betaFeedback

WorksPhysics and Astronomy

Shadows of new physics on Dirac materials, analog GUPs and other amusements

Abstract

We discuss here how, when higher-order effects in the parameter ℓℏ∣p⃗∣\frac{\ell}{\hbar} |\vec{p}|, related to the lattice spacing ℓ\ell, are considered, pristine graphene, and other Dirac materials, can be used as tabletop systems where generalized commutation relations are naturally realized. Such generalized algebras of quantization, which lead to generalized versions of the Heisenberg uncertainty principle, are under intense scrutiny these days, as they could manifest a fundamental length scale of spacetime. Despite the efforts and the many intriguing results, there are no experimental signatures of any generalized uncertainty principle (GUP). Therefore, our results here, which tell how to use tabletop physical systems to test certain GUPs in analog experiments, should be of interest to practitioners of quantum gravity. We identify three different energy regimes that we call ``layers'', where the physics is still of a Dirac type but within precisely described limits. The higher the energy, the more sensitive the Dirac system becomes to the effects of the lattice. Here such lattice plays the role of a discrete space where the Dirac quasi-particles live. With the goals just illustrated, we had to identify the mapping between the high-energy coordinates, XiX^i, and the low-energy ones, xix^i, i.e., those measured in the lab. We then obtained three generalized Heisenberg algebras. For two of them we have the noticeable result that Xi=xiX^i = x^i, and for the third one we obtained an improvement with respect to an earlier work: the generalized coordinates expressed in terms of the standard phase space variables, Xi(x,p)X^i(x,p), and higher order terms.

Introduction

In the search for new physics, it is an established tradition to look for effects of high energy phenomena at our low energy scales. One way to do it is to consider the action for the “high-energy” fields, Φ^a\widehat{\Phi}_a, as an expansion in terms of “low-energy” fields, Φa\Phi_a and a suitable parameter ϵ\epsilon, so that

S(Φ^a)=S(Φa)+S1(ϵ,Φa,∂)+O(ϵ2,∂2).(1)S(\widehat{\Phi}_a) = S(\Phi_a) + S_1(\epsilon,\Phi_a,\partial) + O(\epsilon^2,\partial^2). \tag*{(1)}

Examples from earlier research are many. Two that come to mind are the noncommutative field theories faced with the help of the Seiberg-Witten map [1]

A^μ=Aμ−12θαβAα(∂βAμ+Fβμ)+O(θ2),F^μν=Fμν+θαβ(FμαFνβ−Aα∂βFμν)+O(θ2),(2)\begin{aligned} \widehat{A}_{\mu} &= A_{\mu} - \frac{1}{2}\theta^{\alpha\beta} A_{\alpha}(\partial_{\beta}A_{\mu} + F_{\beta\mu}) + O(\theta^2),\\ \widehat{F}_{\mu\nu} &= F_{\mu\nu} + \theta^{\alpha\beta}(F_{\mu\alpha}F_{\nu\beta} - A_{\alpha}\partial_{\beta}F_{\mu\nu}) + O(\theta^2), \tag*{(2)} \end{aligned}

that for (1) gives

S(A^μ)=−14F^⋅F^=−14(F⋅F−12(θ⋅F)(F⋅F)+2(FθF)⋅F)+O(θ2),(3)S(\widehat{A}_{\mu}) = -\frac{1}{4}\widehat{F}\cdot\widehat{F} = -\frac{1}{4}\left(F\cdot F - \frac{1}{2}(\theta\cdot F)(F\cdot F) + 2(F\theta F)\cdot F\right) + O(\theta^2), \tag*{(3)}

where θ∼ℓP2\theta\sim\ell_P^2 and ℓP\ell_P is a reference length, usually taken to be the Planck length. For more on this see [1, 2, 3, 4] for the theory and, e.g., [5, 6, 7] for some phenomenology.

Another, somehow more general, example is Colladay-Kostelecky’s Standard Model Extension [8], that can be described by

S(Φ^a)=SSM(Φa)+∑k=1∞Cμ…ν(k)(Φa,∂)μ…ν,(4)S(\widehat{\Phi}_a) = S_{SM}(\Phi_a) + \sum_{k=1}^{\infty} C^{(k)}_{\mu\ldots\nu}(\Phi_a,\partial)^{\mu\ldots\nu}, \tag*{(4)}

with C(k)∼ℓPkC^{(k)} \sim\ell_P^k. For a review of both approaches, and a comparison, see, e.g., [9].

These are the days of the generalized uncertainty principles (GUPs) [10, 11, 12]; see also the recent review [13]. Many approaches exist that consider high-energy phase space variables, (x,p)(x,p), vs low-energy variables, (x0,p0)(x_0,p_0). Here we focus on a specific one that has been introduced by Ali, Das and Vagenas (ADV) in [14, 15, 16, 17]

pi=p0i(1−A∣p⃗0∣+2A2∣p⃗0∣2),(5)p_i = p_{0i}\left(1 - A|\vec{p}_0| + 2A^2|\vec{p}_0|^2\right), \tag*{(5)}

where A=A~ℓP/ℏA = \widetilde{A}\ell_P/\hbar, A~\widetilde{A} is a numerical factor and

xi=x0i,(6)x_i = x_{0i}, \tag*{(6)}

with [x0i,p0j]=ıℏδij[x_{0i},p_{0j}] = \imath\hbar\delta_{ij}, but where

[xi,pj]=ıℏ[δij−A∣p⃗∣(δij+pipj∣p⃗∣2)+A2∣p⃗∣2(δij+3pipj∣p⃗∣2)].(7)[x_i,p_j] = \imath\hbar\left[\delta_{ij} - A|\vec{p}|\left(\delta_{ij} + \frac{p_i p_j}{|\vec{p}|^2}\right) + A^2|\vec{p}|^2\left(\delta_{ij} + 3\frac{p_i p_j}{|\vec{p}|^2}\right)\right]. \tag*{(7)}

There is an intense activity to find GUP-corrected physics. Besides the already cited work of [14, 15, 16, 17], other important results that use different approaches are in [18, 19, 20]. A compelling example for this paper, for instance, is in [17] where it is found that the GUP-corrected Dirac equation, in any dimension, is

Hψ=(cα⃗⋅p⃗+βmc2)ψ=(cα⃗⋅p⃗0−cA(α⃗⋅p⃗0)2+βmc2)ψ=Eψ(8)H\psi= \left(c\vec{\alpha}\cdot\vec{p}+\beta mc^2\right)\psi = \left(c\vec{\alpha}\cdot\vec{p}_0-cA\left(\vec{\alpha}\cdot\vec{p}_0\right)^2+\beta mc^2\right)\psi = E\psi \tag*{(8)}

that in two dimensions gives

(−iℏcσ1ddx0+Aℏ2cd2dx02+βmc2)ψ=Eψ.(9)\left(-\mathrm{i}\hbar c\sigma_1\frac{d}{dx_0}+A\hbar^2c\frac{d^2}{dx_0^2}+\beta mc^2\right)\psi=E\psi. \tag*{(9)}

This search for what we might call the “shadows of new physics” at our energy scales is fascinating. Nonetheless, not even one single experiment over the many performed over the decades has seen anything. Fortunately, due to the higher experimental control of condensed matter systems, it is now becoming increasingly popular to reproduce various aspects of fundamental physics in analog systems. Examples include the Hawking phenomenon in Bose–Einstein condensates [21], the Weyl symmetry [22, 23] and the related Hawking/Unruh phenomenon on graphene [24, 25, 26], gravitational and axial anomalies in Weyl semimetals [27], and more.

Henceforth, taking that road, we focus on the proposal of graphene as an analog of high-energy fundamental physics [22, 23, 24, 25, 26, 28, 29, 30, 31, 32], based on the fact that its low-energy excitations are massless Dirac pseudo-relativistic fermions (the matter fields ψ\psi), propagating in a carbon two-dimensional honeycomb lattice [33]. Such behaviour is shared by a wide range of materials, called “Dirac materials” [34].

Here we shall rely heavily on [35] and, to a lesser extent, on [36, 37], and shall take the chance to explain key details of the construction that are not explicitly exposed there.

Variables, Supervariables, Hypervariables

In this problem we are in two-space and one-time dimensions. We have three energy “layers”, each with its own appropriate phase-space variables, that we indicate as

(x,p),(X0,P0),(X,P).(10)(x,p),\quad(X_0,P_0),\quad(X,P). \tag*{(10)}

This notation is discussed at length in [35]. Let us say here that the variables (x,p)(x,p) are the actual phase-space variables of the lowest energy Dirac quasiparticles of graphene that originate from the O(ℓ)O(\ell) dispersion relations, with ℓ\ell the lattice spacing

E±=±vF∣p⃗∣,(11)E_{\pm}=\pm v_F\lvert\vec{p}\rvert, \tag*{(11)}

where vF=3/2 ∣η1∣ ℓ/ℏv_F = 3/2\,|\eta_1|\,\ell/\hbar, with η1≃−2.8 eV\eta_1 \simeq-2.8\,\mathrm{eV} the nearest neighbor hopping energy; see details in [33].

Such variables have a clear physical meaning, as coordinates and momenta of the low-energy conductivity electrons moving around (quantum mechanically) on the membrane.1

There we explain how to obtain the dispersion relations in the usual way (secular equation), based on the knowledge of the functions

Fm(k⃗)≡∑i=1nmeik⃗⋅s⃗i(m),(12)\mathcal{F}_m(\vec{k}) \equiv\sum_{i=1}^{n_m} e^{i\vec{k}\cdot\vec{s}^{(m)}_i}, \tag*{(12)}

encoding the relevant information on the geometry of the lattice relative to the mthm^{\mathrm{th}}-near neighbors. Other details on the derivation of the Fm\mathcal{F}_ms are in [36], where it is explained that the nmn_m vectors s⃗i(m)\vec{s}^{(m)}_i connect any given atom to its mthm^{\mathrm{th}}-near neighbors.

Using the fortuitous occurrence, F2=∣F1∣2−3\mathcal{F}_2 = |\mathcal{F}_1|^2 - 3, and various approximations explained in [35], we arrive at the dispersion relations

E±=η1(±∣F1∣−A~∣F1∣2),(13)E_{\pm} = \eta_1\left(\pm|\mathcal{F}_1|-\widetilde{A}|\mathcal{F}_1|^2\right), \tag*{(13)}

where A~≃0.15\widetilde{A} \simeq0.15.

It is then natural to define two further sets of variables, that we call “supervariables”, (X0,P0)(X_0,P_0), and “hypervariables”, (X,P)(X,P), as both sets refer to functions that appear in the dispersion relations that contain all orders2 in ℓ\ell. In other words (writing things in a dimensionless fashion, i.e., not including a necessary factor of 1/ℓ1/\ell)

∣F1∣∼p+O(ℓ2),∣F1∣∼P0,∣F1∣−A~∣F1∣2∼P.(14)|\mathcal{F}_1| \sim p + O(\ell^2), \qquad|\mathcal{F}_1| \sim P_0, \qquad|\mathcal{F}_1|-\widetilde{A}|\mathcal{F}_1|^2 \sim P. \tag*{(14)}

These relations are key in our analysis and are dictated by the phenomenology of the system.

In (13), if we define the supermomenta as

P⃗0=−ℏℓ(Re⁡F1,Im⁡F1),(15)\vec{P}_0 = -\frac{\hbar}{\ell}\left(\operatorname{Re}\mathcal{F}_1,\operatorname{Im}\mathcal{F}_1\right), \tag*{(15)}

then we can write

E±=VF(±∣P⃗0∣−A∣P⃗0∣2),(16)E_{\pm} = V_F\left(\pm|\vec{P}_0|-A|\vec{P}_0|^2\right), \tag*{(16)}

where VF=∣η1∣ ℓ/ℏV_F = |\eta_1|\,\ell/\hbar and A=ℓ/ℏ A~A = \ell/\hbar\,\widetilde{A}.

Notice that in the following, we shall use a slightly different definition of P⃗0\vec{P}_0, that allows one to expand it as P⃗0∼p⃗+⋯\vec{P}_0 \sim\vec{p}+\cdots, rather than P⃗0∼3/2p⃗+⋯\vec{P}_0 \sim3/2\vec{p}+\cdots. That is P⃗0≡2/3P⃗0\vec{P}_0 \equiv2/3\vec{P}_0, but, to simplify notation with an abuse, we shall write P⃗0→P⃗0\vec{P}_0 \to\vec{P}_0.

Let us now set ℏ=1\hbar=1, unless otherwise stated.

Just like in the literature it is customary to use a standard Dirac Hamiltonian, in relation to the linear dispersion relations (11), we can keep going with the same logic by building-up the appropriate Hamiltonian in terms of the supermomenta P⃗0\vec{P}_0, whose dispersion relations, descending from the secular equation

det⁡(H−E1)=0(17)\det(H-E\mathbf{1})=0 \tag*{(17)}

are indeed (16). To do that one must realize that to obtain, through (17), ∣P⃗0∣|\vec{P}_0| in the dispersion relation, a corresponding term σ⃗⋅P⃗0≡P0\vec{\sigma}\cdot\vec{P}_0\equiv P_0 needs be in the Hamiltonian: ∣P⃗0∣→P0|\vec{P}_0|\to P_0. Thus, the effective Hamiltonian we can use to describe the physics of the “superlayer” (X0,P0)(X_0,P_0) is

H(P0)=VF∑kψk†(P0−AP0P0)ψk=VF∑kψk†(σ⃗⋅P⃗0−A∣P⃗0∣212)ψk.(18)H(P_0)=V_F\sum_k\psi_k^\dagger(P_0-AP_0P_0)\psi_k=V_F\sum_k\psi_k^\dagger\left(\vec{\sigma}\cdot\vec{P}_0-A|\vec{P}_0|^2\mathbf{1}_2\right)\psi_k . \tag*{(18)}

Notice that we use the convention3 ψk†=(bk†,ak†)\psi_k^\dagger=(b_k^\dagger,a_k^\dagger), where aka_k and bkb_k are the annihilation operators of each graphene sublattice, see [22].

Indeed, by considering the 2×22\times2 structure of HH

H(P0)∼VF(−A∣P⃗0∣2P01−iP02P01+iP02−A∣P⃗0∣2),(19)H(P_0)\sim V_F\begin{pmatrix} -A|\vec{P}_0|^2 & P_0^1-iP_0^2\\ P_0^1+iP_0^2 & -A|\vec{P}_0|^2 \end{pmatrix}, \tag*{(19)}

clearly the two solutions of (17)

det⁡(H(P0)−E1)=0∼det⁡(−A∣P⃗0∣2−EP01−iP02P01+iP02−A∣P⃗0∣2−E)=0∼(A∣P⃗0∣2+E)2=∣P⃗0∣2,(20)\det(H(P_0)-E\mathbf{1})=0\sim\det\begin{pmatrix} -A|\vec{P}_0|^2-E & P_0^1-iP_0^2\\ P_0^1+iP_0^2 & -A|\vec{P}_0|^2-E \end{pmatrix}=0\sim(A|\vec{P}_0|^2+E)^2=|\vec{P}_0|^2, \tag*{(20)}

are those given in (13).

We can then move one last step up, in energy and symmetry (see [35]), by defining the hypermomenta

P⃗=P⃗0(1−A∣P⃗0∣),(21)\vec{P}=\vec{P}_0(1-A|\vec{P}_0|), \tag*{(21)}

that is nothing more than a consequence of the above mentioned fortuitous occurrence, F2=∣F1∣2−3\mathcal{F}_2=|\mathcal{F}_1|^2-3, hence a physical fact of this system; see (14) and discussion there.

With this definition we arrive at the most symmetric situation of all, that is a full linearization of the Hamiltonian

H(P)=VF∑kψk†Pψk(22)H(P)=V_F\sum_k\psi_k^\dagger P\psi_k \tag*{(22)}

with dispersion relations

E±=±VF∣P⃗∣(23)E_\pm=\pm V_F\lvert\vec{P}\rvert \tag*{(23)}

which is, in all respects but two, similar to (11). The two differences are crucial, though. First, the slope here is VFV_F not vF≃1.5VFv_F\simeq1.5V_F. Second, the hypermomenta are generalized variables, containing, in principle, all orders in ℓ\ell and pp (see [35] and later here).

To obtain H(P0)H(P_0) from H(P)H(P) one should first use the expansion (21), and then use the prescription ∣P⃗0∣→P0\lvert\vec{P}_0\rvert\to P_0

H(P(P0))=VF∑kψk†σ⃗⋅P⃗ ψk=VF∑kψk†σ⃗⋅(P⃗0(1−A∣P⃗0∣))ψk→VF∑kψk†(σ⃗⋅P⃗0−A(σ⃗⋅P⃗0)(σ⃗⋅P⃗0))ψk=VF∑kψk†(P⃗0−A∣P⃗0∣2)ψk=H(P0).(24)\begin{aligned} H(P(P_0))&=V_F\sum_k\psi_k^\dagger\vec{\sigma}\cdot\vec{P}\,\psi_k\\ &=V_F\sum_k\psi_k^\dagger\vec{\sigma}\cdot\left(\vec{P}_0(1-A\lvert\vec{P}_0\rvert)\right)\psi_k\\ &\to V_F\sum_k\psi_k^\dagger\left(\vec{\sigma}\cdot\vec{P}_0-A(\vec{\sigma}\cdot\vec{P}_0)(\vec{\sigma}\cdot\vec{P}_0)\right)\psi_k\\ &=V_F\sum_k\psi_k^\dagger\left(\vec{P}_0-A\lvert\vec{P}_0\rvert^2\right)\psi_k=H(P_0). \tag*{(24)} \end{aligned}

Now, let us move towards the measurable quantities of the lowest layer (x,p)(x,p). We first need to expand F1\mathcal{F}_1 in powers of ℓ\ell or pp. Indeed, this means, through (15), to expand P0P_0 in powers of ℓ\ell or pp, and through (21), to expand PP in powers of ℓ\ell or pp. The function in point is

F1=eiℓky(1+2ei32ℓkxcos⁡(32ℓkx)).(25)\mathcal{F}_1=e^{i\ell k_y}\left(1+2e^{i\frac{\sqrt{3}}{2}\ell k_x}\cos\left(\frac{\sqrt{3}}{2}\ell k_x\right)\right). \tag*{(25)}

This expansion, in general, needs be done around the two inequivalent Fermi points K±D=(±4π33ℓ,0)K_\pm^D=(\pm\frac{4\pi}{3\sqrt{3}\ell},0), located where F1=0\mathcal{F}_1=0. In our case, though, we are not considering topological defects or other physical set-ups that include, in their description, both Dirac points. For us it is sufficient to do all calculations near one Dirac point, e.g., if ki=K+iD+pik_i=K_{+i}^D+p_i.

Thus, the Hamiltonian H(P0)H(P_0) in (18) expressed in terms of p⃗\vec{p} up to O(ℓ3)O(\ell^3)4, is

H(P0(p))=H(p)=vF∑p⃗ψp⃗†[σ1(p1−ℓ4(p12−p22)−ℓ28p1(p12+p22))+σ2(p2+ℓ2p1p2−ℓ28p2(p12+p22))−32A((p12+p22)−ℓ2p13+3ℓ2p1p22)]ψp⃗.\begin{aligned} H(P_0(p))=H(p)=v_F\sum_{\vec{p}}\psi_{\vec{p}}^\dagger\Bigg[& \sigma_1\left(p_1-\frac{\ell}{4}(p_1^2-p_2^2)-\frac{\ell^2}{8}p_1(p_1^2+p_2^2)\right)\\ &+\sigma_2\left(p_2+\frac{\ell}{2}p_1p_2-\frac{\ell^2}{8}p_2(p_1^2+p_2^2)\right)\\ &-\frac{3}{2}A\left((p_1^2+p_2^2)-\frac{\ell}{2}p_1^3+\frac{3\ell}{2}p_1p_2^2\right)\Bigg]\psi_{\vec{p}}. \end{aligned}

For the sake of complete control of the units, let us now reintroduce ℏ\hbar, and let us write the dispersion relations for this Hamiltonian as

E±=vF[±∣p⃗∣∓ℓ4ℏ∣p⃗∣2cos⁡(3θ)+164ℓ2ℏ2∣p⃗∣3(7+cos⁡(6θ))−32A∣p⃗∣2+34Aℓℏ∣p⃗∣3cos⁡(3θ)](26)E_{\pm}=v_F\left[\pm\lvert\vec{p}\rvert\mp\frac{\ell}{4\hbar}\lvert\vec{p}\rvert^2\cos(3\theta)+\frac{1}{64}\frac{\ell^2}{\hbar^2}\lvert\vec{p}\rvert^3\left(7+\cos(6\theta)\right)-\frac{3}{2}A\lvert\vec{p}\rvert^2+\frac{3}{4}A\frac{\ell}{\hbar}\lvert\vec{p}\rvert^3\cos(3\theta)\right] \tag*{(26)}

where

A=0.15ℓℏ>0(27)A=0.15\frac{\ell}{\hbar}>0 \tag*{(27)}

tan⁡θ=p2/p1\tan\theta=p_2/p_1 and

vF≡32∣η1∣ℓℏ≃c/300≃106m/s.(28)v_F\equiv\frac{3}{2}\lvert\eta_1\rvert\frac{\ell}{\hbar}\simeq c/300\simeq10^6m/s. \tag*{(28)}

Let us give here another writing of the dispersion relations

E±=32∣η1∣[±(ℓℏ∣p⃗∣)+14(ℓℏ∣p⃗∣)2cos⁡(3θ)+164(ℓℏ∣p⃗∣)3(7+cos⁡(6θ))−320.15(ℓℏ∣p⃗∣)2+340.15(ℓℏ∣p⃗∣)3cos⁡(3θ)],(29)\begin{aligned} E_{\pm}={}&\frac{3}{2}\lvert\eta_1\rvert\left[\pm\left(\frac{\ell}{\hbar}\lvert\vec{p}\rvert\right)+\frac{1}{4}\left(\frac{\ell}{\hbar}\lvert\vec{p}\rvert\right)^2\cos(3\theta)+\frac{1}{64}\left(\frac{\ell}{\hbar}\lvert\vec{p}\rvert\right)^3\left(7+\cos(6\theta)\right)\right.\\ &\left.-\frac{3}{2}0.15\left(\frac{\ell}{\hbar}\lvert\vec{p}\rvert\right)^2+\frac{3}{4}0.15\left(\frac{\ell}{\hbar}\lvert\vec{p}\rvert\right)^3\cos(3\theta)\right], \tag*{(29)} \end{aligned}

which makes clear that the expansion parameter is ℓℏ∣p⃗∣\frac{\ell}{\hbar}\lvert\vec{p}\rvert.

Choice of the generalized coordinates and related GUPs

From the previous discussion, it is clear that the interpretation of P0iP_0^i and PiP^i as generalized momenta is suggested by the dispersion relations, hence the functional form of P0i(p)P_0^i(p)

P01=p1+ℓ4((p2)2−(p1)2)−ℓ28p1((p1)2+(p2)2),P_0^1=p^1+\frac{\ell}{4}\left((p^2)^2-(p^1)^2\right)-\frac{\ell^2}{8}p^1\left((p^1)^2+(p^2)^2\right),
P02=p2+ℓ2p1p2−ℓ28p2((p1)2+(p2)2).(30)P_0^2=p^2+\frac{\ell}{2}p^1p^2-\frac{\ell^2}{8}p^2\left((p^1)^2+(p^2)^2\right). \tag*{(30)}

or, equivalently

P01=p1−ℓ4∣p⃗∣2cos⁡2θ−ℓ28∣p⃗∣cos⁡θ,P_0^1=p^1-\frac{\ell}{4}\lvert\vec{p}\rvert^2\cos2\theta-\frac{\ell^2}{8}\lvert\vec{p}\rvert\cos\theta,
P02=p2+ℓ4∣p⃗∣2sin⁡2θ−ℓ28∣p⃗∣sin⁡θ,(31)P_0^2=p^2+\frac{\ell}{4}\lvert\vec{p}\rvert^2\sin2\theta-\frac{\ell^2}{8}\lvert\vec{p}\rvert\sin\theta, \tag*{(31)}

Hence of Pi(p)P^i(p), through (21), are dictated by the physics of the system. The form of the dispersion relations, in turn, tells us how the effective Hamiltonian must look like, with little or no freedom.

What is left to the interpretation is the meaning of the generalized coordinates, X0iX_0^i and XiX^i. Only when this is done the picture is complete, and we have all that is necessary, (H,X,P)(H,X,P), to describe the dynamics.

Although there is room for different interpretations for the generalized coordinates, there are two choices that are very natural, hence stand out as privileged:

  • The first choice is to keep xix^i as the generalized coordinates, xi=X0i=Xix^i = X^i_0 = X^i. This choice, as obvious from [x,P(p)]∼i∂pP[x, P(p)] \sim i\partial_p P, leads to some form of GUP.

  • The second choice is to demand the generalized coordinates to be canonical, [X0,P0]=i=[X,P][X_0, P_0] = i = [X, P]. This choice necessarily leads to both X0i(x,p)X^i_0(x,p) and Xi(x,p)X^i(x,p).

To summarize (see [35] for details): One can move from the standard (canonical) variables that describe the lowest energy physics of the Dirac quasi-particles of graphene, (x,p)(x,p), to the next “layer”, by including higher order contributions in pp to P0P_0, that we call supervariables, (X0,P0)(X_0,P_0). Then, considering that contributions of the same order could come by including next-to-near neighbors in the computations (A≠0A \ne0), one can move one more “layer” up, to the hypervariables (X,P)(X,P).

If, at every step, we only consider the two privileged choices for the generalized coordinates, the GUP and the canonical, the overall options we have are given in the following diagram [35].

Image not extracted

Figure 1.

It is important to notice that:

  • To include terms from O(p2)O(p^2) on, we have to keep A≠0A \ne0, as seen in (27). Therefore, the “layer of reference” is always the “hyperworld”, (X,P)(X,P).

  • The interesting set-ups are those leading to GUPs in the final layer, the one of reference. In the above scheme, these are (V) and (VII).

  • No matter the choice, all the quantities eventually need be expressed in terms of the measurable standard variables, (x,p)(x,p).

With this in mind, the GUP-corrected physics here is given by the Hamiltonian (26), and the related dispersion relations (26). We simply need to read-off all results in terms of the appropriate variables, that for the coordinates are:

  1. the case Xi=xiX^i=x^i (that is the ladder (I)→(III)→(VII)(I)\to(III)\to(VII)), leading to the GUP (VII)

[Xi,Pj]=iFik(P⃗)[δjk−A∣P⃗∣(δjk+PkPj∣P⃗∣2)−A2∣P⃗∣2(δjk+PkPj∣P⃗∣2)](32)[X^i,P_j]=iF^{ik}(\vec{P})\left[\delta^k_j-A|\vec{P}|\left(\delta^k_j+\frac{P^kP_j}{|\vec{P}|^2}\right)-A^2|\vec{P}|^2\left(\delta^k_j+\frac{P^kP_j}{|\vec{P}|^2}\right)\right] \tag*{(32)}

where

Fij(P⃗)=δij+12ℓ(−P1P2P2P1)+12ℓA∣P⃗∣(−P1P2P2P1)−12ℓ2((P1)2P1P2P1P2(P2)2),F^{ij}(\vec{P})=\delta^{ij}+\frac{1}{2}\ell\begin{pmatrix}-P^1 & P^2 \\ P^2 & P^1\end{pmatrix}+\frac{1}{2}\ell A|\vec{P}|\begin{pmatrix}-P^1 & P^2 \\ P^2 & P^1\end{pmatrix}-\frac{1}{2}\ell^2\begin{pmatrix}(P^1)^2 & P^1P^2 \\ P^1P^2 & (P^2)^2\end{pmatrix},

Notice that Fij=FjiF^{ij}=F^{ji}, and Fii(P⃗)=2−ℓ22∣P⃗∣2F^i_i(\vec{P})=2-\frac{\ell^2}{2}|\vec{P}|^2.

  1. and the case Xi=X0i(x,p)X^i=X^i_0(x,p) with

X01(x,p)=[1+ℓ2p1+ℓ28(5(p1)2+3(p2)2)]x1+[−ℓ2p2+ℓ24p1p2]x2,X02(x,p)=[−ℓ2p2+ℓ24p1p2]x1+[1−ℓ2p1+ℓ28(3(p1)2+5(p2)2)]x2.(33)\begin{aligned} X^1_0(x,p)&=\left[1+\frac{\ell}{2}p^1+\frac{\ell^2}{8}\left(5(p^1)^2+3(p^2)^2\right)\right]x^1+\left[-\frac{\ell}{2}p^2+\frac{\ell^2}{4}p^1p^2\right]x^2,\\ X^2_0(x,p)&=\left[-\frac{\ell}{2}p^2+\frac{\ell^2}{4}p^1p^2\right]x^1+\left[1-\frac{\ell}{2}p^1+\frac{\ell^2}{8}\left(3(p^1)^2+5(p^2)^2\right)\right]x^2. \tag*{(33)} \end{aligned}

That is the ladder (I)→(II)→(V)(I)\to(II)\to(V), leading to the GUP (V), an expression coinciding, at first order in AA, with the ADV-like GUP (see (7))

[Xi,Pj]=i[δij−A∣P⃗∣(δij+PiPj∣P⃗∣2)−A2∣P⃗∣2(δij+PiPj∣P⃗∣2)].(34)[X^i,P^j]=i\left[\delta^{ij}-A|\vec{P}|\left(\delta^{ij}+\frac{P^iP^j}{|\vec{P}|^2}\right)-A^2|\vec{P}|^2\left(\delta^{ij}+\frac{P^iP^j}{|\vec{P}|^2}\right)\right]. \tag*{(34)}

The other commutators are trivial.

One does not need any special or exotic setup to have the GUP all-the-way ladder at work. The only thing that is necessary is to include higher order terms in the dispersion relations, and then move to the next-to-near neighbors. The coordinates are always just xix^i.

Since the literature is full of proposals on how GUPs affect a variety of phenomena, see, e.g., [10, 11, 12, 18, 19, 20, 32] and since we have obtained here many such GUPs, some very similar to the ones descending from quantum gravity, then we are now in the position to prove some of those theoretical conjectures in the lab.

Notice that, in going from (I)(I) to (III)(III), we also have a GUP

[X0i,P0j]=iFij(P⃗0),[X0i,X0j]=0=[P0i,P0j],(35)[X_0^i,P_0^j] = iF^{ij}(\vec{P}_0), \qquad[X_0^i,X_0^j] = 0 = [P_0^i,P_0^j], \tag*{(35)}

with Fij(P⃗0)F^{ij}(\vec{P}_0) given in

Fij(P⃗0)=δij+12ℓ(−P01P02P02P01)−12ℓ2((P01)2P01P02P01P02(P02)2).F^{ij}(\vec{P}_0) = \delta^{ij} + \frac{1}{2}\ell \begin{pmatrix} -P_0^1 & P_0^2 \\ P_0^2 & P_0^1 \end{pmatrix} -\frac{1}{2}\ell^2 \begin{pmatrix} (P_0^1)^2 & P_0^1P_0^2 \\ P_0^1P_0^2 & (P_0^2)^2 \end{pmatrix}.

Since, for the reasons explained earlier, we shall always like to include the A≠0A \ne0 effects, we only consider this as an intermediate mathematical step.

Conclusions

We have discussed here how analogs of three kinds of GUPs can be obtained when higher order terms are considered in the dispersion relations of the conductivity electrons of a generic Dirac material.

An analog system cannot behave as the target system in all respects but only in certain specific conditions, for certain types of particles and for particular regimes of the interaction. In fact, one of the aspects we have to give up in an analog is the universality, which should be expected in the presence of a fundamental GUP.

For instance, in the “three graphene worlds” we have explored here, we only have at our disposal a limited set of all the possible particles and fields, compared to those that nature gives us. Essentially, in a Dirac material we have analogs of

  • massless Dirac spinors, ψ\psi, that are the π\pi-electrons quasiparticles;

  • the U(1) gauge fields, AμA_\mu, stemming from straining the material, that could be used to mimic electromagnetic interactions;

  • the nonabelian gauge fields, AμaA_\mu^a, due to the two copies of ψ\psi, one per Dirac point, often described as SU(2) gauge fields, that could be used to mimic internal symmetries;

  • the spin connection, ωμa\omega_\mu^a, that always takes care of the intrinsic curvature associated to disclination defects in the lattice, and may or may not include contorsion, κμa\kappa_\mu^a, that takes care of the intrinsic curvature associated to dislocation defects;

  • classical (and quantum) metric fields, gμνg_{\mu\nu}, that emerge as a collective description of the elastic and unelastic properties of the membrane.

Some more steps could be moved when we adapt (reduce) external fields, e.g., the true electromagnetic field, to the lower dimensional, and vFv_F-relativistic (rather than the cc-relativistic) and specific dynamics of the system. That is all.

Clearly, then, here we cannot have all types of particles and fields propagating in a GUP-compatible space, described by a suitable gμνg_{\mu\nu}, but we can do that for a limited set, for instance, ψ\psi propagating on a flat space with granular structure.

Even for this limited situation, since the recipe here consists in having control on the appropriate phase-space variables, we need to know how to recognize the “shadows of the high energy physics”, so to speak, on what mimics “the low energy GUP-corrected physics”.

With the results of [35], we are now able to face these issues because we now know the form of XiX^{i} and X0iX_{0}^{i}, in terms of xix^{i} and pip^{i}. With those results in our hands, we may safely solve the problem of lack of universality by, e.g., limiting the set-up to only Dirac quasi-particles ψ\psi and setting up a potential well, as many others do routinely in the same analog/effective language.

Such potential VV must be a function of the correct variables, XiX^{i}, i.e. V(X)V(X). But now we know X(x,p)X(x,p), and this simply means Xi=xiX_{i}=x_{i} in the most interesting case. Hence, the very same potential well V(x)V(x) one has for the GUP-non-corrected physics (the standard one) stays for GUP-corrected physics. In an experiment like that, we would prove that the GUP-corrected dynamics can indeed be practically realized in such a simple system, and one could check many theoretical statements in a lab.

To close, a note on the generality of our results. All condensed matter systems have corrections to linear or quadratic dispersion laws. Some kind of GUP may then appear in many systems after the suitable variables change. Notice that our approach may or may not lead to sensible analogs GUPs. In other words, it is not as simple as “any suitable change of variables leads to some sensible GUP”. For instance, in our case, we have seen that only by stopping at F2\mathcal{F}_{2} and making other approximations, it is possible to find GUPs related to the ADV GUPs. If one keeps going in the order of the expansion, no natural ADV or ADV-like structures are recovered. Other systems may have different natural GUPs, none at all, or unknown GUPs (like some of those that we discussed here).

To close, let us stress again that, of course, we did not prove here that graphene, or other materials of that kind, are systems where fundamental GUPs are at work. Nor we are suggesting that, following our procedure, fundamental GUPs can be spotted in other condensed matter systems. We are saying here that those easy-to-produce tabletop real systems, by a simple redefinition of the dynamical variables, can be turned into very efficient analogs of the fundamental scenarios that are supposed to be responsible for the GUPs that the high energy physics community is eagerly looking for. This seems promising, especially in the light of how simple our recipe is and of how difficult it is to run high energy experiments.

Acknowledgements

A. I. and P. P. sincerely thank Thomas Elze and the other organizers of DICE2022, for a wonderful and very productive conference. They gladly acknowledge the financial support of Charles University Research Center (UNCE/SCI/013). P. P. is also supported by Fondo Nacional de Desarrollo Científico y Tecnológico–Chile (Fondecyt Grant No. 3200725). S. M. acknowledges support from COST action CA18108.

References

References

  1. [1]Seiberg N and Witten E 1999 JHEP 1999 032 (Preprint hep-th/9908142)
  2. [2]Madore J, Schraml S, Schupp P and Wess J 2000 Eur. Phys. J. C16 161
  3. [3]Iorio A and Sýkora T 2002 Int. J. Mod. Phys. A17 2369
  4. [4]Iorio A 2008 Phys. Rev. D77 048701
  5. [5]Guralnik Z, Jackiw R, Pi S and Polychronakos A 2001 Phys. Lett. B517 450
  6. [6]Jackiw R 2002 Nucl. Phys. B Proc. Suppl. 108 30
  7. [7]Castorina P, Iorio A and Zappala D 2004 Phys. Rev. D69 065008
  8. [8]Colladay D and Kostelecky V A 1998 Phys. Rev. D 58 116002 (Preprint hep-ph/9809521)
  9. [9]Iorio A 2007 J. Phys. Conf. Ser. 67 012008
  10. [10]Maggiore M 1993 Phys. Lett. B 304 65 (Preprint hep-th/9301067)
  11. [11]Kempf A 1994 J. Math. Phys. (NY) 35 4483 (Preprint hep-th/9311147)
  12. [12]Scardigli F 1999 Phys. Lett. B 452 39 (Preprint hep-th/9904025)
  13. [13]Scardigli F 2022 Uncertainty principle and gravity (Preprint 2212.13776)
  14. [14]Das S and Vagenas E C 2008 Phys. Rev. Lett. 101(22) 221301
  15. [15]Ali A F, Das S and Vagenas E C 2011 Phys. Rev. D84 044013 (Preprint 1107.3164)
  16. [16]Ali A F, Das S and Vagenas E C 2009 Phys. Lett. B678 497
  17. [17]Ali A F, Das S and Vagenas E C 2010 Phys. Lett. B690 407
  18. [18]Buoninfante L, Luciano G G and Petruzziello L 2019 Eur. Phys. J. C79 663 (Preprint 1903.01382)
  19. [19]Petruzziello L and Illuminati F 2021 Nature Commun. 12 4449 (Preprint 2011.01255)
  20. [20]Bosso P and Luciano G G 2021 Eur. Phys. J. C 81 982 (Preprint 2109.15259)
  21. [21]Muñoz de Nova J R, Golubkov K, Kolobov V I and Steinhauer J 2019 Nature 569 688
  22. [22]Iorio A 2011 Ann. Phys. (Amsterdam) 326 1334
  23. [23]Iorio A 2102 Eur. Phys. J. Plus 127 156
  24. [24]Iorio A and Lambiase G 2012 Phys. Lett. B716 334
  25. [25]Iorio A and Lambiase G 2014 Phys. Rev. D 90 025006
  26. [26]Iorio A 2015 Int. J. Mod. Phys. D24 1530013
  27. [27]Gooth J, Niemann A C, Meng T, Grushin A G, Landsteiner K, Gotsmann B, Menges F, Schmidt M, Shekhar C, Süß V, Hühne R, Rellinghaus B, Felser C, Yan B and Nielsch K 2017 Nature 547 324–327
  28. [28]Iorio A and Pais P 2015 Phys. Rev. D 92 125005
  29. [29]Iorio A and Pais P 2022 Phys. Rev. B106 157401
  30. [30]Iorio A and Pais P 2018 Ann. Phys. (Amsterdam) 398 265
  31. [31]Iorio A 2015 Front. Mater. 1 36 (Preprint 1412.5160)
  32. [32]Iorio A, Lambiase G, Pais P and Scardigli F 2020 Phys. Rev. D 101(10) 105002
  33. [33]Castro Neto A H, Guinea F, Peres N M R, Novoselov K S and Geim A K 2009 Rev. Mod. Phys. 81 109
  34. [34]Wehling T, Black-Schaffer A and Balatsky A 2014 Adv. Phys. 63 1–76
  35. [35]Iorio A, Ivetić B, Mignemi S and Pais P 2022 Phys. Rev. D106 116011
  36. [36]Iorio A, Pais P, Elmashad I A, Ali A F, Faizal M and Abou-Salem L I 2018 Int. J. Mod. Phys. D27 1850080 (Preprint 1706.01332)
  37. [37]Iorio A and Pais P 2019 J. Phys. Conf. Ser. 1275 012061 (Preprint 1902.00116)

Paper details

vlnc:W_4rt653xq
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 1

downloads · 0

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

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents