Analytical Separations

Chromatography: colour writing

- chroma (colour) graphein (write)
- first application of column chromatography was separation of
plant pigments by Michael Tswett (1872-1919), Russian botanist.

General Principle: separation of components that are dissolved

in a mobile phase through their selective retention by a stationary
phase.

preparative chromatography- separation and collection of different

components (usually large scale).

analytical chromatography- separation and detection of different

fractions (can use smaller columns and samples)

Classification of Analytical
Separations
Classification by Physics
Thin Layer (planar) chromatography
-stationary phase is held on flat support (ie- paper) and mobile
phase passes over stationary phase through force of gravity or
capillary action
S Column chromatography
- stationary phase is held inside a column (metal, glass or fused
silica). Mobile phase passes through column through external
pressure (pump or gaseous pressure)
S Electrophoresis (both planar and capillary)
- no stationary phase, separation based upon ionic mobility
- force to drive mobile and to facilitate separation phase comes
from applied electric field.
S Electrokinetic chromatography
- electrophoresis with a chromatographic stationary phase

1
 The separation process

Partitioning
Partitioning
Amobile ÅÆ Astationary partitioning of analyte A

each analyte will have some equilibrium partitioning between the

mobile and stationary phase, analogous to the partition coefficient
used in solvent extraction.
Distribution Constant (partition coefficient)
C
K= S
CM
Cs : concentration of analyte in stationary phase
CM : concentration of analyte in mobile phase

Retention Factor (capacity factor)

moles A (stationary ) C V V
k' = = S S =K S
moles A (mobile ) C M VM VM

2
 Retention

tR'

tM = time for mobile phase (unretained component) to travel from injector to detector
tR = time for retained analyte to travel from injector to detector.
tR' = tR-tM = adjusted retention time

How do we get equilibrium information from chromatograph?

Consider the linear velocities:

L
vR = average velocity of a retained component
tR
L average velocity of mobile phase
vM =
tM

L = length of column from injector to detector

BUT, vR = vM x fraction of time analyte is in mobile phase

At any instant, the fraction of time spent by an analyte in the mobile

phase is given by the ratio of moles of analyte in the mobile phase to the
total number of moles of analyte:

C M VM 1 1 1
vR = vM × = vM = vM = vM
CM VM + CSVS C V V 1+ k'
1+ S S 1+ K S
C M VM VM

3
Substitute the velocity equations into the above:

1
vR = vM
1+ k'
L L 1 1 1 1
= × ⇒ = ×
tR tM 1+ k' tR tM 1 + k '
tR tR
1+ k' = ⇒ k' = −1
tM tM
tR − tM tR '
k' = ⇒ k' =
tM tM

Thus we can get the partition for each species directly from the chromatogram just by
measuring the retention times!

(Be aware that tM, t'M, k' are different for each chemical species (component) in a
separation and they can change with conditions. For simplicity, a subscript has not been
included in the above equations.

Selectivity
Selectivity in analytical separations is achieved through selecting proper conditions
such that two (or more) analytes have differential retention (and thus ∆vR and ∆tR).
Selectivity is achieved through differential affinity for the stationary phase, ie- KA≠
KB. Thus the selectivity coefficient for analyte B with respect to analyte A is
defined as:
KB
α= a = selectivity coefficient
KA
Where “B” is the more strongly retained component (longer retention time). Thus
α is defined such that it is always greater than 1.0

k ' B t R ,B − t M t 'B
It is easy to show the following: α= = =
k ' A t R ,A − t M t ' A
Thus, we get selectivity coefficient from chromatogram!

4
 Column Efficiency
The efficiency of a column to achieve a separation is characterized by:

N: number of theoretical plates (effective, not real).

H: height equivalent of a theoretical plate

Real chromatographic peaks will disperse as they travel down the column. One
measure of this dispersion is the variance of the peak, F2, assuming it is guassian.

The height equivalent of a theoretical plate is the ratio of the dispersion, σ2, per
distance travelled on the column, x :

σ2
H=
x

The smaller the better (more efficient separation)

The number of theoretical plates, N, is given by the length of the column, L, divided
by the “plate height”, H.
L Lx
N= =
H σ2
If we wish to measure the dispersion AFTER the peak travels the full length of the
column (from injector to detector), then
x = L, and: L2
N=
σ L2
Other equations are based upon variations of this, depending on how we measure F
of the chromatographic peak.

For example, since distance = v × time, we can express the above equation in terms
of the retention time of a peak and it’s standard deviation measured in the same
time units:

5
 2
tR
N=
σ 2t
2
16t R
N= 2
w
2
5.54t R
N= 2
w1/ 2
W = width of peak at the baseline as measured by the
tangent to the slope of the guassian curve at the
inflection points (ie- calculus can show that W=4F)
W1/2 = width of peak at half peak height

N is a dimensionless quantity. 100 < N < 2 x106

Both N and H can be quoted but must be quoted for a specific
analyte and set of conditions

6
 Band broadening in chromatography
σ 2peak
H=
x
Why do peaks disperse?
- kinetic factors
- thermodynamic factors
Van Deemter Equation
B
H = A+ + Cu u = flow rate
u
A = Eddy diffusion coefficient
B = Longitudinal diffusion coefficient
C = Mass transfer coefficient
(Cs + Cm)µ Cs = stationary phase mass transfer
Cm = mobile phase mass transfer

Peak Broadening Processes

7
 Eddy Diffusion (multiple paths)
For this process, peak broadening occurs because of the
statistical variation in the length of paths that analytes can
take when passing from one end of a column to the other,
through and around packing materials.

A = 2λdp dp = diameter of particles

λ = packing uniformity factor

This term is more problematic in packed columns. At low

flow rates, some diffusional averaging can occur. That reduces
the significance of this term, whereby analyte molecules
sample multiple paths in the column.

Longitudinal Diffusion
Longitudinal diffusion refers to the diffusion of analytes from the center of a
concentrated band to either side of the band. It is a process that can occur in the
absence, or presence, of flow down the column. The longer the migration time , the
larger the effect.

Time
B 2γDm Dm = diffusion coefficient in mobile phase (cm2sec-1)
=
u u γ = obstruction factor (1.0 for open tube, ~0.6 for packed column)

Longitudinal diffusion increases with decreasing flow rate due to the longer time that
analytes spend on the column. It is also more problematic in gas chromatography
compared to liquid chromatography due to higher diffusion coefficients in the gas
phase.

8
 Mass Transfer (resistance to)
The kinetics of mass transfer, from the mobile phase to the stationary phase or visa-
versa, is responsible for this type of dispersion. The resistance to mass transfer means
equilibrium conditions are not always achieved everywhere in the column (ie -
distribution of analyte between mobile and stationary phase) . In picture below,
analyte molecule is “left behind” due to resistance to mass transfer.
Mass transfer in stationary phase
stationary phase
f s (k ')d f2
Csu = u
Ds
u = flow rate
df = SP film thickness
analyte
Ds= diffusion coefficient in stationary phase molecule
Mass transfer in mobile phase

( )
fm k ' d p2
u = flow rate
Cmu = u dp = diameter of stationary phase particles*
DM DM = diffusion coefficient in mobile phase

*(ie- the larger the particles, the wider the mobile phase streams between particles)

Solution to mass transfer problems

We want fast kinetics of mass transfer to lower the time necessary for equilibrium
to be achieved. Rapid dynamic equilibrium is what we are after.
- use low viscosity, very thin layers of stationary phase coated on very small
particles.
- keep k' just large enough (2-10) to achieve separation with an acceptable
resolution. The higher k' (retention), the larger the resistance to mass transfer
contribution to H.

Practical problems (trade-offs)?

- low viscosity films not as stable (bleeding!)
- thin films don’t have high capacity for analyte loading
- small particles require higher pressures
- larger values of k' are needed as the number of components that need to be
resolved increases

9
 van Deepter Plot

A - eddy diffusion
B – longitudinal diffusion
C – mass transfer
Optimum
column TOTAL
efficiency

Comments about Van Deemter Theory

A) Eddy diffusion will be absent in hollow bore columns (ie- capillary GC). It becomes
more significant in HPLC
B) Longitudinal diffusion is a function of the diffusion coefficient of the solute in the
mobile phase. If the mobile phase is a gas, the diffusion coefficient, Dm, is orders of
magnitude higher than in a liquid. Therefore, the B/µ term is more pronounced in
GC.
GC HPLC

H H

flow rate flow rate

C)For hollow columns, Cmµ is smaller in narrower columns since the solute has smaller
distance to diffuse. Therefore we get higher efficiencies in GC with narrower I.D.
columns.

10
 Resolution

Resolution is a measure of the effective separation of adjacent analyte peaks.

Resolution of two peaks is aided by thermodynamic factors, which give rise to
partitioning and separation (ie- K), and it is countered by kinetics, which can give
rise to additional dispersion of a peak, making otherwise well separated peaks merge
with each other. For high resolution, we want high selectivity (αBA = KB/KA) and a
low level of dispersion (ie- combined, we want high N!)

Experimental Measure of Resolution

For 2 adjacent peaks, A & B:

∆t R t R ,B − t R , A 2(t R ,B − t R ,A )
Rs = = =
W  A
W + W B  W A + WB
 
 2 

Example
Chromatogram C
tR,B = 13.4 cm
tR,A = 8.1 cm

WA = 1.8 cm
WB = 3.75 cm

2(t R ,B − t R , A )
Rs =
W A + WB
2(13.4 − 8.1)
=
1.8 + 3.75
= 1.9

Usually one needs

Rs> 0.5 to identify
presence of 2nd peak,
Rs>1.5 to quantify
the 2nd peak

11
Can we define resolution in terms of fundamental parameters (ie- N, K, α, k’, etc.)?

Consider 2 closely spaced peaks that we wish to separate. If they are similar, we can
presume that WA ~ WB.
t R ,B − t R ,A t R2 4t R
Rs = N = 16 ⇒ W =
W W2 N
t R ,B − t R ,A t R ,B − t M k 'B K B
Rs = N α= = =
4t R ,B t R ,A − t M k'A K A
N  α − 1 k 'B
Rs =  
4  α  (1 + k ' B )
If the peaks are very close (not well resolved necessarily),
then α ∼1

Rs =
N
(α − 1) k 'B
4 (1 + k 'B )
1 2 3
Dependence
1) N – physical column characteristics (length, particle size…as well as column
conditions)
2) α – selectivity controlled with column material, dependent on analytes as well
3) k’ – separation conditions, amount of retention. As k' Æ 0, Rs Æ 0.

How to control resolution?

physical column, column material, separation conditions (we could also add detector
to this list)-flow rate, column length, particle size, column id., mobile phase,
gradients, temperature.

12
 The General Elution Problem

It is sometimes difficult to get an optimum separation of ALL components in the

mixture in a reasonable period of time.

all peaks separated BUT…3 hour

separation & 5,6 have poor signal.

increase flow or change column,

good separation time but 1,2 & 3,4
are not resolved.

change flow and column material.

3,4,5,6 OK but early peaks 1,2 still
not resolved.

Solution to General Elution problem

If we could change the separation conditions (ie-K) as the
separation procedes, we could optimize conditions for
separation of early peaks in the first part of the procedure, and
then change conditions to gradually optimize conditions for
successive eluting components, so that they are adequately
separated in a reasonable period of time.

How do we change K as separation proceeds?

Gas Chromatography- temperature programming

Liquid Chromatography – solvent programming

13
 HPLC Solvent Gradient
Temperature Programin GC 100
250

200 80

150 60

100 40

50 20

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30 35 40
Time (min) Time (min)

H2O CH3CN CH3OH

Peak Asymmetry

Asymmetry Factor
where b and a are horizontal
distances measured from the
centroid vertical line to positions
on the right and left portions of
the curve. The horizontal line is
drawn at a defined % of the peak
height (ie-10%).

Sources of Asymmetry
; overloading column (too much
injected)
; N is too large
; wrong analyte/stationary phase
combination
; poor column (spaces in column, loss
b
of stat phase).
Fasym =
a

14
 Quantification in Chromatography
How do we quantify the
amount of material in a ?
chromatographic peak?

We need a reproducible
analytical signal….AND 0.25
we CALIBRATE the a
amount of injected 0.20
material necessary to give c

TAS (AU)
0.15 b f
a unit amount of d g
analytical signal. We 0.10
e
calibrate with injections
of standard material 0.05
under the same
conditions as the samples
are injected 25 30 35 40 45 50
Time (min)

Analytical Signals in Chromatography

glyoxal
Peak height 0.14
height of the peak from the assumed
peak height
TAS (AU)

baseline to the maximum peak height.

Units can be anything…detector signal,
or distance on a printout in cm, mm; as 0.10
long as same units are used in
calibration
0.06
38.0 40.0 42.0 44.0 46.0
Problems Time (min)
- The peak height signal is sensitive to slight changes in conditions (changing
temperature, solvent gradient, etc.) …or to systematic changes over time (degrading
column, etc).
-- peak height will depend on the slope of the baseline…the maximum signal will
appear at a different location
-- peak height signal is more noisy, it does not use the full amount of
information…and material in the chromatographic peak

15
 PEAK AREA
t R ,high

Peak Area = ∫y
t R ,low
i dt R y i = analytical signal

glyoxal
integrated area of the peak above the assumed 0.14
baseline. As with peak height, units can be peak area

TAS (AU)
anything…but are usually in units of (y-axis
units)X(x-axis units); ie absorbance units-
0.10
seconds. Again, same units are used in
calibration

0.06
ADVANTAGES 38.0 40.0 42.0
Time (min)
44.0 46.0

- The peak area is lass sensitive sensitive to changes in conditions (changing

temperature, solvent gradient, etc.) …or to systematic changes over time (degrading
column, etc).
-- peak area intrinsically has greater signal to noise ratios (and therefore lower
detection limits) since it uses ALL of the signal from ALL of the material in the
chromatographic peak.
-- in a side by side comparison of the peak shown above, McLaren found that the
peak area gave S/N ratios that were 2.7 times higher than the peak height.

Gas Chromatography
(Chapter 27 in Skoog, Holler & Neiman)

16
 GC Instrument Components

Carrier Gas: inert gases- H2, He, N2, He most common, pressurized (~2500psi)
Two stage regulator : steps down pressure from cylinder to a safe pressure used at
the instrument inlet (ie- 10-40psi). Note 1 atm = 101.3 kPa = 14.5 psi
Flow Rates: 25-150 mL/min (packed columns)
0.5-25mL/min (open tubular) out
finish

start
in

Rotameter Soap Bubble Flowmeter

Flow proceeds through a tapered channel, Vfinal − V start
ball rises according to flow. Must be Flow =
calibrated, depends on the gas. t final − t start

Carrier Gases

Interesting question: why does H2 give more efficient separations (lower H) at high
flow and worse separations at low flow, when compared to N2 a a carrier gas?

17
 Injection

.samples can be injected as liquids or gases

.in general, GC is used for volatile
components (in contrast, HPLC for
semi-volatiles or non-volatiles)
Direct Injector
. flash injector is held at 150-300o C
. liquid (or gas) is injected through
septum with microliter syringe
. sample quickly volatilizes (boils) and is
swept onto the column by the carrier
gas.

Rotary Valve
. 6-port injector, used for gaseous
samples (see diagram)
. sample is loaded while valve is in load
position- sample fills the sample loop
ranging from 60nL - 10 mL.
. valve is switched to inject, putting the
sample loop into the carrier gas flow,
carrying it to the column.

Peak Compression (peak focussing)

.if the column temperature is kept low to begin (ie- below the boiling points of
the analytes), the sample being injected is concentrated onto the head of the
column. This reduces the contribution of the injection volume to overall peak
dispersion
2 2 2 2 2
σ peak = σ injection + σ column + σ det ector + σ other

Peak dispersion theory already discussed

18
 Autosamplers
¾automatic sampling and injection that has better reproducibility than can be
achieved with manual injection (<0.5% RSD)
¾samples stored in sealed vials in a rotating tray; tray rotates, syringe pierces
septum, flushes loop with sample to reduce “carry-over”, syringe detracts, valve
switches to inject.
¾allows fully automated analysis. Fill tray with 24 samples, program sequence,
results are waiting for interpretation in the morning.

Detectors
Desirable Qualities
- high sensitivity - does not respond to carrier gas
- stable (reproducible: ie-low drift) - able to withstand high temperature
- high dynamic range - non-destructive
- universal response (responds similarly to all analytes or selective response for
certain classes of compounds)

Flame Ionization Detector (FID)

¾ column effluent is mixed with H2/air mixture that is carried to small flame
¾ organic compounds are pyrolized in the flame producing ions and electrons
¾ # of ions produced is proportional to # of reduced carbons (ie- -CH2- as opposed to -
C=O or -C-OH)
¾ high voltage (100-200V) applied between burner tip and collector electrode, current
flows to collector (pA-nA)
¾ small currents are amplified (ie- i to V convertor)
¾ Icollector % gaseous concentration of analyte

19
 FID
Advantages - highly sensitive, large
dynamic range, economic, reproducible.

Disadvantages-destructive of sample, H2O

sensitive, not selective (no molecular
information)

Thermal Conductivity Detector (TCD)

Principle of operation: differences in thermal conductivity of analyte gases will give
change in temperature of a heated filament, which gives rise to a change in R of
filament. The change in R can be detected in a “Wheatstone Bridge” circuit using 4
filaments. One pair (2) is located in column effluent and one pair(2) is located in a split
stream (or before the injector). ie- one pair acts as a reference.
- thermal conductivity of carrier gases (H2 or He) is much higher than organic gases
(analytes). When organic analyte enters the detector, the temperature of the filament
rises, the resistance increases, the imbalance in resistance between the reference and
detector filaments is the analytical signal.
Wheatstone Bridge – voltage drop (IR drop) depends on the resistance of the filament.
The bridge uses an instrumentation amplifier to amplify small differences in the
voltage at the two points shown in the bridge.
Advantages - universal detector, economical, easy replacement
Disadvantages
- low sensitivity,
- not selective
(no molecular
information)

20
 Electron Capture Detector
¾principle of operation: column effluent passes between two electrodes. On the
surface of one electrode is a beta (e-) emitter (Ni63 or tritium, H3).
¾.beta (e-) emitter produces high energy electrons that pass into the gas stream,
resulting in a plasma of thermal electrons and positive ions.
¾current between the electrodes is detected by applying a pulsed voltage to one of the
electrodes.
¾when analyte (electron capturer) passes between the electrodes, they capture
electrons in the plasma, reducing the current. The reduction in current is the
analytical signal.

ECD
¾Analytes: efficient electron
capturers such as halogenated
organics, peroxides and nitro-
substituted compounds.
R-Cl ; R-Br ; -NO2; R1-OO-R2 ;

Applications: environmental, ie-

chlorinated pesticides, PCB’s other
halogenated organics.
Advantages - very sensitive
Disadvantages - small dynamic
range (2 orders of magnitude)

21
 Atomic Emission Detector (AED)

As name implies, emission of atomic species (no

molecular information) in column effluent is the
analytical signal. Uses microwave plasma as source

Application of
AED
Analysis of oxygenates
in gasoline
Carbon chromatogram

Oxygen chromatogram

22
 Mass Selective Detector

mass spectrometry
the most powerful advantage is the excellent abilities to
obtain structural information, m/z for parent ions &
fragments, for unknown analyte peaks.
selected ion mode (SIM) or scan mode.

Other detectors
Sulfur Chemiluminescence Detector
¾responds to S in compounds (ie- mercaptans)
¾eluent is burned in H2/O2 (same as FID)
¾S reacts with O3 in post column flow to produce
chemiluminescent compounds
¾intensity of emitted light is the analytical signal
Flame Photometric Detector
¾sensitive to both S & P
¾eluent is burned in H2/O2
¾forms HPO* º 510-526 nm radiation
¾ S2* º 394 nm radiation

23
 GC Stationary Phases
Polydimethylsiloxane (PDMS) when R = methyl
R = CH3 (methyl)
nonpolar
= C6H5 (phenyl)
= C3H6CN (cyanopropyl)
= C3H6CF3 (perfluoro-
methylpropyl)

Polyethyleneglycol (PEG)
polar

Other GC stationary phase

24
 Adsorption of Polar compounds on
Silica Surfaces
Silanol groups on surface of
fused silica will attract polar
compounds which are
adsorbed. Can be problematic.
Solution:
Silanization of support
(deactivation) using
dimethylchlorosilane.

Methanol wash to give non-

polar methoxy terminated
group.

Columns
Packed columns
2 - 3m in length; 2-4mm i.d.
densely packed with fine material, diatomaceous earth (skeltons of diatom
organisms), coated with a thin layer of non-volatile liquid stationary phase.
Column efficiency increases with decreasing particle size.

Open Tubular columns

first developed in 1950's and are rapidly replacing packed columns for
analytical separations
15 - 100m lengths!; 100 - 600µm i.d.
WCOT - wall coated open tubular (thin layer of liquid)
SCOT - support coated open tubular(thin layer of support material)
PLOT - porous layer open tubular (thin layer of solid stat. phase ie- Al2O3)
FSOT - fused silica open tubular, 1980's (refers to outer column holding)
Open tubular columns are much more efficient than packed columns....up to
1,000,000 theoretical plates, but typical is ~ 100,000

25
 Applications
SPME/GC-MS analysis of oxygenated,
aromatic and biogenic hydrocarbons.
Vancouver - 2000

Cryogenic preconcentration/GC-MS
analysis of oxygenated, aromatic and
biogenic hydrocarbons.
Vancouver -2000

Site details

Mobile unit at Seymour

Conservation area

Sampling through Teflon line

26
 Solid Phase Micro extraction (SPME)

compounds retention timeM/Z for quantif

methanol 5.314 31

Chromatogram
ethanol 5.846 31
propanol 7.912 31
butanol 10.078 31
pentanol 11.942 42
hexanol 13.275 56
isopropanol 5.726 45
Abundance
isobutanol 9.036 43
acetaldehyde 3.4021 29
TIC: AUG24C2.D propanal 4.06 58
320000
butanal 5.089 72
300000 hexanal 8.992 56
280000
acetone 4.368 43
methylethylketone 5.406 43
260000 methylvinylketone 6.129 55
240000
ethylacetate 5.19 43
proppylacetate 6.573 43
220000 butylacetate 8.754 43
200000 benzene 6.113 78
toluene 8.219 91
180000 ethylbenzene 9.992 91
160000 o-xylene 10.142 91
m-xylene 11.144 91
140000
p-xylene 10.29 91
120000 isoprene 3.107 67
a-pinene 7.753 93
100000
acrolein 4.656 55
80000 B-pinene 9.667 93
methacrolein 5.092 41
60000
cis-3-hexen-1-ol 13.667 67
40000 R-(+)-Limonene 11.347 93
20000

0
2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00 18.00 20.00
Time-->

27
 cyclohexadiene
O

Thujone, 91
β-phellandrene limonene
Abundance 3-isopropyl-toluene
TIC: DBFAUG27.D

180000

170000

160000

150000

140000

130000

120000

110000

100000

90000

80000

70000

60000

2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00 18.00 20.00
T ime-->

VOCs emitted by Deer Fern commonly found at Seymour site

Kovats Retention Index (RI)

A method for charterizing retention properties, predicting
retention times of analytes and also for possible peak
identification of unknowns. The retention index predicts
where compounds will appear on a chromatogram with
respect to straight chain hydrocarbons under a given set of
conditions. The RI for straight chain saturated
hydrocarbons is defined to be (regardless of the GC column
or conditions):

RI = 100 x Nc where Nc = number of carbons

28
 Species Formula Nc RI
methane CH4 1 100
ethane CH3CH 3 2 200
propane CH 3CH2CH 3 3 300
n-hexane C6H14 6 600
n-octane C8H18 8 800
n-dodecane C12H 26 12 1200
normal HC CnH 2n+2 n n ×100
It has long been known that a plot of the logarithm of the
adjusted retention (t’r) time is linear with Nc. This is true
for other homologous series as well (ie- alcohols, acids, etc.)

log (tr-tm) = m R.I. + b

Example
separation on a DB-1 (non-polar) column
t’r (min) log t’r R.I.
Hexane 15.0 1.176 600
nitro-propane 16.0 1.204 ?
Octane 25.0 1.398 800
Find the retention index for nitropropane

log t’r = m×R.I. + b

m = )y/)x = (1.398 - 1.176)/(800-600) = 0.00111
b = y-mx (Use one existing point and m)
b = 1.398 - 0.00111(800) = 0.510

29
 Cont’d
log t’r = 0.00111×R.I. + 0.510
For nitro-propane
R.I. = (log t’r - 0.510)/0.00111= (log 16.0 - 0.510)/0.00111

= 625.3

The R.I. of nitropropane under the given conditions is 625.3

HPLC

GC

30
 HPLC
Applications:
In general, applications are for the analysis of mixtures of
non-volatile or thernmally unstable compounds (those that
cannot be analyzed by GC).

Analytes:
Biological compounds - amino acids, proteins, nucleic acids,
metabolites
Pharmaceutical – drugs, antibiotics, steroids
Environmental – non volatile hydrocarbons, pesticides,
PAH’s…

Instruments

31
 Instrument Details
Mobile Phase Delivery
-We require very clean, very pure solvents for HPLC
(usually UV transparent), “HPLC” grade.
-in addition, solvents are usually filtered and ‘degassed’
-degassing (removal of dissolved N2, O2, CO2 …) is achieved
by vacuum devices (vaccum applied to headspace above
solvent), ultrasonication (high frequency vibration drives
gasses out of solvent), heating (decreases solubility of gases)
or He sparging (headspace of solvent exposed to insoluble
gas…He).
-If solvents are not degassed, gases that are soluble at high
pressure (ie- 4,000 psi) become insoluble at low pressure
end (ie at the end of the column or detector). Bubbles can
interfere with separation AND detection,

Elution
Isocratic elution – solvent composition is constant
Gradient elution - solvent composition changes with time
binary gradient – 2 solvents (very common)
tertiary gradient – 3 solvents (somewhat common)
quaternary gradient – 4 solvents (very rarely)
Pumping – supplies required pressure to force solvent
through packed column.
Reciprocating pump – mechanical high pressure liquid
pumping achieved by piston, check valves. Pressure surges
can be somewhat overcome with pulse dampeners.
Displacement pump – continuous pressure applied to large
volume (syringe pump), advantage is lack of pressure
surges, disadvantage is price.

32
 Reciprocating Pump

Injection

Fixed Volume – volume is given by the size of the sample

loop…it must be overfilled with sample.
Variable Volume – partial filling of loop, better for
development, but precision is not as good.

33
 Columns

Material- stainless steel or fused silica

Length - 5-50 cm (analytical)
Diameter
4-6 mm i.d. – conventional
1-2 mm i.d. – micro columns
0.3-1.0 mm i.d. – capillary columns
100 – 300 µm i.d.–nano columns DEVELOPMENT

Efficiency – 10,000 – 200,000 plates per meter

Particle sizes
The most common
particles are 5µm and
3µm silica, with a bonded
stationary phase.

f m (k ')d 2p
Data from Fig 28.2 Data from Fig 28.2
3.0 cm/sec 3.0 cm/sec

Cm u = u
5 5
4

DM
4
H (mm)

3
H (mm)

3
2 2
1 1

0
0 10 20 30 40 50 0 500 1000 1500 2000
Particle Size (um) dp ^2 (um^2)

34
 Detectors for HPLC
What size should a detector be? We want it to be small enough such that it
doesn’t contribute to extra column bandbroadening. As a rule of thumb, if the
detector volume is 20x smaller than the volume of the analyte peak, it should
not broaden the peak significantly. What is the volume of a peak? We can
easily calculate using some empirical observations. Let us calculate for a
conventional column size, 4.6 mm i.d.
N = 20,000 plates (good quality column)
tR = 10.0 min (for example)
Dcolumn = 4.6 mm i.d. …Flow rate = 1.0 mL/min
W = ??, N = 16 (tr/W)2 W = 0.28 min
Peak volume at detector = ?
Vpeak = W x F = 0.28 min x 1.0 mL/min = 0.28mL
Vdetector = ?
Vdetector = Vpeak/20 = 280µL/20 = 14.1 µL

Therfore, for conventional analytical chromatography, Vdetector ~14µL

Detector Volumes
For smaller size columns, for a similar separation with a similar linear velocity,
the flow rate, and therefore the required detector volume, should both scale
down linearly with the cross sectional area of the column…in other words they
should scale down as the square of the diameter of the column.

Area = πr2

2
 d column 
f scale =  
 d conventional 
dcolumn Flow (µL/min) Vdetector (µL)
4.6 mm 1000 14.1 Note the lower solvent use
associated with smaller
2.0 mm 189 2.7 diameter columns
1.0mm 47 0.67
0.3 mm 4.2 0.060 (60 nL)
0.1 mm 0.47 0.007 (6.7 nL)

35
 Detector Types
UV Absorbance
-Variable Wavelength Detector (VWD)
-Diode Array Detector (DAD)

Fluorescence
Electrochemical
Refractive Index
Light scattering
Mass spectrometric
Raman

UV/VIS Absorbance
Standard Flow Cell – VWD or DAD

A = -log T
= εbC

b = pathlength
= 5 -10 mm
Vdetector = 2-20µL

36
 Small Volume Absorbance Flow Cell

b = 12 mm
Vdetector = 30 nL

0.25
a
0.20

c
TAS (AU)

0.15 b f
d g
e
0.10

0.05

25 30 35 40 45 50
Time (min)

37
 Diode Array Detection

Refractive Index detector

Advantages – universal, insensitive to flow, economical, robust

Disadvantages – not sensitive, highly temperature dependent

38
 Partition Chromatography
Liquid-Liquid Phase Chromatography: liquid stationaryphase is retained on
support particles by physical adsorption

Bonded Phase Liquid Chromatography: stationary phase is chemically bonded

to surface of silica particles or other support materials

R = octadecyl (C18)
R = octyl (C8)
R = butyl (C4)
R = ethyl (C2)

Normal/Reverse Phase
Chromatographies
“Normal” and “Reverse phase" terminology refers to the
polar/non-polar or hyrophylic/hydrophobic nature of the
stationary phase and the mobile phase employed in the liquid
chromatographic separation. It will affect the order of elution
of analytes and the quality of the separation depending on the
nature of the analytes. In general terms, one would want to
use normal phase chromatography for very polar analytes and
reverse phase chromatography for non-polar analytes,

39
 Normal Phase Chromatography
Stationary phase is highly polar
mobile phase is non-polar.
Mobile phase gradient (if not isocratic) procedes from
non-polar to progressively more polar.
Stationary Phases – water or triethylene glycol on silica or
alumina particles (highly polar)
Mobile Phases – pentane, hexane, isopropylether
Order of Elution:
“Like dissolves like”. Most polar compounds are highly
retained by the stationary phase. They come off last.
Order of elution is non-polar (or hydrophobic) followed
by polar (hydrophylic).

Reverse Phase Chromatography

Stationary phase is non polar
mobile phase is polar.
Mobile phase gradient (if not isocratic) procedes from
polar to progressively less polar (ie- water -> methanol)
Stationary Phases – bonded ODS (C18), C8, C4
Mobile Phases – water, acetonitrile, methanol,
tetrahydrofuran
Order of Elution:
Non polar compounds ar most highly retained. Order of
elution is polar first, followed by non-polar.

40
 Application of HPLC
Successful application implies baseline separation of all
analytes in the fastest time possible.
This requires choice of stationary phase & mobile phase
to match the analytes
Optimum separations have k’ in the range of 2-5 for less
than ten components in the mixture. For complex
mixtures, with more analytes, it may be necessary to
expand the k’ range (ie- 1.5-10) in order to separate
everything satisfactorily. A more complex separation
usually takes more time.
Once a column is chosen, most of the control in the
separation, k’ and α, is based upon choice of mobile phase,
especially the polarity.

41
 Snyder Polarity Index, P’
Based upon solubility of substance in 3 solvents, dioxane,
nitromethane, and ethanol.

one can change the polarity of the mobile phase by

mixing solvents that have different P’s
P’AB = φAP’A + φB P’B φ = volume fraction
P’ = polarity index
Rule of thumb – a 2 unit change in P’ will give a 10 unit
change in k’ (ie-relative retention time). k’ gets larger with
increase in P’ for reverse phase chromatography, k’ gets
smaller with increase in P’ for normal phase
chromatography.
k2
= 10( P '2 − P '1 ) / 2 reverse-phase separation
k1
k2 normal-phase separation
= 10( P '1 − P '2 ) / 2
k1

42
 to optimize
separations, one first
determines required
polarity to get k’ in
correct range. Then
one can
systematically try
different solvents,
keeping total polarity
(and thus k’) the
same

Ion Chromatography

separations based upon exchange of ions (cations or anions)

on an ion exchange stationary phase (resin) column.
for cation analysis, exchange is with mobile ionizable cations
on a cation exchange resin
Support-R-SO3-H+ (sulphonic acid)
for anion analysis, exchange is with mobile ionizable anions
on an anion exchange resin
Support-N(CH3)3+-OH- (trimethylammonium)

43
Equilibrium (Cation Exchange)

M+ = analyte cations; R-H+ = exchange resin

R-H+ + M+ W RM+ + H+
[R − M + ][H + ]
K exchange =
[R − H + ][ M + ]
Now usually, [H+] >>[M+] and it is maintained constant with a buffer.
Also usually, [R-H+] >>[R-M+] and is ~ constant .

[ M + ]stationary [R − M + ]re sin [R − H + ]

= = K exchange =K
[ M + ]mobile [M + ] [H + ]

The exchange on an ion exchange column behaves like any other chromatographic
partitioning process. The separation depends on the affinity of the analyte cations
for the exchange resin. This affinity will depend on:
i) analyte charge (usually higher charge = higher K, stronger retention)
ii) ionic radius (usually smaller radius, higher K)
iii) polarizability (higher polarizability, higher K)

Equilibrium (Anion Exchange)

Similarly, we have for anion exchange...
A- = analyte anions;
RN(CH3)3+OH- = exchange resin = R+OH-

R+OH- + A- W R+ A- + OH-
Ion Exchange Instrument
- very similar to HPLC except we need an eluent suppressor and we typically use a
conductivity detector. The conductivity detector "sees" an increase in solution
conductivity for an analyte peak. But our mobile phase eluent is usually a buffered
salt, that has high conductivity. To increase sensitivity (decrease background signal
due to eluent conductivity), the instrument contains an eluent supressor that
removes the high conductivity of the eluent ions, while NOT affecting the analyte
ions. This is a second small ion exchange column (of opposite type) that follows the
analytical ion exchange column.
ie – For cation analysis, HClaq is often used as our eluent. Following our cation
exchange column, our suppressor column is a cation exchange column where the
following takes place...
H+ + Cl- + R+OH-resin Æ R+Cl-resin + H2O.
The product of the exchange, H2O , is neutral, and therefore has low conductivity!

44
 Conductance and Conductivity Detector
L
r
V+ Pt Wire Pt Wire V-

For a voltage applied across two electrodes, of radius r, and separation distance L,
the current that flows is given by...
M i N z i 2e 2πr 2V
I= ∑ i
6πη ri L
i = summation over all ions
Mi = concentration of ion
N = Avogadros constant
V = applied voltage
zi = charge on ion
e = charge on electron
ri = radius of ion
η = viscosity

Ion Exchange Resins

DVB Crosslinks

Cations exchange with H+

45
 IC Instrument
Buffer

Exchange column
(anion or cation)

Micromembrane Suppressor
removes background conductivity from buffer, so there is
only significant conductivity from analyte ions.

46
Application of IC to Inorganic Anions
Analysis in Atmospheric Particulates

Eluent: CO32-/HCO3-

Inorganic Cations in Atmospheric

Particulates

47