Anthony J Muscat

Anthony J Muscat

Department Chair, Chemical and Environmental Engineering
Professor, BIO5 Institute
Professor, Chemical and Environmental Engineering
Primary Department
Department Affiliations
(520) 621-6162

Research Interest

Research Interest
Dr. Anthony Muscat's group's research interests are in surface chemistry, specifically the chemical processes used to clean, etch, or deposit on the surfaces of solids, including 2D planes and 3D nanostructures. An understanding of surface chemistry can be used to optimize existing materials used in microelectronics, catalysis, or solar energy conversion and develop new materials with unique properties. The primary research goals are learning how chemical reactions take place on surfaces and how the atom or molecular group terminating a surface affects the types of structures that can be built on it or using it as a building block. Understanding the reaction mechanism provides a means to rationally design interfaces for specific purposes. Current research projects include 1) engineering the surfaces of semiconductors (GaAs, InAs, InGaAs, CuInS2) for advanced electronic, optoelectronic, and solar devices, 2) synthesis and self-assembly of nanoparticles such as quantum dots (clusters of atoms 1-5 nm in diameter), 3) self-assembled monolayer (SAM) formation, and 4) dealloying metal alloys using liquids and supercritical fluids to make nanoporous noble metal films and composites. We approach these problems by using experiments and modeling to understand the mechanisms of the surface chemical reactions that are at the heart of these technologies.


Deng, Z., Lie, F. L., Shen, S., Ghosh, I., Mansuripur, M., & Muscat, A. J. (2010). Water-Based Route to Ligand-Selective Synthesis of ZnSe and Cd-Doped ZnSe Quantum Dots with Tunable Ultraviolet A to Blue Photoluminescence. LANGMUIR, 25(1), 434-442.
BIO5 Collaborators
Indraneel Ghosh, Anthony J Muscat
Thorsness, A. G., & Muscat, A. J. (2003). Moisture absorption and reaction in BPSG thin films. Journal of the Electrochemical Society, 150(12), F219-F228.


As-deposited (AD) and annealed (500, 750, and 900°C) borophosphosilicate glass (BPSG) films were characterized during aging, baking, and etching using transmission Fourier transform infrared spectroscopy and ellipsometry. BPSG films contained oxides such as Si-O, P=O, P-O, and B-O as well as hydroxyl groups such as SiO-H, HOH, PO-H, and BO-H in a variety of local bonding environments, which became more uniform as the annealing temperature was increased. The water content in the BPSG films increased steadily during storage at ambient conditions. Based on bond strength, polarity, thermodynamics, and FTIR data, the B-O bond is the primary site for water adsorption on the surface of the film. Water absorption within the film was consistent with a reaction-limited model. Water reacted readily with P-O groups forming P=O and PO-H, which H bonds strongly within the film. The slower reaction with P=O moieties is proposed as the rate-limiting step for water absorption. Annealing after deposition strengthened the Si-O lattice, which reduced the affinity to absorb water. Etching rates ranged from 1 to 10 Å/s on the films studied. A 200°C bake desorbed water from the surface layer of the films and increased the reaction rate between water and P=O and B-O to form PO-H and BO-H groups. The bulk etching rate was not affected by baking, but the induction time needed to start etching increased to 31 ± 1, 22 ± 2, and 74 ± 24 s for the AD, 500 and 750°C annealed films, respectively, and increased from 45 ± 5 to 72 ± 5 s for the 900°C annealed film.

Xie, B., Montaño-Miranda, G., Finstad, C. C., & Muscat, A. J. (2005). Native oxide removal from SiGe using mixtures of HF and water delivered by aqueous, gas, and supercritical CO2 processes. Materials Science in Semiconductor Processing, 8(1-3 SPEC. ISS.), 231-237.


HF and water delivered using liquid, gas, and supercritical CO2 phases were used to remove native oxides and suboxides from SiGe surfaces containing 15-30% Ge. X-ray photoelectron spectroscopy was used to monitor the Si 2p and Ge 2p3/2 peaks as a function of the process type, HF concentration, and Ge fraction. A larger percentage of Si oxides were removed compared to Ge oxides by all three processes. The threshold HF concentration in aqueous-phase solutions for complete removal of Si and Ge oxides was approximately 28 millimole/l (mM) on Si0.85Ge0.15 and increased by approximately an order of magnitude on the surfaces with the higher Ge concentrations. The threshold concentration increased from 0.007 mM on Si0.85Ge0.15 to at least 0.14 mM on Si 0.70Ge0.30 to obtain significant removal of both Si and Ge oxides using supercritical CO2 to carry the HF and water etching solution. Although air exposure of the liquid water and supercritical CO 2 etched samples provided an opportunity for reoxidization, the trends in the metal and oxide XPS peak areas with HF concentration suggest that chemical treatment had the largest effect. Gas-phase HF/vapor processes removed all of the SiO2, Si suboxides, and GeO2 in less than 30 s offering the best performance of the processes studied without roughening the surface. © 2004 Elsevier Ltd. All rights reserved.

Xie, B., & Muscat, A. J. (2004). Condensation of silanol groups in porous methylsilsesquioxane films using supercritical CO2 and alcohol cosolvents. IEEE Transactions on Semiconductor Manufacturing, 17(4), 544-553.


Fourier transform infrared (FTIR) spectroscopy, goniometry, and electrical measurements were used to investigate the effect of adding alcohol and carboxylic acid cosolvents to supercritical carbon dioxide (scCO2) to condense silanol groups in blanket porous methylsilsesquioxane (p-MSQ) films (JSR LKD 5109). The aliphatic C1-C6 alcohols removed approximately 50% more hydrogen-bonded silanol (SiO-H) groups than pure scCO2, leaving isolated silanol groups on the surfaces of the pores. Acetic acid removed H-bonded silanols but left fewer isolated moieties. On a molar basis, n-propanol, isopropanol, and n-butanol removed the largest percentage of silanols per molecule of cosolvent. These cosolvents were also among the lowest vapor pressure cosolvents studied, making them the most environmentally acceptable. As-received ashed ultralow-k MSQ had a contact angle of less than 10° and a dielectric constant of 3.5±0.1. After processing in a mixture containing 7% n-propanol and scCO2, the contact angle was 15° and the dielectric constant decreased to 3.2±0.1. The surface was hydrophilic after processing in mixtures of cosolvents and scCO2 because of the isolated silanol groups on the surface. A comparison of the trends across the alcohol series indicates that cosolvent addition to scCO 2 increased the solubility of water in the supercritical fluid mixture compared to pure scCO2. Within the same class of molecules, the solubility of the cosolvent in the supercritical fluid is a more important selection criterion than the solubility of water in the cosolvent.

Muscat, A., Jiang, F., & Muscat, A. J. (2012). Ligand-controlled growth of ZnSe quantum dots in water during Ostwald ripening. Langmuir : the ACS journal of surfaces and colloids, 28(36).

A strong ligand effect was observed for the aqueous-phase growth of ZnSe quantum dots (QDs) in the Ostwald ripening (OR) stage. The QDs were made by injecting Se monomer at room temperature followed by a ramp to 100 °C. The ramp produced a second, more gradual increase in the concentrations of both Zn and Se monomers fed by the dissolution of QDs below the critical size. The dissolution process was followed using measurements of the mass of Zn in QDs and in the supernatant by inductively coupled plasma optical emission spectroscopy (ICP-OES). Despite the flux of monomers, there was little growth in the QDs of average size based on UV-vis absorption spectra, until the temperature reached 100 °C, when there was a period of rapid growth followed by a period of linear growth. The linear growth stage is the result of OR as the total mass of Zn in QDs and in the solvent remained constant. The growth data were fit to a continuum model for the limiting case of surface reaction control. The rate is proportional to the equilibrium coefficient for ligand detachment from the QD surface. The ligand 3-mercaptopropionic acid (MPA) was the most tightly bound to the surface and produced the lowest growth rate of (1.5-2) × 10(-3) nm/min in the OR stage, whereas thiolactic acid (TLA) was the most labile and produced the highest growth rate of 3 × 10(-3) nm/min. Methyl thioglycolate (MTG) and thioglycolic acid (TGA) produced rates in between these values. Ligands containing electron-withdrawing groups closer to the S atom and branching promote growth, whereas longer, possibly bidendate, ligands retard it. Mixed ligand experiments confirmed that growth is determined by ligand bonding strength to the QD. Photoluminescence spectroscopy showed that the more labile the ligand, the more facile the repair of surface defects during the exposure of the QDs to room light.