Two-step iron(0)-mediated N-demethylation of N -methyl alkaloids

Article abstract of DOI:10.1021/jo1008492

(Figure Presented) A mild and simple two-step Fe(0)-mediated N-demethylation of a number of tertiary N-methyl alkaloids is described. The tertiary N-methylamine is first oxidized to the corresponding N-oxide, which is isolated as the hydrochloride salt. Subsequent treatment of the N-oxide hydrochloride with iron powder readily provides the N-demethylated amine. Representative substrates include a number of opiate and tropane alkaloids. Key intermediates in the synthesis of semisynthetic 14-hydroxy pharmaceutical opiates such as oxycodone and oxymorphone are also readily N-demethylated using this method.

Full text of DOI:10.1021/jo1008492

pubs.acs.org/joc
Two-Step Iron(0)-Mediated N-Demethylation of N-Methyl Alkaloids
†
Gaik B. Kok, Cory C. Pye, Robert D. Singer, and Peter J. Scammells*
‡
‡
,†
†
Medicinal Chemistry and Drug Action, Monash Institute of Pharmaceutical Sciences, Monash University,
3
‡
81 Royal Parade, Parkville, Victoria, 3052, Australia, and Department of Chemistry, Saint Mary’s
University, Halifax, Nova Scotia B3H 3C3, Canada
peter.scammells@pharm.monash.edu.au
Received April 30, 2010
A mild and simple two-step Fe(0)-mediated N-demethylation of a number of tertiary N-methyl
alkaloids is described. The tertiary N-methylamine is first oxidized to the corresponding N-oxide,
which is isolated as the hydrochloride salt. Subsequent treatment of the N-oxide hydrochloride with
iron powder readily provides the N-demethylated amine. Representative substrates include a number
of opiate and tropane alkaloids. Key intermediates in the synthesis of semisynthetic 14-hydroxy
pharmaceutical opiates such as oxycodone and oxymorphone are also readily N-demethylated using
this method.
3
4
(von Braun reaction), chloroformate esters, and dialkyl
Introduction
5
azodicarboxylates. Many of these methods have drawbacks
Many tertiary N-methylamines have profound pharma-
cological properties. These include alkaloids from the opiate
family, such as the powerful analgesics morphine (1) and
codeine (2), as well as tropanes such as cocaine (3) and the
potent CNS stimulant atropine (4) (Figure 1). Dextro-
methorphan (5) is a potent NMDA antagonist that is used
in a range of cough and cold formulations.
such as the employment of toxic reagents, use of less readily
available and expensive reagents, or vigorous reaction con-
6
7
ditions. Photochemical and biochemical methods for
N-demethylation have also been reported, but these are
generally relatively low yielding. More recently, Hudlicky
and co-workers have found that hydrocodone and a few
tropane-type alkaloids were N-demethylated/N-acylated
N-Demethylation is a chemical transformation important
to the pharmaceutical industry. Besides being tools for both
pharmacological and metabolic studies, N-demethylated
compounds themselves may have intrinsic biological activity.
For example, N-nortropane (6) has been shown to have
increased potency and selectivity over the corresponding
8
using palladium(II).
The Polonovski reaction is a two-step process for the
9
N-demethylation of tertiary amines (Scheme 1). The tertiary
amine is first converted into the N-oxide, which is then
reacted with an activating agent to afford the N-demethy-
lated product. The classical Polonovski reaction employs
acylating agents such as acid chlorides, acid anhydrides, or
chloroformate esters as the activating agent. However,
1
N-methyl analogue as a serotonin transporter ligand. N-De-
methylated compounds also serve as key synthetic precursors
to other important pharmaceuticals. For example, a number of
semisynthetic pharmaceutical opiates such as naloxone (7) and
naltrexone (8) are derived from the corresponding N-nor
opiates, which in turn are synthesized from naturally occurring
N-methyl opiates.
(4) See, for example: (a) Csutoras, C.; Zhang, A.; Bidlack, J. M.;
Neumeyer, J. L. Bioorg. Med. Chem. 2004, 12, 2687–2690. (b) Kraiss, G.;
N ꢀa dor, K. Tetrahedron Lett. 1971, 12, 57–58. (c) Cooley, J. H.; Evain, E. J.
Synthesis 1989, 1–7. (d) Abdel-Monem, M. M.; Portoghese, P. S. J. Med.
Chem. 1972, 15, 208–210. (e) Olofson, R. A.; Martz, J. T.; Senet, J. P.; Piteau,
M.; Malfroot, T. J. Org. Chem. 1984, 49, 2081–2082.
A number of general methods for N-demethylation of
2
tertiary amines have been reported in the literature. Most
notably, these methods include the use of cyanogen bromide
(5) See, for example: (a) Schwab, L. S. J. Med. Chem. 1980, 23, 698–702.
(
b) Merz, H.; Pook, K. -H. Tetrahedron 1970, 26, 1727–1741.
6) Ripper, J. A.; Tiekink, E. R. T.; Scammells, P. J. Bioorg. Med. Chem.
Lett. 2001, 11, 443–445.
7) (a) Kieslich, K. In Microbial Transformations of Non-Steroid Cyclic
(
(
(
1) Emond, P.; Helfenbein, J.; Chalon, S.; Garreau, L.; Vercouillie, J.;
Frangin, Y.; Besnard, J. C.; Guilloteau, D. Bioorg. Med. Chem. 2001, 9,
849–1855.
2) Thavaneswaran, S.; McCamley, K.; Scammells, P. J. Nat. Prod.
Commun. 2006, 1 (10), 885–897.
3) von Braun, J. Ber. Dtsch. Chem. Ges. 1900, 33, 1438–1452.
Compounds; Thieme: Stuttgart, 1976; pp 213-215. (b) Chaudhary, V; Leisch, H.;
Moudra, A.; Allen, B.; De Luca, V.; Cox, D. P.; Hudlicky, T. Collect. Czech.
Chem. Commun. 2009, 74, 1179–1193.
(8) Carroll, R. J.; Leisch, H.; Scocchera, E.; Hudlicky, T.; Cox, D. P. Adv.
Synth. Catal. 2008, 350, 2984–2992.
(9) Grierson, D. Org. React. 1990, 39, 85–295.
1
(
(
4
806 J. Org. Chem. 2010, 75, 4806–4811
Published on Web 06/21/2010
DOI: 10.1021/jo1008492
r 2010 American Chemical Society

Kok et al.
JOCArticle
FIGURE 1. Some examples of natural and semisynthetic opiate and tropane alkaloids.
SCHEME 1. N-Demethylation via the Polonovski Reaction
one of the most abundant metals on earth and is conse-
1
7
quently inexpensive. Furthermore, iron is an essential
metal for living organisms and is typically very environmen-
tally friendly. As a result of these features, it has been
identified as an attractive catalyst for large-scale applica-
tions, such as those conducted by the pharmaceutical
1
8
industry.
N-demethylation of opiate alkaloids has not been successful
1
under “classical Polonovski” conditions.
0,11
Herein we describe a mild and simple two-step Polonovski-
type process for N-demethylation involving the formation and
isolation of the N-oxide hydrochloride and subsequent reac-
tion with iron powder.
In recent years, other variants of the original activating
agent have successfully been employed to N-demethylate
tertiary amine alkaloids, including opiates and tropanes. Our
previous efforts have primarily focused on Fe(II)-mediated
Polonovski-type N-demethylations on a range of opiate and
tropane alkaloids. We have found that a number of Fe(II)-
Results and Discussion
N-Demethylation using iron(0) was initially investigated
using the commercially available substrate dextromethor-
phan (5a). Oxidation of 5a with m-CPBA according to
1
2,13
based reagents including FeSO 7H O
1
and the ferrous
are effective catalysts. Sipos and
4
3
2
4,15
porphyrin, Fe(II)TPPS,
co-workers have also successfully employed FeSO for the
1
2-15
4
published procedures
provided the N-oxide 5b, which
one-pot N-demethylation of apomorphine N-oxide as well as
the acid-catalyzed rearrangement of morphinan N-oxides
was isolated as the corresponding hydrochloride salt. The
N-oxide hydrochloride in an appropriate solvent was treated
with a stoichiometric amount of iron powder, and the result-
ing mixture was stirred at room temperature until starting
material was consumed, as monitored by TLC analysis. A
standard workup followed by purification via column chro-
matography employing a CHCl /MeOH/NH OH gradient
was used to isolate the desired N-nor product 5c from the
corresponding tertiary N-methylamine 5a. The N-demethy-
lation step was initially performed using methanol as the
solvent. However, a number of commonly used solvents
1
6
into noraporphine derivatives. Thus, a process involving
N-oxide formation, isolation of the corresponding N-oxide
hydrochloride, and treatment with the iron(II) reagent has
afforded the desired N-demethylated product, in most cases
with the only byproduct being the parent tertiary amine. This
reaction outcome can be rationalized from a consideration of
3
4
9
the proposed reaction mechanism which involves two suc-
cessive one-electron steps in a Fe(II)/Fe(III) redox system.
Further examination of the different Fe(0)/Fe(II)/Fe(III)
redox potentials suggests that a redox couple involving iron
in the zero oxidation state may prove viable as an alternate
electron source in these Polonovski-type reactions. Iron is
(EtOH, i-PrOH, MeCN, EtOAc, CHCl , and CH Cl ) were
3 2 2
subsequently evaluated. As the results in Table 1 show,
solvent also plays a critical role in product outcome in these
zerovalent iron-mediated reactions. Of the three alcoholic
solvents studied (entries 1-3), the reaction was fastest in
MeOH with the reaction complete in 3 h while the reaction in
i-PrOH (21 h) delivered the best outcome with respect to the
yield of the desired N-nor product 5c (89%). In MeCN, all
substrate N-oxide was consumed within 1.5 h (via TLC
analysis), with the reaction delivering a modest yield of 5c
(
10) Klein, M. In Mass Spectrometry in Drug Metabolism; Frigerio, A.,
Ghisalberti, E. L., Eds.; Plenum Press: New York, 1977; p 449.
11) Allen, A. C.; Moore, J. M.; Cooper, D. A. J. Org. Chem. 1983, 48,
951–3954.
12) McCamley, K.; Ripper, J. A.; Singer, R. D.; Scammells, P. J. J. Org.
Chem. 2003, 68, 9847–9850.
13) Thavaneswaran, S.; Scammells, P. J. Bioorg. Med. Chem. Lett. 2006,
0, 2868–2871.
(
3
(
(
1
(
(
14) Dong, Z.; Scammells, P. J. J. Org. Chem. 2007, 72, 9881–9885.
15) Kok, G.; Ashton, T. D.; Scammells, P. J. Adv. Synth. Catal. 2009,
3
51, 283–286.
16) (a) Sipos, A.; Berenyi, S. Synlett 2008, 11, 1703–1705. (b) Berenyi, S.;
Gyulai, Z.; Udvardy, A.; Sipos, A. Tetrahedron Lett. 2010, 51, 1196–1198.
(17) Bolm, C.; Legros, J.; Le Paih, J.; Zani, L. Chem. Rev. 2004, 104,
6217–6254.
(18) Bauer, E. B. Curr. Org. Chem. 2008, 12, 1341–1369.
(
J. Org. Chem. Vol. 75, No. 14, 2010 4807

Kok et al.
JOCArticle
TABLE 1. Solvent Effects on N-Demethylation of Dextromethorphan
N-Oxide Hydrochloride with Iron Powder
a
b
%
yield
entry
solvent
time (h)
5c
5a
1
2
3
4
5
6
7
a
MeOH
EtOH
i-PrOH
MeCN
EtOAc
3
8
21
1.5
2 (40 °C)
1
1.5
59
62
89
75
72
97
86
24
23
5
21
16
c
CHCl
CH Cl
3
2
2
10
Unless otherwise indicated, reactions employed 1 equiv of iron and
were conducted at room temperature (concentration: 5 mL of solvent
per 100 mg of substrate). Yields following column chromatography.
b
c
Although a trace amount of 5a was observed by TLC, it was not
recovered in the isolation process.
TABLE 2. Effects of Stoichiometry and Temperature on the N-Demethy-
lation of Dextromethorphan N-Oxide Hydrochloride with Iron Powder
FIGURE 2. N-Methyl alkaloid test set.
a
b
reactions on morphine (1a), oripavine (11a), and 13a-16a
were conducted in i-PrOH only. Results are summarized in
Table 3.
%
yield
iron powder
(equiv)
entry solvent
temp (°C) time (h)
5c
5a
For CME (9a), thebaine (10a), and thevinone (12a), solvent
effects were consistent with the previous findings for dextro-
methorphan (5a), with CHCl delivering better outcomes, both
1
2
3
4
5
i-PrOH
i-PrOH
i-PrOH
1
rt
rt
60
rt
21
24
3
1
20
89
88
89
97
97
5
8
7
0.25
0.25
1
3
with respect to reaction time and yield of the N-nor product. The
more polar N-oxides of thebaine (10a) and oripavine (11a)
CHCl
CHCl
3
19
3
0.1
rt
2
a
b
Concentration: 5 mL of solvent per 100 mg of substrate. Isolated via
column chromatography.
were slower to react at lower temperatures, and subsequent
heating at a higher temperature was required to complete
the reactions within a reasonable time frame. After column puri-
fication on silica gel, N-northebaine (10c) and thebaine (10a)
were isolated in 77% and 12% yield, respectively. For oripavine
N-oxide, the respective yields of N-nororipavine (11c) and
oripavine (11a) were 40% and 24%.
Iron powder, using i-PrOH as solvent, was also effective in
the N-demethylation of morphine (1a), delivering, after
column chromatography, a modest yield of the correspond-
ing N-nor product 1c of 58% (entry 1). Four semisynthetic
of 75%. Although there was no apparent product formation
after 3 h at room temperature in ethyl acetate, the reaction
was complete in only 2 h when the temperature was increased
to 40 °C. Chloroform and dichloromethane were found to be
efficient solvents with all starting material being consumed
within 1-1.5 h, delivering the N-nor product 5c in high yields
(
Table 1, entries 6 and 7).
The effects of stoichiometry and temperature were ex-
1
4-hydroxy opiates 13a-16a were also readily N-demethy-
plored to further optimize the reaction conditions for the
N-demethylation of dextromethorphan N-oxide (Table 2).
These experiments were conducted in both 2-propanol and
chloroform. Employing 2-propanol as solvent at room tem-
perature and lowering the catalyst loading from 1 to 0.25
equiv (entries 1 and 2) did not significantly alter the reaction
time (21-24 h) or product yield ratio (5c:5a). By increasing
the temperature to 60 °C (entry 3), the reaction took only 3 h
to complete, without compromising the product yield ratio.
lated (Table 3, entries 12-16). N-Noroxycodeinone (13c)
and N-noroxymorphinone (14c) were found to be unstable to
column chromatography using silica gel and were isolated
via extraction with a suitable solvent or solvent combination.
This extraction protocol was also successfully employed for
the isolation of N-noroxymorphone (16c) from oxymor-
phone (16a). Noroxycodone (15c) was readily isolated from
oxycodone (15a) via column chromatography on silica gel
employing a CHCl /MeOH/NH OH gradient.
3
4
Likewise, in CHCl , lowering the catalyst loading to 0.1
3
molar equiv increased the completion time for the reaction
Iron powder has also proven to be effective for the
N-demethylation of tropane alkaloids (Table 3, entries
(
20 h) without any consequence on the yield of 5c (97%) and
with only a small amount of the parent amine 5a isolated
2%).
To evaluate the scope of this protocol, we have extended
the study to a number of other representative opiate and
tropane alkaloids, including morphine (1a) and atropine
1
7-20). For example, atropine (4a) and tropine (17a) were
readily N-demethylated to afford N-noratropine (4c) and
N-nortropine (17c), respectively. Interestingly, for the tro-
panes, a slightly more favorable product ratio was obtained
for reactions in i-PrOH compared to CHCl3.
(
In summary, in cases where substrate solubility was not a
limiting factor, chloroform was shown to be an effective
solvent across a range of opiate substrates studied. For the
tropane alkaloids 4a and 17a, we have found that reactions in
i-PrOH delivered slightly better outcomes.
(
4a). Other substrates are shown in Figure 2. Notably, these
also include a number of the key pharmaceutical intermedi-
ates; namely, 14-hydroxy semisynthetic opiates 13a-16a.
For CME (9a) and thebaine (10a), reactions were con-
ducted in MeOH, CHCl , and i-PrOH. However, given the
3
less favorable outcomes for 5a, 9a, and 10a using MeOH as
solvent, investigations on thevinone (12a) were performed in
CHCl and i-PrOH only. Due to limited solubility in CHCl ,
(
19) In CHCl
thebaine N-oxide isomers were 0.45 and 0.53; for the two N-oxide isomers
of oripavine, the R values were 0.29 and 0.36.
3 4 f
/MeOH/NH OH (85:15:1), the R values for the two
3
3
f
4
808 J. Org. Chem. Vol. 75, No. 14, 2010

Kok et al.
JOCArticle
a
TABLE 3. N-Demethylation of Other Opiate and Tropane Alkaloids with Iron Powder
b
%
yield
entry
N-oxide hydrochloride of
c
iron (equiv)
solvent
time (h)
temp (°C)
#c
#a
d
1
2
3
4
5
morphine, 1b
CME, 9b
0.5
1.0
1.0
1.0
0.25
1.3
1.3
1.3
2.0
1.0
1.0
1.0
0.5
0.5
2.0
0.5
1.0
1.0
1.0
1.0
i-PrOH
MeOH
CHCl
24
3
2
48
48
48
1
48
48
2
40
58
61
97
88
85
30
86
77
40
44
15
46
40
59
55
42
76
81
66
74
b
36
12
1
CME, 9b
CME, 9b
CME, 9b
3
i-PrOH
i-PrOH
MeOH
CHCl
3
i-PrOH
i-PrOH
7
11
59
13
12
24
26
76
c
6
thebaine, 10b
thebaine, 10b
thebaine, 10b
oripavine, 11b
thevinone, 12b
thevinone, 12b
7
8
e
rt-50
f
g
d
i
9
40-60
e
c
10
CHCl
3
c
h
1
1
2
3
4
5
6
7
8
9
0
a
i-PrOH
i-PrOH
i-PrOH
i-PrOH
i-PrOH
i-PrOH
3
60
40
1
1
1
1
1
1
1
1
2
oxycodeinone, 13b
oxymorphinone, 14b
oxycodone, 15b
oxycodone, 15b
oxymorphone, 16b
atropine, 4b
27
168
120
22
144
2
26i
44
22
26
40
60
40
i
41
CHCl
i-PrOH
CHCl
i-PrOH
3
15
17
11
3
c
j
atropine, 4b
tropine, 17bc
tropine, 17b
50
3
40
2
Unless otherwise specified, reactions were performed at room temperature at a concentration of 5 mL per 100 mg of substrate. Yield via column
c
d
4
chromatography. Concentration: 10 mL per 100 mg of substrate. No NH OH was added during workup; accordingly, the corresponding HCl salts
f g
e
were isolated. The reaction mixture was stirred at rt for 24 h and then at 50 °C for 24 h. Concentration: 20 mL per 100 mg of substrate, Reaction heated
h
i
at 40 °C for 24 h and then 60 °C for 24 h. No apparent reaction after 72 h at rt. Isolated via extraction of an aqueous solution of the crude at pH 2-10
j
with a suitable solvent/solvent mixture. After 48 h at rt, the reaction mixture was heated at 60 °C for 2 h.
TABLE 4. N-Demethylation of Dextromethorphan and Thebaine with
Added Salts
1 equiv offered no further advantages. Added FeSO 7H O
4 2
a
3
did not have a significant effect on the product ratio, but did
accelerate the reaction (cf. entries 1 and 4). On the other
hand, addition of CuCl , Cu(ClO ) and Cu(NO ) resulted
c
%
yield
2
4 2
3 2
iron
entry N-oxide (equiv)
b
in an increase in the proportion of dextromethorphan (5a)
added salt
solvent time (h) #c #a #b
(
entries 5-7).
1
2
3
4
5
6
7
8
9
0
1
a
5b
5b
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.3
1.3
1.3
1.3
MeOH
MeOH
MeOH
120
6
58 21
81 12
9
9
d
d
Further studies on the effect of added anions were con-
CuSO
ZnSO
4
5H
2
O
O
3
3
ducted using thebaine as substrate (Table 4, entries 8-11).
With MeOH as solvent, the reaction of thebaine N-oxide
hydrochloride (10b) with iron powder alone returned more
thebaine (10a) than the N-nor product 10c (% yield of 10c/
10a = 30:59) (vide supra). Interestingly, addition of 0.25
equiv of CuSO 5H O reversed this outcome, delivering N-
northebaine (10c) and thebaine (10a) in respective yields of
72% and 25%. While the reaction outcome with MeOH as
solvent was positively moderated via the addition of Cu-
SO 5H O, in a non-coordinating medium such as chloro-
5b
5b
4
7H
2
120
24
120
120
120
48
48
1
81
9
FeSO
4
7H
2
O
MeOH
MeOH
MeOH
59 25
54 40
47 48
38 55
30 59
72 25
86 13
86 11
3
5b
5b
CuCl
Cu(ClO
Cu(NO
2
4
4
7
4
)
2
3
6H
2
O
1
2 /
5b
3
)
2
2
H
2
O MeOH
MeOH
3
10b
10b
10b
10b
4
3
2
CuSO
CuSO
4
4
3 5H2
3 5H2
O
MeOH
1
1
CHCl
CHCl
3
O
3
1
Unless otherwise stated, reactions were conducted at room tempera-
b
ture at a concentration of 10 mL of solvent per 100 mg of substrate. 0.25
equiv of salt. Yield after column chromatography. No reaction in the
absence of Fe.
4
3
2
c
d
form, addition of CuSO 5H O offered no further advantages
4 2
on the product yield ratio (entries 10 and 11).
3
Previous studies on the iron-catalyzed dealkylation of
trimethylamine oxide to a secondary amine and formalde-
hyde have found that the decomposition rate is dependent on
Conclusions
In conclusion, iron powder has been shown to be an
effective catalyst for the N-demethylation of a number of
tertiary N-methylamines. To the best of our knowledge, this
is the first report of iron(0) effecting a reaction of this type. In
general, these reactions are very high yielding, and the
product is readily separated from the iron residues. Thebaine
(10a) is an important starting material or intermediate in the
synthesis of a number of pharmaceutical opiates, including
opiate antagonists such as naloxone (7) and naltrexone (8)
21,22
as well the mixed agonist-antagonist buprenorphine.
2
0
the anion used. These results suggest that addition of salts
to the iron powder experiment may be useful. To minimize
solubility as a limiting factor, subsequent experiments were
conducted at higher dilution, with 0.25 equiv of added salt.
Interestingly, both CuSO 5H O and ZnSO 7H O en-
4
3
2
4
3
2
hanced the product ratio in favor of the N-nor product 5c
Table 4, entries 2 and 3). Additionally, while added
ZnSO 7H O offered no advantage with respect to reaction
(
4
3
2
rate (cf. entries 1 and 3), all starting material was consumed
within 6 h when CuSO 5H O was added (entry 2). How-
4
3
2
ever, increasing the amount of CuSO 5H O from 0.25 to
4
3
2
(
21) Berenyi, S.; Csutoras, C.; Sipos, A. Curr. Med. Chem. 2009, 16, 3215–
242.
(22) Patel, N. S.; Kilaru, S.; Thennati, R. PCT Patent WO2009/122436,
2009.
3
(20) Ferris, J. P.; Gerwe, R. D.; Gapski, G. R. J. Am. Chem. Soc. 1967, 89,
270–5275.
5
J. Org. Chem. Vol. 75, No. 14, 2010 4809

Kok et al.
JOCArticle
The N-demethylation of thebaine (10a) was previously
to 6-10 with concentrated ammonia before extraction with a
suitable solvent or solvent system.
Method D. The crude reaction mixture was purified via
achieved using both FeSO - and Fe(II)TPPS-mediated
4
Polonovski-type reactions in yields of 74% and 71%, respec-
1
2,14,15
column chromatography on silica gel, eluting with a CHCl3/
tively.
8
The current methodology effects this reaction in
6% yield. Where a direct comparison can be made between
MeOH gradient which isolated the N-nor product from the
tertiary N-methylamine, both obtained as the corresponding
hydrochloride salt.
N-Normorphine (1c). Purified via method D using a gradient
of CHCl /MeOH (49:1-4:1) to give the hydrochloride salt of 1c
as an off-white solid. A small sample of N-normorphine hydro-
chloride was dissolved in H O, and the pH of the solution was
the three methods for the N-demethylation of codeine methyl
ether (9a), the iron(0)-promoted method also afforded a
superior yield of the desired secondary amine. Solvent was
found to play a significant role in the reactivity and selec-
tivity of the reaction, with higher coordinating solvents
delivering a poorer outcome with respect to the yield of
N-nor product. Furthermore, we also found that the coordi-
nation effect of added anions also impacts on the reaction
outcome. Employing dextromethorphan and CME as sub-
strates, it was found that the ratio of secondary to tertiary
amines formed was essentially independent of the stoichio-
metry of the catalyst used. As iron is both inexpensive
and environmentally friendly, this procedure has enormous
potential for large-scale applications such as those used by
the pharmaceutical industry for the preparation of semi-
synthetic opiates.
3
2
adjusted to 8-9 with concentrated ammonia. The precipitate
was isolated via filtration to give N-normorphine as a light
2
8
24
brown solid: mp 272-276 °C (lit. mp 275-277 °C dec); [R]
D
1
-
54 (c 1, 10% HOAc); H NMR (D O/CF CO D, 400 MHz) δ
2 3 2
6
.68-6.55 (m, 2 H), 5.64 (d, J = 9.6 Hz, 1 H), 5.29 (d, J = 9.6
Hz, 1 H), 4.94 (d, J = 5.7 Hz, 1 H), 4.27-4.23 (m, 2 H),
.32-3.22 (m, 1 H), 3.08-2.78 (m, 4 H), 2.15 (ddd, J = 4.2, 13.2
3
13
and 13.2 Hz, 1 H), 2.04 (dd, J = 3.2 and 14.0 Hz, 1 H); C NMR
O/CF CO D, 100 MHz) δ 145.4, 137.8, 133.0, 129.3, 125.7,
23.4, 120.3, 117.4, 90.6, 65.5, 51.5, 42.0, 37.1, 36.7, 31.6, 25.7;
(
D
2
3
2
1
þ
MS (ESI) m/z 272 [M þ H] ; HRMS C H NO calcd for [M þ
1
6
18
3
þ
H] 272.1281, found 272.1286.
N-Noratropine (4c). Purified via method A using a gradient of
CHCl /MeOH/NH OH (90:10:1-85:15:1) to afford a colorless
3
4
Experimental Section
1
oil; H NMR (D O) δ 7.45-7.30 (m, 5 H), 4.98 (t, J = 4.2 Hz,
2
General Methods. Iron powder (NC100.24, 99% Fe) was
purchased from a commercial supplier. All solvents were de-
gassed before use, and reactions were conducted under an
atmosphere of nitrogen. All substrate N-oxides, isolated as the
corresponding hydrochloride salt, were prepared by following
1 H), 4.16 (dd, J = 10.2 and 12.9 Hz, 1 H), 3.94-3.09 (m, 2 H),
1
3
3.51-3.42 (m, 1 H), 3.42-3.33 (m, 1 H), 2.10-1.51 (m, 8 H);
C
3
NMR (CDCl ) δ 172.3, 135.8, 128.8, 128.1, 127.7, 68.6,
63.9, 54.5, 53.1, 53.0, 37.0, 36.7, 28.8, 28.4; ES-MS m/z 276
[M þ H] ; HRMS C16
þ
þ
H
22NO
3
calcd for [M þ H] 276.1594,
1
2-15
the procedure previously described.
Thin-layer chroma-
found 276.1604.
N-Nordextromethorphan (5c). Purified via method A using a
tography (TLC) was performed with 0.25 μm TLC silica gel 60F
aluminum plates with 254 nm fluorescent indicator, and plates
were visualized using both UV light and molybdate stain. Unless
gradient of CHCl
a colorless oil: [R]
/MeOH/NH OH (90:10:1-85:15:1) to afford
4
1
3
2
4
D
þ33 (c 0.82, CHCl
3 3
); H NMR (CDCl ) δ
1
13
otherwise indicated, H and C NMR were recorded at 300 and
7.03 (d, J = 8.4 Hz, 1 H), 6.81 (d, J = 2.6 Hz, 1 H), 6.72 (dd, J =
2.6 and 8.4 Hz, 1 H), 3.22 (s, 3 H), 3.17-3.09 (m, 2 H), 2.82-2.61
(m, 5 H), 2.32 (m, 1 H), 1.83-1.77 (m, 1 H), 1.68-1.57 (m, 2 H),
75 MHz, respectively. Chemical shifts (δ ppm) were referenced
with solvent residual peaks.
1
3
General N-Demethylation Procedure Using Iron Powder. To a
solution of the N-oxide hydrochloride (100 mg) in a solvent such
as MeOH, EtOH, i-PrOH, MeCN, EtOAc, CHCl , or DCM
1.55-1.48 (m, 1 H), 1.44-1.25 (m, 5 H), 1.12-0.99 (m, 1 H);
C
NMR (CDCl , 150 MHz) δ 158.8, 138.9, 129.3, 126.0, 112.0,
3
111.1, 55.1, 51.1, 40.7, 38.3, 37.4, 36.6, 35.4, 27.8, 25.5, 25.4,
21.5; MS (ESI) m/z 258 [M þ H] ; HRMS C17
3
þ
(
typically, 10 mL per 100 mg) was added iron powder (0.1-2.4
H24NO calcd for
þ
molar equiv). The heterogeneous mixture was allowed to stir for
the specified length of time or until complete consumption of
starting material and then filtered through Celite, and the filter
pad was washed with a suitable solvent such as CHCl3 or
[M þ H] 258.1852, found 258.1851.
N-NorCME (9c). Purified via method A using a gradient of
CHCl
/MeOH/NH OH (90:10:1-85:15:1) to afford an off-
4
23 24
3
white solid: mp 97-101 °C (lit. mp 98-100 °C); [R]
D
-170
3 3
(c 0.83, CHCl ); H NMR (CDCl ) δ 6.67 (d, J = 8.0 Hz, 1 H),
1
CHCl /i-PrOH (3:1) (10 mL ꢀ 2). The original filtrate and
3
washings were combined and concentrated to dryness to give a
crude mixture of the N-nor compound and the starting tertiary
N-methylamine. Pure N-nor product was isolated from the
tertiary N-methylamine and/or N-methylamine N-oxide via
method A, B, C, or D.
6.53 (d, J = 8.0 Hz, 1 H), 5.80-5.72 (m, 1 H), 5.30 (ddd, J = 2.4,
2.4, and 9.9 Hz, 1 H), 4.98 (dd, J = 1.2 and 5.7 Hz, 1 H), 3.84 (s,
3 H), 3.83-3.78 (m, 1 H), 3.73-3.67 (m, 1 H), 3.54 (s, 3 H),
1
3
3.08-2.82 (m, 4 H), 2.64-2.59 (m, 1 H), 2.00-1.90 (m, 4 H);
C
NMR (CDCl , 100 MHz) δ 147. 3, 141.8, 130.6, 130.3, 128.5,
3
Method A. To the crude product mixture was added chloro-
126.9, 118.5, 113.0, 89.6, 75.7, 56.9, 56.1, 51.7, 44.1, 41.4, 38.3,
36.5, 31.3; MS (ESI) m/z 300 [M þ H] ; HRMS C18
þ
form (20 mL) or CHCl
washed with either concentrated ammonia in brine (5 mL) or
0% aqueous NaOH (1 mL), dried (Na SO ), filtered, and
concentrated. The remaining residue was purified via column
chromatography on SiO using a CHCl /MeOH/NH OH gra-
3
/i-PrOH (3:1), and the solution was
H22NO
3
þ
calcd for [M þ H] 300.1594, found 300.1586.
1
2
4
N-Northebaine (10c). Purified via method A using a gradient
of CHCl
3 4
/MeOH/NH OH (90:10:1-85:15:1) to afford an off-
24 24
white solid: mp 146-150 °C (lit. mp 157-158 °C); [R]
(c 0.82, 5% MeOH in CHCl
D
-225
-197.9 (c 0.31, 5%
-200 (c 0.1, 5% MeOH in
28
2
3
4
7
b
20
dient (95:5:1-85:15:1) which isolated the N-nor product from
the tertiary N-methylamine.
Method B. To the crude product mixture was added 10%
aqueous NaOH (1 mL), and the mixture was concentrated to
3
) [lit. [R]
D
D
2
5
23
MeOH in CHCl
); lit. [R]
); lit. (þ)-northebaine [R]
H NMR (CDCl ) δ 6.68 (d, J = 8.1 Hz, 1 H), 6.62 (d, J = 8.1 Hz,
3
2
6
CHCl
3
D
þ235.2 (c 0.108, CHCl
3
)];
1
3
dryness. The resulting residue was column purified on SiO
using an ethyl acetate/MeOH/NH OH gradient (140:60:1-
2
(23) Dhar, N. C.; Maehr, R. B.; Masterson, L. A.; Midgley, J. M.;
Stenlake, J. B.; Wastila, W. B. J. Med. Chem. 1996, 39, 556–561.
4
7
N-nor product.
0:30:1) which isolated the tertiary N-methylamine from the
(
(
24) Bartels-Keith, J. R. J. Chem. Soc. 1966, 617–624.
25) Madyastha, K. M.; Reddy, G. V. B.; Nagarajappa, H.; Sridhar,
Method C. To the crude product mixture was added 5%
aqueous HCl (20 mL). The pH of the solution was then adjusted
G. R. Indian J. Chem. 2000, 39B, 377–381.
(26) Rice, K. C.; Newman, A. H. US Patent 5668285, 1997.
4
810 J. Org. Chem. Vol. 75, No. 14, 2010

Kok et al.
JOCArticle
1
1 H), 5.50 (d, J = 6.4 Hz, 1 H), 5.27 (s, 1 H), 5.03 (d, J = 6.4 Hz,
1 H), 3.92 (dd, J = 1.2 and 5.6 Hz, 1 H), 3.86 (s, 3 H), 3.61 (s, 3 H),
3.24-3.07 (m, 3 H), 2.94 (dd, J = 4.3 and 13.7 Hz, 1 H), 1.43 (br,
1 H), 2.08 (ddd, J = 5.1, 12.6, and 12.6 Hz, 1 H), 1.84 (dd, J = 2.6
(c 1.2, 10% HOAc); H NMR (D O/CF CO D) δ 6.95 (d, J =
2 3 2
9.9 Hz, 1 H), 6.77-6.70 (m, 2 H), 6.17 (d, J = 9.9 Hz, 1 H), 4.94
(s, 1 H), 4.05-3.98 (m, 1 H), 3.30-3.16 (m, 3 H), 2.93 (ddd, J =
3.9, 13.2, and 13.2 Hz, 1 H), 2.65 (ddd, J = 5.1, 13.2, and 13.2
13
13
and 12.7 Hz, 1 H); C NMR (CDCl
3
, 100 MHz) δ 152.4, 144.7,
2
Hz, 1 H), 1.86 (dd, J = 3.9 and 14.1 Hz, 1 H); C NMR (D O/
1
5
42.7, 133.9, 133.3, 127.7, 119.1, 112.7, 110.0, 95.7, 89.1, 56.2, 54.8,
3.7, 46.6, 40.5, 38.4, 37.6; MS (ESI) m/z 298 [M þ H] ; HRMS
CF
120.9, 118.6, 85.8, 66.2, 56.2, 46.0, 36.9, 27.2, 24.8; MS (ESI) m/z
3 2
CO D) δ 196.6, 147.5, 142.4, 138.4, 132.7, 128.6, 122.1,
þ
þ
þ
þ
C H NO calcd for [M þ H] 298.1438, found 298.1432.
286 [M þ H] ; HRMS C H NO calcd for [M þ H] 286.1074,
18
20
3
16 16
4
N-Nororipavine (11c). Purified via method D using a gradient
of CHCl /MeOH (24:1-17:3) to give the hydrochloride salt of
found 286.1085.
3
N-Noroxycodone (15c). Purified via method A using a CHCl /
3
2
4
1
1
6
6
1c as an off-white solid: mp >200 °C dec; [R]
0% HOAc); H NMR (D O/CF CO D, 400 MHz) δ 6.71-
-194 (c 0.83,
MeOH/NH OH (90:10:1-85:15:1) gradient to afford 15c as an
D
4
1
off-white solid; a small sample was treated with 10% aqueous
HCl to give the hydrochloride salt, mp >250 °C. 15c: [R]
2
3
2
2
4
.64 (m, 2 H), 5.76 (d, J = 6.8 Hz, 1 H), 5.38 (s, 1 H), 5.06 (d, J =
.8 Hz, 1 H), 4.52 (d, J = 6.6 Hz, 1 H), 3.53 (s, 3 H), 3.37-3.13
D
7
b
20
-123 (c 1.86, 10% HOAc) [lit. [R]
D
-102.7 (c 0.875, MeOH);
2
6
25
1
(
(
m, 4 H), 2.22 (ddd, J = 5.6, 13.4, and 13.4 Hz, 1 H), 1.91-1.85
m, 1 H); C NMR (D
lit. (þ)-noroxycodone [R]
þ100 (c 0.55, MeOH)]; H NMR
D
1
3
2
O/CF
42.8, 138.6, 132.0, 124.4, 124.2, 120.8, 117.3, 117.0, 96.0,
7.7, 55.1, 53.2, 44.6, 37.0, 33.5, 33.1; MS (ESI) m/z 284 [M þ
3
CO
2
D, 100 MHz) δ 153.0,
3
(CDCl ) δ 6.74-6.61 (m, 2 H), 4.66 (s, 1 H), 3.92 (s, 3 H),
3.20-2.93 (m, 4 H), 2.75 (dd, J = 2.7 and 9.3 Hz, 1 H), 2.47-
2.25 (m, 2 H), 1.89 (ddd, J = 3.0, 6.0, and 13.2 Hz, 1 H), 1.80
(br, 2 H), 1.70-1.52 (m, 2 H); C NMR (CDCl , 150 MHz)
δ 208.5, 145.0, 143.0, 129.5, 125.2, 119.3, 114.8, 90.5, 70.2,
57.4, 56.8, 51.1, 37.2, 36.2, 32.9, 31.5, 30.1; MS (ESI) m/z 302
1
8
þ
þ
13
H] ; HRMS C H NO calcd for [M þ H] 284.1281, found
17
18
3
3
2
84.1287.
N-Northevinone (12c). Purified via method A using a gradient
þ
þ
of CHCl /MeOH/NH OH (90:10:1-85:15:1) to afford a pale
[M þ H] ; HRMS C H NO calcd for [M þ H] 302.1387,
3
4
17 20
4
24
27
23
yellow solid: [R]
-234 (c 1.25, CHCl ) [lit. [R]
D
1
)]; H NMR (CDCl
-240 (c
found 302.1396.
3
D
0
.1, CHCl
3
3
) δ 6.65 (d, J = 7.8 Hz, 1 H), 6.55
N-Noroxymorphone (16c). Purified via method C. Thus, 5%
aqueous HCl (20 mL) was added, and the solution was adjusted to
pH 7 with concentrated ammonia and extracted with CHCl3/
i-PrOH (9:1) to isolate oxymorphone. The solution was then
(
d, J = 7.8 Hz, 1 H), 5.93 (d, J = 8.8 Hz, 1 H), 5.56 (d, J = 8.8
Hz, 1 H), 4.56 (s, 1 H), 3.82 (s, 3 H), 3.61 (s, 3 H), 3.54-3.47 (m,
H), 3.18-2.70 (m, 6 H), 2.16 (s, 3 H), 1.90-1.70 (m, 3 H), 1.42
1
(
1
3
dd, J = 6.4 and 12.0 Hz, 1 H); C NMR (CDCl
3
) δ 208.7,
adjusted to pH 8-9 and extracted with CHCl
3
/i-PrOH (3:1) to
24
1
5
47.9, 141.7, 135.7, 133.8, 128.0, 126.1, 119.2, 113.4, 95.6, 80.9,
6.4, 53.4, 52.7, 50.5, 40.1, 42.2, 37.2, 34.3, 33.0, 30.5, 29.4; MS
afford 16c as an off-white solid: mp >300 °C dec; [R]
D
-154 (c
1.2, 10% HOAc); H NMR (D O/CF CO D) δ 6.72-6.67 (m,
1
2
3
2
þ
(
ESI) m/z 368 [M þ H] .
2 H), 4.90 (s, 1 H), 3.77 (dd, J = 2.1 and 4.8 Hz, 1 H), 3.15-3.10
(m, 3 H), 2.89 (ddd, J = 5.1, 14.7, and 14.7 Hz, 1 H), 2.75 (ddd,
J = 3.9, 13.2, and 13.2 Hz, 1 H), 2.54 (ddd, J = 5.1, 13.2, and 13.2
Hz, 1 H), 2.25-2.15 (m, 1 H), 1.90 (ddd, J = 2.7, 4.8, and 14.1 Hz,
N-Noroxycodeinone (13c). Purified via method C. Thus, 5%
aqueous HCl (20 mL) was added, and the solution was adjusted
to pH 7 with concentrated ammonia and extracted with CH Cl
2
2
1
3
(5 mL ꢀ 2) and CHCl
solution was then adjusted to pH 9 and extracted with CHCl
3
(10 mL ꢀ 4) to isolate oxycodeinone. The
2 3 2
1 H), 1.65-1.54 (m, 2 H); C NMR (D O/CF CO D) δ 211.7,
3
/
143.0, 138.6, 127.3, 122.0, 120.9, 118.5, 89.3, 69.4, 57.4, 49.2, 36.7,
34.4, 30.3, 27.5, 25.7; MS (ESI) m/z 288 [M þ H] ; HRMS
24
þ
i-PrOH (3:1) to afford 13c as an off-white solid: [R]
-112
D
1
þ
(
c 1.2, 10% HOAc); H NMR of (D O/CF CO D) δ 6.93 (d,
C H NO calcd for [M þ H] 288.1230, found 288.1226.
2
3
2
16 18
4
J = 10.2 Hz, 1 H), 6.88-6.77 (m, 2 H), 6.14 (d, J = 10.2 Hz,
H), 4.94 (s, 1 H), 4.02-3.98 (m, 1 H), 3.74 (s, 3 H), 3.26-3.15
m, 3 H), 2.90 (ddd, J = 4.0, 13.5, and 13.5 Hz, 1 H), 2.64 (ddd,
J = 5.1, 13.5, and 13.5 Hz, 1 H), 1.82 (dd, J = 4.0 and 13.5 Hz,
N-Nortropine (17c). Purified via method B to give 17c as an
1
off-white solid: H NMR (D
1
(
2
O) δ 4.27-4.20 (m, 1 H),
4.20-4.10 (m, 1 H), 2.55-2.42 (m, 2 H), 2.36-2.30 (m, 1 H),
1
3
2
2.30-2.22 (m, 1 H), 2.22-2.02 (m, 4 H); C NMR (D O, 150
1
3
1
1
2
H); C NMR (D
2
O/CF
3
CO
2
D) δ 196.8, 147.4, 143.1, 142.2,
32.7, 128.2, 122.7, 120.8, 115.2, 85.8, 66.1, 56.2, 56.1, 45.8, 36.8,
MHz) δ 62.1, 61.9, 54.1, 54.0, 35.1, 25.5; MS (ESI) m/z 128 [M þ
þ
þ
H] ; HRMS C
128.1065. A small sample of 17c was dissolved in H O, and 10%
7
H
13NO calcd for [M þ H] 128.1070, found
þ
7.1, 24.8; MS (ESI) m/z 300 [M þ H] ; HRMS C H NO
1
7
18
4
2
þ
calcd for [M þ H] 300.1230, found 300.1240.
aqueous HCl was added to pH 2. The resulting solution was
concentrated to dryness to give N-nortropine hydrochloride as a
N-Noroxymorphinone (14c). Purified via method C. Thus, 5%
aqueous HCl (20 mL) was added, and the solution was adjusted
to pH 7 with concentrated ammonia and extracted with CHCl3/
i-PrOH (100:1) to isolate oxymorphinone. The solution was
4
b
white solid, mp 281-285 °C dec (lit. mp 280-282 °C dec).
Acknowledgment. We thank the ARC Centre for Free
Radical Chemistry and Biotechnology for financial support.
then adjusted to pH 9 and extracted with CHCl
afford 14c as an off-white solid: mp >230 °C dec; [R]
3
/i-PrOH (3:1) to
2
4
-109
D
1
Supporting Information Available: Copies of H and
13
C
NMR spectra for all N-nor compounds. This material is avail-
able free of charge via the Internet at http://pubs.acs.org.
(
27) Madyastha, K. M.; Reddy, G. V. B. J. Chem. Soc., Perkin Trans. 1
994, 911–912.
28) Rice, K. C. J. Org. Chem. 1975, 40, 1850–1851.
1
(
J. Org. Chem. Vol. 75, No. 14, 2010 4811