Home

About Us

Contact Us

Location

Staff info

 

 

 

Request a part

Coatings information & resources

Diamond turned optics

Integrated Optics chips

Success Stories

Tools and capabilities

About pricing

 

 

 

The ANFF is an open access network comprising of eight Nodes across 21 institutions with a portfolio of more than 500 tools valued at over $200 million.

 

See:

http://www.anff.org.au/

 

for full details

 

 

Coatings Information

 This page holds general information on coating design and deposition technology, and specific information related to the tools, processes and materials in use at Optofab ACT where standard processes are established.

 

Coating design software, tips, and materials data

Available materials in Spector IBS

Coating properties:

Thickness dependent film properties

Coating stress

Absorption

Point defects

Uniformity

Repeatability

Laser Induced Damage

      Coating Methods

 

Coming later:

Coating Adhesion

Particulates and cleaning

OH contamination

Interfacial absorption

More

 

 

 

 

 

Coating Design Methods and Software

Optofab ACT has a suite of design and modelling software for interference coatings and can perform design to specifications as part of our service. Some users may however wish to investigate coating design on their own and supply a suggested recipe for their desired device. Here we provide some links to appropriate design resources and materials properties for our calibrated processes. However please note that the final refractive indices etc are not those present inside the coating tool due to annealing and stress relaxation and we will always have to modify the design for deposition monitoring purposes so we can program the layers appropriately to account for these changes and also to enable optical monitoring. All supplied designs will also have to be run through our software suite to determine the tolerance budget and optimised for this if necessary.

 

Coating design is well studied and there are a number of good books on the topic, for example see:

 

http://www.sspectra.com/literature.html

 

The book Design of Optical Interference Coatings by Alfred Thelen was kindly made available free of charge by the author for some time, you may be able to find a freely downloadable copy of it. This book and others provide a conceptual framework for interference coating designs of pretty much all the usual kinds. Having then a framework, software is needed to tailor the starting design to the required specifications. There is free code available to do this from the Polytechnique de Montreal called OpenFilters, see:

 

https://www.polymtl.ca/larfis/en/links

 

This is quite a capable program, though it lacks a little integration and cannot optimise designs with tolerancing in mind as some of the commercial products can. Python source code is also available under the GNU public license arrangements so you can modify it to suit if suitably inclined and motivated. We run it at Optofab ACT as it can be available everywhere, and also to have a common base for design exchange with our customers. We have a range of custom written Labview code for converting materials files, creating design templates, etc for Openfilters that will be made available for Download shortly. Beyond this, there are many commercial options. An incomplete list of some of the more commonly used programs is presented below:

 

TFCALC free version available with limited materials

Optilayer

Film Wizard

Essential Mccleod

Filmstar free version with fixed refractive indices

Freesnell free

 

We use TFCALC routinely, and will invest in Optilayer once we have the broadband monitoring systems running as Optilayer has a module that enables on line correction/compensation to still meet spec if a layer is incorrectly deposited or terminated, and it supports both broadband and ellipsometric monitoring.

 

Before leaving the topic it is also worth mentioning that with any of these optimising software models, how you set the targets is critical to the convergence to a practically useful result. On the one hand you want to use the minimum number of target points to maximise iteration speed, but on the other oscillations in the response mean that being too sparse or having the target points at the wrong wavelength will generate designs that are not what is desired with peaks/troughs not caught by the target points. Therefore measures need to be taken to prevent this (for example stop the optimisation and look at the response and adjust the target wavelengths to the peak/trough positions).

 

Another issue is just putting in very simple targets. For example, there may be a need for a mirror with very high 1064 nm reflectivity and a very tightly controlled 532nm partial transmission. It is tempting to put in targets at just 1064 and 532nm and run the optimiser. This can produce results with steep slopes at the target wavelengths that are then very fabrication sensitive. Better to put in three points at each target wavelength, one at the wavelength and one either side with carefully chosen values to try to force the software to make a peak or trough there which will be more fabrication tolerant. Similar principles apply to bandpass or edgepass filters where extra points should be used to control the roll off shape to enhance fabrication tolerances.

 

Our standard IBS coatings use Silica and Tantala as the low and high refractive index coatings respectively. We run two processes, one at 150C deposition temperature with a post deposition anneal to stabilise the coating and reduce absorption, and one at room temperature with the assist gun to reduce stress and avoid high temperature anneals which the target device (e.g. a deformable mirror) may not be able to tolerate. Representative post anneal or post deposition dispersion curves for the materials are provided below (shortly) for the standard single rotation case, but we note that the precise behaviour of index and stress and layer thickness changes on annealing are subject to the exact coating design due to stress distributions through the coating and the change in sign of the Tantala stress during anneal. For dual position single rotation depositions, contact us as this case is more complex.

 

150C process post anneal tantala dispersion

text file

Openfilters .mat

TFCALC .mat

150C process post anneal silica dispersion

text file

Openfilters .mat

TFCALC .mat

25C process post deposition tantala dispersion

text file

Openfilters .mat

TFCALC .mat

25C process post deposition silica dispersion

text file

Openfilters .mat

TFCALC .mat

 

 

Return to top

 

 

 

 

 

Available materials and typical coating rates - IBS

As the IBS system uses 14 water cooled targets that can cost up to USD15k, then we only hold a limited number of them. The currently available materials are:

 

Silica

RI ~1.46 @ 1064nm

Tantalum (reactive mode for Ta2O5)

RI ~2.08 @ 1064nm

Titanium (reactive mode for TiO2)

RI ~2.48 @ 1064nm

Aluminium Oxide

RI ~1.67 @ 1064nm

Niobium (reactive mode for Nb2O5)

RI ~2.25 @ 1064nm

Hafnium (reactive mode for HfO2)

RI ~2.08 @ 1064nm

Indium Tin Oxide

RI - its complicated! TBD for conditions

 

Currently Only Silica and Tantala are properly characterised. We are happy to attempt others but are not yet able to guarantee performance with them.

 

The metal targets can also be sputtered in the presence of nitrogen to form the corresponding nitrides, but we have not yet developed processes for this. As noted under the tool capabilities, the IBS system is restricted purely to depositing metal oxides and nitrides and whilst fluorides are possible in the presence of NF3, we are not going to contaminate the chamber by doing this. Additional targets can be purchased as required for coating jobs, though if these are very expensive and it looks as if the material in question is very specialised, additional charges may apply. The targets in the tool as standard will be Silica, Tantala, and Titania, and in the long run target change costs will be applied as its about an hours work to change a target. Deposition profiles and rates have so far only been characterised in depth for Silica and Tantala. The image below shows a typical deposition rate distribution measured at the deposition plane for Tantala, and for the standard single rotation cases the deposition rates are ~0.1 nm/sec for Silica and Tantala at 100C substrate temperature.

 

 

The deposition rates are very stable over time, and thick coatings can be deposited though obviously this takes quite some time and will cost rather more. Coatings of 10-20 microns thickness are entirely viable (subject to stress constraints for 150C coatings), just slow. For the standard single rotation deposition, the coating can be quite conformal, the images below showing coatings over steps of sizes as indicated.

 

IBS400nm As2S3-1500nm_TOX_003 (002)vert-1_002

 

 

Return to top

 

 

 

Uniformity

Two types of depositions are possible:

·        single rotation depositions where a plate holding the substrates is located in a fixed position (which can easily be changed) and rotated about its centre

·        Dual rotation where the rotating substrate plate is sequentially positioned at two different locations for each layer to put down two complementary profile parts of each layer that sum up to a high uniformity layer over a large area.

 

For single rotation films, the uniformity pattern depends on the position of the substrate holder, and results from modelling are shown below to indicate the range of possibilities which allow for optimising maximum area, PCD for a given substrate size, or flatness. It is also clear from these plots how two complementary positions can be found for the dual rotation case to absolutely maximise uniformity over large areas.

 

 

Typical measured thickness uniformity results for each type of run are shown below and reflect that the modelled data represents reality reasonably well:

 

Single rotation uniformity plot at different positions showing maximum single substrate flatness or largest PCD for many 1 substrate deposition

 

Dual rotation showing large diameter high uniformity

 

Single rotation is much simpler, faster, and lower cost and sufficient for small substrates where ~0.1% uniformity can theoretically be achieved on many substrates in one run up to 50mm diameter. For substrates beyond 100mm diameter (only 1% uniformity possible for runs of up to 6 substrates in single rotation or better with fewer substrates on smaller pitch circle diameter, see above uniformity data), it may be necessary to go to dual rotation which is rather more complex and expensive but that can offer sub 0.1% uniformity under optimal conditions on 400mm diameter substrates.

 

 

Return to top

 

 

 

 

Repeatability

The typical run to run repeatability in thickness uniformity characterised by the standard deviation on the total number of runs for Silica and Tantala are respectively 3.5 % and 3 %.

It is important to note that deposition rate and uniformity vary with the plume shape and hence strongly depend on the grid set. As our initial grid sets are of unknown heritage we are currently sourcing several new grid sets of nominally the same specification to improve the grid set dependence. More details when we have them

 

Uniformity shape depends on grid set and wear

 

Return to top

 

 

 

Absorption

Absorption losses in annealed coatings are very low, typically <10ppm in coatings up to 5 microns thick. This means extremely high reflectivities and many pass mirror cells are very possible with IBS coatings. We are currently analysing the absorption loss in detail by depositing Silica or Tantala onto high Q factor disk resonators and looking at the Q as a function of annealing temperature to characterise this in great detail. Initial results for all loss sources combined (absorption, surface scatter, volume scatter) have the Tantala at 0.1 dB/cm at 1550nm, and the Silica at 0.05 dB/cm. For silica this equates to 1.1 ppm total loss per micron of coating thickness at 1550 nm. This study will be ongoing and will look at other wavelengths over time, check back later for more data

 

Return to top

 

 

 

Coating Stress

IBS coatings are deposited with considerable kinetic energy as the typical ion energies before neutralisation and target bombardment are in the 1000eV range. It is this energetic nature that gives them their densely packed environmentally insensitive nature. However, this also produces considerable compressive stress in the as deposited coatings. We have quantified the coating stresses by measuring bow on silicon wafers.

 

Assuming the thicknesses of the substrate and the coating are smaller than the lateral dimensions, thickness of the coating is smaller than the thickness of the substrate, the substrate and coating are homogenous, the radius of curvature is equal in all directions (spherical deformation), the thin film residual stress (s) can be deduced from the radius of curvature (R) measured before and after deposition of the film thanks to the Stoney formula (Ardigo M., Ahmed M., Besnard A. (2014). Stoney Formula: Investigation of Curvature Measurements by Optical Profilometer. Advanced Materials Research. 996. 361-366).

 

 

Es is the Youngs modulus of the substrate

n is the Poissons ratio of the substrate

ts is the substrate thickness, tf is the film thickness

R1 is the radius of curvature before deposition (substrate)

R2 is the radius of curvature after deposition (substrate + film)

The measurement of the radius of curvature is realized with a Dektak 504 profilometer that gives a direct access to the sagitta (Z) of the measured sample from which R can be deduced.

                

Where L is the total scan length. The substrates used are 4 wafers Si [100], 525 m thick, p type, the scan length is 55mm. The bow measurements are done on the bare Si substrate first, then on the Si substrate with the deposited film and finally on the annealed Si substrate with the deposited film.

 

The standard IBS Ta2O5 films deposited in the Spector have a typical compressive -195MPa residual stress as deposited that becomes tensile +52 MPa after standard annealing. The standard IBS SiO2 films deposited in the Spector have a typical compressive -551 MPa residual stress as deposited that stays compressive at -184 MPa after standard annealing.

 

We are also currently starting to investigate low stress coatings. This can be achieved by low temperature deposition and by using the assist ion gun running on oxygen. Data will be presented when studies are complete.

 

 Return to top

 

 

 

 

Thickness Dependent Film Properties

 

Index of Refraction Variation

The refractive index of an optical coating depends on a combination of many parameters. Besides stoichiometry, the deposition method as well as thermal history affects the optical properties of the final product. For ion beam sputtered coatings, the variation of beam current, particle energy, direction of travel, substrate temperature, process pressure, background atmospheric composition, and interaction with intra-vacuum objects can all lead to inhomogeneities of the coatings. When ultimate precision is required over large apertures, even small variations can compromise reflective or antireflective coating properties.

 

There is also the question of homogeneity within a layer. With techniques such as e-beam evaporation it is well known that the initial part of a film (first 20nm or so) has different properties to the later part. We have not yet characterised this for the Spector IBS, but data exists to suggest similar effects apply certainly to Tantala layers, for example the data below from CSIRO for Tantala for a film as it grows plotting the effective index vs film thickness.

 

Figure 1b

 

There is also data from optical quantum wells using very thin layers (5nm) that they can be bulk like so there is an open question here as to what influences these properties. For very high performance coatings these thickness dependent properties have to be considered in the design and annealing of the coating.

 

Our IBD system sputters most oxides reactively, that is, it uses a pure semiconductor or metallic material target, and a gas feed maintains a low-pressure oxygen background atmosphere in which oxidation occurs in-flight in the sputter plume. The properties of the film that forms on the substrates take a brief period to stabilise after the start of a new layer, and the growth rates need to be balanced against the oxidation rate, to guarantee full oxidisation in the film.

An exhaustive predictive model for the index of refraction variation in the deposition plane is virtually impossible to maintain. Instead, we periodically map the as-deposited and post-annealing optical properties of our coatings in the deposition plane, which allows us to minimise this effect. We have found that our refractive indices tend to vary on the level of a few parts per thousand. Taking this effect into account in addition to the expected physical thickness allows us to produce variations in coating reflectance and transmittance as low as a few parts in ten thousand over apertures as large as 34 centimetres.

 

Return to top

 

 

 

Point Defects

Point defects in optical coatings can originate from a long list of sources, such as pre-deposition contamination or defects of the substrate surface, embedding of particulate released in the coating chamber during the deposition process, coating damage from handling or thermal shock, onsets of crystallisation during annealing, and more. These defects will in most cases become sources of excess optical scatter or absorption, both of which are detrimental to precision optical and particularly interferometric experiments.

 

Point scatterers increase the local value of the bidirectional reflectance distribution function (BRDF) which describes how much light is scattered per unit solid angle in which direction. This means they represent a source of optical loss to propagating light rays, which can be particularly harmful for high finesse optical cavities, but also provide a channel for ambient light to scatter into the primary measurement modes, which causes wavefront aberrations and phase and amplitude noise. An ensemble of point scatterers can coherently combine to scatter optical power in beamed directions. At this stage we have not measured the BRDF for either defect free films (to estimate the surface/volume scatter) or with point defects. Data will be posted when available.

 

Point absorbers locally decrease the damage threshold of optical coatings, which causes problems for pulsed laser applications that either have large individual pulse energies or high burst energy content. This can cause the optical coatings to rapidly degrade, which accelerates once initial damage has occurred. Even when operated below the damage threshold, point absorbers may locally deform the optic and produce temperature gradients in the substrate material. The result is increased wavefront aberration of both reflected and transmitted fields due to thermal deformation and thermal lensing.

 

We operate our coating equipment in state-of-the-art cleanrooms with particle level monitors and rigorous environmental controls to suppress particulate generation and influx. We have carefully audited cleaning protocols for all components that enter the vacuum chamber and operate our deposition processes with a minimum count of mechanical components in the coating chambers that can be sources of particulate contamination.

 

Return to top

 

 

 

 

Laser Induced Damage

Optical coatings are today usually the limiting factor in the damage threshold of high power laser components. Damage occurs through three distinct sets of mechanisms depending on the exposure duration:

 

Pulse Duration

< 1ns

1ns-100ns

100ns-100ms

>100ms

Damage Mechanism

Nonlinear ionisation

Dielectric Breakdown

Thermal and dielectric breakdown

Thermal

Relevant damage spec

Ultrafast specific

Pulsed

Pulsed and CW

CW

 

There are also wavelength effects (most materials exhibiting enhanced absorption at DUV and MIR wavelengths) and Beam profile effects through the peak intensity present in the beam at a given power level e.g. Gaussian has twice the intensity of flat top of the same power.

 

Thermal effects are driven by absorption in the coating materials themselves or the substrate, or from defects in the coating and lead to heating of the coating and eventual melting, stress induced coating pop off, or recrystalisation. For a Gaussian beam heating a coating by surface absorption, the temperature rise is given by:

 

 

Where bs is the surface absorption, P is the power, k is the thermal conductivity, k is the thermal diffusivity, tI is the irradiation time, w is the 1/e2 beam diameter. When the irradiation time is long compared to the thermal diffusivity time constant (w2/k) the above equation has an asymptotic solution given by:

 

Note that counterintuitively this is characterised by P/w ie W/cm and not intensity (W/cm2). Thus the CW and long pulse damage threshold has linear scaling with power and beam diameter.

 

For shorter pulses the full equation has to be solved and there is a cross over zone where thermal or breakdown damage or a combination of both occurs. In this zone it is common practice to express the LIDT in fluence (J/cm2), which again is counterintuitive as there is then no time dependence in the specification which certainly exists in practice. A better spec is in intensity (W/cm2).

 

However, as the ultrashort pulse regime is entered, a range of nonlinear effects come into play and a totally separate damage specification is required for this. In the best possible case, damage here is defined by the ablation threshold of the materials, e.g. about 2 J/cm2 for silica at 1030nm with pulses of a few hundred femtoseconds. There are a wide range of things that can lower this such as electric field enhancement around defects in the coating or on the substrate surface, doping of the materials, photochemically deposited atmospheric materials, photodarkening, etc.

 

As noted, the wavelength matters, and in the absence of detailed data, the damage threshold is usually assumed to scale inversely with wavelength as long as no absorption bands are encountered. As a single pulse duration is normally only given for the pulsed damage threshold, then as long as the pulse you are interested in is not in the ultrafast regime, the scaling based on the square root of the ratio of the pulse durations is usually applied for pulse lengths below 100ns. Above 100ns, the thermal effects come into play and scaling is more difficult to predict but some idea may be obtained from comparing the predicted LIDT for 100ns with that from the CW regime.

 

IBS coatings are restricted in our system to metal oxides and nitrides on the whole and in the 400nm to potentially 4m region have very low linear absorption if correctly deposited and thermally annealed. To give some concrete numbers, as noted previously early measurements for all scattering and absorption losses summed indicate 0.05 dB/cm loss for silica at 1550nm (1.1 ppm per micron), and 0.1 dB/cm for tantala (2.2 ppm per micron) as measured by waveguide propagation losses by coating very high Q disk resonators with ~2 micron thick films. As the light is now propagating through the film rather than perpendicular to it, then as well as absorption, this figure also includes a significant surface scattering loss that would not be encountered in a standard transmissive or reflective thin film filter stack. We have not yet measured LIDT on these coatings, but expect them to be high. Data will be provided once available.

 

DUV and MIR are usually the domain of IAD E-beam coating. We do not yet have this system running and so have no figures for these regions.

 

Return to top

 

 

 

 

 

A coating method for every application

Optical coatings can be deposited by many methods, each having its own characteristic strengths and weaknesses. Here we provide some generic but detailed information on the usual suspects

E-beam evaporation

Ion assisted E-beam evaporation

Reactive Magnetron sputtering

Ion Beam sputtering

Chemical Vapour Deposition (PECVD, APCVD)

Fast Atomic Layer Deposition (ALD)

Comparison table

 

Return to top

 

 

 

 

Coating Methodologies

  

E-Beam evaporation

E-Beam or Electron beam evaporation is a type of Physical Vapour Deposition process where target material is bombarded with an electron beam through charged tungsten filament under a high vacuum process chamber. The electron beam causes the atoms from target material into gaseous state and then precipitating into the solid form coating onto the substrate inside the vacuum chamber within the line of sight. The electron beam is generated by an electron gun, which produces thermionic emission from the tungsten filament and emitted electrons are accelerated by high voltage potential (kilovolts).

 

The advantage of E-beam coating is the deposition rate of this process ranges from very low (1 nm per minute) to very high (few micrometres per minute) making the material utilization efficiency very high and it also offers structural and morphological control of the film with very high thermal efficiency, high productivity and low contamination. The deposition rate can be measured using in situ by quartz crystal and deposition rate depends on the starting materials and E-Beam power.

The electricity density of the electron beam is very large and deposition of various materials is possible like high melting points, oxide materials and materials which sublime. It is possible to have multiplayer film coating using several sources and crucibles in the process chambers.

Fig.1:E-beam operation principle

 

Return to top

 

 

 

 

Ion Assisted E-Beam evaporation

During the Ion assist E-beam operation, an ion beam typically from argon gas with broad range of energy is targeted towards the substrate and it arrives alongside the evaporant materials to be deposited. The directed Ions impart the energy into atoms of the evaporant materials and increases the surface mobility. The surface mobility of the atoms helped the materials to improve the adhesion, density and structure of the thin film. Because of these improved film quality, this technique provides more repeatable refractive index of the material.

 

Ion assist E-beam evaporation can be achieved by typically using End-hall type of Ion source which provides uniform substrate coverage and control over the current and energy of the Ion beam.

 

Fig.2:Ion Assisted E-beam operation principle

 

Return to top

 

 

 

 

 

Reactive Magnetron Sputtering

Magnetron Sputtering process is another type of Physical Vapour Deposition (PVD) technique to grow thin films on to the substrates because large quantity of films can be produced with relatively high quality and low cost. The magnetron sputtering process involves a gaseous plasma which is generated by introducing inert high molecular weight sputtering gas like Argon or xenon and confined to a source where material target is located. The high energy ions of the plasma erodes the surface of the target and ejected atoms travel through the Vacuum environment and deposited on to the substrates to form thin films.

 

The plasma is initiated by applying a high voltage between the cathode (located behind the target) and anode (commonly connected to the chamber as ground).Electrons from the sputtering gas are accelerated away from the cathode and collide with the nearby atoms from sputtering gas resulting an electrostatic repulsion which knocks off the electrons from the sputtering gas atoms causing ionisation.

 

The positively charged atoms now accelerate towards the negatively charged cathode causing high energy collisions with the target materials. Each of this collision eject atoms from the target materials with good enough kinetic energy to reach to the surface of the substrate start to condense to form a thin film. As more and more such atoms combine together on the substrate, they start binding each other at molecular level forming tightly bound atomic layer. A precise layer of thin film is created by producing one or more such atomic layers depending on the sputtering time.

 

There are different type of magnetron sputtering such as DC magnetron sputtering and RF magnetron sputtering each having different working principles and objectives.

The advantage of RF magnetron sputtering over the DC magnetron sputtering is, it does not require the target as an electrode to be electrically conductive hence any material can be deposited theoretically using RF magnetron sputtering.

 

In case of standard sputtering, a target of whatever pure material is desired and an inert gas usually an argon is used. However in the reactive sputtering, non-inert gases such as oxygen or nitrogen is used either in place of or in addition of inert gas(Argon). In this case the ionised non inert gas chemically react with the vapour of target materials and form a compound molecular layer which is deposited as a thin film on to the substrate.For example a silicon target with non inert gas oxygen can form Silicon di oxide layer or with N2 can form Silicon nitride layer.

 

In many cases it is also possible that the reactive gas ions chemically react with surface of the target materials and does not sputter anything from the target.This state of the target is called poisoning state.To get a good film stoichiometry with high deposition rate requires the target material state to be in fixed state between pure metallic and pure oxide materials.

Fig.3:DC sputtering system Fig.4: Magnetron sputtering system

 

Return to top

 

 

 

 

Ion Beam Sputtering

Ion Beam Sputtering is one of the PVD methods which provides very fine and good quality thin film coating. During the Ion beam deposition, Ions from the Ion source is focused on the Target materials and the sputtered material from the target is deposited on to the substrate as a thin film. The system configuration may include another guided ion source, which focuses the ion beam directly to the substrate assisting highly dense film deposition. Although in the typical IBS system, the main feature of the system is Ion source, a target and a substrate. The system configuration with extra guided ion source is known as Ion Assist Deposition.

Fig.5:Ion Beam Sputtering(Ion assist)

 

Ion Beam deposition method is one of the slowest and most expensive deposition method but it produces very high quality film with good precision.

 

Return to top

 

 

 

 

Chemical Vapour Deposition (PECVD, APCVD)

CVD is a versatile deposition technique that provides the process of growing thin films of elemental and compound semiconductor materials, amorphous or crystalline compound and metal alloys of different stoichiometry. It consists of chemically reactive volatile compound of the materials to be deposited with other gases to produce non-volatile solid on a suitable substrate as an atomic layer.CVD process is a very well established technique and the reactor or process chamber used for the process mainly depends on the precursor, condition of the deposition, kind of energy used to the system to activate the chemical reaction. The following processes are the most established CVD processes in the industry:

 

-      PECVD-Plasma Enhanced Chemical Vapour Deposition(When plasma is used to introduce chemical reactions i.e. Microwave or RF power based plasma)

-      APCVD-Atmospheric Pressured Chemical Vapour Deposition(It is a CVD process which uses atmospheric pressure to provoke chemical reactions)

-      LPCVD-Low Pressure Chemical Vapour Deposition(CVD process at low gas pressure of 0.6 to 1.33mbar)

-      MOCVD-Metal Oxide Chemical Vapour Deposition(In this process of CVD the metal oxides are used as precursors for example trimethylaluminium, trimethylgallium)

 

There are two type of typical CVD reactors, hot wall CVD and cold wall CVD. In the hot wall CVD, the heating is achieved by surrounding the resistive element around the reactor and in the cold wall CVD, the substrate holder is heated up while the chamber walls are air cooled or water cooled.

 

 

PECVD(Plasma Enhanced Chemical Vapour Deposition)

PECVD is a very common and well established CVD process which uses much lower temperature than the typical CVD process to deposit various materials from gas state to a solid state thin film on to the substrate. In this process deposition is achieved by introducing reactive gas between two electrodes one is grounded and another one is RF powered (or AC frequency or DC discharge). The plasma is created by the capacitive coupling of two electrodes which induces a chemical reactions resulting a product being deposited on to the substrate. The substrate which is mounted on the ground electrode is typically heated at 250C to 350C depending upon the specific thin film requirement. During the PECVD process, the electrons acquired sufficient energy from the applied electric field to create reactive species without increasing the gas temperature in the process chamber. And this is the key advantage of PECVD maintaining the lower temperature process comparing to other CVD methods.

 

The energised electrons and gas molecules collide to each other in plasma and decomposes the source gases to form the reactive species such as excited neutrals and free radicals as well as ions and electrons. PECVD could be operated using RF (13.56Mhz), AC (50Hz), Microwave (2.45GHZ) or DC power supplies, with RF being the most common source of power supply in most of the PECVD system.

 

For example, to deposit silicon di oxide (SiO2) film, the plasm decomposes the silicon gas sources SiH4(Silane),Tetramethoxysilane(TMOS) or HMDS to silicon radicals and reacts with O2 radicals from oxygen sources or N2O sources inside the vacuum chamber. When the RF power is applied to generate the plasma, the energised electrons ionise the reactant gases and create more chemically reactive radicals which react to form thin film materials on top of the substrate or sample. The vacuum chamber (Fig.5) is connected to pump system which consists of roughing pump and Turbomolecular pump to pump down the chamber to very low vacuum level. The gas system is connected to the chamber through Mass flow controller. The MFC controls the flow rate of the gas inside the chamber. As mentioned above the plasma is ignited by applying electric field at RF frequency from RF sources.

 

Various kind of inorganic films are produced using the PECVD system in semiconductor or photonics industry such as SiO2, Silicon Nitride, Silicon Ox nitride etc.

Fig.6:PECVD System

 

APCVD(Atmospheric Pressure Chemical Vapour Deposition)

APCVD is a CVD method which deposits different kinds of oxide materials (doped and undoped) at atmospheric pressure. This process is highly suitable for volume production, continuous in line manufacturing, etc as it is vacuum free and can coat many substrates on a belt feed or long rolls of material. It is often used in low cost production like PV cell manufacturing.

 

APCVD process can be used for the following applications:

 

-      Production of compound semiconductors

-      AR coatings on glasses and lenses

-      Silicon-di-oxide and Transparent Conductive Oxide (TCO) coatings.

 

Some of the Transparent Conductive Oxide coatings are on flat panel displays like OLED, LCD, touch screen, solar cells etc. High temperature APCVD process is used for producing thin film in many other technologies like solid state devices, metal oxide semiconductors, graphene based devices and APCVD is one of the best to produce graphene.

 

Return to top

 

 

 

 

Fast Atomic Layer Deposition

Atomic Layer Deposition (ALD) is a type of vapour phase technique used to deposit thin film layer onto the substrate.ALD can produce defect free and angstrom level controlled films. It is a gas phase method based on sequential and self-limiting surface reaction process where each reaction allows only one monolayer of deposition. The nature of interaction between the precursor and surface determines the complete cycle of ALD. Depending on the requirement, the ALD cycle can be performed multiple times to increase the layers of thin films. The process of ALD can be performed in lower temperature very often which is beneficial for the fragile and thermally sensitive materials. With the emergence of nanotechnologies and electronics miniaturisation, the application of ALD has become very popular and vast in terms of coatings nano layers and thin films. In ALD process, gas phase reactions are prevented by injecting the precursor separately and as a result it allows self-limiting reactions on the surface of the substrate. Defects free films are deposited on the substrate comparatively at low temperature using ALD allowing the low-temperature processing as key requirement by some technology manufacturing like display coatings.

 

However throughput is very low in this case due to the long cycle duration of ALD cycle which is approximately 10sec or more at lower temperature.

 

To meet the high volume industrial production and throughput, spatial kind of ALD is developed where precursors and reactants are continuously injected into different spaces, using different distinct zones separated by the purge zones. The complete single ALD cycle obtained using this technique is 1 sec or so. This new technique called as Fast-Atomic Layer Deposition or Spatial-Atomic Layer Deposition is the combination of the conventional technique (defects free and uniform deposition over large area) and high growth rate. As a result we get 5-10 times higher growth rate in Fast-ALD than the conventional one.

Fast-Atomic Layer Deposition Atomic Layer Deposition

 

Fig.7:Atomic Layer Deposition(ALD)

 

Return to top

 

 

 

  

 

Comparison Table

CVD

PVD

PECVD

E- Beam

Advantage:

·         No line of sight deposition

·         High deposition rate

·         Mass producibility of Thick layer

·         Co deposition of multiple layers

 

Disadvantage:

·         High temperature process

·         High Toxic precursor requirement

·         Mostly inorganic materials

 

Advantage:

·         Safer process

·         Atomic level control of Chemical composition

·         No special precursor needed

 

Disadvantage:

·         Low deposition rate

·         Thin layer deposition

·         Line of sight deposition

·         Production of only thin coating layers

·         Annealing of the film is required

 

 

 

Advantage:

·         Avoids the line of sight deposition to certain extent

·         High deposition rate

·         Low temperature deposition

·         Precursor: both organic and inorganic materials

·         Chemical and thermal stability

·         No limitation on substrates type

 

Disadvantage:

·         Instability against humidity and aging

·         Film stress

·         Time consuming for some material structures

·         Toxic and explosive gas used in the plasma process

·         High cost equipment

 

Advantage:

·         Both metal and dielectric materials

·         Low impurity comparing to thermal deposition

·         High deposition rate

 

Disadvantage:

·         High cost

·         High Temperature deposition

 

 

Return to top

 

 

 

  

 

 

 

 

 

 

Quick links to our main areas

 

(Use the regular menus to access all our other pages.)

 

Make a booking

Group Training

 Powered by

News & Links

Optofab ACT newsletters

Latest news -

More details on our

 

more

more

main news page

 

Policy & Pricing

ANFFL Pricing & Access Policy

ACT Optofab Node Pricing details

Safety & Emergency information

                                       

more

more

more

 

Feedback

We run an annual User Feedback Survey but feel free to send feedback at anytime using this form.

 

 

 

 

Providing nano and microfabrication facilities to Australia's researchers

 

 

 

 

Copyright 2021 ANFF Optofab ACT Node. All rights reserved. Page last updated: 30 Jan, 2021