Minimizing isobutylene emissions from large scale tert-butoxycarbonyl deprotections

Article abstract of DOI:10.1021/op049837v

Isobutylene off-gas amounts liberated during the methanesulfonic acid-catalyzed deprotection of N-BOC - pyrrolidine in THF, methanol, ethanol, 2-propanol, toluene, and dichloromethane were measured using on-line gas-phase mass spectroscopy. While one full

Full text of DOI:10.1021/op049837v

Organic Process Research & Development 2005, 9, 39−44
Minimizing Isobutylene Emissions from Large Scale tert-Butoxycarbonyl
Deprotections
Eric L. Dias,* Kevin W. Hettenbach, and David J. am Ende
Process Safety and Reaction Engineering Laboratory, Pfizer Global Research and DeVelopment,
Eastern Point Road, Groton, Connecticut 06340
Abstract:
with scavengers1,7,8 such as thiophenol1,7 to form an unre-
active byproduct, 3a; however, in an industrial setting, this
can be prohibitive in terms of cost, worker exposure, and
added purification.
Isobutylene off-gas amounts liberated during the methane-
sulfonic acid-catalyzed deprotection of N-BOC-pyrrolidine in
THF, methanol, ethanol, 2-propanol, toluene, and dichlo-
romethane were measured using on-line gas-phase mass spec-
troscopy. While one full equivalent of isobutylene was released
as an off-gas when THF was used as the reaction solvent,
emissions were reduced by 65-95% in other solvents. In alcohol
solvents, the corresponding alkyl tert-butyl ethers are formed
as byproducts of the reaction as expected. In dichloromethane
and toluene, oligomers of isobutylene can be formed under the
reaction conditions. These results provided the basis for
developing an effective acid/toluene scrubber for isobutylene
that was successfully employed on the pilot plant scale.
In the absence of a powerful scavenger, 1 can be trapped
in other manners. When trifluoroacetic acid is used, reaction
with the conjugate base can form the tert-butyl trifluoroac-
6,7,8a,9
etate ester, 3b;
however this typically will not occur
with methanesulfonic or other nonnucleophilic strong acids.
It has also been noted that water and alcohols will encourage
2a,11b
formation of tert-butyl alcohol
tert-butyl ethers, 3c,
or the corresponding alkyl
10,11
which may provide a practical
approach to reducing isobutylene emissions but will be
somewhat limited by potential functional group incompat-
ibilities.
Initially, we set out to measure isobutylene emissions from
BOC deprotections in different solvents with two goals. First,
we wanted to identify solvents that produce high isobutylene
emissions, and second, we wished to measure the isobutylene
trapping efficiency of different alcohol solvents. To this end,
we decided to use on-line mass spectrometry to indepen-
dently measure the amounts of both carbon dioxide and
Introduction
The tert-butoxycarbonyl (BOC) fragment is commonly
used throughout organic synthesis as a protecting group for
1,2
amines. The BOC group is typically removed under acidic
conditions1 in a wide variety of solvents, and one potential
-3
2,4
organic byproduct of this process is isobutylene, which is
a VOC subject to regulation by the EPA. Because iso-
5
12
isobutylene released from the reaction in real-time, since
butylene has a boiling point (bp) of -6.9 °C, allowing it to
pass through most process condensers uncontrolled and vent
to the atmosphere in the absence of suitable end-of-line
devices, emissions must be dealt with accordingly before a
process can be run on a manufacturing scale.
t
Scheme 1. Byproducts of BOC deprotection
The generally accepted mechanism for BOC deprotection
is shown below (Scheme 1):2 Protonation of the carbonyl
oxygen of the tert-butyl carbamate results in degradation to
initially produce the unprotected amine, carbon dioxide, and
,6
2
a,6,7
the highly reactive tert-butyl cation, 1,
which can
decompose to isobutylene, 2,2 in the absence of suitable
,4
trapping reagents. On laboratory scale, 1 is often trapped
*
To whom correspondence should be addressed. Email: eric_l_dias@
groton.pfizer.com.
(
1) Greene, T. W.; Wuts, P. G. M. ProtectiVe Groups in Organic Synthesis,
rd ed.; Wiley: New York, 1999; pp 518-525.
3
(2) (a) Lawrence, S. A. Chimica Oggi 1999, 15-20. (b) Carpino, L. A. Acc.
Chem. Res. 1973, 6, 191-198. (c) Carpino, L. A. J. Am. Chem. Soc. 1957,
7
9, 98-101.
3) Strazzolini, P.; Melloni, T.; Giumanini, A. G. Tetrahedron 2001, 57, 9033-
043.
(8) (a) Bodanszky, M.; Bodanszky, A. Int. J. Pept. Protein Res. 1984, 23, 565-
572. (b) Masui, Y.; Chino, N.; Sakakibara, S. Bull. Chem. Soc. Jpn. 1980,
53, 464-468.
(
(
9
4) Kaiser, E., Sr.; Tam, J. P.; Kubiak, T. M.; Merrifield, R. B. Tetrahedron
Lett. 1988, 29, 303-306.
5) C. F. R. 51.100 (s).
(9) Latremouille, G. A.; Eastham, A. M. Can. J. Chem. 1967, 45, 11-16.
(10) (a) Evans, T. W.; Edlund, K. R. Ind. Eng. Chem. 1936, 28, 1186-1188.
(b) Sola, L.; Pericas, M. A.; Cunill, F.; Izquierdo, J. F. Ind. Eng. Chem.
Res. 1997, 36, 2012-2018.
(11) (a) Di Girolamo, M. et al. Ind. Eng. Chem. Res. 1997, 36, 4452-4458. (b)
Tejero, J.; Calderon, A.; Cunill, F.; Izauierdo, J. F.; Iborra, M. React. Funct.
Polym. 1997, 33, 201-209.
(
(
6) Losse, Gl; Zeidler, D.; Grieshaber, T. Liebigs Ann. Chem. 1968, 715, 196-
2
03.
(7) Lundt, B. F.; Johansen, N. L.; Volune, A.; Markussen, J. Int. J. Pept. Protein
Res. 1978, 12, 258-268.
1
0.1021/op049837v CCC: $30.25 © 2005 American Chemical Society
Vol. 9, No. 1, 2005 / Organic Process Research & Development
•
39
Published on Web 12/15/2004

the rate of carbon dioxide production also provides a
convenient measure of reaction rate and completion.
In this study, we report a method for quantifying absolute
amounts of isobutylene and carbon dioxide released during
BOC-deprotections in a variety of solvents using on-line gas
phase mass spectrometry as the analytical method. Based
on these results, a general solution for externally scrubbing
isobutylene from reaction off-gas streams with greater than
result, it was often observed that the isobutylene signal (m/z
) 56) would not return to baseline when the reaction was
performed in alcohol solvents, since isobutylene itself is a
mass fragment of the tert-butyl ethers. To obtain accurate,
quantitative results, the overlap at m/z ) 56 was corrected
by using the unique signals for each ether, e.g., the relative
intensities of the signals at 56 and 73 for pure MTBE were
used to adjust for the signal overlap with isobutylene and
correct the baseline.
99% efficiency was developed and employed successfully
on the pilot plant scale.
One point worth mentioning is that a lag time between
the carbon dioxide and isobutylene flow profiles was
observed in some cases. For example, the flow and total off-
gas profiles for the reaction performed in THF are shown
below (Figure 1). While the carbon dioxide flow profile for
the reaction in THF returns to baseline in approximately 30
min, it takes several hours for the isobutylene emission to
be complete. Interestingly, this effect is not observed in the
case of alcohols. One possible explanation for this behavior
is that isobutylene has a limited, although significant,
solubility in THF. Because the reaction is performed under
flowing helium, the isobutylene in the headspace is constantly
being removed, which slowly drives the remaining isobu-
tylene out of solution.
Results and Discussion
Measurement of Isobutylene Off-Gas from Common
Solvents. Using the methodology described in the Experi-
mental Section, carbon dioxide and isobutylene off-gas
results were obtained for the methanesulfonic acid (MSA)
catalyzed deprotection of N-BOC-pyrrolidine, 4, in several
solvents (Scheme 2) using on-line gas-phase mass spectrom-
1
3
etry. The results are summarized in Table 1.
Scheme 2. Model deprotection reaction
Table 1. Summary of off-gas results
temp
(°C)
isobutylene
(equiv)
CO
2
(equiv)
solvent
THF
toluene
dichloromethane
methanol
ethanol
45
45
20
20
40
45
1.12
0.25
0.15
0.10
0.29
0.31
1.07
0.91
0.99
0.97
1.08
0.94
2-propanol
From Table 1, it is immediately apparent that isobutylene
emissions are greatest when THF is used as the solvent.
Emissions from methanol are the lowest (0.10 equiv, Table
2 2
Figure 1. CO and isobutylene off-gas profiles. CO (dashed
1
), presumably due to formations of methyl tert-butyl ether
lines) and isobutylene (solid lines) flow and total off-gas profiles
for the BOC deprotection (Scheme 2) in THF at 45 °C. The
total off-gas profiles were generated by integration of the
corresponding flow profiles, using the ideal gas law to obtain
molar equivalents.
10,11a
(MTBE).
Isobutylene emissions from ethanol and 2-pro-
panol are approximately 3 times higher (0.29 equiv and 0.31
10
equiv, respectively, Table 1), likely because of steric effects.
Alcohols with less steric bulk are expected to favor nucleo-
philic substitution over elimination, resulting in higher ratios
of the corresponding alkyl tert-butyl ether to isobutylene.
Because these reactions were performed with a constant
stream of flowing helium as a reference gas, often at elevated
temperatures to increase the reaction rate (the measured
amount of isobutylene was relatively unaffected by temper-
ature), the tert-butyl ethers that form in alcohol solvents could
not be accurately quantified by gas chromatography due to
evaporative losses. Evidence for this was observed in the
mass spectra, as unique signals arising from the ethers were
This effect is observed for the reactions performed in
dichloromethane and toluene as well, although to a much
lesser extent. When taken in conjunction with the fact that
isobutylene emissions from dichloromethane and toluene
were relatively low (Table 1), it became apparent that another
process was also occurring. In the case of toluene, formation
of at least some 4-tert-butyltoluene was expected;8
however, GC analysis indicated that this accounted for at
most 5% of the isobutylene. Instead, it was determined that,
under the reaction conditions, the cationic polymerization
a,14
i
detected (m/z ) 73 for MTBE; 59 for ETBE, PTBE). As a
1
1,15
of isobutylene
to form nonvolatile oligomers was in fact
(
12) (a) Walsh, M. R.; LaPack, M. A. ISA Trans. 1995, 67-85. (b) Am Ende,
D. J. et al. Org. Process Res. DeV. 2000, 4, 587-593. (c) Hettenbach, K.
et al. Org Process Res. DeV. 2002, 6, 407-415.
16
responsible for the observed low emissions (Scheme 3).
(13) Hettenbach, K.; am Ende, D. J.; Leeman, K.; Dias, E. et al. Org. Process
Res. DeV. 2002, 6, 407-415.
(14) (a) Schlatter, M. J.; Clark, R. D. J. Am. Chem. Soc. 1953, 75, 361-369. (b)
Gekhtman, B. N. et al. Nefteper. I Neftekhimiya 1974, 11, 80-82.
40
•
Vol. 9, No. 1, 2005 / Organic Process Research & Development

Scheme 3. Oligomerization of isobutylene
Table 2. Laboratory scrubber results
scrubber
run
no.
scrubber
medium
amount
(g)
amount isobutylene efficiency
acid
MSA
(g)
(g)
(%)
a
1
2
3
4
5
6
methanol
198
350
225
225
225
8
10
1.5
1.5
1.5
1.5
0
0
0
99.6
99.6
0
a
a
a
methanol
toluene
toluene
toluene
concd HCl
MSA
MSA
8
8
8
b
36
1.5
a
toluene/methanol 173/40 MSA
a
Based on isobutylene flow ) 10 sccm over 60 min. b Based on isobutylene
flow ) 10 sccm over 24 h.
Using dichloromethane as the solvent, this was verified
by using a standard workup procedure to isolate the organic
soluble matter from the reaction mixture. The pyrrolidine-
MSA salt was extracted with water, and the organic layer
Figure 2. Isobutylene scrubbing experiment with 225 g of
toluene and 8 g of methanesulfonic acid (from Table 2, run 4).
The isobutylene flow was set at 10 sccm and the helium flow
was set at 100 sccm for this experiment. The response ratios
for isobutylene/helium (56/4) at the scrubber inlet (solid line)
and outlet (dashed line) are shown. Integration of the two
profiles gives a scrubbing efficiency of 99.6%.
3
was washed with aq. NaHCO to remove the remaining
MSA. The organic layer was dried and concentrated, leaving
a clear liquid that accounted for 97% of the isobutylene that
was not evolved from the reaction. Combustion analysis and
IR spectroscopy are consistent with the liquid being isobut-
ylene oligomers (see Experimental Section).
Development of an Isobutylene Scrubbing System. As
described above, appropriate choice of solvent can provide
a primary method for reducing isobutylene emissions during
BOC deprotections; however, it is recognized that there will
be instances where solvents such as THF must be used for
optimal reaction conditions. In the absence of a suitable end-
of-line device, the use of an external scrubber will be
necessary to ensure that isobutylene emissions will meet the
environmental regulatory limits. Furthermore, an efficient
external scrubbing system is a generally applicable solution
for all BOC deprotections, regardless of solvent.
Based on the above data, the most promising scrubbing
systems appeared to be either methanol/acid to form MTBE
or toluene/acid to form poly(isobutylene); dichloromethane
was not considered due to its relatively low boiling point.
Taking into account the limitations of the pilot plant gas
scrubbing system, the laboratory studies were all performed
at room temperature, yielding the following results (Table
no reaction with isobutylene to form MTBE was observed
in either case.
Runs 3, 4, and 5 were all performed using toluene as the
solvent. Run 3 was performed as a control with no acid
present to catalyze oligomerization, and there was no
observable scrubbing of isobutylene as expected. Run 4 was
the initial run using methanesulfonic acid (MSA), which
formed a dispersion due to its low miscibility with toluene.
Despite this drawback, greater than 99% isobutylene removal
from the vapor stream was observed (Figure 2). In Figure 2,
the response ratio for isobutylene to the argon reference gas
13
is plotted at the scrubber inlet and the scrubber outlet, and
it can be easily seen that very little isobutylene is present
after the vapor stream has passed through the scrubbing
medium. Weighing of the scrubber bottle after the experiment
indicated that the mass of the solution had increased by
approximately 1.45 g, indicating good agreement with the
1
7
theoretical input. Run 5 was an experiment to determine
the amount of isobutylene that could be taken up by the
scrubber medium before efficiency was reduced, if at all.
After 24 h, approximately 36 g of isobutylene were intro-
duced into the scrubber with little or no reduction in
2).
Runs 1 and 2 were performed using methanol as the
solvent with methanesulfonic acid and concentrated HCl as
the acid source, respectively. Under the scrubbing conditions,
1
8
scrubbing efficiency.
Run 6 was an experiment to determine if methanol could
be added to the toluene/MSA solution to increase miscibility.
(
15) (a) Schmerling, L.; Ipatieff, V. N. AdVances in Catalysis and Related
Subjects; Wiley: New York, 1975; Vol. 2, p 86. (b) Kennedy, J. P. Cationic
Polymerization of Olefins: A Critical InVentory; Wiley: New York, 1975;
p 86. (c) Higashimura, T.; Miyoshi, Y.; Hasegawa, H. J. Appl. Polym. Sci.
(17) Because there is no condenser at the scrubber outlet, we cannot account for
any toluene that may be lost during the experiment.
1
983, 28, 241-251.
16) When the deprotection in toluene was performed at elevated temperatures
70-80 °C), unique signals in the mass spectrum corresponding to
diisobutylene were detected.
(18) It should be mentioned that, as expected, a substantial increase in the volume
of the scrubber solution was noted during the course of this experiment,
which had to be terminated to avoid overflow of the solution into the gas
lines.
(
(
Vol. 9, No. 1, 2005 / Organic Process Research & Development
•
41

Figure 4. Mass spectrometry profiles for the pilot plant
scrubber inlet (solid line) and outlet (dashed line). Isobutylene
off-gas is observed at the scrubber inlet starting at 2 h when
Figure 3. Toluene/MSA scrubber efficiency vs amount of
MSA. Keeping the same amount of toluene (225 g), the scrubber
efficiency is shown to decrease with decreasing amounts of
methanesulfonic acid. At 1.2 g of MSA, the scrubbing efficiency
falls to zero.
2
concentrated HCl is added to the reaction. At 14 h, the N sweep
is increased to purge the remaining gases from the reactor
headspace. Integration of the profiles shows 97% scrubber
efficiency.
Using what was determined to be a minimum amount of
methanol, the results were disappointing as the scrubbing
efficiency was completely destroyed.
outcome for the first attempt to employ this novel scrubber
on scale.
Additional experiments were performed to determine the
minimum concentration of MSA required to maintain good
scrubbing efficiency (Figure 3). As shown in Figure 3,
halving the amount of MSA from 8 g (Table 2, Run 4) to 4
g resulted in a small decrease in scrubbing efficiency from
greater than 99% to around 97%. When the amount of acid
was reduced to 2.0 g, the scrubbing efficiency was further
eroded to 80%. Further reduction of the acid to 1.2 and 0.8
g, however, completely eliminated the scrubbing efficiency,
indicating that there is likely a critical concentration in the
range 1.2-1.7 g of MSA per 225 g toluene that is required
for the oligomerization reaction to occur under these condi-
tions.
Isobutylene Scrubbing on the Pilot Plant Scale. Re-
cently, a reaction was scaled-up in the Pfizer Groton Pilot
Plant facility in which two BOC groups were removed in a
single step using THF as the solvent. To reduce isobutylene
emissions, an external toluene/MSA scrubber using the
conditions from Table 2, Run 4 was employed. The reaction
was run in 3 × 500 gallon batches, and the scrubbing solution
was not changed between runs. Before the third and final
batch, the mass spectrometer was interfaced with the pilot
plant scrubber train so that we could measure the effective-
ness of this scrubber on scale. Using house nitrogen as the
carrier and reference gas, the data in Figure 4 were collected.
Because the house nitrogen flow fluctuated throughout
the course of the experiment, the data in Figure 4 is noisier
than that obtained on the laboratory scale (Figure 2). This
does not impact the results, however, since the flow is equal
at both the scrubber inlet and outlet. From these data, we
can estimate that approximately 97% scrubbing efficiency
was achieved on scale for this run. While these results are
not as exceptional as the >99% efficiency achieved in the
laboratory, possibly due to different mixing of the MSA and
toluene within the pilot plant scrubber system, the 97%
reduction in isobutylene emissions is viewed as a successful
Conclusions
By performing calibrations with isobutylene and carbon
dioxide using helium as a reference gas, we have demon-
strated that gas-phase mass spectrometry can be used as an
on-line analytical tool to measure the rate at which these
off-gases are liberated during BOC deprotections. Simple
integration of the resulting flow profiles produced off-gas
totals, which were used to determine the effectiveness of
common organic solvents at reducing isobutylene emissions.
It was determined that in THF all of the isobutylene is
liberated as a reaction off-gas, such that external scrubbing
will be required if this solvent is used on a manufacturing
scale. In alcohol solvents, the tert-butyl cation is partially
trapped as the corresponding alkyl tert-butyl ethers. Methanol
is most efficient for this purpose showing a 90% trapping
efficiency, while ethanol and 2-propanol are not as efficient
1
0
likely due to steric effects. Surprisingly, isobutylene
emissions from both dichloromethane and toluene were found
to be relatively low as well due to the propensity of
isobutylene to polymerize under the reaction conditions to
1
5
produce nonvolatile oligomers.
One question that naturally arises is why isobutylene
oligomerizes in dichloromethane and toluene but not in THF.
Two reasons for this behavior are offered at this time: (1)
THF is a Lewis base which will favor E1 elimination of the
tert-butyl cation to form isobutylene and (2) the acidity of
methanesulfonic acid is reduced in THF, as the solution will
only be as acidic as the corresponding oxonium ion.9
Evidence for reason (1) is provided by the fact that both
ethanol and 2-propanol exhibit higher isobutylene emissions
than toluene and dichloromethane, while methanol shows
1
0
the greatest trapping efficiency. However, the fact that in
our laboratory scrubbing studies the methanol/MSA solution
showed zero trapping efficiency suggests not only that reason
(2) may contribute as well but also that during a BOC
42
•
Vol. 9, No. 1, 2005 / Organic Process Research & Development

deprotection reaction it is the initially released tert-butyl
cation that reacts with methanol to form MTBE, and not one
formed from direct protonation of isobutylene. This argument
is further supported by the fact that, in our laboratory
scrubbing studies, a toluene/MSA solution provided >99%
scrubbing efficiency; however, the introduction of a mini-
mum amount of methanol to promote miscibility reduced
this to zero. Overall, these results suggest that, in THF, the
equilibrium between isobutylene and the tert-butyl cation
may not be favored enough to promote oligomerization under
the reaction conditions, and since the initially formed tert-
butyl cation cannot react with THF in the manner that it does
with alcohols, the only available route is elimination to form
isobutylene which is inevitably released as an off-gas.
On the pilot plant scale, the toluene/MSA scrubber
developed in the laboratory exhibited 97% efficiency for the
removal of isobutylene from the reactor effluent stream.
While this is successful, it is not as efficient as the >99%
efficiency that was observed on the laboratory scale. It is
suggested that this may be due to differences in mixing of
the relatively immiscible MSA and toluene in the two
different scrubbing systems and may be supported by the
fact that on the laboratory scale a mere 6.5-fold reduction
in the amount of MSA reduces the scrubbing efficiency to
zero. In other words, if the oligomerization reaction is that
sensitive to the concentration of acid, incomplete or even
different degrees of dispersion may lead to the observed
differences in scrubbing efficiency.
dosing port. A constant flow of helium was supplied as the
carrier gas at the inlet. The outlet line was split, with one
line going to the mass spectrometer and the other to an oil
bubbler to ensure that a helium atmosphere is maintained
throughout the system.
Figure 5. Reactor setup for off-gas analysis of BOC depro-
tections.
Calibration curves were generated using known flow
ratios of isobutylene and carbon dioxide relative to the helium
carrier gas (Figure 6). In Figure 6, the response ratio for the
isobutylene ion peak at m/z ) 56 relative to the helium ion
peak are plotted as a function of the input flow ratio. The
slope of the resulting linear fit provides a calibration factor
that relates the ion response ratio to the flow ratio. By using
the helium carrier gas with constant flow as an internal
standard, the flow profiles for carbon dioxide and isobutylene
can then be constructed.
In conclusion, we have demonstrated that isobutylene
emissions from BOC deprotection reactions can be heavily
influenced by choice of solvent, which should be the first
consideration when a reaction is being optimized for produc-
tion scale. In addition, we have developed a novel scrubber
that takes advantage of the fact that, under the right
conditions, isobutylene reacts with itself to form nonvolatile
products. In addition to being chemically efficient, this
scrubbing system may also be economically efficient. Since
there is no need for a stoichiometric trapping reagent, the
scrubbing solution can be recycled and utilized for several
batches with little reduction in isobutylene removal ef-
ficiency.
Experimental Section
Materials and Methods. N-BOC pyrrrolidine and meth-
anesulfonic acid were purchased from Aldrich and used
without further purification. Matheson flow controllers and
a Matheson gas mixing system were used for delivery of
helium, carbon dioxide, and isobutylene at constant flow
rates. An Amatek Dycor ProMaxion multiport mass spec-
trometer was used to obtain gas-phase measurements. FT-
IR spectra were collected using an MT Autochem ReactIR
Figure 6. Isobutylene response calibration plot. The response
ratio for isobutylene (m/z ) 56)-helium (m/z ) 56) is plotted
versus the gas input flow ratio (O) with helium at a constant
flow rate of 5 sccm. Linear regression (solid line, R ) 0.999 27)
provides the calibration factor of 2.597 in this instance.
Sample Calculation. To demonstrate the manner in which
the flow profiles were obtained, the following sample
calculation quantifying the carbon dioxide off-gas for the
BOC deprotection run in methanol at 20 °C is presented.
First, the response ratio profile for the carbon dioxide ion
peak (m/z ) 44) relative to the helium ion peak (m/z ) 4) is
plotted. Dividing the response ratio by the calibration factor
of 4.535 provides the flow ratio profile for carbon dioxide
relative to helium (Figure 7).
4000 with a DiComp ATR probe. Microsoft Excel and
Kaleidagraph for Windows were used for all of the data
analysis.
Measurement of Isobutylene and Carbon Dioxide Off-
Gas. The diagram in Figure 5 details the experimental setup
used for our quantitative off-gas analysis. The reaction was
performed in a pressure-tight HEL Automate Hastelloy vessel
containing one gas inlet port, one gas outlet port, and one
Because the flow rate of helium is fixed at 5 sccm (sccm
) standard cubic centimeters per minute), multiplying the
Vol. 9, No. 1, 2005 / Organic Process Research & Development
•
43

Figure 9. Apparatus for isobutylene scrubber experiments.
Figure 7. Response ratio (solid line) and flow ratio (dashed
line) for carbon dioxide (44) relative to helium (4) for the
reaction performed in methanol at 20 °C.
dichloromethane, the solution was worked up as follows. The
reaction mixture was first extracted with water (3 × 25 mL)
to remove the pyrrolidine-MSA salt. The organic layer was
flow ratio by 5 gives us the actual carbon dioxide flow profile
in sccm (Figure 8). Integration of the flow profile with
respect to time gives us the total amount of carbon dioxide
liberated from the reaction, which is 316 mL (1 scc ) 1 mL
at STP) (Figure 8).
then washed with aq. NaHCO
residual MSA, collected, and dried over Na
3
(2 × 25 mL) to remove
2
SO . After
4
filtration, the solution was concentrated to dryness by rotary
evaporation at or below ambient temperature to minimize
the loss of diisobutylene (bp ) 101-102 °C), producing
0
.814 g of a clear, colorless liquid (83% yield). Anal.
Calcd: C, 85.63%; H, 14.37% based on isobutylene.
Found: C, 85.50%; H, 14.49%; N < 0.10%. FT-IR (neat):
-
1
3
076 cm (w, CdCsH), 3025-2760 (vs, R
3
CsH), 1635
(
w, CdC), 1479 and 1464 (s, R CsH, R CsH
3
2
2
), 1364 (vs,
3
RCsH ), 892 (m, CdCsH).
Isobutylene Scrubber Studies. A 250 mL glass scrubber
bottle equipped with a gas dispersion frit and magnetic stir
bar were employed (Figure 9). Helium and isobutylene were
introduced through the scrubber inlet tube connected to the
dispersion frit. The outlet tube was connected to an oil
bubbler to maintain gastight conditions. The gas flow line
was split immediately before and after the scrubber bottle
so that both the inlet and outlet streams could be monitored
simultaneously by mass spectroscopy.
Typically, 200 mL of solvent and 8 g of methanesulfonic
acid were employed in the scrubber studies with magnetic
stirring. The apparatus was purged with helium at a constant
flow of 15-100 sccm until steady baseline readings were
achieved at the scrubber inlet and outlet. Isobutylene was
introduced as a square wave with a constant flow of 10-50
sccm for approximately 30 min.
Figure 8. Flow profile (solid line) and total (dashed line) for
carbon dioxide off-gas from the reaction performed in methanol
at 20 °C.
Based upon 2.5 g of starting material, the expected amount
of CO liberated is shown in eq 1:
2
1
mol
22.4 L
1 mol
2
.5 g × 1
×
) 0.327 L ) 327 scc
(1)
71.24 g
Therefore, the total amount of CO
2
liberated, as obtained
The response ratio profile for isobutylene to helium as a
function of time at the scrubber inlet and outlet can be used
directly to obtain scrubbing efficiency. By simply integrating
the area under both profiles, the efficiency can be calculated
according to eq 2.
from the mass spectroscopy data, is 316/327 ) 0.97 equiv.
DeprotectionofN-BOC-Pyrrolidine.Typically,N-BOC-
pyrrolidine (2.5 g, 2.6 mL, 14.6 mmol, 1 equiv) and 25 mL
of solvent were charged to the reactor and brought to
temperature. The vessel was then purged with a constant
helium flow of 5 sccm until constant baseline readings for
the atmosphere were obtained on the mass spectrometer.
Methanesulfonic acid (7.0 g, 4.7 mL, 73.0 mmol, 5 equiv)
was introduced via syringe through the dosing port, which
was kept sealed with a PTFE coated silicone rubber septum
seated above a stainless steel ball valve. Upon addition,
vigorous carbon dioxide evolution was often observed at the
Areaoutlet
Efficiency ) 1 -
(2)
Areainlet
Acknowledgment
We thank Angela Cady and Matthew Jorgensen for
analytical support during the course of this investigation.
2
oil bubbler, verified by an increase in the CO signal at the
mass spectrometer.
Isolation of Isobutylene Oligomers from Dichloro-
methane. After performing the deprotection reaction in
Received for review August 27, 2004.
OP049837V
44
•
Vol. 9, No. 1, 2005 / Organic Process Research & Development