Monthly Update

Atmospheric Chemistry Program



Volume 8, Number 4, Summer 1997

Mexico City Peroxyacyl Nitrate and Hydrocarbon Measurements

(Contribution by J. S. Gaffney, N. A. Marley, and P. V. Doskey)

A collaborative experimental effort on air quality in Mexico City was carried out during February and March 1997. The Mexican effort was led by the Instituto Mexicano de Petróleo (IMP); meteorological measurements were added by Argonne National Laboratory (ANL], Los Alamos National Laboratory, Pacific Northwest National Laboratory, and the Environmental Technological Laboratory of the National Oceanic and Atmospheric Administration; and an extensive aerosol sampling campaign was directed by the Desert Research Institute. Here, some surface measurements of peroxyacyl nitrates (PANs, RCO3NO2) and hydrocarbons are described.

The PANs measured were peroxyacetyl nitrate (PAN, R=CH3- ), peroxypropionyl nitrate (PPN, R=CH3CH2- ), and peroxybutyryl nitrate (PBN, R=CH3CH2CH2-). Air samples were analyzed at 30-min intervals from 21 February through 23 March 1997 by use of an automated gas chromatograph equipped with an electron capture detector. The samples were collected at IMP, which is located in the northwestern-central portion of Mexico City. Over 1300 analyses were taken for PAN, which was found as expected to be the predominant member of the PANs observed in Mexico City air.

PANs are formed from the photooxidation of organics in the presence of nitrogen dioxide. Peroxyacyl radicals (RCO3), which are formed primarily from the reaction of OH and ozone with a variety of volatile organics, react with nitrogen dioxide to form PANs. PANs can thermally decompose back to the RCO3 radical and nitrogen dioxide. The chemical equilibrium that is key to the formation and loss of PANs under the conditions at Mexico City is given by the following reaction:

RCO3 + NO2 <--> RCO3NO2 (PANs) [1]
PAN levels were found to exceed 30 ppb on five of the days and exceeded 10 ppb on all but a few days of the study. PPN levels were typically 10% of the PAN levels with 3-5 ppb maximum concentrations observed on a number of days. PBN levels reached 1 ppb and were much more sporadic in their behavior. PANs were found to show a strong diurnal patt ern, consistent with their daytime photochemical formation form the photooxidation of organics. For most of the period sampled there appeared to be a rapid loss of PANs in the late afternoon, with minimal PANs observed during the nighttime periods.

Importance and Chemical Implications of Some Preliminary Results

These data represent the first comprehensive measurements of PANs in the Mexico City urban airshed. The levels of PANs observed were quite large and were comparable to those observed during the 1970s in the Los Angeles Air Basin. Variations in the concentration of PANs correlated well with ozone changes and were consistent with the coupled chemistry of PANs and ozone formation. The PAN levels approached 50% of the NO2 observed during the early to mid afternoon of 25 February 1997. Multiple peaks in the PAN concentrations and the rapid rise in the morning indicated a very rapid formation rate associated with reactive hydrocarbons, primarily aldehydes and alkenes. A rapid loss was usually observed in late afternoon.

As shown in equation (1), PANs are formed from the oxidation of organics and essentially are a trapped peroxy radical. During this study, Argonne National Laboratory (ANL) obtained cannister samples at the IMP site for subsequent analysis for hydrocarbons. They will be examined to evaluate the formation potential for PANs.

The observed rapid loss of PANs can be due to dilution by mixing or from chemical reactions, homogeneous or heterogeneous. Peroxy radicals produced from the gas phase homogeneous thermal decomposition of PANs and the subsequent reactions with NO can lead to the formation of OH radicals and NO2. The chemical pathways involved are given for the case of PAN (R=CH3- ) in the following reaction sequence:

CH3CO3NO2 (PAN) --> CH3CO3 + NO2 [2]
CH3CO3 + NO --> CH3CO2 + NO2 [3]
CH3CO2 + O2 --> CH3O2 + CO2 [4]
CH3O2 + NO --> CH3O + NO2 [5]
CH3O + O2 --> CH2O + HO2 [6]
HO2 + NO --> OH + NO2 [7]
CH2O + OH --> CHO + H2O [8]
CHO + O2 --> CO + HO2 [9]
HO2 + NO --> OH + NO2 [10]

Thus, decomposition of PANs in the presence of fresh NO can lead to the direct formation of OH radicals and conversion of NO to NO2, from solely thermal reactions. This process leads to increased ozone from the photolysis of NO2 to form ozone:

NO2 + hv --> NO + O --> O + O2 + M --> O3 [11]

The importance of these reactions indicate that PANs are an indicator of the ozone forming potential and of the organic oxidizing capacity of the urban air mass.

The OH radicals initiate oxidation of the larger organics and aromatics found in Mexico City to form secondary organic aerosols. The data indicate that a substantial amount of organic aerosols including nitrophenols and nitro-PAH are likely to be formed in the Mexico City air from these oxidation reactions. As well, the data for PANs will be us eful in assessing secondary inorganic nitrate aerosol formation because the two following reactions will be moderated by the PANs as they are a reservoir for NO2 (see equation 1).

OH + NO2 --> HNO3 (gas) [12]
HNO3 + NH3 --> NH4NO3 (aerosol) [13]

Continuing Research

A fairly wide variety of air chemical and meteorological conditions occurred during the study. For example, the pollutant gas concentrations observed and the meteorological, visibility, and aerosol data gathered by other study participants all suggest that the air above Mexico City was relatively clean on 6 March. For many of the days sampled, the PANs observed at the IMP site had multiple maxima, which suggests poor vertical mixing and return flow of polluted air to source areas. Examination of both meteorological and air chemical data will be necessary in the research to be conducted by the various participants in the study.

The hydrocarbon data indicate that mixing and transport during late afternoon could be responsible for up to 50% of the decrease seen then in PAN concentrations. Preliminary estimates for the loss by reaction with NO leave approximately 30% of the loss unaccounted for. This additional loss might be due to heterogeneous reactions of PANs on fine carbonaceous and inorganic aerosols present in the Mexico City air. Laboratory studies performed at Argonne National Laboratory have shown that PANs can react very rapidly with carbon soot and other surfaces. Data on fine aerosol nitrate will be examined and compared to gaseous nitric acid and PAN data to evaluate the relative importance of nitr ic acid and PAN in secondary nitrate aerosol formation.

Ozone and PANs might possibly be involved in the oxidation of SO2 to aerosol sulfate under the conditions in Mexico City. Sulfate and SO2 concentrations will be examined with PAN data for correlations to explore this pathway for fine aerosol formation. 

The Monthly Update is an informal newsletter, distributed for the
dissemination of information to Atmospheric Chemistry Program
scientists and their colleagues.  Submissions to the Monthly Update
are welcome and should be mailed to

M.L. Wesely or R. Spencer
Building 203, ER
Argonne National Laboratory
Argonne, IL  60439
e-mail: mlwesely@anl.gov
630/252-5827