Development of acyl-bonded aminopropyl silica gel fillers with low interfacial adhesion energy in a polystyrene matrix

Abstract

To develop fillers that adhere closely to a matrix polymer and reduce the glass transition temperature (T_g) of filler/polymer composites, N-acyl-3-aminopropyl silica gel fillers (1) were prepared by the reaction of 3-aminopropyl silica gel (SiO₂–NH₂) and succinimidyl alkanoate (2) by controlling the amount and length of alkyl chains. The surface tension (γ) of 1 was measured by the wicking method. The interfacial adhesion energy (W) of composites (1/PS) between 1 and polystyrene (PS) was calculated using the dispersive component (γ^D) and the polar component (γ^P) of γ, according to the Equation () where subscripts m and f indicate matrix and filler, respectively. The W values of 1/PS composites were considerably smaller than those of the SiO₂–NH₂/PS composite. A good correlation between W and T_g showed that the introduction of acyl groups on SiO₂ was effective for the enhancement of adhesion between 1 and the matrix polymer. Moreover, it was confirmed that the addition of less than 30 wt% of 1 to PS did not affect the tensile strength of composites.

Introduction

Modification of polymers with fillers is widely used to enhance the performance of polymers.^{1, 2} Inorganic materials such as talc,³ calcium carbonate,^{4, 5} glass,^{6, 7} silica,^{8, 9} alumina,¹⁰ zeolite^{11, 12, 13} and wollastonite¹⁴ have been widely used as fillers to improve mechanical performance such as the tensile strength and elongation strength of polymers. To enhance adhesion between inorganic fillers and the polymer matrix, organic additives such as maleic anhydride, carboxylic acids and silane-coupling reagents are mixed with the polymer.⁴ However, problems such as the dispersibility of organic additives in the polymer, the increase in glass transition temperatures (T_g) and the elution of organic additives remain in the manufacturing process.

Recently, alkyl group-bonded zeolite¹² and wollastonite¹⁴ fillers have been prepared to prevent the elution of organic additives from the polymer and to improve the handling of organic additives in the manufacturing process.¹⁵ Silica gel filler is a stable and safe material that has a wide surface area and the ability to adsorb various materials. 3-Aminopropyl silica gel (SiO₂–NH₂)^{16, 17} can immobilize functional groups through a covalent bond. To develop a high-performance filler that achieves higher adhesion and a lower T_g of filler/polymer composites, the surface tension of the acyl-bonded SiO₂–NH₂ filler (1) can be modified by controlling the amount and length of alkyl chains (Scheme 1).

Scheme 1

N-Acyl-3-aminopropyl silica gel fillers (1a–g).

Experimental procedure

Instruments

¹H-nuclear magnetic resonance (NMR) (400 MHz) and ¹³C-NMR (100 MHz) spectra were taken on a Bruker AV 400 M spectrometer (Bruker, Silberstrifen, Germany) for CDCl₃ solutions using SiMe4 as an internal standard. High-resolution mass spectra were taken on a Hitachi M2000A (Hitachi, Bremen, Germany) and a Bruker Daltonics Autoflex III TOF/TOF (Bruker) equipped with a matrix-assisted laser desorption/ionization apparatus. Fourier transform infrared was measured on a JASCO Herscel FT/IR-300 (JASCO, Hachioji, Japan) with a Micro-20 spectrometer. Scanning electron microscope images were recorded on a Hitachi S-4100.

Preparation of succinimidyl alkanoate (2)

A volume of 20 ml of CH₂Cl₂ solution containing alkanoic acid (0.5 mmol), N-hydroxy-succinimide (0.07 g, 0.6 mmol), N,N′-dicyclohexylcarbodiimide (0.12 g, 0.6 mmol) and 4-(dimethylamino)pyridine (0.06 g, 0.5 mmol) was stirred for 2 d at ambient temperature (Scheme 1). The reaction was monitored by TLC. The urea derived from N,N′-dicyclohexylcarbodiimide was removed by filtration and the solution was concentrated. The crude product was purified by silica gel column chromatography using CHCl₃–toluene–ethyl acetate (5:3:1) as the eluent and by repeated recrystallization from EtOH, resulting in 2 as a white solid.

Succinimidyl ethanoate (2a)

Yield 30.5%. Mp 131–133 °C. ¹H-NMR δ=2.35 (s, 3H), 2.84 (s, 4H), ¹³C-NMR δ=17.5, 25.6, 165.6, 169.1; HRMS found: 157.0358. HRMS Calcd. for C₆H₇NO₄: 157.0375 [M⁺].

Succinimidyl hexanoate (2b)

Yield 96.0%. Mp 34–37 °C. ¹H-NMR δ =0.91 (t, J=7.1 Hz, 3H), 1.34–1.41 (m, 4H), 1.71–1.79 (m, 2H), 2.60 (t, J=7.5 Hz, 2H), 2.83 (s, 4H); ¹³C-NMR δ=13.7, 22.1, 24.2, 25.4, 25.6, 30.8, 168.7, 169.2; HRMS found: 213.1015. Calcd. for C₁₀H₁₅NO₄: 213.1001 [M⁺].

Succinimidyl decanoate (2c)

Yield 70.5%. Mp 54–56 °C. ¹H-NMR δ=0.88 (t, J=6.8 Hz, 3H), 1.33–1.41 (m, 12H), 1.74 (quint, J=7.6 Hz, 2H), 2.60 (t, J=7.6 Hz, 2H), 2.83 (s, 4H); ¹³C-NMR δ=14.1, 22.6, 24.6, 25.6, 28.8, 29.1, 29.2, 29.3, 30.9, 31.8, 168.7, 169.1; HRMS found: 269.1656. Calcd. for C₁₄H₂₃NO₄: 269.1627 [M⁺].

Succinimidyl dodecanoate (2d)

Yield 82.4%. Mp 61–63 °C. ¹H-NMR; δ=0.88 (t, J=6.8 Hz, 3H), 1.26–1.42 (m, 16H), 1.70–1.78 (quint, J=8.0 Hz, 2H), 2.60 (t, J=7.2 Hz, 2H), 2.83 (s, 4H); ¹³C-NMR δ=14.1, 22.7, 24.6, 25.6, 28.8, 29.1, 29.3, 29.3, 29.5, 29.5, 30.9, 31.9, 168.7, 169.1. MALDI-MS found: 320.1835. Calcd. for C₁₆H₂₇NO₄Na: 297.1838 [M+Na⁺].

Succinimidyl tetraadecanoate (2e)

Yield 74.7%. Mp 81–83 °C. ¹H-NMR δ=0.88 (t, J=7.08 Hz, 3H), 1.26–1.42 (m, 20H), 1.70–1.78 (quint, J=7.4 Hz, 2H), 2.60 (t, J=7.44 Hz, 2H), 2.83 (s, 4H). ¹³C-NMR δ=14.1, 22.7, 24.6, 25.6, 25.6, 28.8, 29.1, 29.3, 29.5, 29.6, 29.6, 29.6, 30.9, 31.9, 168.7, 169.2. HRMS found: 325.2176. Calcd. for C₁₈H₃₁NO₄: 325.2253 [M⁺].

Succinimidyl hexadecanoate (2f)

Yield 62.0%. Mp 82–84 °C. ¹H-NMR δ=0.88 (t, J=6.8 Hz, 3H), 1.26–1.42 (m, 24H), 1.70–1.78 (quint, J=7.6 Hz, 24H), 2.60 (t, J=7.2 Hz, 2H), 2,83 (s, 4H); ¹³C-NMR δ=14.1, 22.7, 24.6, 25.6, 25.6, 28.8, 29.1, 29.3, 29.5, 29.6, 29.6, 29.7, 30.9, 31.9, 168.7, 169.1. MALDI-MS found: 376.2446. Calcd. for C₂₀H₃₅NO₄Na: 376.2464 [M+Na⁺].

Succinimidyl octadecanoate (2g)

Yield 88.0%. Mp 91–92 °C. ¹H-NMR δ =0.88 (t, J=6.9 Hz, 3H), 1.26–1.42 (m, 28H), 1.74 (quint, J=7.51 Hz, 2H), 2.60 (t, J=7.5 Hz, 2H), 2.83 (s, 4H); ¹³C NMR δ 14.1, 22.7, 24.6, 25.6, 28.8, 29.1, 29.3, 29.5, 29.6, 29.6, 29.6, 29.6, 29.6, 29.6, 29.6, 29.7, 30.9, 31.9, 168.7, 169.2; HRMS found: 381.2866. Calcd. for C₂₂H₃₉NO₄: 381.2879 [M⁺].

Quantitative analysis of 1 by infrared absorption spectrophotometry

For calibration lines, RCO₂H was mixed with SiO₂–NH₂ in a given ratio of M_R to W_S, where M_R and W_S denote the molar amounts of RCO₂H and the weight of SiO₂–NH₂, respectively. N-Benzylacetamide and N-phenylhexanamide were used instead of oily acetic acid and hexanoic acid. The M_R/W_S were plotted against the area ratio (A_R/A_S) of the C–H stretching of the RCO– group at 3000–2830 cm^–1 to the Si–O stretching of SiO₂–NH₂ at 800 cm^–1, which were obtained from the absorbance mode in the infrared (IR) spectra of the mixed samples. Compound 1 was ground on a mortar and subjected to IR absorption spectrophotometry. Quantitative analysis was performed using the A_R/A_S of 1 and the calibration line.

Measurement of contact angle

The surface tension (γ) of the solid is related to surface tensions in solid–liquid (γ_SL) and liquid–vapor (γ_LV) using the contact angle (θ), as shown in Figure 1. The surface tension γ of PS was determined from θ, which was measured at 23 °C by the sessile drop method¹⁴ using test liquids such as CH₂I₂, H₂O and formamide. The average θ values of five drops were taken 10 s after drop deposition.

Figure 1

Contact angles (θ) and surface tensions among vapor, liquid and solid surfaces.

The γ of 1 was measured by the wicking method reported by Kim and co-workers.¹² Glass capillaries with a 100 mm length and 1.0 mm diameter were plugged at the bottom with a small wad of cotton wool. Well-dried 1 (40 mg) was filled into the capillary up to a height of 70 mm. The penetration distance (h in cm) of the probe liquids through the column packed with 1 was measured at 23 °C in a given time (t in seconds). Parameter h can be related to t and θ by Washburn's Equation, Equation (1),^{18, 19, 20} using the effective interstitial pore radius (R in cm), viscosity (η in cP) and the interfacial tension of liquid–vapor (γ_LV) of the probe liquid (Electronic Supporting Information available: The following data are included: the properties of probe liquids such as hexane, heptane, octane, water, CH₂I₂ and formamide, the plots of h² versus time 1a as typical example, the plots of 2ηh²/t versus γ_LV for wicking with alkanes in the measurement of the surface tensions, and differential scanning calorimetry (DSC) profiles of 1/PS. See Supplementary Information).

When hexane, heptane and octane, which has a θ value of nearly zero, were used as spreading liquids, Equation (1) could be changed to Equation (2).¹² According to Equation (2), R was determined by the plot of h²/t against the γ_LV/(2η) of solvents. The R-values obtained for 1 are listed in Table 1. Using CH₂I₂, H₂O and formamide as the probe liquids with θ > 0, the plots of h² against t produced a straight line with the slope of Rγ_LV cosθ /(2η), which was plotted against Rγ_LV/(2η) to yield θ-values of 1 toward each of the solvents (Table 1).

Table 1 Effective interstitial pore radius (R), contact angle (θ), surface tension (γ), interfacial adhesion energy (W) and glass tension temperature (T_g) for acyl-bonded silica-gel microbeads

Calculation of surface tension components

According to Equation (3),¹² the dispersive Lifshitz–van der Waals component (γ^D) of 1 was determined using the cosθ value when CH₂I₂ (γ⁻_LV=γ⁺_LV=0, γ_LV=γ^D_LV=50.8 mJ m⁻²) was used as the penetration probe solvent. In addition, the Lewis acid (γ⁺) and Lewis base (γ⁻) components were determined using the cosθ values obtained from the measurements in H₂O and formamide, the surface tension components of which are shown in the supporting information (Electronic Supporting Information available: The following data are included: the properties of probe liquids such as hexane, heptane, octane, water, CH₂I₂ and formamide, the plots of h² versus time 1a as typical example, the plots of 2ηh²/t versus γ_LV for wicking with alkanes in the measurement of the surface tensions, and DSC profiles of 1/PS. See Supplementary Information). The polar component (γ^P) due to the Lewis acid–base contribution was calculated by Equation (4) using γ⁺ and γ⁻. The γ of 1 was obtained by the summation of γ^D and γ^P (Equation (5)). These results are shown in Table 1.

Preparation of 1/PS composites

A polystyrene pellet (PS; ρ=1.04–1, 05 g cm⁻³, Japan Polystyrene G590N) was dissolved in tetrahydrofuran. The given percents by weight % (1, 5, 10 and 30 wt%) of 1 were added to the tetrahydrofuran solution of the PS. The PS solution containing 1 was poured on a Teflon plate that was surrounded by a barrier and dried to yield the 1/PS composite film. As a reference, SiO₂/PS and SiO₂–NH₂/PS composites were prepared in a similar manner. The prepared composites were subjected to scanning electron microscope observation, DSC measurement and tensile testing.

Measurement of the tensile strength (σ) of composites

The 1/PS film was introduced into a handy injection molding machine (Shinko Sellbic, Japan), melted at a temperature of 260 °C, and injected into a dumbbell shape of 65 × 10 × 2 mm to fabricate the test piece of 1/PS composites. Similarly, a test piece of a composite of PS with SiO₂ (SiO₂/PS) was prepared as a reference material. Tensile tests of composites were carried out using a universal testing machine at a crosshead speed of 10 mm min⁻¹ at a temperature of 23 °C. Tensile strength (σ) was determined from stress-strain curves according to ASTM D638.

DSC Measurement of 1/PS composites.

A piece of 1/PS film (5.0 mg) was packed in an aluminum pan and sealed to prepare samples for DSC. In general, the DSC profile was dependent on the preheated temperature. After the samples were preheated up to 190 °C and then cooled to room temperature, DSC charts were recorded at a temperature-increasing rate of 10 °C min⁻¹ on a Rigaku Thermo plus DSC 8240 (Rigaku, Akishima, Japan). The T_g of 1/PS and SiO₂–NH₂/PS is listed in Table 1 and thermograms are shown in the supporting information (Electronic Supporting Information available: The following data are included: the properties of probe liquids such as hexane, heptane, octane, water, CH₂I₂ and formamide, the plots of h² versus time 1a as typical example, the plots of 2ηh²/t versus γ_LV for wicking with alkanes in the measurement of the surface tensions, and DSC profiles of 1/PS. See Supplementary Information).

Results and Discussion

Preparation of acyl-bonded SiO₂–NH₂ fillers (1)

N-Acyl-3-aminopropyl silica gels (1a–g) were prepared by a reaction of succinimidyl alkanoate (2a–g) with SiO₂–NH₂ (average diameter of beads: 8.3 μm; pore volume: 0.59 cm³ g⁻¹; amount of NH₂ (x_A): 1.50 mmol g⁻¹, Fuji Silysia Chemical, Kasugai, Japan) in the presence of imidazole for 3 days. The characteristic absorptions of the IR spectra of 1 were assigned by a comparison with that of N-propyl-4-octadecanamide (3), which was a reference sample prepared by a reaction of 2 g with n-PrNH₂ in the presence of imidazole (Figure 2). Absorption near 1560 cm^–1 can be assigned to an N–H bending of the amide bond (a₁). A C–H bending of the methyl group (a₂) and a C–H stretching mode (a₃) appeared at 1470 and 3000–2850 cm^–1, respectively. The SiO₂–NH₂ exhibited strong Si–O absorption at 800 cm⁻¹ (a₄), as well as at 1700–1660 and 1300–900 cm⁻¹.

Figure 2

IR spectra of 1a–c along with those of SiO₂–NH₂ and N-propyl-4-octadecanamide (3) in the region 600–3800 cm⁻¹. a₁: The N–H bending of the amide bond near 1560 cm⁻¹; a₂: the bending mode of C–H at 1470 cm⁻¹; a₃: the stretching mode of C–H at 3000–2830 cm⁻¹; and a₄: the Si–O stretching mode at 800 cm⁻¹.

The amount of acyl groups (x_R/mmol g^–1) that bonded to SiO₂–NH₂ in 1 was analyzed by IR absorption spectrophotometry at a₃ and a₄. Figure 3 shows plots of x_R versus a given amount of 2b, 2c and 2 g used in the preparation of 1. As the amount of 2 increased, the x_R of 1b, 1c and 1 g increased gradually to reach a maximum point near 1.50 mmol g⁻¹ of 2, which was equal to x_A (=1.50 mmol g⁻¹). At the maximum point, the yields (100x_R/x_A) based on x_A were determined to be 56, 52 and 44% for 1b, 1c and 1 g, respectively. Similarly, the maximum yields of 1a, 1d, 1e and 1f were 31, 59, 58 and 54 %, respectively.

Figure 3

Plots of the amount of acyl groups (x_R) versus the amount of 2b (○), 2c (△) and 2 g (□) used in the reaction of 2 with SiO₂–NH₂.

Morphology and tensile strength (σ) of 1/PS composites

Scanning electron microscope images of the 1/PS and SiO₂/PS composites were observed to examine the adhesion of 1 with the PS matrix (Figure 4). In the scanning electron microscope images of the cross-section of SiO₂/PS composites, a void appeared at the interface between SiO₂ particles and the PS matrix, showing that the adhesion of SiO₂ toward PS was poor. In the case of 1/PS, on the other hand, 1 was well dispersed over the PS matrix, and the PS matrix adhered closely to 1 compared with SiO₂/PS. With an increase in the alkyl chain length from C₆ to C₁₈, the void decreased, resulting in the higher adhesion of 1 in PS.

Figure 4

The scanning electron microscope images of the cross-sections of 1/PS composites and SiO₂/PS composites under × 700 (lower) and × 4000 (upper) magnifications.

The tensile strength (σ) of the SiO₂/PS composite decreased to 37.3 MPa from 46.0 MPa of PS when 30 wt% of SiO₂ was added to the PS (Figure 5). In the case of 1/PS composites, on the other hand, the addition of 30 wt% of 1 maintained the σ-values of 44–45 MPa the same as the σ-value of PS. Thus, the introduction of acyl groups on SiO₂ was effective for the enhancement of adhesion between 1 and the matrix polymer.

Figure 5

The additive effect of SiO₂ (•), 1a (○), 1b (△) and 1c (□) on the tensile strength (σ) of PS.

Interfacial adhesion energy between the filler and matrix polymer

The interfacial adhesion energy (W) between the filler and the matrix polymer is one of the most important properties of composites.¹⁴ Švab et al.¹⁴ have postulated that γ^D and γ^P can be regarded in the solid state as the dispersive and polar components of the surface energy, respectively. Moreover, they proposed that W can be represented by Equation (6) using the surface free energy (surface tension) of the filler (γ_f) and matrix (γ_m) and the interfacial free energy (γ_mf) between the filler and matrix polymer. The γ_mf was obtained by Equation (7) using components γ^D and γ^P of the filler and matrix polymer.^{14, 21} Therefore, W-values can be calculated according to Equation (8), where subscripts m and f indicate matrix and filler, respectively.

The W values of the 1/PS composites were calculated according to Equation (8) using the γ^D and γ^P of 1 and PS listed in Table 1. The dependence of W values on x_R was examined in the cases of 1b, 1c and 1 g (Figure 6a). With an increase in x_R, W values decreased gradually. In addition, W values decreased as the carbon number (n) on the acyl group of 1 increased (Figure 6b). As shown in Table 1, the γ of 1 was strongly affected by the γ^D component, whereas the polar γ^P contributions were small because of the hydrophobic acyl chains on 1. On the other hand, SiO₂–NH₂ had relatively large values of γ^P. These results showed that the interaction energy between 1 and the hydrophobic PS matrix was weaker, as the hydrophobic character of 1 increased because of increases in x_R and n.

Figure 6

(a) Dependence of x_R on W in 1b (○), 1c (△) and 1 g (□) along with SiO₂–NH₂ (•). (b) Dependence of carbon number (n) on W.

Relationship between W and T_g

The T_g is one of the most important characteristics of polymer materials. T_g is used as a parameter to evaluate the adhesion of filler toward the matrix.^{22, 23} The T_g values of 1/PS and SiO₂–NH₂/PS composites measured by DSC are shown in Table 1. The T_g of 1/PS composite was lower than that of the SiO₂–NH₂/PS composite. Moreover, T_g decreased with an increase in x_R of 1b, 1c and 1g. Figure 7 shows the correlation between T_g and W, revealing that the introduction of acyl groups on SiO₂ was effective to enhance the adhesion between 1 and the matrix polymer.

Figure 7

Relationship between T_g and W: SiO₂–NH₂/PS (•), 1a/PS (▴), 1b/PS (○), 1c/PS (△), 1d/PS (▪), 1e/PS (♦), 1f/PS (▾) and 1g/PS (□).

In conclusion, intrafacial adhesion energy (W) can be modified by γ, which is easily controlled by the amount and length of hydrophobic alkyl chains. It was found that lower W values led to close adhesion and lower T_g.