Separation of Closely Boiling Isomers and Identically Boiling Isotopomers via Electron-Transfer-Assisted Extraction

Article abstract of DOI:10.1021/ac980221b

Mixtures of isotopomers with identical molecular weights (ethyl-d₅-benzene + ethylbenzene-d₅ and tert-butyl-d₉-biphenyl + tert-butylbiphenyl-d₉) have been partially resolved by making use of their differing solu

Full text of DOI:10.1021/ac980221b

Anal. Chem. 1998, 70, 3880-3885
S e p a ra t io n o f Clo s e ly Bo ilin g Is o m e rs a n d
Id e n t ic a lly Bo ilin g Is o t o p o m e rs via
Ele c t ro n -Tra n s fe r-As s is t e d Ex t ra c t io n
Che ryl D. Ste ve ns on,* Da nie l J . Mc Elhe ny, Da vid E. Ka ge , J a m e s T. Cis ze w s ki, a nd Ric ha rd C. Re ite r
Department of Chemistry, Illinois State University, Normal, Illinois 61790-4160
Mixtures of isotopomers with identical molecular weights
ethyl-d -benzene + ethylbenzene-d and ter t-butyl-d
biphenyl + ter t-butylbiphenyl-d ) have been partially
method for separating ethylbenzene from m- and p-xylene); (ii)
fractional crystallization5 (this method employs differences in
-12
(
5
5
9
-
freezing points); (iii) fractional crystallization used in conjunction
9
with catalytic isomerization;1
3-16
(iv) membrane permeation;17,18
resolved by making use of their differing solution electron
affinities. EP R analysis of the anion radicals resulting
from the partial potassium reduction of mixtures of the
isotopomers (*R and R, respectively) in tetrahydrofuran
(v) adsorption19,20 (the Universal Oil Products (UOP) “Parex”
process for separating m- and p-xylenes), and the recently
reported electrochemically modulated chromatography,21 the UOP
“Ebex” process for separating ethylbenzene from m- and p-
xylenes;22 (vi) reactive distillation which makes use of chemical
differences (i.e., selectivity) between the isomers and a third
component, the reactive entrainer.3a
•-
+
show that the equilibrium constant for the reaction R ,M
•
-
+
+
*R ) R + *R ,M deviates from unity. These EP R
results predict that a separation of anion radical from
neutral molecule would effect a separation of the isoto-
pomers. Removal of the solvent leaves a solid mixture of
neutral and potassium anion radical salt forms of the
isotopomers. Subsequent dissolution of the neutral ma-
terials from the anion radical salts yields a mixture that
is enriched in the ring-deuterated isotopomer, and reoxi-
dation of the anion radical with iodine yields a mixture
that is enriched in the ring-protiated isotopomer. Analo-
gous experiments were also used to separate closely
boiling geometric isomers.
An example of the reactive distillation process is the separation
of m- and p-xylene isomers.3a Organosodium-exchange reactions
of this type (reaction 1) are reversible and fast and involve
-
2
+
-
1
+
R H + R ,Na h R H + R ,Na
(1)
1
2
organometallic compounds with very low vapor pressures. In
reaction 1, the sodium preferentially binds to the stronger acid
component. The key to attaining a rapid equilibrium is the
introduction of a chelating agent which will complex with the
organosodium compounds. The crown ether, 18-crown-6, was
chosen for this particular purpose. Also, a hydrocarbon solvent,
Due, in part, to industrial needs, the development of non-
destructive separation techniques for closely boiling isomeric
mixtures remains an ongoing challenge in separation science.
Examples of such mixtures include aromatics (i.e., o-xylene,
(
5) Meek, P. D. In Advances in Petroleum Chemistry and Refining; McKetta, J.
p-xylene, ethylbenzene), dichlorinated toluenes, and alkyl-
J., Ed.; Interscience: New York, 1961; Vol. 4.
(6) Wolfe, P. Commun. ACM 1 9 5 9 , 2, 12.
7) Chem. Eng. News 1 9 6 8 , 46, (40, Sept 16), 50.
8) Machall, E. F. U.S. Patent 3 798 282, 1972.
substituted biphenyls.1 Other examples are isotopic in nature,
(
(
2 2 2 2
such as H O/ D O and H
16O/ H 18O.2
One of the most attractive separation processes is extractive
distillation.3a Extractive distillation utilizes the addition of a third
(9) Oden, E. C. U.S. Patent 3 788 819, 1974.
(10) Benris, A. G. U.S. Patent 3 798 283, 1972.
(
(
11) Egan, C. J.; Luthy, R. V. Ind. Eng. Chem. 1 9 5 5 , 47, 250.
component to a mixture of two closely boiling components. This
process is most effective for two closely boiling components that
are different in chemical nature. However, this separation process
cannot be successfully applied to isomers because the entrainer
has about the same effect on each of the components and
therefore does not change their relative volatility.
12) Schaeffer, W. D.; Dorsey, W. S. In Advances in Petroleum Chemistry and
Refining; McKetta, J. J., Ed.; Interscience: New York, 1962; Vol. 6.
13) Hahn, A. The Petrochemical Industry; McGraw-Hill: New York, 1970.
14) Brownstein, A. U. S. Petrochemicals; Petroleum Publishing Co.: Tulsa, OK,
(
(
1972.
(15) Goldstein, R. F.; Waddams, A. L. The Petroleum Chemicals Industry, 3rd ed.;
F. N. Spon Ltd.: London, 1967.
(
16) Waddams, A. L. Chemicals from Petroleum; The Noyes Press: Pearl River,
Alternative techniques must therefore be investigated for the
NY, 1962.
separation of mixtures of isomers. Some alternatives include the
following: (i) mixed extractive entrainers3b,4 (this is an effective
(17) Choo, C. Y. In Advances in Petroleum Chemistry and Refining; McKetta, J.
J., Ed.; Interscience: New York, 1967; Vol. 6.
(
18) Kikic, I.; Alessi, P.; Visalberghi, M. O. Inst. Chem. Eng. Symp. Ser. 1 9 7 8 ,
No. 54, 139.
(
(
(
1) Tyreus, B. D.; Luben, W. L. Chem. Eng. Commun. 1 9 8 2 , 16, 91.
2) Huffman, J. R.; Urey, H. C. Ind. Eng. Chem. 1 9 3 7 , 29, 531.
3) (a) Doherty, M. F.; Sylvester, L. F.; Terrill, D. L. Ind. Eng. Chem. Processes
Des. Dev. 1 9 8 5 , 24, 1062. (b) Berg. L.; Kober, P. J. AICheE J. 1 9 8 0 , 26,
(19) Winnick, J. Chem. Eng. Processes 1 9 9 0 , 41.
(20) Rosset, A. J. U.S. Patent 3 665 046, 1972.
(21) Deinhammerf, R. S.; Ting E. Porter, M. D. Anal. Chem. 1 9 9 5 , 67, 237.
(22) de Rosset, A. J.; Neuzil, R. W.; Tajbl, D. G.; Braband, J. M. Sep. Sci. Technol.
1 9 8 0 , 15, 637.
8
62.
(
4) Berg, L. AIChE J. 1 9 8 3 , 29, 694.
3880 Analytical Chemistry, Vol. 70, No. 18, September 15, 1998
S0003-2700(98)00221-2 CCC: $15.00 © 1998 American Chemical Society
Published on Web 08/08/1998

such as cumene, needs to be introduced in order to dissolve the
crown ether complex. Gau and Marquez have studied these
transmetalation reactions when R
1 2
is m-xylene and R is p-xylene.
The transmetalation reaction reported is shown in reaction 2.23
Reaction 2 illustrates the fact that m-xylene [m-C
more acidic than p-xylene; therefore, the sodium preferentially
binds to the m-xylene, leaving the p-xylene [p-C (CH ] free
6 4 3 2
H (CH ) ] is
Fig u re 1 . Actual (lower) EPR spectrum recorded at 190 K of a
:1.78 ratio of a ethyl-d5-benzene/ethylbenzene-d5 mixture in DME
reduced by a very deficient amount of potassium metal. The simulated
upper) EPR spectrum was generated using a 1.0:0.6 ratio of anion
1
6
H
4
3 2
)
to boil overhead in a distillation column. These results as well
as those of Deinhammerf et al.21 motivated us to investigate the
feasibility, in terms of thermodynamics, of effecting analogous
chemical separations via a simple electron transfer. Such a
process would not involve any chemical change of the substrates
other than the addition of an electron and would rely upon the
nonunity value of the Keq for an electron transfer (reaction 3) as
opposed to a proton transfer as in reaction 1. Hence, the
separation technique investigated here would rely upon the
differences in solution electron affinity as opposed to differences
in acidity.
(
•-
•-
radicals (ethyl-d5-benzene /ethylbenzene-d5 ) suggesting an equi-
librium constant for reaction 5 of 0.34. Coupling constants (in G): used
•
-
for ethyl-d5-benzene simulation, a(H) ) 4.95 (2H), a(H) ) 4.63 (2H),
a(H) ) 0.70 (1H), a(D) ) 0.11 (2D); used for the ethylbenzene-d5•
simulation, a(D) ) 0.82 (2D), a(D) ) 0.78 (2D), a(D) ) 0.114 (1D), a(H)
-
)
0.70 (2H).
differences with the fact that deuteriation at nonequivalent sites
can drastically alter the chemistry of radical anions26 to separate
hydrocarbon isotopomers that would be nearly impossible to
R + R ,Na h R + R •-,Na+
•
-
+
separate by physical means.
(3)
1
2
2
1
Since the thermochemistry of electron attachment to benzene27
and to the polyarenes28 has been shown to be made less favorable
upon isotopic perturbation, we were motivated to utilize this
thermochemical difference to probe the relative solution electron
affinities of alkyl-substituted benzene and biphenyl isotopomers.
We further hoped to take advantage of any electron affinity
differences to develop a means for partial separation of these
isomers.
Over two and a half decades ago Szwarc and co-workers24 used
EPR as an analytical tool for the evaluation of the relative
concentrations of the anion radicals of â-ethylnaphthalene and
R-ethylnaphthalene, in the presence of the neutral molecules. In
tetrahydrofuran (THF), the free energy change for reaction 4
RESULTS AND DISCUSSION
When 5 mL of dimethoxyethane (DME) solution that is 0.100
M in ethyl-d -benzene (C
5
2 5 6 5
D -C H ) and 0.178 M in ethylbenzene-
d
5
(C -C D ) containing a very molar deficient amount of 18-
2
H
5
6 5
crown-6 (18-C-6) is exposed to a freshly distilled potassium metal
mirror under high vacuum, a stable solution of both anion radicals
proved to be -600 cal/ mol with a reported experimental error of
2
4
only ∼10 cal/ mol. The nonunity value of Keq for reaction 4
means that any separation of the anionic material from equilibrated
neutral material would also represent a partial separation of the
two isomers.
A separation scheme based upon solution electron affinity has
another potential application: it could lead to better techniques
for the separation of isotopomers. Hydrocarbon isotopomers,
which have identical masses, are impossible to separate via the
usual physical means since their vapor pressures and diffusion
and effusion properties are nearly identical. Hence chemical
means must be utilized. The thermochemical properties of
isotopomers can be very different. For example, the enthalpy
•
-
2 5 6 5
and C H -C D
•-) and neutral materials is formed.
(
C
2
D
5
-C
6
H
5
The concentration of anion radical cannot exceed the concentra-
tion of 18-C-6 as potassium metal will not reduce ethylbenzene at
ambient temperature in the absence of crown ether. The ESR
spectrum of the anion radical of ethylbenzene has been previously
recorded,29 and our mixture of anion radicals (Figure 1) is best
2 5 6 5 / 2 5
simulated using a ratio of concentrations ([C D -C H [C H -
•-]
(
(
26) Guthrie, R. D.; Shi, B. J. Am. Chem. Soc. 1 9 9 0 , 112, 3136.
27) (a) Stevenson, C. D.; Espe, M. P.; Reiter, R. C. J. Am. Chem. Soc. 1 9 8 6 ,
108, 532. (b) Stevenson, C. D.; Reidy, K. A.; Peters, S. J.; Reiter, R. C. J.
Am. Chem. Soc. 1 9 8 9 , 111, 6578.
(
28) (a) Stevenson, C. D.; Espe, M. P.; Reiter, R. C. J. Am. Chem. Soc. 1 9 8 6 ,
+
+
difference between (CD
3 2 3 2 3 2 3 2
) C -CH(CH ) and (CH ) C -CH(CD )
1
08, 5760. (b) Godnow, T. T.; Kaifer, A. E. J. Phys. Chem. 1 9 9 0 , 94, 7682.
(c) Stevenson, C. D.; Sturgeon, B. E.; Vines, K. S. J. Phys. Chem. 1 9 8 8 , 92,
850. (d) Stevenson, C. D.; Sturgeon, B. E. J. Org. Chem. 1 9 9 0 , 55, 4090.
(e) Stevenson, C. D.; Rice, C. V. J. Am. Chem. Soc. 1 9 9 5 , 117, 10551.
is 419 cal/ mol.25 We have coupled analogous thermochemical
6
(
(
(
23) Gau, C.; Marquez, S. J. Am. Chem. Soc. 1 9 7 6 , 98, 1538.
24) Moshuk, G.; Connor, H. D.; Szwarc, M. J. Phys. Chem. 1 9 7 2 , 76, 1734.
25) Saunders, M.; Cline, G. W. J. Am. Chem. Soc. 1 9 9 0 , 112, 3955.
(29) Bolton, J. R.; Carrington, A.; Forman, A.; Orgel, L. E. Mol. Phys. 1 9 6 2 , 5,
43.
Analytical Chemistry, Vol. 70, No. 18, September 15, 1998 3881

1
Fig u re 3 . GC traces and the tert-butyl region of the H NMR spectra
above) of the original mixture (standard), phase II (anionic phase),
(
and phase I (neutral phase) resulting from a separation process
carried out on a mixture of m- and p-tert-butylbiphenyl based on
Reaction 6. The first GC peak is due to m-tert-butylbiphenyl, and the
second is due to p-tert-butylbiphenyl. Likewise, the low-field NMR
peak is due to the m-tert-butyl group, and the higher field peak is
due to the p-tert-butyl group.
Fig u re 2 . Percentage of C6D5C2H5 in mixtures containing C6D5C2H5
and C6H5C2D5 vs the percentage that the m/z ) 96 peak makes up
of the sum of the intensities of the m/z ) 96 and m/z ) 93 peaks.
The solid circles represent the mass spectral results for phase I and
phase II resulting from the isotopomer separation process. Note that
the percentage of C6D5C2H5 is higher in the neutral phase (phase I)
than it is in the anionic phase (phase II).
{
6 5 2 5 6 5 2 5
[C H C D ]/ [C D C H ](phase II)} for this particular experiment.
If the efficiency of the separation was 100% and the thermodynamic
equilibrium remained the same in the solid phase, the separation
factor (R) would be the same as Keq for reaction 5 at -43 °C.
However, it has been well established that R values obtained from
solid-state separations are larger than the thermodynamic equi-
librium constants found in solution for these anion radical
systems.28d Nevertheless, this is an encouraging separation factor
for a system consisting of isotopomers (with identical masses).30
As indicated in the introduction, we were interested in
extending the possible utility of this technique to geometrical
isomers. The Friedel-Crafts addition of tert-butyl chloride to
biphenyl leads to a mixture of m- and p-tert-butylbiphenyls, which
6 5
C D
•-]) that is 1.0/ 0.60 at 190 K. This ratio implies that the
equilibrium constant for reaction 5 is 0.34. Five such experiments
1
are very difficult to separate but easy to analyze via H NMR. In
our hands, the Friedel-Crafts additions resulted in the expected
mixture of m- and p-tert-biphenyls, which GC and proton NMR
analysis (Figure 3) show to be 31% p-tert-biphenyls. This mixture
was then reduced with an ∼50% molar deficient amount of
potassium metal in THF, which resulted in the equilibrated
electron-transfer reaction (reaction 6). After removal of the solvent
show that Keq ) 0.30 ( 0.03 for reaction 5 at 190 K, where the
error represents the standard deviation in the measured values
for Keq
.
Attempts to separate the anionic material (phase II) from the
neutral material (phase I) after removal of the solvent were
+
severely hampered due to the instability of the K (18-C-6)-
ethylbenzene•- salts toward return of the electron to the metal
leaving a simple mixture of 18-C-6, ethylbenzene, and potassium
metal. In a particular experiment, a mixture of 2.34 mmol of both
2 5 6 5 2 5 6 5
C D -C H and C H -C D were reduced with ultrapure potassium
metal in 20 mL of THF containing 1.73 mmol of 18-C-6. The
solvent was removed under high-vacuum conditions while the
solution was maintained at -43 °C. This is the lowest temperature
at which THF has a sufficient vapor pressure for removal. Hexane
at -43 °C was used to extract the neutral ethylbenzenes, which
were labeled phase I, from the solid anion radical salts. Iodine
in diethyl ether was added to the solid anion radical salts to cause
under reduced pressure, freshly distilled hexanes were added in
order to remove the neutral materials (labeled phase I) from the
anionic material. The anionic materials were subsequently re-
oxidized to the neutral state by the addition of I
diethyl ether and labeled phase II.
2
dissolved in
reoxidation to the ethylbenzenes [1/
•-
+
2 2
I + R ,(18-C-6)K (solid) f
From GC and NMR analysis, it is evident that the neutral phase
(
18-C-6(solid) + KI(solid) + R(liq)], which were labeled phase II. Mass
spectral analysis (Figure 2) of the two phases showed that phase
II was enriched and phase I accordingly depleted in the ethyl-d
benzene. Thus, R ) 0.84 ) {[C ]/ [C
(
phase I) is depleted in p-tert-butylbiphenyl and enriched in the
5
-
(
30) The molecular masses of the isotopomers actually differ very slightly due
to relativistic effects, but this was not considered.
6
H
5
C
2
D
5
6 5 2 5
D C H ](phase I)}/
3882 Analytical Chemistry, Vol. 70, No. 18, September 15, 1998

m-tert-butylbiphenyl (Figure 3). Conversely, phase II is depleted
in the m-tert-butylbiphenyl and enriched in the p-tert-butylbiphenyl.
The proton NMR spectra confirm the GC results (Figure 3), and
both analytical techniques show R ) {[meta]/ [para](phase II)}/
9
an enrichment in tert-butylbiphenyl-d . Again these results cor-
respond to an electron preference for the isotopomer that is not
deuterated in the aromatic moiety. This was the expected result,
as it has been observed that the equilibrium constant for reaction
9 is 0.51 ( 0.01 at 298 K,31 and deuteriation of the â sp hybridized
carbons should not affect the π-electron attachment.
3
{
[meta]/ [para](phase I)} ) 0.36 ( 0.08. Hence, solution electron
affinity differences can provide an efficient mechanism for isomer
separation.
The fact that Keq for reaction 6 is less than unity is explained
by the steric interaction between the m-tert-butyl group and the
protons on the unsubstituted phenyl moiety. A PM3 calculation
predicts that this interaction increases the dihedral angle between
the two phenyl rings to 1.2° in the meta anion radical from 0.079°
in the para anion radical. The consequent PM3 predicted Keq for
reaction 6 (assuming ∆H° ) ∆G° ) -RT ln Keq) is 0.47 at 298 K.
Due to its size, the tert-butyl substituent worked out suitably for
this geometric isomer separation study; it is also an ideal candidate
for further investigation of the isotopomer problem, as both the
ring system and substituent have nine sites for H and D
substitution.
The separation factor for both the meta- and para-substituted
isotopomers is 0.69 ( 0.05, which agrees quite well with the above
arguments. Most likely, the reason that R is greater than Keq for
reaction 9 is because the tert-butyl group actually replaces one of
the deuteriums in the perdeuterated biphenyl.
Friedel-Crafts additions were carried out separately with tert-
butyl-d
biphenyl-d10 to yield two mixtures: (1) p-tert-butyl-d
p-(CD ] + m-tert-butyl-d -biphenyl [m-(CD
and (2) p-tert-butylbiphenyl-d [p-(CH ] + m-tert-butyl-
biphenyl-d [m-(CH
When a mixture (Figure 4) of p-tert-butylbiphenyl-d
butyl-d -biphenyl is reduced with 0.5 equiv of potassium metal in
THF at -78 °C, the equilibria shown in reactions 7 and 8 as well
as the electron-transfer equilibria between the respective para and
meta isomers are established.
9
chloride and biphenyl and with tert-butyl chloride and
SUMMARY
9
-biphenyl
It has been shown that solution electron affinity can discrimi-
[
3
)
3
C-C12
H
9
9
3 3 9
) C-C12H ]
nate between materials of identical molecular weight for both
cases of isotopomers and geometrical isomers. Further, this
difference in solution electron affinities can be used to realize the
physical separation of these materials. To render the solution
electron affinity discrimination of practical value in an industrial
sense, it would necessary to engineer a continuous extractive or
distillation process, which would make use of the very large
differences in the vapor pressures or solubilities of the neutral
and anionic phases, as described above for the separations based
upon relative acidities.
9
3 3 9
) C-C12D
9
3 3 9
) C-C12D ].
9
and p-tert-
9
EXPERIMENTAL SECTION
P hysical Separations. The separations based upon solution
electron affinity differences were all carried out as described below
for a particular example. Carefully weighed portions of the tert-
-
4
butyl-d
9
-biphenyl (150 mg, 6.85 × 10 mol) and tert -butylbiphenyl-
-4
d
9
(150 mg, 6.85 × 10 mol) were placed in bulb A of the
apparatus shown in Figure 5. Tube B was charged with enough
argon-packed, oil-free, potassium metal to reduce approximately
half of the molecules in bulb A. Tube B was sealed at point 1,
and the entire apparatus was evacuated. A fresh potassium mirror
was deposited in bulb A by distillation from tube B. The tube B
was then removed from the apparatus at point 2. Approximately
3
2
0 mL of anhydrous THF, dried over NaK , was distilled into bulb
After removal of the THF under reduced pressure, freshly
distilled hexanes were added in order to remove any neutral
isotopomers (phase I) from the anion radicals (phase II). The
anion radical material (phase II) was reoxidized by introduction
A through the vacuum line. The stopcock was shut, and the
apparatus was removed from the vacuum line. The THF solution
was exposed to the potassium mirror (∼30 min.) until the mirror
was completely dissolved. A dark green solution of the anion
radical was formed immediately. The THF was then redistilled
out of the apparatus through the use of the vacuum line. Next,
2
of a molar excess of I dissolved in diethyl ether. Both phases
were submitted to 300-MHz proton NMR analysis.
The integration of the aromatic region shows the relative
1
0-15 mL of hexane was distilled into bulb A to dissolve the
amount of tert-butylbiphenyl-d
integration of the peaks between 1.3 and 1.7 ppm indicates the
relative amount of tert-butyl-d -biphenyl present. In comparison
to the original mixture of isotopomers (Figure 4), it is evident
that the neutral phase is depleted in tert-butylbiphenyl-d , while
the NMR spectrum of the reoxidized anion radical phase indicates
9
present. On the other hand,
neutral hydrocarbons. The hexane solution was then poured
through the porous frit of the apparatus. This washing procedure
was repeated two more times to ensure that all the neutral material
9
9
(
31) Stevenson, C. D.; Wehrmann, G. C.; Reiter, R. C. J. Phys. Chem. 1 9 9 1 , 95,
901.
Analytical Chemistry, Vol. 70, No. 18, September 15, 1998 3883

Fig u re 4 . (Upper) 1H NMR spectrum of the mixture of m- and p-tert-butylbiphenyls obtained from the Friedel-Crafts addition of tert-butyl
1
chloride to biphenyl-d9. (Middle) H NMR spectrum of the mixture of m- and p-tert-butylbiphenyls obtained from the Friedel-Crafts addition of
1
tert-butyl-d9 chloride to biphenyl. (Lower) H NMR spectrum of the mixture of Friedel-Crafts products used in the separation experiment.
had been separated from the anion radical. Bulbs A and C were
separated, and the anion radical phase (in bulb A) was reoxidized
However, the use of the Friedel-Crafts alkylation reaction with
isotopic compounds requires special precautions in order to avoid
exchange with the isotopic positions. The process is described
2
to the neutral state by addition of I in diethyl ether. The ether
solution was stirred in the presence of sodium thiosulfate and
gravity filtered. The filtered solution was concentrated and labeled
phase II. Both phases were submitted to GC and proton NMR
analysis.
Friedel-Crafts Alkylations. The alkylation of aromatic rings,
called Friedel-Crafts alkylation, is a reaction of very broad scope.16
below for the case of tert-butyl-d
The 250-mL round-bottom flask was charged with 1.100 g of
AlCl ‚6H O, and a stir bar was added. To the 50-mL round-bottom
flask was added 2.00 g (1.29 × 10 mol) of biphenyl. The bulb
was connected to the solid addition arm, and the entire apparatus
was evacuated immediately. A light flame was used to gently
9
-biphenyl.
3
2
-
2
3884 Analytical Chemistry, Vol. 70, No. 18, September 15, 1998

The residual materials were diluted with cyclohexane and filtered
under gravity. The hexanes were evaporated under reduced
pressures, and the residual material was subjected to vacuum
distillation. Approximately 900 mg of a colorless liquid was
collected between the temperatures of 105-112 °C at ∼0.01 Torr
(
17% yield). Mass spectral analysis revealed the liquid to be tert-
9
butyl-d -biphenyl. NMR analysis also revealed the compound to
be fully deuterated in the tert-butyl region (Figure 4). The
analogous procedure can also be used to produce the tert-
butylbiphenyl-d
Materials. The ethylbenzene-d
chased from Aldrich Chemical Co., and the ethyl-d
9
(Figure 4).
Fig u re 5 . Apparatus used to carry out the separations based upon
relative solution electron affinity.
5
and 18-crown-6 were pur-
-benzene was
5
purchased from Cambridge Isotope Laboratories. The perdeu-
terated biphenyl, biphenyl, perdeuterated 2-chloro-2-methylpro-
pane, and 2-chloro-2-methylpropane were purchased from Aldrich
Chemical Co. All of these materials were used without further
purification. Ultrapure, oil-free, potassium metal was purchased
from Aldrich Chemical Co.
Instrumentation. EPR spectra were recorded on a Bruker
ER-200 EPR spectrometer equipped with a Bruker ER-4111
variable-temperature controller. Mass spectral data were obtained
on a Hewlett-Packard 5790/ 5970 gas chromatograph/ mass spec-
trometer system containing a 30-m, 0.25-mm-i.d. capillary column
of methylphenyl silicone. NMR spectra were recorded on a Varian
Gemini 300 MHz spectrometer.
warm the round-bottom bulb with stirring of the aluminum
chloride. As the catalyst was heated, the powder would begin to
bump into the sides due to the molecules of water breaking from
3
their coordination sites with the AlCl . Carefully, heating was
continued until the catalyst stopped bumping. After a cooling
period, the biphenyl in the 50-mL round-bottom flask was poured
over into the 250-mL round-bottom flask. While under vacuum,
the flask was cooled with a dry ice/ acetone solution (-78 °C)
2
and 50 mL of CS was allowed to distill into the flask. After
distillation, the apparatus was isolated from the vacuum line, and
the solution was allowed to warm to room temperature. The
apparatus was then filled with nitrogen gas by use of the bubbler,
and positive pressure was maintained in the apparatus. Ap-
ACKNOWLEDGMENT
We thank the National Science Foundation (Grant CHE-
9617066)
-2
proximately 1.12 g (1.212 × 10 mol) of tert-butyl-d
added dropwise via syringe into the flask, through the rubber
septum, over a 1 h period. After addition of the tert-butyl-d
9
chloride was
9
Received for review February 25, 1998. Accepted July 8,
chloride, the solution was maintained at ambient temperature with
stirring for 10 h. The apparatus was reattached to the vacuum
1998.
line, and the CS
2
was completely redistilled out of the apparatus.
AC980221B
Analytical Chemistry, Vol. 70, No. 18, September 15, 1998 3885