Hi Robert,

I was able to find that data I mentioned on etch rates vs. dissolved
silicon.  What I did was dissolve crushed wafer fragments in a stirred, hot
solution of 25wt% TMAH (from Transene) to "season" the etchant and increase
the dissolved silicon concentration, then measure the [100] etch rate on a
test wafer at a range of dissolved silicon concentrations.  The solution was
initially fresh, and temperatures ranged from 83-86°C.

I found no obvious dependence of Si [100] etch rate (measured via stylus
profilometry on etch coupons) or thermal oxide etch rate (measured via
spectroscopic reflectometry) on dissolved silicon concentration, up to 20
grams dissolved silicon per liter of 25wt% solution (on the order of 2-3
dissolved 4" wafers per liter of etchant).

As I mentioned, temperatures ranged from 83-86°C (this was before we had an
accurate temperature control system for the etch bath).  I recorded Si [100]
etch rates in the range of 30-38 µm/hr and oxide etch rates between 5-9
nm/hr.  There was some instability in the temperature and I attributed the
variation in etch rates to the temperature instability; however, when I
looked at the Si [100] etch rates vs. dissolved silicon concentration and
temperature, it appears as though the etch rate may have been increasing
slightly with increasing dissolved silicon, but don't quote me on that
(there was a not inconsiderable amount of scatter in the data).

Bottom line: I didn't find any strong dependence of Si [100] or thermal
oxide etch rates on dissolved silicon concentration, but you haven't said
how old your etchant is or how much silicon has been dissolved in it.  I
wouldn't be surprised if the etch rate of a lower-TMAH-concentration
solution was more sensitive to dissolved Si than a higher-TMAH-concentration
solution.  In any case, I recommend that you monitor the etch rates over the
life of your etchant bath and replace the bath when appropriate.

I hope this helps; feel free to contact me to discuss this further.

Best regards,

Brian C. Stahl
Graduate Student Researcher
UCSB Materials Research Laboratory
brian.stahl@gmail.com / bstahl@mrl.ucsb.edu
Cell: (805) 748-5839
Office: ò_Ó MRL 3117A


On Mon, Jun 21, 2010 at 9:38 PM, Brian Stahl  wrote:

> Hi Robert,
>
> I used a similar setup to anisotropically wet etch Si.  I also found that
> clear crystals formed upon cooling the solution when the etchant had
> consumed a large quantity of silicon (on the order of grams per liter of
> etchant solution); these crystals would dissolve when the etchant was heated
> again.  Although I have never characterized them directly, I suspect they
> are some kind of water-soluble silicate compound.  It is well established
> that etching of silicon in the presence of hydroxyl-containing solutions
> proceeds according to:
>
> Si + 2(OH-) + 2H2O → SiO2((OH)2)^2- + 2H2
>
> I imagine that the water-soluble dihydroxyl silica product could form a
> complex with the positively charged tetramethylammonium ion (two (TMA)^1+
> ions to one SiO2((OH)2)^2- ion) and crystallizes out when the solution
> becomes supersaturated (either through cooling or an increase in
> concentration).
>
> I believe I have data on the silicon etch rate, and possibly oxide
> selectivity, as a function of dissolved silicon content in a notebook at
> work.  I'll check tomorrow when I get in.
>
>  Best regards,
>
> Brian Stahl
>
> --
> Brian C. Stahl
> Graduate Student Researcher
> UCSB Materials Research Laboratory
> brian.stahl@gmail.com / bstahl@mrl.ucsb.edu
> Cell: (805) 748-5839
> Office: ò_Ó MRL 3117A
>
>
>
> On Mon, Jun 21, 2010 at 3:41 AM, Robert MacDonald <
> robm@shearwaterscientific.com> wrote:
>
>> I have a TMAH bath I use for Si Etching. I run it at 18 percent by weight
>> concentration, and a temperature of 90 deg C. It has a cooled lid to prevent
>> evaporation, and I use a hydrometer to keep the density constant. I find
>> that over time crystals start to form in the bath.
>>
>> Does anyone know what these crystals are?
>>
>> Does anyone know if this effects the etch rate, or selectivity to silicon
>> dioxide?
>>
>> Thanks,
>> Rob MacDonald
>> Shearwater Scientific