Skip to main content
NCBI home page
Wiley Open Access Collection logoLink to Wiley Open Access Collection
. 2025 Jul 16;64(35):e202512254. doi: 10.1002/anie.202512254

Intermolecular Sp3C─H Metalation of Non‐Nucleophilic Brønsted Bases Using Simple Lewis Acids

Anna V Schellbach 1, Dominic R Willcox 1,2, Miriana Guarnaccia 1,3, Gary S Nichol 1, Valerio Fasano 3, Michael J Ingleson 1,✉
PMCID: PMC12377440  PMID: 40567016

Abstract

2,6‐Di‐tert‐butyl substituted pyridines ( t Bu2‐py) are widely used non‐nucleophilic Brønsted bases. Their ubiquity is due to their highly hindered basic site and chemically robust nature. Herein we report that simple M2X6 Lewis acids (M═Al or Ga, X═Cl, Br or I) effect intermolecular sp3C─H metalation of t Bu2‐py bases under mild conditions. The sp3C─H metalated products can be converted in situ into ─BPin, ─iodo, ─bromo and ─hydroxy derivatives for further elaboration. Mechanistic investigations indicate that: i) a frustrated Lewis pair effects sp3C─H heterolysis to form the C─M bond and a protonated pyridine; ii) C─H metalation requires singly halide‐bridged super‐electrophilic M2X6 dimers for sufficiently low barriers. Finally, sp3C─H metalation using M2X6 is not limited to t Bu2‐py bases. Thus, it is important to be aware of this facile sp3C─H functionalisation when using a range of non‐nucleophilic Brønsted bases.

Keywords: Aluminium trichloride, C─H functionalisation, Frustrated Lewis Pairs, Lewis acids, Non‐nucleophilic bases

Simple M2X6 Lewis acids (M═Al or Ga, X═Cl, Br or I) effect the sp3C─H metalation of a number of hindered bases, including the ubiquitous non‐nucleophilic 2,6‐ditertbutyl substituted pyridine/pyrimidines. Mechanistic studies indicate a frustrated Lewis pair mechanism requiring singly halide‐bridged super‐electrophilic M2X6 dimers for sufficiently low C─H metalation barriers.

graphic file with name ANIE-64-e202512254-g003.jpg

Introduction

Sterically hindered Brønsted bases, termed non‐nucleophilic bases, are ubiquitous in chemistry. This is due to their ability to sequester protons while not (or only weakly) interacting with most other electrophiles. There are multiple classes of non‐nucleophilic bases, with the most established neutral examples being based on bulky amines (e.g., Hünig's base, Figure 1a),[ 1 ] amidines/guanidines (e.g., DBU),[ 2 , 3 ] bulky phosphine derivatives (e.g., BEMP)[ 2 ] and 2,6‐ t Bu2‐pyridines (e.g., 2,6‐ t Bu2‐4‐Me‐pyridine, DBMPy), with this family termed t Bu2‐py herein.[ 4 , 5 ] Of these bases, the most chemically robust are the t Bu2‐py class. While the utility of t Bu2‐py bases is long established,[ 4 , 5 ] they remain pervasive in the current literature (>450 references using them in the period 2020 to 2024).[ 6 ] The robust nature of t Bu2‐py bases comes from their low energy HOMO, high aromatic stabilization energy, extreme bulk around N,[ 7 ] and a periphery consisting of unactivated sp3C─H bonds. The functionalisation of the latter is a significant challenge,[ 8 , 9 , 10 , 11 , 12 , 13 ] and to our knowledge, there is only one report on the C─H functionalisation of a t Bu unit in a t Bu2‐py base.[ 14 ] This required high energy radicals generated by photolysis that abstract H• from protonated DBMPy to form primary radical A (Figure 1b). A rearranges to a more stable 3° radical before reacting further. The identification of another sp3C─H functionalisation of t Bu2‐py bases would be notable, particularly if the transformation proceeds using simple commodity reagents under mild conditions.

Figure 1.

Figure 1

Open in a new tab

a) Select non‐nucleophilic bases. b) Radical‐based functionalisation of t Bu2‐py. c) Previous work on main group compounds effecting sp3C─H functionalisation. d) This work on sp3C─H functionalisation using just M2X6.

Conceptually, if a strong Lewis acid (LA) interacts with one of the t Bu sp3C─H σ bonds in t Bu2‐py this could lower the pKa sufficiently to enable deprotonation by the proximal pyridine and concomitant C─(LA) bond formation.[ 15 ] This would represent an intermolecular frustrated Lewis pair (FLP) type C─H cleavage process as a Lewis acid and a Brønsted base are both required (i.e., strong pyridyl coordination to the Lewis acid needs to be disfavoured). While FLP functionalisation of sp and sp2C─H bonds is now well‐documented,[ 16 , 17 , 18 , 19 ] to date FLPs that cleave sp3C─H bonds are rare.[ 20 , 21 ] The examples reported to date have the Brønsted base, Lewis acid and sp3C─H unit all preorganized in one molecule (e.g., compound B, Figure 1c),[ 22 ] involve relatively complex strong Lewis acids (e.g., compounds C and D)[ 23 , 24 ] or use toxic heavy metals (e.g., Tl(O2CCF3)3).[ 25 ] It would be notable if FLP‐type functionalisation of unactivated sp3C─H bonds could be achieved using simple main group Lewis acids such as aluminium trichloride. This would represent the first catalyst free homogenous alumination of an unactivated sp3C‐H bond.[ 26 , 27 ]

Herein we report that t Bu2‐py bases undergo sp3C─H metalation using M2X6 Lewis acids (M═Al or Ga, X═Cl, Br or I) under mild conditions. Furthermore, we show that intermolecular sp3C─H functionalisation using M2X6 proceeds with other less hindered Brønsted bases and that the sp3C─M products (e.g., compound 1) can be converted into useful bench stable compounds.

Results and Discussion

During our studies using pyrazabole derivatives in electrophilic borylation,[ 28 , 29 ] it was observed that heating dichloro‐pyrazabole, Al2Cl6, DBMPy and activated arenes in chlorobenzene (PhCl) led to a new lower symmetry DBMPy species. This was identified as an sp3C─H functionalised product by conversion to the sp3C─BPin species, compound 2 (Scheme 1). Repeating this reaction in the absence of an activated arene led to the formation of 2 as the major product (by NMR spectroscopy). Compound 2 displays the expected NMR resonances, including a δ 11B = 33.5 and aliphatic 1H resonances in a 3:9:6:2 ratio. Given the widespread use of t Bu2‐py bases, including in SEAr reactions mediated by Al2Cl6,[ 30 ] it was assessed if a boron electrophile was required for sp3C─H functionalisation. In the absence of dichloro‐pyrazabole sp3C─H functionalisation still occurred, with an effectively quantitative (>95%, Scheme 1) conversion when using ≥1 equiv. of Al2Cl6. When less than one equiv. of Al2Cl6 was used, unreacted DBMPy remained. Optimal outcomes were observed using a small excess of Al2Cl6 (e.g., 1.25 equiv.). The in situ 1H NMR spectrum of the sp3C‐H functionalised product displayed a N─ H resonance (broad at 10.5 ppm), and aliphatic resonances in a 3:6:9:2 ratio. These observations combined with subsequent calculations (vide infra) led us to assign the C─H metalated product as compound 1 (Scheme 1). This formulation was supported by analysis in CDCl3 (which revealed two inequivalent meta Aryl─ H resonances). Notably, the addition of dichloro‐pyrazabole/Al2Cl6 to 1 led to transmetalation and formation of species assigned as E (Scheme 1). E is not well‐defined as it is a range of pyrazaboles varying in the exocyclic groups, but these convert into 2 on in situ pinacol protection. This suggests the observed sp3C─H borylation proceeds via initial sp3C─H alumination, a hypothesis supported by attempts to borylate DBMPy with pyrazabole activated by [Ph3C][B(C6F5)4] leading to no borylated product post pinacol protection.

Scheme 1.

Scheme 1

Open in a new tab

The formation of compounds 1 and 2. Yields by NMR spectroscopy versus an internal standard.

With the functionalisation of DBMPy using Al2Cl6 established the conditions when this occurs were explored. This revealed that formation of 1 occurs even at ambient temperature (albeit slowly), and at 80 °C in PhCl > 95% conversion to 1 takes only 2 h despite the incomplete dissolution of Al2Cl6 under these conditions. Given the sensitivity of 1 and 2 to protodemetalation during purification, the in situ conversion of 1 into more stable derivatives and potentially useful (e.g., for subsequent derivatisation) was explored. This was achieved by the reaction of crude 1 with N‐iodo‐succinimide (NIS) at room temperature for 18 h. This enabled isolation of 3 in 58% yield (Scheme 2). Note, under these conditions DBMPy does not react with NIS, even in the presence of Al2Cl6,[ 31 ] indicating initial C─H alumination is essential for accessing 3. NBS also reacted analogously affording 4, albeit in a lower yield than when using NIS. The mass balance in these reactions was free DBMPy which is reformed to some extent on addition of NIS/NBS. Attempts to use 1 to alkylate benzophenone were unsuccessful, with this also resulting in the formation of DBMPy, in this case the Ph2C═O→AlCl3 adduct was identified as the by‐product.[ 32 ] Thus, the addition of carbonyl compounds to 1 can lead to protodemetalation of the sp3C─Al unit by the N─H unit leading back to DBMPy. Another sp3C‐functionalised product is accessible post transmetalation by oxidation of E to form alcohol 5 in 67% isolated yield (Scheme 2).

Scheme 2.

Scheme 2

Open in a new tab

One‐pot alumination/halogenation, and one‐pot alumination / transmetalation / oxidation of DBMPy. Yields versus an internal standard with isolated yields in parentheses.

Next, we explored what other Lewis acids effected sp3C─H metalation of DBMPy. The heavier halide congeners, AlBr3 and AlI3 both perform C─H metalation to form 6 and 7, respectively (Figure 2, top). These reactions proceed more rapidly compared to those using Al2Cl6, presumably in part due to the improved solubility of these Lewis acids in PhCl. Note, ≥ 1 equiv. of Al2X6 is required in each case for full consumption of DBMPy. In contrast to Al2X6, Al2Me6, Al2Et6, Al2Me4Cl2, and ethyl aluminium sesquichloride each resulted in no C─H metalation even after heating in PhCl (by in situ NMR spectroscopy). Metalation is not limited to Al2X6, as Ga2Cl6 effects C─H metalation to produce 8 (Figure 2 top). However, other Group 13 Lewis acids, specifically BCl3, BBr3, BI3, InCl3, and InBr3 all led to no C─H functionalisation (by in situ NMR spectroscopy). A number of other metal halides also resulted in no observable sp3C─H metalation (TiCl4, ZrCl4, SiCl4, PCl5, and ZnCl2). While limited to Al/Ga Lewis acids this is still notable as to our knowledge this is the first homogenous alumination/galliation of unactivated sp3C─H bonds.

Figure 2.

Figure 2

Open in a new tab

Top, the formation of compounds 6–8. Top right, solid‐state structure of 8‐GaCl3 , ellipsoids at 50% probability. Bottom, the extended structure of 8‐GaCl3 showing select intermolecular interactions as black dashed lines.

On standing, crystals suitable for single crystal X‐ray diffraction analysis slowly (over weeks) formed from a reaction of Ga2Cl6 and DBMPy. This confirmed formation of an sp3C─H metalated product, specifically 8‐GaCl3 .[ 33 ] This zwitterion contains one four coordinate anionic gallium unit and a protonated pyridine (inset Figure 2). The structure of 8‐GaCl3 contains a Ga‐C bond length (1.980(2) Å) comparable to that in other [RGaCl3]− anions.[ 34 , 35 ] However, there is evidence for steric induced distortions in 8‐GaCl3 . This includes a large C─C─Ga angle (120.22(15)°) and the two eclipsed (with the proximal methyls) chlorides being angled away from the quaternary carbon (C─Ga─Cleclipsed = 117.28(7) and 118.31(7), whereas C─Ga─Clnon‐eclipsed = 106.14(7)). The extended structure of 8‐GaCl3 contains multiple Ga─Cl⋯H─C interactions (see Figure 2a) with the shortest being to a meta‐arylC─H (at 2.7591(6) Å). In contrast, the shortest intermolecular Cl⋯H─N contact (e.g., Figure 2b) is significantly longer (>3.2 Å), consistent with considerable steric shielding of the N─H unit by the two flanking t Bu substituents.

While confirming C─H functionalisation the structure of 8‐GaCl3 is not consistent with the stoichiometry required for complete consumption of DBMPy (≥ 1 equiv. of Ga2Cl6 is required, Figure S100). This suggested that 8‐GaCl3 is not the major product in solution, but instead forms on standing (over weeks) due to it precipitating as a crystalline solid. To assess this hypothesis the thermodynamics of the C─H metalation reactions using Al2Cl6 and Ga2Cl6 were calculated at the MN15/Def2TSVPP (SMD = chlorobenzene) level (Scheme 3); note all calculations herein are performed at this level. These calculations indicated that the bimetallic C─M2X6 products (e.g., 1/8) are thermodynamically favored over the formation of the mono‐metallic derivatives 1‐AlCl3 and 8‐GaCl3 . While the formation of 8‐GaCl3 is also exergonic from DBMPy and 0.5 equiv. Ga2Cl6, the formation of 1‐AlCl3 is endergonic. The latter is consistent with the outcome from the reaction of 1 and benzophenone, which react to form Ph2C═O→AlCl3 and presumably mono‐metallic 1‐AlCl3 . 1‐AlCl3 then would convert to DBMPy (which is observed by NMR spectroscopy) in line with the relative calculated energies. This highlights the importance of the bimetallic unit in 1 for providing an exergonic metalation process when using Al2Cl6.

Scheme 3.

Scheme 3

Open in a new tab

Calculated free energies (kcal mol−1) of C─H metalation products.

Notably, the formation of the cyclometalated compounds F/G and an equiv. of [(DBMPy)H][MCl4] was found to be favoured thermodynamically over the formation of 1/8. However, F/G are not observed in any reaction mixtures using DBMPy, nor is any significant amount of [(DBMPy)H]+ observed (< 5% of [(DBMPy)H]+ is observed which is attributed to trapping of Brønsted acids generated by reaction of M2X6 with trace protic impurities). Furthermore, combining 1/8 with free DBMPy does not lead to formation of F/G and [(DBMPy)H][MCl4]. This indicates a significant kinetic barrier to the interconversion of 1(8) and F(G), presumably due to the steric bulk surrounding the N─H unit in 1/8 which prevents proton transfer to free DBMPy. These observations preclude the intermediacy of F/G in these metalation reactions and thus disfavours a N‐directed C─H metalation mechanism (involving initial pyN→MxCl3x Lewis adduct formation and metalation then proceeding via the cation [(DBMPy)MCl2]+) as this would be expected to form F/G as metalation products.

Attention thus turned to identifying a feasible mechanism for the C─H metalation of DBMPy. First, C─H metalation to form 1 proceeds analogously when performed in the dark. This, coupled with the absence of any rearrangement products (as observed when radical A is an intermediate), and there being no activation of the Aryl‐ CH 3 unit (which contain weaker C─H bonds relative to the t Bu C‐H bonds), disfavours a radical mechanism. Analysis of the Lewis acids effective for metalation of DBMPy reveals some correlation with Lewis acidity, with only the most Lewis acidic aluminium compounds resulting in C─H functionalisation.[ 36 , 37 ] However, it is significant that BX3 Lewis acids do not effect DBMPy sp3C─H functionalisation despite BX3 Lewis acids having comparable (or greater) calculated Lewis acidity toward “soft” nucleophiles″ (e.g., hydride/methide) relative to monomeric MX3 (M═Ga/Al) species.[ 36 ] The major difference between these boron and aluminium/gallium Lewis acids is that boron trihalides exist as monomers in solution, while the latter exist as Al2X6/Ga2X6 dimers in weakly coordinating solvents (such as PhCl). The disparate outcomes observed using dimeric M2X6 Lewis acids versus BX3 could be kinetic and/or thermodynamic in origin, with the latter consistent with the binding of chloride to form bimetallic [M2Cl7]− anions calculated to be more thermodynamically favored than the binding of chloride to form monometallic anions [MCl4]−;[ 38 ] while the former would be consistent with carbonyl‐olefin metathesis reactions having lower barriers when using M2X6 dimeric Lewis acids compared to that catalyzed with MX3.[ 39 ] This was attributed to the greater Lewis acidity of mono‐halide bridged M2X6 Lewis acids relative to MX3. To provide more insight into the disparate reactivity we explored computationally the sp3C─H metalation of DBMPy using dimeric M2Cl6 (M═Al/Ga) and BCl3.

A mechanism was identified (Figure 3) that proceeds via opening of one μ‐Cl to form σ‐complex H (at ΔG + 19.5 kcal mol−1), via a feasible barrier (TSAl1  = +23.4 kcal mol−1). The reaction then proceeds by a concerted metalation/deprotonation process with a barrier of +24.6 kcal mol−1 (TSAl2 ). The structure of H contains a single μ─Cl and an elongated Al⋯ּAl distance relative to that in Al2Cl6 (3.88 Å in H versus 3.20 Å calculated for Al2Cl6). A significant interaction between a t Bu C─H and the proximal Al centre in H was indicated by the compressed ΣCl─Al─Cl angles of 340.1° (relative to 360° for a trigonal planar Al) and a short Al⋯HC distance of 1.96 Å. The C─H unit interacting with the aluminium centre in H is preorganized for deprotonation, with a short N⋯H distance (N⋯H = 1.989 Å). An analogous mechanism was calculated for Ga2Cl6 with comparable intermediate / transition states (with the highest barrier TSGa2  = +24.6 kcal mol−1). It is notable that while homogenous alumination/galliation of unactivated sp3C─H bonds has not been previously reported to our knowledge, alkane C─H cleavage has been reported using heterogeneous systems, e.g., alumina and Ga‐doped zeolites,[ 40 , 41 ] albeit at much higher temperatures than required to form 1/8. While these heterogeneous systems involve ill‐defined Al/Ga based Lewis acids, multiple calculations support an ambiphilic mechanism involving a strongly Lewis acidic Al/Ga centre and a proximal oxygen‐basic site that combined effect C─H heterolysis.[ 40 , 41 ] Thus, the homogeneous sp3C─H metalations of DBMPy using M2X6 represent well‐defined model systems for heterogeneous Al/Ga mediated alkane activation processes.

Figure 3.

Figure 3

Open in a new tab

Calculated sp3C─H metalation mechanism for BCl3, Al2Cl6 and Ga2Cl6 (free energies in kcal mol−1).

Moving to calculations on sp3C─H functionalisation using BCl3, the overall reaction is less energetically favoured than the Al and Ga analogues, being effectively thermoneutral (Figure 3). Furthermore, the calculated mechanism is different with no σ‐complex analogous to H found. Instead sp3C─H metalation using BCl3 has a single transition state involving concerted cleavage of the C─H bond by BCl3 and the pyridyl lone pair. Notably, this barrier is much higher (TSB  = +41.6 kcal mol−1) than those involving Al2Cl6/Ga2Cl6. Thus, there is a dramatic difference when singly bridged bimetallic M2Cl6 complexes are involved in sp3C─H cleavage. We attribute this to the greater Lewis acidity of the mono bridged Cl3M─(μ─Cl)─MCl2 species. This is consistent with Schlinder's studies on carbonyl‐olefin metathesis,[ 39 ] and with alkane activation using heavy main group complexes where more electrophilic complexes have lower barriers to sp3C─H metalation (specifically relative barriers are Pb(IV) < Tl(III) < Hg(II) for the M(O2CCF3)x series).[ 25 ]

With a plausible mechanism identified, our attention turned to identifying other bases that undergo this C─H metalation. Firstly, another widely used non‐nucleophilic base 2,6‐di‐tert‐butyl‐pyridine was found to undergo alumination with Al2Cl6. Based on the DBMPy studies we assign the product as 9 which is formed in 77% yield by in situ NMR spectroscopy (Figure 4). Note, in situ monitoring is facilitated by the C─H aluminated products containing a diagnostic C H 2─Al resonance (see Supporting Information). For this substrate, sp3C─H functionalisation was confirmed by reaction of 9 with NIS to form the para‐H analogue of 3 (termed 3‐p‐H) in 45% isolated yield. Another common non‐nucleophilic base, 2,4,6‐tri‐ t Bu‐pyrimidine, also reacted with Al2Cl6 by t Bu sp3C─H metalation which proceeded in 33% yield after 18 h at 120 °C. C─H alumination was confirmed again by conversion to the sp3C─iodinated pyrimidine product. This pyrimidine undergoes a selective single C─H alumination (for the t Bu group located between both N atoms), with no double C─H alumination observed. The absence of a second C─H alumination is presumably due to the first C─H cleavage quaternising one nitrogen thus reducing the Brønsted basicity of the second N center. The aluminated product is assigned as compound 10 by analogy to the DBMPy alumination product. Incorporating a para phenyl resulted in selective sp3C─H alumination (to form 11), with no sp2C─H alumination of the phenyl unit observed.

Figure 4.

Figure 4

Open in a new tab

Other t Bu2‐py bases amenable to sp3C─H alumination. Yields by NMR spectroscopy versus an internal standard.

Reducing the ortho steric bulk by using 2,6‐diisopropyl‐pyridine resulted in no C─H metalation using Al2Cl6 under a range of conditions. In contrast, 2‐bromo‐6‐ t Bu‐pyridine (termed 2Br‐Py herein) was amenable to C─H alumination (using Al2Br6 to preclude formation of mixtures from C─X/Al─X halide scrambling).[ 42 ] Under a range of conditions, the maximum C─H alumination yield (versus an internal standard) for 2Br‐Py was ca. 45%. In addition, ca. 55% protonated 2Br‐Py was formed. In this case, the lower bulk around the nitrogen centre in 2Br─Py, relative to that in DBMPy, is expected to result in lower barriers for proton transfer steps for 2Br‐Py derivatives. Thus, formation of the 2Br‐Py analogue of F (Scheme 3) is feasible, which would lead to compound 12 (Scheme 4) and [H(2Br‐Py)]+ in a 1:1 ratio (the theoretical maximum yield of 12 of 50% is not achieved due to the presence of trace protic species which leads to formation of an additional ca. 5% of [H(2Br‐Py)]+). Attempts to crystallize 12 to confirm a cyclo‐metalated structure were not successful. However, the in situ NMR data is consistent with this outcome, furthermore 12 is soluble in pentane (in contrast to 1) and only 0.6 equiv. of Al2Br6 is required for full consumption of 2Br‐Py (as 12 is a mono‐metallic species in contrast to bimetallic 1). While formulation as cyclo‐metalated 12 is tentative, C─H alumination was confirmed by reaction of 12 with NIS to form 13. Compound 13 is bench stable enabling isolation and full characterization.

Scheme 4.

Scheme 4

Open in a new tab

sp3C─H alumination of 2Br─Py using Al2Br6. Yield of isolated compound 13 (relative to 12).

With a range of ortho‐ t Bu‐substituted pyridine bases found to react with Al2X6 other hindered nitrogen bases were tested. When Hünig's base was combined with Al2Cl6 no C─H alumination was observed even on prolonged heating. We surmised that this was due to the shorter hydrocarbyl unit in Hünig's base relative to that in tBu2‐py bases, resulting in a more strained (and thus higher energy) transition state for sp3C─H cleavage. To test this hypothesis an alternative hindered tertiary amine was selected, specifically, MesNMe2 as this amine contains a longer hydrocarbyl chain (N─Csp2─Csp2─Csp3) and undergoes C─H cleavage using bora‐triptycene C.[ 24 ] Notably, combination of MesNMe2 with Al2Cl6 led to ca. 46% sp3C─H alumination (by in situ NMR spectroscopy) with the mass balance being [MesN(H)Me2][AlxCl3x+1]. Consistent with this outcome, only 0.6 equiv. of Al2Cl6 was required to fully consume MesNMe2. By analogy to 12 we assign the C─H alumination product derived from MesNMe2 as the cyclo‐metalated compound 14 (Scheme 5). This demonstrates that sp3C─H alumination using just Al2Cl6 is not limited to ortho‐ t Bu‐pyridine bases.

Scheme 5.

Scheme 5

Open in a new tab

C─H alumination of MesNMe2 using Al2Cl6. Yield by NMR spectroscopy versus an internal standard.

Finally, as pyridine units are privileged moieties in active pharmaceutical ingredients the C─H alumination of 2Br‐Py was used to rapidly access functionalized pyridines (Scheme 6). Specifically, 2Br─Py underwent a sequence of alumination / transmetalation / oxidation in one‐pot to form compound 15 in 29% overall yield (note this is a reasonable yield given the multiple steps involved and the fact that only 50% of 2Br─Py can undergo C─H alumination). Subsequent cross coupling of 15 was then straight forward, as exemplified by formation of 16 under standard Suzuki‐Miyaura coupling conditions. This sequence provides an alternative to Minisci chemistry to access these products, which is notable as Minisci‐type hydroxy alkylation of pyridines often gives mixtures of C2/C4 hydroxy‐alkylated products.[ 43 ] In contrast only formation of the C2 t Bu─OH group occurs via this C─H alumination process.

Scheme 6.

Scheme 6

Open in a new tab

Sequential functionalisation of 2Br─Py to form 15 and 16. Isolated yields shown.

Conclusion

The ubiquitous class of non‐nucleophilic Brønsted bases based on 2,6‐di‐tert butyl pyridines/pyrimidines undergo facile sp3C─H functionalisation. This intermolecular sp3C─H functionalisation occurs with simple aluminium and gallium M2X6 Lewis acids and proceeds even at room temperature. Thus, it is surprising that this transformation has gone unnoticed to date. Mechanistic studies indicate that dimeric Lewis acids are essential for sp3C‐H metalation, an observation attributed to the extreme Lewis acidity of mono‐halide bridged M2X6 units which is key to enable a sufficiently low barrier concerted C‐H bond heterolysis. The sp3C‐aluminated products can be converted into a range of useful derivatives, specifically: sp3C─BPin, ‐halogenated and ‐hydroxylated, thus this transformation represents a novel route to access functionalised pyridines. Furthermore, this transformation is not limited to extremely hindered pyridines, with a 2‐bromo‐pyridyl derivative and a hindered tertiary amine, Mesityl‐NMe2, also undergoing intermolecular sp3C─H alumination. Given the widespread use of these hindered bases combined with the facile nature of this sp3C─H functionalisation, it is important to be aware of this transformation when considering using these non‐nucleophilic bases.

Conflict of Interests

The authors declare no conflict of interest.

Supporting information

Supporting Information

ANIE-64-e202512254-s003.pdf (4.9MB, pdf)

Supporting Information

ANIE-64-e202512254-s001.zip (14.9KB, zip)

Supporting Information

ANIE-64-e202512254-s002.cif (681.1KB, cif)

Acknowledgements

This project has received funding from the EPSRC Programme Grant “Boron: Beyond the Reagent” (EP/W007517/1), EPSRC Grants EP/X021858/1 and EP/X035174/1. The authors thank the Mass Spectrometry facility (SIRCAMS) at the University of Edinburgh for carrying out MS analysis. M.G. thanks the European Union for funding through the Erasmus Traineeship program.

Schellbach A. V., Willcox D. R., Guarnaccia M., Nichol G. S., Fasano V., M. J. Ingleson*, Angew. Chem. Int. Ed. 2025, 64, e202512254. 10.1002/anie.202512254

Data Availability Statement

The data that support the findings of this study are available in the Supporting Information of this article.

References

  • 1. Sorgi K. L., in Diisopropylethylamine in Encyclopedia of Reagents for Organic Synthesis, 2001, 10.1002/047084289X.rd254. [DOI] [Google Scholar]
  • 2. Organic Synthesis: Guanidines, Amidines, Phosphazenes and Related Organocatalysts, (ed.: Ishikawa T.), John Wiley & Sons, Ltd, Chichester, United Kingdom: 2009, https://onlinelibrary.wiley.com/doi/book/10.1002/9780470740859?msockid=10247e1d3ed4600e10296b283fae612c. [Google Scholar]
  • 3. Muzart J., ChemistrySelect 2020, 5, 11608–11620. [Google Scholar]
  • 4. Kostikov R. R., Francisco S.‐S., Garranzo M., Murcia M. C., 2,6‐Di‐t‐butyl‐pyridine in Encyclopedia of Reagents for Organic Synthesis, John Wiley & Sons, Ltd, 2010. [Google Scholar]
  • 5. Balaban A. T., 2,6‐Di‐tert‐butyl‐4‐methylpyridine in Encyclopedia of Reagents for Organic Synthesis, John Wiley & Sons, Ltd, 2004. [Google Scholar]
  • 6.SciFinder Scholar Search, March 2025.
  • 7. Saida A. B., Mahaut D., Tumanov N., Wouters J., Champagne B., Vanthuyne N., Robiette R., Berionni G., Angew. Chem. Int. Ed. 2024, 63, e202407503. [DOI] [PubMed] [Google Scholar]
  • 8. Baudoin O., Chem. Soc. Rev. 2011, 40, 4902. [DOI] [PubMed] [Google Scholar]
  • 9. Rouquet G., Chatani N., Angew. Chem. Int. Ed. 2013, 52, 11726–11743. [DOI] [PubMed] [Google Scholar]
  • 10. Hu J., Lv J., Shi Z., Trends Chem. 2022, 4, 685–698. [Google Scholar]
  • 11. Newton C. G., Wang S.‐G., Oliveira C. C., Cramer N., Chem. Rev. 2017, 117, 8908–8976. [DOI] [PubMed] [Google Scholar]
  • 12. Tu J.‐L., Zhu Y., Li P., Huang B., Org. Chem. Front. 2024, 11, 5278–5305. [Google Scholar]
  • 13. Rej S., Amos S. G. E., Kleij A. W., ACS Catal. 2025, 15, 1753–1770. [Google Scholar]
  • 14. Kang Y. C., Treacy S. M., Rovis T., ACS Catal. 2021, 11, 7442–7449. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.(Ed. Kubas G. J.), Metal Dihydrogen and σ‐Bond Complexes: Structure Theory and Reactivity, Springer US, Boston, MA, 2001, pp. 365–415. [Google Scholar]
  • 16. Fontaine F.‐G., Desrosiers V., Synthesis 2021, 53, 4599–4613. [Google Scholar]
  • 17. Ma Y., Lou S.‐J., Hou Z., Chem. Soc. Rev. 2021, 50, 1945–1967. [DOI] [PubMed] [Google Scholar]
  • 18. Ingleson M. J., ACS Catal. 2023, 13, 7691–7697. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19. Soltani Y., Fontaine F.‐G., Frustrated LewisPairs (Eds.: Slootweg C., Jupp A.R.), Springer International Publishing, Cham, Switzerland: 2021, pp. 113–166. [Google Scholar]
  • 20. Ashwathappa P. K. S., Higashi T., Desrosiers V., Omaña A. A., Fontaine F.‐G., Angew. Chem. Int. Ed. 2023, 62, e202309295. [DOI] [PubMed] [Google Scholar]
  • 21.For more on main group compounds that functionalise unactivated C‐H bonds see (and references therein): Pei R., Chang W., He L., Wang T., Zhao Y., Liang Y., Wang X., Nat. Comm. 2024, 15, 7943–7949. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22. Rochette É., Courtemanche M.‐A., Fontaine F.‐G., Chem. ‐ Eur. J. 2017, 23, 3567–3571. [DOI] [PubMed] [Google Scholar]
  • 23. Liu Y., Dong W., Li Z. H., Wang H., Chem 2021, 7, 1843–1851. [Google Scholar]
  • 24. Osi A., Mahaut D., Tumanov N., Fusaro L., Wouters J., Champagne B., Chardon A., Berionni G., Angew. Chem. Int. Ed. 2022, 61, e202112342. [DOI] [PubMed] [Google Scholar]
  • 25. Hashiguchi B. G., Konnick M. M., Bischof S. M., Gustafson S. J., Devarajan D., Gunsalus N., Ess D. H., Periana R. A., Science 2014, 343, 1232–1237. [DOI] [PubMed] [Google Scholar]
  • 26.For a recent example of a catalysed sp3C‐H alumination see: Kanbur U., Paterson A. L., Rodriguez J., Kocen A. L., Yappert R., Hackler R. A., Wang Y.‐Y., Peters B., Delferro M., LaPointe A. M., Coates G. W., Perras F. A., Sadow A. D., J. Am. Chem. Soc. 2023, 145, 2901–2910. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.For a recent paper on direct sp2C‐H alumination see: Kumar Bisai M., Łosiewicz J., Sotorrios L., Nichol G. S., Dominey A. P., Cowley M. J., Thomas S. P., Macgregor S. A., Ingleson M. J., Angew. Chem. Int. Ed. 2024, 63, e202404848. [DOI] [PubMed] [Google Scholar]
  • 28. Millet C. R. P., Noone E., Schellbach A. V., Pahl J., Łosiewicz J., Nichol G. S., Ingleson M. J., Chem. Sci. 2023, 14, 12041–12048. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29. Pahl J., Noone E., Uzelac M., Yuan K., Ingleson M. J., Angew. Chem. Int. Ed. 2022, 61, e202206230. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.For an example in electrophilic C‐H borylation see: Bagutski V., Del Grosso A., Carrillo J. A., Cade I. A., Helm M. D., Lawson J. R., Singleton P. J., Solomon S. A., Marcelli T., Ingleson M. J., J. Am. Chem. Soc. 2013, 135, 474–487. [DOI] [PubMed] [Google Scholar]
  • 31. Racys D. T., Warrilow C. E., Pimlott S. L., Sutherland A., Org. Lett. 2015, 17, 4782–4785. [DOI] [PubMed] [Google Scholar]
  • 32. Branch C. S., Bott S. G., Barron A. R., J. Organomet. Chem. 2003, 666, 23–34. [Google Scholar]
  • 33.Deposition number CCDC 2455501 contain the supplementary crystallographic data for this paper. These data are provided free of charge by the joint Cambridge Crystallographic Data Centre and Fachinformationszentrum Karlsruhe access structures service.
  • 34. El‐Hellani A., Monot J., Tang S., Guillot R., Bour C., Gandon V., Inorg. Chem. 2013, 52, 11493–11502. [DOI] [PubMed] [Google Scholar]
  • 35. van Poppel L. H., Bott S. G., Barron A. R., Polyhedron 2003, 22, 9–17. [Google Scholar]
  • 36. Erdmann P., Greb L., ChemPhysChem 2021, 22, 935–943. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37. Goonesinghe C., Jung H.‐J., Betinol I. O., Gaffen J. R., Garrard C. N., Chang J., Hosseini K., Roshandel H., Nyamayaro K., Patrick B., Ezhova M., Caputo C. B., Baumgartner T., Reid J. P., Mehrkhodavandi P., ACS Catal. 2023, 13, 16148–16157. [Google Scholar]
  • 38. Kraft A., Beck J., Krossing I., Chem. ‐ Eur. J. 2011, 17, 12975–12980. [DOI] [PubMed] [Google Scholar]
  • 39. Albright H., Riehl P. S., McAtee C. C., Reid J. P., Ludwig J. R., Karp L. A., Zimmerman P. M., Sigman M. S., Schindler C. S., J. Am. Chem. Soc. 2019, 141, 1690–1700. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40. Coperet C., Chem. Rev. 2010, 110, 656–680. [DOI] [PubMed] [Google Scholar]
  • 41. Feng Z., Liu X., Wang Y., Meng C., Molecules 2021, 26, 2234. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42. Olah G. A., Tolgyesi W. S., Dear R. E. A., J. Org. Chem. 1962, 27, 3441–3449. [Google Scholar]
  • 43. Kianmehr E., Fardpour M., Khan K. M., Eur. J. Org. Chem. 2017, 2017, 2661–2668. [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supporting Information

ANIE-64-e202512254-s003.pdf (4.9MB, pdf)

Supporting Information

ANIE-64-e202512254-s001.zip (14.9KB, zip)

Supporting Information

ANIE-64-e202512254-s002.cif (681.1KB, cif)

Data Availability Statement

The data that support the findings of this study are available in the Supporting Information of this article.

Articles from Angewandte Chemie (International Ed. in English) are provided here courtesy of Wiley

RESOURCES