Synthesis and molecular dynamics simulation of hyperbranched poly(amine-ester)/neodymium nanocomposites

1 Introduction

Hyperbranched polymers are synthesized through a one-step method to produce highly branched polymers. Hyperbranched polymers have emerged during the recent years as a promising class of versatile materials for a broad range of novel nanoscale applications. Flory first put forward the concept of hyperbranched polymers in 1952 and pointed out that hyperbranched polymers are synthesized from the structure of monomers with AB₂. However, the real development of hyperbranched polymers began in the 1980s, when the polymers research group of DuPont Company successfully synthesized the polymerization of acrylic monomer at room temperature. In recent years, hyperbranched polymers have aroused more and more attention due to their unique structure and excellent performance [1–3]. Hyperbranched polymers do not need careful separation and purification in the process of preparation and have some unique physical and chemical properties, such as the ability of unique molecular internal nanosized holes to chelate ions, adsorb small molecules, or act as catalytic activity sites of small molecules. Hyperbranched polymers are difficult to crystallize, so dissolution performance greatly improves with highly branched structure. Compared with linear molecules with the same molecular weight, hyperbranched polymers have low molten viscosity. Hence, it becomes more and more important in the application of paint, with polymers blending with its low viscosity and many modified end groups [4, 5]. At the same time, hyperbranched polymers also have a big potential in the development and application of functional materials, such as important application in drug carrier, gene carrier, and macromolecular building “block” [6]. Some scholars have begun to explore hyperbranched functional polymers and have obtained initial achievement [5–11]; they have become an important direction in polymer science research in the 21st century [12]. Bao et al. [13] from Jilin University used hyperbranched molecules as a template to prepare Au nanoparticles. However, combination with the aspect of molecular dynamics (MD) simulation was seldom reported.

On the basis of our previous work [14, 15], pentaerythritol was used as nuclear in the direct synthesis of hyperbranched poly(amine-ester) (HPAE) through the chemical reduction method, then HPAE was used as a template and stabilizer to compound with Nd to prepare HPAE/Nd nanocomposites. The microstructure of the nanocomposites was investigated by transmission electron microscopy (TEM), and Virtual Materiale software was applied to HPAE/Nd nanocomposites for MD simulation to research the stability of the composite system and its mechanism in the canonical system (constant NVT) from the perspective of molecular structure and energy change.

2 Materials and methods

All reagents used were of analytic grade and were available commercially. Diethanolamine, methyl acrylate, pentaerythritol, p-toluene sulfonic acid, hydroxy benzoic ether, and acetic anhydride were purchased from Shanghai Chemical Reagents Company. Pyridine, NdCl₃, sodium borohydride, methanol, ethanol, and KOH were purchased from Acros Organics (USA). Deionized water was used for preparation of all aqueous solutions.

N,N-2-hydroxyethyl-3-amino methyl propionate monomer was synthesized according to Lu et al. [16]. Same amounts of diethanolamine and methyl acrylate were mixed. Methanol and hydroxy benzoic ether were added into a four-neck flask and acted as solvent and polymerization inhibitor, respectively, then reacted at room temperature under an atmosphere of nitrogen for 7 h. Vacuum distillation of methanol and production of yellowish oily liquid, i.e., N,N-2-hydroxyethyl-3-amino methyl propionate monomer, were carried out.

The mole ratio of pentaerythritol and N,N-2-hydroxyethyl-3-amino methyl propionate monomer was 1:12, added into a round-bottomed flask with methanol, 0.1 g p-toluenesulfonic acid, as catalyst. This solution was left to react for 3 h with vigorous stirring at 120°C. Decompression to remove methanol and produce a light yellow sticky liquid, i.e., G2 HPAE, was performed. The above steps were repeated to obtain G3 and G4 HPAE according to stoichiometric proportion.

G2, G3, and G4 HPAEs were dissolved in deionized water to prepare a 0.012 mol/l solution. These were then mixed with 0.06 mol/l NdCl₃ aqueous solution with vigorous stirring for 1 h and slowly added to a 0.6 mol/l NaBH₄ solution under nitrogen gas. The color of the solution turned dark after addition of the reductant, and the solution was stirred overnight at room temperature.

3 Results and discussion

Element analysis method was applied to research the content of C, H, and N in the synthesized N,N-2-hydroxyethyl-3-amino methyl propionate monomer. Results show that the theoretical value and the experimental value were consistent, as shown in Table 1, which indicates that the synthetic monomer was pure and reached application requirements.

The distribution of molecular weight was measured by Waters244GPC-LC chromatograph, with tetrahydrofuran as solvent. Flow rate was kept at 1.0 ml/min, polystyrene copolymer microspheres were kept at stationary phase, and apertures were 100, 103, 104, and 105, respectively. The determination results are as follows.

The hydrodynamic radius of the spherical molecules was much smaller than that of linear molecules with the same molecular weight, so the polystyrene sample for the GPC measurement molecular weight was lower than the actual value [17]. However, there is no suitable GPC standard sample, and linear polystyrene is still used as a standard sample. Table 2 shows the hyperbranched molecular characteristics of experimental and theoretical data. Using the GPC method for measurement of molecular weight, the experimental molecular weight was much smaller than the theoretical molecular weight; this may be due to the fact that hyperbranched molecule with special molecular structure induces a small hydrodynamic radius effect.

The sample was dissolved in acetic anhydride/pyridine (v/v=1:5) hybrid solutions and reflowed for 1.5 h for acetylation. Then, 20 ml of deionized water was added after cooling, and titration with 0.5 mol/l KOH-ethanol solution was carried out to measure the hydroxy value of hyperbranched polymers.

The different hydroxy values generated for HPAE are shown in Table 3. The theoretical and experimental hydroxy values decrease with an increase in reproductive generation. As we know, the molecular weight of hyperbranched molecules and end hydroxy number increase with reproductive generation, but growth rate for the former is slightly higher than for the latter. Thus, the hydroxy value of hyperbranched molecules slightly decreases with an increase in molecular weight. In addition, the experimental hydroxy value was also higher than the theoretical hydroxy value, as shown in Table 3. This result indicates that the actual molecular weight is less than the theoretical molecular weight, which illustrates the formation of spherical hyperbranched molecules with defect.

Figure 1 shows the FT-IR spectra of the monomer and G2 HPAE. The infrared spectra of the other generations of HPAE are similar to that of G2 HPAE with same functional groups, so these are not shown here. The wave number is 3391.41 cm^-1 near the wide bond from the spectrum of monomer belonging to the end hydroxy characteristic absorption peak; alkyl characteristic absorption peaks for ester bond and tertiary amine bond appeared at 2951.59 and 2835.55 cm^-1, respectively. Carbonyl characteristic absorption peak for the ester bond appeared at 1732.90 cm^-1, and wave number appeared at 1037 cm^-1 near the absorption peak corresponding to the hydroxy of O-H bending vibration peak. From the spectra, we also can see that methyl acrylate has completely reacted due to the absorption peak being very weak around 1600 cm^-1, and no double bond emerged in the product. End hydroxy characteristic absorption peak weakened to 3368 cm^-1 in the spectrum of G2 HPAE; this is because it is easy to form an intramolecular hydrogen bond with the increase in end hydroxy, thus reducing the frequency of the absorption peak. The peak appearing at 1616 cm^-1 is not the characteristic absorption peak of the double bond because in the reactant monomer and pentaerythritol, no double bond existed. We can thus conclude that the characteristic absorption peak should be benzene from the reaction of p-toluene sulfonic acid. The FT-IR spectrum analysis shows that the synthesis of monomer and hyperbranched composites is basically consistent with the theoretical structure.

Figure 1

FT-IR spectra of (A) N,N-2-hydroxyethyl-3-amino methyl propionate monomer and (B) G2 HPAE.

From Figure 2B, we can see that the peak at 3.7 ppm corresponded to the methylene that connected with the ester group. Multiple peaks that appeared at 3.6 ppm were from methylene that connected with hydroxy. The unimodal peak at 3.42 ppm was from hydroxy and multiple peaks between 2.4 and 2.8 ppm should be from methine that connected with carbonyl and N atoms; there was no characteristic peak of hydrogen bonding. Figure 2A is similar to Figure 2B, but all kinds of peak location shift and the crack of the peaks become more apparent due to the factors of the hydrogen bond and electronic effect enhancement. The peak that appeared at around 3.25 ppm is from the hydrogen bond that formed by N atoms and H atoms from the hydrogen bond.

Figure 2

¹H NMR spectra of (A) G2 HPAE and (B) G3 HPAE.

Figure 3 is the TG spectra of G2 and G3 HPAE. The figure shows that there was no significant weight loss before 180°C; this was the process of dehydration by end carboxyl and its adjacent amide bond. Product weight loss was above 70% from 200°C to 300°C in the broken process of polymeric backbone. On one hand, this result illustrates that hyperbranched macromolecules do not contain small molecules; on the other hand, the product has excellent thermal stability performance. G2 HPAE has a high purity, and G3 shows better thermal stability performance.

Figure 3

Thermogravimetric analysis HPAE.

4 MD simulation of HPAE/Nd nanocomposites

4.1 Structural analysis

We first studied the molecular structure of G2–G4 HPAE because hyperbranched composites have unique internal nanosized holes that can chelate ions and adsorb small molecules. Figure 4 shows that the branched chains of G2 HPAE molecules are stretched and molecular internal cavity is open. G3 HPAE molecular branched chains and end groups have slight bending and obvious cavity in molecules, but sealing ability is poor. The surface of G4 HPAE molecules has great density of end hydroxy and forms closed intramolecular cavity to provide space for combining inorganic molecules or ions. Therefore, G4 HPAE molecules are more suitable as a template for synthesis of nanocomposites. Nd particles are then added to the generation with the same number of end groups. The topological structure shown in Figure 4 indicates that Nd particles can hardly compound with G2 HPAE. In addition, it cannot form a stable nanocomposite system. Almost all Nd particles are packaged in the closed cavity of G4 HPAE, and a stable nanocomposites system is formed, which is consistent with previous results.

Figure 4

Topological structure of G2, G3, and G4 HPAE.

4.2 Energy analysis

From Tables 4 and 5, we can see that the potential energy of HPAE/Nd nanocomposites is lower than that of its corresponding HPAEs. That is, the energy of the system drops and tends to be stable after addition of the rare earth. Figure 5 shows that the variation tendency of angle, torsion, inversion, and VDW is the same with that of potential energy. It also indicates that they contribute largely to changing the total energy.

Table 4

Energies of G2–G4 HPAE systems (KJ/mol).

	EBond	EAngle	ETor	EInver	EVDW	ECoul	PE
G2 HPAE/Nd	492.61	803.98	261.15	20.32	755.43	183.39	2514.97
G2 HPAE	526.15	1121.46	344.19	17.81	804.36	162.15	2982.89
G3 HPAE/Nd	894.42	1748.03	653.46	18.81	1653.69	533.72	5497.71
G3 HPAE	1037.36	2657.92	776.43	42.96	1820.81	592.80	6919.59
G4 HPAE/Nd	2304.26	3711.92	1296.39	89.74	3377.20	1052.12	11,830.9
G4 HPAE	2389.74	5263.13	1515.2	152.31	3724.72	864.03	13,905.63

Table 5

Energy change in G2–G4 HPAE systems (KJ/mol).

Material	≥EBond	≥EAngle	≥ETor	≥EInver	≥EVDW	≥ECoul	≥PE
G2 HPAE	33.54	317.47	83.04	-2.51	48.93	-21.24	467.92
G3 HPAE	142.94	909.89	122.98	24.15	167.12	59.09	1421.89
G4 HPAE	85.48	1551.21	218.81	62.57	347.52	-188.09	2071.10

Figure 5

Relationship between HPAE/Nd energy curves and generation.

The ≥E_Angle of the composite system increases as a whole with the increase in HPAE generation, as shown in Figure 6B. The bond angle bending decreases after insertion of Nd particles as a result of a large deviation from the θ₀ from the bond angle bending formula because the θ is reduced when θ<θ₀ and enlarged when θ>θ₀.

Figure 6

Amplification pictures of HPAE/Nd energy curves with the generation relationship.

(1)EAngle=12C(cosθ-cosθ0)2(θ=bond angle, C=constant) (1)

The increase in bond angle will inevitably lead to a change in the other angle of the bond. This is one of the reasons that HPAE molecular branched chains are stretched from the structure analysis. Bond angle bending can be reduced because Nd particles in the coordination process have a small influence on bonding from space steric hindrance, and bond angle can be changed so that energy reaches its lowest point.

As shown in Figure 6D, the VDW energy of HPAE/Nd is lower than that of corresponding HPAEs, and the VDW energy is reduced greatly with the increase in HPAE generation. We know that the reduction in intramolecular VDW energy is the interatomic distance of the changes. This result indicates that the distance between the atoms tends to approach equilibrium state. The insertion of Nd particles can decrease VDW interaction. Along with the increase in HPAE generation, VDW energy is reduced with the increase in Nd particles.

Figure 6C shows that the bond torsional energy of HPAE/Nd is lower than that of its corresponding HPAEs; the reduction changes with the increase in HPAE generation. Structural analysis indicates that G2 and G3 HPAEs have a very opened structure and are stretched, so the ability to accommodate Nd particles is poor and part of the Nd particles may escape from the composite system. Bond torsional energy is reduced slightly after comparison. G4 HPAE approximate spherical structure, end groups, and branched unit are incurved to form a closed intramolecular nanocavity, which is very suitable for cladding metallic atoms to form a stable compound system. As a result, bond torsional energy is reduced, which is consistent with the structural analysis.

The ≥E_Inversion of the composite system increases as a whole with the increase in HPAE generation, as shown in Figure 6C. We can see from the formula of bond angular external bending energy that Nd particles reduce bond angular external bending energy. Ligands have multiple positions to coordinate with Nd atoms when Nd particles are inserted. As a result, the whole energy will be reduced.

(2)EInversion=12Kω(cosω-cosω0)2(Kω=force constant, ω=angle between IL and IJK) (2)

The dynamics optimization for G2–G4 HPAE and composite system is shown in Figure 6B. The ≥E_Bond decrement of the G3 composite system is largest with the increase in generation. Structure analysis shows that the G2 and G3 HPAEs express a more opened structure, resulting in key expansion in a large space and an arbitrary expansion change in order to reduce energy. However, G4 HPAE approximate spherical structure, end groups, and branched chains are incurved to form a closed intramolecular cavity, so bond expansion is limited and bond expansion energy fluctuation is smaller than in G3 HPAE. Because the ligand gives electrons to share with Nd atoms, the electron density of the ligand decreases, which could lead to ligand covalent bond shortening, which is consistent with the radius of gyration after insertion of Nd particles. It can be concluded that the reason for the reduction in total bond expansion energy lies in the shortening of the HPAE molecules of the covalent bond. At the same time, Nd atoms do not exist significantly in steric hindrance. In order to choose the appropriate position and produce the lowest energy system, the bond expansion can be adjusted freely.

Figure 6D indicates the charges of system redistribution as a whole after insertion of Nd particles. This kind of distribution can happen easily; bond formation, rotation, and distortion can all cause electronic redistribution in the moment. As a result, the generation of electrostatic energy fluctuates after insertion of Nd atoms.

4.3 Gyration radius analysis

The radius of gyration is very important to explore the compact degree of HPAE/Nd nanocomposites structure, and the specific calculated data are shown in Table 6.

Table 6

Gyration radius of G2–G4 HPAE and the corresponding composite system.

Material	Rg (nm)	Material	Rg (nm)	≥Rg
G2 HPAE	1.0142	G2 HPAE/Nd	0.9873	-0.0269
G3 HPAE	1.3447	G3 HPAE/Nd	1.3344	-0.0103
G4 HPAE	2.0469	G4 HPAE/Nd	1.9889	-0.058

The negative value indicates that the gyration radius of HPAE/Nd nanocomposites is smaller than that of HPAEs. From Table 6, it can be seen that the gyration radius of HPAE/Nd nanocomposites is smaller than that of the corresponding HPAEs. In addition, the gyration radius of the G4 HPAE/Nd nanocomposites is largely reduced (Figure 7). This phenomenon indicates low generation of nanocomposites with a loose structural expression, branched chains, and end groups of HPAE occurring in bending along with the increase in generation, leading to the decrease in gyration radius and the structure of nanocomposites becomes more closed and stable. We can conclude that G4 HPAE molecules have excellent cavity that can better clad Nd particles to form a stable HPAE/Nd nanocomposites system. This result is consistent with the experimental results and dynamic energy analysis.

Figure 7

Relationship between ≥Rg and generations.

4.4 Morphology analysis of HPAE/Nd nanocomposites

Figure 8 shows the TEM of G2, G3, and G4 HPAE/Nd nanocomposites. Compared with G2 and G3, G4 HPAE/Nd nanocomposites have the best dispersion and uniform particle size. The size of the particles is also small, as shown in Figure 8A and B, but reunion phenomenon is quite obvious compared with that in Figure 8C. This result demonstrates that the G4 HPAE/Nd compound effect is excellent; the size of the nanocomposites in Figure 8C was 4.8±1.0 nm. Through molecular simulation, it was found that the radius of gyration of G4 HPAE/Nd is 1.989 nm (diameter is 3.98 nm); thus, it can be concluded that part of the Nd atoms adsorbed in the HPAE peripheral did not enter the dendrimer internal cavity. As a result, the HPAE acts as template and stabilizing agent in preparing nanocomposites. The above comparison indicates that G4 HPAE is more suitable to form nanocomposites with Nd particles. Of course, the composite effect is different from the dendrimer PAMAM as template, based on our previous work [14]. This is caused mainly by the fact that the branched degree of hyperbranched molecule is <100%. However, further research on improving the degree of branching is needed.

Figure 8

(A) TEM of G2 HPAE/Nd, (B) G3 HPAE/Nd, and (C) G4 HPAE/Nd.

5 Conclusions

Irregularly branched dendrimers have bad branched degree than regular dendrimers do but still have the typical characteristics of dendrimer, such as proper size, internal cavity, and end groups; all of these factors make it possible for these to be used as a template for synthesis of metallic nanocomposites. Here, HPAE can be used as a template and sodium borohydride can be used as a reductant to obtain Nd particles with small size and homogeneous distribution. Generation of hyperbranched molecules is used for the decision for use as a template or as a stabilizer in the reaction process. Internal cavity can synthesize stable nanoparticles when their chemical structure is approximately spherical. Thus, we can use this method to determine the size of the reservation of rare earth metal nanoparticles. In this work, we applied MD simulation method to the lanthanides and irregular dendritic composite system to obtain the scientific conclusion, which is based on the specific experiment. The experimental result shows that irregular dendrimer also can act as a template to synthesize metallic nanoparticles. At the same time, microscopic image of molecular motion through MD simulation can provide a scientific explanation for the macroscopic experiment phenomena.

Scheme 1

Synthesis of N,N-2-hydroxyethyl-3-amino methyl propionate monomers.

Scheme 2

Schematic of synthesis of G4 HPAE molecules.