Supporting Information

Hydrocarbon Analysis using Discharge-induced Oxidation in Desorption Electrospray Ionization

Chunping Wu1, Kuangnan Qian2, Marcela Nefliu1 and R. Graham Cooks1*

Figure S1. Schematic of typical DESI experiment with discharge oxidation. The distance between the emitter and the inlet capillary was 2 mm or less. The distance between the substrate (a glass slide) and the inlet capillary was less than 0.5 mm. The distance between the substrate and the emitter was 2 mm or less. Hydrocarbon solution to be analyzed was deposited and dried onto the glass microscope slide. Solvent ACN/CHCl3 (1:2) containing 50 ppm betaine aldehyde chloride was sprayed under the influence of a high voltage (5 to 8 kV). Nitrogen was used as the nebulizing gas.

Figure S2. Reactive DESI of (a) cyclopentadecane, (b) squalane, (c) cholesterol and (d) octadecylbenzene. In each case, an aliquot of 1µL of the hydrocarbon solution (200 ppm) dissolved in CHCl3 was spotted on a microscope glass slide.

Figure S3. Reactive DESI of the mixture of 2-naphthol and cyclopentadecanol, acquired on the LTQ-Orbitrap. An aliquot of 1µL of the solution of the mixture (1×10-2 M concentration for each compound) in CHCl3/Acetone (1:1) was spotted on a microscope glass slide. The spot contained 1×10-8 mole 2-naphthol (1.4µg) and 1×10-8 mole cyclopentadecanol (2.3µg). Conditions were optimized to avoid discharge-induce oxidation and allow direct comparison of the reactivity of 2-naphthol and cyclopentadecanol with BA. The [BA]+ peak at m/z 102.09134 was used as the lock-mass for exact mass measurements, so it did not present in the mass spectrum.

Figure S4. Gas chromatogram showing peaks due to cyclopentadecane and its oxidation products.

As shown in the gas chromatogram (Figure S4), besides unreacted cyclopentadecane at peak 1, there were 4 other significant peaks. For peak 2 in the gas chromatogram, a MW of 224 was determined by CI as shown in Figure S5c. The peaks at m/z 253 and 265 were adducts of this compound formed with the reagent methane. The EI spectrum (Figure S5d) confirmed the oxidation product at peak 2 was cyclopentadecanone by comparing the fragmentation pattern to the standard EI spectra on the website of National Institute of Standards and Technology (http://webbook.nist.gov/chemistry/). Similarly, the oxidation product at peak 3 in the gas chromatogram was cyclopentadecanol, determined by CI (Figure S5e) and EI (Figure S5f). Peak 4 and 5 corresponded to multi-oxidized product cyclopentadecane-dione and hydroxy-cyclopentadecanone (Figure S5g to j), but the information about the relative position of the ketone groups in cyclopentadecane-dione and relative position of the hydroxyl group to the ketone group in hydroxy-cyclopentadecanone could not be obtained in GC-MS.

Figure S5. Mass spectra of peaks 1 to 5 in the chromatogram (Figure S4) obtained with CI and EI, as indicated. Methane was used as the reagent gas in CI. Electron ionizing voltage of 70 eV was used in EI.

Figure S6. Gas chromatogram showing the peaks of n-octadecane and its oxidation products.

Figure S7. Mass spectra of peak 1 to 6 in the chromatogram (Figure S6) obtained with CI and EI.

By combing the chromatogram and MS information, the oxidation products of n-octadecane were identified. The retention time is related to the boiling point and also the polarity of the analytes. Normally, alcohols have higher boiling points than the corresponding ketones. The location of the functional group also affects the boiling point. Those with the functional group (hydroxyl and keto) toward the end of the carbon chain has a higher boiling point.

As shown in the gas chromatogram (Figure S6), besides non-reacted n-octadecane at peak 1, there were 5 other significant peaks. In Figure S7b, peak 1 in the chromatogram shows a typical saturated hydrocarbon EI spectrum, corresponding to n-octadecane. CI shows that the MW of oxidation products at peak 2, 3, 5 are 268 Da, and the MW of oxidation products at peak 4 and 6 are 270 Da. For the oxidation product at peak 2, m/z 82 and 96 in Figure S7d indicates it is n-octadecanal (comapred to standard EI spectra from NIST). Peak 3 corresponds to the mixture of n-octadecan-4-one and n-octadecan-5-one, by comparing the characteristic alpha cleavage (m/z 71, 225, 85, 211) and McLafferty rearrangement (m/z 86, 240, 100, 226) peaks with those in the standard EI spectra. DB-1 capillary column used for GC analysis was not able to seperate the peaks of n-octadecan-4-one and n-octadecan-5-one. Peak 4 corresponds to the mixture of n-octadecan-4-ol and n-octadecan-5-ol, since m/z 73 and 87 are the characteristic alpha cleavage peaks of n-octadecan-4-ol and n-octadecan-5-ol. Peak 5 corresponds to n-octadecan-2-one, as indicated by the dominant McLafferty rearrangement peak at m/z 58, and the EI spectrum (Figure S7j) matches well with the standard EI spectrum. Peak 6 corresponds to n-octadecan-2-ol, proved by the characteristic peak at m/z 45 by alpha cleavage, and the EI spectrum (Figure S7l) matches well with the standard EI spectrum. The above assignment is reasonable, because the boiling points of the assigned products match the ordering of elution from the GC column.

GC-MS analysis indicated mono-oxidized species were the major oxidation products of cyclopentadecane and n-octadecane; multiply-oxidized products showed in GC-MS to a much smaller extent than observed in reactive DESI experiments. A possible explaination for this is that the multiply oxidized products are not very stable after collection, and could degrade during heating in the GC analysis.

Figure S8. (a) FI-TOF mass spectrum of a vacuum gas oil saturate II (boiling point between 343 and 540oC) (b) Reactive DESI–Orbitrap mass spectrum of the vacuum gas oil saturates and the expanded mass ranges (c), (d) and (e).

Figure S9. Reactive DESI–Orbitrap mass spectrum of a standard alkanes mixture containing discontinuous carbon numbers (C5H12 to C40H82).

Figure S10. (a) Reactive DESI–Orbitrap mass spectrum of a standard alkanes mixture containing the same amount of each alkane (C21H44 to C30H62) (b) Signal response of different carbon numbers relative to henicosane, derived from the mass spectrum. The signal response of each alkane was calculated as the sum of all the oxidation and dehydrogenation peaks.

Supporting Video:

The video first shows electric discharges between the DESI sprayer tip and the MS inlet, which were positioned 2 mm apart from each other. When the glass slide with a hydrocarbon spot was moved close to the sprayer tip and the MS inlet (~ 1 mm), an electric discharge between the glass slide and the MS inlet was observed. Other conditions used for this video: The spray voltage was set at 7 kV. Spray impact angle was 52o. The nebulizing gas (N2) pressure was 150 psi and the solvent flow rate was 3 µL/min. The spray solvent was acetonitrile (ACN)/chloroform (CHCl3) with 1:2 volume ratio, containing 50 ppm betaine aldehyde chloride.

1