The effect of grounding on electrostatic damage
Electrostatic damage during semiconductor manufacturing is nothing new, but despite considerable advancement in the management of electrostatic problems, ESD damage is still being sustained. Reticles (photomasks) are among the most delicate objects in the semiconductor manufacturing environment and since they are critical to the production of virtually everything electronic, a great deal of attention has been paid to preventing them from being damaged. In studying how reticles become damaged, new insight has been gained that is relevant to the damage mechanisms in other electrostatic-sensitive objects.
Gavin Rider, Microtome, Colorado Springs, CO USA
It is commonly thought that if an electrostatic-sensitive object such as a chip is handled in the vicinity of another charged object, it becomes charged by induction and is subsequently discharged (through a spark) if it is picked up using a grounded tool. While a spark may indeed occur as the chip is contacted by a grounded tool, it is not being discharged, it is being charged. To some this may seem pedantic since all that really matters is whether or not the spark occurs. However, having a clear understanding of the process is essential in finding ways to minimize the risk of damage.
When an electrically isolated chip (or any other electrostatic-sensitive object) is brought close to a charged insulator it is not charged inductively, it will be polarized by the electric field but it will remain electrically neutral because no charge has been added to or taken from it. This is shown in Fig. 1a. Inductive charging occurs if the chip is connected to ground while in the presence of the electric field from the static charge, as shown in Fig. 1b.
|Figure 1. a) Polarization of an electrically neutral chip near to a charged surface, and b) inductive charging of the chip by connection to ground.|
It is apparent that the connection of the chip to ground during handling puts it at risk whether or not the chip itself has been charged. If there is any source of electric field nearby, contacting the chip with a grounded handling tool will cause a current to flow by induction and this may cause damage. If the ground contact is broken before the chip is moved away from the charged surface, it will retain the charge that was placed on it by grounding and this can cause further damage if the charged chip is grounded and discharged at its next handling point.
In fact, simply the proximity of the grounded handling tool affects the field strength within the sensitive object and increases the internal electrical stress, such that damage may be caused without ground contact being made. This means that if static charge is likely to be produced by any process, by handling, or may be present on anything else in the handling environment, the last thing that should be done is to move grounded conductors anywhere near to an electrostatic-sensitive object until the static charge has been neutralized.
Since it is currently standard practice to handle electrostatic-sensitive objects with grounded handling tools (normally via a resistor or static dissipative contacts) the perspective on the electrostatic damage problem presented here may cause some raised eyebrows. It should, because it has been found that grounding – normally applied as an ESD preventive measure – increases the risk posed by electric fields under every scenario that has been studied. It is the fixed electrical potential in the presence of an electric field that is the key factor affecting the probability of field induced damage, not the conductivity of the material.
This issue is now being addressed further through the development of a new SEMI Standard guide for the handling of extremely electrostatic-sensitive objects. For further details, please contact the author or SEMI.
Gavin Rider received his PhD in physics from Southampton U., UK and is VP, Technology and Development at Microtome, 3390 Fillmore Ridge Heights, Colorado Springs, CO 80907 USA; ph.: 719-598-8831; email firstname.lastname@example.org.