The use of immobilized crown ethers as in-situ N-protecting groups in organic synthesis and their application under continuous flow

Article abstract of DOI:10.1016/j.tet.2008.12.052

In addition to their high affinity for inorganic cations, crown ethers have been shown to efficiently sequester ammonium ions, forming a stable adduct via hydrogen bonding. Using this principle, several authors have reported the use of crown ethers as protecting groups for amines however to date, their widespread use has been somewhat precluded by the difficulties associated with removal of the crown ether from the resulting reaction mixture. In order to address this problem, we report the preparation of an immobilized 18-crown-6 ether derivative and its incorporation into a flow reactor, demonstrating the ability to use and recycle the reagent for the chemoselective O-acylation and alkylation of bifunctional compounds such as 4-(2-aminoethyl)phenol and 4-nitrophenol.

Full text of DOI:10.1016/j.tet.2008.12.052

Tetrahedron 65 (2009) 1618–1629
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
The use of immobilized crown ethers as in-situ N-protecting groups in organic
synthesis and their application under continuous flow
,
,
a
Gareth. P. Wild a, Charlotte Wiles a b, Paul Watts a , Stephen J. Haswell
*
a Department of Chemistry, University of Hull, Cottingham Road, Hull, HU6 7RX, UK
b Chemtrix BV, Burgemeester Lemmensstraat 358, 6163JT Geleen, The Netherlands
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 10 October 2008
Received in revised form 26 November 2008
Accepted 11 December 2008
Available online 24 December 2008
In addition to their high affinity for inorganic cations, crown ethers have been shown to efficiently se-
quester ammonium ions, forming a stable adduct via hydrogen bonding. Using this principle, several
authors have reported the use of crown ethers as protecting groups for amines however to date, their
widespread use has been somewhat precluded by the difficulties associated with removal of the crown
ether from the resulting reaction mixture. In order to address this problem, we report the preparation of
an immobilized 18-crown-6 ether derivative and its incorporation into a flow reactor, demonstrating the
ability to use and recycle the reagent for the chemoselective O-acylation and alkylation of bifunctional
compounds such as 4-(2-aminoethyl)phenol and 4-nitrophenol.
Ó 2009 Elsevier Ltd. All rights reserved.
1. Introduction
Bushmann and Mutihac3 compared the degree of ammonium ion
complexation with various functionalized and un-functionalized
When conducting the synthesis of polyfunctionalized molecules
it is often necessary to render one or more of the functionalities
within the molecule temporarily inert in order to allow the selec-
tive reaction of another group. This process is achieved by
appending a blocking group, which is stable to the reaction con-
ditions under investigation whilst being readily removed; such
a moiety is termed as a protecting or protective group.1 While the
use of protecting groups has enabled access to a vast library of
complex molecules over the years, by removing functional group
incompatibilities, their use can be disadvantageous as the in-
troduction and removal of such groups generates additional syn-
thetic steps. This can lead to increased costs and contribute to
reductions in overall yield, along with the need to perform complex
purifications in order to remove any protecting group residues from
the final material. Furthermore, the deprotection strategy must be
carefully selected in order to ensure that the product is obtained in
the desired form i.e. as the free functionality or as a salt.
crown ethers, concluding that 18-crown-6 ethers afforded superior
complexation cf. to smaller crown ethers. As depicted in Figure 1,
unlike metal ions, the ammonium ion is held above the cavity of the
crown ether in a tetrahedral configuration via hydrogen bonding4
with stabilization achieved due to distribution of the positive
charge on the ammonium ion over the hydrogen atoms. The geo-
metry of the complexed ammonium ion also leaves the remainder
of the molecule unhindered and available to take part in sub-
sequent reactions as demonstrated by Kunishima et al.4b
In the early 1990s, Mascagni and Hyde5 exploited this phe-
nomena for the synthesis of peptides and oligomers, proposing that
the non-covalent nature of the interaction between the crown
ether and the ammonium ion would provide a mild protection
strategy for amines. Employing dibenzo-18-crown-6 ether (DB-18-
c-6) and an array of alanine salts (HCl, TFA and p-TSA), the effect of
solvent polarity on a coupling reaction was evaluated. During this
investigation it was observed that tosylate salts formed the most
stable complexes (p-TSA>TFA>HCl) and that complex stability in-
creased with decreasing solvent polarity; an observation that was
1.1. The use of crown ethers in organic synthesis
Cl
Cl
O
R
+
In addition to the propensity of crown ethers to form complexes
with metal ions (Liþ, Naþ, Kþ, Rbþ, Csþ) and promote solubility in
non-polar media, Pederson2 also reported their ability to form
complexes with ammonium ions (R–NHþ3 ). Further work by
H
N
O
O
O
O
H
K+
O
H
O
O
O
O
O
O
* Corresponding author. Tel.: þ44 1482 465471.
E-mail address: p.watts@hull.ac.uk (P. Watts).
Figure 1. Schematic illustrating the modes of complexation observed for ammonium
ions versus metal ions with 18-c-6 ether.
0040-4020/$ – see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2008.12.052

G.P. Wild et al. / Tetrahedron 65 (2009) 1618–1629
1619
attributed to a reduction in competition between solvation of the
cation and coordination by the crown ether.5 Consequently, sub-
sequent coupling reactions were performed in DCM and found to
afford the desired complexed tripeptide in 80% yield, with the re-
mainder comprising of mixed oligomers; the formation of which
was attributed to partial decomplexation of the peptide during the
reaction.6 The authors subsequently reported the decomplexation
of another tripeptide using aq KCl, exploiting the crown ethers af-
finity for metal ions to force out the ammonium ion and leaving the
crown ether complexed with potassium, as depicted in Figure 1.
Although this approach demonstrated the key components of
a protecting group strategy, namely the ability to protect an amino
acid, perform a coupling reaction and deprotect the resulting
peptide, the use of stoichiometric quantities of crown ether proved
disadvantageous due to difficulties associated with its removal
from the reaction product. As such, isolation of the peptide, re-
generation of the crown ether cavity and subsequent re-use of the
crown ether were not demonstrated by the authors.
Based on the preliminary observations made by Mascagni
et al.5,6 it was proposed that the preparation of an immobilized 18-
crown-6 ether derivative would facilitate the process of amine
deprotection, along with product isolation and subsequent crown
ether regeneration. With this in mind, our aim was to develop
a non-covalent protecting group strategy using an immobilized 18-
crown-6 ether derivative and combine it with continuous flow
methodology, developed within our group, to demonstrate efficient
O-acetylation/alkylation of substituted 1ꢀ amines.
labile tert-butoxycarbonyl (Boc) protecting group. As depicted in
Scheme 2, the reaction sequence firstly involved the protection of
tyramine 1, using di-tert-butyl dicarbonate 7 in the presence of
NaHCO3 8, which afforded the intermediate Boc-tyramine 9 as
a white solid (93% yield). Subsequent acylation, with acetyl
chloride 3 (1.2 equiv) in the presence of NaH 2 (1.0 equiv) affor-
ded the desired Boc-tyramine acetate 10 as a pale yellow gum
(69% yield).
OH
O
H2N
CF3CO2H. H2N
O
2
2
1
11
Boc O , NaHCO
7
8
2
3
TFA 38, DCM
H2O, t-BuOH
O
3
O
O
OH
NaH 2,
Cl
BocHN
BocHN
DCM
2
2
10
9
Scheme 2. Schematic of the covalent protecting group strategy employed to syn-
thezise tyramine acetate TFA salt 11.
Although the use of a protecting group afforded a chemo-
selective route to the O-acetylation of tyramine 4, in addition to
increasing the number of reaction steps employed in the synthetic
pathway, removal of the protecting group proved problematic,
affording a mixture of tyramine acetate TFA salt 11 (69%), tyramine
TFA salt 12 (16%) and Boc-tyramine acetate 10 (15%) as determined
by HPLC. In an analogous manner to the direct acetylation of ty-
ramine 1 (Scheme 1), the use of a protecting group afforded
a complex reaction mixture, largely due to problems associated
with the efficient removal of the Boc-protecting group. Further-
more, in order to isolate the desired tyramine acetate 4 it was
necessary to free-base the TFA salt 11 thus incurring additional
reaction steps.
2. Results and discussion
Owing to the highly basic and nucleophilic nature of an amine
group, protection is essential in reactions such as acylations and
alkylations. This is illustrated for the bifunctional compound tyra-
mine 1, which contains both aromatic alcohol and aliphatic amine
functionalities. As depicted in Scheme 1, failure to protect the
amine group and direct the reaction to the phenolic group upon
treatment of tyramine 1 with NaH 2 (1.0 equiv) and acetyl chloride
3 (1.2 equiv) results in a complex mixture containing the desired
tyramine acetate (4-(2-aminoethyl)-phenyl acetate) 4 (23%), along
with tyramine N-acetate (N-[2-(4-hydroxyphenyl)ethyl]-acet-
amide) (12%) 5, tyramine diacetate (acetic acid 4-(2-acetylami-
noethyl)-phenyl ester) (20%) 6 and residual starting material 1
(45%).
2.1. Evaluation of immobilized crown ethers as non-covalent
protecting groups
In order to evaluate the use of an immobilized crown ether as
a protecting group, preliminary investigations were conducted
using the commercially available di-tert-butylcyclohexano-18-
crown-6 ether on an inert chromatographic support
(0.37 mmol gꢁ1) (Eichrom Technologies, France). To assess the po-
tential of di-tert-butylcyclohexano 18-c-6 as a non-covalent pro-
tecting group, the ability of the material to complex tyramine HCl
OH
H2N
1
2
O
NaH 2
13 and subsequently decomplex tyramine
1 was firstly in-
3
Cl
vestigated. This was achieved by stirring the crown ether with ty-
ramine HCl 13 in MeOH, prior to filtration, under suction, to remove
any uncomplexed tyramine HCl 13. To confirm sequestration had
occurred, the immobilized complex was treated with methanolic
KCl 14, to induce decomplexation, and the resulting filtrate ana-
lyzed by HPLC. Initial results were pleasing and detection of tyr-
amine 1 confirmed the ability of the material to complex tyramine
HCl 13 and release tyramine 1 from the crown ether cavity. To
demonstrate re-use of the supported crown ether, the cavity was
regenerated and the above procedure repeated; unfortunately,
upon analysis of the filtrate no tyramine 1 was detected. Employing
a fresh portion of the solid-supported crown ether again confirmed
sequestration of tyramine HCl 13 and release of the amine 1. It was
therefore postulated that during the regeneration step the crown
ether, which was only adsorbed onto the solid support, leached
from the support and therefore could not be efficiently recycled or
employed in a continuous system.
OH
O
H
N
H2N
O
2
2
O
O
O-Acylation
N-Acylation
4
5
H
N
O
2Diacylation
6
O
Scheme 1. Schematic illustrating the potential reaction products obtained when
acetylating tyramine 1.
In order to benchmark the developed non-covalent protecting
group strategy against existing covalent approaches, the acety-
lation of tyramine 1 was subsequently performed using the acid

1620
G.P. Wild et al. / Tetrahedron 65 (2009) 1618–1629
Having identified the presence of active crown ether cavities on
O
O
the solid-support 22, the ability to recycle the material was sub-
sequently investigated, with removal of the potassium cation ach-
ieved using methanolic acetic acid 23. Once successfully
regenerated, the cycle was repeated a further four times and unlike
the commercially available di-tert-butylcyclohexano-18-c-6, all
subsequent cycles yielded tyramine 1, with an 11.4% RSD (n¼5)
(Scheme 4).
O
O
O
O
NH2
N
H
15
O
O
O
O
O
N
H
OH
Cl
O
16
O
O
O
O
R-NH3+ Cl-
13
O
O
O
O
Figure 2. Illustration of two immobilized crown ethers that was found to complex
O
O
O
O
+
metal ions but not ammonium ions.
H
N
H
N
H3N
O
O
MeOH
2.1.1. Preparation of an immobilized crown ether
22
O
5
5
To address if desorption of the crown ether was the problem, we
evaluated the covalent immobilization of three 18-crown-6 ether
derivatives and subsequently evaluated the materials towards the
sequestration of ammonium salts under continuous flow. With this
in mind, several immobilization techniques were employed, with
starting materials including diamino-dibenzo-18-crown-6 ether
and carboxybenzo-18-crown-6 ether to afford crown ethers 15 and
16, respectively, as depicted in Figure 2.
14
KCl
MeOH
5% Acetic acid
23
Cl
O
O
O
O
K+
O
H
N
OH
O
O
Unfortunately, due to the increased ring strain observed as
a result of immobilization, neither material was able to complex
ammonium ions, a process, which is reliant on hydrogen bonding
(Fig. 1), however, both materials successfully sequestered potas-
+
1
5
KCl 14 + H2N
Scheme 4. Reaction scheme illustrating the complexation of tyramine HCl 13, using
immobilized AM-18-c-6 22, and subsequent decomplexation and cavity regeneration.
sium permanganate (0.12 mmol gꢁ1 and 0.98 mmol gꢁ1
, re-
spectively) confirming the presence of active crown ether moieties.
Owing to the reliance of ammonium ion complexation on hy-
drogen bonding (Fig. 1), it is imperative for the immobilized crown
ether to be free of strain in order to form a stable complex. With this
in mind, a further attempt was made to prepare a covalently bound
18-crown-6 ether with no ring strain, this was achieved by
employing aminomethyl-18-crown-6 ether (AM-18-c-6) 17 as
a precursor. As Scheme 3 illustrates, treatment of carbxoypolys-
tyrene 18 with thionyl chloride 19 afforded the immobilized acid
chloride 20, to which was added Et3N 21 followed by AM-18-c-6 17
to afford immobilized AM-18-c-6 22.
Although the material was able to be recycled, illustrating the
robustness required for continuous processing, the irreproducible
quantities of tyramine 1 recovered were still undesirable. In order
to identify the origin of any irreproducibility, an aliquot of immo-
bilized crown ether was taken at each stage of the process and
subjected to elemental analysis. As Table 1 illustrates, the decom-
plexation step was found to be 100% efficient and as such, the
gradual decrease in the proportion of tyramine 1 recovered was
attributed to loss of immobilized crown ether 22 upon filtration
and washing.
O
OH
O
Cl
Table 1
17
O
O
O
O
O
O
Elemental analyses illustrating the efficient complexation and decomplexation
achieved using immobilized crown ether 22
H2N
5
5
SOCl2 19
Toluene
18
20
Polymeric material
N (%)
N (mmol gꢁ1
)
Et3N 21
Toluene
Blank 18
Immobilized AM-18-c-6 22
Complexed
0.00
0.52
0.97
0.52
0.00
0.37
0.69
0.37
O
O
Decomplexed
O
O
O
O
H
N
O
2.1.2. Continuous flow evaluation of crown ether
5
In order to address the irreproducibility associated with
the batch process and develop an efficient, re-useable system for
the non-covalent protection of amines, the incorporation of the
immobilized crown ether into a continuous flow reactor was
22
Scheme 3. Synthetic protocol employed for the covalent immobilization of AM-18-c-6
17 to carboxypolystyrene 18.
To evaluate the materials ability to sequester ammonium
ions, the crown ether 22 was stirred at room temperature in
a methanolic solution of tyramine HCl 13, filtered under suction
and washed with MeOH prior to treatment with methanolic KCl
14 to release any complexed tyramine HCl 13 as the free amine
1. Analysis of the filtrate by HPLC confirmed the presence of
tyramine 1 and elemental analysis of the crown ether 22,
coupled with sequestration of KMnO4 followed by ICP-MS
PTFE Tubing
(800 µm i.d.)
Borosilicate Glass Capillary
(3 mm i.d. x 50 mm length)
Omnifit
Connector
analysis, confirmed the material to have
a loading of
Figure 3. Schematic of the continuous flow reactor used herein for the evaluation of
an immobilized crown ether 22.
0.37 mmol gꢁ1
.

G.P. Wild et al. / Tetrahedron 65 (2009) 1618–1629
1621
0.16 mmol gꢁ1. Prior to performing a reaction, material’s 24 stability
to DMF, THF, DCM and MeCN was evaluated and unlike immobi-
lized crown ether 22, all solvents were able to be pumped through
O
O
O
O
O
O
H
N
O
the reactor at 100 m
L minꢁ1 with no sign of swelling or restricted
24
flow over a period of 8 h. Having demonstrated the materials sta-
bility, its ability to complex tyramine HCl 13 and release tyramine 1
was evaluated. Employing a solvent stream of MeOH, tyramine HCl
13 (0.12 M, 2.4ꢂ10ꢁ2 mmol) was introduced into the flow reactor,
containing SILICA AM-18-c-6 24 (0.15 g, 2.4ꢂ10ꢁ2 mmol) at a flow
3
Figure 4. SILICA AM-18-c-6 24 prepared via the acid chloride.
rate of 100 m
L minꢁ1. Decomplexation was again achieved using KCl
investigated. It was proposed that such an approach would enable
material 22 to be recycled with ease as solutions of the various
reactants would be pumped through crown ether 22 in order to
conduct the desired reaction step; thus removing the main source
of error observed thus far, loss of material upon filtration.
As Figure 3 illustrates, the flow reactor comprised of a borosili-
cate glass capillary (50 mm (length)ꢂ3 mm (i.d.)) packed with
w0.15 g of immobilized crown ether 22. Reactant solutions were
passed through the reactor using a syringe pump and reaction
products collected in a sample vial prior to off-line analysis by
HPLC.7 To increase the reproducibility of the technique further,
reactants were introduced into the system via a Rheodyne valve
14 in MeOH (0.12 M, 2.4ꢂ10ꢁ2 mmol) and the column washed with
5% acetic acid 23 in MeOH to regenerate the crown ether 24. Using
this approach, a loading of 0.16 mmol gꢁ1 was obtained, which was
in agreement with ICP-MS analysis performed on the respective
potassium complex.
2.1.4. Evaluation of complex stability
Once complexed, it was important to determine how stable the
ammonium salt was to various solvents and reactants that may be
employed in the derivatization of the complexed material. To en-
sure the investigation provided general conclusions for this non-
covalent protecting group strategy, three ammonium salts were
investigated (tyramine HCl 13, then tyramine TFA 12 and finally
tyramine p-TSA 25). As before, complexation was achieved by
(200 mL sample loop).
Employing a solvent stream of MeOH, tyramine HCl 13 (0.3 M,
6.0ꢂ10ꢁ2 mmol) was introduced into the flow reactor, containing
immobilized AM-18-c-6 22 (0.15 g, 5.6ꢂ10ꢁ2 mmol) at a flow rate
injecting a 200 mL plug of the tyramine salt under investigation
(0.12 M, 2.4ꢂ10ꢁ2 mmol) into
a continuous MeOH stream
of 100 m
L minꢁ1. Decomplexation was again achieved using KCl 14
(100 m
L minꢁ1) (Scheme 5), which ensured complete washing of the
in MeOH (0.3 M, 6.0ꢂ10ꢁ2 mmol) and the column purged with 5%
acetic acid 23 in MeOH to regenerate the crown ether 22. Using this
approach, 0.35 mmol gꢁ1 of tyramine 1 was released with an RSD of
4.3% (n¼10), demonstrating a dramatic increase in reproducibility
cf. the 11.4% obtained in batch.
resin prior to evaluating the stability of the complex under the
selected reaction condition, followed by decomplexation with KCl
14 in MeOH (0.12 M, 2.4ꢂ10ꢁ2 mmol) and analysis by HPLC (Tables
2–5).
While the use of MeOH was found to be ideal for complex for-
mation, decomplexation and cavity regeneration, the solvent is not
suitable for the reactions under investigation herein. As such, a se-
ries of alternative solvents were investigated, these included DMF,
THF, DCM and MeCN. Unfortunately, when this series of common
organic solvents were employed, the immobilized AM-18-c-6 22
was observed to swell, leading to blockages within the reactor and
inconsistencies in the volume of solution passing through the
packed bed at any one time. This observation was attributed to the
low degree of crosslinking within the carboxypolystyrene 18 (1%
DVB) employed as a solid support and was confirmed by packing
the reactor with polystyrene crosslinked with 2% DVB whereby no
swelling was observed. Regrettably, carboxypolystyrene 18 was not
available with a higher degree of crosslinking and as such, 3-car-
boxypropyl functionalized silica gel was evaluated as an alternative,
non-swelling support material.
2.1.4.1. Solvent stability. To determine the stability of the ammo-
nium salts of tyramines 12, 13 and 25, each solvent illustrated in
Table 2 was pumped through the reactor at 100 m
L minꢁ1 for 5 min
prior to analysis of the reactor effluent by HPLC; all solvents were
evaluated five times and the average resulted is presented. With the
Table 2
Evaluation of complex stability to a series of common organic solvents (n¼5)
Solvent
Stability of complex (%)
Tyramine HCl 13
Tyramine TFA 12
Tyramine p-TSA 25
Acetone
Methanol
Ethanol
Dichloromethane
Diethyl ether
Hexane
Toluene
Water
Acetonitrile
Tetrahydrofuran
DMF
100
100
100
100
100
100
100
100
100
100
18
100
100
100
100
100
100
100
100
100
100
34
100
100
100
100
100
100
100
100
100
100
44
2.1.3. Preparation and evaluation of SILICA AM-18-c-6
Having successfully immobilized AM-18-c-6 17 onto carboxy-
polystyrene 18 via the acid chloride, an analogous approach was
employed for the derivatization of 3-carboxypropyl functionalized
silica gel, as depicted in Figure 4, affording
a loading of
X
O
OH
O
O
O
O
+
H3N
O
NH
O
3
X = Cl, CO2CF3, CH3C6H5SO3
Reactant/Solvent
Tyramine 1
Scheme 5. Set-up used to evaluate the stability of tyramine HCl 13, TFA 12 and p-TSA 25 salts to an array of reactants and solvents.

1622
G.P. Wild et al. / Tetrahedron 65 (2009) 1618–1629
Table 3
exception of DMF, the complexes were found to be stable to all
solvents investigated and interestingly, the proportion of tyramine
1 displaced by DMF ranged from 66 to 82% depending on the am-
monium salt under investigation. The results from this study herein
suggest therefore that the basic nature of DMF induces decom-
plexation, thus removing the protecting capacity of the crown ether
and potentially resulting in undesirable reaction of the free amine
functionality. Mascagni and Hyde5a also observed this trend, p-
TSA>TFA>HCl, suggesting that stability increased as a function of
cation conjugation.
Summary of the stability of tyramine HCl 13, TFA 12 and p-TSA 25 to a variety of
common amines in MeOH (n¼5)
Amine
pKb
Stability of complex (%)
13
12
25
3ꢀ
3ꢀ
3ꢀ
3ꢀ
3ꢀ
Et3N 21
DIEA
N-Methylpiperidine
TMEDA 26
Lutidine
11.06
10.50
10.08
6.10
100
48
51
100
54
53
0
100
69
68
0
0
6.75
100
100
100
2ꢀ
2ꢀ
2ꢀ
2ꢀ
2ꢀ
Diethylamine
Piperazine
Piperidine
Diisopropylamine
Dimethylamine
11.09
9.82
11.22
11.05
10.73
27
37
40
57
50
48
47
54
61
58
89
57
81
73
69
2.1.4.2. Complex stability in the presence of amines. In addition to
the use of organic solvents, many reactions that require the pro-
tection of amine functionalities involve the use of compounds
containing other amine moieties; as such it was important to
identify, which of those reagents were compatible with the pro-
tecting group strategy under investigation. Using the aforemen-
tioned complexation strategy, the effect of an array of amines
(0.12 M, 2.4ꢂ10ꢁ2 mmol) on the complex stability was in-
vestigated, with the data presented in Table 3 highlighting an
obvious trend of increased complex stability towards 3ꢀ amines;
with 1ꢀ amines causing the greater degree of destabilization. This
general trend of complex stability, 1ꢀ<2ꢀ<3ꢀ, has some obvious
exceptions such as lutidine and Et3N 21, which have no destabi-
lizing effect, an observation that is attributed to their steric bulk.
As a result, however, these bases may be useful as reagents in
future reactions, such as in the deprotonation of the phenolic
moiety in the model reaction.
1ꢀ
1ꢀ
1ꢀ
1ꢀ
Aniline
Benzylamine
2-Phenylethylamine
3-Phenylpropylamine
4.63
9.33
9.58
9.68
27
50
39
49
58
64
61
57
78
73
53
63
Table 4
Stability of tyramine HCl 13 complexed SILICA AM-18-c-6 24 when exposed to
a plethora of common reactants
Substrate type
Compound
Stability of complex (%)
Cation/anion
Cation/anion
Cation/anion
KOH
NaOH
LiOH
0
0
0
Interestingly, N,N,N0,N0-tetramethylethylenediamine 26 (TMEDA)
was found to afford quantitative decomplexation for all three salts
12, 13 and 25; an observation that compares favourably with re-
ports by Hyde et al.5b who found diisopropylethylamine (DIEA)
induced decomplexation of an ammonium salt. In comparison to
TMEDA 26, however, DIEA afforded only 31–52% decomplexation,
depending upon the salt employed. As such, the advantages of
TMEDA 26 as a decomplexation agent are discussed in Section
2.1.5.
Cation/anion
Cation/anion
Cation/anion
KCl
NaCl
LiCl
0
0
0
Cation/anion
Cation/anion
Cation/anion
K2CO3
Na2CO3
Li2CO3
46
0
50
Reactant
Reactant
Reactant
Reactant
Reactant
Reactant
Acetyl chloride 3
Acetic anhydride 26
DMAP
DCC
EDCI
100
100
47
45
50
0
Methyl iodide 27
2.1.4.3. Stability to common reactants and by-products. Further to
investigating the complex stability to possible bases and solvents, it
was also important to consider other reagents common to synthetic
reactions that require the protection of amines. Table 4 illustrates
the stability of the tyramine HCl 13 complexed with SILICA AM-18-
c-6 24 when exposed to a plethora of possible reactants (0.12 M,
2.4ꢂ10ꢁ2 mmol) such as those used for peptide couplings, acety-
lations and methylations.
Acid
Acid
Acid
Acid
TFA 38
HCl
Acetic acid 23
Sulfuric acid
100
100
100
100
Table 5
Common reagents for acylation (acetyl chloride 3, acetic anhy-
dride27)andalkylation(methyliodide 28)wereobserved to have no
destabilizing effect, however, reagents employed in coupling re-
actions such as DMAP, DCC and EDCI were all found to cause varying
degrees of decomplexation. This is an observation that would go
some way towards explaining the poor reaction control reported by
Mascagni et al.,5 as upon deprotection, the amino acid could take
part in random solution phase couplings to afford oligomers.
Entries 1–9 (Table 4) were investigated for the purpose of
gaining further understanding into the decomplexation process as
well as identifying possible reagents for use in future reactions; all
solutions evaluated were saturated in MeOH. Of the three metal
salts investigated, lithium was found to have the least affinity for
the 18-c-6 ether due to its small cation size, followed by sodium. In
addition to decomplexation, potassium hydroxide was found to be
unsuitable for use with the solid-supported crown ether 24 as it
was found to cleave AM-18-c-6 17 from the support (confirmed by
ICP-MS analysis). In comparison to the metal alkoxides, carbonates
are weaker inorganic bases, which exhibit relatively poor dissoci-
ation in solution, resulting in a lower concentration of available
Summary of the ammonium salts evaluated under continuous flow and their ability
to form complexes with SILICA AM-18-c-6 24
Effect
Amine
Complexation (%)
Cation
Cation
Cation
No cation
Tyramine p-TSA 25
Tyramine TFA 12
Tyramine HCl 13
Tyramine 1
100
100
100
0
Free amine
Free amine
Free amine
Free amine
Aniline
Benzylamine
2-Phenylethylamine
3-Phenylpropylamine
0
0
0
0
Chain length
Chain length
Chain length
Chain length
Chain length
Aniline HCl
Benzylamine HCl
Phenylethylamine HCl
Phenylpropylamine HCl
4-Aminophenol HCl 35
100
100
100
100
100
Steric
L-
b
-Alanine benzyl
0
ester HCl 29
Steric
(S)-(ꢁ)-2-Amino-3-phenyl-1-propanol HCl 30
0
0
Functionality
Benzamide HCl 31

G.P. Wild et al. / Tetrahedron 65 (2009) 1618–1629
1623
(a)
Cl
O
O
H3N
O
OH
O
O
O
NH
O
3
OH
Excess Reactant
HCl . H2N
13
in MeOH
@ 100 µl min-1
Solvent Wash (THF @ 100 µl min-1)
(b)
(c)
O
O
Cl
O
O
O
O
O
H3N
NH
O
3
O
O
21
Excess Reactants
N
O
27
in THF
@ 100 µl min-1
Solvent Wash (MeOH @ 100 µl min-1)
(d)
(e)
O
Cl
O
O
O
O
K+
O
NH
O
O
3
O
KCl 14 in MeOH
@ 100 µl min-1
4
H2N
Solvent Wash (MeOH @ 100 µl min-1)
(f)
(g)
O
O
O
O
O
O
NH
O
3
Acetic Acid 23 in
MeOH
Acetic Acid
23
in MeOH
@ 100 µl min-1
Solvent Wash (MeOH @ 100 µl min-1)
Scheme 6. Schematic illustrating the reaction steps used to evaluate the synthesis of tyramine acetate 4 under continuous flow.
(h)
metal ions, which leads to the observed increase in complex
stability.
In addition, acid catalyzed reactions are also commonly
employed and thus a range of acids, which are known to remove
inorganic ions from crown ether cavities, were tested for their
effect on ammonium complex stability, this time focusing on the
tyramine HCl 13 complex as it had been shown to be the least
stable of the three salts evaluated thus far. As Table 4 illustrates,
it was pleasing to see that the acids evaluated were found to
have no effect on the complex stability and thus could be

1624
G.P. Wild et al. / Tetrahedron 65 (2009) 1618–1629
chain length, steric hinderance, as well as the effect of counterions.
Si Cl
33
Again MeOH was used as the reaction solvent (100 m
L minꢁ1), re-
N
N
actant concentrations of 0.12 M (2.4ꢂ10ꢁ2 mmol) were employed,
methanolic KCl 14 (0.12 M, 2.4ꢂ10ꢁ2 mmol) afforded decom-
plexation and the reaction products were analyzed off-line by
HPLC.
N Si
NH
32
O
O
H
N
O
O
O
O
MeCN
O
3
24
O
Table 5 summarizes the results obtained and confirms initial
observations whereby no complexation is observed for free amines.
In keeping with previous findings the HCl salts of all amines
investigated, independent of chain length or aromaticity, were
found to complex efficiently. The issue of steric hinderance on
O
O
O
O
H
N
O
34
O
3
complexation was also evaluated, employing L-b-alanine benzyl
Scheme 7. Schematic illustrating the end-capping protocol employed to prepare TMS
SILICA 18-c-6 34.
ester HCl 29 and (S)-(ꢁ)-2-amino-3-phenyl-1-propanol HCl 30, in
both cases no complexation was observed. Benzamide HCl 31 also
failed to complex due to resonance effects experienced by the
carbonyl group, which reduces hydrogen bonding between the
crown ether and results in the formation of an unstable complex.
employed as reactants alongside this non-covalent protecting
group strategy.
2.1.5. The selective acetylation of tyramine
Having successfully developed a supported crown ether 24 ca-
pable of sequestering ammonium ions and subsequently evaluated
2.1.4.4. Generality of complexation. Recognizing the need for the
technique to N-protect compounds other than tyramine 1, a gen-
eral study was undertaken in order to evaluate factors such as
CF3CO2
(a)
OH
O
H3N
O
O
O
O
O
NH
O
3
OH
Excess Reactant
CF3CO2H . H2N
12
in THF
@ 100 µl min-1
Solvent Wash (THF @ 100 µl min-1)
(b)
CF3CO2
(c)
O
O
O
O
O
O
O
H3N
O
NH
O
3
O
O
Excess Reactants
N
21
O
27
in THF
@ 100 µl min-1
Solvent Wash (THF @ 100 µl min-1)
(d)
(e)
O
O
O
O
O
O
NH
O
O
34
3
O
TMEDA 26 in THF
@ 100 µl min-1
4
H2N
Scheme 8. Summary of the protocol employed for the continuous flow acetylation of tyramine 1, using TMS SILICA AM-18-c-6 34.

G.P. Wild et al. / Tetrahedron 65 (2009) 1618–1629
1625
the stability of the complex to a range of common reaction condi-
tions, the final step of the investigation was to perform a reaction
on the exposed phenolic moiety. When conducting the acetylation
of tyramine in batch, NaH 2 was employed as the base (Scheme 1),
however, when employing a crown ether as the protecting group,
the presence of sodium is undesirable due to its ability to deprotect
the amine (Table 4), consequently Et3N 21 and acetic anhydride 27
were selected suitable as reactants for the transformation.
With this in mind, Scheme 6 summarizes the proposed reaction
sequence for the continuous flow acetylation of tyramine 1, com-
prising of N-protection (step (a)), followed by O-acetylation (step
(c)), deprotection (step (e)) and crown ether regeneration (step (g))
all punctuated with solvent wash (steps (b), (d), (f) and (h)). Upon
evaluation of the reaction under continuous flow, it was disap-
pointing to observe that analysis of the reaction products obtained
reactor (0.15 g, 2.4ꢂ10ꢁ2 mmol). At this stage, the reaction se-
quence illustrated in Scheme 6 was repeated, affording 42% con-
version to tyramine acetate 4 and 58% un-reacted tyramine 1;
quantified via internal standardization with biphenyl. Although
a proportion of the N-protected tyramine had been successfully
acetylated and no N- or di-acetylation was observed, the detection
of un-reacted tyramine 1 was again attributed to the presence of
residual MeOH, owing to its use as a reaction solvent for the
complexation and decomplexation steps ((a) and (e)). Conse-
quently, all efforts to remove MeOH from the process were,
therefore, investigated.
To achieve this, two approaches were investigated, the first in-
volved exploring an alternative decomplexation strategy as KCl 14
was found to be insoluble in THF. As Table 3 illustrates, TMEDA 26
(0.12 M, 2.4ꢂ10ꢁ2 mmol) was shown to quantitatively decomplex
all three salts evaluated and owing to its increase solubility in THF
cf. KCl 14 it provided an alternative means of decomplexation (%
RSD¼4.4 (n¼5)). Furthermore, by employing an organic base, the
regeneration steps (f), (g) and (h) illustrated in Scheme 6 became
unnecessary as TMEDA 26 destabilized the complexed amine
without residing within the cavity, thus reducing the reaction cycle
to complexation, reaction and decomplexation. Using the modified
decomplexation strategy, the reaction was repeated and found to
afford 63% conversion to tyramine acetate 4 and 27% residual ty-
ramine 1. At this stage, the complexation step (a) was the only
methanolic step remaining, which proved necessary due to the
insolubility of tyramine HCl 13 in non-polar solvents. To circumvent
this, tyramine TFA 12 was employed, which was found to be ex-
tremely soluble in THF, alongside TMEDA 26 as illustrated in
Scheme 8. Using the refined protocol, complete removal of MeOH
from the reaction enabled 100% conversion of tyramine TFA 12 to
tyramine acetate 4, affording 2.4ꢂ10ꢁ2 mmol reactionꢁ1 (4.3 mg)
(Scheme 8).
from step (c) (Scheme 6), afforded only tyramine
1
(2.4ꢂ10ꢁ2 mmol). Therefore, in order to promote the acetylation
(step (c)), a range of flow rates (10–200 m
L minꢁ1), hence reaction
times, were subsequently investigated; unexpectedly all flow rates
failed to prepare the desired product 4. Based on this observation it
was postulated that the THF flush performed between reaction
steps was not removing MeOH from the reactor prior to the in-
troduction of the acetylating reagents and hence the volume of THF
was increased (from 1 mL to 2 mL), however, all subsequent re-
actions failed, yielding only un-reacted tyramine
1
(2.4ꢂ10ꢁ2 mmol).
As Mascagni5 and the work conduct within our laboratory had
shown analogous reactions to be possible in the solution phase, the
only remaining explanation for the reaction failing was an in-
teraction between the reagents and the solid-support itself,
resulting in either quenching or adsorption of the reactants. Con-
sequently, the SILICA AM-18-c-6 24 was treated with imidazole 32
and chlorotrimethylsilane 33 in MeCN to afford trimethylsilane
(TMS) end-capped SILICA AM-18-c-6 34 (Scheme 7), rendering the
support hydrophobic (represented as a white sphere around the
solid support).8
2.1.6. Further reactions of N-protected compounds
Having demonstrated the ability to complex a bifunctional
compound and selectively O-acetylate it under continuous flow
(Scheme 8), the next step of the investigation was to evaluate the
The TMS modified material 34 was subsequently filtered,
washed with acetone and oven dried prior to packing into the flow
(c)
CF3CO2
O
O
O
O
O
O
H3N
O
NH
O
3
N
N
Excess Reactants
36
in THF
@ 100 µl min-1
CF3CO2
(d)
O
O
O
O
O
H3N
O
O
NH
O
3
CH3I 28 in THF
@ 100 µl min-1
Excess Reactants
Scheme 9. Schematic illustrating the consecutive reaction steps required for the synthesis of 2-(4-methoxyphenyl)ethylamine.

1626
G.P. Wild et al. / Tetrahedron 65 (2009) 1618–1629
OH
(a)
O
O
O
O
H3N
Cl
O
O
O
NH
3
OH
Excess Reactant
HCl.H2N
35
in MeOH
@ 100 µl min-1
O-
Cl
(b)
O
O
O
O
O
H3N
O
NH
O
3
N
N
Excess Base
36
in THF
@ 100 µl min-1
O
O
O
O
O
Cl
(c)
H3N
O
O
NH
O
3
CH3I 28 in THF
100 µl min-1
Excess Reactant
O
O
(d)
O
O
O
O
NH
O
3
O
TMEDA 26 in THF
100 µl min-1
37
H2N
Scheme 10. Summary of the reaction protocol employed for the continuous flow synthesis of 4-methoxyphenylamine 37.
generality of the protecting group strategy. With the ability to
complex other bifunctional compounds previously illustrated
(Table 5), this prompted us to investigate the acetylation of
4-aminophenol. Once again, the use of MeOH as a complexation
solvent for 4-aminophenol HCl 35 (0.12 M, 2.4ꢂ10ꢁ2 mmol) was
found to be problematic, affording only 84% conversion to 4-ami-
nophenyl acetate. Replacing MeOH with THF and employing the
protocol depicted in Scheme 8, quantitative conversion to 4-ami-
nophenyl acetate was obtained, again affording a throughput of
respective methyl esters. However, unlike the acylations, where
a pre-mixed solution was employed to achieve deprotonation and
acetylation, the reagents selected for the alkylation, namely methyl
iodide 28 and 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) 36,9,10
could potentially react together. As such, the steps were performed
separately, as illustrated in Scheme 9, employing DBU 36 (0.12 M,
2.4ꢂ10ꢁ2 mmol) in THF at 100
m
L minꢁ1, followed by methyl iodide
28 (0.12 M, 2.4ꢂ10ꢁ2 mmol) in THF at 100
m
L minꢁ1. Decom-
plexation was again achieved using TMEDA 26 and analysis of the
reaction products by HPLC confirmed quantitative conversion of
tyramine TFA 12 to 2-(4-methoxyphenyl)ethylamine.
2.4ꢂ10ꢁ2 mmol reactionꢁ1
.
2.1.6.1. Alkylations. Having acetylated two bifunctional com-
pounds, the scope of the technique was extended to the O-alkyl-
ation of tyramine TFA 12 and 4-aminophenol HCl 35, to afford the
Using analogous reaction conditions to those reported for the
alkylation of tyramine 1, the methylation of 4-aminophenol HCl 35
(0.12 M, 2.4ꢂ10ꢁ2 mmol) was investigated and afforded 100%

G.P. Wild et al. / Tetrahedron 65 (2009) 1618–1629
1627
conversion of the starting material to 4-methoxyphenylamine 37
(2.4ꢂ10ꢁ2 mmol) (Scheme 10), with no N-methylation or di-
methylation observed.
In addition to the generality observed for both the complex
formation and subsequent reaction of the protected compounds, it
must be noted that a full 12 months after the initial investigation
was performed, the alkylation of 4-aminophenol HCl 35 was re-
peated using the same aliquot of TMS SILICA AM-18-c-6 34 and
found to afford analogous results, demonstrating long term sta-
bility of the immobilized crown ether 34.
H2O) (Fluka, UK), with the exception of DCM and tert-butanol,
which were of laboratory grade (Fisher Scientific, UK). All chro-
matography employed HPLC grade solvents (Fisher Scientific) and
purified water (5 MU
cmꢁ1) was prepared by reverse osmosis and
ion exchange using a water purifier (Elgast, UK) fitted with an
Option 4 cartridge. PTFE tubing (1/1600 o.d.ꢂ800
mm i.d.), femal
luers 10–32 (Tefzel), 1/1600 unions (Tefzel), o-rings (Viton), gas-tight
syringes (5 mL and 10 mL, Hamilton, UK) and Omnifit connectors
employed for the flow reactor system were sourced from Supelco
(UK) and Kinesis (UK). Borosilicate glass capillary (3 mm i.d.)
(DuranÒ, UK) was cut into the desired 50 cm lengths and flame
polished.
3. Conclusions
Owing to the complex nature of conventional covalent pro-
tecting group chemistry, it was the aim of this work to devise a non-
covalent protecting group strategy, which would enable the facile
N-protection of bifunctional compounds, thus removing the se-
lectivity issues associated with the reaction of bifunctional amines.
18-Crown-6 ethers have been shown to efficiently complex am-
monium salts and examples of reactions employing these com-
plexes have been reported within the literature. Application of the
technique to the protection of amines was, however, limited due to
problems largely associated with the removal of the crown ether
from the resulting reaction product. It was, therefore, proposed that
through the immobilization of an 18-crown-6 ether derivative, that
many of the issues that have prevented adoption of this technique
could be overcome.
Having investigated a range of immobilization strategies and
crown ether derivatives, it was found that the covalent immobili-
zation of AM-18-c-6 17 onto 3-carboxypropyl functionalized silica
gel (0.16 mmol gꢁ1) combined the properties of efficient complex-
ation and suitability for use in a continuous flow reactor. Using the
aforementioned material, a plethora of ammonium salts were
complexed (Table 5) and their stability to a wide range of solvents
(Table 3) and reactants (Table 4) was evaluated.
4.2. Instrumentation
Nuclear magnetic resonance (NMR) spectra were recorded at
room temperature as solutions in either deuterated chloroform
(CDCl3) or deuterated MeOH (CD3OD) using TMS as the internal
standard. All spectra were recorded on a Jeol GX400 spectrometer
and the chemical shifts given in parts per million (ppm) with
coupling constants in hertz (Hz). Elemental analyses were per-
formed using a Fisons (UK) Carlo Erba EA1108 analyser, with
measurements repeated until concurrent data was obtained, typi-
cally n¼2. Matrix-Assisted Laser Desorption Ionization (MALDI)-
Mass Spectrometry was performed using a Bruker Reflex 4
instrument operated in reflector mode. Inductively Coupled
Plasma-Mass Spectrometry (ICP-MS) measurements were made at
257.61 nm and 766.49 nm using a Perkin Elmer (UK) Optima
5300DV instrument. Melting points were obtained using a Stuart
Scientific (UK) SMP10 apparatus and are reported uncorrected.
High Performance Liquid Chromatography (HPLC) data was
obtained using a Jasco (UK) modular system comprising of a LV-
1580-03 solvent selector, a DG-1580-53 degasser, two PU-1580
pumps, an HG-1580-32 mixer, a UV-1575 detector and an AS-1555
autosampler. Analytical measurements were made using a Jupiter
Once the scope and limitations of the operating conditions had
been evaluated, reaction of the complexed tyramine salt was per-
formed and found to afford the selective O-acetylation, providing
an efficient route to the synthesis of tyramine acetate 4 cf. the la-
borious route required when employing covalent protecting groups
(Scheme 2). Having demonstrated the quantitative conversion of
tyramine TFA salt 12 to the free tyramine acetate 4, the in-
vestigation was extended to O-alkylation, demonstrating again
the selective synthesis of 2-(4-methoxyphenyl)ethylamine in
quantitative conversion, with no sign of competing di-alkylation or
N-alkylation products. The generality of the technique was sub-
sequently explored using 4-aminophenol HCl 35, which enabled
the facile synthesis of 4-aminophenyl acetate and 4-methoxy-
phenylamine 37 in quantitative yield, respectively. In all cases, the
free amine was afforded and products were obtained in higher
purity than those prepared using conventional N-protecting group
strategies.
10
m
m C18, 300 A (250ꢂ4.60 mm) column (Phenomenx, UK). Re-
agents and solutions were delivered to the continuous flow reactor
using a Harvard syringe pump (UK) capable of delivering liquids at
flow rates ranging from 0.1 to 1000.0 m
L minꢁ1 based on a 5 mL
gas-tight syringe. Where necessary, aliquots of reactants were
introduces into the continuous flow reactor using a Rheodyne
injector valve, model 7125 (Supelco, UK).
4.3. HPLC method
Using a gradient elution at a flow rate of 1.5 mL minꢁ1, the
aqueous portion of the mobile phase was decreased from 60% to
40% over a period of 7 min and then maintained at 40% for the
remaining 30 min of the method. Both the organic phase (MeOH)
and aqueous phases contained 0.1% TFA. An injection volume of
20 mL was employed and biphenyl used as the internal standard.
In summary, the work described herein presents a broad in-
vestigation into the viability of immobilized crown ethers as a re-
placement for traditional covalent N-protecting group chemistry
and combines their use with continuous flow technology to afford
a technique that has potential for future automation.
4.4. Preparation of silica gel immobilized aminomethyl-18-
crown-6 ether 24
Thionyl chloride 19 (0.37 mL, 5.11 mmol) was added to a stirred
solution of oven dried 3-carboxypropyl functionalized silica gel
(1.06 g, 1.6 mmol gꢁ1, 200–400 mesh) in toluene (20 mL) and the
reaction mixture heated to reflux for 3 h. The resulting silica sup-
ported acid chloride was concentrated in vacuo to afford a free
flowing white solid, which was subsequently redispersed in tolu-
ene (20 mL), prior to the addition of 2-aminomethyl-18-crown-6
ether 17 (0.50 g,1.70 mmol), followed by triethylamine 21 (0.26 mL,
1.86 mmol). The reaction mixture was stirred overnight, under N2,
prior to filtration under vacuum. The supported crown ether 24 was
then washed (H2O, acetone and DCM) and oven dried, at 90 ꢀC, to
4. Experimental section
4.1. Reagents and materials
Unless otherwise stated, the chemicals employed herein were
used as received and purchased from Sigma Aldrich, Acros and
Avocado. Where available, reactions were performed using puriss
grade solvents, which were stirred over molecular sieves (<0.005%

1628
G.P. Wild et al. / Tetrahedron 65 (2009) 1618–1629
afford
a
free flowing white powder (1.08 g, 72.2%); ICP-MS
from waste to sample collection; the reaction products were ana-
lyzed off-line by HPLC.
0.162 mmol gꢁ1
.
4.5. Silanisation of SILICA AM-18-c-6 24 to afford TMS SILICA
AM-18-c-6 34
4.7.3. Regeneration of the immobilized crown ether
In the cases where KCl 14 in MeOH was employed for the
decomplexation step, the immobilized crown ether cavity was
Imidazole 32 (1.36 g, 20.00 mmol) and chlorotrimethylsilane 33
(2.51 mL, 20.00 mmol) were added to a stirred solution of SILICA
AM-18-c-6 24 (0.50 g, 0.16 mmol gꢁ1) in MeCN (10 mL) under N2.
After stirring at room temperature for 1 h, the reaction mixture was
filtered under suction and washed with MeCN (20 mL), acetone
(20 mL) and DCM (20 mL) prior to oven drying at 60 ꢀC to afford
a free flowing white powder 34.
regenerated using methanolic acetic acid 23 (200 mL, 5% v/v). In this
case, the reactor effluent was again diverted to waste and the sys-
tem purged with the reaction solvent for 2 min prior to repeating
the aforementioned steps.
4.8. Flow synthesis of 4-(2-aminoethyl)phenyl acetate 4
Using the general flow procedure detailed above, tyramine ac-
4.6. Flow reactor set-up
etate 4 was synthesized under continuous flow. To achieve this,
tyramine TFA 12 (200
m
L, 0.12 M, 2.4ꢂ10ꢁ2 mmol), in THF, was in-
As illustrated in Figure 3, the syringe driver was interfaced to
inlet 2 of the Rheodyne valve (7125) using a series of commercially
available connectors and a length of PTFE tubing (800 mm i.d.). In-
terconnects were made between the PTFE tubing and the luer lock
gas-tight syringes using a female to male 10–32 (Tefzel) luer, a 1/
1600 union (Tefzel) and a 1/1600 HPLC connector (PEEK). The flow
troduced into a continuous stream of THF (100 m
L minꢁ1) via
a Rheodyne valve. To ensure the tyramine salt 12 had passed
through the system containing TMS SILICA AM-18-c-6 34 (0.15 g,
2.4ꢂ10ꢁ2 mmol) before introducing additional reactants, THF was
pumped through the system for 2 min. Reaction of the phenolic
moiety was achieved via introduction of a pre-mixed solution of
reactor comprised of
a
borosilicate glass capillary (50 mm
acetic anhydride 27 and Et3N 21 (200
a flow rate of 100 m
m
L, 0.12 M, 2.4ꢂ10ꢁ2 mmol) at
(length)ꢂ3 mm (i.d.)), packed with w0.15 g of the immobilized
crown ether under investigation, the inlet of which was connected
to the Rheodyne valve via outlet 3. The stainless steel sample loop
L minꢁ1. The system was again purged with THF
(2 min), prior to initiating decomplexation using TMEDA 26
(200
m
L, 0.12 M, 2.4ꢂ10ꢁ2 mmol), in THF, at 100
m
L minꢁ1. At this
(200
mL) was positioned across outlets 1 and 4, with excess re-
point the solvent stream was diverted from waste to sample col-
lection and the reaction products collected prior to off-line analysis
by HPLC; whereby comparison with a synthetic standard10 (HPLC,
tR¼6.3 min) confirmed quantitative conversion of tyramine TFA 12
actants diverted to waste via outlet 6 of the valve and reaction
products collected in a sample vial (1.5 mL) at the reactor outlet.
4.7. General flow reaction protocol
to 4-(2-aminoethyl)phenyl acetate 4 with a throughput of
4.3ꢂ10ꢁ2 g reactionꢁ1 (2.3ꢂ10ꢁ2 mmol).
Prior to performing a flow reaction, the solvent under in-
vestigation (THF, MeOH or DCM) was driven through the packed-
4.9. Flow synthesis of 4-aminophenyl acetate
bed reactor at 100 m
L minꢁ1 to waste.
Using the general flow protocol detailed above, 4-aminophenyl
acetate was synthesized under continuous flow. To achieve this, 4-
4.7.1. Crown ether complexation
To prepare the immobilized crown ether complex, the sample
loop was filled with a solution of the ammonium salt under in-
aminophenol HCl 35 (200
m
L, 0.12 M, 2.4ꢂ10ꢁ2 mmol), in THF, was
introduced into a continuous stream of THF (100 m
L minꢁ1) via
vestigation (200
m
L) with the valve in the load position. Once filled,
a Rheodyne valve. To ensure salt 35 had passed through the system
before introducing additional reagents, THF was pumped through
the system for 2 min. Reaction of the phenolic moiety was achieved
via introduction of a pre-mixed solution of acetic anhydride 27 and
the valve was turned to the inject position and the solvent stream
diverted through the sample loop and the ammonium salt pumped
through the packed bed at 100
diverted to waste.
m
L minꢁ1 and the reactant stream
Et3N 21 (200
100
m
L, 0.12 M, 2.4ꢂ10ꢁ2 mmol) at
a flow rate of
m
L minꢁ1. The system was purged with THF (2 min) prior to
4.7.1.1. Acetylation of the ammonium complex. To acetylate the
immobilized crown ether complex, a pre-mixed solution of acetic
initiating decomplexation using TMEDA 26 (200
m
L, 0.12 M,
2.4ꢂ10ꢁ2 mmol), in THF, at 100
m
L minꢁ1. At this point the solvent
anhydride 27 and Et3N 21 (200
m
L) in THF was pumped through the
stream was diverted from waste to sample collection and the re-
action products collected prior to off-line analysis by HPLC. Com-
parison with a synthetic standard (HPLC, tR¼3.2 min) confirmed
quantitative conversion of 4-aminophenol HCl 35 to 4-amino-
reactor at 100
m
L minꢁ1, which facilitated both deprotonation of the
phenolic moiety and the subsequent acetylation in a single step.
THF was then pumped through the system for 2 min, prior to ini-
tiating decomplexation.
phenyl acetate, with a throughput of 2.3ꢂ10ꢁ2 mmol reactionꢁ1
.
4.7.1.2. Alkylation of the ammonium complex. To alkylate the
4.10. Flow synthesis of 2-(4-methoxyphenyl)ethylamine
immobilized crown ether complex, a solution of DBU 36 (200
THF was pumped through the reactor at a flow rate of 100
L minꢁ1
followed by THF for 2 min prior to the introduction of the alkylating
agent (200 L) in THF at 100
L minꢁ1. This process was found to
mL) in
m
,
Using the general flow procedure detailed above, tyramine TFA 12
(200
m
L, 0.12 M, 2.4ꢂ10ꢁ2 mmol) in THF was introduced into a con-
m
m
tinuous THF stream (100 m
L minꢁ1) via a Rheodyne valve. To ensure
enable deprotonation of the phenolic moiety and subsequent al-
kylation; throughout this step, the reaction products were diverted
to waste.
tyramine salt 12 had passed through the system containing TMS
SILICA AM-18-c-6 34 (0.15 g, 2.4ꢂ10ꢁ2 mmol), before introducing
further reagents, THF (100 m
L minꢁ1) was pumped through for 2 min.
The first of the reaction steps incorporated a 200 mL plug of DBU 36
4.7.2. Decomplexation of the immobilized product
(0.12 M, 2.4ꢂ10ꢁ2 mmol), followed by methyl iodide 28 (200
mL,
The decomplexing agent, KCl 14 in MeOH (200
in THF (200 L), was pumped through the packed bed at a flow rate
of 100
L minꢁ1 and at this point, the reactant stream was diverted
m
L) or TMEDA 26
0.12 M, 2.4ꢂ10ꢁ2 mmol); both of which were passed through the
m
flow reactor at 100 m
L minꢁ1, facilitating the deprotonation of the
phenolic moiety and methylation in consecutive steps. Again 2 min
m

G.P. Wild et al. / Tetrahedron 65 (2009) 1618–1629
1629
was allowed before initiating decomplexation via a 200
m
L plug of
work described herein. Eichrom Technologies (France) are
thanked for their kind donation of di-tert-butylcyclohexano-18-
crown-6 ether. Mr. Bob Knight, Dr. Kevin Welham and Mr. Mike
Bailey (The University of Hull) are also thanked for their assis-
tance with ICP-MS, MS and flow reactor fabrication,
respectively.
TMEDA 26 (0.12 M, 2.4ꢂ10ꢁ2 mmol) at 100
m
L minꢁ1. At this point,
the solvent stream was diverted from waste to sample collection and
the reaction products analyzed off-line by HPLC (tR¼6.1 min)
whereby 100% conversion to 3-(4-methoxyphenyl)ethylamine was
obtained (2.3ꢂ10ꢁ2 mmol reactionꢁ1).
4.11. Flow synthesis of 4-methoxyphenylamine 37
Supplementary data
4-Aminophenol HCl 35 (200
m
L, 0.12 M, 2.4ꢂ10ꢁ2 mmol) in THF
was introduced into a continuous THF stream (100 m
L minꢁ1) via
Supplementary data associated with this article can be found in
the online version, at doi:10.1016/j.tet.2008.12.052.
the Rheodyne valve. To ensure 4-aminophenol salt 35 has passed
through the system containing TMS SILICA AM-18-c-6 34 (0.15 g,
2.4ꢂ10ꢁ2 mmol), before introducing further reagents, THF
(100 m
L minꢁ1) was pumped through the reactor for 2 min. The
References and notes
reaction steps firstly incorporated a 200 mL plug of DBU 36 (0.12 M,
2.4ꢂ10ꢁ2 mmol), followed by methyl iodide 28 (200
mL, 0.12 M,
1. Wuts, P. G. M.; Greene, T. W. Greene’s Protective Groups in Organic Synthesis, 4th
ed.; Wiley Interscience: New Jersey, NJ, 2007.
2. Pederson, C. J. J. Am. Chem. Soc. 1967, 89, 7017.
2.4ꢂ10ꢁ2 mmol); both of which were pumped at a flow rate of
100 m
L minꢁ1, facilitating the deprotonation and methylation of the
3. Buschmann, H.-J.; Mitihac, L. Anal. Chim. Acta 2002, 466.
4. (a) Batinic-Haberele, I.; Crumbliss, A. L.; Spasojevic, I.; Bartsch, R. A. J. Chem. Soc.,
Dalton Trans. 1995, 2503; (b) Kunishima, M.; Hioki, K.; Moriya, T.; Morita, J.;
Ikuta, T.; Tani, S. Angew. Chem., Int. Ed. 2006, 45, 1.
5. Mascagni, P.; Botti, P.; Ball, H. L.; Rizzi, E.; Lucietto, P.; Pinori, M. Tetrahedron
1995, 51, 5447; (a) Mascagni, P.; Hyde, C. B. Tetrahedron Lett. 1990, 31, 339; (b)
Mascagni, P.; Hyde, C. B.; Welham, K. J. J. Chem. Soc., Perkin Trans. 2 1989, 2011;
Mascagni, P.; Hyde, C. B.; Charalambous, M. A.; Welham, K. J. J. Chem. Soc., Perkin
Trans. 2 1987, 323.
complexed phenolic derivative. Again, 2 min was allowed before
initiating decomplexation via a 200 mL plug of TMEDA 26 (0.12 M,
2.4ꢂ10ꢁ2 mmol). At this point the solvent stream was diverted from
waste to sample collection and the reaction products analyzed off-
line by HPLC, affording 100% conversion with a throughput of
2.4ꢂ10ꢁ2 mmol reactionꢁ1; HPLC, tR¼5.7 min.
6. Mascagni, P.; Botti, P.; Ball, H. L.; Rizzi, E.; Lucietto, P.; Pinori, M. Tetrahedron
1995, 51, 5447.
7. The reactor volume was determined experimentally to be 35 mL.
Acknowledgements
8. Verzele, M.; Dewaele, C.; Mussche, P. J. High Resolut. Chromatogr. 1982, 5, 616.
9. Mai, D. Synth. Commun. 1986, 16, 331.
10. See, Supplementary data for additional experimental details.
The authors wish to acknowledge the EPSRC (G.P.W. and
C.W.) (Grant No. GR/S34106/01) for full financial support of the