Skip to main content
NCBI home page
ACS AuthorChoice logoLink to ACS AuthorChoice
. 2026 Feb 20;148(8):8970–8981. doi: 10.1021/jacs.5c22122

How Alkali Metal Alkoxides Initiate Organic Radical Reactions

Seb Tyerman †, Kenneth F Clark †, Alexander J Stewart †, Krystian Kolodziejczak †, Craig M Robertson ‡, Laura Evans §, Alan R Kennedy †, Tell Tuttle †, David J Nelson †,*, John A Murphy †,*
PMCID: PMC12964405  PMID: 41718038

Abstract

Alkali metal alkoxides have long been known to cause hydrodehalogenation of aryl halides; this conversion of aryl halides to arenes happens when the reactions are conducted in appropriate solvents (with weak C–H bonds). More recently, when aryl halides are heated with alkoxides in arene solvents, coupling to arenes occurs. Both of these reaction types are known to involve aryl radical intermediates. The consensus has been that alkali metal alkoxides undergo electron transfer to aryl halides to form radicals, but crucial evidence has been missing. We now refute this proposal and show through deuterium isotope studies that the deprotonation of the substrates leads to benzynes that initiate radical chemistry. Surprisingly, o-, m-, p- and, in appropriate cases, r- (remote) benzynes are simultaneously formed. During reactions with potassium tert-butoxide, we observed for the first time low-level methylation of arenes, resulting from methyl radicals derived from tert-butoxide. Although methyl radicals could, in principle, arise by electron transfer from tert-butoxide ions, followed by known radical fragmentation, we show that a different, previously unreported mechanism applies.

graphic file with name ja5c22122_0012.jpg

graphic file with name ja5c22122_0010.jpg

Introduction

Activation of aryl halides to convert the C–Hal bond to a C–H or C–C or C–heteroatom is central to organic chemistry. Conversion of aryl halides 1 to arenes 4 on heating with alkoxides 3 (Scheme A) has been reported since 1899. In 1992, Bunnett reviewed known cases and proposed a radical mechanism for the propagation steps of the reactions but avoided any comment on the source of the radicals (i.e., the initiation of the reactions). More recently, when aryl halides were heated with an alkali metal tert-alkoxide above 100 °C in an arene solvent, coupling to the arene was observed to give a biaryl. This type of coupling reaction is termed a base-promoted homolytic aromatic substitution (BHAS) reaction (Scheme B) and again proceeds through aryl radical intermediates. Thus, the two types of reactions, hydrodehalogenation and BHAS coupling, are linked by the reaction of alkali metal alkoxides with aryl halides and the formation of aryl radical intermediates.

1. (A) Protodehalogenation of Haloarenes 1 to Arene 4; (B) BHAS Coupling of Aryl Halide 1 to Form Biaryl 9; and (C) Additives 10, 12–17 React with KOtBu to Form Strong Organic Electron Donors.

1

Open in a new tab

Further studies on BHAS reactions showed that they are assisted by any of a wide range of additives. The reactions generally proceed in high yield. − Propagation occurs when aryl radicals 2 add to arene 4 to give 7. Deprotonation affords radical anion 8, which donates an electron to another molecule of 1, thereby propagating the chain and forming the product, biphenyl 9. We previously explored the role of the additives and showed that super electron donors 11, − formed in situ from additive 10, can initiate these BHAS reactions. Additives other than 10 facilitate these reactions, e.g., 12–17, and are converted in situ by KOtBu into strong organic electron donors that initiate the chemistry − (Scheme C). Thus, it was clear for BHAS reactions that the formation of strong electron donor molecules facilitated the reaction by converting the aryl halide substrates into aryl radicals through electron transfer.

But we also showed that a slower coupling of aryl halides to arenes can occur in the absence of additives, i.e., simply by the reaction of KOtBu 18 with the aryl halide in the presence of excess arene. Although not discussed in detail, the literature represents the reaction between KOtBu and an aryl halide as involving direct electron transfer from the alkoxide to the aryl halide, leading to a tert-butoxyl radical, an aryl radical, and a halide anion. − ,− However, alkali metal alkoxides are very poor electron donors. For example, for potassium tert-butoxide, E ox = +0.10 V (vs. saturated calomel electrode (SCE) in dimethylformamide (DMF)), while for Ar–I, E red ≈ −2.3 V (vs. SCE in DMF), making outer sphere electron transfer from KOtBu to ArI impossible.

Besides the problem of incompatible redox potentials, another problem with proposing electron transfer from KOtBu as the route to the initiation of radical chemistry is the absence of evidence of methyl radicals in these reactions. If direct electron transfer from KOtBu to aryl halides occurred, the resulting tert-butoxyl radicals 19 would partition between hydrogen atom abstraction to form tBuOH 23 and fragmentation to acetone 20 and methyl radicals 21 (Scheme A). − No evidence has been published of methyl radicals arising in these reactions (methyl radicals, generated by metal-free methods from peroxides, are known to add to heteroarenes). −

2. (A) The Known Fragmentation of tert-Butoxyl Radicals 19; (B) Potential Initiation of BHAS Coupling through Formation of o-Benzyne 24; and (C) 2,6-Dimethyliodobenzene 32 as a BHAS Substrate.

2

Open in a new tab

In 2014, we tentatively proposed an alternative initiation of BHAS reactions in the absence of recognized electron donors, through benzyne intermediates (Scheme B). Benzynes are known to form when aryl halides are treated with a strong base, and addition of benzynes to alkenes has been proposed to go through diradical intermediates. − In principle, a similar addition to arenes might provide a source of radicals to initiate BHAS chemistry. But disruption of the aromaticity of two benzene rings by a ground-state o-benzyne to form products like 31 is unknown and would be very surprising. In our proposal, the addition of o-benzyne 24 to benzene 4 forms diradical 25. This is both a reactive aryl radical and a less reactive cyclohexadienyl radical. The aryl radical could either add to benzene to form 26 or abstract an H atom from benzene to form 7. Deprotonations and electron transfers would convert 7 and 26 to 9 and 31, respectively. The electron transfer steps would activate substrate 1 as in Scheme B to allow the propagation of the reaction. Byproduct 31 emerging from the proposed initiation pathway has never been reported. That tallies with propagation being much faster than initiation, leading to very low concentrations of such byproducts.

To test the proposal, we reported the outcome of experiments with 2,6-dimethyliodobenzene 32, which cannot form an o-benzyne. , This hindered substrate affords lower yields of products, but hindrance is key to its role as an indicator of mechanism. Using a suitable electron donor, formed from KOtBu and phenanthroline ,, (Scheme C), coupling to benzene occurred, affording 34 via radical 33. The alternative outcome for 33 was that it abstracted an H atom from benzene yielding 35, and the resulting phenyl radical was added to benzene to form biphenyl 9 by a BHAS mechanism. Seeing these products reinforced the importance of electron transfer initiation of the reaction.

When the experiment was repeated without an electron donor, i.e., in the presence of KOtBu alone, we expected that the combined yield of coupled products, 34 + 9, would be 0%. Although the combined yield dropped significantly from 42 to <1%, nevertheless, the fact that the coupled products 34 and 9 were still formed to any extent indicated that an additional mechanism for their formation was in play.

Returning to Scheme B, our motivation was to search for terphenyl byproduct 31 and to see if that might give evidence of the involvement of benzyne in the initiation process. If 31 could be identified, (i) it might simply arise through further BHAS reaction of phenyl radicals 2 with biphenyl 9, or alternatively, (ii) it could arise from the reaction of benzyne 24 with two molecules of benzene solvent. The two routes might be distinguished by reacting deuterated iodobenzene, 1-d 5 , with a base, in benzene as a solvent, and observing the products that are formed (Scheme ). In this case, if labeled phenyl radicals 2-d 5 are produced, then they should routinely lead to labeled biphenyl, 9-d 5 , through reaction with benzene solvent; reaction of another radical 2-d 5 with 9-d 5 would yield equal amounts of terphenyls 31-d 9 and 31-d 10 . On the other hand, if the terphenyl arises from labeled benzyne 24-d 4 reacting with two molecules of benzene solvent, this would afford 31-d 4 . As will be seen below, deuterium isotope studies provide strong evidence in support of o-terphenyl formation from o-benzyne 24, but they will also provide evidence of simultaneous formation of m- and p-benzynes from iodobenzene as a substrate. Further important information about the regiochemistry of benzyne formation will then be explored using isotopologues of 9-bromoanthracene and 9,10-dibromoanthracene. Our strategy also involves studying the outcomes of such reactions with the deuterated solvent C6D6 and deuterated base KOtBu-d 9 to provide further important mechanistic information.

3. Using 1a- d 5 to Distinguish between Routes to o-Terphenyl 31 .

3

Open in a new tab

Results and Discussion

Evidence of Benzynes

As mentioned in the Introduction, our study began with a search for terphenyl product 31 in the reaction of iodobenzene 1a (X = I), with KOtBu in benzene as a solvent. This reaction afforded tert-butoxybenzene, 36 (16.1%), from attack by butoxide on benzyne and traces of triphenylene 37 from the trimerization of benzyne (Scheme a). In addition, biphenyl 9 (27.2%) from the propagation steps in Scheme B starting material, and iodobenzene 1a, X = I (36.7%), were quantified by gas chromatography-flame ionization detection (GCFID) or NMR. Three terphenyls, 31 (0.18%), 38 (0.08%), and 39 (0.05%), were also identified by GC mass spectrometry (GCMS) (see SI, pp S6–S10), quantified by GCFID, and correlated with authentic samples of o-, m-, and p-terphenyl (Scheme A).

4. (A) Products from the Reaction of PhI with KOtBu in PhH; (B) From Reaction of C6D5I with KOtBu in PhH, Terphenyls Might Arise from Attack of Phenyl Radicals 2-d 5 on Biphenyl Product 9- d 5 ; and (C) The Observed o-, m-, and p-Terphenyl Products and How They Arise from o-, m-, and p-Benzynes.

4

Open in a new tab

a Only the o-terphenyl products expected 31-d 9 and 31-d 10 are shown here.

These terphenyls might simply have arisen through the BHAS reaction of the standard propagation product of the reaction, biphenyl 9, with further phenyl radicals in solution. However, this was disproved when the experiment was repeated with iodobenzene-d 5, where the deuteration patterns of the products revealed important information.

In this reaction, phenyl radicals 2-d 5 might react with the biphenyl formed in this reaction, i.e., 9-d 5 ; o-, m-, and p-terphenyls could form in this way. Scheme B focuses only on the o-terphenyl products and shows the two isotopologues that would arise. The radical attack of 2-d 5 can occur on either ring of the biphenyl, ultimately giving a ∼1:1 mixture of o-terphenyls, 31-d 9 and 31-d 10 . Likewise, d 9 and d 10 isotopologues would be seen for the m- and p-terphenyls. However, the observed labeling patterns were quite different; d 9 isotopologues were seen, but d 10 isotopologues were not seen. The other isotopologues seen were d 4 isotopologues. While both these d 4 and d 9 isotopologues were easily seen for the ortho- and para-isomers, 31 and 39, the meta-case formed almost exclusively 38-d 9 (see SI, pp S14 and S15). Scheme C shows how they arise from benzyne intermediates.

First, we focus on o-benzyne, o- 24-d 4 . This led to the terphenyls 31- d 4 and 31- d 9 . As shown in the inset, 31- d 4 would arise from sequential BHAS reactions with two benzene molecules. The route to 31- d 9 , on the other hand, would involve an ipso-substitution , reaction on C6D5I, together with a BHAS reaction with PhH. Similar outcomes for the m- and p-terphenyls would fit with the proposal. Our experiment therefore surprisingly establishes that the terphenyl products formed in this reaction arise because o-, m-, and p-benzynes are being formed in solution. (m- and p-Benzynes may alternatively be referred to as “arenediyls” or as “didehydroarenes”, but we will use the benzyne terminology in this paper.)

These reactions also show the following:

  • (i)

    m- and p-benzynes are formed through base-treatment of a simple halobenzene (until now, only o-benzynes were known to form in this way). Perrin et al. have recently pioneered the nonradical chemistry of p-benzynes and widely investigated the reverse reaction, i.e., addition of halide anion to p-benzynes. ,

  • (ii)

    Addition of a ground state o-benzyne to an arene, benzene, does indeed lead to products derived from diradical intermediates. , Ground state o-benzynes do not routinely show diradical behavior with closed shell molecules, but recent results showed that they can be attacked by persistent radicals like TEMPO. ,

Haloanthracene Substrates

Having observed that the deprotonation of iodobenzene could occur at ortho, meta, and para-sites to lead to o-, m-, and p-benzynes, we were keen to explore the regiochemistry of deprotonation of other substrates, namely, 9-haloanthracenes 40, 40 Cl , and 40 I that cannot form o-benzynes (Scheme A). 9-Iodoanthracene 40 I and 9-bromoanthracene 40 reacted most efficiently, while 9-chloroanthracene 40 Cl was less reactive (see SI, pp S59–S68). For these substrates, while o- and m-benzynes cannot form, a p-benzyne could be possible. Focusing on the reaction with 9-bromoanthracene, this afforded anthracenes 41–44, as well as biphenyl 9, as detailed below. Notably, methylated compounds 45–48 were also identified (Scheme A and see SI, p. S25).

5. (A) Reaction of 9-Bromoanthracene, 40; (B) Formation of the Non-methylated Products from 9-Bromoanthracene 40; and (C) Expected Reactions of Radical 49 with C6D6 .

5

Open in a new tab

a In the actual experiment, the product anthracene 41 was unlabeled.

The methylated products will be discussed below, while Scheme B illustrates the formation of the other products. 9-Anthracenyl radical 49 could (a) abstract an H atom from benzene to afford anthracene 41 and a phenyl radical 2 or (b) add to benzene 4 (shown) or to anthracene 41 (not shown) by the BHAS mechanism or to bromoanthracene 40 by ipso-substitution of the bromine atom (not shown) to afford 9-phenylanthracene 42 and 9,9′-bianthracene 44 (inset in Scheme B), respectively. In turn, further BHAS reaction of phenylanthracene 42 with phenyl radical 2 would form diphenylanthracene 43, while 2 and 41 undergo further reactions to form products 9 and 42.

Isotope Effects with 9-Bromoanthracene 40

The above reaction of 40 in C6H6 was repeated under identical conditions in a side-by-side reaction in C6D6 as a solvent. First, we describe what was expected when the anthracenyl radical 49 reacts with C6D6 in relation to isotope effects. Two possibilities exist:

  • (i)

    The radical adds to C6D6 to form arylcyclohexadienyl radical 50- d 6 and this undergoes loss of D+ to afford 51- d 5 en route to products (Scheme C). Deprotonations of cyclohexadienyl radicals (like 50) to form cyclohexadienyl radical anions in BHAS chemistry are known not to be rate-determining steps. ,,

  • (ii)

    The radical abstracts a D atom from C6D6. This would form 41- d together with a C6D5 radical, 2- d 5 . This breaking of an Ar–D bond is a kinetically challenging step in BHAS chemistry and would give rise to a primary deuterium isotope effect.

The results are shown in Table , entries 1 and 2. Similar products were formed in both reactions, but the labeled products that formed in C6D6 were 9- d 10 , 42- d 5 , and 43- d 10 . Unlabeled anthracene 41 was formed in that reaction rather than 41- d . This implies that 49 abstracts an H atom from a source, R–H, other than deuterobenzene 4- d 6 , and that the resulting radical is responsible for the subsequent formation of the phenyl radical, 2- d 5 . The nature of R–H is discussed later in this paper.

1. Reactions of 40 Using Isotopologues of the Solvent and KOtBu.

graphic file with name ja5c22122_0007.jpg

entry conditions 40 9 or 9-d 10 (%) 41 (%) 42 or 42-d 5 (%) 43 or 43-d 10 (%) 44 (%) 45 or 45-d n (%)
1 KOtBu, C6H6 0 28.2 39.2 20.4 1.4 5.8 3.2
2 KOtBu, C6D6 0 5.2 20.4 20.1 0.4 6.8 6.2
3 KOtBu-d 9, C6H6 0 26.8 29.0 18.9 1.5 4.7 0.7
4 KOtBu-d 9, C6D6 0 5.0 8.3 18.1 0.4 3.8 2.5
Open in a new tab
a

Reactions were conducted in triplicate (entry 1) and in duplicate (entry 2).

b

Indicates deuterated product.

The amount of biphenyl 9- d 10 produced from the C6D6 reaction was greatly decreased compared to that of the corresponding compound in C6H6 (28.2 → 5.2%), reflecting that the formation of biphenyl 9 was occurring through the expected BHAS process. The amount of phenylanthracene 42- d 5 produced remained constant relative to its isotopologue in the C6H6 reaction (20.1 → 20.4%), which would be expected, as no kinetic isotope effect (KIE) should be associated with the deprotonation of radical 50. The same applies to the formation of bianthracene 44 (entries 3 and 4 of Table will be discussed later).

Further information was gathered by preparing a range of selectively deuterated isotopologues of 40, namely, 40- d 1 , 40- d 8 , and 40- d 9 , and comparing their reactivity with that of undeuterated 40. In these experiments, the duration of the reaction is shorter (30 min) to ensure that none of the substrates is fully converted into product and hence that the differences in their reactivity are evident. The substrates were tested under identical conditions in side-by-side reactions ( d 0 , d 1 , d 8 , d 9 ) (Table ).

2. Reactions of Isotopologues of 40 with C6H6 and KOtBu.

graphic file with name ja5c22122_0008.jpg

    % yield ,
entry substrate 40 9 41 42 43 44 45 52
1 40- d 0 45.1 12.4 17.6 7.1 0.5 2.0 1.7 0.8
2 40- d 1 65.2 5.7 8.7 3.3 0.2 1.1 0.9 0.5
3 40- d 8 75.7 2.2 3.5 1.4 0.1 0.5 0.4 0.2
4 40- d 9 81.0 1.8 2.8 1.1 0.1 0.3 0.3 0.2
Open in a new tab
a

Isotopologue yields determined using GCFID calibrations for nondeuterated compounds.

b

Reactions were performed in triplicate; averaged values are presented in this table.

The results show that an increasing level of deuteration in 40 slows the reaction. The amount of the particular isotopologue of 9 (or of anthracene 41) formed in each experiment represents the radical flux. Comparing experiments in Table (entries 1 and 2) shows that the yield of 9 from 40- d 1 (5.7%) is significantly lower than that from 40- d 0 (12.4%). This shows that the amount of phenyl radical being produced in these reactions is dependent on the C–H/C–D bond in the 10-position of the anthracene. This is not at all consistent with the radical chemistry being driven by single electron transfer (SET) from KOtBu 18 (Scheme A). Instead, it suggests that removal by KOtBu of H+/D+ from this C–H/C–D bond controls the level of initiation of radical chemistry. This makes sense if this step is followed by the loss of a bromide anion affording a p-benzyne. More surprising information emerges from the reactions of the d 8 and d 9 isotopologues (entries 3 and 4), which have progressively stronger inhibitory effects. This shows that breaking the C–H at sites other than at position 10 also contributes to the flux of radical chemistry. This suggests that the site of deprotonation and loss of halide do not need to be on the same ring, i.e., that “remote benzynes” (“distal diradicals” resulting from loss of HBr from substrates; we suggest that distal diradicals is the better term) can also trigger these reactions.

Explaining the Methylated Products

Now returning to consider the methylated products from Scheme A, the structure of methylanthracene 45 was confirmed by 1H NMR and by high-resolution MS (HRMS), and its retention time on GCMS was confirmed by comparison with that of authentic 9-methylanthracene. The positions of the methyl groups in the other methylated products 46–48 have not been determined but are proposed by analogy to 45. These compounds arise through the BHAS reaction of methyl radicals 21 with the anthracenes present in the reaction mixture. Methyl radicals might arise through fragmentation of tert-butoxyl radicals 19, which, in turn, might form following electron transfer from KOtBu 18 (Scheme A), and so SET from KOtBu seemed at least a possible explanation.

Comparing the reactions conducted in C6D6 versus C6H6 (Table ), the amount of 9-methylanthracene 45 formed in C6H6 was much lower compared to the reaction in C6D6 (3.2 vs 6.3%). If tert-butoxyl radicals 19 were present and were fragmenting as in Scheme A, this could explain the data. In C6H6, the methyl radicals would (i) add to anthracenes or (ii) abstract an H from the solvent in a defined ratio. In C6D6, it would be more difficult to abstract a D atom, resulting in a higher percentage of addition to anthracenes. However, please note that an alternative rationale for this chemistry is provided below.

As mentioned above, the anthracene 41 generated from the reaction of 9-bromoanthracene 40 with KOtBu in C6D6 was unlabeled, suggesting KOtBu as the most likely source of the abstracted H atom. To investigate this, KOtBu- d 9 was used in a repeat experiment (Table , entry 3; spectra pp S38–S40). Treatment of 9-bromoanthracene 40 with KOtBu-d 9 in C6H6 still led to the formation of unlabeled anthracene 41, suggesting that, in this case, the hydrogen atom transfer (HAT) is solely or predominantly occurring from the solvent. That would mean that the selectivity for abstraction of an H versus a D atom was good, with KOtBu being used as a source of HAT when the solvent is C6D6, and C6H6 being used when KOtBu-d 9 is the base. In view of this, we reacted 40 with a combination of C6D6 and KOtBu-d 9, which yielded predominantly anthracene 41- d 1 along with undeuterated anthracene (Table , entry 4; spectra pp S41–S43). The undeuterated anthracene component shows that intermolecular hydrogen abstraction from anthracene C–H bonds competed with the abstraction of C–D from KOtBu-d 9 and C6D6. In this case, when faced with both deuterated reagents, then labeling did indeed occur; the yield of the anthracene was significantly decreased compared to the parent reaction (entry 1) with KOtBu in C6H6 (39.2 → 8.3%).

The information that KOtBu is used for hydrogen atom transfer (HAT) by the aryl radicals is very important and it raises the question of what happens to the product radicals 54 (Scheme A). The answer to that comes from comparing entries 1 and 3 or 2 and 4 in Table . Comparing entries 1 and 3, we see that using C6H6 as the solvent and changing from KOtBu to KOtBu-d 9 decrease the amount of methylanthracene isotopologue 45 4-fold (3.2 → 0.7%). This results from a 4-fold decrease in the flux of methyl radicals. This number represents a primary kinetic isotope effect associated with C–H/C–D bond cleavage in KOtBu in the rate-determining step (Scheme A). That is completely inconsistent with the picture shown in Scheme A. There, the reactions would simply involve electron transfer from KOtBu or KOtBu-d 9 to substrate 40. This could involve a minor quantitative change but far smaller than what is seen here. Similarly comparing entries 2 and 4 in Table shows that a primary deuterium isotope effect is evident. A primary deuterium isotope effect would align with HAT from KOtBu by diradicals 53 (or aryl radicals) and fragmentation of the resulting alkyl radical 54 to afford methyl radicals 21 and the potassium enolate of acetone, 55. Computational studies show that both the hydrogen atom abstraction from potassium tert-butoxide and the fragmentation of the resulting radical are easily achieved under the reaction conditions (see pp. S144–S146). That therefore means that electron transfer from KOtBu was NOT driving the reaction, but rather that formation of anthracenyl radicals 49 controls the formation of methyl radicals 21, in line with our other evidence for the role of benzynes.

6. (A) A Role for KOtBu in Hydrogen Atom Transfer (HAT) Reactions; (B) TEMPO Studies; and (C) Double BHAS Reactions on Substrates 68, 69, and 71 .

6

Open in a new tab

a Conditions: KOtBu (2 equiv), PhH, 130 °C, 24 h.

Turning now to results in Table , the yield of 9-methylanthracene 45 also decreases with an increasing level of deuteration of the bromoanthracene substrate. Again, this is completely inconsistent with the electron transfer from KOtBu. If that were controlling the initiation, then no notable change in the level of initiation would be expected with the different isotopologues. Instead, the level of radical flux is controlled by the speed of the removal of HBr/DBr from the substrate by KOtBu.

TEMPO Studies

Given the role of radicals in these reactions, we next used TEMPO to probe the intermediates (Scheme B). When bromoanthracene 40 was heated with KOtBu in benzene in the presence of TEMPO, this led to products including anthraquinone 56 and anthrone 57 in addition to TEMPO-trapped products 58 and 59. Anthraquinone arises by thermolysis of bis-TEMPO intermediate 60, which, in turn, comes from trapping of the p-benzyne 53 from the initiation of the reaction. This has clear direct analogy in the literature when Bergman cyclization of o-benzenediyne 62 led to formation of naphthalen-1,4-diyl 63; the initial bis-TEMPO-trapped intermediate 64 thermolyzed in situ to afford p-naphthoquinone 65. The anthrone 57 formed in our experiment would similarly form from thermolysis of 66, which in turn arises from trapping anthracenyl radical 49.

Studies with Different Bases

Given the outcome of the experiments of 9-bromoanthracene with KOtBu in C6H6, we also explored a number of variants that are detailed in the SI (pp S45–S59). The reaction worked well when KOEt was used, establishing that the reaction was not confined to KOtBu as the base (see SI). This was important because it showed that nontertiary alkoxides promote this reaction. Unlike tert-butoxide, ethoxide contains weak H–CO bonds that can quench aryl radicals, and so, in this case, quenching of anthracenyl radicals to form anthracene, a hydrodehalogenation reaction, was more prevalent than in reactions with KOtBu. The hydrodehalogenation reactions reviewed by Bunnett invariably used primary or secondary alkoxides and this therefore links the mechanism for radical formation in those reactions to our observations on BHAS reactions here.

When KOCEt3 was used instead of KOtBu, the reaction was again successful. As expected, KOCEt3 afforded traces of ethylated products rather than the methylated products seen with KOtBu. No reaction was seen with NaOtBu or LiOtBu, but this may have been due to the low solubility in benzene. Adding 15-crown-5 to the NaOtBu experiment allowed conversion to the expected products, including the BHAS products. 15-Crown-5 is an excellent H atom donor, and so anthracene 41 was the major product in this case (see pp S50–S52). A control reaction using KOtBu (but carried out while in darkness (under foil)) also afforded the products seen under normal daylight conditions, so the reaction was not light-dependent.

Dihalo-Substrates

Having studied the reactivities of monohalo substrates, we then explored the reactivity of three dibromo substrates 68, 69, and 71 (Scheme C). 9,10-Dibromoanthracene 68 was of interest because benzyne-type activation of this substrate specifically could only be achieved by deprotonation at a site remote from the C–Br bonds (i.e., not o-, m-, or p- to these bonds). Here, we compared the reaction of the parent to the reaction of its d 8 -isotopologue (Table ). Side-by-side reactions (15 min) of the two isotopologues were conducted under identical conditions. The yield of 9 dropped from 19.1% in the parent substrate to 2.8% in the deuterated substrate. This provides strong backing to the generality of the benzyne activation in this case, where no o-, m-, or p-benzyne can form. In this case, deprotonation at the periphery must lead to an “r-benzyne” (or “r-diyl”).

3. Reaction of 9,10-Dibromoanthracene.

graphic file with name ja5c22122_0009.jpg

    % yield
entry substrate 68 9 40 41 42 43 44 45 52
1 68- d 0 41.0 19.1 22.1 0.7 2.4 0.4 3.7 0.4 3.7
2 68- d 8 85.1 2.8 4.1 0.0 0.3 0.0 0.0 0.0 4.1
Open in a new tab
a

Isotopologue yields determined using GCFID calibrations for nondeuterated compounds.

The remaining substrates that we discuss here are vicinal dibromides 69 and 71. These substrates were interesting as o-dibromides that offered opportunities for intramolecular BHAS reactions. In 69, deprotonation can occur at m- and p- to the bromides, while in 71, deprotonation at a remote ring site is needed. Comparing the outcomes allows for an assessment of any penalty in 71 for requiring deprotonation in a remote ring. The complex multistep reactions lead to similar yields, suggesting that the need for remote deprotonation does not necessarily impede the reaction.

Conclusions

The route by which base-treatment of haloarenes affords aryl radicals, which has puzzled chemists since 1899, is now revealed.

  • (i)

    Deuterium labeling establishes that treatment of iodobenzene with KOtBu in benzene affords simultaneously o-, m-, p-benzynes through deprotonation and loss of halide anion but, starting from C6D5I, the formation of specifically deuterated o-, m-, and p-terphenyl products shows that each type of benzyne is able to trap two benzenes regiospecifically through radical reactions and thereby they act as initiators of radical chain chemistry. As initiators, they only need to be present in minute quantities.

  • (ii)

    A study of specifically deuterated haloanthracenes shows that base-induced removal of H+/D+ does not need to occur in the ring that bears the halogen substituent but can happen in a neighboring ring, thereby giving rise to distal diradicals that are effectively “r-benzynes” (remote benzynes).

  • (iii)

    When KOtBu is used as a base, low levels of methylated arenes are also formed that arise from methyl radicals. These are generated when aryl radicals or diradicals abstract an H atom from a C–H bond in KOtBu and the resulting carbon radical undergoes a previously unreported fragmentation.

The results are completely at odds with suggestions that KOtBu acts as an electron donor to haloarenes. They rationalize the initiation pathway for the protodehalogenations observed since 1899 as reviewed by Bunnett.

Wider Applications?

Questions arise about how widespread our observations may prove to be. Clearly, any case in which alkoxides encounter aryl halides could lead to benzynes that trigger further reactions. Some colleagues report on preferences for sodium alkoxides over potassium alkoxides as bases in synthetic transformations such as Buchwald–Hartwig aminations. It may be that potassium alkoxides are more effective bases on aryl halides and promote byproducts to a greater extent, but this would require detailed analysis. The role of alkoxides in deprotonating arenes is not necessarily confined to aryl halides but can also be manifested in other arenes that may be subject to deprotonation.

One case in which KOtBu was suggested to act as an electron donor to a surprising substrate was reported by Lei and Jutand. They showed that the reaction of KOtBu E ox = +0.10 V (vs. SCE in DMF) with phenanthroline E red ≈ −2.06 V (vs. SCE in DMF) afforded radical anions derived from phenanthroline. They proposed that electron transfer from KOtBu was taking place despite the large gap in potentials. We proposed an alternative explanation, where deprotonation of phenanthroline by KOtBu led to an electron-rich dimeric product and, in turn, to formation of radical anion organic electron donors. , The take-home message is a reassuring one, that KOtBu is a notably poor electron donor in its ground state − and that its radical-forming properties derive from its ability to deprotonate aryl halides and to convert them to benzynes/arenediyls, or to deprotonate other arenes.

Supplementary Material

ja5c22122_si_001.pdf (8.8MB, pdf)

Acknowledgments

We thank (i) EPSRC and GSK for i-CASE studentship (grant number EP/W522260/1 to S.T.), (ii) AZ and University of Strathclyde for a studentship to K.F.C., and (iii) University of Strathclyde for studentships (to A.J.S. and K.K.). We thank Craig Irving, Patricia Keating, Dr. Jessica Bame, and Dr. Graeme Anderson for assistance with spectroscopy. The calculations reported in this manuscript were performed using the ARCHIE-WeSt High-Performance Computer, hosted at the University of Strathclyde (archie-west.ac.uk); we thank J. Buzzard, Dr K. Karina-Ossowska, and Dr R. Martin for their assistance with this facility.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/jacs.5c22122.

  • Experimental procedures and spectroscopic data in support of the results are reported; X-ray data for compound 69 are available (Database ref number is CCDC 2211274) Computational data underpinning this study can be downloaded from the ioChem-BD database , at the following URL: https://doi.org/10.19061/iochem-bd-6-628 (PDF)

∥.

S.T. and K.F.C. contributed equally. The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.

Traces of acetonylanthracene were detected. The potassium enolate of acetone can undergo easy abstraction of an H atom, and the resulting radical may react to form acetonylanthracene or the enolate could undergo electron transfer to the aryl halide (see pp S45 and S46).

The authors declare no competing financial interest.

References

  1. Jackson C. L., Gazzolo P. S.. On Certain Derivatives of Symmetrical Trichlorobenzol. Am. Chem. J. 1899;22:50–60. doi: 10.2307/20020876. [DOI] [Google Scholar]
  2. Bunnett J. F.. Radical-Chain, Electron-Transfer Dehalogenation Reactions. Acc. Chem. Res. 1992;25:2–9. doi: 10.1021/ar00013a001. [DOI] [Google Scholar]
  3. Yanagisawa S., Ueda K., Taniguchi T., Itami K.. Potassium t-Butoxide Alone Can Promote the Biaryl Coupling of Electron-Deficient Nitrogen Heterocycles and Haloarenes. Org. Lett. 2008;10:4673–4676. doi: 10.1021/ol8019764. [DOI] [PubMed] [Google Scholar]
  4. Sun C. L., Li H., Yu D.-G., Yu M., Zhou X., Lu X.-Y., Huang K., Zheng Z.-B., Li S.-F., Shi J.. An efficient organocatalytic method for constructing biaryls through aromatic C-H activation. Nat. Chem. 2010;2:1044–1049. doi: 10.1038/nchem.862. [DOI] [PubMed] [Google Scholar]
  5. Sun C. L., Gu Y.-F., Wang B., Shi Z.-J.. Direct Arylation of Alkenes with Aryl Iodides/Bromides through an Organocatalytic Radical Probe. Chem. - Eur. J. 2011;17:10844–10847. doi: 10.1002/chem.201101562. [DOI] [PubMed] [Google Scholar]
  6. Sun C. L., Gu Y.-F., Huang W.-P., Shi Z.-J.. Neocuproine-KOtBu promoted intramolecular cross coupling to approach fused rings. Chem. Commun. 2011;47:9813–9815. doi: 10.1039/C1CC13907J. [DOI] [PubMed] [Google Scholar]
  7. Shirakawa E., Itoh K.-I., Higashino T., Hayashi T.. tert-Butoxide-Mediated Arylation of Benzene with Aryl Halides in the Presence of a Catalytic 1,10-Phenanthroline Derivative. J. Am. Chem. Soc. 2010;132:15537–15539. doi: 10.1021/ja1080822. [DOI] [PubMed] [Google Scholar]
  8. Shirakawa E., Zhang X., Hayashi T.. Mizoroki-Heck-Type Reaction Mediated by Potassium tert-Butoxide. Angew. Chem., Int. Ed. 2011;50:4671–4674. doi: 10.1002/anie.201008220. [DOI] [PubMed] [Google Scholar]
  9. Shirakawa E., Hayashi T.. E. Transition-metal-free Coupling Reactions of Aryl Halides. Chem. Lett. 2012;41:130–134. doi: 10.1246/cl.2012.130. [DOI] [Google Scholar]
  10. Liu W., Cao H., Zhang H., Zhang H., Chung K. H., He C., Wang H., Kwong F. Y., Lei A.. Organocatalysis in Cross-Coupling: DMEDA-Catalyzed Direct C–H Arylation of Unactivated Benzene. J. Am. Chem. Soc. 2010;132:16737–16740. doi: 10.1021/ja103050x. [DOI] [PubMed] [Google Scholar]
  11. Studer A., Curran D. P.. Organocatalysis and C-H Activation Meet Radical- and Electron-Transfer Reactions. Angew. Chem., Int. Ed. 2011;50:5018–5022. doi: 10.1002/anie.201101597. [DOI] [PubMed] [Google Scholar]
  12. Doni E., Murphy J. A.. Evolution of neutral organic super-electron-donors and their applications. Chem. Commun. 2014;50(60):6073–6087. doi: 10.1039/c3cc48969h. [DOI] [PubMed] [Google Scholar]
  13. Murphy J. A.. Discovery and Development of Organic Super-Electron-Donors. J. Org. Chem. 2014;79:3731–3746. doi: 10.1021/jo500071u. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Murphy J. A., Khan T. A., Zhou S.-Z., Thomson D. W., Mahesh M.. Highly Efficient Reduction of Unactivated Aryl and Alkyl Iodides by a Ground-State Neutral Organic Electron Donor. Angew. Chem., Int. Ed. 2005;44:1356–1360. doi: 10.1002/anie.200462038. [DOI] [PubMed] [Google Scholar]
  15. Zhou S., Anderson G. M., Mondal B., Doni E., Ironmonger V. M., Kranz M., Tuttle T., Murphy J. A.. Organic super-electron-donors: initiators in transition metal-free haloarene–arene coupling. Chem. Sci. 2014;5:476–482. doi: 10.1039/C3SC52315B. [DOI] [Google Scholar]
  16. Zhou S., Doni E., Anderson G. M., Kane R. G., MacDougall S. W., Ironmonger V. M., Tuttle T., Murphy J. A.. Identifying the Roles of Amino Acids, Alcohols and 1,2-Diamines as Mediators in Coupling of Haloarenes to Arenes. J. Am. Chem. Soc. 2014;136:17818–17826. doi: 10.1021/ja5101036. [DOI] [PubMed] [Google Scholar]
  17. Nocera G., Murphy J. A.. Ground State Cross-Coupling of Haloarenes with Arenes Initiated by Organic Electron Donors, Formed in situ: An Overview. Synthesis. 2020;52:327–336. doi: 10.1055/s-0039-1690614. [DOI] [Google Scholar]
  18. Smith A. J., Poole D. L., Murphy J. A.. The role of organic electron donors in the initiation of BHAS base-induced coupling reactions between haloarenes and arenes. Sci. China Chem. 2019;62:1425–1438. doi: 10.1007/s11426-019-9611-x. [DOI] [Google Scholar]
  19. For a review of KOtBu in organic bond formations, see:; Madasu J., Shinde S., Das R., Patel S., Shard A.. Potassium tert-butoxide mediated C-C, C-N, C-O and C-S bond forming reactions. Org. Biomol. Chem. 2020;18:8346–8365. doi: 10.1039/D0OB01382J. [DOI] [PubMed] [Google Scholar]
  20. Bhakuni B. S., Kumar A., Balkrishna S. J., Sheikh J. A., Konar S., Kumar S.. KO t Bu Mediated Synthesis of Phenanthridinones and Dibenzoazepinones. Org. Lett. 2012;14:2838–2841. doi: 10.1021/ol301077y. [DOI] [PubMed] [Google Scholar]
  21. De S., Ghosh S., Bhunia S., Sheikh J. A., Bisai A.. Intramolecular Direct Dehydrohalide Coupling Promoted by KO t Bu: Total Synthesis of Amaryllidaceae Alkaloids Anhydrolycorinone and Oxoassoanine. Org. Lett. 2012;14:4466–4469. doi: 10.1021/ol3019677. [DOI] [PubMed] [Google Scholar]
  22. Wu Y., Wong S. M., Mao F., Chan T. L., Kwong F. Y.. Intramolecular Direct C–H Bond Arylation from Aryl Chlorides: A Transition-Metal-Free Approach for Facile Access of Phenanthridines. Org. Lett. 2012;14:5306–5309. doi: 10.1021/ol302489n. [DOI] [PubMed] [Google Scholar]
  23. Wu Y., Choy P. Y., Kwong F. Y.. A Direct C-H Arylation of Unactivated Arenes Promoted by Mixed Potassium Alkoxides. Asian J. Org. Chem. 2014;3:1262–1265. doi: 10.1002/ajoc.201402181. [DOI] [Google Scholar]
  24. Liu W., Tian F., Wang X., Yu H., Bi Y.. Simple alcohols promoted direct C–H arylation of unactivated arenes with aryl halides. Chem. Commun. 2013;49:2983–2985. doi: 10.1039/c3cc40695d. [DOI] [PubMed] [Google Scholar]
  25. Cuthbertson J., Gray V. J., Wilden J. D.. Observations on transition metal free biaryl coupling: potassium tert-butoxide alone promotes the reaction without diamine or phenanthroline catalysts. Chem. Commun. 2014;50:2575–2578. doi: 10.1039/C3CC49019J. [DOI] [PubMed] [Google Scholar]
  26. Rueping M., Leiendecker M., Das A., Poisson T., Bui L.. Potassium tert-butoxide mediated Heck-type cyclization/isomerization–benzofurans from organocatalytic radical cross-coupling reactions. Chem. Commun. 2011;47:10629–10631. doi: 10.1039/c1cc14297f. [DOI] [PubMed] [Google Scholar]
  27. Yi H., Jutand A., Lei A.. Evidence for the interaction between t BuOK and 1,10-phenanthroline to form the 1,10-phenanthroline radical anion: a key step for the activation of aryl bromides by electron transfer. Chem. Commun. 2015;51:545–548. doi: 10.1039/C4CC07299E. [DOI] [PubMed] [Google Scholar]
  28. Pause L., Robert M., Savéant J.-M.. Can Single-Electron Transfer Break an Aromatic Carbon–Heteroatom Bond in One Step? A Novel Example of Transition between Stepwise and Concerted Mechanisms in the Reduction of Aromatic Iodides. J. Am. Chem. Soc. 1999;121:7158–7159. doi: 10.1021/ja991365q. [DOI] [Google Scholar]
  29. Finn M., Friedline R., Suleman N. K., Wohl C. J., Tanko J. M.. Chemistry of the t-Butoxyl Radical: Evidence that Most Hydrogen Abstractions from Carbon are Entropy-Controlled. J. Am. Chem. Soc. 2004;126:7578–7584. doi: 10.1021/ja0493493. [DOI] [PubMed] [Google Scholar]
  30. Griller D., Ingold K. U.. Free-radical clocks. Acc. Chem. Res. 1980;13:317–323. doi: 10.1021/ar50153a004. [DOI] [Google Scholar]
  31. Walling C., Wagner P. J.. Positive Halogen Compounds. X. Solvent Effects in the Reactions of t-Butoxy Radicals. J. Am. Chem. Soc. 1964;86:3368–3375. doi: 10.1021/ja01070a032. [DOI] [Google Scholar]
  32. Zady M. F., Wong J. L.. Reactivities and electronic aspects of nucleic acid heterocycles. Part 6. Kinetics and mechanism of carbon-8 methylation of purine bases and nucleosides by methyl radical. J. Am. Chem. Soc. 1977;99:5096–5100. doi: 10.1021/ja00457a033. [DOI] [PubMed] [Google Scholar]
  33. Rong X., Jin L., Gu Y., Liang G., Xia Q.. Transition-Metal-Free Radical C–H Methylation of Quinoxalinones with TBHP. Asian J. Org. Chem. 2020;9:185–188. doi: 10.1002/ajoc.201900758. [DOI] [Google Scholar]
  34. Jin S., Yao H., Lim S., You X., Yang Y., Yan Z.. Peroxide-mediated site-specific C–H methylation of imidazo­[1,2-a]­pyridines and quinoxalin-2­(1H)-ones under metal-free conditions. Org. Biomol. Chem. 2020;18:205–210. doi: 10.1039/C9OB02328C. [DOI] [PubMed] [Google Scholar]
  35. Li Z.-L., Jin L.-K., Cai C.. Efficient synthesis of 2-substituted azoles: radical C–H alkylation of azoles with dicumyl peroxide, methylarenes and cycloalkanes under metal-free condition. Org. Chem. Front. 2017;4:2039–2043. doi: 10.1039/C7QO00396J. [DOI] [Google Scholar]
  36. Li G., Yang S., Lv B., Han Q., Ma X., Sun K., Wang Z., Zhao F., Lv Y., Wu H.. Metal-free methylation of a pyridine N-oxide C–H bond by using peroxides. Org. Biomol. Chem. 2015;13:11184–11188. doi: 10.1039/C5OB01900A. [DOI] [PubMed] [Google Scholar]
  37. Zhang P.-Z., Li J.-A., Zhang L., Shoberu A., Zou J.-P., Zhang W.. Metal-free radical C–H methylation of pyrimidinones and pyridinones with dicumyl peroxide. Green Chem. 2017;19:919–923. doi: 10.1039/C6GC03355E. [DOI] [Google Scholar]
  38. Gassman P. G., Benecke H. P., Murphy T. J.. The 1.2 Versus 1,4 addition of “dienophiles” to 1,3-cis,trans-cyclooctadiene. Tetrahedron Lett. 1969;10:1649–1652. doi: 10.1016/S0040-4039(01)87970-3. [DOI] [Google Scholar]
  39. Leitich J.. [2 + 2] Cycloaddition of benzyne to cis,trans and cis,cis −1,5-cyclooctadiene. Tetrahedron Lett. 1980;21:3025–3028. doi: 10.1016/S0040-4039(00)77397-7. [DOI] [Google Scholar]
  40. Bowne A. T., Christopher T. A., Levin R. H.. The cycloaddition of o-benzyne to (E)-deuterio-tert-butylethylene. Tetrahedron Lett. 1976;17:4111–4114. doi: 10.1016/S0040-4039(00)74608-9. [DOI] [Google Scholar]
  41. Cadogan J. I. G., Hall K. A., Sharp J. T., Robertson A. K.. Reactions of bromodurene and related compounds with strong bases: a novel 1,3-dehydrobromination. Chem. Commun. 1971:1273–1274. doi: 10.1039/c29710001273. [DOI] [Google Scholar]
  42. Although mechanism is not discussed, reaction of benzyne with benzene to form biphenyl is known:; Miller R. G., Stiles M.. Reaction of Benzyne with Benzene and Naphthalene. J. Am. Chem. Soc. 1963;85:1798–1800. doi: 10.1021/ja00895a023. [DOI] [Google Scholar]
  43. Clark K. F., Tyerman S., Evans L., Robertson C. M., Murphy J. A.. An assay for aryl radicals using BHAS coupling. Org. Biomol. Chem. 2024;22:1018–1022. doi: 10.1039/D3OB01743E. [DOI] [PubMed] [Google Scholar]
  44. Heaney H., Millar I. T.. Triphenylene. Org. Synth. 1960;40:105. doi: 10.15227/orgsyn.040.0105. [DOI] [Google Scholar]
  45. Traynham J. G.. Ipso substitution in free-radical aromatic substitution reactions. Chem. Rev. 1979;79:323–330. doi: 10.1021/cr60320a002. [DOI] [Google Scholar]
  46. Ueno R., Ikeda Y., Shirakawa E.. tert-Butoxy-Radical-Promoted α-Arylation of Alkylamines with Aryl Halides. Eur. J. Org. Chem. 2017;2017:4188–4193. doi: 10.1002/ejoc.201700548. [DOI] [Google Scholar]
  47. For fluoride anion-induced route to m-benzynes, featuring silyl and hypervalent halogen leaving groups, see; Koyamada K., Miyamoto K., Uchiyama M.. Room-temperature synthesis of m-benzyne. Nat. Synth. 2024;3:1083–1090. doi: 10.1038/s44160-024-00572-y. [DOI] [Google Scholar]
  48. Perrin C. L.. Nonradical Reactions of p-Benzyne Diradicals. J. Org. Chem. 2025;90:8445–8456. doi: 10.1021/acs.joc.5c00517. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Perrin C. L., Rodgers B. L., O’Connor J. M.. Nucleophilic Addition to a p-Benzyne Derived from an Enediyne: A New Mechanism for Halide Incorporation into Biomolecules. J. Am. Chem. Soc. 2007;129:4795–4799. doi: 10.1021/ja070023e. [DOI] [PubMed] [Google Scholar]
  50. Sandhiya L., Jangra H., Zipse H.. Molecule-Induced Radical Formation (MIRF) Reactions - A Reappraisal. Angew. Chem., Int. Ed. 2020;59:6318–6329. doi: 10.1002/anie.201912382. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Rüchardt C., Gerst M., Ebenhoch J.. Uncatalyzed Transfer Hydrogenation and Transfer Hydrogenolysis: Two Novel Types of Hydrogen-Transfer Reactions. Angew. Chem., Int. Ed. 1997;36:1406–1430. doi: 10.1002/anie.199714061. [DOI] [Google Scholar]
  52. Scherübl M., Daniliuc C. G., Studer A.. Arynes as Radical Acceptors: TEMPO-Mediated Cascades Comprising Addition, Cyclization, and Trapping. Angew. Chem., Int. Ed. 2021;60:711–715. doi: 10.1002/anie.202012654. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Lezius L., Horst E. S., Scherübl M., Lammert J., Daniliuc C. G., Mori T., Yamaguchi S., Studer A.. Blatter and Verdazyl Radicals with Arynes: Synthesis and Investigation of N-Chiral, Antiaromatic Triazines, and Tetrazinones. Angew. Chem., Int. Ed. 2026:e20021. doi: 10.1002/anie.20252002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Bergstrom F. W., Wright R. E., Chandler C., Gilkey W. A.. The action of bases on organic halogen compounds. I. The reaction of aryl halides with potassium amide. J. Org. Chem. 1936;01:170–177. doi: 10.1021/jo01231a007. [DOI] [Google Scholar]
  55. Li Z., Shen J., Jiang Q., Shen S., Zhou J., Zeng H.. Directionally Aligned Crown Ethers as Superactive Organocatalysts for Transition-metal Ion-free Arylation of Unactivated Arenes. Chem. - Asian J. 2022;17:e202200303. doi: 10.1002/asia.202200303. [DOI] [PubMed] [Google Scholar]
  56. Grissom J. W., Gunawardena G. U.. Intermolecular reaction of nitroxide radicals with biradical intermediates generated from aromatic enediynes. Tetrahedron Lett. 1995;36:4951–4954. doi: 10.1016/0040-4039(95)00967-H. [DOI] [Google Scholar]
  57. When KOtBu is replaced by NaH, the electron donor still forms.; Nozawa-Kumada K., Iwakawa Y., Onuma S., Shigeno M., Kondo Y.. NaH-mediated direct C–H arylation in the presence of 1,10-phenanthroline. Chem. Commun. 2020;56:7773–7776. doi: 10.1039/D0CC00730G. [DOI] [PubMed] [Google Scholar]
  58. But see; Zhao Y., Zhu H., Sung S., Wink D. J., Zadrozny J. M., Driver T. G.. Counterion Control of t-BuO-Mediated Single Electron Transfer to Nitrostilbenes to Construct N-Hydroxyindoles or Oxindoles. Angew. Chem., Int. Ed. 2021;60:19207–19213. doi: 10.1002/anie.202104319. [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Xu Z., Gao L., Wang L., Gong M., Wang W., Yuan R.. Visible Light Photoredox Catalyzed Biaryl Synthesis Using Nitrogen Heterocycles as Promoter. ACS Catal. 2015;5:45–50. doi: 10.1021/cs5011037. [DOI] [Google Scholar]
  60. Nocera G., Young A., Palumbo F., Emery K. J., Coulthard G., McGuire T., Tuttle T., Murphy J. A.. Electron Transfer Reactions: KOtBu (but not NaOtBu) Photoreduces Benzophenone under Activation by Visible Light. J. Am. Chem. Soc. 2018;140:9751–9757. doi: 10.1021/jacs.8b06089. [DOI] [PubMed] [Google Scholar]
  61. Alvarez-Moreno M., de Graaf C., Lopez N., Maseras F., Poblet J. M., Bo C.. Managing the Computational Chemistry Big Data Problem: The ioChem-BD Platform. J. Chem. Inf. Model. 2015;55:95–103. doi: 10.1021/ci500593j. [DOI] [PubMed] [Google Scholar]
  62. Bo C., Maseras F., Lopez N.. The Role of Computational Results Databases in Accelerating the Discovery of Catalysts. Nat. Catal. 2018;1:809–810. doi: 10.1038/s41929-018-0176-4. [DOI] [Google Scholar]

Associated Data

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

Supplementary Materials

ja5c22122_si_001.pdf (8.8MB, pdf)

Articles from Journal of the American Chemical Society are provided here courtesy of American Chemical Society

RESOURCES