Preparation of ABNO on Scale and Analysis by Quantitative Paramagnetic NMR

Article abstract of DOI:10.1021/acs.oprd.8b00191

A practical, safe, and scalable synthesis of the stable nitro-oxide radical catalyst ABNO was developed. This process is chromatography-free and avoids the Wolff-Kishner reduction. ¹H NMR data for this paramagnetic compound were obtained that allowed an assessment of its chemical purity. Impact sensitivity test data for solid ABNO are also reported.

Full text of DOI:10.1021/acs.oprd.8b00191

Subscriber access provided by University of South Dakota
Full Paper
Preparation of ABNO on Scale and Analysis
by Quantitative Paramagnetic NMR
Zhiguo Jake Song, Guoyue Zhou, Ryan D. Cohen, and Lushi Tan
Org. Process Res. Dev., Just Accepted Manuscript • DOI: 10.1021/acs.oprd.8b00191 • Publication Date (Web): 13 Aug 2018
Downloaded from http://pubs.acs.org on August 13, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 18
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Preparation of ABNO on Scale and Analysis by
Quantitative Paramagnetic NMR
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
†
Zhiguo J. Song*, Guoyue Zhou , Ryan Cohen, Lushi Tan
Department of Process R & D, Merck & Co. Inc., P.O. Box 2000, Rahway NJ 07065, USA
†
Process R&D, WuXi AppTec Co., Ltd., 288 Fute Zhong Road, Shanghai 200131, China
ACS Paragon Plus Environment
1

Organic Process Research & Development
Page 2 of 18
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table of Contents Graphic
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ACS Paragon Plus Environment
2

Page 3 of 18
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
ABSTRACT
A practical, safe, and scalable synthesis of the stable nitroꢀoxide radical catalyst, ABNO, was
developed. This process is chromatography free and avoids the WolffꢀKishner reduction. Proton
NMR of this paramagnetic compound was obtained that allows assessment of its chemical purity.
Impact sensitivity test data of solid ABNO is also reported.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
KEYWORDS
ABNO, ketoꢀABNO, nitroxyl radical, paramagnetic NMR
ACS Paragon Plus Environment
3

Organic Process Research & Development
Page 4 of 18
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
TEMPO ((2,2,6,6ꢀtetramethylpiperidinꢀ1ꢀyl)oxyl) and related nitroꢀoxide radicals are an
important class of catalysts for the oxidation of alcohols. Aerobic oxidation of alcohols to
ketones and aldehydes is particularly attractive for large scale reactions because of the potential
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
to reduce cost and minimize waste. ABNO (9ꢀazabicyclo[3.3.1]nonaneꢀNꢀoxyl) and ketoꢀABNO
1,2
(9ꢀazabicyclo[3.3.1]nonaneꢀ3ꢀoneꢀNꢀoxyl) are the most effective catalysts for such reactions.
Typically, 1ꢀ10 mol% of the catalyst is employed, which would represent significant quantities
on scale. Due to their rather limited commercial availability, we were required to prepare
kilogram quantities to support our ongoing efforts to prepare drug candidates.
We initially evaluated the literature procedure for preparation of these compounds, which is
1
shown in scheme 1. The shortest route to access ABNO and ketoꢀABNO is through compound
3
which can be made directly from 1 and 2 via Mannich reaction followed by decarboxylation.
KetoꢀABNO is made from compound 3 in one step by oxidation. ABNO is made from
compound 3 after deoxygenation of the ketone via WolffꢀKishner reduction followed by a
similar oxidation step. While these sequences were reported for the preparation of gram
quantities of ABNO and ketoꢀABNO, direct scaleꢀup to kilogram quantities pose serious process
challenges and grave processꢀsafety concerns. Herein, we report significant modifications to the
synthetic sequence and process parameters that make the process readily amenable for scale up.
We also developed analytical methods for determining the purity of the final products, which are
challenging to characterize by conventional methods. Operational safety for handling ABNO
and ketoꢀABNO will also be discussed.
Scheme 1.
ACS Paragon Plus Environment
4

Page 5 of 18
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Evaluation of the literature procedure revealed several deficiencies that made it impractical to
1
run on scale. The Mannich reaction between compounds 1, 2, and ammonia was performed in
water, and the product is highly water soluble and thus cannot be easily extracted into organic
solvents for isolation and purification. In the WolffꢀKishner reduction sequence, a very high
o
temperature (220 C) was required, which in the presence of excess hydrazine was unsafe, and
1
,3
the extreme water solubility of the hydrazone intermediate made it impractical to isolate. For
the oxidations of amines 3 and 5 to ketoꢀABNO and ABNO, respectively, we observed that the
use of acetonitrile as reaction solvent following the reported procedure resulted in incomplete
conversion. Finally, practical crystallization procedures for purification of ABNO and ketoꢀ
ABNO besides chromatography or sublimation were needed.
Scheme 2.
ACS Paragon Plus Environment
5

Organic Process Research & Development
Page 6 of 18
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The need for purification of compound 3 was driven by the fact that very often diacid 1 is prone
to decomposition so the reaction to form 3 from 1 and 2 can form a substantial amount of tar.
As an alternative to making compound 3 directly from 1, 2, and ammonia, the reaction of
benzylamine with 1 and 2 gave compound 4, which can be extracted into organic solvents and
offers purification opportunities. Compound 4 can then be deprotected under hydrogenolysis
1
b
conditions to give 3. To avoid the use of a large excess of hydrazine at high temperature in the
WolffꢀKishner reduction, we evaluated different routes for deoxygenation of ketones 3 and 4.
1b
Ketone 4 can be readily reduced to alcohol 6 with sodium borohydride in quantitative yield.
We then found that alcohol 6 can be converted to mesylate 7 and subsequently transformed into
olefin 8 with sodium tꢀbutoxide. More directly, alcohol 6 can be dehydrated with sulfuric acid to
olefin 8. Under hydrogenation conditions, the olefin was reduced with simultaneous benzyl
group deprotection, cleanly generating compound 5. This new threeꢀstep sequence from
compound 4 to 5 via 8 has the same number of chemical transformations as the previous method
but is more economical, much simpler to execute, avoids the hazardous use of hydrazine, and is
more amenable for large scale. The details of each step of the new route will be discussed in the
following sections.
1
b
Preparation of compound 4: In the reported literature procedure, benzylamine hydrochloride
was used in the reaction with 1 and 2 in water in the presence of sodium acetate buffer. After
ACS Paragon Plus Environment
6

Page 7 of 18
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
initial condensation reaction at room temperature, the reaction mixture was heated for
decarboxylation. We evaluated the combination of benzylamine with different acids such as
acetic acid, sulfuric acid, phosphoric acid; only sulfuric acid gave comparable results to the HCl
salt, while other acids gave inferior reaction profiles. We chose sulfuric acid over hydrochloric
acid to minimize the halideꢀcontaining waste. The reaction solvent was changed to a mixture of
water and methanol to alleviate the agitation problem caused by the tar formed during the
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
o
o
reaction. The maximum reaction temperature was also reduced from the reported 50 C to 30 C
to minimize tar formation and to alleviate safety concerns for the decarboxylation step (CO2
evolution). After reaction and adjustment to low pH, the reaction mixture was extracted with
MTBE to remove any organic soluble impurities, and the product was extracted into heptane
under basic conditions. We found this crude product mixture was much more stable in heptane
than in MTBE, although, pure 4 was stable in both solvents. This heptane solution was filtered
through a small amount of silica gel and the solvent distilled off to give the crude 4, which was
used in the next step without purification. The purity at this stage was typically 90ꢀ95% by
HPLC and the isolated yield was 53%. Although the yield is modest, this process has proven to
be reproducible at over 100 kg scale.
Preparation of compound 5 through compounds 6 and 8: The sodium borohydride reduction
of 4 in methanol was straightforward and resulted in essentially quantitative yield. This alcohol
6
was extracted into MTBE from the reaction after acid quench and basification. The crude
product was used in the subsequent elimination directly after removal of solvent. Treatment of
o
alcohol 6 with 70% sulfuric acid at 100 C resulted in a clean elimination to the desired olefin 8.
After pH adjustment, the product was extracted into ethyl acetate. The assay yield and purity by
HPLC was 97%. The crude product was solvent switched into isopropanol for hydrogenation
ACS Paragon Plus Environment
7

Organic Process Research & Development
Page 8 of 18
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
using palladium hydroxide on carbon as catalyst to give 5 in high yield and purity. Product 5 is
fairly volatile and some loss of product to the distillates may occur if distillation of solvent is
required.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Preparation of ABNO: The reported method used acetonitrile as solvent for oxidation of 5 with
1b
urea hydrogen peroxide and catalytic Na WO •2H O. Under these conditions, we observed
2
4
2
sluggish conversion and a large excess of urea hydrogen peroxide was required to drive the
reaction to completion. It was also observed that Na WO •2H O was not completely soluble.
2
4
2
This prompted us to investigate mixtures of water and organic solvents such as methanol,
isopropanol, and acetone. Use of methanol resulted in up to 30% of the formamide of 5, but in
water/isopropanol or water/acetone mixtures, the reaction was clean and proceeded to
completion. We chose a mixture of isopropanol and water because isopropanol was used in the
previous hydrogenation step. It is critical to charge the urea hydrogen peroxide complex to the
reaction over time to control decomposition and associated release of oxygen gas in this step.
This slow, controlled addition also allowed for more efficient use of the hydrogen peroxide and
thus better conversion. After the reaction, the product was extracted into dichloromethane and
crystallized from dichloromethane and heptane to ABNO in 99.5% GC area percent purity and
70% isolated yield.
Preparation of keto-ABNO: Compound 4 was hydrogenated over palladium on carbon in
methanol to cleave the benzyl protecting group to give compound 3. The same oxidation
procedure was used in the preparation of ketoꢀABNO from compound 3, except the reaction
solvent was mostly water with low levels of methanol that was carried over from the previous
step. The same workup and crystallization from dichloromethane/heptane gave ketoꢀABNO in
6
7% isolated yield and 98.5% HPLC area percent purity.
ACS Paragon Plus Environment
8

Page 9 of 18
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
Safety considerations for the handling of ABNO and keto-ABNO: Both compounds exhibit a
o
large exotherm on DSC with an onset temperature of ca. 80 C. The exotherms for ABNO and
ketoꢀABNO are approximately 960 J/g and 1100 J/g, respectively, over a temperature range of
o
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
8
0ꢀ220 C. These are potentially large exotherms at rather low onset temperatures, which are
serious concerns from a process safety perspective. Shock sensitivity evaluation indicated that
solid ABNO gave positive fallꢀhammer results (explosions) in 3 out of 8 drops at an impact
energy level of 30 J, but 0 out of 8 drops at 20 J. ABNO solid is thus considered potentially
impact sensitive so proper precaution is required when handling solid ABNO. The fallꢀhammer
test result for ketoꢀABNO is 0 out of 8 positive at 50 J, so ketoꢀABNO solid is not considered to
be impact sensitive for normal handling. Although exotherms measured on DSC for the process
solution streams containing ABNO and ketoꢀABNO are very modest, process temperature was
o
kept below 45 C as a precaution.
Purity analysis of ABNO and keto-ABNO: Because ABNO and ketoꢀABNO are
paramagnetic, characterization by NMR is not straightforward. When we analyzed ABNO by
+
GC/EIꢀMS, the mass spectrum showed [m] /z = 140 as the most abundant peak but higher than
+
expected intensity of the M+1 peak ([m] /z = 141), which caused us concern that this may be the
coꢀeluting hydroxylamine intermediate as an impurity. Our attempt to screen GC conditions to
separate the presumed hydroxylamine impurity proved fruitless. However, when an acidic eluent
was used with LC/ESIꢀMS, two chromatographically resolved peaks were observed. One peak
exhibits a mass spectrum similar to that observed on GCꢀMS for ABNO but with a higher than
+
expected M+1 peak ([m] /z = 141), while the other exhibits a mass spectrum consistent with the
+
corresponding hydroxylamine ([m] /z = 142 for M+H). This appears to indicate that the
hydroxylamine seen by LCꢀMS and GCꢀMS is an artifact, presumably caused by
ACS Paragon Plus Environment
9

Organic Process Research & Development
Page 10 of 18
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
disproportionation of ABNO during the analysis. Similar acid catalyzed disproportionation of
related compound TEMPO to the corresponding hydroxy amine and the oxoammonium cation
4
a
was known. Fortunately, we were able to confirm the purity of ABNO by quantitative NMR, a
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
gentler and nonꢀdestructive technique, that showed typical batches were made with less than
0
.2% hydroxylamine. Figure 1 shows the proton NMR of a typical sample of ABNO with peak
assignments that were determined from density functional theory (DFT) calculations of chemical
4
b
shifts (see Supporting Information for calculation details). Since the compound is
paramagnetic, a high concentration (ca. 200 mg/mL) of ABNO was dissolved in CD CN to
3
improve T relaxation, and the spectral region was increased to ꢀ100 to +250 ppm. The sharpest
2
resonance of ABNO at 26.2 ppm was used for quantitation versus the impurity signals. To
determine the identity of impurity peaks, 2D NMR experiments (i.e., COSY, HSQC, HMBC, and
NOESY) were performed on a narrow spectral region from ꢀ1 to 14 ppm using standard
acquisition parameters. This NMR technique allowed us to assertain the purity of the prepared
ABNO in conjunction with GC analysis.
For ketoꢀABNO, the hydroxylamine peak on HPLC was resolved and identified by LCꢀMS. The
hydroxylamine in the isolated solid ketoꢀABNO was present at 0.2% by HPLC. Similar proton
NMR was also observed for ketoꢀABNO.
ACS Paragon Plus Environment
10

Page 11 of 18
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 1. NMR spectrum of a typical batch of ABNO. Note that diamagnetic impurities include
the HDO and residual acetonitrileꢀd peaks as the largest signals.
2
In summary, a scalable process was developed for the preparation of ABNO and ketoꢀABNO
and the process steps have been scaled up either on pilot plant or several hundred gram scale.
Quantitative proton and 2D NMR of ABNO allows for structure confirmation and purity
analysis. The potential for impact sensitivity should be taken into consideration when handling
ABNO, especially in its solid form.
EXPERIMENTAL
All reagents and solvents were used as purchased as reagent grade or better except otherwise
noted. Purity of isolated products are reported as LCAP (liquid chromatography area percent),
GCAP (gas chromatography area percent), or wt% (weight %). Reference standards used to
determine wt% were assigned purity based on LCAP or GCAP minus residual moisture by KF
titration and residual solvents by GC confirmed by quantitative NMR when possible.
1b
9
-Benzyl-9-azabicyclo[3.3.1]nonan-3-one (4) To a mixture of water (388 g) and benzylamine
ACS Paragon Plus Environment
11

Organic Process Research & Development
Page 12 of 18
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
o
(
100 g, 0.93 mol, 0.9 eq.) at 0ꢀ10 C was added 18% sulfuric acid (364 g, 0.67 mol, 0.65 eq.)
o
while maintaining the reaction mixture at 0ꢀ10 C. The reaction flask was protected from light as
a general precaution. Then 50% glutaraldehyde (2, 205.5 g, 1.03 mol, 1.0 eq.) was added
followed by acetone dicarboxylic acid (1, 150 g, 1.03 mol, 1.0 eq.) while maintaining the
o
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
reaction mixture at 0ꢀ10 C. Then 9% sodium acetate (363 g, 0.40 mol, 0.4 eq.) solution was
o
o
added over 4 h. The reaction mixture was aged at 0ꢀ10 C for 10 h and then 20ꢀ30 C for 29 h.
Slow gas evolution was observed during the reaction. A sample analyzed by HPLC indicated
complete consumption of acetone dicarboxylic acid but 10/90 ratio of benzylamine to product.
o
Additional 18% sulfuric acid (132 g) was added at 20ꢀ30 C to adjust the batch pH to 2. The
batch was extracted with MTBE (3 x 884 g). Then heptane (3.42 kg), silica gel (150 g) were
added to the aqueous layer followed by 20% Na CO (450 g) to adjust the pH to 8 (Caution: gas
2
3
evolution!). The batch was filtered and cake washed with 2.4 L heptane. The layers were
separated, and the organic layer concentrated under vacuum and solvent switched to methanol
(
about 1.66 kg). A solution of 4 in methanol at about 15 wt% was obtained with an assay amount
of 114 g (53% yield) and 94% LCAP. A pure sample was obtained by silica gel column
purification using ethyl acetate/heptanes as eluent, which resulted in a white solid with mp 72.4
o
1
C. H NMR (400 MHz, CDCl ) δ = 7.45ꢀ7.38 (m, 2H), 7.38ꢀ7.31 (m, 2H), 7.31ꢀ7.25 (m, 1H),
3
3
.94ꢀ3.90 (m, 2H), 3.32 (br. s, 2H), 2.74 (dd, J = 6.6, 16.6 Hz, 2H), 2.27 (d, J = 16.7 Hz, 2H),
13
2
.02ꢀ1.89 (m, 2H), 1.61ꢀ1.46 (m, 4H). C NMR (101 MHz, CDCl ) δ = 211.59, 139.20, 128.35,
3
+
128.28, 127.11, 57.03, 53.49, 42.81, 29.30, 16.53. HRMS (ESI) m/z: [M+H] Calcd for
C H NO 230.1539; Found 230.1540.
15 20
1b
9
-Benzyl-9-azabicyclo[3.3.1]nonan-3-ol (6). To a stirred solution of 4 (48.4 g in 458 g
methanol, 0.21 mol, 1.0 eq.) at 5ꢀ15 °C under nitrogen was added NaBH (50 g, 1.32 mol, 6.3
4
eq.) in portions over 1 h while maintaining the batch temperature at 10 °C. The batch was stirred
at 25 °C for 1 h until HPLC analysis indicated completion of reaction. The batch was then cooled
to 5 °C and quenched by slow addition of 10% hydrochloric acid (511 g over 1 h) to get the
batch pH to 5ꢀ6 (Caution: hydrogen gas evolution was removed by steady nitrogen sweep). The
o
batch was concentrated under vacuum (<50 C) to about 10 L and cooled to 5 °C and then mixed
with MTBE (286 g). Then 20% NaOH solution (73 g) was added at 10 °C to adjust the pH to
1
1ꢀ12. The batch was filtered to remove the solids formed, and the cake was washed with
ACS Paragon Plus Environment
12

Page 13 of 18
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
MTBE (72 g). The layers were separated, and the aqueous layer extracted with MTBE (144 g).
The combined organic layer was washed with 4% aqueous NaHCO solution (200 g),
3
concentrated under reduced pressure at below 40 °C to remove all solvents. Crude compound 6
was obtained as a thick oil (51.0 g), which was used directly in the next step without purification.
A pure sample was obtained by silica gel column purification with ethyl acetate/heptanes as
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
o
1
eluent, which yielded a white solid with mp 70.7 C. H NMR (400 MHz, CDCl ) δ = 7.39ꢀ7.31
3
(m, 4H), 7.29ꢀ7.23 (m, 1H), 4.36ꢀ4.25 (m, 1H), 3.82 (s, 2H), 3.11ꢀ3.04 (m, 2H), 2.46ꢀ2.36 (m,
2
2
3
2
H), 2.31ꢀ2.17 (m, 1H), 2.03ꢀ1.87 (m, 3H), 1.60ꢀ1.52 (m, 1H), 1.42ꢀ1.34 (m, 2H), 1.18ꢀ1.11 (m,
1
3
H); C NMR (101 MHz, CDCl ) δ = 140.45, 128.14, 128.09, 126.60, 64.13, 55.85, 49.60,
3
+
5.49, 25.00, 14.53. HRMS (ESI) m/z: [M+H] Calcd for C H NO 232.1696; Found
15 22
32.1696.
9
-Benzyl-9-azabicyclo[3.3.1]non-3-ene (8) A solution of crude 6 from the previous step (50 g,
theoretical 0.21 mol, 1.0 eq.) in 70% aq. H SO (500 g, 3.6 mol, 17 eq.) was heated to 100 °C
2
4
and stirred for 20 h until >99% conversion by HPLC. The batch was slowly quenched into a
solution of NaOH (288 g in 1200 g water) while maintaining the quench batch temperature at or
o
below 30 C (The end pH was 7ꢀ8). The batch was then extracted with 3x1L ethyl acetate, and
the combined organic layer was washed with 300 mL water. The organic layer was concentrated
to remove all solvents to give 52.7 g oil residue which contained 43.8 g 8 by HPLC assay and
9
6.8% LCAP in 97.3% overall yield for two steps. Pure 8 was obtained as a light yellow oil via
1
silica gel column purification with ethyl acetate/heptanes as eluent. H NMR (400 MHz, CDCl )
3
δ = 7.35ꢀ7.28 (m, 2H), 7.27ꢀ7.19 (m, 2H), 7.19ꢀ7.12 (m, 1H), 5.95 (td, J = 3.5, 10.1 Hz, 1H),
5
1
1
.56ꢀ5.39 (m, 1H), 3.66ꢀ3.54 (m, 2H), 3.07 (br. s, 1H), 2.91ꢀ2.84 (m, 1H), 2.37 (ddd, J = 3.0, 4.3,
1
3
8.8 Hz, 1H), 1.82ꢀ1.57 (m, 4H), 1.45ꢀ1.29 (m, 3H). C NMR (101 MHz, CDCl ) δ = 140.16,
3
28.69, 128.04, 127.86, 127.12, 126.57, 57.79, 52.48, 50.55, 33.38, 28.37, 25.70, 16.09. HRMS
+
(
ESI) m/z: [M+H] Calcd for C H N 214.1590; Found 214.1588.
1
5
20
1
b
9
-Azabicyclo[3.3.1]nonane (5) A solution of 8 (10.0 g x 93.16 wt%, 44 mmol) in 50 mL
isopropanol was purged with argon three times and mixed with Pd(OH) on carbon (20 wt%)
2
(2.0 g). The vessel was then purged with argon and pressurized with hydrogen at 40ꢀ50 psi and
agitated at 40ꢀ50 °C for 48 h until complete reaction by GC. The catalyst was filtered off through
Solka Floc, and the cake washed with 50 ml IPA. The filtrate contained 5.37 g (62.9 g x 8.53
ACS Paragon Plus Environment
13

Organic Process Research & Development
Page 14 of 18
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
wt%) product 5 by GC assay, 98% yield. A pure sample was obtained via silica gel column
o
purification with ethyl acetate/heptanes as eluent, resulting in a light yellow solid, mp ~110 C.
1
H NMR (400 MHz, CDCl ) δ = 3.57 (br. s, 2H), 2.41ꢀ2.18 (m, 4H), 2.05ꢀ1.88 (m, 2H), 1.80ꢀ
3
13
+
1
.59 (m, 6H) C NMR (101 MHz, CDCl ) δ = 46.39, 27.05, 18.65. HRMS (ESI) m/z: [M+H]
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Calcd for C H N 126.1277; Found 126.1279.
8
16
1
b
9
-Azabicyclo[3.3.1]nonane-N-oxyl (ABNO) To a threeꢀneck round bottom flask equipped
with a thermometer, a mechanical stirrer, and a nitrogen inlet for slow sweep, was charged the
solution of compound 5 (57.4 g x 8.53 wt% = 4.9 g, 39.1 mmol, 1.0 eq.) in IPA (13.6 vol.), 40 g
of H O (40 mL) and 1.29 g (3.9 mmol, 0.1 eq.) of Na WO •2H O. The batch was cooled to 0ꢀ10
2
2
4
2
°
C, and a preꢀcooled solution of urea hydrogen peroxide (UHP, 12.5 g, 132.9 mmol, 3.4 eq.) in
0 g water (solution at 0 to 10 °C) was added dropꢀwise over 1 h. A sample was taken after
4
another 0.5 h, and GC analysis indicated about 97% conversion. The batch was extracted with
dichloromethane (DCM) (49 mL x 3). The combined organic layers were concentrated under
reduced pressure at less than 10 °C until 12 mL and then solvent switched to nꢀheptane (100 mL
x 2) to a final volume of 12 mL. The slurry was cooled to ꢀ10 to ꢀ5 °C and stirred overnight (18
h). Crude ABNO was collected by filtration. HPLC indicated 85% LCAP purity with
significant amount of urea present. The wet cake was dissolved with 90 mL DCM, and the
solution was washed with water (50 mL x 2). The aqueous layer was backꢀextracted with 30 mL
DCM, and the combined organic layers were concentrated under reduced pressure to ca. 50 mL.
The solution was then further azeotropically distilled with 100 mL nꢀheptane under reduced
pressure to a final volume of 50 mL. The mixture was filtered and the cake dried in vacuum
below 30 °C for 2 h. ABNO was obtained as a red solid, 3.84 g, 99.5% GCAP in 70% isolated
o
o
yield. DSC showed two melting endotherms, 55 C and 69 C, indicating crystal form
1
conversions. H NMR (600 MHz, CD CN) δ 166.4 (br. s, 2H), 96.1 (br. s, 2H), 73.9 (br. s, 4H),
3
2
6.2 (br. s, 2H), ꢀ45.2 (br. s, 4H). More details on the peak assignments are available in the
Supporting Information.
1
b
9
-Azabicyclo[3.3.1]nonan-3-one (3) A hydrogenation vessel was charged with 112 g 4 (in
about 1.05 kg solution in MeOH, 0.49 mol, 1.0 eq.). The vessel was purged three time with
nitrogen, and 36.9 g of 10% Pd/C (0.035 mol, 0.07 eq.) was charged. The batch was agitated
under H (40 to 50 psi) and at 40 to 50 °C for 27 h until >99% conversion by HPLC analysis.
2
ACS Paragon Plus Environment
14

Page 15 of 18
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
The batch was filtered through diatomite to remove catalyst. GC assay of the filtrate indicated
97.2% GCAP purity for 3 and quantitative yield. This solution was used in the next step without
purification. NMR and MS data of the solid obtained after silica gel column purification
1
matches the reported data. H NMR (400 MHz, CDCl ) δ = 3.60 (br s, 2 H), 2.54ꢀ2.65 (m, 2 H),
3
1
3
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
.36ꢀ2.45 (m, 2 H), 1.71ꢀ1.84 (m, 2 H), 1.61ꢀ1.71 (m, 3 H), 1.45ꢀ1.61 (m, 2 H). C NMR (101
+
MHz, CDCl ) δ = 211.00, 49.21, 47.40, 31.93, 16.67. HRMS (ESI) m/z: [M+H] Calcd for
3
C H NO 140.1070; Found 140.1072.
8
14
1
b
9
-Azabicyclo[3.3.1]nonan-3-one-N-oxyl (keto-ABNO) A methanol solution of 3 from the
last step (50.6 g, 0.36 mol in 765.1 g solution, 1.0 eq.) was concentrated under reduced pressure
below 40 °C and solvent switched to water to a final volume of 650 mL. GC analysis indicated
residue MeOH in the mixture was 3.39 wt%. The batch was cooled to 0ꢀ10 °C, and
Na WO •2H O (6 g, 0.018 mol, 0.05 eq.) was charged. In another vessel, urea hydrogen
2
4
2
peroxide (UHP, 102.6 g, 1.09 mol, 3.0 eq.) was added in portions to 500 mL cold water at 0ꢀ
0 °C. This UHP solution was then added dropꢀwise to the batch at 0ꢀ10 °C over 1.5 h. After
1
aging for another 0.5 h, HPLC analysis indicated the levels of amine 3/hydroxylamine/ketoꢀ
ABNO to be 0% / 1.1% / 87.5% (Caution: prolonged aging for 18 h resulted in partial
conversion of ketoꢀABNO back to hydroxylamine from 1% to 6%). The batch was twice
extracted with 500 mL DCM (500 mL x 2). Assay of the DCM layer indicated purity of ketoꢀ
ABNO was 96.1% LCAP, 44.2 g, 79% assay yield. The combined organic layers were
concentrated under reduced pressure below 10 °C until 200 mL (Caution: it is critical to keep
o
the batch temperature close to 0 C, and the batch should not be stored at high concentration to
minimize decomposition of ketoꢀABNO) and 750 mL nꢀheptane was added dropꢀwise at 0ꢀ5 °C
over 4 h. The slurry was stirred at 0ꢀ5 °C for another 8 h and at ꢀ5 to 0 °C for 1 h. The batch
was filtered, and the cake was dried on the filter under nitrogen flow. KetoꢀABNO was obtained
as a yellow solid, 34.2 g by assay (35.6 g x 96.0 wt %), and 98.5% LCAP in 61% yield. Mp by
1
DSC (decomp.) ~110 °C. H NMR (600 MHz, CD Cl ) δ 130.8 (br. s, 2H), 98.5 (br. s, 2H), 55.8
2
2
(br. s, 2H), 52.0 (br. s, 1H), 15.1 (br. s, 1H), ꢀ32.5 (br. s, 2H), ꢀ53.1 (br. s, 2H). HRMS (ESI) m/z:
+
[
M] Calcd for C H NO 154.0868; Found 154.0861.
8
12
2
ACS Paragon Plus Environment
15

Organic Process Research & Development
Page 16 of 18
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Supporting Information. HPLC or GC analysis conditions for each step, DSC of ABNO and
1
ketoꢀABNO, and density functional theory (DFT) calculations of H NMR chemical shifts for
ABNO and ketoꢀABNO are included in the supporting information. This material is available
free of charge via the internet at http://pubs.acs.org.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
AUTHOR INFORMATION
Corresponding Author
*zhiguo_song@merck.com
ACKNOWLEDGMENT
The authors would like to acknowledge the technical support of Xianghui Wen, Minyi Feng,
Weicheng Wang, Xiugui Wen, Yaping Wang, Haokun Shi, Xuedu Lan, Feng Zhou, Pengju Chen
Wen Leng, Bangbang Cao, Pengju Chen, Qun Xiang, Yipeng Lin, and Xuedu Lan from Wuxi
AppTec and helpful discussions with Qinghao Chen, Jing Xu, Mikhail Reibarkh, Huaming Shen,
RongꢀSheng Yang, Steven Oliver from MSD.
REFERENCES
1
. Synthesis of ABNO: (a) Shibuya, M.; Tomizawa, M.; Sasano, Y.; Iwabuchi, Y. An
Expeditious Entry to 9ꢀAzabicyclo[3.3.1]nonane NꢀOxyl (ABNO): Another Highly
Active Organocatalyst for Oxidation of Alcohols J. Org. Chem. 2009, 74, 4619–4622.
(b) Lauber, M. B.; Stahl S. S. Efficient Aerobic Oxidation of Secondary Alcohols at
Ambient Temperature with an ABNO/NOx Catalyst System ACS Catal. 2013, 3, 2612−
2616.
ACS Paragon Plus Environment
16

Page 17 of 18
Organic Process Research & Development
1
2
3
4
5
6
7
8
9
2
. ABNO catalyzed oxidations: (a) Steves, J. E.; Stahl, S. S. Copper(I)/ABNOꢀCatalyzed
Aerobic Alcohol Oxidation: Alleviating Steric and Electronic Constraints of Cu/TEMPO
Catalyst Systems J. Am. Chem. Soc. 2013, 135, 15742 15745. (b) TEMPO/ABNO
−
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
McCann, S. D.; Stahl, S. S. CopperꢀCatalyzed Aerobic Oxidations of Organic Molecules:
Pathways for TwoꢀElectron Oxidation with a FourꢀElectron Oxidant and a OneꢀElectron
RedoxꢀActive Catalyst Acc. Chem. Res. 2015, 48, 1756 1766. (c) Steves, J. E.; Preger,
−
J.; Martinelli, J. R.; Welch, C. J.; Root, T. W.; Hawkins, J. M.; Stahl, S.S. Process
Development of CuI/ABNO/NMIꢀCatalyzed Aerobic Alcohol Oxidation Org. Process
Res. Dev. 2015, 19, 1548ꢀ1553. (d) Greene, J. F.; Preger, J.; Stahl, S. S.; Root, T.W.
PTFEꢀMembrane Flow Reactor for Aerobic Oxidation Reactions and Its Application to
Alcohol Oxidation Org. Process Res. Dev. 2015, 19, 858 864. (e) Kim, J.; Stahl, S. S.
−
Cu/NitroxylꢀCatalyzed Aerobic Oxidation of Primary Amines into Nitriles at Room
Temperature ACS Catal. 2013, 3, 1652 1656. (f) Qun Cao, Laura M. Dornan, Luke
−
Rogan, N. Louise Hughes and Mark J. Muldoon Aerobic oxidation catalysis with stable
radicals Chem. Commun., 2014, 50, 4524. (g) Luke Rogan, N. Louise Hughes, Qun Cao,
Laura M. Dornan and Mark J. Muldoon Copper(I)/ketoABNO Catalysed Aerobic
Alcohol Oxidation Catal. Sci. Technol., 2014, 4, 1720ꢀ1725. (h) Zultanski, S. L.; Zhao,
J. and Shannon S. Stahl, S. S. Practical Synthesis of Amides via Copper/ABNOꢀ
Catalyzed Aerobic Oxidative Coupling of Alcohols and Amines J. Am. Chem. Soc. 2016,
1
38, 6416ꢀ6419.
3
. WolffꢀKishner reductions under flow chemistry conditions: Newman, S. G.; Gu, L.;
Lesniak, C.; Victor, G.; Meschke, F.; Abahmaneb, L.; Jensen, K. F. Rapid Wolff–
Kishner reductions in a silicon carbide microreactor Green Chem., 2014, 16, 176.
ACS Paragon Plus Environment
17

Organic Process Research & Development
Page 18 of 18
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
4
. (a) Sen, V. D.; Golube, V. A. Kinetics and mechanism for acidꢀcatalyzed
disproportionation of 2,2,6,6ꢀtetramethylpiperidineꢀ1ꢀoxyl J. Phys. Org. Chem. 2009, 22
138–143. (b) Paramagnetic NMR spectra were recorded and resonances assigned by
density functional theory calculations following the method reported: Rastrelli, F.;
Bagno, A. Predicting the NMR Spectra of Paramagnetic Molecules by DFT: Application
to Organic Free Radicals and TransitionꢀMetal Complexes Chem. Eur. J. 2009, 15, 7990
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
–
8004. More details are available in Supporting Information.
5
. One of the major challenges we faced for this step is the facile decomposition of acetone
dicarboxylic acid 1. Even the process of NMR sample preparation and analysis led to
various extent of decarboxylation. Similar decarboxylation was also observed when 1
was analyzed by reverse phase HPLC. Due to the lack of analytical standard of this
compound, the weight percentage purity of the material was not confirmed. We suspect
the different reported yields of this step were largely due to the variable quality of the raw
materials used. In our work we kept compound 1 refrigerated all the time except when
needed for reaction.
6
7
. Whenever possible, both ABNO and ketoꢀABNO solid should be stored refrigerated.
. The mass spectrum for ketoꢀABNO and the corresponding hydroxylamine on GC/MS
showed predominantly the positive molecular ion peak for each so they can be
differentiated.
ACS Paragon Plus Environment
18