Kapitel 4 – Rättssäkerhet i Sverige och EU

4.2 Rättssäkerhet i svensk rätt

4.2.6 Rättssäkerhetsbegreppet inom skatterätten .1 Inledning

first, we investigate the phosphine ligands, employing 1 as substrate. THC 2a was isolated as a single E-isomer in 38% yield using PCy3. Bidentate phosphines yielded 2a with slightly better performances. The use of cheaper PPh3 promoted the catalytic transformation with

moderate yields and high chemoselectivity, even at lower temperature. Then we investigated the palladium source. Pd(dba)2 could be applicable, although to a lower extent, even though, a Pd(II) precursor, such as the

popular Pd(OAc)2, was not a suitable catalyst for this transformation. [127–132]

Buchwald’s ligands and substituted triarylphosphine analogues were unable to improve the catalytic system. [126] Hence, 2a was isolated in 78% yield by increasing the catalyst loading. The reaction without benzoic acid provided traces of 2a, highlighting the crucial role of the former. In all cases, we observed complete control of the alkene configuration, retrieving 2a as a single E-isomer.

We than studied the generality of the reaction. A family of diversely functionalized propargylic tryptamines 1 was then submitted to the optimized catalytic system.

Figure 6: Single cristal XRay diffraction of 2a

Scheme 20 Scope of propargyl fragment

Alkyl- and arylalkyne derivatives were tolerated, leading to the corresponding products 2b-c in 54 and 85% yields, respectively. A broad family of 3-arylprop-2-yn-1-ylindole derivatives 1 was efficiently converted to THCs 2. Thus, the method is suitable for the synthesis of various fluorinated THCs. Substrates bearing EWGs performed better with respect to the corresponding electron-rich ones. Tryptamine derivatives bearing biphenyls, naphthalenes, and heteroarenes, such as thiophene, delivered the corresponding THCs in good yields (2j−l, 50−73%). Reactions involving arylalkynes (2c−l) deliver the desired products only, and unreacted starting materials can be recovered by chromatography.

Then, we studied substituted tryptamines. Protection of tryptamine with either methanesulfonyl (Ms) or p-Cl-benzensulfonyl did not affect the outcome of the

NH NTs

NH NTs

R R1

10 mol % Pd(PPh3)4 30 mol % BzOH 20 mol % PPh3 toluene, 100 °C, 16 h

NH NTs

Me 2a 78 %

NH NTs

Et 2b 54 %

NH NTs

2c 85 % 1 mmol scale 72 %

NH NTs

2d 84 % F

NH NTs

2e 88 %

F NH

NTs

Me 2f 50 %

NH NTs

2g 65 % 1 mmol scale 69 %

F

Me

NH NTs

2h 64 % OMe

NH NTs

2i 69 % CF3

NH NTs

2j 73 %

NH NTs

2k 50 %

NH NTs

2l 71 % S

1a-l 2a-l

catalytic reaction (4a-4b).

Scheme 21 Scope of indole substitution

Electron-donating functional groups such as alkyl, aryl, silyl ethers and ethers at the C(5)- position of the indole were

well tolerated, delivering THCs 4c - g with good to excellent yields (57−93%). The method allowed the synthesis of THCs 4h − k with electron-withdrawing groups located at the C(5)- and C(6)-position with moderate to good yields (51−79%).

The robustness of the sequence was further mirrored by the synthesis of a symmetrical THC dimer. Noteworthy, the synthesis of 4l proceeds smoothly with high chemoselectivity and a synthetically useful yield (56%).

NH NTs

NH NR2

R R1

10 mol % Pd(PPh3)4 30 mol % BzOH

20 mol % PPh3 toluene, 100 °C, 16 h

NH NMs

Me 4a 69 %

NH

NSO2Ar

Et Ar = 4-Cl-C6H4

4b 68 %

NH NTs

Me 4c 63 % Me

NH NTs

Me 4d 62 %

NH NTs

Me 4e 59 % TBSO

NH NTs

Me 4f 57 % MeO

NH NTs

4g 93 % MeO

HN NTs

Me 4h 70 % F

NH NTs

Me 4i 59 % Cl

NH NTs

4j 79 % Cl

NH NTs

Me 4k 51 % R1

3a-k 4a-k

Cl

Figure 7: Single cristal XRay diffraction of 4g

Equation 1 Sinthesis of dimeric THC

Then, different experiments were carried out to rationalize the reaction mechanism and gain insights on its complete chemo- and site-selectivities. THC 4m was isolated in 70% yield using allenamide 3m under typical reaction conditions. This result is consistent with the intermediate formation of an allenamide in the sequence. Then, the influence of indole N-substitution was investigated.

We observed the formation of acyclic diene 4n′ and no traces of THC 4n by testing precursor 3n. This result shows that the free N−H of the indole is crucial for the annulation step and shows once more the virtues of a protecting-group free synthesis. [133] Finally, the attempted intramolecular dearomatization of propargylic precursor 3o, in which the indole C(2) is substituted with a methyl group, failed to provide any product. [134] This result was highly unexpected on the basis of the literature precedents and led us hypothesize that a C−H activation step could operate in our case. We are unawere of the literature precedents on C−H activation/alkylation at C(2) of unprotected indoles. [135–139]

20 mol % Pd(PPh3)4 60 mol % BzOH 40 mol % PPh3 toluene, 100 °C, 16 h TsN

HN NTs

NH MeO

OMe

NTs

NH

TsN MeO HN

OMe

5a 6a 56 %

Scheme 22 Experimental mechanism investigation

Thus, we resorted to DFT modelling to solve the riddle. The investigation began using the M06 functional in combination with either lacvp(d) and Def2-svp(d) basis sets, which proved to be reliable methods to describe elaborate palladium-catalysed sequences (Scheme 19). [140–142] Complete pathways were modelled with PMe3 as ligand both in the gas phase and using toluene as implicit solvent to reduce further the odds of modelling artefacts. Different functionals were also tested and the key steps were reoptimized with PPh3. Overall, coherent results were obtained in all cases, which led us to propose the mechanism of Scheme 23 to account for the formation of THC 2 in these sequences. A reaction of the carboxylic acid with Pd(0) complex I can deliver trans-Pd(II) hydride II. Its cis-peer is less stable and furthermore requires a higher barrier too. Endoergonic phosphine replacement by the substrate yields complex III, which evolves to vinyl species IV via migratory insertion. Barriers for this step were expectedly low (between +2.8 and +6.8 kcal/mol among the different models). The

NH NTs

NH NTs

7a 8a 70 %

standard condition

N NTs

N NTs

7b 8b

standard condition

Me Me Me Me

N NTs

8b' 67 % Me

NH

NTs

N 7c

standard condition

Ph Me

NTs Me Ph

8c

subsequent β-elimination affords allenamide complex V. This proved to be the most energy-demanding barrier of the whole sequence because of the limited rotational flexibility of the allylic methylene (ΔG of +24.9 kcal/mol; up to +28.1 at the B3LYP level). Complex V is more stable than its alkyne peer III (by −4.8 kcal/mol). This suggests that allenamides bind Pd(II) hydrides more strongly than an internal alkyne. A second insertion into the Pd−H bond gives allyl complex VI through a low barrier process (+5.2 kcal/mol in ΔG). Replacement of the acetate ligand by phosphine provides complex VII, in which the metal presents a slipped η2-indole coordination, and the carboxylate is engaged in hydrogen bonding with the indole N−H group. This is crucial to favor the sequential C−H activation, which occurs through an outer-sphere CMD pathway

[143–149] (barriers of +18.2 and +18.8 kcal/mol in ΔG with PMe3 and PPh3, respectively). This is consistent with the positive role of additional ligands on yield. It is worth noting that direct C−H activation of indole is, on the contrary, usually prevented by the presence of a free N−H group, which sunk the basicity of metal-bound carboxylates via hydrogen bonding.

Scheme 23 Most favourable modelled mechanism

Modeling an inner-sphere CMD indeed provided higher barriers (+29.2 kcal/mol in ΔG). Try as we might, we failed to obtain any stationary point for Pictet−Spengler-like pathways from either complex V or VI. Scans of these pathways show a linear increase of E only (up to above 40 kcal/mol), suggesting that indole dearomatization does not lead to any stable intermediate in these cases. These results parallel recent computational ones on C-palladations previously thought to occur via electrophilic aromatic substitutions. [150]

Heck-like insertion on indole from VII provided a sky-high barrier of +50.5 kcal/mol. Taken together within the framework of the energy span model, [151]

these results strongly suggest that indole functionalization occurs via C−H activation in this sequence. The results also highlight the dual nature of the carboxylic acid in this cascade. It initially serves to generate the Pd hydride that triggers alkyne isomerization. The resulting carboxylate then becomes crucial as well. It plays the role of a base assisting the metal in the C−H activation, therefore acting in a catalytic fashion. The desired product is eventually released by C−C

0 10,6

16,1

-9,8 15,1

5,8 11,4

-16,9 -13,1 5,3

-3,8 12,4

-25,1 M06/Def2-svp(d)

CPCM (toluene)

∆Gs @ 298.15 K values in Kcal/mol With R = Me

Pd R3P

PR3 AcO H

II Pd

R3P PR3

I AcOH

N H

NMs

Pd Me R3P

H OAc H

H

III

NH NMs H

H

IV Pd H R3P

N H

NMs

V H

H Me Pd R3P OAc

H

HN

VI H

NMs

Pd H

Me H O O R3P

N

VII H

NMsH

Me H Pd O O H

N VIII

NMsH

Me H Pd

O H

OH R3P PR3

NH NMs

Me product

reductive elimination from metallacycle VIII. This step has a barrier comparable to those of similar Pd(II) complexes (+16.2 kcal/mol in ΔG).

In document Förbudet mot rättsmissbruk i EU-rätten En förändrad avvägning mellan rättssäkerhet och rättvisa i den svenska skatterätten Fritz, Maria (Page 180-192)