arxiv: v2 [quant-ph] 29 Mar 2018

Transcription

1 The Commutation Relation for Cavity Moe Operators arxiv: v2 [quant-ph] 29 Mar 2018 Fesseha Kassahun Department of Physics, Ais Ababa University P. O. Box 33761, Ais Ababa, Ethiopia October 21, 2018 We have obtaine the commutation relation for a cavity moe riven by coherent light an interacting with a two-level atom. We have foun that the commutation relation for a free cavity moe with or without photons is just the usual commutation relation. However, the commutation relation for a cavity moe, which is interacting with a two-level atom, turns out to be ifferent from the usual one. With the operator ĉ efine by ĉ = â+iˆb, the commutation relation for ĉ an ĉ has been also iscusse. It appears that the usual commutation relation [â,â ] = 1 [1] is taken to be the commutation relation for a cavity moe regarless of whether the cavity moe is interacting with an atom or not. In orer to check the valiity of this viewpoint, we seek here to obtain the commutation relation for a cavity moe riven by coherent light an interacting with a two-level atom. The interaction between the cavity moe an the riving coherent light can be escribe by the Hamiltonian Ĥ = iλ(ˆbâ ˆb â), (1) where â the annihilation operator for the cavity moe, ˆb is the annihilation operator for the riving coherent light, an λ is the coupling constant between the cavity moe an the riving coherent light. In aition, the interaction between the cavity moe an the two-level atom can be escribe at resonance by the Hamiltonian [2] where Ĥ = ig(ˆσ â â ˆσ), (2) ˆσ = b a (3) is a lowering atomic operator an g is the coupling constant between the cavity moe an the atom. Here a an b are the upper an lower levels of the two-level atom. We thus see that the Hamiltonian escribing the interaction of the cavity moe with the riving coherent light an the two-level atom has the form Ĥ = iλ(ˆbâ ˆb â)+ig(ˆσ â â ˆσ). (4) We consier the case in which the cavity moe is couple to a vacuum reservoir via a singleport mirror. In aition, we carry out our calculation by taking into account the noise operators associate with the vacuum reservoir. aress: fesseha.kassahun@aau.eu.et.

2 We recall that the operators in the Hamiltonian escribing the interaction of a two-level atom with a single-moe light are fixe at the initial time. It then appears to be appropriate to fix the Hamiltonian given by (4) at the initial time, whether we are working in the Schröinger or Heisenberg picture. The time evolution of the operator â(t) is escribe by the quantum Langevin equation [3] tâ(t) = κ 2â(t) i[â(t),ĥ(0)]+ ˆF(t), (5) where κ is the cavity amping constant an ˆF(t) is a noise operator with vanishing mean. In orer to evaluate the commutator of â(t) an Ĥ(0), we introuce ˆρ(0), which is the ensity operator for the system uner consieration at the initial time. We observe that ˆρ(0)[â(t),Ĥ(0)] is in the Heisenberg picture. This can also be written in the Schröinger picture as Now applying the Hamiltonian given by (4), we fin Hence in view of this result, Eq. (5) takes the form ˆρ(0)[â(t),Ĥ(0)] = ˆρ(t)[â(0),Ĥ(0)]. (6) [â(t),ĥ(0)] = iλˆb(t) igˆσ(t). (7) tâ(t) = κ 2â(t)+λˆb(t) gˆσ(t)+ ˆF(t). (8) With the intention of obtaining the solution of this equation, we replace ˆb(t) by the c-number β, consiere to be real an constant. Thus making use of this replacement, Eq. (8) can be put in the form (9) in which ε = λβ. In aition, employing the relation Â(t) = i [Â(t),Ĥ(0)] (10) t along with Eq. (4), one reaily obtains (with the time argument suppresse) where t ˆσ = g (ˆη b ˆη a )a, (11) t ˆη a = g ˆσ â +g â ˆσ, (12) ˆη a = a a, (13) ˆη b = b b. (14) We see that Eqs. (11) an (12) are nonlinear ifferential equations an hence it is not possible to obtain exact solutions of these equations. We seek to overcome this problem by making use of the so-calle large-time approximation scheme [4]. Hence applying this approximation scheme, we obtain from Eq. (9) the approximately vali relation â(t) = 2g κ ˆσ(t)+ 2ε κ + 2 κ ˆF(t). (15)

3 Now substitution of (15) into the aforementione equations yiels where t ˆσ = 1 2 γ c ˆσ + 2gε κ ˆη b ˆη a + 2g κ (ˆη b ˆη a )ˆF, (16) t ˆη a = γ c ˆη a + 2gε κ ˆσ + ˆσ + 2g κ ˆσ ˆF + ˆF ˆσ, (17) γ c = 4g 2 /κ. (18) Assuming that the atomic an noise operators are not correlate, we have (ˆη b ˆη a )ˆF = ˆη b ˆη a ˆF = 0, (19) ˆσ ˆF = ˆσ ˆF = 0, (20) ˆF ˆσ = ˆF ˆσ = 0. (21) We therefore see that t ˆσ = 1 2 γ c ˆσ + 2gε κ ˆη b ˆη a. (22) t ˆη a = γ c ˆη a + 2gε κ ˆσ + ˆσ. (23) Moreover, we note that the steay-state solutions of Eqs. (22) an (23) have the form Hence on substituting (24) into Eq. (25), we fin Now taking into account the completeness relation we get an applying once more Eq. (27), we have ˆσ = 4gε κγ c ˆη b ˆη a, (24) ˆη a = 2gε κγ c ˆσ + ˆσ. (25) ˆη a = 4ε2 κγ c ˆη b ˆη a. (26) ˆη a + ˆη b = Î, (27) ˆη a = ˆη b = 1 4ε 2 8ε 2 (28) 4ε 2 8ε 2. (29) It is not har to realize that ˆη a an ˆη a represent the probabilities for the two-level atom to be in the upper an lower levels, respectively. Finally, with the ai of (28) an (29), one can put Eq. (24) in the form 4gε ˆσ = 8ε 2. (30) Furthermore, applying the relation t â(t)â (t) = â(t) t â (t) + â(t) â (t) t (31)

4 along with Eq. (9) an its ajoint, we reaily get t â(t)â (t) = κ â(t)â (t) +ε â(t)+â (t) g [ ˆσ(t)â (t) + â(t)ˆσ (t) + ˆF(t)â (t) + â(t)ˆf (t) ]. (32) It proves to be convenient to replace the operators â (t) an â(t) that appear in the secon, thir, fourth, an fifth terms in Eq. (32) by expression (15) an its ajoint. We then fin t â(t)â (t) = κ â(t)â (t) +γ c ˆη b + 4ε2 κ 4gε κ ˆσ(t)+ ˆσ (t) 2g ( κ ˆσ(t)ˆF (t) + ˆF(t)ˆσ (t) ) + ˆF(t)â (t) + â(t)ˆf (t), (33) so that in view of the assumption that the atomic an noise operators are not correlate, there follows t â(t)â (t) = κ â(t)â (t) +γ c ˆη b + 4ε2 κ 4gε κ ˆσ(t)+ ˆσ (t) Moreover, the solution of Eq. (9) can be written as + ˆF(t)â (t) + â(t)ˆf (t). (34) t â(t) = â(0)e κt/2 +e κt/2 e κt /2 [ gˆσ(t )+ε+ ˆF(t )]t, (35) an multiplying this equation on the right by ˆF (t), we have 0 t â(t)ˆf (t) = â(0)ˆf (t) e κt/2 +e κt/2 e κt /2 [ g ˆσ(t )ˆF (t) + ˆF(t )ˆF (t) ]t, (36) in which we have use the fact that ε ˆF (t) = 0. Now in view of the assumptions 0 â(0)ˆf (t) = 0 (37) an Eq. (36) takes the form ˆσ(t )ˆF (t) = 0, (38) t â(t)ˆf (t) = e κt/2 e κt /2 ˆF(t )ˆF (t) t. (39) 0 Hence using the correlation property [5] ˆF(t )ˆF (t) = κδ(t t ), (40) we arrive at â(t)ˆf (t) = κ 2. (41) Therefore, on substituting (41) an its complex conjugate into Eq. (34), there follows t â(t)â (t) = κ â(t)â (t) +γ c ˆη b + 4ε2 κ 4gε κ ˆσ(t)+ ˆσ (t) +κ. (42)

5 The steay-state solution of Eq. (42) has the form an on taking into account (30), we obtain ââ = γ c κ ˆη b + 4ε2 κ 2 4gε κ 2 ˆσ + ˆσ +1 (43) ââ = γ c κ ˆη b + 4ε2 κ 2 γ c 8ε 2 κ 8ε (44) Following the same proceure, we can also reaily establish that at steay state Now employing Eqs. (44) an (45), we easily fin â â = γ c κ ˆη a + 4ε2 κ 2 γ c 8ε 2 κ 8ε 2. (45) [â,â ] = γ c κ ( ˆη b ˆη a )+1. (46) Finally, with the ai of Eqs. (28) an (29), one can put Eq. (46) in the form [â,â ] = γ 2 c 8ε 2 +1 (47) an on account of (28) the mean photon number of the cavity moe has the form n = 4ε2 κ 2 γ c 4ε 2 κ 8ε 2. (48) We realize that Eq. (47) represents the commutation relation for a cavity moe which is interacting with a two-level atom. An we notice that the first term on the right sie of this equation is ue to the interaction of the cavity moe with the two-level atom an the secon term is ue to the vacuum reservoir noise. We now consier the case in which the cavity moe is not interacting with an atom. Then for this case (g=0), Eq. (47) takes the form with [â,â ] = 1, (49) n = 4ε2 κ 2. (50) This represents the commutation relation for a free cavity moe with photons. Alternatively, in the absence of the riving coherent light (ε = 0), Eq. (47) has the form [â,â ] = γ c κ +1, (51) with n = 0. We see from Eq. (29) that the atom in this case is in the lower level. We observe that Eq. (51) represents the commutation relation for a vacuum cavity moe which is interacting with a two-level atom. We may envisage this interaction as a process in which virtual photons are absorbe an emitte, with the atom continuing to be in the lower level. We note that in the absence of this interaction (g=0), Eq. (51) goes over into [â,â ] = 1. (52) We now realize that the usual commutation relation is just the commutation relation for a free cavity moe with or without photons.

6 Furthermore, we wish to iscuss the commutation relation for the superposition of two light moes, represente by the operators â an ˆb. It proves to be very convenient to represent the superpose light moes by the operator ĉ efine by We thus see that ĉ = â+iˆb. (53) [ĉ,ĉ ] = [â,â ]+[ˆb,ˆb ]+i[ˆb,â ] i[â,ˆb ]. (54) We assume that the operators â an ˆb commute. It then follows that [ĉ,ĉ ] = [â,â ]+[ˆb,ˆb ]. (55) We therefore note that the commutator for the superpose light moes is just the sum of the commutators for the separate light moes. Finally, with the intention to etermine the conitions uner which the efinition given by (53) hols, we consier the mean photon number of the superpose light moes. Applying Eq. (53), this mean photon number can be written as ĉ ĉ = â â + ˆb ˆb +i( â ˆb ˆb â ). (56) We expect the mean photon number to be the sum of the mean photon numbers of the separate light moes. We then eman that the last term in Eq. (56) must be zero. To achieve this, we first assert that the superpose light moes must not be correlate. In view of this, the mean photon number can be rewritten as ĉ ĉ = â â + ˆb ˆb +i( â ˆb ˆb â ). (57) We now realize that the last term woul vanish if â = ˆb = 0 or â = ε a an ˆb = ε b, with ε a an ε b taken to be real or equal. [1] P. L. Knight an B. W. Shore, Phys. Rev. A 48, 642 (1993). [2] Fesseha Kassahun, Quantum Analysis of Light, (Kinle Direct Publishing, 2018). [3] M. J. Collett an D. F. Walls, Phys. Rev. A 32, 2887 (1985). [4] Fesseha Kassahun, Opt. Commun. 284, 1357 (2011). [5] C. W. Gariner an M. J. Collett, Phys. Rev. A 31, 3761 (1985).