Quantum Physics-Failures of Quantum Mechanics

8.04 Quantum Physics Lecture IICM fails to describe “macroscopic quantum phenomena.” Phenomena where microscoopi properties carry over into macroscopic world: • superfluidity Helium flows without friction at sufficiently low temperature. • superconductivity Current flows without dissipation. • ferromagnetisim Self-alignment of magnetic moments/spins... Bose-Einstein condensation • Superfluidity of atoms. . . All these problems (and many more) are solved by marvellously successful quantum mechanics, QM. Problems with/failures of QM • No model for some macroscopic quantum behavior (high-critical temperature superconducttivity) this is likely to be solved within QM. • QM is incompatible with general relativity. This is likely to be solved outside QM, within some bigger theory. (string theory?) • Interpretation ambiguities of mathematical structure of QM: -Role of measurement -Determinism vs. probability. . . “The old one does not roll dice.” -Transition from microscopic quantum mechanical to macroscopic quantum behavioor...Schr¨odinger’s cat, “dead and alive”) -Entanglement and hidden-variables. . . (Einstein-Podolsky-Rosen (EPR) paradoox Bell’s inequalities) -Bohr’s (& Born’s) Copenhagen Interpretation vs. “many-worlds-theory” Quantum Mechanics Quantum Mechanics is the “weirdest physics theory you have ever encountered.” The mathematical structure of QM is well-defined, but the physical interpretation of mathematicca results, such as observables and measurement can lead to apparent paradoxa. Math-Massachusetts Institute of Technology II-1 8.04 Quantum Physics Lecture IIematical concept used for describing object, wavefunction, is not the same as object itseelf As seen in Figure I, spreading of particle wavepacket in free space forces on us a probabilistic interpretation of QM. Copenhagen interpretation Wavefunction is probability amplitude (magnitude of wavefunctiio squared is probability or probability density) to find particle in certain state or at certain position. Probabilities for outcomes of measurements are deterministic. They are determined by wave-functions that evolve deterministically. But the outcome of a particcula measurement cannot, in general, be predicted. We can prediic expectation values, variances, standard deviations, but not a particular realization of an experiment. The Schr¨y(t=0)y(t>0)xodinger equation (SE) is a partial differential equation that governs the determinis-Figure I: Wavepacket tic evolution of a wavefunction ψ. ψ contains all the information available to describe the system. The probability for a particular measurement outcome (say, to find the particle between x and x +Δx) is determined by |ψ(x)|2 → (Copenhagen Interpretation). Other interpretations “Many Worlds” interpretation Each time we make a measurement, our world splits into many, one for each possible outcome of the measurement. Absurd? No known way to falsify. “Hidden-Variables” theories Favored by Einstein, Einstein-Podolsky-Rosen (EPR) paradox. Randomness arises because we do not know the complete state of the system. There is some information that is not contained in the wavefunction, but that in principle exists. This would correspond to a randommnes similar to that encountered in classical statistical mechanics. For many decades it was believed that hidden-variables theories are “metaphysics”, i.e., that they cannot be falsiffie or distinguished experimentally from the standard Copenhagen interpretation. Howevver John Bell came up with a possibility to experimentally distinguish hidden-variables theories, local-realism theories from standard QM. Bell’s inequalities Bounds on conditional probabilities that are obeyed by hidden-variables theories, but violaate by standard QM. Experiments performed many times since disagreed with hidden-Massachusetts Institute of Technology II-2 8.04 Quantum Physics Lecture IIvariables theories, but confirm standard QM. When do we need QM?Which historic experiments cannot be adequately described by CM?We will take a logical, not a chronological approach. CM, E&M fail in general at the microscopic level, i.e., when we deal with single fundamental particles (single quanta, such as electrons, atoms, photons, . . . ). When does the classical description of a single particle/object break down? CM fails at small distance d, i.e., when the resolution of your measurement is improved to measure the position of the particle with increasing accuracy Δd. This distance scale d depends on the momentum p of the particle, or more accurately, on the uncertainty of your measurement (or knowledge) of the momentum Δp of the particle. Critical distance scale Δd where CM becomes inadequate is set by quantity Δd Δp (phase space area). · DpDdDdDp

Position dMomentum pFigure II: Phase space area• Units: [Δd · Δp] = m · kg · ms = J · s If Δd Δp � h , then behavior is classical.• · Massachusetts Institute of Technology II-3 8.04 Quantum Physics Lecture IIh = 6.6 × 10−34J s Planck’s constant • ·→ Same quantity that, given a frequency ν, introduces a natural energy scale E = hν.• If Δd Δp ∼ h , then QM is necessary. • · Other formulation [dp] = [r × p] = [l] (2-1)· hIf l= � CM (2-2)||� 2π|l|∼ � QM (2-3)h � == 1.05 × 10−34J s (reduced Planck’s constant) (2-4)2π · → Given an angular frequency ω = 2πν, � sets a natural energy scale E = �ω. Examples 1. Gas in a box at room temperature:d = 1 cm• m = 87mp•mp = 1.67 × 10−27 kg• �p2� = 1kBT (in 1D) (2-5)2m 2→ �� mΔp ∼�p2� = mkBT = 2.5 × 10−23 kg · s (2-6)d ∼ Δd = 1 cm (2-7) 2 Δp Δd = 2.5 × 10−25kg m= 2.5 × 10−25J s = 4 × 108h (2-8)·· s · = CM adequate⇒ 2. Laser cooled gas in a magnetic trap Δp ∼ � mkBT = 1.4 × 10−26 kg · m s (2-9) Δp · Δd = 1.4 × 10−30 J · s ∼ 2000h (2-10) = ⇒ still classical Massachusetts Institute of Technology II-4 � 8.04 Quantum Physics Lecture IId~100 mm+q-qr(a) Magnetic potential with T ∼ (b) Electron in atom 100 µK Figure III: Magnetic potential and atom model 3. Evaporative cooling to T ∼ 1 µK, d ∼ 1 µm Δp ∼ 1.4 × 10−27 kg m (2-11)· s Δp Δd ∼ 2h (2-12)· Bose-Einstein condensation if atoms are boson • • Degenerate Fermi gas if atoms are fermions = QM description is necessary ⇒ 4. Electron in an atom m = 9.1 × 10−31 kg q = 1.6 × 10−19 C r ∼ 1Å = 10−10 m As ε0 = 8.85 × 10−12 · Vm· l = pr = 2m4qπε20 r ∼ 2 × 10−34 J · s ∼ � (2-13) = QM description⇒ Massachusetts Institute of Technology II-5 8.04 Quantum Physics Lecture IIWhy does the classical description break down at this scale? • Heisenberg uncertainty • wavelength of particle To shed some light on this question, let us examine Fermat’s principle of least time. Fermat’s principle of least time ABBeam block here hasno effect. How close to correct path can we place beam block?Figure IV: Straight line path ABSlab of glass n>1Beam block here hasno effect.Figure V: Refraction • How does light travel from A to B? • Straight line. • Why? • How straight? How does light travel? Fig. IV = straight line ⇒ n2Fig. V = at interface it obeys Snell’s law sin θ1 = ⇒ sin θ2 n1 Massachusetts Institute of Technology II-6 8.04 Quantum Physics Lecture IIUnified description Light travels along the path that takes the shortest time. In glass, light travels at c� = nc 2, so correct path takes shorter time than straight line. But: How does light “know” what path takes the shortest time? In some instances, light takes the path that takes the longest time. Improved version of Fermat’s principle Light takes the path where there is no first-order change in the traveling time for nearby paths. Reading Feynman, Lectures on Physics I, Ch. 26. Massachusetts Institute of Technology II-7