Transcript The Chemical Sensor Goal - Stanford Microsystems Group

Selectivity: Lock & Key vs. Fingerprint
Chemical Sensor Review
Stanford University, ME342, July 19th, 2007
Tracy Fung, Yun Seog Lee, Beth Pruitt
Techniques:
Chromatography and Spectrometry
(Separation and Detection)
Source: Albert, Chem Rev,100,2595, 2000.and COINS 2007 Annual Report
Some microfabricated chemical sensors:
Source
Bench-Top Gas Chromatographs
Portable Gas Chromatographs
Micro-Chem-Lab (μChemLab) on a Chip
Ion Mobility Spectrometry
Mass Spectrometry
Electrochemical Sensors
Conductometric Sensors
Polymer-Absorption Chemiresistors
Catalytic Bead Sensors
Metal-Oxide Semiconductor Sensors
Potentiometric and Amperometric Sensors
Mass Sensors
Surface Acoustic Wave Sensors
Microcantilever sensors
Optical Sensors
Fiber Optic Sensors
Colorimetry
Infrared Sensors
Images here
for these
techniques
from sandia
and albert
paper??
Terminology:
LOD – Limit of Detection
VOC – volatile organic compounds
PCA - principle components analysis
TON – threshold odor number
GC – gas chromatography
TCR – thermal conductivity detector
ppb – parts per billion
CA – cluster analysis
Properties or metrics to separate or
evaluate:
The Chemical Sensor Goal:
Source: COINS annual report 2007
What else goes here?
Highly sensitive and selective with low
percentage of false positives. While these
devices have been miniaturized the excellent
ones are still in bulk form. These include
receptor-based and optical sensors such as
mass spectrometry and Gas Chromatography.
Add discussion of need for preconc, and
sample detection limits here from coins?
Preconcentration needed to sample
small quantities in large volumes:
Serial - Serial detection processes separated
Source
components from a combination of gasses past a
single detector. e.g. Gas chromatography.
Parallel - Parallel
processing uses use large arrays of
detectors with combinations of
functionalized coatings. e.g.
microcantilever arrays.
Readout: optical, resistive,
electrical, resonance…
Source
Source: Albert, Chem
Rev,100,2595, 2000.
Source
Source: Albert, Chem Rev,100,2595, 2000.
What goes here?
Principal components analysis (PCA) reduces multidimensional data
sets to lower dimensions for analysis. The first Principal Component
ideally represents the dominant gradient. The second Component,
orthogonal to the first, will explain some residual variation. The third
will be orthogonal to first and second and so on.
Rao, C.R. The Use and Interpretation of Principal Component Analysis in Applied Research. Sankhya A 26, 329 -358 (1964)
A tutorial on Principal Component Analysis, Lindsey Smith (2002):
ttp://www.cs.otago.ac.nz/cosc453/student_tutorials/principal_components.pdf.