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 781

Figure 16.19: Jablonski diagram showing principles of photoluminescence spectroscopy.

Absorption of a photon excites an electron to an excited state from which it decays by a variety of

processes to lower energy states. Observation of the photons emitted provides fundamental

information about the energy levels in the system. See text for more details [57].

For molecular materials, the intensity and proﬁle of the photoluminescence spectra are direct

measures of various important material properties such as the relative energies of the ground

and excited states, electronic transitions and concentrations of the emitting species [58]. The

temporal dependence of the photoluminescence reﬂects the relaxation characteristics of the

excited state, molecular bonding environment, and identiﬁcation and quantiﬁcation of

photoluminescence quenchers. The polarization change of photoluminescence reveals the

electronic transitions involved and the direction of the transition moments. For

lanthanides-based materials and some transition metals, the symmetry is related to band

splitting pattern and for lanthanides, such as Eu(III) and Tb(III), the quenching behavior offers

clues about the vibrational characteristic of its immediate surroundings, such as high-energy

vibrators, O

H and N

H [59, 60]. Fluorescence energy transfer studies allow identiﬁcation of

other ﬂuorophores in the vicinity of a molecule as well as the distance between them and the

nature of their association.

782 Chapter 16

For semiconductor materials, photoluminescence occurs in a similar mechanism.

Photoexcitation of a semiconductor material promotes an electron from the valence band to the

conduction band, leaving a ‘hole’ in the valence band. The excited electron in the conduction

band relaxes quickly, typically within 15–25 ps [61], to the band edge via inter- or

intra-sub-band scattering and then recombines with the hole either radiatively by emitting a

photon or non-radiatively by passing the energy on to one or more phonons, to ‘trapping’

states that are created by impurities, dopants, or defects and lie in between the conduction

band and the valence band (Shockley–Read–Hall recombination), or to another electron or

hole (Auger recombination). For radiative recombination, the emitted photon usually has an

energy equal to the band gap energy, hν = E

, where h is Planck’s constant, ν is the frequency

of the emitted photon, and E

is the bandgap energy. Therefore, photoluminescence spectra

and decay behavior (or lifetime) provide direct measurement of the bandgap energy, which

help identify and quantify the elemental composition of compound semiconductors, and

determine the presence of impurities, dopant concentrations, and structural defects, and

understand the electron–hole recombination mechanisms and evaluation and quality control of

semiconductor materials.

16.3.5.2 Instrumentation

The complexity of photoluminescence instrumentation largely depends on the parameters of

interests. Conventional photoluminescence (more commonly termed ﬂuorescence)

spectrometers typically use xenon lamps as the excitation source. Most xenon lamps are

effective in a broad wavelength range from the UV to near-IR region. The excitation beam is

dispersed through scanning monochromators for sample excitation at different wavelength

(energies). The emitted ﬂuorescence is collected at either a right angle (90

◦

relative to the

excitation beam) or front face (∼30

◦

) and dispersed through a second scanning

monochromator and then detected by a photomultiplier tube. Multiple slit mechanisms are

employed at both the excitation and emission sides to increase the monochromaticity of the

excitation beam as well as the photoluminescence emission. Photoluminescence lifetime

measurements are achieved by replacing the continuous wave xenon lamp with ﬂash xenon or

xenon–mercury lamps or pulsed lasers and recording the photoluminescence intensity change

as a function of time after the pulsed excitation.

The availability of lasers and CCD detectors in the last couple of decades has signiﬁcantly

improved the photoluminescence technique [62]. Lasers provide high-power, coherent

monochromatic light, greatly improving the spectral resolution and detection limit. Pulsed

lasers with pulse width at nanoseconds to femtoseconds not only permit photoluminescence

measurements with lifetimes in a broad range from milliseconds to subpicoseconds, and cover

most photoluminescence materials, but also facilitate a new suite of photoluminescence

measurements, called time-resolved photoluminescence spectroscopy [63]. Photoluminescence

Characterization of Thin Films and Coatings 783

spectra recorded at different delay times after the laser pulse allow temporal resolution of the

emitted ﬂuorescence spectra and thus help with the identiﬁcation of materials with

photoluminescence centers in different chemical environments and the elucidation of the

recombination mechanisms. The use of CCD detectors eliminates the scanning process of the

emission monochromator and thus signiﬁcantly improves the throughput of the measurements.

This is particularly useful when using pulsed laser excitation sources. For these, the entire

photoluminescence measurements can be recorded with a single laser pulse, thus avoiding

measurement errors introduced by the instability and ﬂuctuation of the excitation sources.

16.3.5.3 Photoluminescence Applications in Thin Film Materials and Future Directions

Because of its high sensitivity, providing rich information about the band structure,

electron-hole recombination and relaxation processes, and relative ease of use,

photoluminescence spectroscopy applies to the characterization, photophysical,

photochemical, and mechanistic studies of a wide array of thin ﬁlm materials including

ZnO/II–VI semiconductors [64], III–V semiconductors [65], lanthanide-based optoelectronic

materials [66, 67], inorganic and organic light-emitting diodes (LEDs and OLEDs) [68], thin

ﬁlm photovoltaic materials [69], and radiation detection devices [70]. Interested readers are

referred to many of the previous review articles and book chapters cited here [63, 71, 72].

Zinc oxide is a wide-band (3.37 eV, 60 meV exciton binding energy) semiconductor with

potential applications in photoelectronics, light-emitting devices, gas sensors, and solar cells.

Studies have shown that many properties of ZnO ﬁlms are strongly dependent on the crystal

structure and orientation of ZnO, its crystalline quality and defect states, and thus vary as a

function of ﬁlm preparation method, selection of substrate, and growth conditions. As an

example, for ZnO thick ﬁlms (∼200 nm thickness) deposited on indium tin oxide (ITO glass)

by radio-frequency (RF) magnetron sputtering at a substrate temperature of 400

◦

C, the UV

photoluminescence of the ﬁlm initially increases as the oxygen partial pressure (PO

)

increases from 0% to 60%, but then decreases as PO

increases further (Figure 16.20) [73].

The initial increase probably resulted from the improved stoichiometry of the ﬁlms, associated

with the incorporation of oxygen at oxygen vacancies, while excessive PO

may worsen the

stoichiometry of the ﬁlms by introducing interstitial oxygen or zinc vacancies. Gaussian ﬁtting

of the UV-photoluminescence band gives three peaks situated at 380, 395, and 410 nm,

corresponding to recombination of free excitons through an exciton–exciton collision process,

emission from the ITO layer and O dangling bonds on the ITO surface layer of the interface

between substrate and ITO, respectively. The relative intensity of the 410 nm peak increases

with the excitation intensity.

The versatility of ﬂuorescence detection also makes it possible to integrate photoluminescence

instruments with other experimental techniques such as optical microscopes and scanning

784 Chapter 16

Figure 16.20: Photoluminescence (PL) spectra of ZnO thin ﬁlms on ITO glass with increasing

P(O

) [73].

atomic force microscopes. In addition, the non-contact nature of the techniques allows the

sample to be measured at different temperatures. The current trend in the application of

photoluminescence for thin ﬁlm studies is closely coupled with technical developments in laser

ﬂuorescence spectroscopy and scanning probe microscopy. Examples of such developments

are near-ﬁeld scanning optical microscopy (NSOM) and non-linear optical spectroscopy.

NSOM is based on designs that limit the photoexcitation volume using a tapered, metal-coated

optical ﬁber, or ﬁeld-enhanced excitation by noble metal probes [74]. Using NSOM, spatial

resolution much less than the diffraction limit can be achieved. Combining NSOM with

photoluminescence spectroscopy allows study of the photoluminescence properties of thin

ﬁlm material with ultrahigh spatial resolution. Using this method, Crowell et al. [75] were able

to image photoluminescence in an InGaN/GaN quantum well with a spatial resolution of

approximately 100 nm for temperatures between 50 and 295 K, and directly correlated the

spatial features of the thin ﬁlm material with photoluminescence characteristics (Figure 16.21).

Time-resolved photoluminescence (TRPL) measures photoluminescence intensity as a

function of time. With commercial high-speed detection devices such as time-correlated

single-photon counting and streak camera with femtosecond lasers as excitation sources,

luminescence measurement reaches a resolution of a few picoseconds. A new type of TRPL

technique that is based on sum frequency generation (or frequency up-conversion) has been

demonstrated [76]. This technique offers a time resolution on the order of the laser pulse

width, as short as tens of femtoseconds. Rain

oetal.[77] used this technique to measure the

rise and decay dynamics of the ground and ﬁrst excited states of InAs quantum dots (QDs)

capped with an InGaAs quantum well and found that the higher energy states of the QDs do

Characterization of Thin Films and Coatings 785

Figure 16.21: Images of a 2.5 ×8 ␮m region of the sample at T = 295 K obtained using the

near-ﬁeld technique: (a) topographic image of the InGaN/GaN sample measured using the

shear-force feedback technique; (b) transmission image at the excitation energy of 3.16 eV; (c–e)

photoluminescence images obtained at detection energies of 2.48, 2.88, and 2.95 eV, respectively.

The grayscale for each image is indicated at the right [75].

not act as intermediate stages in the carrier relaxation, while the carriers can cool down to any

lower energy states following relaxation through a continuum background.

Strengths:



Non-contact, non-invasive/non-destructive.



Requires minimum sample preparation.



Sensitive to impurities and defects (low detection limits).



Applies to broad temperature range.



Ease of integration with other techniques.

786 Chapter 16



Offers direct observation of the electronic transition energies, including bandgap

energy.



The energetics and dynamics provide ideal tools in the study of recombination

mechanisms.

Limitations:



Only applies to materials that ﬂuoresce.



Only detects electronic states that relax radiatively.



In some cases, unknown quenching mechanisms prevent quantitative measurement.

16.4 Incident Ion Methods

Ion beams are used in a variety of ways to characterize ﬁlms and surfaces. Most of the

techniques involve either analyzing ions scattered from a sample to extract information about

the sample or using incident ions to remove material from the sample. Both methods are

included in this section. The several techniques involving elastic and inelastic scattering have

several common physical aspects and they are described as a unit, including: (1) low-energy

ion scattering (LEIS); (2) medium-energy ion scattering (MEIS); and (3) high-energy ion

scattering (HEIS) methods such as Rutherford backscattering spectrometry (RBS), nuclear

reaction analysis (NRA), ion channeling, and elastic recoil detection analysis (ERDA). Other

sputtering methods, such as sputter depth proﬁling, secondary ion mass spectrometry (SIMS),

glow discharge mass spectrometry (GDMS) and some applications of focused ion beams

(FIBs) use energetic ion beams. In addition, FIBs are often combined with electron

microscopy as the ion beam can be useful in the collection of high-resolution images. The ﬁnal

section is on sputter depth proﬁling, which is useful for determining depth distributions and

layer thickness. This ﬁnal section also includes a table comparing a variety of approaches for

determining ﬁlm or layer thickness.

16.4.1 Ion Scattering Methods (LEIS, MEIS, HEIS, RBS, NRA, ERDA, Channeling)

V. Shutthanandan and Y. Zhang

Ion beams have been effectively used in material modiﬁcation and material analysis for more

than 40 years. Ions at different energies are used to alter surfaces (or ﬁlms) or to collect

different types of information about them. The picture in Figure 16.22 shows different

capabilities associated with various ion energy ranges. A few hundred eV to a few keV inert

gas ions, in particular argon ions, are extensively used in sputter cleaning of surfaces which are

necessary for surface science studies. In addition, a few keV inert gas ions along with alkali

ions are very useful in determining the structural and chemical properties of surface layers and

adsorbate layers. This capability is called LEIS spectroscopy, which is very surface sensitive

Characterization of Thin Films and Coatings 787

Figure 16.22: Schematic illustration of the uses of ion beams with various energies on Si substrate.

and it is described in detail in the following sections. MEIS spectrometry is a powerful

capability to characterize thin ﬁlms of a few nanometers thickness. In general, about 100 keV

helium or hydrogen ions are used as an incident ion beam and electrostatic analyzers are used

to characterize the backscattered ions at large scattering angles. High-energy ion beams are

generally used for material characterization of both thin and thick ﬁlms. Although several

high-energy ion beam-based techniques are available for material characterization, RBS,

NRA, proton-induced X-ray emission (PIXE), and ERDA are commonly used.

Although not the focus here, ion beams are also important for material deposition and

substrate modiﬁcation. Ions with energies between a few eV and a few tens of eV enable the

‘soft’ landing of larger molecules on the surfaces and the deposition of thin ﬁlms using ion

beam-assisted deposition. Material properties can be modiﬁed using ion implantation and it

has been a very valuable tool in the semiconductor industry. The research community is

investigating this capability for other applications including spintronics and photocatalysis.

Ion implantation is often carried out using ion beams with energies in the range of a few tens

of keV to a few hundreds of keV.

In general, the ion scattering spectroscopies are mass sensitive and mass selected information

can be obtained from the data. As a result, backscattering and forward scattering capabilities

along with nuclear reaction analysis capabilities are isotope speciﬁc. The experimental data

can be simply simulated using classical mechanics principles as discussed in the following

sections. The following elements are important to understanding and modeling ion scattering

from surfaces:



Energy transfer from a projectile to a target nucleus in an elastic two-body collision –

This concept is related to kinematic factor which depends on the mass of the

788 Chapter 16

projectile, target atoms, and the scattering angle. The energy of the outgoing ions is

related to the incoming ions through this factor.



Likelihood of occurrence of such a two-body collision – This factor is deﬁned as the

scattering cross-section and the height of the experimental spectrum is determined by

the scattering factor.



Energy loss of a projectile moving through a dense medium – Since the ion beam loses

energy when it travels through the material, the stopping cross-section which is related

to the energy loss as a function of depth provides depth perception.



Statistical ﬂuctuations in the energy loss of a projectile moving through a dense

medium – When the ion beams travel through the material, the energy spread will be

increasing owing to the interaction with the material as a function of depth. This

‘energy straggling’ needs to be taken into account in interpreting the experimental data

and determining various properties of the material.

Scattering kinematics

The simple solution to the equations associated with energy and momentum conservation

leads to the following equation (Figure 16.23):

K =

⎡

⎢

⎣



− M

sin



1/2

+ M

cosθ

+ M

⎤

⎥

⎦

(16.4)

K is called the kinematic factor and it represents the ratio of the incident energy of the

projectile to the outgoing projectile energy. The masses of the projectile and target atom are

and M

, respectively, and the scattering angle is θ. The outgoing projectile energy,

= KE

, and by measuring the energy of the backscattered ions the composition of the

material can be determined as a function of depth. The variation of K as a function of the mass

Figure 16.23: A cartoon picture depicting the elastic collision between the projectile with mass M

and target atom with mass M

Characterization of Thin Films and Coatings 789

of the target atoms for light projectiles such as hydrogen and helium ions indicates that the

differences in K for light target atoms are larger in comparison to the heavy target atoms. As

such, for the helium ions which are commonly used for RBS, light target atoms can be easily

separated while the separation between heavy target atoms is relatively difﬁcult, in particular

when the elements are close to each other. Heavy target atoms can be separated using heavy

projectiles, but these may damage the sample.

Scattering cross-section

The scattering cross-section, σ, is deﬁned by the following equation:

σ(θ) =







[1 − ((M

)sinθ)

]

+ cosθ



sin





[1 − ((M

)sinθ)

]

K =

⎡

⎢

⎣



− M

sin



1/2

+ M

cosθ

+ M

⎤

⎥

⎦

(16.5)

where Z

and Z

are the atomic number of the projectile and target atom, respectively, and E is

the energy of the projectile. This equation clearly demonstrates that the energy and atomic

number of the target atom depend on the scattering factor as follows:

σ(θ) ∼ Z

σ(θ) ∼ E

−2

Since the scattering cross-section relates to the signal amplitude of the backscattering

spectrum, the backscattering signal would vary as a square of the atomic number of the target

elements. As such, the backscattering signals from the material that consists of heavy elements

will provide larger signals. On the other hand, the signals from the light elements are weak and

the detection and quantiﬁcation of light elements in a heavy element matrix will be difﬁcult.

As such, thin ﬁlms that contain heavy elements on light element substrates can be successfully

characterized using backscattering spectrometry. The energy dependence of the scattering

cross-section exhibits the 1/E

relationship and for E > {1/2} MeV the scattering cross-section

does not vary signiﬁcantly as a function of energy. Less variation in scattering cross-section at

higher energies makes the thin ﬁlm analysis easier and 2 MeV helium ions are generally used

for RBS.

Energy loss and energy straggling

When the energetic ions travel through the material they lose energy owing to nuclear and

electronic interactions with the target atoms. In general, since the radius of the nucleus is

small, nuclear stopping is small in comparison to the electronic stopping. Excitation and

790 Chapter 16

ionization during the inelastic collisions with the electrons in target atoms contribute to the

electronic stopping.

The rate of energy loss is deﬁned as dE/dx, the change in energy (dE) per unit distance (dx) the

ion beams travel inside the material. The stopping cross-section (ε) is deﬁned as the ratio

between the rate of energy loss and the areal density of target atoms within the distance

((1/N)(dE/dX) −N is the areal density of the atoms within the distance dx). Since the stopping

cross-sections have been already established for many materials as a function of energy, the

thickness of the materials can be determined.

Although energy spread (energy straggling) as a function of depth is important for bulk

materials, for thin ﬁlm measurements this can usually be neglected. As such, energy straggling

is not discussed in detail in this chapter.

As discussed in the above sections, the analysis of ion scattering data is straightforward

because the kinetics of energetic ion–atom collisions and scattering can be accurately

described by simple classical mechanics concepts. During the collision, the ions lose some of

their energies to target atoms. The backscattered ions have energies equivalent to the product

of the incident energy and the kinematic factor while the target atoms recoil to the forward

direction owing to the energy gained. As such, the ﬁnal energies of the scattered ions and

forward recoiled atoms can be determined in a particular direction using the mass of the

projectile ion and target atoms. LEIS, MEIS, and HEIS capabilities are brieﬂy described in the

following sections.

16.4.1.1 Low-Energy Ion Scattering

Incident ions with energies ranging from 1 to 10 keV are used in LEIS investigations. Since the

ions have low energies, the ions penetrating beyond the ﬁrst one or two layers are neutralized

and, as a result, not detected. As such, this capability is very surface sensitive. Some of the

ions which interact with the surface layer are also neutralized. In addition to the scattered ions,

the neutrals can be used to obtain information from surfaces. LEIS can be efﬁciently used to

characterize surface composition and surface structure [78–84]. Ion neutralization can be used

to probe the spatial distribution of surface electrons. Since the number of scattered ions is very

small compared to the incident ions (< 5%) neutralization was initially perceived as a feature

that complicates ion scattering analysis. Most of the conventional ion scattering systems were

not equipped with instrumentation able to capture the neutrals and it was difﬁcult to separate

neutralization effects. To overcome the issues associated with neutralization, Niehus and

co-workers [85, 86] developed the use of alkali ions instead of conventionally used inert gas

ions. Alkali ions have low neutralization probabilities and, as a result, the scattered intensity of

the ion is signiﬁcantly higher than that of inert gas ions. However, surface contamination by

alkali ions can be an issue. During the late 1980s and early 1990s time-of-ﬂight (TOF)

methods including mass spectroscopy of recoiled ions (MSRI) and direct recoiling

spectroscopy (DRS) to analyze both scattered ions and recoiled neutrals were developed and