高斯波包的色散：动量为何变得更确定？

Great question—this is one of those subtle intersections of quantum math and intuition where it’s easy to misread the reciprocal space picture. Let’s unpack this step by step, starting with a critical correction to your initial thought.

First: Momentum Uncertainty Doesn’t Actually Improve #

When the position wavefunction spreads, it’s tempting to lean on the Fourier reciprocal relationship and think "wider position distribution = narrower momentum distribution"—but that only holds for static wavefunctions. For a free particle, momentum is a conserved quantity!

In momentum space, the wavefunction’s time evolution is just a phase shift:
φ(p,t) = φ(p,0) e^(-i p² t/(2mℏ))
The key here is that the magnitude of φ(p,t) never changes—only the phase. That means the probability density for momentum, |φ(p,t)|², is completely time-independent. So your sense of "momentum deterministicity提升" (increasing certainty) is a misinterpretation: the position spread isn’t making the momentum distribution narrower—it’s just that the interference between different momentum components (which originally localized the particle in position space) breaks down over time.

The Physical Process Driving Dispersion #

The core physical principles here are momentum conservation and dispersion from the quadratic energy-momentum relation:

• Since no forces act on the free particle, every momentum component in the wavefunction retains its momentum forever—there’s no process "changing" the particle’s momentum.
• But energy depends quadratically on momentum (E = p²/(2m)), so higher-momentum components evolve with faster phase rates. Over time, these differing phase speeds cause the constructive interference that originally concentrated the particle in position space to fall apart. Different momentum components get out of sync, so the position probability spreads out as their contributions no longer align neatly.

This isn’t an active "process" modifying the particle—it’s just the natural time evolution of a quantum state that’s a superposition of momentum eigenstates, each evolving independently with their own phase.

Understanding Dispersion Without Position Wavefunction Spread #

Absolutely—we can frame this entirely outside the position space picture:

• Momentum space phase perspective: Each momentum component has an angular frequency ω(p) = p²/(2mℏ). For a wavepacket with a range of momenta Δp, the phase difference between the highest and lowest momentum components grows linearly with time. As this difference becomes large compared to 2π, the interference pattern in position space (which comes from summing these phase-shifted components) spreads out because peaks and troughs no longer overlap.
• Uncertainty principle re-clarified: The relation ΔxΔp ≥ ℏ/2 doesn’t mandate that Δp decreases when Δx increases—only that their product can’t drop below a minimum. For free particles, Δp is fixed (momentum conservation), so Δx must increase over time to satisfy the principle as the wavepacket evolves. This is a direct result of phase evolution, not improved momentum certainty.
• Loose classical analogy: Think of a classical wavepacket (like a water wave pulse) made of different frequency components. If the medium is dispersive (different frequencies travel at different speeds), the pulse spreads out over time. The quantum free particle works the same way—different momentum components correspond to different speeds (v = p/m), so they drift apart, spreading the packet. No change in the distribution of speeds (momenta), just their relative positions shifting.

内容的提问来源于stack exchange，提问作者Meep