A new look at the β-pinene–ozone reaction using the atmospheric pressure reactor

2.1 The ion trap mass spectrometer

The reactor used in this work contained a flowing reacting atmospheric pressure gas mixture that was admitted to an ion trap mass spectrometer through a flow restrictor for analysis. The custom reactor was coupled to a Varian Saturn 2000 ion trap mass spectrometer via the 35.6 mm outer diameter direct insertion probe interface. The reactor interface was inserted into the vacuum adaptor installed in the GC transfer line inlet port.

The experiments reported here were conducted using the methane chemical ionization. The mass spectra (m/z 30–250) were collected using the ion trap control software in the GC mode. The WSearch reduction program was used to convert the measured mass spectra to the ASCII form for further analysis.

2.2 Static reactor setup

The reactor setup was described in a previously published work (Wedian and Atkinson 2010). Briefly, all reactions were carried out in a 12 L Teflon bag. One side of the bag was connected to mass spectrometer via a custom made interface while the other side was connected to a 3-way valve. In one valve position, the valve admitted a He flow sufficient to supply the mass spectrometer interface and to maintain the bag shape. He served as the gas medium for reactions as well as the carrier gas that delivered the reactants and products into the mass spectrometer. In the second valve position, the bag was connected to a diaphragm pump to evacuate the bag's contents between successive experiments. The vapor phase of β-pinene and ozone, was injected with a syringe through a hole in the side of the bag.

2.3 Static reactor operation

The reactions were performed in this experiment by filling with He then taking the background of the bag, after few minutes, the β-pinene (initial calculated concentration of 6 ppmv) was injected into the bag using a 25 ml syringe until the signal reached the steady state, and then finally, excess amount of purified ozone in nitrogen was introduced into the bag using a 50 mL syringe to complete the reaction. It is expected that the primary and secondary products of ozonation are formed in few seconds. Moreover, it is expected all reactants and products were exposed to similar conditions of wall reactions, dilution, and transport of all products (Wedian and Atkinson 2010). Between experiments, the reactor bag was flushed and filled several times with He to remove the contents from the previous experiment. The ozone concentration in the reactor was determined in separate experiments by exhausting the reactor through a Dasibi ozone instrument.

2.4 Chemicals

Chemicals and gases used in these experiments were: β-pinene (Fluka, 99.5%), acetone (Aldrich, 99%), nopinone (Aldrich, 98%), formaldehyde (Aldrich, 36.5–38.0%), He (commercial grade 99.998, Airgas Inc.), Methane (UHP 99.9995, Airgas Inc.), N₂ (UHP 99.9995, Airgas Inc.), and O₂ (UHP 99.9995, Airgas Inc.).

3.1 β-pinene–ozone reaction

Fig. 1ِA shows the mass spectrum of the reaction mixture at long reaction time, after subtracting the pre-reaction background. The mass spectrum is dominated by ions of unreacted β-pinene and nopinone which produce strong peaks at m/z 57, 59, 69, 83, 97, and 107 and a few weaker fragments. Fig. 1B shows the mass spectrum of β-pinene which is dominated by major fragments (m/z = 67, 81, 93, 107, and 121), Fig. 1C is chemical ionization mass spectrum of pure Nopinone which is dominated by the fragments (m/z = 55, 69, 81, 93, 97, 107, 135 and 137). The CI mass spectrum of pure pinene and nopinone was measured in separate experiments. Multiplying the reference β-pinene and nopinone CI spectra by an appropriate weighting factor, and then subtracting them from the β-pinene–ozone reaction mixture (Fig. 1A) allowed the contribution of nopinone and un-reacted β-pinene to be removed and Fig. 1D shows the resulting mass spectrum. The other products of the β-pinene–ozone reaction are now more easily recognized (m/z 31, 45, 47, 57, 59, 69, 83, and 97).

Close

Figure 1 (A) The mass spectrum of the β-pinene–ozone reaction mixture after subtraction of the ozone background. The spectrum is dominated by β-pinene and nopinone. (B) The mass spectrum of β-pinene. (C) The mass spectrum of nopinone. (D) The net mass spectrum of the β-pinene–ozone mixture after subtracting contributions of ozone, pure β-pinene, and pure nopinone.

Export to PPT

The logistic growth function (1 − exp (−a₁t))a₂, where a₁ is the rise coefficient adjustable in (s⁻¹), and a₂ (the upper limit of the growth), was used to fit the initial growth of product ions vs. time after injection of excess ozone into the unstirred reactor (Fig. 2). This experiment was carried out by injecting excess ozone into the reactor containing β-pinene in order to monitor the behavior of the unreacted β-pinene and the products of β-pinene–ozone reaction (Fig. 1A). The measured growth coefficients and evolution times of all ions (Fig. 1D) were used to characterize the ions as a primary product by showing a short revolution time or as a secondary product by showing a relatively long revolution time (Table 1).

Close

Figure 2 Function (1 − exp (−a₁t)) a₂, with adjustable coefficients, used to fit the initial growth of product ions of β-pinene /O₃ in the static bag reactor: (A) the growth ion profile of m/z 57 and 93 of nopinone. (B) The growth ion profile of m/z 61. (C) The growth ion profile of m/z 55 and 60. (D) The growth ion profiles of m/z 57, 83, and 96.

Export to PPT

It was concluded from the differences in the fitted rise coefficients in Table 1 that an ion with a unique rise coefficient corresponds to a distinct neutral species in the atmospheric-pressure reactor, moreover, the multiple ions with the same rise coefficient in Fig. 1 also belong to a single species and those fragments are produced by the low-pressure chemical ionization of the single species inside the ion source of the ITMS.

The ions in Fig. 1A with (m/z 57, 81, 93, 109, 135 and 137) are associated with nopinone (scheme 1C) which is one of the two primary products of the β-pinene–ozone reaction. Of the ions in Fig. 1D, the growth of m/z = 31, was satisfactorily fitted with the logistic growth function, with an evolution time of 21 ± 7 s, and is associated with protonated formaldehyde which is the other primary product identified by Winterhalter et al. (2000), Atkinson (1997), and Grosjean et al. (1996). This conclusion was supported by MS/MS data.

The growth of m/z = 47, is tentatively associated with HCOOH, which showed a slow evolution time, diagnostic of a secondary product. The MS/MS experiments confirmed this conclusion. Winterhalter et al. (2000) showed that HCOOH is a secondary product resulting from the decomposition of the C₉-Criegee intermediate and other radicals. The growth of m/z = 59 is associated with presumably protonated acetone, which is probably present as an impurity in the reactor or as a product of the reaction of ozone with an olefinic impurity (Grosjean et al., 1993). The MS/MS experiments supported this conclusion.

The ion at m/z = 61 showed a slow evolution time, characteristic of a secondary product which is a contribution of a protonation product of mass 60 such as CH₃COOH. This conclusion was confirmed by the MS/MS data. Winterhalter et al. (2000) who used ion chromatograms were able to identify traces of CH₃COOH as a result of the decomposition of unstable intermediates such as acetaldehyde monopeacetate.

Fig. 1D showed the presence of overlapped multiple fragments which may result from at least two unidentified secondary products. The mathematical fitting was used to correlate the fragments in clusters, Table 1; the first cluster of ions at m/z 45, 55, 56, 60, 69 and 99 showed a slow evolution time of ∼55 ± 19 s which is a characteristic of a secondary product; and another cluster of ions at m/z 57, 71, 72, 82, 83, 96 and 98 are fragments of the other secondary product of the β-pinene–ozone reaction with an induction delay of ∼48 ± 14 s. Based on Winterhalter et al. (2000) work, these fragments are associated with two detected secondary products: 3-hydroxy-nopinone and the cis-pinic acid. The MS/MS experiments for selected m/z from each cluster showed severe fragmentation and produced no unique ion fragment in the daughter mass spectra making it difficult to correlate them to either possible compound. Up to the time of doing these experiments, standard mass spectra of pure 3-hydroxy-nopinone and the cis-pinic acid were not available in our research group which limited the ability to get a positive correlation of ions to their neutral compound.

The main products of the β-pinene–ozone reaction identified by Winterhalter et al. (2000) are HCHO, nopinone, 3-hydroxy-nopinone, CO₂, CO, HCOOH, the secondary ozonide of β-pinene, and cis-pinic acid. There were two primary products of the β-pinene–ozone reaction identified by Winterhalter et al. (2000), Atkinson (1997), and Grosjean et al. (1996): formaldehyde and nopinone. Table 2 lists all primary and secondary products of the β-pinene–ozone reaction and their molar masses identified in the literature.

3.2 New short-lived intermediates

This reactor was able to detect new ions with unique rise coefficients which could have possibly resulted from contributions of thermalized neutral intermediates:

The ion at m/z = 70 showed a slow evolution time, diagnostic for a secondary product which may associate with a neutral parent ion or a short-lived neutral intermediate yielding m/z = 70 on protonation. The MS/MS spectrum of m/z = 70 dominated with (m/z 70, 52 (base peak), and 30) which did not provide enough information to reach a definite conclusion regarding the nature of this ion. The tentative chemical formula of m/z = 70 is C₄H₅O.

The ion at m/z = 97 is associated with a secondary component with an induction delay of ∼42 ± 19 s; The MS/MS spectrum of m/z = 97 dominated with (m/z 97, 79, 69 (base peak), and 44), these data did not provide more information to support whether the m/z = 97 is an unidentified new secondary product or a contribution of short-lived intermediate. It is believed that m/z = 97 may have a tentative chemical formula of C₅H₄O₂.

3.3 Calibrations and yields of the primary products

The calibrations of the primary products and their yields were performed without a scavenger in a similar procedure that was described in detail in previous work (Wedian and Atkinson 2010). The yields of formaldehyde and nopinone were 0.58 (±0.05), and 0.19 (±0.05), respectively. Table 3 shows a comparison of the yields of this work and other previous β-pinene–ozone studies.

The limitations of this reactor in its present state of development lie mainly with the lack of accurate identification of the parent ion mass of fragments; Moreover, the system does not exclude any possible contribution of thermalized neutral fragments. Those limitations could be overcome when a standard mass spectra library of the ion trap mass spectrometer is available for most detected compounds.