Multidimensional Parallel Column Gas Chromatography P. M. Owens and D. W. Loehle Center for Molecular Sciences United States Military Academy West Point,

Multidimensional Parallel Column Gas Chromatography

P. M. Owens and D. W. Loehle

Center for Molecular Sciences

United States Military Academy

West Point, NY 10996

Figure 1. Multidimensional parallel column GC system

GC OVEN

Precolumn

ColumnsAnalytical

Polar SP

Semipolar SP

Nonpolar SP

SwitchDean's

DetectorAnalytical

DetectorMonitor

Injector

using nonpolar, semipolar and polar stationary phases (SP).

Parallel Column GC Systems

System 1 System 2

Precolumns

5m, 0.53mm HP1 15m, 0.50mm UAC-1

Analytical Columns

10m, 0.53mm HP1 15m, 0.25mm UAC-1

10m, 0.53mm HP17 15m, 0.25mm UAC-17

10m, 0.53mm HP20M 15m, 0.25mm UAC-CW

Retention Time Equations

tr = tm + tm k

tm = mobile phase hold-up time

k = partition ratio

t’r = tr - tm

t’r = adjusted retention time

k = (tr - tm) / tm = t’r / tm

Parallel Column Retention EquationsSingle analyte & 3 columns w / diff. stationary phases (SP)

tr (SP1) = tm + tm k (SP1)

tr (SP2) = tm + tm k (SP2)

tr (SP3) = tm + tm k (SP3)

• The k’s result from specific solute-stationary phase interactions and can therefore be used for solute identification

Retention Time Calibration

Day 1 tIS (1) = tm1 + tm1 kIS

Day 1 tAN (1) = tm1 + tm1 kAN

Day 2 tIS (2) = tm2 + F * tm2 kIS

Day 2 tAN (2) = tm2 + F * tm2 kAN

F corrects for changes in k

Retention Time Adjustment

t’AN (2) / t’IS (2) = t’AN (1) / t’IS (1)

( t’ = tr - tm )

Accounts for variations in tm’s

Accounts for linear variations in k (F factor)

Relative Retention () Libraries

A, IS = t’A (2) / t’IS(2) = t’A (1) / t’IS (1)

• Generate GC library to tabulate ’s for each compound on all stationary phases

• Run int. std. with all analyte & library runs

• Since ’s are T-dependent, run all samples with identical temperature programs

Search Algorithms

• Sum of differences

Hit Metric = (LIB )

SP

• Euclidean distance

Hit Metric = [1- (N, LIB )]SP

and N, LIB are normalized vectors from set of ’s for each compound

Search ResultsHP20M HP17 HP1

• Sum of differences

Cyclooctane 1.85 1.77 1.46

Cyclooctadiene 1.93 1.93 1.49

1-Heptanol 1.84 1.84 1.58

• Euclidean Search

Cyclooctane 1.85 1.77 1.46

Nonanoic Acid 3.05 2.87 2.46

Octanoic Acid 2.71 2.57 2.19

Chromatography Relations

KD = k KD = Distribution constant

k = Partition ratio

= Phase ratio (Vg / Vs )

KD depends on three variables: 1) temperature, 2) solute, & 3) stationary phase

Retention & Thermodynamics

KD = k G = -RT ln KD

tr = tm + tm k G = H - T S

ln k + ln = -H / R T + S / R

ln k = - H / R (1 / T) + S / R - ln

-0.1

-0.05

0

0.05

0.1

RE

TE

NT

ION

TIM

E E

RR

OR

(M

IN)

C8 C10 C12 C14 C16COMPOUND

HP20M HP17 HP1

ERRORS IN PREDICTED RETENTION TIMEST PROG OF 12/MIN FROM 15/MIN & 8/MIN

-0.1

-0.05

0

0.05

0.1

RE

TE

NT

ION

TIM

E E

RR

OR

(M

IN)

C8 C10 C12 C14 C16COMPOUND

HP20M HP17 HP1

ERRORS IN PREDICTED RETENTION TIMEST PROG OF 20/MIN FROM 15/MIN & 8/MIN

Retention Prediction Errors (CV)Prediction Used T-progs of 8oC/min & 15oC/min

Sample (C7-C11) 12 C/min 20 C/min

Alkanes (C8-C16) 0.27% 0.51%

Halogenated 0.38% 0.87%

Ketones 0.56% 0.59%

Aldehydes 0.61% 0.62%

Alkanes (C9-C15) 0.33% 0.45%

-0.2

0

0.2

RE

TE

NT

ION

TIM

E E

RR

OR

(M

IN)

C8 C7CL C8CL C9CLC10CLC11BR C16COMPOUND

HP20M HP17 HP1

ERRORS IN PREDICTED RETENTION TIMEST PROG OF 12/MIN FROM 15/MIN & 8/MIN

-0.2

0

0.2 R

ET

EN

TIO

N T

IME

ER

RO

R (

MIN

)

C8 C7AL C8AL C9AL C10AL C16COMPOUND

HP20MHP17HP1

ERRORS IN PREDICTED RETENTION TIMEST PROG OF 20/MIN FROM 15/MIN & 8/MIN

-0.1

0.1

Tr D

IFF

ER

EN

CE

(m

in)

6 8 10 12 14 16 18 20

HP20MHP17 HP1

Tr Variation on GC Stationary PhasesFOR GC RUNS 30 DAYS APART

900

1000

1100

1200

1300

1400

1500 T-P

RO

GR

AM

ME

D R

ET

IN

DE

X

C7CL C8CL C9CL C10CL C11BR

GC1-20M GC1-DB17 GC1-DB1

GC2-20M GC2-DB17 GC2-DB1

RETENTION INDICES FROM 2 INSTRUMENTSWITH .25mm & .53mm DB1/17/20M COLUMNS

LibraryCompounds

Predict Retention for Analyte

GC conditions

Predict AnalyteRetention under

Lib. GC Conditions

Single AnalyteChromatogramX

Parallel Column Gas Chromatography

• Measures interaction on multiple stationary phases - a separate dimension of analyte information

• Requires the use of internal standards to characterize GC operating conditions

• Thermodynamic modeling allows adjustment of library retentions to current operating conditions

Future Areas of Focus

• Interinstrument variability assessment

• Development of calibration procedures to minimize retention prediction errors

• Optimization of stationary phase selection

• Evaluation of an increased number of parallel columns

• Application for complex mixture analysis

Acknowledgements

• Association of Graduates and Army Research Office

• Beverly S. Scott & Rodney S. Gonzalez

• Tony Weaver

• Department of Chemistry, USMA