The role of β-cyclodextrin in mediating regioselective dimethylaminomethylation of phenol

Article abstract of DOI:10.1016/j.tet.2017.12.007

Regioselective reactions with supramolecular control are of great interest. Herein, the para-regioselectivity in the Mannich reaction of phenol with formaldehyde and dimethylamine was achieved with the use of β-cyclodextrin (β-CD), giving 4-(N,N-dimethylaminomethyl)phenol (p-AP) as major product. ¹H NMR and ITC measurements of the binding of β-CD with the reactants and the products o- and p-AP revealed a new mechanism, in which β-CD includes p-AP instead of phenol to control the reaction regioselectivity. This product-inclusion mechanism is remarkably different to the known reactant-inclusion process.

Full text of DOI:10.1016/j.tet.2017.12.007

Tetrahedron 74 (2018) 360e365
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
The role of
b
-cyclodextrin in mediating regioselective
dimethylaminomethylation of phenol
Wen-Hai Wu, Jiang Duan, Ting Wei, Hai-Yang Tu**, Ai-Dong Zhang*
Key Laboratory of Pesticide & Chemical Biology of Ministry of Education, College of Chemistry, Central China Normal University, Wuhan 430079, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
Regioselective reactions with supramolecular control are of great interest. Herein, the para-regiose-
lectivity in the Mannich reaction of phenol with formaldehyde and dimethylamine was achieved with
Received 27 September 2017
Received in revised form
29 November 2017
Accepted 5 December 2017
Available online 6 December 2017
the use of
1H NMR and ITC measurements of the binding of
revealed a new mechanism, in which -CD includes p-AP instead of phenol to control the reaction
b
-cyclodextrin (
b
-CD), giving 4-(N,N-dimethylaminomethyl)phenol (p-AP) as major product.
b
-CD with the reactants and the products o- and p-AP
b
regioselectivity. This product-inclusion mechanism is remarkably different to the known reactant-
inclusion process.
Keywords:
© 2017 Elsevier Ltd. All rights reserved.
Supramolecular chemistry
b-cyclodextrin
Mannich reaction
Regioselectivity
Product-inclusion mechanism
1. Introduction
the presence of
accounted in terms of 1H NMR measurements11 that the inclusion
complex formation of -CD with benzoic acid and thereby the
blocking of the ortho/meta sites was essential to achieve the perfect
regioselectivity. -CD, which has a larger cavity, was reported to
include both cinnamic acid and coumarin, forming a 1:1:1 ternary
inclusion complex that kept nearby the C¼C bonds of the two
partners and afforded a heterodimerization instead of homo-
dimerization product under light irradiation.12
b-CD. The mechanism of this reaction was
Supramolecularly controlled regioselective reactions have been
widely exploited for the synthesis of a variety of compounds that
are unable or difficult to be synthesized using common methods. A
number of host molecules such as crown ethers,1 cucurbiturils,2
and cyclodextrins (CDs)3 are known to have this role to mediate
regioselective reactions through forming supramolecular inclusion
complexes by hydrogen bonding, electrostatic, van der Waals and
b
g
p
-
p
interactions, as well as steric, shape-complementary and hy-
Phenol is a kind of electron-rich aromatic compound with high
reactivity toward electrophiles. Due to the strong ortho/para
orientation of phenolic hydroxyl, the reaction of phenol tends to
give products with non-selective ortho- and para-substitution. A
few works regarding the para-selective reaction of phenol have
been reported using supramolecular methods. For instance,
Komiyama et al. realized the synthesis of 4-hydroxy benzaldehyde
drophilic/hydrophobic effects.4 The inclusion complex is able to
block some reaction sites of the substrate, or stabilize the transition
state or the intermediate, leading to control of a chemio-, regio-,
stereo- and/or enantio-selective reaction toward a desired direc-
tion.5,6 CDs are among the most important supramolecular hosts
and many of their mediated regioselective reactions are known.7e9
For example, Breslow et al.10 utilized
a
-CD to include anisole,
and 4-hydroxybenzoate with high para-selectivity via
b-CD
realizing the para-selective synthesis of 4-chloroanisole in the re-
action with HOCl. Hirai et al.11 succeeded in introducing carboxyl to
the para-site of benzoic acid selectively to yield terephthalic acid
through the reaction of benzoic acid with CCl4 and copper under
controlled formylation13 and carboxylation14 of phenol.
b
-CD was
also usable in achieving the para-Reimer-Tiemann reaction of
phenol to give 4-hydroxybenzaldehyde with high regioselectivity
under light irradiation.15 Further, with the presence of
b-CD, a mild
and efficient protocol was developed for the para-selective
bromination of substituted phenols with Br2/CCl4 in aqueous or
methanol medium.16 Although solid evidence was still needed,
these reports proposed the reactions occurring through a mecha-
* Corresponding author.
** Corresponding author.
E-mail addresses: haiytu@mail.ccnu.edu.cn (H.-Y. Tu), adzhang@mail.ccnu.edu.
cn (A.-D. Zhang).
nism where the encapsulation of phenol by b-CD to block the ortho-
https://doi.org/10.1016/j.tet.2017.12.007
0040-4020/© 2017 Elsevier Ltd. All rights reserved.

W.-H. Wu et al. / Tetrahedron 74 (2018) 360e365
361
sites was the necessary process to give the desired para-substituted
products.
(83%) and the selectivity ratio (65/35) (Entry 5). At a 1.5-fold excess
of -CD, the rate was improved to 95% and the ratio of p-AP/o-AP to
b
The three-component Mannich reaction of phenol, formalde-
hyde and amine could be performed in a one-pot manner but al-
ways produces a mixture of ortho/para mono- or multi-substituted
products.17,18 To overcome this difficulty, two variations of Mannich
reaction for synthesis of ortho-substituted products were reported;
one was the ortho-selective reaction of phenol to give 2-(N,N-
dimethylaminomethyl)phenol (o-AP) with 81% regioselectivity
through use of the Eschenmoser salt19; the other was the ruthe-
nium catalytic dehydrogenation of methanol as C1 source to react
with dimethylamine and phenol to yield o-AP.20 However, the para-
selective Mannich reaction of phenol with formaldehyde and
dimethylamine to give 4-dimethylamino methylphenol (p-AP) has
thus far not been reported. In this work, we demonstrated that,
89/11 (Entry 6). Further increasing the amount of
result in a more profound improvement (Entry 7).
b-CD did not
The above results show that the use of
b-CD as a mediator
enabled the para-selective N,N-dimethylaminomethylation of
phenol for the first time. To highlight the feasibility of this supra-
molecular method, a practical 10-fold scale-up synthesis of p-AP, in
conjunction with a simplified purification procedure, was done and
described in the experiment section. It remains unclear, however,
whether the achievement of para-regioselectivity in this reaction is
a result of the known mechanism that the b-CD cavity includes the
reactant phenol to alter the reaction regioselectivity.22
2.2. 1H NMR measurements
among various CDs,
b-CD was able to achieve the para-regiose-
lectivity of the Mannich reaction. Furthermore, the role of b-CD was
To gain insight into the mechanism, a supramolecular 1H NMR
explored in-depth with 1H NMR and ITC measurements, and the
results revealed a process that is remarkably different to the
existing mechanism.
method23 was used to investigate the interactions of
b-CD with the
reactant phenol, the intermediate N,N-dimethyliminium (DMA)
generated from the reaction of formaldehyde and dimethylamine,
and the products p- and o-AP. The criterion of this method is the
proton signal upfield shift of a guest molecule when included in the
2. Results and discussion
cavity of
magnetic field.23 Scheme 1 depicts the structures and proton labels
of phenol, DMA, and the products, as well as one glucose unit of
CD with numbered protons H1-H6 and their relative positions to the
b-CD host, due to the shielding effect to the applied
2.1. Effect of cyclodextrins on the Mannich reaction
b
-
At the beginning of this work, the Mannich reaction of phenol
with formaldehyde and dimethylamine in the absence and pres-
ence of various CDs serving as a supramolecular host was per-
formed and monitored with 1H NMR. When the reaction was
conducted in the absence of any CD, a low phenol conversion rate of
77% and a non-regioselective formation of p-AP/o-AP in ratio of 53/
b-CD cavity.
According to the molecular model and the literature,24
b-CD
consists of seven a-D-glucose units with 1,4-linkage to form a cone-
type structure, where protons H5 and H3 locate inside the cone with
the former close to the narrow rim (the primary hydroxyl rim) and
the latter to the wide rim (the secondary hydroxyl rim); while the
protons H1, H2, H4 and H6 reside outside the cavity. Since H2O was
the medium in the Mannich reaction, D2O was used for the 1H NMR
measurements, and therefore, the active proton signals of phenol,
47 were observed (Table 1, Entry 1). In the presence of equimolar
CD, the conversion rate was improved whereas the product ratio
had no obvious alteration (Entry 2). When equimolar - or -CD
was used, both the conversion rate and the product ratio were
improved significantly. With use of -CD the conversion rate and
a-
b
g
b
the products and b-CD were invisible due to the DH exchange. The
chemical shift variations (Dd, ppm) of the inner protons H3 and H5
can be used as a measure to evaluate the host-guest binding
strength and the inclusion depth and orientation of the guest in the
the product ratio were raised up to 92% and 75/25, respectively
(Entry 3), while the corresponding rate and ratio were 90% and 63/
37 in the presence of
host molecule in this para-selective reaction. The inferior effects of
- and
-CD might come from their unsuitable cavity sizes.21 In
addition, the employed amount of -CD was optimized. A half equiv
of -CD led the reaction to the decease of both the conversion rate
g-CD (Entry 4), indicating b-CD was the best
cavity.25 The Dd for phenol, DMA, o-AP, p-AP and
b-CD in their in-
a
g
dividual complexation states are summarized in Table 2, and their
1H NMR spectra sets shown in Figs. 1e4.
b
b
The interaction of phenol with b-CD was first examined. For
phenol, small upfield shifts of Hfꢀa (ꢀ0.01 ppm), Hfꢀb (ꢀ0.02 ppm)
and Hfꢀc (ꢀ0.03 ppm) are observed (Fig. 1 and Table 2), indicating
Table 1
Conversion rate of phenol and selectivity toward p-/o-AP.a
that phenol receives a negligible shielding effect from b-CD and
does not insert into the cavity. Relatively, Hfꢀc has a larger Dd
,
Entry
CD
Phenol/CD
Conversion (%)b
p-AP/o-APc
1
2
3
4
5
6
7
none
e
77
83
92
90
83
95
96
53/47
54/46
75/25
63/37
65/35
89/11
90/10
a
b
-CD
-CD
1: 1
1: 1
1: 1
1: 0.5
1: 1.5
1: 2
g-CD
b-CD
b-CD
b-CD
a
Reaction conditions: a molar ratio of 1:1:1 was used for phenol, formaldehyde
and dimethylamine.
b
Conversion rate was calculated based on the recovered amount of phenol.
c
Product ratio was calculated based on the1H NMR signal intensities of methy-
Scheme 1. Structures and proton labels of phenol, DMA, the products and one glucose
unit in Newman projection of the -CD model.
lene in p-AP and o-AP.
b

362
W.-H. Wu et al. / Tetrahedron 74 (2018) 360e365
Table 2
Proton chemical shift variations in complexation (Dd, ppm).a
Phenol/
Phenol
b
-CD
DMA/
b-CD
b
-CD
DMA
b-CD
Hfꢀa (ꢀ0.01)
Hfꢀb (ꢀ0.02)
Hfꢀc (ꢀ0.03)
H1 (ꢀ0.02)
H2 (ꢀ0.03)
H3 (ꢀ0.05)
H4 (ꢀ0.04)
H5 (ꢀ0.11)
H6 (ꢀ0.02)
H1 (ꢀ0.01)
H2 (ꢀ0.02)
H3 (ꢀ0.03)
H4 (ꢀ0.02)
H5 (ꢀ0.03)
H6 (ꢀ0.02)
Hi-e (þ0.02)
Hi-f (þ0.03)
o-AP/
o-AP
b-CD
p-AP/
b-CD
b
-CD
p-AP
b-CD
Hoꢀa (þ0.11)
Hoꢀb (þ0.01)
Hoꢀc (þ0.10)
Hoꢀd (ꢀ0.03)
Hoꢀe (ꢀ0.27)
Hoꢀf (ꢀ0.15)
H1 (ꢀ0.01)
H2 (ꢀ0.02)
H3 (ꢀ0.03)
H4 (ꢀ0.01)
H5 (ꢀ0.05)
H6 (ꢀ0.01)
Hp-a (þ0.03)
Hp-b (ꢀ0.08)
H1 (ꢀ0.04)
H2 (ꢀ0.03)
H3 (ꢀ0.13)
H4 (ꢀ0.02)
H5 (ꢀ0.20)
H6 (ꢀ0.04)
Hp-e (ꢀ0.14)
Hp-f (ꢀ0.20)
a
Signs “þ” and “ꢀ” indicate downfield and upfield shift, respectively.
different shift direction in proton signals (Fig. 2 and Table 2), such
that the small downfield shifts of protons Hi-e (þ0.02 ppm) and Hi-f
(þ0.03 ppm) are observed, as an indicative of the DMA residing in
the deshielding region outside b-CD. Meanwhile, the b-CD inner
protons H3 (ꢀ0.03 ppm) and H5 (ꢀ0.03 ppm) have negligible Dd
,
further suggesting that DMA does not enter the b-CD cavity. This
may be a result of the highly hydrophilic nature of ionic DMA that is
repulsive to the hydrophobic attraction of -CD cavity.
b
Fig. 3 shows the 1H NMR spectra set of o-AP, o-AP/
b-CD mixture
and
b
-CD. The aromatic Hoꢀa (þ0.11 ppm) and Hoꢀc (þ0.10 ppm) in
o-AP exhibit relatively large downfield shifts; whereas Hoꢀb
(þ0.01 ppm) has a negligible downfield Dd and Hoꢀd (ꢀ0.03 ppm) a
.
More significantly, the methylene Hoꢀe
Fig. 1. 1H NMR spectra set of phenol (top), phenol/
b-CD mixture (middle) and b-CD
(bottom) in D2O at 0.1 mmol.
slight upfield Dd
(ꢀ0.27 ppm) and the methyl Hoꢀf (ꢀ0.15 ppm) of N,N-dimethyla-
minomethyl have large upfield Dd. The 1H NMR data manifest that
the different parts of the o-AP molecule suffer different interactions
meaning that it is near to the shielding region of the cavity. As for b-
from the induced magnetic field of
created, where phenyl resides at the interface of the shielding/
deshielding regions near the narrow rim of -CD cone while N,N-
b-CD. Thus a scenario can be
CD, the inner protons H3 (ꢀ0.05 ppm) and H5 (ꢀ0.11 ppm) have
upfield shifts with the latter more significant. Because H5 locates at
the narrow rim, its large upfield shift would be a result caused by
the anisotropic magnetic field shielding of phenol.26 Using circular
dichroism spectroscopy in combination with molecular mechanical
calculation, Marconi et al.27 demonstrated that phenol resides
slantingly near the narrow rim, in line with our observation. This
binding conformation may take use of the hydrogen bonding be-
tween the hydroxyl group of phenol and the primary hydroxyl
b
dimethyl aminomethyl locates inside the b-CD cavity. On the other
hand, the host
b
-CD displays small upfield Dds for H3 (ꢀ0.03 ppm)
and H5 (ꢀ0.05 ppm) in the cavity, presumably due to the minor
effect from the aromatic ring of o-AP; while the relatively larger
upfield Dd for H5 means that o-AP enters the
b-CD cavity through
the narrow rim.
The 1H NMR spectra set of the product p-AP, the p-AP/
b
-CD
group of
CD cavity.
The intermediate DMA in complexation shows a completely
b b-
-CD,28 impeding the entry of the phenol ring into the
mixture and
b-CD is shown in Fig. 4. The phenolic hydroxyl-ortho
Fig. 2. 1H NMR spectra set of DMA (top), DMA/
b-CD mixture (middle) and
b-CD
Fig. 3. 1H NMR spectra set of o-AP (top), o-AP/
b-CD (middle) and b-CD (bottom) in D2O
(bottom) in D2O at 0.1 mmol.
at 0.1 mmol.

W.-H. Wu et al. / Tetrahedron 74 (2018) 360e365
363
Fig. 4. 1H NMR spectra set of p-AP (top), p-AP/
b-CD (middle) and b-CD (bottom) in D2O
at 0.1 mmol.
Hp-a (þ0.03 ppm) and hydroxyl-meta Hp-b (ꢀ0.08 ppm) of p-AP
shift oppositely to downfield and upfield, suggesting the phenolic
Fig. 5. ITC data in 298 K for p-AP (A) and o-AP (B) to complex with b-CD.
hydroxyl group may point outside the b-CD cavity with the phenyl
D
G and K (2 orders higher in both cases for p-AP and o-AP,
ring residing in a half-insertion state because of the small positive
Dd of Hp-a. Moreover, the methylene Hp-e (ꢀ0.14 ppm) and the
methyl Hp-f (ꢀ0.20 ppm) of p-AP exhibit large upfield Dds, thus N,N-
Supporting Information), the thermodynamic parameters (Table 3)
derived from the first step binding are used for further discussion.
The values of
D
H1,
D
S1,
D
G1 and K1 for p-AP/
b
-CD are ꢀ25.68 0.08
dimethylaminomethyl should be inside the
b
-CD cavity with
KJ/mol, 0.066 KJ/mol$K, ꢀ45.35
0.08 KJ/mol and
methyl Hp-f more deeply inserted. As for
b
-CD, the wide rim H3
8.26
1.08 ꢁ 107
M
ꢀ1, respectively. In contrast, the o-AP/
b
-CD
(ꢀ0.13 ppm) and the narrow rim H5 (ꢀ0.20 ppm) in the cavity shift
upfield remarkably, most likely attributed to the impact of the
strong aromatic induced magnetic field of p-AP in the cavity. In
addition, the larger Dd for H5 suggests the phenyl of p-AP locating
generates
the
respective
thermodynamic
parameters
as ꢀ23.57 0.08 KJ/mol, 0.062 KJ/mol$K, ꢀ42.05 0.08 KJ/mol and
2.28 0.29 ꢁ 107 Mꢀ1, respectively. As a result, the
DH1 of p-AP/b-
CD is 2.11 KJ/mol lower and the DS1 0.004 KJ/mol$K higher than that
near to the narrow rim of
b-CD cone.
of o-AP/b-CD, leading to the binding constant of p-AP/b-CD 3.6-fold
higher.
According to the results from 1H NMR data, phenol may not
enter into the b-CD cavity but lies slantingly near the narrow rim as
The differences in thermodynamic parameters reflect that p-AP
binds with -CD more tightly than o-AP, consistent with the results
speculated from the 1H NMR data. This binding preference of
-CD
with p-AP other than o-AP in the competitive aminomethylation of
phenol comes from the larger
G1 lowering (ꢀ45.35 vs ꢀ42.05 KJ/
mol, Table 3), thereby disrupting the non-selectivity as in the case
an outer-sphere complex29 through the hydrophobic and hydrogen
b
bonding effects. By contrast, both the products o-AP and p-AP enter
b
the cavity and form inclusion complexes with
b-CD, though in
different binding modes. Another point worthy of mention is that
the entry of the products into the cavity most likely occurs upon the
D
product formation, implying the
place. Thus we postulated that the
aminomethylation of phenol takes place not through the
complexation of -CD with phenol and blocking the ortho-reaction
sites, but instead through a binding process involving the formation
of the inclusion complexes of -CD with the individual products.
Since p-AP inserts more deeply inside the -CD cavity than o-AP,
b-CD cavity is not the reaction
of without
product.
b-CD and leading to the formation of p-AP as major
b
-CD mediated para-selective
The bindings of p-AP and o-AP with
using UVevis absorption spectroscopy. The measurements were
done by keeping the concentration of p-AP or o-AP at 50 M while
varying the molar ratio of -CD to the guest in the range from 0 to
0.7 with 0.1 intervals. The inclusion of p-AP and o-AP with -CD
resulted in the increase of absorption and both the complexes
gave maximum absorptions at the guest/ -CD molar ratio of 1:0.5
b-CD were also examined
b
m
b
b
b
b
there should be a thermodynamically controlled process favouring
the reaction toward the direction of para-selectivity.
b
with lmax 272 and 275 nm, respectively (Supporting Information),
meaning that the binding ratio was 2:1 for both p-AP and o-AP
over b-CD. This binding ratio is consistent with the result obtained
2.3. ITC and UV absorption analyses
in the exotherm fitting of ITC measurements, where the exo-
therms were best fitted using a complexation number of 2.
To obtain more information about the effect of b-CD on the re-
action regioselectivity, isothermal titration calorimetry (ITC) mea-
surements were performed to determine the changes of enthalpy
(DH), entropy (DS) and Gibbs free energy (DG) with titration of b-
CD by phenol, p-AP and o-AP, respectively and simultaneously, to
obtain the binding constants (K). The exotherm of complexation of
Interestingly, both the bindings did not result in
lmax shift, but
the binding of p-AP with -CD produced a 4.3-fold enhancement
b
in the maximum absorption at the ratio of 2:1; while a 1.4-fold
increase was found for o-AP, a consequence of distinct binding
strengths.30
phenol with b-CD was nearly a horizontal line, demonstrating a
rather weak interaction and giving no reasonable thermodynamic
parameters (data not shown). The exotherms of complexation of p-
2.4. Mechanism discussion
AP (Fig. 5A) and o-AP (Fig. 5B) with
“S” shapes, a consequence of the inclusion complex formation of p-
AP/ -CD and o-AP/ -CD. The exotherm was best fitted using a
complexation number of 2 for p-AP or o-AP over -CD, suggesting
the binding was a two-step process when excess of a guest was
used in the titration. Since the first step binding was dominant in
b-CD show typical exothermic
On the basis of the above results from 1H NMR and ITC, a
b
b
plausible mechanism can be outlined for the
regioselective dimethylaminomethylation of phenol as shown in
Scheme 2. First, phenol poses slantingly near the narrow rim of
CD cone via the hydrophobic and hydrogen bonding interactions to
b-CD mediated para-
b
b-

364
W.-H. Wu et al. / Tetrahedron 74 (2018) 360e365
Table 3
Thermodynamic parameters obtained from ITC measurement.
Inclusion complex
DH1 (KJ/mol)
DS1 (KJ/mol$K)
D
G1 (KJ/mol)
K1 ꢁ 107 (Mꢀ1
)
p-AP-
b
-CD
-CD
ꢀ25.68 0.08
ꢀ23.57 0.08
0.066
0.062
ꢀ45.35 0.08
ꢀ42.05 0.08
8.26 1.08
2.28 0.29
o-AP-
b
3. Conclusions
In summary, we investigated the effect of various cyclodextrins
on the three-component Mannich reaction of phenol with form-
aldehyde and dimethylamine, and demonstrated the use of
most effective to give p-AP as major product. By utilizing 1H NMR
and ITC measurements, we probed in-depth the role of -CD, and
proposed the inclusion conformations of phenol/ -CD, o-AP/ -CD
and p-AP/ -CD. Based on these conformations, the -CD mediated
b-CD is
b
Scheme 2. Plausible mechanism for the
aminomethylation.
b-CD mediated para-selective
b
b
b
b
para-regioselective dimethylaminomethylation of phenol could be
reasonably explained, that is, the preferential formation of the
complex p-AP/b-CD resulted in the change of the reaction pathway
to the desired direction. This product-inclusion process is com-
plementary to the known substrate-inclusion mechanism and may
have an important role in developing regioselective reactions by
supramolecular chemistry.
form an outer-sphere complex (Scheme 2A). The intermediate N,N-
dimethyliminium, which resides in the proton-deshielding region
outside
attack at the para-site of phenol to generate p-AP and subsequently
forming a inclusion complex with -CD (Scheme 2B), or via the
attack at the ortho-site of phenol to give o-AP and followed by the
formation of o-AP/ -CD (Scheme 2C). The positions and orienta-
tions of p-AP and o-AP in -CD are drawn accordingly in line with
b-CD, undergoes an electrophilic addition either via the
b
b
4. Experimental section
b
this process. In addition, there may exist somewhat hydrogen
bonding between the phenolic hydroxyl group of o-AP and the
4.1. Synthesis
primary hydroxyl group at the narrow rim of
b-CD, which is pre-
One solution was firstly prepared by dissolving 2 mmol phenol
and a certain amount of one CD (a-, b- or g-CD) in 15 ml water and
sumably formed upon interaction with -CD by disrupting the
b
intramolecular cyclic hydrogen bond between the hydroxyl and the
stirring for 30 min. Another solution, involving the formation of the
intermediate N,N-dimethyliminium (DMA), was prepared by mix-
ing 37% aqueous solution of formaldehyde (0.16 g, 2 mmol) and 33%
aqueous solution of dimethylamine (0.27 g, 2 mmol) in 10 ml water.
Then the DMA solution was added dropwise to the phenol solution
with magnetic stirring, and the reaction was monitored by TLC.
After stirring for 4 h, the reaction was complete, and the solution
was acidified to pH 2 with 2 M HCl and extracted with ethyl acetate
(3 ꢁ 10 ml). The organic extracts were combined, dried with
anhydrous sodium sulfate, concentrated in vacuum to give
unreacted phenol, whose amount was used to calculate the con-
version rate. The aqueous phase was basified to pH 9e10 with 2 M
NaOH and extracted with ethyl acetate (3 ꢁ 10 ml). The combined
organic phase was dried with anhydrous sodium sulfate, concen-
trated in vacuum to give a product mixture, which was used for the
1H NMR analysis of the composition of o-AP and p-AP. The mixture
was also subjected to purification with flash chromatography on
silica gel with EtOAc/petroleum ether (volumetric ratio 1:1) as the
first eluent to obtain o-AP, then with EtOAc/CH3OH (volumetric
ratio 1:1) as the second eluent to afford p-AP.
amino nitrogen in o-AP. This result can be inferred from the large
methylene chemical shift variation from
free o-AP binds with -CD, and the resulting value of
approximates the methylene chemical shift of 3.83 ppm of free p-
d4.06 to
d3.79 ppm when
b
d3.79 ppm
d
AP, where no intramolecular hydrogen bonding exists. Further-
more, this intermolecular hydrogen bonding may prevent o-AP
from entering into the cavity, resulting in a binding conformation
different to the deep complexation of p-AP/b-CD.
The proposed binding conformations are in agreement with the
1H NMR data to a large extent, except for that of the dimethyla-
minomethyl group. In principle, the more deep insertion of a gust
molecule inside the
b-CD cavity, the larger upfield proton Dds
should be resulted.31,32 The methyl Dds of p- and o-AP are consis-
tent with this expectation (Hp-f (ꢀ0.20 ppm) vs Hoꢀf (ꢀ0.15 ppm));
however, the methylene Dd
s
(Hp-e (ꢀ0.14 ppm) vs Hoꢀe
(ꢀ0.27 ppm)) deviate from this expectation largely. We could
attribute the deviation to a consequence of the transformation of
the intramolecular hydrogen bonding to the intermolecular one as
mentioned above, due to the former induces a larger downfield
shift to the methylene group of o-AP. Another remaining concern is
that the stoichiometry of b-CD in the optimal reaction conditions
needed to be or more than 1.5-fold over that of phenol to give the
high regioselectivity. This might be ascribed to fact that the binding
4.2. Simplified purification procedure
of
b
-CD with p-AP is a thermodynamic process, thus a higher
An alternative simplified purification procedure was developed
to obtain the pure compounds o-AP and p-AP without use of col-
umn chromatography. Briefly, after the reaction mentioned above,
the reaction mixture was first acidified to pH 2 with 2 M HCl and
extracted with ethyl acetate (3 ꢁ 10 ml) to remove the unreacted
phenol. Next, the aqueous phase was basified to about pH 7 using
2 M NaOH and extracted with ethyl acetate (3 ꢁ 10 ml) to give an
organic phase which contained the product o-AP. Finally, the
aqueous phase was further basified to about pH 8e9 and extracted
with ethyl acetate (3 ꢁ 10 ml) to give another organic phase con-
taining p-AP. Both the organic phases were dried with anhydrous
amount of
direction.
b-CD would favor the reaction toward the desired
It is worthy of mention that the plausible mechanism depicted
in Scheme 2, in which -CD includes more efficiently the product p-
AP to realize the para-selectivity, is significantly different to the
known role of -CD on supramolecular chemistry based syntheses.
For example, the para-photo-Reimer-Tiemann reaction of phenol to
yield 4-hydroxy benzaldehyde under light irradiation was reported
b
b
to involve the complexation of
b-CD with phenol and block the
ortho-site other than to affect the product.15

W.-H. Wu et al. / Tetrahedron 74 (2018) 360e365
365
sodium sulfate, concentrated in vacuum to afford the products o-AP
and p-AP, respectively, with adequate purity as determined by 1H
NMR.
National Natural Science Foundation of China (Nos. 21472064 and
21772060).
4.3. Scale-up synthesis of p-AP
Appendix A. Supplementary data
The optimal reaction conditions and the simplified procedure
were employed for the scale-up synthesis of p-AP with a 10-fold
amplification. Thus 2.4 grams of the pure product p-AP in 90%
yield was obtained after recrystallization from acetonitrile.
Supplementary data related to this article can be found at
https://doi.org/10.1016/j.tet.2017.12.007.
4.4. 1H NMR measurements
References
1H NMR spectra of pure o-AP, p-AP and the product mixture in
DMSO obtained in the synthesis procedure were recorded on Var-
ian NMR System 600 MHz and used to calculate the composition of
1. Guilera G, McGrady GS, Steed JW, Jones AL. Organomet. 2006;25:122e127.
2. Uhlenheuer DA, Young JF, Nguyen HD, Scheepstra M, Brunsveld L. Chem
Commun. 2011;47:6798e6900.
3. Fukuhara G, Mori T, Wada T, Inoue Y. Chem Commun. 2006;16:1712e1714.
4. Deraedt C, Astruc D. Coord Chem Rev. 2016;324:106e122.
5. Ramamurthy V, Sivaguru J. Chem Rev. 2016;116:9914e9993.
6. Tung CH, Wu LZ, Zhang LP, Chen B. Acc Chem Res. 2003;36:39e47.
7. Barr L, Dumanski PG, Easton CJ, et al. J Inclusion Phenom Macrocycl Chem.
2004;50:19e24.
o- and p-AP in the mixture. For the 1H NMR analyses of the
b
-CD
complexes, the interactive 1H NMR spectra of
b-CD with phenol,
DMA, p-AP, and o-AP were recorded using the samples prepared by
ultrasonically dissolving 0.1 mmol of the substances separately in
D2O containing 0.1 mmol b-CD.
8. Marchetti L, Levine M. ACS Catal. 2011;1:1090e1118.
9. Davis H, Phipps R. Chem Sci. 2017;8:864e877.
10. Breslow R, Kohn H, Siegel B. Tetrahedron Lett. 1976;17:1645e1646.
11. Hirai H, Shiraishi Y. React Funct Polym. 2007;67:1115e1128.
12. Clements AR, Pattabiraman M. J Photochem Photobiol A. 2015;297:1e7.
13. Komiyama M, Hirai H. J Am Chem Soc. 1983;105:2018e2021.
14. Komiyama M, Hirai H. J Am Chem Soc. 1984;106:174e178.
15. Ravichandran R. J Mol Catal A Chem. 1998;130:L205eL207.
16. Suresh P, Annalakshmi S, Pitchumani K. Tetrahedron. 2007;63:4959e4967.
17. Omura Y, Taruno Y, Irisa Y, Morimoto M, Saimoto H, Shigemasa Y. Tetrahedron
Lett. 2001;42:7273e7275.
4.5. ITC measurements
ITC experiments were carried out on a Microcal VPITC apparatus
at 25 ꢂC with the reference power set at 10 cal sꢀ1. The aqueous
solutions of phenol, o-AP or p-AP at 1 mM were used separately to
titrate 0.1 mM aqueous -CD in the sample cell under stirring
continuously at 310 rpm. For each titration, consecutive 15e20
injections with 15 L each were made at 5-min intervals and data
m
b
18. Matsumoto K, Joho K, Mimori S, Iida H, Hamana H, Kakehi A. Heterocycl.
2008;76:1061e1067.
m
were recorded. Data were analyzed using the ITC associated
software.
19. Bream RN, Ley SV, Procopiou PA. Org Lett. 2002;4:3793e3796.
20. Kim S, Hong SH. Adv Synth Catal. 2017;359:798e810.
ꢀ
21. Zielenkiewicz W, Terekhova IV, Kozbiał Ma, Poznanski J, Kumeev RS. J Phys Org
Chem. 2007;20:656e661.
22. Takahashi K. Chem Rev. 1998;98:2013e2034.
4.6. UVevis titration
23. Djedaïni F, Perly B. J Pharm Sci. 1991;80:1157e1161.
24. Yamamura H, Rekharsky M, Akasaki A, Kawai M, Inoue Y. J Phys Org Chem.
2001;14:416e424.
25. Rekharsky MV, Goldberg RN, Schwarz FP, et al. J Am Chem Soc. 1995;117:
8830e8840.
UVevis titration measurements were performed with a UV-
3600 spectrophotometer (Shimadzu, Japan) at room temperature.
The concentration of p-AP or o-AP was maintained constant at
26. Correia I, Bezzenine N, Ronzani N, Platzer N, Beloeil JC, Doan BT. J Phys Org
Chem. 2002;15:647e659.
27. Marconi G, Monti S, Mayer B, Koehler G. J Phys Chem. 1995;99:3943e3950.
28. Liu L, Guo QX. J Inclusion Phenom Macrocycl Chem. 2002;42:1e14.
29. Takahashi K, Yokomizo H, Ishiyama K, Kitsuta M, Ohashi M. J Inclusion Phenom
Macrocycl Chem. 2006;56:95e99.
50 mM while the molar ratio of b-CD to the guest was varied in the
range from 0 to 0.7 with 0.1 intervals. The spectra were acquired by
scanning from 200 to 500 nm.
Acknowledgments
30. Thordarson P. Chem Soc Rev. 2011;40:1305e1323.
31. Szejtli J. Chem Rev. 1998;98:1743e1754.
The authors gratefully acknowledge the financial supports from
32. Zhao YL, Benítez D, Yoon I, Stoddart JF. Chem Asian J. 2009;4:446e456.