solutions for the fabrication of high value-added heterogeneous components and sys-tems including memories, logic, sensors, actu-ators and wireless communications. Partici-pants in the program include materialproviders, laboratories, research centers andmanufacturers of equipment, componentsand systems.

The project is structured around five themes:· Methodology and evaluation tools to inte-grate elementary components in 3D systems.· 3D technologies and integration processesonto materials and non-silicon substrates.· 3D technologies and integration processesonto silicon, using processes closely derivingfrom microelectronics.· Reliability methodologies and analysis forintegrated 3D systems.· Performance evaluation and equipment val-idation for volume production equipment andgeneric manufacturing.

SINTEF, The Fraunhofer Institute and FEI areparticipants in JEMSiP_3D, and the work pre-sented here was funded in part by the project.

3D INTEGRAT IONWhile the performance and productivity ofmicroelectronics have increased continuouslyover more than four decades due to the enor-mous advances in lithography and device tech-nology, it has now become questionable ifadvances in these areas alone will be able to

B IOGRAPHYMaaike M. V. Takloreceived her PhD in phys-ical electronics from theUniversity of Oslo in 2002for her thesis entitled‘Wafer bonding forMEMS’. From 1998 until2010 she was employed at the Departmentof Microsystems and Nanotechnologywithin SINTEF ICT in Norway where sheworked on MEMS processing and wasresponsible for their wafer level bondingactivities. Maaike is now a senior scientist atSINTEF ICT at the Department of Instrumen-tation and is the group leader for ‘AdvancedPackaging and Interconnects’.

ABSTRACT3D integration schemes connect stackedintegrated circuits using through silicon vias(TSV) and special bonding techniques. Phys-ical characterization of these TSVs andbonds is essential, but their relatively largesize (tens or hundreds of micrometers)requires prohibitively long milling times inthe conventional focused ion beam (FIB) sys-tems typically used for this work. A newplasma-based FIB system can remove mater-ial more than 20 times faster, providing thespeed and precision required to ensurerobust processes and reliable products.

KEYWORDSfocused ion beam, scanning electronmicroscopy, plasma ion source, ion beammilling, 3D integration, through silicon vias

ACKNOWLEDGEMENTSA part of this work has been performed inthe project JEMSiP_3D, which is funded bythe public authorities in France, Germany,Hungary, The Netherlands, Norway andSweden, as well as by the ENIAC JointUndertaking.

AUTHOR DETA I L SDr Maaike M. Visser Taklo, SINTEF ICT, Department of Instrumentation,PO Box 124 Blindern, N-0314 Oslo, NorwayTel: +47 2206 7300Email: MaaikeMargrete.VisserTaklo@sintef.nowww.sintef.no

Microscopy and Analysis 25(7):9-12 (AM), 2011

PFIB IN MICROELECTRONICS

I N TRODUCT ION3D integration schemes, which stack inte-grated circuits and other microelectronic orMEMS devices and interconnect them usingthrough silicon vias (TSV), are likely to be thenext revolution in electronic fabrication. Theycan be used to continue the increases in speedand density of microelectronic systemsdescribed by Moore’s Law (More Moore), butthey may offer even greater benefits whenused to connect devices of different technolo-gies (More than Moore), packing more perfor-mance and functionality into smaller volumes.

In either case, the ability to physically char-acterize the TSVs and mechanical bonds usedin 3D integration is essential for developingrobust manufacturing processes and fabricat-ing reliable products. Focused ion beam (FIB)systems have long provided physical analysisin the manufacture of integrated circuits, butconventional FIB cannot remove material fastenough to analyze these relatively large struc-tures used in 3D integration. The launch of anew plasma-based FIB system now providesthe speed and precision needed to developand deploy these exciting new technologies.

J EMS i P _ 3DThe Joint Equipment and Materials for System-in-Package and 3D Integration (JEMSiP_ 3D) isa project undertaken by a consortium of Euro-pean manufacturers to validate technological

Bonding and TSV in 3D IC Integration:Physical Analysis with a Plasma FIBMaaike M. V. Taklo,1 Armin Klumpp,2 Peter Ramm,2 Laurens Kwakman3 and German Franz 3

1. SINTEF, Oslo, Norway. 2. Fraunhofer EMFT, Munich, Germany. 3. FEI Company, Eindhoven, The Netherlands

Figure 1a: Schematic of a xenon plasma focused ion beam (PFIB) system. A PFIB uses an inductively coupled plasma to deliver high beam current. The source islarger than a liquid metal ion source (LIMS), but delivers a more collimated beam, enabling better beam spot performance at high beam currents.

MICROSCOPY AND ANALYSIS NOVEMBER 2011 9

a

overcome the predicted performance and costproblems of future IC fabrication. The ITRSroadmap predicts 3D integration as a key tech-nology to overcome this so-called ‘wiring crisis’and the solution will most likely be based onTSV technology.

The most promising 3D integration schemescurrently under consideration involve the ver-tical stacking of integrated circuits and otherdevices. These schemes vary in their details butall must solve two central problems: how tobond the integrated layers together and howto create electrical connections among them.Bonding and TSV technologies each have theirown unique set of considerations which oftencenter around how the structure will hold upduring subsequent processing, such as theaddition of another layer:·Will the stresses induced by additional ther-mal processing cause debonding or shifting ofthe existing bonds? · Will the stress and strain cause cracks ordelamination in the TSVs? · What are the best materials and processesto use to minimize these negative effects?

P LASMA FOCUSED ION BEAMFIB systems, which use an ion beam to cut andimage cross sections through subsurface struc-tures with nanoscale precision and imagingresolution, have long been a mainstay of phys-ical analysis for integrated circuits. Althoughthe structures used in 3D integration can beexpected to decrease in size as the technolo-gies evolve, they are much larger than thedimensions of the transistors and intercon-nects used in current integrated circuits, andthe cutting speed of FIBs designed for ICs isgenerally inadequate for TSVs and bondingstructures. A typical 10 µm � 10 µm IC cross-section requires the removal of 1000 µm3 ofmaterial and takes a few minutes. A 100 µm �100 µm TSV cross-section requires the removalof 1,000,000 µm3 of material and would takemost of a day with conventional FIB.

The Vion PFIB system (FEI Company, Hills-boro, Oregon, USA) uses an inductively cou-pled plasma source [1-3] (Figure 1) to providematerial removal rates 20� faster than con-ventional FIBs that use liquid metal ion sources

(LMIS). A LMIS is essentially a point source 50nm in diameter with a low angular intensity.The Vion system’s plasma source is larger, 15µm, but has a much higher angular intensity.Because of its small virtual size, the LMIS is easyto focus into a small spot at low beam currents,but at beam currents above 10 nA sphericalaberration effects severely degrade perfor-mance. The plasma source can deliver currentsin excess of a µA (>20� greater than a typicalLMIS based system) while still maintaining awell focused beam. Since material removalrates are primarily a function of beam current,the PFIB has an advantage of 20� or moreover conventional FIB at high currents, whilestill preserving excellent milling precision andimaging resolution at low beam currents.

The xenon ion beam emitted by the plasmasource has high sputtering yield, high bright-ness and low energy spread. In addition, byintroducing various gases, the PFIB can selec-tively etch specific materials or deposit pat-terned conductors and insulators (similar toconventional FIB systems). The plasma sourcealso offers the potential to use different ionspecies to enhance performance in specificapplications.

CURTA IN INGThe difference in FIB milling rates of the vari-ous materials present in a device (Cu, Si, Sn,dielectrics, polyimides and mold compounds)can cause ‘curtaining’ when milling cross-sec-tions. This milling artifact can make detailed

Figure 2: Curtaining artifacts (upper left), caused by variations in milling rate for different materials, can be effectively suppressed (right) by rocking the sampleto mill in a sequence of alternating angles (lower left).

MICROSCOPY AND ANALYSIS NOVEMBER 201110

Figure 1b: The PFIB maintains excellent spot size performance over a broad range of beam currents.

Figure 1c: At high beam currents the PFIB can remove material twenty times faster than a liquid metal ion source.

b c

PFIB IN MICROELECTRONICS

analysis of the structures difficult or evenimpossible.

Figure 2 shows typical curtaining effects onthe silicon substrate as well as the TSV itself,caused by milling through the overlying roughpoly crystalline metal film. These curtainingeffects can be effectively suppressed by rock-ing the sample during the FIB milling process.Milling in a sequence of alternating incidenceangles creates a clean cross section free of cur-taining artifacts without the need for time-consuming low current polish steps.

EXAMPLES OF APPL ICATIONS OFPLASMA FOCUSED ION BEAM

Through Silicon ViasTSVs are themselves subject to a number ofeffects that can result in defects and failures.For example, the large differences in thermalexpansion between copper via fill and the sur-rounding silicon substrate can cause crackingwithin the copper and delamination from thevia sidewall during thermal processing. ‘Key-holing’ results from incomplete filling of vias(Figure 3).

Solid Liquid Interdiffusion BondingOne of the most difficult issues to address isthe behavior of bonds between chips duringsubsequent processes (Figure 4). For example,it is critical that a bond between the first chipsin the stack not be disturbed by the subse-quent bonding of an additional chip. Solid-liquid-interdiffusion (SLID) [4] is a uniquedirect metal bonding technology that avoidsremelting of existing bonds during the forma-tion of new bonds by using high melting inter-metallic phases. During bond formation, solidmetal diffuses into the liquid phase of a lowermelting metal resulting in high melting pointfinal phase that remains solid during subse-quent bond forming processes.

Anisotropic Conductive AdhesivesAnisotropic conductive adhesives (ACA) can beused [5] to bond wafers together physicallyand electrically using an organic bonding com-pound (benzocyclobutene, BCB) filled with 4-µm sized metal covered polymer spheres(MPS). The BCB assures mechanical strengthwhereas the MPS provide the required electri-cal conductivity at interconnection points. Theconcentration of MPS must be high enough toensure good electrical contact betweenopposed pads and at the same time lowenough to guarantee electrical insulationwhere pads are not present.

To study the bonding in detail, samples werecleaved, then milled with the plasma-FIB toreveal the bonding region and finallyinspected with plasma-FIB imaging. Theplasma-FIB milling speed makes it possible toprepare the sample (~200 � 50 � 600 µm3

material removed) within 30 minutes. Themetal layer covering the polymer spherescould be observed at the bonding interfacewith sufficient resolution to estimate both thelocal MPS density and their compression statebetween the bond pad metal layers. In Figure5 the bonding process is illustrated in the top

Figure 4: The void between these pads is the result of an incomplete SLID bonding process. The various intermetallic phases are clearly visible above, below andto the right of the void.

MICROSCOPY AND ANALYSIS NOVEMBER 2011 11

Figure 3: (a, b) Differing thermal expansionbetween copper via fill and silicon substrate caused delamination shownin this via before (a) and after (b)annealing. (c) Keyholing occurred when this viawas not filled completely with tung-sten.

c

a

Cu

Cu3Sn

Cu6Sn

5

FIB debris

b

MICROSCOPY AND ANALYSIS NOVEMBER 201112

images and the bottom plasma-FIB imagesshow details of the bonding interface andcompressed spheres.

3D Test ChipThe 3D integrated reliability test chip shown inFigure 6 is a 3-level-stack with a modular lay-out designed to permit evaluation of assemblyprocesses between two initial layers and, sub-sequently, the effects of adding a third layer[6]. The PFIB can mill a cross section throughthe entire three layer stack showing criticaldetails of both upper and lower bondingregions and the complete TSV through themiddle layer.

CONCLUS IONSBy combining high-speed milling and deposi-tion with precise control and high qualityimaging, the plasma focused ion beam pro-vides critically needed physical analysis for TSVand bonding processes that are essential tocurrent 3D integration schemes. At high beamcurrents, cross-sections with dimensions ofhundreds of micrometers can be completed inless than an hour, fast enough to provideeffective feedback on process performance. Atlow beam currents, the same system delivershigh resolution imaging for accurate structuralanalysis.

The PFIB provides an effective, practical toolfor a variety of 3D integration applications,including failure analysis of bumps, wirebonds, TSVs, and stacked die; site specificremoval of package and other materials toenable failure analysis and fault isolation onburied die; circuit and package modificationsto test design changes without repeating thefabrication process or creating new masks;process monitoring and development at thepackage level; and defect analysis of packagedparts and MEMS devices.

REFERENCES1. Smith, N. S., Skoczylas, W. P., Kellogg, S.M., Kinion, D.E.,

Tesch, P.P., Sutherland, O., Aanesland, A., Boswell, R.W. HighBrightness Inductively Coupled Plasma Source for HighCurrent Focused Ion Beam Applications. J. Vac. Sci. Technol.B24(6):2902-2906, 2006.

2. Kellogg, M., Schampers, R., Zhang, S.Y., Graupera, A.A.,Miller, T., Laur, W.D., Dirriwachter, A.B. High ThroughputSample Preparation and Analysis using an InductivelyCoupled Plasma (ICP) Focused Ion Beam Source. Microsc.Microanal. 16(Suppl 2):222-223, 2010.

3. Kwakman, L., Franz, G., Taklo, M. M. V., Klumpp, A., Ramm,P. Characterization and Failure Analysis of 3D IntegratedSystems using a novel plasma-FIB system. Proc. InternationalConference on Frontiers of Characterization and Metrologyfor Nanoelectronics, Grenoble, France, 2011.

4. Ramm, P. Method of making a three-dimensional integratedcircuit. US Patent 5,563,084; P. Ramm, A. Klumpp. Methodof vertically integrating microelectronic components. USPatent 6,548,391.

5. Taklo, M. et al. Anisotropic Conductive Adhesive for Wafer-to-Wafer Bonding, Proceedings of 7th Intl Conference andExhibition on Device Packaging, March 2011.

6. Ramm, P., Klumpp, A., Franz, G., Kwakman, L. FailureAnalysis and Reliability of 3D Integrated Systems. Proc.IMAPS Device Packaging Conf., Scottsdale, Arizona, 2011.

©2011 John Wiley & Sons, Ltd

Figure 6: The high milling speed of PFIB permits cross-sections through the full three layer stack of the test chip, revealing both upper and lower bonding regionsand the entire TSV.

Figure 5: Anisotropic conductive adhesives provide mechanical bonding and electrical conductivity. (a,b) Images show metal coated spheres before mixing withBCB (a), and a schematic of how TSVs can be electrically connected to pads on another wafer using BCB filled with such spheres (b). (c-f) The lower four images are a clockwise sequence of increasing magnification with the compressed metal coated spheres clearly visible in the twobottom images.

f e

d

a b

c