Radiant energy – Ion generation – Field ionization type
Reexamination Certificate
1999-09-14
2003-01-07
Anderson, Bruce (Department: 2881)
Radiant energy
Ion generation
Field ionization type
C250S492210, C315S111810
Reexamination Certificate
active
06504159
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention generally relates to ion implantation processes for manufacture of semiconductor integrated circuit devices and, more particularly, to ion implantation for creating buried layers such as silicon-on-insulator (SOI) substrates.
2. Description of the Prior Art
The art of semiconductor electronic device manufacture has become highly sophisticated in recent years to provide a wide range of electrical properties of the devices, often at very high integration density. The capability to determine the electrical properties with high reliability, consistency and manufacturing yield is often limited, as a practical matter, by the tools used for processing the semiconductor material, usually in the form of a wafer. Such tools are often complex and of high precision. Therefore such tools are generally expensive to build and maintain. The principal expense of modern semiconductor devices is thus a portion of the cost of the tools used to produce them and, therefore, varies inversely with tool throughput. Of course, tool cost of each manufactured unit also is proportional to the cost and complexity and maintenance costs of each of the tools used in its fabrication.
An example of such a process is formation of semiconductor-on-insulator (SOI) substrates. SOI structures can be created by depositing or growing an insulator, (e.g. oxide) on a substrate followed by epitaxial growth or deposition of a further layer of semiconductor. Annealing may be required to develop a monocrystalline structure in the further semiconductor layer. This process is complex, proceeds slowly (and is thus expensive) and is of relatively low yield (further increasing unit cost) since the monocrystalline layer must be substantially free of crystal lattice dislocations; a quality which is difficult to achieve over oxide.
A preferred technique is to form the insulator by implantation of ions into a monocrystalline wafer such that the oxygen ions combine with the substrate material at a desired depth within the wafer to form the buried insulator while leaving the surface monocrystalline layer substantially intact. However, to obtain an ion beam of sufficient purity (e.g. free of ions of other than an intended element or radical) mass analysis is often employed and results in loss of a large fraction of the beam current.
Basically, mass analysis involves passing the ion beam through a dipole magnetic field established across a small gap. Assuming substantially the same velocity of the charged particles in the ion beam, the magnetic field will exert a force on each ion in a direction in accordance with the charge thereon perpendicular to both the direction of motion and the magnetic field. This force results in an acceleration inversely proportional to the square root of the ion (or electron) mass and ions having differing charge or differing mass are placed on different trajectories. Thus, undesired ions can be intercepted and removed from the ion beam resulting in a beam populated by only identical ions.
However, this process also causes loss of significant populations of desired ions from the beam through several mechanisms such as spreading of the beam within the dipole magnet in the direction of the pole pieces such that a significant number of ions strike the faces thereof. Further, the mass analysis process results in a beam of relatively small cross-section which must be scanned across the wafer or workpiece and electrical or magnetic deflection of the ion beam results in further loss of ions from the mass-analyzed beam by similar mechanisms. Perhaps most significantly for oxygen (which is the material of choice for forming SOI structures), ions may be produced in several forms which will be affected differently by the mass analysis process: O
−
and O
+
ions will be separated by charge (and, generally, velocity) and O
−
and O
2
−
ions will be separated by mass.
Accordingly, due to the relatively low beam current and the relatively high concentration of ions which must be delivered to form an insulator of significant thickness and adequate integrity, the ion implantation process proceeds slowly and throughput is very low, resulting in very high tool costs which dominate the unit cost of SOI structures. It is estimated that an 80% reduction in such costs could be achieved through increase of ion beam current. Moreover, while SOI structures are known to have some important performance advantages over other known semiconductor technologies, those advantages cannot be economically exploited without achievement of a substantial portion of that cost reduction.
In an effort to obtain increased ion beam currents (and wider beam cross-section to require, at most, mechanical scanning to avoid loss of ions to electrical or magnetic deflection arrangements), a technique referred to as plasma source ion implantation (PSII) has been developed. This technique involves production of a plasma near or in contact with a surface of a workpiece and pulsing a voltage between the workpiece and the plasma to accelerate the ions to achieve the desired implantation energy and implantation depth.
Pulsing of the accelerating voltage is necessary in order to avoid arc breakdown between the workpiece and the plasma. Pulsing, of course, also reduces the duty cycle of the implantation process and thus limits the speed at which the implantation process could proceed. In general, the time period between pulses cannot be reduced beyond a particular limit determined by other parameters of the process such as ion and electron mobility in order to avoid arc breakdown.
Unfortunately, this technique has been found unsuitable for forming SOI structures since it causes excessive damage to the crystal lattice near the surface of the workpiece beyond practical recovery through subsequent annealing. It has been theorized (see “Boron Doping of Silicon by Plasma Source Implantation” by R. J. Matyi et al., Surface and Coating Technology 93 (1997), pp. 247-253) that the damage is due to ions implanted at the beginning and end of the accelerating voltage pulse having lower energy and which are thus implanted at reduced depth. This observation also implies that PSII would also be unsuitable for producing any buried layer in which the concentration of buried layer material in the overlying layer is at all critical to intended device function (e.g. a buried conductive plate in a memory device). Some damage may be done in the surface layer by contact with the plasma, as well.
Electron-cyclotron resonance (ECR) plasma sources are known in the art and operate well at relatively high vacuum levels. In these sources microwave energy is projected through a dielectric window to form an electromagnetic wave in a plasma chamber in which a strong magnetic field is present. A gas to form the plasma is introduced at an inlet and, as the gas is ionized, electrons are directed in circular trajectories around the magnetic field lines. When the electrons are rotating with the same frequency as the microwave power, the microwave power increases their energy. The electrons will absorb more energy at high vacuum (fractions of a mTorr) since fewer collisions will occur. Therefore, ECR plasma sources are used to facilitate generation of a plasma.
Also, due to the high energy of electrons in the ECR system, molecules of gases from which the plasma is formed are readily cracked to their atomic species. It should be recognized that the number of ions striking the chamber, being neutralized, returning to the gas and again being ionized is many times greater than the number of molecules (or atoms) of gas provided to the plasma/reaction chamber. Therefore, there will be many collisions between high energy electrons and gas molecules or molecular ions (e.g. O
2
+
) to cause gas cracking. It is also known that, by virtue of this mechanism of cracking of molecules of gases, when oxygen is used to form the plasma that a preponderance of O
+
ions will be produced.
Plasma sources with surf
Anderson Bruce
Anderson Jay H.
Whitham, Curtis & Whitham
LandOfFree
SOI plasma source ion implantation does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with SOI plasma source ion implantation, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and SOI plasma source ion implantation will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3024690