Laser plasma diagnostics in rubidium vapor cell J.S.Bakos, - PowerPoint PPT Presentation

HAS WIGNER RESEARCH CENTRE FOR PHYSICS Laser plasma diagnostics in rubidium vapor cell J.S.Bakos, G.P.Djotyan, G.Demeter, P.N.Ignácz , M.Á.Kedves , B.Ráczkevi , Zs.Sörlei , J.Szigeti, K.Varga-Umbrich A.Czitrovszky , A. Nagy, P. Dombi, P. Rácz

Motivation and issues found in measurements • Plasma generation in Rb vapor by ultrashort laser pulses • Plasma diagnostics by CW diode lasers: plasma density Why rubidium cell? Why diode laser? Easily vaporized Rb Commercially available Convenient spectral lines Cheap (CD writer 780nm) 780nm Simple to operate Vapor density simply Easy frequency tuning controllable by temperature 2

Another simple plasma source Rb vapor source : getter @ double Rb vapor distribution slit above the slit 3

Direct way of plasma diagnostics: collecting charged particles Langmuir flat probe: 1. Simple design 2. Ions/electrons come from ‘ somewhere ’ , calibration difficulties 4

Results of direct ion detection ionization dependence on laser intensity Slope: ~ 2 Maximum laser intensity: 10 11 W/cm 2 5

Indirect plasma diagnostics: plasma = ‘ lack ’ of neutral atoms Atomic processes: Decay time some 10 ns Population dynamics for a pair of resonant pulses 6

Plasma diagnostics by CW diode lasers Atomic Lorentz model: exp( E  i   i 0 t  )  2 1 ia  exp( 1 sin   t i  1 B )  resonant absorption @ exp(  ia 2 sin t    i 2 B ) exp(   i t 0 ) interferometry 7

Experimental layout Parameters of the Ti:Sa laser Mean wavelength 806 nm Courtesy of Beam Diameter:9 mm (1/e2Gauss) A. Czitrovszky, P. Dombi, Polarisation:Linear, vertical P. Rácz, A. Nagy, I. Márton Repetition Rate 1 kHz Pulse duration (FWHM):35 fs 8 Pulse 3.5 mJ

Experimental layout 9

Vapor cell, heating wires, reflector Temperature Courtesy distribution A. Bendefy (BME) 10

Spectroscopic observations Detection of the radiation of the plasma by a fast spectrograph (Andor Mechelle 5000) High spectral resolution (0.05 nm accuracy) High temporal resolution with intensified camera (~ ns) Spectrograph courtesy of L. Kocsányi (BME), and help with the measurements R. Bolla (WRCP) 11

Observed spectral lines of Rb 12

Time dependence of the spectral emission Temperature: ~ 200 C o Ion relaxation mainly through D2 lines (and D1) 13

Transversal absorption measurements Parameters: 1. Ionizing laser intensity 2. Probe laser detuning 14 3. Vapor density

Tipical transmission signals on microsec scale (different) CW level: Positive peak @ Negative peak and relax. (New Focus 1591NF): 4.5 GHz Very fast peak: AC Stark shift 10 ns decay: atomic relaxation Slow (1-10 microsec) decay: plasma relaxation Decrease of transmission is attributed to reflection on the boundaries of the plasma channel. Initial condition: atoms in the ground state 15

Detuning:Rubidium frequency reference 16

Dependence of the fast peak maxima on the laser frequency 17

Slow relaxation component (negative peak) at different vapor densities 18

Transmission signal vs. vapor density Signal oscillations ? Plasma freq.100 GHz Repeated reflections on the 19 boundaries of the plasma channel

Summary of transmission signal detection Cw level is different for detector 1 and detector 2: - beam divergency, different coupling into the detector fiber - condensed Rb on the windows surface - different vapor temperature and density - CW signal is absorbed close to the resonance lines - Negative peak sygnal is missing far from the atomic lines - ‘ Plasma oscillations ’ 20

Plasma density measurements by longitudinal interferometry 21

Phase variation        I ( t ) I ( t ) I 2 I ( t ) I cos( ( t )) int erf tr ref tr ref 0 1      ( t ) ( 2 / ) L n ( t ) Phase variation 1    ( i ) 2 2 2 N fe   Refractive     j n ( ) 1 p      i ( i ) ( i ) 2 2 index 2 m ( )   i 1 j 1 0 j j Plasma    ( i ) 2 2 2 m   density    j L N ( t ) L ( t ) /[ p ]       * p 1 i 2 2 ( i ) ( i ) 2 2 ( ) fe   i 1 j 1 0 j j length 22

Phase variation @ Doppler broadening 2 /     Nfe m     0 ( ) 1 n     2 2      ( )           n ( ) 1 D ( ) d       0 2 2 ( )        2       Nfe /( m ) ( D ) Comparative function: ( ) ( ) / ( ) N t N t 0 0 p p Absorption coefficient 1 2 /      2 D ( ) e 0   0 Doppler broadening Normalized detuning 23

Results: time dependent fringes   =2.0x10 11 cm -3 7.8% 3.6 rad N =2.4x10 11 cm -3 10.8% 5.4 rad N   N =1.1x 10 12 cm -3 10.6% 24.8rad 24 N =1.0x10 12 cm -3 8.5% 18.3 rad

Plasma relaxation        I ( t ) I ( t ) I 2 I ( t ) I cos( ( t )) int erf tr ref tr ref 0 1 dN 1      t / N N e N 0 N Diffusion model D 0  dt 3 body N dN     2 0 N B N 1   recombination 3 N t dt 0 model dN N     3 0 N N BP    Pitaevski: N 3 1   dt 2 2 N t 0 N        t / 0 ( 1 ) N N e N 0 P. Muggli MPP   PM 0 1 N t 0 25

Curve fitting for diffusion @ 3body model    t / N N e N 0 D 0 N  0 N B 1   3 N t 0 26

Curve fitting for mixed models N        t / 0 N N e ( 1 ) N 0   PM 0 1 N t 0   1          t / N N e ( 1 ) N 0  D 3 BP 0     2 1 2 N t   0 27 Fit courtesy of M. Kedves

Interpretation of decay time Knudsen regime: mean free path ~ characteristic length Intermediate state between molecular flow and viscous flow 1 Mean free path in Rb vapor:  L  2 2 d N  L 4.5cm at 120 C o and 2x10 13 cm -3    10 atomic diameter for Rb d 5 x 10 m L 20.9cm at 95 C o and 4,3x10 12 cm -3 Rb vapor cell below 120 C o : quasi collisionless flight of atoms: Probe beam channel is filled with neutral atoms out of the channel 3 k T  330 m/s at 95 C o and 340 m/s at 120 C o B ~ v rms m dN 1      t / Linear kinetic N N N e N 0 Exponential decay  0 equation dt N 0     l / L N / N ( 1 e ) l 1 cm Ratio of collisions on a characteristic length C 0 0  at 95 C o N C / N 0 . 05 The greater the density 28 0  at 120 C o The greater the decay time N C / N 0 . 2

Thank you for your attention 29