Frequently Asked Questions (FAQ) - Luxel Corporationluxel.com/wp-content/uploads/2011/10/Radak-FAQ-Rev9-28-2011.pdf · Frequently Asked Questions (FAQ) Radak & OLED Furnaces Table

Frequently Asked Questions (FAQ) Radak & OLED Furnaces

Table of Contents:

1. What crucible is recommended for my material? 2. What is the maximum operating temperature of my crucible? 3. Do I need a separate crucible for each material to be deposited? 4. Which power controller do I need? 5. Can the power controller be wired for voltages other than 110V AC? 6. Will my power supply work with the Radak furnace? 7. Can the furnace be run with an empty crucible? 8. Which type of thermocouple should I use? 9. Can I replace the heater coil myself? 10. How do I clean my Radak furnace? 11. What is the difference between OLED and Radak furnaces? 12. When should the radiation baffle be used? 13. How do you suggest cleaning the anodized aluminum “Organic” crucibles if I want to use

them for different organic materials? 14. Where can I find a copy of the Radak Operating Manual? 15. What if my material deposits at temperatures above 1500C? 16. Will the presence of Oxygen damage the furnace during oxide depositions? 17. What are the latest design changes on Radak furnaces? 18. Do you make larger capacity evaporation source than the Radak III? 19. What is the flux distribution from a Radak source? Where should I place my substrate to

achieve a uniform coating? 20. How do I install a Radak furnace in my system? 21. How accurately can you control the deposition rate from a Radak furnace? 22. What are the voltage and current requirements of the furnace? 23. How long does the furnace take to cool down? Can I speed up the cool down?

Crucible Recommendation Table APPEDIX A: Radak and OLED Furnace Technical Info

360.378.4137 | 360.378.4266 fax | [email protected] | www.Luxel.com | P.O. Box 1879, 515 Tucker Ave., Friday Harbor, WA 98250, USA FAQ Rev.9-28-2011

1. What crucible is recommended for my material?

Please see our growing table of crucible / liner recommendations for depositing common materials using Radak sources.

Another helpful resource is the table of deposition techniques on Kurt Lesker’s website here. This table lists the boat or crucible typically used with most materials. The table is based on nominal chemical compatibility and availability of standard crucible/boat materials. The table also lists the melting point and temperature where useful vapor pressures are achieved for deposition. If the evaporation temperature is below the melt temperature the source sublimes, greatly reducing the risk of crucible cracking. If the material melts below the evaporation temperature, a liner may be required to ensure the crucible does not crack.

2. What is the maximum operating temperature of my crucible?

Crucible Maximum Recommended Temperature (°C)Alumina 1700PBN 1500Quartz 1000Arc‐Coated Stainless 900Organic 460

3. Do I need a separate crucible for each material to be deposited?

If your material can be easily removed with solvents or acids that do not attack the crucible then you may be able to use the same crucible for multiple materials. However, because of the risk of cross‐contamination we recommend using separate crucibles for each material to be deposited. If using a liner, be aware that some deposition will occur on the upper part of the crucible due to the shorter height of the liner; for this reason we also recommend a separate crucible/liner pair for each material to avoid cross‐contamination.

4. Which power controller do I need?

We currently offer three models of power controller: the Radak II, II+, and IIP. The basic difference between the controllers is the model II is meant as a stand‐alone controller, the II+ has digital I/O for working with a PLC, deposition controller, etc., and the IIP has the features of the II+ with the addition of a switchable power output allowing the controller to select between one of two attached furnaces.

Please see the feature comparison chart below to determine the right controller for your needs:


http://www.lesker.com/newweb/menu_techinfo.cfm?section=materials&init=skip

Model II Model II+ Model IIP

General

Incorporates Eurotherm Model # 2404 2404 2404 RS 232 Communications Auto‐tune algorithm Output channels 1 1 2*

Programmer

Stored programs 1 4 4 Program segments 8 16 16 Stored programs can be linked Digital front panel program select

I/O

Analog inputs 1 PV† 1 PV, external control

2 PV, external control

Software‐selectable input source ‐ All common thermocouple types ‐ Custom linearization profile ‐ All standard RTDs, pyrometers ‐ Voltage, current inputs User calibration External control inputs Digital I/O 3 input 3 input Relay outputs 2 SPST, 1 SPDT 2 SPST, 1 SPDT Control schemes simple PID simple PID, external simple PID, external

Other

Eurotherm iTools software for configuration and instrument communications

* Allows sequential operation of two furnaces. † PV is the Process Variable, generally a thermocouple input.

All of these controllers have RS232 connections for communication with the Eurotherm 2404 controller which is the heart of the unit. The 2404 is addressable through the Modbus protocol which is documented in the Eurotherm 2000 series MODBUS and EI‐BISYNCH Communications Handbook. The controllers can also be configured and run via Eurotherm’s iTools software which is available for free download.


http://www.eurotherm.com/products/controllers/single-loop/2400/2404/?assetesctl1616844=1833

http://www.eurotherm.com/products/controllers/itools/?assetesctl1460014=30778

5. Can the power controller be wired for line voltages other than 110V AC? What size circuit is recommended?

A custom transformer can be used to accommodate your country’s line voltage (i.e. 100V, 220V, 240V, etc) but this may extend the lead time a bit. The standard 110VAC unit should be connected to a 20 amp 108‐120VAC circuit in your facility.

6. Will my power supply work with the Radak furnace?

Luxel’s power controllers are designed to drive the furnace at low voltage using a high‐current transformer and phase‐angle‐fired SCR. This minimizes the risk of electrical shock, arcing, and ionic conduction through the crucible which becomes problematic at higher voltages. If you decide to use your own power supply we offer the following recommendations:

• Output voltage should not exceed 60VAC at 100% power.

• DC power supplies will also work, but our experience is limited to AC.

• For Radak/OLED I furnaces, output current should not exceed 30 Amps at 100% power.

• For Radak/OLED II furnaces, output current should not exceed 40 Amps at 100% power.

• If depositing a low temperature material (i.e. less than 500C), we suggest limiting the maximum power output to prevent an over‐temperature situation.

• If using PID temperature control, make sure to tune your PID parameters to avoid giving the furnace 100% power when cold.

• After line connections are made, always test the furnace at a low power setting first to ensure there are no shorts in the system, and verify the furnace is heating properly.

7. Can the furnace be run with an empty crucible?

Yes. But there must be a crucible installed. The furnace will run without issue with an empty crucible but the temperature feedback relies on the thermocouple contact with the crucible so do not operate the furnaces without a crucible. If you must operate the furnace without a crucible for some reason, run the furnace on manual with a low (e.g. 10%) power setting.

8. Which type of thermocouple should I use?

Type K thermocouples tend to be more robust than type C thermocouples, but are limited to a maximum of 1200°C. Type C thermocouples will operate up to 2300°C, but please remember that Radak furnaces are only warranted up to 1500°C. Another consideration is that certain automation products do not yet recognize the C‐Type thermocouple standard. Be sure to choose the same thermocouple type as used in your vacuum feedthrus, power controller(s), or other system components that interface with the furnace.

9. Can I replace the heater coil myself?

Yes. In the event that the tungsten heater filament should fail, you can purchase our heater replacement kit and follow the instructions here. To prolong the life of the heater filament, we recommend operating


http://www.luxel.com/PDF_doc/Heater%20Replacement.pdf

the furnace at pressures less than 5x10‐6 Torr. Introduction of gases during deposition is possible, but may substantially shorten the life of the filament.

10. How do I clean my Radak furnace?

The furnace is made of low‐outgassing high‐temperature materials. The furnace body is stainless steel with a molybdenum vapor shield. The interior is tantalum, molybdenum, tungsten, and high purity alumina. Cleaning, when necessary, should be restricted to the stainless‐molybdenum cover and the alumina crucible. Cleaning methods will depend on the material to be removed, but acids, bases, and scouring materials, which do not strongly attack stainless, molybdenum, or alumina can be used. Cleaning of the multi‐layer radiation baffle is not recommended beyond ultrasonic cleaning in hot water, followed by a rinse in clean solvent such as distilled isopropyl alcohol, to remove any water trapped in the multi‐layer structure, followed by oven drying.

A word of caution about using acids to clean the furnace cover ‐ The molybdenum vapor shield will be quickly attacked by most acids; therefore exposure to acid should be minimized.

If there is heavy contamination or material buildup due to a crucible overflow, please contact our technical staff at (360) 378 4137 for specific recommendations or repair work if necessary.

11. What is the difference between OLED and Radak furnaces?

Our OLED furnace is the low temperature version of the Radak furnace. Since late 2008, the only difference is that the OLED furnace is not supplied with an internal radiation baffle.

12. When should the radiation baffle be used?

With the radiation baffle installed on Radak furnaces, it can be hard to get good PID temperature control below about 500°C. Removing the baffle greatly improves radiative cooling, hence improving response at lower temperatures. With a properly tuned PID you should be able to control the furnace to within a degree at low temperatures.

There are no hard and fast rules with the baffle. We recommend using the baffle above 500‐600°C but it depends on your tolerance to heat load in the system vs. control lag /cool‐down time. The furnace controls relatively well at 400°C with the baffle in, but takes quite a bit longer to cool down to 100°C which is a pain if you’re doing batch coating. Above 600°C you’re dumping a lot of heat into the system and your maximum temperature will ultimately be limited by power needed to drive the furnace. The stainless outer cover may not fair so well either if the furnace is run too hot for too long without the baffle.

13. How do you suggest cleaning the anodized aluminum (Organic) crucibles if I want to use them for different organic materials?


Remove as much material as you can with a soft tool, then use an oxygen plasma to remove the remaining organics. The main reason for the anodizing is to prevent shorting of the heater coils. Oxygen plasma may damage the anodizing but will work to passivate the surface so no harm done.

14. Where can I find a copy of the Radak Operating Manual?

Please download the latest version of the Radak operating manual on our website here.

15. What if my material deposits at temperatures above 1500C?

The 1500°C limit includes a considerable safety margin and the furnace design has been tested to 1700°C; however, operation above 1500°C is not covered under warranty and may result in a shortened filament life. If, for example, your material begins to evaporate at 1650°C (Yttrium) you may be not be able to get a usable deposition rate within the range of the Radak furnace.

16. Will the presence of Oxygen damage the furnace during oxide depositions?

Many oxide materials require a background of Oxygen to ensure proper stoichiometry. The oxygen can attack the W filament used to heat the furnace. As a result, the filament may become brittle and prone to breakage. As long as the oxygen level is low enough (around the 10‐3 Torr range), the degradation of the filament should be relatively slow. If the filament is being attacked by the oxygen, WO3 will deposit inside the furnace and will appear as a blue hue on the crucible and ceramics inside of the furnace. If the filament does fail, it can be replaced by the user (with purchase of a heater replacement kit).

17. What are the latest design changes to Radak furnaces?

In response to end‐user requests to operate our furnaces under more demanding environments (specifically CuInGaSe2 solar cell production) we’ve made a few changes in our furnaces. If you’re familiar with our earlier design, the primary difference you'll notice is changing the top from a 45 deg cone to the flat top we use in our OLED sources. The flat top reduces material build up on the vapor shield and has a pressed ring which indexes the crucible for proper alignment. The second change you'll notice is the use of a ceramic plug for the thermocouple. The ceramic plugs make installation and removal much easier. On the Radak/OLED I, the ceramic plug extends slightly from the side of the furnace, which may present an interference problem in some installations. For these cases an optional 90° mounting adapter is available to turn the plug downwards, eliminating protrusion from the side of the furnace. Upon request, the old‐style copper bosses are still available for thermocouple connection. Optional stainless steel boss style terminals are now available for Se compatibility.

Internally, the Radak II furnaces are constructed with a profiled ceramic insulator post eliminating the Ta braces used in the previous design. The profiled ceramic enhances the structure of the cage and eliminates the potential shorting and breakage issues of the Ta supports. Finally, all Ni has been removed from the furnace construction to make the furnaces less susceptible to damage from over‐temperature transients.


http://www.luxel.com/PDF_doc/Radak%20Manual.pdf

Crucibles are now being manufactured with a notched lip. The lip projects slightly beyond the vapor shield to promote better sealing between the vapor shield and crucible to guard against evaporant entering the furnace. The lipped crucibles are compatible with existing furnace designs. These crucibles became standard in 2009.

18. Do you make larger capacity evaporation sources than the Radak III?

Yes! For greater capacities than the Radak III (100cc), a Radak IV is also available for large volume (1 L capacity) low temperature (<500C) applications. Even larger custom furnaces are also available.

Luxel works with customers to develop furnaces to meet their special requirements, including even higher‐volume sources, water‐cooled fast cool‐down furnaces, and many other special requests. Luxel’s technical support staff is ready to help you with your application; please call us to discuss your furnace requirements (360‐378‐4137).

19. What is the flux distribution from a Radak source? Where should I place my substrate to achieve a uniform coating?

We have measured the flux distribution from a Radak II furnace, which closely follows a Cos4 distribution; please see the document titled “Flux distribution and uniformity.pdf” available on our website.

For further detail, an excellent paper describing the theory and modeling of flux distribution from Knudsen cell sources was published in the Journal of Vacuum Science and Technology:

“Monte Carlo calculations of the beam flux distribution from molecular‐beam epitaxy sources”, Adamson, S.; OCarroll, C.; McGilp, J. F., Journal of Vacuum Science & Technology B, May 1989, Volume: 7 Issue:3, Pages: 487 – 490.

20. How do I install a Radak furnace in my system?

Our sources mount with a single 1/4 ‐20 or optional 6 mm bolt (please specify metric with your request for quote). In addition you will need power (40VAC/ 40A) and thermocouple (C or K) feedthroughs. Furnaces are supplied with vacuum side thermocouple lead wire, you will need to supply the vacuum side power leads (10 AWG solid copper recommended). Power controllers ship with both air side power and thermocouple leads. You can find the instructions for installation and operation of our furnaces on our website here.

21. How accurately can you control the deposition rate from a Radak furnace?

The rate at which any material deposits will be affected by several factors including the crucible and liner used, the furnace temperature, as well as the furnace power in some cases. Most notably, quartz crucibles are largely transparent to IR radiation and thus pass heat to the source material without much heating of the crucible; in this case the thermocouple will respond much slower to an increase in power than the deposition rate. However, for most crucible/material combinations the deposition rate is


http://www.luxel.com/PDF_doc/Radak%20Manual.pdf

360.378.4137 | 360.378.4266 fax | [email protected]

directly proportional to the crucible temperature and can thus be accurately controlled with the temperature controller included in our Power Controller II Series.

For applications where rate must be very accurately controlled, or held constant for long periods of time we recommend using our Power Controller II+ with rate input from your deposition controller. Our power controllers can automatically adjust the furnace temperature to respond to changes in deposition rate in order to keep your process running smoothly. For a detailed example of configuring your Power Controller for automation, please see the document titled “Controller Automation IIP.pdf” available on our website.

In response to a customer’s inquiry we tested the deposition rate stability that we could achieve using an OLED I furnace depositing copper phthalocyanine (CuPC). The rate information was recorded from an Inficon IC5 deposition monitor, temperature information and control performed using a Eurotherm 2704 temperature controller. Data was logged via Eurotherm iTools (a free software interface for our Power Controller II series). The system was configured with an OLED I furnace, Luxel “Organic” crucible, QCM rate monitor, and witness substrate. Both the QCM and witness substrate were located ~11" above the source.

The crucible was filled approximately 1/2 full with copper phthalocyanine and installed in the system. The system was evacuated to <5X10^‐6 Torr, and the source heated to ~430C at 50C/sec. Source temperature was maintained manually to control rate for a reported (IC5) total "raw" thickness of 505nm. Post deposition, the furnace was allowed to cool to <100C then the chamber was vented and opened. The witness sample thickness was measured as 600nm via DekTak 3st surface profilometer .

Results are given the table and figure below. As the figure shows, the temperature was very gradually increased in order to keep the rate extremely steady (less than a tenth of an angstrom/second change).

Rate (A/Sec)55 minute CuPe depositonMean 1.296173Standard Error 0.000171Median 1.295915Mode 1.299293Standard Deviation 0.009807Sample Variance 9.62E‐05Range 0.072629Minimum 1.267881Maximum 1.34051

| www.Luxel.com | P.O. Box 1879, 515 Tucker Ave., Friday Harbor, WA 98250, USA FAQ Rev.9-28-2011

http://www.eurotherm.co.uk/products/controllers/itools/


22. What are the voltage and current requirements of the furnace?

Recent test data has been collected on several Radak and OLED furnace models; please Appendix A: Radak and OLED Furnace Technical Info for complete details.

23. How long does the furnace take to cool down? Can I speed up the cooldown process?

The furnace must be allowed to cool to 100C before opening the chamber. Cool‐down time will depend greatly on the type and amount of source material in the crucible. The cool‐down time is shorter for OLED furnaces than Radak furnaces due to the absence of the radiation baffle. For a comparison between furnace models with empty alumina crucibles please refer to the chart “Radak & OLED Furnace Cool‐down Curves” in Appendix A.

Some processes have critical cool‐down requirements for either chemical or high‐throughput reasons. It is possible to backfill your system with an inert gas like Nitrogen to speed the cool‐down after the furnace has cooled for at least 10 minutes and is below 300C. This will typically cut the cool‐down time in half, but DO NOT open the chamber until the furnace is below 100C. For even faster cool‐down times Luxel offers a water‐cooled jacket for the Radak II and III. Please note that this custom option will add lead time to your order, and your application must be discussed with our technical staff prior to ordering.


Crucible Recommendation Chart Important: Please use the crucible recommendation table as a guideline only; Luxel Corporation makes no assurances of chemical compatibility, nor guarantees any results by using this table. Always seek out multiple references to support the use any crucible with any material.

Material Symbol Crucible Liner (if required) Melt °C S / D NOTESALQ3 ALQ3 OrganicAluminum Al PBN 660Antimony Sb Alumina 631 S O,TArsenic As Alumina 814 S O,TBarium Ba Alumina Ta 717 O,P,TBarium Cloride BaCl2 Alumina Ta, Mo 962Beryllium Be Alumina W 1284Bismuth Bi Alumina W, Ta, Mo 271 TBoron Nitride BN Alumina 2973 SCadmium Cd Alumina 321 S P,TCalcium Ca Alumina 810 S OCalcium Fluoride CaF2 Alumina Mo, Ta 1418Cerium Ce Alumina 785 HCesium Iodide CsI Alumina 621Chromium Cr Alumina W 1857 SCobalt Co Alumina 1478Copper Cu Alumina Mo 1083Copper phthalocyanine CuPC OrganicDysprosium Dy Alumina Ta 1409Erbium Er Alumina Ta, W 1522 SEuropium Eu Alumina 822 S OGadolinium Gd Alumina 1312Gallium Ga Alumina, PBN? 30Germanium Ge Alumina 959Gold Au Alumina 1063Holmium Ho Alumina Ta 1470Indium In Alumina 157Indium Tin Oxide ITO Alumina 1800 SIron Fe Alumina 1535Lanthanum La Alumina Ta, W 887Lead Pb Alumina 328 TLithium Li Alumina 179 H,OLithium Fluoride LiF Alumina Ni 845Lutetium Lu Alumina 1656

O



Magnesium Mg Alumina 651 S OMagnesium Fluoride MgF2 Alumina 1261Manganese Mn Alumina 1244 SNeodymium Nd Alumina 1021Nickel Ni Alumina 1455Palladium Pd Alumina 1555 SPhosphorus P Alumina 597 S T,FPolonium Po Quartz 254 TPotassium K Quartz 64 FPotassium Bromide KBr Alumina 734Potassium Chloride KCl Alumina Ni 770 SPotassium Iodide KI Alumina 681Praseodymium Pr Alumina Ta 931Rubidium Rb Quartz 38 FSamarium Sm Alumina 1072 OScandium Sc Alumina 1397 OSelenium Se Arc‐Coated SS 217 TSilicon Si Alumina Ta 1410Silicon Monoxide SiO Alumina Ta 1702 SSilicon Nitride Si3N4 Alumina 1900Silver Ag PBN, Alumina Mo 961Sodium Na Quartz 98 FSodium Fluoride NaF Alumina 993Strontium Sr Alumina Mo, Ta, W 771Sulfur S Quartz 597 S PTellurium Te Alumina Ta, W 450Terbium Tb Alumina 1360Thallium Tl Alumina 304 TThullium Tm Alumina 1545 STin Sn Alumina 232Tin Oxide SnO2 Alumina 1630 STitanium Oxide TiO Alumina W, MoTungsten Oxide WO3 Alumina 1473 SVictawet VW AluminaYtterbium Yb Alumina Ta 824 SYttrium Y Alumina 1477Zinc Zn Alumina 419 S P

O,F

T

360.378.4137 | 360.378.4266 fax | [email protected]

NOTES:H = Hydroscopic filmsO = Oxidizes quickly in airT = ToxicF = Flammable in airP = Poisonous to vacuum systems due to low sticking coefficientS = SublimesD = Decomposes

| www.Luxel.com | P.O. Box 1879, 515 Tucker Ave., Friday Harbor, WA 98250, USA FAQ Rev.9-28-2011

APPENDIX A: Radak & OLED Furnace Technical Information

The following pages contain information gathered on Radak & OLED furnaces during laboratory testing. All furnaces were configured with Type‐C thermocouples and installed in a vacuum system evacuated to <5x10‐6 Torr. Data was logged via National Instruments LabVIEW software and DAQ hardware.

Figures 1, 2, 3, and 4 attempt to show the maximum power requirements of the furnace while it is ramping up to temperature. For Radak furnaces an appropriate ramp rate for most processes is 100C per minute; for OLED furnaces a typical ramp rate may be 50C per minute for most processes. This information may be particularly useful for customers wishing to build or use their own power supply with the furnace.

Figures 5 and 6 show average cool‐down times for Radak and OLED furnaces.

Figures 7 and 8 give an example of the effects of tuned vs. un‐tuned PID temperature control. The example emphasizes the importance of tuning your PID parameters to achieve smooth process control. An auto‐tuning feature is included in all of our Power Controller II series – the autotune function should be performed at your process temperature (i.e. deposition temperature) to achieve best results. Follow the instructions provided with your Power Controller II series.

We hope you find the following information useful as a rough guide only when thinking about your particular process. Please note that different system configurations will produce different results: factors such as crucible/source material combinations and complex or long wire runs will significantly affect performance.

Please feel free to contact the Luxel technical support team regarding your specific application: [email protected]

Luxel Corporation

515 Tucker Ave

Friday Harbor, WA 98250

(360) 378 4137


mailto:[email protected]










EXAMPLE RADAK PID PARAMETERS:

PID Parameter Name Parameter ValueGain Scheduler Setpoint 300

Exam

ple of Un‐tune

d PID Value

s Proportional Band Set 1 15 Integral Time Set 1 30s Derivative Time Set 1 15s Manual Reset Set 1 0 Cutback High Set 1 15 Cutback Low Set 1 20

Exam

ple of Tun

ed PID Value

s Proportional Band Set 2 141 Integral Time Set 2 11s 137ms Derivative Time Set 2 1s 856ms Manual Reset Set 2 0 Cutback High Set 2 225 Cutback Low Set 2 225

These table values were used to generate the data plotted on the previous page.

Follow the tuning instructions in the power controller manual to produce the most appropriate PID values for your operating range.

Documents

Frequently Asked Questions (FAQ) - Luxel Corporationluxel.com/wp-content/uploads/2011/10/Radak-FAQ-Rev9-28-2011.pdf · Frequently Asked Questions (FAQ) Radak & OLED Furnaces Table