Expand this Topic clickable element to expand a topic
Skip to content
Optica Publishing Group

Rapid spectrum measurement at 3 μm over 100 nm wavelength range using mid-infrared difference frequency generation source

Open Access Open Access

Abstract

We demonstrate a broadband rapid scanning light source in the 3-μm region by using difference frequency generation (DFG). The DFG source consists of a module with quasi-phase-matched LiNbO3 ridge waveguides, a 1-μm-band wide swept range laser for the pump source, and a 1.5-μm continuous wave laser for the signal source. The sweep rate and the tuning bandwidth of this source are 20 kHz and 100 nm, respectively. This source enables us to evaluate the temperature dependence of absorbance of methane gas.

© 2016 Optical Society of America

Coherent light sources in the mid-infrared (mid-IR) range are eagerly required for gas sensing because a large number of atmospheric gases have intense absorption caused by fundamental vibrational bands in this wavelength region [1]. Various mid-IR lasers have already been reported and employed for spectroscopic investigations. Quantum cascade lasers (QCLs) have been greatly developed [24], and a distributed feedback QCL operating at room temperature is commercially available [5]. However, the QCL tuning range is limited. An optical parametric oscillator (OPO) provides a high output power and a wide tunable range [6,7], while the wavelength sweep is slow because of the mechanical tuning of the laser cavity, which includes a nonlinear crystal such as LiNbO3 (LN). Another approach is difference frequency generation (DFG) using periodically poled LiNbO3 (PPLN). It generates a 2–5 μm mid-IR wave where the LN crystal is transparent. We have developed PPLN waveguides by using a direct bonding and cutting technique, and achieved a conversion efficiency of 40%/W [8]. Thanks to a significant development in telecommunication technology we can easily obtain a 1.5-μm wide-range tunable laser and apply it to the DFG for wavelength tuning. A 3-μm DFG wave is often generated by a combination of 1.5- and 1.0-μm lasers. However, the phase matching bandwidth for 1.5-μm wave tuning is narrow, whereas that for 1.0-μm tuning is wide. Using this property, we demonstrated a two discrete wavelength 3-μm DFG laser and measured real-time CH4 isotope ratios where the results were consistent with those determined from mass spectroscopy [9].

Investigations of the transient processes in detonations and explosions require a time scale of 101 to 106s. High-speed spectroscopy using rapidly swept laser sources has been developed [10] and it has proven to be a powerful tool for fast real-time spectroscopy, which is one way of evaluating a gas temperature from acquired spectra. A laser that has a narrow linewidth compared with a white source is repeatedly scanned over the spectral range of interest and the time-dependent transmission spectra through the sample gas are recorded with a fast photodetector. When the elapsed time is identified with the source wavelength, the transient signal is converted into an absorption spectrum of the sample. Analyses of the explosion processes typically require a wavelength sweep rate of 4 nm/μs across a 40-nm span at an instantaneous linewidth of 0.1nm [11]. An optical coherence tomography (OCT) source satisfies these demands. OCT has been extensively developed [12] in the 1.0-, 1.3-, and 1.5-μm regions. These light sources have attained a wide tuning range and a fast swept rate through the integration of micro electro mechanical systems (MEMS) optics. On the other hand, there are no such fast scanning light sources for 3 μm, where the CH stretching vibrations are located, and this would be particularly useful for the diagnosis of gasoline engines. The only fast scanning mid-IR light source ever reported is an OPO, which has attained a tuning range over 900cm1 (1.67μm) in 3.36 ms [13].

In this Letter, we report a wide and fast swept mid-IR DFG source that uses an OCT scan laser as the pump source. The sweep span is wider than 100 nm in the 3-μm region, and the sweep rate is 20 kHz (50 μs). We apply the source to sensing of methane (CH4) gas and investigate the temperature dependence of the absorption spectra.

Figure 1 shows schematic drawings of the experimental setup for CH4 gas detection. We employ a 1.0-μm-band swept source including a MEMS-based micro-optic swept laser module (ESS320049-00, Exalos) for the pump and a 1.5-μm-band external-cavity tunable laser diode (TSL-210, Santec) for the signal. The signal wavelength is fixed at 1570.5 nm and the signal wave is amplified to 100 mW with an Er-doped fiber amplifier (ErFA11025, Furukawa). The average power of the pump is 15 mW. The pump and signal lights are combined with a wavelength division multiplexing fiber coupler and injected into a PPLN waveguide. The temperature of the PPLN is kept constant with a temperature controller. A 3-μm band output light is collimated with a collimate lens and injected into an absorption cell made of a test tube with a side arm [14]. The average power of the idler wave is about 10 μW. (A peak power is about 70 μW.) This side arm is connected to a vacuum station and enables us to control the pressure of the sample gas inside the absorption cell. The absorption cell has one window at the top and a retroreflector at the bottom. An incident beam is reflected on three facets of the retroreflector and reflected back to the incident window. The beam is transmitted through an optical bandpass filter to detect only mid-IR light and then focused onto a liquid-nitrogen-cooled InSb detector (J10D-M204-R250 U-10, Judson). The detected signal is sent to a preamplifier and an oscilloscope. The total bandwidth of the detector is 100 MHz. The detected signal is synchronized with the trigger signal provided by the OCT scan source.

 figure: Fig. 1.

Fig. 1. Experimental setup. ECLD, 1.5-μm-band external-cavity laser diode; SM, single mode fiber; FA, Er-doped fiber amplifier; WDM, wavelength division multiplexing fiber coupler; Col., collimate lens; InSb, liquid nitrogen cooled InSb detector; OBPF, optical bandpass filter; Osc., oscilloscope. Dotted lines indicate electric cables.

Download Full Size | PDF

Idler wave power is defined as [15]

Pi[sin(ΔkL/2)(ΔkL/2)]2η,
where Δk is the phase mismatch, L is the length of the LN. Figure 2 shows the pump and signal relationships that satisfy η=1(Δk=0) (center curve) and η=0.5 (upper and lower curves) for a quasi-phase-matched (QPM) bulk LN device. In this calculation, we used the Sellmeier equation found in Ref. [16] and the parameters we employed were L=50mm, QPM period Λ=30.3μm, and temperature T=19°C. The horizontal axis shows the pump wavelength. The calculated DFG wavelength is shown on the upper horizontal axis when the signal at 1570.5 nm is constant. The vertical axes show the signal wavelength and corresponding DFG wavelength when the pump at 1064 nm is constant. When we look at the horizontal straight line at 1570.5 nm, the full width at half-maximum, which is defined as η=0.5 (the width between the points where the 1570.5 nm line cross the lower blue curve), is 129 nm. On the other hand, that of the 1064 nm vertical line is only 7 nm. This figure clearly shows the effectiveness of tuning 1.0 μm pump light in terms of realizing a broadband 3-μm light source using QPM-LN DFG. Figure 3 shows the experimentally obtained phase matching characteristics. The lower horizontal axis shows the idler wavelength and the upper axis shows the acquired time. The QPM-LN module temperature is kept at 18.8°C. A pump source sweep rate of 20 kHz is realized by applying a sinusoidal wave voltage to the MEMS of the pump source. The one-shot recorded time is 25 μs, which satisfies the sweep speed of 4 nm/μs required for explosion analysis [11]. We adjusted the polarization of the pump light by twisting the single mode fiber, because the polarization of the pump source output is not maintained. On the other hand, the gray curve shown in Fig. 3 represents the experimental results for the phase matching curve obtained with a fixed-wavelength pump LD emitting at 1064.5 nm, rather than employing the swept pump source used in Fig. 1. The measured bandwidth is 5 nm. These bandwidths are almost consistent with the calculation in Fig. 2. However, there is some inconsistency between the calculation and the experimental results in regards to the center wavelength because we do not take account of the waveguide dispersion effect in the calculation. When we consider the waveguide dispersion in the calculation with the beam propagation method [17], the bottom point of the Δk=0 curve shifts to a shorter wavelength. The cross in Fig. 2 shows the shifted point. Although there is a possibility of obtaining a wider phase-matching curve by setting the signal at 1572 nm, we chose one peak phase-matching curve by setting the signal at 1570.5 nm.

 figure: Fig. 2.

Fig. 2. Pump and signal wavelength combination curves that satisfy Δk=0 (center, red) and [sin(ΔkL/2)/(ΔkL/2)]2=0.5 (upper and lower, blue) for a QPM bulk LN device with QPM periods of 30.3 μm. LN temperature is 19°C and L=50mm. Horizontal straight line is 1570.5 nm and vertical straight line is 1064 nm. Δλj is the bandwidth in wavelength (j=p,s,i). The subscripts indicate pump, signal and idler, respectively. When the margin of the DFG conversion is half of the peak power, the bandwidth is 129 nm in the 3-μm region when the signal at 1570.5 nm is constant, while that for the bandwidth of 1064 nm is only 7 nm. The cross shows the Δk=0 position against the signal at 1570.5 nm when we take the waveguide dispersion into account.

Download Full Size | PDF

 figure: Fig. 3.

Fig. 3. Measured phase matching spectra. LN temperature is constant at 18.8°C. The thick (red) curve indicates a fixed signal wavelength. Pump wave is repeatedly swept. Swept time is the upper horizontal axis. The gray curve shows the phase matching curve of the fixed pump wavelength. The upper horizontal axis is only for the thick curve.

Download Full Size | PDF

Then, we checked the linewidth of the DFG source. We introduced methane gas into the absorption cell and controlled the pressure at 9.7 kPa (73 Torr). The signal wavelength was 1570.504 nm and the LN temperature was 18.8°C. Figure 4 shows the obtained methane spectrum with an absorption length of 40 cm. We successfully observed the fine structure from P(6) to R(11) of the fundamental band ν3 of methane. The number in parentheses is the rotational angular momentum of the vibrational ground state of methane. This spectrum is averaged over 8 times, however, it should be noted that one sweep time for observing the overall ro-vibrational band is less than 10 μs. The time of the corresponding full width at half-maximum (FWHM) of one R-branch absorption line, R(0), is 18 ns (Fig. 4, inset), which corresponds to 0.6cm1 (18 GHz) (FWHM) by calibrating time with wavenumber from HITRAN2012 [18]. The main linewidth factors of a sample are the pressure and Doppler widths. The pressure width is 460 MHz [19] and the Doppler width is 280 MHz (FWHM) at room temperature. On the other hand, the rise time corresponding to the signal bandwidth of the detection, 100 MHz, is 3.5 ns, which corresponds to a linewidth of 0.12cm1 (3.6 GHz) (FWHM). The coherence length of the source is defined as [20]

lcFWHM=2ln(2)πλ02Δλ=2ln(2)π1Δν˜,
where λ0 is the center wavelength, Δλ is the 3-dB spectral bandwidth (FWHM linewidth) in wavelength, and Δν˜ is FWHM linewidth in wavenumber. The coherence length of the pump source is 8–12 mm from the specification sheet. From Eq. (2), we obtain the linewidth of the pump source as 0.40.6cm1. Therefore, the main linewidth of this spectrum is determined by the pump source and we cannot resolve a rotational state ensemble to each component.

 figure: Fig. 4.

Fig. 4. Fine structure of the fundamental band ν3 of methane. This spectrum was acquired by averaging 8 times. The inset is the arrowed absorption line [R(0)] expanded.

Download Full Size | PDF

Figure 5(a) shows the normalized spectrum of methane at room temperature. This spectrum is obtained by dividing the Fig. 4 spectrum by the spectrum without the sample. Figure 6(a) shows the result when the absorption cell is cooled with liquid nitrogen. The horizontal axis shows the idler wavenumber and is calibrated to the methane lines. We were able to confirm the linearity between the wavenumber and the elapsed time and to assign the entire measured spectrum beyond the 200cm1 wavenumber region to that in the HITRAN2012 database [18]. Silva and Lindsay reported a spectrum over 900cm1 using an OPO [13]. However, the linewidth of the OPO is 4.5cm1, which is 7 times wider than in this work. Figure 5(b) shows the calculated spectra of methane for 296 K. We assumed a Lorentzian lineshape for the spectra and the parameters we used included a linewidth of 0.4cm1 (FWHM), a pressure of 9.3 kPa (70 Torr), and an absorption length of 40 cm. The intensities were calculated using the JavaHAWKS program [21]. Figure 6(b) shows a spectrum calculated on the assumption that the light is transmitted in a sequence through a 30 cm cell at 77 K and a 10 cm cell at 296 K, because the absorption cell was cooled over only 3/4 of its length. When we compare the spectra, we found a good agreement between the experimental results and the calculation, especially in the 2970 to 3100cm1 wavelength region.

 figure: Fig. 5.

Fig. 5. Normalized (a) and calculated (b) spectra for methane ν3 band spectra: (a) is measured at room temperature; (b) is assumed a Lorentzian lineshape with a linewidth 0.2cm1, a pressure of 70 Torr, and an absorption length of 40 cm at 296 K.

Download Full Size | PDF

 figure: Fig. 6.

Fig. 6. Normalized (a) and calculated (b) spectra for methane ν3 band spectra: (a) is measured by cooling the absorption cell with liquid nitrogen; (b) is assumed a Lorentzian lineshape with a linewidth 0.2cm1, a pressure of 70 Torr, and an absorption length of 30 cm at 77 K and 10 cm at 296 K.

Download Full Size | PDF

In summary, we have reported a wide and fast swept mid-IR DFG source in the 3-μm region where we used an OCT scan laser as the pump source. We have demonstrated methane (CH4) gas detection using the developed spectrometer and revealed broadband characteristics as wide as 123 nm with a scan time of less than 10 μs. We also demonstrated a preliminary gas temperature analysis technique using a liquid nitrogen cooled gas cell. In this work, we demonstrated only the broadband property of the light source, however, this source has a fast sweep ability up to 20 kHz. We conclude that this newly developed light source would be beneficial for real time chemical species detection and the analysis of transient combustion processes such as those of engines and gas turbines.

Acknowledgment

We are grateful to Dr H. Sasada of the Keio University for lending us the InSb detector. He also made many suggestions in regards to the experimental setup and analysis. We are also grateful to Mr. H. Miyazawa of the NTT Electronics Corporation for making the high conversion efficiency PPLN and providing us with information about the PPLN.

REFERENCES

1. T. Töpfer, K. P. Petrov, Y. Mine, D. Jundt, R. F. Curl, and F. K. Tittel, Appl. Opt. 36, 8042 (1997). [CrossRef]  

2. S. Bartalini, S. Borri, and P. D. Natale, Opt. Express 17, 7440 (2009). [CrossRef]  

3. R. F. Curl, F. Capasso, C. Gmachl, A. A. Kosterev, B. McManus, R. Lewicki, M. Pusharsky, G. Wysocki, and F. K. Tittel, Chem. Phys. Lett. 487, 1 (2010). [CrossRef]  

4. S. Borri, S. Bartalini, P. Cancio, I. Galli, G. Giusfredi, D. Mazzotti, and P. De Natale, Opt. Eng. 49, 111122 (2010). [CrossRef]  

5. K. Fujita, T. Edamura, N. Akikusa, A. Sugiyama, T. Ochiai, S. Furuta, A. Ito, M. Yamanishi, and H. Kan, Proc. SPIE 7230, 723016 (2009). [CrossRef]  

6. F. Kuhnemann, K. Schneider, A. Hecker, A. Martis, W. Urban, S. Schiller, and J. Mlynek, Appl. Phys. B 66, 741 (1998). [CrossRef]  

7. E. V. Kovalchuk, D. Dekorsy, A. I. Lvovsky, C. Braxmaier, J. Mlynek, A. Peters, and S. Schiller, Opt. Lett. 26, 1430 (2001). [CrossRef]  

8. O. Tadanaga, T. Yanagawa, Y. Nishida, H. Miyazawa, K. Magari, M. Asobe, and H. Suzuki, Appl. Phys. Lett. 88, 061101 (2006). [CrossRef]  

9. M. Abe, S. Kusanagi, Y. Nishida, O. Tadanaga, H. Takenouchi, and H. Sasada, Opt. Express 23, 21786 (2015). [CrossRef]  

10. F. C. D. Lucia, R. S. Harmon, K. L. McNesby, R. J. Winkel, and A. W. Miziolek, Appl. Opt. 42, 6148 (2003). [CrossRef]  

11. L. A. Kranendonk, X. An, A. W. Caswell, R. E. Herold, S. T. Sanders, R. Huber, J. G. Fujimoto, Y. Okura, and Y. Urata, Opt. Express 15, 15115 (2007). [CrossRef]  

12. M. Wu, E. Vail, G. Li, W. Yuen, and C. Chang-Hasnain, Electron. Lett. 31, 1671 (1995). [CrossRef]  

13. A. Silva and I. D. Lindsay, Appl. Phys. B 107, 685 (2012). [CrossRef]  

14. H. Sera, M. Abe, K. Iwakuni, S. Okubo, H. Inaba, F. L. Hong, and H. Sasada, Opt. Lett. 40, 5467 (2015). [CrossRef]  

15. T. Yanagawa, O. Tadanaga, Y. Nishida, H. Miyazawa, K. Magari, M. Asobe, and H. Suzuki, Opt. Lett. 31, 960 (2006). [CrossRef]  

16. D. H. Jundt, Opt. Lett. 22, 1553 (1997). [CrossRef]  

17. O. Tadanaga, T. Yanagawa, Y. Nishida, K. Magari, T. Umeki, M. Asobe, and H. Suzuki, J. Appl. Phys. 102, 033102 (2007). [CrossRef]  

18. L. S. Rothman, I. E. Gordon, Y. Babikov, A. Barbe, D. C. Benner, P. F. Bernath, M. Birk, L. Bizzocchi, V. Boudon, L. R. Brown, A. Campargue, K. Chance, E. A. Cohen, L. H. Coudert, V. M. Devi, B. J. Drouin, A. Fayt, J.-M. Flaud, R. R. Gamache, J. J. Harrison, J.-M. Hartmann, C. Hill, J. T. Hodges, D. Jacquemart, A. Jolly, J. Lamouroux, R. J. Le Roy, G. Li, D. A. Long, O. M. Lyulin, C. J. Mackie, S. T. Massie, S. Mikhailenko, H. S. P. Müller, O. V. Naumenko, A. V. Nikitin, J. Orphal, V. Perevalov, A. Perrin, E. R. Polovtseva, C. Richard, M. A. H. Smith, E. Starikova, K. Sung, S. Tashkun, J. Tennyson, G. C. Toon, V. G. Tyuterev, and G. Wagner, J. Quant. Spectrosc. Radiat. Transfer 130, 4 (2013). [CrossRef]  

19. A. S. Pine, J. Chem. Phys. 97, 773 (1992). [CrossRef]  

20. C. Akcay, P. Parrein, and J. P. Rolland, Appl. Opt. 41, 5256 (2002). [CrossRef]  

21. L. S. Rothman, J. Schroeder, and K. Tang, JavaHAWKS Manual (Harvard Smithsonian Center for Astrophysics, 2004).

Cited By

Optica participates in Crossref's Cited-By Linking service. Citing articles from Optica Publishing Group journals and other participating publishers are listed here.

Alert me when this article is cited.


Figures (6)

Fig. 1.
Fig. 1. Experimental setup. ECLD, 1.5-μm-band external-cavity laser diode; SM, single mode fiber; FA, Er-doped fiber amplifier; WDM, wavelength division multiplexing fiber coupler; Col., collimate lens; InSb, liquid nitrogen cooled InSb detector; OBPF, optical bandpass filter; Osc., oscilloscope. Dotted lines indicate electric cables.
Fig. 2.
Fig. 2. Pump and signal wavelength combination curves that satisfy Δ k = 0 (center, red) and [ sin ( Δ k L / 2 ) / ( Δ k L / 2 ) ] 2 = 0.5 (upper and lower, blue) for a QPM bulk LN device with QPM periods of 30.3 μm. LN temperature is 19°C and L = 50 mm . Horizontal straight line is 1570.5 nm and vertical straight line is 1064 nm. Δ λ j is the bandwidth in wavelength ( j = p , s , i ). The subscripts indicate pump, signal and idler, respectively. When the margin of the DFG conversion is half of the peak power, the bandwidth is 129 nm in the 3-μm region when the signal at 1570.5 nm is constant, while that for the bandwidth of 1064 nm is only 7 nm. The cross shows the Δ k = 0 position against the signal at 1570.5 nm when we take the waveguide dispersion into account.
Fig. 3.
Fig. 3. Measured phase matching spectra. LN temperature is constant at 18.8°C. The thick (red) curve indicates a fixed signal wavelength. Pump wave is repeatedly swept. Swept time is the upper horizontal axis. The gray curve shows the phase matching curve of the fixed pump wavelength. The upper horizontal axis is only for the thick curve.
Fig. 4.
Fig. 4. Fine structure of the fundamental band ν 3 of methane. This spectrum was acquired by averaging 8 times. The inset is the arrowed absorption line [R(0)] expanded.
Fig. 5.
Fig. 5. Normalized (a) and calculated (b) spectra for methane ν 3 band spectra: (a) is measured at room temperature; (b) is assumed a Lorentzian lineshape with a linewidth 0.2 cm 1 , a pressure of 70 Torr, and an absorption length of 40 cm at 296 K.
Fig. 6.
Fig. 6. Normalized (a) and calculated (b) spectra for methane ν 3 band spectra: (a) is measured by cooling the absorption cell with liquid nitrogen; (b) is assumed a Lorentzian lineshape with a linewidth 0.2 cm 1 , a pressure of 70 Torr, and an absorption length of 30 cm at 77 K and 10 cm at 296 K.

Equations (2)

Equations on this page are rendered with MathJax. Learn more.

P i [ sin ( Δ k L / 2 ) ( Δ k L / 2 ) ] 2 η ,
l c FWHM = 2 ln ( 2 ) π λ 0 2 Δ λ = 2 ln ( 2 ) π 1 Δ ν ˜ ,
Select as filters


Select Topics Cancel
© Copyright 2024 | Optica Publishing Group. All rights reserved, including rights for text and data mining and training of artificial technologies or similar technologies.