Rapid microwave-assisted cleavage of methyl phenyl ethers: New method for synthesizing desmethyl precursors and for removing protecting groups

Article abstract of DOI:10.1002/jlcr.577

A new microwave-enhanced method for rapid demethylation of methyl phenyl ethers using neat methanesulfonic acid (CH₃SO₃H) is presented. Using a monomodal microwave cavity, cleavage of anisole (1), used as model compound, to phenol (2) was achieved with high conversions (ca 80%) in very short reaction times (10-20 s). The feasibility of cleaving one or both of two methoxy groups was illustrated with 4-(3-bromoanilino)-6,7-dimethoxyquinazoline (PD153035, 3). High conversions (≥82%) of 3 were attained with four different conditions (i.e. combination of input effect (35-125 W) and time (15s-2min)). 4-(3-Bromoanilino)-7-hydroxy-6-methoxyquinazoline (4), 4-(3-bromoanilino)-6-hydroxy-7-methoxyquinazoline (5) and 4-(3-bromoanilino)-6,7-dihydroxyquinazoline (6), the possible mono- or di-demethylated compounds, were obtained. Methods for rapid demethylations are of interest in radiochemistry for post-labeling deprotections of hydroxyl containing aromatic rings and also provide a more direct route for synthesizing precursor compounds for labeling by alkylation. Copyright

Full text of DOI:10.1002/jlcr.577

JOURNAL OF LABELLED COMPOUNDS AND RADIOPHARMACEUTICALS
J Label Compd Radiopharm 2002; 45: 529–538.
Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/jlcr.577
Research Article
Rapid microwave-assisted cleavage of methyl
phenyl ethers: new method for synthesizing
desmethyl precursors and for removing
protecting groups
Anna Fredriksson* and Sharon Stone-Elander
Karolinska Pharmacy, Karolinska Hospital, SE-171 76 Stockholm, Sweden
Summary
A new microwave-enhanced method for rapid demethylation of methyl phenyl
ethers using neat methanesulfonic acid (CH3SO3H) is presented. Using a
monomodal microwave cavity, cleavage of anisole (1), used as model
compound, to phenol (2) was achieved with high conversions (ca 80%) in
very short reaction times (10–20 s). The feasibility of cleaving one or both of
two methoxy groups was illustrated with 4-(3-bromoanilino)-6,7-dimethoxy-
quinazoline (PD153035, 3). High conversions (582%) of 3 were attained
with four different conditions (i.e. combination of input effect (35–125 W) and
time (15 s–2 min)). 4-(3-Bromoanilino)-7-hydroxy-6-methoxyquinazoline (4),
4-(3-bromoanilino)-6-hydroxy-7-methoxyquinazoline (5) and 4-(3-bromoanili-
no)-6,7-dihydroxyquinazoline (6), the possible mono- or di-demethylated
compounds, were obtained. Methods for rapid demethylations are of interest
in radiochemistry for post-labeling deprotections of hydroxyl containing
aromatic rings and also provide a more direct route for synthesizing precursor
compounds for labeling by alkylation. Copyright # 2002 John Wiley & Sons,
Ltd.
Key Words: demethylation; deprotection; precursor synthesis; PD153035
*Correspondence to: A. Fredriksson, Karolinska Pharmacy, Karolinska Hospital, SE-171 76
Stockholm, Sweden. E-mail: anna.fredriksson@ks.se
Copyright # 2002 John Wiley & Sons, Ltd.
Received 31 December 2001
Revised 14 February 2002
Accepted 16 February 2002

530
A. FREDRIKSSON AND S. STONE-ELANDER
Introduction
The usefulness of microwaves for accelerating organic syntheses has
been recognized since the mid-1980s and the number of synthetic
publications involving microwave techniques is rapidly increasing.
Reviews1À3 are available that summarize microwave applications in
different types of reactions, such as oxidations, condensations,
heterocyclizations and carbon–carbon couplings. Reactions that usually
require conventional heating, often proceed equally well or better, but
more rapidly with microwave heating. The reductions in reaction times
achievable are particularly important for syntheses with short-lived
radionuclides and microwave techniques have been successfully applied
in many radiolabelings.4,5 In addition to the labeling step, many tracer
preparations require additional transformations such as functional
group protection and deprotection that might also profit from the
accelerations achieved with microwave heating. For example, phenols
are often protected as ethers during aromatic radiofluorinations and the
ethers must be removed post-labeling (review on ether clevage).6,7
A
typical example is the synthesis of [6-18F]6-fluorodopa in which the
catechol hydroxyls must be deprotected after the labeling step.8 Such
multi-step syntheses are more common in fluorine-18 chemistry, but
efficient and fast conditions for demethylation of N, O and S could also
be potentially beneficial for carbon-11 radiochemical strategies.
Microwave-assisted deprotections of aromatic methyl ethers have
previously been performed with pyridine hydrochloride,9 KOBu-t and
crown ether10 and with lithium iodide and solid supports.11 These
methods were either not fast enough and/or the reagents used did not
lend themselves well to the handling constraints of fluorine-18 and
carbon-11 chemistry. We have examined here the use of monomodal
microwave heating to speed up the cleavage of methylphenylethers with
methanesulfonic acid12 (CH3SO3H). Since CH3SO3H is very polar, it
would be expected to interact strongly with the oscillating electro-
magnetic field. Its high boiling point (1678C at 10 mm Hg) suggested
that high temperatures could be employed without the pressure increase
typically observed in microwave heating of the more volatile mineral
acids.
We report the use of this deprotection method in a model reaction
(Scheme 1a) in which methyl phenyl ether (anisole, 1) was converted to
phenol (2). Furthermore, the mono- and di-demethylation of PD153035
(4-(3-bromoanilino)-6,7-dimethoxyquinazoline, 3, Scheme 1b), a tyrosine
Copyright # 2002 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2002; 45: 529–538

RAPID DEMETHYLATIONS
531
Scheme 1. Methyl phenyl ether cleavage with methanesulfonic acid was studied
in: (a) the conversion of anisole (1) to phenol (2); and (b) the conversion of
PD153035 (3) to the two monomethoxy compounds (4 and 5) and the dihydroxy
product, 6
kinase inhibitor selective for the epidermal growth factor receptor,13
was also studied. Finding feasible routes for rapidly synthesizing the
desmethyl counterparts of potentially interesting pharmaceuticals can
expedite the tracer developmental process. Directly demethylating the
parent compound avoids the alternative development and optimization
of sometimes complex and lengthy organic pathways. In compounds
with multiple methyl groups, access to all the different desmethyl
compounds may be desirable for labeling in different positions so that
the metabolism of the parent compound can be examined.
Experimental
General
All reagents, unless otherwise specified, were of analytical grade and
commercially available. Anisole (1) and CH3SO3H were obtained from
Aldrich and phenol (2) from Merck. 4-(3-Bromoanilino)-6,7-dimeth-
oxyquinazoline (PD153035, 3) was synthesized as described previously14
based on the method of Rewcastle et al.15 HPLC analyses were
performed using a Shimadzu SPD-6A spectrophotometric detector, a
Shimadzu C-R4AX integrator and two Shimadzu LC-10AD pumps.
LC/MS-analysis were performed using a VG Platform II, Fisons
Copyright # 2002 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2002; 45: 529–538

532
A. FREDRIKSSON AND S. STONE-ELANDER
Instrument. The monomodal microwave cavity used was a Microwell 10
(Personal Chemistry AB, Uppsala, Sweden).
Demethylation procedures
In pyrex vessels16 1 (15 ml, 138 mmol) or 3 (1.5–20 mg,. 4.2–55.5 mmol)
was dissolved in CH3SO3H (200 or 500 ml, respectively). The samples
were vortexed, the tube openings were covered with parafilm (to prevent
loss of sample in case of bumping) and they were subsequently treated
with microwaves of different input power (35–125 watts (W)) for
different times (10 s–2 min). Immediately after irradiation (eor, end of
reaction) the temperature in the reaction mixture was measured using a
mercury thermometer (measuring interval 70–3608C). The demethyla-
tion of 3 was also performed with conventional heating at 200 Æ 58C
(silicon oil bath). Aliquots (5 ml) taken were diluted in distilled water
(dH2O, 200–400 ml depending on the initial concentration of 3) for
HPLC analysis. In the model reaction aliquots (and reference 2) were
diluted in ethanol.
Analysis
The conversion of 1 to 2 was followed at 270 nm using a PRP-1 column
(300 Â 7.0 mm, 10 mm) and a mobile phase of A (acetonitrile (MeCN)
containing 0.085 vol% trifluoroacetic acid (TFA)) and B (dH2O
containing 0.1 vol% TFA). At a flow rate of 2.0 ml/min, the mobile
phase composition was changed by increasing A in a linear gradient
from 40 to 60% during 14 min and then an isocratic elution at 75% A
from 14 to 25 min. The demethylation of 3 was followed at 249 nm using
a C18 mBondapakTM column (Waters, 3.9 Â 300 mm, 10 mm), isocratic
elution with MeCN:dH2O:TFA-30:70:0.1 at 1.5 ml/min.
The product of the demethylation of 1 was identified by coelution on
HPLC with reference 2. The products 4–6 obtained in the demethylation
reaction of 3 were identified by LC/MS-analysis (positive (50–100 V)
electrospray ionization (ESI)) performed at the Swedish Pharmacy’s
Central Laboratory (Kungens Kurva, Sweden). The unfragmented
[M+H]+ ions (m/z) were 359.8, 345.8, 345.8 and 331.8 for 3–6,
respectively, based on the bromine-79 peak.
Conversions were estimated from the total area (area units, AU) of
the peaks deemed to be derived from the starting material. In the
demethylation of 1, the AUs for 1 and 2 were related to the total AUs
Copyright # 2002 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2002; 45: 529–538

RAPID DEMETHYLATIONS
533
Figure 1. HPLC chromatograms from the product distribution analyses (UV-
absorbance) for: (a) 1 (tR ꢀ 21.3 min) cleaved to 2 (tR ꢀ 9.4 min); and (b) 3
cleaved to the two mono demethylated products, 4 and 5 and the didemethylated
product, 6. The approximate tR for 3–6 were 12.4, 9.7, 8.1 and 6.8 min,
respectively. The unidentified hydrophilic byproduct (tR ꢀ 1.8–1.9 min) is marked
with an asterisk
Figure 2. (a) The temperature of CH3SO3H (500 ml) as a function of time
measured after microwave irradiation with different input power or during
conventional heating in an oil bath of 195–2008C; (b) A bar diagram showing the
conversions of 1 to 2 for different microwave input powers and times. Unidentified
byproducts are indicated with an asterisk
calculated (see a typical HPLC chromatogram in Figure 1(a)). The
molar absorptivities at 270 nm of 1 and 2 are nearly the same, with that
of 1 being slightly larger than that of 2. Thus, conversions to 2 might be
slightly underestimated. All peaks other than 1 and 2 were considered as
unidentified byproducts (marked with an asterisk in Figure 2b). Since
independently synthesized references 4–6 were either not available at all
or in very small quantities, the amounts of 3–6 were calculated based on
their area as related to the initial concentration of 3. An unidentified
polar byproduct (marked with an asterisk in Figure 1b) was also
produced in increasing amounts at longer times or higher input power.
Copyright # 2002 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2002; 45: 529–538

534
A. FREDRIKSSON AND S. STONE-ELANDER
To facilitate comparison of the different conditions, the total amounts
of 3–6 together with the unidentified byproduct (*) were normalized to
the original area of 3 (set as 100% with a correction factor range:
1.0 Æ 0.2), without adjustment for individual molar absorptivities at
249 nm.
Results and discussion
This new microwave-assisted method for rapidly cleaving methyl phenyl
ethers to phenols with CH3SO3H gave good conversions in 51 min
reaction time. Conditions were identified which are suitable for
removing methyl protecting groups post-labeling and for rapidly
synthesizing the desmethyl precursors required for labeling with alkyl
halides or triflates.
The temperatures achieved when heating CH3SO3H (500 ml) for
different times and microwave input powers are shown in Figure 2a. The
expected pattern of faster heating at higher input power was observed.
It is noteworthy that the temperature curves for different input powers
are similar, i.e. there seems to be a delay before the temperature
increases at lower power but once the heating has started the rise per
unit time is fairly similar.
The conversion of 1 to 2 proceeded with good yields at very short
reaction times. After only 10 s with 100 W, the conversion of 1 was ca.
75%. In addition to 2 (ca. 60%), other unidentified byproducts were
observed, as shown in Figure 2b. The byproducts eluted at retention
times between those of 1 and 2 on the PRP-1 column and were not
observed when CH3SO3H alone was heated under similar conditions.
Further heating with 100 W did not improve the yield of 2, nor did
prolonged heating at lower input powers. For example, very little 1 (ca.
7%) remained after treatment with 100 W for 30 s, but the ratio of 2 to
byproducts had also decreased (28%/65%). Apparently, once the
optimum reaction temperature is reached, demethylation occurs very
readily, but the phenol formed can be further transformed unless the
microwave heating is quickly stopped.
The demethylating ability of CH3SO3H was further examined with
the more complex PD153035 (3). With its vicinal methoxy groups
(Scheme 1b) this compound provides an opportunity for investigating
the method’s regioselectivity, as well as multiple demethylation
capability. HPLC analysis of reaction aliquots clearly showed the
Copyright # 2002 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2002; 45: 529–538

RAPID DEMETHYLATIONS
535
formation of three new compounds (4–6) (see the chromatogram in
Figure 1b). They were identified by LC/MS analysis as the didemethy-
lated 6 and the two monomethoxy compounds (4 and 5) arising from
the monodemethylation of 3. Compounds 4 and 5 can be easily
separated on the C18-HPLC column (Figure 1b). However, their
fragmentation patterns in the ESI-MS analysis were identical. The
major fragments were [M+H+– CH3], [M+H+– CH3 – CO], [M+H+–
CH3 – CO – CO], [M+H+– CH3 – Br] with m/z of approximately 331,
302, 274 (based on bromine-79) and 251, respectively. The identities of 4
and 5 were established based on coelution with reference compounds on
HPLC. Compound 4 coeluted with 4-(3-bromoanilino)-7-hydroxy-6-
methoxyquinazoline that we have synthesized previously by a multi-step
pathway (data not shown) starting with production of 2-amino-4-
benzyloxy-5-methoxybenzoic acid which was further transformed by
basically the same route as used in the synthesis of 3 and then finishing
with selective cleavage of the benzyl group. Compound 5 coeluted with
the desmethyl precursor compound of 3 synthesized (by demethylation
using BBr3) as previously reported14 and then realkylated with
[11C]iodomethane to afford [11C]-3, and is thus assigned as 4-(3-
bromoanilino)-6-hydroxy-7-methoxyquinazoline.
Figure 3 summarizes the demethylation results at different reaction
times for four different microwave input powers. For each input power,
the reaction time was selected so that an eor temperature of ca.
180–2008C would be achieved (Figure 2a), since initial screening had
Figure 3. Compound distribution (expressed as percent with respect to starting
concentration of 3) after the demethylation reaction of 3 (1.5–2.5 mg) in
CH3SO3H at various microwave input powers and for different times. The
asterisk indicates the unidentified byproduct
Copyright # 2002 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2002; 45: 529–538

536
A. FREDRIKSSON AND S. STONE-ELANDER
indicated that those temperatures were required for demethylations of 3
to occur (data not shown). The actual eor temperatures measured for
each reaction time and effect are indicated in Figure 3. Conversion of 3
was efficient (582%) in all four trials (Figure 3), with the best
conversion (>90%) obtained after 2 min with 35 W. All of the possible
products 4–6 were obtained, though a certain degree of regioselectivity
favoring 5 was observed. Prolonged heating gave the most didemethyla-
tion to 6, along with a more hydrophilic byproduct (tR ꢀ 1.8–1.9 min,
marked with an asterisk in Figure 1b). Both monomethoxy compounds
were obtained in all four trials, with around 8–14% and 26–35% of 4
and 5, respectively, after the first heating period. During the second
heating period, 3–5 decreased while the didemethylated 6 and the
unidentified byproduct increased. Highest yields of 6 (40%) were
obtained with 125 W for 30 s. These conditions can be attractive for use
in radiosyntheses in which multiple demethylations must be performed.
To test whether the reaction could be scaled up, 10 and 20 mg of 3
were also subjected to the same conditions as for 1.5–2.5 mg in the series
above. As seen in Figure 4, similar conversions of 3 were observed for
these 4- to 10-fold increases in concentration, with 5 obtained in the
highest yields under the conditions tested. These results indicate the
feasibility of using this method as a quick preparative route to the
desmethyl compounds.
The reaction of 3 with CH3SO3H was also performed with
conventional heating in an oil bath at 200 Æ 58C. No demethylation
Figure 4. Conversions (expressed as percent with respect to starting concentra-
tion of 3) for: (a and b) scaled-up microwave reactions; and (c) after
conventionally heating 3 (1.8 mg) with CH3SO3H in an oil bath
Copyright # 2002 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2002; 45: 529–538

RAPID DEMETHYLATIONS
537
of 3 was detectable after the first 20 s, in contrast to the microwave
experiments. After 10 min, 30% of 3 had still not reacted and only small
amounts of the didemethylated product, 6, were produced. Other
byproducts were not detected.
The demethylation of 3 was thus faster with microwave heating
(conversions 582% in 42 min) than with conventional heating (ca.
70% at 10 min). Regioselectivity favoring 5 was essentially the same for
both heating methods, while the extent of didemethylation differed
considerably. The temperature of the sample increased much more
slowly when it was convectively heated in the oil bath (Figure 2a), but
had reached 1908C (the eor temperature of the microwave experiments)
by 1.5 min. The sample temperatures were not continuously measured in
the microwave experiments nor were they corrected for possible cooling
of reaction mixture by the thermometer. It is likely that the
temperatures reached during the ‘flash’ microwave heating were
somewhat higher than the temperature measured at eor. It has also
been proposed17 that local ‘hot spots’ may form during microwave
heating and that the rapid rate of heating can contribute to the
accelerations reported and may also have affected the differing extents
of didemethylation and byproduct formation observed here.
Conclusions
Cleavage of methyl phenyl ethers with CH3SO3H has been accom-
plished with the aid of monomodal microwave heating. Multiple
demethylations could be rapidly performed in good yields with
microwave, but not with conventional heating. The speed of this
method makes it attractive for post-labeling deprotections in both
fluorine-18 and carbon-11 labeling chemistry. The capability for quickly
preparing multi-milligram quantities of desmethyl compounds for
labeling in alternative positions may reduce the necessity for alternative,
protracted synthetic procedures. These results are particularly convin-
cing considering our previous experiences that the demethylation
of 3 is difficult to accomplish with more commonly used cleavage
conditions.14
Acknowledgements
Support from the Swedish National Board for Technical Research
(210-1997-494), Personal Chemistry AB and the Karolinska Institute
Copyright # 2002 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2002; 45: 529–538

538
A. FREDRIKSSON AND S. STONE-ELANDER
are gratefully acknowledged. This project was, in part, initiated through
the EU sponsored D10 COST program (Innovative Methods and
Techniques for Chemical Transformations).
References
1. Abramovitch RA. Org Prep Proced Int 1991; 23: 683–711.
2. Caddick S. Tetrahedron 1995; 51: 10 403–10 432.
.
3. Lidstrom P, Tierney J, Wathey B, Westman J Tetrahedron 2001; 57:
9225–9283.
4. Elander N, Jones JR, Lu S-Y, Stone-Elander S. Chem Soc Rev 2000;
29: 239–249; and references therein.
5. Stone-Elander S, Elander N. J Label Compd Radiopharm 2002, in press;
and references therein.
6. Bhatt MV, Kulkarni SU. Synthesis 1983; 4: 249–282.
7. Tiecco M. Synthesis 1988; 10: 749–759.
8. Luxen A, Guillaume M, Melega WP, Pike VW, Solin O, Wagner R. Nucl
Med Biol 1992; 19: 149–158.
9. Kulkarni PP, Kadam AJ, Mane RB, Desai UV, Wadgaonkar PP. J Chem
Res. (S) 1999; 6: 394–395.
.
10. Oussaıd A, Thach LN, Loupy A. Tetrahedron Lett 1997; 38: 2451–2454.
11. Stone-Elander S, Fredriksson A, Elander N. J Label Compd Radiopharm
2001; 44: S1029–S1031.
12. Fujii N, Irie H, Yajima H. J Chem Soc Perkin Trans I 1977; 2288–2289.
13. Fry DW, Kraker AJ, McMichael A, Ambroso LA, Nelson JM, Leopold
WR, Connors RW, Bridges AJ. Science 1994; 265: 1093–1095.
.
14. Johnstrom P, Fredriksson A, Thorell J-O, Stone-Elander S. J Label Compd
Radiopharm 1998; 41: 623–629.
15. Rewcastle GW, Denny WA, Bridges AJ, Zhou H, Cody DR, McMichael
A, Fry DW. J Med Chem 1995; 38: 3482–3487.
(
16. Stone-Elander SA, Elander N, Thorell J-O, Solas G, Svennebrink J. J
Label Compd Radiopharm 1994; 34: 949–960.
17. Galema SA. Chem Soc Rev 1997; 26: 233–238.
Copyright # 2002 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2002; 45: 529–538