Helium Mixture

Article pubs.acs.org/ac

Characterization of the Axial Jet Separator with a CO2/Helium Mixture: Toward GC-AMS Hyphenation G. Salazar,*,† K. Agrios,†,‡ R. Eichler,‡ and S. Szidat† †

Department of Chemistry and Biochemistry & Oeschger Centre for Climate Change Research, University of Bern, 3012 Bern, Switzerland ‡ Paul Scherrer Institute (PSI), 5232 Villigen-PSI, Switzerland S Supporting Information *

ABSTRACT: Development of interfaces for sample introduction from high pressures is important for real-time online hyphenation of chromatographic and other separation devices with mass spectrometry (MS) or accelerator mass spectrometry (AMS). Momentum separators can reduce unwanted low-density gases and introduce the analyte into the vacuum. In this work, the axial jet separator, a new momentum interface, is characterized by theory and empirical optimization. The mathematical model describes the different axial penetration of the components of a jetgas mixture and explains the empirical results for injections of CO2 in helium into MS and AMS instruments. We show that the performance of the new interface is sensitive to the nozzle size, showing good qualitative agreement with the mathematical model. Smaller nozzle sizes are more preferable due to their higher inflow capacity. The CO2 transmission efficiency of the interface into a MS instrument is ∼14% (CO2/helium separation factor of 2.7). The interface receives and delivers flows of ∼17.5 mL/min and ∼0.9 mL/min, respectively. For the interfaced AMS instrument, the ionization and overall efficiencies are 0.7−3% and 0.1−0.4%, respectively, for CO2 amounts of 4−0.6 μg C, which is only slightly lower compared to conventional systems using intermediate trapping. The ionization efficiency depends on to the carbon mass flow in the injected pulse and is suppressed at high CO2 flows. Relative to a conventional jet separator, the transmission efficiency of the axial jet separator is lower, but its performance is less sensitive to misalignments.

I

Accelerator mass spectrometry (AMS) is highly sensitive to radionuclides, thus GC-AMS and HPLC-AMS are powerful tools to study 14C-labeled molecules, for example, for pharmacokinetics studies of drug microdosing in humans.15 However, the conventional AMS offline method takes days to analyze the eluates of a single chromatogram. Real-time AMS hyphenation is such a challenging goal because most of the instruments use a type of SIMS ion source to create negative ions typically from a solid sample as the target and using Cs+ as the sputtering primary ion.16−18 For high-throughput detection of 14C with AMS, gas-accepting ion sources have been developed, which directly convert the carbon from CO2 into C− when the gas interacts with the Cs+ beam on a metal surface.19−21 However, the carrier gas flow preferably should not exceed a few milliliters per minute to keep the pressure of the ion source lower than 1 × 10−3 Pa. AMS-HPLC and AMSGC are still underdeveloped and only a few laboratories have tried real-time analysis for detecting attomole levels of 14Clabeled molecules.22−26 The layout for AMS hyphenation in the literature has consisted of connecting the chromatograph online

n mass spectrometry (MS), ionized or neutral samples at high pressures are introduced into the instrument by direct insertion or by infusion.1 The sample introduction is limited by the utmost need of keeping sufficient vacuum in the device. Instruments equipped with ion sources like matrix-assisted laser desorption ionization (MALDI) or secondary ionization mass spectrometry (SIMS) normally introduce the sample by temporal interruption of the vacuum, setting interlocks and probes.2,3 On the other hand, sample infusion usually injects or funnels the gas-phase sample (neutral or ionized) through an interface that constricts the inflows so that the vacuum pumps can maintain the vacuum. The continuous-flow nature of sample infusion has eased the hyphenation of MS with multiple analytical techniques enabling the usage of MS into diverse fields of science.4 The need of hyphenation has driven the development of multiple interfaces for the transmission from high pressures.5 The group of momentum separators are gastight interfaces and they have been useful for the hyphenation cases where higher efficiency is needed compared to simple open-split injection. Separators like the jet separator and the effusion separator are mostly used for coupling gas chromatography (GC) with MS.6−10 Furthermore, other variations like the particle beam have been used for interfacing HPLC with MS.11−14 © 2015 American Chemical Society

Received: September 21, 2015 Accepted: December 10, 2015 Published: December 14, 2015 1647

DOI: 10.1021/acs.analchem.5b03586 Anal. Chem. 2016, 88, 1647−1653

Article

Analytical Chemistry

Figure 1. Switching injection system and close-up of the axial jet separator. All the gas lines consist of stainless steel tubing with an outer diameter o.d. = 1/16 in and inner diameter i.d. = 0.5 mm. The 6-way valve is switched pneumatically. The inlet of the AMS is located below the ion source (not shown in the picture). The axial jet separator is marked with an ellipse and the inset shows the separation of CO2 from helium (red and green closed circles, respectively) and the transmission into AMS or MS instruments.

■

MATERIALS AND METHODS Gas Injection System. Figure 1 shows the CO2 injection system using helium as a carrier gas. All the parts are purchased from VICI (Houston). A manometer (Keller AG., Winterthur, Switzerland) quantifies the CO2 in the injection loop. Throughout the paper, the CO2 amount is presented as masses in micrograms of carbon. The manometer also quantifies the pressure of the helium/CO2 mixture during the injection (∼0.1 MPa), enabling us to calculate the CO2 mole fraction. Two mass flow meters (MW-500SCCM and M-0.5SCCM Alicat, Tucson) characterize the flows. The 500 sccm meter was set at point a. This device measures the absolute pressure of the inflow gas at a volume with a cross section area of 56.1 mm2 located at its entrance. The pressure at the exit tubing of the 500 sccm meter is calculated using the Bernoulli principle. This pressure corresponds to the stagnation pressure (P0). The Bernoulli calculation requires knowing the pressure and gas velocities at the entrance and exit of the flow meter. The gas velocities (w) are calculated multiplying the gas flow (f) with the cross section areas (A) at the entrance or exit tubing of the flow meter. The inflow capillary of the axial jet separator (fused silica o.d. = 0.36 mm, dn = 0.1 mm or 0.25 mm) is connected to the exit of the flow meter by using an adaptor union. The 0.5 sccm meter is set at point b to measure the background pressure (Pb) and the flow introduced into the MS or AMS instruments. Helium transmission is defined as the ratio of the measured flows at points b and a. Axial Jet Separator. The inflow capillary is set across a union tee and ends inside of a confinement tubing (i.d. = 0.66 mm, length L = 2.5 cm) that is mounted at the front of the tee. The tee is connected to a 90 L/min scroll pump (Agilent, Santa Clara) with a plastic tubing (o.d. = 1/4 in., i.d. = 4 mm) to create a backward pumping. The confinement tubing is connected to the inlet of the MS or AMS by means of a sample introduction tubing (PEEK, o.d. = 1/16, i.d. = 0.25 mm, L = 1 m).

with a combustion system which is interfaced with the ion source. Other conventional gas interfaces for AMS slowly infuse the CO2, requiring an intermediate gas trapping step.19,27 This method is relatively efficient (6−10%), but it is incompatible for real-time hyphenation. The “axial jet separator” is a new momentum gasintroduction interface developed in our laboratory.28 The rationale of this interface is that the axial speeds (wn) of the components of a gas mixture jet are approximately equal to the carrier gas at the time of exiting a nozzle of area (An); however, the components have different momentum flows (Jn) because their densities (ρn) are different (see eq 1 for CO2).29 Jn ,CO2 = A nρn ,CO2 wn 2

(1)

This approximation is correct, if the mole fraction of the components is small. This paper’s hypothesis is that the axial jet separator separates at some degree a heavy gas component from a lighter carrier and this separation depends on the nozzle size and the inflows. The conventional jet separator also requires components with different densities resulting in different radial expansions for each component. For this reason, the precision of the alignment of the jet nozzle with the sampling orifice is crucial (±20 μm).7 In contrast, the intention of the axial jet separator design is to make it largely independent of the radial expansion and crucial alignments. We report here the separation characteristics of the axial jet separator as an alternative interface for AMS and MS. A mathematical model is presented to qualitatively explain the separation. Also, empirical optimizations of the degree of separation of CO2 from helium and CO2 transmission across the axial jet separator coupled to a MS instrument are presented. The overall efficiency is presented for an AMS instrument. This device shall be applied to hyphenate GC, elemental or aerosol analyzers to AMS; and potentially, it could be used as an interface for MS. 1648


Article


Figure 2. Diagram of the main parameters of the axial flow separator. The inflow comes from the right side of the diagram and through a nozzle (inflow capillary) with inner diameter (dn). The flow direction and magnitude are illustrated with the blue arrows. Nozzle positioning uncertainty is 1 mm. Pb is the background pressure near the exit to the introduction tubing. The parameters at the moment of exiting the nozzle, at stagnation condition, and backward pumping flow are denoted with the subscripts n, 0, and p, respectively.

MS System and Signal Processing. A Residual Gas Analyzer (RGA) based on a single-quadrupole mass spectrometer with an inlet capillary (Cirrus, MKS Andover) measures the gas signals. The CO2+ signal is taken at 44 m/z using a selected ion monitoring mode. For the CO 2 transmission efficiency, the MS signal is calibrated with the setup of Figure 1 without including the axial jet separator to ensure 100% transmission at different flows. Then, CO2 is injected into the axial jet separator and the transmission efficiency is calculated. Carbon signals are not taken when the flow meter was set at the position b because the meter could add dead volume. AMS System and Signal Processing. A MICADAS AMS instrument is used to measure the 14C isotope applying an acceleration voltage of 200 kV.30,31 The ion source conditions are the same that we use for routine direct analysis of CO2 gas, and the details can be found elsewhere.19 The ion source inlet is a stainless steel tubing (o.d. = 1/16 in., i.d. = 0.5 mm, L = 1 m). To study the behavior of the axial jet separator, the 12C− signal is detected by a Faraday cup set after the first magnet, just before the entrance of the high acceleration (low energy side).

h = KD Z

(2)

K is a dimensionless empirical constant that depends on the mixing length and shear layer of the gases,33 Z is the ratio that relates the initial momentum of the jet with the momentum of the opposing or retarding flow, and D is the inner diameter of the confinement tubing. We suggest that Z is best represented by the ratio of the jet-to-backward momentum flows32,34 (Jn/Jp) (eq 3). Momentum flow [kg m/s2] is a force calculated by the product of mass flow (Aρw) and velocity (w), so our Z compares the force that brings the gas from stagnation state into sonic speed versus the retarding force of the backward pumping. Z=

A nρn wn 2 A pρp wp 2

(3)

.

The parameters ρp = ρ0,He and wp depend on the backward laminar flow, and the other parameters are explained in the next section. The speed of the backward pumping (wp) is determined from the relation for equilibrium pressure (Pb) of a volume (V) that is receiving a flow throughput (Q [Pa m3/s]) and at the same time, it is being evacuated by an effective pumping speed (Sp) (eq 4).35 Ap is the concentric crosssectional area outside of the nozzle which can be approximated to πD2/4. The flow meter at position a measures Q and the second meter measures Pb. wp is calculated from the slope of Pb vs Q.

■

RESULTS AND DISCUSSION The results of this paper are grouped in two parts, development of a mathematical model to describe the axial jet separator and the empirical optimization of the device. The mathematical model is based on another model describing the penetration distance of a jet-gas in a counterflow. Conjointly, other theories are added to describe the jet-gas momentum as a function of pressure ratios. For the empirical optimization, the transmission efficiency of the axial jet separator is measured with a MS instrument. The MS results are applied for measuring the CO2 overall and ionization efficiencies when the axial jet separator is used with an AMS ion source. The data are fitted with polynomial functions after being smoothed by a 3-point moving average. Jet-Gas in a Counterflow. The axial jet separator can be treated as a jet in a counterflow, which consists of a system, where a secondary flow is directed in opposite direction to the jet. In the axial flow separator, the backward pumping creates the counterflow (Sp) as Figure 2 illustrates. Because of the counterflow and energy loss, the momentum of the jet components dissipate (Jb = 0) at a certain penetration distance (h). Therefore, the transmission across the axial jet separator becomes higher, when the analyte penetration becomes closer to hmax. Several works about jets in counterflow and crossflow have converged into a similar empirical mathematical expression for describing h (eq 2).32,33

Pb =

Q 1 = Q Sp A pwp

(4)

. Jet-Gas Momentum as a Function of Pressure Ratios. The inflow capillary behaves as a convergent jet nozzle and the fluid entering the nozzle is at stagnation conditions because P0 stays constant. For this type of nozzle, the gas mixture should become a well-expanded sonic jet at the nozzle exit if the ratio Xp = Pb/P0 is lower than 0.5.36 Figure 3a shows with the dashed line that the value of the pressure ratio is lower than 0.5 for almost all the experimental conditions; therefore, the axial jet separator works in the sonic jet regime and ρn,CO2, ρn,He, and wn for eq 3 should be calculated with the isentropic expansion theory,37 as shown in eqs 5 and 6. ρn ,CO2 1649

⎛ Pb ⎞1/ γCO2 = ρ0,CO2 ⎜ ⎟ = ρ0,CO2 X p(1/ γCO2) ⎝ P0 ⎠

(5)


Article


The empirical model (eq 7) predicts a notable difference between axial jet separators with different nozzle sizes. Figure 3b shows that the nozzle with higher inner diameter has higher penetration because a jet from a bigger nozzle is more massive than the jet of a smaller nozzle. It should be noted that the gas penetration is limited by the axial confinement. Therefore, we expect that the signal reaches a maximum value when the penetration of CO2 equals hmax (11 ± 1 mm), which corresponds to the inflow ranges of 6.5 ± 1.5 mL/min and 21 ± 4 mL/min for 0.25 mm and 0.10 mm i.d., respectively, as shown in the gray areas and Table 1. At higher carrier flows, the transmission efficiency should drop due to back pressure buildup and turbulence formed at the entrance of the introduction tubing. Table 1. Summary of the Effect of the Inflow Magnitude and Nozzle Size on the Parameters Optimum Values nozzle 0.10 mm inflow

Figure 3. Empirical and theoretical conditions of the jet for two axial jet separators with different nozzles diameters: (a) measured pressure ratios and (b) calculated penetration for the gas components CO2 and helium.

wn =

γHe 2P0 [1 − X p(γHe− 1/ γHe)] ρ0,He γHe − 1

mathematical model gas penetration CO2 transmission eff. into MS He transmission eff. into MS CO2 overall efficiencyc into AMS

(6)

a Inflow in mL/min. measured for 4 μg C.

In eq 5 (density of the jet-component, e.g., CO2); ρ0,CO2 [kg/ m3] is the density of the component at stagnant conditions, γCO2 is the heat capacity ratio, and the parameters Pb, P0 [Pa] are the background and stagnation pressures of the mixture. In eq 6 (speed of the jet at the nozzle); P0/ρ0 can be replaced by RT0. The parameters R (specific gas constant), T0 (temperature [K]), ρ0 and γ should correspond to the CO2/helium mixture. However, these parameters are approximated to the values corresponding to the pure carrier gas because the CO2 mole fraction is kept lower than 4% for all the experimental conditions. The values for RHe, γCO2, γHe, and ρ0,CO2 and ρ0,He are 2.08 × 103 J/kg/K, 1.3, 1.7, 1.84 kg/m3, and 0.17 kg/m3. Finally, eqs 3, 5, and 6 are substituted in eq 2 to obtain eq 7: hCO2 = Kdn

2ρ0,CO2 P0

γHe

(ρ0,He )2 wp 2

γHe − 1

X p(1/ γCO2){1 − X p(γHe− 1/ γHe)}

b

a

21 ± 4 17.5 17.5 14

nozzle 0.25 mm

value

inflowa

value

hmax

6.5 ± 1.5

hmax

14% 5.1% (0.9)b 0.09%

8.0 8.0

17.5% 7.5% (0.6)b 0.12%

7.0

Transmitted flow (mL/min).

c

Efficiency

Empirical Optimization. Transmission Efficiency of the Axial Flow Separator with MS. The calibration of the MS signal for all the efficiency measurements considers the effect of the carrier flow (Figure S1 of the Supporting Information). The sample transmission efficiencies for two axial jet separators (inflow capillary) are shown in Figure 4. The inflows are set

(7)

where the subscript He represents the carrier gas and the subscript CO2 represents the component (helium or carbon dioxide). The drawbacks of this model are (1) The actual jet speed wn is lower than the theoretical value due to turbulent remixing and friction with the walls. (2) The model is based on empirical observations from other works and it is not an exact solution of the physical laws. (3) The model is expected to fail at nonideal gas conditions; however, this is not expected for most GC gas carriers. Analyte−analyte interactions are negligible due to their low concentration. Slight nonideal conditions can be accounted with the empirical constant K and by calculating γ for the corresponding T0.38 For these reasons, the model is used here for qualitative explanation of the results and not for quantitative purposes. As an advantage over the conventional jet separator, the jet-gas components are confined during the axial separation which means setting up is easy; and the efficiency depends more on the gases nature rather than the radial misalignments.

Figure 4. CO2 transmission efficiency (measured by MS) across two axial jet separators with different nozzle diameters for CO2 injections of 4 μg of C. Closed red circles and closed black triangles are the experimental data, and the lines are interpolations. The optimum values are shown with dashed lines.

with two needle valves, which is not a very precise way to control the flows, resulting in relatively instable inflows and creating the degree of scatter in the data. Therefore, the results are presented as approximations and summarized in Table 1. The most notable difference in Figure 4 is that the CO2 transmission efficiency for the smaller nozzle is optimum at a higher inflow than for the bigger nozzle. The optimum momentum flow for separation is reached at lower inflows for bigger nozzles. This flow difference is consistent with the 1650


Article


transmission efficiency with the ionization efficiency and the ion optics guidance (G = 1.0) (see eq 9).

mathematical model depicted in Figure 3b. If the inflow is further increased then the back pressure diminishes the jet-gas penetration and therefore the CO2 transmission. Equations 2 and 7 reveal that the model requires an empirical scaling constant. A value of K = 2.2 scales the calculated CO2 penetration to match hmax at the respective optimum inflows for both nozzles. The value of K is in the same range of other works39 supporting the hypothesis that the jet-in-counterflow model applies to the axial jet separator. Another observation is that the optimum transmission efficiencies for both nozzles are similar, a fact that agrees with the mathematical model. For comparison, the conventional jet separator has demonstrated transmission efficiencies of 30 to 40%, accommodating inflow rates of 5−30 mL/min.40 This work indicates that the axial jet separator only covers the lower and middle range of inflows of the conventional jet separator. Moreover, the transmission efficiency is lower. Further optimizations will be investigated in future studies. We suppose that to work at higher inflows, the backward pumping flow should be increased. On the contrary, the transmission efficiency of the carrier gas decreases because the transmitted amount slowly increases with the change of inflow (Figure 5). Both nozzles present the same

E(%) =

∫ I dt × M / F C0

× 100 (8)

E(%) = transmission E × ionization E × G × 100

(9)

The inflows for optimum overall CO2 efficiency are different depending on the nozzle size (see Figure 5 and Table 1). This behavior is very similar to the CO2 transmission efficiency shown in Figure 4 using a MS instrument. As predicted from the CO2 penetration model, the CO2 transmission is optimum when it reaches a certain momentum flow. This optimum momentum flow is reached at low inflows for bigger nozzles. For all the data in Figures 5−7, the axial jet separator manages

Figure 6. Overall CO2 efficiency of the AMS ion source (calculated with eq 8) for two axial jet separators with different nozzles diameters. Closed red circles and closed black triangles are the experimental data, and the lines are interpolations. CO2 injections of 4 μg of C.

to keep the ion source pressure lower than 6 × 10−4 Pa fulfilling the purpose of separating most of the carrier gas. Comparing with the extreme conditions when introducing 100% of the carrier gas (e.g., 10 mL/min), a pressure of 0.01 Pa is obtained; and when no gas is introduced, a pressure of 9 × 10−5 Pa is registered. Most of the flow-related conditions for our AMS and MS instruments are similar: ion source vacuum, sample introduction lines, and axial jet separator. Therefore, the axial jet separator transmission with both instruments should be the same (∼16% average of both nozzles). For the AMS coupled with the axial jet separator, the overall and ionization efficiencies (calculated from eq 9) are 0.1−0.4% and 0.7−3%, respectively, for CO2 amounts of 4−0.6 μg of C (Figure 7a). These efficiencies are only slightly lower than for conventional systems using intermediate trapping and slow CO2 infusion (∼60 μL/min).19,27 The reasons of the low efficiency at masses higher than 5 μg can be inferred from the ionization mechanism of SIMS.41 It has been proposed that the formation of C− depends on the available amount of electrons near the surface and the limited residence time of the CO2 in contact with certain types of surfaces.20,42 The lack of oxygen in the initial molecule is important, for example, CH4 produces more C− than CO and CO2. The ionization also depends on the type of anion, with O− being more favored than C−.42 Therefore, the oxygen from CO2 can suppress the C− signal when the concentration of CO2 is high. Fahrni et al.19 demonstrated that high CO2 mass flows affect the ionization efficiency using the same type of instrument and ion source as ours. They directly infused specific CO2/helium

Figure 5. Helium transmission efficiency across the axial jet separator and helium introduced flow into the MS for different inflows. The nozzle dn = 0.25 mm. Opened blue circles and closed red circles are the experimental flow and transmission efficiency data, respectively; and the lines are interpolations. The dashed line indicates the inflow for optimum CO2 transmission.

graphical behavior. Table 1 shows the helium transmissions at the CO2 optimum inflow for both nozzle sizes, which correspond to a removal of 94.9% and 92.5%, respectively. The comparison of the transmission efficiencies of CO2 and helium reveals a separation factor of 2.7 and 2.3 for the small and big nozzles, respectively, while for the conventional jet separator, the literature has shown a separation factor of ∼2.0. For vacuum maintenance, however, it is even more important to know the introduced flow rather than the separation factor. At the range of inflows shown in Figure 5, the axial jet separator delivers flows lower or equal to 0.9 mL/min. CO2 Overall and Ionization Efficiency with AMS. After studying the transmission efficiency with the MS instrument, it is important to study the overall efficiency. For AMS, the overall ion production is of large interest, which in turn improves the precision of the measurement of rare radionuclides. The overall efficiency is calculated in eq 8 from the time integration of the 12C− current signal [C/s]; the Faraday constant [F = 9.65 × 104 C/mol] and the carbon molecular weight [M = 12.0 g/mol] relative to the carbon amount in the injected pulse (C0). Furthermore, the overall efficiency can also be calculated as the product of the axial jet separator 1651


Article


calculate the total overall efficiency is multiplying the overall efficiency described in eq 9 (from injection until the low energy side) by the ion guidance from the low until the high energy side (G2). Therefore, 0.6 μg of C presents a total overall efficiency of 0.4% × G2 = 0.16%, which agrees with the value found with the 14C counts. This result reinforces the decision of taking the value of G = 1.0.

■

OUTLOOK In future, we will try to hyphenate a GC with AMS using the axial jet separator. Furthermore, it is worth to try an HPLCAMS coupling using a nanoelectrospray version with the axial jet separator in a similar way as Kientz et al.11 did with a particle beam and a conventional jet separator. A hot desolvation tube can enclose a nanospray tip, funneling the nanodroplets into the inflow capillary of the axial jet separator, separating the water vapor and air from the heavy analytes.

■

CONCLUSIONS The axial jet separator is characterized by a theoretical semiquantitative model and by empirical optimization. It is easy to set up and is not sensitive to the alignment. The mathematical model is based on the different momentum flows of the gas-phase components. The observed effect of the nozzle size on the separation of CO2 from helium shows good agreement with the model, being more preferable smaller nozzles due to their higher inflow capacity. The axial jet separator shows a CO2 transmission efficiency of ∼16% at inflows of 8.0 and 17.5 mL/min, reaching separation factors of 2.7 and 2.3, depending on the nozzle size. The axial jet separator is able to interface an AMS instrument and keep the high vacuum (∼6 × 10−4 Pa). The overall efficiency of the CO2 is 0.1−0.4% with an ionization efficiency of 0.7−3% corresponding to CO2 amounts of 4−0.6 μg of C. The ionization efficiency is optimum at carbon mass flow of ∼2.8 μg C/min in the injected pulse. Although, the efficiencies seem relatively low, the produced 12C− current signal is acceptable. This work paves the way toward applying an axial jet separator to hyphenate AMS and MS with chromatography.

Figure 7. Effect of the carbon amount in injected pulses on the ionization efficiency of the AMS ion source. Axial jet separator nozzle of 0.10 mm and inflow = 12 mL/min. The injected standard gas is a mixture of CO2 in helium (4.5% fraction mole) with 14C/12C concentration of 134.0 pMC or 1.6 × 10−12 mole per mole. The closed symbols are the experimental data. (a) Ionization efficiency calculated for a transmission of 14%. The lines are interpolations. (b) Example of the isotopes peaks for an injection of 600 ng of C corresponding to 0.080 attomoles of 14C and overall efficiency of 0.4%. The lines join each point to the next one.

mixtures at low flows with a syringe pump to control the CO2 mass flow and to ensure 100% transmission. They observed that the efficiency and 12C− current dropped after certain CO2 mass flow, probably because the abundant oxygen suppressed the signal. We reproduced the experiment of Fahrni et al.19 (Figure S2 of the Supporting Information) in order to characterize the ionization efficiency independently if the gas is slowly infused or injected as short pulses coming from the axial flow separator. Figure S2 shows that the CO2 mass flow entering the ion source should be ∼4.2 μg C/min as estimated from the interpolation of the ionization efficiency (∼0.7%). Consequently, the efficiency could be improved if the CO2 amount is reduced until matching the optimum CO2 mass flow. This hypothesis is proved in Figure 7a showing how the ionization efficiency increases very fast for CO2 injections between 0.6 and 2.5 μg of C (marked with dashed arrows). At 3% ionization efficiency, the optimum CO2 mass flow corresponds to ∼2.8 μg C/min as estimated with Figure S2. Not surprisingly, it has been shown20,43 high ionization efficiencies (1−2.5% and 3.5%) for injected CO2 amounts of 0.2−1.2 μg C and 1.8 μg C, respectively. The typical height of the 12C signal peaks at the optimum inflows are 5−15 μA and 2−6 μA measured at the low and high energy sides of the AMS instrument, respectively; which correspond to an ion guidance efficiency G2 = 0.4. Figure 7b shows the carbon peak signal for a relatively small CO2 injection and the integration of the 14C gives 100 counts, detected at the high energy side of the AMS. Comparing with the amount of injected 14C (0.080 amol), we calculate a total overall efficiency of 0.2%. Another method to

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.analchem.5b03586. Effect of the carrier flow for the MS signal calibration and effect of the CO2mass flow on the carbon signal and ionization efficiency in the AMS ion source (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*Phone: +41-31-6314263. E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We acknowledge our technician Michael Battaglia for his support in the construction of the injection system and maintenance of the ion source of the AMS instrument. 1652


Article


■

(34) Kartaev, E.; Emel’kin, V.; Ktalkherman, M.; Kuz’min, V. I.; Aul’chenko, S.; Vashenko, S. P. Chem. Eng. Sci. 2014, 119 (119), 1−9. (35) Hoffman, D. M.; Singh, B.; Thomas, J. H. Handbook of Vacuum Science and Technology; Academic Press: San Diego, CA, 1998; p 214. (36) Sanna, G.; Tomassetti, G. Introduction to Molecular Beams Gas Dynamics; Imperial College: London, 2005; pp 246−249. (37) Sanna, G.; Tomassetti, G. Introduction to Molecular Beams Gas Dynamics; Imperial College: London, 2005; pp 115−119. (38) Christen, W.; Rademann, K. Phys. Rev. A: At., Mol., Opt. Phys. 2008, 77 (1), 012702. (39) Bernero, S. A Turbulent Jet in Counterflow. PhD. Thesis Ing., Technical University of Berlin, Berlin, Germany, 2000; pp 6−7. (40) Watson, J. T.; Sparkman, O. D. Introduction to Mass Spectrometry: Instrumentation, Applications and Strategies for Data Interpretation, 4th ed.; Watson, J. T., Sparkman, O. D., Eds.; John Wiley: Chichester, 2007; p 598. (41) Van der Heide, P. Secondary Ion Mass Spectrometry: An Introduction to Principles and Practices; van der Heide, P.; John Wiley & Sons, Ltd.: Hoboken, NJ, 2014. (42) (a) Middleton, R. Nucl. Instrum. Methods Phys. Res., Sect. B 1984, 5 (2), 193−199. (b) Middleton, R.; Klein, J.; Fink, D. Nucl. Instrum. Methods Phys. Res., Sect. B 1989, 43, 231−239. (43) Agrios, K.; Salazar, G.; Zhang, Y.-L.; Uglietti, C.; Battaglia, M.; Luginbühl, M.; Ciobanu, V. G.; Vonwiller, M.; Szidat, S. Nucl. Instrum. Methods Phys. Res., Sect. B 2015, 361, 288.

REFERENCES

(1) Hoffmann, E. D.; Stroobant, V. Mass Spectrometry: Principles and Applications, 3rd ed.; J. Wiley: Chichester, U.K., 2007. (2) Mahoney, C. M. Mass Spectrom. Rev. 2010, 29 (2), 247−293. (3) McDonnell, L. A.; Heeren, R. M. A. Mass Spectrom. Rev. 2007, 26 (4), 606−643. (4) Donato, P.; Cacciola, F.; Tranchida, P. Q.; Dugo, P.; Mondello, L. Mass Spectrom. Rev. 2012, 31 (5), 523−559. (5) Abian, J. J. Mass Spectrom. 1999, 34 (3), 157−168. (6) Pongpun, N.; Mlynski, V.; Crisp, P. T.; Guilhaus, M. J. Mass Spectrom. 2000, 35 (9), 1105−1111. (7) Kuwana, T. Physical Methods in Modern Chemical Analysis 1; Academic Press: Orlando, FL, 1978−1983; p68. (8) Solouki, T.; Szulejko, J. E.; Bennett, J. B.; Graham, L. B. A. J. Am. Soc. Mass Spectrom. 2004, 15 (8), 1191−1200. (9) Creba, A. S.; Weissfloch, A. N. E.; Krogh, E. T.; Gill, C. G. J. Am. Soc. Mass Spectrom. 2007, 18 (6), 973−979. (10) Amirav, A.; Gordin, A.; Hagooly, Y.; Rozen, S.; Belgorodsky, B.; Seemann, B.; Marom, H.; Gozin, M.; Fialkov, A. B. Tetrahedron 2012, 68 (29), 5793−5799. (11) Kientz, C. E.; Hulst, A. G.; De Jong, A. L.; Wils, E. R. J. Anal. Chem. 1996, 68 (4), 675−681. (12) Cappiello, A. Mass Spectrom. Rev. 1996, 15 (5), 283−296. (13) Palma, P.; Famiglini, G.; Trufelli, H.; Pierini, E.; Termopoli, V.; Cappiello, A. Anal. Bioanal. Chem. 2011, 399 (8), 2683−2693. (14) Castro, J.; Pregibon, T.; Chumanov, K.; Marcus, R. K. Talanta 2010, 82 (5), 1687−1695. (15) Burlingame, A. L. Biological Mass Spectrometry; Methods in Enzymology, Vol. 402; Elsevier Academic Press: Amsterdam, The Netherlands, 2005; p 423. (16) Litherland, A. E.; Zhao, X.-L.; Kieser, W. E. Mass Spectrom. Rev. 2011, 30 (6), 1037−1072. (17) Hellborg, R.; Skog, G. Mass Spectrom. Rev. 2008, 27 (5), 398− 427. (18) Elmore, D.; Phillips, F. M. Science 1987, 236 (4801), 543−550. (19) Fahrni, S. M.; Wacker, L.; Synal, H.-A.; Szidat, S. Nucl. Instrum. Methods Phys. Res., Sect. B 2013, 294, 320−327. (20) Salazar, G.; Ognibene, T. Nucl. Instrum. Methods Phys. Res., Sect. B 2013, 294, 300−306. (21) Ruff, M.; Wacker, L.; Gäggeler, H. W.; Suter, M.; Synal, H.-A.; Szidat, S. Radiocarbon 2007, 49 (2), 307−314. (22) Liberman, R. G.; Hughey, B. J.; Skipper, P. L.; Wishnok, J. S.; Klinkowstein, R. E.; Shefer, R. E.; Tannenbaum, S. R. Nucl. Instrum. Methods Phys. Res., Sect. B 2004, 223−224, 82−86. (23) Thomas, A. T.; Stewart, B. J.; Ognibene, T. J.; Turteltaub, K. W.; Bench, G. Anal. Chem. 2013, 85 (7), 3644−3650. (24) Thomas, A. T.; Ognibene, T.; Daley, P.; Turteltaub, K.; Radousky, H.; Bench, G. Anal. Chem. 2011, 83 (24), 9413−9417. (25) Roberts, M. L.; von Reden, K. F.; Burton, J. R.; McIntyre, C. P.; Beaupré, S. R. Nucl. Instrum. Methods Phys. Res., Sect. B 2013, 294, 296−299. (26) Ramsey, C.; Ditchfield, P.; Humm, M. Radiocarbon 2004, 46 (1), 25−32. (27) Wacker, L.; Fahrni, S. M.; Hajdas, I.; Molnar, M.; Synal, H.-A.; Szidat, S.; Zhang, Y. L. Nucl. Instrum. Methods Phys. Res., Sect. B 2013, 294, 315−319. (28) Salazar, G.; Zhang, Y. L.; Agrios, K.; Szidat, S. Nucl. Instrum. Methods Phys. Res., Sect. B 2015, 361, 163. (29) Sanna, G.; Tomassetti, G. Introduction to Molecular Beams Gas Dynamics; Imperial College: London, 2005; p 334. (30) Synal, H.-A.; Stocker, M.; Suter, M. Nucl. Instrum. Methods Phys. Res., Sect. B 2007, 259 (1), 7−13. (31) Szidat, S.; Salazar, G.; Vogel, E.; Battaglia, M.; Wacker, L.; Synal, H.-A.; Turler, A. Radiocarbon 2014, 56 (2), 561−566. (32) Ktalkherman, M. G.; Emel’kin, V. A.; Pozdnyakov, B. A. J. Eng. Phys. Thermophys. 2010, 83 (3), 539−548. (33) Sivapragasam, M.; Deshpande, M. D.; Ramamurthy, A. S.; White, P. Sadhana 2014, 39 (3), 713−729. 1653


Helium Mixture

Recommend Documents