Below is a series of papers that have been setting the direction of category theory applied to quantum computing. The overview provides an easy reading that can give you (the reader) an idea of why you would want to read any of the chosen publications. The papers were selected in terms of general “impact” (not really clearly defined measure, but how formative the texts were for the field). If you have papers that you would like to see in this list, please do contact me.

Early works by Abramsky, Coecke, and others established the core semantic framework (dagger compact categories) and demonstrated its relevance by reconstructing quantum protocols and no-cloning within that abstract setting. The introduction of classical structures and CPM by 2007–2008 extended the framework to encompass measurements, mixed states, and classical data, enabling a more realistic modeling of quantum algorithms and protocols. Around 2010–2011, the focus shifted to powerful graphical calculi (like ZX) and to connecting the categorical approach with other fields (via expository works and textbook-style treatments). The 2010s saw proofs of completeness (Selinger 2012) and unification of classical/quantum channels (CP* in 2014), showing the framework is mathematically robust and comprehensive. Most recently, achievements like the completeness of the ZX-calculus (2020) illustrate how far the field has come – from abstract theory to practical diagrammatic reasoning that fully captures quantum computing.

A categorical semantics of quantum protocols

This paper introduced the modeling finite-dimensional quantum processes using dagger compact closed categories. It demonstrated that quantum information protocols like teleportation and entanglement swapping can be captured diagrammatically at an abstract categorical level. By recasting the Hilbert space formalism into category-theoretic terms, it laid the foundation for categorical quantum mechanics, showing how key quantum structures (states, processes, tensorial composition) can be defined from purely algebraic axioms. This work is highly influential (hundreds of citations) and is regarded as the starting point for the field, establishing that compact closed categories with biproducts provide a sound semantic framework for quantum computing.

Quantum Measurements Without Sums

This paper develops the categorical semantics of quantum measurement by eliminating the need for direct sum structures. It shows that all aspects of quantum mechanics (including mixed states and measurement outcomes) can be expressed using only the tensor (monoidal) structure, provided one identifies “classical spaces” as those objects that copy and delete data. In categorical terms, the authors axiomatize classical interfaces as special commutative dagger Frobenius algebras within a dagger compact category. This result implies that an orthonormal basis (a classical measurement context) can be captured abstractly as a Frobenius algebra, and that the ability to copy/delete classical information vs. the inability to do so for quantum data underpins the quantum–classical divide. By providing an intuitive graphical calculus for measurements, this work greatly influenced the development of classical structures in categorical quantum mechanics and underlies later quantum protocol semantics.

Dagger Compact Closed Categories and Completely Positive Maps

Selinger extended the categorical framework to mixed states and quantum channels (noisy processes) via the introduction of the CPM construction. This paper presents a graphical language for dagger compact categories and proves it complete for equational reasoning, meaning any valid equation between morphisms can be derived graphically. Crucially, it defines a construction that takes any dagger compact category \(C\) and produces a new category CPM[$C$] whose morphisms represent completely positive maps (quantum channels) on the processes of \(C\). Applying this to the pure-state category (like FdHilb) yields a category of density matrices and CP maps that is again dagger compact. This result showed that the categorical approach is not limited to pure unitary quantum mechanics but also handles probabilistic mixing and decoherence in an elegant, abstract way. By uniting pure and mixed quantum semantics, Selinger’s CPM construction has become a cornerstone of categorical quantum computing, reflected in its widespread use in later work.

Interacting Quantum Observables: Categorical Algebra and Diagrammatics

This highly-cited paper introduced the ZX-calculus, a pivotal graphical language for quantum circuits based on complementary observables. Within the categorical framework of dagger symmetric monoidal categories, the authors axiomatize the notion of complementary quantum observables (such as the $Z$ and $X$ Pauli bases for qubits) as interacting *Frobenius algebras*. Each observable is represented by a special commutative dagger Frobenius algebra (graphically, a red or green node), and mutual unbiasedness (complementarity) is captured by algebraic relations (a form of bialgebra law) between these structures. The paper provides an intuitive graphical calculus (ZX-diagrams) that greatly simplifies reasoning about quantum circuits and protocols, allowing one to derive equalities (e.g. circuit identities, simplifications) by diagrammatic rewriting. The ZX-calculus has proven to be universal (capable of representing any quantum computation) and became an influential tool in quantum computing theory. This work marks a key evolution in the field, demonstrating the power of category-based diagrams to handle complex quantum-algebraic reasoning in a purely visual manner.

A New Description of Orthogonal Bases

This paper solidifies the link between abstract categorical structures and concrete quantum mechanics by showing that *orthogonal bases in Hilbert spaces are equivalently characterized as commutative dagger-Frobenius monoids. An orthonormal basis (a classical set of states that can be copied/cloned) is exactly captured by a special commutative dagger Frobenius algebra in the category FdHilb. The authors prove that the extra condition corresponds to the basis being normalized (orthonormal). This result gives an operational meaning to the algebraic structures: the Frobenius comultiplication “$\delta$” duplicates basis vectors and the counit “$\epsilon$” deletes them, reflecting the fact that classical data can be cloned and discarded while quantum data cannot. By characterizing bases intrinsically (without referring to coordinates or inner products) and showing how classical information emerges as observable structures in a category, this paper provided a rigorous foundation for classical structures in categorical quantum mechanics. It has important implications, for instance in understanding quantum measurement and classical control in purely compositional terms, and it underpins many later developments (including the ZX-calculus and categorical quantum logic).

Finite Dimensional Hilbert Spaces Are Complete for Dagger Compact Closed Categories

This theoretical result addresses the completeness of the categorical axiomatics with respect to standard quantum mechanics. Selinger proved that any equation between morphisms that holds in all finite-dimensional Hilbert spaces (the usual semantics of quantum computing) can already be derived from the abstract axioms of a dagger compact closed category. In other words, the category FdHilb (finite-dimensional Hilbert spaces with linear maps) is a complete model of the dagger compact closed category axioms – there are no “extra” equations in Hilbert space beyond those provable from the categorical framework. This paper ensures that the diagrammatic reasoning developed in categorical quantum mechanics is complete: if two quantum processes are equal (as linear maps), their equality can be proven using only the categorical axioms and graph transformations. Completeness was a non-trivial property to establish and gives strong validation that the chosen axioms fully capture finite-dimensional quantum theory. This work thus solidifies the foundational footing of the entire approach, confirming that one does not lose anything by working abstractly – the categorical semantics is equivalent to the usual Hilbert space semantics for all practical purposes.

Categories of Quantum and Classical Channels

This paper introduced the CP*-construction, a significant generalization of Selinger’s CPM construction to incorporate both quantum and classical information in one category. The authors build a unified categorical framework where objects can represent hybrid classical-quantum systems (interpreted as abstract C*-algebras), and morphisms represent completely positive maps that may carry classical data as well. The CP* construction fully embeds the pure quantum CPM category as a special case while also embedding the category of classical stochastic processes, thereby allowing classical and quantum channels to coexist and interact in a single symmetric monoidal category. This yields elegant abstract notions of state preparation, measurement, and conditional dynamics within one compositional model. In essence, CP* provides a categorical semantics for quantum systems with classical outcomes or controls – a “von Neumann style” category of quantum observables and channels as opposed to purely unitary evolutions. This work has been influential in areas like quantum foundations and quantum protocols, as it places classical-quantum interaction on equal footing and has inspired further developments in categorical probability and operational theories.

Physics, Topology, Logic and Computation: A Rosetta Stone In New Structures for Physics (eds. B. Coecke), Lecture Notes in Physics 813, pp. 95–172. [ArXiv:0903.0340]. This widely cited exposition connected the categorical quantum computing program with broader areas, helping to spread these ideas. Baez and Stay survey how closed symmetric monoidal categories and their diagrammatic calculus provide a common language for quantum physics, logic, and computation. In particular, they explain Abramsky and Coecke’s categorical quantum mechanics in a tutorial style, drawing an analogy between Feynman diagrams, cobordisms in topology, and proof nets in logic. The paper does not introduce new theorems specific to quantum computing, but it synthesizes many concepts (traced monoidal categories, dagger compact categories, etc.) in an accessible way. By acting as a “Rosetta Stone,” it helped researchers in various fields understand the categorical approach to quantum theory and recognize its significance. This work thus contributed to the evolution of the field by broadening its audience and establishing a clear conceptual bridge between category theory and quantum computation, influencing education and further interdisciplinary research.

Completeness of the ZX-Calculus

This recent paper resolves a long-standing open question by proving the ZX-calculus is complete for all pure qubit quantum mechanics. The ZX-calculus (introduced by Coecke & Duncan in 2011) was known to be universal but not fully complete (certain true quantum equations had no diagrammatic proof using the original rules). Jeandel et al. provided the first complete axiomatisation: they show that any equality of linear maps (matrices) on qubits can be derived using ZX-diagram rewrite rules. The proof proceeds by first enriching the ZX-calculus with additional rewrite rules to make it complete for the Clifford+T fragment (a universal gate set), and then leveraging a translation to a related graphical language (the ZW-calculus) to reach completeness for arbitrary quantum operations. The result implies that the diagrammatic calculus now fully matches the expressive power of matrix mathematics – if two quantum circuits are mathematically equivalent, the ZX-calculus can prove it. This accomplishment marks a maturation of the categorical approach: the diagrammatic methods are not only convenient and intuitive but also no less powerful than algebraic methods. It has generated further research and practical tools in quantum circuit optimization, now with graphical proof technique for quantum computing theory.