Measurements of the yield and chlorine utilizationof singlet oxygen generator by useof Raman spectroscopy

Weili Zhao, Liping Duo, Fengting Sang, and Fang Chen

A Raman scattering system was used to measure the O2 �a1�� yield and the chlorine utilization in asinglet oxygen generator of chemical oxygen iodine laser with nitrogen diluent. We present the resultsfrom the tests that conducted on a 0.1-mol�s jet-type singlet oxygen-iodine generator. On the basis of thecurrent reported uncertainty of the Raman cross section, the error in the yield measurement is calculatedto be less than 8%, and the error of the chlorine utilization is 12%. © 2003 Optical Society of America

OCIS codes: 120.0120, 140.3550, 140.1550.

1. Introduction

The chemical oxygen-iodine laser �COIL� is the firstchemical laser based on electronic transitions. Itemits at 1315 nm on the transition between the spin-orbit levels of the ground state configuration of theiodine atom. The upper level is populated by a near-resonant energy transfer from an O2 �a1�� moleculeto an atomic iodine atom1–3:

O2�a1�� � I�2P3�2�3 O2�X

3�� � I*�2P1�2�. (1)

O2 �a1�� molecules are obtained from the chemicalreaction between chlorine and basic hydrogen perox-ide liquor4

Cl2 � H2 O2 � 2KOH3 O2�a1�� � 2H2 O � 2KCl.

(2)

The output power of this laser strongly depends onthe yield of O2 �a1�� in singlet oxygen generator�SOG�5,6:

Y ��O2�a

1���

�O2�a1��� � �O2�X

3���. (3)

The authors are with the Dalian Institute of Chemical Physics,the Chinese Academy of Sciences, P.O. Box 110, Dalian 116023,China. W. Zhao is also with the Dalian Navy Academy, China.W. Zhao’s e-mail address is zhaoweili@yahoo.com.

Received 25 March 2002; revised manuscript received 6 Novem-ber 2002.

0003-6935�03�183524-04$15.00�0© 2003 Optical Society of America

3524 APPLIED OPTICS � Vol. 42, No. 18 � 20 June 2003

To predict COIL performance, it is very significant tomeasure the exact values of the chlorine utilizationand the yield of O2 �a1�� of the SOG. Usually, thechlorine utilization fraction is determined by measur-ing chlorine absorption at 325 nm in the diagnosticduct downstream of the oxygen generator,7 and theO2 �a1�� density is measured by monitoring the O2�a1�� emission at 1270 nm with a calibrated intrinsicGe detector.8 However, this intrinsic Ge detector isdifficult to calibrate, and the results usually exhibitlow absolute accuracy in the yield. Recently, diodelasers were used for measurements of both the yieldand the water vapor fraction by absorption spectros-copy diagnostics in the COIL.9 Nevertheless, thistechnique requires at least one additional measure-ment to obtain the yield. Furthermore, clean win-dows must be required during the experiment toavoid erroneous results. If one wants to monitor thechlorine utilization and yield simultaneously, how-ever, two diagnostic systems are required. Fortu-nately, the spontaneous Raman-scattering techniquehas been developed that can be used to monitor theconcentrations of the O2 �a1�� and the O2 �X3�� si-multaneously10 in the SOG. Many common experi-mental problems, such as fluctuations in laser power,dirty windows, and misalignment, can be ratioed outbecause of monitoring in the same measurement vol-ume.

In this paper, we report the application of sponta-neous Raman spectroscopy to monitor the yield andthe chlorine utilization at the exit of a jet-type SOGfor COIL with a nitrogen diluent. To our knowledge,these are the first experimental results in which thechlorine utilization and yield were measured simul-

taneously in one system. The results are from theexperiments that were conducted on a 0.1-mol�sSOG.

2. Measurement Principle

The intensity of Raman spectrum is decided by scat-tered molecule structures. The intensity is propor-tional to the concentration of molecule in sample. Sowe can get the expressions

a1 �IO2�a1����O2�a1��

IN2��N2

, (4)

a2 �IO2�X3����O2�X3��

IN2��N2

, (5)

where a1 and a2 are the concentration ratio of the O2�a1�� and of the O2 �X3��, respectively, to N2 at theexit of the SOG; I is the area under the respectivepeaks of the Raman spectrum of O2 �a1��, O2 �X3��,and N2; and � is the Raman cross section. Hence theconcentration ratio of the total oxygen to nitrogen atthe exit of the SOG can be obtained by using theexpression

�O2�

�N2�� a1 � a2. (6)

The chlorine Raman signal is filtered because ofthe close wavelength to the pump laser; however, thechlorine utilization can be calculated from the follow-ing expression:

� �NO2

NCl2

��O2�

�N2�b � �a1 � a2�b, (7)

where N is the number of molecules and b is the ratioof nitrogen to chlorine. According to respective Ra-man cross sections,11,12 the chlorine utilizationshould be obtained as

� � �1.9IO2�a1��

IN2

� 0.86IO2�X3��

IN2

�b, (8)

where the O2 �a1�� Raman cross section at 527 nm isexpected to be close to that at 532 nm because of thecloseness of the wavelengths.

Similarly, the yield can be calculated by using theexpression

Y ��O2�a

1���

�O2�a1��� � �O2�X

3���

�IO2�a1����O2�a�

IO2�a1����O2�a� � IO2�X3����O2�X�

�IO2�a1��

IO2�a1�� � IO2�X3����O2�a���O2�X��. (9)

3. Experimental Setup

The square-pipe-array jet-type SOG was described inour previous paper.13 There are some holes �0.7 mm

in diameter� on the undersurface of each square pipe.The specific surface area of the basic hydrogen per-oxide jet is approximately 3.1 cm1.

The schematic illustration of experimental setup isshown in Fig. 1. The doubled Nd:YAG laser wasoperated at 10 Hz. Each laser pulse had a durationof 20 ns and a pulse energy of 400 mJ. The Ramanscattering light, which was collected perpendicularlyto the propagation of the laser beam, was focusedonto an intensified charge-coupled device array by acoated lens, a holographic notch filter, and an f�6.5spectrograph. The time delay of the DG-535 digitaldelay�pulse generator was controlled by computer soaccurately that the photocathode of the image inten-sifier tube was opened when the Raman scatteringsignal arrived. The detector microchannel intensi-fier had a gate width of 25 ns.

4. Results and Discussions

A. Chlorine Utilization

We can obtain the chlorine utilization and yield si-multaneously for a COIL with a nitrogen diluent.The typical Raman spectrum is shown in Fig. 2. TheRaman signal of gas in low pressure is so feeble be-

Fig. 1. Schematic illustration of experimental setup.

Fig. 2. Raman spectrum of a sparger flow.

20 June 2003 � Vol. 42, No. 18 � APPLIED OPTICS 3525

cause the Raman cross sections are extremely low�1034m2sr1�, and it is very important in enhanc-ing the collecting efficiency. In addition, the Ramanlight should be collected at suitable angles.

The values of a series of experiments are presentedin Table 1. The chlorine utilization is calculated tobe in the range of 85–95% in the case of the diluentratio of 1.1:1. This result is coincident with thoseachieved by Furman et al.9

B. Yield

Figure 3 shows the relation between the yield and thepressure in the case of the diluent ratio �N2:Cl2� of1.1:1. Obviously, the yield increases with a reducingof cell pressure in the same flux. This is because theprobability of self-quenching of O2 �a1�� molecules isproportional to the square of O2 �a1�� concentrationas well as to the square of the cell pressure.14 More-over, the increasing velocity makes the singlet oxygenmolecules remain a shorter time at the reaction re-gion and reduces the deactivated probability of sin-glet molecular oxygen when the pressure is reduced.The highest delta oxygen yield in our experiments isapproximately 60%.

For the purpose of studying the influence of thediluent, these experiments were done with heliumdiluent and without diluent, respectively. The ex-perimental results were shown in Figs. 4 and 5. Thediluent made the yield high. The highest yield inFig. 5 is approximately 65% when the ratio of heliumto chlorine is approximately 5.5:1. Obviously, the

kinds of diluent and diluent ratio affect the yield.Hence, it is very significant to choose the optimumdiluent and ratio for obtaining higher-output powerof COIL. The relationship of yield and proportionsof helium to Cl2 is shown in Fig. 6. In the range of

Fig. 3. Relation of yield and pressure �nitrogen�.

Fig. 4. Relation of yield and pressure �no carrier�.

Fig. 5. Relation of yield and pressure �helium�.

Fig. 6. Relation of yield and proportion He:Cl2.

Table 1. Series of Experimental Results on a 0.1-mol Jet-Type SingletOxygen-Iodine Generator

Exp. N2:Cl2 P �Torr� Yield �%�Chlorine

Utilization �%�

1 1.1 45 29 922 1.1 30 41 883 1.1 25 46 894 1.1 20 54 945 1.1 12 62 86

3526 APPLIED OPTICS � Vol. 42, No. 18 � 20 June 2003

our experiments, the yield increases with an increas-ing diluent ratio. The reason is that the gas velocityis higher with helium and hence the singlet oxygenresidence time is shorter. The values of �RT7, de-fined as the product of the oxygen pressure and res-idence time for different diluent ratios, werecalculated and shown in Fig. 7. This product re-duces with increasing dilution ratio. It is worth no-ticing that the oxygen generator will be unstableowing to liquid carryover problems when the diluentratio is increased to some degree.

The relative error in the measurement of chlorineutilization can be calculated by the expression

�d�

�� � �d�IO2�a1�� � IO2�X3����

�IO2�a1�� � IO2�X3����� � �dIN2

IN2

� (10)

and found to be less than 7%, where � is the Ramancross-section ratio of the O2�a1�� and O2 �X3��; theothers had been mentioned in Section 2. Consider-ing the error of the cross-section ratio, the total errorof the chlorine utilization is 12%.

The error in the yield measurement can be writtenas

�dYY � �

IO2�X3���

�IO2�a1�� � IO2�X3���� ��d�

�� � �dIO2�X3��

IO2�X3���

� �dIO2�a1��

IO2�a1���� . (11)

The errors of all applying values we had calculatedby Eq. �11� are less than 8%. This equation complieswith the analysis of McDemott.15

The most important error source is that the ratio ofsignal to noise is lower. Increased sensitivity andaccuracy can be obtained with further research.

5. Conclusions

The spontaneous Raman spectroscopy can be used tomeasure the yield and the chlorine utilization of

chemical oxygen iodine lasers with nitrogen diluent.This technique not only has the advantage of simple-ness but also increases the accuracy of the resultsbecause it allows one to monitor both the O2 �a1�� andO2 �X3�� simultaneously in the same measurementvolume. By using the current reported uncertaintyof the Raman cross-section, one can calculate the er-ror in the yield measurement to be less than 8%, andthe error of chlorine utilization is 12%.

The authors thank Benjie Fang, Yuelong Zhang,and Xiangde Min for their invaluable help with op-eration of the COIL system.

References1. W. E. McDemott, N. R. Pchelkin, D. J. Benard, and R. R.

Bousek, “An electronic transition chemical laser,” Appl. Phys.Lett. 32, 469–470 �1978�.

2. K. A. Truesdell and S. E. Lamberson, “Philips LaboratoryCOIL technology overview,” in 9th International Symposiumon Gas Flow and Chemical Lasers, C. Fotakis, C. Kalpouzos,and G. Papazoglou, eds., Proc. SPIE 1810, 476–492 �1992�.

3. P. V. Avizonis, “Chemical pumped electronic transition lasers,”in Gas and Chemical Lasers, M. Onorato, ed. �Plenum, NewYork, 1982�.

4. F. L. Li, B. L. Yang, and Q. Zhuang, “Theoretical study ofoptimum construction parameters of JSOG,” Chin. J. Quan-tum Electron. 16, 142–146 �1999�, in Chinese.

5. B. D. Barmashenko, A. Elior, E. Lebiush, and S. Rosenwaks,“Modeling of mixing in chemical oxygen-iodine lasers: ana-lytic and numerical solutions and comparison with experi-ments,” J. Appl. Phys. 75, 7653–7655 �1994�.

6. N. N. Yuryshev, “Chemically pumped oxygen-iodine laser,”Quantum Electron. 23, 583–600 �1996�.

7. T. L. Rittenhouse, S. P. Phipps, and C. A. Helms, “Performanceof a high-efficiency 5-cm gain length supersonic chemicaloxygen-iodine laser,” IEEE J. Quantum Electron. 35, 857–866�1999�.

8. A. Elior, B. D. Barmashenko, E. Lebiush, and S. Rosenwaks,“Experiment and modeling of a small-scale, supersonic chem-ical oxygen-iodine laser,” Appl. Phys. B 61, 37–47 �1995�.

9. D. Furman, B. D. Barmashenko, and S. Rosenwaks, “Diode-laser-based absorption spectroscopy diagnostics of a jet-typeO2 �a1�� generator for chemical oxygen-iodine lasers,” IEEE J.Quantum Electron. 35, 540–547 �1999�.

10. V. T. Gylys and L. F. Rubin, “Direct measurement of O2 �a1��and O2 �X3�� in chemical oxygen-iodine lasers with use ofspontaneous Raman imaging,” Appl. Opt. 37, 1026–1031�1998�.

11. E. W. Rothe and P. Andresen, “Application of tunable excimerlasers to combustion diagnostics: a review,” Appl. Opt. 36,3971–4033 �1997�.

12. L. F. Rubin and V. T. Gylys, “Measurement of the Raman crosssection of O2 �a1�g�,” Opt. Lett. 22, 1347–1349 �1997�.

13. B. J. Fang, F. T. Sang, and L. Y. Wei, “BHP surface tempera-ture rise and water vapour content in JSOG,” High PowerLaser Particle Beams 13, 27–30 �2001�, in Chinese.

14. Q. Zhuang, F. T. Sang, and D. Z. Zhou, Short WavelengthChemical Laser �Publishing House of National Defense andIndustry, Beijing, China, 1997�, p. 28, in Chinese.

15. W. E. McDemott, “Measurement of the O2�a1�� O2 �X3��ratio using Raman spectroscopy,” in Gas, Chemical, and Elec-trical Lasers and Intense Beam Control and Applications, S.Basu, S. J. Davis, and E. A. Dorko, eds., Proc. SPIE 3931,131–137 �2000�.

Fig. 7. Relation of xRT and proportion He:Cl2.

20 June 2003 � Vol. 42, No. 18 � APPLIED OPTICS 3527