Request For Comment

[amal banerjee]

Dear All,

I am planning on writing and publishing a lecture notes style reference book
on ALL homogeneous and heterogeneous junction semiconductor devices.
The list includes BJTs, FETs, MOSFETs, HBTs, HEMTs, LEDs, Solar cells,
VCSELs etc. The main reason for writing this book are :
1, Very few(miniscule) of the available books have any material on
heterogeneous junction devices.
2. Available books fall short on the key details - e.g., the sequence of steps
involved in creating the crucial energy band diagrams when e,g., two isolated
oppositely doped semiconductors are brought in contact.
3. Some semiconductor device books are written by physicists who overlook
large signal models and equivalent circuit models. Other books written by
pure electronics people ignore the key quamtum or statistical mechanics that govern semiconductor device operation. No mention of the large signal Angelovc models for HDTs, HEMTS.

A tentative list of topics is as follow - it will be expanded.

A Fundamental Quantum and Statistical Mechanics of Crystalline Solids
-Energy Bands
-Physics of Energy Bands
-Material Classification Using Quantum States
-Crystal Structure and Semiconductor Energy Bands
-Crystal Momentum, Effective Mass, Negative Effective Mass
-Effective Mass Schrodinger’s Wave Equation
-Electron Excitation(Chemical, Electrical, Optical, Thermal) From Valence to Conduction Band – Hole Creation
-Recombination(Band-Band, Auger) and Recombination Lifetime
-Carrier Concentrations
-Thermal Equilibrium and Fermi Dirac Statistics
-Collisions and Scattering
-Fermi Dirac Statistics and Fermi Level, Equilibrium Carrier Concentration

B. Charge Transport(Current Flow) in Semiconductors
-General Concepts
-Relation between Charge, Current and Energy
-Boltzmann Transport Equation
-Method of Moments Solution of Boltzmann's Equation
-Drift-Diffusion Equations
-Hydrodynamic Transport Equations
-Semiconductor Device Design Equations
-Conductivity Electron, Hole Effective Masses
-Drift=Diffusion and Thermal Currents
-Ballistic Transport of Charge Carriers

C. Homo|Heterogeneous Semiconductor Junction(np|Np) Fundamentals
-Homogeneous np Semiconductor Junction
-Homogeneous np Junction with External Bias
-Quasi Fermi Levels
-Heterogeneous Semiconductor Junctions Np and Pn
-Heterogeneous Semiconductor Junction Quasi Fermi Level Splitting

D. Basic Heterogeneous Bipolar|Bijunction Transistor(HBT)
-Types of Heterogeneous Bipolar Transistor and Characteristics
-The Collector and Base Current of a HBT
-Emitter Hole Current in a HBT
-Simple DC Equivalent Circuit Model for Heterogeneous Transistor(HBT)

E. Advanced VBIC, Non-VBIC Heterogeneous Bipolar|
Bijunction Transistor(HBT) - Large Signal Models
- Parameters - Equivalent Electrical Circuits
-The VBIC Model Specification
-VBIC Homogeneous Junction Bipolar Transistor
-VBIC Heterogeneous Junction Bipolar Transistor
-Non-VBIC Heterogeneous Junction Bipolar Transistor

F. Basic High Electron Mobility Transistor(HEMT)
-Metal-Semiconductor Junction, Schottky Diode, Ohmic and Rectifying Junction
-Metal-Semiconductor Field Effect Transistor - Drain Curremt
-High Electron Mobility Transistor(HEMT) - Potential Wells, Undoped Layer Two Dimensional Electron Gas(2DEG) and Properties

All hints, suggestions recommendations are welcome. Thanks in advance.

[Phil Hobbs]

Sounds like quite the tome!

You don't seem to be planning to cover normal devices much--most
current-technology s/c devices are somewhere in the gap between sections
C and D. I'm personally a big user of pHEMTs and SiGe HBTs, but there
are lots of other things out there, such as CMOS and flash EEPROM.
That's not intended as a criticism, but it may restrict your market some.

Things I'd like to read about:

Device design, including process limitations, e.g. planar vs. epitaxial
vs. HBT bipolars and FETs (A lot of low-noise design is limited by the
Johnson noise of various extrinsic resistances, for instance.)

Tradeoffs between breakdown voltage, speed, Early voltage, and so on,
and how HBTs and graded-composition designs can help. Normal RF
transistors have terrible Early voltages, but graded-base SiGe ones can
be amazing.

Large-signal behaviour of HBTs, e.g. what happens when you forward-bias
the gate of a pHEMT. (JL and I have a few tricks to share in that area.)

In general I'd suggest thinking fairly carefully about who your audience
is. I've been out of academia for 35 years, myself, but BITD folks who
were interested in a deep dive into heterojunctions, 2-d electron gas
devices, and that sort of thing probably knew some quantum mechanics and
semiconductor theory already.

Good luck with it--it sounds like a noble endeavour.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com

[jla...@highlandsniptechnology.com]

As a circuit designer, I don't know or care much about semiconductor
physics except where it provides hints to possible undocumented device
behavior, which I can measure and see if it's useful.

So, such a book should include practical stuff, like the
voltage-current-capacitance curves of a part that has typical physics.
In other words, fill a few pages with equations, then specify a
specific case, and describe the device in datasheet terms.

I measure phemts and GaN and SiC for undocumented behavior, like use
as a diode, or step recovery, or Rds-on, and discover useful stuff. If
a book has practical stuff, working engineers might buy it.

Device degradation and damage are important topics but actively
ignored by the part makers.

I have (somebody's) book about microwave PCBs, microstrip and
stripline and such. Some of the cases are pages of equations and pages
of equations inside of equations. Totally useless.

[John May]

Possibly add a chapter on the design tools actually used in device design (TCAD, Sentaurus etc.). Or maybe utilise some of the resources at nanoHUB.

[whit3rd]

On Wednesday, August 3, 2022 at 4:18:40 AM UTC-7, daku...@gmail.com wrote:
> Dear All,
>
> I am planning on writing and publishing a lecture notes style reference book
> on ALL homogeneous and heterogeneous junction semiconductor devices.
> The list includes BJTs, FETs, MOSFETs, HBTs, HEMTs, LEDs, Solar cells,
> VCSELs etc. The main reason for writing this book are :
> 1, Very few(miniscule) of the available books have any material on
> heterogeneous junction devices.
> 2. Available books fall short on the key details - e.g., the sequence of steps
> involved in creating the crucial energy band diagrams when e,g., two isolated
> oppositely doped semiconductors are brought in contact.
> 3. Some semiconductor device books are written by physicists who overlook
> large signal models and equivalent circuit models. Other books written by
> pure electronics people ignore the key quamtum or statistical mechanics that govern semiconductor device operation. No mention of the large signal Angelovc models for HDTs, HEMTS.

This sounds too broad for a single book, alas.
As for heterojunction and other devices not covered in the old Sze book's various editions

<https://www.goodreads.com/book/show/382279.Semiconductor_Devices>

It'd be useful, all right, but device-design and device-utilization issues are very
different; a broad audience wants to know utilization, i.e. a set of model
parameters for real devices, and that's info that changes often.

The physics of heterojunctions is interesting, but to sell copies... maybe just a tiny
book with those topics can be marketable, preferably with some SPICE or other
support. A lecture note base has to prepare naiive students for the real
world; that's too much info to casually read through, but a discussion of
models and parameters for the novel devices is... a right-size information dose.

[Joe Gwinn]

On Wed, 3 Aug 2022 04:18:37 -0700 (PDT), amal banerjee
<daku...@gmail.com> wrote:

>Dear All,
>
>I am planning on writing and publishing a lecture notes style reference book
>on ALL homogeneous and heterogeneous junction semiconductor devices.
>The list includes BJTs, FETs, MOSFETs, HBTs, HEMTs, LEDs, Solar cells,
>VCSELs etc. The main reason for writing this book are :
>1, Very few(miniscule) of the available books have any material on
>heterogeneous junction devices.
>2. Available books fall short on the key details - e.g., the sequence of steps
>involved in creating the crucial energy band diagrams when e,g., two isolated
>oppositely doped semiconductors are brought in contact.
>3. Some semiconductor device books are written by physicists who overlook
>large signal models and equivalent circuit models. Other books written by
>pure electronics people ignore the key quamtum or statistical mechanics that govern semiconductor device operation. No mention of the large signal Angelovc models for HDTs, HEMTS.
>
>A tentative list of topics is as follow - it will be expanded.
>

[snip]

>
>C. Homo|Heterogeneous Semiconductor Junction(np|Np) Fundamentals
>-Homogeneous np Semiconductor Junction
>-Homogeneous np Junction with External Bias
>-Quasi Fermi Levels
>-Heterogeneous Semiconductor Junctions Np and Pn
>-Heterogeneous Semiconductor Junction Quasi Fermi Level Splitting
>
>D. Basic Heterogeneous Bipolar|Bijunction Transistor(HBT)
>-Types of Heterogeneous Bipolar Transistor and Characteristics
>-The Collector and Base Current of a HBT
>-Emitter Hole Current in a HBT
>-Simple DC Equivalent Circuit Model for Heterogeneous Transistor(HBT)
>
>E. Advanced VBIC, Non-VBIC Heterogeneous Bipolar|
> Bijunction Transistor(HBT) - Large Signal Models
>- Parameters - Equivalent Electrical Circuits
>-The VBIC Model Specification
>-VBIC Homogeneous Junction Bipolar Transistor
>-VBIC Heterogeneous Junction Bipolar Transistor
>-Non-VBIC Heterogeneous Junction Bipolar Transistor
>
>F. Basic High Electron Mobility Transistor(HEMT)
>-Metal-Semiconductor Junction, Schottky Diode, Ohmic and Rectifying Junction
>-Metal-Semiconductor Field Effect Transistor - Drain Curremt
>-High Electron Mobility Transistor(HEMT) - Potential Wells, Undoped Layer Two Dimensional Electron Gas(2DEG) and Properties
>
>All hints, suggestions recommendations are welcome. Thanks in advance.

What is often missing is a discussion of what drives noise generation
in the various transistor types. This kind of data is essential in
many practical low-noise applications, and it's useful to know how the
various transistor types compare, and how the noise scales with device
details.

There are two major kinds of noise to addressed, being flat-band (the
noise floor), and flicker noise at low offset frequencies. These
differ in their physical causes.

Both current and voltage noise must be considered. Spurs are ignored
here.

The noise floor is typically Gaussian, and often arises from thermal
and shot-noise sources.

The sources of flicker noise are many and mostly unknown, but cleaner
material and bigger device volume both reduce the flicker level.
Flicker noise follows a sum of kn/f^n terms formula, where n is an
integer from zero to three or four.

One mistake one often sees is that the offset-frequency location where
the Flicker noise curve crosses the noise floor is cited as defining
the flicker noise. The problem is that variation in noise floor level
will change this intersection, even though the flicker noise is
unchanged, so it's best to tabulate the coefficients kn of the kn/f^n
terms.

Joe Gwinn

[amal banerjee]

Thanks for the detailed set of suggestions. I will definitely have short chapters on
bipolar transistors and MOSFETs. The reason I have placed the heterojuncrion devices
before the homogeneous devices is that a miniscule number of authors include these
in there books. Most of the material on heterogeneous junctions is in conference|journal
papers etc., which is good since the reader has access to the latest information.
There will definitely be some chapters on how to design transistors for various applicarions
-e.g., power FET as compared to high speed(as in a microprocessor) or high frequency
(RF|microwave) application. Most definitely, there will be a detailed chapter on noise
sources in these high performance devices. Quantum, statistical mechanics and crystal structure is nor an issue - my basic training is in physics.

[amal banerjee]

Thanks for the feedback, will definitely keep your suggestions in mind while structuring the chapters.

[amal banerjee]

Absolutely, there will be a detailed chapter on noise sources in these high performance
devices and how they are being tackled, or proposed to being tackled. The physics is
confined to the first two chapters only. Without a clear idea of how drift|diffusion work
the reaser would be confused about how depletion regions form. Also, without any idea
of Fermi levels and thermal equilibroum, it is difficult to digest band bending at junctions.

[Mike Monett]

amal banerjee <daku...@gmail.com> wrote:

[...]

Quantum, statistical mechanics and crystal
> structure is nor an issue - my basic training is in physics.

You physics guys must have brain neurons 1/10 the size of normal people - you
have ten times as many:)




--
MRM

[Phil Hobbs]

Nah, we just gave up chasing girls a bit sooner. ;)

Cheers

Phil Hobbs

[Phil Hobbs]

whit3rd wrote:
> On Wednesday, August 3, 2022 at 4:18:40 AM UTC-7, daku...@gmail.com wrote:
>> Dear All,
>>
>> I am planning on writing and publishing a lecture notes style reference book
>> on ALL homogeneous and heterogeneous junction semiconductor devices.
>> The list includes BJTs, FETs, MOSFETs, HBTs, HEMTs, LEDs, Solar cells,
>> VCSELs etc. The main reason for writing this book are :
>> 1, Very few(miniscule) of the available books have any material on
>> heterogeneous junction devices.
>> 2. Available books fall short on the key details - e.g., the sequence of steps
>> involved in creating the crucial energy band diagrams when e,g., two isolated
>> oppositely doped semiconductors are brought in contact.
>> 3. Some semiconductor device books are written by physicists who overlook
>> large signal models and equivalent circuit models. Other books written by
>> pure electronics people ignore the key quamtum or statistical mechanics that govern semiconductor device operation. No mention of the large signal Angelovc models for HDTs, HEMTS.
>
>
> This sounds too broad for a single book, alas.
> As for heterojunction and other devices not covered in the old Sze book's various editions
>
> <https://www.goodreads.com/book/show/382279.Semiconductor_Devices>

<snip>

I really like Sze's second edition--the new co-author apparently tore
out all the BJT stuff, which is a shame.

[John Larkin]

On Thu, 04 Aug 2022 15:31:32 -0400, Joe Gwinn <joeg...@comcast.net>
wrote:

Most phemt type parts are specified for RF use. So one gets no clue,
from data sheets, about actual gate currents, Rds-on, breakdown
voltages, diode behavior, none of the good stuff.

You are very lucky to get any DC curves. "Adjust the gate bias until
it works."

More people would buy the book if it includes this kind of stuff. How
about sidebars about device details and quirks? Write a classic.



--

John Larkin Highland Technology, Inc trk

The cork popped merrily, and Lord Peter rose to his feet.
"Bunter", he said, "I give you a toast. The triumph of Instinct over Reason"

[Phil Hobbs]

A sidebar about stabilizing them in built-up circuits would be pretty
cool too. Specifically, the Murata BLM18- and BLM15BA050 and -100 (5
and 10 ohms, in 0603 and 0402 sizes) will turn a magic but unusably
unstable 14 GHz pHEMT or 40 GHz SiGe BJT into a magic and super stable
2-GHz.

pHEMT+bead gets you stuff like 0.3 nV 1-Hz noise in the flatband and a
transconductance within a factor of 2 of the thermodynamic limit.

SiGe BJT + bead gets you almost equally low noise plus a beta of 250 and
an Early voltage > 250V, so you can get a voltage gain of 50 or more in
one stage.

I'm just finishing up a bootstrapped preamp for a 3000-pF SiPM array
that uses a Mini-Circuits SAV-331+ depletion pHEMT. At 14 mA, its
output impedance is about 3 ohms, which is 10x lower than the closest
JFET. Its 1/f corner is 20 MHz or so, but it's still quieter than the
JFET (0.8 nV/sqrt(Hz)) in the 7 MHz circuit bandwidth.

[DecadentLinux...@decadence.org]

Phil Hobbs <pcdhSpamM...@electrooptical.net> wrote in
news:48d6402c-fcc7-917d...@electrooptical.net:

I think you have more on the ball than anyone else in this group,
Mr.(Dr.) Hobbs.

[John Larkin]

I like the enhancement parts like SAV541 for switching, like
discharging timing ramps, or as pulse generator output stages. (GaN is
better for slower, higher voltage apps.)

>
>Cheers
>
>Phil Hobbs

You and I could write little sidebar boxes for Amal's book.