strenghth of doped si wafer

[niteng]

Recently, I'm intrested in the strength of doped Si wafer. Because I have found that light doped Si wafer is more likely going to broke in experiment.
I dont know the exact reason, but i want to find it out. Does anybody has some ideas?
I want to download an article which discuss this problem named "Micromechanical and tribological characterization of doped single-crystal silicon and polysilicon films for microelectromechanical systems devices", could anyone access this article? Hope you can email me one copy. Thank you so much.

 

[dick_freebird]

I'd be surprised if typical starting-material dopings have
any real effect; even near-degenerate doping is sub-%
dopant to silicon ratio. Typical starting materials I use
are maybe 1ppm-ish dopant (E17/cm3 Nd vs E23/cm3 Si).

How many wafers is your sample set, to draw such a
conclusion? I'd think to be confident that it's real, you
would have broken at least tens of them and none of
the doped ones, under some controlled conditions.

 

[niteng]

I'd be surprised if typical starting-material dopings have
any real effect; even near-degenerate doping is sub-%
dopant to silicon ratio. Typical starting materials I use
are maybe 1ppm-ish dopant (E17/cm3 Nd vs E23/cm3 Si).

How many wafers is your sample set, to draw such a
conclusion? I'd think to be confident that it's real, you
would have broken at least tens of them and none of
the doped ones, under some controlled conditions.

Thank you for your reply at first!
1. I have two kinds of Boron-doped p-type 4-inch silicon wafer:
A:high-resistivity 1500ohm·cm(light doped);
B: low-resistivity 10ohm·cm(heavily doped). I found high-resistivity wafer is more likely to be broken.
You said "even near-degenerate doping is sub-%dopant to silicon ratio."
I don't really catch it. Is that the condition of my wafer B?
2. I'm sorry to say I only broke two wafers.
My tutor has told me he once also had encountered such situation, and he suggests me to find out whether there is a relationship between the resistivity of the silicon wafer and the mechanical strength of that.

 

[dick_freebird]

I consider neither of these "heavily doped" - 3-5 ohm-cm
is what we used to use for "40V" bipolar technology and
good to over 300V blocking.

Degenerate is something like 1E20/cm3, very low resistivity.

I would suspect something like a thermal strain, perhaps
an uneven cooling event (we used to see wafer breakage
at RTA steps if the profile was too abrupt) more than the
chemical content. If the wafers are from different sources
or have been subjected to different thermal profiles then
you have to consider more than chemistry.

 

[niteng]

The first time I broke one wafer is when I washed it
using deionized water, which may be explained
by too much mechanical stress.

The second situation drove me to consider why the high resistivity-wafer
broke again. 7 wafers were placed in a box. And I droped it on the floor uncarefully.
The 4 10ohm·cm wafers were undamaged，but again one 1500ohm·cm wafer was broken to two pieces.

I guess there is a discrepancy of lattice structure between the two kinds of wafer. And they can stand different mechanic stress.
Am i reasonable? I study microelectronics, but am not familiar with mechanics.

 

[dick_freebird]

Metallurgically, sometimes small chemistry variations
do impact hardness, toughness, etc. but this is also
coupled to thermomechanical treatment and constraints
(hardening, quenching of steels). But these are really
heterogeneous metal composites unlike what should be
a single crystal structure with very low impurities.

I'd bet, in fact, that the 1500 ohm-cm wafer is intrinsic
and the 1500 ohm-cm spec is very wide, perhaps a one
sided limit with no maximum? Could be that undoped, vs
the implant-and-activation thermal processing and/or
epi growth applies a very different thermal stress/strain.
Or, who knows, in-situ doping for a bulk doped wafer
could modify the crystalography somehow. It's just that
I don't see sub-ppm doping levels having that much
mechanical influence.

Then again, thermal conductivity can be in large part
electron transport, maybe mechanical shock resembles
that (both being vibration of a sort) and a conductive
wafer spreads impact energy better than a semi-insulating
one.

 

[niteng]

Metallurgically, sometimes small chemistry variations
do impact hardness, toughness, etc. but this is also
coupled to thermomechanical treatment and constraints
(hardening, quenching of steels). But these are really
heterogeneous metal composites unlike what should be
a single crystal structure with very low impurities.

I'd bet, in fact, that the 1500 ohm-cm wafer is intrinsic
and the 1500 ohm-cm spec is very wide, perhaps a one
sided limit with no maximum? Could be that undoped, vs
the implant-and-activation thermal processing and/or
epi growth applies a very different thermal stress/strain.
Or, who knows, in-situ doping for a bulk doped wafer
could modify the crystalography somehow. It's just that
I don't see sub-ppm doping levels having that much
mechanical influence.

Then again, thermal conductivity can be in large part
electron transport, maybe mechanical shock resembles
that (both being vibration of a sort) and a conductive
wafer spreads impact energy better than a semi-insulating
one.

1. I agree with your saying-- “Could be that undoped, vs
the implant-and-activation thermal processing and/or
epi growth applies a very different thermal stress/strain.
Or, who knows”. It's indeed that we can hardly estimate
the influence of fabrication parameters eg. doping temperature,
and time on the thermal stress.
2. The wafer supplier has given a spec -- ρ>1500 ohm-cm.
I measure ρ≈1.6×E4 ohm-cm using 4-point probe, which is equivalent to
E12 cm-³. Why did you say it's intrinsic?
And I don't know the wafer is in-situ doping or not.
Does the doping mode(in-situ or not) have large impact on
the crystallography?
3. You said " a conductive wafer spreads impact energy
better than a semi-insulating one." What is your basis？
Do you mean the doped wafer can bear more strong mechanical shock,
if there really lies relationship between doping level and mechanical influence?
4. I noticed that you have mentioned thermal conductivity several times. And I don't see
how that concept is involved in this matter!


I know my sample is so meager. Maybe it is because some accidental factors. But I just
have been curious about that why I crumble the same kind of wafer two times.
I'm going to look for the relationship between crystallography and mechanical properties
via papers and books. Hoping I can find a explanation!
Thank you for your kind help!