Lecture Notes: Introduction to Quantum Mechanics for Quantum Computing
Introduction
This document provides a structured overview of the fundamental concepts of quantum mechanics (QM), essential for understanding quantum computing. We begin by exploring the historical context that necessitated the development of quantum theory, focusing on key experimental observations that classical physics failed to explain. We introduce the Bohr model as a significant early, albeit incomplete, attempt to incorporate quantization into atomic physics, particularly for explaining the hydrogen spectrum. This leads into the formal development of modern QM through Heisenberg’s matrix mechanics and Schrödinger’s wave mechanics in the mid-1920s. The core of the lecture focuses on the mathematical formalism underpinning QM, including Hilbert spaces and function spaces, followed by a detailed presentation of the fundamental postulates. We discuss quantum dynamics as governed by the Schrödinger equation and the crucial role of unitary operators, which are mathematically equivalent to the quantum gates used in quantum computation. Key consequences of the theory, such as the Heisenberg Uncertainty Principle, and illustrative examples like the quantum harmonic oscillator and the quantization of angular momentum (including intrinsic spin), are presented. Finally, we connect these abstract concepts to the concrete realization of qubits, particularly using spin-1/2 systems, and address the conceptually profound and practically vital topic of quantum measurement. The objective is to establish a foundational understanding of the quantum principles governing the systems and operations utilized in quantum computers.
Motivation: Failures of Classical Physics
Around the turn of the 20th century, classical physics—encompassing Newtonian mechanics, Maxwell’s electromagnetism, and classical statistical mechanics—faced several crises when confronted with experimental results at the atomic scale. These failures highlighted the need for a new theoretical framework.
Black Body Radiation
A black body is an idealized object that absorbs all incident electromagnetic radiation and emits radiation based on its temperature. Classical physics (specifically, the Rayleigh-Jeans law derived from classical statistical mechanics and electromagnetism) predicted that the intensity of the emitted radiation should increase indefinitely as the frequency increases (or wavelength decreases), leading to an "ultraviolet catastrophe"—infinite energy emission at high frequencies. Experimental measurements, however, showed a characteristic spectrum that peaked at a specific frequency determined by the temperature and fell off at higher frequencies.
Planck’s Quantization Hypothesis Max Planck (1900) resolved this discrepancy by postulating that the energy of the oscillators within the black body cavity could not take continuous values but was quantized. He proposed that energy is emitted and absorbed in discrete packets, or quanta, with energy proportional to the frequency: \[E = h\nu \label{eq:planck_energy}\] where \(h \approx 6.626 \times 10^{-34}\) J\(\cdot\)s is Planck’s constant and \(\nu\) is the frequency of the radiation. This assumption led to a formula that perfectly matched the experimental spectrum.
Photoelectric Effect
When light illuminates a metal surface, electrons can be ejected. Classical wave theory predicted that:
Electron emission should occur for any frequency of light, provided the intensity is sufficiently high (allowing energy to accumulate).
The kinetic energy of the emitted electrons should increase with the light intensity.
Experiments contradicted these predictions, showing instead that:
Electron emission occurs only if the light’s frequency \(\nu\) is above a specific threshold frequency \(\nu_0\), characteristic of the metal, regardless of the light’s intensity.
Below the threshold frequency (\(\nu < \nu_0\)), no electrons are emitted, no matter how intense the light.
Above the threshold frequency (\(\nu > \nu_0\)), the maximum kinetic energy (\(K_{max}\)) of the emitted electrons increases linearly with the frequency and is independent of the intensity. Increased intensity only increases the *number* of emitted electrons.
Einstein’s Explanation Albert Einstein (1905) explained the photoelectric effect by extending Planck’s quantization idea to light itself. He proposed that light consists of discrete particles, later called photons, each carrying energy \(E = h\nu\). An electron is ejected only if it absorbs a single photon with sufficient energy to overcome the metal’s work function \(\Phi\) (the minimum energy required to remove an electron). The energy balance is: \[K_{max} = h\nu - \Phi \label{eq:photoelectric}\] This equation explains the threshold frequency (\(\nu_0 = \Phi/h\)) and the linear dependence of \(K_{max}\) on \(\nu\). Einstein was awarded the 1921 Nobel Prize in Physics primarily for this explanation, not for his theory of relativity.
Atomic Spectra and Stability
Classical electromagnetism predicted that an electron orbiting a nucleus, being an accelerating charge, should continuously radiate energy, causing it to lose energy and rapidly spiral into the nucleus. This contradicted the observed stability of atoms. Furthermore, when atoms (like hydrogen gas in a discharge tube) are excited, they emit light not as a continuous spectrum (like a black body) but only at specific, discrete frequencies or wavelengths. This results in a characteristic line spectrum for each element. Classical physics could not explain why atoms were stable or why their emission spectra were discrete.
The Bohr Model of the Hydrogen Atom (c. 1913)
Niels Bohr proposed a semi-classical model for the hydrogen atom that successfully explained its line spectrum by incorporating quantization rules into classical mechanics.
Postulates and Quantization of Angular Momentum
Bohr’s model was based on postulates that departed from classical physics:
Electrons orbit the nucleus in specific circular orbits ("stationary states") without radiating energy.
The angular momentum \(L\) of the electron in these allowed orbits is quantized, restricted to integer multiples of the reduced Planck constant \(\hbar = h/(2\pi)\): \[L = m_e v r = n \hbar, \quad n = 1, 2, 3, \dots \label{eq:bohr_L}\] where \(m_e\) is the electron mass, \(v\) is its orbital speed, \(r\) is the orbit radius, and \(n\) is the principal quantum number.
An electron can transition between stationary states by absorbing or emitting a photon whose energy \(h\nu\) exactly matches the energy difference between the initial and final states.
Derivation of Energy Levels and Spectra
Bohr combined his quantization condition [eq:bohr_L] with classical mechanics for an electron orbiting a proton:
Force Balance: The electrostatic Coulomb attraction between the proton (charge \(+e\)) and the electron (charge \(-e\)) provides the centripetal force required for circular motion: \[\frac{1}{4\pi\epsilon_0} \frac{e^2}{r^2} = \frac{m_e v^2}{r} \label{eq:bohr_force}\] where \(\epsilon_0\) is the permittivity of free space.
Total Energy: The total energy \(E\) is the sum of the electron’s kinetic energy \(K = \frac{1}{2}m_e v^2\) and its potential energy \(U = -\frac{1}{4\pi\epsilon_0} \frac{e^2}{r}\): \[E = K + U = \frac{1}{2}m_e v^2 - \frac{1}{4\pi\epsilon_0} \frac{e^2}{r} \label{eq:bohr_classical_E}\] From the force balance equation [eq:bohr_force], we find \(\frac{1}{2}m_e v^2 = \frac{1}{2} \frac{1}{4\pi\epsilon_0} \frac{e^2}{r}\), so \(E = -\frac{1}{2} \frac{1}{4\pi\epsilon_0} \frac{e^2}{r}\).
By solving equations [eq:bohr_L] and [eq:bohr_force] for \(v\) and \(r\), and substituting into the energy expression, Bohr found that both the allowed radii and the corresponding energy levels are quantized, depending on the integer \(n\): \[\begin{aligned} r_n &= n^2 \frac{4\pi\epsilon_0 \hbar^2}{m_e e^2} = n^2 a_0 \label{eq:bohr_rn} \\ E_n &= - \frac{m_e e^4}{2 (4\pi\epsilon_0)^2 \hbar^2} \frac{1}{n^2} = - \frac{E_R}{n^2}, \quad n = 1, 2, 3, \dots \label{eq:bohr_En}\end{aligned}\] where \(a_0 \approx 0.0529\) nm is the Bohr radius (radius of the \(n=1\) orbit) and \(E_R \approx 13.6\) eV is the Rydberg energy.
Explanation of Hydrogen Spectrum
According to Bohr’s third postulate, when an electron transitions from a higher energy level \(n_i\) to a lower energy level \(n_f\) (\(n_i > n_f\)), it emits a photon with energy equal to the difference \(\Delta E = E_{n_i} - E_{n_f}\). The frequency \(\nu\) and wavelength \(\lambda\) of the emitted photon are given by: \[h\nu = \frac{hc}{\lambda} = \Delta E = E_{n_i} - E_{n_f} = \left(-\frac{E_R}{n_i^2}\right) - \left(-\frac{E_R}{n_f^2}\right) = E_R \left( \frac{1}{n_f^2} - \frac{1}{n_i^2} \right) \label{eq:bohr_transition}\] This formula, often written in terms of the inverse wavelength (wavenumber \(\bar{\nu} = 1/\lambda\)), precisely matched the empirical Rydberg formula for the hydrogen spectrum: \[\frac{1}{\lambda} = R_H \left( \frac{1}{n_f^2} - \frac{1}{n_i^2} \right) \label{eq:rydberg_formula}\] where \(R_H = E_R / (hc)\) is the Rydberg constant. Bohr’s model provided a theoretical derivation for \(R_H\) that agreed remarkably well with the experimental value.
Hydrogen Spectrum Lines The formula [eq:bohr_transition] or [eq:rydberg_formula] explains the observed spectral series:
Lyman series (ultraviolet): Transitions ending at \(n_f=1\) (from \(n_i=2, 3, \dots\)).
Balmer series (visible/near UV): Transitions ending at \(n_f=2\) (from \(n_i=3, 4, \dots\)). For instance, the transition \(n_i=3 \to n_f=2\) yields the H-\(\alpha\) line with \(\lambda \approx 656\) nm. The transition \(n_i=4 \to n_f=2\) yields the H-\(\beta\) line (\(\approx 486\) nm), and \(n_i=5 \to n_f=2\) yields H-\(\gamma\) (\(\approx 434\) nm).
Paschen series (infrared): Transitions ending at \(n_f=3\) (from \(n_i=4, 5, \dots\)).
The model successfully predicted the wavelengths of these numerous spectral lines using a simple formula based on integer quantum numbers.
Limitations of the Bohr Model Despite its success with hydrogen, the Bohr model had significant limitations. It could not correctly predict the spectra of atoms with more than one electron, failed to explain the relative intensities of spectral lines, could not account for the splitting of spectral lines in magnetic fields (Zeeman effect), and was based on an inconsistent mix of classical and quantum ideas (e.g., why were only certain orbits stable?). It represented a crucial stepping stone, demonstrating the power of quantization, but a more comprehensive theory was needed. This paved the way for the development of modern quantum mechanics in the 1920s.
Formal Development of Modern Quantum Mechanics
The successes and failures of the Bohr model, along with other experimental results and theoretical insights (like Louis de Broglie’s hypothesis of matter waves in 1924), spurred intense activity in the mid-1920s to develop a consistent and comprehensive quantum theory. This led to two initially distinct but ultimately equivalent formulations.
Heisenberg’s Matrix Mechanics (1925)
Werner Heisenberg, collaborating with Max Born and Pascual Jordan, developed matrix mechanics. This approach abandoned the classical concepts of electron orbits and focused directly on observable quantities, such as the frequencies and intensities of spectral lines emitted by atoms.
Observables as Matrices: Physical quantities like position (\(\hat{X}\)), momentum (\(\hat{P}\)), and energy (Hamiltonian \(\hat{H}\)) were represented not by numbers, but by infinite-dimensional matrices.
Dynamics and Commutation Relations: The time evolution of these matrices was governed by equations analogous to classical Hamiltonian mechanics, but crucially, matrix multiplication is generally non-commutative. Heisenberg discovered that the matrices representing position and momentum must satisfy the canonical commutation relation: \[[\hat{X}, \hat{P}] = \hat{X}\hat{P} - \hat{P}\hat{X} = i\hbar \hat{I} \label{eq:heisenberg_commutation}\] where \(\hat{I}\) is the identity matrix. This non-commutativity is a fundamental departure from classical physics (where \(xp - px = 0\)) and is the mathematical root of the uncertainty principle.
Quantization Emerges Naturally: The possible energy values (spectral lines) emerged as the eigenvalues obtained by diagonalizing the Hamiltonian matrix \(\hat{H}\).
Matrix mechanics provided a mathematically rigorous framework that successfully reproduced known quantum phenomena, but its reliance on abstract infinite matrices made it difficult for many physicists to grasp intuitively at first.
Schrödinger’s Wave Mechanics (1926)
Inspired by de Broglie’s idea that particles could exhibit wave-like properties (\(\lambda = h/p\)), Erwin Schrödinger developed wave mechanics.
The Wave Function: Schrödinger proposed that the state of a quantum system (like an electron) is described by a complex-valued wave function, \(\psi(\mathbf{r}, t)\), which evolves in space and time.
The Schrödinger Equation: He formulated a differential equation, now known as the time-dependent Schrödinger equation, that governs the evolution of the wave function: \[i\hbar \frac{\partial}{\partial t} \psi(\mathbf{r}, t) = \hat{H} \psi(\mathbf{r}, t) \label{eq:tdse_intro}\] where \(\hat{H}\) is the Hamiltonian operator, constructed from classical energy expressions using specific correspondence rules (where \(\hat{H}\) is the Hamiltonian operator, constructed from classical energy expressions using specific correspondence rules (e.g., replacing momentum \(p_x\) with the differential operator \(-i\hbar \frac{\partial}{\partial x}\)).
Quantization from Boundary Conditions: For bound systems (like the electron in an atom), solving the time-independent version of the Schrödinger equation (\(\hat{H}\psi = E\psi\)) subject to appropriate boundary conditions (e.g., \(\psi \to 0\) at infinity) naturally yielded discrete energy eigenvalues \(E\), corresponding to the quantized energy levels observed experimentally.
Schrödinger’s approach, using familiar concepts from classical wave theory and differential equations, was initially perceived as more intuitive and visually appealing than matrix mechanics.
Equivalence and the Copenhagen Interpretation
Shortly after both theories were proposed, Schrödinger himself, Carl Eckart, and Paul Dirac independently demonstrated that matrix mechanics and wave mechanics were mathematically equivalent. They described the same physics using different mathematical languages.
Unifying Framework: Hilbert Space: The underlying mathematical structure for both formulations is the abstract Hilbert space. State vectors (kets \(\ket{\psi}\)) live in this space, and observables are represented by linear operators acting on these vectors. Schrödinger’s wave functions can be seen as the representation of the state vector in the position basis, while Heisenberg’s matrices are representations of operators in some chosen basis (often the energy eigenbasis).
The Copenhagen Interpretation: As the mathematical formalism solidified, the physical interpretation of the theory became a central focus. The Copenhagen interpretation, developed primarily by Niels Bohr and Werner Heisenberg, with significant contributions from Max Born, Wolfgang Pauli, and Pascual Jordan, became the standard view for many decades. Its key tenets include:
Probabilistic Nature (Born Rule): Max Born interpreted the square of the magnitude of the wave function, \(|\psi(\mathbf{r}, t)|^2\), as the probability density of finding the particle at position \(\mathbf{r}\) at time \(t\). More generally, the probability of measuring a specific eigenvalue is related to the square of the amplitude of the corresponding eigenvector in the state’s expansion (see 4.4).
Wave-Particle Duality and Complementarity: Quantum objects exhibit both wave-like and particle-like properties, depending on the experimental setup. Bohr’s principle of complementarity states that these aspects are mutually exclusive but necessary for a complete description; an experiment designed to precisely measure the particle aspect (e.g., position) necessarily obscures the wave aspect (e.g., momentum), and vice versa.
Measurement and Collapse: The act of measurement plays a crucial role. Before measurement, a system may exist in a superposition of states. Upon measurement, the system instantaneously "collapses" into one of the eigenstates corresponding to the measured observable, with the outcome determined probabilistically. The nature and mechanism of this collapse remain subjects of debate (the measurement problem).
While the Copenhagen interpretation provides a consistent operational framework, philosophical debates about the meaning of quantum mechanics and alternative interpretations (like the Many-Worlds interpretation or Bohmian mechanics) continue to this day. However, the mathematical formalism established in the 1920s remains the undisputed foundation for virtually all applications of quantum theory, including quantum computing.
Mathematical Formalism of Quantum Mechanics
Modern quantum mechanics is formulated within the mathematical framework of linear algebra over complex numbers, specifically using Hilbert spaces. This provides a rigorous and abstract structure capable of describing both discrete (like spin) and continuous (like position) quantum systems.
Hilbert Spaces
The state of a quantum system is postulated to reside in a Hilbert space associated with that system.
Hilbert Space A Hilbert space \(\mathcal{H}\) is a complex vector space equipped with an inner product \(\braket{\cdot}{\cdot}: \mathcal{H} \times \mathcal{H} \to \mathbb{C}\), such that \(\mathcal{H}\) is complete with respect to the norm induced by the inner product. The norm of a vector \(\ket{\psi} \in \mathcal{H}\) is defined as \(\norm{\psi} = \sqrt{\braket{\psi}{\psi}}\).
The Inner Product
The inner product \(\braket{\phi}{\psi}\) between two vectors \(\ket{\phi}\) and \(\ket{\psi}\) in \(\mathcal{H}\) satisfies the following properties for any \(\ket{\phi}, \ket{\psi}, \ket{\chi} \in \mathcal{H}\) and \(a, b \in \mathbb{C}\):
Linearity in the second argument (ket): \(\braket{\phi}{a\psi + b\chi} = a \braket{\phi}{\psi} + b \braket{\phi}{\chi}\)
Conjugate symmetry: \(\braket{\phi}{\psi} = \braket{\psi}{\phi}^*\) (where \(*\) denotes complex conjugation)
Anti-linearity (or conjugate linearity) in the first argument (bra): \(\braket{a\phi + b\chi}{\psi} = a^* \braket{\phi}{\psi} + b^* \braket{\chi}{\psi}\) (This follows from properties 1 and 2)
Positive-definiteness: \(\braket{\psi}{\psi} \ge 0\), and \(\braket{\psi}{\psi} = 0\) if and only if \(\ket{\psi} = \ket{0}\) (the zero vector).
The inner product allows us to define geometric concepts like orthogonality and projections in the abstract vector space. Two vectors \(\ket{\phi}\) and \(\ket{\psi}\) are orthogonal if \(\braket{\phi}{\psi} = 0\).
Completeness
Completeness means that every Cauchy sequence of vectors in \(\mathcal{H}\) converges to a vector within \(\mathcal{H}\). This property is crucial for ensuring that limits exist, allowing for the application of calculus and analysis within the Hilbert space, which is particularly important for infinite-dimensional spaces.
Dirac Notation (Bra-Ket Notation)
Paul Dirac introduced a convenient and powerful notation widely used in quantum mechanics:
Ket vectors \(\ket{\psi}\): Represent state vectors as elements of the Hilbert space \(\mathcal{H}\). They can be thought of as analogous to column vectors in finite dimensions.
Bra vectors \(\bra{\phi}\): Represent elements of the dual space \(\mathcal{H}^*\), which is the space of linear functionals mapping vectors in \(\mathcal{H}\) to complex numbers. For each ket \(\ket{\phi}\), there is a corresponding bra \(\bra{\phi}\). Bras can be thought of as analogous to row vectors (specifically, the conjugate transpose of the corresponding column vector).
Inner Product \(\braket{\phi}{\psi}\): The inner product is formed by combining a bra \(\bra{\phi}\) and a ket \(\ket{\psi}\). It represents the action of the linear functional \(\bra{\phi}\) on the vector \(\ket{\psi}\), yielding a complex number.
Outer Product \(\dyad{\psi}{\phi}\): This forms an operator (specifically, a linear map from \(\mathcal{H}\) to \(\mathcal{H}\)). When applied to a ket \(\ket{\chi}\), it yields \(\dyad{\psi}{\phi}\ket{\chi} = \ket{\psi}\braket{\phi}{\chi}\), which is the vector \(\ket{\psi}\) scaled by the complex number \(\braket{\phi}{\chi}\). Projection operators are often written in this form (e.g., \(\hat{P}_\psi = \dyad{\psi}{\psi}\) for a normalized state \(\ket{\psi}\)).
State Vectors and Normalization
The physical state of a quantum system is represented by a normalized vector \(\ket{\psi}\) in the appropriate Hilbert space \(\mathcal{H}\), meaning \(\braket{\psi}{\psi} = 1\).
Physical States as Rays Technically, a physical state corresponds to a *ray* in Hilbert space, meaning that \(\ket{\psi}\) and \(e^{i\theta}\ket{\psi}\) (for any real phase \(\theta\)) represent the same physical state. This overall phase factor does not affect measurement probabilities, which depend on \(|\braket{\phi}{\psi}|^2\). Normalization to unity (\(\braket{\psi}{\psi}=1\)) is a convention that simplifies the probabilistic interpretation (Born rule, see 4.4), where probabilities are given by the squared magnitudes of amplitudes.
Finite vs. Infinite-Dimensional Hilbert Spaces
The dimension of the Hilbert space depends on the system being described:
Finite-Dimensional: Systems with a finite number of discrete states are described by finite-dimensional Hilbert spaces. The most important example in quantum computing is a single qubit, whose state space is the 2-dimensional complex Hilbert space \(\mathcal{H} = \mathbb{C}^2\). An \(N\)-level system (qudit) uses \(\mathbb{C}^N\). A system of \(n\) qubits uses the \(2^n\)-dimensional Hilbert space \(\mathcal{H} = (\mathbb{C}^2)^{\otimes n}\). In these cases, kets can be represented as column vectors and bras as conjugate-transpose row vectors.
Infinite-Dimensional: Systems with continuous degrees of freedom (like position or momentum) or an infinite number of discrete states (like the energy levels of a harmonic oscillator) require infinite-dimensional Hilbert spaces. A common example is the space of square-integrable functions.
Example: Function Spaces (\(L^2\))
For systems described by continuous variables, such as the position \(\mathbf{r}\) of a particle in \(d\) dimensions, the relevant Hilbert space is often the space of square-integrable complex-valued functions, denoted \(L^2(\mathbb{R}^d)\).
Square Integrable Functions
The space \(L^2(\mathbb{R}^d)\) consists of all complex-valued functions \(\psi(\mathbf{r})\) defined on \(\mathbb{R}^d\) such that the integral of their squared magnitude over all space is finite: \[L^2(\mathbb{R}^d) = \left\{ \psi: \mathbb{R}^d \to \mathbb{C} \;\middle|\; \int_{\mathbb{R}^d} |\psi(\mathbf{r})|^2 d^d\mathbf{r} < \infty \right\} \label{eq:L2_space}\] The condition \(\int |\psi(\mathbf{r})|^2 d^d\mathbf{r} < \infty\) ensures that the function is normalizable, consistent with the probabilistic interpretation where \(|\psi(\mathbf{r})|^2\) represents the probability density. The state of a spinless particle in 3D space is typically described by a wave function \(\psi(\mathbf{r}, t) \in L^2(\mathbb{R}^3)\) for any given time \(t\).
Inner Product for Functions
The inner product between two functions (state vectors) \(\phi(\mathbf{r})\) and \(\psi(\mathbf{r})\) in \(L^2(\mathbb{R}^d)\) is defined by the integral: \[\braket{\phi}{\psi} = \int_{\mathbb{R}^d} \phi^*(\mathbf{r}) \psi(\mathbf{r}) d^d\mathbf{r} \label{eq:inner_product_L2}\] Here, \(\phi^*(\mathbf{r})\) is the complex conjugate of \(\phi(\mathbf{r})\). This definition satisfies all the properties of an inner product listed in 3.1.1. The normalization condition \(\braket{\psi}{\psi}=1\) translates to \(\int_{\mathbb{R}^d} |\psi(\mathbf{r})|^2 d^d\mathbf{r} = 1\).
Wave Functions as Representations
It’s important to connect the abstract state vector \(\ket{\psi}\) with its concrete representation as a wave function \(\psi(\mathbf{r})\). The wave function \(\psi(\mathbf{r})\) can be understood as the "component" of the abstract state vector \(\ket{\psi}\) along the continuous basis vector \(\ket{\mathbf{r}}\) corresponding to the position \(\mathbf{r}\). That is: \[\psi(\mathbf{r}) = \braket{\mathbf{r}}{\psi} \label{eq:psi_as_projection}\] The position "eigenvectors" \(\ket{\mathbf{r}}\) are not strictly in the Hilbert space \(L^2\) themselves (they correspond to Dirac delta functions), but they form a useful continuous basis. The inner product [eq:inner_product_L2] can then be seen as arising from the abstract inner product via an insertion of the identity operator expressed as an integral over the position basis: \(\hat{I} = \int d^d\mathbf{r} \dyad{\mathbf{r}}{\mathbf{r}}\). \[\braket{\phi}{\psi} = \bra{\phi} \hat{I} \ket{\psi} = \bra{\phi} \left( \int d^d\mathbf{r} \dyad{\mathbf{r}}{\mathbf{r}} \right) \ket{\psi} = \int d^d\mathbf{r} \braket{\phi}{\mathbf{r}} \braket{\mathbf{r}}{\psi} = \int d^d\mathbf{r} \phi^*(\mathbf{r}) \psi(\mathbf{r})\]
Basis and Expansion
Just like in finite-dimensional vector spaces, we can often find a set of basis vectors that span the Hilbert space \(\mathcal{H}\). If the basis vectors \(\{\ket{u_i}\}\) are orthonormal, meaning \(\braket{u_i}{u_j} = \delta_{ij}\) (where \(\delta_{ij}\)is the Kronecker delta), any state vector \(\ket{\psi}\) can be expanded as a linear combination of these basis vectors: \[\ket{\psi} = \sum_i c_i \ket{u_i} \label{eq:basis_expansion_discrete}\] The coefficients \(c_i\) are complex numbers given by the projection of \(\ket{\psi}\) onto the basis vector \(\ket{u_i}\): \[c_i = \braket{u_i}{\psi} \label{eq:expansion_coeffs}\] The normalization condition \(\braket{\psi}{\psi}=1\) implies \(\sum_i |c_i|^2 = 1\). This expansion is fundamental for understanding measurements, as the basis is often chosen to be the set of eigenvectors of the observable being measured. For continuous bases like position \(\ket{\mathbf{r}}\), the sum becomes an integral: \(\ket{\psi} = \int d^d\mathbf{r} \, \psi(\mathbf{r}) \ket{\mathbf{r}}\), where \(\psi(\mathbf{r}) = \braket{\mathbf{r}}{\psi}\).
Postulates of Quantum Mechanics
The behavior and description of quantum systems are governed by a set of fundamental postulates, which form the axiomatic basis of the theory.
Postulate 1: State Space Description
State Space Associated with any isolated physical system is a complex vector space with an inner product (i.e., a Hilbert space \(\mathcal{H}\)) known as the state space of the system. The system is completely described at any time \(t\) by its state vector \(\ket{\psi(t)}\), which is a normalized vector in \(\mathcal{H}\) (\(\braket{\psi(t)}{\psi(t)} = 1\)).
Completeness and Superposition The phrase "completely described" is a strong statement, implying that the state vector \(\ket{\psi(t)}\) contains all possible information about the system at time \(t\). A direct consequence of the vector space structure is the principle of superposition: if \(\ket{\psi_1}\) and \(\ket{\psi_2}\) are two possible states of the system, then any linear combination \[\ket{\psi} = c_1 \ket{\psi_1} + c_2 \ket{\psi_2} \quad (c_1, c_2 \in \mathbb{C}) \label{eq:superposition}\] is also a valid state (after normalization). This principle is responsible for many non-classical phenomena, including interference. Applying superposition to macroscopic systems leads to conceptual challenges, famously illustrated by the Schrödinger’s Cat paradox (8.2.1), where a cat can seemingly exist in a superposition of "alive" and "dead" states before observation.
Postulate 2: Observables and Operators
Observables Every measurable physical quantity \(O\) (an observable) is associated with a linear Hermitian operator \(\hat{O}\) acting on the Hilbert space \(\mathcal{H}\) of the system.
Hermitian Operator An operator \(\hat{O}\) is Hermitian (or self-adjoint in the finite-dimensional case, with subtleties in infinite dimensions) if it is equal to its adjoint (conjugate transpose), \(\hat{O} = \hat{O}^\dagger\). A key property of Hermitian operators is that their eigenvalues are always real numbers, consistent with the fact that measurement outcomes are real quantities.
Common Observables Examples include:
Position operator \(\hat{X}\)
Momentum operator \(\hat{P}_x = -i\hbar \frac{\partial}{\partial x}\)
Energy operator (Hamiltonian) \(\hat{H}\)
Orbital angular momentum operators \(\hat{L}_x, \hat{L}_y, \hat{L}_z, \hat{L}^2\)
Spin angular momentum operators \(\hat{S}_x, \hat{S}_y, \hat{S}_z, \hat{S}^2\)
For a qubit, the Pauli operators \(\sigma_x, \sigma_y, \sigma_z\) (often denoted \(\hat{X}, \hat{Y}, \hat{Z}\)) are Hermitian and represent measurements along the respective axes on the Bloch sphere.
Postulate 3: Measurement Outcomes
Measurement Outcomes The only possible results that can be obtained when measuring an observable \(O\) are the eigenvalues \(\lambda_i\) of the corresponding Hermitian operator \(\hat{O}\).
The eigenvalues \(\lambda_i\) are the values that satisfy the eigenvalue equation: \[\hat{O} \ket{\lambda_i} = \lambda_i \ket{\lambda_i} \label{eq:eigenvalue_eq}\] where \(\ket{\lambda_i}\) are the eigenvectors corresponding to the eigenvalue \(\lambda_i\).
Spectral Theorem A fundamental result for Hermitian operators (the spectral theorem) guarantees that their eigenvectors \(\{\ket{\lambda_i}\}\) can be chosen to form a complete orthonormal basis for the Hilbert space \(\mathcal{H}\). This means any state \(\ket{\psi}\) can be expressed as a linear combination of these eigenvectors (see [eq:state_expansion]). This property is crucial for the interpretation of measurement probabilities.
Postulate 4: Measurement Probabilities and Statistics
This postulate quantifies the probabilistic nature of quantum measurements.
Probabilities (Born Rule)
Born Rule If a quantum system is in a normalized state \(\ket{\psi}\) at the time an observable \(O\) is measured, the probability \(P(\lambda_i)\) of obtaining the specific eigenvalue \(\lambda_i\) as the measurement outcome is given by the squared magnitude of the projection of the state vector onto the eigenspace corresponding to \(\lambda_i\). If \(\lambda_i\) is non-degenerate with normalized eigenvector \(\ket{\lambda_i}\), the probability is: \[P(\lambda_i) = |\braket{\lambda_i}{\psi}|^2 \label{eq:born_rule_discrete}\] If \(\lambda_i\) is degenerate, let \(\hat{P}_i\) be the projection operator onto the eigenspace \(V_i\) associated with \(\lambda_i\). Then the probability is: \[P(\lambda_i) = \bra{\psi} \hat{P}_i \ket{\psi} = \norm{\hat{P}_i \ket{\psi}}^2 \label{eq:born_rule_degenerate}\]
If we expand the state vector \(\ket{\psi}\) in the orthonormal basis of eigenvectors \(\{\ket{\lambda_i}\}\) (assuming non-degeneracy for simplicity): \[\ket{\psi} = \sum_j c_j \ket{\lambda_j}, \quad \text{where } c_j = \braket{\lambda_j}{\psi} \label{eq:state_expansion}\] Then the probability of measuring \(\lambda_i\) is simply the squared magnitude of the corresponding expansion coefficient: \[P(\lambda_i) = |c_i|^2 = |\braket{\lambda_i}{\psi}|^2 \label{eq:prob_ci_squared}\] The normalization condition \(\braket{\psi}{\psi}=1\) ensures that the probabilities sum to unity: \(\sum_i P(\lambda_i) = \sum_i |c_i|^2 = 1\).
Expectation Values
If one performs many measurements of the observable \(O\) on identically prepared systems, each in the state \(\ket{\psi}\), the average value of the results obtained will converge to the expectation value \(\expval{\hat{O}}_\psi\).
Expectation Value The expectation value of an observable \(\hat{O}\) for a system in state \(\ket{\psi}\) is given by: \[\expval{\hat{O}}_\psi = \braket{\psi}{\hat{O}|\psi} \label{eq:expectation_value}\]
Using the expansion [eq:state_expansion] and the eigenvalue equation [eq:eigenvalue_eq], we can verify the consistency with the probabilistic interpretation: \[\begin{aligned} \expval{\hat{O}}_\psi &= \bra{\psi} \hat{O} \left( \sum_j c_j \ket{\lambda_j} \right) = \bra{\psi} \sum_j c_j (\hat{O} \ket{\lambda_j}) \nonumber \\ &= \bra{\psi} \sum_j c_j \lambda_j \ket{\lambda_j} = \left( \sum_i c_i^* \bra{\lambda_i} \right) \left( \sum_j c_j \lambda_j \ket{\lambda_j} \right) \nonumber \\ &= \sum_i \sum_j c_i^* c_j \lambda_j \braket{\lambda_i}{\lambda_j} = \sum_i \sum_j c_i^* c_j \lambda_j \delta_{ij} \nonumber \\ &= \sum_i |c_i|^2 \lambda_i = \sum_i P(\lambda_i) \lambda_i \label{eq:expectation_derivation}\end{aligned}\] This confirms that the expectation value is the weighted average of the possible outcomes \(\lambda_i\), with weights given by their respective probabilities \(P(\lambda_i)\).
Measuring \(\sigma_z\) on a Superposition State Consider a qubit in the state \(\ket{\psi} = \frac{1}{\sqrt{2}}(\ket{0} + \ket{1})\). We want to measure the observable corresponding to the Pauli-Z operator, \(\hat{Z} = \sigma_z\). The eigenvectors of \(\sigma_z\) are \(\ket{0}\) (eigenvalue \(\lambda_0 = +1\)) and \(\ket{1}\) (eigenvalue \(\lambda_1 = -1\)). The probabilities are: \[\begin{aligned} P(+1) &= |\braket{0}{\psi}|^2 = \left| \braket{0}{\frac{1}{\sqrt{2}}(\ket{0} + \ket{1})} \right|^2 = \left| \frac{1}{\sqrt{2}} (\braket{0}{0} + \braket{0}{1}) \right|^2 = \left| \frac{1}{\sqrt{2}}(1+0) \right|^2 = \frac{1}{2} \\ P(-1) &= |\braket{1}{\psi}|^2 = \left| \braket{1}{\frac{1}{\sqrt{2}}(\ket{0} + \ket{1})} \right|^2 = \left| \frac{1}{\sqrt{2}} (\braket{1}{0} + \braket{1}{1}) \right|^2 = \left| \frac{1}{\sqrt{2}}(0+1) \right|^2 = \frac{1}{2} \end{aligned}\] The expectation value is: \[\begin{aligned} \expval{\hat{Z}}_\psi &= \braket{\psi}{\hat{Z}|\psi} = \left( \frac{1}{\sqrt{2}}(\bra{0} + \bra{1}) \right) \hat{Z} \left( \frac{1}{\sqrt{2}}(\ket{0} + \ket{1}) \right) \\ &= \frac{1}{2} (\bra{0} + \bra{1}) (\hat{Z}\ket{0} + \hat{Z}\ket{1}) \\ &= \frac{1}{2} (\bra{0} + \bra{1}) ((+1)\ket{0} + (-1)\ket{1}) \\ &= \frac{1}{2} (\braket{0}{0} - \braket{0}{1} + \braket{1}{0} - \braket{1}{1}) \\ &= \frac{1}{2} (1 - 0 + 0 - 1) = 0 \end{aligned}\] Alternatively, using the probabilities: \(\expval{\hat{Z}}_\psi = P(+1)\lambda_0 + P(-1)\lambda_1 = \frac{1}{2}(+1) + \frac{1}{2}(-1) = 0\). An average value of 0 makes sense, as the outcomes +1 and -1 are equally likely.
Superposition vs. Statistical Mixture
It is crucial to understand that a system in a superposition state, like \(\ket{\psi} = \frac{1}{\sqrt{2}}(\ket{0} + \ket{1})\), is fundamentally different from a statistical ensemble (a mixture) where the system is either in state \(\ket{0}\) with 50
Superposition: The system genuinely exists in a combination of states *before* measurement. This allows for quantum interference effects when measuring observables whose eigenstates are different from \(\ket{0}\) and \(\ket{1}\) (e.g., measuring \(\sigma_x\)).
Mixture: The system is definitely in one specific state, but we have classical uncertainty about which one. No interference effects characteristic of superposition occur. Mixtures are described by density operators, not single state vectors.
Quantum mechanics asserts that the pre-measurement state *is* the superposition. The probabilistic outcome arises only upon measurement. The Schrödinger’s Cat paradox arises precisely because applying this principle to macroscopic objects seems to imply they can exist in paradoxical superpositions until observed.
Postulate 5: Time Evolution and Measurement Collapse
This postulate describes how quantum states change over time, distinguishing between evolution between measurements and the effect of measurement itself.
Unitary Time Evolution
Time Evolution Between measurements, the time evolution of the state vector \(\ket{\psi(t)}\) of an isolated quantum system is governed by the time-dependent Schrödinger equation (TDSE): \[i\hbar \dv{t} \ket{\psi(t)} = \hat{H} \ket{\psi(t)} \label{eq:tdse_postulate}\] where \(\hat{H}\) is the Hamiltonian operator of the system (the observable corresponding to the total energy).
The Hamiltonian \(\hat{H}\) is the generator of time translations. If \(\hat{H}\) is time-independent, the formal solution to the TDSE is given by applying a unitary time evolution operator \(\hat{U}(t, t_0)\): \[\ket{\psi(t)} = \hat{U}(t, t_0) \ket{\psi(t_0)} = e^{-i\hat{H}(t-t_0)/\hbar} \ket{\psi(t_0)} \label{eq:unitary_evolution_op_post}\] This evolution is deterministic, continuous, and reversible (unitary operators preserve inner products and norms). Quantum gates in quantum computing are implementations of such unitary operators.
Measurement Collapse (Projection Postulate)
Measurement Collapse Upon measurement of an observable \(O\) on a system in state \(\ket{\psi}\), if the measurement yieldsUpon measurement of an observable \(O\) on a system in state \(\ket{\psi}\), if the measurement yields the result \(\lambda_i\), the state of the system immediately after the measurement collapses (or is projected) onto the normalized state corresponding to that outcome. If \(\hat{P}_i\) is the projector onto the eigenspace of \(\lambda_i\), the post-measurement state \(\ket{\psi'}\) is: \[\ket{\psi'} = \frac{\hat{P}_i \ket{\psi}}{\sqrt{\bra{\psi} \hat{P}_i \ket{\psi}}} = \frac{\hat{P}_i \ket{\psi}}{\sqrt{P(\lambda_i)}} \label{eq:collapse_postulate}\]
This process is fundamentally different from the unitary evolution described by the Schrödinger equation:
It is probabilistic (the outcome \(\lambda_i\) is random, governed by the Born rule).
It is non-unitary and generally irreversible. Information about the initial superposition coefficients (relative phases) is lost.
It appears instantaneous.
The Measurement Problem The nature of measurement and the collapse postulate are perhaps the most controversial aspects of quantum mechanics. Questions arise about what constitutes a "measurement," whether it requires a conscious observer, and how the deterministic unitary evolution transitions into the probabilistic collapse. This is known as the measurement problem, and various interpretations (Copenhagen, Many-Worlds, Decoherence, etc.) attempt to address it. For practical purposes in quantum computing, we typically apply the collapse rule as stated.
Dynamics and Evolution in Quantum Mechanics
Having established the state space and the representation of observables, we now turn to how quantum states change over time and how the quantum description connects to classical mechanics.
The Correspondence Principle and Operator Construction
A key step in applying quantum mechanics is constructing the appropriate quantum operators, especially the Hamiltonian, for a given physical system. The correspondence principle provides guidance, suggesting that quantum mechanics should reproduce classical mechanics in the appropriate limit (e.g., for large quantum numbers or macroscopic systems). More practically, it involves promoting classical variables to quantum operators.
Canonical Quantization
For systems with classical analogues described by coordinates \(q_i\) and conjugate momenta \(p_i\), canonical quantization provides a standard procedure:
Replace classical variables \(q_i, p_i\) with corresponding operators \(\hat{Q}_i, \hat{P}_i\).
Impose the canonical commutation relations: \[\begin{aligned} [\hat{Q}_i, \hat{P}_j] &= i\hbar \delta_{ij} \hat{I} \\ [\hat{Q}_i, \hat{Q}_j] &= 0 \\ [\hat{P}_i, \hat{P}_j] &= 0 \end{aligned}\]
Construct the quantum Hamiltonian operator \(\hat{H}\) by substituting the operators \(\hat{Q}_i, \hat{P}_i\) into the classical Hamiltonian function \(H(q_i, p_i)\). Ambiguities can arise due to non-commutativity (e.g., does \(qp\) become \(\hat{Q}\hat{P}\) or \(\hat{P}\hat{Q}\)?); symmetrization rules (like Weyl ordering) are often employed.
Position and Momentum Operators in Wave Mechanics
In the context of Schrödinger’s wave mechanics, where states are functions \(\psi(\mathbf{r})\) of position \(\mathbf{r}=(x,y,z)\), the canonical quantization rules are realized by specific differential and multiplicative operators:
Position Operators (\(\hat{X}, \hat{Y}, \hat{Z}\)): Act by multiplication. \[\hat{X} \psi(\mathbf{r}) = x \psi(\mathbf{r}), \quad \hat{Y} \psi(\mathbf{r}) = y \psi(\mathbf{r}), \quad \hat{Z} \psi(\mathbf{r}) = z \psi(\mathbf{r}) \label{eq:op_XYZ}\]
Momentum Operators (\(\hat{P}_x, \hat{P}_y, \hat{P}_z\)): Act as derivatives. \[\hat{P}_x \psi(\mathbf{r}) = -i\hbar \pdv{}{x} \psi(\mathbf{r}), \quad \hat{P}_y \psi(\mathbf{r}) = -i\hbar \pdv{}{y} \psi(\mathbf{r}), \quad \hat{P}_z \psi(\mathbf{r}) = -i\hbar \pdv{}{z} \psi(\mathbf{r}) \label{eq:op_Pxyz}\] These can be compactly written using the gradient operator \(\nabla\): \(\hat{\mathbf{P}} = -i\hbar \nabla\).
One can verify that these operators satisfy the canonical commutation relations, e.g., \([\hat{X}, \hat{P}_x] = i\hbar \hat{I}\) (see 6.1.2).
Eigenstates of Position and Momentum The eigenstates of the position operator \(\hat{X}\) corresponding to eigenvalue \(x_0\) are Dirac delta functions \(\delta(x-x_0)\). These are not functions in the Hilbert space \(L^2(\mathbb{R})\) but are distributions. Similarly, the eigenstates of the momentum operator \(\hat{P}_x\) corresponding to eigenvalue \(p_0\) are plane waves \(e^{ip_0x/\hbar}\), which are also not in \(L^2(\mathbb{R})\) (they are not square-integrable over all space). This necessitates a more sophisticated mathematical framework (rigged Hilbert spaces or Gel’fand triples) for a fully rigorous treatment, but the basic formalism and physical interpretations remain valid.
Constructing the Hamiltonian
Using the correspondence rules, the Hamiltonian operator is constructed from the classical expression for total energy. For a non-relativistic particle of mass \(m\) moving in a potential \(V(\mathbf{r})\):
Classical Hamiltonian: \(H = \frac{\mathbf{p}^2}{2m} + V(\mathbf{r}) = \frac{p_x^2 + p_y^2 + p_z^2}{2m} + V(x,y,z)\)
Quantum Hamiltonian Operator: \[\begin{aligned} \hat{H} &= \frac{\hat{\mathbf{P}}^2}{2m} + V(\hat{\mathbf{R}}) = \frac{\hat{P}_x^2 + \hat{P}_y^2 + \hat{P}_z^2}{2m} + V(\hat{X},\hat{Y},\hat{Z}) \nonumber \\ &= \frac{(-i\hbar \nabla) \cdot (-i\hbar \nabla)}{2m} + V(x,y,z) \nonumber \\ &= -\frac{\hbar^2}{2m} \nabla^2 + V(x,y,z) \label{eq:op_H_particle_3D} \end{aligned}\] where \(\nabla^2 = \frac{\partial^2}{\partial x^2} + \frac{\partial^2}{\partial y^2} + \frac{\partial^2}{\partial z^2}\) is the Laplacian operator.
This operator \(\hat{H}\) is Hermitian if the potential \(V(\mathbf{r})\) is real-valued.
The Schrödinger Equation
The Schrödinger equation is the fundamental equation governing the time evolution of a quantum state (Postulate 5a).
Time-Dependent Schrödinger Equation (TDSE)
As stated in [eq:tdse_postulate], the TDSE is: \[i\hbar \pdv{}{t} \ket{\psi(t)} = \hat{H} \ket{\psi(t)} \label{eq:tdse_repeat}\] This equation connects the temporal change of the state vector to the total energy operator \(\hat{H}\). It can be motivated by the energy correspondence \(E \leftrightarrow i\hbar \frac{\partial}{\partial t}\) applied to the state vector, equating it to the action of the Hamiltonian operator \(\hat{H}\) (representing energy) on the state vector. In the position representation for a particle in a potential, it becomes the partial differential equation: \[i\hbar \pdv{}{t} \psi(\mathbf{r}, t) = \left( -\frac{\hbar^2}{2m} \nabla^2 + V(\mathbf{r}) \right) \psi(\mathbf{r}, t) \label{eq:tdse_wavefunction}\] Solving this equation, given an initial state \(\psi(\mathbf{r}, t_0)\), determines the state \(\psi(\mathbf{r}, t)\) at all future times.
Time-Independent Schrödinger Equation (TISE)
If the Hamiltonian \(\hat{H}\) does not explicitly depend on time, we can look for solutions of a special form called stationary states. These are states whose probability density \(|\psi(\mathbf{r},t)|^2\) is constant in time. They can be found using the method of separation of variables, assuming \(\ket{\psi(t)} = \ket{\phi} f(t)\). Substituting into the TDSE yields: \[i\hbar \ket{\phi} \dv{f}{t} = (\hat{H}\ket{\phi}) f(t)\] Dividing by \(\ket{\phi}f(t)\) gives \(\frac{i\hbar}{f(t)}\dv{f}{t} = \frac{\hat{H}\ket{\phi}}{\ket{\phi}}\). Since the left side depends only on \(t\) and the right side depends only on the spatial variables (or other degrees of freedom acted upon by \(\hat{H}\)), both sides must equal a constant, which we call \(E\). This leads to two equations:
Time-Independent Schrödinger Equation (TISE): \[\hat{H} \ket{\phi} = E \ket{\phi} \label{eq:tise}\] This is an eigenvalue equation for the Hamiltonian operator. Its solutions \(\ket{\phi_n}\) are the energy eigenvectors (stationary states) with corresponding energy eigenvalues \(E_n\).
Time Evolution Equation: \(i\hbar \dv{f}{t} = E f(t)\), which has the solution \(f(t) = e^{-iEt/\hbar}\) (up to a constant factor).
Therefore, the stationary state solutions to the TDSE are of the form: \[\ket{\psi_n(t)} = \ket{\phi_n} e^{-iE_n t/\hbar} \label{eq:stationary_state}\] These states have a definite energy \(E_n\), and their probability density \(|\psi_n(\mathbf{r},t)|^2 = |\phi_n(\mathbf{r}) e^{-iE_n t/\hbar}|^2 = |\phi_n(\mathbf{r})|^2\) is indeed time-independent. Any general solution \(\ket{\psi(t)}\) to the TDSE can be expressed as a superposition of these stationary states (since the eigenvectors \(\ket{\phi_n}\) form a basis): \[\ket{\psi(t)} = \sum_n c_n \ket{\psi_n(t)} = \sum_n c_n \ket{\phi_n} e^{-iE_n t/\hbar} \label{eq:general_solution_tdse}\] where the coefficients \(c_n = \braket{\phi_n}{\psi(0)}\) are determined by the initial state \(\ket{\psi(0)}\).
Time Evolution Operator
The evolution described by the Schrödinger equation can be encapsulated by a time evolution operator.
Definition and Formal Solution
For a time-independent Hamiltonian \(\hat{H}\), the operator \(\hat{U}(t, t_0)\) that evolves the state from time \(t_0\) to time \(t\) is given by the exponential of the Hamiltonian: \[\hat{U}(t, t_0) = e^{-i\hat{H}(t-t_0)/\hbar} \label{eq:time_evolution_op}\] such that \(\ket{\psi(t)} = \hat{U}(t, t_0) \ket{\psi(t_0)}\). The exponential of an operator \(\hat{A}\) is formally defined by its Taylor series expansion: \[e^{\hat{A}} = \sum_{k=0}^{\infty} \frac{\hat{A}^k}{k!} = \hat{I} + \hat{A} + \frac{\hat{A}^2}{2!} + \frac{\hat{A}^3}{3!} + \dots \label{eq:operator_exponential}\] This definition is particularly useful when \(\hat{H}\) can be diagonalized or has simple properties (like Pauli matrices where \(\sigma_k^2 = \hat{I}\)).
Unitary Nature of Time Evolution
A crucial property of the time evolution operator \(\hat{U}(t, t_0)\) for a Hermitian Hamiltonian \(\hat{H}\) is that it is unitary.
Unitary Operator An operator \(\hat{U}\) is unitary if its adjoint \(\hat{U}^\dagger\) is also its inverse: \[\hat{U}^\dagger \hat{U} = \hat{U} \hat{U}^\dagger = \hat{I} \label{eq:unitary_def}\]
Unitary Property of \(\hat{U}(t, t_0)\) If the Hamiltonian \(\hat{H}\) is Hermitian, the time evolution operator \(\hat{U}(t, t_0) = e^{-i\hat{H}(t-t_0)/\hbar}\) is unitary.
Proof. Proof. Let \(\Delta t = t-t_0\). \(\hat{U}(\Delta t)^\dagger = \left( e^{-i\hat{H}\Delta t/\hbar} \right)^\dagger\). Using the property \((e^{\hat{A}})^\dagger = e^{\hat{A}^\dagger}\), we get: \(\hat{U}(\Delta t)^\dagger = e^{(-i\hat{H}\Delta t/\hbar)^\dagger} = e^{i\hat{H}^\dagger \Delta t/\hbar}\). Since \(\hat{H}\) is Hermitian, \(\hat{H}^\dagger = \hat{H}\). Therefore: \(\hat{U}(\Delta t)^\dagger = e^{i\hat{H}\Delta t/\hbar}\). Now, consider the product: \(\hat{U}(\Delta t)^\dagger \hat{U}(\Delta t) = e^{i\hat{H}\Delta t/\hbar} e^{-i\hat{H}\Delta t/\hbar}\). Since \(\hat{H}\) commutes with itself, \(e^{\hat{A}}e^{\hat{B}} = e^{\hat{A}+\hat{B}}\) holds. \(\hat{U}(\Delta t)^\dagger \hat{U}(\Delta t) = e^{(i\hat{H}\Delta t/\hbar) + (-i\hat{H}\Delta t/\hbar)} = e^0 = \hat{I}\). Similarly, \(\hat{U}(\Delta t) \hat{U}(\Delta t)^\dagger = \hat{I}\). Thus, \(\hat{U}(\Delta t)\) is unitary. ◻
The unitarity of time evolution has profound physical consequences:
Conservation of Probability: Unitary operators preserve the norm of vectors: \(\norm{\psi(t)}^2 = \braket{\psi(t)}{\psi(t)} = \bra{\psi(t_0)} \hat{U}^\dagger \hat{U} \ket{\psi(t_0)} = \braket{\psi(t_0)}{\psi(t_0)} = \norm{\psi(t_0)}^2\). If a state starts normalized, it remains normalized. Total probability is conserved.
Preservation of Inner Products: Unitary operators preserve inner products between states: \(\braket{\phi(t)}{\psi(t)} = \braket{\phi(t_0)}{\psi(t_0)}\). This implies that angles and orthogonality relationships are maintained during evolution.
Reversibility: Since \(\hat{U}\) has an inverse (\(\hat{U}^{-1} = \hat{U}^\dagger\)), the evolution is reversible. Knowing the state at time \(t\), one can determine the state at time \(t_0\) by applying \(\hat{U}^\dagger\).
Unitary evolution corresponds geometrically to a rotation (possibly combined with reflections, though time evolution is typically continuous from the identity) in the Hilbert space.
Connection to Quantum Gates
The concept of unitary time evolution is central to quantum computing.
Quantum Gates as Unitary Operators: The operations performed on qubits in a quantum computer are represented by unitary operators (quantum gates). Examples include the Pauli gates (\(\hat{X}, \hat{Y}, \hat{Z}\)), Hadamard gate (\(\hat{H}\)), phase gates (\(\hat{S}, \hat{T}\)), and multi-qubit gates like CNOT.
Physical Implementation: Implementing a specific quantum gate \(\hat{U}_{gate}\) physically involves engineering a system Hamiltonian \(\hat{H}_{control}\) (often time-dependent, controlled by external fields like lasers or microwaves) and applying it for a precise duration \(\tau\) such that the resulting time evolution operator is the desired gate: \[\hat{U}_{gate} = \mathcal{T} \exp\left(-\frac{i}{\hbar} \int_0^\tau \hat{H}_{control}(t') dt'\right) \label{eq:gate_implementation}\] where \(\mathcal{T}\) denotes time-ordering for time-dependent Hamiltonians. For a constant control Hamiltonian \(\hat{H}_{control}\), this simplifies to \(\hat{U}_{gate} = e^{-i\hat{H}_{control}\tau/\hbar}\).
Quantum Circuits: A quantum algorithm is represented as a sequence of quantum gates applied to an initial state, followed by a measurement. The overall transformation before measurement is the product of the unitary gate operators: \[\ket{\psi_{final}} = \hat{U}_k \dots \hat{U}_2 \hat{U}_1 \ket{\psi_{initial}} \label{eq:quantum_circuit}\] The entire computation (excluding measurement) is a unitary transformation.
Understanding the dynamics dictated by the Schrödinger equation and the properties of the resulting unitary evolution is therefore fundamental to understanding how quantum computers operate.
Further Concepts and Examples
Building upon the postulates, we explore some key consequences and illustrative examples that are fundamental to quantum mechanics and relevant to quantum computing.
The Heisenberg Uncertainty Principle
One of the most striking features distinguishing quantum from classical mechanics is the existence of fundamental limits on the precision with which certain pairs of physical properties can be simultaneously known.
Non-commuting Observables and Simultaneous Measurability
Recall that observables are represented by Hermitian operators (Postulate 2). The order in which operations are performed matters in quantum mechanics if the corresponding operators do not commute. The commutator of two operators \(\hat{A}\) and \(\hat{B}\) is defined as: \[[\hat{A}, \hat{B}] = \hat{A}\hat{B} - \hat{B}\hat{A} \label{eq:commutator}\]
Two observables \(\hat{A}\) and \(\hat{B}\) possess a complete set of common eigenvectors (meaning they can be simultaneously diagonalized and thus measured with arbitrary precision simultaneously) if and only if they commute, i.e., \([\hat{A}, \hat{B}] = 0\).
If \([\hat{A}, \hat{B}] \neq 0\), the observables are said to be incompatible. Measuring one observable with high precision inherently introduces uncertainty into the value of the other.
Position and Momentum Uncertainty
The canonical example involves the position operator \(\hat{X}\) and the corresponding momentum operator \(\hat{P}_x\). As derived previously (5.1.2), in the position representation \(\hat{X} = x\) and \(\hat{P}_x = -i\hbar \pdv{}{x}\). Their commutator is: \[\begin{aligned} [\hat{X}, \hat{P}_x] \psi(x) &= (\hat{X}\hat{P}_x - \hat{P}_x\hat{X}) \psi(x) \nonumber \\ &= x \left(-i\hbar \pdv{\psi}{x}\right) - \left(-i\hbar \pdv{}{x} (x \psi(x))\right) \nonumber \\ &= -i\hbar x \pdv{\psi}{x} + i\hbar \left( \psi(x) \cdot 1 + x \pdv{\psi}{x} \right) \quad \text{(using product rule for differentiation)} \nonumber \\ &= -i\hbar x \pdv{\psi}{x} + i\hbar \psi(x) + i\hbar x \pdv{\psi}{x} \nonumber \\ &= i\hbar \psi(x) \label{eq:xp_commutator_derivation}\end{aligned}\] Since this holds for any arbitrary state function \(\psi(x)\), the operator relation is: \[[\hat{X}, \hat{P}_x] = i\hbar \hat{I} \label{eq:xp_commutator}\] Because the commutator is non-zero, position and momentum are incompatible observables.
The General Uncertainty Principle (Robertson Relation)
The quantitative statement of the uncertainty principle for any pair of observables \(\hat{A}\) and \(\hat{B}\) was derived by H. P. Robertson.
For any quantum state \(\ket{\psi}\) and any two Hermitian operators \(\hat{A}\) and \(\hat{B}\), the product of the uncertainties (standard deviations) in the measurement outcomes satisfies: \[\Delta A \, \Delta B \ge \frac{1}{2} |\expval{[\hat{A}, \hat{B}]}_\psi| \label{eq:uncertainty_principle}\] where the uncertainty \(\Delta O\) is defined as the standard deviation: \[\Delta O = \sqrt{\expval{\hat{O}^2}_\psi - (\expval{\hat{O}}_\psi)^2} \label{eq:std_dev_def}\] and \(\expval{\cdot}_\psi = \braket{\psi}{\cdot|\psi}\).
Applying this to position and momentum, using \([\hat{X}, \hat{P}_x] = i\hbar \hat{I}\): \[\Delta X \, \Delta P_x \ge \frac{1}{2} |\expval{i\hbar \hat{I}}_\psi| = \frac{1}{2} |i\hbar \braket{\psi}{\hat{I}|\psi}| = \frac{1}{2} |i\hbar \cdot 1| = \frac{\hbar}{2} \label{eq:xp_uncertainty_relation}\] This famous result implies a fundamental trade-off: the more precisely the position of a particle is determined (\(\Delta X \to 0\)), the less precisely its momentum can be known (\(\Delta P_x \to \infty\)), and vice versa.
It is crucial to understand that this uncertainty is not a limitation of our measurement apparatus or techniques; it is an intrinsic property of quantum systems stemming from the non-commutativity of observables. Even with perfect instruments, this fundamental limit cannot be overcome. For macroscopic objects, the value of \(\hbar \approx 1.054 \times 10^{-34}\) J\(\cdot\)s is so small that the uncertainty product is negligible compared to the scales involved, explaining why we don’t observe this effect in everyday life. However, at the atomic and subatomic scales, \(\hbar\) is significant, and the uncertainty principle plays a critical role.
A similar relation, \(\Delta E \, \Delta t \ge \hbar/2\), is often quoted. However, its interpretation is more subtle because time \(t\) is treated as a parameter, not an operator, in standard non-relativistic quantum mechanics. It typically relates the uncertainty in the energy of a state (\(\Delta E\)) to the characteristic time scale (\(\Delta t\)) over which the state changes significantly, or the lifetime of an unstable state.
The Quantum Harmonic Oscillator
The harmonic oscillator is a ubiquitous model system in physics, describing phenomena ranging from molecular vibrations and lattice vibrations in solids (phonons) to modes of the electromagnetic field (photons). Its importance stems from the fact that any system near a point of stable equilibrium can often be approximated as a harmonic oscillator for small displacements.
Hamiltonian
The classical Hamiltonian for a one-dimensional harmonic oscillator (mass \(m\), angular frequency \(\omega\)) is \(H_{cl} = \frac{p_x^2}{2m} + \frac{1}{2} m \omega^2 x^2\). Applying canonical quantization yields the quantum Hamiltonian operator: \[\hat{H}_{QHO} = \frac{\hat{P}_x^2}{2m} + \frac{1}{2} m \omega^2 \hat{X}^2 = -\frac{\hbar^2}{2m} \dv[2]{}{x} + \frac{1}{2} m \omega^2 x^2 \label{eq:qho_H}\]
Quantized Energy Levels
Solving the time-independent Schrödinger equation \(\hat{H}_{QHO} \psi_n(x) = E_n \psi_n(x)\) reveals that the allowed energy levels are discrete and uniformly spaced: \[E_n = \hbar \omega \left( n + \frac{1}{2} \right), \quad n = 0, 1, 2, \dots \label{eq:qho_En}\] Key features of the QHO energy spectrum:
Quantization: Only discrete energy values are allowed.
Zero-Point Energy: The lowest possible energy (ground state, \(n=0\)) is \(E_0 = \frac{1}{2}\hbar\omega\), not zero. This reflects the uncertainty principle: the particle cannot be perfectly still (\(\Delta P_x = 0\)) at the exact minimum of the potential (\(\Delta X = 0\)).
Equal Spacing: The energy difference between adjacent levels is constant: \(\Delta E = E_{n+1} - E_n = \hbar \omega\).
The discrete, well-defined energy levels of confined quantum systems like the harmonic oscillator make them suitable candidates for encoding quantum information. A common approach is to use the two lowest energy levels to represent a qubit: \[\ket{0} \equiv \ket{\psi_0} \quad (\text{ground state, energy } E_0) \\ \ket{1} \equiv \ket{\psi_1} \quad (\text{first excited state, energy } E_1 = E_0 + \hbar\omega)\] Transitions between these states can be precisely controlled by applying electromagnetic radiation (e.g., microwaves) with frequency \(\nu = \Delta E / h = \omega / (2\pi)\), allowing for the implementation of quantum gates. Superconducting circuits often utilize engineered potentials that approximate harmonic oscillators (or anharmonic oscillators to isolate the lowest two levels).
Angular Momentum and Spin
Angular momentum is another fundamental quantity that exhibits quantization in quantum mechanics.
Orbital Angular Momentum
Classically, the orbital angular momentum of a particle with position \(\mathbf{r}\) and momentum \(\mathbf{p}\) is \(\mathbf{L} = \mathbf{r} \times \mathbf{p}\). In quantum mechanics, this becomes an operator \(\hat{\mathbf{L}} = \hat{\mathbf{R}} \times \hat{\mathbf{P}}\), where \(\hat{\mathbf{R}} = (\hat{X}, \hat{Y}, \hat{Z})\) and \(\hat{\mathbf{P}} = (\hat{P}_x, \hat{P}_y, \hat{P}_z)\). The components are, for example: \[\hat{L}_z = \hat{X}\hat{P}_y - \hat{Y}\hat{P}_x = -i\hbar \left( x \pdv{}{y} - y \pdv{}{x} \right)\] The components of \(\hat{\mathbf{L}}\) do not commute with each other, satisfying the commutation relations: \[[\hat{L}_x, \hat{L}_y] = i\hbar \hat{L}_z, \quad [\hat{L}_y, \hat{L}_z] = i\hbar \hat{L}_x, \quad [\hat{L}_z, \hat{L}_x] = i\hbar \hat{L}_y \label{eq:L_commutation}\] However, the square of the total angular momentum, \(\hat{L}^2 = \hat{L}_x^2 + \hat{L}_y^2 + \hat{L}_z^2\), commutes with each component: \([\hat{L}^2, \hat{L}_i] = 0\) for \(i=x,y,z\). Therefore, we can find simultaneous eigenstates of \(\hat{L}^2\) and one component, conventionally chosen as \(\hat{L}_z\). The eigenvalues are quantized: \[\begin{aligned} \hat{L}^2 \ket{l, m_l} &= l(l+1)\hbar^2 \ket{l, m_l}, \quad l = 0, 1, 2, \dots \label{eq:L2_eigenvalues} \\ \hat{L}_z \ket{l, m_l} &= m_l \hbar \ket{l, m_l}, \quad m_l = -l, -l+1, \dots, l-1, l \label{eq:Lz_eigenvalues}\end{aligned}\] Here, \(l\) is the orbital angular momentum quantum number, and \(m_l\) is the magnetic quantum number. For a given \(l\), there are \(2l+1\) possible values (orientations) for the projection of the angular momentum onto the z-axis.
Spin: Intrinsic Angular Momentum
Experiments, most notably the Stern-Gerlach experiment (1922), revealed that particles like electrons possess an additional, intrinsic form of angular momentum called spin, denoted by \(\hat{\mathbf{S}}\). Spin is not related to the particle’s motion through space (orbital angular momentum) but is an inherent quantum property, like mass or charge.
Spin operators \(\hat{S}_x, \hat{S}_y, \hat{S}_z\) obey the same commutation relations as orbital angular momentum: \[[\hat{S}_x, \hat{S}_y] = i\hbar \hat{S}_z, \quad (\text{and cyclic permutations}) \label{eq:S_commutation}\]
The square of the total spin, \(\hat{S}^2 = \hat{S}_x^2 + \hat{S}_y^2 + \hat{S}_z^2\), commutes with each component: \([\hat{S}^2, \hat{S}_i] = 0\).
The eigenvalues of \(\hat{S}^2\) and \(\hat{S}_z\) are analogous to orbital angularThe eigenvalues of \(\hat{S}^2\) and \(\hat{S}_z\) are analogous to orbital angular momentum, but the quantum number \(s\) can be a half-integer (0, 1/2, 1, 3/2, ...): \[\begin{aligned} \hat{S}^2 \ket{s, m_s} &= s(s+1)\hbar^2 \ket{s, m_s} \label{eq:S2_eigenvalues} \\ \hat{S}_z \ket{s, m_s} &= m_s \hbar \ket{s, m_s}, \quad m_s = -s, -s+1, \dots, s-1, s \label{eq:Sz_eigenvalues} \end{aligned}\]
Electrons, protons, and neutrons are all spin-1/2 particles (\(s=1/2\)). For these:
Eigenvalue of \(\hat{S}^2\): \(s(s+1)\hbar^2 = \frac{1}{2}(\frac{1}{2}+1)\hbar^2 = \frac{3}{4}\hbar^2\).
Eigenvalues of \(\hat{S}_z\): \(m_s \hbar\), where \(m_s = -\frac{1}{2}, +\frac{1}{2}\).
The two possible states for the \(z\)-component of spin are fundamental for qubits. They are commonly denoted as:
Spin-up: \(\ket{\uparrow}\) or \(\ket{s=1/2, m_s=+1/2}\), eigenvalue \(+\hbar/2\).
Spin-down: \(\ket{\downarrow}\) or \(\ket{s=1/2, m_s=-1/2}\), eigenvalue \(-\hbar/2\).
The state space for the spin of a spin-1/2 particle is a 2-dimensional Hilbert space, \(\mathbb{C}^2\).
The Stern-Gerlach Experiment
This crucial experiment provided direct evidence for spin quantization.
In the experiment, a beam of neutral silver atoms (which have a net spin due to their electron configuration) was passed through a magnetic field with a strong gradient (\(\partial B_z / \partial z \neq 0\)). A particle with magnetic moment \(\vec{\mu}\) experiences a force \(\vec{F} = \nabla(\vec{\mu} \cdot \vec{B})\). For spin, \(\vec{\mu} \propto \hat{\mathbf{S}}\). The force in the z-direction is approximately \(F_z \approx \mu_z (\partial B_z / \partial z)\). Classically, the magnetic moments (and thus spins) could point in any direction, leading to a continuous range of forces and a smeared-out pattern on the detection screen. Instead, Stern and Gerlach observed the beam splitting into two distinct spots, corresponding to the two quantized values of the spin component \(S_z\): \(m_s = +1/2\) and \(m_s = -1/2\). This provided compelling evidence for:
The existence of spin angular momentum.
The quantization of angular momentum components ("space quantization").
Spin Operators and Pauli Matrices
For a spin-1/2 system, the operators \(\hat{S}_x, \hat{S}_y, \hat{S}_z\) acting on the 2D Hilbert space spanned by \(\ket{\uparrow}\) and \(\ket{\downarrow}\) can be represented by matrices proportional to the Pauli matrices: \[\hat{S}_x = \frac{\hbar}{2} \sigma_x, \quad \hat{S}_y = \frac{\hbar}{2} \sigma_y, \quad \hat{S}_z = \frac{\hbar}{2} \sigma_z \label{eq:spin_pauli}\] where the Pauli matrices are: \[\sigma_x = \begin{pmatrix} 0 & 1 \\ 1 & 0 \end{pmatrix}, \quad \sigma_y = \begin{pmatrix} 0 & -i \\ i & 0 \end{pmatrix}, \quad \sigma_z = \begin{pmatrix} 1 & 0 \\ 0 & -1 \end{pmatrix} \label{eq:pauli_matrices}\] In the standard computational basis where \(\ket{0} \equiv \ket{\uparrow} = \begin{pmatrix} 1 \\ 0 \end{pmatrix}\) and \(\ket{1} \equiv \ket{\downarrow} = \begin{pmatrix} 0 \\ 1 \end{pmatrix}\), the \(\sigma_z\) matrix is diagonal with eigenvalues +1 and -1, corresponding to measuring \(S_z\) and obtaining \(\pm \hbar/2\). The Pauli matrices \(\sigma_x, \sigma_y, \sigma_z\) (often denoted \(\hat{X}, \hat{Y}, \hat{Z}\) in quantum computing) are themselves fundamental quantum gates.
Spin-1/2 Systems as Qubits
The two discrete states (\(\ket{\uparrow}, \ket{\downarrow}\)) of a spin-1/2 particle form a natural basis for a qubit: \[\ket{0} \leftrightarrow \ket{\uparrow}, \quad \ket{1} \leftrightarrow \ket{\downarrow} \label{eq:spin_qubit_map}\] A general qubit state \(\ket{\psi} = \alpha \ket{0} + \beta \ket{1}\) corresponds to the spin being in a superposition of spin-up and spin-down states. Physical systems used include electron spins (in quantum dots, nitrogen-vacancy centers in diamond) and nuclear spins (in NMR quantum computing).
Other Qubit Implementations
While spin-1/2 systems are a prime example, other quantum systems with two distinct states can serve as qubits. Another important example mentioned in the transcript is photon polarization. A photon’s polarization can be described in different bases:
Linear polarization: Horizontal (\(\ket{H}\)) and Vertical (\(\ket{V}\)). We can map \(\ket{0} \leftrightarrow \ket{H}\) and \(\ket{1} \leftrightarrow \ket{V}\).
Circular polarization: Right-circular (\(\ket{R}\)) and Left-circular (\(\ket{L}\)).
These different bases are related by superposition (e.g., \(\ket{R} = \frac{1}{\sqrt{2}}(\ket{H} + i\ket{V})\)). Photons are used in linear optics quantum computing and quantum communication. Other platforms include superconducting circuits (transmons, flux qubits), trapped ions, and topological systems.
Qubits in the Quantum Mechanical Framework
Having established the general principles of quantum mechanics, we now focus on how these apply to the fundamental unit of quantum information: the qubit.
The Abstract Qubit State
Mathematically, a qubit is a quantum system described by a 2-dimensional complex Hilbert space, \(\mathcal{H} \cong \mathbb{C}^2\). Its state is represented by a normalized vector in this space. A standard orthonormal basis for this space is the computational basis, denoted \(\{\ket{0}, \ket{1}\}\). A general state of a single qubit \(\ket{\psi}\) can be written as a superposition of these basis states: \[\ket{\psi} = \alpha \ket{0} + \beta \ket{1} \label{eq:qubit_state}\] where \(\alpha, \beta \in \mathbb{C}\) are complex amplitudes satisfying the normalization condition: \[|\alpha|^2 + |\beta|^2 = 1 \label{eq:qubit_normalization}\] According to the Born rule (Postulate [post:born_rule]), if we measure the qubit in the computational basis (e.g., by measuring the observable corresponding to \(\sigma_z\)), the probability of obtaining the outcome associated with \(\ket{0}\) is \(|\alpha|^2\), and the probability of obtaining the outcome associated with \(\ket{1}\) is \(|\beta|^2\).
Physical Realizations and Full State Description
The abstract qubit state \(\ket{\psi} = \alpha \ket{0} + \beta \ket{1}\) must be encoded in a physical system. This requires identifying a physical property of a quantum object that has exactly two distinguishable states, which can be mapped to \(\ket{0}\) and \(\ket{1}\). Examples include:
Spin-1/2 particles (electrons, nuclei): \(\ket{0} \leftrightarrow \ket{\uparrow}\), \(\ket{1} \leftrightarrow \ket{\downarrow}\) (6.3.5).
Photon polarization: \(\ket{0} \leftrightarrow \ket{H}\) (horizontal), \(\ket{1} \leftrightarrow \ket{V}\) (vertical).
Two energy levels of an atom or artificial atom (like a superconducting circuit): \(\ket{0} \leftrightarrow \ket{E_0}\) (ground state), \(\ket{1} \leftrightarrow \ket{E_1}\) (first excited state).
Separating Qubit Dynamics from Other Degrees of Freedom
Physical systems inevitably possess other degrees of freedom besides the one used to encode the qubit. For instance, an electron used as a spin qubit also has a position, momentum, and potentially exists within a complex material environment. The full Hilbert space \(\mathcal{H}_{total}\) of the physical system is typically a tensor product of the qubit’s 2D Hilbert space \(\mathcal{H}_{qubit}\) and the Hilbert space describing all other degrees of freedom \(\mathcal{H}_{other}\): \[\mathcal{H}_{total} = \mathcal{H}_{qubit} \otimes \mathcal{H}_{other} \label{eq:total_hilbert_space}\] A general state in \(\mathcal{H}_{total}\) might be an entangled state between the qubit and other degrees of freedom. However, often we are interested in situations where the state is approximately separable, or we focus only on operations acting within \(\mathcal{H}_{qubit}\). The full state might look like: \[\ket{\Psi_{total}(t)} = \sum_{i=0,1} \ket{i}_{\text{qubit}} \otimes \ket{\phi_i(t)}_{\text{other}} \label{eq:full_state_general}\] In many quantum computing contexts, a crucial simplification is made: we assume that the operations (gates) primarily affect the qubit subspace \(\mathcal{H}_{qubit}\) and that the state of the other degrees of freedom \(\mathcal{H}_{other}\) either remains constant, evolves independently, or its influence can be treated as noise or decoherence. We often implicitly assume a product state where \(\ket{\phi_0(t)} \approx \ket{\phi_1(t)} \approx \ket{\phi_{env}(t)}\), allowing us to write: \[\ket{\Psi_{total}(t)} \approx (\alpha(t) \ket{0} + \beta(t) \ket{1})_{\text{qubit}} \otimes \ket{\phi_{env}(t)}_{\text{other}} \label{eq:full_state_product_approx}\] This allows us to focus solely on the evolution of the qubit state \(\ket{\psi(t)} = \alpha(t) \ket{0} + \beta(t) \ket{1}\). However, it’s vital to remember that this is an idealization. Physical implementation requires careful control over the entire system to minimize unwanted entanglement and decoherence. Sending a qubit physically means transmitting the particle (e.g., photon, trapped ion) that carries the quantum state.
Example: Manipulating Spin Qubits with Magnetic Fields
Let’s revisit the example of a spin-1/2 particle (like an electron or certain nuclei) used as a qubit, where \(\ket{0} \equiv \ket{\uparrow}\) and \(\ket{1} \equiv \ket{\downarrow}\).
Magnetic Moment and Interaction Hamiltonian
Spin-1/2 particles possess an intrinsic magnetic dipole moment \(\vec{\mu}\) proportional to their spin angular momentum \(\hat{\mathbf{S}}\): \[\vec{\mu} = \gamma \hat{\mathbf{S}} = \frac{\gamma \hbar}{2} \vec{\sigma} \label{eq:magnetic_moment}\] where \(\gamma\) is the gyromagnetic ratio (a property of the particle) and \(\vec{\sigma} = (\sigma_x, \sigma_y, \sigma_z)\) is the vector of Pauli matrices. When placed in an external magnetic field \(\vec{B}\), the potential energy of interaction is given by the Hamiltonian: \[\hat{H}_{int} = - \vec{\mu} \cdot \vec{B} = - \frac{\gamma \hbar}{2} \vec{\sigma} \cdot \vec{B} \label{eq:H_int_spin}\] This Hamiltonian governs the evolution of the spin state according to the Schrödinger equation \(i\hbar \dv{t}\ket{\psi} = \hat{H}_{int} \ket{\psi}\).
Larmor Precession
Consider a static magnetic field along the z-axis, \(\vec{B} = B_0 \hat{z}\). The Hamiltonian becomes: \[\hat{H}_0 = - \frac{\gamma \hbar B_0}{2} \sigma_z \label{eq:H0_spin}\] The energy eigenstates are \(\ket{0}\) (spin up, energy \(E_0 = -\gamma \hbar B_0 / 2\)) and \(\ket{1}\) (spin down, energy \(E_1 = +\gamma \hbar B_0 / 2\)). The energy difference is \(\Delta E = E_1 - E_0 = \gamma \hbar B_0\). The time evolution operator is \(\hat{U}_0(t) = e^{-i\hat{H}_0 t/\hbar} = e^{i (\gamma B_0 t / 2) \sigma_z}\). Using \(e^{i\theta \sigma_z} = \cos\theta \hat{I} + i\sin\theta \sigma_z\), this becomes: \[\hat{U}_0(t) = \cos(\omega_0 t / 2) \hat{I} + i \sin(\omega_0 t / 2) \sigma_z = \begin{pmatrix} e^{i\omega_0 t / 2} & 0 \\ 0 & e^{-i\omega_0 t / 2} \end{pmatrix} \label{eq:U0_spin}\] where \(\omega_0 = \gamma B_0\) is the Larmor frequency. If the spin starts in a superposition state like \(\ket{+} = \frac{1}{\sqrt{2}}(\ket{0}+\ket{1})\), its state evolves as: \[\ket{\psi(t)} = \hat{U}_0(t) \ket{+} = \frac{1}{\sqrt{2}} (e^{i\omega_0 t / 2} \ket{0} + e^{-i\omega_0 t / 2} \ket{1})\] This corresponds to the spin vector precessing around the z-axis (the direction of \(B_0\)) at the Larmor frequency \(\omega_0\). This precession forms the basis of magnetic resonance techniques.
Implementing Gates with Oscillating Fields (Magnetic Resonance)
To manipulate the qubit state (i.e., perform rotations away from the z-axis), we need to apply magnetic fields perpendicular to the main field \(B_0 \hat{z}\). Typically, this is done using an oscillating magnetic field in the x-y plane, \(\vec{B}_1(t) = B_1 (\cos(\omega t) \hat{x} - \sin(\omega t) \hat{y})\), rotating at frequency \(\omega\). When the driving frequency \(\omega\) matches the Larmor frequency \(\omega_0\) (resonance condition), the effect of this oscillating field becomes significant, especially when viewed in a reference frame rotating at frequency \(\omega_0\) around the z-axis. In this rotating frame, the effect of \(\hat{H}_0\) is removed, and the Hamiltonian due to the resonant field \(\vec{B}_1\) appears (approximately) as a static field along, say, the rotating x’-axis: \[\hat{H}_{rot} \approx - \frac{\gamma \hbar B_1}{2} \sigma_x' \label{eq:H_rot_spin}\] The evolution under this Hamiltonian causes rotations around the x’-axis in the rotating frame. The time evolution operator in the rotating frame is \(\hat{U}_{rot}(t) = e^{-i\hat{H}_{rot} t/\hbar} = e^{i (\gamma B_1 t / 2) \sigma_x'}\). Using the identity \(e^{i(\theta/2)\sigma_k} = \cos(\theta/2) \hat{I} + i \sin(\theta/2) \sigma_k\), we get: \[\hat{U}_{rot}(t) = \cos\left(\frac{\Omega_1 t}{2}\right) \hat{I} + i \sin\left(\frac{\Omega_1 t}{2}\right) \sigma_x' \label{eq:Urot_spin_x_expanded}\] where \(\Omega_1 = \gamma B_1\) is the Rabi frequency, determining the speed of rotation around the x’-axis. By controlling the duration \(t\) of the applied resonant field (a "pulse"), we can achieve specific rotation angles \(\theta = \Omega_1 t\):
\(\pi\)-pulse (\(t = \pi / \Omega_1\)): Here \(\theta/2 = \pi/2\). \(\hat{U}_{rot} = \cos(\pi/2)\hat{I} + i\sin(\pi/2)\sigma_x' = i \sigma_x'\). This implements the \(\sigma_x\) gate (up to phase \(i\)), flipping \(\ket{0}\) to \(\ket{1}\) and vice versa.
\(\pi/2\)-pulse (\(t = \pi / (2\Omega_1)\)): Here \(\theta/2 = \pi/4\). \(\hat{U}_{rot} = \cos(\pi/4)\hat{I} + i\sin(\pi/4)\sigma_x' = \frac{1}{\sqrt{2}}(\hat{I} + i \sigma_x')\). This creates superpositions. For example, applied to \(\ket{0}\), it yields \(\frac{1}{\sqrt{2}}(\ket{0} + i\ket{1})\).
By applying pulses along different axes (by changing the phase of the oscillating field) and controlling their durations, arbitrary single-qubit rotations (gates) can be implemented. This is the principle behind Nuclear Magnetic Resonance (NMR) and Electron Spin Resonance (ESR) quantum control.
Early demonstrations of quantum algorithms (like Shor’s algorithm for factoring 15) were performed using liquid-state NMR, manipulating the nuclear spins of molecules in solution. While powerful for small numbers of qubits, scaling NMR quantum computing faces significant challenges.
Summary: Qubit as a Controlled Quantum System
The quantum mechanical framework allows us to:
Describe the state of a qubit using a vector in \(\mathbb{C}^2\).
Identify physical systems (like spin-1/2 particles) that realize this 2-level structure.
Understand how to manipulate the qubit state using controlled interactions (like magnetic fields), described by Hamiltonians.
Realize quantum gates as specific unitary time evolutions generated by these Hamiltonians.
Predict the probabilistic outcomes of measurements on the qubit state using the Born rule and projection postulate.
While we often abstract away the underlying physical details and work directly with state vectors and unitary gates, it’s essential to remember the connection to the underlying quantum mechanical principles and physical realizations.
Quantum Measurement
Measurement plays a uniquely central and perplexing role in quantum mechanics. It is the process through which we extract information about a quantum system, but it also fundamentally alters the system’s state in a way distinct from unitary evolution.
The Measurement Problem and State Collapse
As formalized in Postulate [post:measurement_collapse] (4.5.2), the standard description of measurement involves the "collapse of the wave function" (or state vector reduction). When an observable \(\hat{O}\) is measured on a system in a superposition state \(\ket{\psi} = \sum c_i \ket{\lambda_i}\) (where \(\ket{\lambda_i}\) are eigenstates of \(\hat{O}\)), the measurement yields one specific eigenvalue \(\lambda_k\) with probability \(P(\lambda_k) = |c_k|^2\). Immediately after the measurement, the system’s state is no longer the superposition \(\ket{\psi}\) but collapses to the corresponding eigenstate \(\ket{\lambda_k}\) (or the projection onto the eigenspace if degenerate).
This process has several characteristics that distinguish it sharply from the unitary evolution described by the Schrödinger equation:
Non-Unitary: State collapse is not described by a unitary operator. It does not preserve inner products in the same way (e.g., the overlap between the initial state and a state orthogonal to the outcome state becomes zero).
Probabilistic: While unitary evolution is deterministic, the outcome of a measurement on a superposition state is fundamentally random, governed by the Born rule probabilities.
Irreversible: Unitary evolution is reversible (\(\hat{U}^{-1} = \hat{U}^\dagger\)), but measurement collapse is generally irreversible. Once the state collapses to \(\ket{\lambda_k}\), information about the original superposition amplitudes \(c_j\) (especially their relative phases) is lost.
Instantaneous (apparently): The collapse is typically treated as occurring instantaneously upon measurement.
The Measurement Problem
The stark difference between these two modes of state evolution (unitary vs. collapse) leads to the measurement problem. Key questions include:
What constitutes a "measurement"? Does it require a macroscopic apparatus, a conscious observer, or simply an interaction that records information? Where is the dividing line (the "Heisenberg cut") between the quantum system evolving unitarily and the classical apparatus causing collapse?
How does the collapse happen? What physical process underlies this non-unitary, probabilistic transition? Is it a fundamental process, or an effective description arising from a more complex underlying reality?
Consistency with Universal Quantum Mechanics: If quantum mechanics is supposed to be a universal theory applicable to all physical systems, including measurement devices and observers, shouldn’t the interaction between the system and the apparatus also be described by unitary evolution? If so, instead of collapse, the system and apparatus should become entangled, leading to macroscopic superpositions. Why don’t we observe these?
Paradoxes and Interpretations
The conceptual difficulties surrounding measurement have led to various thought experiments and interpretations.
Schrödinger’s Cat
This famous thought experiment, proposed by Erwin Schrödinger in 1935, vividly illustrates the absurdity that seems to arise when applying quantum superposition and unitary evolution to macroscopic objects.
Inside a sealed box, a cat’s life depends on the state of a single radioactive atom. If the atom decays within a certain time (a probabilistic quantum event), it triggers a detector, which releases poison, killing the cat. If the atom does not decay, the cat remains alive. According to a naive application of unitary quantum evolution to the entire system (atom + detector + poison + cat), before an external observer opens the box (performs a measurement), the system should evolve into an entangled state: \[\ket{\Psi_{system}} = \alpha(t) \ket{\text{undecayed}} \otimes \ket{\text{detector off}} \otimes \ket{\text{poison intact}} \otimes \ket{\text{cat alive}} + \beta(t) \ket{\text{decayed}} \otimes \ket{\text{detector on}} \otimes \ket{\text{poison released}} \otimes \ket{\text{cat dead}} \label{eq:cat_state}\] This suggests the cat exists in a superposition of being simultaneously alive and dead until observed. This clashes violently with our everyday experience of macroscopic objects having definite states. The paradox forces us to confront the measurement problem and the quantum-to-classical transition.
Interpretations of Quantum Mechanics
Various interpretations attempt to resolve the measurement problem and provide a coherent picture:
Copenhagen Interpretation (and variants): Often pragmatic, it accepts the collapse postulate as a fundamental rule governing the interaction between quantum systems and classical measurement devices. It emphasizes the role of the observer or measurement apparatus in defining outcomes.
Decoherence Theory: Explains how macroscopic superpositions rapidly decay into effectively classical mixtures due to entanglement with the environment. While it explains the *appearance* of collapse and the emergence of classicality, it doesn’t fully solve the measurement problem (e.g., why one specific outcome is realized).
Many-Worlds Interpretation (MWI): Proposed by Hugh Everett III, MWI denies the collapse postulate. It suggests that the entire universe (system + apparatus + observer + environment) evolves unitarily at all times. Every measurement causes the universe to branch into multiple parallel universes, each corresponding to one possible outcome. In the cat example, the universe splits into one branch where the atom decayed and the cat is dead, and another where the atom didn’t decay and the cat is alive.
Hidden Variable Theories (e.g., Bohmian Mechanics): Propose that quantum mechanics is incomplete and there exist underlying ("hidden") variables that determine the measurement outcomes deterministically. Bell’s theorem and subsequent experiments have placed strong constraints on local hidden variable theories.
Objective Collapse Theories (e.g., GRW, Penrose): Modify the Schrödinger equation to include spontaneous, random collapse processes that become significant for macroscopic systems, thus avoiding macroscopic superpositions without invoking observers.
While these interpretations offer different conceptual frameworks, they generally agree on the predictive rules of quantum mechanics for experimental outcomes, as described by the postulates (including Born rule and effective collapse). For most practical purposes in quantum computing, the standard formalism with the collapse postulate is sufficient.
Generalized Measurements (POVM)
The description of measurement via Hermitian operators and projectors (Postulate [post:measurement_collapse]) is known as projective measurement. While fundamental, it is a special case of a more general framework called Positive Operator-Valued Measures (POVMs). POVMs are particularly useful when considering measurements on subsystems of larger systems or when the measurement process itself might provide incomplete information or involve interaction with an auxiliary system (ancilla).
Measurement Operators
A generalized measurement on a Hilbert space \(\mathcal{H}\) is described by a collection of measurement operators \(\{ \hat{M}_m \}\), acting on \(\mathcal{H}\). The index \(m\) labels the possible outcomes of the measurement. These operators are required to satisfy the completeness relation: \[\sum_m \hat{M}_m^\dagger \hat{M}_m = \hat{I} \label{eq:povm_completeness}\] where \(\hat{I}\) is the identity operator on \(\mathcal{H}\). Note that the operators \(\hat{M}_m\) themselves are not necessarily Hermitian or projectors.
Measurement Probabilities
If the system is in the state \(\ket{\psi}\) immediately before the measurement, the probability \(P(m)\) of obtaining the outcome labeled by \(m\) is given by: \[P(m) = \bra{\psi} \hat{M}_m^\dagger \hat{M}_m \ket{\psi} = \norm{\hat{M}_m \ket{\psi}}^2 \label{eq:povm_probability}\] The completeness relation [eq:povm_completeness] ensures that the probabilities sum to one: \[\sum_m P(m) = \sum_m \bra{\psi} \hat{M}_m^\dagger \hat{M}_m \ket{\psi} = \bra{\psi} \left( \sum_m \hat{M}_m^\dagger \hat{M}_m \right) \ket{\psi} = \bra{\psi} \hat{I} \ket{\psi} = \braket{\psi}{\psi} = 1\] (assuming \(\ket{\psi}\) is normalized). The operators \(\hat{E}_m = \hat{M}_m^\dagger \hat{M}_m\) are called the POVM elements. They are positive semi-definite operators (\(\bra{\phi}\hat{E}_m\ket{\phi} \ge 0\) for all \(\ket{\phi}\)) and sum to the identity (\(\sum_m \hat{E}_m = \hat{I}\)).
Post-Measurement State
If the measurement yields outcome \(m\), the state of the system immediately after the measurement collapses to: \[\ket{\psi'_m} = \frac{\hat{M}_m \ket{\psi}}{\sqrt{\bra{\psi} \hat{M}_m^\dagger \hat{M}_m \ket{\psi}}} = \frac{\hat{M}_m \ket{\psi}}{\sqrt{P(m)}} \label{eq:povm_post_measurement_state}\] This resulting state is correctly normalized.
Projective Measurements as a Special Case of POVM
Projective measurements, as described in the standard postulates, correspond to a specific type of POVM. Let \(\hat{O}\) be an observable with spectral decomposition \(\hat{O} = \sum_i \lambda_i \hat{P}_i\), where \(\lambda_i\) are the distinct eigenvalues and \(\hat{P}_i\) are orthogonal projectors onto the corresponding eigenspaces (\(\hat{P}_i^\dagger = \hat{P}_i\), \(\hat{P}_i^2 = \hat{P}_i\), \(\hat{P}_i \hat{P}_j = \delta_{ij} \hat{P}_i\), and \(\sum_i \hat{P}_i = \hat{I}\)). This measurement can be described within the POVM formalism by choosing the measurement operators to be the projectors themselves: \[\hat{M}_i = \hat{P}_i\] Let’s check the POVM conditions:
Completeness: \(\sum_i \hat{M}_i^\dagger \hat{M}_i = \sum_i \hat{P}_i^\dagger \hat{P}_i = \sum_i \hat{P}_i^2 = \sum_i \hat{P}_i = \hat{I}\). This holds.
Probability: \(P(i) = \bra{\psi} \hat{M}_i^\dagger \hat{M}_i \ket{\psi} = \bra{\psi} \hat{P}_i \ket{\psi}\). This matches the Born rule for projective measurements [eq:born_rule_degenerate].
Post-Measurement State: \(\ket{\psi'_i} = \frac{\hat{M}_i \ket{\psi}}{\sqrt{P(i)}} = \frac{\hat{P}_i \ket{\psi}}{\sqrt{\bra{\psi} \hat{P}_i \ket{\psi}}}\). This matches the collapse rule [eq:collapse_postulate].
Thus, projective measurements are POVMs where the measurement operators are orthogonal projectors. POVMs allow for more general scenarios, including measurements that might not fully distinguish between certain states or measurements that extract partial information.
Measuring a qubit \(\ket{\psi} = \alpha \ket{0} + \beta \ket{1}\) in the computational basis corresponds to the projective measurement with operators \(\hat{M}_0 = \hat{P}_0 = \dyad{0}{0}\) and \(\hat{M}_1 = \hat{P}_1 = \dyad{1}{1}\). The probabilities are: \[\begin{aligned} P(0) &= \bra{\psi} \hat{M}_0^\dagger \hat{M}_0 \ket{\psi} = \bra{\psi} (\dyad{0}{0})^\dagger (\dyad{0}{0}) \ket{\psi} = \bra{\psi} (\dyad{0}{0}) (\dyad{0}{0}) \ket{\psi} \\ &= \bra{\psi} \dyad{0}{0} \ket{\psi} = \braket{\psi}{0}\braket{0}{\psi} = |\braket{0}{\psi}|^2 = |\alpha|^2 \\ P(1) &= \bra{\psi} \hat{M}_1^\dagger \hat{M}_1 \ket{\psi} = \bra{\psi} \dyad{1}{1} \ket{\psi} = |\braket{1}{\psi}|^2 = |\beta|^2 \end{aligned}\] The post-measurement states are: \[\begin{aligned} \text{If outcome 0:} \quad \ket{\psi'_0} &= \frac{\hat{M}_0 \ket{\psi}}{\sqrt{P(0)}} = \frac{\dyad{0}{0} (\alpha\ket{0} + \beta\ket{1})}{\sqrt{|\alpha|^2}} = \frac{\alpha \ket{0}}{|\alpha|} = e^{i\arg(\alpha)} \ket{0} \\ \text{If outcome 1:} \quad \ket{\psi'_1} &= \frac{\hat{M}_1 \ket{\psi}}{\sqrt{P(1)}} = \frac{\dyad{1}{1} (\alpha\ket{0} + \beta\ket{1})}{\sqrt{|\beta|^2}} = \frac{\beta \ket{1}}{|\beta|} = e^{i\arg(\beta)} \ket{1} \end{aligned}\] The post-measurement states are \(\ket{0}\) (if outcome 0) and \(\ket{1}\) (if outcome 1), up to an overall phase factor, which is consistent with the standard collapse rule (as overall phases don’t change the physical state).
In most quantum circuit diagrams, the final measurement step, often depicted with a meter symbol, typically implies a projective measurement in the computational basis \(\{\ket{0}, \ket{1}\}\) for each qubit. The outcomes (0 or 1 for each qubit) are then read out classically.
Conclusion and Outlook
This lecture provided an essential introduction to the principles of quantum mechanics, establishing the theoretical foundation necessary for understanding quantum computation and information processing. We began by examining the historical context, highlighting key experimental phenomena—black body radiation, the photoelectric effect, and atomic spectra—that classical physics could not explain, thereby necessitating the development of quantum theory. Early attempts at quantization, notably Bohr’s model of the hydrogen atom, demonstrated the power of discrete energy levels and angular momentum but were ultimately incomplete.
The development of modern quantum mechanics in the mid-1920s, through the equivalent formalisms of Heisenberg’s matrix mechanics and Schrödinger’s wave mechanics, provided a consistent and comprehensive framework. The core of this framework rests on several key postulates:
States of quantum systems are represented by vectors in a complex Hilbert space, allowing for the crucial concept of superposition.
Physical observables correspond to Hermitian operators, whose real eigenvalues represent the possible outcomes of measurements.
The evolution of an isolated quantum system between measurements is deterministic, continuous, and reversible, governed by the Schrödinger equation and described by unitary operators.
Measurement is a distinct process that is probabilistic (with probabilities given by the Born rule), non-unitary, and results in the collapse of the state vector onto an eigenstate corresponding to the measurement outcome.
We explored fundamental consequences of these postulates, including the Heisenberg Uncertainty Principle, which arises from the non-commutativity of certain observables, and the quantization of energy levels in systems like the quantum harmonic oscillator. The concept of angular momentum, particularly intrinsic spin, was introduced, revealing how spin-1/2 systems naturally provide the two-level structure required for implementing qubits. We saw how the manipulation of these physical qubits, for instance, using magnetic fields to control spin states, directly corresponds to applying unitary transformations.
The direct link between the abstract formalism of quantum mechanics and the practical elements of quantum computing is crucial:
Qubit states (\(\alpha\ket{0} + \beta\ket{1}\)) are specific examples of quantum state vectors.
Quantum gates (\(\hat{X}, \hat{H}, \hat{CNOT}\), etc.) are specific unitary operators \(\hat{U}\) derived from controlled time evolution \(\hat{U} = e^{-i\hat{H}t/\hbar}\).
Reading out the result of a quantum computation involves performing a quantum measurement, typically in the computational basis, yielding probabilistic outcomes according to the Born rule.
While deep conceptual challenges remain, particularly concerning the nature of measurement and the quantum-to-classical transition (the measurement problem), the operational framework provided by the postulates has proven extraordinarily successful in describing the microscopic world and provides the predictive power needed to design and operate quantum computers.
Outlook: Topics for Subsequent Lectures Building on this foundation, future lectures will delve deeper into concepts essential for quantum computation:
Multi-Qubit Systems: Describing systems of multiple qubits using the tensor product formalism.
Entanglement: Exploring this uniquely quantum correlation, its properties, and its role as a resource in quantum algorithms.
Quantum Gates and Circuits: Detailed analysis of common single- and multi-qubit gates and how they are combined to build quantum circuits.
Density Operators: Introducing a formalism to describe quantum states that may be part of a larger system or statistical mixtures (relevant for noise and decoherence).
Decoherence: Understanding how interaction with the environment leads to the loss of quantum coherence and poses a major challenge for building quantum computers.
Quantum Algorithms: Exploring specific algorithms (e.g., Deutsch-Jozsa, Grover, Shor) that demonstrate the potential power of quantum computation.
Physical Implementations: Briefly reviewing various experimental platforms being pursued for building quantum computers (superconducting circuits, trapped ions, photonics, etc.).