Substrates and Epitaxy in III-V Manufacturing Rober obert Y t - - PowerPoint PPT Presentation

Substrates and Epitaxy in III-V Manufacturing

Rober

• bert Y

t Yank anka 20 2010 CS 0 CS MANTECH W MANTECH Workshop shop

2010 CS MANTECH Workshop Session 2 2

Ov Over ervie view

• Substrat

Substrate Manuf Manufacturing Ov cturing Over ervie view

• Metal-Organic Chemical V

Metal-Organic Chemical Vapor Deposition (MOCVD) por Deposition (MOCVD) Basics Basics

• Molecular Beam Epitaxy (MBE)

Molecular Beam Epitaxy (MBE) Basics Basics

• Epiw

Epiwaf afer Charact er Characterization rization

• RFMD Epiw

RFMD Epiwaf afer Pr er Production Ov

• duction Over

ervie view

2010 CS MANTECH Workshop Session 2 3

Substrate Manufacturing Overview

2010 CS MANTECH Workshop Session 2 4

Pr Process Flo

• cess Flow

w of

• f GaAs

GaAs Substrat Substrate Pr e Production

• duction

Single Crystal Growth Cylindrical Grinding Slicing Edge Rounding Polishing Characterization & Inspection Laser Marking

2010 CS MANTECH Workshop Session 2 5

Substrat Substrates – es – Boule Gro

• ule Growth

wth Techniques niques

As As

Growing Cooling

SolidLiquid Solid

Temperature Profile

low

Melting Point

Seed Seed

Czochralski (Cz) Gr Czochralski (Cz) Growth wth Horiz Horizontal Bridgman (HB) Gr

• ntal Bridgman (HB) Growth

th

2010 CS MANTECH Workshop Session 2 6

Substrat Substrates – es – LEC vs. V EC vs. Vertical Boat ical Boat

Ver ertical Boat ical Boat Liq Liquid Encapsulat id Encapsulated ed Czochralski (LEC) Czochralski (LEC)

Boule Mo Boule Moves Fu Furnace M Moves

2010 CS MANTECH Workshop Session 2 7

Substrat Substrate – e – VB VB vs.

• vs. VGF

GF

Longer Longer Ingo Ingot Shor Shorter Ingo Ingot

Ver ertical Boat ical Boat vs. vs. Ver ertical ical Gradient Gradient Freeze reeze

2010 CS MANTECH Workshop Session 2 8

Substrat Substrates – es – Grinding Grinding and and Slicing Slicing

As sliced Wafers Wire Saw

2010 CS MANTECH Workshop Session 2 9

Substrates – Edge Beveling and Polishing

La Laser Ma r Marki rking ng Double-Side P Double-Side Polishing lishing Edge Be Edge Beveling ling

2010 CS MANTECH Workshop Session 2 10

Substrat Substrates – es – Inspection and Inspection and Charact Characterization rization

Sur Surface Inspection ace Inspection Substr Substrat ate Flatness e Flatness

2010 CS MANTECH Workshop Session 2 11

Substrat Substrates – es – Inspection and Inspection and Charact Characterization rization

• Electrical Charact

Electrical Characterization rization

– Hall and Hall and non-contact resistivit non-contact resistivity measur y measurements are ements are per performed on the seed and

• rmed on the seed and tail t

tail to de determine the ermine the electrical charact electrical characteristics of eristics of the boule the boule

• Etch Pit

Etch Pit Decoration (K Decoration (KOH) OH)

– Chemical e Chemical etch used t tch used to det determine dislocation density rmine dislocation density

2010 CS MANTECH Workshop Session 2 12

MOCVD Basics

2010 CS MANTECH Workshop Session 2 13

Epitaxial Cr Epitaxial Crystal ystal Gr Growth wth

• Epitaxy

Epitaxy

– is is deriv derived fr ed from the Greek w

• m the Greek word meaning “or

d meaning “ordered upon” dered upon” – Epitaxy is Epitaxy is the gr the growth of wth of thin thin single cr single crys ystals tals of one mat

• f one material

erial on

• n the

the cr crystal f ystal face ce of

• f the same

the same (homoepitaxy) or (homoepitaxy) or ano another (he ther (heter eroepitaxy)

• epitaxy)

mat material, such that the tw erial, such that the two mat

• materials ha

erials have a e a def defined, relativ ined, relative e structural orientation structural orientation

• Two main R
• main Rules of

les of epitaxy epitaxy

– Matc Matching of hing of Symme Symmetr try be y betw tween the subs een the substrat trate and e and the epila the epilayer er – Misf Misfit be it betw tween lattice cons een lattice constants of subs tants of substrat trate and e and epila epilayer should be er should be minimal minimal

• Exam

Examples of ples of Epitaxy:- Epitaxy:-

– GaAs GaAs on

• n GaAs

GaAs (Homoepitaxy) (misf (Homoepitaxy) (misfit v it value = lue = 0) 0) – AlGaAs on GaAs AlGaAs on GaAs (He (Heter eroepitaxy)

• epitaxy) (misf

(misfit it value > 0) lue > 0) – InGaAs InGaAs on

• n GaAs

GaAs (He (Heter eroepitaxy) (misf

• epitaxy) (misfit v

it value >> lue >> 0) 0)

2010 CS MANTECH Workshop Session 2 14

Metal Organic Chemical V Metal Organic Chemical Vapor apor Deposition Deposition

• First document

First documented in 1 ed in 1963 b 963 by Har y Harold Manase ld Manasevit f vit for silicon r silicon

• n sapphire at R
• n sapphire at Rockw

ckwell Corporation ell Corporation

• Manase

Manasevit gre vit grew GaAs on sapphire in 1 GaAs on sapphire in 1968 968

• Gr

Growth in th involv lves the transpor es the transport of t of me metal-organic and tal-organic and hydride dride precursors t precursors to heat heated substrat ed substrates where the es where they p pyrolize, lea lize, leaving ing the gr the growth species on the substrat th species on the substrate sur e surface ace

• Also ref

Also referr erred t ed to as (LP-, AP-) OMCVD, MO as (LP-, AP-) OMCVD, MOVPE, OMVPE VPE, OMVPE

2010 CS MANTECH Workshop Session 2 15

Metal Organic Chemical V Metal Organic Chemical Vapor apor Deposition Deposition

• MOCVD is capable of

MOCVD is capable of high gr high growth th rat rates, making es, making it the most it the most widely used t widely used techniq echnique f ue for the manuf r the manufacture of cture of optical

• ptical

de devices such as LED’s vices such as LED’s

• At lo

t lower gr er growth rat th rates, MOCVD can es, MOCVD can pr produce abrupt

• duce abrupt

int inter erfaces, allo ces, allowing it’s application t wing it’s application to RF RF de devices such as vices such as HBT’s and HBT’s and FET’s FET’s

2010 CS MANTECH Workshop Session 2 16

Metal Organic Chemical V Metal Organic Chemical Vapor apor Deposition Deposition

• Sour

Source mat ce materials f rials for MOCVD are e r MOCVD are extremely hazar tremely hazardous, dous, req requiring rigor uiring rigorous monit us monitoring and

• ring and handling pr

handling procedures

• cedures

– Me Metal-organics are tal-organics are toxic and xic and pyr yrophoric phoric – Hydrides are Hydrides are highly t highly toxic xic – Hydr Hydrogen

• gen carrier gas is e

carrier gas is explosiv xplosive

• Maint

Maintenance cy enance cycles driv cles driven b en by coating of coating of reaction chamber reaction chamber

– Req equires relativ uires relatively freq ely frequent cleaning uent cleaning to remo remove mat material fr erial from

• m

syst system com em components and

• nents and reduce

reduce par particulat ticulates in es in the chamber the chamber – Syst System em do downtime wntime is is on

• n the or

the order of hour der of hours

2010 CS MANTECH Workshop Session 2 17

MOCVD – MOCVD – Syst ystem Block Diagram m Block Diagram

2010 CS MANTECH Workshop Session 2 18

MOCVD – MOCVD – Syst ystem Pho m Photograph

• graph

2010 CS MANTECH Workshop Session 2 19

MOCVD – MOCVD – Syst ystem Schematic m Schematic

2010 CS MANTECH Workshop Session 2 20

MOCVD – MOCVD – Flo low w Contr Control

• l
• Mass Flo

Mass Flow Contr Controller (MFC) ller (MFC)

– A A por portion of the gas tion of the gas flo flow passes thr passes through a ugh a tube incorporating tw tube incorporating two

• temperature sensor

erature sensors – A A tem emperature dif erature difference results that is erence results that is pr propor

• portional t

tional to the mass the mass flo flow thr through the contr ugh the controller ller

2010 CS MANTECH Workshop Session 2 21

MOCVD – MOCVD – Sour

• urces

es

MO Bubbler MO Bubbler

Carrier gas (H2) flowing through the bubbler picks up MO vapor from the liquid

Typical MOCVD Sour pical MOCVD Sources ces

Group III: Tri-methyl gallium, Tri-methyl aluminum, Tri-methyl indium Group V: Arsine, Phosphine, Ammonia, Tertiary-butyl arsine (TBAs), Tertiary-butyl phosphine (TBP) Dopants: Carbon-tetrabromide, Disilane, Di-methyl zinc

2010 CS MANTECH Workshop Session 2 22

TMGa

δ0

H2

V

Diffusion

CH3 radical AsH3

J

MOCVD - MOCVD - Growth wth Dynamics Dynamics

• Sour

Source molecules mix ce molecules mixed int d into a a carr carrie ier gas (h gas (hydr ydrogen gen) with ) with mid-str mid-stream v velocity elocity V V

• Frictio

iction reduce reduces gas v s gas velocity t locity to near zer near zero at sur at surfaces aces

• Sour

Sources must dif ces must diffuse thr use through the near st ugh the near stagnant bo nant boundar undary l layer a er above s e subs bstrat ate e

• Precursor decom

Precursor decomposition

• sition
• Sur

Surface adsor ace adsorption & tion & dif diffusion usion incor incorporation

• ration and

and gr growth th

• Gr

Growth wth rat rate depen depends upon s upon pressure, flo pressure, flow rat w rate and t and temperatu erature

2010 CS MANTECH Workshop Session 2 23

MOCVD – MOCVD – Growth wth Regimes gimes

A.

• A. Desorption /

Desorption / Gas Gas Phase Phase Limit Limited Gr ed Growth th B.

• B. Mass T

Mass Transpor ansport Limit t Limited ed Gr Growth (Dif th (Diffusion thr fusion through ugh boundar boundary la layer). er). C.

• C. Sur

Surface Kine ace Kinetic tic Limit Limited ed Gr Growth th Gr Growth rat th rate e ~ ~ tem emperature independent in erature independent in mass mass transpor transport limit t limited regime ed regime

2010 CS MANTECH Workshop Session 2 24

MOCVD – MOCVD – Gr Growth wth Rat Rate Contr Control

• l

Gr Growth Rat th Rate: The Mass T e: The Mass Transpor ansport Limit t Limited R d Regime egime

2010 CS MANTECH Workshop Session 2 25

MOCVD – MOCVD – In-situ Monit In-situ Monitoring

• ring
• Pyr

Pyrome metr try

– Substrat Substrate t e tem emperature measurement erature measurement – Multi-w Multi-wavelength f length for r emissivity emissivity correction correction

• Re

Reflecto tometry

– Gr Growth rat th rate e and and ternar ernary com y composition data

• sition data

– Waf afer cur er curvature measureme ature measurement

Transparent Film ansparent Film Ab Absor sorbing Fi ing Film

2010 CS MANTECH Workshop Session 2 26

MBE Basics

2010 CS MANTECH Workshop Session 2 27

Molecular Beam Epitaxy Molecular Beam Epitaxy

• Molecular Beam Epitaxy w

Molecular Beam Epitaxy was de s developed b eloped by Alfred Cho and y Alfred Cho and John Ar John Arthur thur in 1 in 1970 at Bell Labs 70 at Bell Labs

• Originally applied t

Originally applied to the gr the growth of th of GaAs lasers GaAs lasers

• Gr

Growth in th involv lves the e es the evaporation of aporation of high purity elemental high purity elemental sour sources in an ultra-high v ces in an ultra-high vacuum en cuum envir vironment

• nment
• The resulting “molecular beams” im

The resulting “molecular beams” impinge heat pinge heated ed substrat substrates pr es producing epitaxial gr

• ducing epitaxial growth

th

2010 CS MANTECH Workshop Session 2 28

Molecular Beam Epitaxy Molecular Beam Epitaxy

• MBE typically utilizes slo

MBE typically utilizes slower gr er growth rat th rates (~1µm/hr) s (~1µm/hr) resul resulting ing in high q in high quality ality, at atomically abrupt int

• mically abrupt inter

erfaces.

• ces. This

This is an adv is an advantage f ntage for superlattice and/or q r superlattice and/or quantum antum w well based ell based applications. applications.

• Relativ

elatively sim ly simple gr le growth kine th kinetics and precise la tics and precise layer contr er control l ha have made MBE a e made MBE a popular t popular techniq echnique f e for in r investigating no estigating novel el de devices structures in R&D. vices structures in R&D. Ho However, MBE is also widely MBE is also widely used in the manuf used in the manufactur cture of pHEMT’s and HBT’s. e of pHEMT’s and HBT’s.

• MBE sour

MBE source mat ce materials rials are relativ are relatively saf ly safe, as the e, as they are lo y are low w vapor pressur por pressure solids at solids at room t

• om tem

emperatur erature

2010 CS MANTECH Workshop Session 2 29

Molecular Beam Epitaxy Molecular Beam Epitaxy

• Maint

Maintenance cy enance cycles are driv cles are driven b en by the need t y the need to replenish replenish sour sources ces

• Ultra-high v

Ultra-high vacuum en cuum envir vironment req

• nment requir

uires e es extensiv nsive cleaning e cleaning and baking of the syst and baking of the system em

– Maint Maintenance enance cy cycles req cles require se uire several w eral weeks t eeks to com comple lete

• Cam

Campaign lengths betw aign lengths between clea een cleaning can range fr ning can range from six t

• m six to

nine months nine months

2010 CS MANTECH Workshop Session 2 30

MBE – MBE – Block Diagram Block Diagram

Wafer ers are are intr introduced fr

• duced from atmosphere
• m atmosphere

int into the

• the load lock

load lock. Ne Next the xt they mo move int into

• the prep

the prep chamber f chamber for

• r
• utgassing.
• utgassing.

Finally Finally, the , they are are transf transferred int erred into the the deposition chamber f deposition chamber for r gr growth. th.

2010 CS MANTECH Workshop Session 2 31

MBE - MBE - Syst ystem m Pho Photograph

• graph

Load Locks Load Locks Prep/T Prep/Tran ransf sfer Chamb r Chamber er Depo Chamber Depo Chamber Sour Source Flange ce Flange Cr Cryo Pum Pumps Veeco Gen2000

2010 CS MANTECH Workshop Session 2 32

MBE – MBE – Syst ystem m Sc Schematic hematic

Cr Cryopan panel maint l maintain ins lo s low w im impurity purity back backgr ground in the depo chamber

• und in the depo chamber.

Substrate ro rotation i improve roves unif uniformity

• rmity.

2010 CS MANTECH Workshop Session 2 33

MBE – MBE – Ef Effusion Cell usion Cell

Typically ypically used f used for Gr r Group

• up III’s and

III’s and Dopants Dopants Ga, Ga, In, Al In, Al Be, Be, Si Si Cell stability pla Cell stability plays a ys a major major role in

• le in pr

product

• duct variability

ariability

2010 CS MANTECH Workshop Session 2 34

MBE – MBE – Valv alved Crac d Cracker

Crack Cracking Zone Zone Subli Sublimator

• r

(Sour (Source Load) e Load) Va Valve A Actuator Gr Group V’s consum

• up V’s consumed at a

ed at a high rat high rate due t e due to the flux req the flux requirements uirements Tem emperat perature ram ure ramps are im ps are impractic practical f for the r the large mat large materia rial loads req loads required ired Valv lved ed cells pr cells provid ide flux con e flux control at a l at a constant t nstant temper erture ture Valv alve contr e controls flo ls flow betw between een sublim sublimator and crack

• r and cracking zone

ng zone

2010 CS MANTECH Workshop Session 2 35

MBE – MBE – Growth Dynamics wth Dynamics

Gr Growth only when Gr wth only when Group

• up III

III present present Gr Growth wth rat rate pr e propor

• portional

tional to Gr Group

• up III flux

III flux at at optimized gr

• ptimized growth t

th tem emperature erature Sur Surface ace adsorption adsorption and disassociation and disassociation of Gr

• f Group V dimer
• up V dimers

Arriving Gr Arriving Group

• up III at

III atoms react on

• ms react on the sur

the surface f ace forming the com

• rming the compound
• und

Substrat bstrate t e temperature must erature must be be suf sufficient t icient to pr promot

• mote l

e layer-by-la

• layer g

er growth th High v High vapor pressure Gr apor pressure Group

• up V’s req

V’s require high flux t ire high flux to stabilize the sur

• stabilize the surface

ace

2010 CS MANTECH Workshop Session 2 36

MBE – MBE – Gr Growth Rat wth Rate

• Gr

Growth rat th rate pr e propor

• portional t

tional to Gr Group III flux

• up III flux

– Follo llows Arrhenius relationship ws Arrhenius relationship – Doping Doping is is contr controlled in lled in the same the same manner manner as as gr growth rat wth rate

2010 CS MANTECH Workshop Session 2 37

MBE – MBE – In-situ In-situ Monit Monitoring

• ring
• Ion gauge f

Ion gauge for measuri r measuring sour g source mat ce material flux rial flux

• Reflection High Energy Electr

flection High Energy Electron Dif

• n Diffraction (RHEED) f

fraction (RHEED) for r gr growth rat th rate and e and structur structural inf al information

• rmation
• Optical p

Optical pyrometr

• metry f

y for substrat r substrate t e temperature erature

• Band edge thermometr

Band edge thermometry f y for substrat r substrate t e temperature erature

• Laser reflect

Laser reflectome metr try f y for gr r growth rat th rate

• Atomic absor
• mic absorption flux monit

tion flux monitoring ring

2010 CS MANTECH Workshop Session 2 38

Epiwafer Characterization

2010 CS MANTECH Workshop Session 2 39

Epiwa wafe fer C Characte terization

• Charact

Characterization Cat rization Categories gories

– Sur Surface mor ace morphology hology – Electrical pr Electrical proper

• perties

ties – Structur Structural pr al proper

• perti

ties es

• Destructiv

Destructive vs. Non-Destructiv e vs. Non-Destructive

• Pr

Product W

• duct Wafers vs. Pr

s vs. Process Monit

• cess Monitors
• rs

2010 CS MANTECH Workshop Session 2 40

Epiw Epiwaf afer Charact er Characterization – rization – Sur urface Morphology face Morphology

Sur Surface r ace roughness, par ughness, particulat iculates, gr es, growth def th defects, scratches cts, scratches Typicall pically non-destr y non-destructiv ctive

• Visual Inspection

Visual Inspection

– Looking f Looking for gr

• r gross def
• ss defect

ects such such as as haze, scratche haze, scratches, large par large particles ticles

• Automat

mated Inspection d Inspection

– Commer Commercial t cial tools such

• ols such as

as Tencor’s Sur encor’s SurfScan or Scan or Candela Candela – Pr Provides w ides waf afer ers maps maps for r haze, point haze, point def defects and ects and scratches scratches

• Optical Micr

Optical Microscop

• scopy

– Standar Standard and d and phase-contrast phase-contrast – Usually f Usually for diagnostic purposes, t r diagnostic purposes, to indentify the type of def indentify the type of defects ects

2010 CS MANTECH Workshop Session 2 41

Epiw Epiwaf afer Charact er Characterization – rization – Electrical Proper Electrical Properties ties

Doping Doping le levels, doping unif els, doping uniformity

• rmity, doping depth pr

, doping depth profile

• file

Carrier concentration, mobility Carrier concentration, mobility, shee , sheet resistance t resistance

Destr Destructiv ctive or non-destr e or non-destructiv ctive depending upon t e depending upon techniq echnique ue

• Electr

Electro-chemical CV Pr

• -chemical CV Prof
• filing

iling

– Usually ref Usually referred t erred to as as Polar laron fr n from the commer

• m the commercial syst

cial system em – Pr Provides a ides a carrier carrier concentration depth pr concentration depth prof

• file but

ile but is is destructiv destructive

• Hall Measurement

Hall Measurement

– Pr Provides a ides average carrier erage carrier concentration and concentration and mobility o mobility over a er a sam sample le – Typically destructiv pically destructive (non-destructiv e (non-destructive appr e approaches e

• aches exist)

ist)

• Non-contact Sheet R

Non-contact Sheet Resistance sistance

– Pr Provides non-destructiv ides non-destructive shee e sheet resistance mapping t resistance mapping

2010 CS MANTECH Workshop Session 2 42

Epiw Epiwaf afer Charact er Characterization – rization – Structural Proper tructural Properties ties

La Layer thickness, com r thickness, composition, unif

• sition, uniformity and
• rmity and cr

crystal q ystal quality ality Techniq echniques are typically non-destr es are typically non-destructiv ctive

• Sur

Surface Pr ace Prof

• filome

ilometry

– Sim Simple st le step ep height measurement f height measurement for r thickness / thickness / gr growth rat th rate

• X-ra

ray Dif y Diffraction / R fraction / Reflection eflection

– La Layer thickness and r thickness and com composition mapping

• sition mapping

– Def Defect density estimat ect density estimates es

• Phot

Photoluminescence

• luminescence

– Optical t Optical techniq chnique ue for pr

• r probing bandgap
• bing bandgap

– Pr Provides com ides composition mapping

• sition mapping and

and mat material q erial quality data uality data

2010 CS MANTECH Workshop Session 2 43

RFMD Epiwafer Production Overview

2010 CS MANTECH Workshop Session 2 44

RFMD MBE Operations RFMD MBE Operations

• Nearl

Nearly 100% of y 100% of AlGaAs AlGaAs HBT and pHEMT HBT and pHEMT starts are from tarts are from MBE systems in Greensbor MBE systems in Greensboro, NC , NC

• Minimal number of

Minimal number of device device structur structures es

– Easily more MBE Easily more MBE systems than SKU’s systems than SKU’s – Allows an Allows an MBE MBE tool tool to produce a to produce a single SKU for weeks or months single SKU for weeks or months

• Heavy focus on production metrics

Heavy focus on production metrics

– System uptime, throughput and System uptime, throughput and yield yield – Manpower efficiency Manpower efficiency – Reproducibility Reproducibility across across different types of MBE different types of MBE systems systems – Wafer uniformity and Wafer uniformity and run-to-run reproducibility run-to-run reproducibility

2010 CS MANTECH Workshop Session 2 45

RFMD MBE Operations RFMD MBE Operations

• MBE Operation’s production fleet consists of

MBE Operation’s production fleet consists of 14 14 – 7x6” x6” MBE system MBE systems

– Veeco Veeco GEN2000 GEN2000 – Riber Riber MBE 7000 BE 7000

• Three additional system are dedicated

Three additional system are dedicated to process to process development development

− Riber Riber MBE 6000 BE 6000 − VG V100 VG V100

• Development work is also carried out on production tools

Development work is also carried out on production tools

− Facilitates Facilitates transition of new processes transition of new processes into production into production

2010 CS MANTECH Workshop Session 2 46

RFMD MBE RFMD MBE Operations – Operations – 24/7 Utilization 4/7 Utilization

• Production technicians in a

Production technicians in a 12 12 hour day hour day shift maintain 24 shift maintain 24 hours of hours of production on MBE systems production on MBE systems

– Documentation, training Documentation, training and and discipline discipline become critical become critical to to effective effective shift passdowns shift passdowns

• Leverage automation on production MBE systems to

Leverage automation on production MBE systems to reduce reduce manpower demands manpower demands

• Dual load locks allow contin

Dual load locks allow continuous operation (complimentary uous operation (complimentary

• peration)
• peration)

2010 CS MANTECH Workshop Session 2 47

RFMD MBE Operations RFMD MBE Operations

MBE Operations Equipment R&M Production Equipment Engineers R&M Technicians Shift B Shift A Component Rebuilding Process Engineers Shift A Shift B Process Development Production Technicians Shift A Shift B

2010 CS MANTECH Workshop Session 2 48

RFMD MBE RFMD MBE Operations – Operations – Syst System Management m Management

Original Source Load Quantities Know n Consum ption Rate per Epiw afer Run Num ber of Runs Since Last Reload

2010 CS MANTECH Workshop Session 2 49

RFMD MBE RFMD MBE Operations – Operations – Maint aintenance Cy enance Cycle cle

• Maintenance activities are tracked to provide feedback for

Maintenance activities are tracked to provide feedback for minimizing system turn-around times minimizing system turn-around times

2010 CS MANTECH Workshop Session 2 50

RFMD MBE Operations – RFMD MBE Operations – Epi Epi Charact Characterization rization

• Epiwafer

Epiwafer characterization data characterization data is monitored using online is monitored using online process control charts process control charts

• Daily

Daily check of SPC check of SPC data provides data provides feedback to alert feedback to alert Production technicians and Production technicians and engineers of engineers of any need for any need for process adjustments process adjustments

• Appropriate process adjustments are

Appropriate process adjustments are calculated with calculated with standardized tools standardized tools

− Input characterization data from recent runs Input characterization data from recent runs − Input Input MBE MBE process parameters from same process parameters from same runs runs − Outputs include setpoint Outputs include setpoint correction correction

2010 CS MANTECH Workshop Session 2 51

RFMD MBE Operations – RFMD MBE Operations – HBT “Quic HBT “Quick T Turn” rn”

• Periodic sampling from ev

Periodic sampling from every MBE tool growing HBT’s ery MBE tool growing HBT’s

• Essential for determining if de

Essential for determining if device parameters vice parameters are on target are on target

• Additional information on doping levels of

Additional information on doping levels of individual layers individual layers can eliminate the need fo can eliminate the need for dedicated calibration runs r dedicated calibration runs

− Base sheet Base sheet resistance resistance − Emitter sheet resistance Emitter sheet resistance − BVebo BVebo

• Process

Process adjustm adjustments ents calculated with standardized tools calculated with standardized tools

2010 CS MANTECH Workshop Session 2 52

RFMD MBE Operations – RFMD MBE Operations – Fab Data ab Data

• MBE

MBE process engineers monitor online trend charts process engineers monitor online trend charts

− Cont Contac act resistanc t resistance − Doping levels and breakdown voltages Doping levels and breakdown voltages − Current gain (HBT) Current gain (HBT) − Turn-on volt Turn-on voltage age (HBT or (HBT or pHEMT) pHEMT)

• Sampling of

Sampling of recent production epiwafers recent production epiwafers keeps in-process test data eeps in-process test data current (Modified FIFO) current (Modified FIFO)

• MBE

MBE process engineers held accountable for process engineers held accountable for MBE MBE related wafer fab related wafer fab yield loss yield loss

− Wafer fab Wafer fab correcti

• rrective action reports automa

e action reports automatic tically assigned y assigned − MBE MBE engineer may scrap additional engineer may scrap additional wafers from epiwafer wafers from epiwafer invent inventory

• ry

− Creat Create a a culture in which MBE culture in which MBE personnel realize wafer fab personnel realize wafer fab scrap is scrap is more more costly to costly to business than epiwafer business than epiwafer scrap scrap

2010 CS MANTECH Workshop Session 2 53

RFMD MBE Operations – RFMD MBE Operations – Continuous Impr Continuous Improvement ement

• Ongoing efforts to improve

Ongoing efforts to improve yield and reduce variation yield and reduce variation

− MBE epiwafer MBE epiwafer yield yield − Wafer fab Wafer fab line ine yield yield − Final product test yield Final product test yield

• Pareto analysis of

Pareto analysis of scrap causes during monthly and scrap causes during monthly and quarterly reviews. (MBE yield and quarterly reviews. (MBE yield and fab fab yield) ield)

• Common databases provides traceabilit

Common databases provides traceability from GaAs from GaAs substrates through MBE and substrates through MBE and fab fab processing rocessing

− Invaluable in Invaluable in determining correlations between MBE determining correlations between MBE process process parameters, epiwafer parameters, epiwafer characterization and characterization and device performance device performance − Identify Identify sources of sources of variation variation that can that can be be reduced to reduced to achieve specific achieve specific goals in goals in product performance or product performance or yield yield

2010 CS MANTECH Workshop Session 2 54

Ackno knowledgements wledgements

• Thanks t

Thanks to the f the follo llowing people and wing people and com companies f anies for r their assistance in their assistance in pr producing this presentation:

• ducing this presentation:

– Phil Grunes, Sumit Phil Grunes, Sumitomo Electric Semiconduct mo Electric Semiconductor r Mat Materials, Inc. rials, Inc. – Kevin St vin Stevens, Cha ens, Charles L les Lutz a tz and d Eric R Eric Rehder hder, Kopin Corporation pin Corporation – Molly Doran and Molly Doran and Nat Nate Gr e Groneberg

• neberg, V

, Veeco Instr eeco Instrument ments – Tom R m Rogers, RFMD gers, RFMD