5-Hexenyl bromide

One experimental test for the involvement of radical intermediates is to study 5-hexenyl systems and look for the characteristic cyclization to cyclopentane derivatives (see Part A, Section 11.2.3). When 5-hexenyl bromide or iodide reacts with LiAlH4, no cyclization products are observed. However, the more hindered 2,2-dimethyl-5-hexenyl iodide gives mainly cyclic product.164... 

Small amounts of cyclized products are obtained after the preparation of Grignard reagents from 5-hexenyl bromide.9 This indicates that cyclization of the intermediate radical competes to a small extent with combination of the radical with the metal. Quantitative kinetic models that compare competing processes are consistent with diffusion of the radicals from the surface.10 Alkyl radicals can be trapped with high efficiency by the nitroxide radical TMPO.11 Nevertheless, there remains disagreement about the extent to which the radicals diffuse away from the metal surface.12... 

Reaction (7.13) shows the simplest and most popular type of hex-5-en-l-yl radical cyclization mediated by (TMS)3SiH. Thus, a 50 mmol solution of 5-hexenyl bromide and silane in the reported conditions leads to a 24 1 ratio of cyclized vs uncyclized products [6]. Similar results were obtained by replacing the bromide with isocyanide [7]. 

The intramolecular C—C bond formation (or cyclization) mediated by (TMS SiH has been the subject of numerous publications. Scheme 5 presents the simplest and more popular type of 5-hexenyl radical cyclization. Thus, a 50 mmolar solution of 5-hexenyl bromide and silane (or tin hydride) lead to a 24 1 (or 6 1) ratio of cyclized versus uncyclized products61. Therefore, under the same conditions the silane gives higher yields of cyclization than the stannane. 

Hexenyl bromide 22 can be activated similarly to 5-hexenyl radical 22A, which cyclizes to cyclopentylmethyl radical 22B with a rate constant of 2.3 x 105 s 1 [91]. For this slower reaction, a competition between direct trapping of 22A to 24 and 5-exo cyclization to 22B followed by coupling to 23 is often found. These unimolcular reaction steps are in comparison to bimolecular transition metal-centered transformations faster or at least as fast. 

To calculate the product distribution for Grignard reagent formation from 5-hexenyl bromide. Garst et al. [93b,c] used the following equations and compared the results with the experimental data of Bodewitz et al. [76] for the reaction of 5-hexenyl bromide with magnesium (see Sec. III.E.2). 

Table 13 Product Distribution for the Reaction of 5-Hexenyl Bromide with Magnesium in Ether at 40 "C...

Calculations for Grignard reagent formation from 5-hexenyl bromide and product distribution in other solvents such as THF, DBE, DPE, using similar assumptions and neglecting the reactions of radicals with solvents were also compared in each case with the experimental results of Bodewitz et al. [76] to support the argument for the greater predictive competence of the D-model. 

More recently, the authors of the D-model investigated the reaction of bromocyclopropane 133 with magnesium [93e], using again the equations V=l4kJ3(kfD) ]v, A = (D//cJ mentioned earlier to calculate the product distribution, but substituting another composite parameter F for A and using the same values, questioned earlier for 5-hexenyl bromide, now for cyclopropyl... 

Bickclhaiipt and co-workers have provided extensive data sets lor Grignard reactions ol 5-hexenyl bromide in DHF. at - 40 ( carried out in NMR tubes at high concentrations. [RBrlu — 2.1 M [4X. They deiermined the yields of RMgBr. QMgBi. RR. RQ. and Qt (Q... 

Table 7.1 contains derived rale parameters for the preceding analyses of data for the Grignard reaction of 5-hexenyl bromide in several solvents [83. Two shortcomings should be noted. First, the v iseosities used in the calculations are for the pure solvents. They are not necessarily identical with or proportional to the viscosities of the product mixtures in which the major parts of the reactions take place. Second, for DBF and DPF.. the rela-tivciv poor hi introduces considerable uncertainty into 1 i values. Those of A are better determined hy the data. 

By adjusting l and A. D-model calculations can reproduce the data 66. l or the unstirred reaction ol AdBr in Did. reasonable agreement is obtained with V 5.06 and A = 2.55 (compare V = 2.50 x I01 and A — 82.6 for 5-hexenyl bromide). The relatively small values for AdBr could rellect the high viscosity of the AdAd deposit. 

Treating Mg as if it were a molecule in solution. we estimate these probabilities. Let D, the relative diffusion coefficient of Mg and R. he 5 x 10" A s 1. Let k for Mg be. 1 x 10 As. the value derived from experimental data lor reactions of hulk Mg/ with 5-hexenyl bromide (Section 7.2.8). For spherical reactant molecules, the probability d of geminate reaction is related to these parameters by liquation 7.A.8 (Appendix). 

Reactions of l-iodo-l-methyl-2,2-diphenyl-eyclopropane and 5-hexenyl bromide are among those exhibiting characteristic radical isomerisations. l-lodo-l-methyl-2,2-diphenylcyclopropanc gives nearly racemic RMgl, which does not itself racemize (Equation 7.4) (20). Similarly, along with RMgBr, 5-hexenyl bromide gives the cyclized... 

Indeed, X appears to dominate. Suppose that Tr were 3 x 10 s (value for 5-hexenyl bromide in DF.E) for the pathway R (D-model) contribution in reactions of aryl halides and that the only product of pathway X were RMgX. Then the implied extent of pathway X for the reactions of 2-(.3-butenyl)phenyl halides would be >97f 11 12],... 

For reactions of. 5-hexenyl bromide in THF at 22 C with sonication, for example, these predictions have been confirmed. The molarralio Mg/RBi was varied from 1(1 1 to 1 1. with IRBijo constant, without affecting the product distribution I6. ... 

Fig. 7.21. Obscivcd vs calculated yields and yield-derived parameters tor Grignard reactions of 5-hexenyl bromide (initially 2 1 M) in DBE (solid circles) and DPE (open circles) al -40 C [8.3].

SS) =4%, and (CpS) = 2%. When there is no radical isomerization, the D-model product distribution is determined by two composite parameters, V and A (equations (7.40) and (7.41)). The observed product distribution is approximated by a D-model calculation (Figure 7.25) with V = 2.58 and A = 2.71, corresponding to 8 = 0.0070 A, ks = 2.0 X 10 s, and all other parameters the same as those used in D-model calculations for 5-hexenyl bromide (102], Taking disproportionation into account will bring the observed and calculated values of (CpH). (CpCp), (SS), and (CpS) even closer. Thus, with plausible parameter values, the D model cati describe this product distribution. 

In fact, such calculations fail for phenyl and 2-(3-butenyl)phenyl halides [112. Cyclopropyl and 5-hexenyl halides are not good models for the corresponding aryl halides. Grignard reactions of the aryl halides give far fewer by-products than predicted from their aliphatic tnodels in D-model calculations (equation 2.6.7) using Tk = 3 x 10 s, a value that describes reactions in DEE and THF of cyclopropyl and 5-hexenyl bromides (Sections 7.2.8-7.2.9). 

An excellent measure of radical involvement for certain types of studies relies upon the rearrangement rates of organic radicals that might be formed under the reaction conditions. Thus, if a free 5-hexenyl radical is formed in the oxidative addition of 5-hexenyl bromide, the product will contain a cyclopentylmethyl group. The rearrangement of the radical occurs at a rate of 10 s so that if the radical has a lifetime of more than about lO " s, it will rearrange. This technique for the demonstration of radical paths has been used in a number of cases... 

---