硅晶片的润湿性和老化行为

Abstract

This paper reports on the wettability and aging behaviors of the silicon wafers that hadbeen cleaned using a piranha (3:1 mixture of sulfuric acid (HSO4, 96%) and hydrogenperoxide (H202, 30%), 120 C), SC1 (1:1:5 mixture of NH40H,H202 and H20, at 80 C) or HFsolution (6 parts of 40% NH4F and 1 part of 49% HF, at room temperature) solution, andtreated with gaseous plasma. The silicon wafers cleaned using the piranha or SC1 solutionwere hydrophilic, and the water contact angles on the surfaces would increase along withaging time, until they reached the saturated points of around 70°. The contact angleincrease rate of these wafers in a vacuum was much faster than that in the open airbecause of loss of water, which was physically adsorbed on the wafer surfaces. The siliconwafers cleaned with the HF solution were hydrophobic, Their contact angle decreased inthe atmosphere, while it increased in the vacuum up to 95°. Gold thin films deposited onthe hydrophilic wafers were smoother than that deposited on the hydrophobic wafersbecause the numerous oxygen groups formed on the hydrophilic surfaces would reactwith gold adatoms in the sputtering process to form a continuous thin film at thenucleation stage. The argon, nitrogen, oxygen gas plasma treatments could change thesilicon wafer surfaces from hydrophobic to hydrophilic by creating a thin (around 2.5 nmsilicon dioxide film, which could be utilized to improve the roughness and adhesion of thecold thin fim.

1. Introduction

The terms of“hydrophilic” and “hydrophobic”are usedto describe the surface wettability by water or non-polarliquids, respectively. A surface is hydrophilic if it tends toadsorb water or be wetted by water with a low contactangle; otherwise it is hydrophobic. The surface hydropho-bicity of a silicon wafer is important in many applications(1-8]. For example, the silicon hydrophilic property playsan important role in the process of direct wafer bonding [1,2], silicon-on-insulator (SOI) products [3], silicon-basedsensors [2,4], biochemical systems [5,6], and microelectromechanical systems (MEMS)[8].A silicon surface is hydrophilic because the siliconsurface is generally covered by a thin native oxide layer.which links many silanol (Si-OH) groups, and physicallyadsorbs 1 or 2 monolayers of water on top [7,9]. The nativesilicon oxide (SiOz) on a silicon surface could be formed inthe presence of water molecules, which might be extractedfrom moisture or from polymerization of Si-OH on thesurface [1]. The singular and associated -OH groups are relatively stable on a silicon dioxide surface during heating (2,10]A variety of cleaning methods could be utilized to modifythe silicon surface wettability.

The silicon wafer cleaned using a piranha solution is hydrophilic, but the siliconwafer cleaned by a HF solution is hydrophobic, and thehydrophilicity could be restored after long time exposureto moisture or by a plasma activation treatment (11].However, most of these studies were carried out in atmo-sphere environments, and few works systematically inves-tigated the silicon wafer surface wettability and agingbehaviors under a vacuum condition and the impact onthe noble-metal thin-film deposition and growth. Forexample, the influence of the silicon surface wettabilityon the morphology of the sputtered gold film is still notclear. In this work, by measuring the surface contact angle,and using X-ray photoelectron spectroscopy (XPS)atomic force microscopy (AFM) and high-resolution trans-mission electron microscopy (TEM), we investigated silicon surface wettability and its aging process, and thesputtered gold thin-film growth mechanism on siliconwafers. In-situ plasma treatments were performed toimprove the roughness and adhesion of the sputteredgold thin film on silicon wafers. It is a low cost andeffective technique, especially for some applications, such as magnetic-nanowire gold thin-film coating, in which a titanium or chromium adhesion layer is strictly prohibited.

2.Experimental

P-type silicon wafers were used in the experiments, andthey were purchased from Semiconductor Wafer, Inc. Weused a piranha solution, an SC1 solution or HF solution toclean the silicon wafers in the study. The piranha solution is a3:1 mixture of sulfuric acid (HSO4, 96%) and hydrogenperoxide (H202, 30%), and is heated up to 120 C. The SC1solution used is a mixture of NH OH, HOz and HO with aratio of 1:1:5 at 80 C (12] The HF solution is a mixture of 6parts of 40% NH4F and 1 part of 49% HF The wafers wererinsed in deionized (DI) water for three cleaning cycles,which lasted for ten minutes with a continuous DI waterpurge, and then were dried in a spun drier with a hot Npurge. A vacuum oven (DZ-1BZ) was used to provide avacuum condition with a pressure less than 1 Torr, and itwas placed in a class 100 cleaning room with 42% humidityat 23 °C A plasma treatment was carried out in a sputtering chamber (RD4, ESC, UK) for the silicon surface wettabilitymodification at a pressure of 15 mTorr and RF power of100 W with argon, nitrogen or oxygen gas. The base pressureof the plasma treatment chamber was around 3.0e-8 Torr.The duration of the plasma treatment was three minutes. Wemeasured the water contact angles using a contact anglesystem (DSA100, KRUSS) with a sessile mode. At least threemeasurements were conducted for each sample, and theiraverage value was taken to be the contact angle for thatsample. Every measurement was performed at a fresh site.An X-ray photoelectron spectroscope (XPS，KRATOS AxisUltra DLD, UK) was used to analyze the wafer surfacecomposition. It is equipped with a monochromatic Al KaX-ray source (hu=1486.6 ev) operating at 150 W, a multichannel plate and a delay line detector in a 1.0 x 10-9 Torrvacuum. XPS surveys were collected at the fixed analyzerpass energies of 160 eV. Charge neutralization was performed for all samples. Binding energies were referencedto the C 1s binding energy of adventitious carbon contamination, which was taken to be 284.8 eV.

The gold thin films were deposited in a magnetronsputtering system (RD4, ESC, UK) with a process pressureof 5 mTorr at the room temperature. Before film deposition.the chamber was pumped to a base pressure of around3.0e- 8 Torr, and the pre-sputtering time was three minutesto remove contaminants on the target surfaces. The thicknessof the gold thin film was measured using a profilometer(Dektak 8，Advanced Development Profiler, US), and thesurface roughness was examined using an atomic forcemicroscope (AFM, D3100, Topometrix Instrument). A transmission electron microscope (TEM, FEI Titan 80-300 s-twin)was also employed to inspect the cross-section of the goldthin film texture and the interface laver between siliconsubstrate and gold film.