Thienopyrimidine complexities, Electrophilic aromatic quandaries and Suzuki concerns

21 Sep
Published by JimCronshaw

Jim Cronshaw from the Todd Laboratory at the University of Sydney here again for my two monthly update.

Previously I had reported puzzlement at why the proton at the 6-position of the thieno[3,2-d]pyrimidine (aka: the α position) was proving difficult to remove with BuLi. Both n and s BuLi had been utilised at different temperatures to no effect: That proton was not going to removed. When the nitrogen was masked with a morpholine group a different story emerged. Treatment of this substrate with n-BuLi followed by iodine lead to the iodination of the desired position alongside extensive decomposition (Scheme 1).




Scheme 1: Iodonated Morpholine derivative


Interestingly, the ESI mass spectrum of this compound showed a peak corresponding to the following compound:

Figure 1: Iodonated Thienopyrimidine

Based upon this information, it was thought that perhaps the thienopyrimidine with a hydroxyl group at the 4-position might be suitable for iodination. This turned out to be the case, and the iodinated product (6-iodothieno[3,2-d]pyrimidin-4-ol) was synthesised in 30% yield (Scheme 2).

Scheme 2: Iodonated Thienopyrimdine

The ability of this reaction to proceed shed some light on why the earlier reaction, involving proton removal in the presence of the amine, didn’t work. Assuming that the above thienopyrimidine ring exists in its tautomeric keto form (as shown), then the pKa’s suggest that the proton on the 1-position of the thienopyrimidine ring shown in Figure 2 ought to be more acidic than the amine protons of thieno[3,2-d]pyrimidin-4-amine (Figure 2).

Figure 2: pKa's of relevant protons

Though the iodinated thienopyrimidine was now in my possession, I was still unsure as to where the substitution had taken place. Conventional wisdom would dictate that the substitution occurred at the 6-position, but a summary of the chemistry of thienopyrimidines indicated that thieno[3,2-d]pyrimidines are particularly reactive at the 7-position.  Two lines of investigation were employed to determine where the substitution had occurred.

First, 2D NMR studies (HSQC and HMBC) were utilised. Known values for the chemical shifts of substituted carbons in thiophene rings were necessary to make a definitive judgement on where the substitution had occurred. Without these, no judgement with any certainty could be passed. An X-ray crystal structure would be needed.

An alternative way of iodinating the thiophene ring was desired to avoid the harsh conditions of the above reaction and to improve the yield. A series of experiments (Figure 3) based on reactions found in the literature were attempted but to no avail: The BuLi reaction was the only means of iodinating the thiophene ring.

Figure 3: Alternative methods of halogenation

The next step in the synthetic plan involved treating this iodinated thienopyrimidine with POCl3, but unfortunately this reaction lead to decomposition of starting material. It was thought that using bromine in place of iodine might lead to better stability. In addition, brominated aryl rings are more frequently used in Suzuki couplings (the next step in the synthetic plan). This offered additional incentive to pursue a brominated thienopyrimidine.

Surprisingly, treating 6-iodothieno[3,2-d]pyrimidin-4-ol with bromine lead to the recovery of starting material. So too did attempts utilising 1,2-dibromoethane.  Treating 4-chlorothieno[3,2-d]pyrimidine with n-BuLi and bromine lead to the synthesis of a brominated thienopyrimidine ring in 50% yield (Scheme 3).

Scheme 3: Brominated 4-chloro thienopyrimdine

Suzuki reactions were attempted next, and these form the crux of my problem right now. Reactions substrates and conditions are shown below in Table 1.

Table 1: Suzuki Reactions

Once purified, only two of these reactions have yielded products that appear interesting (reactions C and F).

It’s interesting that none of the para-isomers of the boronic ester have given products that appear reasonable for this reaction. It’s also quite fortunate, since the compound that I’m interested in (reaction F) involves the meta-isomer.  The typical problematic 1H-NMR spectrum is shown below. (Figure 4). The top spectrum shows peaks that are associated with the desired product (Reaction C), and the bottom spectrum shows peaks that are associated with an undesired product (Reaction D).

The primary question that needs to be solved right now is,

Why aren’t these Suzuki reactions giving the desired products? What can be done to improve the situation? What is shown in that 1H-NMR spectrum?


JimCronshaw's picture

There's a suggestion from the people at ChemForums already that a Buchwald-Hartwig reaction might be interfering.
There's also a mistake in Scheme 2. The pyrimidine ring on the right hand side should also be on the left hand side.

An iodinating reagent you may wish to try is acetyl hypoiodite. It is generated in situ from bis(acetoxy)iodobenzene and iodine and reacts similarly (at least for the systems in which I have tried both) to iodine(I) chloride but is much more friendly to use. I have used it to iodinate intractable aromatic rings in the past and found that it reacted cleanly and quickly.
Most literature examples are for the addition of iodine and some nucleophile across an alkene (e.g. For this purpose a 2:1 ratio of bis(acetoxy)iodobenzene to iodine is generally used but I find a 1:1 ratio is best for aromatic systems. The acetonitrile doesn't need to be particularly dry but the reagent is very light sensitive.
I hope it helps!

JimCronshaw's picture

Thanks Carl. I'll give it a try. I switched the brominating the substrates since I found that far easier than iodinating them, but with this method, perhaps iodination will become viable again.