Experimental investigation of thermo-physical properties of graphene oxide nanofluid for solar thermal applications

Abstract This article explores thermophysical properties along with the tribological behavior of graphene oxide (GO) based nanofluids and evaluates their potential applications in solar thermal systems. The experiments were designed using the Taguchi method to optimize the performance parameters, respective models developed. GO nanofluids samples (S1, S2, S3, etc.) of 0.1%, 0.2%, 0.3%, 0.4% and 0.5% volume concentration ratios were prepared via two-step method. GO nanoparticles dispersed in the distilled water followed by magnetic stirring and sonication which ensured uniform dispersion of nanoparticles. The Raman spectroscopy analysis confirmed the good crystallography and structural properties of the GO nanoparticles. Zeta potential measurements indicated that the prepared GO nanofluids remained highly stable up to the 41st day, with some samples maintaining moderate to good stability even on the 95th day. The stability also confirmed after 155th day by UV- Vis spectroscopy, demonstrating the superior stability of GO-based nanofluids compared to other nanofluids. The thermal conductivity test results reported an enhancement in thermal conductivity from 5.35% to 39.11% with variation of nanoparticle concentration over a temperature range of 25 °C to 65 °C. Notably, nanoparticle concentration exhibited a stronger influence on thermal conductivity, particularly at lower temperature regimes. Rheological analysis revealed that the viscosity of GO nanofluids increased by 155.62% as the GO concentration roses from 0.1% to 0.5% at ambient temperature. Additionally, a slight reduction in specific heat capacity as increase in concentration ratio and a marginal increase in density were observed compared to the base fluid. Based on experimental studies regression model developed. ANOVA confirms that confidence level higher than 95% satisfy, with larger F value and P < 0.001, MSE < 5% and noted that the regression models are robust, reliable, and suitable for prediction and inference.