betaFeedback

WorksMathematics

An isolated complex Hadamard matrix with nonzero defect

Abstract

We exhibit a complex Hadamard matrix of nonzero defect which is isolated among complex Hadamard matrices, up to Hadamard equivalence. This counterexample is realized by a Butson-type Hadamard matrix of size 1313, with entries roots of order 66 of unity. In the language of commuting squares, this shows that the span condition, previously introduced by the second author as a sufficient condition for isolation, is not a necessary condition for isolation.

Introduction

A complex Hadamard matrix is a matrix H∈Mn(C)H \in M_n(\mathbb{C}) whose entries have absolute value 1 and whose rows are mutually orthogonal. Equivalently, u=n−1/2Hu = n^{-1/2}H is unitary and all its entries have absolute value n−1/2n^{-1/2}. If P=DnP = D_n is the algebra of diagonal matrices and Q=uDnu∗Q = uD_nu^*, the Hadamard condition is equivalent to the commuting square condition for

C(H)=(P⊂Mn(C)∪∪CI⊂Q),τ=1nTr⁡.(1)\mathcal{C}(H)= \left( \begin{array}{ccc} P & \subset& M_n(\mathbb{C})\\ \cup& & \cup\\ \mathbb{C}I & \subset& Q \end{array} \right),\qquad\tau=\frac{1}{n}\operatorname{Tr}. \tag*{(1)}

Here the commuting square condition means that τ(pq)=τ(p)τ(q)\tau(pq)=\tau(p)\tau(q) for p∈Pp \in P, q∈Qq \in Q.

The span condition was introduced in [8] as a sufficient condition for a commuting square to be isolated, modulo isomorphisms. In the case (1), it asks that

dim⁡C[P,Q]=(n−1)2, where [P,Q]=span⁡C{pq−qp:p∈P, q∈Q}.(2)\dim_{\mathbb{C}}[P,Q]=(n-1)^2,\text{ where }[P,Q]=\operatorname{span}_{\mathbb{C}}\{pq-qp:p\in P,\ q\in Q\}. \tag*{(2)}

The defect of a complex Hadamard matrix, introduced and studied by Tadej and Życzkowski [10, 11], measures the dimension of the dephased first order deformation space. These two notions are related by

d(H)=(n−1)2−dim⁡C[P,Q].(3)d(H)=(n-1)^2-\dim_{\mathbb{C}}[P,Q]. \tag*{(3)}

Thus defect zero implies isolation. The converse asks whether every isolated Hadamard matrix has defect zero.

This converse was formulated explicitly by Banica, Özteke and Pittau [1 Conjecture 1.8]: an isolated partial Hadamard matrix should have minimal undephased defect. For a square matrix of order nn, their undephased defect is d(H)+2n−1d(H)+2n-1, and its minimal value is 2n−12n-1. Thus their conjecture asserts that isolation implies d(H)=0d(H)=0. It was still recorded as open in the 2026 review of McNulty and Weigert [5 Appendix B, following Lemma B.1], where they write: “It is presently unknown if isolation is equivalent to the span condition”.

There is also an earlier claim which must be distinguished from this open question. In his 2014 dissertation, Motta [6 Section 2.4.3, Proposition 16, p. 118] asserted that the order 9 matrix B9(0)B_{9}^{(0)}, whose defect is 2, is isolated. The section is entitled “An Isolated Matrix with Positive Defect.” Karlsson subsequently placed B9(0)B_{9}^{(0)} in a non-affine two-parameter family [2 Section 10]. The later paper of Motta and Shipman [MS, Section 4] explicitly records that “the defect arises from a non-affine 2-dimensional family” containing this matrix. Thus the order 9 example does not disprove the converse.

In this paper we give an example which does. In the following we will use the notation BH(n,q)BH(n,q) for the Hadamard matrices of order nn whose entries are qq-th roots of unity.

Theorem 1.1. There exists a complex Hadamard matrix H∈BH(13,6)H \in BH(13,6) which is isolated up to equivalence and satisfies d(H)=1d(H)=1. For u=13−1/2Hu=13^{-1/2}H, its spin model commuting square is isolated up to commuting square isomorphism, and

dim⁡C[D13,uD13u∗]=143<144.\dim_{\mathbb{C}}[D_{13},uD_{13}u^{*}]=143<144.

The matrix is given explicitly in Section 4. It is already present in the dataset of Lampio, Östergård and Szöllősi [4], accompanying their classification paper [3]. Our result concerns the local isolation of that matrix in the moduli space of non-equivalent complex Hadamard matrices, thus providing a counterexample to [1 Conjecture 1.8].

Our proof follows the method of sequential deformations used in [8, 9]. Assuming that HH is not isolated, we take a sequence of inequivalent Hadamard matrices converging to HH, and then a limit of the normalized differences of the corresponding unitaries. A relation between commutators gives a second order restriction on this direction of convergence. This restriction is the one in [9 Corollary 3.3(ii)] for a general commuting square. We include in this paper a direct proof for commuting squares based on Hadamard matrices. For the matrix in Theorem 1.1, the first order space is one-dimensional and its generator violates the restriction. This gives sequential isolation, without any differentiability assumptions.

AI disclosure. The AI model Astra was used to search for possible candidates for a counterexample among known databases of complex Hadamard matrices, and for verifying finite computations given by exact integer and cyclotomic identities.

Preliminaries

We denote by Tr⁡\operatorname{Tr} the trace on Mn(C)M_n(\mathbb{C}), and by τ=n−1Tr⁡\tau=n^{-1}\operatorname{Tr} the normalized trace. We use the notation ∥x∥2=τ(x∗x)1/2\lVert x\rVert_2=\tau(x^{*}x)^{1/2} for the norm 2 on matrices. For matrices S,HS,H of the same size, S∘HS\circ H denotes their entrywise product (also called Hadamard product).

Two Hadamard matrices are equivalent if one can be obtained from the other by permutations of rows and columns and multiplication of rows and columns by scalars of absolute value 1. A matrix is dephased if its first row and first column consist of 1’s. Every Hadamard matrix is equivalent to a dephased one. More precisely, the map

D(K)ij=KijKi1‾K1j‾K11(4)\mathcal{D}(K)_{ij}=K_{ij}\overline{K_{i1}}\overline{K_{1j}}K_{11} \tag*{(4)}

is continuous on matrices with unimodular entries, takes values in the dephased matrices, and fixes every dephased matrix.

A Hadamard matrix is isolated up to equivalence if some neighborhood contains no inequivalent Hadamard matrix. For a dephased matrix, this is equivalent to being isolated among dephased Hadamard matrices: a fixed equivalence class has only finitely many dephased representatives, obtained by dephasing its row and column permutations.

Fix a dephased Hadamard matrix HH of order nn. Put

TH={R∈Mn(R):R1j=Ri1=0,∑j=1nHijHkj‾(Rij−Rkj)=0(i≠k)}.(5)\mathcal{T}_H=\left\{R\in M_n(\mathbb{R}):\begin{array}{l} R_{1j}=R_{i1}=0,\\ \sum_{j=1}^{n}H_{ij}\overline{H_{kj}}(R_{ij}-R_{kj})=0\quad(i\ne k) \end{array}\right\}. \tag*{(5)}

The dephased defect is d(H)=dim⁡RTHd(H)=\dim_{\mathbb{R}}\mathcal{T}_H. This is the dimension of the solution space of the linearized Hadamard equations; we do not assume that a vector in this space is the derivative of a deformation. For any real matrix RR, write

hR=1n(R∘H)H∗.(6)h_R=\frac{1}{n}(R\circ H)H^*. \tag*{(6)}

Then hRH=R∘Hh_RH=R\circ H, and the equations in the second line of (5) are equivalent to hR=hR∗h_R=h_R^*.

Let u=n−1/2Hu=n^{-1/2}H, P=DnP=D_n, Q=uDnu∗Q=uD_nu^*, and pi=eiip_i=e_{ii}. For R∈Mn(R)R\in M_n(\mathbb{R}) define

qi(R)=udiag⁡(Ri1,…,Rin)u∗∈Qsa.(7)q_i(R)=u\operatorname{diag}(R_{i1},\ldots,R_{in})u^*\in Q_{\mathrm{sa}}. \tag*{(7)}

A direct computation gives

∑ipiqi(R)=hR,∑i[pi,qi(R)]=hR−hR∗.(8)\sum_i p_iq_i(R)=h_R,\qquad\sum_i[p_i,q_i(R)]=h_R-h_R^*. \tag*{(8)}

Thus every R∈THR\in\mathcal{T}_H gives a relation between commutators.

For completeness, we show that these formulas also prove (3). The real linear map R↦∑i[pi,qi(R)]R\mapsto\sum_i[p_i,q_i(R)] has image span⁡R[Psa,Qsa]\operatorname{span}_{\mathbb{R}}[P_{\mathrm{sa}},Q_{\mathrm{sa}}], whose real dimension is dim⁡C[P,Q]\dim_{\mathbb{C}}[P,Q]: this real image consists of skew-adjoint matrices, and its complexification is [P,Q][P,Q]. Its kernel is the undephased solution space of the first order equations. The matrices

Rij=ai+bj,ai,bj∈R,R_{ij}=a_i+b_j,\qquad a_i,b_j\in\mathbb{R},

form a (2n−1)(2n-1)-dimensional subspace of this kernel. Subtracting Ri1+R1j−R11R_{i1}+R_{1j}-R_{11} from RijR_{ij} gives its unique dephased representative. Hence the kernel is the direct sum of this subspace and TH\mathcal{T}_H, and rank-nullity gives (3).

Since PP and QQ are abelian, [P,Q][P,Q] is orthogonal to P+QP+Q with respect to the trace inner product. Also P∩Q=CIP\cap Q=\mathbb{C}I, by the commuting square condition, so dim⁡(P+Q)=2n−1\dim(P+Q)=2n-1. Therefore (2) is equivalent to [P,Q]+P+Q=Mn(C)[P,Q]+P+Q=M_n(\mathbb{C}), the span condition for (1).

Directions of convergence and a second order condition

We first recall the restriction on a direction of convergence that we will use. It is the scalar-corner case of [9 Corollary 3.3(ii)]. For completeness, we reprove that result here in the case of spin model (i.e., Hadamard-based) commuting squares.

Lemma 3.1. Let P,Q⊂Mn(C)P,Q\subset M_n(\mathbb{C}) be unital ∗*-subalgebras satisfying τ(pq)=τ(p)τ(q)\tau(pq)=\tau(p)\tau(q) for p∈Pp\in P, q∈Qq\in Q. Suppose Um≠IU_m\ne I are unitaries converging to II such that

τ(pUmqUm∗)=τ(p)τ(q),p∈P, q∈Q,\tau(pU_mqU_m^*)=\tau(p)\tau(q),\qquad p\in P,\ q\in Q,

and, with tm=∥Um−I∥2t_m=\|U_m-I\|_2,

Um−Iitm⟶h.(9)\frac{U_m-I}{it_m}\longrightarrow h. \tag*{(9)}

Then h=h∗h=h^*, ∥h∥2=1\|h\|_2=1. Moreover, every relation

∑i=1r[pi,qi]=0,pi∈Psa,qi∈Qsa,\sum_{i=1}^{r}[p_i,q_i]=0,\qquad p_i\in P_{\mathrm{sa}},\quad q_i\in Q_{\mathrm{sa}},

satisfies

∑i=1rτ(piqih2)=∑i=1rτ(pihqih).(10)\sum_{i=1}^{r}\tau(p_iq_ih^2)=\sum_{i=1}^{r}\tau(p_ihq_ih). \tag*{(10)}

Proof. Put Vm=Um−IV_m=U_m-I. Unitarity gives

Vm+Vm∗=−Vm∗Vm.(11)V_m+V_m^*=-V_m^*V_m. \tag*{(11)}

Dividing by tmt_m and using (9), we obtain h=h∗h=h^*. The equality ∥h∥2=1\lVert h\rVert_2=1 follows from the definition of tmt_m.

Let A=∑ipiqiA=\sum_i p_iq_i. The commutator relation gives A=∑iqipiA=\sum_iq_ip_i, and hence A=A∗A=A^*. By the two commuting square relations and the trace property,

0=∑iτ(piUmqiUm∗−piqi)=τ(A(Vm+Vm∗))+∑iτ(piVmqiVm∗)=−τ(AVm∗Vm)+∑iτ(piVmqiVm∗).\begin{aligned} 0=\sum_i\tau(p_iU_mq_iU_m^*-p_iq_i) \\ &=\tau\left(A(V_m+V_m^*)\right)+\sum_i\tau(p_iV_mq_iV_m^*) \\ &=-\tau(AV_m^*V_m)+\sum_i\tau(p_iV_mq_iV_m^*). \end{aligned}

Divide this by tm2t_m^2 and let mm tend to infinity. Since Vm/(itm)→h=h∗V_m/(it_m)\to h=h^*, the resulting equality is (10).

We now give a new interpretation to the restriction from the previous lemma. For R∈THR\in\mathcal{T}_H, define

OH(R)=τ(hR3)−1n2∑i,jRij3.(12)\mathcal{O}_H(R)=\tau(h_R^3)-\frac{1}{n^2}\sum_{i,j}R_{ij}^3. \tag*{(12)}

This is a real homogeneous cubic on TH\mathcal{T}_H.

Proposition 3.2. If a dephased Hadamard matrix HH is not isolated up to equivalence, there exists 0≠R∈TH0\ne R\in\mathcal{T}_H such that OH(R)=0\mathcal{O}_H(R)=0.

Proof. Choose Hadamard matrices Hm→HH_m\to H which are not equivalent to HH. Applying (4), we may assume that each HmH_m is dephased. Set

Um=1nHmH∗,tm=∥Um−I∥2.U_m=\frac{1}{n}H_mH^*,\qquad t_m=\lVert U_m-I\rVert_2.

Then UmU_m is unitary, Um→IU_m\to I, and tm>0t_m>0. The sequence (Um−I)/(itm)(U_m-I)/(it_m) has norm 1. By finite dimensionality, after passing to a subsequence we may assume that it converges to a matrix hh as in Lemma 3.1.

Write hm=(Um−I)/(itm)h_m=(U_m-I)/(it_m). Since Hm=UmHH_m=U_mH, we have

Hm=H+itmhmH.(13)H_m=H+it_mh_mH. \tag*{(13)}

The equality ∣(Hm)ij∣=∣Hij∣=1\lvert(H_m)_{ij}\rvert=\lvert H_{ij}\rvert=1 gives

2Re⁡(i(hmH)ijH‾ij)+tm∣(hmH)ij∣2=0.2\operatorname{Re}\left(i(h_mH)_{ij}\overline{H}_{ij}\right)+t_m\lvert(h_mH)_{ij}\rvert^2=0.

Taking limits shows that Rij=(hH)ijH‾ijR_{ij}=(hH)_{ij}\overline{H}_{ij} is real. The first row and column of Hm−HH_m-H vanish, so (13) also gives R1j=Ri1=0R_{1j}=R_{i1}=0. Thus hH=R∘HhH=R\circ H and h=hRh=h_R. Since h=h∗h=h^*, we have R∈THR\in\mathcal{T}_H; since ∥h∥2=1\lVert h\rVert_2=1, we have R≠0R\ne0.

Take pi=eiip_i=e_{ii} and qi=qi(R)q_i=q_i(R) as in (7). By (8), ∑i[pi,qi]=0\sum_i[p_i,q_i]=0 and ∑ipiqi=h\sum_i p_iq_i=h. The Hadamard condition on HmH_m says precisely that PP and UmQUm∗U_mQU_m^* satisfy the hypothesis of Lemma 3.1. Consequently,

τ(h3)=∑iτ(pihqih).(14)\tau(h^3)=\sum_i\tau(p_i hq_i h). \tag*{(14)}

Now hu=R∘uhu=R\circ u and h=h∗h=h^*, so

Tr⁡(pihqih)=((hu)diag⁡(Ri1,…,Rin)(hu)∗)ii=1n∑jRij3.\operatorname{Tr}(p_i hq_i h)=\left((hu)\operatorname{diag}(R_{i1},\ldots,R_{in})(hu)^*\right)_{ii} =\frac{1}{n}\sum_jR_{ij}^3.

After dividing by nn and summing over ii, (14) becomes OH(R)=0\mathcal{O}_H(R)=0.

:::

Corollary 3.3. Suppose TH=RR\mathcal{T}_H=\mathbb{R}R with R≠0R\ne0. If OH(R)≠0\mathcal{O}_H(R)\ne0, then HH is isolated up to equivalence.

Proof. Otherwise Proposition 3.2 gives a nonzero S∈THS\in\mathcal{T}_H with OH(S)=0\mathcal{O}_H(S)=0. Write S=cRS=cR, where c∈R∖{0}c\in\mathbb{R}\setminus\{0\}. Then 0=OH(S)=c3OH(R)0=\mathcal{O}_H(S)=c^3\mathcal{O}_H(R), a contradiction. □\square

Remark 3.4. The argument above excludes arbitrary convergent sequences of inequivalent Hadamard matrices. In particular, it excludes differential deformations whose first derivative vanishes at the initial matrix.

An isolated matrix of order thirteen

Let w=exp⁡(πi/3)w=\exp(\pi i/3) and define Hij=wEijH_{ij}=w^{E_{ij}}, where

E=(0000000000000000002223344400013445142230024413430152003320255242502303503125530253542250023025410533431103135315033510340242030240041433513201504315243104200432115154143).(15)E=\begin{pmatrix} 0&0&0&0&0&0&0&0&0&0&0&0&0\\ 0&0&0&0&0&2&2&2&3&3&4&4&4\\ 0&0&0&1&3&4&4&5&1&4&2&2&3\\ 0&0&2&4&4&1&3&4&3&0&1&5&2\\ 0&0&3&3&2&0&2&5&5&2&4&2&5\\ 0&2&3&0&3&5&0&3&1&2&5&5&3\\ 0&2&5&3&5&4&2&2&5&0&0&2&3\\ 0&2&5&4&1&0&5&3&3&4&3&1&1\\ 0&3&1&3&5&3&1&5&0&3&3&5&1\\ 0&3&4&0&2&4&2&0&3&0&2&4&0\\ 0&4&1&4&3&3&5&1&3&2&0&1&5\\ 0&4&3&1&5&2&4&3&1&0&4&2&0\\ 0&4&3&2&1&1&5&1&5&4&1&4&3 \end{pmatrix}. \tag*{(15)}

This is the matrix on line 394 of matrices/BH-13-6.txt in [4]. Our argument uses this matrix, but does not depend on the completeness of the classification in [3].

Consider the real matrix

R=(00000000000000−2−1−1−12−4−1−422−4−10−1−21−21−2−2−4−11−2−20−11411−21−4−11−210−1−21−21−2−2−4−11−2−20−1−2−2−2−2−2−2−4−1−2−2−20−1−21−21−2−2−4−11−2−20−1−2−2−21−5−2−4−11−5−20−4−4−4−4−4−4−4−8−4−4−4−402−1−1−12−4−1−4−22−4−10−1−2−2−21−5−2−4−11−5−20−1−21−21−2−2−4−11−2−20−1−2−2−2−2−2−2−4−1−2−2−2).(16)R=\begin{pmatrix} 0&0&0&0&0&0&0&0&0&0&0&0&0\\ 0&-2&-1&-1&-1&2&-4&-1&-4&2&2&-4&-1\\ 0&-1&-2&1&-2&1&-2&-2&-4&-1&1&-2&-2\\ 0&-1&1&4&1&1&-2&1&-4&-1&1&-2&1\\ 0&-1&-2&1&-2&1&-2&-2&-4&-1&1&-2&-2\\ 0&-1&-2&-2&-2&-2&-2&-2&-4&-1&-2&-2&-2\\ 0&-1&-2&1&-2&1&-2&-2&-4&-1&1&-2&-2\\ 0&-1&-2&-2&-2&1&-5&-2&-4&-1&1&-5&-2\\ 0&-4&-4&-4&-4&-4&-4&-4&-8&-4&-4&-4&-4\\ 0&2&-1&-1&-1&2&-4&-1&-4&-2&2&-4&-1\\ 0&-1&-2&-2&-2&1&-5&-2&-4&-1&1&-5&-2\\ 0&-1&-2&1&-2&1&-2&-2&-4&-1&1&-2&-2\\ 0&-1&-2&-2&-2&-2&-2&-2&-4&-1&-2&-2&-2 \end{pmatrix}. \tag*{(16)}

Explicit calculations

Proposition 4.1. The matrices (15) and (16) satisfy

HH∗=13I,TH=RR(17)HH^*=13I,\qquad\mathcal{T}_H=\mathbb{R}R \tag*{(17)}
∑i,jRij3=−3018,Tr⁡(((R∘H)H∗)3)=−538434.(18)\sum_{i,j}R_{ij}^{3}=-3018,\qquad\operatorname{Tr}\left(((R\circ H)H^*)^3\right)=-538434. \tag*{(18)}

In particular,

τ(hR3)=−3186169,OH(R)=−168169≠0.(19)\tau(h_R^3)=-\frac{3186}{169},\qquad\mathcal{O}_H(R)=-\frac{168}{169}\ne0. \tag*{(19)}

Proof. All computations below use exact arithmetic. In Z[w]\mathbb{Z}[w] we use w2=w−1w^2=w-1 and w‾=1−w\overline{w}=1-w. The six powers of ww are

1, w, w−1, −1, −w, 1−w.1,\ w,\ w-1,\ -1,\ -w,\ 1-w.

Using these expressions in the row inner products gives HH∗=13IHH^*=13I.

To determine the defect, set

a=(2,1,−1,−2,−1,1),b=(0,1,1,0,−1,−1).a=(2,1,-1,-2,-1,1),\qquad b=(0,1,1,0,-1,-1).

For e=0,…,5e=0,\ldots,5, these are the values of 2Re⁡(we)2\operatorname{Re}(w^e) and 2Im⁡(we)/32\operatorname{Im}(w^e)/\sqrt{3}, respectively. Let LL be the 156×144156\times144 integer matrix obtained from the equations in (5) as follows. Its columns are indexed lexicographically by (r,j)(r,j), 2≤r,j≤132\le r,j\le13. Order the pairs (i,k)(i,k) by

(2,1),(3,1),(3,2),…,(13,12).(2,1),(3,1),(3,2),\ldots,(13,12).

For each pair include first the aa row and then the bb row, where

L(i,k),(r,j)a=a(Eij−Ekj) mod 6(δir−δkr),L(i,k),(r,j)b=b(Eij−Ekj) mod 6(δir−δkr).(20)\begin{aligned} L^{a}_{(i,k),(r,j)}&=a_{(E_{ij}-E_{kj})\bmod6}(\delta_{ir}-\delta_{kr}),\\ L^{b}_{(i,k),(r,j)}&=b_{(E_{ij}-E_{kj})\bmod6}(\delta_{ir}-\delta_{kr}). \tag*{(20)} \end{aligned}

The real kernel of LL is precisely the space of cores of matrices in TH\mathcal{T}_H. Direct integer multiplication gives Lvec⁡(Rcore)=0L\operatorname{vec}(R_{\mathrm{core}})=0.

Delete column 144 and rows 132, 145, 146, …\ldots, 156 from LL, with indices starting at 1, and keep the remaining rows and columns in their original order. Gaussian elimination over F101\mathbb{F}_{101} on this 143×143143\times143 matrix MM gives

det⁡M≡87(mod101).(21)\det M\equiv87\pmod{101}. \tag*{(21)}

Thus det⁡M≠0\det M\ne0 as an integer, and rank⁡RL≥143\operatorname{rank}_{\mathbb{R}}L\ge143. Since R≠0R\ne0 is in the kernel, we conclude that rank⁡RL=143\operatorname{rank}_{\mathbb{R}}L=143 and TH=RR\mathcal{T}_H=\mathbb{R}R.

Finally, put A=(R∘H)H∗A=(R\circ H)H^*. Multiplication in Z[w]\mathbb{Z}[w] gives

A=A∗,AH=13(R∘H),Tr⁡(A3)=−538434,A=A^*,\qquad AH=13(R\circ H),\qquad\operatorname{Tr}(A^3)=-538434,

and summing the cubes of the entries of (16) gives −3018-3018. Since hR=A/13h_R=A/13 and τ=Tr⁡/13\tau=\operatorname{Tr}/13, it follows that

OH(R)=−538434134−−3018132=−168169.\mathcal{O}_H(R)=\frac{-538434}{13^4}-\frac{-3018}{13^2}=-\frac{168}{169}.

□

Proof of isolation

Proof of Theorem 1.1. Proposition 4.1 gives d(H)=1d(H)=1 and OH(R)≠0\mathcal{O}_H(R)\ne0. Corollary 3.3 proves isolation of HH. Equation (3) gives dim⁡C[D13,uD13u∗]=144−1=143\dim_{\mathbb{C}}[D_{13},uD_{13}u^*]=144-1=143.

We also verify the assertion about the commuting square. We first note that it is nondegenerate: if fj=uejju∗f_j=ue_{jj}u^*, then

τ((pifj)∗pkfl)=δikδjln2,\tau\left((p_if_j)^*p_kf_l\right)=\frac{\delta_{ik}\delta_{jl}}{n^2},

so the products pifjp_if_j form a basis of Mn(C)M_n(\mathbb{C}), with n=13n=13.

By the local reduction in [9 Lemma 2.5], failure of isolation of C(H)\mathcal{C}(H) would give unitaries Um→IU_m\to I such that the commuting squares C(H)(Um)\mathcal{C}(H)(U_m), obtained by replacing QQ with UmQUm∗U_mQU_m^*, are non-isomorphic to C(H)\mathcal{C}(H). Then UmHU_mH are Hadamard matrices converging to HH. If UmHU_mH were equivalent to HH, the corresponding squares would be isomorphic: right monomial multiplication leaves uD13u∗uD_{13}u^* unchanged, and left monomial multiplication conjugates the square by a unitary normalizing D13D_{13}. This contradicts the isolation already proved. □

For the explicit commutator relation, take pi=eiip_i=e_{ii} and qi=qi(R)q_i=q_i(R) from (7). Equations (8) and (19) give

∑i[pi,qi]=0,∑ipiqi=hR,∑iτ(piqihR2−pihRqihR)=−168169.(22)\sum_i [p_i,q_i]=0,\qquad\sum_i p_iq_i=h_R,\qquad\sum_i \tau\left(p_iq_i h_R^2-p_i h_Rq_i h_R\right)=-\frac{168}{169}. \tag*{(22)}

Thus the relation between commutators that makes the span condition fail is incompatible with the second order condition on any nonzero element satisfying the first order condition.

References

References

  1. [1]T. Banica, D. Özteke and L. Pittau, Isolated partial Hadamard matrices and related topics, Open Syst. Inf. Dyn. 25 (2018), no. 2, 1850009, 27 pp. doi:10.1142/S1230161218500099. See also arXiv:1706.00986v2, Conjecture 1.8.DOI
  2. [2]B. R. Karlsson, BCCB complex Hadamard matrices of order 9, and MUBs, Linear Algebra Appl. 504 (2016), 309–324. doi:10.1016/j.laa.2016.04.012.DOI
  3. [3]P. H. J. Lampio, P. R. J. Östergård and F. Szöllősi, Orderly generation of Butson Hadamard matrices, Math. Comp. 89 (2020), 313–331. doi:10.1090/mcom/3453.DOI
  4. [4]P. Lampio, P. Östergård and F. Szöllősi, Dataset for Orderly Generation of Butson Hadamard Matrices, Zenodo, March 13, 2019. doi:10.5281/zenodo.2585766. File matrices/BH-13–6.txt, line 394. License: CC BY 4.0.arxiv.org/abs/1707.02287
  5. [5]D. McNulty and S. Weigert, Mutually unbiased bases in composite dimensions—a review, Quantum 10 (2026), 2051. doi:10.22331/q-2026-04-01-2051. Appendix B, following Lemma B.1; see also arXiv:2410.23997v2.DOI
  6. [6]F. C. Motta, Discrete-time topological dynamics, complex Hadamard matrices, and oblique-incidence ion bombardment, Ph.D. dissertation, Colorado State University, 2014. Section 2.4.3, pp. 114–118, especially Proposition 16, p. 118. hdl:10217/83793.
  7. [8]R. Nicoară, A finiteness result for commuting squares of matrix algebras, J. Operator Theory 55 (2006), 295–310. Journal article.DOI
  8. [9]R. Nicoară, Limit points of commuting squares, Indiana Univ. Math. J. 60 (2011), 847–858. doi:10.1512/iumj.2011.60.4294.DOI
  9. [10]W. Tadej and K. Życzkowski, A concise guide to complex Hadamard matrices, Open Syst. Inf. Dyn. 13 (2006), 133–177. doi:10.1007/s11080-006-8220-2.DOI
  10. [11]W. Tadej and K. Życzkowski, Defect of a unitary matrix, Linear Algebra Appl. 429 (2008), 447–481. doi:10.1016/j.laa.2008.02.036.DOI

Paper details

vlnc:W_km6gm93n
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Mathematics

Not recorded for this paper

View statement
Looking for related papers…
Contents