Carbon Nanotube Fiber Ionization Mass Spectrometry: A Fundamental Study of a Multi-Walled Carbon Nanotube Functionalized Corona Discharge Pin for Polycyclic Aromatic Hydrocarbons Analysis
- 268 Downloads
Mass spectrometry continues to tackle many complicated tasks, and ongoing research seeks to simplify its instrumentation as well as sampling. The desorption electrospray ionization (DESI) source was the first ambient ionization source to function without extensive gas requirements and chromatography. Electrospray techniques generally have low efficiency for ionization of nonpolar analytes and some researchers have resorted to methods such as direct analysis in real time (DART) or desorption atmospheric pressure chemical ionization (DAPCI) for their analysis. In this work, a carbon nanotube fiber ionization (nanoCFI) source was developed and was found to be capable of solid phase microextraction (SPME) of nonpolar analytes as well as ionization and sampling similar to that of direct probe atmospheric pressure chemical ionization (DP-APCI). Conductivity and adsorption were maintained by utilizing a corona pin functionalized with a multi-walled carbon nanotube (MWCNT) thread. Quantitative work with the nanoCFI source with a designed corona discharge pin insert demonstrated linearity up to 0.97 (R2) of three target PAHs with phenanthrene internal standard.
KeywordsPolycylic aromatic hydrocarbons APCI Corona discharge Carbon nanotube Carbon fiber ionization SPME
Mass spectrometry has been historically performed in the laboratory; however, with the advent of desorption electrospray ionization (DESI), it has become a simplified as well as portable technique . The DESI source has the ability to be used outside the mass spectrometer for preparationless analysis of samples at ambient conditions. Considering the mechanism of ionization by desorbed droplets, DESI has primary utility for polar molecules. Other methods such as direct analysis in real time (DART) and desorption atmospheric pressure chemical ionization (DAPCI)  have been similarly used for the determination of nonpolar chemical compounds, including chemical warfare agents  as well as polycyclic aromatic hydrocarbons(PAH’s) .
Uses of mass spectrometry extend to an array of disciplines with applications ranging from trace detection of chemical warfare agents to determination of extraterrestrial gas compositions. Recent space expeditions aim at determination of organic compounds in Martian soil . In particular, polycyclic aromatic hydrocarbons (PAHs) have been of interest since PAHs  were quantified from meteorites possibly originating from Mars . In the past, volatile compounds were thermally desorbed, separated by gas chromatography, and ionized by electron impact ionization . It is planned that the ongoing missions to Mars will employ a laser desorption ionization source with a linear ion trap for trace detection of organic compounds in order to greatly simplify the traditional analytical scheme .
Recent years have also shown much progress in the area of compound extraction prior to instrumental analysis, and one of the most promising techniques is solid-phase microextraction (SPME). SPME is an equilibrium process by which analytes undergo absorption or adsorption onto the surface of the fiber as a means of preconcentration as well as isolation from the sample matrix . This sample preparation technique has been employed in combination with DESI for imaging metabolites  as well as detection of drugs analytes from SPME fibers . Limits of detection in the ppt range have been reported .
Multi-walled carbon nanotubes (MWCNTs) have shown great promise in many of these novel analytical applications. MWCNTs are large surface area molecules that are chemically inert and they are well known for their electrical conductivity. CNTs have been used as stationary coatings in SPME  as well as directly involved in ionization in CNT functionalized paper spray [13, 14], while carbon fibers have been used in carbon fiber ionization (CFI) . CFI as well as DAPCI abide by a similar mechanism of ionization as atmospheric pressure chemical ionization (APCI) .
In this work, assembly of MWCNTs into fibers was investigated for the dual purpose of serving as an SPME extraction fiber and for analyte ionization. Hence, this source would be capable of single fiber SPME experiments, low gas consumption, as well as little sample preparation. PAHs were utilized as a standard to assess limits of detection and linearity. Considering that CFI occurs by corona discharge and there are known discharges utilizing carbon dioxide , this would be a practical extractive ionization source for a Martian atmosphere consisting primarily of carbon dioxide [17, 18]. Because many NASA experiments must be consolidated into a small space, the usage of our MWCNT nanoCFI source may be utilized in DAPCI experiments, but would also be capable of coupling with other analytical techniques. These techniques include cyclic voltammetry , anodic stripping voltammetry , biochemical sensors , as well as use as an extractive scaffold for DAPPI  or thermal desorption SPME for online GC-MS. The usage of a MWCNT fiber is ideal not only as a stand-alone technique, but as the centerpiece for a suite of analytical techniques in a small space.
HPLC-grade acetonitrile (ACN), formic acid, toluene, dichloromethane (MeCl), and water were purchased from Fisher Scientific (Waltham, MA, USA). ACS grade benzyl alcohol was also purchased from Fisher Scientific. ACS-grade fluorene, benz(a)anthracene, phenanthrene, and naphthalene were purchased from Fisher Scientific. The State of Pennsylvania Polynuclear Aromatic Hydrocarbons Extractables Mix standard was purchased from Fisher Scientific. This standard includes 2000 ppm benz(a)anthracene, benzo(a)pyrene, fluorene, naphthalene, and phenanthrene in MeCl.
The mass spectrometer was a Thermo MSQ equipped with a single quadrupole mass filter. The Thermo MSQ was run in both selected ion monitoring (SIM) and full scan between 100–300 m/z in positive ion mode. The psuedomolecular ion was monitored within 0.2 m/z. The cone voltage was set for 40 V and the probe heater set to 350 °C with a corona discharge current of 5 μA. The corona discharge voltage was 3.2–3.3 kV VDC for corona discharge pin insert and the corona discharge voltage was 3.0 kV VDC for the standard corona discharge pin. Data was collected over a 5 min period, where integrations were performed for 3 min. In the minute prior and the minute after the integration, the corona discharge current and probe heater were shut off. The corona discharge current and temperature were left off for the first min of the run, turned on after the first min, kept on until the fourth min, and the run was concluded after 5 min. Peak areas were taken after subtraction of blank integrations.
Functionalized Corona Discharge Pin Prototype
Functionalized corona discharge pins were produced by attaching a ~5 mm length of 170 micron thick MWCNT fiber to the corona pin with silver epoxy, as shown in the Supplementary Information. The functionalized corona pin was cured at 60 °C for 1 h to establish a connection between the pin and the MWCNT thread. Validation of connection was established by appearance of the corona discharge compared with a standard corona discharge.
Corona Discharge Pin Insert and Seat
Results and Discussion
Functionalized Corona Discharge Pin
Prior to analysis, the functionalized corona discharge pin was compared with a standard corona pin. The fine point on the standard corona discharge pin is shown in Figure 2a. The corona discharge was visually observed at 10 uA with both the standard discharge pin and the functionalized corona discharge pin. Visual inspection proved the ionization process was not consistent with field desorption or CNT functionalized paper spray, but rather that the ionization mechanism was consistent with the corona discharge.
The experimental design was streamlined regarding extraction, apparatus, and data analysis. Samples were prepared as PAHs in methylene chloride, and immersion extractions were performed by exposing the entire length of the MWCNT for 10 min periods, after which the functionalized pin was immediately placed inside the mass spectrometer. In order to disprove a probe electrospray-based process, only nitrogen gas flow was employed for the first min to promote desolvation. Thereafter, the heater and corona discharge current were turned on to enable thermal desorption as well as ionization by corona discharge for 3 min, followed by an additional min without the heat or corona current.
Figure 1c shows the results of analysis of a 15 ppm sample of phenanthrene in methylene chloride. Considering the respectively lower proton affinity of this solvent, the protonated molecular ion peaks were targeted for determination. The molecular weight of phenanthrene is 178 and so the protonated mass of 179 was monitored. The black trace was obtained using the standard corona discharge pin, and the results using the functionalized corona discharge pin are shown in red. The immersed modified corona pin showed a signal signature consistent with thermal desorption. In between analyses, the functionalized pin was immersed in methylene chloride for 1 min to ensure no carryover.
Optimum instrumental parameters for analysis were investigated by analysis of a 100 ppm PAH solution under varying conditions. The PAH solution included 3 PAHs of interest: naphthalene ([M+H]+=129 m/z), fluorene ([M+H]+=167 m/z), and benz(a)anthracene ([M+H]+=229 m/z). Using a heater temperature of 350 C and a 10 μA corona current, the cone voltage was varied from 10 to 50 V in 10 V increments; optimal voltage was found to be 40 V. In a similar fashion, the corona current was varied from 5 to 50 μA, and 10 μA was found to deliver optimum signal for all three protonated molecular ion peaks with minimal in-source fragmentation.
In order to assess the linearity as well as limits of detection, standards from 10 to 100 ppm of the PAH mixture were tested at the optimized parameters. Limits of detection were calculated as 3.3 times the standard deviation divided by the slope and ranged from 8 μg mL–1 for fluorene to 155 μg mL–1 for naphthalene. Linearity of calibration curves was as follows: naphthalene (R2 = 0.357), fluorene (R2 = 0.980), and benz(a)anthracene (R2 = 0.985). Naphthalene had the poorest linearity, likely due to known difficulties in ionization as well as due to 2-piperazinoethylamine ([M]+=129 m/z) interference. This compound is a component of the silver paste used to attach the MWCNT fiber to the corona discharge needle.
Functionalized Corona Discharge Pin Insert
Two measures were taken to remove this interference on naphthalene as well as other extraneous peaks that may have resulted from the silver paste used to attach the MWCNT to the corona discharge needle. First, the prototype was redesigned as detailed in the Experimental section (Figure 2b). Summarily, a MWCNT thread was pulled through a 2 mm o.d., 150 mm long borosilicate capillary; the open flat end of the capillary was filled with silver paste and connected to the corona discharge needle. Second, the sample solvent was changed to acetonitrile (ACN) to reduce the solubility of the silver paste. In addition, ACN is safer and a more common solvent in high performance liquid chromatographic analysis. As a means to further improve on the method, an appropriate wash method was developed to prevent cross-contamination and improve reproducibility (Supplementary Figure S3a). Extraction time was again investigated and 5 min was found to be the optimal time (Supplementary Figure S3b).
Figures of merit were investigated and the calibration curve was expanded to range from 0.1 to 150 μg mL–1, and regression coefficients were calculated by both external and internal standard methods (phenanthrene designated as internal standard, 1 μg mL–1). Linearity of calibration curves was improved from previous results. When including the use of phenanthrene as internal standard, the following figures of merit were obtained: naphthalene (R2 = 0.7361), fluorene (R2 = 0.8685), and benz(a)anthracene (R2 = 0.9693). Reproducibility of the MWCNT pin inserts was shown to have at lowest 4% RSD over three replicates, but over 10 replicates would be as high as 34% for fluorene (166 m/z).
Although this source was capable of qualitative results, quantitation by the nanoCFI source with a designed corona pin insert was capable of linearity up to 0.97 with phenanthrene internal standard. Future work aims at producing a nanoCFI source with varying MWCNT fiber dimensions, such as length and diameter, to improve linearity as well as limit of detection. Experiments will also extend to utilization of the functionalized corona discharge pin insert in a carbon dioxide environment.
Professor Elke Buschbek is acknowledged for her assistance with the MWCNT assembly within a borosilicate capillary using a puller.
- 5.Mahaffy, P.R., Webster, C.R., Cabane, M., Conrad, P.G., Coll, P., Atreya, S.K., Arvey, R., Barciniak, M., Benna, M., Bleacher, L., Brinckerhoff, W.B., Eigenbrode, J.L., Carignan, D., Cascia, M., Chalmers, R.A., Dworkin, J.P., Errigo, T., Everson, P., Franz, H., Farley, R., Feng, S., Frazier, G., Freissinet, C., Glavin, D.P., Harpold, D.N., Hawk, D., Holmes, V., Johnson, C.S., Jones, A., Jordan, P., Kellogg, J., Lewis, J., Lyness, E., Malespin, C.A., Martin, D.K., Maurer, J., McAdam, A.C., McLennan, D., Nolan, T.J., Noriega, M., Pavlov, A.A., Prats, B., Raaen, E., Sheinman, O., Sheppard, D., Smith, J., Stern, J.C., Tan, F., Trainer, M., Ming, D.W., Morris, R.V., Jones, J., Gundersen, C., Steele, A., Wray, J., Botta, O., Leshin, L.A., Owen, T., Battel, S., Jakosky, B.M., Manning, H., Squyres, S., Navarro-Gonzalez, R., McKay, C.P., Raulin, F., Sternberg, R., Buch, A., Sorensen, P., Kline-Schoder, R., Coscia, D., Szopa, C., Teinturier, S., Baffes, C., Feldman, J., Flesch, G., Forouhar, S., Garcia, R., Keymeulen, D., Woodward, S., Block, B.P., Arnett, K., Miller, R., Edmonson, C., Gorevan, S., Mumm, E.: The sample analysis at Mars Investigation and Instrument Suite. Space Sci. Rev. 170(1/4), 401–478 (2012)CrossRefGoogle Scholar
- 9.Goetz, W., Brinckerhoff, W.B., Arevalo, R., Freissinet, C., Getty, S., Glavin, D.P., Siljestrom, S., Buch, A., Stalport, F., Grubisic, A., Li, X., Pinnick, V., Danell, R., van Amerom, F.H.W., Goesmann, F., Steininger, H., Grand, N., Raulin, F., Szopa, C., Meierhenrich, U., Brucato, J.R.: Team, MS: MOMA: the challenge to search for organics and biosignatures on Mars. Int. J. Astrobiol. 15(3), 239–250 (2016)CrossRefGoogle Scholar
- 17.Skalny, J.D., Stoica, A., Orszagh, J., Vladoiu, R., Mason, N.J.: Positive and negative corona discharges in flowing carbon dioxide. J. Phys. D Appl. Phys. 41,(17), 1-7 (2008)Google Scholar
- 18.Mahaffy, P.R., Webster, C.R., Atreya, S.K., Franz, H., Wong, M., Conrad, P.G., Harpold, D., Jones, J.J., Leshin, L.A., Manning, H., Owen, T., Pepin, R.O., Squyres, S., Trainer, M., Team, M.S.: Abundance and isotopic composition of gases in the Martian atmosphere from the Curiosity Rover. Science. 341(6143), 263–266 (2013)CrossRefGoogle Scholar
- 19.Koo, Y., Shanov, V.N., Yun, Y.: Carbon nanotube paper-based electroanalytical devices. Micromachines (Basel) 7(4), 1-9 (2016)Google Scholar