Thermal transport plays a pivotal role across diverse disciplines, yet the intricate relationship between amorphous network structures and thermal conductance properties remains elusive due to the absence of a reliable and comprehensive network’s dataset to be investigated. In this study, we have created a dataset comprising multiple amorphous network structures of varying sizes, generated through a combination of the node disturbance method and Delaunay triangulation, to fine-tune an initially random network toward both increased and decreased thermal conductance C. The tuning process is guided by the simulated annealing algorithm. Our findings unveil that C is inversely dependent on the normalized average shortest distance L n o r m connecting heat source nodes and sink nodes, which is determined by the network topological structure. Intuitively, the amorphous network with increased C is associated with an increased number of bonds oriented along the thermal transport direction, which shortens the heat transfer distance from the source to sink node. Conversely, thermal transport encounters impedance with an augmented number of bonds oriented perpendicular to the thermal transport direction, which is demonstrated by the increased L n o r m. This relationship can be described by a power law C = L n o r m α, applicable to the diverse-sized amorphous networks we have investigated.

The thermal transport property of a material is a crucial factor determining the performance and reliability of various devices’ applications, including thermal management1 and thermoelectric conversion.2 A high thermal conductivity is preferred in electronics, as it minimizes temperature rise during device operation, thereby enhancing device performance and lifespan.2 Conversely, a low thermal conductivity is valuable in thermoelectric materials, as it gives rise to substantial temperature gradients, a crucial factor for the effective functioning of such devices.3 Amorphous networks are prevalent in various systems, including biological networks,4,5 dynamical networks,6 polymers,7,8 and nanotube/nanowires,9,10 and they exhibit great potential of flexible manipulation of its thermal transport properties.11 

Understanding thermal transport in these networks is crucial for optimizing heat dissipation, designing efficient materials, and comprehending the behavior of amorphous systems.12–15 However, the relationship between amorphous network structures and their thermal transport property is still an open question.16 The thermal transport properties of such systems are influenced by a variety of network characteristics, such as node masses,17 average degree,18 assortativity coefficient,19,20 clustering coefficient,21 small-worldness,22 and closeness centrality.23 It is also due to the complicated topological patterns of amorphous networks that we lack a sufficient, representative dataset to investigate the thermal transport of network systems.24 

To unveil the fundamental physical mechanism dictating heat transport within amorphous networks, a substantial number of network structures serving as the dataset are imperative. By scrutinizing these amorphous structures, one can delve into the inherent thermal transport mechanisms at play. To establish a comprehensive dataset, it is essential to employ a versatile method for modifying the network structure. One common approach altering the topological configuration is achieved by selectively removing links.25–28 However, this method is constrained to reducing thermal conductance as cutting bonds impedes thermal transport pathways, leading to a decrease in the thermal conductance. An alternative tool is the rewire method,29,30 which introduces bond crossing areas within the network. Another promising method entails perturbing nodes within amorphous networks toward nearby positions, followed by the application of Delaunay triangulation to update the topological patterns. This particular method has exhibited potential in both augmenting and diminishing thermal conductance. Nevertheless, despite its promise, the tuning procedure lacks an efficient mechanism for guiding node distortion. Improving the guidance efficiency of node distortion remains a pivotal aspect in refining the tuning process for generating amorphous networks with significantly increased or decreased thermal conductance. Machine learning and optimization algorithms are promising to overcome the design impedance in materials research.31 

In this study, we employed the simulated annealing (SA) algorithm to systematically generate a comprehensive dataset of amorphous networks, encompassing configurations that both enhance and reduce thermal conductance. This versatile tuning methodology was applied across networks of diverse sizes, ranging from 16 × 16 to 32 × 32. The study delves into the generated amorphous networks with a variety of thermal conductance and tries to discover the fundamental physics governing thermal transport within these amorphous networks. Throughout the tuning process, the networks consistently maintained their amorphous state, as indicated by the order parameter ϕ 6, which consistently hovered around 0.4. Networks with enhanced thermal conductance exhibited bonds predominantly parallel to the heat flow direction, while networks with reduced thermal conductance showed a tendency for bonds to orient perpendicular to the heat flow direction. This observed orientation tendency is quantified through the average shortest path distance connecting heat source and sink nodes: a reduction in the average path distance from high-temperature nodes to low-temperature nodes enhances thermal transport, whereas an increase in the effective path length impedes thermal transport along the direction from high-temperature to low-temperature nodes.

We choose the random networks for investigation due to their completely stochastic structures, anticipating that the underlying thermal transport mechanisms within such a system would be more universally applicable to various amorphous network structures. Initially, a set of nodes is randomly distributed within a square region with a side length of N. Along each boundary of the square region, we randomly positioned N nodes, ensuring that the total number of nodes for each network equals to N t o t = N × N, as illustrated in Fig. 1(a). Subsequently, bonds connecting these nodes are established through Delaunay triangulation, resulting in the formation of an amorphous network with identical degree of connectivity, as depicted in Fig. 1(b).

FIG. 1.

(a)–(c) Generation and modification of a random network. For an amorphous network composed of N t o t = N × N nodes, a fixed number of N nodes are uniformly distributed on each side length of a square regime with length N, and the square region is marked as a red dashed line. Next, these nodes are connected through the Delaunay triangulation method. To modify the configuration of a random network, a selected node marked with a red circle is displaced to its nearby position, and a new pattern of network is generated by Delaunay triangulation again. Note that the local structure of the red square region within (b) and (c) is completely different. The different structures give rise to different thermal conductance values in (b) C = 2.02 and (c) C = 1.96, and (d) the simulated annealing algorithm can drive the solution jumping out of local optima and converges to global optimum.

FIG. 1.

(a)–(c) Generation and modification of a random network. For an amorphous network composed of N t o t = N × N nodes, a fixed number of N nodes are uniformly distributed on each side length of a square regime with length N, and the square region is marked as a red dashed line. Next, these nodes are connected through the Delaunay triangulation method. To modify the configuration of a random network, a selected node marked with a red circle is displaced to its nearby position, and a new pattern of network is generated by Delaunay triangulation again. Note that the local structure of the red square region within (b) and (c) is completely different. The different structures give rise to different thermal conductance values in (b) C = 2.02 and (c) C = 1.96, and (d) the simulated annealing algorithm can drive the solution jumping out of local optima and converges to global optimum.

Close modal

To manipulate the configuration of the random network, we applied a combination of the node distortion method and Delaunay triangulation, generating multiple random networks with varying thermal conductance values to form a dataset. A specific node, denoted by the red circle in Fig. 1(b), is initially displaced to a nearby position. Subsequently, Delaunay triangulation is reapplied, yielding a distinct random network pattern as shown in Fig. 1(c). Notably, the local structure highlighted by the red square in Figs. 1(b) and 1(c) is entirely different. This configuration change resulted in a modified thermal conductance value, transitioning from 2.02 to 1.96. We have observed that manipulating a specific node within the network has the potential to either enhance or diminish the overall conductance of the amorphous network. However, achieving the global optimal values of thermal conductance necessitates ensuring that the node distortion consistently aligns with the trend of continuously increasing or decreasing C. Therefore, a generalized approach is required to monitor whether the manipulation of node distortion can be accepted during the tuning process. This oversight is facilitated by the simulated annealing (SA) algorithm.32–34 

In this section, we provide the original version of SA algorithm and the modified formula we used in our case to tune the thermal conductance of amorphous networks. The SA algorithm emulates the gradual cooling process of metals. The algorithm employs the Metropolis criteria to either accept or reject a new state, and this iterative process persists until convergence is achieved. A system in thermal equilibrium at temperature T can be found with energy E with a probability as P ( E ) = exp ( E / ( k B T ) ), where k B is the Boltzmann constant. At low temperatures, there is a small chance that the system is in a high energy state. This plays a crucial role in SA since an increase in the energy allows escape from local minima and converge to global minimum, and the transition probability from a low energy state E l o w to high energy state E h i g h is
P = ( E l o w E h i g h ) = exp ( E h i g h E l o w ) / k B T = exp Δ E / k B T .
(1)
If we envision the solution space as a pitted plane, then the optimal node distortion strategy is analogous to a ball that falls into a pit. The goal of finding the globally optimal strategy is to guide the ball to the deepest pit on the plane. However, without a reliable algorithm, the ball may end up getting stuck in local areas, as depicted in Fig. 1(d). Applying the acceptance probability and the Metropolis principle is analogous to shake the solution space that forces the ball to jump out of the local deepest pit and continue exploring the solution space until it eventually falls into the global deepest pit. Once the ball reaches the global deepest pit, it remains there even under oscillation.
We have modified the original acceptance probability in SA algorithm based on our demand to tune C. Consider minimizing the thermal conductance C to its global minima. The algorithm comprises both an external and an internal loop. The external loop is governed by a tunable hyperparameter Γ, which is a unitless quantity used to tune C. Γ can be understood as the “virtual” temperature similar to the temperature governing the actual annealing process, and it decreases between two consecutive iteration steps t and t + 1 as Γ t + 1 = α Γ t with a decreasing rate α and α = 0.95. The tuning process ceases when the virtual temperature reaches final value Γ f. Within the internal loop, the Metropolis principle comes into play. We distorted the nodes for L k times at each virtual temperature Γ t. After each node distortion step, the network’ configuration changes from π t to π t + 1. Whether the node disturbance is accepted is governed by our modified acceptance probability P encoding the thermal conductance difference between C ( π t ) and C ( π t + 1 ), which can be summarized as
P = { 1 if  C ( π t + 1 ) < C ( π t ) , e x p ( C ( π t + 1 ) C ( π t ) / Γ t ) if  C ( π t + 1 ) > C ( π t ) .
(2)
Accordingly, if C decreases after node distortion, then this operation will undoubtedly be accepted ( P = 1). Conversely, if C increases, the algorithm has a chance to accept this worse solution by introducing the Metropolis principle: generate a random number ε within [ 0 , 1 ] and compare it with the exponential function P. If ε P, the strategy is accepted. Otherwise, go directly to the next iteration. Likewise, for increasing C, P for a bad strategy becomes exp ( ( C ( π t + 1 ) C ( π t ) ) / Γ t ) if C ( π t + 1 ) < C ( π t ).

We ignore the impact from contact resistance and only focus on the influence from the network topology to obtain the temperature distribution and heat flux among the amorphous networks. In this case, the topology of an amorphous network can be expressed as an adjacency matrix A, with its element A i j = 1 if nodes i and j are connected and 0 otherwise. Define the A i i = 0 for avoiding self-connection. Denote F as the temperature field vector of N t o t components. We consider that there are m modes with temperature F h as the heat source and n nodes with temperature F l as the heat sink.

The temperature distribution fields of source nodes are written as a vector F h = ( T h 1 , T h 2 , , T h m ), and the temperature fields of sink nodes are encoded into a vector F l = ( T l 1 , T l 2 , , T l n ). Thus, the entire temperature vector F for each node of an amorphous network can be generalized as F = ( F h , F l , T ( N t o t m n + 1 ) , , T N t o t ) and it can be divided into block-1 with F 1 = ( F h , F l ), which is a ( m + n)-dimensional vector and block-2 F 2 = ( T ( N t o t m n + 1 ) , , T N t o t ), which is a ( N t o t m n )-dimensional vector. The discrete Laplace operator on the network L ( N t o t × N t o t-dimensional) is an analog of 2 and can be expressed as L = D A, where D is the diagonal degree matrix. Accordingly, the matrix L can be divided into four submatrices: ( m + n ) × ( m + n )-dimensional L 11, ( m + n ) × ( N t o t m n )-dimensional L 12, ( N t o t m n ) × ( m + n )-dimensional L 21 and ( N t o t m n ) × ( N t o t m n )-dimensional L 22. In the steady state, we can obtain L F = J, where J is the external flux vector J = ( J h , J l , J ( N t o t m n + 1 ) , , J N t o t ). Note that this is the discrete analog of κ 2 T ( r ) = j ( r ) with F playing the role of κ T ( r ) and J playing the role of j ( r ), where j ( r ) is the heat flux of the vector field. Except for the source and sink nodes, the equilibrium conditions require zero net heat flux, i.e., J = 0 for the nodes i = N t o t m n + 1 , , N t o t. In other words, heat generation is zero for the node outside the source and sink node. Therefore, we have
{ L 11 F 1 + L 12 F 2 = J , L 21 F 1 + L 22 F 2 = 0 ,
(3)
where J = ( J h , J l ) represents the heat flux on each source node. J h = J 01 , J 02 , , J 0 m represents heat flux on node applied with high temperature, and J l = J 01 , J 02 , , J 0 n demonstrates the heat flux on each node with applied low temperatures. According to Kirchhoff’s law, the total amount of heat flow entering the network should be equal to the amount of heat flow outgoing the network, p = 1 m J 0 p = q = 1 n J 0 q = J 0.

The second equation can be expressed as L 22 F 2 = L 21 F 1. As F 1 and the elements of all submatrices L 11, L 12, L 21, and L 22 are known for a network, so L 21 F 1 is a ( N t o t m n)-dimensional vector and the equation L 22 F 2 = L 21 F 1 has a unique nonzero solution F 2. Then, substituting F 2 into the first equation, we can obtain J 0. Thus, J 0 can be analytically solved. Finally, we can calculate the thermal conductance C of the entire amorphous network as C = J 0 / ( T h T l ). Note that these quantities are dimensionless.

By employing the node distortion method in conjunction with Delaunay triangulation, we can systematically mold the initially random network to attain specific configurations characterized by both enhanced and decreased thermal conductance. The initial configuration, delineated in Fig. 2(a), represents the temperature field at steady state with the color on each node marked the temperature distribution. Nodes along the bottom boundary are applied with high temperatures, and those along the top boundary are set with low temperatures. Therefore, the setting of boundary conditions will cause the overall heat flow to flow from the bottom to the top. In Fig. 2(b), the network configuration illustrates an augmentation in thermal conductance, whereas Fig. 2(c) showcases the ultimate adjusted network with reduced thermal conductance. Intriguingly, the network with enhanced thermal conductance exhibits a prevalence of bonds oriented along the heat flow direction, while the random network with diminished thermal conductance showcases numerous bonds aligned perpendicularly to the flow direction. This observation emphasizes the important role of bonds oriented along the direction of heat flow in increasing the thermal conductivity of the network. Conversely, bonds oriented perpendicular to the direction of heat flow impede heat transport from high-temperature nodes to low-temperature nodes.

FIG. 2.

(a)–(c) Temperature distribution of the original amorphous network (a), and the tuned network structure with increased (b) and decreased (c) thermal conductance. (c) Hexagonal order parameter ϕ 6 as a function of thermal conductance of networks with varying sizes from 16 × 16 to 32 × 32. (e) Clustering coefficient E with respect to the thermal conductance for different sized amorphous networks.

FIG. 2.

(a)–(c) Temperature distribution of the original amorphous network (a), and the tuned network structure with increased (b) and decreased (c) thermal conductance. (c) Hexagonal order parameter ϕ 6 as a function of thermal conductance of networks with varying sizes from 16 × 16 to 32 × 32. (e) Clustering coefficient E with respect to the thermal conductance for different sized amorphous networks.

Close modal

The amorphous networks maintained their amorphous state throughout the tuning process, as quantified by the order parameter ϕ 6 illustrated in Fig. 2(d). The local hexagonal order parameter for a given node j is quantified by ψ 6 ( j ) = 1 N j n = 1 N j = exp ( i 6 θ n j ), where N j is the number of n nearest neighbors of j and θ n j is the angle between the horizontal axis and the bond linking j to n.35 We considered the average of the absolute values of local the hexagonal order parameter from each node inside the amorphous network as ϕ 6 = | ψ 6 ( j ) | . In a perfect hexagonal lattice, ϕ 6 = 1, while values of ϕ 6 = 0.5 or less indicate an amorphous structure.36  Figure 2(d) depicts the variations in the order parameter as thermal conductance increased and decreased, respectively, with different network sizes ranging from 16 × 16 to 32 × 32. Notably, throughout the tuning process, the order parameter consistently hovered around 0.4, a value below the critical threshold of 0.5, indicating that the amorphous networks remained in an amorphous state during tuning.

Another order parameter, clustering coefficient E, quantifies the heterogeneity of the amorphous network.37,38 It is defined as E = 1 N t o t 2 δ i k i ( k i 1 ), where δ i is the number of bonds connected surrounding the node i, and k i is the number of neighbor nodes around i, which is the degree of i. E represents the probability that two neighbors of a node are also connected to each other. For larger E, the links form groups and the network is more likely to generate local clusters. The heterogeneity also influences the heat flux through the network. As shown in Fig. 2(c), the clustering coefficient for networks tuned to enhanced and decreased thermal conductance values is kept at around 0.1, independent with the network sizes from 16 × 16 to 32 × 32. This indicates that the inherent thermal transport tailored by our method is not dependent on the network clustering coefficient.

The average shortest path length is a critical property in the general analysis of amorphous networks. It refers to the total length of bonds connecting any two nodes within a network, representing the shortest route from a source node to a target node. The average shortest distance, denoted as L, is influenced by both network size and structural characteristics of the network, impacting heat conduction dynamics.39–41 

In thermal transport, a shorter distance between the heat source and sink nodes promotes more efficient heat transfer, while a longer transport distance indicates inefficiency in guiding heat in the desired direction. We have observed that thermal conductance is inversely proportional to the average shortest distance between connecting edges of hot and cold nodes. For analysis, we define pairs of heat source and sink nodes located on the bottom and upper boundaries, sharing the same x coordinate, as p (where p = 1 , 2 , , N for N nodes on each boundary). The Floyd algorithm42,43 is employed to find all edges generating the shortest path distance l p between each source and sink node pair. As illustrated in Fig. 3(a), the left panel represents a configuration of the maximized thermal conductance, while the right panel represents a configuration of the minimized thermal conductance. Both configurations are derived from the same original amorphous network using our node displacement and Delaunay triangulation method.

FIG. 3.

(a) and (b) A pair of nodes with the same x coordinate on top and bottom boundaries are selected, and the edges passing from the heat source node (the bottom one) to the sink node (the top one) with the shortest total path distance are marked with red. (c) The logarithmic plot of the thermal conductance C as a function of normalized average shortest path distance L n o r m for amorphous networks of different sizes N × N. Notably, these curves merge closely together, demonstrating that the thermal conductance is governed by a general underlying mechanism that is correlated to the L n o r m, derived from the topological structure of amorphous networks. (d) Thermal conductance C as a function of normalized angle summation θ n o r m of bonds along the shortest path.

FIG. 3.

(a) and (b) A pair of nodes with the same x coordinate on top and bottom boundaries are selected, and the edges passing from the heat source node (the bottom one) to the sink node (the top one) with the shortest total path distance are marked with red. (c) The logarithmic plot of the thermal conductance C as a function of normalized average shortest path distance L n o r m for amorphous networks of different sizes N × N. Notably, these curves merge closely together, demonstrating that the thermal conductance is governed by a general underlying mechanism that is correlated to the L n o r m, derived from the topological structure of amorphous networks. (d) Thermal conductance C as a function of normalized angle summation θ n o r m of bonds along the shortest path.

Close modal

The disparity in the shortest distance l p between the left and right panels is apparent. On the left panel, the red curve forms an almost straight line connecting the selected pair of source and sink nodes, resulting in a smaller l p. Conversely, the red curve on the right panel meanders through more convoluted paths, yielding a larger l p. As we know, the right panel exhibits a smaller thermal conductance compared to the left.

To quantify this observation, we computed the average shortest path distance for each pair of source and sink nodes along both the bottom and upper boundaries. We define the average shortest path distance, denoted as L ¯, as L ¯ = 1 N p = 1 N l p. Here, p represents each pair of heat source and sink nodes on the opposite boundaries, and l p is the shortest distance between the respective pairs. Given that we are dealing with networks of different sizes characterized by varying boundary lengths, denoted as N × N, the average shortest path distance L ¯ is expected to increase with the network boundary length N. To account for the impact of network size, the average shortest path distance L ¯ for each sized network is further normalized by dividing it by the network size N, resulting in L n o r m = L ¯ / N. In our exploration, we successfully tuned amorphous networks with dimensions ranging from 16 × 16 to 32 × 32 to exhibit both increased and decreased thermal conductance. As shown in Fig. 3(c), the scatter log–log plot reported the logarithmic values of C and L n o r m for different sized amorphous networks. It is noted that these scatters merge together to form a line, indicating the C and L n o r m following the power law. Furthermore, these networks can be accurately fitted with a red linear regression line, expressing the power law relationship between C and L n o r m as approximately C = L n o r m α, with α = 21.46. The negative power α provides additional evidence that an increase in the path distance through which heat energy travels along the network leads to a reduction in the overall thermal conductance. Therefore, we have proved that the entire thermal conductance of the amorphous network is inversely related to the average shortest path distance between high-temperature and sink nodes. This observation underscores the significance of average path distance as a crucial metric for comprehending variations in heat transfer properties within our amorphous network systems.

We have further uncovered that thermal conductance C adheres to a power law concerning the normalized average shortest distance L n o r m, as depicted in the log–log plot in Fig. 3(c). Notably, diverse-sized amorphous networks exhibit a convergence, indicating a common underlying physics. It is noteworthy that the thermal conductance of the amorphous network is highly sensitive to changes in the average shortest distance from high-temperature to sink nodes. Even a slight variation in L n o r m, such as a decrease from approximately 1.07 to 1.02, constituting only 0.05 in changes and roughly 5 % change ratio, can result in tuning C from around 1 to 3.5. Therefore, it can be anticipated that finding a method to further decrease L n o r m significantly would lead to a more substantial increase in C.

The variation in the shortest path distance can be interpreted as the curvature degree of the path. The presence of more zigzags in the shortest path indicates a more winding trajectory, resulting in a longer path length. This concept is visually represented in Fig. 3(d). For each shortest path, the curvature degree can be quantified by calculating the summation of the angle θ (representing the angle between the bond orientation and the y-axis). To articulate this, we express the summation angle for each bond as θ b, where b ranges from 1 to the total number of bonds along the shortest path. Subsequently, we derive the average value of each angle summation p = 1 N ( θ b ) and normalize it based on the largest angle summation θ b within each network, denoted as θ n o r m. A higher θ n o r m indicates a tendency for bonds to align perpendicularly to the heat transport direction, resulting in a decrease in the thermal conductance C. Conversely, a smaller θ n o r m implies a more direct alignment of bonds with the heat transport direction, leading to an increase in the thermal conductance C for the amorphous networks.

The planar network graphs can be further transformed into tangible physical entities, specifically metallic networks, as depicted in Fig. 4. Copper was chosen as the material for the entire network, with each bond having a uniform thickness of 0.1 cm. This conversion process resulted in two-dimensional metallic networks, showcased in the first row of Fig. 4. Subsequently, the planar graphs were extended by 1 cm along the z direction to create three-dimensional metallic networks, depicted in the second row of Fig. 4. The effective thermal conductivity ( κ e f f) of each network is recorded atop each panel. Due to the presence of multiple voids within the material, the effective thermal conductance values are inherently smaller than the thermal conductivity of a pure copper plate. Notably, the ratio of κ e f f between the optimized networks (both increased and decreased) and the original network, for both two-dimensional and three-dimensional calculations, is approximately 1.30. Similarly, the ratio of κ e f f between the original and decreased networks, for both two-dimensional and three-dimensional calculations, also hovers around 1.30. Moreover, the theoretically calculated ratios of thermal conductance further corroborate these findings, with C i n c r e a s e / C o r i g i n a l = 1.67 and C o r i g i n a l / C d e c r e a s e = 1.63. These results validate the optimization of network structures, demonstrating that the metallic networks exhibit enhanced or reduced κ e f f, corresponding to increased or decreased thermal conductance values. This validation underscores the efficacy of our optimization method in designing metallic network systems.

FIG. 4.

(a)–(c) Temperature profiles of two-dimensional and three-dimensional metallic network.

FIG. 4.

(a)–(c) Temperature profiles of two-dimensional and three-dimensional metallic network.

Close modal

By employing the SA algorithm, we have successfully customized the thermal conductance of amorphous networks, achieving both increased and decreased values. Throughout the tuning process, the algorithm generated a diverse set of network structures, forming the dataset instrumental in exploring the physical mechanisms governing the heat transfer properties of amorphous networks. Specifically, we have unveiled the crucial role of network bonds as pathways guiding heat flow within the network. This underscores the role of machine learning in the study of heat transport in complex networks. More importantly, this process can be reversed so that the revealed physics mechanism can reciprocally enhance the performance of machine learning model. Recent studies have exemplified this philosophy by employing the Laplacian matrix or Dynamical matrix as input to train the convolutional neural network, which has greatly improved the training and prediction efficiencies.44,45

With the rapid advancements in nanotechnology, the fabrication of nanowires/nanotubes in the laboratory has become more accessible.46–50 Metallic seamless nanowire networks, in particular, have garnered increasing interest due to their fast lithographic techniques for fabrication and unique advantages, including being interwire-contact-junction free.51,52 While their thermal performance has been extensively evaluated based on factors such as nanowires’ aspect ratio,53,54 network area fraction,55 and network density,56,57 few studies have delved into the impact of geometrical topology on thermal transport properties due to the lack of efficient design protocols. Our results offer insights into the preparation principles of seamless nanowires concerning the topology of the entire network structure. As a result, our discovery linking amorphous network topology to thermal transport behavior may contribute valuable guidance to the fabrication of nanowires/nanotubes with desired thermal properties.

In summary, the presented research unveils a comprehensive methodology for tuning the thermal conductance of amorphous networks, showcasing the potential for broader applications beyond thermal properties. The devised tuning approach not only successfully enhances but also diminishes the thermal conductance of networks, providing a versatile tool for tailoring various characteristics within amorphous systems. Crucially, the tuned networks exhibit distinctive bond orientation tendencies aligned with the predefined thermal transport direction. These tendencies indicate that the thermal conductance of the amorphous network is determined by the average shortest path length between the hottest and the coldest nodes. The observed bond orientation tendencies also provide valuable insights for the experimental creation of nanotubes and nanowires, offering a pathway for the deliberate construction of nanoscale networks with predetermined thermal transfer properties.

This work was supported in part by the National Natural Science Foundation of China (NNSFC) (Grant Nos. 12205138 and 52250191) and Shenzhen Science and Technology Innovation Committee (SZSTI) (Grant/Award No. JCYJ20220530113206015).

The authors have no conflicts to disclose.

Changliang Zhu: Data curation (lead); Formal analysis (equal); Methodology (equal); Writing – original draft (lead); Writing – review & editing (equal). Tianlin Luo: Formal analysis (equal); Methodology (equal); Writing – review & editing (equal). Baowen Li: Funding acquisition (equal); Supervision (supporting); Writing – review & editing (equal). Xiangying Shen: Conceptualization (lead); Funding acquisition (equal); Supervision (equal); Writing – review & editing (equal). Guimei Zhu: Funding acquisition (equal); Supervision (equal); Writing – review & editing (equal).

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

1.
Z.
Ling
,
Z.
Zhang
,
G.
Shi
,
X.
Fang
,
L.
Wang
,
X.
Gao
,
Y.
Fang
,
T.
Xu
,
S.
Wang
, and
X.
Liu
, “
Review on thermal management systems using phase change materials for electronic components, li-ion batteries and photovoltaic modules
,”
Renew. Sustain. Energy Rev.
31
,
427
438
(
2014
).
2.
X.
Xu
,
J.
Chen
,
J.
Zhou
, and
B.
Li
, “
Thermal conductivity of polymers and their nanocomposites
,”
Adv. Mater.
30
,
1705544
(
2018
).
3.
L.
Shi
,
J.
Jiang
,
G.
Zhang
, and
B.
Li
, “
High thermoelectric figure of merit in silicon-germanium superlattice structured nanowires
,”
Appl. Phys. Lett.
101
,
233114
(
2012
).
4.
H.
Yan
,
S. H.
Park
,
G.
Finkelstein
,
J. H.
Reif
, and
T. H.
LaBean
, “
DNA-templated self-assembly of protein arrays and highly conductive nanowires
,”
Science
301
,
1882
1884
(
2003
).
5.
J.
Ma
,
X.
Zhang
,
C. Y.
Ung
,
Y. Z.
Chen
, and
B.
Li
, “
Metabolic network analysis revealed distinct routes of deletion effects between essential and non-essential genes
,”
Mol. Biosys.
8
,
1179
1186
(
2012
).
6.
W.-X.
Wang
,
J.
Ren
,
Y.-C.
Lai
, and
B.
Li
, “
Reverse engineering of complex dynamical networks in the presence of time-delayed interactions based on noisy time series
,”
Chaos
22
,
033131
(
2012
).
7.
B.
Liu
,
T.
Lu
,
B.
Wang
,
J.
Liu
,
T.
Nakayama
,
J.
Zhou
, and
B.
Li
, “
Thermoelectric transport in hybrid materials incorporating metallic nanowires in polymer matrix
,”
Appl. Phys. Lett.
110
,
113102
(
2017
).
8.
L.
Dong
,
Q.
Xi
,
D.
Chen
,
J.
Guo
,
T.
Nakayama
,
Y.
Li
,
Z.
Liang
,
J.
Zhou
,
X.
Xu
, and
B.
Li
, “
Dimensional crossover of heat conduction in amorphous polyimide nanofibers
,”
Natl. Sci. Rev.
5
,
500
506
(
2018
).
9.
J.-W.
Jiang
,
J.-S.
Wang
, and
B.
Li
, “
A nonequilibrium Green’s function study of thermoelectric properties in single-walled carbon nanotubes
,”
J. Appl. Phys.
109
,
014326
(
2011
).
10.
L.
Shi
,
J.
Chen
,
G.
Zhang
, and
B.
Li
, “
Thermoelectric figure of merit in Ga-doped [0001] ZnO nanowires
,”
Phys. Lett. A
376
,
978
981
(
2012
).
11.
G.
Yan
,
J.
Ren
,
Y.-C.
Lai
,
C.-H.
Lai
, and
B.
Li
, “
Controlling complex networks: How much energy is needed?
,”
Phys. Rev. Lett.
108
,
218703
(
2012
).
12.
X.
Shen
,
G.
Zhu
, and
B.
Li
, “
Non-affinity: The emergence of networks from amorphous planar graphs
,”
Sci. China, Ser. G: Phys., Mech. Astron.
66
,
260501
(
2023
).
13.
S.
Wang
,
C.
Zeng
,
G.
Zhu
,
H.
Wang
, and
B.
Li
, “
Controlling heat ratchet and flow reversal with simple networks
,”
Phys. Rev. Res.
5
,
043009
(
2023
).
14.
Q.
Xi
,
J.
Zhong
,
J.
He
,
X.
Xu
,
T.
Nakayama
,
Y.
Wang
,
J.
Liu
,
J.
Zhou
, and
B.
Li
, “
A ubiquitous thermal conductivity formula for liquids, polymer glass, and amorphous solids
,”
Chin. Phys. Lett.
37
,
104401
(
2020
).
15.
S.
Kumar
,
J.
Murthy
, and
M.
Alam
, “
Percolating conduction in finite nanotube networks
,”
Phys. Rev. Lett.
95
,
066802
(
2005
).
16.
K.
Xiong
and
Z.
Liu
, “
Temperature dependence of heat conduction coefficient in nanotube/nanowire networks
,”
Chin. Phys. B
26
,
098904
(
2017
).
17.
K.
Xiong
,
Z.
Yan
,
Y.
Xie
, and
Z.
Liu
, “
Regulating heat conduction of complex networks by distributed nodes masses
,”
Sci. Rep.
11
,
5501
(
2021
).
18.
K.
Xiong
,
C.
Zeng
,
Z.
Liu
, and
B.
Li
, “
Influence of the degree of a complex network on heat conduction
,”
Phys. Rev. E
98
,
022115
(
2018
).
19.
K.
Xiong
,
C.
Zeng
, and
Z.
Liu
, “
Effect of degree correlation on the thermal transport in complex networks
,”
Nonlinear Dyn.
94
,
3067
3075
(
2018
).
20.
M. E.
Newman
, “
Assortative mixing in networks
,”
Phys. Rev. Lett.
89
,
208701
(
2002
).
21.
X.
Wu
and
Z.
Liu
, “
How community structure influences epidemic spread in social networks
,”
Physica A
387
,
623
630
(
2008
).
22.
D. J.
Watts
and
S. H.
Strogatz
, “
Collective dynamics of ‘small-world’ networks
,”
Nature
393
,
440
442
(
1998
).
23.
K.
Xiong
,
J.
Ren
,
F.
Marchesoni
, and
J.
Huang
, “
Phononic band gap in random spring networks
,”
Phys. Rev. E
108
,
044306
(
2023
).
24.
Z.
Liu
and
B.
Li
, “
Heat conduction in simple networks: The effect of interchain coupling
,”
Phys. Rev. E
76
,
051118
(
2007
).
25.
X.
Shen
,
C.
Fang
,
Z.
Jin
,
H.
Tong
,
S.
Tang
,
H.
Shen
,
N.
Xu
,
J. H. Y.
Lo
,
X.
Xu
, and
L.
Xu
, “
Achieving adjustable elasticity with non-affine to affine transition
,”
Nat. Mater.
20
,
1635
1642
(
2021
).
26.
K.
Xiong
,
Z.
Yan
,
Y.
Xie
,
Y.
Wang
,
C.
Zeng
, and
Z.
Liu
, “
Heat flux localization and abnormal size effect induced by multi-body vibration in complex networks
,”
Nonlinear Dyn.
110
,
2771
2779
(
2022
).
27.
Z.
Jin
,
C.
Fang
,
X.
Shen
, and
L.
Xu
, “
Designing amorphous networks with adjustable Poisson ratio from a simple triangular lattice
,”
Phys. Rev. Appl.
18
,
054052
(
2022
).
28.
I.
Kryven
, “
Bond percolation in coloured and multiplex networks
,”
Nat. Commun.
10
,
404
(
2019
).
29.
B. J.
Kim
, “
Performance of networks of artificial neurons: The role of clustering
,”
Phys. Rev. E
69
,
045101
(
2004
).
30.
K.
Xiong
,
J.
Zhou
,
M.
Tang
,
C.
Zeng
, and
Z.
Liu
, “
Control of thermal conduction and rectification in a model of complex networks with two asymmetric parts
,”
Phys. Rev. E
98
,
062144
(
2018
).
31.
C.
Zhu
,
E. A.
Bamidele
,
X.
Shen
,
G.
Zhu
, and
B.
Li
, “
Machine learning aided design and optimization of thermal metamaterials
,”
Chem. Rev.
124
,
4258
–4331 (
2024
).
32.
D.
Bertsimas
and
J.
Tsitsiklis
, “
Simulated annealing
,”
Stat. Sci.
8
,
10
15
(
1993
).
33.
L.
Ingber
, “
Simulated annealing: Practice versus theory
,”
Math. Computer Modell.
18
,
29
57
(
1993
).
34.
C.
Tsallis
and
D. A.
Stariolo
, “
Generalized simulated annealing
,”
Physica A
233
,
395
406
(
1996
).
35.
H.
Zhang
,
F.
Liu
,
G.
Ungar
,
Z.
Zheng
,
Q.
Sun
, and
Y.
Han
, “
A regime beyond the Hall–Petch and inverse-Hall–Petch regimes in ultrafine-grained solids
,”
Commun. Phys.
5
,
329
(
2022
).
36.
R.
Zuñiga
,
G.
Varas
, and
S.
Job
, “
Geometry-controlled phase transition in vibrated granular media
,”
Sci. Rep.
12
,
14989
(
2022
).
37.
S.
Boccaletti
,
V.
Latora
,
Y.
Moreno
,
M.
Chavez
, and
D.-U.
Hwang
, “
Complex networks: Structure and dynamics
,”
Phys. Rep.
424
,
175
308
(
2006
).
38.
Z.
Liu
,
X.
Wu
,
H.
Yang
,
N.
Gupte
, and
B.
Li
, “
Heat flux distribution and rectification of complex networks
,”
New J. Phys.
12
,
023016
(
2010
).
39.
S. N.
Dorogovtsev
,
A. V.
Goltsev
, and
J. F.
Mendes
, “
Critical phenomena in complex networks
,”
Rev. Mod. Phys.
80
,
1275
(
2008
).
40.
K.
Xiong
,
Z.
Liu
,
C.
Zeng
, and
B.
Li
, “
Thermal-siphon phenomenon and thermal/electric conduction in complex networks
,”
Natl. Sci. Rev.
7
,
270
277
(
2020
).
41.
M.
Verdier
,
D.
Lacroix
, and
K.
Termentzidis
, “
Thermal transport in two-and three-dimensional nanowire networks
,”
Phys. Rev. B
98
,
155434
(
2018
).
42.
S.
Hougardy
, “
The Floyd–Warshall algorithm on graphs with negative cycles
,”
Inf. Process. Lett.
110
,
279
281
(
2010
).
43.
A.
Aini
and
A.
Salehipour
, “
Speeding up the Floyd–Warshall algorithm for the cycled shortest path problem
,”
App. Math. Lett.
25
,
1
5
(
2012
).
44.
C.
Zhu
,
X.
Shen
,
G.
Zhu
, and
B.
Li
, “
Prediction of thermal conductance of complex networks with deep learning
,”
Chin. Phys. Lett.
40
,
124402
(
2023
).
45.
C.
Zhu
,
C.
Fang
,
Z.
Jin
,
B.
Li
,
X.
Shen
, and
L.
Xu
, “A cyclical route linking fundamental mechanism and AI algorithm: An example from Poisson’s ratio in amorphous networks,” arXiv:2312.03404 (2023).
46.
C. W.
Chang
,
D.
Okawa
,
A.
Majumdar
, and
A.
Zettl
, “
Solid-state thermal rectifier
,”
Science
314
,
1121
1124
(
2006
).
47.
Q.
Cao
and
J. A.
Rogers
, “
Ultrathin films of single-walled carbon nanotubes for electronics and sensors: A review of fundamental and applied aspects
,”
Adv. Mater.
21
,
29
53
(
2009
).
48.
D. R.
Kauffman
and
A.
Star
, “
Carbon nanotube gas and vapor sensors
,”
Angew. Chem. Int. Ed.
47
,
6550
6570
(
2008
).
49.
R. S.
Prasher
,
X.
Hu
,
Y.
Chalopin
,
N.
Mingo
,
K.
Lofgreen
,
S.
Volz
,
F.
Cleri
, and
P.
Keblinski
, “
Turning carbon nanotubes from exceptional heat conductors into insulators
,”
Phys. Rev. Lett.
102
,
105901
(
2009
).
50.
M. F. P.
Wagner
,
K.-O.
Voss
,
C.
Trautmann
, and
M.
Toimil-Molares
, “
Three-dimensional nanowire networks fabricated by ion track nanotechnology and their applications
,”
EPJ Tech. Instrum.
10
,
2
(
2023
).
51.
K.
Esteki
,
D.
Curic
,
H.
Manning
,
E.
Sheerin
,
M.
Ferreira
,
J.
Boland
, and
C.
Rocha
, “
Thermo-electro-optical properties of seamless metallic nanowire networks for transparent conductor applications
,”
Nanoscale
15
,
10394
10411
(
2023
).
52.
D.
Ma
,
H.
Ding
,
H.
Meng
,
L.
Feng
,
Y.
Wu
,
J.
Shiomi
, and
N.
Yang
, “
Nano-cross-junction effect on phonon transport in silicon nanowire cages
,”
Phys. Rev. B
94
,
165434
(
2016
).
53.
H. G.
Manning
,
P. F.
Flowers
,
M. A.
Cruz
,
C. G. D.
Rocha
,
C. O.
Callaghan
,
M. S.
Ferreira
,
B. J.
Wiley
, and
J. J.
Boland
, “
The resistance of cu nanowire–nanowire junctions and electro-optical modeling of Cu nanowire networks
,”
Appl. Phys. Lett.
116
,
251902
(
2020
).
54.
Z.
Wei
,
G.
Wehmeyer
,
C.
Dames
, and
Y.
Chen
, “
Geometric tuning of thermal conductivity in three-dimensional anisotropic phononic crystals
,”
Nanoscale
8
,
16612
16620
(
2016
).
55.
K.
Esteki
,
H. G.
Manning
,
E.
Sheerin
,
M. S.
Ferreira
,
J. J.
Boland
, and
C. G.
da Rocha
, “
Tuning the electro-optical properties of nanowire networks
,”
Nanoscale
13
,
15369
15379
(
2021
).
56.
A.
Kumar
and
G.
Kulkarni
, “
Evaluating conducting network based transparent electrodes from geometrical considerations
,”
J. Appl. Phys.
119
,
015102
(
2016
).
57.
E.
Mulazimoglu
,
S.
Coskun
,
M.
Gunoven
,
B.
Butun
,
E.
Ozbay
,
R.
Turan
, and
H. E.
Unalan
, “
Silicon nanowire network metal-semiconductor-metal photodetectors
,”
Appl. Phys. Lett.
103
,
083114
(
2013
).