Alcohols as electrophiles: Iron-catalyzed Ritter reaction and alcohol addition to alkynes

Article abstract of DOI:10.1016/j.tet.2014.03.072

A simple, iron-based catalytic system allows for a straightforward method for the synthesis of primary, secondary, and tertiary amides. The system also allows the addition of benzyl alcohols across phenylacetylene to produce substituted phenyl ketones. This transformation improves and expands the substrate scope beyond that previously reported and proceeds under mild reaction conditions, tolerating air and moisture.

Full text of DOI:10.1016/j.tet.2014.03.072

Tetrahedron xxx (2014) 1e4
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Alcohols as electrophiles: iron-catalyzed Ritter reaction and alcohol
addition to alkynes
*
Latisha R. Jefferies, Silas P. Cook
Indiana University, Department of Chemistry, 800 East Kirkwood Avenue, Bloomington, IN 47405, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
A simple, iron-based catalytic system allows for a straightforward method for the synthesis of primary,
secondary, and tertiary amides. The system also allows the addition of benzyl alcohols across phenyl-
acetylene to produce substituted phenyl ketones. This transformation improves and expands the sub-
strate scope beyond that previously reported and proceeds under mild reaction conditions, tolerating air
and moisture.
Received 11 February 2014
Received in revised form 17 March 2014
Accepted 21 March 2014
Available online xxx
Ó 2014 Published by Elsevier Ltd.
Keywords:
Iron catalysis
Alcohols
Ritter reaction
Atom economy
1. Introduction
The Ritter reaction, discovered at New York University in 1948,1,2
offers a particularly atom-economical approach to the synthesis of
amides (Scheme 1). In the traditional mechanistic paradigm, the
reaction proceeds through the generation of a stable carbocation
(e.g., 3) followed by attack of the nitrile. The newly formed nitrilium
ion (e.g., 4) is then quenched by water to form the amide (e.g., 5)
after a tautomerization event. Since the Ritter reaction generally
requires carbocation formation, it works best for the formation of
sterically encumbered amides. Unfortunately, the traditionally
harsh reaction conditions needed to form carbocations (e.g., stoi-
chiometric sulfuric acid) limits the substrate scope of the reaction.
Despite this limitation, the Ritter reaction has found widespread
use in synthesis. For example, the Ritter reaction enabled the
synthesis of aristotelone,3 isocyanoallopupukeanane,4 and
CrixivanÔ.5 Based on these critical applications, more broadly
useful Ritter variants would be highly valuable to expand its use in
synthesis.
Scheme 1. The Ritter reaction.
as p-nitro, and secondary and tertiary aliphatic alcohols were
unreactive under these conditions. In 2009, Cossy and co-workers
reported an inexpensive, environmentally friendly Ritter reaction
based on FeCl3$6H2O (Scheme 2).7 This Ritter reaction provided the
target amides in good yields, but the starting materials were lim-
ited to benzyl alcohols and t-butyl acetate as substrates. These re-
actions also required relatively high temperatures (150 ꢀC).
We recently reported the powerful dehydration properties of
a FeCl3/AgSbF6 system in a formal FriedeleCrafts alkylation re-
action.8 Under our conditions, unactivated secondary alcohols were
competent electrophiles in arene alkylation reactions for the first
time and can provide enantioenriched products.9 In an effort to
further elucidate the utility of this catalytic system, we applied our
conditions to the Ritter reaction. Here we report a general catalytic
system for the Ritter reaction with acetonitrile.
The search for mild conditions capable of effecting a catalytic
Ritter reaction resulted in the first Lewis acid-catalyzed Ritter re-
action reported in 1994.6 The amidation of secondary benzylic al-
cohols was catalyzed by 0.1e0.4 equiv of boron trifluoride etherate
complex (BF3$OEt2) in good-to-excellent yields (Scheme 2). Sec-
ondary benzylic alcohols with electron-withdrawing groups, such
* Corresponding author. Tel.: þ1 812 856 3273; fax: þ1 812 855 3000; e-mail
address: sicook@indiana.edu (S.P. Cook).
0040-4020/$ e see front matter Ó 2014 Published by Elsevier Ltd.
http://dx.doi.org/10.1016/j.tet.2014.03.072
Please cite this article in press as: Jefferies, L. R.; Cook, S. P., Tetrahedron (2014), http://dx.doi.org/10.1016/j.tet.2014.03.072

2
L.R. Jefferies, S.P. Cook / Tetrahedron xxx (2014) 1e4
established that AgSbF6 salts can dramatically increase the dehy-
drating power of FeCl3,8 the reaction was run with FeCl3 in the
presence of AgSbF6 (entry 13, Table 1). To our delight, we observed
the formation of amide 8ac in 36% yield. Unfortunately, additional
optimization studies did not provide higher yields.
With the optimum conditions in hand, we explored the scope of
the reaction of acetonitrile with various activated and unactivated
alcohols (Table 2). We were pleased to find that our reaction con-
ditions provided the desired Ritter products across a range of cyclic,
secondary alcohols (entries 1e5, Table 2). Although benzyl alcohols
represent a common substrate for the Ritter reaction,10 primary
benzylic alcohols remain rare.6,10 Consequently, we were interested
in comparing the reactivity of both primary and secondary benzylic
alcohols. Both benzyl alcohol (7f) and 1-phenylethanol (7g) per-
formed well in the reaction, returning products 8bf and 8bg in 83%
and 87% yield, respectively. In cases where the reaction provides
modest yields (e.g., entry 3, Table 3), increasing the catalyst loading
to 25 mol % provides synthetically useful yields (e.g., 63% yield of
8bc). Unfortunately, catalyst loadings higher than 25% did not
provide any further improvements in yield. Consequently, these
conditions provide a promising method for use in the Ritter re-
action with activated and unactivated alcohols.
Scheme 2. Examples of Lewis acid-catalyzed Ritter reactions.
2. Results and discussion
2.1. The Ritter reaction
Table 2
Substrate scope of the Ritter reaction
FeCl3 (0.15 equiv),
O
AgSbF6 (0.45 equiv)
To test the viability of unactivated secondary alcohols in the
Ritter reaction, cyclohexanol (7c) was treated with a variety of
Lewis acids in dichloroethane at 80 ꢀC (Table 1). As can be seen in
Table 1, only iron(III) chloride (entry 1 and 2, Table 1), bismuth(III)
triflate (entry 8, Table 1), and aluminum(III) chloride (entry 9, Table
1) provided detectable quantities of amide 8ac. Since we have
CH3CN
ROH
R
H3C
N
H
(3 equiv)
DCE, 0.1 M, 100 °C, 24h
6b
7
8bx
Entry
1
7
Product (8)
Isolated yield (%)
Trace
H
OH
CH3
N
8ba
8bb
7a
7b
O
Table 1
Examination of Lewis acids in the Ritter reaction
OH
OH
H
CH3
N
OH
N
2
3
34
O
Catalyst (0.15 equiv)
O
Ph
N
H
DCE, 0.1M,
80 °C, 24 h
H
3 equiv
N
36
CH3
7c
7d
8bc
8bd
63a
6a
7c
8ac
O
Entry
Catalyst
Yield %
1
FeCl3
6
9
36
0
0
0
CH3
OH
2
3
4
FeCl3$6H2O
FeCl3 w/3AgSbF6
FeCl2
HN
4
5
37
23
O
5
6
7
Fe(BF4)2$3H2O
FeF3$3H2O
BiCl3
CH3
O
OH
0
HN
8
Bi(OTf)3
19
<5
0
0
0
<5
18
<5
<5
0
0
7
26
7
<5
<5
<5
12
<5
9
AlCl3
7e
8be
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
CuOTf
CuCl2
ZnCl2
AgSbF6
FeCl3 w/3AgAsF6
FeCl3 w/3AgPF6
FeCl3 w/3AgOTf
FeCl3 w/3AgNO3
FeCl3 w/3AgOAc
FeCl2 w/3AgSbF6
FeCl3$6H2O w/3AgSbF6
FeF3$3H2O w/3AgSbF6
HCla
O
OH
6
83
87
7f
8bf
N
H
CH3
OH
CH3 7g
CH3
O
7
8bg
CH3
N
H
TfOH
pTSA$H2O
H2SO4
a
25 mol % FeCl3 and 75 mol % AgSbF6 used.
b
HSbF6
a
To explore the Ritter reaction in the context of a medicinally
important small molecule, we chose a new route to adamantadine.
4.0 M in 1,4-dioxane.
In 65e70% aqueous solution.
b
Please cite this article in press as: Jefferies, L. R.; Cook, S. P., Tetrahedron (2014), http://dx.doi.org/10.1016/j.tet.2014.03.072

L.R. Jefferies, S.P. Cook / Tetrahedron xxx (2014) 1e4
3
Table 3
Substrate scope of the alcohol addition to alkynes
FeCl3 (0.15 equiv),
AgSbF6 (0.45 equiv)
O
R
ROH
DCE, 0.1 M, 80 °C, 24h
15
7
16
OH
OH
OH
CH3
7g
OH
CH3
CH3
Br
I
7h
7i
7j
OH
OH
CF3
OH
OH
7f
7k
7c
10
Scheme 3. The synthesis of amide 13.
Entry
1
ROH
Product
CH3
Isolated yield
2.2. Benzyl alcohol addition to alkynes
O
7g
7h
60%
92%
16g
Another atom-economical reaction is the alcohol addition to
alkynes (Scheme 4). In this reaction, loss of water leads to the for-
mation of a new carbonecarbon bond followed by a hydration of an
alkyne to provide ketone products. Such a reaction seemed ideally
suited to demonstrate the capability of our new dehydration sys-
tem. Cognizant of the recent report by Jana and co-workers de-
scribing the addition of benzyl alcohols to alkynes using FeCl3 as
the catalyst,16 we were pleased to find our system provided sig-
nificantly better yields.
2
16h
O
CH3
O
16i
16j
3
4
7i
7j
95%
73%
Br
Scheme 4. The addition of alcohols to alkynes.
CH3
O
As can be seen in Table 3, the reaction worked well with phe-
nylacetylene and various benzylic alcohols. For substrates 7g and 7h
(entries 1 and 2, Table 3), the addition of AgSbF6 proved critical for
high yields.16 When election poor substrate 7k was employed, the
reaction recovered starting material. As a testament to the power of
this new dehydration system, primary alcohol 7f produced ketone
16f in 50% yield (entry 6, Table 3) in contrast to previous reports
where 7f was unreactive.16 Based on our experience with unac-
tivated secondary alcohols,8 cyclohexanol (7c) was subjected to the
reaction conditions (entry 7, Table 3). Surprisingly, none of desired
ketone 16c was detectable in the reaction. While explanations for
failure abound, one possibility is that the secondary alcohols do not
pass through carbocation intermediates while benzylic alcohols
access carbocations en route to product. Finally, adamantanol (10)
also performed well in the reaction, providing adamatyl phenyl
ketone 16l in 72% yield (entry 8, Table 3).
I
O
O
5
6
7
7f
50%
0%
16f
CF3
7k
7c
16k
O
0%
16c
O
As part of a cursory mechanistic investigation, enantioenriched
alcohols (R)-7g and (S)-7g were subjected to the optimized re-
action conditions (Scheme 5). The resulting phenyl ketone 16g was
isolated as a racemic mixture. The erosion of enantioenrichment is
consistent with the intermediacy of benzylic carbocations. Based
on these results, a mechanistic hypothesis is provided in Scheme 6.
8
10
72%
16l
Adamantadine (14), originally approved for use against the Asian
flu in 1966, is now used widely as both an antiviral and anti-
parkinsonian drug.11e14 While the drug has been prepared through
the Ritter reaction, previous syntheses utilize 1-adamantyl bromide
(11) and adamantyl-1-nitrate (12) as starting materials (Scheme 2).
With the technology described above, we envisioned the use of 1-
adamantol (12) directly in the synthesis of adamantadine. We were
delighted to find the treatment of 1-adamantol (10) with catalytic
FeCl3/AgSbF6 in the presence of 3 equiv of acetonitrile produced
amide 14 in 86% yield (Scheme 3). This reaction avoids the use of
oleum and large excess of acetonitrile. Due to the greater scarcity of
acetonitrile supplies, this has become a considerable issue.
Scheme 5. Racemization of enantioenriched 7g under standard conditions.
Please cite this article in press as: Jefferies, L. R.; Cook, S. P., Tetrahedron (2014), http://dx.doi.org/10.1016/j.tet.2014.03.072

4
L.R. Jefferies, S.P. Cook / Tetrahedron xxx (2014) 1e4
(10 mL) and allowed to stir until FeCl3 was completely dissolved. To
this solution was added phenylacetylene (0.17 mL, 1.5 mmol) im-
mediately followed by the AgSbF6 (154.6 mg, 0.45 mmol), which
was then capped and put into a mechanical shaker at 80 ꢀC for 24 h.
The reaction was quenched with water (30 mL) extracted with DCM
(3ꢁ30 mL). The organic extracts were combined, dried (MgSO4),
filtered, and concentrated to give the residue. The residue was
purified by silica flash chromatography using hexane and ethyl
acetate (9:1) to give desired product 2-(adamantan-1-yl)-1-
phenylethan-1-one 16l (184.2 mg, 72% yield) as a yellow oil; Rf
(hexanes:EtOAc¼9:1)¼0.60; nmax (liquid film) 2900, 2848,
1674 cmꢂ1
;
1H NMR (400 MHz CDCl3)
d
7.94 (d, J¼7.3, 1.7 Hz, 2H),
Scheme 6. A mechanistic hypothesis for alcohol addition to alkynes.
7.58e7.50 (m, 1H), 7.45 (dd, J¼8.4, 6.9 Hz, 2H), 2.72 (s, 2H), 1.94 (q,
J¼3.1 Hz, 3H), 1.74e1.59 (m, 12H); 13C NMR (101 MHz, CDCl3) 200.3,
138.9, 132.7, 128.4, 128.3, 51.2, 43.0, 36.7, 33.9, 28.7; HRMS: C15H22
requires 254.1671, found: 254.1666.
3. Conclusions
In conclusion, we present a simple method to access primary,
secondary, and tertiary amides via the Ritter reaction. The powerful
dehydration conditions also permit the addition of benzyl alcohols
across phenylacetylene in significantly higher yields than pre-
viously reported. The reaction is simple and based on a low-cost
iron catalyst. Further, it provides an atom-economical synthesis of
amides and phenyl ketones.
Acknowledgements
We gratefully acknowledge Eli Lilly and Company for recent fi-
nancial support through the Lilly Grantee Award. We acknowledge
funding from the American Chemical Society Petroleum Research
Fund (PRF52233-DNI1) and Indiana University for support of this
work.
4. Experimental section
Supplementary data
4.1. N-Cyclohexylbenzamide (8ac)
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.tet.2014.03.072.
To an oven dried 20 mL vial containing FeCl3 (24.3 mg,
0.15 equiv) was added a solution of cyclohexanol (1 mmol) in DCE
(10 mL) and allowed to stir until FeCl3 was completely dissolved
(10e15 min). To this solution was added the phenylnitrile (3 mmol,
References and notes
3
equiv) immediately followed by the AgSbF6 (154.6 mg,
1. Ritter, J.; Kalish, J. J. Am. Chem. Soc. 1948, 70, 4048e4050.
2. Ritter, J.; Minieri, P. P. J. Am. Chem. Soc. 1948, 70, 4045e4048.
3. Stoermer, D.; Heathcock, C. H. J. Org. Chem. 1993, 58, 564e568.
4. Ho, T.-L.; Kung, L.-R.; Chein, R.-J. J. Org. Chem. 2000, 65, 5774e5779.
5. Senanayake, C. H.; Roberts, F. E.; DiMichele, L. M.; Ryan, K. M.; Liu, J.; Fre-
denburgh, L. E.; Foster, B. S.; Douglas, A. W.; Larsen, R. D.; Verhoeven, T. R.;
Reider, P. J. Tetrahedron Lett. 1995, 36, 3993e3996.
0.45 equiv), which was then capped and put into mechanical shaker
at 80 ꢀC for 24 h. The reaction was quenched with water (30 mL)
extracted with DCM (3ꢁ30 mL). The organic extracts were com-
bined, dried (MgSO4), filtered, and silica chromatography
(hexanes:EtOAc¼9:1) to yield 8ac (72.8 mg, 36%) as a yellow oil
whose 1H and 13C NMR matches the reported spectra.15 1H NMR
6. Firouzabadi, H.; Sardarian, A. R.; Badparva, H. Synth. Commun. 1994, 24,
601e607.
(500 MHz, CDCl3)
d
7.76 (d, J¼7.4 Hz, 2H), 7.48 (t, J¼7.3 Hz, 1H), 7.41
ꢀ
7. Anxionnat, B.; Guerinot, A.; Reymond, S.; Cossy, J. Tetrahedron Lett. 2009, 50,
(t, J¼7.5 Hz, 2H), 6.13 (s, 1H), 4.12e3.83 (m, 1H), 2.03 (d, J¼9.4 Hz,
2H), 1.76 (dd, J¼10.0, 3.6 Hz, 2H), 1.66 (dd, J¼9.3, 3.7 Hz, 1H),
1.49e1.35 (m, 2H), 1.33e1.15 (m, 3H). 13C NMR (126 MHz, CDCl3)
3470e3473.
8. Jefferies, L. R.; Cook, S. P. Org. Lett. 2014, http://dx.doi.org/10.1021/ol500606d
9. For the only example of a nonracemic arylation of activated enantioenriched
alcohols, see: Vicennati, P.; Cozzi, P. G. Eur. J. Org. Chem. 2007, 2248e2253.
d
166.7, 135.1, 131.2, 128.5, 126.9, 48.7, 33.2, 25.6, 24.9.
ꢀ
10. Guerinot, A.; Reymond, S.; Cossy, J. Eur. J. Org. Chem. 2012, 19e28.
11. Stetter, H.; Mayer, J.; Schwarz, M.; Wulff, K. Chem. Ber. 1960, 93, 226e230.
12. Maugh, T. H. Science 1976, 192, 130e131.
13. Moiseev, I. K.; Doroshenko, R. I.; Ivanova, V. I. Pharm. Chem. J. 1976, 10, 450e451.
14. Maugh, T. Science 1979, 206, 1058e1060.
15. Ren, W.; Yamane, M. J. Org. Chem. 2010, 75, 8410e8415.
16. Jana, U.; Biswas, S.; Maiti, S. Eur. J. Org. Chem. 2008, 5798e5804.
4.2. 2-(Adamantan-1-yl)-1-phenylethan-1-one (17i)
To an oven dried vial containing FeCl3 (24.3 mg, 0.15 mmol) was
added a solution of adamantanol (152.3 mg, 1 mmol) in DCE
Please cite this article in press as: Jefferies, L. R.; Cook, S. P., Tetrahedron (2014), http://dx.doi.org/10.1016/j.tet.2014.03.072