80
4 Ultrathin Layers
largest technological relevance. Nitride layers can be produced by the reaction of
metal chlorides and ammonia, while elemental metallic layers can also be produced
at higher temperatures in various catalytic processes. Mixed oxides and multilayered oxide structures are also available with ALD with a thoughtful combination of
the precursor compounds. Various reviews of this topic are available for interested
readers [1–3].
An important feature of the ALD process is that the reactions leading to the
formation of the surface layer are surface-are-limited.
1 This means that once the
surface is covered with the newly introduced reactant, the process stops and the
thickness increment in each cycle is constant, regardless of the excess of the available
reactant in the gas phase. A direct consequence of the surface-area-limited nature
of the ALD process is that it leads to a highly conformal deposit on the surface,
practically independently of the shape of the workpiece. Beside the variability of the
ALD process concerning the surface layer composition, the conformal nature makes
it ideal to cover objects with a complex shape, especially as compared to physical
methods where the deposit is grown by a process having a preferential direction (like
evaporation and sputtering).
The experimental setup of ALD devices includes reservoirs of the reactants, a
tank for an inert gas, a pump system and valves. The entire system can be automated.
The reaction of the surface with one of the reactants is followed by the removal of
the reactant gas, flushing of the system with an inert gas and introduction of the new
reactant. Since the full coverage of the surface requires a minor amount of the reactant, the pressure is usually much below the atmospheric conditions. The exposure
time must be high enough so that the reactant gas can reach all surface sites of the
workpiece even at the bottom of the cavities and pores (it is which usually happens
within a few minutes). Hence, the ALD process is quite time-consuming. Another
drawback of the ALD process is that the leftover precursor materials introduced to
the system in a particular cycle cannot be used in the next cycle, which makes the
process inefficient with respect to the usage of the precursor material.
An ALD-based process using liquid phases is the so-called successive ionic layer
adsorption and reaction (SILAR) method. This is used mostly for the deposition of
oxide and chalcogenide materials by immersing the substrates alternately in reactantcontaining and rinsing solutions. The sequence of the steps in the SILAR processes
is essentially the same as for the ALD method with the difference that ALD always
takes the advantage of the reaction of molecules while SILAR uses dissolved ions as
reactants in a liquid phase. A deeper insight into the current state of SILAR-based
research is available in various reviews [4, 5].
1 Throughout this book, the term “surface-area-limited” will be used in order to indicate that we
mean the limitation of the number of active sites, not only the zone of the reaction. When the term
“surface-limited” is retained for sake of the compliance with the literature (i.e., in the name of some
methods that are commonly used elsewhere), the maximum extent of the reaction is also determined
by the surface area, but it will not be mentioned again.
4 Ultrathin Layers
largest technological relevance. Nitride layers can be produced by the reaction of
metal chlorides and ammonia, while elemental metallic layers can also be produced
at higher temperatures in various catalytic processes. Mixed oxides and multilayered oxide structures are also available with ALD with a thoughtful combination of
the precursor compounds. Various reviews of this topic are available for interested
readers [1–3].
An important feature of the ALD process is that the reactions leading to the
formation of the surface layer are surface-are-limited.
1 This means that once the
surface is covered with the newly introduced reactant, the process stops and the
thickness increment in each cycle is constant, regardless of the excess of the available
reactant in the gas phase. A direct consequence of the surface-area-limited nature
of the ALD process is that it leads to a highly conformal deposit on the surface,
practically independently of the shape of the workpiece. Beside the variability of the
ALD process concerning the surface layer composition, the conformal nature makes
it ideal to cover objects with a complex shape, especially as compared to physical
methods where the deposit is grown by a process having a preferential direction (like
evaporation and sputtering).
The experimental setup of ALD devices includes reservoirs of the reactants, a
tank for an inert gas, a pump system and valves. The entire system can be automated.
The reaction of the surface with one of the reactants is followed by the removal of
the reactant gas, flushing of the system with an inert gas and introduction of the new
reactant. Since the full coverage of the surface requires a minor amount of the reactant, the pressure is usually much below the atmospheric conditions. The exposure
time must be high enough so that the reactant gas can reach all surface sites of the
workpiece even at the bottom of the cavities and pores (it is which usually happens
within a few minutes). Hence, the ALD process is quite time-consuming. Another
drawback of the ALD process is that the leftover precursor materials introduced to
the system in a particular cycle cannot be used in the next cycle, which makes the
process inefficient with respect to the usage of the precursor material.
An ALD-based process using liquid phases is the so-called successive ionic layer
adsorption and reaction (SILAR) method. This is used mostly for the deposition of
oxide and chalcogenide materials by immersing the substrates alternately in reactantcontaining and rinsing solutions. The sequence of the steps in the SILAR processes
is essentially the same as for the ALD method with the difference that ALD always
takes the advantage of the reaction of molecules while SILAR uses dissolved ions as
reactants in a liquid phase. A deeper insight into the current state of SILAR-based
research is available in various reviews [4, 5].
1 Throughout this book, the term “surface-area-limited” will be used in order to indicate that we
mean the limitation of the number of active sites, not only the zone of the reaction. When the term
“surface-limited” is retained for sake of the compliance with the literature (i.e., in the name of some
methods that are commonly used elsewhere), the maximum extent of the reaction is also determined
by the surface area, but it will not be mentioned again.