Synthesis and investigation of water-soluble anticoagulant warfarin/ferulic acid grafted rare earth oxide nanoparticle materials

Abstract

To overcome the poor solubility of anticoagulant nano-rare earth (RE) oxides and warfarin/ferulic acid in water, two series of water-soluble anticoagulant materials, RE₂O₃-TDI-PEG-warfarin and RE₂O₃-TDI-PEG-ferulic acid (TDI = toluene 2,4-diisocyanate, RE = La, Eu, Nd, Sc, Sm, Dy), were prepared via a grafting method. The novel materials were characterized by IR, TG, XRD, SEM, TEM, ¹H and ¹³C NRM and particle size tests. The results confirmed that warfarin and ferulic acid were successfully modified onto the surfaces of the rare earth nano-oxides. The anticoagulant properties were evaluated on the basis of coagulation time (CT), recalcification time (RT), activated partial thromboplastin time (APTT) and prothrombin time (PT). It was demonstrated that the novel hybrid materials have better cell compatibility and anticoagulant action than that of warfarin sodium or ferulic acid, due to the increased solubility (>10 mg mL⁻¹) in water and synergy between nano-RE₂O₃ and warfarin/ferulic acid. It was also found that the anticoagulant time of the hybrid materials depends on the concentration i.e., the higher the concentration of hybrid materials, the longer the anticoagulant time. Additionally, Sc₂O₃-TDI-PEG-warfarin and Sc₂O₃-TDI-PEG-ferulic acid exhibited the best anticoagulant properties, which indicate potential application in the medicinal area. These results also provide the opportunity for the development of anticoagulant complexes and the potential for application in other related fields.

---

1. Introduction

Warfarin (W) and ferulic acid (FA) have long been the major drugs for anticoagulation in clinical applications^1–3 due to their advantages of oral availability, convenient application, inexpensive price, and lasting effect (Scheme 1). However, Shouming Wang⁴ pointed out that there are a few drawbacks from the extended use of warfarin compounds as anticoagulant drugs, such as spontaneous bleeding and skeletal development retardation. Meanwhile, FA is widely studied in combination therapies in treating cancer, myocardial infarction, atherosclerosis,⁵ acute cerebral infarction⁶ and for kidney protection.⁷ However, the application of FA in clinics is limited by its difficulty in penetrating the biological membrane bilayer, and the poor solubility in water. Therefore, it is important to find a substitute that has lower toxicity, better anticoagulant properties and fewer side effects compared to warfarin/ferulic acid and their derivatives. One useful strategy to improve the efficiency of warfarin and ferulic acid is to fabricate hybrid materials by introducing new components. Studies on the properties of warfarin complexes composited with rare earth complexes⁸ or transition metals⁹ by the groups of Jiao and Dong have demonstrated that these complexes exhibit better hemorrhagic anticoagulant properties, lower toxicity and high stability. Meanwhile, complexes of ferulic acid with transition metals or rare earth ions were found to prolong the anticoagulant time and recalcification time.¹⁰ However, the poor water solubility still limits their applications in the clinical sense.

Scheme 1 The structure of warfarin and ferulic acid.

Rare earth complexes including rare earth nano-oxides as anticoagulants have been broadly investigated.^11,12 For example, recently, Wang and co-workers¹³ presented the anticoagulant performance of heparin/rare earth nano-oxide hybrids. Ni and Zhou^14,15 et al. reported that the rare earth elements can replace Ca²⁺ to inhibit the coagulation of blood.

Though the rare earth nano-oxide ferulic acid and warfarin complexes have a significant anticoagulant effect, they suffer from poor water solubility along with the problem of spontaneous bleeding due to large amounts of drug usage. As a result, it is necessary to develop novel anticoagulant materials with high water solubility and only a small amount of drug usage. On this basis, herein two series of water soluble RE₂O₃-TDI-PEG-Warfarin and RE₂O₃-TDI-PEG-Ferulic acid complexes were prepared (Fig. 1) by combining water soluble H₂N–PEG–NH₂ with RE₂O₃ nano-particles and ferulic acid/warfarin. These novel hybrid materials were characterized by infrared spectroscopy (IR), thermo-gravimetric analysis (TG), scanning electron microscopy (SEM), powder X-ray diffractometry (powder-XRD), transmission electron microscopy (TEM), ¹H and ¹³C NMR spectroscopy and particle size analysis. The anticoagulant properties were evaluated on the basis of anticoagulant time (CT), recalcification time (RT), activated partial thromboplastin time (APTT) and prothrombin time (PT). The results demonstrated that the novel hybrid materials have longer anticoagulation times than that of ferulic acid or warfarin, and significantly improved water solubility (>10 mg mL⁻¹) compared with bare ferulic acid (<1 mg mL⁻¹) and warfarin (<1 mg mL⁻¹). In addition, the cell adhesion of the materials remained the same as untreated control samples.

Fig. 1 Preparation of the hybrid materials.

2. Experimental

2.1 Materials and measurement

RE₂O₃ nano-oxides were purchased from the Shanghai Yuelong Plant (China). Warfarin sodium was acquired from the Shanghai Sixteenth Pharmaceutical factory (China) (m.p. 200–205 °C), and was treated with HCl to produce warfarin (m.p. 160.5–161.0 °C, literature:⁸ 159.5–160.5 °C). Ferulic acid and PEG-4000 were obtained from the Shanghai Zhongqin Chemical Reagent Limited Company (China). The H₂N–PEG–NH₂ was prepared according to the literature method.¹⁶ Toluene 2,4-diisocyanate (TDI) was acquired from the Shanghai Kangda Amino acid factory (China). All other chemicals and solvents were of analytical grade.

FT-IR spectra of the materials were obtained on a Nicolet NEXUS 670 FT-IR spectrometer with a wave number range of 4000–400 cm⁻¹ from KBr pellets. Thermogravimetric analysis (TG) was performed through a Perkin-Elmer TGA7 instrument in flowing N₂ and a heating rate at 10° min⁻¹ in the 30–700° range. The morphological features of all materials were identified through scanning electron microscopy (SEM) (XL-20 (Philips, Netherlands)). Transmission electron microscope (TEM) was used to determine the morphology of the hybrid materials. Powder X-ray diffraction (powder-XRD) analysis was obtained with a Philips XPert Pro diffractometer using Ni-filtered and Cu Kα radiation. Particle size analysis was performed on a Zetasizer Nano-Zs laser particle size instrument (Malvern, UK). ¹H and ¹³C NMR spectra were recorded using a Bruker 400 MHz and d₆-DMSO as a solvent.

2.2 Modification of the surfaces of rare earth nano-oxides

The rare earth nano-oxides (RE₂O₃) were dried in a vacuum oven at 120 °C for 7–8 h before use. RE₂O₃ (5 g), TDI (5 g) and 100 mL anhydrous acetone were added into a flask and sonicated under N₂ atmosphere for 2 h. To this solution, an appropriate amount of dibutyltin dilaurate was added and the solution was sonicated at 80 °C for 2 h. Finally, the reaction mixture was separated by filtration and the obtained solid was washed with anhydrous acetone (30 mL). The resulting materials were then heated in a vacuum oven at 80 °C. The obtained solid was labeled as RE₂O₃-TDI and was slightly soluble in water.

2.3 Preparation of warfarin/ferulic acid rare earth nano-oxide hybrid materials

RE₂O₃-TDI (20 mg) was dissolved in chloroform (50 mL) and sonicated for 0.5 h. Then, H₂N–PEG–NH₂ (100 mg) was added and the mixture was stirred at room temperature for 12 h before the addition of warfarin (10 mg) and anhydrous acetone (5 mL). The resulting mixture was refluxed for 12 h under N₂ at 70 °C. Upon completion of the reaction, the solid was separated by filtration, washed with diethyl ether (15 mL) and roasted in a vacuum oven at 45 °C for 24 h. The obtained solid was marked as RE₂O₃-TDI-PEG-W and was soluble in water (>10 mg mL⁻¹) as well as some organic solvents, such as anhydrous ethanol, acetone, DMSO, etc.

In a similar manner, ferulic acid (13 mg), dimethylformamide (DMF, 15 mL), N,N′-dicyclohexylcarbodiimide (DCC, 13 mg) and N-hydroxyl succinimide (NHS, 11 mg) were added subsequently to the RE₂O₃-TDI and H₂N–PEG–NH₂ mixed solution. The mixture was stirred for 12 h under N₂ at room temperature. Finally, the solid was separated by filtration and washed with diethyl ether (15 mL) and heated in a vacuum oven for 24 h at 45 °C. The obtained solid was marked as RE₂O₃-TDI-PEG-FA and was soluble in water (>10 mg mL⁻¹) and some organic solvents.

2.4 Cell compatibility

The cell compatibility was measured by the MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) method.¹⁷ The HepG2 cells were routinely cultured in tissue culture boxes with 1640 culture medium (containing 10% fetal bovine serum), and incubated at 37 °C in a humidified atmosphere with 95% air and 5% CO₂. After several generations of cells, the solution of hybrid materials (100 μL, 10 mg mL⁻¹, by diluting with 1640 culture medium) was added to the corresponding cell in the well, and then co-cultured at 37 °C for 48 h in a humidified atmosphere with 5% CO₂. After harvesting the cells, the MTT solution (150 μL, 5 mg mL⁻¹, PBS buffer solution) was added and the cell was cultured for 4 h at 37 °C. After removing the MTT solution, DMSO (150 μL) was added in each well at 37 °C, and the cell lysate was measured in a Bio-Rad Microplate Reader (model 1680) at excitation and emission wavelengths of 570 nm. The cell activity was expressed as a percentage of the optical absorbance of the test group vs. the control group.¹⁷

2.5 Test of anticlotting activity

2.5.1 Whole blood coagulation tests (CT). Clotting time was determined visually at 37 °C using eight groups of clean test tubes (three tubes in each group). The first group was blank and the other seven groups contained 10 mg of warfare sodium, ferulic acid, RE₂O₃, RE₂O₃-TDI or RE₂O₃-TDI-PEG, RE₂O₃-TDI-PEG-W, RE₂O₃-TDI-PEG-FA. All of the test tubes were maintained at a constant temperature of 37 °C. Healthy human blood (1 mL) was added to each tube. After 2 min, the test tubes were slightly warmed about 30 °C every 30 s until no more blood flow was observed at the constant temperature. The time between the addition of blood and no more blood flow was recorded as the CT. To analyze the results from the test of anticlotting activity, the SPSS (Statistical Product and Service Solution) softword version 16.0 was used. Standard deviation has been included for each sample and the value of significance of the differences (P) have been annotated in the tables.

2.5.2 Recalcification time tests (RT). The hybrid materials were dispersed in NaCl solution (0.154 mol L⁻¹) for 24 h and then collected. Eight groups of clean test tubes (three tubes in each group) were used. The first group was blank and the rest of the seven groups contained samples (10 mg each). All samples were incubated at 37 °C in a water bath for 2 min, then the pre-warmed mixture of blood plasma (1.0 mL) and CaCl₂ solution (0.3 mL, 0.025 mol L⁻¹) were added into each test tube. A stainless steel crotchet was inserted in the solution and agitated slowly until the fibrin was formed (at 37 °C). The recalcification time is defined as the difference between the addition of the mixture and the formation of fibrin.

2.5.3 Anticoagulant activity test. The activated partial thromboplastin time (APTT) and prothrombin time (PT) indicate the influence of hybrid materials on the coagulant.¹⁸ Blood plasma (0.5 mL) was added to the test tubes containing 10 mg of RE₂O₃-TDI-PEG-W or RE₂O₃-TDI-PEG-FA. The solution of CaCl₂ (0.25 mL, 0.025 mol L⁻¹) was then added to these tubes and incubated at 37 °C until the temperature was constant. The time of fibrin formation was recorded as the APTT. The time of blood plasma clotting was recorded as the PT, when the PT agent was added to replace CaCl₂.

3. Results and discussion

3.1 Anticlotting properties

3.1.1 Whole blood coagulation tests (CT). Fig. 2 shows the CT of blank, nano-RE₂O₃, RE₂O₃-TDI, RE₂O₃-TDI-PEG, RE₂O₃-TDI-PEG-W and RE₂O₃-TDI-PEG-FA. Under identical conditions, the CT of the blank sample was the shortest one (about 10.24 min). The CT time of RE₂O₃-TDI-PEG-W (13.1–15.0 min) was longer than that of warfarin sodium (12.3 min), and the CT time of RE₂O₃-TDI-PEG-FA (14 ± 0.5 min) was almost the same as that of ferulic acid (14.0 min). It was concluded that 6% (w%, M_w/(M_RE2O3 + M_TDI + M_PEG + M_w)) warfarin or 4% ferulic acid in hybrid materials can achieve better or at least the same anticoagulant performance compared with the corresponding bare drug. Hence, the hybrid materials are promising candidates for application in the medical field. It is interesting to note that the CT time of the samples was in the order of nano RE₂O₃ > RE₂O₃-TDI > RE₂O₃-TDI-PEG-W/RE₂O₃-TDI-PEG-FA > RE₂O₃-TDI-PEG. This may be because of the significant decrease in RE₂O₃ content by the introduction of PEG.

Fig. 2 Coagulation time (CT) of different compounds.

The anticoagulant properties of RE₂O₃-TDI and RE₂O₃-TDI-PEG can be attributed to nano RE₂O₃. RE₂O₃-TDI-PEG-W/RE₂O₃-TDI-PEG-FA have longer CT times than that of RE₂O₃-TDI-PEG, due to the existence of warfarin/ferulic acid. The CT time of Sc₂O₃-TDI-PEG-W and Sc₂O₃-TDI-PEG-FA was the longest in all of the hybrid materials because of the similar chemical properties of Sc and Ca. The two elements are neighbours in the same period in the periodic table and the effective nuclear charge of Sc³⁺ is higher than the other rare earth ions. Therefore, Sc³⁺ can replace Ca²⁺ more easily than the other rare earth ions.

3.1.2 Test of recalcification time. Fig. 3 shows the recalcification time (RT) of warfarin sodium, ferulic acid and their respective hybrid materials. Under identical conditions, the RT of RE₂O₃-TDI-PEG-W (11.0–13.1 min) and RE₂O₃-TDI-PEG-FA (13.0–13.5 min) are significantly enhanced compared with the control sample (10 min), warfarin sodium and ferulic acid (11–12 min). The RT of RE₂O₃-TDI-PEG-W/RE₂O₃-TDI-PEG-FA is the longest in the group of nano-RE₂O₃ (11.2–12.0 min), RE₂O₃-TDI (10.2–12.0 min) and RE₂O₃-TDI-PEG (10.2–12.0 min). It is known that the longer the RT, the better the anticoagulant property. It can be therefore concluded here that the hybrid materials have better anticoagulant properties than that of RE₂O₃, warfarin sodium and ferulic acid. Additionally, Sc₂O₃-TDI-PEG-FA exhibited the best recalcification time, possibly due to its small radius and the highly effective nuclear charge of Sc³⁺.

Fig. 3 Recalcification time (RT) of various complexes.

The CT and RT of the mixture of drugs and rare earth nano-oxides were also investigated. The results (Table 1) show that the CT and RT of the mixture were shorter than those of the hybrid materials. As there were about 50% (w%) drugs in the mixture, it could not achieve the same anticoagulant performance compared with the hybrid materials (6% warfarin or 4% ferulic acid).

Table 1 CT (min) and RT (min) of the drug mixture (5 mg, RE₂O₃, 5 mg W/FA)

Mixture	CT/min	P	RT/min	P
Sc2O3 + W	9.20(±0.02)	<0.05	6.20(±0.02)	<0.05
Sc2O3 + FA	9.50(±0.02)	<0.05	6.70(±0.02)	<0.05
La2O3 + W	9.00(±0.02)	<0.05	6.10(±0.02)	<0.05
La2O3 + FA	9.50(±0.02)	<0.05	6.50(±0.02)	<0.05
Nd2O3 + W	10.50(±0.02)	<0.05	6.80(±0.02)	<0.05
Nd2O3 + FA	10.70(±0.02)	<0.05	7.30(±0.02)	<0.05
Sm2O3 + W	9.80(±0.02)	<0.05	7.00(±0.02)	<0.05
Sm2O3 + FA	10.10(±0.02)	<0.05	7.50(±0.02)	<0.05
Eu2O3 + W	9.10(±0.02)	<0.05	6.10(±0.02)	<0.05
Eu2O3 + FA	9.20(±0.02)	<0.05	6.20(±0.02)	<0.05
Dy2O3 + W	9.30(±0.02)	<0.05	7.10(±0.02)	<0.05
Dy2O3 + FA	9.20(±0.02)	<0.05	7.30(±0.02)	<0.05

---

3.1.3 Anticoagulant activity test. APTT and PT are multi-purpose investigation tools for assessing the coagulation activity in vivo and in vitro. Prolonged APTT and PT indicate that there is an unusual congenital clotting factor or a lacking of multiple clotting factors in the blood. The data in Table 2 shows: (1) the APTT and PT of hybrid materials were longer than that of the blank sample, which means the hybrid materials had better anticoagulant properties; (2) the values of APTT and PT of the hybrid materials were within the normal scope (APTT: 25-27, PT: 10-13), which indicates there was no effect on the blood property when a small amount of hybrid materials were present. Thus, these kinds of hybrid materials clearly show potential for application in the medical field.

Table 2 Anticoagulant activity of the hybrid materials (10 mg)

Compound	APTT/s	P	PT/s	P
Blank	27.50(±1)	<0.05	12.40(±1)	<0.05
Sc2O3-TDI-PEG-W	34.20(±2)	<0.05	13.60(±1)	<0.05
Sc2O3-TDI-PEG-FA	33.70(±2)	<0.05	13.30(±1)	<0.05
La2O3-TDI-PEG-W	34.70(±2)	<0.05	13.50(±1)	<0.05
La2O3-TDI-PEG-FA	34.00(±2)	<0.05	13.20(±1)	<0.05
Nd2O3-TDI-PEG-W	36.50(±2)	<0.05	13.80(±1)	<0.05
Nd2O3-TDI-PEG-FA	33.20(±2)	<0.05	13.10(±1)	<0.05
Sm2O3-TDI-PEG-W	33.20(±2)	<0.05	13.20(±1)	<0.05
Sm2O3-TDI-PEG-FA	33.00(±2)	<0.05	13.20(±1)	<0.05
Eu2O3-TDI-PEG-W	34.70(±2)	<0.05	13.50(±1)	<0.05
Eu2O3-TDI-PEG-FA	34.20(±2)	<0.05	13.10(±1)	<0.05
Dy2O3-TDI-PEG-W	34.50(±2)	<0.05	13.60(±1)	<0.05
Dy2O3-TDI-PEG-FA	33.50(±2)	<0.05	13.10(±1)	<0.05
Reference	25-37		10-13

---

3.1.4 Cell compatibility. Fig. 4 indicates the HepG2 cell viability after seeding for 48 h in the solution of nano-Nd₂O₃, Nd₂O₃-TDI, Nd₂O₃-TDI-PEG, Nd₂O₃-TDI-PEG-W and Nd₂O₃-TDI-PEG-FA. It can be seen that all cells had similar activation compared to the blank sample. Additionally, it was found that the cell activity was highest in the presence of nano-Nd₂O₃ and Nd₂O₃-TDI. It indicated the cell amplification was effectively promoted by nano-Nd₂O₃ and Nd₂O₃-TDI. After PEG was grafted onto the surface of Nd₂O₃-TDI, the cell activity decreased. The data of cell activity showed that cell toxicity was reduced and cell compatibility was enhanced after warfarin and ferulic acid were grafted onto RE₂O₃-TDI-PEG. This illustrates these hybrid materials have low toxicity and higher cellular compatibility. The results of the HepG2 cell viability (Table 3) after 2 days of the other hybrid materials seeding show no more effects of the materials on the cell activity, which displays that the materials have as low toxicity as the blank sample.

Fig. 4 HepG2 cell viability after 2 days of Nd hybrid material seeding.

Table 3 HepG2 cell viability after 2 days of the hybrid material seeding

Compound	Cell viability (%)	P
Sc2O3-TDI-PEG-W	94	<0.05
Sc2O3-TDI-PEG-FA	92	<0.05
La2O3-TDI-PEG-W	95	<0.05
La2O3-TDI-PEG-FA	89	<0.05
Nd2O3-TDI-PEG-W	99	<0.05
Nd2O3-TDI-PEG-FA	96	<0.05
Sm2O3-TDI-PEG-W	93	<0.05
Sm2O3-TDI-PEG-FA	88	<0.05
Eu2O3-TDI-PEG-W	90	<0.05
Eu2O3-TDI-PEG-FA	94	<0.05
Dy2O3-TDI-PEG-W	96	<0.05
Dy2O3-TDI-PEG-FA	94	<0.05
Blank	96	<0.05

---

3.2 Characterization

3.2.1 FT-IR spectra. The novel hybrid materials were characterized by FT-IR spectroscopy. Due to similarity of the RE₂O₃ materials, nano-Eu₂O₃ was chosen as the sample for detailed interpretation. The hydroxyl group in nano-Eu₂O₃ had vibration absorptions at 1620 and 3400 cm⁻¹. The characteristic absorption of the –NCO group of TDI appeared at 2270–2280 cm⁻¹. If the reaction occurs between the –NCO group of TDI and the hydroxyl in the surface of nano-Eu₂O₃, the absorption of –NH should appear at 3500–3300 cm⁻¹.

Fig. 5(A) shows the IR spectra of Eu₂O₃-TDI. There was a vibration peak at 3300 cm⁻¹, which indicated the reaction had occurred. However, the presence of a characteristic absorption of –NCO at 2273 cm⁻¹ suggests that one –NCO in the TDI was preserved. This group was used to react with H₂N–PEG–NH₂.¹⁹ The IR spectra of Eu₂O₃-TDI revealed the presence of the carbonates absorption (1640 and 1543 cm⁻¹),²⁰ the benzene absorption (1605 cm⁻¹), the anti-symmetric and symmetric stretching vibration peaks of the C–H of the TDI methyl (2922 and 2857 cm⁻¹), all of which confirm the successful grafting of TDI onto the surface of nano-Eu₂O₃.

Fig. 5 IR spectra of (A) Eu₂O₃-TDI, (B) Eu₂O₃-TDI-PEG, (C) warfarin (D) Eu₂O₃-TDI-PEG-W, (E) Ferulic Acid and (F) Eu₂O₃-TDI-PEG-FA.

Fig. 5(B) shows the IR spectra of Eu₂O₃-TDI-PEG. The disappearance of the characteristic absorption of –NCO indicates that the –NCO left in the Eu₂O₃-TDI surface reaction with –NH of PEG. The characteristic vibrations of PEG appeared at 2882 cm⁻¹ (ν_(C–C–H)) and 1109 cm⁻¹ (ν_{(–C–O–C)}), which confirmed the formation of Eu₂O₃-TDI-PEG.

Fig. 5(C) reveals the IR spectra of warfarin. Two carbonyl absorption peaks appeared at 1718 and 1664 cm⁻¹.²¹ The former belongs to the ketone carbonyl stretching vibration; the latter is the vibration of the aromatic ring of carbonyl stretching, an O–H stretching vibration appeared at 3283–3500 cm⁻¹.

In the IR spectra of Eu₂O₃-TDI-PEG-W (Fig. 5(D)), the vibration of carbonyl stretching was absent, and –C=N absorption appeared at 1631 cm⁻¹, which suggested that the reaction between –C=O and –NH₂ had occurred and that –C=N was formed consequently.

Fig. 5(E) and (F) illustrate the IR spectra of ferulic acid and the hybrid material, Eu₂O₃-TDI-PEG-FA. The spectra showed that the vibration peak of carboxylic carbonyl appeared at 1686 cm⁻¹ in the hybrid material. This number shifted toward the lower wave number reaching 1635 cm⁻¹, which indicated the formation of an amide bond between ferulic acid and –NH₂ of PEG. The results clearly suggest that ferulic acid has been grafted to the surface of nano-Eu₂O₃. The absorbance of the signal at 3425 cm⁻¹ was the stretching vibration of the phenol hydroxyl group. After forming the hybrid materials Eu₂O₃-TDI-PEG-FA, it became a smooth wide peak (3300–3600 cm⁻¹), which indicated the successful formation of ferulic acid hybrid materials. The results of the IR spectra of other samples were in approximate agreement with that of Eu₂O₃-TDI-PEG-W and Eu₂O₃-TDI-PEG-FA (Table 4).

Table 4 IR spectra (cm⁻¹)

Compound	–C=O	–C=N
Warfarin	1718	—
Ferulic acid	1686	—
Sc2O3-TDI-PEG-W	—	1627
Sc2O3-TDI-PEG-FA	1639	—
La2O3-TDI-PEG-W	—	1641
La2O3-TDI-PEG-FA	1651	—
Nd2O3-TDI-PEG-W	—	1647
Nd2O3-TDI-PEG-FA	1658	—
Sm2O3-TDI-PEG-W	—	1644
Sm2O3-TDI-PEG-FA	1656	—
Eu2O3-TDI-PEG-W	—	1631
Eu2O3-TDI-PEG-FA	1635	—
Dy2O3-TDI-PEG-W	—	1638
Dy2O3-TDI-PEG-FA	1646	—

---

3.2.2 Thermal analysis. Fig. 6 shows the weight loss curves of Eu₂O₃, Eu₂O₃-TDI, Eu₂O₃-TDI-PEG and Eu₂O₃-TDI-PEG-FA. The results show that the weight loss of TDI was about 31.9% (theory: 33.1%) at 200–300 °C. At 300–600 °C, the weight loss was mainly due to the decomposition and melting oxide of PEG and ferulic acid in the hybrid materials. The thermo-gravimetric curves of Eu₂O₃-TDI-PEG-FA reveal two significant weight loss steps, which were most likely due to the burning of TDI and PEG-FA. This result supported the conclusion obtained from IR spectra that the surface of RE₂O₃ was coated by PEG and warfarin/ferulic acid molecules. The curve of the weight loss of Eu₂O₃-TDI-PEG-W and the other hybrid materials was very similar to Eu₂O₃-TDI-PEG-FA. The TGA-DTA data is listed in Table 5. There is a remarkable resemblance between Eu₂O₃-TDI-PEG-W, Eu₂O₃-TDI-PEG-FA and the other compounds during weightlessness.

Fig. 6 TGA curves.

Table 5 TG-DTA of the materials

Compounds	TDI-PEG-W losing temperature/°C, losing rate/% found	TDI-PEG-FA losing temperature/°C, losing rate/% found	Residual weight/% found (calc.)
Eu2O3-TDI-PEG-W	300–600, 66.7		33.3
Eu2O3-TDI-PEG-FA		300–600, 65.4	34.6
Sc2O3-TDI-PEG-W	200–650, 77.5		12.5
Sc2O3-TDI-PEG-FA		200–650, 78.3	13.3
La2O3-TDI-PEG-W	4200–550, 70.0		29.6
La2O3-TDI-PEG-FA		200–550, 74.4	25.6
Nd2O3-TDI-PEG-W	200–550, 68.8		31.2
Nd2O3-TDI-PEG-FA		200–550, 67.4	32.6
Sm2O3-TDI-PEG-W	200–600, 67.6		32.4
Sm2O3-TDI-PEG-FA		200–600, 72.5	26.5
Dy2O3-TDI-PEG-W	200–550, 72.2		27.8
Dy2O3-TDI-PEG-FA		200–550, 71.2	26.8

---

3.2.3 ¹H NMR analysis. ¹H and ¹³C NMR of the warfarin, ferulic acid, H₂N–PEG–NH₂, and La-materials were investigated using d₆-DMSO as a solvent. Here, the ¹H NMR chemical shifts of the main groups list as follows:

Ferulic acid: 3.83 ppm. (–OCH₃), 6.37 ppm (–CH=C–benzene), 7.54 ppm (–benzene–CH=C), 9.57 ppm (–OH), 12.16 ppm. (–COOH); warfarin: 2.12 ppm (–CH₃), 4.94 ppm (–CH₂), 11.39 ppm. (–OH); H₂N–PEG–NH₂: 2.49 ppm (–NH₂), 3.33 ppm (N–CH₂), 3.49 ppm. (–CH₂); La-TDI-PEG-FA: 2.32 ppm. (–NH), 6.45 ppm (–CH=C–benzene), 7.54 ppm(–benzene–CH=C), 3.79 ppm (–OCH₃), 10.51 ppm (–OH); La-TDI-PEG-W: 2.32 ppm (–CH₃), 3.39 ppm (CH–N), 3.49 ppm (–CH₂) and 10.45 ppm (–OH).

In La-TDI-PEG-FA, the peak at 12.38 ppm disappeared, which was attributed to the –COOH group in the ferulic acid;²² whereas the signals of –OCH₃ of ferulic acid and N–CH₂ of H₂N–PEG–NH₂ protons shifted slightly upfield; in La-TDI-PEG-W, the signals of –CH₃ of warfarin at 2.12 ppm and N–CH₂ of H₂N–PEG–NH₂ at 3.33 ppm were moved downfield. Simultaneously, the peak of the –OH group of warfarin at 11.39 ppm shifted upfield to 10.45 ppm. Hence, suggesting the successful preparation of ferulic acid or warfarin was grafted onto nano-RE₂O₃ modified by H₂N–PEG–NH₂ and TDI.

The ¹³C NMR chemical shifts of the main groups of ligands and La-complex list as follows: ferulic acid: 116.74 ppm (C=C–), 145.37 ppm (–C=C), 170.01 ppm (–COOH); warfarin: 23.15 ppm (–C(=O)–CH₃), 27.12 ppm (–CH₃), 47.41 ppm (–CH₂); H₂N–PEG–NH₂: 40.10 ppm (N–CH₂), 70.20 ppm (–CH₂); La₂O₃-TDI-PEG-W: 31.11 ppm (–CH₃), 36.20 ppm (–C(=O)–CH₃), 40.17 ppm (N–CH₂), 70.21 ppm (–CH₂), 162.75 ppm (N–C(=O)–N); La₂O₃-TDI-PEG-FA: 39.95 ppm (N–CH₂), 70.21 ppm (–CH₂), 115.09 ppm (C=C–), 138.18 ppm (–C=C) and 162.72 ppm (N–C(=O)–N).

Obviously, when RE₂O₃-TDI-PEG-FA was formed, the N–CH₂ peak at 40.10 ppm of H₂N–PEG–NH₂, C=C– and –C=C of ferulic acid, at 116.74 and 145.37 ppm, respectively, shifted weakly upfield. The peak of –COOH was attributed to ferulic acid at 170.01 ppm, which moved to 162.72 ppm because it formed a acylamino group; in La-TDI-PEG-W, the signals of N–CH₂ at 40.10 ppm of H₂N–PEG–NH₂, –C(=O)–CH₃ and –CH₃ of warfarin at 23.15 and 27.12 ppm, respectively, shifted downfield. All of these phenomena show that RE₂O₃-TDI-PEG-FA and RE₂O₃-TDI-PEG-W were successfully synthesized.

3.2.4 SEM analysis. The SEM images (Fig. 7) illustrate the morphology of the rare earth nano-oxides particles, PEG, La₂O₃-TDI-PEG-W and La₂O₃-TDI-PEG-FA. As shown in the SEM image, the nano-oxide particles had a spherical structure, which was significantly different from the plated PEG, flower structure of La₂O₃-TDI-PEG-W and La₂O₃-TDI-PEG-FA. It implies that the graft reaction has occurred on the surface of the rare earth nano-oxides and PEG-W/PEG-FA had completely enveloped the rare earth nano-oxides. However, in the La₂O₃ image of SEM, there are different size particles present, which show some aggregation existing in these particles.

Fig. 7 SEM images of compounds.

3.2.5 TEM analysis. In the TEM images (Fig.8), the spherical black spots are metal particles and the white stripes are organic substances, the organic substance formed a skeleton and the metal particles were evenly dispersed there. It can be seen that the metal particles are packaged by organic molecules from the enlarged picture, which indicates that the title compound has been formed. In addition, it can be observed that different size particles existed in the TEM images, which due to the RE₂O₃ nano-particles gathered and the organic substance parceled these particles loosely or tightly.

Fig. 8 TEM images of compounds.

3.2.6 XRD analysis. Powder X-ray diffractometry is a useful method for the detection of compounds in powder or microcrystalline states. Fig. 9 shows the X-ray diffraction patterns of warfarin sodium, ferulic acid, Dy₂O₃-TDI, Dy₂O₃-TDI-PEG-W and Dy₂O₃-TDI-PEG-FA. The XRD pattern of FA showed intense, sharp peaks at the diffraction angles (2θ) of 8.68°, 15.24°, 17.12°, 20.88°, and 26.16°, that proved the typical crystalline nature of the compound. Warfarin had the strongest diffraction peaks at 7.84°, 15.64°, 21.2°, 25.44° and 29.92°, implying its crystalline nature. On the other hand, the powder diffraction pattern of Dy₂O₃-TDI displayed several sharp peaks at 20.3°, 28.88°, 33.5°, 43.1°, 48.2°, 52.78°, and 57.2°, which indicates the compound exists in a crystalline form. In the meantime, the powder diffraction pattern of Dy₂O₃-TDI-PEG-W and Dy₂O₃-TDI-PEG-FA were all lacking crystalline peaks, which means they are amorphous.²³ This may be due to the larger size of the hybrid material of Dy₂O₃-TDI-PEG-W and Dy₂O₃-TDI-PEG-FA, so they cannot easily form a regular structure.

Fig. 9 XRD of compounds.

3.2.7 Particle size analysis. To confirm the successful surface modification, the particle size of these hybrid materials was tested and analyzed by the Malvern DTS software (version 6.0). When using this method to analyse samples, two values are given, a mean value for size (Z-average), and a width parameter known as the polydispersity index (PDI). PDI is the particle size distribution coefficient and Z-average (nm) is the average value of the particle diameter. The PDI and Z-average values of Nd₂O₃-TDI and the hybrid materials are shown in Table 6.

Table 6 PDI and Z-average value of hybrid materials

Compounds	PDI	Z-Average (d, nm)	P
Eu2O3-TDI-PEG-W	0.533	181	<0.5
Eu2O3-TDI-PEG-FA	0.479	157	<0.5
Sc2O3-TDI-PEG-W	0.588	134	<0.5
Sc2O3-TDI-PEG-FA	0.578	120	<0.5
La2O3-TDI-PEG-W	0.680	174	<0.5
La2O3-TDI-PEG-FA	0.494	156	<0.5
Nd2O3-TDI-PEG-W	0.491	238	<0.5
Nd2O3-TDI-PEG-FA	0.446	214	<0.5
Sm2O3-TDI-PEG-W	0.468	224	<0.5
Sm2O3-TDI-PEG-FA	0.442	198	<0.5
Dy2O3-TDI-PEG-W	0.577	268	<0.5
Dy2O3-TDI-PEG-FA	0.469	247	<0.5
PEG-4000	0.494	116	<0.5

---

Fig. 10 present the intensity (%) of the particle size for Nd₂O₃-TDI, Nd₂O₃-TDI-PEG, Nd₂O₃-TDI-PEG-W and Nd₂O₃-TDI-PEG-FA (0.01 mg mL⁻¹, water), respectively. The value of the Z-average is 8.90 nm for Nd₂O₃-TDI, 200 nm for Nd₂O₃-TDI-PEG and 238/214 nm for Nd₂O₃-TDI-PEG-W/Nd₂O₃-TDI-PEG-FA. From Nd₂O₃, to hybrid materials, the Z-average value becomes larger and larger, so warfarin (W) or ferulic acid (FA) has been successfully grafted onto the nano-RE₂O₃ by PEG.

Fig. 10 Particle size distribution of compounds.

Generally, it is believed when the value of PDI < 0.2, it is mainly a monodisperse system.²⁴ However, the PDI values of all the samples are mostly over 0.2, which suggests that there are different sizes of particle existing in the test samples. This phenomenon is supported by the results of SEM and TEM images.

4. Conclusions

A class of water-soluble anticoagulant warfarin/ferulic acid grafted rare earth nano-oxide hybrid materials was reported by using TDI and PEG as connection reagents. The products were characterized by IR spectroscopy, TGA, ¹H and ¹³C NMR spectroscopy, XRD, SEM, TEM and particle size testing. The results showed that warfarin/ferulic acid were successfully grafted onto the surface of the nano rare earth oxides. It was demonstrated that hybrid materials have better cell compatibility and anticoagulant properties. The anticoagulant time, the recalcification time, the activated partial thromboplastin time and prothrombin time depend on the concentration and therefore the higher the concentration of the hybrid material, the better the anticoagulant property. When a small amount of hybrid material was added to the blood, no obvious variation of the properties of the blood was observed. These novel hybrid materials can serve as a foundation for clinical screening of anticoagulant drugs.

References

1. E. Nomun, A. Kashiwada and A. Hosoda, Bioorg. Med. Chem., 2003, 11, 3807.
2. Y. J. Hu, O. Y. Yu and C. M. Dai, Biomacromolecules, 2010, 11, 106.
3. T. B. Lu, M. Thomas and K. Shelley, J. Med. Chem., 2010, 53, 1843.
4. S. M. Wang, B. Richard and B. Tob, J. Med. Chem., 2010, 53, 1465.
5. S. Y. Ou and K. C. Kwok, J. Sci. Food Agric., 2004, 84, 1261.
6. C. Anselmi, M. Centini and M. Ricci, et al., J. Pharm. Biomed. Anal., 2006, 40, 875.
7. B. H. Wang and J. P. Ou-Yang, Cardiovasc. Drug Rev., 2005, 23, 161.
8. T. Q. Jiao, J. G. Wu and F. L. Zeng, Synth. React. Inorg. Met.-Org. Chem., 1999, 29, 725.
9. Y. L. Dong, N. N. Luan, H. Y. Wang and Y. M. Song, Acta Chim. Sin., 2008, 66, 1497 in Chinese.
10. P. Y. Wang, B. Wu and C. X. Bian, Chin. J. Inorg. Chem., 2012, 28, 1609 in Chinese.
11. T. Funakoshi, K. Furushima and H. Shimada, Biochem. Int., 1992, 28, 113.
12. L. J. Fu, J. X. Li, X. G. Yang and K. Wang, J. Biol. Inorg. Chem., 2009, 14, 219.
13. K. J. Wang, H. X. Li and Y. M. Song, Biopolymers, 2010, 93, 887.
14. N. Xiang, X. M. Zhang and L. Y. Tian, J. Biol., 2009, 26, 64.
15. X. B. Zhou and Y. Z. Wei, Chin. Bull. Life Sci., 1999, 11, 70 in Chinese.
16. B. D. Wang and C. J. Xu, J. Am. Chem. Soc., 2008, 130, 14437.
17. X. F. Hu and J. Ji, Acta Polym. Sin., 2009, 8, 828.
18. S. Kunar, N. Alnasif and E. Fleige, et al., Eur. J. Pharm. Biopharm., 2014, 88, 625.
19. C. M. Wu, T. W. Xu and W. H. Yang, J. Membr. Sci., 2003, 216, 269.
20. H. R. Li, J. Lin and H. J. Zhang, Chem. Mater., 2002, 14, 3651.
21. E. J. Valente, E. C. Lingafeler and W. R. Porter, J. Med. Chem., 1997, 20, 1489.
22. M. Bunzel, J. Ralph, C. Funk and H. Steinhart, Eur. Food Res. Technol., 2003, 217, 128.
23. G. Zingone and F. Rubessa, Int. J. Pharm., 2005, 291, 5.
24. H. B. Bohidar and M. Behboudnia, Colloids Surf., A, 2001, 178, 313.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c4ra14633f

---