Coating Materials News Volume 11 Issue 4
December, 2001
The mission of Coating Material News is to acquaint our readers with advances in materials processes, properties and applications. To this end, we offer three reviews of recent pertinent research in this issue that directly present these advances or illustrate them by example. It is our objective that CMN lend some assistance to new readers in solving materials deposition problems or inspire new approaches to solutions for veteran readers. We always welcome comments and suggestions for future topics of current general interest.
Ion-Beam Assisted DepositionIBAD or ion plating processes are based on bombarding a growing film with energetic particles for the purpose of modifying its mechanical, electrical, or optical properties. Many variations of IBAD have been implemented, but the basic physics is similar in most cases. Reactive or non-reactive ions can be supplied by a gun-type source that encloses a discharge plasma from which energetic particles are extracted and accelerated toward the substrate. Alternative implementations generate a plasma using microwave or RF excitation of gases. Growth-related effects in film layers that are caused by ion bombardment include increased surface mobility of arriving adatoms, sputtering of loosely bound species, and interfacial mixing. These effects discourage growth nuclei and suppress columnar growth, thus resulting in densification of the microstructure. In a modified structure zone model of microstructure, the high temperature parameter required for high packing density is replaced by bombarding ion energy.
Two types of IBAD are distinguished by the nature of their reactive species. In the production of oxides and nitrides for example, the plasma contains activated oxygen or nitrogen ions or atoms that chemically react to form the desired compound and simultaneously, by momentum transfer, densify the film. In non-reactive IBAD, the important parameter is the energy delivered to the growing film per deposited adatom, which consists of the ion energy and the ratio of ion flux density to adatom flux density. Within limits, these components are interchangeable: similar results can be obtained for energies of a few hundred eVs and flux ratio 1:10 ions:condensing atoms or <10 ev="" and="" 10:1="" ratio.="" the="" applicability="" of="" this="" reciprocal="" relationship="" is="" material="" dependent="" especially="" for="" the="" case="" of="" reactive="" ibad.="" further,="" low="" energies="" are="" useful="" to="" increase="" surface="" mobility,="" while="" high="" energies="" promote="" densely="" packed="" fine-grained="" structures.="" it="" has="" been="" observed="" [1]="" that="" high="" energies="" lead="" to="" high="" values="" of="" index,="" density,="" hardness,="" and="" compressive="" stress.="" low="" stress="" films="" that="" are="" also="" hard,="" dense,="" and="" chemically="" stable="" are="" deposited="" at="" low="" energies="" (10="" –="" 30="" ev)="" and="" high="" ion:atom="" flux="" ratio,="" ~1="" –="">
It is difficult, if not impossible, with typical ion sources to gain independent control of ion energy and ion:atom flux ratio. However, a dual-mode plasma source consisting of a microwave source and a RF generator provides that control in a plasma enhanced chemical vapor deposition process (PECVD). The MW source influences the ion:atom flux ratio while the RF-induced negative dc bias at the substrate establishes the energy. Other advantages of this PECVD process are the ability to form most oxide and nitride compounds including carbon films and graded-index layers for optical and mechanical purposes at substrate temperatures below 250° C. This is substantially below CVD and sol-gel processes, and PECVD is likely to gain in popularity and importance for future thin film applications [1]. In the article below on Sputter Deposition, we see how independent control of important variables can be achieved in sputter deposition processes.
Excimer Damage Thresholds of MgF2 Films
The operational lifetimes of optical systems employing high power excimer lasers are limited by the laser-induced damage threshold (LIDT) of the coated components in the system. LIDT is the fluence at which irreversible damage is created. The damage might take the form of ablative cratering, cracking, or pit formation. It is important, therefore to explore the influences of deposition process, substrate properties, etc on the LIDT. In fact, many thin film parameters affect the LIDT, among them are the refractive index and absorption of the film, its stress, adhesion, density, thermal characteristics, and film microstructure, the latter which includes defect density and size distribution.
There are few low-index materials suitable for pairing with high-index layers to produce laser mirrors and AR coatings for the UV (KrF at 248 nm wavelength) excimer region. Candidates with proven acceptability include SiO2, MgF2, CaF2, LaF3, and AlF3. In a recent study, the LIDTs of MgF2 films deposited by e-beam and resistance-heated evaporation and by ion-beam sputtering were determined [2]. Fused silica and calcium fluoride substrates that are transparent to 248 nm energy were coated with a 200 nm thickness by the three techniques. Analysis showed that the e-beamed MgF2 on calcium fluoride had the highest LIDT: 9 J/cm2. The resistance-heated film had LIDTs of 6.5 – 8, and the sputtered films were 2. The e-beamed film on silica had a lower LIDT because the thermal expansion coefficient difference was large, while the coefficients. are nearly equal between CaF2 and MgF2. Ablation and cracking were, respectively, the failure mechanisms for the calcium fluoride and silica substrates. The ion-beam sputtered film showed greater mechanical strength in that no cracking was produced. However, localized highly absorbing defects became sites of damage pits. It is also known that high energy processes involving ion beams cause fluorine reduction and consequent absorption in the UV region. We can conclude that the moderate energy provided by e-beam deposition produces superior MgF2 LIDT films for excimer lasers.
Sputter Deposition: Influence of Process Parameters on Stress and Microstructure
In addition to the wide utility of the sputtering for the deposition of metal and dielectric thin films, much has been learned about the fundamental nature of the growth of thin solid films through the unique control of parameters that are critical to that growth. The sputter deposition process has been discussed in past CMN articles [CMN V10 Issue 3 Sept, 2000; CMN V9 Issue 2 June 1999; CMN V6 Issue 3 July-Sept 1996; CMN V2 Issue 3 July-Sept 1992; CMN V1 Issue 4 Oct.- Dec. 1991].
The availability of additional process parameters in sputter deposition allows the control of residual film stress and microstructure as well as mechanical properties such as hardness, adhesion, and strength. In the large scheme of things, these characteristics are not independent; however, control of parameters that influence one property more than another permits some degree of understanding of those specific influences. The film properties are also affected by the presence of impurities and structural defects.
It has been established that film stress can be altered from tensile, created at low working gas pressure, to compressive at higher pressures. Similarly, stress can be increased from low to higher compression with increasing bias voltage. Simultaneous with a transition from tensile to compressive stress, the microstructure changes from low packing density columnar to a denser structure. Thermal coefficient differences between film layer and substrate also cause tensile stress when the film is grown on a heated substrate of lower TCE. Other mechanisms less well understood contribute to the development of tensile stress. These include grain boundary formation and others. Compressive stresses are often attributed to the incorporation of gas and the generation of defects in the lattice.
The example reviewed is based on magnetron sputtering of Ti onto Si wafers, where the working gas pressure and applied substrate bias voltage, the two process parameters known to affect residual stress and microstructure, are varied [3]. These process parameters essentially are the energy and species bombarding the growing film layer. The gas pressure was held constant at 3 mTorr while the substrate bias voltage was varied between 0 and –300 V. Alternatively, the bias voltage was held constant at 0 V (ground) while the pressure was varied between 1 and 8 mTorr. Without an applied bias but at varied Ar pressure, the bombarding species are Ti atoms. With an applied bias, the species are Ar+ ions. The authors demonstrated that a dramatic transition from tensile to compressive stress occurs with decreasing pressure and also with increasing negative bias voltage, due to increasing bombardment energy. The transition in both cases occurs near an energy of 15 eV per deposited Ti atom. For the 0 V case, the transition point is 2 mTorr; for the 3 mTorr case, it is -75 V.
Microstructural changes accompanied the stress transitions: tensile films had a low packing density columnar structure, while compressive films had a higher packing density. As bias voltage was increased, the columnar structure became more equiaxied and thereby denser. The higher energies offset the low adatom mobility that leads to columnar microstructure through self-shadowing. Compressive stress can be due to the presence of impurities or growth defects. In this case where the gas density is low, defects are apparently created by the bombarding Ti atoms. Since it was found that the critical bombarding energy at the stress transitions was the same, it was concluded that the same mechanism is responsible for the generation of compressive stresses, namely the knock-on-induced lattice damage. The reader is encouraged to obtain and study this excellent paper [3] for more details.
SummaryA great deal can be learned about the properties of films and mechanisms that produce them, as well as methods for controlling such properties as stress and microstructure, simply by studying the work of dedicated researchers.
References
- Ludvik Martinu and Daniel Poitras, J. Vac. Sci. Technol. A 18(6), Nov/Dec 2000.
- Protopapa, De Tomasi, Perrone, Piegari, Masetti, Ristau, Quesnel, and Duparre’, J. Vac. Sci. Technol. A 19(2), Mar/Apr 2001.
- H. Ljungcrantz, L. Hutman, J.-E Sundgren, S. Johansson, N. Kristensen, J. –Å Schweitz, and C. J. Shute, J. Vac. Sci. Technol. A11(3) 543 (1993).
*Newsletters beginning with Vol. 6 are available from the CMN archives page of this web site. Contact Materion Advanced Chemicals for printed copies of referenced newsletters prior to Vol. 6.
If you have a question or a topic you would like us to consider for a future issue of CMN, contact Materion Advanced Chemicals or fax requests to 414-289-9804. We also encourage contributions from other writers. Contact the Materion Advanced Chemicals marketing department via e-mail for more details on submitting an article.
(S.F. Pellicori is available for private consulting on matters concerning optical thin films. Please contact him directly for more information)
Editor:
Russ DeLong
Materion Advanced Chemicals
P.O. Box 1178 | Milwaukee, WI 53201
Phone: 414-289-9800 | FAX: 414-289-9805
e-mail: AdvancedChemicals@Materion.com
Principal Contributor:
Samuel Pellicori
Pellicori Optical Consulting
P.O. Box 60723 | Santa Barbara, CA 93160
Phone/FAX: 805-682-1922
e-mail: pellopt@silcom.com