Update on Hardware since last year30 min

Beam width probesolve the scaling factor

Converging aperturea true PM2.5 instrument

Most important developmentsare going to be...

Summary Statementfrom 2002 Users Meeting…

Converging nozzle to improve collection of micron size particles

A fabrication challenge…

•Two nozzles have been made (not quite to spec).• Measurement results not as expected, no improvement in large particle transmission

1.2

1.0

0.8

0.6

0.4

0.2

0.0

Col

lect

ion

Effic

ienc

y

102 3 4 5 6 7 8 9

1002 3 4 5 6 7 8 9

10002

Calc'd Mass from CPC (ug/m^3)

Particle Transmission Curve

Beam Width Probe (BWP) Assembly

•There have been two field deployments which have used the BWP.

• Mexico MCMA (Jose and Mobile Lab) and Finland (UMIST)

• Already providing good information

• Redesign likely

High Throughput Lens~2.5 cc/s flow rate

Several units have been built.Requires additional pumping (Alcatel hybrid and V301 pump).Performance appears to be qualitatively the same as standard lens.

Integrated Instrument Rack System

Advantages-Ease of shipping -Fast start-up-Don’t need to break vacuum-Don’t need to re-make cable

connections

DisadvantageMust be moved in one piece

AMS Shipping container

Adjustable Inlet with Micrometer heads

Starrett 460MB McMasterCarr #8578A16(need 0.236 reamer)

Small Volume Ionizer

Contains vapor plume…higher pressure…more ions~2x enhancement in effective ionization efficiency

Signals as a function of emission current increase linearly up to 5mA then appear to saturate, except tungsten [log scale] (small contribution of surface ionization?).

Ionizer was retuned at each new emission current setting.

1200

1000

800

600

400

200

0

IPP

(30+

46)

654321

Emission Current (mA)

10x106

8

6

4

2

0

Air Beam (H

z)

102

103

104

105

106

107

Tung

sten

Sig

nal (

Hz)

6x10-6

5

4

3

2

1

Ionization Eff. (Ions/Molec.) AB_Hz (28) W_Sig (m183) IE (350nm NH4NO3) IPP_30_46 (350nm NH4NO3)

High Emission Current OperationUsers Meeting 2002

Action item

• Find a filament material that produces more electrons at lower temperature

• Tested – 2% thoriated tungsten (SIS)– Yttrium oxide coated iridium (Pfeiffer)

High Emission Current Operation Tungsten versus Yttrium Oxide-Iridium

0.01

0.1

1

10

100

1000

Sign

al

300280260240220200180160140120100806040200

AMU

5 mA emission Back-to-Back spectranormalized to m28

MSSClosed_R394 W MSSClosed_R393 Yr

Mass 89 (Yt), 105 (YtO) and 190, 193 (Ir) are probably the Yt-Ox/Ir from the filament

0.01

0.1

1

10

100

1000

Sign

al

1009080706050403020100

AMU

5 mA emission Back-to-Back spectranormalized to m28

MSSClosed_R394 W MSSClosed_R393 Yr

High Emission Current Operation Tungsten versus Yttrium Oxide-Iridium

• Additional 16 (O), 35, 37 (Cl) may be electron induced desorption of ions from the stainless steel parts of the ion source

Light Scattering Particle Detection Module

Eben Cross has made a lot of progress in further developing this unit

Yttrium Coated Filament Summary

• Will elevated background decrease in time?• Will vaporized (and deposited) yttrium cause ionizer instabilities/surface charging. Pfeiffer has some indication that there may be an issue with this.• Need to start using this filament, longevity testing.

Future plans15 min

Key Issues• Laboratory quantification of lens transmission…close the

loop between FLUENT predictions and efforts in the lab to measure transmission.

• Avoid calibrating the AMS to other instruments from field measurements

Collection efficiency!!For mass measurementFor a complete understanding of lens,

to correct data for lens transmission.

Things under development

• Piezo motor to replace chopper servo and beam width probe servo motor• Ultrasonic (electrically driven) atomizer, calibrate at the flip of a switch.• Cryopump to reduce pump out time.• HTP lens• Converging orifice, PM 2.5 inlet• Humidity probe now available• ambient pressure probe available

Ricor MicroStarTM Cryopumpfast pump out / background reduction in ARI Aerosol Mass

Spectrometer ionizer chamber

• Cryoshiled surrounds ionizer• Mounted between P5 (V301) and chamber

http://www.ricor.com

•1000 Ls-1 pumping speed for water• 17 kg (38 lbs)• 400 Watts start-up at 50 VDC

(~200W normal load)

Custom cryoshield surrounds ionizer

Comparison of Mass Spec with and without Cryopump

102

103

104

105

106

107

Ion

Rat

e (H

z)

100908070605040302010AMU

Normal Background Background with Ricor Pump

120K

Water vapor drops ~15x; organics drop between 2 and 10x

Mass Spectrum with and without cryopumping

101

102

103

104

105

106

107

Ion

Rat

e (H

z)

30028026024022020018016014012010080604020AMU

Normal Background Background with Ricor Pump

120K

Ricor Microstar Cryopumpperformance in AMS

102

103

104

105

106

107

Ion

Sign

al (H

z)

6:20 PM9/18/2003

6:45 PM

dat

9:00 AM9/24/2003

10:00 AM

dat

400

350

300

250

200

150

100

50

Cryopum

p Temp, K

10:00 AM9/26/2003

12:00 PM

dat

Water (m18) Organic (m55+m57) Cryopump Temp

Test 1 - Pump from vent condition Cryopump on from start

Test 2 - Cryopump on after 4 days of 'normal' pumping

Test 4 - Ionizer off overnight, turbos on Cryopump on later

Ionizer on

Cryopump on

Description of tests performed• Test 1: After installing cryopump, system was evacuated. Approximately 30 min after tubos

reached full speed ionizer was turned on and cyopump was turned on. Data logging was started at 1 min intervals. Cryopump temperature and skin temperature were manually recorded. At full power, the cryopump skin temperature reached 55C and then cooling air (60 cfm fan) was applied. This was not quite sufficient (note fluctuation in plotted temperature as controller reduced cooling power to match Tskin warning). Water vapor started pumping at ~160K. Note that watervapor only increased ~2x from the end of test 1 to start of test 2. Duration of this test was ~1hr to reach 118K set point at which time I shut off the cryopump and left the system under turbo only pumping over the weekend.

• Water cooling used for Tests 2 and 4, skin temperature never exceed 35C.

• Test 2: To evaluate the effect of the cyopump for a “typical” AMS vacuum state. It was started after four days of turbo only pumping (with ionizer on). Organics (m55 and m57) drop ~2x initially then water vapor drops at about ~15x from 260K to 120K, different temp range than observed from Test 1. After set point of 120K was reached, temperature was lowered to 70K to investigate cooling power. Seemed to be no problem getting to 70K, ss power ~120W. Next turned off ionizer and observed power drop slowly to ~80-100W…ionizer heat load.

• Test 4: Ionizer and cryopump off overnight, turbos left running. Next morning ionizer turned on and data logging started. Note high water and organics. Organics continue to pump out. Cryopump turned on and organics drop soon after ~3x. Similar to Test 2 result. Water vapor later drops a factor of ten. Not temperature record.

Summary of CryopumpPerformance

• Rapid pump out of chamber (from a vented state) in ~1 hour to levels equivalent to several days of turbo pump only pumping.

• Water vapor levels can be reduced by a factor of ~15x.• Organics (m55 and m57) can be reduced by a factor of ~3.• Reducing set point temperature from 120K to 70K does not lead to

further reduction in background.• This pump provides plenty of cooling power even with custom

cryoshield that surrounds the ionizer. Free run to ~50K and condensed N2! (as well as O2 and CO2).

• Vibration of cryopump does not appear to significantly affect signal.• Water cooling of pump will likely be required for operation in ambient

temperatures above 25C (most field deployments).• Need to evaluate pump capacity, how long before regeneration.• Evaluate engineering issues associated with mounting to AMS. • Significant weight and power consumption.

Problems encountered

• Preamp solder joint• bent pins on NI connector • SEV 217 multiplier excessive dark current• wires on servo stage broke due to too much cycling• premature chopper servo failure, too much friction on ball joint linkage• Filaments/Turbo pumps• QMG422 reverts to “some other” settings• Failure of IS420 ionizer controller-emission current regulation.• Quads come loose

Addition of Alcatel reduces work load on P2-V301 ~0.7 Amps. Provides small pressure reduction in skimmer chamber (lens exit region).

6

4

2

0

Turb

o Pu

mp

Amps

4.54.03.53.02.52.01.51.00.50.0

Flow into AMS, cc/s

0.001

0.01

0.1

Skim

mer

Pre

ss

2.0

1.5

1.0

0.5

0.0

Inle

t P, t

orr

Measured with AlcatelMeasured w/o Alcatel

V301 P2 V70 P3

Alcatel

100 um120 um 130 um

140 um

Users suggestions for improvements15

• Manual needs fix’n!

• Shorter chamber • Better system to mount the critical orifice• Add another 3-way valve to connect the 10 Torr

Baratron to the skimmer chamber. • More restrictive vacuum interlock• Inlet manifold supplied with the instrument• A 3-way valve connected to a 1000 torr baratron to

sample the atmospheric pressure in the room or in the line.

One users request for changes

IE/AB comparisons across instruments

10 min

Air Beam and Ionization Efficiency

• Ionization efficiency, IE (ions/molecule)– Sample a particle of known size (no. of

molecules) and measure the number of ions produced. Typically a few ions generated from 106 molecules.

• Air Beam, AB (Hz, ions / second)– The signal at N2 (or O2), use signal strength

as a measure of instrument sensitivity.Typically a few million ions per second.

8x106

6

4

2

0

Air Beam (H

z)

Cal

tech

BC-L

abAR

I IU

MIS

TU

C-T

oohe

ySU

NY

ASU

CEH MPI

ARI I

IU

. Tok

yoU

MIS

T Fl

ight

NO

AAEn

v C

anU

C-J

imen

ezJu

lich

DO

E - G

1N

IES

JCAP

Uta

hU

MR

Wyo

mm

ing

Mai

nz S

ampl

ing

DO

E G

1SA

NYU

JAR

I

8x10-6

6

4

2

0

Ioni

zatio

n Ef

ficie

ncy

Summary of AMS Performance at Time of DeliveryHigh efficiency ionizer

High performance quad

2.0x10-12

1.5

1.0

0.5

0.0

IE/A

B

Cal

tech

BC-L

ab

ARI I

UM

IST

UC

-Too

hey

SUN

Y

ASU

CEH MPI

ARI I

I

U. T

okyo

UM

IST

Flig

ht

NO

AA

Env

Can

UC

-Jim

enez

Julic

h

DO

E - G

1

NIE

S

JCAP

Uta

h

UM

R

Wyo

mm

ing

Mai

nz S

ampl

ing

DO

E G

1

SAN

YU

JAR

I

Group

Average Value = 0.8 x 10-12

Variability in Ionization Efficiency and Air Beam

• Position of filament with respect to slit in ion reference region. IE/AB constant

• Multiplier collection efficiency. IE/AB constant• Pressure in TOF chamber (small internal leaks

can have a measurable effect on AB intensity). IE/AB larger

• Positioning of vaporizer inside ionizer volume IE/AB smaller

• Machining imperfections in skimmer channel. AB could be off-axis. IE/AB larger

Egbert AMS Calibrations

0.00E+00

1.00E-06

2.00E-06

3.00E-06

4.00E-06

5.00E-06

6.00E-06

13-Jul-02 1-Sep-02 21-Oct-02 10-Dec-02 29-Jan-03 20-Mar-03 9-May-03 28-Jun-03 17-Aug-03 6-Oct-03 25-Nov-03

IE

0.00E+00

5.00E+05

1.00E+06

1.50E+06

2.00E+06

2.50E+06

3.00E+06

3.50E+06

4.00E+06

AirB

eam

IEAirbeam (MS)

Environment Canada IE and AB History

AMS Calibration History

0.00E+00

5.00E-13

1.00E-12

1.50E-12

2.00E-12

2.50E-12

13-Jul-02 1-Sep-02 21-Oct-02 10-Dec-02 29-Jan-03 20-Mar-03 9-May-03 28-Jun-03 17-Aug-03 6-Oct-03 25-Nov-03

IE/A

B

IE/Airbeam (as displayed)IE/Airbeam (calculated w/ TOF)

Environment Canada IE/AB History

An alternate calibration procedure

20

Schematic of DMA-less calibration setup

AtomizerElectrically driven

Impactor/Drier

D<2umDilution AMS

Velocity –Size calibration

100

101

102

103

104

Ions

/Par

ticle

106 107 108 109 1010 1011

Molecules/Particle

Particle Diameter (µm)0.2 0.5 1.0 2.0 5.0

• Sample all single particles• Determine IE and linearity in one measurement• Check particle transmission

Ion

Sign

al

0.0080.0060.0040.0020.000

Time of Flight, sec

DOP_TOF.itx

6040200

2 3 4 5 6 7 8 90.01

TOF (s)

1000800600400200

0

1200800400

0

600400200

0

600400200

0

Single particles M46_Part M30_Part M17_Part M16_Part M15_Part Thres

IGOR Procedure under development to analyze single particles to get a multi-point calibration for ionization efficiency

100

101

102

103

104

IPP

- bin

ave

rage

1072 3 4 5 6 7 8 9

1082 3 4 5 6 7 8 9

1092 3 4 5 6 7 8 9

1010

Molecules - bin average

m15AvgIPP m16AvgIPP m17AvgIPP m30AvgIPP m46AvgIPP 3:1 line

NO3 = 2.4e-6NH4 = 3.1e-6

Benefits• Soft, low-velocity particle production, total flexibility with flow systemdesign • Pressure-less operation, smaller, no large pump needed.• No clogging • Number densities can be controlled electronically…without changing flow rates • Size distribution appears to be controllable by varying solution density• Electronically controlled, easily be automated.

Potential Disadvantages (needs further investigation)• Electrical power warms solution.• Liquid level over transducer may be important for stable production

rate…liquid level control• Probably limited to solutions with viscosities similar to water.

Application Goal• Routine poly-disperse calibration source for velocity calibration (PSLs) and ionization efficiency (NH4NO3)• Abandon DMA…!!

Ultrasonic atomization

The average particle size is related to the surface tension (T), density (p) and the frequency (f) of the liquid.

For water, T=.0729N/m, p=1000kg/cu. m and f=2.4 MHz, the size of the particles should center around 1.7 microns.

Ultrasonic Atomizer

3/1

273.0 ⎟⎟⎠

⎞⎜⎜⎝

⎛=

fTDρ

Mode size can be controlled by varying solution concentration

1.5

1.0

0.5

0.0

NO

2+ Ion

Sig

nal

2 3 4 5 6 7 8 9100

2 3 4 5 6 7 8 91000

2

Dva

30

20

10

0

1.0

0.8

0.6

0.4

0.2

0.0

gm salt / gm water 1e-3 1e-5 1.3e-4

10

2

3

456

100

2

3

456

1000

Dry

Dia

m (n

m)

10-7 10-6 10-5 10-4 10-3 10-2 10-1

Conc (gm salt/gm water)

assumes 1.7 um Wet Size, spec. assumes 4.5 um size

Dry particle size is related to solution concentration

NH4NO3 Number Density as a Function of Applied Voltage

10-1

101

103

105

107

CPC

Num

ber

2826242220181614

Transducer Voltage

5

0

Dynamic range of ~103 in control of number concentration

Beam width probe design20

Beam Width ProbePrinciple of operation

1.0

0.8

0.6

0.4

0.2

Tran

smis

sion

-0.4 -0.2 0.0 0.2 0.4

Wire Position (mm)

75 µ

m d

iam

eter

wire

350 nm NH4NO3 DOP

wires.pxp

Half widths from Gaussian fits are 0.13 DOP and 0.23mm NH4NO3 Jayne et al, 2000

18 mm

450 mm

403 mm

378 mm

140 mmPivot point

Channel skimmer1 mm ID x 25.4 L

Channel aperture3.8mm ID x 20 mm L

Channel aperture3.8mm ID x 10 mm L

0.15”(3.8mm) OD Heater

Distances and Aperturesfor 255-xxx AMS Chamber

395 mm

178 mm

Chopper

353 mmBeam Probe

Aug. 2003

Beam Width Probe

Probe mounts here

Easily installedSelf aligning

Run Inventor

Beam Width Probe Program

Run AutoDesk

20

15

10

5

0

Mas

s C

once

ntra

tion

(µg

m)

5:00 AM4/2/2003

6:00 AM 7:00 AM 8:00 AM 9:00 AM

Date and Time

2.0

1.5

1.0

0.5

0.0

Water Ammonium Nitrate Sulphate Organics Chloride wPos

5.5x106

5.0

4.5

4.0

N2

AB (H

z)

10:00 PM4/1/2003

11:00 PM 12:00 AM4/2/2003

1:00 AM 2:00 AM 3:00 AM 4:00 AM

Date and Time

1.0

0.8

0.6

0.4

0.2

0.0 amuc28 wPos

5 min data from Mexico

Issues with Current Beam Probe• Is there enough position resolution?• Is the precision/reproducibility sufficient?• Is Air Beam really centered at aperture?

Exploring Piezo motor technology• Micron resolution.• Better vacuum compatibility. • Probably need an encoder system. • More complex… motor and control circuitry currently being evaluated.

AMS Users meetingHardware developments

-capacitors on preamp and multiplier-heater control

Problems encountered by usersManualsUsers suggestions for hardware improvementsIE, AB, IE/AB comparison across instrumentsConverging inlet1000 torr headVaisala probe Hummiter 50Y…RH/Temp for inlet, RH-fragmentRicor MicroStarTM cryopumpPreamp ranging issueHigh temp heater, metal vapor deposits short heater bias voltageElectronic generation of aerosol, polydisperse calibration approachPlots of Alcatel power for HTP lens system…point out that its never been tested under “harsh” field conditions.

Some of Johns comments• multiplier calibration, no one understands it first time• concept of SI