We report here straightforward procedures for the synthesis of isomeric isoprene β-epoxydiols (β-IEPOX) in high yield from inexpensive, readily available starting compounds. The syntheses do not require the protection/deprotection steps or time-consuming purification of intermediates as used in past studies (Cole-Filipiak et al., 2010; Zhang et al., 2012; Bates et al., 2014, 2016; Chase et al., 2015) and can readily be scaled up to yield the target IEPOX isomers in gram quantities.

Yearly global emissions of isoprene (2-methyl-1,3-butadiene, C5H8), primarily from deciduous vegetation, are estimated to be between 500 and 600 Tg and constitute the largest source of nonmethane atmospheric hydrocarbons (Kanakidou et al., 2005; Guenther et al., 2006; Hallquist et al., 2009; St. Clair et al., 2016). In the gas phase under low-nitric-oxide (NO) conditions, the atmospheric hydroxyl radical (OH) adds rapidly to isoprene almost exclusively at C1 and C4, followed by addition of O2 to yield β- or δ-hydroperoxyl radicals (Hallquist et al., 2009). The major sink (>90%) for the peroxyl radicals is reaction with the hydroperoxyl radical (HO2) to give closed-shell isoprene hydroxyhydroperoxides (ISOPOOHs). ISOPOOHs then undergo sequential addition with OH and O2, followed by the elimination of OH to yield a ∼2:1 mixture of cis- and trans-(2-methyloxirane-2,3-diyl)dimethanol (cis- and trans-β-IEPOX). The IEPOX isomers account for about 80 % of the closed-shell hydroxyperoxy products (St. Clair et al., 2016; Wennberg et al., 2018; Paulot et al., 2009) and are rapidly taken up onto acidic aerosols (Lin et al., 2012; Gaston et al., 2014; Riedel et al., 2015). IEPOX isomers thus make a significant mass contribution to secondary organic aerosol (SOA) (Surratt et al., 2010; Riva et al., 2019) and the resulting atmospheric fine particulate matter (PM2.5) (Lin et al., 2013; Budisulistiorini et al., 2015, 2016; Rattanavaraha et al., 2016). PM2.5 is known to be a major factor in climate forcing (Hallquist et al., 2009) and adversely affects respiratory and cardiovascular systems of exposed populations (Pope and Dockery, 2006; Pye et al., 2021). Advancing the understanding of the impacts of PM2.5 requires the ability to predict PM2.5 composition and distribution, both in regional- and global-scale atmospheric chemistry and climate models (Pye et al., 2013; McNeill et al., 2015; Marais et al., 2016; Jo et al., 2019), which in turn depends crucially on the accuracy of identification and quantitation of uptake product formation. As major precursors of PM2.5, cis- and trans-β-IEPOX have been the focus of considerable effort to elucidate mechanisms underlying PM2.5 formation and aging (Lin et al., 2012, 2014; Gaston et al., 2014; Nguyen et al., 2014; Zhang et al., 2018; Riva et al., 2019; Armstrong et al., 2022; Cooke et al., 2022). Underpinning such efforts is the availability of authentic cis- and trans-β-IEPOX in high purity and in quantity.

The reactions described below should be performed under a fume hood.

The esterification and diisobutylaluminum hydride (DIBAL-H) reduction of mesaconic and citraconic acids generally followed the procedure reported by Klimovica et al. (2011). Epoxidation of the E- and Z-2-methylbut-2-ene-1,4-diols followed the procedure reported by Zhang et al. (2012). 1H nuclear magnetic resonance (NMR) spectra of all isolated products, as well as 13C NMR for target cis- and trans-β-IEPOX isomers, are provided in the Supplement. Mass spectra of cis- and trans-β-IEPOX, also provided in the Supplement, were obtained using the Agilent 1200 Series HPLC (high-performance liquid chromatography) system equipped with an ESI (electrospray ionization) source interfaced to an Agilent 6250 Series Accurate-Mass Q-TOF MS (quadrupole time-of-flight mass spectrometer) operated in negative ion (–) mode using instrumental conditions described elsewhere (Cui et al., 2018).

Several synthetic routes to the β-IEPOX isomers have been published to date. Procedures for the synthesis of trans-β-IEPOX followed the three strategies given in Figs. 1–3.

Figure 1, the first published route to trans-β-IEPOX (Cole-Filipiak et al., 2010), yielded a mixture of cis and trans products in four steps. The procedure is lengthy, and consecutive vacuum distillations for isolation of 1,4-dibromoisoprene from the 1,4-diol were required. The mixture was not separated, and the combined overall yield was 11 %.

The approach in Fig. 2 (Zhang et al., 2012) has been used in most syntheses reported subsequent to publication in 2012. Figure 2 targets synthesis of the trans isomer starting with prenol (3-methyl-2-buten-1-ol). SeO2 oxidation of the trisubstituted olefin yielded E-2-methylbut-2-ene-1,4-diol. Deprotection of the diol, followed by epoxidation with m-CPBA gave the target trans-β-IEPOX in an overall yield of 43 %. The expected trans geometry of the ultimate IEPOX isomer (Trachtenberg, et al., 1970; Sharpless and Lauer, 1972) was confirmed by the absence of a nuclear Overhauser effect correlation between the methyl group and the oxirane proton in the 1D NOESY (nuclear Overhauser effect spectroscopy) spectrum. An overall yield of ∼11% can be calculated for synthesis by Fig. 2 in the only other report citing yields (Bates et al., 2014). The SeO2 oxidation–NaBH4 reduction sequence to generate 2-methylbut-2-ene-1,4-diol appears largely responsible for the discrepancy in yields. Isolation of the diol from the NaBH4 reduction step yields a mixture from which the isolation of a product is challenging and is most likely the source of the difference.

More recently, Fig. 3, a route to trans-β-IEPOX-d2, has been reported that could also serve as a route to the protio compound by substituting LAH4 for LAD4 as the reducing agent (Chase et al., 2015).

This route also involves a problematic metal hydride reduction step, and an overall yield of 31 % was reported. Figures 1–3 have steps that are difficult to accomplish in common, such as vacuum distillation, or they require carefully controlled conditions for the protection/deprotection of labile substituents, with the best reported yield being 43 % for Fig. 2 (Zhang et al., 2012).

Here we report a procedure for the synthesis of pure racemic trans-β-IEPOX that is efficient and simple and provides the target IEPOX with an overall yield of 62 %. No protection/deprotection steps, which add steps and can decrease yields, are involved, and no specialized glassware or instrumentation is required. The strategy for synthesis follows Fig. 4, which is based on inexpensive, readily available mesaconic acid as the starting material.

Mesaconic acid can be esterified to the dimethyl ester in high yield by refluxing in methanol containing 2 % concentrated sulfuric acid. The diester is reduced by diisobutylaluminum hydride (DIBAL-H) in methylene chloride to E-2-methylbut-2-en-1,4-diol, which is epoxidized by m-CPBA in acetonitrile. Key to the procedure is the efficient extraction of E-2-methylbut-2-ene-1,4-diol from the DIBAL-H reduction reaction with ethyl acetate, which allows for recovery of the diol in 70 % yield. The route in Fig. 4 will make trans-β-IEPOX readily available to laboratories without sophisticated synthesis capabilities. The procedure is particularly attractive because it can readily be scaled up to produce trans-β-IEPOX in gram quantities.

The isolation of E-2-methylbut-2-ene-1,4-diol from the DIBAL-H reduction reaction in high yield by extraction with ethyl acetate led us to revisit published syntheses of cis-β-IEPOX in which metal hydrides were used as reduction reagents (Bates et al., 2014, 2016; Zhang et al., 2012). In Figs. 5 and 6, citraconic anhydride was the starting material, and the reducing agent was DIBAL-H.

The reported overall yield of cis-β-IEPOX from Fig. 5 was 12 % (Bates et al., 2014). Figure 6 streamlined the synthesis through bypassing steps 2 and 3 of Fig. 5 with direct reduction of citraconic anhydride to Z-2-methylbut-2-en-1,4-diol (Bates et al., 2016). Reduction of the anhydride required forcing conditions (five equivalents of DIBAL-H were used) to achieve reduction of citraconic anhydride, and possibly less efficient recovery of the diol resulted in the same overall yield reported for Fig. 5.

3-Methylfuran-2(5H)-one was the starting material for Fig. 7 and was reduced directly to Z-2-methylbut-2-en-1,4-diol by LAH (Zhang et al., 2012).

LAH is a powerful reducing agent leading to some unavoidable over-reduction of the furanone to the saturated diol, and the overall yield was 19 %. We repeated the synthesis of cis-β-IEPOX using either citraconic acid or 3-methylfuran-2(5H)-one as the starting point (Fig. 8). Because reduction of anhydrides to diols generally appears to be more difficult and requires forcing conditions (Bloomfield and Lee, 1967), we selected citraconic acid rather than the anhydride as the starting point for the synthesis. Citraconic acid requires esterification prior to reduction. Although the dimethyl ester is commercially available, the esterification reaction is very straightforward with a nearly quantitative yield, and the savings in cost are substantial.

Under the assumption that sterically hindered cis diester would nevertheless not proceed as readily as reduction of the trans diester, five equivalents of DIBAL-H were used for the reduction of citraconic diester. With a yield of 74.9 %, the reduction of the diester proved to be much more efficient than that of citraconic anhydride, for which the reported yield was 28 % (Bates et al., 2016). The overall yield for this route was 57 %, which represents a significant improvement over the yields reported by the routes in Figs. 5 and 6.

3-Methylfuran-2(5H)-one is a more expensive starting compound than citraconic acid or citraconic anhydride, but the procedure is streamlined to two steps, and the overall yield, at 69 %, is much higher than for any of the published routes. Cost of reagents would largely dictate the choice of citraconic acid or 3-methylfuran-2(5H)-one starting material. We note that trans- and cis-β-IEPOX were isolated directly following lyophilization. The purity of trans- and cis-β-IEPOX was >98% by NMR (Figs. S3–S4 and S8–S9, respectively).