Highly Efficient Multigram Synthesis of Dibenzoazacyclooctyne (DBCO) without Chromatography

Article abstract of DOI:10.1021/acs.oprd.9b00406

The synthesis of 4-[11,12-didehydrodibenzo[b,f]azocin-5(6H)-yl]-4-oxobutanoic acid, also known as dibenzoazacyclooctyne (DBCO) or aza-dibenzocyclooctyne (ADIBO), was optimized for large-scale preparations of at least 10 g with an overall yield of 42%.

Full text of DOI:10.1021/acs.oprd.9b00406

Article
pubs.acs.org/OPRD
Cite This: Org. Process Res. Dev. XXXX, XXX, XXX−XXX
Highly Efficient Multigram Synthesis of Dibenzoazacyclooctyne
(DBCO) without Chromatography
Stuart A. McNelles, Julia L. Pantaleo, and Alex Adronov*
Department of Chemistry & Chemical Biology, McMaster University, 1280 Main Street West, Hamilton, Ontario, L8S 4M1 Canada
S
* Supporting Information
ABSTRACT: The synthesis of 4-[11,12-didehydrodibenzo[b,f ]azocin-5(6H)-yl]-4-oxobutanoic acid, also known as
dibenzoazacyclooctyne (DBCO) or aza-dibenzocyclooctyne (ADIBO), was optimized for large-scale preparations of at least
10 g with an overall yield of 42%.
KEYWORDS: strained cyclooctyne, SPAAC, click chemistry, DBCO, ADIBO, bioorthogonal
INTRODUCTION
RESULTS AND DISCUSSION
■
■
When planning this new preparation of DBCO−COOH, we
first had to identify the critical bottlenecks in our previous
work. These bottlenecks are indicated in red in Scheme 1, and
the key changes from this preparation are indicated in blue.
In our original synthesis, the initial conversion of
dibenzosuberenone to the corresponding oxime 2 was
performed using pyridine in ethanol at reflux using hydroxyl-
amine hydrochloride. While this reaction is clean and allows
for quantitative conversion over the course of 24 h, the workup
required a large volume of solvent to fully remove all the
pyridine by extraction, particularly due to the need to
sufficiently dilute the ethanol to allow for phase separation.
We attempted to replace pyridine with an inorganic water-
soluble base that could be removed by a simple precipitation
step into water, and sodium carbonate was found to be a
suitable reagent for this purpose. The removal of pyridine as a
co-solvent substantially reduced the solubility of dibenzosuber-
enone in ethanol at room temperature, but it remained soluble
at reflux. The workup required only cooling the reaction
mixture to room temperature and precipitation into ice-cold 1
M H3PO4 and subsequent filtration to give the product as a
white powder in a nearly quantitative yield.
To perform the Beckmann rearrangement to form amide 3,
we screened conditions such as acetic acid/acetic anhydride,
polyphosphoric acid, and DIBAL-H rearrangement/reduction,
but none of these provided a clean conversion to the amide.
Instead, we returned to Eaton’s reagent (7.7 wt % P2O5 in
methanesulfonic acid), which results in very clean conversion
from starting material to product with only minor impurities
(ca. 3% by high-performance liquid chromatography (HPLC)
area). Since Eaton’s reagent is one of the costliest reagents
used in the synthesis and used here as a solvent, we
investigated the effect of increasing the concentration of 2 in
the reaction mixture to determine if we could reduce the
amount of Eaton’s reagent required for this step. We varied the
concentration of 2 from 166 to 332, 500, 664, 996, and
ultimately 1600 mg/mL. At concentrations up to 500 mg/mL,
The strain-promoted alkyne-azide cycloaddition (SPAAC)
reaction, originally developed by Blomquist1 and popularized
by Bertozzi and co-workers,2 has seen extensive use in
chemical biology.3 The high rate of reaction, water and air
tolerance, and bioorthogonality make it exceptionally useful for
protein modification,4−6 gel formation,7−9 and numerous other
applications.10−12 Despite the utility of SPAAC, the cost of the
strained cyclooctyne reagents has precluded its use for larger-
scale applications. Currently, strained cyclooctyne reagents
such as cyclooctynol, BCN, DBCO−COOH, and DBCO−
NHS still cost over USD 1000/g, which is unacceptably high
for use in polymer or materials chemistry. Although work has
been done to develop efficient scalable methods for less
reactive strained cyclooctynes such as (2-cyclooctyne-1-yloxy)
acetic acid,13 there are currently no published reports on the
syntheses of more reactive strained cyclootynes that circum-
vent chromatographic purification.
Dibenzoazacyclooctyne (DBCO or ADIBO) is a strained
cyclooctyne that has one of the highest second-order rate
constants of commonly used strained cyclooctynes (∼0.35
M−1 s−1).14 Originally, this was prepared using two distinct
methods by van Delft and Popik.14,15 In a previous report,16
our group has modified the first several steps of the Popik
route, although this synthesis still required several chromato-
graphic steps and arduous workup procedures. This limited the
scale of DBCO synthesis to approximately 1 g (Scheme 1).
However, our group is interested in the use of DBCO for
polymer and materials chemistry, which required substantially
larger quantities of material, and so we have sought out an
improved method of DBCO synthesis which is efficient, high
yielding, and requires no chromatography.
The synthetic route described herein has the same synthetic
sequence as the original work by Popik and our previous
modification, although substantial changes have been made in
workup and purification to allow for the facile isolation of each
intermediate, without any chromatographic purification steps.
The overall yield is 42% over seven steps, and we were able to
prepare 10 g of material.
Received: September 23, 2019
© XXXX American Chemical Society
A
DOI: 10.1021/acs.oprd.9b00406
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Scheme 1. Synthesis of DBCO−COOH
the reaction proceeded rapidly and cleanly to give 3. However,
at higher concentrations, the reaction began to produce several
side products (data not shown), and so we settled on a
concentration of 500 mg/mL for this scale up, which resulted
in 66% reduction in the amount of Eaton’s reagent required. At
scales above ∼25 g, the magnetic stirrer was unable to
sufficiently stir the reaction, and invariably the solid 2 was
found to coalesce into a viscous clump, which resulted in the
formation of several byproducts by thin-layer chromatography
(TLC). To avoid this, mechanical stirring is required for scales
larger than ∼20 g. The workup was also changed from the
previous quench with water and serial extraction with hot ethyl
acetate, which would take up to 10 extraction cycles to recover
3 from the aqueous phase. Instead, the hot reaction mixture
was poured directly onto crushed ice, which precipitated the
product as an off-white slurry. This was filtered through a glass
fritted funnel and washed with water until the filtrate was no
longer acidic, indicating the complete removal of residual
Eaton’s reagent. Amide 3 was found to be challenging to dry,
which was a concern for use in the subsequent LiAlH4
reduction step. While room-temperature vacuum drying was
insufficient to remove water, vacuum drying at 80 °C overnight
resulted in material having less than 500 ppm of moisture as
analyzed by the Karl−Fisher analysis. For bulk drying, the use
of a vacuum oven is inconvenient, and we found that the use of
an in-house fluidized bed dryer resulted in a material with
similar dryness as obtained from the vacuum oven after only 4
h.
issue is compounded by the release of hydrogen gas and the
evaporation of solvent due to the exothermal reaction. This
results in splattering of LiAlH4 slurry and poor mixing of the
water for quenching, which can ignite the mixture under worst-
case conditions. Previously, this was avoided by the dilution of
the reaction mixture into a large volume of dichloromethane
(DCM), which ensured that magnetic stirring continued. We
instead turned to mechanical stirring, and this was found to be
much more resistant to stalling in the presence of precipitates.
Indeed, vigorous mechanical stirring allowed us to use 75% less
of the total solvent for this reaction and resulted in a far more
homogeneous addition of water to the quench, which made the
subsequent filtration and solvent removal steps more
straightforward as well. The original workup only quenched
with water, resulting in a very fine precipitate that readily
formed foams and was challenging to filter. To address this, the
Fieser workup17 (water, NaOH, water) was used instead to
form large particles of alumina, which were straightforward to
remove by filtration. It should be noted that if the reaction is
left for more than 48 h without workup, minor impurities are
detectable by TLC, although the product is easily purified by
recrystallization in methanol.
The acylation of 4 with methyl 4-chloro-4-oxybutyrate (5)
to prepare 6 presented two challenges: the high cost of 5 from
commercial sources and the chromatographic purification of 6,
which was a major bottleneck for the scale up of the overall
synthesis. Our original method produced 6 in adequate yield
and purity, although several other products were visible by
TLC analysis after isolation. We first chose to replace Et3N
with 2 M KOH and ran the reaction under Schotten−
Baumann conditions. Several solvents were tried for the
organic phase such as toluene and methyl tert-butyl ether
(MTBE), although dichloromethane provided a much cleaner
conversion. We then decided to vary the stoichiometry of 5
and found that using this as the limiting reagent, rather than 4,
The reduction of 3 to amine 4 was accomplished using
LiAlH4, although we found that the use of 4 equiv of LiAlH4
would allow for the full conversion to 4 overnight, rather than
the previously used 20 equiv. Previously, workup for this
reaction had been particularly challenging, as the large amount
of solid which forms a fine precipitate during the aqueous
quenching of LiAlH4 would prevent magnetic stirring. This
B
DOI: 10.1021/acs.oprd.9b00406
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
produced 6 in considerably higher purity than when 4 was
used as the limiting reagent. We determined that using a 1−2
mol % excess of 4 gave the best compromise of yield and
purity, as this discarded a minimal amount of 4 as an excess but
still retained excellent purity. The organic layer was then
simply separated from the aqueous base later by extraction, and
rotary evaporation gave the crude material as a slightly brown
solid. Methanol, ethanol, acetone, isopropanol, and 2-butanone
were investigated as solvents for recrystallization, and
ultimately ethanol proved to be most effective. We were
initially concerned with the possibility of the trans-
esterification of methyl ester to form ethyl ester, but 1H
NMR did not show evidence of any significant amount
(<0.5%) of the ethyl ester. In addition, the following step is the
hydrolysis of this ester, and so small amounts of trans-
esterification should not affect the overall synthetic pathway
significantly.
sequence. We initially screened the base used and found that
amine bases such as Et3N and DBU were unacceptably slow to
perform the second elimination, even if the temperature of the
reaction was elevated to 40 °C. Stronger, bulky bases such as
KHMDS, LiHMDS, and LDA were screened and, while 9 was
formed, it was not the major product (multiple additional
peaks were visible by HPLC, data not shown). We therefore
returned to the original choice of KOtBu, which was found to
work best when used in excess (3.5−4 equiv), resulting in the
full conversion of 8 to 9 at −78 °C with only minor impurities
(<2%) detectable by HPLC. We corroborated the HPLC
1
results using H NMR on the crude material, and the identity
and purity of the compound were confirmed. However,
workup of the reaction mixture proved challenging, as many
workup procedures appear to substantially decompose 9.
Ultimately, we determined that dilution of the reaction mixture
with Et2O and doing an aqueous extraction with H3PO4
sufficiently quenched the KOtBu. We had hoped to find a
recrystallization solvent that would provide 9 in pure form,
although solvents such as MeOH, EtOH, iPrOH, n-PrOH,
acetone, butanone, tetrahydrofuran (THF), CH2Cl2, and
CHCl3 were all found to either not purify the material
sufficiently, or heating to the boiling point of the solvent would
decompose the reaction mixture. Ultimately, recovery of pure
9 from this solution was performed by rotary evaporation
without heating to keep the crude mixture cold throughout
evaporation. After visible crystal formation, the flask was
removed from the rotary evaporator and an equal volume of
diethyl ether was added to the crude mixture. This was left to
cool and crystallize at −20 °C in a refrigerator overnight. The
resulting crystals are then collected on a filter and washed with
a minimum of cold Et2O, giving 9 with excellent purity as
Since commercial sourcing of 5 on a laboratory scale is
expensive, we sought to prepare this on scale as well. The
preparation of monomethyl hydrogen succinate from succinic
anhydride has been reported in the literature, typically via
heating to reflux in methanol.18 However, in our experience,
we found considerable formation of dimethyl succinate, as well
as the formation of succinic acid if methanol was not rigorously
dried. To circumvent this, we prepared monomethyl hydrogen
succinate by 4-dimethylaminopyridine (DMAP) catalyzed ring
opening of succinic anhydride in methanol at room temper-
ature. After a short induction period of ∼15 min, the reaction
is quite vigorous and is completed within a further 15 min.
After filtering the reaction mixture through a short plug of
DOWEX-WX2 resin to remove the DMAP catalyst, the
1
resulting material exhibited >99% purity by H NMR and did
1
determined by HPLC and H NMR with a yield of 53%.
not require purification before the formation of the acid
chloride. Conversion to acid chloride was performed by
heating monomethyl hydrogen succinate with SOCl2 and then
fractional distillation to give 5.
Multiple attempts to isolate and crystallize additional crops of
product from the mother liquor were made, but all crops
subsequent to the initial one were found to be unacceptably
impure (purity of approximately 50% by NMR analysis). Thus,
the labor involved in the isolation of additional crops of
material was deemed unwarranted.
Our original saponification procedure of 6 to 7 was
performed at a relatively low concentration (ca. 30 mg/mL),
as the solubility of 6 in 2:1 MeOH/H2O is poor. We suspected
that the reaction would proceed with less solvent despite the
poor solubility, as there should be an equilibrium between the
undissolved solid phase and the liquid phase. We therefore
increased the concentration to 500 mg/mL. Although 6 was
mostly undissolved in 2:1 MeOH/H2O even at reflux, we
found that after 1 h of stirring at reflux with 6 equiv of LiOH,
the product had fully saponified to 7. In this case, the poor
solubility of 7 was exploited for the workup, as the base was
quenched by the addition of 1 M H3PO4, and the product
could be isolated by extraction into CH2Cl2 and solvent
removal, without having to remove a large amount of MeOH.
We also investigated our previous choice of LiOH·xH2O as the
base and found that when using the higher-concentration
procedure, there was no practical difference between LiOH,
NaOH, or KOH. The increased concentration also served to
increase the rate of reaction, decreasing the reaction time from
16 to 1 h.
CONCLUSIONS
■
We have optimized the synthesis of DBCO−COOH for the
preparation on a multigram scale, with an overall seven-step
yield of 42%. Key workup and purification steps of previous
syntheses such as serial extractions with hot solvent, large-
volume extractions, and silica gel chromatography have been
circumvented with easier, more streamlined approaches. This
dramatic cost reduction of 9 has enabled our research group
and others to translate of use of strained cyclooctynes into
polymer chemistry and material chemistry work. We believe
that enabling access to large amounts of cyclooctynes at low
cost should enable many researchers to find interesting new
uses that are currently economically impractical.
EXPERIMENTAL SECTION
■
Bromination of 7 was performed similarly to our previous
work, with the key distinction of using a stoichiometric amount
of Na2SO3 to quench excess Br2, rather than a large molar
excess. This considerably reduced the formation of an
emulsion during the extraction, as much less 7 was
deprotonated by the Na2SO3 solution. The elimination of 8
however proved to be the most challenging step of the entire
Chemicals were sourced from Chem-Impex unless otherwise
noted. Eaton’s reagent, LiAlH4, KOtBu, and solvents were
sourced from Sigma-Aldrich. Solvents were either purchased
anhydrous or were dried with 3 Å molecular sieves or activated
1
neutral alumina. H NMR spectroscopy was performed on a
Bruker Avance AV600 or AV700 spectrometer equipped with a
cryoprobe at 600 and 700 MHz, respectively. Residual
C
DOI: 10.1021/acs.oprd.9b00406
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
nondeuterated solvent signals were used as internal chemical
shift references. All chemical shifts are reported in ppm.
Mechanical stirring was performed with a Heidolph model
RZR 2071 explosion-proof stirrer. HPLC analysis was
performed using a Waters Alliance 2695 separations module
equipped with a column heater and a Waters 2996 PDA
detector. All HPLC purity is presented in area % at 254 nm.
Eluent A consisted of 5% acetonitrile in water, eluent B
consisted of 100% HPLC grade acetonitrile, eluent C consisted
of 5% acetonitrile and 0.1% trifluoroacetic acid in water, and
eluent D consisted of 0.1% trifluoroacetic acid in HPLC grade
acetonitrile. HPLC analysis was performed using two methods.
Method 1 used a 3.0 × 100 mm Phenomenex Kinetex column
(2.7 μm particle size) at 40 °C using the following gradient
profile: flow rate of 0.76 mL/min, 100% A for 1 min, ramp to
100% B over 5 min, hold 100% B for 2 min, ramp to 100% A
over 10 s, then equilibrate with 100% A for 2 min. Method 2
used a 4.6 × 150 mm Phenomenex Luna Phenyl Hexyl column
(3 μm particle size) using the following gradient profile: flow
rate of 1.5 mL/min, 100% C for 1 min, ramp to 100% D over 9
min, hold 100% D for 2 min, ramp to 100% C over 1 min, then
equilibrate with 100% C for 3 min.
5,6-Dihydrodibenzo[b,f ]azocine (4). Compound 3
(40.013 g, 180.84 mmol) was added to a 1 L glass reactor
equipped with a paddle-type Teflon mechanical stirrer and
stirred at 400 rpm. This was suspended in 300 mL of
anhydrous ether and then the headspace was purged under a
strong flow of N2 gas for 10 min. Pellets of LiAlH4 (34.154 g,
899.9 mmol) were added slowly against a counterflow of N2
gas, and the reaction began to slowly boil during the addition.
The reactor was then sealed and kept under an atmosphere of
N2. The reactor was then placed in an oil bath at 35 °C and
stirred for 24 h at a gentle reflux. At this point, the reaction was
diluted with 300 mL of CH2Cl2 and cooled to 0 °C in an ice
bath for 30 min and was quenched by the dropwise addition of
40 mL of water. During the quench the stirring speed was
increased to approximately 800 rpm to ensure thorough mixing
of the water into the organic layer. The reactor was left to cool
for 15 min; then, 40 mL of 4 M NaOH was added dropwise,
which was extremely exothermic. The addition rate was
controlled manually to prevent excessive boiling of the solvent,
and the total addition time was approximately 30 min. At this
point, a further 120 mL of deionized water was added to the
reactor, and this was left to stir at 400 rpm for 45 min. Sodium
sulfate was then added until the liquid layer appeared clear and
was left to stir for 15 more min. The reaction mixture was then
filtered through a glass fritted funnel and the solids were
washed with 3 × 100 mL EtOAc to remove adsorbed 4.
Solvent was then removed by rotary evaporation to give the
product as a bright yellow powder (38.412 g, 96%).
5H-Dibenzo[a,d]cyclohepten-5-one oxime (2). To a 2
L round-bottom flask equipped with a magnetic stirrer,
dibenzosuberenone (100.032 g, 485.03 mmol) was added,
and this was then diluted with 600 mL of anhydrous ethanol.
To this, hydroxylamine hydrochloride (134.872 g, 1941 mmol)
and sodium carbonate (77.109 g, 727.92 mmol) were added. A
reflux condenser was attached, and the reaction was then
heated in an oil bath at 90 °C for 24 h, at which point the TLC
analysis (5% MeOH/CH2Cl2) indicated full conversion. The
reaction mixture was then cooled to room temperature and
poured into 1 L of ice cold 1 M HCl, forming an off-white
slurry. The solution was then filtered through a glass frit and
washed with 3 × 200 mL of 1 M HCl and 3 × 200 mL of
deionized water. The filter cake was left to dry on the frit for 1
h and then dried in vacuo overnight to give the product as an
off-white solid (105.262 g, 98%).
1H NMR (600 MHz; CDCl3): δ 7.06−6.97 (m, 3H), 6.91
(d, J = 7.7, 1H), 6.88 (d, J = 7.1, 1H), 6.81−6.78 (m, 1H),
6.57−6.54 (m, 1H), 6.40 (d, J = 13.2, 1H), 6.16 (d, J = 13.1,
1H), 6.10 (d, J = 8.0, 1H), 4.23 (s, 2H).
HPLC purity: 97.8% (area%).
Methyl 4-dibenzo[b,f ]azoncin-5(6H)-yl-4-oxobuta-
noate (6). Compound 4 (34.350 g, 165.7 mmol) was added
to a 1 L round-bottom flask and then dissolved in 300 mL of
CH2Cl2. Two hundred seventy milliliters of 2 M KOH was
added, and the reaction was magnetically stirred at 600 rpm.
The flask was then equipped with an addition funnel, to which
acid chloride 5 (27.5 mL, 223 mmol) was added and then
diluted with CH2Cl2 to a total volume of 80 mL. This was
added dropwise to the stirred reaction mixture over 30 min,
until the yellow color of the solution disappeared. The TLC
analysis indicated the nearly complete conversion of 4, and so
the reaction mixture was transferred to a separatory funnel and
the aqueous phase was removed and washed with 3 × 100 mL
of CH2Cl2. The combined organic layers were dried with
Na2SO4, filtered, and the solvent removed by rotary
evaporation. When nearing complete evaporation, 200 mL of
hexanes were added to allow the crude product to precipitate
as an off-white solid. This was filtered onto a fritted funnel and
washed with small aliquots of hexanes. The crude was then
recrystallized twice in a minimum of hot ethanol, which gave
the product as white crystals (47.719g, 87%).
1H NMR (700 MHz; CDCl3): δ 9.26 (s, 1H), 7.67 (dd, J =
7.6, 1.5, 1H), 7.60−7.59 (m, 1H), 7.45−7.39 (m, 5H), 7.36
(dd, J = 5.1, 3.9, 1H), 6.92 (q, J = 9.8, 2H).
HPLC purity: 96.0% (area %).
Dibenzo[b,f ]azocin-6(5H)-one (3). Compound 2
(100.088 g, 452.35 mmol) was added to a 1 L glass reactor
equipped with a paddle-type Teflon mechanical stirrer and
stirred at 400 rpm. To this, Eaton’s reagent (200 mL) was
added, and the reaction began turning a dark red color. The
reaction was heated in an oil bath to 100 °C for 30 min, at
which point TLC analysis (5% MeOH/CH2Cl2) indicated
complete conversion. The hot reaction mixture was poured
directly onto 500 g of crushed ice and mixed with a glass
spatula to fully quench the Eaton’s reagent. The slurry was
then filtered through a glass fritted funnel and washed with 3 ×
400 mL of deionized water until the filtrate was neutral as
measured by pH paper. The filter cake was dried on the frit for
1 h, and the resulting caked material was dried using an in-
house fluidized bed drying apparatus. After 4 h of drying, the
product was collected as a gray powder (97.082 g, 97%).
1H NMR (600 MHz; DMSO-d6): δ 9.85 (s, 1H), 7.33−7.31
(m, 2H), 7.26−7.20 (m, 2H), 7.17−7.09 (m, 4H), 7.00 (d, J =
11.6, 1H), 6.89 (d, J = 11.6, 1H).
1H NMR (600 MHz; CDCl3): δ 7.26−7.24 (m, 5H), 7.17−
7.11 (m, 3H), 6.79 (d, J = 12.9, 1H), 6.61 (d, J = 12.9, 1H),
5.51 (d, J = 15.1, 1H), 4.25 (d, J = 15.1, 1H), 3.61 (s, 3H),
2.62−2.57 (m, 1H), 2.48−2.39 (m, 2H), 2.03−1.98 (m, 1H).
HPLC purity: 99.3% (area %).
4-Dibenzo[b,f ]azoncin-5(6H)-yl-4-oxobutanoic Acid
(7). Compound 6 (54.637 g, 170.01 mmol) was added to a
300 mL round-bottom flask and then suspended in 165 mL of
a 2:1 mixture of MeOH/H2O. LiOH·xH2O (24.440 g, 1020
HPLC purity: 96.0% (area %).
D
DOI: 10.1021/acs.oprd.9b00406
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
mmol) was added, and the reaction vessel fitted with a reflux
condenser. The flask was purged with N2 gas for 10 min and
then heated to reflux. After 30 min, TLC analysis (3:2 EtOAc/
Hex) indicated the complete conversion to 7, and the reaction
mixture was acidified with 300 mL of 1 M HCl. This was then
extracted with 3 × 200 mL of CH2Cl2, dried with MgSO4,
filtered, and the solvent removed by rotary evaporation. To
obtain a convenient free-flowing crystalline product, the
material was redissolved in 100 mL of CH2Cl2 and 100 mL
of hexanes was added. Solvent was removed by rotary
evaporation without heating, and as the CH2Cl2 was removed
and 7 precipitated as fine white crystals. The product was then
filtered through a Buchner funnel and washed with small
aliquots of hexanes to give 7 as fine white crystals (51.723 g,
99%).
HPLC purity: 97.3%.
4-Methoxy-4-oxobutanoic Acid. To a 500 mL round-
bottom flask equipped with a stir bar and reflux condenser,
succinic anhydride (68.198 g, 681.5 mmol) and methanol (145
mL, 3.584 mol) were added, along with DMAP (0.850 g, 6.958
mmol). These were stirred for 15 min at room temperature, by
which time the succinic anhydride was fully dissolved and the
reaction exotherm had heated the solution to boiling. Upon
boiling, the reaction mixture was cooled in an ice−water bath,
although reflux continued for several minutes. After a further
15 min, 1H NMR spectroscopy indicated the full conversion to
the product, and the reaction mixture was left to fully cool
down. This was then washed through ∼5 cm packed bed of
DOWEX 50WX2-100 to capture the DMAP catalyst. Solvent
was removed from the filtrate by rotary evaporation then dried
in vacuo to give the product as white crystals (65.2 g, 73%).
1H NMR (600 MHz; CDCl3): δ 2.67 (ddd, J = 7.3, 6.1, 1.2,
2H), 2.61 (ddd, J = 7.3, 6.1, 1.2, 2H).
1H NMR (600 MHz; CDCl3): δ 7.30−7.22 (m, 5H), 7.18−
7.11 (m, 3H), 6.80 (d, J = 12.8, 1H), 6.61 (d, J = 12.8, 1H),
5.52 (d, J = 15.1, 1H), 4.29 (d, J = 15.1, 1H), 2.61 (ddd, J =
16.8, 8.2, 5.6, 1H), 2.51 (ddd, J = 16.8, 7.0, 5.6, 1H), 2.41
(ddd, J = 17.0, 8.2, 5.5, 1H), 2.04 (ddd, J = 17.0, 6.8, 5.8, 1H).
HPLC purity: 99.6%.
4-Chloro-4-methyl-oxobutyrate (5). To a 500 mL
round-bottom flask equipped with a stir bar, 4-methoxy-4-
oxobutanoic acid (71.642 g, 542.2 mmol) was added. Using an
addition funnel, thionyl chloride (52 mL, 705 mmol) was
added dropwise to the carboxylic acid. After completing the
addition of the thionyl chloride, the addition funnel was
removed and replaced with a water condenser. The condenser
was capped with a rubber septum, and a teflon tube was run
from the top of the condenser to an Erlenmeyer flask with 500
L of 4 M KOH to quench the HCl and SO2 emitted from the
reaction. The reaction mixture was then heated to 50 °C for 40
min, at which time the reaction mixture had ceased bubbling.
After the cessation of bubbling, the reaction was left to cool to
room temperature for 1 h, at which point the condenser was
removed and a 30 cm Vigreaux column and short path
distillation head were affixed to the flask. The thionyl chloride
was distilled off, and once no more thionyl chloride could be
removed, 60 mL of dry toluene was added and used to co-
distill residual thionyl chloride. This was repeated three total
times. Residual toluene was then distilled off, and the product
was distilled at reduced pressure (∼1 mbar) to give it as a clear
colorless liquid (74.353 g, 91%).
4-[11,12-Didehydrodibenzo[b,f ]azocin-5(6H)-yl]-4-
oxobutanoic Acid (DBCO−COOH) (9). Compound 7
(15.003 g, 48.814 mmol) was added to a 500 mL round-
bottom flask and then dissolved in 200 mL of CH2Cl2 and
cooled to 0 °C. Separately, Br2 (3.251 mL, 63.4 mmol) was
dissolved in 10 mL of CH2Cl2, and this was added dropwise to
1
the reaction mixture. After 20 min of reaction time, H NMR
analysis indicated the absence of the alkene protons at ∼6.7
ppm, and the reaction was worked up by quenching with 1 mL
of saturated Na2SO3 and then washed with 3 × 100 mL of
water. The organic layer was washed with 1 × 100 mL of 1 M
H3PO4, 1 × 100 mL brine, dried with MgSO4, solvent was
removed by rotary evaporation then dried in vacuo for 1 h to
give 8 as a slightly green foam (22.744 g, 99%).
Since compound 8 is unstable, it should be eliminated on
the same day as it is prepared. The flask containing 8 was
charged with 100 mL of anhydrous THF. The flask was then
cooled to −78 °C in a dry ice−acetone bath for 5 min with
stirring. Fifity milliliters of a 4 M solution of anhydrous KOtBu
in THF was separately chilled to −78 °C in a dry ice−acetone
bath, and this was cannulated into the solution of 8. The
1H NMR (600 MHz; CDCl3): δ 3.20 (t, J = 6.6, 2H), 2.67
(t, J = 6.6, 2H).
1
reaction was stirred at −78 °C for 1 h, at which point, H
NMR spectroscopy and HPLC analysis indicated full
conversion to 9. The reaction mixture was then diluted with
200 mL of diethyl ether and washed with 3 × 200 mL 1 M
H3PO4, 1 × 100 mL brine, dried with MgSO4, and filtered.
Solvent was removed by rotary evaporation without a heating
bath, as the cooling of the flask is essential to isolate 9. Once
most of the solvent has been removed, 9 began to precipitate
into the solution as off-white crystals. At the onset of crystal
formation, the flask was removed from the rotary evaporator
and 200 mL of diethylether was added, and the product was
left to crystallize at −20 °C overnight. The product was then
filtered through a Buchner funnel and then washed with small
aliquots of ice-cold ether, giving 9 as white crystals (7.942 g,
53%).
ASSOCIATED CONTENT
■
S
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.oprd.9b00406.
1H NMR spectra and HPLC traces for each compound
synthesized (PDF)
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: adronov@mcmaster.ca.
1H NMR (600 MHz; DMSO-d6): δ 7.66 (dd, J = 7.7, 1.2,
1H), 7.62 (d, J = 7.3, 1H), 7.52−7.45 (m, 3H), 7.38 (td, J =
7.5, 1.5, 1H), 7.34 (td, J = 7.4, 1.2, 1H), 7.29 (dd, J = 7.5, 1.3,
1H), 5.03 (d, J = 14.1, 1H), 3.62 (d, J = 14.0, 1H), 2.59 (dt, J =
16.7, 7.1, 1H), 2.30 (ddd, J = 17.2, 7.5, 6.5, 1H), 2.19 (dt, J =
17.2, 6.6, 1H), 1.78 (dt, J = 16.7, 6.5, 1H).
ORCID
Stuart A. McNelles: 0000-0003-0855-3307
Alex Adronov: 0000-0002-0770-3118
Notes
The authors declare no competing financial interest.
E
DOI: 10.1021/acs.oprd.9b00406
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Migrations during an Asymmetric Intermolecular Heck Reaction. J.
Org. Chem. 2007, 72, 7253−7259.
ACKNOWLEDGMENTS
■
The authors gratefully acknowledge the Natural Science and
Engineering Research Council of Canada (NSERC) for
generous financial support. S.A.M. is grateful for support
through the Ontario Graduate Scholarships (OGS) program.
REFERENCES
■
(1) Blomquist, A. T.; Liu, L. H. Many-Membered Carbon Rings. VII.
Cyclooctyne. J. Am. Chem. Soc. 1953, 75, 2153−2154.
(2) Agard, N. J.; Prescher, J. A.; Bertozzi, C. R. A Strain-Promoted
[3 + 2] Azide-Alkyne Cycloaddition for Covalent Modification of
Biomolecules in Living Systems. J. Am. Chem. Soc. 2004, 126, 15046−
15047.
(3) Thirumurugan, P.; Matosiuk, D.; Jozwiak, K. Click Chemistry for
Drug Development and Diverse Chemical-Biology Applications.
Chem. Rev. 2013, 113, 4905−4979.
(4) van Hest, J. C. M.; van Delft, F. L. Protein Modification by
Strain-Promoted Alkyne-Azide Cycloaddition. ChemBioChem 2011,
12, 1309−1312.
(5) Kurra, Y.; Odoi, K. A.; Lee, Y. J.; Yang, Y.; Lu, T.; Wheeler, S. E.;
Torres-Kolbus, J.; Deiters, A.; Liu, W. R. Two Rapid Catalyst-Free
Click Reactions for in Vivo Protein Labeling of Genetically Encoded
Strained Alkene/Alkyne Functionalities. Bioconjugate Chem. 2014, 25,
1730−1738.
(6) Jang, S.; Sachin, K.; Lee, H. J.; Kim, D. W.; Lee, H. S.
Development of a Simple Method for Protein Conjugation by
Copper-Free Click Reaction and Its Application to Antibody-Free
Western Blot Analysis. Bioconjugate Chem. 2012, 23, 2256−2261.
(7) Xu, J.; Filion, T. M.; Prifti, F.; Song, J. Cytocompatible
Poly(Ethylene Glycol)-Co-Polycarbonate Hydrogels Cross-Linked by
Copper-Free, Strain-Promoted Click Chemistry. Chem. - Asian J.
2011, 6, 2730−2737.
(8) Ono, R. J.; Lee, A. L. Z.; Voo, Z. X.; Venkataraman, S.; Koh, B.
W.; Yang, Y. Y.; Hedrick, J. L. Biodegradable Strain-Promoted Click
Hydrogels for Encapsulation of Drug-Loaded Nanoparticles and
Sustained Release of Therapeutics. Biomacromolecules 2017, 18,
2277−2285.
(9) Hodgson, S. M.; McNelles, S. A.; Abdullahu, L.; Marozas, I. A.;
Anseth, K. S.; Adronov, A. Reproducible Dendronized PEG
Hydrogels via SPAAC Cross-Linking. Biomacromolecules 2017, 18,
4054−4059.
(10) Baskin, J. M.; Prescher, J. A.; Laughlin, S. T.; Agard, N. J.;
Chang, P. V.; Miller, I. A.; Lo, A.; Codelli, J. A.; Bertozzi, C. R.
Copper-Free Click Chemistry for Dynamic in Vivo Imaging. Proc.
Natl. Acad. Sci. U.S.A. 2007, 104, 16793−16797.
(11) Fong, D.; Yeung, J.; McNelles, S. A.; Adronov, A. Decoration of
Polyfluorene-Wrapped Carbon Nanotubes via Strain-Promoted
Azide−Alkyne Cycloaddition. Macromolecules 2018, 51, 755−762.
(12) Kardelis, V.; Chadwick, R. C.; Adronov, A. Click Functionaliza-
tion of a Dibenzocyclooctyne-Containing Conjugated Polyimine.
Angew. Chem., Int. Ed. 2016, 55, 945−949.
́
(13) Burk, M.; Rothstein, S.; Dube, P. Enabling the Multigram
Synthesis of (2-Cyclooctyn-1-Yloxy)Acetic Acid. Org. Process Res. Dev.
2018, 22, 108−110.
(14) Debets, M. F.; van Berkel, S. S.; Schoffelen, S.; Rutjes, F. P. J.
T.; van Hest, J. C. M.; van Delft, F. L. Aza-Dibenzocyclooctynes for
Fast and Efficient Enzyme PEGylation via Copper-Free (3+2)
Cycloaddition. Chem. Commun. 2010, 46, 97−99.
(15) Kuzmin, A.; Poloukhtine, A.; Wolfert, M. A.; Popik, V. V.
Surface Functionalization Using Catalyst-Free Azide−Alkyne Cyclo-
addition. Bioconjugate Chem. 2010, 21, 2076−2085.
(16) Chadwick, R. C.; Van Gyzen, S.; Liogier, S.; Adronov, A.
Scalable Synthesis of Strained Cyclooctyne Derivatives. Synthesis
2014, 46, 669−677.
(17) Pirrung, M. C. The Synthetic Organic Chemist’s Companion;
John Wiley & Sons, Ltd: Hoboken, New Jersey, 2006.
(18) Wheatley, B. M. M.; Keay, B. A. Use of Deuterium Labeling
Studies to Determine the Stereochemical Outcome of Palladium
F
DOI: 10.1021/acs.oprd.9b00406
Org. Process Res. Dev. XXXX, XXX, XXX−XXX