jjjohn_psd9_pimms_v2

Download Report

Transcript jjjohn_psd9_pimms_v2

PImMS
a fast event-triggered pixel detector
with storage of multiple timestamps
Mark Brouard, Ewen Campbell, Edward Halford, Alex Johnsen, Jason Lee,
Craig Slater, Claire Vallance, Edward Wilman, Benjamin Winter, Weihao Yuen
Chemistry, University of Oxford
Iris Friedli, Laura Hill, Jaya John John, Andrei Nomerotski, Robert Pisarczyk
Physics, University of Oxford
Andy Clark, Jamie Crooks, Iain Sedgwick, Renato Turchetta
STFC Rutherford Appleton Laboratory
14 September 2011 - PSD9, Aberystwyth
1
Outline

PImMS: Pixel Imaging Mass Spectrometry

PImMS1 sensor
 Context
/ requirements
 Design
 Initial results

Future directions
PImMS 1
2
Context: time of flight mass spectrometry
Mass spectrum for
human plasma
~ 100 µs duration
3
Context: ion imaging

Use a position sensitive detector to obtain x-y distributions
– learn about reaction dynamics

Need to tune the timing to select one ion
S atom ion images for OCS photodissociation at 248nm
4
Pixel Imaging Mass Spectrometry

Combines time of
flight MS with 2D
ion imaging

Takes advantage
of recent
advances in
silicon to image
multiple ions in
one cycle
5
Initial proof of concept

Proof of concept experiments with a fast framing camera
(Dalsa CCD) in 2007-8 for dimethyldisulfide
CH3S2CH3

Required prior knowledge of timing of mass peaks
M. Brouard, E.K. Campbell, A.J. Johnsen, C. Vallance, W.H. Yuen, and A. Nomerotski, Rev. Sci. Instrum. 79, 123115, (2008)
6
Towards sensor requirements
1 register
2 registers
To record both early and late
ions, need multiple memories.
How many? Simulate:
4 registers
1
p (ion detected)

Sparse events  consider
time-stamping approach
3 registers
0.95
0.9
ion13
0.85
0.8
0.75
0.7
0.65
0.6
0
50000
100000
150000
200000
n (ions flown)
all 40 ions simulated
p (ion detected)

Want a fast sensor, flexible to
analyse any mass spectrum
p (ion detected)

1
ion39
0.95
0.9
0.85
0.8
0.75
0.7
0.65
0.6
0
n (ions flown)
50000
100000
150000
n (ions flown)
200000
7
PImMS1 sensor: specifications

72 by 72 pixel array

70 µm by 70 µm pixel

5 mm x 5 mm active area

< 50 ns timing resolution

12 bit time stamp storage

4 memories per pixel

adjustable experimental period, up to ~1ms

programmable threshold and trim – 4 bits per pixel

one test pixel with access to intermediate analogue points
PImMS 1
8
Overview of the sensor
Column bias and
timecode distribution
Pixel
configuration
+ row
addressing
Pixel
configuration
read-back
72 x 72 pixel array
Digital Sense Amplifiers
Digital Readout Path
12
Analogue Readout Path
9
PImMS1 sensor: technology

Light is detected in the
thin epitaxial layer,
< 20µm

With only NMOS
transistors, obtain
limited functionality

PMOS transistors
would compete for
charge

INMAPS process developed at RAL

Isolated N-well Monolithic Active
Pixel Sensors – p+ shield

Gain full CMOS capabilities
10
PImMS1 sensor: technology

0.18µm CMOS
fabrication

INMAPS process

615 transistors per pixel

over 3 million transistors
in all
7.2 mm
Sensor design:
Andy Clark and Jamie Crooks,
STFC Rutherford Appleton Laboratory
11
PImMS pixel
Preamplifier
Shaper
Charge
Collection
Diodes
hit
Comparator
12-bit timecodes
12
Pixel operation
diode
preamplifier
shaper
crossing => hit
crossing => hit
comparator inputs
hit indicator
timecode 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
memory 1 0 0 0 0 0 6 6 6 6 6 6 6 6 6 6 6 6 6
memory 2 0 0 0 0 0 0 0 0 0 0 0 0 0 14 14 14 14 14
13
Pixel layout
14
Readout: camera

USB control and
readout

F-mount SLR lens

Cooling system

Option for
nitrogen/dry air
15
Readout: software

The camera is controlled and read out by bespoke
LabView software.

Data can be saved to disk for offline analysis.

A growing library of online and offline visualisation
tools is available.
Software design:
Jason Lee,
Oxford Chemistry
16
Analogue readout

Corresponds to the
output of the preamplifier

Represents the total
charge stored in each
pixel, cumulative for all
hits during a given
experiment

Mainly used for
focusing an optical lens
onto the phosphor
screen
17
Digital readout – multiple hits
5 laser hits, 30µs apart
A = analogue image,
integrated over all
hits
1 = 1st memory
2 = 2nd memory
3 = 3rd memory
4 = 4th memory
A
1
2
3
4
18
# of hits
40µs apart = 800 timecodes (50ns/timecode)
# of hits

# of hits
5 laser pulses, 25ns long, at 405nm
Timecode

1st four pulses:
# of hits
Digital readout – 3D visualisation
Timecode
19
Threshold

The threshold for experiments is set by two adjustable
analogue voltages (generated on the camera)

This shows the spot produced by a defocused class 1 laser
at increasing threshold levels.
400 mV
200 mV
100 mV
55 mV
0.1 mV
20
Sensor calibration

Maximum trim ~50mV

Dispersion (sigma) before
and after calibration:
12.5 -> 4.5 mV (this plot).

With subsequent
improvements to the
software, the current
dispersion is 2.3 mV.
net threshold
Each pixel has 4 bits of
trim and can be masked
Rising Edge, VthP-VthN wrt.
DAQ15

Pixel response to trim
60
a5
40
e8
20
i11
0
-20
1
6
11
o14
u17
-40
aa20
-60
-80
16
ae23
trim_naive
ai26
trim
21
Optical testing
Quantum
efficiency 8-9%
for visible light
 Max @ 470 nm
 Fill factor 20% for
front illuminated
 Full well capacity
24,000 e
Photon Transfer Curve (on Log - Log scale)
1.6
Variance - Dark Variance (log(DN))
1.4
1.2
1
0.8
0.6
0.4
0.2
0
1
1.5
2
2.5
3
3.5
4
Signal - Dark Signal (log(DN))
22
Mass spectrometry rig (Oxford Chemistry)
23
Comparison of PImMS and PMT

Same mass peaks seen with PImMS as with a
photomultiplier tube (PMT)

2 fragments of CHCA
24
Velocity map imaging

N,N-dimethylformamide (DMF) is a prototype
molecule for studying peptide bond cleavage.

Early PImMS data on the 193 nm fragmentation
of DMF is shown below.
25
First PImMS spatial imaging results
# of hits

Comparison:
conventional
camera
to PImMS
1
2
3
4
timecode
26
PImMS2

Larger array: 324 by 324 pixels

23 mm by 23 mm active area

380 experiments/sec

Potential 400,000
measurements per
experimental cycle

Designed to also work directly
after MCP – reduced pin count
for vacuum applications

Improved power supply, routing
and trim

Submission this autumn, ready
by early 2012
27
Future work and directions

Sensor characterisation:

Currently working on: noise, time resolution

Next: spatial resolution, time walk versus light power

In Chemistry, further spatial and velocity map imaging

Possible new applications:


Atomic probe tomography (alloy analysis)

Fluorescence imaging
Larger, improved sensor PImMS2

Submission this autumn; testing in early 2012
28
Summary

PImMS is both a new technique in mass spectrometry
and a specialised sensor for MS

The first sensor has been proven for mass spectrometry

Adding 2D sensing to a time-of-flight mass spectrum
adds structural information and can increase throughput

Multiple memories capture different mass peaks within
one experimental cycle

The second generation sensor will be ready in early 2012
29
Acknowledgements
The support of the EPSRC through Grant EP/G00224X/1, of the STFC through
PNPAS award, of the RC-UK through MI-3 programme (GR/S85733/01) and a
`proof of concept' grant from ISIS Innovation Ltd. are gratefully acknowledged.
Thank you
30
Back-up material
31
Sensor calibration
Number of hits
Pixels are
characterised by
plotting threshold
voltage versus
number of noise hits.
Number of hits
Limited floor space
and manufacturing
tolerances mean that
pixel responses vary.
Global mean
32
Number of hits
Sensor calibration
Number of hits
After
trimming of
pixels
Global mean
33