Impacts of air pollution and climate on materials in Athens, Greece

For more than 10 years now the National and Kapodistrian University of Athens,, Greece, contributes to the UN/ECE ICP Materials programme for monitoring of the corrosion/soiling levels of different kind of materials due to environmental air-quality parameters. In this paper we present the results obtained from the analysis of such observational 10 data that were collected in Athens during the period 2003-2012. According to these results the corrosion/soiling of the particular exposed materials tend to decrease over the years, except for the case of copper. Based on this long experimental database applicable to multi-pollutant situation of the Athens basin we present dose response functions (DRFs) considering, that “dose” stands for the air pollutant concentration, “response” for the material mass loss (normally per annum) and the “function” the relationship derived by the best statistical fit to the data. 15

materials' degradation due to atmospheric pollution and climate parameters. The Athens, Greece with significant cultural heritage monuments (UNESCO Cultural Heritage site: Acropolis, Parthenon) has been involved in ICP Materials since 2002 as a targeted field exposure test site, participating also in the EU project MULTI-ASSESS (Model for multi pollutant impact and assessment of threshold levels for cultural heritage: http://www.corr-institute.se/multi-assess/web/page.aspx).
An important contribution to this effort is the development of dose response functions (DRFs) for particular materials. DRFs 5 are relationships between the corrosion or soiling rates and the levels or loads of pollutants in combination with climatic parameters. The corrosion is mainly caused by chemical reactions on the material surface involving air pollutants (e.g., SO 2 , NO x and O 3 ), while soiling is principally depicted as loss of reflectance (Watt et al., 2008). Concerning the latter, the incorporation of PM 10 concentration in the above mentioned relationship allows for the generation of empirical doseresponse functions for soiling (Brimblecombe and Grossi, 2005).

10
The DRFs are used for the assessment of pollution tolerable levels and to recommend target levels to be implemented in the future development of measures on urban air quality in order to minimise the pollution effects on historic and cultural objects. In addition, they can be used in sites where there are no experimental results in order to make estimations of corrosion/soiling rates. According to previous studies implemented in Athens, carbon steel has been proven that is the material which suffers more from corrosion than the others exposed metals/alloys. On the contrary, copper is the most 15 durable (Tzanis et al., 2011). Another study has revealed that the greatest part of the deposited particle mass is not water soluble, while in the water soluble part of it there is an unbalance between the cations and anions with the cations to surpass anions (Tzanis et al., 2009a).
In this study we present the most recent results from the UNECE/ICP Materials trend exposure programme 2011-2012 obtained in Athens, Greece test site, along with the corresponding measurements from previous exposure periods for 20 comparison reasons. We also demonstrate the comparison between experimental results and theoretical corrosion/soiling estimations by employing the newly developed DRFs for the campaigns conducted in Athens, Greece.

Experimental
For the purpose of MULTI-ASSESS and UNECE ICP Materials trend exposure programmes, a station is installed in central Athens, Greece (37°59´57´´ N, 23°43´59´´ E), since 2003. The main rack -field exposure site with exposure samples and the 25 carousel on rack along with sheltered samples enclosed in a box under the rack, for the last exposure period, are shown in Atmos. Chem. Phys. Discuss., doi:10.5194/acp-2016-196, 2016 Manuscript under review for journal Atmos. Chem. Phys. Published: 8 June 2016 c Author(s) 2016. CC-BY 3.0 License.
The withdrawn specimens were sent to the responsible subcentres in Europe (see Table 1) for further analysis and evaluation of soiling or corrosion attack.
In particular, for the determination of multi-pollutant effects on materials, chemical analysis of the specimens was conducted and basic parameters as the weight change, mass loss, surface recession, haze, the total deposited mass of particles per surface unit of glass (TP/S) were calculated. For comparison reasons, as also indicated in Introduction, the corrosion and 5 soiling values for the exposure period 2011-2012 was complemented with the available data collected previously (2003-2004, 2005-2006 and 2008-2009) in the frame of MULTI-ASSESS and UNECE ICP Materials programmes, in which the Athens station has been involved.
In addition, the diffusive passive samplers for the surface air-pollutants (SO 2 , HNO 3 , HCOOH, CH 3 COOH, HCl and HF) measurements and the passive particle collector (aerosols) that were used (shown also in Fig. 1), were prepared at Swedish 10 Environmental Research Institute (IVL). The samplers were mounted under a metal disc ca 2m above the ground in order to protect them from rain and direct sunshine and after the exposure, they were returned to IVL for analysis. The main aim of these measurements was to correlate the pollutants concentrations with the degradation rate of the exposed material specimens.

15
As mentioned before, in order to study the corrosion of structural metals (copper, zinc, carbon and weathering steel), the parameters weight change and mass loss were evaluated. Figures 2-7 present the weight change and mass loss values obtained after the analysis of the exposed specimens. In these figures the experimental results of previous expositions are also presented. It should be mentioned that the presented values are the mean values obtained for the three specimens of each structural metal exposed during the aforementioned exposure periods.

20
The parameter "weight change" describes the difference in specimen's mass after the exposure minus its initial mass. If the specimen was exposed under sheltered conditions this parameter is expected to be positive due to uptake processes (e.g. deposition) and the lack of any mass loss mechanism. In the case of unsheltered exposition, weight change can be positive or negative depending on the balance among uptake and loss mechanisms. According to the results obtained for the case of copper ( Fig. 2), mean weight change of samples exposed during 2011-2012 period is almost 1.5 times greater than that of the 25 samples exposed during 2003-2004(Tidblad et al., 2013. The parameter "mass loss" expresses the difference in specimen's initial mass minus the specimen's mass after removing its corroded part. It should be mentioned here that both the weight change and mass loss parameters are affected by the run-off and the chemical composition of the corrosion layer (Horalek et al., 2005). The experimental results of the mass loss, for copper, zinc and carbon steel, are presented in Figs. 3, 5 and 7, respectively. According to these results, mass loss of copper

5
The carbon and weathering steel arises to be the most sensitive metals, among the exposed ones, to the mass loss, while copper is the most durable. That means that steel is the most sensitive material to the corrosion while copper suffered less by atmospheric corrosion. Considering climate change future projections it is expected an increase in temperature, relative humidity and precipitation (IPCC, 2013) factors which favour corrosion rate. However, corrosion rate is also affected by pollutants levels which generally are decreasing. So the question "how much climate change affects materials corrosion?" 10 needs very careful approach.
In the case of zinc samples, chemical analyses were performed to water solutions of the corrosion products. These solutions were analysed for inorganic acids, formate and acetate. The aim was the identification of corrosive media which affected metal surface. The results can not be used for quantitative analysis but they are useful for qualitative conclusions about the substances which mainly corroded zinc samples (Tidblad et al., 2013). The analysis showed that chloride ions, water-soluble 15 sulphate and nitrates are involved in the corrosion processes of the exposed zinc samples in Athens. No traces of formate and acetate were found.
For the evaluation of corrosion of limestone specimens exposed in unsheltered positions, surface recession, was calculated. This parameter is defined by the formula where W 0 is sample's weight before the exposure, W 1 is sample's weight after the exposure, A is the total surface area of sample and ρ is the density of the limestone. The results of surface 20 recession for the limestone specimens exposed, under unsheltered conditions, for one year are presented in Fig. 8 along with the same results obtained during previous exposure periods. Generally, the recession of limestone has decreased slightly after the period 2005-2006 due possible to the reduced pollution levels. It is also obvious from this figure that recession during last exposure period (2011-2012) is slightly higher than the previous one, perhaps due to a small increase in NO 2 concentration during this period.

25
Another material studied during this exposure period was modern glass. This one is not part of historic and cultural monuments but it is a material which is used widely in synchronous art as well as in other kind of modern constructions. In addition to that, modern glass is also an ideal material for soiling studies because it is transparent, flat, non-porous and chemically inert. Due to these properties modern glass does not affect particles deposition and accumulation (Lombardo et al., 2010).

30
In order to evaluate soiling two parameters are investigated; the total deposited mass of particles per surface unit of glass (TP/S) in μg cm -2 and haze defined as the ratio, expressed in percentage, of the diffuse to direct transmitted light. Modern glass samples were exposed under sheltered conditions during all exposure periods.
Atmos. Chem. Phys. Discuss., doi: 10.5194/acp-2016-196, 2016 Manuscript under review for journal Atmos. Chem. Phys. Greece. The obtained values of these statistical parameters are given in Table 2. For copper and weathering steel the available data were not adequate for developing new DRFs. All the presented below DRFs (Eqs 1, 2, 3, 4, 5,7,8,9)   21%. This range of differences may indicate that for the Athens, Greece case the parameters used in DRF for the modern glass are not sufficient and more experimental data are needed in order to specify the factors which affect haze. In Fig. 16 are presented the percentage contribution of each Athens DRF factor to the total corrosion/soiling of each material for all 25 exposure periods.