文献阅读笔记：Bhave-2002-EST

基于现场的方法确定 ATOFMS 仪器对铵和硝酸盐的敏感性。

Information for the paper

Title: A Field-Based Approach for Determining ATOFMS Instrument Sensitivities to Ammonium and Nitrate

Author: Bhave, Prakash V.

Year: 2002

Journal: Environmental Science & Technology

URL: https://doi.org/10.1021/es015823i

Introduction

1. A commonly cited limitation of single-particle mass spectrometry instruments is that the chemical composition measurements are not quantitative. There are two main obstacles to quantitation. First, the ion signal intensities produced by laser ablation/ionization of nominally identical particles vary greatly from shot-to-shot, primarily because of inhomogeneities in the ablation/ionization laser beam. Second, instrument sensitivities to different aerosol-phase chemical species are largely unknown.

2. In the present work, instrument sensitivity is defined as the ion signal intensity per unit mass of a chemical species, averaged over a particle ensemble.

3. It has been reported that shot-to-shot variations in the ion signal intensities can be mitigated by using very high laser irradiances, but molecular information is lost due to fragmentation of polyatomic ions. For example, laser irradiances > 2 × 10¹⁰ W cm^-2 have been shown to fragment pure ammonium sulfate particles into monatomic N, H, S, and O. To retain molecular information, most single particle mass spectrometry techniques use moderate laser irradiances (10⁷ — 10⁹ W cm^-2).

4. Instrument sensitivities can vary dramatically from one chemical species to another, due to chemically specific differences in ionization efficiency.

5. To date, all efforts to determine instrument sensitivities have been based on particles generated in laboratory environments. These laboratory-generated particles are typically monodisperse, spherical, and have nominally identical chemical compositions.

6. Recent studies revealed that instrument sensitivities can be affected substantially by the size of the individual particle being sampled, trace impurities in the particle matrix, and relative humidity of the background gas.

7. Although a field-based approach for determining instrument sensitivity is appealing, it is subject to three limitations which are not encountered in a laboratory-based approach. First, a field-based approach requires quantitative reference measurements of the chemical species of interest to be taken in parallel with the single-particle measurements, because the chemical composition of an atmospheric aerosol is unknownat the time of sampling. Consequently, the accuracy of instrument sensitivities determined from a field-based approach is limited by the precision of the reference measurements. Second, a field-based approach requires the collection of a much larger number of single-particle spectra than are needed for most laboratory-based approaches, to obtain a statistically significant number of nominally identical particles from the complex mixture of particle types in the atmosphere. Third, particle detection efficiencies of the single-particle instrument must be well characterized to ensure the success of a field-based approach.

Experimental Method

1. Operating principles of the ATOFMS instrument stationed at Riverside during the IOPs are described in detail elsewhere, so only a brief overview is given here.

2. The time difference between scattering pulses is also used to actuate the firing of a high power pulse from a Nd:YAG laser, operating at 266 nm wavelength and 2 × 10⁷ — 4 × 10⁸ W cm^-2 irradiance, upon the particle's arrival in the source region of a time-of- flight mass spectrometer.

ATOFMS Data Treatment

Before ATOFMS and impactor measurements can be compared with each other, the measurements which best represent ATOFMS instrument responses to NH₄⁺ and NO₃^- must be selected, and the ATOFMS data must be corrected for certain sampling biases.

ATOFMS Response Functions

1. To compare ATOFMS data with quantitative measurements of NH₄⁺ and NO₃^-, a measure of the ATOFMS instrument’s response to NH₄⁺ and NO₃^- must be precisely defined.

• Ion signals indicating the presence of NH₄⁺ in an individual particle are detected most often at mass-to-charge (m/z) ratio 18, when sampling Riverside aerosols by ATOFMS. The ion signal at m/z 30 (NO⁺) is an established measure of aerosol nitrate at Riverside. Ion signals at other m/z ratios (e.g. m/z 35 (NH₄NH₃⁺), m/z 46 (NO₂⁺), and m/z 108 (Na₂NO₃⁺)) also indicate the presence of NH₄⁺ and NO₃^- in atmospheric particles.

• In ATOFMS positive ion spectra, the presence of particulate H₂O is typically indicated by a peak at m/z 19 (H₃O⁺) and therefore does not augment the NH₄⁺ signal at m/z 18.

• Ion signals at m/z 18 and 30 are also detected when the ablation/ionization laser fragments certain nitrogen-containing organic compounds. However, NH₄NO₃ typically comprises the largest fraction of fine particle mass sampled at Riverside, so we expect the contributions of fragmented nitrogen-containing organic compounds to the ion signals at m/z 18 and 30 to be negligible relative to the contributions from NH₄⁺ and NO₃^-.

2. Although there are several possible measures of ion signal intensity, absolute area and relative area are the most appropriate for quantification of mass spectrometry data.

• In laboratory ATOFMS studies of nominally identical particles, shot-to-shot variations caused the absolute areas of specific ion signals to vary by an average of 59%. During the same studies, relative areas, defined as the absolute area of the ion signal of interest divided by the total area of the mass spectrum, varied by an average of only 16%. This evidence suggests that relative areas should be used for quantification of ATOFMS data.

• However, when sampling a polydisperse multicomponent aerosol, such as that found in an urban atmosphere, relative area measurements can be affected greatly by the presence of additional chemical species in the particle. For example, the relative area of an ion signal at m/z 18 produced from ablation/ionization of a pure NH₄NO₃ particle will likely be larger than the relative area at m/z 18 measured from an identical particle that also contains a trace amount of potassium, because potassium is efficiently ionized and will therefore increase the total area of the mass spectrum.

• Hence, relative area is not a stable measure of ion signal intensity when determining instrument sensitivities by a field-based approach. Instead, absolute area is selected as the measure of ion signal intensity in the present work.

Results and discussion

1. The ratio of a data point's vertical coordinate to its horizontal coordinate is generally largest for particle ensembles in the 0.32-0.56 μm D_a range and smallest for ensembles in the 1.0-1.8 μm range, for both NH₄⁺ and NO₃^-.

2. This suggests that the ion signal intensity produced by laser ablation/ionization of a unit mass of either species (NH₄⁺ or NO₃^-) decreases as particle aerodynamic diameter increases over the 0.32-1.8 μm range. In other words, ATOFMS instruments are more sensitive to NH₄⁺ and NO₃^- when sampling smaller particles.

3. The increased instrument sensitivity to chemical species in smaller particles is presumed to be due to (1) a greater volume fraction of small particles being vaporized by the ablation/ionization laser relative to larger particles and (2) a lower probability of positive-negative charge recombination in the ablation plume of small particles relative to larger ones.

4. Similar trends have been reported in single particle mass spectrometry analyses of pure, laboratory-generated RbNO₃, (NH₄)₂SO₄, NaCl, and KCl particles, but this is the first such observation in atmospheric aerosol measurements.

Particle Size-Dependent Parametrization of Instrument Sensitivity

5. A result of \(0< \delta_k <3\) might suggest that small particles are completely vaporized, while particles at the upper end of the 0.32-1.8 μm Da range are only partially vaporized ... Hence, we expect best-fit values of NH₄⁺ and NO₃^- to fall in the intermediate range (\(0< \delta_k <3\)) ... When a particle is partially vaporized by the ablation/ionization laser, the ATOFMS instrument is more likely to detect material near the particle surface than material in the particle core.

6. NH₄⁺ and NO₃^- are believed to have similar spatial distributions within the individual particle matrices studied here, because the origin of these two species in Riverside aerosols is largely attributed to the condensation of gas-phase ammonia and nitric acid molecules on the surface of pre-existing particles.

7. Relative sensitivity factors (RSFs) are typically defined on a molar basis and are often used to correct for differences between the instrument sensitivities to two chemical species of interest, when analyzing the composition of a multicomponent sample (Gross, et.al 2000).

\[ RSF \left( \frac{NH_4^+}{NO_3^-} \right) = \frac{18}{62} \cdot \frac{\gamma_{NH_4^+}}{\gamma_{NO_3^-}} \]

8. The RSF of NH₄⁺ versus NO₃^- under sampling conditions is 0.5 and lies in the 0.4-0.7 range with 95% confidence. This implies that ATOFMS measurements of particles containing equimolar concentrations of NH₄⁺ and NO₃^- should yield larger ion signals at m/z 30 than at m/z 18, by a factor of approximately two.

Scaled ATOFMS Measurements of NH₄⁺ and NO₃^-

These high correlation coefficients indicate that the most influential factor governing ATOFMS instrument sensitivities to NH₄⁺ and NO₃^- is particle aerodynamic diameter, under the sampling conditions encountered at Riverside.

Residual Analysis

Influence of Particle Detection Efficiency Corrections

Both subplots show statistically significant positive correlations (R²=0.35 for NH₄⁺ and R²=0.36 for NO₃^-), indicating that approximately 35% of the variance in $ _i $ can be explained by a linear relationship with \(\epsilon _i\). In other words, approximately one-third of the error in the instrument sensitivities to NH₄⁺ and NO₃^- is attributable to uncertainty in the ATOFMS particle detection efficiencies.

Influence of Gas-Phase Properties

1. Although our data set provided no clear evidence that instrument sensitivities to NH₄⁺ and NO₃^- are affected by ambient relative humidity over the 21-69% range, statistically significant negative correlations of \(\epsilon_{ik}\) with ambient temperature (R²=0.14 for NH₄⁺ and R²=0.39 for NO₃^-) indicate that scaled ATOFMS measurements of NH₄⁺ and NO₃^- tend to exceed the corresponding impactor measurements when sampling at high temperatures.

2. This apparent temperature effect is most likely due to the condensation of gas-phase NH₃ and HNO₃ upstream of the ATOFMS instrument, which was stationed in an air conditioned laboratory (T ≈ 22-25 °C) and drawing ambient air through a ~5 m long sampling line at a relatively low flowrate.

3. The effect on ATOFMS NH₄⁺ measurements is less pronounced than on NO₃^- measurements because NH₃ has a higher vapor pressure than HNO₃, making it less likely to condense in the sampling line, and because three of the NH₄⁺ measurements taken during high-temperature IOPs were excluded from the entire analysis, for reasons given above.

4. This suggests that the instrument sensitivity parametrization for NH₄⁺ is stable over the 23-35 °C temperature range.

Influence of Bulk Aerosol Composition

1. Laboratory studies of single-particle mass spectrometry instruments reveal that the presence of certain chemical species in a particle can dramatically affect the instrument response to other species present in the same particle. These phenomena, commonly referred to as matrix effects, have not yet been assessed under ambient sampling conditions.

2. Based on these calculations, we conclude that aerosol chemical composition had an insignificant influence on the sensitivity of ATOFMS instruments to NH₄⁺ and NO₃^-, when averaged over the size-segregated particle ensembles sampled at Riverside.

3. It is important to note that NH₄⁺ and NO₃^- comprise a large fraction of the aerosols studied in this work, so the ATOFMS instrument response to NH₄⁺ and NO₃^- may be less influenced by matrix effects than the instrument response to other species.

Influence of Single-Particle Composition

1. Ion signals at 27 different m/z ratios exhibit statistically significant negative correlations with \(\epsilon_{i,NH_4^+}\), indicating that NH₄⁺ concentrations are overestimated when ion signals at these m/z ratios are abundant.

2. Negative correlationsmayimply that (1) some fraction of the ion signals at m/z 18 resulted from aerosol species other than NH₄⁺ and (2) the presence of other species in the aerosols increased the ionization efficiency of NH₄⁺ (i.e. a matrix effect).

3. Ion signals at eachm/z ratio are uncorrelated with \(\epsilon_{i,NO_3^+}\), indicating that ions at m/z ratios other than 30 do not significantly influence scaled ATOFMS NO₃^- measurements in the Riverside aerosols.

Conclusion

1. The comparison reveals that ATOFMS instrument sensitivities to both NH₄⁺ and NO₃^- decline with increasing particle aerodynamic diameter over a 0.32-1.8 µm calibration range. The strong influence of particle size on instrument sensitivities is an important conclusion of the present work.

2. Aside from particle aerodynamic diameter, few factors significantly influenced the instrument sensitivities to NH₄⁺ and NO₃^-. The second most pronounced influence is attributed to uncertainties in the ATOFMS particle detection efficiency. Sampling artifacts at high ambient temperatures contributed a significant fraction of the variance in \(\epsilon_{i,NO_3^+}\). Finally, a small fraction of the variance in \(\epsilon_{i,NH_4^+}\) may be attributed to interfering ion signals at m/z 18 resulting from the fragmentation of organic amines, and to matrix effects that enhance the ionization of NH₄⁺.

3. Volatilization of NH₄NO₃ from impaction substrates during sampling is favored at high temperatures and low relative humidities and has been shown to result in 7-8% losses of fine particulate nitrate under hot (35°C) and dry (18% relative humidity) conditions.

4. In laboratory-based ATOFMS studies, Angelino et al. discovered that ion signals at m/z 18 are commonly detected when sampling individual particles that contain organic amines.

总结与思考

1. 在 0.32-1.8 µm 粒径范围内，随着颗粒物粒径增大，仪器对 NH₄⁺ (m/z 18) and NO₃^- (m/z 30) 的灵敏度减小。其余影响因素还有仪器检测效率、环境温度（高温气象条件影响 NO₃^- 的响应）、颗粒物化学组分（有机胺物质可能产生 m/z 18）等。
2. 这篇文章发表于2002年，差不多20年了，也不知最新的仪器有没有针对性的改进。
3. 文章用了很多公式，计算过程复杂，不是很好理解。
4. 行文过于啰嗦。对比之下，Rehbein-2001-EST 这篇文章就详略得当、言简意赅。
5. 文章的验证方法有参考价值，但是吧……这作者调门起的太高了点，Introduction 部分恨不得把这方法捧上天……一股浓浓的营销味儿……

扩展阅读

• Liu-2000-AST: Nitrate-Containing Particles

• Fergenson-2001-AC: Quantification of ATOFMS Data

• Gross-2000-AC: Relative Sensitivity Factors for Alkali Metal and Ammonium Cations in ATOFMS

• Reilly-2000-AST: Matrix Effects