Octanoate

All crystal structures have been deposited into the Protein Data Bank (PDB, www.pdb.org) with the following PDB codes: N5-laninamivir - 3TI8, p09N1-zanamivir - 3TI5, p09N1-laninamivir - 3TI3, p09N1-laninamivir octanoate - 3TI4, p09N1-oseltamivir - 3TI6, p57N2-zanamivir - 3TIC, p57N2-laninamivir - 3TIA, and p57N2-laninamivir octanoate - 3TIB.

The dust samples were sieved using a 100 mesh stainless steel sieve to remove the coarse (>150 μm) fraction of dust. To cover the broad suite of organic analytes of interest, we used three different extraction methods, each combined with GC/MS in the multiple ion detection mode, for detection and quantification. Sample batches consisted of 15 field samples plus three quality control (QC) samples: a lab duplicate, a solvent method blank, and a duplicate spiked sample (spiked with either 50 ng or 250 ng) or solvent spiked sample (100 ng/analyte). For analysis of metals, conventional microwave-assisted acid digestion was combined with ICP/MS. Sample batches consisted of 35 to 37 field samples, one SRM (NIST 2583), a method blank, a duplicate, and a duplicate spiked sample (spiked to give 2.5 μg/L in the extract for all analytes except zinc, which was spiked to give 25 μg/L); in addition, three digested samples were analyzed in duplicate and then spiked and reanalyzed.
Resource constraints precluded us from performing all analyses on every sample, so we selected 15 sample pairs for both neutral and acid extractions, 25 for neutral extractions only, and 5 for acid extractions only (for a total of 40 sample pairs extracted for neutrals and 20 for acids). We chose a larger sample size for the neutral extractions because there were 10-fold more neutral-extractable analytes (n = 64) than acid-extractable analytes (n = 6). We extracted metals from all samples with sufficient dust remaining after the neutral and/or acid extractions (n = 19).
For the hexane:acetone (H:A) extraction method, a 0.5 g aliquot of fine dust was spiked with 250 ng each of 14 surrogate recovery standards (SRSs), most of these being C13 labelled analogs of one of the analytes. The SRSs represent all major compound classes of the analytes and are added as a sample-by-sample check on method performance for the compound classes being analyzed. The dust was extracted with 12 mL of 1:1 H:A in an ultrasonic bath (Branson 5210) for 10 minutes. After centrifugation (Forma Scientific), a 10 mL aliquot was removed and concentrated to 1 mL. The extract was solvent exchanged into hexane and applied to a silica SPE cartridge (6 mL, 1000 mg loading; Baker) that had been conditioned in sequence with 20% acetone in ethyl acetate, dichloromethane, 15% diethyl ether in hexane, and hexane. The analytes were then eluted with these solvents in reverse order. The first three fractions were collected as one and concentrated to 1 mL; the final eluent was collected separately and concentrated to 1 mL. Two internal standards (IS) were added: p,p = -dibromophenyl (for pesticide and PCB analyses) and d₁₂-benzo(e)pyrene (for PAH analyses).
The dichloromethane (DCM) method was similar to the H:A method, except the SRS mixture had two components rather than 14, the extraction was performed with DCM rather than H:A, the concentrated extract was solvent exchanged into ethyl acetate, there was no SPE cleanup step, and there was only one IS (p,p = -dibromophenyl).
For the herbicide acid (Acid) method, a 0.5 g aliquot of dust was weighed into a 60 mL centrifuge tube and 250 ng of one SRS was added. A 25 mL aliquot of the extraction solvent, 70:30 acetonitrile:phosphate buffer (0.1 M sodium acid phosphate) at pH = 3, was added to the dust. The dust was sonicated and centrifuged. A 20 mL aliquot of the extract was transferred to a separatory funnel containing deionized water, and the pH was adjusted to 1 with concentrated HCl. The aqueous layer was applied to a C18 SPE cartridge (6 mL, 500 mg; Baker) that had been conditioned just prior to use with methanol, deionized water, and 1:10 acetonitrile:0.025 M phosphoric acid. The extract was applied to the SPE cartridge and the cartridge was dried for two hours. The cartridge was eluted with 1:1 hexane:diethyl ether. The eluent was concentrated to near dryness under a stream of dry N₂, resuspended in 5% methanol in methyl-t-butyl ether, and then methylated using ethereal diazomethane generated in-situ from Diazald, carbitol, and 37% aqueous potassium hydroxide. After the solutions were allowed to stand for 30 minutes, they were purged of excess diazomethane and the IS (p,p = -dibromobiphenyl) was added.
The H:A sample extracts were analyzed using an RTx-5 MS column (30 M, 0.25 mm id, 0.25 μm film) with a GC oven temperature programmed from 130–220°C @2°/min and then 220–330°C @10°/min. The DCM and Acid sample extracts were analyzed using a DB-1701 column (30 M, 0.25 mm id, 0.15 μm film) with the GC oven temperature programmed 130–220°C @2°C/min and then 220–280°C @10°/min for the DCM method, and programmed 140–280°C @20°/min for the Acid method. Typically two ions were monitored for each analyte, although for selected compounds (e.g., malathion) three ions were monitored for identification. An 8-point calibration curve, spanning the range of 2–750 ng/mL for analytes and 10–300 ng/mL for SRSs, plus an instrument blank, was analyzed concurrently with each sample set for each analytical method. Linear regression analysis was used to establish the calibration curve for each analyte.
For analysis of metals, 0.2 g of dust was digested in 10 mL of ultra-pure nitric acid in a Teflon microwave digestion vessel. Once capped, the vessels were heated slowly to 150°C and allowed to digest for three hours. After cooling, the digestate was transferred to a 50 mL conical tube and diluted to 50 mL with deionized water. Further dilutions were an additional 10X and 100X from the 50 ml volume. Solutions were analyzed in reverse order of dilution (e.g., 100X first) to obtain data without matrix effects and matched to the calibration range. The ICP/MS was calibrated daily using an 8 to 11 point calibration curve ranging in concentrations from 0.1 to 2,500 μg/mL. Internal standards, added in-line to samples and standards, were used for quantification and to correct for variations in instrument response. Quantification was performed using a linear regression analysis of the calibration curve data.
The analytes for which each extraction method was used are given in Table 1. Several additional analytes were tested for but not detected in these samples: beta-endosulfan, heptachlor, azinphos-methyl, dimethoate, methidathion, metolachlor, ethafluralin, cyanazine, butylate, and pebulate, with detection limits 10 to 100 ng/g in the H:A method; and prometryn, phorate, tribufos, and bromoxynil octanoate, with detection limits 10 to 50 ng/g in the DCM method. These analytes are not listed on Table 1 and are not considered further in this paper. Spiking of four randomly selected dust samples with 250 ng (500 ng/g) of each organic analyte showed that among the 64 organic analytes detected in at least one household, mean recoveries generally ranged from 81% to 125%. Analytes with lower recovery means were dicamba (33 ± 20%), chlorothalonil (48 ± 55%), acephate (69 ± 21%), mecoprop (MCPP) (69 ± 35%), and piperonyl butoxide (73 ± 11%); and analytes with higher recovery means were allethrin 1 (127 ± 21%), allethrin 2 (136 ± 31%), carbaryl (130 ± 36%), cyfluthrin 4 (131 ± 27%), malathion (134 ± 26%), iprodione (139 ± 27%), phosmet (140 ± 21%), cypermethrin 2 (141 ± 19%), cypermethrin 4 (141 ± 24%), methoxychlor (145 ± 42%), and dichlorodiphenyltrichloroethane (p,p'-DDT) (163 ± 26%). Three additional analytes had mean recovery standard deviations exceeding 40% (propargite: 111 ± 51%, deltamethrin: 125 ± 52%, and dicofol: 103 ± 86%). The SRS recoveries averaged from 85 to127% in these QC samples, as well as in field samples, with the exception of ¹³C₁₂-p,p'-DDT (139 ± 13%), F₄-terephthalonitrile (53 ± 14%), and d₄-cotinine (65 ± 26%). Laboratory spikes with metals at 60 μg/g showed mean recoveries ranging from 86 to 100%, with mean standard deviations of 3 to 7%. We did not adjust reported analyte levels in dust for spike recoveries or SRS recoveries in this analysis. In addition to the spiked dust samples, up to seven samples were analyzed in duplicate. The average relative percent difference for duplicates was typically 10 to 30% for organics and 1 to 10% for the metals, indicating very good agreement between pairs of samples.

The quantitative analyses were run in the same GC instrument as the qualitative ones, configured with a Flame Ionization Detector (FID) and equipped with an Agilent Technologies 7683 series autoinjector (Little Falls, DE, USA).
The analytical conditions were the same described for the qualitative analyses, but with a different thermal program. In fact, with DB-5ms column, the initial temperature of 50 °C was kept for 1 min, followed by a thermal gradient of 3 °C/min until 180 °C, then a second thermal gradient of 15 °C/min until 250 °C. The final temperature was maintained for 15 min. For what concerns the analysis on HP-INNOWax, the same GC method as DB-5ms was applied, except for the final temperature that only reached 230 °C. The FID was alimented with a mixture of hydrogen and air, at the flow of 30 mL/min and 300 mL/min respectively. The detector was set at the temperature of 250 °C. In order to quantify the analytes, a relative response factor (RRF) was calculated for each component, according to the respective combustion enthalpy [37 (link),38 (link)]. In this respect, A. Chaintreau and colleagues demonstrated that the RRF of an organic compound, analyzed by FID, only depends, with good approximation, on its molecular formula and number of aromatic rings. According to this principle, they described a mathematical formula [38 (link)], that permits to estimate the RRF toward a quantification standard (usually methyl octanoate). In our case, a modified method was actually applied, since isopropyl caproate was used instead of methyl octanoate and two calibration curves (one for each column) have been used instead of a single point internal standard. The isopropyl caproate was prepared by synthesis in one of the authors’ laboratory (G.G.) and its purity was calculated by GC as 97%. For calibration curves construction, six calibration standard dilutions were prepared, dissolving 0.6, 1.8, 4.3, 8.3, 16.8, and 34.3 mg of isopropyl caproate in 10 mL of cyclohexane respectively. As usual, an amount of 7.0 mg of n-nonane was used as internal standard inside each dilution. Both calibration curves generated a correlation coefficient of 0.995.

Synthesis of piceid octanoate was previously reported by Larrosa et al. [46 (link)]. Briefly, PIC-OCT was prepared via enzymatic acylation of piceid using Novozym 435, lipase immobilized on macroporous acrylic resin. The reaction was carried out in tert-butyl alcohol (10 mL for 1 g of piceid) by adding vinyl octanoate (15 mL, 3 eq). The mixture was stirred in an orbital shaker at 60 °C for 16 h. The enzyme was decanted and separated. The solvent was evaporated and the product was purified via flash column chromatography (CH2Cl2/MeOH from 95:5 to 85:15) to obtain the corresponding PIC-OCT as a white solid (yield 83%, purity 98.5%).

The intermediate 8‐(2‐hydroxybenzamido)octanoic acid, weighing 5 g, was placed into a small amount of isopropanol and stirred, and then a sodium hydroxide solution of 0.7 g was added dropwise. Subsequently, the remaining isopropanol (totaling 8 V) was also added. The mixture was slowly cooled with stirring to induce crystallization, and then filtered under suction. The resulting filter cake was rinsed with isopropanol and subsequently dried at 60°C for 2–4 h, yielding a 90.8% yield.