Synthesis of 18 O-Labelled Alcohols from Unlabelled Alcohols

The synthesis of primary, secondary and tertiary 18 O-enriched alcohols from readily available 16 O-alcohols via a Mitsunobu esterification and hydrolysis is described. The method is further exemplified in the labelling of the active pharmaceutical ingredient, Dropropizine and is shown to be tolerant of modern, separation friendly Mitsunobu reagents. isotope-labelled compounds organic they are used to probe reaction mechanisms through the determination of kinetic isotope effects and isotope experiments. 1 In pharmaceutical standards for mass-spectrometry based analysis of pharmacokinetics and pharmacodynamics. the metabolic stability of active pharmaceutical We began by investigating the synthesis of 4-nitrobenzoic acid-[ 18 O] 2 ( 2 ) by an acid-mediated hydrolysis of 4-nitrobenzonitrile ( 4 ) using 18 O-enriched water. 4-Nitrobenzoic acid was chosen as a suitable carboxylic acid as it is a very commonly used in the Mitsunobu reaction. 10 A short optimisation sequence (see supplementary information, Table S1 and S2) revealed that the hydrolysis of 4-nitrobenzonitrile in 4 M HCl in dioxane afforded 89% of the corresponding doubly 18 O-labelled carboxylic acid with an isotopic purity of 93%. The only other isotopologue detected was the singly labelled acid, which accounted for the remaining 7%. In this protocol, 4-nitrobenzoic acid-[ 18 O] 2 was isolated by filtration and used directly in the Mitsunobu esterification step without further purification.

We began by investigating the synthesis of 4-nitrobenzoic acid-[ 18 O]2 (2) by an acid-mediated hydrolysis of 4-nitrobenzonitrile (4) using 18 O-enriched water. 4-Nitrobenzoic acid was chosen as a suitable carboxylic acid as it is a very commonly used in the Mitsunobu reaction. 10 A short optimisation sequence (see supplementary information, Table S1 and S2) revealed that the hydrolysis of 4-nitrobenzonitrile in 4 M HCl in dioxane afforded 89% of the corresponding doubly 18 O-labelled carboxylic acid with an isotopic purity of 93%. The only other isotopologue detected was the singly labelled acid, which accounted for the remaining 7%. In this protocol, 4-nitrobenzoic acid-[ 18 O]2 was isolated by filtration and used directly in the Mitsunobu esterification step without further purification.
With a convenient method to access 4-nitrobenzoic acid-[ 18 O]2 (2) in hand, we explored the scope of the labelling protocol with a range of alcohol substrates. Firstly, benzylic alcohols substrates were examined and both 1-naphthalenemethanol (1a) and sulfone-containing substrate 1b were converted into the corresponding isotopologues with excellent levels of 18 O enrichment. Additionally, the 18 O isotopologues for indole-containing alcohol 1c and amino acid derivative 1d were accessed in moderate-to excellent yields and with high levels of 18 O enrichment. Ferrocenemethanol (1e) and sugar derivative 1f were also efficiently labelled-the slightly lower level of 18 O enrichment for 1e being possibly attributable to post hydrolysis SN1-type hydroxyl exchange. The enrichment of monoterpenoids, namely geraniol (1g) and (1R)-(−)-myrtenol (1h) was also investigated, and the 18 O-enriched products (3g and 3h) were isolated with good yields and excellent levels of 18 O incorporation. 18 O-enriched (±)-neomenthol (3i) was obtained when (±)-menthol (1i) was subjected to the reaction conditions, demonstrating the expected inversion of stereochemistry associated with the Mitsunobu esterification. Similarly, epi-cholesterol-[ 18 O] (3k) was obtained from natural cholesterol (1k). Finally, the 18 O-isotopologue of acetal-protected adenosine 1l was produced with excellent 18 O enrichment albeit with a poor yield. Although the yields of different substrates were variable the 18 O enrichment for each example remained excellent throughout, ranging from 87-96%.
We next sought to extend this method to encompass active pharmaceutical ingredients. In the context of deuterium labelling, recent developments in hydrogen/deuterium exchange reactions have enabled straightforward access to deuterium-labelled pharmaceuticals. 11 This approach, whereby certain hydrogen atoms of a pharmaceutical are exchanged for deuterium, offers a cost-and time-efficient alternative to de novo synthesis. We therefore wished to apply this principle to 18 O-enriched active pharmaceutical ingredients. Generally, for stable-isotope analogues of active pharmaceutical ingredients to serve as internal standards, the most abundant isotopologue is required to be at least four mass units heavier than the unlabelled isotopologue ([M0 + 4]). In the context of 18 O labels, the most abundant isotopologue of an internal standard must contain a minimum of two 18 O labels. Furthermore, the internal standard should contain no greater than 0.5% of the parent, unlabelled isotopologue ([M0]) in order to avoid cross-signal interferences. 12 The cough suppressant, Dropropizine, was chosen as an appropriate example, since it contains two hydroxyl groups that are reactive towards a Mitsunobu esterification reaction. Using our standard protocol, we were able to obtain Dropropizine-[ 18 O]2 in an overall yield of 56%. Mass spectrometry analysis of this sample indicated that the doubly-labelled isotopologue ([M0 + 4]) accounted for 90.3% of the sample, the singly-labelled isotopologue ([M0 + 2]) accounted for 9.4%, and only 0.3% of the unlabelled substrate ([M0]) was present, thus demonstrating that our method can be effectively used as a means to access these mass spectrometry internal standards.
The classical Mitsunobu conditions (DIAD and triphenylphosphine), used in this study suffer from well-documented drawbacks, specifically the production of triphenylphosphine oxide and hydrazine by-products. Furthermore, DIAD is thermally unstable and its use should be avoided on scale. 13 Therefore we explored the use of second-generation Mitsunobu reagents 14 in our 18 O-labelling protocol.
Firstly, the use of di-tert-butyl azodicarboxylate (DTBAD) in conjunction with diphenyl-2-pyridylphosphine, 15 followed by hydrolysis, enabled the 18 O-enrichment of 1-naphthalenemethanol with comparable efficiency to the DIAD/triphenylphosphine system (Fig. 3A, Entry  2). Under these conditions, addition of hydrochloric acid coverts the hydrazine by-product into gaseous products and allows the phosphine oxide to be removed by phase-separation. Similarly, the use of di-(4-chlorobenzyl)azodicarboxylate (DCAD) developed by Lipshutz 16 with polystyrene-supported triphenylphosphine (PS-PPh3), followed by hydrolysis, afforded 1-naphthalenemethanol-[ 18 O] with comparable 18 O enrichment (92%) but with a lower yield of 49% (Fig. 3A, Entry 3). Both the hydrazine and phosphine oxide by-products of this reaction are insoluble in dichloromethane and were separated from the reaction mixture by filtration.
Finally, we turned our attention to the synthesis of 18 O-enriched tertiary alcohols, which are hard to prepare by direct Mitsunobu reactions. 19 Oxidation of 18 O-enriched alcohol 3j to the corresponding ketone using Dess-Martin periodinane, 20 followed by a Grignard addition using 2-methoxyphenylmagnesium bromide afforded the tertiary alcohol 5 in a yield of 67% over two steps with minimal erosion of the 18 O enrichment (Fig. 3B).
In conclusion, an operationally simple synthesis of 18 O-enriched primary, secondary and tertiary alcohols has been developed. The Mitsunobu coupling approach allows abundant unlabelled alcohols to be used as substrates whilst also providing stereocontrol. The conditions were shown to be tolerant of a wide range of functional groups and were also applied to the synthesis of an isotopically labelled active pharmaceutical ingredient. Furthermore, we have demonstrated that alternative Mitsunobu reagents, which simplify purification, are compatible with this protocol. We envision that this method will be useful as a general and practical means of accessing a broad range of 18 O-enriched alcohols.

Conflicts of interest
There are no conflicts to declare.