Investigations on the Multicentered Active Sites of Ziegler–Natta Polypropylene Catalysts through Multi-Models of the MgCl2 Support Authors: Shuxuan Tang, Haitao Liu, Zhuang Zhuang, Xiaoxia Cai, Xiaofan Zhang, Qilong Zhou Published: 2026 ABSTRACT The Ziegler–Natta (ZN) catalyst has unshakable importance to the polypropylene (PP) industry. Until now, theoretical research on the distribution of active centers in ZN catalysts is quite rare. It remains a challenge to understand the relationship between internal donor (ID) and molecular weight distribution (MWD) of the PP material. In this work, a series of MgCl2 models were designed, and both diether- (IDs 1 and 2) and diester-type (IDs 3 and 4) IDs were selected for theoretical research. The ID adsorption energies of MgCl2–ID and MgCl2–ID-TiCl4 clusters were obtained by DFT calculations. The results revealed that the structural flexibility and steric hindrance of the ID molecule and the Lewis acidity and stereoeffect of MgCl2 play a critical role in the distribution of multicentered active sites in the ZN catalyst. Moreover, the results of ID adsorption energies showed a surprising relationship with the experimental MWD. Our work provides a systematic study on the ID adsorption of the ZN catalyst and a unique perspective in understanding the relationship between the ID structure and MWD of PP. Introduction Polypropylene (PP) is one of the most important synthetic resin materials. − Nowadays, more than 90% of industrial PP is produced using a Ziegler–Natta (ZN) catalyst, revealing the unshakable importance of the ZN catalyst to the PP industry. − The ZN catalyst is composed of a magnesium-containing support, TiCl4, an electron donor, and a cocatalyst. The magnesium-containing support, such as MgCl2, provides adsorption sites for TiCl4. , During propylene polymerization, TiCl4 would be activated by the cocatalyst and form the active sites. , The electron donor includes two categories: internal electron donor (ID) and external electron donor (ED). , The ID would be added during the catalyst preparation, while the ED would be added during polymerization in conjunction with the cocatalyst such as triethyl aluminum (TEA). − As Lewis bases, both the ID and ED have the ability to adsorb on the surface of the support and influence the Ti active sites. , In this case, the ID can strongly affect the performances of the ZN catalyst such as catalytic activity, stereospecificity, and molecular weight distribution (MWD) of the PP product. ,, Over the past 70 years, the development of electron donors has significantly promoted the comprehensive performance of the ZN catalyst. , The phthalates/alkoxysilanes have been applied as electron donors in the fourth-generation ZN catalyst. The diethers/alkoxysilanes and dioldiesters/alkoxysilanes have been used as electron donors in the nonphthalate fifth-generation ZN catalyst. , As a heterogeneous catalyst with multicentered active sites, direct exploration on the mechanism and microstructure of the ZN catalyst is quite difficult. Until now, spectroscopic techniques such as nuclear magnetic resonance, − electron paramagnetic resonance, , X-ray absorption/emission spectroscopy, , and other techniques such as inelastic neutron scattering technique have been applied in the research studies of the ZN catalyst. Moreover, beyond experimental research methods, DFT calculations have also proven to be a powerful tool in the fundamental research studies of the ZN catalyst. ,, Theoretical research studies on the MgCl2 support, − Ti adsorption sites (i.e., the polymerization active centers), − polymerization reaction, , and the structure–property relationship between the ID and catalytic performance of the ZN catalyst − have already been carried out through DFT calculations. However, most of the previous research studies, which explained the influences of ID on the performances of the ZN catalyst through one or fewer specific ID-support models (such as (110)-facet Corradini model , ), have actually simplified the multicentered active sites of the ZN catalyst to single active sites. As a result, explorations on the mechanism associated with multicentered active sites of the ZN catalyst become difficult. Recently, Taniike et al. reported the nonempirical structure determination on the ZN catalyst with ID including diethyl phthalate and 9,9-bis­(methoxymethyl)­fluorene. Their research reveals that the copresence of TiCl4 and ID leads to the preferential formation of specific surface motifs, and the multiplicity of the ID adsorption mode has a direct consequence on the structural diversity of active sites. In fact, the diversity and distribution of active sites in the ZN catalyst play a critical role in the macroproperties of the PP material, such as MWD. ,, And the MWD of PP strongly affects its mechanical property and processing method. ,, Nevertheless, until now, explanation on the relationship between the ID and macroproperties of PP remains a challenge. This is due to the multicentered active site nature of the ZN catalyst, which hinders the consideration of all possible microstructures in ZN catalysts during DFT model design. Taking into account these difficulties, designing a series of model systems that are appropriate to describe the possible adsorption modes for different IDs and relatively small to decrease the computational cost could be an effective research approach. Meanwhile, exploration on the influences of ID on the distribution of multicentered active sites of ZN catalysts is quite necessary. To gain further insight into the distribution of active centers in the ZN catalyst, herein, a systematic study on the adsorption of different IDs on the MgCl2 support was carried out. The ID compounds studied in this work were selected from an industrial ZN catalyst including the diether IDs 1 and 2, the phthalate ID 3, and the dioldiester ID 4 (Scheme ), , some of which have been investigated by experimental characterization methods. A series of MgCl2 crystallite corner models were designed and applied in DFT calculations in conjunction with (104)- and (110)-facet models. The different adsorption modes of ID were also considered during theoretical calculation, and a detailed result of adsorption energies for ID molecules in MgCl2–ID and MgCl2–ID-TiCl4 clusters was obtained and discussed. Our research reveals that ID has great influences on the distribution of multicentered active sites in ZN catalysts, which finally affects the MWD of PP. Structures of the ID Compounds Experiment and Calculation Section Experiment All experiments were carried out under a nitrogen atmosphere using a glovebox and standard Schlenk techniques. Propylene, hydrogen, MgCl2, and ID 4 were obtained from SINOPEC. Epoxy chloropropane (ECP), tributyl phosphate (TBP), phthalic anhydride (PA), TEA, TiCl4, IDs 1–3, and cyclohexylmethyldimethoxysilane (CHMDMS) were purchased from J&K Scientific without further purification. Toluene and n-hexane were purchased from Sinopharm Chemical Reagent Co. Ltd. and dried with 4 Å molecular sieves. Before being used, TEA and CHMDMS were diluted in n-hexane to 0.5 and 0.1 mol/L, respectively. Synthesis of the Catalyst To a nitrogen-protected five-necked reactor, MgCl2 (4.8 g), toluene (95 mL), ECP (4 mL), and TBP (12.5 mL) were added successively. The reaction mixture was heated up to 60 °C with stirring and maintained at 60 °C for 1.5 h, followed by the addition of PA (1.5 g). The reaction mixture was maintained at 60 °C for 1 h and then cooled down to −25 °C. TiCl4 (56 mL) was introduced slowly to the reaction mixture at −25 °C. Then the mixture was heated to 80 °C, followed by adding 6 mmol of ID. The mixture obtained was stirred for 1 h, and then the supernatant was removed. The residue was washed with toluene and treated two times with the TiCl4/toluene mixed solution (40 mL of TiCl4 and 60 mL of toluene) at 100 °C for 1 h. After sedimentation and filtration of the reaction mixture, the residue was washed four times with n-hexane (60 mL) and dried under vacuum to get the solid catalyst. Propylene Polymerization Propylene polymerization was carried out in a 5 L stainless steel reactor. The reactor was filled with propylene gas before its use. Typical procedure for the bulk polymerization of propylene was as follows: TEA solution (5 mL, 0.5 mol/L in n-hexane), CHMDMS (1 mL, 0.1 mol/L in n-hexane), the solid catalyst (about 9 mg), and 1.2 SL (standard liter) of H2 were added into the reactor. Then 2.3 L of liquid propylene was added. The reactor was heated to 70 °C, and the polymerization was processed at 70 °C for 1 h under a stirring rate of 300 rpm. The PP product (resin powder) was obtained after reducing the temperature and pressure of the reactor. Characterization X-ray diffraction (XRD) patterns were collected on a PANalytical B.V. Empyrean diffractometer operated at 40 kV and 40 mA with the Cu Kα radiation (k = 0.154 nm, step size = 0.013°). All samples were protected by the polyethylene (PE) film in a glovebox before being transferred to the XRD instrument, to prevent the samples from air and moisture. The ID content of catalysts was measured by gas chromatography (GC, ID 1) and liquid chromatography (LC, IDs 2, 3, and 4). The catalysts were dissolved in dilute sulfuric acid, respectively, and then extracted with ethyl acetate. Furthermore, GC separation was performed on Agilent 7890A with the column of HP-INNOWAX (60 m × 0.530 mm × 1.0 μm). The absolute content of ID 1 in the catalyst was calculated by the internal standard method. LC separations were performed on Waters-600E UPLC H-Class with the column of ACQUITY UPLC BEH C18 (50 mm × 2.1 mm, 1.7 μm). The absolute content of IDs 2, 3, and 4 in the corresponding catalyst was calculated by using a standard curve. The ID content results of Cat-1, Cat-2, Cat-3, and Cat-4 are 7.8%, 10.8%, 10.0%, and 11.3% (11.2% for the mesomer isomer and 0.1% for the racemate isomer), respectively. The catalytic activity of the catalyst was calculated as the ratio between the weight of PP and the weight of the corresponding catalyst used in propylene polymerization. The isotactic index (I.I.) of PP was characterized using the following method: 2 g of dried PP sample was extracted with boiling heptane in an extractor for 6 h. Then, the residual substance was dried to constant weight. The ratio of residual polymer weight (g) to 2 (g) was defined as the I.I. The molecular weight and MWD of PP were measured by a Waters Alliance GPCV2000 gel permeation chromatograph (GPC) equipped with a refractive index detector, using three Polymer Laboratory MIXED-B columns and 1,2,4-trichlorobenzene as solvent at 150 °C. The number-average and weight-average molecular weight (M n and M w, respectively) were evaluated with reference to a polystyrene standard calibration. The characterization and polymerization results for the prepared ZN catalysts are listed in Table . 1Characterization and Polymerization ResultscatalystIDcatalytic activity (kgPP/gCat)I.I. (%) M n M w MWD Cat-1 1 48.698.3761003924005.16Cat-2 2 32.797.1639003343005.23Cat-3 3 38.398.5563003820006.78Cat-4 4 40.397.0319002829008.87aMWD = M w/M n. Characterization and Polymerization Results MWD = M w/M n. Calculation Details All DFT calculations were carried out using the Gaussian16 program. All geometry optimizations were carried out at the PBE0-D3/6-311G­(d,p) level in combination with Becke–Johnson (BJ) damping correlation. − The obtained stationary points were characterized by frequency calculations to confirm that the structures have been properly optimized (no imaginary frequencies). To avoid the errors associated with low frequency, Grimme’s quasi-RRHO model was applied in Gibbs free energy calculations using the Shermo-2.6 program. , All of the MgCl2 models used in this work were cut from α-MgCl2 crystals (Figure S9 in Supporting Information), with Mg atoms fixed while other atoms unfixed during structure optimization. The Mg···Mg distance was 3.6363 Å. The ID adsorption energy ΔG(ID) was calculated according to the following equations: ΔG(ID)=G(Mg−ID)−G(ID)−G(Mg) ΔG(ID)=G(Mg−ID−Ti)−G(ID)−G(Mg−Ti) where G(Mg–ID) is the Gibbs free energy of the MgCl2 model adsorbed by the ID molecule (i.e., MgCl2–ID clusters). G(ID) and G(Mg) are the Gibbs free energies of the ID molecule and MgCl2 model, respectively. G­(Mg–ID–Ti) is the Gibbs free energy of the MgCl2 model adsorbed by ID and TiCl4 (i.e., MgCl2–ID–TiCl4 clusters). G­(Mg–Ti) is the Gibbs free energy of the MgCl2 model adsorbed by TiCl4. The Ti adsorption energy ΔG(Ti) was calculated according to the following equation: ΔG(Ti)=G(Mg−ID−Ti)−G(Ti)−G(Mg−ID) where G(Mg–ID–Ti) is the Gibbs free energy of the MgCl2 model adsorbed by the ID molecule and TiCl4. G(Ti) is the Gibbs free energy of the discrete TiCl4 molecule, and G(Mg-ID) is the Gibbs free energy of the MgCl2 model adsorbed by the ID molecule. Result and Discussion Propylene Polymerization Considering the influences of ID on the catalytic performances of the ZN catalyst, four representative IDs used in an industrial ZN propylene catalyst were selected in this work (Scheme ). Both 1 and 2 are diether-type ID. The phthalate ID 3 has been widely used in the fourth-generation ZN catalyst. Compound 4 is a dioldibenzoate-type ID, which has four different optical isomers due to the two chiral carbon atoms in its chemical structure, and only one of the mesomer isomers was chosen to be studied in this work (4 is the RS isomer). , The ZN catalysts were prepared using IDs 1–4, respectively, followed by propylene polymerization experiments. The results of propylene polymerization reveal that IDs have great influences on the catalytic activity and stereospecificity (i.e., I.I. of PP) of the ZN catalyst, also including the molecular weight and MWD of PP. PP produced with catalysts containing diether-IDs 1 and 2 exhibits narrower MWD, while PP produced with the catalyst containing dioldibenzoate-ID 4 reveals broader MWD. These results reveal that IDs have great influences on the active center of the ZN catalyst. As Lewis bases, IDs adsorbed on the MgCl2 support during catalyst preparation. ,, Therefore, explorations on the adsorption of ID are quite necessary for the study on the ZN catalyst. Design of MgCl2 Models Due to the multicentered active site feature of the ZN catalyst, the adsorption of ID molecules on the MgCl2 support is quite complicated, indicating that the ID adsorption on the MgCl2 support cannot be sufficiently studied through the classical regular (104)- and (110)-facet models. Moreover, a large amount of crystal defects on the surface of MgCl2 would be produced during the catalyst preparation process. The XRD results of the α-MgCl2 raw material and the catalysts prepared in this work are listed in Figure . Compared with the α-MgCl2 raw material, XRD patterns of the catalysts all exhibit broader diffraction peaks at 12–17°, 27–40°, and 48–53°, confirming the presence of crystal defects and disordered lattices in the MgCl2 support of ZN catalysts, , which has also been observed in the previous related studies. Powder XRD patterns of the α-MgCl2 raw material and the prepared catalysts. The diffraction peaks at 21.3° and 23.6° were due to the PE film to protect the samples from air and moisture. Inspired by the reported studies of ID adsorption on the defective MgCl2 support and small MgCl2 particle models, ,, herein we designed a series of MgCl2 crystallite corner models (Figure ) to provide a platform for DFT calculations. Considering the computational cost and rationality, the size of MgCl2 models was limited to a reasonable level. The surface size of the MgCl2 model should be sufficient for the coordination of ID and TiCl4 and minimized as much as possible for improving computational efficiency. To ensure that the corner models are representative to describe the diversity of the MgCl2 crystallite corner, the corner models were designed according to the different included angles between neighboring Mg atoms. From the α-MgCl2 crystal model (Figure S9 in Supporting Information), one Mg atom was chosen as the vertex of the crystallite corner and named as Mg0. Then the (110)-facet containing Mg0 was cut for the coordination of TiCl4 and treated as one side of the corner. The other side of the corner was chosen from the possible (104)- or (110)-facet containing Mg0 and finally resulted in corner models A, B, D, and E. Situation for the existence of a short (104)-facet at the vertex of the crystallite corner (i.e., Cl atom as the vertex) was also considered, which resulted in models C and F. The corresponding dihedral angle and facet for each kind of corner model are shown in Figure . Models A, B, and C belong to the convex corner of the MgCl2 crystallite, while models D, E, and F belong to the concave corner of the MgCl2 crystallite. The interfacial angles from the corner models are (110)–(104) for models A and D, (110)–(110) for models B and E, and (110)–(104)–(110) for models C and F, respectively. (104)- and (110)-facet models, in conjunction with the MgCl2 crystallite corner models A–F. The blue dotted lines represent the (104)-facet, while the red dotted lines represent the (110)-facet. Models A–F include diversity situation for the crystallite corner and provide a platform for studying the adsorption of ID and TiCl4 on the MgCl2 support. It should be noticed that the actual MgCl2 surface is far more complicated than the theoretical models. And other kinds of models have also been applied in the research studies on the ZN catalyst. ,− ,, Therefore, the development of a novel and rational MgCl2 model would continue to be the research focus of the ZN catalyst. The corner models A–F, in conjunction with (104)- and (110)-facet models, were applied in DFT calculations to provide systematic exploration on the adsorption of ID molecules on the MgCl2 support. All of the MgCl2 models were calculated first to obtain the corresponding Gibbs free energies. Then TiCl4 was introduced to be adsorbed on the (110)-facet part of each model, coordinating with two Cl atoms of the MgCl2 surface. The Ti-adsorbed MgCl2 models were then calculated to obtain their Gibbs free energies. Adsorption of ID The ID molecules in Scheme are Lewis bases featuring two ether bonds or an ester group. The adsorption mode of IDs 1–4 includes three patterns: monodentate adsorption, chelate adsorption, and bridge adsorption (Figure ). , According to the previous studies on the Corradini site and ID adsorption modes on the MgCl2 support, , in this work, only chelate adsorption and bridge adsorption were considered due to its bidentate coordination, which is more favorable in stability compared with monodentate adsorption. For each kind of ID compound, its adsorption on different MgCl2 models in both the chelate mode and bridge mode was considered in DFT calculations. Moreover, the adsorption conformation of the ID molecule was also considered. The MgCl2–ID cluster structures were calculated first to obtain the ID adsorption energies. Then TiCl4 was introduced to MgCl2–ID clusters and adsorbed on the (110)-facet of each model. The adsorption energies of ID and TiCl4 in MgCl2–ID-TiCl4 models were then calculated. The adsorption modes of ID in MgCl2–ID clusters and the corresponding ID adsorption energies, obtained from DFT calculations, were listed in Table . Adsorption modes of diether- and diester-type IDs on the MgCl2 support. 2Models, Adsorption Modes, and Calculated ID Adsorption Energies (kcal·mol–1) of MgCl2–ID Clusters ID adsorption energycluster model adsorption mode coordinating Mg atom 1234104bridge–13.96–20.77–39.21–43.71110–1chelate-a–48.05–43.58–43.87–42.56110–1chelate-b–48.09–43.41–43.01110–2bridge-a–32.86–32.68–45.42–52.61110–2bridge-b–45.50–53.53AbridgeMg0···Mg1–26.07–28.01–38.57–45.10B-1chelate-aMg0–59.14–52.36–56.66–60.65B-1chelate-bMg0–58.98–59.97–60.54B-2bridgeMg0···Mg1–21.15 –29.82 –41.76–48.02CbridgeMg1’···Mg1–32.96–34.39–46.39–52.44D-1bridgeMg0···Mg1–27.87–23.60–42.08–47.33D-2bridgeMg1···Mg2–20.67–25.50–41.81–46.31D-3bridgeMg0···Mg2–27.86E-1bridgeMg0···Mg1–26.71 –19.42 –44.06–53.74E-2bridgeMg1···Mg2–18.69–18.21–44.71–54.26E-3chelate-aMg1–49.03–42.13–47.86–47.41E-3chelate-bMg1–50.16–46.78–47.98F-1bridgeMg1···Mg2–28.47 –16.38 –33.99–49.96F-2bridgeMg1’···Mg1–31.40–23.29–39.43–52.36aCluster models in this table were named according to the name of the MgCl2 model (shown in Figure ), the adsorption modes of ID, and the coordinating Mg atoms.bLetters “a” and “b” following the adsorption mode correspond to the different adsorption conformations of ID in MgCl2–ID clusters.cSee the Mg atom labels in Figure .dThe ID molecule in the optimized structure was monodentate coordinating with the MgCl2 support due to the stereoeffect. Models, Adsorption Modes, and Calculated ID Adsorption Energies (kcal·mol–1) of MgCl2–ID Clusters Cluster models in this table were named according to the name of the MgCl2 model (shown in Figure ), the adsorption modes of ID, and the coordinating Mg atoms. Letters “a” and “b” following the adsorption mode correspond to the different adsorption conformations of ID in MgCl2–ID clusters. See the Mg atom labels in Figure . The ID molecule in the optimized structure was monodentate coordinating with the MgCl2 support due to the stereoeffect. The calculation results of the ID adsorption energy reveal the obvious differences between different MgCl2–ID cluster models. For IDs 1, 2, and 3, its adsorption on the (104)-facet is less stable than that on the (110)-facet with both chelate and bridge modes. For ID 4, its adsorption energy on the (104)-facet is close to that on the (110)-facet with the chelate mode and less stable than that on the (110)-facet with the bridge mode. Moreover, on the (110)-facet, chelate adsorption for 1 and 2 is more stable than bridge adsorption. In contrast, the adsorption energies for chelate and bridge adsorption of 3 on the (110)-facet are close to each other. For ID 4, its bridge adsorption on the (110)-facet is more stable than chelate adsorption. These results were probably due to the different molecular structures of ID, especially the structural flexibility between the two coordinating atoms and groups. Fewer chemical bonds between the two coordinating O atoms in 1, 2, and 3 increased the difficulty for molecular conformation adjustment and hindered the formation of bridge adsorption. Moreover, formation of a six-membered ring in the (110)-chelate adsorption model of 1 and 2 also increased the stability of MgCl2–ID clusters (see Tables S7 and S8 in Supporting Information). Interestingly, among MgCl2 corner models A–F, the B-1 model features the highest ID adsorption energies (i.e., the most negative adsorption energies) for IDs 1–4. Among the convex corner models (models A, B, and C), the ID adsorption energies of the A model are lower (i.e., less negative than other adsorption energies) than that of other models. The differences of ID adsorption energies among models A, B, and C were probably due to the different Lewis acidity of coordinating Mg atoms. The Lewis acidity of a three-coordinated Mg atom is higher than that of four- or five-coordinated Mg atoms, which rationalizes the higher adsorption energies of ID in the B-1 model. Among the concave corner models (models D, E, and F), the E-3 model exhibits the most stable adsorption for IDs 1, 2, and 3, while the E-2 model shows the highest adsorption energy for ID 4. Because of the concave structure for models D, E, and F, the steric hindrance of ID has great influences on its adsorption energies. Moreover, in some of the MgCl2–ID cluster models, the ID molecules in the optimized structures were monodentate coordinating with the MgCl2 support, due to the steric effect (see Table ). Distribution of Multicentered Active Sites and Their Relationship with MWD To gain further insight into the active centers of the ZN catalyst, TiCl4 was introduced to the MgCl2–ID clusters and adsorbed on the (110)-facet part of each model. The adsorption energies of ID and TiCl4 in MgCl2–ID-TiCl4 clusters were calculated, and the calculation results are listed in Table and Table S5 in Supporting Information, respectively. 3Models, Adsorption Modes, and Calculated ID Adsorption Energies (kcal·mol–1) of MgCl2–ID-TiCl4 Clusters ID adsorption energycluster model adsorption mode coordinating Mg atom 1234110–1chelate-a–47.66–42.73–46.54–43.51110–1chelate-b–50.55–43.22–45.91110–2bridge-a–36.65–35.75–47.51–54.54110–2bridge-b–43.56–51.05AbridgeMg0···Mg1–30.65–24.05–39.33–50.27B-1chelate-aMg0–55.32–48.59–53.62–53.80B-1chelate-bMg0–54.63–54.98–55.55B-2bridgeMg0···Mg1–32.11 –21.36 –37.10–42.75CbridgeMg1’···Mg1–34.48–31.88–45.55–56.13D-1bridgeMg0···Mg1–28.12–24.20–42.14–47.82D-2bridgeMg1···Mg2–30.45–26.64–44.66–50.71D-3bridgeMg0···Mg2–27.70E–1bridgeMg0···Mg1–27.97 –20.66 –46.38–55.31E–2bridgeMg1···Mg2–19.83–18.88–46.27–56.13E–3chelate-aMg1–50.73–47.17–48.04–47.07E–3chelate-bMg1–47.54–48.03–42.15F-1bridgeMg1···Mg2–31.23 –20.68–40.22–47.81F-2bridgeMg1’···Mg1–32.30–23.64–41.18–52.65aCluster models in this table were named the same as those in Table .bLetters “a” and “b” following the adsorption mode correspond to the different adsorption conformations of ID in MgCl2–ID-TiCl4 clusters.cSee the Mg atom labels in Figure .dThe ID molecule in the optimized structure was monodentate coordinating with the MgCl2 support due to the stereoeffect. Models, Adsorption Modes, and Calculated ID Adsorption Energies (kcal·mol–1) of MgCl2–ID-TiCl4 Clusters Cluster models in this table were named the same as those in Table . Letters “a” and “b” following the adsorption mode correspond to the different adsorption conformations of ID in MgCl2–ID-TiCl4 clusters. See the Mg atom labels in Figure . The ID molecule in the optimized structure was monodentate coordinating with the MgCl2 support due to the stereoeffect. The calculation results of ID adsorption energies for MgCl2–ID-TiCl4 clusters (Table ) show similar tendency as that for MgCl2–ID clusters (Table ). The B-1 model features the highest ID adsorption energies among all MgCl2 models for IDs 1–3. For ID 4, the ID adsorption energy of the B-1 model is also close to the highest ID adsorption energy (E-2 and C model). These results indicate the importance and significance for the consideration of the MgCl2 crystallite corner in the exploration of ID adsorption. The crystallite corner of MgCl2 particles is strongly related to the crystal defect and crystal size of MgCl2, which could be affected by the preparation procedure and condition for the ZN catalyst. In this case, the multicentered active sites of the ZN catalyst are influenced by both catalyst synthesis (i.e., formation of the MgCl2 support) and ID molecules (i.e., different adsorption modes on the MgCl2 surface). To get clear insight into the distribution of ID adsorption energies for MgCl2–ID-TiCl4 clusters, further processing on the calculation results of the ID adsorption energy was carried out as follows: (i) For different adsorption conformations in one same model, only the higher adsorption energy would be selected in the following processing. For example, the ID adsorption energy of conformation “chelate-b” for the [ID 1 110–1] model is more negative than that of conformation “chelate-a”. In this case, only the ID adsorption energy of conformation “chelate-b” would be selected from [ID 1 110–1] models for the following process. Moreover, the D-3 cluster model was also excluded because of its low ID adsorption energy. (ii) For the ID adsorption energies from different cluster models of each ID, the highest adsorption energy was selected to be the fiducial value. (iii) The difference values (D value) between the fiducial value and the ID adsorption energy of each model were calculated, and the results were arranged in descending order. The further processing result of ID adsorption energies was listed in Figure . And the optimized structures for MgCl2–ID-TiCl4 clusters correspond to the fiducial values of IDs 1–4 (i.e., MgCl2–ID-TiCl4 cluster featuring the most negative ID adsorption energy), are listed in Figure . Further processing result of ID adsorption energies for MgCl2–ID-TiCl4 clusters. The table on the right side shows the specific MgCl2–ID-TiCl4 cluster models of each ID, which are corresponding to the model ranking. Optimized structures for the selected MgCl2–ID-TiCl4 clusters. The result in Figure clearly shows the differences of ID adsorption energies among different MgCl2–ID-TiCl4 clusters, revealing that the structure and adsorption mode of ID had great influences on the stability of MgCl2–ID-TiCl4 cluster models, which finally affected the distribution of active centers in the ZN catalyst. It is noteworthy that the distribution of multicentered active sites in the ZN catalyst has strong influences on the MWD of the PP product. , The advantage of ID adsorption energy for the most stable MgCl2–ID-TiCl4 cluster models of IDs 1–4 (Figure ) would lead to a promotion for its population among the diverse models. On the other hand, the lower ID adsorption energy gap of the diverse MgCl2–ID-TiCl4 models would give rise to more even distribution for the models. Moreover, the diversity and distribution of active sites in the ZN catalyst would influence the formation of PP chains during the polymerization process, which includes polymerization activity, chain transfer rate, and hydrogen response of active centers, and finally affect the MWD of PP. In this case, the more severe decline of D values in Figure would result in narrower MWD, while the slower descending curve of D values would lead to broader MWD. Compared with diester IDs 3 and 4, D values of diether IDs 1 and 2 exhibit more severe decline, indicating that MgCl2–ID-TiCl4 cluster models of 1 and 2 featuring higher ID adsorption energies would have greater advantages in their formation and population. From 1–4, D values of 4 exhibit the slowest descending curve, revealing that the active sites from different cluster models of ID 4 formed a relative even distribution than other IDs. The result of Figure shows a surprising relationship with the experimental MWD results of the PP product produced by ZN catalysts containing corresponding IDs (Table ), revealing the potential relationship between ID structures, distribution of active sites, and macroproperties of the PP material. It should be noticed that the microstructure of active centers in actual ZN catalysts and the polymerization kinetics are quite complicated. The MWD of PP would also be influenced by other experimental conditions, not only limited to the structure of ID. Furthermore, the proposed mechanism in our study is based on theoretical calculations rather than a validation result from experiments and characterizations. In conclusion, our research provides a novel perspective in explaining the influences of different IDs on the MWD of the PP material. 4Average D Values for IDs 1–4 and the Corresponding MWD of PP Produced by ZN Catalysts Containing Corresponding IDsIDID typeaverage D valueMWD 1 diether–19.85.16 2 diether–18.85.23 3 phthalate–10.36.78 4 dioldibenzoate–5.18.87 Average D Values for IDs 1–4 and the Corresponding MWD of PP Produced by ZN Catalysts Containing Corresponding IDs Conclusions In this work, a series of MgCl2 models have been designed and studied by DFT calculations. Diether- (1 and 2) and diester-type (3 and 4) IDs were selected to be adsorbed on MgCl2 models in chelate or bridge adsorption modes, forming MgCl2–ID clusters. The ID adsorption energies of MgCl2–ID clusters were obtained and discussed. Then TiCl4 was introduced to MgCl2–ID clusters and adsorbed on the (110)-facet of each model, resulting in MgCl2–ID-TiCl4 clusters. The ID adsorption energies of MgCl2–ID-TiCl4 clusters were also obtained, which reveal similar tendency as that of MgCl2–ID clusters. Further processing was carried out on the calculation data, the result of which shows a surprising relationship with the experimental MWD results of PP. The main conclusions we obtained are as follows:iThe structure of IDs, especially the structural flexibility between the coordinating groups and the steric hindrance effect, has a significant impact on the adsorption modes and adsorption energies of ID molecules.iiThe Lewis acidity of coordinating Mg atoms and the stereoeffect of MgCl2 corner models also have great influences on ID adsorption energies.iiiThe descending curve of D values has shown an obvious relationship with the distribution of multicentered active sites in the ZN catalyst, which finally affects the MWD of PP.ivThe consideration of MgCl2 crystallite corner models is quite necessary for the theoretical exploration of ID adsorption in the ZN catalyst. iThe structure of IDs, especially the structural flexibility between the coordinating groups and the steric hindrance effect, has a significant impact on the adsorption modes and adsorption energies of ID molecules. The structure of IDs, especially the structural flexibility between the coordinating groups and the steric hindrance effect, has a significant impact on the adsorption modes and adsorption energies of ID molecules. iiThe Lewis acidity of coordinating Mg atoms and the stereoeffect of MgCl2 corner models also have great influences on ID adsorption energies. The Lewis acidity of coordinating Mg atoms and the stereoeffect of MgCl2 corner models also have great influences on ID adsorption energies. iiiThe descending curve of D values has shown an obvious relationship with the distribution of multicentered active sites in the ZN catalyst, which finally affects the MWD of PP. The descending curve of D values has shown an obvious relationship with the distribution of multicentered active sites in the ZN catalyst, which finally affects the MWD of PP. ivThe consideration of MgCl2 crystallite corner models is quite necessary for the theoretical exploration of ID adsorption in the ZN catalyst. The consideration of MgCl2 crystallite corner models is quite necessary for the theoretical exploration of ID adsorption in the ZN catalyst. Our work provides a systematic study on the ID adsorption at the MgCl2 support of the ZN catalyst, using a series of MgCl2 support models. This research also provides a unique perspective in understanding the relationship between IDs, distribution of active centers in ZN catalysts, and MWD of the PP material. These results have great significance in the R&D of novel ID technology and high-performance ZN catalysts. More research on the mechanism of the ZN catalyst is currently ongoing in our laboratory. Supplementary Material