Jordan–Wigner, Bravyi–Kitaev and second quantization.
Electrons anticommute; qubits on different sites commute. Every quantum simulation of fermions must bridge that gap, and the bridge you pick sets the Pauli weight of every term, the circuit depth, the qubit count, and which symmetries you can exploit. Jordan–Wigner pays with strings of length O(M), Bravyi–Kitaev with O(logM), and nothing ancilla-free does better than logarithmic.
What you’ll learn
Prove that representing M fermionic modes with Pauli operators requires at least M qubits, via the Majorana (Clifford-algebra) picture.
Prove the canonical anticommutation relations for the Jordan–Wigner transformation and compute the Pauli images of hopping terms.
Construct the parity and Bravyi–Kitaev encodings as linear maps on occupation vectors, and derive their update, parity and flip sets.
Taper qubits using Z2 symmetries with an explicit Clifford transformation, and avoid the wrong-particle-number pitfall.
Anticommuting modes on commuting qubits
M fermionic modes carry operators aj obeying {aj,ak†}=δjk and {aj,ak}=0. It is cleanest to trade them for 2M Hermitian Majorana operators
so that aj=21(γ2j+iγ2j+1). A fermion-to-qubit mapping (without ancillas) is a choice of 2M pairwise anticommuting Pauli strings on n qubits to play the role of the γμ, together with the identification of the vacuum.
Jordan–Wigner, precisely
Jordan–Wigner stores occupations directly, ∣n0⋯nM−1⟩↦∣n0⟩⊗⋯⊗∣nM−1⟩, and sets
Write Z<k=Z<jZjZ(j,k), where Z(j,k) is the string strictly between j and k. The two copies of Z<j square to I, and operators on different sites commute. The products differ only in the order of σj− and Zj.
{aj,ak†}={σj−,Zj}Z(j,k)σk+=0
Site j supplies the relative minus sign. The same computation gives {aj,ak}=0.
{aj,aj†}=Z<j2{σj−,σj+}=I■
On a single mode the strings cancel and the qubit algebra does the rest.
Hopping (p<q) has weight q−p+1; number and Coulomb terms are local. A double excitation plus its conjugate, ap†aq†aras+h.c. with distinct indices, expands into 8 Pauli strings carrying Z strings over the intervening index ranges.
Check your understanding
Under Jordan–Wigner, what is a2†a5+a5†a2?
21(X2X5+Y2Y5)
21(X2X5+Y2Y5)Z3Z4
21(X2X5+Y2Y5)Z0Z1
21(X2X5−Y2Y5)Z3Z4
Show the answer
B.21(X2X5+Y2Y5)Z3Z4
The strings Z<2 are common to both operators and cancel; the modes strictly between, 3 and 4, remain, and σ2+σ5−+σ2−σ5+=21(X2X5+Y2Y5).
Parity and Bravyi–Kitaev encodings
Jordan–Wigner, parity and Bravyi–Kitaev are all linear encodings: the occupation vector n∈F2M is stored as b=An(mod2) for an invertible binary matrix A. To implement aj† you need three pieces of information, each a set of qubits:
the update setU(j): qubits whose stored value contains nj and must flip with it;
the parity setP(j): qubits whose sum is n0+⋯+nj−1, giving the fermionic sign;
the flip setF(j): qubits that, with qubit j, determine nj.
The Pauli weight of aj† is about ∣U(j)∣+∣P(j)∣+1.
The three encodings for M=4. Jordan–Wigner stores occupations, the parity encoding stores prefix parities bj=n0+⋯+nj, and Bravyi–Kitaev stores parities of dyadic blocks (a Fenwick tree).
In the parity encoding the sign is local (P(j)={j−1}) but the update set is everything above j:
aj†=21(Zj−1Xj−iYj)Xj+1⋯XM−1.
Jordan–Wigner has the opposite profile. Bravyi–Kitaev balances them: for M=2m, qubit j stores nk summed over the dyadic block ending at j (for M=8: b0=n0, b1=n0+n1, b2=n2, b3=n0+⋯+n3, b4=n4, b5=n4+n5, b6=n6, b7=n0+⋯+n7).
DerivationWorked example: a₅† in the Bravyi–Kitaev encoding (M = 8)
U(5)={7},P(5)={3,4},F(5)={4}
n5 is contained in b5 and b7; the prefix n0+⋯+n4=b3+b4; and n5=b5⊕b4.
Check: if b4=0 the operator needs b5=0 (so n5=0) and sets b5=1; if b4=1 it needs b5=1 (again n5=0) and sets b5=0. Either way n5 goes 0→1.
sign=(−1)b3⋅(−1)b4=(−1)n0+⋯+n4,X7:b7↦b7⊕1
The Z3 factor and the ± from the b4 branch give exactly the Jordan–Wigner sign, and X7 updates the total parity. Weight 4, versus 6 under Jordan–Wigner; the gap grows to O(logM) versus O(M).
Mapping
Qubits
Worst-case weight of aj
Notes
Jordan–Wigner
M
O(M)
simplest; local for 1D chains
Parity
M
O(M)
two qubits removable for free (spin-sector parities)
Bravyi–Kitaev
M
O(logM)
Fenwick-tree bookkeeping
Ternary tree
M
⌈log3(2M+1)⌉ per Majorana
optimal on average (Jiang et al. 2020)
Verstraete–Cirac, BK superfast, compact
>M
O(1) for lattice hopping
ancillas plus stabilizer constraints
Lower Pauli weight means shallower circuits for each Pauli exponential: e−iθP/2 for a weight-w string costs 2(w−1) CNOTs as a parity ladder.
Symmetries and qubit tapering
Molecular Hamiltonians conserve the electron numbers N↑ and N↓, hence their parities, and often point-group symmetries. After mapping, such a symmetry is a Pauli string τ with [H,τ]=0. Since H=∑jcjPj and τPjτ=±Pj, commuting with H forces every term to commute with τ.
For Z-type symmetries (the usual case for parities), Gaussian elimination over F2 yields generators and pivot qubits q(i) where τi acts as Z and every other generator acts trivially. General Pauli symmetries are first rotated to Z-type by a Clifford.
Ui2=21(2I+{Xq(i),τi})=I
Ui is Hermitian and squares to the identity, so it is unitary (and a Clifford: it maps Paulis to Paulis).
Abbreviating X=Xq(i) and τ=τi, and using XτX=−τ and τ2=I.
[Pj,τi]=0⇒[UPjU†,Xq(i)]=0
Conjugation preserves commutation, so every transformed term acts on qubit q(i) as I or X. In each symmetry sector Xq(i)=±1, and the qubit drops out.
With the parity encoding the Clifford is unnecessary. Order spin-orbitals as all spin-up, then all spin-down. Then bM/2−1=N↑mod2 and bM−1=Nmod2, and no term ever flips them: they appear only as Z's, and two qubits disappear at once. Try it on a two-electron state of H₂ in this ordering (modes: bonding↑, antibonding↑, bonding↓, antibonding↓), prepared as an equal superposition of the Hartree–Fock determinant ∣1010⟩ and the double excitation ∣0101⟩.
Check your understanding
A 12-qubit Hamiltonian commutes with 3 independent, pairwise commuting Pauli strings (none of them ±I, and −I not in the group they generate). What does tapering produce?
One Hamiltonian on 9 qubits
Three Hamiltonians on 9 qubits
Eight Hamiltonians on 9 qubits, one per symmetry sector
Eight Hamiltonians on 6 qubits
Show the answer
C. Eight Hamiltonians on 9 qubits, one per symmetry sector
Each independent Z2 symmetry removes one qubit and splits the problem into two sectors: 12−3=9 qubits and 23=8 sectors. You then pick the sector containing the state you want (the right electron-number parities).
Key takeaways
2M Majoranas must map to pairwise anticommuting Pauli strings; the Clifford-algebra argument forces at least M qubits and makes some nonlocality unavoidable.
Jordan–Wigner is exact and simple, with Z strings of length O(M) (local in 1D, O(N) in 2D snake order).
Linear encodings b=An trade update sets against parity sets: parity has local signs but long updates; Bravyi–Kitaev achieves O(logM), and ternary trees reach the optimal ≈log3(2M).
Z2 symmetries of H can be rotated onto single-qubit X's by a Clifford and removed, one qubit per symmetry; the parity encoding removes two for free.
The qubit Hamiltonian lives on all of Fock space: always fix the particle-number sector you mean.
Check yourself
In the interactive lesson these are scored with your confidence, and what you miss comes back for review.
Question 1
Why can't aj be represented by an operator acting only on qubit j?
Because aj is not Hermitian
Because a single qubit has only two states, while a mode has more
Because for j=k, aj and ak must anticommute, while operators on different qubits commute
It can: aj=σj− works because the signs cancel in the Hamiltonian
Show the answer
C. Because for j=k, aj and ak must anticommute, while operators on different qubits commute
The algebra forces nonlocality. Local operators on disjoint qubits commute, so {aj,ak}=0 with ajak=0 is impossible for them; strings (or ancillas) are needed.
Question 2
For M=1024 modes, roughly what Pauli weight does a1023† (the last mode) have under Jordan–Wigner and under Bravyi–Kitaev?
About 1024 and about 10
About 10 and about 1024
About 1024 under both
1 under both
Show the answer
A. About 1024 and about 10
Jordan–Wigner needs Z0⋯Z1022 plus X or Y on qubit 1023: weight 1024. Bravyi–Kitaev's update, parity and flip sets are each of size at most log21024=10, so the weight is O(10).
Question 3
You diagonalize the 4-qubit Jordan–Wigner Hamiltonian of H₂ exactly. Is its smallest eigenvalue guaranteed to be the energy of the neutral two-electron ground state?
Yes, by the variational principle
No: the qubit Hamiltonian acts on every electron number from 0 to 4, so the lowest eigenvalue may belong to a different N
Yes, because Jordan–Wigner conserves particle number
No, because Jordan–Wigner introduces approximation errors
Show the answer
B. No: the qubit Hamiltonian acts on every electron number from 0 to 4, so the lowest eigenvalue may belong to a different N
Second-quantized Hamiltonians conserve N but act on all sectors. You must restrict to the N=2 sector (and the desired spin), for example by tapering parities plus number-conserving states or penalty terms.
Question 4
In the parity encoding with spin-orbitals ordered all spin-up then all spin-down, why can qubits M/2−1 and M−1 be removed without any Clifford transformation?
They store N↑mod2 and Nmod2, which no term changes, so they appear only as I or Z
They are never used by the Hartree–Fock state
They store occupations of the highest orbitals, which are always empty
Because the parity encoding has O(logM) weight
Show the answer
A. They store N↑mod2 and Nmod2, which no term changes, so they appear only as I or Z
bM/2−1=n0+⋯+nM/2−1=N↑mod2 and bM−1=Nmod2. Number- and spin-conserving terms never flip them, so they act diagonally and can be replaced by their eigenvalues.
Further reading
P. Jordan & E. Wigner, "Über das Paulische Äquivalenzverbot", Z. Phys. 47, 631 (1928).
S. B. Bravyi & A. Yu. Kitaev, "Fermionic quantum computation", Ann. Phys. 298, 210 (2002).
J. T. Seeley, M. J. Richard & P. J. Love, "The Bravyi–Kitaev transformation for quantum computation of electronic structure", J. Chem. Phys. 137, 224109 (2012).
S. Bravyi, J. M. Gambetta, A. Mezzacapo & K. Temme, "Tapering off qubits to simulate fermionic Hamiltonians", arXiv:1701.08213 (2017).
Z. Jiang, A. Kalev, W. Mruczkiewicz & H. Neven, "Optimal fermion-to-qubit mapping via ternary trees with applications to reduced quantum states learning", Quantum 4, 276 (2020).