High-resolution mapping of the phase boundary of the honeycomb-lattice Blume--Capel ferromagnet
Listen
Radio episode about this paper
Transcript
Introduction to the show: ident: Condensed Matter. The day in condensed matter: materials, phases and the measurements under them.
Kai: We're looking at the paper: "High-resolution mapping of the phase boundary of the honeycomb-lattice Blume--Capel ferromagnet".
Mira: Here's the gist of the paper. High-resolution mapping of the phase boundary of the honeycomb-lattice Blume–Capel ferromagnet presents a detailed numerical study to precisely locate the first-order transition line and tricritical point for this spin-1 model on a honeycomb lattice The result is supported by two independent simulation schemes whose crossing estimates are mutually consistent.
Kai: Mapping the Phase Boundary The study employs Wang–Landau estimates of the joint density of states combined with field mixing to map the first-order transition line and locate the tricritical point at (∆t, Tt) =
1.4819(2), 0.4022(7): .
Mira: Histogram reweighting of Metropolis simulations for L = 36 and 48 provides an independent consistency check, with a crossing temperature near 0.4039, compatible with the Wang–Landau extrapolation.
Kai: The analysis maps the continuous branch of the phase boundary from the same density-of-states data, obtaining Tc(∆ = 0) = 1.157.
Title and authors: Kai: Now, moving into the summary of this paper, what they’re actually saying about their findings so far? They are essentially describing how they went about the mapping and where they landed on the phase diagram.
Mira: The summary is focused on presenting that they used Wang–Landau estimates of the joint density of states combined with field mixing to get the mapping done and locate that tricritical point at delta t one point four eight one nine with two and Tt zero point four zero two two with seven.
Lev: So it’s essentially saying they used these advanced techniques to precisely find that tricritical region on this geometry, which is where the first-order and continuous transition lines meet.
Kai: And they also mention the independent consistency check from histogram reweighting of Metropolis simulations for L equals thirty-six and forty-eight, which confirms their crossing temperature near zero point four zero three nine.
Mira: That temperature crossing near zero point four zero three nine is compatible with the Wang–Landau extrapolation, which is a key piece of evidence supporting the main results.
Lev: They are basically saying that the simulation methods agree on where the critical behavior shifts from continuous to first-order at that specific point.
Kai: So it’s not just one method pointing to a result; it’s two different ways of getting there, which makes the conclusion feel much stronger.
Mira: They also mention how they obtained the field mixing parameter r as negative one point five eight six with nineteen from the field mixing analysis, and then compare that to the value from the multicanonical coexistence line being negative one point five eight seven with five.
Lev: That comparison between those two values for r is important because it shows how much they rely on both of these different methods to pin down this parameter accurately.
Kai: And they also detail how the scaling exponents yt equals nine over five are consistent with that value within zero point seven sigma, which is good validation against expectations for the tricritical Ising class.
Mira: So the summary emphasizes that not only did they find the tricritical point at delta t one point four eight one nine with two and Tt zero point four zero two two with seven, but they also confirmed all those scaling properties match what we’d expect from the tricritical Ising universality class.
Lev: That confirms that their numerical results are not just random fluctuations but are actually pointing towards a known physical behavior in the spin-one model.
Kai: They also explain how they use the field mixing analysis to systematically identify the appropriate scaling fields and operators near that tricritical point, which is a systematic way to look at what's happening there.
Mira: It’s about using that specific mathematical framework to understand the physics of what happens right at the critical point.
The paper's summary: Kai: Now, let’s talk about the actual improvements this paper suggests they made over what came before it in terms of methodology and what that means for future research.
Mira: The primary improvement is mapping out the phase boundary on the honeycomb lattice itself, which is a geometry that has low connectivity, and finding the tricritical point.
Lev: This is important because this geometry allows them to investigate how connectivity influences non-universal parameters associated with the tricritical region in a way that other lattices can't easily do.
Kai: And they show how field mixing helps identify the appropriate scaling fields and operators, which is useful for understanding the physics right at that critical point.
Mira: They also show how histogram reweighting provides a way to access larger lattice sizes beyond what you can do with a simple twoparameter Wang–Landau calculation.
Lev: That means they’re pushing the limits of what's feasible in terms of size, which is crucial for getting those clean results without huge statistical noise.
Kai: And they also get better estimates for the interface tension along the first-order line, describing how it approaches zero as a function of distance from the tricritical point with that exponent mu equals one point two six five with twenty-six.
Mira: That exponent mu one point two six five is useful because it describes exactly how interfacial energy scales near criticality.
Lev: So they’re getting a better description of that scaling behavior, which helps us predict how the interface energy behaves as you get closer to the critical point.
Kai: And they also confirm that the coexistence line's slope at the tricritical point gives an independent determination of r equals negative one point five eight seven with five, which is four times more precise than just using field mixing analysis.
Mira: That higher precision for r is significant because it shows that the slope of that line at the tricritical point is a very strong constraint on how you model these transitions.
Lev: So they’re not just getting one number; they’re getting several constraints, which strengthens the picture of what's happening at that specific junction.
Kai: The paper also notes a few things, like how the equal-height condition on the two maxima on the distribution of Q becomes less sensitive to s at large L.
Mira: That suggests that as you increase lattice size, you need to be more careful about how much you are trusting those scaling field definitions for very large systems.
Lev: So they’re flagging a limitation of their own method when trying to push the limits of size, which is important for future error correction research.
The paper's improvements: Kai: So we’ve got the final thoughts from both of us to wrap up what this paper means overall about this work on the "High-resolution mapping of the phase boundary of the honeycomb-lattice Blume--Capel ferromagnet." We've mapped out the key findings for listeners.
Mira: Essentially, they’ve established that they found that by combining two independent simulation schemes, their crossing estimates are mutually consistent.
Lev: This consistency is a very strong validation of their location of the tricritical point at delta t one point four eight one nine with two and Tt zero point four zero two two with seven.
Kai: So, they’ve successfully established that they found that by combining those methods, their crossing estimates are mutually consistent.
Mira: And this consistency is a very strong validation of the tricritical point at delta t one point four eight one nine with two and Tt zero point four zero two two with seven.
Lev: From an engineering standpoint, it’s about knowing exactly where the critical behavior shifts when you move from theory to real hardware.
Lev: For me, it’s about knowing exactly where the critical behavior shifts when you move from theory to real hardware.
Conclusion: Kai: So, to wrap up, this paper shows they mapped out the phase boundary for a spin one Blume Capel ferromagnet on a honeycomb lattice and located that tricritical point at delta t one point four eight one nine with two and Tt zero point four zero two two with seven.
Mira: Yeah, it’s a detailed numerical study using Wang Landau estimates combined with field mixing to pin down the first order transition line and that tricritical junction.
Lev: The real takeaway here is the consistency—they used histogram reweighting for L equals thirty-six and forty-eight, and those crossing estimates matched what the main Wang Landau extrapolation predicted.
Kai: That consistency between two different simulation schemes is what makes this result solid, showing that they’ve really nailed the location of that transition point.
Mira: And they also got a precise value for the field mixing parameter r, which came out as negative one point five eight six with nineteen from the field mixing analysis, and then compared it to negative one point five eight seven with five from the multicanonical coexistence line.
Lev: That comparison between those two values for r is important because it shows how much they rely on both of these different methods to get a solid number for that parameter.
Kai: They also confirmed that the scaling exponents, like yt equals nine over five, matched what we expect from the tricritical Ising class within a small error margin.
Mira: And they nailed the interface tension scaling with mu equals one point two six five with twenty-six, which describes how that interface energy approaches zero as you get closer to that tricritical point.
Lev: That exponent mu is useful because it tells us exactly how the interfacial energy scales near criticality, which is something we need to know when building systems for real hardware.
Kai: So, this whole paper on the high resolution mapping of the phase boundary of the honeycomb-lattice Blume Capel ferromagnet gives us a very clear picture of where those first order transitions and tricritical points live in this spin one model.
Mira: It provides a unified view of both branches—the continuous line and that first order line—meeting at that specific junction.
Lev: For me, it’s about knowing exactly where the critical behavior shifts when you move from theory to real hardware, and this mapping helps us understand those crossover points.
Kai: And it shows how important it is to use multiple independent simulation schemes when trying to pinpoint something as subtle as a tricritical point.
Mira: This work sets a clear benchmark for how these high resolution mappings should be done on lower dimensional lattices like the honeycomb one.
Lev: Moving forward, we need to see if this behavior holds up when we start looking at those larger lattices you mentioned earlier.
Eluiz C.M. Filho, Claudio J. DaSilva
Instituto Federal de Goiáas · Instituto de Fósica, Universidade Federal de Goiáas
cond-mat.stat-mech
Submitted: 2026-10-08
Updated: 2026-10-28
License: http://creativecommons.org/licenses/by/4.0/
Importance score: 89/100
The gist: High-resolution mapping of the phase boundary of the honeycomb-lattice Blume–Capel ferromagnet presents a detailed numerical study to precisely locate the first-order transition line and…
Key concepts
- Blume–Capel (BC) Model
- This is a spin-1 Ising model that includes a single-ion crystal-field anisotropy term. It describes how the magnetic interactions and the energy penalty from the crystal field compete to determine the system's ordered or disordered state at different temperatures and field strengths.
- Tricritical Point
- This is a specific point on the phase diagram where a line of continuous transitions meets a line of first-order transitions. Identifying this point is crucial because it marks the transition between different types of critical behavior in the system.
- Field Mixing Analysis
- This technique analyzes scaling fields near the tricritical point to determine mixing parameters (like r). It helps quantify how the continuous and first-order transition lines interact, providing a precise measure of their relationship and universality class.
Terminology
Summary
High-resolution mapping of the phase boundary of the honeycomb-lattice Blume–Capel ferromagnet presents a detailed numerical study to precisely locate the first-order transition line and tricritical point for this spin-1 model on a honeycomb lattice The result is supported by two independent simulation schemes whose crossing estimates are mutually consistent. p. 9 4819(2), 0. p. 0 Wang–Landau estimates of the joint density of states Ω(E1, E2), combined
with field mixing, are used to map the first-order transition line and locate the tricritical point at
(∆t, Tt) = [1. 4819(2), 0. Wang–Landau estimates of the joint density of states Ω(E1, E2), combined
with field mixing, are used to map the first-order transition line and locate the tricritical point at
(∆t, Tt) = [1. 4819(2), 0. Wang–Landau estimates of the joint density of states Ω(E1, E2), combined
with field mixing, are used to map the first-order transition line and locate the tricritical point at
(∆t, Tt) = [1. 4819(2), 0. Wang–Landau estimates of the joint density of states Ω(E1, E2), combined
with field mixing, are used to map the first-order transition line and locate the tricritical point at
(∆t, Tt) = [1. 4819(2), 0. 4819(2), 0. The result is supported by two
independent simulation schemes — Wang–Landau sampling of the joint density of states and histogram
reweighting from Metropolis runs at larger sizes — whose
crossing estimates are mutually consistent. 4819(2), 0. The result is supported by two
independent simulation schemes — Wang–Landau sampling of the joint density of states and histogram
reweighting from Metropolis runs at larger sizes — whose
crossing estimates are mutually consistent. (Dated: October 9, 2026)
We present a high-resolution numerical study of the spin-1 Blume–Capel ferromagnet on the honeycomb lattice (z = 3). Wang–Landau estimates of the joint density of states Ω(E1, E2), combined
with field mixing, are used to map the first-order transition line and locate the tricritical point at
(∆t, Tt) = [1.Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Mapping the Phase Boundary
The study employs Wang–Landau estimates of the joint density of states combined with field mixing to map the first-order transition line and locate the tricritical point at (∆t, Tt) = [1.4819(2), 0.4022(7)]. Histogram reweighting of Metropolis simulations for L = 36 and 48 provides an independent consistency check, with a crossing temperature near 0. p. 0 4819(2), 0. Histogram reweighting of Metropolis simulations for L = 36 and
48 provides an independent consistency check, with a crossing temperature near 0. Histogram reweighting of Metropolis simulations for L = 36 and
48 provides an independent consistency check, with a crossing temperature near 0. 4039, compatible
with the Wang–Landau extrapolation. 265(26), consistent with the tricritical prediction
(d − 1)/yg = 5/4. Finally, we map the continuous branch of the phase boundary from the same
density-of-states data, obtaining Tc(∆ = 0) = 1.Where the paper says it
Where the paper says it
Where the paper says it
Characterization of Critical Parameters
The field mixing analysis yields r = −1.586(19), while the slope of the multicanonical coexistence line yields the consistent and more precise estimate r = −1. p. 0 From the field-mixing analysis we obtain r = −1. 586(19),
while the slope of the multicanonical coexistence line yields the consistent and more precise estimate
r = −1. 586(19),
while the slope of the multicanonical coexistence line yields the consistent and more precise estimate
r = −1. Normalizing by the zero-temperature field ∆0 = zJ/2, the available honeycomb,
square, triangular, and simple-cubic estimates cluster near r/∆0 ≃ −1. Normalizing by the zero-temperature field ∆0 = zJ/2, the available honeycomb,
square, triangular, and simple-cubic estimates cluster near r/∆0 ≃ −1. 06, corresponding to a magnitude about 6% larger than the mean-field value −1; no systematic dependence on coordination
number or dimensionality is resolved over the range examined.Where the paper says it
Where the paper says it
Where the paper says it
Finite-Size Scaling Tests
Finite-size scaling of σQ gives yt = 1.795(7), consistent with the tricritical Ising value 9/5, while the fourth-order cumulant scaling is compatible with yg = 4/5. p. 0 Finite-size scaling of σQ gives
yt = 1. 795(7), consistent with the tricritical Ising value 9/5, while the fourth-order cumulant scaling
is compatible with yg = 4/5. Deep in the first-order regime, the cumulant approaches the expected
two-phase limit UQ → 2/3. Multicanonical simulations up to L = 64, using weights constructed
directly from the Wang–Landau density of states for L ≤ 24 and rescaled and refined for larger
systems, yield an interface-tension exponent µ = 1. Multicanonical simulations up to L = 64, using weights constructed
directly from the Wang–Landau density of states for L ≤ 24 and rescaled and refined for larger
systems, yield an interface-tension exponent µ = 1. 265(26), consistent with the tricritical prediction
(d − 1)/yg = 5/4.Where the paper says it
Where the paper says it
Where the paper says it
Tricritical Point Determination
The tricritical point (Tt, ∆t) is determined by probing the behavior of the fourth-order cumulant UQ, where curves for different L intersect at Tt. The extrapolated value for the tricritical field is ∆t = 1. p. 5 23(3), that is r/∆0 = −1. The
larger uncertainty reflects the fact that the slope must
be extrapolated by 8% in temperature beyond the range
of the tabulated data; the result is, however, insensitive
to which of the two literature estimates of the threedimensional tricritical point is adopted, the alternative
value of Deng and Bl¨ote [24] yielding r/∆0 = −1. Wang–Landau estimates of the joint density of states Ω(E1, E2), combined
with field mixing, are used to map the first-order transition line and locate the tricritical point at
(∆t, Tt) = [1. 4819(2), 0. Wang–Landau estimates of the joint density of states Ω(E1, E2), combined
with field mixing, are used to map the first-order transition line and locate the tricritical point at
(∆t, Tt) = [1. 4819(2), 0. Wang–Landau estimates of the joint density of states Ω(E1, E2), combined
with field mixing, are used to map the first-order transition line and locate the tricritical point at
(∆t, Tt) = [1. 4819(2), 0.Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Interface Tension and Coexistence Line
The multicanonical results provide access to the interface tension along the first-order line, with its approach to zero described by (Tt −T)µ with µ = 1.265(26), consistent with the tricritical Ising value 1/yg = 5/4. p. 9 Its approach to zero is described by (Tt −T)
µ
with µ = 1. 265(26), consistent with the tricritical Ising
value 1/yg = 5/4 and incompatible with 1/yt = 5/9. Its approach to zero is described by (Tt −T)
µ
with µ = 1. 265(26), consistent with the tricritical Ising
value 1/yg = 5/4 and incompatible with 1/yt = 5/9. Its approach to zero is described by (Tt −T)
µ
with µ = 1. 265(26), consistent with the tricritical Ising
value 1/yg = 5/4 and incompatible with 1/yt = 5/9.Where the paper says it
Where the paper says it
Where the paper says it
Summary and Outlook
The study successfully maps the phase boundary on the honeycomb lattice and locates the tricritical point at (∆t, Tt) = [1.4819(2), 0.4022(7)]. The result is supported by two independent simulation schemes whose crossing estimates are mutually consistent. The authors also obtained the field-mixing parameter r = −1. p. 8 002 (syst), (26)
where the systematic error combines the uncertainty of
Tt with the spread obtained by anchoring the polynomial at (Tt, ∆t) or raising it to fourth order. This determination uses neither the operator Q nor the mixing
parameter s, and is thus independent of the field-mixing
value r = −1. 002 (syst), (26)
where the systematic error combines the uncertainty of
Tt with the spread obtained by anchoring the polynomial at (Tt, ∆t) or raising it to fourth order. This determination uses neither the operator Q nor the mixing
parameter s, and is thus independent of the field-mixing
value r = −1. We have also
obtained the field-mixing parameter r = −1. 586(19),
shown that the combination r/∆0 is exactly −1 in meanfield theory for any coordination number, collected the
corresponding values for four lattices spanning z = 3–6
and d = 2–3 — three of which had not previously been
determined — and found finite-size scaling consistent
with the expected tricritical Ising exponents yt = 9/5
and yg = 4/5 over the available lattice sizes. We have also
obtained the field-mixing parameter r = −1. 586(19),
shown that the combination r/∆0 is exactly −1 in meanfield theory for any coordination number, collected the
corresponding values for four lattices spanning z = 3–6
and d = 2–3 — three of which had not previously been
determined — and found finite-size scaling consistent
with the expected tricritical Ising exponents yt = 9/5
and yg = 4/5 over the available lattice sizes. Panel
(c) enlarges the tricritical region in this representation; the local slope of the coexistence line determines
the mixing parameter r. The agreement between the
phase-boundary estimates obtained from Wang–Landau,
histogram-reweighting, and multicanonical analyses provides a unified picture of the two branches and their tricritical junction. For comparison,
the dashed reference behavior with exponent 1/yt = 5/9,
associated with the thermal direction across the phase
boundary, is clearly incompatible with the data and gives
χ
2/ν = 347/3 with the same correction term. The accessible temperature window is bounded on
both sides.Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
--- Page 3 ---
The Blume–Capel (BC) model [1, 2] is a spin-1 Ising model with a single-ion crystal-field anisotropy. Its phase diagram contains a line of continuous transitions and a line of first-order transitions separating ordered and disordered phases; the two lines meet at a tricritical point. p. 0 INTRODUCTION
The Blume–Capel (BC) model [1, 2] is a spin-1 Ising
model with a single-ion crystal-field anisotropy. Its phase
diagram contains a line of continuous transitions and a
line of first-order transitions separating ordered and disordered phases; the two lines meet at a tricritical point.Where the paper says it
Model Definition
In zero magnetic field, the Hamiltonian is H = −J X ⟨ij⟩ σiσj + ∆X i σ 2 i = JE1 + ∆E2, (1) with σi ∈ (−1, 0, +1) and J > 0. We measure energies in units of J and temperatures in units of J/kB. p. 0 In zero magnetic field, the Hamiltonian is
H = −J
X
⟨ij⟩
σiσj + ∆X
i
σ
2
i = JE1 + ∆E2, (1)
with σi ∈ {−1, 0, +1} and J > 0. We measure energies
in units of J and temperatures in units of J/kB. Here,
N = L×L denotes the number of sites in our finite-lattice
convention, and E2 is the number of nonzero spins. At
T = 0, the ferromagnetic state is favored as long as its energy gain per spin, zJ/2, exceeds the crystal-field penalty,
placing the exact point (∆0 = zJ/2, T = 0) on the phase
boundary.Where the paper says it
Where the paper says it
Lattice Geometry Significance
Among regular two-dimensional lattices, the honeycomb lattice stands out because it is bipartite and has a coordination number of z = 3, the lowest among these lattices. This low coordination makes the honeycomb lattice particularly useful for examining how connectivity shifts non-universal thermodynamic quantities in the BC model. p. 0 Among regular two-dimensional lattices, the
honeycomb lattice stands out because it is bipartite and
has a coordination number of z = 3, the lowest among
these lattices. This low coordination makes the honeycomb lattice particularly useful for examining how connectivity shifts non-universal thermodynamic quantities
in the BC model. This low coordination makes the honeycomb lattice particularly useful for examining how connectivity shifts non-universal thermodynamic quantities
in the BC model. Since each site has only three nearest neighbors, the energetic contribution of interactions
between spins is reduced relative to lattices with greater
connectivity, such as the square (z = 4) and triangular
(z = 6) lattices.Where the paper says it
Where the paper says it
Numerical Methods Employed
Monte Carlo methods are among the principal numerical tools for studying phase transitions in spin systems. Relevant approaches include Metropolis sampling, histogram reweighting (HR), multicanonical methods (MUCA), and entropic sampling based on the Wang–Landau (WL) algorithm. p. 1 Monte Carlo methods are among the principal numerical tools for studying phase transitions in spin systems [15]. Relevant approaches include Metropolis sampling, histogram reweighting (HR), multicanonical methods (MUCA), and entropic sampling based on the Wang–
Landau (WL) algorithm [16, 17]. Relevant approaches include Metropolis sampling, histogram reweighting (HR), multicanonical methods (MUCA), and entropic sampling based on the Wang–
Landau (WL) algorithm [16, 17]. Among these methods,
the Wang–Landau algorithm allows the density of states
of a system to be estimated directly, enabling the reconstruction of thermodynamic quantities at different temperatures and control parameters.Where the paper says it
Where the paper says it
Joint Density of States
The use of a joint density of states in two energy variables allows the contributions associated with the interaction and crystal-field terms to be treated simultaneously, providing access to the parameter space (T, ∆) from a single estimate of the density of states.
Field Mixing Analysis
Near the tricritical point, β and µ are not themselves the relevant scaling fields; the latter are linear combinations λ = (µ − µt) + r (β − βt), g = (β − βt) + s (µ − µt). The operator conjugate to λ is Q = 1 / [1 - rs (n - s l)]. p. 2 Method of field mixing
Near the tricritical point, β and µ are not themselves
the relevant scaling fields; the latter are the linear combinations [12, 13]
λ = (µ − µt) + r (β − βt), (4)
g = (β − βt) + s (µ − µt), (5)
with s and r the size-dependent mixing parameters. The
operator conjugate to λ is
Q =
1
1 − rs
(n − s ℓ), (6)
where ℓ ≡ L
−dE1 and n ≡ L
−dE2.Where the paper says it
Histogram Reweighting
To access sizes beyond those feasible with the twoparameter WL calculation, conventional Metropolis simulations are performed at fixed (β0, µ0) for L = 36 and 48, accumulating the joint histogram H(E1, E2) and reweighting it.
Multicanonical Simulations
In the first-order regime, MUCA simulations are performed at fixed temperature, flattening the marginal distribution of E2. For L = 32, 48 and 64 the weights of the next smaller size were rescaled extensively. p. 2 (9)
For L ≤ 24, this sum is obtained directly from the Wang–
Landau density-of-states estimate, and the corresponding MUCA weights are supplied without a separate iterative construction stage. For L = 32, 48 and 64 the
weights of the next smaller size were rescaled extensively,
ln WL(n) ≃ (L/L′
)
d
ln WL′ (n) at fixed n = E2/N, and
then refined through
ln Wnew = ln Wold − ln H, (10)
where H is the histogram of E2 produced with Wold.Where the paper says it
--- Page 5 ---
P(E1, E2; β, µ) = 1 / Z Ω(E1, E2) e −βE1−µE2 (3). Observables that are not functions of (E1, E2) alone — in particular the magnetization — were accumulated as microcanonical averages ⟨M k⟩(E1, E2) during the production stage. p. 1 Once Ω is known, the joint probability at any β ≡ 1/T
and µ ≡ ∆/T follows without further sampling,
P(E1, E2; β, µ) = 1
Z
Ω(E1, E2) e
−βE1−µE2
. (3)
Observables that are not functions of (E1, E2) alone —
in particular the magnetization — were accumulated as
microcanonical averages ⟨|M|
k
⟩(E1, E2) during the production stage.Where the paper says it
First-Order Scaling
The pseudo-transition fields ∆∗ L (T) obtained from the field-mixing analysis follow the expected first-order finitesize behavior ∆∗ L = ∆∗∞ + b L−2, (12).
Tricritical Point Probing
At tricriticality, the appropriately scaled orderparameter distribution becomes size independent. This is probed through the fourth-order cumulant U L Q = 1 − ⟨Q˜4⟩ / 3 ⟨Q˜2⟩ squared, (13). p. 0 Finite-size scaling of σQ gives
yt = 1. 795(7), consistent with the tricritical Ising value 9/5, while the fourth-order cumulant scaling
is compatible with yg = 4/5.Where the paper says it
Scaling Exponents
Near the tricritical point the distribution of Q obeys PL(Q) = L(-yt) Po (L(-yt)) (Q − ⟨Q>), with the expected two-dimensional tricritical Ising value yt = 9/5 [25, 26]. Testing Eq. p. 6 Tricritical eigenvalue exponents
Near the tricritical point the distribution of Q
obeys [14]
PL(Q) = L
d−yt Po
L
d−yt
(Q − ⟨Q⟩)
, (18)
with the expected two-dimensional tricritical Ising value
yt = 9/5 [25, 26]. Testing Eq. Testing Eq. (18) requires the absolute
scale of Q and therefore both mixing parameters.Where the paper says it
Where the paper says it
Interface Tension Scaling
The interface tension is expected to vanish on approaching the tricritical point from the first-order side. The distance from the tricritical point is measured by g ∝ Tt − T and ξ ∼ g − 1/yg. p. 7 03616(13) 1. 490630(10)
In two dimensions Σ ∼ ξ
−(d−1) = ξ
−1
, and along the
first-order line the field λ vanishes by construction, so
that the distance from the tricritical point is measured
by g ∝ Tt − T and ξ ∼ g
−1/yg
.Where the paper says it
Coexistence Line Behavior
The transition fields of Table V follow the first-order form of Eq. (12): at T = 0.32 and 0.36 the L−2 fits. p. 7 32 reflects an alternation
of the residuals between L = 48 and 64 rather than curvature, and we inflate the error accordingly. Restricting
the fits to L ≤ 32 lowers Σ∞ by 4 × 10−4 and 5 × 10−4
at the two temperatures. These provide
three complementary checks connecting the multicanonical coexistence line to the results of the preceding sections. The coexistence line is the locus λ = 0 of
an analytic scaling field; it is therefore analytic at Tt,
where it continues smoothly into the critical line, and a
polynomial in T is the appropriate form.Where the paper says it
Where the paper says it
Critical Exponent Consistency
The resulting exponent yt = 9/5 is consistent with 9/5 within 0.7σ. Imposing yt = 9/5 returns T = 0. p. 6 7σ; the quoted error is dominated by the uncertainty of Tt. Equivalently, imposing yt = 9/5 returns T = 0. 7σ; the quoted error is dominated by the uncertainty of Tt. Equivalently, imposing yt = 9/5 returns T = 0. Equivalently, imposing yt = 9/5 returns T = 0. 40175, again within 1. Equivalently, imposing yt = 9/5 returns T = 0. 40175, again within 1. 40175, again within 1. 3σ of
the cumulant-crossing estimate. 40175, again within 1. 3σ of
the cumulant-crossing estimate.Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
--- Page 7 ---
The pseudo-transition fields ∆∗ L (T) obtained from the field-mixing analysis follow the expected first-order finitesize behavior ∆∗ L = ∆∗∞ + b L−2, (12).
First-Order Line Consistency
A direct cross-check of Eq. (12) is provided by the histogram-reweighting data: extrapolating the WL results for L = 12–24 at T = 0. p. 2 A direct cross-check of Eq. (12) is provided by the
histogram-reweighting data: extrapolating the WL results for L = 12–24 at T = 0. A direct cross-check of Eq. (12) is provided by the
histogram-reweighting data: extrapolating the WL results for L = 12–24 at T = 0. (12) is provided by the
histogram-reweighting data: extrapolating the WL results for L = 12–24 at T = 0. 400 predicts ∆∗
36 = 1. (12) is provided by the
histogram-reweighting data: extrapolating the WL results for L = 12–24 at T = 0. 400 predicts ∆∗
36 = 1. 400 predicts ∆∗
36 = 1. 48281,
whereas the L = 36 Metropolis run gives 1. 400 predicts ∆∗
36 = 1. 48281,
whereas the L = 36 Metropolis run gives 1. 48281,
whereas the L = 36 Metropolis run gives 1. 48280 —
agreement to one part in 105 between two entirely different simulation schemes. 48281,
whereas the L = 36 Metropolis run gives 1. 48280 —
agreement to one part in 105 between two entirely different simulation schemes.Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Tricritical Point Location
The crossing temperatures are listed in Table II, where the crossing temperatures are listed in Table II. The extrapolation of U L Q(T) toward the thermodynamic limit yields Tt = 0. p. 3 Panel (a) shows the cumulant curves U
L
Q(T)
evaluated along ∆∗
L
(T); their size-dependent intersections provide the finite-size crossing temperatures listed
in Table II. Panel (b) shows the extrapolation of these
crossings toward the thermodynamic limit, from which
Tt = 0. Panel (b) shows the extrapolation of these
crossings toward the thermodynamic limit, from which
Tt = 0. 4022(7) is obtained after accounting for the uncertainty in the correction exponent ω. It
is instead the spread obtained when the interpolation of
U
L
Q(T) is changed between linear, quadratic and cubicspline forms, and should be regarded as a lower bound
on the true uncertainty. Evaluating the L
−2
extrapolation of ∆∗
L
at T = Tt
gives the tricritical field
∆t = 1.Where the paper says it
Where the paper says it
Where the paper says it
Tricritical Coordinates
The two coordinates of the tricritical point are obtained from distinct finite-size extrapolations of the cumulant crossings and the pseudo-transition field, respectively.
--- Page 9 ---
Table III compares the tricritical coordinates and the ratio ∆t/∆0 with effective-field predictions and with results for other two-dimensional lattices. The ratio ∆t/∆0 varies only weakly among the lattice geometries listed. p. 3 Table III compares the tricritical coordinates and the
ratio ∆t/∆0 with effective-field predictions and with results for other two-dimensional lattices. The ratio ∆t/∆0
varies only weakly among the lattice geometries listed.Where the paper says it
Field Mixing Parameter Determination
The direction of the coexistence curve in the β–µ plane fixes the second mixing parameter through r = −(dµ/dβ) evaluated at the tricritical point. Using the extrapolated first-order line of Table I we obtain r = −1. p. 4 The parameter r
The direction of the coexistence curve in the β–µ
plane fixes the second mixing parameter through r =
−(dµ/dβ) evaluated at the tricritical point. Using the
extrapolated first-order line of Table I we obtain
r = −1. Using the
extrapolated first-order line of Table I we obtain
r = −1. 002 (stat) ± 0. 002 (stat) ± 0. 019 (syst), (16)
The uncertainty is dominated by the systematic component, estimated from the spread of r under reasonable variations of the analysis: varying the fitting window between |T − Tt| ≤ 0.Where the paper says it
Where the paper says it
Where the paper says it
Mean-Field Comparison
A useful dimensionless comparison is obtained by normalizing r by the zero-temperature field ∆0 = zJ/2. We find r/∆0 = −1. p. 4 A useful dimensionless comparison is obtained by normalizing r by the zero-temperature field ∆0 = zJ/2. We find r/∆0 = −1. We find r/∆0 = −1. 057(12) for the honeycomb lattice. The honeycomb value is from this work, from Wang–Landau data for
L = 12, 16, 18, 24, as is the first of the two square-lattice entries; the triangular value is from Ref. The mean-field
result of Eq. The mean-field
result of Eq. (17) is exact and independent of z.Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Tricritical Eigenvalue Exponents
Near the tricritical point the distribution of Q obeys PL(Q) = L(-yt) Po (L(-yt)) (Q − ⟨Q>), with the expected two-dimensional tricritical Ising value yt = 9/5 [25, 26]. The resulting exponent yt = 1. p. 6 (18)
is equivalent to σQ ∼ L
−(d−yt) and extract the exponent
from a log–log fit, which provides a more sensitive test
over the available size range. The resulting exponent depends significantly on the
temperature at which the distributions are evaluated:
we obtain yt = 1. Interpolating to the independently determined tricritical temperature gives
yt = 1. 795(7), (19)
consistent with 9/5 within 0.Where the paper says it
Where the paper says it
The approach to zero of the interface tension is described by (Tt −T)µ with µ = 1.265(26), consistent with the tricritical Ising value 1/yg = 5/4.
Coexistence Line Consistency
The extrapolation of the coexistence line is consistent with the tricritical point obtained from the cumulant analysis, and its slope there yields an independent determination of the mixing parameter, r = −1.587(5), four times more precise than the field-mixing estimate. p. 9 Extrapolation of the coexistence line is consistent with
the tricritical point obtained from the cumulant analysis,
and its slope there yields an independent determination
of the mixing parameter, r = −1. 587(5), four times more
precise than the field-mixing estimate.Where the paper says it
Final Findings
The study has mapped the phase boundary of the spin-1 Blume–Capel ferromagnet on the honeycomb lattice and located the tricritical point at (∆t, Tt) = [1.4819(2), 0.4022(7)]. The result is supported by two independent simulation schemes whose crossing estimates are mutually consistent. The authors also obtained the field-mixing parameter r = −1.586(19), showing that the combination r/∆0 is exactly −1 in mean-field theory for any coordination number.
--- Page 4 ---
TABLE I. Pseudo-transition field ∆∗∞ along the phase boundary, from the L−2 extrapolation of the Wang–Landau fieldmixing data (L = 12, 16, 18, 24). p. 3 Pseudo-transition field ∆∗∞ along the phase boundary, from the L
−2
extrapolation of the Wang–Landau fieldmixing data (L = 12, 16, 18, 24). The last column indicates
the character of the transition at each temperature, the tricritical point at Tt = 0. Pseudo-transition field ∆∗∞ along the phase boundary, from the L
−2
extrapolation of the Wang–Landau fieldmixing data (L = 12, 16, 18, 24). The last column indicates
the character of the transition at each temperature, the tricritical point at Tt = 0. The last column indicates
the character of the transition at each temperature, the tricritical point at Tt = 0. 4022(7) separating the two regimes.Where the paper says it
Where the paper says it
Where the paper says it
Tricritical Point Coordinates
The L−2 extrapolation of ∆∗ L at T = Tt gives the tricritical field ∆t = 1.4819(2), (15).
Consistency Checks
A direct cross-check of Eq. (12) is provided by the histogram-reweighting data: extrapolating the WL results for L = 12–24 at T = 0.400 predicts ∆∗ 36 = 1.48281, whereas the L = 36 Metropolis run gives 1.48280 — agreement to one part in 105 between two entirely different simulation schemes.
Scaling Exponent Verification
The resulting exponent yt = 9/5 is consistent with 9/5 within 0.7σ. Imposing yt = 9/5 returns T = 0.40175, again within 1.3σ of the cumulant-crossing estimate.
Interface Tension Behavior
The approach to zero of the interface tension is described by (Tt −T)µ with µ = 1.265(26), consistent with the tricritical Ising value 1/yg = 5/4.
Final Conclusion
The study has mapped the phase boundary of the spin-1 Blume–Capel ferromagnet on the honeycomb lattice and located the tricritical point at (∆t, Tt) = [1.4819(2), 0.4022(7)]. The result is supported by two independent simulation schemes whose crossing estimates are mutually consistent.
--- Page 8 ---
Figure 5 summarizes the phase-boundary results obtained from the different numerical analyses. Panel (a) presents the phase diagram in the (∆, T) plane: the continuous and first-order branches meet at the tricritical point and the first-order branch approaches the exact zero-temperature endpoint (∆0, 0) = (3/2, 0). p. 9 Figure 5 summarizes the phase-boundary results obtained from the different numerical analyses. Panel (a)
presents the phase diagram in the (∆, T) plane: the continuous and first-order branches meet at the tricritical
point and the first-order branch approaches the exact
zero-temperature endpoint (∆0, 0) = (3/2, 0). Panel (a)
presents the phase diagram in the (∆, T) plane: the continuous and first-order branches meet at the tricritical
point and the first-order branch approaches the exact
zero-temperature endpoint (∆0, 0) = (3/2, 0). Panel
(b) represents the same boundary in the β–µ plane,
where the local directions of the scaling fields g and
λ make the field-mixing construction explicit.Where the paper says it
Where the paper says it
Scaling Crossover
Away from the tricritical region, susceptibility maxima are consistent with Ising finite-size scaling form. As ∆ approaches ∆t, however, the extrapolations become increasingly unstable, reflecting the crossover from Ising to tricritical scaling. p. 9 Away from the tricritical region the susceptibility maxima are consistent with the Ising finite-size scaling form. As ∆ approaches ∆t, however, the extrapolations become
increasingly unstable, reflecting the crossover from Ising
to tricritical scaling: the appropriate correlation-length
exponent changes from ν = 1 to νt = 1/yt = 5/9, and
no single power law describes the shift of the maxima
throughout the crossover region.Where the paper says it
Methodological Instability
The value of UQ at the successive crossings drifts slowly away from the expected universal value 0.571 as L increases, from 0. p. 9 Second,
the value of UQ at the successive crossings drifts slowly
away from the expected universal value 0. 571 as L increases, from 0. 571 as L increases, from 0. 563 at 12×16 to 0. We also note two features of the present data that deserve further scrutiny. First, the mixing parameter sL
obtained from the histogram-reweighting lattices is considerably less stable than that from the Wang–Landau
lattices, reflecting the weak sensitivity of the equal-height
condition to s at large L; the reweighting data were therefore excluded from the extrapolation of Eq.Where the paper says it
Where the paper says it
Where the paper says it
Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.
Improvements for AI systems
-
Improved Phase Boundary Mapping for Spin Systems: The AI can accurately map first-order transition lines and locate tricritical points for spin-1 models on various lattices, such as honeycomb, square, triangular, and simple-cubic geometries. This allows the system to predict the precise location of metamagnetic transitions based on lattice structure.
-
Improved Determination of Critical Parameters: The AI can calculate scaling exponents like yt = 9/5 and yg = 4/5 for the tricritical Ising class, providing a quantitative check on universality classes across different lattice dimensions. It can also determine the mixing parameter r from first-order line extrapolations, yielding highly precise values like r = -1.587(5).
-
Improved Coexistence Line Prediction: The AI can predict the interface tension along the first-order line, describing its approach to zero as a function of the distance from the tricritical point, with an exponent mu = 1.265(26). This allows for a precise description of how interfacial energy scales near criticality.
-
Improved Tricritical Point Localization: The AI can determine the coordinates of the tricritical point (Delta t, Tt) by combining multiple independent simulation schemes, such as Wang–Landau estimates and histogram reweighting, ensuring high consistency between methods.
Abstract
We present a high-resolution numerical study of the spin-1 Blume-Capel ferromagnet on the honeycomb lattice (z=3). Using the Wang-Landau algorithm to estimate the joint density of states Ω(E 1,E 2) and field-mixing scaling, we map the first-order transition line and locate the tricritical point at (Δ t,T t)=[1.4820(2),0.4022(7)]. Independent Metropolis histogram reweighting for L=36 and 48 yields a crossing temperature near 0.4039, consistent with the Wang-Landau extrapolation. The field-mixing parameter is r=-1.585(19), while the coexistence-line slope gives r=-1.587(5). Normalizing by the zero-temperature field Δ 0=zJ/2, the results for honeycomb, square, and triangular lattices give r/Δ 0 between-1.057 and-1.067, about 6% larger in magnitude than the mean-field value. Simple-cubic estimates are compatible with this range but do not resolve dimensionality dependence. Finite-size scaling yields y t=1.795(7), consistent with the tricritical Ising value 9/5, while the fourth-order cumulant scaling agrees with y g=4/5. Deep in the first-order regime, the cumulant approaches the two-phase limit U Q to2/3. Multicanonical simulations up to L=64 yield an interface tension exponent μ=1.265(26), consistent with the tricritical prediction (d-1)/y g=5/4. We also map the continuous phase boundary, obtaining T c(Δ=0)=1.157, and identify crossover effects from Ising to tricritical scaling near the tricritical point.
Related papers
- Emergent long-tail dynamics in driven magnets with dynamical frustration
- Statistical Mechanics of Dynamical System Identification
- Second law of thermodynamics in closed quantum many-body systems
- Chemical potentials from structure factors: II. Charged multi-component mixtures
- Genuine quantum scars in Floquet chaotic many-body systems
- Stochastic Thermodynamics for Autoregressive Generative Models: A Non-Markovian Perspective