How to decap an IC

[Grant Stockly]

This is a collection of instructions from the web on decapping ICs. I have not successfully decapped any ICs yet, but when/if I do I will post my results.

[Grant Stockly]

From: Jon Hiller : hiller-at-anl.gov
Date: Thu, 24 Jan 2002 19:26:21 -0600
Subject: IC package removal

Contents Retrieved from Microscopy Listserver Archives
http://www.microscopy.com/MicroscopyListserver/MicroscopyArchives.html

Body of Message


Does anybody have a chemical solution for removing IC packaging without
damaging the internal components? Simple grinding from the top down is not
suitable because I need all the interconnects intact. Any help in this
matter is greatly appreciated.

Sincere regards,

Jon Hiller
================================================== ================
Jon M. Hiller
Argonne National Laboratory
Materials Science Division
Electron Microscopy Center
Tel: 630-252-9558
Fax: 630-252-4798
Email: hiller-at-anl.gov
================================================== ================


From daemon Thu Jan 24 19:34:57 2002

[Grant Stockly]

From: Diane.Ciaburri-at-gd-ais.com
Date: Fri, 25 Jan 2002 08:50:58 -0500
Subject: Re: IC package removal (LONG)

Contents Retrieved from Microscopy Listserver Archives
http://www.microscopy.com/MicroscopyListserver/MicroscopyArchives.html

Body of Message

Jon,

I assume you're speaking about plastic packages since you want a chemical removal technique. Below your message is a summary of responses that I received when I asked a similar question. I can't comment on them, because my project dried up after I asked the question and I never got to try any of the techniques, but there look like some good ones.

Diane Ciaburri
General Dynamics
Pittsfield, MA


<Note from logjam, the following posts until noted by me are a part of this e-mail. I'm separating them for readability. I'm also fixing the line break issues that make the post very hard to read and doing some minor formatting>

[Grant Stockly]

<Part of the last post>

Here's the summary (long) for all those interested in deencapsulating plastic encapsulated ICs. I have no preferences as I haven't tried any yet, but thought the fuming sulfuric acid might be 'fun'. Thanks again!

[Grant Stockly]

<Part of the previous post>

-- [ From: Garber, Charles A. * EMC.Ver #3.1 ] --
Hi Diane,

The way this is generally done is to mill the plastic down on a grinding wheel to the point where only a fairly thin layer of plastic remains.

Then, using a plasma etcher, and a mixture of oxygen to CF4 (for example, 30% oxygen/70% CF4), whereby the oxygen etches away the plastic and the CF4 etches away the glass frit that is usually found in the plastic, you can remove the remaining plastic (package) without damaging the device itself. Different people like to use different gas ratios, of course, and that is probably a function, at least to some degree, of the concentration of glass frit in their particular plastic.


The SPI Plasma Prep II unit, as shown on URL http://www.2spi.com/catalog/instruments/etchers1.html in the world, is probably the most widely used unit for doing this type of operation. It is inexpensive and highly reliable, and requires virtually no maintenance.

Chuck

[Grant Stockly]

<Part of the previous post>

The ion beam approach works well. I have not used it recently on finer pitch ICs. With as-built feature sizes of 2-4u, it is fine. It will stop at the passivation and leave the Al bond wires intact. The resulting package looks like it has a V-shaped pit in it (which it does). The extent of the pit depends on the size of the die and if you want to blast down to the lead frame or substrate.

I have not done this on finer pitch devices. I would be a bit skeptical about these mostly because of the smaller bond pads. The etching would still stop at the passivation.

There are numerous places in Silicon Valley that do this on an outsource basis. Typical costs are about $75 per IC. I can get some contacts for you if you'd like.

gary g.

[Grant Stockly]

<Part of the previous post>

Diane: attached is a text document outlining the procedure my FA lab uses. Yellow fuming nitric is usually the acid of choice. If you can get a few extra parts to practice on, that would be best. And you're right, plasma takes FOREVER. If I can be of any more help, please let me know.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Becky Holdford (r-holdford-at-ti.com)
972-598-1291 (pager)
KFAB Physical Analysis Labs--SEM/FIB/FA
Kilby Center West
Texas Instruments, Inc.
Dallas, TX
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

TOOLS, EQUIPMENT, & SUPPLIES



Milling machine and appropriate end mills
Stabilo SuperFine OH pen or equivalent
Fume hood properly equipped for exhausting acid fumes and solvent vapors (see reference 3.4)
Explosion proof hot plate (see reference 3.4)
Ultrasonic cleaning apparatus
An optical microscope capable of 100X to 500X magnification, equipped with a lighting system.
Chemical resistant latex gloves
Chemical resistant laboratory coat
Chemical resistant safety glasses (see reference 3.5)
Hand tools (tweezers, scalpels and etc.)
Plastic micro-pipette
Fuming red nitric acid
Yellow nitric acid
Methyl alcohol
Acetone

1. RECORDING OF PACKAGE MARKINGS
Record all of the device markings that are on the top and bottom sides of the devices prior to starting any of the decap operations.



2. CAVITY MILLING
Determine the exact location of the chip within the package and mark the top of the device package showing the chip perimeter, using a Stabilo SuperFine OH pen and a straight edge. A SAM plot or X-ray image may be used to help determine the exact location of the chip and also to determine the thickness of the mold compound covering the chip.


Note: This should be done on devices having large chips. Devices with small chips (less than 0.125 inches in their longest dimension) do not require this step.

Mill a cavity out of the plastic package that is centered over the chip. The size of the milled cavity should typically be .050 to .100 inches larger than the length and width dimensions of the chip. The depth of the milled cavity depends on the thickness of the mold compound and the location and loop height of the bond wires. During the milling operation use a vacuum line to pick up the loose plastic particles generated.

Caution: do not mill into the bond wires or the chip. Mill counter bores on devices with chip dimensions greater than 0.400 inches on a side. These counter bores should be made on one or more levels within the bond pad perimeter and at the outermost corners of the cavity (this is necessary to facilitate etching of the mold compound at the corners of the chip before the sides are exposed and subsequent damage to the leadframe). Care must be taken during the milling operation to avoid excessive pressure on the mill resulting in filler induced damage to the chip P.O. The end mill should not bind, bend, or "smoke" during the milling operation.



3. PACKAGE ETCHING
All etching must be performed in a chemical hood that meets the requirements defined in references #.3 and 4. Heating of acid or device prior to application of acid must be done using an explosion proof hot plate that meets the requirements of reference #4. Obtain the appropriate acid for use on the mold compound being removed. Following are the acids that have been identified for the removal of the various mold compounds. Mold Compound Acid/Temperature Shinetsu Red fuming nitric acid at 140-150 degrees Celsius Plascon & Sumitomo Red fuming or yellow nitric acid 140-150 degrees Celsius


Note: Fuming sulfuric acid reacts with exposed aluminum bond pad metallization and may result in ball bond discontinuity thus hampering further analysis.




When using red fuming nitric acid it may be helpful to start the etching process using a mixture of red and yellow nitric acids in order to slow down the etch process until a "residue crust" is formed over the cavity and then switch to the red nitric acid.


Apply the acid in drops using a plastic micro-pipette. The drops should be placed in the center and at the corners of the cavity in approximately a 1:1 ratio.


Allow the acid to react with the mold compound and form a crust of dissolved compound. Caution: Do not allow the crust to dry out completely before adding additional drops of acid.


Remove the dissolved material using cotton swabs or by rinsing with acetone when the dissolved materials threaten to spill over the cavity. Caution: Rinse the device immediately with acetone if acid spills onto the package pins.


Soak the device in acetone for a minimum of 10 minutes, followed by a spray of methanol to remove loose residue and to clean the residue from the cavity rim.


Perform a thorough microscopic inspection to determine whether all necessary areas of the chip are exposed.


If dried mold compound residue persists on the chip surface, use the following in the order shown to attempt removal:


Solvent bath (such as methyl alcohol) in ultrasonic cleaner


Several drops of room temperature fuming sulfuric acid applied to the chip (with the chip at room temperature) for several seconds then rinse the device in DI water.


Several drops of fuming sulfuric acid applied to the chip with the chip on a 100 degree Celsius hot plate. Note: Fuming sulfuric acid will attack aluminum bond pads and is therefore the method of last resort.

[Grant Stockly]

<Part of the previous post>

Can't comment on sulfuric, but I have used red fuming nitric at near boiling temperature. Apply acid, let react. Flush witn more acid, let react, etc.
Woody White

[Grant Stockly]

<Part of the previous post>
Diane,

Yes, hot fuming sulfuric and/or hot fuming nitric are used as standard procedures for removing plastic from IC's. The process is not quite that simple. For example, water rinses will almost certainly etch the bond pads on the IC and thus removing connection to the outside world. Additionally, the plastic contains fire retardants which some regions don't like being washed down the drain. There is more detailed help through EDFAS.org (one of ASM's branches). B&G International sells a very safe, effective etcher which performs decapsulation automatically in minutes.

I have no association with B&G International.

David Saxon
Analytical Microscope Services
11826 Reservoir Rd. E.
Puyallup, WA 98374
253-848-7701 voice & fax
email: info-at-analyticalmicroscope.com
website: www.analyticalmicroscope.com

[Grant Stockly]

<Part of the previous post>

Diane,

I used to do failure analysis on semi conductor memories which were starting to be made of plastic/epoxy with glass rods about 20 years ago.

I have some technics and possible help but its too much to write. Basically you drill a small hole about 0.1" deep then heat the IC on a hot plate and then you drop your acid to remove the plastic. I dont know chemistry, I'm and Electronics Technician. I did this work with a meterial sicentist, my mentor.

We used fumming sulfuric acid and fuimg nitric acid, also some type of organic pink and blue solutions to stop some of the acids etching effect.

The company back then was Burroughs Corp. today is Unysis.

I presently work for the U S Department of Energy in New York City. My phone number is 212 6203650, I'll be happy to walk through some ideas and things I learned.

Regards
Peter Roiz

[Grant Stockly]

<Last part of the previous post>

Perhaps it's time to comment on this thread.

Dichloromethane and dimethylformamide are relatively effective disrupters of most epoxies but their action is accompanied by great swelling because the polymer becomes engorged with the liquid before any significant solvation takes place. This will destroy wire bonds on an IC.

Fuming (essentially anhydrous) sulfuric acid acts by the completely different process of sulfonating reactive groups that remain on the polymer. The depolymerized and sulfonated byproducts are quite soluble not only in the acid but usually in water as well. The worst thing that you could do in this relatively straightforward process is to wash with water at intervals because this would initiate almost instantaneous corrosion. It would be advisable for a chemist, as someone trained in the handlingof reactive materials, to carry this out or at least to establish procedures and train others with less experience. The action of sulfuric acid in this regard is quite different than that of nitric. Nearly anhydrous nitric acid (completely anhydrous is extremely difficult to prepare) is a very powerful oxidizer and could lead to unstable, dangerous byproducts whereas the sulfonates resulting from the sulfuric acid reaction are relatively stable. Water must, of course, be prevented from splashing into any concentrated acid, especially sulfuric.

A very strong acid such as sulfuric behaves completely differently in the absence of water. Since most acids are highly hygroscopic and are sold as water solutions, most people do not observe this other side of their behavior. Without water to create an ionized electrolyte, corrosion of metals will not take place. I have de-encapsulated ICs for failure analysis in 200 degree sulfuric acid and been able to operate the IC without replacing the .001" aluminum wirebonds that it came with. I recall one instance where our company built prototype hybrid microelectronic circuits out of such de-encapsulated ICs when their supplier was late getting a new design on the market and the only ones available were already encapsulated.

The key is to realize that water must be excluded until the sulfonating acid has been completely rinsed away by a non-aqueous liquid. As Mr. Saxon said, there are simple and safe devices available for doing this operation. However, with proper care and protective gear it can be done in a beaker on a hot plate in a fume hood. A few ml.s of sulfuric acid are heated to drive off water until heavy vapors are observed over the liquid (which may darken during heating due to trace impurities). The IC is carefully lowered into the hot acid and a vigorous reaction ensues with the epoxy almost instantly washing into the solution. After a few seconds the IC is then quickly lifted out and held over a receiving vessel and flooded with a stream of ethanol. Only after this is a final rinse in deionized water carried out, followed by fresh electronic grade ethanol and forced drying in warm air.

The ready made devices which carry out the operation are typically a small bowl with a hinged lid from which air is withdrawn by a gentle vacuum. An inert metal feeder tube leads from a heated reservoir for the sulfuric acid and passes through the wall of the bowl to a position where the encapsulated device is secured. When the lid is closed and the slight vacuum applied, the hot acid is pulled into the bowl over the device. It is somewhat self-limiting in that, if the lid is opened, there is no driving force to bring more acid into the container. Naturally, the vacuum source needs to be protected by a trap and all waste products properly handled no matter how the procedure is carried out.

John Twilley
Conservation Scientist
(formerly, Manager of the Reliability Analysis Center, Teledyne Microelectronics)

From daemon Fri Jan 25 08:35:53 2002

[Grant Stockly]

From: Darrell Miles : milesd-at-US.ibm.com
Date: Fri, 25 Jan 2002 11:45:46 -0500
Subject: IC package removal

Contents Retrieved from Microscopy Listserver Archives
http://www.microscopy.com/MicroscopyListserver/MicroscopyArchives.html

Body of Message

Jon,

If it is the black "plastic" (glass filled epoxy), nitric acid works on some, and sulfuric acid works on most. There are also some chemicals called "Dynasolve", that we use various types of. The differences are in the speed with which the material is removed, and what damage is done to interconnects, etc. If you watch the time you have the sample in sulfuric acid, it works for most parts.

Hope this helped.
Regards,
Darrell

[Grant Stockly]

I have sent one PAL (Apple calls them HALs on the Mac boards) to MEFAS. You can find them at http://www.mefas.com

They offer a decapsulation service at only $45 per part with a 2-3 day turnaround. I have sent them the ASG PAL since its the only one I don't have equations for. I have requested the service of decapsulation with cleaning options for photography. This means they don't pay attention to the bond wires and attempt to clean the surface.

Their website describes their decapsulation service as follows.

Decapsulation is a technique for exposing the internal components of a package device. There was a time when a hot plate and concentrated nitric acid were capable of decapping integrated devices (IC’s). However, advancements in device packaging introduced new plastic encapsulants and the array of solder bumped external leads package designs. Our jet etching systems are capable of exposing any package die without melting solder bumps, etching bond pads, or destroying external leads. The system sprays high temperature nitric and/or sulfuric acid through a fine diameter nozzle head. By computer control of the temperature, pressure, and time of the etching process a uniform etch cavity is created which exposes the IC.

Of course not all devices need the selectivity of our automatic decapsulator. For these devices we x-ray the device to measure the internal die, cut the proper die cavity in the package, and finally eye drop concentrated nitric acid in the cavity to remove the plastic from the IC.

Once decapped, analyses such as inspection, microprobe, FIB rework, delayering, or even retesting can be accomplished. The device will remain functional.



I liked how they described the entry level procedure... "eye drop concentrated nitric acid in the cavity to remove the plastic from the IC."

Most companies I've found don't go into detail about how they do their service, and wouldn't admit it could be done with a hotplate and eye dropper.

[erik@nisene.com]

Decapping by hand is not only extremely dangerous, but it is done successfully by very few people. Nisene Technology Group Manufactures the equipment you would need to do this. Email me at erik@nisene.com and I will set you up with a quote and resources on getting out equipment.

[Grant Stockly]

The hard part is getting acid good enough. After an hour of work with some over-the-counter nitric acid I had a 68000 done, but the water content was so high that it left a residue on the chip. Its still fun to look at. MEFAS at $35-40 is worth it for sure! :)

[elrob1981]

Guys,

I just joined this forum and would really appreicate your help here.

What are the concentrations of the sulfuric and nitric acids you use for
wet chemical decapsulation. I'm in the process of purchasing chemicals for our chemical lab and need to know the concentration of acid in the solution that will yield the best results. I was thinking 90% Nitric Acid and 96% sulfuric acid would be ok to use but need you to confirm.

thank you

Roberto

[erik@nisene.com]

Guys,

I just joined this forum and would really appreicate your help here.

What are the concentrations of the sulfuric and nitric acids you use for
wet chemical decapsulation. I'm in the process of purchasing chemicals for our chemical lab and need to know the concentration of acid in the solution that will yield the best results. I was thinking 90% Nitric Acid and 96% sulfuric acid would be ok to use but need you to confirm.

thank you

Roberto


There are several variations of sulfuric and nitric acid on the market. Our equipment uses 20% fuming sulfuric and 98% fuming nitric. I could give you a few pointers if you would like to discuss further. erik@nisene.com or 831-761-7980

[Joe_Montana]

I work in a company that specialize in decap and test discret parts and IC's, to detect counterfeit components, we are thinking in buy a X-Ray machine, does anyone in this forum knows what kind i need to buy for this particular application?