TeraDox Wet Clean Process and System Advanced Surface Preparation - - PowerPoint PPT Presentation

Altay, Inc.

http://www.altayinc.com info@altayinc.com +1 855 717 3932

Advanced Surface Preparation HTechnology

TeraDox Wet Clean Process and System

TeraDox Wet Clean Process and System

What are the main factors that influence undesirable native oxide formation during and after wet processing to achieve pristine, stable, “oxide free” and ideal Si-Hx terminated silicon surfaces?

2

➢ Silicon wafer material and electrical specifications that affect the minority carrier lifetime & surface recombination velocity (ie. non-silicon impurities, dopants, crystallinity, defects, surface morphology) ➢ Dissolved impurities (ie. Oxygen, CO2, silica) in the oxide removal wet process chemistry, which include the DI water, HF and HCl, as well as impurities in the N2 purge and drying gas. ➢ Impurities in the wet process equipment materials and components ➢ Gaseous (ie. O2, CO2) permeation through wet processing equipment piping and materials which contaminate the wet process chemistry and environment that comes in direct contact with the wafer surfaces ➢ Wet process recipe conditions for the wet etching, rinsing and drying steps ➢ Bare silicon surface exposure to airborne contamination (ie. oxygen, moisture, CO2,organics), light and heat ➢ Queue time: the exposure time in air after the bare silicon surface has been prepared and before it placed in an inert environment (typically for the subsequent process)

TeraDox Wet Clean Process and System

An enhanced version of Altay’s NEO FRD system which integrates chemical processing, insitu-rinsing and a world class drying technology into a single vessel wet processor (“dry in / dry out”) for wafers and other substrates. This system... ➢ incorporates five patented “oxide free” methods and system features ➢ is comprised of a main wet process unit which is paired in series with a stand-alone Dox60 DI water degas unit ➢ maximizes the purity of the wet process chemistry and minimizies permeation, which are the critical parameters that are optimized for achieving the TeraDox’s superior “oxide free” process capabilities ➢ is offered with semi-automated or fully-automated wafer handling (including SMIF & FOUP), cassette & cassetteless configurations, and process vessels designed for batches of 3, 25 or 50 of 100-300mm diameter wafers. ➢ is also available in a Recirculating Filtered Etch Bath (RFEB) style version for R&D and low volume manufacturing applications.

What is the TeraDox System ?

3

DIMENSIONS:1350mm W - 1700mm D - 2000mm H WEIGHT: 1000kg Typical NEO-2000 TeraDox Main Unit Layout

TeraDox Wet Clean System

4

Typical Dox60 UPW Degas Unit Layout

TeraDox Wet Clean System

DIMENSIONS: 900mm W - 1500mm D - 2000mm H WEIGHT: 400kg 5

TeraDox Wet Clean System

WAFER TRANSFER / DRYER HOOD OVER THE PROCESS VESSEL WAFER CASSETTE LOAD / UNLOAD DOCK PROCESS VESSEL, CASSETTELESS WAFER CARRIER & ELEVATOR WAFER TRANSFER / DRYER HOOD OVER THE WAFER CASSETTE DOCK

PHOTOS OF THE PROCESS DECK STATIONS ON A TYPICAL CASSETTELESS 200mm SYSTEM 7

Robot FOUP Process Bath Dox60 DI water degas unit Monitor Exhaust FOUP Robot Electric Area TeraDox Space Changer & Aligner QDR Additional Chemical Bath Additional Chemical Bath Electric Area Degassed DI water Wafers

TeraDox Wet Clean System

Example of an Altay 300mm TeraDox system configuration with FOUPS & additional cleans 7

➢ Stand-alone UPW treatment unit that can be installed remotely and easily interfaces with any wet processing tool ➢ Utilizes membrane contactor technology (combination of vacuum and N2 gas sweep) to degas UPW ➢ Reduces dissolved O2 as well as other gases, TOC, metals and colloidal solids like silica ➢ Achieves > 99.999% dissolved oxygen removal efficiency ➢ Capable of providing UPW with <100ppt Dissolved Oxygen (DO) for up to 60 lpm of DI water These UPW degassing capabilities aid to prevent: ➢ watermarks and allows for wafer drying without IPA on bare silicon surfaces ➢ surface roughening / faceting on bare silicon surfaces ➢ corrosion ➢ bacterial growth in the UPW supply

Altay TeraDox Wet Clean System

Dox 60 UPW Degas Unit … it starts with purifying the water to PPT quality

8

➢

Anti-scrap features to prevent wafer breakage ➢ HCl and NH4OH (option) chemical supply network to complement the HF ➢ Chemical degassing using membrane contactor technology ➢ Fully integrated filtration and purification of the DI water and N2 ➢ Heated N2 wafer drying to minimize organic residues on the wafer surface ➢ Ionizer for the neutralization of electrostatic charge on the wafer surface (option) ➢ PdM module for reducing H2O2 to < 1ppb (option) ➢ TeraZone UHP DIO3 module (option) ➢ Mini-bulk chemical delivery module for chemical supplies (option) ➢ Mini-batch TeraDox system for 300mm wafers (alternative to full batch system)

Altay TeraDox Wet Clean System

Additional System Features and Options

9

Chemical Required

DI Water H2SO4 HF HCL NH4OH H2O2 O3 N2 IPA DIW

** DHF and/or DHCL are used based on surface to be Cleaned

Typical Full WET Clean Sequence

** DHF/ DHCL QDR HPM QDR DHF/ DHCL QDR DRY SPM

• r

SOM QDR APM QDR

Competitor WET Cleaning System & Process Flow

For Hydro Phobic Surface Only

TeraDox Wet Clean System

10

* DHF and/or DHCL are used base on surface to be cleaned (Si, SiGe, Ge)

HCL Supply UPW Degas UPW Supply HF Supply

DRY QDR DIO3 DIO3 QDR

* dHF/ dHCL

QDR

Apet FRD with Options (Degas, DIO3, dHF, dHCL)

For Hydrophilic Surface Only

Full WET Clean Sequence

Fill DIO3 Mixer Drain O3 Gas Supply CHEM Degas Chemical Required

• DI Water

HF HCL O3(g) N2(g) IPA N2, IPA

Process Vessel Dryer Dome

H1 H2 H3

TeraDox Wet Clean System

11

Understanding Oxide Thickness, Monolayers and encapsulated SIMS Areal Oxide Density (AOD) when assessing “oxide free” surface preparation capabilities

Description SiOx Thickness(Å) SiOx Monolayers SIMS AOD (atom/cm2) SiOx / SiHx Coverage (%) Reference 1 0.29 2.10E+14 29% / 71% 3.5 1.0 7.20E+14 100% / 0% Typical Native Oxide 7 2.0 1.50E+15 100% / 0% Detection Limit of XPS 0.1 0.029 2.10E+13 2.9% / 97.1% Detection Limit of SIMS 0.0005 0.00014 1.00E+11 0.014% / 99.986%

Assume the silicon wafer surface is terminated with either SiOx or SiHx species

Oxygen Hydrogen 3.5 Å SiOx 1 Å SiOx 0.014 Å SiOx Wafer surface 100% covered by 1 monolayer of Oxygen Typical HF Last process 29% Oxygen 71% Hydrogen Encapsulated SIMS DL 0.014% Oxygen 99.986% Hydrogen

TeraDOx + 650C NB Si cap ES Non-detectable <0.00014 <1.00E+11 <0.014% / 99.986%

12

TeraDox Wet Clean Process

Study to measure the effect of Dissolved Oxygen concentration in the UPW

• f the TeraDox “oxide free” process vs. areal oxygen density (AOD)

using the encapsulated SIMS method (2009) ➢ Si wafers are prepared using the TeraDox wet clean process to remove the native

• xide with the UPW supply DO degassed to 1 ppb and 0.1 ppb.

➢ An unprocessed control wafer is included in the test plan as a reference for the initial native oxide on the wafers being wet cleaned. ➢ The front surface of the three wafers are encapsulated by depositing a thin (80- 150nm) silicon layer using the standardized “ASM 650 No Bake SiH4 deposition” recipe in an ASM E2000 epi reactor. ➢ Dynamic SIMS characterization is used to measure the Areal Oxygen Density (AOD), which has the units of atoms/ cm^2, at the Si cap /Si wafer interface. ➢ The AOD data also quantifies the effect of the DO in the wet clean UPW on the efficiency of the oxide removal process .

13

unprocessed control DO= 1 ppb DO= 0.1 ppb *

* The detection limit of this DO meter was O.1ppb (current DO meters can now detect down to 10 ppt)

7.267 E15 at/cm2 2.078 E13 at/cm2 2.627 E12 at/cm2

Results for the effect of DO concentration in the UPW vs. encapsulated SIMS AOD

TeraDox Wet Clean Process

… as it can be seen, a lower DO in the UPW allows for a lower AOD

14

➢ Slides 16 & 17 show encapsulated SIMS profiles for three 200mm wafers using the Altay TeraDox (batch wet processor) dHF last wet clean

• the wafers were dried with heated N2 only (no IPA)
• oxygen peaks were non-detectable (< 1 E11 at/cm^2) on all three

wafers for the center and edge ➢ Slide 18 shows encapsulated SIMS profiles for two 200mm wafers using customer T’s process of record dHF last wet clean on a DNS single wafer wet processor.

• oxygen peaks > 1 E13 at/cm^2
• edge AOD ~3x higher than the center

Note: A significant number of process and hardware improvements have been made on the Altay TeraDox since this demo that provide even lower O and C contamination.

Customer T’s demo: encapsulated SIMS results using the “ASM 650 No Bake SiH4 deposition” recipe (2010)

TeraDox Wet Clean Process

15

O

1E+15 1E+16 1E+17 1E+18 1E+19 1E+20 1E+21 1E+22 1E+23 200 400 600 800 1000 1200 1400 DEPTH (Å) O CONCENTRATION (atoms/cc) O 1/11/2011 CF957_YR_11 Slot 20 ID D65603-09 Wf 11 PBL Spree (O) Analog Devices Inc: Slot 20 ID D65603.09 Wf 11 PBL Spree (O)

O

1E+15 1E+16 1E+17 1E+18 1E+19 1E+20 1E+21 1E+22 20 40 60 80 100 120 140 160 180 200 DEPTH (nm) O CONCENTRATION (atoms/cc) O 2/16/2011 CM639_YA_06 Devices Slot 20 ID D65603-09 Wf11 PBL Spree, edge (O) Analog Devices Slot 20 ID D65603.09 Wf11 PBL Spree, edge (O)

O

1E+15 1E+16 1E+17 1E+18 1E+19 1E+20 1E+21 1E+22 1E+23 200 400 600 800 1000 1200 1400 DEPTH (Å) O CONCENTRATION (atoms/cc) O 1/11/2011 CF957_YR_15 Slot 21 ID D65603-09 Wf 10 PBL Spree (O) Analog Devices Inc: Slot 21 ID D65603.09 Wf 10 PBL Spree (O)

O

1E+15 1E+16 1E+17 1E+18 1E+19 1E+20 1E+21 1E+22 20 40 60 80 100 120 140 160 180 200 DEPTH (nm) O CONCENTRATION (atoms/cc) O 2/16/2011 CM639_YA_02 Devices Slot 21 ID D65603-09 Wf11 PBL Spree, edge (O) Analog Devices Slot 21 ID D65603.09 Wf11 PBL Spree, edge (O)

Encapsulated SIMS (Center & Edge) for Altay’s TeraDox Wet Clean Process

Center Edge 16

O

1E+15 1E+16 1E+17 1E+18 1E+19 1E+20 1E+21 1E+22 1E+23 200 400 600 800 1000 1200 1400 DEPTH (Å) O CONCENTRATION (atoms/cc) O 1/11/2011 CF957_YR_16 Slot 25 ID D65603-09 Wf 12 PBL Spree (O) Analog Devices Inc: Slot 25 ID D65603.09 Wf 12 PBL Spree (O)

O

1E+15 1E+16 1E+17 1E+18 1E+19 1E+20 1E+21 1E+22 20 40 60 80 100 120 140 160 180 200 DEPTH (nm) O CONCENTRATION (atoms/cc) O 2/16/2011 CM639_YA_04 Devices Slot 25 ID D65603-09 Wf12 PBL Spree, edge (O) Analog Devices Slot 25 ID D65603.09 Wf12 PBL Spree, edge (O)

Encapsulated SIMS (Center & Edge) for Altay’s TeraDox Wet Clean Process

Center Edge O is non-detectable 17

O

1E+17 1E+18 1E+19 1E+20 1E+21 1E+22 1E+23 10 20 30 40 50 60 70 80 90 100 DEPTH (nm) O CONCENTRATION (atoms/cc) O Depth Areal Density (nm) (atoms/cm²) O 60.5-81.0 3.33E+13 2/8/2011 cf957_YA_52 Devices slot 13 (O) Analog Devices slot 13 (O)

Depth Areal Density (nm) (atoms/cm²) O 60.5-81.0 3.33E+13

O

1E+17 1E+18 1E+19 1E+20 1E+21 1E+22 1E+23 10 20 30 40 50 60 70 80 90 100 DEPTH (nm) O CONCENTRATION (atoms/cc) O Depth Areal Density (nm) (atoms/cm²) O 59.4-80.7 9.69E+13 2/8/2011 cf957_YA_58 Devices slot 15 (O) Analog Devices slot 15 (O)

Depth Areal Density (nm) (atoms/cm²) O 59.4-80.7 9.69E+13

Encapsulated SIMS (Center Only) for Customer T’s HF Last Wet Clean Process of Record using Competitor D’s Wet Process System Center Edge

18

An ideal in-line measurement method/system to assess the quality of “as processed” bare semiconductor wafer surfaces should provide these features and capabilities: ➢ Isolation of surface properties from those of the bulk ➢ Nondestructive & noncontact (e.g. PL & m-PCD) ➢ Meaningful and quantitative parameter (atom density, lifetimes, SRV etc.) ➢ Sensitive & accurate ➢ Imaging ➢ Fast ➢ Simple to use After researching the details for all of the measurable semiconductor surface parameters the Altay opinion is that Surface Recombination Velocity (SRV) provides the most meaningful and encompassing information since it is sensitive to all impurities and defects that can affect the electrical performance of the semiconductor device structures being fabricated. A 4-year focused effort to find the ideal method/system to measure SRV led Altay to Q-LIC in 2017, which was invented and developed by the University of Toronto’s CADIPT department.

19

20

Lock-in Carrierography (LIC) is a camera-based imaging extension of PCR ➢ Infrared camera speed (2-kHz frame rate is state-of-the-art) does not allow capturing high frequency information, which, however, is indispensable for bulk and surface separation. ➢ Heterodyne LIC (HeLIC) features a slow enough beat frequency component via non- linear mixing of two adjacent high

• frequencies. HeLIC allows full frequency

camera pixel profile acquisition without upper frequency limitations.

University of Toronto CADIPT Department

Q-LIC Method/System for the Evaluation of Surface Transport Parameters

PhotoCarrier Radiometry (PCR) ➢PCR is a form of spectrally-gated modulated photoluminescence (PL). ➢Lock-in detection. Frequency-domain. ➢Quantitative. The key is that PCR is a frequency-domain

• technique. The modulation frequency can

control the carrier diffusion (penetration) depth. SPCR=f(ω,S1,S2,τb, NT, D, β). There are

models and algorithms to process f-domain data and extract the bulk lifetime and the surface recombination velocities separately as well as other transport parameters.

• A. Mandelis, Diffusion-Wave Fields, Springer 2001. Chapter 9.

21

f1 f2

808 nm Laser 1 Iris Diffuser 808 nm Laser 2 Lock-in Amplifier 1 Lock-in Amplifier 2 NI 6259 Computer Collimator Mirror InGaAs detector Sample InGaAs Camera Sample holder Longpass filter Longpass filter

Df=f2-f1 f2

Quantitative Lock-in Carrierography (Q-LIC): Experimental set up (HeLIC, HoLIC and PCR)

• In one measurement, one can simultaneously obtain the homodyne (HoLIC) signal and the heterodyne (HeLIC) signal

from a single detector, and their image counterparts from the camera.

• Single detector measurements up to 5 MHz can be achieved.
• Imaging from DC to 2 kHz with HoLIC and up to 5 MHz for HeLIC are feasible with subsequent quantitative analysis,

generation and display of quantitative images of key transport parameters (SRV, lifetime, carrier diffusivity).

• Max laser power up to 40 W for each laser.
• Illuminated area up to 30cm x 30cm
• Duration of simultaneously measurements is <30 s for single detector PCR and Ho- HeLIC image at one frequency.
• Duration of measurements is <20 s for Ho-HeLIC image at one frequency
• Duration of complete frequency scans and data analysis depends on total number of points, typically within

just a few minutes.

22

The statistical distributions of pixel numbers of the five SRV images clearly show how the SRV changes with queue time. SRV values increased with Q-time as the native oxide layer grew on the silicon surfaces due to adsorption/absorption of impurities from air exposure.

Case study: SRV evolution vs. queue time (Q-time)

• Appl. Phys. Lett. 112, 012105 (2018)

23

Case study: SRV evolution vs. queue time (Q-time) for various treatments

Sample Apparatus DO level

• Nos. 1 & 2

Altay 400 ppb

• Nos. 3 & 4

FSI > 2 ppm

• No. 5

Altay 40 ppt

• No. 6

R&D > 2ppm

• No. 7

None N/A

• Nos. 8 & 9

Altay 100 ppt

• Nos. 10 & 11

FSI > 2 ppm

• Nos. 12

R&D 2% HF > 2 ppm

1 10

10

1

10

2

10

3

No.8 No.9 No.10 No.11 No.12

SRV(sm/s) Time after etching (hrs)

The Altay vs. the other wet processes/systems evaluated clearly demonstrates a significantly lower number of the recombination centers/defects on the surface.

• Appl. Phys. Lett. 112, 012105 (2018)

TeraDox Endorsement

“ Based on my 40 years of experience with research and development of semiconductor characterization technologies and their applications, along with the TeraDox vs. the competition studies that we have done over the past two years, I think that the TeraDox process technology

• utperforms other wet surface preparation technologies towards producing pristine bare silicon

surfaces by large multiples. While there are several critical and enabling components to the TeraDox’s outstanding performance it is primarily achieved by being able to provide DO levels in the process chemistry from at least ~25 times lower (with a resulting threefold decrease in surface recombination velocities) to up to ~50,000 times (with a resulting tenfold decrease in surface recombination velocities) over other established wet cleans.” Andreas Mandelis, FRSC, FCAE, FAPS, FSPIE, FAAAS, FASME, PhD Professor and Canada Research Chair (Tier 1) Director, Center for Advanced Diffusion-Wave and Photoacoustic Technologies

• Dept. of Mechanical and Industrial Engineering
• Dept. of Electrical and Computer Engineering

Institute of Biomaterials and Biomedical Engineering University of Toronto 5 King's College Road Toronto, ON M5S 3G8 CANADA 24

Final Comments The TeraDox technology has proven to be capable to achieve the current and future stringent requirements for semiconductor wafer surface preparation. Altay offers unique flexibility (not a cookie cutter approach) to customize a system (hardware, software and process) to satisfy each customer’s particular needs. Altay is a small company but still provides customers the expected attention and technical support required by a global equipment supplier. Please give is the opportunity to prove these claims. Thank you!

TeraDox Wet Clean System

25