A novel technique based on laser induced plasma imaging is proposed to measure residual pressure in sealed containers with transparent walls, e.g. high voltage vacuum interrupter in this paper. The images of plasma plumes induced on a copper target at pressure of ambient air between 10−2Pa and 105Pa were acquired at delay times of 200ns, 400ns, 600ns and 800ns. All the plasma images at specific pressures and delay times showed a good repeatability. It was found that ambient gas pressure significantly affects plasma shape, plasma integral intensities and expansion dynamics. A subsection characteristic method was proposed to extract pressure values from plasma images. The method employed three metrics for identification of high, intermediate and low pressures: the distance between the target and plume center, the integral intensity of the plume, and the lateral size of the plume, correspondingly. The accuracy of the method was estimated to be within 15% of nominal values in the entire pressure range between 10−2Pa and 105Pa. The pressure values can be easily extracted from plasma images in the whole pressure range, thus making laser induced plasma imaging a promising technique for gauge-free pressure detection.
© 2018 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
Vacuum apparatuses are widely used in civilian and industrial applications, military facilities, scientific research, and space exploration [1–5]. Therefore, low pressure measurement is an ubiquitous technique. At room temperature and normal atmospheric pressure, there are approximately 2.7 × 1019 molecules per cubic centimeter which move randomly according to the Maxwell velocity distribution. The momentum imparted to the walls (or liquid surface) creates a force of 14.7 pounds for every square inch of wall area. Based on this, the unit of pressure is defined as force divided by area in pound/inch2 by the American National Standard  and in Newton/meter2 or Pa by SI . So, in the very beginning, the procedure of pressure measurement was directly or indirectly related to the measurement of force exerted on a known area by pressure gauges, such as Bourdon gauge , McLeod gauge  and capacitance manometer . At the same time, there are also many other techniques using pressure-dependent gas properties to measure pressure. The most frequently-used physical properties are thermal conductivity and number of ions on which Pirani gauge  and ionization gauge  are based, respectively. The above low pressure gauges are widely used in practice. In recent years, several new-type pressure measurement tools have been proposed based on light-matter interaction technique. The most famous are Fabry-Perot refractometer  and tunable diode laser absorption spectrometer (TDLAS) . They characterize the pressure using gas refractive index and absorption spectra, respectively. The Fabry-Perot refractometer was proposed by the National Institute of Standards and Technology (NIST) and aimed to replace mercury-based pressure standards to improve accuracy. The TDLAS-based technique is used for measurements of partial pressures. The present low pressure measurement techniques are summarized in Table 1. All the above techniques assume a contact of a low pressure gas with the measurement unit, that is very inconvenient in engineering applications. An example is the vacuum interrupter, which uses vacuum to quench the electrical arc. These compact devices operating at high voltage are widely used in power systems . It is hard to detect the vacuum quality with the existing low pressure measurement techniques because inserted component can distort the electric field distribution which can result in discharge phenomenon. Therefore, the gauge-free low pressure detection techniques become urgent in vacuum measurement field.
In our previous work, we proposed a type of the pressure detection technique based on laser induced breakdown spectroscopy which adopts spectral line intensities of ambient hydrogen and oxygen as pressure-dependent signals . However, it is hard to monitor the gas components in engineering application and the intensities are very susceptible to ambient environment, such as humidity. So, to solve this problem, we propose a new type of gauge-free low pressure online detection technique based on laser induced plasma imaging. Laser induced plasma technique is versatile for quantitative determination of elemental composition . Now it has been extensively applied in many fields such as environment monitoring , material analysis , forensic/biological identification , sputter coating  for its advantages of easiness of use, real time operation and minimal sample damage. Many studies demonstrated that ambient gas pressure has significant influence on laser induced plasma, including plasma intensities , morphology , temperature and density . Because of the ambient gas, the laser-target interaction, laser-plasma interaction and plasma expansion will differ at different ambient pressures. On one hand, with the increase of the ambient pressure, the confinement effect will become more and more pronounced. What’s more, the presence of ambient gas also leads to the formation of an ambient plasma . On the other hand, more laser energy will be absorbed by the plasma induced at high pressure, which can significantly affect the laser energy coupled to the target. Thanks to the strong coupling of laser-target, laser-plasma and plasma-ambient interactions, the laser induced plasma dynamics at an ambient pressure becomes very complicated. Free expansion, plume splitting and sharpening, hydrodynamic instability, and stagnation of the plume had been observed at different pressure levels [26, 27]. And also, the plasma plume length is related to the mass of the gas species and thermal conductivity . In general, the plasma plume at high pressure tends to become smaller than at low pressure; and more laser energy is deposited to the plasma rather than to the target. This had been verified both in experiments  and simulations . This motivates us to develop a novel low pressure on line detection technique using laser induced plasma imaging. This method is fundamentally different from previous low pressure measurement techniques both in principles and applications. It can be used as a gauge-free measurement technique without guiding the low pressure gas into a detection unit.
2. Experimental setup
The schematic of the experimental setup is shown in Fig. 1. A Q-switched laser (Brilliant Eazy, Quantel, France) with 1064nm fundamental wavelength, 5ns pulse duration and 10Hz repetition rate was used to ablate the material and form the plasma. The laser energy was set at 5mJ. This is the low energy as compared to common laser induced plasma applications; it is chosen for the purpose of minimizing the effect of laser ablation on the vacuum quality and allowing the portability and inexpensiveness of the equipment. The ablated crater area is 1.25 × 10−3cm2 in 105Pa gas, which gives the 4J/cm2 fluence. To guide the laser beam into the vacuum chamber, a laser reflection mirror was employed and the plasma was formed after focusing the laser beam by a 15cm focal length, 1-inch diameter plano-convex lens through a quartz window. The focused laser beam was perpendicular to the copper target which was placed on a XYZ-stepping motor stage in the vacuum chamber. The XYZ-stepping motor stage ensured a fresh surface to be ablated with each laser shot. The volume of the whole vacuum chamber is 48L; it is evacuated by the two-stage pump system which consists of a mechanical pump (Ecodry M, Leybold Vakuum, Germany) and turbo molecular pump (Turbovac 361, Leybold Vakuum, Germany). The pressure in the vacuum chamber was monitored by a wide range hot ion combi vacuum gauge (Ionvac ITR90, Leybold Vakuum, Germany) which consists of a hot cathode ionization measurement system and Pirani measurement system; it can measure pressure from 10−8Pa to 105Pa. The plasma images were captured by an intensified CCD (ICCD) camera (DH734, Andor, UK) using a 15cm focal length, 1-inch diameter objective lens set at a 90° angle with respect to the laser beam and then displayed and recorded by the computer. The ICCD can cover the wavelength range from 180nm to 850nm. The timing between laser and ICCD was controlled by a delay generator (DG645, SRS, USA). During the experiment, all plasma images were recorded using a 100ns gate and 0 gain. To investigate the time-dependent evolution of plasma shape, time resolved plasma imaging was performed at delay time of 200ns, 400ns, 600ns and 800ns. The experiments were performed in a wide pressure range from 10−2Pa to 105Pa with a step of one order of magnitude.
Figure 2 shows the plasma images at 200ns, 400ns, 600ns and 800ns delay time and different pressures from 10−2Pa to 105Pa in steps of one order of magnitude. The plasma intensities at pressures above 10Pa are much higher than those at lower pressures. So, the scale in the top right corner of Fig. 2 shows the color code for only low pressures from 10−2Pa to 10Pa. In each picture, the laser beam was incident from below. Firstly, we can see that the plasma sizes are much smaller at high pressures because of the strong confinement effect by the ambient gas. The laser induced plasma slowly expands in the 105Pa gas and propagates in the direction of laser incidence when pressure decreases. With the decrease of pressure from 105Pa to 10Pa, the confinement effect becomes weak, and that is revealed by the larger plasma plume area as compared to that at higher pressures. When the pressure is above 100Pa, the leading edge of the plasma plume is much flatter than that at lower pressures; the effect is especially strong at pressures from 104Pa to 105Pa, as if the plasma were a moving liquid obstructed by a slab. In order to see this more clearly, the enlarged images are shown in the red rectangle in Fig. 2. The flat leading edge can be explained by the fact that at the beginning of plasma formation, the normal component of the velocity of plasma particles is much larger than the tangent component. So, the compression of the ablated copper material and ambient gas is stronger in the normal direction than in the tangent direction; this results in a denser particle region at the leading edge of the plasma, thus making the plasma to move tangentially. This phenomenon agrees with our previous simulation results  that the pressure gradient is higher along smaller dimension (cigar-like) that makes plume to expand more tangentially than normally.
When the pressure decreases from 10Pa to 1Pa, the plasma size also decreases and the plasma shape changes from spherical to tear-like. With further decrease of pressure down to 10−2Pa, the plasma size and plasma intensities increase again. Such behavior of the plasma intensity agrees with the results by Xiao . This may possibly occur due to the decrease in the electron densities at low pressures from 1Pa to 10−2Pa , that causes less laser energy to be absorbed by the plasma plume, and, correspondingly, more target atoms to be ablated. Therefore, there appears a large plasma region at lower pressures below 1Pa. Furthermore, we can clearly see a tear-like shape plasma when pressure is lower than 1Pa. This is due to the laser induced detonation effect ; the initially formed plasma absorbs the energy of the laser pulse trailing edge, resulting in faster expansion as compared to the later plasma; at the same time, due to the insufficient collision frequency, the after-breakdown plasma moves very slowly, which means near to the target; so, the high-speed moving particles and low-speed moving particles contribute to a tear-like shape; this is obvious from Fig. 2 for the delay times exceeding 400ns.
A series of fast images taken by the ICCD camera visualizes the pulse-to-pulse fluctuation of plasma emission which is crucial for pressure value extraction from plasma images. Therefore, to investigate the repeatability of the laser induced plasma shape and signal intensities, three single shot images are acquired at each specific pressure and delay time. The single-shot plasma images taken at 600 ns delay time and pressures from 10−2Pa to 105Pa in one order of magnitude step are shown in Fig. 3. All three repetitive images at each pressure are very similar. We thus conclude that the plasmas are shot-to-shot reproducible. All the plasma images show tear-like shapes when the pressure is below 1Pa. With increasing the pressure up to 103Pa, plasmas tend to an ellipsoidal shape. When pressure exceeds 104Pa, all the images show a flat leading edge. In the whole pressure range, the position of plasma center, plasma shape and plasma intensity are very reproducible at each specific pressure; the maximum RSD of plasma center and plasma integral intensities at all the pressures are 6.5% and 5% respectively, which means this low pressure detection technique is robust.
To extract pressure information and quantify it, a subsection characterization method is used, which is illustrated in Fig. 4. Firstly, the distance from the barycenter of plasma emission intensity to the target surface is adopted as a measuring metric when pressures are higher than 10Pa. To get the relationship between the measured image size in pixels and physical length, a ruler was placed at the plasma formation region and projected to the ICCD giving 50 pixels per 1mm. The distance to the target surface was plotted in red in Fig. 4; it varies from 0.3mm to 4.5mm when pressure decreases from 105Pa to 10Pa and the variation is almost linear. The standard deviation of the distance calculated from the three-shot statistics decreases significantly with increasing the pressure. This indicates the pressure is easy to measure when it is in the range between 10Pa and 105Pa. The distance metric, however, cannot be used in the pressure range between 1Pa and 10Pa because the barycenter is much closer to the target at 1Pa than at 10Pa. Instead, the integral plasma intensity is used as a metric in this pressure range. To find the transition point between the two metrics, the plasma shape at 5Pa was acquired. The integral plasma intensities were plotted by the dash purple line in Fig. 4. One sees that the integral intensity increases from 0.8 × 107 to 4.5 × 107 counts when the pressure increases from 1Pa to 10Pa. The integral intensities become even higher when the pressure exceeds 10Pa which makes it very easy to distinguish the pressure above or below 10Pa. Third, the plume length metric is used to define the pressure when it drops below 1Pa; this region was plotted by the black line in Fig. 4. The plume length is calculated as below: first, the background noise is subtracted from the plasma signal; then, the intensity profile of the line going through the barycenter is fitted with Gaussian curve; afterwards; getting the difference value of abscissa in 10% Gaussian curve peak; finally, conversing pixel length to physical length. The plume length changes from 4.3mm to 6.8mm when the pressure decreases from 1Pa to 10−2Pa with the RSD below 1.5%.
Considering the situation that just one image is used to deduce the pressure in a practical application, the following strategy can be applied, which is shown in Fig. 5 in the form of flowchart:
- 1) Obtaining the integral plasma intensity, comparing with integral intensity at 10Pa which is obtained in advance. If the intensity is higher than 10Pa’s, which means the pressure above 10Pa, distance to the target surface is used to pressure quantification, otherwise the pressure is lower than 10Pa;
- 2) If no tear-like shape appears, which means the pressure is between 1Pa and 10Pa, the integral plasma intensity is used for pressure quantification;
- 3) If a tear-like shape appears, which means the pressure is below 1Pa, the plasma plume length is used for pressure quantification.
In order to verify the accuracy of this method, we performed the experiments at 5 × 104Pa, 5 × 103Pa, 5 × 102Pa, 5 × 101Pa, 7Pa, 3Pa, 5 × 10−1Pa and 5 × 10−2Pa. Those pressures are in the middle of the calibration points in Fig. 4. Accuracy is calculated using Eq. (1) and calculated results are shown in Fig. 6.
The , and n represent the measured pressure by our method, measured pressure by hot ion combi gauge and number of measurements.
The abscissa and ordinate of Fig. 6(a) represent the pressure values measured by hot ion combi gauge and our method, respectively. Figure 6(b) represents the calculated accuracy using the above equation; the error bars represent the fluctuation of five time measurement. In order to get the pressure values, the distance to the target surface, integral intensity and plasma plume length were used for the pressure above 10Pa, and 1Pa-10Pa and below 1Pa, respectively, as prompted by Fig. 4. The calculated results indicate that all the pressures measured by our method and vacuum gauge deviate insignificantly from a 45° slope, like corresponding to 100% accuracy analysis. What is more, the bias doesn’t show a strong change rule in the whole pressure range in Fig. 6(b); but the average accuracies are 15% approximately from 10−2Pa to 105Pa.
In conclusion, a novel gauge-free low pressure online detection technique was proposed in this paper. In order to realize the low pressure measurement, the plasma expansion dynamics was studied at pressures from 10−2Pa to 105Pa in an order of magnitude step and the delay times of 200ns, 400ns, 600ns and 800ns, respectively. The experimental results indicated that the ambient gas pressure has a significant influence on the shapes and intensities of laser induced plasma images. The laser-target and, laser-plasma interactions and plasma expansionare markedly different at pressures ranging from 10−2Pa to 105Pa, that is reflected in complicated plasma shapes and intensity distributions. To extract pressure information from the enclosure with an accessible ablation surface, plasma imaging was applied. The subsection characterization method was developed for different pressure ranges, which adopted the metrics of (i) the distance from the plume barycenter to the target surface, (ii) the integral plasma intensity, and (iii) the plume length. The plasma images at 5 × 104Pa, 5 × 103Pa, 5 × 102Pa, 5 × 101Pa, 7Pa, 3Pa, 5 × 10−1Pa and 5 × 10−2Pa were acquired to assess the method’s accuracy. Pressure determined by the imaging method coincided with those measured by the pressure gauge within 15%. Compared with the previous pressure detection method based on the force exerted on a known area or pressure-dependent gas properties, this technique is fundamentally different and allows pressure measurement in the wide pressure range. It is unnecessary to bring the low pressure gas in contact with the detection unit that makes this technique very convenient for engineering applications. As discussed in the introduction session, the typical application can be used to vacuum interrupter of which the critical pressure is 0.1Pa. Due to the limitation of our vacuum system, the minimal reachable vacuum degree was 10−2Pa; however, this technique might be usable even at pressures lower than 10−2Pa.
National Natural Science Foundation of China (NSFC) (51777154, 51521065) and China Scholarship Council.
References and links
1. A. N. Rider, A. A. Baker, C. H. Wang, and G. Smith, “An Enhanced Vacuum Cure Technique for On-Aircraft Repair of Carbon-Bismaleimide Composites,” Appl. Compos. Mater. 18(3), 231–251 (2011). [CrossRef]
2. W. Hopper, “One Mill’s Experience Using MAC Testing to Evaluate Vacuum Interrupter Integrity in 15 kV Vacuum Circuit Breakers,” IEEE Trans. Ind. Appl. 53(1), 774–779 (2017). [CrossRef]
3. P. Kothary, B. M. Phillips, S. Y. Leo, and J. Peng, “Bioinspired broadband midwavelength infrared antireflection coatings on silicon,” Vac. Sci. Technol., B 34(4), 041807 (2016). [CrossRef]
4. K. Belczynski, D. E. Holz, T. Bulik, and R. O. Shaughnessy, “The first gravitational-wave source from the isolated evolution of two 40-100 Msun stars,” Nature 534(7608), 512 (2016). [CrossRef] [PubMed]
5. A. G. Bagmutand and V. M. Beresnev, “Kinetics of the electron beam induced crystallization of amorphous ZrO2 films obtained via ion-plasma and laser sputtering,” Phys. Solid State 59(1), 151–155 (2017). [CrossRef]
6. D. J. Pacey, Measurement of vacuum. In Instrumentation Reference Book: Third Edition (Elsevier Inc., 2003).
7. N. Barry, Taylor, and Ambler Thompson, “The International System of Units (SI),” (National Institute of Standards and Technology, 2008).
8. S. G. Foord, “An improved Bourdon gauge,” J. Sci. Instrum. 11(4), 126–127 (2002). [CrossRef]
9. C. Meinke and G. Reich, “Influence of diffusion on the measurement of low pressure with the McLeod vacuum gauge. Based on a paper by Gaede,” Vacuum 13(12), 579–581 (1963). [CrossRef]
10. G. Loriotand and T. Moran, “Reliability of a capacitance manometer in the range 2×10−4-5×10−6Torr,” Rev. Sci. Instrum. 46(2), 140–143 (1975). [CrossRef]
11. J. S. Shie, B. C. S. Chou, and Y. Chen, “High performance Pirani vacuum gauge,” J. Vac. Sci. Technol. A 13(6), 2972–2979 (1998). [CrossRef]
12. C. K. Dong and G. R. Myneni, “Carbon nanotube electron source based ionization vacuum gauge,” Appl. Phys. Lett. 84(26), 5443–5445 (2004). [CrossRef]
13. P. F. Egan, J. A. Stone, J. E. Ricker, and J. H. Hendricks, “Comparison measurements of low-pressure between a laser refractometer and ultrasonic manometer,” Rev. Sci. Instrum. 87(5), 053113 (2016). [CrossRef] [PubMed]
14. J. Karl, G. Padilla-Víquezand, and T. Bock, “Investigation of tunable diode laser absorption spectroscopy for its application as primary standard for partial pressure measurements,” J. Phys. Conf. Ser. 100, 0920059 (2008).
15. M. Weuffel, D. Gentschand, and P. G. Nikolic, “Influence of Current Interruption Operations on Internal Pressure in Vacuum Interrupters,” IEEE Trans. Plasma Sci. 45(8), 2144–2149 (2017). [CrossRef]
16. X. H. Wang, H. Yuan, D. X. Liu, A. J. Yang, P. Liu, L. Gao, H. B. Ding, W. T. Wang, and M. Z. Rong, “A pilot study on the vacuum degree online detection of vacuum interrupter using laser-induced breakdown spectroscopy,” J. Phys. D Appl. Phys. 49(44), 44LT01 (2016). [CrossRef]
17. S. V. Shabanov and B.I. Gornushkin, “Geometrical effects in data collection and processing for calibration-free laser-induced breakdown spectroscopy,” J. Quant. Spectrosc. Radiat. Transf. 204, 190–205 (2018). [CrossRef]
18. A. Moosakhani, P. Parvin, A. Majdabadiand, and M. M. Hashemi, “Radon decay monitoring in air using characteristic emission of species in metal-assisted LIBS,” Radiat. Meas. 92, 39 (2016). [CrossRef]
19. J. E. Haddad, B. Bousquet, L. Canioniand, and P. Mounaix, “Good practices in LIBS analysis: Review and advices,” Spectrochim. Acta B At. Spectrosc. 101(3), 171–182 (2014). [CrossRef]
20. B. Busser, S. Moncayo, J. L. Coll, L. Sanceyand, and V. Motto-Ros, “Elemental imaging using laser-induced breakdown spectroscopy: A new and promising approach for biological and medical applications,” Coord. Chem. Rev. 358, 70–79 (2018). [CrossRef]
21. H. Yeom, B. Hauch, G. Cao, B. Garcia-Diaz, M. Martinez-Rodriguez, H. Colon-Mercado, L. Olsonand, and K. Sridharan, “Laser surface annealing and characterization of Ti2AlC plasma vapor deposition coating on zirconium-alloy substrate,” Thin Solid Films 615(6), 202–209 (2016). [CrossRef]
22. M. López-Claros, M. Dell Aglio, R. Gaudiuso, A. Santagata, A. D. Giacomo, F. J. Fortes, and J. J. Laserna, “Double pulse laser induced breakdown spectroscopy of a solid in water: Effect of hydrostatic pressure on laser induced plasma, cavitation bubble and emission spectra,” Spectrochim. Acta B At. Spectrosc. 133, 63–67 (2017). [CrossRef]
23. X. M. Lin, H. R. Sun, and J. J. Lin, “Comparison of SP-LIBS and DP-LIBS on metal and non-metal testing based on LIBS,” Proc. SPIE 10457, 104571V (2017).
24. N. Faridah, N. Salwanie, S. Z. H. Rizvi, K. T. Chaudary, M. S. A. Aziz, and J. Ali, “Laser induced graphite plasma kinetic spectroscopy under different ambient pressures,” AIP Conf. Proc. 1824, 030007 (2017). [CrossRef]
25. N. Farid, S. S. Harilal, H. Ding, and A. Hassanein, “Emission features and expansion dynamics of nanosecond laser ablation plumes at different ambient pressures,” J. Appl. Phys. 115(3), 033107 (2014). [CrossRef]
26. S. S. Harilal, C. V. Bindhu, M. S. Tillack, F. Najmabadi, and A. C. Gaeris, “Internal structure and expansion dynamics of laser ablation plumes into ambient gases,” J. Appl. Phys. 93(5), 2380–2388 (2003). [CrossRef]
27. S. Amoruso, B. Toftmann, and J. Schou, “Thermalization of a UV laser ablation plume in a background gas: From a directed to a diffusionlike flow,” Phys. Rev. E 69(5), 056403 (2004). [CrossRef] [PubMed]
28. M. S. Dawood, A. Hamdan, and J. Margot, “Influence of surrounding gas, composition and pressure on plasma plume dynamics of nanosecond pulsed laser-induced aluminum plasmas,” AIP Adv. 5(10), 107143 (2015). [CrossRef]
29. G. Cristoforetti, S. Legnaioli, V. Palleschi, A. Salvettiand, and E. Tognoni, “Influence of ambient gas pressure on laser-induced breakdown spectroscopy technique in the parallel double-pulse configuration,” Spectrochim. Acta B At. Spectrosc. 59(12), 1907–1917 (2004). [CrossRef]
30. Z. Y. Chen, D. Bleiner, and A. Bogaerts, “Effect of ambient pressure on laser ablation and plume expansion dynamics: A numerical simulation,” J. Appl. Phys. 99(6), 063304 (2006). [CrossRef]
31. S. V. Shabanov and I. B. Gornushkin, “Two-dimensional axisymmetric models of laser induced plasmas relevant to laser induced breakdown spectroscopy,” Spectrochim. Acta B At. Spectrosc. 100, 147–172 (2014). [CrossRef]
32. Q. M. Xiao, “Analysis of ITER-like first wall materials in tokamak by laser induced breakdown and mass spectroscopy,” http://cdmd.cnki.com.cn/Article/CDMD-10141-1015573855.htm.
33. S. S. Harilal, N. Farid, J. R. Freeman, P. K. Diwakar, N. L. Lahaye, and A. Hassanein, “Background gas collisional effects on expanding fs and ns laser ablation plumes,” Appl. Phys., A Solids Surf. 117(1), 319–326 (2014). [CrossRef]
34. M. Ushio, K. Komurasaki, K. Kawamura, and Y. Arakawa, “Effect of laser supported detonation wave confinement on termination conditions,” Shock Waves 18(1), 35–39 (2008). [CrossRef]