34194

ATMOSPHERIC ELEMENTAL AND ORGANIC CARBON MONITORING WITH HIS-NIR SPECTROSCOPY

Saioa Elcoroaristizabal 1 ; Jose Manuel Amigo 2 ; Jose Antoio García 3 ; Nieves Durana 3 ; Iñaki Elorduy 3 ; Lucio Alonso 3
Download
Favoritar este trabalho

Carbonaceous particulate matter, usually classified into two categories, organic carbon (OC) and elemental carbon (EC), constitutes an important component of the atmospheric aerosol, which have gained importance due to their influence on global warming, visibility degradation, and human health. The OC mixture can be of both primary and secondary origin, comprising compounds like polycyclic aromatic hydrocarbons, known for their potential to induce carcinogenic and mutagenic effects. In contrast, EC is exclusively of primary origin. Therefore, it is essential to know which fraction of EC and OC are emitted to the atmosphere in order to effectively implement mitigation strategies. Usually, the protocol used for the determination of OC and EC involves the destructive measurements of the filters employed for collecting the suspended particles, such as thermal-optical transmission (TOT). Apart from being destructive, TOT is time consuming and it only analyzes a small portion of the filters. Moreover, a fast and accurate measurement of the air condition is needed to explore their dynamics and monitor the air quality. In this sense, near infrared spectroscopy coupled with a hyperspectral detector (HSI-NIR) together with the proper multivariate data analysis comes as alternative.

Therefore, the aim of this work is to join the power of NIR-Hyperspectral Imaging, multivariate data analysis and validation with TOT for developing a strategy of real-time monitoring of OC, EC in atmospheric particulate matter. This can be divided into two different sub-objectives:
- To quantitatively determine the different ratios of organic carbon by using HIS-NIR and compare them with more standard methods.
- To study how OC and EC attach in the filter when they are collected.

PM10 and PM2.5 aerosol sampling was performed using a low-volume sampling system (Derenda LVS3.1 sampler, Berlin, Germany) at a flow rate of 2.3 m3 h-1. Samples of PM10 and PM2.5 were captured onto 47 mm diameter quartz-fiber filters. A total of 12 samples were taken: 2 , 6 and 10 h average samples for PM10 samples; 4, 8 and 12 h average samples for PM2.5 samples. The filters were then analyzed for OC and EC using a laboratory OCEC analyzer (Sunset Laboratory Inc.) with a multistep temperature programme. HIS-NIR measurements were taken using the UmBio inspector (UmBio AS, Sweeden). The pixel resolution was set to 300 μm; while the spectral wavelength range was 1000 – 2700 nm. Principal component analysis (PCA), K-means clustering and Partial Least Squares (PLS) methods applied were implemented in Matlab (v.7.0, The Mathworks, MA, USA).

The PLS models developed for quantification of OC and EC with HIS-NIR demonstrate a strong correlation between the HIS-NIR information and the TOT technique, demonstrating the feasibility of using HIS-NIR for quantitative purposes in air pollution monitoring. Moreover, looking at the individual filters by using PCA and K-means it can be concluded that there are little inhomogeneous areas depending of the filter and the sampling period. This is another important achievement since TOT techniques only analyze a very small part of the filter, having the risk of obtaining a wrong estimation of the carbonaceous matter.