1063-7397/02/3104- $27.00 © 2002 MAIK “Nauka /Interperiodica”
Russian Microelectronics, Vol. 31, No. 4, 2002, pp. 248–253. Translated from Mikroelektronika, Vol. 31, No. 4, 2002, pp. 292–298.
Original Russian Text Copyright © 2002 by Snitovskii.
The design of high-speed active components is still
a central problem in microelectronics. With respect to
bipolar transistors, this implies reducing the transit
time through the base region [1, 2].
With growing injection level, the lateral portions of
the emitter–base junction are known to exert an
increasingly harmful effect on transistor performance
in common-emitter connection, lowering the gain fac-
tor and the gain–bandwidth product
, as indicated by
calculations and measurements .
This paper presents the results of an experiment that
demonstrated a positive effect of lateral injection on the
parameters and output current–voltage characteristic of
chip n–p–n transistors that were designed for pulsed
conditions and showed
300 MHz. Some process
techniques for such transistors are described as well.
For bipolar transistors, the level of injection across
the emitter–base junction is known to be the major
quantity that determines what factors limit the gain.
With low-level injection, these are
(1) the minority-carrier generation and the recombi-
nation of injected minority carriers in the neutral por-
tion of the active base region,
(2) the recombination in the depletion layers of the
(3) the recombination in the inactive base region.
The third process occurs in the oxide–semiconduc-
tor depletion layer, at the surface states, and inside the
region. The three processes appreciably contribute to
the base current.
Some transistors operate under conditions when the
injected-carrier concentration in the base is comparable
with or much higher than the base doping level . In
that case, the above-mentioned factors act in combina-
tion with the following effects in the base region [3, 4]:
(1) current crowding (Fig. 1) as a result of the lateral
voltage drop beneath the emitter region due to the lat-
eral base current (along the
-axis in Fig. 1);
(2) lateral injection, i.e., the injection at the lateral
portions of the emitter–base junction (Fig. 1);
(3) conductivity modulation;
(4) the Kirk effect; and
(5) the thickness modulation of the neutral base
In view of the above, we propose this:
(1) Transistors should be designed in such a way
that the collector–base junction follow the emitter–base
junction to a larger extent (Figs. 2a, 2b), with the junc-
tions separated by a distance equal to the thickness
of the active base region, which would increase the
effective area of the emitter junction and, thus, the gain
and the gain–bandwidth product.
(2) The conﬁguration of the two junctions should be
such that the lateral portions of the emitter–base junc-
tion have injection efﬁciencies as high as that of the
front portion, which would increase the number of car-
riers injected into the base region across the lateral por-
tions and, hence, increase the injection efﬁciency of the
emitter junction as a whole.
To test the above concept, we fabricated transistors
implementing the new (Fig. 2a) and a baseline (Fig. 2b)
design. The transistors used epitaxial structures grown
on Si(111) that complied with the
7KEF1.5/380EKESO,01 speciﬁcations. A maximum
effort was made to follow identical procedures in mak-
ing the two types of transistor. However, for the new
design, (1) the base region was formed by ion implan-
Lateral Injection Utilized for Improving the Performance
of Microwave Bipolar Transistors
Yu. P. Snitovskii
Belarussian State University of Information Science and Electronics, Belarus
Received November 30, 2001
—An experimental evaluation is presented concerning the common-emitter parameters and output
current–voltage characteristics of chip n–p–n transistors that are designed for pulsed conditions and have gain–
, higher than 300 MHz. The conﬁguration of the emitter and collector junctions essen-
tially embodies a new concept whereby injection efﬁciency is increased by lateral injection. It is shown that the
new approach enables one to improve transistor performance. Some process techniques for the transistors are