Sources of Measurement Imprecision Possible Areas for

Sources of Measurement Imprecision

Possible Areas for Measurement Errors • Definition of what is to be measured “WIDTH or Dimater or Critical Dimension” • Environmental Influences – Vibration, Stray Fields • Instrument Maintenance • Specimen Charging • Electron Detector Position and Type • Accelerating Voltage Affects • Sample Contamination • Sample Dimensional Changes • Errors due to specimen tilt • Resolution of the measurement system

Possible Areas for Measurement Errors (cond. ) • • • SEM Magnification calibration Raster rotation Lens Hysteresis Temperature variations Conductive coating Noise on the signal Repeatability of low voltage operation Interpretation of the signal profile Choice of measurement algorithm Acoustic Coupling

Environmental Variables Influencing the Measurement • Electromagnetic Interference – Tool Design • instrument shielding • electron source/column design • Mechanical Vibrations – Robust design – Supplementary vibration isolation • Acoustic Coupling – E-beam column sensitivity – Dampened enclosures • Temperature changes

Material Variables Influencing the Measurement • Dielectrics – polymers, ceramics • Conductors – metals, doped silicon • Edge Roughness • Surface Roughness • Wall angle • Wall shape • Contact hole issues

Human Variables • Operator – Training – Skill level – Experience • Automation – Minimizes operator variability • Interpretation of the data – Where are the problems? – Are there any problems? – Is the instrument working properly?

GOAL • The goal of this portion of the tutorial/ course is to make you aware of some of the metrology pitfalls so that you might avoid them.

Resolution of the Measurement System

Pixel Resolution (512 Pixels; 115 mm field width)

Effect of Pixel Size on the Measurement • NIST RM 8090 sample measured in a commercial CD-SEM system • 0. 2 micrometer pitch structures and 50 micrometer pitch structures • Instrument operating the same in both instances, the sample is the same only the magnification (= pixel size) differs. – 512 pixels

Measurement Data SRM 2090 Prototype • . 2 µm lines – Maximum = 0. 205 µm – Minimum = 0. 198 µm – Mean = 0. 202 µm – 3σ - 0. 006 µm (1σ - 0. 002 µm = 100, 000 x pixel size) • 50 µm lines – Maximum = 50. 440 µm – Minimum = 50. 280 µm – Mean = 50. 370 µm – 3σ - 0. 136 µm (1σ - 0. 045 µm = 5, 000 x pixel size)

Edge Detection

Types of Measurement • Pitch measurement – No need to know the position of the “true” edge. – Self compensating measurement. – Requires that the lines have the same edge shape. – Requires that the same definition of “edge” be applied to both measured edges. • Width Measurement – Position of the “true” edge is needed.

Types of Measurement • Why is the type of measurement important? • What is the effects of the electron beam interaction? • How does the signal mechanism used make a difference?

SE vs. BSE Metrology • SE images of the same structure consistently measure larger than the BSE image • Experimental verification is difficult and good samples are needed. • Further modeling must account for this difference.

Signal to Noise Ratio

SEM Signal and Contrast • Signal is what we choose to collect. • Signal can be: – Secondary electrons – Backscattered electrons – Transmitted electrons – etc. • This signal is collected by an appropriate detector.

SEM Signal and Contrast • All systems have noise, the higher the signal relative to the noise the better the signal quality and image. • The signal quality can be expressed by the signal-to-noise ratio:

SEM Signal and Contrast • Therefore, as the mean number of counts increases, the S/N ratio improves. • Signal integration is primary to effective SEM and especially SEM metrology. • Field emission instruments and digital frame averaging are two improvements that have helped to a great extent to improve the S/N ratio and thus SEM imaging and measurements.

Environmental Influences

Environmental Influences • Stray fields – AC Interference • Vibration – Instrument induced – Site induced – Operator induced

AC Interference • Mains frequency produced by AC current flowing in a loop • Field Strength – Generally proportional to the current and inversely proportional to the area of the loop

Sources of AC Interference • Coils of large electric motors or transformers • AC Voltage regulators • Main trunk power cables • Mistakes in building wiring – Neutral return erroneously attached to the ground wire on a piece of equipment which is also grounded independently • ground strap – Creates a large loop

Sources of AC Interference • Two possible current paths between any two nodes in house wiring • Instrument itself – internal wiring faults – loose connectors – incorrectly installed accessories – mechanical pump motors – TV Monitors – Improperly shielded computers – Ion pumps

Methods of Dealing with Stray Fields

AC Interference Solutions • Field effects diminish with at least the third power of the distance – Attack the source first – Move source – Rotate source – Move instrument • Shielding – Shield source – Shield instrument • permalloy, Mu Metal

AC Interference Solutions • There are consulting companies who will come to your lab and develop a plan to solve these problems. – field canceling coils etc. • REMEMBER: these problems affect all types of instruments - because the SEM is primarily an imaging system it is able to show these problems to you.

The Influence of Vibration

The Influence of Vibration • The SEM should be isolated from all forms of vibration – air flow – acoustically induced – mechanical • Vibration will be included in the measurement

Measurement where tilt is present

Line Tilt/Line Rotation Errors • For precise measurements the specimen should be: – Viewed flat (00 tilt) – Lines should be physically rotated to yield a symmetrical profile (may not be necessary in digitally addressable scan systems) • Instrument should be calibrated at that working distance and the final lens current measured. • Any further focusing should be done with “Z”

Effects of Raster Rotation

Effects of Raster Rotation • Electronic raster rotation is often used to straighten structures on the viewing screen prior to measurement. • Raster rotation can distort the image. • If raster rotation is routinely used by the operator it should be tested.

Effects of Electronic Scan Shift

Effects of Electronic Scan Shift • The SEM generally does not do flat field scanning. • The scan is relatively linear at the high magnifications used for submicrometer work. • Moving the scan to the far portions of the field by electronic scan shift can induce a distortion. • This will affect the measurements across the field.

Stage Drift

Stage Drift • SEM stages can move large distances at a rapid speed. • Once the stage stops there is the potential for drift during the measurement process. • If the image is frame averaged this will result in larger than expected measurements.

Stage Drift • Temperature variations of only a couple of degrees can cause the stage mechanism to expand. • Measurements may vary because of this expansion. • Due to: – Stage motor heating – Room temperature variations – Etc.

Sample Drift • Improperly adhered samples can move when irradiated by the electron beam. • Can look like vibration or drift. • Sample can become charged and fly off the stub – and usually lands where you do not want to to be.

Specimen Contamination

Contamination • Specimen contamination remains one of the major SEM issues to be overcome. • Contamination results from many places: – Vacuum system – Specimen surface – Stage lubricants – Specimen handling – Processing chemistry – Specimen itself

Specimen Cleaning (Contamination Removal)

Specimen Cleaning • NIST has been investigation cleaning techniques for the removal of electron beam induced contamination. • There are tools now available that will “scrub” the specimen chamber and the sample to remove contamination. • Oxygen plasma is a successful technique and others are being pursued.

Specimen Damage

Specimen Damage • Irradiation of the sample with the electron beam can: – Electrically damage fragile devices. – Induce dimensional change in sensitive samples. – Damage the structures through melting or heating. • Low accelerating voltage mode results in the depositing of most of the energy in the structure - not the substrate.

Specimen Charging

Specimen Charging • A full understanding of the effects of charging on image formation and metrology in an SEM has not been fully developed. • Charging is a function of a number of parameters – accelerating voltage – material type – surface tilt – beam current – magnification • Because of instrumental design differences, a sample that will charge in one instrument may not charge in another.

Specimen Charging • Surface Tilt – Beam perpendicular to the sample gives the smallest electron yield – Charging potential increases – Increase tilt and the electron collection improves and the potential for charge build-up decreases • Magnification – The beam current density increases as the magnification increases. • more electrons/area scanned

Operator Factors • Instrument Operation – focus – astigmatism – misalignments – misinterpretations • Instrument maintenance – contamination