An operator-algebraic approach for generalized Cardano polynomials Leonard Mada, Maria Anastasia Jivulescu Abstract . We develop an operator‑algebraic framework for generalized Cardano polynomials and show how their structure naturally leads to an operator formulation of Cardano’s method that is compatible with tools and concepts from quantum information theory. The generalized Cardano polynomials are constructed as a generalization of classical theory of Cardano formula for cubic equation, as well as through the spectral properties of the circular operator W = p ​ Z n + q ​ Z n − 1 , W=pZ_{n}+qZ_{n}^{-1}, where Z n Z_{n} is the n-dimensional clock operator. This representation embeds Cardano‑type identities into the spectral theory of circulant operators. The construction clarifies the algebraic structure and solvability of a family of two-parameters odd-order polynomials, classically and through operator methods familiar in QIT, including Fourier transforms and spectral calculus on operator algebras. As applications, we show connections to Cebyshev polynomials and the solution of the quartic order Ferrari equation. Keywords: Cardano’s formula, Weyl-Heisenberg operators, Quantum Fourier transform. 1. Introduction The study of algebraic equations has a long history and classical formulas for solving cubic equations are well-known due to Cardano [ Katz2009 ] . The method consists in reducing the original polynomial x^{3}+ax^{2}+bx+c into a simpler form, called depressed, by replacing x\mapsto x-(a/3) . The famous Cardano formula refers to the solutions of the equation (1) x^{3}+3sx+t=0,\text{ where }s=\frac{b}{3}-\frac{a^{2}}{9}\text{ and }t=c-\frac{ab}{3}+\frac{2a^{3}}{27}, that are given as (2) x_{1}=u_{0}+v_{0},\,x_{1}=\omega u_{0}+\omega^{2}v_{0}\text{ and }x_{2}=\omega^{2}u_{0}+\omega v_{0}, where \omega=e^{\frac{2\pi i}{3}} and u_{0}, v_{0} are primitive solutions of equations (3) u^{3}=\frac{-t+\sqrt{t^{2}+4s^{3}}}{2}\text{ and }v^{3}=\frac{-t-\sqrt{t^{2}+4s^{3}}}{2},\text{ where }\Delta:=t^{2}+4s^{3}\geq 0. Over the past several years, Cardano formulas have been the subject of further studies [ Pfeifer2025 ] . We mention here simplifications of Cardano’s solutions [ Liao2023 ] , adaptations of Cardano’s approach to Vieta-Lucas polynomials and Vieta-Finonacci functions [ Witula2010 ] , as well as more modern approaches, such as the use of superposition principle of quantum mechanics [ Fujii2013 ] . In Fujii’s reinterpretations of Cardano’s formula, an operator formalism has been developed, where spectral properties of low-dimensional operators are used to reproduce the classical solutions. In addition, generalization to Cardano polynomials and its applications to radical reductions are known due to [ Osler2001 ] . In this paper we generalize Cardano’s approach to recover an odd order (n=2m+1) generalized polynomial and its roots. Also, Fujii’s operator techniques for the cubic and quartic case are extended to a family of two parameters odd degree polynomials, based on the observation that, given two real numbers c,d , then two variables p and q, that satisfy the relations pq=c,\,\,p^{n}+q^{n}=2d , form the root x=p+q of a generalized Cardano polynomial. The Fujii operator W=pZ_{n}+qZ_{n}^{-1}, where Z_{n} is the n-dimensional clock operator, satisfies the general odd-degree Cardano-operator polynomial W^{n}=2dI+\sum\limits_{i=0}^{m-1}B_{m,j}c^{m-j}W^{2j+1}, where the coefficients B_{m,j} are derived explicitly. Our operator formalism is based on the observation that the conjugation of the Fujii operator W with the discrete Fourier transform F_{n} , transforms it into X=F_{n}^{+}WF_{n}, , a circulant operator of the form X=pX_{n}+qX_{n}^{-1}, where X_{n} is the cycle shift operator. Circulant operators play a central role in quantum Fourier analysis, quantum walks and Hamiltonian simulations and their spectral properties are completely determined by the Fourier transform. Indeed, the eigenvalues of X are \lambda_{k}=p\omega^{k}+q\omega^{-k},\,\omega=e^{2\pi i/n}, which are exactly the roots of generalized Cardano polynomials. Therefore, we extend the operator Cardano framework to two-parameter family of odd degree operator polynomials, using an algebra generated by clock, shift and Fourier operators [ Gosson2006 ] . Our approach provides a new connection as well as a quantum formulation of the classical algebraic theory of two-parameters family of odd polynomials, using structures specific to Quantum Information Theory [ NC10 ] , such as the operator algebra generated by clock and shift operators. Moreover, quantum circuit realization can be taken into account using qudit phase shifts and Fourier transforms, enabling circuits that realize polynomial spectral transformations. The paper is organized as follows. In Section ( 2 ) we introduce algebraically the odd-degree Cardano polynomials, presenting the general theory, examples, as well as its connections to Chebyshev polynomials. In Section ( 3 ) we consider an operator version of Cardano polynomials and show how this structure encodes the solutions in the operators W as the eigenvalues of the Cardano operator. We extend the operator formalism to Ferrari equations. 2. Generalised Cardano polynomials 2.1. Algebraic approach to Cardano polynomials. Lemma 2.1 . Given c and d, two real parameters( D:=d^{2}-c^{n}\geq 0 ) and n, an odd natural number, the system of equations (4) \begin{cases}x^{n}+y^{n}=2d\\ xy=c\end{cases} has the set of solutions \{(p\omega^{j},\,q\omega^{-j})\},\,j\in\{0,\pm 1,\ldots\pm[(n-1)/2]\}, where (5) p:=\sqrt[n]{d+\sqrt{d^{2}-c^{n}}},\,\,q:=\sqrt[n]{d-\sqrt{d^{2}-c^{n}}}, and \omega=e^{\frac{2\pi i}{n}} is the n-th order root of unity, ( \omega^{n}=1 ). Proof. We notice that, given that the discriminant D , defined as D:=d^{2}-c^{n}\geq 0 , the parameters p and q are well defined (the case D<0 will be treated separately, in Subsection( 2.4 )). Also, the system ( 4 ), in the unknowns x and y , is a symmetric system. Straightforward computations show that the pairs (p\omega^{j},\,q\omega^{-j}),j\in\{0,\pm 1,\ldots\pm[(n-1)/2]\}, satisfy the equations ( 4 ). In this way we recover one real solution and [(n-1)/2] pairs of complex conjugate solutions. ∎ Let us denote (6) x[j]:=p\omega^{j}+q\omega^{-j},j\in\{0,\pm 1,\ldots\pm[(n-1)/2]\}\}. We aim to recover the n-th order equation of real coefficients (for odd n), whose solutions are given by x[j],\,j\in\{0,\pm 1,\ldots\pm[(n-1)/2]\} . To this aim, we introduce the procedure for simple cases, such as n=3 and n=5 . 2.2. 3-rd and 5-th order Cardano polynomials For n=3 , it is easy to show that x[0]=p+q is a solution of the depressed form of the cubic equation (7) x^{3}-3cx-2d=0. Indeed, for x=x[0]=p+q and using straightforward algebraic computation, it follows that (8) \displaystyle x^{3}=(p+q)^{3}=p^{3}+q^{3}+3pq(p+q)\Rightarrow x^{3}=3cx+2d. Therefore, x[0]=p+q is a solution of the cubic equation x^{3}-3cx-2d=0 . In a similar way, it can be shown that x[1]=p\omega+q\omega^{-1} and x[-1]=p\omega^{-1}+q\omega are both solutions of eq. ( 7 ). Hence, for given c and d , the cubic polynomial of Cardano is C_{3,c,d}(x)=x^{3}-3cx-2d and its roots are \{p+q,p\omega+q\omega^{2},p\omega^{2}+q\omega\} . In a similar way we find the 5-th order polynomial equation; indeed, by setting x=x[0]=p+q, \displaystyle x^{5}=(p+q)^{5}=p^{5}+q^{5}+5pq(p^{3}+q^{3})+10p^{2}q^{2}(p+q) Using the relations p^{5}+q^{5}=2d,pq=c and p^{3}+q^{3}=x^{3}-3cx , it follows that x^{5}=5cx^{3}-5c^{2}x+2d. Again, for given c and d , the fifth polynomial of Cardano is C_{5,c,d}(x)=x^{5}-5cx^{3}+5c^{2}x-2d and its roots are \{p+q,p\omega+q\omega^{4},p\omega^{4}+q\omega,p\omega^{2}+q\omega^{3},p\omega^{3}+q\omega^{2}\} . We understand that the previous procedure can be extended to any arbitrary odd natural number n ; that is, based on the real numbers c and d and the associated parameters p and q , as given by ( 5 ), and their properties ( 4 ), we can find the odd order polynomial equations satisfied by x=p+q , using a simplified procedure, similar to the Cardano formula. More exactly, based on the binomial formula of n-th order for (p+q) , we can express all the terms of decompositions in terms of the sum p+q , as well as of product pq=c and, in this way, we will recover the n-th order Cardano polynoimial. Let’s denote it by C_{n,c,d}(x) . We present here in brief the first few polynomials of Cardano (for given parameters c and d ): (9) \displaystyle C_{3,c,d}(x)=x^{3}-3cx-2d (10) \displaystyle C_{5,c,d}(x)=x^{5}-5cx^{3}+5c^{2}x-2d (11) \displaystyle C_{7,c,d}(x)=x^{7}-7cx^{5}+14c^{2}x^{3}-7c^{3}x-2d. 2.3. The n-th order Cardano polynomial In the following, we aim to find the polynomial equations of odd order ( n=2m+1 ), using a similar procedure, as shown above for n=3 and n=5 . To this aim, we are going to state the following Theorem 2.2 . Consider n, an odd natural number ( n=2m+1 ), and the real parameters p and q, as given by ( 5 ). The n-th order polynomial equation whose solutions are x[j]=p\omega^{j}+q\omega^{-j},j\in\{0,\pm 1,\ldots,\pm[(n-1)/2]\} is (12) \displaystyle x^{2m+1}=2d+\sum\limits_{j=0}^{m-1}B_{m,j}c^{m-j}x^{2j+1},\text{ where } (13) \displaystyle B_{m,j}=(-1)^{m-1-j}\frac{2m+1}{2j+1}\begin{pmatrix}m+j\\ 2j\end{pmatrix},0\leq j\leq m-1. The polynomial (14) C_{n,c,d}(x)=x^{2m+1}-\sum\limits_{j=0}^{m-1}B_{m,j}c^{m-j}x^{2j+1}-2d, associated to the poynomial equation ( 12 ) will be called the n-th order generalized Cardano polynomial. Proof. Using the binomial formula for computing x^{n}, where x=p+q and n=2m+1, as well as Kummer’s formula [ Ma2005 ] to express terms of the form p^{r}+q^{r},\,r\geq 1 , we express x^{n} as in ( 12 ) in terms of coefficients ( 13 ). For a complete proof, see Appendix ( 5 ). Hence, we obtain the n-th order polynomial equation as given by ( 12 ), whose solution is p+q=x[0] . Moreover, also x[j]=p\omega^{j}+q\omega^{-j},j\in\{0,\pm 1,\ldots,\pm m)\} are solutions of ( 12 ) by the simple fact that p\omega^{j} and q\omega^{-j} satisfy the same properties ( 4 ) as p and q , with respect to their (2m+1) -th power sum and their product, expressions that appear also in eqs. ( 42 ) and ( 41 ). We conclude that, given equation ( 12 ), we can compute their solutions x[\pm j],\,\,j\in\{0,\pm 1,\ldots,\pm m\}, as given by ( 6 ). ∎ Remark 2.3 . The values x[j] will be referred as the j-th branches of the solution, as replacing j by j+k(\text{ mod }n) will only permute the set of roots. In the following, we present a couple of examples of Cardano polynomials that can be recovered using ( 14 ). Straightforward computations enable us to write the cubic Cardano polynomials, C_{n,c,d} , for given c, d, real parameters. Indeed, for n=3 , we write C_{3,c,d}(x)=x^{3}-B_{1,0}cx-2d, where B_{1,0}=3 (as given by ( 13 ), for m=1,\,j=0 ). So, C_{3,c,d}(x)=x^{3}-3cx-2d . In a similar manner, for n=5, the polynomial C_{5,c,d}(x)=x^{5}-\sum\limits_{j=0}^{1}B_{2,j}c^{2-j}x^{2j+1}-2d, where B_{2,0}=(-1)5\begin{pmatrix}2\\ 0\end{pmatrix}=-5, and B_{2,1}=\frac{5}{3}\begin{pmatrix}3\\ 2\end{pmatrix}=5. Therefore, C_{5,c,d}(x)=x^{5}-5cx^{3}+5c^{2}x-2d . Similar computations allow us to recover high order Cardano polynomials, as given by ( 14 ). Moreover, our procedure allows to recover also the roots of the Cardano polynomials in a simpler manner, as given by ( 6 ). To this aim, we illustrate a few simple examples, focusing both on the construction of Cardano polynomial of a given order, starting from assessing explicit values of the parameters c and d , as well as on the method to find the complete set of solutions, for a given polynomial equation which can be cast in Cardano form. Example 1. Consider c=d=1 . The values p=q=1 are obtained by straightforward computations, which also provide the equation for the associated cubic polynomial x^{3}-3x-2=0 . Its solutions are x[0]=2,\,x[1]=\omega+\omega^{-1}=-1 and x[-1]=\omega^{-1}+\omega=-1 . For the same values of the parameters c,d,p and q , the fifth order polynomial equation is x^{5}-5x^{3}+5x-2=0 and the set of solutions is x[0]=2,x[\pm 1]=\omega+\omega^{4} and x[\pm 2]=\omega^{2}+\omega^{3} . Example 2. Let us consider the cubic polynomial equation x^{3}-6x-9=0 . The identification with the form ( 7 ) gives us the following values for the parameters c=2,\,d=9/2 ; therefore, the parameters p and q, given by eqs. ( 5 ), are p=2,\,q=1 . The solutions of the equation are given by ( 6 ), as x[j]=2\omega^{j}+\omega^{-j},\,j\in\{0,\pm 1\}. Straightforward computations provide the complete set of solutions x[0]=3,\,x[1]=2\omega+\omega^{2}=\frac{-3+i\sqrt{3}}{2} and x[-1]=2\omega^{2}+\omega=\frac{-3-i\sqrt{3}}{2} . 2.4. Negative discriminant case In the following we study the case of negative discriminant D=d^{2}-c^{n}<0 . We aim to find a closed form for the roots of the n-th order Cardano polynomial. In this case, the parameters p and q, as defined by eqs. ( 5 ), are complex conjugates and are the solution of the system of equations given by ( 4 ). Given that the parameter d can be rewritten as d=c^{n/2}\cos(\alpha) , where \alpha=\arccos(\frac{d}{c^{n/2}}) (this is possible, given that D<0 ), it holds that D=d^{2}-c^{n}=c^{n}\cos(\alpha)^{2}-c^{n}=-c^{n}sin^{2}(\alpha). Therefore, (15) \sqrt{D}=ic^{n/2}\sin(\alpha). Applying de Moivre’s formula, the parameters p and q are (16) \displaystyle p=\sqrt{c}(\cos\frac{\alpha}{n}+i\sin\frac{\alpha}{n}),\,\,q=\sqrt{c}(\cos\frac{\alpha}{n}-i\sin\frac{\alpha}{n}) Therefore, the j-th root of Cardano generalized polynomials, for the case D<0 , given by the eq ( 6 ), can be written as (17) \displaystyle x[j]=p\omega^{j}+q\omega^{-j}=2\sqrt{c}\cos(\frac{\alpha+2\pi j}{n}),\,\forall j\in\{0,\pm 1\ldots,\pm[(n-1)/2]\}. We notice that the case of n-th order Cardano polynomial, for which D<0 , corresponds to real roots. For j=0 we have that (18) x[0]=p+q=2c^{1/2}\cos[\alpha/n] Example Consider the case d=2 and c=3 and n=5 . The associated 5-th order Cardano polynomial equation is given by x^{5}-15x^{3}+45x-4=0 , with D=-239 . Its solutions are x[0]=(2+i\sqrt{239})^{1/5}+(2-i\sqrt{239})^{1/5}\approx 3.321 Using the trig-identities ( 17 ), we recover the same set of solutions, which can be written in a more accessible form (19) x[j]=2\sqrt{3}\cos[(\arccos(2/3^{5/2})+2\pi j)/5],\,\,j\in\{0,1,2\}. For j=0, we have \displaystyle x[0]=2\sqrt{3}\cos[\arccos(2/3^{5/2})/5]\approx 3.321 2.5. Generalized Chebyshev Polynomials In the following we aim to connect the theory of generalized Cardano polynomials to that of Chebyshev polynomials. We recall that the so-called modified Chebyshev polynomials of the first kind are defined as: (20) \Omega_{n}(x):=\sum\limits_{k=0}^{n}(-1)^{k}\frac{n}{n-k}\begin{pmatrix}n-k\\ k\end{pmatrix}x^{n-2k},\,n\in\mathbb{N}. The polynomial \Omega_{n}(x) , also called the n-th order Vieta-Lucas polynomial [ Witula2007 ] , satisfies the identity: (21) \Omega_{n}(x)=2T_{n}(\frac{x}{2}), where T_{n}(\cos\theta)=\cos(n\theta) is the n-th Chebyshev polynomial of first kind. The modified Chebyshev polynomials of the first kind, \Omega_{n}(x) , have the following properties: \displaystyle\Omega_{1}(x)=x,\,\Omega_{2}(x)=x^{2}-2 (22) \displaystyle\Omega_{n+2}(x)=x\Omega_{n+1}(x)-\Omega_{n}(x),n\in\mathbb{N}. Using Theorem 4.1, from [ Witula2010 ] , we notice that Chebyshev polynomials of the first kind can be written as follows: (23) \displaystyle(\sqrt{c})^{n}\Omega_{n}(\frac{x}{\sqrt{c}})-2d=\prod\limits_{k=0}^{n-1}(x-p\omega^{k}-q\omega^{-k}) The right hand side term of ( 23 ) is, actually, the decomposition of the n-th order Cardano generalized polynomial ( 12 ), C_{n}[x] , we establish the following representation of modified Chebyshev polynomials of the first kind (n=2m+1) : (24) \displaystyle\Omega_{n}(x)=c^{-n/2}C_{n,c,d}(\sqrt{c}x)+2dc^{-n/2}. Using the recurrence relation ( 2.5 ), we recover the following recurrence relation between generalized Cardano polynomials (25) C_{{n+2},c,d}(x)-xC_{{n+1},c,d}(x)+cC_{n,c,d}(x)=2d(x-c-1),n\geq 1, which can be used to recover the generalized even-order Cardano polynomials, in terms of odd order Cardano polynomials, C_{n,c,d}(x), as given by ( 14 ). Moreover, the class of two-parameters Cardano polynomials can be further connected to Dickson polynomials [ Dickson1897 ] , D(x,a) given their relation to Chebyshev polynomials, as given by D(x,a)=2(\sqrt{a})^{n}T_{n}(\frac{x}{2\sqrt{a}}). 2.6. Ferrari polynomial equation. We recall now that Ferrari quartic equation [ Fujii2013 ] (26) x^{4}+ax^{2}+bx+c=0 can be related to cubic Cardano polynomial equation, by writing the polynomial of eq. ( 26 ) as a difference of two square polynomials, as follows x^{4}+ax^{2}+bx+c=(x^{2}+y)^{2}-(\alpha x+\beta)^{2}. Consequently, solving eq. ( 26 ) means to find the roots of the cubic equation 4(y^{2}-c)(2y-a)-b^{2}=0. By writing the depressed form of the previous cubic equation and by using the classical approach of cubic Cardano polynomial, the solution of Ferrari equation eq.( 26 ) is given by the following (27) \displaystyle x=\frac{\alpha\pm\sqrt{\alpha^{2}-4(y-\beta)}}{2}\text{ and }x=\frac{-\alpha\pm\sqrt{\alpha^{2}-4(y-\beta)}}{2}. We exemplify with the following equation x^{4}+6x^{2}+8x+3=0. By writing x^{4}+6x^{2}+8x+3=(x^{2}+y-\alpha x-\beta)(x^{2}+y+\alpha x+\beta) it follows that the parameters y,\alpha and \beta have to satisfy (28) \begin{cases}2y-\alpha^{2}=8\\ -2\alpha\beta=8\\ y^{2}-\beta^{2}=3.\end{cases} By reducing \alpha and \beta , it follows the third order equation y^{3}-3y^{2}-3y+1=0 which, by substitution y=z+1 , transforms into z^{3}-6z-4=0. Using that it has the form( 7 ), for c=d=2, and n=3 , we compute p,q=\sqrt[3]{2\pm i\sqrt{2}}. Given that D=-4, the solutions are z[j]=2\sqrt{2}\cos(\frac{\pi/4+2\pi j}{3}) . The set of solution for the cubic equation in y are y[j]=z[j]+1 and the solution of initial Ferrari equation follows naturally using ( 27 ). 3. Cardano operator In the following we are going to generalize the results introduced in [ Fujii2013 ] that consist in finding the roots of cubic Cardano polynomial using operator setting. 3.1. Fujii operator Firstly let’s recall few important notions from Quantum Information Theory (QIT) [ NC10 ] . In a given n-dimensional Hilbert space \mathcal{H}=\mathbb{C}^{n}, the computational basis is given by \{|0\rangle,|1\rangle,\ldots,|n-1\rangle\}, where Dirac notation is used to denote the j-th vector |j\rangle=(0\,\ldots 1\ldots 0)^{T} of the computational basis. The generalized Pauli operators are defined as: (29) \displaystyle\text{ Shift operator: }X_{n}|j\rangle=|j+1\rangle(\text{ modulo }n), (30) \displaystyle\text{ Clock operator: }Z_{n}|j\rangle=\omega^{j}|j\rangle, where \omega=e^{2\pi i/n} and n\geq 2 . For example, for n=2 we recover the well-known Pauli operators X_{2}=\sigma_{x}=\begin{pmatrix}0&1\\ 1&0\end{pmatrix} and Z_{2}=\sigma_{z}=\begin{pmatrix}1&0\\ 0&-1\end{pmatrix}, whereas for n=3 we have X_{3}=\begin{pmatrix}0&0&1\\ 1&0&0\\ 0&1&0\end{pmatrix} and Z_{3}=\text{ diag }[1,\omega,\omega^{2}] . Also, we have that Z_{n}X_{n}=\omega X_{n}Z_{n},\,\forall n\geq 2 . The discrete Fourier transform is defined as (31) F_{n}|j\rangle:=\frac{1}{\sqrt{n}}\sum\limits_{i=0}^{n-1}\omega^{ij}|i\rangle,\,\forall n\geq 2 It holds that F_{n}^{+}Z_{n}F_{n}=X_{n} and F_{n}^{+}Z_{n}^{-1}F_{n}=X_{n}^{-1}, where (\cdot)^{+} is the conjugate transpose operation. We call Fourier basis the orthonormal basis obtained by applying the discrete Fourier F_{n} transform (DFT) to the computational basis of \mathbb{C}^{n} . The Fourier basis vectors |\tilde{j}\rangle are the vectors of computational basis transformed by DFT, i.e. (32) |\tilde{j}\rangle:=F_{n}^{+}|j\rangle=\frac{1}{\sqrt{n}}\sum\limits_{i=0}^{n-1}\omega^{-ij}|i\rangle,\,j\in\{0,1,\ldots,n-1\} Given that F_{n}|\tilde{j}\rangle=|j\rangle, the Fourier basis diagonalizes the shift operator as X_{n}|\tilde{j}\rangle=\omega^{j}|\tilde{j}\rangle . Now, using the basic notions from QIT introduced previously, we focus on developing the operator approach for generalized Cardano polynomials. Proposition 3.1 . Let n\geq 3 be an integer and c,d reals parameters( as well as p and q as given by ( 5 )). The Fujii operator W , as defined by (33) W:=pZ_{n}+qZ_{n}^{-1}, has the following properties: i) the standard computational basis vector |j\rangle is an eigenvector of W with eigenvalue \lambda_{j}=p\omega^{j}+q\omega_{j}^{-1}, for each j\in\{0,1,\ldots,n-1\} . ii) the operator W is normal and, for any polynomal p\in\mathbb{C}(x), it holds that p(W)=\sum\limits_{j}p(\lambda_{j})|j\rangle\langle j| iii) It holds the operator identity C_{n,c,d}(W)=0, where C_{n,c,d}(x) is the n-th order generalized Cardano polynomial. Proof. i) The operator is diagonal by definition, as Z_{n}=\text{ diag }[1,\omega,\omega^{2},\ldots\omega^{n}],\,\forall n\geq 3 . Moreover, Z_{n} is unitary and Z_{n}^{-1}=Z_{n}^{+}, so Z_{n}^{-1}|j\rangle=\omega^{-j}|j\rangle . Consequently, the operator is also diagonal, as linear combination of two diagonal operators, that is W=pZ_{n}+qZ_{n}^{-1}=\text{ diag }[p\omega^{j}+q\omega^{-j}],\,j\in\{0,1,\ldots,n-1\}. Therefore, the operator ( 33 ) has eigenvalues \lambda_{j}=p\omega^{j}+q\omega^{-j} . Indeed, (34) \displaystyle W|j\rangle=(pZ_{n}+qZ_{n}^{-1})|j\rangle=(p\omega^{j}+q\omega^{-j})|j\rangle. ii) The operator W is a normal operator ( WW^{+}=W^{+}W .) Given a polynom p\in\mathbb{C}(T), it holds that p(W)=\sum_{j}p(\lambda_{j})|j\rangle\langle j|, so p(W) is diagonal in the same basis and the eigenvalues are p(\lambda_{j}) . iii) Given that \lambda_{j}=p\omega^{j}+q\omega^{-j},\,\forall j, are the roots x[j] of C_{n,c,d}(x), it follows that C_{n,c,d}(W)=0, where the operator identity is defined by functional calculus on diagonal matrices. ∎ The operator polynomial C_{n,c,d}(W)=0 , that is (35) W^{2m+1}=2dI+\sum\limits_{j=0}^{m-1}B_{m,j}c^{m-j}W^{2j+1}, is the generalization of Fujii’s cubic and quartic operators [ Fujii2013 ] and its roots are x[j],\in j\in\{0,\pm 1,\ldots,\pm m\} . We notice that the operator W encodes the roots structure of generalized Cardano polynomials C_{n,c,d}(x) . We will use the operator W to define another operator, which is defined by Fourier conjugation of W . 3.2. Cardano operator. Given the operator W, associated to n,c and d as in ( 33 ), we define the operator (36) X:=F_{n}^{+}WF_{n},\,n\geq 3 and called it Cardano operator. It holds that Proposition 3.2 . The Cardano operator X has the followings properties: a) It can be written as X=pX_{n}+qX_{n}^{-1} ( circular form), where X_{n} is the shift operator. b) The operator X has the same spectrum as W . c) The operator X satisfies the operator equation C_{n,c,d}(X)=0 , where C_{n,c,d}(X) is the n-th order Cardano- polynomial for given real parameters c and d , as given by eq.( 12 ). Proof. a) Given that F_{n}^{+}Z_{n}F_{n}=X_{n} and F_{n}^{+}Z_{n}^{-1}F_{n}=X_{n}^{-1}, it follows the the operator X can be written as X=pX_{n}+qX_{n}^{-1} . Given the structure of the operators X_{n} and X_{n}^{-1}, for which every row is a cycle shift of the first, we conclude that X is a circulant matrix of the form X=\operatorname{circ}(0,q,0,\ldots,0,p) . b) Given its circular structure as well as the fact that is unitarily equivalent to W , the operator X is diagonalizable in the Fourier eigenbasis and its eigenvalues are \lambda_{\tilde{j}}=p\omega^{j}+q\omega^{-j}. c) As F_{n} is a unitary matrix, it holds that X^{k}=F_{n}^{+}W^{k}F_{n}=F_{n}^{+}\text{diag }[\lambda^{k}[j]]F_{n} . We have that C_{n}(X)=F_{n}^{+}C_{n}(W)F_{n}=0 . ∎ Let’s explain now in details Fujii’s construction [ Fujii2013 ] , that corresponds for the case n=3 . Suppose that for given real parameters c and d ( and consequently p and q ), we construct the third order Cardano polynomial C_{3,c,d}(x)=x^{3}-3cx-2d, whose roots are denoted by x[j],\,j\in\{0,\pm 1\} . Now  relying on the operator approach, we introduced the operator Z_{3}=\text{diag }[1,\omega,\omega^{2}] and W=pZ_{3}+qZ_{3}^{-1}=\text{diag }[p+q,p\omega+q\omega^{-1},p\omega^{2}+q\omega^{-2}]=\text{diag }[\lambda_{j}], where \lambda_{j},\,j\in\{0,1,2\} denotes the eigenvalues of W . Suppose that W=\text{ diag}[x[0],x[1],x[-1]], where x[j]\,j\in\{0,\pm 1\} are the roots of C_{3,c,d}(x), as given by ( 6 ). Using the discrete quantum Fourier transform, F_{3} , the operator X:=F_{3}^{+}WF_{3} has the following form (37) X=\frac{1}{3}\begin{pmatrix}x[0]+x[1]+x[-1]&x[0]+x[1]\omega+x[2]\omega^{2}&x[0]+x[1]\omega^{2}+x[-1]\omega\\ x[0]+x[1]\omega^{2}+x[-1]\omega&x[0]+x[1]+x[-1]&x[0]+x[1]\omega+x[-1]\omega^{2}\\ x[0]+x[1]\omega+x[-1]\omega^{2}&x[0]+x[1]\omega^{2}+x[-1]\omega&x[0]+x[1]+x[-1]\end{pmatrix} In the same time, X=pX_{3}+qX_{3}^{-1}=\begin{pmatrix}0&q&p\\ p&0&q\\ q&p&0\end{pmatrix}=\operatorname{circ}(0,q,p), where X_{3} is generalized Pauli matrix. Consequently, from the equality of the two forms of X, it follows that (38) \begin{pmatrix}0&q&p\end{pmatrix}=\begin{pmatrix}x[0]&x[1]&x[-1]\end{pmatrix}\frac{1}{3}\begin{pmatrix}1&1&1\\ 1&\omega&\omega^{2}\\ 1&\omega^{2}&\omega\end{pmatrix}\Leftrightarrow\begin{pmatrix}0&q&p\end{pmatrix}=\begin{pmatrix}x[0]&x[1]&x[-1]\end{pmatrix}\frac{1}{\sqrt{3}}F_{3} Therefore, the vector of the roots x[j],\,j\in\{0,1,2\} is given by (39) \begin{pmatrix}x[0]&x[1]&x[-1]\end{pmatrix}=\begin{pmatrix}0&q&p\end{pmatrix}\cdot\sqrt{3}F_{3}^{+}=\begin{pmatrix}p+q&p\omega+q\omega^{2}&p\omega^{2}+q\omega\end{pmatrix}. Consequently, we obtain the set of roots for the cubic Cardano polynomial, as given by ( 6 ), for j\in\{0,\pm 1\} and that they are equal to the eigenvalues of the operator W, as well as of X (given that the two matrices are similar). Indeed, the above results follows naturally using ( 3.2 ). The Cardano operator cubic polynomial equation is X^{3}-3cX-2dI=0 and the eigenvalues of the Cardano operator X , as given by ( 36 ), are the roots of the polynomial C_{3,c,d}(x). The above construction can be extended for any n odd natural number. The Cardano operator X=F_{n}^{+}WF_{n} has the eigenvalues given by the roots ( 6 ) of the Cardano polynomial C_{n,c,d}(x), associated to the polynomial equation ( 43 ). 3.3. Ferrari operator. In this section we show that the operator‑theoretic Ferrari construction for quartic equations is a special case of the general spectral method developed earlier for Cardano polynomials. The following procedure reconsider in the operator setting the procedure presented in section( 2.6 ), for solving quartic Ferrari equations. Indeed, Ferrari’s method reduces the quartic operator equations (40) X^{4}+A_{2}X^{2}+A_{1}X+A_{0}I=0, to the cubic Cardano equation Z^{3}-3CZ-2DI=0, where C and D are functions of initial real coefficients A_{i},\,i\in\{0,1,2\} . Using the Cardano operator W=pZ_{3}+qZ_{3}^{-1}, where p and q are associated to the parameters C and D, we can solve the cubic depressed Cardano equation and therefore find the solution of Ferrari operator equation using a similar procedure as presented in section ( 2.6 ), by transforming the scalar equations into operator identities. 4. Conclusions This paper introduces a generalization of Cardano’s formula for a subclass of higher-order polynomials. Although the technique can be used to solve all polynomials of order 3, only a subset of polynomials of higher order can be solve. We construct this family of two-parameters odd-order polynomials and called them generalized Cardano polynomials. We also use our method to extend Fujii’s operator formulation [ Fujii2013 ] of Cardano method beyond cubic case. Our approach show that the algebraic structure of two-parameters odd order polynomials can be expressed entirely within the algebra generated by the clock operator Z_{n} . The resulting operator W satisfies the general odd-degree Cardano polynomial identity ( 35 ) and it provides in a single, closed form a family operator polynomials, that are connected to a family of two-parameters odd-degree equations. We unify the classical algebraic methods to that of operator formulation which, based on spectral analysis, offers similar results. The operator Cardano approach offers a bridge between classical algebra and quantum operator theory and a practical way to encode this particular polynomial structure into the finite-dimensional operators. We show that based on Cardano operator we can find an mechanism to solve a class of two-parameters odd‑degree polynomials equations and highlights structural connections between algebraic solvability, circulant operators, and quantum‑inspired operator calculus. Future work may explore the applications of operators identities in quantum computations procedures, such as quantum algorithms, spectral engineering, where polynomials relations and Fourier diagonalization play central role. 5. Appendix In the following we are going to give a complete proof for the polynomial equation that characterizes the Cardano polynomial C_{n,c,d}(x) . The binomial formula applied to x=p+q, as given by ( 4 ) is (41) \displaystyle x^{n}=(p+q)^{n}=p^{n}+q^{n}+\sum\limits_{k=1}^{m}\begin{pmatrix}2m+1\\ k\end{pmatrix}(pq)^{k}(p^{2(m-k)+1}+q^{2(m-k)+1}). The term p^{r}+q^{r},\,\forall r\in\mathbb{N^{*}} can be written in terms of x=p+q and xy=c as given by Kummer formula [ Ma2005 ] (42) p^{r}+q^{r}=\sum\limits_{i=0}^{[r/2]}(-1)^{i}\frac{r}{r-i}\begin{pmatrix}r-i\\ i\end{pmatrix}(pq)^{i}(p+q)^{r-2i}=\sum\limits_{i=0}^{[r/2]}(-1)^{i}\frac{r}{r-i}\begin{pmatrix}r-i\\ i\end{pmatrix}c^{i}x^{r-2i}, Consequently, eq. ( 41 ) becomes (43) \displaystyle x^{n}=2d+\sum\limits_{k=1}^{m}\sum\limits_{i=0}^{m-k}(-1)^{i}\begin{pmatrix}2m+1\\ k\end{pmatrix}\frac{2(m-k)+1}{2(m-k)+1-i}\begin{pmatrix}2(m-k)+1-i\\ i\end{pmatrix}c^{i+k}x^{2(m-k)+1-2i} By changing the index i\mapsto j , where j=m-k-i, we recover the following form (44) \displaystyle x^{n}=2d+\sum\limits_{k=1}^{m}\sum\limits_{j=0}^{m-k}(-1)^{m-k-j}\begin{pmatrix}2m+1\\ k\end{pmatrix}\frac{2(m-k)+1}{m-k+1+j}\begin{pmatrix}m-k+1+j\\ m-j-k\end{pmatrix}c^{m-j}x^{1+2j}. For fixed j\in\{0,\ldots,m-1\}, we will denote the coefficient of c^{m-j}x^{2j+1} as follows (45) \displaystyle B_{m,j}=\sum\limits_{k=1}^{m-j}(-1)^{m-j-k}\begin{pmatrix}2m+1\\ k\end{pmatrix}\frac{2(m-k)+1}{m-k+1+j}\begin{pmatrix}m-k+1+j\\ 1+2j\end{pmatrix}. Let’s denote t=m-j-k . Then, (46) \displaystyle B_{m,j}=\frac{1}{2j+1}\sum\limits_{t=0}^{m-j-1}(-1)^{t}\begin{pmatrix}2m+1\\ m-j-t\end{pmatrix}[2(j+t)+1]\begin{pmatrix}2j+t\\ t\end{pmatrix}. Let’s complete the summation in B_{m,j} up to (m-j) and denote it by S_{m,j} . Then, (47) B_{m,j}=S_{m,j}-(-1)^{m-j}\frac{2m+1}{2j+1}\begin{pmatrix}m+j\\ m-j\end{pmatrix}. We are going to show the the term S_{m,j}=\frac{1}{2j+1}\sum\limits_{t=0}^{m-j}(-1)^{t}\begin{pmatrix}2m+1\\ m-j-t\end{pmatrix}[2(j+t)+1]\begin{pmatrix}2j+t\\ t\end{pmatrix} reduces to 0. To this aim, we notice that the term (-1)^{t}(2j+2t+1)\begin{pmatrix}2j+t\\ t\end{pmatrix} can be found by using the expansion \sum\limits_{t=0}^{\infty}\begin{pmatrix}2j+t\\ t\end{pmatrix}(-x)^{t}=(1+x)^{-(2j+1)},\,|x|<1 by applying the differential operator 2x\frac{d}{dx}+(2j+1). Indeed, by applying the the differential operator on the left-hand side of the expansion we get \displaystyle[2x\frac{d}{dx}+(2j+1)]\left(\sum\limits_{t=0}^{\infty}\begin{pmatrix}2j+t\\ t\end{pmatrix}(-x)^{t}\right)=\sum\limits_{t=0}^{\infty}(2j+2t+1)\begin{pmatrix}2j+t\\ t\end{pmatrix}(-x)^{t}, whereas, applying on the right-hand side of the expansion, it yields [2x\frac{d}{dx}+(2j+1)]\left((1+x)^{-(2j+1)}\right)=\frac{(2j+1)(1-x)}{(1+x)^{2j+2}}. Therefore, the sum S_{m,j} is the coefficient of x^{n},\,(n=m-j) in the product of the above generating function and the generating function for the remaining binomial coefficient of (1+x)^{2m+1} . Here [x^{n}][P(x)] denotes the coefficient of x^{n} from the polynomial P(x) . Hence, we have \displaystyle S_{m,j}=\frac{1}{2j+1}[x^{n}]\cdot\left[(1+x)^{2m+1}\frac{(2j+1)(1-x)}{(1+x)^{2j+2}}\right]. Consequently, \displaystyle S_{m,j}=[x^{n}]\left[(1-x)(1+x)^{2n-1}\right]=[x^{n}]\left[(1+x)^{2n-1}\right]-[x^{n-1}]\left[(1+x)^{2n-1}\right], which are the following binomial coefficients S_{m,j}=\begin{pmatrix}2n-1\\ n\end{pmatrix}-\begin{pmatrix}2n-1\\ n-1\end{pmatrix}=0 . Therefore, eq.( 47 ) is going to be equal to ( 13 ).