Todd:The Lewis Acid-Catalysed Pictet-Spengler Formation of Substituted 1,2,3,4-Tetrahydroisoquinolines

Home        Contact        Internal        People        Papers        Research        Teaching        Links        News        Construction       

This is an undergraduate project running from March-June 2012. Current plan of action here. Final document due: 17:00 +10 GMT (AEST), 15 June 2012.

Abstract

The Pictet-Spengler (PS) reaction is a useful carbon-carbon bond forming reaction. It is efficiently used in the synthesis of tetrahydroisoquinoline- (THIQ), tetrahydro-β-carboline- and more recently, quinoxaline-derived moieties. (SOMETHING ELSE). There are no known examples of the asymmetric Lewis acid-catalysed PS reaction. There are also no reported examples of the catalytic asymmetric PS formation of THIQ frameworks. This project... Ytterbium(III) triflate catalysed acyl-PS reaction to give the THIQ scaffold.

Introduction and Background

The Pictet-Spengler Reaction in Organic Synthesis and Nature

The first example of the synthesis of tetrahydroisoquinoline (THIQ) from β-phenethylamine and formaldehyde was reported by Amé Pictet and Theodor Spengler in 1911.^[1] The reaction now known as the Pictet-Spengler (PS) reaction typically occurs via the condensation of a β-arylamine and an aldehyde or ketone to give an electrophilic iminium species. This is subsequently attacked by the neighbouring aryl group (Scheme 1).

Scheme 1: The typical Pictet-Spengler reaction based on the original synthesis of tetrahydroisoquinoline (red).

The PS reaction has been found to occur in nature. The recently discovered class of enzymes known as Pictet-Spenglerases are instrumental in the synthesis of 3α(S)-strictosidine and (S)-norcoclaurine, which are key intermediates in the biosynthesis of numerous plant alkaloids (Figure 1).^[2,3] Additionally, studies of a biosynthetic pathway to the potent antibiotic saframycin A have revealed that the bacterium nonribosomal peptide synthetase SfmC constructs the scaffold via two PS reactions (Figure 1).^[4]

Figure 1: Natural PS products bearing the tetrahydro-β-carboline (blue) and tetrahydroisoquinoline moieties (red).

The PS reaction is a powerful tool in the synthesis of complex, biologically active polycyclic heterocycles.^{[3, 8-9]}

It is commonly applied typically associated with plant alkaloids.^[3] Other notable variations include the acyl-PS reaction and the modified PS to give fused quinoxalines (Scheme 2).^[5-7]

Scheme 2: The tetrahydro-β-carboline, acyl-, alkyl- and quinoxaline Pictet-Spengler variations.

The Limits of the Pictet-Spengler Reaction in Organic Synthesis

A limitation of the PS reaction is the catalytic formation of non-activated THIQs. Ironically, the first PS formation of the THIQ scaffold appears to be the most difficult to produce catalytically via the PS route, at least by traditional homogenous catalytic methods.^[B,6] Two examples are known where the PS cyclisation of the unactivated β-phenethylamine with various aldehydes is effected by zeolite and acidic montomorillonite frameworks.^[D] A biosynthetic example of the PS formation of the unactivated THIQ scaffold is yet to be found.

Recent reviews of the literature have revealed a lack of Lewis acid-catalysed asymmetric PS reactions.^{[3, H]} Furthermore, there are no known examples of the asymmetric PS formation of the THIQ scaffold. Instead, the focus is on the organocatalysed asymmetric PS formation of tetrahydro-β-carbolines using catalysts such as the chiral thioureas^[ref] and binapthyl-derived phosphoric acids.^[ref]

Still, there is precedent for the development of a catalytic asymmetric PS reaction to yield the THIQ scaffold. Rather than using chiral organocatalysts, as is the case with the synthesis of the tetrahydro-β-carboline moiety, the potential for the catalytic asymmetric formation to give the THIQ scaffold possibly lies with Lewis acid-catalysis. The The low reactivity of the imines - Cyclisation to give THIQ is difficult because the imines are rather unreactive (ref [5] mentions their low reactivity). Lewis acids have stronger electron withdrawing character than Brønsted acids(?), which should make the iminium more electrophilic. Existing examples of the asymmetric Friedel-Crafts, which is mechanistically similar, effected by a combination of a Lewis acid-catalyst and chiral ligands of the bisoxazoline type.

Project Summary

A method for the asymmetric Lewis acid-catalysed PS reaction to give the THIQ scaffold first requires a racemic "template". In comparison to the number of reported examples of the Lewis acid-catalysed formation of the tetrahydro-β-carboline scaffold, the scope of the synthesis of THIQs using Lewis acid catalysts is limited. (FIX) Three examples of the Lewis acid-catalysed PS reaction to give the THIQ scaffold use ytterbium(III) triflate,^[A] calcium hexafluoroisopropoxide^[B] and a gold(III) chloride-silver triflate combination.^[6]

The focus of this project was the development of a method for the achiral Lewis acid-catalysed PS formation of the THIQ scaffold that would be adaptable to an asymmetric model (Scheme 3). In order to minimise complications arising from the one-pot procedures using the aldehyde and arylamine starting materials, the corresponding isolatable imines were prepared for use as model substrates (Scheme 3, 1a-d).

Scheme 3: Project outline for the PS formation of THIQs (2a-c and 3) from simple imines (1a-d).

Brønsted acids only effected cyclisation of the imine substrates at high temperatures or when present in superstoichiometric amounts. This was consistent with the known literature that, in the PS-syntheses of the activated THIQ scaffold, describes the use of strong Brønsted acids in vast excess.^[Gao,X] Lewis acids revealed a similar ineffectiveness at inducing the PS formation of the THIQs 2b and 2c from the corresponding imines.
Adaptation of the model to the acyl-PS reaction produced more promising results. The ytterbium(III) triflate Lewis acid catalyst was found to be highly efficient at inducing the acyl-PS cyclisation of the activated THIQ scaffold. The reaction optimisation process was aided by a ¹H-NMR spectroscopic assay that involved development of a post-reaction work-up procedure.

Results and Discussion

Synthesis of the Model Substrate Materials and the Brønsted Acid-Mediated Cyclisations

The first phase of the project required synthesis of the imine model substrates, which were easily prepared in excellent yield (Scheme 4, Table 1). The condensation reactions of the arylamines with the corresponding aldehydes was facile. The nitro-substituted imines (1c and 1d) did not require post-reaction addition of a dehydrating agent but readily crystallised from the crude reaction mixtures.^[C1]

Scheme 4: Synthesis of the substrates for the Pictet-Spengler model.

Table 1: Results of the synthesis of the substrates for the Pictet-Spengler model.

The literature procedures for the PS formation of activated-THIQs describe harsh conditions: excess strong Brønsted acids and high temperatures.^[Gao] In contrast, there are a number of examples describing the Brønsted acid-catalysed PS formation of the tetrahydro-β-carboline scaffold from tryptamine and various aryl- and alkyl-aldehydes.[ref] It was therefore appropriate to evaluate the reactivity of the model imine substrates under excess and catalytic Brønsted acid reaction conditions.

Attempts to cyclise imine 1a using neat methanesulfonic acid and elevated temperatures were largely unsuccessful (Table 2, Entries 1-3). Of the activated imine, 1b, ¹H-NMR spectroscopic analysis indicated good conversion when a large excess of trifluoroacetic acid was used (Table 2, Entry 5). Lowering the acid loading and increasing the reaction temperature resulted in moderate conversion. Under catalytic conditions, no conversion of 1b was observed after 5 days and no conversion of 1d was observed after 3 days (Table 2, Entries 7 & 9).

Scheme 5: Ideal outcomes for the Brønsted Acid-mediated Pictet-Spengler cyclisations.

Table 2: Results of the Brønsted acid-mediated Pictet-Spengler cyclisations.

The conditions for monitoring the reaction by thin layer chromatography (TLC) required a highly polar eluent (1:5, methanol/dichloromethane, v/v). Use of a ninhydrin stain facilitated visualisation of the characteristic secondary amine, indicative of the THIQ formation. Conversion for the attempts to give 2a and 2b were determined by ¹H-NMR spectroscopy of the crude products of known mass. The conversion and overall yield was calculated by the total mass of the isolated crude product and the relative peak integrals of the imine protons and the C1-THIQ protons (Figure 2).

Figure 2: ¹H-NMR of the...

Despite the considerable conversion, by ¹H-NMR spectroscopy, of imine 1b to the corresponding THIQ (Table 2, Entry 5), no product was isolated. The high polarity of the substrates and products made observation and purification of the reaction mixtures and products by chromatographic techniques difficult. The imine substrate materials did not behave as expected by TLC, despite appearing clean by ¹H-NMR spectroscopy. Stambulli and coworkers made a comment on the high polarities and difficulties in handling the THIQ products.^[B] They suggested, based on IR spectroscopy, the substrate and products existed as zwitterions. The difficulties in purification of the THIQ by chromatography or recrystallisation were consistent with these literature comments.

The conversion of imine 1d to the THIQ 2c was only qualitatively assessable by TLC and ¹H-NMR spectroscopy (Table 2, Entries 8-9). The ¹H-NMR spectrum of the crude reaction mixture exhibited the characteristic C1-proton of the cyclised product, however comparison of the peak with the imine proton was complex (Table 2, Entry 8). The imine was situated in a region of high coincidence (~7.5-9.0 ppm) so precise integrations were improbable. The product 2c was somewhat purified under polar chromatographic conditions in moderate yield (Table 2, Entry 8). No conversion was observed in the reaction under a catalytic acid load (Table 2, Entry 9).

The results of the Brønsted acid-catalysed reactions were not entirely unexpected given the low reactivity of the imines and lack of examples and in the literature. The inability to efficiently monitor conversion and yield for the reactions in the presence of Brønsted acids encouraged movement onto the next stage of the project.

The Lewis Acid-Catalyst Screen

The next phase of the project involved the evaluation of a number of Lewis acids with potential to effect the formation of the THIQ scaffold. Given the literature procedures for the formation of the THIQ scaffold required excess amounts of strong Brønsted acids,^[Gao] it was expected that strong Lewis acids would be required for the PS reaction under catalytic conditions.
The lanthanoid (Ln) triflates are potent Lewis acids.^[Ln] The triflate counterion is highly effective at augmenting the Lewis acidity of the Ln(III) metals.^[I] Ytterbium(III) is particularly acidic, with its f¹³ configuration; it is one of the most Lewis acidic metals of the lanthanoid series.^[S] Ytterbium(III) triflate and other metal triflates, such as Cu(OTf)₂ and Zn(OTf)₂ are also also catalytically active in the PS reaction.[Gan] However, only Yb(OTf)₃ has been shown to effect formation of the THIQ scaffold.^[A] Metal triflates were an attractive choice for the Lewis acid-catalysed PS reaction. They also provided a link to the broader goal of this project: asymmetric catalysis. Metal triflates coordinate to chiral ligands, such as derivatives of bisoxazoline and binapthol (Figure X).^{[Ln, Acc]} These chiral complexes have been effectively used in asymmetric catalysis in cycloaddition reactions and the mechanistically similar Friedel-Crafts reaction.^{[Tang, Ln, py]}

Figure X: Ligands used in chiral metal triflate complexes (M = metal).

As shown previously, the electron-rich model substrates 1b and 1d were most likely to undergo the cyclisation reaction. The conditions for the Lewis acid screen for the PS reaction of these imines were adapted from Kobayashi and co-workers conditions for the PS reaction to give the THIQ scaffold using m-tyramine as a substrate Scheme X).^[A]

Scheme X: The conditions for the Pictet-Spengler reaction from m-tyramine and benzaldehyde reported by Kobayashi and co-workers.^[A]

Under the conditions employed, the four metal triflates were ineffective at forming the Pictet-Spengler products (2b-c) from the corresponding imines (Scheme 6). Starting material was recovered in all reactions.

Scheme 6: The Lewis acid-catalyst screen for the PS formation of 2b and 2c.

Like many examples of the PS formation of THIQ, the substrate materials in the template protocol were the aryl-amine and the aldehyde. The lack of cyclisation of the imine substrates prompted re-screening of Yb(OTf)₃ in reactions from phenethylamine and benzaldehyde. This also also resulted in recovery of the corresponding 1b imine (Scheme 7, Table 3).

Scheme 7: Related to Table 3.

Table 3: Results of the Yb(OTf)₃ catalysed PS formation of 2b.

Using an altered method and m-tyramine and benzaldehyde as substrates, Stambuli and co-workers reported similar a yield to Kobayashi and co-workers for the Yb(OTf)₃ catalysed formation of the THIQ scaffold.^[B] The altered protocol, which involved substitution of the dichloromethane (DCM) solvent for toluene and an increase in reaction temperature (25 °C to 110 °C) was also attempted with phenethylamine and aldehyde as substrates (Table 3: Entry 2). Again, 1b imine was recovered and zero conversion to the THIQ 2b was observed. In all attempts, the recovered material from the attempts at the Yb(OTf)₃ catalysed PS reactions from phenethylamine and benzaldehyde was identified, by ¹H-NMR spectroscopy as the corresponding imine 1b. These results suggested greater substrate specificity for the Lewis acid catalysed PS formation of the substituted THIQ.

The Lewis Acid-Catalysed Acyl-Pictet-Spengler Reaction

The lack of activity of the imines to catalytic PS conditions prompted adaptation of the model system to the acyl-PS variation. This variation involves formation of a highly electrophilic N-acylminium species. The electron withdrawing effects of the acyl group vastly increase the electrophilicity of the intermediate relative to the non-acylated iminium intermediate.^[Acyl] Nonetheless, there are limited known examples of the acyl-PS reaction effected by catalytic Lewis acids.^[ref]

One example of the Lewis acid-catalysed acyl-PS reaction to give the THIQ describes use of an AuCl₃ and AgOTf catalyst combination.^[6] This reaction was an ideal starting point for the Lewis acid catalysed acyl-PS reaction, particularly due to the identical starting materials and expected products to this project.

With minimal adaptation, the THIQ was produced with moderate yield (Scheme 8, Table 4: Entry 1). Presumably, the reduced yield in comparison to the reported yield was partially due to the use of HAuCl₄·3H₂O instead of AuCl₃. Following purification, the product, 3, and by-products, 4-nitrobenzaldehyde (4) and N-(3,4-dimethoxyphenethyl)acetamide (5) were isolated.

Scheme 8: The HAuCl₄·3H₂O/AgOTf catalysed acyl-PS reaction and associated by-products.

Table 4: Results of the Lewis acid-catalysed acyl-PS reactions.

The procedure was performed in the absence of acylating conditions resulting in recovery of the imine starting material, which was consistent with the literature.^[6]
This suggested the acyl group was essential for the reaction to proceed.

The protocol was adapted and carried using Yb(OTf)₃ Lewis acid catalyst (Scheme 9).

Scheme 9: The Yb(OTf)₃ catalysed acyl-PS formation of THIQ.

¹H-NMR spectroscopic analyses of the reaction product and by-products revealed sufficient separation of integrable peaks (Figure 3). Unlike monitoring of the conversion of the non-acyl-PS reactions, conversion and yield were gauged by comparison of the THIQ product and the hydrolysis by-products. Ideally, conversion was by use of 1,1,2,2,-tetrachloroethane (CDCl₃ δ 5.97 ppm) as an internal standard and peak integrals of the aldehyde proton of 4, the acyl- or CH2 environment of 5 (Figure 3, inset 2) and either the stereogenic proton or the aromatic proton of 3 (Figure 3, inset 1).

Error creating thumbnail: File with dimensions greater than 12.5 MP

Figure 3: Overlay of the acyl-PS THIQ product, 3 (black) and associated by-products, 4 (red) and 5 (blue). Black arrows indicate peaks integrated in the assay.

The first attempts to gauge the yield by ¹H-NMR spectroscopy of the crude product (in CDCl₃) were inconsistent with the isolated product yield. Spectroscopic analysis suggested a substrate to product conversion of 81%, while the isolated yield after chromatography was 51%. There were two problems associated with the spectroscopic assay: the presence of a CDCl₃ insoluble white solid and peak interference by 2,6-lutidine. The nature of the assay meant that complete dissolution of the crude product into the CDCl₃ solvent was crucial (see section 5.12). Incomplete dissolution of the crude product meant that the mass of the crude product assayed and the mass of the internal standard were non-comparable by ¹H-NMR spectroscopy. Furthermore, the 2,6-lutidine signals were partially coincidental to the integrable peaks of interest.

A post-reaction work-up procedure was developed in order to address the problems with the ¹H-NMR assay. The 2,6-lutidine peaks were minimised by including a citric acid wash of the crude product. The presence of chloride in the crude reaction mixture exacerbated the formation of a suspected hydrochloride salt in subsequent attempts to dissolve the extracted product in CDCl₃. This was successfully resolved by carrying out an alkaline wash following the acid-workup.

Use of the spectroscopic assay enabled a more efficient reaction optimisation process. The reaction outcomes (i.e. product and byproduct yields) were quickly determined without chromatography. As a result, the effects of the altered reaction conditions were taken into consideration for subsequent reactions (Table 4, Entries 3-5).

The protocol adapted from the HAuCl₄·H₂O/AgOTf co-catalysed reaction gave highly promising results (Scheme 9, Table 4, Entry 2). Anhydrous conditions were employed to minimise competing hydrolysis reactions. The first attempts at the Yb(OTf)₃ catalysed acyl-PS reaction suggested the rapid formation of the expected product and hydrolysis by-products. The extent of hydrolysis was minimised by reducing the initial reaction temperature and altering the order in which the reaction components were added (see section 5.11).

Conclusions and Future Work

The Yb(OTf)₃ catalysed acyl-PS reaction is an efficient route to the dimethoxy substituted THIQ scaffold. The high Lewis acidity of the Yb(OTf)₃ catalyst in conjunction with the acyl-PS reaction... . Required less preparation and was more atom efficient than the HAuCl₄·H₂O/AgOTf co-catalysed reaciton.

Due to time constraints, minimal optimisation of the Yb(OTf)₃ catalysed acyl-PS reaction was achieved. The and effects of solvent, reaction temperature, acylating agents and duration effects were not adequately explored. The primary solvents used in catalytic asymmetric PS reactions are toluene and dichloromethane, which are notably different from the polar acetonitrile that was used in the reactions.[ref]

The spectroscopic assay made evaluation of reaction conditions on the results more efficeint. Some improvements on the assay could involve construction of a standard curve of the product relative to the standard, so the crude product doesn't need to be dry. The current method is dependent on the mass of crude product assayed. If the sample wasn't dry, the calculated yield was inaccurate, due to the presence of solvent.

The post-reaction work-up procedure also needs further attention for yield optimisation.

The lack of imine reactivity in both the Yb(OTf)₃ catalysed and HAuCl₄·H₂O/AgOTf co-catalysed PS formation of the THIQ suggested the N-acyliminium intermediate was primarily responsible cyclisation. This makes sense because of the electron withdrawing properties N-acyliminium increasing the electrophilicity of the imine carbon.

A screen the catalytic potential of other metal triflates in the acyl-PS reaction is necessary. While Yb(OTf)₃ may be one of the most Lewis acidic catalysts, the d-block transition metal triflates are more known to do the asymmetric thing. But there is also a report of Yb(III)-pyBOX complexes.*Evaluation of other MOTfs in acyl PS reaction (Esp. CuOTf2). **enantioselective - pyBOX. Yb(OTf)3

• Since the AuCl3/AgOTf effected no cyclisation in the non-acyl PS reaction, it was suspected the successful cyclisation was effected by the acyl group. Evaluation of the achiral Brønsted acids in the acyl-PS reaction would also be appropriate.

• NMR kinetics thing

• Comment on robustness(?) of model system.

Experimental

The CDCl₃ and DMSO-d₆ solvents used in ¹H-NMR and ¹³C-NMR spectroscopic analyses were obtained from the Cambridge Isotope Laboratories. All melting points were recorded using on a Standford Research Systems OptiMelt (capillaries: ø = X-X mm, 90mm; ramp rate 1 °C min^-1). Glassware used in anhydrous reactions were dried >2 hours at 130 °C then cooled under inert gas before use. All molecular sieves were microwave activated and cooled under nitrogen before immediate use. Silica used in column chromatography procedures was X. Thin layer chromatography was performed on X.

N-benzylidene-2-phenylethanamine (1a)

To a stirring solution of benzaldehyde (4 mL, 40 mmol, 1 equiv.) in diethyl ether (10 mL) was slowly added 2-phenylethanamine (5 mL, 40 mmol, 1 equiv.). The clear yellow solution was stirred at room temperature for 5 hours, dried (MgSO₄) and concentrated under reduced pressure to yield a yellow oil that solidified on standing to give a yellow crystalline solid (8.3 g, 99%). M.p. 33-35 °C. ¹H-NMR (300 MHz; DMSO-d₆): δ 2.93 (t, J = 7.3 Hz, 2H), 3.81 (t, J = 7.3 Hz, 2H), 7.45-7.16 (m, 10H), 7.71 (dd, J = 6.6, 2.9 Hz, 2H), 8.26 (s, 1H). ¹³C-NMR (75 MHz; CDCl₃): δ 37.5, 63.2, 126.1, 128.1, 128.3, 128.6, 129.0, 130.6, 136.2, 139.9, 161.5. Relevant lab book entries: KAB18-1, KAB18-2.

N-benzylidene-2-(3,4-dimethoxyphenyl)ethanamine (1b)

To a stirring solution of benzaldehyde (3.1 mL, 30 mmol, 1 equiv.) in diethyl ether (10 mL) was slowly added 2-(3,4-dimethoxyphenyl)ethanamine (5.0 mL, 30 mmol, 1 equiv.). The mixture was stirred at ambient temperature for 5 hours, diluted with diethyl ether (20 mL), dried (MgSO₄) and concentrated under reduced pressure to yield a yellow oil that crystallised on standing (7.3 g, 90%). M.p. 31-33 °C. ¹H-NMR (300 MHz; CDCl₃): δ 8.15 (s, 1H), 7.73 (dd, J = 6.7, 3.0 Hz, 2H), 7.43 (dt, J = 5.3, 2.6 Hz, 3H), 6.83-6.77 (m, 3H), 3.89-3.84 (m, 8H), 2.99 (t, J = 7.2 Hz, 2H). ¹³C-NMR (75 MHz; CDCl₃): δ 161.5, 148.7, 147.4, 136.2, 132.6, 130.6, 128.6, 128.0, 120.9, 112.6, 111.2, 77.6, 77.1, 76.7, 63.3, 55.89, 55.71, 37.0. Relevant lab book entries: KAB19-1, KAB19-2.

N-(4-nitrobenzylidene)-2-phenylethanamine (1c)

To a stirring suspension of 4-nitrobenzaldehyde (6.0 g, 40 mmol, 1 equiv.) in diethyl ether (40 mL) was slowly added 2-phenylethanamine (5.0 mL, 40 mmol, 1 equiv.). The mixture was stirred at room temperature for 1 hour before a yellowish solid precipitated. The mixture was concentrated under reduced pressure and the residue recrystallised from diethyl ether to afford the pure product as pale yellow needles. M.p. 71-72 °C. ¹H-NMR (300 MHz; CDCl₃): δ 8.25 (d, J = 8.7 Hz, 2H), 8.20 (s, 1H), 7.85 (d, J = 8.7 Hz, 2H), 7.31-7.26 (m, 2H), 7.20-7.18 (m, 1H), 3.93 (t, J = 7.2 Hz, 2H), 3.04 (t, J = 7.3 Hz, 2H). ¹H-NMR (300 MHz; DMSO-d₆): δ 8.41 (s, 1H), 8.28 (d, J = 8.7 Hz, 2H), 7.96 (d, J = 8.7 Hz, 2H), 7.31-7.16 (m, 5H), 3.89 (t, J = 7.2 Hz, 2H), 2.96 (t, J = 7.3 Hz, 2H).¹³C-NMR (75 MHz; DMSO-d₆): δ 160.1, 149.0, 142.1, 140.1, 129.30, 129.26, 128.7, 126.5, 124.4, 62.4, 37.1. Relevant lab book entry: KAB22-1.

2-(3,4-dimethoxyphenyl)-N-(4-nitrobenzylidene)ethanamine (1d)

4-Nitrobenzaldehyde (3.7 g, 24 mmol) was suspended in diethyl ether (50 mL). 2-(3,4-Dimethoxyphenyl)ethanamine (4.0 mL, 25 mmol) was added dropwise, with stirring. The mixture was left to stir at ambient temperature for 6 hours resulting in the precipitation of a fine light yellow solid. The solvent was removed under reduced pressure to give the crude product as a fine, yellow powder (8.0 g, 103%). Recrystallisation of the crude product from ethanol (~200 mL) afforded the pure product as yellow needles (7.1 g, 23 mmol, 92%). M.p. 123-124 °C. ¹H-NMR (200 MHz; CDCl₃): δ 8.29-8.23 (m, 2H), 8.19 (s, 1H), 7.86 (d, J = 8.8 Hz, 2H), 6.78-6.72 (m, 3H), 3.91 (td, J = 7.1, 1.1 Hz, 2H), 3.85 (s, 3H), 3.81 (s, 3H), 2.99 (t, J = 7.1 Hz, 2H). ¹H-NMR (300 MHz; DMSO-d₆): δ 8.40 (s, 1H), 8.29 (d, J = 8.7 Hz, 2H), 7.98 (d, J = 8.7 Hz, 2H), 6.84 (dd, J = 4.8, 3.3 Hz, 2H), 6.75 (dd, J = 8.2, 1.4 Hz, 1H), 3.86 (t, J = 7.1 Hz, 2H), 3.69 (d, J = 2.1 Hz, 6H), 2.90 (t, J = 7.2 Hz, 2H). ¹³C-NMR (75 MHz; DMSO-d₆): δ 160.0, 148.9, 147.6, 142.1, 132.5, 129.2, 124.4, 121.1, 113.3, 112.2, 62.7, 55.94, 55.80, 36.6. Relevant lab book entries: KAB23-1, KAB23-2.

Attempts at the synthesis of 1-phenyl-1,2,3,4-tetrahydroisoquinoline (2a)

Typical procedure (e.g. Entry 2): To a stirring solution of methanesulfonic acid (3.0 mL, 46 mmol) at 0 °C was added N-benzylidene-N-phenethylamine (1a) (0.48 g, 2.3 mmol). The now yellow solution was heated to 60 °C. After 30 minutes the solution had turned dark red. Stirring was continued at 60 °C for 68 hours. The mixture was poured over an ice water slurry (~15 mL) and made alkaline by the addition of sodium hydroxide solution (5 M), resulting in the formation of a white solid. The mixture was extracted with diethyl ether (3 × 30 mL). The organic fractions were combined, dried (MgSO₄) and concentrated under reduced pressure to yield a brown oil (410 mg). ¹H-NMR of the isolated material was of the 1a starting material.
In the case of entry 3, the yield was calculated by comparing the integrals of the 1a imine proton (CDCl₃ δ 8.22 ppm) with the 2a C1 proton (CDCl₃ δ 5.14 ppm).[ref] Relevant lab book entries: KAB20-2, KAB20-3, KAB20-4.

Brønsted acid synthesis of 6,7-dimethoxy-1-phenyl-1,2,3,4-tetrahydroisoquinoline (2b)

Methanesulfonic Acid

To solution of N-[2-(3,4-Dimethoxyphenyl)ethyl]-1-phenylmethanimine (1b) (418 mg, 1.55 mmol) in toluene (40 mL) was added methanesulfonic acid (0.10 mL, 1.6 mmol) at 0 °C. The mixture was stirred at ambient temperature for 3.5 hours. Ethyl acetate (50 mL) was added. The mixture was quenched with saturated sodium bicarbonate solution (50 mL). The organic layer was separated and the aqueous layer was extracted with ethyl acetate (2 × 50 mL). The organic layers were combined, dried (MgSO₄) and concentrated under reduced pressure yielding a yellow oil (>100%). ¹H-NMR spectroscopy of the isolated material confirmed the presence of 6,7-dimethoxy-1-phenyl-1,2,3,4-tetrahydroisoquinoline (2b).^[6] Relevant lab book entry: KAB21-1.

Trifluoroacetic Acid

To a stirring solution of N-[2-(3,4-Dimethoxyphenyl)ethyl]-1-phenylmethanimine (1b) (1.9 g, 7.1 mmol) in toluene (40 mL) was added trifluoroacetic acid (30 mL, 0.36 mol). The dark yellow solution was refluxed for 22 hours. etc. etc. etc. ¹H-NMR of the isolated material confirmed the presence of 6,7-Dimethoxy-1-phenyl-1,2,3,4-tetrahydroisoquinoline (2b).^[6] Relevant lab book entries: KAB21-2, KAB21-3, KAB21-4.

Brønsted acid synthesis of 6,7-dimethoxy-1-(4-nitrophenyl)-1,2,3,4-tetrahydroisoquinoline (2c)

To a pale yellow suspension of partially dissolved 2-(3,4-dimethoxyphenyl)-N-(4-nitrobenzylidene)ethanamine (1d) (401 mg, 1.28 mmol) in toluene (40 mL) was slowly added trifluoroacetic acid (0.20 mL, 2.6 mmol). The now clear, orange-red solution was reflux heated to 110 °C for 6.5 h. The mixture was allowed to cool then made alkaline with sodium hydroxide solution (6 M) to pH 9-10. The organic layer was isolated and the aqueous layer was extracted with ethyl acetate (3 × 30 mL). The organic layers were combined, dried (MgSO₄) and concentrated under reduced pressure to give a dark yellow oil (524 mg, 131%). Purification of the oil by silica gel column chromatography (5% methanol/dichloromethane) yielded 6,7-dimethoxy-1-(4-nitrophenyl)-1,2,3,4-tetrahydroisoquinoline (2c) as a yellow solid (112 mg, 28%).Relevant lab book entries: KAB24-1, KAB24-2, KAB24-3, KAB24-4.

Procedure for the Lewis Acid-Catalyst Screen

Substrate stock solutions (0.20 M in dichloromethane) were prepared N-benzylidene-2-(3,4-dimethoxyphenyl)ethanamine (0.20 M) and 2-(3,4-dimethoxyphenyl)-N-(4-nitrobenzylidene)ethanamine. To a mixture of the metal triflates (15-40 mg) at 20 mol% and microwave activated 3 Å powdered molecular sieves were the substrate stock solutions added. The mixtures were stirred at ambient temperature for 24 hours. The mixtures were quenched with saturated sodium bicarbonate solution and extracted with ethyl acetate. The organic layers were combined, dried (MgSO₄) and concentrated under reduced pressure. ¹H-NMR spectroscopy of the crude reaction mixtures indicated the products obtained consisted mostly of the corresponding starting material. Relevant lab book entry: KAB21-5, KAB21-6, KAB21-7, KAB21-8, KAB24-5, KAB24-6, KAB24-7 & KAB24-8.

Attempts at the Yb(OTf)₃ catalysed synthesis of 6,7-dimethoxy-1-(4-nitrophenyl)-1,2,3,4-tetrahydroisoquinoline (2c)

Procedure 1^[A]

To a mixture of Yb(OTf)₃ (121 mg, 0.194 mmol, 0.2 equiv.) and microwave activated 3 Å powdered molecular sieves (~20 mg) was added dry dichloromethane (30 mL). Benzaldehyde (0.10 mL, 0.97 mmol, 1 equiv.) and 2-(3,4-dimethoxyphenyl)ethanamine (0.16 mL, 0.97 mmol, 1 equiv.) were added. The reaction mixture was stirred under nitrogen for 24 hours. Saturated sodium bicarbonate solution (30 mL) was added to quench the reaction. The organic layer was separated and the alkaline aqueous fraction was extracted with ethyl acetate (3 × 50 mL). The organic fractions were combined, dried (MgSO₄) and concentrated under reduced pressure yielding a yellow oil (390 mg, 150%). ¹H-NMR spectroscopy of the oil indicated the isolated product contained water and a 1:0.15 mixture of imine 1d and 4-nitrobenzaldehyde (5). Relevant lab book entries: KAB25-1.

Procedure 2^[B]

Relevant lab book entry: KAB25-2 To a mixture of Yb(OTf)₃ (120 mg, 0.194 mmol) and microwave activated 3 Å powdered molecular sieves (~20 mg) was added dry toluene (30 mL) under a nitrogen atmosphere. Benzaldehyde (0.10 mL, 0.97 mmol) and 2-(3,4-dimethoxyphenyl)ethanamine (0.16 mL, 0.97 mmol) were added. The reaction mixture was reflux heated to 110 °C for 96 hours. The mixture was made alkaline by the addition of saturated sodium bicarbonate solution. The organic layer was separated and the aqueous fraction was extracted with ethyl acetate (3 × 30 mL). The organic fractions were combined, dried (MgSO₄) and concentrated under reduced pressure to yield a yellow solid (360 mg, 130%). ¹H-NMR spectroscopy of the oil indicated the isolated product was mostly composed of imine 1b.

HAuCl₄·3H₂O/AgOTf catalysed synthesis of 1-(6,7-dimethoxy-1-(4-nitrophenyl)-3,4-dihydroisoquinolin-2(1H)-yl)ethanone (3)^[6]

A solution of gold(III) chloride trihydrate (31 mg, 0.079 mmol, 0.01 equiv.) and silver(I) trifluoromethanesulfonate (30 mg, 0.12 mmol, 0.02 equiv.) in acetonitrile (15 mL) was vigorously stirred at ambient temperature (~17 °C) for 1 hour. To the now yellow reaction mixture was added a pale yellow solution of 2-(3,4-dimethoxyphenyl)-N-(4-nitrobenzylidene)ethanamine (1d) (1.9 g, 6.1 mmol, 1 equiv.), acetyl chloride (0.40 mL, 6.1 mmol, 1 equiv.) and 2,6-lutidine (0.70 mL, 6.1 mmol, 1 equiv.) in acetonitrile (250 mL). The reaction mixture was stirred for 14 hours at ambient temperature (~12 °C), concentrated under reduced pressure and purified by silica gel column chromatography (50-100% ethyl acetate/hexane, v/v). The expected product (3d) was isolated (0.86 g, 41%) in addition to the byproducts,4 and 5. Relevant lab book entries: KAB26-1 & KAB26-2, KAB26-3, KAB26-10.

1-(6,7-dimethoxy-1-(4-nitrophenyl)-3,4-dihydroisoquinolin-2(1H)-yl)ethanone (3d)

M.p. 177-179 °C. Two amide rotamers (91:9). Signals corresponding to the major rotamer: ¹H-NMR (200 MHz; CDCl₃): δ 8.12 (d, J = 8.7 Hz, 2H), 7.42 (d, J = 8.7 Hz, 2H), 6.90 (s, 1H), 6.69 (s, 1H), 6.48 (s, 1H), 3.89 (s, 3H), 3.76 (s, 3H), 3.74-3.70 (m, 1H), 3.42-3.27 (m, 1H), 2.95 (ddt, J = 15.9, 10.7, 5.2 Hz, 1H), 2.81-2.71 (m, 1H), 2.18 (s, 3H). Signals corresponding to the minor rotamer: ¹H-NMR (200 MHz; CDCl₃): δ 8.17 (s, 2H), 6.60 (s, 1H), 5.94 (s, 1H), 2.32 (s, 3H). Spectroscopic data matched those in the literature.^[6]
4-nitrobenzaldehyde (4)
M.p. 103.2 - 104.3 °C. ¹H-NMR (300 MHz; CDCl₃): δ 10.18 (s, 1H), 8.42 (d, J = 8.6 Hz, 2H), 8.10 (d, J = 8.5 Hz, 2H).¹³C-NMR (75 MHz; CDCl₃): δ 190.2, 140.0, 130.5, 124.3. Spectroscopic data matched those in the literature.^[6]

N-(3,4-dimethoxyphenethyl)acetamide (5)

M.p. 77-79 °C. ¹H-NMR (300 MHz; CDCl₃): δ 6.82-6.79 (m, 1H), 6.74-6.71 (m, 2H), 5.66 (s, 1H), 3.86 (s, 3H), 3.86 (s, 3H), 3.48 (q, J = 6.6 Hz, 2H), 2.76 (t, J = 7.0 Hz, 2H), 1.94 (s, 3H). ¹³C-NMR (75 MHz; CDCl₃): δ 170.1, 149.0, 147.7, 131.4, 120.6, 114.7, 111.9, 111.4, 55.91, 55.86, 40.8, 35.2, 23.3.

Yb(OTf)₃ catalysed synthesis of 1-(6,7-dimethoxy-1-(4-nitrophenyl)-3,4-dihydroisoquinolin-2(1H)-yl)ethanone (3)

The acetonitrile (HPLC grade) was dried over microwave activated 3 Å molecular sieves (2.5-5.0 mm, 30 %(w/v)) for >24 hours. All glassware was ovendried (130 °C) for >2 hours prior to use. 2,6-lutidine was dried over 3 Å molecular sieves (diameter, 2.5-5.0 mm, 50 %(w/v)).

Procedure 1

2-(3,4-dimethoxyphenyl)-N-(4-nitrobenzylidene)ethanamine (X g, X mmol, 1 equiv.) was dissolved in anhydrous acetonitrile (X mL). Relevant lab book entries: KAB26-4

Procedure 2

Added citric acid workup. Relevant lab book entry: KAB26-5

Procedure 3

Acetonitrile/liquid N₂ bath. KAB26-6 KAB26-7 KAB26-9

Procedure 4

To a mixture of 3 Å molecular sieves (~30 g) in dry acetonitrile (160 mL), under nitrogen, was added 2-(3,4-dimethoxyphenyl)-N-(4-nitrobenzylidene)ethanamine (1.50 g, 4.77 mmol, 1 equiv.). Once dissolved, the mixture was cooled in a brine ice bath. Acetyl chloride (0.34 mL, 4.8 mmol, 1 equiv.) and 2,6-lutidine (0.55 mL, 4.8 mmol, 1 equiv.) were added, dropwise. Yb(OTf)₃ (0.034 g, 0.048 mmol, 0.01 equiv.) was added. Thereaction mixture was allowed to warm to ambient temperature (~12 °C) and stirred under argon for 23 hours. The mixture was filtered through a bed of Celite, eluting with ethyl acetate (~50 mL). The filtrate was washed with saturated sodium bicarbonate solution (40 mL). The aqueous layer was extracted with ethyl acetate (3 × 40 mL). The organic fractions were combined, dried (MgSO₄) and concentrated under reduced pressure to yield a yellow oil that partially crystallised on standing (1.8 g, 106%). The crude product was dissolved in hot methanol, dry loaded onto a silica gel column (ø = 6.5 cm, 15 cm) and purified by chromatography (70-100% ethyl acetate/hexane) yielding the expected product as a yellow powder (1.3 g, 77%). Relevant lab book entry: KAB26-11.

Typical procedure for the ¹H-NMR assays of the Yb(OTf)₃ catalysed acyl-Pictet-Spengler reactions

Tetrachloroethane (8.0 mg, 4.8 × 10^-5 mol) was added to a known amount of crude product (e.g. 10 mg). To the mixture was added CDCl₃. The peaks were integrated and normalised based on the integrated signal and number of protons. The number of moles of product or by-product was determined by comparing the ratio of integrals with the known moles of added tetrachloroethane. The amount in moles was converted to mass, which was then divided by the amount of crude product dissolved in the CDCl₃.

References

[1] Über die Bildung von Isochinolin-derivaten durch Einwirkung von Methylal auf Phenyl-äthylamin, Phenyl-alanin und Tyronsin, A. Pictet and T. Spengler, Ber. Dtsch. Chem. Ges. 1911, 44, 2030-2036. (DOI: 10.1002/cber.19110440309) Paper

[2] Strictosidine Synthase: Mechanism of a Pictet−Spengler Catalyzing Enzyme, J. J. Maresh, L.-A. Giddings, A. Friedrich, E. A. Loris, S. Panjikar, B. L. Trout, J. Stöckigt, B. Peters and S. E. O'Connor, J. Am. Chem. Soc. 2007, 130, 710-723. (DOI: 10.1021/ja077190z) Paper

[3] The Pictet–Spengler Reaction in Nature and in Organic Chemistry, J. Stöckigt, A. P. Antonchick, F. Wu and H. Waldmann, Angew. Chem., Int. Ed. 2011, 50, 8538-8564. (DOI: 10.1002/anie.201008071) Paper

[4] Reconstruction of the saframycin core scaffold defines dual Pictet-Spengler mechanisms, K. Koketsu, K. Watanabe, H. Suda, H. Oguri and H. Oikawa, Nat. Chem. Biol. 2010, 6, 408-410. (DOI: 10.1038/nchembio.365) Paper

[5] [Jac04] Highly Enantioselective Catalytic Acyl-Pictet-Spengler Reactions, M. S. Taylor and E. N. Jacobsen, J. Am. Chem. Soc. 2004, 126, 10558-10559. (DOI: 10.1021/ja046259p) Paper

[6] [C] Development of the Pictet-Spengler Reaction Catalyzed by AuCl₃/AgOTf, S. W. Youn, J. Org. Chem 2006, 71, 2521-2523. (DOI: 10.1021/jo0524775) Paper

[7] [Q] Lewis Acid-Catalyzed Selective Synthesis of Diversely Substituted Indolo- and Pyrrolo[1,2-a]quinoxalines and Quinoxalinones by Modified Pictet–Spengler Reaction, A. K. Verma, R. R. Jha, V. K. Sankar, T. Aggarwal, R. P. Singh and R. Chandra, Eur. J. Org. Chem. 2011, 34, 6998-7070. (DOI: 10.1002/ejoc.201101013) Paper

[8] Chemistry and Biology of the Tetrahydroisoquinoline Antitumor Antibiotics, J. D. Scott and R. M. Williams, Chem. Rev. 2002, 102, 1669-1730. (DOI: 10.1021/cr010212u) Paper

[9] [NMDA] Affinity of 1-aryl-1,2,3,4-tetrahydroisoquinoline derivatives to the ion channel binding site of the NMDA receptor complex, M. Ludwig, C. E. Hoesl, G. Höfner and K. T. Wanner, Eur. J. Med. Chem. 2006, 41, 1003-1010. (DOI: 10.1016/j.ejmech.2006.03.005) Paper

The rest. $ = still needs to be formatted correctly (section will be deleted on completion of the report)

[C1] Imine 1d showed trace amounts of the 4-nitrobenzaldhyde starting material, which promted recrystallisation of the pure imine from ethanol to yield the final product. Imine 1c required no further purification.

[I] Evaluation of the relative Lewis acidities of lanthanoid(III) compounds by tandem mass spectrometry, H. Tsuruta, K. Yamaguchi and T. Imamoto, Chem. Commun. 1999, 1703-1704. (DOI: 10.1039/A905569J) Paper

[Gan] Highly efficient Lewis acid-catalysed Pictet-Spengler reactions discovered by parallel screening, N. Srinivasan and A. Ganesan, Chem. Commun. 2003, 916-917. (DOI: 10.1039/B212063A) Paper

[A] Catalytic Pictet-Spengler reactions using Yb(OTf)₃, K. Manabe, D. Nobutou and S. Kobayashi, Bioorg. Med. Chem. 2005, 13, 5154-5158. (DOI: 10.1016/j.bmc.2005.05.018) Paper

[S] Principal properties (PPs) for lanthanide triflates as Lewis-acid catalysts, C. G. Fortuna, G. Musumarra, M. Nardi, A. Procopio and G. Sindona, S. Scirè, J. Chemom. 2006, 20, 418-424. (DOI: 10.1002/cem.1016) Paper

[Tang] Trisoxazoline/Cu(II)-catalyzed asymmetric intramolecular Friedel-Crafts alkylation reaction of indoles, J.-L. Zhou, M.-C. Ye, X.-L. Sun and Y. Tang, Tetrahedron 2009 65, 6877-6881. (DOI: 10.1016/j.tet.2009.06.071) Paper

[B] Calcium-Promoted Pictet-Spengler Reactions of Ketones and Aldehydes, M. J. V. Eynden, K. Kunchithapatham and J. P. Stambuli, J. Org. Chem. 2010, 75, 8542-8549. (DOI: 10.1021/jo1019283) Paper

[D] (i) Pictet-Spengler condensation reactions catalyzed by a recyclable H(+)-montmorillonite as a heterogeneous Brønsted acid, Wang, Y., Z. Song, et al. Science China-Chemistry 2010, 53(8), 562-568. (DOI: 10.1007/s11426-010-0073-4) Paper; (ii) One-step preparation of 1-substituted tetrahydroisoquinolines via the Pictet–Spengler reaction using zeolite catalysts, A. Hegedüs and Z. Hell, TetLett 2004, 45, 8553–8555. (DOI: 10.1016/j.tetlet.2004.09.097) Paper

[Acyl] Cyclizations of N-Acyliminium Ions, B. E. M, H.-C. Zhang, J. H. Cohen, I. J. Turchi and C. A. Maryanoff, Chem. Rev. 2004, 104, 1431–1628. (DOI:10.1021/cr0306182) Paper

[Ln] Rare-Earth Metal Triflates in Organic Synthesis, S. Kobayashi, M. Sugiura, H. Kitagawa and W. W.-L. Lam, Chem. Rev. 2002, 102, 2227–2302. (DOI: 10.1021/cr010289i) Paper

[H] Badiola, K. A., Robertson, M. N., Tarselli, M. A., Todd, M. H. 2012, The Catalytic, Asymmetric Pictet-Spengler Reaction wiki, openwetware, 9 June 2012, http://openwetware.org/wiki/Todd:Catalytic%2C_Asymmetric_Pictet-Spengler_Reaction.

[Gao] (i) Synthesis of carbon-11 and fluorine-18 labeled N-acetyl-1-aryl-6,7-dimethoxy-1,2,3,4-tetrahydroisoquinoline derivatives as new potential PET AMPA receptor ligands, M. Gao, D. Kong, A. Clearfield, Q.-H. Zheng, Bioorg. Med. Chem. Lett. 2006, 16, 2229-2233. (DOI: 10.1016/j.bmcl.2006.01.042) Paper; (ii) Synthesis of stable azomethine ylides by the rearrangement of 1,3-dipolar cycloadducts of 3,4-dihydroisoquinoline-2-oxides with DMAD, N. Coşkun and S. Tunçman, Tetrahedron 2006, 62, 1345-1350. (DOI: 10.1016/j.tet.2005.11.040) Paper; (iii) Synthesis, antibacterial activity and QSAR studies of 1,2-disubstituted-6,7-dimethoxy-1,2,3,4-tetrahydroisoquinolines, R. K. Tiwari, D. Singh, J. Singh, A. K. Chhillar, R. Chandra and A. K. Verma, Eur. J. Med. Chem. 2006, 41, 40-49. (DOI: 10.1016/j.ejmech.2005.10.010) Paper

[Acc] Chiral Bis(oxazoline) Copper(II) Complexes: Versatile Catalysts for Enantioselective Cycloaddition, Aldol, Michael, and Carbonyl Ene Reactions, J. S. Johnson and D. A. Evans, Acc. Chem. Res. 2000, 33, 325-335. (DOI: 10.1021/ar960062n) Paper

[py] Chiral 2,6-Bis(oxazolinyl)pyridine−Rare Earth Metal Complexes as Catalysts for Highly Enantioselective 1,3-Dipolar Cycloaddition Reactions of 2-Benzopyrylium-4-olates, H. Suga, K. Inoue, S. Inoue, A. Kakehi and M. Shiro, J. Org. Chem. 2005, 70, 47-56. (DOI: 10.1021/jo049007f) Paper