Melting and re-entrant melting of polydisperse hard disks

Melting and re-entrant melting of polydisperse hard disks

Pablo Sampedro Ruiz School of Chemical and Biomedical Engineering, Nanyang Technological University,
62 Nanyang Drive, 637459, Singapore
   Qun-li Lei School of Chemical and Biomedical Engineering, Nanyang Technological University,
62 Nanyang Drive, 637459, Singapore
   Ran Ni School of Chemical and Biomedical Engineering, Nanyang Technological University,
62 Nanyang Drive, 637459, Singapore

Because of long-wavelength fluctuations, the nature of solids and phase transitions in 2D are different from those in 3D systems, and have been heavily debated in past decades, in which the focus was on the existence of hexatic phase. Here, by using large scale computer simulations, we investigate the melting transition in 2D systems of polydisperse hard disks. We find that, with increasing the particle size polydispersity, the melting transition can be qualitatively changed from the recently proposed two-stage process to the Kosterlitz-Thouless-Halperin-Nelson-Young scenario with significantly enlarged stability range for hexatic phase. Moreover, re-entrant melting transitions are found in high density systems of polydisperse hard disks, which were proven impossible in 3D polydisperse hard-sphere systems. These suggest a new fundamental difference between phase transitions in polydisperse systems in 2D and 3D.

I Introduction

Melting of two-dimensional solids has been heavily discussed since it was proven that long-wavelength thermal fluctuations prevent the long-range positional order in 2D systems Strandburg (1988); Dash (1999); Gasser (2009). A significant attempt to reach the general description of 2D melting was the Kosterlitz-Thouless-Halperin-Nelson-Young (KTHNY) theory, that predicts a new phase of matter, i.e., the hexatic phase, which has quasi-long-range orientational order and short-range positional order. The hexatic phase is predicted to be interposed between the usual solid and fluid phases. Therefore, the melting in 2D was predicted to undergo two continuous transitions from solid to hexatic and hexatic to fluid, respectively Kosterlitz and Thouless (1973); Halperin and Nelson (1978); Nelson and Halperin (1979); Young (1979); Berezinskii (1971). However, the KTHNY theory does not rule out the first-order transition due to other effects Binder et al. (2002), e.g., the grain-boundary induced melting Chui (1982); Saito (1982). In arguably the simplest benchmarking model system in 2D, i.e., the system of monodisperse hard disks, the physics of melting transition has been debated for long Zahn et al. (1999); Karnchanaphanurach et al. (2000); Han et al. (2008); Rice (2009); Murray and Van Winkle (1987); Marcus and Rice (1996); von Grünberg et al. (2004); Keim et al. (2007); Xu and Rice (2008). It was recently settled that the melting of solids in the system of monodisperse hard disk undergoes a two-stage process consisting of a continuous solid-hexatic transition followed by a first-order hexatic-fluid transition Bernard and Krauth (2011); Engel et al. (2013), and the shape and softness of particles also play important roles in the 2D melting Kapfer and Krauth (2015); Anderson et al. (2017); Li and Ciamarra (2018).

It was found that pinning a small fraction of particles can change the melting transition in hard-disk systems significantly Deutschländer et al. (2013); Qi and Dijkstra (2015). Moreover, simulations of binary hard-disk mixtures showed that the presence of tiny amounts of small particles can eliminate the stability of hexatic phase Russo and Wilding (2017). These highlight that the melting transition in 2D is subtle. A recent experiment with colloidal hard spheres in 2D Thorneywork et al. (2017) appears to support the two-stage melting found in simulation Bernard and Krauth (2011); Engel et al. (2013). However, most of experimental systems possess certain degree of (continuous) particle size polydispersity, and polydisperse hard disks have also been widely employed as a model system to investigate the glass transition Kawasaki et al. (2007); Kawasaki and Tanaka (2011); Russo and Tanaka (2015). Yet the effect of polydispersity on the nature of phase transitions in 2D remains unknown. To this end, we investigate the equilibrium phase behaviour of a 2D polydisperse hard-disk system (PHDS) with Gaussian-like particle size polydispersity. We find that with increasing the polydispersity, the first-order hexatic-fluid transition becomes weaker, and completely switches to be continuous following the KTHNY scenario in PHDS with around size polydispersity. Concurrently, the packing fraction range for stable hexatic phase increases significantly by one order of magnitude compared to that in monodisperse hard-disk systems. More surprisingly, in PHDS with slightly higher polydispersity, we observe re-entrant solid-hexatic and hexatic-fluid transitions at high density, which were proven impossible in 3D polydisperse hard-sphere systems Fasolo and Sollich (2003).

Figure 1: Phase behaviour of polydisperse hard disks.(a): Equation of state (EOS) for polydisperse hard disk systems (PHDS) with various polydispersity parameter to in the representation of vs , where and are the pressure and density of the hexatic phase at the fluid-hexatic transition, respectively, and the solid lines are fits of the EOS using 5th order polynomials. (b): as functions of for systems with and 0.08. and are the pressure of the system and the density of particles in the system, respectively. (c): EOSs for PHDS of various numbers of particles at . The solid symbols are the obtained fluid-hexatic transition points. (d): Phase diagram of the PHDS in the representation of and , in which the dashed lines are the interpolated phase boundaries for re-entrant melting of solid and hexatic phases. The phase boundaries are obtained from MC simulations for PHDS with to 0.0835. Inset: the enlarged view of the region of phase diagram at . The dotted lines connect the re-entrant transitions at , , and 0.0835 from left to right.

Ii Results

ii.1 Model

To model the effect of polydispersity, we consider a 2D system of volume containing polydisperse hard disks based on the semigrand canonical ensemble, in which the chemical potential difference between particles of different size is fixed Kofke and Glandt (1988); Bolhuis and Kofke (1996); Kofke and Bolhuis (1999); Pronk and Frenkel (2004). In this work, we use the following function for the distribution of chemical potential difference


where is the particle diameter changing from 0 to with and the Boltzmann constant and temperature of the system, respectively. is the polydispersity parameter, and in the ideal gas limit Eq. (1) gives a Gaussian-like particle size distribution centered around with the standard deviation . This models PHDS in equilibrium with the dilute reservoir of Gaussian-like particle size distribution, e.g., the very top region in sedimentation experiments Thorneywork et al. (2017).

Figure 2: Stabilization of hexatic phase. (a): The fractions of topological defects and dislocations in the hexatic phase as functions of for systems of various polydispersity parameter . Here and are the density of the system and the lowest density of stable hexatic phase, respectively. The vertical dashed lines indicate the hexatic-solid transition points. (b): Low pressure phase diagram of polydisperse hard disks in the representation of vs , where is the packing fraction of the system with the diameter of particle . The state points obtained from simulations at each from 0.005 to 0.08 are shown as the symbols, which are color coded with . The error bars are smaller than the symbols.
Figure 3: Re-entrant melting transitions. (a,b): The pair correlation function along the major axis of the system (a) and the six fold orientation correlation function (b) for polydisperse hard disk systems (PHDS) with at various densities. (c): and as functions of density for PHDS with , 0.083, and 0.0835. (d): Contour plot of the probability density distribution of particle size for systems of different densities at . (e): Mean square displacement in PHDS of at densities , 2.292, and 3.565 as marked in the inset, where is the time unit of molecular dynamics simulations with the mass of particles. Inset: the diffusion coefficient as a function of density in PHDS with .

ii.2 Phase diagram

By performing Monte Carlo (MC) simulations, we calculate the equation of state (EOS) for a system of hard disks with various from 0.005 to . As shown in Fig. 1a, one can see that when the polydispersity parameter is small, i.e., , there is clearly a Mayer-Wood loop in the EOS Mayer and Wood (1965) implying a first-order transition from fluid with increasing the density of the system. Similar to the monodisperse hard-disk system, i.e., , the coexisting phase with fluid in PHDS is a hexatic phase Bernard and Krauth (2011); Engel et al. (2013); Pronk and Frenkel (2004). To characterize the structural difference between fluid and hexatic phases, we calculate the six-fold bond orientation order parameter with , where is the angle between the vector connecting particle with its neighbour and a chosen fixed reference vector. is the number of first neighbours for particle based on the Voronoi tessellation of the system. As shown in Fig. 1a, the density difference between the coexisting hexatic and fluid phases becomes smaller with increasing , and when , the density jump from fluid to hexatic phase disappears, while increases significantly at (Fig. 1b). Here we note that in an infinitely large system, should be zero for hexatic phase, while in our large finite system it can be a positive number indicating the existence of some orientational order in the system. However, we do not use to identify the existence of hexatic phase, for which we always check the change of correlation functions to confirm. This suggests that the first-order fluid-hexatic transition becomes weaker with increasing , and changes to be continuous at high polydispersity, i.e., , following the celebrated KTHNY scenario Kosterlitz and Thouless (1973); Halperin and Nelson (1978); Nelson and Halperin (1979); Young (1979). This, to some extent, is similar to the melting of soft spheres Kapfer and Krauth (2015). Because with increasing the particle size polydispersity, the distribution of distance between the nearest neighbors becomes wider, which is similar to introducing a “soft” repulsion between particles, and it has been also found in the melting of soft spheres in 2D, the transition type can switch from a first order transition to a continuous KTHNY scenario with increasing the “softness” of the potential  Kapfer and Krauth (2015).

ii.3 Finite size effect on the melting in polydisperse hard disks

It was shown that the system finite size effect is important in studying 2D melting Bernard and Krauth (2011); Gribova et al. (2011). Therefore, we perform MC simulations for PHDS of various numbers of particles from to at , and the EOSs for different system sizes are shown in Fig. 1c. When the system size is small, i.e., , there is a pronounced Mayer-Wood loop in EOS, signaturing a first-order fluid-hexatic transition consistent with Ref Pronk and Frenkel (2004). Interestingly, with increasing the system size, the first-order fluid-hexatic transition becomes weaker, and changes to be continuous at . Further increasing does not change EOS significantly, which suggests the finite size effect negligible in our system of particles. The change from a first order like transition to a continuous transition with increasing system size can be interpreted as following. A finite 2D system with periodic boundary conditions can be seen as a system wrapped on a torus, and this increases the cooperation between neighbouring particles as a result of extra “communication” via paths encircling the torus, which was shown being able to change the transition type from continuous to first order like in small systems Ferdinand and Fisher (1969). By contrast, this effect does not appear in systems with open free boundaries, while in real experiments, the open boundary walls can induce order in the fluid. Moreover, for , we ensure that most of the coexisting densities do not change significantly with further increasing the system size to , while the only exception is for systems at (see Supplementary Figure 1), of which the exact boundary is out of the reach with our present computation capability.

ii.4 Stabilization of the hexatic phase

With increasing the density of the system, the hexatic phase solidifies, and the hexatic-solid transition point can be obtained by checking the pair correlation function along the major axis of system switching from an exponential decay to a power law decay upon solidification Bernard and Krauth (2011). The resulting phase diagram is plotted in the representation of vs in Fig. 1d, where is the actual particle size polydispersity in the system. One can see that with increasing the polydispersity of the system, the density range for stable hexatic phase increases substantially for above 0.07.

To understand the physics behind this enhanced stability of hexatic phase in PHDS, we calculate the fraction of topological defects in the system using the method in Ref Anderson et al. (2017). Particle is identified as a topological defect if its disclination charge . These topological defects form clusters which can be classified based on the total disclination charge and Burgers vector Anderson et al. (2017). Defect clusters having nonzero Burgers vectors and zero disclination charges are called dislocations, and the dislocations as well as the defect clusters with nonzero disclination charges can destroy the bond orientational order in 2D solids. Although the total amount of topological defects is not directly related to the stability of hexatic phase, the dislocations are evidence for the hexatic phase along with few defect clusters with nonzero disclination charges Anderson et al. (2017). The fraction of particles in defect clusters with nonzero disclination charge is much smaller, i.e., more than one order of magnitude smaller, than that of dislocations in the hexatic phase found in our simulations, although the changing trends of these two types of defects are very similar (see Supplementary Figure 2).

The calculated fractions of topological defects and dislocations in our simulations with various are plotted as functions of in Fig 2a, where is the lowest density of stable hexatic phase, and the vertical dashed lines locate hexatic-solid transitions. One can see that typically when the fraction of dislocations decreases to below about , the system solidifies. When the polydispersity is small, i.e., , the dependence of both fractions of defects and dislocations on density do not change significantly with increasing , and the density ranges for stable hexatic phase below the dashed lines in Fig. 2a are almost the same. However, at , the fraction of topological defects increases to around at , and also the fraction of dislocations increases to about at . Along with increasing the density at and 0.08, the fraction of dislocations decreases to below the threshold for solidification at much higher density compared with the system of . The fraction of defect clusters with nonzero disclination charges also changes similarly in systems of different polydispersity (Supplementary Figure 2). Therefore, this suggests that the size polydispersity of hard disks creates more topological defects, which could subsequently increase the fraction of dislocations as well as the defect clusters with nonzero disclination charges in the system. This destroys the quasi-long-range positional correlation in the solid driving the formation of hexatic phase. As the chemical potential difference is fixed in our simulations, the average particle size in the system decreases with increasing density. Therefore, the packing fraction of the system does not linearly increase with the density, while the actual particle size distribution remains very close to a Gaussian-like distribution centered around (see Supplementary Figure 3). At very high density, when the polydispersity is high enough, e.g., , the packing fraction of the system can even decrease with increasing the density in ensemble, which different from the ensemble Santen and Krauth (2000). This decreasing packing fraction at high density signatures a re-entrant melting transition, which we explain later. In experiments, one of the most relevant parameters is the packing fraction of the system. Thus we plot the low pressure phase diagram of PHDS in the representation of vs in Fig. 2b. Here the low pressure means the range of pressure close to the fluidhexatic and hexaticsolid transitions below the re-entrant transitions. One can see that in systems of larger size polydispersity, the average particle size is smaller, and the actual packing fraction range for stable hexatic increases from in the monodisperse hard-disk system to about in the PHDS with . Moreover, at fixed , increasing the density of the system along with the fluid-hexatic and hexatic-solid transitions decreases the actual polydispersity in the system, which is due to the formation of more ordered structures eliminating the particle size fluctuation in the system.

ii.5 Re-entrant meltings

Another interesting feature of the phase diagram in Fig. 1d is that when the size polydispersity of hard disks is around 0.08 to 0.1, at very high density, increasing density triggers re-entrant melting transitions from solid to hexatic and hexatic to fluid. As shown in Fig. 3a, for PHDS with , with increasing the density from to 1.6424, the system transforms from a hexatic phase with the short-range positional correlation, i.e., an exponential decay , to a solid with quasi-long-range positional order, i.e., a power law decay with , where the six fold orientation correlation function  Qi et al. (2014) remains almost the same (Fig. 3b). At much higher density, increasing the density from to 3.67, changes again from a power law decay to an exponential decay with almost the same (Fig. 3b), which suggests a re-entrant transition from solid to hexatic phase. For hard-disk systems with higher polydispersity, i.e., , we plot as functions of in Fig. 3c, and one can see that with increasing , first increases from 0 to around 0.8 indicating the formation of an ordered phase, and further increasing leads to the drop of to 0 implying a re-entrant melting transition into a disordered phase. By checking the scaling of in the system, we ensure that the ordered phase formed is a hexatic phase.

To understand the mechanism behind the re-entrant melting of hexatic phase, we plot as functions of in the lower panel of Fig. 3c, from which one can find a clear correlation between the change of and . Here we ensure that our simulations at high density have been well equilibrated and independent with the initial configurations (Supplementary Figure 5). Actually, it has been shown that the equilibration in systems of continuously polydisperse particles is much faster than that in monodisperse systems Berthier et al. (2016); Ninarello et al. (2017); Ikeda et al. (2017). At high density, during the decrease of , increases significantly to the value even higher than that in the low density fluid. This suggests that by increasing the density of the system, the standard deviation of particle size increases, and the combinatorial entropy in the system associated with the variation of particle size increases, which stabilizes the amorphous structures against the ordered ones in the system.

Moreover, along with the decrease of with density at high pressure, we do not observe any Mayer-Wood loops in EOS (Supplementary Figure 4). This implies that the re-entrant melting of hexatic phase at high density is continuous, which is similar to that in soft particle systems Zu et al. (2016); Dudalov et al. (2014). Furthermore, it was shown that the strong particle size fractionation in polydisperse hard spheres eliminates the possibility of re-entrant melting in 3D polydisperse hard-sphere solids Fasolo and Sollich (2003). Therefore, in Fig. 3d, we plot the probability density distribution of particle size for PHDS of various density with . One can see that within the whole density range, there is always a single peak in at fixed , indicating no particle size fractionation in the system. This also suggests that re-entrant melting can indeed exist in 2D polydisperse particle systems, if there is no particle size fractionation Bartlett and Warren (1999). Here we note that although all our simulations starting with different particle size distributions converge to the same result at fixed polydispersity and density, there still exist a possibility that the system may posses an equilibrium multimodal particle size distribution which is not accessible with direct simulations Bartlett and Warren (1999).

In addition, to check whether the amorphous phase formed at high density is a fluid or glass, we perform event driven molecular dynamics (EDMD) simulations starting from the equilibrated configurations obtained from our MC simulations at . In the EDMD simulation, the system evolves via a time-ordered sequence of elastic collision events, which are described by Newton’s equations of motion. The spheres move at constant velocities between collisions, and the velocities of the respective particles are updated when a collision occurs. All collisions are elastic and preserve energy and momentum Rapaport (2004). The calculated mean square displacement (MSD) for various densities are shown in Fig. 3e. One can see that not only the diffusion coefficient increases by nearly two orders of magnitude along with the re-entrant melting of hexatic phase (Fig. 3e inset), but also the plateau in MSD almost disappears. This suggests that the high density amorphous phase is a diffusive fluid, and in the equilibrium sedimentation of such PHDS of Gaussian-like particle size distribution, there should be a “floating hexatic phase” sandwiched in two fluids. For PHDS with , with increasing the density, we do not observe any ordered phase in our simulations.

Iii Discussions

By performing large scale computer simulations, we investigate the effect of particle size polydispersity on the phase behaviour of hard-disk systems. We find that with increasing the size polydispersity of hard disks from zero, the first-order melting transition from hexatic phase to fluid becomes weaker, and completely switches to be continuous at following the celebrated KTHNY scenario. Simultaneously, the density range for stable hexatic phase gets substantially enlarged. Compared with monodisperse hard-disk systems, where hexatic phase is only stable in a very small range of packing fraction of  Bernard and Krauth (2011), the packing fraction range for stable hexatic phase in PHDS with is about . This suggests new directions in searching for the hexatic phase in 2D polydisperse particle systems. More interestingly, in PHDS with even higher polydispersity, i.e., , we find that at very high density, increasing density of the system can trigger re-entrant transitions from solid to hexatic and hexatic to fluid phases depending on . In polydisperse systems, re-entrant transitions, especially the re-entrant melting, were originally predicted theoretically for 3D hard-sphere crystals Bartlett and Warren (1999), but proven impossible due to the strong fractionation that was not taken into account in the theory Fasolo and Sollich (2003). We find that the absence of strong fractionation in 2D polydisperse systems can indeed induce re-entrant transitions at high density. These show a new difference between phase transitions in polydisperse hard-sphere systems in 2D and 3D, and suggest a new direction in investigating phase transitions in 2D systems by considering the particle size polydispersity, which was overlooked in the past.

Iv Methods

We perform Monte Carlo simulations in the semigrand canonical ensemble () using a square simulation box with periodic boundary conditions in both directions, in which each particle can randomly change its diameter under the control of Eq. (1). To accelerate the equilibration and sampling, we implement the rejection free event-chain MC algorithm, with which the pressure in the system can be calculated using the mean excess chain displacement Michel et al. (2013)


where and are the initial and final positions of each event chain along the direction of the chain, respectively. and are the chain length and number density of the particles, respectively, with calculating the average over all event chains. For each simulation, we perform about MC sweeps, and each MC sweep on average consists of 1 event chain with the chain length of and 500 trials of randomly changing the diameter of a randomly selected particle.

V Author Contributions

R.N. conceived the research. P.S.R. performed the event chain Monte Carlo simulations. Q.-L.L. performed the event driven molecular dynamics simulations. All authors analysed the results and wrote the manuscript.

Vi Competing interests

The authors declare no competing interests.

Vii Code availability

Codes used for the numeric simulations are available on request from R Ni at Nanyang Technological University, Singapore.

Viii Data availability

All data that support the findings of this study are available from the corresponding author on reasonable request.

We thank Dr. Saurish Chakrabarty for helpful discussions. This work is supported by Nanyang Technological University Start-Up Grant (NTU-SUG: M4081781.120), the Academic Research Fund from Singapore Ministry of Education (M4011616.120 and M4011873.120), and the Advanced Manufacturing and Engineering Young Individual Research Grant (A1784C0018) by the Science and Engineering Research Council of Agency for Science, Technology and Research Singapore. We are grateful to the National Supercomputing Centre (NSCC) of Singapore for supporting the numerical calculations.


  • Strandburg (1988) Katherine J. Strandburg, “Two-dimensional melting,” Rev. Mod. Phys. 60, 161–207 (1988).
  • Dash (1999) J. G. Dash, “History of the search for continuous melting,” Rev. Mod. Phys. 71, 1737–1743 (1999).
  • Gasser (2009) U. Gasser, “Crystallization in three- and two-dimensional colloidal suspensions,” Journal of Physics: Condensed Matter 21, 203101 (2009).
  • Kosterlitz and Thouless (1973) J.M. Kosterlitz and D.J. Thouless, “Ordering, metastability and phase transitions in two-dimensional systems,” Journal of Physics C: Solid State Physics 6, 1181 (1973).
  • Halperin and Nelson (1978) B. I. Halperin and David R. Nelson, “Theory of two-dimensional melting,” Phys. Rev. Lett. 41, 121–124 (1978).
  • Nelson and Halperin (1979) David R. Nelson and B. I. Halperin, “Dislocation-mediated melting in two dimensions,” Phys. Rev. B 19, 2457–2484 (1979).
  • Young (1979) A. P. Young, “Melting and the vector coulomb gas in two dimensions,” Phys. Rev. B 19, 1855–1866 (1979).
  • Berezinskii (1971) V. Berezinskii, “Destruction of long range order in one-dimensional and two-dimensional systems having a continuous symmetry group. i. classical systems,” Sov. Phys. JETP 32, 493 (1971).
  • Binder et al. (2002) Kurt Binder, Surajit Sengupta,  and Peter Nielaba, “The liquid-solid transition of hard discs: first-order transition or kosterlitz-thouless-halperin-nelson-young scenario?” Journal of Physics: Condensed Matter 14, 2323 (2002).
  • Chui (1982) S. T. Chui, “Grain-boundary theory of melting in two dimensions,” Phys. Rev. Lett. 48, 933–935 (1982).
  • Saito (1982) Y. Saito, “Melting of dislocation vector systems in two dimensions,” Phys. Rev. Lett. 48, 1114–1117 (1982).
  • Zahn et al. (1999) K. Zahn, R. Lenke,  and G. Maret, “Two-stage melting of paramagnetic colloidal crystals in two dimensions,” Phys. Rev. Lett. 82, 2721–2724 (1999).
  • Karnchanaphanurach et al. (2000) Pallop Karnchanaphanurach, Binhua Lin,  and Stuart A. Rice, “Melting transition in a quasi-two-dimensional colloid suspension: Influence of the colloid-colloid interaction,” Phys. Rev. E 61, 4036–4044 (2000).
  • Han et al. (2008) Y. Han, N. Y. Ha, A. M. Alsayed,  and A. G. Yodh, “Melting of two-dimensional tunable-diameter colloidal crystals,” Phys. Rev. E 77, 041406 (2008).
  • Rice (2009) Stuart A. Rice, “Structure in confined colloid suspensions,” Chemical Physics Letters 479, 1 – 13 (2009).
  • Murray and Van Winkle (1987) C. A. Murray and D. H. Van Winkle, “Experimental observation of two-stage melting in a classical two-dimensional screened coulomb system,” Phys. Rev. Lett. 58, 1200–1203 (1987).
  • Marcus and Rice (1996) Andrew H. Marcus and Stuart A. Rice, ‘‘Observations of first-order liquid-to-hexatic and hexatic-to-solid phase transitions in a confined colloid suspension,” Phys. Rev. Lett. 77, 2577–2580 (1996).
  • von Grünberg et al. (2004) H. H. von Grünberg, P. Keim, K. Zahn,  and G. Maret, “Elastic behavior of a two-dimensional crystal near melting,” Phys. Rev. Lett. 93, 255703 (2004).
  • Keim et al. (2007) P. Keim, G. Maret,  and H. H. von Grünberg, “Frank’s constant in the hexatic phase,” Phys. Rev. E 75, 031402 (2007).
  • Xu and Rice (2008) Xinliang Xu and Stuart A. Rice, “Liquid-to-hexatic phase transition in a quasi-two-dimensional colloid system,” Phys. Rev. E 78, 011602 (2008).
  • Bernard and Krauth (2011) Etienne P. Bernard and Werner Krauth, “Two-step melting in two dimensions: First-order liquid-hexatic transition,” Phys. Rev. Lett. 107, 155704 (2011).
  • Engel et al. (2013) Michael Engel, Joshua A. Anderson, Sharon C. Glotzer, Masaharu Isobe, Etienne P. Bernard,  and Werner Krauth, “Hard-disk equation of state: First-order liquid-hexatic transition in two dimensions with three simulation methods,” Phys. Rev. E 87, 042134 (2013).
  • Kapfer and Krauth (2015) Sebastian C. Kapfer and Werner Krauth, “Two-dimensional melting: From liquid-hexatic coexistence to continuous transitions,” Phys. Rev. Lett. 114, 035702 (2015).
  • Anderson et al. (2017) Joshua A. Anderson, James Antonaglia, Jaime A. Millan, Michael Engel,  and Sharon C. Glotzer, “Shape and symmetry determine two-dimensional melting transitions of hard regular polygons,” Phys. Rev. X 7, 021001 (2017).
  • Li and Ciamarra (2018) Yan-Wei Li and Massimo Pica Ciamarra, “Role of cell deformability in the two-dimensional melting of biological tissues,” Phys. Rev. Materials 2, 045602 (2018).
  • Deutschländer et al. (2013) Sven Deutschländer, Tobias Horn, Hartmut Löwen, Georg Maret,  and Peter Keim, “Two-dimensional melting under quenched disorder,” Phys. Rev. Lett. 111, 098301 (2013).
  • Qi and Dijkstra (2015) Weikai Qi and Marjolein Dijkstra, “Destabilisation of the hexatic phase in systems of hard disks by quenched disorder due to pinning on a lattice,” Soft Matter 11, 2852 (2015).
  • Russo and Wilding (2017) John Russo and Nigel B. Wilding, ‘‘Disappearance of the hexatic phase in a binary mixture of hard disks,” Phys. Rev. Lett. 119, 115702 (2017).
  • Thorneywork et al. (2017) Alice L. Thorneywork, Joshua L. Abbott, Dirk G. A. L. Aarts,  and Roel P. A. Dullens, “Two-dimensional melting of colloidal hard spheres,” Phys. Rev. Lett. 118, 158001 (2017).
  • Kawasaki et al. (2007) T. Kawasaki, T. Araki,  and H. Tanaka, ‘‘Correlation between dynamic heterogeneity and medium-range order in two-dimensional glass-forming liquids,” Phys. Rev. Lett. 99, 215701 (2007).
  • Kawasaki and Tanaka (2011) T. Kawasaki and H. Tanaka, “Structural signature of slow dynamics and dynamic heterogeneity in two-dimensional colloidal liquids: Glassy structural order,” J Phys. Condens. Matter. 23, 194121 (2011).
  • Russo and Tanaka (2015) J. Russo and H. Tanaka, “Assessing the role of static length scales behind glassy dynamics in polydisperse hard disks,” Proc. Natl Acad. Sci. USA 112, 6920 (2015).
  • Fasolo and Sollich (2003) Moreno Fasolo and Peter Sollich, “Equilibrium phase behavior of polydisperse hard spheres,” Phys. Rev. Lett. 91, 068301 (2003).
  • Kofke and Glandt (1988) David A. Kofke and Eduardo D. Glandt, “Monte carlo simulation of multicomponent equilibria in a semigrand canonical ensemble,” Molecular Physics 64, 1105–1131 (1988).
  • Bolhuis and Kofke (1996) Peter G. Bolhuis and David A. Kofke, “Monte carlo study of freezing of polydisperse hard spheres,” Phys. Rev. E 54, 634–643 (1996).
  • Kofke and Bolhuis (1999) David A. Kofke and Peter G. Bolhuis, “Freezing of polydisperse hard spheres,” Phys. Rev. E 59, 618–622 (1999).
  • Pronk and Frenkel (2004) Sander Pronk and Daan Frenkel, “Melting of polydisperse hard disks,” Phys. Rev. E 69, 066123 (2004).
  • Mayer and Wood (1965) Joseph E. Mayer and Wm. W. Wood, “Interfacial tension effects in finite, periodic, two‐dimensional systems,” The Journal of Chemical Physics 42, 4268–4274 (1965).
  • Gribova et al. (2011) Nadezhda Gribova, Axel Arnold, Tanja Schilling,  and Christian Holm, “How close to two dimensions does a lennard-jones system need to be to produce a hexatic phase?” The Journal of Chemical Physics 135, 054514 (2011).
  • Ferdinand and Fisher (1969) Arthur E. Ferdinand and Michael E. Fisher, “Bounded and inhomogeneous ising models. i. specific-heat anomaly of a finite lattice,” Phys. Rev. 185, 832–846 (1969).
  • Santen and Krauth (2000) L. Santen and W. Krauth, “Absence of thermodynamic phase transition in a model glass former,” Nature 405, 500 (2000).
  • Qi et al. (2014) Weikai Qi, A.P. Gantapara,  and Marjolein Dijkstra, ‘‘Two-stage melting induced by dislocations and grain boundaries in monolayers of hard spheres,” Soft Matter 10, 5449 (2014).
  • Berthier et al. (2016) Ludovic Berthier, Daniele Coslovich, Andrea Ninarello,  and Misaki Ozawa, “Equilibrium sampling of hard spheres up to the jamming density and beyond,” Phys. Rev. Lett. 116, 238002 (2016).
  • Ninarello et al. (2017) Andrea Ninarello, Ludovic Berthier,  and Daniele Coslovich, ‘‘Models and algorithms for the next generation of glass transition studies,” Phys. Rev. X 7, 021039 (2017).
  • Ikeda et al. (2017) Harukuni Ikeda, Francesco Zamponi,  and Atsushi Ikeda, “Mean field theory of the swap monte carlo algorithm,” The Journal of Chemical Physics 147, 234506 (2017).
  • Zu et al. (2016) Mengjie Zu, Jun Liu, Hua Tong,  and Ning Xu, “Density affects the nature of the hexatic-liquid transition in two-dimensional melting of soft-core systems,” Phys. Rev. Lett. 117, 085702 (2016).
  • Dudalov et al. (2014) D.E. Dudalov, Yu.D. Fomin, E.N. Tsiok,  and V.N. Ryzhov, “Anomalous melting scenario of the two-dimensional core-softened system,” arXiv:1311.7534v4  (2014).
  • Bartlett and Warren (1999) Paul Bartlett and Patrick B. Warren, “Reentrant melting in polydispersed hard spheres,” Phys. Rev. Lett. 82, 1979–1982 (1999).
  • Rapaport (2004) D. C. Rapaport, The Art of Molecular Dynamics Simulation (Cambridge University Press, New York, 2004).
  • Michel et al. (2013) M. Michel, S.C. Karpfer,  and W. Krauth, “Generalized event-chain monte carlo: Constructing rejection-free global-balance algorithms from in nitesimal steps,” J. Chem. Phys , 054116 (2013).
Comments 0
Request Comment
You are adding the first comment!
How to quickly get a good reply:
  • Give credit where it’s due by listing out the positive aspects of a paper before getting into which changes should be made.
  • Be specific in your critique, and provide supporting evidence with appropriate references to substantiate general statements.
  • Your comment should inspire ideas to flow and help the author improves the paper.

The better we are at sharing our knowledge with each other, the faster we move forward.
The feedback must be of minimum 40 characters and the title a minimum of 5 characters
Add comment
Loading ...
This is a comment super asjknd jkasnjk adsnkj
The feedback must be of minumum 40 characters
The feedback must be of minumum 40 characters

You are asking your first question!
How to quickly get a good answer:
  • Keep your question short and to the point
  • Check for grammar or spelling errors.
  • Phrase it like a question
Test description