Steven,
your philosophy is mine too, if my car is reluctant to my desire (yes, am speaking car), I don't start with pulling out injection puter for examination, but check, by way of introduction, if I have runout of petrol, if battery is alive, if coolant leakage is at hand etc. I even use my nose in the daily merlin work.
My q was exclusively tracking down e.g. under-SiN-layer failures or deeper, for instance gate-drain migrations or shorts under bridges. You can't remove the bridge or etch it without destroing what's beneath, therefore my little passage about thinning the whole die from backside until just the metallisation system remains. We do it, but fuming acids give me headache and red eyes (am kidding, sorry BB, won't do that again). You may not have experience from this field, couldn't resist to try if the door was open...
Your method to use LOMs and darkf/coax is perfomance on highest level, need many years of experience. My favorite is a 15 year old Wild M3 with only three magnifications! But I also have a Nikon L150 with diode illumination and UV lamp + digital camera and a lot of stuff, but the M3 is used more frequently!
A strong side beam in parallel with sample is excellent for decamoflaging superthin contaminations, agree, but you need a skilled eye and a little imagination + intuition.
About sticking picking, we had troubles a few years ago, learned that discs had to come fresh and right into N2 cabins, and consumed asap. The jelly ages rapidly and like 200,000 viscosity is snaily, sooner or later it bites and sets. My two drakma (US is great this year, but still China a horselength in front.
Mastic>etch resist>u misunderstanding>skip q
friend busy with pull-ejection-pick-click-q as allday work. If you have spec q or loosing temper and hair, plz shout and I'll bounce to him...
Thx, don't sniff to much aceton, and use no cigars in the cleanroom
Ingemar
PS. offline about backetch, if you want...
-----Original Message-----
From: TechNet [mailto:[log in to unmask]]On Behalf Of Creswick, Steven
Sent: den 20 augusti 2004 13:58
To: [log in to unmask]
Subject: Re: [TN] Picking Die
Ingemar,
I am not totally sure that I understand your question, so if my response does not adequately respond to your inquiry, please let us have another try at it.
As I mentioned to Joyce, I would do this [destructive test] on a few chips that had stuck in the waffle pack. My objective is simple - just like your use of an eraser to investigate wire bonding surface characteristics - I try to understand the source of the problem, so that I can reasonably figure out what I may need to do to apply a remedy. I may not know exactly WHAT it is, but knowing that SOMETHING is, or is not there, goes a long way to understanding the problem [or re-directing the investigation].
My pseudo-analysis technique in this case would initially be VISUAL only. I can see a great deal under an oblique light, with a low power microscope. If I don't see anything, I switch to coaxial lighting, and/or greater magnification. If I can not see anything in the range of 10X-200X with oblique and coaxial illumination - it either isn't there, or it is SEM time [or I have not yet had my spot o' tea].
Note: My pet peeve is when you give something to someone to look at, and the first thing they do is put it under the microscope and crank the magnification to maximum. As the old song goes, "they can't see the forest for the trees". They have zoomed in so tightly that they have lost all sense of the broader picture.
Anyway, I have a SiC chip. Being essentially transparent, if there is a dot of mastic being transferred to the back of the chip during die attach [just going directly from the membrane to the end product with no waffle packs involved], it is really difficult to see the mastic on the back of the chip [even when in the product, and silhouetted against the conductive adhesive. However, if I cure the adhesive, THEN look through the chip I will see a amber/brown spec if the mastic is present. Ditto, if I take the chip and flip it belly-up [before sticking it down into the adhesive] and simulate an adhesive cure.
Since some of my chips are still 'scored and broken', this can add another variable to consider.
Now, with Si, or all the other opaque materials such as GaAs, AlInGaP, InGaN, etc the illumination and angle of observation becomes critical. There is no secret that I am aware of, other than trial and error.
Finally getting to what I believe is your question - would the mastic act as an etch resist? Don't know, but is interesting thought. I wonder if there is actually enough physical mastic present to act as an etch resist. It would be interesting to try. I don't even know if we have HF acid here. The chemistry folks have just about every other chemical under the sun, but they are not to up to speed on semiconductor type stuff. I am not really sure whether HF would etch some of the non-silicon based devices or not. Maybe there is someone out there from an analysis lab that can chime in.
There is one other aspect from the original posting that we have not asked yet - how large are the die that you are having problems sticking in the waffle packs? Teeny tiny 0.008-0.012" diodes, or chips as large as your thumbnail?
Granted, it does not take much to cause any device to stick in the cell ... but small die probably stick more frequently than large ones :-)
If the ultimate cause of "device sticking in waffle pack cell" turns out to be the mastic from the tape, then:
[1] Is it from the ejection needle? - mastic ideally in the center of the chip. Indicating improper set-up or needle selection.
[2] Is it from something else in the pick process? - like the pick up tool not being well centered about the ejection needle and maybe causing the chip to slide slightly during pick-up (maybe picking up some mastic as it moves)
[3] Is the face of the pick up tool orthogonal to the plane of the wafer and fixturing? - if not, maybe the device slides again.
[4] Is it from the dicing/sawing operation? - mastic along the edge/s of the device.
[5] Is it really new/old tape?
We have probably beat this to a pulp already with the various "what if's".
My general philosophy on this issue would be that unless the chips are extremely expensive, or there are a really large number of them sticking in the waffle pack - to just not use them!
If you have the time [or the need] to try to figure out the root cause fine ... maybe you can pass it along to an intern or something. I am sure that we have all experienced missed-picks before. I just have not been pushed to/over my threshold for action yet. That is not to say that someone else has not, however.
Watch, the problem was likely found to be a dirty waffle pack :-)
Hope that you are enjoying a nice weekend.
Steve Creswick - Gentex Corp
-----Original Message-----
From: Ingemar Hernefjord (KC/EMW)
[mailto:[log in to unmask]]
Sent: Friday, August 20, 2004 3:15 AM
To: [log in to unmask]
Subject: Re: [TN] Picking Die
Speaking dice,
you are two top experts on the stage, I take the opportunity to ask you about this one: do you practice the analysis way by backetching the chips and study them from 'underside'? What acidous process is best for GaAs resp. Si?
Ingemar Hernefjord
-----Original Message-----
From: TechNet [mailto:[log in to unmask]]On Behalf Of Creswick, Steven
Sent: den 19 augusti 2004 18:19
To: [log in to unmask]
Subject: Re: [TN] Picking Die
Joyce,
My suggestion for use of hotplate to determine presence/or not, of adhesive was considered to be a destructive test. In no way would I do this as a 100% 'fix'
Any die tested in this manner would be considered as scrap
Steve
-----Original Message-----
From: Joyce Koo [mailto:[log in to unmask]]
Sent: Thursday, August 19, 2004 9:20 AM
To: TechNet E-Mail Forum; Creswick, Steven
Subject: RE: [TN] Picking Die
depend on type of chip... adhesive residue may cause high electrical
resistance if the back is used as ground. cook on the heat plate must be a
short time... otherwise, you may cause some degradation of wirebondability
(depend upon how clean is your hot plate... you can varify cleanliness of
your chip surface before and after hot plate using DI contact angle
measurement, you will see the difference)...If I were you, I'll check all
the vacuum pick, process parameter, etc. If everything is OK as per
qualification, you still got stuck die, I'll reject the lot for sure...(not
that make me popular...)
-my cold 1.52 cents..
jk
-----Original Message-----
From: TechNet [mailto:[log in to unmask]]On Behalf Of Creswick, Steven
Sent: Thursday, August 19, 2004 6:35 AM
To: [log in to unmask]
Subject: Re: [TN] Picking Die
Karen,
Bruce is correct in that the tape mastic should not be on the backside of
the device IF properly ejected, however, sometimes that is a big "if". Many
individuals feel that you have to poke THROUGH the tape in order to eject
the device. I find that tends to transfer a small amount of adhesive, so I
elect NOT to poke through the tape. It is all a game of tip diameter,
needle travel, time, etc.
I have had similar experiences, but never to the magnitude that caused undo
grief. On occassion, I have washed waffle packs out with clean IPA and this
has helped. For those [empty] waffle packs that seem prone to die sticking
[naturally, assuming the cavity size is large enough, etc], I just trash
them.
I have a few devices which are essentially optically clear. If there is any
mastic transfer during the die ejection process, it shows up as a small
amber/brown dot after the die attach adhesive has been cured [can see it
from the top]. I guess from a silver lining stand-point this allows me to
verify just exactly where the ejection needle is contacting the device - but
that is something I can generally do without :-)
I wonder .... if your device is large enough ... could you carefully remove
it from the waffle pack with tweezers, flip it upside down on a 200-250°C
heat stage for 10-15 minutes, char the mastic a bit [if it is there], then
pull it off and look at it under the microscope. I am guessing that if
mastic were there, it might show up if the lighting were just right.
I have had situations where the mastic is along the very edges of the
device. Believe this was due to cutting too far into the nitto tape during
dicing. Like Bruce said, this has been fixed with a better vacuum.
Just goes to show that even the little things are not so trivial.
Have a good day!
Steve Creswick - Gentex Corp
-----Original Message-----
From: Karen Walters [mailto:[log in to unmask]]
Sent: Wednesday, August 18, 2004 2:14 PM
To: [log in to unmask]
Subject: [TN] Picking Die
Has anyone ever experience die sticking to the waffle paks during die
attach. Please let me know what your solution was to this problem. We
believe it is Nitto tape residue that sticks to the back of the die when
the die are picked from the wafer ring to the waffle pak
Please let me know your experience with this.
(Embedded image moved to file: pic26500.pcx)
---------------------------------------------------
Technet Mail List provided as a service by IPC using LISTSERV 1.8e
To unsubscribe, send a message to [log in to unmask] with following text in
the BODY (NOT the subject field): SIGNOFF Technet
To temporarily halt or (re-start) delivery of Technet send e-mail to
[log in to unmask]: SET Technet NOMAIL or (MAIL)
To receive ONE mailing per day of all the posts: send e-mail to
[log in to unmask]: SET Technet Digest
Search the archives of previous posts at: http://listserv.ipc.org/archives
Please visit IPC web site http://www.ipc.org/contentpage.asp?Pageid=4.3.16
for additional information, or contact Keach Sasamori at [log in to unmask] or
847-509-9700 ext.5315
-----------------------------------------------------
---------------------------------------------------
Technet Mail List provided as a service by IPC using LISTSERV 1.8e
To unsubscribe, send a message to [log in to unmask] with following text in
the BODY (NOT the subject field): SIGNOFF Technet
To temporarily halt or (re-start) delivery of Technet send e-mail to
[log in to unmask]: SET Technet NOMAIL or (MAIL)
To receive ONE mailing per day of all the posts: send e-mail to
[log in to unmask]: SET Technet Digest
Search the archives of previous posts at: http://listserv.ipc.org/archives
Please visit IPC web site http://www.ipc.org/contentpage.asp?Pageid=4.3.16
for additional information, or contact Keach Sasamori at [log in to unmask] or
847-509-9700 ext.5315
-----------------------------------------------------
---------------------------------------------------
Technet Mail List provided as a service by IPC using LISTSERV 1.8e
To unsubscribe, send a message to [log in to unmask] with following text in
the BODY (NOT the subject field): SIGNOFF Technet
To temporarily halt or (re-start) delivery of Technet send e-mail to [log in to unmask]: SET Technet NOMAIL or (MAIL)
To receive ONE mailing per day of all the posts: send e-mail to [log in to unmask]: SET Technet Digest
Search the archives of previous posts at: http://listserv.ipc.org/archives
Please visit IPC web site http://www.ipc.org/contentpage.asp?Pageid=4.3.16 for additional information, or contact Keach Sasamori at [log in to unmask] or 847-509-9700 ext.5315
-----------------------------------------------------
---------------------------------------------------
Technet Mail List provided as a service by IPC using LISTSERV 1.8e
To unsubscribe, send a message to [log in to unmask] with following text in
the BODY (NOT the subject field): SIGNOFF Technet
To temporarily halt or (re-start) delivery of Technet send e-mail to [log in to unmask]: SET Technet NOMAIL or (MAIL)
To receive ONE mailing per day of all the posts: send e-mail to [log in to unmask]: SET Technet Digest
Search the archives of previous posts at: http://listserv.ipc.org/archives
Please visit IPC web site http://www.ipc.org/contentpage.asp?Pageid=4.3.16 for additional information, or contact Keach Sasamori at [log in to unmask] or 847-509-9700 ext.5315
-----------------------------------------------------
---------------------------------------------------
Technet Mail List provided as a service by IPC using LISTSERV 1.8e
To unsubscribe, send a message to [log in to unmask] with following text in
the BODY (NOT the subject field): SIGNOFF Technet
To temporarily halt or (re-start) delivery of Technet send e-mail to [log in to unmask]: SET Technet NOMAIL or (MAIL)
To receive ONE mailing per day of all the posts: send e-mail to [log in to unmask]: SET Technet Digest
Search the archives of previous posts at: http://listserv.ipc.org/archives
Please visit IPC web site http://www.ipc.org/contentpage.asp?Pageid=4.3.16 for additional information, or contact Keach Sasamori at [log in to unmask] or 847-509-9700 ext.5315
-----------------------------------------------------
---------------------------------------------------
Technet Mail List provided as a service by IPC using LISTSERV 1.8e
To unsubscribe, send a message to [log in to unmask] with following text in
the BODY (NOT the subject field): SIGNOFF Technet
To temporarily halt or (re-start) delivery of Technet send e-mail to [log in to unmask]: SET Technet NOMAIL or (MAIL)
To receive ONE mailing per day of all the posts: send e-mail to [log in to unmask]: SET Technet Digest
Search the archives of previous posts at: http://listserv.ipc.org/archives
Please visit IPC web site http://www.ipc.org/contentpage.asp?Pageid=4.3.16 for additional information, or contact Keach Sasamori at [log in to unmask] or 847-509-9700 ext.5315
-----------------------------------------------------
|
