Commun. Theor. Phys. (Beijing, China) 44 (2005) pp. 455–458 c International Academic Publishers Vol. 44, No. 3, September 15, 2005 Continuous Quantum Nondemolition Measurements of a Particle in Electromagnetic and Gravitational Fields∗ ZHU Chun-Hua1,† and ZHA Chao-Zheng1,2 1 Department of Physics, Xinjiang University, Wulumuqi 830046, China 2 CCAST (World Laboratory), P.O. Box 8730, Beijing 100080, China (Received January 27, 2005) Abstract The detection of a particle in electromagnetic plus gravitational fields is investigated. We obtain a set of quantum nondemolition variables. The continuous measurements of these nondemolition parameters are analyzed in the framework of restricted path integral formalism. We manipulate the corresponding propagators, and deduce the probabilities associated with the possible measurement outputs. PACS numbers: 03.65.-w Key words: continuous quantum nondemolition measurements, restricted path integral 1 Introduction The problem of quantum measurement has provoked a deep interest since the outset of quantum theory. From the ordinary quantum mechanics we know that the measurement on a quantum system obeys the von Neumann’s reduction postulate. Nevertheless, this approach has the unpleasant feature of introducing an extra assumption in the theory and deals only with instantaneous and perfect measurement. During the past two decades many authors attempted to solve these problems via various approaches and a number of models have been proposed. In this paper we investigate the quantum nondemolition (QND) measurement of a particle moving in an electromagnetic plus gravitational field by means of the restricted path integral (RPI) formalism, which was proposed by Feynman[1] to describe continuous quantum measurement (CQM), and was developed latter by Mensky.[2] The measurement system is considered in the RPI theory of measurements as an open system, and the back influence of the environment onto the measured system is taken into account by restricting the path integral. The restriction is determined by the information which the measurement imposes on the measured system and is described by a weight functional. The RPI approach is one of phenomenological descriptions of CQM, and does not require an explicit model of the measuring device. The RPI method has important applications in many fields. For instance, it has been applied to the analysis of QND measurements.[3−5] QND measurements were introduced by Braginsky, Vorontsov, and Khalili[6] and discussed later from different points of view.[7] Caves etc. defined the QND measurements of  as a sequence of precise measurements of  such that the result of each measurement (after the first) is completely predictable from the ∗ The result of the preceding measurement. In such a measurement one can monitor an observable with arbitrarily small error, in principle, without quantum limit. Camacho discussed QND measurement of a particle in electric and gravitational fields.[8,9] In his model he ignored the magnetic fields. In this work we shall find a set of QND variables for the case of a particle moving in an electromagnetic and gravitational fields of a spherical body, for instance, in the Earth’s electromagnetic and gravitational fields. In this sense, we would say our model is a more realistic scenario. We will consider the continuous monitoring of these nondemolition parameters, and calculate the corresponding propagators along the ideas of the RPI formalism. Using these results the probabilities associated with the possible measurement outputs are evaluated. 2 Our Model Suppose we have a spherical body with mass M , radius R, and electric charge Q. We assume that the sphere rotates around an axis of a sphere with an angular velocity ω0 . For the sake of simplicity, we suppose that the sphere is provided with a constant electric charge density. Now let us consider a particle whose mass and charge are m and q, respectively. The Hamiltonian of the particle moving in our electromagnetic and gravitational fields is given by 1 ~ )2 + k qQ − G mM , ( p~ − q A (1) 2m r r where r denotes the distance from the center of the sphere, ~ is the vecp~ is canonical momentum of the particle, and A tor potential of the electromagnetic field. Considering the H= project supported by National Natural Science Foundation of China under Grant No. 10265003 xdzch2002@yahoo.com.cn † E-mail: 456 ZHU Chun-Hua and ZHA Chao-Zheng ~ is then written as symmetry of the system, A 2 ~ = µ0 ω0 QR sin θ ~ey , A 20πr2 where θ is the angle between rotational axis and vector ~r. The corresponding Lagrangian is written as p~ 2 qQ n2 q 2 Q2 R4 mM −k − +G , (2) 2m 2mr4 r r where n = µ0 ω0 sin(θ)/20π Let r = R + z, where z is the height over the sphere. When the height is much smaller than the radius of the sphere z R, the Hamiltonian and Lagrangian become L= p~ 2 nqQ 2nqQ 1 − py + py z − F z + mω 2 z 2 , 2m m mR 2 2 p~ 1 L= + F z − mω 2 z 2 , 2m 2 respectively, where H= (3) (4) qQ mM 2n2 q 2 Q2 +k 2 −G 2 , (5) mR R R 5n2 q 2 Q2 qQ M + k − G . (6) ω2 = 2 m2 R2 mR3 R3 From Eq. (4), it can be easily seen that the particle in our model is a harmonic oscillator subjected to force F . Compare our model with that of a particle moving in electric and gravitational fields[8] in which the condition for ω 2 > 0 is qQ > GM m/k, and the charge are only of the same sign, or, if the charges are of opposite sign (qQ < 0), the harmonic oscillator is then provided with a complex frequency (ω 2 < 0). We find that in our model the effect of magnetic field on the particle are represented by 2n2 q 2 Q2 /mR and 5n2 q 2 Q2 /m2 R2 in Eqs. (5) and (6), so the corresponding conditions is changed. Thus the condition for ω 2 > 0 is qQ > GM m/k − 5n2 q 2 Q2 R/mk and the frequency r 5n2 q 2 Q2 qQ M |ω| = 2 + k − G m2 R2 mR3 R3 is larger than that of the previous model. For ω 2 < 0, the charge is of opposite sign. Moreover, not only the charges should be opposite in sign for ω 2 < 0 in our model, the condition 5n2 q 2 Q2 R/mk < GM m/k + |qQ| should also be satisfied. It is worthy of mentioning that the previous model is a one-dimensional system (having z component only). While the present model is a two-dimensional system (having y component as well as z component) due to ~ the vector potential A. F = 3 Continuous Quantum Nondemolition Variables Now we introduce the operator A(t) = ρq + σp , (7) where ρ, σ are real functions of time. Suppose now that the continuous monitoring of the observable A is carried Vol. 44 out during the time interval [0, T ], and the result of this measurement being expressed in terms of function a[t]. An effective Hamiltonian H̃ can be introduced, ih̄ (A − a)2 , H̃ = H − T ∆a2 where ∆a is the measurement error. In the case of multidimensional system, the condition for A to be a QND variable is that the equation of motion for à = ρq̃ + σ p̃ (corresponding to the effective Hamiltonian H̃) coincides with the equation for A (corresponding to the Hamiltonian H), all terms containing h̄/T ∆a2 being cancelled.[3] We can apply this criterion in our case. Our model is a two-dimensional systems because of the magnetic field. So if ω 2 > 0, we obtain a set of QND observable, m A1 = σ(t) − y + py , (8) t A2 = σ(t)(mω tan(ωt)z + pz ) . (9) And if ω 2 < 0, we can also have m A1 = σ(t) − y + py , t A2 = σ(t)(−m|ω| tanh(|ω|t)z + pz ) . (10) (11) Any choice for function σ(t) gives a set of QND variables. In this work, we shall only consider the case ω 2 > 0, and for ω 2 < 0 we can also deal with in the same way. For the sake of simplicity we assume σ(t) = 1 . So we obtain a set of QND variables, m (12) A1 = − y + py , t A2 = mω tan(ωt)z + pz . (13) Such a set of QND variables have the following properties. The variables can be measured simultaneously with arbitrary precision, and in a sequence of precise, simultaneously measuring of all the variables, the results of each set of measurements can be predicted from the results of the preceding set.[7] 4 Propagator and Probability We have already learned that a continuously measured quantum system can be described by RPI method.[2,10] The whole quantum theory of measurements can be derived from quantum mechanics by means of path integral in the framework of the RPI approach. From the behavior of a set of QND variables we know that the variables A1 and A2 can be measured simultaneously with arbitrary precision. First, let us consider the situation in which A2 (t) is continuously monitored. The measuring progress is taken into account through a weight functional in the path integral. We choose a weight functional in gaussian form, Z T n o 1 2 [A (t) − a (t)] dt . (14) ω[a2 (t)] [ p, q] = exp − 2 2 T ∆a22 0 No. 3 Continuous Quantum Nondemolition Measurements of a Particle in Electromagnetic and Gravitational Fields Here a2 (t) is the measurement output, and ∆a2 is the measurement error. According to RPI approach, if the Z U[a2 (t)] (W, T ; N, 0) = n d[ p]d[q] exp − 1 T ∆a22 Z 0 T 457 particle goes from point N to point W the corresponding propagator is given by n i Z T p~ 2 o 1 [A2 (t) − a2 (t)] dt exp + F z − mω 2 z 2 dt . h̄ 0 2m 2 2 o Considering expression (13), we obtain the propagator Z T o n im o n m 1 a22 (t)dt exp [(xW − xN )2 + (yW − yN )2 ] exp − U[a2 (t)] (W, T ; N, 0) = 2 2πih̄T 2h̄T T ∆a2 0 Z Z n i T h 1 i o 2ih̄ ih̄ 2 × d[z]d[ pz ] exp pz − (a2 − mω tan(ωt))z)pz dt + 2 2 h̄ 0 2m T ∆a2 T ∆a2 n i Z T h ih̄ 1 (mω tan(ωt))2 − mω 2 z 2 × exp 2 h̄ 0 T ∆a2 2 2a ih̄ i o 2 − (mω tan(ωt)) − F z dt . T ∆a22 (15) (16) This last path integral is gaussian in pz and z, then it can be calculated, Z T n im o n o T ∆a22 − 2imh̄ 2 U[a2 (t)] (W, T ; N, 0) = N2 exp [(xW − xN )2 + (yW − yN )2 ] exp − a (t)dt 2h̄T (T ∆a22 )2 + (2mh̄)2 0 2 ia2 F F2 a22 (T ∆a22 − 2imh̄)2 β 2 2 − (T ∆a − 2imh̄) − ((T ∆a22 )2 + (2mh̄)2 ) ) 2 2 (T ∆a2 )2 + (2mh̄)2 h̄ 4h̄2 dt . (17) imω 2 0 (T ∆a22 − 2imh̄)β 2 + ((T ∆a22 )2 + (2mh̄)2 ) 2h̄ For the sake of implicity, here we have introduced N2 , which is a normalized constant, and set β = mω tan(ωt). The probability density, associated with the different measurement outputs, is given as (according to the expression P[a2 (t)] = |U[a2 (t)] |2 ) Z T n o 2T ∆a22 2 P[a2 (t)] = exp − a (t)dt (T ∆a22 )2 + (2mh̄)2 0 2 (Z × exp T (Z × exp 0 T 16a22 m2 h̄2 ω 2 T ∆a22 β 2 8h̄2 a22 T ∆a22 β 4 − − 4a2 mω 2 F T ∆a22 − 2F 2 T ∆a22 β 2 ) (T ∆a22 )2 + (2mh̄)2 (T ∆a22 )2 + (2mh̄)2 dt . 4h̄2 β 4 − 8m2 h̄2 ω 2 β 2 + m2 ω 2 ((T ∆a22 )2 + (2mh̄)2 ) (18) Second, if A1 (t) is measured simultaneously, we obtain the corresponding propagator with the same method, n im o n h imω 2 U[a1 (t)] (W, T ; N, 0) = N1 exp (xW − xN )2 exp cos(ωT )(zW + zN2 ) − 2zW zN 2h̄T 2h̄ sin(ωT ) 2 F 2 T sin(ωT ) io 2F 2 (1 − cos(ωT ))(z + z ) − F (1 − cos(ωT )) + , (19) + W N mω 2 m2 ω 4 m2 ω 3 where N1 is the normalized constant. And probability is given as P[a1 (t)] = |U[a1 (t)] |2 = 1 . (20) 5 Conclusions In this work we investigate the motion of a charged particle moving in the electromagnetic and gravitational fields of a spherical body, and find a set of QND variables. Then it is assumed that the set of QND parameters are continuously monitored simultaneously. Along the ideas of RPI approach, the propagators and probabilities associated with the possible measurement outputs are manip- ulated. Looking at expression (18), the mass m of the particle appears in the relation mh̄, which is consistent with the result of QND measurement in electric and gravitational fields.[8] It differs from the result of Colella, Werner, and Overhauser experiment,[11] and differs from the case of a quantum demolition measurement,[12] which has the combination m/h̄. Obviously, we can see from expression (20) that there is no standard quantum limit. Now let us consider two different limits for expression (18). First, if ∆a2 → 0, we obtain P[a2 (t)] → 1, namely, the probability does not de- 458 ZHU Chun-Hua and ZHA Chao-Zheng Vol. 44 pend upon the measurement output a2 (t), and all the possible measurement outputs have the same probability. We may notice that there is also no standard quantum limit, in other words, we can measure variable A2 with an arbitrarily small error, and all the necessary information can be extracted. Clearly, this quantum feature is associated with QND measurement. Second, the opposite case, that is, the limit of rough measurements, ∆a2 → ∞, renders also the result P[a2 (t)] → 1. In this case weight functional (which restricts the integration domain) ω[a2 (t)] → 1, that is, all the paths have the same probability in path integral. In this sense this case is as if there were no measuring process, and all the possible measurement outputs have the same probability. From our results (18) and (20), we draw a conclusion that a set of QND variables A1 and A2 could be monitored simultaneously with arbitrary precision, which is consistent with the property of a set of QND variables. To finish, it is worthy to be mentioned that our case was a two-dimensional system because we considered not only the electric and gravitational fields but also magnetic field. So we obtained a set of QND variables, expressed by Eqs. (12) and (13). This obviously differs from the onedimensional case when only considering the electric and gravitational fields,[8] where what we obtain is only one QND variable. References [7] C.M. Caves, K.S. Thorne, R.W.P. Drever, V.D. Sandberg, and M. Zimmermann, Rev. Mod. Phys. 52 (1980) 341, [8] A. Camacho and A. Camacho-Galvan, Int. J. Mod. Phys. D 10 (2001) 859, gr-qc/0104082. [9] A. Camacho, Gen. Rel. Grav. 33 (2001) 901, quantph/9911106. [10] M.B. Mensky, Int. J. Theor. Phys. 37 (1998) 273. [11] R. Colella, A.W. Overhauser, and S.A. Werner, Phys. Rev. Lett. 34 (1975) 1472. [1] R.P. Feynman, Rev. Mod. Phys. 20 (1948) 367. [2] M.B. Mensky, Physics-Uspekhi. 41 (1998) 923, quantph/9812017. [3] G.A. Golubssova and M.B. Mensky, Applied Physics A 4 (1992) 2733. [4] M.B. Mensky and J. Audretsch, Int. J. Mod. Phys. A 277 (1997) 129. [5] A. Camacho, Phys. Lett. A 277 (2002) 7. [6] V.B. Braginsky, Y.I. Vorontsov, and F.Y. Khalili, Zh. Eksp. Teor. Fiz. 73 (1997) 1340. [12] A. Camacho and A. Camacho-Galvan, Phys. Lett. A 247 (1992) 373.