Synthesis of annulated pyridines by intramolecular inverse-electron-demand hetero-diels-alder reaction under superheated continuous flow conditions

Article abstract of DOI:10.1002/ejoc.201101538

Pyrimidine alkynes can be transformed into the corresponding annulated pyridines efficiently in flow. The superheating of organic solvents far beyond their boiling point enables toxic and difficult to workup solvents such as nitrobenzene or chlorobenzene,

Full text of DOI:10.1002/ejoc.201101538

SHORT COMMUNICATION
DOI: 10.1002/ejoc.201101538
Synthesis of Annulated Pyridines by Intramolecular Inverse-Electron-Demand
Hetero-Diels–Alder Reaction under Superheated Continuous Flow Conditions
Rainer E. Martin,*[a] Falk Morawitz,[a] Christoph Kuratli,[a] André M. Alker,[b] and
Alexander I. Alanine[a]
Keywords: Cycloaddition / Nitrogen heterocycles / Flow chemistry / Superheated solvents
Pyrimidine alkynes can be transformed into the correspond-
ing annulated pyridines efficiently in flow. The superheating
of organic solvents far beyond their boiling point enables
toxic and difficult to workup solvents such as nitrobenzene
or chlorobenzene, which are usually employed for these re-
actions, to be replaced by less harmful ones like toluene. The
relative rate of reactivity for a series of structurally close
starting materials was investigated and a scalable flow pro-
cess was developed, providing facile access to a series of
novel annulated pyridine building blocks. The effect of ther-
mal volume expansion of solvents under superheated condi-
tions was found to be significant and influenced the resi-
dence times considerably. To obtain meaningful and accurate
residence times, flow rates need to be corrected for volume
expansion. To avoid confusion with residence times, tR, cal-
culated from nominal flow rates, we would propose to use for
such corrected residence times the term effective residence
time, tR,eff
.
Introduction
Results and Discussion
One of these “neglected reactions” that we have recently
become interested in and started to reinvestigate is the
transformation of pyrimidine alkynes of type 1a to the cor-
responding annulated pyridine systems 2a by an inverse-
electron-demand hetero-Diels–Alder (HDA) reaction and
subsequent cycloreversion reaction cascade (Scheme 1).[4]
Pyridines are one of the most important heterocyclic struc-
tural motifs and are found in numerous natural products,
active pharmaceuticals, and functional materials, and a
number of synthetic methodologies for their preparation
have been documented.[5] The HDA reaction is assumed to
occur via an intermediate tricyclic adduct which results
from an intramolecular [4+2] cycloaddition across the C2
and C5 position of the pyrimidine ring followed by subse-
quent elimination of hydrogen cyanide.[6]
Flow chemistry has attracted considerable attention over
recent years as an enabling and innovative technology that
significantly enhances the classical process window typically
accessible with standard batch laboratory equipment.[1]
Flow chemical approaches have been found to be advan-
tageous, as they allow not only for automation of opera-
tions, but also lead to better control of chemical processes
due to improved heat and mass transfer.[2] Unconventional
and harsh reaction conditions such as greatly elevated tem-
peratures and pressures can be generated easily, allowing
the superheating of organic solvents far beyond their boil-
ing point in a controlled and safe manner. Therefore, toxic
and difficult to workup solvents can be replaced by less
harmful, environmentally benign solvents. Furthermore, re-
actions which were formerly avoided due to serious safety
concerns can now be conducted conveniently in flow, as
only small quantities of reactive materials or hazardous in-
termediates are within the reactor at a given time.[3] Flow
chemistry thus offers an ideal opportunity to revisit chal-
lenging, difficult to operate, or dangerous reactions that
have been underutilized in the past.
[a] Chemistry Technology and Innovation, Discovery Chemistry,
Pharmaceuticals Division, F. Hoffmann-La Roche Ltd
Grenzacherstrasse, 4070 Basel, Switzerland
Fax: +41-61-6886965
E-mail: rainer_e.martin@roche.com
[b] Biostructure Section, Discovery Technologies,
Pharmaceuticals Division, F. Hoffmann-La Roche Ltd
Grenzacherstrasse, 4070 Basel, Switzerland
Scheme 1. Classical batch synthesis of dihydro-5H-[1]pyridine 2a
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejoc.201101538.
by using nitrobenzene as solvent.[4]
Eur. J. Org. Chem. 2012, 47–52
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
47

R. E. Martin, F. Morawitz, C. Kuratli, A. M. Alker, A. I. Alanine
SHORT COMMUNICATION
Typically, these reactions are favored by strong electron- wash bottles equipped with a mixture of sodium hydroxide
withdrawing groups on the heteroaromatic azadiene and and sodium hypochlorite (bleach), which enabled this pro-
electron-rich dienophiles, with the presumed cycloaddition cess to be run safely even on a larger scale. A schematic
reaction being the rate-determining step in this sequence.[7] representation of the flow configuration is shown in Fig-
In cases where A ϶ C but a heteroatom such as O or N, ure 1. Due to the harsh reaction conditions required in
the reactivity of these systems is reduced due to both stereo- batch for this transformation we started the evaluation pro-
electronic and conformational reasons. On the one hand, cess at 210 °C up to the maximum temperature of 250 °C
heteroatoms donate additional electron density into the achievable in this system, varying the residence time tR from
aza-aryl system, and on the other hand, side chains con- 20 min to 50 min.
nected through heteroatoms (A ϶ C) adopt an unfavorable
According to 1H NMR spectroscopic analysis, heating of
in-plane conformation (the N–C–A–C torsion angle shows 1a in toluene at 210 °C for 20 min resulted in the formation
a strong preference around 0°), which is not optimally of the product in only 1%, which was improved consider-
aligned for the HDA reaction.[8] On the contrary, in cases ably to 49% by increasing the heating temperature to
with substituents A = C, SO2, or NAc, a productive, per- 250 °C. Extending the residence time at the same tempera-
pendicular orientation of the side chain to the azadiene ring ture to 50 min resulted in a further significant improvement
system is strongly favored, thereby allowing the reaction to nearly quantitative conversion of 96%. Encouraged by
partners to approach one another better. As a consequence, these optimization results we processed a 600 mg batch of
in cases with heteroatom linkers, rather high reaction tem- 1a, which provided, after aqueous workup and MPLC
peratures and extended reaction times are required to chromatography, desired product 2a in an excellent 94%
achieve acceptable conversion rates. Therefore these reac- isolated yield. On the basis of these results we aimed for a
tions have classically been conducted in high-boiling sol- further scale up to multigram quantities of 1a by using the
vents such as nitrobenzene, dichlorobenzene, or diphenyl same equipment and identical processing parameters. By
ether that are not only toxic, but also difficult to remove making use of both pumps of the Vapourtec R2+/R4 flow
during workup.[4,6] Moreover, during the second retro-HDA system and two SSTFR coils setup in parallel, the theoreti-
step, hydrogen cyanide is produced in stoichiometric cal processing time of 54 h was reduced down to 27 h. How-
amounts, which poses a considerable safety risk requiring ever, after an initial processing time of 6–8 h (system pres-
specific precautionary measures and thus further limits the sure of about 20 bar), a gradual pressure increase was ob-
scalability of this procedure within a conventional research served, which eventually led to a complete blockage of the
environment.
reactor coils within 2–4 h resulting in an automated shut-
Our initial efforts focused on the development of a stable down of the flow system as the maximum pressure of 50 bar
and scalable flow process for the conversion of chloropyr- was reached. Attempts to clean the reactor coils by repeated
imidine 1a into the corresponding dihydro-5H-[1]pyridine flushing with a number of different solvents of varying po-
2a. The Vapourtec R2+/R4 flow system equipped with a larity ranging from acetone to methanol and finally water
high-temperature 316 stainless steel tube flow reactor failed. In addition, dilute acids and bases were not success-
(SSTFR, 10 mL) and a 250 psi back-pressure regulator ful in dissolving the blockage either. Repeating the flow ex-
(BPR) was used to conduct this step, allowing the use of periment under the same reaction conditions by using a
toluene as a solvent safely under superheated conditions.[9] new reactor coil resulted in a similar observation with rapid
The starting material was introduced as a 0.14 m solution internal pressure onset after a short, initially stable, pro-
through
a reagent loop (1 mL) at a flow rate of cessing period. To identify the cause of the obstruction, a
200 μLmin–1. The crude reaction stream was collected in reactor tube was cut open and the content visually in-
an aqueous solution of sodium hydroxide and HCN was spected. A black polymer-like substance could be scratched
purged by a stream of nitrogen through two subsequent out, which was poorly soluble in both water and organic
Figure 1. Illustration of the flow reactor configuration used for the transformation of pyrimidine 1a into the corresponding annulated
pyridine 2a by employing the commercial Vapourtec R2+/R4 flow system.
48
www.eurjoc.org
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2012, 47–52

Synthesis of Annulated Pyridines
solvents and most likely is the polymerization product of ably preventing pressure increase and reactor fouling. In ad-
HCN eliminated during the retro-HDA reaction (see the dition, we used a short silica plug in an OmniFit column
Supporting Information). HCN is known to form sponta- to protect the back-pressure regulator from blockage. This
neously oligomers and polymers in nonaqueous solvents slight modification to the processing conditions allowed us
and in water even under much milder reaction conditions, to keep the process stable for several hours, enabling the
a process which is known to be very difficult to control.[10] production of 21 g (137 mmol) of building block 2a in a
Analysis of cross- and longitudinal section cuts of the reac- yield of 84% after MPLC purification. The structure of
tor revealed that the entire tube was not blocked by polyme- compound 2a was also confirmed by single-crystal X-ray
rization products, but that small plugs were formed ran- analysis (Figure 2).[11]
domly that started to aggregate and ultimately caused ob-
struction of the flow. Interestingly, there was no observed
damage to the stainless steel surface, as trace amounts of
water (toluene was not dried before use) in combination
with HCN might have caused corrosion of the reactor ma-
terial. Attempts to reduce polymer formation by addition
of bases such as triethylamine and isopropyl ethylamine or
acids like acetic acid were not successful. Moreover, intro-
ducing alternating washing cycles after 1 h processing time
to rinse the reactor coil did not help considerably to dimin-
ish the gradual pressure increase or to reduce the number
of blockage events. However, pentan-3-one [1% (v/v) in tol-
uene] was finally identified as the most effective HCN trap-
Figure 2. Single-crystal X-ray diffraction analysis of 3-chloro-6,7-
dihydro-5H-[1]pyridine (2a). The ORTEP drawing depicts thermal
ellipsoids at a 30% probability level.
On the basis of these encouraging results we decided to
ping agent (likely by cyanohydrin adduct formation), which further examine the scope and limitations of this interesting
enabled stable, continuous processing over many hours, reli- transformation on a number of closely related building
Table 1. Synthesis of annulated pyridines 2a–u in flow.
Entry
Starting
material
Product
A
R1
R2
R3
R4
R5
R6
% Yield[a]
1
2
3
4
5
6
7
8
1a
1b
1c
1d
1e
1f
1g
1h
1i
2a
2b
2c
2d
2e
2f
C
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
H
Cl
Cl
Cl
H
H
H
H
H
H
Cl
OCH3
Cl
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
CH3
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
CH3
H
H
H
H
H
H
TMS
CH3
H
H
H
H
H
H
H
H
H
H
H
H
H
95
70
75
31
52
65
60
64
Cl
Cl
Cl
H
H
H
H
H
H
H
H
H
H
CH3
Cl
Cl
H
H
H
F
2g
2h
2i
Cl
Br
NO2
CN
H
H
H
H
H
H
H
9
60
41
10
11
12
13
14
15
16
17
18
19
20
21
1j
1k
1l
2j
2k ϵ 2b
72[b]
23[b]
16[b]
55[b]
41
2l
1m
1n
1o
1p
1q
1r
1s
1t
2m
2n
2o
2p
2q
2r
O(CH2)2CϵCH
Ph
CH3
Cl
Cl
H
H
H
31
59
21
95
O
NH
S
SO2
NAc
Cl
Cl
Cl
Cl
H
H
H
H
2s
2t
75
1u
2u
H
H
83[c]
[a] Yield of isolated product after aqueous extraction and purification by MPLC. [b] From two possible isomers, the one following
expulsion of HCN is formed exclusively. [c] Processing temperature 230 °C; at 310 °C decomposition is observed.
Eur. J. Org. Chem. 2012, 47–52
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
49

R. E. Martin, F. Morawitz, C. Kuratli, A. M. Alker, A. I. Alanine
Using this corrected flow rate equivalent to a fixed, effec-
SHORT COMMUNICATION
blocks. We were particularly interested in more challenging
substrates with non-carbon-linked side chains, where rate tive processing time of tR,eff = 30 min, the conversion of
acceleration by superheating under flow conditions would pyrimidine 1b into the corresponding ring-annulated com-
be even more important. Substrates with A = O or NH are pound 2b was assessed between a temperature range of 250
known to be much less reactive in the HDA reaction than and 320 °C by using 1H NMR spectroscopy (Figure 3).
their carbon counterparts and often require significantly ex- Whereas at 250 °C only a modest conversion of 15% to 2b
tended periods of heating to obtain acceptable yields.[12] was observed, the rate of conversion increased considerably,
Again, toluene was considered to be an ideal solvent for reaching a maximum at 310 °C of 95%.
this transformation, as it allows facile superheating, easy
removal during workup, and exhibits reduced toxicity. To
heat toluene significantly beyond 300 °C pressures of up to
70 bar are required. Therefore, we switched from the
Vapourtec R2+/R4 flow system to a self-made system con-
sisting of a Knauer K-501 pump with internal pressure sen-
sor (upper pressure limit of 400 bar) connected to a gas
chromatography oven from a HP 6890 Series system, which
allowed us to reach temperatures of up to 400 °C quickly
and safely (see the Supporting Information). As a reactor
coil, stainless steel tubing from Supelco with an ID of
2.1 mm and 15.2 m length was used, providing a reactor
volume of 53 mL, which allows for facile upscaling.[13] The
Figure 3. Temperature dependency of the HDA reaction of O-
linked dichloropyrimidine 1b to the corresponding annulated pyr-
idine 2b. Conversions of the crude reaction mixture were deter-
mined by 1H NMR spectroscopic analysis after a fixed effective
residence time of tR,eff = 30 min.
reactor outlet was equipped with a small Omnifit glass
chromatography column[14] filled with a short silica plug
and glass wool to protect the 750 psi back-pressure regula-
tor, which is required to keep toluene in the liquid state
under these superheated conditions.
As an ideal model compound to examine the tempera-
ture dependency of the HDA reaction at a fixed reaction
From these optimization experiments, we concluded that
time of 30 min we selected symmetrical dichloropyrimidine an effective residence time of tR,eff = 30 min and a tempera-
compound 1b (Table 1). Due to the presence of two chlorine ture of 310 °C are ideal processing parameters to compare
substituents the relative reactivity compared to the unsub- the relative rate of reactivity in a series of structurally close
stituted counterpart should be enhanced and thus compen- starting materials and to demonstrate the versatility of this
sate for the reduced activity due to the alkoxy side chain. procedure.
The starting material was introduced as a 0.1 m solution
Table 1 summarizes the isolated yields in a systematic
[containing 1% (v/v) pentan-3-one] by passing directly series of annulated pyridine products that were obtained by
through the HPLC solvent pump. To obtain a residence using these optimized reaction conditions. As expected,
time of tR = 30 min with a reactor volume of 53 mL, a substrate 1a with an all-carbon side chain and a chloride
theoretical, nominal flow rate of 1766 μLmin–1 would be atom para to the linker atom A afforded, after MPLC puri-
required. However, this simple calculation does not take fication, product 2a in 95% yield, which is significantly bet-
into account the thermal volume expansion of toluene, ter than that obtained under batch conditions.[4] Dichlo-
which at 310 °C is significant, an estimated 31% according ropyrimidine 1b provided 2b in 70% isolated yield, which is
to literature data.[15] Our own in-house measurements of somewhat lower, but given the O-linkage and the limited
the thermal volume expansion of toluene at 310 °C by using reaction time of 30 min it is an acceptable yield. Interest-
two independent methods gave a slightly higher value of ingly, trimethylsilyl-protected alkyne 1c provided even
37%.[16] In fact, expansion of the volume has a direct influ- slightly better results for 2c (75%) than the corresponding
ence on processing time of the reactants in the heated reac- free acetylene 1b. The introduction of an alkylsilyl group at
tor zone, which results in an effective residence time of only position R6 in final product 2c (ortho to the annulated ring)
tR,eff = 22 min. Therefore, taking the thermal volume ex- is of high synthetic value, as it allows further manipulation
pansion effect in this case into account a corrected flow rate of a position that is traditionally difficult to modify with
of 1290 μLmin–1 (= 1766 μLmin–1/1.37) is obtained, which the HDA reaction. In comparison, methyl-terminated
consequently was used for the following optimization study. acetylene 1d afforded rather moderate yields of 2d (31%).
Interestingly, the effect of volume expansion and the conse- Compared to unsubstituted compound 1e, which afforded
quences associated with it when running high-temperature the corresponding reaction product 2e in 52% yield, elec-
reactions seems to have been largely ignored by the flow tron-withdrawing substituents para to the linker atom A
chemistry community apart from very few exceptions.[17] such as fluorine, chlorine, bromine, or nitro groups are un-
Therefore, published residence times with respect to high- doubtedly beneficial in boosting reactivity. This has been
temperature flow chemistry should be interpreted with cau- exemplified by 1f–i, which gave product yields for the corre-
tion.
sponding annulated products 2f–i in the range of 60–65%.
50
www.eurjoc.org
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2012, 47–52

Synthesis of Annulated Pyridines
Interestingly, para-cyano compound 1j formed 2j in only reduced, unoptimized yield of 46% in comparison with its
moderate yields of 41%, probably due to subsequent side five-membered analogue 2a (95%). As expected, switching
reactions of either starting material or reaction product un- from carbon compound 1v to oxygen analogue 1w resulted
der these superheated conditions.
in a reduction in the yield for 2w (24%) despite the presence
The influence of the substitution pattern on the reactivity of an additional chlorine atom. However, an increase in re-
can be evaluated by direct comparison of para-substituted sidence time to 1 or even 2 h should provide a significant
compound 1g with its meta-substituted congener 1k, which improvement in isolable yields.
provided the corresponding products in 60 and 72% iso-
lated yield, respectively. Surprisingly, the presence of one or
two meta-chlorine substituents afforded similar results
when comparing 1k (72%) with 1b (70%). In contrast, elec-
tron-donating side chains such as alkoxy-substituted deriva-
tives 1l and 1m showed significantly reduced reactivity, as
evidenced by the products that were isolated in yields of
23 and 16%, respectively. Phenyl-substituted compound 1n
afforded the corresponding product in 55% yield. It should
be noted that for all substrates carrying a single meta-sub-
stituent (R1 ϶ H) like 1k–n only a single product was
formed with retention of the substituent in the reaction
products 2k–n. This demonstrates that in cases of ambigu-
ity, the expulsion of HCN from the tricyclic transition state
Scheme 2. Flow synthesis of dihydro-5H-[1]pyridines 2v and 2w.
Conclusions
In summary we have demonstrated that pyrimidine al-
is preferred over the loss of R1CN (R1 ϶ H). However, kynes can be transformed in flow into the corresponding
examples bearing two symmetrical meta-methyl groups annulated pyridines in good to excellent yields by using
such as 1o are also useful substrates, albeit providing the short reaction periods through a cascade involving an in-
corresponding product 2o in a moderate yield of 41%. The verse-electron-demand HDA reaction and subsequent cy-
reduced yield can be rationalized with the combined unfa- cloreversion. The use of flow technology enables organic
vorable effects of the electron-donating properties of the solvents to be superheated far beyond their boiling points,
methyl group as well as the high-energy barrier for the ex- and thus toxic and difficult to workup solvents such as
pulsion of acetonitrile. In the case of dichloropyrimidine 1b nitrobenzene or chlorobenzene, which are typically em-
we explored the influence of methyl substitution in the alkyl ployed for these reactions, can be replaced by less harmful
side chain. Whereas one methyl group at the R4 position in ones. In addition, cyano-containing side products (gener-
1p leads to a rather significant reduction in yield (2p: 31%) ated during this transformation in stoichiometric amounts)
compared with 1b (2b: 70%), dimethylation of the R5 posi- can be trapped continuously in a controlled and safe man-
tion as in 1q provided a considerable improvement in yield ner, avoiding accumulation. The addition of 1% (v/v) of
to 59% for 2q, probably due to the gem-dialkyl (Thorpe– pentan-3-one to the toluene solvent stream was critical to
Ingold) effect.[7] Whereas switching of the linker atom A prevent gradual blockage of the flow reactor coil and a key
from C (2a: 95%) to O and NH resulted in a clear down- success factor in making this process scalable. We envisage
ward trend with respect to yields as demonstrated by 2,3- that this method can be readily extended to other syntheti-
dihydrofuro[2,3-b]pyridine 2g (60%) and 2,3-dihydro-1H- cally important building blocks requiring harsh reaction
pyrrolo[2,3-b]pyridine 2r (21%), the corresponding sulfur conditions, as it enables the preparation of multigram quan-
compound 2,3-dihydro-thieno[2,3-b]pyridine 2s provided tities of product, overcoming limitations associated with
with 95% amazingly good results. To our surprise, sulfone- classical microwave processing. When conducting flow
linked derivative 2t was formed in slightly lower yields of chemistry at such elevated temperatures, thermal volume
75% than its sulfur counterpart 2s, despite the greater elec- expansion of solvents becomes significant and needs to be
tron-withdrawing nature of the former. However, this might taken into account. The calculation of unadjusted, nominal
be attributed to some decomposition of reaction product 2t flow rates leads to considerable errors. To obtain meaning-
due to retro-1,3-chelatotropic side reactions, resulting in the ful and accurate residence times, flow rates need to be cor-
expulsion of SO2. Compared with free aniline 1r, acetylated rected for volume expansion. To avoid confusion with resi-
derivative 1u provided significantly enhanced yields (83%). dence times, tR, calculated from nominal flow rates, we
Owing to predominant decomposition at 310 °C in this spe- would propose to use for such corrected residence times the
cific case, the processing temperature was reduced to term effective residence time, tR,eff. The prototype flow reac-
230 °C.
tor system reported herein by using commercially available
The HDA reaction with an unsubstituted alkyl chain was standard HPLC pumps, stainless steel tubing, and a cost-
also extended to the formation of six-membered ring sys- effective gas chromatography oven expands the ability of
tems, as shown in Scheme 2. A direct comparison between the synthetic chemist to perform high-temperature flow
1a and homologated analogue 1v shows that the corre- chemistry in a safe, scalable, and controlled fashion. It
sponding six-membered ring 2v can be synthesized success- should allow many synthetic organic laboratories to adopt
fully by using the same reaction conditions, although in a this technology and help to significantly expand their cur-
Eur. J. Org. Chem. 2012, 47–52
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
51

R. E. Martin, F. Morawitz, C. Kuratli, A. M. Alker, A. I. Alanine
SHORT COMMUNICATION
[3]
a) A. Polyzos, M. O’Brien, T. P. Petersen, I. R. Baxendale, S. V.
Ley, Angew. Chem. Int. Ed. 2011, 50, 1190–1193; b) M.
O’Brien, I. R. Baxendale, S. V. Ley, Org. Lett. 2010, 12, 1596–
1598; c) B. Gutmann, J.-P. Roduit, D. Roberge, C. O. Kappe,
Angew. Chem. Int. Ed. 2010, 49, 7101–7105; d) J. C. Brandt,
S. C. Elmore, R. I. Robinson, T. Wirth, Synlett 2010, 20, 3099–
3103; e) A. A. Kulkami, V. S. Kalyani, R. A. Joshi, R. R. Joshi,
Org. Process Res. Dev. 2009, 13, 999–1002.
rent process window and to conduct underrepresented reac-
tions more efficiently. It limits the risks and hazards associ-
ated with the use of traditional high-temperature and high-
pressure batch equipment and enables the re-exploration of
“neglected” reactions usually avoided due to lack of suit-
able processing tools.
[4]
[5]
A. E. Frissen, A. T. M. Marcelis, G. Geurtsen, D. A. de Brie,
H. C. van der Plas, Tetrahedron 1989, 45, 5151–5162.
a) J. A. Joule, K. Mills, Heterocyclic Chemistry, 4th ed.,
Blackwell Science, Oxford, 2000, pp. 63–120; b) K. Kral, M.
Hapke, Angew. Chem. Int. Ed. 2011, 50, 2434–2435; c) M. D.
Hill, Chem. Eur. J. 2010, 16, 12052–12062.
a) A. E. Frissen, A. T. M. Marcelis, H. C. van der Plas, Tetrahe-
dron 1989, 45, 803–812; b) W. A. W. Stolle, A. T. M. Marcelis,
A. Koetsier, H. C. van der Plas, Tetrahedron 1989, 45, 6511–
6520; c) D. M. Vandenberghe, G. J. Hoornaert, Bull. Soc. Chim.
Belg. 1994, 103, 185–186; d) T. Koike, Y. Hoashi, T. Takai, O.
Uchikawa, Tetrahedron Lett. 2011, 52, 3009–3011; e) D. L.
Boger, S. M. Weinreb, “Hetero Diels–Alder Methodology in
Organic Synthesis” in Organic Chemistry (Ed.: H. H. Wasser-
man), Academic Press, New York, 1987.
W. A. W. Stolle, A. E. Frissen, A. T. M. Marcelis, H. C.
van der Plas, Y. Wang, L. Häming, C. H. Stam, J. Org. Chem.
1991, 56, 2411–2416.
K. A. Brameld, B. Kuhn, D. C. Reuter, M. Stahl, J. Chem. Inf.
Model. 2008, 48, 1–24.
The Vapourtec R2+/R4 flow system is commercially available
from Vapourtec. Website: http://www.vapourtec.co.uk.
a) I. Mamajanov, J. Herzfeld, J. Chem. Phys. 2009, 130, 134503/
1–134503/6; b) I. Mamajanov, J. Herzfeld, J. Chem. Phys. 2009,
130, 134504/1–134504/5; c) C. N. Matthews, R. D. Minard,
Faraday Discuss. 2006, 133, 393–401; d) C. N. Matthews, Cell.
Origin Life Extreme Habitats Astrobiol. 2004, 6, 121–135.
CCDC-826667 contains the supplementary crystallographic
data for this paper. These data can be obtained free of charge
from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif.
Experimental Section
Experimental Setup: Knauer K-501 HPLC solvent pump, HP 6890
Series Gas Chromatography (GC) oven equipped with a 53-mL
homemade reactor coil by using Supelco 304 stainless steel tubing
(length ϫ ID = 15.2 m ϫ2.1 mm), an Omnifit glass chromatog-
raphy column [filled with a short silica plug (ca. 1 cm, silica 60 Å)
and glass wool to provide a large surface for absorption of oligo-
meric and polymeric material] and a 750 psi back-pressure regula-
tor.
[6]
General Flow Procedure for the Synthesis of Annulated Pyridines
2a–w: A 0.1 m solution (25 mL) of pyrimidines 1a–w in toluene
containing 1% (v/v) of pentan-3-one was pumped at a flow rate of
1290 μLmin–1 (effective residence time tR,eff = 30 min) through the
reactor coil (volume = 53 mL) heated to 310 °C. The reactant
stream was collected in water containing 1 m NaOH (except for 2c
and 2u) and gaseous cyanide side products were pushed by a gentile
stream of nitrogen into two traps filled with a 1:1 mixture of 10%
NaOH and sodium hypochlorite (13% active chlorine). The solu-
tion was extracted with ethyl acetate (3ϫ), and the organic solvent
was dried with anhydrous MgSO4 and evaporated under reduced
pressure. The crude material was purified by medium-pressure li-
quid chromatography (MPLC) over silica. The fume hood air space
was monitored during the entire experiment by a HCN detector.
[7]
[8]
[9]
[10]
[11]
Supporting Information (see footnote on the first page of this arti-
cle): Determination of the thermal volume expansion of toluene,
pictures of the cross- and longitudinal section cuts of the reactor
coil and the experimental setup used for meso-scale production,
detailed experimental procedures, and compound characterization
data.
[12]
[13]
A. E. Frissen, A. T. M. Marcelis, H. C. van der Plas, Tetrahe-
dron Lett. 1987, 28, 1589–1592.
Supelco premium grade 304 stainless steel tubing; dimensions:
length ϫ OD (ID) = 15.2 m ϫ3.2 mm (2.1 mm); product refer-
ence number: 2–0526-U.
[14]
[15]
Commercially available Omnifit glass chromatography column
with adjustable height-end piece (plunger) of 6.6 mm
bore and 100 mm length (max 900 psi). Website: http://
www.omnifit.com.
J. S. Yadav, D. V. K. Sharma, Thermochim. Acta 2009, 496,
166–172. The thermal expansion coefficient α for toluene is
given as 1.085ϫ103 (K–1). Based on a temperature difference
between room temperature (23 °C) and reaction temperature
(310 °C) of 289 K a volume expansion of 31% is calculated.
However, it should be noted that the thermal expansion coeffi-
cient α is only valid within a certain temperature range, as the
volume expansion as a function of temperature is not strictly
linear.
The thermal volume expansion of toluene was determined by
measuring the excess amount of solvent that leaves the reactor
during heatup between room temperature and the reaction
temperature as well as by an independent flow marker method.
Both methods revealed with 37 and 38% volume expansion,
respectively, similar values. For more detailed information, see
the Supporting Information.
Acknowledgments
We are grateful to Mrs. Diana Monteiro for the preparation of the
starting materials, Mr. Daniel Zimmerli for technical assistance, Dr.
Thomas Wellauer (DSM Nutritional Products AG, 4334 Sisseln,
Switzerland) for corrosion analysis, and Dr. Thomas Woltering and
Dr. Luke Green for helpful discussions.
[1] a) T. Illg, P. Löb, V. Hessel, Bioorg. Med. Chem. 2010, 18, 3707–
3719; b) V. Hessel, Chem. Eng. Technol. 2009, 32, 1655–1681;
c) C. Wiles, P. Watts, Future Med. Chem. 2009, 1, 1593–1612;
d) F. Paviou, Pharm. Technol. Eur. 2009, 21, 22–32; e) B. P.
Mason, K. E. Price, J. L. Steinbacher, A. R. Bogdan, D. T.
McQuade, Chem. Rev. 2007, 107, 2300–2318.
[2] a) C. Wiles, P. Watts, Chem. Commun. 2011, 47, 6512–6535; b)
P. Watts, C. Wiles (Eds.), Micro Reaction Technology in Organic
Synthesis, CRC Press, Boca Raton, 2011; c) T. Wirth (Ed.),
Microreactors in Organic Synthesis and Catalysis, Wiley-VCH,
Weinheim, 2008; d) S. Ceylan, A. Kirschning, “Organic Syn-
thesis in Mini Flow Reactors Using Immobilized Catalysts” in
Recoverable and Recyclable Catalysts (Ed.: M. Benaglia), John
Wiley & Sons, New York, 2009, pp. 379–410; e) V. Hessel, C.
Knobloch, H. Löwe, Recent Pat. Chem. Eng. 2008, 1, 1–16.
[16]
[17]
U. Tilstam, T. Defrance, T. Girard, Org. Process Res. Dev. 2009,
13, 321–323.
Received: October 20, 2011
Published Online: November 24, 2011
52
www.eurjoc.org
© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2012, 47–52