Next Article in Journal
A 3D-Printed Sole Design Bioinspired by Cat Paw Pad and Triply Periodic Minimal Surface for Improving Paratrooper Landing Protection
Previous Article in Journal
Research Progress on Emerging Polysaccharide Materials Applied in Tissue Engineering
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

A Simple Stochastic Reaction Model for Heterogeneous Polymerizations

State Key Laboratory for Chemistry and Molecular Engineering of Medicinal Resources, Key Laboratory for Chemistry and Molecular Engineering of Medicinal Resources (Ministry of Education of China), Collaborative Innovation Center for Guangxi Ethnic Medicine, School of Chemistry and Pharmaceutical Sciences, Guangxi Normal University, Guilin 541004, China
*
Authors to whom correspondence should be addressed.
Polymers 2022, 14(16), 3269; https://doi.org/10.3390/polym14163269
Submission received: 14 July 2022 / Revised: 3 August 2022 / Accepted: 5 August 2022 / Published: 11 August 2022
(This article belongs to the Section Polymer Chemistry)

Abstract

:
The stochastic reaction model (SRM) treats polymerization as a pure probability‐based issue, which is widely applied to simulate various polymerization processes. However, in many studies, active centers were assumed to react with the same probability, which cannot reflect the heterogeneous reaction microenvironment in heterogeneous polymerizations. Recently, we have proposed a simple SRM, in which the reaction probability of an active center is directly determined by the local reaction microenvironment. In this paper, we compared this simple SRM with other SRMs by examining living polymerizations with randomly dispersed and spatially localized initiators. The results confirmed that the reaction microenvironment plays an important role in heterogeneous polymerizations. This simple SRM provides a good choice to simulate various polymerizations.

Graphical Abstract

1. Introduction

Heterogeneous polymerizations are polymerization processes with two or more phases during or after polymerization, as well as polymerizations at the surface or interface [1,2]. Typical heterogeneous polymerizations, such as suspension, emulsion, dispersion, precipitation polymerizations, etc., are widely applied for the synthesis of high-performance materials such as paints, coating, and many others. Other systems include polymerization-induced phase separation (PISA) [3,4,5], surface-initiated polymerization (SIP) [6,7], and polymerizations in living cells [8,9]. It is challenging to characterize the polymerization kinetics since the distributions of species are inhomogeneous in space and variant with time.
Today, computer simulation is an important tool in the study of polymerization, which can provide both microscopic structural and time-dependent information about the system. Due to the broad coverage of the time and length scales of polymerization systems, coarse-grained models are usually utilized to describe the polymerization kinetics, such as coarse-grained molecular dynamics simulation (CGMD) [10,11,12,13,14,15,16,17,18], Brownian dynamic simulation [19], the dynamic lattice liquid model (DLL) [20,21,22], the bond fluctuation model (BFM) [23,24,25,26,27], the reaction-event-driven kinetic Monte Carlo (kMC) model [28,29], and the dissipative particle dynamics (DPD) method [30,31,32,33,34,35], etc. Various strategies, based on the distance, probability, or reactive force field, have been applied to manipulate the polymerization [15,36].
We are interested in the stochastic reaction model (SRM) [37,38], which treats polymerization as a pure probability-based issue. Due to the simple idea, various SRMs have been proposed and applied in CG simulations to study polymerizations in homogeneous as well as heterogeneous systems [10,12,19,25,27,31,32,33,34,35]. However, as pointed out by Arraez et al., most SRMs are based on simplified constant reaction probabilities, which fail to capture the kinetics of heterogeneous polymerizations since the concentration variations are inevitable [29].
To better understand the dilemma of the SRM, we would like to briefly introduce how to simulate a living polymerization with different versions of the SRM. As known, the polymerization rate Rp of a living polymerization is
R p = d [ M ] d t = k p [ M ] [ I ] 0 ,
where kp is the reaction rate constant, and [M] and [I]0 are the concentrations of free monomers and active centers, respectively. A successful SRM should yield the correct polymerization kinetics.
In the SRM, usually, an active center reacts with a randomly selected free monomer [27,33] or a closed one [39,40] in the given reaction radius with a probability Pr (Figure 1a). The number of consumed monomers in one reaction step is I0Pr. Pr is an important characteristic for controlling the kinetics of the process. According to Equation (1), Pr should be proportional to the concentration of free monomers. In a very popular version of the SRM (Version I), Pr is simply calculated as [27]
P r = P 0 [ M ] / [ M ] 0
where [M] and [M]0 are the instantaneous and initial concentration of free monomers, respectively, and the initial addition probability P0 is a constant. Equation (2) implies that the system is well-mixed, i.e., all species are homogeneously distributed in space so that all active centers react with the same Pr in Version I.
In Version I, the reaction probability Pr decreases with the decreasing of the monomer concentration [M] and should be calculated at every reaction step. When the monomer conversion is low, the variance of [M] can be ignored [36]. There is no need to calculate Pr at every reaction step. In some simulations (Version II), Pr was further simplified as a constant, which was independent of [M] [31,33,41]. It is a type of polymerization with a constant reaction rate Rp instead of a constant reaction rate constant kp.
Version I differs from Version II, as the influence of the monomer concentration has been considered in Version I. However, in both versions, all active centers react with the same Pr. The divergence of the local reaction microenvironment is not considered. To simulate heterogeneous polymerizations, a step forward is required to consider the reaction microenvironment of each active center [29].
In the DPDChem software [42], another strategy has been applied. An active center sequentially reacts with nearby monomers according to the distance with a fixed probability of P0 until a bond is created or there are no more unchecked monomers (Figure 1b). We can calculate the probability of the active center to create a bond, which is
P r = 1 ( 1 P 0 ) m ,
where m is the number of monomers in the reaction radius. When P0 is small, Pr can be expressed as mP0. In this version (Version III), each active center reacts with its Pr, and the effect of the reaction microenvironment is well considered. DPDChem has been widely applied to simulate various homogeneous and heterogeneous polymerizations [35,43,44,45,46,47].
In lattice Monte Carlo (MC) simulations, such as BFM and DLL [20,21,22,23,24,25,26,48,49], another version (Version IV) has been applied (Figure 1c). For an active center, one of the neighboring sites is randomly selected. If this site is occupied by a free monomer, a reaction is tried with a fixed probability P0. Attention must be paid that Version IV (select a site and try to react if it is a monomer) is essentially different from Versions I and II (select a monomer and try to react). The reaction microenvironment is considered in Version IV since the probability to find a free monomer at a given site is determined by the local concentration of free monomers. On the contrary, it is not considered in Versions I and II.
Recently, we proposed a simple SRM (Version V) for heterogeneous polymerizations [38]. When an active center tries to react with a randomly selected free monomer in the reaction radius, the reaction probability Pr is calculated according to the number of local free monomers m as mP0 (Figure 1a). This idea follows the work by Akkermans et al. [10], in which the probability was suggested to be m/mmax, where mmax was a preset value and always larger than m. To our understanding, however, it was the average number of free monomers <m> that was applied in [10], which approached a constant value for not short chains (Equation (4) in [10]). We applied version V as well as Version I for comparison to study surface-initiated polymerization (SIP) [38]. Essentially, SIP is a heterogeneous polymerization that is easy to be ignored because free monomers are gradually distributed in the system [38]. The results suggested that the reaction microenvironment plays an important role in SIP.
To our understanding, the influence of the reaction microenvironment has not attracted enough attention in many simulations with SRM. For example, inappropriate versions have been applied in the simulations of SIP and PISA [27,32,33]. Sometimes, the description of the SRM is too simple to judge whether the heterogeneous reaction microenvironment has been considered or not. To provoke enough attention to the reaction microenvironment and promote the application of the SRM in heterogeneous polymerizations, we compared the new SRM with Versions I, II, and III in this paper. Version IV is not examined here as it can be only applied in lattice simulations. Theoretically, it should obtain the same results as Version V. The chance that a randomly selected neighbor site is occupied by a free monomer is proportional to the number of local free monomers. The algorithms of different versions of the SRM are introduced in Section 2. The results of living polymerizations with randomly dispersed and spatially localized initiators are shown in Section 3. A brief conclusion is given in Section 4.

2. Models and Simulation Methods

2.1. Lattice Monte Carlo Simulation

The simulation was carried out in a simple cubic lattice with a volume V = L × L × L with the periodic boundary condition in three directions. The Larson-type bond fluctuation model has been adopted [38,50,51,52] since the corresponding theoretical polymerization kinetics can be easily obtained to guide the simulation, as shown later. In this model, each monomer (or initiator) occupies one lattice, and the permitted bond length is 1 or   2 . During relaxation, a monomer randomly selects one of 18 nearest and next nearest neighbor sites and tries to move. In any elementary movement, a bond intersection is forbidden. Meanwhile, each lattice can be occupied only once; thus, the excluded volume effect is well considered in this simulation. The stimulation time is measured in units of MC steps (MCs), which is defined as all monomers are tried to move once on average.

2.2. Implementation of Stochastic Reaction Model

While implementing SRM, one question is when to model a reaction. In the work by Genzer [27], a decision was made by a comparison of a generated random number with a probability of choosing motion over reaction. While studying the influence of diffusion with BFM, Lu and Ding applied a probability to reduce the movement of a monomer to a vacancy [24]. Here, a characteristic delay time, or reaction interval time, τ was applied to separate two successive reaction steps [33,38,53]. The effect of diffusion can be simply tuned by adjusting the value of τ.
When an active center tries to react with a free monomer in a given reaction radius Rcut, the value of Rcut influences the polymerization kinetics. A bigger Rcut can speed up the polymerization, as there are more reaction candidates. In simulations such as MD, however, instabilities might be provoked due to the strong force after the creation of a new bond [15]. A proper reaction radius is needed to achieve the balance between the polymerization speed and the simulation stability. A larger reaction interval time is helpful for the dissipation of the new bond energy and the simulation stability. Here, Rcut was set as   2 , the same as the longest bond length in the Larson-type bond fluctuation model. The maximum reaction candidates mmax around an active center is 18.
Section 1 has introduced the basic idea of different SRMs. In this simulation, the procedures of different versions are as follows: In Version III, an active center sequentially reacted with nearby monomers with a reaction probability Pr,seq = P0, where P0 is a constant and defined as the reaction probability between one active center and one free monomer. In Version V, an active center tried to react with a randomly selected free monomer in the reaction radius with a reaction probability Pr,loc = mP0, where m is the number of monomers in the reaction radius of the active center. Here, P0 is restricted to be no larger than 1/mmax, i.e., Pr,loc is no larger than 1; thus, the effect of the concentration of free monomers could be correctly considered.
In the literature [27,36], the reaction probability of Version I was calculated according to the instantaneous concentration of free monomers [M] and scaled by the initial concentration [M]0 (Equation (2)). It must be pointed out that [M]0 in Equation (2) should be the maximum concentration of monomers in a bulk polymerization. Otherwise, it will fail to compare the polymerization kinetics of systems with different initial monomer concentrations. In this study, the reaction probability of Version I, Pr,av, was calculated according to the average number of free monomers around an active center
P r , a v = ( m max 1 ) [ M ] t P 0 .
Here, mmax − 1 instead of mmax is used as one nearby position is occupied by the previous monomer of the same chain. For Version II, the reaction probability Pr,const was a constant, which was calculated according to the initial concentration of free monomers as
P r , c o n s t = ( m max 1 ) [ M ] 0 P 0 .
Once the reaction probability (Pr,av, Pr,const, P0, and Pr,loc in Versions I, II, III, and V, respectively) between the reaction pair is determined, a random number is generated. If the randomly generated number is no larger than the reaction probability, the reaction is accepted, and the free monomer turns out to be the active center for future reaction. The information such as the chain length will be updated. The illustrations of one reaction cycle of different versions are shown in Figure 1.

2.3. Polymerization Kinetics

The polymerization kinetics of a living polymerization has been deduced to guide the simulation [38]. Since the time between two reaction steps is τ, the concentration change of free monomers during polymerization can be calculated as
d [ M ] d t = [ I ] 0 P r / τ .
According to Equation (4), the above equation can be written as
d [ M ] d t = ( m max 1 ) [ I ] 0 [ M ] t P 0 / τ .
The theoretical monomer conversion of a homogeneous living polymerization is
C = 1 [ M ] t [ M ] 0 = 1 exp ( ( m max 1 ) [ I ] 0 P 0 t / τ ) .

3. Results and Discussion

3.1. Homogeneous Polymerization

Firstly, different versions of the SRM were applied to examine a living polymerization system with randomly dispersed free monomers and initiators. The length of the cubic lattice was L = 60, the initial monomer concentration was [M]0 = 0.4 monomer per lattice, the number of initiators was I0 = 1000, the reaction interval time τ was 10 MCs, the simulation time was 106 MCs, and the reaction probability between one active center and one monomer was P0 = 0.001. In the reaction step, an initiator is randomly selected, then it tries to react with a nearby free monomer with a certain reaction probability. If the reaction is accepted, the initiator transfers the reactivity to the free monomer and itself becomes the first monomer of the polymer chain. Meanwhile, the newly reacted free monomer becomes an active center for future reactions. In this way, linear polymer chains are obtained. The results were averaged over 20 independent runs.
Living polymerization characters were obtained with all versions, and the results are almost identical (Figure 2). For example, the linear relationship between the number-average molecular weight Mn and the monomer conversion C was observed (Figure 2a). The dispersity (Đ = Mw/Mn) slightly increased with C at the start of the simulation and further decreased to reach a plateau (Figure 2b). Figure 2c shows the molecular weight distributions during the polymerization, which can be described by the Poisson distribution
P ( N ) = M n N N ! exp ( M n ) .
We studied the number of unreacted initiators I during polymerization. It was predicted that I can be calculated as [33]
I = I 0 ( 1 P r ) t / τ
when Pr is a constant. In the early stage of the polymerization, the monomer conversion is very low, e.g., about 3% when t = 4000 MCs. The influence of monomer concentration on the reaction probability can be ignored. Figure 2d shows that ln(I/I0) decays linearly, consistent with the theoretical prediction very well.
The differences between different versions can be observed when the monomer conversion C and reaction rate Rp are concerned. As expected, the monomer conversion C of Version II increased linearly with time and quickly reached a plateau due to the constant reaction probability (Figure 3a). The results of other versions followed the polymerization kinetics of a living polymerization as predicted by Equation (8). However, a deviation could be clearly observed for Version I in the late stage of the simulation. The reason is that the well-mixed assumption adopted in Version I is no longer held. When the concentration of free monomers [M] is very low (mmax[M] < 1), a monomer cannot be found in the reaction radius of an active center. Both the chance to find a free monomer and the reaction probability Pr,av are proportional to [M]. The effect of monomer concentration was considered twice, and a slow polymerization is observed (also shown in Figure 3b). Due to the same reason, the polymerization rate of Version II was no more a constant at the late stage of polymerization (Figure 3b). Both Versions I and II should better study polymerization systems with a concentration of free monomers higher than 1/mmax [38]. There is no such limitation for Versions III and V, and the results are consistent with the theoretical predictions even when the concentration of free monomers is very low (Figure 3a,b).

3.2. Heterogeneous Polymerization

We examined a heterogeneous polymerization system with 1000 initiators spatially localized in a 10 × 10 × 10 cube at the beginning of the simulation (as illustrated by the inset in Figure 4a), which corresponds to an experiment with powered initiators. The initiators are allowed to diffuse apart during the polymerization. We can expect that the polymerization should be influenced by the localization of the initiators since the inner and outer initiators react differently.
When the same parameters as those in Figure 2 were applied (P0 = 0.001 and τ = 10 MCs), the obtained results are similar to those of the homogeneous systems (Figure S1). According to the snapshots (Figure S2), the localized initiators became quickly dispersed after several reaction steps due to the slow polymerization and small size of the initiators. The effect of the localization of the initiators was not obvious.
When the polymerization is very fast (P0 = 0.01), the influence of the localization of the initiators can be observed. All versions suggest that the polymerization of the heterogeneous system (Figure 4a) is slower than that of the corresponding homogeneous system (Figure 4b). It is easy to explain. Due to the large P0, the initiators reacted with free monomers before they diffused apart from each other. The outer initiators have more chance to react with free monomers, the inner ones are trapped. Due to the trapping effect, the initiators should react slower, and a broader molecular weight distribution should be observed. Such a slower conversion of initiators was observed with Versions III and V (Figure 4c). Meanwhile, Versions I and II obtained a conversion similar to the theoretical prediction. Figure 4e suggests that the molecular weight distributions of the heterogeneous polymerization system are very broad and cannot be described by the Poisson distribution. For comparison, the results of the homogeneous polymerization system still follow the theoretical predictions even when the polymerization is very fast (Figure 4d,f). It confirms that the location of the initiators influences the polymerization. Firstly, the access of free monomers to an inner initiator (active center) is limited as the area around the initiator (active center) is occupied by other initiators (active centers) and reacted monomers. Secondly, the delivery of free monomers might be blocked by outer initiators and active centers due to the large P0. As a result, inner initiators (active centers) might be in a starve state, i.e., they fail to find a free monomer in the reaction range. It is the reason that a broad distribution was still obtained with Versions I and II. Therefore, the distribution revealed by Versions III and V is more reliable since the reaction microenvironment is well considered.

3.3. A Further Comparison between Versions III and V

In Versions III and V, the influence of the reaction microenvironment was considered as the reaction probability of an active center is determined by the local reaction microenvironment. As shown, both versions can be applied to study both homogeneous and heterogeneous polymerizations, and the obtained results are almost identical. Strictly speaking, the consideration of reaction microenvironment is different in Versions III and V. It is directly considered in Version V as the reaction probability Pr,loc = mP0, while indirectly considered in Version III according to Equation (3). One question is whether this difference can be ignored or not.
The difference between Versions III and V can be estimated by the reaction probability. According to Equation (3), the reaction probability of Version III, Pr,seq, can be expressed as mP0 when P0 and m are small. Figure 5a shows Pr,seq as a function of P0 while fixing m = mmax. Pr,seq gradually deviates from the theoretical reaction probability Pth = mP0 with increasing P0. We calculated the relative difference between the reaction probability
Δ P / P t h = ( P t h P r , s e q ) / P t h .
The relative difference is 8.1% when P0 = 0.01 and m = 18. For a given P0, the relative difference decreases with decreasing m. As the concentration of free monomers [M] is 0.4 in this study, the relative difference is only 3.1% at the start of the stimulation and decreases with polymerization (decreasing m). Thus, no significant difference between Versions III and V can be observed in Figure 4. When P0 is extremely large (P0 = 0.05 and [M] = 0.4), the relative difference reaches 14.2%. Figure 5b shows that the polymerization of Version III is slower than that of Version V. However, the difference between the molecular weight distributions is still ignorable (Figure S3).
According to the communication with Berezkin, the reaction probability in the simulation should be sufficiently small as the reaction probability in the real world is far smaller than that in simulations. Otherwise, the overall kinetics becomes unrealistic because it will be diffusion controlled. Typically, the reaction probability is 0.001, or even smaller in simulations. With such a small value, polymerization can be freely studied by either Version III or V.

4. Conclusions

Due to the simple idea, various stochastic reaction models were proposed to simulate different polymerization processes, which can be divided into two types according to the reaction microenvironment considered or not. In the first type (Versions I and II), all active centers react with the same probability, ignoring the divergence of the reaction microenvironment, in the second type (Versions III, IV, and V), each active center reacts individually according to the local reaction environment, considering the influence of reaction microenvironment.
In this paper, we applied different versions of the SRM to study a homogeneous polymerization system with randomly dispersed initiators and a heterogeneous polymerization system with spatially localized initiators. For the homogeneous polymerization system, all versions obtained typical characteristics of a living polymerization such as the linear increase in Mn with the monomer conversion C and the Poisson molecular weight distribution. For the heterogeneous polymerization system with spatially localized initiators, the expected behaviors, such as the slow conversion of initiators and broader molecular weight distribution, were observed with the second type of the SRM (Versions III and V). Though Versions I and II obtained broad MWDs when polymerization is very fast, the results are no more reliable. Similar conclusions were also obtained from the study of surface-initiated polymerization [38] that the properties such as MWD and dispersity are strongly influenced by the reaction microenvironment.
In a summary, caution must be paid while studying heterogeneous polymerization with a stochastic reaction model. Each active center should react with its own probability, which is determined by the active center’s local reaction microenvironment. As far as the three versions (III, IV, and V) are concerned, we recommend Version V due to its simplicity, popularity for both lattice and off-lattice simulations, and convenience for theoretical analysis.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/polym14163269/s1, Figure S1: comparison of living polymerization systems with spatially localized (left column) and randomly dispersed (right column) initiators when the polymerization is slow; Figure S2: Snapshots of the heterogeneous polymerization system with spatially localized initiators with given simulation times; Figure S3: The molecular weight distributions of the homogeneous polymerization obtained with versions III and V.

Author Contributions

S.J., B.-P.J., and X.-C.S. designed the research and wrote the paper. J.M., J.L., B.Y., and S.L. carried out the simulations and analyzed the data. S.J. interpreted the results. All authors have read and agreed to the published version of the manuscript.

Funding

This work was supported by the “Overseas 100 Talents Program” of Guangxi Higher Education, the Natural Science Foundation of China (No. 21464004); Guangxi Natural Science Foundation of China (Nos. 2014GXNSFAA118038 and 2015GXNSFCB139005).

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

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

Acknowledgments

We thank Hong Liu and Anatoly V. Berezkin for very helpful discussions.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Arshady, R. Suspension, emulsion, and dispersion polymerization: A methodological survey. Colloid Polym. Sci. 1992, 270, 717–732. [Google Scholar] [CrossRef]
  2. Qiu, J.; Charleux, B.; Matyjaszewski, K. Controlled/living radical polymerization in aqueous media: Homogeneous and heterogeneous systems. Prog. Polym. Sci. 2001, 26, 2083–2134. [Google Scholar] [CrossRef]
  3. Charleux, B.; Delaittre, G.; Rieger, J.; D’Agosto, F. Polymerization-induced self-assembly: From soluble macromolecules to block copolymer nano-objects in one step. Macromolecules 2012, 45, 6753–6765. [Google Scholar] [CrossRef]
  4. Derry, M.J.; Fielding, L.A.; Armes, S.P. Polymerization-induced self-assembly of block copolymer nanoparticles via RAFT non-aqueous dispersion polymerization. Prog. Polym. Sci. 2016, 52, 1–18. [Google Scholar] [CrossRef] [Green Version]
  5. Cornel, E.J.; Jiang, J.; Chen, S.; Du, J. Principles and characteristics of polymerization-induced self-assembly with various polymerization techniques. CCS Chem. 2021, 3, 2104–2125. [Google Scholar] [CrossRef]
  6. Chen, W.-L.; Cordero, R.; Tran, H.; Ober, C.K. 50th anniversary perspective: Polymer brushes: Novel surfaces for future materials. Macromolecules 2017, 50, 4089–4113. [Google Scholar] [CrossRef]
  7. Turgman-Cohen, S.; Genzer, J. Simultaneous bulk- and surface-initiated controlled radical polymerization from planar substrates. J. Am. Chem. Soc. 2011, 133, 17567–17569. [Google Scholar] [CrossRef]
  8. Dai, Y.; Li, T.; Zhang, Z.; Tan, Y.; Pan, S.; Zhang, L.; Xu, H. Oxidative polymerization in living cells. J. Am. Chem. Soc. 2021, 143, 10709–10717. [Google Scholar] [CrossRef]
  9. Geng, J.; Li, W.; Zhang, Y.; Thottappillil, N.; Clavadetscher, J.; Lilienkampf, A.; Bradley, M. Radical polymerization inside living cells. Nat. Chem. 2019, 11, 578–586. [Google Scholar] [CrossRef] [Green Version]
  10. Akkermans, R.L.C.; Toxvaerd, S.R.; Briels, W.J. Molecular dynamics of polymer growth. J. Chem. Phys. 1998, 109, 2929–2940. [Google Scholar] [CrossRef]
  11. Deng, B.; Shi, Y. A reactive coarse-grained model for polydisperse polymers. Polymer 2016, 98, 88–99. [Google Scholar] [CrossRef] [Green Version]
  12. de Buyl, P.; Nies, E. A parallel algorithm for step- and chain-growth polymerization in molecular dynamics. J. Chem. Phys. 2015, 142, 134102. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  13. Biedermann, M.; Diddens, D.; Heuer, A. rs@md: Introducing reactive steps at the molecular dynamics simulation level. J. Chem. Theory Comput. 2021, 17, 1074–1085. [Google Scholar] [CrossRef] [PubMed]
  14. Zhang, Z.; DuBay, K.H. The sequence of a step-growth copolymer can be influenced by its own persistence length. J. Phys. Chem. B 2021, 125, 3426–3437. [Google Scholar] [CrossRef] [PubMed]
  15. Farah, K.; Karimi-Varzaneh, H.A.; Müller-Plathe, F.; Böhm, M.C. Reactive molecular dynamics with material-specific coarse-grained potentials: Growth of polystyrene chains from styrene monomers. J. Phys. Chem. B 2010, 114, 13656–13666. [Google Scholar] [CrossRef]
  16. Zhang, Z.; Krajniak, J.; Samaey, G.; Nies, E. A Parallel Multiscale Simulation Framework for Complex Polymerization: AB2-Type Monomer Hyperbranched Polymerization as an Example. Adv. Theory Simul. 2019, 2, 1800102. [Google Scholar] [CrossRef]
  17. Gao, J. An efficient method of generating dense polymer model melts by computer simulation. J. Chem. Phys. 1995, 102, 1074–1077. [Google Scholar] [CrossRef]
  18. Krajniak, J.; Zhang, Z.; Pandiyan, S.; Nies, E.; Samaey, G. Coarse-grained molecular dynamics simulations of polymerization with forward and backward reactions. J. Comput. Chem. 2018, 39, 1764–1778. [Google Scholar] [CrossRef]
  19. Xu, J.; Xue, Y.-H.; Cui, F.-C.; Liu, H.; Lu, Z.-Y. Simultaneous polymer chain growth with the coexistence of bulk and surface initiators: Insight from computer simulations. Phys. Chem. Chem. Phys. 2018, 20, 22576–22584. [Google Scholar] [CrossRef]
  20. Polanowski, P.; Hałagan, K.; Pietrasik, J.; Jeszka, J.K.; Matyjaszewski, K. Growth of polymer brushes by “grafting from” via ATRP–Monte Carlo simulations. Polymer 2017, 130, 267–279. [Google Scholar] [CrossRef]
  21. Halagan, K.; Banaszak, M.; Jung, J.; Polanowski, P.; Sikorski, A. Polymerization and structure of opposing polymer brushes studied by computer simulations. Polymers 2021, 13, 4294. [Google Scholar] [CrossRef] [PubMed]
  22. Polanowski, P.; Jeszka, J.K.; Matyjaszewski, K. Polymer brush relaxation during and after polymerization–Monte Carlo simulation study. Polymer 2019, 173, 190–196. [Google Scholar] [CrossRef]
  23. Wang, L.; He, X. Reaction kinetics of nonideal hyperbranched polymerizations: Influences of chain rigidity and reaction reversibility. J. Polym. Sci. Part A Polym. Chem. 2012, 50, 2705–2714. [Google Scholar] [CrossRef]
  24. Lu, W.; Ding, J. Dynamic Monte Carlo simulation of chain growth polymerization and its concentration effect. Sci. China Ser. B 2005, 48, 459–465. [Google Scholar] [CrossRef]
  25. Lu, W.; Ding, J. Dynamic Monte Carlo simulation of polymerization of amphiphilic macromers in a selective solvent and associated chemical gelation. Macromolecules 2006, 39, 7433–7440. [Google Scholar] [CrossRef]
  26. Wang, L.; He, X.; Chen, Y. Diffusion-limited hyperbranched polymers with substitution effect. J. Chem. Phys. 2011, 134, 104901. [Google Scholar] [CrossRef] [PubMed]
  27. Genzer, J. In Silico Polymerization:  Computer simulation of controlled radical polymerization in bulk and on flat Surfaces. Macromolecules 2006, 39, 7157–7169. [Google Scholar] [CrossRef]
  28. Arraez, F.J.; Van Steenberge, P.H.; D’hooge, D.R. The competition of termination and shielding to evaluate the success of surface-initiated reversible deactivation radical polymerization. Polymers 2020, 12, 1409. [Google Scholar] [CrossRef] [PubMed]
  29. Arraez, F.J.; Van Steenberge, P.H.M.; Sobieski, J.; Matyjaszewski, K.; D’hooge, D.R. Conformational variations for surface-initiated reversible deactivation radical polymerization: From flat to curved nanoparticle surfaces. Macromolecules 2021, 54, 8270–8288. [Google Scholar] [CrossRef]
  30. Lísal, M.; Brennan, J.K.; Smith, W.R. Mesoscale simulation of polymer reaction equilibrium: Combining dissipative particle dynamics with reaction ensemble Monte Carlo. I. Polydispersed polymer systems. J. Chem. Phys. 2006, 125, 164905. [Google Scholar] [CrossRef]
  31. Yan, Y.-D.; Xue, Y.-H.; Zhao, H.-Y.; Liu, H.; Lu, Z.-Y.; Gu, F.-L. Insight into the polymerization-induced self-assembly via a realistic computer simulation strategy. Macromolecules 2019, 52, 6169–6180. [Google Scholar] [CrossRef]
  32. Lu, H.; Song, W.-Y.; Zou, Y.-Y.; Xu, W.-S.; Yan, Y.-D.; Liu, H.; Ma, L.-J. Kinetics and morphologies in polymerization-induced cooperative assembly: A computer simulation investigation. Polym. Int. 2021, 71, 359–365. [Google Scholar] [CrossRef]
  33. Liu, H.; Li, M.; Lu, Z.-Y.; Zhang, Z.-G.; Sun, C.-C. Influence of surface-initiated polymerization rate and initiator density on the properties of polymer brushes. Macromolecules 2009, 42, 2863–2872. [Google Scholar] [CrossRef]
  34. Liu, H.; Qian, H.-J.; Zhao, Y.; Lu, Z.-Y. Dissipative particle dynamics simulation study on the binary mixture phase separation coupled with polymerization. J. Chem. Phys. 2007, 127, 144903. [Google Scholar] [CrossRef] [PubMed]
  35. Berezkin, A.V.; Kudryavtsev, Y.V. Simulation of end-coupling reactions at a polymer-polymer interface: The mechanism of interfacial roughness development. Macromolecules 2011, 44, 112–121. [Google Scholar] [CrossRef]
  36. Liu, H.; Zhu, Y.L.; Lu, Z.Y.; Müller-Plathe, F. A kinetic chain growth algorithm in coarse-grained simulations. J. Comput. Chem. 2016, 37, 2634–2646. [Google Scholar] [CrossRef] [PubMed]
  37. Zhu, Y.L.; Liu, H.; Li, Z.W.; Qian, H.J.; Milano, G.; Lu, Z.Y. GALAMOST: GPU-accelerated large-scale molecular simulation toolkit. J. Comput. Chem. 2013, 34, 2197–2211. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  38. Yang, B.; Liu, S.; Ma, J.; Yang, Y.; Li, J.; Jiang, B.-P.; Ji, S.; Shen, X.-C. A Monte Carlo simulation of surface-initiated polymerization: Heterogeneous reaction environment. Macromolecules 2022, 55, 1970–1980. [Google Scholar] [CrossRef]
  39. Wang, J.; Li, J.; Wang, Y.; Li, Z.; Zhang, J. Polymerization-induced self-assembly of comb-like amphiphilic copolymers into onion-like vesicles. Macromolecules 2021, 54, 7448–7459. [Google Scholar] [CrossRef]
  40. Zhang, Z.; Wang, L.; Wang, Z.; He, X.; Chen, Y.; Müller-Plathe, F.; Böhm, M.C. A coarse-grained molecular dynamics-reactive Monte Carlo approach to simulate hyperbranched polycondensation. RSC Adv. 2014, 4, 56625–56636. [Google Scholar] [CrossRef]
  41. Turgman-Cohen, S.; Genzer, J. Computer simulation of controlled radical polymerization: Effect of chain confinement due to initiator grafting density and solvent quality in “grafting from” method. Macromolecules 2010, 43, 9567–9577. [Google Scholar] [CrossRef]
  42. DPDChem Software. Available online: https://www.researchgate.net/project/DPDChem-Software (accessed on 1 August 2022).
  43. Berezkin, A.V.; Kudryavtsev, Y.V. End-coupling reactions in incompatible polymer blends: From droplets to complex micelles through interfacial instability. Macromolecules 2013, 46, 5080–5089. [Google Scholar] [CrossRef]
  44. Guseva, D.V.; Kudryavtsev, Y.V.; Berezkin, A.V. Simulation of heterogeneous end-coupling reactions in polydisperse polymer blends. J. Chem. Phys. 2011, 135, 112. [Google Scholar] [CrossRef]
  45. Gavrilov, A.A.; Chertovich, A.V. Copolymerization of partly incompatible monomers: An insight from computer simulations. Macromolecules 2017, 50, 4677–4685. [Google Scholar] [CrossRef] [Green Version]
  46. Gavrilov, A.A.; Chertovich, A.V. Self-assembly in thin films during copolymerization on patterned surfaces. Macromolecules 2013, 46, 4684–4690. [Google Scholar] [CrossRef]
  47. Kozhunova, E.Y.; Gavrilov, A.A.; Zaremski, M.Y.; Chertovich, A.V. Copolymerization on selective substrates: Experimental test and computer simulations. Langmuir 2017, 33, 3548–3555. [Google Scholar] [CrossRef]
  48. Shao, J.; Tang, W.; Xia, R.; Feng, X.; Chen, P.; Qian, J.; Song, C. Monte Carlo simulation on kinetics of batch and semi-batch free radical polymerization. Macromol. Res. 2015, 23, 1042–1050. [Google Scholar] [CrossRef] [Green Version]
  49. Hałagan, K.; Banaszak, M.; Jung, J.; Polanowski, P.; Sikorski, A. Dynamics of opposing polymer brushes: A computer simulation study. Polymers 2021, 13, 2758. [Google Scholar] [CrossRef] [PubMed]
  50. Larson, R.; Scriven, L.; Davis, H. Monte Carlo simulation of model amphiphile-oil–water systems. J. Chem. Phys. 1985, 83, 2411–2420. [Google Scholar] [CrossRef]
  51. Larson, R. Monte Carlo lattice simulation of amphiphilic systems in two and three dimensions. J. Chem. Phys. 1988, 89, 1642–1650. [Google Scholar] [CrossRef]
  52. Ji, S.; Ding, J. Nonequilibrium Monte Carlo simulation of lattice block copolymer chains subject to oscillatory shear flow. J. Chem. Phys. 2005, 122, 164901. [Google Scholar] [CrossRef] [PubMed]
  53. Müller-Plathe, F. A simple nonequilibrium molecular dynamics method for calculating the thermal conductivity. J. Chem. Phys. 1997, 106, 6082–6085. [Google Scholar] [CrossRef]
Figure 1. Illustration of one reactive cycle in different stochastic reaction models. (a) An active center reacts with a free monomer in the reaction radius with probability Pr. The methods to calculate Pr in Versions I, II, and V are shown. (b) In Version III, an active center sequentially reacts with nearby free monomers. (c) In Version IV, one neighboring direction of an active center is selected in a lattice simulation; if it is occupied by a monomer, a reaction is tried.
Figure 1. Illustration of one reactive cycle in different stochastic reaction models. (a) An active center reacts with a free monomer in the reaction radius with probability Pr. The methods to calculate Pr in Versions I, II, and V are shown. (b) In Version III, an active center sequentially reacts with nearby free monomers. (c) In Version IV, one neighboring direction of an active center is selected in a lattice simulation; if it is occupied by a monomer, a reaction is tried.
Polymers 14 03269 g001
Figure 2. A living polymerization with randomly dispersed free monomers and initiators studied by different versions of stochastic reaction model. (a) Number-average molecular weight Mn and (b) dispersity (Đ) as a function of the monomer conversion C. (c) Molecular weight distribution with Mn = 30 and 60 respectively. The dotted lines are drawn according to a Poisson distribution. (d) The ratio of unreacted initiators during polymerization. The reaction interval time τ = 10 MCs, and the reaction probability between one active center and one monomer P0 = 0.001.
Figure 2. A living polymerization with randomly dispersed free monomers and initiators studied by different versions of stochastic reaction model. (a) Number-average molecular weight Mn and (b) dispersity (Đ) as a function of the monomer conversion C. (c) Molecular weight distribution with Mn = 30 and 60 respectively. The dotted lines are drawn according to a Poisson distribution. (d) The ratio of unreacted initiators during polymerization. The reaction interval time τ = 10 MCs, and the reaction probability between one active center and one monomer P0 = 0.001.
Polymers 14 03269 g002
Figure 3. (a) Monomer conversion C as a function of simulation time. (b) Polymerization rate Rp as a function of monomer conversion C. The parameters are the same as those in Figure 2.
Figure 3. (a) Monomer conversion C as a function of simulation time. (b) Polymerization rate Rp as a function of monomer conversion C. The parameters are the same as those in Figure 2.
Polymers 14 03269 g003
Figure 4. (Left column) Heterogeneous polymerization with spatially localized initiators; (right column) homogeneous polymerization with randomly dispersed initiators. (a,b) Monomer conversion C. (c,d) the ratio of unreacted initiators I/I0. (e,f) molecular weight distribution with Mn = 60. The dotted lines are Poisson distributions. The inlets are the 2D illustrations of the spatially localized and randomly distributed initiators, respectively. The cyan cycles: free monomers; hollow cycles: initiators. The number of initiators I0 = 1000, the reaction interval time τ = 10 MCs, and the reaction probability between one active center and one monomer P0 = 0.01.
Figure 4. (Left column) Heterogeneous polymerization with spatially localized initiators; (right column) homogeneous polymerization with randomly dispersed initiators. (a,b) Monomer conversion C. (c,d) the ratio of unreacted initiators I/I0. (e,f) molecular weight distribution with Mn = 60. The dotted lines are Poisson distributions. The inlets are the 2D illustrations of the spatially localized and randomly distributed initiators, respectively. The cyan cycles: free monomers; hollow cycles: initiators. The number of initiators I0 = 1000, the reaction interval time τ = 10 MCs, and the reaction probability between one active center and one monomer P0 = 0.01.
Polymers 14 03269 g004
Figure 5. (a) The reaction probability Pr,seq, Pth and the relative difference as a function of P0. The number of monomers around each active center is m = 18. (b) The conversion of free monomers during the polymerization of a system with randomly distributed initiators. The reaction interval time τ = 10 MCs, and the reaction probability between one active center and one monomer P0 = 0.05. Other parameters are the same as those in Figure 2.
Figure 5. (a) The reaction probability Pr,seq, Pth and the relative difference as a function of P0. The number of monomers around each active center is m = 18. (b) The conversion of free monomers during the polymerization of a system with randomly distributed initiators. The reaction interval time τ = 10 MCs, and the reaction probability between one active center and one monomer P0 = 0.05. Other parameters are the same as those in Figure 2.
Polymers 14 03269 g005
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Share and Cite

MDPI and ACS Style

Ma, J.; Li, J.; Yang, B.; Liu, S.; Jiang, B.-P.; Ji, S.; Shen, X.-C. A Simple Stochastic Reaction Model for Heterogeneous Polymerizations. Polymers 2022, 14, 3269. https://doi.org/10.3390/polym14163269

AMA Style

Ma J, Li J, Yang B, Liu S, Jiang B-P, Ji S, Shen X-C. A Simple Stochastic Reaction Model for Heterogeneous Polymerizations. Polymers. 2022; 14(16):3269. https://doi.org/10.3390/polym14163269

Chicago/Turabian Style

Ma, Jiashu, Jiahao Li, Bingbing Yang, Siwen Liu, Bang-Ping Jiang, Shichen Ji, and Xing-Can Shen. 2022. "A Simple Stochastic Reaction Model for Heterogeneous Polymerizations" Polymers 14, no. 16: 3269. https://doi.org/10.3390/polym14163269

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop