The N-butyl amide group, CONHBu, has been found to be an effective promoter of the [1,2]-Wittig rearrangement of aryl benzyl ethers and thus allow the two-step synthesis of isomerically pure substituted diarylmethanols starting from simple hydroxybenzoic acid derivatives. The method is compatible with a wide range of functional groups including methyl, methoxy, and fluoro, although not with nitro and, unexpectedly, is applicable to meta as well as ortho and para isomeric series.
The N-butyl amide group, CONHBu, has been found to be an effective promoter of the [1,2]-Wittig rearrangement of aryl benzyl ethers and thus allow the two-step synthesis of isomerically pure substituted diarylmethanols starting from simple hydroxybenzoic acid derivatives. The method is compatible with a wide range of functional groups including methyl, methoxy, and fluoro, although not with nitro and, unexpectedly, is applicable to meta as well as ortho and para isomeric series.
Directed ortho-metalation is now well established
as a powerful method for regioselective functionalization of aromatic
compounds,[1] and subsequent reaction with
an electrophile such as an aromatic aldehyde allows facile construction
of ortho-substituted diarylmethanols (Scheme ). In contrast, methods for
the analogous meta- or para-functionalization
are nowhere near so well developed despite some recent progress.[2] In view of the low cost and ready availability
of the three isomeric hydroxybenzoic acids, an attractive alternative
strategy to access specifically substituted diarylmethanols would
be to transform the carboxylic acid into a suitable activating group, O-benzylate the phenolic OH, and then conduct a [1,2]-Wittig
rearrangement (Scheme ). Although this is a well-known aromatic rearrangement,[3] it has not been widely exploited in synthesis,
most likely due to the strongly basic conditions required, which limit
functional group compatibility, and there have been few recent examples
of its use.[4] Some highlights in recent
use of the Wittig rearrangement include rearrangement of benzyl butadienyl
ethers,[5] benzyl pyridyl ethers,[6] tandem anion translocation—Wittig rearrangement
and tandem Wittig rearrangement—aldol reaction,[7] as well as the study of systems where there is competition
between 1,2- and 2,3- or 1,2- and 1,4-Wittig rearrangements.[8] The Wittig rearrangement has also been used to
access a range of chiral binaphthyl ligands that have been applied
in asymmetric catalysis.[9] The rearrangement
has been carried out in an enantioselective way by adding a chiral
bis(oxazoline) catalyst,[10] and diastereoselective
Wittig rearrangements have been reported directed by adjacent carbohydrate
and α-alkoxyamide functions.[11] Finally,
it is also possible to suppress the Wittig rearrangement if desired
in order to make use of the unrearranged α-lithiobenzyloxy group
in the synthesis of benzofurans and other heterocycles.[12] In this paper, we describe the discovery and
development of the facile Wittig rearrangement of isomeric benzyloxy-N-butylbenzamides to furnish the corresponding diarylmethanols
in the ortho-, meta-, and para-series.
Scheme 1
Alternative Approaches to Diarylmethanols
Results and Discussion
Our entry
into this area came from a serendipitous discovery during
attempted ring bromination of the 2,4-bis(benzyloxy)phenyloxazoline 1. Treatment with n-butyllithium followed
by bromine gave not the expected product 2 but instead
a product identified by spectroscopic methods and X-ray diffraction
(see Supporting Information) as 4, presumably formed by Wittig rearrangement to afford the intermediate 3, which was then oxidized by bromine[13] to the ketone (Scheme ).
Scheme 2
Unexpected Reaction Leading to 4
We then examined the reactivity of the simpler 2-benzyloxyphenyl
compound 5 and found that, depending upon the reaction
conditions, varying mixtures of the Wittig rearrangement product 6, the 3-aminobenzofuran 7 resulting from intramolecular
nucleophilic ring opening of the oxazoline by the benzyl anion, and
the O-dealkylation product 8 were formed
(Scheme ). As we have
already reported elsewhere,[14] the process
could be optimized toward the formation of 7 using n-butyllithium/potassium tert-butoxide.
However, optimized conditions for the formation of 6,
namely 2.2 equiv n-BuLi in THF at rt for 1 h, resulted
in an isolated yield after chromatographic purification of just 29%.
Under the same conditions, the para-isomer 9 gave the rearrangement product 10 but in only
11% isolated yield, and the meta-isomer 11 was recovered unchanged. It was clear that the 4,4-dimethyloxazoline
group was not an optimal promoter of the Wittig rearrangement.
Scheme 3
Behavior of Mono(benzyloxy)phenyloxazolines
We next investigated carbamates and tertiary amides, both classic ortho-directing groups,[1] as promoters
of Wittig rearrangement, but this was uniformly unsuccessful. The
diethylcarbamate 12 suffered a nucleophilic attack at
carbonyl to give 13 and 14 in low yield
while the phenylcarbamate 15 was recovered unchanged
(Scheme ). The N,N-diethyl amide 16 reacted at the carbonyl
group to give the ketone 17. In a previous case where
an undesired attack of butyllithium at a diethyl amide was encountered,[15] this could be suppressed by changing to the N,N-diisopropyl amide but the reaction of 18 took a different course, giving 2-hydroxybenzil 19 in
THF, but the 3-aminobenzofuran 20 in toluene. There are
only a few other synthetic routes to substituted 3-aminobenzofurans.
The formation of both these products involves an initial intramolecular
nucleophilic attack of benzyloxy anion on the amide carbonyl with
hydrolysis and dehydration giving 20, while loss of diisopropylamine,
hydrolysis, and oxidation affords 19. The closely analogous
formation of 19 by base treatment and oxidation of methyl
2-benzyloxybenzoate has already been reported.[16]
Scheme 4
Reaction of Carbamates and Tertiary Benzamides
Success was finally achieved by moving to the
secondary N-butylbenzamides and reaction of 21a with
3.3 equiv n-butyllithium in THF at rt for 2 h followed
by workup gave an essentially quantitative yield of the diarylmethanol 22a (Scheme ). This product was, however, found to be unstable and slowly cyclized
over a period of weeks to give mainly 3-phenylphthalide 23a (80%) accompanied by a low yield of anthraquinone 24a resulting from an alternative mode of cyclization and subsequent
oxidation. Alternatively, treating the crude product 22a with p-toluenesulfonic acid in boiling toluene
for 1 h,[17] followed by aqueous workup and
chromatographic purification led directly to 23a in 90%
isolated yield. The need for 3 equiv of n-BuLi is, we believe, due
to the first two equivalents reacting to deprotonate the NH and then
bring about an amide-directed ortho-metalation. Only
with the third equivalent of base is the benzyl group deprotonated.
Use of less n-BuLi resulted in progressively lower yields and recovery
of unreacted starting material.[18] With
these optimized conditions in hand, the scope was now explored, and
analogues 21b–w were prepared in good yield by O-alkylation of N-butylsalicylamide 25 with benzylic and other halides and potassium carbonate
in DMF (Scheme ).
Scheme 5
Wittig Rearrangement of 2-Benzyloxy-N-butylbenzamide
Scheme 6
Preparation of ortho-substituted N-butylbenzamides
These were now subjected to the Wittig rearrangement conditions
used for 21a and a varied pattern of reactivity emerged
(Scheme ). The systems
containing methyl, methoxy, and fluoro substituents all reacted to
give the corresponding secondary alcohols 22b–j, which were fully characterized (see Supporting Information) but cyclized upon storage or p-toluenesulfonic acid treatment to give the corresponding phthalides 23. In the cases of the fluorobenzyl compounds 21h and 21j, the corresponding anthraquinone products 24h and 24j were also isolated in low yield.
The thienyl compound 21q, the α-methylbenzyl compound 21s, and the prenyl compound 21u also underwent
rearrangement. However, the remaining compounds bearing more electron-withdrawing
substituents did not, and compounds 21k–p, r, v, and w either underwent decomposition
or were recovered unchanged. While all the products shown in Scheme were obtained by
spontaneous cyclization, those marked *, in addition to 23a, were also prepared by the p-toluenesulfonic acid
method.
Scheme 7
Products from Wittig Rearrangement of ortho-Substituted
Benzamides
The case of the allyl compound 21t was particularly
interesting. It afforded an inseparable mixture of the expected rearrangement
product 22t and an isomer, which proved to be the 3-ethyl-3-hydroxyisoindolinone 27. Over a period of months, the mixture converted entirely
into the latter, whose structure was confirmed by X-ray diffraction
(Scheme ).
Scheme 8
Isomerization
of 22t
We believe this to
involve double-bond migration in 22t to give the enol,
which tautomerizes to ketone 26,
which then undergoes ring closure. Such isomerization of allyl carbinols
to ethyl ketones occurs under a range of basic conditions.[19]Attention was now turned to the isomeric para-system,
and a range of substrates 29a–p were prepared
in good yield by O-alkylation of N-butyl-p-hydroxybenzamide 28 (Scheme ).
Scheme 9
Preparation of para-substituted N-butylbenzamides
These were subjected to the normal rearrangement
conditions, and
a similar pattern emerged as for the ortho series.
The unsubstituted system as well as those with methyl, methoxy, and
fluoro substituents rearranged to give the diarylmethanols 30a–h mostly in good yield (Scheme ). The α-methylbenzyl compound 29o and the prenyl compound 29p also reacted to give 30o and 30p although with a low yield in the
latter case. In the case of 30c, the molecular structure
was confirmed by X-ray diffraction and the crystal structure featured
each molecule involved in two donor and two acceptor interactions,
with head-to-tail hydrogen-bonded dimers linked by C=O...H–O
interactions, which were then further linked by N–H...O(H)–C
interactions with adjacent molecules (see Supporting Information). Again the three isomeric nitro compounds 29j–l as well as the pentafluorophenyl compound 29m, the diphenylmethyl compound 29n and the
methoxynaphthyl system 29i either decomposed or were
recovered unchanged.
Scheme 10
Products from Wittig Rearrangement of para-Substituted
Benzamides
Having established the viability
of the rearrangement for both ortho and para isomeric systems, it was
of interest to compare the relative ease of the two processes, and
for this, the 2,4-bis(benzyloxy)benzamide 31 was prepared.
When this was treated with 3.3 equiv n-BuLi, the product was mainly 32 resulting from rearrangement of the ortho group with just a trace of the isomer 33 from the reaction
of the para group (Scheme ). This appears to be the first example
of an amide directing lithiation onto an ortho-alkoxy
group. Storage of compound 32 over a period of months
resulted in spontaneous cyclization to give phthalide 34 in good yield. On the other hand, treatment of 31 with
4.4 equiv of n-BuLi resulted in rearrangement of both groups to give
the diol 35 as a 1:1 mixture of diastereomers.
Scheme 11
Reactivity
of bis(benzyloxy) Compound 31
Based on the earlier finding that the meta-benzyloxy
oxazoline 11 did not react, in contrast to ortho and para-isomers 5 and 9, and assuming that the activating effect of the amide group in the
Wittig rearrangement would be via a spiro anionic intermediate (see
below, Scheme ),
we expected that meta-benzyloxy N-butylbenzamides 37 would not react. However, to our
surprise, the unsubstituted compound 37a did rearrange
under the normal conditions to afford 38a, albeit in
a somewhat lower yield than for ortho or para-isomers. The structure of 38a was confirmed
by X-ray diffraction and again the crystal structure featured each
molecule involved in two donor and two acceptor interactions, but
in contrast to 30c, this involved head-to-tail hydrogen-bonded
dimers linked by N–H...O(H)–C interactions, which were
then further linked by C=O...H–O interactions with adjacent
molecules (see Supporting Information).
Based on this result, a range of substituted examples 37b–n were prepared in good-to-moderate yield by O-alkylation
of 36 (Scheme ). When these were subjected to the standard rearrangement
conditions, a somewhat more restricted pattern of reactivity emerged
with successful rearrangement only being observed for alkyl, methoxy,
and fluoro substituents as well as the thienyl compound 37m and, in all cases, the isolated yields were lower than for the isomeric
systems (Scheme ).
Scheme 17
Opportunities for Charge Stabilization
in the Three Isomeric Systems
Scheme 12
Preparation of meta-substituted N-butylbenzamides
Scheme 13
Products from Wittig
Rearrangement of meta-Substituted
Benzamides
So far substituent effects
have only been examined in the benzyl
as opposed to the aryl ring. To further examine the scope, we returned
to the ortho system and investigated the effect of
substituents on the other aryl ring. Starting from the 5-nitrosalicylamide 39, a range of three O-benzyl derivatives 40a–c were prepared, while the corresponding 5-dimethylamino
compound 41 led to derivatives 42a–c. However, when these six compounds were subjected to the usual rearrangement
conditions, only in the cases of 42a and 42c were the corresponding rearrangement products 43a and 43c obtained (Scheme ). It is clear that the presence of a nitro group on either
ring is sufficient to prevent the reaction. Since this is most likely
due to incompatibility of the nitro function with the reaction conditions,
additional evidence was sought from isomeric compounds with methoxy
and fluoro substituents, both of which were compatible with the reaction
conditions. Thus the 6-fluoro compound 44 in which ortho-metalation is impossible was treated with 2.2 equiv
n-BuLi to give the expected rearrangement product 45,
isolated after p-toluenesulfonic acid-mediated cyclization
as the 7-fluoro-3-phenylphthalide (27%), together with the n-butyl compound 47 resulting from nucleophilic
aromatic substitution in low yield (Scheme ). Methoxy substituents in either the 4-
or 5-position were also compatible with the rearrangement and 48 reacted to form 49, isolated as the phthalide 50 while 51 reacted via 52 to give
the phthalide product 53 in moderate yield. In contrast
to this, the 5-fluoro compound 54 (Scheme ) underwent decomposition
under the normal rearrangement conditions, perhaps due to intervention
of an aryne process, while the 3,5-dimethyl-4-benzyloxy compound 55 in which the supposed spiro anion intermediate is sterically
disfavored, was recovered largely unchanged.
Scheme 14
Nitro and Dimethylamino
Substituted Systems
Scheme 15
Rearrangement with
4-, 5-, or 6-Substituents
Scheme 16
Further Mechanistic Probes Examined
To further probe the mechanism, the 4-trimethylsilyl compound 56 was prepared and was found to undergo rearrangement readily
with n-BuLi to afford the alcohol 57. Finally, compounds 58 and 59 in which the benzyloxy group of 21a and 29a is replaced by the isomeric phenoxymethyl
group were prepared, and these were also found to readily undergo
the rearrangement under the normal conditions, giving, respectively, 22a and 30a.Although there have been a
good number of mechanistic studies on
the [1,2]-Wittig rearrangement, some early suggestions involving the
intermediacy of arynes and carbonyl compounds were later disproven.[20] The most recent and detailed mechanistic study
on aryl benzyl ethers employing both experimental and theoretical
methods,[21] quite clearly points to two
major mechanistic possibilities: an anionic mechanism via a spiro-epoxide
intermediate that is the normal route for neutral and electron-poor
aryl rings, and a radical dissociation/recombination route that is
more likely to be important for electron-rich aryl systems. In our
system, we envisage initial amide NH deprotonation and ortho-metalation in each case before the third equivalent of n-BuLi deprotonates
the benzyl group. The resulting intermediates 60, 62, and 64 can each cyclize to the spiro epoxides
implicated in the Wittig rearrangement (Scheme ) but, while 61 and 65 are be stabilized by the negative charge
being on nitrogen, this is not possible for 63 derived
from the meta compound thus explaining the lower
yields obtained in that case.The recent mechanistic study[21] was focused
on substituent effects on the aryl ring rather than the benzyl ring,
and substituent effects on the benzyl ring do not seem to have been
examined in detail until now. Overall our results on the three isomeric
benzyloxybenzamide systems show that a single N-butylbenzamide
group on the aryl ring is sufficient to facilitate the rearrangement
of ortho and para benzyloxy systems,
presumably via the anionic spiro-epoxide intermediates. The meta system is slightly less prone to rearrangement presumably
reflecting the reduced capacity for delocalization of the negative
charge to a meta disposed electron-withdrawing amide.
In all three series, the reaction proceeds with a wide range of both
electron-donating and electron-withdrawing benzyl substituents. Only
nitro compounds were uniformly unsuccessful due to the incompatibility
of that group with butyllithium. The fact that the silyl compound 56 rearranges much more rapidly than benzyl phenyl ether is
consistent with the anionic mechanism in that case where the stabilizing
effect of silicon upon the α-anion is key. For the isomeric
phenoxymethyl compounds 58 and 59 where
the spiro-epoxide intermediate cannot be stabilized, the amide group
nevertheless promotes the reaction perhaps by favoring the benzylic
deprotonation, but the rearrangement must necessarily proceed by the
radical route in these cases.The fact that 48 gives
a substantially lower yield
than the isomer 51 implies that the rearrangement is
discouraged by the presence of a second inductively electron-withdrawing
(but mesomerically electron-donating) group in the para position to the rearranging group. The occurrence of the rearrangement
for 42a and 42c, which also have an inductively
electron-withdrawing group para to the reaction site,
can perhaps be taken to indicate that the electron-donating mesomeric
effect outweighs the inductive effect in these cases.Finally,
we examined briefly whether the process could be extended
from the ethers to the corresponding sulfides or amines since both
thia- and aza-[1,2]-Wittig rearrangements are known.[22,23] The 2-benzylthio-N-butylbenzamide 66 was readily prepared, but upon treatment with butyllithium, it underwent
dehydrative cyclization to afford the 3-aminobenzothiophene 67 in moderate yield, in a reaction analogous to the formation
of 20 from 18 (Scheme ). In an attempt to suppress this process,
the more bulky tert-butylbenzamide 68 was prepared, but
it was recovered unreacted from BuLi treatment as was the 2-(benzylmethylamino)
analogue 69.
Scheme 18
Attempted Extension to thia- and aza-Analogues
Experimental Section
1H and 13C NMR spectra were recorded in CDCl3 unless otherwise stated with internal TMS as reference. IR
spectra were recorded using the ATR technique. HRMS measurements were
made either using ESI with TOF analyzer or NSI with an ion trap analyzer.The following procedures are illustrative; full experimental procedures
and characterization data are given in the Supporting Information.
2-(Benzyloxy)-N-butylbenzamide 21a
A solution of 2-(benzyloxy)benzoyl chloride[13] (10.85 g, 44.0 mmol) in toluene (60 mL) was
added dropwise to a stirred 0 °C solution of n-butylamine (12.1 mL, 8.95 g, 0.122 mol) in toluene (60 mL). Once
the addition was complete, the reaction mixture was allowed to warm
to rt for 1 h before being poured into water and washed with 2 M NaOH
and brine. The organic layer was dried and evaporated to afford, after
recrystallization (EtOAc/hexane), 21a (9.34 g, 75%) as
colorless crystals, mp 52–54 °C; IR 3380, 1648, 1599,
1558, 1292, 1238, 1164, 1101, 1005, 865, 752, 700 cm–1; 1H NMR (500 MHz) δ 8.25 (dd, J = 7.8, 1.8 Hz,1H, ArH), 7.88 (br s, 1H, NH), 7.47–7.38 (m,
6H, ArH and Ph), 7.12–7.08 (m, 1H, ArH), 7.06 (d, J = 8.5 Hz, 1H, ArH), 5.15 (s, 2H, OCH2), 3.34 (td, J = 7.0, 5.5 Hz, 2H, NCH2), 1.35–1.29
(m, 2H, NCH2CH2), 1.19–1.11
(m, 2H, CH2CH3), 0.80 (t, J = 7.3 Hz, 3 H, CH3); 13C NMR (75
MHz) δ 164.9(C), 156.8(C), 135.5(C), 132.5 (CH), 132.4 (CH),
128.9 (2CH), 128.8 (CH), 128.2 (2CH), 122.0(C), 121.6 (CH), 112.4
(CH), 71.4 (OCH2), 39.4 (NCH2), 31.2 (CH2), 20.0 (CH2), and 13.7 (CH3); HRMS
(ESI+) m/z [M + Na+] calcd for C18H21NaNO2 306.1465,
found 306.1455.
N-Butyl-2-(hydroxy(phenyl)methyl)benzamide 22a, Anthraquinone 24a and 3-Phenylphthalide 23a
Under a nitrogen atmosphere, n-butyllithium (2.6 mL, 6.50 mmol) was added dropwise to a stirred
solution of 2-(benzyloxy)-N-butylbenzamide 21a (0.5678 g, 2.00 mmol) in dry THF (20 mL). After stirring
at rt for 2 h, the reaction mixture was quenched by addition of sat.
aq. NH4Cl and extracted with Et2O (×3).
The combined organic layers were washed with 2 M NaOH and water before
being dried and evaporated to give 22a as a pale yellow
oil: IR 3296, 3064, 2931, 1635, 1540, 1450, 1303, 1228, 1104, 1024,
757, 699 cm–1; 1H NMR (500 MHz) δ
7.38 (t, J = 7.3 Hz, 2H, ArH), 7.29–7.17 (m,
7H, ArH and Ph), 6.31 (t, J = 5.3 Hz, 1H, NH), 5.79
(s, 1H, CHOH), 3.21–3.14 (m, 1H, NCH2), 3.12–3.05 (m, 1H, NCH), 1.29–1.23
(m, 2H, NCH2CH2), 1.22–1.15
(m, 2H, CH2CH3), 0.85 (t, J = 7.0 Hz, 3H, CH3); 13C NMR (125
MHz) δ 170.8(C), 143.1(C), 142.7(C), 135.8(C), 130.6 (CH), 129.9
(CH), 127.74 (CH), 127.71 (2CH), 127.69 (CH), 126.7 (CH), 126.2 (2CH),
74.9 (CHOH), 39.7 (NCH2), 31.1 (CH2), 19.9 (CH2), 13.6 (CH3); HRMS (ESI+) m/z [M + Na+] calcd for C18H21NaNO2 306.1465, found 306.1456.On
standing at rt in EtOAc solution for 2–3 months, an intramolecular
cyclization occurred to give, after purification by column chromatography
(SiO2, Et2O/hexane 2:3), at Rf 0.80, 24a (17.5 mg, 4%) as yellow needles,
mp 275–279 °C (lit.[24] 275 °C); 1H NMR (500 MHz) δ 8.34–8.30 (m, 4 H, ArH), 7.83–7.79
(m, 4 H, ArH). The 1H NMR spectral data were in accordance
with those previously reported.[25]This was followed by a second fraction to give, at Rf 0.55, 23a (0.3350 g, 80%) as tan-colored
crystals, mp 113–116 °C (lit.[26] 115.5 °C); 1H NMR (500 MHz) δ 7.96 (d, J = 7.5 Hz, 1H, ArH), 7.65 (td, J = 7.5,
1.0 Hz, 1H, ArH), 7.55 (t, J = 7.5 Hz, 1H, ArH),
7.39–7.36 (m, 3H, ArH), 7.33 (dd, J = 7.8,
0.8 Hz, 1H, ArH), 7.29–7.26 (m, 2H, ArH), 6.41 (s, 1H, CHPh).
The 1H NMR spectral data were in accordance with those
previously reported.[27]Alternatively,
the following literature procedure[28] may
be employed: A mixture of N-butyl-2-(hydroxy(phenyl)methyl)benzamide 22a (prepared as above from 1.14 g 21a, assuming
4.02 mmol) and p-toluenesulfonic acid monohydrate
(1.55 g, 8.15 mmol) in toluene (80 mL) was heated at reflux for 1
h. After cooling to rt, the reaction mixture was washed with water
(50 mL), 2 M NaOH (50 mL), and brine (50 mL) before being dried and
evaporated. The crude residue was purified by column chromatography
(SiO2, gradient elution, Et2O/hexane 1:4 to
Et2O) to give 23a (0.76 g, 90%) as tan-colored
crystals.
N-Butyl-4-hydroxybenzamide 28
A mixture of methyl 4-hydroxybenzoate (30.43 g, 0.20 mol) and n-butylamine (100 mL, 74.00 g, 1.01 mol) was heated at reflux
for 4 d before being concentrated in vacuo. The residue
was acidified to pH 1 by the addition of 2 M HCl and extracted with
Et2O (3 × 100 mL). The combined organic layers were
washed with water (100 mL) before being dried and evaporated. The
crude residue was recrystallized (EtOAc/PhMe) to give 28 (28.01 g, 72%) as colorless crystals, mp 118–120 °C
(lit.[29] 118.5–119.5 °C); 1H NMR (500 MHz) δ 7.95 (br s, 1H, OH), 7.62 (d, J = 8.8 Hz, 2H, 2,6-H), 6.86 (d, J = 8.8
Hz, 2H, 3,5-H), 6.17 (t, J = 5.5 Hz, 1H, NH), 3.44
(td, J = 7.0, 5.5 Hz, 2H, NCH2), 1.62–1.56
(m, 2H, NCH2CH2), 1.44–1.36
(m, 2H, CH2CH3), 0.94 (t, J = 7.3 Hz, 3H, CH3). The 1H NMR spectral
data were in accordance with those previously reported.[30]
4-(Benzyloxy)-N-butylbenzamide 29a
N-Butyl-4-hydroxybenzamide 28 (3.87 g, 20.0 mmol) was added to a stirred suspension of
sodium
hydride (60% in mineral oil, prewashed with hexane, 0.82 g, 20.5 mmol)
in DMF (20 mL), and the mixture was stirred at rt for 15 min before
benzyl bromide (2.4 mL, 3.45 g, 20.2 mmol) was added. After stirring
for 18 h at rt, the reaction mixture was poured into water and extracted
with CH2Cl2 followed by Et2O (×3).
The combined organic layers were washed with brine (×5) and 2
M NaOH before being dried and evaporated. Recrystallization of the
residue (EtOAc/hexane) gave 29a (4.68 g, 82%) as colorless
crystals, mp 126–128 °C; (lit.[31] 119.1–119.7 °C); 1H NMR (500 MHz) δ
7.72 (d, J = 8.8 Hz, 2H, 2,6-H), 7.44–7.38
(m, 4H, Ph), 7.35–7.32 (m, 1H, Ph), 6.99 (d, J = 8.8 Hz, 2H, 3,5-H), 6.03 (t, J = 5.5 Hz, 1H,
NH), 5.11 (s, 2H, OCH2), 3.44 (td, J =
7.3, 5.5 Hz, 2H, NCH2), 1.62–1.56 (m, 2H, NCH2CH2), 1.45–1.37 (m, 2H,
CH2CH3), and 0.95 (t, J = 7.3 Hz, 3H, CH3). The 1H NMR spectral
data were in accordance with those previously reported.[31]
N-Butyl-4-(hydroxy(phenyl)methyl)benzamide 30a
Under a nitrogen atmosphere, n-butyllithium (2.5 M in hexane, 6.6 mL, 16.5 mmol) was added dropwise
to a stirred solution of 4-(benzyloxy)-N-butylbenzamide 29a (1.41 g, 4.98 mmol) in dry THF (50 mL). After stirring
at rt for 2 h, the reaction mixture was quenched by addition of sat.
aq. NH4Cl and extracted with Et2O (×3).
The combined organic layers were washed with 2 M NaOH and water before
being dried and evaporated. Purification of the residue by column
chromatography (SiO2, gradient elution, Et2O/hexane
7:3 to Et2O) and subsequent recrystallization (EtOAc/hexane)
gave 30a (1.11 g, 79%) as colorless crystals, mp 113–114
°C; IR 3432, 3337, 2953, 2926, 1616, 1542, 1448, 1303, 1228,
1045, 736, 694 cm–1; 1H NMR (400 MHz)
δ 7.67 (d, J = 8.2 Hz, 2H, ArH), 7.41 (d, J = 8.2 Hz, 2H, ArH), 7.36–7.30 (m, 4H, Ph), 7.28–7.24
(m, 1H, Ph), 6.14 (t, J = 5.6 Hz, 1H, NH), 5.85 (s,
1H, CHOH), 3.41 (td, J = 7.2, 5.6
Hz, 2H, NCH2), 2.76 (d, J = 3.2 Hz, 1H,
OH), 1.61–1.53 (m, 2H, NCH2CH2), 1.43–1.34 (m, 2H, CH2CH3), 0.94 (t, J = 7.4 Hz, 3H, CH3); 13C NMR (125 MHz) δ 167.3(C), 147.1(C),
143.4(C), 133.8(C), 128.6 (2CH), 127.8 (CH), 126.9 (2CH), 126.6 (2CH),
126.5 (2CH), 75.7 (CHOH), 39.8 (NCH2), 31.7 (CH2), 20.1 (CH2), 13.8 (CH3); HRMS (NSI+) m/z [M + H+] calcd
for C18H22NO2 284.1645, found 284.1644.
3-(Benzyloxy)-N-butylbenzamide 37a
To
a stirred solution of 3-(benzyloxy)benzoyl chloride[32] (6.82 g, 27.6 mmol) in CH2Cl2 (100
mL) at 0 °C, Et3N (3.85 mL, 27.6 mmol) was added dropwise.
The solution was stirred for 5 min, and then n-butylamine
(2.73 mL, 27.6 mmol) was added dropwise and the mixture stirred at
rt for 18 h. The reaction mixture was poured into H2O,
extracted (×3) with CH2Cl2, and the combined
organic fractions were dried over MgSO4 and concentrated
to afford, after recrystallization (EtOH) 37a (5.80 g,
74%) as off-white crystals, mp 83–86 °C; IR 3298, 3229,
2961, 1626, 1603, 1580,1553, 1016, 698 cm–1; 1H NMR (400 MHz) δ 7.47–7.38 (m, 5H, ArH), 7.38–7.32
(m, 2H, ArH), 7.32–7.28 (m, 1H, ArH), 7.09 (ddd, J = 7.8, 2.6, 1.5 Hz, 1H, ArH), 6.07 (br s, 1H, NH), 5.11 (s, 2H,
OCH2), 3.45 (td, J = 7.1, 5.7 Hz, 2H,
NHCH2), 1.65–1.53 (m, 2H, NHCH2CH2), 1.51–1.30 (m, 2H,
CH2CH3), 0.96 (t, J = 7.3 Hz, 3H, CH2CH3); 13C NMR (100 MHz) δ 167.2(C), 158.9(C), 136.5(C), 136.4(C),
129.6 (CH), 128.6 (CH), 128.1 (CH), 127.5 (CH), 118.8 (CH), 118.1
(CH), 113.3 (CH), 70.1 (OCH2), 39.8 (NHCH2),
31.7 (CH2), 20.1 (CH2), 13.8 (CH3); HRMS (ESI+) m/z [M
+ H+] calcd for C18H22NO2 284.1651, found 284.1636.
N-Butyl-3-(hydroxy(phenyl)methyl)benzamide 38a
Under a nitrogen atmosphere, n-butyllithium (2.5 M in hexane, 2.64 mL, 6.60 mmol) was added dropwise
to a stirred solution of 3-(benzyloxy)-N-butylbenzamide 37a (567 mg, 2.0 mmol) in dry THF (20 mL). After stirring
at rt for 2 h, the reaction mixture was quenched by addition of sat.
aq. NH4Cl and extracted with Et2O (×3).
The combined organic layers were dried and evaporated. Purification
of the residue by column chromatography (gradient elution, Et2O/hexane 1:1 to Et2O/hexane 7:3) gave 38a (298 mg, 53%) as colorless crystals, mp 97–100 °C; IR
3298, 3229, 2961, 2932, 2857, 1626, 1553, 1418, 1327, 1016, 746, 698
cm–1; 1H NMR (400 MHz) δ 7.75 (t, J = 1.8 Hz, 1H, ArH), 7.57 (dt, J = 7.7,
1.5 Hz, 1H, ArH), 7.42–7.35 (m, 1H, ArH), 7.31–7.24
(m, 5H, ArH), 7.24–7.21 (m, 1H, ArH) 6.43 (t, J = 5.7 Hz, 1H, NH), 5.74 (s, 1H, CHOH), 3.71 (s,
1H, CHOH), 3.31 (td, J = 7.2, 5.7
Hz, 2H, NHCH2), 1.54–1.45 (m, 2H,
NHCH2CH2), 1.39–1.27
(m, 2H, CH2CH3), 0.90 (t, J = 7.3 Hz, 3H, CH2CH3); 13C NMR (100 MHz) δ 167.6(C), 144.5(C), 143.5(C),
134.7(C), 129.5 (CH), 128.5 (CH), 128.4 (2CH), 127.5 (CH), 126.5 (2CH),
125.9 (CH), 124.7 (CH), 75.6 (CHOH), 39.8 (NHCH2), 31.5
(CH2), 20.1 (CH2), 13.7 (CH3); HRMS
(ESI+) m/z [M + H+] calcd for C18H22NO2 284.1651,
found 284.1637.
Authors: Samuel Martinez-Erro; Amparo Sanz-Marco; Antonio Bermejo Gómez; Ana Vázquez-Romero; Mårten S G Ahlquist; Belén Martín-Matute Journal: J Am Chem Soc Date: 2016-09-30 Impact factor: 15.419
Authors: Luo-Yan Liu; Jennifer X Qiao; Kap-Sun Yeung; William R Ewing; Jin-Quan Yu Journal: Angew Chem Int Ed Engl Date: 2020-07-09 Impact factor: 15.336