The MIR experiment: status and perspectives C. Braggio on behalf of the MIR collaboration C. Braggio, G. Carugno, G. Galeazzi, G. Ruoso (PADOVA) F. Della Valle, G. Messineo (TRIESTE) F. Massa, D. Zanello (ROMA) A. Agnesi, F. Pirzio, G. Reali (PAVIA)
Outline Introduction to the MIR experimental scheme Optimization of the experimental parameters to maximize the gain Parametric excitation of a pre-charged field in the cavity Data analysis and discussion Perspectives and Conclusions
MIR experimental scheme Modulation of the surface CONDUCTIVITY of a semiconductor slab inside a microwave cavity C. Braggio et al Europhys. Lett. Vol. 70, 2005 Accelerating reference frame for electromagnetic waves in a rapidly growing plasma: Unruh-Davies-De Witt radiation and the nonadiabatic Casimir effect E. Yablonovitch Phys. Rev. Lett. Vol. 62, 1989 Parametric excitation of vacuum by use of femtosecond pulses Y. E. Lozovik, V.G. Tsvetus and E. A. Vinogradov Physica Scripta Vol. 52, 1995 Quantum phenomena in nonstationary media V. V. Dodonov, A B Klimov, D Nikonov Phys. Rev. A. Vol. 47, 1993
MIR experimental scheme Requirements: RF cavity geometry: highest frequency shift, smallest illumination area, f 0 2.5 GHz, Q 0 104-10 6 ; laser system: high frequency repetition rate (f rep 5 GHz, stability better than the cavity BW 1 khz), variable f rep, 10 ps pulses duration, Epulse few µj, 780-820 nm output wavelength; semiconductor: high mobility (1 m 2 /Vs), recombination time of a few ps; low noise microwave receiver: sensitivity of a few hundreds of 10-4 ev energy photons (i.e. 10-23 W/Hz) C. Braggio et al NIM Vol. A603, 2009
Theory V. V. Dodonov realistic MIR experimental conditions Results: a significant amount of Casimir photons (>10 3 ) can be produced in MIR experiment with feasible parameters calibration of the apparatus with a pre-charged field or with thermal photons thermal or external field E 0 N ph (n) E 0 e 2 χ max F(A 0 )n n number of pulses F(A 0 ) gain coefficient χ max connected to the frequency shift (0.005-0.008 if d = 600 µm) Dynamical Casimir effect at finite temperature G. Plunien, R. Schuetzhold, G. Soff Phys. Rev. Lett. Vol. 84, 2000 kt N th = = 8. 5T h ν meas N = G( 1+ 2Nth ) simulation
MIR experimental scheme 30 l 80 l LHe vessel, T = (0.8 9) K
microwave cavity Cavity geometries optimized for photon production Nb superconducting cavity- (80 x 90 x 9) mm 3 Cylindrical reentrant cavity 1. Smaller amount of E pulse E, H field profiles E, H field profiles Note that: - Semiconductor at x (E 0) (rectangular cavity); at x (E=E max ) (reentrant cavity) - Opposite sign of the frequency shift 2. Simplified optical scheme for a uniform illumination of the semiconductor
Semiconductor- 1 high mobility (1 m 2 /Vs), recombination time of a few ps R&D on a new material, starting from semi-insulating (SI) gallium arsenide SI GaAs irradiated with nuclear reactor neutrons (Italy, USA) --- Foulon et al J. Appl. Phys. 88 (2000) SI GaAs irradiated with Au, Br ions from Tandem in LNL, Padova Italy Mangeney et al Appl. Phys. Lett 80 (2002) SI GaAs irradiated with 1-5 MeV protons from CN accelerator in LNL Stelmakh et al Appl. Phys. 73 (2000) Measurement of the recombination time and mobility of the irradiated samples Optical pump-thz probe setup in SELITEC Vilnius, Lituania (prof Krotkus group) 1. Same concentration of free carriers produced as in the plasma mirror (n 10 17 cm -3 ); 2. Measurements are conducted at different temperatures in the range 300-10 K in a cryocooler
Semiconductor- 2 Measurements of the recombination time with the optical pump-thz probe setup: the THz transmission signal is connected to the modulation of the GaAs conductivity 240 MeV Br 14+ ions: 20 µm thickness of irradiated material Recombination time at different temperatures ions/cm 2 1-5 MeV protons: 100 µm thickness of irradiated material Recombination time ad different irradiation doses MOBILITY is inferred by comparison between the THz transmitted amplitude through the non-irradiated sample and the same sample after the irradiation procedure After the irradiation the mobility is 80% of the substrate value at the same temperature (highest at 80 K)
Laser system - 1 Epulse 10 µj; since the average power of a CW mode-locked laser having this value of energy per pulse would be too high, we developed a laser delivering a macropulse of T=350-450 ns duration. The highest output macropulse to generate the greatest number of pulses. FINAL SPECS - high frequency repetition rate (f rep better than the cavity BW 1 khz), - variable f rep, - 10 ps pulses duration, - Epulse few µj, - 780-820 nm output wavelength 5 GHz, stability Optics Express 13, 5302 (2005) Optics Express 14, 9244 (2006) Optics Express 16, 15811 (2008)
Laser system - 2 LASER repetition rate stability and variation Active control of the M_OSC length: the feedback system locks the repetition frequency of the laser to a reference microwave generator Short term stability < Nb cavity linewidth (1 KHz) Long term stability
Laser system - 3 OPO train of pulses uniformity 20 GHz sampling oscilloscope; 20 ps rise time photodiode uniform illumination of the slab OPO field profile
Frequency shift - 1 The total number of produced photons depends exponentially on the frequency shift N ph ( n) E 0 e 2 χ max F ( A 0 ) n Maximizing the frequency shift: results from Ansoft HFSS (stationary boundary conditions) L cavity length D semiconductor thickness
Frequency shift - 2 GEOMETRY A GEOMETRY B
Frequency shift - 3 Measuring the frequency shift Δν = ν 0 - ν ill Non-irradiated semiconductor disk, characterized by τ (room temperature) few ns at T = 290 K at T=77 K the free carriers plasma lasts for Δt 100 µs ν ill at T=77K persists for at least 100 µs Δt τ E 500 ns ν 0 = 2.3384 ± 14 khz; ν ill = 2.3272 ± 44 khz Δν =ν 0 - ν ill = 11 MHz Q 0 = 6200 ± 100; Q ill = 1500 ± 100 ν 0 = 2.3295 ± 50 khz; ν ill = 2.30308 ± 50 khz Δν =ν 0 - ν ill = 26.4 MHz Q 0 = 8500 ± 100; Q ill = 2600 ± 100 This measurement is not sufficient to characterize the moving mirror
Frequency shift - 4 In the previous measurement the free carriers are present in the overall GaAs volume. During the experiment we use instead a short τ very thin film of photoexcited carriers Thickness of the plasma is determined by: I=I 0 exp(-αx) at T=77K for λ= (810 ±10) nm, the absorption coefficient is α -1 =1µm Thus the definitive proof that the mirror is efficient is given by comparisons of ν 0 and Q 0 in these two cases: ν 0 = 2.32845 ± 100 khz ν Cu = 2.3005 ± 100 khz Q 0 5000 ± 100 Q Cu 5000 ± 100 Δν = ν 0 - ν Cu 28 MHz We obtain the same Δν foreseen by the simulations and measured with the volume plasma
Parametric excitation of an external field - 1 Is our parametric amplifier characterized by a gain allowing the observation of the vacuum contribution? 1. The cavity unperturbed mode ν 0 is critically coupled to the transmission line; the cavity is pre-charged. 2. the radiofrequency at frequency ν 0 of the unperturbed cavity is switched off and the EM field starts to decay with decay time τ E (free oscillations); 3. after 100 ns the external generator has been switched off, during the field decay, the laser train of pulses impinges on the semiconductor surface A(laser OFF ) GAIN of the parametric amplifier=a(laser ON )/A(laser OFF ) A(laser ON )
Experiment simulation: RLC circuit - 1 pulse generator train of pulses this circuit generates a voltage V2 which is then used to mathematically modify C T exc excitation period temporal shape of the excitation (rise/decay time - duration) TD delay line to change the phase between the excitation and the free oscillations RLC circuit unperturbed cavity ν 0 free oscillations frequency =1/ 2π(LC) 1/2 Q 0 quality factor of the unperturbed cavity =R/ω 0 L RLC with variable capacitance par.excited cavity coefficient of V(2) determines ΔC, corresponding to a frequency shift in the cavity of Δν =1/2 ν ΔC Q exc quality factor of the perturbed cavity =R(parallel)/ω 0 L
Experiment simulation: RLC circuit - 2 Transient analysis: evolution of an initial applied voltage level pre-charged field Example FREE OSCILLATIONS ν 0 = 2516.51 MHz; Q 0 = 5000 MODULATION frequency shift = 8.5 MHz 2 ps rise time - 8 ps duration - 2 ps decay time at parametric resonance exponential growth of the field amplitude GAIN = ratio between the initial voltage and its amplitude after 400 ns G = 300V/V IC = 1V number of pulses
Experiment simulation: RLC circuit - 3 Texc = varied between 198.52 ps to 199.06 ps 1. exponential growth at the parametric resonance; decaying oscillations when detuned 2. higher order maxima in the gain vs delta f plot are observed. These peaks arise due to modulations of the field in the cavity at exactly f mod = f laser - 2ν 0 3. the maximum of the amplification is found with f laser =2 ν 0 + Δf (in our experiment the excitation is not a pure harmonic signal); 4. If the phase between the excitation and the field in the cavity differs, at every laser shot a different gain is observed; In agreement with V.V. Dodonov theoretical results Phys. Scr. T143 (2011)
Parametric excitation of an external field - 2 Is our parametric amplifier characterized by a gain allowing the observation of the vacuum contribution? To test the apparatus and measure its gain we operate the cavity at 77 K. The amplitude of the signal strongly depends on the phase The average value of the output amplitude (with respect to the input phase) and its standard deviation as functions of detuning from the resonance frequency show decaying oscillations.
Parametric excitation of an external field - 3 EXPERIMENTAL RESULTS + Br ions-irradiated sample τ = 7.2 ps - frequency shift 11 MHz - tuning the cavity, laser f rep fixed G = A(lasON)/A(lasOFF) < 1 ---> losses Maximal standard deviation is observed for the same resonance frequency as the average amplitude itself
Parametric excitation condition Matching the parametric resonance condition: cavity tuning: a saffire rod changes the cavity resonance frequency laser tuning: active control of the M_OSC length
Parametric excitation of an external field - 4 EXPERIMENTAL RESULTS + frequency shift 28 MHz + sapphire substrate + mainly tuning the laser - proton-irradiated sample τ = 1 ps The average value of the output amplitude and its standard deviation as functions of detuning from the resonance frequency show decaying oscillations.
Parametric excitation of an external field - 5 DATA ANALYSIS AND INTERPRETATION We fit the experimental curves with recent theoretical results obtained by V.V. Dodonov x(i) = (δ(i Res ) δ(i))/((w/2 δ(i)) δ(i Res ) /w/2)(1 δ(i Res ) /w/2 )) δ(i)= w/2 w o (i) w Res (i)/2= w o (i)/(1+ϕ/π) ϕ= 2π w o (i Res ) χ τ w is the laser frequency; w 0 is the cavity frequency ϕ phase, χ τ depending on the semiconductor LOSSES Both the samples had no AR coating more than 30% of the light is diffused in the cavity continuous background of free carriers during the laser excitation
Coming soon Irradiation procedure - better control of the proton or ion beam Optical pump-thz probe setup in our laboratory Apply the antireflection coating to the good irradiated samples and repeat the gain measurement;
Conclusions and perspectives The preliminary results of the parametric amplification of a classical field agree with theoretical predictions Once a G>1 amplification of a classical field is observed, the system is in principle ready to detect the quantum effect varying the cavity temperature in the range 0.8 K < T < T c. Study of the thermal photons statistics simulation
The microwave receiver TE 101 mode, ν 101 = c/2 (b-2 + L -2 ) ~ 2.5 GHz E q = 10-5 ev energy of a single photon RADIOMETER R H Dicke (1946) 1. Antenna (dipole/ inductive loop) matched to the line* 2. Thermal equilibrium 3. Johnson noise in the resistor = blackbody rad from the walls P = 4R k Δf G (T R +Tn) Receiver sensitivity (2 ± 0.2) X 10-22 W/Hz
Chi-squared method hp: the vacuum term does not exist Principle of measurement graph (fake N and fake errors!) ( ) 2 χ 2 7 N = m 2N 0 T i G = i=1 P(χ 2 > c) = 0.005 7 1 χ 2 > 19 σ 2 degrees of freeedom 7 G 2 i=1 σ 2 2 2 σ = σ s + σ σ 2 s 2 stat σ 2 stat receiver sensitivity error thermal and vacuum photon statistics error If the chi-square is low, we can decrease the receiver sensitivity error via AVERAGING: Es: σ avg = σ s N Poisson distribution; G=100 AVG 150 for each temperature If the statistics is super-poissonian?