betaFeedback

WorksPhysics and Astronomy

Perimeter shell properties of branched polymer adsorbents in various conformational states

Abstract

The structural accessibility and perimeter shell properties of branched macromolecules are studied using Monte Carlo simulations on a simple cubic lattice. Constituent polymer chains of these macromolecules are constructed in an algorithmic way as: a set of collinear monomers; a self-avoiding walk (SAW); tigthly packed monomers. These mimic the ground states for a rigid rod, flexible chain, and collapsed chain conformational states, respectively. We analyzed the set of observables that are related to adsorption, such as perimeter shell size (number of available sites for particle adsorption) per monomer ⟨nˉa⟩\langle \bar{n}_a \rangle, mean coordination number of perimeter shell ⟨nbond⟩\langle n_{\rm bond} \rangle (average number of contacts per adsorbed particle), and coordination accessibility per monomer ⟨nˉs⟩\langle \bar{n}_s \rangle. Our results demonstrate that branching reduces the perimeter shell size while it increases mean coordination number of perimeter shell across all conformational states, due to higher local monomer density near branching cores and formation of shared adsorption sites or microcages. Stiff extended structures yield the highest values for ⟨nˉa⟩\langle \bar{n}_a \rangle (approaching 44 in the limit of long chains), whereas the flexible ones provide the largest values for ⟨nbond⟩\langle n_{\rm bond} \rangle. We can note that the collinearity of monomers in constituent chains works in the same direction as reduction of branching (⟨nˉa⟩\langle \bar{n}_a \rangle grows), whereas the { tight packing of them} – in the same direction as the increase of branching (⟨nˉa⟩\langle \bar{n}_a \rangle decreases)

Introduction

Selective adsorption of organic and inorganic pollutants from air, water and soil, aimed at their cleaning and/or purification, is of great practical importance [1, 2]. Despite the differences between the chemical and physical mechanisms in each particular case, the adsorption effectiveness dictates rather similar requirements to the adsorbent: mechanical stability, maximum surface area, specific arrangement and accessibility of adsorbing chemical groups, etc. These requirements can be addressed more readily for the polymer-based adsorbents [3, 4], where the polymer macromolecules serve as stationary adsorbent medium (matrix) for external targets. The examples include such super-adsorbents as macroporous resins, biopolymer-based adsorbents, hydrogels, all representing the same class of highly crosslinked polymer networks [5, 6, 7]. To achieve the desired structure of a polymer adsorbent, and, consequently, its high effectiveness, one can vary the molecular mass and flexibility of its constituent chains, the amount and type of molecular branching, and other details. Clarification of the interdependence between all these factors could lead to prediction of the optimal molecular structure of a polymeric adsorbent. Comparison between linear and branched polymer chelating absorbers for removal of heavy metal ions and hazardous dyes have been done in a range of experiments [8, 9, 10, 11, 12]. Branching was found to enhance the effectiveness of chelation because of two factors: natural formation of the chelation-adsorption-friendly cages around the branching points, and spatial separation of individual linear spacer-chains which increases their accessibility for pollutants [13].

Typical polymer networks and gels are crosslinked randomly and, therefore, are characterized by a polydispersity in both branching points multiplicity and polymer spacer lengths. The effects of the branching points of a particular kind, and of the spacers of a given length, are all mixed up and cannot be separated easily. On the other hand, the macromolecules of a regular branching pattern such as: star-like, pom-pom, snowflake, dendrimer, etc. [14, 15, 16, 17], allow such a possibility. Therefore, such molecular architectures are logical candidates for studying the effects of various types of branching on the effectiveness of adsorption. Star polymer represents the simplest example of a branching molecular architecture with, typically, a single branching point [18, 19]. It contains ff linear branches (the star functionality) that radiate outwards from a central core (Figure 1a), thus reconstituting a single linear chain at f=1f=1. Such polymer exhibits a more closely packed object than its linear counterpart of the same molecular weight [20, 21, 22, 23, 24]. A snowflake polymer topology [25, 26, 27], that can be interpreted as a first-generation dendrimer, is of particular interest. It consists of a star polymer of the functionality ff with additional gg twigs radiating from the terminal points of each arm (Figure 1b).

Figure 1. Monte Carlo simulation snapshots demonstrating the spatial conformations of representative polymer structures on a simple cubic lattice. (a), (c), and (e): a star polymer with f=6f=6 branches; (b), (d), and (f): a snowflake polymer with f=4f=4 branches and g=3g=3 twigs. (a) and (b): polymers comprising rigid rod branches; (c) and (d): flexible polymers; (e) and (f): tightly packed polymers.

Besides the type of branching of a particular polymer, there is another factor that affects its structure: the conformation state of its constituent chains, reflected in the scaling law for their mean-square radius of gyration ⟨Rg2⟩\langle R_g^2\rangle. In general,

⟨Rg2⟩∼N2ν,(1)\langle R_g^2\rangle\sim N^{2\nu}, \tag*{(1)}

where NN is the number of monomers in a chain, and ν\nu is the critical index for a correlation length. RgR_g can be measured directly via static light scattering [28, 29], and distinguishes between different conformational states of constituent polymer chains. Additional characteristic is the asymmetry measure of typical polymer configurations, such as the average asphericity ⟨A⟩\langle A\rangle [30, 31]. This quantity is bound between a maximum value of one for a completely rigid, rod-like conformation, and a minimum value of zero for an ideally spherical shape, thereby obeying the inequality 0≤⟨A⟩≤10 \le\langle A\rangle\le1. Hereafter, ⟨…⟩\langle\ldots\rangle denotes an ensemble average over all possible polymer conformations.

The SAW mimics a polymer chain in a good solvent, where it adopts the shape of a coil [32, 33], and, assuming that the adjacent monomers are separated by at least the Kuhn segment length, the scaling index is ν≈0.588\nu\approx0.588 in D=3D=3 (DD being the space dimensionality) [34]. We note that this law holds also for the macromolecule, composed of flexible chains, as a whole. The asphericity of a macromolecule AA, however, is influenced strongly by the molecular topology. For instance, for the case of the ff-functional star, the limit case of a linear chain exhibits significant spatial anisotropy with ⟨A(f=1)⟩=0.431(2)\langle A(f=1)\rangle= 0.431(2) [35], whereas the star with f>1f>1 is characterized by a gradual decrease of its asphericity with the increase of its functionality, ff: A(f=3)=0.306(1)A(f=3)=0.306(1), A(f=4)=0.227(1)A(f=4)=0.227(1), A(f=5)=0.177(1)A(f=5)=0.177(1), A(f=6)=0.140A(f=6)=0.140 [36].

Tightly packed polymer chains can be associated with the case when they are surrounded by a poor solvent, in which case they collapse into a dense globular structure. According to the polymer physics and the scaling theory, it can have two different forms [37, 38]: a disordered globule with high uniform concentration, and a highly symmetric crystalline or semi-crystalline form (e.g. lamellae). This dense compact state is driven by a short-range, solvent-mediated monomer-monomer attraction, whereas conformation entropy works against the polymer collapse. From structural point of view, one may say that such polymers tend to form the mostly compact, tightly packed structure with maximum number of monomer-monomer contacts, which implies possessing a high level of bending. The effective size measure (1) of a polymer in this state scales with ν=1/D\nu=1/D. The shape of linear and star-branched polymers in compact state have been questioned by means of the Monte Carlo simulations in Refs. [39, 40], indicating a substantial decrease in the asphericity parameter ⟨A⟩\langle A\rangle, driven by the higher symmetry of compact globules as compared to crumpled coils in good solvent regimes.

The collinear polymer chain represent a semiflexible polymer in a limit case of infinite rigidity. The latter was introduced by Flory [41] in a study of polymer crystallization. In this case, bending of two adjacent bonds causes the energy penalty, quantified via a stiffness parameter. Depending on solvent conditions, semi-flexible polymers can adopt a variety of ordered morphologies, including rod-like or racquet-like conformations, as the result of the competition between the monomers attraction and chain stiffness [42, 43, 44, 45]. In particular, the polyelectrolytes are more rigid because of long-range electrostatic interactions between the monomers, and these tend to behave like a semi-flexible rods in good or theta-solvents [44]. The rigid conformations of polymers with exceptional stiffness are often found in aromatic or hetero-cyclic backbones [46, 47, 48]. In this regime, the radius of gyration scales linearly with NN, so that ν=1\nu=1. Rigid-rod polymers are of interest because of their high mechanical and thermal/oxidative stability, as well as environmental resilience. Experimental studies of semiflexible and rigid star polymers reveal their unique features in molecular recognition and self-assembly behavior [49, 50, 51]. The rigid branched dendrimeric structures were synthesized in Ref. [52] based on poly([1.1.1]propellane) and in Ref. [53], based on fibrin polymerization.

In the present study, we focus on the perimeter shell properties of branched macromolecules that may be related to their ability to adsorb small particles. Their constituent chains are constructed in an algorithmic way as: a set of collinear monomers; a self-avoiding walk (SAW); tightly packed monomers. These mimic the ground states for the limit cases of a rigid rod, flexible chain, and collapsed chain conformational states, respectively. Our analysis is performed using the discrete model of a growing polymer chain on regular lattices within the framework of self-avoiding random walk - SAW [54, 55, 56] which captures the essential structural properties of macromolecules on a macroscopic scale without requiring a detailed description of intermolecular potential. Within the framework of this model, each consecutive step of SAW trajectory represents a single monomer of the polymer structure. This approach has been successfully applied in our previous studies [57, 58] focusing on the properties that can be related to adsorption processes involving linear and branched polymer architectures. Here, we focus on the perimeter of SAW, as given by the set of sites occupied by SAW trajectory with at least one vacant lattice site in their neighborhood. The perimeter shell is thus defined as the set of vacant lattice sites adjacent to the perimeter sites; related to the concept of “atmosphere” of lattice clusters [59, 60, 61, 62]. We focus on the following structural characteristics of perimeter shell:

  • perimeter shell size per monomer nˉa\bar{n}_a: The average number of vacant sites that are adjacent to at least one monomer of the perimeter, normalized by the total number of monomers NN.

  • coordination accessibility per monomer nˉs\bar{n}_s: The average total number of open neighboring directions for all the monomers of the perimeter normalized by NN;

  • mean coordination number of perimeter shell nbondn_{\mathrm{bond}}: The average number of polymer monomers adjacent to a single perimeter shell site.

By interpreting the perimeter shell as an adsorbent matrix, these parameters are related to experimentally observable adsorption characteristics as follows. nˉa\bar{n}_a represents the theoretical maximum number of particles that can be potentially adsorbed at the neighborhood of a single monomer and can be associated with maximum adsorption capacity; nˉs\bar{n}_s gives the maximum number of potential coordination bonds between a single monomer and adsorbed particles; nbondn_{\mathrm{bond}} corresponds to the average number of bonds established between a single adsorbed particle and the adjacent monomers of a perimeter, related in the same way to the concept of adsorption coordination strength (an analogue of binding energy in the case of discrete lattice or off-lattice models with explicit pairwise attractive potential between monomers and particles).

These parameters can be mapped onto realistic disordered environment where the adsorbent polymer matrix captures target particles (pollutants) from a surrounding solution. Within the formalism of a lattice model, we can introduce point-like defects with a concentration cc (0≤c≤10 \le c \le1). At smaller cc, the obstacles fill the available sites randomly with probability proportional to cc. Thus, the average number of actual adsorbed particles per monomer na(c)n_a(c) is assumed to scale as na(c)∼cnˉan_a(c) \sim c\bar{n}_a, while the average number ns(c)n_s(c) of active bonds formed between the monomer and the adsorbed particles: ns(c)∼cnˉsn_s(c) \sim c\bar{n}_s. It is important to mention that assumed linear scaling with the defect concentration cc is a phenomenologically observed in the dilute low-density limit of a non-interacting adsorbate e.g. in our previous study [57]. Our lattice parameters nˉa\bar{n}_a and nˉs\bar{n}_s represent strictly the geometric upper bounds (per the total number of available sites and coordination directions in a site-filling scenario), rather than describing the entire functional range of a thermodynamic adsorption isotherm. Note that the mean coordination number nbondn_{\mathrm{bond}} is an intrinsic geometric property of the polymer’s local conformation and remains invariant with respect to the defect concentration cc.

The remainder of this paper is organized as follows. Section 2 introduces the lattice model of the growing polymer chain and details of the calculation of observables. Section 3 provides a thorough analysis of the data obtained for branched polymer structures in rigid, fully flexible, and tightly packed regimes. Finally, Section 4 presents our conclusions and outlook.

The method

To apply the polymer growth algorithm strengthened by pruned-enriched population control (PERM) [63, 64] for constructing the structures shown in Figure 1, we consider a three-dimensional (3D3D) simple cubic lattice. The first monomer of a polymer structure is chosen at the point with coordinates R⃗0\vec{R}_0. Next, a branched structure is generated by growing a predefined number ff of the first generation chains (branches) that radiate from this core. The first monomer (n=1n = 1) of each linear branch is placed at a randomly selected unoccupied nearest-neighbor site of the central core {x0,y0,z0}\{x_0, y_0, z_0\}, which is shared by all ff branches. Consequently, for a simple cubic lattice, the number of branches is geometrically constrained to 1≤f≤61 \le f \le6.

Each kk-th branch (k=1,…,fk = 1, \ldots, f) grows sequentially: its nn-th monomer (n=1,…,Nn = 1, \ldots, N) is placed at a site with coordinates R⃗nk=(xnk,ynk,znk)\vec{R}^k_n = (x^k_n, y^k_n, z^k_n) chosen from the unoccupied nearest neighbors of the (n−1)(n - 1)-th monomer (out of up to 5 available sites). To construct a snowflake configuration with gg twigs, each terminal monomer of the primary branches (R⃗Nk\vec{R}^k_N with k=1,…,Nk = 1, \ldots, N) is treated as a new starting core for l=1,…,gl = 1, \ldots, g growing linear chains, such that R⃗Nk=R⃗0k,l\vec{R}^k_N = \vec{R}^{k,l}_0. The generation process then proceeds for each twig reaches its specified length NN with the monomer coordinates denoted as R⃗nk,l=(xnk,l,ynk,l,znk,l)\vec{R}^{k,l}_n = (x^{k,l}_n, y^{k,l}_n, z^{k,l}_n). At this point it is important to note that our approach focuses on the three distinct conformational regimes (states) of polymer chains, generated exclusively by using qualitatively distinct algorithmic growth rules as listed below:

  • rigid polymer structures: Each branch is straight and runs along one of a randomly chosen spatial axis (chain bending is prohibited).

  • flexible polymer structures: The nn-th monomer is any unoccupied nearest neighbor of the n−1n-1-th one, resulting in the SAW construction algorithm for each chain.

  • tightly packed polymer structures: For each available neighbor site ii of (n−1)(n-1)th monomer, its number of contacts ziz_i is calculated. The site with the highest ziz_i is selected for nnth monomer. In the case when two or more sites share the same maximum number of contacts, one of them is chosen at random with equal probability.

A weight WnW_n is assigned to each trial configuration at the nnth step, given by the Rosenbluth weights [63]

WnRosenbluth=∏l=1nml.(2)W_n^{\mathrm{Rosenbluth}} = \prod_{l=1}^{n} m_l. \tag*{(2)}

Here, mlm_l counts the number of unoccupied nearest neighbor sites in the restricted subset of sites that satisfy the corresponding algorithmic “filters” at step ll. Namely, mlm_l is equal to: (i) the number of non-occupied sites along the chosen spatial axis in the rigid polymer regime; (ii) all unoccupied nearest neighbors in the flexible polymer regime; and (iii) only the subset of unoccupied sites that corresponds to the degenerate maximum number of contacts ziz_i in the tightly packed regime.

Since the algorithm chooses one site uniformly at random from the mlm_l allowed options at each step ll, the conditional probability of making this choice is 1/ml1/m_l. The total probability of generating the configuration at nnth step is the product over all sequential growth steps:

Pn=∏l=1n1ml.(3)P_n = \prod_{l=1}^{n} \frac{1}{m_l}. \tag*{(3)}

Thus, the Rosenbluth weight defined in Eq. (2) satisfies the condition Wn=1/PnW_n = 1/P_n. The growth process continues until the required total number of monomers in each branch, NN, is reached. The total number of monomers in constructed branched structures are thus given by Ntotal=1+fgNN_{\mathrm{total}} = 1 + f_gN and Ntotal=1+(f+fg)NN_{\mathrm{total}} = 1 + (f + f_g)N for star-like and snowflake topologies, respectively.

According to PERM population control, configurations with too small weights are pruned, while the sample is being enriching with copies of the high-weight configurations [64]. The thresholds values Wn<W_n^{<} and Wn>W_n^{>} are chosen; if the current weight WnW_n of an nn-monomer chain is less than Wn<W_n^{<}, the chain is either discarded with probability 1/21/2, or it is kept and its weight is doubled. If WnW_n exceeds Wn>W_n^{>}, the configuration is doubled and the weight of each copy is taken as half the original weight. The pruning-enrichment control parameters Wn<W_n^{<} and Wn>W_n^{>} are adjusted in such a way that on average 10 chains of total length NtotalN_{\mathrm{total}} are generated per each iteration.

The size and the shape characteristics of a polymer are obtained conveniently from its gyration tensor, Q\mathbf{Q}, with the components defined as

Qab=12Ntotal2∑n=1N∑k=1f∑l=1g(ank,l−amr,s)(bnk,l−bmr,s),(4)Q_{ab} = \frac{1}{2N_{\mathrm{total}}^2}\sum_{n=1}^{N}\sum_{k=1}^{f}\sum_{l=1}^{g}(a_n^{k,l}-a_m^{r,s})(b_n^{k,l}-b_m^{r,s}), \tag*{(4)}

where {a,b}={x,y,z}\{a,b\} = \{x,y,z\}.

The squared radius of gyration Rg2R_g^2 is thus given by the trace of Q\mathbf{Q}:

Rg2=12Ntotal2∑n=1N∑k=1f∑l=1g(ank,l−amr,s)2.(5)R_g^2 = \frac{1}{2N_{\mathrm{total}}^2}\sum_{n=1}^{N}\sum_{k=1}^{f}\sum_{l=1}^{g}(a_n^{k,l}-a_m^{r,s})^2. \tag*{(5)}

The asphericity parameter AA is evaluated based on the eigenvalues λi\lambda_i (i=1,2,3i = 1, 2, 3) of the tensor Q\mathbf{Q} according to [30, 31]:

Ad=1d(d−1)∑i=1d(λi−λˉ)2λˉ2,(6)A_d = \frac{1}{d(d-1)}\sum_{i=1}^{d}\frac{(\lambda_i-\bar{\lambda})^2}{\bar{\lambda}^2}, \tag*{(6)}

with λˉ≡Tr⁡Q/d\bar{\lambda} \equiv\operatorname{Tr}\mathbf{Q}/d being the mean arithmetic eigenvalue of the gyration tensor eigenvalues.

Once a configuration with the specified parameters, NtotalN_{\mathrm{total}}, ff, and gg, is constructed, we assign a label O(x(j),y(j),z(j))O(x(j),y(j),z(j)) to each jjth lattice site with coordinates {x(j),y(j),z(j)}\{x(j),y(j),z(j)\} such that

O(x(j),y(j),z(j))={1,if jth site is occupied by monomer,0,otherwise.O(x(j),y(j),z(j)) = \begin{cases} 1, & \text{if }j\text{th site is occupied by monomer},\\ 0, & \text{otherwise}. \end{cases}

Then, the accessibility (i.e., the total number of potential adsorption sites) of the polymer structure is given by

Ns(Ntotal)=∑jδO(x(j),y(j),z(j)),1[δO(x(j)±1,y(j),z(j)),0+δO(x(j),y(j)±1,z(j)),0+δO(x(j),y(j),z(j)±1),0].(7)N_s(N_{\mathrm{total}}) = \sum_j \delta_{O(x(j),y(j),z(j)),1}\left[\delta_{O(x(j)\pm1,y(j),z(j)),0}+\delta_{O(x(j),y(j)\pm1,z(j)),0}+\delta_{O(x(j),y(j),z(j)\pm1),0}\right]. \tag*{(7)}

so that the normalized value nˉs=Ns(Ntotal)/Ntotal\bar{n}_s = N_s(N_{\mathrm{total}})/N_{\mathrm{total}} represents the average mean total coordination accessibility (the potential number of outgoing active bonds) per single monomer.

To estimate the total number Na(Ntotal)N_a(N_{\mathrm{total}}) of unoccupied sites within the perimeter shell of a polymer structure with NtotalN_{\mathrm{total}} monomers, we assign another label, v(x(j),y(j),z(j))v(x(j),y(j),z(j)), to each jjth lattice site, initializing its value to zero. Then, the following condition is checked for each jj-th site:

if O(x(j),y(j),z(j))=1O(x(j),y(j),z(j)) = 1, then

  • if O(xj±1,yj,zj)=0O(x_j \pm1,y_j,z_j) = 0, then v(xj±1,yj,zj)=v(xj±1,yj,zj)+1v(x_j \pm1,y_j,z_j) = v(x_j \pm1,y_j,z_j) + 1; if v(xj±1,yj,zj)=1v(x_j \pm1,y_j,z_j) = 1 then $N_a(N_{\mathrm{total}}) = N_a(N_{\mathrm{total}}) + 1

  • if O(xj,yj±1,zj)=0O(x_j,y_j \pm1,z_j) = 0, then v(xj,yj±1,zj)=v(xj,yj±1,zj)+1v(x_j,y_j \pm1,z_j) = v(x_j,y_j \pm1,z_j) + 1; if v(xj,yj±1,zj)=1v(x_j,y_j \pm1,z_j) = 1 then $N_a(N_{\mathrm{total}}) = N_a(N_{\mathrm{total}}) + 1

  • if O(xj,yj,zj±1)=0O(x_j,y_j,z_j \pm1)=0, then v(xj,yj,zj±1)=v(xj,yj,zj±1)+1v(x_j,y_j,z_j \pm1)=v(x_j,y_j,z_j \pm1)+1; if v(xj,yj,zj±1)=1v(x_j,y_j,z_j \pm1)=1 then Na(Ntotal)=Na(Ntotal)+1N_a(N_{\mathrm{total}})=N_a(N_{\mathrm{total}})+1

so that at the end of the procedure, Na(Ntotal)N_a(N_{\mathrm{total}}) counts the total number of vacant sites within a perimeter shell of the polymer structure. The normalized value, nˉa=Na(Ntotal)/Ntotal\bar{n}_a=N_a(N_{\mathrm{total}})/N_{\mathrm{total}}, representing the perimeter shell size (that can be related to the maximum adsorption capacity) per single monomer, is introduced in a similar way.

Finally, the values of v(x(i),y(i),z(i))v(x(i),y(i),z(i)) give the mean coordination number of perimeter shell nbondn_{\mathrm{bond}} (that can be interpreted as the adsorption coordination strength), the number of bonds that can be established between a particle located at site ii and the monomers of the polymer structure.

The configurational averaging for any observable of interest OO is given by

Oˉ=∑confWNconfOZN(8)\bar{O}=\frac{\sum_{\mathrm{conf}}W_N^{\mathrm{conf}}O}{Z_N} \tag*{(8)}

with WNconfW_N^{\mathrm{conf}} being the weight of an NN-monomer chain in a given configuration given by (2) and ZN=∑confWNconfZ_N=\sum_{\mathrm{conf}}W_N^{\mathrm{conf}} is the partition sum.

To provide a clear geometric view on the introduced parameters, a schematic example of a polymer chain of length N=8N=8 on a square lattice (for easier visualization) is given in Figure 2. The perimeter shell is formed exclusively by the adjacent unoccupied lattice sites (shown with open circles), yielding a total shell size of Na=14N_a=14. The outgoing dashed lines illustrate the open, unoccupied neighboring directions from each monomer, contributing to the total coordination accessibility Ns=18N_s=18. To calculate the mean coordination number ⟨nbond⟩\langle n_{\mathrm{bond}}\rangle, we notice that out of the 14 shell sites, 11 sites are adjacent to exactly one polymer monomer (single contact), 2 sites share contacts with two monomers (double contacts), and 1 mostly “trapped” perimeter site is enclosed by three monomers (triple contact). Summing these structural interactions gives ⟨nbond⟩=(11×1+2×2+1×3)/14=18/14\langle n_{\mathrm{bond}}\rangle=(11\times1+2\times2+1\times3)/14=18/14.

Schematic representation of the 2D lattice polymer chain

Figure 2. Schematic representation of the 2D lattice polymer chain with N=8N = 8 monomers, shown by green filled circles. The vacant lattice sites adjacent to the monomers, shown with Na=14N_a = 14 open circles, constitute the perimeter shell. Dashed lines denote the available coordination directions that determine the total coordination accessibility (Ns=18N_s = 18). The resulting normalized metrics thus are: the perimeter shell size per monomer nˉa=14/8\bar{n}_a = 14/8, the mean total coordination accessibility per monomer nˉs=18/8\bar{n}_s = 18/8, the mean coordination number per shell site ⟨nbond⟩=18/14\langle n_{\mathrm{bond}}\rangle= 18/14.

One easily notes that the coordination accessibility nˉs\bar{n}_s, perimeter shell size nˉa\bar{n}_a, and the mean coordination number of perimeter shell, nbondn_{\mathrm{bond}}, are related by the identity nˉs=nˉa⋅nbond\bar{n}_s=\bar{n}_a\cdot n_{\mathrm{bond}} [57], thus these quantities are not fully independent. Therefore, in the subsequent discussions and presentation of results of our analysis, we primarily concentrate on the two key independent observables, nˉa\bar{n}_a and nbondn_{\mathrm{bond}}, which together provide a comprehensive description of the adsorption performance.

Results

Rigid polymers of various architecture.

We begin by examining the observables of interest for the case of the polymers with the rod-like constituent chains. These can be found among polyelectrolytes [44], aromatic or heterocyclic backbones [46, 47, 48], polysaccharides [65, 66], etc. This case allows for simple analytical treatment. For a rod-like linear polymer chain of length NtotalN_{\mathrm{total}}, there are 2 end monomers with 5 available connections each, and Ntotal−2N_{\mathrm{total}}-2 inner monomers with 4 connections each. Consequently, the mean total coordination

accessibility per monomer for a single monomer is given by:

nˉs,f=1stiff(Ntotal)=5⋅2+4(Ntotal−2)Ntotal=2+4NtotalNtotal.(9)\bar{n}_{s,f=1}^{\mathrm{stiff}}(N_{\mathrm{total}})=\frac{5\cdot2+4(N_{\mathrm{total}}-2)}{N_{\mathrm{total}}}=\frac{2+4N_{\mathrm{total}}}{N_{\mathrm{total}}}. \tag*{(9)}

Following the same reasoning, we have nˉa,f=1stiff(Ntotal)=nˉs,f=1stiff(Ntotal)\bar{n}_{a,f=1}^{\mathrm{stiff}}(N_{\mathrm{total}})=\bar{n}_{s,f=1}^{\mathrm{stiff}}(N_{\mathrm{total}}) and nbondstiff(Ntotal)=1n_{\mathrm{bond}}^{\mathrm{stiff}}(N_{\mathrm{total}})=1.

For a star-shaped structure of the same total molecular weight, NtotalN_{\mathrm{total}}, consisting of ff rod-like branches each of length N=(Ntotal−1)/fN=(N_{\mathrm{total}}-1)/f, we introduce the sequential numbering for the monomers along each kk-th branch (k=1,…,fk=1,\ldots,f) as nk=1,…,Nn_k=1,\ldots,N, starting from the central core. In this configuration, there are (6−f)(6-f) available connections for the central monomer, 5 connections for each of the ff endpoints (where nk=Nn_k=N), and 4 connections for the remaining Ntotal−1−fN_{\mathrm{total}}-1-f inner monomers (where nk=1,…,N−1n_k=1,\ldots,N-1). This leads to the following result:

nˉs,f>1stiff(Ntotal)=(6−f)+5f+4f(N−1)Ntotal=nˉs,f=1stiff(Ntotal).(10)\bar{n}_{s,f>1}^{\mathrm{stiff}}(N_{\mathrm{total}})=\frac{(6-f)+5f+4f(N-1)}{N_{\mathrm{total}}}=\bar{n}_{s,f=1}^{\mathrm{stiff}}(N_{\mathrm{total}}). \tag*{(10)}

The evaluation of the perimeter shell size nˉa,f>1stiff(Ntotal)\bar{n}^{\mathrm{stiff}}_{a,f>1}(N_{\mathrm{total}}) has some nuances. While the central monomer can adsorb 6−f6-f particles, and each of the end monomers (nk=Nn_k=N) and inner monomers (nk=2,…,N−1n_k=2,\ldots,N-1) has coordination accessibility of 5 and 4, respectively, some particles are shared by monomers with nk=1n_k=1 (those adjacent to the central monomer), depending on the mutual orientation of the branches. Indeed, each pair of branches positioned at a 90∘90^\circ angle leads to a decrease in nˉa,f>1stiff\bar{n}^{\mathrm{stiff}}_{a,f>1} by one compared to nˉs,f>1stiff\bar{n}^{\mathrm{stiff}}_{s,f>1}, since in this case, one site is shared by the first two monomers of these branches. In general, we have:

nˉa,f>1stiff(Ntotal)=nˉs,f>1stiff(Ntotal)−CfNtotal(11)\bar{n}^{\mathrm{stiff}}_{a,f>1}(N_{\mathrm{total}})=\bar{n}^{\mathrm{stiff}}_{s,f>1}(N_{\mathrm{total}})-\frac{C_f}{N_{\mathrm{total}}} \tag*{(11)}

with CfC_f calculated on the base of combinatorics: C2=12/15C_2=12/15, C3=8/3C_3=8/3, C4=14/3C_4=14/3, C5=8C_5=8, C6=12C_6=12. As a result, mean coordination number of perimeter shell can be evaluated as

nˉbond,f>1stiff(Ntotal)=nˉs,f>1stiff(Ntotal)nˉa,f>1stiff(Ntotal)=2+4Ntotal2+4Ntotal−Cf.(12)\bar{n}^{\mathrm{stiff}}_{\mathrm{bond},f>1}(N_{\mathrm{total}})=\frac{\bar{n}^{\mathrm{stiff}}_{s,f>1}(N_{\mathrm{total}})}{\bar{n}^{\mathrm{stiff}}_{a,f>1}(N_{\mathrm{total}})}=\frac{2+4N_{\mathrm{total}}}{2+4N_{\mathrm{total}}-C_f}. \tag*{(12)}

Our simulation data, presented in Figure 3a, Figure 4a, agree with these theoretical results well, thereby validating our simulation approach. Based on these theoretical and simulation results, we can draw the following conclusions. In such extended, rigid polymer structures, branching leads to a slight decrease in perimeter shell size nˉa,f>1,g>1stiff<nˉa,f>1,g=1stiff<nˉa,f=1stiff\bar{n}^{\mathrm{stiff}}_{a,f>1,g>1}<\bar{n}^{\mathrm{stiff}}_{a,f>1,g=1}<\bar{n}^{\mathrm{stiff}}_{a,f=1} and an increase of mean coordination number of perimeter shell nˉbond,f>1,g>1stiff>nˉbond,f>1,g=1stiff>nˉbond,f=1stiff\bar{n}^{\mathrm{stiff}}_{\mathrm{bond},f>1,g>1}>\bar{n}^{\mathrm{stiff}}_{\mathrm{bond},f>1,g=1}>\bar{n}^{\mathrm{stiff}}_{\mathrm{bond},f=1}. These effects for rigid structures can be detected only at relatively small values of NN and NtotalN_{\mathrm{total}} (for Ntotal→∞N_{\mathrm{total}}\to\infty we have nˉs,f,gstiff=nˉa,f,gstiff=4\bar{n}^{\mathrm{stiff}}_{s,f,g}=\bar{n}^{\mathrm{stiff}}_{a,f,g}=4, nˉbond,f,gstiff=1\bar{n}^{\mathrm{stiff}}_{\mathrm{bond},f,g}=1 independent of ff and gg).

Figure 3. Perimeter shell size, ⟨nˉa⟩\langle\bar{n}_a\rangle, for three representative polymer structures: comprising rigid rod chains (a), flexible (b), and tightly packed (c) regimes. Solid lines in (a) represent analytical evaluation given by Eqs. (9), (11), whereas in (c) the solid line gives a result of rough approximation of Eq. (13).

Figure 4. Mean coordination number of perimeter shell, ⟨nbond⟩\langle n_{\mathrm{bond}}\rangle, for three representative polymer structures comprising rigid chains (a), flexible (b), and tightly packed (c) regimes. Solid line in (a) represent analytical evaluation as given by Eq. (12).

Note that hereafter, in the majority of figures, we present results for only three representative architectures: the linear chain (f=1f=1), the star polymer with the maximum possible functionality (f=6f=6), and the snowflake polymer with f=4f=4 and g=3g=3 twigs. This choice is motivated by the fact that the intermediate architectures, e.g. star polymers with f=3f=3–4, have rather subtle effect on the adsorption-related observables.

We calculated also the gyration radii for the branched structures in rigid regime, presented in Figure 5a. As expected, radius of gyration scales linearly with NtotalN_{\mathrm{total}} independently of the polymer topology, whereas the value of RgR_g decreases with increasing branching of polymer structure at fixed NtotalN_{\mathrm{total}}. Additionally, we calculated the averaged asphericity, as defined by (6), of a star and snowflake structures in the rigid regime. While a stiff linear chain, f=1f=1, is characterized by the maximum possible asphericity, A=1A=1, a star with f>1f>1 branches shows gradual decrease of its asphericity with the increase of ff (Figure 6). The case of f=6f=6 is characterized by a perfect global spherical symmetry because all arms run straight along one of the three spatial axes of a 3D lattice.

Figure 5. The averaged gyration radius, ⟨Rg2⟩\langle R_g^2\rangle, for three representative polymer structures comprising rigid rod chains (a), flexible (b), and tightly packed (c) regimes as function of NtotalN_{\mathrm{total}} in double logarithmic scale.

Figure 6. Averaged asphericity ⟨A⟩\langle A\rangle for a set of star-like (a) and snowflake-like (b) structures in rigid regime. Panel (a) omits the trivial limiting cases f=1f=1 (with A=1A=1) and f=6f=6 (with A=0A=0).

Tightly packed state of polymers of various architectures

Real-life polymers are often found in collapsed, maximally compact, state, e.g. in the case of a flexible polymer in a poor solvent. In this case, collapse is driven by strong, solvent-mediated, attraction between monomers. Typical core of the micellar structure [67, 68], or collapsed polymer brush [69] is exactly in this state. To mimic collapse of real-life polymers in our simulation scenario, the compact conformational state is generated using specific algorithmic growth rules rather than a continuous variation of physical parameters within a single thermodynamic Hamiltonian. This regime also allows for a simplified analytical treatment. The equilibrium shape of a tightly packed polymer “globule” is dictated by the minimization of its surface area, resulting in a shape with the highest possible monomer-monomer contacts. In continuous 3D space, this shape is a sphere. However, in a lattice polymer model, the monomer arrangement is restricted to a grid of discrete positions determined by the lattice symmetry, which affects the globule’s shape. The analogue of a sphere of radius rr on a discrete cubic lattice is most naturally considered to be a cube with an edge length nedge=2rn_{\mathrm{edge}}=2r, as the simplest irreducible closed volume for the Chebyshev metric [70]. Indeed, it is defined by the condition for the maximum coordinates: max⁡(∣x∣,∣y∣,∣z∣)≤r\max(|x|,|y|,|z|)\leq r. For the case of a polymer composed of NtotalN_{\mathrm{total}} monomers we should take into account, that perfect cube can be formed only if the edge

length as given by nedge=Ntotal1/3n_{\mathrm{edge}}=N_{\mathrm{total}}^{1/3} is integer, otherwise the compact rectangular cuboid, octahedron or rhombododecahedron will be formed. The perfect cubic ground state for lattice polymers in 3D are observed e.g. in Ref. [71] at very low temperature limit (extremely poor solvent). The temperature fluctuations and entropic effects lead to “smearing” of this highly symmetric crystalline structure.

To establish a basic theoretical background for analyzing the perimeter properties in the maximally compact state, we consider a simple limiting geometric reference: a perfectly filled, non-fluctuating cubic cluster composed of NtotalN_{\mathrm{total}} monomers. It is important to emphasize that this cuboidal configuration does not represent the true thermodynamic equilibrium ground state of a collapsed branched polymer structure under nearest-neighbor attraction, which is highly non-trivial. Instead, it serves purely as a rigid geometric upper bound for a completely packed discrete volume. Under this strict limiting assumption of a perfect cube with an integer edge length nedge=Ntotal1/3n_{\mathrm{edge}}=N_{\mathrm{total}}^{1/3}, the distribution of monomers is as follows: there are 8 corner particles with 3

available connections each, 12(nedge−2)12(n_{\mathrm{edge}}-2) particles along the edges with 2 connections, 6(nedge−2)26(n_{\mathrm{edge}}-2)^2 particles on the faces with only 1 open connection, and (nedge−2)3(n_{\mathrm{edge}}-2)^3 interior particles that are completely ’screened’. Inside the globule, the polymer is packed without internal voids, representing a ’frozen’ or ’globular’ phase. As a result:

nˉsglobule=8⋅3+12(nedge−2)⋅2+6(nedge−2)2Ntotal=6Ntotal1/3.(13)\bar{n}_{s}^{\mathrm{globule}}=\frac{8\cdot3+12(n_{\mathrm{edge}}-2)\cdot2+6(n_{\mathrm{edge}}-2)^2}{N_{\mathrm{total}}}=\frac{6}{N_{\mathrm{total}}^{1/3}}. \tag*{(13)}

This value slowly approaches 0 in the limit of infinite NN. Straightforwardly we obtain nˉaglobule=nˉsglobule\bar{n}_{a}^{\mathrm{globule}}=\bar{n}_{s}^{\mathrm{globule}}, and correspondingly nbondglobule=1n_{\mathrm{bond}}^{\mathrm{globule}}=1. One should note that this idealized cuboidal configuration serves purely as a basis for comparison with our numerical data.

Now we proceed to the simulation results. The gyration radii for the branched structures under consideration are presented in Figure 7.

Figure 7. Averaged bounding-box anisotropy index ⟨E⟩\langle E\rangle for representative polymer structures in tightly packed regime.

5c. The fitting of resulting curves restores the value of scaling exponent ν=0.334(1)\nu= 0.334(1) known for the compact globular polymer state [33], independently of the topology of considered polymer structures. Let us begin with some general considerations about the shape of the cluster formed by polymer structures in the compact regime. The linear dimensions of the cluster (“bounding box”) denoted as sizexsize_x, sizeysize_y, and sizezsize_z, define the minimum orthogonal volume that completely encompasses the set of NtotalN_{\mathrm{total}} monomers within the three-dimensional cubic lattice. Due to the discrete nature of the grid, these parameters represent the total number of lattice sites along each coordinate axis and are computed using the extreme coordinates of the set of monomers of polymer structure in collapsed states. To quantitatively assess the spatial anisotropy and geometrical distortion of the shape of collapsed polymer structure, we introduce an integrated scalar metric designated as the bounding-box anisotropy index (EE):

E=(r1−1)2+(r2−1)2+(r3−1)23,(14)E = \sqrt{\frac{(r_1-1)^2+(r_2-1)^2+(r_3-1)^2}{3}}, \tag*{(14)}

with the axial aspect ratios r1=sizex/sizeyr_1 = size_x/size_y, r2=sizex/sizezr_2 = size_x/size_z, and r3=sizey/sizezr_3 = size_y/size_z. This non-negative metric provides a rigorous, coordinate-independent measure of shape variation; a value of E=0E = 0 corresponds to a cubic shape, whereas progressively higher values indicate an increasing degree of structural elongation towards a cuboidal shape. Note that parameter (14) serves primarily as a convenient operational tool to quantify how the bounding volume of the collapsed cluster deviates from a perfect cube (the natural irreducible volume in Chebyshev metric on a cubic grid) toward a cuboidal shape. Our simulation data for the averaged value of this parameter, as given in Figure 7, demonstrate a deviation of the shape of compact cluster from the perfect cube, more pronounced for the linear polymer structures. This tendency is further supported by evaluating the asphericity, AA, of the corresponding polymer structures (Figure 8). In its collapsed state, a linear polymer chain with f=1f = 1 forms the most “unspherical” (or rather, within the framework of a regular lattice, the most “uncubical”) cluster. Increasing the branching leads to enhanced shape symmetry in the resulting cluster. Thus both independent metrics, based on based on bounding-box anisotropy index EE and rotationally invariant asphericity AA, are in agreement: the linear chain (f=1f = 1) forms the most elongated/least spherical structure in the tightly packed state, whereas branching drives the system toward globally symmetric shape with asphericity value close to zero.

Figure 8. Averaged asphericity ⟨A⟩\langle A\rangle for a set of star-like (a) and snowflake-like (b) structures in tightly packed regime.

The internal structure and effective local cavity structure of the collapsed polymer adsorbent strongly influence its adsorption efficiency. While a full characterization of local cavity structure involves complex cavity size distributions and interconnectivity, we can estimate local site availability by analyzing the distribution of the mean coordination number nbondn_{\mathrm{bond}}. The perimeter shell sites can be occupied by adsorbed particles, making their distribution (shown in Figure 9a) a direct measure of structural accessibility. Note that only data for nbond≥3n_{\mathrm{bond}} \ge3 are shown here, to concentrate specifically on the probability of forming local geometric cavities specifically, unoccupied lattice sites surrounded by 4 or 5 monomers. This agrees with earlier simulations [40], which demonstrated that even in the collapsed state, each arm of a star polymer remains relatively extended due to the high monomer density near the central branching point. We should mention that our model evaluates the static local geometric accessibility based on frozen snapshots of the polymer matrix, which does not account for the kinetics of the adsorption process. In real polymer systems, the size of spatial cavities within collapsed or branched architectures will depend on the competition between the timescales of adsorbate diffusion and polymer chain fluctuations.

Figure 9. Probability distribution of the number of contacts, nbondn_{\mathrm{bond}} with nbond≥3n_{\mathrm{bond}} \ge3 between adsorbed particles and monomers, for various polymer structures in the tightly packed state (a) and flexible state (b) at Ntotal=49N_{\mathrm{total}} = 49.

The general tendency is preserved for the geometry characteristics of polymer structures in tightly packed state that are related to their adsorbing abilities, observed earlier for their counterparts comprising rod-like chains: branching leads to a slight decrease of perimeter shell size: nˉa,f>1,g>1globule<nˉa,f>1,g=1globule<nˉa,f=1globule\bar{n}^{\mathrm{globule}}_{a,f>1,g>1} < \bar{n}^{\mathrm{globule}}_{a,f>1,g=1} < \bar{n}^{\mathrm{globule}}_{a,f=1} and an increase of mean coordination number of perimeter shell nbond,f>1,g>1globule>nbond,f>1,g=1globule>nbond,f=1globulen^{\mathrm{globule}}_{\mathrm{bond},f>1,g>1} > n^{\mathrm{globule}}_{\mathrm{bond},f>1,g=1} > n^{\mathrm{globule}}_{\mathrm{bond},f=1} (cf. also our simulation data presented in Figure 3c, Figure 4c).

Flexible coils of various architecture

Typical real-life polymers are often found in a swollen state (flexible polymer in a good solvent, etc.). Our simulation results for this case are presented in Figure 3b and Figure 4b. Regarding the impact of branching on the observables, the main tendencies remain consistent with those observed previously: the maximum perimeter shell size nˉacoil\bar{n}^{\mathrm{coil}}_{a} decreases with increasing branching, while the mean coordination number of perimeter shell ⟨nbondcoil⟩\langle n^{\mathrm{coil}}_{\mathrm{bond}}\rangle conversely increases. These tendencies can be clearly followed in Figure 10, which summarizes our data obtained for a set of star-like and snowflake-like structures at fixed total molecular weight Ntotal=121N_{\mathrm{total}} = 121 for comparison.

Figure 10. Perimeter shell size ⟨nˉa⟩\langle\bar{n}_{a} \rangle (a) and mean coordination number of perimeter shell ⟨nbond⟩\langle n_{\mathrm{bond}} \rangle (b) for a set of star and snowflake structures at fixed Ntotal=121N_{\mathrm{total}} = 121 in flexible regime as function of ff.

In this study we do not evaluate local cavity structure of polymers explicitly, via distribution of pore sizes, their interconnectivity, etc. Instead we consider related characteristic such as the distribution function of the number of bonds between unoccupied lattice sites and monomers, nbondn_{\mathrm{bond}}, as presented in Figure 9b. The general tendency is inline with the case of tightly packed regime (cf. Fig. 9a: an increased degree of branching supports the higher probability of forming cavities, unoccupied lattice sites surrounded by more than 4 (up to 6) monomers; however this effect is less pronounced comparing to the tightly packed state.

For completeness, we also analyzed the gyration radii for the representative set of polymer topologies, as presented in Figure 5b. The fitting of resulting curves restores the value of scaling exponent ν=0.586(1)\nu= 0.586(1) known for the scaling of polymers in good solvent regime, as given e.g. in [34]. The data for the asphericity parameter ⟨A⟩\langle A \rangle of the polymer structures under consideration in the good solvent regime are presented in Figure 11a, they agree well with the corresponding values obtained earlier for star-like polymers with up to f=6f = 6 in Ref. [36]. The results for a representative snowflake polymer structure with a fixed f=4f = 4 and various gg (Figure 11b) support the general tendency toward an increasing symmetry of the typical polymer configuration with a higher degree of branching.

Figure 11. Averaged asphericity ⟨A⟩\langle A \rangle for a set of star-like (a) and snoflake-like (b) structures in flexible regime.

Comparison between the adsorbents of various architectures comprising rod-like, flexible, or tightly packed chains

Let us overview the three regimes discussed above, starting with the data for perimeter shell size presented in Figure 3. Polymers comprising rigid-rod chains are the most effective in terms of the geometric accessibility of their monomers for adsorption of particles: as the total molecular weight NtotalN_{\mathrm{total}} increases, ⟨nˉarod⟩\langle\bar{n}_{a}^{\mathrm{rod}} \rangle reaches a value of 4 in the large-NN limit for all branching topologies (Figure 3a). In the case of the structures comprising flexible coiled chains (Figure 3b), the values of ⟨nˉacoil⟩\langle\bar{n}_{a}^{\mathrm{coil}} \rangle for each topology are considerably smaller compared to those in the rigid state. In the case of tightly packed polymeric structures, the perimeter shell size is, as expected, the lowest. At small NN, the tightly packed structures (globules) formed by branched structures are more ’porous’ and thus characterized by smaller values of ⟨nˉaglobule⟩\langle\bar{n}_{a}^{\mathrm{globule}} \rangle compared to linear chains. With increasing NN, the differences between the ⟨nˉaglobule⟩\langle\bar{n}_{a}^{\mathrm{globule}} \rangle values for various topologies practically disappear.

We continue with the results for the mean coordination number of perimeter shell shown in Figure 4. Structures comprising rod-like chains are clearly the weakest in this regard: as the total molecular weight NtotalN_{\mathrm{total}} increases, ⟨nbondrod⟩\langle n_{\mathrm{bond}}^{\mathrm{rod}}\rangle reaches a limiting value of 1 in the large-NN limit for any branching topology (Figure 4a). The structures in a flexible regime (Figure 4b) are characterized by the highest values of ⟨nbondcoil⟩\langle n_{\mathrm{bond}}^{\mathrm{coil}}\rangle. Note that while the mean coordination number of perimeter shell of a linear flexible chain slightly increases with NN, whereas the values for branched structures comprising flexible chains, decrease with NN. This decrease is due to the diminishing role of the mutual crowding effect of branches in the vicinity of the branching cores. In the tightly packed regime, the larger values of ⟨nˉbondglobule⟩\langle\bar{n}_{\mathrm{bond}}^{\mathrm{globule}}\rangle of branched structures comparing with that of linear one is caused by more pronounced “local cavity structure” of clusters formed by them; this difference diminishes with increasing NN. The zigzag-like behavior of the number of available open sites (pronounced for nbondn_{\mathrm{bond}} as a function of arm length NN, as shown in Figure 4), is primarily due to the discrete, parity-dependent nature of the lattice, combined with the excluded volume effect at the crowded central node. This occurs because the number of available adjacent lattice sites depends heavily on whether NN is odd or even, reflecting the bipartite nature of the

cubic lattice; monomers on even steps often face different crowding constraints than those on odd steps [72, 73]. For large NN, this zigzag effect is gradually vanishing.

As one can see from Figure 3 and Figure 4, the density of branching in polymer structures strongly affects its geometric properties related to adsorption. To analyze this effect quantitatively, we introduce a dimensionless branch-point density that takes into account both the number of branching centers and their functionalities, according to

ρf,g=(f−2)+(g−1)Ntotal.(15)\rho_{f,g} = \frac{(f-2)+(g-1)}{N_{\mathrm{total}}}. \tag*{(15)}

Thus, one can compare structures with various branching topology and fixed NtotalN_{\mathrm{total}}. Whereas for the linear chain we have the minimum value ρf=1,g=2=0\rho_{f=1,g=2} = 0, it increases for star structures with increasing ff (e.g. for f=6f=6-branch star ρf=6,g=0=4/Ntotal\rho_{f=6,g=0} = 4/N_{\mathrm{total}}), and grows further for the snowflake topologies (e.g. at f=4f=4, g=3g=3 one has ρf=4,g=3=10/Ntotal\rho_{f=4,g=3} = 10/N_{\mathrm{total}}). The systematic increase of ρf,g\rho_{f,g} enhances local monomer crowding, and thus reduces the perimeter shell size ⟨nˉa⟩\langle\bar{n}_a\rangle, but increases the mean coordination number of perimeter shell, ⟨nbond⟩\langle n_{\mathrm{bond}}\rangle. The effect is very prominent for the case of systems comprising rod-like chains, frames (a) in both Figure 3 and Figure 4, and can be more easily visualized for this case. Mutual arrangement of a polymer and of all possible impurities adsorbed on it, is shown in Figure 12 for the 2D case and the same number of monomers in all three cases equal to 25. Polymer sites are shown in blue, adsorbed impurities connected via a single bond in green, and adsorbed impurities connected via a double bond in red. It is clear, that in the vicinity of each branching point there are sites, available for adsorption, that are neighbors to more than one chain. As the result, the total number of adsorbed impurities for branched architectures decreases and is equal to 52, 48, and 36 in frames (a), (b), and (c) of Figure 12, respectively. However, the impurities located at these sites (shown in red) form more than one link with the polymer, and their fraction for branched architectures grows and is: 0, 0.083, and 0.333 for the same respective cases. This explains the geometric origins for the effect of branching on the adsorption characteristics as seen in frames (a) of Figure 3 and Figure 4.

Figure 12. Two-dimensional (2D) visualization of the spatial arrangement of perimeter shell of polymer structure comprising rod-like chains. Panels show: (a) linear polymer, (b) a star-like polymer, and (c) a partial snowflake polymer. Perimeter sites are depicted in blue, perimeter shell sites connected via a single bond in green, and those connected via a double bond in red. The total polymer length is identical in all three cases and equals 25 monomers.

The effect of polymer flexibility is illustrated in Figure 13 for the linear polymer of 25 monomers. One can see that the self-avoidance of a flexible chain, see frame (a), has the similar effect on both ⟨nˉa⟩\langle\bar{n}_a\rangle and ⟨nbond⟩\langle n_{\mathrm{bond}}\rangle as chemical branching. Indeed, bending of a chain promotes formation of internal geometric cavity, where the restricted number of impurities can be adsorbed, albeit with their strong bonding with two or three monomers. For the case of a tightly packed state, see frame (b), we depicted the case of an ideal 2D packing with no voids in the interior of a chain. The situation is different in 3D, but the general trend for both ⟨nˉa⟩\langle\bar{n}_a\rangle and ⟨nbond⟩\langle n_{\mathrm{bond}}\rangle to stay on a low side, as is seen Figs. 3 and 4, is the same. The experimental adsorption capacity is affected by both availability of adsorbing sites (characterized by ⟨nˉa⟩\langle\bar{n}_a\rangle), and by the mean coordination number of perimeter shell, ⟨nbond⟩\langle n_{\mathrm{bond}}\rangle. The importance of the latter can be envisaged in the case, when the minimum mean coordination number of perimeter shell, required for stable binding of impurities over prolonged time, can not be provided by a single monomer-impurity bond, but requires double or triple bonding. We note, that despite the similarities between the effect of branching and coiling of a flexible linear chain on ⟨nbond⟩\langle n_{\mathrm{bond}}\rangle, the effect from the latter is always weaker, see Figure 4. Branched polymer is characterized by a fixed set of branching points resulting in formation of geometric cavities, adjacent to each of them.

Figure 13. Visualization of the effect of flexibility for the case of a linear chain with 25 monomers. (a) a flexible constituent chains, (b) tightly packed state.

These allow for double- or triple-bonds for impurities, thus increasing the average value of ⟨nbond⟩\langle n_{\mathrm{bond}}\rangle. The effect is subtle for the set of regular architectures, each containing just a few branching points, discussed here. But it will be amplified enormously for the case of a hyperbranched polymer, with a huge number of branching points. In the case of a long linear chain, formation of such number of cavities requires very specific conformations that are improbable due to entropic considerations. This interpretation may serve as an explanation of the experimentally observed higher adsorption capacity of branched polymers compared with the linear ones, both in chelation [8, 9], and flocculation [10].

Conclusions and Discussions

We studied the perimeter shell properties for branched macromolecules (star and snowflake polymers) that can be associated with their adsorbing abilities. The study involves both theoretical calculations and Monte Carlo simulations on a cubic lattice. We compared three representative classes of polymer conformations: polymers comprising rod-like chains, flexible coil states, and tightly packed states related to the collapsed globular regime in poor solvent. The quantitative results for a set of adsorption characteristics, such as perimeter shell size per monomer nˉa\bar{n}_a and mean coordination number of perimeter shell nbondn_{\mathrm{bond}} are obtained. These lattice geometry properties allow us to connect the microscopic properties to real-world experimental observables. Indeed, the perimeter shell size defines the theoretical maximum adsorption capacity in the limit of non-interacting adsorbate: the actual amount of captured particles is expected to scale linearly with their low concentration cc, in accordance with the low-density phenomenological assumption, limited to low concentration cc and supported by some previous simulations. This mapping allows us to interpret experimental data for disordered or defect-rich polymer matrices and helps in designing responsive materials.

Our results quantitatively demonstrate that both polymer topology (degree of branching) and conformational state of their constituent chains are important in determining the geometric accessibility of its monomers that is related to the overall adsorption efficiency of these macromolecules. Across all regimes, higher branching (from linear chains to stars and snowflake architectures) leads to a moderate decrease in the perimeter shell size nˉa\bar{n}_a while simultaneously enhancing the mean coordination number of perimeter shell nbondn_{\mathrm{bond}}. This is caused by the higher local monomer density near branching cores, leading to formation of local geometric cavities and shared adsorption sites between neighboring branches. In the case of polymers comprising rigid chains, the highest perimeter shell size is observed, approaching the theoretical limit of 4 available sites per monomer for large NN in 3D with excellent agreement between simulations and analytical predictions. Polymer structures comprising flexible chains show intermediate perimeter shell size but the strongest particle bonding due to their crumpled conformations that facilitate multi-monomer contacts. However, branched structures form more porous globules (which we observe indirectly, via a set of evaluated characteristics), making their internal sites more accessible than those of linear chains. Our simulations confirm that this internal local cavity structure enhances site accessibility.

We also calculated the average asphericity parameter ⟨A⟩\langle A\rangle across all three regimes. In the rigid regime, the linear chain is fully anisotropic (A=1A=1). Increasing the core functionality ff makes the molecule more symmetric, until the structure is characterized by zero asphericity (A=0A=0) at f=6f=6. A similar trend holds for the tightly packed compact state, where branched architectures form more isotropic and compact clusters than linear chains. In the flexible regime, branching likewise promotes sphericity, which agrees with earlier theoretical and simulation studies. These shape characteristics explain the molecular mechanisms behind the observed adsorption behavior: more symmetric conformations enhance the multi-contact binding but slightly reduce the total accessible surface area.

It can be worth questioning whether our main conclusions and findings, observed in the frameworks of lattice model, are applicable to off-lattice (continuous space) polymer representations. Indeed, a simple cubic lattice model presents serious limitations, such as discrete coordination numbers and strict directional bonding. However our main conclusions rely on universal topological features. In an off-lattice continuous model, the central physical mechanisms, such as crowding of monomers around the branching cores and steric competition between branches, are topologically driven. Thus, though we expect the quantitative values of observables (such nˉa\bar{n}_a and nbondn_{\mathrm{bond}}) to differ from their counterparts for the case of continuous models, the qualitative tendencies of the impact of branched architecture and the role of chain flexibility will remain invariant.

Our forthcoming studies will extend this analysis to higher-generation dendrimers and explore the dynamics of the adsorption process.

Acknowledgments

The work was supported by the National Research Foundation of Ukraine Project “Functional efficiency of complex multi-branched polymers in adsorption-based applications: computer simulations and predictions” No. 2025.07/0366.

References

  1. [1]Kingsley Eghonghon Ukhurebor (Ed.) 2023, September). Adsorption applications for environmental sustainability, edited by Kingsley Eghonghon Ukhurebor. London, England: Institute of Physics Publishing.DOI
  2. [2]Boena Czech. “Adsorption Materials and Their Applications.” Materials 18, no. 23 (2025): 5370.DOI
  3. [3]Bingjun Pan, Bingcai Pan, Weiming Zhang, Lu Lv, Quanxing Zhang, and Shourong Zheng. “Development of polymeric and polymer-based hybrid adsorbents for pollutants removal from waters.” Chemical Engineering Journal 151, no. 1-3 (2009): 1929.DOI
  4. [4]Huda Alkhaldi, Sarah Alharthi, Salha Alharthi, Hind A. AlGhamdi, Yasmeen M. AlZahrani, Safwat A. Mahmoud, Lamia Galal Amin, Nora Hamad Al-Shaalan, Waleed E. Boraie, Mohamed S. Attia, Samera Ali Al-Gahtany, Nadiah Aldaleeli, Mohamed Mohamady Ghabashy, A. I. Sharshir, Mohamed Madani, Reem Darwesh, and Sana F. Abaza. “Sustainable polymeric adsorbents for adsorption-based water remediation and pathogen deactivation: a review.” RSC Advances 14, no. 45 (2024): 3314333190.DOI
  5. [5]Claudia F. Martín, Ev Stöckel, Rob Clowes, Dave J. Adams, Andrew I. Cooper, Jose J. Pis, Fernando Rubiera, and Cova Pevida. “Hypercrosslinked organic polymer networks as potential adsorbents for pre-combustion CO2 capture.” Journal of Materials Chemistry 21, no. 14 (2011): 5475.DOI
  6. [6]Hadiseh Masoumi (Sahar) 2024. Types of polymeric adsorbents, pp. 146. Elsevier.DOI
  7. [7]Yujie Cheng, and et al.. “A Supramolecular Polymer Network Based on Chitosan and Porphyrin and Its Application as an Adsorbent for Selective Removal of Dyes from Water.” Advanced Sustainable Systems 8, no. 11.DOI
  8. [8]Shiro Kobayashi, Kazuhisa Hiroishi, Masazumi Tokunoh, and Takeo Saegusa. “Chelating properties of linear and branched poly(ethylenimines).” Macromolecules 20, no. 7 (1987): 14961500.DOI
  9. [9]Tan Yi, Zongheng Cen, Yiwei Ji, Junlong Huang, Min Liang,, and Shaohong Liu. “Amidoximebased Star Polymer Adsorbent with Ultrahigh Uranyl Affinity for Extremely Fast Uranium Extraction from Natural Seawater.” Advanced Functional Materials 34, no. 41.
  10. [10]Ángeles Blanco, Elena Fuente, Noelia Monte,, and Carlos Negro. “Polymeric Branched Flocculant Effect on the Flocculation Process of Pulp Suspensions in the Papermaking Industry.” Industrial & Engineering Chemistry Research 48, no. 10 (2009): 48264836.DOI
  11. [11]Zhi Liu, Max Purro, Jing Qiao,, and May P. Xiong. “Multifunctional Polymeric Micelles for Combining Chelation and Detection of Iron in Living Cells.” Advanced Healthcare Materials 6, no. 17.
  12. [12]Joon Hee Jo, Seung Su Shin, Sungkweon Jeon, Sung-Joon Park, Hosik Park, You-In Park,, and Jung-Hyun Lee. “Star polymer-assembled adsorptive membranes for effective Cr(VI) removal.” Chemical Engineering Journal449 (2022): 137883.DOI
  13. [13]Abdel Wahab Mouhamad 2023, Jan). “Chelating innovative nanomaterials : elaboration, characterization and potential.” NNT: 2022UPASF052, tel-03917154, https://theses.hal.science/tel-03917154v1.DOI
  14. [14]C Gao, and D Yan. “Hyperbranched polymers: from synthesis to applications.” Progress in Polymer Science 29, no. 3 (2004): 183275.DOI
  15. [15]Weinan Xu, Petr A. Ledin, Valery V. Shevchenko,, and Vladimir V. Tsukruk. “Architecture, Assembly, and Emerging Applications of Branched Functional Polyelectrolytes and Poly(ionic liquids).” ACS Applied Materials & Interfaces 7, no. 23 (2015): 1257012596.DOI
  16. [16]In-Yup Jeon, Hyuk-Jun Noh,, and Jong-Beom Baek. “Hyperbranched Macromolecules: From Synthesis to Applications.” Molecules 23, no. 3 (2018): 657.DOI
  17. [17]Soyoung E. Seo, and Craig J. Hawker. “The Beauty of Branching in Polymer Science.” Macromolecules 53, no. 9 (2020): 32573261.DOI
  18. [18]Bruno H. Zimm, and Walter H. Stockmayer. “The Dimensions of Chain Molecules Containing Branches and Rings.” The Journal of Chemical Physics 17, no. 12 (1949): 1301–1314.DOI
  19. [19]Christos N. Likos. “Effective interactions in soft condensed matter physics.” Physics Reports 348, no. 45 (2001): 267439.DOI
  20. [20]Akira Miyake, and Karl F. Freed. “Excluded volume in star polymers: chain conformation space renormalization group.” Macromolecules 16, no. 7 (1983): 12281241.DOI
  21. [21]Gaoyuan Wei. “Shapes and Sizes of Gaussian Macromolecules. 2. Stars and Combs in Three Dimensions.” Macromolecules 30, no. 7 (1997): 21302134.DOI
  22. [22]C. von Ferber, and Yu. Holovatch. “Copolymer networks and stars: Scaling exponents.” Physical Review E 56, no. 6 (1997): 63706386.DOI
  23. [23]C. von Ferber, and Yu. Holovatch. (Eds.) 2002. Condensed Matter Physics, Special Issue Star Polymers, edited by C. von Ferber and Yu. Holovatch, Vol 5. Institute for Condensed Matter Physics of Ukrainian Academy of Sciences.
  24. [24]O Kalyuzhnyi, J M Ilnytskyi, Yu Holovatch,, and C von Ferber. “Universal shape characteristics for the mesoscopic star-shaped polymer via dissipative particle dynamics simulations.” Journal of Physics: Condensed Matter 30, no. 21 (2018): 215101.DOI
  25. [25]Y. Liu, S. Bai, T. Wu, C.C. Chen, Y. Liu, X. Chao,, and Y. Bai. “Dendronized Arm Snowflake Polymer as a Highly Branched Scaffold for Cellular Imaging and Delivery.” Biomacromolecules22 (2021): 37913799.DOI
  26. [26]T.D. Leitner, E. Mena-Osteritz,, and P. Bäuerle. “Spacial structures induced by sterical hindrance of large substituents: A dendritic macromolecular ”snowflake” molecule.” J. Polym. Sci.60 (2022): 1845–1854.DOI
  27. [27]Khristine Haydukivska, Viktoria Blavatska,, and Jarosaw Paturej. “The size and shape of snowflake-shaped polymers in dilute solution: Analytical and numerical approaches.” Journal of Molecular Liquids392 (2023): 123430.DOI
  28. [28]Fabio Ferri, Maria Greco,, and Mattia Rocco. “On the determination of the average molecular weight, radius of gyration, and mass/length ratio of polydisperse solutions of polymerizing rod-like macromolecular monomers by multi-angle static light scattering.” Macromolecular Symposia 162, no. 1 (2000): 2344.
  29. [29]Detlef-M. Smilgiels, and Ewa Folta-Stogniew. “Molecular weight-gyration radius relation of globular proteins: a comparison of light scattering, small-angle X-ray scattering and structure-based data.” Journal of Applied Crystallography 48, no. 5 (2015): 16041606.
  30. [30]J.A. Aronovitz, and D.R. Nelson. “Universal features of polymer shapes.” Journal de Physique 47, no. 9 (1986): 1445–1456.DOI
  31. [31]J Rudnick, and G Gaspari. “The asphericity of random walks.” Journal of Physics A: Mathematical and General 19, no. 4 (1986): L191–L193.DOI
  32. [32]P. de Gennes 1979. Scaling concepts in polymer physics, Ithaca [u.a.]: Cornell Univ. Pr.DOI
  33. [33]Jaques des Cloizeaux, and Gerard Jannink. 1991. Polymers in Solution: their modelling and structure, Clarendon Press:Oxford.
  34. [34]R Guida, and J Zinn-Justin. “Critical exponents of theN-vector model.” Journal of Physics A: Mathematical and General 31, no. 40 (1998): 8138121.DOI
  35. [35]O. Jagodzinski, E. Eisenriegler,, and K. Kremer. “Universal shape properties of open and closed polymer chains: renormalization group analysis and Monte Carlo experiments.” Journal de Physique I 2, no. 12 (1992): 22432279.DOI
  36. [36]Marvin Bishop, Julian H. R. Clarke, Antonio Rey,, and Juan J. Freire. “The shape of linear and star polymers with and without excluded volume.” The Journal of Chemical Physics 94, no. 5 (1991): 40094011.DOI
  37. [37]H. Yamakawa 1971. Modern Theory of Polymer Solution, New York: Harper and Row.
  38. [38]K. Binder, and W. Paul. “Recent Developments in Monte Carlo Simulations of Lattice Models for Polymer Systems.” Macromolecules 41, no. 13 (2008): 45374550.DOI
  39. [39]Marvin Bishop, and Julian H. R. Clarke. “Brownian dynamics study of the shape of star and linear polymers in different regimes.” The Journal of Chemical Physics 90, no. 11 (1989): 66476651.DOI
  40. [40]Andrzej Sikorski, and Piotr Romiszowski. “Shape of star-branched polymers at various solvent conditions. A computer simulation study.” The Journal of Chemical Physics 109, no. 14 (1998): 61696174.DOI
  41. [41]P. J. Flory. “Statistical thermodynamics of semi-flexible chain molecules.” Proceedings of the Royal Society of London. Series A. Mathematical and Physical Sciences 234, no. 1196 (1956): 6073.DOI
  42. [42]Alberto Montesi, Matteo Pasquali,, and F. C. MacKintosh. “Collapse of a semiflexible polymer in poor solvent.” Physical Review E 69, no. 2.
  43. [43]Victor A. Bloomfield. “Condensation of DNA by multivalent cations: Considerations on mechanism.” Biopolymers 31, no. 13 (1991): 14711481.pubmed.ncbi.nlm.nih.gov/1814499
  44. [44]Takeaki Araki. “Conformational changes of polyelectrolyte chains in solvent mixtures.” Soft Matter 12, no. 28 (2016): 61116119.DOI
  45. [45]Nasir Amiri, Jonathan P. Singer,, and Xin Yong. “Self-Assembly of Crowded Semiflexible Polymers under Dynamic and Deformable Confinement.” Macromolecules 59, no. 8 (2026): 47444752.
  46. [46]XiaoDong Hu, Shawn E. Jenkins, Byung G. Min, Malcolm B. Polk,, and Satish Kumar. “RigidRod Polymers: Synthesis, Processing, Simulation, Structure, and Properties.” Macromolecular Materials and Engineering 288, no. 11 (2003): 823843.
  47. [47]Jun Guan, Zijjing Zhang, and Richard M. Laine. “Synthesis and Characterization of Rigid-Rod Polymers with Silsesquioxanes in the Main Chain.” Macromolecules 55, no. 13 (2022): 54035411.DOI
  48. [48]Slimane Abdous, Mehdi Derradji, Karim Khiari, Oussama Mehelli, Abdelmalek Habes, Tintaoucinne Mhamed El Hadi,, and Abdelwahab Berrouane. “Innovative Reactive Structural Materials: Rigid Polymers for Enhanced Ammunition Performance.” . Propellants, Explosives, Pyrotechnics.
  49. [49]Toshitaka Miyabe, Hiroki Iida, Motonori Banno, Tomoko Yamaguchi,, and Eiji Yashima. “Synthesis and Visualization of a Core Cross-Linked Star Polymer Carrying Optically Active Rigid-Rod Helical Polyisocyanide Arms and Its Chiral Recognition Ability.” Macromolecules 44, no. 21 (2011): 86878692.DOI
  50. [50]Moriya Kikuchi, Atsushi Takahara,, and Seigou Kawaguchi. “Dimensional Characterizations from Rod Stars to Brushes of Polymers with a Low Degree of Polymerization.” Macromolecules 50, no. 1 (2016): 324331.DOI
  51. [51]Hirokazu Hasegawa, Yuuya Nagata, Ken Terao,, and Michinori Suginome. “Synthesis and Solution Properties of a Rigid Helical Star Polymer: Three-Arm Star Poly(quinoline-2, 3-diyl).” Macromolecules 50, no. 19 (2017): 74917497.
  52. [52]R. Freudenberger, W. Claussen, A.-D. Schlüter,, and H. Wallmeier. “Functionalized rod-like polymers: one-dimensional rigid matrices.” Polymer 35, no. 21 (1994): 44964501.DOI
  53. [53]Dominik Konkolewicz, Angus Gray-Weale,, and Sébastien Perrier. “The structure of randomly branched polymers synthesized by living radical methods.” Polymer Chemistry 1, no. 7 (2010): 1067.DOI
  54. [54]Kurt Kremer, and Kurt Binder. “Monte Carlo simulation of lattice models for macromolecules.” Computer Physics Reports 7, no. 6 (1988): 259310.DOI
  55. [55]K. Binder 2012. Monte Carlo Simulations in Polymer Science, pp. 461474. Elsevier.DOI
  56. [56]Wyatt Hooper, and Alexander R. Klotz. “Trapping in self-avoiding walks with nearest-neighbor attraction.” Physical Review E 102, no. 3.
  57. [57]Viktoria Blavatska, Jaroslav Ilnytskyi,, and Erkki Lähderanta. “Mapping self-avoiding walk on obstacle-ridden lattice onto chelation of heavy metal ions: Monte Carlo study.” Physical Review E 110, no. 6.
  58. [58]Viktoria Blavatska, Jaroslav Ilnytskyi,, and Erkki Lähderanta. “Coagulation-flocculation process on a lattice: Monte Carlo simulations.” Journal of Physics A: Mathematical and Theoretical 57, no. 31 (2024): 315002.DOI
  59. [59]A Rechnitzer, and E J Janse van Rensburg. “Generalized atmospheric Rosenbluth methods (GARM).” Journal of Physics A: Mathematical and Theoretical 41, no. 44 (2008): 442002.DOI
  60. [60]E J Janse van Rensburg, and A Rechnitzer. “Atmospheres of polygons and knotted polygons.” Journal of Physics A: Mathematical and Theoretical 41, no. 10 (2008): 105002.DOI
  61. [61]A L Owczarek, and T Prellberg. “Scaling of the atmosphere of self-avoiding walks.” Journal of Physics A: Mathematical and Theoretical 41, no. 37 (2008): 375004.DOI
  62. [62]J Alvarez, M Gara, E J Janse van Rensburg,, and A Rechnitzer. “Atmospheric collapse in self-avoiding walks: a numerical study using GARM.” Journal of Statistical Mechanics: Theory and Experiment 2009, no. 12 (2009): P12005.
  63. [63]Marshall N. Rosenbluth, and Arianna W. Rosenbluth. “Monte Carlo Calculation of the Average Extension of Molecular Chains.” The Journal of Chemical Physics 23, no. 2 (1955): 356–359.DOI
  64. [64]Peter Grassberger. “Pruned-enriched Rosenbluth method: Simulations of polymers of chain length up to 1 000 000.” Phys. Rev. E56 (1997): 3682–3693.DOI
  65. [65]Ikbel Benalava, Gilberto Alves, João Lopes,, and Luís R. Silva. “A Review of Natural Polysaccharides: Sources, Characteristics, Properties, Food, and Pharmaceutical Applications.” International Journal of Molecular Sciences 25, no. 2 (2024): 1322.
  66. [66]Mohammad Salehi, and Ali Rashidinejad. “Multifaceted roles of plant-derived bioactive polysaccharides: A review of their biological functions, delivery, bioavailability, and applications within the food and pharmaceutical sectors.” International Journal of Biological Macromolecules290 (2025): 138855.DOI
  67. [67]Irina Negut, and Bogdan Bita. “Polymeric Micellar SystemsA Special Emphasis on Smart Drug Delivery.” Pharmaceutics 15, no. 3 (2023): 976.DOI
  68. [68]Vikrant Abbot, Poonam Sharma, Abha Sharma, Deepak Sharma,, and Tamanna Sharma. “Micelles transforming modern medicine: a comprehensive review on the role of micellar physicochemical properties in improving the flavonoid bioavailability and neomycin efficacy.” Discover Chemistry 2, no. 1.
  69. [69]Mittal Gunjan Lokesh, and Amit Kumar Tiwari. “A concise review on polymer brushes and its interaction with surfactants: An approach towards smart materials.” Journal of Molecular Liquids 407 (2024): 125168.DOI
  70. [70]J. Serra 1997. “Cube, Cube-octahedron or Rhombodo-D shape Descriptions.” In Advances in Visual Form Analysis, , edited by Carlo Arcelli, Luigi P. Cordella, and Gabriella Sanniti di Baja, 502–519. World Scientific.
  71. [71]Thomas Vogel, Michael Bachmann,, and Wolfhard Janke. “Freezing and collapse of flexible polymers on regular lattices in three dimensions.” Physical Review E 76, no. 6.DOI
  72. [72]S. Campbell, and E. J. Janse van Rensburg. “Numerical estimates of square lattice star vertex exponents.” Physical Review E 103, no. 5.DOI
  73. [73]M. K. Wilkinson, D. S. Gaunt, J. E. G. Lipson,, and S. G. Whittington. “Lattice models of branched polymers: statistics of uniform stars.” Journal of Physics A: Mathematical and General 19, no. 5 (1986): 789796.

Paper details

vlnc:W_99cfgusb
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents