Nicolas J. Cerf,Chris Adami
Abstract:We present a quantum information theory that allows for the consistent description of quantum entanglement. It parallels classical (Shannon) information theory but is based entirely on density matrices, rather than probability distributions, for the description of quantum ensembles. We find that, unlike in Shannon theory, conditional entropies can be negative when considering quantum entangled systems such as an Einstein-Podolsky-Rosen pair, which leads to a violation of well-known bounds of classical information theory. Negative quantum entropy can be traced back to ``conditional'' density matrices which admit eigenvalues larger than unity. A straightforward definition of mutual quantum entropy, or ``mutual entanglement'', can also be constructed using a ``mutual'' density matrix. Such a unified information-theoretic description of classical correlation and quantum entanglement clarifies the link between them: the latter can be viewed as ``super-correlation'' which can induce classical correlation when considering a ternary or larger system.
What problem does this paper attempt to address?
The problems that this paper attempts to solve mainly focus on the following aspects:
1. **Characteristics of Bose - Einstein Condensation (BEC) in a harmonic oscillator potential well**:
- The paper explores the Bose - Einstein condensation phenomena in different dimensions in an external harmonic oscillator potential well. In particular, in the thermodynamic limit, Bose - Einstein condensation occurs spatially in systems of two or more dimensions, while in one - dimensional systems, there is no true phase transition, but there will be a pseudo - phase transition at a "pseudo - critical" temperature.
2. **BEC in finite systems**:
- For the finite - particle - number systems used in actual experiments, the paper studies the conditions for the occurrence of BEC. Even in the absence of a true phase transition, BEC can still occur in a finite system below a certain "pseudo - critical" temperature.
3. **Condensation fractions in momentum space and position space**:
- The paper analyzes in detail the condensation fractions in momentum space and position space and finds that in the thermodynamic limit, the condensation fraction in momentum space approaches zero as the number of particles \( N \) increases, while in position space, a certain condensation proportion can be maintained.
4. **Reconciliation of the Hohenberg theorem and the CPO theorem**:
- The paper discusses that in a two - dimensional harmonic oscillator potential well, although the Hohenberg theorem states that there is no BEC with zero momentum in a two - dimensional system, BEC in position space can still exist. This needs to be explained by the conditions of the CPO theorem, that is, the density diverges at the origin, so that the CPO theorem is not applicable.
5. **The influence of interactions on BEC**:
- The paper also briefly mentions the influence of interactions on BEC, pointing out that in the case of considering hard - core interactions, BEC in a two - dimensional system will disappear in the thermodynamic limit, but in actual experiments, due to the small number of particles, the ideal gas model is still a reasonable approximation.
### Specific problem analysis
#### 1. Does a true phase transition occur?
- The BEC observed in experiments is not a true phase transition because a true phase transition requires that the number of particles and volume tend to infinity while keeping the density constant. In actual experiments, the number of particles is finite and the potential well is fixed, so there is no true non - analytic phase transition.
#### 2. Is a pseudo - phase transition possible?
- Even if there is no true phase transition in the thermodynamic limit, BEC can still occur in a finite system below a certain "pseudo - critical" temperature. This pseudo - phase transition is manifested as particles rapidly occupying the lowest energy level, similar to a true phase transition.
#### 3. How is BEC in two - dimensional systems explained?
- Although the Hohenberg theorem states that there is no BEC with zero momentum in a two - dimensional system, BEC in position space can still exist. This is because the density diverges at the origin, making the CPO theorem inapplicable, thus allowing BEC in position space.
### Conclusion
Through detailed theoretical analysis and numerical calculations, the paper explores the BEC phenomena in different dimensions in a harmonic oscillator potential well, especially the characteristics of BEC in finite systems and the relationship between the condensation fractions in momentum space and position space. These studies are of great significance for understanding the BEC phenomena in actual experiments.