Open Access
Add to BibSonomy
Abstract
We demonstrate a method based on an ICL with tunable wavelength covering ethanol absorption peak, water absorption peak and a reference point around 3.345um to make a stand-off detection of ethanol vapor in the space. The detection model is established using the ratios of reference signal and detection signal at three target wavelengths, which help to eliminate the influence of laser power and the cross interference from water vapor in the space. The intrinsic error caused by detectors and optical elements have been corrected, and availability of this approach has been proved both in theory and in experiment.
© 2018 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Stand-off detection of ethanol vapor is a special case of ethanol detection methods with the feature of detecting target space away from a distance without any disturbances, which is particularly useful to the ethanol concentration monitoring in the air exhaled by people. A device based on this technology could be widely applied to detect ethanol vapors at entrance where people are prohibited under the influence of ethanol, to monitor the interior air condition of vehicles on the road, which could contribute to eradication of the drunken drivers, and other real-time remote sensing situations.
As a result of such useful applications, this technology draws more and more attention of scientists and institutions, and a variety of methods and strategies have been reported [1–5]. The system based on Fourier transform infrared spectroscopy (FTIR) could applied to detect ethanol vapor with high selectivity but suffer the disadvantages of bulkiness, complexity, uncontrolled background source radiance variability and poor sensitivity [6, 7]. Tunable diode laser absorption spectroscopy(TDLAS) based on the 1.392um tunable diode laser does not give a sufficient measurement sensitivity of ethanol, not only because of the low absorption strength of ethanol vapor at that wavelength, but also the presence of strong interference signal caused by the water vapor, which absorbs strongly at this wavelength [8, 9]. Then a commercial available 3.345um tunable interband cascade laser (ICL) is used and the sensitivity of ethanol vapors detection improve to 1ppm [10]. However, the background subtraction method used in the system seems to be unreliable for the reason that the water vapor concentration in the atmosphere has great impact on the background signal, which is likely to change during the actual measurement process. Ethanol concentration exhaled by human can also been determined indirectly by measuring the acetaldehyde [11, 12]. Acetaldehyde is present in the exhaled breath after drinking alcohol since oxidizing in the liver, but the relationship between the acetaldehyde concentration and ethanol concentration in the breath varies a lot among individuals because of the differences in ethanol conversion efficiency [13]. Differential absorption lidar (DIAL) is the most often used technology in ethanol stand-off detection applications. It makes use of the physical phenomenon of “difference absorption” using at least two laser beams at different wavelengths. The on-absorption wavelengths that are often chosen to measure the ethanol vapor are 2.74um, 2.75um, 3.38um, 3.39um, 3.45um and 3.49um, which are close to the stronger absorption feature of ethanol [14–22]. The corresponding off-absorption reference wavelength often locates on the region where the absorption of ethanol is weaker, but it can’t be too far from the on-absorption wavelength in the system since the transmissivity varies with the wavelength. Limited by commercially available laser sources, the wavelength interval between on-absorption and off-absorption wavelengths is at least 50nm in these DIALs, which may introduce large detection error in long distance sensing applications. Moreover, the cross absorption interference from other gases, such as carbon dioxide, water vapors and some organic components, also can obscure the absorption signal from ethanol and finally lower the measurement precision. A DIAL system with extra light sources for other interference gases sensing can help to overcome this problem, but this could make the system more complex and expensive.
With the development of mid-infrared laser manufacturing technology, there are more and more commercial available laser operating around 3 to 4μm wavelengths. With the help of a tunable mid-infrared laser operating around 3.345um wavelength, a method for stand-off detection of ethanol vapors is presented in this paper. In order to improve the sensitivity and accuracy of ethanol remote sensing and reduce the environmental disturbances, a detection model based on the ratios of reference signal and detection signal at three target wavelengths is developed to determine the concentration level of ethanol vapor in the space. This approach cannot only remove the cross absorption interference of water vapor around the ethanol absorption feature, but also eliminate the influence of the variation of light power caused by the fluctuations of laser power, light beam alignment and background reflection, which make it more suitable for remote sensing of ethanol vapors in different humidity and different distance conditions.
2. Spectral characteristics and basic method to stand-off detection of ethanol vapors
Ethanol has stronger fundamental absorption feature in the 3-4um wavelength region as shown in Fig. 1(a), which is based on Pacific Northwest National Laboratory (PNNL) database. It is obvious that ethanol show two narrow features at 3345nm (2989cm−1) and 3447nm (2900cm-1) with a half width around 0.55nm and 2.3nm respectively on a broader absorption feature. The absorption peak at 3345nm (2989cm−1) is narrower and the absorption intensity is stronger. The peak absorbance (base-10) is close to 3 × 10−4 for a gas concentration of 1 ppm × meter, which makes this feature suitable for lower concentration detection of ethanol vapor. Unfortunately, according to the details of spectra shown in Figs. 1(b) and 1(c), there is a weak absorption feature of water around this wavelength. The absorption strength of water is quite weak compared with ethanol, which is approach to 5 × 10−8 for water vapor of 1 ppm × meter. In practical measurement applications, the distance of stand-off sensing and weather condition can’t be foreseen. Thus, the absorption interference of water vapor could be quite large in some scenarios. Assuming an ambient temperature of 25°C, the concentration of water vapors in the atmosphere varies between 0 and 3% [23–25]. If the distance of stand-off detection is 4m, then the optical path length is 8m and the maximum light absorption rate approach to 1%. If this part of light loss is mistaken for ethanol absorption, the measuring error would be as high as 40ppm∙m. In fact, the error would be even larger if the absorption strength of ethanol at the reference wavelength is not equal to zero.
Fig. 1 (a) Absorption spectra of ethanol in wavelength region from 3 to 4μm based on PNNL database; (b) Absorption spectra of water around 3.345um; (c) Absorption spectra of ethanol around 3.345um; (d)Measured transmission spectra of ethanol and water based on ICL source.
To solve this problem, we present a method to remove the influence of water with a tunable ICL laser operating at 3345nm wavelength. According to the Beer-Lambert Law, the relationship between incident intensity and transmitted intensity of the light beam after interacting with ethanol and water can be written as:
Where and are the absorption coefficients of water and ethanol at wavelength respectively, and are the averaging concentrations of water and ethanol in the light path L.Considering the transmission loss caused by scattering and beam steering, the incident intensity can be expressed as:
where is the light intensity assigned to sensing path, is the transmission rate at wavelength in sensing space without target gases, and represents the intensity loss cause by optical system structure, which is wavelength independent.When a beam splitter is applied to divide the laser beam into two parts, each of the beam can be expressed as:
where represents the original intensity of laser, is the reflection rate, is the transmission rate and is the absorption rate of beam splitter. is the intensity of sensing beam and is the intensity of reference beam. Then the intensity of sensing beam can be described as:Substituting Eq. (2) and Eq. (6) into Eq. (1), the transmission intensity can be expressed as:and are detected by different detectors. Assuming linear characteristics of detectors, the reference signal and detection signal can be written as:
where and are the electric signals from detection channel and reference channel respectively, is the responsivity of detector in sensing path and is the responsivity of detector in reference path. Then, the relationship between reference signal and detection signal can be expressed as:If the devices used in the sensing system are determined, the values of , , , are all confirmed at the certain wavelength. Therefore, we can define
Figure 1(d) shows the transmission spectra of water and ethanol measured by the tunable ICL and three wavelengths around 3345nm are chosen as target wavelengths, which are ethanol absorption feature peak , water absorption feature peak and reference wavelength. It is worth noting that the wavelength of laser can be modulated to cover the three target wavelengths by changing the injected current. Since, and are close to each other, the maximum difference in wavelength between them is less than 3nm, so we can assume. By defining, the relationship between reference signal and detection signal at three target wavelengths can be shown as:
Solving the Eqs. (12) to (14) with introducing the following expressions:
one can get the expressions of ethanol vapor concentration and water vapor concentration as: When there is no ethanol vapors and water vapors in sensing space, one can achieve: Then the Eq. (19) and Eq. (20) can be expressed as:3. Experiment and results
The investigations of the stand-off detection of ethanol vapor are carried out in the setup presented in Fig. 2. A tunable laser (Nanoplus S/N: 1740/1-30) generating wavelength around 3.345μm is employed as the excitation source and the features are shown in Fig. 3. A coated aspherical lens (Thorlabs model C036TME) is used to collimate the laser beam and then an isolator (Thorlabs Model IO-4-3400-WG) is applied to eliminate the influence of reflection light. Another laser generating wavelength at 0.63μm is used as an optical path label to track the ICL beam. Two laser beams are combined into one single beam by a glass plate with appropriate thin-layer coating, which could also divide the beam of ICL into two beams. One beam is directly detected to monitor the intensity of ICL by a reference detector. The other is interacting with the target gases in the sensing space and then collected by a spherical mirror to focus onto the sensing detector. To perform quantitative measurement, a gas pipe with length of (100cm) and volume of (2290.221ml) is used. The windows of pipe have an angle of (8 degree) to diminish the optical interference. The ICL is controlled by a precise integrated driver. The output current resolution of the driver is 0.01mA, and the temperature stability is up to 0.001°C.
Fig. 3 (a) Emission wavelength of laser for different temperatures and driving currents; (b) Measured P-I curve for ICL laser operating at 293K.
The absorption coefficients of ethanol and water at three target wavelengths can be calculated with the data in PNNL. The method to compute the absorption coefficient with data in PNNL database has been described in literature [26]. The values are shown in Table 1.
Table 1. Absorption coefficients of ethanol and water based on PNNL
To prepare the required concentrations of the ethanol vapor, the appropriate volumes of saturated ethanol vapor are injected into the gas pipe using a syringe. Figure 4 shows the relationship between saturated ethanol vapor and temperature [27]. At the temperature of 20.5°C, the molecular pressure of saturated ethanol vapor is equal to 5.976kPa. Thus, the concentration of the ethanol vapor with unit of particle per million (ppm) in mole fraction could be described as:
Introducing the volume of ethanol vapor that injected into gas pipe, the concentration of ethanol vapor can be expressed as:
To ensure the stability of excitation source, the ICL is driven by a periodic ramp current with output wavelengths covering the three target wavelengths in each scanning period. The detection signals and reference signals are shown in Figs. 5(a) and 5(b). The measurements are performed at distance of 2m and 4m respectively, which help to change the integral concentration of water in the experiment. The volumes of ethanol vapors injected into the gas pipe are 0ml and 4ml, and the corresponding concentration levels of ethanol vapor in the gas pipe are 0 and 103ppm. It is obvious that the baseline, detection signal without target gases, is not fixed. It is mainly impacted by sensing distance and relevant position of reflector and sensor. When the injected current of ICL is equal to 58.9 mA, the ethanol vapors have maximum light absorbance in the current scanning period. The water absorption and reference points locate at 42.4mA and 46.7mA respectively as shown in Fig. 5(a). Compared the normalized red and blue curves in Fig. 5(a), a weak water absorption feature is found locating around the ethanol absorption peak. As the sensing distance or ambient humidity increases, the cross interference of water vapor becomes more and more serious. Additionally the position of reference channel are fixed. Hence, the reference signal is almost the same except for some minor fluctuations caused by ICL power and system noise variation. Figure 5(b) shows the reference signals measured at the same time.
Fig. 5 (a) detection signal versus injected current of laser and detail of weak water absorption feature around ethanol absorption peak; (b) reference signal versus injected current of laser.
In order to ensure the accuracy of measurement, the influence of detectors and beam splitter need to be eliminated with the method mentioned in section 2. When there is no ethanol vapor in the gas pipe, the value of correction factor is measured in different conditions. The measurements are operated for 4 times with different distance and original baseline, and the results are shown in Fig. 6. Even though the measuring conditions are changed, the test data of correction factor are close to each other and show better time stability in a period time. The mean value of correction factor of ethanol concentration is equal to 18. The value of can be determined with the same approach. However, it is not easy to create an environment without water vapors to complete the experiment. Moreover, this sensor aim at detecting ethanol vapor in the space and the concentration of water is used as a reference target to prove the availability of this method. Therefore, there is no necessary to measuring the value of .
Fig. 6 (a) measured correction factor of ethanol under different conditions; (b) sensing signal curves correspond to different conditions.
During the experiment, the volumes of ethanol vapor injected in to the gas pipe change from 0.3ml to 11ml. These correspond to the concentration levels of ethanol vapor in the gas pipe from 7.7ppm to 283ppm. The test performed at distance of 2m and 4m respectively. When there is no correction operation, the measurement results are show in Fig. 7(a). Obviously, the measuring data show good linearity but all diverge from actual concentration levels, which show large error between measuring and actual concentration levels. Taking the correction factor of ethanol vapor into consideration, the error of measuring data have been reduced as shown in Fig. 7(b).
Fig. 7 (a) Measured ethanol value without correction versus actual ethanol concentration in 2m and 4m respectively; (b) Measured ethanol value with correction versus actual ethanol concentration.
Figure 8 shows the stability of the ethanol remote sensing system. A series of ethanol vapor with different concentration levels have been continuously measured over a period of 1 minute at distance of 2m and 4m. Figure 8(a) is the ethanol vapor measuring result. The variation of ethanol measuring result for each concentration level, which mainly caused by system random noise, is nearly 10ppm. In addition, for a certain ethanol concentration, there may be some difference between the 2m and 4m measuring results because the volumes of standard ethanol vapor injected into the gas pipe are operated manually. So it is likely to introduce error among actual concentration levels when operated at different times. The water concentration without correction measured at the same time are shown in Fig. 8(b). As the sensing distance increased from 2m to 4m, the absorbance caused by water also increased. According to Beer-Lambert rule, the equivalent water concentration measured at 4m should be twice times than the data measured at 2m in theory. It is clearly proved in the experiment as shown in Fig. 8(b), which also show the ability of the method to remove the interference of water vapor.
Fig. 8 (a) Measured ethanol concentration versus measurement time for 0.3, 0.6, 0.9, 1.5, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11 ml of saturated ethanol vapors injected into the gas pipe. (b) Reference water concentration levels measured at the same time.
4. Conclusion
In summary, we demonstrate an approach to stand-off detection of ethanol vapor in space. This method use only one tunable ICL source to detect the ethanol vapor and remove the interference of water vapor in the air at the same time. We analyze the spectra of ethanol and water around 3.345um and choose ethanol absorption peak, water absorption peak and a reference point as target wavelengths. Then a detection model is established based on the ratios of the reference signal and the detection signal at the three target wavelengths. By solving the model, the concentration of ethanol vapor in the space is calculated and the reference water concentration is obtained. As the setup of reference channel, the variation of optical power in light source and the optical path is eliminated. The wavelength intervals among the three target wavelengths are less than 3nm, so the measuring error caused by differences of air transmissions is much lower. In addition, the background signal caused by optical elements is measured to corrected the measuring result and improve detection accuracy. The experiment results prove the availability of the approach and make it a suitable strategy for remote measurement of ethanol vapor.
Funding
Ministry of Science and Technology of China (MSTC) (2017YFB0405304).
References and links
1. X. Guo, A. Mandelis, Y. Liu, B. Chen, Q. Zhou, and F. Comeau, “Noninvasive in-vehicle alcohol detection with wavelength-modulated differential photothermal radiometry,” Biomed. Opt. Express 5(7), 2333–2340 (2014). [CrossRef] [PubMed]
2. A. Nadezhdinskii, A. Berezin, Y. Bugoslavsky, O. Ershov, and V. Kutnyak, “Application of near-IR diode lasers for measurement of ethanol vapor,” Spectrochim. Acta A 55(10), 2049–2055 (1999). [CrossRef]
3. A. A. Kosterev, R. F. Curl, F. K. Tittel, C. Gmachl, F. Capasso, D. L. Sivco, J. N. Baillargeon, A. L. Hutchinson, and A. Y. Cho, “Effective utilization of quantum-cascade distributed-feedback lasers in absorption spectroscopy,” Appl. Opt. 39(24), 4425–4430 (2000). [CrossRef] [PubMed]
4. A. G. Berezin, O. V. Ershov, and A. I. Nadezhdinskii, “Trace complex-molecule detection using near-IR diode lasers,” Appl. Phys. B 75(2-3), 203–214 (2002). [CrossRef]
5. S. Jie, Q. J. Tang, C. Cheng, and Z. Y. Li, “Remote Detection of Alcohol Concentration in Vehicle Based on TDLAS,” in Symposium on Photonics & Optoelectronic, (IEEE, 2010), 1–3.
6. P. O. Idwasi, G. W. Small, R. J. Combs, R. B. Knapp, and R. T. Kroutil, “Multiple filtering strategy for the automated detection of ethanol by passive Fourier transform infrared spectrometry,” Appl. Spectrosc. 55(11), 1544–1552 (2001). [CrossRef]
7. T. Tarumi, G. W. Small, R. J. Combs, and R. T. Kroutil, “Remote detection of heated ethanol plumes by airborne passive Fourier transform infrared spectrometry,” Appl. Spectrosc. 57(11), 1432–1441 (2003). [CrossRef] [PubMed]
8. M. Azzazy, T. Chau, M. Wu, and T. Tanbun-Ek, “Remote sensor to detect alcohol impaired drivers,” in Lasers and Electro-Optics Society Annual Meeting,1995.8th Annual Meeting Conference Proceedings,Volume 1., IEEE, (IEEE, 1995), pp. 320–321. [CrossRef]
9. H. Geng, J. G. Liu, Y. J. Zhang, R. F. Kan, Z. Y. Xu, L. Yao, and R. Jun, “Ethanol vapor measurement based on tunable diode laser absorption spectroscopy,” Physics (College Park Md.) 63, 43301 (2014).
10. P. Kluczynski and S. Lundqvist, “Method and apparatus for remote detection of alcohol vapor in the atmosphere,” US patent 9068885 B2 (Jun 30 2015).
11. O. Ahmed, “System and method for determining a vehicle driver's blood/alcohol level,” US patent 7292153 B1 (Nov 6 2007).
12. P. C. Kamat, C. B. Roller, K. Namjou, J. D. Jeffers, A. Faramarzalian, R. Salas, and P. J. McCann, “Measurement of acetaldehyde in exhaled breath using a laser absorption spectrometer,” Appl. Opt. 46(19), 3969–3975 (2007). [CrossRef] [PubMed]
13. K. Mitsubayashi, H. Matsunaga, G. Nishio, S. Toda, and Y. Nakanishi, “Bioelectronic sniffers for ethanol and acetaldehyde in breath air after drinking” (2005), retrieved 8, 20.
14. G. H. Atkinson, M. A. Wolperdinger, and J. S. Pilgrim, “ILS sensors for alcohol detection within vehicles,” US patent 5907407 (May 25 1999).
15. N. Shuji, “Alcohol detector in vehicle,” Parent JP2000230900 (A) (2000).
16. M. Schuetz, J. Bufton, and C. R. Prasad, “A Mid-IR DIAL System Using Interband Cascade Laser Diodes,” in Conference on Lasers and Electro-Optics/Quantum Electronics and Laser Science Conference and Photonic Applications Systems Technologies, OSA Technical Digest (CD) (Optical Society of America, 2007), 1–2.
17. T. A. Alobaidi and D. W. Hill, “A helium-neon laser infrared analyser for alcohol vapour in the breath,” J. Phys. Educ. 8(1), 30–32 (1975). [CrossRef] [PubMed]
18. M. T. Azzazy and A. Dabiri, “Method and apparatus for detecting vehicle occupants under the influence of alcohol,” US parent 5349187 (Sep 20 1994).
19. J. Kubicki, J. Mlynczak, and K. Kopczynski, “Application of modified difference absorption method to stand-off detection of alcohol in simulated car cabins,” J. Appl. Remote Sens. 7(1), 1–13 (2013). [CrossRef]
20. J. Mlynczak, J. Kubicki, and K. Kopczynski, “Stand-off detection of alcohol in car cabins,” J. Appl. Remote Sens. 8(1), 083627 (2014). [CrossRef]
21. J. Mlynczak, J. Kubicki, K. Kopczynski, and J. Mierczyk, “Assessment of the application of cascade lasers to stand-off detection of alcohol vapors in moving cars,” J. Appl. Remote Sens. 10(4), 046010 (2016). [CrossRef]
22. K. Kopczyński, J. Kubicki, J. Młyńczak, J. Mierczyk, and K. Hackiewicz, “Stand-off detection of alcohol vapours in moving cars,” in Laser Technology 2016: Progress and Applications of Lasers, (International Society for Optics and Photonics, 2016), 101590Z.
23. O. Bridgeman and E. Aldrich, “Vapor pressure tables for water,” J. Heat Transfer 86(2), 279–286 (1964). [CrossRef]
24. D. R. Stull, “Vapor pressure of pure substances. Organic and inorganic compounds,” Ind. Eng. Chem. 39(4), 517–540 (1947). [CrossRef]
25. A. Gubkov, N. Fermor, and N. Smirnov, “Vapor pressure of mono-poly systems,” Zh. Prikl. Khim 37, 2204–2210 (1964).
26. J. J. Harrison and P. F. Bernath, “Infrared absorption cross sections for propane (C3H8) in the 3 mu m region,” J. Quant. Spectrosc. Ra. 111(9), 1282–1288 (2010). [CrossRef]
27. D. Ambrose and C. Sprake, “Thermodynamic properties of organic oxygen compounds XXV. Vapour pressures and normal boiling temperatures of aliphatic alcohols,” J. Chem. Thermodyn. 2(5), 631–645 (1970). [CrossRef]
