Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

Chapter 24 - Gas Chromatography

Download as pdf or txt
Download as pdf or txt
You are on page 1of 28

Gas Chromatography Harris Chapter 24

Gaseous Mobile Phase, Solid or Liquid Stationary Phases

11/22/2005

Chem 253 - Chapter 24

Schematic & Major Components (fig 24-1)

11/22/2005

Chem 253 - Chapter 24

Carrier Gas Reservoir Gas Flow Make-up Gas (optional) Injection Port Sample Introduction Separation column Injection Port Oven Temperature T1 Detector Oven Temperature T3
3

Column Oven Temperature T2


Chem 253 - Chapter 24

11/22/2005

Summary of GC components.
Injection port Syringe containing sample is introduced through a septum, injection port oven temperature heated to temperatures that ensures fast volatilization of sample, i.e. above the b.p. of all sample components, usually 275 0C. Keep in mind these terms: split and splitess injection Carrier Gas Sweeps analyte to separation column. Usually, He, or N2, sometimes H2. Column Either packed (old) or capillary (newer) columns. Each type requires its own set of plumbing components (threads, etc.) Interestingly all GC and HPLC plumbing components in the English system of measurements.

11/22/2005

Chem 253 - Chapter 24

Summary of GC components.
Column T May be below b.p. of sample components, but high enough to keep a significant quantity in the gas phase. Temperature may be ramped up to get separation based on b.p. differences of components along with chromatographic separation. Makeup Gas Sometimes required since detector volume is too large for carrier gas flow. Remember that makeup gas is usually required for capillary columns. Detector Ideally, want universal detector, sensitive to anything that passes by. Wide dynamic range, low D.L. the usual wants. We want to keep all sample components volatilized, detector oven temperature usually 300 0C.

11/22/2005

Chem 253 - Chapter 24

Effects of the Column and Mobile Phase Flow in GC.


Why are modern GC based on a capillary column? Back to the van Deemter Eqn. H = A + B/u + Cu Remember, u = flow rate,

A = multiple paths B/u = longitudinal diffusion effects Cu = MT effects

We want to minimize H as much as possible.

11/22/2005

Chem 253 - Chapter 24

11/22/2005

Chem 253 - Chapter 24

11/22/2005

Chem 253 - Chapter 24

11/22/2005

Chem 253 - Chapter 24

Older packed columns


usually 1/8 (3.2 mm OD, 2.2 mm I.D.) diameter, 1 2 m length design impedes gas flow max flow rate about 1 mL/min or 8 cm/min.

11/22/2005

Chem 253 - Chapter 24

10

Newer capillary columns


0.25 mm inner diameter. major point, less restricted flow path. flow rate about 20 ml/min or 20 cm/min

11/22/2005

Chem 253 - Chapter 24

11

We want to minimize H as much as possible.


H = A + B/u + Cu Which of the above affects H the most in GC?

B/u effects! Diffusion coefficients are large in the gas phase.


L = 2D u

Simply increasing the flow rate partially addresses the B/u effects in GC
11/22/2005 Chem 253 - Chapter 24 12

Other major factors that influence the performance of capillary columns relative to packed ones are evident in the table below.
Typical 1/8 packed I.D. df 2.2 mm 5 m Typical Capillary 0.25 mm 0.1-0.53 mm 0.25 m Comments No packing material less restricted flow
MT for s.p. part of Cu
Cs f (k ' )d 2 f Ds

L N

1-2 m 4,000

10-60 m 180,000 Higher sep. efficiency Faster sep. Better for complex mixtures
Chem 253 - Chapter 24

HN = L

Advantages Lower cost Easy to make Larger samples

11/22/2005

13

Other advantages of tubular over packed GC columns:

No A term, straight path.

11/22/2005

Chem 253 - Chapter 24

14

Example of the separation efficiency of capillary vs. packed column.

11/22/2005

Chem 253 - Chapter 24

15

11/22/2005

Chem 253 - Chapter 24

16

GC stationary phases.
Older packed columns uniform silica particles (150-250 m) required to ensure uniform path lengths (the A term in the van Deemter eqn.) Surfaces are chemically modified (see below). The columns themselves were either glass or stainless steel. Capillary columns fused silica which like the particles in the packed column require chemical modification (below).The stationary phase surface (silica) is a hydroylated surface. This caused problems with nonpolar stationary phases. When polar or mildly polar species partition into the s.p. they stand a good chance of being trapped, causing a excessive bandbroadening beyond what is expected from van Deemter considerations.

11/22/2005

Chem 253 - Chapter 24

17

Trapped polar solute Mobile Phase

Stationary Phase
OH OH OH OH OH OH OH OH

Silica Support

11/22/2005

Chem 253 - Chapter 24

18

Surface modification with trimethylchlorosilane reduces surface polarity.

H3C

CH3 CH3 Si Cl + H3C OH O CH3 CH3 Si O + HCl O

OH O

OH O

O Si O Si O Si O Si O

11/22/2005

Chem 253 - Chapter 24

19

Effects can be dramatic:


A = EtOH B = methylethyl ketone C = benzene D = cyclohexane Chromatogram 1 = unmodified Chromatogram 2 = silinazied
11/22/2005 Chem 253 - Chapter 24 20

10

Liquid stationary phases

11/22/2005

Chem 253 - Chapter 24

21

Important Notes:

All columns have a Tmax, never exceed this! Thermal degradation, or s.p. simply boils off. Most s.p. are air sensitive and easily oxidize at high temperatures. Capillary columns cost $150-600 (2002 dollars)
22

11/22/2005

Chem 253 - Chapter 24

11

Effect of Stationary Phase Polarity on Retention Times.


#1 n-heptane #2 tetrahydrofuran #3 2-butanone #4 n-propanol We can reason this order by considering the polarity of the analytes and of the s.p. In GC the s.p. characteristics have the largest effect on the order of elution.

11/22/2005

Chem 253 - Chapter 24

23

Column Temperature and Temperature Programming.


The oven T must be controlled to within 0.5 0C from RT to 400 0C. A fan circulates air throughout the oven ensuring uniform heating. Constant temperature isothermal separations are good for simple mixtures. Typically T programming is required. Slowly ramping T throughout the separation provides a basis for the separation of sample components based on BP.
11/22/2005 Chem 253 - Chapter 24 24

12

Isothermal vs. Temp Programming


The more volatile cpds elute very close together under isothermal conditions (upper). Ramping T from 50 to 250 oC at 8 oC/min separates out those volatile cpds (lower).

11/22/2005

Chem 253 - Chapter 24

25

How does T programming work?


T programming is an example of the control of the capacity
' r factor: k A = t m

t tm

which in turn controls the resolution term:

Rs =

N 4

' 1 k B ' 1 + k B

Note that T programming does not change the order of elution.

11/22/2005

Chem 253 - Chapter 24

26

13

Carrier Gases

O2 is usually avoided since it will oxidize the s.p. 3 most common gases N2, H2, He. Each have very different Hmin characteristics. The lighter gases He and H2 require faster analysis flow rates 20-50 cm/min.

Diffusion in He and H2 causes more band-broadening in m.p. Greater MT effects decrease band-broadening at faster flow rates H = A + B/u + Cu

N2 Hmin occurs at about 10 cm/min.


Chem 253 - Chapter 24 27

11/22/2005

11/22/2005

Chem 253 - Chapter 24

28

14

Sample Injection

11/22/2005

Chem 253 - Chapter 24

29

Injection Port for Modern GCs

11/22/2005

Chem 253 - Chapter 24

30

15

GC Injection Syringe
It is important to rapidly vaporize the sample. Slow vaporization increases band broadening, by increasing the sample plug. Injection port temperature is usually held 50 0C higher than the BP of the least volatile cpd.

11/22/2005

Chem 253 - Chapter 24

31

GC injection and band broadening and anomalies.


Manual injections takes practice and patience. Extremely slow injections will cause band-broadening, wide sample plug. Jerky injections may cause double peaks for the same analyte species. Auto-injectors robotized carousels of sample vials. No practice needed. Can be set to make measurements all night long. Rare in academe, frequent in industry.
11/22/2005 Chem 253 - Chapter 24 32

16

11/22/2005

Chem 253 - Chapter 24

33

11/22/2005

Chem 253 - Chapter 24

34

17

11/22/2005

Chem 253 - Chapter 24

35

Split vs. Splitless Injection


Sample injection is done by a syringe 1 to 5 L or ngs of analyte for the average capillary column. Capillary columns usually require split in injections, a sample reduction method. Depending on the spilt ratio (adjustable) only 0.2 to 2 % of the sample injection makes its way to the column. The rest is discarded.

11/22/2005

Chem 253 - Chapter 24

36

18

Effect of split versus splitless injections on separation quality


The chromatogram (upper left) demonstrates ideal separation characteristics. Upper right illustrates what happens when too much analyte solute is injected onto a capillary column.

11/22/2005

Chem 253 - Chapter 24

37

Solid-Phase Microextraction
Analyte preconcentration method, remember stripping voltammetry. Typical derivatizing agents: C18COO-Si-(O)3-

11/22/2005

Chem 253 - Chapter 24

38

19

11/22/2005

Chem 253 - Chapter 24

39

SPME of Nerve Agents


Fig 24-21. Advantages of SPME Green method compared to solvent extraction Fibers are reusable 20-30x 102-103 LOD enhancement Disadvantage lots of technique and method development.

11/22/2005

Chem 253 - Chapter 24

40

20

Purge and Trap


Major Goal: Remove all analyte from sample. Volatile components are removed from sample by gentle heating and a stream of purge gas, e.g. He. Adsorbents trap targeted analytes, e.g. molecular sieves. After complete purge of sample, contents of trap tube are injected into GC. A purge and trap module is on the Cassini probe to Saturn and moons looking for hydrocarbons, H2, He, NH3. http://saturn.jpl.nasa.gov/home/index.cfm

11/22/2005

Chem 253 - Chapter 24

41

GC Detectors
Ideal detector characteristics, for flowing systems (e.g. GC) large dynamic/linear range response independent of flow rate, i.e. fast response times. Universal detection, responds to all species. Keep this in mind when we discuss HPLC and CE. Additional requirements for GC operation from RT to 400 oC detector response independent of detector oven T

11/22/2005

Chem 253 - Chapter 24

42

21

Thermal Conductivity Detector-TCD


Measures heat loss from a hot filament nearly universal filament heated to const T when only carrier gas flows heat loss to metal block is constant, filament T remains constant when an analyte species flows past the filament generally thermal conductivity goes down, T of filament will rise. (resistance of the filament will rise).

11/22/2005

Chem 253 - Chapter 24

43

Flame Ionization Detector (FID).


Sensitive towards organics Analyte is burned in H2/air, which produces CH and CHO+, radicals, remember our discussion regarding the blue cone in AA.
CHO+ radicals are reduced at a cathode which produces a current proportional to the radical quantity. About 10-12 A Specific for organic carbon, insensitive to inorganics, CO2, SO2 etc. Generally DL 100x less than TCD about pg/s (flow rate dependent) Dynamic range of 107 Response to specific organic depends on the number of organic carbons.
11/22/2005 Chem 253 - Chapter 24 44

22

11/22/2005

Chem 253 - Chapter 24

45

Electron Capture Detector (ECD)


Sensitive to electron withdrawing groups especially towards organics conatining F, -Cl, -Br, -I also, -CN, NO2 Nickel-63 source emits energetic electrons collides with N2 (introduced as make-up gas or can be used as carrier gas) producing more electrons:

Ni-63 => ee- + N2 => 2e- + N2+ The result is a constant current that is detected by the electron collector (anode).

11/22/2005

Chem 253 - Chapter 24

46

23

As an analyte flows through past the Ni-63 source electron capture is possible by electron-withdrawing species: A + e- => ACurrent decreases as a result of e- capture by analyte. This is one of the few instances in which a signal is produced by a decrease in detectable phenomenon.

11/22/2005

Chem 253 - Chapter 24

47

ECD Characrtersitics (the good)


VERY low DL for detected species 10-15g/s for many halogenated substances (PCB, DDT etc). OK dynamic range of 104.

(The bad)
Radioactive Ni-63 source EASILY contaminated with O2, H2O, sample overloading. High maintenance device. Highly variable response to halogenated substances, see table below: Can be a real headache when method developing a specific analysis, e.g. CH2Cl2 in the presence of CCl4.

Sometimes complementary information from FID helps.

11/22/2005

Chem 253 - Chapter 24

48

24

11/22/2005

Chem 253 - Chapter 24

49

Other GC detectors
Nitrogen-Phosphorous Detector (NPD) Also know as the thermionic detector (TID) or alkali flame detector. It is an FID tweaked for N-P cpds, and organics. Flame Photometric Detector (FPD) FID tweaked for S containing cpds. Photoionization Detector (PID) UV ionization of organic analyte, coupled with high voltage cathode and analode results in current proportional to ionized organics.

11/22/2005

Chem 253 - Chapter 24

50

25

FT-IR weve discussed before. (because of generally low in A =bc, most analytes have a modest DL with FT-IR) Mass Spec
51

11/22/2005

Chem 253 - Chapter 24

Qualitative Analysis in GC.


GC-MS offers structural determinations (remember organic?) With other detectors identification is possible with retention times of analyte and standard, however its best if another method is used as a confirmation.
11/22/2005 Chem 253 - Chapter 24 52

26

Quantitative Analysis in GC.


Calibration curve Standard Addition generally matrix effects are not as pronounced as in GC as I have personally noticed in LC. Internal Standard best used with CC and SA methods. Almost always required since sample injection is not always a reproducible phenomenon, even with auto-samplers.
11/22/2005 Chem 253 - Chapter 24 53

Method of Internal Standards


Internal standards are often used in chromatographic methods, but useful in other techniques where sample sizes may vary from rune to run Known quantities of both non-interfering species and analyte species are subjected to the same analytical procedure and the signal is measured: [X] = 0.0837 mM [S] = 0.0666 mM Signal Analyte, X IS, S

Independent variable

11/22/2005

Chem 253 - Chapter 24

54

27

F-ratio
The signal of the IS is 347, and that of the analyte is 423. We now can define a response factor F, defined as

A 423 347 AX =F S = =F [X ] [S ] 0.0837 0.0666


Solving for F = 0.9700

We should realize that F will not vary even if sample sizes may do so.

11/22/2005

Chem 253 - Chapter 24

55

Example - Now consider an unknown, if 10.0 ml of 0.146 M IS is added to 10.0 ml of unknown then diluted to 25.0 ml, what is the unknown analyte concentration is As = 553 and Ax is 582?

[S] = 0.146 M (10.0 ml/25.0 ml) = 0.0584 M

553 582 = 0.9700 [X ] 0.0584

[X] = 0.05721 M [X]sample = 0.05721 M (25.0 ml/10.0 ml) = 0.143 M

11/22/2005

Chem 253 - Chapter 24

56

28

You might also like