Anal. Chem. 2003, 75, 5317-5327
Chemical Ionization Mass Spectrometer Instrument for the Measurement of Tropospheric HO2 and RO2 Gavin D. Edwards,† Christopher A. Cantrell,*,† Sherry Stephens,† Brian Hill,†,| Olusegun Goyea,†,‡ Richard E. Shetter,† R. Leon Mauldin, III,† Edward Kosciuch,† David J. Tanner,§ and Fred L. Eisele†,§
Atmospheric Chemistry Division, National Center for Atmospheric Research, Boulder, Colorado 80305, City College of New York (CUNY), New York, New York 10031, and Georgia Institute of Technology, Atlanta, Georgia 30332
Laboratory characterizations of the peroxy radical chemical ionization mass spectrometer (PerCIMS) instrument have been performed. The instrument functions by drawing ambient air through a 50-µm-diameter orifice into an inlet held at low pressure. Peroxy radicals (HO2 and RO2) within this air are detected by amplified chemical conversion into a unique ion (HSO4-) via the chemistry initiated by the addition of NO and SO2 to the inlet. HSO4- ions are then quantified by a quadrupole filter mass spectrometer. PerCIMS provides measurements of the sum of peroxy radicals, HO2 + RO2 (HOxROx mode), or the HO2 component only (HO2 mode), achieved through the control of concentration of NO and SO2 added to the instrument. The characterization and response of this instrument have been evaluated through modeling of inlet chemistry and laboratory experiments and have also been demonstrated through successful deployment during field campaigns. The performance of PerCIMS with respect to calibration pressure and relative humidity is reported, as are the sensitivities of the instrument to organic peroxy radicals with different hydrocarbon groups. These data show PerCIMS to be a practical field instrument for the fast and accurate evaluation of the concentration of peroxy radicals over a variety of atmospheric conditions. The estimated accuracy of the derived [HOxROx] concentrations is (35% (at the 95% confidence interval), while [HO2] measurements have accuracies of (41% (at the 95% confidence interval). Typical precision of measurements well above the detection limit is 10%, and typical detection limits are 1 × 107 radicals cm-3 for 15-s averaging times. Photochemistry is the driving force for almost all the radical chemistry that occurs in the troposphere.1,2 The primary source * To whom correspondence should be addressed. Tel: 303-497-1479. Fax: 303-497-1492. E-mail:
[email protected]. † National Center for Atmospheric Research. ‡ City College of New York. § Georgia Institute of Technology. | Current address: Intel Corp., Jones Farm Campus Bldg. 4, 2111 NE 25th Ave., Hillsboro, OR 97124. (1) Levy, H. Planet. Space Sci. 1973, 21, 575. (2) Crutzen, P. Pure Appl. Geophys. 1974, 106, 1385. 10.1021/ac034402b CCC: $25.00 Published on Web 09/09/2003
© 2003 American Chemical Society
of tropospheric OH radicals is from the reaction of O(1D), produced in the photolysis of ozone in the ultraviolet-B spectral region (290-340 nm), with water vapor, while the oxidative removal of hydrocarbons and CO are the dominant sink processes for this radical;
O3 + hν f O(1D) + O2(1∆,3Σ)
(1)
O(1D) + H2O f OH + OH
(2)
RH + OH f R + H2O
(3)
R + O2 + M f RO2 + M
(4)
CO + OH f CO2 + H
(5)
H + O2 + M f HO2 + M
(6)
Where M in reactions 4 and 6 refers to any collision partner that carries away energy, usually molecular oxygen and nitrogen in the troposphere, and R in reactions 3 and 4 designates an organic carbon fragment. Reactions 4 and 6 produce peroxy radicals, HO2 and RO2. Measurements of [HO2 + RO2], [HO2], or both provide important information about atmospheric processing of hydrocarbons and improve our understanding of the partitioning of peroxy radicals between the hydro and organic peroxy radical forms. In environments where concentrations of NO and NO2 (known collectively as NOx) are low, the dominant sinks for peroxy radicals are via self- and cross reactions which form peroxides.3,4
HO2 + HO2 f H2O2 + O2
(7)
RO2 + HO2 f ROOH + O2
(8)
The self-reaction of RO2 with RO2 does occur, but is much slower than cross-reaction of RO2 with HO2.5 In the presence of high (3) Cantrell, C. A.; Shetter, R. E.; Gilpin, T. M.; Calvert, J. G.; Eisele, F.; Tanner, D. J. J. Geophys. Res. 1996, 101, 14653. (4) Monks, P. S.; Salisbury, G.; Holland, G.; Penkett, S. A.; Ayers, G. P. Atmos. Environ. 2000, 34, 2547. (5) Carpenter, L. C.; Monks, P. S.; Bandy, B. J.; Penkett, S. A.; Galbally, I. E.; Meyer, C. P. J. Geophys. Res. 1997, 102, 25/417.
Analytical Chemistry, Vol. 75, No. 20, October 15, 2003 5317
Table 1. Characteristics of the Peroxy Radical Chemical Ionization Mass Spectrometer description
detail
quantity measured inlet pressure chamber pressure reaction scheme concentration of reagent gases reaction time chain length detection method detection limit precision calibration wall loss signal/background ratio electronic noise cross-calibration accuracy
either [HO2] or [HO2 + RO2] (HOxROx) variable, but typically held at 2 × 104 Pa differentially pumped, ranging from 1.0 to k23[NO]) and the alkoxy radical formed as shown in reaction 20 can react with O2 to form HO2 (reaction 22). Where [NO] is relatively high (i.e., k23[NO] > k22[O2]), the chemical conversion of alkoxy radicals to HO2 must compete with reaction 23. This reaction can, however, be exploited to perform separate measurements of [HO2] and [HO2 + RO2]. This process of conversion of RO2 radicals into HO2 within the PerCIMS inlet is a competition for RO radical between reaction 22, reaction 23, and decomposition or isomerization prior to conversion into HO2.22 Inlet reagent gas conditions can be adjusted in order to favor one reaction pathway over the other, although even in HOxROx measurement mode there is still some potential loss of radicals via reactions 21 and 23 prior to the conversion into HO2. If kdecomp is the effective rate coefficient of these decomposition/isomerization processes, and k(RO+SO2) is the rate of RO reaction with SO2 leading to HO2 (see later), the conversion efficiency of RO2 into HO2 is given by R, where,
(
k20 k20 + k21
RRO2 )
kdecomp
(k
(
){[
23[NO]
)(
k20 k [O ] + k(RO+SO2)[SO2] / k20 + k21 22 2
23[NO]
[k
(k22[O2] + k(RO+SO2)[SO2]) +
)
]
)
+ k22[O2] + k(RO+SO2)[SO2] /
]
+ k22[O2] + k(RO+SO2)[SO2] + kdecomp
}
(24)
The first term on the right-hand side of eq 24 accounts for nitrate formation in the reaction of RO2 radicals with NO (reaction 21), while the second term accounts for nitrite formation in the reaction of RO with NO (reaction 23). Considering the yield of H2SO4 per RO2 sampled, R should be multiplied by the right-hand side of the HO2 yield (eq 19) so that
[H2SO4]t [RO2]0
(
) RRO2
)(
k14[SO2] k17[NO]
1+
)
Ce-Dt - De-Ct G1/2
(25)
where C, D, and G are as defined in eq 19. (22) Tyndall, G. S.; Cox, R. A.; Granier, C.; Lesclaux, R.; Moortgat, G. K.; Pilling, M. J.; Ravishankara, A. R.; Wallington, T. J. J. Geophys. Res. 2001, 106, 12157.
Figure 4. Measured and modeled changes in RCH3O2 (see text) versus [NO]. In the HOxROx mode, inlet [NO] ) 5 ppmv and R is large (green triangle) ([SO2] for these conditions is 300 ppmv). For HO2 mode, inlet [NO] ) 2300 ppmv and R is small (yellow triangle) ([SO2] for these conditions is 50 000 ppmv). Modeling of the PerCIMS inlet chemistry shows that R is influenced by the inclusion of RO + SO2 chemistry. Under conditions of high [NO], R ∼ 0 without the inclusion of RO + SO2 (dashed blue line) while R ∼ 0.15 if RO + SO2 is considered (red solid line). These changes in RCH3O2 are close to the trends observed in experimental data (back circles).
In the “background” troposphere, away from significant anthropogenic and biogenic hydrocarbon sources, the most abundant RO2 radical is the methylperoxy radical.3,9 The relative conversion efficiencies (RRO2) for different inlet NO concentrations for CH3O2 were measured and calculated (Figure 4). At low [NO], the loss of RO from reaction 23 is small, and R approaches unity. For [NO] >1000 ppmv, loss of radicals via reaction 23 is large and RCH3O2 is reduced. Experimental values also plotted in Figure 4 compare favorably to these calculated values. These results suggest that this method of using high/low concentrations of reagent gases in the inlet is effective in enhancing or limiting RRO2. (3) “Background” H2SO4 Measurements. H2SO4 present in the ambient air when drawn into the PerCIMS inlet can also lead to HSO4- ions. PerCIMS measurements must therefore be able to account for ambient [H2SO4] and other inlet artifacts during its measurement cycle. Quantification of these factors is achieved by adding SO2 at different locations of the instrument inlet. In the “background” portion of the measurement cycle, NO is added to the front injector while SO2 is added to the rear injector with the aid of solenoid valves (see Figure 1). In this case, the chemistry proceeds so that OH radicals formed by reactions 11 react with NO as shown in reaction 17 and all ambient peroxy radicals are converted to HONO. Inlet residence times and reagent gas concentrations are such that this chemistry is completed before introduction of SO2 in this “background mode”. Alternatively, when in the “total measurement” portion of the cycle, the valve control allows SO2 to be added at the same location as NO. Then, OH radicals react either with SO2 (reaction 14) or with NO (reaction 17). For typical operation, an equal portion of the measurement cycle is devoted to the measurement for the evaluation of “background”, and “total measurement” signals. The Analytical Chemistry, Vol. 75, No. 20, October 15, 2003
5321
true radical signal is given by total measurement minus background signals. (4) Ion Chemistry. Following the neutral reaction chemistry region of the PerCIMS inlet is the rear portion of the inlet where ion chemistry occurs. Here, a fraction of H2SO4 molecules formed in the neutral region are converted to the HSO4- ion in a manner similar to the ion scheme described by Eisele and Tanner17 and Hanke et al.20 Conversion is achieved through reaction of gasphase H2SO4 with NO3- ions. A mixture of nitric acid vapor in air is allowed to enter the system at a point where neutral chain chemistry is complete and provides a sheath flow within the ion chemistry region as shown in Figure 1. HNO3 within the sheath flow air is ionized by a radioactive source (241Am) to produce NO3ions, which in turn form HSO4- through a proton-transfer reaction:
NO3- + H2SO4 f HNO3 + HSO4-
PerCIMS inlet. Reiner et al.18 discussed other potentially important reactions in their instrument, which is similar to PerCIMS. They concluded that the rates of these side reactions are typically much slower than those of normal inlet chemistry (e.g., peroxide formation within the inlet) or have tropospheric concentration small enough to be insignificant sources of radicals (e.g., peroxyacetyl nitrate (CH3C(O)O2NO2) or peroxynitric acid (HO2NO2) decomposition leading to radical formation). Optimization of inlet conditions such as residence times and reagent gas concentrations for PerCIMS operation is usually sufficient to make these effects negligible. Reiner et al.,18 also showed that wall effects were unimportant. Equation 19 was modified to include the effect of a first-order wall loss, whose rate coefficient is kw:
[H2SO4]t
(26)
[HO2]0 The inlet volume of the ion chemistry region is ∼150 cm3, and the total gas flow is ∼5 SLPM for normal operation conditions leading to a reaction time 0.35 s. This is combined with the neutral chemistry that takes 0.13 s (see above), giving a total reaction time for the conversion of ambient radicals into measurable ions of
