Lithography - Chemical Engineering IIT Madras

Download Report

Transcript Lithography - Chemical Engineering IIT Madras

Introduction
 Scope
 Testing (quality control) is done at various steps
 Our focus: Only electrical tests
 Other tests: Defectivity test, etc
 Electrical Testing
 Process check (Only overall condition)
 PARAMETRIC or E-Test
 Product check (Pass/Fail for each chip)
 BIN or SORT
 RELIABILITY
2
7-Jul-15
Process Test (parametric)
 Process Check
 Simple structures, made between the ‘real’ chips
 If the structures are very bad (all shorted, all open/ broken)
 Chips are also likely to be ‘dead’
 Not worth processing further (if the wafers are in the midst
of processing)
 Not worth testing the chips (if the wafers have completed
processing)
 So, process check is done
 in the ‘line’, at some standard steps (after M1 CMP and so on...)
 and at the end, just before the chips are tested for pass/fail
 The ‘simple structures’ used for process checks are normally called
as ‘Scribe line’ or ‘Kerf’
 Process check is called “scribeline test’ or ‘kerf test’
3
7-Jul-15
Review from Mask: Basics
 A mask may be 100 mm x 100 mm (for example)
 So, one ‘print’ will be about 25 mm by 25 mm, on the
wafer.
 A chip may be only 5 mm by 4 mm
 So, one mask will have perhaps 20 chips, if the chip
is small
 The gaps between the ‘real chips’
are used for
 alignment marks (TBD)
 test structures (TBD)
4
7-Jul-15
Process Test (parametric)
 Basic check for shots/opens:
 Metal lines: Snakes, Combs
 Snake & Comb for both open and
shorts
 At each layer
 Many structures
 ==> many tests, time
 few structures
 ==> lower ability to predict fails
 low resolution
 if the snake is too long, cannot
differentiate between real open vs
marginal open vs good structure
 if comb is too large, cannot
7-Jul-15
distinguish between leakage vs shorts
5
Process Test (parametric)
 Example of
shorts:
 What is expected in
the electrical test
 Very low resistance (or high leakage
current)
CDF plot
Leak
between
combs
© micro magazine
Snake
resistance
6
7-Jul-15
Process Test (parametric)
 Plot of leakage current (CDF or QQ-Norm)
 sorted values
 log scale to identify ppm or ppb level defects
 Stop processing the wafers with shorts/opens
 Helps isolate problematic modules
7
7-Jul-15
Process Test (parametric)
 VDP
 Proximity structures
 Kelvin
 Misalign Structures
 To find alignment margin
 Border structures (via, contact)
 Size structures
 Capacitor (low-k specific)
 intra layer
 inter layer
 Copper specific (Dishing/Erosion)
 Concept of Short Loop, POR, split
7-Jul-15
8
Tester
©Advantest
Ability to test at high temp
some may have low temp capability also
Die level vs package level
9
7-Jul-15
Probe
©Advantest
Tungsten tip (for example)
7-Jul-15
Auto probe vs Manual Probe
10
Test Program
 C or script like
 Normally well commented, reasonably readable
 information on voltage applied (for example), store
the data in test name etc..
 DUT (Device Under Test)
 alignment and x,y movement
 C-V test (for oxide) OR I-V results
 time consuming
 usually one result in a DUT in a die
 TDDB, NBTI, ESD, EM
Temp. Dependent Dielectric Breakdown, Negative Bias Temperature Instability
 Summary file (for water, for lot ...)
11
7-Jul-15
Pads/Pins: Some examples
Pads in product Test
Pads in “test chip”
and perhaps in scribe line
Usually
Parametric (analog) tester = slower
Digital tester = faster
Parallel testing => Faster even for parametric
7-Jul-15
12
Pads/Pins: Some examples
Too long a pad-group
==> probe alignment/
probe contact issues
Pads in “test chip”
and perhaps in scribe line
7-Jul-15
13
Product Test
 Binning
 Soft bin, hard bin
 initial stages, test the least failing part
 some times COF
 production stage, test the most failing part
 always SOF
 Continuity (leakage, open)
 Built In Self Test (BIST)
Functional
 SCAN
 Memory
 Repair (yield can improve dramatically)
 Fail Bit Map (FBM)
7-Jul-15 WL ==> M1 short, BL fail ==> via fail (for
14
Product Test
 IDDQ test ( Direct Drain Quiescent current)
 some devices will pass the logic test, but will
degrade in the field
 likely to be captured by IDDQ test
 Circuit must be CMOS and other parts must be
isolated. Tested part should not give significant
current, by design
 (does not run at the specified frequency, so it
won’t replace the function test, for example)
 will not detect opens/ high resistance /
prob
interconnect capacitance effects
 DC, AC, functional IDDQ...
 QBD - charge (Q) to break down
7-Jul-15  Transistor (gate oxide)
i spec
15