*3.4. Characterization Methods*

In order to examine the microstructure of the hybrid specimens the samples were analyzed using a Zeiss Neon 40 High Resolution Scanning Electron Microscope (SEM). Images were acquired at diverse magnifications while the microscope was operated at 10 or 20 kV. Energy Dispersive Spectroscopy (EDS) experiments were conducted in conjunction with the SEM using the EDAX equipment with an Apollo 10 silicon drift detector (SDD). Data was collected and analyzed using Genesis Spectrum software.

A Netzsch STA 449 FE Jupiter, operated in a Temperature Programmed Oxidation (TPO) mode, was used to study the thermal stability of the samples. The samples were exposed to an Ar/O2, 80%/20% atmosphere, with a total flow of 120 mL minute<sup>í</sup><sup>1</sup> , from RT to 1000 °C at a heating rate of 10 °C minute<sup>í</sup><sup>1</sup> .

The XRD utilized was a Philips 1830 PAnalytical X-ray Diffractometer. The X-ray tube contained a copper source and the X-rays utilized had a primary wavelength, or K-Alpha, of 1.54 Å. The samples were placed into a silicon low background sample holder and the diffraction patterns recorded between 5–70° (2 theta) with 0.020 degrees step size and one second per step.

A JEOL 2010F FASTEM field emission gun scanning transmission electron microscope (STEM/TEM) equipped with Gatan GIF image filtering system was employed. Samples were prepared by dispersing the powders in a few ml of ethanol and a drop of the dispersion was placed in a copper holey-carbon TEM grid where the ethanol was allowed to evaporate.

A Perkin Elmer ICP 5300 DV-AES Inductive Coupled Plasma Emission Spectrometer, was used to determine the elemental composition of the carbon nanofiber base filler materials before their addition to the epoxy matrix (hybrids CNF/IF-WS2).

Brunauer Emmet Teller (BET) surface area analysis was performed employing a Quantachrome Nova 4200. A 300 °C degas step was conducted prior to the analysis; samples were then allowed to cool down to room temperature and then transferred to the analysis station. The measurements were done using nitrogen atmosphere.

Nanoindentation was used to measure the composite mechanical properties (elastic modulus and hardness) of epoxy composites filled with mixtures of CNFs and IF-WS2 particles. The samples were prepared by mixing the specified amounts of nanophase material with Struers Speci-Fix 20 two-part epoxy in a 28 mm diameter mold and then allowing the mixture to cure for 24 h. After curing, the surface of the epoxy composite was polished using standard metallographic techniques, including diamond suspension polishing using suspended aluminum oxide particles of 1 ȝm and 0.05 ȝm diameters. The indentations were performed using an Agilent G200 nanoindenter. We performed two types of experiments with this instrument.

The first experiment was a quasi-static indentation to a set depth, 2 ȝm for all samples. Other indentation parameters can be found in Table 2. This experiment used a diamond, Berkovich indenter tip with a nominal tip radius of 150 nm, calibrated using a fused silica standard. A grid of 20 indentation points spaced by 50 ȝm was measured for each epoxy nanocomposite. The Young's modulus and hardness were calculated using the approach of Oliver and Pharr [69,70].

The second experiment was dynamic mechanical analysis using a 50 ȝm diameter flat punch. This experiment allowed the measurement of both the storage and loss moduli of the epoxy composites. These measurements were performed for five frequency values between 1 Hz and 45 Hz on the neat epoxy, CNF and IF-WS2 samples. Other parameters for the measurement can be found in Table 3. A grid of 20 measurements with a 100 ȝm separation between indentations was used for each specimen. The storage modulus, loss modulus, and tan į properties for each specimen were calculated using the measurement parameters in Table 3 and the methods of Hay and Herbert [71].


**Table 2.** Parameters used for quasi-static indentation measurements.

**Table 3.** Parameters used for dynamic mechanical analysis.


### **4. Conclusions**

Novel hybrid CNF/IF-WS2 with diverse IF loadings were generated using an *in situ* protocol that allowed the integration of the two phases into a tridimensional architecture, producing homogeneous dispersions at the nanoscale in the absence of a polymeric matrix. CNF 3D structures loaded with IF-WS2 could only be fabricated using a two stage process that involved: (a) the carbon nanofiber growth from a mixture of metal catalyst with tungsten oxide nanoparticles, using ethylene as carbon source and moderate temperatures to render CNF/WO3, followed by (b) the sulfurization of the sample to convert the tungsten precursor into IF-WS2.

In contrast, Graphene/IF-WS2 hybrids were easily obtained either by mixing graphene and tungsten oxide followed by a sulfurization step or by direct dispersion of the layered graphene structure with existing IF particles using solvents.

The thermal stability of the CNF/IF-WS2 hybrid samples is much higher than those observed for IF-WS2 by itself or mixed with polymeric components.

Epoxy composites with 1% weight loadings of hybrid CNF/IF-WS2 showed drastic improvements in the Young's modulus and hardness values, with approximately 100 and 250% increase respectively, over the bare epoxy values. The CNF/IF-WS2 inclusions seem to have a much greater impact in the mechanical properties of the composite than the Graphene/IF-WS2 based counterparts.
