The pixelated array detector (PAD) is a planar array of complementary metal oxide/semiconductor (CMOS) charge-collecting electrodes and readout circuitry for measuring the electric charges on particles in a charge-detection mass spectrometer (CDMS). In comparison with a single-Faraday-cup detector occupying the same total area, the PAD offers advantages, as explained below.

The CDMS approach offers a faster, cheaper alternative to pulsed gel electrophoresis and other techniques for measuring the masses of a variety of large molecules (masses > 106 daltons) and similarly sized particles. Examples of particles amenable to CDMS analysis include polymer molecules, bacteria, viruses, and airborne contaminant particles. Whereas a typical analysis by pulsed gel electrophoresis takes days, a typical analysis by CDMS takes minutes.

One Pixel Circuit and part of the circuit common to all pixels in a column is depicted in this abbreviated schematic diagram.

In a CDMS, a particle to be analyzed is first subjected to electrospray ionization, causing it to bear an electric charge as high as hundreds of thousands of fundamental (electron) units. The particle is then accelerated electrostatically into a Faraday tube connected to circuitry for measuring the time of flight of the particle along the tube. Upon leaving the Faraday tube, the particle impinges on a charge-collecting electrode or else enters a Faraday cup, the electrode or cup being connected to a charge-sensitive amplifier. The time of flight through the Faraday tube is related in a known way to the charge-to-mass ratio of the particle. Thus, the mass of the particle can be calculated from the Faraday-tube time-of-flight measurement and the measurement of collected charge.

The collection area and thus the fraction of ions collected increases with the size of the charge-collecting electrode, but the capacitance and associated electronic noise also increase with size. This leads to the basic idea of the PAD, which is to use a sufficiently large total charge-collection area to obtain the desired charge-collection efficiency while apportioning the area among multiple small electrodes or Faraday cups for measurement of the charge deposited into them. Because each electrode occupies a small fraction of the total charge-collection area, the capacitance and electronic noise are reduced accordingly.

A proposed CDMS containing a PAD would include an optional electrostatic deflector, which, when triggered by the passage of an ion through the tubes, would prevent additional ions from reaching the PAD until the PAD had been read out and reset. The ion-passage-induced trigger signal would also be used to gate the readout of the pad to reduce the noise associated with collection of dark current. The time needed to clock the flight of a single ion and measure its charge should be no longer than a few milliseconds; thus, it should be possible to measure a hundred or more ion masses per second and to accumulate a complete mass spectrum with data on thousands of ions in less than 30 seconds.

A prototype PAD contains a 28-by-28 array of pixels with a 40-by-40-µm pitch. The charge-collection electrode of each pixel is a 36-by-36-µm metal patch on its surface. Each pixel (see figure) contains a source-follower input transistor, a row-selection transistor, and a row-reset transistor. At the bottom of each column of pixels there is (a) a load transistor, (b) an output branch containing a sampling switch (SHS) and sample-and-hold capacitor (CS) for storing signal levels, and (c) a similar output branch with a switch (SHR) and sample-and-hold capacitor (CR) for storing reset signals. There are also source followers with a column-selection switches (COL) at both ends of each column. The reset and signal levels are read out differentially, suppressing fixed pattern noise. If signal levels are read out twice - once before and once after integrating charge - then kTC noise (where k is Boltzmann's constant,T is absolute temperature, and C is capacitance) is also suppressed. In tests, the prototype PAD exhibited a noise floor of 90 electrons root mean square at room temperature.

This work was done by Stephen D. Fuerstenau and George A. Soli of Caltech for NASA's Jet Propulsion Laboratory. NPO-20128