Diastereoselective Synthesis of (–)-6,7-Dimethoxy-1,2,3,4-tetrahydroisoquinoline-1-carboxylic Acid via Morpholinone Derivatives

A simple and convenient synthesis of (–)-6,7-dimethoxy-1,2,3,4-tetrahydroisoquinoline-1-carboxylic acid is described, applying a combination of two synthetic methods: the Petasis reaction and Pomeranz–Fritsch–Bobbitt cyclization. The diastereomeric morpholinone derivative N-(2,2-diethoxyethyl)-3-(3,4-dimethoxyphenyl)-5-phenyl-1,4-oxazin-2-one formed in the Petasis reaction was further transformed into 1,2,3,4-tetrahydroisoquinoline-1-carboxylic acid via Pomeranz–Fritsch–Bobbitt cyclization, a classical method of synthesis leading to the tetrahydroisoquinoline core. We review important examples of applications of the Pomeranz–Fritsch process and its modifications in the synthesis of chiral tetrahydroisoquinoline derivatives that have been published in the past two decades.

Considerable effort has been devoted to asymmetric synthesis of tetrahydroisoquinolines [13,14]. Among the variety of novel methods of the synthesizing isoquinoline core, an important role is played by the modification of traditional methods such as Pictet-Spengler cyclization, Bischler-Napieralski cyclization/reduction, or Pomeranz-Fritsch-Bobbitt cyclization.
The term "Pomeranz-Fritsch-Bobbitt synthesis" refers to a variety of synthetic strategies in which the nitrogen-containing heterocyclic ring B is built including formation of C4-C4a bond of benzylamines, bearing a two-carbon C-C chain at the nitrogen atom with a good alkoxy leaving group (Scheme 1) [13,14].
A modified procedure for the classical Pomeranz-Fritsch protocol was presented by the Lumb Group [43]. They showed that the strong acids and elevated temperatures used in the Pomeranz-Fritsch cyclization step can be replaced by a combination of silyl triflate and a sterically encumbered pyridine base [44], which allows acetal activation under milder, more chemoselective conditions. This modification tolerates acid-sensitive functional groups in substrates and facilitates the synthesis of diverse 1,2-dihydroisoquinoline derivatives of type 32. From acetals of type 31. Further functionalization of these compounds led to reduced 1,2,3,4-tetrahydroisoquinolines of type 33 (Scheme 8). Enantioselective synthesis of (S)-salsolidine (28) was part of the work carried out in our laboratory [41]. Additions of methyllithium to the imine 7, carried out in the presence of several oxazoline chiral ligands, type 29, led to aminoacetal 30 of known (S) configuration. The yields differed from 44-92% and enantioselectivity was up to 76% e.e., depending on the type of oxazoline 29 used (Scheme 7). The Pomeranz-Fritsch-Bobbitt cyclization of aminoacetal 30 led to (S)-(-)-salsolidine (28) [42]. Scheme 7. Enantioselective synthesis of (S)-salsolidine (28) by addition of MeLi to imine 7 catalyzed by chiral oxazolidine 29.
A modified procedure for the classical Pomeranz-Fritsch protocol was presented by the Lumb Group [43]. They showed that the strong acids and elevated temperatures used in the Pomeranz-Fritsch cyclization step can be replaced by a combination of silyl triflate and a sterically encumbered pyridine base [44], which allows acetal activation under milder, more chemoselective conditions. This modification tolerates acid-sensitive functional groups in substrates and facilitates the synthesis of diverse 1,2-dihydroisoquinoline derivatives of type 32. From acetals of type 31. Further functionalization of these compounds led to reduced 1,2,3,4-tetrahydroisoquinolines of type 33 (Scheme 8).
Enantioselective synthesis of (S)-salsolidine (28) was part of the work carried out in our laboratory [41]. Additions of methyllithium to the imine 7, carried out in the presence of several oxazoline chiral ligands, type 29, led to aminoacetal 30 of known (S) configuration. The yields differed from 44-92% and enantioselectivity was up to 76% e.e., depending on the type of oxazoline 29 used (Scheme 7). The Pomeranz-Fritsch-Bobbitt cyclization of aminoacetal 30 led to (S)-(-)-salsolidine (28) [42]. A modified procedure for the classical Pomeranz-Fritsch protocol was presented by the Lumb Group [43]. They showed that the strong acids and elevated temperatures used in the Pomeranz-Fritsch cyclization step can be replaced by a combination of silyl triflate and a sterically encumbered pyridine base [44], which allows acetal activation under milder, more chemoselective conditions. This modification tolerates acid-sensitive functional groups in substrates and facilitates the synthesis of diverse 1,2-dihydroisoquinoline derivatives of type 32. From acetals of type 31. Further functionalization of these compounds led to reduced 1,2,3,4-tetrahydroisoquinolines of type 33 (Scheme 8).

11.
Retrosynthetic approach for the synthesis of (-)-6,7-dimethoxy-1,2,3,4-tetrahydroisoquinoline-1-carboxylic acid. Chiral aminoacetal 47 incorporating (R)-phenylglycinol moiety was synthesized in the reaction of (R)-phenylglycinol (45) with excess of 2-bromo-1,1-diethoxyethane (46) carried out in dry DMF in the presence of anhydrous K2CO3 at 110 °C. After purification by column chromatography, the product was isolated in 67% yield. When the same reaction was carried with an equimolar ratio of (R)-phenylglycinol (45) and 46, low conversion was observed as indicated by TLC. The next step of the synthesis was the Petasis reaction with boronic acid 38, glyoxylic acid monohydrate (39), and aminoacetaldehyde acetal 47 (Scheme 12). Scheme 11. Retrosynthetic approach for the synthesis of (-)-6,7-dimethoxy-1,2,3,4-tetrahydroisoquinoline-1-carboxylic acid. Chiral aminoacetal 47 incorporating (R)-phenylglycinol moiety was synthesized in the reaction of (R)-phenylglycinol (45) with excess of 2-bromo-1,1-diethoxyethane (46) carried out in dry DMF in the presence of anhydrous K 2 CO 3 at 110 • C. After purification by column chromatography, the product was isolated in 67% yield. When the same reaction was carried with an equimolar ratio of (R)-phenylglycinol (45) and 46, low conversion was observed as indicated by TLC. The next step of the synthesis was the Petasis reaction with boronic acid 38, glyoxylic acid monohydrate (39), and aminoacetaldehyde acetal 47 (Scheme 12). The substrates 38, 39, and 47 were stirred in DCM at room temperature for 78 h, then the inorganic solid was removed by filtration. The filtrate was evaporated in vacuo to give an oily residue consisting of diastereomeric oxazin-2-ones 48 and 49 with 3:1 d.r. It was necessary to carry out this reaction for at least 78 h, because at shorter reaction times substantial amounts of unreacted substrates 38 and 47 were present in the reaction mixture as well as two unidentified products, considerably more polar than 48 and 49, probably being their noncyclized precursors (not shown).
The diastereomeric composition of oxazin-2-ones was established by 1 H NMR spectroscopic analysis of the crude reaction mixture. The diastereomeric ratio of 3:1 for compounds 48 and 49 was deduced by comparing the integration values for the H-3 oxazin-2-one ring proton signals at 4.91 and 5.05 ppm, respectively, and integration values for the methyl groups of the acetal moiety signals at 1.21 and 1.12 ppm.
After purification of the crude product by column chromatography, pure major diastereomer 48 (51%) and minor diastereomer 49 (16%) were isolated as colourless oils. The substrates 38, 39, and 47 were stirred in DCM at room temperature for 78 h, then the inorganic solid was removed by filtration. The filtrate was evaporated in vacuo to give an oily residue consisting of diastereomeric oxazin-2-ones 48 and 49 with 3:1 d.r. It was necessary to carry out this reaction for at least 78 h, because at shorter reaction times substantial amounts of unreacted substrates 38 and 47 were present in the reaction mixture as well as two unidentified products, considerably more polar than 48 and 49, probably being their noncyclized precursors (not shown).
The diastereomeric composition of oxazin-2-ones was established by 1 H NMR spectroscopic analysis of the crude reaction mixture. The diastereomeric ratio of 3:1 for compounds 48 and 49 was deduced by comparing the integration values for the H-3 oxazin-2-one ring proton signals at 4.91 and 5.05 ppm, respectively, and integration values for the methyl groups of the acetal moiety signals at 1.21 and 1.12 ppm.
The major diastereomer 48 was the key intermediate for the synthesis of 1.
Debenzylation of oxazinone 48 to aminoacetal 50 (racemic 50 ref. [46]) was performed by hydrogenolysis conducted under atmospheric pressure (using a reaction balloon filled with hydrogen gas) in the presence of palladium hydroxide on charcoal (Pearlman's catalyst) (Scheme 13). The crude product was pure enough for the next reaction step, at approximately 90% as determined by 1 H NMR, but contained some minor unidentified impurities. Therefore, it was purified by column chromatography to give the pure (R)-(-)-N-(2,2-diethoxyethyl)-3,4-dimethoxyphenylglycine (50) in 66% yield as a yellowish solid.
was necessary to carry out this reaction for at least 78 h, because at shorter reaction times substantial amounts of unreacted substrates 38 and 47 were present in the reaction mixture as well as two unidentified products, considerably more polar than 48 and 49, probably being their noncyclized precursors (not shown).
The diastereomeric composition of oxazin-2-ones was established by 1 H NMR spectroscopic analysis of the crude reaction mixture. The diastereomeric ratio of 3:1 for compounds 48 and 49 was deduced by comparing the integration values for the H-3 oxazin-2-one ring proton signals at 4.91 and 5.05 ppm, respectively, and integration values for the methyl groups of the acetal moiety signals at 1.21 and 1.12 ppm.
The major diastereomer 48 was the key intermediate for the synthesis of 1.

General
Melting points were determined by using open glass capillaries in a Büchi melting point B-545 apparatus and are reported uncorrected. IR spectra were recorded on a Bruker FT-IR IFS 113V spectrophotometer or Jasco FT-IR 4600 spectrophotometer with ATR PRO ONE using a diamond crystal. The 1 H and 13 C NMR spectra were recorded on a Bruker ASCEND 400 spectrometer. NMR spectra are reported in parts per million (ppm) and were measured relative to the signals for residual solvent peak CDCl 3 or DMSO-d 6 , or using tetramethylsilane as an internal reference. Mass spectra (EI) were measured using an AMD402 spectrometer. High-resolution mass spectra (HRMS) were measured using an Impact HD (Bruker Daltonics, Bremen, Germany) spectrometer. High-resolution ESI-MS spectra were recorded on a quadrupole time-of-flight mass spectrometer (QTOF, Impact HD, Bruker Daltonics, Bremen, Germany) in positive mode. Merck DC-Alufolien Kieselgel 60 254 were used for TLC, and silica gel (100-200 mesh ASTM) for column chromatography. Optical rotation was measured using a Perkin-Elmer polarimeter 242B at 20 • C. All reagents and solvents were purchased from commercial suppliers and used as received. (R)-phenylglycinol (45) (798 mg, 5.82 mmol), anhydrous K 2 CO 3 (1465 mg, 10.6 mmol) and DMF (6 mL) followed by 2-bromo-1,1-diethoxyethane (46) (1380 mg, 1.05 mL, 7 mmol) were added to a round-bottomed flask equipped with a magnetic stirrer and reflux condenser with drying tube attached. The reaction mixture was heated at 110 • C for 24 h. During that time, an additional quantity of 46 (1380 mg, 1.05 mL, 7 mmol) was added. After completion of the reaction, the mixture was poured onto ice and after reaching rt was extracted four times with ethyl ether. The organic extracts were dried over anhydrous MgSO 4 and the solvent was evaporated to afford crude 47, which after purification by silica gel column chromatography (DCM-MeOH 100:0 to 98:2 v/v) yielded aminoacetal 47 as a solidifying oil (988 mg, 67%).