新型选择性湿法加工

Abstract

A new selective processing technique based on a confineddynamic liquid dropmeniscus is presented. This approach isrepresented by the localized wet treatment of silicon wafersusing dynamic liquid drop that while is in contact with thewafer forms a dynamic liquid meniscus. The main scientificinnovation and relevance introduced by this work have beenapplied to industrial solar cell production and on silicon wafermetal bumps formation for the IC interconnection (i.e. copperpillars). Such new technique allows to touch in specificdefined positions the silicon wafer in order to perform anykind of wet processing (e.g. etching, cleaning and/or platingwithout the need of any protective resist. To investigate onpendant dynamic liquid drops and dynamic liquid meniscususe of computational fluid dynamic technique (i.e. numericaltechniques to accurately predict fluid flows) was followed andis presented. An experimental setup has been built to validatethe calculations. Numerical results showed a good agreementwith experimental ones. Prototypes heads, using stereo-lithography systems, were developed and localized selectiveplating without the need of lithography step was performed onsilicon.

Introduction

As in any other sector of the semiconductor and of thephotovoltaic (PV) industry, the main driver is to increaseperformances attended by cost reduction. n the photovoltaicsector this means cheaper Watt per dollar and higherefficiency silicon solar cell (i.e. > 19%). In the electronicspackaging industry this means to look for better electricalperformance and higher number of I/O (Input/Output) countusing area array packages with the aim to be “smaller, fasterand cheaper.'

In silicon solar cell module, the cost per watt compositionis given for approximately 32 percent from the semiconductorpart (i) (i.e. Si feedstock, saw wires, saw slurry, equipmentlabor, cost of capital, manufacturing margin, etc.), about 28percent from the cell cost (ii) (i.e. metallization, SiNx, dopantchemicals, equipment, labor, cost of capital, manufacturingmargin, etc.) and approximately 40 percent from the moduleassembly part (iii) (i.e. glass, EVA, metal frame, j-boxequipment, labor, cost of capital, manufacturing margin, etc.)In the cell cost (ii), the raw material accounts for more than 40percent and nowadays, the silver paste consumption perstandard silicon solar cell (i.e. 156x156mm2) influences formore than 30 percent of the total cell cost (ii). Accounting forabout 12 percent of the overall silicon solar cell cost (i+ii+iii)The main effort for solar cell metallization is to remove thesilver paste and use cheaper metals (i.e. Copper).

The International Technology Roadmap for Photovoltaics (ITRPV), in March 2012 [1] estimates a reduction of silver paste for the 156x156 mm2 cells from 200mg/cell in 2013down to less than 25mg/cell in 2020. The beginning of suchreplacement on a large-scale basis must start in 2015. Beforethe introduction of Copper altemnative metallizationtechniques, technical issues in reliability and adhesion have tobe solved and also appropriate equipment has to be availableas well. Following actual prediction no industrial solution isstill known for such large-scale replacement.

The solution that many companies are still chasing andtrying to fix is light-induced plating (LIP) techniques. LIPhave been well characterized and used for many decades(2,3,4]. LIP is referred to be a self-aligned metallizationscheme due to the fact that “in principle metal plates only inareas where the silicon is exposed [5]. In practice the mainissue ofLIP technique is the plating in pin-holes and scratchesthat are always present in industrial solar cell production lineand the only solution to overcome such problem is to useinkjet patterning [6,7,8] by a resist material that will protectthe cell during the LIP process.