PhD With Me 🔬

Etch rate test

Reactive Sputtering and Wet Etching of TiN Films: Process Insights and Optimization

Abstract

We present an investigation into the deposition and patterning of titanium nitride (TiN) films via reactive sputtering and wet etching. The deposition process was monitored using plasma emission colour as an in-situ diagnostic, and subsequent wet etching experiments were performed to tailor film thickness and assess surface uniformity. Our results highlight key process parameters and underscore the need for improved equipment and protocols to achieve reproducible film properties.

1. Introduction

Titanium nitride (TiN) is a material of significant interest in microelectronics and protective coatings due to its high hardness, chemical stability, and electrical conductivity. In this study, we explore two complementary techniques:

Our objectives are to correlate plasma characteristics with film quality and to quantify etch rates under varying conditions.

2. Experimental Methods

2.1 Reactive Sputtering Deposition

TiN films were deposited using reactive sputtering under varying conditions. Notably, the plasma colour served as an indicator of the deposition regime:

Note: Initial deposition rates were found to be unreliable due to issues with the substrate shutter not fully closing. Future experiments should consider continuous timing or upgrading equipment to support a substrate temperature of (600^\circ\text{C}).

Table 1. Deposition Parameters Summary

Parameter Condition/Value Remarks
Gas Composition Ar/N(_2) mixture Plasma colour indicative of composition
DC Power Varied Lower DC power yields a white plasma
N(_2) Flow Rate Varied Higher flow yields pink plasma
Substrate Temperature Target (600^\circ\text{C}) Future equipment upgrade required

2.2 Wet Etching Procedures

Two methods were employed for wet etching TiN films:

  1. Direct Etching:
    Samples of various initial thicknesses (e.g., 52±3 nm, 127±5 nm) were immersed in a 30 wt.% Hydrogen Peroxide solution heated to 50ºC (with a later increase to 60ºC). For the thinnest film, an iterative protocol was used: the sample was removed every 60 seconds, rinsed with deionized water, and dried with nitrogen, allowing for a stepwise assessment of the etch progression.

  2. Protective Layer Assisted Etching:
    A protective photoresist (Ma-N 405) was spin-coated and baked for 60 seconds to create defined regions prior to etching. In one instance, a red-coloured sample with an initial thickness of (228\pm4) nm was etched for approximately 14 minutes. It was observed that the photoresist began to degrade (react with H(_2)O(_2)) after around 12 minutes, complicating complete removal.

Surface roughness was assessed with a Bruker Profiler and subsequently confirmed via Atomic Force Microscopy (AFM). For example, a sample initially measuring (127\pm5) nm showed a reduced thickness of (122.5\pm10.6) nm after etching, corresponding to an etch rate of approximately 7 nm/min ((\pm12%)). This is notably lower than the 40 nm/min reported in the literature ([doi:10.1063/5.0179924, \ https://doi.org/10.1557/PROC-477-447]).

3. Results and Discussion

3.1 Deposition Observations

The plasma colour during sputtering served as an effective proxy for film composition:

The reliability of deposition rates was compromised by a mechanical issue (substrate shutter malfunction), underscoring the need for enhanced process control.

3.2 Etching Analysis

In the direct etching approach: