Determination of carbon dioxide concentration in the atmosphere from absorption spectra

135 JOURNAL OF SCIENCE OF HNUE DOI: 10.18173/2354-1059.2017-0041 Mathematical and Physical Sci. 2017, Vol. 62, Iss. 8, pp. 135-141 This paper is available online at DETERMINATION OF CARBON DIOXIDE CONCENTRATION IN THE ATMOSPHERE FROM ABSORPTION SPECTRA Nguyen Thi Thuy, Nguyen Thi Huyen Trang and Ngo Ngoc Hoa Faculty of Physics, Hanoi National University of Education Abstract. Carbon dioxide (CO2) gas is one of the main greenhouse gases (following water vapor) and has a significant impact on the Earth’s climate. Determation of greenhouse gas concentration (of CO2 in particular) will improve our understanding of the Earth’s climate changes. In this study, we use absorption spectra to determine the concentration of CO2 in the atmosphere. The absorption spectra of CO2 corresponding to three transitions in the mid- infrared region were recorded at room temperature using a diode laser system. Ambient air was charged into the absorption cell (White-type) with an optical path of 24 to 40 m and total pressure of 4 Torr to 300 Torr. The absorption spectra were analyzed using a line by line process and the Voigt profile. Spectroscopic parameters (line intensity) were used with the 2012 HITRAN databases to determine the concentration of CO2 in the laboratory atmosphere. The results show that the concentration of CO2 in the atmosphere exceeded 400 ppm. Keywords: Atmosphere, intensity, concentration of carbon dioxide, voigt profile. 1. Introduction The rapid development of industrialization and mechanization has resulted in non-stop air pollution. High concentrations of pollutants such as CO2, NO2, hydrocarbons cause significant health and the environmental problems. Of these carbon dioxide is one of the most important elements in the Earth’s atmosphere, the second most prevalent greenhouse gas following water vapor in the atmosphere, and therefore, it contributes greatly to planet warming and has a strong impact on global climate change [1, 2]. As an inevitable consequence of human activities (combustion of fossil fuels), the amount of CO2 in the atmosphere has rapidly increased since the beginning of the industrial revolution. According to annual statistical data, the concentration of CO2 increased about 3 parts per million (ppm) between (February 2014, 397 ppm and February 2015, 400 ppm) [3]. The current global atmospheric CO2 level is greater than 400 ppm [4, 5], which is a problem of global concern. The absorption or emission spectra of CO2 can give information about the characteristics of this molecule. For example the reversion of absorption spectra shows the concentration of CO2 in the atmosphere. These factors have promoted the study and determination of the concentration of CO2 in the atmosphere from studies of CO2 spectra in the lab, such as Fourier transform interferometer Received May 3, 2017. Accepted August 15, 2017. Contact Ngo Ngoc Hoa, email: hoa.nn@hnue.edu.vn Nguyen Thi Thuy, Nguyen Thi Huyen Trang and Ngo Ngoc Hoa 136 [6, 7], laser diode system [8], CRDS system [9] or by satellite for remote sensing methods OCO (The Orbit Carbon Observatory) [10]. To determie the CO2 concentration in the atmosphere, we need to know the spectroscopic parameters (position line, line intensity, line broadening and line-shape). To describe the experimental line shape, the Voigt profile is used in most available studies, the Voigt profile being a convolution of two broadening mechanism, one of which produces a Gaussian profile (usually, as a result of Doppler broadening), the other would producing a Lorentz profile (as a result of pressure broadening). When performing the calculations with Voigt profile, the Doppler width being generally fixed to its theoretical value. In this paper, we use absorption spectra to determine the concentration of CO2 in the laboratory atmosphere. It is organized into 4 sections as follows: introduction, theoretical foundations and experimental process, results and discussion, and conclusions and future studies. 2. Content 2.1. Theoretical foundations 2.1.1. Integrated intensity Absorption spectra are used as an analytical method for determination of certain substances present in a sample. There are many experimental techniques for measuring absorption spectra. The most common method is direct measurement by comparing the intensity of radiation before I0 (ω) and after IL (ω) through the sample. The intensity IL (ω) of the beam after through the sample with optical path L depends on the interaction between matter and radiation. The relation between I(ω) and Io(ω) follows the Beer-Lambert law [11, 12]. -αL o I =e , I (1) where, α is the absorption coefficient and L is the optical path. The absorption coefficient can be written as [11, 13]:  oα=β×f ω-ω , (2) where, is a normalization function describing the line-shape spectra:   + o 0 f ω-ω dω=1,   (3) Integrating both sides of the equation (1), we have:   + + o 0 0 αdω= f ω-ω ×βdω=β,     (4) Thus, β called the integrated intensity depends on many factors such as pressure of activated molecules, optical path and transition. To make application easier, integrated intensity is generally calculated for one unit concentration and one unit optical path and is called the standard integrated intensity. Standard integrated intensities (S) for many kinds of molecules and many transitions are stocked in HITRAN (https://www.cfa.harvard.edu/hitran/). The HITRAN database 2012 contains spectroscopic parameters for thousand of transitions and for 34 different kinds of molecules [14]. Standard integrate intensity is determined by equation [13]: β S= , L×n (5) Determination of carbon dioxide concentration in the atmosphere from absorption spectra 137 If the integrated intensity β is determined by (cm-1), the concentration n of activated molecules is in molecule/cm 3 , optical path L is in cm then standard integrated intensity S will be in cm -1 molec -1 cm 2 [14]. 2.1.2. Line shape We then take into account both broadening mechanisms, Doppler and Lorentzian broadening, with the hypothesis that the two mechanisms are completely independent (but not taking into account the Dicke narrowing effect and the speed dependence of the collisional parameters due to our experimental uncertainties). The corresponding absorption coefficients for an isolated line are given as [11]:     VP o D D β ln2 α σ Re W σ-σ ,Γ ,Γ , Γπ  (6) where 0 and D are the unperturbed spectral position of the transition and the Doppler width,  is the collisional half-width (HWHM). The complex probability function W is given by:     2-t o D o D D i e W σ-σ ,Γ ,Γ dt π ln2 ln2 σ-σ -t+iΓ Γ Γ     (7) And (in cm -1 ) is given by expressions [11]: 7 D o T Γ 3.58 10 σ , M   (8) where 0 is number wave of unperturbed spectral position of transition (cm -1 ), M is molar mass (g/mol) and T is temperature (K). In this work, we used the Voigt profile to fit the absorption spectra of CO2, the integrated intensity and the standard integrated intensity in HITRAN database 2012 [14] to determine the concentration of CO2 in atmosphere. 2.2. Measurement procedure The absorption spectra of CO2 corresponding to three transitions in the mid-infrared region were measured at room temperature by using a diode laser system described in detail in [15, 13]. Ambient air was charged into the absorption cell (white-type) with an optical path of 20 to 40 m at a total pressure from 4 to 300 Torr. The obtained spectra are fitted line by line using the Voigt profile. Three transitions and the experimental conditions are shown in Table 1. Table 1. Experimental conditions for measurement of CO2 absorption spectra Transition (cm -1 ) [14] Standard integrated intensity (cm -1 .molec - 1 .cm 2 ) [12] Total pressure (Torr) Optical length (cm) Temperature (K) 2251.681043 cm -1 1,050.10 -20 4 – 50 4000 295 2257.508986 cm -1 1,876.10 -20 3 – 250 2400 295 2261.277760 cm -1 2,507.10 -20 4 – 200 2400 295 Nguyen Thi Thuy, Nguyen Thi Huyen Trang and Ngo Ngoc Hoa 138 2.3. Results and discussion 2.3.1. The fitting results of the experimental spectra The fitting results of experimental spectra are given by Figure 1. Figure (1a) and (1b) respectively show results for transition at wave number 2251.681043 cm -1 and 2257.508986 cm -1 . The lowest panels are ten times the difference between the experimental spectra and the fitting results using Voigt profile. We used a straight line for the base line, half width Doppler broadening is fixed to its theoretical values calculated from equation (8), we have determined the spectroscopic parameters for each transition: peak position, half-width Lorentzian, integrated intensity together with two parameters describing the base line. 0.0 0.5 1.0 1.5 0.0 0.5 1.0 1.5 2.0 -0.06 -0.03 0.00 0.03 0.06 -0.3 0.0 0.3 -0.04 -0.02 0.00 0.02 0.04 -0.3 0.0 0.3 60.530 Torr 42.510 Torr 20.510 Torr 15.150 Torr 10.624 Torr (b)(a) A b s o rp ti o n 50.000 Torr 45.040 Torr 35.060 Torr 17.980 Torr 10.058 Torr 1 0 x (o b s -c a lc ) Relative Wavenumber (cm -1 ) Figure 1. The absorption spectrum of atmospheric CO2 at room temperature measured different pressures corresponding to two transitions at wave number 2251.681043 cm -1 (a) and 2257.508986 cm -1 (b) The lowest panels are ten times the difference between experimental spectra and fitting results using Voigt function The fitting results show that the difference between the experimental spectra and calculations is clearly higher than the noise levels of our measurements. The W form of residues (figure 1(a)) shows the narrowing effect of the experimental lines. In other words, the Voigt profile neglects the other kinetic process which may change line-shape as the Dicke narrowing effect [11, 16] or speed dependence of collisional parameters [17, 18]. In figure (1b), we also observed two transitions near the studied transition. In the fitting process, we did not calculate these two transitions. Therefore the results of integrated intensity, concentration of gas may be affected by these near transitions. Determination of carbon dioxide concentration in the atmosphere from absorption spectra 139 2.3.2. Results determine concentration CO2 in atmosphere From formula (5): B r βk Tβ S= , L×n L×P  (9) where P is a particular pressure of CO2 in atmosphere, kB is the Boltzmann constant. With mr is the molecular ratio of total CO2 molecules in the atmosphere, formula (9) can be written as: r B L.m β=S P, k T (10) P representing total pressure. Therefore, when T is constant, the integrated intensity is linearly proportional to total pressure. Figure 2. Integrated intensity is expressed as total pressure to transition at wave number2251.681043 cm -1 (a) and 2261.277760 cm -1 (b) Figure 2 describes the dependence of integrated intensity on the total pressure for two transitions at wave number 2251.681043 cm -1 and 2261.277760 cm -1 . The results show a linear dependence of integrated intensity via the total pressure. After the linear fitting process, from the slope together with the parameters S, L, T shows in Table 1, we have determined the molecular ratio of total CO2 molecules in the atmosphere corresponding to three transitions, the results shown in Table 2. Table 2. The molecules ratio were determined from three transitions in this work Transition Molecule ratio mr(ppm) 2251.681043 cm -1 4,74.10 -6 cm -1 /Pa 459 2257.508986 cm -1 4,22.10 -6 cm -1 /Pa 381 2261.277760 cm -1 6,38.10 -6 cm -1 /Pa 432 Average 420  30 Nguyen Thi Thuy, Nguyen Thi Huyen Trang and Ngo Ngoc Hoa 140 The results in Table 2 show that the ratio of CO2 molecules determined from the transition at 2257.508986 cm -1 has the smallest value and differs with the results of the other two transitions. This may be due to the influence of two near transitions of which we have not taken into account in the fitting process for this transition. The average value of CO2 molecules ratio in the atmosphere is higher than the value from [3]. This result may be because, the values of standard integrated intensity S from [14] is usually smaller than the value determined from diode laser system by about 5% [13] and therefore the value of mr will be higher than the real value by about 5%. 3. Conclusions In this work, we used absorption spectra of three different transitions of O 16 C 12 O 16 in the atmosphere in a laboratory, the absorption spectra recorded at room temperature to determine CO2 molecules ratio in the atmosphere. The results show that this value was about 420 ppm. The fitting result from experimental spectra corresponding to the transition at 2257.508986 cm -1 may be affected by two near transitions. Therefore, in future study, we will take into account these transitions. The results also show that there were other molecular kinetic mechanisms affecting the line- shape spectrum. Therefore, in future studies, we will use different models that take into account molecular dynamic mechanisms such as: Dicke narrow effect and the speed dependence of collisional parameters to describe refined line-shape absorption spectra. Acknowledgments. The authors are pleased to acknowledge the financial support of this research by the Vietnam Ministry of Education & Training for project No. B2016-SPH-21. REFERENCES [1] Donald M. Hunten, 1993. Atmospheric evolution of the terrestrial planets. Science, 259 915. [2] I. Pouchet, V. Zéninari, B. Parvitte, G. Durry, 2004. Diode laser spectroscopy of CO2 in the 1.6μm region for the in situ sensing of the middle atmosphere. Journal of Quantitative Spectroscopy and Radiative Transfer 83 619. [3] [4] NOAA, Record annual increase of carbon dioxide observed at Mauna Loa for 2015 [5] NIES, Whole-atmosphere monthly mean CO2 concentration based on GOSAT observations - Recent data- [6] A. Predoi-Cross, A.V. Unni, W. Liu, I. Schofield, C. Holladay, A.R.W. McKellar, D. Hurtmans, 2007. Line shape parameters measurement and computations for self-broadened carbon dioxide transitions in the 30012  00001 and 30013  00001 bands, line mixing, and speed dependence. Journal of Molecular Spectroscopy 245 34. [7] N.H. Ngo, X. Landsheere, E. Pangui, S.B. Morales, J.-M. Hartmann, 2014. Self- broadening of 16 O 12 C 16 O3-band lines. Journal of Molecular Spectroscopy 306 33. [8] I. Pouchet, V. Zéninari, B. Parvitte, G. Durry, 2004. Diode laser spectroscopy of CO2 in the 1.6 μm region for the in situ sensing of the middle atmosphere, Journal of Quantitative Spectroscopy and Radiative Transfer 83 619. [9] B.V. Perevalov, A. Campargue, B. Gao, S. Kassi, S.A.Tashkun, V.I. Perevalov, 2008. New CW-CRDS measurements and global modeling of 12 C 16 O2 absolute line intensities in the 1.6 m region. Journal of Molecular Spectroscopy 252 190. Determination of carbon dioxide concentration in the atmosphere from absorption spectra 141 [10] Z. Kuang, et al., 2002. Spaceborne measurements of atmospheric CO2 by highresolution NIR spectrometry of reflected sunlight: an introductory study Geophys. Res. Lett. 29 1716. [11] J. M. Hartmann, C. Boulet and D. Robert, 2008. Collisional effects on molecular spectra.Laboratory experiments and models, consequences for applications, Elsevier, Amsterdam. [12] A.P. Thorne, U. Litzén, and S. Johansson, 1999. Spectrophysics: Principles and Applications. Springer. [13] N. H. Ngo, N. Ibrahim, X. Landsheere, H. Tran, P. Chelin, M. Schwell, and J. M. Hartmann, 2012. Intensities and shapes of H2O lines in the near-infrared by tunable diode laser spectroscopy. Journal of Quantitative Spectroscopy and Radiative Transfer 113 870. [14] Rothman LS et al., 2013. The HITRAN 2012 molecular spectroscopic database. Journal of Quantitative Spectroscopy and Radiative Transfer 130 4. [15] N. Ibrahim, P. Chelin, J. Orphal, Y. I. Baranov, 2008. Line parameters of H2O around 0.8 m studied by tuneable diode laser. Journal of Quantitative Spectroscopy and Radiative Transfer 109 2523-2536. [16] H. Ngo, H. Tran, and R. R. Gamache, 2012. A pure H2O isolated line-shape model based on classical molecular dynamics simulations of velocity changes and semi-classical calculations of speed-dependent collisional parameters. J. Chem. Phys. 136 154310. [17] F. Rohart, H. Mader and H. W. Nicolaisen, 1994. Speed dependence of rotational relaxation induced by foreign gas collisions: Studies on CH3F by millimeter wave coherent transients. The Journal of Chemical Physics 101 6475-6486. [18] H. Tran, D. Bermejo, J.-L.Domenech, P. Joubert, R. R. Gamache, and J.-M.Hartmann, 2007. Collisional parameters of H2O lines: Velocity effects on the line-shape. Journal of Quantitative Spectroscopy & Radiative Transfer 108 126-145.