Controlling Chemoselectivity of Catalytic Hydroboration with Light

Abstract The ability to selectively react one functional group in the presence of another underpins efficient reaction sequences. Despite many designer catalytic systems being reported for hydroboration reactions, which allow introduction of a functional handle for cross‐coupling or act as mild method for reducing polar functionality, these platforms rarely deal with more complex systems where multiple potentially reactive sites exist. Here we demonstrate, for the first time, the ability to use light to distinguish between ketones and carboxylic acids in more complex molecules. By taking advantage of different activation modes, a single catalytic system can be used for hydroboration, with the chemoselectivity dictated only by the presence or absence of visible light.


Materials and Methods
Unless otherwise stated, all reactions were performed utilizing standard Schlenk techniques. All reagents and starting materials were purchased at reagent grade and used as received. Anhydrous solvents were dried using an Innovative Technology PS-MD-5 solvent purification system. Thin layer chromatography (TLC) was performed on Merck Kieselgel 60 F254 aluminum plates with unmodified silica and visualized either under UV light or stained with potassium permanganate, vanillin, or cerium ammonium molybdate (Hanessian's stain). Column chromatography was performed with Merck silica gel 60 (35 -70 mesh).
High-resolution mass spectrometry (HRMS) was performed using a Thermo Scientific LTQ Orbitrap XL spectrometer. Infrared (IR) spectra were recorded on a Perkin Elmer Spektrum 100 FT-IR spectrometer Spectrum 100 spectrometer with an UATR Diamond/KRS-5 crystal with attenuated total reflectance (ATR) and signals reported as wavenumbers in reciprocal centimeters. The ozonolysis reactions were carried out using the Fisher 502 Ozone Generator. Some 1 H NMR spectra have an impurity at 4.25-4.0 ppm. This signal arises from the oxidised form of the ligand: POPh(OEt)2. In some cases, when the amount was not negligible, it is stated how much it contributes to the isolated mass, and in all the cases the amount of the ligand is taken into account before calculating the isolated yields. In a lot of cases, it can be removed by preparative TLC, however due to the small scale of the reactions, it is hard to detect with UV light and so this second purification can lead to lower isolated yields.
All the photochemical reactions were carried out using a blue LED stripe (24 V, 19.2 W, λmax = 467 nm) purchased from ledxon® GmbH. Reaction vials were placed approximately 2 cm from the light source, using a cork ring wrapped in aluminum foil as a spacer to ensure that all the vial are at the same distance from the light source (vide infra).

Synthesis and Characterization Data of CoH[PPh(OEt)2]4
Procedure Anhydrous CoCl2 (650 mg, 5 mmol, 1 eq) was dissolved in dry ethanol (30 ml) inside an oven dried Schlenk flask under an argon atmosphere, forming a blue solution. This solution was then heated in an oil bath at 65°C. PPh(OEt)2 (4.9 ml, 25 mmol, 5 eq) was added and the solution turned green. Then NaBH4 (430 mg, 11 mmol, 2.2 eq) was slowly added dropwise as suspension in dry ethanol (20 ml), during the addition the solution becomes bright yellow and H2 is produced. Once the gas evolution has stopped, the solution was filtered under argon into a clean Schlenk flask and around half of the solvent was removed under vacuum. The remaining solution was cooled to 0°C with an ice bath to induce precipitation of the product. After 20 minutes the precipitate (2.25 g, 2.65 mmol, 53%), a bright yellow solid, was filtered, dried under high vacuum, and weighed.

Selected optimization reactions
All optimization reactions were carried out on 0.1 mmol scale and analyzed by 1 H NMR spectroscopy with CHBr3 as internal standard.

General procedure for the synthesis of 1,5 ketoacids: Friedel-Crafts acylation (A)
Powdered, anhydrous AlCl3 (3.0 equiv, 15.0 mmol) was suspended in the appropriate substituted arene (5.0 ml) and stirred for 30 minutes. After this time glutaric anhydride (1.0 equiv, 5.0 mmol) was added and the resulting mixture was stirred overnight at 80°C. The reaction was then cooled to 0°C and quenched by slow addition of 1 M HCl(aq) (caution: exothermic reaction with production of fumes). The crude was diluted with EtOAc and the two layers were separated. The organic layer was basified (pH ~ 14) by addition of 10% w/w KOH(aq), the resulting aqueous layer was separated and acidified (pH ~ 2) with 12 M HCl(aq). At this point, the solution turned into a suspension, and it was extracted three times with EtOAc. The combined organic layers were washed with brine, dried (MgSO4), filtered and concentrated under vacuum. The crude was purified through column chromatography (pentane:EtOAc 70:30 + 1% of AcOH).

5-Oxo-5-(p-tolyl)pentanoic acid (1b)
Prepared following the general procedure A starting from toluene, 1b was obtained after purification by column chromatography and recrystallized from EtOAc as a white solid (456 mg, 2.21 mmol, 44%). These data are in agreement with those reported previously in the literature. 1

5-(4-Fluorophenyl)-5-oxopentanoic acid (1c)
Prepared following the general procedure A starting from fluorobenzene, 1c was obtained after purification by column chromatography and recrystallized from EtOAc as a pale-yellow solid (256 mg, 1.22 mmol, 24%). These data are in agreement with those reported previously in the literature. 1

5-(4-Methoxyphenyl)-5-oxopentanoic acid (1f)
Prepared following the general procedure A starting from anisole, 1f was obtained after purification by column chromatography as a pale brown solid (455 mg, 2.05 mmol, 41%). 1  These data are in agreement with those reported previously in the literature. 2

General procedure for the synthesis of ketoacids (B)
Step 1: The appropriate cyclic ketone (1.0 equiv) was dissolved in anhydrous THF (1.0 M) under an argon atmosphere and the resulting solution was cooled to -78 °C (dry ice/acetone). PhMgBr (1.2 equiv, 1 M in THF) was then added. The mixture was allowed to reach room temperature and left stirring overnight. The reaction was quenched by addition of saturated NH4Cl(aq) and diluted with Et2O. The layers were separated, and the aqueous layer was washed one more time with Et2O. The combined organic layers were washed with brine, dried (MgSO4), filtered and concentrated under vacuum. The resulting crude was directly dissolved in AcOH (25 ml) and refluxed (115°C) in an oil bath for 2 hours. The reaction was quenched by slow addition of saturated NaHCO3(aq). Once the CO2 evolution has stopped, Et2O was added, and the layers were separated. The organic layer was subsequently washed with water and brine, dried (MgSO4), filtered and concentrated under vacuum.
Step 2: The alkene prepared in the previous step was dissolved in DCM in a Schlenk tube capped with a gas bubbler. The solution was flushed with argon for 5 minutes and then cooled to -78°C. Ozone was then bubbled in the solution until a pale blue color appeared. The ozone stream was then interrupted, and the solution was flushed with argon for 30 minutes and allowed to reach room temperature. Et3N (2.0 equiv) was added, and the solution turned yellow. After two hours of stirring the crude was washed twice with 1 M HCl(aq) Step 3: The obtained crude aldehyde was dissolved in water (20 equiv) and tert-butyl alcohol (52 equiv). To this was added 2-methyl-2-butene (10 equiv), NaClO2 (3.0 equiv) and NaH2PO4 (5.0 equiv). The mixture was left to stir during 2h at room temperature. Then, EtOAc was added to the mixture, and extraction was done 3 times. The combined organic layers were washed with brine, dried (MgSO4), filtered and concentrated under vacuum.

N-methoxy-N-methylundec-10-enamide (1j-I)
Step 1: To a solution of 10-undecenoic acid (1.0 equiv, 10 mmol, 1.84 g) in DCM (40 mL) was added EDCI (1.5 equiv, 15 mmol, 2.87 g), DMAP (1.5 equiv, 15 mmol, 1.83 g) and N,Odimethylhydroxyamine hydrochloride (1.5 eq, 15 mmol, 1.46 g). The resulting mixture was stirred at room temperature for 16 h before being quenched with brine. The phases were separated, and the aqueous phase was extracted with DCM. The combined organic phases were washed with 2M HCl(aq) and brine before being dried (MgSO4) and concentrated under vacuum giving a pale-yellow oil. The desired amide was obtained clean. The isolated yield is 73% (1. These data are in agreement with those previously reported in the literature. 7 Step 2: The Weinreb amide (1.0 equiv, 1.63 g, 7.3 mmol) was dissolved in THF (25 mL) under argon atmosphere. nBuLi (1.3 equiv, 9.5 mmol) was added to this solution at -78°C (acetone/dry ice). The mixture was left to stir at the same temperature for 45 minutes. The reaction mixture was warmed to room temperature and was then quenched with saturated NH4Cl(aq). The crude was extracted with Et2O, the organic layers were dried (MgSO4), filtered and the solvent was removed under vacuum. The desired product was used without further purification.
Step 3: The obtained product was dissolved in DCM and treated with ozone at -78°C (acetone/dry ice) until the appearance of a pale blue color. When the reaction was finished, the mixture was flushed with argon for 30 minutes, and then Et3N (2 equiv, 14.6 mmol) was added to the mixture at RT. The solution became yellow. It was left to stir for 1 hour. The mixture was then diluted with DCM and washed twice with water, the organic layer was dried (MgSO4), filtered and the solvent was removed under vacuum. The obtained aldehyde was used in the next step without further purification.
Step 4: The aldehyde was dissolved in water (20 equiv) and tert-butyl alcohol (52 equiv). To this was added 2-methyl-2-butene (10 equiv), NaClO2 (3.0 equiv) and NaH2PO4 (5.0 equiv). The mixture was left to stir for 2 hours at room temperature. The mixture was then diluted with EtOAc, and it was washed three times with water. The combined organic layers were washed with brine, dried (MgSO4), filtered and concentrated under vacuum. The desired product was purified by column chromatography (DCM: AcOH= 99:1) and obtained as a white solid in 22% yield (390.9 mg, 1.61 mmol) These data are in agreement with those previously reported in the literature. 8

General procedure 1
In an oven-dried 4 ml vial were introduced the cobalt catalyst (5 mol%, 10 μmol, 8.6 mg), the starting material (1.0 equiv, 0.2 mmol) and a stirring bar. A cap with rubber septum was used to close the vial and the system was then purged with argon. 2-MeTHF (C = 1.0 M, 0.2 mL), HBpin (5.0 equiv, 1.0 mmol, 144 µL) were then added successively and the vial was placed in the dark overnight. After this time, the solution was diluted with Et2O and washed two times with distilled water, the combined aqueous layers were extracted two times with Et2O and the combined organic layers were washed with brine, dried (MgSO4), filtered and concentrated under vacuum. The resulting crude was then purified by column chromatography as detailed for the single compounds.

General procedure 2
In an oven-dried 4 ml vial were introduced the cobalt catalyst (5 mol%, 10 μmol, 8.6 mg), the starting material (1.0 equiv, 0.2 mmol) when solid and a stirring bar. A cap with rubber septum was used to close the vial and the system was then purged with argon. 2-MeTHF (C = 0.1 M, 2.0 mL), HBpin (3.0 equiv, 0.6 mmol, 86 µL) were then added successively and the vial was exposed to blue LEDs overnight. After this time, the solution was diluted with Et2O and washed two times with distilled water, the combined aqueous layers were extracted two times with Et2O and the combined organic layers were washed with brine, dried (MgSO4), filtered and concentrated under vacuum.

Variation for 1,5 ketoacids 1a-1f:
After stirring overnight under irradiation, TFA (5.0 equiv, 1.0 mmol, 77 µL) was added and the mixture was stirred for two hours to induce the lactonization of the products. The solvent was then removed under reduced pressure. The resulting crude was then purified by column chromatography as detailed for the single compounds.

5-Hydroxy-1-phenylpentan-1-one (2a)
Prepared following general procedure 1, 2a was obtained after purification by column chromatography (DCM: acetone 95:5) as a colorless oil ( These data are in agreement with those reported previously in the literature. 9 Hydroxy-1-(p-tolyl) These data are in agreement with those reported previously in the literature. 9

7-Hydroxy-1-phenylheptan-1-one (2h)
Prepared following general procedure 1, 2h was obtained after purification by column chromatography (DCM: acetone 95:5) as a colorless oil (18.9  These data are in agreement with those reported previously in the literature.

n-Pent-4-enyl alcohol (2n)
Prepared following general procedure 1, 2n was obtained crude ( 1 H NMR yield 45% , with respect to CHBr3 as the internal standard). These data are in agreement with those reported previously in the literature. 8

10-Undecen-1-ol (2o)
Prepared following general procedure 1, 2o was obtained crude ( 1 H NMR yield 50%, with respect to CHBr3 as the internal standard). These data are in agreement with those reported previously in the literature. 13

1 H NMR monitoring in dark conditions
For experiments under the dark conditions, we have carried out a preliminary analysis using 1 H NMR monitoring of the reaction. During the initial period of the reaction, we observe the acidic proton (Ha) disappearing and the protons at the α-position to the carboxylic acid becoming deshielded (Hb) -which we believe is consistent with what we outlined in Scheme 1c of the manuscript (reaction of the more nucleophilic carboxylic acid species with the borane). The peak corresponding to the hydride of CoH[PPh(OEt)2]4 (He) also decreases during this time. As product formation starts (Hd), there is an increase also in the quantity of He (at around 350-400 minutes) which matches with our suggestion that CoH[PPh(OEt)2]4 is generated during the catalytic cycle of product formation.

NMR monitoring in light conditions
When our benchmark substrate (spectrum 1) was treated with a sub-stoichiometric amount of cobalt hydride and no HBpin, by 1 H NMR we can clearly observe disappearance of the peaks corresponding to the starting material and appearance of peaks that correspond to the free PPh(OEt)2 ligand. The spectrum 2 represent the reaction mixture before irradiation; the spectra 3 and 4 after 30 minutes and 16 hours of irradiation, respectively. This suggests possible formation of (paramagnetic) cobalt carboxylate species which is likely an unproductive pathway that leads to catalyst decomposition. This implies that the reaction does not necessarily proceed as initially proposed in Scheme 1 of the paper with the 'free' acid.
On the other hand, when we carried out the reaction under the optimised conditions, we observe a peak (-12.36 ppm) that is consistent with the reports of {CoH2[PPh(OEt)2]4} + in the 1 H NMR. 21 We also carried out the standard reaction under light irradiation, measuring the 1 H NMR at partial conversion. Here, along with formation of the product, we also observe that the peak for the protons at the α-position of the carboxylic moiety in the remaining starting material is significantly shifted (peak C, 2.3 to 2.5 ppm), as highlighted in the comparison below. This provides further strong evidence that the substrate initially reacts with one equivalent of HBpin, forming a mixed anhydride.

Deuterium incorporation experiments
• The cobalt deuteride CoD[PPh(OEt)2]4 was synthesised following the procedure described in Section 2, using EtOD and NaBD4 and tested in the dark conditions. No deuterium incorporation in the product was observed.
• Carrying out the benchmark reaction in the light with 20 mol% of CoD[PPh(OEt)2]4 led to good conversion but no detectable deuterium incorporation.

Experiments with TEMPO
The addition of TEMPO in catalytic (5 mol%) as well as stochiometric amounts (2 eq.) to the reaction mixture led to a decrease in acid reduction from 68% in optimized conditions to 25% and 7%, respectively. Unfortunately, despite several attempts, we were not able to identify the product from TEMPO trapping of the boraketyl radical.

The role of HBpin
To further support the dual role of HBpin acting both to protect the carboxylic acid functionality as well as the hydride source, we replaced it with isopropanol as an alternative hydride source which should not react with the carboxylic acid. In this case, no reduction occurs and only starting material is recovered which is consistent with HBpin playing a unique role for ketoacid substrates.

Alternative proposed mechanism
We therefore suggest that an alternative mechanism could be operative whereby the cobalt has a dual role.
Here, the cobalt catalyst would both protect the acid as the anhydride via a mechanism that proceeds via {CoH2[PPh(OEt)2]4} + as well as undergoing ligand photodissociation and subsequent insertion into the ketone.
This suggestion is based on: 1) The work of Onishi describing the reactivity of CoH[PPh(OEt)2]4 with formic acid; 21 2) Our observation of the corresponding {CoH2[PPh(OEt)2]4} + species (section 8.1.2); 3) The experiments which suggest the 'free' carboxylic acid functionality is not compatible with the ketone hydroboration catalytic cycle (sections 8.1.2 and 8.4); 4) NMR data consistent with formation of a mixed boron anhydride (section 8.1.2); However, this is a preliminary suggestion and, at this stage, we cannot rule out other possibilities.