In a MOSFET, the oxide plays the role of an insulator. Is oxide synonymous with to SiO2 or O2?
What is the intrinsic reason why the oxide is an insulator?
\$\begingroup\$
\$\endgroup\$
11
-
4\$\begingroup\$ Are you asking why SiO2 is an insulator? Why oxides in general are often insulators? From a materials perspective? Or why we want an insulator at all? Many of these would be off topic here. \$\endgroup\$– MattNov 14, 2021 at 13:45
-
\$\begingroup\$ @Matt : I know tha twe need an insulator. The question is why SiO2 is an insulator ? Why oxides in general are insulators ? What is an oxyde formula ? Is an oxyde formula a formula that contains O2 ? \$\endgroup\$– Mathieu KrisztianNov 14, 2021 at 13:48
-
3\$\begingroup\$ O2 is not an oxide (unless you consider oxygen monoxide as one). Oxides in general are compounds of something with oxygen (eg. SiO2). Due to the tendency of oxygen to form strong tight ionic bonds with most metal atoms, most oxides are solid insulators. \$\endgroup\$– tobaltNov 14, 2021 at 13:48
-
\$\begingroup\$ @tobalt : thank you \$\endgroup\$– Mathieu KrisztianNov 14, 2021 at 13:50
-
6\$\begingroup\$ Oxides are not always insulators. Indium tin oxide is conductive, and used in flat panel displays, for instance. \$\endgroup\$– HearthNov 14, 2021 at 14:37
|
Show 6 more comments
3 Answers
\$\begingroup\$
\$\endgroup\$
13
The "oxide" used in semiconductors is most often Silicon dioxide which is commonly known as glass.
It's an insulator because the molecule has filled valence orbitals which makes it stable. The molecule thus has no need to either give away electrons or accept electrons. So there is little electron mobility which would allow it to conduct. Therefore, it's an insulator and an excellent one at that.
-
9\$\begingroup\$ Silicon dioxide is not glass, as a category, and it does not have molecules. The silicon and oxygen atoms are covalently bonded. In terms of valence bond theory, the valence band is almost entirely full, the conduction band is almost entirely empty and the band gap is large, so silicon dioxide is an insulator: it does not have significant numbers of charge carriers than can move in response to an applied electric field. \$\endgroup\$– Ed VNov 14, 2021 at 15:18
-
3\$\begingroup\$ It just is. Mostly found in nature as quartz, though. Also, most glass contains only about 75% SiO2. \$\endgroup\$– ocrduNov 14, 2021 at 16:56
-
3\$\begingroup\$ glass is not a precise term. in application context it's a transparent solid of which there are many (e.g. diamond, al2o3 and gazillion others). In material science context, glass means mainly amorphous condensed matter. Again, any material can assume a glass structure given low enough temperature when being formed. And Alpha-SiO2 would not be considered a glass. And large anorganic compounds are almost never formed of individual molecules. \$\endgroup\$– tobaltNov 14, 2021 at 17:15
-
6\$\begingroup\$ @jwh20 It's what's known as a covalent network solid. Other examples include diamond, boron nitride, and elemental silicon. \$\endgroup\$– HearthNov 14, 2021 at 18:04
-
6
\$\begingroup\$
\$\endgroup\$
2
Wiki page on bandgap.
It is the energy required to promote a valence electron bound to an atom to become a conduction electron
The bandgap of SiO2 is 9.3eV link.
Compared to that, the bandgaps of other semiconductors are:
Si == 1.12eV
Ge == 0.7eV
GaAs == 1.5ev
That's why SiO2 can be considered an insulator for the range of voltages commonly encountered in semiconductors.
EDIT-1 & II
Mechanisms of dielectric breakdown have been discussed by Klootwijk et al, in a series of papers I and II.
-
1\$\begingroup\$ This is a very important point! While band gaps of amorphous materials are a little tricky, the gaps and the conduction bands that define them are still there, and one can potentially have conduct through them if the incident electron energy is high enough. e.g. Diamond is a wide-bandgap semiconductor (Egap = 5.47 eV) with tremendous potential as an electronic device material in both active devices, such as high-frequency field-effect transistors (FETs) and high-power switches, and passive devices, such as Schottky diodes. \$\endgroup\$– uhohNov 15, 2021 at 2:55
-
\$\begingroup\$ @tobalt My choice of language was poor. I have removed my comment. \$\endgroup\$ Nov 15, 2021 at 15:32
\$\begingroup\$
\$\endgroup\$
5
SiO2 is formed in the clean room during manufacturing.
SiO2 is an excellent insulator and is one of the keys of Silicon success.
SiO2 is an insulator because there are no free electrons in the outer orbitals of the crystal.
answered Nov 14, 2021 at 16:44
-
1\$\begingroup\$ What are "outer orbitals of the crystal" ? There isn't even a crystal to begin with in SiO2 used as oxide in Si manufacturing..Its amorphous. \$\endgroup\$– tobaltNov 14, 2021 at 17:17
-
\$\begingroup\$ Ok it's amorphous but still there aren't free electrons that can move. \$\endgroup\$ Nov 14, 2021 at 17:48
-
1\$\begingroup\$ Agreed but then why call it outer orbitals of the crystal ? maybe you mean the other atomic orbitals that are close to the fermi energy? \$\endgroup\$– tobaltNov 14, 2021 at 17:52
-
\$\begingroup\$ By outer orbitals I mean the orbitals that are more distant from the nucleolus. \$\endgroup\$ Nov 14, 2021 at 21:30
-
\$\begingroup\$ @ElliotAlderson Do you have anything to back that up ? Usually SiO2 grown either natively or thermally (wet or dry) at up to many 100°C will be amorphous even when on single crystalline silicon. Also please look at Fig. 1.5. in this PDF that I just googled: iwailab.ep.titech.ac.jp/pdf/201103dthesiskawanago.pdf -> Clearly amorphous image taken from a "production" 90nm node product. If gate oxide was epitaxial, then people would surely also use epitaxial semiconductors on top :) \$\endgroup\$– tobaltNov 15, 2021 at 13:32