To catalyze or not to catalyze? Insight into direct amide bond formation from amines and carboxylic acids under thermal and catalyzed conditions

Article abstract of DOI:10.1002/adsc.200606018

Kinetic studies show that the direct formation of amides from amines and carboxylic acids without catalyst does occur under relatively low temperature conditions, but is highly substrate dependent. Boric and boronic acid-based catalysts improve the reacti

Full text of DOI:10.1002/adsc.200606018

COMMUNICATIONS
DOI: 10.1002/adsc.200606018
To Catalyze or not to Catalyze? Insight into Direct Amide Bond
Formation from Amines and Carboxylic Acids under Thermal
and Catalyzed Conditions
Kenny Arnold,a Bryan Davies,b Richard L. Giles,a Christophe Grosjean,a
Gillian E. Smith,b Andrew Whitinga,*
a
Department of Chemistry, Durham University, Science Laboratories, South Road, Durham DH1 3LE, United Kingdom
Fax: (þ44)-191-386 1127, e-mail: andy.whiting@durham.ac.uk
b
GlaxoSmithKline Research & Development Ltd., Gunnels Wood Road, Stevenage, Hertfordshire, SG1 2NY,
United Kingdom
Received: January 13, 2006; Accepted: March 9, 2006
Supporting Information for this article is available on the WWW under http://asc.wiley-vch.de/home/.
Abstract: Kinetic studies show that the direct forma-
tion of amides from amines and carboxylic acids
without catalyst does occur under relatively low tem-
perature conditions, but is highly substrate depen-
dent. Boric and boronic acid-based catalysts improve
the reaction, especially for less reactive acids, and in-
itial results indicate that bifunctional catalysts show
even greater potential.
some reactions carried out in the absence of a catalyst
gave surprising amounts of amide product under mild
conditions. We therefore undertook an investigation
into the kinetics of both thermal and boric or boronic
acid-catalyzed reactions, in order to compare these re-
sults with those derived from bifunctional systems. In
this communication we report our unexpected findings.
A parallel series of experiments was carried out as de-
tailed by Scheme 1, in which a carboxylic acid 1 (either
1a or b) was reacted with an amine 2 (either 2a or b) in
refluxing toluene in the presence of activated 3 Š mo-
lecular sieves (Soxhlet). The thermal reaction was com-
pared with the addition of 1 mol % of each of the cata-
lysts shown in Scheme 1, with direct monitoring from
the reaction being carried out at 2 hourly intervals
over 22 hours by HPLC. The results are shown in
Keywords: amines; bifunctional catalysis; boronate
catalysis; carboxylic acids; direct amide formation;
green catalysis
Amide bonds are generally formed from carboxylic acid Table 1.
derivatives such as acyl halides or anhydrides, including
Reaction 1 (1aþ2a, Table 1) shows the highest level
mixed anhydrides, or active esters.[1] Equivalent reactive of uncatalyzed, thermal reaction (blue line), resulting
entities can also be generated in situ using diimide deriv- in a surprisingly high ca. 60% yield of 3a after 22 hours.
atives.[1] Such reagents, however, represent poor atom In comparison, and even more surprisingly, all the cata-
economy,[2] as does the addition of activating agents or lyzed reactions are only incrementally better than the
acyl transfer agents, etc.[3] The most desirable method thermalreaction, with boric acid, 4 and 5 being essential-
for preparing amides would be a direct condensation be- ly identical. These results immediately raised the ques-
tween a carboxylic acid and amine, which is generally tion as to whether much of the catalytic activity ob-
understood to be impossible due to the formation of served, albeit only 20% of the yield (i.e., over the ther-
an unreactive carboxylate-ammonium salt. Although mal reaction), was actually due to the formation of boric
the direct formation of amide bonds has been known acid under these reaction conditions. Boronic acids are
since 1858,[4] this process has found little synthetic utility known to undergo facile proto-deboronation in the
with a few exceptions.[5] However, the use of boron re- presence of carboxylic acids, especially alkyl carboxylic
agents to assist amide formation directly from carboxyl- acids which is a faster process than for more electron-de-
ic acids and amines[6,7] has been punctuated by the re- ficient arylboronic acids.[10] However, boronic acids 4, 5
port[8] that boric acid is similarly effective.
and 6a do not seem to show substantial proto-deborona-
Our recent interests in the development of bifunction- tion under the reaction conditions reported herein. In
al catalysts[9] led us to examine amine-boronic acids as comparison, reactions with 1b showed a much less im-
potential amide bond forming catalysts; however, pressive thermal amide formation (see Reactions 2
Adv. Synth. Catal. 2006, 348, 813 – 820
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
813

Kenny Arnold et al.
COMMUNICATIONS
Scheme 1. General equation for direct amide formation and catalyst structures.
and 4, Table 1). For the reaction of 1b with 2a (Reaction Table 2). The similarity of the different catalysts for Re-
2, Table 1), the catalyzed reactions are also poor, howev- actions 1 and 3 (Table 2) shows that, for certain substrate
er, boric acid does show ca. 70% yield over 22 hours, and combinations, catalyst choice is less important. Howev-
the only other effective catalyst is the bifunctional sys- er, once the substrate combination becomes inherently
tem 6a, which shows 40% yield. Since the catalyst 6a is less reactive (Reactions 2 and 4, Table 2), the possible
stable to proto-deboronation under these conditions, advantage of bifunctional catalysts becomes evident.
this result raises the question as to whether with this Hence, in Reaction 2 (Table 2), 6a is the most active cat-
combination of substrates, there is benefit to using the alyst leading to ca. 80% conversion over 48 hours, with
bifunctional catalyst, compared with 4 or 5. However, the thermal reaction being non-existent over the same
the differences again disappear when applied to more time period. Next most reactive is 4, with boric acid
reactive systems, i.e., as in Reaction 3, 1a and 2b (Ta- and 5 being similar. Interestingly, using the less basic
ble 1), which again shows a substantial thermal reaction and less hindered catalyst 6b results in a drop in activity
À
(ca. 70%), with all catalysts adding only incremental im- to a similar level to 4. It is known that 6a shows no B N
provement.
chelation in both solution and solid states, whereas 6b
Because of the substantial amide formation observed only shows B N chelation under these conditions.[9b] A
in the thermal reactions (Reactions 1 and 3, Table 1), it similar trend is seen in Reaction 4 (Table 2), where again
was necessary to separate actual catalytic effects from there is essentially no thermal reaction and the more ba-
any other processes. This necessarily meant looking sic and hindered bifunctional catalyst 6a is the most re-
for reaction conditions which minimized the thermal re- active catalyst, providing ca. 60% yield of amide over
action. Hence, a second series of parallel reactions was 48 hours.
À
carried out, but at lower temperature in refluxing fluoro-
The findings demonstrated by Tables 1 and 2 highlight
benzene (858C), while maintaining azeotropic removal many issues which have not been fully addressed in the
of water. These reactions were carried out as described literature to date. Firstly, thermolysis of carboxylate-
above and formulated by Scheme 1. Under these lower ammonium salts[11] in non-polar solvent conditions
temperature conditions, catalyst loading had to be in- alone does produce amide products, however, the effi-
creased to 10 mol % in order to obtain reasonable reac- ciency is highly substrate- and temperature-dependent.
tion rates. Each of these reactions was monitored by Presumably, there is a critical balance between the abil-
HPLC over 24 to 48 hours. The results of these experi- ity of the ammonium salt to act as a general Brønsted
ments are shown in Table 2.
acid, and the free amine to act as a reasonable nucleo-
Even in fluorobenzene, the thermal conversion of 1a phile on either the protonated carboxylic acid or the an-
and 2a to 3a is reasonable, i.e., ca. 60% over 24 hours hydride (vide infra). This balance seems to result in a
(Reaction 1, Table 2). However, under these reaction “benzylic effect”, i.e., benzylamine shows higher amide
conditions where there is no evidence of proto-deboro- formation potential than the more nucleophilic amine
nation, the arylboronic acids all show rate improvement 2b, especially when coupled with the more reactive
over the thermal reaction, with 4 being particularly im- acid 1a.[12]
pressive, showing an initial rapid reaction which levels
Secondly, the more electron-deficient catalyst 4 is an
off at ca. 80% yield. Boric acid is generally less effective effective catalyst for amide formation reactions,[7] and
at the lower temperature, except with 1a and 2b where it indeed, similar to boric acid, which is also a good general
is similarly active to the arylboronic acids (Reaction 3, catalyst under the higher temperature conditions.
814
asc.wiley-vch.de
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2006, 348, 813 – 820

Direct Amide Bond Formation from Amines and Carboxylic Acids
COMMUNICATIONS
Table 1. Table of yields versus time and rate constants for different carboxylic acid-amine combinations carried out in reflux-
ing toluene.
Adv. Synth. Catal. 2006, 348, 813 – 820
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
815

Kenny Arnold et al.
COMMUNICATIONS
Table 2. Table of yields versus time and rate constants for different carboxylic acid-amine combinations carried out in reflux-
ing fluorobenzene.
816
asc.wiley-vch.de
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2006, 348, 813 – 820

Direct Amide Bond Formation from Amines and Carboxylic Acids
COMMUNICATIONS
Thirdly, the general reactivity of boric acid at higher lating species are produced in these amide formation re-
temperature raises the question as to why this occurs. actions. However, using ambient, soft ionization electro-
It is known that boric acid reacts readily under acylating spray mass spectrometric techniques, we can gain some
conditions to form tetraacyldiborate systems, i.e., of insight into the species which are being produced in sol-
type 7,[13] however, triacylboranes have also been shown ution[16] (see Supplementary Information). Hence, for
to react via amine complex intermediates to provide the reaction of catalyst 6a with 1a, there is evidence of
amides.[14] It is possible that formation of these species several species, including boroxine 10, diboronate 9
is essential for the catalytic activity of boric acid and and diacyloxyboronate 12, and the absence of either a
that the higher temperature (refluxing toluene) assists monoacyloxyboronate 11 or diacyloxydiboronate (i.e.,
catalyst recycling.
by diacylation of 9, see Scheme 2).[17] Overall, this and
related evidence[7a] does not categorically point to which
is the active acylating species, however, literature prece-
dent[13,14] does suggest that diacyloxyboronate systems in
general are acylating agents, i.e., 12 (Scheme 2).
Finally, from Tables 1 and 2, the change in amide con-
centration with time was found to follow first-order ki-
netics for 1a. In the case of 1b, where the reaction was
Fourthly, we find clear evidence of bifunctional activ- too slow to be fitted to a particular model, initial rate
ity assisting certain substrate-dependent reactions. This constants were calculated. Data were fitted to a first or-
is shown most graphically at lower temperature (Reac- der model[18] and initial rate constants were calculated
tion 2, Table 2). We observe a clear added advantage by regression.[19]
of the more hindered bifunctional system 6a over the
In the case of 1a, first-order kinetics for the catalyzed
less hindered 6b, the non-bifunctional system 5 and the amide formation are followed (see Tables 1 and 2, en-
more electron deficient system 4. It may well be the tries 1 and 3), which is consistent with the reaction pro-
case that the combination of both electron deficiency ceeding through an intermediate. These are likely to be
and an intramolecular base may provide more active, either 11 or 12 derived from acylation[20,21] of the boronic
generally applicable, catalysts in the future. In the inter- acids (a similar intermediate presumably is involved in
im, we have been able to show that bifunctionality is at the case of boric acid) (Scheme 2). In contrast, under
the root of the activity of 6a, as shown by the fact that ad- thermal conditions (see Scheme 3), the likely intermedi-
dition of 10 mol % of diisopropylethylamine to the phe- ate is the anhydride 14 formed by thermolysis,[22] which
nylboronic acid 5-catalyzed reaction (Reaction 4, Ta- would explain the subsequent formation of the amide.[23]
ble 2), does not lead to significant enhanced reactivity Indeed, under thermal conditions, the reaction of benzo-
which approaches that of the bifunctional catalyst 6a ic anhydride with 2a shows that the acylation of anhy-
[ki ¼(2.32ꢀ0.04)ꢁ10À7 c.f. (2.03ꢀ0.04)ꢁ10À7 mol drides by amines is a fast process (0.5 equivs. benzoic an-
dmÀ3 sÀ1 (Table 2)].[15]
hydride, 1 equiv. 2a produces 100% 3d formation in less
Fifthly, NMR (1H, 13C and 11B) and IR[7a] spectroscopic than 2 mins). Hence, in the present study, amine acyla-
studiesare not sufficient to determine exactly which acy- tion is clearly not the rate-determining step on the basis
Scheme 2. Proposed overall mechanism for amide formation involving either boric or arylboronic acid catalysis.
Adv. Synth. Catal. 2006, 348, 813 – 820
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
817

Kenny Arnold et al.
COMMUNICATIONS
of the large difference in rate between amide formation exact nature of which is yet to be fully elucidated. How-
from the pre-formed anhydride and the thermal reac- ever, for more difficult amidations this catalyst is superi-
tions shown in Tables 1 and 2. Therefore, for the thermal or to other monofunctional boronic acid catalysts. Fur-
reaction, formation of the anhydride 14 is likely to be ther studies are underway to examine this process in
rate-limiting. It is therefore also possible, by analogy more detail and to develop improved catalytic systems
with the thermal reactions, that for the catalyzed reac- in the future.
tions, amine acylation is also not rate-determining,
hence carboxylate activation is likely to be rate-deter-
mining and is possibly achieved by formation of either
Experimental Section
the monoacyloxyboronic acid 11 as proposed by Yama-
moto et al.[7a] or, as favoured by us, the diacyloxyboro-
General Remarks
All 1H and 13C NMR were recorded with either a Varian Mer-
cury-400, Bruker Avance-400 or Varian Inova-500 spectrome-
nate 12 (vide supra). It is also of interest to note that
Brown[20] reported the influence of the substituents on
the formation of acyloxyboranes from carboxylic acids
and boranes, where carboxylic acids with lower pKa val- ter. 11B NMR were recorded with the Bruker Avance-400 at a
frequency of 128 MHz. Chemical shifts are expressed as parts
per million (ppm) downfield from the internal standard TMS.
Mass spectra were performed with a Micromass Autospec.
IR spectra were recorded with a Perkin-Elmer 1615 FTIR
spectrometer. Elemental analysis was performed using an Ex-
eter Analytical E-440 Elemental Analyser. Melting points
were determined using an electrothermal melting point appa-
ratus. All reagents were obtained from Aldrich or Lancaster
and were used as received. Molecular sieves were activated
by heating to 450 8C under vacuum (<2 mmHg). Reactions
ues were shown to slow the formation of acyloxyborane.
This can be related to the difference in the rate of reac-
tion between carboxylic acids 1b and 1a towards amide
formation (pKa ¼4.19 and 4.76, respectively). There-
fore, the more electron-rich carboxylic acid 1a is inher-
ently more reactive towards formation of the acyloxy-
boronate, and hence, undergoes amide formation
more readily. It is also likely that this reactivity is paral-
leled in the monofunctional catalysts: more electron-de-
ficient boron catalysts would be expected to have higher were stirred using a magnetic stirrer bar. All parallel reactions
were performed on a Gilson 215 Synthesis Workstation equip-
ped with ReactArray racks and heating block, carried out using
ReactArrray Control Software (version 3,0,0,3048) and HPLC
data were analyzed either directly using Gilson Unipoint (ver-
sion 5.11) or using ReactArray DataManager (version
1,1,33,0). HPLC conditions were under Gilson Unipoint (ver-
sion 5.11) control and injections carried out in conjunction
with ReactArray DataManager. The HPLC system consisted
of Gilson 322 Pump, Gilson 402 Syringe Pump, Agilent 1100
Series UV Diode Array Detector and Phenomenex Gemini
reactivity towards acylation and, hence, will behave as
superior catalysts.[21] This is manifested by the observa-
tion that the trifluorophenylboronic acid 4 is a superior
catalyst to phenylboronic acid 5.
In summary, the efficiency of amide formation under
thermal and catalyzed conditions is highly substrate-de-
pendent. For alkyl carboxylic acids such as 1a, the addi-
tion of boron-based catalysts does improve reaction rate
and yield of amide despite the competing thermal reac-
tion. There is, however, even greater improvement with C18 5 mm, 150 mmꢁ4.60 mm column.
aryl carboxylic acids such as 1b, where the thermal reac-
tion is essentially non-existent. Indeed, the use of the
General Direct Amide Formation Procedure
(Toluene)
amino-boronate catalyst 6a clearly improves amide for-
mation particularly for aryl carboxylic acids and at lower
reaction temperatures. This catalyst has also been dem-
onstrated to act through a bifunctional mechanism, the
To the reaction vessel was added catalyst (1 mol %), followed
by assembly of a micro-Soxhlet apparatus loaded with activat-
Scheme 3. Proposed overall mechanism for thermal amide formation.
818
asc.wiley-vch.de
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2006, 348, 813 – 820

Direct Amide Bond Formation from Amines and Carboxylic Acids
ed 3 Š molecular sieves. The amine (3.27 mmol) and toluene
COMMUNICATIONS
N-4-Phenylbutyl-4-phenylbutyramide (3b): Yield: 1.03 g
(5.6 mL) were added and the mixture heated at 120 8C for 22
hours under argon. Samples (100 mL) were removed automati-
cally at 2 h intervals, diluted with MeCN (900 mL), mixed and
further diluted sequentially in 2ꢁ10-fold steps (total 1000-
fold dilution). Samples were analyzed by online HPLC
[MeCN (0.05% TFA)/water (0.05% TFA) 70:30 for 10 min
or gradient MeCN (0.05% TFA)/water (0.05% TFA) 0:100
to 100:0 over 15 min; 1 mL minÀ1]. Calibration of HPLC-UV
response was achieved using external calibration curves and
checked by normalization of starting material consumption
against product formation. Four or more (typically six) solu-
tions of analyte acid and analyte amide with graduated concen-
trations covering the concentration range of the amide bond
forming experiment in question were produced. Samples
were analyzed by the same method as used in the original ac-
quisition and the results plotted to give a calibration curve
against which the samples could be calibrated. Peak area
data was extracted using Gilson Unipoint peak recognition
software.
(70%); HPLC (gradient 19 min, tr ¼14.21 min); mp 59–60 8C;
1H NMR (400 MHz, CDCl3): d¼1.47–1.56 (m, 2H, CH2),
1.60–1.69 (m, 2H, CH2), 1.92–2.01 (m, 2H, CH2), 2.15 (t, J¼
7.4 Hz, 2H, CH2), 2.64 (m, 4H, 2ꢁArCH2), 3.26 (m, 2H, CH2
N), 5.44 (br s, 1H, CONH), 7.15–7.22 (m, 6H, ArH), 7.25–
7.31 (m, 4H; ArH); 13C NMR (100.6 MHz, CDCl3): d¼27.3
(CH2), 28.8 (CH2), 29.3 (CH2), 35.3 (CH2), 35.6 (CH2), 36.0
(CH2), 39.4 (CH2), 125.9 (ArC), 126.1 (ArC), 128.45 (ArC),
128.49 (ArC), 128.51 (ArC), 128.6 (ArC), 141.6 (ArC-CH2),
˜
142.2 (ArC-CH2), 172.7 (CONH); IR (film): n¼3304, 2925,
2861, 1635 (s), 1535 (s), 741 (s), 693 (vs) cmÀ1; MS (ES): m/z
(%)¼318.3 (100) [MþNa]þ, 296.3 (60) [MþH]þ; elemental
analysis (%): calcd. for C20H25NO: C 81.31, H 8.53, N 4.74;
found: C 81.24, H 8.57, N 4.69.
N-Benzylbenzamide (3c):[26] Yield: 0.53 g (50%); HPLC
(gradient 15 mins, tr ¼9.12 min).
Preparation of N-4-Phenylbutylbenzamide (3d)
To a stirred solution of benzoyl chloride (0.12 mL, 1 mmol) in
dry Et2O (6 mL) under Ar at 08C, was added triethylamine
(0.21 mL, 1.5 mmol) and 4-phenylbutylamine (0.16 mL,
1 mmol). The reaction mixture was allowed to warm to room
temperature, stirred for 2 hours and then quenched with 5%
(w/v) HCl (5 mL). The organic layer was separated and washed
again with 5% (w/v) HCl (5 mL), then brine (5 mL), 5% (w/v)
NaOH (2ꢁ5 mL), brine (5 mL), dried over MgSO4, and con-
centrated under vacuum to afford N-4phenylbutylbenzamide
as a white solid; yield: 0.25 g (99%); HPLC (gradient 15 min,
General Direct Amide Formation Procedure
(Fluorobenzene)
The appropriate catalyst (0.233 mmol, 10 mol %) was manual-
ly weighed into each reaction vessel, followed by assembly of a
micro-Soxhlet apparatus loaded with activated 3 Š molecular
sieves under argon. Solid reagents were added using the React-
Array as standard solutions (0.5 M in fluorobenzene). Naph-
thalene (0.35 mmol, 15 mol %) and amine (2.33 mmol) were
added to the reaction vessels at ambient temperature. The ap-
propriate amount of fluorobenzene was then added to each re-
action vessel in order to give a final reaction volume of 10 mL.
After heating to reflux, carboxylic acid (2.33 mmol) was added
to the stirred solution. Reactions were sampled (50 mL) at 2 or
4 h intervals (24 or 48 h reaction time respectively). Samples
were quenched with MeCN (950 mL), diluted once (50 mL in
950 mL MeCN) mixed and analyzed by HPLC [gradient
MeCN (0.05% TFA)/water (0.05% TFA) 0:100 to 100:0
over 15 or 19 minutes; 1 mL minÀ1). Naphthalene was used
as an internal standard, with response factors calculated auto-
matically by ReactArray DataManager.
1
tr ¼10.83 min); mp 76–77 8C; H NMR (400 MHz, CDCl3):
d¼1.60–1.78 (m, 4H, 2ꢁCH2), 2.66 (t, J¼7.4 Hz, 2H,
ArCH2), 3.47 (m, 2H, CH2N), 6.21 (br s, 1H, CONH), 7.19 (t,
J¼7.1 Hz, 3H, ArH), 7.28 (t, J¼7.6 Hz, 2H, ArH), 7.41 (t,
J¼7.2 Hz, 2H, ArH), 7.49 (t, J¼7.4 Hz, 1H, ArH), 7.75 (dt,
2H, J¼7.4, 1.2 Hz, ArH); 13C NMR (100.6 MHz, CDCl3): d¼
28.8 (CH2), 29.4 (CH2), 35.6 (CH2), 40.0 (CH2N), 126.0
(ArC), 126.9 (ArC), 128.47 (ArC), 128.54 (ArC), 128.7
À
À
(ArC), 131.5 (ArC), 134.8 (ArC CH2), 142.2 (ArC CONH),
˜
167.7(CONH); IR (film): n¼3328, 2936, 2859, 1633 (s), 1521
(s), 691 (vs) cmÀ1; MS (ES): m/z (%)¼276.5 (100) [MþH]þ;
elemental analysis (%): calcd. for C17H19NO: C 80.60, H 7.56,
N 5.53; found: C 80.50, H 7.59, N 5.44.
Acknowledgements
General Procedure for Isolation of Amides
We thank the EPSRC for a research grant (GR/R24722/01), for
DTA studentships (for KA and RG) and the National Mass
Spectrometry Service at Swansea, and GlaxoSmithKline for
CASE funding (KA and RG), S. M. Lynn (GSK, Tonbridge)
for information retrieval.
A 2-necked round-bottomed flask was equipped with stirrer
bar, pressure equalizing dropping funnel (in vertical neck)
with a Soxhlet thimble containing CaH2 (ꢂ1 g) inside, fol-
lowed by a condenser. The appropriate carboxylic acid
(5 mmol), followed by fluorobenzene (50 mL), and amine
(5 mmol) were added, followed by catalyst 6a (117.6 mg,
0.5 mmol, 10 mol %). The mixture was allowed to stir at reflux
for 24 h, before being concentrated under vacuum. The residue
was then redissolved in DCM (25 mL), washed with brine
(25 mL), 5% (w/v) HCl (25 mL), brine (25 mL), 5% (w/v)
NaOH (25 mL), brine (25 mL), dried over MgSO4, and the sol-
vent evaporated under vacuum.
References and Notes
[1] B. M. Trost, I. Fleming, in: Comprehensive Organic Syn-
thesis, (Ed.: E. Winterfield), Pergamon, Oxford, 1991,
Vol. 6.
N-Benzyl-4-phenylbutyramide (3a):[25] Yield: 0.86 g (68%);
HPLC (gradient 15 min, tr ¼10.28 min).
Adv. Synth. Catal. 2006, 348, 813 – 820
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
819

Kenny Arnold et al.
COMMUNICATIONS
[2] a) B. M. Trost, Science 1991, 254, 1471; b) B. M. Trost,
Angew. Chem. 1995, 107, 285; Angew. Chem. Int. Ed.
Engl. 1995, 34, 259.
[3] a) C. A. G. N. Montalbetti, V. Falque, Tetrahedron 2005,
61, 10827; b) Y. R. Mahajan, S. M. Weinreb, in Science
of Synthesis 2005, Vol. 21, Thieme, Stuttgart, New
York, p 17.
[4] a) J. A. Mitchell, E. E. Reid, J. Am. Chem. Soc. 1931, 53,
1879 and references cited therein; b) B. S. Jursic, Z.
Zdravkovski, Synth. Commun. 1993, 23, 2761.
[5] L. Perreux, A. Loupy, F. Volatron, Tetrahedron 2002, 58,
2155.
[13] A. Dal Negro, L. Ungaretti, A. Perotti, J. Chem. Soc.
Dalton Trans. 1972, 1639.
[14] A. Pelter, T. E. Levitt, P. Nelson, Tetrahedron 1970, 26,
1899.
[15] Diisopropylethylamine alone does not catalyze this reac-
tion.
[16] Electrospray mass spectrometry is recognized as being a
sufficiently mild technique to allow the observation of
solution phase intermediates which are not a facet of
the technique, i. e., caused through heating effects, see:
a) A. P. Bruins, T. R. Covey, J. D. Henion, Anal. Chem.
1987, 59, 2642; b) C. M. Whitehouse, R. N. Dreyer, M.
Yamashita, J. B. Fenn, Anal. Chem. 1985, 57, 675.
[17] See Supporting Information.
[6] a) A. Pelter, T. E. Levitt, P. Nelson, Tetrahedron 1970, 26,
1539; b) R. Latta, G. Springsteen, B. Wang, Synthesis
2001, 1611; c) W. Yang, X. Gao, G. Springsteen, B. [18] Micromathꢁ Scientistꢁ for Windowsꢁ, Version 2.02.
Wang, Tetrahedron Lett. 2002, 43, 6339.
[19] Microsoftꢁ Excel, 2003.
[7] a) K. Ishihara, S. Ohara, H. Yamamoto, J. Org. Chem.
1996, 61, 4196; b) K. Ishihara, S. Ohara, H. Yamamoto,
Macromolecules 2000, 33, 3511; c) K. Ishihara, S. Kondo,
H. Yamamoto, Synlett 2001, 1371; d) K. Ishihara, S.
Ohara, H. Yamamoto, Org. Synth. 2002, 79, 176; e) T.
Maki, K. Ishihara, H. Yamamoto, Org. Lett. 2005, 7,
5043.
[20] H. C. Brown, T. P. Stocky, J. Am. Chem. Soc. 1977, 99,
8218.
[21] A. Pelter, M. G. Hutchings, T. E. Levitt, K. W. Smith,
Chem. Commun. 1970, 347.
[22] D. Davidson, P. Newman, J. Am. Chem. Soc. 1952, 74,
1515.
[23] T. Higuchi, T. Miki, A. C. Shah, A. K. Herd, J. Am.
Chem. Soc. 1963, 85, 3655.
[8] P. Tang, Org. Synth. 2005, 81, 262.
[9] a) R. L. Giles, J. A. K. Howard, L. G. F. Patrick, M. R.
Probert, G. E. Smith, A. Whiting, J. Organomet. Chem.
2003, 680, 257; b) S. W. Coghlan, R. L. Giles, J. A. K. Ho-
ward, M. R. Probert, G. E. Smith, A. Whiting, J. Organo-
met. Chem. 2005, 490, 4784.
[10] a) L. H. Toporcer, R. E. Dessy, S. I. E. Green, J. Am.
Chem. Soc. 1965, 87, 1236; b) K. V. Nahabedian, H. G.
Kuivila, J. Am. Chem. Soc. 1961, 83, 2167.
[24] It has been reported that 4 Š molecular sieves can cata-
lyze amide formation by direct heating of carboxylic
acids and amines between 140 and 1808C, however, the
comparative reactions without sieves was not reported;
see: J. Cossy, C. Pale-Grosdemange, Tetrahedron Lett.
1989, 30, 2771. We have found no evidence of such mo-
lecular sieve-catalyzed reactions, however, thermal reac-
tions alone do occur as reported herein.
[11] The formation of the carboxylate-ammonium salt is im-
mediate for all the combinations reported herein, as
shown by reaction in fluorobenzene at room tempera-
ture. Precipitation of the salt is rapid and quantitative.
[12] Note that aniline is unreactive under all of the relatively
ambient reaction conditions reported herein, presumably
because both its basicity and nucleophilicity are too low.
[25] R. Verma, S. K. Ghosh, J. Chem. Soc. Perkin Trans. 1
1998, 2377.
[26] N. Shangguan, S. Katukojvala, R. Greenberg, L. J. Wil-
liams, J. Am. Chem. Soc. 2003, 125, 7754.
820
asc.wiley-vch.de
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2006, 348, 813 – 820