent phase
(4.7M DEB) 18 exposed to a high radiation dose equivalent to 2 watt hr/1.
Fig 3 and Table 3 Indicate that between 0.15 and 0.3% of the DEB would
be degraded with a release of 500 to 1000 cal from each liter of organs.c
phase (uaing 40 kcal/mole DEB). Some of this heat will be dissipated
in the aqueous phase with which the organic phase 18 being mixed, but
P
2
!.
**
*
.
Y
-
Fr.
Y
21
neglecting this and assuming a specific heat of 0.4 for the organic
phase, its temperature could increase by 1.5 to 3 degrees. Further
assume that 0.5 ml of a IM TBP/DEB organic phase is entrained in 1
liter of aqueous phase which is to be concentrated 10-fold by evap-
oration. The heat evolved if the DEB were totally decomposed would
amount to -1600 cal per liter of concentrate, an amount which 18 small
compared to the quantity of heat being supplied to sustain the evap-
oration.
6. STABILITY OF OTHER COMMERCIALLY AVAILABLE ALKYLBENZENES
Two additional a lkylbenzene products are available.commercially
in quantity, i.e., "detergent alkyla tes" (alkylbenzenes prepared from
a-olefins in the Cg to C24 range) at less than $1.00/gallon, and d1180-
propylbenzene at less than $1.40/gallon. Their poor stability against
attack by HNO3 does not at this time make attractive their use as di-
luents in radiochemi :al processing of nitrate solutions, but this does
not preclude their use in other systems.
Stability of Alkylbenzenes made from Alpha-Olefins. Production
of biodegradable detergents has made a lkylbenzenes (alkylates) derived
from straight-chained a-olefins available in quantity at prices compe-
titive with n-dodecane or DEB. Three commercially available products,
each prepared from a-olefins in the Cg to C14 range were examined with
regard to their stability toward nitric acid degradation. As indicated
by a high organic phase nitrogen concentration and a contribution to
high 101Hf extraction, each of these products was found to degrade
22
severely when boiled for 4 hours with 2 HNO3. This behavior 18 in
contrast to that of g-octylbenzene which had good stability under the
same conditions (Table 6). Extraction of 18Hf considerably higher
than those given in Table 6 was obtained when the degraded alkylates
were contacted with the solid Ca (OH)2 for a time longer than the 30 min
specified in the calcium test or when the materials were scrubbed re-
peatedly with IM LLOH. In the case of alkylate D-111 extraction in-
creased from D=0.07 after the first contact to D=14 after the fourth
contact. It is apparent that the rate at which the degradation pro-
ducts are converted to their most extracting form 18 much lower than
that for DEB degradation products.
Examination by gas chroma tography of the alkylates used in these
tests showed each to have at least 17 components. Other workers have
shown that while a-olefins were used in the synthesis of the alkylates,
most of the products were secondary alkylbenzenes and that a high per-
centage of these had phenyl substitution on the interior carbon atoms
DI21
of the olefin. No n-alkylbenzenes were observed and a significant
amount (-10%) of more highly branched isomers and other hydrocarbons
were found. Such branching provides focal points for attack by HNO3
and explains the observed instability. It is possible that some pre-
treatment, e.g., heating with HNO, and distillation, might improve the
stability of the subsequent product, but the practicality of such trea t-
ment would necessarily depend upon the relative amounts of stable and
uns table components.
Stability of Commercial Diisopropylbenzenes. A commercially avail
:
able d11sopropylbenzene (66% meta, 34% para, <0.5% ortho) was among the
..
TABLE 6. STABILITY OF ALKYLBENZENES MADE FROM
Cg TO C24 O-OLEFINS
181Hf
Alkylbenzene
Distribution
Coefficient
· (Calcium Test)º
Organic-Phase
Nitrogen
Uptake
(M)
-
...
-
-
-
- -
0.40
1
Detergent alkylate D-109
Detergent alkylate D-111
Detergent alkylate D-110
0.39
0.32
0.34
.
0.35
0.26
..
,
n-octyl (pure) D-4
0.002
0.01
IM TBP Solutions in alkylbenzene boiled under reflux for
4 hr with an equal volume of 2M HNO3.
See "Experimental Methods."
original diluents tested in 1962 (Table 1). Samples of meta- and para-
dilsopropyl benzenes have since been tested and GLC examinations, nitro-
gen analysis of the degraded materials, and DTA (15) confirm the Insta-
bility which was indicated by the extraction tests. Boiling each of the
isomers (14 TBP) with 2M ANOg for 4 hr degraded them about 25%, and the
organic phases contained about im nitrogen. GLC chroma tograms showed
only 1 major degradation product, and by ana logy with DEB degradation,
this product is expected to be:
i c-(CH3)2
'?-G3 Hy
. SUMMARY OF ORNL DATA DESCRIBING STABILITIES OF SOME
SELECTED SOLVENT EXTRACTION REAGENTS
The preceding pages have described solvent degradation almost ex-
clusively in terms of the diluent portion of the organic phase even
though TBP was present during many of the degradations. TBP, however,
has high stability in the area of its current application and its prin-
ciple degradation products, when forned, are usually removed by simple
a lkaline scrubbing. The stabilities of some of the newer reagents which
.
.
.
.
.
have been proposed for use are not so well known, however, and the fol-
.
lowing section constitutes a brief summary of the data which have been
accumulated at ORNL pertaining to some of these reagents and gives re-
ference to publications where the data are treated in more detail.
.
It is interesting that while there 18 variety in the functional
groups contained in these reagents, e.8., amine, phenol, amide, the
points of attack by acids are frequently at carbon atoms whose hydro-
gens are activated. In this respect the stabilities of both reagent
and diluent against degradation by HNO3 and Hc1 are very much related.
Stability of Di-sec-butyl Phenylphosphonate. Undiluted di-sec-
butyl phenylphosphonate (DSBPP), a solvent extraction reagent with good
hydrolytic stability and excellent ability to separa te uránium from
thorium, 2 was irradiated (Ⓡoco gamma-ray source) to 100 whr/1iter while
being stirred with 2M HNO3. The fission product extraction power of
the irradiated material was less than that of irradiated tributyl phos-
phate in a comparable test. Each reagent was nitrated to the extent
of about 0.05 mole of nitrogen per mole of reagent. It was interesting
to find that about one-third the nitrogen was removed from both reagents.
by scrubbing with strong alkalies (e.8., IM LLOH) and at the same time
the fission product extraction abilities were reduced to very low levels.
The degree of reagent nitration noted in these tests was not predicted
from previous testsº in which the reagent had been irradiated with a
limited amount of nitric acid dissolved in the organic phase and fur-
ther tests are required to explain these results.
In further studies, solutions of TBP or DSBPP (each 1M) in DEB were
Irradiated to 70 whr/1 while being stirred with 2M HNO3. Nitrogen up-
take in the organic phases was 0.1M and 0.0%, respectively, (Table 7)
and GLC ana lyses Indicated that the nitrogen was present almost com-
pletely in DEB-nitro compounds of Type 5. Extraction of 184Hf from
26
TABLE 7. COMPARISON OF HAFNIUM EXTRACTIONS BY DEGRADED
DI-8ec-BUTYL PHENYLPHOSPHONATE (DSBPP) AND TBP
Hafnium Extraction from 2M HNO, with
Degraded
Solution
Nitrogen
Uptake
181Hf Tracer
Distribution
Coefficient
0.1M Hf, Moles
Complexed/1 Organic
Phase After
Calcium
Scrub
Test
(calcium test)
IM DSBPP/DEB
0.07
1.4
0.0002
0.0006
IM TBP/DEB
0.10
1.7
0.0002
0.0023
Irradiated while stirring with 2M HNO3 In 80Co source, 70 wht/1 exposure
see "Experimental Methods"
2M HNO3 (calcium test) was slightly lower with DSBPP solution than with
that of TBP(D=1.4 vs D=1.7). Organic phase ha fnium concentrations after
· extraction from 0. IM Hf solution (traced with 181Hf) were low (Table 7).
Phenyl substituted organophosphorus reagents have been reported10
to be susceptible to formation of phenol groups on the benzene ring
ortho to the attached phosphorus a tom, and it has been reportedlo that
the products extract fission products strongly. However, this 18 con-
tradicted by IR and tracer studies. Infrared absorption by the irra-
dlated 100% DSBPP in the 3200 wave number region attributable to phenol
groups of this type was negligible in a method where the lower limit of
analysis 18 1 part in 250. A similar examination of the irradiated IM
DSBPP was made, but diluent absorption was too strong to permit phenol
.
detection.
It is difficult to determine relative hafnium extractions
by reagent and diluent degradation products, but the low extractions
from the strong 0. IM hafnium solution indicate low concentrations of
any materials which extract hafnium strongly.
Stability of Amides. Amides, particularly those of low molecular
weight, are subject to hydrolysis in both acid and alkaline media. Two
commercially available amides,
1
9-CH2O-NCR
and
9-C12Heste-Nets
CH
CH3
were each boiled (undiluted) with an equal volume of 2M HNO2 under total
reflux for 4 hr. The degraded amides were diluted to 1M (based on com-
position of the original sample) with DEB. These solutions were scrubbed
several times with water and once with 0.21 Na 2003 to remove
28
:
nitric acid. The scrubbed solutions were titrated with sodium hydrox-
ide and were used to extract uranium and thorium from 2M HNO3 (Table 8).
The decrease in uranium distribution coefficient, Do[amide)?, indicated
a decrease of amide concentrations by 40% and 30% respectively. The
concentrations of carboxylic acid found (about 20% of Initial amide) were
!
in fair agreement with this, considering the probable loss of some car-...
boxylic acid to the sodium carbonate scrub. In a similar test with TBP
the uranium extraction coefficient indicated loss of 13% of the reagent.
Sidda 1122 reports that other amides have greater stability against
acidic hydrolysis than does TBP, 2.8., compounds of the following struc-
tural type:
:
".
..
.
RYCH-E-NR.
Stability of Amines. An important application for amines 18 in
the transuranium element (TRU) recovery program. In that program the
Tramex process uses 0.6M tertiary amine hydrochloride (a lamine 336-HC1)
in DEB for TRU recovery from chloride solutions. Counter-current tests
with 242cm at power densities varying from 5-10 watts/1 (reagent dose
~5 whr/1) showed no adverse effect on golvent extraction due to a irra
diation. 12
In nitrate systems, relatively high selectivity for plutonium ni-
trate over fission products in amine extraction, and high resistance of
amines to radiation damage, have been noted in general by various
TABLE 8. STABILITY OF AMIDES
Dimethyl
p-octyl amide
Dimethyl
n-dodecyl amide
a-octyl anide
TBP
2-dodecyl amide
Analysis
Initial
Degraded
Initial
Degraded
Initial Degraded
:
24
1
Titration (wt% acid)
U Distribution coefficient
Th Distribution Coefficient
2
5
0.1
24
2
0.02
20
3
0.04
-
24
-
18
.
7
0.11
.
Uadiluted amide or TBP boiled 4 hr under reflux with an equal volume of 2M HNO3. Degraded amide
diluted to lM with fresh DEB. IM Solution scrubbed with water and 0.2M Nazcoz.
"Titration was with aqueous sodium hydroxide. Percent acid calculated as weight percent of
corresponding carboxylic acid.
CUranium and thorium extraction with 1M reagent in DEB from 2M HNO3 ; equal phase ratio; room
temperature; 10 min contact.
labora tories. The following tests show that some zirconium-extracting
material 18 produced on prolonged contact of the amine extractants with
nitric acid. 13
to
A number of amines (Alamine 336 and trilaurylamines from several
sources) in fresh 0.3M solutions extracted from ~10 to as high as ~200
c/sec ml from tracer solution at ~104 c/sec ml in 2M HNO3. Even the .
lowest of the corresponding apparent extraction coefficients, ~10-3, 18 .:
higher than the coefficient of ~10-4 found in macro zirconium extrac-
tion. (This suggests that all the fresh amine solutions contained vary-
ing traces of zirconium-extracting impurities, although it 18 also pos-
sible that the traceż solution contained a higher proportion of extracto
able zirconium species than did the macro solutions, since the methods ....
of preparation were not identical and zirconium behavior can be highly .
dependent on its history.) Simple acid and basic scrubs removed the
extracted zirconium in rather sma 11 increments, e.8., usually 10 to 70%
per scrub, corresponding to decontamination factors of only wz or less.
The "Fresh TLA-Fresh Amsco" entry in Table 9 18 typical. In contrast
to the fairly low extraction by the fresh extractants, somé amine solu-
tions that had been used in a nitrate extraction system and then stored
for approximately one year (containing extracted nitric acid) extracted
up to 80% of the tracer activity from solutions at ~104 c/sec ml. Amine
solutions (0.3M) treated with 10M HNO2 for 2 days extracted ~15-20% of
the activity, and undiluted amines treated with 10M HNO3 for 5-7 days
and then diluted to 0.3M extracted ~3-5%, or 300-500 c/sec ml. The ex-
tracted izirconium, like the smaller amounts extracted by fresa amine solum
tions, was removed only slowly by acid and basic scrubs.