Natural radioactivity as an easy and quick parameter for describing the dynamic of the Planetary Boundary Layer

Abstract

This work describes a methodological approach based on natural radioactivity measurements aimed at interpreting air pollution episodes in urban air. The use of such parameters helps in the understanding of the temporal behaviors of seasonal primary (benzene and carbon monoxide) and secondary (nitrogen dioxide and ozone) pollutants. A comparison between the daily concentrations of primary and secondary pollutants and the natural radioactivity trends, considered as an index of the dynamic of the low atmospheric boundary layers, evidences that acute episodes of air pollution in downtown Rome occur in wintertime due to high atmospheric stability (primary pollution) and in summertime because of the strong diurnal atmospheric mixing (secondary pollution).

---

Introduction

The interpretation of atmospheric pollution phenomena is really complex due to the interplay of emission processes and physical–chemical pollutant reactions associated with meteorological processes (turbulence, diffusion, transport).^1–3 Over these problems a further issue is added because the pollutant measurements are carried out at ground level. This offers practical advantages, but creates many problems in interpreting data because the atmospheric processes are dependent on the dynamic evolution of the Planetary Boundary Layer (PBL). Therefore, the identification of parameters capable of describing the PBL behavior and its temporal evolution is of fundamental importance. The meteorological classic observations, normally detected in monitoring networks, provide information on the stability classes, but are not sufficient to characterize the boundary layer. In urban areas, the ground orography and the thermal variations due to anthropogenic activities, make even more difficult the interpretation of these processes. The complexity of such systems can be simplified by the parametrization of the dynamic PBL evolution in order to describe the spatial–temporal evolution of a pollutant. Techniques such as acoustic surveys, radiosonde and analysis of mechanical turbulence with a sonic anemometer, can provide much information but their use is limited by their cost and complexity. On the other hand, the meteorological measurements normally performed in the monitoring networks, do not furnish sufficient information to describe the evolution of the atmospheric low boundary layer. Recently, different research groups have relied on a simple parameter able to precisely describe the pollutant behavior and the dynamic properties of the boundary layer, i.e. the natural beta radiation.^4–12

For a given geographical location and for weeks of observation, however, the emission flux of radon can be considered to be constant and the air concentration of radon and ²²²radon short-lived daughters (²¹⁸Po, ²¹⁴Pb, ²¹⁴Bi and ²¹⁴Po) can be assumed to depend only on the dilution factor.¹³ The dilution properties of the lower atmosphere can, therefore, be characterized by monitoring natural radioactivity due to radon progeny absorbed to atmospheric particles,^5,11 and the dilution properties can be used to analyze primary pollutants pollution.^12,14,15

Radon and its decay products (beta radiation) do not undergo any chemical transformation, then their atmospheric concentrations depend only on the dynamic of the boundary layer. In other words, simple natural radioactivity measurements easily describe the remixing properties of the low boundary layer. For instance, in case of vertical mixing or intense advection radon and its decay products do not accumulate in the low troposphere so that the natural radioactivity assumes low values and is scarcely modulated. On the other hand, in case of atmospheric stability the radon dilution is hampered, consequently high natural radioactivity levels are measured.

The relevance of the natural radioactivity is evident in air quality studies where local or remote emission sources are investigated^16–18 as well as for occurrence of photochemical episodes in atmosphere.^19–21

In this paper primary and secondary air pollution episodes in Rome are discussed and interpreted by means of natural radioactivity as a tracer of PBL dynamic.

Experimental part

Sampling site

Rome, located in the Mediterranean area with almost 3.0 million inhabitants (21 m a.s.l.; 41°53′35′′ N and 12°28′58′′ E), is characterized by different meteorological and anthropogenic conditions such as typical local wind from the sea and presence of intensive human activities (i.e., autovehicular traffic, about 2.5 million among cars, motorcycles and bus, source from Automobil Club Italia; domestic heating; industrial emissions), respectively. As regards natural source (marine aerosol and mass transport coming from the Sahara desert, desert dust) it is necessary to specify that these episodes in Rome are not as rare as believed.²² Saharan dust was advected over Rome on about 30% of the days of 2001; mean contribution of Saharan dust transport events to daily PM10 levels was of the order of 20 mg m⁻³.²³

Two monitoring stations were located in Rome, one at INAIL building (40 m-height) in downtown Rome close Piazza Venezia, the second inside the green park, a location not directly influenced by city emission flux and representative of a homogeneous and background pollution.^9,24–28 The first site is located in an area characterized by high traffic density, in a two-lane street whose aspect ratio H/W (H: building height, W: street width) is about 3: the traffic density during the experimental analysis was equal to more than 70 vehicles min⁻¹ (considering cars, pullman, buses and motorbikes) which is typical of a large street in downtown Rome.

Instrumentation

Natural radioactivity was assessed using a PBL Mixing Monitor (FAI Instruments, Fonte Nuova, Italy). The instrument samples atmospheric PM on 47 mm membrane filters: a Geiger detector measures the β-radioactivity of short-lived decay products of radon with 1 h time resolution.^15,18 The instrument performs sampling of atmospheric particulate matter over an interval of 60 minutes and then measures its beta activity. The characteristics of the instrument, which operates on two membranes, one in sampling phase while the other in measurement phase, make possible to obtain 24 average values of natural radioactivity, each relating to 1 hour-interval. The instrument is equipped with a series of automatic quality checks on both sampling and measurement phase: in this way it manages to ensure a good accuracy of the results obtained.

A Benzene–Toluene–Xylene (BTX, mod. 955, Syntech Spectra, The Netherlands) was used for continuous off-line measurements of aromatic hydrocarbon concentrations with 15 min time intervals. The GC was equipped with a capillary column AT-5 (13 m × 0.32 mm ID) and with a Photo Ionization Detector (PID).

The total mass of particulate matter in atmosphere was measured continuously with an analyzer provided by an oscillating microbalance (mod. TEOM, Rupprecht & Patashnick Co Inc., Albany, NY, USA), as the granulometric splitting up the analyzer was equipped with an inertial impact separation system PM10 (R&P).

The separation and quantification of Elemental Carbon (EC) and Organic Carbon (OC) were carried out by means an Ambient Carbon Particulate Monitor (Rupprecht & Patashnick Co Inc., NY). By means a non-dispersive infrared detector (NDIR) the instrument measures the CO₂ amount released when a PM sample collected in a collector is oxidized at elevated temperatures. The instrument cycle is made up of two parts: the collection phase during which the sample is gathered in a collector, and the analysis phase during which the entire collector with its collected particulate matter is elevated in temperature to perform oxidation. The conditions for the collection phase were a collection period of one hour and a collection temperature at the collector of 50 °C. The sampling flow rate is 16.7 L min⁻¹. The temperature of the collector is then raised to 340 °C for a period of 13 min during which the instrument measures the CO₂ concentrations in the analysis loop: in this condition the instrument measures the OC concentration. Then, a final burn of 8 min at 750 °C takes place to burn off the high-temperature carbon that was not oxidized at 340 °C setting for measuring the total carbon (TC) concentration in the sample. Finally, EC is calculated as difference between TC and OC.

A Differential Optical Absorption Spectrometer (DOAS, Opsis, Sweden) has been used for the measurement of the gaseous pollutant. The DOAS's analytical method is based on the absorption of light in the near UV and IR regions of the pollutants with fine vibrational structures.^29–32 The Lambert–Beer's law regulates the relationship between the intensity of absorbed light and the compound concentration. The DOAS system consists of an emitter (a xenon lamp at high pressure), a receiver, a spectrophotometer equipped with an optical fiber and a computer for the system management (data elaboration and data storage). The absorption spectra of each monitored chemical species are acquired at their relative typical wavelength ranges; subsequently, the interferences are eliminated by comparison with the reference spectra. The distance between emitter and receiver is about 280 m: this parameter is important because it influences the sensitivity of the measures. The absorbance of light from the emitter is continuously measured within the wavelength range 240–350 nm to determine several compounds. The aromatic hydrocarbons are detected in the wavelength range between 250 and 290 nm where the major interfering gases are oxygen, ozone and sulphur dioxide; around 100 spectra per second are collected in this wavelength range and stored in a register with 1000 channels with a resolution of better than 0.05 nm. Spectra are required on an average time of 7 min for the system located in downtown Rome. The specification provided by the instrument manufacturers are: (i) minimum detectable concentrations over 280 m = 5 μg m⁻³; (ii) zero point stability +10 μg m⁻³ per month; (iii) linearity +10 μg m⁻³ in the range 0–100 μg m⁻³. The DOAS system allows measuring the average integrated concentrations of SO₂, NO₂, O₃, HNO₂, HCHO, benzene and toluene. It should be mentioned that these measurements are related to a defined portion of atmospheric environment (from one hundred meter to one kilometer).

Results and discussion

Typical radon trends

Fig. 1 shows a typical daily trend of the natural radioactivity in a cold period (January) in downtown Rome: it shows inhomogeneous structures that are characteristic of a winter period. The comparison with Fig. 2 describing the natural radioactivity trend in a hot period (June), shows that during summer, period usually characterized by persistent high pressure conditions, the natural radioactivity has a well-modulated daily behavior with little differences between days.

Fig. 1 Typical daily trend of the natural radioactivity in a cold period.

Fig. 2 Typical daily trend of the natural radioactivity in a hot period.

The night atmospheric stability conditions and the daytime convective mixing occurring during warmer months allow radon accumulation during the night and dispersion during daylight hours. On the contrary, in winter high pressure periods are infrequent while advection episodes often occur (pollutants undergo horizontal transport due to the wind speed): during these latter episodes the natural radioactivity assumes a scarce-modulated trend and the values remains consistently low.

The time interval during which natural radioactivity levels are minimum and then the dilution capacity is maximum (mixing window) is related to the duration and intensity of the solar radiation (convective remixing) whereas its amplitude varies over the year. Fig. 3 shows the temporal trends of the natural radioactivity in the two different (winter and summer) periods characterized by high pressure.

Fig. 3 Comparison between temporal remixing “windows” in cold/hot periods.

The comparison between the two trends clearly shows the different mixing windows: in summer the window runs from 8 until 22, whereas in winter it is limited between 10 and 18. It should also be noted that the natural radioactivity values are extremely low at daytime in summer (great mixing of the PBL), while in winter the minimum values are less pronounced (low height of the PBL during the day). In fact, in summer the strong convective drive due to the intense solar radiation has the effect of increasing the PBL height, whereas in winter such effect is limited.

Radon emission on large scale

Radon emission rate varies from one place to another according to soil composition, moisture content, porosity and permeability, but the variations can be considered to be negligible in a time scale of some days and a space scale of some kilometers.¹⁶ Fig. 4 shows the simultaneous trends of natural radioactivity in three Italian locations distant up to more than 600 km: as it can be seen the trends are almost similar. It means that the radon behavior is spatially almost constant for long distance, the only difference regards the quantitative levels of the radon accumulation depending on the orographic conditions of each territory.

Fig. 4 Simultaneous radon concentration trends in two Italian locations (500 km-distance).

Radon as tracer of the Planetary Boundary Layer

The study of the radon concentrations as tracer of the dynamic properties of the boundary layer helps to understand the dynamic of pollutant behavior. A very clear example is reported in Fig. 5: it is reported a typical occurrence of natural and anthropogenic events recorded in few consecutive days in Rome. Two different pollution episodes can be evaluated due to different sources. The first episode is from 14^th to 16^th of February: an atmospheric stability condition causes high PM10 levels due essentially to a fine PM event (anthropogenic sources).³³ On the other hand, during the 21^st of February there are strong mixing conditions due to a coarse event (Saharan dust): the natural sources are the only responsible for the almost 90 μg m⁻³ of PM10 (Fig. 6). The first episode could be explained by the high anthropogenic sources present in the area whereas the second episode is not easy to justify. The natural radioactivity modulations of the two periods help to understand the reasons of these occurrences. During the first episode, from 14^th to 16^th, the high radon concentration values means steady atmospheric conditions (no atmospheric remixing): this condition is favorable for pollutant accumulation (i.e., pollutants from anthropogenic activities). On 21^st the radon concentration values are low: apparently, this could be a good circumstance for pollutant dispersion (e.g., wind, remixing) but, on the contrary, PM10 levels remain still high meaning the presence of some natural sources in the investigated area. In this latter case it can be supposed the influence of the long-range mass transport from the desert confirmed also by the high peaks of coarse fraction (Fig. 6), also recorded in close location by other author²² can be studied.

Fig. 5 Typical anthropogenic and natural events of PM10 concentration values (above) in downtown Rome investigated throughout the modulation of natural radioactivity (below).

Fig. 6 Highlights of the period reported in Fig. 5 (bold fine fraction, line coarse fraction).

A similar situation occurred in other period (December) (Fig. 7): the interpretation is still the same. Also in this case two different episodes can be identified and evaluated in terms of anthropogenic (17^th–21^st) and natural (27^th) events.

Fig. 7 Trends of daily PM10 (a), two different granulometric fractions (b) and natural radioactivity concentrations (c) during December.

PBL dynamic in presence of primary/secondary pollutants

The temporal evolution of the radon concentration and its partial derivative depend on the dynamics of the boundary layer and can be formally defined by the following equation:¹⁵where α is a parameter that links the PBL properties with the emissive source intensity, β{C_R} is a term taking into account the properties of the atmosphere mixing, ϕ_R is the emission flux of radon, Adν is the advection term. From the analysis of the radon concentration and its derivative it is possible to obtain information on α and β, then recognize both the stability/instability conditions and transition phases.

The concentration temporal variation of primary pollutants at low reactivity, i.e. those directly emitted from the sources and/or those that do not undergo chemical–physical relevant processes (e.g., CO, benzene), can be formally described by the following equation:

The eqn (2) differs from eqn (1) only for the term ϕ that takes into account the emission flow of the pollutant investigated. The availability of atmospheric dispersion capacity measure (α and β) through the natural radioactivity determination, allows identifying changes in primary air pollutant concentrations (due, for instance, to changes in emission flows, e.g. increasing of motor vehicle traffic or regulations for limiting traffic).

For secondary pollutants at high reactivity such as NO₂, eqn (2) is:

where ϕ_NO2 is negligible, P_NO2 is related to formation processes and L_NO2 to removal processes and L_s to removal rate. As an example, it is evident that understanding the NO₂ trend is a really complex issue because it requires the knowledge of all the processes involved.^34,35

The NO₂ formation steps are:

NO + O₃ = NO₂ (4)

NO + RO₂ = NO₂ + RO (5)

NO + HO₂ = NO₂ + OH (6)

The removal processes are:

NO₂ + hν = NO + O (7)

NO₂ + OH = HNO₃ (8)

NO₃ + NO₂ = N₂O₃ (9)

The process of removing surface is:

2NO₂ + H₂O = HNO₂ + HNO₃ (10)

The main production and removal NO₂ processes directly involve O₃. Similar considerations can be made for those ozone processes directly involving NO₂.

A similar equation formally describes the temporal O₃ evolution:

where P_O3 = L_NO2 due to NO₂ photolysis and P_NO2 = L_O3 due to titration reaction. Assuming O_x variable as sum¹⁵ of NO₂ + O₃, the eqn (11) is:where L′_s is the removal rate related to NO₂ and O₃.

During strongly advective conditions the O_x partial derivative with respect to time can be considered constant, and consequently O_x is a constant (NO₂ + O₃ = K). In fact, during instability conditions the radical processes are negligible and the reactions between NO₂ and O₃ are dominant and complementary. On the other hand, during atmospheric stability conditions the O_x variable shows a well-defined trend due to both the presence of radical oxidative processes and the dynamic properties of the low boundary layer. During the night, in presence of high atmospheric stability the O_x derivative is approximately 0 and its value at ground coincides almost with NO₂ values. At sunrise O_x increases rapidly, reaches its maximum value around 10–11 a.m. and decreases in the afternoons up to minimum levels at night.

Gaseous primary pollutants

The concentration temporal evolution of primary air pollutants at low reactivity such as benzene and CO, depends exclusively on the pollutant amount emitted instantly into atmosphere and the atmospheric dispersion capacity.

In wintertime the primary pollution events reach the maximum intensity; in this period the mixing window is shorter and the diurnal atmospheric convective mixing occurs in the late morning and significantly drops in the early afternoon. Time periods characterized by intense traffic emission coincide with atmospheric stability periods (between 6 a.m. and 9 a.m. and between 16 p.m. and 21 p.m.) causing a clear increase in primary pollutant concentrations. Fig. 8 shows the temporal trends of benzene and CO in a cold period along with the natural radioactivity concentrations: it could be noted that the atmospheric diffusion processes are preponderant, rather than the traffic emission. During atmospheric instability conditions the benzene and CO trends show no regular behavior. In days characterized by atmospheric stability they show a modulation with peaks in the morning and in the evening. This analysis has shown the existence of a very close relationship between the atmospheric dynamic properties and the primary pollutant concentrations such as benzene and CO. During the summer period, the primary pollution events are less intense than in winter. In fact, in summer the mixing window is maximum, the convective mixing starts from the early hours in the morning and operates until late in the afternoon. Consequently, the temporal overlap between maximum emission fluxes and atmospheric stability is reduced, therefore the primary air pollutant concentration reaches values on average lower than those recorded during winter period.

Fig. 8 Typical daily trends of CO and benzene (above) and natural radioactivity (below) in January.

Fig. 9 shows the temporal trends of benzene and CO in a hot period. Basically, benzene and CO levels are lower than in winter period. The cause of such lower primary pollution during summertime has to therefore be found in the wider atmospheric mixing window and greater PBL mixing height (Fig. 3).

Fig. 9 Typical daily trends of CO and benzene in September.

Gaseous secondary pollutants

For the secondary pollution the study of temporal NO₂ and O₃ trends is of considerable importance: these two pollutants are simultaneously the main products of oxidative processes and precursors of OH radicals. NO₂ is essentially a secondary pollutant resulting from NO oxidation by O₃ and HO₂ and RO₂ radicals. The ozone formation process is the reaction between O₂ and atomic oxygen provided by NO₂ photolysis.

If the reaction of NO to NO₂ was due solely to reaction with ozone, oxidative cycle would reach a steady state and there would not be ozone accumulation. In polluted atmosphere, in presence of reactive hydrocarbons (i.e., VOCs) and OH radicals, radicals such as HO₂ and RO₂ are formed and rapidly oxidize NO. It is thus triggered a process that leads to ozone accumulation.

In order to interpret this complex phenomenon of secondary pollution, and therefore to assess the contribution of radical oxidative processes in NO₂ and O₃ concentration variations, the use of the variable O_x is of fundamental importance. Indeed in absence of radical oxidative processes, in condition of strong mixing (advection) the O_x variable trend assumes a extremely simple shape with a slight modulation. The NO₂ and O₃ trends are complementary, i.e. one is the specular image of the other. Such behavior is shown in Fig. 10 where daily trends of NO₂ and O₃ (a), O_x (b) and natural radioactivity (c) are reported in wintertime.

Fig. 10 Daily trends of NO₂ and O₃ (a), O_x (b) and natural radioactivity (c) in November.

In high-pressure conditions on synoptic scale and low wind speed, the mechanical advective transport processes are negligible and therefore the radical oxidative processes are predominant. Fig. 11 shows the trends of both NO₂ and O₃ (a), O_x (b) and natural radioactivity (c) related to a summer period.

Fig. 11 Daily trends of NO₂ and O₃ (a), O_x (b) and natural radioactivity (c) in August.

In presence of radical processes the NO₂ and O₃ trends are no longer the specular image, the variable O_x assumes a behavior with higher daily modulation and minimum and maximum more pronounced. During the day the presence of oxidative processes is highlighted by the fact that the O_x value quickly exceeds its background value. This allows easily to put in evidence the presence of radical oxidative phenomena.

Particulate matter pollutant

The approach based on natural radioactivity can be also successfully applied to the particulate matter pollutant issue. An interesting application regards the carbonaceous fraction, a major PM component, which is constituted by primary and secondary components. In order to understand its temporal behavior we have applied the method based on the identification of key variables: this allows to disclose total carbon (TC)^36–38 concentration variations due to both sources/sinks and vertical/horizontal mass transport. According eqn (1), the temporal TC evolution at ground level can be formally described by the following equation:where C is mixing ratio near the ground, α stability term, ϕ_(t) primary emission flux, β{C} vertical mass exchange due to eddy diffusion, Adν advection term, L_s removal rate from dry deposition. The term Adν can be deduced from knowledge of the wind intensity and direction at ground level. The time trend of radon daughters concentration can be used in order to characterize the terms α and β.

Fig. 12a shows the temporal trends of TC and Total Suspended Particulate (TSP) whereas Fig. 12c the trend of radon concentration. It can be seen that the trends of TC and TSP are very similar and that they have the same modulation as radon concentration.

Fig. 12 Temporal TC and TSP (a), TC and benzene (b) and natural radioactivity (c) trends.

Fig. 12b shows temporal trends of benzene in gas phase and total carbon particles. Benzene is a primary low-reactive pollutant. The trends of these two primary pollutants are similar and they have the same modulation of radon concentration (Fig. 12c). The quantitative TC/benzene ratios vary according the meteorological conditions. In fact, the temporal trend of the ratio TC/benzene (Fig. 13a) is not constant but it shows a pattern depending on the term L_s, eqn (13). The term L_s, which takes into account the deposition losses, is very important for carbonaceous compounds associated to particulate matter, expecially during unstable periods, in which turbulent deposition is very effective, whereas the term L_s for benzene is generally negligible. If we compare the derivative of radon concentration with TC/benzene trend (Fig. 13b), it appears that high values of the ratio occur when stability condition are present (December 8^th–10^th), instead minimum values occur during unstable periods (December 5^th–7^th and 10^th). The stability and instability conditions are shown from the shape of the radon time derivative trend.

Fig. 13 Daily TC/benzene (a) and radon derivative (b) trends.

Conclusions

The measurement of the natural radioactivity as tracer of the PBL dynamic allows to understand the complex (gaseous and particulate) phenomena in atmosphere in simple, accurate and rapid way. The ability of both “following” only one parameter and correlating pollution phenomena to the atmosphere dynamic is very useful for a fast interpretation of what occurring in atmosphere; basically, the air quality models, usually involved in such study, require various weather parameters simultaneously and therefore of dedicated, complex and expensive software.

Finally, the variable O_x, sum of O₃ and NO₂, helps the understanding of the oxidative processes occurring in atmosphere: also in this issue the study of the natural radioactivity gives an important help for the knowledge of the phenomena occurring in atmosphere.

Acknowledgements

This research was performed under the grant INAIL/P20L01.

References

1. D. Pasquil, Meteorol. Mag., 1961, 90, 33.
2. F. Pasquill and F. B. Smith, Atmospheric Diffusion, Ellis Horwood Ltd., John Wiley & Sons, Chichester, 3^rd edn, 1983, p. 437.
3. M. Cassiani, A. Stohl and S. Eckhardt, Atmos. Chem. Phys., 2013, 13, 9975.
4. D. J. Jacob, M. J. Prather, P. J. Rasch, R.-L. Shia, Y. J. Balkanski, S. R. Beagley, D. J. Bergmann, W. T. Blackshear, M. Brown, M. Chiba, M. P. Chipperfield, J. de Grandpré, J. E. Dignon, J. Feichter, C. Genthon, W. L. Grose, P. S. Kasibhatla, I. Kohler, M. A. Kritz, K. Law, J. E. Penner, M. Ramonet, C. E. Reeves, D. A. Rotman, D. Z. Stockwell, P. F. J. Van Velthoven, G. Verver, O. Wild, H. Yang and P. Zimmennan, J. Geophys. Res., 1997, 102, 5953.
5. C. Perrino, A. Pietrodangelo and A. Febo, Atmos. Environ., 2001, 35, 5235.
6. A. Pasini and F. Ameli, Geophys. Res. Lett., 2003, 30, 1386.
7. L. Sesana, E. Caprioli and G. M. Marcazzan, J. Environ. Radioact., 2003, 65, 147.
8. L. Sesana, B. Ottobrini, G. Polla and U. Facchini, J. Environ. Radioact., 2006, 86, 271.
9. C. Perrino, M. Catrambone and A. Pietrodangelo, Environ. Int., 2008, 34, 621.
10. C. Papastefanou, Aerosol Air Qual. Res., 2009, 9, 385.
11. Z. Zhang, F. Wang, F. Costabile, I. Allegrini, F. Liu and W. Hong, Environ. Sci. Pollut. Res., 2012, 19, 3421.
12. F. Wang, Z. Zhang, M. P. Ancora, X. Deng and H. Zhang, Sci. World J., 2013, 626989.
13. J. E. Pearson and G. E. Jones, J. Geophys. Res., 1965, 70, 5279–5285.
14. I. Allegrini, A. Febo, A. Pasini and S. Schiarini, J. Geophys. Res., 1994, 99, 18765.
15. A. Febo, C. Perrino, C. Giliberti and I. Allegrini, in Urban Air Pollution: Monitoring and Control Strategies, ed. I. Allegrini and F. De Santis, NATO ASI Series, Springer-Verlag, Berlin, 1996, pp. 295–317.
16. A. Febo, F. Guglielmi, M. Manigrasso, V. Ciambottini and P. Avino, Atmos. Pollut. Res., 2010, 1, 141.
17. M. Manigrasso, A. Febo, F. Guglielmi, V. Ciambottini and P. Avino, Environ. Pollut., 2012, 170, 43.
18. P. Avino, D. Brocco, L. Lepore and S. Pareti, Ann. Chim., 2003, 93, 589.
19. R. M. Harrison, M. Dall'Osto, D. C. S. Beddows, A. J. Thorpe, W. J. Bloss, J. D. Allan, H. Coe, J. R. Dorsey, M. Gallagher, C. Martin, J. Whitehead, P. I. Williams, R. L. Jones, J. M. Langridge, A. K. Benton, S. M. Ball, B. Langford, C. N. Hewitt, B. Davison, D. Martin, K. F. Petersson, S. J. Henshaw, I. R. White, D. E. Shallcross, J. F. Barlow, T. Dunbar, F. Davies, E. Nemitz, G. J. Phillips, C. Helfter, C. F. Di Marco and S. Smith, Atmos. Chem. Phys., 2012, 12, 3065.
20. A. L. Robinson, N. M. Donahue, M. K. Shrivastava, E. A. Weitkamp, A. M. Sage, A. P. Grieshop, T. E. Lane, J. R. Pierce and S. N. Pandis, Science, 2007, 315, 1259.
21. L. D. Yee, J. S. Craven, C. L. Loza, K. A. Schilling, N. L. Ng, M. R. Canagaratna, P. J. Ziemann, R. C. Flagan and J. H. Seinfeld, Atmos. Chem. Phys., 2013, 13, 11121.
22. A. Febo, Proc. Accademia Lincei “Ecosistema Roma”, Rome, 14–16 April, 2005, pp. 69–72.
23. G. P. Gobbi, F. Barnaba and L. Ammannato, Atmos. Environ., 2007, 41, 261.
24. P. Avino, D. Brocco, L. Lepore and I. Ventrone, J. Aerosol Sci., 2000, 31, S364.
25. A. Monod, B. C. Sive, P. Avino, T. Chen, D. R. Blake and F. S. Rowland, Atmos. Environ., 2001, 35, 135.
26. P. Avino, D. Brocco, A. Cecinato, L. Lepore and C. Balducci, Ann. Chim., 2002, 92, 333.
27. P. Avino, G. Capannesi and A. Rosada, Toxicol. Environ. Chem., 2006, 88, 633.
28. P. Avino, G. Capannesi and A. Rosada, Microchem. J., 2008, 88, 97.
29. U. Platt and D. Perner, J. Geophys. Res., 1980, 85, 7453.
30. U. Platt, in Air monitoring by spectroscopic techniques, ed. M. W. Sigrist, Chemical Analysis Series 127, John Wiley & Sons, Inc., 1994, pp. 27–84, ISBN: 0-471-55875-3.
31. R. Volkamer, T. Etzkorn, A. Geyer and U. Platt, Atmos. Environ., 1998, 32, 3731.
32. P. Avino and M. Manigrasso, Atmos. Environ., 2008, 42, 4138.
33. D. Ji, Y. Wang, L. Wang, L. Chen, B. Hub, G. Tang, J. Xin, T. Song, T. Wen, Y. Sun, Y. Pan and Z. Liu, Atmos. Environ., 2012, 50, 338.
34. R. Atkinson, Atmos. Environ., 2000, 34, 2063.
35. K. Acker, A. Febo, S. Trickc, C. Perrino, P. Bruno, P. Wiesend, D. Möller, W. Wieprecht, R. Auel, M. Giusto, A. Geyer, U. Platt and I. Allegrini, Atmos. Environ., 2006, 40, 3123.
36. P. Avino, D. Brocco and L. Lepore, Anal. Lett., 2001, 34, 967.
37. X. Querol, A. Alastuey, M. Viana, T. Moreno, C. Reche, M. C. Minguillón, A. Ripoll, M. Pandolfi, F. Amato, A. Karanasiou, N. Pérez, J. Pey, M. Cusack, R. Vázquez, F. Plana, M. Dall'Osto, J. de la Rosa, A. Sánchez de la Campa, R. Fernández-Camacho, S. Rodríguez, C. Pio, L. Alados-Arboledas, G. Titos, B. Artíñano, P. Salvador, S. García Dos Santos and R. Fernández Patier, Atmos. Chem. Phys., 2013, 13, 6185.
38. P. Avino, M. Manigrasso, A. Rosada and A. Dodaro, Environ. Sci.: Processes Impacts, 2015, 17, 300.

---