Tip Sonication Power and Time Govern Graphite Exfoliation into Graphene Nanoplatelets: A Quantitative Study

Background

Graphene’s exceptional mechanical, thermal, optical, and electronic properties—Young’s modulus ≈ 1 TPa, thermal conductivity ≈ 5000 W m^–1 K^–1, optical transmittance 97.7 %, intrinsic mobility ≈ 200 000 cm² V^–1 s^–1, and gas‑tightness—make it a prime candidate for sensors, electronics, composites, energy storage, photonics, catalysis, and biomedicine [1–6]. These applications have spurred intensive research into scalable, high‑quality graphene production methods.

Available routes include micromechanical cleavage, chemical reduction of graphene oxide, chemical vapor deposition (CVD), and liquid‑phase exfoliation (LPE). Micromechanical cleavage delivers pristine, large‑area sheets but suffers from low throughput. Reduction of graphene oxide is inexpensive but leaves residual oxygen groups that degrade performance. CVD yields monolayer or few‑layer graphene with high quality but requires high temperature, vacuum, and costly substrates. LPE, pioneered by Coleman et al. [11], offers a low‑cost, scalable alternative that can produce large volumes of GNPs.

LPE typically follows three steps: dispersion of graphite in a suitable solvent, exfoliation by mechanical or acoustic means, and purification of the resulting GNPs. Over 60 solvents—including organics, low‑boiling liquids, surfactants, ionic liquids, polymers, and amphiphilic biomolecules—have been evaluated for their exfoliation efficiency [16–22]. Surface tension matching and Hansen solubility parameters help predict optimal solvents. Sonication, high‑shear mixing, ball milling, and high‑pressure homogenisation are common exfoliation techniques, with sonication—especially high‑power tip sonication—emerging as the most effective for large‑scale GNP production [23–35].

While several studies have explored sonication parameters, a systematic investigation of how tip‑sonication power and duration influence GNP yield and quality remains scarce. This study addresses that gap by evaluating the effect of power (60–300 W) and time (10–180 min) on GNP concentration, lateral size, defect density, and colloidal stability for three graphite grades. The results provide practical recommendations for industrial‑scale GNP synthesis.

Methods/Experimental

Selecting the Dispersing Liquid Medium

GNP dispersion stability depends on the interfacial free energy between the graphene sheets and the liquid. The Gibbs free energy change for exfoliation can be expressed as:

ΔG=2Nγ^GL−2γ^GL=2(N−1)γ^GL

with γ^GL = γ^GV + γ^LV – 2√(γ^GVγ^LV) = (√γ^GV–√γ^LV)². Because γ^GV is constant for graphene, matching γ^LV to γ^GV minimises ΔG and maximises dispersion efficiency [11,16].

To identify an optimal solvent, we mixed ethanol and ultrapure water at varying ratios to produce surface tensions between 22 and 50 mJ m^–2. Three flaked graphite grades—G10 (~10 µm), G30 (~30 µm), and G100 (~100 µm)—were dispersed in 40 mL of each mixture, sonicated (tip diameter 6 mm) for 10 min, and centrifuged (1000 rpm, 30 min) to remove aggregates. GNP concentrations were measured by UV–Vis absorbance at 600 nm. The 45 vol % ethanol / 55 vol % water mixture (γ_LV ≈ 30 mJ m^–2) yielded the highest concentrations and was selected for all subsequent experiments.

Graphite Exfoliation at Various Tip Sonication Parameters

Using the 30 mJ m^–2 solvent, 4 mg of each graphite grade was sonicated in 40 mL at 60, 100, 200, or 300 W for 10, 30, 60, 90, 120, or 180 min. A thermostated water bath maintained 20 °C throughout. After sonication, samples were centrifuged (1000 rpm, 30 min) and the supernatant was collected for characterisation.

Characterization of the Produced GNPs

GNP concentration was determined by UV–Vis absorbance at 600 nm. Lateral size was measured by SEM (Nova NanoSEM 430, 10 kV) on 100 GNPs per sample. Raman spectra (514 nm laser) assessed defect density via the I_D/I_G ratio. TEM (Tecnai F30, 200 kV) revealed layer number and confirmed crystallinity via electron diffraction. Sedimentation behaviour was monitored by tracking absorbance over 96 h.

Results and Discussion

Exfoliating Graphite into GNPs in Liquid Media with Different Surface Tensions

Figure 1 illustrates how GNP concentration varies with solvent surface tension. All graphite grades dispersed most efficiently in the 45 % ethanol / 55 % water mixture (γ_LV ≈ 30 mJ m^–2), consistent with prior reports [17]. Consequently, this mixture was chosen as the dispersing medium.

a Optical density and mass concentration of graphene dispersions produced by exfoliating G10, G30, and G100 as a function of the surface tension of ethanol–water solvent mixtures. b Relationship between surface tension and volume fractions of water (orange) and ethanol (blue)

Concentrations of GNP Dispersion Produced Using Various Sonication Powers and Times

UV–Vis data (Figure 2) show that GNP concentration increases with both power and time. For a fixed power, the concentration rises rapidly at first and then plateaus around 120 min. G100 could not be exfoliated at 60 or 100 W. The highest yields were obtained at 300 W, regardless of graphite grade.

Concentrations of GNP dispersions produced by exfoliating (a1) G10, (b1) G30, and (c1) G100 using different sonication powers and times. (a2)–(c2) Concentration versus sonication energy input (E = power × time).

Plotting concentration against the square root of energy input (E) reveals a robust relationship:

C_g = a E^½

with a = 1.612 × 10^–4, 4.175 × 10^–4, and 1.061 × 10^–4 mg mL^–1 kJ^–½ for G10, G30, and G100, respectively. This scaling matches earlier findings for bath sonication [23,37].

Size of GNPs Produced Using Various Sonication Powers and Times

SEM analysis (Figure 3) shows that lateral dimensions range from 1 to 3 µm across all graphite grades and sonication conditions. Increasing power or time slightly reduces size, but the effect is modest compared to the concentration trend.

Mean GNP size for (a1) G10, (b1) G30, and (c1) G100 under varying sonication conditions. SEM images of (a2)–(c2) graphite flake origins and (a3)–(c5) GNPs exfoliated at 300 W for 60, 120, and 180 min.

Defect Density of GNPs Produced Using Various Sonication Powers and Times

Raman spectra (Figure 4) indicate that I_D/I_G values increase linearly with energy input, ranging from 0.1 to 0.3 for all samples—demonstrating low defect densities. The trend is steeper for G30, suggesting superior starting material quality.

I_D/I_G for (a1) G10, (b1) G30, and (c1) G100; (a2)–(c2) versus sonication energy input.

Sedimentation Behavior of GNPs in a Liquid Medium

Sedimentation profiles (Figure 5) reveal that GNP concentrations drop sharply during the first 12 h and then stabilise. After 96 h, ~70 % of the initial concentration remains for all graphite grades and sonication conditions, indicating good colloidal stability.

Sedimentation curves for (a) G10, (b) G30, and (c) G100 sonicated at 300 W.

Implications for Selecting the Suitable Tip Sonication Parameters

Considering yield, defect density, and stability, 300 W for 120 min emerges as the optimal condition for producing few‑layer, high‑quality GNPs from all graphite grades. TEM and electron diffraction (Figure 6) confirm monolayer or few‑layer structures at these settings.

TEM images of (a) G10, (b) G30, and (c) G100 exfoliated at 300 W for 120 min. (d) Electron diffraction pattern from (b). (e) Intensity profile along the diffraction ring, showing I_{1100}/I_{2110} ≈ 1.30, indicative of few‑layer graphene.

Conclusions

We quantified how tip‑sonication power and time govern GNP yield, size, defect density, and colloidal stability. Concentration scales with the square root of sonication energy input, while lateral size decreases modestly with increased power or time. Defect density rises linearly with energy input but remains low (I_D/I_G = 0.1–0.3). Sedimentation tests show ~70 % of the initial concentration is retained after 96 h. TEM confirms that 300 W for 120 min produces few‑layer GNPs. These results provide a clear, data‑driven recipe for high‑yield, high‑quality graphene production via tip sonication.