Z. Phys. D - Atoms, Molecules and Clusters 21,209-213 (1991)

Atoms. Molecules ze,,=,,, fLir Physik D and Clusters © Springer-Verlag 1991

Lifetime measurements of the 2p 3d 3D°1, 2, 3 and 2p 3d leO levels in NII by beam-foil-laser spectroscopy P.-D. Dumont, H.-P. Garnir, and Y. Baudinet-Robinet* Institut de Physique Nucl6aire Exp6rimentale, Universit~ de Liege, Sart Tihnan B 15, B-4000 Liege, Belgium Received 13 March 1991

Abstract. Accurate lifetimes measured by means of the cascade-free method based on laser excitation of a fast ion beam preexcited in a carbon foil are reported for four 2p3d levels in NII. The lifetime results are:

r(2p3d3D°)=O.209+_O.OO7ns,

r(2p3d3D°)=0.219

± 0.007 ns, r (2p 3 d 3D°) = 0.217 _+0.005 ns, and z (2p 3d 1Pl°) = 0.410 -4-0.017 ns. These results are compared to theoretical and experimental lifetimes reported previously. PACS: 32.70. Fw; 32.70. Cs; 32.90. + a

1. Introduction Accurate lifetimes can be measured using the Beam-FoilLaser (BFL) method which combines a first nonselective excitation of a fast ion beam in a carbon foil with a second selective excitation by a tunable cw dye laser beam. The detailed description of the method is given in [ 1, 2] and will not be repeated here. This method has been applied previously to measure accurate lifetimes in CIII, CIV [2], NII [3], and OIII [4] ions. Lifetime values for the 2p3d3D°,2,3 and 2p3d ipO levels have been calculated - some very recently - using different methods [5-9]. Lifetimes have been measured for these levels by the standard beam-foil method only [10, 111. However, it is well known that beam-foil lifetimes can be unreliable in the presence of cascading. In the present work, cascade-free decay curves reducing to single exponentials are recorded for the 2p 3d 3D°1 , 2 , 3 and 2p 3d ipO levels in N I I using BFL spectroscopy. Accurate lifetimes are deduced from the analyses of these decay curves.

2. Experiment The experimental arrangement is similar to that described in detail in [ 1,2]. A N2+ beam - supplied by a 2 MV Van * Senior Research Associate of the Belgian FNRS

de Graaff accelerator - crosses at right angles successively a carbon foil and the intracavity beam of a cw dye laser pumped by a 20 W argon laser. The dye laser wavelength situated in the visible region is tuned to resonance with a transition in an ion emerging from the foil. In the present work, population changes produced by the laser field in the upper level of the induced transition are measured. These measurements are made by recording the difference of intensities, with and without laser excitation, for a line emitted in the VUV region from the upper level of the laser induced transition. This difference of intensities is measured as a function of the distance travelled by the ions downstream from the laser interaction region. The whole experiment is controlled by a network of small computers and all the data are collected and analysed on line [12]. In our experiment, the recorded photons are analysed with a Seya-Namioka type spectrometer and detected by channeltron detectors. The distance between the foil and the laser beam is chosen to obtain strong laser effects (strong population changes of the levels studied in the laser field) [3]. The intracavity power of the dye laser is controlled during the measurements so that the number o f recorded photons can be normalized to take into account small variations of this power. The experimental conditions are summarized in Table 1. For the 2p 3d 3D°3, 2p 3d 3D°2, and 2p 3d 3D° lifetime measurements, the dye laser is tuned to resonance with the 2p3p 3P2-2p3d3D° 3 transition (2 =594.165nm), 2p3p 3pl-2p3d 3D° transition (2 = 593.178 nm), and 2p3p 3Po-2p3d3D° transition (2 =592.781 nm), respectively. The laser-on and laser-off intensities o f the 2p 2 3p-2p 3d 3D° multiplet line at 53.4 nm are recorded in the three cases (the multiplet components of the 53.4 nm line were not resolved). Figure 1 shows the transitions involved in the 2p 3d 3D° 2. 3 lifetime measurements, together with approximate lifetime values of the levels coupled by the laser field. For the 2p 3d 1pO lifetime measurement, the dye laser is tuned to resonance with the 2p 3p 1D2-2p 3d ipo transition at 2 = 628.43 nm. The laser-on and laser-off intensities of the 2p 2 1So-2p3dlP°1 line at 63.52nm are

210 Table 1. Summary of experimental conditions N + beam energy (MeV) Ion beam cross section (mm2) Dye Dye intracavity power (W) Dy laser line FWHM (nm) Dye laser beam diameter (mm) Carbon foil thickness (txg/crn2) Carbon foil size (ram~) Ion beam portion analysed (mm) Distance between foil and laser beam axis (ram)

recorded. Figure 2 shows these transitions and approximate lifetime values of the coupled levels. In both experiments, the lifetime of the upper level of the laser induced transition is much shorter than the lifetime of the lower level of this transition. Thus, the population of the upper level decreases strongly on the distance ( ~ 3 m m ) between the foil and the laser beam while the population of the lower level is nearly constant. In this case, the upper level studied is strongly repopulated in the laser field. However, to record a significant BFL decay curve, it is necessary to accumulate several BFL decay curves, each obtained in a sufficiently short time for maintaining the experimental conditions (laser power, foil thickness .... ) unchanged. For each lifetime measurement, m a n y B F L decay curves of approximately equal statistical weights are recorded. In order to cancel small systematic errors which could be due to changes of the foil during the irradiation time required to record a single decay curve, half of the curves are taken where the foil-laser interaction region is moved upstream along the ion beam direction and the other half downstream. No significant difference has been observed between these two groups of curves. For each curve, twelve data points along a total dis-

NII 2p3d3D °

NII 2p3d 1po

1.0, (1.5) 0.6 × 6.0 Rhodamine 6G ~ 20 ~ 0.1 ~1 ~20, ~ 12 1.5 × 7.0 ~0.6 ~3

1.0 0.6 × 6.0 DCM ~ 20 ~ 0.1 ~1 ~ 15 1.5 × 7.0 ~0.6 ~3

tance of 2.2 m m are taken but for the 2p 3d 3D° lifetime measurement, a few curves are measured along a distance of 1.8 m m only instead of 2.2 mm. The first decay curves for the 2p 3 d 3D° lifetime measurement are obtained with 1.5 MeV N ~ ions. However, more statistically significant intensity changes (laser on - laser off) are observed by accelerating N + ions to 1 MeV instead of 1.5 MeV. Thus, in the remaining of the work, 1 MeV N2+ ions are employed. We utilize carbon foils of thickness between 10 and 20 g g / c m 2 prepared by ethylene cracking at the University of Bochum (Germany). These foils have a lifetime much longer than the lifetime of the carbon foils prepared by vapor deposition and used in our B F L lifetime measurements reported previously [2, 3, 4].

3. Results and discussion 3.1. 2p 3d 3D 1,2,3 ° levels in N H 3.1.1. Lifetime result. 2 p 3 d 3D°3 . We have recorded 31 BFL decay curves for measuring the 2p 3d 3D° lifetime. F o r each curve, the numbers of photons counted - at the

N II

N II

2p 3d ]P~ =0.4ns

2p3d3D0 =0,2ns

594.165nm(2~3) J / 593,178nm(1~2) / 592,781n m / 2p3p3p

/

=6ns

2p23p

/

Fig. 1. Observed transitions and laser induced transitions involved in the lifetime measurements of the 2p3d3D °, 2p3d 3D°, and 2p 3d 3D° levels

/

63.52nm

=53.4nm

2p2 ]So

Fig. 2. Observed transition and laser induced transition involved in the lifetime measurement of the 2p 3d ip0 level

211 where the quoted error represents the standard deviation given by the least-squares fitting program. The mean value of the lifetimes estimated from single exponential fits of the 39 B F L decay curves is r = 0.220 _4-0.005 ns "

where the error represents the standard deviation of the mean of the sample. This lifetime value is in very good agreement with the lifetime estimated from the cumulated curve. The mean value of the two lifetime results is presented in Table 2.

1000

100

~

~

J

~

I 0.5

~

~

~

~

~

~

~

i

1

i

I

I

~

~

1.5

Distance (ram) Fig. 3. BFL decay curve for the NII 53.37 nm line recorded using 1 MeV N + ions and 2p 3p 3p2-2p 3d 3D° laser induced transition (see text). The solid line is a single exponential fit to the data. The error bars represent the statistical errors in the data (one standard deviation)

exit of the laser intraction region - with and without laser interaction were ~ 1000 and ~ 500, respectively. All these curves have been well fitted to a single exponential. A typical cumulated BFL decay curve is shown in Fig. 3. This curve results from the accumulation of all the BFL decay curves (21 curves) recorded with 1 MeV N + ions. The lifetime resulting from the adjustment of this cumulated BFL decay curve to a single exponential (X 2 = 1 . 4 ) is -c = 0.217 ± 0.004 ns The quoted error represents the standard deviation given by the least-squares fitting program. The mean value of the 21 lifetimes estimated from the fits of the individual BFL decay curves to single exponentials is r = 0.218 _+ 0.004 ns The error represents the standard deviation of the mean value deduced from the dispersion of the sample. As expected, the two analyses lead to the same results. The mean value of the lifetimes estimated from single exponential adjustments of all the B F L decay curves recorded (both 1 and 1.5 MeV N2+ ions are used) is given in Table 2. This value is in perfect agreement with the lifetime value estimated from decay curves recorded with 1 MeV N + ions only.

2p 3d 3D°2 . F o r each of the 39 B F L decay curves recorded for the 2 p 3 d 3D° lifetime measurement, the number of photons counted - at the exit of the laser interaction region - with and without laser interaction were ~ 9 0 0 and ~ 550, respectively. All these curves have been well fitted to a single exponential. The lifetime estimated from the fitting of the cumulated B F L decay curve (curve resulting from the summation of the 39 curves) to a single exponential (Z 2 = 1.3) is r = 0.218 _+ 0.005 ns

2 p 3 d 3D°. For the 2 p 3 d 30o lifetime measurement, we have recorded 69 BFL decay curves containing each ~ 400 photon counts (,~ 2000 photon counts when the laser is on and ~, 1600 when the laser is off) - at the exit of the laser interaction region. The lifetime estimated from the adjustment of the cumulated BFL decay curve (summation of the 69 curves) to a single exponential 0C 2 = 1.1) is r = 0.207 _+0.005 ns where the quoted error represents the standard deviation given by the fitting program. The mean value of the 69 lifetimes estimated from single exponential fits of the individual decay curves is r = 0.212 _%0.006 ns where the error represents the standard deviation of the mean of the sample. The lifetime values obtained by the two methods are in agreement, within statistical errors. In Table 2, the quoted lifetime represents the mean value of the two lifetime results.

3.1.2. Discussion. The theoretical [5-9] and experimental [10, 11] lifetime results available for the 2 p 3 d 3 D °1,2,3 levels (or 2 p 3 d 3 D ° term) in N I I are presented in Table 2. The theoretical values are very scattered. Our results are in very good agreement with Fawcett's calculated values [8]. The previously measured lifetimes [10, 11] - for the 2 p 3 d 3 D ° term only - have been obtained by standard beam-foil spectroscopy. These lifetimes were deduced from two- or three-exponential fits of decay curves recorded for the 53.4 nm line. With the B F L method we have been able to measure the lifetimes of the three levels of the 2p 3d 3D° term. These lifetime values are not significantly different. The lifetime of the 2p 3d 3D° term deduced from our BFL lifetimes is shorter and more accurate than previously reported beam-foil values.

3.2. 2 p 3 d tpo level in N H 3.2.1. Lifetime result. Due to the weak intensity of the observed 63.53 nm line, it was necessary to record a great number ( ~ 2 0 0 ) of B F L decay curves to obtain an accurate value for the lifetime of the 2p 3d 1pO level. Each curve contained ~ 5 0 photon counts ( ~ 3 0 0 photon

212 Table 2. Lifetime results for the 2p 3d aDo,

2p 3d 3D°, and 2p 3d 3D~ levels in NII

Lifetime (ns)

Method

2p 3d 3D°

2p 3d 3D°

2p 3d

0.27 0.257 0.2P 0.12

0.27 0.257 0.23 a 0.12

0.27 0.257 0.21a 0.12

, , 0.209 _+0.007 °

0.24 - 0 . 0 4 b

0.241 + 0.012 b 0.219 +_0.007 c

3D3°

, '

0.217 _+0.005c

Theory Restricted Hartree-Fock [5, 6] Polarized frozen-core H.-F. [7] Hartree-Fock relativistic [8] Model potential [9] Experiment Beam-foil [10] Beam-foil [11] Beam-foil-laser (this work)

Neglecting the very weak 3p 3p, 3D_3d3D0 transition probabilities [6, 7] not given in [8] b Unresolved multiplet components ° The quoted error represents the statistical error (one standard deviation) combined with a 2% uncertainty in the ion velocity

a

counts when the laser is on and ~ 2 5 0 when the laser is off) - at the exit o f the laser interaction region. A sample o f 14 B F L decay curves was then o b t a i n e d by accumulating several individual decay curves. Each o f these 14 B F L decay curves has been well adjusted to a single exponential. The weighted m e a n lifetime deduced f r o m the sample o f the 14 estimated lifetimes is

Table 3. Lifetime results for the 2p 3d 1po level in NII Lifetime (ns)

Method

0.50 0.483 0.32 a 0.20

Theory Restricted Hartree-Fock [5, 6] Polarized frozen-core H.-F. [7] Hartree-Fock relativistic [8] Model potential [9]

0.41 Jr 0.06 0.473 + 0.024 0.410 + 0.017 b

Experiment Beam-foil [ 10] Beam-foil [ 11] Beam-foil-laser (this work)

r = 0.412__ 0.010 ns where the q u o t e d error represents the standard deviation o f the m e a n value. We have also s u m m e d all the B F L decay curves recorded and obtained the cumulated curve given in Fig. 4. The lifetime estimated f r o m the fit o f this curve to a single exponential (Z 2 = 0.5) leads to

a Neglecting the very weak 2p3p IS, ~P, ID-2p3d 1p0 transition probabilities [7, 9] not given in [8] b The quoted error represents the statistical error (one standard deviation) combined with a 2% uncertainty in the ion velocity

z = 0.407 +_ 0.015 ns The lifetimes obtained by the two methods are in agreement, within statistical errors. In Table 3, the quoted result is the m e a n value o f the two lifetime results.

10000

3.2.2. Discussion. The theoretical and experimental lifetime values available for the 2p 3d ~pO level are presented in Table 3. The theoretical values vary between 0.2 and 0.5 ns and our result is n o t in agreement with any o f these values. T w o beam-foil lifetimes have been reported for the 2p 3d 1P1° level. These two results were obtained f r o m two-exponential fits o f beam-foil decay curves. O u r B F L lifetime is shorter than the m o r e recent and precise beamfoil value reported in [11 ].

4. Conclusions g

1000

i

,

i

i

L 0.5

. . . .

i 1

. . . .

i

. . . .

1.5

i 2

. . . .

2.5

Distance (ram)

Fig. 4. BFL decay curve for the NIt 63.53 nm line recorded using 1 MeV N + ions (see text). The solid line is a single exponential fit to the data. The error bars represent the statistical errors in the data (one standard deviation)

We have measured the first accurate lifetimes for the 2P 3d 3D°, 2P 3d 3D°2, 2p 3d 3D3,0 and 2p 3d 1pO levels in N i l . These lifetime values were obtained using the cascade-free B F L method. The B F L lifetimes measured in this w o r k are shorter than those measured previously with the standard beam-foil m e t h o d (see Tables 2 and 3). N o t e that the lifetimes o f the different J-levels o f the 2p 3d 3D° term have not been measured separately by standard beam-foil spectroscopy. The 2p 3d 3D°, 2/9 3d 3D°, and 2p 3d 3D° lifetime values o f the present w o r k are not significantly different.

213 They are in good agreement with the recent values of Fawcett [8] calculated with a Hartree-Fock-Relativistic program package (see Table 2). The 2p 3d 1pO B F L lifetime result is situated between the value of McEachran and Cohen [7] calculated by the polarized frozen-core version of the Hartree-Fock approximation and the value of Fawcett [8] calculated with a Hartree-Fock-Relativistic program package (see Table 3). We thank Dr. H.H. Bukow (University of Bochum, Germany) for preparing long-lived carbon foils by ethylene cracking. This work was supported by the Belgian Institut Interuniversitaire des Sciences Nucldaires and the Universit6 de Liege.

References 1. Dumont, P.D., Garnir, H.P., Baudinet-Robinet, Y., E1 Himdy, A.: Nucl. Instrum. Methods B35, 191 (1988)

2. Baudinet-Robinet, Y., Dumont, P.D., Garnir, H.P., E1 Himdy, A.: Phys. Rev. A40, 6321 (1989) 3. Baudinet-Robinet, Y., Garnir, H.P., Dumont, P.D., R~simont, J.: Phys. Rev. A42, 1080 (1990) 4. Baudinet-Robinet, Y., Dumont, P.D., Garnir, H.P. : Phys. Rev. A43, 4022 (1991) 5. Kelly, P.S.: Astrophys. J. 140, 1247 (1964) 6. Wiese, W.L., Smith, M.W., Glennon, B.M. : Atomic transition probabilities, Natl. Stand. Ref. Data Ser., Natl. Bur. Stand. (U.S.) 4, vol. 1: U.S. GPO, Washington, D.C. (1966) 7. McEachran, R.P., Cohen, MJ.: Quant. Spectros. Radiat. Transfer 27, 119 (1982) 8. Fawcett, B.C. : At. Data Nucl. Data Tables 37, 411 (1987) 9. Victor, G.A., Escalante, V. : At. Data Nucl. Data Tables 40, 227 (1988) 10. Buchet, J.P., Poulizac, M.C., Carrd, M.: J. Opt. Soc. Am. 62, 623 (1972) 11. Kernahan, J.A., Livingston, A.E., Pinnington, E.H.: Can. J. Phys. 52, 1895 (1974) 12. Monjoie, F.S., Garnir, H.P. : Comput. Phys. Commun., 61,267 (1990)