Utility of iron nanoparticles and a solution-phase iron species for the: N-demethylation of alkaloids

Article abstract of DOI:10.1039/c7gc00436b

The N-demethylation of selected N-methylalkaloids using a modified Polonovski reaction can be accomplished using a novel green methodology employing nanoscale zero-valent iron, nZVI, in isopropanol. Use of nZVI promotes a much faster conversion to N-demethylated products due to much higher surface area on the metal surface as shown by SEM analysis. Rates of conversion can be further enhanced using catalytic quantities of the solubilised iron(0) species triiron dodecacarbonyl, Fe₃(CO)₁₂.

Full text of DOI:10.1039/c7gc00436b

View Article Online
View Journal
Green
Chemistry
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: J. K. Awalt, R.
Lam, B. Kellam, B. Graham, P. J. Scammells and R. D. Singer, Green Chem., 2017, DOI:
10.1039/C7GC00436B.
This is an Accepted Manuscript, which has been through the
Volume 18 Number
7 7 April 2016 Pages 1821–2242
Royal Society of Chemistry peer review process and has been
accepted for publication.
Green
Chemistry
Accepted Manuscripts are published online shortly after
Cutting-edge research for a greener sustainable future
www.rsc.org/greenchem
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
author guidelines.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the ethical guidelines, outlined
in our author and reviewer resource centre, still apply. In no
event shall the Royal Society of Chemistry be held responsible
for any errors or omissions in this Accepted Manuscript or any
consequences arising from the use of any information it contains.
ISSN 1463-9262
CRITICAL REVIEW
G. Chatel et al.
Heterogeneous catalytic oxidation for lignin valorization into valuable
chemicals: what results? What limitations? What trends?
rsc.li/green-chem

Please do not adjust margins
Page 1 of 8
Green Chemistry
View Article Online
DOI: 10.1039/C7GC00436B
Green Chemistry
ARTICLE
Utility of iron nanoparticles and a solution-phase iron species for
the N-demethylation of alkaloids
Jon Kyle Awalt,a,‡ Raymond Lam,b,c,‡ Barrie Kellam,c Bim Graham,b Peter J. Scammellsb,* and
Received 00th February 2017,
Accepted 00th January 20xx
Robert D. Singera,*
DOI: 10.1039/x0xx00000x
The N-demethylation of selected N-methylalkaloids using a modified Polonovski reaction can be accomplished using a
novel green methodology employing nanoscale zero-valent iron, nZVI, in isopropanol. Use of nZVI promotes a much faster
conversion to N-demethylated products due to much higher surface area on the metal surface as shown by SEM analysis.
Rates of conversion can be further enhanced using catalytic quantities of the solubilised iron(0) species triiron
www.rsc.org/
dodecacarbonyl, Fe3(CO)12
.
industrial scale. The most commonly used methods for the N-
demethylation of tertiary N-methylamines involved the use of
cyanogen bromide (von Braun reaction)6 and chloroformate
reagents.7 Both of these methods have significant drawbacks;
cyanogen bromide is a highly toxic reagent,8 whilst the more
effective chloroformate reagents, such as vinyl chloroformate,
are quite expensive. In both cases, the resulting product must
be further processed in order to return the N-nor product; in
the case of the von Braun reaction to hydrolyse the resulting
cyanamide group, and with the Polonovski reaction to cleave
the resulting carbamate.
The N-demethylation of a range of alkaloids has also been
achieved using the non-classical Polonovski reaction.
Galanthamine was N-demethylated in good yield by first
oxidising this alkaloid with mCPBA and subsequently treating
the resultant N-methylamine oxide with ferrous sulfate.9 Iron
catalysts have also been successfully investigated and
developed into efficient catalysts for the N-demethylation of
opiate alkaloids utilizing what has been proposed as a non-
classical Polonovski-type mechanism.10 Further development of
this work showed that the use of solid iron powder as the
catalyst allowed its easy removal from the reaction mixture via
filtration.11
Introduction
The opiate alkaloids morphine and codeine, which are extracted
from the opium poppy, Papaver somniferum, are common
analgesics that have been used for thousands of years for the
treatment and control of pain.1 In addition to being used
clinically as therapeutic agents in their own right, they can be
used as starting materials in the production of semi-synthetic
opioid pharmaceuticals, such as naltrexone, naloxone and
buprenorphine.2 These particular opioids are useful as
therapeutic agents for the treatment of opiate or alcohol
dependence, opiate overdose and in pain management.
Specifically, naloxone hydrochloride is used to reverse the acute
symptoms of opioid overdose and studies have found it to be a
cost-effective strategy.3 The rate of drug overdose deaths
increased by 140% between the years 2000 and 2014, an
increase driven by opioid overdose and, since 2013, in large part
due to synthetic opioids including fentanyl.4 A key feature in the
synthesis of semi-synthetic opioid pharmaceuticals is the loss of
the N-methyl group, which is replaced by a variety of alkyl
groups, such as a cyclopropylmethyl group in naltrexone and
buprenorphine, or an allyl group in naloxone (Figure 1).
Inclusion of a non-methyl N-alkyl group is rationalized by the
“message-address” concept, where varying this “message”
group typically allows control of agonist or antagonist activity.5
Despite the medical utility of these semi-synthetic opioids,
N-demethylation has traditionally been a difficult step on the
a.Atlantic Centre for Green Chemistry, Department of Chemistry, Saint Mary’s
University, Halifax, Canada B3H 3C3.
b.Medicinal Chemistry, Monash Institute of Pharmaceutical Sciences, Monash
University, Parkville, Victoria 3052, Australia.
Fig. 1. Semi-synthetic opioid pharmaceuticals used for
treatment of opioid overdose.
In addition to this, the reaction returns the free N-nor
compound without the need for further reaction, therefore
avoiding additional steps. This system has been subsequently
c. School of Pharmacy, Centre for Biomolecular Sciences, University of Nottingham,
Nottingham NG7 2RD, U.K.
‡
J.A. and R.L. contributed equally to this work.
Electronic Supplementary Information (ESI) available: synthetic procedures. See
DOI: 10.1039/x0xx00000x
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 1
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 2 of 8
ARTICLE
Journal Name
adapted by Nakano et al. in a flow process using iron/sand Reaction kinetics, thermodynamics and yield may be affected
View Article Online
DOI: 10.1039/C7GC00436B
columns, again demonstrating the utility of using these by the balance of solvent polarity and protic nature. More
catalysts.12
specifically in the case of N-demethylations, molecular oxygen
The optimised process requires the formation of the alkaloid dissolved in the solvent can also impact on reaction rate and
N-oxide, which can be readily accomplished using a variety of yield.11 The work previously conducted by our group has
oxidants, such as hydrogen peroxide or mCPBA on a laboratory extensively explored various iron powders, iron salts, as well as
scale.10 This is then converted to the hydrochloride salt, which steel alloy powders under a variety of conditions.10,11,14,15,17,18 In
was previously found to be an important factor, as the the present study, we present the use of two new iron species
additional proton is essential in the proposed mechanism.10 as alternatives for the N-demethylation of alkaloids whilst also
Furthermore, the anion used in the salt formation is also exploring the use of “greener” solvents for this reaction.
important, as departure from the chloride anion results in a
Nanoscale zero-valent iron (nZVI) has been prepared using a
reduction in yield.13 While the reaction will still proceed using variety of methods and employed in several different types of
the free N-oxide, the yields obtained are suboptimal.14 Solvent reactions.19,20,21 The majority of literature reports involving nZVI
also plays an important role, with chloroform (CHCl3) giving the describe the remediation of ground water and the removal of
highest yield of the solvents tested.15
contaminants such as arsenic(III),22 chromium(VI),23 and E. coli
The precise nature of the catalytic system is also important. bacteria.24 There are a number of reports where nZVI has been
The optimised loading is described as 13 mol% Fe(0) powder used in catalytic applications.21,25,26 The efficacy of these
with an additional 2 mol% Fe3+ in the form of FeCl3.11 It was methods is due in part to the extremely high surface area-to-
demonstrated that while increasing the loading of Fe(0) has volume ratio of nZVI. This property, along with our experiences
little impact on the yield, increasing the loading of Fe3+, or even using iron(0) as a catalyst in the modified Polonovski reaction,
switching to an Fe2+ species, results in suboptimal yields. Finally, prompted us to investigate its use for N-demethylation
allowing the reaction to proceed in open air increases the reactions. A number of “green” advantages are also presented
reaction rate, while under an inert nitrogen atmosphere the through the use of nZVI in these systems. The high surface area
reaction rate is adversely affected.11 All these factors may be of nZVI was anticipated to lead to faster reaction rates for N-
implemented in an industrial setting for cost-effective, demethylation of the N-oxides used in these reactions.
relatively safe, and efficient N-demethylation of opiates for Furthermore, the preparation of nZVI can be accomplished in
conversion to other clinically useful drugs (Scheme 1).
water using minimally toxic sodium borohydride (NaBH4) and
While previously described advances have been extremely subsequent catalytic demethylations can be conducted in
useful, we have a continued interest in the development of iron alcoholic solvents such as isopropanol (i-PrOH).
catalysts in N-demethylations, either for industrial or synthetic
A number of procedures have been reported for the
purposes. While the Fe(0)/Fe(III) system we described earlier is preparation of nZVI. The most common technique involves
efficient, both in terms of yield and cost, the use of CHCl3 is reduction of an iron(II) or iron(III) salt, such as iron(II) sulfate or
undesirable as it is toxic and a potential carcinogen.16 iron(III) chloride, to iron(0) using NaBH4 in aqueous media. This
Furthermore, it is potentially fatal if absorbed or ingested in procedure is relatively non-toxic, inexpensive, uncomplicated
larger quantities.
and can yield nanoscale iron(0) particles in the size range of 10–
50 nm.27 The effects of iron and NaBH4 concentration,
atmospheric versus inert conditions, the rate of agitation during
reaction, the rate of NaBH4 addition, and how the iron is dried
and stored have been extensively studied and optimised.
Triiron dodecacarbonyl (Fe3(CO)12) is an unusual iron
species, typically used as a source of reactive Fe(0) for the
synthesis of various iron complexes, or as a reagent. Its utility as
a catalyst has, to our knowledge, not been previously described.
Previously investigated catalysts tend to be heterogeneous due
to the insolubility of iron and iron salts in the tested solvents. In
contrast to these, Fe3(CO)12 is soluble in non-polar organic
solvents, with limited solubility in more polar solvents.
Furthermore, the carbonyl ligands are easily displaced by Lewis
bases, potentially allowing for better interaction with the
substrate and hence higher rates of reaction.
Synthesis of iron nanoparticles can also be achieved using
Fe3(CO)12. Previous examples demonstrate that under relatively
harsh conditions, Fe3(CO)12 may undergo controlled
decomposition to form either nZVI or magnetite
nanoparticles.28,29 A solution of Fe3(CO)12 and PPh3 in hexane
may form graphite coated iron nanoparticles under strong UV
irradiation over an extended period of time.28 Fe3(CO)12 may
Scheme 1. N-Demethylation of DXM, 4a
, via N-oxide
hydrochloride, 4b
.
Therefore, its use on an industrial scale is highly undesirable.
Seeking a solvent replacement is not always a simple task.
2 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 3 of 8
Green Chemistry
Please do not adjust margins
Green Chemistry
ARTICLE
form magnetite nanoparticles upon prolonged reflux in
diethylene glycol, diethyl ether, and oleic acid.29 Both of these
methods result in nanoparticles that have an organic coating. As
we are interested in “bare” nZVI, these methods were not
utilised.
View Article Online
DOI: 10.1039/C7GC00436B
Industrial adoption of any new method for the N-
demethylation of alkaloids depends on appropriate selection of
a green solvent. The environmental impact and waste involved
in the synthesis of active pharmaceutical ingredients (API’s) is
of considerable concern to the pharmaceutical industry.30 With
solvents contributing 80–90% of the non-aqueous waste
generated, reduction in toxic waste solvent is critical to
reducing health and environmental damage. Solvents,
nonetheless, are still required for the production of API’s, and Fig. 2. SEM image of synthesised nZVI.
therefore need to be replaced with greener options. To that
end, GSK have assembled a toolkit to assist in the selection of
The pseudo-opioid dextromethorphan (DXM), 4a, was
solvents suitable for large-scale industrial applications.30 It has chosen as an ideal substrate for preliminary reactions
also been suggested that initial reaction development could be investigating optimisation of reaction conditions. DXM,
guided by this solvent selection process, as this would save both contained in many common cough medications due to its anti-
time and money further down the line when optimising such tussive properties, is structurally similar to naturally occurring
reactions for industrial purposes. With this in mind, we have opiates possessing an N-methyl group, and is a member of the
chosen to approach the problem primarily using greener morphinan family of compounds. Unlike naturally occurring
options.
opiates, it is not a highly controlled substance and is hence
Herein, we report the use of nZVI and Fe3(CO)12 as Fe(0) readily available as its hydrobromide salt from commercial
sources for the modified Polonovski reaction in which the N- sources. DXM hydrobromide is an atmospherically stable solid
demethylation of various alkaloids in green solvents is at room temperature making it easy to manipulate in a
demonstrated. The reaction systems have been optimised for laboratory setting. DXM is easily transformed into its N-oxide
the use of i-PrOH, a green solvent that is directly compared to and N-oxide hydrochloride derivative, 4b, in high yields. N-
the use of the “non-green” solvent CHCl3 as well as other Oxidation of 4a and isolation of the hydrochloride salt 4b was
solvent combinations. The use of nZVI represents an improved achieved by treatment of 4a with either mCPBA or H2O2,
green approach to this reaction in that the rate of reaction is followed by addition of aqueous HCl using a variation of a
significantly enhanced versus iron dust while maintaining ease literature procedure.14 In general, mCPBA in CHCl3 was
of isolation of products and the use of a green iron(0) catalyst. employed as the oxidant as it afforded the desired N-oxide after
Fe3(CO)12, although less green than using nZVI, has favourable a short reaction time (~ 1 h). However, H2O2 in MeOH has
solubility properties, allowing for catalytic amounts to be previously been used as a greener alternative for this
10,13,18
employed in conjunction with the use of i-PrOH. The extension transformation
.
Despite this, literature procedures tend to
of the methodology to a selected array of N-methylalkaloids is vary significantly in equivalents added and the reaction time
also reported.
required. Initial screening using titrated H2O2 showed that 10 eq. was
sufficient to afford conversion of 4a to 4b overnight. Excess H2O2 was
then degraded to water using minimal MnO2. After filtration and
removal of the organic solvent in vacuo, the HCl salt was
subsequently generated by adding dilute acid and removing the
excess water using a freeze-dryer. Increasing the molar ratio of H2O2
to 20 eq. and 30 eq. offered no significant advantage in terms of
reaction time. Overall, this procedure gave comparable yields of the
N-oxide and eliminated the need for chlorinated solvents, replacing
them with MeOH and water.
The N-oxide 4b was subsequently converted to
corresponding N-nor derivative 4c. Typically, 4b was dissolved
or suspended in the selected solvent system and nZVI or
Fe3(CO)12 was then added as a solution/suspension in the
appropriate solvent and the reaction was stirred at room
temperature for the prescribed period of time. The results of
these experiments are summarised in Tables 1 and 2. Reaction
of 4b in i-PrOH gave a good yield of the N-demethylated
product, 4c, using a stoichiometric amount of nZVI (Table 1,
Entry 1). Use of this green solvent afforded slightly lower yields
Results and discussion
nZVI was prepared using a previously reported procedure in
which ferric sulfate heptahydrate was reduced in aqueous
medium using NaBH4 with mechanical stirring, washed with
50% aqueous methanol via repetitive centrifugation,
decantation and replacement of solvent.27 The isolated nZVI
was then suspended in the reaction solvent to be used for the
modified Polonovski reactions. nZVI prepared using this method
could be carefully isolated under inert atmosphere to prevent
oxidation, as far as possible, followed by characterisation using
scanning electron microscopy (SEM) (Figure 2). The SEM image
shows spherical, monodisperse particles of approximately 75
nm in diameter, forming chain-like secondary structures.
Comparison with the SEM of iron dust used in previous studies
shows distinctly smaller particle size (i.e. 75 nm vs. 40 m) and
therefore significantly higher surface area of the iron(0) surface
upon which reaction can occur.
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 3
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 4 of 8
ARTICLE
Journal Name
than when CHCl3 was used (Table 1, Entry 2), likely due to the addition of MeOH to increase the solubility of tVhieewsAurtbicsletOrantlinee.
DOI: 10.1039/C7GC00436B
differences in solubility of the substrate molecules. Switching Ester solvents (Table 2, Entries 4, 6, 7 and 8) afforded generally
the solvent to methanol resulted in a noticeable decrease in good results for the reaction with 4b. Dimethyl carbonate
yield of 4c (Table 1, Entry 3). It has previously been postulated (DMC) (Table 2, Entry 5) proved to be the exception, with very
that iron acts as a catalyst in the modified Polonovski reaction.10 low recovery of the product. Although DMC is known to be a
When a catalytic amount of nZVI (i.e. 10 mol%) was employed, methylating agent, its ability to alkylate is only evident under
there was a pronounced decrease in the yield of the N- relatively harsh conditions. Therefore, we suspect that
demethylated product, 4c (Table 1, Entries 4 – 6), regardless of methylation of 4c under the tested conditions is unlikely to
the solvent used. We believe this is due to catalyst occur. Rather, this solvent may simply not be amenable to this
poisoning/deactivation as the N-demethylated product builds particular reaction. The effect of substituting i-PrOH for MeOH
up in the reaction. The coordination of the Lewis basic product, in these solvent mixtures was also explored. EtOAc/i-PrOH
4c
, to iron prevents further substrate molecules from (Table 2, Entry 8) gave higher yields compared to the
interacting with the iron catalyst. Conversion of 4b to 4c was EtOAc/MeOH mixture (Table 2, Entry 7). Indeed, when used as
unsuccessful when nZVI were employed using water as a the bulk solvent, i-PrOH gave surprisingly good results (Table 2,
solvent.
Entry 3). Although reaction times were longer, the yields
obtained were improved.
Table 1. N-Demethylation of various alkaloid N-oxide
hydrochlorides using nZVI.
Table 2. N-Demethylation of various alkaloid N-oxide
hydrochlorides using Fe3(CO)12.
Entry
1
nZVI
(mol %)
100
Substrate (N-
oxide)
DXM (4b)
Solvent /
ConditionsA
i-PrOH
Time (h)
% yieldB
(N-nor)
4c, 84
Entry
1
Fe3(CO)12
(mol %)
5
Substrate (N-
oxide)
DXM (4b)
Solvent /
ConditionsA
CHCl3
Time
(h)
0.1
% yieldB
(N-nor)
4c, 86
0.75
1
2
100
100
10
DXM (4b)
DXM (4b)
CHCl3
MeOH
i-PrOH
CHCl3
4c, 92
4c, 59
4c, 78
4c, 67
2
10
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
DXM (4b)
DXM (4b)
CHCl3
0.1
0.7
0.1
0.1
0.1
0.1
0.1
24
4c, 73
4c, 92
4c, 76
4c, 15
4c, 83
4c, 74
4c, 81
5c, 28
5c, 26
5c, 30
3
3
3
i-PrOH
4
DXM (4b)
21
1
4
DXM (4b)
MeOAc/MeOH
(9:1)
DMC/MeOH
(9:1)
5
10
DXM (4b)
5
DXM (4b)
6
10
DXM (4b)
MeOH
i-PrOH
i-PrOH
i-PrOH
i-PrOH
i-PrOH
3
4c, 34
6
DXM (4b)
tBuOAc/MeOH
_C
7
100
100
100
100
100
Noscapine (5b
)
24
3
(9:1)
EtOAc/MeOH
(9:1)
7
DXM (4b)
8
Atropine (6b
)
6c, 85
8
DXM (4b)
EtOAc/i-PrOH
C
9
Tropine (7b
)
24
14
14
-
(9:1)
CHCl3
9
Noscapine (5b
Noscapine (5b
Noscapine (5b
Noscapine (5b
)
)
)
)
10
11
Pivaloyltropine
8b
Benzoyltropine
9b
8c, 59
9c, 58
(
)
10
11
12
13
14
15
16
17
18
19
EtOAc/MeOH
(9:1)
EtOAc/i-PrOH
(9:1)
17
(
)
2
A 10 mL of solvent per 100 mg substrate at ambient temperature (measured to be
21oC on average) under an atmosphere of argon. Reactions were stirred using a
12 mm magnetic bead at 900 RPM.
C
i-PrOH
120
22
-
B Isolated yield following column chromatography.
Atropine (6b
)
)
EtOAc/MeOH
(9:1)
EtOAc/i-PrOH
(9:1)
6c, 66
6c, 79
C NMR of crude product mixture showed no evidence of any conversion of N-oxide
starting material to N-nor product.
Atropine (6b
19
C
Tropine (7b
)
i-PrOH
2.5
2
-
Fe3(CO)12 proved to be an efficient catalyst for the N-
demethylation of 4b. In CHCl3, good yields were achievable with
low catalyst loadings and drastically reduced reaction times. As
seen in Table 2, the optimal yield was obtained using 5 mol%
Fe3(CO)12 (Table 2, Entry 1), with increases in catalyst loading
resulting in slightly reduced yield. When less than 5 mol% of
Fe3(CO)12 is used, the reaction never goes to completion, even
if left for extended periods of time. We again rationalise this to
be due to poisoning of the catalyst. We suspect that 4c may be
Pivaloyltropine
8b
Pivaloyltropine
8b
Pivaloyltropine
8b
Pivaloyltropine
8b
CHCl3
8c, 83
8c, 63
8c, 66
8c, 63
(
)
i-PrOH
17
(
)
EtOAc/MeOH
(9:1)
EtOAc/i-PrOH
(9:1)
0.1
0.1
(
)
(
)
A 10 mL of solvent per 100 mg substrate at ambient temperature (measured to be
21oC on average) under an atmosphere of argon. Reactions were stirred using a 12
mm magnetic bead at 750 RPM.
forming stronger coordination bonds to iron than the native B Isolated yield following column chromatography.
C NMR of crude product mixture showed no evidence of any conversion of N-oxide
carbonyl ligand at the conclusion of the catalytic cycle.
starting material to N-nor product.
Following optimisation of the reaction in CHCl3, we
proceeded to assess various greener solvents in order to
ascertain the ability of Fe3(CO)12 to perform under alternative
conditions. To this end, various ester solvents were tested, with
Following the demonstration of nZVI and Fe3(CO)12 as
effective catalysts for the demethylation of 4b, we sought to
establish whether the methodology could be applied to other
4 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 5 of 8
Green Chemistry
Please do not adjust margins
Green Chemistry
ARTICLE
biologically interesting alkaloids. We therefore investigated the
demethylation of the N-oxide hydrochlorides of noscapine (5a),
View Article Online
DOI: 10.1039/C7GC00436B
atropine
(6a), tropine
(7a), pivaloyltropine
(8a), and
benzoyltropine (9a) (Figure 3) using nZVI and Fe3(CO)12
.
Compound 5a is interesting as a potential starting material for
synthesis. Its complex array of carbocycles and heterocycles
makes it an attractive starting point for more complex products.
Furthermore, 5a has gained interest as a starting material for
the synthesis of potential anti-cancer agents.31,32 Previous
studies in our group have already indicated the potential use of
N-substituted noscapine analogues as more potent anti-cancer
agents, which obviously necessitates the preparation of N-
nornoscapine, 5c, as an intermediate.31 Compound 6a, like
many opioids, has been a historically important drug, present in
various plant sources, particularly Atropa belladonna, from
which the drug gets its name. It is also used as a starting
material for various pharmaceuticals; N-noratropine, 6c, is an
important intermediate for the synthesis of ipratropium
Fig. 3. Alternative alkaloid substrates for N-demethylations.
bromide,
a
bronchodilator.33 Improved N-demethylation
methods for both 5a and 6a could be utilized for rapid and easy
access to the N-nor compounds. Compound 7a was tested as an
interesting synthon that may be included in various drugs (such
as Granisetron, Hyoscyamine and Tropisetron), while 8a and 9a
represent more “drug-like” compounds with increased organic
solubility.
The more soluble 8b gave variable results compared to those
previously obtained using nZVI or Fe3(CO)12. Whereas previous
substrates gave good yields in a mixture of 9:1 (v/v) EtOAc/i-
PrOH, conversion to 8c was suboptimal in this system compared
to in CHCl3 (Table 2, Entry 16). Switching to 9b did not result in
any appreciable increase in yield when using nZVI (Table 1, Entry
11). The yields are, nevertheless, comparable to those obtained
for previous attempts at N-demethylation of tropane alkaloids,
but with a significantly reduced reaction time.33 Although some
optimisation may be required, these results highlight the
possibility of using greener solvent substitutes without a
significant compromise in yield.
Reactions using nZVI and Fe3(CO)12 were observed to
proceed much faster in CHCl3 than in greener solvents such as i-
PrOH. Furthermore, these two iron sources seemed to
considerably enhance the rate of the N-demethylation
reactions to a far greater degree than other iron sources
investigated earlier by our research group. For example, N-
demethylation of 4b appeared complete within 5 min when
conducted using Fe3(CO)12, and within 1 h when using nZVI in
CHCl3. Hence, we endeavoured to measure the relative rates of
conversion for each of the iron sources. In order to make direct
comparisons and also to emphasize the compatibility of our
systems in green solvent systems, the study of relative
conversions of 4b to 4c, using the different iron catalysts, were
conducted in i-PrOH. Ferric sulfate did not give 100% conversion
into nor-DXM, 4c, even after 30 h (Figure 4). As we reported
earlier, the use of iron(0) dust improved this reaction and
resulted in essentially 100% conversion after almost 30 h.15
N-Demethylation of these alternative test substrates gave
variable, but promising results (Table 1, Entries 7 – 11, Table 2,
Entries 9 – 19). Compound 5b, for example, failed to give
increased yields of 5c when Fe3(CO)12 was used as the catalyst
compared to previously established literature methods.31
However, the reaction did go to completion (Table 2, Entries 9-
11). Using i-PrOH, no product was observed when either nZVI or
Fe3(CO)12 were used (Table 1, Entry 7 and Table 2, Entry 12).
Once again, we believe this to be due to the limited solubility of
the substrate in the selected solvents. Furthermore, 5a may be
a particularly difficult substrate for this particular process as the
N-methyl group is quite sterically hindered. Upon oxidation to
5b, further steric hindrance may occur, preventing approach of
the catalyst to the required site for the radical oxidation to
occur.
Using 6b as a test substrate, very good yields were obtained
when nZVI was used in i-PrOH (Table 1, Entry 8), while only
moderate yields were observed when using Fe3(CO)12 in
EtOAc/i-PrOH (Table 2, Entry 13 & 14). Lower yields were
observed when using MeOH instead of i-PrOH in the mixture,
which is consistent with previous results. Conversion of 7b to N-
nortropine proved unsuccessful under all tested conditions,
including using i-PrOH as the solvent (Table 1, Entry 9 & Table 2,
Entry 15).
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 5
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 6 of 8
ARTICLE
Journal Name
according to Gottlieb et al.34 Alkaloids were purchased from the
View Article Online
DOI: 10.1039/C7GC00436B
supplier and used as provided (Sigma-Aldrich, Alfa Aesar),
except for dextromethorphan, which was washed with base (2
x 5 % aqueous NaOH) to return the free base before N-oxidation
according to literature procedure.14 Triirondodecacarbonyl and
ferric sulfate heptahydrate were purchased from Sigma-Aldrich.
General procedure for N-oxidation of alkaloids using mCPBA.
Alkaloid N-oxide hydrochloride salts were generated using
variations of a literature procedure.14 The alkaloid free amine
was dissolved in CHCl3 and cooled to (-5–0° C, typically 20 mL
per gram of alkaloid). mCPBA (typically 1.2 eq.) was then added
in a single portion and the reaction allowed to stir until reaction
completion as determined by TLC analysis, after which time the
mixture was worked up using one of the following procedures:
A. The volume of CHCl3 was doubled and i-PrOH was added to
make the solution a 3:1 (v/v) mixture of CHCl3/i-PrOH. This was
then washed successively with 1:4 (v/v) 2 M NaOH/brine (5 × 5
mL), followed by 1:4 (v/v) 2 M HCl/brine (5 × 5 mL). The organic
phase was dried over anhydrous Na2SO4, filtered, and the
solvent removed in vacuo to give the alkaloid N-oxide
hydrochloride salt.
Fig. 4. Percentage conversion versus time for various iron
catalysts. FeSO4 (), Fe(0) dust (), nZVI (), Fe3(CO)12 ().
As anticipated, the use of nZVI, with far greater surface area,
afforded 100% conversion after only 40 min. Lastly, the use of
Fe3(CO)12 afforded the fastest rate of conversion for all iron
sources trialled, with 100% conversion of 4b to 4c in just over
30 min.
B. The reaction mixture was extracted with 1 M HCl (3 × 10 mL).
The combined aqueous extracts were washed with CHCl3 (2 × 10
mL) and the resulting aqueous phase reduced in vacuo to give
the alkaloid N-oxide hydrochloride salt.
Procedure for N-oxidation of DXM using H2O2
DXM free amine (366 mg, 1.35 mmol, 1 eq.) was dissolved
in MeOH (5 mL) and cooled to 0oC. H2O2 (26% w/w in H2O, 1.766
g, 13.5 mmol, 10 eq.) was then added dropwise and the reaction
allowed to stir at room temperature for 18 h. The reaction was
then diluted with MeOH and excess H2O2 deactivated with
MnO2. The reaction was filtered through Celite and the solvent
removed in vacuo. Aqueous HCl (0.1 M, 50 mL) was then added,
and the resulting solution freeze-dried to give 4b as the
hydrochloride salt (412 mg, 94%).
Conclusions
The green attributes of the modified Polonovski reaction,
through which the N-demethylation of various alkaloids is
achieved, have been significantly improved in comparison to
previously reported methodologies. The use of nZVI as a green
catalyst and i-PrOH as a green solvent afforded enhanced rates
of conversion to nor-alkaloid products. Although less green, the
use of Fe3(CO)12 as a catalyst affords even faster rates of
conversion to demethylated, nor-alkaloid products. The
solubility properties of this compound allow for genuinely
catalytic amounts to be employed, while still permitting the use
of i-PrOH. This methodology has been successfully applied to a
number of N-methylated alkaloids, with the possibility of being
further extended to the N-demethylation of opiates of social
significance, used in the commercial preparation of naltrexone,
naloxone and buprenorphine.
General procedure for N-demethylation of alkaloid N-oxides using
nZVI.
Following a modified procedure,27 an appropriate amount of
FeSO4
(two parts deionized water, one part methanol, 8 mL per 0.50
mmol FeSO4 7H2O) and stirred at 450 RPM using a mechanical
⋅7H2O (see Table 1) was dissolved in aqueous methanol
⋅
stirrer in an appropriately sized round-bottom flask while a 0.12
M aqueous solution of NaBH4 was added dropwise at a rate
of approximately 1 drop every 5 sec (0.8 mL per 0.50 mmol
Experimental
FeSO4⋅7H2O). As drops were added, the solution turned an
All solvents were used as provided by the manufacturer (Fisher
Scientific, Alfa Aesar, Sigma-Aldrich). TLC was performed using
aluminium-backed silica plates (Merk, Silica gel 60Å f254), with
compounds visualised using a UV lamp, or via polymolybdate or
KMnO4 staining. 1H and 13C NMR spectra were recorded at 400
and 101 MHz or 300 and 75 MHz (Bruker Avance Nanobay III
400 MHz Ultrashield Plus or Bruker Avance III HD), respectively.
Chemical shifts were referenced to residual solvent peaks
increasingly darker green until finally a black precipitate was
formed. The suspension of black precipitate was stirred for an
additional 5 min before being transferred into a 15 mL
centrifuge tube. The washing process next consisted of five
repetitions of centrifugation (5 min each at 3200 RPM) and
decantation of the supernatant fluid. Before each
centrifugation, the tubes were shaken vigorously such that the
iron particles were re-suspended in the solvent. After the initial
6 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 7 of 8
Green Chemistry
Please do not adjust margins
Green Chemistry
ARTICLE
3
4
5
A. K. Clark, C. M. Wilder and E. L. Winstanley, J.VAiedwdAircttic.leMOenldin.e,
2014, , 153–63.
R. A. Rudd, N. Aleshire, J. E. Zibbell and R. M. Gladden, Morb.
Mortal. Wkly. Rep., 2016, 64, 1378-1382.
P. S. Portoghese, M. Sultana, H. Nagase and A. E. Takemori, J.
Med. Chem., 1988, 31, 1986–1987.
J. von Braun, Ber. Dtsch. Chem. Ges. 1900, 33, 1438-1452.
J. H. Cooley and J. E. Evain, Synthesis, 1989, 1-7.
W. E. Luttrell, J. Chem. Health Saf., 2009, 16, 29-30.
A. Mary, D. Z. Renko, C. Guillou and C. Thal, Tetrahedron Lett.,
1997, 38, 5151–5152.
centrifugation and decantation, the tube was filled to the 5 mL
mark with 2 mL of water and 3 mL of methanol, centrifuged and
decanted. The same washing process was then repeated. The
tube was then filled with 5 mL of the solvent to be used in the
N-demethylation reaction, centrifuged and decanted. The same
washing process was repeated one more time. At this point the
nZVI was suspended in the appropriate solvent (5 mL) and
added to the N-demethylation reaction flask. The alkaloid N-
oxide hydrochloride salt (0.50 mmol) was previously dissolved
or partially dissolved in the selected solvent (10 mL) in the
reaction flask, giving a total of 15 mL of solvent, along with the
N-oxide hydrochloride salt and suspended nZVI. This mixture
was stirred for the appropriate amount of time (see Table 1) at
room temperature. Upon completion of reaction, the solvent
was removed under reduced pressure and the residue dissolved
in 3:1 (v/v) CHCl3/i-PrOH (26 mL). The resulting organic phase
was washed with 2 M NaOH (2 × 2 mL), dried over anhydrous
Na2SO4, filtered, and the solvent removed in vacuo. The
resulting residue was purified by silica column chromatography
to give the corresponding N-nor compounds.
8
DOI: 10.1039/C7GC00436B
6
7
8
9
10 K. McCamley, J. A. Ripper, R. D. Singer and P. J. Scammells, J.
Org. Chem., 2003, 68, 9847–9850.
11 G. B. Kok and P. J. Scammells, Aust. J. Chem., 2011, 64, 1515–
1521.
12 Y. Nakano, G. P. Savage, S. Saubern, P. J. Scammells and A.
Polyzos, Aust. J. Chem., 2013, 66, 178–182.
13 S. Thavaneswaran and P. J. Scammells, Bioorg. Med. Chem.
Lett., 2006, 16, 2868–2871.
14 G. B. Kok, T. D. Ashton and P. J. Scammells, Adv. Synth. Catal.,
2009, 351, 283–286.
15 G. B. Kok, C. C. Pye, R. D. Singer and P. J. Scammells, J. Org.
Chem., 2010, 75, 4806–4811.
16 Toxicological
Profile
for
Chloroform,
https://www.atsdr.cdc.gov/toxprofiles/tp.asp?id=53&tid=16,
(accessed January 2017).
17 G. B. Kok and P. J. Scammells, Org. Biomol. Chem., 2011, 9,
1008–11.
General procedure for N-demethylation of alkaloid N-oxides using
Fe3(CO)12.
18 Z. Dong and P. J. Scammells, J. Org. Chem., 2007, 72, 9881–
9885.
19 R. Dey, N. Mukherjee, S. Ahammed and B. C. Ranu, Chem.
Commun., 2012, 48, 7982–7984.
20 J. F. Sonnenberg, N. Coombs, P. A. Dube and R. H. Morris, J.
Am. Chem. Soc., 2012, 134, 5893–5899.
21 V. Kelsen, B. Wendt, S. Werkmeister, K. Junge, M. Beller and
B. Chaudret, Chem. Commun., 2013, 49, 3416–3418.
22 S. R. Kanel, B. Manning, L. Charlet and H. Choi, Environ. Sci.
Technol., 2005, 39, 1291–1298.
The alkaloid N-oxide hydrochloride salt (100 mg) was
dissolved or partially dissolved in the selected solvent (9 mL)
and an appropriate amount of Fe3(CO)12 (see Table 2) was
added as a solution/suspension in the solvent (1 mL). The
reaction was allowed to stir until reaction completion (see Table
2). Where solvent mixtures with 10% MeOH or i-PrOH were
used, the alkaloid was dissolved in the appropriate alcohol (1
mL) first, followed by the bulk solvent (8 mL). The iron species
was then delivered as a suspension/solution in the bulk solvent
(1 mL). Upon reaction completion, the solvent was removed in
vacuo and the residue dissolved in 3:1 (v/v) CHCl3/i-PrOH (26
mL). The resulting organic phase was washed with 2 M NaOH (2
× 2 mL), dried over anhydrous Na2SO4, filtered, and the solvent
removed in vacuo. The resulting residue was purified by silica
column chromatography and evaporated from DCM to give the
corresponding N-nor compounds.
23 X. Q. Li, J. Cao and W. X. Zhang, Ind. Eng. Chem. Res., 2008, 47
2131–2139.
,
24 Z. Li, K. Greden, P. J. J. Alvarez, K. B. Gregory and G. V. Lowry,
Environ. Sci. Technol., 2010, 44, 3462–3467.
25 J. F. Sonnenberg, N. Coombs, P. A. Dube and R. H. Morris, J.
Am. Chem. Soc., 2012, 134, 5893–5899.
26 L. Tan, S. Lu, Z. Fang, W. Cheng and E. Tsang, Appl. Catal., B,
2017, 200, 200-210.
27 C. M. Cirtiu, T. Raychoudhury, S. Ghoshal and A. Moores,
Colloids Surf., A, 2011, 390, 95–104.
28 E. Ye, B. Liu and W. Fan, Chem. Mater. 2007, 3, 3845–3849.
29 D. Amara, I. Felner, I. Nowik, and S. Margel, Colloids Surf., A,
2009, 339, 106–110.
Acknowledgements
30 R. K. Henderson, C. Jiménez-González, D. J. C. Constable, S. R.
Alston, G. G. A. Inglis, G. Fisher, J. Sherwood, S. P. Binksa and
A. D. Curzonsf, Green Chem., 2011, 13, 854–862.
31 A. Debono, J. Xie, S. Ventura, C. Pouton, B. Capuano and P. J.
Scammells, ChemMedChem, 2012, 7, 2122–2133.
32 A. DeBono, S. Mistry, J. Xie, D. Muthiah, J. Phillips, S. Ventura,
R. Callaghan, C. Pouton, B. Capuano and P. J. Scammells,
ChemMedChem, 2014, 9, 399–410.
The authors wish to thank the Natural Sciences and Engineering
Research Council of Canada (NSERC Discovery to R.D.S.) for
financial support for this research. R.L. would like to
acknowledge financial support from an Australian Government
Research Training Program Scholarship. Financial contributions
from Saint Mary’s University, Monash University and the
University of Nottingham are also gratefully acknowledged.
33 D. D. Do Pham, G. F. Kelso, Y. Yang and M. T. W. Hearn, Green
Chem., 2012, 14, 1189–1195.
34 H. E. Gottlieb, V. Kotlyar and A. Nudelman, A. J. Org. Chem.,
1997, 62, 7512–7515.
Notes and references
1
P. R. Blakemore and J. D. White, Chem. Commun., 2002, 11,
1159-1168
2
S. Berenyi, C. Csutoras, and A. Sipos, Curr. Med. Chem., 2009,
16, 3215–3242.
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 7
Please do not adjust margins

Green Chemistry
Page 8 of 8
View Article Online
DOI: 10.1039/C7GC00436B
N-demethylation of selected N-methylalkaloids using a modified Polonovski reaction can be accomplished employing nanoscale zero-valent
iron, nZVI, in isopropanol.
O
O
Cl
N
H
H
"Fe(0)"
Solvent
HN
HO
4c
4b