Anal. Chem. 1998, 70, 1030-1032

Technical Notes

Gas Chromatograph Injection Liner for Continuous Analyte Admission into a Mass Spectrometer Thomas N. Corso, Colleen K. Van Pelt, and J. Thomas Brenna*

Division of Nutritional Sciences, Cornell University, Savage Hall, Ithaca, New York 14853

An inexpensive modification to a gas chromatography injector liner is reported that facilitates continuous admission of analyte into a gas chromatograph/mass spectrometer (GC/MS) for methods development. The MS methods development liner can be made by making simple modifications to commercially available liners and fits into standard injectors in place of the normal liners without any need to break vacuum in the MS. The injector temperature and gas flow rates are adjusted to provide appropriate analyte levels in the MS, which can be admitted under conditions identical with those of real analyses, including co-admission of column bleed. The device is particularly useful for development of tandem MS methods in GC/MS/MS instruments, which are configured with the GC as the sole sample inlet. Tandem mass spectrometry (MSn) is one of the most powerful methods for structural and quantitative chemical analysis. The recent commercial availability of MSn capabilities for low-cost tabletop gas chromatograph/mass spectrometers (GC/MS), based on quadrupole ion traps, has widely expanded the range of applications of MSn.1-3 Optimization of ionization and detection parameters is straightforward for single-stage MS experiments, so that tuning based on standards such as perfluorotributylamine provides satisfactory results for a wide range of analytes. MSn parameters involve isolation and dissociation steps, which are closely linked to analyte chemical characteristics, and consequently often require considerable effort to obtain optimal results.1 High-performance magnetic sector and quadrupole MS instruments are typically equipped with multiple sample inlets to the MS, including a heated inlet or probe from which a constant stream of analyte can be admitted for tuning purposes. Low-cost ion trap-based GC/MS instruments are frequently equipped with the GC as the sole sample inlet to the MS. As a result, MSn methods must be developed by optimizing tuning conditions during the short period of time that analyte is

eluting from the GC column. Numerous parameters are relevant for ion trap tandem MS analysis, including rf amplitude during ionization, isolation mass window and time, collision-induced dissociation (CID) amplitude, rf, time, type, and bandwidth, rough and fine ejection amplitudes, prescan type, modulation range and rate, and low and high isolation rf offsets. At least one commercial manufacturer has incorporated software-based routines to rapidly cycle through some of the settings for several parameters so that optimization can be done on eluting GC peaks. However, cycling through several parameters is time-consuming because of the requirement for multiple injections, and interpretation of results is uncertain because analyte concentration changes continuously. We report here a simple and inexpensive GC injection liner that makes it possible to admit sample continuously into any GC/ MS instrument which requires no instrumentation hardware modifications. Because the intact GC injector system is used, actual experimental conditions, including column bleed and flow rates, are maintained during parameter optimization.

* To whom correspondence should be addressed. Phone: (607) 255-9182. Fax: (607) 255-1033. E-mail: [email protected]. (1) March, R. E., Todd, J. F. J., Eds. Practical aspects of ion trap mass spectrometry, Volume III: Chemical, Environmental, and Biomedical Applications; CRC: Boca Raton, FL, 1995. (2) Strife, R. J.; Simms, J. R. J. Am. Soc. Mass Spectrom. 1992, 3, 372-377. (3) Johnson, J. V.; Yost, R. A.; Kelley, P. E.; Bradford, D. C. Anal. Chem. 1990, 62, 2162-2172.

EXPERIMENTAL SECTION Hardware. Electron impact (EI), chemical ionization (CI), and collision-induced dissociation (CID) were performed on a Varian (Walnut Creek, CA) Saturn III QISMS ion trap equipped with a wave board and MSn software. The chromatograph setup consisted of an HP5890 Series I GC (Hewlett-Packard) equipped with a split/splitless injector and a DB5 fused-silica capillary column (J&W, Folsom, CA), 30 m × 0.25 mm, and 0.5-µm film thickness. Samples and Conditions. Analytes were purchased from Sigma Chemical Co. (St. Louis, MO) and used without further purification. In the first experiment, 100 mg of >98% methyl myristate (methyl tetradecanoate, Me14:0) was deposited into the modified development liner. For the second experiment, approximately 200 mg of >95% R-tocopherol was deposited into the modified development liner. These analyte amounts were chosen for convenience only, and significantly lower amounts have been used. The R-tocopherol was analyzed in EI mode, and the Me14:0 was analyzed in CI mode with ethanol as the CI reagent. Conditions common to all analyses were as follows: electron multiplier voltage, 1.4 kV; emission current, 10 mA; manifold (ion trap) temperature, 295 °C; axial modulation, 4.0 V; and acquisition time, 1 s.

1030 Analytical Chemistry, Vol. 70, No. 5, March 1, 1998

S0003-2700(97)00928-1 CCC: $15.00

© 1998 American Chemical Society Published on Web 01/10/1998

Figure 1. (a) Schematic of a conventional commercial split/splitless injector liner and the modified experimental MS method development liner. Modifications are a seal at the bottom of the liner to hold the analyte and a hole for the exit of analyte gas. (b) Schematic of the modified liner in a generic split/splitless injector. Injector heating increases analyte vapor pressure, and the resulting gas escapes out the hole and is simultaneously swept onto the column and out of the split vent by the carrier gas. Control of the sample entering the MS is achieved by injector temperature and the split flow.

Injector Modifications. A conventional SGE (Austin, TX) commercial split/splitless liner was modified in-house in two ways, as shown in Figure 1a. First, the bottom of the conventional commercial liner was sealed to hold the analyte. Second, a small (1-5 mm) hole was made in the liner, allowing the analyte gas to escape into the injector body and be swept by the carrier gas. Prior to installation, the sample is deposited into the modified liner either as a neat compound, in a mixture, or from solution. Caution should always be taken when changing injector liners as the injector normally operates at high temperatures. The GC injector assembly should be at room temperature before beginning installation. The replacement procedure is equivalent to that for any standard liner. Because the column is not removed and vacuum is not broken, the ion trap need not be shut down for the installation. Figure 1b displays the modified liner in a generic split/splitless injector. During heating, analyte gas escapes out of the hole into the injector body, where it is picked up by carrier gas and is simultaneously swept onto the column and out of the split vent. Optimal injector temperature was chosen by slowly increasing its temperature until signal for the specific analyte was observed in the ion trap. The split flow was used for fine adjustment of signal level. The oven temperature was chosen on the basis of its value at the elution time of a normal GC run for the test compounds. For R-tocopherol, the injector temperature was 235 °C with a split flow of 10 mL/min, while the GC oven temperature was 280 °C. For Me14:0, the injector temperature was 70 °C with a split flow of 60 mL/min, while the GC oven temperature was 150 °C. RESULTS AND DISCUSSION The MS methods development liner was first used to determine the optimal parameters for MS/MS of Me14:0. The first stage of ionization employed ethanol CI to minimize the MH+ ion

Figure 2. Total ion current for the continuous inlet of Me14:0 in CI/MS/MS mode while cycling the CID amplitude from 0 to 0.9 V at 0.1-V increments. The left inset shows the spectrum with the CID at 0 V, resulting in the isolated MH+ ion observed at m/z 243. With a CID amplitude of 0.4 V, many product ions are observed in addition to the MH+ ion as seen in the right inset.

Figure 3. (a) Total ion current of the isolated molecular ion of R-tocopherol in EI mode while cycling through 10 different CID amplitudes (0-0.9 V at 0.1-V increments). The inset to the left is performed with a CID amplitude of 0 V, resulting in no fragmentation of the isolated M+ ion at m/z 430. The inset to the right is performed with a CID amplitude of 0.4 V, resulting in product ions at m/z 165 and 205 in addition to the M+ ion at m/z 430. (b) The MS/MS/MS spectrum of R-tocopherol is shown. The parent ion m/z 430 was isolated and dissociated with a CID amplitude 0.4 V. The product ion m/z 165 was isolated and then dissociated into the ions m/z 109, 119, and 137 using a CID amplitude of 0.3 V.

fragmentation. We have found that the easiest manner to determine optimal conditions with the constant analyte concentration is to manually cycle through parameter settings while monitoring signal. For illustrative purposes, we set the instrument to cycle through 10 CID amplitudes from 0 to 0.9 V at 0.1-V increments automatically using Varian’s automated methods development (AMD) routine with all other conditions constant. Figure 2 displays the total ion current of the isolated MH+ ion of Me14:0 during this experiment. Three peaks are presented, Analytical Chemistry, Vol. 70, No. 5, March 1, 1998

1031

corresponding to little or no dissociation. The left inset shows MS/MS performed with a CID voltage of 0 V, resulting in the MH+ ion of m/z 243 as the sole ion observed in the spectrum. The right inset is a display of the mass spectrum observed when the CID voltage is 0.4 V. Both the parent and product ions are present. Because the analyte concentration in the ion trap is constant, optimal CID voltage is readily read off the plot. For R-tocopherol, the development liner was used to determine the optimal parameters for both MS/MS and MS/MS/MS in EI mode. Figure 3a displays the total ion current of the isolated molecular ion of R-tocopherol while cycling through 10 different CID amplitudes (0-0.9 V at 0.1-V intervals) using AMD. The left inset shows MS/MS performed with a CID voltage of 0 V, where again the isolated M+ ion at m/z 430 (430.7) is the sole ion observed. With a CID voltage of 0.4 V, the parent and product ions m/z 165 and 205 are observed, as the MS/MS spectrum in the right inset reveals. The product ion m/z 165 from this spectrum was then isolated and subsequently fragmented with a CID voltage of 0.3 V to yield m/z 109, 119, and 137, as shown in the MS/MS/MS spectrum of Figure 3b. The CID voltage for the second collision stage was obtained manually in instrument control mode. It should be noted that AMD is not available for MS>2 in our version of the vendor’s software. Since the sample is continuously admitted to the trap, the time required to optimize the MS/MS parameters is limited only by the time required to adjust the parameters in the software. Once

1032 Analytical Chemistry, Vol. 70, No. 5, March 1, 1998

the liner was installed, the optimization of parameters for each compound required under 15 min, which is less than the time for a single GC run. This was also the case for development of the MS/MS/MS method for R-tocopherol. In our hands, method development can require tens of GC runs to optimize the numerous parameters required even for MS/MS, so that the time saved with the liner is substantial. Analyte concentration entering the MS was consistent and stable over the several hours the development liner was installed. Some residual analyte memory was observed after the removal of the liner, but after several hours no traces of the analytes could be detected. Baking would be expected to decrease the time to desorb all residual analyte from the column and the injector. The liner is reusable by simple cleaning and can be used with oncolumn injectors with appropriate modifications. In total, it represents a convenient, low-cost alternative to secondary MS inlets for methods development. ACKNOWLEDGMENT This work was supported by NIH Grant GM49209. T.N.C. acknowledges predoctoral support from National Institutes of Health Training Grant DK07158. Received for review August 25, 1997. Accepted December 5, 1997. AC970928I