n-[3-(Triethoxysilyl)propyl]-2-carbomethoxyaziridine

• CAS: 193417-26-4
• Molecular Formula: C13H27NO5Si
• Molecular Weight: 305.446
• SMILES: CCO[Si](CCCN1CC1C(=O)OC)(OCC)OCC
• InChI: InChI=1S/C13H27NO5Si/c1-5-17-20(18-6-2,19-7-3)10-8-9-14-11-12(14)13(15)16-4/h12H,5-11H2,1-4H3
• InChIKey: BDUJZVMOZLQWNL-UHFFFAOYSA-N

N-[3-(Triethoxysilyl)propyl]-2-carbomethoxyaziridine (CAS 193417-26-4) is a bifunctional organosilane that combines a hydrolyzable triethoxysilyl anchoring group with a 2-carbomethoxy-substituted aziridine ring, linked through a three-carbon propyl spacer. The compound serves as a precursor to fulleropyrrolidine-functionalized silica HPLC stationary phases via thermal aziridine ring-opening and 1,3-dipolar cycloaddition with C60, a transformation that proceeds with 36% isolated yield (71% based on C60 conversion).

N-[3-(Triethoxysilyl)propyl]-2-carbomethoxyaziridine (CAS 193417-26-4) — Procurement-Grade Organosilane Aziridine for Fullerene-Functionalized HPLC Stationary Phases and Elastomer Coupling

N-[3-(Triethoxysilyl)propyl]-2-carbomethoxyaziridine (CAS 193417-26-4) is a bifunctional organosilane that combines a hydrolyzable triethoxysilyl anchoring group with a 2-carbomethoxy-substituted aziridine ring, linked through a three-carbon propyl spacer [1]. The compound serves as a precursor to fulleropyrrolidine-functionalized silica HPLC stationary phases via thermal aziridine ring-opening and 1,3-dipolar cycloaddition with C60, a transformation that proceeds with 36% isolated yield (71% based on C60 conversion) [1]. It is also disclosed as a member of a broader class of aziridine-containing hydrolysable silanes useful as coupling agents for diene elastomer compositions, where the aziridine ring reacts with elastomer C=C bonds through cycloaddition [2]. Key physicochemical identifiers include molecular formula C13H27NO5Si, molecular weight 305.44 g/mol, and characteristic GC-MS spectral signatures catalogued in the Wiley SpectraBase library [3].

• [1] Bianco, A.; Gasparrini, F.; Maggini, M.; Misiti, D.; Polese, A.; Prato, M.; Scorrano, G.; Toniolo, C.; Villani, C. Molecular Recognition by a Silica-Bound Fullerene Derivative. J. Am. Chem. Soc. 1997, 119 (32), 7550–7554. DOI: 10.1021/ja970502l View Source
• [2] Backer, M.; Chosse, T.; De Bever, O.; Grofils, S. Hydrolysable Silanes. U.S. Patent Application Publication US 2014/0364627 A1, December 11, 2014. (Also published as WO 2010/139473 A1.) View Source
• [3] Wiley SpectraBase. N-[3-(Triethoxysilyl)propyl]-2-carbomethoxy aziridine, Compound ID DNJTHBoh9lI. Spectra: 2 MS (GC). John Wiley & Sons, Inc. View Source

Why Generic Silane Coupling Agents Cannot Replace N-[3-(Triethoxysilyl)propyl]-2-carbomethoxyaziridine for Fullerene-Immobilized Chromatographic Phases

Conventional silane coupling agents such as 3-aminopropyltriethoxysilane (APTES), 3-glycidoxypropyltrimethoxysilane (GPTMS), or 3-(triethoxysilyl)propyl isocyanate (TEPI) lack the aziridine-2-carboxylate pharmacophore that enables thermal generation of an azomethine ylide for stereospecific 1,3-dipolar cycloaddition with C60 [1]. The earlier chlorosilane-C60 approach (Nagashima et al., J. Org. Chem. 1995) required pre-synthesis of a chlorosilane-fullerene adduct as a separate intermediate before silica attachment, whereas the aziridine-carbomethoxy architecture of 193417-26-4 permits a convergent, one-pot strategy: first, fulleropyrrolidine formation in solution, then direct covalent grafting to silica microparticles via the triethoxysilyl terminus [1][2]. Closely related aziridine-silanes such as 1-(2-triethoxysilylethyl)aziridine (CAS 18297-22-8) or 1-[3-(trimethoxysilyl)propyl]aziridine lack the 2-carbomethoxy substituent and consequently cannot generate the stabilized azomethine ylide required for efficient fullerene cycloaddition [3]. Furthermore, the triethoxysilyl hydrolysis rate is substantially slower than that of trimethoxysilyl analogs — literature evidence demonstrates that trimethoxysilyl-terminated polymers are 10–30-fold more reactive toward hydrolytic condensation than their triethoxysilyl counterparts — offering the triethoxy variant superior shelf stability and more controllable surface-modification kinetics in moisture-sensitive workflows [4].

• [1] Bianco, A.; Gasparrini, F.; Maggini, M.; Misiti, D.; Polese, A.; Prato, M.; Scorrano, G.; Toniolo, C.; Villani, C. Molecular Recognition by a Silica-Bound Fullerene Derivative. J. Am. Chem. Soc. 1997, 119 (32), 7550–7554. View Source
• [2] Nagashima, H.; Terasaki, H.; Saito, Y.; Jinno, K.; Itoh, K. Chlorosilanes and Silyl Triflates Containing C60 as a Partial Structure. A Versatile Synthetic Entry Linking the C60 Moieties with Alcohols, Phenols, and Silica. J. Org. Chem. 1995, 60 (16), 4966–4967. View Source
• [3] Backer, M.; Chosse, T.; De Bever, O.; Grofils, S. Hydrolysable Silanes. U.S. Patent Application Publication US 2014/0364627 A1, December 11, 2014. (Comparative examples of aziridine-silane structures lacking 2-carbomethoxy substitution.) View Source
• [4] Kricheldorf, H. R.; Behnken, G. Telechelic polylactones functionalized with trimethoxysilyl groups. Polymer 2005, 46 (24), 10782–10787. (Reports 10–30-fold higher reactivity of Si(OMe)₃ vs. Si(OEt)₃ in crosslinking experiments.) View Source

Quantitative Differentiation Evidence for N-[3-(Triethoxysilyl)propyl]-2-carbomethoxyaziridine vs. Closest Analogs and Alternative Approaches

Fulleropyrrolidine-Functionalized Silica Loading and Surface Coverage vs. Chlorosilane-C60 Approach

The target compound enables preparation of fulleropyrrolidine-grafted HPLC silica (designated 'phase 3') with a quantified C60 loading of 0.07 mmol fulleropyrrolidine per gram of silica (72 mg/g), corresponding to a surface coverage of 0.4 µmol/m², as determined by elemental analysis based on 5.86% carbon content [1]. In contrast, the prior chlorosilane-C60 approach (Nagashima et al., 1995) required the separate synthesis of chlorosilane-C60 adducts 3 and 4, followed by reaction with silica in refluxing toluene with pyridine — a multi-step protocol for which comparable quantitative loading and surface-coverage metrics were not reported in the original disclosure [2]. The Bianco method achieves chemically homogeneous coverage in a single grafting step from a preformed, well-characterized trialkoxysilane-fulleropyrrolidine intermediate [1].

Evidence Dimension	C60 loading on silica (fulleropyrrolidine per gram)
Target Compound Data	0.07 mmol/g (72 mg/g); surface coverage 0.4 µmol/m²
Comparator Or Baseline	Nagashima chlorosilane-C60 method: multi-step protocol; quantitative loading and coverage data not reported for direct comparison
Quantified Difference	Quantified loading achieved for target; no comparable metric available for the chlorosilane-C60 approach
Conditions	Elemental analysis (5.86% C) of vacuum-dried silica 3; HPLC spherical silica particles, refluxing toluene

Why This Matters

For procurement decisions involving C60-functionalized chromatographic supports, the documented loading density of 0.07 mmol/g provides a verifiable benchmark absent from the earlier chlorosilane-based method, enabling reproducible column preparation.

• [1] Bianco, A.; Gasparrini, F.; Maggini, M.; Misiti, D.; Polese, A.; Prato, M.; Scorrano, G.; Toniolo, C.; Villani, C. J. Am. Chem. Soc. 1997, 119 (32), 7550–7554. View Source
• [2] Nagashima, H.; Terasaki, H.; Saito, Y.; Jinno, K.; Itoh, K. J. Org. Chem. 1995, 60 (16), 4966–4967. View Source

Size-Selective Calixarene Recognition on Fulleropyrrolidine-Silica Phase vs. Bare Silica Baseline

HPLC column packed with the fulleropyrrolidine-silica phase derived from 193417-26-4 exhibits extraordinary size-selectivity for tert-butylcalix[n]arenes in CH₂Cl₂/i-PrOH (99.5/0.5) eluent, with a marked preference for calix[8]arene over the four- and six-membered oligomers [1]. In 100% toluene, retention times were 2.0 min (calix[4]arene), 7.3 min (calix[6]arene), and 7.9 min (calix[8]arene) [1]. The selectivity is entirely attributable to the grafted fulleropyrrolidine: on bare silica under identical CH₂Cl₂/i-PrOH conditions, all calixarenes are front-eluted with no differential retention [1]. By comparison, the chlorosilane-derived C60-bonded phase reported by Jinno et al. exhibited lower retention for calixarenes than conventional ODS phases, indicating weaker host–guest interactions with that stationary phase architecture [2].

Evidence Dimension	Calixarene retention selectivity (capacity factor differential)
Target Compound Data	tR calix[4]arene 2.0 min, calix[6]arene 7.3 min, calix[8]arene 7.9 min (toluene); marked calix[8]arene preference (CH₂Cl₂/i-PrOH)
Comparator Or Baseline	Bare silica: all calixarenes front-eluted, zero selectivity. Jinno C60-bonded phase: retention lower than ODS, weak host–guest interaction
Quantified Difference	Calix[8]/calix[4] retention time ratio ≈ 4:1 on target-derived phase; no measurable selectivity on bare silica
Conditions	250 × 1.8 mm stainless steel column, Hypersil 5 µm silica, UV detection at 280 nm, flow rate 0.3 mL/min (toluene); CH₂Cl₂/i-PrOH (99.5/0.5) for size-selectivity experiment

Why This Matters

This compound is the only documented precursor to a fulleropyrrolidine HPLC phase with demonstrated calixarene size-selectivity, a capability that cannot be replicated with generic C18, ODS, or even earlier C60-bonded phases for supramolecular host–guest screening.

• [1] Bianco, A.; Gasparrini, F.; Maggini, M.; Misiti, D.; Polese, A.; Prato, M.; Scorrano, G.; Toniolo, C.; Villani, C. J. Am. Chem. Soc. 1997, 119 (32), 7550–7554. View Source
• [2] Jinno, K.; Tanabe, K.; Saito, Y.; Nagashima, H.; Trengove, R. D. Retention Behavior of Calixarenes with Various C60 Bonded Silica Phases in Microcolumn Liquid Chromatography. J. High Resolut. Chromatogr. 1997, 20 (8), 451–456. View Source

Aqueous-Media Cyclodextrin Baseline Separation: Fulleropyrrolidine Phase vs. Conventional Reversed-Phase Columns

The fulleropyrrolidine-silica column derived from 193417-26-4 achieves baseline resolution of α-, β-, and γ-cyclodextrins in water-rich media (H₂O/MeOH/THF 80/10/10 for 3 min, then linear gradient to 40/30/30 in 25 min), as confirmed by evaporative light scattering detection [1]. Column efficiency was measured at N/m = (4–5) × 10⁴ for simple aromatic solutes in both organic and water-rich eluents, demonstrating dual-mode (reversed-phase-like) retention behavior [1]. Conventional reversed-phase C18 columns do not exhibit size-based cyclodextrin selectivity under comparable conditions, as cyclodextrins typically co-elute or show minimal differential retention on alkyl-bonded phases [2]. This performance is a direct consequence of the curved fulleroid surface of the immobilized C60, which enables multipoint contacts with cyclic oligosaccharides that are inaccessible to planar or alkyl-chain stationary phases.

Evidence Dimension	Cyclodextrin isomer separation (α-CD, β-CD, γ-CD) and column efficiency
Target Compound Data	Baseline separation of α-, β-, γ-cyclodextrins; N/m = (4–5) × 10⁴ theoretical plates per meter
Comparator Or Baseline	Conventional ODS/C18 columns: no size-based cyclodextrin selectivity; co-elution typically observed
Quantified Difference	Three cyclodextrin isomers baseline-resolved on target-derived phase; 0 achieved on standard C18 for size-based separation
Conditions	250 × 1.8 mm column, water/methanol/THF gradient, 0.25 mL/min, 25 °C, evaporative light scattering detector

Why This Matters

Procurement of this compound uniquely enables fabrication of HPLC columns capable of separating cyclodextrin homologues in aqueous media — a capability not offered by any commercial alkyl-bonded silica phase, making it essential for cyclodextrin quality control and supramolecular host screening.

• [1] Bianco, A.; Gasparrini, F.; Maggini, M.; Misiti, D.; Polese, A.; Prato, M.; Scorrano, G.; Toniolo, C.; Villani, C. J. Am. Chem. Soc. 1997, 119 (32), 7550–7554. View Source
• [2] Szente, L.; Szemán, J. Cyclodextrins in Analytical Chemistry: Host–Guest Type Molecular Recognition. Anal. Chem. 2013, 85 (17), 8024–8030. (Documents limited cyclodextrin discrimination on conventional RP phases.) View Source

Fulleropyrrolidine Cycloaddition Yield: Aziridine-2-Carboxylate Route vs. Alternative Azomethine Ylide Generation Methods

Thermal ring-opening of N-[3-(triethoxysilyl)propyl]-2-carbomethoxyaziridine in the presence of C60 in refluxing toluene produces fulleropyrrolidine derivative 2 in 36% isolated yield (71% based on unconverted C60), as confirmed by IR, UV-vis, MALDI-MS, ¹H and ¹³C NMR, and elemental analysis [1]. This yield is achieved via a convergent strategy: the aziridine serves simultaneously as the azomethine ylide precursor and carries the triethoxysilyl group for subsequent silica grafting, eliminating the need for a separate aldehyde or N-substituted glycine component required in the classical Prato reaction [1][2]. The classical Prato method (sarcosine + aldehyde + C60) typically yields fulleropyrrolidines in 25–55% isolated yield depending on the aldehyde, but requires post-synthetic introduction of a silyl anchoring group for surface immobilization [2].

Evidence Dimension	Fulleropyrrolidine synthesis: isolated yield and step count to silylated product
Target Compound Data	36% isolated yield (71% based on C60 conversion); single-step conversion to silylated fulleropyrrolidine ready for silica grafting
Comparator Or Baseline	Classical Prato reaction: 25–55% yield; requires separate silylation step for surface attachment
Quantified Difference	Comparable or superior net efficiency when accounting for the elimination of a separate silylation step; 71% yield based on recovered C60 indicates efficient fullerene utilization
Conditions	Refluxing toluene, C60, 1.2 equiv aziridine precursor, thermal ring-opening/cycloaddition

Why This Matters

For laboratories synthesizing fullerene-functionalized materials, this compound provides the most direct route to a silica-graftable fulleropyrrolidine, reducing synthetic steps from two to one relative to the classical Prato pathway.

• [1] Bianco, A.; Gasparrini, F.; Maggini, M.; Misiti, D.; Polese, A.; Prato, M.; Scorrano, G.; Toniolo, C.; Villani, C. J. Am. Chem. Soc. 1997, 119 (32), 7550–7554. View Source
• [2] Maggini, M.; Scorrano, G.; Prato, M. Addition of Azomethine Ylides to C60: Synthesis, Characterization, and Functionalization of Fullerene Pyrrolidines. J. Am. Chem. Soc. 1993, 115 (21), 9798–9799. View Source

Triethoxysilyl Hydrolysis Kinetics: Controlled Surface Modification vs. Trimethoxysilyl Analogs

The triethoxysilyl terminus of 193417-26-4 hydrolyzes substantially slower than the trimethoxysilyl group present in closely related aziridine-silanes such as 1-[3-(trimethoxysilyl)propyl]aziridine. Quantitative crosslinking studies on telechelic polymers have established that trimethoxysilyl end-groups are 10–30-fold more reactive toward hydrolytic condensation than their triethoxysilyl counterparts [1]. This kinetic differential is functionally significant: the slower hydrolysis of triethoxysilanes permits longer pot life in coating and surface-modification formulations, reduces premature gelation during silica grafting reactions, and provides more uniform monolayer coverage on hydroxylated surfaces [1][2]. The patent literature explicitly notes that ethoxy is the most preferred alkoxy leaving group for hydrolysable aziridine-silanes due to favorable hydrolysis kinetics relative to methoxy variants [2].

Evidence Dimension	Hydrolysis/condensation rate of alkoxysilyl group
Target Compound Data	Triethoxysilyl (Si(OEt)₃): slower, controlled hydrolysis kinetics suitable for reproducible surface grafting and longer formulation pot life
Comparator Or Baseline	Trimethoxysilyl analogs (e.g., 1-[3-(trimethoxysilyl)propyl]aziridine): 10–30-fold faster hydrolysis/condensation
Quantified Difference	Trimethoxysilyl 10–30× more reactive than triethoxysilyl in crosslinking experiments [1]
Conditions	Oligo(ethylene glycol) crosslinking, polymer end-group reactivity comparison (literature class-level data)

Why This Matters

Procurement of the triethoxy variant (193417-26-4) over a trimethoxy analog is justified when reproducible, controlled surface modification is required, as the slower hydrolysis reduces batch-to-batch variability in grafting density and minimizes premature condensation.

• [1] Kricheldorf, H. R.; Behnken, G. Telechelic polylactones functionalized with trimethoxysilyl groups. Polymer 2005, 46 (24), 10782–10787. View Source
• [2] Backer, M.; Chosse, T.; De Bever, O.; Grofils, S. Hydrolysable Silanes. U.S. Patent Application Publication US 2014/0364627 A1, December 11, 2014. (¶ [0054]: ethoxy preferred as most preferred alkoxy group.) View Source

Diene Elastomer Coupling: Aziridine-C=C Cycloaddition Mechanism vs. Conventional Sulfur-Silane or Epoxy-Silane Coupling Agents

Patent disclosure WO 2010/139473 (US 2014/0364627) establishes that hydrolysable silanes bearing an aziridine ring — including the class to which 193417-26-4 belongs — bond to diene elastomers through a cycloaddition mechanism between the aziridine ring and the elastomer C=C bonds under typical rubber compounding conditions [1]. This is mechanistically distinct from conventional sulfur-functional silanes (e.g., bis[3-(triethoxysilyl)propyl]tetrasulfide, TESPT) that couple via sulfidic bridges, and from epoxy-silanes (e.g., GPTMS) that require nucleophilic ring-opening. The aziridine-based coupling is described as 'strong' and proceeds through covalent C–C and C–N bond formation with the polymer backbone, whereas sulfur-silane coupling efficiency depends on vulcanization conditions and can be compromised by sulfur blooming or reversion [1][2]. A related patent (US 2014/0329976) further discloses that aziridine-containing silanes of this class bond to carbon-based fillers (carbon black, carbon nanotubes, graphene) through the same aziridine-C=C cycloaddition mechanism [2].

Evidence Dimension	Coupling mechanism and bond type to diene elastomer
Target Compound Data	Covalent C–C/C–N bond formation via aziridine cycloaddition to elastomer C=C; independent of sulfur vulcanization chemistry
Comparator Or Baseline	TESPT (sulfur-silane): sulfidic bridge coupling, dependent on vulcanization kinetics and subject to reversion. GPTMS (epoxy-silane): nucleophilic ring-opening required
Quantified Difference	Mechanistic differentiation only; quantitative coupling efficiency data (e.g., bound rubber content) not reported in patent for this specific compound
Conditions	Elastomer compounding at elevated temperatures; patent scope covers diene elastomer compositions containing inorganic filler

Why This Matters

For formulators of silica-filled diene rubber compounds (e.g., tire tread), the aziridine-based coupling mechanism offers a vulcanization-independent pathway for filler–polymer bonding, potentially reducing the sulfur-silane loading required and mitigating cure-reversion issues.

• [1] Backer, M.; Chosse, T.; De Bever, O.; Grofils, S. Hydrolysable Silanes. U.S. Patent Application Publication US 2014/0364627 A1, December 11, 2014. View Source
• [2] Backer, M.; Chosse, T.; De Bever, O.; Grofils, S. Treatment of Filler with Silane. U.S. Patent Application Publication US 2014/0329976 A1, November 6, 2014. View Source

Procurement-Driven Application Scenarios for N-[3-(Triethoxysilyl)propyl]-2-carbomethoxyaziridine (CAS 193417-26-4)

Fabrication of Fullerene-Immobilized HPLC Stationary Phases for Supramolecular Host–Guest Screening

This compound is the only documented precursor for preparing a covalently bound fulleropyrrolidine HPLC stationary phase that exhibits size-selective recognition of calixarenes (calix[8]arene retention time 7.9 min vs. calix[4]arene 2.0 min in toluene) and baseline separation of α-, β-, and γ-cyclodextrins in aqueous media at column efficiencies of N/m = (4–5) × 10⁴ [1]. The prepared phase achieved a verified C60 loading of 0.07 mmol/g silica (0.4 µmol/m² surface coverage) with thermal stability to 950 °C [1]. This application is not addressable by generic silane coupling agents (APTES, GPTMS, TEPI) or by earlier C60-bonded phases derived from chlorosilane-C60 adducts, which lack comparable size-selectivity for cyclic oligomers [1][2].

One-Pot Synthesis of Silica-Graftable Fulleropyrrolidine Derivatives for Materials Science

Researchers synthesizing fullerene-polymer hybrids, sol-gel fullerene glasses, or fullerene-functionalized nanoparticles benefit from the convergent reactivity of 193417-26-4: thermal aziridine ring-opening in the presence of C60 yields the fulleropyrrolidine (36% isolated, 71% based on C60 conversion), with the triethoxysilyl group already installed for immediate grafting to silica, metal oxides, or glass surfaces [1]. This eliminates the separate silylation step required when using the classical Prato azomethine ylide method (sarcosine/aldehyde route), reducing total synthetic steps and improving atom economy [1][3].

Silica-Filled Diene Elastomer Composites with Vulcanization-Independent Filler–Polymer Coupling

In tire tread and industrial rubber formulations, aziridine-functional silanes of the class represented by 193417-26-4 bond to diene elastomers through aziridine-C=C cycloaddition, creating covalent C–C and C–N linkages that are independent of the sulfur vulcanization system [4]. This mechanism is orthogonal to conventional sulfur-silane coupling (e.g., TESPT), potentially enabling reduced sulfur-silane loading, lower hysteresis (tan δ), and improved reversion resistance in silica-filled natural rubber and SBR compounds [4][5]. The triethoxysilyl group bonds to silica filler through hydrolysis/condensation, while the aziridine simultaneously reacts with the polymer backbone during compounding [4].

Controlled Surface Functionalization of Carbon-Based Fillers (Carbon Black, CNTs, Graphene)

Patent US 2014/0329976 discloses that aziridine-containing hydrolysable silanes of the formula G-OC(O)-(Az)-J, encompassing 193417-26-4, bond to carbon-based fillers through aziridine cycloaddition with surface C=C unsaturation [5]. This enables direct covalent modification of carbon black, carbon fiber, carbon nanotubes, graphene, and expandable graphite without requiring pre-oxidation or plasma treatment to introduce surface hydroxyl groups [5]. The triethoxysilyl terminus subsequently permits dispersion in or coupling to polymer matrices, addressing a long-standing challenge in carbon filler surface engineering where conventional silanes require hydroxylated surfaces for attachment [5].

• [1] Bianco, A.; Gasparrini, F.; Maggini, M.; Misiti, D.; Polese, A.; Prato, M.; Scorrano, G.; Toniolo, C.; Villani, C. Molecular Recognition by a Silica-Bound Fullerene Derivative. J. Am. Chem. Soc. 1997, 119 (32), 7550–7554. View Source
• [2] Jinno, K.; Tanabe, K.; Saito, Y.; Nagashima, H.; Trengove, R. D. Retention Behavior of Calixarenes with Various C60 Bonded Silica Phases in Microcolumn Liquid Chromatography. J. High Resolut. Chromatogr. 1997, 20 (8), 451–456. View Source
• [3] Maggini, M.; Scorrano, G.; Prato, M. Addition of Azomethine Ylides to C60: Synthesis, Characterization, and Functionalization of Fullerene Pyrrolidines. J. Am. Chem. Soc. 1993, 115 (21), 9798–9799. View Source
• [4] Backer, M.; Chosse, T.; De Bever, O.; Grofils, S. Hydrolysable Silanes. U.S. Patent Application Publication US 2014/0364627 A1, December 11, 2014. View Source
• [5] Backer, M.; Chosse, T.; De Bever, O.; Grofils, S. Treatment of Filler with Silane. U.S. Patent Application Publication US 2014/0329976 A1, November 6, 2014. View Source