Comparison of Cu+, Ag+, and Au+ Ions as Ionization Agents of Volatile Organic Compounds at Subatmospheric Pressure

Ionization of volatile organic compounds (VOCs) by coinage metal ions (Cu+, Ag+, and Au+) generated by laser desorption and ionization (LDI) of a metal nanolayer in subatmospheric conditions is explored. The study was performed in a commercial subatmospheric dual MALDI/ESI ion source. Five compounds representing different VOC classes were chosen for a detailed study of the metal ionization mechanism: ethanol, acetone, acetic acid, xylene, and cyclohexane. In the gas phase, ion molecular complexes of all three metal ions were formed, typically with two ligand molecules. The successful detection of the metal complexes with VOCs strongly depended on the applied voltages across the ion source, minimizing the in-source fragmentation. The employed orbital trap with ultrahigh resolving power and sub-parts-per-million mass accuracy allowed unambiguous identification of the formed complexes based on their molecular formulas. The detection limits of the studied compounds in the gas were in the range 0.1–1.4 nmol/L. Compared to Cu+ and Ag+ ions, Au+ ions exhibited the highest reactivity, often complicating spectra by side products of reactions. On the other hand, they also allowed detecting saturated hydrocarbons, which did not produce any signals with Ag+ and Cu+.


Sputtering of metal nanolayers in the laboratory-built magnetron chamber
Ag and Cu films were prepared on microscopic slides using a laboratory-built magnetron sputtering device in a vacuum chamber provided by Activair, Czech Republic.The vacuum pumping system consisted of rotary and turbomolecular pumps, giving <10 -4 Torr base pressure.The 3" TORUS magnetron gun (Kurt J. Lesker, USA) with Ag or Cu target (99.99%pure, Camex s.r.o., Czech Republic) operated at 1.1 mTorr Pa pressure in Ar (99.996% pure) at 50 W or 150 W DC power, respectively, supplied by PD500X high-voltage power supply (Kurt J. Lesker, USA).The deposition rates were measured with the mechanical profilometer DektakXT using films deposited for 6, 12, and 18 min (Ag) and 3.5, 7, and 10.5 min (Cu), see Figure S1.A scratch was made on each film, and the height profile was measured three times.Based on the data, linear curves were obtained, and deposition rates of 10.5 and 4.6 nm/min were calculated for Ag and Cu, respectively.

Optimization of metal layer thickness
To select the optimal thickness of metal layers, i.e., the layers which yield the maximal amount of M + ions during LDI, several layers with calculated thicknesses of 1, 4, 6, 8, 10, and 20 nm (estimated by the deposition time and known deposition rate) were investigated.The average signal of M + was recorded in a 30 s linear laser scan over the substrate.Based on the data shown in Figure S2, the optimal layer thickness of 8, 10, and 6 nm was selected for Cu, Ag, and Au layers, respectively.Xylene:  S-10

Calibration curves for mixtures of acetone and ethanol
Determination of LODs was carried out with a mixture of two VOCs to address the effect of competition for M + ions.One VOC had a constant concentration, and the concentration of the other VOC was gradually increased.In the first case, solutions were introduced with a constant acetone concentration of 6.76 mmol/L and increasing ethanol concentrations of 17.1 μmol/L -30.8 mmol/L.For the second experiment, the ethanol concentration was 6.85 mmol/L, and the acetone concentration increased between 13.5 μmol/L and 28.S1; these values are comparable to those for quantification from solutions of individual compounds.Of course, there is competition between VOCs for metal ions; if any VOCs were in excess, the quantitation of the others may not be possible.
For example, acetone has a higher affinity for gold ions compared to ethanol, and the LODs for acetone are lower compared to ethanol.When increasing acetone concentration at a constant ethanol concentration (Figure S11), the signal intensity of acetone-containing ions grows rapidly, and the signal intensity of ethanol-containing ions drops sharply.In the opposite case, where ethanol concentration was increasing and acetone concentration was held constant, the intensity of acetone-containing ions decreased slowly (Figure S10).

Figure S1 :
Figure S1: Deposition rate of A) Ag and B) Cu layer in the laboratory-built magnetron chamber.

Figure S3 :
Figure S3: SubAP LDI mass spectra of A) xylene and B) cyclohexane ionized with Au + .

Figure S4 :
Figure S4: SubAP LDI mass spectra of A, C) 10 nm silver nanolayer and B, D) 8 nm copper nanolayer with ambient air A, B) and cyclohexane C, D) entering the ion source through ESI capillary.

Figure S5 :Figure S6 :Figure S7 :
Figure S5: Dependence of intensities of ion-molecular complexes of A) ethanol, B) acetic acid, C) acetone, and D) xylene with Au on voltage V 1 .

Figure S10 :Figure S12 :
Figure S10: Calibration curves for mixture of acetone and ethanol showing dependence of A) [Au+C 2 H 6 O+H 2 O] + , B) [Au+C 3 H 6 O+H 2 O] + , and C) [Au+C 2 H 6 O+H 2 O] + ion intensities on ethanol concentration.The concentration of acetone was kept constant; the amount corresponding to a single spectrum acquisition was 563 pmol.The amount of ethanol corresponding to a single spectrum acquisition ranged from 1.4 pmol to 2.6 nmol.
4 mmol/L.Figures S10 and S11 show data for ions containing only one type of VOCs, [Au+C 2 H 6 O+H 2 O] + , [Au+C 3 H 6 O+H 2 O] + , and the ion containing both VOCs, [Au+C 3 H 6 O+C 2 H 6 O] + .The estimated LODs are shown in Table

Table S1 :
Limits of detection (LOD) calculated for ethanol or acetone in the mixture for ionization by Au + .