Diazene-Catalyzed Oxidative Alkyl Halide-Olefin Metathesis

Abstract

The first platform for oxidative alkyl halide-olefin metathesis is described. The procedure employs diazenes as catalysts, which effect cyclization of alkenyl alkyl halides to generate cyclic olefins and carbonyl products. The synthesis of phenanthrene, coumarin, and quinolone derivatives is demonstrated, as well as the potential to apply this strategy to other electrophiles.

Graphical Abstract

Metathesis reactions can accelerate the construction of complex products from relatively simple starting materials through the catalytic exchange of their component functional groups. While olefin¹ and alkyne² metathesis are the most well-developed reactions in this category, other isofunctional reactions including carbonyl-olefin metathesis (COM) have been the focus of increasing attention (Figure 1A).^3,4 However, despite notable progress, many functional group metathesis reactions remain unknown. For example, alkyl halides and olefins are two abundant functional groups in organic chemistry, but there is currently no way to achieve their metathetical exchange (Figure 1B). Alkyl halides and olefins can be combined through radical pathways⁵ or via Heck reactions,⁶ but fragment exchange is not a typical outcome. Indeed, it is not obvious how the known strategies for functional group-olefin metathesis chemistries might be expanded to include alkyl halides. Nevertheless, the realization of an alkyl-halide olefin metathesis (AHOM) reaction could enable a variety of novel synthetic strategies for small molecule and materials synthesis. Here, we introduce the first strategy to achieve AHOM using diazene catalysis.⁷

Figure. 1.

Alkyl Halide-Olefin Metathesis: (A) Established functional group-olefin metathesis reactions. (B) Generic structures of alkyl halides and olefins. (C) General design of diazene-catalyzed alkyl halide-olefin metathesis. Colored circles and “R” represent generic substituents. [M] = metal with associated ligands.

The inspiration for this chemistry arose from our work in catalytic carbonyl-olefin metathesis (COM),^4a,8 which has resulted in the development of a number of ring-opening,^8,9 ring-closing,¹⁰ and ROMP¹¹ reactions. The key intermediate underpinning these reactions is the hydrazonium ion (i.e. a protonated azomethine imine), which is formed by condensation of a 1,2-dialkyhydrazine catalyst with the carbonyl component and which achieves metathesis by reversible [3+2]-cycloaddition with the olefin component. The key realization for the current work was that the same hydrazonium intermediate might instead be accessed by alkylation of a diazene followed by tautomerization.¹² If viable, this alternative hydrazonium synthesis could enable the engagement of a broad range of alkylating agents in metathesis reactions with olefins.

The blueprint for this AHOM idea is shown in Figure 1C. An appropriately substituted diazene catalyst 3 can undergo alkylation by an alkyl halide 1 to form diazenium 6. N-Alkyldiazeniums can tautomerize to form the corresponding hydrazoniums 7,¹² thus intersecting with our carbonyl-olefin metathesis strategy. [3+2]-cycloaddition of hydrazonium 7 with the olefin substrate 2 produces an intermediate pyrazolidine cycloadduct 8.¹³ Due to the symmetry of this intermediate, cycloreversion to alternative, metathesized fragments in the form of olefin 4 and hydrazonium 9 can occur. In theory, tautomerization of 9 to diazenium 10 followed by dealkylation by the halide counteranion would lead to the new alkyl halide product 11 and regenerate the diazene 3. In practice, however, we have found that hydrolysis of 9 outcompetes this pathway, which leads instead to the formation of a carbonyl product 5. Although this hydrolysis necessarily returns the catalyst in its reduced form as hydrazine 12, the autoxidation of hydrazines is usually quite facile.¹⁴ Thus, adventitious oxygen can lead to oxidation to reform the diazene 3. Given this behavior, we have termed this reaction oxidative alkyl halide-olefin metathesis.

We began our investigation of this concept by examining the ring-closing metathesis of alkenyl bromide 13c (Figure 2).¹⁵ For reference, 13c spontaneously undergoes a Prins-type cyclization under thermal conditions, leading to the formation of 9-isopropylphenanthrene (14). However, in the presence of 2,3-diazabicyclo[2.2.2]oct-2-ene (17) or benzo[c]cinnoline (18) under various conditions described below, 13c converted instead to phenanthrene (15) and acetone (16c), the products of the oxidative ring-closing alkyl halide-olefin metathesis.

Figure 2.

Divergent pathways for substrate 13c.

In the development of this AHOM reaction, we first examined the effect of the alkene substituents by screening several different olefins, 13a-c (Table 1). Using 1 equiv diazene 17 and a reaction temperature of 65 °C in MeCN, all three substrates converted to phenanthrene in nearly quantitative yield (entries 1–3), with the isopropylidene substrate 13c requiring the shortest reaction time (entry 3). Unfortunately, we were unable to achieve catalyst turnover with 17, which we believe was due to the difficulty of reoxidizing the protonated hydrazinium intermediate (cf. 12, Figure 1C). To remedy this situation, and to identify a catalyst with greater potential for electronic tuning,¹⁶ we examined benzo[c]cinnoline 18 (BCC) as an alternative diazene. We had previously demonstrated the utility of BCC as a catalyst for the oxidation of alkyl halides to aldehydes,¹⁷ in which turnover was readily achieved by autoxidation with adventitious O₂, and we reasoned that the spontaneous oxidation of BCC should enable a catalytic reaction. Olefins 13a-c were again screened with stoichiometric 18 to determine the most reactive olefin (entries 4–6). In this case, olefin 13b was found to be optimal, leading to phenanthrene (15) in 85% yield. Interestingly, the cis-isomer was more reactive than the trans, which we suspect may be due to steric strain between the benzocinnoline framework and the trans olefin substituent inhibiting the cycloaddition step. While olefin 13c was determined to be the most reactive substrate with diazene 17, the use of 18 led primarily to 9-isopropylphenanthrene (14) as the major product instead of the metathesis adduct 15. We speculate that 18, being less electron-rich, has a slower rate of alkylation, which thus leads to the competitive formation of the undesired side-product.

Table 1.

Optimization of olefin moiety for the oxidative alkyl halide-olefin metathesis to form phenanthrene 15.^a

entry	diazene	substrate	temp (°C)	time (h)	yield (%)
1	17	13a	65	20	96
2	17	13b	65	8	96
3	17	13c	65	5	96
4	18	13a	100	18	54
5	18	13b	100	18	85
6	18	13c	100	18	38

^aConditions: substrate (1.0 equiv) and diazene (1.0 equiv) were dissolved in MeCN (0.2 M) and heated in a sealed vial until starting material could no longer be detected by TLC. Yields were determined on purified products.

After identifying the most productive olefin substrate, we examined the possibility of catalysis with 18 (Table 2). Treatment of 13b with 20 mol% 18 at 100–120 °C for 48 h led to the formation of phenanthrene (15) in varying yields (entries 1–5), with MeCN being optimal (entry 6). In all cases, adventitious water and O₂ were sufficient to induce catalyst turnover. Reducing catalyst loading (entries 7 and 8) was detrimental to yield.

Table 2.

Development of a catalytic oxidative AHOM of 13b.^a

entry	18 (mol%)	solvent	temp (°C)	time (h)	yield (%)
1	20	THF	100	48	54
2	20	DMF	100	48	50
3	20	DMSO	100	48	0
4	20	PhMe	120	48	25
5	20	MeCN	120	48	70
6	20	MeCN	120	16	74
7	10	MeCN	120	48	39
8	5	MeCN	120	48	13

^aConditions: substrate 13b (1.0 equiv) and diazene (0.05 – 0.2 equiv) were dissolved in organic solvent (0.2 M) and heated in a sealed vial until starting material could no longer be detected by TLC. Yields were determined on purified products.

To probe the impact of catalyst structure on reaction rate and efficiency, we prepared BCC derivatives 21 and 22 bearing electron-donating and electron-withdrawing substituents, respectively, and compared their capacities to achieve the ring-closing alkyl halide-olefin metathesis of 19 to form coumarin (20) (Table 3). At 20 mol% loading in wet acetonitrile at 120 °C, BCC (18) generated 20 in 41% yield. Surprisingly, BCCs 21 and 22 did not produce any amount of 20 (entries 2 and 3). Because there was no observed change to 21 or the substrate, we conclude that 21 is too electron-poor to react with the substrate at a reasonable rate (entry 2). On the other hand, 22 was alkylated readily by the substrate, but the resulting benzocinnolinium did not convert to product (entry 3). We rationalized that the 1,3-prototropic shift was inefficient with catalyst 22, and so we examined the addition of base to facilitate this step. While pyridine was ineffective because it underwent alkylation by the substrate (entry 4), both 2,6-lutidine (entry 5) and i-Pr₂NEt (entry 6) led to appreciable increases in yield with catalyst 18. Even with base, catalyst 21 failed to deliver any product (entry 7), but the methoxy-substituted catalyst 22 with base produced coumarin in 76% yield.

Table 3.

Exploration of catalyst structure and the impact of base for diazene-catalyzed AHOM of 19.^a

entry	catalyst	additive	yield (%)
1	18	-	41
2	21	-	0
3	22	-	0
4	18	pyridine	0
5	18	2,6-lutidine	52
6	18	i-Pr2NEt	54
7	21	i-Pr2NEt	0
8	22	i-Pr2NEt	76

^aConditions: substrate 19 (1.0 equiv) and diazene (0.2 equiv) were dissolved in MeCN (0.2 M) and heated in a sealed vial for 12 h. Yields were determined on purified products.

Using the optimized conditions from entry 8, we found it was possible to prepare several coumarin 20-24 and quinolone 25-27 derivatives (Figure 3). Both electron-withdrawing and electron-donating substituents were found to be well tolerated, with all coumarin and quinolone derivates being delivered in good to excellent yields.

Figure 3.

Coumarin and quinolone synthesis via AHOM.

Finally, because the key step in initiating the catalytic cycle is an alkylation, we reasoned that other electrophiles should also be subject to metathesis with olefins under this paradigm. For example, we hypothesized that alcohol 28 might alkylate a diazene catalyst under ionizing conditions. Indeed, we found that treatment of 28 with 1.0 equiv diazene 17 and 2.0 equiv TFA in a wet acetonitrile solution at 120 °C for 12 h gave rise to phenanthrene (15) in 78% yield (eq 1). Benzo[c]cinnoline (18) furnished the product in more modest yield (48%); however, we were unable to achieve catalyst turnover. Neither benzo[c]cinnolines 21 nor 22 led to any product, which we attribute to the poor nucleophilicity of 21 and the difficulty of 22 undergoing the requisite 1,3-prototropic shift in the absence of base.

(1)

In summary, we have developed the first strategy for metathesis of alkyl halides and olefins using diazene catalysis. As with other metathesis chemistries, a number of unique synthetic applications can be envisioned for this type of exchange. We anticipate that further development of this chemistry could arise from the development of diazene catalysts with improved rates of alkylation, cycloaddition, and cycloreversion, and through the identification of alternative electrophiles to give entry to the key hydrazonium intermediate.