Low Temperature Catalytic Combustion Reactors for High Precision

2 days ago - Activated Research Company, 7561 Corporate Way, Eden Prairie , Minnesota 55344 , United States ... [email protected]. Phone: ...
0 downloads 0 Views 2MB Size
Download PDF
Article pubs.acs.org/ac

Cite This: Anal. Chem. XXXX, XXX, XXX−XXX

Low Temperature Catalytic Combustion Reactors for High Precision Carbon Isotope Measurements in Gas Chromatography Combustion Isotope Ratio Mass Spectrometry Herbert J. Tobias,*,† Andrew Jones,‡ Charlie Spanjers,‡ Larry Bowers,§ and J. Thomas Brenna†,⊥

Downloaded via LMU MUENCHEN on February 7, 2019 at 12:21:24 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

†

Dell Pediatric Research Institute, Dell Medical School, University of Texas at Austin, 1400 Barbara Jordan Boulevard, Austin, Texas 78723, United States ‡ Activated Research Company, 7561 Corporate Way, Eden Prairie, Minnesota 55344, United States § LDBowers, LLC, Southern Pines, North Carolina 28387, United States ⊥ Department of Chemistry, College of Natural Sciences, University of Texas at Austin, Austin, Texas 78712-1224, United States S Supporting Information *

ABSTRACT: Metal oxide-filled reactors constructed with ceramic tubes or fused silica capillary are widely used for combustion in gas chromatography combustion isotope ratio mass spectrometry (GCCIRMS). However, they tend to be easily cracked or broken and prone to leaks at operating temperatures of ∼950 °C. Here we introduce a modified commercially available catalytic combustion/reduction methanizer to quantitatively convert organics to CO2 for δ13C analysis while retaining chromatographic resolution. These modified “ARC” reactors operate with a transition-metal catalyst that requires a flowing O2 gas to enable complete conversion to CO2 at lower temperature (620 °C) with acceptable reactor life, reduced complexity, and improved robustness. Performance of two versions of the ARC reactors with different combustion volumes was characterized by analysis of steroid and alkane isotopic standard materials. Linearity of steroid isotopic standards ranged from 0.02 to 0.60 ‰/V in the range of 25 to 200 ng of each steroid injected. Precisions and accuracies of measurements for steroids and alkanes had average standard deviations of SD(δ13C) less than ±0.18 ‰ and average accuracy of better than 0.19 ‰ δ13CVPDB. Peak width expansion within both devices were similar to that in traditionally used metal oxide reactors. These data demonstrate for the first time that novel combustion schemes enable operation at lower temperatures as an alternative approach comparable to high temperature techniques to yield high precision δ13C data with GCC-IRMS.

■

INTRODUCTION

δ13CStandard =

Measurements of the natural stable isotopic variability of light elements C, H, N, O, and S are widely used to establish sources and fates of chemical compounds and complex natural mixtures. Sourcing is used in many scientific disciplines,1,2 with applications in geochemistry,3 ecology,4 food authentication,5 environmental,6 and forensic sciences.7 Isotope ratios are also used as nonradioactive labels in tracer studies.8 Isotope ratio mass spectrometry (IRMS)9−11 is the technique used to measure the stable isotope ratios of these light elements, specifically 13C/12C, 2H/1H, 15N/14N, 18O/16O, 34S/32S. Molecules of interest native to the sample must be converted to a single largely inert gas for isotopic analysis, typically CO2, H2, N2, CO, and SO2, that reflect the isotopic composition of the original target analyte. Carbon isotope ratio (CIR) data is expressed in units of parts per thousand (‰) of the difference between the CIR of a sample and a standard traceable to the international standard Vienna PeeDee Belemnite (VPDB), δ13CVPDB (‰) and is defined in eq 1:12 © XXXX American Chemical Society

R Sample R Standard

−1

(1)

where R is the measured ratio13C/12C of either the sample or a standard.13 When individual compounds in mixtures are the target analytes, gas chromatography coupled to mass spectrometry is used (GC-IRMS) and referred to as compound specific isotope analysis (CSIA). For CIR measurements, the technique requires an online chemical combustion reactor between the GC and IRMS to convert the separated individual organic compounds to CO2 and is known as GC combustion IRMS (GCC-IRMS).9−11 These systems conventionally employ reactors made of a ceramic tube or to a much lesser extent for high resolution applications, a fused silica capillary held in place in a ceramic tube.14−17 They are filled with Cu or Ni Received: November 1, 2018 Accepted: January 28, 2019

A

DOI: 10.1021/acs.analchem.8b05043 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

Figure 1. Images of the two versions of the ARC reactors built and characterized in this work where (A) ARC1 is displayed prior to installation and (B) ARC2 is installed on top of GC2. ARC2 has narrower channel inner dimensions than ARC1. (C) Depicted is the general schematic of the ARC reactor and its catalytic chemistry. Input and output tubing/capillaries are positioned on the bottom side for the He carrier gas and analytes, connecting the gas flow path between the GC and IRMS. On the top is the input for an oxidation gas, such as Zero Air or He/1%O2, that supplies reagent combustion oxidant to the reactor. The flow rate for this gas was normally around 0.3 mL/min.

were obtained from Steraloids (Newport RI) at >98% purity, which were subsequently used to create steroid isotopic standards (SIS) consisting of either single steroids or steroid mixtures.25,26 These SIS were characterized by GCC-IRMS for their δ13CVPDB using the CH4 and C2H6 contained in the NIST RM8559 natural gas isotopic standard. The SIS used were CU/PCC 40−1, 41−1, and 42−1 each containing a single steroid, either androsterone (A), androsterone acetate (A-AC), or cholestane (Cne), respectively. The CU/PCC 44−1 contained A-AC, 5β-androstanediol diacetate (5βA-diAC), Cne, and 5β-pregnandiol diacetate (5βP-diAC). These were created in 2014−201525 at Cornell University supported by the Partnership for Clean Competition (PCC). These standards are available to the WADA antidoping laboratories through the National Measurement Institute (NMI) in Sydney, Australia. The n-alkanes (C16−C30) in the stable isotopic standard A2 were determined at Indiana University by combusting multiple aliquots of each pure n-alkane off-line, and CIR are reported relative to the VPDB scale where NBS 19 and L-SVEC are defined as +1.95 and −46.6 ‰.27 ARC Reactors. Conventional GCC-IRMS employs combustion reactors made of a ceramic tube or fused silica capillary filled with Cu or Ni wires, which are oxidized with a flowing stream of O2 before use. Most often Pt wire is included as a catalyst. The tubular metal oxide reactor is commonly held in a high temperature furnace at ∼950 °C. However, these reactors are well-known to be fragile and leak-free connections to them are difficult to make and maintain when operating at high temperatures and temperature gradients. PolyARC reactors were designed as methanizers, where the components in the effluent from the GC are combusted to CO2 and then reduced to CH4 using combustion and reduction chambers with transition metal catalysts.21−24 This provides enhanced and consistent signal detection based on absolute carbon number of each molecule using flame ionization detectors (FID) at relatively low temperature (450−500 °C), and facilitates straightforward quantitative calibration of detector response. Here, the reactor was redesigned to contain only a catalytic combustion microchamber operating at a lower temperature (620 °C) than conventional IRMS combustion reactors. Combustion catalysts, such as Pd on Al2O3,23 are widely discussed in the literature and are used industrially in commercial microreactors to improve gas chromatography workflow by converting analytes to CO2 with >99.9% conversion and subsequently to CH4.29 In this work, a custom modified Polyarc reactor was developed to efficiently combust steroids and hydrocarbons to CO2 with an optimized flow path and catalyst design that minimizes losses of resolution and sensitivity due to peak

wires oxidized using an O2 stream before any samples are analyzed. The oxidized metal serves as a source of reagent O2 for combustion via the chemical reaction MOx → M + O2. Pt wire is most commonly added as a catalyst to assist in achieving complete combustion. Quantitative combustion is essential for accurate and precise CIR measurements because incomplete combustion leads to significant isotopic fractionation that is difficult to calibrate. All combustion reactors used to date are held at 850−1150 °C, depending on the stability of the analytes and metal oxide used.11,18 In addition, minimal peak broadening to retain chromatographic resolution postcolumn is a challenge due to the additional flow pathways. Reactor structural materials must withstand the high temperatures required for combustion. Materials for reactor connections in the transition from GC oven temperatures as low as 30 to 360 °C, are limited, particularly those used in commercially available systems. They must have low dead volumes and enabling facile, leak-free connections. Reduction of flow path dimensions for robust fast GC and comprehensive 2D GC (GC × GC) is particularly difficult. Alternative designs for the combustion reactor that improve robustness and mitigate these issues are desirable and have previously been explored with limited success.19 In this work, we customized, developed, and characterized low temperature catalytic reactors for GCC-IRMS based on the original PolyARC or “ARC” reactor (Activated Research Company, Eden Prairie, MN).20−24 Performance of the reactors was evaluated with steroids that are normally analyzed by GCC-IRMS in doping control for detection of synthetic steroid use in sport. The reactor capabilities were determined using well- characterized pure steroid isotopic standards that were previously developed in our laboratory.25,26 In addition to natural gas to explore combustion efficiency and sample linearity, here by directly injecting various amounts of CH4, we analyzed an isotopically standardized n-alkane mixture to demonstrate the method’s usefulness in both oil and gas exploration applications (e.g., tracing geological history of hydrocarbons).27,28

■

EXPERIMENTAL SECTION Chemicals and Isotopic Standards. High purity He (99.999%), high purity Zero Air, high purity CH4, and high purity He/1% O2 balance were purchased from Airgas Inc. (Austin, Texas). Acetone (ACS Certified) used for the dry ice/ acetone trap, heptane (HPLC grade), and 2-propanol (ACS Certified Plus) were purchased from Fisher Scientific (Waltham, MA) as solvents for n-alkanes and steroids, respectively. In previous work, all the steroids in this study B

DOI: 10.1021/acs.analchem.8b05043 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

the narrow, full width at half-maximum (fwhm) of ∼120 ms wide CO2 peaks delivered for subsequent peak broadening experiments is presented in the top panel (Figure 2A). Here, a 2 m × 0.1 mm i.d. capillary was connected to a 0.7 m × 0.1 mm i.d. transfer line to the open split of the IRMS. Figure 2B was the setup for the experiments involving chromatography for steroids or alkanes, where a 20 m × 0.15 mm i.d. × 0.15 μm Rxi-1 ms GC column (Restek, Bellefonte, PA) was used. The flow rate was set to a total of 1.3 mL/min for normal steroid and alkane separation experiments using a 2.0 m × 0.10 mm i.d. transfer line with ARC1 and a 1.6 m × 0.10 mm i.d. transfer line with ARC2. The length difference is due to greater flow restriction in ARC2. A programmable temperature vaporization (PTV) inlet15,17,31,32 was used for sample injection and removal of solvent before the column and ARC, which would potentially deplete the oxygen supply of the ARC reactor. The added auxiliary oxidation gas (either Zero Air or He/1%O2 blend gases) that feeds into the ARC was set to 20 and 28 psi for the ARC1 and ARC2, respectively. The PTV was held at 50 °C and operated at a high split ratio (500:1 to 800:1 depending on head pressure and GC temperature) so that low volatility steroids or alkanes condense in the inlet while most of the solvent, in this case 2-propanol or heptane respectively, is vented for 1 min. Subsequently, the inlet was rapidly heated to vaporize analytes onto the GC column. The column in GC1 (heated zone 1) was initially held at 100 °C for 3.5 min, ramped at 30 °C/min to 280 °C, and then held for 6 min for steroids. For alkanes, the program was the same except the final temperature was 300 °C, and then held for 6 min. Head pressure was adjusted to achieve desired flow rate out of the capillary end before the open split, ranging from 0.7 to 4.5 mL/ min for peak broadening characterization experiments. The flow rate was set to total of 1.3 mL/min for the normal steroid and alkane separation experiments. Calibrations of CO2 gas from the Conflo III were done using the same standard materials analyzed at the highest mass injections (i.e., 200 ng) and applied to subsequent analyses of the analytes. Direct injections of CH4 for combustion and linearity tests were done with both ovens at 30 °C. When simulating GC × GC peaks, a hot jet and cold jet, from a Zoex X2 system (Houston TX) on the integrated Shimadzu/Zoex comprehensive GC × GC fast-qMS system, were directed past the end of the column before the ARC. The cold jet remained on (N2 gas from LN2 tank at 90−110 °C), trapping analyte during a set time period (in this case 2 s), and then the hot jet was pulsed (at 350 °C for 0.35 s) to release the trapped analytes. In experiments where steroids or alkanes were injected (except CO2 injections), a loop of the transfer capillary was made after the exit from the reactor to the open split inlet of the IRMS. This loop was immersed in a dry ice/acetone bath to trap water at −78 °C to avoid protonation of CO2, which would produce HCO2+ in the ion source.33 The capillary was then removed automatically between sample runs by a homebuilt robot device triggered via a GC program relay to warm to room temperature and bleed off the condensed water. Periodic water removal prevents capillary occlusion by trapped ice evident after a few sample analyses. The loop was automatically immersed again at the beginning of each analysis. A MAT 253 IRMS (Thermo Fisher Scientific, Waltham MA) was used with a custom-made high speed data acquisition data logger board and software to acquire data at up to 37 Hz (Thermo Finnigan, Bremen, Germany) enabling sufficient

shape distortions. Full understanding of the kinetics and thermodynamics of the catalytic combustion on transition metals, particularly in flowing systems, remain under active study and details are controversial.30 The proprietary catalysts used here were specifically formulated to support use in highprecision carbon isotope ratio measurements for potential future commercialization. Figure 1A and B show images of the two versions of the reactors “ARC1” and “ARC2” respectively. Figure 1C depicts a schematic of the device and its catalytic chemistry.21−24 Input and output tubing to the ARC are on the bottom consisting of deactivated metal input/output capillaries in the flow path between the GC and IRMS and use robust leak-free metal ferrule unions. On the top is the input for the oxidation gas, such as Zero Air or He/1%O2 blend, that mixes with the GC effluent to maintain oxidative capacity for the catalyst. The reactor operates with combustion volumes consisting of a transition-metal catalyst.21−24 The channels and volumes are microfabricated and the ARC2 has a two-fold narrower flow path or effective diameter than ARC1, to minimize peak broadening as analytes travel through the reactor. Conventional GCC-IRMS Instrument Configuration. The instrument setup for GCC-IRMS using a high temperature metal oxide tubular reactor is described in the Supporting Information. A schematic of the system is depicted in Supporting Information Figure S-1. The reactor contained CuO, NiO, and Pt held at 950 °C. GC-ARC-IRMS Instrument Configuration. Instrument setup used to characterize ARC reactors in this work (Figure 2) consist of a Shimadzu QP 2010 GC-MS (Shimadzu, Kyoto Japan) with 2 GC ovens, ARC reactor, IRMS, and remaining interface. The ARC reactor was mounted atop the GC2 (heated zone 2). The system schematic used to characterize

Figure 2. Instrumentation schematic. Setup for (A) characterizing the narrow CO2 peak shapes delivered for peak broadening experiments. A 2 m × 0.1 mm i.d. capillary was connected to a 0.7 m × 0.1 mm i.d. transfer line to the open split of the IRMS. (B) Main setup, including the ARC reactor. For peak broadening experiments, a 2 m × 0.1 mm i.d. capillary was used for the ARC. For experiments involving chromatography, a 20 m × 0.15 mm i.d. × 0.15 μm Rxi-1 ms GC column was used. C

DOI: 10.1021/acs.analchem.8b05043 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

GCC-IRMS, by far the most common CSIA technique used in the CIR analysis community. Future work will involve further design changes to facilitate analysis of narrower peaks at lower flow rates. Combustion and Linearity of CH4. Complete combustion of methane using low temperature catalysts has been reported in the literature;36,37 however, that work did not involve online reactors suitable for GC applications and has been done without any stable isotopic measurements. Measurement of CH4 is a common way to initially assess combustion capacity of online CIR with IRMS and is an important molecule for CIR measurements in mud gas isotope logging (MGIL), a natural gas fingerprinting technique that uses isotopic analyses of the mud stream in drilling wells to access the extent of the reservoir.38 Using the ARC2 reactor for combustion with He/1% O2 oxidation gas, linearity of the CIR of CH4 gas was measured over the range of 5−85 μL CO2 injections, resulting in 525−8866 mV intensity (0.57−9.7 Vs area) for m/z 44. A full system with 2 m × 0.1 mm i.d. capillary, ARC2 at 620 °C, 28 psi He/1%O2, and 1.6 m × 0.1 mm i.d. transfer line was used. For the ARC reactors, while higher temperatures increase the reaction kinetics, overly elevated temperatures reduce reactor durability and accelerate materials degradation. We found 620 °C to give the optimum performance/longevity. This resulted in a linearity of 0.05 ‰/V with standard deviations of SD(δ13C) ± 0.28 ‰ over that range (Figure 3). Comparison of 10 μL injections of CO2

acquisition speed for some fast GC peak characterization and CIR analysis. Commercially offered MAT 253 IRMS systems acquire only up to 10 Hz. The feedback resistors and capacitors were reconfigured to speed up and match the resistor-capacitor time constants of the Faraday cup detectors appropriate for the fast GC peaks.17,32 The linearity of the IRMS using CO2 pulses from a Conflo III device was well within the manufacturer’s spec of 0.060 ‰/V (in this work averaging 0.020 ‰/V) at 2000−8000 mV of m/z 44 signal from CO2 and an absolute sensitivity of approximately 1300 molecules/ion in continuous flow mode.

■

RESULTS AND DISCUSSION Peak Broadening. Three general types of GC front end separation systems are used or in development for coupling to IRMS: traditional GC, Fast GC, and GC × GC. The traditional GC yields separations with peakwidths on the order of seconds (fwhm), while the latter two techniques require minimal peak broadening through the combustion reactor to attain fwhm peaks on the order of few 100 ms. With these requirements in mind, we characterized peak broadening through two versions of ARC. With the addition of oxidation gas, optimal column flow rate for conventional GCC-IRMS is around 1.3 mL/min, striking a balance between chromatographic performance and analyte sensitivity due to the IRMS acceptance flow of about 0.3 mL/ min. Peak broadening through the ARC reactors was determined using injections of 5 μL CO2 at high split (300:1 to 500:1). Figure S-2 in the Supporting Information shows the peak shapes of CO2 gas injections at 1.0 mL/min flow into the IRMS. The Figure S-2A setup was for a 2 m × 0.1 mm i.d. capillary to the IRMS with no ARC and resulted in 123 ms fwhm peaks. Figure S-2B depicts peak shape when ARC2 inline with a 2 m × 0.1 mm i.d. capillary and a 0.7 m × 0.1 mm i.d. transfer line to the IRMS open split was used; here the fwhm was 890 ms. Figure S-2C is the same setup as Figure S-2B, but with ARC1 inline and resulted in 2840 ms wide peaks. The ARC2 narrower internal dimensions produce less peak broadening than ARC1. Table 1 lists the peak widths at different flow rates for both ARC devices. When the reactor is hot, flow rates are higher and the peaks are narrower. The goal of 250 ms fwhm peak widths sufficient for fast GC or GC × GC34,35 was achieved at 4.5 mL/min flow rate with ARC1, higher than the 1.0−1.3 mL/min required for optimal overall sensitivity. Both ARC1 and ARC2 work well for conventional

Figure 3. Example of linearity of the δ13C measurement of CH4 gas injections over a large dynamic range using the ARC2 reactor for combustion with He/1% O2 oxidation gas.

and CH4 and the m/z 16 (unconverted CH4) trace from CH4 combustion indicated at least >99.999% combustion efficiency. Precision of methane isotope measurements were SD(δ13C) ± 0.09 ‰ (n = 6) in line with expectations for conventional 950 °C combustion systems. Moreover, in the Activated Research Company laboratory, a 0.1 wt % each of dodecane and eicosane in 1 μL of heptane through the ARC reactors yielded no signal when the primary ion m/z 57 was monitored by a quadrupole mass spectrometer or by a flame ionization detector (FID), the latter of which is not responsive to CO2, but is responsive to all hydrocarbons. δ13C of Steroids. The precision and linearity of the CIR measurements for the steroid isotopic standards (SIS) was explored using ARC1 and ARC2 with Zero Air and He/1%O2 as oxidation gases. Isotopic analysis of 13C in steroids is a standard method for confirmation of an adverse analytical screening that indicates the use of banned39 synthetic versions of endogenous steroids, particularly testosterone (T), that athletes use to enhance endogenous steroid levels.40−44 Figure

Table 1. Fwhm Peak Widths of ∼120 ms (at 1 mL/min) That Were Directed through ARC Reactors at Conditions Listed Resulting in Various Amounts of Peak Broadeninga ARC temp. (°C)

ARC gas (ARC 1,2 psi)

total FLoq rate (mL/min)

ARC 1 Peak Widths (ms)

ARC 2 peak widths (ms)

25 25 620 25 620 25 620

0,0 NA,26 20,26 NA,26 NA,26 20,26 20,26

0.7 1.0 1.0 1.3 1.3 4.5* 4.5*

77700 ± 183 ND 2840 ± 42 ND ND 1221 ± 20 665 ± 15

1740 1079 886 890 610 295 250

± ± ± ± ± ± ±

42 13 13 14 12 5 10

CO2 gas was injected at high split (500:1 for all flow rates or 300:1 @ 4.5 mL/min*). ARC2 results in less peak broadening. Peaks are narrower when reactor is hot. a

D

DOI: 10.1021/acs.analchem.8b05043 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry 4A presents an example chromatogram of 50 ng of each steroid from CU/PCC 44−1. These chromatograms are similar to

Figure 4. GC-ARC1-IRMS chromatograms of (A) 50 ng of each steroid in CU/PCC 44−1 injected onto a 20 m × 0.15 mm i.d. × 0.15 μm Rxi-1 ms GC column with a 2.0 m × 0.10 mm i.d. transfer line and (B) the GC × GC simulation of 25 ng of each steroid in CU/ PCC 44−1 injected and modulated at 2 s intervals at end of the same GC column (no second column) at 4.5 mL/min with the same transfer line. (C) Chromatogram of the analysis of ∼500 ng each nalkane in the Indiana University certified isotope standard A2: equal mass n-alkane mixture. Analysis of the alkanes was performed using ARC2 at 1.3 mL/min flow with 28 psi zero air oxidation gas and 2 m × 0.1 mm i.d. transfer line.

Figure 5. Linearity of response for CIR measurement of steroid isotopic standard compounds. ARC1 analysis of CU/PCC 44−1 using (A) Zero Air and (B) He/1%O2 as oxidation gas. Linearities range from 0.02 to 0.40 δ13C (‰/V) in the intervals of 25 to 200 ng steroid injected.

diAC had a LLQ between 25 ng to 50 ng. For ARC2 using Zero Air, the steroids A, A-AC, and Cne had LLQ between 12 ng to 25 ng. The performance of the GC-ARC-IRMS system was also compared to conventional GCC-IRMS using a CuO, NiO, and Pt filled high temperature tubular reactor. The performance of the GC-ARC-IRMS system was comparable to conventional GCC-IRMS using a CuO, NiO, and Pt filled high-temperature tubular reactor (Table 2). For 100 ng of each steroid injected, the ARC reactor is comparable to the metal oxide reactor. The ARC1 using Zero Air or He/1% O2 and the ARC2 using Zero Air resulted in measurements with precisions that averaged SD(δ13C) ± 0.15 ‰ and average absolute deviation from certified values was equal to 0.17 ‰. We also tested the capability of doing GC × GC type modulations and CIR measurement with ARC1. The setup and parameters were identical to those used for all other steroid analyses with the exception of the flow rate set to 4.5 mL/min to produce suitable resolution. Effluent at the end of the column was trapped for 2 s by a cold jet, then released by hot jet at 350 °C for 350 ms. A sample chromatogram is shown in Figure 4B. The resulting CIR of the steroids from CU/PCC 44−1 here in elution order were, δ13C of −32.18 ± 0.45 ‰, −30.15 ± 0.70 ‰, −25.59 ± 0.70 ‰, and −21.63 ± 0.70 ‰, with accuracies of 0.60 ‰, 0.04‰, 0.24‰, and −0.47‰, respectively. Compared to our previous work, precisions of SD(δ13C) < ± 1.0 ‰ were expected.17,32 δ13C of Alkanes. The ARC2 system was also tested by measurement of the Indiana University certified alkane mixture

traditional GCC-IRMS chromatograms. GC-ARC-IRMS, using a 20 m × 0.15 mm i.d. × 0.15 μm Rxi-1 ms GC column with a 2.0 m × 0.10 mm i.d. transfer line, was used for this work in determining the CIR and linearity of steroid analysis. The plots of the CIR measured for steroids in CU/PCC 44−1 for 25, 50, 100, and 200 ng each steroid injected are depicted in Figure 5. In Figure 5A, Zero Air was used, where the linearity ranged from 0.02 to 0.09 ‰/V for the steroids, while in Figure 5B, He/1%O2 was used, where the linearity ranged from 0.04 to 0.40 ‰/V. Figure S-3 presents the plots of the CIR measured for steroids in CU/PCC 40−1, 41−1, 42−1, also for 25, 50, 100, and 200 ng each steroid injected. In Figure S-3A, Zero Air was used with ARC1, where the linearity ranged from 0.10 to 0.20 ‰/V for the steroids, while in Figure S-3B, ARC2 was used with Zero Air, where the linearity ranged from 0.04 to 0.60 ‰/V for the steroids. For the most part, these linearities are comparable to manufacturers specification for CO2 pulse analysis at