WorksPhysics and Astronomy
Apparent contact angle equation for a dielectric droplet on a conductive hydrophilic surface in an electric field
Abstract
A dielectric droplet resting on a conductive substrate elongates when a voltage is applied across the surrounding gap without touching the upper electrode. We construct a closed form theory of the equilibrium shapes by approximating the interface as a truncated spheroid with a single shape parameter and obtain the dimensionless electrostatic prefactor that corrects the free-space depolarization, based on the mirror symmetry across the substrate. Minimizing the resulting free energy along voltage ramps reproduces the boundary-element-method (BEM) equilibria, both the droplet height and the contact-line radius, with and without gravity, at negligible computational cost. An explicit apparent contact angle equation is found that collapses all the numerical calculation data. The configuration shows a behavior of fixed contact angle and an increasing apparent angle with the applied voltage, which is completely different from that of the Young–Lippmann equation.
Introduction
The interplay of electric fields and liquid interfaces has been studied for more than a century, from droplet deformation and disintegration in uniform fields [1, 2, 3, 4, 5] to electrohydrodynamic patterning and jet printing [6, 7, 8, 9]; on substrates, electrically actuated droplets underpin digital microfluidics [10, 11]. For a conductive droplet on a dielectric-coated electrode, the Young–Lippmann equation [12, 13] lowers the effective solid–liquid tension, so the measured angle decreases with voltage, yet the microscopic angle at the contact line retains the Young value [14, 15], with only an apparent angle responding [16, 17].
The dual configuration—a dielectric droplet on a conductive substrate in a parallel-plate gap it does not bridge—is qualitatively different: the field penetrates and polarizes the liquid, the electrostatic energy is stored in the droplet volume, and the primary response is bulk elongation rather than a change in contact-angle [18, 19, 20]. Realized in non-contact electrostatic manipulation [21, 22], this problem has been addressed by experiments, asymptotics, and direct simulation [19, 23, 24, 25], but a closed-form theory of the equilibrium shape is still lacking. This theory is desirable because the configuration exhibits voltage-driven wetting transitions and critical voltages [25, 26, 27] and a generic loss of equilibrium in strong fields [28, 29, 30], whose prediction requires the coupled shape-field problem at every voltage.
Here, we construct such a theory, with its proof relying on three observations. First, the true contact angle is pinned at the Young angle by the Young equation at the contact line [14, 16], a fact we verify with a variational model in which the contact angle is a free variable
(Sect. II); the equilibrium shapes are thus confined to a low-dimensional family, which truncated spheroids of aspect ratio already capture quantitatively. Second, the dimensionless electrostatic prefactor of this family has a closed form, obtained by mapping the droplet and its mirror image onto a lens-shaped spheroid and correcting the free-space depolarization factor; the result matches BEM calculations within 2.7% over the investigated range , . The free energy is then explicit, with each equilibrium found by one-dimensional minimization without solving any field equation. Third, the theory predicts that an experimental measurement would yield an apparent contact angle that varies with voltage and measurement scale, even though the microscopic angle does not.
The paper is organized as follows. Section II defines the model, the free-energy construction, and the two numerical implementations used in this work. Section III derives the closed-form electrostatic prefactor. Section IV compares the theoretical equilibria with the BEM solutions and maps the equilibrium shape on the plane. Section V establishes the apparent-angle scaling law, and Sec. VI concludes.
Model and Methods
We consider an axisymmetric dielectric droplet of volume , surface tension , density contrast and relative permittivity , sessile on the lower plate of a parallel-plate capacitor of gap , across which a voltage is applied. Here is a material parameter; the results below use (water), and is used in Figure 5 to test the generality of the scaling law Figure 1. The far field is uniform, , and the Bond number reflects the ratio of the contributions made by gravity and surface tension; we report results with and without gravity throughout. Throughout the paper, lengths are measured in units of , and the applied voltage is measured in units of the characteristic voltage

Figure 1. Schematic of the system. A dielectric droplet sits on the grounded lower plate of a parallel-plate capacitor of gap ; and are the semi-axes of the truncated spheroid, the height, and the contact-line radius. Left, zero voltage; right, voltage applied. The contact angle stays at the Young value while the drop elongates in the field ; the droplet and its mirror image form the lens construction of Sec. III.

Figure 5. Collapse of different parameter combinations onto the scaling law Eq. (13) (dotted line), for and , and . Markers: open triangles (Bo = 0, ), filled triangles (Bo = 0.82, ), open circles (Bo = 0, ), and filled circles (Bo = 0.82, ). in , if scaled by , can be uniquely determined by .
At the sharp-interface level, the contact line is a boundary condition rather than a variational degree of freedom. The Young equation fixes the microscopic contact angle at [14], and the electrostatic contribution to the contact-line force vanishes because the wedge singularity of the field is integrable [16, 17]. We test this assumption independently with a variational (spectral) model of the full profile in which the contact angle is a free optimization variable: minimizing the total energy over , returns within , while the height increases by up to 22%. The free-boundary BEM solver described in the following therefore imposes as its contact-line boundary condition and serves as the benchmark for the shape. The equilibrium problem therefore reduces to shape variations at fixed . Guided by this fact, we approximate the interface by a truncated spheroid, i.e., the cap of an ellipsoid of equatorial and polar semi-axes and , cut by the substrate such that the contact angle is equal to .
In the DC voltage ensemble, the total free energy reads
where , , are dimensionless prefactors of the electrostatic, surface, and gravitational contributions. The geometric prefactors and are analytic functions of the ellipsoid geometry and are given in Appendix A. The electrostatic prefactor is the nontrivial input and is derived in Sec. III.
In this work, two numerical implementations are used. The first is the theory itself. At each voltage, the volume constraint Eq. (14) fixes the scale for a given , so that the free energy Eq. (1) becomes a function of alone; the equilibrium is located by the bounded one-dimensional minimization of the free energy with respect to , followed by a continuation in from the zero-voltage state, with the converged at one voltage serving as the initial guess in the next.
The second implementation is an independent sharp-interface free-boundary solver that serves as the benchmark. The electric problem is solved by an axisymmetric boundary-element method. The potential is represented as with the applied uniform potential , and the interface conditions of continuous and reduce to a second-kind Fredholm equation for the polarization charge density ,
where , is the Green’s half-space function of the grounded substrate, and the azimuthally integrated kernels are expressed through complete elliptic integrals. The interface is discretized into arc-length segments with a piecewise-linear charge density, collocated at interior Gauss points to avoid the spurious singularity of the segment normals. From the solution, the Maxwell stress along the outward normal is
with and the normal and tangential field components on the liquid side. The electrostatic free energy, which defines the prefactor of Eq. (1), is evaluated from the polarization charge as
where with the applied potential , so that the BEM solution determines . The shape obeys the Young–Laplace equation in a strong form, integrated as an initial-value problem in arc length from the apex,
where is the tangent angle and is the local pressure jump. The three shooting parameters, namely the apex height , the pressure jump , and the arc length of the contact line , are fixed by the conditions , , and at the contact line. The solver is iterated with under-relaxation until self-consistent, and the branch is built by continuation in . The solver is validated against the classical depolarization result for spheres and spheroids (0.08–0.25%) and against independent finite-element (COMSOL) energies (to within 0.3%), as detailed in Appendix B.
Closed-Form Electrostatic Prefactor
For a complete spheroid in a uniform far field, the polarization energy is classical [31],
with the depolarization factor,
which equals for a sphere. For a truncated spheroid on a conductive substrate, such a closed form does not exist.

Figure 2. (a) Effective depolarization factor given by the lens model, Eqs. (8) and (9) (solid lines), compared with BEM calculations (open circles) for five Young angles. (b) Electrostatic prefactor given by the lens model (solid lines), compared with the same BEM data.
The grounded substrate acts as a mirror, so the droplet and its image form a double-capped lens of aspect ratio
Equilibrium Shapes
We validated the theory against BEM free-boundary solutions on 32 equilibrium branches spanning –, with and without gravity, each followed by a continuation in .
We take the effective depolarization factor of the truncated spheroid as
which at recovers and the half-spheroid value exactly. Figure 2 compares Eqs. (8) and (9) with BEM calculations of on a grid spanning – and –, showing that the lens model reproduces well both the depolarization factor and the prefactor over the whole range.
The BEM values agree with the independent finite-element (COMSOL) energies to within (see the Appendix), indicating that the residual error is attributable to the model rather than the numerics.
With given by Eqs. (6), (8) and (9) and , analytic (Appendix A), the entire free energy Eq. (1) becomes explicit and no field equation is needed to be solved at any stage of the equilibrium calculation.
Figure 3 shows four representative branches at , indicating that the theory closely follows the benchmark for height and radius of the contact line.

Figure 3. (a) Equilibrium height and (b) contact-line radius as functions of the dimensionless voltage at and for , comparing the theory (solid lines) with the BEM free-boundary solutions (open circles).
The precision is controlled by the prefactor rather than the shape family: recomputing with the uncorrected free-space factor degrades the height error to 8% at large angles, and the branches terminate – early. The cost is negligible—each equilibrium is a one-dimensional minimization of an analytic function, whereas the benchmark needs tens of coupled free-boundary solves per voltage.
Figure 4 assembles the validated theory into a map of the equilibrium shape on the plane with gravity. The response strongly depends on the Young angle. At , increasing to 3.4 raises the height by 12% and reduces the radius of the contact line by 4%, while at the same ramp raises the height by 22% and shrinks the contact line by 9%. In this whole map, the theory deviates from the benchmark by no more than the few percent documented in Figure 8. The largest deviations occur near the upper-right corner, where the equilibrium branch approaches its terminal turning point. No adjustable parameter enters either map.

Figure 4. Equilibrium shape over the parameter plane at and : (a) droplet height and (b) contact-line radius as functions of the Young angle and the dimensionless voltage .

Figure 8. Maximum relative error along the equilibrium branches versus for the height (circles) and the contact-line radius (squares), with filled symbols and without (open symbols) gravity; the dotted line marks 1%.
Apparent Contact Angle Equation
The pinned contact angle does not imply a null measurement: a goniometer fits the visible profile, which does change. In electrowetting, the measured angle is the spherical cap sharing the drop’s volume and contact radius, extrapolated to the substrate; the two definitions coincide because the field distorts only a small region near the contact line [16, 17]. Here, the distortion spans the whole profile, so no unique macroscopic angle exists.
We therefore define the local tangent angle at the arc length from the contact line and the field-induced change relative to the zero-voltage profile. We also introduce the cosine change .
The change in is positive, linear in in the observable range, and increases quadratically with . Integrating the Young–Laplace equation [Eq. (5)] from the contact line gives
At zero voltage, the profile is a spherical cap, on which is constant. Near the contact line () the azimuthal-curvature term therefore changes only at relative to the leading electrostatic term and is removed in what follows. Subtracting the zero-voltage profile and linearizing the cosine at the zero-field tangent angle, in leading order , and all data collapse onto
where is the electro-capillary number and is the radius of curvature of the zero-field spherical cap. The dependence of on the Young angle can be fixed as follows. In the contact line limit, the balance of electrostatic driving against capillary stiffness gives , where and is the second derivative with respect to ; the resulting change in the meridional curvature of the contact line gives . Combined with (11), this yields
where is the half-spheroid prefactor. A least-squares fit of the BEM data over fixes the remaining numerical coefficient to , so that at acute and right angles,
Figure 5 shows that different combinations of parameters collapse onto the scaling law Eq. (13), establishing it as a numerically established scaling law. Rescaling is insensitive to gravity: the Bo = 0.82 data (dashed) fall onto the same master curve as the Bo = 0 data within the investigated range. Figure 6 further compares the theory with the BEM free-boundary benchmark for the cosine change at four Young’s contact angles, confirming that Eq. (13) describes the apparent contact angle of a dielectric droplet on a conductive hydrophilic surface in an electric field.

Figure 6. Cosine change of the local tangent angle versus at Bo = 0 for four Young angles and four voltages each, the color encoding the voltage, comparing the theory (solid lines) with the BEM free-boundary benchmark (open circles).
Conclusion
We have shown that the equilibrium of a sessile dielectric droplet in a non-contact electric field admits an accurate closed-form description by approximating the interface as a truncated spheroid with a fixed contact angle and a single shape parameter. The dimensionless electrostatic prefactor corrects the free-space polarization well. Based on this, an explicit apparent contact angle equation is found that does not depend on gravity and collapses all the numerical calculation data. It also demonstrates a behavior different from that of the Young-Lippmann equation that increases with the applied voltage.
Acknowledgments
We acknowledge the financial support from the China National Natural Science Foundation under Grant
No. 12574232 and the 111 project B16029.
The data supporting the findings of this article are available from the authors upon reasonable request.
Analytic geometric prefactors
For the truncated-spheroid family, the geometry is fixed by the aspect ratio , the truncation parameter defined in Eq. (8), and the volume constraint. With the vertical coordinate measured in units of from the spheroid center, the cap spans , and the volume constraint is given by
so that and .
The liquid–gas area of the cap is
with the elementary antiderivative
where and for . The solid–liquid area is , and the surface prefactor of Eq. (1) is
The centroid height above the substrate is
and the gravitational prefactor is .
Accuracy of the closed-form theory
The BEM solver is validated against independent finite-element (COMSOL) calculations. Figure 7(a) compares the electrostatic prefactor obtained by the two methods on – and –; the largest relative difference is 0.3%. Figure 7(b) reports the maximum relative error of the prefactor on the BEM grid of § 3: the lens model stays at or below 2.7% for every Young angle. The remaining equilibrium-shape

Figure 7. (a) Validation of the BEM electrostatic prefactor against independent finite-element (COMSOL) calculations. Solid lines of different colors: BEM for ; open circles of the same color: COMSOL at the same points over –. The largest difference is 0.3%. (b) Maximum relative error of the electrostatic prefactor over the sampled plane for the lens model.
error is small and, at acute angles, is dominated by the prefactor rather than the truncated-spheroid family. Figure 8 reports the maximum relative error of height and contact-line radius along each of the 32 branches of § 4, with and without gravity. The height error stays below 2% for – and grows only near the highest angle termination points, the radius of the contact line following the same trend with smaller errors at acute angles.
References
- [1]C. T. O’Konski and H. C. Thacher, The Journal of Physical Chemistry 57, 955 (1953).
- [2]G. I. Taylor, Proceedings of the Royal Society of London. Series A 280, 383 (1964).
- [3]G. I. Taylor, A. D. McEwan, and L. N. J. de Jong, Proceedings of the Royal Society of London. Series A 291, 159 (1966).
- [4]K. J. Cheng and J. B. Chaddock, Physics Letters A 106, 51 (1984).
- [5]K. Adamiak and J. M. Floryan, IEEE Transactions on Industry Applications 47, 2374 (2011).
- [6]R. T. Collins, J. J. Jones, M. T. Harris, and O. A. Basaran, Nature Physics 4, 149 (2008).
- [7]J. U. Park, M. Hardy, S. J. Kang, K. Barton, K. Adair, D. K. Mukhopadhyay, C. Y. Lee, M. S. Strano, A. G. Alleyne, J. G. Georgiadis, P. M. Ferreira, and J. A. Rogers, Nature Materials 6, 782 (2007).
- [8]J. B. Fenn, M. Mann, C. K. Meng, S. F. Wong, and C. M. Whitehouse, Science 246, 64 (1989).
- [9]P. M. Vlahovska, Annual Review of Fluid Mechanics 51, 305 (2019).
- [10]K. Choi, A. H. C. Ng, R. Fobel, and A. R. Wheeler, Annual Review of Analytical Chemistry 5, 413 (2012).
- [11]P. Zhu and L. Wang, Lab on a Chip 17, 34 (2016).
- [12]G. Lippmann, Annales de Chimie et de Physique 5, 494 (1875).
- [13]F. Mugele and J. C. Baret, Journal of Physics: Condensed Matter 17, R705 (2005).
- [14]T. Young, Philosophical Transactions of the Royal Society of London 95, 65 (1805).
- [15]P. G. de Gennes, Reviews of Modern Physics 57, 827 (1985).
- [16]J. Buehrle, S. Herminghaus, and F. Mugele, Physical Review Letters 91, 086101 (2003).
- [17]F. Mugele and J. Buehrle, Journal of Physics: Condensed Matter 19, 375112 (2007).
- [18]C. Tsakonas, L. T. Corson, I. C. Sage, and C. V. Brown, Journal of Electrostatics 72, 437 (2014).
- [19]L. T. Corson, C. Tsakonas, B. R. Duffy, N. J. Mottram, I. C. Sage, C. V. Brown, and S. K. Wilson, Physics of Fluids 26, 122106 (2014).
- [20]V. Vancauwenberghe, P. Di Marco, and D. Brutin, Colloids and Surfaces A: Physicochemical and Engineering Aspects 432, 50 (2013).
- [21]N. Sinn, M. T. Schuer, and S. Hardt, Applied Physics Letters 114, 213704 (2019).
- [22]J. Hartmann, M. T. Schür, and S. Hardt, Nature Communications 13, 27879 (2022).
- [23]D. S. Pillai and K. C. Sahu, Physical Review E 109, L013101 (2024).
- [24]D. Langemann, Mathematics and Computers in Simulation 63, 529 (2003).
- [25]J. Li, K. Wen, X. Xiao, X. Chen, and C.-X. Wu, Journal of Applied Physics 137, 054701 (2025).
- [26]K. Xiao, X. Chen, and C.-X. Wu, Physical Review Research 3, 033277 (2021).
- [27]K. Xiao and C.-X. Wu, Physical Review E 105, 064609 (2022).
- [28]J. Beroz, A. J. Hart, and J. W. M. Bush, Physical Review Letters 122, 244501 (2019).
- [29]Q. Vo and T. Tran, Physical Review Letters 123, 024502 (2019).
- [30]A. Cavalli, D. J. Preston, E. Tio, D. W. Martin, N. Miljkovic, E. N. Wang, F. Blanchette, and J. W. M. Bush, Physics of Fluids 28, 022101 (2016).
- [31]L. D. Landau and E. M. Lifshitz, Electrodynamics of Continuous Media, 2nd ed. (Pergamon Press, Oxford, 1984).