Mass Spectrometer Autosampler Jake Ahrens Stephen Pearson Kyle Sala Seth Yellin Project Mentor: Professor Jane Hill.

Mass Spectrometer AutosamplerJake AhrensStephen PearsonKyle SalaSeth Yellin

Project Mentor: Professor Jane Hill

Project Background

•Mass Spectrometer – Highly sensitive device that analyzes volatile organic compounds (VOC)

•Pathogen bacteria – Mass spectrometry is used to identify different pathogen bacteria

•Identification – Every pathogen has a specific VOC profile

Project Background

•Pathogen Library – Build a database of pathogen bacteria profile under various conditions.

•Unknown identification – Ultimately will be used to identify unknown bacteria in human lungs, by comparing the mass spectra of the humans breath, to the pathogen bacteria database

Project Motivation• Currently, each sample must be connected by hand,

tested, and then disconnected and cleaned.

• To build a large database, the sampling process needs to be automated

• The sampler will reduce bacterial aerosol & person contact

• Reduce sample contamination

• Increase data reproducibility

+ = ?

•Design an autosampler to be integrated with a mass spectrometer. This device would allow the user to hook up multiple samples, and then test them all sequentially.

Problem Objective

Project Goals

•Maintain temperature of samples at 37°C•Sample bottles need be 100ml•System needs to be air tight•Flow rate needs to remain constant at 2

L/min•System flushing and headspace

replacement•Stay within $5000 budget

Method of Sampling

Before the Autosampler…

Project Progression

1. Cartesian Robot

2. Valves, Needles, and Septa Bottle Tops

3. Manually Connect Bottles to Valves

4. Incubator

Cost

Fluid Dynamics in Needles

Familiar Set-up with Current Bottles and Tops

Acquired Instead of Built an Enclosure

Our Device

Our Device

Circuit Board

Major Features

•Valve System•Microcontrollers

▫Code•Incubator

Valve System

•3-Way Solenoid pinch valves

•Used to control the path of Air and CO2

Microcontrollers

•2 Independent microcontrollers

•Push button switches used to open all valves to replace tubing for cleaning

Valve Code

•Written in C

•Uses subroutines for each phase▫Flushing▫Collect CO2 and medium spectra

•Constants for these times defined in beginning

•Easy for user to change time constants depending on the tests they are running

Incubator•Temperature accurate

within 1% of target temp

•Samples need to be at constant temperature for growth and testing

2 Sample Schematic

Valves

Initial Flush of Main Line

Collect CO2 Spectrum

Collect Medium Spectrum

Sample 1

Air Replacement

Flushing/Collect CO2 Spectrum

Collect Medium Spectrum

Sample 2

Operation of device

•Attach sample bottles to the valve system (one inlet and one outlet per bottle)

•Turn on CO2 and Air supplies

•Turn the power switch on

•Save spectrum graphs for the samples

Operation•Loading of sample bottles

Design Demonstration

•Three samples, all contain water

• ‘Bubbles’ depict the CO2 flow into and out of the sample bottles

•Three subsequent tests▫Testing with standard compounds that

bacteria could produce in natural environment. The two we used are acetone and indole.

Test #1•Volumetric flow rate (2 L/min)•Ran auto-sampler with all 9

sample bottles in place▫Testing the code that controls the

valves and subsequent gas flow•Flushing and sampling sequences

▫Simulated in complete fashion

•Successful Test!

Test #2• Demonstration of TIC (total ion

chromatograph) of sampling process• Having the mass spectrometer effectively read

samples from the autosampler• Three sample test

▫One background control and two standard compounds

▫Water and two samples of acetone Acetone – Commonly produced by bacteria

• Successful Test!

Test #2 ResultsDemonstration TIC (Total ion chromatograph) of sampling process of one background control and two standard compounds

TIC of +Q1: from Sample 14 (PM Test 4) of 04-28.wiff (Turbo Spray) Max. 2.0e8 cps.

0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5Time, min

0.0

1.0e7

2.0e7

3.0e7

4.0e7

5.0e7

6.0e7

7.0e7

8.0e7

9.0e7

1.0e8

1.1e8

1.2e8

1.3e8

1.4e8

1.5e8

1.6e8

1.7e8

1.8e8

1.9e8

2.0e8

Inte

ns

ity, c

ps

5.212.71

5.08

4.13

7.22 7.70 7.95

Step 3: Start collect sample 1 Step 5: Start collect sample 2

Step 4: CO2 flush

Step 6: CO2 flushStep 1: Water background collection

Step 2: CO2 flush

Time (min)

Intensity (%)

Mass spectrum of step 1 (signal response to bottle 1 – water)

+Q1: 1.029 min from Sample 14 (PM Test 4) of 04-28.wiff (Turbo Spray) Max. 1.7e6 cps.

20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200m/z, amu

5%

10%

15%

20%

25%

30%

35%

40%

45%

50%

55%

60%

65%

70%

75%

80%

85%

90%

95%

100%

Re

l. Int. (%

)

149.2

57.0

43.1

129.1185.155.0

59.0

29.241.1 184.2

69.0 111.1 170.1121.245.0 139.283.1 105.1 119.356.1 71.1 150.2101.0 171.1147.391.2

88.2 167.272.1 109.0107.1 187.2157.181.2 96.9 169.0113.1 173.2132.9130.239.1 46.0 145.369.9 199.258.0 108.773.1 156.799.090.6 180.7120.1 188.7177.4158.731.2 80.1 93.952.9 132.086.040.1 134.251.030.2

Intensity (%)

Mass/charge

Mass spectrum of step 3 and 5 (signal response to bottles 2 and 3 – acetone)

+Q1: 2.885 min from Sample 14 (PM Test 4) of 04-28.wiff (Turbo Spray) Max. 3.8e6 cps.

20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200m/z, amu

5%

10%

15%

20%

25%

30%

35%

40%

45%

50%

55%

60%

65%

70%

75%

80%

85%

90%

95%

100%

Re

l. Int. (%

)

58.7

41.2

60.0

39.243.0

44.1

57.0

31.1

149.1

42.1

129.1185.1

55.061.0 69.0 184.1

111.145.183.2 170.2139.1121.340.3 88.1 101.0 150.271.0 167.246.038.3 105.093.172.929.2 119.3 130.2136.1 186.2 199.3177.2169.1143.385.167.056.0 84.2 160.651.027.2

+Q1: 5.343 min from Sample 14 (PM Test 4) of 04-28.wiff (Turbo Spray) Max. 3.8e6 cps.

20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200m/z, amu

5%

10%

15%

20%

25%

30%

35%

40%

45%

50%

55%

60%

65%

70%

75%

80%

85%

90%

95%

100%

Re

l. Int. (%

)

58.8

43.2

41.1

39.1

149.2

44.1 57.0

60.0

31.2

42.1 55.0129.1

185.1

184.169.0

45.1 111.1 170.2121.183.0 139.1128.6101.071.129.2 150.161.0 90.9 167.240.2 56.0 81.1 109.3 173.346.0 113.3 187.2131.289.1 123.3 199.038.1 143.2 181.199.2 151.365.032.2 68.151.1 76.127.1

Test #3•Second demonstration of TIC (total ion

chromatograph) of sampling process•Sampling the more volatile compound of

indole in varying bottle positions▫Indole – characteristic compound from

E.Coli bacteria•Able to get a momentary reading in the

first bottle position and a successful reading in the last (ninth) bottle position

•Why is the mass spectrometer not reading this compound?

Test #3 ResultsIndole test at valve position 1

TIC of +Q1: from Sample 15 (PM test 4) of 04-28.wiff (Turbo Spray) Max. 2.1e8 cps.

0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0Time, min

0.0

1.0e7

2.0e7

3.0e7

4.0e7

5.0e7

6.0e7

7.0e7

8.0e7

9.0e7

1.0e8

1.1e8

1.2e8

1.3e8

1.4e8

1.5e8

1.6e8

1.7e8

1.8e8

1.9e8

2.0e8

2.1e8

In

te

ns

ity

, c

ps

5.19

4.944.69

2.18

0.684.44

+Q1: 2.157 min from Sample 15 (PM test 4) of 04-28.wiff (Turbo Spray) Max. 2.6e6 cps.

20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200m/z, amu

5%

10%

15%

20%

25%

30%

35%

40%

45%

50%

55%

60%

65%

70%

75%

80%

85%

90%

95%

100%

Re

l. In

t. (%

)

118.1

149.2

91.1

57.0

129.1

43.2 55.1 59.0 185.2117.2

119.129.2

184.190.569.1 121.1111.1

41.3 45.1 170.156.0 139.3105.183.271.1 101.1 150.1 171.2109.1 167.193.1 133.288.2 107.165.1 77.139.1 72.0 130.2113.1 187.1145.3 173.2161.044.1 46.1 58.0 199.1143.399.0 174.231.1 80.0

Short time response of Indole (peak 118) in position 1

+Q1: 1.304 min from Sample 18 (PM test 6 position 9) of 04-28.wiff (Turbo Spray) Max. 1.9e6 cps.

20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200m/z, amu

5%

10%

15%

20%

25%

30%

35%

40%

45%

50%

55%

60%

65%

70%

75%

80%

85%

90%

95%

100%

Re

l. In

t. (%

)

118.1

149.1

59.0

55.0 91.157.0

43.1

129.1185.1

69.0

41.229.1

45.1184.0117.2119.2

83.1 111.1170.1121.256.042.1 139.2

105.071.139.3 100.9 150.2 171.2147.3 167.144.0 87.981.0 107.1109.1 187.0133.1130.1 173.265.1 77.0 89.158.031.1 159.1143.1145.1 199.246.1 181.2103.0 189.1120.2 163.0122.253.0 115.187.684.1 137.078.640.2 65.954.330.1

TIC of +Q1: from Sample 18 (PM test 6 position 9) of 04-28.wiff (Turbo Spray) Max. 1.5e8 cps.

0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0Time, min

0.0

1.0e7

2.0e7

3.0e7

4.0e7

5.0e7

6.0e7

7.0e7

8.0e7

9.0e7

1.0e8

1.1e8

1.2e8

1.3e8

1.4e8

1.5e8

In

te

ns

ity

, c

ps

1.151.012.56 4.09 4.30

4.58

4.925.243.53

Indole test at valve position 9

Stable response of Indole (peak 118) in position 9

Test Results• Auto-sampler mechanically works and

operates like it was designed too!!• Effectively samples and flushes bottles• Maintains volumetric flow rate (2 L/min)• Able to send samples to mass spectrometer for

identification• But…• Some compounds, as seen with indole, cannot

be successfully read▫Possible problem could be chemical interactions

with the valves’ silicone tubing

Desirable Attributes Outcome

• Effectiveness▫ Temperature Control – Incubator (~37°C)▫ Constant CO2 flow rate – 2 L/min▫ Automation – effectively samples nine bottles sequentially,

with some user interaction• Ease of Use

▫ Setting up sampler – familiar process of manually connecting the tubing

▫ Simplicity to operate – flipping a switch• Economy

▫ Cost of parts – under budget (< $5,000)▫ Replacement cheapness (cost to replace the parts that did

not break because of the ones that did)– instructions from an annotated user manual, easy to replace parts

Project Outcome•Sampling process automated, to build large

library

•The sampler will reduce bacterial aerosol & person contact

•Reduce sample contamination

•Increase data reproducibility

Future Development•Reduce dead space tubing for better

results

•More samples to be taken

•Tube track to prevent any pinch points and allow for easier tube exchanges

•More chemically inert tubing

•Higher levels of automation▫Computer automation

Future DevelopmentComputer automation:

• Develop a program to allow interface with sampler and mass spec

• Mass spec uses Analyst software that can communicate with other windows programs to create databases of samples

Beta

Acknowledgements

•Yin-Ming Kuo – Post Doctoral Fellow in The Hill Lab

•Jiangjiang Zhu – Ph.D student in The Hill Lab•Professor Stephen Titcomb•Paul Sala

Questions and Discussion