betaFeedback

WorksPhysics and Astronomy

Optically- and thermally-driven huge lattice orbital and spin angular momenta from spinning fullerenes

Abstract

Lattice vibration in solids may carry angular momentum. But unlike the intrinsic spin of electrons, the lattice vibration is rarely rotational. To induce angular momentum, one needs to find a material that can accommodate a twisted normal mode, two orthogonal modes or excitation of magnons. If excitation is too strong, one may exceed the Lindemann limit, so the material melts. Therefore these methods are not ideal. Here, we theoretically propose a new route to phonon angular momentum in a molecular crystal C60\rm C_{60}. We find that a single laser pulse is able to inject a significant amount of angular momentum to C60\rm C_{60}, and the momentum transfer is helicity-dependent. Changing from right-circularly polarized light to left-circularly polarized light switches the direction of phonon angular momentum. On the ultrafast time scale, the orbital angular momentum change closely resembles the displacive excitation of coherent phonons, with a cosine-function dependence on time, different from the spin counterpart. Atomic displacements, even under strong laser excitation, remain far below the Lindemann criterion. Under thermal excitation, spinning C60\rm C_{60} even at room temperature generates a huge angular momentum close to several hundred ℏ\hbar. Our finding opens the door to a large group of fullerenes, from C60\rm C_{60}, C70_{70} to their endohedral derivatives, where angular momentum can be generated through light or temperature. This paves the way to the phononic control electronic spin and harvesting thermal energy through phonon angular momentum.

Nuclear vibration is ubiquitous and important to biological, chemical, and physical processes, ranging from photosynthesis[1], photoisomerization[2] to spin manipulation[3, 4, 5]. Microscopically, each material has its own vibrational normal modes. However, due to the symmetry constraint[6], normal modes rarely have a rotational eigenvector, so the phonon angular momentum (PAM) is feeble. The concept of phonon spin is not new[7, 8], and reemerged[9]. McLellan[10] investigated PAM in a rotationally invariant harmonic model. Interest in PAM is more recent[9, 11, 12, 13, 14]. It was shown that one may use an electric field[15] or temperature gradient[16] to control PAM. By coupling PAM to the electronic spin degrees of freedom, one can use PAM to control spin devices[17]. However, the main obstacle is how to generate PAM in the first place. It was predicted that one could introduce chiral phonons in monolayer molybdenum disulfide[18], but the real calculation[19] showed that the effect is tiny in transition metal dichalcogenides. The same thing is true for Pt3\mathrm{Pt}_{3} and Pt5\mathrm{Pt}_{5}[20], where the most of vibrational modes has nearly zero angular momentum, with the maximum close to 0.094(ℏ/2)0.094(\hbar/2). Experimentally, it was demonstrated that one could construct a superimposed vibrational state which consists of two orthogonal vibrational modes to generate an artificial circular phonon mode[21]. One may also couple the phonon to magnons in a magnetic material[22, 23, 24, 25] to induce spin angular momentum of phonons, but one must first have magnons. In general, these methods are not easy.

Molecular crystals are different. The weak bonding between neighboring sites render atom motifs rotating freely. Buckminsterfullerene C60\mathrm{C}_{60}[26] is a prime example. At room temperature, C60\mathrm{C}_{60} is spinning rapidly[27, 28, 29] around its equilibrium lattice positions. Spinning is temperature-dependent. As temperature cools down below 261 K, the rotation slows down, where 13C{}^{13}\mathrm{C} NMR spectra of solid C60\mathrm{C}_{60} show a significant broadening at a chemical shift of 143 ppm[27, 30, 31]. Its rotation can also be manipulated through substrates[32, 33], the tip of scanning tunneling microscope[34] and encapsulation inside carbon nanotubes[35]. C60\mathrm{C}_{60} rotation also affects its energy spectrum[36] and vibrational spectra[37]. C60\mathrm{C}_{60} and C70\mathrm{C}_{70} are not the only fullerenes that spin. Endohedral fullerenes[38], with additional atoms encapsulated inside fullerenes, offer additional kinds of rotational dynamics[39], with applications to organic photovoltaic devices[40] and quantum computing[41]. Recently, possible light-induced superconductivity was reported in K3C60\mathrm{K}_{3}\mathrm{C}_{60}[42]. It is conceivable that light can inject angular momentum into K3C60\mathrm{K}_{3}\mathrm{C}_{60}. These readily available molecular crystals represent a new frontier for phonon angular momentum investigation.

In this Letter, instead of relying on artificial rotational motion of atoms to search for phonon angular momentum, we start with a material that rotates in the first place at room temperature. Specifically, we employ C60\mathrm{C}_{60} as an example, and we show that a laser pulse can induce a significant orbital and spin angular momentum change. PAM depends on the laser helicity. Circularly polarized light injects more angular momentum to the system than linearly polarized light. Changing from right- to left-circularly polarized light switches angular momentum direction. The total angular momentum exactly follows the energy change, thus reaffirming that it is physical. The orbital and spin angular momenta depend on time differently. The orbital has a cosine dependence, very similar to displacive excitation of coherent phonons[43]. The lattice distortion of C60\mathrm{C}_{60} is small even under strong laser excitation, with the atom displacement far below the Lindemann criterion[44]. The global rotation of C60\mathrm{C}_{60} generates several hundred of ℏ\hbar. Our finding here represents a different direction for phonon angular momentum research, and is expected to motivate experimental and theoretical investigations.

We define the total angular momentum of C60\mathrm{C}_{60} as

J=∑iji(t)=∑iri(t)×pi(t),(1)\mathbf{J}=\sum_i\mathbf{j}_i(t)=\sum_i\mathbf{r}_i(t)\times\mathbf{p}_i(t), \tag*{(1)}

where pi(t)\mathbf{p}_i(t) is its momentum of atom ii at time tt and ri(t)\mathbf{r}_i(t) is its position vector. The reason why we use J\mathbf{J} instead of L\mathbf{L} will become clear below. Although we do not quantize the atomic vibration formally here, we still use phonons below. To find the initial coordinates {ri(0)}\{\mathbf{r}_i(0)\} of carbon atoms in C60\mathrm{C}_{60}, we use the spiral generation method proposed by Fowler and Manolopoulos[45] for the topological problem of fullerenes. Here, in brief, we first construct the adjacency matrix, and diagonalize it. The three lowest eigenvectors correspond to the xx, yy and zz coordinates of the carbon atoms, and then we properly rescale them according to the radius of C60\mathrm{C}_{60}.

To optimize the geometry, we employ the Su-Schrieffer-Heeger (SSH) model[46] often used in conjugated polymers, where only π\pi electrons are treated quantum mechanically, and the nuclear motions are described by the classical potentials[47] since the mass of carbon atoms is more than three orders of magnitude larger than that of the electron. Recently, the SSH model finds an important application in topological insulators[48]. The Hamiltonian reads[49, 50, 51]

H0=−∑⟨ij⟩,σtij(ci,σ†cj,σ+h.c.)+K12∑⟨i,j⟩(rij−d0)2+K22∑iδθi,p2+K32∑i(δθi,h,12+δθi,h,22),(2)H_0=-\sum_{\langle ij\rangle,\sigma}t_{ij}\left(c_{i,\sigma}^{\dagger}c_{j,\sigma}+\mathrm{h.c.}\right)+\frac{K_1}{2}\sum_{\langle i,j\rangle}(r_{ij}-d_0)^2+\frac{K_2}{2}\sum_i\delta\theta_{i,p}^2+\frac{K_3}{2}\sum_i\left(\delta\theta_{i,h,1}^2+\delta\theta_{i,h,2}^2\right), \tag*{(2)}

where ci,σ†c^{\dagger}_{i,\sigma} is the electron creation operator at site ii with spin σ(=↑↓)\sigma(=\uparrow\downarrow) [52] and the summation ⟨ij⟩\langle ij\rangle over i(j)i(j) runs from 1 to 60 with i≠ji\ne j. The first term on the right-hand side represents the electron hopping between nearest-neighbor atoms at positions ri\mathbf{r}_i and rj\mathbf{r}_j, tij=t0−α(∣ri−rj∣−d0)t_{ij}=t_0-\alpha(|\mathbf{r}_i-\mathbf{r}_j|-d_0), where rij=∣ri−rj∣r_{ij}=|\mathbf{r}_i-\mathbf{r}_j|, t0t_0 is the average hopping constant, and α\alpha is the electron-lattice coupling constant. The last three terms are the lattice stretching, pentagon-hexagon and hexagon-hexagon bending energies, respectively, (see further details in [53]). You et al.[54] parametrized the Hamiltonian by fitting the energy gap, bond lengths and 174 normal mode frequencies to their respective experimental values. They found that t0=1.91t_0=1.91 eV, α=5.0\alpha=5.0 eV/Å, K1=42K_1=42 eV/Å2^2, K2=8K_2=8 eV/rad2^2, K3=7K_3=7 eV/rad2^2 and d0=1.5532d_0=1.5532 Å. These parameters are consistent with the literature values[55, 56, 57, 58, 59, 60, 61, 62]. With You’s six parameters and with the electrons initially occupying 30 lowest energy levels, we optimize the structure. Figure 1(a) shows a two-dimensional structure of our optimized C60_{60}, where the zz axis is out of the page. The C60_{60} molecule has the highest IhI_h point symmetry. 60 carbon atoms form 90 bonds, 60 single bonds around 12 pentagons, and 30 single bonds shared by 20 hexagons (see Figure 1(a)). Our double bond length is 1.403 Å, and the single bond length is 1.443 Å, both of which indeed match the experimental values of 1.40±\pm0.015Å and 1.45±\pm0.015Å [63, 64].

C60 structure and energy level scheme

Figure 1. (a) C60_{60} structure. The empty circles denote sixty carbon atoms. The laser polarization (the vector) is in the xyxy plane (the front pentagon) and controlled by the helicity ϵ\epsilon. (b) Energy level scheme for C60_{60}. The double-arrow denotes the transition between the second lowest unoccupied molecular orbital (LUMO+1) and highest occupied molecular orbital (HOMO).

Figure 1(b) shows the energy spectrum with the degeneracy. The highest occupied molecular orbital (HOMO) is a HuH_u state, while the lowest unoccupied molecular orbital (LUMO) is a T1uT_{1u} state. The first dipole-allowed transition is between HOMO and LUMO+1, which is highlighted by the double arrow. All these features are fully consistent with the prior study[54]. The challenge is to fit 174 normal mode frequencies[65] with only six parameters. We find that the calculated frequencies within our theory slightly deviate from the experimental values[65] within 3%. For instance, our breathing mode Ag(1)A_g(1) frequency is 481 cm−1^{-1}, while the experimental one is between 492[28] and 496[65] cm−1^{-1}. Our pentagonal pinch mode Ag(2)A_g(2)’s frequency is 1477 cm−1^{-1}, while the experimental one is 1470[28, 65] cm−1^{-1}. Here the error is 0.5%. Because our Hamiltonian does not include high-order terms[66], these small differences are expected and do not affect our main conclusion of the paper. For this reason, we do not adjust those parameters.

As seen above, there are varieties of ways to induce angular momentum change. We perturb the system with an ultrafast laser pulse, whose interaction with C60_{60} is described by

HI=−e∑iσE(t)⋅ri niσ(3)H_I=-e\sum_{i\sigma}\mathbf{E}(t)\cdot\mathbf{r}_i\,n_{i\sigma} \tag*{(3)}

where E(t)\mathbf{E}(t) is the electric field of the laser and niσn_{i\sigma} is the electron number operator at site ii.

We choose a Gaussian pulse for E(t)\mathbf{E}(t): E(t)=A0exp⁡(−t2/τ2)[cos⁡(ωt)x^+ϵsin⁡(ωt)y^]/1+ϵ2\mathbf{E}(t)=A_0\exp(-t^2/\tau^2)[\cos(\omega t)\hat{x}+\epsilon\sin(\omega t)\hat{y}]/\sqrt{1+\epsilon^2}, where A0A_0, ω\omega, τ\tau, tt, ϵ\epsilon and τ\tau are the field amplitude, laser frequency, pulse duration or width, time, helicity and pulse duration, respectively. x^\hat{x} and y^\hat{y} are the unit vectors along the xx and yy axes, respectively. We ensure that A0A_0 is large enough, so the quantum effect of light is small and the classical treatment of our laser field is adequate. We treat the electrons quantum mechanically through the time-dependent Liouville equation[51, 52],

iℏ∂ρijσ∂t=⟨[H,ρijσ]⟩,(4)i\hbar\frac{\partial\rho_{ij}^{\sigma}}{\partial t}=\langle[H,\rho_{ij}^{\sigma}]\rangle, \tag*{(4)}

where H=H0+HIH=H_0+H_I, and ρijσ\rho_{ij}^{\sigma} is the density matrix. The lattice vibrations are described by the Newtonian equation. This classical treatment of nuclear motion is reasonable if the de Broglie wavelength is smaller than the system size, but if the de Broglie wavelength becomes comparable to the system size, then our approach is less accurate. We solve the coupled Liouville and Newtonian equations numerically.

We excite C60_{60} with a 60-fs laser pulse with electric field amplitude of 0.01 V/Å. The photon energy is 2.76 eV, which is tuned to be resonant with the transition between HOMO and LUMO+1 (see Figure 1b). Figure 2(a) displays the total energy (thin line), lattice kinetic energy (thick line) and lattice potential energy (thick dashed line) as a function of time. The curve around zero denotes the laser pulse. Upon laser excitation, energy enters the electronic system first. The peak in the total energy is mainly due to the electron energy. Through the electron-lattice interaction and laser excitation, the lattice starts to vibrate. Most of the lattice energy is in the potential energy, and the kinetic energy is very small. In the figure, we multiply the kinetic energy by 100. Carbon atoms move at a speed of 10−3−10−210^{-3}-10^{-2} Å/fs, or 100-1000 m/s, very typical for nuclear vibrations. The oscillation is due to the energy exchange between the electron and lattice subsystems (for details, see[53]). Since we want to check the energy conservation as well as vibrational oscillations, we do not dampen our system. We find that the oscillation has a period of 69.3 fs, which exactly matches the frequency of the breathing mode Ag_g(1)[67, 68]. Such a coherent phonon excitation has been observed experimentally[67] and in other systems as well[43]. The reason why this Ag_g(1) is excited strongly is due to our chosen laser parameters[50, 53]. Our interest is in the lattice angular momentum. Figure 2(b) shows the lattice angular momentum for three laser helicities (ϵ\epsilon): right (σ+\sigma^+), left (σ−\sigma^-) and linear (π\pi) pulses, with the polarization in the xyxy plane (see Figure 1(a)). There is a general trend as to how the angular momentum is transferred to the lattice. Upon laser excitation, the flow of angular momentum to the lattice occurs on the 100 fs time scale, which corresponds to the total energy absorbed. Although each atom still vibrates and exchanges its angular momentum with the rest of the atoms, the total angular momentum must remain constant and is conserved in the absence of an external field. This is rigorously reproduced in our calculation. The angular momentum shows a strong dependence on ϵ\epsilon. For σ−\sigma^{-}, the lattice angular momentum is along the +z+z axis, while for σ+\sigma^{+}, it is along the −z-z axis. The π\pi pulse induces a much weaker angular momentum (see the long-dashed line). This is expected because the angular momentum of light is directly related to the helicity of light. When it interacts with C60\mathrm{C}_{60}, the lattice angular momentum has the hallmark of incident light. Figure 2(c), (Figure 2d) and (Figure 2e) show the detailed change in JxJ_x, JyJ_y and JzJ_z as a function of helicity ϵ\epsilon. When ϵ=0\epsilon= 0, the laser polarization is linear along the xx axis. One can see that only JyJ_y differs from zero, because the atom moves mainly along the xx axis. Going from ϵ=0\epsilon= 0 to 1, our pulse changes from a linearly polarized pulse to elliptically, and finally to a circularly polarized pulse. Both JxJ_x and JzJ_z increase significantly as the atom also moves along the yy axis. JzJ_z is negative as expected. When we change ϵ\epsilon from 0 to -1, we have a left elliptically to circularly polarized pulse, so JzJ_z flips its sign, fully consistent with our finding in Figure 2(b).

Four-panel plots of energy change and angular momentum

Figure 2. (a) Total energy change as a function of time under a 60-fs σ+\sigma^{+} pulse. Here ϵ=1\epsilon= 1. The laser field amplitude is 0.01 V/A˚\mathring{\mathrm{A}}, and the photon energy is ℏω=2.76\hbar\omega= 2.76 eV. Thin solid, thick solid, and dashed lines denote the system total energy, lattice kinetic and potential energy, respectively. The lattice kinetic energy is multiplied by 100 to have an easy view. To demonstrate the energy conservation, the system is not damped. The red thin line around 0 is the laser electric field. The oscillation period in the lattice potential is 69.3 fs, which exactly matches the frequency of the Raman-active Ag(1)A_g(1) mode. (b) Total angular momentum J\mathbf{J} under σ+\sigma^{+} (solid line), σ−\sigma^{-} (dotted line) and π\pi pulses (dashed line). Only the strongest component of J\mathbf{J} is shown. Linearly polarized light transfers a small angular momentum. The left- and right-circularly polarized light transfer opposite angular momenta. All the pulses have the same laser parameters as (a). (c), (d) and (e) show the final xx, yy and zz components of J\mathbf{J} as a function of ellipticity ϵ\epsilon.

What is unknown or less familiar is whether and how the lattice angular momentum J\mathbf{J} can characterize a system. This is an important conceptual question because angular momentum is rarely used to characterize the dynamics of a system, except in atoms and some simple structures. Prior studies[9, 11, 12, 13, 14] are all based on this assumption. We compare J\mathbf{J} against the absorbed energy ΔE\Delta E into C60\mathrm{C}_{60}. When the laser photon energy is off resonance, the energy change follows the perturbative path and has a simple dependence on the laser field amplitude A0A_0[69], so the agreement between the angular momentum J\mathbf{J} and ΔE\Delta E could be coincidental. The harder question is what happens if the laser is on resonance. Figure 3(a) shows the energy absorbed as a function of laser field amplitude where the photon energy is tuned to the dipole-allowed transition between the HOMO and LUMO+1 states (Figure 1(b)). It is clear that the dependence on A0A_0 is highly nonlinear. When A0A_0 is far below 0.005 V/Å, the change is linear, but as far as it closes to 0.005 V/Å, it peaks, after which it starts to decrease. Because resonant excitation activates multiple real electronic excitation, this opens other channels and ΔE\Delta E decreases. Once A0A_0 exceeds 0.01 V/Å, ΔE\Delta E increases again, which is consistent with our prior study[49]. Figure 3(b) is our angular momentum. One can see that its change with A0A_0 matches the absorbed energy change, down to some minor details. This proves that J\mathbf{J} is physical, and it represents the acquired angular momentum from light.

Four-panel plots of absorbed energy, total angular momentum, and orbital and spin angular momenta

Figure 3. (a) Absorbed energy ΔE\Delta E as a function of laser field amplitude A0A_0. The photon energy and pulse duration are the same as Figure 2. (b) Total angular momentum as a function of laser field amplitude A0A_0, which matches the total energy change. This demonstrates that J\mathbf{J} is physical. (c) The orbital LzL_z and spin SzS_z angular momenta as a function of time. SzS_z is very small when ℏω=2.76\hbar\omega= 2.76 eV. In the figure, it is multiplied by 10. (d) The orbital LzL_z and spin SzS_z angular momenta as a function of time with ℏω=0.185\hbar\omega= 0.185 eV. SzS_z reaches a comparable magnitude as LzL_z. The dependence of SzS_z on time follows a sine function, while that of LzL_z follows a cosine function, very similar to DECP[43].

It has been proposed that if one rewrites ri(t)\mathbf{r}_i(t) as a sum of the equilibrium position ri(0)\mathbf{r}_i(0) and the displacement ui(t)\mathbf{u}_i(t), ri(t)=ri(0)+ui(t)\mathbf{r}_i(t) = \mathbf{r}_i(0) + \mathbf{u}_i(t), the angular momentum for atom ii can be separated into two terms:

ji(t)=ri(0)×pi(t)+ui(t)×pi(t)≡li(t)+si(t).(5)\mathbf{j}_i(t) = \mathbf{r}_i(0) \times\mathbf{p}_i(t) + \mathbf{u}_i(t) \times\mathbf{p}_i(t) \equiv\mathbf{l}_i(t) + \mathbf{s}_i(t). \tag*{(5)}

li(t)\mathbf{l}_i(t) is called the orbital angular momentum of atom ii, while si(t)\mathbf{s}_i(t) is called the spin angular momentum[3, 9, 10, 12] because si\mathbf{s}_i is independent of the initial position of an atom. By separating orbital from spin angular momentum, we can directly investigate the interplay between spin and orbital angular momenta on an ultrafast time scale.

Figure 3(c) compares the spin angular momentum with the orbital counterpart. We employ a 60-fs laser pulse with ℏω=2.76\hbar\omega= 2.76 eV and A0=0.01A_0 = 0.01 V/Å. We notice that the angular moment mainly enters the orbital part. From Eq. (5), we see that it is the momentum pi\mathbf{p}_i that is mainly responsible for this increase. SzS_z is very small and magnified by 10 in the figure. What is interesting is that the spin angular momentum does not follow the laser pulse, very different from the orbital angular momentum. The reason for this difference is straightforward. SzS_z depends on both ui(t)\mathbf{u}_i(t) and pi\mathbf{p}_i. Although the pulse peaks at 0 fs, the atom cannot follow instantaneously and needs time to respond. The behavior of LzL_z is like a cosine function, with its maximum at 0 fs, similar to displacive excitation of coherent phonons (DECP)[43]. But SzS_z is like a sine function. The sum of SzS_z and LzL_z is JzJ_z, which is constant after the laser field is gone because the system does not have a way to exchange angular momentum. This demonstrates that our definition of angular momentum through Eq. (5) is physically sound.

It has been a big challenge to induce phonon spin angular momentum in real materials[19, 21, 22]. We can increase spin angular momentum by tuning the laser photon energy to vibrations directly[17]. We choose ℏω=0.185\hbar\omega= 0.185 eV, or 44.7 THz, to be resonant with nuclear vibrations. The laser pulse duration is also 60 fs and the field amplitude is increased to 0.15 V/Å. We caution that our current treatment of nuclear vibration may not be adequate to compute the infrared spectrum since we do not include the nuclear dipole moment and the nuclear vibration is not treated quantum mechanically. Figure 3(d) shows that SzS_z reaches 0.5ℏ0.5\hbar. Under our laser field, the structure of C60\mathrm{C}_{60} is well maintained, far below the Lindemann criterion[44] for melting. This criterion has been a major obstacle for a prior study[19], where the atomic displacements for BaO and LiNbO3_3 can be as large as 0.10.1 Å. Therefore, an experimental test is difficult.

The top figure of Figure 4 shows atomic displacements projected on the xyxy plane for five atoms on the front pentagon. The scales for the xx and yy axes are increased by 100 for an easy view. One can see the displacement is well below the Lindemann criterion. This paves the way to experimental testing. What is even more interesting is that besides these internal rotations, under thermal excitation, the entire molecule spins rapidly, although along arbitrary axes at room temperature. The angular momentum is L=IΩL = I\Omega, where II is the moment of inertia of C60\mathrm{C}_{60} and Ω\Omega is the angular speed. Assuming C60\mathrm{C}_{60} spins along the zz axis through two pentagons (see Figure 1), the moment of inertia is Iz=m∑i=160(xi2+yi2)I_z = m\sum_{i=1}^{60}(x_i^2+y_i^2), where mm is the mass of carbon atom and xix_i and yiy_i are the coordinates of atom ii. Note that the moment of inertia is almost identical along other axes because C60\mathrm{C}_{60} is highly spherical. An early estimate puts Ω\Omega between 1.5/ps1.5/\mathrm{ps} and 2.8/ps2.8/\mathrm{ps}[70]. According to the latest theoretical estimate[29], its average angular frequency is 0.340.34 rad/ps. Since our present study only contains one single C60\mathrm{C}_{60} and has no intermolecular interaction between several fullerenes, we are unable to verify their average angular frequency. Instead, we use their number and find ∣J∣|J| for C60\mathrm{C}_{60} to be 322ℏ322\hbar. This angular momentum is from the global rotation of C60\mathrm{C}_{60}. Because our Hamiltonian in Eq. (2) only depends on the distance between neighboring atoms, the rapid global rotation of C60\mathrm{C}_{60} does not enter our calculation directly and only implicitly through the laser field polarization. However, due to the high symmetry of C60\mathrm{C}_{60}, different polarizations yield almost identical results. Therefore, the global rotation must be treated separately in our formalism. To see the thermal effect of the global rotation on the spin angular momentum, we set the initial velocities of the carbon atoms to kBT/m\sqrt{k_BT/m}, where kBk_B is the Boltzmann constant and mm is the carbon mass. We allow a random distribution of the initial velocity directions. We use four temperatures, T=0T = 0, 10, 150 and 300 K.

Traces of displacements and phonon spin angular momentum

Figure 4. (Top) Traces of displacements of five carbon atoms on the front pentagon in the earlier stage of laser excitation. The change is only 0.01 Å, which is well below the Lindemann criterion[44]. The arrow on the top left denotes the rotation of entire C60_{60}. (Bottom) Phonon spin angular momentum at (a) 0 K, (b) 10 K, (c) 150 K and (d) 300 K, where the solid, dotted and long-dashed lines denote SxS_x, SyS_y and SzS_z, respectively. Its magnitude becomes larger as temperature increases.

The results are shown in the bottom figures of Figure 4. Figure 4(a) is our result with 0 K, but with a small laser field amplitude of 0.01 V/Å. The solid, dotted and long-dashed lines denote SxS_x, SyS_y, and SzS_z, respectively, and are the same for the rest of the figures. We see that only SzS_z has a sizable value. But as we increase temperature to 10 K, we find that all the other components are increased to 0.5ℏ0.5 \hbar. The amplitude is comparable among all the three components. At 150 K, the spin increases to 5ℏ5\hbar. At room temperature of 300 K, Figure 4(d) shows the spin angular momentum increases to 10ℏ10\hbar. We also check whether our initial configuration of velocity direction affects our results. Among all the configurations we investigate, we do not see any qualitative difference. They should be observable experimentally. Spinning C60_{60} could be a natural phonon angular momentum generator. As a large group of other fullerenes are available and even its endohedral forms already exist, such as N@C60_{60} and P@C60_{60}, our finding opens a new route to phonon angular momentum. The future study should focus on how to harvest such a huge momentum for technological applications.

In conclusion, we have proposed a route to generate phonon orbital and spin angular momenta from C60_{60}, without resorting to more complicated synthetic methods[18, 19, 21, 22]. We show that a single laser pulse can transfer a significant amount of angular momentum to C60_{60}. The phonon angular momentum that C60_{60} receives shows a strong dependence on laser helicity. The circularly polarized light injects more momentum than the linearly polarized light. We demonstrate that the total angular momentum change follows the total energy change faithfully, thus establishing that even in the time domain PAM is a valid concept and can be used to characterize the system property. The orbital angular momentum is normally larger than the spin counterpart, and it also has a different dependence on time. Similar to DECP[43], the orbital has a cosine function dependence, while the spin has a sine function dependence. We find that the atomic displacement is spiral and its amplitude is far below the Lindemann criterion, which has been a big obstacle to a prior study[19]. The angular momentum of the entire C60_{60} reaches 322ℏ322\hbar at room temperature. Since endohedral fullerenes such as N@C60_{60} and P@C60_{60} are readily available, our study points out a large group of materials suitable for phonon angular momentum generation. It is expected that our finding will motivate experimental and theoretical investigations in this field and beyond.

Acknowledgments

This work was solely supported by the U.S. Department of Energy under Contract No. DE-FG02-06ER46304. Part of the work was done on Indiana State University’s high performance Quantum and Obsidian clusters. The research used resources of the National Energy Research Scientific Computing Center, which is supported by the Office of Science of the U.S. Department of Energy under Contract No. DE-AC02-05CH11231.

References

References

  1. [1]T. Brixner, J. Stenger, H. M. Vaswani, M. Cho, R. E. Blankenship, and G. R. Fleming, Two-dimensional spectroscopy of electronic couplings in photosynthesis, Nature 434, 625 (2005).
  2. [2]F. Gai, K. C. Hasson, J. C. McDonald, and P. A. Anfinrud, Chemical dynamics in proteins: the photoisomerization of retinal in bacteriorhodopsin, Science 279, 1886 (1998).
  3. [3]D. A. Garanin and E. M. Chudnovsky, Angular momentum in spin-phonon processes, Phys. Rev. B 92, 024421 (2015).DOI
  4. [4]M. Fähnle, Th. Tsatsoulis, C. Illg, M. Haag, B. Y. Müller, and L. Zhang, Ultrafast demagnetization after femtosecond laser pulses: Transfer of angular momentum from the electronic system to magnetoelastic spin-phonon modes, J. Supercond. Nov. Magn 30, 1381 (2017).
  5. [5]D. Zahn, F. Jakobs, Y. W. Windsor, H. Seiler, T. Vasileiadis, T. A. Butcher, Y. Qi, D. Engel, U. Atxitia, J. Vorberger and R. Ernstorfer, Lattice dynamics and ultrafast energy flow between electrons, spins, and phonons in a 3d ferromagnet, arXiv:2008.04611v1 (2020).DOI
  6. [6]S. Coh, Classification of materials with phonon angular momentum and microscopic origin of angular momentum, arXiv:1911.05064v2 (2019).DOI
  7. [7]S. V. Vonsovskii and M. S. Svirskii, Phonon spin, Sov. Phys. Solid State 3, 1568 (1962).
  8. [8]A. D. Levine, A note concerning the spin of the phonon, Nuovo Cimento 26, 190 (1962).
  9. [9]L. Zhang and Q. Niu, Angular momentum of phonons and the Einstein-de Haas effect, Phys. Rev. Lett. 112, 085503 (2014).DOI
  10. [10]A. G. McLellan, Angular momentum states for phonons and a rotationally invariant development of lattice dynamics, J. Phys. C: Solid State Phys. 21, 1177 (1988).DOI
  11. [11]J. J. Nakane and H. Kohno, Angular momentum of phonons and its application to single-spin relaxation, Phys. Rev. B 97, 174403 (2018).DOI
  12. [12]A. Rückriegel and R. A. Duine, Long-range phonon spin transport in ferromagnet-nonmagnetic insulator heterostructures, Phys. Rev. Lett. 124, 117201 (2020).DOI
  13. [13]S. Park and B.-J. Yang, Phonon angular momentum Hall effect, Nano Lett. 20, 7694 (2020).
  14. [14]S. Streib, Difference between angular momentum and pseudoangular momentum, Phys. Rev. B 103, L100409 (2021).
  15. [15]K. Moseni, R. Wilson, and S. Coh, Electric field control of phonon angular momentum in perovskite BaTiO₃, arXiv:2103.06316v1 (2021).DOI
  16. [16]M. Hamada, E. Minamitani, M. Hirayama, and S. Murakami, Phonon angular momentum induced by the temperature gradient, Phys. Rev. Lett. 121, 175301 (2018).DOI
  17. [17]A. Stupakiewicz, C. S. Davies, K. Szerencos, D. Afanasiev, K. S. Rabinovich, A. V. Boris, A. V. Boris, A. V. Kimel, and A. Kirilyuk, Ultrafast phononic switching of magnetization, Nat. Phys. 17, 489 (2021).
  18. [18]L. Zhang and Q. Niu, Chiral phonons at high-symmetry points in monolayer hexagonal lattices, Phys. Rev. Lett. 115, 115502 (2015).DOI
  19. [19]D. M. Juraschek and N. A. Spaldin, Orbital magnetic moments of phonons, Phys. Rev. Materials 3, 064405 (2019).
  20. [20]O. Bistoni, F. Mauri, and M. Calandra, Intrinsic vibrational angular momentum from nonadiabatic effects in noncollinear magnetic molecules, Phys. Rev. Lett. 126, 225703 (2021).DOI
  21. [21]T. F. Nova, A. Cartella, A. Cantaluppi, M. Först, D. Bossini, R. V. Mikhaylovskiy, A. V. Kimel, R. Merlin, and A. Cavalleri, An effective magnetic field from optically driven phonons, Nat. Phys. 13, 132 (2017).DOI
  22. [22]J. Holanda, D. S. Maior, A. Azevedo, and S. M. Rezende, Detecting the phonon spin in magnon–phonon conversion experiments, Nat. Phys. 14, 500 (2018).
  23. [23]S. Streib, H. Keshತgar, and G. E. W. Bauer, Damping of magnetization dynamics by phonon pumping, Phys. Rev. Lett. 121, 027202 (2018).
  24. [24]K. An et al., Coherent long-range transfer of angular momentum between magnon Kittel modes by phonons, Phys. Rev. B 101, 060407(R) (2020).DOI
  25. [25]A. Rückriegel, S. Streib, G. E. W. Bauer, and R. A. Duine, Angular momentum conservation and phonon spin in magnetic insulators, Phys. Rev. B 101 104402 (2020).DOI
  26. [26]H. W. Kroto, J. R. Heath, S. C. O’Brien, R. F. Curl, and R. E. Smalley, C₆₀: Buckminster-fullerene, Nature 318, 162 (1985).
  27. [27]R. D. Johnson, C. S. Yannoni, H. C. Dorn, J. R. Salem, and D. S. Bethune, C₆₀ rotation in the solid state: Dynamics of a faceted spherical top, Science 255, 1235 (1992).
  28. [28]M. S. Dresselhaus, G. Dresselhaus, and P. C. Eklund, Science of Fullerenes and Carbon Nanotubes, Academic Press, San Diego, California (1996).
  29. [29]A. M. Bubenchikhov, M. A. Bubenchikhov, D. V. Mamontov, and A. V. Lun-Fu, MD-simulation of fullerene rotations in molecular crystal fullerite, Crystals 9, 496 (2019).DOI
  30. [30]R. Tycko, R. C. Haddon, G. Dabbagh, S. H. Glarum, D. C. Douglass, and A. M. Mujtaba, Solid-state magnetic resonance spectroscopy of fullerenes, J. Phys. Chem. 95, 518 (1991).
  31. [31]R. Tycko, G. Dabbagh, R. M. Fleming, R. C. Haddon, A. V. Makhija, and S. M. Zahurak, Molecular dynamics and the phase transition in solid C₆₀, Phys. Rev. Lett. 67, 1886 (1991).
  32. [32]L.-L. Wang and H.-P. Cheng, Rotation, translation, charge transfer, and electronic structure of C₆₀ on Cu(111) surface, Phys. Rev. B 69, 045404 (2004).DOI
  33. [33]S. I. Bozhko, S. A. Krasnikov, O. Lübben, B. E. Murphy, K. Radican, V. N. Semenov, H. C. Wu, B. Bulfin, and I. Shvets, Rotational transitions in a C₆₀ monolayer on the WO₂/W(110) surface, Phys. Rev. B 84, 195412 (2011).
  34. [34]N. Neel, L. Limot, J. Kröger, and R. Berndt, Rotation of C₆₀ in a single-molecule contact, Phys. Rev. B 77, 125431 (2008).DOI
  35. [35]Y. Zou, B. Liu, L. Wang, D. Liu, S. Yu, P. Wang, T. Wang, M. Yao, Q. Li, B. Zou, T. Cui, G. Zou, T. Wagberg, B. Sundqvist, and H.-K. Mao, Rotational dynamics of confined C₆₀ from near-infrared Raman studies under high pressure, PNAS 106, 22135 (2009).
  36. [36]J. R. F. Lima, J. Brandao, M. M. Cunha, and F. Moraes, Effects of rotation in the energy spectrum of C₆₀, Eur. Phys. J. D 68, 94 (2014).DOI
  37. [37]R. Saito, G. Dresselhaus, and M. S. Dresselhaus, Hindered rotation of solid ¹²C₆₀ and ¹³C₆₀, Phys. Rev. B 50, 5680 (1994).
  38. [38]D. S. Bethune, R. D. Johnson, J. R. Salem, M. S. de Vries, and C. S. Yannoni, Atoms in carbon cages: the structure and properties of endohedral fullerenes, Nature 366, 123 (1993).DOI
  39. [39]J. Hernandez-Rojas, J. Breton, and J. M. G. Llorente, Rotational dynamics of endohedral C₆₀ fullerene complexes, J. Phys. Chem. Solids 58, 1689 (1997).DOI
  40. [40]R. B. Ross, C. M. Cardona, D. M. Guldi, S. G. Sankaranarayanan, M. O. Reese, N. Kopidakis, J. Peet, B. Walker, G. C. Bazan, E. Van Keuren, B. C. Holloway, and M. Drees, Endohedral fullerenes for organic photovoltaic devices, Nat. Mater. 8, 208 (2009).DOI
  41. [41]J. Twamley, Quantum-cellular-automata quantum computing with endohedral fullerenes, Phys. Rev. A 67, 052318 (2003).DOI
  42. [42]M. Mitrano, A. Cantaluppi, D. Nicoletti, S. Kaiser, A. Perucchi, S. Lupi, P. Di Pietro, D. Pontiroli, M. Riccò, S. R. Clark, D. Jaksch, and A. Cavalieri, Possible light-induced supercon- ductivity in K₃C₆₀ at high temperature, Nature 530, 461 (2016).
  43. [43]H. J. Zeiger, J. Vidal, T. K. Cheng, E. P. Ippen, G. Dresselhaus, and M. S. Dresselhaus, Theory for displacive excitation of coherent phonons, Phys. Rev. B 45, 768 (1992).
  44. [44]F. A. Lindemann, Über die Berechnung molekularer Eigenfrequenzen, Phys. Z. 11, 609 (1910).
  45. [45]P. W. Fowler and D. E. Manolopoulos, An Atlas of Fullerenes, Clarendon Press, Oxford (1995).
  46. [46]W. P. Su, J. R. Schrieffer, and A. J. Heeger, Solitons in polyacetylene, Phys. Rev. Lett. 42, 1698 (1979).
  47. [47]A. J. Heeger, S. Kivelson, J. R. Schrieffer, and W. P. Su, Solitons in conducting polymers, Rev. Mod. Phys. 60, 781 (1988).
  48. [48]J. K. Asboth, L. Oroszlany and A. Palyi, A Short course on topological insulators: Band structure and edge states in one and two dimensions, Lecture Notes in Physics, Vol. 919. Springer (2016).DOI
  49. [49]G. P. Zhang, X. Sun, and T. F. George, Laser-induced ultrafast dynamics in C₆₀, Phys. Rev. B 68, 165410 (2003).
  50. [50]G. P. Zhang and T. F. George, Controlling vibrational excitations in C₆₀ by laser pulse durations, Phys. Rev. Lett. 93, 147401 (2004).
  51. [51]G. P. Zhang, Optical high harmonic generations in C₆₀, Phys. Rev. Lett. 95, 047401 (2005).
  52. [52]G. P. Zhang, Hartree-Fock dynamical electron-correlation effects in C₆₀ after laser excitation, Phys. Rev. Lett. 91, 176801 (2003).
  53. [53]See Supplementary Material.
  54. [54]W. M. You, C. L. Wang, F. C. Zhang, and Z. B. Su, Application of a Su-Schrieffer-Heeger-like model to the intramolecular electron-phonon coupling in C₆₀ clusters, Phys. Rev. B 47, 4765 (1993).
  55. [55]G. B. Adams, J. B. Page, O. F. Sankey, K. Sinha, J. Menendez, and D. R. Huffman, First-principles quantum molecular-dynamics study of the vibrations of icosahedral C₆₀, Phys. Rev. B 44, 4052(R) (1991).
  56. [56]M. Menon and K. R. Subbaswamy, Universal Parameter Tight-Binding Molecular Dynamics: Application to C₆₀, Phys. Rev. Lett. 67, 3487 (1991).
  57. [57]J. L. Feldman, J. Q. Broughton, L. L. Boyer, D. E. Reich, and M. D. Kluge, Intramolecular-force-constant model for C₆₀, Phys. Rev. B 46, 12731 (1992).
  58. [58]J. C. R. Faulhaber, D. Y. K. Ko, and P. R. Briddon, Vibronic coupling in C₆₀ and C₆₀⁻³, Phys. Rev. B 48, 661 (1992).
  59. [59]B. Friedman, Infrared absorption of polarons in C₆₀, Phys. Rev. B 48, 17551 (1993).
  60. [60]K. Harigaya and S. Abe, Optical-absorption spectra in fullerenes C₆₀ and C₇₀: Effects of Coulomb interactions, lattice fluctuations, and anisotropy, Phys. Rev. B 49, 16746 (1994).
  61. [61]W.-Z. Wang, C.-L. Wang, Z.-B. Su, and L. Yu, Lattice relaxation study on self-trapped exciton and biexciton in neutral and charged fullerenes, Phys. Rev. Lett. 72, 3550 (1994).
  62. [62]J. Liu, S. Iwata, and B. Gu, Structural properties of the endohedral complex Na⁺@C₆₀, J. Phys.: Condens. Matter 6, L253 (1994).
  63. [63]K. Hedberg, L. Hedberg, D. S. Bethune, C. A. Brown, H. C. Dorn, R. D. Johnson, and M. de Vries, Bond lengths in free molecules of Buckminsterfullerene, C₆₀, from gas-phase electron diffraction, Science 254, 410 (1991).
  64. [64]C. S. Yannoni, P. P. Bernier, D. S. Bethune, G. Meijer and J. R. Salem. NMR determination of the bond lengths in C₆₀, J. Am. Chem. Soc. 113, 3190 (1991).
  65. [65]J. Menendez and J. B. Page, Vibrational spectroscopy of C₆₀, in Light Scattering in Solids VIII Fullerenes, Semiconductor Surfaces, Coherent Phonons, edited by M. Cardona and G. Güntherodt, Springer (2000).
  66. [66]Z. C. Wu, D. A. Jelski, and T. F. George, Vibrational motions of Buckminsterfullerene, Chem. Phys. Lett. 137, 291 (1987).DOI
  67. [67]S. L. Dexheimer, D. M. Mittleman, R. W. Schoenlein, W. Vareka, X.-D. Xiang, A. Zettl, and C. V. Shank, Ultrafast Dynamics of Photoexcited C₆₀, in Ultrafast Pulse Generation and Spectroscopy, edited by T. R. Gosnell, A. J. Taylor, K. A. Nelson, and M. C. Downer, SPIE Proc. 1861, 328 (1993).
  68. [68]G. P. Zhang and T. F. George, Normal-mode selectivity in ultrafast Raman excitations in C₆₀, Phys. Rev. B 73, 035422 (2006).DOI
  69. [69]Y. R. Shen, The Principles of Nonlinear Optics, John Wiley & Sons, Inc., Hoboken, New Jersey (2003).
  70. [70]Q.-M. Zhang, J.-Y. Yi, and J. Bernholc, Structure and dynamics of solid C₆₀, Phys. Rev. Lett. 66, 2633 (1991).

Paper details

vlnc:W_efxfscsy
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.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents