🚀 go-pugleaf

In article <1992Sep28.071622.3040@trl.oz.au> jdell@trl.OZ.AU (John Dell - Photonics) writes:

>Anyway, the problem (at last) is that we need to measure lowish temperatures
>now (of the order of 300-500 C) and we would like some recommendations of how
>to do this using optical pyrometry.  What have been peoples' experiences, how
>do you calibrate the system, what products are available, how accurate
>are they, how precise are they, do you look at the block or the semiconductor,
>have you had to worry about the window (ie do you need a sapphire window
>or will a standard window do and simply change the effective emissivity), 
>do you need to heat the window to prevent condensation of As, how do you
>limit the field of view .....? 

Last spring I worked with an Optical Pyrometer attempting to measure
(accurately) the maximum temperatures on Si using an Nd:YAG (1.06um)
beam, with very little success.  Although we were looking for
temperatures near 1600K, we still learned quite a bit about optical
pyrometry.

We found the biggest consideration to be how much area on the sample
that you need to see.  We needed a resolution such that the window that
the pyrometer saw was about 300um wide - which is very difficult to
attain.  The problems in viewing an area this small come from the fact
that the lens must be _very_ close to the sample.  Since we were trying
to look at the surface temperature, our situation was  a little
different than yours because we needed to look at the sample head on.

Our pyrometer was made by Inframetrics (I don't remember the model
number offhand) but we found it to be very accurate in other (larger
scale) situations.  Simply looking into a tube furnace revealed only a
2-3 degree difference between the pyrometer and the furnace controller;
Calibration was performed by Inframetrics.

Please send me email and I can dig up some old manuals and notes to give
you more information.


Ian D. Bjorhovde
ianbjor@gas.uug.arizona.edu  -or-  ianbjor@convx1.ccit.arizona.edu
Department of Materials Science & Engineering
University of Arizona, Tucson AZ

In article <1992Sep28.071622.3040@trl.oz.au> jdell@trl.OZ.AU (John Dell - Photonics) writes:
>
>Anyway, the problem (at last) is that we need to measure lowish temperatures
>now (of the order of 300-500 C) and we would like some recommendations of how
>to do this using optical pyrometry.  What have been peoples' experiences, how
>do you calibrate the system, what products are available, how accurate
>are they, how precise are they, do you look at the block or the semiconductor,
>have you had to worry about the window (ie do you need a sapphire window
>or will a standard window do and simply change the effective emissivity), 
>do you need to heat the window to prevent condensation of As, how do you
>limit the field of view .....? 
>
	I'll tell you what I did as a grad student at the U. of
Illinois.  The moly blocks can be cleaned with conc. HCl.  The wafers
themselves should first be degreased thoroughly using several rinses of
TCA (trichloroethane) followed by acetone and methanol and then DI
water.  Then they should be etched using a warm solution of
H2SO4:H202:H20 (I think).  This can be found in the literature in
several articles by Al Cho (Bell Labs) or several other people.  I can
find the references if you want to though I doubt that, working in MBE,
you would be unaware of them.  
After etching the wafers, they are
mounted as usual with In solder on the moly  blocks.  Be *extremely*
careful and patient with this.  Use the edge of a pair of tweezers to
slide the wafer around on the block to ensure that the backside of the
wafer is uniformly coated with the In.  One of the biggest causes of
temperature variations is that the In does not coat the backsides of the
wafer uniformly.
	Finally, get yourself a RHEED setup.  Once you get a properly
As-stabilized surface, you are ready to go.  Typically, we heated the
wafer to about 590 degrees C or so to get rid of the oxides, waited to
get the As-stabilized surface RHEED pattern (with the As oven shutter
open, of course), opened the Ga shutter and proceeded.
	We never had a problem with the pyrometer, though it got
increasingly off target (never more than 10 or 20 degrees) as the
window started getting coated with As and Sb.  Eventually, we got to the
point where we monitored oxide desorption and growth initiation almost
solely by RHEED and chamber pressure and used temperature to ensure that
we were roughly at 580-590 which if I remember right is optimum of
growing GaAs. 
	If you are having problems with temp variations, I would look at
your In-wafer interface first, and think of getting RHEED next.
	Boy, this stuff brings back a lot of memories.  It is not often
that you see compound semiconductor types posting to the net!
 

--
                                  | "I am not a satanist!  I find it hard
Sam Sivakumar                     | enough to conjure myself out of bed
ssivakum@hopi.intel.com           | every morning, leave alone conjure up
                		  | the devil!"  --  Ozzy Osbourne

    Hellow, I am a student in Japan.   And I use ANELVA S830 MBE
system.

>   Some one must have done this - low substrate temperature measurment for MBE.
>
>   We have a much modified (and now very old) Varian 360 MBE which we use for
>   growth of InGaAs/GaAs layers (MQWs for modulators and 980 nm lasers).
>   Currently we measure the substrate temperature in two ways -

I think you grow InGaAs/GaAs at Tsub=450 degree C or so.

>   1) We have a thermocouple attached to a sapphire bead which is supposed
>   to be in contact with the back of the Mo block (we still indium solder
>   our wafers but put silicon nitride or dioxide on the back which overcomes
>   a lot of the problems).  In practice, the sapphire bead doesn't touch the
>   back and we rely on radiative coupling to the bead (which isn't too bad at
>   high temps).

Thermocouples at substrate heater do not touch enough also in our
system.    I do not believe the temperature of the thermocouple.
It changes from +100 to -100 degree C while a substrate is rotated.

>   The system is calibrated using the Al/Si eutectic temperature.  We have
>   looked at using the oxide desorption temperature but seem to have a problem
>   with our substrate cleaning so that the desorption temperature appears
>   to vary by about 50 C from run to run.  (Anyone willing to tell me 
>   EXACTLY what their favourite cleaning recipe is?? - I've seen lots of
>   very brief descriptions but no real details.  We currently use a sulphuric
>   acid / hydrogen peroxide etch.)

I think the reason of the errors of your pyrometer cannot be put
the blame on your pyrometer.     The temperature showed by
pyrometer dependes on solid angle to the substrates so that the angle
of the substarates are not constant in each run.

>   Anyway, the problem (at last) is that we need to measure lowish temperatures
>   now (of the order of 300-500 C) and we would like some recommendations of how

In our laboratory, Tsub is calibrated by Al/Si for high
temperature (for GaAs and AlAs growth) and Au/Si for low
temperature (for InAs growth).  And I get the relation between the
Tsub and the heater current.  The lineality of the heater current
and Tsub is also checked with pyrometer in the region between 
the eutectic temepratures of Al/Si and Au/Si .

In practice, windows get more and more dirty with As(GaAs?) in
every runs.    I think pyrometer cannot bring true results in any
time.    What I believe is the heater current and RHEED pattern.

If you want to get more detail about our fasion, please send
e-mail to me.   ( But I cannot say 'leave it to me!'. )

--
===================================================
$@5~ETBg3X9)3XItEE5$9)3X65<<(J  $@:4!9LZ8&(J$@5f<<(J  $@1'Ln(J $@OB9T(J
                    e-mail : uno@kuee.kyoto-u.ac.jp