betaFeedback

WorksPhysics and Astronomy

Nanorod Pair Complexes Manipulated via Magnetic Casimir Forces

Abstract

Controlling nanoscale interactions to suppress aggregation from short-range attractive forces is a key problem in nanoengineering. Here, we demonstrate a route to modulate Casmir-Lifshitz interactions between anisotropic nanoparticles with the magnetic fluids. By semi-classical quantum electrodynamics, we study ground state dispersion forces for cylindrical dielectric nanorods made of polystyrene (PS), and zinc oxide (ZnO) embedded in toluene-based host media with gold-coated magnetite nanoparticles and also predict magnetic contributions to the non-retarded excited state interaction. The variation in magnetic permeability enables tuning between repulsive and attractive interaction and a thermally unstable and measurable magnetic Casimir traps are predicted between a pair of ZnO-PS nanoparticles whose equilibrium position can be modulated over an order of magnitude with a small variation in the size of the magnetite nanoparticle. This provides an alternative magnetic Casimir-effect pathway to reversibly tune quantum electromagnetic forces at the nanoscale for assembly and enhancement of colloidal stability.

What if the very surrounding media in space itself could be engineered into a tool for nano-manipulation? For instance, can we take the background media as a magnetic fluid, in which electromagnetic fluctuations exert forces that can pull, push or most interestingly trap matter at the nanoscale level? Such fluctuation-induced interactions, which are referred to as van der Waals [1, 2] (vdW) and Casimir-Lifshitz (CL) [3, 4] forces together, typically act to bring microscopic bodies into contact, leading to unwanted self-aggregation in nanofabrication and soft-matter systems. [5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15] An important practical issue in nano and micro-devices is to overcome close contact attractions that causes building blocks to cluster together and an experimental solution is to engineer systems where the interaction turns repulsive at small distances [7, 8, 9, 10, 11, 12, 14, 16] Beyond simply mitigating stiction, these fluctuations could also be harnessed to produce stable and controllable interaction potentials. Fundamental investigations of the influence on CL forces by a magnetic surrounding media have been largely restricted to simple geometries. [17, 18, 19] Exploiting such a foundation, we have recently expanded the theoretical approach by including the contributions of transverse magnetic (TM) and transverse electric (TE) modes within a model described by the proximity force theorem [20, 21] in such a way that both energy contributions due to TM and TE modes would partially act as candidates to enable quantum trapping between two quasi-planar surfaces. [22]

In this letter, we go beyond these simple geometries and study an experimentally feasible setup which comprises of pairs of cylindrical nanorods made of polystyrene (PS) and zinc oxide (ZnO). The use of the full cylindrical geometry permits a more realistic description of nanoscale behavior, albeit at the price of considerable added theoretical complexity. This stimulates further development of analytical and numerical studies of Lifshitz and vdW interaction between elongated nanoparticles. [23, 24, 25, 26, 27, 28] One of us previously derived a few asymptotic limits for this kind of a system. [26] Here, we examine the magnetic contributions associated with a surrounding ferrofluid for the long-range and high-temperature limits of ground-state dispersion interactions between cylindrical nanoparticles. We also predict a new kind of magnetic contribution to the retarded excited-state interaction. In an illustrative example, we study the CL interaction between a ZnO nanorod and a PS nanorod suspended in a toluene and gold-coated magnetite nanoparticle (5-7%) fluid mixtures. In the case of two unequal nanorods, we find that subtle change of nanoparticle concentration into the ferrofluid can induce a transition from short-range repulsion to long-range attraction between them. It is the fluid’s magnetic permeability and its dependence upon the magnetite nanoparticle size that dictates both interaction strength and sign, in contrast to the fluid dielectric function which remains nearly constant. Such a device could enable controlled, magnetic-induced entropic CL trapping. Early studies of CL interactions between magnetic fluida [18, 19] were mainly concentrated on the observation of a repulsive force by changing the optical properties of the medium. By contrast, we have implemented here a way in which this zero-frequency, purely entropically driven Casimir trapping is realized exclusively by tuning the fluid’s magnetic permittivity, an effect not investigated earlier.

To lay foundations for such a framework, we develop, within the semi-classical quantum electrodynamics, the general theory for ground and excited state dispersion interactions for dielectric cylindrical nanorods embedded in a magnetic fluid. In the subsequent sections, we present analytical results for the excited state interaction between thin cylinders and the corresponding free energy, followed by numerical analysis of the ground-state forces. We briefly describe the essential optical and magnetic properties of the interacting nanorods and background fluid in the main text and at length in the Supplementary Information. Lastly we demonstrate here that the retardation substantially diminishes the overall interaction strength as well as amplifies the relative weight of magnetic response from the long-range, classical entropic contributions.

Here, we thus propose a novel experimentally feasible approach to manipulate nanorod pair complexes by suspending them in a magnetically diluted fluid, where the interplay between repulsive non-magnetic and attractive magnetic contributions to the CL force can be continuously tuned to form magnetic CL traps appearing as equilibria whose strength and range derive not from mechanical considerations but from spectral properties of quantum electromagnetic fluctuations.

We consider a system of two parallel cylindrical objects of radii aa with dielectric constant and magnetic permittivity, (ε1,μ1)(\varepsilon_1,\mu_1) and (ε2,μ2)(\varepsilon_2,\mu_2), respectively, embedded in a ferrofluid of dielectric constant ε3\varepsilon_3 and permittivity μ3\mu_3 shown in Figure 1. The cylinders are separated by a center-to-center distance RR. We assume that the cylinders are infinitely long in the zz direction, and we will work in the quasi-static limit where the wavelength of light is much larger than any other length scale in the problem. The dispersion interaction in this system is obtained using scattering approach, [26] where the interaction energy per unit length (for a cylinder with length LL) in finite temperature TT is (an explicit derivation is given in Sec. (S.2) of the Supplementary Information),

Image not extracted

Figure 1. Color online: Two cylindrical nanorods, one composed of ZnO with permittivity ε1\varepsilon_1 and permeability μ1\mu_1, and the other of PS with permittivity ε2\varepsilon_2 and permeability μ2\mu_2, each having an identical radius aa, are immersed in a magnetic liquid. The magnetic liquid is composed of toluene and it contains dispersed gold coated (thickness RAuR_{\mathrm{Au}}) magnetite nanoparticles of diameter DD and volume fraction ϕ\phi and it has effective permittivity ε3\varepsilon_3 and permeability μ3\mu_3. The nanorods are separated by a center-to-center distance RR.

G(a,R,T)=kBTπ∑p=0∞′∫0∞dk ln⁡D(iξp)(1)G(a,R,T)=\frac{k_{\mathrm{B}}T}{\pi}\sum_{p=0}^{\infty}{}^{\prime}\int_{0}^{\infty}dk\,\ln\mathcal{D}(i\xi_p) \tag*{(1)}

where the prime in the summation means that the term for which p=0p=0 must be weighted by 1/21/2. The above expression is given in Matsubara frequencies ξp=2πkBTp/ℏ\xi_p=2\pi k_{\mathrm{B}}Tp/\hbar, where pp is an integer. Here, D(iξp)\mathcal{D}(i\xi_p) is expressed as the determinant of the scattering matrix of the interacting cylinders. In general, such a matrix is infinite, but significant simplifications can be made in the “thin cylinder” approximation where the separation distance between the cylinders is much larger than their radii (R≫aR\gg a). In this case, terms containing Bessel functions[29] with n≥2n\ge2 make negligible contributions (details in Sec. (S.2) in the Supplementary Information) as they rapidly decay with RR. Thus, we find that (see Sec. (S.2) of the Supplementary Information),

ln⁡D(iξp)=f0K02(ν3R)+f1K12(ν3R)+f2(K02(ν3R)+K22(ν3R))(2)\ln\mathcal{D}(i\xi_p)=f_0K_0^2(\nu_3R)+f_1K_1^2(\nu_3R)+f_2\left(K_0^2(\nu_3R)+K_2^2(\nu_3R)\right) \tag*{(2)}

where K0K_0, K1K_1 and K2K_2 are modified Bessel functions of zeroth, first and second orders respectively, να2(iξp)=ξp2c2μα(iξp)εα(iξp)+k2\nu_\alpha^2(i\xi_p)=\frac{\xi_p^2}{c^2}\mu_\alpha(i\xi_p)\varepsilon_\alpha(i\xi_p)+k^2 for α=1,2,3\alpha=1,2,3, and the coefficients are as follows:

f0=−a44[(ε3−ε1)(ε3−ε2)ε32+(μ3−μ1)(μ3−μ2)μ32]ν34(3)f_0=-\frac{a^4}{4}\left[\frac{(\varepsilon_3-\varepsilon_1)(\varepsilon_3-\varepsilon_2)}{\varepsilon_3^2}+\frac{(\mu_3-\mu_1)(\mu_3-\mu_2)}{\mu_3^2}\right]\nu_3^4 \tag*{(3)}
f1=−a4{ν12ν34(ε3+ε1)(μ3+μ1)[ε3−ε2ε3(ε3μ3ν32−ε1(μ1+μ3)+ε3(μ3−μ1)2ν12)+μ3−μ2μ3(ε3μ3ν32−ε3(μ1+μ3)+ε1(μ1−μ3)2ν12)]+ν22ν34(ε3+ε2)(μ3+μ2)[ε3−ε1ε3(ε3μ3ν32−ε2(μ2+μ3)+ε3(μ3−μ2)2ν22)+μ3−μ1μ3(ε3μ3ν32−ε3(μ2+μ3)+ε2(μ2−μ3)2ν22)]}(4)\begin{aligned} f_1={}&-a^4\left\{\frac{\nu_1^2\nu_3^4}{(\varepsilon_3+\varepsilon_1)(\mu_3+\mu_1)} \left[\frac{\varepsilon_3-\varepsilon_2}{\varepsilon_3}\left(\frac{\varepsilon_3\mu_3}{\nu_3^2}-\frac{\varepsilon_1(\mu_1+\mu_3)+\varepsilon_3(\mu_3-\mu_1)}{2\nu_1^2}\right)\right.\right.\\ &\left.\left.+\frac{\mu_3-\mu_2}{\mu_3}\left(\frac{\varepsilon_3\mu_3}{\nu_3^2}-\frac{\varepsilon_3(\mu_1+\mu_3)+\varepsilon_1(\mu_1-\mu_3)}{2\nu_1^2}\right)\right]\right.\\ &\left.+\frac{\nu_2^2\nu_3^4}{(\varepsilon_3+\varepsilon_2)(\mu_3+\mu_2)} \left[\frac{\varepsilon_3-\varepsilon_1}{\varepsilon_3}\left(\frac{\varepsilon_3\mu_3}{\nu_3^2}-\frac{\varepsilon_2(\mu_2+\mu_3)+\varepsilon_3(\mu_3-\mu_2)}{2\nu_2^2}\right)\right.\right.\\ &\left.\left.\qquad+\frac{\mu_3-\mu_1}{\mu_3}\left(\frac{\varepsilon_3\mu_3}{\nu_3^2}-\frac{\varepsilon_3(\mu_2+\mu_3)+\varepsilon_2(\mu_2-\mu_3)}{2\nu_2^2}\right)\right]\right\} \tag*{(4)} \end{aligned}
f2=−a4ν12ν22ν34(ε3+ε1)(ε3+ε2)(μ3+μ1)(μ3+μ2)[(ε3μ3ν32−ε1(μ1+μ3)+ε3(μ3−μ1)2ν12)(ε3μ3ν32−ε2(μ2+μ3)+ε3(μ3−μ2)2ν22)+(ε3μ3ν32−ε3μ3ν12)(ε3μ3ν32−ε2μ2ν22)+(ε3μ3ν32−ε3μ3ν22)(ε3μ3ν32−ε1μ1ν12)+(ε3μ3ν32−ε3(μ1+μ3)+ε1(μ1−μ3)2ν12)(ε3μ3ν32−ε3(μ2+μ3)+ε2(μ2−μ3)2ν22)](5)f_2=-\frac{a^4\nu_1^2\nu_2^2\nu_3^4}{(\varepsilon_3+\varepsilon_1)(\varepsilon_3+\varepsilon_2)(\mu_3+\mu_1)(\mu_3+\mu_2)} \left[ \left(\frac{\varepsilon_3\mu_3}{\nu_3^2}-\frac{\varepsilon_1(\mu_1+\mu_3)+\varepsilon_3(\mu_3-\mu_1)}{2\nu_1^2}\right) \left(\frac{\varepsilon_3\mu_3}{\nu_3^2}-\frac{\varepsilon_2(\mu_2+\mu_3)+\varepsilon_3(\mu_3-\mu_2)}{2\nu_2^2}\right) +\left(\frac{\varepsilon_3\mu_3}{\nu_3^2}-\frac{\varepsilon_3\mu_3}{\nu_1^2}\right) \left(\frac{\varepsilon_3\mu_3}{\nu_3^2}-\frac{\varepsilon_2\mu_2}{\nu_2^2}\right) +\left(\frac{\varepsilon_3\mu_3}{\nu_3^2}-\frac{\varepsilon_3\mu_3}{\nu_2^2}\right) \left(\frac{\varepsilon_3\mu_3}{\nu_3^2}-\frac{\varepsilon_1\mu_1}{\nu_1^2}\right) +\left(\frac{\varepsilon_3\mu_3}{\nu_3^2}-\frac{\varepsilon_3(\mu_1+\mu_3)+\varepsilon_1(\mu_1-\mu_3)}{2\nu_1^2}\right) \left(\frac{\varepsilon_3\mu_3}{\nu_3^2}-\frac{\varepsilon_3(\mu_2+\mu_3)+\varepsilon_2(\mu_2-\mu_3)}{2\nu_2^2}\right) \right] \tag*{(5)}

where cc is the speed of light in vacuum.

Here we are also interested if a dispersion interaction in excited state could be realized between two equal cylindrical nanoparticles. The interaction energy is due to the separation dependence of the excited energy, such that Eres(R)=ℏ[ωr(R)−ωr(∞)]E^{\mathrm{res}}(R)=\hbar[\omega_r(R)-\omega_r(\infty)], where ωr(R)\omega_r(R) represents the resonance frequency at interatomic separation RR. The interaction energy of the excited state is, [30, 31] Eres(R)=s2T(R∣ωj)∝R−3E^{\mathrm{res}}(R)=s^2T(R|\omega_j)\propto R^{-3}, where ss is the magnitude of the transition dipole moment of the excitation and T(R∣ωj)T(R|\omega_j) is the dipole–dipole coupling tensor (from the electromagnetic Green’s function) at frequency ωj\omega_j which accounts for how field from one oscillating dipole influences another at distance RR and it typically scales as R−3R^{-3} in the non-retarded limit. This can be compared with the ground state van der Waals interaction which decays with distance as R−6R^{-6}. [32]

This analogy implies that two identical cylindrical nanoparticles can similarly form a long lasting antisymmetric excited state by virtue of a long-range resonant dispersion interaction. A corresponding expression for the retarded excited-state interaction energy, obtained semi-classically in Sec. (S.3) of the Supplementary Information, reads

Gres(a,R,T)=−2kBTπ∑p=0∞∫0∞dk (t0+t1+t2+t3)K0(ν3R)(6)G^{\mathrm{res}}(a,R,T)=-\frac{2k_{\mathrm{B}}T}{\pi}\sum_{p=0}^{\infty}\int_{0}^{\infty}dk\,(t_0+t_1+t_2+t_3)K_0(\nu_3R) \tag*{(6)}

where

t0=[ε3−μ12ε3]a2ν32;t1=[μ3−ε12μ3]a2ν32t3=[−ε1(μ1+μ3)+ε3(μ3−μ1)ν12+2ε3μ3ν32]a2ν12ν32(ε1+ε3)(μ1+μ3)t4=[−ε3(μ1+μ3)+ε1(μ1−μ3)ν12+2ε3μ3ν32]a2ν12ν32(ε1+ε3)(μ1+μ3)(7)\begin{aligned} t_0 &= \left[\frac{\varepsilon_3-\mu_1}{2\varepsilon_3}\right]a^2\nu_3^2; \qquad t_1 = \left[\frac{\mu_3-\varepsilon_1}{2\mu_3}\right]a^2\nu_3^2 \\ t_3 &= \left[-\frac{\varepsilon_1(\mu_1+\mu_3)+\varepsilon_3(\mu_3-\mu_1)}{\nu_1^2}+\frac{2\varepsilon_3\mu_3}{\nu_3^2}\right]\frac{a^2\nu_1^2\nu_3^2}{(\varepsilon_1+\varepsilon_3)(\mu_1+\mu_3)} \\ t_4 &= \left[-\frac{\varepsilon_3(\mu_1+\mu_3)+\varepsilon_1(\mu_1-\mu_3)}{\nu_1^2}+\frac{2\varepsilon_3\mu_3}{\nu_3^2}\right]\frac{a^2\nu_1^2\nu_3^2}{(\varepsilon_1+\varepsilon_3)(\mu_1+\mu_3)} \tag*{(7)} \end{aligned}

The entropic contribution (p=0p = 0) suggests that Eq. (6) provides R−3R^{-3} dependence that is much longer ranged than the ground state interaction between two equal thin cylindrical nanoparticles that is proportional to R−5R^{-5}. [33] It gives both magnetic and non-magnetic contributions. Notably, this previously neglected magnetic contribution in Eq. (6) has the same power law as its non-magnetic contribution. However, the significance of this contribution should be judged by the low temperature/short distance non-magnetic excited state interaction. The non-retarded excited state interaction becomes attenuated by retardation due to the finite velocity of light. In the retarded limit additional effects potentially impact the interaction via real photon exchange. Ultimately, in the limit of large distances or high temperatures, the interaction is expected to be dominated by the non-magnetic and magnetic zero frequency contributions [for more detail see Sec. (S.7) in the Supplementary Information]. However, it is important to emphasize that when the interaction occurs in a fluid the excited state lifetime will be significantly reduced. This will impact the potential existence of excited interactions in a fluid at large separations.

The nanoparticles we consider have large enough dimensions to approximate their effective dielectric functions with those of bulk materials. The dielectric functions at imaginary Matsubara frequencies are obtained through the Kramers–Kronig relation

ε(iξp)=1+2π∫0∞dω ωε′′(ω)ω2+ξp2.(8)\varepsilon(i\xi_p)=1+\frac{2}{\pi}\int_{0}^{\infty}d\omega\,\frac{\omega\varepsilon''(\omega)}{\omega^2+\xi_p^2}. \tag*{(8)}

The geometries of the nanoparticles and nanorods are described within our theoretical model, whereas the material properties are macroscopic attributes. We use available roomtemperature experimental data for PS as provided by van Zwol, et al. [34] For ZnO we calculate the imaginary part of the dielectric function within the density functional theory (DFT) as implemented in the Vienna Ab initio Simulation Package. [35] To correctly mimic the experimental band-gap energy of 3.4 eV, we employ the hybrid functional with 37.5% Fock exact exchange, which is a widely adopted parameter choice for this oxide. The GWGW-type pseudopotentials ensure better accuracy in describing high-frequency optical transitions. Convergence with respect to kk-space mesh is considered. [36] Further computational details, together with additional materials information, are described in Sec. (S.5) of the Supplementary Information.

The optical properties of the magnetic fluid are estimated by taking into account the volume fraction of commercially available gold-coated magnetite particles in a toluene mixture. We use the mixing model presented by Chen Chen et al. 37`` as,

ε3=ε3,3(1+2Λ3)1−Λ3;Λ3=(ϕ1+ϕ2)εs−ε3,3εs+2ε3,3(9)\varepsilon_{3} = \frac{\varepsilon_{3,3}(1 + 2\Lambda_{3})}{1 - \Lambda_{3}}; \qquad\Lambda_{3} = (\phi_{1} + \phi_{2})\frac{\varepsilon_{s} - \varepsilon_{3,3}}{\varepsilon_{s} + 2\varepsilon_{3,3}} \tag*{(9)}

where the effective dielectric function of the sphere (εs\varepsilon_{s}) is

εs=ε3,1(1−2G)1+2G;G=ε3,1−ε3,22ε3,1+ε3,2ϕ2ϕ1+ϕ2.(10)\varepsilon_{s} = \frac{\varepsilon_{3,1}(1 - 2G)}{1 + 2G}; \qquad G = \frac{\varepsilon_{3,1} - \varepsilon_{3,2}}{2\varepsilon_{3,1} + \varepsilon_{3,2}}\frac{\phi_{2}}{\phi_{1} + \phi_{2}}. \tag*{(10)}

Here, ϕi\phi_{i} and ε3,i\varepsilon_{3,i} are the volume fraction and dielectric function of toluene (i=3i = 3), magnetite (i=2i = 2) and gold (i=1i = 1) for modeling the magnetic fluid where DD is the diameter of the magnetite core and the thickness RAuR_{\mathrm{Au}} of the gold coating, as illustrated in Figure 1. The radius of the gold coated nanoparticles is defined as R1=D/2+RAuR_{1} = D/2 + R_{\mathrm{Au}}.

In the current work, we consider the case where the amount of magnetite is 5% and 7% (ϕ2=0.05,0.07\phi_{2} = 0.05, 0.07). The corresponding volume fraction of the gold coating is therefore given by ϕ1=ϕ2×[(2R1/D)3−1]\phi_{1} = \phi_{2} \times[(2R_{1}/D)^{3} - 1]. We consider four typical cases where the average diameter (DD) of the magnetite nanoparticles is 10, 13, 15 and 17 nm. We use experimental data for toluene at T=298.15T = 298.15 K from the literature. [34] The dielectric functions of magnetite and gold are from DFT calculations in our recent work, [38] where room-temperature thermal electronic excitation was considered for the small-gap magnetite compound. In the main text, we report some important results that are obtained for pure magnetite particles (i.e. for R1=D/2R_1 = D/2). Most of the results obtained using gold-coating are included in Sec. (S.8) of the Supplementary Information. The size of the magnetite nanoparticles and gold-coated magnetite nanoparticle can be adjusted to control trapping behavior for pairs of unequal nanorods.

The permeability of different ferrofluids was described by Klimchitskaya et al. [18, 19] where numerical calculations were presented for a water-based ferrofluid containing a 5 % volume fraction of 10 nm diameter magnetite nanoparticles. The fluid was enclosed between two dissimilar plates, Au and SiO2_2. It was found that the magnetic particles caused the CL force to be repulsive for all the gap widths considered, between 100 nm and 2 μ\mum. Both the Drude and the plasma dispersive models for the Au plate were calculated, and no found difference. These results are of obvious technological interest. For microdevices the stiction effect may be a serious difficulty, and the magnetically tuned repulsiveness may therefore be helpful.

For water (close to being a non-magnetic fluid) with a fraction ϕ\phi of magnetite Klimchitskaya et al. estimated the permeability to be [18, 19]

μ3(0)=1+2π2ϕMs2D39kBT.(11)\mu_3(0) = 1 + \frac{2\pi^2 \phi M_s^2 D^3}{9 k_B T}. \tag*{(11)}

Toluene is only a little more magnetic (μ=0.99996≈1.00\mu= 0.99996 \approx1.00) so within the level of accuracy in our work the permeability of both water and toluene-based mixtures can be estimated with the approximation used by Klimchitskaya and Mostepanenko for water-based mixtures. [18, 19] Using Eq. (11) we derive the permeability for a range of volume fractions (from 0.05 to 0.07) and diameters (from 10 to 17 nm). The cylindrical nanoparticles we consider are non-magnetic (μ1(0)≈μ2(0)≈1\mu_1(0) \approx\mu_2(0) \approx1).

We recently investigated the interaction between a planar PTFE surface and a spherical PS surface in the presence of a magnetic fluid. [22] In this case both the non-retarded and retarded interaction could be tuned to be repulsive while zero frequency contributions to the CL interaction (both TM and TE) where attractive. This led us to conclude that it is possible to have a magnetic contribution to CL trapping. [22] However, trapping still occurred when the magnetic effects were ignored (i.e. the magnetic permeability set to one).

We will in the current work explore a more versatile system consisting of nanorod pair complexes where the materials are described by the dielectric functions shown in Figure 2. Here the intermediate magnetic fluid has a dielectric function in between those of ZnO and PS. This indicates that the non-retarded and retarded CL interaction between a ZnO and PS nanorod will be repulsive. [4]

Image not extracted

Figure 2. (Color online): The dielectric functions on the imaginary axis as functions of frequency ξ\xi. The parametrized data for PS and toluene are based on optical experiments given by van Zwol et al. [34] The magnetic fluid consists of a mixture of toluene and magnetite particles, using Eq. (9) for a volume fraction of ϕ=5%\phi= 5\%. The response functions of gold and magnetite are DFT results presented in [38], whereas the data for ZnO are from present DFT calculations.

In Figure 3 we show that in the case of a pair of ZnO and PS nanorods in a magnetic fluid with toluene mixed with 5-7% magnetite it is possible to obtain trapping that is entirely controlled by the magnetic contribution to the CL force. When the magnetic permeability of the fluid is set to one (μ3=1\mu_{3} = 1), the system only experiences repulsion. This dependence of trapping on the magnetic permeability has not been possible to achieve in the past. Notably, when the diameter of the magnetite nanoparticles decreases with fixed volume fraction of magnetite, or a weak magnetic field is applied, leading to reduced magnetic permeability of

Image not extracted

Figure 3. (Color online:) The CL force per unit length between two cylindrical objects (ZnO and PS respectively) with the same radius a=20a = 20 nm as a function of center-to-center distance RR when immersed in a ferrofluid with 5% and 7% concentrations of magnetite nanoparticles of diameter 10, 13, 15 and 17 nm.

the fluid, [39] the trapping occurs at increasingly larger nanorod pair separations. The trapping can also be controlled by keeping the diameter constant and varying the volume fraction.

As illustrated in Figure 5, the trend is that larger ϕ\phi shifts the force minimum towards larger separations and making it more shallow. Ultimately, in the absence of magnetic permeability effects or very small volume fraction, there is no trapping at all and the nanorod complex is split into two separate nanorods. Replacing the magnetite nanoparticles with commercially

Image not extracted

Figure 5. (Color online:) The CL force per unit length between two cylindrical objects (ZnO and PS respectively) with the same radius a=20a = 20 nm as a function of center-to-center distance RR when immersed in a ferrofluid with different concentrations of (ϕ=0.01\phi=0.01-0.15) magnetite nanoparticles of diameter 17 nm.

available gold-coated magnetite nanoparticles leads to an additional mechanism to tune the trapping distance as can be seen in Figure 4. With changes in the gold coating the dielectric function of the background media changes and the trapping distance can be shifted.

Image not extracted

Figure 4. (Color online): The CL force per unit length between two ZnO and PS cylindrical objects was evaluated for volume fraction ϕ=5%\phi= 5\%, 7%7\% and a nanoparticle diameter D=17D = 17 nm. The terms “without gold” and “with gold” refer to magnetite nanoparticles without and with a gold coating, respectively. For gold coating, layer thickness was taken to be, RAu=2R_{\mathrm{Au}} = 2 nm. Total1_1 is the fully retarded calculation for all frequencies without gold coating whereas Total2_2 reflects the retarded force with gold coating and p=0p = 0 denotes the zero frequency contribution.

Non-magnetic Casimir forces have been exploited, both theoretically and experimentally, to produce Casimir trapping and non-adhesion, via repulsive forces, in NEMS and MEMS.[7, 8, 9, 10, 11, 12, 14, 16] We have focused in the current work on the novel results where only the zero-frequency TE term changes, as trapping occurs over a range of different ϕ\phi values. The repulsion occurs when ϕ\phi is tuned such that the dielectric function for the fluid is between the values of the dielectric functions of the interacting nanorods for a wide range of frequencies. Our work demonstrates that the trapping can be controlled by both changes in the size of the magnetic nanoparticles and the concentration in the fluid. The exact values for these quantities can be adjusted to achieve trapping for different separations and for different material combinations. The proposed framework suggests a novel route for actively tuning the interacting forces between two unequal nanorods where a purely attractive potential can be driven into repulsive–attractive trapping potentials and even into regimes of complete repulsion. By a careful choice of the material parameters, one can tailor finite nanosystems that give rise to reversible adhesive behavior, stable tunable trapping and on-demand destabilization of clusters. These results demonstrate a general platform for nanocluster design in magnetic colloids where adjustable magnetic Casimir forces can be exploited for NEMS/MEMS applications.

Note that the proof-of-principle theory presented here assumes that trapping takes place at separations much larger than the diameter of the nanorod. In this situation, the trapping minima are very weak and can be affected by thermal noise Sec. (S.6) in the Supplementary Information. Nevertheless, this restriction can be mitigated by applying the proximity force approximation (PFA), valid in the short-distance regime, where enhanced stability of the trapping potential is expected.

Going forward, the framework we have developed here indicates some exciting prospects for future extension of tunable dispersion-force control [7, 8, 9, 10, 11, 12, 14, 16] to biomolecular materials. In particular, these ideas could be extended to direct biopolymer complexation with artificial nanothreads via magnetic nanoparticles in biofluids with a suitable choice of materials. This has a great potential in a variety of future diagnostic and therapeutic nanomedicine research. [40] It also provides a route to field-controlled assembly in hybrid soft–magnetic nanostructures and could also find utility for biosensing, targeted delivery, reconfigurable nanoarchitectures, and beyond.

Data availability: The code and data that support the findings of this study are available in Github.

Acknowledgement

SP and UDG acknowledge support from the Marie Skłodowska-Curie Doctoral Network TIMES, grant No. 101118915 and SPARKLE grant No. 101169225. MB’s contributions to this research are part of the project No. 2022/47/P/ST3/01236 co-funded by the National Science Centre and the European Union’s Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement No. 945339. Institutional and infrastructural support for the ENSEMBLE3 Centre of Excellence was provided through the ENSEMBLE3 project (MAB/2020/14) delivered within the Foundation for Polish Science International Research Agenda Programme and co-financed by the European Regional Development Fund and the Horizon 2020 Teaming for Excellence initiative (Grant Agreement No. 857543), as well as the Ministry of Education and Science initiative “Support for Centres of Excellence in Poland under Horizon 2020” (MEiN/2023/DIR/3797). We gratefully acknowledge Poland’s high-performance computing infrastructure PLGrid (HPC Centers: ACK Cyfronet AGH) for providing computer facilities and support within computational grant no. PLG/2023/016228 and for awarding this project access to the LUMI supercomputer, owned by the EuroHPC Joint Undertaking, hosted by CSC (Finland) and the LUMI consortium through grant no. PLL/2023/4/016319. CP acknowledges access to high-performance computing resources from NAISS, provided by NSC and PDC. LMW acknowledges financial support from the US Department of Energy under Grant No. DE-FG02-06ER46297.

References

  1. [1]Wang, S. C. The problem of the normal hydrogen molecule in the new quantum mechanics. Phys. Rev. 1928, 31, 579.DOI
  2. [2]London, F. Zur theorie und systematik der molekularkräfte. Zeitschrift für Physik 1930, 63, 245–279.
  3. [3]Casimir, H. B. G. On the attraction between two perfectly conducting plates. Proceedings of the KNAW. 1948; pp 793–795.
  4. [4]Dzyaloshinskii, I.; Lifshitz, E.; Pitaevskii, L. The general theory of van der Waals forces. Adv. Phys. 1961, 10, 165–209.
  5. [5]Parsegian, V. A.; Ninham, B. W. Application of the Lifshitz Theory to the Calculation of Van der Waals Forces across Thin Lipid Films. Nature 1969, 224, 1197–1198.pubmed.ncbi.nlm.nih.gov/5360550
  6. [6]Dobson, J.; White, A.; Rubio, A. Asymptotics of the dispersion interaction: analytic benchmarks for van der Waals energy functionals. Phys. Rev. Lett. 2006, 96, 073201.
  7. [7]Munday, J. N.; Capasso, F.; Parsegian, V. A. Measured long-range repulsive Casimir-Lifshitz forces. Nature 2009, 457, 170 – 173.DOI
  8. [8]Esteso, V.; Carretero-Palacios, S.; Míguez, H. Nanolevitation phenomena in real plane-parallel systems due to the balance between Casimir and gravity forces. J. Phys. Chem. C 2015, 119, 5663–5670.pubmed.ncbi.nlm.nih.gov/26405466
  9. [9]Woods, L. M.; Dalvit, D. A. R.; Tkatchenko, A.; Rodriguez-Lopez, P.; Rodriguez, A. W.; Podgornik, R. Materials perspective on Casimir and van der Waals interactions. Rev. Mod. Phys. 2016, 88, 045003.DOI
  10. [10]Zhao, R.; Li, L.; Yang, S.; Bao, W.; Xia, Y.; Ashby, P.; Wang, Y.; Zhang, X. Stable Casimir equilibria and quantum trapping. Science 2019, 364, 984–987.DOI
  11. [11]Rodriguez-Lopez, P.; Popescu, A.; Fialkovsky, I.; Khusnutdinov, N.; Woods, L. M. Signatures of complex optical response in Casimir interactions of type I and II Weyl semimetals. Commun. Mater. 2020, 1, 14.
  12. [12]Esteso, V.; Carretero-Palacios, S.; Míguez, H. Effect of Spatial Inhomogeneity on Quantum Trapping. J. Phys. Chem. Lett. 2022, 13, 4513–4519.DOI
  13. [13]Spreng, B.; Gong, T.; Munday, J. N. Recent developments on the Casimir torque. Int. J. Mod. Phys. A 2022, 37, 2241011.
  14. [14]Shelden, C.; Spreng, B.; Munday, J. N. Opportunities and challenges involving repulsive Casimir forces in nanotechnology. Appl. Phys. Rev. 2024, 11, 041325.DOI
  15. [15]Lee, D.-N.; Rodriguez-Lopez, P.; Woods, L. M. Phonon-assisted Casimir interactions between piezoelectric materials. Commun. Mater. 2024, 5, 260.
  16. [16]Williamson, C.; Besley, E. Casimir Force in Layered Materials and Control of the Stable Equilibrium. J. Phys. Chem. Lett. 2025, 16, 12470–12476, PMID: 41287959.DOI
  17. [17]Richmond, P.; Ninham, B. W. A note on the extension of the Lifshitz theory of van der Waals forces to magnetic media. J. Phys. C: Solid State Phys. 1971, 4, 1988–1993.DOI
  18. [18]Klimchitskaya, G. L.; Mostepanenko, V. M.; Nepomnyashchaya, E. K.; Velichko, E. N. Impact of magnetic nanoparticles on the Casimir pressure in three-layer systems. Phys. Rev. B 2019, 99, 045433.DOI
  19. [19]Velichko, E. N.; Klimchitskaya, G. L.; Nepomnyashchaya, E. K. Casimir repulsion though a water-based ferrofluid. Mod. Phys. Lett. A 2020, 35, 2040016.DOI
  20. [20]Derjaguin, B. Untersuchungen Über die Reibung und Adhäsion, IV. Kolloid-Zeitschrift 1934, 69, 155–164.
  21. [21]Blocki, J.; Swiatecki, W. J. A generalization of the Proximity Force Theorem. Ann. Phys. 1981, 132, 53–65.DOI
  22. [22]Pal, S.; Inacio, L.; Woods, L.; De Giovannini, U.; Persson, C.; Boström, M. Trapping in a Casimir Force Set-Up Controlled by Solution Permeability. arXiv preprint arXiv:2505.10977 2025,arxiv.org/abs/2505.10977
  23. [23]Richmond, P.; Davies, B.; Ninham, B. Van der Waals attraction between conducting molecules. Phys. Lett. A 1972, 39, 301–302.DOI
  24. [24]Mitchell, D.; Ninham, B.; Richmond, P. Van der Waals forces between cylinders: I. Nonretarded forces between thin isotropic rods and finite size corrections. Biophys. J. 1973, 13, 359–369.pubmed.ncbi.nlm.nih.gov/4696762
  25. [25]Davies, B.; Ninham, B.; Richmond, P. Van der Waals forces between thin cylinders: New features due to conduction processes. J. Chem. Phys. 1973, 58, 744–750.DOI
  26. [26]Nyland, G. H.; Brevik, I. Van der Waals interaction between two cylinders in a magnetic fluid. Physica A: Stat. Mech. Appl. 1994, 202, 81–93.DOI
  27. [27]Pal, S.; Brevik, I.; Boström, M. Dispersion Interaction Between Thin Conducting Cylinders. Phys. Chem. Phys. Chem. 2024, 26, 17969–17978.arxiv.org/abs/2406.08047
  28. [28]Wang, J.; Seidel, G. D.; Cheng, S. Analytical interaction potential for Lennard-Jones rods. Phys. Rev. E 2025, 111, 015403.DOI
  29. [29]Watson, G. N. Treatise on the Theory of Bessel Functions; Cambridge University Press, 1958.
  30. [30]McLachlan, A. Resonance transfer of molecular excitation energy. Mol. Phys. 1964, 8, 409–423.DOI
  31. [31]Boström, M.; Longdell, J.; Mitchell, D.; Ninham, B. Resonance interaction between one excited and one ground state atom. Eur. Phys. J. D 2003, 22, 47–52.DOI
  32. [32]Parsegian, V. A. Van Der Waals Forces: A Handbook for Biologists, Chemists, Engineers, and Physicists; Cambridge University Press: New York, 2006; p 398.
  33. [33]Langbein, D. Van der Waals attraction between cylinders, rods or fibers. Phys. Kondens. Mater. 1972, 15, 61–86.DOI
  34. [34]van Zwol, P. J.; Palasantzas, G. Repulsive Casimir forces between solid materials with high-refractive-index intervening liquids. Phys. Rev. A 2010, 81, 062502.DOI
  35. [35]Kresse, G.; Joubert, D. From ultrasoft pseudopotentials to the projector augmented-wave method. Phys. Rev. B 1999, 59, 1758–1775.DOI
  36. [36]Crovetto, A.; Chen, R.; Ettlinger, R. B.; Cazzaniga, A. C.; Schou, J.; Persson, C.; Hansen, O. Dielectric function and double absorption onset of monoclinic CuSn₃: Origin of experimental features explained by first-principles calculations. Sol. Energy Mater. Sol. Cells 2016, 154, 121–129.
  37. [38]Carretero-Palacios, S.; Esteso, V.; Li, Y.; Kuthe, S.; Brevik, I.; Iordanidou, K.; Malyi, O. I.; Glaser, B.; Persson, C.; Boström, M. Impact of metal oxidation on ice growth and melting. Phys. Rev. B 2025, 111, 085407.
  38. [39]Zhang, Y.; Zhang, H.; Wang, X.; Wang, Y.; Liu, Y.; Li, S.; Zhang, T.; Fan, C.; Zeng, C. Magnetic-field tuning of the Casimir force. Nature Physics 2024, 20, 1282–1287.DOI
  39. [40]Coene, A.; Leliaert, J. Magnetic nanoparticles in theranostic applications. J. Appl. Phys. 2022, 131, 160902.DOI

Paper details

vlnc:W_jnmtk6yy
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