The following information may be useful for your research

Ultrahigh performance cooling is one of the important needs of many industries. However, low thermalconductivity is a primary limitation in developing energy-efficient heat transfer fluids that are required for coolingpurposes. Nanofluids are engineered by suspending nanoparticles with average sizes below 100 nm in heat

transfer fluids such as water, oil, diesel, ethylene glycol, etc. Innovative heat transfer fluids are produced bysuspending metallic or nonmetallic nanometer-sized solid particles. Experiments have shown that nanofluids havesubstantial higher thermal conductivities compared to the base fluids. These suspended nanoparticles can changethe transport and thermal properties of the base fluid. As can be seen from the literature, extensive research hasbeen carried out in alumina-water and CuO-water systems besides few reports in Cu-water-, TiO2-, zirconia-,diamond-, SiC-, Fe3O4-, Ag-, Au-, and CNT-based systems. The aim of this review is to summarize recentdevelopments in research on the stability of nanofluids, enhancement of thermal conductivities, viscosity, and heattransfer characteristics of alumina (Al2O3)-based nanofluids. The Al2O3 nanoparticles varied in the range of 13 to 302 nm to prepare nanofluids, and the observed enhancement in the thermal conductivity is 2% to 36%. 

Ref:Al2O3-based nanofluids: a review

Veeranna Sridhara and Lakshmi Narayan Satapathy

Nanoscale Research Letters 2011, 6:456

www.nanoscalereslett.com/content/pdf/1556-276x-6-456.pdf

Stable suspension of nano-particles can be attained by the help of the suitable surfactant (that depends on the HLB number), this stability should be improved if the mixing device (as ultrasonicator) is developed a homogeneous dispersed mixture (mono-dispersion) where agglomeration rate is reduced. While searching about mixing of nano-materials for friend I found the attached files useful; I hope they are so for you.

Thank you, yes I have had a very difficult time finding a surfactant that works. I've tried PEG and PVP so far. I'll take a look at these articles.

Commercial sources rarely fully disclose stabilizing agents on their colloids, which is a problem for applications in which subsequent surface modifications are required to control stability and reactivity for a specific purpose/environment.  An article just was published that describes the preparation of sub-50 nm alumina nanoparticles via a ball milling technique:  Razali, et. al., "Large Scale Production and Characterization of Biocompatible Colloidal Nanoalumina", Langmuir 2014, 30(50), 15091-15101.

The technique use high-energy ball milling at T < 100 C to convert alpha phase alumina microparticles to corresponding nanoparticles without phase change. A method is also described to remove any ZrO2 contamination (from the balls and mill chamber) in the alumina nanoparticles obtained.  The particles themselves are stable as aqueous dispersions, exhibiting +40 mV (low pH) to -40 mV (high pH) zeta potentials, with agglomeration only near the pI = ~8.4.  The authors further describe a simple method for chemisorption of aminopropyltriethoxysilane onto the alumina surface to produce cationic alumina nanoparticles that form stable aqueous dispersions in near neutral and acidic solutions (pI > 10).  The aminosiloxane-modified alumina nanoparticles can later be conjugated to biomolecules by reaction with the amine.  One could also attach a carboxylic acid-terminated PEG if desired to further control stability (though this is not discussed in the article), with PEG-COOH available in a range of molecular weights from Aldrich.  One could also conceivably attach the PEG-COOH to the alumina surface (sans the aminosiloxane) by simple sonication to chemisorb PEG-COOH via the carboxylic acid function.  Either of these methods should be straightforward to perform.

Hi

I think ultrasonic is useful for decreasing the agglomeration after you add additives and surfactants.