Volume 33 Issue 1
Apr.  2020
Turn off MathJax
Article Contents

Bo Fang, Na-na Yang, Chun-hui Wang, Wei-xiong Zhao, Xue-zhe Xu, Yang Zhang, Wei-jun Zhang. Detection of Nitric Oxide with Faraday Rotation Spectroscopy at 5.33 μm[J]. Chinese Journal of Chemical Physics , 2020, 33(1): 37-42. doi: 10.1063/1674-0068/cjcp1910182
Citation: Bo Fang, Na-na Yang, Chun-hui Wang, Wei-xiong Zhao, Xue-zhe Xu, Yang Zhang, Wei-jun Zhang. Detection of Nitric Oxide with Faraday Rotation Spectroscopy at 5.33 μm[J]. Chinese Journal of Chemical Physics , 2020, 33(1): 37-42. doi: 10.1063/1674-0068/cjcp1910182

Detection of Nitric Oxide with Faraday Rotation Spectroscopy at 5.33 μm

doi: 10.1063/1674-0068/cjcp1910182
More Information
  • Corresponding author: Wei-xiong Zhao, E-mail: wxzhao@aiofm.ac.cn; Wei-jun Zhang, E-mail: wjzhang@aiofm.ac.cn
  • Part of the special topic on "The 3rd Asian Workshop on Molecular Spectroscopy"
  • Received Date: 2019-10-21
  • Accepted Date: 2019-11-13
  • Publish Date: 2020-02-27
  • We report the development of a static magnetic field Faraday rotation spectrometer for NO detection. A 5.33 μm continuous-wave quantum cascade laser was used as the probing laser. Line absorption at 1875.81 cm$^{-1}$ ($^2\Pi_{3/2}$Q(3/2), $v$=1$\leftarrow$0) was chosen for the detection. By using a Chernin type multipass cell, a detection precision of 1.15 ppbv (1$\sigma$, 1s) was achieved with an absorption pathlength of 108 m. This value was reduced to 0.43 ppbv by increasing the data-acquisition time to 15 s.
  • Part of the special topic on "The 3rd Asian Workshop on Molecular Spectroscopy"
  • 加载中
  • [1] S. W. Kim, B. C. McDonald, S. Baidar, S. S. Brown, B. Dube, R. A. Ferrare, G. J. Frost, R. A. Harley, J. S. Holloway, H. J. Lee, S. A. McKeen, J. A. Neuman, J. B. Nowak, H. Oetjen, I. Ortega, I. B. Pollack, J. M. Roberts, T. B. Ryerson, A. J. Scarino, C. J. Senff, R. Thalman, M. Trainer, R. Volkamer, N. Wagner, R. A. Washenfelder, E. Waxman, and C. J. Young, J. Geophys. Res. 121, 1340 (2016)
    [2] P. J. Crutzen, Ann. Rev. Earth Planet. Sci. 7, 443 (1979). doi:  10.1146/annurev.ea.07.050179.002303
    [3] U. Platt, Phys. Chem. Chem. Phys. 1, 5409 (1999). doi:  10.1039/a906810d
    [4] J. Wojtas, A. Gluszek, A. Hudzikowski, and F. K. Tittel, Sensors 17, 1 (2017). doi:  10.1109/JSEN.2017.2761499
    [5] G. Dooly, C. Fitzpatrick, and E. Lewis, Sensor. Actuat. B 134, 317 (2008). doi:  10.1016/j.snb.2008.05.011
    [6] Y. Yu, N. P. Sanchez, F. Yi, C. Zheng, W. Ye, H. Wu, R. J. Griffin, and F. K. Tittel, Appl. Phys. B 123, 1 (2017).
    [7] M. R. McCurdy, Y. Bakhirkin, G. Wysocki, and F. K. Tittel, J. Biomed. Opt. 12, 034034-1 (2007).
    [8] I. Ventrillard, P. Gorrotxategi-Carbajo, and D. Romanini, Appl. Phys. B 123, 1 (2017).
    [9] S. Zhou, Y. Han, and B. Li, Appl. Phys. B 124, 1 (2018).
    [10] R. Lewicki, J. H. Doty Ⅲ, R. F. Curl, F. K. Tittel, and G. Wysocki, PNAS 106, 12587 (2009). doi:  10.1073/pnas.0906291106
    [11] G. Litfin, C. R. Pollock, R. F. Curl, and F. K. Tittel, J. Chem. Phys. 72, 6602 (1980). doi:  10.1063/1.439117
    [12] A. Hinz, J. Pfeiffer, W. Bohle, and W. Urban, Mol. Phys. 45, 1131 (1982). doi:  10.1080/00268978200100871
    [13] M. C. McCarthy, J. C. Bloch, and R. W. Field, J. Chem. Phys. 100, 6331 (1994). doi:  10.1063/1.467095
    [14] L. Lathdavong, J. Westberg, J. Shao, C. M. Dion, P. Kluczynski, S. Lundqvist, and O. Axner, Appl. Opt. 49, 5597 (2010). doi:  10.1364/AO.49.005597
    [15] R. J. Brecha, L. M. Pedrotti, and D. Krause, J. Opt. Soc. Am. B 14, 1921 (1997). doi:  10.1364/JOSAB.14.001921
    [16] H. Adams, D. Reinert, P. Kalkert, and W. Urban, Appl. Phys. B 34, 179 (1984).
    [17] M. Havenith, M. Schneider, W. Bohle, and W. Urban, Mol. Phys. 72, 1149 (1991). doi:  10.1080/00268979100100821
    [18] T. A. Blake, C. Chackerian, Jr., and J. R. Podolske, Appl. Opt. 35, 973 (1996). doi:  10.1364/AO.35.000973
    [19] H. Ganser, M. Horstjann, C. V. Suschek, P. Hering, and M. Mürtz, Appl. Phys. B 78, 513 (2004).
    [20] T. Fritsch, M. Horstjann, D. Halmer, Sabana, P. Hering, and M. Mürtz, Appl. Phys. B 93, 713 (2008). doi:  10.1007/s00340-008-3223-z
    [21] K. Wu, Y. Feng, J. Li, G. Yu, L. Liu, Y. Xiong, and F. Li, Opt. Express 25, 30916 (2017). doi:  10.1364/OE.25.030916
    [22] G. Wysocki, R. Lewicki, X. Huang, R. F. Curl, and F. K. Tittel, Proc. of SPIE 7222, 72220M-1 (2009). doi:  10.1117/12.823724
    [23] J. Shao, L. Lathdavong, J. Westberg, P. Kluczynski, S. Lundqvist, and O. Axner, Appl. Opt. 49, 5614 (2010). doi:  10.1364/AO.49.005614
    [24] P. Kluczynski, S. Lundqvist, J. Westberg, and O. Axner, Appl. Phys. B 103, 451 (2011). doi:  10.1007/s00340-010-4336-8
    [25] Y. Wang, M. Nikodem, J. Hoyne, and G. Wysocki, Proc. of SPIE 8268, 82682F-1 (2012).
    [26] S. M. Cristescu, D. Marchenko, J. Mandon, K. Hebelstrup, G. W. Griffith, L. A. J. Mur, and F. J. M. Harren, Appl. Phys. B 110, 203 (2013). doi:  10.1007/s00340-012-5050-5
    [27] Y. Wang, M. Nikodem, and G. Wysocki, Opt. Express 21, 740 (2013). doi:  10.1364/OE.21.000740
    [28] J. Westberg and O. Axner, Appl. Phys. B 116, 467 (2014). doi:  10.1007/s00340-013-5721-x
    [29] J. Westberg, J. Wang, and O. Axner, J. Quant. Spectrosc. Ra. 133, 244 (2014). doi:  10.1016/j.jqsrt.2013.08.008
    [30] E. N. Chesnokov, V. V. Kubarev, P. V. Koshlyakov, Y. V. Getmanov, and O. A. Shevchenko, Chem. Phys. Lett. 636, 203 (2015). doi:  10.1016/j.cplett.2015.07.043
    [31] E. Zhang, S. Huang, Q. Ji, M. Silvernagel, Y. Wang, B. Ward, D. Sigman, and G. Wysocki, Sensors 15, 25992 (2015). doi:  10.3390/s151025992
    [32] M. Gianella, T. H. P. Pinto, X. Wu, and G. A. D. Ritchie, J. Chem. Phys. 147, 054201-1 (2017).
    [33] K. Wu, L. Liu, Z. Chen, X. Cheng, Y. Yang, and F. Li, J. Phys. B: At. Mol. Opt. Phys. 50, 035401-1 (2017).
    [34] Y. Wang, M. Nikodem, E. Zhang, F. Cikach, J. Barnes, S. Comhair, R. A. Dweik, C. Kao, and G. Wysocki, Sci. Rep. 5, 1 (2015).
    [35] S. G. So, E. Jeng, and G. Wysocki, Appl. Phys. B 102, 279 (2011). doi:  10.1007/s00340-010-4002-1
    [36] W. Zhao, L. Deng, X. Xu, W. Chen, X. Gao, W. Huang, and W. Zhang, Imaging and Applied Optics JTu4A.31 (2014).
    [37] W. Zhao, G. Wysocki, W. Chen, E. Fertein, D. L. Coq, D. Petitprez, and W. Zhang, Opt. Express 19, 2493 (2011). doi:  10.1364/OE.19.002493
    [38] W. Zhao, B. Fang, X. Lin, Y. Gai, W. Zhang, W. Chen, Z. Chen, H. Zhang, and W. Chen, Anal. Chem. 90, 3958 (2018). doi:  10.1021/acs.analchem.7b04949
    [39] D. R. Glowacki, A. Goddard, and P. W. Seakins, Appl. Opt. 46, 7872 (2007). doi:  10.1364/AO.46.007872
    [40] D. R. Glowacki, A. Goddard, K. Hemavibool, T. L. Malkin, R. Commane, F. Anderson, W. J. Bloss, D. E. Heard, T. Ingham, M. J. Pilling, and P. W. Seakins, Atmos. Chem. Phys. 7, 5371 (2007). doi:  10.5194/acp-7-5371-2007
    [41] X. Yang, W. Zhao, L. Tao, X. Gao, and W. Zhang, Acta Phys. Sin-Ch. Ed. 59, 5554 (2010).
  • 加载中
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Figures(7)  / Tables(1)

Article Metrics

Article views(73) PDF downloads(43) Cited by()

Proportional views
Related

Detection of Nitric Oxide with Faraday Rotation Spectroscopy at 5.33 μm

doi: 10.1063/1674-0068/cjcp1910182

Abstract: We report the development of a static magnetic field Faraday rotation spectrometer for NO detection. A 5.33 μm continuous-wave quantum cascade laser was used as the probing laser. Line absorption at 1875.81 cm$^{-1}$ ($^2\Pi_{3/2}$Q(3/2), $v$=1$\leftarrow$0) was chosen for the detection. By using a Chernin type multipass cell, a detection precision of 1.15 ppbv (1$\sigma$, 1s) was achieved with an absorption pathlength of 108 m. This value was reduced to 0.43 ppbv by increasing the data-acquisition time to 15 s.

Part of the special topic on "The 3rd Asian Workshop on Molecular Spectroscopy"
Bo Fang, Na-na Yang, Chun-hui Wang, Wei-xiong Zhao, Xue-zhe Xu, Yang Zhang, Wei-jun Zhang. Detection of Nitric Oxide with Faraday Rotation Spectroscopy at 5.33 μm[J]. Chinese Journal of Chemical Physics , 2020, 33(1): 37-42. doi: 10.1063/1674-0068/cjcp1910182
Citation: Bo Fang, Na-na Yang, Chun-hui Wang, Wei-xiong Zhao, Xue-zhe Xu, Yang Zhang, Wei-jun Zhang. Detection of Nitric Oxide with Faraday Rotation Spectroscopy at 5.33 μm[J]. Chinese Journal of Chemical Physics , 2020, 33(1): 37-42. doi: 10.1063/1674-0068/cjcp1910182
  • Atmospheric nitric oxide (NO) is an important compound of atmospheric reactive nitrogen. It is mainly formed in the combustion process carried out by human activities [1], and plays important roles in controlling the photochemical production of ozone (O$_3$), in determining the concentration of the hydroxyl radical (OH$\cdot$), and in contributing to the formation of secondary organic aerosols (SOA) as well as acid precipitation [2, 3]. High-precision measurement of ambient NO is thus critical for environmental pollution and atmospheric chemistry research.

    Laser spectroscopy technique provides an attractive and powerful chemical-free tool for NO measurement with high time resolution and good precision. A wide range of laser spectroscopy methods, such as quartz-enhanced photoacoustic spectroscopy [4], differential optical absorption spectroscopy [5], tunable diode laser absorption spectroscopy [6], cavity based techniques (integrated cavity output spectroscopy [7], cavity-enhanced absorption spectroscopy [8], cavity ring down spectroscopy [9]), and Faraday rotation spectroscopy (FRS) [10], have been used for NO measurement. The reported detection precision ranged from several tens of pptv (parts per trillion by volume) to several ppbv (parts per billion by volume) levels.

    FRS is a spectroscopic technique that relies on the magneto-optic effect (Zeeman split) for paramagnetic species [11-13]. The background signals from the absorption of diamagnetic compunds are largely suppressed [14], thus providing a useful method for high sensitive measurement of NO [10, 14-34]. There are two ways to modulate the Zeeman splitting of the absorption line: (ⅰ) an alternating magnetic field (AC-field) produces varying magnetic circular birefringence; (ⅱ) a static magnetic field (DC-field) combined with wavelength modulation spectroscopy (WMS) of the laser frequency to effectively vary the magnetic circular birefringence [35]. The FRS signal is then generated from the demodulation of the modulated magnetic circular birefringence with a phase-sensitive lock-in amplifier. Compared with AC-FRS method, DC-FRS method provides an alternative FRS scheme with excellent performance. The use of permanent magnet instead of AC magnetic coil has potential to reduce the power consumption, and the low frequency noise can be reduced by using high frequency modulation (in AC-FRS method, the demodulation frequency is usually limited by the resonant frequency of RLC circule) [36].

    In this work, DC-FRS technique was studied for NO detection at 5.33 µm (1875.81 cm$^{-1}$, $^2\Pi_{3/2}$Q(3/2), $v$=1$\leftarrow$0) with a continuous-wave (CW) quantum cascade laser (QCL). A Chernin type multipass cell was used to increase the absorption pathlength. Performance evaluation is presented and discussed.

  • For weak absorption and small rotation angle ($\varphi$, the angle offset the cross position between two polarizers), the laser power ($P$) emerging from the analyzer can be expressed as [11]:

    where $P_0$ is the laser incident power, $L$ is the absorption path length inside the magnetic field, and $R_\Delta$ is the differential between the refraction indices ($n$) of the medium for the right- (RHCP, +) and left-handed (LHCP, $-$) circularly polarized light. $R_\Delta$=$k_0$($n_+$$-$$n_-$), where $k_0$ is the wave vector, responsible for the FRS signal ($F$):

    $R_\Delta$ contains line shape information, which can be calculated as the sum of all allowed Zeeman sublevel transition components (as shown in FIG. 1) [37]:

    Figure 1.  Zeeman splitting energy pattern for Q$_{3/2}$(3/2) of NO molecule. The $\Delta M_J$=$\pm$1 transitions are indicated by arrows

    where $N$ is the molecule concentration in molecule/cm$^3$, $S$ is the line absorption intensity in cm$^{-1}\cdot$molecule$^{-1}\cdot$cm$^{2}$, $\gamma_\textrm{D}$ is the Doppler width (HWHM) in cm$^{-1}$, $M_J'$ and $M_J''$ are the magnetic quantum numbers for the upper and lower states, and $Z(z)$ is the plasma dispersion function with $z$=$x$+$iy$:

    where $\gamma_\textrm{C}$ is the collisional width in cm$^{-1}$, $\nu$ is the laser frequency, and $\nu_{M'M''}$ is the line center frequency in the magnetic field.

    The total noise can be expressed as the following by adding the noise term with the extinction ratio ($\xi$) term of the analyzer [18]:

    which is a function of the offset angle $\varphi$ [37]:

    Among these noise sources, the detector noise $N_0$ is independent of $\varphi$. For small $\varphi$, the shot noise $N_1$($\sin^2\varphi$+$\xi$)$^{1/2}$ is proportional to $\varphi$, and the laser noise $N_2$($\sin^2\varphi$+$\xi$) is proportional to $\varphi^2$. As the FRS signal is proportional to $\varphi$, there is a maximum signal-to-noise ratio (SNR) at an optimal rotation angle $\varphi_{\rm{opt}}$ depending on the contribution of each noise source.

  • The schematic diagram and the corresponding photograph of the experimental setup are shown in FIG. 2. A 5.33 µm room temperature continuous-wave (CW) quantum cascade laser (QCL, Thorlabs), controlled by a laser diode controller (ITC4002QCL, Thorlabs), was used for probing the Faraday rotation effect via measurement of the Q$_{3/2}$(3/2) line of NO at 1875.81 cm$^{-1}$ (with a line strength of 3.76$\times$10$^{-20}$ cm$^{-1}\cdot$molecule$^{-1}\cdot$cm$^{2}$). The output of the laser was collimated by an aspheric lens with an effective focal length of 1.873 mm. The collimated laser beam was then directed to FRS setup: (ⅰ) a Rochon type polarizer used to "clean-up" a polarization state of the probe laser, (ⅱ) a Chernin type multipass cell used for increasing the absorption pathlength and improving the detection sensitivity, and (ⅲ) a second Rochon type polarizer acted as an polarization analyzer. The laser intensity emerging from the analyzer was detected with a thermoelectrically cooled photodetector (PIP-4TE-8, Vigo System).

    Figure 2.  Schematic diagram and corresponding photograph of the experimental setup. DAQ: data acquisition, PC: personal computer

    Laser wavelength scan was realized by feeding an external voltage ramp from a function generator (Agilent 33622A) to the injection laser diode current at a rate of 100 Hz. An internal reference sinusoidal signal ($f_{\rm{m}}$=33 kHz) from a lock-in amplifier (SR850, Stanford Research) was added to the ramp for wavelength modulation. The FRS signal was demodulated (second harmonic, $2f$ detection) by the lock-in amplifier with a time constant of 100 µs.

  • The light path diagram and actual light spot pattern of the Chernin type optical multipass cell [38-41] are shown in FIG. 3 (a) and (b) respectively. The cell consisted of two rectangular filed mirrors (FIG. 3(c), with dimensions of 75 mm$\times$15 mm and 105 mm$\times$90 mm), and three circular objective mirrors (FIG. 2(d), 45 mm in diameter). All mirrors had the same radius of curvature (ROC=1500 mm), which was equal to the base length of the cell. The mirrors were coated with protected silver with a reflectivity of $\sim$97%. Field and objective mirrors were mounted on their own back plates. Each mirrors could be adjusted independently, which had good stability to vibrations [41]. Arbitrary rows with even columns spot patterns on the field mirrors were easily achievable (as shown in FIG. 2(b), an example of 6 rows$\times$6 columns spot pattern). This configuration offered an effective use of the surface of the field mirror compared with the common used White type cell.

    Figure 3.  The Chernin type optical multipass cell. (a) Ray tracing simulation of the cell using TracePro. (b) 6 rows$\times$6 columns spot pattern. A pathlength of 108 m was achieved under this pattern. (c) Three-dimension rendering models of the field and (d) objective mirrors. The golden color of the mirror surfaces is only used to distinguish between the mirrors and the mount

  • The static magnetic field was provided by a superconducting wires (NbTi) wrapped coil [38]. In order to maintain superconducting sate, the magnet coil was sealed in a Dewar vessel (with 1280 mm long, and 500 mm inner diameter), and was cooled to temperature below 5 K with a He-cycle cryocooler (SHI F-50, Sumitomo Industries). The intensity of the magnetic field intensity was adjustable (17.9 Gauss/A) with a resolution of 2 Gauss. The maximum field intensity tested was about 1800 Gauss, which was limited by the current source. The intensity could be further increased as the excitation current increased.

  • To maximize the FRS signal in the experimental pressure of 100 mbar, a series of experiments were performed to determine the optimum magnetic field strength ($B_{\rm{opt}}$, as shown in FIG. 4(a)). The intensity of $2f$ signal (peak-to-peak value of the spectrum) initially increased with the increment of $B$, and when it reached a certain value, it began to decrease with the increment of $B$. The largest FRS $2f$ signal was found at 107 Gauss. The FRS signals at three different magnetic field strengths ($B$) are shown in FIG. 4(b). The magnet current of 0 A meant a zero $B$, which corresponded to WMS only. A current of 6 A corresponded to the optimum field. With a current of 13 A, split of the $2f$ signal was observed, which was caused by the large Zemman splitting energy level. The measured total noise of the system and SNR as a function of $\varphi$ is shown in FIG. 5. The maximum SNR was achieved at a rotation angle of 7$^\circ$ with a total noise of 3.70 µV$\cdot$Hz$^{-1/2}$.

    Figure 4.  (a) The largest FRS $2f$ signal was found at an current of 6 A, which corresponded to an optimum magnetic field strength $B_{\rm{opt}}$ of 107 Gauss. (b) The FRS $2f$ signals at three different magnetic field strengths $B$. Split of the $2f$ signal was observed at high $B$ condition

    Figure 5.  (a) The measured total noise and (b) SNR as a function of rotation angle

  • The linear relationship between FRS $2f$ signal and NO concentration is shown in FIG. 6. The linear fit uncertainty was less than 3%. Different NO concenrations were obtained by diluting 1000 ppmv NO cylinder gas with high purity N$_2$. A commercial NO-NO$_2$-NO$_x$ analyzer (Thermo Modeal 42i) was used for the measurement of the concentration of the mixed gas. The linear relationship with a correlation coefficient $R^2$ of 0.998 demonstrates that the FRS system has a good response for NO measurement.

    Figure 6.  Linear relationship between FRS-$2f$ signal and NO concentration

    Performance comparison between FRS and WMS was taken to depict the improvement of DC-FRS, as shown in FIG. 7. Continuous time series measurement of NO with the two methods are shown in FIG. 7(a, b). The time resolution of the data was 1 s (wavelength scanning with a 100 Hz ramp, and 100 spectral averaging). Measurement fluctuations ($1\sigma$ standard deviation) over 5000 s were 1.45 ppbv for FRS, and 7.06 ppbv for WMS, respectively.

    Figure 7.  Comparison between (a, c, e) FRS and (b, d, f) WMS. Time series measurement of NO concentration with (a) FRS and (b) WMS. Frequency distribution of the mixing ratio with (c) FRS and (d) WMS. Allan deviation plots for NO measurement with (e) FRS and (f) WMS

    A histogram plot of time series depicting an approximate normal distribution around the mean value is plotted in FIG. 7(c, d), which was used to assess the measurement repeatability. A Gaussian profile was fitted to the distribution histogram, resulting in a half-width at half-maximum (HWHM) of 1.71 ppbv and 8.32 ppbv, and a $\sigma_{\rm{Gaussian}}$ (a measure of actual precision) of 1.45 ppbv and 7.51 ppbv for FRS and WMS respectively.

    The stability and precision were investigated using Allan deviation analysis, which is shown in FIG. 7(e, f). For FRS, the measurement precision was 1.15 ppbv with a 1 s data acquisition time, and was improved further to 0.43 ppbv with averaging time of 15 s. For WMS, the precision over 1 s was 3.12 ppbv, and may be improved to 1.28 ppbv in 150 s. The precision for FRS was several times better than WMS.

    A comparison of the detection precision with some literature report results is shown in Table Ⅰ. Though further improvement can be made, the achieved precision in this work with DC-FRS was comparable to AC-FRS methods combined with laser frequency locking [10, 22], heterodyne-enhanced [23], dual-modulation [34], and cavity enhanced methods [32].

    Table Ⅰ.  Comparison of the NO detection precision of some FRS systems

  • An experimental study was carried out on NO detection with DC-FRS method. By using a Chernin type multipass cell, a precision of 1.15 ppbv in 1 s data-acquisition time was achieved. This precision was reduced to 0.43 ppbv by increasing the sampling time to 15 s. The experimental system in this work can be further miniaturized by using a compact multipass cell and a small solenoid magnet or a permanent magnet to make it suitable for field application.

  • This work was supported by the National Key Research and Development Program of China (No.2016YFC0202205), the National Natural Science Foundation of China (No.41805104, No.41875151, and No.41627810), the Natural Science Foundation of Anhui Province (No.1508085J03), the Youth Innovation Promotion Association CAS (No.2016383), and the CASHIPS Director's Fund (YZJJ2018QN7, BJPY2019B02).

Reference (41)

Catalog

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return