Skip to content. | Skip to navigation

Sections
Personal tools
You are here: Home / Equipment / Dry Etching / STS Deep RIE Etcher, stsetch (semi-clean) / STS Deep RIE Etcher, stsetch

STS Deep RIE Etcher, stsetch

This is a IPC (Inductive Charged Plasma) Deep Reactive Ion etcher from Surface Technology Systems. The platform is single-chamber, manual loadlock system. The etch process is based on the patented Laermer and Schlip process, commonly referred to as the Bosch Process. The etch process alternates between the passivating C4F8 plasma and the silicon etching SF6 plasma.

Picture

 

STS Deep RIE Etcher

 

Background


The following information is provided to SNF labmembers as an aid in making the decision to use or get training on the STS RIE tool. The information provided here is the result of several years of working with the tool and encountering many different types of inquiries about the machine. It is also based on the mistakes of past users of the tool.

 

 What the tool is and is not

Whether to use the STS etcher or not is based on what the final structure will be. Just because it etches Si very quickly does not mean it is the best tool for the desired outcome. For example; it is not an alternative to CMP, it is not a wafer scriber, nor is it a resist asher. It may not be the best choice for trench work either; users wanting to put electrically active devices near the trench should look into using another method of etching.

It is not a finesse machine. The Bosch process etches though Si leaving scalloped sidewalls. Some structures may exhibit a swirling etch pattern at the inside corners of the device. The bottom surface of the etched area is not smooth. Larger areas etch faster than smaller ones. And very large open areas of Si may show grass or micromasking.

Although the ideal amount of exposed Si is about 20% (evenly distributed) for the standard programs, any questions about mask layout should be directed to staff. There are some designs which will guarantee the wafer will not survive the etching.

 

A brief description of the machine chuck and clamp

The SNF STS machine uses a mechanical clamp to hold the wafer to a chuck. This clamp is made up of eight ceramic 'fingers' which bear down near the edge of the wafer to hold it onto the chuck. The chuck provides He backside cooling to the wafers. This keeps the wafer cool enough during the plasma process so that the resist does not etch very quickly (selectivity to Si is about 75:1) and also minimizes resist burning.

The wafer to chuck interface is aided by a lipseal (like an o-ring), which stops the He from leaking into the chamber and changing the etch rate. The backsides of the wafers must be clean of resist, debris, tape or any other material which would break the seal. It is best to think of this seal as being similar to a vacuum seal.

Additionally, the wafer must free of resist developer. As with all the etchers in the facility, the STS chuck may be corroded by the developer. To avoid damage to the chuck all wafers should go through a full six-cycle rinse after development and be inspected before placing into the etcher.

Helium is delivered to the bottom of the wafer at about 10 torr of pressure. Coupled with the downward pressure of the clamp fingers, it is easy to see that the wafer is subjected to a large amount of force. This force must be kept in mind when etching, for as the etch goes deeper, the wafer gets weaker. In many cases wafers with 400µm of remaining Si thickness have broken in the chamber. We therefore require that for etches that will leave less than 400µm of Si remaining on the process wafer be bonded (using photoresist) to a support or carrier wafer using the procedure as outlined in the following link; stsetch/stsprep .Wafers must be bonded and not simply placed over another wafer. The bonding aids in thermal transfer enough to keep the resist from burning.

This obviously includes wafers with a starting substrate thickness of less that 400µm. Support wafers should be the standard 500-550µm thick.

Exceptions to these policies are decided on a case-by-case basis by responsible staff members. Labmembers must consult with staff and not assume that another labmember's process is the same as theirs.

 

  Final comments on STS Etcher Use

It should be noted that the STS etcher is one of, if not the, most heavily used pieces of equipment in the lab. Any labmember trying to make a reservation on the machine will attest to that fact. Accordingly, any downtime on the machine affects subsequent users and can set back timetables or cause the loss of hard fought for reservations.

A user should use the machine shutdown option if a wafer is broken in the chamber, even if the broken piece is not on or around the lipseal. As pieces continue to etch they may become light enough to be sucked into the vacuum system.

A shutdown should be initiated if the turbo pump fails or the software freezes. In addition to alerting the maintenance personnel, the shutdown informs off-site labmembers there is a problem which may affect their reservation.

Obviously, if a labmember observes the etcher is operating in an unsafe fashion they should shut it down. Any deviance from the norm of 1) chamber pressure, 2) gas flow, 3) He pressure or flow or 4) RF conditions should be reported to staff via the machine problem option on coral.


If there are any questions regarding layout design, previous processing steps completed before etching, machine capabilities, resist bonding issues, Si etching issues or general machine issues, please consult Nancy Latta , latta@snf.stanford.edu .

 

 Stsetch Frequently Asked Questions

  1. What factors determine whether or not the STS Deep RIE etcher is the appropriate tool for the work I want to perform?

Deep etching refers to long etches, typically >60 mins, which attempt to etch a significant fraction of the wafer thickness (>1/4). For standard 400-500um thick wafers, this addresses etch depths >100um. Key issues include;

    1. wafer structural integrity
    2. etch rate and uniformity.
    3. mask selectivity etch rate and uniformity.

Imagine the wafer when the etch is complete. Can it withstand a force caused by 10 torr backside pressure and 30 mtorr frontside? Can it withstand the pressure of the mechanical clamp fingers on its edges? Are there any small chips which could come loose? Any scribe lanes which might form a cleave plane?

Mask design considerations should include the number and size of the features to be etched and their proximity to the edge of the wafer.

A good rule of thumb is to etch no deeper than 100um without the special holder. Using the holder and guard ring greatly reduces the chances that the wafer will break in the chamber requiring about a full day's shutdown of the etcher for removal and cleaning of the chamber.

 

2) What is the etch rate of Si? Oxide? Nitride? Photoresist?

The etch rate of Si depends on a few things; feature size (larger features etch faster than smaller ones), the amount of Si that is exposed (10% or less exposed Si for a given mask slows things down, but the etch is more uniform across the wafer) and the etch recipe used. Typical recipes are :

Recipe Name

Etch Rate Range

Comments

DEEP

1.8-4 um/min

Std Bosch process

FASTDEEP

Up to 5 um/in

Sidewalls not smooth

SMOOSHAL

0.5-2 um/min

For depths of 5-20 um-micromasking (grass) may occur at longer etches

SMOODEEP

1.8-4 um/min

For smoother sidewalls

The selectivity of oxide to Si is about 100:1, depending on structures being etched.

The selectivity of nitride to Si is about 75-125:1.

Photoresist etches at about 250A-600A/min, depending on post-exposure treatment. Thick resist (7um or more) is recommended for etches lasting more that 30 mins.

 

  1. What are good etch masking materials?

Photoresist, oxide and nitride are the most commonly used etch masks. Others, such as polyimide tape (kapton), are used with staff approval of process and material compatibility.

Standard SNF metals (Al, Al/Si, Ti, and W) may be used as etch stops, but not etch masks. Ideally, metal is covered in photoresist or oxide.

 

  1. What are recommended photoresist treatments before STS etching?

For etches lasting more that 30 mins or about 100um deep, SPR220-7 must be used in a thickness of 7um, 10um or 18um. Please consult the SPR220-7 page in the Photolithography section of the Process Library. Extra hotplate bakes may be required.

For all thicknesses of SPR220-7 a post exposure bake is needed. Using the ovens a bake of 110C for 30 mins or 90C for 45-60 mins. This bake is needed to harden the resist to the long etches. The choice of higher temp is based on whether the user can tolerate reflow or misshapen enlargement of the device features. For larger features (50um or more) this is probably not an issue.

Resist thickness of 10um of SPR220-7 , when baked properly should be able to withstand about 5 hrs of etching using the DEEP recipe.

 

  1. What is the required procedure for etching through a wafer?

When etching through a wafer it is highly recommended for most cases to bond the device wafer to a blank support wafer. A user would choose not to use a support wafer if the device wafer had underlying etch stop material layers sufficient to a) structurally support the wafer, and/or b) not be completely etched away exposing the chuck to the plasma. This is somewhat risky and staff should be consult if there is any question.

The procedure for mounting a wafer to a support wafer follows:

  • Singe wafers at 150C for 30 mins.
  • Coat a backing (support) wafer with 1.6um 3612 resist using Program 8 in manual (you don't want to prebake the wafer).
  • Place your device wafer on top of the support wafer and align the flats as best you can.
  • Place the wafers on a hotplate set to 105C and place foil over the wafer to protect the surface. Put the weight on top of the foil (the weights are near the hotplate).
  • Bake on the hotplate for 1 hour.
  • Coat the wafers with SVGCOAT Program 1 on the prime station and the program for the desired thickness on the coat station (for ex. Program 4 for 7um). You may choose not to use the edge bead removal steps.
  • Expose, develop and postbake.
  • Do an additional post bake in the 90 (45-60 mins) or 110C (30 mins) to further harden the resist.

 

6. When is the special aluminum holder and guard ring used?

The aluminum holder and guard ring were developed to address issues of wafer structural integrity and machine chuck protection. The holder and guard ring are used in place of mounting device wafers to a support wafer. The holder and guard ring have the following features:

  • Wafers with bottom side materials (such as resist) are separated from the wafer chuck so that they do not heat up and stick, causing catastrophic unloads.
  • Through-etches which accidentally punch through their etch-stop membranes will not attack the wafer chuck.
  • A hole in the holder plate allows helium gas to reach the back of the wafer, so that wafers are cooled and etch at the same rate as they do without the holder.
  • The surface of the wafer on the outside edge (~outer 1cm), including the wafer thick vertical circumference edge of the wafer, etch much faster than the center of the wafer (>20% faster in some cases), so the ring guard is used to protect the wafer against breakage. Make sure the pins are aligned correctly and snug and the o-ring is seated correctly.
  • However, the backside of the wafer (~outer 1cm) is etched in the wafer holder, so many users coat the non-etch side of the wafer with resist or oxide for protection.

 

7. What are good references for additional information?

a) A .A. Ayon, R. Braff, C.C. Lin, H.H. Sawin and M.A. Schmidt, J. Electrochem. Soc., 146, 339 (1999).

 

 

 

Contact List and How to Become a User

 

Please refer to the Training Calendar to sign up for the most convenient session.

 

Operating Instructions

 

Basic System and Process Description

 

The STS is an ICP (Inductively Charged Plasma) etch system which uses the ‘Bosch’ process to etch deeply into silicon or polysilicon.  During the etching process gases toggle in user set cycles of SF6 (to etch anisotropically) and C4F8 (to deposit a sidewall protecting polymer or passivate the wafer surface).

 

A description of the tool may be found at:

 

stsetchgeninfo

 

Labmembers requesting training on stsetch must be familiar with this information.

 

Labmembers should also have read over the following Frequently Asked Questions prior to hands-on training with a staff member:

 

stsetch stsfaqs

 

 

Description of Software and Definition of Action Buttons

 

The software displays four windows in normal run mode; two informational schematic windows (ICP and Machine View) and two task windows (Transfer and Process Control-ICP).  Let’s go over each of the windows;

 

ICP View; this window show the requested and actual set points for gas flow, RF power and matching values while to system is etching a wafer.

 

Machine View; this window displays the wafer status, load-lock and chamber pressure.

 

Transfer; this window is where the users requests via buttons that a wafer be loaded, unloaded, the load-lock be vented, or the loading process be aborted.  There is also a close button which will close the window altogether.  To restore the window after an accidental close go to the toolbar and look for the window in either Control Panel or Window.

 

Process Control-ICP; this is the window where the users loads, edits or creates a recipe and controls the actual etch process.  These are the buttons and there functions;

  • Chamber LU; this is a leak-up test used mostly used by the maintenance staff.
  • Process; this button starts the recipe selected.
  • Abort; is a hard stop of a running process.  The tool goes into idle mode without going through the usual pump and purge cycles.
  • Hold; suspends the recipe by turning off the clock, RF and gases.  If pressed it will turn into Resume to start the recipe (and count down clock).
  • Skip; moves the recipe to the next step.  It is a kinder gentler way of aborting a program as it will allow the tool to go through the usual pump/purge steps.
  • Close; is defined above in Transfer.

 

Stand By Conditions

 

  1. Verify the turbo pump is on by checking the electronics rack located behind the pump.  The third panel from the top marked Pfeiffer Vacuum has a digital read out window.  It should read 390-400Hz.  If it drops below 390, please use coral to report it as a problem.
  2. Look around the loading deck for any notes.  There may be a message asking the next user of the tool remove a wafer from the chamber.  A proper storage container should be available for the wafer.
  3. The load-lock should be left vented.

 

Load and Edit a Program 

 

1) After enabling stsetch, click on the aqua colored recipe name found on the far left-hand side of the Process-ICP window.  This will open the file that contains all the recipes.  Search the file until you find the desired recipe.  Be careful; many recipes have very similar names.  Be sure to choose the correct one.

 

► This loads the recipe.

 

2) Click on ‘Recipe’ on the right-hand side of the Process-ICP window.  This will bring up the chosen recipe.

3) The file should open on the page containing the Standby Step parameters.  This step is run automatically before and after the etching step.  Most recipes have this step set to these values;

 

Standby Gas-ARGON

Flow Rate           0sccm                         (Ar is not even plumbed to the tool)

 

Pumpdown Time  00:00:20

Purge Time           00:00:10

Pump Out Time    00:00:30

 

Base Pressure   4mT

Pressure Trip   94mT                             (The pressure operating range)

4) Click on the recipe name on the left-hand side and then on the tab labeled General to get to these input fields;

 

Pumpdown Time   00:00:30

Gas Stabilization   00:00:10  

Process Time        (variable)                 (User defined time)

Pump Out Time     00:00:30

 

Parameter Switching         (The box must be checked for the Bosch process)

○ Etch First

○ Pass First                           (User chooses whether to etch or passivation first)

Etch Time

Pass Time

Overrun                       (This is where both gases, SF6 and C4F8 are on at the same time)

 

5) Click on Pressure to get to the pressure page.  APC (automatic Pressure Control) is set to manual.  It’s a long story, but we lost the ability to use APC when the turbo pump was upgraded.  The APC setting is a percentage of how open the valve is.  A one percent change is significant.

 

Base Pressure     1mT             (This is a change from the Standby page)

Pressure Trip    94mT

 

6) Click on the Gases to get to the gases page

 

SF6   Flow   Tol%      (Tolerance should be set to around 20% to avoid an error)

C4F8 Flow   Tol%

 

7) Click on RF to get this page.  There are two generator sources, Coil and Platen (chuck). The coil generator is on for both the etch and passivation steps, but the platen generator is on only during the etch step.  This gives directionality to the plasma and creates approximately 90° sidewalls.

 

Platen Generator

                        Etch                 Passivate

Power          (Typical value is about 120W.  This is actually only 12W as there 10X scale)

Tolerance

 

Coil Generator

                        Etch                 Passivate

Power           (Typical value is 600W.  The 10X scale is off)

Tolerance

 

In both cases the matching should be chosen and the Load and Tune set to 50%.

 

8) Click Endpoint to get to the endpoint page.  We do not have an endpoint detector on the system, so the disabled box is checked.

 

► This completes the review of a recipe.

 

NOTE: We strongly encourage to user to check his/her recipe before every use.  Any user can change any parameter of any recipe at any time.  Please protect your work by verifying the recipe is correct.

 

If you change a parameter of a recipe, for example Time, please use the EXIT option on the toolbar to exit.  As you do this you will be asked if you want to saves changes.  Answer yes.  You will then be shown the file you are about to change.  Answer yes.  You will be prompted that the file exists and asked if you want to over write it.  Answer yes.

 

We also encourage users to condition the chamber by running 10-15 mins of the desired recipe.  Additionally users may want to run the O2CLEAN recipe for 20 mins prior to the conditioning run to strip polymer in the chamber.  This may be important to users concerned with grass (micromasking) formation.

 

►This completes the editing of a recipe.  For more information about recipes, please refer to the More on Recipes section of these instructions.

 

 

Starting an Etch

  • The load lock should be vented.  Check the loading area for any notes from the previous user.  They may have left a wafer in the system and have requested that you remove it.  In that case VENT and UNLOAD the wafer.
  • Load your wafer into the load lock with the flat at the 9:00 position if the wafer indent on the loading plate is imagined to be a clock face.  This assures that the flat of the wafer will not be under a clamp finger and avoids breaking the wafer, or worse, breaking the clamp finger.
  • Click on LOAD in the transfer window.  This will start pumping down the load lock and will position the wafer on the chuck in the process chamber.
  • Once loading has occurred, push PROCESS in the Process Control-ICP window.  Following the stand-by step, the plasma will ignite and etching has started.  You can verify this by looking into the chamber.  Note; the plasma will appear to flicker.  This is good.  What you are seeing is the gases and RF changing as per the Bosch process.

At the Completion of the Etch

  • Once the etching is complete the stand-by step will again pump and purge the chamber.  When completed, UNLOAD will become an option in the transfer window.  Click it to remove the wafer from the chamber.
  • Once the wafer is back in the load lock click on VENT to vent the load lock and remove the wafer.
  • You are now back to tool’s stand-by condition.  You may disable the etcher.

 

The Paperwork

 

Users of the tool are asked to fill out the log form located in the binder near the tool.  You will be asked for the following information;

Date, Coral ID and Recipe used.

He Flow (after wafer loads) and He Flow (after plasma starts)- these are two checks of the He backside cooling system.   He flow should be between 3.7-4.0 Sccm.  There will be a diffenece from no wafer to a wafer loaded and no plasma to a wafer with the plasma on.  Recording this gives us a check of how the He cooling system doing.

The other thing that is being checked here is that those numbers are steady and not fluctuating.  If they are, there is a He leak.  This means the He is escaping and your wafer is getting hot, you lost control of the etch and it is no longer anisotropic. 

To fix this problem do the following;

  • Stop the etching by clicking on SKIP in the Process Control-CP window. 
  • UNLOAD and VENT. 
  • Take a look at the back of the wafer for any resist, particles, or pieces that might interfere with the maintaining a good pressure seal on the chuck. 
  • Remove to make sure the back of the wafer is clean and re-LOAD the wafer. 
  • Check to see that the He pressure and flow are steady.

You may have to do this procedure more that once.  Examine the wafer carefully; there may be small cracks or you may have punched through a feature.

 

Chamber Pressure Passivation/Etch- you are asked to note the chamber pressure during the etching as a check status of the vacuum system.  This can be an indicator that the lines are beginning to constrict and needs attention.

 

Wafer #’s- in order to look at the log of the etching run we will need to have the wafer number.  There is no time/date stamp.  The number is located in the transfer window.

 

Comments- this is the place to note comments during the etch.  Obviously, equipment problems should be noted in Coral.

 

 

 

More on Recipes

 

There are four staff maintained recipes; DEEP, SMOODEEP, SMOOSHAL, FASTDEEP and O2CLEAN.  This means that users are welcome to use these recipes, but if any parameter is changed besides Time, the user is requested to write their own recipe.

 

Here are the four recipes with a basic explanation of why they were created and when a user might want to use them;

 

 

Recipe Name

Etch

Passivation

DEEP

SF6 Flow- 130 sccm

C4F8 Flow- 85 sccm

 

RF Coil- 600W

RF Coil- 600W

 

RF Platen- 120W

RF Platen-0W

 

Time- 12 secs

Time- 7 secs

This is a basic Bosch process recipe.  It has been the starting point for many users who have gone on to customize and write their own recipes.  The etch rate is very much dependant on the size of the features to be etched and the amount of wafers surface that is exposed silicon.  Recent testing using patterns with about 15-20% exposed silicon has the etch rate at about 1.8-2.0µ/min. 

 

Users are very strongly encouraged to establish the etch rate of critical features for each patterns (mask level) etched.

 

 

 

Recipe Name

Etch

Passivation

SMOODEEP

SF6 Flow- 130 sccm

C4F8 Flow- 85 sccm

 

RF Coil- 600W

RF Coil- 600W

 

RF Platen- 120W

RF Platen- 0W

 

Time- 9 secs

Time- 7 secs

 

This recipe was developed in an effort to smooth out the ‘scalloped’ side walls when using the DEEP recipe.  In DEEP there are roughly 10 scallops per micron of etch depth.  SMOODEEP has a faster gas toggle rate and hence there are more scallops per micron.

 

 

 

Recipe Name

Etch

Passivation

SMOOSHAL

SF6 Flow- 130 sccm

C4F8 Flow- 120 sccm

 

RF Coil- 600W

RF Coil- 600W

 

RF Platen- 120W

RF Platen- 0W

 

Time- 6 secs

Time- 5 secs

 

SMOOSHAL is a recipe to be used for etch depths of less than 20µ.  Micromasking (grass) may occur at deeper etch depths.

 

 

 

Recipe Name

Etch

 

O2CLEAN

O2 Flow- 40 sccm

 

 

RF Coil- 600W

 

 

RF Platen- 80W

 

 

Time- 30 mins

 

 

 

O2CLEAN is a recipe that is used primarily after the chamber has been exposed to air, say, after a broken wafer has been cleaned up.  Users are encouraged to use it before critical etches, but need to be advised to also run the desired etch program for 10-15 mins after O2CLEAN to condition or season the chamber.  The clean is also done before running the process quals.

 

 

 

 

Training Checklist

 

£ Be familiar with basic equipment and Bosch process description.

 

£ Be aware of the proper use of the mechanical clamp and its’ limitations.

 

£ Have read the general information and FAQs.

 

£ Know the machine stand-by conditions.

 

£ Know the four windows of the software and the functions of action buttons in the task windows.

 

£ Demonstrate correct loading and editing of recipes.

 

£ Know how to position and load a wafer into the process chamber.

 

£ Know how to start a recipe and how to stop the process before completion.

 

£ Know when to use the holder vs using a carrier wafer.

 

£ Know what faults should be recorded in coral as a problem or a shutdown.

 

£ Demonstrate properly filling out the log sheet.

 

 

 

 Process Monitoring and Machine Qualification

 

 

Purpose: to provide verification and trend of the DEEP program including etch rates of silicon and photoresist, selectivity of those materials, and within-a-wafer uniformity of etch.

 

Frequency of the test: to be completed after major maintenance such as cleaning or gas line replacement or on a set schedule to be determined or as needed based on user feedback.

 

Documentation of results: to be posted on the log sheet located in the assigned book at the tool, posted in the equipment archive for amtetcher on the SNF website and in a file or webpage available to users in data or chart format.

 

Use wafers with 3 or 7 um layer of SPR220 photoresist.  The wafer should have the resolution mask used in litho area.

  1. Measure photoresist thickness using one of the Nanospecs.  Be sure to use reference wafer before testing to calibrate the tool.  Take readings for the Center, Top, Bottom, Right and Flat positions of the wafer.  Readings should be taken about 20mm from the edge.  Use Statistics feature to get mean value.  See Fig.1.
  2. Season the chamber for 10 minutes using the DEEP.  Be sure to check the program parameters before starting.
  3. Etch the wafer for 10 minutes.  Monitor gas flows, RF power and chamber pressure during passivation and etch.  Record on Process Qualification Log Form.
  4. Measure photoresist thickness using the same Nanospec.  Be sure to use reference wafer before testing to calibrate the tool.  Take readings for the Center, Top, Bottom, Right and Flat positions of the wafer.  Use Statistics feature to get mean value.  Readings should be taken about 20mm from the edge.  Record in the Log.
  5. Subtract the mean values of the pre-etch measurements from the post-etch measurements.  This gives you the amount of photoresist etched.
  6. Measure the depth of the etch using one of the calibrated microscopes.  Focus first on the photoresist then, using a smooth motion, focus on the bottom of the etch. Choose a larger feature and focus in a corner.   This may be somewhat difficult as the etch is not very deep and not too roughed up.  Two tricks: 1) look for texture or 2) ‘crawl’ down the sidewall to the bottom. 
  7. Take readings in about the same 5 places as you took readings for the photoresist. Once you get consistent readings in each spot, record the measurement in the Log.
  8. Subtract the photoresist measurement from Step 6 from the etch measurement to get the etch depth.
  9. Record results on the Monitor Log. Record results in data file to get within-a-wafer uniformity.
    • To get the total amount of photoresist etched, subtract the average of the second reading from the average of the first reading.
    • To get the etch rate per minute divide the etch measurements by 10.
    • To get the selectivity of silicon to photoresist divide the etch rate of Si by the etch rate of PR.

 

   Recent Results

 

Graphs of recent results are posted near the tool and updated as needed.

 

DEEP Recipe for 00:10:00 is used unless otherwise noted.

 

ate Si Etch Rate
PR Etch Rate
SI : PR Selectivity
Comments
12/07/2005 3.36um/min - - Before annual PM
01/03/2006 3.11um - - After annual PM
04/21/2006 2.18um 285A/min  76 : 1
New resolution mask used- Opposite polarity as before (20% exposed Si)
05/10/2006  2.45um  257A  95 : 1
 
 06/01/2006  1.7um  268A  63 : 1
 
 09/19/2006  2.18um  -  -  Nanospecs down- unable to measure PR
 11/17/2006  2.37um  269A  88 : 1
 
 12/06/2006  2.30um  269A  85 : 1
 
 01/03/2007  -  251A  -  Opps- wafers etched again before Si measured
 02/02/2007  -  266A  -  Double opps- did it again...
 05/22/07  2.9um  216A 125 : 1
 Test after C4F8 change out
 07/05/2007  2.6um  199A  130 : 1
 After pump package change
 08/21/2007  2.49um  232A  107 : 1
 Test after C4F8 change out.  Used Zygo for Si measurement
 10/10/2007 2.4um
 230A  104 : 1
 Zygo
 11/01/2007  2.8um  212A  127 : 1
 Zygo
 01/03/2008  2.43um  216A  123 : 1
 After PM- Zygo
 02/14/2008  2.37um  203A  117 : 1
 Zygo
 03/04/2008  2.54um  207A  123 : 1
 After C4F8 mfc's replaced- Zygo
 05/06/2008  2.80um  201A  139 : 1
 Zygo
 07/15/2008  2.40um  226A  106 : 1
 Zygo
 09/23/2008  2.69um  235A  115 : 1
 Zygo
 11/18/2008  1.88um  248A  76 : 1
 After turbo pump replacement
 1/27/2009  -  246A  -  
 5/20/2009  1.6um  254A  65 : 1  Before SF6 calibration (old 130 sccm setting, actual +65% (214 sccm))
 5/20/2009  1.5um  239A  64 : 1
 Before calibration (set to actual 130 sccm, 45 sccm)
 6/2/2009  2.0um  233A  86 : 1
 Re-test after SF6 calibration, 213 sccm ('old' 130 sccm)
 6/2/2009  1.9um  270A  70 : 1
 Re-test after SF6 calibration, 310 sccm (actual!)
 8/11/2009  -  234A    Ck after wafer slice removed
 8/12/2009  -  234A  -  Oxide etch ck- 72A/min, sel = 3 : 1, ox : pr
 9/1/2009  2.12um  296A  72 : 1
 Wafer #1- ck before chamber clean and PM
 9/1/2009  1.91um  245A  80 : 1
 Wafer #2- ck before chamber clean and PM
 9/2/2009  1.97um  293A  67 : 1
  Wafer #1- ck after chamber clean and PM
 9/2/2009  2.21um  323A  68 : 1
  Wafer #2- ck after chamber clean and PM
 12/11/09 2.75um
 254A  108 : 1  Before annual shutdown 2009
 1/5/10  2.41um  262A  92 : 1
 After annual shutdown
 2/23/10  2.25um  228A  98.1 : 1
 After He valve repair
 3/23/10  1.81um  245A  73.9 : 1
 
 3/23/10  2.02um  206A  98.1 : 1
 Used holder- reports of slower ER's
 4/2/10  1.81um  211A  85.8 : 1
 Check after chuck rebuild
 4/2/10  2.09um  198A  105.6 : 1
 Check after chuck rebuild- Used holder
 4/30/10  2.41um  238A  101 : 1
 Routine ER check
 5/27/10  1.71um  185A  92 : 1
 Low etch rates reported- He flow and pressure off
 6/8/10  2.02um  212A  95.3 : 1
 After field service visit.  Spring loaded 'pips' removed from lip seal.  Will be replaced by less stiff springs.
 6/15/10  1.97um  208A  95 : 1
 Pip springs replaced and clamp finger dot replaced.  Uniformity ok.
 7/8/10  2.06um  252A  82 : 1
 After actuator switch on He valve adjusted.
 7/8/10  1.73um  169A  102 : 1
 Temp dot test for 60 mins, did not reach 130C.  Ave depth 104um.
 7/13/10  2.04um  119A  171 : 1
 Reinstalled the electrode teflon spacers (pips).  Adjusted the He bypass flow and also adjusted the He flow restriction valve to the chamber
 8/18/10  1.93um  140A  138 : 1
 Before lip seal change
 8/18/10  1.98um  119A  166 : 1
 After lip seal change
 8/19/10  1.68um  192A  88 : 1
 Investigations of reports of decreased Si ER and increased PR ER.  Also looking into the SI ER decresing with each run. Pr ok, but Si is low. 10 min etch.  Zygo
 8/19/10  1.7um  144A  118 : 1
 See above.  60 min etch.  Si read by microscope
 8/19/10  1.57um  145A  108 : 1
 See above.  10 min etch.  Zygo
 8/20/10  1.48um  151A  98 : 1
 Used temp dots, less than 54C.  60 min etch, read by microscope.
 8/24/10  -  -  -  Used DEP TEST recipe to get dep rate, 10 mins.  Dep rate = 1172A/min, nanspec with oxide program.
 8/24/10  1.73um  146A  118 : 1
 Looking into heat as a cause of decreased Si ER.  Run DEEP for 60 mins, then ran Dep rate test, and two 10 min quals.
 8/24/10  1.74um  136A  128 : 1
 See above.
 8/27/10  -  -  -  DEP-TEST run again, same as 8/24 except APC set to 66%, chamber press = 17mT, Dep rate = 1050A/min
 9/30/10  2.12um  205A
 103 : 1
 Tests after He metering valve replaced; He flow = 3.0max, read out 2.4sccm, He press = 9.79.
Etcher idle more that 10 hours, 20 DEEP condition followed by six wafers, DEEP 00:10:00.  Wafers run one right after another.
  9/30/10  2.02um  200A  101 : 1
 
  9/30/10  1.9um  187A  102 : 1
 
  9/30/10  1.9um  188A  101 : 1
 
  9/30/10  1.78um  177A  101 : 1
 
  9/30/10  1.74um  176A   99 : 1
 
 10/5/10  1.81um  185A   98 : 1
 He flow = 3.8max, read out 3.66, He press = 9.77 Wafers run one right after another.
 10/5/10  1.77um  182A   97 : 1
 
 10/6/10  1.92um  210A   91 : 1
  He flow = 3.0max, read out 2.47, He press = 9.78 Wafers run one right after another.
 10/6/10  1.84um  192A   96 : 1
 
 10/7/10  1.83um  234A   78 : 1
 Tests done after SF6 reservoir by-pass installed to track down decreasing Si etch rate as wafers run sequentially.  Tool idle for 15 hours, 20 min conditioning, DEEP 00:10:00.  Slight increase in PR etch rate may be due to SPR220.  Earlier tests done using 3612.
 10/7/10  1.68um  230A   73 : 1
 
 10/7/10  1.72um  227A   76 : 1
 
 10/7/10  1.65um  216   76 : 1
 
 10/13/10  1.67um  246A   68 : 1
 Tests after turbo replaced and exhaust lines cleaned.  SF6 by-pass still in place, 20 min conditioning, DEEP 00:10:00.
 10/13/10  1.63um  258A   63 : 1
 
 10/13/10  1.67um  248A   68 : 1
 
 10/13/10  1.67um  236A   71 : 1
 
 10/14/10  1.74um  243A   72 : 1
 Dry pump replaced, SF6 by-pass still in place, 30 min conditioning, DEEP 00:10:00.
 10/14/10  1.74um  243A   72 : 1
 
 10/14/10  1.71um  243A   70 : 1
 
 1/11/11  2.0um  85A  235 : 1
 After annual winter shutdown.  Three wafers run for 00:10:00 using DEEP.
 1/11/11  2.0um  73A  274 : 1
 
 1/11/11  1.96um  69A  284 : 1
 
 1/11/11  1.75um  171A  102 : 1
 Wafer run for 01:00:00 using DEEP
 2/10/11  1.90um  207A  92 : 1
 Monthly qual- wf #1
 2/10/11  1.86um  206A  90 : 1
 Wf #2
 2/10/11  1.92um  204A  94 : 1
 Wf #3
 2/11/11  1.95um  210A  93 : 1
 After turbo replaced
 3/14/11  1.82um  199A  91 : 1
 March qual- 3 wafers
 3/14/11  1.88um  199A  94 : 1
 Wf #2
 3/14/11  1.82um  210A  87 : 1
 Wf #3
 4/1/11  2.13um  211A  101 : 1
 April qual- one wafer quick check
 5/11/11  1.63um  186A  88 : 1
 After lip seal replaced. chips removed from chamber
 5/12/11  1.60um  204A  78 : 1
 After tool use- looking for improved ER's
 5/26/11  2.10um  217A  97 : 1
 One wafer quick check
 8/10/11  1.69um  148A  114 : 1
 Test after reports of undercutting and burnt resist
 8/11/11  2.07um  172A  120 : 1
 Test after repairs; He pressure turned back to 9.9T, lip seal replaced and chips removed from around wafer lifter.
 8/18/11  2.1um  198A  106 : 1
 Test after report of lower than usual Si ER.  Test ok.
 9/15/11  2.01um  224A  90 : 1
 Monthly qual
 10/4/2011  1.81um  194A  93 : 1
 Monthly qual and user complaints of wf breakage and fast etch rates
 10/11/11  1.57um  174A  90 : 1
 Qual after lifter burr polished off.  Wafer breakage has gone down.
 11/1/11  2.01um  256A  78 : 1
 Monthly qual.
 2/14/12  2.11um  236A  89 : 1
 Quals after annual shutdown and lab renovations.  Wafer #1
 2/14/12  2.11um  232A  90 : 1
 Wafer #2
 3/2/12  2.3um  259A  89 : 1
 
 3/29/12  2.12um  285A  74 : 1
 Monthly qual
 4/26/12  2.38um  287A  83 : 1
 After He flow set back to 3.75 from 5.0
 6/8/12  1.98um  234A  85 :1
 
 6/27/12  2.24um  232A  97 : 1
 After C4F8 replaced
 7/6/12  1.96um  221A  89 : 1
 After C4F8 line flushed
 7/27/12  2.27um  241A  94 : 1
 He flow a bit high- 4.03
 10/2/12  2.06um  240A  86 : 1
  Monthly qual
 11/10/12  2.38um  233A  102 : 1
  Monthly qual
 1/8/13  1.96um  258A  76 : 1
 Qual after annual Winter shutdown
 10/17/13  2.16um  175A  123 : 1
 
 12/5/13  2.6um  259A  100 : 1
 Qual before Winter Shutdown
 1/10 14
 2.57um  236A  109 : 1
 Qual after Winter Shutdown
 2/7/14  1.99um  285A  69 : 1
 Qual after user report of low ER.  SF6 unstable during etch cycle.

 

                

 

 Process Modules

Preparing Wafers for Through Etching

If you etch holes completely through the STS, you need a support wafer attached to the back. Follow the steps below:

Note: Good resist adhesion is dependent on a dry surface. If the wafers haven't recently come out of a high temperature furnace (out for more than 1 hour), they should get a singe at 150C for 30 minutes.

  1. Singe wafers @150C for 30 minutes.
  2. Coat a backing (support) wafer with 1.65um 3612 resist using Program 8 in manual mode (you don't want to prebake the wafer).
  3. Place your device wafer on top of the support wafer and align the flats as best you can.
  4. Place the wafers on a hotplate (set at 105C) and place foil over the wafer to protect the surface. Put the weight on top of the foil (the weights are near the hotplate).
  5. Bake on the hotplate for 1 hour.
  6. Coat your wafer with the SVGCOAT using program 1 on the prime station and program for the desired thickness on the coat station ( for example Program 4 for 7um). You may choose not to use the edge bead removal steps.
  7. Expose, develop and postbake.
  8. Do an additional post bake in the 90C (45-60 mins) or 110C (30 mins) to further harden the resist.

 

Document Actions