Continuous flow macrocyclization at high concentrations: Synthesis of macrocyclic lipids

Article abstract of DOI:10.1039/c3gc40872h

A phase separation/continuous flow macrocyclization protocol eliminates the need for high-dilution conditions and can be used to prepare gram quantities of biologically relevant macrocyclic lipid structures. The method presents several green advantages towards macrocycle synthesis: (1) the prevention of unwanted oligomers and waste, (2) a reduction in the large quantities of toxic, volatile organic solvents and (3) the use of PEG as an environmentally benign reaction media. Macrocycles could be synthesized in high yields (up to 99%) in short reaction times (1.5 h) and on gram scales without the need to alter the reaction conditions.

Full text of DOI:10.1039/c3gc40872h

Green Chemistry
View Article Online
View Journal | View Issue
PAPER
Continuous flow macrocyclization at high
concentrations: synthesis of macrocyclic lipids†
Cite this: Green Chem., 2013, 15, 1962
Anne-Catherine Bédard, Sophie Régnier and Shawn K. Collins*
A phase separation/continuous flow macrocyclization protocol eliminates the need for high-dilution con-
ditions and can be used to prepare gram quantities of biologically relevant macrocyclic lipid structures.
The method presents several green advantages towards macrocycle synthesis: (1) the prevention of
unwanted oligomers and waste, (2) a reduction in the large quantities of toxic, volatile organic solvents
and (3) the use of PEG as an environmentally benign reaction media. Macrocycles could be synthesized in
high yields (up to 99%) in short reaction times (1.5 h) and on gram scales without the need to alter the
reaction conditions.
Received 8th March 2013,
Accepted 20th May 2013
DOI: 10.1039/c3gc40872h
www.rsc.org/greenchem
Continuous flow strategies for green chemical synthesis have The reaction rates of the macrocyclizations reported were accel-
impacted every area of organic synthesis, from academic to erated through judicious choice of ligands and the exploitation
industrial research.1 Flow chemistry represents a powerful of copper tubing as the flow reactor. The efficient heat transfer
technology whose advantages include precise control of reac- that can be achieved in flow would seem to provide a solution
tion time, temperature, concentration and stoichiometry. Con- to the challenge of macrocyclization efficiency, however the
sequently, many of the principles of green chemistry are challenge of reaction dilution is often more daunting.4
fulfilled such as reduced energy requirements, minimized
Consequently, our group has recently reported a general,
exposure to hazardous chemicals or intermediates and efficient and green macrocyclization protocol via oxidative
reduced amounts of unwanted by-products and waste. Given Glaser–Hay coupling at high concentrations through the use of
these advantages, flow chemistry is an ideal technique to use a “phase separation” strategy (Scheme 1).5 The concentration
in tackling the challenges associated with macrocycle syn- effects of the macrocyclization reaction were controlled
thesis. Macrocycles are important structural motifs found in through the aggregation properties of poly(ethylene)glycol400
compounds with applications in pharmaceutical, agrochemi- (PEG400). The “phase separation” technique allowed for signifi-
cal, cosmetic and material sciences.2 Their synthesis is chal- cantly higher concentrations (150–500×), a reduction of the
lenging, often requiring significant amounts of solvent to amounts of toxic organic solvents that are eventually converted
control concentration effects and prevent the formation of to waste, and replacement of the majority of the volatile
undesirable oligomers. When the dilution requirements and organic solvents by the environmentally benign PEG. Given
inconvenient set-up are considered, macrocyclization reactions the advantages presented by the “phase separation” strategy,
in general would benefit from improved protocols employing we sought to apply it to the development of a continuous flow
flow techniques.
process to synthesize important medicinally relevant classes of
Macrocyclization via continuous flow presents several chal-
lenges, including: (1) the need to accelerate the normally slow
cyclization reactions to improve their efficiency and most impor-
tantly, (2) the need to prevent oligomerization, as these by-
products have low solubilities and could block the flow reactor.
Recently, James and co-workers have reported the first flow-
macrocyclization reaction that involved the synthesis of con-
strained macrocycles via alkyne–azide cycloaddition chemistry.3
Département de Chimie, Centre for Green Chemistry and Catalysis, Université de
Montréal, CP 6128 Station Downtown, Montréal, Québec, Canada, H3C 3J7.
E-mail: shawn.collins@umontreal.ca
†Electronic supplementary information (ESI) available: Experimental procedure
and characterization data for all new compounds. See DOI: 10.1039/c3gc40872h
Scheme 1 Advantages of the “phase separation” strategy in macrocyclization.
1962 | Green Chem., 2013, 15, 1962–1966
This journal is © The Royal Society of Chemistry 2013

View Article Online
Green Chemistry
Paper
macrocycles. Consequently, herein we report on the develop- control. It was found that by using lower flow rates, pressure
ment of efficient continuous flow macrocyclization for the control was easily maintained and helped to provide more
preparation of macrocyclic lipids that mimic those found in reproducible results. Gratifyingly, it was found that flow rates
Archaeal membranes.6
of 0.22 mL min−1 and a residence time of 1.5 h afforded a
The investigations into a continuous flow/macrocyclization nearly quantitative yield (97%) of the macrocyclic product 4.
protocol began by transposing the previously optimized reac- The high yield of macrocycle 4 achieved at relatively high con-
tion conditions for the Glaser–Hay coupling utilizing micro- centration ([0.03 M]) denotes the high efficiencies that are
wave heating.7 The preliminary studies utilized the formation possible when combining the two green chemistry techniques
of macrolactone 4, as it provides a valid comparison for the of phase separation and continuous flow synthesis. The calcu-
continuous flow protocol versus other existing macrocyclization lated Emac (a recently proposed meter for grading the efficiency
strategies (Table 1). TMEDA was chosen as the ligand as its of macrocyclizations)9 of ∼7.4 for the macrocyclization (3→4)
bidentate nature provides increased solubility and stability of is high and similar to those obtained in James’ previously
the resulting transition metal complexes at high temperatures.5 reported continuous flow synthesis of constrained macrocycles
Intermolecular Glaser–Hay couplings have been previously via alkyne–azide cycloadditions.
studied in continuous flow settings, where semi-permeable
In an effort to evaluate the generality of the optimized con-
Teflon AF-2400 membranes8 were used to increase contact with tinuous flow conditions, the macrocyclization of three other
oxygen gas. In the present study, the solutions for macrocycliza- diynes was performed (Table 2, entries 2→4). In each case, the
tion were only sparged with O2 for 5 min before injection into results of the cyclization using the continuous flow conditions
the flow apparatus. First, we investigated the effect of the temp- were compared with the isolated yields obtained utilizing the
erature on the yields of the desired 21-membered macrolactone previously developed microwave heating protocol. First, the
4 (Table 1, entries 1→4). When the reaction was performed at isolated yield of the 21-membered macrolactone 4 obtained via
the same temperature as in the microwave heated reaction continuous flow (97%) was found to be much better than that
(120 °C), 4 was formed in higher yield using a flow strategy obtained via microwave heating (81%). Given the excellent
(85% vs. 81%, entry 3). Lower temperatures resulted in lower yield obtained for 4, the macrocyclization to afford 4 was
conversions and isolated yields. Increasing the temperature scaled to 1 mmol and a 93% yield was obtained.
(140 °C) resulted in a decrease of selectivity (macrocyclization
The macrocyclization of other diynes was then investigated.
vs. oligomerization) as the yield of 4 decreased (72%) and oligo- When the cyclization to form the 20-membered macrocyclic
mers could be observed. Changing the ratio of PEG400/MeOH diester 5 was investigated, a 72% yield of the product was
(Table 1, entries 4→6) did not result in significant changes in obtained utilizing continuous flow. A much lower yield (47%)
isolated yield, although larger ratios of PEG result in better solu- of 5 was obtained using the microwave heating protocol. The
bility of the organic substrates. Higher ratios of PEG400/MeOH preparation of macrocyclic ether 6 was initially problematic, as
were not explored as previous work has indicated that catalyst its corresponding diyne precursor was partially insoluble in
inhibition becomes problematic at higher ratios.5
The influence of reaction time and flow/rate were also inves- ments demonstrated very little difference in yield between
tigated (Table 1, entries 7→9). Due to the viscosity of PEG400 PEG400/MeOH (2 : 1) and (1 : 1) mixtures, the cyclization to
the PEG400/MeOH (1 : 1) mixture. However, as previous experi-
,
high flow rates were found to cause irregularities in pressure form 6 was conducted using the higher ratio of PEG400/MeOH
(2 : 1). The macrocycle 6 was once again obtained in higher
yields (71%) when using the continuous flow strategy.
However, it was noted that when the synthesis of a smaller and
significantly more strained macrocycle 2 was investigated, the
trend reversed. The 16-membered macrocycle 2 was obtained
in higher yield using microwave heating (75%) versus the con-
tinuous flow (58%).
Table 1 Optimization of reaction conditions for macrocyclic Glaser–Hay coup-
ling of 3
Encouraged by the preliminary optimization of a continu-
ous flow-macrocyclization protocol employing a “phase sepa-
ration” strategy, the optimal conditions were also explored
with a macrocyclic lipid precursor 7 which bears a protected
glycerol motif (Table 3). The macrocyclic Archaeal lipids are
typically composed of polyprenyl chains with glycerol or polyol
head groups.6 Within the structure of the lipids, macrocycliza-
tion can occur through the alkyl chains of the same head
group or between two lipids, producing dimers. As such, 36- or
72-membered rings can be formed.10 Recently, much attention
has been given to macrocyclic lipids for potential anti-cancer
activities or in the design of novel vehicles for liposomal drug
delivery.11 When the benzyl-protected glycerol derivative 7 was
PEG400
/
Flow rate (mL min−1);
Entry Temp. (°C) MeOH (ratio) residence time (h)
Yield (%)
1
2
3
4
5
6
7
8
9
80
100
120
140
120
120
120
120
120
2 : 1
2 : 1
2 : 1
2 : 1
1 : 1
1 : 2
1 : 1
1 : 1
1 : 1
5; 3
5; 3
2225; 3
5; 3
5; 3
74
73
85
72
86
82
96
91
97
5; 3
5; 1.5
1; 1.5
0.22; 1.5
This journal is © The Royal Society of Chemistry 2013
Green Chem., 2013, 15, 1962–1966 | 1963

View Article Online
Paper
Green Chemistry
subjected to the optimized conditions from Table 2, it was
found that 7 was poorly soluble in the 1 : 1 PEG400/MeOH
solvent mixture (Table 3, entry 1). However, when the ratio of
PEG400/MeOH was increased (2 : 1), the 16-membered macro-
cyclic lipid 9 was isolated in 71% yield. Variations in the reac-
tion temperature followed the same trends that were observed
for the synthesis of macrolactone 4 (Table 3). Lower tempera-
tures resulted in lower conversions and isolated yields,
although excellent selectivity was observed and the remaining
mass balance was recovered starting diyne 7 (Table 3, entry 3).
Increasing the reaction temperature once again lowered selecti-
vity, resulting in the formation of oligomers and decreased iso-
lated yields (63%).
Table 2 Macrocyclic lactones and ethers synthesized via continuous flow-
macrocyclization using “phase separation”
Entry
1
Product
Yield: flow (%)
97a
Yield: microwave (%)
81
With optimized continuous flow conditions in hand for the
synthesis of macrocyclic lipids, the macrocyclization of 3 other
substrates were performed (Table 4). The isolated yield of the
16-membered macrolipid 8 was found to be slightly higher
2
58
75
Table 4 Macrocyclic lipids synthesized via continuous flow-macrocyclization
using “phase separation”
3
4
72
47
65
71b
Entry
1
Product
Yield: flow (%)
71a
Yield: microwave (%)
65
a On a 1 mmol scale, 4 was isolated in 93% yield. b The corresponding
diyne precursor of 7 was insoluble in PEG400/MeOH mixtures. As such,
a PEG400/MEOH (2 : 1) ratio was used.
2
3
4
78
99
45
62
92
42
Table 3 Optimization of reactions conditions for the synthesis of a protected
macrocyclic lipid 8
Entry
Temp. (°C)
PEG400/MeOH (ratio)
Yield (%)
a
1
2
3
4
120
120
100
140
1 : 1
2 : 1
2 : 1
2 : 1
—
71
66
63
a Diyne 7 precipitates from the solution when a PEG400/MeOH (1 : 1) a When cyclization was performed on 3.8 mmol scale, the macrocycle
ratio is used.
was obtained in 66% yield.
1964 | Green Chem., 2013, 15, 1962–1966
This journal is © The Royal Society of Chemistry 2013

View Article Online
Green Chemistry
Paper
under continuous flow conditions (71%) compared to cycliza-
tion under microwave heating (65%) (Table 4, entry 1). As one
of the advantages of continuous flow methods is the facile
scale-up, the synthesis of 8 was repeated on ∼4 mmol scale.
The yield of the 16-membered macrolipid 8 did not change sig-
nificantly (66% isolated yield on ∼4 mmol scale). Similar
macrolipids based upon a glycerol head group were then pre-
pared having different ring sizes. When the 26-membered
macrocycle 9 was prepared via continuous flow, the isolated
yield of 78% was again higher than what was obtained via
microwave heating (62%). Similarly, the 30-membered macro-
cycle 10 was isolated in 99% yield after synthesis via continu-
ous flow (92% isolated yield was obtained using microwave
heating). As many macrocyclic lipids can be prepared with
ester linkages, the macrocycle 11 was prepared by continuous
flow synthesis and was isolated in 45% yield. In all of the
macrocyclizations, the products were obtained in higher yield
when using continuous flow versus microwave heating. In
general, the yields of the larger 21-, 26- and 30-membered
macrocyclic lipids were higher than smaller macrocycles,
perhaps due to the ring strain caused by the incorporation of a
linear 1,3-diyne within the cyclic motif.
Scheme 2 Synthesis of a dimeric macrocyclic lipid 13 as a mixture of head-to-
tail and head-to-head isomers.
In an effort to further improve the continuous flow process,
the use of other PEG solvents were investigated with the goal
of reducing the catalyst loadings (Table 5).5 When the diyne 7
had been previously cyclised under optimum conditions using
25 mol% of catalysts and PEG400 as a co-solvent, the yield of
the corresponding lipid macrocycle 8 was 71%. When the
solvent PPG425 was used as a substitute for PEG400 at identical
catalyst loadings a similar yield was observed (Table 5, entry 2).
However, when the catalyst loading was dropped to 10 mol%,
continuous flow conditions could be developed so that the
desired macrocycle 8 could be isolated in 92% yield. These pre-
liminary results demonstrate that the catalyst loadings could
be lowered and the yields increased through judicious choice
of the PEG co-solvent in the continuous flow conditions.
Further reductions in the catalyst loading to 5 mol% also pro-
vided good yields of macrocycle 8 (78%) and excellent selecti-
vity for macrocyclization was observed, as the remaining mass
Scheme 3 Conversion of Bn-protected macrocycle 8 into novel macrocyclic
phosphonate containing lipid 14.
balance was recovered diyne 7. Longer reaction times may be
necessary for effective use of reduced catalyst loadings.
As dimeric macrocyclic lipids are found in Nature,6 the
macrocyclization of the diyne 12 was investigated under the
optimized reaction conditions (Scheme 2). Acyclic diyne 12
cannot undergo intramolecular cyclization due to ring strain
but selectively forms the dimer 13 in 55% yield via continuous
flow (46% using microwave heating).
To demonstrate the utility of the novel macrocycles pre-
pared in Table 5 as lipids, the synthesis of the 16-membered
macrocyclic diyne 8 was performed on a multigram scale
(∼4 mmol) and similar yields were obtained even when the
reaction was scaled by a factor of 30× (Scheme 3). Upon iso-
lation, macrocyclic lipid 8 was functionalized with a polar
phosphonate group. A global hydrogenation (Pd/C, H2 (1 atm))
of macrocycle 8 results in cleavage of the Bn protecting group
and exhaustive hydrogenation of the 1,3-diyne to afford the
corresponding saturated macrocycle in 52% yield. Subsequent
phosphonation provided the saturated macrocyclic lipid 14 in
58% yield.
Table 5 Reduction of catalyst loadings when using PPG425 for the synthesis of
a protected macrocyclic lipid 8
Conclusions
Flow rate (mL min−1);
residence time (h)
Yield
(%)
The above studies demonstrate that a continuous flow-macro-
cyclization protocol utilizing “phase separation” eliminates the
need for extremely high-dilution conditions and can be used
to prepare meaningful quantities of biologically relevant
macrocyclic lipid structures. The combination of the “phase
Entry
Mol%
PEG
1
2
3
4
25
25
10
5
PEG400
PPG425
PPG425
PPG425
0.22; 1.5
0.22; 1.5
0.11; 3
71
65
92
78
0.11; 3
This journal is © The Royal Society of Chemistry 2013
Green Chem., 2013, 15, 1962–1966 | 1965

View Article Online
Paper
Green Chemistry
separation” strategy with a continuous flow synthesis presents
several green advantages towards macrocycle synthesis: (1) the
precise control of reaction time and temperature prevent the
formation of unwanted oligomers and waste, (2) the high con-
centrations reduce the large quantities of solvents that would
have to be disposed of in environmentally damaging processes
and (3) the use of PEG as a solvent represents a non-volatile
and non-toxic alternative to traditional organic solvents.
Macrocycles could be synthesized in high yields (up to 99%),
in short reaction times (1.5 h) and on large scales without the
need to alter the reactions conditions. Preliminary results
demonstrate that reduced catalyst loadings may be possible
when using modified PEG co-solvents. The demonstration of
the viability of the phase separation/continuous flow synthetic
strategy should find applicability in the synthesis of other
macrocycles in academia and industry.
used in flow see: M. Nagel, G. Frater and H.-J. Hansen,
Chimia, 2003, 57, 196.
5 A.-C. Bédard and S. K. Collins, J. Am. Chem. Soc., 2011, 133,
19976; A.-C. Bédard and S. K. Collins, Chem. Commun.,
2012, 48, 6420; A.-C. Bédard and S. K. Collins, Chem.–Eur.
J., 2013, 19, 2108; A.-C. Bédard and S. K. Collins, ACS
Catal., 2013, 3, 773.
6 M. De Rosa and A. Gambacorta, Prog. Lipid Res., 1988, 27,
153; M. Kates, Membrane Lipids of Archaeal, Elsevier,
Oxford, 1993, p. 261; P. B. Comita, R. B. Gagosian, H. Pang
and C. E. Costello, J. Biol. Chem., 1984, 259, 15234.
7 T. N. Glasnov and C. O. Kappe, Chem.–Eur. J., 2011, 17,
11956.
8 T. P. Petersen, A. Polyzos, M. O’Brien, T. Ulven,
I. R. Baxendale and S. V. Ley, ChemSusChem, 2012, 5, 274.
9 An Emac above ∼7.1 suggests a macrocyclization is within
the top 10% of the most efficient macrocyclizations
reported to date. For a discussion on grading the efficiency
of macrocyclization reactions through the use of the Emac
factor see: J. C. Collins and K. James, MedChemComm,
2012, 3, 1489.
10 T. Eguchi, T. Terachi and K. Kakinuma, J. Chem. Soc.,
Chem. Commun., 1994, 137; T. Eguchi, K. Arakawa,
T. Terachi and K. Kakinuma, J. Org. Chem., 1997, 62, 1924;
K. Arakawa, T. Eguchi and K. Kakinuma, Chem. Lett., 1998,
901.
The authors acknowledge the Natural Sciences and Engi-
neering Research Council of Canada (NSERC), Université de
Montreal and the Centre for Green Chemistry and Catalysis
(CGCC) for generous funding. A.-C. B. thanks NSERC (Vanier
Graduate Scholarship) and CGCC for graduate scholarships,
and S. R. thanks CGCC for an undergraduate fellowship.
Notes and references
11 H. Liu, C. E. Olsen and S. B. Christensen, J. Nat. Prod.,
2004, 67, 1439; T. Eguchi, K. Arakawa, K. Kakinuma,
G. Rapp, S. Ghosh, Y. Nakatani and G. Ourisson,
Chem.–Eur. J., 2000, 6, 3351; O. Dannenmuller, K. Arakawa,
T. Eguchi, K. Kakinuma, S. Blanc, A.-M. Albrecht,
M. Schmutz, Y. Nakatani and G. Ourisson, Chem.–Eur. J.,
2000, 6, 645; W. P. D. Goldring, E. Jubeli, R. A. Downs,
A. J. S. Johnston, N. Abdul Khalique, L. Raju, D. Wafadari
and M. D. Pungente, Bioorg. Med. Chem. Lett., 2012, 22,
4686; M. Shibakami, S. Miyoshi, M. Nakamura and
R. Goto, Synlett, 2009, 2651; S. Bhattacharya and J. Biswas,
Langmuir, 2010, 26, 4642; Q.-D. Huang, H. Chen,
L.-H. Zhou, J. Huang, J. Wu and X.-Q. Yu, Chem. Biol. Drug
Des., 2008, 71, 224; M.-L. Miramon, N. Mignet and
J. Herscovici, J. Org. Chem., 2004, 69, 6949.
1 P. Watts and C. Wiles, J. Chem. Res., 2012, 36, 181; C. Wiles,
PharmaChem, 2011, 10, 15; L. Malet-Sanz and F. Susanne,
J. Med. Chem., 2012, 55, 4062.
2 J. C. Roxburgh, Tetrahedron, 1995, 51, 9767; E. M. Driggers,
S. P. Hale, J. Lee and N. K. Terrett, Nat. Rev. Drug Discovery,
2008, 7, 608; E. Marsault and M. L. Peterson, J. Med. Chem.,
2011, 54, 1961; H. Matsuda, S. Watanabe and K.
Yamamoto, Chem. Biodiversity, 2004, 1, 1985.
3 A. R. Bogdan, S. V. Jerome, K. N. Houk and K. James, J. Am.
Chem. Soc., 2012, 134, 2127; A. R. Bogdan and K. James,
Org. Lett., 2011, 13, 4060; A. R. Bogdan and K. James,
Chem.–Eur. J., 2010, 16, 14506.
4 For an alternative synthetic route employing ring expansion
to similar macrocycles that avoid high dilution and can be
1966 | Green Chem., 2013, 15, 1962–1966
This journal is © The Royal Society of Chemistry 2013