2

Ion implanters

An ion beam source generates a flux of ions with high energy, which is translated to high velocity in a low-pressure atmosphere, prior to impacting a surface to be ion-implanted. The history of ion beam sources began with Lord Ernest Rutherford’s experiments. With a view to applications, two kinds of industrial approaches are considered here: conventional beam line ion implanters and plasma immersion ion implantation (PIII). The effects of the dynamics the matrix sheath and the Child’s Law sheath in plasma immersion ion implantation are explained. Methods to eliminate problems with surface charging when treating the surfaces of insulators, such as the use of additional mesh electrodes, are examined. Methods of fluence measurement and estimation for polymer targets are outlined. PIII treatment of inner tube surfaces, 3D shapes, particles, wet chemical post-treatment and in-situ measurements are also considered.

Keywords

ion beam implanter; plasma immersion ion implantation; sheath; fluence; charging effects

Ion beam treatment requires a source of energetic ions (e.g., an ion beam accelerator). An ion beam source generates a flux of ions with high energy, which is translated to high velocity in a low-pressure atmosphere, prior to impacting the surface to be treated. The ion beam system includes an ion source (e.g., glow discharge in a gas or magnetron sputtering target), which produces ions typically as part of a plasma discharge. The ions are then typically accelerated in an electrical field to provide sufficient energy for implantation. The energy of ions is then significantly higher than the kinetic energy of the plasma species or atoms in the ion source.

Development of ion implanters

Historically, ion beam sources were developed as a consequence of experiments designed to probe atomic structure (Lord Ernest Rutherford’s experiments). The first experiments relied on radioactive isotopes to release the high-energy charged particles used. Later experimenters developed accelerators based on forces derived from electrical fields. In 1931, Robert Van de Graaff built an electrostatic generator, which is still used today to provide high-energy ions. In 1932, a cascade generator was developed that produced a 1 MeV proton beam, and the first nuclear reaction in Lithium bombarded by artificially accelerated protons was carried out. Later (1932–1944) accelerators based on synchronizing energy delivery with a resonant trajectory were developed to provide 10–20 MeV proton energies. Since 1950, accelerators are created based on cyclotron resonance, and energies up to some hundred GeV are now available. Such accelerators are used in high-energy physics to accelerate electrons or protons. The development of ion accelerators for heavy ions began in the 1950s for the division of uranium and plutonium isotopes required for nuclear weapons. Subsequently, in the 1960–1970s, heavy-ion accelerators were used for the modification of materials, including polymers (Los Alamos National Laboratory, Sandia National Laboratories, Kurchatov Institute of Nuclear Physics). In the 1970–1980s, new ion beam sources were developed for industrial processing of metals and semiconductors. The ion beams with high current density were applied for the modification of alloys, steels, and glasses, and for doping technologies of silicon and germanium wafers. Such sources then began to be used for polymer modification. But, the cost of ion beam accelerators was too high for industrial applications, and polymers were treated only as part of curiosity-driven research projects. In the 1990s, the price of ion beam sources decreased together with an increase in the ion beam diameter, and ion beam implantation became profitable for polymer materials.

Here, we do not consider all of the types of accelerators that were used for ion beam implantation of polymers. Detailed reviews on ion beam accelerators can be found in the literature [111]. Here, we will focus on ion beam sources used for industrial applications related to polymer materials. There are two types of ion beam sources frequently used for polymer modification: ion beam implanters and plasma immersion ion implanters.

The first method, often referred to as conventional beam line ion implantation, relies on the use of biased grids to accelerate a beam of ions out of a plasma source (Figure 2.1). There are two kinds of ion beam implanters: continuous and pulsed ion beam implanters. In first case, the accelerating field and plasma source works continuously; due to space charge effects in the beam, the ion currents are typically limited to the range of microamperes (μA). In the second case, a pulsed accelerating electrical field with a continuous plasma source—or alternatively a continuous accelerating electrical field with a pulsed plasma source—generates a periodically pulsed high-current ion beam. The current density during a short pulse can reach some A/cm2 but the average current density is typically some μA/cm2.

image
Figure 2.1 Simple schematic diagram of an ion beam implanter.

In all cases, the ion implantation is achieved by placing the material to be modified in the path of the extracted ion beam. The method is intrinsically line of sight, so that uniform treatment of complex 3D surfaces can only be achieved by mechanically rotating the object to allow the ion beam to paint over the entire surface. The advantages of this method include the straightforward selection of a monoenergetic ion beam of a single species for implantation; the ion energy can be up to MeV. Disadvantages of the method include the complexity of the motion required to give a uniform treatment over the surface of a complex 3D form.

An example of the first kind of accelerator is the “Pulsar” ion beam accelerator at the Institute of Electrophysics, Ekaterinburg, Russia. The accelerator includes a plasma source based on an arc discharge, or on a hollow cathode in a magnetic field. An electro-optic system with three high-voltage electrodes produces an ion beam from the plasma discharge. The accelerator generates a beam of gaseous ions (N2, O2, Ar, C3H8, and others), with energy from 2 to 40 keV. The cross-section of the beam is 100–200 cm2, with a current density deviation on a beam diameter no higher than 10%. The beam is pulsed with pulse duration of 0.03–1 ms and pulse repetition frequency of 0.1–100 Hz. The current density can be varied between 10 μA/cm2 and 10 mA/cm2. The source is connected to a low-pressure (up to 10−3 Pa) vacuum chamber. The source is compact and used in technological processes.

The second method of supplying energetic ions to a surface is known as plasma immersion ion implantation (PIII), and it has many advantages as a method for the ion modification of polymers. It is more cost effective than beam line implantation, providing higher fluxes of ions per unit time and eliminating the need to rotate the object being implanted to achieve modification of a 3D surface. But the energy of ions is limited to about 50–60 keV. The PIII method was developed in the late 1980s [12] to implant nitrogen into steel surfaces, for hardening. The basic principle is to immerse the object to be treated (workpiece) directly into a plasma, and to bias it with high voltage (in the kilovolt range) to draw ions out of the plasma and accelerate them into the surface of the workpiece (Figure 2.2).

image
Figure 2.2 Schematic diagram of a plasma immersion ion implanter.

An example of this second kind of ion implanter is the plasma immersion ion implanter at the School of Physics, University of Sydney, Australia (Figure 2.3), developed in-house. An inductive-coupled antenna is placed on the outside of the glass tube to provide power for generating plasma. The glass tube is connected to an aluminium chamber in which the ion implantation is performed. A vacuum of up to 10−5 Pa is provided by turbo-molecular and scroll oil-free pumps. A gas-flow controller controls the rate at which gas flows into the system, to provide stable pressure of the working gas. A radiofrequency (13.75 MHz) power supply provides a power of 50–400 W to the antenna. A matching box provides a stable plasma power by matching the radiofrequency power supply impedance to the plasma discharge. The aluminium chamber has current-carrying coils wound around it that create an axial magnetic field in the chamber. This allows a plasma density in the range of 108–1011 ions/cm3 to be maintained in this chamber. A high-voltage electrode is inserted into the chamber. High voltages from 1 to 40 keV are applied in pulsed mode: pulse duration times are from 5 to 100 μs, at a pulse frequency of 20–400 Hz. The high-voltage pulses are provided by a power supply produced by the Australian Nuclear Science and Technology Organisation (ANSTO). This high-voltage power supply can be operated using a computer. The implantation parameters are monitored using a Langmuir probe to measure the plasma density, using an Ocean Optics spectrometer to record plasma emission lines, and by recording the ion current on the high-voltage electrode. A polymer sample is placed on the high-voltage electrode in the vacuum chamber. This polymer sample can be up to 150 mm in diameter.

image
Figure 2.3 Plasma immersion ion implanter at the School of Physics, University of Sydney.

A mechanical arm can manipulate the polymer sample in the vacuum chamber. This allows multistage experiments, including some in situ characterization, to be performed without contact with the atmosphere. With this mechanical arm, the polymer sample can be moved through an air load lock without opening the vacuum chamber to the atmosphere. It can be placed on the germanium crystal of an infrared spectrometer under vacuum, and Fourier transform infrared-attenuated total reflection (FTIR-ATR) spectra of the sample can be recorded without contact with the atmosphere. A polymer sample can also be posttreated by a variety of aggressive media in a chemical attachment chamber isolated from the main vacuum chamber. This posttreatment can be carried out without contact of the treated polymer sample with atmospheric gases.

Most conventional plasma chambers can be converted into PIII systems by the addition of a high-voltage power supply and a high-voltage electrode. This affords the opportunity of acquiring a PIII system without the expense associated with an ion beam implanter.

Sheath dynamics in plasma immersion ion implantation

The ion beam implantation process in the case of insulating materials, such as polymers, is different when an ion beam source is used, as compared to using plasma immersion ion implantation. In the case of the ion beam source, the beam is generated in a chamber remote from the polymer to be treated. Therefore, the presence of the nonconducting polymer does not influence the ion beam. In the case of plasma immersion ion implantation, the ion flux is generated near the nonconducting polymer target. In this case, the presence of the polymer target does influence the electric field near the polymer surface, and affects the ion beam formation.

As soon as the bias is applied to the high-voltage electrode, a sheath forms around the high-voltage electrode as electrons are repelled from its boundaries on a nanosecond timescale. This leaves a space charge of ions, which shield the rest of the plasma from the high bias, close to the high-voltage electrode. The ions’ higher inertia prevents them from moving on timescales typical of electron motion. The sheath that exists on these short timescales is known as the matrix sheath [12, p. 115], because the ions maintain the matrix associated with their locations in the quasi-neutral plasma prior to electron motion. The matrix sheath forms around the high-voltage electrode and more or less conforms to its shape. The width of this sheath is given by:

sM=(2ε0V0en)1/2 (Eq. 2.1)

image (Eq. 2.1)

where ε0 is the permittivity of free space, e is the electronic charge, V0 is the bias potential applied to the high-voltage electrode, and n is the plasma density. A strong electric field drops across the sheath. Ions, already located within the sheath as well as those that drift into the region from outside, are accelerated toward the high-voltage electrode. The timescale for ion motion (the inverse ion plasma frequency), given by:

τi=2π(ε0Me2n)1/2 (Eq. 2.2)

image (Eq. 2.2)

where M is the ion mass, is typically in the range of microseconds.

In the absence of collisions, the ions gain kinetic energy equal to the potential through which they fall on their way to the high-voltage electrode. They then implant into its surface, with energies up to their charge multiplied by the applied bias voltage. As the ions are accelerated and implanted into the high-voltage electrode, their density in the sheath is reduced and the sheath must expand in order to contain enough positive charge to continue to screen the applied bias from the bulk plasma. This process continues until an equilibrium density profile, consistent with the acceleration of ions in the sheath, is formed. Equilibrium occurs when the ion current entering the sheath is equal to the space charge limited current that flows through it. At this point, the sheath is known as an equilibrium or Child’s Law sheath [12, pp. 115–117], and its width on a planar high-voltage electrode is given by:

sCL=23ε0(2eM)1/4V03/4env (Eq. 2.3)

image (Eq. 2.3)

where v is the component, normal to the sheath-plasma boundary, of the velocity with which the ions enter the sheath. In a nondrifting plasma, v is the Bohm speed, given by

eTeM

image

where Te is the plasma electron temperature in electron volts.

Typical laboratory plasmas range in density between 107–1011 cm−3, with matrix sheath dimensions for a planar high-voltage electrode biased at 10 kV, ranging from 30 cm to 3 mm, respectively, and Child’s Law sheath dimensions ranging from 1.4 m to 1.4 cm, respectively. Clearly, for the low-plasma density range where a background gas is present, collisions will not be negligible. Any collisions with neutrals in the sheath will reduce the energy with which the ions impact the high-voltage electrode. Such effects must be taken into account when determining the depth of ion modification achieved.

Both drift velocities [13,14] of ions greater than the Bohm speed and substrate curvature [15] act to reduce the Child’s Law sheath dimensions. Both of these effects are important to note, because they place limits on the maximum plasma density in which PIII with a particular bias can be applied, for a given high-voltage electrode geometry and plasma drift geometry. If the sheath becomes too thin in any one location, the electric field strength will exceed the vacuum breakdown limit, and arcing across the sheath will result [13]. Aside from interrupting the ion implantation process, arcs can also cause damage through ablation of the polymer target, and, thus, must be avoided.

In the absence of collisions in the Child’s Law sheath, the energies of implanted ions are typically higher than those implanted from the matrix sheath, since all implanting ions are accelerated through the entire sheath width in the equilibrium phase. In this phase, it is in principle possible to keep implanting ions at a steady rate, their supply limited only by their generation rate in the bulk plasma. However, practical limits are imposed by heating and charging of the ion-implanted surface. Both of these effects are much more limiting in the case of polymers than for other target materials.

It is usually necessary to apply the bias in a pulsed mode in order to control the heat load imposed on the high-voltage electrode by the implanting ions and to ensure a good supply of ions. Depending on the nature of the plasma source, the ions are often extracted from the plasma at a faster rate than they are generated, so the pulse off time is required to replenish them.

The two most important aspects of the ion implantation process for determining the modification achieved in the target surface are the fluence (total number of ions per unit area implanted) and the energy distribution of those ions. The fluence varies linearly with the time over which the PIII process is applied. The spread of energies of the ions impacting on the surface is determined by (i) the shape of the voltage pulses applied, (ii) reduction in the voltage appearing on the surface due to the dielectric constant of the polymer and surface charging, and (iii) the extent to which ions collide with other species, including background gas molecules, in the sheath.

Since the ion density in the sheath region is highest in the matrix sheath and most of the matrix sheath ions are not accelerated over the whole sheath width, there is a substantial ion flux in any PIII process, which has energy below that corresponding to the applied bias. This is not the case in beam line implantation, where all ions are delivered to the surface with the same energy.

The proportion of low-energy ions implanted for a given bias pulse length will be substantially higher if the pulse rise time is slower than the characteristic response time of the ions, as given by Eq. (2.2). For most currently available pulsed power supplies, the rise time is between several nanoseconds and several microseconds, so in many practical cases, the rise time will exceed the time taken to form the Child’s Law sheath. In such cases, the spread of energies of ions implanted into a conducting surface using the PIII method is primarily determined by the rise time of the applied voltage pulse [16,17].

Plasma immersion ion implantation of insulators

For insulating surfaces, such as polymers, the buildup of charge on the surface can be a major problem that, if left unchecked, reduces the energy with which subsequent ions enter the surface [18]. This can be an issue with both beam line ion implantation and the PIII method. The presence of surface charge reduces the bias voltage appearing at the polymer surface, and the energies of implanted ions correspondingly. Emmert presented modified Child’s Law sheath equations that allow the effect to be estimated [19].

In the case of polymeric materials, their high secondary electron emission coefficient increases the rate of charging. Because of the electric field present at the polymer surface, secondary electrons are repelled from the surface, leaving behind an equivalent net positive charge. As the charge builds up, the potential at the negatively biased surface rises. If the process continues long enough, the surface will attain the floating potential [20, p. 112] (i.e., the potential that appears on an unbiased object, insulated from ground, and immersed in a plasma—usually a few tens of volts).

As the surface potential rises, the potential difference between the polymer surface and the plasma is reduced, causing a corresponding reduction in the energies of ions implanted into the surface. The reduction in the voltage across the sheath, associated with the buildup of surface charge, also causes a sheath contraction (as per Eq. (2.3)). Oates et al. measured this collapse of the sheath using Langmuir probes [21] for PIII from a filtered cathodic vacuum arc plasma. The measurements showed that applied voltages of several kilovolts were fully compensated on microsecond timescales by the accumulation of surface charge. This indicates that surface charging is a severe problem for the implantation of thick polymer targets.

In cases where the target is a thin polymeric film with a low dielectric breakdown strength compared to the PIII bias being applied, this issue can be neglected. Breakdown from the top of the polymer to the high-voltage electrode occurs by two mechanisms: through the film and across the polymer surface. In these cases, continuous cycles of charging and breakdown occur during the bias pulse. This will cause the potential to oscillate between the applied bias and the applied bias less the breakdown strength, with corresponding fluctuations imposed on the energies of implanting ions. The reductions in ion energy associated with voltage droop-induced sheath collapse can be mitigated by use of pulse durations that are much shorter than the characteristic time of the droop. The flow of plasma electrons to the surface in between pulses serves to neutralize the accumulated charge between pulses. In some cases, the charging effect is not serious. For example, a polyethylene film of 20 μm thickness has maximal breakdown strength of 1000 V, so maximum variations of bias of 5% are therefore expected, if the applied bias is 20 kV.

The charge on the polymer surface may also lead to the problem of arc discharging to the plasma sheath. When the charge on the top of the polymer target becomes high enough, the electric field is concentrated in the polymer. The electric field between the polymer and the plasma sheath decreases or becomes zero. When breakdown through the polymer or across the polymer surface occurs, it generates a number of charge carriers (electrons and ions) with high local concentration. After breakdown, the electrical field in the polymer becomes zero and the electric field between the top of the polymer and the plasma (i.e., the field in the sheath) becomes high again. However, the high concentration of charge carriers in the local area near the polymer surface may then cause breakdown across the plasma sheath. The breakdown has an avalanche character and continues until the end of high-voltage pulse. This is observed as arcing with light emission and high current, and the damping of high voltage. The polymer burns locally and the ion fluence distribution is nonuniform on the polymer surface. Such a strongly nonuniform ion beam fluence observed experimentally by the mapping of UV spectra is shown in Figure 2.4. The observed absorbance distribution depends on local conductive properties of the polymer sample and shows substantial nonuniformity.

image
Figure 2.4 Fluence distribution on a polyethylene film target after PIII without mesh. The highly nonuniform fluence with local spikes is a result of arcing due to the buildup of surface charge on the polymer.

Mesh-assisted PIII is a method often adopted to reduce the charging problem in polymers [2224]. With this approach, a conducting mesh, which is biased in the same way as the high-voltage electrode, is placed a small distance in front of the polymer (Figure 2.5). The plasma sheath then forms around the mesh and the space between the mesh and the polymer is free of substantial electric fields. The electric field, which accelerates the implanting ions, forms in the sheath between the mesh surface and the bulk plasma. The ions are then accelerated to the mesh surface and those that pass through the mesh holes continue on to the polymer target. Although the implanted ions still create an accumulating surface charge, the increases in charging rate due to the emission of secondary electrons are usually avoided. As the surface becomes slightly more positive than the negatively biased mesh, the secondary electrons generated at the polymer surface are attracted back to the polymer, resulting in no net charging from this source. This charge compensation from secondary electrons is best achieved with conformal meshes held close to the surface and completely enclosing the polymer. Disadvantages associated with the use of meshes include surface contamination arising from material sputtered from the mesh. In some cases, implanted fluence surface distributions are modulated by an image of the mesh, which can be excluded by selection of appropriate mesh hole size and mesh distance from polymer.

image
Figure 2.5 (A) Plasma immersion ion implanter with additional mesh electrode. (B) Mesh electrode for PIII. The diameter of the mesh is 100 mm, the cell size of mesh is 0.7 mm and its distance from the polymer surface is 30–45 mm. The diameter of the resulting uniform dose distribution area on the polymer surface is about 50 mm.

Figure 2.6 demonstrates the reduction in the fluence distribution deviations achieved by using mesh-assisted ion implantation. The figure shows a map of the UV absorbance for a polyethylene film after PIII with a mesh of 100 mm diameter placed in front of the polymer target and electrically connected to the substrate holder. Without the mesh, the fluence distribution was highly nonuniform, showing variations of up to 500% (Figure 2.4). With the mesh in place, the deviation of the fluence on the polymer surface is reduced to about 10%. In this geometry, the mesh provides an area of uniform fluence distribution on the polyethylene surface, with a diameter of about 50–70 mm (Figure 2.6). This distribution does not depend on the properties of the polymer but rather the edge effects of the fluence distribution are caused by shadowing and focusing of the ion beam by the support walls of the mesh cap.

image
Figure 2.6 Fluence distribution on a polyethylene film target after PIII with a mesh electrode as measured by UV absorbance.

Treatment of 3D insulating objects is significantly more difficult than the PIII treatment of analogously shaped electrically conducting objects. Using a mesh can be beneficial, although when the 3D polymer object has negative curvature or cavities, the fluence on the surface will vary due to nonnormal incidence of the ions. A calculation of the ion trajectories and adjustment of the mesh shape to compensate can help to reduce this fluence variation.

The treatment of polymer objects with particular complex shapes can be achieved with specially tailored approaches. For example, the inner surface of a long elastic tube can be treated after inverting the tube so that the inner surface becomes the outer surface. If the tube cannot be inverted (long tube, hard material, thick wall tube), the inner surface can be treated using a tailored PIII treatment electrode configuration (Figure 2.7). The working gas is pumped through the tube, the plasma is generated in the tube, and the high-voltage electrode surrounds the tube downstream with respect to the gas flow. In this case, the gas is ionized and the ions are accelerated inside the tube. This approach works for very thin tubes, and fluence uniformity can be improved by motion of the tube with respect to the electrodes. We have treated tubes of 1–2 mm inner diameter in this way. As the plasma sheath expands, it will soon consume all of the plasma, due to small tube diameter, so ion implantation cannot be sustained over long pulses. Desired fluences can be achieved by using short pulses but increased number of pulses. The implantation time can be maintained for a given fluence by increasing the pulsing frequency to compensate for the reduction in pulse length.

image
Figure 2.7 A specialized electrode arrangement for the PIII treatment of the inner surface of small-diameter polymer tube. Working gas is flowing through the tube at low pressure. Plasma is generated in the tube and transported along the tube. High-voltage bias is applied to accelerate the ions toward the inner surface of the tube. The A–A′ region of the tube becomes treated.

Fluence nonuniformity can also arise due to nonuniformity of the plasma density. Figure 2.8 shows an example of this situation, where a long tubular ePTFE vascular graft was treated in a plasma with a density gradient along the length of the tube. In some cases, a second treatment, with the sample remounted to reverse the direction of the gradient as a function of position on the sample, can eliminate the fluence gradient. Rotation of the samples during treatment can also be used to improve the uniformity of the fluence. This approach was applied to achieve uniform treatment of balloons for vascular stents (see Figure 2.9).

image
Figure 2.8 Image showing the PIII treatment of an ePTFE vascular graft. The ePTFE tube is mounted on a metal bar in the PIII chamber. The rf-powered plasma generating antenna is underneath this chamber. The plasma density therefore decreases gradually from the bottom to the top of the chamber. The PIII fluence along the tube is proportional to the plasma density. The uniformity of the fluence can be restored with a second treatment, where the tube is mounted the other way up.
image
Figure 2.9 A schematic diagram and images showing the PIII treatment of stent balloons. The balloons (of length 20 mm and end diameter of less than 1 mm) were fixed on a holder and turned during the treatment. The fluence is uniformly distributed over the balloon surface (bottom photo of three untreated and one treated balloons). The experiments were done in cooperation with Irina Kondyurina and supported by Boston Scientific.

Estimating fluence and practical process considerations

The fluence of implanting ions per unit time can be calculated from the pulse parameters according to:

F=j·τ·fe·n (Eq. 2.4)

image (Eq. 2.4)

where j is the current density during the voltage pulse, τ is the pulse duration, f is the pulse frequency, e is the electron charge, and n is the mean charge state of the ions. For an ion beam implanter, this formula can be used, if all parameters are known. In most cases, the mean charge state of the ions and the current density during the pulse cannot be calculated from the ion source parameters. The current density can be measured using a Faraday cap, and the ion types being implanted can be determined using a mass spectrometer or optical emission spectrometry.

A Faraday cup is typically used for the measurement of current density in ion implanters (Figure 2.10). The Faraday cap consists of two metal cylinders: the internal cylinder is connected to a measuring device (current integrator or oscilloscope); the external cylinder is under a negative DC potential to exclude the effect of secondary electrons sputtered from the internal cylinder. The ions impinge on the base of the internal cylinder and generate a current. The current measured in the internal cylinder divided by the area of the cylinder base gives the current density.

image
Figure 2.10 Faraday cup scheme for current density measurements. The ions are flying to the inner electrode. The falling ions bring a charge to the electrode and generate electrical current, which is measured (I). The secondary electrons sputtered from the inner electrode reflected back with the potential (U) of the external electrode. If the charge of ions is known, the current corresponds to the number of ions.

In the case of PIII treatment, the Faraday cup method is not applicable. An indirect method of measurement of fluence according to structural transformations that occur for known polymers is described in Chapter 4.

Monitoring of the plasma source and the ion beam with electrical and optical measurements during implantation is very useful, because the implantation effects are strongly influenced by contamination of the chamber environment. Residual gases, degraded components of polymers, and volatile components of hermetic sealing, as well as residual vacuum oil and grease, can significantly influence the polymer surface layer during and after ion beam treatment.

It is important to clean the chamber when changing working gas, because the previously used gas has been implanted into the chamber walls and it may constitute 5–10% of the process gas on the first operation after a gas change. The cleaning is carried out by generating plasma, with the new gas and maintaining it for a period of time. The plasma parameters and the time needed for cleaning depends on the parameters and geometry of the chamber. In our PIII system, a plasma treatment lasting 30–40 min can clean the chamber if, for example, nitrogen is changed to argon. In the case of chemically active silicon or carbon containing gases, the cleaning process is more complicated and requires understanding of the particular chemistry involved. For example, when acetylene gas is used, an oxygen plasma should be used to reactively etch the carbon coating from the system, and then an argon plasma should be used to remove residual oxygen. Severe cases of contamination may require polishing of the chamber walls.

An oil-containing vacuum pump may create contamination problems, as well. The surface of polymer materials is extremely active after PIII treatment, and any active gases can react with the surface. The oil from diffusion or rotary pumps can diffuse into the chamber and be chemically bonded to the modified surface. Oil-free pumps are preferable when the activity and final chemistry of the surface are critical.

Contamination of the chamber may also occur when polymers with volatile components are treated. For example, PTFE-like polymers degrade strongly in the ion beam and a large amount of fluorine is released. The presence of fluorine ions alters the plasma density (increases it when in a nitrogen plasma) and, therefore, also the fluence of the treatment. When fluorine-free polymers are treated after PTFE-like polymers, a residual amount of fluorine can be found in the treated surfaces. Therefore, the chamber should be properly cleaned with plasma when another type of polymer is to be treated.

If the polymers to be treated contain contaminants, there is a possibility that the chamber materials are degraded. The plasma discharges may etch all materials that they contact. We have observed degradation of glass and metal walls, which become porous and absorb residuals. When plasma is subsequently generated, the residual ions appear in the plasma and affect the polymer surface treatment process. In this case, the degraded parts of the chamber must be replaced.

In some technological applications, gas contamination is not important and a residual atmosphere (nitrogen and oxygen mixture) is sufficient for an effective ion treatment. For such applications, a rotary vacuum pump is sufficient to evacuate the chamber. An example of such a process is the scheme shown in Figure 2.11, used to modify polyethylene film for hardening of the surface layer. Multistage pumping with simple rotary pumps is used to achieve a pressure suitable for plasma and ion beam transportation. The polyethylene film is progressed through a set of volumes delineated by rollers that unwind untreated film from one side, pass it into the treatment region, and then wind it back onto another roll after treatment. The rollers serve also to separate the system into a number of adjacent volumes, to allow multistage pumping to the working pressure and back to atmospheric pressure again. A number of ion beam implanters are used for the modification. The winding speed of the film and the ion current density from the implanters are parameters that can be used to adjust the fluence and optimize the treatment.

image
Figure 2.11 A schematic diagram of an ion beam implantation system for roll-to-roll surface modification of polyethylene film. The atmospheric pressure from 105 Pa is progressively reduced with stages P1=103 Pa, P2=101 Pa, P3=10−1 Pa, and P4=10−2 Pa to the pressure (P5=10−3 Pa) at which the ion modification takes place. Separate rotary pumps are used to evacuate each of the volumes.

For some applications (such as printing), posttreatment of the polymer-modified surface in aggressive gases is needed without exposure to the atmosphere. In this case, a manipulator unit can be used to move the treated polymer into a separate chemical chamber joined to the plasma chamber. An example of this manipulator scheme is presented in Figure 2.12. Periodic chemical and ion processing can be applied for polymer surface modification using a robotic manipulator unit. Separating the chemical chamber from the plasma chamber prevents corrosion of the plasma chamber and polymer contamination before modification.

image
Figure 2.12 A schematic diagram of a PIII chamber connected with a chemical chamber through a sample manipulator unit. Periodic PIII and chemical treatment processes with posttreatment by aggressive chemicals after PIII are possible with this arrangement.

Chemical posttreatment of the polymer can also be done in liquid substances. In this case, the chemical chamber is equipped with a liquid cell and additional manipulator (Figure 2.13). Before the experiment, the liquid cell is pumped to the boiling point of the liquid and filled with inert gas—this operation is repeated until the required residual pressure is achieved. Then the liquid cell is separated with a shutter from the chemical chamber, and the chemical chamber is further evacuated. The polymer is treated, moved into the chemical chamber, and fixed there with the manipulator. The shutter is closed to separate the plasma chamber and the chemical chamber. The chemical chamber is then filled with inert gas to match the pressure in the liquid chamber. Then the liquid cell is opened and the polymer is immersed into the liquid. This operation can be automated.

image
Figure 2.13 Schematic diagram of the chemical chamber for posttreatment of the PIII-treated samples without contact with atmosphere.

Plasma and ion beam treatments have a sterilizing effect when applied to medical devices and tools, and provide for sterile conditions inside the treatment chamber. For biological and medical applications, the sterile polymer devices can be treated in a plasma chamber under sterile conditions. The plasma chamber can be placed into a sterile room or covered with a sterile shield to provide for transfer of the polymer devices into and from the chamber under sterile conditions. Our experience shows that the sterile conditions in the chamber, coupled with appropriate transfer conditions, are sufficient to provide sterile ion-treated medical devices for in vivo implantation.

Ions can also be used for the surface modification of small particles. Figure 2.14 shows a schematic diagram for a suitable system configuration. To ensure that the particles do not enter the pumping system or contaminate the vacuum chamber, the particles to be treated should be placed in a cap and covered with a mesh. For low pumping and venting speeds when the gas stream is not directed at the particles, the particles remain in the cap and damage to the vacuum and high-voltage systems is avoided. To provide the most uniform treatment across the particle surfaces, the cap containing the particles should be vibrated during the treatment, to promote good mixing and rotation of the particles. One possible solution is to have the cap fixed on a shaker head. The shaker motor and mechanical transmitter should be vacuum compatible.

image
Figure 2.14 Schematic diagram of the high-voltage electrode for PIII treatment of polymer particles. The metal cup is installed on the shaker bar and covered with the mesh. The particles are placed into the cup. High-voltage bias could be applied to the cup. The particles are shaking during the ion beam treatment.

The fluence (number of ions per unit surface area) for spherical particles can be calculated from the cap diameter (D), mass of particles being treated (M), particle density (ρ), and particle diameter (d). Assuming that the particles are well dispersed and well mixed by the vibrations applied during the ion treatment time (t), the average fluence for surface treatment of the particles (Fp) can be calculated by the following equation:

Fp=π·F·d·ρ·D224·M (Eq. 2.5)

image (Eq. 2.5)

where F is the fluence of flat substrate (as in Figure 2.4) at the same ion beam and plasma parameters. This calculation gives the average fluence over all of the particles; however, an individual particle can receive more or less fluence. A vibration regime that provides good mixing together with a low ion flux will produce the most uniform fluence distribution on the particles.

The examples of accessories given above are not complete, due to ongoing research and development in this area. The ongoing penetration of ion beam methods into the polymer industry will continue to provide more examples of specific regimes, constructions, and designs of chambers and ion beam sources specially developed for polymer processing.

References

1. Marletta G. In: Bernas H, ed. Materials science with ion beams. Springer 2010;345–371.

2. Friedbacher G, Bubert H, eds. Surface and thin film analysis. Wiley 2011.

3. Hellborg R, Whitlow HJ, Zhang Y, eds. Ion beams in nanoscience and technology. Springer 2009.

4. Avasthi DK, Mehta GK, eds. Swift heavy ions for materials engineering and nanostructuring. Springer 2011.

5. Cuomo JJ, Rossnagel SM, Kaufman HR. Handbook of ion beam processing technology William Andrew Publishing 1989; 438 pp.

6. Byeli AV, Kukareko VA, Lobodaeva OV, Taran II, Shikh SK. The ion-beam processing of metals, alloys and ceramic materials Minsk, Belarus: Physical-Engineering Institute of NASB; 1998; 220 pp.

7. Orloff J, Swanson LW, Utlaut M. 303 pp High resolution focused ion beams: fib and its applications Kluwer Academic Publishers 2002.

8. Bystritskii VM, Didenko AN, editors [Agyei A, Trans.]. High-power ion beams. New York: Springer-Verlag, Incorporated; 1990, 163 pp.

9. Forrester AT. Large ion beams: fundamentals of generation and propagation Wiley, John & Sons, Incorporated 1988; 325 pp.

10. Carter G, Grant WA, Collison JS. Applications of ion beams to materials 1975 Institute of Physics Publishing 1976; 364 pp.

11. Brewer GR, Wilson RG. Ion beams: with applications to ion implantation Krieger Publishing Company 1979; 512 pp.

12. Conrad JR, Radtke JL, Dodd RA, Worzala FJ, Tran NC. Plasma source ion-implantation technique for surface modification of materials. J Appl Phys. 1987;62:4591–4596.

13. Anders A. Width, structure and stability of sheaths in metal plasma immersion ion implantation and deposition: measurements and analytical considerations. Surf Coat Technol. 2001;136:85–92.

14. Brown IG, Monteiro OR, Bilek MMM. High voltage sheath behavior in a drifting plasma. Appl Phys Lett. 1999;74:2426–2428.

15. Bilek MMM. Effect of sheath evolution on metal ion implantation in a vacuum arc plasma source. J Appl Phys. 2001;89:923–927.

16. Kwok DTK, Bilek MMM, McKenzie DR, Chu PK. The importance of bias pulse rise time for determining shallow implanted dose in plasma immersion ion implantation. Appl Phys Lett. 2003;82:1827–1829.

17. Tian XB, Kwok DTK, Chu PK. Modeling of incident particle energy distribution in plasma immersion ion implantation. J Appl Phys. 2000;88:4961–4966.

18. Tian XB, Fu RKY, Chen JY, Chu PK, Brown IG. Charging of dielectric substrate materials during plasma immersion ion implantation. Nucl Instrum Methods Phys Res Sect B. 2002;187:485–491.

19. Emmert G. Model for expanding sheaths and surface charging at dielectric surfaces during plasma source ion implantation. J Vac Sci Technol B: Microelectronics Process Phenomena. 1994;12:880–883.

20. Anders A, ed. Handbook of plasma immersion ion implantation and deposition. New York, NY: Wiley; 2000.

21. Oates T, Pigott J, McKenzie D, Bilek M. Electric probe measurements of high-voltage sheath collapse in cathodic arc plasmas due to surface charging of insulators. IEEE Trans Plasma Sci. 2003;31:438–443.

22. Matossian JN, Schumacher RW, Pepper DM. Surface potential control in plasma processing of materials. Hughes Aircraft Company. USA Patent 995864 [5374456], 20.12.1994, Los Angeles, CA, 1992.

23. Tian XB, Chu PK. Multiple ion-focusing effects in plasma immersion ion implantation. Appl Phys Lett. 2002;81:3744–3746.

24. Tian XB, Fu KY, Chu PK, Yang SQ. Plasma immersion ion implantation of insulating materials. Surf Coat Technol. 2005;196:162–166.

..................Content has been hidden....................

You can't read the all page of ebook, please click here login for view all page.
Reset
18.188.96.232