Martin P.M. Handbook of Deposition Technologies for Films and Coatings, Third Edition: Science, Applications and Technology

Подождите немного. Документ загружается.

Characterization of Thin Films and Coatings 811

beams. Sample topography differences, and orientation of individual polycrystalline or

nanocrystalline grains, are expected to strongly affect the imaging contrast. The crystalline

surface which is aligned with the direction of ion beam is expected to be channeled by the ion

beam and produce low backscattered electrons as opposed to other polycrystalline

orientations. Therefore channeled crystal will appear much lighter than surrounding darker

grains, if the surrounding grains are oriented in the off-normal axis with respect to the

incoming ion beam channeling direction. By using ion beam-induced imaging in conjunction

with conventional secondary electron imaging, one can make interpretations about the

microstructure and grain-boundary formation mechanisms in any polycrystalline sample. This

effect is even more magniﬁed when elements with a high Z value are involved in the analysis.

Strengths:



3D analysis.



High phase contrast ion imaging.



Selected area sample preparation and extraction.

Limitations:



Requires vacuum.



Damages the sample.



Ga deposition and induced damage.

16.4.5 Sputter Depth Proﬁles and Sample Thickness Measurements

M.H. Engelhard

In many circumstances, it is important to obtain information about the elemental or molecular

distribution of components within a ﬁlm. It can also be essential to know the thickness of a

ﬁlm or the layers within a ﬁlm. Sputter depth proﬁling can be a useful way to address both

these requirements. In this section, information is provided both about sputter depth proﬁling

and some methods to determine ﬁlm thickness. The depth proﬁling section provides some

general background and information about some of the more recent advances and new

directions in sputter proﬁling.

16.4.5.1 Composition Proﬁles

It is often important to understand the composition proﬁles of thin ﬁlms or coatings as they are

synthesized, as they age, or when failure occurs. Depending on the thickness and required

resolution, a variety of destructive or non-destructive approaches can be used. Ion-induced

sputter depth proﬁling has been found to be very useful for many different applications.

Sputter depth proﬁling has been well reviewed, including several useful reviews by Hofmann

812 Chapter 16

Table 16.5: Typical ions used in common sputter depth proﬁling methods, with suitable

ion energy ranges

Method Typical ions

used

Typical ion

energies(keV)

Comments

TOF-SIMS Cs

0.2–10 Detect material removed

0.2–10

5–40

Dynamic SIMS Cs

0.5–20 Detect material removed

GDMS Ar

0.3–5 Detect material removed

AES Ar

0.5–5 Detect material remaining on surface

XPS Ar

0.5–5 Detect material remaining on surface

(some C

)

[110, 111] that examine both general considerations about general aspects of proﬁling and the

methods for obtaining information about interfaces. Sputter proﬁling can be done by several

different methods, including AES, SIMS, XPS, and GDMS. As shown in Table 16.5, there are

both similarities and differences among these methods. As highlighted by Hofmann [111],

some methods examine material removed from the samples (SIMS and GDMS) while other

methods examine material remaining on the surface (AES and XPS). A few factors important

to ion sputtering depth proﬁling are noted below:



Beam energy, angle of incidence, ﬂatness of the sample, sample rotation, and other

conditions signiﬁcantly impact the rate and overall quality of sputter depth proﬁles

[112, 113]. An example of the impact of sample rotation on resolution of the depth

proﬁle of a multilayer NiCr sample is shown in Figure 16.40.



Higher beam energies usually produce faster sputter rates, but have decreased

interfacial resolution.



AES and SIMS are known as the methods for which sputter proﬁling has been very

common. The development of higher spatial resolution XPS capability has allowed

increased use of XPS for sputter depth proﬁling.



Many factors inﬂuence the interfacial resolution. These include all of the factors listed

above for depth proﬁle quality, plus (for XPS and AES) the kinetic energy of the

electrons used for elemental analysis. Lower kinetic energy electrons have shorter

IMFP and allow increased interfacial resolution [113].



Among the recent developments in ion sputtering is the increased use of cluster ion

beams (C

,SF

, etc.) [114, 115]. They are especially useful for depth proﬁling of

organic materials, biological tissues, and polymer materials. As an example, the

Characterization of Thin Films and Coatings 813

Figure 16.40: Auger electron spectroscopy sputter depth proﬁles of a NiCr multilayer ﬁlm (30 nm

layers) generated using 3 kV Ar

beam with and without sample rotation. The data show that

rotation can signiﬁcantly improve the depth resolution and allow high-quality data to be collected

from the sample [113].

differences between Ar

and C

sputtering and XPS analysis of plasma-treated

ethylene-tetraﬂuoroethylene-copolymer (ETFE) [116] are highlighted in Figure 16.41.

There is little apparent damage with the C

sputtering and almost immediate damage

with the Ar

. In addition, TOF-SIMS experiments shows that molecular information

can be preserved after C

sputtering in the ice matrix or trahelose matrix [117, 118].

It should be noted, however, that C

or any form of cluster sputtering does not work

ideally in all circumstances and that the impact of sputtering by many types of cluster

is subject to a good deal of current research [119]. Seah [120] has proposed a

framework for understanding the yield dependence of cluster sputtering depending on

the ion mass and atomic number. Mahoney and co-workers [121] have examined the

814 Chapter 16

Figure 16.41: XPS sputter proﬁles generated using Ar

and C

sputtering of plasma-treated

ethylene-tetraﬂuoroethylene-copolymer (ETFE). Note that there is a rapid variation in the F and C

signals for Ar

, while the C

ion sputtering is more consistent with expected composition of

ETFE [116].

temperature dependence of damage, while Yu et al. [122] suggest that a combination of

and Ar

sputtering can extend the useful range for sputter proﬁling organic ﬁlms.

16.4.5.2 Film Thickness

Many different methods can be used to determine ﬁlm thickness. The importance of the

different methods will vary with the nature and thickness of the ﬁlms and general comments

may not apply in all circumstances. However, some of the common methods to determine

thickness are listed in Table 16.6 along with the nominal thickness ranges where they most

commonly apply. Some of the issues or concerns related to applying the methods are

noted. It is important to realize that some methods require a vacuum and some are

destructive.

Particularly for ultrathin ﬁlms, which are of increasing importance for some electronic

materials and other applications, there has been signiﬁcant work conducted considering the

advantages, limitations, and overall consistency of the various thickness measurement

methods [123]. It seems particularly important to note that there are apparently some small but

systematic offsets among the different measurement methods that fundamentally relate to what

Characterization of Thin Films and Coatings 815

Table 16.6: Nominal thickness ranges for different techniques

Technique Nominal thickness

range

Comments

Ion sputter proﬁle – XPS 3 nm < d < ␮m Composition and thickness,

destructive

Ion sputter proﬁle – AES 3 nm < d < ␮m Composition and thickness,

destructive

Ion sputter proﬁle – static SIMS 3 nm < d < ␮m Composition, thickness, dopant

proﬁles, destructive

Ion sputter proﬁle – dynamic SIMS 3 nm < d < ␮m Composition, thickness, dopant

proﬁles, destructive

Angle resolved XPS d < ∼10 nm Composition, thickness,

non-destructive, model dependent,

requires ﬂat samples

Peak to background XPS d < ∼10 nm Composition, thickness,

non-destructive, model dependent

TEM 1nm<d < ∼␮m Based on cr oss-section

RBS 0.1 nm < d < ∼1 ␮m Best for heavy elements on light

substrate

NRA – resonant 8 nm < d <3␮m H-NRA

NRA – non-resonant 50 nm < d <3␮m

ERDA 0.1 nm < d < 500 nm H ERDA and heavy ion ERDA

MEIS 0.1 nm < d < 0.2 ␮m

GDMS 0.5 ␮m<t <10␮m Trace element sensitivity

XRR 5nm<t < ∼100 nm Thickness and roughness

information

Stylus proﬁlometer 50 nm < t < 500 ␮m Measures height variations, needs

appropriate edge to determine

thickness; provide measure of

roughness; stylus usually

non-destructive, but varies with

material

Optical proﬁlometer 50 nm < t < 500 ␮m Opaque sample requir ement,

needs edge to determine thickness,

Estimated roughness analysis,

non-destructive

816 Chapter 16

is being measured and how the measurement is understood [124]. Especially for very thin

ﬁlms, knowing about such differences can be important in obtaining a consistent

understanding of ﬁlms.

Strengths:



Quantitative depth proﬁles.



Accurate ﬁlm thickness measurements.

Limitations:



Requires vacuum.



Extensive sample damage.

16.5 Incident Electron Methods

Methods involving incident electrons include Auger electron spectroscopy (AES), scanning

electron microscopy (SEM), energy dispersive X-ray spectroscopy (EDXS), transmission

electron microscopy (TEM), and electron diffraction (low-energy electron diffraction (LEED)

and reﬂection high-energy electron diffraction (RHEED)).

16.5.1 Auger Electron Spectroscopy

A.S. Lea

Because of its ability to focus an electron beam, AES has the highest spatial resolution of any

conventional surface analysis tool. AES has been highly important for electronic materials,

analysis of particles and grain boundaries, and for use in obtaining sputter proﬁles of many

types of materials. AES relies on the Auger process to probe the elemental composition of the

outermost layers of surfaces. First discovered by Auger and Meitner in the 1920s, the Auger

process results from an energetic electron or photon, typically 2–20 keV, stimulating the atom

by creating a hole in a core shell through ejection of an ionizing electron [125, 126]. The atom

relaxes when an outer shell electron drops down to ﬁll the core level hole and the excess

energy is then released by one of two competing processes: through the release of a photon or

the release of a third outer shell electron, the so-called Auger electron. Since electron shell

energy levels are unique to each atom, the energy of this Auger electron is characteristic of the

Characterization of Thin Films and Coatings 817

Figure 16.42: Schematic of the Auger process showing that a core shell vacancy in the K level is

ﬁlled by an electron from a higher energy level and a third electron, the Auger electron, is emitted.

atom from which it originated and depends on the electronic transitions involved. For the

process involving the energy levels depicted in Figure 16.42, the kinetic energy of the Auger

electron is given by Eq. (16.6):

2,3

= E

− E

2,3

− φ (16.6)

where E

is the energy of the ionized core-level electron, E

is the energy of the outer level

electron that drops down to ﬁll the hole, E

2,3

is the energy level of the Auger electron, and φ is

the spectrometer work function, i.e., the energy necessary to promote an electron from the

Fermi level to the vacuum level.

While it may appear that a plethora of transitions are possible yielding numerous Auger

transitions in the energy spectrum, particularly for high Z elements, only a few occur with

sufﬁcient probability that they are visible in the Auger spectra. This minimizes the appearance

of overlapping peaks and makes AES spectra relatively easy to interpret. This is due to both

the electron-impact cross-section (how easily core holes are created) and the Auger and

ﬂuorescence yields. The energy level of the core hole often determines the transitions that will

be favored. For a K shell hole, transitions can occur from the L levels, producing strong KLL

peaks in the energy spectrum. Higher level transitions, such as those from M levels, may also

818 Chapter 16

Figure 16.43: Plot of Auger electron emission (solid lines) and X-ray ﬂuorescence (dashed lines)

yields as a function of atomic number (Z) for each core shell vacancy using data from Hubbell

et al. [127]. AES has good sensitivity for low Z elements.

occur, but are much less likely to do so. The strongest electron–electron interactions are

between levels that are close in energy, and this gives rise to the characteristic KLL, LMM,

MNN peaks seen in Auger spectra. The probabilities of Auger yield and ﬂuorescence (X-ray)

yield for the various transitions are shown in Figure 16.43 [127]. In the case of a K shell hole,

the Auger yield is highest for low Z elements, whereas the ﬂuorescence yield is greater for the

high Z elements and so AES is much more sensitive to low Z elements than electron probe

microanalysis.

The surface sensitivity of AES, as well as the other electron spectroscopies, is due to the low

IMFP of electrons. Auger electrons, with kinetic energies ranging from about 30 to 2500 eV,

have material-dependent IMFPs (λ) ranging from about 0.2–2 nm. The information depth for

AES is ∼3λ and so the Auger peaks visible in the spectra originate from only 0.5–5 nm

beneath the surface. While the energetic primary beam can penetrate micrometers below the

surface, those Auger electrons that originate at depths below the information depth lose their

characteristic energy owing to inelastic collisions and contribute to the background observed

in AES spectra. The low IMFP also necessitates the use of ultrahigh vacuum conditions to

minimize scattering off residual gas atoms and has the additional beneﬁt of keeping the surface

clean from adsorbates that may collect on the surface at higher pressures.

The highly energetic incident electron beam penetrates deep into the sample and initiates an

electron cascade, whereby inelastic collision of the primary electrons with surrounding atoms

causes these electrons to lose a portion of the primary energy. The energy transfer can also

propagate additional electrons through ionization of atoms within the substrate. These

mechanisms contribute to the broad featureless background observed in the Auger spectra.

Characterization of Thin Films and Coatings 819

Figure 16.44: Raw Auger spectrum of a natural hematite sample. Shown in the ﬁgure are the

Auger peaks on top of a large background that must be removed for accurate quantiﬁcation.

Low-energy electrons emitted from the surface are designated as secondary electrons and are

captured by the secondary electron detector, allowing one to obtain SEM images of the surface

at high resolution during the AES analysis.

Figure 16.44 shows a plot of a typical Auger spectrum with the Auger peaks superimposed on

top of a large background of backscattered primaries, inelastically scattered Auger electrons,

and secondary electrons. This large background, while easily removed through differentiation,

still limits the sensitivity that can be achieved with this technique.

As seen from Eq. (16.6), the kinetic energy of the Auger electron is independent of the energy

of the incident radiation. It is only necessary to impart sufﬁcient energy to the atom to generate

an electron hole in a core-level shell. The probability with which this occurs is greatest when

the energy of the incident radiation is two to three times the energy of the core shell and so the

Auger process is most efﬁcient when the incident radiation is 2–20 keV. Whereas it is possible

to utilize either photons or electrons to initiate the process, electron beams are used in most

instruments since they are more easily focused than X-rays and can provide higher spatial

resolution [128]. While both LaB

ﬁlaments and ﬁeld emitters are high-current sources

utilized in Auger spectrometers, newer generation instrumentation are migrating towards ﬁeld

emission electron guns through which beam sizes below 20 nm are now achievable. This small

beam size combined with the short IMFP provides AES with the smallest analytical volume of

any standard surface analysis equipment.

The schematic of a typical Auger spectrometer is depicted in Figure 16.45. In this set-up,

electron optics focus the electrons generated from either a ﬁeld emission source or an LaB

820 Chapter 16

Figure 16.45: Schematic diagram of an Auger spectrometer including the electron gun, focusing

lenses, sample, and cylindrical mirror electron analyzer.

ﬁlament onto the sample. Electrons produced from the sample enter the coaxial cylindrical

mirror analyzer (CMA) and are focused into the electron multiplier and detector. The voltage

on the outside plates of the CMA is swept such that at any given time, only electrons that fall

within a narrow energy window are able to reach the detector. In this manner, the spectrometer

can scan the energy spectrum and measure the intensity of incident electrons as a function of

their kinetic energy. The spectrometer is often equipped with an ion source for depth proﬁling

experiments, and in the case of newer generation instruments, charge neutralization. Most

spectrometers are also equipped with a secondary electron detector which permits acquisition

of high-resolution SEM images of the surface during AES analysis.

As alluded to earlier, the Auger peaks visible in an intensity versus energy [N(E)vsE]

spectrum sit on top of a large background of backscattered primary electrons, secondary

electrons, and inelastically scattered Auger electrons, and are typically only a small percentage

of the total electrons that reach the electron detector. In order to perform quantitative analysis,

it is necessary to derivatize the spectrum [d(N(E))/dE vs E] to remove the background and

utilize the intensity of the derivatized Auger peaks in the quantiﬁcation. The underlying

equation for quantiﬁcation is given by Eq. (16.7):

Atomic concentration of element i = (I

/RSF

)×100/(

I/RSF) (16.7)