Anal. Chem. 2001, 73, 5365-5369
Enhancing the Detection of Sulfate Particles for Laser Ablation Aerosol Mass Spectrometry David B. Kane and Murray V. Johnston*
Department of Chemistry and Biochemistry, University of Delaware, Newark, Delaware 19716
A major limitation to the application of laser ablation aerosol mass spectrometry for the detection of particles less than 200 nm in diameter is a low ablation efficiency for sulfate particles. (Ablation efficiency is the probability that an ablated particle produces a detectable ion signal.) A method is described here to enhance the ablation efficiency of sulfate particles by coating them with a UVabsorbing compound. The method can be applied in-line with the aerosol mass spectrometer in a manner that does not significantly alter the aerosol size distribution. It is shown that a 12-nm coating of 1-naphthyl acetate increases the ablation efficiency of 136-nm ammonium sulfate particles by at least a factor of 20, while similar coatings on oleic acid and ammonium nitrate particles do not significantly alter the ablation efficiency. The results suggest that “undetected” particles, presumably sulfate, in ambient aerosol can be assessed. Sulfur is an important component of atmospheric aerosols. In the troposphere sulfur as sulfuric acid or ammonium sulfate may contribute to nucleation bursts, leading to the formation of new particles.1,2 Sulfate aerosols influence the global climate by increasing the albedo of the atmosphere3 and acidifying particles and cloud droplets4 and may pose a health risk.5 Because of their great importance in atmospheric chemistry and the short time scale over which atmospheric events occur, methods for rapid identification of sulfate in ultrafine aerosol particles are needed. Aerosol mass spectrometry is an on-line method for the sampling and chemical characterization of airborne particulate matter on a particle-by-particle basis.6-8 A significant number of particles can be sampled in a few minutes, and each particle spends only about 10 ms between sampling and analysis. Aerosol mass spectrometry makes it possible to track changes in the * To whom correspondence should be addressed. Telephone: 302-831-8014. Fax: 302-831-6335. E-mail:
[email protected]. (1) Grenfell, J. L.; Harrison, R. M.; Allen, A. G.; Shi, J. P.; Penkett, S. A.; O’dowd, C. D.; Smith, M. H.; Hill, M. K.; Robertson, L.; Hewitt, C. N.; Davison, B.; Lewis, A. C.; Creasey, D. J.; Heard, D. E.; Hebestreit, K.; Alicke, B.; James, J. J. Geophys. Res., [Atmos.] 1999, 104 (D11), 13771-13780. (2) Kulmala, M.; Pirjola, L.; Makela, J. M. Nature 2000, 404, 66-69. (3) Charlson, R. J.; Wigley, T. M. Sci. Am. 1994, 270, 48-57. (4) Seinfeld, J. H.; Pandis, S. N. Atmospheric Chemistry and Physics: From Air Pollution to Climate Change; John Wiley and Sons: New York, 1998. (5) Dockery, D. W.; Pope, C. A.; Xu, X.; Spengler, J. D.; Ware, J. H.; Fay, M. E.; Ferris, B. G.; Speizer, F. E. N. Engl. J. Med. 1993, 329 (24), 1753-1759. (6) Johnston, M. V. J. Mass Spectrom. 2000, 35, 585-595. (7) Suess, D. T.; Prather, K. A. Chem. Rev. 1999, 99, 3007-3035. (8) Johnston, M. V.; Wexler, A. S. Anal. Chem. 1995, 67, 721A-726A. 10.1021/ac010469s CCC: $20.00 Published on Web 10/06/2001
© 2001 American Chemical Society
chemical composition of particles with high temporal resolution and minimal loss due to evaporation. The recent development of aerosol mass spectrometers for the analysis of particles less than 200 nm in diameter9-16 makes it possible to use these instruments to study particles resulting from photochemically induced nucleation processes, such as those resulting in a nucleation burst. Understanding these processes is essential to determining the environmental and health-related effects of sulfate aerosols. However, aerosol mass spectrometers that rely upon laser ablation are severely limited in their ability to detect sulfate in particles less than 200 nm in diameter.12,14 The ablation threshold, the laser fluence required to produce ions from a particle, and ablation efficiency, the probability that an ablated particle produces a detectable ion signal, are known to be dependent on the composition of the material being ionized.14,17,18 Variations in the absorption cross section and ionization energy of the precursor particle and the material ablated from it are thought to cause these dependencies. For some constituents of atmospheric aerosol particles, such as ammonium sulfate, limited absorption at the wavelengths typically used for laser ablation (193, 248, 266 nm) and high ionization energies combine to give low ion yields from these species. Insufficient ion yields may cause some species within a particle or the entire particle to go undetected. We present a method for increasing the hit rate of ammonium sulfate particles that can be used in-line with an aerosol mass spectrometer. The method uses a thin coating of an aromatic compound to enhance the ablation efficiency of ammonium sulfate particles. The coating is achieved by mixing the aerosol with a supersaturated vapor of an aromatic compound. This study focuses on coating particles with 1-naphthyl acetate. The effect of the (9) Reents, W. D.; Downey, S. W.; Emerson, A. B.; Mujsce, A. M.; Muller, A. M.; Siconolfi, D. J.; Sinclair, J. D.; Swanson, A. G. Plasma Sources Sci. Technol. 1994, 3, 369-372. (10) Reents, W. D.; Ge, Z. Aerosol Sci. Technol. 2000, 33 (1-2), 122-134. (11) Carson, P. G.; Johnston, M. V.; Wexler, A. S. Rapid Commun. Mass Spectrom. 1997, 11, 993-996. (12) Ge, Z.; Wexler, A. S.; Johnston, M. V. Environ. Sci. Technol. 1998, 32 (20), 3218-3223. (13) Mallina, R. V.; Wexler, A. S.; Rhoads, K. P.; Johnston, M. V. Aerosol Sci. Technol. 2000, 33 (1-2), 87-104. (14) Kane, D. B.; Johnston, M. V. Environ. Sci. Technol. 2000, 34 (24), 48874893. (15) Kane, D. B.; Oktem, B.; Johnston, M. Aerosol Sci. Technol. 2001, 34 (6), 520-527. (16) Jayne, J. T.; Leard, D. C.; Zang, X.; Davidovits, P.; Smith, K. A.; Kolb, C. E.; Worsnop, D. R. Aerosol Sci. Technol. 2000, 33 (1-2), 49-70. (17) Thomson, D. S.; Middlebrook, A. M.; Murphy, D. M. Aerosol Sci. Technol. 1997, 26, 544-559. (18) Thomson, D. S.; Murphy, D. M. Appl. Opt. 1993, 32 (33), 6818-6826.
Analytical Chemistry, Vol. 73, No. 22, November 15, 2001 5365
coating on the ablation efficiency of ammonium sulfate particles and other particle compositions representative of ambient aerosols is presented. The detection of ammonium sulfate particles in ambient aerosol is also discussed. EXPERIMENTAL DETAILS For these experiments, test aerosols were generated by atomizing a solution containing the compound of interest and size selecting a monodisperse fraction (∼125-nm diameter) of the aerosol with a differential mobility analyzer. Ambient aerosols were sampled directly from laboratory room air at the University of Delaware. Depending on the experiment, the aerosols were either drawn directly into the aerosol mass spectrometer or coated prior to being sampled by the mass spectrometer. A scanning mobility particle sizer (model 3934; TSI, Inc., St. Paul, MN) was used inline with the mass spectrometer to measure the size distribution of the sampled aerosol. Details of the design and operation of the aerosol mass spectrometer are available elsewhere.14,15 For the experiments described here, an ArF excimer laser (model PSX100; MPB Technologies, Dorval, PQ, Canada) at 193 nm with a 8-ns, 2.5-mJ pulse was used for simultaneous ablation and ionization of the particles. The laser was focused with a 20-cm lens to give a fluence of 2.0 × 104 J/m2 in the mass spectrometer source region. The negative ion mass spectra presented here were obtained on the same instrument in the linear TOF mode with a postacceleration detector. Two types of measurements were made with the aerosol mass spectrometer: the particle hit rate, or the rate at which singleparticle mass spectra were collected, and the particle beam profile. The particle beam profile is measured by translating the ablation laser beam across the ion source of the mass spectrometer and measuring the particle hit rate as a function of laser position. Combining these two measurements with the aerosol concentration measured by the scanning mobility particle sizer makes it possible to determine the ablation efficiency of the particles using the expression for the hit rate.14,15
HR ) naerosolVT(AL/AP)LEa
(1)
where naerosol is the number density of the aerosol, V is the volume flow of aerosol through the inlet, T is inlet transmission efficiency, the ratio AL/AP gives the overlap of the particle beam area (AP) by the laser beam area (AL), L is the laser duty factor, and Ea is the ablation efficiency. The particle beam area (AP) is determined by deconvoluting the particle beam diameter from the laser beam diameter in the particle beam profile.15 Aerosol particles were coated by mixing them with a supersaturated vapor of the coating material in a flow diffusion cloud chamber (FCC).19 A schematic of the FCC is shown in Figure 1. The FCC consists of a saturator, a preheating-mixing region, and a cooled condensation region. Air flows through the saturator filled with the coating material at a volume flow rate of 0.5 L/min. The saturated vapor is drawn into a preheating region where it is mixed with the aerosol flowing at 1.0 L/min. The preheating-mixing region is kept at a slightly higher temperature than the saturator, and the aerosol is given sufficient time to warm to the temperature (19) Vohra, V.; Heist, R. H. J. Chem. Phys. 1991, 104, 382-395.
5366 Analytical Chemistry, Vol. 73, No. 22, November 15, 2001
Figure 1. Schematic diagram of the flow diffusion cloud chamber.
of the vapor before mixing to prevent condensation. From the preheating region of the FCC, the vapor-aerosol mixture flows into the cooled condensation region. Rapid cooling of the vapor on entering the condensation region causes the vapor to supersaturate. Aerosol particles are sampled from the cloud chamber at a flow rate of 1 L/min, and the excess flow is pumped away through an exhaust port. Independent temperature controls are provided for the saturator (Ts), the gas flow at the exit of the preheating-mixing region (T0), and the walls of the condenser (Tw). The degree of vapor supersaturation is controlled by the temperature difference between the saturator and the condensation region.19 If the supersaturation is large enough for heterogeneous nucleation, the vapor will condense on the aerosol particles. In the analysis that follows, it is assumed that the vapor forms a uniform coating on the surface of the particles. However, this may not be the case. Depending on the interfacial surface tension between the coating and the particle and the miscibility of the coating material and the particle, the coating may form a separate domain on the surface of the particle or may dissolve into the particle. The thickness of the coating on the particles can be controlled by the supersaturation of the vapor and the residence time of the aerosol particles in the supersaturated vapor. Figure 2 shows the effect of coating on the particle size distribution. Figure 2a is the size distribution of monodisperse ammonium sulfate particles. These particles have a mean diameter of 136 nm with a geometric standard deviation of 1.15. Figure 2b is the particle size distribution of the same monodisperse ammonium sulfate aerosol after being coated with 1-naphthyl acetate in the FCC. The mean diameter of the particles has increased to 161 nm with a geometric standard deviation of 1.15. The change in the mobility diameter corresponds to an average coating thickness of 12.5 nm on the surface of the particle. As mentioned previously, this value assumes a uniform coating on the surface of the particle. While the flow through the FCC is laminar, aerosol sampling from the condensation region is not isokinetic. Therefore, particles sampled from the FCC travel along different streamlines through the condensation region. Since the velocity profile of the gas in the FCC is parabolic, particles sampled from different streamlines are exposed to different degrees of vapor supersaturation for different lengths of time. This results in some particles having a slightly thicker coating than others. In addition, because of a greater surface-to-volume ratio, the coating will grow faster on small particles than on larger ones. RESULTS AND DISCUSSION Coating Materials. The cloud chamber coating method is very versatile. Coatings of several organic compounds (including
Figure 3. Average of 100 ammonium sulfate single-particle mass spectra from the size distribution in Figure 2a.
Figure 2. (a) Particle size distribution of a monodisperse ammonium sulfate aerosol. (b) Particle size distribution of the monodisperse aerosol after coating with 1-naphthyl acetate. Coating conditions: saturator temperature Ts ) 39 °C, preheater-mixing region temperature To ) 53 °C, and condenser wall temperature Tw ) 15 °C.
benzoic acid, 3-nitroaniline, and oleic acid) have been grown on a variety of particle compositions from salts to organics. However, the choice of a coating material for enhancing the detection rate of sulfate particles in ambient air requires a number of considerations. The primary consideration is that the coating must have a strong absorption at the wavelength of the ablation laser. This is necessary if the coating is to increase the ablation efficiency of the particle. However, the coating must not have too great an effect on the ablation efficiency. If it does, then the hit rates of particles that have ablation efficiencies large enough to be detected without the coating will also be increased. While there are benefits to enhancing the ablation of all particle compositions to a uniform level, this would make it more difficult to detect the presence of pure ammonium sulfate particles. For example, the difference in the number of particles containing sulfate with and without a coating can be used to distinguish the presence of pure sulfate particles from particles containing sulfate mixed with other compounds that would have higher ablation efficiencies. If the coating enhances both the pure sulfate particles and the more easily ablated mixed particles, then this would not be possible. Therefore, a coating that increases the ablation efficiency of sulfate particles to approximately the ablation efficiency of other particle compositions, without increasing the ablation efficiency of the other compositions, is desired. A second consideration is the vapor pressure of the coating material, since this will determine the temperatures required in
the saturator and preheating-mixing regions. If the vapor pressure of the coating material is too low, then a high temperature in the preheating-mixing region will cause evaporation of semivolatile components from the sampled particles. Deleterious effects would include the loss of small particles, changes in the particle size distribution, and the undercounting of some particle compositions. If the vapor pressure of the coating material is too large, then the high partial pressure of the coating material in the gas phase will add significantly to the background ion signal of the aerosol mass spectrometer, making it difficult to distinguish the ion signal of particles from the background. This would also lead to a miscounting of particles. Coating materials with a variety of properties were investigated. In early work, it was found that benzoic acid enhanced particle hit rates with 266-nm laser ablation.20 However, the ablation efficiencies of most compositions typical of ambient particles of less than a 200-nm diameter are very low at 266 nm. Since the hit rates of many particle types are enhanced, it is not possible to use benzoic acid coatings with 266-nm laser ablation to give information on sulfate particles alone. For ablation at 193 nm, which is preferred for aerosol mass spectrometry of ultrafine particles, 1-naphthyl acetate offers the best balance of physical properties and ablation efficiency enhancement. Coating Ammonium Sulfate Particles. A laser ablation mass spectrum of 136-nm-diameter ammonium sulfate particles is shown in Figure 3. This spectrum is the average of 100 single-particle spectra. While the single-particle spectra have a significant amount of variation, the average mass spectrum shows the major ions typically observed. The dominant ions are C+ and NO+, indicating that particles containing organic and nitrate impurities are detected. While these ions dominate the single-particle mass spectra, the impurities may be present only in trace quantities. Middlebrook et al.21 reported that organic impurities could be detected in sulfuric acid particles at concentrations as low as 0.02 wt %. Kane and Johnston14 have shown that particles having only 10% ammonium nitrate exhibit a high ablation efficiency similar to that of pure ammonium nitrate. The presence of impurity ions in the single-particle mass spectra suggests that the ammonium (20) Kane, D. B.; Oktem, B.; Johnston, M. V. Nucleation and Atmospheric Aerosols 2000; AIP Conference Proceeding 534; AIP: Mellville, NY 2000. (21) Middlebrook, A. M.; Thomson, D. S.; Murphy, D. M. Aerosol Sci. Technol. 1997, 27, 293-307.
Analytical Chemistry, Vol. 73, No. 22, November 15, 2001
5367
Figure 5. Comparison of the particle beam profiles of the ammonium sulfate aerosol in Figure 2a and the same aerosol after coating with 1-naphthyl acetate, Figure 2b. Table 1. Ablation Efficiencies of Typical Particle Compositions, and the Same Compositions Coated with 1-Naphthyl Acetate
Figure 4. Average of 100 single-particle mass spectra of ammonium sulfate coated with 1-naphthyl acetate: (a) positive ion mode; (b) negative ion mode. The size distribution is given in Figure 2b.
compound
diameter (nm)
ablation efficiency
ammonium sulfate ammonium nitrate oleic acid 1-naphthyl acetate 1-naphthyl acetate 1-naphthyl acetate
136 114 139 73 106 145
0.014a 0.84 0.20 0.15 0.32 0.42
coating thickness (nm)
ablation efficiency of coated particles
12.5 10 13.5
0.28 0.85 0.23
a For ammonium sulfate, the ablation efficiency depends on the level of impurities in the particles.
sulfate particles are only detected when impurities are present in sufficient amounts. Positive and negative ion laser ablation mass spectra of 136nm ammonium sulfate particles coated with 12.5 nm of 1-naphthyl acetate are shown in Figure 4. These spectra are also the average of 100 single-particle spectra. Not surprisingly, the positive ion spectrum shown in Figure 4a is dominated by ions from the coating material. However, the dominant peak in the negative ion mass spectrum is the HSO4- ion (m/z ) 97) from ammonium sulfate. This ion has been observed in the laser ablation mass spectra of 3-µm ammonium sulfate particles and provides unambiguous identification of sulfate.22 Figure 5 shows the increase in particle hit rate that results from coating ammonium sulfate particles with 1-naphthyl acetate. From the particle beam profiles, it is evident that the maximum hit rate across the beam profile increases by a factor of 20 when the particles are coated. The number of particles that contain impurities determines the hit rate of the uncoated ammonium sulfate particles. The ablation efficiency of the particles can be determined from the particle beam profile and the particle hit rate using the procedure described by Kane and Johnston.14 The ablation efficiency increases from 0.014 for the uncoated particles to 0.28 for the coated particles. The actual increase in the ablation efficiency of “pure” ammonium sulfate particles is greater than (22) Neubauer, K. R.; Sum, S. T.; Johnston, M. V. J. Gerophys. Res. 1996, 101 (D13), 18, 701-18, 707.
5368 Analytical Chemistry, Vol. 73, No. 22, November 15, 2001
the 20-fold increase observed, because the ablation efficiency of the uncoated particles is enhanced by the presence of impurities. Coating Other Particle Compositions. To determine the concentration of ammonium sulfate particles that are undetected in a conventional aerosol mass spectrometry experiment, it is important to know how the coating will affect the ablation efficiency of other composition particles that may be present. Oleic acid and ammonium nitrate are chosen to represent ambient organic and salt particles. In previous work, we determined that polycyclic aromatic hydrocarbons have the highest ablation efficiencies, followed by inorganic salts and then aliphatic organic compounds.14 Since particles composed of polycyclic aromatic compounds have ablation efficiencies near unity, they are not expected to be affected by the 1-naphthyl acetate coating. The ablation efficiencies of oleic acid and ammonium nitrate particles have been measured with and without an approximately 12-nm coating of 1-naphthyl acetate. The data are shown in Table 1. Also given for comparison are the ablation efficiencies of coated and uncoated ammonium sulfate particles and 1-naphthyl acetate particles from 73 to 145 nm in diameter. Table 1 shows that only the ablation efficiency of ammonium sulfate is significantly enhanced by the 1-naphthyl acetate coating. The ablation efficiency of ammonium nitrate, which is much greater than that of a pure 1-naphthyl acetate particle of the same
size, does not change within experimental error when coated. The ablation efficiency of oleic acid, which is about half that of a similar size 1-naphthyl acetate particle shows only a slight enhancement. Since the enhancement of the ablation efficiency decreases as the ablation efficiency increases, we suggest that only particle compositions with ablation efficiencies less than ∼0.1 will be significantly enhanced by the coating. In previous work, it has been observed that the ablation efficiency of an internally mixed two-component particle is approximately the same as that of a single-component particle with the same mass of the more easily ablated material.14 However, the ablation efficiencies observed here are greater than would be expected from this observation. This suggests that the geometry of the coating may play a role in the ablation efficiency of the particle. From molecular dynamics simulations of the laser ablation of particles with transparent inclusions, Schoolcraft et al.23 found that the inclusion enhances the ablation of the surrounding material. The relatively high ablation efficiency of the coated particles may be a manifestation of this effect. Coating Ambient Particles. The cloud chamber has been used to coat ambient particles. Particle size distributions measured with the scanning mobility particle sizer show that lower saturator and preheater temperatures are needed to apply a ∼10-nm layer of 1-naphthyl acetate on ambient aerosols than are needed for the previously discussed laboratory-generated aerosols. This is due to the lower concentration of ambient particles, ∼103/cm3, relative to the laboratory-generated particles ∼104/cm3, and suggests that vapor depletion limits the growth of the coating at higher particle concentrations. Mixing the aerosol with the supersaturated vapor in the FCC results in a dilution of the aerosol by 33% for the flow rates used in these experiments. However, this dilution does not dramatically change the time required to sample the aerosol. Furthermore, for the coating conditions used, no significant change in the ambient particle size distribution was observed when aerosol was passed through the cloud chamber at operating temperatures without the saturated vapor. This suggests that no significant evaporation occurred in the FCC and that particle losses through the cloud chamber are minimal. Coatings of 1-naphthyl acetate between 3 and 20 nm thick have been applied to ambient aerosols. As discussed previously, this is an average coating thickness assuming the particles are uniformly coated. These coatings resulted in a 20% increase in the normalized hit rate, indicating that some particles are indeed “undetected” in a conventional experiment. In principle, an increase in the normalized hit rate may be due to an increase in the ablation efficiency or a change in the geometry of the particles. Changes in the particle size or shape can alter the overlap of the particle beam with the ablation laser beam.14 If the ablation laser samples a greater portion of the coated particle beam, then the normalized hit rate will increase. However, (23) Schoolcraft, T. A.; Constable, G. S.; Zhigilei, L. V. Anal. Chem. 2000, 72 (21), 5143-5150.
measurements of the particle beam profile, such as that shown in Figure 5, demonstrate that the diameter of the particle beam is not significantly affected by an approximately 10-nm-thick coating of 1-naphthyl acetate. Therefore, increases in the normalized hit rate are due primarily to increases in the ablation efficiencies of the particles. Because the ablation efficiency of sulfate particles will be enhanced, the increases in the hit rate can be used to estimate the concentration of pure sulfate particles. While the preliminary experiments presented here are for particles with a nominal diameter of 100 nm, the particle coating is expected to have similar results for particles of different diameters. The ablation efficiency of coated particles is dependent on the mass of the 1-naphthyl acetate coating. Since particles of different diameters should be coated to approximately the same final diameter, the ablation efficiencies of the coated particles should be mostly independent of the initial particle diameter. Sizedependent particle losses in the cloud chamber are another factor, which may affect the application of coating to ambient particles. Diffusion to the walls is expected to be the primary mechanism for particle loss in the laminar flow through the cloud chamber. In the cloud chamber, these losses are expected to be minimal for particles of >50 nm in diameter. To establish the feasibility of the method, an aerosol mass spectrometer capable of only single polarity ion detection was used. The positive ion mass spectra of coated particles are dominated by carbon clusters and hydrocarbon fragment ions from the coating material. For example, in the laser ablation mass spectra of ammonium sulfate particles coated with 1-naphthyl acetate (Figure 4a), ions from the coating obscure the presence of ions expected from the ammonium sulfate, S+ (m/z ) 32) and SO+ (m/z ) 48). Therefore, it is not possible to definitively determine the presence sulfate in the coated particles in positive ion mode. Initial work done in negative ion mode shows an enhancement of the HSO4- ion (m/z ) 97) in the mass spectra of coated ammonium sulfate particles (Figure 4b). This suggests that simultaneous positive and negative ion detection in combination with particle coating may permit unambiguous identification of sulfate in particles that are undetected by conventional laser ablation aerosol mass spectrometry. It should be noted that the use of negative ion or simultaneous positive-negative ion detection without coating does not overcome limitations in the detection of sulfate. “Pure” ammonium sulfate particles have a low ablation efficiency independent of the polarity of ions detected. The application of the 1-naphthyl acetate coating to ambient particle characterization will be the subject of future work. ACKNOWLEDGMENT This research was supported by NSF Grant CHE-9629672 and EPA STAR Grant R82-6769-010. Received for review April 26, 2001. Accepted August 13, 2001. AC010469S
Analytical Chemistry, Vol. 73, No. 22, November 15, 2001
5369
